diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2020-01-28 16:02:33 -0800 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2020-01-28 16:02:33 -0800 |
| commit | bd2463ac7d7ec51d432f23bf0e893fb371a908cd (patch) | |
| tree | 3da32c23be83adb9d9bda7e51b51fa39f69f2447 /drivers/net/ethernet | |
| parent | a78208e2436963d0b2c7d186277d6e1a9755029a (diff) | |
| parent | f76e4c167ea2212e23c15ee7e601a865e822c291 (diff) | |
| download | linux-bd2463ac7d7ec51d432f23bf0e893fb371a908cd.tar.bz2 | |
Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next
Pull networking updates from David Miller:
1) Add WireGuard
2) Add HE and TWT support to ath11k driver, from John Crispin.
3) Add ESP in TCP encapsulation support, from Sabrina Dubroca.
4) Add variable window congestion control to TIPC, from Jon Maloy.
5) Add BCM84881 PHY driver, from Russell King.
6) Start adding netlink support for ethtool operations, from Michal
Kubecek.
7) Add XDP drop and TX action support to ena driver, from Sameeh
Jubran.
8) Add new ipv4 route notifications so that mlxsw driver does not have
to handle identical routes itself. From Ido Schimmel.
9) Add BPF dynamic program extensions, from Alexei Starovoitov.
10) Support RX and TX timestamping in igc, from Vinicius Costa Gomes.
11) Add support for macsec HW offloading, from Antoine Tenart.
12) Add initial support for MPTCP protocol, from Christoph Paasch,
Matthieu Baerts, Florian Westphal, Peter Krystad, and many others.
13) Add Octeontx2 PF support, from Sunil Goutham, Geetha sowjanya, Linu
Cherian, and others.
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next: (1469 commits)
net: phy: add default ARCH_BCM_IPROC for MDIO_BCM_IPROC
udp: segment looped gso packets correctly
netem: change mailing list
qed: FW 8.42.2.0 debug features
qed: rt init valid initialization changed
qed: Debug feature: ilt and mdump
qed: FW 8.42.2.0 Add fw overlay feature
qed: FW 8.42.2.0 HSI changes
qed: FW 8.42.2.0 iscsi/fcoe changes
qed: Add abstraction for different hsi values per chip
qed: FW 8.42.2.0 Additional ll2 type
qed: Use dmae to write to widebus registers in fw_funcs
qed: FW 8.42.2.0 Parser offsets modified
qed: FW 8.42.2.0 Queue Manager changes
qed: FW 8.42.2.0 Expose new registers and change windows
qed: FW 8.42.2.0 Internal ram offsets modifications
MAINTAINERS: Add entry for Marvell OcteonTX2 Physical Function driver
Documentation: net: octeontx2: Add RVU HW and drivers overview
octeontx2-pf: ethtool RSS config support
octeontx2-pf: Add basic ethtool support
...
Diffstat (limited to 'drivers/net/ethernet')
487 files changed, 37290 insertions, 19696 deletions
diff --git a/drivers/net/ethernet/3com/3c509.c b/drivers/net/ethernet/3com/3c509.c index 3da97996bdf3..8cafd06ff0c4 100644 --- a/drivers/net/ethernet/3com/3c509.c +++ b/drivers/net/ethernet/3com/3c509.c @@ -196,7 +196,7 @@ static struct net_device_stats *el3_get_stats(struct net_device *dev); static int el3_rx(struct net_device *dev); static int el3_close(struct net_device *dev); static void set_multicast_list(struct net_device *dev); -static void el3_tx_timeout (struct net_device *dev); +static void el3_tx_timeout (struct net_device *dev, unsigned int txqueue); static void el3_down(struct net_device *dev); static void el3_up(struct net_device *dev); static const struct ethtool_ops ethtool_ops; @@ -689,7 +689,7 @@ el3_open(struct net_device *dev) } static void -el3_tx_timeout (struct net_device *dev) +el3_tx_timeout (struct net_device *dev, unsigned int txqueue) { int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/3com/3c515.c b/drivers/net/ethernet/3com/3c515.c index b15752267c8d..1e233e2f0a5a 100644 --- a/drivers/net/ethernet/3com/3c515.c +++ b/drivers/net/ethernet/3com/3c515.c @@ -371,7 +371,7 @@ static void corkscrew_timer(struct timer_list *t); static netdev_tx_t corkscrew_start_xmit(struct sk_buff *skb, struct net_device *dev); static int corkscrew_rx(struct net_device *dev); -static void corkscrew_timeout(struct net_device *dev); +static void corkscrew_timeout(struct net_device *dev, unsigned int txqueue); static int boomerang_rx(struct net_device *dev); static irqreturn_t corkscrew_interrupt(int irq, void *dev_id); static int corkscrew_close(struct net_device *dev); @@ -961,7 +961,7 @@ static void corkscrew_timer(struct timer_list *t) #endif /* AUTOMEDIA */ } -static void corkscrew_timeout(struct net_device *dev) +static void corkscrew_timeout(struct net_device *dev, unsigned int txqueue) { int i; struct corkscrew_private *vp = netdev_priv(dev); diff --git a/drivers/net/ethernet/3com/3c574_cs.c b/drivers/net/ethernet/3com/3c574_cs.c index 3044a6f35f04..ef1c3151fbb2 100644 --- a/drivers/net/ethernet/3com/3c574_cs.c +++ b/drivers/net/ethernet/3com/3c574_cs.c @@ -234,7 +234,7 @@ static void update_stats(struct net_device *dev); static struct net_device_stats *el3_get_stats(struct net_device *dev); static int el3_rx(struct net_device *dev, int worklimit); static int el3_close(struct net_device *dev); -static void el3_tx_timeout(struct net_device *dev); +static void el3_tx_timeout(struct net_device *dev, unsigned int txqueue); static int el3_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static void set_rx_mode(struct net_device *dev); static void set_multicast_list(struct net_device *dev); @@ -690,7 +690,7 @@ static int el3_open(struct net_device *dev) return 0; } -static void el3_tx_timeout(struct net_device *dev) +static void el3_tx_timeout(struct net_device *dev, unsigned int txqueue) { unsigned int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/3com/3c589_cs.c b/drivers/net/ethernet/3com/3c589_cs.c index 2b2695311bda..d47cde6c5f08 100644 --- a/drivers/net/ethernet/3com/3c589_cs.c +++ b/drivers/net/ethernet/3com/3c589_cs.c @@ -173,7 +173,7 @@ static void update_stats(struct net_device *dev); static struct net_device_stats *el3_get_stats(struct net_device *dev); static int el3_rx(struct net_device *dev); static int el3_close(struct net_device *dev); -static void el3_tx_timeout(struct net_device *dev); +static void el3_tx_timeout(struct net_device *dev, unsigned int txqueue); static void set_rx_mode(struct net_device *dev); static void set_multicast_list(struct net_device *dev); static const struct ethtool_ops netdev_ethtool_ops; @@ -526,7 +526,7 @@ static int el3_open(struct net_device *dev) return 0; } -static void el3_tx_timeout(struct net_device *dev) +static void el3_tx_timeout(struct net_device *dev, unsigned int txqueue) { unsigned int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/3com/3c59x.c b/drivers/net/ethernet/3com/3c59x.c index 8785c2ff3825..a2b7f7ab8170 100644 --- a/drivers/net/ethernet/3com/3c59x.c +++ b/drivers/net/ethernet/3com/3c59x.c @@ -776,7 +776,7 @@ static void set_rx_mode(struct net_device *dev); #ifdef CONFIG_PCI static int vortex_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); #endif -static void vortex_tx_timeout(struct net_device *dev); +static void vortex_tx_timeout(struct net_device *dev, unsigned int txqueue); static void acpi_set_WOL(struct net_device *dev); static const struct ethtool_ops vortex_ethtool_ops; static void set_8021q_mode(struct net_device *dev, int enable); @@ -1548,7 +1548,7 @@ vortex_up(struct net_device *dev) struct vortex_private *vp = netdev_priv(dev); void __iomem *ioaddr = vp->ioaddr; unsigned int config; - int i, mii_reg1, mii_reg5, err = 0; + int i, mii_reg5, err = 0; if (VORTEX_PCI(vp)) { pci_set_power_state(VORTEX_PCI(vp), PCI_D0); /* Go active */ @@ -1605,7 +1605,7 @@ vortex_up(struct net_device *dev) window_write32(vp, config, 3, Wn3_Config); if (dev->if_port == XCVR_MII || dev->if_port == XCVR_NWAY) { - mii_reg1 = mdio_read(dev, vp->phys[0], MII_BMSR); + mdio_read(dev, vp->phys[0], MII_BMSR); mii_reg5 = mdio_read(dev, vp->phys[0], MII_LPA); vp->partner_flow_ctrl = ((mii_reg5 & 0x0400) != 0); vp->mii.full_duplex = vp->full_duplex; @@ -1877,7 +1877,7 @@ leave_media_alone: iowrite16(FakeIntr, ioaddr + EL3_CMD); } -static void vortex_tx_timeout(struct net_device *dev) +static void vortex_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct vortex_private *vp = netdev_priv(dev); void __iomem *ioaddr = vp->ioaddr; diff --git a/drivers/net/ethernet/3com/typhoon.c b/drivers/net/ethernet/3com/typhoon.c index be823c186517..14fce6658106 100644 --- a/drivers/net/ethernet/3com/typhoon.c +++ b/drivers/net/ethernet/3com/typhoon.c @@ -2013,7 +2013,7 @@ typhoon_stop_runtime(struct typhoon *tp, int wait_type) } static void -typhoon_tx_timeout(struct net_device *dev) +typhoon_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct typhoon *tp = netdev_priv(dev); diff --git a/drivers/net/ethernet/8390/8390.c b/drivers/net/ethernet/8390/8390.c index 78f3e532c600..0e0aa4016858 100644 --- a/drivers/net/ethernet/8390/8390.c +++ b/drivers/net/ethernet/8390/8390.c @@ -36,9 +36,9 @@ void ei_set_multicast_list(struct net_device *dev) } EXPORT_SYMBOL(ei_set_multicast_list); -void ei_tx_timeout(struct net_device *dev) +void ei_tx_timeout(struct net_device *dev, unsigned int txqueue) { - __ei_tx_timeout(dev); + __ei_tx_timeout(dev, txqueue); } EXPORT_SYMBOL(ei_tx_timeout); diff --git a/drivers/net/ethernet/8390/8390.h b/drivers/net/ethernet/8390/8390.h index 3e2f2c2e7b58..529c728f334a 100644 --- a/drivers/net/ethernet/8390/8390.h +++ b/drivers/net/ethernet/8390/8390.h @@ -32,7 +32,7 @@ void NS8390_init(struct net_device *dev, int startp); int ei_open(struct net_device *dev); int ei_close(struct net_device *dev); irqreturn_t ei_interrupt(int irq, void *dev_id); -void ei_tx_timeout(struct net_device *dev); +void ei_tx_timeout(struct net_device *dev, unsigned int txqueue); netdev_tx_t ei_start_xmit(struct sk_buff *skb, struct net_device *dev); void ei_set_multicast_list(struct net_device *dev); struct net_device_stats *ei_get_stats(struct net_device *dev); @@ -50,7 +50,7 @@ void NS8390p_init(struct net_device *dev, int startp); int eip_open(struct net_device *dev); int eip_close(struct net_device *dev); irqreturn_t eip_interrupt(int irq, void *dev_id); -void eip_tx_timeout(struct net_device *dev); +void eip_tx_timeout(struct net_device *dev, unsigned int txqueue); netdev_tx_t eip_start_xmit(struct sk_buff *skb, struct net_device *dev); void eip_set_multicast_list(struct net_device *dev); struct net_device_stats *eip_get_stats(struct net_device *dev); diff --git a/drivers/net/ethernet/8390/8390p.c b/drivers/net/ethernet/8390/8390p.c index 6cf36992a2c6..6834742057b3 100644 --- a/drivers/net/ethernet/8390/8390p.c +++ b/drivers/net/ethernet/8390/8390p.c @@ -41,9 +41,9 @@ void eip_set_multicast_list(struct net_device *dev) } EXPORT_SYMBOL(eip_set_multicast_list); -void eip_tx_timeout(struct net_device *dev) +void eip_tx_timeout(struct net_device *dev, unsigned int txqueue) { - __ei_tx_timeout(dev); + __ei_tx_timeout(dev, txqueue); } EXPORT_SYMBOL(eip_tx_timeout); diff --git a/drivers/net/ethernet/8390/axnet_cs.c b/drivers/net/ethernet/8390/axnet_cs.c index 0b6bbf63f7ca..aeae7966a082 100644 --- a/drivers/net/ethernet/8390/axnet_cs.c +++ b/drivers/net/ethernet/8390/axnet_cs.c @@ -83,7 +83,7 @@ static netdev_tx_t axnet_start_xmit(struct sk_buff *skb, struct net_device *dev); static struct net_device_stats *get_stats(struct net_device *dev); static void set_multicast_list(struct net_device *dev); -static void axnet_tx_timeout(struct net_device *dev); +static void axnet_tx_timeout(struct net_device *dev, unsigned int txqueue); static irqreturn_t ei_irq_wrapper(int irq, void *dev_id); static void ei_watchdog(struct timer_list *t); static void axnet_reset_8390(struct net_device *dev); @@ -903,7 +903,7 @@ static int ax_close(struct net_device *dev) * completed (or failed) - i.e. never posted a Tx related interrupt. */ -static void axnet_tx_timeout(struct net_device *dev) +static void axnet_tx_timeout(struct net_device *dev, unsigned int txqueue) { long e8390_base = dev->base_addr; struct ei_device *ei_local = netdev_priv(dev); diff --git a/drivers/net/ethernet/8390/lib8390.c b/drivers/net/ethernet/8390/lib8390.c index c9c55c9eab9f..babc92e2692e 100644 --- a/drivers/net/ethernet/8390/lib8390.c +++ b/drivers/net/ethernet/8390/lib8390.c @@ -251,7 +251,7 @@ static int __ei_close(struct net_device *dev) * completed (or failed) - i.e. never posted a Tx related interrupt. */ -static void __ei_tx_timeout(struct net_device *dev) +static void __ei_tx_timeout(struct net_device *dev, unsigned int txqueue) { unsigned long e8390_base = dev->base_addr; struct ei_device *ei_local = netdev_priv(dev); diff --git a/drivers/net/ethernet/adaptec/starfire.c b/drivers/net/ethernet/adaptec/starfire.c index 816540e6beac..165d18405b0c 100644 --- a/drivers/net/ethernet/adaptec/starfire.c +++ b/drivers/net/ethernet/adaptec/starfire.c @@ -576,7 +576,7 @@ static int mdio_read(struct net_device *dev, int phy_id, int location); static void mdio_write(struct net_device *dev, int phy_id, int location, int value); static int netdev_open(struct net_device *dev); static void check_duplex(struct net_device *dev); -static void tx_timeout(struct net_device *dev); +static void tx_timeout(struct net_device *dev, unsigned int txqueue); static void init_ring(struct net_device *dev); static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); @@ -1105,7 +1105,7 @@ static void check_duplex(struct net_device *dev) } -static void tx_timeout(struct net_device *dev) +static void tx_timeout(struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem *ioaddr = np->base; diff --git a/drivers/net/ethernet/agere/et131x.c b/drivers/net/ethernet/agere/et131x.c index 174344c450af..cb6a761d5c11 100644 --- a/drivers/net/ethernet/agere/et131x.c +++ b/drivers/net/ethernet/agere/et131x.c @@ -3651,15 +3651,6 @@ static int et131x_close(struct net_device *netdev) return del_timer_sync(&adapter->error_timer); } -static int et131x_ioctl(struct net_device *netdev, struct ifreq *reqbuf, - int cmd) -{ - if (!netdev->phydev) - return -EINVAL; - - return phy_mii_ioctl(netdev->phydev, reqbuf, cmd); -} - /* et131x_set_packet_filter - Configures the Rx Packet filtering */ static int et131x_set_packet_filter(struct et131x_adapter *adapter) { @@ -3811,7 +3802,7 @@ drop_err: * specified by the 'tx_timeo" element in the net_device structure (see * et131x_alloc_device() to see how this value is set). */ -static void et131x_tx_timeout(struct net_device *netdev) +static void et131x_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct et131x_adapter *adapter = netdev_priv(netdev); struct tx_ring *tx_ring = &adapter->tx_ring; @@ -3899,7 +3890,7 @@ static const struct net_device_ops et131x_netdev_ops = { .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, .ndo_get_stats = et131x_stats, - .ndo_do_ioctl = et131x_ioctl, + .ndo_do_ioctl = phy_do_ioctl, }; static int et131x_pci_setup(struct pci_dev *pdev, diff --git a/drivers/net/ethernet/allwinner/sun4i-emac.c b/drivers/net/ethernet/allwinner/sun4i-emac.c index 0537df06a9b5..22cadfbeedfb 100644 --- a/drivers/net/ethernet/allwinner/sun4i-emac.c +++ b/drivers/net/ethernet/allwinner/sun4i-emac.c @@ -207,19 +207,6 @@ static void emac_inblk_32bit(void __iomem *reg, void *data, int count) readsl(reg, data, round_up(count, 4) / 4); } -static int emac_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - struct phy_device *phydev = dev->phydev; - - if (!netif_running(dev)) - return -EINVAL; - - if (!phydev) - return -ENODEV; - - return phy_mii_ioctl(phydev, rq, cmd); -} - /* ethtool ops */ static void emac_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) @@ -407,7 +394,7 @@ static void emac_init_device(struct net_device *dev) } /* Our watchdog timed out. Called by the networking layer */ -static void emac_timeout(struct net_device *dev) +static void emac_timeout(struct net_device *dev, unsigned int txqueue) { struct emac_board_info *db = netdev_priv(dev); unsigned long flags; @@ -791,7 +778,7 @@ static const struct net_device_ops emac_netdev_ops = { .ndo_start_xmit = emac_start_xmit, .ndo_tx_timeout = emac_timeout, .ndo_set_rx_mode = emac_set_rx_mode, - .ndo_do_ioctl = emac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = emac_set_mac_address, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/alteon/acenic.c b/drivers/net/ethernet/alteon/acenic.c index 46b4207d3266..f366faf88eee 100644 --- a/drivers/net/ethernet/alteon/acenic.c +++ b/drivers/net/ethernet/alteon/acenic.c @@ -437,7 +437,7 @@ static const struct ethtool_ops ace_ethtool_ops = { .set_link_ksettings = ace_set_link_ksettings, }; -static void ace_watchdog(struct net_device *dev); +static void ace_watchdog(struct net_device *dev, unsigned int txqueue); static const struct net_device_ops ace_netdev_ops = { .ndo_open = ace_open, @@ -1542,7 +1542,7 @@ static void ace_set_rxtx_parms(struct net_device *dev, int jumbo) } -static void ace_watchdog(struct net_device *data) +static void ace_watchdog(struct net_device *data, unsigned int txqueue) { struct net_device *dev = data; struct ace_private *ap = netdev_priv(dev); diff --git a/drivers/net/ethernet/amazon/ena/ena_ethtool.c b/drivers/net/ethernet/amazon/ena/ena_ethtool.c index fc96c66b44cb..b4e891d49a94 100644 --- a/drivers/net/ethernet/amazon/ena/ena_ethtool.c +++ b/drivers/net/ethernet/amazon/ena/ena_ethtool.c @@ -740,7 +740,9 @@ static int ena_set_channels(struct net_device *netdev, struct ena_adapter *adapter = netdev_priv(netdev); u32 count = channels->combined_count; /* The check for max value is already done in ethtool */ - if (count < ENA_MIN_NUM_IO_QUEUES) + if (count < ENA_MIN_NUM_IO_QUEUES || + (ena_xdp_present(adapter) && + !ena_xdp_legal_queue_count(adapter, channels->combined_count))) return -EINVAL; return ena_update_queue_count(adapter, count); diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.c b/drivers/net/ethernet/amazon/ena/ena_netdev.c index 948583fdcc28..894e8c1a8cf1 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.c +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.c @@ -36,7 +36,6 @@ #include <linux/cpu_rmap.h> #endif /* CONFIG_RFS_ACCEL */ #include <linux/ethtool.h> -#include <linux/if_vlan.h> #include <linux/kernel.h> #include <linux/module.h> #include <linux/numa.h> @@ -47,6 +46,7 @@ #include <net/ip.h> #include "ena_netdev.h" +#include <linux/bpf_trace.h> #include "ena_pci_id_tbl.h" static char version[] = DEVICE_NAME " v" DRV_MODULE_VERSION "\n"; @@ -78,7 +78,37 @@ static void check_for_admin_com_state(struct ena_adapter *adapter); static void ena_destroy_device(struct ena_adapter *adapter, bool graceful); static int ena_restore_device(struct ena_adapter *adapter); -static void ena_tx_timeout(struct net_device *dev) +static void ena_init_io_rings(struct ena_adapter *adapter, + int first_index, int count); +static void ena_init_napi_in_range(struct ena_adapter *adapter, int first_index, + int count); +static void ena_del_napi_in_range(struct ena_adapter *adapter, int first_index, + int count); +static int ena_setup_tx_resources(struct ena_adapter *adapter, int qid); +static int ena_setup_tx_resources_in_range(struct ena_adapter *adapter, + int first_index, + int count); +static int ena_create_io_tx_queue(struct ena_adapter *adapter, int qid); +static void ena_free_tx_resources(struct ena_adapter *adapter, int qid); +static int ena_clean_xdp_irq(struct ena_ring *xdp_ring, u32 budget); +static void ena_destroy_all_tx_queues(struct ena_adapter *adapter); +static void ena_free_all_io_tx_resources(struct ena_adapter *adapter); +static void ena_napi_disable_in_range(struct ena_adapter *adapter, + int first_index, int count); +static void ena_napi_enable_in_range(struct ena_adapter *adapter, + int first_index, int count); +static int ena_up(struct ena_adapter *adapter); +static void ena_down(struct ena_adapter *adapter); +static void ena_unmask_interrupt(struct ena_ring *tx_ring, + struct ena_ring *rx_ring); +static void ena_update_ring_numa_node(struct ena_ring *tx_ring, + struct ena_ring *rx_ring); +static void ena_unmap_tx_buff(struct ena_ring *tx_ring, + struct ena_tx_buffer *tx_info); +static int ena_create_io_tx_queues_in_range(struct ena_adapter *adapter, + int first_index, int count); + +static void ena_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct ena_adapter *adapter = netdev_priv(dev); @@ -123,6 +153,448 @@ static int ena_change_mtu(struct net_device *dev, int new_mtu) return ret; } +static int ena_xmit_common(struct net_device *dev, + struct ena_ring *ring, + struct ena_tx_buffer *tx_info, + struct ena_com_tx_ctx *ena_tx_ctx, + u16 next_to_use, + u32 bytes) +{ + struct ena_adapter *adapter = netdev_priv(dev); + int rc, nb_hw_desc; + + if (unlikely(ena_com_is_doorbell_needed(ring->ena_com_io_sq, + ena_tx_ctx))) { + netif_dbg(adapter, tx_queued, dev, + "llq tx max burst size of queue %d achieved, writing doorbell to send burst\n", + ring->qid); + ena_com_write_sq_doorbell(ring->ena_com_io_sq); + } + + /* prepare the packet's descriptors to dma engine */ + rc = ena_com_prepare_tx(ring->ena_com_io_sq, ena_tx_ctx, + &nb_hw_desc); + + /* In case there isn't enough space in the queue for the packet, + * we simply drop it. All other failure reasons of + * ena_com_prepare_tx() are fatal and therefore require a device reset. + */ + if (unlikely(rc)) { + netif_err(adapter, tx_queued, dev, + "failed to prepare tx bufs\n"); + u64_stats_update_begin(&ring->syncp); + ring->tx_stats.prepare_ctx_err++; + u64_stats_update_end(&ring->syncp); + if (rc != -ENOMEM) { + adapter->reset_reason = + ENA_REGS_RESET_DRIVER_INVALID_STATE; + set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags); + } + return rc; + } + + u64_stats_update_begin(&ring->syncp); + ring->tx_stats.cnt++; + ring->tx_stats.bytes += bytes; + u64_stats_update_end(&ring->syncp); + + tx_info->tx_descs = nb_hw_desc; + tx_info->last_jiffies = jiffies; + tx_info->print_once = 0; + + ring->next_to_use = ENA_TX_RING_IDX_NEXT(next_to_use, + ring->ring_size); + return 0; +} + +/* This is the XDP napi callback. XDP queues use a separate napi callback + * than Rx/Tx queues. + */ +static int ena_xdp_io_poll(struct napi_struct *napi, int budget) +{ + struct ena_napi *ena_napi = container_of(napi, struct ena_napi, napi); + u32 xdp_work_done, xdp_budget; + struct ena_ring *xdp_ring; + int napi_comp_call = 0; + int ret; + + xdp_ring = ena_napi->xdp_ring; + xdp_ring->first_interrupt = ena_napi->first_interrupt; + + xdp_budget = budget; + + if (!test_bit(ENA_FLAG_DEV_UP, &xdp_ring->adapter->flags) || + test_bit(ENA_FLAG_TRIGGER_RESET, &xdp_ring->adapter->flags)) { + napi_complete_done(napi, 0); + return 0; + } + + xdp_work_done = ena_clean_xdp_irq(xdp_ring, xdp_budget); + + /* If the device is about to reset or down, avoid unmask + * the interrupt and return 0 so NAPI won't reschedule + */ + if (unlikely(!test_bit(ENA_FLAG_DEV_UP, &xdp_ring->adapter->flags))) { + napi_complete_done(napi, 0); + ret = 0; + } else if (xdp_budget > xdp_work_done) { + napi_comp_call = 1; + if (napi_complete_done(napi, xdp_work_done)) + ena_unmask_interrupt(xdp_ring, NULL); + ena_update_ring_numa_node(xdp_ring, NULL); + ret = xdp_work_done; + } else { + ret = xdp_budget; + } + + u64_stats_update_begin(&xdp_ring->syncp); + xdp_ring->tx_stats.napi_comp += napi_comp_call; + xdp_ring->tx_stats.tx_poll++; + u64_stats_update_end(&xdp_ring->syncp); + + return ret; +} + +static int ena_xdp_tx_map_buff(struct ena_ring *xdp_ring, + struct ena_tx_buffer *tx_info, + struct xdp_buff *xdp, + void **push_hdr, + u32 *push_len) +{ + struct ena_adapter *adapter = xdp_ring->adapter; + struct ena_com_buf *ena_buf; + dma_addr_t dma = 0; + u32 size; + + tx_info->xdpf = convert_to_xdp_frame(xdp); + size = tx_info->xdpf->len; + ena_buf = tx_info->bufs; + + /* llq push buffer */ + *push_len = min_t(u32, size, xdp_ring->tx_max_header_size); + *push_hdr = tx_info->xdpf->data; + + if (size - *push_len > 0) { + dma = dma_map_single(xdp_ring->dev, + *push_hdr + *push_len, + size - *push_len, + DMA_TO_DEVICE); + if (unlikely(dma_mapping_error(xdp_ring->dev, dma))) + goto error_report_dma_error; + + tx_info->map_linear_data = 1; + tx_info->num_of_bufs = 1; + } + + ena_buf->paddr = dma; + ena_buf->len = size; + + return 0; + +error_report_dma_error: + u64_stats_update_begin(&xdp_ring->syncp); + xdp_ring->tx_stats.dma_mapping_err++; + u64_stats_update_end(&xdp_ring->syncp); + netdev_warn(adapter->netdev, "failed to map xdp buff\n"); + + xdp_return_frame_rx_napi(tx_info->xdpf); + tx_info->xdpf = NULL; + tx_info->num_of_bufs = 0; + + return -EINVAL; +} + +static int ena_xdp_xmit_buff(struct net_device *dev, + struct xdp_buff *xdp, + int qid, + struct ena_rx_buffer *rx_info) +{ + struct ena_adapter *adapter = netdev_priv(dev); + struct ena_com_tx_ctx ena_tx_ctx = {0}; + struct ena_tx_buffer *tx_info; + struct ena_ring *xdp_ring; + u16 next_to_use, req_id; + int rc; + void *push_hdr; + u32 push_len; + + xdp_ring = &adapter->tx_ring[qid]; + next_to_use = xdp_ring->next_to_use; + req_id = xdp_ring->free_ids[next_to_use]; + tx_info = &xdp_ring->tx_buffer_info[req_id]; + tx_info->num_of_bufs = 0; + page_ref_inc(rx_info->page); + tx_info->xdp_rx_page = rx_info->page; + + rc = ena_xdp_tx_map_buff(xdp_ring, tx_info, xdp, &push_hdr, &push_len); + if (unlikely(rc)) + goto error_drop_packet; + + ena_tx_ctx.ena_bufs = tx_info->bufs; + ena_tx_ctx.push_header = push_hdr; + ena_tx_ctx.num_bufs = tx_info->num_of_bufs; + ena_tx_ctx.req_id = req_id; + ena_tx_ctx.header_len = push_len; + + rc = ena_xmit_common(dev, + xdp_ring, + tx_info, + &ena_tx_ctx, + next_to_use, + xdp->data_end - xdp->data); + if (rc) + goto error_unmap_dma; + /* trigger the dma engine. ena_com_write_sq_doorbell() + * has a mb + */ + ena_com_write_sq_doorbell(xdp_ring->ena_com_io_sq); + u64_stats_update_begin(&xdp_ring->syncp); + xdp_ring->tx_stats.doorbells++; + u64_stats_update_end(&xdp_ring->syncp); + + return NETDEV_TX_OK; + +error_unmap_dma: + ena_unmap_tx_buff(xdp_ring, tx_info); + tx_info->xdpf = NULL; +error_drop_packet: + + return NETDEV_TX_OK; +} + +static int ena_xdp_execute(struct ena_ring *rx_ring, + struct xdp_buff *xdp, + struct ena_rx_buffer *rx_info) +{ + struct bpf_prog *xdp_prog; + u32 verdict = XDP_PASS; + + rcu_read_lock(); + xdp_prog = READ_ONCE(rx_ring->xdp_bpf_prog); + + if (!xdp_prog) + goto out; + + verdict = bpf_prog_run_xdp(xdp_prog, xdp); + + if (verdict == XDP_TX) + ena_xdp_xmit_buff(rx_ring->netdev, + xdp, + rx_ring->qid + rx_ring->adapter->num_io_queues, + rx_info); + else if (unlikely(verdict == XDP_ABORTED)) + trace_xdp_exception(rx_ring->netdev, xdp_prog, verdict); + else if (unlikely(verdict > XDP_TX)) + bpf_warn_invalid_xdp_action(verdict); +out: + rcu_read_unlock(); + return verdict; +} + +static void ena_init_all_xdp_queues(struct ena_adapter *adapter) +{ + adapter->xdp_first_ring = adapter->num_io_queues; + adapter->xdp_num_queues = adapter->num_io_queues; + + ena_init_io_rings(adapter, + adapter->xdp_first_ring, + adapter->xdp_num_queues); +} + +static int ena_setup_and_create_all_xdp_queues(struct ena_adapter *adapter) +{ + int rc = 0; + + rc = ena_setup_tx_resources_in_range(adapter, adapter->xdp_first_ring, + adapter->xdp_num_queues); + if (rc) + goto setup_err; + + rc = ena_create_io_tx_queues_in_range(adapter, + adapter->xdp_first_ring, + adapter->xdp_num_queues); + if (rc) + goto create_err; + + return 0; + +create_err: + ena_free_all_io_tx_resources(adapter); +setup_err: + return rc; +} + +/* Provides a way for both kernel and bpf-prog to know + * more about the RX-queue a given XDP frame arrived on. + */ +static int ena_xdp_register_rxq_info(struct ena_ring *rx_ring) +{ + int rc; + + rc = xdp_rxq_info_reg(&rx_ring->xdp_rxq, rx_ring->netdev, rx_ring->qid); + + if (rc) { + netif_err(rx_ring->adapter, ifup, rx_ring->netdev, + "Failed to register xdp rx queue info. RX queue num %d rc: %d\n", + rx_ring->qid, rc); + goto err; + } + + rc = xdp_rxq_info_reg_mem_model(&rx_ring->xdp_rxq, MEM_TYPE_PAGE_SHARED, + NULL); + + if (rc) { + netif_err(rx_ring->adapter, ifup, rx_ring->netdev, + "Failed to register xdp rx queue info memory model. RX queue num %d rc: %d\n", + rx_ring->qid, rc); + xdp_rxq_info_unreg(&rx_ring->xdp_rxq); + } + +err: + return rc; +} + +static void ena_xdp_unregister_rxq_info(struct ena_ring *rx_ring) +{ + xdp_rxq_info_unreg_mem_model(&rx_ring->xdp_rxq); + xdp_rxq_info_unreg(&rx_ring->xdp_rxq); +} + +void ena_xdp_exchange_program_rx_in_range(struct ena_adapter *adapter, + struct bpf_prog *prog, + int first, + int count) +{ + struct ena_ring *rx_ring; + int i = 0; + + for (i = first; i < count; i++) { + rx_ring = &adapter->rx_ring[i]; + xchg(&rx_ring->xdp_bpf_prog, prog); + if (prog) { + ena_xdp_register_rxq_info(rx_ring); + rx_ring->rx_headroom = XDP_PACKET_HEADROOM; + } else { + ena_xdp_unregister_rxq_info(rx_ring); + rx_ring->rx_headroom = 0; + } + } +} + +void ena_xdp_exchange_program(struct ena_adapter *adapter, + struct bpf_prog *prog) +{ + struct bpf_prog *old_bpf_prog = xchg(&adapter->xdp_bpf_prog, prog); + + ena_xdp_exchange_program_rx_in_range(adapter, + prog, + 0, + adapter->num_io_queues); + + if (old_bpf_prog) + bpf_prog_put(old_bpf_prog); +} + +static int ena_destroy_and_free_all_xdp_queues(struct ena_adapter *adapter) +{ + bool was_up; + int rc; + + was_up = test_bit(ENA_FLAG_DEV_UP, &adapter->flags); + + if (was_up) + ena_down(adapter); + + adapter->xdp_first_ring = 0; + adapter->xdp_num_queues = 0; + ena_xdp_exchange_program(adapter, NULL); + if (was_up) { + rc = ena_up(adapter); + if (rc) + return rc; + } + return 0; +} + +static int ena_xdp_set(struct net_device *netdev, struct netdev_bpf *bpf) +{ + struct ena_adapter *adapter = netdev_priv(netdev); + struct bpf_prog *prog = bpf->prog; + struct bpf_prog *old_bpf_prog; + int rc, prev_mtu; + bool is_up; + + is_up = test_bit(ENA_FLAG_DEV_UP, &adapter->flags); + rc = ena_xdp_allowed(adapter); + if (rc == ENA_XDP_ALLOWED) { + old_bpf_prog = adapter->xdp_bpf_prog; + if (prog) { + if (!is_up) { + ena_init_all_xdp_queues(adapter); + } else if (!old_bpf_prog) { + ena_down(adapter); + ena_init_all_xdp_queues(adapter); + } + ena_xdp_exchange_program(adapter, prog); + + if (is_up && !old_bpf_prog) { + rc = ena_up(adapter); + if (rc) + return rc; + } + } else if (old_bpf_prog) { + rc = ena_destroy_and_free_all_xdp_queues(adapter); + if (rc) + return rc; + } + + prev_mtu = netdev->max_mtu; + netdev->max_mtu = prog ? ENA_XDP_MAX_MTU : adapter->max_mtu; + + if (!old_bpf_prog) + netif_info(adapter, drv, adapter->netdev, + "xdp program set, changing the max_mtu from %d to %d", + prev_mtu, netdev->max_mtu); + + } else if (rc == ENA_XDP_CURRENT_MTU_TOO_LARGE) { + netif_err(adapter, drv, adapter->netdev, + "Failed to set xdp program, the current MTU (%d) is larger than the maximum allowed MTU (%lu) while xdp is on", + netdev->mtu, ENA_XDP_MAX_MTU); + NL_SET_ERR_MSG_MOD(bpf->extack, + "Failed to set xdp program, the current MTU is larger than the maximum allowed MTU. Check the dmesg for more info"); + return -EINVAL; + } else if (rc == ENA_XDP_NO_ENOUGH_QUEUES) { + netif_err(adapter, drv, adapter->netdev, + "Failed to set xdp program, the Rx/Tx channel count should be at most half of the maximum allowed channel count. The current queue count (%d), the maximal queue count (%d)\n", + adapter->num_io_queues, adapter->max_num_io_queues); + NL_SET_ERR_MSG_MOD(bpf->extack, + "Failed to set xdp program, there is no enough space for allocating XDP queues, Check the dmesg for more info"); + return -EINVAL; + } + + return 0; +} + +/* This is the main xdp callback, it's used by the kernel to set/unset the xdp + * program as well as to query the current xdp program id. + */ +static int ena_xdp(struct net_device *netdev, struct netdev_bpf *bpf) +{ + struct ena_adapter *adapter = netdev_priv(netdev); + + switch (bpf->command) { + case XDP_SETUP_PROG: + return ena_xdp_set(netdev, bpf); + case XDP_QUERY_PROG: + bpf->prog_id = adapter->xdp_bpf_prog ? + adapter->xdp_bpf_prog->aux->id : 0; + break; + default: + return -EINVAL; + } + return 0; +} + static int ena_init_rx_cpu_rmap(struct ena_adapter *adapter) { #ifdef CONFIG_RFS_ACCEL @@ -164,7 +636,8 @@ static void ena_init_io_rings_common(struct ena_adapter *adapter, u64_stats_init(&ring->syncp); } -static void ena_init_io_rings(struct ena_adapter *adapter) +static void ena_init_io_rings(struct ena_adapter *adapter, + int first_index, int count) { struct ena_com_dev *ena_dev; struct ena_ring *txr, *rxr; @@ -172,13 +645,12 @@ static void ena_init_io_rings(struct ena_adapter *adapter) ena_dev = adapter->ena_dev; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = first_index; i < first_index + count; i++) { txr = &adapter->tx_ring[i]; rxr = &adapter->rx_ring[i]; - /* TX/RX common ring state */ + /* TX common ring state */ ena_init_io_rings_common(adapter, txr, i); - ena_init_io_rings_common(adapter, rxr, i); /* TX specific ring state */ txr->ring_size = adapter->requested_tx_ring_size; @@ -188,14 +660,20 @@ static void ena_init_io_rings(struct ena_adapter *adapter) txr->smoothed_interval = ena_com_get_nonadaptive_moderation_interval_tx(ena_dev); - /* RX specific ring state */ - rxr->ring_size = adapter->requested_rx_ring_size; - rxr->rx_copybreak = adapter->rx_copybreak; - rxr->sgl_size = adapter->max_rx_sgl_size; - rxr->smoothed_interval = - ena_com_get_nonadaptive_moderation_interval_rx(ena_dev); - rxr->empty_rx_queue = 0; - adapter->ena_napi[i].dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; + /* Don't init RX queues for xdp queues */ + if (!ENA_IS_XDP_INDEX(adapter, i)) { + /* RX common ring state */ + ena_init_io_rings_common(adapter, rxr, i); + + /* RX specific ring state */ + rxr->ring_size = adapter->requested_rx_ring_size; + rxr->rx_copybreak = adapter->rx_copybreak; + rxr->sgl_size = adapter->max_rx_sgl_size; + rxr->smoothed_interval = + ena_com_get_nonadaptive_moderation_interval_rx(ena_dev); + rxr->empty_rx_queue = 0; + adapter->ena_napi[i].dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_EQE; + } } } @@ -285,16 +763,13 @@ static void ena_free_tx_resources(struct ena_adapter *adapter, int qid) tx_ring->push_buf_intermediate_buf = NULL; } -/* ena_setup_all_tx_resources - allocate I/O Tx queues resources for All queues - * @adapter: private structure - * - * Return 0 on success, negative on failure - */ -static int ena_setup_all_tx_resources(struct ena_adapter *adapter) +static int ena_setup_tx_resources_in_range(struct ena_adapter *adapter, + int first_index, + int count) { int i, rc = 0; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = first_index; i < first_index + count; i++) { rc = ena_setup_tx_resources(adapter, i); if (rc) goto err_setup_tx; @@ -308,11 +783,20 @@ err_setup_tx: "Tx queue %d: allocation failed\n", i); /* rewind the index freeing the rings as we go */ - while (i--) + while (first_index < i--) ena_free_tx_resources(adapter, i); return rc; } +static void ena_free_all_io_tx_resources_in_range(struct ena_adapter *adapter, + int first_index, int count) +{ + int i; + + for (i = first_index; i < first_index + count; i++) + ena_free_tx_resources(adapter, i); +} + /* ena_free_all_io_tx_resources - Free I/O Tx Resources for All Queues * @adapter: board private structure * @@ -320,10 +804,10 @@ err_setup_tx: */ static void ena_free_all_io_tx_resources(struct ena_adapter *adapter) { - int i; - - for (i = 0; i < adapter->num_io_queues; i++) - ena_free_tx_resources(adapter, i); + ena_free_all_io_tx_resources_in_range(adapter, + 0, + adapter->xdp_num_queues + + adapter->num_io_queues); } static int validate_rx_req_id(struct ena_ring *rx_ring, u16 req_id) @@ -495,8 +979,8 @@ static int ena_alloc_rx_page(struct ena_ring *rx_ring, rx_info->page = page; rx_info->page_offset = 0; ena_buf = &rx_info->ena_buf; - ena_buf->paddr = dma; - ena_buf->len = ENA_PAGE_SIZE; + ena_buf->paddr = dma + rx_ring->rx_headroom; + ena_buf->len = ENA_PAGE_SIZE - rx_ring->rx_headroom; return 0; } @@ -513,7 +997,9 @@ static void ena_free_rx_page(struct ena_ring *rx_ring, return; } - dma_unmap_page(rx_ring->dev, ena_buf->paddr, ENA_PAGE_SIZE, + dma_unmap_page(rx_ring->dev, + ena_buf->paddr - rx_ring->rx_headroom, + ENA_PAGE_SIZE, DMA_FROM_DEVICE); __free_page(page); @@ -620,8 +1106,8 @@ static void ena_free_all_rx_bufs(struct ena_adapter *adapter) ena_free_rx_bufs(adapter, i); } -static void ena_unmap_tx_skb(struct ena_ring *tx_ring, - struct ena_tx_buffer *tx_info) +static void ena_unmap_tx_buff(struct ena_ring *tx_ring, + struct ena_tx_buffer *tx_info) { struct ena_com_buf *ena_buf; u32 cnt; @@ -675,7 +1161,7 @@ static void ena_free_tx_bufs(struct ena_ring *tx_ring) tx_ring->qid, i); } - ena_unmap_tx_skb(tx_ring, tx_info); + ena_unmap_tx_buff(tx_ring, tx_info); dev_kfree_skb_any(tx_info->skb); } @@ -688,7 +1174,7 @@ static void ena_free_all_tx_bufs(struct ena_adapter *adapter) struct ena_ring *tx_ring; int i; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = 0; i < adapter->num_io_queues + adapter->xdp_num_queues; i++) { tx_ring = &adapter->tx_ring[i]; ena_free_tx_bufs(tx_ring); } @@ -699,7 +1185,7 @@ static void ena_destroy_all_tx_queues(struct ena_adapter *adapter) u16 ena_qid; int i; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = 0; i < adapter->num_io_queues + adapter->xdp_num_queues; i++) { ena_qid = ENA_IO_TXQ_IDX(i); ena_com_destroy_io_queue(adapter->ena_dev, ena_qid); } @@ -723,6 +1209,32 @@ static void ena_destroy_all_io_queues(struct ena_adapter *adapter) ena_destroy_all_rx_queues(adapter); } +static int handle_invalid_req_id(struct ena_ring *ring, u16 req_id, + struct ena_tx_buffer *tx_info, bool is_xdp) +{ + if (tx_info) + netif_err(ring->adapter, + tx_done, + ring->netdev, + "tx_info doesn't have valid %s", + is_xdp ? "xdp frame" : "skb"); + else + netif_err(ring->adapter, + tx_done, + ring->netdev, + "Invalid req_id: %hu\n", + req_id); + + u64_stats_update_begin(&ring->syncp); + ring->tx_stats.bad_req_id++; + u64_stats_update_end(&ring->syncp); + + /* Trigger device reset */ + ring->adapter->reset_reason = ENA_REGS_RESET_INV_TX_REQ_ID; + set_bit(ENA_FLAG_TRIGGER_RESET, &ring->adapter->flags); + return -EFAULT; +} + static int validate_tx_req_id(struct ena_ring *tx_ring, u16 req_id) { struct ena_tx_buffer *tx_info = NULL; @@ -733,21 +1245,20 @@ static int validate_tx_req_id(struct ena_ring *tx_ring, u16 req_id) return 0; } - if (tx_info) - netif_err(tx_ring->adapter, tx_done, tx_ring->netdev, - "tx_info doesn't have valid skb\n"); - else - netif_err(tx_ring->adapter, tx_done, tx_ring->netdev, - "Invalid req_id: %hu\n", req_id); + return handle_invalid_req_id(tx_ring, req_id, tx_info, false); +} - u64_stats_update_begin(&tx_ring->syncp); - tx_ring->tx_stats.bad_req_id++; - u64_stats_update_end(&tx_ring->syncp); +static int validate_xdp_req_id(struct ena_ring *xdp_ring, u16 req_id) +{ + struct ena_tx_buffer *tx_info = NULL; - /* Trigger device reset */ - tx_ring->adapter->reset_reason = ENA_REGS_RESET_INV_TX_REQ_ID; - set_bit(ENA_FLAG_TRIGGER_RESET, &tx_ring->adapter->flags); - return -EFAULT; + if (likely(req_id < xdp_ring->ring_size)) { + tx_info = &xdp_ring->tx_buffer_info[req_id]; + if (likely(tx_info->xdpf)) + return 0; + } + + return handle_invalid_req_id(xdp_ring, req_id, tx_info, true); } static int ena_clean_tx_irq(struct ena_ring *tx_ring, u32 budget) @@ -786,7 +1297,7 @@ static int ena_clean_tx_irq(struct ena_ring *tx_ring, u32 budget) tx_info->skb = NULL; tx_info->last_jiffies = 0; - ena_unmap_tx_skb(tx_ring, tx_info); + ena_unmap_tx_buff(tx_ring, tx_info); netif_dbg(tx_ring->adapter, tx_done, tx_ring->netdev, "tx_poll: q %d skb %p completed\n", tx_ring->qid, @@ -1037,6 +1548,33 @@ static void ena_set_rx_hash(struct ena_ring *rx_ring, } } +int ena_xdp_handle_buff(struct ena_ring *rx_ring, struct xdp_buff *xdp) +{ + struct ena_rx_buffer *rx_info; + int ret; + + rx_info = &rx_ring->rx_buffer_info[rx_ring->ena_bufs[0].req_id]; + xdp->data = page_address(rx_info->page) + + rx_info->page_offset + rx_ring->rx_headroom; + xdp_set_data_meta_invalid(xdp); + xdp->data_hard_start = page_address(rx_info->page); + xdp->data_end = xdp->data + rx_ring->ena_bufs[0].len; + /* If for some reason we received a bigger packet than + * we expect, then we simply drop it + */ + if (unlikely(rx_ring->ena_bufs[0].len > ENA_XDP_MAX_MTU)) + return XDP_DROP; + + ret = ena_xdp_execute(rx_ring, xdp, rx_info); + + /* The xdp program might expand the headers */ + if (ret == XDP_PASS) { + rx_info->page_offset = xdp->data - xdp->data_hard_start; + rx_ring->ena_bufs[0].len = xdp->data_end - xdp->data; + } + + return ret; +} /* ena_clean_rx_irq - Cleanup RX irq * @rx_ring: RX ring to clean * @napi: napi handler @@ -1048,23 +1586,27 @@ static int ena_clean_rx_irq(struct ena_ring *rx_ring, struct napi_struct *napi, u32 budget) { u16 next_to_clean = rx_ring->next_to_clean; - u32 res_budget, work_done; - struct ena_com_rx_ctx ena_rx_ctx; struct ena_adapter *adapter; + u32 res_budget, work_done; + int rx_copybreak_pkt = 0; + int refill_threshold; struct sk_buff *skb; int refill_required; - int refill_threshold; - int rc = 0; + struct xdp_buff xdp; int total_len = 0; - int rx_copybreak_pkt = 0; + int xdp_verdict; + int rc = 0; int i; netif_dbg(rx_ring->adapter, rx_status, rx_ring->netdev, "%s qid %d\n", __func__, rx_ring->qid); res_budget = budget; + xdp.rxq = &rx_ring->xdp_rxq; do { + xdp_verdict = XDP_PASS; + skb = NULL; ena_rx_ctx.ena_bufs = rx_ring->ena_bufs; ena_rx_ctx.max_bufs = rx_ring->sgl_size; ena_rx_ctx.descs = 0; @@ -1082,12 +1624,22 @@ static int ena_clean_rx_irq(struct ena_ring *rx_ring, struct napi_struct *napi, rx_ring->qid, ena_rx_ctx.descs, ena_rx_ctx.l3_proto, ena_rx_ctx.l4_proto, ena_rx_ctx.hash); + if (ena_xdp_present_ring(rx_ring)) + xdp_verdict = ena_xdp_handle_buff(rx_ring, &xdp); + /* allocate skb and fill it */ - skb = ena_rx_skb(rx_ring, rx_ring->ena_bufs, ena_rx_ctx.descs, - &next_to_clean); + if (xdp_verdict == XDP_PASS) + skb = ena_rx_skb(rx_ring, + rx_ring->ena_bufs, + ena_rx_ctx.descs, + &next_to_clean); - /* exit if we failed to retrieve a buffer */ if (unlikely(!skb)) { + if (xdp_verdict == XDP_TX) { + ena_free_rx_page(rx_ring, + &rx_ring->rx_buffer_info[rx_ring->ena_bufs[0].req_id]); + res_budget--; + } for (i = 0; i < ena_rx_ctx.descs; i++) { rx_ring->free_ids[next_to_clean] = rx_ring->ena_bufs[i].req_id; @@ -1095,6 +1647,8 @@ static int ena_clean_rx_irq(struct ena_ring *rx_ring, struct napi_struct *napi, ENA_RX_RING_IDX_NEXT(next_to_clean, rx_ring->ring_size); } + if (xdp_verdict == XDP_TX || xdp_verdict == XDP_DROP) + continue; break; } @@ -1188,9 +1742,14 @@ static void ena_unmask_interrupt(struct ena_ring *tx_ring, struct ena_ring *rx_ring) { struct ena_eth_io_intr_reg intr_reg; - u32 rx_interval = ena_com_get_adaptive_moderation_enabled(rx_ring->ena_dev) ? - rx_ring->smoothed_interval : - ena_com_get_nonadaptive_moderation_interval_rx(rx_ring->ena_dev); + u32 rx_interval = 0; + /* Rx ring can be NULL when for XDP tx queues which don't have an + * accompanying rx_ring pair. + */ + if (rx_ring) + rx_interval = ena_com_get_adaptive_moderation_enabled(rx_ring->ena_dev) ? + rx_ring->smoothed_interval : + ena_com_get_nonadaptive_moderation_interval_rx(rx_ring->ena_dev); /* Update intr register: rx intr delay, * tx intr delay and interrupt unmask @@ -1203,8 +1762,9 @@ static void ena_unmask_interrupt(struct ena_ring *tx_ring, /* It is a shared MSI-X. * Tx and Rx CQ have pointer to it. * So we use one of them to reach the intr reg + * The Tx ring is used because the rx_ring is NULL for XDP queues */ - ena_com_unmask_intr(rx_ring->ena_com_io_cq, &intr_reg); + ena_com_unmask_intr(tx_ring->ena_com_io_cq, &intr_reg); } static void ena_update_ring_numa_node(struct ena_ring *tx_ring, @@ -1222,22 +1782,82 @@ static void ena_update_ring_numa_node(struct ena_ring *tx_ring, if (numa_node != NUMA_NO_NODE) { ena_com_update_numa_node(tx_ring->ena_com_io_cq, numa_node); - ena_com_update_numa_node(rx_ring->ena_com_io_cq, numa_node); + if (rx_ring) + ena_com_update_numa_node(rx_ring->ena_com_io_cq, + numa_node); } tx_ring->cpu = cpu; - rx_ring->cpu = cpu; + if (rx_ring) + rx_ring->cpu = cpu; return; out: put_cpu(); } +static int ena_clean_xdp_irq(struct ena_ring *xdp_ring, u32 budget) +{ + u32 total_done = 0; + u16 next_to_clean; + u32 tx_bytes = 0; + int tx_pkts = 0; + u16 req_id; + int rc; + + if (unlikely(!xdp_ring)) + return 0; + next_to_clean = xdp_ring->next_to_clean; + + while (tx_pkts < budget) { + struct ena_tx_buffer *tx_info; + struct xdp_frame *xdpf; + + rc = ena_com_tx_comp_req_id_get(xdp_ring->ena_com_io_cq, + &req_id); + if (rc) + break; + + rc = validate_xdp_req_id(xdp_ring, req_id); + if (rc) + break; + + tx_info = &xdp_ring->tx_buffer_info[req_id]; + xdpf = tx_info->xdpf; + + tx_info->xdpf = NULL; + tx_info->last_jiffies = 0; + ena_unmap_tx_buff(xdp_ring, tx_info); + + netif_dbg(xdp_ring->adapter, tx_done, xdp_ring->netdev, + "tx_poll: q %d skb %p completed\n", xdp_ring->qid, + xdpf); + + tx_bytes += xdpf->len; + tx_pkts++; + total_done += tx_info->tx_descs; + + __free_page(tx_info->xdp_rx_page); + xdp_ring->free_ids[next_to_clean] = req_id; + next_to_clean = ENA_TX_RING_IDX_NEXT(next_to_clean, + xdp_ring->ring_size); + } + + xdp_ring->next_to_clean = next_to_clean; + ena_com_comp_ack(xdp_ring->ena_com_io_sq, total_done); + ena_com_update_dev_comp_head(xdp_ring->ena_com_io_cq); + + netif_dbg(xdp_ring->adapter, tx_done, xdp_ring->netdev, + "tx_poll: q %d done. total pkts: %d\n", + xdp_ring->qid, tx_pkts); + + return tx_pkts; +} + static int ena_io_poll(struct napi_struct *napi, int budget) { struct ena_napi *ena_napi = container_of(napi, struct ena_napi, napi); struct ena_ring *tx_ring, *rx_ring; - int tx_work_done; int rx_work_done = 0; int tx_budget; @@ -1247,6 +1867,9 @@ static int ena_io_poll(struct napi_struct *napi, int budget) tx_ring = ena_napi->tx_ring; rx_ring = ena_napi->rx_ring; + tx_ring->first_interrupt = ena_napi->first_interrupt; + rx_ring->first_interrupt = ena_napi->first_interrupt; + tx_budget = tx_ring->ring_size / ENA_TX_POLL_BUDGET_DIVIDER; if (!test_bit(ENA_FLAG_DEV_UP, &tx_ring->adapter->flags) || @@ -1322,8 +1945,7 @@ static irqreturn_t ena_intr_msix_io(int irq, void *data) { struct ena_napi *ena_napi = data; - ena_napi->tx_ring->first_interrupt = true; - ena_napi->rx_ring->first_interrupt = true; + ena_napi->first_interrupt = true; napi_schedule_irqoff(&ena_napi->napi); @@ -1398,10 +2020,12 @@ static void ena_setup_io_intr(struct ena_adapter *adapter) { struct net_device *netdev; int irq_idx, i, cpu; + int io_queue_count; netdev = adapter->netdev; + io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = 0; i < io_queue_count; i++) { irq_idx = ENA_IO_IRQ_IDX(i); cpu = i % num_online_cpus(); @@ -1529,45 +2153,64 @@ static void ena_disable_io_intr_sync(struct ena_adapter *adapter) synchronize_irq(adapter->irq_tbl[i].vector); } -static void ena_del_napi(struct ena_adapter *adapter) +static void ena_del_napi_in_range(struct ena_adapter *adapter, + int first_index, + int count) { int i; - for (i = 0; i < adapter->num_io_queues; i++) - netif_napi_del(&adapter->ena_napi[i].napi); + for (i = first_index; i < first_index + count; i++) { + /* Check if napi was initialized before */ + if (!ENA_IS_XDP_INDEX(adapter, i) || + adapter->ena_napi[i].xdp_ring) + netif_napi_del(&adapter->ena_napi[i].napi); + else + WARN_ON(ENA_IS_XDP_INDEX(adapter, i) && + adapter->ena_napi[i].xdp_ring); + } } -static void ena_init_napi(struct ena_adapter *adapter) +static void ena_init_napi_in_range(struct ena_adapter *adapter, + int first_index, int count) { - struct ena_napi *napi; + struct ena_napi *napi = {0}; int i; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = first_index; i < first_index + count; i++) { napi = &adapter->ena_napi[i]; netif_napi_add(adapter->netdev, &adapter->ena_napi[i].napi, - ena_io_poll, + ENA_IS_XDP_INDEX(adapter, i) ? ena_xdp_io_poll : ena_io_poll, ENA_NAPI_BUDGET); - napi->rx_ring = &adapter->rx_ring[i]; - napi->tx_ring = &adapter->tx_ring[i]; + + if (!ENA_IS_XDP_INDEX(adapter, i)) { + napi->rx_ring = &adapter->rx_ring[i]; + napi->tx_ring = &adapter->tx_ring[i]; + } else { + napi->xdp_ring = &adapter->tx_ring[i]; + } napi->qid = i; } } -static void ena_napi_disable_all(struct ena_adapter *adapter) +static void ena_napi_disable_in_range(struct ena_adapter *adapter, + int first_index, + int count) { int i; - for (i = 0; i < adapter->num_io_queues; i++) + for (i = first_index; i < first_index + count; i++) napi_disable(&adapter->ena_napi[i].napi); } -static void ena_napi_enable_all(struct ena_adapter *adapter) +static void ena_napi_enable_in_range(struct ena_adapter *adapter, + int first_index, + int count) { int i; - for (i = 0; i < adapter->num_io_queues; i++) + for (i = first_index; i < first_index + count; i++) napi_enable(&adapter->ena_napi[i].napi); } @@ -1582,7 +2225,7 @@ static int ena_rss_configure(struct ena_adapter *adapter) rc = ena_rss_init_default(adapter); if (rc && (rc != -EOPNOTSUPP)) { netif_err(adapter, ifup, adapter->netdev, - "Failed to init RSS rc: %d\n", rc); + "Failed to init RSS rc: %d\n", rc); return rc; } } @@ -1620,7 +2263,9 @@ static int ena_up_complete(struct ena_adapter *adapter) /* enable transmits */ netif_tx_start_all_queues(adapter->netdev); - ena_napi_enable_all(adapter); + ena_napi_enable_in_range(adapter, + 0, + adapter->xdp_num_queues + adapter->num_io_queues); return 0; } @@ -1653,7 +2298,7 @@ static int ena_create_io_tx_queue(struct ena_adapter *adapter, int qid) if (rc) { netif_err(adapter, ifup, adapter->netdev, "Failed to create I/O TX queue num %d rc: %d\n", - qid, rc); + qid, rc); return rc; } @@ -1672,12 +2317,13 @@ static int ena_create_io_tx_queue(struct ena_adapter *adapter, int qid) return rc; } -static int ena_create_all_io_tx_queues(struct ena_adapter *adapter) +static int ena_create_io_tx_queues_in_range(struct ena_adapter *adapter, + int first_index, int count) { struct ena_com_dev *ena_dev = adapter->ena_dev; int rc, i; - for (i = 0; i < adapter->num_io_queues; i++) { + for (i = first_index; i < first_index + count; i++) { rc = ena_create_io_tx_queue(adapter, i); if (rc) goto create_err; @@ -1686,7 +2332,7 @@ static int ena_create_all_io_tx_queues(struct ena_adapter *adapter) return 0; create_err: - while (i--) + while (i-- > first_index) ena_com_destroy_io_queue(ena_dev, ENA_IO_TXQ_IDX(i)); return rc; @@ -1731,13 +2377,15 @@ static int ena_create_io_rx_queue(struct ena_adapter *adapter, int qid) netif_err(adapter, ifup, adapter->netdev, "Failed to get RX queue handlers. RX queue num %d rc: %d\n", qid, rc); - ena_com_destroy_io_queue(ena_dev, ena_qid); - return rc; + goto err; } ena_com_update_numa_node(rx_ring->ena_com_io_cq, ctx.numa_node); return rc; +err: + ena_com_destroy_io_queue(ena_dev, ena_qid); + return rc; } static int ena_create_all_io_rx_queues(struct ena_adapter *adapter) @@ -1764,7 +2412,8 @@ create_err: } static void set_io_rings_size(struct ena_adapter *adapter, - int new_tx_size, int new_rx_size) + int new_tx_size, + int new_rx_size) { int i; @@ -1798,14 +2447,24 @@ static int create_queues_with_size_backoff(struct ena_adapter *adapter) * ones due to past queue allocation failures. */ set_io_rings_size(adapter, adapter->requested_tx_ring_size, - adapter->requested_rx_ring_size); + adapter->requested_rx_ring_size); while (1) { - rc = ena_setup_all_tx_resources(adapter); + if (ena_xdp_present(adapter)) { + rc = ena_setup_and_create_all_xdp_queues(adapter); + + if (rc) + goto err_setup_tx; + } + rc = ena_setup_tx_resources_in_range(adapter, + 0, + adapter->num_io_queues); if (rc) goto err_setup_tx; - rc = ena_create_all_io_tx_queues(adapter); + rc = ena_create_io_tx_queues_in_range(adapter, + 0, + adapter->num_io_queues); if (rc) goto err_create_tx_queues; @@ -1829,7 +2488,7 @@ err_setup_tx: if (rc != -ENOMEM) { netif_err(adapter, ifup, adapter->netdev, "Queue creation failed with error code %d\n", - rc); + rc); return rc; } @@ -1852,7 +2511,7 @@ err_setup_tx: new_rx_ring_size = cur_rx_ring_size / 2; if (new_tx_ring_size < ENA_MIN_RING_SIZE || - new_rx_ring_size < ENA_MIN_RING_SIZE) { + new_rx_ring_size < ENA_MIN_RING_SIZE) { netif_err(adapter, ifup, adapter->netdev, "Queue creation failed with the smallest possible queue size of %d for both queues. Not retrying with smaller queues\n", ENA_MIN_RING_SIZE); @@ -1871,10 +2530,11 @@ err_setup_tx: static int ena_up(struct ena_adapter *adapter) { - int rc, i; + int io_queue_count, rc, i; netdev_dbg(adapter->netdev, "%s\n", __func__); + io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues; ena_setup_io_intr(adapter); /* napi poll functions should be initialized before running @@ -1882,7 +2542,7 @@ static int ena_up(struct ena_adapter *adapter) * interrupt, causing the ISR to fire immediately while the poll * function wasn't set yet, causing a null dereference */ - ena_init_napi(adapter); + ena_init_napi_in_range(adapter, 0, io_queue_count); rc = ena_request_io_irq(adapter); if (rc) @@ -1913,7 +2573,7 @@ static int ena_up(struct ena_adapter *adapter) /* schedule napi in case we had pending packets * from the last time we disable napi */ - for (i = 0; i < adapter->num_io_queues; i++) + for (i = 0; i < io_queue_count; i++) napi_schedule(&adapter->ena_napi[i].napi); return rc; @@ -1926,13 +2586,15 @@ err_up: err_create_queues_with_backoff: ena_free_io_irq(adapter); err_req_irq: - ena_del_napi(adapter); + ena_del_napi_in_range(adapter, 0, io_queue_count); return rc; } static void ena_down(struct ena_adapter *adapter) { + int io_queue_count = adapter->num_io_queues + adapter->xdp_num_queues; + netif_info(adapter, ifdown, adapter->netdev, "%s\n", __func__); clear_bit(ENA_FLAG_DEV_UP, &adapter->flags); @@ -1945,7 +2607,7 @@ static void ena_down(struct ena_adapter *adapter) netif_tx_disable(adapter->netdev); /* After this point the napi handler won't enable the tx queue */ - ena_napi_disable_all(adapter); + ena_napi_disable_in_range(adapter, 0, io_queue_count); /* After destroy the queue there won't be any new interrupts */ @@ -1963,7 +2625,7 @@ static void ena_down(struct ena_adapter *adapter) ena_disable_io_intr_sync(adapter); ena_free_io_irq(adapter); - ena_del_napi(adapter); + ena_del_napi_in_range(adapter, 0, io_queue_count); ena_free_all_tx_bufs(adapter); ena_free_all_rx_bufs(adapter); @@ -2053,23 +2715,47 @@ int ena_update_queue_sizes(struct ena_adapter *adapter, ena_close(adapter->netdev); adapter->requested_tx_ring_size = new_tx_size; adapter->requested_rx_ring_size = new_rx_size; - ena_init_io_rings(adapter); + ena_init_io_rings(adapter, + 0, + adapter->xdp_num_queues + + adapter->num_io_queues); return dev_was_up ? ena_up(adapter) : 0; } int ena_update_queue_count(struct ena_adapter *adapter, u32 new_channel_count) { struct ena_com_dev *ena_dev = adapter->ena_dev; + int prev_channel_count; bool dev_was_up; dev_was_up = test_bit(ENA_FLAG_DEV_UP, &adapter->flags); ena_close(adapter->netdev); + prev_channel_count = adapter->num_io_queues; adapter->num_io_queues = new_channel_count; + if (ena_xdp_present(adapter) && + ena_xdp_allowed(adapter) == ENA_XDP_ALLOWED) { + adapter->xdp_first_ring = new_channel_count; + adapter->xdp_num_queues = new_channel_count; + if (prev_channel_count > new_channel_count) + ena_xdp_exchange_program_rx_in_range(adapter, + NULL, + new_channel_count, + prev_channel_count); + else + ena_xdp_exchange_program_rx_in_range(adapter, + adapter->xdp_bpf_prog, + prev_channel_count, + new_channel_count); + } + /* We need to destroy the rss table so that the indirection * table will be reinitialized by ena_up() */ ena_com_rss_destroy(ena_dev); - ena_init_io_rings(adapter); + ena_init_io_rings(adapter, + 0, + adapter->xdp_num_queues + + adapter->num_io_queues); return dev_was_up ? ena_open(adapter->netdev) : 0; } @@ -2253,7 +2939,7 @@ error_report_dma_error: tx_info->skb = NULL; tx_info->num_of_bufs += i; - ena_unmap_tx_skb(tx_ring, tx_info); + ena_unmap_tx_buff(tx_ring, tx_info); return -EINVAL; } @@ -2268,7 +2954,7 @@ static netdev_tx_t ena_start_xmit(struct sk_buff *skb, struct net_device *dev) struct netdev_queue *txq; void *push_hdr; u16 next_to_use, req_id, header_len; - int qid, rc, nb_hw_desc; + int qid, rc; netif_dbg(adapter, tx_queued, dev, "%s skb %p\n", __func__, skb); /* Determine which tx ring we will be placed on */ @@ -2303,50 +2989,17 @@ static netdev_tx_t ena_start_xmit(struct sk_buff *skb, struct net_device *dev) /* set flags and meta data */ ena_tx_csum(&ena_tx_ctx, skb); - if (unlikely(ena_com_is_doorbell_needed(tx_ring->ena_com_io_sq, &ena_tx_ctx))) { - netif_dbg(adapter, tx_queued, dev, - "llq tx max burst size of queue %d achieved, writing doorbell to send burst\n", - qid); - ena_com_write_sq_doorbell(tx_ring->ena_com_io_sq); - } - - /* prepare the packet's descriptors to dma engine */ - rc = ena_com_prepare_tx(tx_ring->ena_com_io_sq, &ena_tx_ctx, - &nb_hw_desc); - - /* ena_com_prepare_tx() can't fail due to overflow of tx queue, - * since the number of free descriptors in the queue is checked - * after sending the previous packet. In case there isn't enough - * space in the queue for the next packet, it is stopped - * until there is again enough available space in the queue. - * All other failure reasons of ena_com_prepare_tx() are fatal - * and therefore require a device reset. - */ - if (unlikely(rc)) { - netif_err(adapter, tx_queued, dev, - "failed to prepare tx bufs\n"); - u64_stats_update_begin(&tx_ring->syncp); - tx_ring->tx_stats.prepare_ctx_err++; - u64_stats_update_end(&tx_ring->syncp); - adapter->reset_reason = ENA_REGS_RESET_DRIVER_INVALID_STATE; - set_bit(ENA_FLAG_TRIGGER_RESET, &adapter->flags); + rc = ena_xmit_common(dev, + tx_ring, + tx_info, + &ena_tx_ctx, + next_to_use, + skb->len); + if (rc) goto error_unmap_dma; - } netdev_tx_sent_queue(txq, skb->len); - u64_stats_update_begin(&tx_ring->syncp); - tx_ring->tx_stats.cnt++; - tx_ring->tx_stats.bytes += skb->len; - u64_stats_update_end(&tx_ring->syncp); - - tx_info->tx_descs = nb_hw_desc; - tx_info->last_jiffies = jiffies; - tx_info->print_once = 0; - - tx_ring->next_to_use = ENA_TX_RING_IDX_NEXT(next_to_use, - tx_ring->ring_size); - /* stop the queue when no more space available, the packet can have up * to sgl_size + 2. one for the meta descriptor and one for header * (if the header is larger than tx_max_header_size). @@ -2393,7 +3046,7 @@ static netdev_tx_t ena_start_xmit(struct sk_buff *skb, struct net_device *dev) return NETDEV_TX_OK; error_unmap_dma: - ena_unmap_tx_skb(tx_ring, tx_info); + ena_unmap_tx_buff(tx_ring, tx_info); tx_info->skb = NULL; error_drop_packet: @@ -2572,6 +3225,7 @@ static const struct net_device_ops ena_netdev_ops = { .ndo_change_mtu = ena_change_mtu, .ndo_set_mac_address = NULL, .ndo_validate_addr = eth_validate_addr, + .ndo_bpf = ena_xdp, }; static int ena_device_validate_params(struct ena_adapter *adapter, @@ -2951,7 +3605,9 @@ static void check_for_missing_completions(struct ena_adapter *adapter) struct ena_ring *tx_ring; struct ena_ring *rx_ring; int i, budget, rc; + int io_queue_count; + io_queue_count = adapter->xdp_num_queues + adapter->num_io_queues; /* Make sure the driver doesn't turn the device in other process */ smp_rmb(); @@ -2966,7 +3622,7 @@ static void check_for_missing_completions(struct ena_adapter *adapter) budget = ENA_MONITORED_TX_QUEUES; - for (i = adapter->last_monitored_tx_qid; i < adapter->num_io_queues; i++) { + for (i = adapter->last_monitored_tx_qid; i < io_queue_count; i++) { tx_ring = &adapter->tx_ring[i]; rx_ring = &adapter->rx_ring[i]; @@ -2974,7 +3630,8 @@ static void check_for_missing_completions(struct ena_adapter *adapter) if (unlikely(rc)) return; - rc = check_for_rx_interrupt_queue(adapter, rx_ring); + rc = !ENA_IS_XDP_INDEX(adapter, i) ? + check_for_rx_interrupt_queue(adapter, rx_ring) : 0; if (unlikely(rc)) return; @@ -2983,7 +3640,7 @@ static void check_for_missing_completions(struct ena_adapter *adapter) break; } - adapter->last_monitored_tx_qid = i % adapter->num_io_queues; + adapter->last_monitored_tx_qid = i % io_queue_count; } /* trigger napi schedule after 2 consecutive detections */ @@ -3560,6 +4217,9 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) adapter->num_io_queues = max_num_io_queues; adapter->max_num_io_queues = max_num_io_queues; + adapter->xdp_first_ring = 0; + adapter->xdp_num_queues = 0; + adapter->last_monitored_tx_qid = 0; adapter->rx_copybreak = ENA_DEFAULT_RX_COPYBREAK; @@ -3573,7 +4233,10 @@ static int ena_probe(struct pci_dev *pdev, const struct pci_device_id *ent) "Failed to query interrupt moderation feature\n"); goto err_netdev_destroy; } - ena_init_io_rings(adapter); + ena_init_io_rings(adapter, + 0, + adapter->xdp_num_queues + + adapter->num_io_queues); netdev->netdev_ops = &ena_netdev_ops; netdev->watchdog_timeo = TX_TIMEOUT; diff --git a/drivers/net/ethernet/amazon/ena/ena_netdev.h b/drivers/net/ethernet/amazon/ena/ena_netdev.h index bffd778f2ce3..094324fd0edc 100644 --- a/drivers/net/ethernet/amazon/ena/ena_netdev.h +++ b/drivers/net/ethernet/amazon/ena/ena_netdev.h @@ -36,6 +36,7 @@ #include <linux/bitops.h> #include <linux/dim.h> #include <linux/etherdevice.h> +#include <linux/if_vlan.h> #include <linux/inetdevice.h> #include <linux/interrupt.h> #include <linux/netdevice.h> @@ -142,6 +143,18 @@ #define ENA_MMIO_DISABLE_REG_READ BIT(0) +/* The max MTU size is configured to be the ethernet frame size without + * the overhead of the ethernet header, which can have a VLAN header, and + * a frame check sequence (FCS). + * The buffer size we share with the device is defined to be ENA_PAGE_SIZE + */ + +#define ENA_XDP_MAX_MTU (ENA_PAGE_SIZE - ETH_HLEN - ETH_FCS_LEN - \ + VLAN_HLEN - XDP_PACKET_HEADROOM) + +#define ENA_IS_XDP_INDEX(adapter, index) (((index) >= (adapter)->xdp_first_ring) && \ + ((index) < (adapter)->xdp_first_ring + (adapter)->xdp_num_queues)) + struct ena_irq { irq_handler_t handler; void *data; @@ -155,6 +168,8 @@ struct ena_napi { struct napi_struct napi ____cacheline_aligned; struct ena_ring *tx_ring; struct ena_ring *rx_ring; + struct ena_ring *xdp_ring; + bool first_interrupt; u32 qid; struct dim dim; }; @@ -180,6 +195,17 @@ struct ena_tx_buffer { /* num of buffers used by this skb */ u32 num_of_bufs; + /* XDP buffer structure which is used for sending packets in + * the xdp queues + */ + struct xdp_frame *xdpf; + /* The rx page for the rx buffer that was received in rx and + * re transmitted on xdp tx queues as a result of XDP_TX action. + * We need to free the page once we finished cleaning the buffer in + * clean_xdp_irq() + */ + struct page *xdp_rx_page; + /* Indicate if bufs[0] map the linear data of the skb. */ u8 map_linear_data; @@ -258,10 +284,13 @@ struct ena_ring { struct ena_adapter *adapter; struct ena_com_io_cq *ena_com_io_cq; struct ena_com_io_sq *ena_com_io_sq; + struct bpf_prog *xdp_bpf_prog; + struct xdp_rxq_info xdp_rxq; u16 next_to_use; u16 next_to_clean; u16 rx_copybreak; + u16 rx_headroom; u16 qid; u16 mtu; u16 sgl_size; @@ -379,6 +408,10 @@ struct ena_adapter { u32 last_monitored_tx_qid; enum ena_regs_reset_reason_types reset_reason; + + struct bpf_prog *xdp_bpf_prog; + u32 xdp_first_ring; + u32 xdp_num_queues; }; void ena_set_ethtool_ops(struct net_device *netdev); @@ -390,8 +423,48 @@ void ena_dump_stats_to_buf(struct ena_adapter *adapter, u8 *buf); int ena_update_queue_sizes(struct ena_adapter *adapter, u32 new_tx_size, u32 new_rx_size); + int ena_update_queue_count(struct ena_adapter *adapter, u32 new_channel_count); int ena_get_sset_count(struct net_device *netdev, int sset); +enum ena_xdp_errors_t { + ENA_XDP_ALLOWED = 0, + ENA_XDP_CURRENT_MTU_TOO_LARGE, + ENA_XDP_NO_ENOUGH_QUEUES, +}; + +static inline bool ena_xdp_queues_present(struct ena_adapter *adapter) +{ + return adapter->xdp_first_ring != 0; +} + +static inline bool ena_xdp_present(struct ena_adapter *adapter) +{ + return !!adapter->xdp_bpf_prog; +} + +static inline bool ena_xdp_present_ring(struct ena_ring *ring) +{ + return !!ring->xdp_bpf_prog; +} + +static inline int ena_xdp_legal_queue_count(struct ena_adapter *adapter, + u32 queues) +{ + return 2 * queues <= adapter->max_num_io_queues; +} + +static inline enum ena_xdp_errors_t ena_xdp_allowed(struct ena_adapter *adapter) +{ + enum ena_xdp_errors_t rc = ENA_XDP_ALLOWED; + + if (adapter->netdev->mtu > ENA_XDP_MAX_MTU) + rc = ENA_XDP_CURRENT_MTU_TOO_LARGE; + else if (!ena_xdp_legal_queue_count(adapter, adapter->num_io_queues)) + rc = ENA_XDP_NO_ENOUGH_QUEUES; + + return rc; +} + #endif /* !(ENA_H) */ diff --git a/drivers/net/ethernet/amd/7990.c b/drivers/net/ethernet/amd/7990.c index ab30761003da..cf3562e82ca9 100644 --- a/drivers/net/ethernet/amd/7990.c +++ b/drivers/net/ethernet/amd/7990.c @@ -527,7 +527,7 @@ int lance_close(struct net_device *dev) } EXPORT_SYMBOL_GPL(lance_close); -void lance_tx_timeout(struct net_device *dev) +void lance_tx_timeout(struct net_device *dev, unsigned int txqueue) { printk("lance_tx_timeout\n"); lance_reset(dev); diff --git a/drivers/net/ethernet/amd/7990.h b/drivers/net/ethernet/amd/7990.h index 741cdc392c6b..8266b3c1fefc 100644 --- a/drivers/net/ethernet/amd/7990.h +++ b/drivers/net/ethernet/amd/7990.h @@ -243,7 +243,7 @@ int lance_open(struct net_device *dev); int lance_close(struct net_device *dev); int lance_start_xmit(struct sk_buff *skb, struct net_device *dev); void lance_set_multicast(struct net_device *dev); -void lance_tx_timeout(struct net_device *dev); +void lance_tx_timeout(struct net_device *dev, unsigned int txqueue); #ifdef CONFIG_NET_POLL_CONTROLLER void lance_poll(struct net_device *dev); #endif diff --git a/drivers/net/ethernet/amd/a2065.c b/drivers/net/ethernet/amd/a2065.c index 212fe72a190b..2f808dbc8b0e 100644 --- a/drivers/net/ethernet/amd/a2065.c +++ b/drivers/net/ethernet/amd/a2065.c @@ -118,10 +118,6 @@ struct lance_private { int auto_select; /* cable-selection by carrier */ unsigned short busmaster_regval; -#ifdef CONFIG_SUNLANCE - struct Linux_SBus_DMA *ledma; /* if set this points to ledma and arch=4m */ - int burst_sizes; /* ledma SBus burst sizes */ -#endif struct timer_list multicast_timer; struct net_device *dev; }; @@ -522,7 +518,7 @@ static inline int lance_reset(struct net_device *dev) return status; } -static void lance_tx_timeout(struct net_device *dev) +static void lance_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct lance_private *lp = netdev_priv(dev); volatile struct lance_regs *ll = lp->ll; @@ -551,11 +547,10 @@ static netdev_tx_t lance_start_xmit(struct sk_buff *skb, if (!lance_tx_buffs_avail(lp)) goto out_free; -#ifdef DEBUG /* dump the packet */ - print_hex_dump(KERN_DEBUG, "skb->data: ", DUMP_PREFIX_NONE, - 16, 1, skb->data, 64, true); -#endif + print_hex_dump_debug("skb->data: ", DUMP_PREFIX_NONE, 16, 1, skb->data, + 64, true); + entry = lp->tx_new & lp->tx_ring_mod_mask; ib->btx_ring[entry].length = (-skblen) | 0xf000; ib->btx_ring[entry].misc = 0; diff --git a/drivers/net/ethernet/amd/am79c961a.c b/drivers/net/ethernet/amd/am79c961a.c index 0842da492a64..1c53408f5d47 100644 --- a/drivers/net/ethernet/amd/am79c961a.c +++ b/drivers/net/ethernet/amd/am79c961a.c @@ -422,7 +422,7 @@ static void am79c961_setmulticastlist (struct net_device *dev) spin_unlock_irqrestore(&priv->chip_lock, flags); } -static void am79c961_timeout(struct net_device *dev) +static void am79c961_timeout(struct net_device *dev, unsigned int txqueue) { printk(KERN_WARNING "%s: transmit timed out, network cable problem?\n", dev->name); diff --git a/drivers/net/ethernet/amd/amd8111e.c b/drivers/net/ethernet/amd/amd8111e.c index 573e88fc8ede..0f3b743425e8 100644 --- a/drivers/net/ethernet/amd/amd8111e.c +++ b/drivers/net/ethernet/amd/amd8111e.c @@ -1569,7 +1569,7 @@ static int amd8111e_enable_link_change(struct amd8111e_priv *lp) * failed or the interface is locked up. This function will reinitialize * the hardware. */ -static void amd8111e_tx_timeout(struct net_device *dev) +static void amd8111e_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct amd8111e_priv *lp = netdev_priv(dev); int err; diff --git a/drivers/net/ethernet/amd/ariadne.c b/drivers/net/ethernet/amd/ariadne.c index 4b6a5cb85dd2..5e0f645f5bde 100644 --- a/drivers/net/ethernet/amd/ariadne.c +++ b/drivers/net/ethernet/amd/ariadne.c @@ -530,7 +530,7 @@ static inline void ariadne_reset(struct net_device *dev) netif_start_queue(dev); } -static void ariadne_tx_timeout(struct net_device *dev) +static void ariadne_tx_timeout(struct net_device *dev, unsigned int txqueue) { volatile struct Am79C960 *lance = (struct Am79C960 *)dev->base_addr; diff --git a/drivers/net/ethernet/amd/atarilance.c b/drivers/net/ethernet/amd/atarilance.c index d3d44e07afbc..4e36122609a3 100644 --- a/drivers/net/ethernet/amd/atarilance.c +++ b/drivers/net/ethernet/amd/atarilance.c @@ -346,7 +346,7 @@ static int lance_rx( struct net_device *dev ); static int lance_close( struct net_device *dev ); static void set_multicast_list( struct net_device *dev ); static int lance_set_mac_address( struct net_device *dev, void *addr ); -static void lance_tx_timeout (struct net_device *dev); +static void lance_tx_timeout (struct net_device *dev, unsigned int txqueue); /************************* End of Prototypes **************************/ @@ -727,7 +727,7 @@ static void lance_init_ring( struct net_device *dev ) /* XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX */ -static void lance_tx_timeout (struct net_device *dev) +static void lance_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct lance_private *lp = netdev_priv(dev); struct lance_ioreg *IO = lp->iobase; diff --git a/drivers/net/ethernet/amd/au1000_eth.c b/drivers/net/ethernet/amd/au1000_eth.c index 307e402db8c9..089a4fbc61a0 100644 --- a/drivers/net/ethernet/amd/au1000_eth.c +++ b/drivers/net/ethernet/amd/au1000_eth.c @@ -1014,7 +1014,7 @@ static netdev_tx_t au1000_tx(struct sk_buff *skb, struct net_device *dev) * The Tx ring has been full longer than the watchdog timeout * value. The transmitter must be hung? */ -static void au1000_tx_timeout(struct net_device *dev) +static void au1000_tx_timeout(struct net_device *dev, unsigned int txqueue) { netdev_err(dev, "au1000_tx_timeout: dev=%p\n", dev); au1000_reset_mac(dev); @@ -1053,23 +1053,12 @@ static void au1000_multicast_list(struct net_device *dev) writel(reg, &aup->mac->control); } -static int au1000_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - - if (!dev->phydev) - return -EINVAL; /* PHY not controllable */ - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static const struct net_device_ops au1000_netdev_ops = { .ndo_open = au1000_open, .ndo_stop = au1000_close, .ndo_start_xmit = au1000_tx, .ndo_set_rx_mode = au1000_multicast_list, - .ndo_do_ioctl = au1000_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_tx_timeout = au1000_tx_timeout, .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, diff --git a/drivers/net/ethernet/amd/declance.c b/drivers/net/ethernet/amd/declance.c index dac4a2fcad6a..7282ce55ffb8 100644 --- a/drivers/net/ethernet/amd/declance.c +++ b/drivers/net/ethernet/amd/declance.c @@ -608,7 +608,7 @@ static int lance_rx(struct net_device *dev) len = (*rds_ptr(rd, mblength, lp->type) & 0xfff) - 4; skb = netdev_alloc_skb(dev, len + 2); - if (skb == 0) { + if (!skb) { dev->stats.rx_dropped++; *rds_ptr(rd, mblength, lp->type) = 0; *rds_ptr(rd, rmd1, lp->type) = @@ -884,7 +884,7 @@ static inline int lance_reset(struct net_device *dev) return status; } -static void lance_tx_timeout(struct net_device *dev) +static void lance_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct lance_private *lp = netdev_priv(dev); volatile struct lance_regs *ll = lp->ll; diff --git a/drivers/net/ethernet/amd/lance.c b/drivers/net/ethernet/amd/lance.c index f90b454b1642..aff44241988c 100644 --- a/drivers/net/ethernet/amd/lance.c +++ b/drivers/net/ethernet/amd/lance.c @@ -306,7 +306,7 @@ static irqreturn_t lance_interrupt(int irq, void *dev_id); static int lance_close(struct net_device *dev); static struct net_device_stats *lance_get_stats(struct net_device *dev); static void set_multicast_list(struct net_device *dev); -static void lance_tx_timeout (struct net_device *dev); +static void lance_tx_timeout (struct net_device *dev, unsigned int txqueue); @@ -913,7 +913,7 @@ lance_restart(struct net_device *dev, unsigned int csr0_bits, int must_reinit) } -static void lance_tx_timeout (struct net_device *dev) +static void lance_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct lance_private *lp = (struct lance_private *) dev->ml_priv; int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/amd/ni65.c b/drivers/net/ethernet/amd/ni65.c index c6c2a54c1121..c38edf6f03a3 100644 --- a/drivers/net/ethernet/amd/ni65.c +++ b/drivers/net/ethernet/amd/ni65.c @@ -254,7 +254,7 @@ static int ni65_lance_reinit(struct net_device *dev); static void ni65_init_lance(struct priv *p,unsigned char*,int,int); static netdev_tx_t ni65_send_packet(struct sk_buff *skb, struct net_device *dev); -static void ni65_timeout(struct net_device *dev); +static void ni65_timeout(struct net_device *dev, unsigned int txqueue); static int ni65_close(struct net_device *dev); static int ni65_alloc_buffer(struct net_device *dev); static void ni65_free_buffer(struct priv *p); @@ -1133,7 +1133,7 @@ static void ni65_recv_intr(struct net_device *dev,int csr0) * kick xmitter .. */ -static void ni65_timeout(struct net_device *dev) +static void ni65_timeout(struct net_device *dev, unsigned int txqueue) { int i; struct priv *p = dev->ml_priv; diff --git a/drivers/net/ethernet/amd/nmclan_cs.c b/drivers/net/ethernet/amd/nmclan_cs.c index 9c152d85840d..023aecf6ab30 100644 --- a/drivers/net/ethernet/amd/nmclan_cs.c +++ b/drivers/net/ethernet/amd/nmclan_cs.c @@ -407,7 +407,7 @@ static int mace_open(struct net_device *dev); static int mace_close(struct net_device *dev); static netdev_tx_t mace_start_xmit(struct sk_buff *skb, struct net_device *dev); -static void mace_tx_timeout(struct net_device *dev); +static void mace_tx_timeout(struct net_device *dev, unsigned int txqueue); static irqreturn_t mace_interrupt(int irq, void *dev_id); static struct net_device_stats *mace_get_stats(struct net_device *dev); static int mace_rx(struct net_device *dev, unsigned char RxCnt); @@ -837,7 +837,7 @@ mace_start_xmit failed, put skb back into a list." ---------------------------------------------------------------------------- */ -static void mace_tx_timeout(struct net_device *dev) +static void mace_tx_timeout(struct net_device *dev, unsigned int txqueue) { mace_private *lp = netdev_priv(dev); struct pcmcia_device *link = lp->p_dev; diff --git a/drivers/net/ethernet/amd/pcnet32.c b/drivers/net/ethernet/amd/pcnet32.c index f5ad12c10934..dc7d88227e76 100644 --- a/drivers/net/ethernet/amd/pcnet32.c +++ b/drivers/net/ethernet/amd/pcnet32.c @@ -314,7 +314,7 @@ static int pcnet32_open(struct net_device *); static int pcnet32_init_ring(struct net_device *); static netdev_tx_t pcnet32_start_xmit(struct sk_buff *, struct net_device *); -static void pcnet32_tx_timeout(struct net_device *dev); +static void pcnet32_tx_timeout(struct net_device *dev, unsigned int txqueue); static irqreturn_t pcnet32_interrupt(int, void *); static int pcnet32_close(struct net_device *); static struct net_device_stats *pcnet32_get_stats(struct net_device *); @@ -2455,7 +2455,7 @@ static void pcnet32_restart(struct net_device *dev, unsigned int csr0_bits) lp->a->write_csr(ioaddr, CSR0, csr0_bits); } -static void pcnet32_tx_timeout(struct net_device *dev) +static void pcnet32_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct pcnet32_private *lp = netdev_priv(dev); unsigned long ioaddr = dev->base_addr, flags; diff --git a/drivers/net/ethernet/amd/sunlance.c b/drivers/net/ethernet/amd/sunlance.c index ebcbf8ca4829..b00e00881253 100644 --- a/drivers/net/ethernet/amd/sunlance.c +++ b/drivers/net/ethernet/amd/sunlance.c @@ -1097,7 +1097,7 @@ static void lance_piozero(void __iomem *dest, int len) sbus_writeb(0, piobuf); } -static void lance_tx_timeout(struct net_device *dev) +static void lance_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct lance_private *lp = netdev_priv(dev); diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c index 98f8f2033154..b71f9b04a51e 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c @@ -2152,7 +2152,7 @@ static int xgbe_change_mtu(struct net_device *netdev, int mtu) return 0; } -static void xgbe_tx_timeout(struct net_device *netdev) +static void xgbe_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct xgbe_prv_data *pdata = netdev_priv(netdev); diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c index 128cd648ba99..46c3c1ca38d6 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-phy-v2.c @@ -1227,7 +1227,7 @@ static bool xgbe_phy_sfp_verify_eeprom(u8 cc_in, u8 *buf, unsigned int len) for (cc = 0; len; buf++, len--) cc += *buf; - return (cc == cc_in) ? true : false; + return cc == cc_in; } static int xgbe_phy_sfp_read_eeprom(struct xgbe_prv_data *pdata) diff --git a/drivers/net/ethernet/apm/xgene-v2/main.c b/drivers/net/ethernet/apm/xgene-v2/main.c index 02b4f3af02b5..c48f60996761 100644 --- a/drivers/net/ethernet/apm/xgene-v2/main.c +++ b/drivers/net/ethernet/apm/xgene-v2/main.c @@ -575,7 +575,7 @@ static void xge_free_pending_skb(struct net_device *ndev) } } -static void xge_timeout(struct net_device *ndev) +static void xge_timeout(struct net_device *ndev, unsigned int txqueue) { struct xge_pdata *pdata = netdev_priv(ndev); diff --git a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c index d8612131c55e..6aee2f0fc0db 100644 --- a/drivers/net/ethernet/apm/xgene/xgene_enet_main.c +++ b/drivers/net/ethernet/apm/xgene/xgene_enet_main.c @@ -859,7 +859,7 @@ static int xgene_enet_napi(struct napi_struct *napi, const int budget) return processed; } -static void xgene_enet_timeout(struct net_device *ndev) +static void xgene_enet_timeout(struct net_device *ndev, unsigned int txqueue) { struct xgene_enet_pdata *pdata = netdev_priv(ndev); struct netdev_queue *txq; @@ -2020,7 +2020,7 @@ static int xgene_enet_probe(struct platform_device *pdev) int ret; ndev = alloc_etherdev_mqs(sizeof(struct xgene_enet_pdata), - XGENE_NUM_RX_RING, XGENE_NUM_TX_RING); + XGENE_NUM_TX_RING, XGENE_NUM_RX_RING); if (!ndev) return -ENOMEM; diff --git a/drivers/net/ethernet/apple/macmace.c b/drivers/net/ethernet/apple/macmace.c index 8d03578d5e8c..95d3061c61be 100644 --- a/drivers/net/ethernet/apple/macmace.c +++ b/drivers/net/ethernet/apple/macmace.c @@ -91,7 +91,7 @@ static int mace_set_address(struct net_device *dev, void *addr); static void mace_reset(struct net_device *dev); static irqreturn_t mace_interrupt(int irq, void *dev_id); static irqreturn_t mace_dma_intr(int irq, void *dev_id); -static void mace_tx_timeout(struct net_device *dev); +static void mace_tx_timeout(struct net_device *dev, unsigned int txqueue); static void __mace_set_address(struct net_device *dev, void *addr); /* @@ -600,7 +600,7 @@ static irqreturn_t mace_interrupt(int irq, void *dev_id) return IRQ_HANDLED; } -static void mace_tx_timeout(struct net_device *dev) +static void mace_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mace_data *mp = netdev_priv(dev); volatile struct mace *mb = mp->mace; diff --git a/drivers/net/ethernet/arc/emac_main.c b/drivers/net/ethernet/arc/emac_main.c index 6f2c867785fe..17bda4e8cc45 100644 --- a/drivers/net/ethernet/arc/emac_main.c +++ b/drivers/net/ethernet/arc/emac_main.c @@ -781,18 +781,6 @@ static int arc_emac_set_address(struct net_device *ndev, void *p) return 0; } -static int arc_emac_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - - if (!dev->phydev) - return -ENODEV; - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - - /** * arc_emac_restart - Restart EMAC * @ndev: Pointer to net_device structure. @@ -857,7 +845,7 @@ static const struct net_device_ops arc_emac_netdev_ops = { .ndo_set_mac_address = arc_emac_set_address, .ndo_get_stats = arc_emac_stats, .ndo_set_rx_mode = arc_emac_set_rx_mode, - .ndo_do_ioctl = arc_emac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = arc_emac_poll_controller, #endif diff --git a/drivers/net/ethernet/atheros/ag71xx.c b/drivers/net/ethernet/atheros/ag71xx.c index 60ba69db48c6..e95687a780fb 100644 --- a/drivers/net/ethernet/atheros/ag71xx.c +++ b/drivers/net/ethernet/atheros/ag71xx.c @@ -1394,14 +1394,6 @@ err_drop: return NETDEV_TX_OK; } -static int ag71xx_do_ioctl(struct net_device *ndev, struct ifreq *ifr, int cmd) -{ - if (!ndev->phydev) - return -EINVAL; - - return phy_mii_ioctl(ndev->phydev, ifr, cmd); -} - static void ag71xx_oom_timer_handler(struct timer_list *t) { struct ag71xx *ag = from_timer(ag, t, oom_timer); @@ -1409,7 +1401,7 @@ static void ag71xx_oom_timer_handler(struct timer_list *t) napi_schedule(&ag->napi); } -static void ag71xx_tx_timeout(struct net_device *ndev) +static void ag71xx_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct ag71xx *ag = netdev_priv(ndev); @@ -1618,7 +1610,7 @@ static const struct net_device_ops ag71xx_netdev_ops = { .ndo_open = ag71xx_open, .ndo_stop = ag71xx_stop, .ndo_start_xmit = ag71xx_hard_start_xmit, - .ndo_do_ioctl = ag71xx_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = ag71xx_tx_timeout, .ndo_change_mtu = ag71xx_change_mtu, .ndo_set_mac_address = eth_mac_addr, diff --git a/drivers/net/ethernet/atheros/alx/main.c b/drivers/net/ethernet/atheros/alx/main.c index d4bbcdfd691a..1dcbc486eca9 100644 --- a/drivers/net/ethernet/atheros/alx/main.c +++ b/drivers/net/ethernet/atheros/alx/main.c @@ -1553,7 +1553,7 @@ static netdev_tx_t alx_start_xmit(struct sk_buff *skb, return alx_start_xmit_ring(skb, alx_tx_queue_mapping(alx, skb)); } -static void alx_tx_timeout(struct net_device *dev) +static void alx_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct alx_priv *alx = netdev_priv(dev); diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c index 2b239ecea05f..4c0b1f8551dd 100644 --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c @@ -350,7 +350,7 @@ static void atl1c_del_timer(struct atl1c_adapter *adapter) * atl1c_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure */ -static void atl1c_tx_timeout(struct net_device *netdev) +static void atl1c_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct atl1c_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c index 4f7b65825c15..e0d89942d537 100644 --- a/drivers/net/ethernet/atheros/atl1e/atl1e_main.c +++ b/drivers/net/ethernet/atheros/atl1e/atl1e_main.c @@ -251,7 +251,7 @@ static void atl1e_cancel_work(struct atl1e_adapter *adapter) * atl1e_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure */ -static void atl1e_tx_timeout(struct net_device *netdev) +static void atl1e_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct atl1e_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/atheros/atlx/atl2.c b/drivers/net/ethernet/atheros/atlx/atl2.c index 3aba38322717..b81a4e0c5b57 100644 --- a/drivers/net/ethernet/atheros/atlx/atl2.c +++ b/drivers/net/ethernet/atheros/atlx/atl2.c @@ -1001,7 +1001,7 @@ static int atl2_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) * atl2_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure */ -static void atl2_tx_timeout(struct net_device *netdev) +static void atl2_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct atl2_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/atheros/atlx/atlx.c b/drivers/net/ethernet/atheros/atlx/atlx.c index 505a22c703f7..0941d07d0833 100644 --- a/drivers/net/ethernet/atheros/atlx/atlx.c +++ b/drivers/net/ethernet/atheros/atlx/atlx.c @@ -183,7 +183,7 @@ static void atlx_clear_phy_int(struct atlx_adapter *adapter) * atlx_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure */ -static void atlx_tx_timeout(struct net_device *netdev) +static void atlx_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct atlx_adapter *adapter = netdev_priv(netdev); /* Do the reset outside of interrupt context */ diff --git a/drivers/net/ethernet/aurora/nb8800.c b/drivers/net/ethernet/aurora/nb8800.c index 30b455013bf3..bc273e0db7ff 100644 --- a/drivers/net/ethernet/aurora/nb8800.c +++ b/drivers/net/ethernet/aurora/nb8800.c @@ -1005,18 +1005,13 @@ static int nb8800_stop(struct net_device *dev) return 0; } -static int nb8800_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static const struct net_device_ops nb8800_netdev_ops = { .ndo_open = nb8800_open, .ndo_stop = nb8800_stop, .ndo_start_xmit = nb8800_xmit, .ndo_set_mac_address = nb8800_set_mac_address, .ndo_set_rx_mode = nb8800_set_rx_mode, - .ndo_do_ioctl = nb8800_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_validate_addr = eth_validate_addr, }; diff --git a/drivers/net/ethernet/broadcom/b44.c b/drivers/net/ethernet/broadcom/b44.c index ec25fd81985d..a780b7215021 100644 --- a/drivers/net/ethernet/broadcom/b44.c +++ b/drivers/net/ethernet/broadcom/b44.c @@ -948,7 +948,7 @@ irq_ack: return IRQ_RETVAL(handled); } -static void b44_tx_timeout(struct net_device *dev) +static void b44_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct b44 *bp = netdev_priv(dev); diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c index d6b1a153f9df..f07ac0e0af59 100644 --- a/drivers/net/ethernet/broadcom/bcmsysport.c +++ b/drivers/net/ethernet/broadcom/bcmsysport.c @@ -1354,7 +1354,7 @@ out: return ret; } -static void bcm_sysport_tx_timeout(struct net_device *dev) +static void bcm_sysport_tx_timeout(struct net_device *dev, unsigned int txqueue) { netdev_warn(dev, "transmit timeout!\n"); @@ -2428,6 +2428,14 @@ static int bcm_sysport_probe(struct platform_device *pdev) if (!of_id || !of_id->data) return -EINVAL; + ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(40)); + if (ret) + ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); + if (ret) { + dev_err(&pdev->dev, "unable to set DMA mask: %d\n", ret); + return ret; + } + /* Fairly quickly we need to know the type of adapter we have */ params = of_id->data; diff --git a/drivers/net/ethernet/broadcom/bgmac.c b/drivers/net/ethernet/broadcom/bgmac.c index 148734b166f0..1bb07a5d82c9 100644 --- a/drivers/net/ethernet/broadcom/bgmac.c +++ b/drivers/net/ethernet/broadcom/bgmac.c @@ -1248,14 +1248,6 @@ static int bgmac_set_mac_address(struct net_device *net_dev, void *addr) return 0; } -static int bgmac_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd) -{ - if (!netif_running(net_dev)) - return -EINVAL; - - return phy_mii_ioctl(net_dev->phydev, ifr, cmd); -} - static const struct net_device_ops bgmac_netdev_ops = { .ndo_open = bgmac_open, .ndo_stop = bgmac_stop, @@ -1263,7 +1255,7 @@ static const struct net_device_ops bgmac_netdev_ops = { .ndo_set_rx_mode = bgmac_set_rx_mode, .ndo_set_mac_address = bgmac_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = bgmac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, }; /************************************************** diff --git a/drivers/net/ethernet/broadcom/bnx2.c b/drivers/net/ethernet/broadcom/bnx2.c index fbc196b480b6..dbb7874607ca 100644 --- a/drivers/net/ethernet/broadcom/bnx2.c +++ b/drivers/net/ethernet/broadcom/bnx2.c @@ -6575,7 +6575,7 @@ bnx2_dump_state(struct bnx2 *bp) } static void -bnx2_tx_timeout(struct net_device *dev) +bnx2_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct bnx2 *bp = netdev_priv(dev); diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c index 5e037a305b83..ee9e9290f112 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c @@ -4970,7 +4970,7 @@ int bnx2x_set_features(struct net_device *dev, netdev_features_t features) return 0; } -void bnx2x_tx_timeout(struct net_device *dev) +void bnx2x_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct bnx2x *bp = netdev_priv(dev); diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.h index 3f63ffd7561b..6f1352d51cb2 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.h +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.h @@ -617,7 +617,7 @@ int bnx2x_set_features(struct net_device *dev, netdev_features_t features); * * @dev: net device */ -void bnx2x_tx_timeout(struct net_device *dev); +void bnx2x_tx_timeout(struct net_device *dev, unsigned int txqueue); /** bnx2x_get_c2s_mapping - read inner-to-outer vlan configuration * c2s_map should have BNX2X_MAX_PRIORITY entries. diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index 61fa32cdd3e3..1c26fa962233 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -15410,6 +15410,7 @@ int bnx2x_configure_ptp_filters(struct bnx2x *bp) REG_WR(bp, rule, BNX2X_PTP_TX_ON_RULE_MASK); break; case HWTSTAMP_TX_ONESTEP_SYNC: + case HWTSTAMP_TX_ONESTEP_P2P: BNX2X_ERR("One-step timestamping is not supported\n"); return -ERANGE; } diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index e6f18f6070ef..483935b001c8 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -944,6 +944,7 @@ static struct sk_buff *bnxt_rx_page_skb(struct bnxt *bp, dma_addr -= bp->rx_dma_offset; dma_unmap_page_attrs(&bp->pdev->dev, dma_addr, PAGE_SIZE, bp->rx_dir, DMA_ATTR_WEAK_ORDERING); + page_pool_release_page(rxr->page_pool, page); if (unlikely(!payload)) payload = eth_get_headlen(bp->dev, data_ptr, len); @@ -6997,7 +6998,6 @@ static int __bnxt_hwrm_func_qcaps(struct bnxt *bp) pf->fw_fid = le16_to_cpu(resp->fid); pf->port_id = le16_to_cpu(resp->port_id); - bp->dev->dev_port = pf->port_id; memcpy(pf->mac_addr, resp->mac_address, ETH_ALEN); pf->first_vf_id = le16_to_cpu(resp->first_vf_id); pf->max_vfs = le16_to_cpu(resp->max_vfs); @@ -7288,6 +7288,7 @@ static int bnxt_hwrm_ver_get(struct bnxt *bp) bp->hwrm_max_ext_req_len = HWRM_MAX_REQ_LEN; bp->chip_num = le16_to_cpu(resp->chip_num); + bp->chip_rev = resp->chip_rev; if (bp->chip_num == CHIP_NUM_58700 && !resp->chip_rev && !resp->chip_metal) bp->flags |= BNXT_FLAG_CHIP_NITRO_A0; @@ -9063,7 +9064,7 @@ static int bnxt_update_phy_setting(struct bnxt *bp) /* The last close may have shutdown the link, so need to call * PHY_CFG to bring it back up. */ - if (!netif_carrier_ok(bp->dev)) + if (!bp->link_info.link_up) update_link = true; if (!bnxt_eee_config_ok(bp)) @@ -9975,7 +9976,7 @@ static void bnxt_reset_task(struct bnxt *bp, bool silent) } } -static void bnxt_tx_timeout(struct net_device *dev) +static void bnxt_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct bnxt *bp = netdev_priv(dev); @@ -10040,6 +10041,13 @@ static void bnxt_timer(struct timer_list *t) bnxt_queue_sp_work(bp); } +#ifdef CONFIG_RFS_ACCEL + if ((bp->flags & BNXT_FLAG_RFS) && bp->ntp_fltr_count) { + set_bit(BNXT_RX_NTP_FLTR_SP_EVENT, &bp->sp_event); + bnxt_queue_sp_work(bp); + } +#endif /*CONFIG_RFS_ACCEL*/ + if (bp->link_info.phy_retry) { if (time_after(jiffies, bp->link_info.phy_retry_expires)) { bp->link_info.phy_retry = false; @@ -10050,7 +10058,8 @@ static void bnxt_timer(struct timer_list *t) } } - if ((bp->flags & BNXT_FLAG_CHIP_P5) && netif_carrier_ok(dev)) { + if ((bp->flags & BNXT_FLAG_CHIP_P5) && !bp->chip_rev && + netif_carrier_ok(dev)) { set_bit(BNXT_RING_COAL_NOW_SP_EVENT, &bp->sp_event); bnxt_queue_sp_work(bp); } @@ -10568,7 +10577,7 @@ static void bnxt_set_dflt_rss_hash_type(struct bnxt *bp) VNIC_RSS_CFG_REQ_HASH_TYPE_TCP_IPV4 | VNIC_RSS_CFG_REQ_HASH_TYPE_IPV6 | VNIC_RSS_CFG_REQ_HASH_TYPE_TCP_IPV6; - if (BNXT_CHIP_P4(bp) && bp->hwrm_spec_code >= 0x10501) { + if (BNXT_CHIP_P4_PLUS(bp) && bp->hwrm_spec_code >= 0x10501) { bp->flags |= BNXT_FLAG_UDP_RSS_CAP; bp->rss_hash_cfg |= VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV4 | VNIC_RSS_CFG_REQ_HASH_TYPE_UDP_IPV6; @@ -10822,6 +10831,7 @@ static void bnxt_fw_reset_task(struct work_struct *work) smp_mb__before_atomic(); clear_bit(BNXT_STATE_IN_FW_RESET, &bp->state); bnxt_ulp_start(bp, rc); + bnxt_dl_health_recovery_done(bp); bnxt_dl_health_status_update(bp, true); rtnl_unlock(); break; @@ -11099,6 +11109,7 @@ static int bnxt_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, struct ethhdr *eth = (struct ethhdr *)skb_mac_header(skb); int rc = 0, idx, bit_id, l2_idx = 0; struct hlist_head *head; + u32 flags; if (!ether_addr_equal(dev->dev_addr, eth->h_dest)) { struct bnxt_vnic_info *vnic = &bp->vnic_info[0]; @@ -11138,8 +11149,9 @@ static int bnxt_rx_flow_steer(struct net_device *dev, const struct sk_buff *skb, rc = -EPROTONOSUPPORT; goto err_free; } - if ((fkeys->control.flags & FLOW_DIS_ENCAPSULATION) && - bp->hwrm_spec_code < 0x10601) { + flags = fkeys->control.flags; + if (((flags & FLOW_DIS_ENCAPSULATION) && + bp->hwrm_spec_code < 0x10601) || (flags & FLOW_DIS_IS_FRAGMENT)) { rc = -EPROTONOSUPPORT; goto err_free; } @@ -11376,8 +11388,8 @@ int bnxt_get_port_parent_id(struct net_device *dev, if (!BNXT_PF(bp) || !(bp->flags & BNXT_FLAG_DSN_VALID)) return -EOPNOTSUPP; - ppid->id_len = sizeof(bp->switch_id); - memcpy(ppid->id, bp->switch_id, ppid->id_len); + ppid->id_len = sizeof(bp->dsn); + memcpy(ppid->id, bp->dsn, ppid->id_len); return 0; } @@ -11433,9 +11445,9 @@ static void bnxt_remove_one(struct pci_dev *pdev) bnxt_sriov_disable(bp); bnxt_dl_fw_reporters_destroy(bp, true); - bnxt_dl_unregister(bp); pci_disable_pcie_error_reporting(pdev); unregister_netdev(dev); + bnxt_dl_unregister(bp); bnxt_shutdown_tc(bp); bnxt_cancel_sp_work(bp); bp->sp_event = 0; @@ -11469,6 +11481,9 @@ static int bnxt_probe_phy(struct bnxt *bp, bool fw_dflt) rc); return rc; } + if (!fw_dflt) + return 0; + rc = bnxt_update_link(bp, false); if (rc) { netdev_err(bp->dev, "Probe phy can't update link (rc: %x)\n", @@ -11482,9 +11497,6 @@ static int bnxt_probe_phy(struct bnxt *bp, bool fw_dflt) if (link_info->auto_link_speeds && !link_info->support_auto_speeds) link_info->support_auto_speeds = link_info->support_speeds; - if (!fw_dflt) - return 0; - bnxt_init_ethtool_link_settings(bp); return 0; } @@ -11858,7 +11870,7 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) if (BNXT_PF(bp)) { /* Read the adapter's DSN to use as the eswitch switch_id */ - bnxt_pcie_dsn_get(bp, bp->switch_id); + rc = bnxt_pcie_dsn_get(bp, bp->dsn); } /* MTU range: 60 - FW defined max */ @@ -11905,11 +11917,14 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) bnxt_init_tc(bp); } + bnxt_dl_register(bp); + rc = register_netdev(dev); if (rc) - goto init_err_cleanup_tc; + goto init_err_cleanup; - bnxt_dl_register(bp); + if (BNXT_PF(bp)) + devlink_port_type_eth_set(&bp->dl_port, bp->dev); bnxt_dl_fw_reporters_create(bp); netdev_info(dev, "%s found at mem %lx, node addr %pM\n", @@ -11919,7 +11934,8 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) return 0; -init_err_cleanup_tc: +init_err_cleanup: + bnxt_dl_unregister(bp); bnxt_shutdown_tc(bp); bnxt_clear_int_mode(bp); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.h b/drivers/net/ethernet/broadcom/bnxt/bnxt.h index f14335433a64..cabef0b4f5fb 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.h @@ -1457,6 +1457,8 @@ struct bnxt { #define CHIP_NUM_58804 0xd804 #define CHIP_NUM_58808 0xd808 + u8 chip_rev; + #define BNXT_CHIP_NUM_5730X(chip_num) \ ((chip_num) >= CHIP_NUM_57301 && \ (chip_num) <= CHIP_NUM_57304) @@ -1846,7 +1848,7 @@ struct bnxt { enum devlink_eswitch_mode eswitch_mode; struct bnxt_vf_rep **vf_reps; /* array of vf-rep ptrs */ u16 *cfa_code_map; /* cfa_code -> vf_idx map */ - u8 switch_id[8]; + u8 dsn[8]; struct bnxt_tc_info *tc_info; struct list_head tc_indr_block_list; struct notifier_block tc_netdev_nb; diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c index 3eedd4477218..eec0168330b7 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.c @@ -21,6 +21,7 @@ bnxt_dl_flash_update(struct devlink *dl, const char *filename, const char *region, struct netlink_ext_ack *extack) { struct bnxt *bp = bnxt_get_bp_from_dl(dl); + int rc; if (region) return -EOPNOTSUPP; @@ -31,7 +32,18 @@ bnxt_dl_flash_update(struct devlink *dl, const char *filename, return -EPERM; } - return bnxt_flash_package_from_file(bp->dev, filename, 0); + devlink_flash_update_begin_notify(dl); + devlink_flash_update_status_notify(dl, "Preparing to flash", region, 0, + 0); + rc = bnxt_flash_package_from_file(bp->dev, filename, 0); + if (!rc) + devlink_flash_update_status_notify(dl, "Flashing done", region, + 0, 0); + else + devlink_flash_update_status_notify(dl, "Flashing failed", + region, 0, 0); + devlink_flash_update_end_notify(dl); + return rc; } static int bnxt_fw_reporter_diagnose(struct devlink_health_reporter *reporter, @@ -89,7 +101,7 @@ static int bnxt_fw_reset_recover(struct devlink_health_reporter *reporter, return -EOPNOTSUPP; bnxt_fw_reset(bp); - return 0; + return -EINPROGRESS; } static const @@ -116,7 +128,7 @@ static int bnxt_fw_fatal_recover(struct devlink_health_reporter *reporter, else if (event == BNXT_FW_EXCEPTION_SP_EVENT) bnxt_fw_exception(bp); - return 0; + return -EINPROGRESS; } static const @@ -262,11 +274,25 @@ void bnxt_dl_health_status_update(struct bnxt *bp, bool healthy) health->fatal = false; } +void bnxt_dl_health_recovery_done(struct bnxt *bp) +{ + struct bnxt_fw_health *hlth = bp->fw_health; + + if (hlth->fatal) + devlink_health_reporter_recovery_done(hlth->fw_fatal_reporter); + else + devlink_health_reporter_recovery_done(hlth->fw_reset_reporter); +} + +static int bnxt_dl_info_get(struct devlink *dl, struct devlink_info_req *req, + struct netlink_ext_ack *extack); + static const struct devlink_ops bnxt_dl_ops = { #ifdef CONFIG_BNXT_SRIOV .eswitch_mode_set = bnxt_dl_eswitch_mode_set, .eswitch_mode_get = bnxt_dl_eswitch_mode_get, #endif /* CONFIG_BNXT_SRIOV */ + .info_get = bnxt_dl_info_get, .flash_update = bnxt_dl_flash_update, }; @@ -333,6 +359,136 @@ static void bnxt_copy_from_nvm_data(union devlink_param_value *dst, dst->vu8 = (u8)val32; } +static int bnxt_hwrm_get_nvm_cfg_ver(struct bnxt *bp, + union devlink_param_value *nvm_cfg_ver) +{ + struct hwrm_nvm_get_variable_input req = {0}; + union bnxt_nvm_data *data; + dma_addr_t data_dma_addr; + int rc; + + bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_NVM_GET_VARIABLE, -1, -1); + data = dma_alloc_coherent(&bp->pdev->dev, sizeof(*data), + &data_dma_addr, GFP_KERNEL); + if (!data) + return -ENOMEM; + + req.dest_data_addr = cpu_to_le64(data_dma_addr); + req.data_len = cpu_to_le16(BNXT_NVM_CFG_VER_BITS); + req.option_num = cpu_to_le16(NVM_OFF_NVM_CFG_VER); + + rc = hwrm_send_message_silent(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT); + if (!rc) + bnxt_copy_from_nvm_data(nvm_cfg_ver, data, + BNXT_NVM_CFG_VER_BITS, + BNXT_NVM_CFG_VER_BYTES); + + dma_free_coherent(&bp->pdev->dev, sizeof(*data), data, data_dma_addr); + return rc; +} + +static int bnxt_dl_info_get(struct devlink *dl, struct devlink_info_req *req, + struct netlink_ext_ack *extack) +{ + struct bnxt *bp = bnxt_get_bp_from_dl(dl); + union devlink_param_value nvm_cfg_ver; + struct hwrm_ver_get_output *ver_resp; + char mgmt_ver[FW_VER_STR_LEN]; + char roce_ver[FW_VER_STR_LEN]; + char fw_ver[FW_VER_STR_LEN]; + char buf[32]; + int rc; + + rc = devlink_info_driver_name_put(req, DRV_MODULE_NAME); + if (rc) + return rc; + + sprintf(buf, "%X", bp->chip_num); + rc = devlink_info_version_fixed_put(req, + DEVLINK_INFO_VERSION_GENERIC_ASIC_ID, buf); + if (rc) + return rc; + + ver_resp = &bp->ver_resp; + sprintf(buf, "%X", ver_resp->chip_rev); + rc = devlink_info_version_fixed_put(req, + DEVLINK_INFO_VERSION_GENERIC_ASIC_REV, buf); + if (rc) + return rc; + + if (BNXT_PF(bp)) { + sprintf(buf, "%02X-%02X-%02X-%02X-%02X-%02X-%02X-%02X", + bp->dsn[7], bp->dsn[6], bp->dsn[5], bp->dsn[4], + bp->dsn[3], bp->dsn[2], bp->dsn[1], bp->dsn[0]); + rc = devlink_info_serial_number_put(req, buf); + if (rc) + return rc; + } + + if (strlen(ver_resp->active_pkg_name)) { + rc = + devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW, + ver_resp->active_pkg_name); + if (rc) + return rc; + } + + if (BNXT_PF(bp) && !bnxt_hwrm_get_nvm_cfg_ver(bp, &nvm_cfg_ver)) { + u32 ver = nvm_cfg_ver.vu32; + + sprintf(buf, "%X.%X.%X", (ver >> 16) & 0xF, (ver >> 8) & 0xF, + ver & 0xF); + rc = devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_PSID, buf); + if (rc) + return rc; + } + + if (ver_resp->flags & VER_GET_RESP_FLAGS_EXT_VER_AVAIL) { + snprintf(fw_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->hwrm_fw_major, ver_resp->hwrm_fw_minor, + ver_resp->hwrm_fw_build, ver_resp->hwrm_fw_patch); + + snprintf(mgmt_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->mgmt_fw_major, ver_resp->mgmt_fw_minor, + ver_resp->mgmt_fw_build, ver_resp->mgmt_fw_patch); + + snprintf(roce_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->roce_fw_major, ver_resp->roce_fw_minor, + ver_resp->roce_fw_build, ver_resp->roce_fw_patch); + } else { + snprintf(fw_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->hwrm_fw_maj_8b, ver_resp->hwrm_fw_min_8b, + ver_resp->hwrm_fw_bld_8b, ver_resp->hwrm_fw_rsvd_8b); + + snprintf(mgmt_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->mgmt_fw_maj_8b, ver_resp->mgmt_fw_min_8b, + ver_resp->mgmt_fw_bld_8b, ver_resp->mgmt_fw_rsvd_8b); + + snprintf(roce_ver, FW_VER_STR_LEN, "%d.%d.%d.%d", + ver_resp->roce_fw_maj_8b, ver_resp->roce_fw_min_8b, + ver_resp->roce_fw_bld_8b, ver_resp->roce_fw_rsvd_8b); + } + rc = devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_APP, fw_ver); + if (rc) + return rc; + + if (!(bp->flags & BNXT_FLAG_CHIP_P5)) { + rc = devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_MGMT, mgmt_ver); + if (rc) + return rc; + + rc = devlink_info_version_running_put(req, + DEVLINK_INFO_VERSION_GENERIC_FW_ROCE, roce_ver); + if (rc) + return rc; + } + return 0; +} + static int bnxt_hwrm_nvm_req(struct bnxt *bp, u32 param_id, void *msg, int msg_len, union devlink_param_value *val) { @@ -475,15 +631,48 @@ static const struct devlink_param bnxt_dl_params[] = { static const struct devlink_param bnxt_dl_port_params[] = { }; -int bnxt_dl_register(struct bnxt *bp) +static int bnxt_dl_params_register(struct bnxt *bp) { - struct devlink *dl; int rc; - if (bp->hwrm_spec_code < 0x10600) { - netdev_warn(bp->dev, "Firmware does not support NVM params"); - return -ENOTSUPP; + if (bp->hwrm_spec_code < 0x10600) + return 0; + + rc = devlink_params_register(bp->dl, bnxt_dl_params, + ARRAY_SIZE(bnxt_dl_params)); + if (rc) { + netdev_warn(bp->dev, "devlink_params_register failed. rc=%d", + rc); + return rc; } + rc = devlink_port_params_register(&bp->dl_port, bnxt_dl_port_params, + ARRAY_SIZE(bnxt_dl_port_params)); + if (rc) { + netdev_err(bp->dev, "devlink_port_params_register failed"); + devlink_params_unregister(bp->dl, bnxt_dl_params, + ARRAY_SIZE(bnxt_dl_params)); + return rc; + } + devlink_params_publish(bp->dl); + + return 0; +} + +static void bnxt_dl_params_unregister(struct bnxt *bp) +{ + if (bp->hwrm_spec_code < 0x10600) + return; + + devlink_params_unregister(bp->dl, bnxt_dl_params, + ARRAY_SIZE(bnxt_dl_params)); + devlink_port_params_unregister(&bp->dl_port, bnxt_dl_port_params, + ARRAY_SIZE(bnxt_dl_port_params)); +} + +int bnxt_dl_register(struct bnxt *bp) +{ + struct devlink *dl; + int rc; if (BNXT_PF(bp)) dl = devlink_alloc(&bnxt_dl_ops, sizeof(struct bnxt_dl)); @@ -510,40 +699,23 @@ int bnxt_dl_register(struct bnxt *bp) if (!BNXT_PF(bp)) return 0; - rc = devlink_params_register(dl, bnxt_dl_params, - ARRAY_SIZE(bnxt_dl_params)); - if (rc) { - netdev_warn(bp->dev, "devlink_params_register failed. rc=%d", - rc); - goto err_dl_unreg; - } - devlink_port_attrs_set(&bp->dl_port, DEVLINK_PORT_FLAVOUR_PHYSICAL, - bp->pf.port_id, false, 0, - bp->switch_id, sizeof(bp->switch_id)); + bp->pf.port_id, false, 0, bp->dsn, + sizeof(bp->dsn)); rc = devlink_port_register(dl, &bp->dl_port, bp->pf.port_id); if (rc) { netdev_err(bp->dev, "devlink_port_register failed"); - goto err_dl_param_unreg; + goto err_dl_unreg; } - devlink_port_type_eth_set(&bp->dl_port, bp->dev); - rc = devlink_port_params_register(&bp->dl_port, bnxt_dl_port_params, - ARRAY_SIZE(bnxt_dl_port_params)); - if (rc) { - netdev_err(bp->dev, "devlink_port_params_register failed"); + rc = bnxt_dl_params_register(bp); + if (rc) goto err_dl_port_unreg; - } - - devlink_params_publish(dl); return 0; err_dl_port_unreg: devlink_port_unregister(&bp->dl_port); -err_dl_param_unreg: - devlink_params_unregister(dl, bnxt_dl_params, - ARRAY_SIZE(bnxt_dl_params)); err_dl_unreg: devlink_unregister(dl); err_dl_free: @@ -560,12 +732,8 @@ void bnxt_dl_unregister(struct bnxt *bp) return; if (BNXT_PF(bp)) { - devlink_port_params_unregister(&bp->dl_port, - bnxt_dl_port_params, - ARRAY_SIZE(bnxt_dl_port_params)); + bnxt_dl_params_unregister(bp); devlink_port_unregister(&bp->dl_port); - devlink_params_unregister(dl, bnxt_dl_params, - ARRAY_SIZE(bnxt_dl_params)); } devlink_unregister(dl); devlink_free(dl); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h index 6db6c3dac472..95f893f2a74d 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_devlink.h @@ -38,6 +38,10 @@ static inline void bnxt_link_bp_to_dl(struct bnxt *bp, struct devlink *dl) #define NVM_OFF_IGNORE_ARI 164 #define NVM_OFF_DIS_GRE_VER_CHECK 171 #define NVM_OFF_ENABLE_SRIOV 401 +#define NVM_OFF_NVM_CFG_VER 602 + +#define BNXT_NVM_CFG_VER_BITS 24 +#define BNXT_NVM_CFG_VER_BYTES 4 #define BNXT_MSIX_VEC_MAX 1280 #define BNXT_MSIX_VEC_MIN_MAX 128 @@ -58,6 +62,7 @@ struct bnxt_dl_nvm_param { void bnxt_devlink_health_report(struct bnxt *bp, unsigned long event); void bnxt_dl_health_status_update(struct bnxt *bp, bool healthy); +void bnxt_dl_health_recovery_done(struct bnxt *bp); void bnxt_dl_fw_reporters_create(struct bnxt *bp); void bnxt_dl_fw_reporters_destroy(struct bnxt *bp, bool all); int bnxt_dl_register(struct bnxt *bp); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c index 08d56ec7b68a..6171fa8b3677 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c @@ -1462,15 +1462,15 @@ static int bnxt_get_link_ksettings(struct net_device *dev, ethtool_link_ksettings_add_link_mode(lk_ksettings, advertising, Autoneg); base->autoneg = AUTONEG_ENABLE; - if (link_info->phy_link_status == BNXT_LINK_LINK) + base->duplex = DUPLEX_UNKNOWN; + if (link_info->phy_link_status == BNXT_LINK_LINK) { bnxt_fw_to_ethtool_lp_adv(link_info, lk_ksettings); + if (link_info->duplex & BNXT_LINK_DUPLEX_FULL) + base->duplex = DUPLEX_FULL; + else + base->duplex = DUPLEX_HALF; + } ethtool_speed = bnxt_fw_to_ethtool_speed(link_info->link_speed); - if (!netif_carrier_ok(dev)) - base->duplex = DUPLEX_UNKNOWN; - else if (link_info->duplex & BNXT_LINK_DUPLEX_FULL) - base->duplex = DUPLEX_FULL; - else - base->duplex = DUPLEX_HALF; } else { base->autoneg = AUTONEG_DISABLE; ethtool_speed = @@ -2707,7 +2707,7 @@ static int bnxt_disable_an_for_lpbk(struct bnxt *bp, return rc; fw_speed = PORT_PHY_CFG_REQ_FORCE_LINK_SPEED_1GB; - if (netif_carrier_ok(bp->dev)) + if (bp->link_info.link_up) fw_speed = bp->link_info.link_speed; else if (fw_advertising & BNXT_LINK_SPEED_MSK_10GB) fw_speed = PORT_PHY_CFG_REQ_FORCE_LINK_SPEED_10GB; diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c index 0a8624be44a9..e50a15397e11 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c @@ -2,7 +2,7 @@ /* * Broadcom GENET (Gigabit Ethernet) controller driver * - * Copyright (c) 2014-2017 Broadcom + * Copyright (c) 2014-2019 Broadcom */ #define pr_fmt(fmt) "bcmgenet: " fmt @@ -508,8 +508,8 @@ static int bcmgenet_set_link_ksettings(struct net_device *dev, return phy_ethtool_ksettings_set(dev->phydev, cmd); } -static int bcmgenet_set_rx_csum(struct net_device *dev, - netdev_features_t wanted) +static void bcmgenet_set_rx_csum(struct net_device *dev, + netdev_features_t wanted) { struct bcmgenet_priv *priv = netdev_priv(dev); u32 rbuf_chk_ctrl; @@ -521,7 +521,7 @@ static int bcmgenet_set_rx_csum(struct net_device *dev, /* enable rx checksumming */ if (rx_csum_en) - rbuf_chk_ctrl |= RBUF_RXCHK_EN; + rbuf_chk_ctrl |= RBUF_RXCHK_EN | RBUF_L3_PARSE_DIS; else rbuf_chk_ctrl &= ~RBUF_RXCHK_EN; priv->desc_rxchk_en = rx_csum_en; @@ -535,12 +535,10 @@ static int bcmgenet_set_rx_csum(struct net_device *dev, rbuf_chk_ctrl &= ~RBUF_SKIP_FCS; bcmgenet_rbuf_writel(priv, rbuf_chk_ctrl, RBUF_CHK_CTRL); - - return 0; } -static int bcmgenet_set_tx_csum(struct net_device *dev, - netdev_features_t wanted) +static void bcmgenet_set_tx_csum(struct net_device *dev, + netdev_features_t wanted) { struct bcmgenet_priv *priv = netdev_priv(dev); bool desc_64b_en; @@ -549,7 +547,7 @@ static int bcmgenet_set_tx_csum(struct net_device *dev, tbuf_ctrl = bcmgenet_tbuf_ctrl_get(priv); rbuf_ctrl = bcmgenet_rbuf_readl(priv, RBUF_CTRL); - desc_64b_en = !!(wanted & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)); + desc_64b_en = !!(wanted & NETIF_F_HW_CSUM); /* enable 64 bytes descriptor in both directions (RBUF and TBUF) */ if (desc_64b_en) { @@ -563,21 +561,27 @@ static int bcmgenet_set_tx_csum(struct net_device *dev, bcmgenet_tbuf_ctrl_set(priv, tbuf_ctrl); bcmgenet_rbuf_writel(priv, rbuf_ctrl, RBUF_CTRL); - - return 0; } static int bcmgenet_set_features(struct net_device *dev, netdev_features_t features) { - netdev_features_t changed = features ^ dev->features; - netdev_features_t wanted = dev->wanted_features; - int ret = 0; + struct bcmgenet_priv *priv = netdev_priv(dev); + u32 reg; + int ret; + + ret = clk_prepare_enable(priv->clk); + if (ret) + return ret; - if (changed & (NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM)) - ret = bcmgenet_set_tx_csum(dev, wanted); - if (changed & (NETIF_F_RXCSUM)) - ret = bcmgenet_set_rx_csum(dev, wanted); + /* Make sure we reflect the value of CRC_CMD_FWD */ + reg = bcmgenet_umac_readl(priv, UMAC_CMD); + priv->crc_fwd_en = !!(reg & CMD_CRC_FWD); + + bcmgenet_set_tx_csum(dev, features); + bcmgenet_set_rx_csum(dev, features); + + clk_disable_unprepare(priv->clk); return ret; } @@ -857,6 +861,9 @@ static const struct bcmgenet_stats bcmgenet_gstrings_stats[] = { STAT_GENET_SOFT_MIB("alloc_rx_buff_failed", mib.alloc_rx_buff_failed), STAT_GENET_SOFT_MIB("rx_dma_failed", mib.rx_dma_failed), STAT_GENET_SOFT_MIB("tx_dma_failed", mib.tx_dma_failed), + STAT_GENET_SOFT_MIB("tx_realloc_tsb", mib.tx_realloc_tsb), + STAT_GENET_SOFT_MIB("tx_realloc_tsb_failed", + mib.tx_realloc_tsb_failed), /* Per TX queues */ STAT_GENET_Q(0), STAT_GENET_Q(1), @@ -1218,18 +1225,6 @@ static void bcmgenet_power_up(struct bcmgenet_priv *priv, } } -/* ioctl handle special commands that are not present in ethtool. */ -static int bcmgenet_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - - if (!dev->phydev) - return -ENODEV; - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static struct enet_cb *bcmgenet_get_txcb(struct bcmgenet_priv *priv, struct bcmgenet_tx_ring *ring) { @@ -1483,6 +1478,7 @@ static void bcmgenet_tx_reclaim_all(struct net_device *dev) static struct sk_buff *bcmgenet_put_tx_csum(struct net_device *dev, struct sk_buff *skb) { + struct bcmgenet_priv *priv = netdev_priv(dev); struct status_64 *status = NULL; struct sk_buff *new_skb; u16 offset; @@ -1495,12 +1491,15 @@ static struct sk_buff *bcmgenet_put_tx_csum(struct net_device *dev, * enough headroom for us to insert 64B status block. */ new_skb = skb_realloc_headroom(skb, sizeof(*status)); - dev_kfree_skb(skb); if (!new_skb) { + dev_kfree_skb_any(skb); + priv->mib.tx_realloc_tsb_failed++; dev->stats.tx_dropped++; return NULL; } + dev_consume_skb_any(skb); skb = new_skb; + priv->mib.tx_realloc_tsb++; } skb_push(skb, sizeof(*status)); @@ -1516,24 +1515,19 @@ static struct sk_buff *bcmgenet_put_tx_csum(struct net_device *dev, ip_proto = ipv6_hdr(skb)->nexthdr; break; default: - return skb; + /* don't use UDP flag */ + ip_proto = 0; + break; } offset = skb_checksum_start_offset(skb) - sizeof(*status); tx_csum_info = (offset << STATUS_TX_CSUM_START_SHIFT) | - (offset + skb->csum_offset); + (offset + skb->csum_offset) | + STATUS_TX_CSUM_LV; - /* Set the length valid bit for TCP and UDP and just set - * the special UDP flag for IPv4, else just set to 0. - */ - if (ip_proto == IPPROTO_TCP || ip_proto == IPPROTO_UDP) { - tx_csum_info |= STATUS_TX_CSUM_LV; - if (ip_proto == IPPROTO_UDP && - ip_ver == htons(ETH_P_IP)) - tx_csum_info |= STATUS_TX_CSUM_PROTO_UDP; - } else { - tx_csum_info = 0; - } + /* Set the special UDP flag for UDP */ + if (ip_proto == IPPROTO_UDP) + tx_csum_info |= STATUS_TX_CSUM_PROTO_UDP; status->tx_csum_info = tx_csum_info; } @@ -1744,7 +1738,6 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, unsigned int bytes_processed = 0; unsigned int p_index, mask; unsigned int discards; - unsigned int chksum_ok = 0; /* Clear status before servicing to reduce spurious interrupts */ if (ring->index == DESC_INDEX) { @@ -1795,9 +1788,15 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, dmadesc_get_length_status(priv, cb->bd_addr); } else { struct status_64 *status; + __be16 rx_csum; status = (struct status_64 *)skb->data; dma_length_status = status->length_status; + rx_csum = (__force __be16)(status->rx_csum & 0xffff); + if (priv->desc_rxchk_en) { + skb->csum = (__force __wsum)ntohs(rx_csum); + skb->ip_summed = CHECKSUM_COMPLETE; + } } /* DMA flags and length are still valid no matter how @@ -1840,18 +1839,12 @@ static unsigned int bcmgenet_desc_rx(struct bcmgenet_rx_ring *ring, goto next; } /* error packet */ - chksum_ok = (dma_flag & priv->dma_rx_chk_bit) && - priv->desc_rxchk_en; - skb_put(skb, len); if (priv->desc_64b_en) { skb_pull(skb, 64); len -= 64; } - if (likely(chksum_ok)) - skb->ip_summed = CHECKSUM_UNNECESSARY; - /* remove hardware 2bytes added for IP alignment */ skb_pull(skb, 2); len -= 2; @@ -2886,9 +2879,10 @@ static int bcmgenet_open(struct net_device *dev) init_umac(priv); - /* Make sure we reflect the value of CRC_CMD_FWD */ - reg = bcmgenet_umac_readl(priv, UMAC_CMD); - priv->crc_fwd_en = !!(reg & CMD_CRC_FWD); + /* Apply features again in case we changed them while interface was + * down + */ + bcmgenet_set_features(dev, dev->features); bcmgenet_set_hw_addr(priv, dev->dev_addr); @@ -3055,7 +3049,7 @@ static void bcmgenet_dump_tx_queue(struct bcmgenet_tx_ring *ring) ring->cb_ptr, ring->end_ptr); } -static void bcmgenet_timeout(struct net_device *dev) +static void bcmgenet_timeout(struct net_device *dev, unsigned int txqueue) { struct bcmgenet_priv *priv = netdev_priv(dev); u32 int0_enable = 0; @@ -3216,7 +3210,7 @@ static const struct net_device_ops bcmgenet_netdev_ops = { .ndo_tx_timeout = bcmgenet_timeout, .ndo_set_rx_mode = bcmgenet_set_rx_mode, .ndo_set_mac_address = bcmgenet_set_mac_addr, - .ndo_do_ioctl = bcmgenet_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_features = bcmgenet_set_features, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = bcmgenet_poll_controller, @@ -3327,19 +3321,15 @@ static void bcmgenet_set_hw_params(struct bcmgenet_priv *priv) if (GENET_IS_V5(priv) || GENET_IS_V4(priv)) { bcmgenet_dma_regs = bcmgenet_dma_regs_v3plus; genet_dma_ring_regs = genet_dma_ring_regs_v4; - priv->dma_rx_chk_bit = DMA_RX_CHK_V3PLUS; } else if (GENET_IS_V3(priv)) { bcmgenet_dma_regs = bcmgenet_dma_regs_v3plus; genet_dma_ring_regs = genet_dma_ring_regs_v123; - priv->dma_rx_chk_bit = DMA_RX_CHK_V3PLUS; } else if (GENET_IS_V2(priv)) { bcmgenet_dma_regs = bcmgenet_dma_regs_v2; genet_dma_ring_regs = genet_dma_ring_regs_v123; - priv->dma_rx_chk_bit = DMA_RX_CHK_V12; } else if (GENET_IS_V1(priv)) { bcmgenet_dma_regs = bcmgenet_dma_regs_v1; genet_dma_ring_regs = genet_dma_ring_regs_v123; - priv->dma_rx_chk_bit = DMA_RX_CHK_V12; } /* enum genet_version starts at 1 */ @@ -3535,9 +3525,11 @@ static int bcmgenet_probe(struct platform_device *pdev) priv->msg_enable = netif_msg_init(-1, GENET_MSG_DEFAULT); - /* Set hardware features */ - dev->hw_features |= NETIF_F_SG | NETIF_F_IP_CSUM | - NETIF_F_IPV6_CSUM | NETIF_F_RXCSUM; + /* Set default features */ + dev->features |= NETIF_F_SG | NETIF_F_HIGHDMA | NETIF_F_HW_CSUM | + NETIF_F_RXCSUM; + dev->hw_features |= dev->features; + dev->vlan_features |= dev->features; /* Request the WOL interrupt and advertise suspend if available */ priv->wol_irq_disabled = true; @@ -3574,6 +3566,14 @@ static int bcmgenet_probe(struct platform_device *pdev) bcmgenet_set_hw_params(priv); + err = -EIO; + if (priv->hw_params->flags & GENET_HAS_40BITS) + err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(40)); + if (err) + err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); + if (err) + goto err; + /* Mii wait queue */ init_waitqueue_head(&priv->wq); /* Always use RX_BUF_LENGTH (2KB) buffer for all chips */ @@ -3689,6 +3689,9 @@ static int bcmgenet_resume(struct device *d) genphy_config_aneg(dev->phydev); bcmgenet_mii_config(priv->dev, false); + /* Restore enabled features */ + bcmgenet_set_features(dev, dev->features); + bcmgenet_set_hw_addr(priv, dev->dev_addr); if (priv->internal_phy) { diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.h b/drivers/net/ethernet/broadcom/genet/bcmgenet.h index a5659197598f..61a6fe9f4cec 100644 --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.h +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.h @@ -144,6 +144,8 @@ struct bcmgenet_mib_counters { u32 alloc_rx_buff_failed; u32 rx_dma_failed; u32 tx_dma_failed; + u32 tx_realloc_tsb; + u32 tx_realloc_tsb_failed; }; #define UMAC_HD_BKP_CTRL 0x004 @@ -251,6 +253,7 @@ struct bcmgenet_mib_counters { #define RBUF_CHK_CTRL 0x14 #define RBUF_RXCHK_EN (1 << 0) #define RBUF_SKIP_FCS (1 << 4) +#define RBUF_L3_PARSE_DIS (1 << 5) #define RBUF_ENERGY_CTRL 0x9c #define RBUF_EEE_EN (1 << 0) @@ -663,7 +666,6 @@ struct bcmgenet_priv { bool desc_rxchk_en; bool crc_fwd_en; - unsigned int dma_rx_chk_bit; u32 dma_max_burst_length; u32 msg_enable; diff --git a/drivers/net/ethernet/broadcom/sb1250-mac.c b/drivers/net/ethernet/broadcom/sb1250-mac.c index f991537818fe..5b4568c2ad1c 100644 --- a/drivers/net/ethernet/broadcom/sb1250-mac.c +++ b/drivers/net/ethernet/broadcom/sb1250-mac.c @@ -294,7 +294,7 @@ static int sbmac_set_duplex(struct sbmac_softc *s, enum sbmac_duplex duplex, enum sbmac_fc fc); static int sbmac_open(struct net_device *dev); -static void sbmac_tx_timeout (struct net_device *dev); +static void sbmac_tx_timeout (struct net_device *dev, unsigned int txqueue); static void sbmac_set_rx_mode(struct net_device *dev); static int sbmac_mii_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static int sbmac_close(struct net_device *dev); @@ -2419,7 +2419,7 @@ static void sbmac_mii_poll(struct net_device *dev) } -static void sbmac_tx_timeout (struct net_device *dev) +static void sbmac_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct sbmac_softc *sc = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/ethernet/broadcom/tg3.c b/drivers/net/ethernet/broadcom/tg3.c index ca3aa1250dd1..88466255bf66 100644 --- a/drivers/net/ethernet/broadcom/tg3.c +++ b/drivers/net/ethernet/broadcom/tg3.c @@ -7645,7 +7645,7 @@ static void tg3_poll_controller(struct net_device *dev) } #endif -static void tg3_tx_timeout(struct net_device *dev) +static void tg3_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct tg3 *tp = netdev_priv(dev); @@ -7874,8 +7874,8 @@ static netdev_tx_t tg3_start_xmit(struct sk_buff *, struct net_device *); static int tg3_tso_bug(struct tg3 *tp, struct tg3_napi *tnapi, struct netdev_queue *txq, struct sk_buff *skb) { - struct sk_buff *segs, *nskb; u32 frag_cnt_est = skb_shinfo(skb)->gso_segs * 3; + struct sk_buff *segs, *seg, *next; /* Estimate the number of fragments in the worst case */ if (unlikely(tg3_tx_avail(tnapi) <= frag_cnt_est)) { @@ -7898,12 +7898,10 @@ static int tg3_tso_bug(struct tg3 *tp, struct tg3_napi *tnapi, if (IS_ERR(segs) || !segs) goto tg3_tso_bug_end; - do { - nskb = segs; - segs = segs->next; - nskb->next = NULL; - tg3_start_xmit(nskb, tp->dev); - } while (segs); + skb_list_walk_safe(segs, seg, next) { + skb_mark_not_on_list(seg); + tg3_start_xmit(seg, tp->dev); + } tg3_tso_bug_end: dev_consume_skb_any(skb); diff --git a/drivers/net/ethernet/brocade/bna/bfa_ioc.c b/drivers/net/ethernet/brocade/bna/bfa_ioc.c index 4042c2185e98..e17bfc87da90 100644 --- a/drivers/net/ethernet/brocade/bna/bfa_ioc.c +++ b/drivers/net/ethernet/brocade/bna/bfa_ioc.c @@ -1124,11 +1124,10 @@ bfa_nw_ioc_sem_release(void __iomem *sem_reg) static void bfa_ioc_fwver_clear(struct bfa_ioc *ioc) { - u32 pgnum, pgoff, loff = 0; + u32 pgnum, loff = 0; int i; pgnum = PSS_SMEM_PGNUM(ioc->ioc_regs.smem_pg0, loff); - pgoff = PSS_SMEM_PGOFF(loff); writel(pgnum, ioc->ioc_regs.host_page_num_fn); for (i = 0; i < (sizeof(struct bfi_ioc_image_hdr) / sizeof(u32)); i++) { diff --git a/drivers/net/ethernet/cadence/macb.h b/drivers/net/ethernet/cadence/macb.h index 19fe4f4867c7..dbf7070fcdba 100644 --- a/drivers/net/ethernet/cadence/macb.h +++ b/drivers/net/ethernet/cadence/macb.h @@ -630,10 +630,17 @@ #define GEM_CLK_DIV96 5 /* Constants for MAN register */ -#define MACB_MAN_SOF 1 -#define MACB_MAN_WRITE 1 -#define MACB_MAN_READ 2 -#define MACB_MAN_CODE 2 +#define MACB_MAN_C22_SOF 1 +#define MACB_MAN_C22_WRITE 1 +#define MACB_MAN_C22_READ 2 +#define MACB_MAN_C22_CODE 2 + +#define MACB_MAN_C45_SOF 0 +#define MACB_MAN_C45_ADDR 0 +#define MACB_MAN_C45_WRITE 1 +#define MACB_MAN_C45_POST_READ_INCR 2 +#define MACB_MAN_C45_READ 3 +#define MACB_MAN_C45_CODE 2 /* Capability mask bits */ #define MACB_CAPS_ISR_CLEAR_ON_WRITE 0x00000001 diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c index f7d87c71aaa9..7a2fe63d1136 100644 --- a/drivers/net/ethernet/cadence/macb_main.c +++ b/drivers/net/ethernet/cadence/macb_main.c @@ -337,11 +337,30 @@ static int macb_mdio_read(struct mii_bus *bus, int mii_id, int regnum) if (status < 0) goto mdio_read_exit; - macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_SOF) - | MACB_BF(RW, MACB_MAN_READ) - | MACB_BF(PHYA, mii_id) - | MACB_BF(REGA, regnum) - | MACB_BF(CODE, MACB_MAN_CODE))); + if (regnum & MII_ADDR_C45) { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_ADDR) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(DATA, regnum & 0xFFFF) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + + status = macb_mdio_wait_for_idle(bp); + if (status < 0) + goto mdio_read_exit; + + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_READ) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + } else { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C22_SOF) + | MACB_BF(RW, MACB_MAN_C22_READ) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, regnum) + | MACB_BF(CODE, MACB_MAN_C22_CODE))); + } status = macb_mdio_wait_for_idle(bp); if (status < 0) @@ -370,12 +389,32 @@ static int macb_mdio_write(struct mii_bus *bus, int mii_id, int regnum, if (status < 0) goto mdio_write_exit; - macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_SOF) - | MACB_BF(RW, MACB_MAN_WRITE) - | MACB_BF(PHYA, mii_id) - | MACB_BF(REGA, regnum) - | MACB_BF(CODE, MACB_MAN_CODE) - | MACB_BF(DATA, value))); + if (regnum & MII_ADDR_C45) { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_ADDR) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(DATA, regnum & 0xFFFF) + | MACB_BF(CODE, MACB_MAN_C45_CODE))); + + status = macb_mdio_wait_for_idle(bp); + if (status < 0) + goto mdio_write_exit; + + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C45_SOF) + | MACB_BF(RW, MACB_MAN_C45_WRITE) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, (regnum >> 16) & 0x1F) + | MACB_BF(CODE, MACB_MAN_C45_CODE) + | MACB_BF(DATA, value))); + } else { + macb_writel(bp, MAN, (MACB_BF(SOF, MACB_MAN_C22_SOF) + | MACB_BF(RW, MACB_MAN_C22_WRITE) + | MACB_BF(PHYA, mii_id) + | MACB_BF(REGA, regnum) + | MACB_BF(CODE, MACB_MAN_C22_CODE) + | MACB_BF(DATA, value))); + } status = macb_mdio_wait_for_idle(bp); if (status < 0) diff --git a/drivers/net/ethernet/calxeda/xgmac.c b/drivers/net/ethernet/calxeda/xgmac.c index af04a2c81adb..05a3d067c3fc 100644 --- a/drivers/net/ethernet/calxeda/xgmac.c +++ b/drivers/net/ethernet/calxeda/xgmac.c @@ -1251,7 +1251,7 @@ static int xgmac_poll(struct napi_struct *napi, int budget) * netdev structure and arrange for the device to be reset to a sane state * in order to transmit a new packet. */ -static void xgmac_tx_timeout(struct net_device *dev) +static void xgmac_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct xgmac_priv *priv = netdev_priv(dev); schedule_work(&priv->tx_timeout_work); diff --git a/drivers/net/ethernet/cavium/liquidio/lio_main.c b/drivers/net/ethernet/cavium/liquidio/lio_main.c index 7f3b2e3b0868..eab05b5534ea 100644 --- a/drivers/net/ethernet/cavium/liquidio/lio_main.c +++ b/drivers/net/ethernet/cavium/liquidio/lio_main.c @@ -2562,7 +2562,7 @@ lio_xmit_failed: /** \brief Network device Tx timeout * @param netdev pointer to network device */ -static void liquidio_tx_timeout(struct net_device *netdev) +static void liquidio_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct lio *lio; diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c index 370d76822ee0..7a77544a54f5 100644 --- a/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c +++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_main.c @@ -1628,7 +1628,7 @@ lio_xmit_failed: /** \brief Network device Tx timeout * @param netdev pointer to network device */ -static void liquidio_tx_timeout(struct net_device *netdev) +static void liquidio_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct lio *lio; diff --git a/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c index f3f2e71431ac..600de587d7a9 100644 --- a/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c +++ b/drivers/net/ethernet/cavium/liquidio/lio_vf_rep.c @@ -31,7 +31,7 @@ static int lio_vf_rep_open(struct net_device *ndev); static int lio_vf_rep_stop(struct net_device *ndev); static netdev_tx_t lio_vf_rep_pkt_xmit(struct sk_buff *skb, struct net_device *ndev); -static void lio_vf_rep_tx_timeout(struct net_device *netdev); +static void lio_vf_rep_tx_timeout(struct net_device *netdev, unsigned int txqueue); static int lio_vf_rep_phys_port_name(struct net_device *dev, char *buf, size_t len); static void lio_vf_rep_get_stats64(struct net_device *dev, @@ -172,7 +172,7 @@ lio_vf_rep_stop(struct net_device *ndev) } static void -lio_vf_rep_tx_timeout(struct net_device *ndev) +lio_vf_rep_tx_timeout(struct net_device *ndev, unsigned int txqueue) { netif_trans_update(ndev); diff --git a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c index cdd7e5da4a74..e9575887a4f8 100644 --- a/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c +++ b/drivers/net/ethernet/cavium/octeon/octeon_mgmt.c @@ -790,9 +790,7 @@ static int octeon_mgmt_ioctl(struct net_device *netdev, case SIOCSHWTSTAMP: return octeon_mgmt_ioctl_hwtstamp(netdev, rq, cmd); default: - if (netdev->phydev) - return phy_mii_ioctl(netdev->phydev, rq, cmd); - return -EINVAL; + return phy_do_ioctl(netdev, rq, cmd); } } diff --git a/drivers/net/ethernet/cavium/thunder/nicvf_main.c b/drivers/net/ethernet/cavium/thunder/nicvf_main.c index f28409279ea4..016957285f99 100644 --- a/drivers/net/ethernet/cavium/thunder/nicvf_main.c +++ b/drivers/net/ethernet/cavium/thunder/nicvf_main.c @@ -1741,7 +1741,7 @@ static void nicvf_get_stats64(struct net_device *netdev, } -static void nicvf_tx_timeout(struct net_device *dev) +static void nicvf_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct nicvf *nic = netdev_priv(dev); diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h index becee29f5df7..8b7d156f79d3 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4.h @@ -56,6 +56,7 @@ #include <asm/io.h> #include "t4_chip_type.h" #include "cxgb4_uld.h" +#include "t4fw_api.h" #define CH_WARN(adap, fmt, ...) dev_warn(adap->pdev_dev, fmt, ## __VA_ARGS__) extern struct list_head adapter_list; @@ -68,6 +69,16 @@ extern struct mutex uld_mutex; #define ETHTXQ_STOP_THRES \ (1 + DIV_ROUND_UP((3 * MAX_SKB_FRAGS) / 2 + (MAX_SKB_FRAGS & 1), 8)) +#define FW_PARAM_DEV(param) \ + (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) | \ + FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_##param)) + +#define FW_PARAM_PFVF(param) \ + (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) | \ + FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_##param) | \ + FW_PARAMS_PARAM_Y_V(0) | \ + FW_PARAMS_PARAM_Z_V(0)) + enum { MAX_NPORTS = 4, /* max # of ports */ SERNUM_LEN = 24, /* Serial # length */ diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c index 4144c230dc97..9d1f2f88b945 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_debugfs.c @@ -3174,14 +3174,12 @@ static const struct file_operations mem_debugfs_fops = { static int tid_info_show(struct seq_file *seq, void *v) { - unsigned int tid_start = 0; struct adapter *adap = seq->private; - const struct tid_info *t = &adap->tids; - enum chip_type chip = CHELSIO_CHIP_VERSION(adap->params.chip); - - if (chip > CHELSIO_T5) - tid_start = t4_read_reg(adap, LE_DB_ACTIVE_TABLE_START_INDEX_A); + const struct tid_info *t; + enum chip_type chip; + t = &adap->tids; + chip = CHELSIO_CHIP_VERSION(adap->params.chip); if (t4_read_reg(adap, LE_DB_CONFIG_A) & HASHEN_F) { unsigned int sb; seq_printf(seq, "Connections in use: %u\n", @@ -3193,9 +3191,9 @@ static int tid_info_show(struct seq_file *seq, void *v) sb = t4_read_reg(adap, LE_DB_SRVR_START_INDEX_A); if (sb) { - seq_printf(seq, "TID range: %u..%u/%u..%u", tid_start, + seq_printf(seq, "TID range: %u..%u/%u..%u", t->tid_base, sb - 1, adap->tids.hash_base, - t->ntids - 1); + t->tid_base + t->ntids - 1); seq_printf(seq, ", in use: %u/%u\n", atomic_read(&t->tids_in_use), atomic_read(&t->hash_tids_in_use)); @@ -3204,14 +3202,14 @@ static int tid_info_show(struct seq_file *seq, void *v) t->aftid_base, t->aftid_end, adap->tids.hash_base, - t->ntids - 1); + t->tid_base + t->ntids - 1); seq_printf(seq, ", in use: %u/%u\n", atomic_read(&t->tids_in_use), atomic_read(&t->hash_tids_in_use)); } else { seq_printf(seq, "TID range: %u..%u", adap->tids.hash_base, - t->ntids - 1); + t->tid_base + t->ntids - 1); seq_printf(seq, ", in use: %u\n", atomic_read(&t->hash_tids_in_use)); } @@ -3219,8 +3217,8 @@ static int tid_info_show(struct seq_file *seq, void *v) seq_printf(seq, "Connections in use: %u\n", atomic_read(&t->conns_in_use)); - seq_printf(seq, "TID range: %u..%u", tid_start, - tid_start + t->ntids - 1); + seq_printf(seq, "TID range: %u..%u", t->tid_base, + t->tid_base + t->ntids - 1); seq_printf(seq, ", in use: %u\n", atomic_read(&t->tids_in_use)); } @@ -3243,6 +3241,9 @@ static int tid_info_show(struct seq_file *seq, void *v) seq_printf(seq, "SFTID range: %u..%u in use: %u\n", t->sftid_base, t->sftid_base + t->nsftids - 2, t->sftids_in_use); + if (t->nhpftids) + seq_printf(seq, "HPFTID range: %u..%u\n", t->hpftid_base, + t->hpftid_base + t->nhpftids - 1); if (t->ntids) seq_printf(seq, "HW TID usage: %u IP users, %u IPv6 users\n", t4_read_reg(adap, LE_DB_ACT_CNT_IPV4_A), diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c index 1d39fca11810..2a2938bbb93a 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_filter.c @@ -361,20 +361,22 @@ static int get_filter_count(struct adapter *adapter, unsigned int fidx, tcb_base = t4_read_reg(adapter, TP_CMM_TCB_BASE_A); if (is_hashfilter(adapter) && hash) { - if (fidx < adapter->tids.ntids) { - f = adapter->tids.tid_tab[fidx]; - if (!f) - return -EINVAL; - } else { + if (tid_out_of_range(&adapter->tids, fidx)) return -E2BIG; - } + f = adapter->tids.tid_tab[fidx - adapter->tids.tid_base]; + if (!f) + return -EINVAL; } else { - if ((fidx != (adapter->tids.nftids + - adapter->tids.nsftids - 1)) && - fidx >= adapter->tids.nftids) + if ((fidx != (adapter->tids.nftids + adapter->tids.nsftids + + adapter->tids.nhpftids - 1)) && + fidx >= (adapter->tids.nftids + adapter->tids.nhpftids)) return -E2BIG; - f = &adapter->tids.ftid_tab[fidx]; + if (fidx < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[fidx]; + else + f = &adapter->tids.ftid_tab[fidx - + adapter->tids.nhpftids]; if (!f->valid) return -EINVAL; } @@ -480,6 +482,7 @@ int cxgb4_get_free_ftid(struct net_device *dev, int family) ftid -= n; } spin_unlock_bh(&t->ftid_lock); + ftid += t->nhpftids; return found ? ftid : -ENOMEM; } @@ -507,6 +510,24 @@ static int cxgb4_set_ftid(struct tid_info *t, int fidx, int family, return 0; } +static int cxgb4_set_hpftid(struct tid_info *t, int fidx, int family) +{ + spin_lock_bh(&t->ftid_lock); + + if (test_bit(fidx, t->hpftid_bmap)) { + spin_unlock_bh(&t->ftid_lock); + return -EBUSY; + } + + if (family == PF_INET) + __set_bit(fidx, t->hpftid_bmap); + else + bitmap_allocate_region(t->hpftid_bmap, fidx, 1); + + spin_unlock_bh(&t->ftid_lock); + return 0; +} + static void cxgb4_clear_ftid(struct tid_info *t, int fidx, int family, unsigned int chip_ver) { @@ -522,33 +543,58 @@ static void cxgb4_clear_ftid(struct tid_info *t, int fidx, int family, spin_unlock_bh(&t->ftid_lock); } +static void cxgb4_clear_hpftid(struct tid_info *t, int fidx, int family) +{ + spin_lock_bh(&t->ftid_lock); + + if (family == PF_INET) + __clear_bit(fidx, t->hpftid_bmap); + else + bitmap_release_region(t->hpftid_bmap, fidx, 1); + + spin_unlock_bh(&t->ftid_lock); +} + bool cxgb4_filter_prio_in_range(struct net_device *dev, u32 idx, u32 prio) { + struct filter_entry *prev_fe, *next_fe, *tab; struct adapter *adap = netdev2adap(dev); - struct filter_entry *prev_fe, *next_fe; + u32 prev_ftid, next_ftid, max_tid; struct tid_info *t = &adap->tids; - u32 prev_ftid, next_ftid; + unsigned long *bmap; bool valid = true; + if (idx < t->nhpftids) { + bmap = t->hpftid_bmap; + tab = t->hpftid_tab; + max_tid = t->nhpftids; + } else { + idx -= t->nhpftids; + bmap = t->ftid_bmap; + tab = t->ftid_tab; + max_tid = t->nftids; + } + /* Only insert the rule if both of the following conditions * are met: * 1. The immediate previous rule has priority <= @prio. * 2. The immediate next rule has priority >= @prio. */ spin_lock_bh(&t->ftid_lock); + /* Don't insert if there's a rule already present at @idx. */ - if (test_bit(idx, t->ftid_bmap)) { + if (test_bit(idx, bmap)) { valid = false; goto out_unlock; } - next_ftid = find_next_bit(t->ftid_bmap, t->nftids, idx); - if (next_ftid >= t->nftids) + next_ftid = find_next_bit(bmap, max_tid, idx); + if (next_ftid >= max_tid) next_ftid = idx; - next_fe = &adap->tids.ftid_tab[next_ftid]; + next_fe = &tab[next_ftid]; - prev_ftid = find_last_bit(t->ftid_bmap, idx); + prev_ftid = find_last_bit(bmap, idx); if (prev_ftid >= idx) prev_ftid = idx; @@ -558,13 +604,13 @@ bool cxgb4_filter_prio_in_range(struct net_device *dev, u32 idx, u32 prio) * accordingly. */ if (CHELSIO_CHIP_VERSION(adap->params.chip) < CHELSIO_T6) { - prev_fe = &adap->tids.ftid_tab[prev_ftid & ~0x3]; + prev_fe = &tab[prev_ftid & ~0x3]; if (!prev_fe->fs.type) - prev_fe = &adap->tids.ftid_tab[prev_ftid]; + prev_fe = &tab[prev_ftid]; } else { - prev_fe = &adap->tids.ftid_tab[prev_ftid & ~0x1]; + prev_fe = &tab[prev_ftid & ~0x1]; if (!prev_fe->fs.type) - prev_fe = &adap->tids.ftid_tab[prev_ftid]; + prev_fe = &tab[prev_ftid]; } if ((prev_fe->valid && prio < prev_fe->fs.tc_prio) || @@ -579,11 +625,16 @@ out_unlock: /* Delete the filter at a specified index. */ static int del_filter_wr(struct adapter *adapter, int fidx) { - struct filter_entry *f = &adapter->tids.ftid_tab[fidx]; struct fw_filter_wr *fwr; + struct filter_entry *f; struct sk_buff *skb; unsigned int len; + if (fidx < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[fidx]; + else + f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids]; + len = sizeof(*fwr); skb = alloc_skb(len, GFP_KERNEL); @@ -609,10 +660,15 @@ static int del_filter_wr(struct adapter *adapter, int fidx) */ int set_filter_wr(struct adapter *adapter, int fidx) { - struct filter_entry *f = &adapter->tids.ftid_tab[fidx]; struct fw_filter2_wr *fwr; + struct filter_entry *f; struct sk_buff *skb; + if (fidx < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[fidx]; + else + f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids]; + skb = alloc_skb(sizeof(*fwr), GFP_KERNEL); if (!skb) return -ENOMEM; @@ -762,10 +818,14 @@ int delete_filter(struct adapter *adapter, unsigned int fidx) struct filter_entry *f; int ret; - if (fidx >= adapter->tids.nftids + adapter->tids.nsftids) + if (fidx >= adapter->tids.nftids + adapter->tids.nsftids + + adapter->tids.nhpftids) return -EINVAL; - f = &adapter->tids.ftid_tab[fidx]; + if (fidx < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[fidx]; + else + f = &adapter->tids.ftid_tab[fidx - adapter->tids.nhpftids]; ret = writable_filter(f); if (ret) return ret; @@ -811,12 +871,22 @@ void clear_all_filters(struct adapter *adapter) struct net_device *dev = adapter->port[0]; unsigned int i; + if (adapter->tids.hpftid_tab) { + struct filter_entry *f = &adapter->tids.hpftid_tab[0]; + + for (i = 0; i < adapter->tids.nhpftids; i++, f++) + if (f->valid || f->pending) + cxgb4_del_filter(dev, i, &f->fs); + } + if (adapter->tids.ftid_tab) { struct filter_entry *f = &adapter->tids.ftid_tab[0]; unsigned int max_ftid = adapter->tids.nftids + - adapter->tids.nsftids; + adapter->tids.nsftids + + adapter->tids.nhpftids; + /* Clear all TCAM filters */ - for (i = 0; i < max_ftid; i++, f++) + for (i = adapter->tids.nhpftids; i < max_ftid; i++, f++) if (f->valid || f->pending) cxgb4_del_filter(dev, i, &f->fs); } @@ -1319,17 +1389,17 @@ out_err: * filter specification in order to facilitate signaling completion of the * operation. */ -int __cxgb4_set_filter(struct net_device *dev, int filter_id, +int __cxgb4_set_filter(struct net_device *dev, int ftid, struct ch_filter_specification *fs, struct filter_ctx *ctx) { struct adapter *adapter = netdev2adap(dev); - unsigned int chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip); - unsigned int max_fidx, fidx; - struct filter_entry *f; + unsigned int max_fidx, fidx, chip_ver; + int iq, ret, filter_id = ftid; + struct filter_entry *f, *tab; u32 iconf; - int iq, ret; + chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip); if (fs->hash) { if (is_hashfilter(adapter)) return cxgb4_set_hash_filter(dev, fs, ctx); @@ -1338,7 +1408,7 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, return -EINVAL; } - max_fidx = adapter->tids.nftids; + max_fidx = adapter->tids.nftids + adapter->tids.nhpftids; if (filter_id != (max_fidx + adapter->tids.nsftids - 1) && filter_id >= max_fidx) return -E2BIG; @@ -1353,6 +1423,13 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, if (iq < 0) return iq; + if (fs->prio) { + tab = &adapter->tids.hpftid_tab[0]; + } else { + tab = &adapter->tids.ftid_tab[0]; + filter_id = ftid - adapter->tids.nhpftids; + } + /* IPv6 filters occupy four slots and must be aligned on * four-slot boundaries. IPv4 filters only occupy a single * slot and have no alignment requirements but writing a new @@ -1373,9 +1450,8 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, else fidx = filter_id & ~0x1; - if (fidx != filter_id && - adapter->tids.ftid_tab[fidx].fs.type) { - f = &adapter->tids.ftid_tab[fidx]; + if (fidx != filter_id && tab[fidx].fs.type) { + f = &tab[fidx]; if (f->valid) { dev_err(adapter->pdev_dev, "Invalid location. IPv6 requires 4 slots and is occupying slots %u to %u\n", @@ -1399,7 +1475,7 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, */ for (fidx = filter_id + 1; fidx < filter_id + 4; fidx++) { - f = &adapter->tids.ftid_tab[fidx]; + f = &tab[fidx]; if (f->valid) { dev_err(adapter->pdev_dev, "Invalid location. IPv6 requires 4 slots and an IPv4 filter exists at %u\n", @@ -1415,7 +1491,7 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, return -EINVAL; /* Check overlapping IPv4 filter slot */ fidx = filter_id + 1; - f = &adapter->tids.ftid_tab[fidx]; + f = &tab[fidx]; if (f->valid) { pr_err("%s: IPv6 filter requires 2 indices. IPv4 filter already present at %d. Please remove IPv4 filter first.\n", __func__, fidx); @@ -1427,36 +1503,35 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, /* Check to make sure that provided filter index is not * already in use by someone else */ - f = &adapter->tids.ftid_tab[filter_id]; + f = &tab[filter_id]; if (f->valid) return -EBUSY; - fidx = filter_id + adapter->tids.ftid_base; - ret = cxgb4_set_ftid(&adapter->tids, filter_id, - fs->type ? PF_INET6 : PF_INET, - chip_ver); + if (fs->prio) { + fidx = filter_id + adapter->tids.hpftid_base; + ret = cxgb4_set_hpftid(&adapter->tids, filter_id, + fs->type ? PF_INET6 : PF_INET); + } else { + fidx = filter_id + adapter->tids.ftid_base; + ret = cxgb4_set_ftid(&adapter->tids, filter_id, + fs->type ? PF_INET6 : PF_INET, + chip_ver); + } + if (ret) return ret; /* Check t make sure the filter requested is writable ... */ ret = writable_filter(f); - if (ret) { - /* Clear the bits we have set above */ - cxgb4_clear_ftid(&adapter->tids, filter_id, - fs->type ? PF_INET6 : PF_INET, - chip_ver); - return ret; - } + if (ret) + goto free_tid; if (is_t6(adapter->params.chip) && fs->type && ipv6_addr_type((const struct in6_addr *)fs->val.lip) != IPV6_ADDR_ANY) { ret = cxgb4_clip_get(dev, (const u32 *)&fs->val.lip, 1); - if (ret) { - cxgb4_clear_ftid(&adapter->tids, filter_id, PF_INET6, - chip_ver); - return ret; - } + if (ret) + goto free_tid; } /* Convert the filter specification into our internal format. @@ -1487,7 +1562,7 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, f->fs.mask.vni, 0, 1, 1); if (ret < 0) - goto free_clip; + goto free_tid; f->fs.val.ovlan = ret; f->fs.mask.ovlan = 0x1ff; @@ -1501,21 +1576,22 @@ int __cxgb4_set_filter(struct net_device *dev, int filter_id, */ f->ctx = ctx; f->tid = fidx; /* Save the actual tid */ - ret = set_filter_wr(adapter, filter_id); - if (ret) { + ret = set_filter_wr(adapter, ftid); + if (ret) + goto free_tid; + + return ret; + +free_tid: + if (f->fs.prio) + cxgb4_clear_hpftid(&adapter->tids, filter_id, + fs->type ? PF_INET6 : PF_INET); + else cxgb4_clear_ftid(&adapter->tids, filter_id, fs->type ? PF_INET6 : PF_INET, chip_ver); - clear_filter(adapter, f); - } - - return ret; -free_clip: - if (is_t6(adapter->params.chip) && f->fs.type) - cxgb4_clip_release(f->dev, (const u32 *)&f->fs.val.lip, 1); - cxgb4_clear_ftid(&adapter->tids, filter_id, - fs->type ? PF_INET6 : PF_INET, chip_ver); + clear_filter(adapter, f); return ret; } @@ -1537,7 +1613,7 @@ static int cxgb4_del_hash_filter(struct net_device *dev, int filter_id, netdev_dbg(dev, "%s: filter_id = %d ; nftids = %d\n", __func__, filter_id, adapter->tids.nftids); - if (filter_id > adapter->tids.ntids) + if (tid_out_of_range(t, filter_id)) return -E2BIG; f = lookup_tid(t, filter_id); @@ -1590,11 +1666,11 @@ int __cxgb4_del_filter(struct net_device *dev, int filter_id, struct filter_ctx *ctx) { struct adapter *adapter = netdev2adap(dev); - unsigned int chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip); + unsigned int max_fidx, chip_ver; struct filter_entry *f; - unsigned int max_fidx; int ret; + chip_ver = CHELSIO_CHIP_VERSION(adapter->params.chip); if (fs && fs->hash) { if (is_hashfilter(adapter)) return cxgb4_del_hash_filter(dev, filter_id, ctx); @@ -1603,21 +1679,31 @@ int __cxgb4_del_filter(struct net_device *dev, int filter_id, return -EINVAL; } - max_fidx = adapter->tids.nftids; + max_fidx = adapter->tids.nftids + adapter->tids.nhpftids; if (filter_id != (max_fidx + adapter->tids.nsftids - 1) && filter_id >= max_fidx) return -E2BIG; - f = &adapter->tids.ftid_tab[filter_id]; + if (filter_id < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[filter_id]; + else + f = &adapter->tids.ftid_tab[filter_id - adapter->tids.nhpftids]; + ret = writable_filter(f); if (ret) return ret; if (f->valid) { f->ctx = ctx; - cxgb4_clear_ftid(&adapter->tids, filter_id, - f->fs.type ? PF_INET6 : PF_INET, - chip_ver); + if (f->fs.prio) + cxgb4_clear_hpftid(&adapter->tids, + f->tid - adapter->tids.hpftid_base, + f->fs.type ? PF_INET6 : PF_INET); + else + cxgb4_clear_ftid(&adapter->tids, + f->tid - adapter->tids.ftid_base, + f->fs.type ? PF_INET6 : PF_INET, + chip_ver); return del_filter_wr(adapter, filter_id); } @@ -1842,11 +1928,18 @@ void filter_rpl(struct adapter *adap, const struct cpl_set_tcb_rpl *rpl) max_fidx = adap->tids.nftids + adap->tids.nsftids; /* Get the corresponding filter entry for this tid */ if (adap->tids.ftid_tab) { - /* Check this in normal filter region */ - idx = tid - adap->tids.ftid_base; - if (idx >= max_fidx) - return; - f = &adap->tids.ftid_tab[idx]; + idx = tid - adap->tids.hpftid_base; + if (idx < adap->tids.nhpftids) { + f = &adap->tids.hpftid_tab[idx]; + } else { + /* Check this in normal filter region */ + idx = tid - adap->tids.ftid_base; + if (idx >= max_fidx) + return; + f = &adap->tids.ftid_tab[idx]; + idx += adap->tids.nhpftids; + } + if (f->tid != tid) return; } diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c index 0dedd3e9c31e..649842a8aa28 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c @@ -804,6 +804,26 @@ static int setup_ppod_edram(struct adapter *adap) return 0; } +static void adap_config_hpfilter(struct adapter *adapter) +{ + u32 param, val = 0; + int ret; + + /* Enable HP filter region. Older fw will fail this request and + * it is fine. + */ + param = FW_PARAM_DEV(HPFILTER_REGION_SUPPORT); + ret = t4_set_params(adapter, adapter->mbox, adapter->pf, 0, + 1, ¶m, &val); + + /* An error means FW doesn't know about HP filter support, + * it's not a problem, don't return an error. + */ + if (ret < 0) + dev_err(adapter->pdev_dev, + "HP filter region isn't supported by FW\n"); +} + /** * cxgb4_write_rss - write the RSS table for a given port * @pi: the port @@ -1427,8 +1447,8 @@ static void mk_tid_release(struct sk_buff *skb, unsigned int chan, static void cxgb4_queue_tid_release(struct tid_info *t, unsigned int chan, unsigned int tid) { - void **p = &t->tid_tab[tid]; struct adapter *adap = container_of(t, struct adapter, tids); + void **p = &t->tid_tab[tid - t->tid_base]; spin_lock_bh(&adap->tid_release_lock); *p = adap->tid_release_head; @@ -1480,13 +1500,13 @@ static void process_tid_release_list(struct work_struct *work) void cxgb4_remove_tid(struct tid_info *t, unsigned int chan, unsigned int tid, unsigned short family) { - struct sk_buff *skb; struct adapter *adap = container_of(t, struct adapter, tids); + struct sk_buff *skb; - WARN_ON(tid >= t->ntids); + WARN_ON(tid_out_of_range(&adap->tids, tid)); - if (t->tid_tab[tid]) { - t->tid_tab[tid] = NULL; + if (t->tid_tab[tid - adap->tids.tid_base]) { + t->tid_tab[tid - adap->tids.tid_base] = NULL; atomic_dec(&t->conns_in_use); if (t->hash_base && (tid >= t->hash_base)) { if (family == AF_INET6) @@ -1518,6 +1538,7 @@ static int tid_init(struct tid_info *t) struct adapter *adap = container_of(t, struct adapter, tids); unsigned int max_ftids = t->nftids + t->nsftids; unsigned int natids = t->natids; + unsigned int hpftid_bmap_size; unsigned int eotid_bmap_size; unsigned int stid_bmap_size; unsigned int ftid_bmap_size; @@ -1525,12 +1546,15 @@ static int tid_init(struct tid_info *t) stid_bmap_size = BITS_TO_LONGS(t->nstids + t->nsftids); ftid_bmap_size = BITS_TO_LONGS(t->nftids); + hpftid_bmap_size = BITS_TO_LONGS(t->nhpftids); eotid_bmap_size = BITS_TO_LONGS(t->neotids); size = t->ntids * sizeof(*t->tid_tab) + natids * sizeof(*t->atid_tab) + t->nstids * sizeof(*t->stid_tab) + t->nsftids * sizeof(*t->stid_tab) + stid_bmap_size * sizeof(long) + + t->nhpftids * sizeof(*t->hpftid_tab) + + hpftid_bmap_size * sizeof(long) + max_ftids * sizeof(*t->ftid_tab) + ftid_bmap_size * sizeof(long) + t->neotids * sizeof(*t->eotid_tab) + @@ -1543,7 +1567,9 @@ static int tid_init(struct tid_info *t) t->atid_tab = (union aopen_entry *)&t->tid_tab[t->ntids]; t->stid_tab = (struct serv_entry *)&t->atid_tab[natids]; t->stid_bmap = (unsigned long *)&t->stid_tab[t->nstids + t->nsftids]; - t->ftid_tab = (struct filter_entry *)&t->stid_bmap[stid_bmap_size]; + t->hpftid_tab = (struct filter_entry *)&t->stid_bmap[stid_bmap_size]; + t->hpftid_bmap = (unsigned long *)&t->hpftid_tab[t->nhpftids]; + t->ftid_tab = (struct filter_entry *)&t->hpftid_bmap[hpftid_bmap_size]; t->ftid_bmap = (unsigned long *)&t->ftid_tab[max_ftids]; t->eotid_tab = (struct eotid_entry *)&t->ftid_bmap[ftid_bmap_size]; t->eotid_bmap = (unsigned long *)&t->eotid_tab[t->neotids]; @@ -1578,6 +1604,8 @@ static int tid_init(struct tid_info *t) bitmap_zero(t->eotid_bmap, t->neotids); } + if (t->nhpftids) + bitmap_zero(t->hpftid_bmap, t->nhpftids); bitmap_zero(t->ftid_bmap, t->nftids); return 0; } @@ -4359,6 +4387,7 @@ static int adap_init0_config(struct adapter *adapter, int reset) "HMA configuration failed with error %d\n", ret); if (is_t6(adapter->params.chip)) { + adap_config_hpfilter(adapter); ret = setup_ppod_edram(adapter); if (!ret) dev_info(adapter->pdev_dev, "Successfully enabled " @@ -4668,16 +4697,6 @@ static int adap_init0(struct adapter *adap, int vpd_skip) /* * Grab some of our basic fundamental operating parameters. */ -#define FW_PARAM_DEV(param) \ - (FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_DEV) | \ - FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_DEV_##param)) - -#define FW_PARAM_PFVF(param) \ - FW_PARAMS_MNEM_V(FW_PARAMS_MNEM_PFVF) | \ - FW_PARAMS_PARAM_X_V(FW_PARAMS_PARAM_PFVF_##param)| \ - FW_PARAMS_PARAM_Y_V(0) | \ - FW_PARAMS_PARAM_Z_V(0) - params[0] = FW_PARAM_PFVF(EQ_START); params[1] = FW_PARAM_PFVF(L2T_START); params[2] = FW_PARAM_PFVF(L2T_END); @@ -4695,6 +4714,16 @@ static int adap_init0(struct adapter *adap, int vpd_skip) adap->sge.ingr_start = val[5]; if (CHELSIO_CHIP_VERSION(adap->params.chip) > CHELSIO_T5) { + params[0] = FW_PARAM_PFVF(HPFILTER_START); + params[1] = FW_PARAM_PFVF(HPFILTER_END); + ret = t4_query_params(adap, adap->mbox, adap->pf, 0, 2, + params, val); + if (ret < 0) + goto bye; + + adap->tids.hpftid_base = val[0]; + adap->tids.nhpftids = val[1] - val[0] + 1; + /* Read the raw mps entries. In T6, the last 2 tcam entries * are reserved for raw mac addresses (rawf = 2, one per port). */ @@ -4706,6 +4735,9 @@ static int adap_init0(struct adapter *adap, int vpd_skip) adap->rawf_start = val[0]; adap->rawf_cnt = val[1] - val[0] + 1; } + + adap->tids.tid_base = + t4_read_reg(adap, LE_DB_ACTIVE_TABLE_START_INDEX_A); } /* qids (ingress/egress) returned from firmware can be anywhere @@ -5058,8 +5090,6 @@ static int adap_init0(struct adapter *adap, int vpd_skip) } adap->params.crypto = ntohs(caps_cmd.cryptocaps); } -#undef FW_PARAM_PFVF -#undef FW_PARAM_DEV /* The MTU/MSS Table is initialized by now, so load their values. If * we're initializing the adapter, then we'll make any modifications diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c index 0fa80bef575d..bb5513bdd293 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_flower.c @@ -672,10 +672,14 @@ int cxgb4_tc_flower_replace(struct net_device *dev, * 0 to driver. However, the hardware TCAM index * starts from 0. Hence, the -1 here. */ - if (cls->common.prio <= adap->tids.nftids) + if (cls->common.prio <= (adap->tids.nftids + + adap->tids.nhpftids)) { fidx = cls->common.prio - 1; - else + if (fidx < adap->tids.nhpftids) + fs->prio = 1; + } else { fidx = cxgb4_get_free_ftid(dev, inet_family); + } /* Only insert FLOWER rule if its priority doesn't * conflict with existing rules in the LETCAM. diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c index 6d485803ddbe..1b7681a4eb32 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_matchall.c @@ -204,7 +204,7 @@ static int cxgb4_matchall_alloc_filter(struct net_device *dev, * -1 here. 1 slot is enough to create a wildcard matchall * VIID rule. */ - if (cls->common.prio <= adap->tids.nftids) + if (cls->common.prio <= (adap->tids.nftids + adap->tids.nhpftids)) fidx = cls->common.prio - 1; else fidx = cxgb4_get_free_ftid(dev, PF_INET); @@ -223,6 +223,8 @@ static int cxgb4_matchall_alloc_filter(struct net_device *dev, fs = &tc_port_matchall->ingress.fs; memset(fs, 0, sizeof(*fs)); + if (fidx < adap->tids.nhpftids) + fs->prio = 1; fs->tc_prio = cls->common.prio; fs->tc_cookie = cls->cookie; fs->hitcnts = 1; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c index 133f8623ba86..269b8d9e25e0 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32.c @@ -176,7 +176,7 @@ int cxgb4_config_knode(struct net_device *dev, struct tc_cls_u32_offload *cls) /* Only insert U32 rule if its priority doesn't conflict with * existing rules in the LETCAM. */ - if (filter_id >= adapter->tids.nftids || + if (filter_id >= adapter->tids.nftids + adapter->tids.nhpftids || !cxgb4_filter_prio_in_range(dev, filter_id, cls->common.prio)) { NL_SET_ERR_MSG_MOD(extack, "No free LETCAM index available"); @@ -199,6 +199,8 @@ int cxgb4_config_knode(struct net_device *dev, struct tc_cls_u32_offload *cls) memset(&fs, 0, sizeof(fs)); + if (filter_id < adapter->tids.nhpftids) + fs.prio = 1; fs.tc_prio = cls->common.prio; fs.tc_cookie = cls->knode.handle; @@ -355,6 +357,7 @@ int cxgb4_delete_knode(struct net_device *dev, struct tc_cls_u32_offload *cls) unsigned int filter_id, max_tids, i, j; struct cxgb4_link *link = NULL; struct cxgb4_tc_u32_table *t; + struct filter_entry *f; u32 handle, uhtid; int ret; @@ -363,8 +366,15 @@ int cxgb4_delete_knode(struct net_device *dev, struct tc_cls_u32_offload *cls) /* Fetch the location to delete the filter. */ filter_id = TC_U32_NODE(cls->knode.handle) - 1; - if (filter_id >= adapter->tids.nftids || - cls->knode.handle != adapter->tids.ftid_tab[filter_id].fs.tc_cookie) + if (filter_id >= adapter->tids.nftids + adapter->tids.nhpftids) + return -ERANGE; + + if (filter_id < adapter->tids.nhpftids) + f = &adapter->tids.hpftid_tab[filter_id]; + else + f = &adapter->tids.ftid_tab[filter_id - adapter->tids.nhpftids]; + + if (cls->knode.handle != f->fs.tc_cookie) return -ERANGE; t = adapter->tc_u32; @@ -445,7 +455,7 @@ void cxgb4_cleanup_tc_u32(struct adapter *adap) struct cxgb4_tc_u32_table *cxgb4_init_tc_u32(struct adapter *adap) { - unsigned int max_tids = adap->tids.nftids; + unsigned int max_tids = adap->tids.nftids + adap->tids.nhpftids; struct cxgb4_tc_u32_table *t; unsigned int i; diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h index 861b25d28ed6..d9d27bc1ae67 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h +++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_uld.h @@ -99,6 +99,7 @@ struct eotid_entry { */ struct tid_info { void **tid_tab; + unsigned int tid_base; unsigned int ntids; struct serv_entry *stid_tab; @@ -111,6 +112,11 @@ struct tid_info { unsigned int natids; unsigned int atid_base; + struct filter_entry *hpftid_tab; + unsigned long *hpftid_bmap; + unsigned int nhpftids; + unsigned int hpftid_base; + struct filter_entry *ftid_tab; unsigned long *ftid_bmap; unsigned int nftids; @@ -147,9 +153,15 @@ struct tid_info { static inline void *lookup_tid(const struct tid_info *t, unsigned int tid) { + tid -= t->tid_base; return tid < t->ntids ? t->tid_tab[tid] : NULL; } +static inline bool tid_out_of_range(const struct tid_info *t, unsigned int tid) +{ + return ((tid - t->tid_base) >= t->ntids); +} + static inline void *lookup_atid(const struct tid_info *t, unsigned int atid) { return atid < t->natids ? t->atid_tab[atid].data : NULL; @@ -171,7 +183,7 @@ static inline void *lookup_stid(const struct tid_info *t, unsigned int stid) static inline void cxgb4_insert_tid(struct tid_info *t, void *data, unsigned int tid, unsigned short family) { - t->tid_tab[tid] = data; + t->tid_tab[tid - t->tid_base] = data; if (t->hash_base && (tid >= t->hash_base)) { if (family == AF_INET6) atomic_add(2, &t->hash_tids_in_use); diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h index ac4fb43bdec6..accad1101ad1 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h +++ b/drivers/net/ethernet/chelsio/cxgb4/t4fw_api.h @@ -1321,6 +1321,7 @@ enum fw_params_param_dev { FW_PARAMS_PARAM_DEV_RDMA_WRITE_WITH_IMM = 0x21, FW_PARAMS_PARAM_DEV_PPOD_EDRAM = 0x23, FW_PARAMS_PARAM_DEV_RI_WRITE_CMPL_WR = 0x24, + FW_PARAMS_PARAM_DEV_HPFILTER_REGION_SUPPORT = 0x26, FW_PARAMS_PARAM_DEV_OPAQUE_VIID_SMT_EXTN = 0x27, FW_PARAMS_PARAM_DEV_HASHFILTER_WITH_OFLD = 0x28, FW_PARAMS_PARAM_DEV_DBQ_TIMER = 0x29, diff --git a/drivers/net/ethernet/cirrus/cs89x0.c b/drivers/net/ethernet/cirrus/cs89x0.c index c9aebcde403a..33ace3307059 100644 --- a/drivers/net/ethernet/cirrus/cs89x0.c +++ b/drivers/net/ethernet/cirrus/cs89x0.c @@ -1128,7 +1128,7 @@ net_get_stats(struct net_device *dev) return &dev->stats; } -static void net_timeout(struct net_device *dev) +static void net_timeout(struct net_device *dev, unsigned int txqueue) { /* If we get here, some higher level has decided we are broken. There should really be a "kick me" function call instead. */ diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c index acb2856936d2..bbd7b3175f09 100644 --- a/drivers/net/ethernet/cisco/enic/enic_main.c +++ b/drivers/net/ethernet/cisco/enic/enic_main.c @@ -1095,7 +1095,7 @@ static void enic_set_rx_mode(struct net_device *netdev) } /* netif_tx_lock held, BHs disabled */ -static void enic_tx_timeout(struct net_device *netdev) +static void enic_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct enic *enic = netdev_priv(netdev); schedule_work(&enic->tx_hang_reset); diff --git a/drivers/net/ethernet/cortina/gemini.c b/drivers/net/ethernet/cortina/gemini.c index 2814b96751b4..f30fa8e6ef80 100644 --- a/drivers/net/ethernet/cortina/gemini.c +++ b/drivers/net/ethernet/cortina/gemini.c @@ -1298,7 +1298,7 @@ out_drop: return NETDEV_TX_OK; } -static void gmac_tx_timeout(struct net_device *netdev) +static void gmac_tx_timeout(struct net_device *netdev, unsigned int txqueue) { netdev_err(netdev, "Tx timeout\n"); gmac_dump_dma_state(netdev); diff --git a/drivers/net/ethernet/davicom/dm9000.c b/drivers/net/ethernet/davicom/dm9000.c index cce90b5925d9..1ea3372775e6 100644 --- a/drivers/net/ethernet/davicom/dm9000.c +++ b/drivers/net/ethernet/davicom/dm9000.c @@ -964,7 +964,7 @@ dm9000_init_dm9000(struct net_device *dev) } /* Our watchdog timed out. Called by the networking layer */ -static void dm9000_timeout(struct net_device *dev) +static void dm9000_timeout(struct net_device *dev, unsigned int txqueue) { struct board_info *db = netdev_priv(dev); u8 reg_save; diff --git a/drivers/net/ethernet/dec/tulip/de2104x.c b/drivers/net/ethernet/dec/tulip/de2104x.c index 7852a4308194..d305d1b24b0a 100644 --- a/drivers/net/ethernet/dec/tulip/de2104x.c +++ b/drivers/net/ethernet/dec/tulip/de2104x.c @@ -1436,7 +1436,7 @@ static int de_close (struct net_device *dev) return 0; } -static void de_tx_timeout (struct net_device *dev) +static void de_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct de_private *de = netdev_priv(dev); const int irq = de->pdev->irq; diff --git a/drivers/net/ethernet/dec/tulip/dmfe.c b/drivers/net/ethernet/dec/tulip/dmfe.c index 0efdbd1a4a6f..32d470d4122a 100644 --- a/drivers/net/ethernet/dec/tulip/dmfe.c +++ b/drivers/net/ethernet/dec/tulip/dmfe.c @@ -2214,15 +2214,16 @@ static int __init dmfe_init_module(void) if (cr6set) dmfe_cr6_user_set = cr6set; - switch(mode) { - case DMFE_10MHF: + switch (mode) { + case DMFE_10MHF: case DMFE_100MHF: case DMFE_10MFD: case DMFE_100MFD: case DMFE_1M_HPNA: dmfe_media_mode = mode; break; - default:dmfe_media_mode = DMFE_AUTO; + default: + dmfe_media_mode = DMFE_AUTO; break; } diff --git a/drivers/net/ethernet/dec/tulip/tulip_core.c b/drivers/net/ethernet/dec/tulip/tulip_core.c index 3e3e08698876..9e9d9eee29d9 100644 --- a/drivers/net/ethernet/dec/tulip/tulip_core.c +++ b/drivers/net/ethernet/dec/tulip/tulip_core.c @@ -255,7 +255,7 @@ MODULE_DEVICE_TABLE(pci, tulip_pci_tbl); const char tulip_media_cap[32] = {0,0,0,16, 3,19,16,24, 27,4,7,5, 0,20,23,20, 28,31,0,0, }; -static void tulip_tx_timeout(struct net_device *dev); +static void tulip_tx_timeout(struct net_device *dev, unsigned int txqueue); static void tulip_init_ring(struct net_device *dev); static void tulip_free_ring(struct net_device *dev); static netdev_tx_t tulip_start_xmit(struct sk_buff *skb, @@ -534,7 +534,7 @@ free_ring: } -static void tulip_tx_timeout(struct net_device *dev) +static void tulip_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct tulip_private *tp = netdev_priv(dev); void __iomem *ioaddr = tp->base_addr; diff --git a/drivers/net/ethernet/dec/tulip/uli526x.c b/drivers/net/ethernet/dec/tulip/uli526x.c index b1f30b194300..117ffe08800d 100644 --- a/drivers/net/ethernet/dec/tulip/uli526x.c +++ b/drivers/net/ethernet/dec/tulip/uli526x.c @@ -1809,8 +1809,8 @@ static int __init uli526x_init_module(void) if (cr6set) uli526x_cr6_user_set = cr6set; - switch (mode) { - case ULI526X_10MHF: + switch (mode) { + case ULI526X_10MHF: case ULI526X_100MHF: case ULI526X_10MFD: case ULI526X_100MFD: diff --git a/drivers/net/ethernet/dec/tulip/winbond-840.c b/drivers/net/ethernet/dec/tulip/winbond-840.c index 70cb2d689c2c..7f136488e67c 100644 --- a/drivers/net/ethernet/dec/tulip/winbond-840.c +++ b/drivers/net/ethernet/dec/tulip/winbond-840.c @@ -331,7 +331,7 @@ static void netdev_timer(struct timer_list *t); static void init_rxtx_rings(struct net_device *dev); static void free_rxtx_rings(struct netdev_private *np); static void init_registers(struct net_device *dev); -static void tx_timeout(struct net_device *dev); +static void tx_timeout(struct net_device *dev, unsigned int txqueue); static int alloc_ringdesc(struct net_device *dev); static void free_ringdesc(struct netdev_private *np); static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); @@ -921,7 +921,7 @@ static void init_registers(struct net_device *dev) iowrite32(0, ioaddr + RxStartDemand); } -static void tx_timeout(struct net_device *dev) +static void tx_timeout(struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem *ioaddr = np->base_addr; diff --git a/drivers/net/ethernet/dlink/dl2k.c b/drivers/net/ethernet/dlink/dl2k.c index 55e720d2ea0c..26c5da032b1e 100644 --- a/drivers/net/ethernet/dlink/dl2k.c +++ b/drivers/net/ethernet/dlink/dl2k.c @@ -66,7 +66,7 @@ static const int multicast_filter_limit = 0x40; static int rio_open (struct net_device *dev); static void rio_timer (struct timer_list *t); -static void rio_tx_timeout (struct net_device *dev); +static void rio_tx_timeout (struct net_device *dev, unsigned int txqueue); static netdev_tx_t start_xmit (struct sk_buff *skb, struct net_device *dev); static irqreturn_t rio_interrupt (int irq, void *dev_instance); static void rio_free_tx (struct net_device *dev, int irq); @@ -696,7 +696,7 @@ rio_timer (struct timer_list *t) } static void -rio_tx_timeout (struct net_device *dev) +rio_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem *ioaddr = np->ioaddr; diff --git a/drivers/net/ethernet/dlink/sundance.c b/drivers/net/ethernet/dlink/sundance.c index 4a37a69764ce..b91387c456ba 100644 --- a/drivers/net/ethernet/dlink/sundance.c +++ b/drivers/net/ethernet/dlink/sundance.c @@ -432,7 +432,7 @@ static int mdio_wait_link(struct net_device *dev, int wait); static int netdev_open(struct net_device *dev); static void check_duplex(struct net_device *dev); static void netdev_timer(struct timer_list *t); -static void tx_timeout(struct net_device *dev); +static void tx_timeout(struct net_device *dev, unsigned int txqueue); static void init_ring(struct net_device *dev); static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static int reset_tx (struct net_device *dev); @@ -969,7 +969,7 @@ static void netdev_timer(struct timer_list *t) add_timer(&np->timer); } -static void tx_timeout(struct net_device *dev) +static void tx_timeout(struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem *ioaddr = np->base; diff --git a/drivers/net/ethernet/dnet.c b/drivers/net/ethernet/dnet.c index e24979010969..5f8fa1145db6 100644 --- a/drivers/net/ethernet/dnet.c +++ b/drivers/net/ethernet/dnet.c @@ -725,19 +725,6 @@ static struct net_device_stats *dnet_get_stats(struct net_device *dev) return nstat; } -static int dnet_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - struct phy_device *phydev = dev->phydev; - - if (!netif_running(dev)) - return -EINVAL; - - if (!phydev) - return -ENODEV; - - return phy_mii_ioctl(phydev, rq, cmd); -} - static void dnet_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { @@ -759,7 +746,7 @@ static const struct net_device_ops dnet_netdev_ops = { .ndo_stop = dnet_close, .ndo_get_stats = dnet_get_stats, .ndo_start_xmit = dnet_start_xmit, - .ndo_do_ioctl = dnet_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, }; diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c index 39eb7d525043..56f59db6ebf2 100644 --- a/drivers/net/ethernet/emulex/benet/be_main.c +++ b/drivers/net/ethernet/emulex/benet/be_main.c @@ -1417,7 +1417,7 @@ drop: return NETDEV_TX_OK; } -static void be_tx_timeout(struct net_device *netdev) +static void be_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct be_adapter *adapter = netdev_priv(netdev); struct device *dev = &adapter->pdev->dev; diff --git a/drivers/net/ethernet/ethoc.c b/drivers/net/ethernet/ethoc.c index c6e74ae0ff0d..a817ca661c1f 100644 --- a/drivers/net/ethernet/ethoc.c +++ b/drivers/net/ethernet/ethoc.c @@ -869,7 +869,7 @@ static int ethoc_change_mtu(struct net_device *dev, int new_mtu) return -ENOSYS; } -static void ethoc_tx_timeout(struct net_device *dev) +static void ethoc_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct ethoc *priv = netdev_priv(dev); u32 pending = ethoc_read(priv, INT_SOURCE); diff --git a/drivers/net/ethernet/faraday/ftgmac100.c b/drivers/net/ethernet/faraday/ftgmac100.c index 8ed85037f021..4572797f00d7 100644 --- a/drivers/net/ethernet/faraday/ftgmac100.c +++ b/drivers/net/ethernet/faraday/ftgmac100.c @@ -1536,16 +1536,7 @@ static int ftgmac100_stop(struct net_device *netdev) return 0; } -/* optional */ -static int ftgmac100_do_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) -{ - if (!netdev->phydev) - return -ENXIO; - - return phy_mii_ioctl(netdev->phydev, ifr, cmd); -} - -static void ftgmac100_tx_timeout(struct net_device *netdev) +static void ftgmac100_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ftgmac100 *priv = netdev_priv(netdev); @@ -1597,7 +1588,7 @@ static const struct net_device_ops ftgmac100_netdev_ops = { .ndo_start_xmit = ftgmac100_hard_start_xmit, .ndo_set_mac_address = ftgmac100_set_mac_addr, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = ftgmac100_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = ftgmac100_tx_timeout, .ndo_set_rx_mode = ftgmac100_set_rx_mode, .ndo_set_features = ftgmac100_set_features, diff --git a/drivers/net/ethernet/fealnx.c b/drivers/net/ethernet/fealnx.c index c24fd56a2c71..84f10970299a 100644 --- a/drivers/net/ethernet/fealnx.c +++ b/drivers/net/ethernet/fealnx.c @@ -428,7 +428,7 @@ static void getlinktype(struct net_device *dev); static void getlinkstatus(struct net_device *dev); static void netdev_timer(struct timer_list *t); static void reset_timer(struct timer_list *t); -static void fealnx_tx_timeout(struct net_device *dev); +static void fealnx_tx_timeout(struct net_device *dev, unsigned int txqueue); static void init_ring(struct net_device *dev); static netdev_tx_t start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t intr_handler(int irq, void *dev_instance); @@ -1191,7 +1191,7 @@ static void reset_timer(struct timer_list *t) } -static void fealnx_tx_timeout(struct net_device *dev) +static void fealnx_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem *ioaddr = np->mem; diff --git a/drivers/net/ethernet/freescale/Makefile b/drivers/net/ethernet/freescale/Makefile index 6a93293d31e0..67c436400352 100644 --- a/drivers/net/ethernet/freescale/Makefile +++ b/drivers/net/ethernet/freescale/Makefile @@ -25,4 +25,5 @@ obj-$(CONFIG_FSL_DPAA_ETH) += dpaa/ obj-$(CONFIG_FSL_DPAA2_ETH) += dpaa2/ obj-$(CONFIG_FSL_ENETC) += enetc/ +obj-$(CONFIG_FSL_ENETC_MDIO) += enetc/ obj-$(CONFIG_FSL_ENETC_VF) += enetc/ diff --git a/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c b/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c index a301f0095223..09dbcd819d84 100644 --- a/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c +++ b/drivers/net/ethernet/freescale/dpaa/dpaa_eth.c @@ -288,7 +288,7 @@ static int dpaa_stop(struct net_device *net_dev) return err; } -static void dpaa_tx_timeout(struct net_device *net_dev) +static void dpaa_tx_timeout(struct net_device *net_dev, unsigned int txqueue) { struct dpaa_percpu_priv *percpu_priv; const struct dpaa_priv *priv; diff --git a/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c b/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c index 6437fe6b9abf..cc1b7f85e433 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c +++ b/drivers/net/ethernet/freescale/dpaa2/dpaa2-ptp.c @@ -27,6 +27,20 @@ static int dpaa2_ptp_enable(struct ptp_clock_info *ptp, mc_dev = to_fsl_mc_device(dev); switch (rq->type) { + case PTP_CLK_REQ_EXTTS: + switch (rq->extts.index) { + case 0: + bit = DPRTC_EVENT_ETS1; + break; + case 1: + bit = DPRTC_EVENT_ETS2; + break; + default: + return -EINVAL; + } + if (on) + extts_clean_up(ptp_qoriq, rq->extts.index, false); + break; case PTP_CLK_REQ_PPS: bit = DPRTC_EVENT_PPS; break; @@ -96,6 +110,12 @@ static irqreturn_t dpaa2_ptp_irq_handler_thread(int irq, void *priv) ptp_clock_event(ptp_qoriq->clock, &event); } + if (status & DPRTC_EVENT_ETS1) + extts_clean_up(ptp_qoriq, 0, true); + + if (status & DPRTC_EVENT_ETS2) + extts_clean_up(ptp_qoriq, 1, true); + err = dprtc_clear_irq_status(mc_dev->mc_io, 0, mc_dev->mc_handle, DPRTC_IRQ_INDEX, status); if (unlikely(err)) { diff --git a/drivers/net/ethernet/freescale/dpaa2/dprtc-cmd.h b/drivers/net/ethernet/freescale/dpaa2/dprtc-cmd.h index 4ac05bfef338..96ffeb948f08 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dprtc-cmd.h +++ b/drivers/net/ethernet/freescale/dpaa2/dprtc-cmd.h @@ -9,9 +9,11 @@ /* Command versioning */ #define DPRTC_CMD_BASE_VERSION 1 +#define DPRTC_CMD_VERSION_2 2 #define DPRTC_CMD_ID_OFFSET 4 #define DPRTC_CMD(id) (((id) << DPRTC_CMD_ID_OFFSET) | DPRTC_CMD_BASE_VERSION) +#define DPRTC_CMD_V2(id) (((id) << DPRTC_CMD_ID_OFFSET) | DPRTC_CMD_VERSION_2) /* Command IDs */ #define DPRTC_CMDID_CLOSE DPRTC_CMD(0x800) @@ -19,7 +21,7 @@ #define DPRTC_CMDID_SET_IRQ_ENABLE DPRTC_CMD(0x012) #define DPRTC_CMDID_GET_IRQ_ENABLE DPRTC_CMD(0x013) -#define DPRTC_CMDID_SET_IRQ_MASK DPRTC_CMD(0x014) +#define DPRTC_CMDID_SET_IRQ_MASK DPRTC_CMD_V2(0x014) #define DPRTC_CMDID_GET_IRQ_MASK DPRTC_CMD(0x015) #define DPRTC_CMDID_GET_IRQ_STATUS DPRTC_CMD(0x016) #define DPRTC_CMDID_CLEAR_IRQ_STATUS DPRTC_CMD(0x017) diff --git a/drivers/net/ethernet/freescale/dpaa2/dprtc.h b/drivers/net/ethernet/freescale/dpaa2/dprtc.h index 311c184e1aef..05c413719e55 100644 --- a/drivers/net/ethernet/freescale/dpaa2/dprtc.h +++ b/drivers/net/ethernet/freescale/dpaa2/dprtc.h @@ -20,6 +20,8 @@ struct fsl_mc_io; #define DPRTC_IRQ_INDEX 0 #define DPRTC_EVENT_PPS 0x08000000 +#define DPRTC_EVENT_ETS1 0x00800000 +#define DPRTC_EVENT_ETS2 0x00400000 int dprtc_open(struct fsl_mc_io *mc_io, u32 cmd_flags, diff --git a/drivers/net/ethernet/freescale/enetc/Kconfig b/drivers/net/ethernet/freescale/enetc/Kconfig index edad4ca46327..fe942de19597 100644 --- a/drivers/net/ethernet/freescale/enetc/Kconfig +++ b/drivers/net/ethernet/freescale/enetc/Kconfig @@ -2,6 +2,7 @@ config FSL_ENETC tristate "ENETC PF driver" depends on PCI && PCI_MSI && (ARCH_LAYERSCAPE || COMPILE_TEST) + select FSL_ENETC_MDIO select PHYLIB help This driver supports NXP ENETC gigabit ethernet controller PCIe diff --git a/drivers/net/ethernet/freescale/enetc/Makefile b/drivers/net/ethernet/freescale/enetc/Makefile index d0db33e5b6b7..74f7ac253b8b 100644 --- a/drivers/net/ethernet/freescale/enetc/Makefile +++ b/drivers/net/ethernet/freescale/enetc/Makefile @@ -3,7 +3,7 @@ common-objs := enetc.o enetc_cbdr.o enetc_ethtool.o obj-$(CONFIG_FSL_ENETC) += fsl-enetc.o -fsl-enetc-y := enetc_pf.o enetc_mdio.o $(common-objs) +fsl-enetc-y := enetc_pf.o $(common-objs) fsl-enetc-$(CONFIG_PCI_IOV) += enetc_msg.o fsl-enetc-$(CONFIG_FSL_ENETC_QOS) += enetc_qos.o diff --git a/drivers/net/ethernet/freescale/enetc/enetc.c b/drivers/net/ethernet/freescale/enetc/enetc.c index 17739906c966..1f79e36116a3 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc.c +++ b/drivers/net/ethernet/freescale/enetc/enetc.c @@ -149,11 +149,21 @@ static int enetc_map_tx_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb, if (enetc_tx_csum(skb, &temp_bd)) flags |= ENETC_TXBD_FLAGS_CSUM | ENETC_TXBD_FLAGS_L4CS; + else if (tx_ring->tsd_enable) + flags |= ENETC_TXBD_FLAGS_TSE | ENETC_TXBD_FLAGS_TXSTART; /* first BD needs frm_len and offload flags set */ temp_bd.frm_len = cpu_to_le16(skb->len); temp_bd.flags = flags; + if (flags & ENETC_TXBD_FLAGS_TSE) { + u32 temp; + + temp = (skb->skb_mstamp_ns >> 5 & ENETC_TXBD_TXSTART_MASK) + | (flags << ENETC_TXBD_FLAGS_OFFSET); + temp_bd.txstart = cpu_to_le32(temp); + } + if (flags & ENETC_TXBD_FLAGS_EX) { u8 e_flags = 0; *txbd = temp_bd; @@ -227,6 +237,8 @@ static int enetc_map_tx_buffs(struct enetc_bdr *tx_ring, struct sk_buff *skb, enetc_bdr_idx_inc(tx_ring, &i); tx_ring->next_to_use = i; + skb_tx_timestamp(skb); + /* let H/W know BD ring has been updated */ enetc_wr_reg(tx_ring->tpir, i); /* includes wmb() */ @@ -1503,6 +1515,8 @@ int enetc_setup_tc(struct net_device *ndev, enum tc_setup_type type, return enetc_setup_tc_taprio(ndev, type_data); case TC_SETUP_QDISC_CBS: return enetc_setup_tc_cbs(ndev, type_data); + case TC_SETUP_QDISC_ETF: + return enetc_setup_tc_txtime(ndev, type_data); default: return -EOPNOTSUPP; } diff --git a/drivers/net/ethernet/freescale/enetc/enetc.h b/drivers/net/ethernet/freescale/enetc/enetc.h index 7ee0da6d0015..dd4a227ffc7a 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc.h +++ b/drivers/net/ethernet/freescale/enetc/enetc.h @@ -72,6 +72,7 @@ struct enetc_bdr { struct enetc_ring_stats stats; dma_addr_t bd_dma_base; + u8 tsd_enable; /* Time specific departure */ } ____cacheline_aligned_in_smp; static inline void enetc_bdr_idx_inc(struct enetc_bdr *bdr, int *i) @@ -256,8 +257,10 @@ int enetc_send_cmd(struct enetc_si *si, struct enetc_cbd *cbd); int enetc_setup_tc_taprio(struct net_device *ndev, void *type_data); void enetc_sched_speed_set(struct net_device *ndev); int enetc_setup_tc_cbs(struct net_device *ndev, void *type_data); +int enetc_setup_tc_txtime(struct net_device *ndev, void *type_data); #else #define enetc_setup_tc_taprio(ndev, type_data) -EOPNOTSUPP #define enetc_sched_speed_set(ndev) (void)0 #define enetc_setup_tc_cbs(ndev, type_data) -EOPNOTSUPP +#define enetc_setup_tc_txtime(ndev, type_data) -EOPNOTSUPP #endif diff --git a/drivers/net/ethernet/freescale/enetc/enetc_ethtool.c b/drivers/net/ethernet/freescale/enetc/enetc_ethtool.c index 880a8ed8bb47..301ee0dde02d 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_ethtool.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_ethtool.c @@ -579,6 +579,7 @@ static int enetc_get_ts_info(struct net_device *ndev, (1 << HWTSTAMP_FILTER_ALL); #else info->so_timestamping = SOF_TIMESTAMPING_RX_SOFTWARE | + SOF_TIMESTAMPING_TX_SOFTWARE | SOF_TIMESTAMPING_SOFTWARE; #endif return 0; diff --git a/drivers/net/ethernet/freescale/enetc/enetc_hw.h b/drivers/net/ethernet/freescale/enetc/enetc_hw.h index 51f543ef37a8..62554f28ce07 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_hw.h +++ b/drivers/net/ethernet/freescale/enetc/enetc_hw.h @@ -200,6 +200,7 @@ enum enetc_bdr_type {TX, RX}; #define ENETC_PFPMR 0x1900 #define ENETC_PFPMR_PMACE BIT(1) #define ENETC_PFPMR_MWLM BIT(0) +#define ENETC_EMDIO_BASE 0x1c00 #define ENETC_PSIUMHFR0(n, err) (((err) ? 0x1d08 : 0x1d00) + (n) * 0x10) #define ENETC_PSIUMHFR1(n) (0x1d04 + (n) * 0x10) #define ENETC_PSIMMHFR0(n, err) (((err) ? 0x1d00 : 0x1d08) + (n) * 0x10) @@ -358,6 +359,7 @@ union enetc_tx_bd { u8 l4_csoff; u8 flags; }; /* default layout */ + __le32 txstart; __le32 lstatus; }; }; @@ -378,11 +380,14 @@ union enetc_tx_bd { }; #define ENETC_TXBD_FLAGS_L4CS BIT(0) +#define ENETC_TXBD_FLAGS_TSE BIT(1) #define ENETC_TXBD_FLAGS_W BIT(2) #define ENETC_TXBD_FLAGS_CSUM BIT(3) +#define ENETC_TXBD_FLAGS_TXSTART BIT(4) #define ENETC_TXBD_FLAGS_EX BIT(6) #define ENETC_TXBD_FLAGS_F BIT(7) - +#define ENETC_TXBD_TXSTART_MASK GENMASK(24, 0) +#define ENETC_TXBD_FLAGS_OFFSET 24 static inline void enetc_clear_tx_bd(union enetc_tx_bd *txbd) { memset(txbd, 0, sizeof(*txbd)); @@ -615,3 +620,7 @@ struct enetc_cbd { /* Port time gating capability register */ #define ENETC_QBV_PTGCAPR_OFFSET 0x11a08 #define ENETC_QBV_MAX_GCL_LEN_MASK GENMASK(15, 0) + +/* Port time specific departure */ +#define ENETC_PTCTSDR(n) (0x1210 + 4 * (n)) +#define ENETC_TSDE BIT(31) diff --git a/drivers/net/ethernet/freescale/enetc/enetc_mdio.c b/drivers/net/ethernet/freescale/enetc/enetc_mdio.c index 149883c8f0b8..48c32a171afa 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_mdio.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_mdio.c @@ -1,41 +1,56 @@ // SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) /* Copyright 2019 NXP */ +#include <linux/fsl/enetc_mdio.h> #include <linux/mdio.h> #include <linux/of_mdio.h> #include <linux/iopoll.h> #include <linux/of.h> -#include "enetc_mdio.h" +#include "enetc_pf.h" -#define ENETC_MDIO_REG_OFFSET 0x1c00 #define ENETC_MDIO_CFG 0x0 /* MDIO configuration and status */ #define ENETC_MDIO_CTL 0x4 /* MDIO control */ #define ENETC_MDIO_DATA 0x8 /* MDIO data */ #define ENETC_MDIO_ADDR 0xc /* MDIO address */ -#define enetc_mdio_rd(hw, off) \ - enetc_port_rd(hw, ENETC_##off + ENETC_MDIO_REG_OFFSET) -#define enetc_mdio_wr(hw, off, val) \ - enetc_port_wr(hw, ENETC_##off + ENETC_MDIO_REG_OFFSET, val) -#define enetc_mdio_rd_reg(off) enetc_mdio_rd(hw, off) +static inline u32 _enetc_mdio_rd(struct enetc_mdio_priv *mdio_priv, int off) +{ + return enetc_port_rd(mdio_priv->hw, mdio_priv->mdio_base + off); +} + +static inline void _enetc_mdio_wr(struct enetc_mdio_priv *mdio_priv, int off, + u32 val) +{ + enetc_port_wr(mdio_priv->hw, mdio_priv->mdio_base + off, val); +} -#define ENETC_MDC_DIV 258 +#define enetc_mdio_rd(mdio_priv, off) \ + _enetc_mdio_rd(mdio_priv, ENETC_##off) +#define enetc_mdio_wr(mdio_priv, off, val) \ + _enetc_mdio_wr(mdio_priv, ENETC_##off, val) +#define enetc_mdio_rd_reg(off) enetc_mdio_rd(mdio_priv, off) #define MDIO_CFG_CLKDIV(x) ((((x) >> 1) & 0xff) << 8) #define MDIO_CFG_BSY BIT(0) #define MDIO_CFG_RD_ER BIT(1) +#define MDIO_CFG_HOLD(x) (((x) << 2) & GENMASK(4, 2)) #define MDIO_CFG_ENC45 BIT(6) /* external MDIO only - driven on neg MDC edge */ #define MDIO_CFG_NEG BIT(23) +#define ENETC_EMDIO_CFG \ + (MDIO_CFG_HOLD(2) | \ + MDIO_CFG_CLKDIV(258) | \ + MDIO_CFG_NEG) + #define MDIO_CTL_DEV_ADDR(x) ((x) & 0x1f) #define MDIO_CTL_PORT_ADDR(x) (((x) & 0x1f) << 5) #define MDIO_CTL_READ BIT(15) #define MDIO_DATA(x) ((x) & 0xffff) #define TIMEOUT 1000 -static int enetc_mdio_wait_complete(struct enetc_hw *hw) +static int enetc_mdio_wait_complete(struct enetc_mdio_priv *mdio_priv) { u32 val; @@ -46,12 +61,11 @@ static int enetc_mdio_wait_complete(struct enetc_hw *hw) int enetc_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 value) { struct enetc_mdio_priv *mdio_priv = bus->priv; - struct enetc_hw *hw = mdio_priv->hw; u32 mdio_ctl, mdio_cfg; u16 dev_addr; int ret; - mdio_cfg = MDIO_CFG_CLKDIV(ENETC_MDC_DIV) | MDIO_CFG_NEG; + mdio_cfg = ENETC_EMDIO_CFG; if (regnum & MII_ADDR_C45) { dev_addr = (regnum >> 16) & 0x1f; mdio_cfg |= MDIO_CFG_ENC45; @@ -61,44 +75,44 @@ int enetc_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 value) mdio_cfg &= ~MDIO_CFG_ENC45; } - enetc_mdio_wr(hw, MDIO_CFG, mdio_cfg); + enetc_mdio_wr(mdio_priv, MDIO_CFG, mdio_cfg); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; /* set port and dev addr */ mdio_ctl = MDIO_CTL_PORT_ADDR(phy_id) | MDIO_CTL_DEV_ADDR(dev_addr); - enetc_mdio_wr(hw, MDIO_CTL, mdio_ctl); + enetc_mdio_wr(mdio_priv, MDIO_CTL, mdio_ctl); /* set the register address */ if (regnum & MII_ADDR_C45) { - enetc_mdio_wr(hw, MDIO_ADDR, regnum & 0xffff); + enetc_mdio_wr(mdio_priv, MDIO_ADDR, regnum & 0xffff); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; } /* write the value */ - enetc_mdio_wr(hw, MDIO_DATA, MDIO_DATA(value)); + enetc_mdio_wr(mdio_priv, MDIO_DATA, MDIO_DATA(value)); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; return 0; } +EXPORT_SYMBOL_GPL(enetc_mdio_write); int enetc_mdio_read(struct mii_bus *bus, int phy_id, int regnum) { struct enetc_mdio_priv *mdio_priv = bus->priv; - struct enetc_hw *hw = mdio_priv->hw; u32 mdio_ctl, mdio_cfg; u16 dev_addr, value; int ret; - mdio_cfg = MDIO_CFG_CLKDIV(ENETC_MDC_DIV) | MDIO_CFG_NEG; + mdio_cfg = ENETC_EMDIO_CFG; if (regnum & MII_ADDR_C45) { dev_addr = (regnum >> 16) & 0x1f; mdio_cfg |= MDIO_CFG_ENC45; @@ -107,86 +121,56 @@ int enetc_mdio_read(struct mii_bus *bus, int phy_id, int regnum) mdio_cfg &= ~MDIO_CFG_ENC45; } - enetc_mdio_wr(hw, MDIO_CFG, mdio_cfg); + enetc_mdio_wr(mdio_priv, MDIO_CFG, mdio_cfg); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; /* set port and device addr */ mdio_ctl = MDIO_CTL_PORT_ADDR(phy_id) | MDIO_CTL_DEV_ADDR(dev_addr); - enetc_mdio_wr(hw, MDIO_CTL, mdio_ctl); + enetc_mdio_wr(mdio_priv, MDIO_CTL, mdio_ctl); /* set the register address */ if (regnum & MII_ADDR_C45) { - enetc_mdio_wr(hw, MDIO_ADDR, regnum & 0xffff); + enetc_mdio_wr(mdio_priv, MDIO_ADDR, regnum & 0xffff); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; } /* initiate the read */ - enetc_mdio_wr(hw, MDIO_CTL, mdio_ctl | MDIO_CTL_READ); + enetc_mdio_wr(mdio_priv, MDIO_CTL, mdio_ctl | MDIO_CTL_READ); - ret = enetc_mdio_wait_complete(hw); + ret = enetc_mdio_wait_complete(mdio_priv); if (ret) return ret; /* return all Fs if nothing was there */ - if (enetc_mdio_rd(hw, MDIO_CFG) & MDIO_CFG_RD_ER) { + if (enetc_mdio_rd(mdio_priv, MDIO_CFG) & MDIO_CFG_RD_ER) { dev_dbg(&bus->dev, "Error while reading PHY%d reg at %d.%hhu\n", phy_id, dev_addr, regnum); return 0xffff; } - value = enetc_mdio_rd(hw, MDIO_DATA) & 0xffff; + value = enetc_mdio_rd(mdio_priv, MDIO_DATA) & 0xffff; return value; } +EXPORT_SYMBOL_GPL(enetc_mdio_read); -int enetc_mdio_probe(struct enetc_pf *pf) +struct enetc_hw *enetc_hw_alloc(struct device *dev, void __iomem *port_regs) { - struct device *dev = &pf->si->pdev->dev; - struct enetc_mdio_priv *mdio_priv; - struct device_node *np; - struct mii_bus *bus; - int err; - - bus = devm_mdiobus_alloc_size(dev, sizeof(*mdio_priv)); - if (!bus) - return -ENOMEM; - - bus->name = "Freescale ENETC MDIO Bus"; - bus->read = enetc_mdio_read; - bus->write = enetc_mdio_write; - bus->parent = dev; - mdio_priv = bus->priv; - mdio_priv->hw = &pf->si->hw; - snprintf(bus->id, MII_BUS_ID_SIZE, "%s", dev_name(dev)); - - np = of_get_child_by_name(dev->of_node, "mdio"); - if (!np) { - dev_err(dev, "MDIO node missing\n"); - return -EINVAL; - } - - err = of_mdiobus_register(bus, np); - if (err) { - of_node_put(np); - dev_err(dev, "cannot register MDIO bus\n"); - return err; - } + struct enetc_hw *hw; - of_node_put(np); - pf->mdio = bus; + hw = devm_kzalloc(dev, sizeof(*hw), GFP_KERNEL); + if (!hw) + return ERR_PTR(-ENOMEM); - return 0; -} + hw->port = port_regs; -void enetc_mdio_remove(struct enetc_pf *pf) -{ - if (pf->mdio) - mdiobus_unregister(pf->mdio); + return hw; } +EXPORT_SYMBOL_GPL(enetc_hw_alloc); diff --git a/drivers/net/ethernet/freescale/enetc/enetc_mdio.h b/drivers/net/ethernet/freescale/enetc/enetc_mdio.h deleted file mode 100644 index 60c9a3889824..000000000000 --- a/drivers/net/ethernet/freescale/enetc/enetc_mdio.h +++ /dev/null @@ -1,12 +0,0 @@ -/* SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) */ -/* Copyright 2019 NXP */ - -#include <linux/phy.h> -#include "enetc_pf.h" - -struct enetc_mdio_priv { - struct enetc_hw *hw; -}; - -int enetc_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 value); -int enetc_mdio_read(struct mii_bus *bus, int phy_id, int regnum); diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pci_mdio.c b/drivers/net/ethernet/freescale/enetc/enetc_pci_mdio.c index fbd41ce01f06..ebc635f8a4cc 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pci_mdio.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_pci_mdio.c @@ -1,7 +1,8 @@ // SPDX-License-Identifier: (GPL-2.0+ OR BSD-3-Clause) /* Copyright 2019 NXP */ +#include <linux/fsl/enetc_mdio.h> #include <linux/of_mdio.h> -#include "enetc_mdio.h" +#include "enetc_pf.h" #define ENETC_MDIO_DEV_ID 0xee01 #define ENETC_MDIO_DEV_NAME "FSL PCIe IE Central MDIO" @@ -13,17 +14,29 @@ static int enetc_pci_mdio_probe(struct pci_dev *pdev, { struct enetc_mdio_priv *mdio_priv; struct device *dev = &pdev->dev; + void __iomem *port_regs; struct enetc_hw *hw; struct mii_bus *bus; int err; - hw = devm_kzalloc(dev, sizeof(*hw), GFP_KERNEL); - if (!hw) - return -ENOMEM; + port_regs = pci_iomap(pdev, 0, 0); + if (!port_regs) { + dev_err(dev, "iomap failed\n"); + err = -ENXIO; + goto err_ioremap; + } + + hw = enetc_hw_alloc(dev, port_regs); + if (IS_ERR(hw)) { + err = PTR_ERR(hw); + goto err_hw_alloc; + } bus = devm_mdiobus_alloc_size(dev, sizeof(*mdio_priv)); - if (!bus) - return -ENOMEM; + if (!bus) { + err = -ENOMEM; + goto err_mdiobus_alloc; + } bus->name = ENETC_MDIO_BUS_NAME; bus->read = enetc_mdio_read; @@ -31,13 +44,14 @@ static int enetc_pci_mdio_probe(struct pci_dev *pdev, bus->parent = dev; mdio_priv = bus->priv; mdio_priv->hw = hw; + mdio_priv->mdio_base = ENETC_EMDIO_BASE; snprintf(bus->id, MII_BUS_ID_SIZE, "%s", dev_name(dev)); pcie_flr(pdev); err = pci_enable_device_mem(pdev); if (err) { dev_err(dev, "device enable failed\n"); - return err; + goto err_pci_enable; } err = pci_request_region(pdev, 0, KBUILD_MODNAME); @@ -46,13 +60,6 @@ static int enetc_pci_mdio_probe(struct pci_dev *pdev, goto err_pci_mem_reg; } - hw->port = pci_iomap(pdev, 0, 0); - if (!hw->port) { - err = -ENXIO; - dev_err(dev, "iomap failed\n"); - goto err_ioremap; - } - err = of_mdiobus_register(bus, dev->of_node); if (err) goto err_mdiobus_reg; @@ -62,12 +69,14 @@ static int enetc_pci_mdio_probe(struct pci_dev *pdev, return 0; err_mdiobus_reg: - iounmap(mdio_priv->hw->port); -err_ioremap: pci_release_mem_regions(pdev); err_pci_mem_reg: pci_disable_device(pdev); - +err_pci_enable: +err_mdiobus_alloc: + iounmap(port_regs); +err_hw_alloc: +err_ioremap: return err; } diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.c b/drivers/net/ethernet/freescale/enetc/enetc_pf.c index e7482d483b28..fc0d7d99e9a1 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pf.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.c @@ -2,6 +2,7 @@ /* Copyright 2017-2019 NXP */ #include <linux/module.h> +#include <linux/fsl/enetc_mdio.h> #include <linux/of_mdio.h> #include <linux/of_net.h> #include "enetc_pf.h" @@ -749,6 +750,52 @@ static void enetc_pf_netdev_setup(struct enetc_si *si, struct net_device *ndev, enetc_get_primary_mac_addr(&si->hw, ndev->dev_addr); } +static int enetc_mdio_probe(struct enetc_pf *pf) +{ + struct device *dev = &pf->si->pdev->dev; + struct enetc_mdio_priv *mdio_priv; + struct device_node *np; + struct mii_bus *bus; + int err; + + bus = devm_mdiobus_alloc_size(dev, sizeof(*mdio_priv)); + if (!bus) + return -ENOMEM; + + bus->name = "Freescale ENETC MDIO Bus"; + bus->read = enetc_mdio_read; + bus->write = enetc_mdio_write; + bus->parent = dev; + mdio_priv = bus->priv; + mdio_priv->hw = &pf->si->hw; + mdio_priv->mdio_base = ENETC_EMDIO_BASE; + snprintf(bus->id, MII_BUS_ID_SIZE, "%s", dev_name(dev)); + + np = of_get_child_by_name(dev->of_node, "mdio"); + if (!np) { + dev_err(dev, "MDIO node missing\n"); + return -EINVAL; + } + + err = of_mdiobus_register(bus, np); + if (err) { + of_node_put(np); + dev_err(dev, "cannot register MDIO bus\n"); + return err; + } + + of_node_put(np); + pf->mdio = bus; + + return 0; +} + +static void enetc_mdio_remove(struct enetc_pf *pf) +{ + if (pf->mdio) + mdiobus_unregister(pf->mdio); +} + static int enetc_of_get_phy(struct enetc_ndev_priv *priv) { struct enetc_pf *pf = enetc_si_priv(priv->si); diff --git a/drivers/net/ethernet/freescale/enetc/enetc_pf.h b/drivers/net/ethernet/freescale/enetc/enetc_pf.h index 10dd1b53bb08..59e65a6f6c3e 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_pf.h +++ b/drivers/net/ethernet/freescale/enetc/enetc_pf.h @@ -49,7 +49,3 @@ struct enetc_pf { int enetc_msg_psi_init(struct enetc_pf *pf); void enetc_msg_psi_free(struct enetc_pf *pf); void enetc_msg_handle_rxmsg(struct enetc_pf *pf, int mbox_id, u16 *status); - -/* MDIO */ -int enetc_mdio_probe(struct enetc_pf *pf); -void enetc_mdio_remove(struct enetc_pf *pf); diff --git a/drivers/net/ethernet/freescale/enetc/enetc_qos.c b/drivers/net/ethernet/freescale/enetc/enetc_qos.c index 2e99438cb1bf..0c6bf3a55a9a 100644 --- a/drivers/net/ethernet/freescale/enetc/enetc_qos.c +++ b/drivers/net/ethernet/freescale/enetc/enetc_qos.c @@ -36,7 +36,6 @@ void enetc_sched_speed_set(struct net_device *ndev) case SPEED_10: default: pspeed = ENETC_PMR_PSPEED_10M; - netdev_err(ndev, "Qbv PSPEED set speed link down.\n"); } priv->speed = speed; @@ -156,6 +155,11 @@ int enetc_setup_tc_taprio(struct net_device *ndev, void *type_data) int err; int i; + /* TSD and Qbv are mutually exclusive in hardware */ + for (i = 0; i < priv->num_tx_rings; i++) + if (priv->tx_ring[i]->tsd_enable) + return -EBUSY; + for (i = 0; i < priv->num_tx_rings; i++) enetc_set_bdr_prio(&priv->si->hw, priv->tx_ring[i]->index, @@ -192,7 +196,6 @@ int enetc_setup_tc_cbs(struct net_device *ndev, void *type_data) u32 hi_credit_bit, hi_credit_reg; u32 max_interference_size; u32 port_frame_max_size; - u32 tc_max_sized_frame; u8 tc = cbs->queue; u8 prio_top, prio_next; int bw_sum = 0; @@ -250,7 +253,7 @@ int enetc_setup_tc_cbs(struct net_device *ndev, void *type_data) return -EINVAL; } - tc_max_sized_frame = enetc_port_rd(&si->hw, ENETC_PTCMSDUR(tc)); + enetc_port_rd(&si->hw, ENETC_PTCMSDUR(tc)); /* For top prio TC, the max_interfrence_size is maxSizedFrame. * @@ -298,3 +301,33 @@ int enetc_setup_tc_cbs(struct net_device *ndev, void *type_data) return 0; } + +int enetc_setup_tc_txtime(struct net_device *ndev, void *type_data) +{ + struct enetc_ndev_priv *priv = netdev_priv(ndev); + struct tc_etf_qopt_offload *qopt = type_data; + u8 tc_nums = netdev_get_num_tc(ndev); + int tc; + + if (!tc_nums) + return -EOPNOTSUPP; + + tc = qopt->queue; + + if (tc < 0 || tc >= priv->num_tx_rings) + return -EINVAL; + + /* Do not support TXSTART and TX CSUM offload simutaniously */ + if (ndev->features & NETIF_F_CSUM_MASK) + return -EBUSY; + + /* TSD and Qbv are mutually exclusive in hardware */ + if (enetc_rd(&priv->si->hw, ENETC_QBV_PTGCR_OFFSET) & ENETC_QBV_TGE) + return -EBUSY; + + priv->tx_ring[tc]->tsd_enable = qopt->enable; + enetc_port_wr(&priv->si->hw, ENETC_PTCTSDR(tc), + qopt->enable ? ENETC_TSDE : 0); + + return 0; +} diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index 9294027e9d90..4432a59904c7 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -1141,7 +1141,7 @@ fec_stop(struct net_device *ndev) static void -fec_timeout(struct net_device *ndev) +fec_timeout(struct net_device *ndev, unsigned int txqueue) { struct fec_enet_private *fep = netdev_priv(ndev); diff --git a/drivers/net/ethernet/freescale/fec_mpc52xx.c b/drivers/net/ethernet/freescale/fec_mpc52xx.c index 30cdb246d020..7a3f066e611d 100644 --- a/drivers/net/ethernet/freescale/fec_mpc52xx.c +++ b/drivers/net/ethernet/freescale/fec_mpc52xx.c @@ -84,7 +84,7 @@ static int debug = -1; /* the above default */ module_param(debug, int, 0); MODULE_PARM_DESC(debug, "debugging messages level"); -static void mpc52xx_fec_tx_timeout(struct net_device *dev) +static void mpc52xx_fec_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mpc52xx_fec_priv *priv = netdev_priv(dev); unsigned long flags; @@ -785,16 +785,6 @@ static const struct ethtool_ops mpc52xx_fec_ethtool_ops = { }; -static int mpc52xx_fec_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - struct phy_device *phydev = dev->phydev; - - if (!phydev) - return -ENOTSUPP; - - return phy_mii_ioctl(phydev, rq, cmd); -} - static const struct net_device_ops mpc52xx_fec_netdev_ops = { .ndo_open = mpc52xx_fec_open, .ndo_stop = mpc52xx_fec_close, @@ -802,7 +792,7 @@ static const struct net_device_ops mpc52xx_fec_netdev_ops = { .ndo_set_rx_mode = mpc52xx_fec_set_multicast_list, .ndo_set_mac_address = mpc52xx_fec_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = mpc52xx_fec_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = mpc52xx_fec_tx_timeout, .ndo_get_stats = mpc52xx_fec_get_stats, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/freescale/fman/mac.c b/drivers/net/ethernet/freescale/fman/mac.c index f0806ace1ae2..55f2122c3217 100644 --- a/drivers/net/ethernet/freescale/fman/mac.c +++ b/drivers/net/ethernet/freescale/fman/mac.c @@ -692,7 +692,7 @@ static int mac_probe(struct platform_device *_of_dev) mac_dev->res = __devm_request_region(dev, fman_get_mem_region(priv->fman), - res.start, res.end + 1 - res.start, + res.start, resource_size(&res), "mac"); if (!mac_dev->res) { dev_err(dev, "__devm_request_mem_region(mac) failed\n"); @@ -701,7 +701,7 @@ static int mac_probe(struct platform_device *_of_dev) } priv->vaddr = devm_ioremap(dev, mac_dev->res->start, - mac_dev->res->end + 1 - mac_dev->res->start); + resource_size(mac_dev->res)); if (!priv->vaddr) { dev_err(dev, "devm_ioremap() failed\n"); err = -EIO; diff --git a/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c b/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c index 3981c06f082f..add61fed33ee 100644 --- a/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c +++ b/drivers/net/ethernet/freescale/fs_enet/fs_enet-main.c @@ -641,7 +641,7 @@ static void fs_timeout_work(struct work_struct *work) netif_wake_queue(dev); } -static void fs_timeout(struct net_device *dev) +static void fs_timeout(struct net_device *dev, unsigned int txqueue) { struct fs_enet_private *fep = netdev_priv(dev); @@ -882,14 +882,6 @@ static const struct ethtool_ops fs_ethtool_ops = { .set_tunable = fs_set_tunable, }; -static int fs_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - extern int fs_mii_connect(struct net_device *dev); extern void fs_mii_disconnect(struct net_device *dev); @@ -907,7 +899,7 @@ static const struct net_device_ops fs_enet_netdev_ops = { .ndo_start_xmit = fs_enet_start_xmit, .ndo_tx_timeout = fs_timeout, .ndo_set_rx_mode = fs_set_multicast_list, - .ndo_do_ioctl = fs_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/freescale/gianfar.c b/drivers/net/ethernet/freescale/gianfar.c index 72868a28b621..f7e5cafe89a9 100644 --- a/drivers/net/ethernet/freescale/gianfar.c +++ b/drivers/net/ethernet/freescale/gianfar.c @@ -2093,7 +2093,7 @@ static void gfar_reset_task(struct work_struct *work) reset_gfar(priv->ndev); } -static void gfar_timeout(struct net_device *dev) +static void gfar_timeout(struct net_device *dev, unsigned int txqueue) { struct gfar_private *priv = netdev_priv(dev); @@ -2205,13 +2205,17 @@ static void gfar_clean_tx_ring(struct gfar_priv_tx_q *tx_queue) skb_dirtytx = tx_queue->skb_dirtytx; while ((skb = tx_queue->tx_skbuff[skb_dirtytx])) { + bool do_tstamp; + + do_tstamp = (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP) && + priv->hwts_tx_en; frags = skb_shinfo(skb)->nr_frags; /* When time stamping, one additional TxBD must be freed. * Also, we need to dma_unmap_single() the TxPAL. */ - if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS)) + if (unlikely(do_tstamp)) nr_txbds = frags + 2; else nr_txbds = frags + 1; @@ -2225,7 +2229,7 @@ static void gfar_clean_tx_ring(struct gfar_priv_tx_q *tx_queue) (lstatus & BD_LENGTH_MASK)) break; - if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS)) { + if (unlikely(do_tstamp)) { next = next_txbd(bdp, base, tx_ring_size); buflen = be16_to_cpu(next->length) + GMAC_FCB_LEN + GMAC_TXPAL_LEN; @@ -2235,7 +2239,7 @@ static void gfar_clean_tx_ring(struct gfar_priv_tx_q *tx_queue) dma_unmap_single(priv->dev, be32_to_cpu(bdp->bufPtr), buflen, DMA_TO_DEVICE); - if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_IN_PROGRESS)) { + if (unlikely(do_tstamp)) { struct skb_shared_hwtstamps shhwtstamps; u64 *ns = (u64 *)(((uintptr_t)skb->data + 0x10) & ~0x7UL); diff --git a/drivers/net/ethernet/freescale/ucc_geth.c b/drivers/net/ethernet/freescale/ucc_geth.c index f839fa94ebdd..0d101c00286f 100644 --- a/drivers/net/ethernet/freescale/ucc_geth.c +++ b/drivers/net/ethernet/freescale/ucc_geth.c @@ -3545,7 +3545,7 @@ static void ucc_geth_timeout_work(struct work_struct *work) * ucc_geth_timeout gets called when a packet has not been * transmitted after a set amount of time. */ -static void ucc_geth_timeout(struct net_device *dev) +static void ucc_geth_timeout(struct net_device *dev, unsigned int txqueue) { struct ucc_geth_private *ugeth = netdev_priv(dev); diff --git a/drivers/net/ethernet/fujitsu/fmvj18x_cs.c b/drivers/net/ethernet/fujitsu/fmvj18x_cs.c index 1eca0fdb9933..a7b7a4aace79 100644 --- a/drivers/net/ethernet/fujitsu/fmvj18x_cs.c +++ b/drivers/net/ethernet/fujitsu/fmvj18x_cs.c @@ -93,7 +93,7 @@ static irqreturn_t fjn_interrupt(int irq, void *dev_id); static void fjn_rx(struct net_device *dev); static void fjn_reset(struct net_device *dev); static void set_rx_mode(struct net_device *dev); -static void fjn_tx_timeout(struct net_device *dev); +static void fjn_tx_timeout(struct net_device *dev, unsigned int txqueue); static const struct ethtool_ops netdev_ethtool_ops; /* @@ -774,7 +774,7 @@ static irqreturn_t fjn_interrupt(int dummy, void *dev_id) /*====================================================================*/ -static void fjn_tx_timeout(struct net_device *dev) +static void fjn_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct local_info *lp = netdev_priv(dev); unsigned int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/google/gve/gve_main.c b/drivers/net/ethernet/google/gve/gve_main.c index 9b7a8db9860f..e032563ceefd 100644 --- a/drivers/net/ethernet/google/gve/gve_main.c +++ b/drivers/net/ethernet/google/gve/gve_main.c @@ -845,7 +845,7 @@ static void gve_turnup(struct gve_priv *priv) gve_set_napi_enabled(priv); } -static void gve_tx_timeout(struct net_device *dev) +static void gve_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct gve_priv *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/hisilicon/hip04_eth.c b/drivers/net/ethernet/hisilicon/hip04_eth.c index 150a8ccfb8b1..d9718b87279d 100644 --- a/drivers/net/ethernet/hisilicon/hip04_eth.c +++ b/drivers/net/ethernet/hisilicon/hip04_eth.c @@ -779,7 +779,7 @@ static int hip04_mac_stop(struct net_device *ndev) return 0; } -static void hip04_timeout(struct net_device *ndev) +static void hip04_timeout(struct net_device *ndev, unsigned int txqueue) { struct hip04_priv *priv = netdev_priv(ndev); diff --git a/drivers/net/ethernet/hisilicon/hisi_femac.c b/drivers/net/ethernet/hisilicon/hisi_femac.c index 90ab7ade44c4..57c3bc4f7089 100644 --- a/drivers/net/ethernet/hisilicon/hisi_femac.c +++ b/drivers/net/ethernet/hisilicon/hisi_femac.c @@ -675,18 +675,6 @@ static void hisi_femac_net_set_rx_mode(struct net_device *dev) } } -static int hisi_femac_net_ioctl(struct net_device *dev, - struct ifreq *ifreq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - - if (!dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, ifreq, cmd); -} - static const struct ethtool_ops hisi_femac_ethtools_ops = { .get_link = ethtool_op_get_link, .get_link_ksettings = phy_ethtool_get_link_ksettings, @@ -697,7 +685,7 @@ static const struct net_device_ops hisi_femac_netdev_ops = { .ndo_open = hisi_femac_net_open, .ndo_stop = hisi_femac_net_close, .ndo_start_xmit = hisi_femac_net_xmit, - .ndo_do_ioctl = hisi_femac_net_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_mac_address = hisi_femac_set_mac_address, .ndo_set_rx_mode = hisi_femac_net_set_rx_mode, }; diff --git a/drivers/net/ethernet/hisilicon/hix5hd2_gmac.c b/drivers/net/ethernet/hisilicon/hix5hd2_gmac.c index 247de9105d10..4fb776920a93 100644 --- a/drivers/net/ethernet/hisilicon/hix5hd2_gmac.c +++ b/drivers/net/ethernet/hisilicon/hix5hd2_gmac.c @@ -893,7 +893,7 @@ static void hix5hd2_tx_timeout_task(struct work_struct *work) hix5hd2_net_open(priv->netdev); } -static void hix5hd2_net_timeout(struct net_device *dev) +static void hix5hd2_net_timeout(struct net_device *dev, unsigned int txqueue) { struct hix5hd2_priv *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/hisilicon/hns/hns_enet.c b/drivers/net/ethernet/hisilicon/hns/hns_enet.c index eb69e5c81a4d..c117074c16e3 100644 --- a/drivers/net/ethernet/hisilicon/hns/hns_enet.c +++ b/drivers/net/ethernet/hisilicon/hns/hns_enet.c @@ -1483,7 +1483,7 @@ static int hns_nic_net_stop(struct net_device *ndev) static void hns_tx_timeout_reset(struct hns_nic_priv *priv); #define HNS_TX_TIMEO_LIMIT (40 * HZ) -static void hns_nic_net_timeout(struct net_device *ndev) +static void hns_nic_net_timeout(struct net_device *ndev, unsigned int txqueue) { struct hns_nic_priv *priv = netdev_priv(ndev); @@ -1497,20 +1497,6 @@ static void hns_nic_net_timeout(struct net_device *ndev) } } -static int hns_nic_do_ioctl(struct net_device *netdev, struct ifreq *ifr, - int cmd) -{ - struct phy_device *phy_dev = netdev->phydev; - - if (!netif_running(netdev)) - return -EINVAL; - - if (!phy_dev) - return -ENOTSUPP; - - return phy_mii_ioctl(phy_dev, ifr, cmd); -} - static netdev_tx_t hns_nic_net_xmit(struct sk_buff *skb, struct net_device *ndev) { @@ -1958,7 +1944,7 @@ static const struct net_device_ops hns_nic_netdev_ops = { .ndo_tx_timeout = hns_nic_net_timeout, .ndo_set_mac_address = hns_nic_net_set_mac_address, .ndo_change_mtu = hns_nic_change_mtu, - .ndo_do_ioctl = hns_nic_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_features = hns_nic_set_features, .ndo_fix_features = hns_nic_fix_features, .ndo_get_stats64 = hns_nic_get_stats64, diff --git a/drivers/net/ethernet/hisilicon/hns3/Makefile b/drivers/net/ethernet/hisilicon/hns3/Makefile index d01bf536eb86..7aa2fac76c5e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/Makefile +++ b/drivers/net/ethernet/hisilicon/hns3/Makefile @@ -3,6 +3,8 @@ # Makefile for the HISILICON network device drivers. # +ccflags-y += -I$(srctree)/$(src) + obj-$(CONFIG_HNS3) += hns3pf/ obj-$(CONFIG_HNS3) += hns3vf/ diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h index 3b5e2d7251e7..a3e4081b84ba 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h +++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h @@ -164,11 +164,7 @@ enum hnae3_reset_type { HNAE3_IMP_RESET, HNAE3_UNKNOWN_RESET, HNAE3_NONE_RESET, -}; - -enum hnae3_flr_state { - HNAE3_FLR_DOWN, - HNAE3_FLR_DONE, + HNAE3_MAX_RESET, }; enum hnae3_port_base_vlan_state { @@ -575,8 +571,7 @@ struct hnae3_ae_algo { const struct pci_device_id *pdev_id_table; }; -#define HNAE3_INT_NAME_EXT_LEN 32 /* Max extra information length */ -#define HNAE3_INT_NAME_LEN (IFNAMSIZ + HNAE3_INT_NAME_EXT_LEN) +#define HNAE3_INT_NAME_LEN 32 #define HNAE3_ITR_COUNTDOWN_START 100 struct hnae3_tc_info { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c index 6b328a259efc..1d4ffc5f408a 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_debugfs.c @@ -176,7 +176,7 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf) return -EINVAL; } - ring = &priv->ring[q_num]; + ring = &priv->ring[q_num]; value = readl_relaxed(ring->tqp->io_base + HNS3_RING_TX_RING_TAIL_REG); tx_index = (cnt == 1) ? value : tx_index; @@ -209,10 +209,10 @@ static int hns3_dbg_bd_info(struct hnae3_handle *h, const char *cmd_buf) le16_to_cpu(tx_desc->tx.bdtp_fe_sc_vld_ra_ri)); dev_info(dev, "(TX)mss: %u\n", le16_to_cpu(tx_desc->tx.mss)); - ring = &priv->ring[q_num + h->kinfo.num_tqps]; + ring = &priv->ring[q_num + h->kinfo.num_tqps]; value = readl_relaxed(ring->tqp->io_base + HNS3_RING_RX_RING_TAIL_REG); rx_index = (cnt == 1) ? value : tx_index; - rx_desc = &ring->desc[rx_index]; + rx_desc = &ring->desc[rx_index]; addr = le64_to_cpu(rx_desc->addr); dev_info(dev, "RX Queue Num: %u, BD Index: %u\n", q_num, rx_index); @@ -297,8 +297,8 @@ static ssize_t hns3_dbg_cmd_read(struct file *filp, char __user *buffer, if (!buf) return -ENOMEM; - len = snprintf(buf, HNS3_DBG_READ_LEN, "%s\n", - "Please echo help to cmd to get help information"); + len = scnprintf(buf, HNS3_DBG_READ_LEN, "%s\n", + "Please echo help to cmd to get help information"); uncopy_bytes = copy_to_user(buffer, buf, len); kfree(buf); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c index b3deb5e5ce29..acb796cc10d0 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c @@ -24,6 +24,12 @@ #include "hnae3.h" #include "hns3_enet.h" +/* All hns3 tracepoints are defined by the include below, which + * must be included exactly once across the whole kernel with + * CREATE_TRACE_POINTS defined + */ +#define CREATE_TRACE_POINTS +#include "hns3_trace.h" #define hns3_set_field(origin, shift, val) ((origin) |= ((val) << (shift))) #define hns3_tx_bd_count(S) DIV_ROUND_UP(S, HNS3_MAX_BD_SIZE) @@ -129,18 +135,21 @@ static int hns3_nic_init_irq(struct hns3_nic_priv *priv) continue; if (tqp_vectors->tx_group.ring && tqp_vectors->rx_group.ring) { - snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1, - "%s-%s-%d", priv->netdev->name, "TxRx", - txrx_int_idx++); + snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN, + "%s-%s-%s-%d", hns3_driver_name, + pci_name(priv->ae_handle->pdev), + "TxRx", txrx_int_idx++); txrx_int_idx++; } else if (tqp_vectors->rx_group.ring) { - snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1, - "%s-%s-%d", priv->netdev->name, "Rx", - rx_int_idx++); + snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN, + "%s-%s-%s-%d", hns3_driver_name, + pci_name(priv->ae_handle->pdev), + "Rx", rx_int_idx++); } else if (tqp_vectors->tx_group.ring) { - snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN - 1, - "%s-%s-%d", priv->netdev->name, "Tx", - tx_int_idx++); + snprintf(tqp_vectors->name, HNAE3_INT_NAME_LEN, + "%s-%s-%s-%d", hns3_driver_name, + pci_name(priv->ae_handle->pdev), + "Tx", tx_int_idx++); } else { /* Skip this unused q_vector */ continue; @@ -157,6 +166,8 @@ static int hns3_nic_init_irq(struct hns3_nic_priv *priv) return ret; } + disable_irq(tqp_vectors->vector_irq); + irq_set_affinity_hint(tqp_vectors->vector_irq, &tqp_vectors->affinity_mask); @@ -175,6 +186,7 @@ static void hns3_mask_vector_irq(struct hns3_enet_tqp_vector *tqp_vector, static void hns3_vector_enable(struct hns3_enet_tqp_vector *tqp_vector) { napi_enable(&tqp_vector->napi); + enable_irq(tqp_vector->vector_irq); /* enable vector */ hns3_mask_vector_irq(tqp_vector, 1); @@ -374,18 +386,6 @@ static int hns3_nic_net_up(struct net_device *netdev) if (ret) return ret; - /* the device can work without cpu rmap, only aRFS needs it */ - ret = hns3_set_rx_cpu_rmap(netdev); - if (ret) - netdev_warn(netdev, "set rx cpu rmap fail, ret=%d!\n", ret); - - /* get irq resource for all vectors */ - ret = hns3_nic_init_irq(priv); - if (ret) { - netdev_err(netdev, "init irq failed! ret=%d\n", ret); - goto free_rmap; - } - clear_bit(HNS3_NIC_STATE_DOWN, &priv->state); /* enable the vectors */ @@ -398,22 +398,15 @@ static int hns3_nic_net_up(struct net_device *netdev) /* start the ae_dev */ ret = h->ae_algo->ops->start ? h->ae_algo->ops->start(h) : 0; - if (ret) - goto out_start_err; - - return 0; - -out_start_err: - set_bit(HNS3_NIC_STATE_DOWN, &priv->state); - while (j--) - hns3_tqp_disable(h->kinfo.tqp[j]); + if (ret) { + set_bit(HNS3_NIC_STATE_DOWN, &priv->state); + while (j--) + hns3_tqp_disable(h->kinfo.tqp[j]); - for (j = i - 1; j >= 0; j--) - hns3_vector_disable(&priv->tqp_vector[j]); + for (j = i - 1; j >= 0; j--) + hns3_vector_disable(&priv->tqp_vector[j]); + } - hns3_nic_uninit_irq(priv); -free_rmap: - hns3_free_rx_cpu_rmap(netdev); return ret; } @@ -510,11 +503,6 @@ static void hns3_nic_net_down(struct net_device *netdev) if (ops->stop) ops->stop(priv->ae_handle); - hns3_free_rx_cpu_rmap(netdev); - - /* free irq resources */ - hns3_nic_uninit_irq(priv); - /* delay ring buffer clearing to hns3_reset_notify_uninit_enet * during reset process, because driver may not be able * to disable the ring through firmware when downing the netdev. @@ -736,6 +724,8 @@ static int hns3_set_tso(struct sk_buff *skb, u32 *paylen, /* get MSS for TSO */ *mss = skb_shinfo(skb)->gso_size; + trace_hns3_tso(skb); + return 0; } @@ -1140,6 +1130,7 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv, desc->tx.bdtp_fe_sc_vld_ra_ri = cpu_to_le16(BIT(HNS3_TXD_VLD_B)); + trace_hns3_tx_desc(ring, ring->next_to_use); ring_ptr_move_fw(ring, next_to_use); return HNS3_LIKELY_BD_NUM; } @@ -1163,6 +1154,7 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv, desc->tx.bdtp_fe_sc_vld_ra_ri = cpu_to_le16(BIT(HNS3_TXD_VLD_B)); + trace_hns3_tx_desc(ring, ring->next_to_use); /* move ring pointer to next */ ring_ptr_move_fw(ring, next_to_use); @@ -1288,6 +1280,14 @@ static bool hns3_skb_need_linearized(struct sk_buff *skb, unsigned int *bd_size, return false; } +void hns3_shinfo_pack(struct skb_shared_info *shinfo, __u32 *size) +{ + int i = 0; + + for (i = 0; i < MAX_SKB_FRAGS; i++) + size[i] = skb_frag_size(&shinfo->frags[i]); +} + static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring, struct net_device *netdev, struct sk_buff *skb) @@ -1299,8 +1299,10 @@ static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring, bd_num = hns3_tx_bd_num(skb, bd_size); if (unlikely(bd_num > HNS3_MAX_NON_TSO_BD_NUM)) { if (bd_num <= HNS3_MAX_TSO_BD_NUM && skb_is_gso(skb) && - !hns3_skb_need_linearized(skb, bd_size, bd_num)) + !hns3_skb_need_linearized(skb, bd_size, bd_num)) { + trace_hns3_over_8bd(skb); goto out; + } if (__skb_linearize(skb)) return -ENOMEM; @@ -1308,8 +1310,10 @@ static int hns3_nic_maybe_stop_tx(struct hns3_enet_ring *ring, bd_num = hns3_tx_bd_count(skb->len); if ((skb_is_gso(skb) && bd_num > HNS3_MAX_TSO_BD_NUM) || (!skb_is_gso(skb) && - bd_num > HNS3_MAX_NON_TSO_BD_NUM)) + bd_num > HNS3_MAX_NON_TSO_BD_NUM)) { + trace_hns3_over_8bd(skb); return -ENOMEM; + } u64_stats_update_begin(&ring->syncp); ring->stats.tx_copy++; @@ -1454,6 +1458,7 @@ out: (ring->desc_num - 1); ring->desc[pre_ntu].tx.bdtp_fe_sc_vld_ra_ri |= cpu_to_le16(BIT(HNS3_TXD_FE_B)); + trace_hns3_tx_desc(ring, pre_ntu); /* Complete translate all packets */ dev_queue = netdev_get_tx_queue(netdev, ring->queue_index); @@ -1562,6 +1567,37 @@ static int hns3_nic_set_features(struct net_device *netdev, return 0; } +static netdev_features_t hns3_features_check(struct sk_buff *skb, + struct net_device *dev, + netdev_features_t features) +{ +#define HNS3_MAX_HDR_LEN 480U +#define HNS3_MAX_L4_HDR_LEN 60U + + size_t len; + + if (skb->ip_summed != CHECKSUM_PARTIAL) + return features; + + if (skb->encapsulation) + len = skb_inner_transport_header(skb) - skb->data; + else + len = skb_transport_header(skb) - skb->data; + + /* Assume L4 is 60 byte as TCP is the only protocol with a + * a flexible value, and it's max len is 60 bytes. + */ + len += HNS3_MAX_L4_HDR_LEN; + + /* Hardware only supports checksum on the skb with a max header + * len of 480 bytes. + */ + if (len > HNS3_MAX_HDR_LEN) + features &= ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK); + + return features; +} + static void hns3_nic_get_stats64(struct net_device *netdev, struct rtnl_link_stats64 *stats) { @@ -1875,7 +1911,7 @@ static bool hns3_get_tx_timeo_queue_info(struct net_device *ndev) return true; } -static void hns3_nic_net_timeout(struct net_device *ndev) +static void hns3_nic_net_timeout(struct net_device *ndev, unsigned int txqueue) { struct hns3_nic_priv *priv = netdev_priv(ndev); struct hnae3_handle *h = priv->ae_handle; @@ -1976,6 +2012,7 @@ static const struct net_device_ops hns3_nic_netdev_ops = { .ndo_do_ioctl = hns3_nic_do_ioctl, .ndo_change_mtu = hns3_nic_change_mtu, .ndo_set_features = hns3_nic_set_features, + .ndo_features_check = hns3_features_check, .ndo_get_stats64 = hns3_nic_get_stats64, .ndo_setup_tc = hns3_nic_setup_tc, .ndo_set_rx_mode = hns3_nic_set_rx_mode, @@ -2057,10 +2094,8 @@ static int hns3_probe(struct pci_dev *pdev, const struct pci_device_id *ent) int ret; ae_dev = devm_kzalloc(&pdev->dev, sizeof(*ae_dev), GFP_KERNEL); - if (!ae_dev) { - ret = -ENOMEM; - return ret; - } + if (!ae_dev) + return -ENOMEM; ae_dev->pdev = pdev; ae_dev->flag = ent->driver_data; @@ -2503,8 +2538,8 @@ void hns3_clean_tx_ring(struct hns3_enet_ring *ring) rmb(); /* Make sure head is ready before touch any data */ if (unlikely(!is_valid_clean_head(ring, head))) { - netdev_err(netdev, "wrong head (%d, %d-%d)\n", head, - ring->next_to_use, ring->next_to_clean); + hns3_rl_err(netdev, "wrong head (%d, %d-%d)\n", head, + ring->next_to_use, ring->next_to_clean); u64_stats_update_begin(&ring->syncp); ring->stats.io_err_cnt++; @@ -2590,6 +2625,12 @@ static void hns3_nic_alloc_rx_buffers(struct hns3_enet_ring *ring, writel_relaxed(i, ring->tqp->io_base + HNS3_RING_RX_RING_HEAD_REG); } +static bool hns3_page_is_reusable(struct page *page) +{ + return page_to_nid(page) == numa_mem_id() && + !page_is_pfmemalloc(page); +} + static void hns3_nic_reuse_page(struct sk_buff *skb, int i, struct hns3_enet_ring *ring, int pull_len, struct hns3_desc_cb *desc_cb) @@ -2604,7 +2645,7 @@ static void hns3_nic_reuse_page(struct sk_buff *skb, int i, /* Avoid re-using remote pages, or the stack is still using the page * when page_offset rollback to zero, flag default unreuse */ - if (unlikely(page_to_nid(desc_cb->priv) != numa_mem_id()) || + if (unlikely(!hns3_page_is_reusable(desc_cb->priv)) || (!desc_cb->page_offset && page_count(desc_cb->priv) > 1)) return; @@ -2674,6 +2715,9 @@ static int hns3_gro_complete(struct sk_buff *skb, u32 l234info) skb->csum_start = (unsigned char *)th - skb->head; skb->csum_offset = offsetof(struct tcphdr, check); skb->ip_summed = CHECKSUM_PARTIAL; + + trace_hns3_gro(skb); + return 0; } @@ -2794,7 +2838,6 @@ static bool hns3_parse_vlan_tag(struct hns3_enet_ring *ring, static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, unsigned char *va) { -#define HNS3_NEED_ADD_FRAG 1 struct hns3_desc_cb *desc_cb = &ring->desc_cb[ring->next_to_clean]; struct net_device *netdev = ring_to_netdev(ring); struct sk_buff *skb; @@ -2811,6 +2854,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, return -ENOMEM; } + trace_hns3_rx_desc(ring); prefetchw(skb->data); ring->pending_buf = 1; @@ -2820,7 +2864,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, memcpy(__skb_put(skb, length), va, ALIGN(length, sizeof(long))); /* We can reuse buffer as-is, just make sure it is local */ - if (likely(page_to_nid(desc_cb->priv) == numa_mem_id())) + if (likely(hns3_page_is_reusable(desc_cb->priv))) desc_cb->reuse_flag = 1; else /* This page cannot be reused so discard it */ put_page(desc_cb->priv); @@ -2838,33 +2882,19 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, desc_cb); ring_ptr_move_fw(ring, next_to_clean); - return HNS3_NEED_ADD_FRAG; + return 0; } -static int hns3_add_frag(struct hns3_enet_ring *ring, struct hns3_desc *desc, - bool pending) +static int hns3_add_frag(struct hns3_enet_ring *ring) { struct sk_buff *skb = ring->skb; struct sk_buff *head_skb = skb; struct sk_buff *new_skb; struct hns3_desc_cb *desc_cb; - struct hns3_desc *pre_desc; + struct hns3_desc *desc; u32 bd_base_info; - int pre_bd; - /* if there is pending bd, the SW param next_to_clean has moved - * to next and the next is NULL - */ - if (pending) { - pre_bd = (ring->next_to_clean - 1 + ring->desc_num) % - ring->desc_num; - pre_desc = &ring->desc[pre_bd]; - bd_base_info = le32_to_cpu(pre_desc->rx.bd_base_info); - } else { - bd_base_info = le32_to_cpu(desc->rx.bd_base_info); - } - - while (!(bd_base_info & BIT(HNS3_RXD_FE_B))) { + do { desc = &ring->desc[ring->next_to_clean]; desc_cb = &ring->desc_cb[ring->next_to_clean]; bd_base_info = le32_to_cpu(desc->rx.bd_base_info); @@ -2899,9 +2929,10 @@ static int hns3_add_frag(struct hns3_enet_ring *ring, struct hns3_desc *desc, } hns3_nic_reuse_page(skb, ring->frag_num++, ring, 0, desc_cb); + trace_hns3_rx_desc(ring); ring_ptr_move_fw(ring, next_to_clean); ring->pending_buf++; - } + } while (!(bd_base_info & BIT(HNS3_RXD_FE_B))); return 0; } @@ -3069,28 +3100,23 @@ static int hns3_handle_rx_bd(struct hns3_enet_ring *ring) if (ret < 0) /* alloc buffer fail */ return ret; - if (ret > 0) { /* need add frag */ - ret = hns3_add_frag(ring, desc, false); + if (!(bd_base_info & BIT(HNS3_RXD_FE_B))) { /* need add frag */ + ret = hns3_add_frag(ring); if (ret) return ret; - - /* As the head data may be changed when GRO enable, copy - * the head data in after other data rx completed - */ - memcpy(skb->data, ring->va, - ALIGN(ring->pull_len, sizeof(long))); } } else { - ret = hns3_add_frag(ring, desc, true); + ret = hns3_add_frag(ring); if (ret) return ret; + } - /* As the head data may be changed when GRO enable, copy - * the head data in after other data rx completed - */ + /* As the head data may be changed when GRO enable, copy + * the head data in after other data rx completed + */ + if (skb->len > HNS3_RX_HEAD_SIZE) memcpy(skb->data, ring->va, ALIGN(ring->pull_len, sizeof(long))); - } ret = hns3_handle_bdinfo(ring, skb); if (unlikely(ret)) { @@ -3596,26 +3622,25 @@ static void hns3_nic_uninit_vector_data(struct hns3_nic_priv *priv) if (!tqp_vector->rx_group.ring && !tqp_vector->tx_group.ring) continue; - hns3_get_vector_ring_chain(tqp_vector, &vector_ring_chain); + /* Since the mapping can be overwritten, when fail to get the + * chain between vector and ring, we should go on to deal with + * the remaining options. + */ + if (hns3_get_vector_ring_chain(tqp_vector, &vector_ring_chain)) + dev_warn(priv->dev, "failed to get ring chain\n"); h->ae_algo->ops->unmap_ring_from_vector(h, tqp_vector->vector_irq, &vector_ring_chain); hns3_free_vector_ring_chain(tqp_vector, &vector_ring_chain); - if (tqp_vector->irq_init_flag == HNS3_VECTOR_INITED) { - irq_set_affinity_hint(tqp_vector->vector_irq, NULL); - free_irq(tqp_vector->vector_irq, tqp_vector); - tqp_vector->irq_init_flag = HNS3_VECTOR_NOT_INITED; - } - hns3_clear_ring_group(&tqp_vector->rx_group); hns3_clear_ring_group(&tqp_vector->tx_group); netif_napi_del(&priv->tqp_vector[i].napi); } } -static int hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv) +static void hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv) { struct hnae3_handle *h = priv->ae_handle; struct pci_dev *pdev = h->pdev; @@ -3627,11 +3652,10 @@ static int hns3_nic_dealloc_vector_data(struct hns3_nic_priv *priv) tqp_vector = &priv->tqp_vector[i]; ret = h->ae_algo->ops->put_vector(h, tqp_vector->vector_irq); if (ret) - return ret; + return; } devm_kfree(&pdev->dev, priv->tqp_vector); - return 0; } static void hns3_ring_get_cfg(struct hnae3_queue *q, struct hns3_nic_priv *priv, @@ -4030,6 +4054,18 @@ static int hns3_client_init(struct hnae3_handle *handle) goto out_reg_netdev_fail; } + /* the device can work without cpu rmap, only aRFS needs it */ + ret = hns3_set_rx_cpu_rmap(netdev); + if (ret) + dev_warn(priv->dev, "set rx cpu rmap fail, ret=%d\n", ret); + + ret = hns3_nic_init_irq(priv); + if (ret) { + dev_err(priv->dev, "init irq failed! ret=%d\n", ret); + hns3_free_rx_cpu_rmap(netdev); + goto out_init_irq_fail; + } + ret = hns3_client_start(handle); if (ret) { dev_err(priv->dev, "hns3_client_start fail! ret=%d\n", ret); @@ -4051,6 +4087,9 @@ static int hns3_client_init(struct hnae3_handle *handle) return ret; out_client_start: + hns3_free_rx_cpu_rmap(netdev); + hns3_nic_uninit_irq(priv); +out_init_irq_fail: unregister_netdev(netdev); out_reg_netdev_fail: hns3_uninit_phy(netdev); @@ -4088,15 +4127,17 @@ static void hns3_client_uninit(struct hnae3_handle *handle, bool reset) goto out_netdev_free; } + hns3_free_rx_cpu_rmap(netdev); + + hns3_nic_uninit_irq(priv); + hns3_del_all_fd_rules(netdev, true); hns3_clear_all_ring(handle, true); hns3_nic_uninit_vector_data(priv); - ret = hns3_nic_dealloc_vector_data(priv); - if (ret) - netdev_err(netdev, "dealloc vector error\n"); + hns3_nic_dealloc_vector_data(priv); ret = hns3_uninit_all_ring(priv); if (ret) @@ -4423,17 +4464,32 @@ static int hns3_reset_notify_init_enet(struct hnae3_handle *handle) if (ret) goto err_uninit_vector; + /* the device can work without cpu rmap, only aRFS needs it */ + ret = hns3_set_rx_cpu_rmap(netdev); + if (ret) + dev_warn(priv->dev, "set rx cpu rmap fail, ret=%d\n", ret); + + ret = hns3_nic_init_irq(priv); + if (ret) { + dev_err(priv->dev, "init irq failed! ret=%d\n", ret); + hns3_free_rx_cpu_rmap(netdev); + goto err_init_irq_fail; + } + ret = hns3_client_start(handle); if (ret) { dev_err(priv->dev, "hns3_client_start fail! ret=%d\n", ret); - goto err_uninit_ring; + goto err_client_start_fail; } set_bit(HNS3_NIC_STATE_INITED, &priv->state); return ret; -err_uninit_ring: +err_client_start_fail: + hns3_free_rx_cpu_rmap(netdev); + hns3_nic_uninit_irq(priv); +err_init_irq_fail: hns3_uninit_all_ring(priv); err_uninit_vector: hns3_nic_uninit_vector_data(priv); @@ -4483,6 +4539,8 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle) return 0; } + hns3_free_rx_cpu_rmap(netdev); + hns3_nic_uninit_irq(priv); hns3_clear_all_ring(handle, true); hns3_reset_tx_queue(priv->ae_handle); @@ -4490,9 +4548,7 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle) hns3_store_coal(priv); - ret = hns3_nic_dealloc_vector_data(priv); - if (ret) - netdev_err(netdev, "dealloc vector error\n"); + hns3_nic_dealloc_vector_data(priv); ret = hns3_uninit_all_ring(priv); if (ret) @@ -4658,7 +4714,7 @@ static int __init hns3_init_module(void) pr_info("%s: %s\n", hns3_driver_name, hns3_copyright); client.type = HNAE3_CLIENT_KNIC; - snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH - 1, "%s", + snprintf(client.name, HNAE3_CLIENT_NAME_LENGTH, "%s", hns3_driver_name); client.ops = &client_ops; diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h index 9d47abd5c37c..abefd7a179f7 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h @@ -673,4 +673,5 @@ void hns3_dbg_init(struct hnae3_handle *handle); void hns3_dbg_uninit(struct hnae3_handle *handle); void hns3_dbg_register_debugfs(const char *debugfs_dir_name); void hns3_dbg_unregister_debugfs(void); +void hns3_shinfo_pack(struct skb_shared_info *shinfo, __u32 *size); #endif diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c index 6e0212b79438..c03856e63320 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c @@ -423,9 +423,8 @@ static void *hns3_update_strings(u8 *data, const struct hns3_stats *stats, data[ETH_GSTRING_LEN - 1] = '\0'; /* first, prepend the prefix string */ - n1 = snprintf(data, MAX_PREFIX_SIZE, "%s%d_", - prefix, i); - n1 = min_t(uint, n1, MAX_PREFIX_SIZE - 1); + n1 = scnprintf(data, MAX_PREFIX_SIZE, "%s%d_", + prefix, i); size_left = (ETH_GSTRING_LEN - 1) - n1; /* now, concatenate the stats string to it */ diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h b/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h new file mode 100644 index 000000000000..7bddcca148a5 --- /dev/null +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_trace.h @@ -0,0 +1,139 @@ +/* SPDX-License-Identifier: GPL-2.0+ */ +/* Copyright (c) 2018-2019 Hisilicon Limited. */ + +/* This must be outside ifdef _HNS3_TRACE_H */ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM hns3 + +#if !defined(_HNS3_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ) +#define _HNS3_TRACE_H_ + +#include <linux/tracepoint.h> + +#define DESC_NR (sizeof(struct hns3_desc) / sizeof(u32)) + +DECLARE_EVENT_CLASS(hns3_skb_template, + TP_PROTO(struct sk_buff *skb), + TP_ARGS(skb), + + TP_STRUCT__entry( + __field(unsigned int, headlen) + __field(unsigned int, len) + __field(__u8, nr_frags) + __field(__u8, ip_summed) + __field(unsigned int, hdr_len) + __field(unsigned short, gso_size) + __field(unsigned short, gso_segs) + __field(unsigned int, gso_type) + __field(bool, fraglist) + __array(__u32, size, MAX_SKB_FRAGS) + ), + + TP_fast_assign( + __entry->headlen = skb_headlen(skb); + __entry->len = skb->len; + __entry->nr_frags = skb_shinfo(skb)->nr_frags; + __entry->gso_size = skb_shinfo(skb)->gso_size; + __entry->gso_segs = skb_shinfo(skb)->gso_segs; + __entry->gso_type = skb_shinfo(skb)->gso_type; + __entry->hdr_len = skb->encapsulation ? + skb_inner_transport_offset(skb) + inner_tcp_hdrlen(skb) : + skb_transport_offset(skb) + tcp_hdrlen(skb); + __entry->ip_summed = skb->ip_summed; + __entry->fraglist = skb_has_frag_list(skb); + hns3_shinfo_pack(skb_shinfo(skb), __entry->size); + ), + + TP_printk( + "len: %u, %u, %u, cs: %u, gso: %u, %u, %x, frag(%d %u): %s", + __entry->headlen, __entry->len, __entry->hdr_len, + __entry->ip_summed, __entry->gso_size, __entry->gso_segs, + __entry->gso_type, __entry->fraglist, __entry->nr_frags, + __print_array(__entry->size, MAX_SKB_FRAGS, sizeof(__u32)) + ) +); + +DEFINE_EVENT(hns3_skb_template, hns3_over_8bd, + TP_PROTO(struct sk_buff *skb), + TP_ARGS(skb)); + +DEFINE_EVENT(hns3_skb_template, hns3_gro, + TP_PROTO(struct sk_buff *skb), + TP_ARGS(skb)); + +DEFINE_EVENT(hns3_skb_template, hns3_tso, + TP_PROTO(struct sk_buff *skb), + TP_ARGS(skb)); + +TRACE_EVENT(hns3_tx_desc, + TP_PROTO(struct hns3_enet_ring *ring, int cur_ntu), + TP_ARGS(ring, cur_ntu), + + TP_STRUCT__entry( + __field(int, index) + __field(int, ntu) + __field(int, ntc) + __field(dma_addr_t, desc_dma) + __array(u32, desc, DESC_NR) + __string(devname, ring->tqp->handle->kinfo.netdev->name) + ), + + TP_fast_assign( + __entry->index = ring->tqp->tqp_index; + __entry->ntu = ring->next_to_use; + __entry->ntc = ring->next_to_clean; + __entry->desc_dma = ring->desc_dma_addr, + memcpy(__entry->desc, &ring->desc[cur_ntu], + sizeof(struct hns3_desc)); + __assign_str(devname, ring->tqp->handle->kinfo.netdev->name); + ), + + TP_printk( + "%s-%d-%d/%d desc(%pad): %s", + __get_str(devname), __entry->index, __entry->ntu, + __entry->ntc, &__entry->desc_dma, + __print_array(__entry->desc, DESC_NR, sizeof(u32)) + ) +); + +TRACE_EVENT(hns3_rx_desc, + TP_PROTO(struct hns3_enet_ring *ring), + TP_ARGS(ring), + + TP_STRUCT__entry( + __field(int, index) + __field(int, ntu) + __field(int, ntc) + __field(dma_addr_t, desc_dma) + __field(dma_addr_t, buf_dma) + __array(u32, desc, DESC_NR) + __string(devname, ring->tqp->handle->kinfo.netdev->name) + ), + + TP_fast_assign( + __entry->index = ring->tqp->tqp_index; + __entry->ntu = ring->next_to_use; + __entry->ntc = ring->next_to_clean; + __entry->desc_dma = ring->desc_dma_addr; + __entry->buf_dma = ring->desc_cb[ring->next_to_clean].dma; + memcpy(__entry->desc, &ring->desc[ring->next_to_clean], + sizeof(struct hns3_desc)); + __assign_str(devname, ring->tqp->handle->kinfo.netdev->name); + ), + + TP_printk( + "%s-%d-%d/%d desc(%pad) buf(%pad): %s", + __get_str(devname), __entry->index, __entry->ntu, + __entry->ntc, &__entry->desc_dma, &__entry->buf_dma, + __print_array(__entry->desc, DESC_NR, sizeof(u32)) + ) +); + +#endif /* _HNS3_TRACE_H_ */ + +/* This must be outside ifdef _HNS3_TRACE_H */ +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_FILE hns3_trace +#include <trace/define_trace.h> diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c index 940ead3970d1..7f509eff562e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c @@ -479,19 +479,6 @@ static void hclge_cmd_uninit_regs(struct hclge_hw *hw) hclge_write_dev(hw, HCLGE_NIC_CRQ_TAIL_REG, 0); } -static void hclge_destroy_queue(struct hclge_cmq_ring *ring) -{ - spin_lock(&ring->lock); - hclge_free_cmd_desc(ring); - spin_unlock(&ring->lock); -} - -static void hclge_destroy_cmd_queue(struct hclge_hw *hw) -{ - hclge_destroy_queue(&hw->cmq.csq); - hclge_destroy_queue(&hw->cmq.crq); -} - void hclge_cmd_uninit(struct hclge_dev *hdev) { spin_lock_bh(&hdev->hw.cmq.csq.lock); @@ -501,5 +488,6 @@ void hclge_cmd_uninit(struct hclge_dev *hdev) spin_unlock(&hdev->hw.cmq.crq.lock); spin_unlock_bh(&hdev->hw.cmq.csq.lock); - hclge_destroy_cmd_queue(&hdev->hw); + hclge_free_cmd_desc(&hdev->hw.cmq.csq); + hclge_free_cmd_desc(&hdev->hw.cmq.crq); } diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h index d97da67f07a1..96498d9b4754 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h @@ -6,6 +6,7 @@ #include <linux/types.h> #include <linux/io.h> #include <linux/etherdevice.h> +#include "hnae3.h" #define HCLGE_CMDQ_TX_TIMEOUT 30000 #define HCLGE_DESC_DATA_LEN 6 @@ -63,6 +64,7 @@ enum hclge_cmd_status { struct hclge_misc_vector { u8 __iomem *addr; int vector_irq; + char name[HNAE3_INT_NAME_LEN]; }; struct hclge_cmq { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c index 112df34b3869..67fad80035d3 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c @@ -73,8 +73,6 @@ static struct hclge_dbg_reg_type_info hclge_dbg_reg_info[] = { static int hclge_dbg_get_dfx_bd_num(struct hclge_dev *hdev, int offset) { -#define HCLGE_GET_DFX_REG_TYPE_CNT 4 - struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT]; int entries_per_desc; int index; @@ -886,8 +884,8 @@ static void hclge_dbg_dump_mng_table(struct hclge_dev *hdev) } } -static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage, - bool sel_x, u32 loc) +static int hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage, + bool sel_x, u32 loc) { struct hclge_fd_tcam_config_1_cmd *req1; struct hclge_fd_tcam_config_2_cmd *req2; @@ -912,7 +910,7 @@ static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage, ret = hclge_cmd_send(&hdev->hw, desc, 3); if (ret) - return; + return ret; dev_info(&hdev->pdev->dev, " read result tcam key %s(%u):\n", sel_x ? "x" : "y", loc); @@ -931,16 +929,76 @@ static void hclge_dbg_fd_tcam_read(struct hclge_dev *hdev, u8 stage, req = (u32 *)req3->tcam_data; for (i = 0; i < 5; i++) dev_info(&hdev->pdev->dev, "%08x\n", *req++); + + return ret; +} + +static int hclge_dbg_get_rules_location(struct hclge_dev *hdev, u16 *rule_locs) +{ + struct hclge_fd_rule *rule; + struct hlist_node *node; + int cnt = 0; + + spin_lock_bh(&hdev->fd_rule_lock); + hlist_for_each_entry_safe(rule, node, &hdev->fd_rule_list, rule_node) { + rule_locs[cnt] = rule->location; + cnt++; + } + spin_unlock_bh(&hdev->fd_rule_lock); + + if (cnt != hdev->hclge_fd_rule_num) + return -EINVAL; + + return cnt; } static void hclge_dbg_fd_tcam(struct hclge_dev *hdev) { - u32 i; + int i, ret, rule_cnt; + u16 *rule_locs; + + if (!hnae3_dev_fd_supported(hdev)) { + dev_err(&hdev->pdev->dev, + "Only FD-supported dev supports dump fd tcam\n"); + return; + } + + if (!hdev->hclge_fd_rule_num || + !hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1]) + return; - for (i = 0; i < hdev->fd_cfg.rule_num[0]; i++) { - hclge_dbg_fd_tcam_read(hdev, 0, true, i); - hclge_dbg_fd_tcam_read(hdev, 0, false, i); + rule_locs = kcalloc(hdev->fd_cfg.rule_num[HCLGE_FD_STAGE_1], + sizeof(u16), GFP_KERNEL); + if (!rule_locs) + return; + + rule_cnt = hclge_dbg_get_rules_location(hdev, rule_locs); + if (rule_cnt <= 0) { + dev_err(&hdev->pdev->dev, + "failed to get rule number, ret = %d\n", rule_cnt); + kfree(rule_locs); + return; } + + for (i = 0; i < rule_cnt; i++) { + ret = hclge_dbg_fd_tcam_read(hdev, 0, true, rule_locs[i]); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to get fd tcam key x, ret = %d\n", ret); + kfree(rule_locs); + return; + } + + ret = hclge_dbg_fd_tcam_read(hdev, 0, false, rule_locs[i]); + if (ret) { + dev_err(&hdev->pdev->dev, + "failed to get fd tcam key y, ret = %d\n", ret); + kfree(rule_locs); + return; + } + } + + kfree(rule_locs); } void hclge_dbg_dump_rst_info(struct hclge_dev *hdev) @@ -976,6 +1034,14 @@ void hclge_dbg_dump_rst_info(struct hclge_dev *hdev) dev_info(&hdev->pdev->dev, "hdev state: 0x%lx\n", hdev->state); } +static void hclge_dbg_dump_serv_info(struct hclge_dev *hdev) +{ + dev_info(&hdev->pdev->dev, "last_serv_processed: %lu\n", + hdev->last_serv_processed); + dev_info(&hdev->pdev->dev, "last_serv_cnt: %lu\n", + hdev->serv_processed_cnt); +} + static void hclge_dbg_get_m7_stats_info(struct hclge_dev *hdev) { struct hclge_desc *desc_src, *desc_tmp; @@ -1227,6 +1293,8 @@ int hclge_dbg_run_cmd(struct hnae3_handle *handle, const char *cmd_buf) hclge_dbg_dump_reg_cmd(hdev, &cmd_buf[sizeof(DUMP_REG)]); } else if (strncmp(cmd_buf, "dump reset info", 15) == 0) { hclge_dbg_dump_rst_info(hdev); + } else if (strncmp(cmd_buf, "dump serv info", 14) == 0) { + hclge_dbg_dump_serv_info(hdev); } else if (strncmp(cmd_buf, "dump m7 info", 12) == 0) { hclge_dbg_get_m7_stats_info(hdev); } else if (strncmp(cmd_buf, "dump ncl_config", 15) == 0) { diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c index dc66b4e13377..c85b72dc44d2 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_err.c @@ -505,7 +505,7 @@ static const struct hclge_hw_error hclge_ssu_mem_ecc_err_int[] = { static const struct hclge_hw_error hclge_ssu_port_based_err_int[] = { { .int_msk = BIT(0), .msg = "roc_pkt_without_key_port", - .reset_level = HNAE3_GLOBAL_RESET }, + .reset_level = HNAE3_FUNC_RESET }, { .int_msk = BIT(1), .msg = "tpu_pkt_without_key_port", .reset_level = HNAE3_GLOBAL_RESET }, { .int_msk = BIT(2), .msg = "igu_pkt_without_key_port", @@ -599,7 +599,7 @@ static const struct hclge_hw_error hclge_ssu_ets_tcg_int[] = { static const struct hclge_hw_error hclge_ssu_port_based_pf_int[] = { { .int_msk = BIT(0), .msg = "roc_pkt_without_key_port", - .reset_level = HNAE3_GLOBAL_RESET }, + .reset_level = HNAE3_FUNC_RESET }, { .int_msk = BIT(9), .msg = "low_water_line_err_port", .reset_level = HNAE3_NONE_RESET }, { .int_msk = BIT(10), .msg = "hi_water_line_err_port", @@ -1898,10 +1898,8 @@ static int hclge_handle_all_hw_msix_error(struct hclge_dev *hdev, bd_num = max_t(u32, mpf_bd_num, pf_bd_num); desc = kcalloc(bd_num, sizeof(struct hclge_desc), GFP_KERNEL); - if (!desc) { - ret = -ENOMEM; - goto out; - } + if (!desc) + return -ENOMEM; ret = hclge_handle_mpf_msix_error(hdev, desc, mpf_bd_num, reset_requests); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c index 13dbd249f35f..ec5f6eeb639b 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c @@ -72,6 +72,8 @@ static int hclge_set_default_loopback(struct hclge_dev *hdev); static struct hnae3_ae_algo ae_algo; +static struct workqueue_struct *hclge_wq; + static const struct pci_device_id ae_algo_pci_tbl[] = { {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_GE), 0}, {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_25GE), 0}, @@ -416,7 +418,7 @@ static int hclge_mac_update_stats_defective(struct hclge_dev *hdev) { #define HCLGE_MAC_CMD_NUM 21 - u64 *data = (u64 *)(&hdev->hw_stats.mac_stats); + u64 *data = (u64 *)(&hdev->mac_stats); struct hclge_desc desc[HCLGE_MAC_CMD_NUM]; __le64 *desc_data; int i, k, n; @@ -453,7 +455,7 @@ static int hclge_mac_update_stats_defective(struct hclge_dev *hdev) static int hclge_mac_update_stats_complete(struct hclge_dev *hdev, u32 desc_num) { - u64 *data = (u64 *)(&hdev->hw_stats.mac_stats); + u64 *data = (u64 *)(&hdev->mac_stats); struct hclge_desc *desc; __le64 *desc_data; u16 i, k, n; @@ -802,7 +804,7 @@ static void hclge_get_stats(struct hnae3_handle *handle, u64 *data) struct hclge_dev *hdev = vport->back; u64 *p; - p = hclge_comm_get_stats(&hdev->hw_stats.mac_stats, g_mac_stats_string, + p = hclge_comm_get_stats(&hdev->mac_stats, g_mac_stats_string, ARRAY_SIZE(g_mac_stats_string), data); p = hclge_tqps_get_stats(handle, p); } @@ -815,8 +817,8 @@ static void hclge_get_mac_stat(struct hnae3_handle *handle, hclge_update_stats(handle, NULL); - mac_stats->tx_pause_cnt = hdev->hw_stats.mac_stats.mac_tx_mac_pause_num; - mac_stats->rx_pause_cnt = hdev->hw_stats.mac_stats.mac_rx_mac_pause_num; + mac_stats->tx_pause_cnt = hdev->mac_stats.mac_tx_mac_pause_num; + mac_stats->rx_pause_cnt = hdev->mac_stats.mac_rx_mac_pause_num; } static int hclge_parse_func_status(struct hclge_dev *hdev, @@ -860,9 +862,7 @@ static int hclge_query_function_status(struct hclge_dev *hdev) usleep_range(1000, 2000); } while (timeout++ < HCLGE_QUERY_MAX_CNT); - ret = hclge_parse_func_status(hdev, req); - - return ret; + return hclge_parse_func_status(hdev, req); } static int hclge_query_pf_resource(struct hclge_dev *hdev) @@ -880,12 +880,12 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) } req = (struct hclge_pf_res_cmd *)desc.data; - hdev->num_tqps = __le16_to_cpu(req->tqp_num); - hdev->pkt_buf_size = __le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S; + hdev->num_tqps = le16_to_cpu(req->tqp_num); + hdev->pkt_buf_size = le16_to_cpu(req->buf_size) << HCLGE_BUF_UNIT_S; if (req->tx_buf_size) hdev->tx_buf_size = - __le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S; + le16_to_cpu(req->tx_buf_size) << HCLGE_BUF_UNIT_S; else hdev->tx_buf_size = HCLGE_DEFAULT_TX_BUF; @@ -893,7 +893,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) if (req->dv_buf_size) hdev->dv_buf_size = - __le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S; + le16_to_cpu(req->dv_buf_size) << HCLGE_BUF_UNIT_S; else hdev->dv_buf_size = HCLGE_DEFAULT_DV; @@ -901,10 +901,10 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) if (hnae3_dev_roce_supported(hdev)) { hdev->roce_base_msix_offset = - hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee), + hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee), HCLGE_MSIX_OFT_ROCEE_M, HCLGE_MSIX_OFT_ROCEE_S); hdev->num_roce_msi = - hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number), HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S); /* nic's msix numbers is always equals to the roce's. */ @@ -917,7 +917,7 @@ static int hclge_query_pf_resource(struct hclge_dev *hdev) hdev->roce_base_msix_offset; } else { hdev->num_msi = - hnae3_get_field(__le16_to_cpu(req->pf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->pf_intr_vector_number), HCLGE_PF_VEC_NUM_M, HCLGE_PF_VEC_NUM_S); hdev->num_nic_msi = hdev->num_msi; @@ -1331,11 +1331,7 @@ static int hclge_get_cap(struct hclge_dev *hdev) } /* get pf resource */ - ret = hclge_query_pf_resource(hdev); - if (ret) - dev_err(&hdev->pdev->dev, "query pf resource error %d.\n", ret); - - return ret; + return hclge_query_pf_resource(hdev); } static void hclge_init_kdump_kernel_config(struct hclge_dev *hdev) @@ -2619,30 +2615,21 @@ static int hclge_mac_init(struct hclge_dev *hdev) hdev->hw.mac.duplex = HCLGE_MAC_FULL; ret = hclge_cfg_mac_speed_dup_hw(hdev, hdev->hw.mac.speed, hdev->hw.mac.duplex); - if (ret) { - dev_err(&hdev->pdev->dev, - "Config mac speed dup fail ret=%d\n", ret); + if (ret) return ret; - } if (hdev->hw.mac.support_autoneg) { ret = hclge_set_autoneg_en(hdev, hdev->hw.mac.autoneg); - if (ret) { - dev_err(&hdev->pdev->dev, - "Config mac autoneg fail ret=%d\n", ret); + if (ret) return ret; - } } mac->link = 0; if (mac->user_fec_mode & BIT(HNAE3_FEC_USER_DEF)) { ret = hclge_set_fec_hw(hdev, mac->user_fec_mode); - if (ret) { - dev_err(&hdev->pdev->dev, - "Fec mode init fail, ret = %d\n", ret); + if (ret) return ret; - } } ret = hclge_set_mac_mtu(hdev, hdev->mps); @@ -2665,31 +2652,27 @@ static int hclge_mac_init(struct hclge_dev *hdev) static void hclge_mbx_task_schedule(struct hclge_dev *hdev) { - if (!test_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state) && + if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) && !test_and_set_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state)) - queue_work_on(cpumask_first(&hdev->affinity_mask), system_wq, - &hdev->mbx_service_task); + mod_delayed_work_on(cpumask_first(&hdev->affinity_mask), + hclge_wq, &hdev->service_task, 0); } static void hclge_reset_task_schedule(struct hclge_dev *hdev) { if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) && !test_and_set_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state)) - queue_work_on(cpumask_first(&hdev->affinity_mask), system_wq, - &hdev->rst_service_task); + mod_delayed_work_on(cpumask_first(&hdev->affinity_mask), + hclge_wq, &hdev->service_task, 0); } void hclge_task_schedule(struct hclge_dev *hdev, unsigned long delay_time) { - if (!test_bit(HCLGE_STATE_DOWN, &hdev->state) && - !test_bit(HCLGE_STATE_REMOVING, &hdev->state) && - !test_and_set_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state)) { - hdev->hw_stats.stats_timer++; - hdev->fd_arfs_expire_timer++; + if (!test_bit(HCLGE_STATE_REMOVING, &hdev->state) && + !test_bit(HCLGE_STATE_RST_FAIL, &hdev->state)) mod_delayed_work_on(cpumask_first(&hdev->affinity_mask), - system_wq, &hdev->service_task, + hclge_wq, &hdev->service_task, delay_time); - } } static int hclge_get_mac_link_status(struct hclge_dev *hdev) @@ -2748,6 +2731,10 @@ static void hclge_update_link_status(struct hclge_dev *hdev) if (!client) return; + + if (test_and_set_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state)) + return; + state = hclge_get_mac_phy_link(hdev); if (state != hdev->hw.mac.link) { for (i = 0; i < hdev->num_vmdq_vport + 1; i++) { @@ -2761,6 +2748,8 @@ static void hclge_update_link_status(struct hclge_dev *hdev) } hdev->hw.mac.link = state; } + + clear_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state); } static void hclge_update_port_capability(struct hclge_mac *mac) @@ -2831,6 +2820,12 @@ static int hclge_get_sfp_info(struct hclge_dev *hdev, struct hclge_mac *mac) return ret; } + /* In some case, mac speed get from IMP may be 0, it shouldn't be + * set to mac->speed. + */ + if (!le32_to_cpu(resp->speed)) + return 0; + mac->speed = le32_to_cpu(resp->speed); /* if resp->speed_ability is 0, it means it's an old version * firmware, do not update these params @@ -2906,7 +2901,7 @@ static int hclge_get_status(struct hnae3_handle *handle) static struct hclge_vport *hclge_get_vf_vport(struct hclge_dev *hdev, int vf) { - if (pci_num_vf(hdev->pdev) == 0) { + if (!pci_num_vf(hdev->pdev)) { dev_err(&hdev->pdev->dev, "SRIOV is disabled, can not get vport(%d) info.\n", vf); return NULL; @@ -2940,6 +2935,9 @@ static int hclge_get_vf_config(struct hnae3_handle *handle, int vf, ivf->trusted = vport->vf_info.trusted; ivf->min_tx_rate = 0; ivf->max_tx_rate = vport->vf_info.max_tx_rate; + ivf->vlan = vport->port_base_vlan_cfg.vlan_info.vlan_tag; + ivf->vlan_proto = htons(vport->port_base_vlan_cfg.vlan_info.vlan_proto); + ivf->qos = vport->port_base_vlan_cfg.vlan_info.qos; ether_addr_copy(ivf->mac, vport->vf_info.mac); return 0; @@ -2998,8 +2996,6 @@ static u32 hclge_check_event_cause(struct hclge_dev *hdev, u32 *clearval) /* check for vector0 msix event source */ if (msix_src_reg & HCLGE_VECTOR0_REG_MSIX_MASK) { - dev_info(&hdev->pdev->dev, "received event 0x%x\n", - msix_src_reg); *clearval = msix_src_reg; return HCLGE_VECTOR0_EVENT_ERR; } @@ -3172,8 +3168,10 @@ static int hclge_misc_irq_init(struct hclge_dev *hdev) hclge_get_misc_vector(hdev); /* this would be explicitly freed in the end */ + snprintf(hdev->misc_vector.name, HNAE3_INT_NAME_LEN, "%s-misc-%s", + HCLGE_NAME, pci_name(hdev->pdev)); ret = request_irq(hdev->misc_vector.vector_irq, hclge_misc_irq_handle, - 0, "hclge_misc", hdev); + 0, hdev->misc_vector.name, hdev); if (ret) { hclge_free_vector(hdev, 0); dev_err(&hdev->pdev->dev, "request misc irq(%d) fail\n", @@ -3247,7 +3245,8 @@ static int hclge_notify_roce_client(struct hclge_dev *hdev, static int hclge_reset_wait(struct hclge_dev *hdev) { #define HCLGE_RESET_WATI_MS 100 -#define HCLGE_RESET_WAIT_CNT 200 +#define HCLGE_RESET_WAIT_CNT 350 + u32 val, reg, reg_bit; u32 cnt = 0; @@ -3264,8 +3263,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev) reg = HCLGE_FUN_RST_ING; reg_bit = HCLGE_FUN_RST_ING_B; break; - case HNAE3_FLR_RESET: - break; default: dev_err(&hdev->pdev->dev, "Wait for unsupported reset type: %d\n", @@ -3273,20 +3270,6 @@ static int hclge_reset_wait(struct hclge_dev *hdev) return -EINVAL; } - if (hdev->reset_type == HNAE3_FLR_RESET) { - while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) && - cnt++ < HCLGE_RESET_WAIT_CNT) - msleep(HCLGE_RESET_WATI_MS); - - if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) { - dev_err(&hdev->pdev->dev, - "flr wait timeout: %u\n", cnt); - return -EBUSY; - } - - return 0; - } - val = hclge_read_dev(&hdev->hw, reg); while (hnae3_get_bit(val, reg_bit) && cnt < HCLGE_RESET_WAIT_CNT) { msleep(HCLGE_RESET_WATI_MS); @@ -3352,7 +3335,19 @@ static int hclge_set_all_vf_rst(struct hclge_dev *hdev, bool reset) return 0; } -static int hclge_func_reset_sync_vf(struct hclge_dev *hdev) +static void hclge_mailbox_service_task(struct hclge_dev *hdev) +{ + if (!test_and_clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state) || + test_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state) || + test_and_set_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state)) + return; + + hclge_mbx_handler(hdev); + + clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state); +} + +static void hclge_func_reset_sync_vf(struct hclge_dev *hdev) { struct hclge_pf_rst_sync_cmd *req; struct hclge_desc desc; @@ -3363,26 +3358,28 @@ static int hclge_func_reset_sync_vf(struct hclge_dev *hdev) hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_QUERY_VF_RST_RDY, true); do { + /* vf need to down netdev by mbx during PF or FLR reset */ + hclge_mailbox_service_task(hdev); + ret = hclge_cmd_send(&hdev->hw, &desc, 1); /* for compatible with old firmware, wait * 100 ms for VF to stop IO */ if (ret == -EOPNOTSUPP) { msleep(HCLGE_RESET_SYNC_TIME); - return 0; + return; } else if (ret) { - dev_err(&hdev->pdev->dev, "sync with VF fail %d!\n", - ret); - return ret; + dev_warn(&hdev->pdev->dev, "sync with VF fail %d!\n", + ret); + return; } else if (req->all_vf_ready) { - return 0; + return; } msleep(HCLGE_PF_RESET_SYNC_TIME); hclge_cmd_reuse_desc(&desc, true); } while (cnt++ < HCLGE_PF_RESET_SYNC_CNT); - dev_err(&hdev->pdev->dev, "sync with VF timeout!\n"); - return -ETIME; + dev_warn(&hdev->pdev->dev, "sync with VF timeout!\n"); } void hclge_report_hw_error(struct hclge_dev *hdev, @@ -3462,12 +3459,6 @@ static void hclge_do_reset(struct hclge_dev *hdev) set_bit(HNAE3_FUNC_RESET, &hdev->reset_pending); hclge_reset_task_schedule(hdev); break; - case HNAE3_FLR_RESET: - dev_info(&pdev->dev, "FLR requested\n"); - /* schedule again to check later */ - set_bit(HNAE3_FLR_RESET, &hdev->reset_pending); - hclge_reset_task_schedule(hdev); - break; default: dev_warn(&pdev->dev, "Unsupported reset type: %d\n", hdev->reset_type); @@ -3483,10 +3474,15 @@ static enum hnae3_reset_type hclge_get_reset_level(struct hnae3_ae_dev *ae_dev, /* first, resolve any unknown reset type to the known type(s) */ if (test_bit(HNAE3_UNKNOWN_RESET, addr)) { + u32 msix_sts_reg = hclge_read_dev(&hdev->hw, + HCLGE_VECTOR0_PF_OTHER_INT_STS_REG); /* we will intentionally ignore any errors from this function * as we will end up in *some* reset request in any case */ - hclge_handle_hw_msix_error(hdev, addr); + if (hclge_handle_hw_msix_error(hdev, addr)) + dev_info(&hdev->pdev->dev, "received msix interrupt 0x%x\n", + msix_sts_reg); + clear_bit(HNAE3_UNKNOWN_RESET, addr); /* We defered the clearing of the error event which caused * interrupt since it was not posssible to do that in @@ -3551,23 +3547,6 @@ static void hclge_clear_reset_cause(struct hclge_dev *hdev) hclge_enable_vector(&hdev->misc_vector, true); } -static int hclge_reset_prepare_down(struct hclge_dev *hdev) -{ - int ret = 0; - - switch (hdev->reset_type) { - case HNAE3_FUNC_RESET: - /* fall through */ - case HNAE3_FLR_RESET: - ret = hclge_set_all_vf_rst(hdev, true); - break; - default: - break; - } - - return ret; -} - static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable) { u32 reg_val; @@ -3581,6 +3560,19 @@ static void hclge_reset_handshake(struct hclge_dev *hdev, bool enable) hclge_write_dev(&hdev->hw, HCLGE_NIC_CSQ_DEPTH_REG, reg_val); } +static int hclge_func_reset_notify_vf(struct hclge_dev *hdev) +{ + int ret; + + ret = hclge_set_all_vf_rst(hdev, true); + if (ret) + return ret; + + hclge_func_reset_sync_vf(hdev); + + return 0; +} + static int hclge_reset_prepare_wait(struct hclge_dev *hdev) { u32 reg_val; @@ -3588,10 +3580,7 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev) switch (hdev->reset_type) { case HNAE3_FUNC_RESET: - /* to confirm whether all running VF is ready - * before request PF reset - */ - ret = hclge_func_reset_sync_vf(hdev); + ret = hclge_func_reset_notify_vf(hdev); if (ret) return ret; @@ -3611,16 +3600,9 @@ static int hclge_reset_prepare_wait(struct hclge_dev *hdev) hdev->rst_stats.pf_rst_cnt++; break; case HNAE3_FLR_RESET: - /* to confirm whether all running VF is ready - * before request PF reset - */ - ret = hclge_func_reset_sync_vf(hdev); + ret = hclge_func_reset_notify_vf(hdev); if (ret) return ret; - - set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state); - set_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - hdev->rst_stats.flr_rst_cnt++; break; case HNAE3_IMP_RESET: hclge_handle_imp_error(hdev); @@ -3672,6 +3654,8 @@ static bool hclge_reset_err_handle(struct hclge_dev *hdev) hclge_dbg_dump_rst_info(hdev); + set_bit(HCLGE_STATE_RST_FAIL, &hdev->state); + return false; } @@ -3747,10 +3731,9 @@ static int hclge_reset_stack(struct hclge_dev *hdev) return hclge_notify_client(hdev, HNAE3_RESTORE_CLIENT); } -static void hclge_reset(struct hclge_dev *hdev) +static int hclge_reset_prepare(struct hclge_dev *hdev) { struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); - enum hnae3_reset_type reset_level; int ret; /* Initialize ae_dev reset status as well, in case enet layer wants to @@ -3761,45 +3744,41 @@ static void hclge_reset(struct hclge_dev *hdev) /* perform reset of the stack & ae device for a client */ ret = hclge_notify_roce_client(hdev, HNAE3_DOWN_CLIENT); if (ret) - goto err_reset; - - ret = hclge_reset_prepare_down(hdev); - if (ret) - goto err_reset; + return ret; rtnl_lock(); ret = hclge_notify_client(hdev, HNAE3_DOWN_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); - - ret = hclge_reset_prepare_wait(hdev); if (ret) - goto err_reset; + return ret; - if (hclge_reset_wait(hdev)) - goto err_reset; + return hclge_reset_prepare_wait(hdev); +} + +static int hclge_reset_rebuild(struct hclge_dev *hdev) +{ + struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); + enum hnae3_reset_type reset_level; + int ret; hdev->rst_stats.hw_reset_done_cnt++; ret = hclge_notify_roce_client(hdev, HNAE3_UNINIT_CLIENT); if (ret) - goto err_reset; + return ret; rtnl_lock(); - ret = hclge_reset_stack(hdev); + rtnl_unlock(); if (ret) - goto err_reset_lock; + return ret; hclge_clear_reset_cause(hdev); ret = hclge_reset_prepare_up(hdev); if (ret) - goto err_reset_lock; + return ret; - rtnl_unlock(); ret = hclge_notify_roce_client(hdev, HNAE3_INIT_CLIENT); /* ignore RoCE notify error if it fails HCLGE_RESET_MAX_FAIL_CNT - 1 @@ -3807,24 +3786,23 @@ static void hclge_reset(struct hclge_dev *hdev) */ if (ret && hdev->rst_stats.reset_fail_cnt < HCLGE_RESET_MAX_FAIL_CNT - 1) - goto err_reset; + return ret; rtnl_lock(); - ret = hclge_notify_client(hdev, HNAE3_UP_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); + if (ret) + return ret; ret = hclge_notify_roce_client(hdev, HNAE3_UP_CLIENT); if (ret) - goto err_reset; + return ret; hdev->last_reset_time = jiffies; hdev->rst_stats.reset_fail_cnt = 0; hdev->rst_stats.reset_done_cnt++; ae_dev->reset_type = HNAE3_NONE_RESET; + clear_bit(HCLGE_STATE_RST_FAIL, &hdev->state); /* if default_reset_request has a higher level reset request, * it should be handled as soon as possible. since some errors @@ -3835,10 +3813,22 @@ static void hclge_reset(struct hclge_dev *hdev) if (reset_level != HNAE3_NONE_RESET) set_bit(reset_level, &hdev->reset_request); + return 0; +} + +static void hclge_reset(struct hclge_dev *hdev) +{ + if (hclge_reset_prepare(hdev)) + goto err_reset; + + if (hclge_reset_wait(hdev)) + goto err_reset; + + if (hclge_reset_rebuild(hdev)) + goto err_reset; + return; -err_reset_lock: - rtnl_unlock(); err_reset: if (hclge_reset_err_handle(hdev)) hclge_reset_task_schedule(hdev); @@ -3939,34 +3929,18 @@ static void hclge_reset_subtask(struct hclge_dev *hdev) hdev->reset_type = HNAE3_NONE_RESET; } -static void hclge_reset_service_task(struct work_struct *work) +static void hclge_reset_service_task(struct hclge_dev *hdev) { - struct hclge_dev *hdev = - container_of(work, struct hclge_dev, rst_service_task); - - if (test_and_set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state)) + if (!test_and_clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state)) return; - clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state); + down(&hdev->reset_sem); + set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); hclge_reset_subtask(hdev); clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); -} - -static void hclge_mailbox_service_task(struct work_struct *work) -{ - struct hclge_dev *hdev = - container_of(work, struct hclge_dev, mbx_service_task); - - if (test_and_set_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state)) - return; - - clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state); - - hclge_mbx_handler(hdev); - - clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static void hclge_update_vport_alive(struct hclge_dev *hdev) @@ -3986,29 +3960,62 @@ static void hclge_update_vport_alive(struct hclge_dev *hdev) } } -static void hclge_service_task(struct work_struct *work) +static void hclge_periodic_service_task(struct hclge_dev *hdev) { - struct hclge_dev *hdev = - container_of(work, struct hclge_dev, service_task.work); + unsigned long delta = round_jiffies_relative(HZ); - clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state); + /* Always handle the link updating to make sure link state is + * updated when it is triggered by mbx. + */ + hclge_update_link_status(hdev); - if (hdev->hw_stats.stats_timer >= HCLGE_STATS_TIMER_INTERVAL) { - hclge_update_stats_for_all(hdev); - hdev->hw_stats.stats_timer = 0; + if (time_is_after_jiffies(hdev->last_serv_processed + HZ)) { + delta = jiffies - hdev->last_serv_processed; + + if (delta < round_jiffies_relative(HZ)) { + delta = round_jiffies_relative(HZ) - delta; + goto out; + } } - hclge_update_port_info(hdev); - hclge_update_link_status(hdev); + hdev->serv_processed_cnt++; hclge_update_vport_alive(hdev); + + if (test_bit(HCLGE_STATE_DOWN, &hdev->state)) { + hdev->last_serv_processed = jiffies; + goto out; + } + + if (!(hdev->serv_processed_cnt % HCLGE_STATS_TIMER_INTERVAL)) + hclge_update_stats_for_all(hdev); + + hclge_update_port_info(hdev); hclge_sync_vlan_filter(hdev); - if (hdev->fd_arfs_expire_timer >= HCLGE_FD_ARFS_EXPIRE_TIMER_INTERVAL) { + if (!(hdev->serv_processed_cnt % HCLGE_ARFS_EXPIRE_INTERVAL)) hclge_rfs_filter_expire(hdev); - hdev->fd_arfs_expire_timer = 0; - } - hclge_task_schedule(hdev, round_jiffies_relative(HZ)); + hdev->last_serv_processed = jiffies; + +out: + hclge_task_schedule(hdev, delta); +} + +static void hclge_service_task(struct work_struct *work) +{ + struct hclge_dev *hdev = + container_of(work, struct hclge_dev, service_task.work); + + hclge_reset_service_task(hdev); + hclge_mailbox_service_task(hdev); + hclge_periodic_service_task(hdev); + + /* Handle reset and mbx again in case periodical task delays the + * handling by calling hclge_task_schedule() in + * hclge_periodic_service_task(). + */ + hclge_reset_service_task(hdev); + hclge_mailbox_service_task(hdev); } struct hclge_vport *hclge_get_vport(struct hnae3_handle *handle) @@ -4079,7 +4086,7 @@ static int hclge_put_vector(struct hnae3_handle *handle, int vector) vector_id = hclge_get_vector_index(hdev, vector); if (vector_id < 0) { dev_err(&hdev->pdev->dev, - "Get vector index fail. vector_id =%d\n", vector_id); + "Get vector index fail. vector = %d\n", vector); return vector_id; } @@ -4654,7 +4661,7 @@ static int hclge_map_ring_to_vector(struct hnae3_handle *handle, int vector, vector_id = hclge_get_vector_index(hdev, vector); if (vector_id < 0) { dev_err(&hdev->pdev->dev, - "Get vector index fail. vector_id =%d\n", vector_id); + "failed to get vector index. vector=%d\n", vector); return vector_id; } @@ -6562,7 +6569,7 @@ static int hclge_set_serdes_loopback(struct hclge_dev *hdev, bool en, hclge_cfg_mac_mode(hdev, en); - ret = hclge_mac_phy_link_status_wait(hdev, en, FALSE); + ret = hclge_mac_phy_link_status_wait(hdev, en, false); if (ret) dev_err(&hdev->pdev->dev, "serdes loopback config mac mode timeout\n"); @@ -6620,7 +6627,7 @@ static int hclge_set_phy_loopback(struct hclge_dev *hdev, bool en) hclge_cfg_mac_mode(hdev, en); - ret = hclge_mac_phy_link_status_wait(hdev, en, TRUE); + ret = hclge_mac_phy_link_status_wait(hdev, en, true); if (ret) dev_err(&hdev->pdev->dev, "phy loopback config mac mode timeout\n"); @@ -6734,6 +6741,19 @@ static void hclge_reset_tqp_stats(struct hnae3_handle *handle) } } +static void hclge_flush_link_update(struct hclge_dev *hdev) +{ +#define HCLGE_FLUSH_LINK_TIMEOUT 100000 + + unsigned long last = hdev->serv_processed_cnt; + int i = 0; + + while (test_bit(HCLGE_STATE_LINK_UPDATING, &hdev->state) && + i++ < HCLGE_FLUSH_LINK_TIMEOUT && + last == hdev->serv_processed_cnt) + usleep_range(1, 1); +} + static void hclge_set_timer_task(struct hnae3_handle *handle, bool enable) { struct hclge_vport *vport = hclge_get_vport(handle); @@ -6742,12 +6762,12 @@ static void hclge_set_timer_task(struct hnae3_handle *handle, bool enable) if (enable) { hclge_task_schedule(hdev, round_jiffies_relative(HZ)); } else { - /* Set the DOWN flag here to disable the service to be - * scheduled again - */ + /* Set the DOWN flag here to disable link updating */ set_bit(HCLGE_STATE_DOWN, &hdev->state); - cancel_delayed_work_sync(&hdev->service_task); - clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state); + + /* flush memory to make sure DOWN is seen by service task */ + smp_mb__before_atomic(); + hclge_flush_link_update(hdev); } } @@ -7483,7 +7503,6 @@ void hclge_uninit_vport_mac_table(struct hclge_dev *hdev) struct hclge_vport *vport; int i; - mutex_lock(&hdev->vport_cfg_mutex); for (i = 0; i < hdev->num_alloc_vport; i++) { vport = &hdev->vport[i]; list_for_each_entry_safe(mac, tmp, &vport->uc_mac_list, node) { @@ -7496,7 +7515,6 @@ void hclge_uninit_vport_mac_table(struct hclge_dev *hdev) kfree(mac); } } - mutex_unlock(&hdev->vport_cfg_mutex); } static int hclge_get_mac_ethertype_cmd_status(struct hclge_dev *hdev, @@ -8257,7 +8275,6 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev) struct hclge_vport *vport; int i; - mutex_lock(&hdev->vport_cfg_mutex); for (i = 0; i < hdev->num_alloc_vport; i++) { vport = &hdev->vport[i]; list_for_each_entry_safe(vlan, tmp, &vport->vlan_list, node) { @@ -8265,7 +8282,6 @@ void hclge_uninit_vport_vlan_table(struct hclge_dev *hdev) kfree(vlan); } } - mutex_unlock(&hdev->vport_cfg_mutex); } static void hclge_restore_vlan_table(struct hnae3_handle *handle) @@ -8277,7 +8293,6 @@ static void hclge_restore_vlan_table(struct hnae3_handle *handle) u16 state, vlan_id; int i; - mutex_lock(&hdev->vport_cfg_mutex); for (i = 0; i < hdev->num_alloc_vport; i++) { vport = &hdev->vport[i]; vlan_proto = vport->port_base_vlan_cfg.vlan_info.vlan_proto; @@ -8303,8 +8318,6 @@ static void hclge_restore_vlan_table(struct hnae3_handle *handle) break; } } - - mutex_unlock(&hdev->vport_cfg_mutex); } int hclge_en_hw_strip_rxvtag(struct hnae3_handle *handle, bool enable) @@ -9256,6 +9269,7 @@ static void hclge_state_init(struct hclge_dev *hdev) set_bit(HCLGE_STATE_DOWN, &hdev->state); clear_bit(HCLGE_STATE_RST_SERVICE_SCHED, &hdev->state); clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + clear_bit(HCLGE_STATE_RST_FAIL, &hdev->state); clear_bit(HCLGE_STATE_MBX_SERVICE_SCHED, &hdev->state); clear_bit(HCLGE_STATE_MBX_HANDLING, &hdev->state); } @@ -9269,38 +9283,57 @@ static void hclge_state_uninit(struct hclge_dev *hdev) del_timer_sync(&hdev->reset_timer); if (hdev->service_task.work.func) cancel_delayed_work_sync(&hdev->service_task); - if (hdev->rst_service_task.func) - cancel_work_sync(&hdev->rst_service_task); - if (hdev->mbx_service_task.func) - cancel_work_sync(&hdev->mbx_service_task); } static void hclge_flr_prepare(struct hnae3_ae_dev *ae_dev) { -#define HCLGE_FLR_WAIT_MS 100 -#define HCLGE_FLR_WAIT_CNT 50 - struct hclge_dev *hdev = ae_dev->priv; - int cnt = 0; +#define HCLGE_FLR_RETRY_WAIT_MS 500 +#define HCLGE_FLR_RETRY_CNT 5 - clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - clear_bit(HNAE3_FLR_DONE, &hdev->flr_state); - set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request); - hclge_reset_event(hdev->pdev, NULL); + struct hclge_dev *hdev = ae_dev->priv; + int retry_cnt = 0; + int ret; - while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) && - cnt++ < HCLGE_FLR_WAIT_CNT) - msleep(HCLGE_FLR_WAIT_MS); +retry: + down(&hdev->reset_sem); + set_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + hdev->reset_type = HNAE3_FLR_RESET; + ret = hclge_reset_prepare(hdev); + if (ret) { + dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n", + ret); + if (hdev->reset_pending || + retry_cnt++ < HCLGE_FLR_RETRY_CNT) { + dev_err(&hdev->pdev->dev, + "reset_pending:0x%lx, retry_cnt:%d\n", + hdev->reset_pending, retry_cnt); + clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); + msleep(HCLGE_FLR_RETRY_WAIT_MS); + goto retry; + } + } - if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state)) - dev_err(&hdev->pdev->dev, - "flr wait down timeout: %d\n", cnt); + /* disable misc vector before FLR done */ + hclge_enable_vector(&hdev->misc_vector, false); + set_bit(HCLGE_STATE_CMD_DISABLE, &hdev->state); + hdev->rst_stats.flr_rst_cnt++; } static void hclge_flr_done(struct hnae3_ae_dev *ae_dev) { struct hclge_dev *hdev = ae_dev->priv; + int ret; + + hclge_enable_vector(&hdev->misc_vector, true); + + ret = hclge_reset_rebuild(hdev); + if (ret) + dev_err(&hdev->pdev->dev, "fail to rebuild, ret=%d\n", ret); - set_bit(HNAE3_FLR_DONE, &hdev->flr_state); + hdev->reset_type = HNAE3_NONE_RESET; + clear_bit(HCLGE_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static void hclge_clear_resetting_state(struct hclge_dev *hdev) @@ -9342,21 +9375,17 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) hdev->mps = ETH_FRAME_LEN + ETH_FCS_LEN + 2 * VLAN_HLEN; mutex_init(&hdev->vport_lock); - mutex_init(&hdev->vport_cfg_mutex); spin_lock_init(&hdev->fd_rule_lock); + sema_init(&hdev->reset_sem, 1); ret = hclge_pci_init(hdev); - if (ret) { - dev_err(&pdev->dev, "PCI init failed\n"); + if (ret) goto out; - } /* Firmware command queue initialize */ ret = hclge_cmd_queue_init(hdev); - if (ret) { - dev_err(&pdev->dev, "Cmd queue init failed, ret = %d.\n", ret); + if (ret) goto err_pci_uninit; - } /* Firmware command initialize */ ret = hclge_cmd_init(hdev); @@ -9364,11 +9393,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) goto err_cmd_uninit; ret = hclge_get_cap(hdev); - if (ret) { - dev_err(&pdev->dev, "get hw capability error, ret = %d.\n", - ret); + if (ret) goto err_cmd_uninit; - } ret = hclge_configure(hdev); if (ret) { @@ -9383,12 +9409,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) } ret = hclge_misc_irq_init(hdev); - if (ret) { - dev_err(&pdev->dev, - "Misc IRQ(vector0) init error, ret = %d.\n", - ret); + if (ret) goto err_msi_uninit; - } ret = hclge_alloc_tqps(hdev); if (ret) { @@ -9397,31 +9419,22 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) } ret = hclge_alloc_vport(hdev); - if (ret) { - dev_err(&pdev->dev, "Allocate vport error, ret = %d.\n", ret); + if (ret) goto err_msi_irq_uninit; - } ret = hclge_map_tqp(hdev); - if (ret) { - dev_err(&pdev->dev, "Map tqp error, ret = %d.\n", ret); + if (ret) goto err_msi_irq_uninit; - } if (hdev->hw.mac.media_type == HNAE3_MEDIA_TYPE_COPPER) { ret = hclge_mac_mdio_config(hdev); - if (ret) { - dev_err(&hdev->pdev->dev, - "mdio config fail ret=%d\n", ret); + if (ret) goto err_msi_irq_uninit; - } } ret = hclge_init_umv_space(hdev); - if (ret) { - dev_err(&pdev->dev, "umv space init error, ret=%d.\n", ret); + if (ret) goto err_mdiobus_unreg; - } ret = hclge_mac_init(hdev); if (ret) { @@ -9477,8 +9490,6 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) timer_setup(&hdev->reset_timer, hclge_reset_timer, 0); INIT_DELAYED_WORK(&hdev->service_task, hclge_service_task); - INIT_WORK(&hdev->rst_service_task, hclge_reset_service_task); - INIT_WORK(&hdev->mbx_service_task, hclge_mailbox_service_task); /* Setup affinity after service timer setup because add_timer_on * is called in affinity notify. @@ -9512,6 +9523,8 @@ static int hclge_init_ae_dev(struct hnae3_ae_dev *ae_dev) dev_info(&hdev->pdev->dev, "%s driver initialization finished.\n", HCLGE_DRIVER_NAME); + hclge_task_schedule(hdev, round_jiffies_relative(HZ)); + return 0; err_mdiobus_unreg: @@ -9534,7 +9547,7 @@ out: static void hclge_stats_clear(struct hclge_dev *hdev) { - memset(&hdev->hw_stats, 0, sizeof(hdev->hw_stats)); + memset(&hdev->mac_stats, 0, sizeof(hdev->mac_stats)); } static int hclge_set_mac_spoofchk(struct hclge_dev *hdev, int vf, bool enable) @@ -9895,7 +9908,6 @@ static void hclge_uninit_ae_dev(struct hnae3_ae_dev *ae_dev) mutex_destroy(&hdev->vport_lock); hclge_uninit_vport_mac_table(hdev); hclge_uninit_vport_vlan_table(hdev); - mutex_destroy(&hdev->vport_cfg_mutex); ae_dev->priv = NULL; } @@ -10157,10 +10169,8 @@ static int hclge_get_dfx_reg_bd_num(struct hclge_dev *hdev, int *bd_num_list, u32 type_num) { -#define HCLGE_DFX_REG_BD_NUM 4 - u32 entries_per_desc, desc_index, index, offset, i; - struct hclge_desc desc[HCLGE_DFX_REG_BD_NUM]; + struct hclge_desc desc[HCLGE_GET_DFX_REG_TYPE_CNT]; int ret; ret = hclge_query_bd_num_cmd_send(hdev, desc); @@ -10273,10 +10283,8 @@ static int hclge_get_dfx_reg(struct hclge_dev *hdev, void *data) buf_len = sizeof(*desc_src) * bd_num_max; desc_src = kzalloc(buf_len, GFP_KERNEL); - if (!desc_src) { - dev_err(&hdev->pdev->dev, "%s kzalloc failed\n", __func__); + if (!desc_src) return -ENOMEM; - } for (i = 0; i < dfx_reg_type_num; i++) { bd_num = bd_num_list[i]; @@ -10611,6 +10619,12 @@ static int hclge_init(void) { pr_info("%s is initializing\n", HCLGE_NAME); + hclge_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, HCLGE_NAME); + if (!hclge_wq) { + pr_err("%s: failed to create workqueue\n", HCLGE_NAME); + return -ENOMEM; + } + hnae3_register_ae_algo(&ae_algo); return 0; @@ -10619,6 +10633,7 @@ static int hclge_init(void) static void hclge_exit(void) { hnae3_unregister_ae_algo(&ae_algo); + destroy_workqueue(hclge_wq); } module_init(hclge_init); module_exit(hclge_exit); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h index ebb4c6e9aed3..f78cbb4cc85e 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.h @@ -139,6 +139,8 @@ #define HCLGE_PHY_MDIX_STATUS_B 6 #define HCLGE_PHY_SPEED_DUP_RESOLVE_B 11 +#define HCLGE_GET_DFX_REG_TYPE_CNT 4 + /* Factor used to calculate offset and bitmap of VF num */ #define HCLGE_VF_NUM_PER_CMD 64 @@ -208,13 +210,14 @@ enum HCLGE_DEV_STATE { HCLGE_STATE_NIC_REGISTERED, HCLGE_STATE_ROCE_REGISTERED, HCLGE_STATE_SERVICE_INITED, - HCLGE_STATE_SERVICE_SCHED, HCLGE_STATE_RST_SERVICE_SCHED, HCLGE_STATE_RST_HANDLING, HCLGE_STATE_MBX_SERVICE_SCHED, HCLGE_STATE_MBX_HANDLING, HCLGE_STATE_STATISTICS_UPDATING, HCLGE_STATE_CMD_DISABLE, + HCLGE_STATE_LINK_UPDATING, + HCLGE_STATE_RST_FAIL, HCLGE_STATE_MAX }; @@ -454,11 +457,7 @@ struct hclge_mac_stats { u64 mac_rx_ctrl_pkt_num; }; -#define HCLGE_STATS_TIMER_INTERVAL (60 * 5) -struct hclge_hw_stats { - struct hclge_mac_stats mac_stats; - u32 stats_timer; -}; +#define HCLGE_STATS_TIMER_INTERVAL 300UL struct hclge_vlan_type_cfg { u16 rx_ot_fst_vlan_type; @@ -549,7 +548,7 @@ struct key_info { /* assigned by firmware, the real filter number for each pf may be less */ #define MAX_FD_FILTER_NUM 4096 -#define HCLGE_FD_ARFS_EXPIRE_TIMER_INTERVAL 5 +#define HCLGE_ARFS_EXPIRE_INTERVAL 5UL enum HCLGE_FD_ACTIVE_RULE_TYPE { HCLGE_FD_RULE_NONE, @@ -712,7 +711,7 @@ struct hclge_dev { struct hnae3_ae_dev *ae_dev; struct hclge_hw hw; struct hclge_misc_vector misc_vector; - struct hclge_hw_stats hw_stats; + struct hclge_mac_stats mac_stats; unsigned long state; unsigned long flr_state; unsigned long last_reset_time; @@ -723,6 +722,7 @@ struct hclge_dev { unsigned long reset_request; /* reset has been requested */ unsigned long reset_pending; /* client rst is pending to be served */ struct hclge_rst_stats rst_stats; + struct semaphore reset_sem; /* protect reset process */ u32 fw_version; u16 num_vmdq_vport; /* Num vmdq vport this PF has set up */ u16 num_tqps; /* Num task queue pairs of this PF */ @@ -774,8 +774,6 @@ struct hclge_dev { unsigned long service_timer_previous; struct timer_list reset_timer; struct delayed_work service_task; - struct work_struct rst_service_task; - struct work_struct mbx_service_task; bool cur_promisc; int num_alloc_vfs; /* Actual number of VFs allocated */ @@ -811,7 +809,8 @@ struct hclge_dev { struct hlist_head fd_rule_list; spinlock_t fd_rule_lock; /* protect fd_rule_list and fd_bmap */ u16 hclge_fd_rule_num; - u16 fd_arfs_expire_timer; + unsigned long serv_processed_cnt; + unsigned long last_serv_processed; unsigned long fd_bmap[BITS_TO_LONGS(MAX_FD_FILTER_NUM)]; enum HCLGE_FD_ACTIVE_RULE_TYPE fd_active_type; u8 fd_en; @@ -825,8 +824,6 @@ struct hclge_dev { u16 share_umv_size; struct mutex umv_mutex; /* protect share_umv_size */ - struct mutex vport_cfg_mutex; /* Protect stored vf table */ - DECLARE_KFIFO(mac_tnl_log, struct hclge_mac_tnl_stats, HCLGE_MAC_TNL_LOG_SIZE); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c index 0b433ebe6a2d..a3c0822191a9 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c @@ -86,10 +86,12 @@ static int hclge_send_mbx_msg(struct hclge_vport *vport, u8 *msg, u16 msg_len, int hclge_inform_reset_assert_to_vf(struct hclge_vport *vport) { struct hclge_dev *hdev = vport->back; - enum hnae3_reset_type reset_type; + u16 reset_type; u8 msg_data[2]; u8 dest_vfid; + BUILD_BUG_ON(HNAE3_MAX_RESET > U16_MAX); + dest_vfid = (u8)vport->vport_id; if (hdev->reset_type == HNAE3_FUNC_RESET) @@ -635,7 +637,6 @@ static void hclge_handle_link_change_event(struct hclge_dev *hdev, #define LINK_STATUS_OFFSET 1 #define LINK_FAIL_CODE_OFFSET 2 - clear_bit(HCLGE_STATE_SERVICE_SCHED, &hdev->state); hclge_task_schedule(hdev, 0); if (!req->msg[LINK_STATUS_OFFSET]) @@ -798,13 +799,11 @@ void hclge_mbx_handler(struct hclge_dev *hdev) hclge_get_link_mode(vport, req); break; case HCLGE_MBX_GET_VF_FLR_STATUS: - mutex_lock(&hdev->vport_cfg_mutex); hclge_rm_vport_all_mac_table(vport, true, HCLGE_MAC_ADDR_UC); hclge_rm_vport_all_mac_table(vport, true, HCLGE_MAC_ADDR_MC); hclge_rm_vport_all_vlan_table(vport, true); - mutex_unlock(&hdev->vport_cfg_mutex); break; case HCLGE_MBX_GET_MEDIA_TYPE: ret = hclge_get_vf_media_type(vport, req); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c index af2245e3bb95..f38d236ebf4f 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c @@ -443,7 +443,7 @@ void hclgevf_cmd_uninit(struct hclgevf_dev *hdev) { spin_lock_bh(&hdev->hw.cmq.csq.lock); spin_lock(&hdev->hw.cmq.crq.lock); - clear_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state); + set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state); hclgevf_cmd_uninit_regs(&hdev->hw); spin_unlock(&hdev->hw.cmq.crq.lock); spin_unlock_bh(&hdev->hw.cmq.csq.lock); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c index 25d78a5aaa34..d6597206e692 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c @@ -16,6 +16,8 @@ static int hclgevf_reset_hdev(struct hclgevf_dev *hdev); static struct hnae3_ae_algo ae_algovf; +static struct workqueue_struct *hclgevf_wq; + static const struct pci_device_id ae_algovf_pci_tbl[] = { {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_VF), 0}, {PCI_VDEVICE(HUAWEI, HNAE3_DEV_ID_100G_RDMA_DCB_PFC_VF), 0}, @@ -440,6 +442,9 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state) struct hnae3_client *rclient; struct hnae3_client *client; + if (test_and_set_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state)) + return; + client = handle->client; rclient = hdev->roce_client; @@ -452,6 +457,8 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state) rclient->ops->link_status_change(rhandle, !!link_state); hdev->hw.mac.link = link_state; } + + clear_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state); } static void hclgevf_update_link_mode(struct hclgevf_dev *hdev) @@ -1309,14 +1316,13 @@ static int hclgevf_set_vlan_filter(struct hnae3_handle *handle, msg_data[0] = is_kill; memcpy(&msg_data[1], &vlan_id, sizeof(vlan_id)); memcpy(&msg_data[3], &proto, sizeof(proto)); - ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN, - HCLGE_MBX_VLAN_FILTER, msg_data, - HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0); - /* when remove hw vlan filter failed, record the vlan id, * and try to remove it from hw later, to be consistence * with stack. */ + ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_SET_VLAN, + HCLGE_MBX_VLAN_FILTER, msg_data, + HCLGEVF_VLAN_MBX_MSG_LEN, true, NULL, 0); if (is_kill && ret) set_bit(vlan_id, hdev->vlan_del_fail_bmap); @@ -1404,32 +1410,6 @@ static int hclgevf_notify_client(struct hclgevf_dev *hdev, return ret; } -static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev) -{ - struct hclgevf_dev *hdev = ae_dev->priv; - - set_bit(HNAE3_FLR_DONE, &hdev->flr_state); -} - -static int hclgevf_flr_poll_timeout(struct hclgevf_dev *hdev, - unsigned long delay_us, - unsigned long wait_cnt) -{ - unsigned long cnt = 0; - - while (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state) && - cnt++ < wait_cnt) - usleep_range(delay_us, delay_us * 2); - - if (!test_bit(HNAE3_FLR_DONE, &hdev->flr_state)) { - dev_err(&hdev->pdev->dev, - "flr wait timeout\n"); - return -ETIMEDOUT; - } - - return 0; -} - static int hclgevf_reset_wait(struct hclgevf_dev *hdev) { #define HCLGEVF_RESET_WAIT_US 20000 @@ -1440,11 +1420,7 @@ static int hclgevf_reset_wait(struct hclgevf_dev *hdev) u32 val; int ret; - if (hdev->reset_type == HNAE3_FLR_RESET) - return hclgevf_flr_poll_timeout(hdev, - HCLGEVF_RESET_WAIT_US, - HCLGEVF_RESET_WAIT_CNT); - else if (hdev->reset_type == HNAE3_VF_RESET) + if (hdev->reset_type == HNAE3_VF_RESET) ret = readl_poll_timeout(hdev->hw.io_base + HCLGEVF_VF_RST_ING, val, !(val & HCLGEVF_VF_RST_ING_BIT), @@ -1516,7 +1492,8 @@ static int hclgevf_reset_stack(struct hclgevf_dev *hdev) /* clear handshake status with IMP */ hclgevf_reset_handshake(hdev, false); - return 0; + /* bring up the nic to enable TX/RX again */ + return hclgevf_notify_client(hdev, HNAE3_UP_CLIENT); } static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev) @@ -1525,18 +1502,10 @@ static int hclgevf_reset_prepare_wait(struct hclgevf_dev *hdev) int ret = 0; - switch (hdev->reset_type) { - case HNAE3_VF_FUNC_RESET: + if (hdev->reset_type == HNAE3_VF_FUNC_RESET) { ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_RESET, 0, NULL, 0, true, NULL, sizeof(u8)); hdev->rst_stats.vf_func_rst_cnt++; - break; - case HNAE3_FLR_RESET: - set_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - hdev->rst_stats.flr_rst_cnt++; - break; - default: - break; } set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state); @@ -1591,11 +1560,12 @@ static void hclgevf_reset_err_handle(struct hclgevf_dev *hdev) set_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state); hclgevf_reset_task_schedule(hdev); } else { + set_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state); hclgevf_dump_rst_info(hdev); } } -static int hclgevf_reset(struct hclgevf_dev *hdev) +static int hclgevf_reset_prepare(struct hclgevf_dev *hdev) { struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); int ret; @@ -1605,61 +1575,64 @@ static int hclgevf_reset(struct hclgevf_dev *hdev) */ ae_dev->reset_type = hdev->reset_type; hdev->rst_stats.rst_cnt++; - rtnl_lock(); + rtnl_lock(); /* bring down the nic to stop any ongoing TX/RX */ ret = hclgevf_notify_client(hdev, HNAE3_DOWN_CLIENT); - if (ret) - goto err_reset_lock; - rtnl_unlock(); - - ret = hclgevf_reset_prepare_wait(hdev); if (ret) - goto err_reset; + return ret; - /* check if VF could successfully fetch the hardware reset completion - * status from the hardware - */ - ret = hclgevf_reset_wait(hdev); - if (ret) { - /* can't do much in this situation, will disable VF */ - dev_err(&hdev->pdev->dev, - "VF failed(=%d) to fetch H/W reset completion status\n", - ret); - goto err_reset; - } + return hclgevf_reset_prepare_wait(hdev); +} + +static int hclgevf_reset_rebuild(struct hclgevf_dev *hdev) +{ + struct hnae3_ae_dev *ae_dev = pci_get_drvdata(hdev->pdev); + int ret; hdev->rst_stats.hw_rst_done_cnt++; rtnl_lock(); - /* now, re-initialize the nic client and ae device */ ret = hclgevf_reset_stack(hdev); + rtnl_unlock(); if (ret) { dev_err(&hdev->pdev->dev, "failed to reset VF stack\n"); - goto err_reset_lock; + return ret; } - /* bring up the nic to enable TX/RX again */ - ret = hclgevf_notify_client(hdev, HNAE3_UP_CLIENT); - if (ret) - goto err_reset_lock; - - rtnl_unlock(); - hdev->last_reset_time = jiffies; ae_dev->reset_type = HNAE3_NONE_RESET; hdev->rst_stats.rst_done_cnt++; hdev->rst_stats.rst_fail_cnt = 0; + clear_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state); + + return 0; +} + +static void hclgevf_reset(struct hclgevf_dev *hdev) +{ + if (hclgevf_reset_prepare(hdev)) + goto err_reset; + + /* check if VF could successfully fetch the hardware reset completion + * status from the hardware + */ + if (hclgevf_reset_wait(hdev)) { + /* can't do much in this situation, will disable VF */ + dev_err(&hdev->pdev->dev, + "failed to fetch H/W reset completion status\n"); + goto err_reset; + } + + if (hclgevf_reset_rebuild(hdev)) + goto err_reset; + + return; - return ret; -err_reset_lock: - rtnl_unlock(); err_reset: hclgevf_reset_err_handle(hdev); - - return ret; } static enum hnae3_reset_type hclgevf_get_reset_level(struct hclgevf_dev *hdev, @@ -1722,25 +1695,60 @@ static void hclgevf_set_def_reset_request(struct hnae3_ae_dev *ae_dev, set_bit(rst_type, &hdev->default_reset_request); } +static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en) +{ + writel(en ? 1 : 0, vector->addr); +} + static void hclgevf_flr_prepare(struct hnae3_ae_dev *ae_dev) { -#define HCLGEVF_FLR_WAIT_MS 100 -#define HCLGEVF_FLR_WAIT_CNT 50 +#define HCLGEVF_FLR_RETRY_WAIT_MS 500 +#define HCLGEVF_FLR_RETRY_CNT 5 + struct hclgevf_dev *hdev = ae_dev->priv; - int cnt = 0; + int retry_cnt = 0; + int ret; - clear_bit(HNAE3_FLR_DOWN, &hdev->flr_state); - clear_bit(HNAE3_FLR_DONE, &hdev->flr_state); - set_bit(HNAE3_FLR_RESET, &hdev->default_reset_request); - hclgevf_reset_event(hdev->pdev, NULL); +retry: + down(&hdev->reset_sem); + set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + hdev->reset_type = HNAE3_FLR_RESET; + ret = hclgevf_reset_prepare(hdev); + if (ret) { + dev_err(&hdev->pdev->dev, "fail to prepare FLR, ret=%d\n", + ret); + if (hdev->reset_pending || + retry_cnt++ < HCLGEVF_FLR_RETRY_CNT) { + dev_err(&hdev->pdev->dev, + "reset_pending:0x%lx, retry_cnt:%d\n", + hdev->reset_pending, retry_cnt); + clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); + msleep(HCLGEVF_FLR_RETRY_WAIT_MS); + goto retry; + } + } - while (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state) && - cnt++ < HCLGEVF_FLR_WAIT_CNT) - msleep(HCLGEVF_FLR_WAIT_MS); + /* disable misc vector before FLR done */ + hclgevf_enable_vector(&hdev->misc_vector, false); + hdev->rst_stats.flr_rst_cnt++; +} - if (!test_bit(HNAE3_FLR_DOWN, &hdev->flr_state)) - dev_err(&hdev->pdev->dev, - "flr wait down timeout: %d\n", cnt); +static void hclgevf_flr_done(struct hnae3_ae_dev *ae_dev) +{ + struct hclgevf_dev *hdev = ae_dev->priv; + int ret; + + hclgevf_enable_vector(&hdev->misc_vector, true); + + ret = hclgevf_reset_rebuild(hdev); + if (ret) + dev_warn(&hdev->pdev->dev, "fail to rebuild, ret=%d\n", + ret); + + hdev->reset_type = HNAE3_NONE_RESET; + clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } static u32 hclgevf_get_fw_version(struct hnae3_handle *handle) @@ -1767,62 +1775,37 @@ static void hclgevf_get_misc_vector(struct hclgevf_dev *hdev) void hclgevf_reset_task_schedule(struct hclgevf_dev *hdev) { - if (!test_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state) && - !test_bit(HCLGEVF_STATE_REMOVING, &hdev->state)) { - set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state); - schedule_work(&hdev->rst_service_task); - } + if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) && + !test_and_set_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, + &hdev->state)) + mod_delayed_work(hclgevf_wq, &hdev->service_task, 0); } void hclgevf_mbx_task_schedule(struct hclgevf_dev *hdev) { - if (!test_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state) && - !test_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state)) { - set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state); - schedule_work(&hdev->mbx_service_task); - } + if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) && + !test_and_set_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, + &hdev->state)) + mod_delayed_work(hclgevf_wq, &hdev->service_task, 0); } -static void hclgevf_task_schedule(struct hclgevf_dev *hdev) +static void hclgevf_task_schedule(struct hclgevf_dev *hdev, + unsigned long delay) { - if (!test_bit(HCLGEVF_STATE_DOWN, &hdev->state) && - !test_and_set_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state)) - schedule_work(&hdev->service_task); + if (!test_bit(HCLGEVF_STATE_REMOVING, &hdev->state) && + !test_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state)) + mod_delayed_work(hclgevf_wq, &hdev->service_task, delay); } -static void hclgevf_deferred_task_schedule(struct hclgevf_dev *hdev) -{ - /* if we have any pending mailbox event then schedule the mbx task */ - if (hdev->mbx_event_pending) - hclgevf_mbx_task_schedule(hdev); - - if (test_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state)) - hclgevf_reset_task_schedule(hdev); -} - -static void hclgevf_service_timer(struct timer_list *t) -{ - struct hclgevf_dev *hdev = from_timer(hdev, t, service_timer); - - mod_timer(&hdev->service_timer, jiffies + - HCLGEVF_GENERAL_TASK_INTERVAL * HZ); - - hdev->stats_timer++; - hclgevf_task_schedule(hdev); -} - -static void hclgevf_reset_service_task(struct work_struct *work) +static void hclgevf_reset_service_task(struct hclgevf_dev *hdev) { #define HCLGEVF_MAX_RESET_ATTEMPTS_CNT 3 - struct hclgevf_dev *hdev = - container_of(work, struct hclgevf_dev, rst_service_task); - int ret; - - if (test_and_set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state)) + if (!test_and_clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state)) return; - clear_bit(HCLGEVF_STATE_RST_SERVICE_SCHED, &hdev->state); + down(&hdev->reset_sem); + set_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); if (test_and_clear_bit(HCLGEVF_RESET_PENDING, &hdev->reset_state)) { @@ -1836,12 +1819,8 @@ static void hclgevf_reset_service_task(struct work_struct *work) hdev->last_reset_time = jiffies; while ((hdev->reset_type = hclgevf_get_reset_level(hdev, &hdev->reset_pending)) - != HNAE3_NONE_RESET) { - ret = hclgevf_reset(hdev); - if (ret) - dev_err(&hdev->pdev->dev, - "VF stack reset failed %d.\n", ret); - } + != HNAE3_NONE_RESET) + hclgevf_reset(hdev); } else if (test_and_clear_bit(HCLGEVF_RESET_REQUESTED, &hdev->reset_state)) { /* we could be here when either of below happens: @@ -1882,42 +1861,29 @@ static void hclgevf_reset_service_task(struct work_struct *work) hclgevf_reset_task_schedule(hdev); } + hdev->reset_type = HNAE3_NONE_RESET; clear_bit(HCLGEVF_STATE_RST_HANDLING, &hdev->state); + up(&hdev->reset_sem); } -static void hclgevf_mailbox_service_task(struct work_struct *work) +static void hclgevf_mailbox_service_task(struct hclgevf_dev *hdev) { - struct hclgevf_dev *hdev; - - hdev = container_of(work, struct hclgevf_dev, mbx_service_task); + if (!test_and_clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state)) + return; if (test_and_set_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state)) return; - clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state); - hclgevf_mbx_async_handler(hdev); clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state); } -static void hclgevf_keep_alive_timer(struct timer_list *t) -{ - struct hclgevf_dev *hdev = from_timer(hdev, t, keep_alive_timer); - - schedule_work(&hdev->keep_alive_task); - mod_timer(&hdev->keep_alive_timer, jiffies + - HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ); -} - -static void hclgevf_keep_alive_task(struct work_struct *work) +static void hclgevf_keep_alive(struct hclgevf_dev *hdev) { - struct hclgevf_dev *hdev; u8 respmsg; int ret; - hdev = container_of(work, struct hclgevf_dev, keep_alive_task); - if (test_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state)) return; @@ -1928,19 +1894,32 @@ static void hclgevf_keep_alive_task(struct work_struct *work) "VF sends keep alive cmd failed(=%d)\n", ret); } -static void hclgevf_service_task(struct work_struct *work) +static void hclgevf_periodic_service_task(struct hclgevf_dev *hdev) { - struct hnae3_handle *handle; - struct hclgevf_dev *hdev; + unsigned long delta = round_jiffies_relative(HZ); + struct hnae3_handle *handle = &hdev->nic; - hdev = container_of(work, struct hclgevf_dev, service_task); - handle = &hdev->nic; + if (time_is_after_jiffies(hdev->last_serv_processed + HZ)) { + delta = jiffies - hdev->last_serv_processed; - if (hdev->stats_timer >= HCLGEVF_STATS_TIMER_INTERVAL) { - hclgevf_tqps_update_stats(handle); - hdev->stats_timer = 0; + if (delta < round_jiffies_relative(HZ)) { + delta = round_jiffies_relative(HZ) - delta; + goto out; + } } + hdev->serv_processed_cnt++; + if (!(hdev->serv_processed_cnt % HCLGEVF_KEEP_ALIVE_TASK_INTERVAL)) + hclgevf_keep_alive(hdev); + + if (test_bit(HCLGEVF_STATE_DOWN, &hdev->state)) { + hdev->last_serv_processed = jiffies; + goto out; + } + + if (!(hdev->serv_processed_cnt % HCLGEVF_STATS_TIMER_INTERVAL)) + hclgevf_tqps_update_stats(handle); + /* request the link status from the PF. PF would be able to tell VF * about such updates in future so we might remove this later */ @@ -1950,9 +1929,27 @@ static void hclgevf_service_task(struct work_struct *work) hclgevf_sync_vlan_filter(hdev); - hclgevf_deferred_task_schedule(hdev); + hdev->last_serv_processed = jiffies; - clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state); +out: + hclgevf_task_schedule(hdev, delta); +} + +static void hclgevf_service_task(struct work_struct *work) +{ + struct hclgevf_dev *hdev = container_of(work, struct hclgevf_dev, + service_task.work); + + hclgevf_reset_service_task(hdev); + hclgevf_mailbox_service_task(hdev); + hclgevf_periodic_service_task(hdev); + + /* Handle reset and mbx again in case periodical task delays the + * handling by calling hclgevf_task_schedule() in + * hclgevf_periodic_service_task() + */ + hclgevf_reset_service_task(hdev); + hclgevf_mailbox_service_task(hdev); } static void hclgevf_clear_event_cause(struct hclgevf_dev *hdev, u32 regclr) @@ -2010,11 +2007,6 @@ static enum hclgevf_evt_cause hclgevf_check_evt_cause(struct hclgevf_dev *hdev, return HCLGEVF_VECTOR0_EVENT_OTHER; } -static void hclgevf_enable_vector(struct hclgevf_misc_vector *vector, bool en) -{ - writel(en ? 1 : 0, vector->addr); -} - static irqreturn_t hclgevf_misc_irq_handle(int irq, void *data) { enum hclgevf_evt_cause event_cause; @@ -2189,16 +2181,31 @@ static int hclgevf_init_vlan_config(struct hclgevf_dev *hdev) false); } +static void hclgevf_flush_link_update(struct hclgevf_dev *hdev) +{ +#define HCLGEVF_FLUSH_LINK_TIMEOUT 100000 + + unsigned long last = hdev->serv_processed_cnt; + int i = 0; + + while (test_bit(HCLGEVF_STATE_LINK_UPDATING, &hdev->state) && + i++ < HCLGEVF_FLUSH_LINK_TIMEOUT && + last == hdev->serv_processed_cnt) + usleep_range(1, 1); +} + static void hclgevf_set_timer_task(struct hnae3_handle *handle, bool enable) { struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle); if (enable) { - mod_timer(&hdev->service_timer, jiffies + HZ); + hclgevf_task_schedule(hdev, 0); } else { - del_timer_sync(&hdev->service_timer); - cancel_work_sync(&hdev->service_task); - clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state); + set_bit(HCLGEVF_STATE_DOWN, &hdev->state); + + /* flush memory to make sure DOWN is seen by service task */ + smp_mb__before_atomic(); + hclgevf_flush_link_update(hdev); } } @@ -2245,16 +2252,12 @@ static int hclgevf_set_alive(struct hnae3_handle *handle, bool alive) static int hclgevf_client_start(struct hnae3_handle *handle) { - struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle); int ret; ret = hclgevf_set_alive(handle, true); if (ret) return ret; - mod_timer(&hdev->keep_alive_timer, jiffies + - HCLGEVF_KEEP_ALIVE_TASK_INTERVAL * HZ); - return 0; } @@ -2267,27 +2270,18 @@ static void hclgevf_client_stop(struct hnae3_handle *handle) if (ret) dev_warn(&hdev->pdev->dev, "%s failed %d\n", __func__, ret); - - del_timer_sync(&hdev->keep_alive_timer); - cancel_work_sync(&hdev->keep_alive_task); } static void hclgevf_state_init(struct hclgevf_dev *hdev) { - /* setup tasks for the MBX */ - INIT_WORK(&hdev->mbx_service_task, hclgevf_mailbox_service_task); clear_bit(HCLGEVF_STATE_MBX_SERVICE_SCHED, &hdev->state); clear_bit(HCLGEVF_STATE_MBX_HANDLING, &hdev->state); + clear_bit(HCLGEVF_STATE_RST_FAIL, &hdev->state); - /* setup tasks for service timer */ - timer_setup(&hdev->service_timer, hclgevf_service_timer, 0); - - INIT_WORK(&hdev->service_task, hclgevf_service_task); - clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state); - - INIT_WORK(&hdev->rst_service_task, hclgevf_reset_service_task); + INIT_DELAYED_WORK(&hdev->service_task, hclgevf_service_task); mutex_init(&hdev->mbx_resp.mbx_mutex); + sema_init(&hdev->reset_sem, 1); /* bring the device down */ set_bit(HCLGEVF_STATE_DOWN, &hdev->state); @@ -2298,18 +2292,8 @@ static void hclgevf_state_uninit(struct hclgevf_dev *hdev) set_bit(HCLGEVF_STATE_DOWN, &hdev->state); set_bit(HCLGEVF_STATE_REMOVING, &hdev->state); - if (hdev->keep_alive_timer.function) - del_timer_sync(&hdev->keep_alive_timer); - if (hdev->keep_alive_task.func) - cancel_work_sync(&hdev->keep_alive_task); - if (hdev->service_timer.function) - del_timer_sync(&hdev->service_timer); - if (hdev->service_task.func) - cancel_work_sync(&hdev->service_task); - if (hdev->mbx_service_task.func) - cancel_work_sync(&hdev->mbx_service_task); - if (hdev->rst_service_task.func) - cancel_work_sync(&hdev->rst_service_task); + if (hdev->service_task.work.func) + cancel_delayed_work_sync(&hdev->service_task); mutex_destroy(&hdev->mbx_resp.mbx_mutex); } @@ -2383,8 +2367,10 @@ static int hclgevf_misc_irq_init(struct hclgevf_dev *hdev) hclgevf_get_misc_vector(hdev); + snprintf(hdev->misc_vector.name, HNAE3_INT_NAME_LEN, "%s-misc-%s", + HCLGEVF_NAME, pci_name(hdev->pdev)); ret = request_irq(hdev->misc_vector.vector_irq, hclgevf_misc_irq_handle, - 0, "hclgevf_cmd", hdev); + 0, hdev->misc_vector.name, hdev); if (ret) { dev_err(&hdev->pdev->dev, "VF failed to request misc irq(%d)\n", hdev->misc_vector.vector_irq); @@ -2611,11 +2597,11 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev) if (hnae3_dev_roce_supported(hdev)) { hdev->roce_base_msix_offset = - hnae3_get_field(__le16_to_cpu(req->msixcap_localid_ba_rocee), + hnae3_get_field(le16_to_cpu(req->msixcap_localid_ba_rocee), HCLGEVF_MSIX_OFT_ROCEE_M, HCLGEVF_MSIX_OFT_ROCEE_S); hdev->num_roce_msix = - hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number), HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S); /* nic's msix numbers is always equals to the roce's. */ @@ -2628,7 +2614,7 @@ static int hclgevf_query_vf_resource(struct hclgevf_dev *hdev) hdev->roce_base_msix_offset; } else { hdev->num_msi = - hnae3_get_field(__le16_to_cpu(req->vf_intr_vector_number), + hnae3_get_field(le16_to_cpu(req->vf_intr_vector_number), HCLGEVF_VEC_NUM_M, HCLGEVF_VEC_NUM_S); hdev->num_nic_msix = hdev->num_msi; @@ -2725,16 +2711,12 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) int ret; ret = hclgevf_pci_init(hdev); - if (ret) { - dev_err(&pdev->dev, "PCI initialization failed\n"); + if (ret) return ret; - } ret = hclgevf_cmd_queue_init(hdev); - if (ret) { - dev_err(&pdev->dev, "Cmd queue init failed: %d\n", ret); + if (ret) goto err_cmd_queue_init; - } ret = hclgevf_cmd_init(hdev); if (ret) @@ -2742,11 +2724,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) /* Get vf resource */ ret = hclgevf_query_vf_resource(hdev); - if (ret) { - dev_err(&hdev->pdev->dev, - "Query vf status error, ret = %d.\n", ret); + if (ret) goto err_cmd_init; - } ret = hclgevf_init_msi(hdev); if (ret) { @@ -2756,13 +2735,11 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) hclgevf_state_init(hdev); hdev->reset_level = HNAE3_VF_FUNC_RESET; + hdev->reset_type = HNAE3_NONE_RESET; ret = hclgevf_misc_irq_init(hdev); - if (ret) { - dev_err(&pdev->dev, "failed(%d) to init Misc IRQ(vector0)\n", - ret); + if (ret) goto err_misc_irq_init; - } set_bit(HCLGEVF_STATE_IRQ_INITED, &hdev->state); @@ -2779,10 +2756,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) } ret = hclgevf_set_handle_info(hdev); - if (ret) { - dev_err(&pdev->dev, "failed(%d) to set handle info\n", ret); + if (ret) goto err_config; - } ret = hclgevf_config_gro(hdev, true); if (ret) @@ -2807,6 +2782,8 @@ static int hclgevf_init_hdev(struct hclgevf_dev *hdev) dev_info(&hdev->pdev->dev, "finished initializing %s driver\n", HCLGEVF_DRIVER_NAME); + hclgevf_task_schedule(hdev, round_jiffies_relative(HZ)); + return 0; err_config: @@ -2838,7 +2815,6 @@ static void hclgevf_uninit_hdev(struct hclgevf_dev *hdev) static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev) { struct pci_dev *pdev = ae_dev->pdev; - struct hclgevf_dev *hdev; int ret; ret = hclgevf_alloc_hdev(ae_dev); @@ -2853,10 +2829,6 @@ static int hclgevf_init_ae_dev(struct hnae3_ae_dev *ae_dev) return ret; } - hdev = ae_dev->priv; - timer_setup(&hdev->keep_alive_timer, hclgevf_keep_alive_timer, 0); - INIT_WORK(&hdev->keep_alive_task, hclgevf_keep_alive_task); - return 0; } @@ -3213,6 +3185,12 @@ static int hclgevf_init(void) { pr_info("%s is initializing\n", HCLGEVF_NAME); + hclgevf_wq = alloc_workqueue("%s", WQ_MEM_RECLAIM, 0, HCLGEVF_NAME); + if (!hclgevf_wq) { + pr_err("%s: failed to create workqueue\n", HCLGEVF_NAME); + return -ENOMEM; + } + hnae3_register_ae_algo(&ae_algovf); return 0; @@ -3221,6 +3199,7 @@ static int hclgevf_init(void) static void hclgevf_exit(void) { hnae3_unregister_ae_algo(&ae_algovf); + destroy_workqueue(hclgevf_wq); } module_init(hclgevf_init); module_exit(hclgevf_exit); diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h index 2f4c81bf4169..fee8d97f323c 100644 --- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h +++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h @@ -142,12 +142,13 @@ enum hclgevf_states { HCLGEVF_STATE_REMOVING, HCLGEVF_STATE_NIC_REGISTERED, /* task states */ - HCLGEVF_STATE_SERVICE_SCHED, HCLGEVF_STATE_RST_SERVICE_SCHED, HCLGEVF_STATE_RST_HANDLING, HCLGEVF_STATE_MBX_SERVICE_SCHED, HCLGEVF_STATE_MBX_HANDLING, HCLGEVF_STATE_CMD_DISABLE, + HCLGEVF_STATE_LINK_UPDATING, + HCLGEVF_STATE_RST_FAIL, }; struct hclgevf_mac { @@ -220,6 +221,7 @@ struct hclgevf_rss_cfg { struct hclgevf_misc_vector { u8 __iomem *addr; int vector_irq; + char name[HNAE3_INT_NAME_LEN]; }; struct hclgevf_rst_stats { @@ -251,6 +253,7 @@ struct hclgevf_dev { unsigned long reset_state; /* requested, pending */ struct hclgevf_rst_stats rst_stats; u32 reset_attempts; + struct semaphore reset_sem; /* protect reset process */ u32 fw_version; u16 num_tqps; /* num task queue pairs of this PF */ @@ -283,12 +286,7 @@ struct hclgevf_dev { struct hclgevf_mbx_resp_status mbx_resp; /* mailbox response */ struct hclgevf_mbx_arq_ring arq; /* mailbox async rx queue */ - struct timer_list service_timer; - struct timer_list keep_alive_timer; - struct work_struct service_task; - struct work_struct keep_alive_task; - struct work_struct rst_service_task; - struct work_struct mbx_service_task; + struct delayed_work service_task; struct hclgevf_tqp *htqp; @@ -298,7 +296,8 @@ struct hclgevf_dev { struct hnae3_client *nic_client; struct hnae3_client *roce_client; u32 flag; - u32 stats_timer; + unsigned long serv_processed_cnt; + unsigned long last_serv_processed; }; static inline bool hclgevf_is_reset_pending(struct hclgevf_dev *hdev) diff --git a/drivers/net/ethernet/huawei/hinic/hinic_main.c b/drivers/net/ethernet/huawei/hinic/hinic_main.c index 2411ad270c98..02a14f5e7fe3 100644 --- a/drivers/net/ethernet/huawei/hinic/hinic_main.c +++ b/drivers/net/ethernet/huawei/hinic/hinic_main.c @@ -766,7 +766,7 @@ static void hinic_set_rx_mode(struct net_device *netdev) queue_work(nic_dev->workq, &rx_mode_work->work); } -static void hinic_tx_timeout(struct net_device *netdev) +static void hinic_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct hinic_dev *nic_dev = netdev_priv(netdev); diff --git a/drivers/net/ethernet/i825xx/82596.c b/drivers/net/ethernet/i825xx/82596.c index 92929750f832..bef676d93339 100644 --- a/drivers/net/ethernet/i825xx/82596.c +++ b/drivers/net/ethernet/i825xx/82596.c @@ -363,7 +363,7 @@ static netdev_tx_t i596_start_xmit(struct sk_buff *skb, struct net_device *dev); static irqreturn_t i596_interrupt(int irq, void *dev_id); static int i596_close(struct net_device *dev); static void i596_add_cmd(struct net_device *dev, struct i596_cmd *cmd); -static void i596_tx_timeout (struct net_device *dev); +static void i596_tx_timeout (struct net_device *dev, unsigned int txqueue); static void print_eth(unsigned char *buf, char *str); static void set_multicast_list(struct net_device *dev); @@ -1019,7 +1019,7 @@ err_irq_dev: return res; } -static void i596_tx_timeout (struct net_device *dev) +static void i596_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct i596_private *lp = dev->ml_priv; int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/i825xx/ether1.c b/drivers/net/ethernet/i825xx/ether1.c index bb3b8adbe4f0..a0bfb509e002 100644 --- a/drivers/net/ethernet/i825xx/ether1.c +++ b/drivers/net/ethernet/i825xx/ether1.c @@ -66,7 +66,7 @@ static netdev_tx_t ether1_sendpacket(struct sk_buff *skb, static irqreturn_t ether1_interrupt(int irq, void *dev_id); static int ether1_close(struct net_device *dev); static void ether1_setmulticastlist(struct net_device *dev); -static void ether1_timeout(struct net_device *dev); +static void ether1_timeout(struct net_device *dev, unsigned int txqueue); /* ------------------------------------------------------------------------- */ @@ -650,7 +650,7 @@ ether1_open (struct net_device *dev) } static void -ether1_timeout(struct net_device *dev) +ether1_timeout(struct net_device *dev, unsigned int txqueue) { printk(KERN_WARNING "%s: transmit timeout, network cable problem?\n", dev->name); diff --git a/drivers/net/ethernet/i825xx/lib82596.c b/drivers/net/ethernet/i825xx/lib82596.c index f9742af7f142..b03757e169e4 100644 --- a/drivers/net/ethernet/i825xx/lib82596.c +++ b/drivers/net/ethernet/i825xx/lib82596.c @@ -351,7 +351,7 @@ static netdev_tx_t i596_start_xmit(struct sk_buff *skb, struct net_device *dev); static irqreturn_t i596_interrupt(int irq, void *dev_id); static int i596_close(struct net_device *dev); static void i596_add_cmd(struct net_device *dev, struct i596_cmd *cmd); -static void i596_tx_timeout (struct net_device *dev); +static void i596_tx_timeout (struct net_device *dev, unsigned int txqueue); static void print_eth(unsigned char *buf, char *str); static void set_multicast_list(struct net_device *dev); static inline void ca(struct net_device *dev); @@ -936,7 +936,7 @@ out_remove_rx_bufs: return -EAGAIN; } -static void i596_tx_timeout (struct net_device *dev) +static void i596_tx_timeout (struct net_device *dev, unsigned int txqueue) { struct i596_private *lp = netdev_priv(dev); diff --git a/drivers/net/ethernet/i825xx/sun3_82586.c b/drivers/net/ethernet/i825xx/sun3_82586.c index 1a86184d44c0..4564ee02c95f 100644 --- a/drivers/net/ethernet/i825xx/sun3_82586.c +++ b/drivers/net/ethernet/i825xx/sun3_82586.c @@ -125,7 +125,7 @@ static netdev_tx_t sun3_82586_send_packet(struct sk_buff *, struct net_device *); static struct net_device_stats *sun3_82586_get_stats(struct net_device *dev); static void set_multicast_list(struct net_device *dev); -static void sun3_82586_timeout(struct net_device *dev); +static void sun3_82586_timeout(struct net_device *dev, unsigned int txqueue); #if 0 static void sun3_82586_dump(struct net_device *,void *); #endif @@ -965,7 +965,7 @@ static void startrecv586(struct net_device *dev) WAIT_4_SCB_CMD_RUC(); /* wait for accept cmd. (no timeout!!) */ } -static void sun3_82586_timeout(struct net_device *dev) +static void sun3_82586_timeout(struct net_device *dev, unsigned int txqueue) { struct priv *p = netdev_priv(dev); #ifndef NO_NOPCOMMANDS diff --git a/drivers/net/ethernet/ibm/ehea/ehea_main.c b/drivers/net/ethernet/ibm/ehea/ehea_main.c index 13e30eba5349..0273fb7a9d01 100644 --- a/drivers/net/ethernet/ibm/ehea/ehea_main.c +++ b/drivers/net/ethernet/ibm/ehea/ehea_main.c @@ -2786,7 +2786,7 @@ out: return; } -static void ehea_tx_watchdog(struct net_device *dev) +static void ehea_tx_watchdog(struct net_device *dev, unsigned int txqueue) { struct ehea_port *port = netdev_priv(dev); diff --git a/drivers/net/ethernet/ibm/emac/core.c b/drivers/net/ethernet/ibm/emac/core.c index 2e40425d8a34..b7fc17756c51 100644 --- a/drivers/net/ethernet/ibm/emac/core.c +++ b/drivers/net/ethernet/ibm/emac/core.c @@ -776,7 +776,7 @@ static void emac_reset_work(struct work_struct *work) mutex_unlock(&dev->link_lock); } -static void emac_tx_timeout(struct net_device *ndev) +static void emac_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct emac_instance *dev = netdev_priv(ndev); diff --git a/drivers/net/ethernet/ibm/ibmvnic.c b/drivers/net/ethernet/ibm/ibmvnic.c index 830791ab4619..c75239d8820f 100644 --- a/drivers/net/ethernet/ibm/ibmvnic.c +++ b/drivers/net/ethernet/ibm/ibmvnic.c @@ -2282,7 +2282,7 @@ err: return -ret; } -static void ibmvnic_tx_timeout(struct net_device *dev) +static void ibmvnic_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct ibmvnic_adapter *adapter = netdev_priv(dev); diff --git a/drivers/net/ethernet/intel/e100.c b/drivers/net/ethernet/intel/e100.c index a65d5a9ba7db..1b8d015ebfb0 100644 --- a/drivers/net/ethernet/intel/e100.c +++ b/drivers/net/ethernet/intel/e100.c @@ -2316,7 +2316,7 @@ static void e100_down(struct nic *nic) e100_rx_clean_list(nic); } -static void e100_tx_timeout(struct net_device *netdev) +static void e100_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct nic *nic = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/e1000/e1000_main.c b/drivers/net/ethernet/intel/e1000/e1000_main.c index aca97b084003..2bced34c19ba 100644 --- a/drivers/net/ethernet/intel/e1000/e1000_main.c +++ b/drivers/net/ethernet/intel/e1000/e1000_main.c @@ -134,7 +134,7 @@ static int e1000_mii_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd); static void e1000_enter_82542_rst(struct e1000_adapter *adapter); static void e1000_leave_82542_rst(struct e1000_adapter *adapter); -static void e1000_tx_timeout(struct net_device *dev); +static void e1000_tx_timeout(struct net_device *dev, unsigned int txqueue); static void e1000_reset_task(struct work_struct *work); static void e1000_smartspeed(struct e1000_adapter *adapter); static int e1000_82547_fifo_workaround(struct e1000_adapter *adapter, @@ -3488,7 +3488,7 @@ exit: * e1000_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void e1000_tx_timeout(struct net_device *netdev) +static void e1000_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct e1000_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/e1000e/netdev.c b/drivers/net/ethernet/intel/e1000e/netdev.c index 7c5b18d87b49..db4ea58bac82 100644 --- a/drivers/net/ethernet/intel/e1000e/netdev.c +++ b/drivers/net/ethernet/intel/e1000e/netdev.c @@ -4721,7 +4721,7 @@ int e1000e_close(struct net_device *netdev) e1000_free_irq(adapter); /* Link status message must follow this format */ - pr_info("%s NIC Link is Down\n", netdev->name); + netdev_info(netdev, "NIC Link is Down\n"); } napi_disable(&adapter->napi); @@ -5071,12 +5071,13 @@ static void e1000_print_link_info(struct e1000_adapter *adapter) u32 ctrl = er32(CTRL); /* Link status message must follow this format for user tools */ - pr_info("%s NIC Link is Up %d Mbps %s Duplex, Flow Control: %s\n", - adapter->netdev->name, adapter->link_speed, - adapter->link_duplex == FULL_DUPLEX ? "Full" : "Half", - (ctrl & E1000_CTRL_TFCE) && (ctrl & E1000_CTRL_RFCE) ? "Rx/Tx" : - (ctrl & E1000_CTRL_RFCE) ? "Rx" : - (ctrl & E1000_CTRL_TFCE) ? "Tx" : "None"); + netdev_info(adapter->netdev, + "NIC Link is Up %d Mbps %s Duplex, Flow Control: %s\n", + adapter->link_speed, + adapter->link_duplex == FULL_DUPLEX ? "Full" : "Half", + (ctrl & E1000_CTRL_TFCE) && (ctrl & E1000_CTRL_RFCE) ? "Rx/Tx" : + (ctrl & E1000_CTRL_RFCE) ? "Rx" : + (ctrl & E1000_CTRL_TFCE) ? "Tx" : "None"); } static bool e1000e_has_link(struct e1000_adapter *adapter) @@ -5319,7 +5320,7 @@ static void e1000_watchdog_task(struct work_struct *work) adapter->link_speed = 0; adapter->link_duplex = 0; /* Link status message must follow this format */ - pr_info("%s NIC Link is Down\n", adapter->netdev->name); + netdev_info(netdev, "NIC Link is Down\n"); netif_carrier_off(netdev); netif_stop_queue(netdev); if (!test_bit(__E1000_DOWN, &adapter->state)) @@ -5940,7 +5941,7 @@ static netdev_tx_t e1000_xmit_frame(struct sk_buff *skb, * e1000_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void e1000_tx_timeout(struct net_device *netdev) +static void e1000_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct e1000_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c index 68baee04dc58..0637ccadee79 100644 --- a/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c +++ b/drivers/net/ethernet/intel/fm10k/fm10k_netdev.c @@ -696,21 +696,24 @@ static netdev_tx_t fm10k_xmit_frame(struct sk_buff *skb, struct net_device *dev) /** * fm10k_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure + * @txqueue: the index of the Tx queue that timed out **/ -static void fm10k_tx_timeout(struct net_device *netdev) +static void fm10k_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct fm10k_intfc *interface = netdev_priv(netdev); + struct fm10k_ring *tx_ring; bool real_tx_hang = false; - int i; - -#define TX_TIMEO_LIMIT 16000 - for (i = 0; i < interface->num_tx_queues; i++) { - struct fm10k_ring *tx_ring = interface->tx_ring[i]; - if (check_for_tx_hang(tx_ring) && fm10k_check_tx_hang(tx_ring)) - real_tx_hang = true; + if (txqueue >= interface->num_tx_queues) { + WARN(1, "invalid Tx queue index %d", txqueue); + return; } + tx_ring = interface->tx_ring[txqueue]; + if (check_for_tx_hang(tx_ring) && fm10k_check_tx_hang(tx_ring)) + real_tx_hang = true; + +#define TX_TIMEO_LIMIT 16000 if (real_tx_hang) { fm10k_tx_timeout_reset(interface); } else { diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c index 2c5af6d4a6b1..8c3e753bfb9d 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_main.c +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c @@ -301,43 +301,24 @@ void i40e_service_event_schedule(struct i40e_pf *pf) * device is munged, not just the one netdev port, so go for the full * reset. **/ -static void i40e_tx_timeout(struct net_device *netdev) +static void i40e_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct i40e_netdev_priv *np = netdev_priv(netdev); struct i40e_vsi *vsi = np->vsi; struct i40e_pf *pf = vsi->back; struct i40e_ring *tx_ring = NULL; - unsigned int i, hung_queue = 0; + unsigned int i; u32 head, val; pf->tx_timeout_count++; - /* find the stopped queue the same way the stack does */ - for (i = 0; i < netdev->num_tx_queues; i++) { - struct netdev_queue *q; - unsigned long trans_start; - - q = netdev_get_tx_queue(netdev, i); - trans_start = q->trans_start; - if (netif_xmit_stopped(q) && - time_after(jiffies, - (trans_start + netdev->watchdog_timeo))) { - hung_queue = i; - break; - } - } - - if (i == netdev->num_tx_queues) { - netdev_info(netdev, "tx_timeout: no netdev hung queue found\n"); - } else { - /* now that we have an index, find the tx_ring struct */ - for (i = 0; i < vsi->num_queue_pairs; i++) { - if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) { - if (hung_queue == - vsi->tx_rings[i]->queue_index) { - tx_ring = vsi->tx_rings[i]; - break; - } + /* with txqueue index, find the tx_ring struct */ + for (i = 0; i < vsi->num_queue_pairs; i++) { + if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) { + if (txqueue == + vsi->tx_rings[i]->queue_index) { + tx_ring = vsi->tx_rings[i]; + break; } } } @@ -363,14 +344,14 @@ static void i40e_tx_timeout(struct net_device *netdev) val = rd32(&pf->hw, I40E_PFINT_DYN_CTL0); netdev_info(netdev, "tx_timeout: VSI_seid: %d, Q %d, NTC: 0x%x, HWB: 0x%x, NTU: 0x%x, TAIL: 0x%x, INT: 0x%x\n", - vsi->seid, hung_queue, tx_ring->next_to_clean, + vsi->seid, txqueue, tx_ring->next_to_clean, head, tx_ring->next_to_use, readl(tx_ring->tail), val); } pf->tx_timeout_last_recovery = jiffies; - netdev_info(netdev, "tx_timeout recovery level %d, hung_queue %d\n", - pf->tx_timeout_recovery_level, hung_queue); + netdev_info(netdev, "tx_timeout recovery level %d, txqueue %d\n", + pf->tx_timeout_recovery_level, txqueue); switch (pf->tx_timeout_recovery_level) { case 1: diff --git a/drivers/net/ethernet/intel/i40e/i40e_xsk.c b/drivers/net/ethernet/intel/i40e/i40e_xsk.c index f73cd917c44f..42058fad6a3c 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_xsk.c +++ b/drivers/net/ethernet/intel/i40e/i40e_xsk.c @@ -269,7 +269,7 @@ static bool i40e_alloc_buffer_zc(struct i40e_ring *rx_ring, bi->handle = xsk_umem_adjust_offset(umem, handle, umem->headroom); - xsk_umem_discard_addr(umem); + xsk_umem_release_addr(umem); return true; } @@ -306,7 +306,7 @@ static bool i40e_alloc_buffer_slow_zc(struct i40e_ring *rx_ring, bi->handle = xsk_umem_adjust_offset(umem, handle, umem->headroom); - xsk_umem_discard_addr_rq(umem); + xsk_umem_release_addr_rq(umem); return true; } diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 8e16be960e96..62fe56ddcb6e 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -159,7 +159,7 @@ void iavf_schedule_reset(struct iavf_adapter *adapter) * iavf_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void iavf_tx_timeout(struct net_device *netdev) +static void iavf_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct iavf_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/ice/Makefile b/drivers/net/ethernet/intel/ice/Makefile index 7cb829132d28..59544b0fc086 100644 --- a/drivers/net/ethernet/intel/ice/Makefile +++ b/drivers/net/ethernet/intel/ice/Makefile @@ -17,7 +17,8 @@ ice-y := ice_main.o \ ice_lib.o \ ice_txrx_lib.o \ ice_txrx.o \ - ice_flex_pipe.o \ + ice_flex_pipe.o \ + ice_flow.o \ ice_ethtool.o ice-$(CONFIG_PCI_IOV) += ice_virtchnl_pf.o ice_sriov.o ice-$(CONFIG_DCB) += ice_dcb.o ice_dcb_nl.o ice_dcb_lib.o diff --git a/drivers/net/ethernet/intel/ice/ice.h b/drivers/net/ethernet/intel/ice/ice.h index f972dce8aebb..cb10abb14e11 100644 --- a/drivers/net/ethernet/intel/ice/ice.h +++ b/drivers/net/ethernet/intel/ice/ice.h @@ -174,6 +174,8 @@ struct ice_sw { struct ice_pf *pf; u16 sw_id; /* switch ID for this switch */ u16 bridge_mode; /* VEB/VEPA/Port Virtualizer */ + struct ice_vsi *dflt_vsi; /* default VSI for this switch */ + u8 dflt_vsi_ena:1; /* true if above dflt_vsi is enabled */ }; enum ice_state { @@ -275,6 +277,7 @@ struct ice_vsi { u8 current_isup:1; /* Sync 'link up' logging */ u8 stat_offsets_loaded:1; u8 vlan_ena:1; + u16 num_vlan; /* queue information */ u8 tx_mapping_mode; /* ICE_MAP_MODE_[CONTIG|SCATTER] */ @@ -462,12 +465,13 @@ static inline void ice_set_ring_xdp(struct ice_ring *ring) static inline struct xdp_umem *ice_xsk_umem(struct ice_ring *ring) { struct xdp_umem **umems = ring->vsi->xsk_umems; - int qid = ring->q_index; + u16 qid = ring->q_index; if (ice_ring_is_xdp(ring)) qid -= ring->vsi->num_xdp_txq; - if (!umems || !umems[qid] || !ice_is_xdp_ena_vsi(ring->vsi)) + if (qid >= ring->vsi->num_xsk_umems || !umems || !umems[qid] || + !ice_is_xdp_ena_vsi(ring->vsi)) return NULL; return umems[qid]; diff --git a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h index 5421fc413f94..4459bc564b11 100644 --- a/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h +++ b/drivers/net/ethernet/intel/ice/ice_adminq_cmd.h @@ -232,6 +232,13 @@ struct ice_aqc_get_sw_cfg_resp { */ #define ICE_AQC_RES_TYPE_VSI_LIST_REP 0x03 #define ICE_AQC_RES_TYPE_VSI_LIST_PRUNE 0x04 +#define ICE_AQC_RES_TYPE_HASH_PROF_BLDR_PROFID 0x60 +#define ICE_AQC_RES_TYPE_HASH_PROF_BLDR_TCAM 0x61 + +#define ICE_AQC_RES_TYPE_FLAG_SCAN_BOTTOM BIT(12) +#define ICE_AQC_RES_TYPE_FLAG_IGNORE_INDEX BIT(13) + +#define ICE_AQC_RES_TYPE_FLAG_DEDICATED 0x00 /* Allocate Resources command (indirect 0x0208) * Free Resources command (indirect 0x0209) @@ -1849,6 +1856,7 @@ enum ice_adminq_opc { /* package commands */ ice_aqc_opc_download_pkg = 0x0C40, + ice_aqc_opc_update_pkg = 0x0C42, ice_aqc_opc_get_pkg_info_list = 0x0C43, /* debug commands */ diff --git a/drivers/net/ethernet/intel/ice/ice_base.c b/drivers/net/ethernet/intel/ice/ice_base.c index 77d6a0291e97..d8e975cceb21 100644 --- a/drivers/net/ethernet/intel/ice/ice_base.c +++ b/drivers/net/ethernet/intel/ice/ice_base.c @@ -93,7 +93,8 @@ static int ice_pf_rxq_wait(struct ice_pf *pf, int pf_q, bool ena) * @vsi: the VSI being configured * @v_idx: index of the vector in the VSI struct * - * We allocate one q_vector. If allocation fails we return -ENOMEM. + * We allocate one q_vector and set default value for ITR setting associated + * with this q_vector. If allocation fails we return -ENOMEM. */ static int ice_vsi_alloc_q_vector(struct ice_vsi *vsi, int v_idx) { @@ -108,6 +109,8 @@ static int ice_vsi_alloc_q_vector(struct ice_vsi *vsi, int v_idx) q_vector->vsi = vsi; q_vector->v_idx = v_idx; + q_vector->tx.itr_setting = ICE_DFLT_TX_ITR; + q_vector->rx.itr_setting = ICE_DFLT_RX_ITR; if (vsi->type == ICE_VSI_VF) goto out; /* only set affinity_mask if the CPU is online */ @@ -299,6 +302,7 @@ int ice_setup_rx_ctx(struct ice_ring *ring) if (ring->vsi->type == ICE_VSI_PF) { if (!xdp_rxq_info_is_reg(&ring->xdp_rxq)) + /* coverity[check_return] */ xdp_rxq_info_reg(&ring->xdp_rxq, ring->netdev, ring->q_index); @@ -323,7 +327,9 @@ int ice_setup_rx_ctx(struct ice_ring *ring) dev_info(&vsi->back->pdev->dev, "Registered XDP mem model MEM_TYPE_ZERO_COPY on Rx ring %d\n", ring->q_index); } else { + ring->zca.free = NULL; if (!xdp_rxq_info_is_reg(&ring->xdp_rxq)) + /* coverity[check_return] */ xdp_rxq_info_reg(&ring->xdp_rxq, ring->netdev, ring->q_index); @@ -674,10 +680,6 @@ void ice_cfg_itr(struct ice_hw *hw, struct ice_q_vector *q_vector) if (q_vector->num_ring_rx) { struct ice_ring_container *rc = &q_vector->rx; - /* if this value is set then don't overwrite with default */ - if (!rc->itr_setting) - rc->itr_setting = ICE_DFLT_RX_ITR; - rc->target_itr = ITR_TO_REG(rc->itr_setting); rc->next_update = jiffies + 1; rc->current_itr = rc->target_itr; @@ -688,10 +690,6 @@ void ice_cfg_itr(struct ice_hw *hw, struct ice_q_vector *q_vector) if (q_vector->num_ring_tx) { struct ice_ring_container *rc = &q_vector->tx; - /* if this value is set then don't overwrite with default */ - if (!rc->itr_setting) - rc->itr_setting = ICE_DFLT_TX_ITR; - rc->target_itr = ITR_TO_REG(rc->itr_setting); rc->next_update = jiffies + 1; rc->current_itr = rc->target_itr; diff --git a/drivers/net/ethernet/intel/ice/ice_common.c b/drivers/net/ethernet/intel/ice/ice_common.c index fb1d930470c7..0207e28c2682 100644 --- a/drivers/net/ethernet/intel/ice/ice_common.c +++ b/drivers/net/ethernet/intel/ice/ice_common.c @@ -4,28 +4,10 @@ #include "ice_common.h" #include "ice_sched.h" #include "ice_adminq_cmd.h" +#include "ice_flow.h" #define ICE_PF_RESET_WAIT_COUNT 200 -#define ICE_PROG_FLEX_ENTRY(hw, rxdid, mdid, idx) \ - wr32((hw), GLFLXP_RXDID_FLX_WRD_##idx(rxdid), \ - ((ICE_RX_OPC_MDID << \ - GLFLXP_RXDID_FLX_WRD_##idx##_RXDID_OPCODE_S) & \ - GLFLXP_RXDID_FLX_WRD_##idx##_RXDID_OPCODE_M) | \ - (((mdid) << GLFLXP_RXDID_FLX_WRD_##idx##_PROT_MDID_S) & \ - GLFLXP_RXDID_FLX_WRD_##idx##_PROT_MDID_M)) - -#define ICE_PROG_FLG_ENTRY(hw, rxdid, flg_0, flg_1, flg_2, flg_3, idx) \ - wr32((hw), GLFLXP_RXDID_FLAGS(rxdid, idx), \ - (((flg_0) << GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_S) & \ - GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_M) | \ - (((flg_1) << GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_1_S) & \ - GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_1_M) | \ - (((flg_2) << GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_2_S) & \ - GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_2_M) | \ - (((flg_3) << GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_3_S) & \ - GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_3_M)) - /** * ice_set_mac_type - Sets MAC type * @hw: pointer to the HW structure @@ -348,88 +330,6 @@ ice_aq_get_link_info(struct ice_port_info *pi, bool ena_lse, } /** - * ice_init_flex_flags - * @hw: pointer to the hardware structure - * @prof_id: Rx Descriptor Builder profile ID - * - * Function to initialize Rx flex flags - */ -static void ice_init_flex_flags(struct ice_hw *hw, enum ice_rxdid prof_id) -{ - u8 idx = 0; - - /* Flex-flag fields (0-2) are programmed with FLG64 bits with layout: - * flexiflags0[5:0] - TCP flags, is_packet_fragmented, is_packet_UDP_GRE - * flexiflags1[3:0] - Not used for flag programming - * flexiflags2[7:0] - Tunnel and VLAN types - * 2 invalid fields in last index - */ - switch (prof_id) { - /* Rx flex flags are currently programmed for the NIC profiles only. - * Different flag bit programming configurations can be added per - * profile as needed. - */ - case ICE_RXDID_FLEX_NIC: - case ICE_RXDID_FLEX_NIC_2: - ICE_PROG_FLG_ENTRY(hw, prof_id, ICE_FLG_PKT_FRG, - ICE_FLG_UDP_GRE, ICE_FLG_PKT_DSI, - ICE_FLG_FIN, idx++); - /* flex flag 1 is not used for flexi-flag programming, skipping - * these four FLG64 bits. - */ - ICE_PROG_FLG_ENTRY(hw, prof_id, ICE_FLG_SYN, ICE_FLG_RST, - ICE_FLG_PKT_DSI, ICE_FLG_PKT_DSI, idx++); - ICE_PROG_FLG_ENTRY(hw, prof_id, ICE_FLG_PKT_DSI, - ICE_FLG_PKT_DSI, ICE_FLG_EVLAN_x8100, - ICE_FLG_EVLAN_x9100, idx++); - ICE_PROG_FLG_ENTRY(hw, prof_id, ICE_FLG_VLAN_x8100, - ICE_FLG_TNL_VLAN, ICE_FLG_TNL_MAC, - ICE_FLG_TNL0, idx++); - ICE_PROG_FLG_ENTRY(hw, prof_id, ICE_FLG_TNL1, ICE_FLG_TNL2, - ICE_FLG_PKT_DSI, ICE_FLG_PKT_DSI, idx); - break; - - default: - ice_debug(hw, ICE_DBG_INIT, - "Flag programming for profile ID %d not supported\n", - prof_id); - } -} - -/** - * ice_init_flex_flds - * @hw: pointer to the hardware structure - * @prof_id: Rx Descriptor Builder profile ID - * - * Function to initialize flex descriptors - */ -static void ice_init_flex_flds(struct ice_hw *hw, enum ice_rxdid prof_id) -{ - enum ice_flex_rx_mdid mdid; - - switch (prof_id) { - case ICE_RXDID_FLEX_NIC: - case ICE_RXDID_FLEX_NIC_2: - ICE_PROG_FLEX_ENTRY(hw, prof_id, ICE_RX_MDID_HASH_LOW, 0); - ICE_PROG_FLEX_ENTRY(hw, prof_id, ICE_RX_MDID_HASH_HIGH, 1); - ICE_PROG_FLEX_ENTRY(hw, prof_id, ICE_RX_MDID_FLOW_ID_LOWER, 2); - - mdid = (prof_id == ICE_RXDID_FLEX_NIC_2) ? - ICE_RX_MDID_SRC_VSI : ICE_RX_MDID_FLOW_ID_HIGH; - - ICE_PROG_FLEX_ENTRY(hw, prof_id, mdid, 3); - - ice_init_flex_flags(hw, prof_id); - break; - - default: - ice_debug(hw, ICE_DBG_INIT, - "Field init for profile ID %d not supported\n", - prof_id); - } -} - -/** * ice_init_fltr_mgmt_struct - initializes filter management list and locks * @hw: pointer to the HW struct */ @@ -882,9 +782,6 @@ enum ice_status ice_init_hw(struct ice_hw *hw) if (status) goto err_unroll_fltr_mgmt_struct; - - ice_init_flex_flds(hw, ICE_RXDID_FLEX_NIC); - ice_init_flex_flds(hw, ICE_RXDID_FLEX_NIC_2); status = ice_init_hw_tbls(hw); if (status) goto err_unroll_fltr_mgmt_struct; @@ -1601,6 +1498,114 @@ void ice_release_res(struct ice_hw *hw, enum ice_aq_res_ids res) } /** + * ice_aq_alloc_free_res - command to allocate/free resources + * @hw: pointer to the HW struct + * @num_entries: number of resource entries in buffer + * @buf: Indirect buffer to hold data parameters and response + * @buf_size: size of buffer for indirect commands + * @opc: pass in the command opcode + * @cd: pointer to command details structure or NULL + * + * Helper function to allocate/free resources using the admin queue commands + */ +enum ice_status +ice_aq_alloc_free_res(struct ice_hw *hw, u16 num_entries, + struct ice_aqc_alloc_free_res_elem *buf, u16 buf_size, + enum ice_adminq_opc opc, struct ice_sq_cd *cd) +{ + struct ice_aqc_alloc_free_res_cmd *cmd; + struct ice_aq_desc desc; + + cmd = &desc.params.sw_res_ctrl; + + if (!buf) + return ICE_ERR_PARAM; + + if (buf_size < (num_entries * sizeof(buf->elem[0]))) + return ICE_ERR_PARAM; + + ice_fill_dflt_direct_cmd_desc(&desc, opc); + + desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD); + + cmd->num_entries = cpu_to_le16(num_entries); + + return ice_aq_send_cmd(hw, &desc, buf, buf_size, cd); +} + +/** + * ice_alloc_hw_res - allocate resource + * @hw: pointer to the HW struct + * @type: type of resource + * @num: number of resources to allocate + * @btm: allocate from bottom + * @res: pointer to array that will receive the resources + */ +enum ice_status +ice_alloc_hw_res(struct ice_hw *hw, u16 type, u16 num, bool btm, u16 *res) +{ + struct ice_aqc_alloc_free_res_elem *buf; + enum ice_status status; + u16 buf_len; + + buf_len = struct_size(buf, elem, num - 1); + buf = kzalloc(buf_len, GFP_KERNEL); + if (!buf) + return ICE_ERR_NO_MEMORY; + + /* Prepare buffer to allocate resource. */ + buf->num_elems = cpu_to_le16(num); + buf->res_type = cpu_to_le16(type | ICE_AQC_RES_TYPE_FLAG_DEDICATED | + ICE_AQC_RES_TYPE_FLAG_IGNORE_INDEX); + if (btm) + buf->res_type |= cpu_to_le16(ICE_AQC_RES_TYPE_FLAG_SCAN_BOTTOM); + + status = ice_aq_alloc_free_res(hw, 1, buf, buf_len, + ice_aqc_opc_alloc_res, NULL); + if (status) + goto ice_alloc_res_exit; + + memcpy(res, buf->elem, sizeof(buf->elem) * num); + +ice_alloc_res_exit: + kfree(buf); + return status; +} + +/** + * ice_free_hw_res - free allocated HW resource + * @hw: pointer to the HW struct + * @type: type of resource to free + * @num: number of resources + * @res: pointer to array that contains the resources to free + */ +enum ice_status +ice_free_hw_res(struct ice_hw *hw, u16 type, u16 num, u16 *res) +{ + struct ice_aqc_alloc_free_res_elem *buf; + enum ice_status status; + u16 buf_len; + + buf_len = struct_size(buf, elem, num - 1); + buf = kzalloc(buf_len, GFP_KERNEL); + if (!buf) + return ICE_ERR_NO_MEMORY; + + /* Prepare buffer to free resource. */ + buf->num_elems = cpu_to_le16(num); + buf->res_type = cpu_to_le16(type); + memcpy(buf->elem, res, sizeof(buf->elem) * num); + + status = ice_aq_alloc_free_res(hw, num, buf, buf_len, + ice_aqc_opc_free_res, NULL); + if (status) + ice_debug(hw, ICE_DBG_SW, "CQ CMD Buffer:\n"); + + kfree(buf); + return status; +} + +/** * ice_get_num_per_func - determine number of resources per PF * @hw: pointer to the HW structure * @max: value to be evenly split between each PF @@ -3510,7 +3515,10 @@ enum ice_status ice_replay_vsi(struct ice_hw *hw, u16 vsi_handle) if (status) return status; } - + /* Replay per VSI all RSS configurations */ + status = ice_replay_rss_cfg(hw, vsi_handle); + if (status) + return status; /* Replay per VSI all filters */ status = ice_replay_vsi_all_fltr(hw, vsi_handle); return status; diff --git a/drivers/net/ethernet/intel/ice/ice_common.h b/drivers/net/ethernet/intel/ice/ice_common.h index b22aa561e253..b5c013fdaaf9 100644 --- a/drivers/net/ethernet/intel/ice/ice_common.h +++ b/drivers/net/ethernet/intel/ice/ice_common.h @@ -34,10 +34,18 @@ enum ice_status ice_acquire_res(struct ice_hw *hw, enum ice_aq_res_ids res, enum ice_aq_res_access_type access, u32 timeout); void ice_release_res(struct ice_hw *hw, enum ice_aq_res_ids res); +enum ice_status +ice_alloc_hw_res(struct ice_hw *hw, u16 type, u16 num, bool btm, u16 *res); +enum ice_status +ice_free_hw_res(struct ice_hw *hw, u16 type, u16 num, u16 *res); enum ice_status ice_init_nvm(struct ice_hw *hw); enum ice_status ice_read_sr_buf(struct ice_hw *hw, u16 offset, u16 *words, u16 *data); enum ice_status +ice_aq_alloc_free_res(struct ice_hw *hw, u16 num_entries, + struct ice_aqc_alloc_free_res_elem *buf, u16 buf_size, + enum ice_adminq_opc opc, struct ice_sq_cd *cd); +enum ice_status ice_sq_send_cmd(struct ice_hw *hw, struct ice_ctl_q_info *cq, struct ice_aq_desc *desc, void *buf, u16 buf_size, struct ice_sq_cd *cd); diff --git a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c index d3d3ec29def9..0664e5b8d130 100644 --- a/drivers/net/ethernet/intel/ice/ice_dcb_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_dcb_lib.c @@ -396,6 +396,12 @@ dcb_error: prev_cfg->etscfg.tcbwtable[0] = ICE_TC_MAX_BW; prev_cfg->etscfg.tsatable[0] = ICE_IEEE_TSA_ETS; memcpy(&prev_cfg->etsrec, &prev_cfg->etscfg, sizeof(prev_cfg->etsrec)); + /* Coverity warns the return code of ice_pf_dcb_cfg() is not checked + * here as is done for other calls to that function. That check is + * not necessary since this is in this function's error cleanup path. + * Suppress the Coverity warning with the following comment... + */ + /* coverity[check_return] */ ice_pf_dcb_cfg(pf, prev_cfg, false); kfree(prev_cfg); } diff --git a/drivers/net/ethernet/intel/ice/ice_devids.h b/drivers/net/ethernet/intel/ice/ice_devids.h index f8d5c661d0ba..ce63017c56c7 100644 --- a/drivers/net/ethernet/intel/ice/ice_devids.h +++ b/drivers/net/ethernet/intel/ice/ice_devids.h @@ -11,5 +11,23 @@ #define ICE_DEV_ID_E810C_QSFP 0x1592 /* Intel(R) Ethernet Controller E810-C for SFP */ #define ICE_DEV_ID_E810C_SFP 0x1593 +/* Intel(R) Ethernet Connection E822-C for backplane */ +#define ICE_DEV_ID_E822C_BACKPLANE 0x1890 +/* Intel(R) Ethernet Connection E822-C for QSFP */ +#define ICE_DEV_ID_E822C_QSFP 0x1891 +/* Intel(R) Ethernet Connection E822-C for SFP */ +#define ICE_DEV_ID_E822C_SFP 0x1892 +/* Intel(R) Ethernet Connection E822-C/X557-AT 10GBASE-T */ +#define ICE_DEV_ID_E822C_10G_BASE_T 0x1893 +/* Intel(R) Ethernet Connection E822-C 1GbE */ +#define ICE_DEV_ID_E822C_SGMII 0x1894 +/* Intel(R) Ethernet Connection E822-X for backplane */ +#define ICE_DEV_ID_E822X_BACKPLANE 0x1897 +/* Intel(R) Ethernet Connection E822-L for SFP */ +#define ICE_DEV_ID_E822L_SFP 0x1898 +/* Intel(R) Ethernet Connection E822-L/X557-AT 10GBASE-T */ +#define ICE_DEV_ID_E822L_10G_BASE_T 0x1899 +/* Intel(R) Ethernet Connection E822-L 1GbE */ +#define ICE_DEV_ID_E822L_SGMII 0x189A #endif /* _ICE_DEVIDS_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_ethtool.c b/drivers/net/ethernet/intel/ice/ice_ethtool.c index 9ebd93e79aeb..90c6a3ca20c9 100644 --- a/drivers/net/ethernet/intel/ice/ice_ethtool.c +++ b/drivers/net/ethernet/intel/ice/ice_ethtool.c @@ -4,6 +4,7 @@ /* ethtool support for ice */ #include "ice.h" +#include "ice_flow.h" #include "ice_lib.h" #include "ice_dcb_lib.h" @@ -283,12 +284,15 @@ out: */ static bool ice_active_vfs(struct ice_pf *pf) { - struct ice_vf *vf = pf->vf; int i; - for (i = 0; i < pf->num_alloc_vfs; i++, vf++) + ice_for_each_vf(pf, i) { + struct ice_vf *vf = &pf->vf[i]; + if (test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states)) return true; + } + return false; } @@ -2531,6 +2535,243 @@ done: } /** + * ice_parse_hdrs - parses headers from RSS hash input + * @nfc: ethtool rxnfc command + * + * This function parses the rxnfc command and returns intended + * header types for RSS configuration + */ +static u32 ice_parse_hdrs(struct ethtool_rxnfc *nfc) +{ + u32 hdrs = ICE_FLOW_SEG_HDR_NONE; + + switch (nfc->flow_type) { + case TCP_V4_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_IPV4; + break; + case UDP_V4_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_IPV4; + break; + case SCTP_V4_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_SCTP | ICE_FLOW_SEG_HDR_IPV4; + break; + case TCP_V6_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_IPV6; + break; + case UDP_V6_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_IPV6; + break; + case SCTP_V6_FLOW: + hdrs |= ICE_FLOW_SEG_HDR_SCTP | ICE_FLOW_SEG_HDR_IPV6; + break; + default: + break; + } + return hdrs; +} + +#define ICE_FLOW_HASH_FLD_IPV4_SA BIT_ULL(ICE_FLOW_FIELD_IDX_IPV4_SA) +#define ICE_FLOW_HASH_FLD_IPV6_SA BIT_ULL(ICE_FLOW_FIELD_IDX_IPV6_SA) +#define ICE_FLOW_HASH_FLD_IPV4_DA BIT_ULL(ICE_FLOW_FIELD_IDX_IPV4_DA) +#define ICE_FLOW_HASH_FLD_IPV6_DA BIT_ULL(ICE_FLOW_FIELD_IDX_IPV6_DA) +#define ICE_FLOW_HASH_FLD_TCP_SRC_PORT BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_SRC_PORT) +#define ICE_FLOW_HASH_FLD_TCP_DST_PORT BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_DST_PORT) +#define ICE_FLOW_HASH_FLD_UDP_SRC_PORT BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_SRC_PORT) +#define ICE_FLOW_HASH_FLD_UDP_DST_PORT BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_DST_PORT) +#define ICE_FLOW_HASH_FLD_SCTP_SRC_PORT \ + BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT) +#define ICE_FLOW_HASH_FLD_SCTP_DST_PORT \ + BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_DST_PORT) + +/** + * ice_parse_hash_flds - parses hash fields from RSS hash input + * @nfc: ethtool rxnfc command + * + * This function parses the rxnfc command and returns intended + * hash fields for RSS configuration + */ +static u64 ice_parse_hash_flds(struct ethtool_rxnfc *nfc) +{ + u64 hfld = ICE_HASH_INVALID; + + if (nfc->data & RXH_IP_SRC || nfc->data & RXH_IP_DST) { + switch (nfc->flow_type) { + case TCP_V4_FLOW: + case UDP_V4_FLOW: + case SCTP_V4_FLOW: + if (nfc->data & RXH_IP_SRC) + hfld |= ICE_FLOW_HASH_FLD_IPV4_SA; + if (nfc->data & RXH_IP_DST) + hfld |= ICE_FLOW_HASH_FLD_IPV4_DA; + break; + case TCP_V6_FLOW: + case UDP_V6_FLOW: + case SCTP_V6_FLOW: + if (nfc->data & RXH_IP_SRC) + hfld |= ICE_FLOW_HASH_FLD_IPV6_SA; + if (nfc->data & RXH_IP_DST) + hfld |= ICE_FLOW_HASH_FLD_IPV6_DA; + break; + default: + break; + } + } + + if (nfc->data & RXH_L4_B_0_1 || nfc->data & RXH_L4_B_2_3) { + switch (nfc->flow_type) { + case TCP_V4_FLOW: + case TCP_V6_FLOW: + if (nfc->data & RXH_L4_B_0_1) + hfld |= ICE_FLOW_HASH_FLD_TCP_SRC_PORT; + if (nfc->data & RXH_L4_B_2_3) + hfld |= ICE_FLOW_HASH_FLD_TCP_DST_PORT; + break; + case UDP_V4_FLOW: + case UDP_V6_FLOW: + if (nfc->data & RXH_L4_B_0_1) + hfld |= ICE_FLOW_HASH_FLD_UDP_SRC_PORT; + if (nfc->data & RXH_L4_B_2_3) + hfld |= ICE_FLOW_HASH_FLD_UDP_DST_PORT; + break; + case SCTP_V4_FLOW: + case SCTP_V6_FLOW: + if (nfc->data & RXH_L4_B_0_1) + hfld |= ICE_FLOW_HASH_FLD_SCTP_SRC_PORT; + if (nfc->data & RXH_L4_B_2_3) + hfld |= ICE_FLOW_HASH_FLD_SCTP_DST_PORT; + break; + default: + break; + } + } + + return hfld; +} + +/** + * ice_set_rss_hash_opt - Enable/Disable flow types for RSS hash + * @vsi: the VSI being configured + * @nfc: ethtool rxnfc command + * + * Returns Success if the flow input set is supported. + */ +static int +ice_set_rss_hash_opt(struct ice_vsi *vsi, struct ethtool_rxnfc *nfc) +{ + struct ice_pf *pf = vsi->back; + enum ice_status status; + struct device *dev; + u64 hashed_flds; + u32 hdrs; + + dev = ice_pf_to_dev(pf); + if (ice_is_safe_mode(pf)) { + dev_dbg(dev, "Advanced RSS disabled. Package download failed, vsi num = %d\n", + vsi->vsi_num); + return -EINVAL; + } + + hashed_flds = ice_parse_hash_flds(nfc); + if (hashed_flds == ICE_HASH_INVALID) { + dev_dbg(dev, "Invalid hash fields, vsi num = %d\n", + vsi->vsi_num); + return -EINVAL; + } + + hdrs = ice_parse_hdrs(nfc); + if (hdrs == ICE_FLOW_SEG_HDR_NONE) { + dev_dbg(dev, "Header type is not valid, vsi num = %d\n", + vsi->vsi_num); + return -EINVAL; + } + + status = ice_add_rss_cfg(&pf->hw, vsi->idx, hashed_flds, hdrs); + if (status) { + dev_dbg(dev, "ice_add_rss_cfg failed, vsi num = %d, error = %d\n", + vsi->vsi_num, status); + return -EINVAL; + } + + return 0; +} + +/** + * ice_get_rss_hash_opt - Retrieve hash fields for a given flow-type + * @vsi: the VSI being configured + * @nfc: ethtool rxnfc command + */ +static void +ice_get_rss_hash_opt(struct ice_vsi *vsi, struct ethtool_rxnfc *nfc) +{ + struct ice_pf *pf = vsi->back; + struct device *dev; + u64 hash_flds; + u32 hdrs; + + dev = ice_pf_to_dev(pf); + + nfc->data = 0; + if (ice_is_safe_mode(pf)) { + dev_dbg(dev, "Advanced RSS disabled. Package download failed, vsi num = %d\n", + vsi->vsi_num); + return; + } + + hdrs = ice_parse_hdrs(nfc); + if (hdrs == ICE_FLOW_SEG_HDR_NONE) { + dev_dbg(dev, "Header type is not valid, vsi num = %d\n", + vsi->vsi_num); + return; + } + + hash_flds = ice_get_rss_cfg(&pf->hw, vsi->idx, hdrs); + if (hash_flds == ICE_HASH_INVALID) { + dev_dbg(dev, "No hash fields found for the given header type, vsi num = %d\n", + vsi->vsi_num); + return; + } + + if (hash_flds & ICE_FLOW_HASH_FLD_IPV4_SA || + hash_flds & ICE_FLOW_HASH_FLD_IPV6_SA) + nfc->data |= (u64)RXH_IP_SRC; + + if (hash_flds & ICE_FLOW_HASH_FLD_IPV4_DA || + hash_flds & ICE_FLOW_HASH_FLD_IPV6_DA) + nfc->data |= (u64)RXH_IP_DST; + + if (hash_flds & ICE_FLOW_HASH_FLD_TCP_SRC_PORT || + hash_flds & ICE_FLOW_HASH_FLD_UDP_SRC_PORT || + hash_flds & ICE_FLOW_HASH_FLD_SCTP_SRC_PORT) + nfc->data |= (u64)RXH_L4_B_0_1; + + if (hash_flds & ICE_FLOW_HASH_FLD_TCP_DST_PORT || + hash_flds & ICE_FLOW_HASH_FLD_UDP_DST_PORT || + hash_flds & ICE_FLOW_HASH_FLD_SCTP_DST_PORT) + nfc->data |= (u64)RXH_L4_B_2_3; +} + +/** + * ice_set_rxnfc - command to set Rx flow rules. + * @netdev: network interface device structure + * @cmd: ethtool rxnfc command + * + * Returns 0 for success and negative values for errors + */ +static int ice_set_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd) +{ + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_vsi *vsi = np->vsi; + + switch (cmd->cmd) { + case ETHTOOL_SRXFH: + return ice_set_rss_hash_opt(vsi, cmd); + default: + break; + } + return -EOPNOTSUPP; +} + +/** * ice_get_rxnfc - command to get Rx flow classification rules * @netdev: network interface device structure * @cmd: ethtool rxnfc command @@ -2551,6 +2792,10 @@ ice_get_rxnfc(struct net_device *netdev, struct ethtool_rxnfc *cmd, cmd->data = vsi->rss_size; ret = 0; break; + case ETHTOOL_GRXFH: + ice_get_rss_hash_opt(vsi, cmd); + ret = 0; + break; default: break; } @@ -3585,6 +3830,53 @@ ice_set_q_coalesce(struct ice_vsi *vsi, struct ethtool_coalesce *ec, int q_num) } /** + * ice_is_coalesce_param_invalid - check for unsupported coalesce parameters + * @netdev: pointer to the netdev associated with this query + * @ec: ethtool structure to fill with driver's coalesce settings + * + * Print netdev info if driver doesn't support one of the parameters + * and return error. When any parameters will be implemented, remove only + * this parameter from param array. + */ +static int +ice_is_coalesce_param_invalid(struct net_device *netdev, + struct ethtool_coalesce *ec) +{ + struct ice_ethtool_not_used { + u32 value; + const char *name; + } param[] = { + {ec->stats_block_coalesce_usecs, "stats-block-usecs"}, + {ec->rate_sample_interval, "sample-interval"}, + {ec->pkt_rate_low, "pkt-rate-low"}, + {ec->pkt_rate_high, "pkt-rate-high"}, + {ec->rx_max_coalesced_frames, "rx-frames"}, + {ec->rx_coalesce_usecs_irq, "rx-usecs-irq"}, + {ec->rx_max_coalesced_frames_irq, "rx-frames-irq"}, + {ec->tx_max_coalesced_frames, "tx-frames"}, + {ec->tx_coalesce_usecs_irq, "tx-usecs-irq"}, + {ec->tx_max_coalesced_frames_irq, "tx-frames-irq"}, + {ec->rx_coalesce_usecs_low, "rx-usecs-low"}, + {ec->rx_max_coalesced_frames_low, "rx-frames-low"}, + {ec->tx_coalesce_usecs_low, "tx-usecs-low"}, + {ec->tx_max_coalesced_frames_low, "tx-frames-low"}, + {ec->rx_max_coalesced_frames_high, "rx-frames-high"}, + {ec->tx_max_coalesced_frames_high, "tx-frames-high"} + }; + int i; + + for (i = 0; i < ARRAY_SIZE(param); i++) { + if (param[i].value) { + netdev_info(netdev, "Setting %s not supported\n", + param[i].name); + return -EINVAL; + } + } + + return 0; +} + +/** * __ice_set_coalesce - set ITR/INTRL values for the device * @netdev: pointer to the netdev associated with this query * @ec: ethtool structure to fill with driver's coalesce settings @@ -3600,6 +3892,9 @@ __ice_set_coalesce(struct net_device *netdev, struct ethtool_coalesce *ec, struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_vsi *vsi = np->vsi; + if (ice_is_coalesce_param_invalid(netdev, ec)) + return -EINVAL; + if (q_num < 0) { int v_idx; @@ -3804,6 +4099,7 @@ static const struct ethtool_ops ice_ethtool_ops = { .set_priv_flags = ice_set_priv_flags, .get_sset_count = ice_get_sset_count, .get_rxnfc = ice_get_rxnfc, + .set_rxnfc = ice_set_rxnfc, .get_ringparam = ice_get_ringparam, .set_ringparam = ice_set_ringparam, .nway_reset = ice_nway_reset, diff --git a/drivers/net/ethernet/intel/ice/ice_flex_pipe.c b/drivers/net/ethernet/intel/ice/ice_flex_pipe.c index cbd53b586c36..99208946224c 100644 --- a/drivers/net/ethernet/intel/ice/ice_flex_pipe.c +++ b/drivers/net/ethernet/intel/ice/ice_flex_pipe.c @@ -3,6 +3,87 @@ #include "ice_common.h" #include "ice_flex_pipe.h" +#include "ice_flow.h" + +static const u32 ice_sect_lkup[ICE_BLK_COUNT][ICE_SECT_COUNT] = { + /* SWITCH */ + { + ICE_SID_XLT0_SW, + ICE_SID_XLT_KEY_BUILDER_SW, + ICE_SID_XLT1_SW, + ICE_SID_XLT2_SW, + ICE_SID_PROFID_TCAM_SW, + ICE_SID_PROFID_REDIR_SW, + ICE_SID_FLD_VEC_SW, + ICE_SID_CDID_KEY_BUILDER_SW, + ICE_SID_CDID_REDIR_SW + }, + + /* ACL */ + { + ICE_SID_XLT0_ACL, + ICE_SID_XLT_KEY_BUILDER_ACL, + ICE_SID_XLT1_ACL, + ICE_SID_XLT2_ACL, + ICE_SID_PROFID_TCAM_ACL, + ICE_SID_PROFID_REDIR_ACL, + ICE_SID_FLD_VEC_ACL, + ICE_SID_CDID_KEY_BUILDER_ACL, + ICE_SID_CDID_REDIR_ACL + }, + + /* FD */ + { + ICE_SID_XLT0_FD, + ICE_SID_XLT_KEY_BUILDER_FD, + ICE_SID_XLT1_FD, + ICE_SID_XLT2_FD, + ICE_SID_PROFID_TCAM_FD, + ICE_SID_PROFID_REDIR_FD, + ICE_SID_FLD_VEC_FD, + ICE_SID_CDID_KEY_BUILDER_FD, + ICE_SID_CDID_REDIR_FD + }, + + /* RSS */ + { + ICE_SID_XLT0_RSS, + ICE_SID_XLT_KEY_BUILDER_RSS, + ICE_SID_XLT1_RSS, + ICE_SID_XLT2_RSS, + ICE_SID_PROFID_TCAM_RSS, + ICE_SID_PROFID_REDIR_RSS, + ICE_SID_FLD_VEC_RSS, + ICE_SID_CDID_KEY_BUILDER_RSS, + ICE_SID_CDID_REDIR_RSS + }, + + /* PE */ + { + ICE_SID_XLT0_PE, + ICE_SID_XLT_KEY_BUILDER_PE, + ICE_SID_XLT1_PE, + ICE_SID_XLT2_PE, + ICE_SID_PROFID_TCAM_PE, + ICE_SID_PROFID_REDIR_PE, + ICE_SID_FLD_VEC_PE, + ICE_SID_CDID_KEY_BUILDER_PE, + ICE_SID_CDID_REDIR_PE + } +}; + +/** + * ice_sect_id - returns section ID + * @blk: block type + * @sect: section type + * + * This helper function returns the proper section ID given a block type and a + * section type. + */ +static u32 ice_sect_id(enum ice_block blk, enum ice_sect sect) +{ + return ice_sect_lkup[blk][sect]; +} /** * ice_pkg_val_buf @@ -158,6 +239,176 @@ ice_pkg_enum_section(struct ice_seg *ice_seg, struct ice_pkg_enum *state, return state->sect; } +/* Key creation */ + +#define ICE_DC_KEY 0x1 /* don't care */ +#define ICE_DC_KEYINV 0x1 +#define ICE_NM_KEY 0x0 /* never match */ +#define ICE_NM_KEYINV 0x0 +#define ICE_0_KEY 0x1 /* match 0 */ +#define ICE_0_KEYINV 0x0 +#define ICE_1_KEY 0x0 /* match 1 */ +#define ICE_1_KEYINV 0x1 + +/** + * ice_gen_key_word - generate 16-bits of a key/mask word + * @val: the value + * @valid: valid bits mask (change only the valid bits) + * @dont_care: don't care mask + * @nvr_mtch: never match mask + * @key: pointer to an array of where the resulting key portion + * @key_inv: pointer to an array of where the resulting key invert portion + * + * This function generates 16-bits from a 8-bit value, an 8-bit don't care mask + * and an 8-bit never match mask. The 16-bits of output are divided into 8 bits + * of key and 8 bits of key invert. + * + * '0' = b01, always match a 0 bit + * '1' = b10, always match a 1 bit + * '?' = b11, don't care bit (always matches) + * '~' = b00, never match bit + * + * Input: + * val: b0 1 0 1 0 1 + * dont_care: b0 0 1 1 0 0 + * never_mtch: b0 0 0 0 1 1 + * ------------------------------ + * Result: key: b01 10 11 11 00 00 + */ +static enum ice_status +ice_gen_key_word(u8 val, u8 valid, u8 dont_care, u8 nvr_mtch, u8 *key, + u8 *key_inv) +{ + u8 in_key = *key, in_key_inv = *key_inv; + u8 i; + + /* 'dont_care' and 'nvr_mtch' masks cannot overlap */ + if ((dont_care ^ nvr_mtch) != (dont_care | nvr_mtch)) + return ICE_ERR_CFG; + + *key = 0; + *key_inv = 0; + + /* encode the 8 bits into 8-bit key and 8-bit key invert */ + for (i = 0; i < 8; i++) { + *key >>= 1; + *key_inv >>= 1; + + if (!(valid & 0x1)) { /* change only valid bits */ + *key |= (in_key & 0x1) << 7; + *key_inv |= (in_key_inv & 0x1) << 7; + } else if (dont_care & 0x1) { /* don't care bit */ + *key |= ICE_DC_KEY << 7; + *key_inv |= ICE_DC_KEYINV << 7; + } else if (nvr_mtch & 0x1) { /* never match bit */ + *key |= ICE_NM_KEY << 7; + *key_inv |= ICE_NM_KEYINV << 7; + } else if (val & 0x01) { /* exact 1 match */ + *key |= ICE_1_KEY << 7; + *key_inv |= ICE_1_KEYINV << 7; + } else { /* exact 0 match */ + *key |= ICE_0_KEY << 7; + *key_inv |= ICE_0_KEYINV << 7; + } + + dont_care >>= 1; + nvr_mtch >>= 1; + valid >>= 1; + val >>= 1; + in_key >>= 1; + in_key_inv >>= 1; + } + + return 0; +} + +/** + * ice_bits_max_set - determine if the number of bits set is within a maximum + * @mask: pointer to the byte array which is the mask + * @size: the number of bytes in the mask + * @max: the max number of set bits + * + * This function determines if there are at most 'max' number of bits set in an + * array. Returns true if the number for bits set is <= max or will return false + * otherwise. + */ +static bool ice_bits_max_set(const u8 *mask, u16 size, u16 max) +{ + u16 count = 0; + u16 i; + + /* check each byte */ + for (i = 0; i < size; i++) { + /* if 0, go to next byte */ + if (!mask[i]) + continue; + + /* We know there is at least one set bit in this byte because of + * the above check; if we already have found 'max' number of + * bits set, then we can return failure now. + */ + if (count == max) + return false; + + /* count the bits in this byte, checking threshold */ + count += hweight8(mask[i]); + if (count > max) + return false; + } + + return true; +} + +/** + * ice_set_key - generate a variable sized key with multiples of 16-bits + * @key: pointer to where the key will be stored + * @size: the size of the complete key in bytes (must be even) + * @val: array of 8-bit values that makes up the value portion of the key + * @upd: array of 8-bit masks that determine what key portion to update + * @dc: array of 8-bit masks that make up the don't care mask + * @nm: array of 8-bit masks that make up the never match mask + * @off: the offset of the first byte in the key to update + * @len: the number of bytes in the key update + * + * This function generates a key from a value, a don't care mask and a never + * match mask. + * upd, dc, and nm are optional parameters, and can be NULL: + * upd == NULL --> udp mask is all 1's (update all bits) + * dc == NULL --> dc mask is all 0's (no don't care bits) + * nm == NULL --> nm mask is all 0's (no never match bits) + */ +static enum ice_status +ice_set_key(u8 *key, u16 size, u8 *val, u8 *upd, u8 *dc, u8 *nm, u16 off, + u16 len) +{ + u16 half_size; + u16 i; + + /* size must be a multiple of 2 bytes. */ + if (size % 2) + return ICE_ERR_CFG; + + half_size = size / 2; + if (off + len > half_size) + return ICE_ERR_CFG; + + /* Make sure at most one bit is set in the never match mask. Having more + * than one never match mask bit set will cause HW to consume excessive + * power otherwise; this is a power management efficiency check. + */ +#define ICE_NVR_MTCH_BITS_MAX 1 + if (nm && !ice_bits_max_set(nm, len, ICE_NVR_MTCH_BITS_MAX)) + return ICE_ERR_CFG; + + for (i = 0; i < len; i++) + if (ice_gen_key_word(val[i], upd ? upd[i] : 0xff, + dc ? dc[i] : 0, nm ? nm[i] : 0, + key + off + i, key + half_size + off + i)) + return ICE_ERR_CFG; + + return 0; +} + /** * ice_acquire_global_cfg_lock * @hw: pointer to the HW structure @@ -205,6 +456,31 @@ static void ice_release_global_cfg_lock(struct ice_hw *hw) } /** + * ice_acquire_change_lock + * @hw: pointer to the HW structure + * @access: access type (read or write) + * + * This function will request ownership of the change lock. + */ +static enum ice_status +ice_acquire_change_lock(struct ice_hw *hw, enum ice_aq_res_access_type access) +{ + return ice_acquire_res(hw, ICE_CHANGE_LOCK_RES_ID, access, + ICE_CHANGE_LOCK_TIMEOUT); +} + +/** + * ice_release_change_lock + * @hw: pointer to the HW structure + * + * This function will release the change lock using the proper Admin Command. + */ +static void ice_release_change_lock(struct ice_hw *hw) +{ + ice_release_res(hw, ICE_CHANGE_LOCK_RES_ID); +} + +/** * ice_aq_download_pkg * @hw: pointer to the hardware structure * @pkg_buf: the package buffer to transfer @@ -253,6 +529,54 @@ ice_aq_download_pkg(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf, } /** + * ice_aq_update_pkg + * @hw: pointer to the hardware structure + * @pkg_buf: the package cmd buffer + * @buf_size: the size of the package cmd buffer + * @last_buf: last buffer indicator + * @error_offset: returns error offset + * @error_info: returns error information + * @cd: pointer to command details structure or NULL + * + * Update Package (0x0C42) + */ +static enum ice_status +ice_aq_update_pkg(struct ice_hw *hw, struct ice_buf_hdr *pkg_buf, u16 buf_size, + bool last_buf, u32 *error_offset, u32 *error_info, + struct ice_sq_cd *cd) +{ + struct ice_aqc_download_pkg *cmd; + struct ice_aq_desc desc; + enum ice_status status; + + if (error_offset) + *error_offset = 0; + if (error_info) + *error_info = 0; + + cmd = &desc.params.download_pkg; + ice_fill_dflt_direct_cmd_desc(&desc, ice_aqc_opc_update_pkg); + desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD); + + if (last_buf) + cmd->flags |= ICE_AQC_DOWNLOAD_PKG_LAST_BUF; + + status = ice_aq_send_cmd(hw, &desc, pkg_buf, buf_size, cd); + if (status == ICE_ERR_AQ_ERROR) { + /* Read error from buffer only when the FW returned an error */ + struct ice_aqc_download_pkg_resp *resp; + + resp = (struct ice_aqc_download_pkg_resp *)pkg_buf; + if (error_offset) + *error_offset = le32_to_cpu(resp->error_offset); + if (error_info) + *error_info = le32_to_cpu(resp->error_info); + } + + return status; +} + +/** * ice_find_seg_in_pkg * @hw: pointer to the hardware structure * @seg_type: the segment type to search for (i.e., SEGMENT_TYPE_CPK) @@ -287,6 +611,44 @@ ice_find_seg_in_pkg(struct ice_hw *hw, u32 seg_type, } /** + * ice_update_pkg + * @hw: pointer to the hardware structure + * @bufs: pointer to an array of buffers + * @count: the number of buffers in the array + * + * Obtains change lock and updates package. + */ +static enum ice_status +ice_update_pkg(struct ice_hw *hw, struct ice_buf *bufs, u32 count) +{ + enum ice_status status; + u32 offset, info, i; + + status = ice_acquire_change_lock(hw, ICE_RES_WRITE); + if (status) + return status; + + for (i = 0; i < count; i++) { + struct ice_buf_hdr *bh = (struct ice_buf_hdr *)(bufs + i); + bool last = ((i + 1) == count); + + status = ice_aq_update_pkg(hw, bh, le16_to_cpu(bh->data_end), + last, &offset, &info, NULL); + + if (status) { + ice_debug(hw, ICE_DBG_PKG, + "Update pkg failed: err %d off %d inf %d\n", + status, offset, info); + break; + } + } + + ice_release_change_lock(hw); + + return status; +} + +/** * ice_dwnld_cfg_bufs * @hw: pointer to the hardware structure * @bufs: pointer to an array of buffers @@ -767,6 +1129,169 @@ enum ice_status ice_copy_and_init_pkg(struct ice_hw *hw, const u8 *buf, u32 len) return status; } +/** + * ice_pkg_buf_alloc + * @hw: pointer to the HW structure + * + * Allocates a package buffer and returns a pointer to the buffer header. + * Note: all package contents must be in Little Endian form. + */ +static struct ice_buf_build *ice_pkg_buf_alloc(struct ice_hw *hw) +{ + struct ice_buf_build *bld; + struct ice_buf_hdr *buf; + + bld = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*bld), GFP_KERNEL); + if (!bld) + return NULL; + + buf = (struct ice_buf_hdr *)bld; + buf->data_end = cpu_to_le16(offsetof(struct ice_buf_hdr, + section_entry)); + return bld; +} + +/** + * ice_pkg_buf_free + * @hw: pointer to the HW structure + * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc()) + * + * Frees a package buffer + */ +static void ice_pkg_buf_free(struct ice_hw *hw, struct ice_buf_build *bld) +{ + devm_kfree(ice_hw_to_dev(hw), bld); +} + +/** + * ice_pkg_buf_reserve_section + * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc()) + * @count: the number of sections to reserve + * + * Reserves one or more section table entries in a package buffer. This routine + * can be called multiple times as long as they are made before calling + * ice_pkg_buf_alloc_section(). Once ice_pkg_buf_alloc_section() + * is called once, the number of sections that can be allocated will not be able + * to be increased; not using all reserved sections is fine, but this will + * result in some wasted space in the buffer. + * Note: all package contents must be in Little Endian form. + */ +static enum ice_status +ice_pkg_buf_reserve_section(struct ice_buf_build *bld, u16 count) +{ + struct ice_buf_hdr *buf; + u16 section_count; + u16 data_end; + + if (!bld) + return ICE_ERR_PARAM; + + buf = (struct ice_buf_hdr *)&bld->buf; + + /* already an active section, can't increase table size */ + section_count = le16_to_cpu(buf->section_count); + if (section_count > 0) + return ICE_ERR_CFG; + + if (bld->reserved_section_table_entries + count > ICE_MAX_S_COUNT) + return ICE_ERR_CFG; + bld->reserved_section_table_entries += count; + + data_end = le16_to_cpu(buf->data_end) + + (count * sizeof(buf->section_entry[0])); + buf->data_end = cpu_to_le16(data_end); + + return 0; +} + +/** + * ice_pkg_buf_alloc_section + * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc()) + * @type: the section type value + * @size: the size of the section to reserve (in bytes) + * + * Reserves memory in the buffer for a section's content and updates the + * buffers' status accordingly. This routine returns a pointer to the first + * byte of the section start within the buffer, which is used to fill in the + * section contents. + * Note: all package contents must be in Little Endian form. + */ +static void * +ice_pkg_buf_alloc_section(struct ice_buf_build *bld, u32 type, u16 size) +{ + struct ice_buf_hdr *buf; + u16 sect_count; + u16 data_end; + + if (!bld || !type || !size) + return NULL; + + buf = (struct ice_buf_hdr *)&bld->buf; + + /* check for enough space left in buffer */ + data_end = le16_to_cpu(buf->data_end); + + /* section start must align on 4 byte boundary */ + data_end = ALIGN(data_end, 4); + + if ((data_end + size) > ICE_MAX_S_DATA_END) + return NULL; + + /* check for more available section table entries */ + sect_count = le16_to_cpu(buf->section_count); + if (sect_count < bld->reserved_section_table_entries) { + void *section_ptr = ((u8 *)buf) + data_end; + + buf->section_entry[sect_count].offset = cpu_to_le16(data_end); + buf->section_entry[sect_count].size = cpu_to_le16(size); + buf->section_entry[sect_count].type = cpu_to_le32(type); + + data_end += size; + buf->data_end = cpu_to_le16(data_end); + + buf->section_count = cpu_to_le16(sect_count + 1); + return section_ptr; + } + + /* no free section table entries */ + return NULL; +} + +/** + * ice_pkg_buf_get_active_sections + * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc()) + * + * Returns the number of active sections. Before using the package buffer + * in an update package command, the caller should make sure that there is at + * least one active section - otherwise, the buffer is not legal and should + * not be used. + * Note: all package contents must be in Little Endian form. + */ +static u16 ice_pkg_buf_get_active_sections(struct ice_buf_build *bld) +{ + struct ice_buf_hdr *buf; + + if (!bld) + return 0; + + buf = (struct ice_buf_hdr *)&bld->buf; + return le16_to_cpu(buf->section_count); +} + +/** + * ice_pkg_buf + * @bld: pointer to pkg build (allocated by ice_pkg_buf_alloc()) + * + * Return a pointer to the buffer's header + */ +static struct ice_buf *ice_pkg_buf(struct ice_buf_build *bld) +{ + if (!bld) + return NULL; + + return &bld->buf; +} + /* PTG Management */ /** @@ -951,6 +1476,48 @@ enum ice_sid_all { ICE_SID_OFF_COUNT, }; +/* Characteristic handling */ + +/** + * ice_match_prop_lst - determine if properties of two lists match + * @list1: first properties list + * @list2: second properties list + * + * Count, cookies and the order must match in order to be considered equivalent. + */ +static bool +ice_match_prop_lst(struct list_head *list1, struct list_head *list2) +{ + struct ice_vsig_prof *tmp1; + struct ice_vsig_prof *tmp2; + u16 chk_count = 0; + u16 count = 0; + + /* compare counts */ + list_for_each_entry(tmp1, list1, list) + count++; + list_for_each_entry(tmp2, list2, list) + chk_count++; + if (!count || count != chk_count) + return false; + + tmp1 = list_first_entry(list1, struct ice_vsig_prof, list); + tmp2 = list_first_entry(list2, struct ice_vsig_prof, list); + + /* profile cookies must compare, and in the exact same order to take + * into account priority + */ + while (count--) { + if (tmp2->profile_cookie != tmp1->profile_cookie) + return false; + + tmp1 = list_next_entry(tmp1, list); + tmp2 = list_next_entry(tmp2, list); + } + + return true; +} + /* VSIG Management */ /** @@ -999,6 +1566,117 @@ static u16 ice_vsig_alloc_val(struct ice_hw *hw, enum ice_block blk, u16 vsig) } /** + * ice_vsig_alloc - Finds a free entry and allocates a new VSIG + * @hw: pointer to the hardware structure + * @blk: HW block + * + * This function will iterate through the VSIG list and mark the first + * unused entry for the new VSIG entry as used and return that value. + */ +static u16 ice_vsig_alloc(struct ice_hw *hw, enum ice_block blk) +{ + u16 i; + + for (i = 1; i < ICE_MAX_VSIGS; i++) + if (!hw->blk[blk].xlt2.vsig_tbl[i].in_use) + return ice_vsig_alloc_val(hw, blk, i); + + return ICE_DEFAULT_VSIG; +} + +/** + * ice_find_dup_props_vsig - find VSI group with a specified set of properties + * @hw: pointer to the hardware structure + * @blk: HW block + * @chs: characteristic list + * @vsig: returns the VSIG with the matching profiles, if found + * + * Each VSIG is associated with a characteristic set; i.e. all VSIs under + * a group have the same characteristic set. To check if there exists a VSIG + * which has the same characteristics as the input characteristics; this + * function will iterate through the XLT2 list and return the VSIG that has a + * matching configuration. In order to make sure that priorities are accounted + * for, the list must match exactly, including the order in which the + * characteristics are listed. + */ +static enum ice_status +ice_find_dup_props_vsig(struct ice_hw *hw, enum ice_block blk, + struct list_head *chs, u16 *vsig) +{ + struct ice_xlt2 *xlt2 = &hw->blk[blk].xlt2; + u16 i; + + for (i = 0; i < xlt2->count; i++) + if (xlt2->vsig_tbl[i].in_use && + ice_match_prop_lst(chs, &xlt2->vsig_tbl[i].prop_lst)) { + *vsig = ICE_VSIG_VALUE(i, hw->pf_id); + return 0; + } + + return ICE_ERR_DOES_NOT_EXIST; +} + +/** + * ice_vsig_free - free VSI group + * @hw: pointer to the hardware structure + * @blk: HW block + * @vsig: VSIG to remove + * + * The function will remove all VSIs associated with the input VSIG and move + * them to the DEFAULT_VSIG and mark the VSIG available. + */ +static enum ice_status +ice_vsig_free(struct ice_hw *hw, enum ice_block blk, u16 vsig) +{ + struct ice_vsig_prof *dtmp, *del; + struct ice_vsig_vsi *vsi_cur; + u16 idx; + + idx = vsig & ICE_VSIG_IDX_M; + if (idx >= ICE_MAX_VSIGS) + return ICE_ERR_PARAM; + + if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use) + return ICE_ERR_DOES_NOT_EXIST; + + hw->blk[blk].xlt2.vsig_tbl[idx].in_use = false; + + vsi_cur = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi; + /* If the VSIG has at least 1 VSI then iterate through the + * list and remove the VSIs before deleting the group. + */ + if (vsi_cur) { + /* remove all vsis associated with this VSIG XLT2 entry */ + do { + struct ice_vsig_vsi *tmp = vsi_cur->next_vsi; + + vsi_cur->vsig = ICE_DEFAULT_VSIG; + vsi_cur->changed = 1; + vsi_cur->next_vsi = NULL; + vsi_cur = tmp; + } while (vsi_cur); + + /* NULL terminate head of VSI list */ + hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi = NULL; + } + + /* free characteristic list */ + list_for_each_entry_safe(del, dtmp, + &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) { + list_del(&del->list); + devm_kfree(ice_hw_to_dev(hw), del); + } + + /* if VSIG characteristic list was cleared for reset + * re-initialize the list head + */ + INIT_LIST_HEAD(&hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst); + + return 0; +} + +/** * ice_vsig_remove_vsi - remove VSI from VSIG * @hw: pointer to the hardware structure * @blk: HW block @@ -1117,6 +1795,215 @@ ice_vsig_add_mv_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig) return 0; } +/** + * ice_find_prof_id - find profile ID for a given field vector + * @hw: pointer to the hardware structure + * @blk: HW block + * @fv: field vector to search for + * @prof_id: receives the profile ID + */ +static enum ice_status +ice_find_prof_id(struct ice_hw *hw, enum ice_block blk, + struct ice_fv_word *fv, u8 *prof_id) +{ + struct ice_es *es = &hw->blk[blk].es; + u16 off, i; + + for (i = 0; i < es->count; i++) { + off = i * es->fvw; + + if (memcmp(&es->t[off], fv, es->fvw * sizeof(*fv))) + continue; + + *prof_id = i; + return 0; + } + + return ICE_ERR_DOES_NOT_EXIST; +} + +/** + * ice_prof_id_rsrc_type - get profile ID resource type for a block type + * @blk: the block type + * @rsrc_type: pointer to variable to receive the resource type + */ +static bool ice_prof_id_rsrc_type(enum ice_block blk, u16 *rsrc_type) +{ + switch (blk) { + case ICE_BLK_RSS: + *rsrc_type = ICE_AQC_RES_TYPE_HASH_PROF_BLDR_PROFID; + break; + default: + return false; + } + return true; +} + +/** + * ice_tcam_ent_rsrc_type - get TCAM entry resource type for a block type + * @blk: the block type + * @rsrc_type: pointer to variable to receive the resource type + */ +static bool ice_tcam_ent_rsrc_type(enum ice_block blk, u16 *rsrc_type) +{ + switch (blk) { + case ICE_BLK_RSS: + *rsrc_type = ICE_AQC_RES_TYPE_HASH_PROF_BLDR_TCAM; + break; + default: + return false; + } + return true; +} + +/** + * ice_alloc_tcam_ent - allocate hardware TCAM entry + * @hw: pointer to the HW struct + * @blk: the block to allocate the TCAM for + * @tcam_idx: pointer to variable to receive the TCAM entry + * + * This function allocates a new entry in a Profile ID TCAM for a specific + * block. + */ +static enum ice_status +ice_alloc_tcam_ent(struct ice_hw *hw, enum ice_block blk, u16 *tcam_idx) +{ + u16 res_type; + + if (!ice_tcam_ent_rsrc_type(blk, &res_type)) + return ICE_ERR_PARAM; + + return ice_alloc_hw_res(hw, res_type, 1, true, tcam_idx); +} + +/** + * ice_free_tcam_ent - free hardware TCAM entry + * @hw: pointer to the HW struct + * @blk: the block from which to free the TCAM entry + * @tcam_idx: the TCAM entry to free + * + * This function frees an entry in a Profile ID TCAM for a specific block. + */ +static enum ice_status +ice_free_tcam_ent(struct ice_hw *hw, enum ice_block blk, u16 tcam_idx) +{ + u16 res_type; + + if (!ice_tcam_ent_rsrc_type(blk, &res_type)) + return ICE_ERR_PARAM; + + return ice_free_hw_res(hw, res_type, 1, &tcam_idx); +} + +/** + * ice_alloc_prof_id - allocate profile ID + * @hw: pointer to the HW struct + * @blk: the block to allocate the profile ID for + * @prof_id: pointer to variable to receive the profile ID + * + * This function allocates a new profile ID, which also corresponds to a Field + * Vector (Extraction Sequence) entry. + */ +static enum ice_status +ice_alloc_prof_id(struct ice_hw *hw, enum ice_block blk, u8 *prof_id) +{ + enum ice_status status; + u16 res_type; + u16 get_prof; + + if (!ice_prof_id_rsrc_type(blk, &res_type)) + return ICE_ERR_PARAM; + + status = ice_alloc_hw_res(hw, res_type, 1, false, &get_prof); + if (!status) + *prof_id = (u8)get_prof; + + return status; +} + +/** + * ice_free_prof_id - free profile ID + * @hw: pointer to the HW struct + * @blk: the block from which to free the profile ID + * @prof_id: the profile ID to free + * + * This function frees a profile ID, which also corresponds to a Field Vector. + */ +static enum ice_status +ice_free_prof_id(struct ice_hw *hw, enum ice_block blk, u8 prof_id) +{ + u16 tmp_prof_id = (u16)prof_id; + u16 res_type; + + if (!ice_prof_id_rsrc_type(blk, &res_type)) + return ICE_ERR_PARAM; + + return ice_free_hw_res(hw, res_type, 1, &tmp_prof_id); +} + +/** + * ice_prof_inc_ref - increment reference count for profile + * @hw: pointer to the HW struct + * @blk: the block from which to free the profile ID + * @prof_id: the profile ID for which to increment the reference count + */ +static enum ice_status +ice_prof_inc_ref(struct ice_hw *hw, enum ice_block blk, u8 prof_id) +{ + if (prof_id > hw->blk[blk].es.count) + return ICE_ERR_PARAM; + + hw->blk[blk].es.ref_count[prof_id]++; + + return 0; +} + +/** + * ice_write_es - write an extraction sequence to hardware + * @hw: pointer to the HW struct + * @blk: the block in which to write the extraction sequence + * @prof_id: the profile ID to write + * @fv: pointer to the extraction sequence to write - NULL to clear extraction + */ +static void +ice_write_es(struct ice_hw *hw, enum ice_block blk, u8 prof_id, + struct ice_fv_word *fv) +{ + u16 off; + + off = prof_id * hw->blk[blk].es.fvw; + if (!fv) { + memset(&hw->blk[blk].es.t[off], 0, + hw->blk[blk].es.fvw * sizeof(*fv)); + hw->blk[blk].es.written[prof_id] = false; + } else { + memcpy(&hw->blk[blk].es.t[off], fv, + hw->blk[blk].es.fvw * sizeof(*fv)); + } +} + +/** + * ice_prof_dec_ref - decrement reference count for profile + * @hw: pointer to the HW struct + * @blk: the block from which to free the profile ID + * @prof_id: the profile ID for which to decrement the reference count + */ +static enum ice_status +ice_prof_dec_ref(struct ice_hw *hw, enum ice_block blk, u8 prof_id) +{ + if (prof_id > hw->blk[blk].es.count) + return ICE_ERR_PARAM; + + if (hw->blk[blk].es.ref_count[prof_id] > 0) { + if (!--hw->blk[blk].es.ref_count[prof_id]) { + ice_write_es(hw, blk, prof_id, NULL); + return ice_free_prof_id(hw, blk, prof_id); + } + } + + return 0; +} + /* Block / table section IDs */ static const u32 ice_blk_sids[ICE_BLK_COUNT][ICE_SID_OFF_COUNT] = { /* SWITCH */ @@ -1374,16 +2261,85 @@ void ice_fill_blk_tbls(struct ice_hw *hw) } /** + * ice_free_prof_map - free profile map + * @hw: pointer to the hardware structure + * @blk_idx: HW block index + */ +static void ice_free_prof_map(struct ice_hw *hw, u8 blk_idx) +{ + struct ice_es *es = &hw->blk[blk_idx].es; + struct ice_prof_map *del, *tmp; + + mutex_lock(&es->prof_map_lock); + list_for_each_entry_safe(del, tmp, &es->prof_map, list) { + list_del(&del->list); + devm_kfree(ice_hw_to_dev(hw), del); + } + INIT_LIST_HEAD(&es->prof_map); + mutex_unlock(&es->prof_map_lock); +} + +/** + * ice_free_flow_profs - free flow profile entries + * @hw: pointer to the hardware structure + * @blk_idx: HW block index + */ +static void ice_free_flow_profs(struct ice_hw *hw, u8 blk_idx) +{ + struct ice_flow_prof *p, *tmp; + + mutex_lock(&hw->fl_profs_locks[blk_idx]); + list_for_each_entry_safe(p, tmp, &hw->fl_profs[blk_idx], l_entry) { + list_del(&p->l_entry); + devm_kfree(ice_hw_to_dev(hw), p); + } + mutex_unlock(&hw->fl_profs_locks[blk_idx]); + + /* if driver is in reset and tables are being cleared + * re-initialize the flow profile list heads + */ + INIT_LIST_HEAD(&hw->fl_profs[blk_idx]); +} + +/** + * ice_free_vsig_tbl - free complete VSIG table entries + * @hw: pointer to the hardware structure + * @blk: the HW block on which to free the VSIG table entries + */ +static void ice_free_vsig_tbl(struct ice_hw *hw, enum ice_block blk) +{ + u16 i; + + if (!hw->blk[blk].xlt2.vsig_tbl) + return; + + for (i = 1; i < ICE_MAX_VSIGS; i++) + if (hw->blk[blk].xlt2.vsig_tbl[i].in_use) + ice_vsig_free(hw, blk, i); +} + +/** * ice_free_hw_tbls - free hardware table memory * @hw: pointer to the hardware structure */ void ice_free_hw_tbls(struct ice_hw *hw) { + struct ice_rss_cfg *r, *rt; u8 i; for (i = 0; i < ICE_BLK_COUNT; i++) { - hw->blk[i].is_list_init = false; + if (hw->blk[i].is_list_init) { + struct ice_es *es = &hw->blk[i].es; + + ice_free_prof_map(hw, i); + mutex_destroy(&es->prof_map_lock); + ice_free_flow_profs(hw, i); + mutex_destroy(&hw->fl_profs_locks[i]); + + hw->blk[i].is_list_init = false; + } + ice_free_vsig_tbl(hw, (enum ice_block)i); devm_kfree(ice_hw_to_dev(hw), hw->blk[i].xlt1.ptypes); devm_kfree(ice_hw_to_dev(hw), hw->blk[i].xlt1.ptg_tbl); devm_kfree(ice_hw_to_dev(hw), hw->blk[i].xlt1.t); @@ -1397,10 +2353,26 @@ void ice_free_hw_tbls(struct ice_hw *hw) devm_kfree(ice_hw_to_dev(hw), hw->blk[i].es.written); } + list_for_each_entry_safe(r, rt, &hw->rss_list_head, l_entry) { + list_del(&r->l_entry); + devm_kfree(ice_hw_to_dev(hw), r); + } + mutex_destroy(&hw->rss_locks); memset(hw->blk, 0, sizeof(hw->blk)); } /** + * ice_init_flow_profs - init flow profile locks and list heads + * @hw: pointer to the hardware structure + * @blk_idx: HW block index + */ +static void ice_init_flow_profs(struct ice_hw *hw, u8 blk_idx) +{ + mutex_init(&hw->fl_profs_locks[blk_idx]); + INIT_LIST_HEAD(&hw->fl_profs[blk_idx]); +} + +/** * ice_clear_hw_tbls - clear HW tables and flow profiles * @hw: pointer to the hardware structure */ @@ -1415,6 +2387,13 @@ void ice_clear_hw_tbls(struct ice_hw *hw) struct ice_xlt2 *xlt2 = &hw->blk[i].xlt2; struct ice_es *es = &hw->blk[i].es; + if (hw->blk[i].is_list_init) { + ice_free_prof_map(hw, i); + ice_free_flow_profs(hw, i); + } + + ice_free_vsig_tbl(hw, (enum ice_block)i); + memset(xlt1->ptypes, 0, xlt1->count * sizeof(*xlt1->ptypes)); memset(xlt1->ptg_tbl, 0, ICE_MAX_PTGS * sizeof(*xlt1->ptg_tbl)); @@ -1443,6 +2422,8 @@ enum ice_status ice_init_hw_tbls(struct ice_hw *hw) { u8 i; + mutex_init(&hw->rss_locks); + INIT_LIST_HEAD(&hw->rss_list_head); for (i = 0; i < ICE_BLK_COUNT; i++) { struct ice_prof_redir *prof_redir = &hw->blk[i].prof_redir; struct ice_prof_tcam *prof = &hw->blk[i].prof; @@ -1454,6 +2435,9 @@ enum ice_status ice_init_hw_tbls(struct ice_hw *hw) if (hw->blk[i].is_list_init) continue; + ice_init_flow_profs(hw, i); + mutex_init(&es->prof_map_lock); + INIT_LIST_HEAD(&es->prof_map); hw->blk[i].is_list_init = true; hw->blk[i].overwrite = blk_sizes[i].overwrite; @@ -1547,3 +2531,1580 @@ err: ice_free_hw_tbls(hw); return ICE_ERR_NO_MEMORY; } + +/** + * ice_prof_gen_key - generate profile ID key + * @hw: pointer to the HW struct + * @blk: the block in which to write profile ID to + * @ptg: packet type group (PTG) portion of key + * @vsig: VSIG portion of key + * @cdid: CDID portion of key + * @flags: flag portion of key + * @vl_msk: valid mask + * @dc_msk: don't care mask + * @nm_msk: never match mask + * @key: output of profile ID key + */ +static enum ice_status +ice_prof_gen_key(struct ice_hw *hw, enum ice_block blk, u8 ptg, u16 vsig, + u8 cdid, u16 flags, u8 vl_msk[ICE_TCAM_KEY_VAL_SZ], + u8 dc_msk[ICE_TCAM_KEY_VAL_SZ], u8 nm_msk[ICE_TCAM_KEY_VAL_SZ], + u8 key[ICE_TCAM_KEY_SZ]) +{ + struct ice_prof_id_key inkey; + + inkey.xlt1 = ptg; + inkey.xlt2_cdid = cpu_to_le16(vsig); + inkey.flags = cpu_to_le16(flags); + + switch (hw->blk[blk].prof.cdid_bits) { + case 0: + break; + case 2: +#define ICE_CD_2_M 0xC000U +#define ICE_CD_2_S 14 + inkey.xlt2_cdid &= ~cpu_to_le16(ICE_CD_2_M); + inkey.xlt2_cdid |= cpu_to_le16(BIT(cdid) << ICE_CD_2_S); + break; + case 4: +#define ICE_CD_4_M 0xF000U +#define ICE_CD_4_S 12 + inkey.xlt2_cdid &= ~cpu_to_le16(ICE_CD_4_M); + inkey.xlt2_cdid |= cpu_to_le16(BIT(cdid) << ICE_CD_4_S); + break; + case 8: +#define ICE_CD_8_M 0xFF00U +#define ICE_CD_8_S 16 + inkey.xlt2_cdid &= ~cpu_to_le16(ICE_CD_8_M); + inkey.xlt2_cdid |= cpu_to_le16(BIT(cdid) << ICE_CD_8_S); + break; + default: + ice_debug(hw, ICE_DBG_PKG, "Error in profile config\n"); + break; + } + + return ice_set_key(key, ICE_TCAM_KEY_SZ, (u8 *)&inkey, vl_msk, dc_msk, + nm_msk, 0, ICE_TCAM_KEY_SZ / 2); +} + +/** + * ice_tcam_write_entry - write TCAM entry + * @hw: pointer to the HW struct + * @blk: the block in which to write profile ID to + * @idx: the entry index to write to + * @prof_id: profile ID + * @ptg: packet type group (PTG) portion of key + * @vsig: VSIG portion of key + * @cdid: CDID portion of key + * @flags: flag portion of key + * @vl_msk: valid mask + * @dc_msk: don't care mask + * @nm_msk: never match mask + */ +static enum ice_status +ice_tcam_write_entry(struct ice_hw *hw, enum ice_block blk, u16 idx, + u8 prof_id, u8 ptg, u16 vsig, u8 cdid, u16 flags, + u8 vl_msk[ICE_TCAM_KEY_VAL_SZ], + u8 dc_msk[ICE_TCAM_KEY_VAL_SZ], + u8 nm_msk[ICE_TCAM_KEY_VAL_SZ]) +{ + struct ice_prof_tcam_entry; + enum ice_status status; + + status = ice_prof_gen_key(hw, blk, ptg, vsig, cdid, flags, vl_msk, + dc_msk, nm_msk, hw->blk[blk].prof.t[idx].key); + if (!status) { + hw->blk[blk].prof.t[idx].addr = cpu_to_le16(idx); + hw->blk[blk].prof.t[idx].prof_id = prof_id; + } + + return status; +} + +/** + * ice_vsig_get_ref - returns number of VSIs belong to a VSIG + * @hw: pointer to the hardware structure + * @blk: HW block + * @vsig: VSIG to query + * @refs: pointer to variable to receive the reference count + */ +static enum ice_status +ice_vsig_get_ref(struct ice_hw *hw, enum ice_block blk, u16 vsig, u16 *refs) +{ + u16 idx = vsig & ICE_VSIG_IDX_M; + struct ice_vsig_vsi *ptr; + + *refs = 0; + + if (!hw->blk[blk].xlt2.vsig_tbl[idx].in_use) + return ICE_ERR_DOES_NOT_EXIST; + + ptr = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi; + while (ptr) { + (*refs)++; + ptr = ptr->next_vsi; + } + + return 0; +} + +/** + * ice_has_prof_vsig - check to see if VSIG has a specific profile + * @hw: pointer to the hardware structure + * @blk: HW block + * @vsig: VSIG to check against + * @hdl: profile handle + */ +static bool +ice_has_prof_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl) +{ + u16 idx = vsig & ICE_VSIG_IDX_M; + struct ice_vsig_prof *ent; + + list_for_each_entry(ent, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) + if (ent->profile_cookie == hdl) + return true; + + ice_debug(hw, ICE_DBG_INIT, + "Characteristic list for VSI group %d not found.\n", + vsig); + return false; +} + +/** + * ice_prof_bld_es - build profile ID extraction sequence changes + * @hw: pointer to the HW struct + * @blk: hardware block + * @bld: the update package buffer build to add to + * @chgs: the list of changes to make in hardware + */ +static enum ice_status +ice_prof_bld_es(struct ice_hw *hw, enum ice_block blk, + struct ice_buf_build *bld, struct list_head *chgs) +{ + u16 vec_size = hw->blk[blk].es.fvw * sizeof(struct ice_fv_word); + struct ice_chs_chg *tmp; + + list_for_each_entry(tmp, chgs, list_entry) + if (tmp->type == ICE_PTG_ES_ADD && tmp->add_prof) { + u16 off = tmp->prof_id * hw->blk[blk].es.fvw; + struct ice_pkg_es *p; + u32 id; + + id = ice_sect_id(blk, ICE_VEC_TBL); + p = (struct ice_pkg_es *) + ice_pkg_buf_alloc_section(bld, id, sizeof(*p) + + vec_size - + sizeof(p->es[0])); + + if (!p) + return ICE_ERR_MAX_LIMIT; + + p->count = cpu_to_le16(1); + p->offset = cpu_to_le16(tmp->prof_id); + + memcpy(p->es, &hw->blk[blk].es.t[off], vec_size); + } + + return 0; +} + +/** + * ice_prof_bld_tcam - build profile ID TCAM changes + * @hw: pointer to the HW struct + * @blk: hardware block + * @bld: the update package buffer build to add to + * @chgs: the list of changes to make in hardware + */ +static enum ice_status +ice_prof_bld_tcam(struct ice_hw *hw, enum ice_block blk, + struct ice_buf_build *bld, struct list_head *chgs) +{ + struct ice_chs_chg *tmp; + + list_for_each_entry(tmp, chgs, list_entry) + if (tmp->type == ICE_TCAM_ADD && tmp->add_tcam_idx) { + struct ice_prof_id_section *p; + u32 id; + + id = ice_sect_id(blk, ICE_PROF_TCAM); + p = (struct ice_prof_id_section *) + ice_pkg_buf_alloc_section(bld, id, sizeof(*p)); + + if (!p) + return ICE_ERR_MAX_LIMIT; + + p->count = cpu_to_le16(1); + p->entry[0].addr = cpu_to_le16(tmp->tcam_idx); + p->entry[0].prof_id = tmp->prof_id; + + memcpy(p->entry[0].key, + &hw->blk[blk].prof.t[tmp->tcam_idx].key, + sizeof(hw->blk[blk].prof.t->key)); + } + + return 0; +} + +/** + * ice_prof_bld_xlt1 - build XLT1 changes + * @blk: hardware block + * @bld: the update package buffer build to add to + * @chgs: the list of changes to make in hardware + */ +static enum ice_status +ice_prof_bld_xlt1(enum ice_block blk, struct ice_buf_build *bld, + struct list_head *chgs) +{ + struct ice_chs_chg *tmp; + + list_for_each_entry(tmp, chgs, list_entry) + if (tmp->type == ICE_PTG_ES_ADD && tmp->add_ptg) { + struct ice_xlt1_section *p; + u32 id; + + id = ice_sect_id(blk, ICE_XLT1); + p = (struct ice_xlt1_section *) + ice_pkg_buf_alloc_section(bld, id, sizeof(*p)); + + if (!p) + return ICE_ERR_MAX_LIMIT; + + p->count = cpu_to_le16(1); + p->offset = cpu_to_le16(tmp->ptype); + p->value[0] = tmp->ptg; + } + + return 0; +} + +/** + * ice_prof_bld_xlt2 - build XLT2 changes + * @blk: hardware block + * @bld: the update package buffer build to add to + * @chgs: the list of changes to make in hardware + */ +static enum ice_status +ice_prof_bld_xlt2(enum ice_block blk, struct ice_buf_build *bld, + struct list_head *chgs) +{ + struct ice_chs_chg *tmp; + + list_for_each_entry(tmp, chgs, list_entry) { + struct ice_xlt2_section *p; + u32 id; + + switch (tmp->type) { + case ICE_VSIG_ADD: + case ICE_VSI_MOVE: + case ICE_VSIG_REM: + id = ice_sect_id(blk, ICE_XLT2); + p = (struct ice_xlt2_section *) + ice_pkg_buf_alloc_section(bld, id, sizeof(*p)); + + if (!p) + return ICE_ERR_MAX_LIMIT; + + p->count = cpu_to_le16(1); + p->offset = cpu_to_le16(tmp->vsi); + p->value[0] = cpu_to_le16(tmp->vsig); + break; + default: + break; + } + } + + return 0; +} + +/** + * ice_upd_prof_hw - update hardware using the change list + * @hw: pointer to the HW struct + * @blk: hardware block + * @chgs: the list of changes to make in hardware + */ +static enum ice_status +ice_upd_prof_hw(struct ice_hw *hw, enum ice_block blk, + struct list_head *chgs) +{ + struct ice_buf_build *b; + struct ice_chs_chg *tmp; + enum ice_status status; + u16 pkg_sects; + u16 xlt1 = 0; + u16 xlt2 = 0; + u16 tcam = 0; + u16 es = 0; + u16 sects; + + /* count number of sections we need */ + list_for_each_entry(tmp, chgs, list_entry) { + switch (tmp->type) { + case ICE_PTG_ES_ADD: + if (tmp->add_ptg) + xlt1++; + if (tmp->add_prof) + es++; + break; + case ICE_TCAM_ADD: + tcam++; + break; + case ICE_VSIG_ADD: + case ICE_VSI_MOVE: + case ICE_VSIG_REM: + xlt2++; + break; + default: + break; + } + } + sects = xlt1 + xlt2 + tcam + es; + + if (!sects) + return 0; + + /* Build update package buffer */ + b = ice_pkg_buf_alloc(hw); + if (!b) + return ICE_ERR_NO_MEMORY; + + status = ice_pkg_buf_reserve_section(b, sects); + if (status) + goto error_tmp; + + /* Preserve order of table update: ES, TCAM, PTG, VSIG */ + if (es) { + status = ice_prof_bld_es(hw, blk, b, chgs); + if (status) + goto error_tmp; + } + + if (tcam) { + status = ice_prof_bld_tcam(hw, blk, b, chgs); + if (status) + goto error_tmp; + } + + if (xlt1) { + status = ice_prof_bld_xlt1(blk, b, chgs); + if (status) + goto error_tmp; + } + + if (xlt2) { + status = ice_prof_bld_xlt2(blk, b, chgs); + if (status) + goto error_tmp; + } + + /* After package buffer build check if the section count in buffer is + * non-zero and matches the number of sections detected for package + * update. + */ + pkg_sects = ice_pkg_buf_get_active_sections(b); + if (!pkg_sects || pkg_sects != sects) { + status = ICE_ERR_INVAL_SIZE; + goto error_tmp; + } + + /* update package */ + status = ice_update_pkg(hw, ice_pkg_buf(b), 1); + if (status == ICE_ERR_AQ_ERROR) + ice_debug(hw, ICE_DBG_INIT, "Unable to update HW profile\n"); + +error_tmp: + ice_pkg_buf_free(hw, b); + return status; +} + +/** + * ice_add_prof - add profile + * @hw: pointer to the HW struct + * @blk: hardware block + * @id: profile tracking ID + * @ptypes: array of bitmaps indicating ptypes (ICE_FLOW_PTYPE_MAX bits) + * @es: extraction sequence (length of array is determined by the block) + * + * This function registers a profile, which matches a set of PTGs with a + * particular extraction sequence. While the hardware profile is allocated + * it will not be written until the first call to ice_add_flow that specifies + * the ID value used here. + */ +enum ice_status +ice_add_prof(struct ice_hw *hw, enum ice_block blk, u64 id, u8 ptypes[], + struct ice_fv_word *es) +{ + u32 bytes = DIV_ROUND_UP(ICE_FLOW_PTYPE_MAX, BITS_PER_BYTE); + DECLARE_BITMAP(ptgs_used, ICE_XLT1_CNT); + struct ice_prof_map *prof; + enum ice_status status; + u32 byte = 0; + u8 prof_id; + + bitmap_zero(ptgs_used, ICE_XLT1_CNT); + + mutex_lock(&hw->blk[blk].es.prof_map_lock); + + /* search for existing profile */ + status = ice_find_prof_id(hw, blk, es, &prof_id); + if (status) { + /* allocate profile ID */ + status = ice_alloc_prof_id(hw, blk, &prof_id); + if (status) + goto err_ice_add_prof; + + /* and write new es */ + ice_write_es(hw, blk, prof_id, es); + } + + ice_prof_inc_ref(hw, blk, prof_id); + + /* add profile info */ + prof = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*prof), GFP_KERNEL); + if (!prof) + goto err_ice_add_prof; + + prof->profile_cookie = id; + prof->prof_id = prof_id; + prof->ptg_cnt = 0; + prof->context = 0; + + /* build list of ptgs */ + while (bytes && prof->ptg_cnt < ICE_MAX_PTG_PER_PROFILE) { + u32 bit; + + if (!ptypes[byte]) { + bytes--; + byte++; + continue; + } + + /* Examine 8 bits per byte */ + for_each_set_bit(bit, (unsigned long *)&ptypes[byte], + BITS_PER_BYTE) { + u16 ptype; + u8 ptg; + u8 m; + + ptype = byte * BITS_PER_BYTE + bit; + + /* The package should place all ptypes in a non-zero + * PTG, so the following call should never fail. + */ + if (ice_ptg_find_ptype(hw, blk, ptype, &ptg)) + continue; + + /* If PTG is already added, skip and continue */ + if (test_bit(ptg, ptgs_used)) + continue; + + set_bit(ptg, ptgs_used); + prof->ptg[prof->ptg_cnt] = ptg; + + if (++prof->ptg_cnt >= ICE_MAX_PTG_PER_PROFILE) + break; + + /* nothing left in byte, then exit */ + m = ~((1 << (bit + 1)) - 1); + if (!(ptypes[byte] & m)) + break; + } + + bytes--; + byte++; + } + + list_add(&prof->list, &hw->blk[blk].es.prof_map); + status = 0; + +err_ice_add_prof: + mutex_unlock(&hw->blk[blk].es.prof_map_lock); + return status; +} + +/** + * ice_search_prof_id_low - Search for a profile tracking ID low level + * @hw: pointer to the HW struct + * @blk: hardware block + * @id: profile tracking ID + * + * This will search for a profile tracking ID which was previously added. This + * version assumes that the caller has already acquired the prof map lock. + */ +static struct ice_prof_map * +ice_search_prof_id_low(struct ice_hw *hw, enum ice_block blk, u64 id) +{ + struct ice_prof_map *entry = NULL; + struct ice_prof_map *map; + + list_for_each_entry(map, &hw->blk[blk].es.prof_map, list) + if (map->profile_cookie == id) { + entry = map; + break; + } + + return entry; +} + +/** + * ice_search_prof_id - Search for a profile tracking ID + * @hw: pointer to the HW struct + * @blk: hardware block + * @id: profile tracking ID + * + * This will search for a profile tracking ID which was previously added. + */ +static struct ice_prof_map * +ice_search_prof_id(struct ice_hw *hw, enum ice_block blk, u64 id) +{ + struct ice_prof_map *entry; + + mutex_lock(&hw->blk[blk].es.prof_map_lock); + entry = ice_search_prof_id_low(hw, blk, id); + mutex_unlock(&hw->blk[blk].es.prof_map_lock); + + return entry; +} + +/** + * ice_vsig_prof_id_count - count profiles in a VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: VSIG to remove the profile from + */ +static u16 +ice_vsig_prof_id_count(struct ice_hw *hw, enum ice_block blk, u16 vsig) +{ + u16 idx = vsig & ICE_VSIG_IDX_M, count = 0; + struct ice_vsig_prof *p; + + list_for_each_entry(p, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) + count++; + + return count; +} + +/** + * ice_rel_tcam_idx - release a TCAM index + * @hw: pointer to the HW struct + * @blk: hardware block + * @idx: the index to release + */ +static enum ice_status +ice_rel_tcam_idx(struct ice_hw *hw, enum ice_block blk, u16 idx) +{ + /* Masks to invoke a never match entry */ + u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }; + u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFE, 0xFF, 0xFF, 0xFF, 0xFF }; + u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 }; + enum ice_status status; + + /* write the TCAM entry */ + status = ice_tcam_write_entry(hw, blk, idx, 0, 0, 0, 0, 0, vl_msk, + dc_msk, nm_msk); + if (status) + return status; + + /* release the TCAM entry */ + status = ice_free_tcam_ent(hw, blk, idx); + + return status; +} + +/** + * ice_rem_prof_id - remove one profile from a VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @prof: pointer to profile structure to remove + */ +static enum ice_status +ice_rem_prof_id(struct ice_hw *hw, enum ice_block blk, + struct ice_vsig_prof *prof) +{ + enum ice_status status; + u16 i; + + for (i = 0; i < prof->tcam_count; i++) + if (prof->tcam[i].in_use) { + prof->tcam[i].in_use = false; + status = ice_rel_tcam_idx(hw, blk, + prof->tcam[i].tcam_idx); + if (status) + return ICE_ERR_HW_TABLE; + } + + return 0; +} + +/** + * ice_rem_vsig - remove VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: the VSIG to remove + * @chg: the change list + */ +static enum ice_status +ice_rem_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, + struct list_head *chg) +{ + u16 idx = vsig & ICE_VSIG_IDX_M; + struct ice_vsig_vsi *vsi_cur; + struct ice_vsig_prof *d, *t; + enum ice_status status; + + /* remove TCAM entries */ + list_for_each_entry_safe(d, t, + &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) { + status = ice_rem_prof_id(hw, blk, d); + if (status) + return status; + + list_del(&d->list); + devm_kfree(ice_hw_to_dev(hw), d); + } + + /* Move all VSIS associated with this VSIG to the default VSIG */ + vsi_cur = hw->blk[blk].xlt2.vsig_tbl[idx].first_vsi; + /* If the VSIG has at least 1 VSI then iterate through the list + * and remove the VSIs before deleting the group. + */ + if (vsi_cur) + do { + struct ice_vsig_vsi *tmp = vsi_cur->next_vsi; + struct ice_chs_chg *p; + + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), + GFP_KERNEL); + if (!p) + return ICE_ERR_NO_MEMORY; + + p->type = ICE_VSIG_REM; + p->orig_vsig = vsig; + p->vsig = ICE_DEFAULT_VSIG; + p->vsi = vsi_cur - hw->blk[blk].xlt2.vsis; + + list_add(&p->list_entry, chg); + + vsi_cur = tmp; + } while (vsi_cur); + + return ice_vsig_free(hw, blk, vsig); +} + +/** + * ice_rem_prof_id_vsig - remove a specific profile from a VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: VSIG to remove the profile from + * @hdl: profile handle indicating which profile to remove + * @chg: list to receive a record of changes + */ +static enum ice_status +ice_rem_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl, + struct list_head *chg) +{ + u16 idx = vsig & ICE_VSIG_IDX_M; + struct ice_vsig_prof *p, *t; + enum ice_status status; + + list_for_each_entry_safe(p, t, + &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) + if (p->profile_cookie == hdl) { + if (ice_vsig_prof_id_count(hw, blk, vsig) == 1) + /* this is the last profile, remove the VSIG */ + return ice_rem_vsig(hw, blk, vsig, chg); + + status = ice_rem_prof_id(hw, blk, p); + if (!status) { + list_del(&p->list); + devm_kfree(ice_hw_to_dev(hw), p); + } + return status; + } + + return ICE_ERR_DOES_NOT_EXIST; +} + +/** + * ice_rem_flow_all - remove all flows with a particular profile + * @hw: pointer to the HW struct + * @blk: hardware block + * @id: profile tracking ID + */ +static enum ice_status +ice_rem_flow_all(struct ice_hw *hw, enum ice_block blk, u64 id) +{ + struct ice_chs_chg *del, *tmp; + enum ice_status status; + struct list_head chg; + u16 i; + + INIT_LIST_HEAD(&chg); + + for (i = 1; i < ICE_MAX_VSIGS; i++) + if (hw->blk[blk].xlt2.vsig_tbl[i].in_use) { + if (ice_has_prof_vsig(hw, blk, i, id)) { + status = ice_rem_prof_id_vsig(hw, blk, i, id, + &chg); + if (status) + goto err_ice_rem_flow_all; + } + } + + status = ice_upd_prof_hw(hw, blk, &chg); + +err_ice_rem_flow_all: + list_for_each_entry_safe(del, tmp, &chg, list_entry) { + list_del(&del->list_entry); + devm_kfree(ice_hw_to_dev(hw), del); + } + + return status; +} + +/** + * ice_rem_prof - remove profile + * @hw: pointer to the HW struct + * @blk: hardware block + * @id: profile tracking ID + * + * This will remove the profile specified by the ID parameter, which was + * previously created through ice_add_prof. If any existing entries + * are associated with this profile, they will be removed as well. + */ +enum ice_status ice_rem_prof(struct ice_hw *hw, enum ice_block blk, u64 id) +{ + struct ice_prof_map *pmap; + enum ice_status status; + + mutex_lock(&hw->blk[blk].es.prof_map_lock); + + pmap = ice_search_prof_id_low(hw, blk, id); + if (!pmap) { + status = ICE_ERR_DOES_NOT_EXIST; + goto err_ice_rem_prof; + } + + /* remove all flows with this profile */ + status = ice_rem_flow_all(hw, blk, pmap->profile_cookie); + if (status) + goto err_ice_rem_prof; + + /* dereference profile, and possibly remove */ + ice_prof_dec_ref(hw, blk, pmap->prof_id); + + list_del(&pmap->list); + devm_kfree(ice_hw_to_dev(hw), pmap); + +err_ice_rem_prof: + mutex_unlock(&hw->blk[blk].es.prof_map_lock); + return status; +} + +/** + * ice_get_prof - get profile + * @hw: pointer to the HW struct + * @blk: hardware block + * @hdl: profile handle + * @chg: change list + */ +static enum ice_status +ice_get_prof(struct ice_hw *hw, enum ice_block blk, u64 hdl, + struct list_head *chg) +{ + struct ice_prof_map *map; + struct ice_chs_chg *p; + u16 i; + + /* Get the details on the profile specified by the handle ID */ + map = ice_search_prof_id(hw, blk, hdl); + if (!map) + return ICE_ERR_DOES_NOT_EXIST; + + for (i = 0; i < map->ptg_cnt; i++) + if (!hw->blk[blk].es.written[map->prof_id]) { + /* add ES to change list */ + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), + GFP_KERNEL); + if (!p) + goto err_ice_get_prof; + + p->type = ICE_PTG_ES_ADD; + p->ptype = 0; + p->ptg = map->ptg[i]; + p->add_ptg = 0; + + p->add_prof = 1; + p->prof_id = map->prof_id; + + hw->blk[blk].es.written[map->prof_id] = true; + + list_add(&p->list_entry, chg); + } + + return 0; + +err_ice_get_prof: + /* let caller clean up the change list */ + return ICE_ERR_NO_MEMORY; +} + +/** + * ice_get_profs_vsig - get a copy of the list of profiles from a VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: VSIG from which to copy the list + * @lst: output list + * + * This routine makes a copy of the list of profiles in the specified VSIG. + */ +static enum ice_status +ice_get_profs_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, + struct list_head *lst) +{ + struct ice_vsig_prof *ent1, *ent2; + u16 idx = vsig & ICE_VSIG_IDX_M; + + list_for_each_entry(ent1, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) { + struct ice_vsig_prof *p; + + /* copy to the input list */ + p = devm_kmemdup(ice_hw_to_dev(hw), ent1, sizeof(*p), + GFP_KERNEL); + if (!p) + goto err_ice_get_profs_vsig; + + list_add_tail(&p->list, lst); + } + + return 0; + +err_ice_get_profs_vsig: + list_for_each_entry_safe(ent1, ent2, lst, list) { + list_del(&ent1->list); + devm_kfree(ice_hw_to_dev(hw), ent1); + } + + return ICE_ERR_NO_MEMORY; +} + +/** + * ice_add_prof_to_lst - add profile entry to a list + * @hw: pointer to the HW struct + * @blk: hardware block + * @lst: the list to be added to + * @hdl: profile handle of entry to add + */ +static enum ice_status +ice_add_prof_to_lst(struct ice_hw *hw, enum ice_block blk, + struct list_head *lst, u64 hdl) +{ + struct ice_prof_map *map; + struct ice_vsig_prof *p; + u16 i; + + map = ice_search_prof_id(hw, blk, hdl); + if (!map) + return ICE_ERR_DOES_NOT_EXIST; + + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), GFP_KERNEL); + if (!p) + return ICE_ERR_NO_MEMORY; + + p->profile_cookie = map->profile_cookie; + p->prof_id = map->prof_id; + p->tcam_count = map->ptg_cnt; + + for (i = 0; i < map->ptg_cnt; i++) { + p->tcam[i].prof_id = map->prof_id; + p->tcam[i].tcam_idx = ICE_INVALID_TCAM; + p->tcam[i].ptg = map->ptg[i]; + } + + list_add(&p->list, lst); + + return 0; +} + +/** + * ice_move_vsi - move VSI to another VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsi: the VSI to move + * @vsig: the VSIG to move the VSI to + * @chg: the change list + */ +static enum ice_status +ice_move_vsi(struct ice_hw *hw, enum ice_block blk, u16 vsi, u16 vsig, + struct list_head *chg) +{ + enum ice_status status; + struct ice_chs_chg *p; + u16 orig_vsig; + + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), GFP_KERNEL); + if (!p) + return ICE_ERR_NO_MEMORY; + + status = ice_vsig_find_vsi(hw, blk, vsi, &orig_vsig); + if (!status) + status = ice_vsig_add_mv_vsi(hw, blk, vsi, vsig); + + if (status) { + devm_kfree(ice_hw_to_dev(hw), p); + return status; + } + + p->type = ICE_VSI_MOVE; + p->vsi = vsi; + p->orig_vsig = orig_vsig; + p->vsig = vsig; + + list_add(&p->list_entry, chg); + + return 0; +} + +/** + * ice_prof_tcam_ena_dis - add enable or disable TCAM change + * @hw: pointer to the HW struct + * @blk: hardware block + * @enable: true to enable, false to disable + * @vsig: the VSIG of the TCAM entry + * @tcam: pointer the TCAM info structure of the TCAM to disable + * @chg: the change list + * + * This function appends an enable or disable TCAM entry in the change log + */ +static enum ice_status +ice_prof_tcam_ena_dis(struct ice_hw *hw, enum ice_block blk, bool enable, + u16 vsig, struct ice_tcam_inf *tcam, + struct list_head *chg) +{ + enum ice_status status; + struct ice_chs_chg *p; + + /* Default: enable means change the low flag bit to don't care */ + u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 }; + u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x00, 0x00, 0x00, 0x00, 0x00 }; + u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x01, 0x00, 0x00, 0x00, 0x00 }; + + /* if disabling, free the TCAM */ + if (!enable) { + status = ice_free_tcam_ent(hw, blk, tcam->tcam_idx); + tcam->tcam_idx = 0; + tcam->in_use = 0; + return status; + } + + /* for re-enabling, reallocate a TCAM */ + status = ice_alloc_tcam_ent(hw, blk, &tcam->tcam_idx); + if (status) + return status; + + /* add TCAM to change list */ + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), GFP_KERNEL); + if (!p) + return ICE_ERR_NO_MEMORY; + + status = ice_tcam_write_entry(hw, blk, tcam->tcam_idx, tcam->prof_id, + tcam->ptg, vsig, 0, 0, vl_msk, dc_msk, + nm_msk); + if (status) + goto err_ice_prof_tcam_ena_dis; + + tcam->in_use = 1; + + p->type = ICE_TCAM_ADD; + p->add_tcam_idx = true; + p->prof_id = tcam->prof_id; + p->ptg = tcam->ptg; + p->vsig = 0; + p->tcam_idx = tcam->tcam_idx; + + /* log change */ + list_add(&p->list_entry, chg); + + return 0; + +err_ice_prof_tcam_ena_dis: + devm_kfree(ice_hw_to_dev(hw), p); + return status; +} + +/** + * ice_adj_prof_priorities - adjust profile based on priorities + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: the VSIG for which to adjust profile priorities + * @chg: the change list + */ +static enum ice_status +ice_adj_prof_priorities(struct ice_hw *hw, enum ice_block blk, u16 vsig, + struct list_head *chg) +{ + DECLARE_BITMAP(ptgs_used, ICE_XLT1_CNT); + struct ice_vsig_prof *t; + enum ice_status status; + u16 idx; + + bitmap_zero(ptgs_used, ICE_XLT1_CNT); + idx = vsig & ICE_VSIG_IDX_M; + + /* Priority is based on the order in which the profiles are added. The + * newest added profile has highest priority and the oldest added + * profile has the lowest priority. Since the profile property list for + * a VSIG is sorted from newest to oldest, this code traverses the list + * in order and enables the first of each PTG that it finds (that is not + * already enabled); it also disables any duplicate PTGs that it finds + * in the older profiles (that are currently enabled). + */ + + list_for_each_entry(t, &hw->blk[blk].xlt2.vsig_tbl[idx].prop_lst, + list) { + u16 i; + + for (i = 0; i < t->tcam_count; i++) { + /* Scan the priorities from newest to oldest. + * Make sure that the newest profiles take priority. + */ + if (test_bit(t->tcam[i].ptg, ptgs_used) && + t->tcam[i].in_use) { + /* need to mark this PTG as never match, as it + * was already in use and therefore duplicate + * (and lower priority) + */ + status = ice_prof_tcam_ena_dis(hw, blk, false, + vsig, + &t->tcam[i], + chg); + if (status) + return status; + } else if (!test_bit(t->tcam[i].ptg, ptgs_used) && + !t->tcam[i].in_use) { + /* need to enable this PTG, as it in not in use + * and not enabled (highest priority) + */ + status = ice_prof_tcam_ena_dis(hw, blk, true, + vsig, + &t->tcam[i], + chg); + if (status) + return status; + } + + /* keep track of used ptgs */ + set_bit(t->tcam[i].ptg, ptgs_used); + } + } + + return 0; +} + +/** + * ice_add_prof_id_vsig - add profile to VSIG + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsig: the VSIG to which this profile is to be added + * @hdl: the profile handle indicating the profile to add + * @chg: the change list + */ +static enum ice_status +ice_add_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsig, u64 hdl, + struct list_head *chg) +{ + /* Masks that ignore flags */ + u8 vl_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0xFF, 0xFF, 0xFF }; + u8 dc_msk[ICE_TCAM_KEY_VAL_SZ] = { 0xFF, 0xFF, 0x00, 0x00, 0x00 }; + u8 nm_msk[ICE_TCAM_KEY_VAL_SZ] = { 0x00, 0x00, 0x00, 0x00, 0x00 }; + struct ice_prof_map *map; + struct ice_vsig_prof *t; + struct ice_chs_chg *p; + u16 i; + + /* Get the details on the profile specified by the handle ID */ + map = ice_search_prof_id(hw, blk, hdl); + if (!map) + return ICE_ERR_DOES_NOT_EXIST; + + /* Error, if this VSIG already has this profile */ + if (ice_has_prof_vsig(hw, blk, vsig, hdl)) + return ICE_ERR_ALREADY_EXISTS; + + /* new VSIG profile structure */ + t = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*t), GFP_KERNEL); + if (!t) + return ICE_ERR_NO_MEMORY; + + t->profile_cookie = map->profile_cookie; + t->prof_id = map->prof_id; + t->tcam_count = map->ptg_cnt; + + /* create TCAM entries */ + for (i = 0; i < map->ptg_cnt; i++) { + enum ice_status status; + u16 tcam_idx; + + /* add TCAM to change list */ + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), GFP_KERNEL); + if (!p) + goto err_ice_add_prof_id_vsig; + + /* allocate the TCAM entry index */ + status = ice_alloc_tcam_ent(hw, blk, &tcam_idx); + if (status) { + devm_kfree(ice_hw_to_dev(hw), p); + goto err_ice_add_prof_id_vsig; + } + + t->tcam[i].ptg = map->ptg[i]; + t->tcam[i].prof_id = map->prof_id; + t->tcam[i].tcam_idx = tcam_idx; + t->tcam[i].in_use = true; + + p->type = ICE_TCAM_ADD; + p->add_tcam_idx = true; + p->prof_id = t->tcam[i].prof_id; + p->ptg = t->tcam[i].ptg; + p->vsig = vsig; + p->tcam_idx = t->tcam[i].tcam_idx; + + /* write the TCAM entry */ + status = ice_tcam_write_entry(hw, blk, t->tcam[i].tcam_idx, + t->tcam[i].prof_id, + t->tcam[i].ptg, vsig, 0, 0, + vl_msk, dc_msk, nm_msk); + if (status) + goto err_ice_add_prof_id_vsig; + + /* log change */ + list_add(&p->list_entry, chg); + } + + /* add profile to VSIG */ + list_add(&t->list, + &hw->blk[blk].xlt2.vsig_tbl[(vsig & ICE_VSIG_IDX_M)].prop_lst); + + return 0; + +err_ice_add_prof_id_vsig: + /* let caller clean up the change list */ + devm_kfree(ice_hw_to_dev(hw), t); + return ICE_ERR_NO_MEMORY; +} + +/** + * ice_create_prof_id_vsig - add a new VSIG with a single profile + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsi: the initial VSI that will be in VSIG + * @hdl: the profile handle of the profile that will be added to the VSIG + * @chg: the change list + */ +static enum ice_status +ice_create_prof_id_vsig(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl, + struct list_head *chg) +{ + enum ice_status status; + struct ice_chs_chg *p; + u16 new_vsig; + + p = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*p), GFP_KERNEL); + if (!p) + return ICE_ERR_NO_MEMORY; + + new_vsig = ice_vsig_alloc(hw, blk); + if (!new_vsig) { + status = ICE_ERR_HW_TABLE; + goto err_ice_create_prof_id_vsig; + } + + status = ice_move_vsi(hw, blk, vsi, new_vsig, chg); + if (status) + goto err_ice_create_prof_id_vsig; + + status = ice_add_prof_id_vsig(hw, blk, new_vsig, hdl, chg); + if (status) + goto err_ice_create_prof_id_vsig; + + p->type = ICE_VSIG_ADD; + p->vsi = vsi; + p->orig_vsig = ICE_DEFAULT_VSIG; + p->vsig = new_vsig; + + list_add(&p->list_entry, chg); + + return 0; + +err_ice_create_prof_id_vsig: + /* let caller clean up the change list */ + devm_kfree(ice_hw_to_dev(hw), p); + return status; +} + +/** + * ice_create_vsig_from_lst - create a new VSIG with a list of profiles + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsi: the initial VSI that will be in VSIG + * @lst: the list of profile that will be added to the VSIG + * @chg: the change list + */ +static enum ice_status +ice_create_vsig_from_lst(struct ice_hw *hw, enum ice_block blk, u16 vsi, + struct list_head *lst, struct list_head *chg) +{ + struct ice_vsig_prof *t; + enum ice_status status; + u16 vsig; + + vsig = ice_vsig_alloc(hw, blk); + if (!vsig) + return ICE_ERR_HW_TABLE; + + status = ice_move_vsi(hw, blk, vsi, vsig, chg); + if (status) + return status; + + list_for_each_entry(t, lst, list) { + status = ice_add_prof_id_vsig(hw, blk, vsig, t->profile_cookie, + chg); + if (status) + return status; + } + + return 0; +} + +/** + * ice_find_prof_vsig - find a VSIG with a specific profile handle + * @hw: pointer to the HW struct + * @blk: hardware block + * @hdl: the profile handle of the profile to search for + * @vsig: returns the VSIG with the matching profile + */ +static bool +ice_find_prof_vsig(struct ice_hw *hw, enum ice_block blk, u64 hdl, u16 *vsig) +{ + struct ice_vsig_prof *t; + enum ice_status status; + struct list_head lst; + + INIT_LIST_HEAD(&lst); + + t = kzalloc(sizeof(*t), GFP_KERNEL); + if (!t) + return false; + + t->profile_cookie = hdl; + list_add(&t->list, &lst); + + status = ice_find_dup_props_vsig(hw, blk, &lst, vsig); + + list_del(&t->list); + kfree(t); + + return !status; +} + +/** + * ice_add_prof_id_flow - add profile flow + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsi: the VSI to enable with the profile specified by ID + * @hdl: profile handle + * + * Calling this function will update the hardware tables to enable the + * profile indicated by the ID parameter for the VSIs specified in the VSI + * array. Once successfully called, the flow will be enabled. + */ +enum ice_status +ice_add_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl) +{ + struct ice_vsig_prof *tmp1, *del1; + struct ice_chs_chg *tmp, *del; + struct list_head union_lst; + enum ice_status status; + struct list_head chg; + u16 vsig; + + INIT_LIST_HEAD(&union_lst); + INIT_LIST_HEAD(&chg); + + /* Get profile */ + status = ice_get_prof(hw, blk, hdl, &chg); + if (status) + return status; + + /* determine if VSI is already part of a VSIG */ + status = ice_vsig_find_vsi(hw, blk, vsi, &vsig); + if (!status && vsig) { + bool only_vsi; + u16 or_vsig; + u16 ref; + + /* found in VSIG */ + or_vsig = vsig; + + /* make sure that there is no overlap/conflict between the new + * characteristics and the existing ones; we don't support that + * scenario + */ + if (ice_has_prof_vsig(hw, blk, vsig, hdl)) { + status = ICE_ERR_ALREADY_EXISTS; + goto err_ice_add_prof_id_flow; + } + + /* last VSI in the VSIG? */ + status = ice_vsig_get_ref(hw, blk, vsig, &ref); + if (status) + goto err_ice_add_prof_id_flow; + only_vsi = (ref == 1); + + /* create a union of the current profiles and the one being + * added + */ + status = ice_get_profs_vsig(hw, blk, vsig, &union_lst); + if (status) + goto err_ice_add_prof_id_flow; + + status = ice_add_prof_to_lst(hw, blk, &union_lst, hdl); + if (status) + goto err_ice_add_prof_id_flow; + + /* search for an existing VSIG with an exact charc match */ + status = ice_find_dup_props_vsig(hw, blk, &union_lst, &vsig); + if (!status) { + /* move VSI to the VSIG that matches */ + status = ice_move_vsi(hw, blk, vsi, vsig, &chg); + if (status) + goto err_ice_add_prof_id_flow; + + /* VSI has been moved out of or_vsig. If the or_vsig had + * only that VSI it is now empty and can be removed. + */ + if (only_vsi) { + status = ice_rem_vsig(hw, blk, or_vsig, &chg); + if (status) + goto err_ice_add_prof_id_flow; + } + } else if (only_vsi) { + /* If the original VSIG only contains one VSI, then it + * will be the requesting VSI. In this case the VSI is + * not sharing entries and we can simply add the new + * profile to the VSIG. + */ + status = ice_add_prof_id_vsig(hw, blk, vsig, hdl, &chg); + if (status) + goto err_ice_add_prof_id_flow; + + /* Adjust priorities */ + status = ice_adj_prof_priorities(hw, blk, vsig, &chg); + if (status) + goto err_ice_add_prof_id_flow; + } else { + /* No match, so we need a new VSIG */ + status = ice_create_vsig_from_lst(hw, blk, vsi, + &union_lst, &chg); + if (status) + goto err_ice_add_prof_id_flow; + + /* Adjust priorities */ + status = ice_adj_prof_priorities(hw, blk, vsig, &chg); + if (status) + goto err_ice_add_prof_id_flow; + } + } else { + /* need to find or add a VSIG */ + /* search for an existing VSIG with an exact charc match */ + if (ice_find_prof_vsig(hw, blk, hdl, &vsig)) { + /* found an exact match */ + /* add or move VSI to the VSIG that matches */ + status = ice_move_vsi(hw, blk, vsi, vsig, &chg); + if (status) + goto err_ice_add_prof_id_flow; + } else { + /* we did not find an exact match */ + /* we need to add a VSIG */ + status = ice_create_prof_id_vsig(hw, blk, vsi, hdl, + &chg); + if (status) + goto err_ice_add_prof_id_flow; + } + } + + /* update hardware */ + if (!status) + status = ice_upd_prof_hw(hw, blk, &chg); + +err_ice_add_prof_id_flow: + list_for_each_entry_safe(del, tmp, &chg, list_entry) { + list_del(&del->list_entry); + devm_kfree(ice_hw_to_dev(hw), del); + } + + list_for_each_entry_safe(del1, tmp1, &union_lst, list) { + list_del(&del1->list); + devm_kfree(ice_hw_to_dev(hw), del1); + } + + return status; +} + +/** + * ice_rem_prof_from_list - remove a profile from list + * @hw: pointer to the HW struct + * @lst: list to remove the profile from + * @hdl: the profile handle indicating the profile to remove + */ +static enum ice_status +ice_rem_prof_from_list(struct ice_hw *hw, struct list_head *lst, u64 hdl) +{ + struct ice_vsig_prof *ent, *tmp; + + list_for_each_entry_safe(ent, tmp, lst, list) + if (ent->profile_cookie == hdl) { + list_del(&ent->list); + devm_kfree(ice_hw_to_dev(hw), ent); + return 0; + } + + return ICE_ERR_DOES_NOT_EXIST; +} + +/** + * ice_rem_prof_id_flow - remove flow + * @hw: pointer to the HW struct + * @blk: hardware block + * @vsi: the VSI from which to remove the profile specified by ID + * @hdl: profile tracking handle + * + * Calling this function will update the hardware tables to remove the + * profile indicated by the ID parameter for the VSIs specified in the VSI + * array. Once successfully called, the flow will be disabled. + */ +enum ice_status +ice_rem_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl) +{ + struct ice_vsig_prof *tmp1, *del1; + struct ice_chs_chg *tmp, *del; + struct list_head chg, copy; + enum ice_status status; + u16 vsig; + + INIT_LIST_HEAD(©); + INIT_LIST_HEAD(&chg); + + /* determine if VSI is already part of a VSIG */ + status = ice_vsig_find_vsi(hw, blk, vsi, &vsig); + if (!status && vsig) { + bool last_profile; + bool only_vsi; + u16 ref; + + /* found in VSIG */ + last_profile = ice_vsig_prof_id_count(hw, blk, vsig) == 1; + status = ice_vsig_get_ref(hw, blk, vsig, &ref); + if (status) + goto err_ice_rem_prof_id_flow; + only_vsi = (ref == 1); + + if (only_vsi) { + /* If the original VSIG only contains one reference, + * which will be the requesting VSI, then the VSI is not + * sharing entries and we can simply remove the specific + * characteristics from the VSIG. + */ + + if (last_profile) { + /* If there are no profiles left for this VSIG, + * then simply remove the the VSIG. + */ + status = ice_rem_vsig(hw, blk, vsig, &chg); + if (status) + goto err_ice_rem_prof_id_flow; + } else { + status = ice_rem_prof_id_vsig(hw, blk, vsig, + hdl, &chg); + if (status) + goto err_ice_rem_prof_id_flow; + + /* Adjust priorities */ + status = ice_adj_prof_priorities(hw, blk, vsig, + &chg); + if (status) + goto err_ice_rem_prof_id_flow; + } + + } else { + /* Make a copy of the VSIG's list of Profiles */ + status = ice_get_profs_vsig(hw, blk, vsig, ©); + if (status) + goto err_ice_rem_prof_id_flow; + + /* Remove specified profile entry from the list */ + status = ice_rem_prof_from_list(hw, ©, hdl); + if (status) + goto err_ice_rem_prof_id_flow; + + if (list_empty(©)) { + status = ice_move_vsi(hw, blk, vsi, + ICE_DEFAULT_VSIG, &chg); + if (status) + goto err_ice_rem_prof_id_flow; + + } else if (!ice_find_dup_props_vsig(hw, blk, ©, + &vsig)) { + /* found an exact match */ + /* add or move VSI to the VSIG that matches */ + /* Search for a VSIG with a matching profile + * list + */ + + /* Found match, move VSI to the matching VSIG */ + status = ice_move_vsi(hw, blk, vsi, vsig, &chg); + if (status) + goto err_ice_rem_prof_id_flow; + } else { + /* since no existing VSIG supports this + * characteristic pattern, we need to create a + * new VSIG and TCAM entries + */ + status = ice_create_vsig_from_lst(hw, blk, vsi, + ©, &chg); + if (status) + goto err_ice_rem_prof_id_flow; + + /* Adjust priorities */ + status = ice_adj_prof_priorities(hw, blk, vsig, + &chg); + if (status) + goto err_ice_rem_prof_id_flow; + } + } + } else { + status = ICE_ERR_DOES_NOT_EXIST; + } + + /* update hardware tables */ + if (!status) + status = ice_upd_prof_hw(hw, blk, &chg); + +err_ice_rem_prof_id_flow: + list_for_each_entry_safe(del, tmp, &chg, list_entry) { + list_del(&del->list_entry); + devm_kfree(ice_hw_to_dev(hw), del); + } + + list_for_each_entry_safe(del1, tmp1, ©, list) { + list_del(&del1->list); + devm_kfree(ice_hw_to_dev(hw), del1); + } + + return status; +} diff --git a/drivers/net/ethernet/intel/ice/ice_flex_pipe.h b/drivers/net/ethernet/intel/ice/ice_flex_pipe.h index 37eb282742d1..c7b5e1a6ea2b 100644 --- a/drivers/net/ethernet/intel/ice/ice_flex_pipe.h +++ b/drivers/net/ethernet/intel/ice/ice_flex_pipe.h @@ -18,6 +18,13 @@ #define ICE_PKG_CNT 4 +enum ice_status +ice_add_prof(struct ice_hw *hw, enum ice_block blk, u64 id, u8 ptypes[], + struct ice_fv_word *es); +enum ice_status +ice_add_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl); +enum ice_status +ice_rem_prof_id_flow(struct ice_hw *hw, enum ice_block blk, u16 vsi, u64 hdl); enum ice_status ice_init_pkg(struct ice_hw *hw, u8 *buff, u32 len); enum ice_status ice_copy_and_init_pkg(struct ice_hw *hw, const u8 *buf, u32 len); @@ -26,4 +33,6 @@ void ice_free_seg(struct ice_hw *hw); void ice_fill_blk_tbls(struct ice_hw *hw); void ice_clear_hw_tbls(struct ice_hw *hw); void ice_free_hw_tbls(struct ice_hw *hw); +enum ice_status +ice_rem_prof(struct ice_hw *hw, enum ice_block blk, u64 id); #endif /* _ICE_FLEX_PIPE_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_flex_type.h b/drivers/net/ethernet/intel/ice/ice_flex_type.h index 5d5a7eaffa30..0fb3fe3ff3ea 100644 --- a/drivers/net/ethernet/intel/ice/ice_flex_type.h +++ b/drivers/net/ethernet/intel/ice/ice_flex_type.h @@ -3,6 +3,9 @@ #ifndef _ICE_FLEX_TYPE_H_ #define _ICE_FLEX_TYPE_H_ + +#define ICE_FV_OFFSET_INVAL 0x1FF + /* Extraction Sequence (Field Vector) Table */ struct ice_fv_word { u8 prot_id; @@ -105,37 +108,57 @@ struct ice_buf_hdr { sizeof(struct ice_buf_hdr) - (hd_sz)) / (ent_sz)) /* ice package section IDs */ +#define ICE_SID_XLT0_SW 10 +#define ICE_SID_XLT_KEY_BUILDER_SW 11 #define ICE_SID_XLT1_SW 12 #define ICE_SID_XLT2_SW 13 #define ICE_SID_PROFID_TCAM_SW 14 #define ICE_SID_PROFID_REDIR_SW 15 #define ICE_SID_FLD_VEC_SW 16 +#define ICE_SID_CDID_KEY_BUILDER_SW 17 +#define ICE_SID_CDID_REDIR_SW 18 +#define ICE_SID_XLT0_ACL 20 +#define ICE_SID_XLT_KEY_BUILDER_ACL 21 #define ICE_SID_XLT1_ACL 22 #define ICE_SID_XLT2_ACL 23 #define ICE_SID_PROFID_TCAM_ACL 24 #define ICE_SID_PROFID_REDIR_ACL 25 #define ICE_SID_FLD_VEC_ACL 26 +#define ICE_SID_CDID_KEY_BUILDER_ACL 27 +#define ICE_SID_CDID_REDIR_ACL 28 +#define ICE_SID_XLT0_FD 30 +#define ICE_SID_XLT_KEY_BUILDER_FD 31 #define ICE_SID_XLT1_FD 32 #define ICE_SID_XLT2_FD 33 #define ICE_SID_PROFID_TCAM_FD 34 #define ICE_SID_PROFID_REDIR_FD 35 #define ICE_SID_FLD_VEC_FD 36 +#define ICE_SID_CDID_KEY_BUILDER_FD 37 +#define ICE_SID_CDID_REDIR_FD 38 +#define ICE_SID_XLT0_RSS 40 +#define ICE_SID_XLT_KEY_BUILDER_RSS 41 #define ICE_SID_XLT1_RSS 42 #define ICE_SID_XLT2_RSS 43 #define ICE_SID_PROFID_TCAM_RSS 44 #define ICE_SID_PROFID_REDIR_RSS 45 #define ICE_SID_FLD_VEC_RSS 46 +#define ICE_SID_CDID_KEY_BUILDER_RSS 47 +#define ICE_SID_CDID_REDIR_RSS 48 #define ICE_SID_RXPARSER_BOOST_TCAM 56 +#define ICE_SID_XLT0_PE 80 +#define ICE_SID_XLT_KEY_BUILDER_PE 81 #define ICE_SID_XLT1_PE 82 #define ICE_SID_XLT2_PE 83 #define ICE_SID_PROFID_TCAM_PE 84 #define ICE_SID_PROFID_REDIR_PE 85 #define ICE_SID_FLD_VEC_PE 86 +#define ICE_SID_CDID_KEY_BUILDER_PE 87 +#define ICE_SID_CDID_REDIR_PE 88 /* Label Metadata section IDs */ #define ICE_SID_LBL_FIRST 0x80000010 @@ -152,6 +175,19 @@ enum ice_block { ICE_BLK_COUNT }; +enum ice_sect { + ICE_XLT0 = 0, + ICE_XLT_KB, + ICE_XLT1, + ICE_XLT2, + ICE_PROF_TCAM, + ICE_PROF_REDIR, + ICE_VEC_TBL, + ICE_CDID_KB, + ICE_CDID_REDIR, + ICE_SECT_COUNT +}; + /* package labels */ struct ice_label { __le16 value; @@ -234,6 +270,13 @@ struct ice_prof_redir_section { u8 redir_value[1]; }; +/* package buffer building */ + +struct ice_buf_build { + struct ice_buf buf; + u16 reserved_section_table_entries; +}; + struct ice_pkg_enum { struct ice_buf_table *buf_table; u32 buf_idx; @@ -248,6 +291,12 @@ struct ice_pkg_enum { void *(*handler)(u32 sect_type, void *section, u32 index, u32 *offset); }; +struct ice_pkg_es { + __le16 count; + __le16 offset; + struct ice_fv_word es[1]; +}; + struct ice_es { u32 sid; u16 count; @@ -280,6 +329,35 @@ struct ice_ptg_ptype { u8 ptg; }; +#define ICE_MAX_TCAM_PER_PROFILE 32 +#define ICE_MAX_PTG_PER_PROFILE 32 + +struct ice_prof_map { + struct list_head list; + u64 profile_cookie; + u64 context; + u8 prof_id; + u8 ptg_cnt; + u8 ptg[ICE_MAX_PTG_PER_PROFILE]; +}; + +#define ICE_INVALID_TCAM 0xFFFF + +struct ice_tcam_inf { + u16 tcam_idx; + u8 ptg; + u8 prof_id; + u8 in_use; +}; + +struct ice_vsig_prof { + struct list_head list; + u64 profile_cookie; + u8 prof_id; + u8 tcam_count; + struct ice_tcam_inf tcam[ICE_MAX_TCAM_PER_PROFILE]; +}; + struct ice_vsig_entry { struct list_head prop_lst; struct ice_vsig_vsi *first_vsi; @@ -329,6 +407,13 @@ struct ice_xlt2 { u16 count; }; +/* Profile ID Management */ +struct ice_prof_id_key { + __le16 flags; + u8 xlt1; + __le16 xlt2_cdid; +} __packed; + /* Keys are made up of two values, each one-half the size of the key. * For TCAM, the entire key is 80 bits wide (or 2, 40-bit wide values) */ @@ -371,4 +456,31 @@ struct ice_blk_info { u8 is_list_init; }; +enum ice_chg_type { + ICE_TCAM_NONE = 0, + ICE_PTG_ES_ADD, + ICE_TCAM_ADD, + ICE_VSIG_ADD, + ICE_VSIG_REM, + ICE_VSI_MOVE, +}; + +struct ice_chs_chg { + struct list_head list_entry; + enum ice_chg_type type; + + u8 add_ptg; + u8 add_vsig; + u8 add_tcam_idx; + u8 add_prof; + u16 ptype; + u8 ptg; + u8 prof_id; + u16 vsi; + u16 vsig; + u16 orig_vsig; + u16 tcam_idx; +}; + +#define ICE_FLOW_PTYPE_MAX ICE_XLT1_CNT #endif /* _ICE_FLEX_TYPE_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_flow.c b/drivers/net/ethernet/intel/ice/ice_flow.c new file mode 100644 index 000000000000..a05ceb59863b --- /dev/null +++ b/drivers/net/ethernet/intel/ice/ice_flow.c @@ -0,0 +1,1275 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2019, Intel Corporation. */ + +#include "ice_common.h" +#include "ice_flow.h" + +/* Describe properties of a protocol header field */ +struct ice_flow_field_info { + enum ice_flow_seg_hdr hdr; + s16 off; /* Offset from start of a protocol header, in bits */ + u16 size; /* Size of fields in bits */ +}; + +#define ICE_FLOW_FLD_INFO(_hdr, _offset_bytes, _size_bytes) { \ + .hdr = _hdr, \ + .off = (_offset_bytes) * BITS_PER_BYTE, \ + .size = (_size_bytes) * BITS_PER_BYTE, \ +} + +/* Table containing properties of supported protocol header fields */ +static const +struct ice_flow_field_info ice_flds_info[ICE_FLOW_FIELD_IDX_MAX] = { + /* IPv4 / IPv6 */ + /* ICE_FLOW_FIELD_IDX_IPV4_SA */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_IPV4, 12, sizeof(struct in_addr)), + /* ICE_FLOW_FIELD_IDX_IPV4_DA */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_IPV4, 16, sizeof(struct in_addr)), + /* ICE_FLOW_FIELD_IDX_IPV6_SA */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_IPV6, 8, sizeof(struct in6_addr)), + /* ICE_FLOW_FIELD_IDX_IPV6_DA */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_IPV6, 24, sizeof(struct in6_addr)), + /* Transport */ + /* ICE_FLOW_FIELD_IDX_TCP_SRC_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_TCP, 0, sizeof(__be16)), + /* ICE_FLOW_FIELD_IDX_TCP_DST_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_TCP, 2, sizeof(__be16)), + /* ICE_FLOW_FIELD_IDX_UDP_SRC_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_UDP, 0, sizeof(__be16)), + /* ICE_FLOW_FIELD_IDX_UDP_DST_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_UDP, 2, sizeof(__be16)), + /* ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_SCTP, 0, sizeof(__be16)), + /* ICE_FLOW_FIELD_IDX_SCTP_DST_PORT */ + ICE_FLOW_FLD_INFO(ICE_FLOW_SEG_HDR_SCTP, 2, sizeof(__be16)), + +}; + +/* Bitmaps indicating relevant packet types for a particular protocol header + * + * Packet types for packets with an Outer/First/Single IPv4 header + */ +static const u32 ice_ptypes_ipv4_ofos[] = { + 0x1DC00000, 0x04000800, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Packet types for packets with an Innermost/Last IPv4 header */ +static const u32 ice_ptypes_ipv4_il[] = { + 0xE0000000, 0xB807700E, 0x80000003, 0xE01DC03B, + 0x0000000E, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Packet types for packets with an Outer/First/Single IPv6 header */ +static const u32 ice_ptypes_ipv6_ofos[] = { + 0x00000000, 0x00000000, 0x77000000, 0x10002000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Packet types for packets with an Innermost/Last IPv6 header */ +static const u32 ice_ptypes_ipv6_il[] = { + 0x00000000, 0x03B80770, 0x000001DC, 0x0EE00000, + 0x00000770, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* UDP Packet types for non-tunneled packets or tunneled + * packets with inner UDP. + */ +static const u32 ice_ptypes_udp_il[] = { + 0x81000000, 0x20204040, 0x04000010, 0x80810102, + 0x00000040, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Packet types for packets with an Innermost/Last TCP header */ +static const u32 ice_ptypes_tcp_il[] = { + 0x04000000, 0x80810102, 0x10000040, 0x02040408, + 0x00000102, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Packet types for packets with an Innermost/Last SCTP header */ +static const u32 ice_ptypes_sctp_il[] = { + 0x08000000, 0x01020204, 0x20000081, 0x04080810, + 0x00000204, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, + 0x00000000, 0x00000000, 0x00000000, 0x00000000, +}; + +/* Manage parameters and info. used during the creation of a flow profile */ +struct ice_flow_prof_params { + enum ice_block blk; + u16 entry_length; /* # of bytes formatted entry will require */ + u8 es_cnt; + struct ice_flow_prof *prof; + + /* For ACL, the es[0] will have the data of ICE_RX_MDID_PKT_FLAGS_15_0 + * This will give us the direction flags. + */ + struct ice_fv_word es[ICE_MAX_FV_WORDS]; + DECLARE_BITMAP(ptypes, ICE_FLOW_PTYPE_MAX); +}; + +#define ICE_FLOW_SEG_HDRS_L3_MASK \ + (ICE_FLOW_SEG_HDR_IPV4 | ICE_FLOW_SEG_HDR_IPV6) +#define ICE_FLOW_SEG_HDRS_L4_MASK \ + (ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_SCTP) + +/** + * ice_flow_val_hdrs - validates packet segments for valid protocol headers + * @segs: array of one or more packet segments that describe the flow + * @segs_cnt: number of packet segments provided + */ +static enum ice_status +ice_flow_val_hdrs(struct ice_flow_seg_info *segs, u8 segs_cnt) +{ + u8 i; + + for (i = 0; i < segs_cnt; i++) { + /* Multiple L3 headers */ + if (segs[i].hdrs & ICE_FLOW_SEG_HDRS_L3_MASK && + !is_power_of_2(segs[i].hdrs & ICE_FLOW_SEG_HDRS_L3_MASK)) + return ICE_ERR_PARAM; + + /* Multiple L4 headers */ + if (segs[i].hdrs & ICE_FLOW_SEG_HDRS_L4_MASK && + !is_power_of_2(segs[i].hdrs & ICE_FLOW_SEG_HDRS_L4_MASK)) + return ICE_ERR_PARAM; + } + + return 0; +} + +/** + * ice_flow_proc_seg_hdrs - process protocol headers present in pkt segments + * @params: information about the flow to be processed + * + * This function identifies the packet types associated with the protocol + * headers being present in packet segments of the specified flow profile. + */ +static enum ice_status +ice_flow_proc_seg_hdrs(struct ice_flow_prof_params *params) +{ + struct ice_flow_prof *prof; + u8 i; + + memset(params->ptypes, 0xff, sizeof(params->ptypes)); + + prof = params->prof; + + for (i = 0; i < params->prof->segs_cnt; i++) { + const unsigned long *src; + u32 hdrs; + + hdrs = prof->segs[i].hdrs; + + if (hdrs & ICE_FLOW_SEG_HDR_IPV4) { + src = !i ? (const unsigned long *)ice_ptypes_ipv4_ofos : + (const unsigned long *)ice_ptypes_ipv4_il; + bitmap_and(params->ptypes, params->ptypes, src, + ICE_FLOW_PTYPE_MAX); + } else if (hdrs & ICE_FLOW_SEG_HDR_IPV6) { + src = !i ? (const unsigned long *)ice_ptypes_ipv6_ofos : + (const unsigned long *)ice_ptypes_ipv6_il; + bitmap_and(params->ptypes, params->ptypes, src, + ICE_FLOW_PTYPE_MAX); + } + + if (hdrs & ICE_FLOW_SEG_HDR_UDP) { + src = (const unsigned long *)ice_ptypes_udp_il; + bitmap_and(params->ptypes, params->ptypes, src, + ICE_FLOW_PTYPE_MAX); + } else if (hdrs & ICE_FLOW_SEG_HDR_TCP) { + bitmap_and(params->ptypes, params->ptypes, + (const unsigned long *)ice_ptypes_tcp_il, + ICE_FLOW_PTYPE_MAX); + } else if (hdrs & ICE_FLOW_SEG_HDR_SCTP) { + src = (const unsigned long *)ice_ptypes_sctp_il; + bitmap_and(params->ptypes, params->ptypes, src, + ICE_FLOW_PTYPE_MAX); + } + } + + return 0; +} + +/** + * ice_flow_xtract_fld - Create an extraction sequence entry for the given field + * @hw: pointer to the HW struct + * @params: information about the flow to be processed + * @seg: packet segment index of the field to be extracted + * @fld: ID of field to be extracted + * + * This function determines the protocol ID, offset, and size of the given + * field. It then allocates one or more extraction sequence entries for the + * given field, and fill the entries with protocol ID and offset information. + */ +static enum ice_status +ice_flow_xtract_fld(struct ice_hw *hw, struct ice_flow_prof_params *params, + u8 seg, enum ice_flow_field fld) +{ + enum ice_prot_id prot_id = ICE_PROT_ID_INVAL; + u8 fv_words = hw->blk[params->blk].es.fvw; + struct ice_flow_fld_info *flds; + u16 cnt, ese_bits, i; + u16 off; + + flds = params->prof->segs[seg].fields; + + switch (fld) { + case ICE_FLOW_FIELD_IDX_IPV4_SA: + case ICE_FLOW_FIELD_IDX_IPV4_DA: + prot_id = seg == 0 ? ICE_PROT_IPV4_OF_OR_S : ICE_PROT_IPV4_IL; + break; + case ICE_FLOW_FIELD_IDX_IPV6_SA: + case ICE_FLOW_FIELD_IDX_IPV6_DA: + prot_id = seg == 0 ? ICE_PROT_IPV6_OF_OR_S : ICE_PROT_IPV6_IL; + break; + case ICE_FLOW_FIELD_IDX_TCP_SRC_PORT: + case ICE_FLOW_FIELD_IDX_TCP_DST_PORT: + prot_id = ICE_PROT_TCP_IL; + break; + case ICE_FLOW_FIELD_IDX_UDP_SRC_PORT: + case ICE_FLOW_FIELD_IDX_UDP_DST_PORT: + prot_id = ICE_PROT_UDP_IL_OR_S; + break; + case ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT: + case ICE_FLOW_FIELD_IDX_SCTP_DST_PORT: + prot_id = ICE_PROT_SCTP_IL; + break; + default: + return ICE_ERR_NOT_IMPL; + } + + /* Each extraction sequence entry is a word in size, and extracts a + * word-aligned offset from a protocol header. + */ + ese_bits = ICE_FLOW_FV_EXTRACT_SZ * BITS_PER_BYTE; + + flds[fld].xtrct.prot_id = prot_id; + flds[fld].xtrct.off = (ice_flds_info[fld].off / ese_bits) * + ICE_FLOW_FV_EXTRACT_SZ; + flds[fld].xtrct.disp = (u8)(ice_flds_info[fld].off % ese_bits); + flds[fld].xtrct.idx = params->es_cnt; + + /* Adjust the next field-entry index after accommodating the number of + * entries this field consumes + */ + cnt = DIV_ROUND_UP(flds[fld].xtrct.disp + ice_flds_info[fld].size, + ese_bits); + + /* Fill in the extraction sequence entries needed for this field */ + off = flds[fld].xtrct.off; + for (i = 0; i < cnt; i++) { + u8 idx; + + /* Make sure the number of extraction sequence required + * does not exceed the block's capability + */ + if (params->es_cnt >= fv_words) + return ICE_ERR_MAX_LIMIT; + + /* some blocks require a reversed field vector layout */ + if (hw->blk[params->blk].es.reverse) + idx = fv_words - params->es_cnt - 1; + else + idx = params->es_cnt; + + params->es[idx].prot_id = prot_id; + params->es[idx].off = off; + params->es_cnt++; + + off += ICE_FLOW_FV_EXTRACT_SZ; + } + + return 0; +} + +/** + * ice_flow_create_xtrct_seq - Create an extraction sequence for given segments + * @hw: pointer to the HW struct + * @params: information about the flow to be processed + * + * This function iterates through all matched fields in the given segments, and + * creates an extraction sequence for the fields. + */ +static enum ice_status +ice_flow_create_xtrct_seq(struct ice_hw *hw, + struct ice_flow_prof_params *params) +{ + struct ice_flow_prof *prof = params->prof; + enum ice_status status = 0; + u8 i; + + for (i = 0; i < prof->segs_cnt; i++) { + u8 j; + + for_each_set_bit(j, (unsigned long *)&prof->segs[i].match, + ICE_FLOW_FIELD_IDX_MAX) { + status = ice_flow_xtract_fld(hw, params, i, + (enum ice_flow_field)j); + if (status) + return status; + } + } + + return status; +} + +/** + * ice_flow_proc_segs - process all packet segments associated with a profile + * @hw: pointer to the HW struct + * @params: information about the flow to be processed + */ +static enum ice_status +ice_flow_proc_segs(struct ice_hw *hw, struct ice_flow_prof_params *params) +{ + enum ice_status status; + + status = ice_flow_proc_seg_hdrs(params); + if (status) + return status; + + status = ice_flow_create_xtrct_seq(hw, params); + if (status) + return status; + + switch (params->blk) { + case ICE_BLK_RSS: + /* Only header information is provided for RSS configuration. + * No further processing is needed. + */ + status = 0; + break; + default: + return ICE_ERR_NOT_IMPL; + } + + return status; +} + +#define ICE_FLOW_FIND_PROF_CHK_FLDS 0x00000001 +#define ICE_FLOW_FIND_PROF_CHK_VSI 0x00000002 +#define ICE_FLOW_FIND_PROF_NOT_CHK_DIR 0x00000004 + +/** + * ice_flow_find_prof_conds - Find a profile matching headers and conditions + * @hw: pointer to the HW struct + * @blk: classification stage + * @dir: flow direction + * @segs: array of one or more packet segments that describe the flow + * @segs_cnt: number of packet segments provided + * @vsi_handle: software VSI handle to check VSI (ICE_FLOW_FIND_PROF_CHK_VSI) + * @conds: additional conditions to be checked (ICE_FLOW_FIND_PROF_CHK_*) + */ +static struct ice_flow_prof * +ice_flow_find_prof_conds(struct ice_hw *hw, enum ice_block blk, + enum ice_flow_dir dir, struct ice_flow_seg_info *segs, + u8 segs_cnt, u16 vsi_handle, u32 conds) +{ + struct ice_flow_prof *p, *prof = NULL; + + mutex_lock(&hw->fl_profs_locks[blk]); + list_for_each_entry(p, &hw->fl_profs[blk], l_entry) + if ((p->dir == dir || conds & ICE_FLOW_FIND_PROF_NOT_CHK_DIR) && + segs_cnt && segs_cnt == p->segs_cnt) { + u8 i; + + /* Check for profile-VSI association if specified */ + if ((conds & ICE_FLOW_FIND_PROF_CHK_VSI) && + ice_is_vsi_valid(hw, vsi_handle) && + !test_bit(vsi_handle, p->vsis)) + continue; + + /* Protocol headers must be checked. Matched fields are + * checked if specified. + */ + for (i = 0; i < segs_cnt; i++) + if (segs[i].hdrs != p->segs[i].hdrs || + ((conds & ICE_FLOW_FIND_PROF_CHK_FLDS) && + segs[i].match != p->segs[i].match)) + break; + + /* A match is found if all segments are matched */ + if (i == segs_cnt) { + prof = p; + break; + } + } + mutex_unlock(&hw->fl_profs_locks[blk]); + + return prof; +} + +/** + * ice_flow_find_prof_id - Look up a profile with given profile ID + * @hw: pointer to the HW struct + * @blk: classification stage + * @prof_id: unique ID to identify this flow profile + */ +static struct ice_flow_prof * +ice_flow_find_prof_id(struct ice_hw *hw, enum ice_block blk, u64 prof_id) +{ + struct ice_flow_prof *p; + + list_for_each_entry(p, &hw->fl_profs[blk], l_entry) + if (p->id == prof_id) + return p; + + return NULL; +} + +/** + * ice_flow_add_prof_sync - Add a flow profile for packet segments and fields + * @hw: pointer to the HW struct + * @blk: classification stage + * @dir: flow direction + * @prof_id: unique ID to identify this flow profile + * @segs: array of one or more packet segments that describe the flow + * @segs_cnt: number of packet segments provided + * @prof: stores the returned flow profile added + * + * Assumption: the caller has acquired the lock to the profile list + */ +static enum ice_status +ice_flow_add_prof_sync(struct ice_hw *hw, enum ice_block blk, + enum ice_flow_dir dir, u64 prof_id, + struct ice_flow_seg_info *segs, u8 segs_cnt, + struct ice_flow_prof **prof) +{ + struct ice_flow_prof_params params; + enum ice_status status; + u8 i; + + if (!prof) + return ICE_ERR_BAD_PTR; + + memset(¶ms, 0, sizeof(params)); + params.prof = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*params.prof), + GFP_KERNEL); + if (!params.prof) + return ICE_ERR_NO_MEMORY; + + /* initialize extraction sequence to all invalid (0xff) */ + for (i = 0; i < ICE_MAX_FV_WORDS; i++) { + params.es[i].prot_id = ICE_PROT_INVALID; + params.es[i].off = ICE_FV_OFFSET_INVAL; + } + + params.blk = blk; + params.prof->id = prof_id; + params.prof->dir = dir; + params.prof->segs_cnt = segs_cnt; + + /* Make a copy of the segments that need to be persistent in the flow + * profile instance + */ + for (i = 0; i < segs_cnt; i++) + memcpy(¶ms.prof->segs[i], &segs[i], sizeof(*segs)); + + status = ice_flow_proc_segs(hw, ¶ms); + if (status) { + ice_debug(hw, ICE_DBG_FLOW, + "Error processing a flow's packet segments\n"); + goto out; + } + + /* Add a HW profile for this flow profile */ + status = ice_add_prof(hw, blk, prof_id, (u8 *)params.ptypes, params.es); + if (status) { + ice_debug(hw, ICE_DBG_FLOW, "Error adding a HW flow profile\n"); + goto out; + } + + INIT_LIST_HEAD(¶ms.prof->entries); + mutex_init(¶ms.prof->entries_lock); + *prof = params.prof; + +out: + if (status) + devm_kfree(ice_hw_to_dev(hw), params.prof); + + return status; +} + +/** + * ice_flow_rem_prof_sync - remove a flow profile + * @hw: pointer to the hardware structure + * @blk: classification stage + * @prof: pointer to flow profile to remove + * + * Assumption: the caller has acquired the lock to the profile list + */ +static enum ice_status +ice_flow_rem_prof_sync(struct ice_hw *hw, enum ice_block blk, + struct ice_flow_prof *prof) +{ + enum ice_status status; + + /* Remove all hardware profiles associated with this flow profile */ + status = ice_rem_prof(hw, blk, prof->id); + if (!status) { + list_del(&prof->l_entry); + mutex_destroy(&prof->entries_lock); + devm_kfree(ice_hw_to_dev(hw), prof); + } + + return status; +} + +/** + * ice_flow_assoc_prof - associate a VSI with a flow profile + * @hw: pointer to the hardware structure + * @blk: classification stage + * @prof: pointer to flow profile + * @vsi_handle: software VSI handle + * + * Assumption: the caller has acquired the lock to the profile list + * and the software VSI handle has been validated + */ +static enum ice_status +ice_flow_assoc_prof(struct ice_hw *hw, enum ice_block blk, + struct ice_flow_prof *prof, u16 vsi_handle) +{ + enum ice_status status = 0; + + if (!test_bit(vsi_handle, prof->vsis)) { + status = ice_add_prof_id_flow(hw, blk, + ice_get_hw_vsi_num(hw, + vsi_handle), + prof->id); + if (!status) + set_bit(vsi_handle, prof->vsis); + else + ice_debug(hw, ICE_DBG_FLOW, + "HW profile add failed, %d\n", + status); + } + + return status; +} + +/** + * ice_flow_disassoc_prof - disassociate a VSI from a flow profile + * @hw: pointer to the hardware structure + * @blk: classification stage + * @prof: pointer to flow profile + * @vsi_handle: software VSI handle + * + * Assumption: the caller has acquired the lock to the profile list + * and the software VSI handle has been validated + */ +static enum ice_status +ice_flow_disassoc_prof(struct ice_hw *hw, enum ice_block blk, + struct ice_flow_prof *prof, u16 vsi_handle) +{ + enum ice_status status = 0; + + if (test_bit(vsi_handle, prof->vsis)) { + status = ice_rem_prof_id_flow(hw, blk, + ice_get_hw_vsi_num(hw, + vsi_handle), + prof->id); + if (!status) + clear_bit(vsi_handle, prof->vsis); + else + ice_debug(hw, ICE_DBG_FLOW, + "HW profile remove failed, %d\n", + status); + } + + return status; +} + +/** + * ice_flow_add_prof - Add a flow profile for packet segments and matched fields + * @hw: pointer to the HW struct + * @blk: classification stage + * @dir: flow direction + * @prof_id: unique ID to identify this flow profile + * @segs: array of one or more packet segments that describe the flow + * @segs_cnt: number of packet segments provided + * @prof: stores the returned flow profile added + */ +static enum ice_status +ice_flow_add_prof(struct ice_hw *hw, enum ice_block blk, enum ice_flow_dir dir, + u64 prof_id, struct ice_flow_seg_info *segs, u8 segs_cnt, + struct ice_flow_prof **prof) +{ + enum ice_status status; + + if (segs_cnt > ICE_FLOW_SEG_MAX) + return ICE_ERR_MAX_LIMIT; + + if (!segs_cnt) + return ICE_ERR_PARAM; + + if (!segs) + return ICE_ERR_BAD_PTR; + + status = ice_flow_val_hdrs(segs, segs_cnt); + if (status) + return status; + + mutex_lock(&hw->fl_profs_locks[blk]); + + status = ice_flow_add_prof_sync(hw, blk, dir, prof_id, segs, segs_cnt, + prof); + if (!status) + list_add(&(*prof)->l_entry, &hw->fl_profs[blk]); + + mutex_unlock(&hw->fl_profs_locks[blk]); + + return status; +} + +/** + * ice_flow_rem_prof - Remove a flow profile and all entries associated with it + * @hw: pointer to the HW struct + * @blk: the block for which the flow profile is to be removed + * @prof_id: unique ID of the flow profile to be removed + */ +static enum ice_status +ice_flow_rem_prof(struct ice_hw *hw, enum ice_block blk, u64 prof_id) +{ + struct ice_flow_prof *prof; + enum ice_status status; + + mutex_lock(&hw->fl_profs_locks[blk]); + + prof = ice_flow_find_prof_id(hw, blk, prof_id); + if (!prof) { + status = ICE_ERR_DOES_NOT_EXIST; + goto out; + } + + /* prof becomes invalid after the call */ + status = ice_flow_rem_prof_sync(hw, blk, prof); + +out: + mutex_unlock(&hw->fl_profs_locks[blk]); + + return status; +} + +/** + * ice_flow_set_fld_ext - specifies locations of field from entry's input buffer + * @seg: packet segment the field being set belongs to + * @fld: field to be set + * @type: type of the field + * @val_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of the value to match from + * entry's input buffer + * @mask_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of mask value from entry's + * input buffer + * @last_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of last/upper value from + * entry's input buffer + * + * This helper function stores information of a field being matched, including + * the type of the field and the locations of the value to match, the mask, and + * and the upper-bound value in the start of the input buffer for a flow entry. + * This function should only be used for fixed-size data structures. + * + * This function also opportunistically determines the protocol headers to be + * present based on the fields being set. Some fields cannot be used alone to + * determine the protocol headers present. Sometimes, fields for particular + * protocol headers are not matched. In those cases, the protocol headers + * must be explicitly set. + */ +static void +ice_flow_set_fld_ext(struct ice_flow_seg_info *seg, enum ice_flow_field fld, + enum ice_flow_fld_match_type type, u16 val_loc, + u16 mask_loc, u16 last_loc) +{ + u64 bit = BIT_ULL(fld); + + seg->match |= bit; + if (type == ICE_FLOW_FLD_TYPE_RANGE) + seg->range |= bit; + + seg->fields[fld].type = type; + seg->fields[fld].src.val = val_loc; + seg->fields[fld].src.mask = mask_loc; + seg->fields[fld].src.last = last_loc; + + ICE_FLOW_SET_HDRS(seg, ice_flds_info[fld].hdr); +} + +/** + * ice_flow_set_fld - specifies locations of field from entry's input buffer + * @seg: packet segment the field being set belongs to + * @fld: field to be set + * @val_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of the value to match from + * entry's input buffer + * @mask_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of mask value from entry's + * input buffer + * @last_loc: if not ICE_FLOW_FLD_OFF_INVAL, location of last/upper value from + * entry's input buffer + * @range: indicate if field being matched is to be in a range + * + * This function specifies the locations, in the form of byte offsets from the + * start of the input buffer for a flow entry, from where the value to match, + * the mask value, and upper value can be extracted. These locations are then + * stored in the flow profile. When adding a flow entry associated with the + * flow profile, these locations will be used to quickly extract the values and + * create the content of a match entry. This function should only be used for + * fixed-size data structures. + */ +static void +ice_flow_set_fld(struct ice_flow_seg_info *seg, enum ice_flow_field fld, + u16 val_loc, u16 mask_loc, u16 last_loc, bool range) +{ + enum ice_flow_fld_match_type t = range ? + ICE_FLOW_FLD_TYPE_RANGE : ICE_FLOW_FLD_TYPE_REG; + + ice_flow_set_fld_ext(seg, fld, t, val_loc, mask_loc, last_loc); +} + +#define ICE_FLOW_RSS_SEG_HDR_L3_MASKS \ + (ICE_FLOW_SEG_HDR_IPV4 | ICE_FLOW_SEG_HDR_IPV6) + +#define ICE_FLOW_RSS_SEG_HDR_L4_MASKS \ + (ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_SCTP) + +#define ICE_FLOW_RSS_SEG_HDR_VAL_MASKS \ + (ICE_FLOW_RSS_SEG_HDR_L3_MASKS | \ + ICE_FLOW_RSS_SEG_HDR_L4_MASKS) + +/** + * ice_flow_set_rss_seg_info - setup packet segments for RSS + * @segs: pointer to the flow field segment(s) + * @hash_fields: fields to be hashed on for the segment(s) + * @flow_hdr: protocol header fields within a packet segment + * + * Helper function to extract fields from hash bitmap and use flow + * header value to set flow field segment for further use in flow + * profile entry or removal. + */ +static enum ice_status +ice_flow_set_rss_seg_info(struct ice_flow_seg_info *segs, u64 hash_fields, + u32 flow_hdr) +{ + u64 val; + u8 i; + + for_each_set_bit(i, (unsigned long *)&hash_fields, + ICE_FLOW_FIELD_IDX_MAX) + ice_flow_set_fld(segs, (enum ice_flow_field)i, + ICE_FLOW_FLD_OFF_INVAL, ICE_FLOW_FLD_OFF_INVAL, + ICE_FLOW_FLD_OFF_INVAL, false); + + ICE_FLOW_SET_HDRS(segs, flow_hdr); + + if (segs->hdrs & ~ICE_FLOW_RSS_SEG_HDR_VAL_MASKS) + return ICE_ERR_PARAM; + + val = (u64)(segs->hdrs & ICE_FLOW_RSS_SEG_HDR_L3_MASKS); + if (val && !is_power_of_2(val)) + return ICE_ERR_CFG; + + val = (u64)(segs->hdrs & ICE_FLOW_RSS_SEG_HDR_L4_MASKS); + if (val && !is_power_of_2(val)) + return ICE_ERR_CFG; + + return 0; +} + +/** + * ice_rem_vsi_rss_list - remove VSI from RSS list + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * + * Remove the VSI from all RSS configurations in the list. + */ +void ice_rem_vsi_rss_list(struct ice_hw *hw, u16 vsi_handle) +{ + struct ice_rss_cfg *r, *tmp; + + if (list_empty(&hw->rss_list_head)) + return; + + mutex_lock(&hw->rss_locks); + list_for_each_entry_safe(r, tmp, &hw->rss_list_head, l_entry) + if (test_and_clear_bit(vsi_handle, r->vsis)) + if (bitmap_empty(r->vsis, ICE_MAX_VSI)) { + list_del(&r->l_entry); + devm_kfree(ice_hw_to_dev(hw), r); + } + mutex_unlock(&hw->rss_locks); +} + +/** + * ice_rem_vsi_rss_cfg - remove RSS configurations associated with VSI + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * + * This function will iterate through all flow profiles and disassociate + * the VSI from that profile. If the flow profile has no VSIs it will + * be removed. + */ +enum ice_status ice_rem_vsi_rss_cfg(struct ice_hw *hw, u16 vsi_handle) +{ + const enum ice_block blk = ICE_BLK_RSS; + struct ice_flow_prof *p, *t; + enum ice_status status = 0; + + if (!ice_is_vsi_valid(hw, vsi_handle)) + return ICE_ERR_PARAM; + + if (list_empty(&hw->fl_profs[blk])) + return 0; + + mutex_lock(&hw->fl_profs_locks[blk]); + list_for_each_entry_safe(p, t, &hw->fl_profs[blk], l_entry) + if (test_bit(vsi_handle, p->vsis)) { + status = ice_flow_disassoc_prof(hw, blk, p, vsi_handle); + if (status) + break; + + if (bitmap_empty(p->vsis, ICE_MAX_VSI)) { + status = ice_flow_rem_prof_sync(hw, blk, p); + if (status) + break; + } + } + mutex_unlock(&hw->fl_profs_locks[blk]); + + return status; +} + +/** + * ice_rem_rss_list - remove RSS configuration from list + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @prof: pointer to flow profile + * + * Assumption: lock has already been acquired for RSS list + */ +static void +ice_rem_rss_list(struct ice_hw *hw, u16 vsi_handle, struct ice_flow_prof *prof) +{ + struct ice_rss_cfg *r, *tmp; + + /* Search for RSS hash fields associated to the VSI that match the + * hash configurations associated to the flow profile. If found + * remove from the RSS entry list of the VSI context and delete entry. + */ + list_for_each_entry_safe(r, tmp, &hw->rss_list_head, l_entry) + if (r->hashed_flds == prof->segs[prof->segs_cnt - 1].match && + r->packet_hdr == prof->segs[prof->segs_cnt - 1].hdrs) { + clear_bit(vsi_handle, r->vsis); + if (bitmap_empty(r->vsis, ICE_MAX_VSI)) { + list_del(&r->l_entry); + devm_kfree(ice_hw_to_dev(hw), r); + } + return; + } +} + +/** + * ice_add_rss_list - add RSS configuration to list + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @prof: pointer to flow profile + * + * Assumption: lock has already been acquired for RSS list + */ +static enum ice_status +ice_add_rss_list(struct ice_hw *hw, u16 vsi_handle, struct ice_flow_prof *prof) +{ + struct ice_rss_cfg *r, *rss_cfg; + + list_for_each_entry(r, &hw->rss_list_head, l_entry) + if (r->hashed_flds == prof->segs[prof->segs_cnt - 1].match && + r->packet_hdr == prof->segs[prof->segs_cnt - 1].hdrs) { + set_bit(vsi_handle, r->vsis); + return 0; + } + + rss_cfg = devm_kzalloc(ice_hw_to_dev(hw), sizeof(*rss_cfg), + GFP_KERNEL); + if (!rss_cfg) + return ICE_ERR_NO_MEMORY; + + rss_cfg->hashed_flds = prof->segs[prof->segs_cnt - 1].match; + rss_cfg->packet_hdr = prof->segs[prof->segs_cnt - 1].hdrs; + set_bit(vsi_handle, rss_cfg->vsis); + + list_add_tail(&rss_cfg->l_entry, &hw->rss_list_head); + + return 0; +} + +#define ICE_FLOW_PROF_HASH_S 0 +#define ICE_FLOW_PROF_HASH_M (0xFFFFFFFFULL << ICE_FLOW_PROF_HASH_S) +#define ICE_FLOW_PROF_HDR_S 32 +#define ICE_FLOW_PROF_HDR_M (0x3FFFFFFFULL << ICE_FLOW_PROF_HDR_S) +#define ICE_FLOW_PROF_ENCAP_S 63 +#define ICE_FLOW_PROF_ENCAP_M (BIT_ULL(ICE_FLOW_PROF_ENCAP_S)) + +#define ICE_RSS_OUTER_HEADERS 1 + +/* Flow profile ID format: + * [0:31] - Packet match fields + * [32:62] - Protocol header + * [63] - Encapsulation flag, 0 if non-tunneled, 1 if tunneled + */ +#define ICE_FLOW_GEN_PROFID(hash, hdr, segs_cnt) \ + (u64)(((u64)(hash) & ICE_FLOW_PROF_HASH_M) | \ + (((u64)(hdr) << ICE_FLOW_PROF_HDR_S) & ICE_FLOW_PROF_HDR_M) | \ + ((u8)((segs_cnt) - 1) ? ICE_FLOW_PROF_ENCAP_M : 0)) + +/** + * ice_add_rss_cfg_sync - add an RSS configuration + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @hashed_flds: hash bit fields (ICE_FLOW_HASH_*) to configure + * @addl_hdrs: protocol header fields + * @segs_cnt: packet segment count + * + * Assumption: lock has already been acquired for RSS list + */ +static enum ice_status +ice_add_rss_cfg_sync(struct ice_hw *hw, u16 vsi_handle, u64 hashed_flds, + u32 addl_hdrs, u8 segs_cnt) +{ + const enum ice_block blk = ICE_BLK_RSS; + struct ice_flow_prof *prof = NULL; + struct ice_flow_seg_info *segs; + enum ice_status status; + + if (!segs_cnt || segs_cnt > ICE_FLOW_SEG_MAX) + return ICE_ERR_PARAM; + + segs = kcalloc(segs_cnt, sizeof(*segs), GFP_KERNEL); + if (!segs) + return ICE_ERR_NO_MEMORY; + + /* Construct the packet segment info from the hashed fields */ + status = ice_flow_set_rss_seg_info(&segs[segs_cnt - 1], hashed_flds, + addl_hdrs); + if (status) + goto exit; + + /* Search for a flow profile that has matching headers, hash fields + * and has the input VSI associated to it. If found, no further + * operations required and exit. + */ + prof = ice_flow_find_prof_conds(hw, blk, ICE_FLOW_RX, segs, segs_cnt, + vsi_handle, + ICE_FLOW_FIND_PROF_CHK_FLDS | + ICE_FLOW_FIND_PROF_CHK_VSI); + if (prof) + goto exit; + + /* Check if a flow profile exists with the same protocol headers and + * associated with the input VSI. If so disassociate the VSI from + * this profile. The VSI will be added to a new profile created with + * the protocol header and new hash field configuration. + */ + prof = ice_flow_find_prof_conds(hw, blk, ICE_FLOW_RX, segs, segs_cnt, + vsi_handle, ICE_FLOW_FIND_PROF_CHK_VSI); + if (prof) { + status = ice_flow_disassoc_prof(hw, blk, prof, vsi_handle); + if (!status) + ice_rem_rss_list(hw, vsi_handle, prof); + else + goto exit; + + /* Remove profile if it has no VSIs associated */ + if (bitmap_empty(prof->vsis, ICE_MAX_VSI)) { + status = ice_flow_rem_prof(hw, blk, prof->id); + if (status) + goto exit; + } + } + + /* Search for a profile that has same match fields only. If this + * exists then associate the VSI to this profile. + */ + prof = ice_flow_find_prof_conds(hw, blk, ICE_FLOW_RX, segs, segs_cnt, + vsi_handle, + ICE_FLOW_FIND_PROF_CHK_FLDS); + if (prof) { + status = ice_flow_assoc_prof(hw, blk, prof, vsi_handle); + if (!status) + status = ice_add_rss_list(hw, vsi_handle, prof); + goto exit; + } + + /* Create a new flow profile with generated profile and packet + * segment information. + */ + status = ice_flow_add_prof(hw, blk, ICE_FLOW_RX, + ICE_FLOW_GEN_PROFID(hashed_flds, + segs[segs_cnt - 1].hdrs, + segs_cnt), + segs, segs_cnt, &prof); + if (status) + goto exit; + + status = ice_flow_assoc_prof(hw, blk, prof, vsi_handle); + /* If association to a new flow profile failed then this profile can + * be removed. + */ + if (status) { + ice_flow_rem_prof(hw, blk, prof->id); + goto exit; + } + + status = ice_add_rss_list(hw, vsi_handle, prof); + +exit: + kfree(segs); + return status; +} + +/** + * ice_add_rss_cfg - add an RSS configuration with specified hashed fields + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @hashed_flds: hash bit fields (ICE_FLOW_HASH_*) to configure + * @addl_hdrs: protocol header fields + * + * This function will generate a flow profile based on fields associated with + * the input fields to hash on, the flow type and use the VSI number to add + * a flow entry to the profile. + */ +enum ice_status +ice_add_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u64 hashed_flds, + u32 addl_hdrs) +{ + enum ice_status status; + + if (hashed_flds == ICE_HASH_INVALID || + !ice_is_vsi_valid(hw, vsi_handle)) + return ICE_ERR_PARAM; + + mutex_lock(&hw->rss_locks); + status = ice_add_rss_cfg_sync(hw, vsi_handle, hashed_flds, addl_hdrs, + ICE_RSS_OUTER_HEADERS); + mutex_unlock(&hw->rss_locks); + + return status; +} + +/* Mapping of AVF hash bit fields to an L3-L4 hash combination. + * As the ice_flow_avf_hdr_field represent individual bit shifts in a hash, + * convert its values to their appropriate flow L3, L4 values. + */ +#define ICE_FLOW_AVF_RSS_IPV4_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_OTHER) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_FRAG_IPV4)) +#define ICE_FLOW_AVF_RSS_TCP_IPV4_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_TCP_SYN_NO_ACK) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_TCP)) +#define ICE_FLOW_AVF_RSS_UDP_IPV4_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_UNICAST_IPV4_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_MULTICAST_IPV4_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_UDP)) +#define ICE_FLOW_AVF_RSS_ALL_IPV4_MASKS \ + (ICE_FLOW_AVF_RSS_TCP_IPV4_MASKS | ICE_FLOW_AVF_RSS_UDP_IPV4_MASKS | \ + ICE_FLOW_AVF_RSS_IPV4_MASKS | BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_SCTP)) + +#define ICE_FLOW_AVF_RSS_IPV6_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_OTHER) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_FRAG_IPV6)) +#define ICE_FLOW_AVF_RSS_UDP_IPV6_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_UNICAST_IPV6_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_MULTICAST_IPV6_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_UDP)) +#define ICE_FLOW_AVF_RSS_TCP_IPV6_MASKS \ + (BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_TCP_SYN_NO_ACK) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_TCP)) +#define ICE_FLOW_AVF_RSS_ALL_IPV6_MASKS \ + (ICE_FLOW_AVF_RSS_TCP_IPV6_MASKS | ICE_FLOW_AVF_RSS_UDP_IPV6_MASKS | \ + ICE_FLOW_AVF_RSS_IPV6_MASKS | BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_SCTP)) + +/** + * ice_add_avf_rss_cfg - add an RSS configuration for AVF driver + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @avf_hash: hash bit fields (ICE_AVF_FLOW_FIELD_*) to configure + * + * This function will take the hash bitmap provided by the AVF driver via a + * message, convert it to ICE-compatible values, and configure RSS flow + * profiles. + */ +enum ice_status +ice_add_avf_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u64 avf_hash) +{ + enum ice_status status = 0; + u64 hash_flds; + + if (avf_hash == ICE_AVF_FLOW_FIELD_INVALID || + !ice_is_vsi_valid(hw, vsi_handle)) + return ICE_ERR_PARAM; + + /* Make sure no unsupported bits are specified */ + if (avf_hash & ~(ICE_FLOW_AVF_RSS_ALL_IPV4_MASKS | + ICE_FLOW_AVF_RSS_ALL_IPV6_MASKS)) + return ICE_ERR_CFG; + + hash_flds = avf_hash; + + /* Always create an L3 RSS configuration for any L4 RSS configuration */ + if (hash_flds & ICE_FLOW_AVF_RSS_ALL_IPV4_MASKS) + hash_flds |= ICE_FLOW_AVF_RSS_IPV4_MASKS; + + if (hash_flds & ICE_FLOW_AVF_RSS_ALL_IPV6_MASKS) + hash_flds |= ICE_FLOW_AVF_RSS_IPV6_MASKS; + + /* Create the corresponding RSS configuration for each valid hash bit */ + while (hash_flds) { + u64 rss_hash = ICE_HASH_INVALID; + + if (hash_flds & ICE_FLOW_AVF_RSS_ALL_IPV4_MASKS) { + if (hash_flds & ICE_FLOW_AVF_RSS_IPV4_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV4; + hash_flds &= ~ICE_FLOW_AVF_RSS_IPV4_MASKS; + } else if (hash_flds & + ICE_FLOW_AVF_RSS_TCP_IPV4_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV4 | + ICE_FLOW_HASH_TCP_PORT; + hash_flds &= ~ICE_FLOW_AVF_RSS_TCP_IPV4_MASKS; + } else if (hash_flds & + ICE_FLOW_AVF_RSS_UDP_IPV4_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV4 | + ICE_FLOW_HASH_UDP_PORT; + hash_flds &= ~ICE_FLOW_AVF_RSS_UDP_IPV4_MASKS; + } else if (hash_flds & + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_SCTP)) { + rss_hash = ICE_FLOW_HASH_IPV4 | + ICE_FLOW_HASH_SCTP_PORT; + hash_flds &= + ~BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_SCTP); + } + } else if (hash_flds & ICE_FLOW_AVF_RSS_ALL_IPV6_MASKS) { + if (hash_flds & ICE_FLOW_AVF_RSS_IPV6_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV6; + hash_flds &= ~ICE_FLOW_AVF_RSS_IPV6_MASKS; + } else if (hash_flds & + ICE_FLOW_AVF_RSS_TCP_IPV6_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV6 | + ICE_FLOW_HASH_TCP_PORT; + hash_flds &= ~ICE_FLOW_AVF_RSS_TCP_IPV6_MASKS; + } else if (hash_flds & + ICE_FLOW_AVF_RSS_UDP_IPV6_MASKS) { + rss_hash = ICE_FLOW_HASH_IPV6 | + ICE_FLOW_HASH_UDP_PORT; + hash_flds &= ~ICE_FLOW_AVF_RSS_UDP_IPV6_MASKS; + } else if (hash_flds & + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_SCTP)) { + rss_hash = ICE_FLOW_HASH_IPV6 | + ICE_FLOW_HASH_SCTP_PORT; + hash_flds &= + ~BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_SCTP); + } + } + + if (rss_hash == ICE_HASH_INVALID) + return ICE_ERR_OUT_OF_RANGE; + + status = ice_add_rss_cfg(hw, vsi_handle, rss_hash, + ICE_FLOW_SEG_HDR_NONE); + if (status) + break; + } + + return status; +} + +/** + * ice_replay_rss_cfg - replay RSS configurations associated with VSI + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + */ +enum ice_status ice_replay_rss_cfg(struct ice_hw *hw, u16 vsi_handle) +{ + enum ice_status status = 0; + struct ice_rss_cfg *r; + + if (!ice_is_vsi_valid(hw, vsi_handle)) + return ICE_ERR_PARAM; + + mutex_lock(&hw->rss_locks); + list_for_each_entry(r, &hw->rss_list_head, l_entry) { + if (test_bit(vsi_handle, r->vsis)) { + status = ice_add_rss_cfg_sync(hw, vsi_handle, + r->hashed_flds, + r->packet_hdr, + ICE_RSS_OUTER_HEADERS); + if (status) + break; + } + } + mutex_unlock(&hw->rss_locks); + + return status; +} + +/** + * ice_get_rss_cfg - returns hashed fields for the given header types + * @hw: pointer to the hardware structure + * @vsi_handle: software VSI handle + * @hdrs: protocol header type + * + * This function will return the match fields of the first instance of flow + * profile having the given header types and containing input VSI + */ +u64 ice_get_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u32 hdrs) +{ + struct ice_rss_cfg *r, *rss_cfg = NULL; + + /* verify if the protocol header is non zero and VSI is valid */ + if (hdrs == ICE_FLOW_SEG_HDR_NONE || !ice_is_vsi_valid(hw, vsi_handle)) + return ICE_HASH_INVALID; + + mutex_lock(&hw->rss_locks); + list_for_each_entry(r, &hw->rss_list_head, l_entry) + if (test_bit(vsi_handle, r->vsis) && + r->packet_hdr == hdrs) { + rss_cfg = r; + break; + } + mutex_unlock(&hw->rss_locks); + + return rss_cfg ? rss_cfg->hashed_flds : ICE_HASH_INVALID; +} diff --git a/drivers/net/ethernet/intel/ice/ice_flow.h b/drivers/net/ethernet/intel/ice/ice_flow.h new file mode 100644 index 000000000000..5558627bd5eb --- /dev/null +++ b/drivers/net/ethernet/intel/ice/ice_flow.h @@ -0,0 +1,207 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright (c) 2019, Intel Corporation. */ + +#ifndef _ICE_FLOW_H_ +#define _ICE_FLOW_H_ + +#define ICE_FLOW_ENTRY_HANDLE_INVAL 0 +#define ICE_FLOW_FLD_OFF_INVAL 0xffff + +/* Generate flow hash field from flow field type(s) */ +#define ICE_FLOW_HASH_IPV4 \ + (BIT_ULL(ICE_FLOW_FIELD_IDX_IPV4_SA) | \ + BIT_ULL(ICE_FLOW_FIELD_IDX_IPV4_DA)) +#define ICE_FLOW_HASH_IPV6 \ + (BIT_ULL(ICE_FLOW_FIELD_IDX_IPV6_SA) | \ + BIT_ULL(ICE_FLOW_FIELD_IDX_IPV6_DA)) +#define ICE_FLOW_HASH_TCP_PORT \ + (BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_SRC_PORT) | \ + BIT_ULL(ICE_FLOW_FIELD_IDX_TCP_DST_PORT)) +#define ICE_FLOW_HASH_UDP_PORT \ + (BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_SRC_PORT) | \ + BIT_ULL(ICE_FLOW_FIELD_IDX_UDP_DST_PORT)) +#define ICE_FLOW_HASH_SCTP_PORT \ + (BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT) | \ + BIT_ULL(ICE_FLOW_FIELD_IDX_SCTP_DST_PORT)) + +#define ICE_HASH_INVALID 0 +#define ICE_HASH_TCP_IPV4 (ICE_FLOW_HASH_IPV4 | ICE_FLOW_HASH_TCP_PORT) +#define ICE_HASH_TCP_IPV6 (ICE_FLOW_HASH_IPV6 | ICE_FLOW_HASH_TCP_PORT) +#define ICE_HASH_UDP_IPV4 (ICE_FLOW_HASH_IPV4 | ICE_FLOW_HASH_UDP_PORT) +#define ICE_HASH_UDP_IPV6 (ICE_FLOW_HASH_IPV6 | ICE_FLOW_HASH_UDP_PORT) + +/* Protocol header fields within a packet segment. A segment consists of one or + * more protocol headers that make up a logical group of protocol headers. Each + * logical group of protocol headers encapsulates or is encapsulated using/by + * tunneling or encapsulation protocols for network virtualization such as GRE, + * VxLAN, etc. + */ +enum ice_flow_seg_hdr { + ICE_FLOW_SEG_HDR_NONE = 0x00000000, + ICE_FLOW_SEG_HDR_IPV4 = 0x00000004, + ICE_FLOW_SEG_HDR_IPV6 = 0x00000008, + ICE_FLOW_SEG_HDR_TCP = 0x00000040, + ICE_FLOW_SEG_HDR_UDP = 0x00000080, + ICE_FLOW_SEG_HDR_SCTP = 0x00000100, +}; + +enum ice_flow_field { + /* L3 */ + ICE_FLOW_FIELD_IDX_IPV4_SA, + ICE_FLOW_FIELD_IDX_IPV4_DA, + ICE_FLOW_FIELD_IDX_IPV6_SA, + ICE_FLOW_FIELD_IDX_IPV6_DA, + /* L4 */ + ICE_FLOW_FIELD_IDX_TCP_SRC_PORT, + ICE_FLOW_FIELD_IDX_TCP_DST_PORT, + ICE_FLOW_FIELD_IDX_UDP_SRC_PORT, + ICE_FLOW_FIELD_IDX_UDP_DST_PORT, + ICE_FLOW_FIELD_IDX_SCTP_SRC_PORT, + ICE_FLOW_FIELD_IDX_SCTP_DST_PORT, + /* The total number of enums must not exceed 64 */ + ICE_FLOW_FIELD_IDX_MAX +}; + +/* Flow headers and fields for AVF support */ +enum ice_flow_avf_hdr_field { + /* Values 0 - 28 are reserved for future use */ + ICE_AVF_FLOW_FIELD_INVALID = 0, + ICE_AVF_FLOW_FIELD_UNICAST_IPV4_UDP = 29, + ICE_AVF_FLOW_FIELD_MULTICAST_IPV4_UDP, + ICE_AVF_FLOW_FIELD_IPV4_UDP, + ICE_AVF_FLOW_FIELD_IPV4_TCP_SYN_NO_ACK, + ICE_AVF_FLOW_FIELD_IPV4_TCP, + ICE_AVF_FLOW_FIELD_IPV4_SCTP, + ICE_AVF_FLOW_FIELD_IPV4_OTHER, + ICE_AVF_FLOW_FIELD_FRAG_IPV4, + /* Values 37-38 are reserved */ + ICE_AVF_FLOW_FIELD_UNICAST_IPV6_UDP = 39, + ICE_AVF_FLOW_FIELD_MULTICAST_IPV6_UDP, + ICE_AVF_FLOW_FIELD_IPV6_UDP, + ICE_AVF_FLOW_FIELD_IPV6_TCP_SYN_NO_ACK, + ICE_AVF_FLOW_FIELD_IPV6_TCP, + ICE_AVF_FLOW_FIELD_IPV6_SCTP, + ICE_AVF_FLOW_FIELD_IPV6_OTHER, + ICE_AVF_FLOW_FIELD_FRAG_IPV6, + ICE_AVF_FLOW_FIELD_RSVD47, + ICE_AVF_FLOW_FIELD_FCOE_OX, + ICE_AVF_FLOW_FIELD_FCOE_RX, + ICE_AVF_FLOW_FIELD_FCOE_OTHER, + /* Values 51-62 are reserved */ + ICE_AVF_FLOW_FIELD_L2_PAYLOAD = 63, + ICE_AVF_FLOW_FIELD_MAX +}; + +/* Supported RSS offloads This macro is defined to support + * VIRTCHNL_OP_GET_RSS_HENA_CAPS ops. PF driver sends the RSS hardware + * capabilities to the caller of this ops. + */ +#define ICE_DEFAULT_RSS_HENA ( \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_SCTP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_TCP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_OTHER) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_FRAG_IPV4) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_TCP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_SCTP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_OTHER) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_FRAG_IPV6) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV4_TCP_SYN_NO_ACK) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_UNICAST_IPV4_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_MULTICAST_IPV4_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_IPV6_TCP_SYN_NO_ACK) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_UNICAST_IPV6_UDP) | \ + BIT_ULL(ICE_AVF_FLOW_FIELD_MULTICAST_IPV6_UDP)) + +enum ice_flow_dir { + ICE_FLOW_RX = 0x02, +}; + +enum ice_flow_priority { + ICE_FLOW_PRIO_LOW, + ICE_FLOW_PRIO_NORMAL, + ICE_FLOW_PRIO_HIGH +}; + +#define ICE_FLOW_SEG_MAX 2 +#define ICE_FLOW_FV_EXTRACT_SZ 2 + +#define ICE_FLOW_SET_HDRS(seg, val) ((seg)->hdrs |= (u32)(val)) + +struct ice_flow_seg_xtrct { + u8 prot_id; /* Protocol ID of extracted header field */ + u16 off; /* Starting offset of the field in header in bytes */ + u8 idx; /* Index of FV entry used */ + u8 disp; /* Displacement of field in bits fr. FV entry's start */ +}; + +enum ice_flow_fld_match_type { + ICE_FLOW_FLD_TYPE_REG, /* Value, mask */ + ICE_FLOW_FLD_TYPE_RANGE, /* Value, mask, last (upper bound) */ + ICE_FLOW_FLD_TYPE_PREFIX, /* IP address, prefix, size of prefix */ + ICE_FLOW_FLD_TYPE_SIZE, /* Value, mask, size of match */ +}; + +struct ice_flow_fld_loc { + /* Describe offsets of field information relative to the beginning of + * input buffer provided when adding flow entries. + */ + u16 val; /* Offset where the value is located */ + u16 mask; /* Offset where the mask/prefix value is located */ + u16 last; /* Length or offset where the upper value is located */ +}; + +struct ice_flow_fld_info { + enum ice_flow_fld_match_type type; + /* Location where to retrieve data from an input buffer */ + struct ice_flow_fld_loc src; + /* Location where to put the data into the final entry buffer */ + struct ice_flow_fld_loc entry; + struct ice_flow_seg_xtrct xtrct; +}; + +struct ice_flow_seg_info { + u32 hdrs; /* Bitmask indicating protocol headers present */ + u64 match; /* Bitmask indicating header fields to be matched */ + u64 range; /* Bitmask indicating header fields matched as ranges */ + + struct ice_flow_fld_info fields[ICE_FLOW_FIELD_IDX_MAX]; +}; + +struct ice_flow_prof { + struct list_head l_entry; + + u64 id; + enum ice_flow_dir dir; + u8 segs_cnt; + + /* Keep track of flow entries associated with this flow profile */ + struct mutex entries_lock; + struct list_head entries; + + struct ice_flow_seg_info segs[ICE_FLOW_SEG_MAX]; + + /* software VSI handles referenced by this flow profile */ + DECLARE_BITMAP(vsis, ICE_MAX_VSI); +}; + +struct ice_rss_cfg { + struct list_head l_entry; + /* bitmap of VSIs added to the RSS entry */ + DECLARE_BITMAP(vsis, ICE_MAX_VSI); + u64 hashed_flds; + u32 packet_hdr; +}; + +enum ice_status ice_flow_rem_entry(struct ice_hw *hw, u64 entry_h); +void ice_rem_vsi_rss_list(struct ice_hw *hw, u16 vsi_handle); +enum ice_status ice_replay_rss_cfg(struct ice_hw *hw, u16 vsi_handle); +enum ice_status +ice_add_avf_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u64 hashed_flds); +enum ice_status ice_rem_vsi_rss_cfg(struct ice_hw *hw, u16 vsi_handle); +enum ice_status +ice_add_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u64 hashed_flds, + u32 addl_hdrs); +u64 ice_get_rss_cfg(struct ice_hw *hw, u16 vsi_handle, u32 hdrs); +#endif /* _ICE_FLOW_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h index e8f32350fed2..f2cababf2561 100644 --- a/drivers/net/ethernet/intel/ice/ice_hw_autogen.h +++ b/drivers/net/ethernet/intel/ice/ice_hw_autogen.h @@ -60,15 +60,6 @@ #define PRTDCB_GENS_DCBX_STATUS_M ICE_M(0x7, 0) #define GL_PREEXT_L2_PMASK0(_i) (0x0020F0FC + ((_i) * 4)) #define GL_PREEXT_L2_PMASK1(_i) (0x0020F108 + ((_i) * 4)) -#define GLFLXP_RXDID_FLAGS(_i, _j) (0x0045D000 + ((_i) * 4 + (_j) * 256)) -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_S 0 -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_M ICE_M(0x3F, 0) -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_1_S 8 -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_1_M ICE_M(0x3F, 8) -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_2_S 16 -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_2_M ICE_M(0x3F, 16) -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_3_S 24 -#define GLFLXP_RXDID_FLAGS_FLEXIFLAG_4N_3_M ICE_M(0x3F, 24) #define GLFLXP_RXDID_FLX_WRD_0(_i) (0x0045c800 + ((_i) * 4)) #define GLFLXP_RXDID_FLX_WRD_0_PROT_MDID_S 0 #define GLFLXP_RXDID_FLX_WRD_0_PROT_MDID_M ICE_M(0xFF, 0) diff --git a/drivers/net/ethernet/intel/ice/ice_lan_tx_rx.h b/drivers/net/ethernet/intel/ice/ice_lan_tx_rx.h index 0997d352709b..878e125d8b42 100644 --- a/drivers/net/ethernet/intel/ice/ice_lan_tx_rx.h +++ b/drivers/net/ethernet/intel/ice/ice_lan_tx_rx.h @@ -199,6 +199,14 @@ enum ice_rxdid { /* Receive Flex Descriptor Rx opcode values */ #define ICE_RX_OPC_MDID 0x01 +/* Receive Descriptor MDID values that access packet flags */ +enum ice_flex_mdid_pkt_flags { + ICE_RX_MDID_PKT_FLAGS_15_0 = 20, + ICE_RX_MDID_PKT_FLAGS_31_16, + ICE_RX_MDID_PKT_FLAGS_47_32, + ICE_RX_MDID_PKT_FLAGS_63_48, +}; + /* Receive Descriptor MDID values */ enum ice_flex_rx_mdid { ICE_RX_MDID_FLOW_ID_LOWER = 5, diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c index e7449248fab4..1874c9f51a32 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.c +++ b/drivers/net/ethernet/intel/ice/ice_lib.c @@ -3,6 +3,7 @@ #include "ice.h" #include "ice_base.h" +#include "ice_flow.h" #include "ice_lib.h" #include "ice_dcb_lib.h" @@ -493,7 +494,28 @@ bool ice_is_safe_mode(struct ice_pf *pf) } /** - * ice_rss_clean - Delete RSS related VSI structures that hold user inputs + * ice_vsi_clean_rss_flow_fld - Delete RSS configuration + * @vsi: the VSI being cleaned up + * + * This function deletes RSS input set for all flows that were configured + * for this VSI + */ +static void ice_vsi_clean_rss_flow_fld(struct ice_vsi *vsi) +{ + struct ice_pf *pf = vsi->back; + enum ice_status status; + + if (ice_is_safe_mode(pf)) + return; + + status = ice_rem_vsi_rss_cfg(&pf->hw, vsi->idx); + if (status) + dev_dbg(ice_pf_to_dev(pf), "ice_rem_vsi_rss_cfg failed for vsi = %d, error = %d\n", + vsi->vsi_num, status); +} + +/** + * ice_rss_clean - Delete RSS related VSI structures and configuration * @vsi: the VSI being removed */ static void ice_rss_clean(struct ice_vsi *vsi) @@ -507,6 +529,11 @@ static void ice_rss_clean(struct ice_vsi *vsi) devm_kfree(dev, vsi->rss_hkey_user); if (vsi->rss_lut_user) devm_kfree(dev, vsi->rss_lut_user); + + ice_vsi_clean_rss_flow_fld(vsi); + /* remove RSS replay list */ + if (!ice_is_safe_mode(pf)) + ice_rem_vsi_rss_list(&pf->hw, vsi->idx); } /** @@ -817,12 +844,23 @@ static int ice_vsi_init(struct ice_vsi *vsi, bool init_vsi) ctxt->info.valid_sections |= cpu_to_le16(ICE_AQ_VSI_PROP_RXQ_MAP_VALID); - /* Enable MAC Antispoof with new VSI being initialized or updated */ - if (vsi->type == ICE_VSI_VF && pf->vf[vsi->vf_id].spoofchk) { + /* enable/disable MAC and VLAN anti-spoof when spoofchk is on/off + * respectively + */ + if (vsi->type == ICE_VSI_VF) { ctxt->info.valid_sections |= cpu_to_le16(ICE_AQ_VSI_PROP_SECURITY_VALID); - ctxt->info.sec_flags |= - ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF; + if (pf->vf[vsi->vf_id].spoofchk) { + ctxt->info.sec_flags |= + ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF | + (ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << + ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S); + } else { + ctxt->info.sec_flags &= + ~(ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF | + (ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << + ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S)); + } } /* Allow control frames out of main VSI */ @@ -1076,6 +1114,115 @@ ice_vsi_cfg_rss_exit: } /** + * ice_vsi_set_vf_rss_flow_fld - Sets VF VSI RSS input set for different flows + * @vsi: VSI to be configured + * + * This function will only be called during the VF VSI setup. Upon successful + * completion of package download, this function will configure default RSS + * input sets for VF VSI. + */ +static void ice_vsi_set_vf_rss_flow_fld(struct ice_vsi *vsi) +{ + struct ice_pf *pf = vsi->back; + enum ice_status status; + struct device *dev; + + dev = ice_pf_to_dev(pf); + if (ice_is_safe_mode(pf)) { + dev_dbg(dev, "Advanced RSS disabled. Package download failed, vsi num = %d\n", + vsi->vsi_num); + return; + } + + status = ice_add_avf_rss_cfg(&pf->hw, vsi->idx, ICE_DEFAULT_RSS_HENA); + if (status) + dev_dbg(dev, "ice_add_avf_rss_cfg failed for vsi = %d, error = %d\n", + vsi->vsi_num, status); +} + +/** + * ice_vsi_set_rss_flow_fld - Sets RSS input set for different flows + * @vsi: VSI to be configured + * + * This function will only be called after successful download package call + * during initialization of PF. Since the downloaded package will erase the + * RSS section, this function will configure RSS input sets for different + * flow types. The last profile added has the highest priority, therefore 2 + * tuple profiles (i.e. IPv4 src/dst) are added before 4 tuple profiles + * (i.e. IPv4 src/dst TCP src/dst port). + */ +static void ice_vsi_set_rss_flow_fld(struct ice_vsi *vsi) +{ + u16 vsi_handle = vsi->idx, vsi_num = vsi->vsi_num; + struct ice_pf *pf = vsi->back; + struct ice_hw *hw = &pf->hw; + enum ice_status status; + struct device *dev; + + dev = ice_pf_to_dev(pf); + if (ice_is_safe_mode(pf)) { + dev_dbg(dev, "Advanced RSS disabled. Package download failed, vsi num = %d\n", + vsi_num); + return; + } + /* configure RSS for IPv4 with input set IP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_FLOW_HASH_IPV4, + ICE_FLOW_SEG_HDR_IPV4); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for ipv4 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for IPv6 with input set IPv6 src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_FLOW_HASH_IPV6, + ICE_FLOW_SEG_HDR_IPV6); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for ipv6 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for tcp4 with input set IP src/dst, TCP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_HASH_TCP_IPV4, + ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_IPV4); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for tcp4 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for udp4 with input set IP src/dst, UDP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_HASH_UDP_IPV4, + ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_IPV4); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for udp4 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for sctp4 with input set IP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_FLOW_HASH_IPV4, + ICE_FLOW_SEG_HDR_SCTP | ICE_FLOW_SEG_HDR_IPV4); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for sctp4 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for tcp6 with input set IPv6 src/dst, TCP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_HASH_TCP_IPV6, + ICE_FLOW_SEG_HDR_TCP | ICE_FLOW_SEG_HDR_IPV6); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for tcp6 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for udp6 with input set IPv6 src/dst, UDP src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_HASH_UDP_IPV6, + ICE_FLOW_SEG_HDR_UDP | ICE_FLOW_SEG_HDR_IPV6); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for udp6 flow, vsi = %d, error = %d\n", + vsi_num, status); + + /* configure RSS for sctp6 with input set IPv6 src/dst */ + status = ice_add_rss_cfg(hw, vsi_handle, ICE_FLOW_HASH_IPV6, + ICE_FLOW_SEG_HDR_SCTP | ICE_FLOW_SEG_HDR_IPV6); + if (status) + dev_dbg(dev, "ice_add_rss_cfg failed for sctp6 flow, vsi = %d, error = %d\n", + vsi_num, status); +} + +/** * ice_add_mac_to_list - Add a MAC address filter entry to the list * @vsi: the VSI to be forwarded to * @add_list: pointer to the list which contains MAC filter entries @@ -1636,22 +1783,14 @@ int ice_cfg_vlan_pruning(struct ice_vsi *vsi, bool ena, bool vlan_promisc) ctxt->info = vsi->info; - if (ena) { - ctxt->info.sec_flags |= - ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << - ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S; + if (ena) ctxt->info.sw_flags2 |= ICE_AQ_VSI_SW_FLAG_RX_VLAN_PRUNE_ENA; - } else { - ctxt->info.sec_flags &= - ~(ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << - ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S); + else ctxt->info.sw_flags2 &= ~ICE_AQ_VSI_SW_FLAG_RX_VLAN_PRUNE_ENA; - } if (!vlan_promisc) ctxt->info.valid_sections = - cpu_to_le16(ICE_AQ_VSI_PROP_SECURITY_VALID | - ICE_AQ_VSI_PROP_SW_VALID); + cpu_to_le16(ICE_AQ_VSI_PROP_SW_VALID); status = ice_update_vsi(&pf->hw, vsi->idx, ctxt, NULL); if (status) { @@ -1661,7 +1800,6 @@ int ice_cfg_vlan_pruning(struct ice_vsi *vsi, bool ena, bool vlan_promisc) goto err_out; } - vsi->info.sec_flags = ctxt->info.sec_flags; vsi->info.sw_flags2 = ctxt->info.sw_flags2; kfree(ctxt); @@ -1899,8 +2037,10 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi, * receive traffic on first queue. Hence no need to capture * return value */ - if (test_bit(ICE_FLAG_RSS_ENA, pf->flags)) + if (test_bit(ICE_FLAG_RSS_ENA, pf->flags)) { ice_vsi_cfg_rss_lut_key(vsi); + ice_vsi_set_rss_flow_fld(vsi); + } break; case ICE_VSI_VF: /* VF driver will take care of creating netdev for this type and @@ -1924,8 +2064,10 @@ ice_vsi_setup(struct ice_pf *pf, struct ice_port_info *pi, * receive traffic on first queue. Hence no need to capture * return value */ - if (test_bit(ICE_FLAG_RSS_ENA, pf->flags)) + if (test_bit(ICE_FLAG_RSS_ENA, pf->flags)) { ice_vsi_cfg_rss_lut_key(vsi); + ice_vsi_set_vf_rss_flow_fld(vsi); + } break; case ICE_VSI_LB: ret = ice_vsi_alloc_rings(vsi); @@ -2402,6 +2544,97 @@ int ice_vsi_release(struct ice_vsi *vsi) } /** + * ice_vsi_rebuild_update_coalesce - set coalesce for a q_vector + * @q_vector: pointer to q_vector which is being updated + * @coalesce: pointer to array of struct with stored coalesce + * + * Set coalesce param in q_vector and update these parameters in HW. + */ +static void +ice_vsi_rebuild_update_coalesce(struct ice_q_vector *q_vector, + struct ice_coalesce_stored *coalesce) +{ + struct ice_ring_container *rx_rc = &q_vector->rx; + struct ice_ring_container *tx_rc = &q_vector->tx; + struct ice_hw *hw = &q_vector->vsi->back->hw; + + tx_rc->itr_setting = coalesce->itr_tx; + rx_rc->itr_setting = coalesce->itr_rx; + + /* dynamic ITR values will be updated during Tx/Rx */ + if (!ITR_IS_DYNAMIC(tx_rc->itr_setting)) + wr32(hw, GLINT_ITR(tx_rc->itr_idx, q_vector->reg_idx), + ITR_REG_ALIGN(tx_rc->itr_setting) >> + ICE_ITR_GRAN_S); + if (!ITR_IS_DYNAMIC(rx_rc->itr_setting)) + wr32(hw, GLINT_ITR(rx_rc->itr_idx, q_vector->reg_idx), + ITR_REG_ALIGN(rx_rc->itr_setting) >> + ICE_ITR_GRAN_S); + + q_vector->intrl = coalesce->intrl; + wr32(hw, GLINT_RATE(q_vector->reg_idx), + ice_intrl_usec_to_reg(q_vector->intrl, hw->intrl_gran)); +} + +/** + * ice_vsi_rebuild_get_coalesce - get coalesce from all q_vectors + * @vsi: VSI connected with q_vectors + * @coalesce: array of struct with stored coalesce + * + * Returns array size. + */ +static int +ice_vsi_rebuild_get_coalesce(struct ice_vsi *vsi, + struct ice_coalesce_stored *coalesce) +{ + int i; + + ice_for_each_q_vector(vsi, i) { + struct ice_q_vector *q_vector = vsi->q_vectors[i]; + + coalesce[i].itr_tx = q_vector->tx.itr_setting; + coalesce[i].itr_rx = q_vector->rx.itr_setting; + coalesce[i].intrl = q_vector->intrl; + } + + return vsi->num_q_vectors; +} + +/** + * ice_vsi_rebuild_set_coalesce - set coalesce from earlier saved arrays + * @vsi: VSI connected with q_vectors + * @coalesce: pointer to array of struct with stored coalesce + * @size: size of coalesce array + * + * Before this function, ice_vsi_rebuild_get_coalesce should be called to save + * ITR params in arrays. If size is 0 or coalesce wasn't stored set coalesce + * to default value. + */ +static void +ice_vsi_rebuild_set_coalesce(struct ice_vsi *vsi, + struct ice_coalesce_stored *coalesce, int size) +{ + int i; + + if ((size && !coalesce) || !vsi) + return; + + for (i = 0; i < size && i < vsi->num_q_vectors; i++) + ice_vsi_rebuild_update_coalesce(vsi->q_vectors[i], + &coalesce[i]); + + for (; i < vsi->num_q_vectors; i++) { + struct ice_coalesce_stored coalesce_dflt = { + .itr_tx = ICE_DFLT_TX_ITR, + .itr_rx = ICE_DFLT_RX_ITR, + .intrl = 0 + }; + ice_vsi_rebuild_update_coalesce(vsi->q_vectors[i], + &coalesce_dflt); + } +} + +/** * ice_vsi_rebuild - Rebuild VSI after reset * @vsi: VSI to be rebuild * @init_vsi: is this an initialization or a reconfigure of the VSI @@ -2411,6 +2644,8 @@ int ice_vsi_release(struct ice_vsi *vsi) int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi) { u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 }; + struct ice_coalesce_stored *coalesce; + int prev_num_q_vectors = 0; struct ice_vf *vf = NULL; enum ice_status status; struct ice_pf *pf; @@ -2423,6 +2658,11 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi) if (vsi->type == ICE_VSI_VF) vf = &pf->vf[vsi->vf_id]; + coalesce = kcalloc(vsi->num_q_vectors, + sizeof(struct ice_coalesce_stored), GFP_KERNEL); + if (coalesce) + prev_num_q_vectors = ice_vsi_rebuild_get_coalesce(vsi, + coalesce); ice_rm_vsi_lan_cfg(vsi->port_info, vsi->idx); ice_vsi_free_q_vectors(vsi); @@ -2535,6 +2775,9 @@ int ice_vsi_rebuild(struct ice_vsi *vsi, bool init_vsi) return ice_schedule_reset(pf, ICE_RESET_PFR); } } + ice_vsi_rebuild_set_coalesce(vsi, coalesce, prev_num_q_vectors); + kfree(coalesce); + return 0; err_vectors: @@ -2549,6 +2792,7 @@ err_rings: err_vsi: ice_vsi_clear(vsi); set_bit(__ICE_RESET_FAILED, pf->state); + kfree(coalesce); return ret; } @@ -2740,3 +2984,121 @@ cfg_mac_fltr_exit: ice_free_fltr_list(&vsi->back->pdev->dev, &tmp_add_list); return status; } + +/** + * ice_is_dflt_vsi_in_use - check if the default forwarding VSI is being used + * @sw: switch to check if its default forwarding VSI is free + * + * Return true if the default forwarding VSI is already being used, else returns + * false signalling that it's available to use. + */ +bool ice_is_dflt_vsi_in_use(struct ice_sw *sw) +{ + return (sw->dflt_vsi && sw->dflt_vsi_ena); +} + +/** + * ice_is_vsi_dflt_vsi - check if the VSI passed in is the default VSI + * @sw: switch for the default forwarding VSI to compare against + * @vsi: VSI to compare against default forwarding VSI + * + * If this VSI passed in is the default forwarding VSI then return true, else + * return false + */ +bool ice_is_vsi_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi) +{ + return (sw->dflt_vsi == vsi && sw->dflt_vsi_ena); +} + +/** + * ice_set_dflt_vsi - set the default forwarding VSI + * @sw: switch used to assign the default forwarding VSI + * @vsi: VSI getting set as the default forwarding VSI on the switch + * + * If the VSI passed in is already the default VSI and it's enabled just return + * success. + * + * If there is already a default VSI on the switch and it's enabled then return + * -EEXIST since there can only be one default VSI per switch. + * + * Otherwise try to set the VSI passed in as the switch's default VSI and + * return the result. + */ +int ice_set_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi) +{ + enum ice_status status; + struct device *dev; + + if (!sw || !vsi) + return -EINVAL; + + dev = ice_pf_to_dev(vsi->back); + + /* the VSI passed in is already the default VSI */ + if (ice_is_vsi_dflt_vsi(sw, vsi)) { + dev_dbg(dev, "VSI %d passed in is already the default forwarding VSI, nothing to do\n", + vsi->vsi_num); + return 0; + } + + /* another VSI is already the default VSI for this switch */ + if (ice_is_dflt_vsi_in_use(sw)) { + dev_err(dev, + "Default forwarding VSI %d already in use, disable it and try again\n", + sw->dflt_vsi->vsi_num); + return -EEXIST; + } + + status = ice_cfg_dflt_vsi(&vsi->back->hw, vsi->idx, true, ICE_FLTR_RX); + if (status) { + dev_err(dev, + "Failed to set VSI %d as the default forwarding VSI, error %d\n", + vsi->vsi_num, status); + return -EIO; + } + + sw->dflt_vsi = vsi; + sw->dflt_vsi_ena = true; + + return 0; +} + +/** + * ice_clear_dflt_vsi - clear the default forwarding VSI + * @sw: switch used to clear the default VSI + * + * If the switch has no default VSI or it's not enabled then return error. + * + * Otherwise try to clear the default VSI and return the result. + */ +int ice_clear_dflt_vsi(struct ice_sw *sw) +{ + struct ice_vsi *dflt_vsi; + enum ice_status status; + struct device *dev; + + if (!sw) + return -EINVAL; + + dev = ice_pf_to_dev(sw->pf); + + dflt_vsi = sw->dflt_vsi; + + /* there is no default VSI configured */ + if (!ice_is_dflt_vsi_in_use(sw)) + return -ENODEV; + + status = ice_cfg_dflt_vsi(&dflt_vsi->back->hw, dflt_vsi->idx, false, + ICE_FLTR_RX); + if (status) { + dev_err(dev, + "Failed to clear the default forwarding VSI %d, error %d\n", + dflt_vsi->vsi_num, status); + return -EIO; + } + + sw->dflt_vsi = NULL; + sw->dflt_vsi_ena = false; + + return 0; +} diff --git a/drivers/net/ethernet/intel/ice/ice_lib.h b/drivers/net/ethernet/intel/ice/ice_lib.h index 6e31e30aba39..68fd0d4505c2 100644 --- a/drivers/net/ethernet/intel/ice/ice_lib.h +++ b/drivers/net/ethernet/intel/ice/ice_lib.h @@ -103,4 +103,12 @@ enum ice_status ice_vsi_cfg_mac_fltr(struct ice_vsi *vsi, const u8 *macaddr, bool set); bool ice_is_safe_mode(struct ice_pf *pf); + +bool ice_is_dflt_vsi_in_use(struct ice_sw *sw); + +bool ice_is_vsi_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi); + +int ice_set_dflt_vsi(struct ice_sw *sw, struct ice_vsi *vsi); + +int ice_clear_dflt_vsi(struct ice_sw *sw); #endif /* !_ICE_LIB_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c index 69bff085acf7..5ae671609f98 100644 --- a/drivers/net/ethernet/intel/ice/ice_main.c +++ b/drivers/net/ethernet/intel/ice/ice_main.c @@ -13,7 +13,7 @@ #define DRV_VERSION_MAJOR 0 #define DRV_VERSION_MINOR 8 -#define DRV_VERSION_BUILD 1 +#define DRV_VERSION_BUILD 2 #define DRV_VERSION __stringify(DRV_VERSION_MAJOR) "." \ __stringify(DRV_VERSION_MINOR) "." \ @@ -379,25 +379,29 @@ static int ice_vsi_sync_fltr(struct ice_vsi *vsi) clear_bit(ICE_VSI_FLAG_PROMISC_CHANGED, vsi->flags); if (vsi->current_netdev_flags & IFF_PROMISC) { /* Apply Rx filter rule to get traffic from wire */ - status = ice_cfg_dflt_vsi(hw, vsi->idx, true, - ICE_FLTR_RX); - if (status) { - netdev_err(netdev, "Error setting default VSI %i Rx rule\n", - vsi->vsi_num); - vsi->current_netdev_flags &= ~IFF_PROMISC; - err = -EIO; - goto out_promisc; + if (!ice_is_dflt_vsi_in_use(pf->first_sw)) { + err = ice_set_dflt_vsi(pf->first_sw, vsi); + if (err && err != -EEXIST) { + netdev_err(netdev, + "Error %d setting default VSI %i Rx rule\n", + err, vsi->vsi_num); + vsi->current_netdev_flags &= + ~IFF_PROMISC; + goto out_promisc; + } } } else { /* Clear Rx filter to remove traffic from wire */ - status = ice_cfg_dflt_vsi(hw, vsi->idx, false, - ICE_FLTR_RX); - if (status) { - netdev_err(netdev, "Error clearing default VSI %i Rx rule\n", - vsi->vsi_num); - vsi->current_netdev_flags |= IFF_PROMISC; - err = -EIO; - goto out_promisc; + if (ice_is_vsi_dflt_vsi(pf->first_sw, vsi)) { + err = ice_clear_dflt_vsi(pf->first_sw); + if (err) { + netdev_err(netdev, + "Error %d clearing default VSI %i Rx rule\n", + err, vsi->vsi_num); + vsi->current_netdev_flags |= + IFF_PROMISC; + goto out_promisc; + } } } } @@ -472,7 +476,7 @@ ice_prepare_for_reset(struct ice_pf *pf) ice_vc_notify_reset(pf); /* Disable VFs until reset is completed */ - for (i = 0; i < pf->num_alloc_vfs; i++) + ice_for_each_vf(pf, i) ice_set_vf_state_qs_dis(&pf->vf[i]); /* clear SW filtering DB */ @@ -840,8 +844,7 @@ ice_link_event(struct ice_pf *pf, struct ice_port_info *pi, bool link_up, ice_vsi_link_event(vsi, link_up); ice_print_link_msg(vsi, link_up); - if (pf->num_alloc_vfs) - ice_vc_notify_link_state(pf); + ice_vc_notify_link_state(pf); return result; } @@ -1291,7 +1294,7 @@ static void ice_handle_mdd_event(struct ice_pf *pf) } /* check to see if one of the VFs caused the MDD */ - for (i = 0; i < pf->num_alloc_vfs; i++) { + ice_for_each_vf(pf, i) { struct ice_vf *vf = &pf->vf[i]; bool vf_mdd_detected = false; @@ -2330,7 +2333,8 @@ static void ice_set_netdev_features(struct net_device *netdev) NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_CTAG_RX; - tso_features = NETIF_F_TSO; + tso_features = NETIF_F_TSO | + NETIF_F_GSO_UDP_L4; /* set features that user can change */ netdev->hw_features = dflt_features | csumo_features | @@ -3568,6 +3572,15 @@ static const struct pci_device_id ice_pci_tbl[] = { { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_BACKPLANE), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_QSFP), 0 }, { PCI_VDEVICE(INTEL, ICE_DEV_ID_E810C_SFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822C_BACKPLANE), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822C_QSFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822C_SFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822C_10G_BASE_T), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822C_SGMII), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822X_BACKPLANE), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822L_SFP), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822L_10G_BASE_T), 0 }, + { PCI_VDEVICE(INTEL, ICE_DEV_ID_E822L_SGMII), 0 }, /* required last entry */ { 0, } }; @@ -4670,6 +4683,13 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) goto err_init_ctrlq; } + if (pf->first_sw->dflt_vsi_ena) + dev_info(dev, + "Clearing default VSI, re-enable after reset completes\n"); + /* clear the default VSI configuration if it exists */ + pf->first_sw->dflt_vsi = NULL; + pf->first_sw->dflt_vsi_ena = false; + ice_clear_pxe_mode(hw); ret = ice_get_caps(hw); @@ -4825,7 +4845,7 @@ static int ice_change_mtu(struct net_device *netdev, int new_mtu) } } - netdev_info(netdev, "changed MTU to %d\n", new_mtu); + netdev_dbg(netdev, "changed MTU to %d\n", new_mtu); return 0; } @@ -5060,42 +5080,23 @@ ice_bridge_setlink(struct net_device *dev, struct nlmsghdr *nlh, * ice_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure */ -static void ice_tx_timeout(struct net_device *netdev) +static void ice_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ice_netdev_priv *np = netdev_priv(netdev); struct ice_ring *tx_ring = NULL; struct ice_vsi *vsi = np->vsi; struct ice_pf *pf = vsi->back; - int hung_queue = -1; u32 i; pf->tx_timeout_count++; - /* find the stopped queue the same way dev_watchdog() does */ - for (i = 0; i < netdev->num_tx_queues; i++) { - unsigned long trans_start; - struct netdev_queue *q; - - q = netdev_get_tx_queue(netdev, i); - trans_start = q->trans_start; - if (netif_xmit_stopped(q) && - time_after(jiffies, - trans_start + netdev->watchdog_timeo)) { - hung_queue = i; - break; - } - } - - if (i == netdev->num_tx_queues) - netdev_info(netdev, "tx_timeout: no netdev hung queue found\n"); - else - /* now that we have an index, find the tx_ring struct */ - for (i = 0; i < vsi->num_txq; i++) - if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) - if (hung_queue == vsi->tx_rings[i]->q_index) { - tx_ring = vsi->tx_rings[i]; - break; - } + /* now that we have an index, find the tx_ring struct */ + for (i = 0; i < vsi->num_txq; i++) + if (vsi->tx_rings[i] && vsi->tx_rings[i]->desc) + if (txqueue == vsi->tx_rings[i]->q_index) { + tx_ring = vsi->tx_rings[i]; + break; + } /* Reset recovery level if enough time has elapsed after last timeout. * Also ensure no new reset action happens before next timeout period. @@ -5110,19 +5111,19 @@ static void ice_tx_timeout(struct net_device *netdev) struct ice_hw *hw = &pf->hw; u32 head, val = 0; - head = (rd32(hw, QTX_COMM_HEAD(vsi->txq_map[hung_queue])) & + head = (rd32(hw, QTX_COMM_HEAD(vsi->txq_map[txqueue])) & QTX_COMM_HEAD_HEAD_M) >> QTX_COMM_HEAD_HEAD_S; /* Read interrupt register */ val = rd32(hw, GLINT_DYN_CTL(tx_ring->q_vector->reg_idx)); netdev_info(netdev, "tx_timeout: VSI_num: %d, Q %d, NTC: 0x%x, HW_HEAD: 0x%x, NTU: 0x%x, INT: 0x%x\n", - vsi->vsi_num, hung_queue, tx_ring->next_to_clean, + vsi->vsi_num, txqueue, tx_ring->next_to_clean, head, tx_ring->next_to_use, val); } pf->tx_timeout_last_recovery = jiffies; - netdev_info(netdev, "tx_timeout recovery level %d, hung_queue %d\n", - pf->tx_timeout_recovery_level, hung_queue); + netdev_info(netdev, "tx_timeout recovery level %d, txqueue %d\n", + pf->tx_timeout_recovery_level, txqueue); switch (pf->tx_timeout_recovery_level) { case 1: diff --git a/drivers/net/ethernet/intel/ice/ice_nvm.c b/drivers/net/ethernet/intel/ice/ice_nvm.c index 57c73f613f32..7525ac50742e 100644 --- a/drivers/net/ethernet/intel/ice/ice_nvm.c +++ b/drivers/net/ethernet/intel/ice/ice_nvm.c @@ -289,6 +289,18 @@ enum ice_status ice_init_nvm(struct ice_hw *hw) nvm->eetrack = (eetrack_hi << 16) | eetrack_lo; + /* the following devices do not have boot_cfg_tlv yet */ + if (hw->device_id == ICE_DEV_ID_E822C_BACKPLANE || + hw->device_id == ICE_DEV_ID_E822C_QSFP || + hw->device_id == ICE_DEV_ID_E822C_10G_BASE_T || + hw->device_id == ICE_DEV_ID_E822C_SGMII || + hw->device_id == ICE_DEV_ID_E822C_SFP || + hw->device_id == ICE_DEV_ID_E822X_BACKPLANE || + hw->device_id == ICE_DEV_ID_E822L_SFP || + hw->device_id == ICE_DEV_ID_E822L_10G_BASE_T || + hw->device_id == ICE_DEV_ID_E822L_SGMII) + return status; + status = ice_get_pfa_module_tlv(hw, &boot_cfg_tlv, &boot_cfg_tlv_len, ICE_SR_BOOT_CFG_PTR); if (status) { diff --git a/drivers/net/ethernet/intel/ice/ice_protocol_type.h b/drivers/net/ethernet/intel/ice/ice_protocol_type.h new file mode 100644 index 000000000000..71647566964e --- /dev/null +++ b/drivers/net/ethernet/intel/ice/ice_protocol_type.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Copyright (c) 2019, Intel Corporation. */ + +#ifndef _ICE_PROTOCOL_TYPE_H_ +#define _ICE_PROTOCOL_TYPE_H_ +/* Decoders for ice_prot_id: + * - F: First + * - I: Inner + * - L: Last + * - O: Outer + * - S: Single + */ +enum ice_prot_id { + ICE_PROT_ID_INVAL = 0, + ICE_PROT_IPV4_OF_OR_S = 32, + ICE_PROT_IPV4_IL = 33, + ICE_PROT_IPV6_OF_OR_S = 40, + ICE_PROT_IPV6_IL = 41, + ICE_PROT_TCP_IL = 49, + ICE_PROT_UDP_IL_OR_S = 53, + ICE_PROT_SCTP_IL = 96, + ICE_PROT_META_ID = 255, /* when offset == metadata */ + ICE_PROT_INVALID = 255 /* when offset == ICE_FV_OFFSET_INVAL */ +}; +#endif /* _ICE_PROTOCOL_TYPE_H_ */ diff --git a/drivers/net/ethernet/intel/ice/ice_status.h b/drivers/net/ethernet/intel/ice/ice_status.h index c01597885629..a9a8bc3aca42 100644 --- a/drivers/net/ethernet/intel/ice/ice_status.h +++ b/drivers/net/ethernet/intel/ice/ice_status.h @@ -26,6 +26,7 @@ enum ice_status { ICE_ERR_IN_USE = -16, ICE_ERR_MAX_LIMIT = -17, ICE_ERR_RESET_ONGOING = -18, + ICE_ERR_HW_TABLE = -19, ICE_ERR_NVM_CHECKSUM = -51, ICE_ERR_BUF_TOO_SHORT = -52, ICE_ERR_NVM_BLANK_MODE = -53, diff --git a/drivers/net/ethernet/intel/ice/ice_switch.c b/drivers/net/ethernet/intel/ice/ice_switch.c index b5a53f862a83..431266081a80 100644 --- a/drivers/net/ethernet/intel/ice/ice_switch.c +++ b/drivers/net/ethernet/intel/ice/ice_switch.c @@ -50,42 +50,6 @@ static const u8 dummy_eth_header[DUMMY_ETH_HDR_LEN] = { 0x2, 0, 0, 0, 0, 0, ((n) * sizeof(((struct ice_sw_rule_vsi_list *)0)->vsi))) /** - * ice_aq_alloc_free_res - command to allocate/free resources - * @hw: pointer to the HW struct - * @num_entries: number of resource entries in buffer - * @buf: Indirect buffer to hold data parameters and response - * @buf_size: size of buffer for indirect commands - * @opc: pass in the command opcode - * @cd: pointer to command details structure or NULL - * - * Helper function to allocate/free resources using the admin queue commands - */ -static enum ice_status -ice_aq_alloc_free_res(struct ice_hw *hw, u16 num_entries, - struct ice_aqc_alloc_free_res_elem *buf, u16 buf_size, - enum ice_adminq_opc opc, struct ice_sq_cd *cd) -{ - struct ice_aqc_alloc_free_res_cmd *cmd; - struct ice_aq_desc desc; - - cmd = &desc.params.sw_res_ctrl; - - if (!buf) - return ICE_ERR_PARAM; - - if (buf_size < (num_entries * sizeof(buf->elem[0]))) - return ICE_ERR_PARAM; - - ice_fill_dflt_direct_cmd_desc(&desc, opc); - - desc.flags |= cpu_to_le16(ICE_AQ_FLAG_RD); - - cmd->num_entries = cpu_to_le16(num_entries); - - return ice_aq_send_cmd(hw, &desc, buf, buf_size, cd); -} - -/** * ice_init_def_sw_recp - initialize the recipe book keeping tables * @hw: pointer to the HW struct * diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c index 2c212f64d99f..fd17ace6b226 100644 --- a/drivers/net/ethernet/intel/ice/ice_txrx.c +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c @@ -1071,13 +1071,16 @@ static int ice_clean_rx_irq(struct ice_ring *rx_ring, int budget) ice_put_rx_buf(rx_ring, rx_buf); continue; construct_skb: - if (skb) + if (skb) { ice_add_rx_frag(rx_ring, rx_buf, skb, size); - else if (ice_ring_uses_build_skb(rx_ring)) - skb = ice_build_skb(rx_ring, rx_buf, &xdp); - else + } else if (likely(xdp.data)) { + if (ice_ring_uses_build_skb(rx_ring)) + skb = ice_build_skb(rx_ring, rx_buf, &xdp); + else + skb = ice_construct_skb(rx_ring, rx_buf, &xdp); + } else { skb = ice_construct_skb(rx_ring, rx_buf, &xdp); - + } /* exit if we failed to retrieve a buffer */ if (!skb) { rx_ring->rx_stats.alloc_buf_failed++; @@ -1925,6 +1928,7 @@ int ice_tso(struct ice_tx_buf *first, struct ice_tx_offload_params *off) } ip; union { struct tcphdr *tcp; + struct udphdr *udp; unsigned char *hdr; } l4; u64 cd_mss, cd_tso_len; @@ -1958,10 +1962,18 @@ int ice_tso(struct ice_tx_buf *first, struct ice_tx_offload_params *off) /* remove payload length from checksum */ paylen = skb->len - l4_start; - csum_replace_by_diff(&l4.tcp->check, (__force __wsum)htonl(paylen)); - /* compute length of segmentation header */ - off->header_len = (l4.tcp->doff * 4) + l4_start; + if (skb_shinfo(skb)->gso_type & SKB_GSO_UDP_L4) { + csum_replace_by_diff(&l4.udp->check, + (__force __wsum)htonl(paylen)); + /* compute length of UDP segmentation header */ + off->header_len = sizeof(l4.udp) + l4_start; + } else { + csum_replace_by_diff(&l4.tcp->check, + (__force __wsum)htonl(paylen)); + /* compute length of TCP segmentation header */ + off->header_len = (l4.tcp->doff * 4) + l4_start; + } /* update gso_segs and bytecount */ first->gso_segs = skb_shinfo(skb)->gso_segs; diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.h b/drivers/net/ethernet/intel/ice/ice_txrx.h index a84cc0e6dd27..a86270696df1 100644 --- a/drivers/net/ethernet/intel/ice/ice_txrx.h +++ b/drivers/net/ethernet/intel/ice/ice_txrx.h @@ -341,6 +341,12 @@ struct ice_ring_container { u16 itr_setting; }; +struct ice_coalesce_stored { + u16 itr_tx; + u16 itr_rx; + u8 intrl; +}; + /* iterator for handling rings in ring container */ #define ice_for_each_ring(pos, head) \ for (pos = (head).ring; pos; pos = pos->next) diff --git a/drivers/net/ethernet/intel/ice/ice_type.h b/drivers/net/ethernet/intel/ice/ice_type.h index c4854a987130..b361ffabb0ca 100644 --- a/drivers/net/ethernet/intel/ice/ice_type.h +++ b/drivers/net/ethernet/intel/ice/ice_type.h @@ -13,6 +13,7 @@ #include "ice_controlq.h" #include "ice_lan_tx_rx.h" #include "ice_flex_type.h" +#include "ice_protocol_type.h" static inline bool ice_is_tc_ena(unsigned long bitmap, u8 tc) { @@ -41,6 +42,7 @@ static inline u32 ice_round_to_num(u32 N, u32 R) #define ICE_DBG_QCTX BIT_ULL(6) #define ICE_DBG_NVM BIT_ULL(7) #define ICE_DBG_LAN BIT_ULL(8) +#define ICE_DBG_FLOW BIT_ULL(9) #define ICE_DBG_SW BIT_ULL(13) #define ICE_DBG_SCHED BIT_ULL(14) #define ICE_DBG_PKG BIT_ULL(16) @@ -559,6 +561,10 @@ struct ice_hw { /* HW block tables */ struct ice_blk_info blk[ICE_BLK_COUNT]; + struct mutex fl_profs_locks[ICE_BLK_COUNT]; /* lock fltr profiles */ + struct list_head fl_profs[ICE_BLK_COUNT]; + struct mutex rss_locks; /* protect RSS configuration */ + struct list_head rss_list_head; }; /* Statistics collected by each port, VSI, VEB, and S-channel */ diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c index edb374296d1f..82b1e7a4cb92 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.c @@ -35,37 +35,6 @@ static int ice_check_vf_init(struct ice_pf *pf, struct ice_vf *vf) } /** - * ice_err_to_virt err - translate errors for VF return code - * @ice_err: error return code - */ -static enum virtchnl_status_code ice_err_to_virt_err(enum ice_status ice_err) -{ - switch (ice_err) { - case ICE_SUCCESS: - return VIRTCHNL_STATUS_SUCCESS; - case ICE_ERR_BAD_PTR: - case ICE_ERR_INVAL_SIZE: - case ICE_ERR_DEVICE_NOT_SUPPORTED: - case ICE_ERR_PARAM: - case ICE_ERR_CFG: - return VIRTCHNL_STATUS_ERR_PARAM; - case ICE_ERR_NO_MEMORY: - return VIRTCHNL_STATUS_ERR_NO_MEMORY; - case ICE_ERR_NOT_READY: - case ICE_ERR_RESET_FAILED: - case ICE_ERR_FW_API_VER: - case ICE_ERR_AQ_ERROR: - case ICE_ERR_AQ_TIMEOUT: - case ICE_ERR_AQ_FULL: - case ICE_ERR_AQ_NO_WORK: - case ICE_ERR_AQ_EMPTY: - return VIRTCHNL_STATUS_ERR_ADMIN_QUEUE_ERROR; - default: - return VIRTCHNL_STATUS_ERR_NOT_SUPPORTED; - } -} - -/** * ice_vc_vf_broadcast - Broadcast a message to all VFs on PF * @pf: pointer to the PF structure * @v_opcode: operation code @@ -78,10 +47,11 @@ ice_vc_vf_broadcast(struct ice_pf *pf, enum virtchnl_ops v_opcode, enum virtchnl_status_code v_retval, u8 *msg, u16 msglen) { struct ice_hw *hw = &pf->hw; - struct ice_vf *vf = pf->vf; int i; - for (i = 0; i < pf->num_alloc_vfs; i++, vf++) { + ice_for_each_vf(pf, i) { + struct ice_vf *vf = &pf->vf[i]; + /* Not all vfs are enabled so skip the ones that are not */ if (!test_bit(ICE_VF_STATE_INIT, vf->vf_states) && !test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states)) @@ -121,26 +91,6 @@ ice_set_pfe_link(struct ice_vf *vf, struct virtchnl_pf_event *pfe, } /** - * ice_set_pfe_link_forced - Force the virtchnl_pf_event link speed/status - * @vf: pointer to the VF structure - * @pfe: pointer to the virtchnl_pf_event to set link speed/status for - * @link_up: whether or not to set the link up/down - */ -static void -ice_set_pfe_link_forced(struct ice_vf *vf, struct virtchnl_pf_event *pfe, - bool link_up) -{ - u16 link_speed; - - if (link_up) - link_speed = ICE_AQ_LINK_SPEED_100GB; - else - link_speed = ICE_AQ_LINK_SPEED_UNKNOWN; - - ice_set_pfe_link(vf, pfe, link_speed, link_up); -} - -/** * ice_vc_notify_vf_link_state - Inform a VF of link status * @vf: pointer to the VF structure * @@ -160,13 +110,17 @@ static void ice_vc_notify_vf_link_state(struct ice_vf *vf) pfe.severity = PF_EVENT_SEVERITY_INFO; /* Always report link is down if the VF queues aren't enabled */ - if (!vf->num_qs_ena) + if (!vf->num_qs_ena) { ice_set_pfe_link(vf, &pfe, ICE_AQ_LINK_SPEED_UNKNOWN, false); - else if (vf->link_forced) - ice_set_pfe_link_forced(vf, &pfe, vf->link_up); - else - ice_set_pfe_link(vf, &pfe, ls->link_speed, ls->link_info & - ICE_AQ_LINK_UP); + } else if (vf->link_forced) { + u16 link_speed = vf->link_up ? + ls->link_speed : ICE_AQ_LINK_SPEED_UNKNOWN; + + ice_set_pfe_link(vf, &pfe, link_speed, vf->link_up); + } else { + ice_set_pfe_link(vf, &pfe, ls->link_speed, + ls->link_info & ICE_AQ_LINK_UP); + } ice_aq_send_msg_to_vf(hw, vf->vf_id, VIRTCHNL_OP_EVENT, VIRTCHNL_STATUS_SUCCESS, (u8 *)&pfe, @@ -331,7 +285,7 @@ void ice_free_vfs(struct ice_pf *pf) usleep_range(1000, 2000); /* Avoid wait time by stopping all VFs at the same time */ - for (i = 0; i < pf->num_alloc_vfs; i++) + ice_for_each_vf(pf, i) if (test_bit(ICE_VF_STATE_QS_ENA, pf->vf[i].vf_states)) ice_dis_vf_qs(&pf->vf[i]); @@ -991,10 +945,17 @@ static void ice_cleanup_and_realloc_vf(struct ice_vf *vf) /* reallocate VF resources to finish resetting the VSI state */ if (!ice_alloc_vf_res(vf)) { + struct ice_vsi *vsi; + ice_ena_vf_mappings(vf); set_bit(ICE_VF_STATE_ACTIVE, vf->vf_states); clear_bit(ICE_VF_STATE_DIS, vf->vf_states); - vf->num_vlan = 0; + + vsi = pf->vsi[vf->lan_vsi_idx]; + if (ice_vsi_add_vlan(vsi, 0)) + dev_warn(ice_pf_to_dev(pf), + "Failed to add VLAN 0 filter for VF %d, MDD events will trigger. Reset the VF, disable spoofchk, or enable 8021q module on the guest", + vf->vf_id); } /* Tell the VF driver the reset is done. This needs to be done only @@ -1023,7 +984,7 @@ ice_vf_set_vsi_promisc(struct ice_vf *vf, struct ice_vsi *vsi, u8 promisc_m, struct ice_hw *hw; hw = &pf->hw; - if (vf->num_vlan) { + if (vsi->num_vlan) { status = ice_set_vlan_vsi_promisc(hw, vsi->idx, promisc_m, rm_promisc); } else if (vf->port_vlan_id) { @@ -1070,7 +1031,7 @@ static bool ice_config_res_vfs(struct ice_pf *pf) ice_irq_dynamic_ena(hw, NULL, NULL); /* Finish resetting each VF and allocate resources */ - for (v = 0; v < pf->num_alloc_vfs; v++) { + ice_for_each_vf(pf, v) { struct ice_vf *vf = &pf->vf[v]; vf->num_vf_qs = pf->num_vf_qps; @@ -1113,10 +1074,10 @@ bool ice_reset_all_vfs(struct ice_pf *pf, bool is_vflr) return false; /* Begin reset on all VFs at once */ - for (v = 0; v < pf->num_alloc_vfs; v++) + ice_for_each_vf(pf, v) ice_trigger_vf_reset(&pf->vf[v], is_vflr, true); - for (v = 0; v < pf->num_alloc_vfs; v++) { + ice_for_each_vf(pf, v) { struct ice_vsi *vsi; vf = &pf->vf[v]; @@ -1161,7 +1122,7 @@ bool ice_reset_all_vfs(struct ice_pf *pf, bool is_vflr) dev_warn(dev, "VF reset check timeout\n"); /* free VF resources to begin resetting the VSI state */ - for (v = 0; v < pf->num_alloc_vfs; v++) { + ice_for_each_vf(pf, v) { vf = &pf->vf[v]; ice_free_vf_res(vf); @@ -1273,7 +1234,7 @@ static bool ice_reset_vf(struct ice_vf *vf, bool is_vflr) */ if (test_bit(ICE_VF_STATE_UC_PROMISC, vf->vf_states) || test_bit(ICE_VF_STATE_MC_PROMISC, vf->vf_states)) { - if (vf->port_vlan_id || vf->num_vlan) + if (vf->port_vlan_id || vsi->num_vlan) promisc_m = ICE_UCAST_VLAN_PROMISC_BITS; else promisc_m = ICE_UCAST_PROMISC_BITS; @@ -1301,7 +1262,7 @@ void ice_vc_notify_link_state(struct ice_pf *pf) { int i; - for (i = 0; i < pf->num_alloc_vfs; i++) + ice_for_each_vf(pf, i) ice_vc_notify_vf_link_state(&pf->vf[i]); } @@ -1385,9 +1346,10 @@ static int ice_alloc_vfs(struct ice_pf *pf, u16 num_alloc_vfs) goto err_pci_disable_sriov; } pf->vf = vfs; + pf->num_alloc_vfs = num_alloc_vfs; /* apply default profile */ - for (i = 0; i < num_alloc_vfs; i++) { + ice_for_each_vf(pf, i) { vfs[i].pf = pf; vfs[i].vf_sw_id = pf->first_sw; vfs[i].vf_id = i; @@ -1396,7 +1358,6 @@ static int ice_alloc_vfs(struct ice_pf *pf, u16 num_alloc_vfs) set_bit(ICE_VIRTCHNL_VF_CAP_L2, &vfs[i].vf_caps); vfs[i].spoofchk = true; } - pf->num_alloc_vfs = num_alloc_vfs; /* VF resources get allocated with initialization */ if (!ice_config_res_vfs(pf)) { @@ -1535,7 +1496,7 @@ void ice_process_vflr_event(struct ice_pf *pf) !pf->num_alloc_vfs) return; - for (vf_id = 0; vf_id < pf->num_alloc_vfs; vf_id++) { + ice_for_each_vf(pf, vf_id) { struct ice_vf *vf = &pf->vf[vf_id]; u32 reg_idx, bit_idx; @@ -1918,6 +1879,89 @@ error_param: } /** + * ice_set_vf_spoofchk + * @netdev: network interface device structure + * @vf_id: VF identifier + * @ena: flag to enable or disable feature + * + * Enable or disable VF spoof checking + */ +int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena) +{ + struct ice_netdev_priv *np = netdev_priv(netdev); + struct ice_pf *pf = np->vsi->back; + struct ice_vsi_ctx *ctx; + struct ice_vsi *vf_vsi; + enum ice_status status; + struct device *dev; + struct ice_vf *vf; + int ret = 0; + + dev = ice_pf_to_dev(pf); + if (ice_validate_vf_id(pf, vf_id)) + return -EINVAL; + + vf = &pf->vf[vf_id]; + + if (ice_check_vf_init(pf, vf)) + return -EBUSY; + + vf_vsi = pf->vsi[vf->lan_vsi_idx]; + if (!vf_vsi) { + netdev_err(netdev, "VSI %d for VF %d is null\n", + vf->lan_vsi_idx, vf->vf_id); + return -EINVAL; + } + + if (vf_vsi->type != ICE_VSI_VF) { + netdev_err(netdev, + "Type %d of VSI %d for VF %d is no ICE_VSI_VF\n", + vf_vsi->type, vf_vsi->vsi_num, vf->vf_id); + return -ENODEV; + } + + if (ena == vf->spoofchk) { + dev_dbg(dev, "VF spoofchk already %s\n", ena ? "ON" : "OFF"); + return 0; + } + + ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); + if (!ctx) + return -ENOMEM; + + ctx->info.sec_flags = vf_vsi->info.sec_flags; + ctx->info.valid_sections = cpu_to_le16(ICE_AQ_VSI_PROP_SECURITY_VALID); + if (ena) { + ctx->info.sec_flags |= + ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF | + (ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << + ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S); + } else { + ctx->info.sec_flags &= + ~(ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF | + (ICE_AQ_VSI_SEC_TX_VLAN_PRUNE_ENA << + ICE_AQ_VSI_SEC_TX_PRUNE_ENA_S)); + } + + status = ice_update_vsi(&pf->hw, vf_vsi->idx, ctx, NULL); + if (status) { + dev_err(dev, + "Failed to %sable spoofchk on VF %d VSI %d\n error %d", + ena ? "en" : "dis", vf->vf_id, vf_vsi->vsi_num, status); + ret = -EIO; + goto out; + } + + /* only update spoofchk state and VSI context on success */ + vf_vsi->info.sec_flags = ctx->info.sec_flags; + vf->spoofchk = ena; + +out: + kfree(ctx); + return ret; +} + +/** * ice_vc_get_stats_msg * @vf: pointer to the VF info * @msg: pointer to the msg buffer @@ -2409,6 +2453,83 @@ static bool ice_can_vf_change_mac(struct ice_vf *vf) } /** + * ice_vc_add_mac_addr - attempt to add the MAC address passed in + * @vf: pointer to the VF info + * @vsi: pointer to the VF's VSI + * @mac_addr: MAC address to add + */ +static int +ice_vc_add_mac_addr(struct ice_vf *vf, struct ice_vsi *vsi, u8 *mac_addr) +{ + struct device *dev = ice_pf_to_dev(vf->pf); + enum ice_status status; + + /* default unicast MAC already added */ + if (ether_addr_equal(mac_addr, vf->dflt_lan_addr.addr)) + return 0; + + if (is_unicast_ether_addr(mac_addr) && !ice_can_vf_change_mac(vf)) { + dev_err(dev, "VF attempting to override administratively set MAC address, bring down and up the VF interface to resume normal operation\n"); + return -EPERM; + } + + status = ice_vsi_cfg_mac_fltr(vsi, mac_addr, true); + if (status == ICE_ERR_ALREADY_EXISTS) { + dev_err(dev, "MAC %pM already exists for VF %d\n", mac_addr, + vf->vf_id); + return -EEXIST; + } else if (status) { + dev_err(dev, "Failed to add MAC %pM for VF %d\n, error %d\n", + mac_addr, vf->vf_id, status); + return -EIO; + } + + /* only set dflt_lan_addr once */ + if (is_zero_ether_addr(vf->dflt_lan_addr.addr) && + is_unicast_ether_addr(mac_addr)) + ether_addr_copy(vf->dflt_lan_addr.addr, mac_addr); + + vf->num_mac++; + + return 0; +} + +/** + * ice_vc_del_mac_addr - attempt to delete the MAC address passed in + * @vf: pointer to the VF info + * @vsi: pointer to the VF's VSI + * @mac_addr: MAC address to delete + */ +static int +ice_vc_del_mac_addr(struct ice_vf *vf, struct ice_vsi *vsi, u8 *mac_addr) +{ + struct device *dev = ice_pf_to_dev(vf->pf); + enum ice_status status; + + if (!ice_can_vf_change_mac(vf) && + ether_addr_equal(mac_addr, vf->dflt_lan_addr.addr)) + return 0; + + status = ice_vsi_cfg_mac_fltr(vsi, mac_addr, false); + if (status == ICE_ERR_DOES_NOT_EXIST) { + dev_err(dev, "MAC %pM does not exist for VF %d\n", mac_addr, + vf->vf_id); + return -ENOENT; + } else if (status) { + dev_err(dev, "Failed to delete MAC %pM for VF %d, error %d\n", + mac_addr, vf->vf_id, status); + return -EIO; + } + + if (ether_addr_equal(mac_addr, vf->dflt_lan_addr.addr)) + eth_zero_addr(vf->dflt_lan_addr.addr); + + vf->num_mac--; + + return 0; +} + +/** * ice_vc_handle_mac_addr_msg * @vf: pointer to the VF info * @msg: pointer to the msg buffer @@ -2419,23 +2540,23 @@ static bool ice_can_vf_change_mac(struct ice_vf *vf) static int ice_vc_handle_mac_addr_msg(struct ice_vf *vf, u8 *msg, bool set) { + int (*ice_vc_cfg_mac) + (struct ice_vf *vf, struct ice_vsi *vsi, u8 *mac_addr); enum virtchnl_status_code v_ret = VIRTCHNL_STATUS_SUCCESS; struct virtchnl_ether_addr_list *al = (struct virtchnl_ether_addr_list *)msg; struct ice_pf *pf = vf->pf; enum virtchnl_ops vc_op; - enum ice_status status; struct ice_vsi *vsi; - struct device *dev; - int mac_count = 0; int i; - dev = ice_pf_to_dev(pf); - - if (set) + if (set) { vc_op = VIRTCHNL_OP_ADD_ETH_ADDR; - else + ice_vc_cfg_mac = ice_vc_add_mac_addr; + } else { vc_op = VIRTCHNL_OP_DEL_ETH_ADDR; + ice_vc_cfg_mac = ice_vc_del_mac_addr; + } if (!test_bit(ICE_VF_STATE_ACTIVE, vf->vf_states) || !ice_vc_isvalid_vsi_id(vf, al->vsi_id)) { @@ -2443,14 +2564,15 @@ ice_vc_handle_mac_addr_msg(struct ice_vf *vf, u8 *msg, bool set) goto handle_mac_exit; } + /* If this VF is not privileged, then we can't add more than a + * limited number of addresses. Check to make sure that the + * additions do not push us over the limit. + */ if (set && !ice_is_vf_trusted(vf) && (vf->num_mac + al->num_elements) > ICE_MAX_MACADDR_PER_VF) { - dev_err(dev, + dev_err(ice_pf_to_dev(pf), "Can't add more MAC addresses, because VF-%d is not trusted, switch the VF to trusted mode in order to add more functionalities\n", vf->vf_id); - /* There is no need to let VF know about not being trusted - * to add more MAC addr, so we can just return success message. - */ v_ret = VIRTCHNL_STATUS_ERR_PARAM; goto handle_mac_exit; } @@ -2462,70 +2584,22 @@ ice_vc_handle_mac_addr_msg(struct ice_vf *vf, u8 *msg, bool set) } for (i = 0; i < al->num_elements; i++) { - u8 *maddr = al->list[i].addr; + u8 *mac_addr = al->list[i].addr; + int result; - if (ether_addr_equal(maddr, vf->dflt_lan_addr.addr) || - is_broadcast_ether_addr(maddr)) { - if (set) { - /* VF is trying to add filters that the PF - * already added. Just continue. - */ - dev_info(dev, - "MAC %pM already set for VF %d\n", - maddr, vf->vf_id); - continue; - } else { - /* VF can't remove dflt_lan_addr/bcast MAC */ - dev_err(dev, - "VF can't remove default MAC address or MAC %pM programmed by PF for VF %d\n", - maddr, vf->vf_id); - continue; - } - } - - /* check for the invalid cases and bail if necessary */ - if (is_zero_ether_addr(maddr)) { - dev_err(dev, - "invalid MAC %pM provided for VF %d\n", - maddr, vf->vf_id); - v_ret = VIRTCHNL_STATUS_ERR_PARAM; - goto handle_mac_exit; - } - - if (is_unicast_ether_addr(maddr) && - !ice_can_vf_change_mac(vf)) { - dev_err(dev, - "can't change unicast MAC for untrusted VF %d\n", - vf->vf_id); - v_ret = VIRTCHNL_STATUS_ERR_PARAM; - goto handle_mac_exit; - } + if (is_broadcast_ether_addr(mac_addr) || + is_zero_ether_addr(mac_addr)) + continue; - /* program the updated filter list */ - status = ice_vsi_cfg_mac_fltr(vsi, maddr, set); - if (status == ICE_ERR_DOES_NOT_EXIST || - status == ICE_ERR_ALREADY_EXISTS) { - dev_info(dev, - "can't %s MAC filters %pM for VF %d, error %d\n", - set ? "add" : "remove", maddr, vf->vf_id, - status); - } else if (status) { - dev_err(dev, - "can't %s MAC filters for VF %d, error %d\n", - set ? "add" : "remove", vf->vf_id, status); - v_ret = ice_err_to_virt_err(status); + result = ice_vc_cfg_mac(vf, vsi, mac_addr); + if (result == -EEXIST || result == -ENOENT) { + continue; + } else if (result) { + v_ret = VIRTCHNL_STATUS_ERR_ADMIN_QUEUE_ERROR; goto handle_mac_exit; } - - mac_count++; } - /* Track number of MAC filters programmed for the VF VSI */ - if (set) - vf->num_mac += mac_count; - else - vf->num_mac -= mac_count; - handle_mac_exit: /* send the response to the VF */ return ice_vc_send_msg_to_vf(vf, vc_op, v_ret, NULL, 0); @@ -2744,17 +2818,6 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) goto error_param; } - if (add_v && !ice_is_vf_trusted(vf) && - vf->num_vlan >= ICE_MAX_VLAN_PER_VF) { - dev_info(dev, - "VF-%d is not trusted, switch the VF to trusted mode, in order to add more VLAN addresses\n", - vf->vf_id); - /* There is no need to let VF know about being not trusted, - * so we can just return success message here - */ - goto error_param; - } - for (i = 0; i < vfl->num_elements; i++) { if (vfl->vlan_id[i] > ICE_MAX_VLANID) { v_ret = VIRTCHNL_STATUS_ERR_PARAM; @@ -2771,6 +2834,17 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) goto error_param; } + if (add_v && !ice_is_vf_trusted(vf) && + vsi->num_vlan >= ICE_MAX_VLAN_PER_VF) { + dev_info(dev, + "VF-%d is not trusted, switch the VF to trusted mode, in order to add more VLAN addresses\n", + vf->vf_id); + /* There is no need to let VF know about being not trusted, + * so we can just return success message here + */ + goto error_param; + } + if (vsi->info.pvid) { v_ret = VIRTCHNL_STATUS_ERR_PARAM; goto error_param; @@ -2785,7 +2859,7 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) u16 vid = vfl->vlan_id[i]; if (!ice_is_vf_trusted(vf) && - vf->num_vlan >= ICE_MAX_VLAN_PER_VF) { + vsi->num_vlan >= ICE_MAX_VLAN_PER_VF) { dev_info(dev, "VF-%d is not trusted, switch the VF to trusted mode, in order to add more VLAN addresses\n", vf->vf_id); @@ -2796,12 +2870,20 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) goto error_param; } - if (ice_vsi_add_vlan(vsi, vid)) { + /* we add VLAN 0 by default for each VF so we can enable + * Tx VLAN anti-spoof without triggering MDD events so + * we don't need to add it again here + */ + if (!vid) + continue; + + status = ice_vsi_add_vlan(vsi, vid); + if (status) { v_ret = VIRTCHNL_STATUS_ERR_PARAM; goto error_param; } - vf->num_vlan++; + vsi->num_vlan++; /* Enable VLAN pruning when VLAN is added */ if (!vlan_promisc) { status = ice_cfg_vlan_pruning(vsi, true, false); @@ -2837,21 +2919,29 @@ static int ice_vc_process_vlan_msg(struct ice_vf *vf, u8 *msg, bool add_v) */ int num_vf_vlan; - num_vf_vlan = vf->num_vlan; + num_vf_vlan = vsi->num_vlan; for (i = 0; i < vfl->num_elements && i < num_vf_vlan; i++) { u16 vid = vfl->vlan_id[i]; + /* we add VLAN 0 by default for each VF so we can enable + * Tx VLAN anti-spoof without triggering MDD events so + * we don't want a VIRTCHNL request to remove it + */ + if (!vid) + continue; + /* Make sure ice_vsi_kill_vlan is successful before * updating VLAN information */ - if (ice_vsi_kill_vlan(vsi, vid)) { + status = ice_vsi_kill_vlan(vsi, vid); + if (status) { v_ret = VIRTCHNL_STATUS_ERR_PARAM; goto error_param; } - vf->num_vlan--; + vsi->num_vlan--; /* Disable VLAN pruning when the last VLAN is removed */ - if (!vf->num_vlan) + if (!vsi->num_vlan) ice_cfg_vlan_pruning(vsi, false, false); /* Disable Unicast/Multicast VLAN promiscuous mode */ @@ -3165,65 +3255,6 @@ ice_get_vf_cfg(struct net_device *netdev, int vf_id, struct ifla_vf_info *ivi) } /** - * ice_set_vf_spoofchk - * @netdev: network interface device structure - * @vf_id: VF identifier - * @ena: flag to enable or disable feature - * - * Enable or disable VF spoof checking - */ -int ice_set_vf_spoofchk(struct net_device *netdev, int vf_id, bool ena) -{ - struct ice_pf *pf = ice_netdev_to_pf(netdev); - struct ice_vsi *vsi = pf->vsi[0]; - struct ice_vsi_ctx *ctx; - enum ice_status status; - struct device *dev; - struct ice_vf *vf; - int ret = 0; - - dev = ice_pf_to_dev(pf); - if (ice_validate_vf_id(pf, vf_id)) - return -EINVAL; - - vf = &pf->vf[vf_id]; - if (ice_check_vf_init(pf, vf)) - return -EBUSY; - - if (ena == vf->spoofchk) { - dev_dbg(dev, "VF spoofchk already %s\n", - ena ? "ON" : "OFF"); - return 0; - } - - ctx = kzalloc(sizeof(*ctx), GFP_KERNEL); - if (!ctx) - return -ENOMEM; - - ctx->info.valid_sections = cpu_to_le16(ICE_AQ_VSI_PROP_SECURITY_VALID); - - if (ena) { - ctx->info.sec_flags |= ICE_AQ_VSI_SEC_FLAG_ENA_MAC_ANTI_SPOOF; - ctx->info.sw_flags2 |= ICE_AQ_VSI_SW_FLAG_RX_PRUNE_EN_M; - } - - status = ice_update_vsi(&pf->hw, vsi->idx, ctx, NULL); - if (status) { - dev_dbg(dev, - "Error %d, failed to update VSI* parameters\n", status); - ret = -EIO; - goto out; - } - - vf->spoofchk = ena; - vsi->info.sec_flags = ctx->info.sec_flags; - vsi->info.sw_flags2 = ctx->info.sw_flags2; -out: - kfree(ctx); - return ret; -} - -/** * ice_wait_on_vf_reset * @vf: The VF being resseting * @@ -3344,28 +3375,18 @@ int ice_set_vf_trust(struct net_device *netdev, int vf_id, bool trusted) int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state) { struct ice_pf *pf = ice_netdev_to_pf(netdev); - struct virtchnl_pf_event pfe = { 0 }; - struct ice_link_status *ls; struct ice_vf *vf; - struct ice_hw *hw; if (ice_validate_vf_id(pf, vf_id)) return -EINVAL; vf = &pf->vf[vf_id]; - hw = &pf->hw; - ls = &pf->hw.port_info->phy.link_info; - if (ice_check_vf_init(pf, vf)) return -EBUSY; - pfe.event = VIRTCHNL_EVENT_LINK_CHANGE; - pfe.severity = PF_EVENT_SEVERITY_INFO; - switch (link_state) { case IFLA_VF_LINK_STATE_AUTO: vf->link_forced = false; - vf->link_up = ls->link_info & ICE_AQ_LINK_UP; break; case IFLA_VF_LINK_STATE_ENABLE: vf->link_forced = true; @@ -3379,15 +3400,7 @@ int ice_set_vf_link_state(struct net_device *netdev, int vf_id, int link_state) return -EINVAL; } - if (vf->link_forced) - ice_set_pfe_link_forced(vf, &pfe, vf->link_up); - else - ice_set_pfe_link(vf, &pfe, ls->link_speed, vf->link_up); - - /* Notify the VF of its new link state */ - ice_aq_send_msg_to_vf(hw, vf->vf_id, VIRTCHNL_OP_EVENT, - VIRTCHNL_STATUS_SUCCESS, (u8 *)&pfe, - sizeof(pfe), NULL); + ice_vc_notify_vf_link_state(vf); return 0; } diff --git a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h index 88aa65d5cb31..4647d636ed36 100644 --- a/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h +++ b/drivers/net/ethernet/intel/ice/ice_virtchnl_pf.h @@ -40,6 +40,9 @@ #define ICE_DFLT_INTR_PER_VF (ICE_DFLT_QS_PER_VF + 1) #define ICE_MAX_VF_RESET_WAIT 15 +#define ice_for_each_vf(pf, i) \ + for ((i) = 0; (i) < (pf)->num_alloc_vfs; (i)++) + /* Specific VF states */ enum ice_vf_states { ICE_VF_STATE_INIT = 0, /* PF is initializing VF */ @@ -91,7 +94,6 @@ struct ice_vf { unsigned long vf_caps; /* VF's adv. capabilities */ u8 num_req_qs; /* num of queue pairs requested by VF */ u16 num_mac; - u16 num_vlan; u16 num_vf_qs; /* num of queue configured per VF */ u16 num_qs_ena; /* total num of Tx/Rx queue enabled */ }; diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c index cf9b8b22d24f..149dca0012ba 100644 --- a/drivers/net/ethernet/intel/ice/ice_xsk.c +++ b/drivers/net/ethernet/intel/ice/ice_xsk.c @@ -414,7 +414,8 @@ ice_xsk_umem_enable(struct ice_vsi *vsi, struct xdp_umem *umem, u16 qid) if (vsi->type != ICE_VSI_PF) return -EINVAL; - vsi->num_xsk_umems = min_t(u16, vsi->num_rxq, vsi->num_txq); + if (!vsi->num_xsk_umems) + vsi->num_xsk_umems = min_t(u16, vsi->num_rxq, vsi->num_txq); if (qid >= vsi->num_xsk_umems) return -EINVAL; @@ -555,7 +556,7 @@ ice_alloc_buf_fast_zc(struct ice_ring *rx_ring, struct ice_rx_buf *rx_buf) rx_buf->handle = handle + umem->headroom; - xsk_umem_discard_addr(umem); + xsk_umem_release_addr(umem); return true; } @@ -591,7 +592,7 @@ ice_alloc_buf_slow_zc(struct ice_ring *rx_ring, struct ice_rx_buf *rx_buf) rx_buf->handle = handle + umem->headroom; - xsk_umem_discard_addr_rq(umem); + xsk_umem_release_addr_rq(umem); return true; } @@ -1019,8 +1020,8 @@ bool ice_clean_tx_irq_zc(struct ice_ring *xdp_ring, int budget) s16 ntc = xdp_ring->next_to_clean; struct ice_tx_desc *tx_desc; struct ice_tx_buf *tx_buf; - bool xmit_done = true; u32 xsk_frames = 0; + bool xmit_done; tx_desc = ICE_TX_DESC(xdp_ring, ntc); tx_buf = &xdp_ring->tx_buf[ntc]; diff --git a/drivers/net/ethernet/intel/igb/igb.h b/drivers/net/ethernet/intel/igb/igb.h index ca54e268d157..49b5fa9d4783 100644 --- a/drivers/net/ethernet/intel/igb/igb.h +++ b/drivers/net/ethernet/intel/igb/igb.h @@ -661,6 +661,7 @@ void igb_configure_tx_ring(struct igb_adapter *, struct igb_ring *); void igb_configure_rx_ring(struct igb_adapter *, struct igb_ring *); void igb_setup_tctl(struct igb_adapter *); void igb_setup_rctl(struct igb_adapter *); +void igb_setup_srrctl(struct igb_adapter *, struct igb_ring *); netdev_tx_t igb_xmit_frame_ring(struct sk_buff *, struct igb_ring *); void igb_alloc_rx_buffers(struct igb_ring *, u16); void igb_update_stats(struct igb_adapter *); diff --git a/drivers/net/ethernet/intel/igb/igb_ethtool.c b/drivers/net/ethernet/intel/igb/igb_ethtool.c index 445fbdce3e25..f96ffa83efbe 100644 --- a/drivers/net/ethernet/intel/igb/igb_ethtool.c +++ b/drivers/net/ethernet/intel/igb/igb_ethtool.c @@ -396,6 +396,7 @@ static int igb_set_pauseparam(struct net_device *netdev, struct igb_adapter *adapter = netdev_priv(netdev); struct e1000_hw *hw = &adapter->hw; int retval = 0; + int i; /* 100basefx does not support setting link flow control */ if (hw->dev_spec._82575.eth_flags.e100_base_fx) @@ -428,6 +429,13 @@ static int igb_set_pauseparam(struct net_device *netdev, retval = ((hw->phy.media_type == e1000_media_type_copper) ? igb_force_mac_fc(hw) : igb_setup_link(hw)); + + /* Make sure SRRCTL considers new fc settings for each ring */ + for (i = 0; i < adapter->num_rx_queues; i++) { + struct igb_ring *ring = adapter->rx_ring[i]; + + igb_setup_srrctl(adapter, ring); + } } clear_bit(__IGB_RESETTING, &adapter->state); diff --git a/drivers/net/ethernet/intel/igb/igb_main.c b/drivers/net/ethernet/intel/igb/igb_main.c index 98346eb064d5..b46bff8fe056 100644 --- a/drivers/net/ethernet/intel/igb/igb_main.c +++ b/drivers/net/ethernet/intel/igb/igb_main.c @@ -146,7 +146,7 @@ static int igb_poll(struct napi_struct *, int); static bool igb_clean_tx_irq(struct igb_q_vector *, int); static int igb_clean_rx_irq(struct igb_q_vector *, int); static int igb_ioctl(struct net_device *, struct ifreq *, int cmd); -static void igb_tx_timeout(struct net_device *); +static void igb_tx_timeout(struct net_device *, unsigned int txqueue); static void igb_reset_task(struct work_struct *); static void igb_vlan_mode(struct net_device *netdev, netdev_features_t features); @@ -4468,6 +4468,37 @@ static inline void igb_set_vmolr(struct igb_adapter *adapter, } /** + * igb_setup_srrctl - configure the split and replication receive control + * registers + * @adapter: Board private structure + * @ring: receive ring to be configured + **/ +void igb_setup_srrctl(struct igb_adapter *adapter, struct igb_ring *ring) +{ + struct e1000_hw *hw = &adapter->hw; + int reg_idx = ring->reg_idx; + u32 srrctl = 0; + + srrctl = IGB_RX_HDR_LEN << E1000_SRRCTL_BSIZEHDRSIZE_SHIFT; + if (ring_uses_large_buffer(ring)) + srrctl |= IGB_RXBUFFER_3072 >> E1000_SRRCTL_BSIZEPKT_SHIFT; + else + srrctl |= IGB_RXBUFFER_2048 >> E1000_SRRCTL_BSIZEPKT_SHIFT; + srrctl |= E1000_SRRCTL_DESCTYPE_ADV_ONEBUF; + if (hw->mac.type >= e1000_82580) + srrctl |= E1000_SRRCTL_TIMESTAMP; + /* Only set Drop Enable if VFs allocated, or we are supporting multiple + * queues and rx flow control is disabled + */ + if (adapter->vfs_allocated_count || + (!(hw->fc.current_mode & e1000_fc_rx_pause) && + adapter->num_rx_queues > 1)) + srrctl |= E1000_SRRCTL_DROP_EN; + + wr32(E1000_SRRCTL(reg_idx), srrctl); +} + +/** * igb_configure_rx_ring - Configure a receive ring after Reset * @adapter: board private structure * @ring: receive ring to be configured @@ -4481,7 +4512,7 @@ void igb_configure_rx_ring(struct igb_adapter *adapter, union e1000_adv_rx_desc *rx_desc; u64 rdba = ring->dma; int reg_idx = ring->reg_idx; - u32 srrctl = 0, rxdctl = 0; + u32 rxdctl = 0; /* disable the queue */ wr32(E1000_RXDCTL(reg_idx), 0); @@ -4499,19 +4530,7 @@ void igb_configure_rx_ring(struct igb_adapter *adapter, writel(0, ring->tail); /* set descriptor configuration */ - srrctl = IGB_RX_HDR_LEN << E1000_SRRCTL_BSIZEHDRSIZE_SHIFT; - if (ring_uses_large_buffer(ring)) - srrctl |= IGB_RXBUFFER_3072 >> E1000_SRRCTL_BSIZEPKT_SHIFT; - else - srrctl |= IGB_RXBUFFER_2048 >> E1000_SRRCTL_BSIZEPKT_SHIFT; - srrctl |= E1000_SRRCTL_DESCTYPE_ADV_ONEBUF; - if (hw->mac.type >= e1000_82580) - srrctl |= E1000_SRRCTL_TIMESTAMP; - /* Only set Drop Enable if we are supporting multiple queues */ - if (adapter->vfs_allocated_count || adapter->num_rx_queues > 1) - srrctl |= E1000_SRRCTL_DROP_EN; - - wr32(E1000_SRRCTL(reg_idx), srrctl); + igb_setup_srrctl(adapter, ring); /* set filtering for VMDQ pools */ igb_set_vmolr(adapter, reg_idx & 0x7, true); @@ -6184,7 +6203,7 @@ static netdev_tx_t igb_xmit_frame(struct sk_buff *skb, * igb_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void igb_tx_timeout(struct net_device *netdev) +static void igb_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct igb_adapter *adapter = netdev_priv(netdev); struct e1000_hw *hw = &adapter->hw; diff --git a/drivers/net/ethernet/intel/igbvf/netdev.c b/drivers/net/ethernet/intel/igbvf/netdev.c index 6003dc3ff5fd..5b1800c3ba82 100644 --- a/drivers/net/ethernet/intel/igbvf/netdev.c +++ b/drivers/net/ethernet/intel/igbvf/netdev.c @@ -2375,7 +2375,7 @@ static netdev_tx_t igbvf_xmit_frame(struct sk_buff *skb, * igbvf_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void igbvf_tx_timeout(struct net_device *netdev) +static void igbvf_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct igbvf_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/igc/Makefile b/drivers/net/ethernet/intel/igc/Makefile index 88c6f88baac5..49fb1e1965cd 100644 --- a/drivers/net/ethernet/intel/igc/Makefile +++ b/drivers/net/ethernet/intel/igc/Makefile @@ -8,4 +8,4 @@ obj-$(CONFIG_IGC) += igc.o igc-objs := igc_main.o igc_mac.o igc_i225.o igc_base.o igc_nvm.o igc_phy.o \ -igc_ethtool.o +igc_ethtool.o igc_ptp.o diff --git a/drivers/net/ethernet/intel/igc/igc.h b/drivers/net/ethernet/intel/igc/igc.h index 0868677d43ed..52066bdbbad0 100644 --- a/drivers/net/ethernet/intel/igc/igc.h +++ b/drivers/net/ethernet/intel/igc/igc.h @@ -10,6 +10,9 @@ #include <linux/vmalloc.h> #include <linux/ethtool.h> #include <linux/sctp.h> +#include <linux/ptp_clock_kernel.h> +#include <linux/timecounter.h> +#include <linux/net_tstamp.h> #include "igc_hw.h" @@ -45,11 +48,15 @@ extern char igc_driver_version[]; #define IGC_REGS_LEN 740 #define IGC_RETA_SIZE 128 +/* flags controlling PTP/1588 function */ +#define IGC_PTP_ENABLED BIT(0) + /* Interrupt defines */ #define IGC_START_ITR 648 /* ~6000 ints/sec */ #define IGC_FLAG_HAS_MSI BIT(0) #define IGC_FLAG_QUEUE_PAIRS BIT(3) #define IGC_FLAG_DMAC BIT(4) +#define IGC_FLAG_PTP BIT(8) #define IGC_FLAG_NEED_LINK_UPDATE BIT(9) #define IGC_FLAG_MEDIA_RESET BIT(10) #define IGC_FLAG_MAS_ENABLE BIT(12) @@ -100,6 +107,20 @@ extern char igc_driver_version[]; #define AUTO_ALL_MODES 0 #define IGC_RX_HDR_LEN IGC_RXBUFFER_256 +/* Transmit and receive latency (for PTP timestamps) */ +/* FIXME: These values were estimated using the ones that i210 has as + * basis, they seem to provide good numbers with ptp4l/phc2sys, but we + * need to confirm them. + */ +#define IGC_I225_TX_LATENCY_10 9542 +#define IGC_I225_TX_LATENCY_100 1024 +#define IGC_I225_TX_LATENCY_1000 178 +#define IGC_I225_TX_LATENCY_2500 64 +#define IGC_I225_RX_LATENCY_10 20662 +#define IGC_I225_RX_LATENCY_100 2213 +#define IGC_I225_RX_LATENCY_1000 448 +#define IGC_I225_RX_LATENCY_2500 160 + /* RX and TX descriptor control thresholds. * PTHRESH - MAC will consider prefetch if it has fewer than this number of * descriptors available in its onboard memory. @@ -370,6 +391,8 @@ struct igc_adapter { struct timer_list dma_err_timer; struct timer_list phy_info_timer; + u32 wol; + u32 en_mng_pt; u16 link_speed; u16 link_duplex; @@ -430,6 +453,20 @@ struct igc_adapter { unsigned long link_check_timeout; struct igc_info ei; + + struct ptp_clock *ptp_clock; + struct ptp_clock_info ptp_caps; + struct work_struct ptp_tx_work; + struct sk_buff *ptp_tx_skb; + struct hwtstamp_config tstamp_config; + unsigned long ptp_tx_start; + unsigned long last_rx_ptp_check; + unsigned long last_rx_timestamp; + unsigned int ptp_flags; + /* System time value lock */ + spinlock_t tmreg_lock; + struct cyclecounter cc; + struct timecounter tc; }; /* igc_desc_unused - calculate if we have unused descriptors */ @@ -513,6 +550,16 @@ int igc_add_filter(struct igc_adapter *adapter, int igc_erase_filter(struct igc_adapter *adapter, struct igc_nfc_filter *input); +void igc_ptp_init(struct igc_adapter *adapter); +void igc_ptp_reset(struct igc_adapter *adapter); +void igc_ptp_stop(struct igc_adapter *adapter); +void igc_ptp_rx_rgtstamp(struct igc_q_vector *q_vector, struct sk_buff *skb); +void igc_ptp_rx_pktstamp(struct igc_q_vector *q_vector, void *va, + struct sk_buff *skb); +int igc_ptp_set_ts_config(struct net_device *netdev, struct ifreq *ifr); +int igc_ptp_get_ts_config(struct net_device *netdev, struct ifreq *ifr); +void igc_ptp_tx_hang(struct igc_adapter *adapter); + #define igc_rx_pg_size(_ring) (PAGE_SIZE << igc_rx_pg_order(_ring)) #define IGC_TXD_DCMD (IGC_ADVTXD_DCMD_EOP | IGC_ADVTXD_DCMD_RS) diff --git a/drivers/net/ethernet/intel/igc/igc_base.c b/drivers/net/ethernet/intel/igc/igc_base.c index db289bcce21d..5a506440560a 100644 --- a/drivers/net/ethernet/intel/igc/igc_base.c +++ b/drivers/net/ethernet/intel/igc/igc_base.c @@ -212,6 +212,7 @@ static s32 igc_get_invariants_base(struct igc_hw *hw) case IGC_DEV_ID_I225_I: case IGC_DEV_ID_I220_V: case IGC_DEV_ID_I225_K: + case IGC_DEV_ID_I225_BLANK_NVM: mac->type = igc_i225; break; default: diff --git a/drivers/net/ethernet/intel/igc/igc_defines.h b/drivers/net/ethernet/intel/igc/igc_defines.h index f3788f0b95b4..58efa7a02c68 100644 --- a/drivers/net/ethernet/intel/igc/igc_defines.h +++ b/drivers/net/ethernet/intel/igc/igc_defines.h @@ -10,6 +10,37 @@ #define IGC_CTRL_EXT_DRV_LOAD 0x10000000 /* Drv loaded bit for FW */ +/* Definitions for power management and wakeup registers */ +/* Wake Up Control */ +#define IGC_WUC_PME_EN 0x00000002 /* PME Enable */ + +/* Wake Up Filter Control */ +#define IGC_WUFC_LNKC 0x00000001 /* Link Status Change Wakeup Enable */ +#define IGC_WUFC_MC 0x00000008 /* Directed Multicast Wakeup Enable */ + +#define IGC_CTRL_ADVD3WUC 0x00100000 /* D3 WUC */ + +/* Wake Up Status */ +#define IGC_WUS_EX 0x00000004 /* Directed Exact */ +#define IGC_WUS_ARPD 0x00000020 /* Directed ARP Request */ +#define IGC_WUS_IPV4 0x00000040 /* Directed IPv4 */ +#define IGC_WUS_IPV6 0x00000080 /* Directed IPv6 */ +#define IGC_WUS_NSD 0x00000400 /* Directed IPv6 Neighbor Solicitation */ + +/* Packet types that are enabled for wake packet delivery */ +#define WAKE_PKT_WUS ( \ + IGC_WUS_EX | \ + IGC_WUS_ARPD | \ + IGC_WUS_IPV4 | \ + IGC_WUS_IPV6 | \ + IGC_WUS_NSD) + +/* Wake Up Packet Length */ +#define IGC_WUPL_MASK 0x00000FFF + +/* Wake Up Packet Memory stores the first 128 bytes of the wake up packet */ +#define IGC_WUPM_BYTES 128 + /* Physical Func Reset Done Indication */ #define IGC_CTRL_EXT_LINK_MODE_MASK 0x00C00000 @@ -187,6 +218,7 @@ #define IGC_ICR_RXDMT0 BIT(4) /* Rx desc min. threshold (0) */ #define IGC_ICR_RXO BIT(6) /* Rx overrun */ #define IGC_ICR_RXT0 BIT(7) /* Rx timer intr (ring 0) */ +#define IGC_ICR_TS BIT(19) /* Time Sync Interrupt */ #define IGC_ICR_DRSTA BIT(30) /* Device Reset Asserted */ /* If this bit asserted, the driver should claim the interrupt */ @@ -209,6 +241,7 @@ #define IGC_IMS_DRSTA IGC_ICR_DRSTA /* Device Reset Asserted */ #define IGC_IMS_RXT0 IGC_ICR_RXT0 /* Rx timer intr */ #define IGC_IMS_RXDMT0 IGC_ICR_RXDMT0 /* Rx desc min. threshold */ +#define IGC_IMS_TS IGC_ICR_TS /* Time Sync Interrupt */ #define IGC_QVECTOR_MASK 0x7FFC /* Q-vector mask */ #define IGC_ITR_VAL_MASK 0x04 /* ITR value mask */ @@ -249,6 +282,10 @@ #define IGC_TXD_STAT_TC 0x00000004 /* Tx Underrun */ #define IGC_TXD_EXTCMD_TSTAMP 0x00000010 /* IEEE1588 Timestamp packet */ +/* IPSec Encrypt Enable */ +#define IGC_ADVTXD_L4LEN_SHIFT 8 /* Adv ctxt L4LEN shift */ +#define IGC_ADVTXD_MSS_SHIFT 16 /* Adv ctxt MSS shift */ + /* Transmit Control */ #define IGC_TCTL_EN 0x00000002 /* enable Tx */ #define IGC_TCTL_PSP 0x00000008 /* pad short packets */ @@ -281,12 +318,21 @@ #define IGC_RCTL_RDMTS_HALF 0x00000000 /* Rx desc min thresh size */ #define IGC_RCTL_BAM 0x00008000 /* broadcast enable */ +/* Split Replication Receive Control */ +#define IGC_SRRCTL_TIMESTAMP 0x40000000 +#define IGC_SRRCTL_TIMER1SEL(timer) (((timer) & 0x3) << 14) +#define IGC_SRRCTL_TIMER0SEL(timer) (((timer) & 0x3) << 17) + /* Receive Descriptor bit definitions */ #define IGC_RXD_STAT_EOP 0x02 /* End of Packet */ #define IGC_RXD_STAT_IXSM 0x04 /* Ignore checksum */ #define IGC_RXD_STAT_UDPCS 0x10 /* UDP xsum calculated */ #define IGC_RXD_STAT_TCPCS 0x20 /* TCP xsum calculated */ +/* Advanced Receive Descriptor bit definitions */ +#define IGC_RXDADV_STAT_TSIP 0x08000 /* timestamp in packet */ +#define IGC_RXDADV_STAT_TS 0x10000 /* Pkt was time stamped */ + #define IGC_RXDEXT_STATERR_CE 0x01000000 #define IGC_RXDEXT_STATERR_SE 0x02000000 #define IGC_RXDEXT_STATERR_SEQ 0x04000000 @@ -323,6 +369,61 @@ #define I225_RXPBSIZE_DEFAULT 0x000000A2 /* RXPBSIZE default */ #define I225_TXPBSIZE_DEFAULT 0x04000014 /* TXPBSIZE default */ +#define IGC_RXPBS_CFG_TS_EN 0x80000000 /* Timestamp in Rx buffer */ + +/* Time Sync Interrupt Causes */ +#define IGC_TSICR_SYS_WRAP BIT(0) /* SYSTIM Wrap around. */ +#define IGC_TSICR_TXTS BIT(1) /* Transmit Timestamp. */ +#define IGC_TSICR_TT0 BIT(3) /* Target Time 0 Trigger. */ +#define IGC_TSICR_TT1 BIT(4) /* Target Time 1 Trigger. */ +#define IGC_TSICR_AUTT0 BIT(5) /* Auxiliary Timestamp 0 Taken. */ +#define IGC_TSICR_AUTT1 BIT(6) /* Auxiliary Timestamp 1 Taken. */ + +#define IGC_TSICR_INTERRUPTS IGC_TSICR_TXTS + +/* PTP Queue Filter */ +#define IGC_ETQF_1588 BIT(30) + +#define IGC_FTQF_VF_BP 0x00008000 +#define IGC_FTQF_1588_TIME_STAMP 0x08000000 +#define IGC_FTQF_MASK 0xF0000000 +#define IGC_FTQF_MASK_PROTO_BP 0x10000000 + +/* Time Sync Receive Control bit definitions */ +#define IGC_TSYNCRXCTL_VALID 0x00000001 /* Rx timestamp valid */ +#define IGC_TSYNCRXCTL_TYPE_MASK 0x0000000E /* Rx type mask */ +#define IGC_TSYNCRXCTL_TYPE_L2_V2 0x00 +#define IGC_TSYNCRXCTL_TYPE_L4_V1 0x02 +#define IGC_TSYNCRXCTL_TYPE_L2_L4_V2 0x04 +#define IGC_TSYNCRXCTL_TYPE_ALL 0x08 +#define IGC_TSYNCRXCTL_TYPE_EVENT_V2 0x0A +#define IGC_TSYNCRXCTL_ENABLED 0x00000010 /* enable Rx timestamping */ +#define IGC_TSYNCRXCTL_SYSCFI 0x00000020 /* Sys clock frequency */ +#define IGC_TSYNCRXCTL_RXSYNSIG 0x00000400 /* Sample RX tstamp in PHY sop */ + +/* Time Sync Receive Configuration */ +#define IGC_TSYNCRXCFG_PTP_V1_CTRLT_MASK 0x000000FF +#define IGC_TSYNCRXCFG_PTP_V1_SYNC_MESSAGE 0x00 +#define IGC_TSYNCRXCFG_PTP_V1_DELAY_REQ_MESSAGE 0x01 + +/* Immediate Interrupt Receive */ +#define IGC_IMIR_CLEAR_MASK 0xF001FFFF /* IMIR Reg Clear Mask */ +#define IGC_IMIR_PORT_BYPASS 0x20000 /* IMIR Port Bypass Bit */ +#define IGC_IMIR_PRIORITY_SHIFT 29 /* IMIR Priority Shift */ +#define IGC_IMIREXT_CLEAR_MASK 0x7FFFF /* IMIREXT Reg Clear Mask */ + +/* Immediate Interrupt Receive Extended */ +#define IGC_IMIREXT_CTRL_BP 0x00080000 /* Bypass check of ctrl bits */ +#define IGC_IMIREXT_SIZE_BP 0x00001000 /* Packet size bypass */ + +/* Time Sync Transmit Control bit definitions */ +#define IGC_TSYNCTXCTL_VALID 0x00000001 /* Tx timestamp valid */ +#define IGC_TSYNCTXCTL_ENABLED 0x00000010 /* enable Tx timestamping */ +#define IGC_TSYNCTXCTL_MAX_ALLOWED_DLY_MASK 0x0000F000 /* max delay */ +#define IGC_TSYNCTXCTL_SYNC_COMP_ERR 0x20000000 /* sync err */ +#define IGC_TSYNCTXCTL_SYNC_COMP 0x40000000 /* sync complete */ +#define IGC_TSYNCTXCTL_START_SYNC 0x80000000 /* initiate sync */ +#define IGC_TSYNCTXCTL_TXSYNSIG 0x00000020 /* Sample TX tstamp in PHY sop */ /* Receive Checksum Control */ #define IGC_RXCSUM_CRCOFL 0x00000800 /* CRC32 offload enable */ @@ -363,6 +464,7 @@ /* PHY Status Register */ #define MII_SR_LINK_STATUS 0x0004 /* Link Status 1 = link */ #define MII_SR_AUTONEG_COMPLETE 0x0020 /* Auto Neg Complete */ +#define IGC_PHY_RST_COMP 0x0100 /* Internal PHY reset completion */ /* PHY 1000 MII Register/Bit Definitions */ /* PHY Registers defined by IEEE */ diff --git a/drivers/net/ethernet/intel/igc/igc_ethtool.c b/drivers/net/ethernet/intel/igc/igc_ethtool.c index 455c1cdceb6e..ee07011e13e9 100644 --- a/drivers/net/ethernet/intel/igc/igc_ethtool.c +++ b/drivers/net/ethernet/intel/igc/igc_ethtool.c @@ -1600,6 +1600,39 @@ static int igc_set_channels(struct net_device *netdev, return 0; } +static int igc_get_ts_info(struct net_device *dev, + struct ethtool_ts_info *info) +{ + struct igc_adapter *adapter = netdev_priv(dev); + + if (adapter->ptp_clock) + info->phc_index = ptp_clock_index(adapter->ptp_clock); + else + info->phc_index = -1; + + switch (adapter->hw.mac.type) { + case igc_i225: + info->so_timestamping = + SOF_TIMESTAMPING_TX_SOFTWARE | + SOF_TIMESTAMPING_RX_SOFTWARE | + SOF_TIMESTAMPING_SOFTWARE | + SOF_TIMESTAMPING_TX_HARDWARE | + SOF_TIMESTAMPING_RX_HARDWARE | + SOF_TIMESTAMPING_RAW_HARDWARE; + + info->tx_types = + BIT(HWTSTAMP_TX_OFF) | + BIT(HWTSTAMP_TX_ON); + + info->rx_filters = BIT(HWTSTAMP_FILTER_NONE); + info->rx_filters |= BIT(HWTSTAMP_FILTER_ALL); + + return 0; + default: + return -EOPNOTSUPP; + } +} + static u32 igc_get_priv_flags(struct net_device *netdev) { struct igc_adapter *adapter = netdev_priv(netdev); @@ -1847,6 +1880,7 @@ static const struct ethtool_ops igc_ethtool_ops = { .get_rxfh_indir_size = igc_get_rxfh_indir_size, .get_rxfh = igc_get_rxfh, .set_rxfh = igc_set_rxfh, + .get_ts_info = igc_get_ts_info, .get_channels = igc_get_channels, .set_channels = igc_set_channels, .get_priv_flags = igc_get_priv_flags, diff --git a/drivers/net/ethernet/intel/igc/igc_hw.h b/drivers/net/ethernet/intel/igc/igc_hw.h index 20f710645746..90ac0e0144d8 100644 --- a/drivers/net/ethernet/intel/igc/igc_hw.h +++ b/drivers/net/ethernet/intel/igc/igc_hw.h @@ -21,8 +21,7 @@ #define IGC_DEV_ID_I225_I 0x15F8 #define IGC_DEV_ID_I220_V 0x15F7 #define IGC_DEV_ID_I225_K 0x3100 - -#define IGC_FUNC_0 0 +#define IGC_DEV_ID_I225_BLANK_NVM 0x15FD /* Function pointers for the MAC. */ struct igc_mac_operations { diff --git a/drivers/net/ethernet/intel/igc/igc_main.c b/drivers/net/ethernet/intel/igc/igc_main.c index 9700527dd797..d9d5425fe8d9 100644 --- a/drivers/net/ethernet/intel/igc/igc_main.c +++ b/drivers/net/ethernet/intel/igc/igc_main.c @@ -8,6 +8,7 @@ #include <linux/tcp.h> #include <linux/udp.h> #include <linux/ip.h> +#include <linux/pm_runtime.h> #include <net/ipv6.h> @@ -44,31 +45,13 @@ static const struct pci_device_id igc_pci_tbl[] = { { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_I), board_base }, { PCI_VDEVICE(INTEL, IGC_DEV_ID_I220_V), board_base }, { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_K), board_base }, + { PCI_VDEVICE(INTEL, IGC_DEV_ID_I225_BLANK_NVM), board_base }, /* required last entry */ {0, } }; MODULE_DEVICE_TABLE(pci, igc_pci_tbl); -/* forward declaration */ -static void igc_clean_tx_ring(struct igc_ring *tx_ring); -static int igc_sw_init(struct igc_adapter *); -static void igc_configure(struct igc_adapter *adapter); -static void igc_power_down_link(struct igc_adapter *adapter); -static void igc_set_default_mac_filter(struct igc_adapter *adapter); -static void igc_set_rx_mode(struct net_device *netdev); -static void igc_write_itr(struct igc_q_vector *q_vector); -static void igc_assign_vector(struct igc_q_vector *q_vector, int msix_vector); -static void igc_free_q_vector(struct igc_adapter *adapter, int v_idx); -static void igc_set_interrupt_capability(struct igc_adapter *adapter, - bool msix); -static void igc_free_q_vectors(struct igc_adapter *adapter); -static void igc_irq_disable(struct igc_adapter *adapter); -static void igc_irq_enable(struct igc_adapter *adapter); -static void igc_configure_msix(struct igc_adapter *adapter); -static bool igc_alloc_mapped_page(struct igc_ring *rx_ring, - struct igc_rx_buffer *bi); - enum latency_range { lowest_latency = 0, low_latency = 1, @@ -76,6 +59,16 @@ enum latency_range { latency_invalid = 255 }; +/** + * igc_power_down_link - Power down the phy/serdes link + * @adapter: address of board private structure + */ +static void igc_power_down_link(struct igc_adapter *adapter) +{ + if (adapter->hw.phy.media_type == igc_media_type_copper) + igc_power_down_phy_copper_base(&adapter->hw); +} + void igc_reset(struct igc_adapter *adapter) { struct pci_dev *pdev = adapter->pdev; @@ -110,11 +103,14 @@ void igc_reset(struct igc_adapter *adapter) if (!netif_running(adapter->netdev)) igc_power_down_link(adapter); + /* Re-enable PTP, where applicable. */ + igc_ptp_reset(adapter); + igc_get_phy_info(hw); } /** - * igc_power_up_link - Power up the phy/serdes link + * igc_power_up_link - Power up the phy link * @adapter: address of board private structure */ static void igc_power_up_link(struct igc_adapter *adapter) @@ -128,16 +124,6 @@ static void igc_power_up_link(struct igc_adapter *adapter) } /** - * igc_power_down_link - Power down the phy/serdes link - * @adapter: address of board private structure - */ -static void igc_power_down_link(struct igc_adapter *adapter) -{ - if (adapter->hw.phy.media_type == igc_media_type_copper) - igc_power_down_phy_copper_base(&adapter->hw); -} - -/** * igc_release_hw_control - release control of the h/w to f/w * @adapter: address of board private structure * @@ -176,43 +162,6 @@ static void igc_get_hw_control(struct igc_adapter *adapter) } /** - * igc_free_tx_resources - Free Tx Resources per Queue - * @tx_ring: Tx descriptor ring for a specific queue - * - * Free all transmit software resources - */ -void igc_free_tx_resources(struct igc_ring *tx_ring) -{ - igc_clean_tx_ring(tx_ring); - - vfree(tx_ring->tx_buffer_info); - tx_ring->tx_buffer_info = NULL; - - /* if not set, then don't free */ - if (!tx_ring->desc) - return; - - dma_free_coherent(tx_ring->dev, tx_ring->size, - tx_ring->desc, tx_ring->dma); - - tx_ring->desc = NULL; -} - -/** - * igc_free_all_tx_resources - Free Tx Resources for All Queues - * @adapter: board private structure - * - * Free all transmit software resources - */ -static void igc_free_all_tx_resources(struct igc_adapter *adapter) -{ - int i; - - for (i = 0; i < adapter->num_tx_queues; i++) - igc_free_tx_resources(adapter->tx_ring[i]); -} - -/** * igc_clean_tx_ring - Free Tx Buffers * @tx_ring: ring to be cleaned */ @@ -274,6 +223,43 @@ static void igc_clean_tx_ring(struct igc_ring *tx_ring) } /** + * igc_free_tx_resources - Free Tx Resources per Queue + * @tx_ring: Tx descriptor ring for a specific queue + * + * Free all transmit software resources + */ +void igc_free_tx_resources(struct igc_ring *tx_ring) +{ + igc_clean_tx_ring(tx_ring); + + vfree(tx_ring->tx_buffer_info); + tx_ring->tx_buffer_info = NULL; + + /* if not set, then don't free */ + if (!tx_ring->desc) + return; + + dma_free_coherent(tx_ring->dev, tx_ring->size, + tx_ring->desc, tx_ring->dma); + + tx_ring->desc = NULL; +} + +/** + * igc_free_all_tx_resources - Free Tx Resources for All Queues + * @adapter: board private structure + * + * Free all transmit software resources + */ +static void igc_free_all_tx_resources(struct igc_adapter *adapter) +{ + int i; + + for (i = 0; i < adapter->num_tx_queues; i++) + igc_free_tx_resources(adapter->tx_ring[i]); +} + +/** * igc_clean_all_tx_rings - Free Tx Buffers for all queues * @adapter: board private structure */ @@ -771,6 +757,51 @@ static void igc_setup_tctl(struct igc_adapter *adapter) } /** + * igc_rar_set_index - Sync RAL[index] and RAH[index] registers with MAC table + * @adapter: address of board private structure + * @index: Index of the RAR entry which need to be synced with MAC table + */ +static void igc_rar_set_index(struct igc_adapter *adapter, u32 index) +{ + u8 *addr = adapter->mac_table[index].addr; + struct igc_hw *hw = &adapter->hw; + u32 rar_low, rar_high; + + /* HW expects these to be in network order when they are plugged + * into the registers which are little endian. In order to guarantee + * that ordering we need to do an leXX_to_cpup here in order to be + * ready for the byteswap that occurs with writel + */ + rar_low = le32_to_cpup((__le32 *)(addr)); + rar_high = le16_to_cpup((__le16 *)(addr + 4)); + + /* Indicate to hardware the Address is Valid. */ + if (adapter->mac_table[index].state & IGC_MAC_STATE_IN_USE) { + if (is_valid_ether_addr(addr)) + rar_high |= IGC_RAH_AV; + + rar_high |= IGC_RAH_POOL_1 << + adapter->mac_table[index].queue; + } + + wr32(IGC_RAL(index), rar_low); + wrfl(); + wr32(IGC_RAH(index), rar_high); + wrfl(); +} + +/* Set default MAC address for the PF in the first RAR entry */ +static void igc_set_default_mac_filter(struct igc_adapter *adapter) +{ + struct igc_mac_addr *mac_table = &adapter->mac_table[0]; + + ether_addr_copy(mac_table->addr, adapter->hw.mac.addr); + mac_table->state = IGC_MAC_STATE_DEFAULT | IGC_MAC_STATE_IN_USE; + + igc_rar_set_index(adapter, 0); +} + +/** * igc_set_mac - Change the Ethernet Address of the NIC * @netdev: network interface device structure * @p: pointer to an address structure @@ -850,7 +881,7 @@ static void igc_tx_ctxtdesc(struct igc_ring *tx_ring, /* set bits to identify this as an advanced context descriptor */ type_tucmd |= IGC_TXD_CMD_DEXT | IGC_ADVTXD_DTYP_CTXT; - /* For 82575, context index must be unique per ring. */ + /* For i225, context index must be unique per ring. */ if (test_bit(IGC_RING_FLAG_TX_CTX_IDX, &tx_ring->flags)) mss_l4len_idx |= tx_ring->reg_idx << 4; @@ -957,6 +988,11 @@ static inline int igc_maybe_stop_tx(struct igc_ring *tx_ring, const u16 size) return __igc_maybe_stop_tx(tx_ring, size); } +#define IGC_SET_FLAG(_input, _flag, _result) \ + (((_flag) <= (_result)) ? \ + ((u32)((_input) & (_flag)) * ((_result) / (_flag))) : \ + ((u32)((_input) & (_flag)) / ((_flag) / (_result)))) + static u32 igc_tx_cmd_type(struct sk_buff *skb, u32 tx_flags) { /* set type for advanced descriptor with frame checksum insertion */ @@ -964,6 +1000,14 @@ static u32 igc_tx_cmd_type(struct sk_buff *skb, u32 tx_flags) IGC_ADVTXD_DCMD_DEXT | IGC_ADVTXD_DCMD_IFCS; + /* set segmentation bits for TSO */ + cmd_type |= IGC_SET_FLAG(tx_flags, IGC_TX_FLAGS_TSO, + (IGC_ADVTXD_DCMD_TSE)); + + /* set timestamp bit if present */ + cmd_type |= IGC_SET_FLAG(tx_flags, IGC_TX_FLAGS_TSTAMP, + (IGC_ADVTXD_MAC_TSTAMP)); + return cmd_type; } @@ -1131,6 +1175,100 @@ dma_error: return -1; } +static int igc_tso(struct igc_ring *tx_ring, + struct igc_tx_buffer *first, + u8 *hdr_len) +{ + u32 vlan_macip_lens, type_tucmd, mss_l4len_idx; + struct sk_buff *skb = first->skb; + union { + struct iphdr *v4; + struct ipv6hdr *v6; + unsigned char *hdr; + } ip; + union { + struct tcphdr *tcp; + struct udphdr *udp; + unsigned char *hdr; + } l4; + u32 paylen, l4_offset; + int err; + + if (skb->ip_summed != CHECKSUM_PARTIAL) + return 0; + + if (!skb_is_gso(skb)) + return 0; + + err = skb_cow_head(skb, 0); + if (err < 0) + return err; + + ip.hdr = skb_network_header(skb); + l4.hdr = skb_checksum_start(skb); + + /* ADV DTYP TUCMD MKRLOC/ISCSIHEDLEN */ + type_tucmd = IGC_ADVTXD_TUCMD_L4T_TCP; + + /* initialize outer IP header fields */ + if (ip.v4->version == 4) { + unsigned char *csum_start = skb_checksum_start(skb); + unsigned char *trans_start = ip.hdr + (ip.v4->ihl * 4); + + /* IP header will have to cancel out any data that + * is not a part of the outer IP header + */ + ip.v4->check = csum_fold(csum_partial(trans_start, + csum_start - trans_start, + 0)); + type_tucmd |= IGC_ADVTXD_TUCMD_IPV4; + + ip.v4->tot_len = 0; + first->tx_flags |= IGC_TX_FLAGS_TSO | + IGC_TX_FLAGS_CSUM | + IGC_TX_FLAGS_IPV4; + } else { + ip.v6->payload_len = 0; + first->tx_flags |= IGC_TX_FLAGS_TSO | + IGC_TX_FLAGS_CSUM; + } + + /* determine offset of inner transport header */ + l4_offset = l4.hdr - skb->data; + + /* remove payload length from inner checksum */ + paylen = skb->len - l4_offset; + if (type_tucmd & IGC_ADVTXD_TUCMD_L4T_TCP) { + /* compute length of segmentation header */ + *hdr_len = (l4.tcp->doff * 4) + l4_offset; + csum_replace_by_diff(&l4.tcp->check, + (__force __wsum)htonl(paylen)); + } else { + /* compute length of segmentation header */ + *hdr_len = sizeof(*l4.udp) + l4_offset; + csum_replace_by_diff(&l4.udp->check, + (__force __wsum)htonl(paylen)); + } + + /* update gso size and bytecount with header size */ + first->gso_segs = skb_shinfo(skb)->gso_segs; + first->bytecount += (first->gso_segs - 1) * *hdr_len; + + /* MSS L4LEN IDX */ + mss_l4len_idx = (*hdr_len - l4_offset) << IGC_ADVTXD_L4LEN_SHIFT; + mss_l4len_idx |= skb_shinfo(skb)->gso_size << IGC_ADVTXD_MSS_SHIFT; + + /* VLAN MACLEN IPLEN */ + vlan_macip_lens = l4.hdr - ip.hdr; + vlan_macip_lens |= (ip.hdr - skb->data) << IGC_ADVTXD_MACLEN_SHIFT; + vlan_macip_lens |= first->tx_flags & IGC_TX_FLAGS_VLAN_MASK; + + igc_tx_ctxtdesc(tx_ring, first, vlan_macip_lens, + type_tucmd, mss_l4len_idx); + + return 1; +} + static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, struct igc_ring *tx_ring) { @@ -1140,6 +1278,7 @@ static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, u32 tx_flags = 0; unsigned short f; u8 hdr_len = 0; + int tso = 0; /* need: 1 descriptor per page * PAGE_SIZE/IGC_MAX_DATA_PER_TXD, * + 1 desc for skb_headlen/IGC_MAX_DATA_PER_TXD, @@ -1162,15 +1301,45 @@ static netdev_tx_t igc_xmit_frame_ring(struct sk_buff *skb, first->bytecount = skb->len; first->gso_segs = 1; + if (unlikely(skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP)) { + struct igc_adapter *adapter = netdev_priv(tx_ring->netdev); + + /* FIXME: add support for retrieving timestamps from + * the other timer registers before skipping the + * timestamping request. + */ + if (adapter->tstamp_config.tx_type == HWTSTAMP_TX_ON && + !test_and_set_bit_lock(__IGC_PTP_TX_IN_PROGRESS, + &adapter->state)) { + skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS; + tx_flags |= IGC_TX_FLAGS_TSTAMP; + + adapter->ptp_tx_skb = skb_get(skb); + adapter->ptp_tx_start = jiffies; + } else { + adapter->tx_hwtstamp_skipped++; + } + } + /* record initial flags and protocol */ first->tx_flags = tx_flags; first->protocol = protocol; - igc_tx_csum(tx_ring, first); + tso = igc_tso(tx_ring, first, &hdr_len); + if (tso < 0) + goto out_drop; + else if (!tso) + igc_tx_csum(tx_ring, first); igc_tx_map(tx_ring, first, hdr_len); return NETDEV_TX_OK; + +out_drop: + dev_kfree_skb_any(first->skb); + first->skb = NULL; + + return NETDEV_TX_OK; } static inline struct igc_ring *igc_tx_queue_mapping(struct igc_adapter *adapter, @@ -1269,6 +1438,10 @@ static void igc_process_skb_fields(struct igc_ring *rx_ring, igc_rx_checksum(rx_ring, rx_desc, skb); + if (igc_test_staterr(rx_desc, IGC_RXDADV_STAT_TS) && + !igc_test_staterr(rx_desc, IGC_RXDADV_STAT_TSIP)) + igc_ptp_rx_rgtstamp(rx_ring->q_vector, skb); + skb_record_rx_queue(skb, rx_ring->queue_index); skb->protocol = eth_type_trans(skb, rx_ring->netdev); @@ -1388,6 +1561,12 @@ static struct sk_buff *igc_construct_skb(struct igc_ring *rx_ring, if (unlikely(!skb)) return NULL; + if (unlikely(igc_test_staterr(rx_desc, IGC_RXDADV_STAT_TSIP))) { + igc_ptp_rx_pktstamp(rx_ring->q_vector, va, skb); + va += IGC_TS_HDR_LEN; + size -= IGC_TS_HDR_LEN; + } + /* Determine available headroom for copy */ headlen = size; if (headlen > IGC_RX_HDR_LEN) @@ -1485,7 +1664,6 @@ static bool igc_can_reuse_rx_page(struct igc_rx_buffer *rx_buffer) * igc_is_non_eop - process handling of non-EOP buffers * @rx_ring: Rx ring being processed * @rx_desc: Rx descriptor for current buffer - * @skb: current socket buffer containing buffer in progress * * This function updates next to clean. If the buffer is an EOP buffer * this function exits returning false, otherwise it will place the @@ -1565,9 +1743,56 @@ static void igc_put_rx_buffer(struct igc_ring *rx_ring, rx_buffer->page = NULL; } +static inline unsigned int igc_rx_offset(struct igc_ring *rx_ring) +{ + return ring_uses_build_skb(rx_ring) ? IGC_SKB_PAD : 0; +} + +static bool igc_alloc_mapped_page(struct igc_ring *rx_ring, + struct igc_rx_buffer *bi) +{ + struct page *page = bi->page; + dma_addr_t dma; + + /* since we are recycling buffers we should seldom need to alloc */ + if (likely(page)) + return true; + + /* alloc new page for storage */ + page = dev_alloc_pages(igc_rx_pg_order(rx_ring)); + if (unlikely(!page)) { + rx_ring->rx_stats.alloc_failed++; + return false; + } + + /* map page for use */ + dma = dma_map_page_attrs(rx_ring->dev, page, 0, + igc_rx_pg_size(rx_ring), + DMA_FROM_DEVICE, + IGC_RX_DMA_ATTR); + + /* if mapping failed free memory back to system since + * there isn't much point in holding memory we can't use + */ + if (dma_mapping_error(rx_ring->dev, dma)) { + __free_page(page); + + rx_ring->rx_stats.alloc_failed++; + return false; + } + + bi->dma = dma; + bi->page = page; + bi->page_offset = igc_rx_offset(rx_ring); + bi->pagecnt_bias = 1; + + return true; +} + /** * igc_alloc_rx_buffers - Replace used receive buffers; packet split - * @adapter: address of board private structure + * @rx_ring: rx descriptor ring + * @cleaned_count: number of buffers to clean */ static void igc_alloc_rx_buffers(struct igc_ring *rx_ring, u16 cleaned_count) { @@ -1725,52 +1950,6 @@ static int igc_clean_rx_irq(struct igc_q_vector *q_vector, const int budget) return total_packets; } -static inline unsigned int igc_rx_offset(struct igc_ring *rx_ring) -{ - return ring_uses_build_skb(rx_ring) ? IGC_SKB_PAD : 0; -} - -static bool igc_alloc_mapped_page(struct igc_ring *rx_ring, - struct igc_rx_buffer *bi) -{ - struct page *page = bi->page; - dma_addr_t dma; - - /* since we are recycling buffers we should seldom need to alloc */ - if (likely(page)) - return true; - - /* alloc new page for storage */ - page = dev_alloc_pages(igc_rx_pg_order(rx_ring)); - if (unlikely(!page)) { - rx_ring->rx_stats.alloc_failed++; - return false; - } - - /* map page for use */ - dma = dma_map_page_attrs(rx_ring->dev, page, 0, - igc_rx_pg_size(rx_ring), - DMA_FROM_DEVICE, - IGC_RX_DMA_ATTR); - - /* if mapping failed free memory back to system since - * there isn't much point in holding memory we can't use - */ - if (dma_mapping_error(rx_ring->dev, dma)) { - __free_page(page); - - rx_ring->rx_stats.alloc_failed++; - return false; - } - - bi->dma = dma; - bi->page = page; - bi->page_offset = igc_rx_offset(rx_ring); - bi->pagecnt_bias = 1; - - return true; -} - /** * igc_clean_tx_irq - Reclaim resources after transmit completes * @q_vector: pointer to q_vector containing needed info @@ -1942,6 +2121,1128 @@ static bool igc_clean_tx_irq(struct igc_q_vector *q_vector, int napi_budget) return !!budget; } +static void igc_nfc_filter_restore(struct igc_adapter *adapter) +{ + struct igc_nfc_filter *rule; + + spin_lock(&adapter->nfc_lock); + + hlist_for_each_entry(rule, &adapter->nfc_filter_list, nfc_node) + igc_add_filter(adapter, rule); + + spin_unlock(&adapter->nfc_lock); +} + +/* If the filter to be added and an already existing filter express + * the same address and address type, it should be possible to only + * override the other configurations, for example the queue to steer + * traffic. + */ +static bool igc_mac_entry_can_be_used(const struct igc_mac_addr *entry, + const u8 *addr, const u8 flags) +{ + if (!(entry->state & IGC_MAC_STATE_IN_USE)) + return true; + + if ((entry->state & IGC_MAC_STATE_SRC_ADDR) != + (flags & IGC_MAC_STATE_SRC_ADDR)) + return false; + + if (!ether_addr_equal(addr, entry->addr)) + return false; + + return true; +} + +/* Add a MAC filter for 'addr' directing matching traffic to 'queue', + * 'flags' is used to indicate what kind of match is made, match is by + * default for the destination address, if matching by source address + * is desired the flag IGC_MAC_STATE_SRC_ADDR can be used. + */ +static int igc_add_mac_filter(struct igc_adapter *adapter, + const u8 *addr, const u8 queue) +{ + struct igc_hw *hw = &adapter->hw; + int rar_entries = hw->mac.rar_entry_count; + int i; + + if (is_zero_ether_addr(addr)) + return -EINVAL; + + /* Search for the first empty entry in the MAC table. + * Do not touch entries at the end of the table reserved for the VF MAC + * addresses. + */ + for (i = 0; i < rar_entries; i++) { + if (!igc_mac_entry_can_be_used(&adapter->mac_table[i], + addr, 0)) + continue; + + ether_addr_copy(adapter->mac_table[i].addr, addr); + adapter->mac_table[i].queue = queue; + adapter->mac_table[i].state |= IGC_MAC_STATE_IN_USE; + + igc_rar_set_index(adapter, i); + return i; + } + + return -ENOSPC; +} + +/* Remove a MAC filter for 'addr' directing matching traffic to + * 'queue', 'flags' is used to indicate what kind of match need to be + * removed, match is by default for the destination address, if + * matching by source address is to be removed the flag + * IGC_MAC_STATE_SRC_ADDR can be used. + */ +static int igc_del_mac_filter(struct igc_adapter *adapter, + const u8 *addr, const u8 queue) +{ + struct igc_hw *hw = &adapter->hw; + int rar_entries = hw->mac.rar_entry_count; + int i; + + if (is_zero_ether_addr(addr)) + return -EINVAL; + + /* Search for matching entry in the MAC table based on given address + * and queue. Do not touch entries at the end of the table reserved + * for the VF MAC addresses. + */ + for (i = 0; i < rar_entries; i++) { + if (!(adapter->mac_table[i].state & IGC_MAC_STATE_IN_USE)) + continue; + if (adapter->mac_table[i].state != 0) + continue; + if (adapter->mac_table[i].queue != queue) + continue; + if (!ether_addr_equal(adapter->mac_table[i].addr, addr)) + continue; + + /* When a filter for the default address is "deleted", + * we return it to its initial configuration + */ + if (adapter->mac_table[i].state & IGC_MAC_STATE_DEFAULT) { + adapter->mac_table[i].state = + IGC_MAC_STATE_DEFAULT | IGC_MAC_STATE_IN_USE; + adapter->mac_table[i].queue = 0; + } else { + adapter->mac_table[i].state = 0; + adapter->mac_table[i].queue = 0; + memset(adapter->mac_table[i].addr, 0, ETH_ALEN); + } + + igc_rar_set_index(adapter, i); + return 0; + } + + return -ENOENT; +} + +static int igc_uc_sync(struct net_device *netdev, const unsigned char *addr) +{ + struct igc_adapter *adapter = netdev_priv(netdev); + int ret; + + ret = igc_add_mac_filter(adapter, addr, adapter->num_rx_queues); + + return min_t(int, ret, 0); +} + +static int igc_uc_unsync(struct net_device *netdev, const unsigned char *addr) +{ + struct igc_adapter *adapter = netdev_priv(netdev); + + igc_del_mac_filter(adapter, addr, adapter->num_rx_queues); + + return 0; +} + +/** + * igc_set_rx_mode - Secondary Unicast, Multicast and Promiscuous mode set + * @netdev: network interface device structure + * + * The set_rx_mode entry point is called whenever the unicast or multicast + * address lists or the network interface flags are updated. This routine is + * responsible for configuring the hardware for proper unicast, multicast, + * promiscuous mode, and all-multi behavior. + */ +static void igc_set_rx_mode(struct net_device *netdev) +{ + struct igc_adapter *adapter = netdev_priv(netdev); + struct igc_hw *hw = &adapter->hw; + u32 rctl = 0, rlpml = MAX_JUMBO_FRAME_SIZE; + int count; + + /* Check for Promiscuous and All Multicast modes */ + if (netdev->flags & IFF_PROMISC) { + rctl |= IGC_RCTL_UPE | IGC_RCTL_MPE; + } else { + if (netdev->flags & IFF_ALLMULTI) { + rctl |= IGC_RCTL_MPE; + } else { + /* Write addresses to the MTA, if the attempt fails + * then we should just turn on promiscuous mode so + * that we can at least receive multicast traffic + */ + count = igc_write_mc_addr_list(netdev); + if (count < 0) + rctl |= IGC_RCTL_MPE; + } + } + + /* Write addresses to available RAR registers, if there is not + * sufficient space to store all the addresses then enable + * unicast promiscuous mode + */ + if (__dev_uc_sync(netdev, igc_uc_sync, igc_uc_unsync)) + rctl |= IGC_RCTL_UPE; + + /* update state of unicast and multicast */ + rctl |= rd32(IGC_RCTL) & ~(IGC_RCTL_UPE | IGC_RCTL_MPE); + wr32(IGC_RCTL, rctl); + +#if (PAGE_SIZE < 8192) + if (adapter->max_frame_size <= IGC_MAX_FRAME_BUILD_SKB) + rlpml = IGC_MAX_FRAME_BUILD_SKB; +#endif + wr32(IGC_RLPML, rlpml); +} + +/** + * igc_configure - configure the hardware for RX and TX + * @adapter: private board structure + */ +static void igc_configure(struct igc_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + int i = 0; + + igc_get_hw_control(adapter); + igc_set_rx_mode(netdev); + + igc_setup_tctl(adapter); + igc_setup_mrqc(adapter); + igc_setup_rctl(adapter); + + igc_nfc_filter_restore(adapter); + igc_configure_tx(adapter); + igc_configure_rx(adapter); + + igc_rx_fifo_flush_base(&adapter->hw); + + /* call igc_desc_unused which always leaves + * at least 1 descriptor unused to make sure + * next_to_use != next_to_clean + */ + for (i = 0; i < adapter->num_rx_queues; i++) { + struct igc_ring *ring = adapter->rx_ring[i]; + + igc_alloc_rx_buffers(ring, igc_desc_unused(ring)); + } +} + +/** + * igc_write_ivar - configure ivar for given MSI-X vector + * @hw: pointer to the HW structure + * @msix_vector: vector number we are allocating to a given ring + * @index: row index of IVAR register to write within IVAR table + * @offset: column offset of in IVAR, should be multiple of 8 + * + * The IVAR table consists of 2 columns, + * each containing an cause allocation for an Rx and Tx ring, and a + * variable number of rows depending on the number of queues supported. + */ +static void igc_write_ivar(struct igc_hw *hw, int msix_vector, + int index, int offset) +{ + u32 ivar = array_rd32(IGC_IVAR0, index); + + /* clear any bits that are currently set */ + ivar &= ~((u32)0xFF << offset); + + /* write vector and valid bit */ + ivar |= (msix_vector | IGC_IVAR_VALID) << offset; + + array_wr32(IGC_IVAR0, index, ivar); +} + +static void igc_assign_vector(struct igc_q_vector *q_vector, int msix_vector) +{ + struct igc_adapter *adapter = q_vector->adapter; + struct igc_hw *hw = &adapter->hw; + int rx_queue = IGC_N0_QUEUE; + int tx_queue = IGC_N0_QUEUE; + + if (q_vector->rx.ring) + rx_queue = q_vector->rx.ring->reg_idx; + if (q_vector->tx.ring) + tx_queue = q_vector->tx.ring->reg_idx; + + switch (hw->mac.type) { + case igc_i225: + if (rx_queue > IGC_N0_QUEUE) + igc_write_ivar(hw, msix_vector, + rx_queue >> 1, + (rx_queue & 0x1) << 4); + if (tx_queue > IGC_N0_QUEUE) + igc_write_ivar(hw, msix_vector, + tx_queue >> 1, + ((tx_queue & 0x1) << 4) + 8); + q_vector->eims_value = BIT(msix_vector); + break; + default: + WARN_ONCE(hw->mac.type != igc_i225, "Wrong MAC type\n"); + break; + } + + /* add q_vector eims value to global eims_enable_mask */ + adapter->eims_enable_mask |= q_vector->eims_value; + + /* configure q_vector to set itr on first interrupt */ + q_vector->set_itr = 1; +} + +/** + * igc_configure_msix - Configure MSI-X hardware + * @adapter: Pointer to adapter structure + * + * igc_configure_msix sets up the hardware to properly + * generate MSI-X interrupts. + */ +static void igc_configure_msix(struct igc_adapter *adapter) +{ + struct igc_hw *hw = &adapter->hw; + int i, vector = 0; + u32 tmp; + + adapter->eims_enable_mask = 0; + + /* set vector for other causes, i.e. link changes */ + switch (hw->mac.type) { + case igc_i225: + /* Turn on MSI-X capability first, or our settings + * won't stick. And it will take days to debug. + */ + wr32(IGC_GPIE, IGC_GPIE_MSIX_MODE | + IGC_GPIE_PBA | IGC_GPIE_EIAME | + IGC_GPIE_NSICR); + + /* enable msix_other interrupt */ + adapter->eims_other = BIT(vector); + tmp = (vector++ | IGC_IVAR_VALID) << 8; + + wr32(IGC_IVAR_MISC, tmp); + break; + default: + /* do nothing, since nothing else supports MSI-X */ + break; + } /* switch (hw->mac.type) */ + + adapter->eims_enable_mask |= adapter->eims_other; + + for (i = 0; i < adapter->num_q_vectors; i++) + igc_assign_vector(adapter->q_vector[i], vector++); + + wrfl(); +} + +/** + * igc_irq_enable - Enable default interrupt generation settings + * @adapter: board private structure + */ +static void igc_irq_enable(struct igc_adapter *adapter) +{ + struct igc_hw *hw = &adapter->hw; + + if (adapter->msix_entries) { + u32 ims = IGC_IMS_LSC | IGC_IMS_DOUTSYNC | IGC_IMS_DRSTA; + u32 regval = rd32(IGC_EIAC); + + wr32(IGC_EIAC, regval | adapter->eims_enable_mask); + regval = rd32(IGC_EIAM); + wr32(IGC_EIAM, regval | adapter->eims_enable_mask); + wr32(IGC_EIMS, adapter->eims_enable_mask); + wr32(IGC_IMS, ims); + } else { + wr32(IGC_IMS, IMS_ENABLE_MASK | IGC_IMS_DRSTA); + wr32(IGC_IAM, IMS_ENABLE_MASK | IGC_IMS_DRSTA); + } +} + +/** + * igc_irq_disable - Mask off interrupt generation on the NIC + * @adapter: board private structure + */ +static void igc_irq_disable(struct igc_adapter *adapter) +{ + struct igc_hw *hw = &adapter->hw; + + if (adapter->msix_entries) { + u32 regval = rd32(IGC_EIAM); + + wr32(IGC_EIAM, regval & ~adapter->eims_enable_mask); + wr32(IGC_EIMC, adapter->eims_enable_mask); + regval = rd32(IGC_EIAC); + wr32(IGC_EIAC, regval & ~adapter->eims_enable_mask); + } + + wr32(IGC_IAM, 0); + wr32(IGC_IMC, ~0); + wrfl(); + + if (adapter->msix_entries) { + int vector = 0, i; + + synchronize_irq(adapter->msix_entries[vector++].vector); + + for (i = 0; i < adapter->num_q_vectors; i++) + synchronize_irq(adapter->msix_entries[vector++].vector); + } else { + synchronize_irq(adapter->pdev->irq); + } +} + +void igc_set_flag_queue_pairs(struct igc_adapter *adapter, + const u32 max_rss_queues) +{ + /* Determine if we need to pair queues. */ + /* If rss_queues > half of max_rss_queues, pair the queues in + * order to conserve interrupts due to limited supply. + */ + if (adapter->rss_queues > (max_rss_queues / 2)) + adapter->flags |= IGC_FLAG_QUEUE_PAIRS; + else + adapter->flags &= ~IGC_FLAG_QUEUE_PAIRS; +} + +unsigned int igc_get_max_rss_queues(struct igc_adapter *adapter) +{ + unsigned int max_rss_queues; + + /* Determine the maximum number of RSS queues supported. */ + max_rss_queues = IGC_MAX_RX_QUEUES; + + return max_rss_queues; +} + +static void igc_init_queue_configuration(struct igc_adapter *adapter) +{ + u32 max_rss_queues; + + max_rss_queues = igc_get_max_rss_queues(adapter); + adapter->rss_queues = min_t(u32, max_rss_queues, num_online_cpus()); + + igc_set_flag_queue_pairs(adapter, max_rss_queues); +} + +/** + * igc_reset_q_vector - Reset config for interrupt vector + * @adapter: board private structure to initialize + * @v_idx: Index of vector to be reset + * + * If NAPI is enabled it will delete any references to the + * NAPI struct. This is preparation for igc_free_q_vector. + */ +static void igc_reset_q_vector(struct igc_adapter *adapter, int v_idx) +{ + struct igc_q_vector *q_vector = adapter->q_vector[v_idx]; + + /* if we're coming from igc_set_interrupt_capability, the vectors are + * not yet allocated + */ + if (!q_vector) + return; + + if (q_vector->tx.ring) + adapter->tx_ring[q_vector->tx.ring->queue_index] = NULL; + + if (q_vector->rx.ring) + adapter->rx_ring[q_vector->rx.ring->queue_index] = NULL; + + netif_napi_del(&q_vector->napi); +} + +/** + * igc_free_q_vector - Free memory allocated for specific interrupt vector + * @adapter: board private structure to initialize + * @v_idx: Index of vector to be freed + * + * This function frees the memory allocated to the q_vector. + */ +static void igc_free_q_vector(struct igc_adapter *adapter, int v_idx) +{ + struct igc_q_vector *q_vector = adapter->q_vector[v_idx]; + + adapter->q_vector[v_idx] = NULL; + + /* igc_get_stats64() might access the rings on this vector, + * we must wait a grace period before freeing it. + */ + if (q_vector) + kfree_rcu(q_vector, rcu); +} + +/** + * igc_free_q_vectors - Free memory allocated for interrupt vectors + * @adapter: board private structure to initialize + * + * This function frees the memory allocated to the q_vectors. In addition if + * NAPI is enabled it will delete any references to the NAPI struct prior + * to freeing the q_vector. + */ +static void igc_free_q_vectors(struct igc_adapter *adapter) +{ + int v_idx = adapter->num_q_vectors; + + adapter->num_tx_queues = 0; + adapter->num_rx_queues = 0; + adapter->num_q_vectors = 0; + + while (v_idx--) { + igc_reset_q_vector(adapter, v_idx); + igc_free_q_vector(adapter, v_idx); + } +} + +/** + * igc_update_itr - update the dynamic ITR value based on statistics + * @q_vector: pointer to q_vector + * @ring_container: ring info to update the itr for + * + * Stores a new ITR value based on packets and byte + * counts during the last interrupt. The advantage of per interrupt + * computation is faster updates and more accurate ITR for the current + * traffic pattern. Constants in this function were computed + * based on theoretical maximum wire speed and thresholds were set based + * on testing data as well as attempting to minimize response time + * while increasing bulk throughput. + * NOTE: These calculations are only valid when operating in a single- + * queue environment. + */ +static void igc_update_itr(struct igc_q_vector *q_vector, + struct igc_ring_container *ring_container) +{ + unsigned int packets = ring_container->total_packets; + unsigned int bytes = ring_container->total_bytes; + u8 itrval = ring_container->itr; + + /* no packets, exit with status unchanged */ + if (packets == 0) + return; + + switch (itrval) { + case lowest_latency: + /* handle TSO and jumbo frames */ + if (bytes / packets > 8000) + itrval = bulk_latency; + else if ((packets < 5) && (bytes > 512)) + itrval = low_latency; + break; + case low_latency: /* 50 usec aka 20000 ints/s */ + if (bytes > 10000) { + /* this if handles the TSO accounting */ + if (bytes / packets > 8000) + itrval = bulk_latency; + else if ((packets < 10) || ((bytes / packets) > 1200)) + itrval = bulk_latency; + else if ((packets > 35)) + itrval = lowest_latency; + } else if (bytes / packets > 2000) { + itrval = bulk_latency; + } else if (packets <= 2 && bytes < 512) { + itrval = lowest_latency; + } + break; + case bulk_latency: /* 250 usec aka 4000 ints/s */ + if (bytes > 25000) { + if (packets > 35) + itrval = low_latency; + } else if (bytes < 1500) { + itrval = low_latency; + } + break; + } + + /* clear work counters since we have the values we need */ + ring_container->total_bytes = 0; + ring_container->total_packets = 0; + + /* write updated itr to ring container */ + ring_container->itr = itrval; +} + +static void igc_set_itr(struct igc_q_vector *q_vector) +{ + struct igc_adapter *adapter = q_vector->adapter; + u32 new_itr = q_vector->itr_val; + u8 current_itr = 0; + + /* for non-gigabit speeds, just fix the interrupt rate at 4000 */ + switch (adapter->link_speed) { + case SPEED_10: + case SPEED_100: + current_itr = 0; + new_itr = IGC_4K_ITR; + goto set_itr_now; + default: + break; + } + + igc_update_itr(q_vector, &q_vector->tx); + igc_update_itr(q_vector, &q_vector->rx); + + current_itr = max(q_vector->rx.itr, q_vector->tx.itr); + + /* conservative mode (itr 3) eliminates the lowest_latency setting */ + if (current_itr == lowest_latency && + ((q_vector->rx.ring && adapter->rx_itr_setting == 3) || + (!q_vector->rx.ring && adapter->tx_itr_setting == 3))) + current_itr = low_latency; + + switch (current_itr) { + /* counts and packets in update_itr are dependent on these numbers */ + case lowest_latency: + new_itr = IGC_70K_ITR; /* 70,000 ints/sec */ + break; + case low_latency: + new_itr = IGC_20K_ITR; /* 20,000 ints/sec */ + break; + case bulk_latency: + new_itr = IGC_4K_ITR; /* 4,000 ints/sec */ + break; + default: + break; + } + +set_itr_now: + if (new_itr != q_vector->itr_val) { + /* this attempts to bias the interrupt rate towards Bulk + * by adding intermediate steps when interrupt rate is + * increasing + */ + new_itr = new_itr > q_vector->itr_val ? + max((new_itr * q_vector->itr_val) / + (new_itr + (q_vector->itr_val >> 2)), + new_itr) : new_itr; + /* Don't write the value here; it resets the adapter's + * internal timer, and causes us to delay far longer than + * we should between interrupts. Instead, we write the ITR + * value at the beginning of the next interrupt so the timing + * ends up being correct. + */ + q_vector->itr_val = new_itr; + q_vector->set_itr = 1; + } +} + +static void igc_reset_interrupt_capability(struct igc_adapter *adapter) +{ + int v_idx = adapter->num_q_vectors; + + if (adapter->msix_entries) { + pci_disable_msix(adapter->pdev); + kfree(adapter->msix_entries); + adapter->msix_entries = NULL; + } else if (adapter->flags & IGC_FLAG_HAS_MSI) { + pci_disable_msi(adapter->pdev); + } + + while (v_idx--) + igc_reset_q_vector(adapter, v_idx); +} + +/** + * igc_set_interrupt_capability - set MSI or MSI-X if supported + * @adapter: Pointer to adapter structure + * @msix: boolean value for MSI-X capability + * + * Attempt to configure interrupts using the best available + * capabilities of the hardware and kernel. + */ +static void igc_set_interrupt_capability(struct igc_adapter *adapter, + bool msix) +{ + int numvecs, i; + int err; + + if (!msix) + goto msi_only; + adapter->flags |= IGC_FLAG_HAS_MSIX; + + /* Number of supported queues. */ + adapter->num_rx_queues = adapter->rss_queues; + + adapter->num_tx_queues = adapter->rss_queues; + + /* start with one vector for every Rx queue */ + numvecs = adapter->num_rx_queues; + + /* if Tx handler is separate add 1 for every Tx queue */ + if (!(adapter->flags & IGC_FLAG_QUEUE_PAIRS)) + numvecs += adapter->num_tx_queues; + + /* store the number of vectors reserved for queues */ + adapter->num_q_vectors = numvecs; + + /* add 1 vector for link status interrupts */ + numvecs++; + + adapter->msix_entries = kcalloc(numvecs, sizeof(struct msix_entry), + GFP_KERNEL); + + if (!adapter->msix_entries) + return; + + /* populate entry values */ + for (i = 0; i < numvecs; i++) + adapter->msix_entries[i].entry = i; + + err = pci_enable_msix_range(adapter->pdev, + adapter->msix_entries, + numvecs, + numvecs); + if (err > 0) + return; + + kfree(adapter->msix_entries); + adapter->msix_entries = NULL; + + igc_reset_interrupt_capability(adapter); + +msi_only: + adapter->flags &= ~IGC_FLAG_HAS_MSIX; + + adapter->rss_queues = 1; + adapter->flags |= IGC_FLAG_QUEUE_PAIRS; + adapter->num_rx_queues = 1; + adapter->num_tx_queues = 1; + adapter->num_q_vectors = 1; + if (!pci_enable_msi(adapter->pdev)) + adapter->flags |= IGC_FLAG_HAS_MSI; +} + +/** + * igc_update_ring_itr - update the dynamic ITR value based on packet size + * @q_vector: pointer to q_vector + * + * Stores a new ITR value based on strictly on packet size. This + * algorithm is less sophisticated than that used in igc_update_itr, + * due to the difficulty of synchronizing statistics across multiple + * receive rings. The divisors and thresholds used by this function + * were determined based on theoretical maximum wire speed and testing + * data, in order to minimize response time while increasing bulk + * throughput. + * NOTE: This function is called only when operating in a multiqueue + * receive environment. + */ +static void igc_update_ring_itr(struct igc_q_vector *q_vector) +{ + struct igc_adapter *adapter = q_vector->adapter; + int new_val = q_vector->itr_val; + int avg_wire_size = 0; + unsigned int packets; + + /* For non-gigabit speeds, just fix the interrupt rate at 4000 + * ints/sec - ITR timer value of 120 ticks. + */ + switch (adapter->link_speed) { + case SPEED_10: + case SPEED_100: + new_val = IGC_4K_ITR; + goto set_itr_val; + default: + break; + } + + packets = q_vector->rx.total_packets; + if (packets) + avg_wire_size = q_vector->rx.total_bytes / packets; + + packets = q_vector->tx.total_packets; + if (packets) + avg_wire_size = max_t(u32, avg_wire_size, + q_vector->tx.total_bytes / packets); + + /* if avg_wire_size isn't set no work was done */ + if (!avg_wire_size) + goto clear_counts; + + /* Add 24 bytes to size to account for CRC, preamble, and gap */ + avg_wire_size += 24; + + /* Don't starve jumbo frames */ + avg_wire_size = min(avg_wire_size, 3000); + + /* Give a little boost to mid-size frames */ + if (avg_wire_size > 300 && avg_wire_size < 1200) + new_val = avg_wire_size / 3; + else + new_val = avg_wire_size / 2; + + /* conservative mode (itr 3) eliminates the lowest_latency setting */ + if (new_val < IGC_20K_ITR && + ((q_vector->rx.ring && adapter->rx_itr_setting == 3) || + (!q_vector->rx.ring && adapter->tx_itr_setting == 3))) + new_val = IGC_20K_ITR; + +set_itr_val: + if (new_val != q_vector->itr_val) { + q_vector->itr_val = new_val; + q_vector->set_itr = 1; + } +clear_counts: + q_vector->rx.total_bytes = 0; + q_vector->rx.total_packets = 0; + q_vector->tx.total_bytes = 0; + q_vector->tx.total_packets = 0; +} + +static void igc_ring_irq_enable(struct igc_q_vector *q_vector) +{ + struct igc_adapter *adapter = q_vector->adapter; + struct igc_hw *hw = &adapter->hw; + + if ((q_vector->rx.ring && (adapter->rx_itr_setting & 3)) || + (!q_vector->rx.ring && (adapter->tx_itr_setting & 3))) { + if (adapter->num_q_vectors == 1) + igc_set_itr(q_vector); + else + igc_update_ring_itr(q_vector); + } + + if (!test_bit(__IGC_DOWN, &adapter->state)) { + if (adapter->msix_entries) + wr32(IGC_EIMS, q_vector->eims_value); + else + igc_irq_enable(adapter); + } +} + +static void igc_add_ring(struct igc_ring *ring, + struct igc_ring_container *head) +{ + head->ring = ring; + head->count++; +} + +/** + * igc_cache_ring_register - Descriptor ring to register mapping + * @adapter: board private structure to initialize + * + * Once we know the feature-set enabled for the device, we'll cache + * the register offset the descriptor ring is assigned to. + */ +static void igc_cache_ring_register(struct igc_adapter *adapter) +{ + int i = 0, j = 0; + + switch (adapter->hw.mac.type) { + case igc_i225: + /* Fall through */ + default: + for (; i < adapter->num_rx_queues; i++) + adapter->rx_ring[i]->reg_idx = i; + for (; j < adapter->num_tx_queues; j++) + adapter->tx_ring[j]->reg_idx = j; + break; + } +} + +/** + * igc_poll - NAPI Rx polling callback + * @napi: napi polling structure + * @budget: count of how many packets we should handle + */ +static int igc_poll(struct napi_struct *napi, int budget) +{ + struct igc_q_vector *q_vector = container_of(napi, + struct igc_q_vector, + napi); + bool clean_complete = true; + int work_done = 0; + + if (q_vector->tx.ring) + clean_complete = igc_clean_tx_irq(q_vector, budget); + + if (q_vector->rx.ring) { + int cleaned = igc_clean_rx_irq(q_vector, budget); + + work_done += cleaned; + if (cleaned >= budget) + clean_complete = false; + } + + /* If all work not completed, return budget and keep polling */ + if (!clean_complete) + return budget; + + /* Exit the polling mode, but don't re-enable interrupts if stack might + * poll us due to busy-polling + */ + if (likely(napi_complete_done(napi, work_done))) + igc_ring_irq_enable(q_vector); + + return min(work_done, budget - 1); +} + +/** + * igc_alloc_q_vector - Allocate memory for a single interrupt vector + * @adapter: board private structure to initialize + * @v_count: q_vectors allocated on adapter, used for ring interleaving + * @v_idx: index of vector in adapter struct + * @txr_count: total number of Tx rings to allocate + * @txr_idx: index of first Tx ring to allocate + * @rxr_count: total number of Rx rings to allocate + * @rxr_idx: index of first Rx ring to allocate + * + * We allocate one q_vector. If allocation fails we return -ENOMEM. + */ +static int igc_alloc_q_vector(struct igc_adapter *adapter, + unsigned int v_count, unsigned int v_idx, + unsigned int txr_count, unsigned int txr_idx, + unsigned int rxr_count, unsigned int rxr_idx) +{ + struct igc_q_vector *q_vector; + struct igc_ring *ring; + int ring_count; + + /* igc only supports 1 Tx and/or 1 Rx queue per vector */ + if (txr_count > 1 || rxr_count > 1) + return -ENOMEM; + + ring_count = txr_count + rxr_count; + + /* allocate q_vector and rings */ + q_vector = adapter->q_vector[v_idx]; + if (!q_vector) + q_vector = kzalloc(struct_size(q_vector, ring, ring_count), + GFP_KERNEL); + else + memset(q_vector, 0, struct_size(q_vector, ring, ring_count)); + if (!q_vector) + return -ENOMEM; + + /* initialize NAPI */ + netif_napi_add(adapter->netdev, &q_vector->napi, + igc_poll, 64); + + /* tie q_vector and adapter together */ + adapter->q_vector[v_idx] = q_vector; + q_vector->adapter = adapter; + + /* initialize work limits */ + q_vector->tx.work_limit = adapter->tx_work_limit; + + /* initialize ITR configuration */ + q_vector->itr_register = adapter->io_addr + IGC_EITR(0); + q_vector->itr_val = IGC_START_ITR; + + /* initialize pointer to rings */ + ring = q_vector->ring; + + /* initialize ITR */ + if (rxr_count) { + /* rx or rx/tx vector */ + if (!adapter->rx_itr_setting || adapter->rx_itr_setting > 3) + q_vector->itr_val = adapter->rx_itr_setting; + } else { + /* tx only vector */ + if (!adapter->tx_itr_setting || adapter->tx_itr_setting > 3) + q_vector->itr_val = adapter->tx_itr_setting; + } + + if (txr_count) { + /* assign generic ring traits */ + ring->dev = &adapter->pdev->dev; + ring->netdev = adapter->netdev; + + /* configure backlink on ring */ + ring->q_vector = q_vector; + + /* update q_vector Tx values */ + igc_add_ring(ring, &q_vector->tx); + + /* apply Tx specific ring traits */ + ring->count = adapter->tx_ring_count; + ring->queue_index = txr_idx; + + /* assign ring to adapter */ + adapter->tx_ring[txr_idx] = ring; + + /* push pointer to next ring */ + ring++; + } + + if (rxr_count) { + /* assign generic ring traits */ + ring->dev = &adapter->pdev->dev; + ring->netdev = adapter->netdev; + + /* configure backlink on ring */ + ring->q_vector = q_vector; + + /* update q_vector Rx values */ + igc_add_ring(ring, &q_vector->rx); + + /* apply Rx specific ring traits */ + ring->count = adapter->rx_ring_count; + ring->queue_index = rxr_idx; + + /* assign ring to adapter */ + adapter->rx_ring[rxr_idx] = ring; + } + + return 0; +} + +/** + * igc_alloc_q_vectors - Allocate memory for interrupt vectors + * @adapter: board private structure to initialize + * + * We allocate one q_vector per queue interrupt. If allocation fails we + * return -ENOMEM. + */ +static int igc_alloc_q_vectors(struct igc_adapter *adapter) +{ + int rxr_remaining = adapter->num_rx_queues; + int txr_remaining = adapter->num_tx_queues; + int rxr_idx = 0, txr_idx = 0, v_idx = 0; + int q_vectors = adapter->num_q_vectors; + int err; + + if (q_vectors >= (rxr_remaining + txr_remaining)) { + for (; rxr_remaining; v_idx++) { + err = igc_alloc_q_vector(adapter, q_vectors, v_idx, + 0, 0, 1, rxr_idx); + + if (err) + goto err_out; + + /* update counts and index */ + rxr_remaining--; + rxr_idx++; + } + } + + for (; v_idx < q_vectors; v_idx++) { + int rqpv = DIV_ROUND_UP(rxr_remaining, q_vectors - v_idx); + int tqpv = DIV_ROUND_UP(txr_remaining, q_vectors - v_idx); + + err = igc_alloc_q_vector(adapter, q_vectors, v_idx, + tqpv, txr_idx, rqpv, rxr_idx); + + if (err) + goto err_out; + + /* update counts and index */ + rxr_remaining -= rqpv; + txr_remaining -= tqpv; + rxr_idx++; + txr_idx++; + } + + return 0; + +err_out: + adapter->num_tx_queues = 0; + adapter->num_rx_queues = 0; + adapter->num_q_vectors = 0; + + while (v_idx--) + igc_free_q_vector(adapter, v_idx); + + return -ENOMEM; +} + +/** + * igc_init_interrupt_scheme - initialize interrupts, allocate queues/vectors + * @adapter: Pointer to adapter structure + * @msix: boolean for MSI-X capability + * + * This function initializes the interrupts and allocates all of the queues. + */ +static int igc_init_interrupt_scheme(struct igc_adapter *adapter, bool msix) +{ + struct pci_dev *pdev = adapter->pdev; + int err = 0; + + igc_set_interrupt_capability(adapter, msix); + + err = igc_alloc_q_vectors(adapter); + if (err) { + dev_err(&pdev->dev, "Unable to allocate memory for vectors\n"); + goto err_alloc_q_vectors; + } + + igc_cache_ring_register(adapter); + + return 0; + +err_alloc_q_vectors: + igc_reset_interrupt_capability(adapter); + return err; +} + +/** + * igc_sw_init - Initialize general software structures (struct igc_adapter) + * @adapter: board private structure to initialize + * + * igc_sw_init initializes the Adapter private data structure. + * Fields are initialized based on PCI device information and + * OS network device settings (MTU size). + */ +static int igc_sw_init(struct igc_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + struct pci_dev *pdev = adapter->pdev; + struct igc_hw *hw = &adapter->hw; + + int size = sizeof(struct igc_mac_addr) * hw->mac.rar_entry_count; + + pci_read_config_word(pdev, PCI_COMMAND, &hw->bus.pci_cmd_word); + + /* set default ring sizes */ + adapter->tx_ring_count = IGC_DEFAULT_TXD; + adapter->rx_ring_count = IGC_DEFAULT_RXD; + + /* set default ITR values */ + adapter->rx_itr_setting = IGC_DEFAULT_ITR; + adapter->tx_itr_setting = IGC_DEFAULT_ITR; + + /* set default work limits */ + adapter->tx_work_limit = IGC_DEFAULT_TX_WORK; + + /* adjust max frame to be at least the size of a standard frame */ + adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN + + VLAN_HLEN; + adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN; + + spin_lock_init(&adapter->nfc_lock); + spin_lock_init(&adapter->stats64_lock); + /* Assume MSI-X interrupts, will be checked during IRQ allocation */ + adapter->flags |= IGC_FLAG_HAS_MSIX; + + adapter->mac_table = kzalloc(size, GFP_ATOMIC); + if (!adapter->mac_table) + return -ENOMEM; + + igc_init_queue_configuration(adapter); + + /* This call may decrease the number of queues */ + if (igc_init_interrupt_scheme(adapter, true)) { + dev_err(&pdev->dev, "Unable to allocate memory for queues\n"); + return -ENOMEM; + } + + /* Explicitly disable IRQ since the NIC can be in any state. */ + igc_irq_disable(adapter); + + set_bit(__IGC_DOWN, &adapter->state); + + return 0; +} + /** * igc_up - Open the interface and prepare it to handle traffic * @adapter: board private structure @@ -2163,18 +3464,6 @@ static void igc_nfc_filter_exit(struct igc_adapter *adapter) spin_unlock(&adapter->nfc_lock); } -static void igc_nfc_filter_restore(struct igc_adapter *adapter) -{ - struct igc_nfc_filter *rule; - - spin_lock(&adapter->nfc_lock); - - hlist_for_each_entry(rule, &adapter->nfc_filter_list, nfc_node) - igc_add_filter(adapter, rule); - - spin_unlock(&adapter->nfc_lock); -} - /** * igc_down - Close the interface * @adapter: board private structure @@ -2398,105 +3687,6 @@ igc_features_check(struct sk_buff *skb, struct net_device *dev, return features; } -/** - * igc_configure - configure the hardware for RX and TX - * @adapter: private board structure - */ -static void igc_configure(struct igc_adapter *adapter) -{ - struct net_device *netdev = adapter->netdev; - int i = 0; - - igc_get_hw_control(adapter); - igc_set_rx_mode(netdev); - - igc_setup_tctl(adapter); - igc_setup_mrqc(adapter); - igc_setup_rctl(adapter); - - igc_nfc_filter_restore(adapter); - igc_configure_tx(adapter); - igc_configure_rx(adapter); - - igc_rx_fifo_flush_base(&adapter->hw); - - /* call igc_desc_unused which always leaves - * at least 1 descriptor unused to make sure - * next_to_use != next_to_clean - */ - for (i = 0; i < adapter->num_rx_queues; i++) { - struct igc_ring *ring = adapter->rx_ring[i]; - - igc_alloc_rx_buffers(ring, igc_desc_unused(ring)); - } -} - -/** - * igc_rar_set_index - Sync RAL[index] and RAH[index] registers with MAC table - * @adapter: address of board private structure - * @index: Index of the RAR entry which need to be synced with MAC table - */ -static void igc_rar_set_index(struct igc_adapter *adapter, u32 index) -{ - u8 *addr = adapter->mac_table[index].addr; - struct igc_hw *hw = &adapter->hw; - u32 rar_low, rar_high; - - /* HW expects these to be in network order when they are plugged - * into the registers which are little endian. In order to guarantee - * that ordering we need to do an leXX_to_cpup here in order to be - * ready for the byteswap that occurs with writel - */ - rar_low = le32_to_cpup((__le32 *)(addr)); - rar_high = le16_to_cpup((__le16 *)(addr + 4)); - - /* Indicate to hardware the Address is Valid. */ - if (adapter->mac_table[index].state & IGC_MAC_STATE_IN_USE) { - if (is_valid_ether_addr(addr)) - rar_high |= IGC_RAH_AV; - - rar_high |= IGC_RAH_POOL_1 << - adapter->mac_table[index].queue; - } - - wr32(IGC_RAL(index), rar_low); - wrfl(); - wr32(IGC_RAH(index), rar_high); - wrfl(); -} - -/* Set default MAC address for the PF in the first RAR entry */ -static void igc_set_default_mac_filter(struct igc_adapter *adapter) -{ - struct igc_mac_addr *mac_table = &adapter->mac_table[0]; - - ether_addr_copy(mac_table->addr, adapter->hw.mac.addr); - mac_table->state = IGC_MAC_STATE_DEFAULT | IGC_MAC_STATE_IN_USE; - - igc_rar_set_index(adapter, 0); -} - -/* If the filter to be added and an already existing filter express - * the same address and address type, it should be possible to only - * override the other configurations, for example the queue to steer - * traffic. - */ -static bool igc_mac_entry_can_be_used(const struct igc_mac_addr *entry, - const u8 *addr, const u8 flags) -{ - if (!(entry->state & IGC_MAC_STATE_IN_USE)) - return true; - - if ((entry->state & IGC_MAC_STATE_SRC_ADDR) != - (flags & IGC_MAC_STATE_SRC_ADDR)) - return false; - - if (!ether_addr_equal(addr, entry->addr)) - return false; - - return true; -} - /* Add a MAC filter for 'addr' directing matching traffic to 'queue', * 'flags' is used to indicate what kind of match is made, match is by * default for the destination address, if matching by source address @@ -2597,159 +3787,20 @@ int igc_del_mac_steering_filter(struct igc_adapter *adapter, IGC_MAC_STATE_QUEUE_STEERING | flags); } -/* Add a MAC filter for 'addr' directing matching traffic to 'queue', - * 'flags' is used to indicate what kind of match is made, match is by - * default for the destination address, if matching by source address - * is desired the flag IGC_MAC_STATE_SRC_ADDR can be used. - */ -static int igc_add_mac_filter(struct igc_adapter *adapter, - const u8 *addr, const u8 queue) +static void igc_tsync_interrupt(struct igc_adapter *adapter) { struct igc_hw *hw = &adapter->hw; - int rar_entries = hw->mac.rar_entry_count; - int i; - - if (is_zero_ether_addr(addr)) - return -EINVAL; - - /* Search for the first empty entry in the MAC table. - * Do not touch entries at the end of the table reserved for the VF MAC - * addresses. - */ - for (i = 0; i < rar_entries; i++) { - if (!igc_mac_entry_can_be_used(&adapter->mac_table[i], - addr, 0)) - continue; - - ether_addr_copy(adapter->mac_table[i].addr, addr); - adapter->mac_table[i].queue = queue; - adapter->mac_table[i].state |= IGC_MAC_STATE_IN_USE; - - igc_rar_set_index(adapter, i); - return i; - } - - return -ENOSPC; -} - -/* Remove a MAC filter for 'addr' directing matching traffic to - * 'queue', 'flags' is used to indicate what kind of match need to be - * removed, match is by default for the destination address, if - * matching by source address is to be removed the flag - * IGC_MAC_STATE_SRC_ADDR can be used. - */ -static int igc_del_mac_filter(struct igc_adapter *adapter, - const u8 *addr, const u8 queue) -{ - struct igc_hw *hw = &adapter->hw; - int rar_entries = hw->mac.rar_entry_count; - int i; - - if (is_zero_ether_addr(addr)) - return -EINVAL; - - /* Search for matching entry in the MAC table based on given address - * and queue. Do not touch entries at the end of the table reserved - * for the VF MAC addresses. - */ - for (i = 0; i < rar_entries; i++) { - if (!(adapter->mac_table[i].state & IGC_MAC_STATE_IN_USE)) - continue; - if (adapter->mac_table[i].state != 0) - continue; - if (adapter->mac_table[i].queue != queue) - continue; - if (!ether_addr_equal(adapter->mac_table[i].addr, addr)) - continue; - - /* When a filter for the default address is "deleted", - * we return it to its initial configuration - */ - if (adapter->mac_table[i].state & IGC_MAC_STATE_DEFAULT) { - adapter->mac_table[i].state = - IGC_MAC_STATE_DEFAULT | IGC_MAC_STATE_IN_USE; - adapter->mac_table[i].queue = 0; - } else { - adapter->mac_table[i].state = 0; - adapter->mac_table[i].queue = 0; - memset(adapter->mac_table[i].addr, 0, ETH_ALEN); - } + u32 tsicr = rd32(IGC_TSICR); + u32 ack = 0; - igc_rar_set_index(adapter, i); - return 0; + if (tsicr & IGC_TSICR_TXTS) { + /* retrieve hardware timestamp */ + schedule_work(&adapter->ptp_tx_work); + ack |= IGC_TSICR_TXTS; } - return -ENOENT; -} - -static int igc_uc_sync(struct net_device *netdev, const unsigned char *addr) -{ - struct igc_adapter *adapter = netdev_priv(netdev); - int ret; - - ret = igc_add_mac_filter(adapter, addr, adapter->num_rx_queues); - - return min_t(int, ret, 0); -} - -static int igc_uc_unsync(struct net_device *netdev, const unsigned char *addr) -{ - struct igc_adapter *adapter = netdev_priv(netdev); - - igc_del_mac_filter(adapter, addr, adapter->num_rx_queues); - - return 0; -} - -/** - * igc_set_rx_mode - Secondary Unicast, Multicast and Promiscuous mode set - * @netdev: network interface device structure - * - * The set_rx_mode entry point is called whenever the unicast or multicast - * address lists or the network interface flags are updated. This routine is - * responsible for configuring the hardware for proper unicast, multicast, - * promiscuous mode, and all-multi behavior. - */ -static void igc_set_rx_mode(struct net_device *netdev) -{ - struct igc_adapter *adapter = netdev_priv(netdev); - struct igc_hw *hw = &adapter->hw; - u32 rctl = 0, rlpml = MAX_JUMBO_FRAME_SIZE; - int count; - - /* Check for Promiscuous and All Multicast modes */ - if (netdev->flags & IFF_PROMISC) { - rctl |= IGC_RCTL_UPE | IGC_RCTL_MPE; - } else { - if (netdev->flags & IFF_ALLMULTI) { - rctl |= IGC_RCTL_MPE; - } else { - /* Write addresses to the MTA, if the attempt fails - * then we should just turn on promiscuous mode so - * that we can at least receive multicast traffic - */ - count = igc_write_mc_addr_list(netdev); - if (count < 0) - rctl |= IGC_RCTL_MPE; - } - } - - /* Write addresses to available RAR registers, if there is not - * sufficient space to store all the addresses then enable - * unicast promiscuous mode - */ - if (__dev_uc_sync(netdev, igc_uc_sync, igc_uc_unsync)) - rctl |= IGC_RCTL_UPE; - - /* update state of unicast and multicast */ - rctl |= rd32(IGC_RCTL) & ~(IGC_RCTL_UPE | IGC_RCTL_MPE); - wr32(IGC_RCTL, rctl); - -#if (PAGE_SIZE < 8192) - if (adapter->max_frame_size <= IGC_MAX_FRAME_BUILD_SKB) - rlpml = IGC_MAX_FRAME_BUILD_SKB; -#endif - wr32(IGC_RLPML, rlpml); + /* acknowledge the interrupts */ + wr32(IGC_TSICR, ack); } /** @@ -2779,114 +3830,28 @@ static irqreturn_t igc_msix_other(int irq, void *data) mod_timer(&adapter->watchdog_timer, jiffies + 1); } + if (icr & IGC_ICR_TS) + igc_tsync_interrupt(adapter); + wr32(IGC_EIMS, adapter->eims_other); return IRQ_HANDLED; } -/** - * igc_write_ivar - configure ivar for given MSI-X vector - * @hw: pointer to the HW structure - * @msix_vector: vector number we are allocating to a given ring - * @index: row index of IVAR register to write within IVAR table - * @offset: column offset of in IVAR, should be multiple of 8 - * - * The IVAR table consists of 2 columns, - * each containing an cause allocation for an Rx and Tx ring, and a - * variable number of rows depending on the number of queues supported. - */ -static void igc_write_ivar(struct igc_hw *hw, int msix_vector, - int index, int offset) -{ - u32 ivar = array_rd32(IGC_IVAR0, index); - - /* clear any bits that are currently set */ - ivar &= ~((u32)0xFF << offset); - - /* write vector and valid bit */ - ivar |= (msix_vector | IGC_IVAR_VALID) << offset; - - array_wr32(IGC_IVAR0, index, ivar); -} - -static void igc_assign_vector(struct igc_q_vector *q_vector, int msix_vector) -{ - struct igc_adapter *adapter = q_vector->adapter; - struct igc_hw *hw = &adapter->hw; - int rx_queue = IGC_N0_QUEUE; - int tx_queue = IGC_N0_QUEUE; - - if (q_vector->rx.ring) - rx_queue = q_vector->rx.ring->reg_idx; - if (q_vector->tx.ring) - tx_queue = q_vector->tx.ring->reg_idx; - - switch (hw->mac.type) { - case igc_i225: - if (rx_queue > IGC_N0_QUEUE) - igc_write_ivar(hw, msix_vector, - rx_queue >> 1, - (rx_queue & 0x1) << 4); - if (tx_queue > IGC_N0_QUEUE) - igc_write_ivar(hw, msix_vector, - tx_queue >> 1, - ((tx_queue & 0x1) << 4) + 8); - q_vector->eims_value = BIT(msix_vector); - break; - default: - WARN_ONCE(hw->mac.type != igc_i225, "Wrong MAC type\n"); - break; - } - - /* add q_vector eims value to global eims_enable_mask */ - adapter->eims_enable_mask |= q_vector->eims_value; - - /* configure q_vector to set itr on first interrupt */ - q_vector->set_itr = 1; -} - -/** - * igc_configure_msix - Configure MSI-X hardware - * @adapter: Pointer to adapter structure - * - * igc_configure_msix sets up the hardware to properly - * generate MSI-X interrupts. - */ -static void igc_configure_msix(struct igc_adapter *adapter) +static void igc_write_itr(struct igc_q_vector *q_vector) { - struct igc_hw *hw = &adapter->hw; - int i, vector = 0; - u32 tmp; - - adapter->eims_enable_mask = 0; - - /* set vector for other causes, i.e. link changes */ - switch (hw->mac.type) { - case igc_i225: - /* Turn on MSI-X capability first, or our settings - * won't stick. And it will take days to debug. - */ - wr32(IGC_GPIE, IGC_GPIE_MSIX_MODE | - IGC_GPIE_PBA | IGC_GPIE_EIAME | - IGC_GPIE_NSICR); - - /* enable msix_other interrupt */ - adapter->eims_other = BIT(vector); - tmp = (vector++ | IGC_IVAR_VALID) << 8; + u32 itr_val = q_vector->itr_val & IGC_QVECTOR_MASK; - wr32(IGC_IVAR_MISC, tmp); - break; - default: - /* do nothing, since nothing else supports MSI-X */ - break; - } /* switch (hw->mac.type) */ + if (!q_vector->set_itr) + return; - adapter->eims_enable_mask |= adapter->eims_other; + if (!itr_val) + itr_val = IGC_ITR_VAL_MASK; - for (i = 0; i < adapter->num_q_vectors; i++) - igc_assign_vector(adapter->q_vector[i], vector++); + itr_val |= IGC_EITR_CNT_IGNR; - wrfl(); + writel(itr_val, q_vector->itr_register); + q_vector->set_itr = 0; } static irqreturn_t igc_msix_ring(int irq, void *data) @@ -2961,49 +3926,6 @@ err_out: } /** - * igc_reset_q_vector - Reset config for interrupt vector - * @adapter: board private structure to initialize - * @v_idx: Index of vector to be reset - * - * If NAPI is enabled it will delete any references to the - * NAPI struct. This is preparation for igc_free_q_vector. - */ -static void igc_reset_q_vector(struct igc_adapter *adapter, int v_idx) -{ - struct igc_q_vector *q_vector = adapter->q_vector[v_idx]; - - /* if we're coming from igc_set_interrupt_capability, the vectors are - * not yet allocated - */ - if (!q_vector) - return; - - if (q_vector->tx.ring) - adapter->tx_ring[q_vector->tx.ring->queue_index] = NULL; - - if (q_vector->rx.ring) - adapter->rx_ring[q_vector->rx.ring->queue_index] = NULL; - - netif_napi_del(&q_vector->napi); -} - -static void igc_reset_interrupt_capability(struct igc_adapter *adapter) -{ - int v_idx = adapter->num_q_vectors; - - if (adapter->msix_entries) { - pci_disable_msix(adapter->pdev); - kfree(adapter->msix_entries); - adapter->msix_entries = NULL; - } else if (adapter->flags & IGC_FLAG_HAS_MSI) { - pci_disable_msi(adapter->pdev); - } - - while (v_idx--) - igc_reset_q_vector(adapter, v_idx); -} - -/** * igc_clear_interrupt_scheme - reset the device to a state of no interrupts * @adapter: Pointer to adapter structure * @@ -3016,48 +3938,6 @@ static void igc_clear_interrupt_scheme(struct igc_adapter *adapter) igc_reset_interrupt_capability(adapter); } -/** - * igc_free_q_vectors - Free memory allocated for interrupt vectors - * @adapter: board private structure to initialize - * - * This function frees the memory allocated to the q_vectors. In addition if - * NAPI is enabled it will delete any references to the NAPI struct prior - * to freeing the q_vector. - */ -static void igc_free_q_vectors(struct igc_adapter *adapter) -{ - int v_idx = adapter->num_q_vectors; - - adapter->num_tx_queues = 0; - adapter->num_rx_queues = 0; - adapter->num_q_vectors = 0; - - while (v_idx--) { - igc_reset_q_vector(adapter, v_idx); - igc_free_q_vector(adapter, v_idx); - } -} - -/** - * igc_free_q_vector - Free memory allocated for specific interrupt vector - * @adapter: board private structure to initialize - * @v_idx: Index of vector to be freed - * - * This function frees the memory allocated to the q_vector. - */ -static void igc_free_q_vector(struct igc_adapter *adapter, int v_idx) -{ - struct igc_q_vector *q_vector = adapter->q_vector[v_idx]; - - adapter->q_vector[v_idx] = NULL; - - /* igc_get_stats64() might access the rings on this vector, - * we must wait a grace period before freeing it. - */ - if (q_vector) - kfree_rcu(q_vector, rcu); -} - /* Need to wait a few seconds after link up to get diagnostic information from * the phy */ @@ -3109,7 +3989,7 @@ bool igc_has_link(struct igc_adapter *adapter) /** * igc_watchdog - Timer Call-back - * @data: pointer to adapter cast into an unsigned long + * @t: timer for the watchdog */ static void igc_watchdog(struct timer_list *t) { @@ -3282,6 +4162,8 @@ no_wait: wr32(IGC_ICS, IGC_ICS_RXDMT0); } + igc_ptp_tx_hang(adapter); + /* Reset the timer */ if (!test_bit(__IGC_DOWN, &adapter->state)) { if (adapter->flags & IGC_FLAG_NEED_LINK_UPDATE) @@ -3294,149 +4176,6 @@ no_wait: } /** - * igc_update_ring_itr - update the dynamic ITR value based on packet size - * @q_vector: pointer to q_vector - * - * Stores a new ITR value based on strictly on packet size. This - * algorithm is less sophisticated than that used in igc_update_itr, - * due to the difficulty of synchronizing statistics across multiple - * receive rings. The divisors and thresholds used by this function - * were determined based on theoretical maximum wire speed and testing - * data, in order to minimize response time while increasing bulk - * throughput. - * NOTE: This function is called only when operating in a multiqueue - * receive environment. - */ -static void igc_update_ring_itr(struct igc_q_vector *q_vector) -{ - struct igc_adapter *adapter = q_vector->adapter; - int new_val = q_vector->itr_val; - int avg_wire_size = 0; - unsigned int packets; - - /* For non-gigabit speeds, just fix the interrupt rate at 4000 - * ints/sec - ITR timer value of 120 ticks. - */ - switch (adapter->link_speed) { - case SPEED_10: - case SPEED_100: - new_val = IGC_4K_ITR; - goto set_itr_val; - default: - break; - } - - packets = q_vector->rx.total_packets; - if (packets) - avg_wire_size = q_vector->rx.total_bytes / packets; - - packets = q_vector->tx.total_packets; - if (packets) - avg_wire_size = max_t(u32, avg_wire_size, - q_vector->tx.total_bytes / packets); - - /* if avg_wire_size isn't set no work was done */ - if (!avg_wire_size) - goto clear_counts; - - /* Add 24 bytes to size to account for CRC, preamble, and gap */ - avg_wire_size += 24; - - /* Don't starve jumbo frames */ - avg_wire_size = min(avg_wire_size, 3000); - - /* Give a little boost to mid-size frames */ - if (avg_wire_size > 300 && avg_wire_size < 1200) - new_val = avg_wire_size / 3; - else - new_val = avg_wire_size / 2; - - /* conservative mode (itr 3) eliminates the lowest_latency setting */ - if (new_val < IGC_20K_ITR && - ((q_vector->rx.ring && adapter->rx_itr_setting == 3) || - (!q_vector->rx.ring && adapter->tx_itr_setting == 3))) - new_val = IGC_20K_ITR; - -set_itr_val: - if (new_val != q_vector->itr_val) { - q_vector->itr_val = new_val; - q_vector->set_itr = 1; - } -clear_counts: - q_vector->rx.total_bytes = 0; - q_vector->rx.total_packets = 0; - q_vector->tx.total_bytes = 0; - q_vector->tx.total_packets = 0; -} - -/** - * igc_update_itr - update the dynamic ITR value based on statistics - * @q_vector: pointer to q_vector - * @ring_container: ring info to update the itr for - * - * Stores a new ITR value based on packets and byte - * counts during the last interrupt. The advantage of per interrupt - * computation is faster updates and more accurate ITR for the current - * traffic pattern. Constants in this function were computed - * based on theoretical maximum wire speed and thresholds were set based - * on testing data as well as attempting to minimize response time - * while increasing bulk throughput. - * NOTE: These calculations are only valid when operating in a single- - * queue environment. - */ -static void igc_update_itr(struct igc_q_vector *q_vector, - struct igc_ring_container *ring_container) -{ - unsigned int packets = ring_container->total_packets; - unsigned int bytes = ring_container->total_bytes; - u8 itrval = ring_container->itr; - - /* no packets, exit with status unchanged */ - if (packets == 0) - return; - - switch (itrval) { - case lowest_latency: - /* handle TSO and jumbo frames */ - if (bytes / packets > 8000) - itrval = bulk_latency; - else if ((packets < 5) && (bytes > 512)) - itrval = low_latency; - break; - case low_latency: /* 50 usec aka 20000 ints/s */ - if (bytes > 10000) { - /* this if handles the TSO accounting */ - if (bytes / packets > 8000) - itrval = bulk_latency; - else if ((packets < 10) || ((bytes / packets) > 1200)) - itrval = bulk_latency; - else if ((packets > 35)) - itrval = lowest_latency; - } else if (bytes / packets > 2000) { - itrval = bulk_latency; - } else if (packets <= 2 && bytes < 512) { - itrval = lowest_latency; - } - break; - case bulk_latency: /* 250 usec aka 4000 ints/s */ - if (bytes > 25000) { - if (packets > 35) - itrval = low_latency; - } else if (bytes < 1500) { - itrval = low_latency; - } - break; - } - - /* clear work counters since we have the values we need */ - ring_container->total_bytes = 0; - ring_container->total_packets = 0; - - /* write updated itr to ring container */ - ring_container->itr = itrval; -} - -/** * igc_intr_msi - Interrupt Handler * @irq: interrupt number * @data: pointer to a network interface device structure @@ -3513,424 +4252,6 @@ static irqreturn_t igc_intr(int irq, void *data) return IRQ_HANDLED; } -static void igc_set_itr(struct igc_q_vector *q_vector) -{ - struct igc_adapter *adapter = q_vector->adapter; - u32 new_itr = q_vector->itr_val; - u8 current_itr = 0; - - /* for non-gigabit speeds, just fix the interrupt rate at 4000 */ - switch (adapter->link_speed) { - case SPEED_10: - case SPEED_100: - current_itr = 0; - new_itr = IGC_4K_ITR; - goto set_itr_now; - default: - break; - } - - igc_update_itr(q_vector, &q_vector->tx); - igc_update_itr(q_vector, &q_vector->rx); - - current_itr = max(q_vector->rx.itr, q_vector->tx.itr); - - /* conservative mode (itr 3) eliminates the lowest_latency setting */ - if (current_itr == lowest_latency && - ((q_vector->rx.ring && adapter->rx_itr_setting == 3) || - (!q_vector->rx.ring && adapter->tx_itr_setting == 3))) - current_itr = low_latency; - - switch (current_itr) { - /* counts and packets in update_itr are dependent on these numbers */ - case lowest_latency: - new_itr = IGC_70K_ITR; /* 70,000 ints/sec */ - break; - case low_latency: - new_itr = IGC_20K_ITR; /* 20,000 ints/sec */ - break; - case bulk_latency: - new_itr = IGC_4K_ITR; /* 4,000 ints/sec */ - break; - default: - break; - } - -set_itr_now: - if (new_itr != q_vector->itr_val) { - /* this attempts to bias the interrupt rate towards Bulk - * by adding intermediate steps when interrupt rate is - * increasing - */ - new_itr = new_itr > q_vector->itr_val ? - max((new_itr * q_vector->itr_val) / - (new_itr + (q_vector->itr_val >> 2)), - new_itr) : new_itr; - /* Don't write the value here; it resets the adapter's - * internal timer, and causes us to delay far longer than - * we should between interrupts. Instead, we write the ITR - * value at the beginning of the next interrupt so the timing - * ends up being correct. - */ - q_vector->itr_val = new_itr; - q_vector->set_itr = 1; - } -} - -static void igc_ring_irq_enable(struct igc_q_vector *q_vector) -{ - struct igc_adapter *adapter = q_vector->adapter; - struct igc_hw *hw = &adapter->hw; - - if ((q_vector->rx.ring && (adapter->rx_itr_setting & 3)) || - (!q_vector->rx.ring && (adapter->tx_itr_setting & 3))) { - if (adapter->num_q_vectors == 1) - igc_set_itr(q_vector); - else - igc_update_ring_itr(q_vector); - } - - if (!test_bit(__IGC_DOWN, &adapter->state)) { - if (adapter->msix_entries) - wr32(IGC_EIMS, q_vector->eims_value); - else - igc_irq_enable(adapter); - } -} - -/** - * igc_poll - NAPI Rx polling callback - * @napi: napi polling structure - * @budget: count of how many packets we should handle - */ -static int igc_poll(struct napi_struct *napi, int budget) -{ - struct igc_q_vector *q_vector = container_of(napi, - struct igc_q_vector, - napi); - bool clean_complete = true; - int work_done = 0; - - if (q_vector->tx.ring) - clean_complete = igc_clean_tx_irq(q_vector, budget); - - if (q_vector->rx.ring) { - int cleaned = igc_clean_rx_irq(q_vector, budget); - - work_done += cleaned; - if (cleaned >= budget) - clean_complete = false; - } - - /* If all work not completed, return budget and keep polling */ - if (!clean_complete) - return budget; - - /* Exit the polling mode, but don't re-enable interrupts if stack might - * poll us due to busy-polling - */ - if (likely(napi_complete_done(napi, work_done))) - igc_ring_irq_enable(q_vector); - - return min(work_done, budget - 1); -} - -/** - * igc_set_interrupt_capability - set MSI or MSI-X if supported - * @adapter: Pointer to adapter structure - * - * Attempt to configure interrupts using the best available - * capabilities of the hardware and kernel. - */ -static void igc_set_interrupt_capability(struct igc_adapter *adapter, - bool msix) -{ - int numvecs, i; - int err; - - if (!msix) - goto msi_only; - adapter->flags |= IGC_FLAG_HAS_MSIX; - - /* Number of supported queues. */ - adapter->num_rx_queues = adapter->rss_queues; - - adapter->num_tx_queues = adapter->rss_queues; - - /* start with one vector for every Rx queue */ - numvecs = adapter->num_rx_queues; - - /* if Tx handler is separate add 1 for every Tx queue */ - if (!(adapter->flags & IGC_FLAG_QUEUE_PAIRS)) - numvecs += adapter->num_tx_queues; - - /* store the number of vectors reserved for queues */ - adapter->num_q_vectors = numvecs; - - /* add 1 vector for link status interrupts */ - numvecs++; - - adapter->msix_entries = kcalloc(numvecs, sizeof(struct msix_entry), - GFP_KERNEL); - - if (!adapter->msix_entries) - return; - - /* populate entry values */ - for (i = 0; i < numvecs; i++) - adapter->msix_entries[i].entry = i; - - err = pci_enable_msix_range(adapter->pdev, - adapter->msix_entries, - numvecs, - numvecs); - if (err > 0) - return; - - kfree(adapter->msix_entries); - adapter->msix_entries = NULL; - - igc_reset_interrupt_capability(adapter); - -msi_only: - adapter->flags &= ~IGC_FLAG_HAS_MSIX; - - adapter->rss_queues = 1; - adapter->flags |= IGC_FLAG_QUEUE_PAIRS; - adapter->num_rx_queues = 1; - adapter->num_tx_queues = 1; - adapter->num_q_vectors = 1; - if (!pci_enable_msi(adapter->pdev)) - adapter->flags |= IGC_FLAG_HAS_MSI; -} - -static void igc_add_ring(struct igc_ring *ring, - struct igc_ring_container *head) -{ - head->ring = ring; - head->count++; -} - -/** - * igc_alloc_q_vector - Allocate memory for a single interrupt vector - * @adapter: board private structure to initialize - * @v_count: q_vectors allocated on adapter, used for ring interleaving - * @v_idx: index of vector in adapter struct - * @txr_count: total number of Tx rings to allocate - * @txr_idx: index of first Tx ring to allocate - * @rxr_count: total number of Rx rings to allocate - * @rxr_idx: index of first Rx ring to allocate - * - * We allocate one q_vector. If allocation fails we return -ENOMEM. - */ -static int igc_alloc_q_vector(struct igc_adapter *adapter, - unsigned int v_count, unsigned int v_idx, - unsigned int txr_count, unsigned int txr_idx, - unsigned int rxr_count, unsigned int rxr_idx) -{ - struct igc_q_vector *q_vector; - struct igc_ring *ring; - int ring_count; - - /* igc only supports 1 Tx and/or 1 Rx queue per vector */ - if (txr_count > 1 || rxr_count > 1) - return -ENOMEM; - - ring_count = txr_count + rxr_count; - - /* allocate q_vector and rings */ - q_vector = adapter->q_vector[v_idx]; - if (!q_vector) - q_vector = kzalloc(struct_size(q_vector, ring, ring_count), - GFP_KERNEL); - else - memset(q_vector, 0, struct_size(q_vector, ring, ring_count)); - if (!q_vector) - return -ENOMEM; - - /* initialize NAPI */ - netif_napi_add(adapter->netdev, &q_vector->napi, - igc_poll, 64); - - /* tie q_vector and adapter together */ - adapter->q_vector[v_idx] = q_vector; - q_vector->adapter = adapter; - - /* initialize work limits */ - q_vector->tx.work_limit = adapter->tx_work_limit; - - /* initialize ITR configuration */ - q_vector->itr_register = adapter->io_addr + IGC_EITR(0); - q_vector->itr_val = IGC_START_ITR; - - /* initialize pointer to rings */ - ring = q_vector->ring; - - /* initialize ITR */ - if (rxr_count) { - /* rx or rx/tx vector */ - if (!adapter->rx_itr_setting || adapter->rx_itr_setting > 3) - q_vector->itr_val = adapter->rx_itr_setting; - } else { - /* tx only vector */ - if (!adapter->tx_itr_setting || adapter->tx_itr_setting > 3) - q_vector->itr_val = adapter->tx_itr_setting; - } - - if (txr_count) { - /* assign generic ring traits */ - ring->dev = &adapter->pdev->dev; - ring->netdev = adapter->netdev; - - /* configure backlink on ring */ - ring->q_vector = q_vector; - - /* update q_vector Tx values */ - igc_add_ring(ring, &q_vector->tx); - - /* apply Tx specific ring traits */ - ring->count = adapter->tx_ring_count; - ring->queue_index = txr_idx; - - /* assign ring to adapter */ - adapter->tx_ring[txr_idx] = ring; - - /* push pointer to next ring */ - ring++; - } - - if (rxr_count) { - /* assign generic ring traits */ - ring->dev = &adapter->pdev->dev; - ring->netdev = adapter->netdev; - - /* configure backlink on ring */ - ring->q_vector = q_vector; - - /* update q_vector Rx values */ - igc_add_ring(ring, &q_vector->rx); - - /* apply Rx specific ring traits */ - ring->count = adapter->rx_ring_count; - ring->queue_index = rxr_idx; - - /* assign ring to adapter */ - adapter->rx_ring[rxr_idx] = ring; - } - - return 0; -} - -/** - * igc_alloc_q_vectors - Allocate memory for interrupt vectors - * @adapter: board private structure to initialize - * - * We allocate one q_vector per queue interrupt. If allocation fails we - * return -ENOMEM. - */ -static int igc_alloc_q_vectors(struct igc_adapter *adapter) -{ - int rxr_remaining = adapter->num_rx_queues; - int txr_remaining = adapter->num_tx_queues; - int rxr_idx = 0, txr_idx = 0, v_idx = 0; - int q_vectors = adapter->num_q_vectors; - int err; - - if (q_vectors >= (rxr_remaining + txr_remaining)) { - for (; rxr_remaining; v_idx++) { - err = igc_alloc_q_vector(adapter, q_vectors, v_idx, - 0, 0, 1, rxr_idx); - - if (err) - goto err_out; - - /* update counts and index */ - rxr_remaining--; - rxr_idx++; - } - } - - for (; v_idx < q_vectors; v_idx++) { - int rqpv = DIV_ROUND_UP(rxr_remaining, q_vectors - v_idx); - int tqpv = DIV_ROUND_UP(txr_remaining, q_vectors - v_idx); - - err = igc_alloc_q_vector(adapter, q_vectors, v_idx, - tqpv, txr_idx, rqpv, rxr_idx); - - if (err) - goto err_out; - - /* update counts and index */ - rxr_remaining -= rqpv; - txr_remaining -= tqpv; - rxr_idx++; - txr_idx++; - } - - return 0; - -err_out: - adapter->num_tx_queues = 0; - adapter->num_rx_queues = 0; - adapter->num_q_vectors = 0; - - while (v_idx--) - igc_free_q_vector(adapter, v_idx); - - return -ENOMEM; -} - -/** - * igc_cache_ring_register - Descriptor ring to register mapping - * @adapter: board private structure to initialize - * - * Once we know the feature-set enabled for the device, we'll cache - * the register offset the descriptor ring is assigned to. - */ -static void igc_cache_ring_register(struct igc_adapter *adapter) -{ - int i = 0, j = 0; - - switch (adapter->hw.mac.type) { - case igc_i225: - /* Fall through */ - default: - for (; i < adapter->num_rx_queues; i++) - adapter->rx_ring[i]->reg_idx = i; - for (; j < adapter->num_tx_queues; j++) - adapter->tx_ring[j]->reg_idx = j; - break; - } -} - -/** - * igc_init_interrupt_scheme - initialize interrupts, allocate queues/vectors - * @adapter: Pointer to adapter structure - * - * This function initializes the interrupts and allocates all of the queues. - */ -static int igc_init_interrupt_scheme(struct igc_adapter *adapter, bool msix) -{ - struct pci_dev *pdev = adapter->pdev; - int err = 0; - - igc_set_interrupt_capability(adapter, msix); - - err = igc_alloc_q_vectors(adapter); - if (err) { - dev_err(&pdev->dev, "Unable to allocate memory for vectors\n"); - goto err_alloc_q_vectors; - } - - igc_cache_ring_register(adapter); - - return 0; - -err_alloc_q_vectors: - igc_reset_interrupt_capability(adapter); - return err; -} - static void igc_free_irq(struct igc_adapter *adapter) { if (adapter->msix_entries) { @@ -3947,62 +4268,6 @@ static void igc_free_irq(struct igc_adapter *adapter) } /** - * igc_irq_disable - Mask off interrupt generation on the NIC - * @adapter: board private structure - */ -static void igc_irq_disable(struct igc_adapter *adapter) -{ - struct igc_hw *hw = &adapter->hw; - - if (adapter->msix_entries) { - u32 regval = rd32(IGC_EIAM); - - wr32(IGC_EIAM, regval & ~adapter->eims_enable_mask); - wr32(IGC_EIMC, adapter->eims_enable_mask); - regval = rd32(IGC_EIAC); - wr32(IGC_EIAC, regval & ~adapter->eims_enable_mask); - } - - wr32(IGC_IAM, 0); - wr32(IGC_IMC, ~0); - wrfl(); - - if (adapter->msix_entries) { - int vector = 0, i; - - synchronize_irq(adapter->msix_entries[vector++].vector); - - for (i = 0; i < adapter->num_q_vectors; i++) - synchronize_irq(adapter->msix_entries[vector++].vector); - } else { - synchronize_irq(adapter->pdev->irq); - } -} - -/** - * igc_irq_enable - Enable default interrupt generation settings - * @adapter: board private structure - */ -static void igc_irq_enable(struct igc_adapter *adapter) -{ - struct igc_hw *hw = &adapter->hw; - - if (adapter->msix_entries) { - u32 ims = IGC_IMS_LSC | IGC_IMS_DOUTSYNC | IGC_IMS_DRSTA; - u32 regval = rd32(IGC_EIAC); - - wr32(IGC_EIAC, regval | adapter->eims_enable_mask); - regval = rd32(IGC_EIAM); - wr32(IGC_EIAM, regval | adapter->eims_enable_mask); - wr32(IGC_EIMS, adapter->eims_enable_mask); - wr32(IGC_IMS, ims); - } else { - wr32(IGC_IMS, IMS_ENABLE_MASK | IGC_IMS_DRSTA); - wr32(IGC_IAM, IMS_ENABLE_MASK | IGC_IMS_DRSTA); - } -} - -/** * igc_request_irq - initialize interrupts * @adapter: Pointer to adapter structure * @@ -4056,25 +4321,10 @@ request_done: return err; } -static void igc_write_itr(struct igc_q_vector *q_vector) -{ - u32 itr_val = q_vector->itr_val & IGC_QVECTOR_MASK; - - if (!q_vector->set_itr) - return; - - if (!itr_val) - itr_val = IGC_ITR_VAL_MASK; - - itr_val |= IGC_EITR_CNT_IGNR; - - writel(itr_val, q_vector->itr_register); - q_vector->set_itr = 0; -} - /** - * igc_open - Called when a network interface is made active + * __igc_open - Called when a network interface is made active * @netdev: network interface device structure + * @resuming: boolean indicating if the device is resuming * * Returns 0 on success, negative value on failure * @@ -4164,8 +4414,9 @@ static int igc_open(struct net_device *netdev) } /** - * igc_close - Disables a network interface + * __igc_close - Disables a network interface * @netdev: network interface device structure + * @suspending: boolean indicating the device is suspending * * Returns 0, this is not allowed to fail * @@ -4199,6 +4450,24 @@ static int igc_close(struct net_device *netdev) return 0; } +/** + * igc_ioctl - Access the hwtstamp interface + * @netdev: network interface device structure + * @ifreq: interface request data + * @cmd: ioctl command + **/ +static int igc_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) +{ + switch (cmd) { + case SIOCGHWTSTAMP: + return igc_ptp_get_ts_config(netdev, ifr); + case SIOCSHWTSTAMP: + return igc_ptp_set_ts_config(netdev, ifr); + default: + return -EOPNOTSUPP; + } +} + static const struct net_device_ops igc_netdev_ops = { .ndo_open = igc_open, .ndo_stop = igc_close, @@ -4210,6 +4479,7 @@ static const struct net_device_ops igc_netdev_ops = { .ndo_fix_features = igc_fix_features, .ndo_set_features = igc_set_features, .ndo_features_check = igc_features_check, + .ndo_do_ioctl = igc_ioctl, }; /* PCIe configuration access */ @@ -4345,32 +4615,26 @@ static int igc_probe(struct pci_dev *pdev, struct net_device *netdev; struct igc_hw *hw; const struct igc_info *ei = igc_info_tbl[ent->driver_data]; - int err; + int err, pci_using_dac; err = pci_enable_device_mem(pdev); if (err) return err; - err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(64)); + pci_using_dac = 0; + err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64)); if (!err) { - err = dma_set_coherent_mask(&pdev->dev, - DMA_BIT_MASK(64)); + pci_using_dac = 1; } else { - err = dma_set_mask(&pdev->dev, DMA_BIT_MASK(32)); + err = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32)); if (err) { - err = dma_set_coherent_mask(&pdev->dev, - DMA_BIT_MASK(32)); - if (err) { - dev_err(&pdev->dev, "igc: Wrong DMA config\n"); - goto err_dma; - } + dev_err(&pdev->dev, + "No usable DMA configuration, aborting\n"); + goto err_dma; } } - err = pci_request_selected_regions(pdev, - pci_select_bars(pdev, - IORESOURCE_MEM), - igc_driver_name); + err = pci_request_mem_regions(pdev, igc_driver_name); if (err) goto err_pci_reg; @@ -4433,6 +4697,9 @@ static int igc_probe(struct pci_dev *pdev, goto err_sw_init; /* Add supported features to the features list*/ + netdev->features |= NETIF_F_SG; + netdev->features |= NETIF_F_TSO; + netdev->features |= NETIF_F_TSO6; netdev->features |= NETIF_F_RXCSUM; netdev->features |= NETIF_F_HW_CSUM; netdev->features |= NETIF_F_SCTP_CRC; @@ -4446,6 +4713,9 @@ static int igc_probe(struct pci_dev *pdev, netdev->hw_features |= NETIF_F_NTUPLE; netdev->hw_features |= netdev->features; + if (pci_using_dac) + netdev->features |= NETIF_F_HIGHDMA; + /* MTU range: 68 - 9216 */ netdev->min_mtu = ETH_MIN_MTU; netdev->max_mtu = MAX_STD_JUMBO_FRAME_SIZE; @@ -4512,6 +4782,9 @@ static int igc_probe(struct pci_dev *pdev, /* carrier off reporting is important to ethtool even BEFORE open */ netif_carrier_off(netdev); + /* do hw tstamp init after resetting */ + igc_ptp_init(adapter); + /* Check if Media Autosense is enabled */ adapter->ei = *ei; @@ -4532,8 +4805,7 @@ err_sw_init: err_ioremap: free_netdev(netdev); err_alloc_etherdev: - pci_release_selected_regions(pdev, - pci_select_bars(pdev, IORESOURCE_MEM)); + pci_release_mem_regions(pdev); err_pci_reg: err_dma: pci_disable_device(pdev); @@ -4554,6 +4826,8 @@ static void igc_remove(struct pci_dev *pdev) struct net_device *netdev = pci_get_drvdata(pdev); struct igc_adapter *adapter = netdev_priv(netdev); + igc_ptp_stop(adapter); + set_bit(__IGC_DOWN, &adapter->state); del_timer_sync(&adapter->watchdog_timer); @@ -4580,105 +4854,216 @@ static void igc_remove(struct pci_dev *pdev) pci_disable_device(pdev); } -static struct pci_driver igc_driver = { - .name = igc_driver_name, - .id_table = igc_pci_tbl, - .probe = igc_probe, - .remove = igc_remove, -}; - -void igc_set_flag_queue_pairs(struct igc_adapter *adapter, - const u32 max_rss_queues) +static int __igc_shutdown(struct pci_dev *pdev, bool *enable_wake, + bool runtime) { - /* Determine if we need to pair queues. */ - /* If rss_queues > half of max_rss_queues, pair the queues in - * order to conserve interrupts due to limited supply. - */ - if (adapter->rss_queues > (max_rss_queues / 2)) - adapter->flags |= IGC_FLAG_QUEUE_PAIRS; + struct net_device *netdev = pci_get_drvdata(pdev); + struct igc_adapter *adapter = netdev_priv(netdev); + u32 wufc = runtime ? IGC_WUFC_LNKC : adapter->wol; + struct igc_hw *hw = &adapter->hw; + u32 ctrl, rctl, status; + bool wake; + + rtnl_lock(); + netif_device_detach(netdev); + + if (netif_running(netdev)) + __igc_close(netdev, true); + + igc_clear_interrupt_scheme(adapter); + rtnl_unlock(); + + status = rd32(IGC_STATUS); + if (status & IGC_STATUS_LU) + wufc &= ~IGC_WUFC_LNKC; + + if (wufc) { + igc_setup_rctl(adapter); + igc_set_rx_mode(netdev); + + /* turn on all-multi mode if wake on multicast is enabled */ + if (wufc & IGC_WUFC_MC) { + rctl = rd32(IGC_RCTL); + rctl |= IGC_RCTL_MPE; + wr32(IGC_RCTL, rctl); + } + + ctrl = rd32(IGC_CTRL); + ctrl |= IGC_CTRL_ADVD3WUC; + wr32(IGC_CTRL, ctrl); + + /* Allow time for pending master requests to run */ + igc_disable_pcie_master(hw); + + wr32(IGC_WUC, IGC_WUC_PME_EN); + wr32(IGC_WUFC, wufc); + } else { + wr32(IGC_WUC, 0); + wr32(IGC_WUFC, 0); + } + + wake = wufc || adapter->en_mng_pt; + if (!wake) + igc_power_down_link(adapter); else - adapter->flags &= ~IGC_FLAG_QUEUE_PAIRS; -} + igc_power_up_link(adapter); -unsigned int igc_get_max_rss_queues(struct igc_adapter *adapter) -{ - unsigned int max_rss_queues; + if (enable_wake) + *enable_wake = wake; - /* Determine the maximum number of RSS queues supported. */ - max_rss_queues = IGC_MAX_RX_QUEUES; + /* Release control of h/w to f/w. If f/w is AMT enabled, this + * would have already happened in close and is redundant. + */ + igc_release_hw_control(adapter); - return max_rss_queues; + pci_disable_device(pdev); + + return 0; } -static void igc_init_queue_configuration(struct igc_adapter *adapter) +#ifdef CONFIG_PM +static int __maybe_unused igc_runtime_suspend(struct device *dev) { - u32 max_rss_queues; - - max_rss_queues = igc_get_max_rss_queues(adapter); - adapter->rss_queues = min_t(u32, max_rss_queues, num_online_cpus()); - - igc_set_flag_queue_pairs(adapter, max_rss_queues); + return __igc_shutdown(to_pci_dev(dev), NULL, 1); } -/** - * igc_sw_init - Initialize general software structures (struct igc_adapter) - * @adapter: board private structure to initialize - * - * igc_sw_init initializes the Adapter private data structure. - * Fields are initialized based on PCI device information and - * OS network device settings (MTU size). - */ -static int igc_sw_init(struct igc_adapter *adapter) +static void igc_deliver_wake_packet(struct net_device *netdev) { - struct net_device *netdev = adapter->netdev; - struct pci_dev *pdev = adapter->pdev; + struct igc_adapter *adapter = netdev_priv(netdev); struct igc_hw *hw = &adapter->hw; + struct sk_buff *skb; + u32 wupl; - int size = sizeof(struct igc_mac_addr) * hw->mac.rar_entry_count; + wupl = rd32(IGC_WUPL) & IGC_WUPL_MASK; - pci_read_config_word(pdev, PCI_COMMAND, &hw->bus.pci_cmd_word); + /* WUPM stores only the first 128 bytes of the wake packet. + * Read the packet only if we have the whole thing. + */ + if (wupl == 0 || wupl > IGC_WUPM_BYTES) + return; - /* set default ring sizes */ - adapter->tx_ring_count = IGC_DEFAULT_TXD; - adapter->rx_ring_count = IGC_DEFAULT_RXD; + skb = netdev_alloc_skb_ip_align(netdev, IGC_WUPM_BYTES); + if (!skb) + return; - /* set default ITR values */ - adapter->rx_itr_setting = IGC_DEFAULT_ITR; - adapter->tx_itr_setting = IGC_DEFAULT_ITR; + skb_put(skb, wupl); - /* set default work limits */ - adapter->tx_work_limit = IGC_DEFAULT_TX_WORK; + /* Ensure reads are 32-bit aligned */ + wupl = roundup(wupl, 4); - /* adjust max frame to be at least the size of a standard frame */ - adapter->max_frame_size = netdev->mtu + ETH_HLEN + ETH_FCS_LEN + - VLAN_HLEN; - adapter->min_frame_size = ETH_ZLEN + ETH_FCS_LEN; + memcpy_fromio(skb->data, hw->hw_addr + IGC_WUPM_REG(0), wupl); - spin_lock_init(&adapter->nfc_lock); - spin_lock_init(&adapter->stats64_lock); - /* Assume MSI-X interrupts, will be checked during IRQ allocation */ - adapter->flags |= IGC_FLAG_HAS_MSIX; + skb->protocol = eth_type_trans(skb, netdev); + netif_rx(skb); +} - adapter->mac_table = kzalloc(size, GFP_ATOMIC); - if (!adapter->mac_table) - return -ENOMEM; +static int __maybe_unused igc_resume(struct device *dev) +{ + struct pci_dev *pdev = to_pci_dev(dev); + struct net_device *netdev = pci_get_drvdata(pdev); + struct igc_adapter *adapter = netdev_priv(netdev); + struct igc_hw *hw = &adapter->hw; + u32 err, val; - igc_init_queue_configuration(adapter); + pci_set_power_state(pdev, PCI_D0); + pci_restore_state(pdev); + pci_save_state(pdev); + + if (!pci_device_is_present(pdev)) + return -ENODEV; + err = pci_enable_device_mem(pdev); + if (err) { + dev_err(&pdev->dev, + "igc: Cannot enable PCI device from suspend\n"); + return err; + } + pci_set_master(pdev); + + pci_enable_wake(pdev, PCI_D3hot, 0); + pci_enable_wake(pdev, PCI_D3cold, 0); - /* This call may decrease the number of queues */ if (igc_init_interrupt_scheme(adapter, true)) { dev_err(&pdev->dev, "Unable to allocate memory for queues\n"); return -ENOMEM; } - /* Explicitly disable IRQ since the NIC can be in any state. */ - igc_irq_disable(adapter); + igc_reset(adapter); - set_bit(__IGC_DOWN, &adapter->state); + /* let the f/w know that the h/w is now under the control of the + * driver. + */ + igc_get_hw_control(adapter); - return 0; + val = rd32(IGC_WUS); + if (val & WAKE_PKT_WUS) + igc_deliver_wake_packet(netdev); + + wr32(IGC_WUS, ~0); + + rtnl_lock(); + if (!err && netif_running(netdev)) + err = __igc_open(netdev, true); + + if (!err) + netif_device_attach(netdev); + rtnl_unlock(); + + return err; +} + +static int __maybe_unused igc_runtime_resume(struct device *dev) +{ + return igc_resume(dev); +} + +static int __maybe_unused igc_suspend(struct device *dev) +{ + return __igc_shutdown(to_pci_dev(dev), NULL, 0); } +static int __maybe_unused igc_runtime_idle(struct device *dev) +{ + struct net_device *netdev = dev_get_drvdata(dev); + struct igc_adapter *adapter = netdev_priv(netdev); + + if (!igc_has_link(adapter)) + pm_schedule_suspend(dev, MSEC_PER_SEC * 5); + + return -EBUSY; +} +#endif /* CONFIG_PM */ + +static void igc_shutdown(struct pci_dev *pdev) +{ + bool wake; + + __igc_shutdown(pdev, &wake, 0); + + if (system_state == SYSTEM_POWER_OFF) { + pci_wake_from_d3(pdev, wake); + pci_set_power_state(pdev, PCI_D3hot); + } +} + +#ifdef CONFIG_PM +static const struct dev_pm_ops igc_pm_ops = { + SET_SYSTEM_SLEEP_PM_OPS(igc_suspend, igc_resume) + SET_RUNTIME_PM_OPS(igc_runtime_suspend, igc_runtime_resume, + igc_runtime_idle) +}; +#endif + +static struct pci_driver igc_driver = { + .name = igc_driver_name, + .id_table = igc_pci_tbl, + .probe = igc_probe, + .remove = igc_remove, +#ifdef CONFIG_PM + .driver.pm = &igc_pm_ops, +#endif + .shutdown = igc_shutdown, +}; + /** * igc_reinit_queues - return error * @adapter: pointer to adapter structure diff --git a/drivers/net/ethernet/intel/igc/igc_phy.c b/drivers/net/ethernet/intel/igc/igc_phy.c index f4b05af0dd2f..8e1799508edc 100644 --- a/drivers/net/ethernet/intel/igc/igc_phy.c +++ b/drivers/net/ethernet/intel/igc/igc_phy.c @@ -173,6 +173,7 @@ s32 igc_check_downshift(struct igc_hw *hw) s32 igc_phy_hw_reset(struct igc_hw *hw) { struct igc_phy_info *phy = &hw->phy; + u32 phpm = 0, timeout = 10000; s32 ret_val; u32 ctrl; @@ -186,6 +187,8 @@ s32 igc_phy_hw_reset(struct igc_hw *hw) if (ret_val) goto out; + phpm = rd32(IGC_I225_PHPM); + ctrl = rd32(IGC_CTRL); wr32(IGC_CTRL, ctrl | IGC_CTRL_PHY_RST); wrfl(); @@ -195,7 +198,18 @@ s32 igc_phy_hw_reset(struct igc_hw *hw) wr32(IGC_CTRL, ctrl); wrfl(); - usleep_range(1500, 2000); + /* SW should guarantee 100us for the completion of the PHY reset */ + usleep_range(100, 150); + do { + phpm = rd32(IGC_I225_PHPM); + timeout--; + udelay(1); + } while (!(phpm & IGC_PHY_RST_COMP) && timeout); + + if (!timeout) + hw_dbg("Timeout is expired after a phy reset\n"); + + usleep_range(100, 150); phy->ops.release(hw); diff --git a/drivers/net/ethernet/intel/igc/igc_ptp.c b/drivers/net/ethernet/intel/igc/igc_ptp.c new file mode 100644 index 000000000000..693506587198 --- /dev/null +++ b/drivers/net/ethernet/intel/igc/igc_ptp.c @@ -0,0 +1,716 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (c) 2019 Intel Corporation */ + +#include "igc.h" + +#include <linux/module.h> +#include <linux/device.h> +#include <linux/pci.h> +#include <linux/ptp_classify.h> +#include <linux/clocksource.h> + +#define INCVALUE_MASK 0x7fffffff +#define ISGN 0x80000000 + +#define IGC_SYSTIM_OVERFLOW_PERIOD (HZ * 60 * 9) +#define IGC_PTP_TX_TIMEOUT (HZ * 15) + +/* SYSTIM read access for I225 */ +static void igc_ptp_read_i225(struct igc_adapter *adapter, + struct timespec64 *ts) +{ + struct igc_hw *hw = &adapter->hw; + u32 sec, nsec; + + /* The timestamp latches on lowest register read. For I210/I211, the + * lowest register is SYSTIMR. Since we only need to provide nanosecond + * resolution, we can ignore it. + */ + rd32(IGC_SYSTIMR); + nsec = rd32(IGC_SYSTIML); + sec = rd32(IGC_SYSTIMH); + + ts->tv_sec = sec; + ts->tv_nsec = nsec; +} + +static void igc_ptp_write_i225(struct igc_adapter *adapter, + const struct timespec64 *ts) +{ + struct igc_hw *hw = &adapter->hw; + + /* Writing the SYSTIMR register is not necessary as it only + * provides sub-nanosecond resolution. + */ + wr32(IGC_SYSTIML, ts->tv_nsec); + wr32(IGC_SYSTIMH, ts->tv_sec); +} + +static int igc_ptp_adjfine_i225(struct ptp_clock_info *ptp, long scaled_ppm) +{ + struct igc_adapter *igc = container_of(ptp, struct igc_adapter, + ptp_caps); + struct igc_hw *hw = &igc->hw; + int neg_adj = 0; + u64 rate; + u32 inca; + + if (scaled_ppm < 0) { + neg_adj = 1; + scaled_ppm = -scaled_ppm; + } + rate = scaled_ppm; + rate <<= 14; + rate = div_u64(rate, 78125); + + inca = rate & INCVALUE_MASK; + if (neg_adj) + inca |= ISGN; + + wr32(IGC_TIMINCA, inca); + + return 0; +} + +static int igc_ptp_adjtime_i225(struct ptp_clock_info *ptp, s64 delta) +{ + struct igc_adapter *igc = container_of(ptp, struct igc_adapter, + ptp_caps); + struct timespec64 now, then = ns_to_timespec64(delta); + unsigned long flags; + + spin_lock_irqsave(&igc->tmreg_lock, flags); + + igc_ptp_read_i225(igc, &now); + now = timespec64_add(now, then); + igc_ptp_write_i225(igc, (const struct timespec64 *)&now); + + spin_unlock_irqrestore(&igc->tmreg_lock, flags); + + return 0; +} + +static int igc_ptp_gettimex64_i225(struct ptp_clock_info *ptp, + struct timespec64 *ts, + struct ptp_system_timestamp *sts) +{ + struct igc_adapter *igc = container_of(ptp, struct igc_adapter, + ptp_caps); + struct igc_hw *hw = &igc->hw; + unsigned long flags; + + spin_lock_irqsave(&igc->tmreg_lock, flags); + + ptp_read_system_prets(sts); + rd32(IGC_SYSTIMR); + ptp_read_system_postts(sts); + ts->tv_nsec = rd32(IGC_SYSTIML); + ts->tv_sec = rd32(IGC_SYSTIMH); + + spin_unlock_irqrestore(&igc->tmreg_lock, flags); + + return 0; +} + +static int igc_ptp_settime_i225(struct ptp_clock_info *ptp, + const struct timespec64 *ts) +{ + struct igc_adapter *igc = container_of(ptp, struct igc_adapter, + ptp_caps); + unsigned long flags; + + spin_lock_irqsave(&igc->tmreg_lock, flags); + + igc_ptp_write_i225(igc, ts); + + spin_unlock_irqrestore(&igc->tmreg_lock, flags); + + return 0; +} + +static int igc_ptp_feature_enable_i225(struct ptp_clock_info *ptp, + struct ptp_clock_request *rq, int on) +{ + return -EOPNOTSUPP; +} + +/** + * igc_ptp_systim_to_hwtstamp - convert system time value to HW timestamp + * @adapter: board private structure + * @hwtstamps: timestamp structure to update + * @systim: unsigned 64bit system time value + * + * We need to convert the system time value stored in the RX/TXSTMP registers + * into a hwtstamp which can be used by the upper level timestamping functions. + **/ +static void igc_ptp_systim_to_hwtstamp(struct igc_adapter *adapter, + struct skb_shared_hwtstamps *hwtstamps, + u64 systim) +{ + switch (adapter->hw.mac.type) { + case igc_i225: + memset(hwtstamps, 0, sizeof(*hwtstamps)); + /* Upper 32 bits contain s, lower 32 bits contain ns. */ + hwtstamps->hwtstamp = ktime_set(systim >> 32, + systim & 0xFFFFFFFF); + break; + default: + break; + } +} + +/** + * igc_ptp_rx_pktstamp - retrieve Rx per packet timestamp + * @q_vector: Pointer to interrupt specific structure + * @va: Pointer to address containing Rx buffer + * @skb: Buffer containing timestamp and packet + * + * This function is meant to retrieve the first timestamp from the + * first buffer of an incoming frame. The value is stored in little + * endian format starting on byte 0. There's a second timestamp + * starting on byte 8. + **/ +void igc_ptp_rx_pktstamp(struct igc_q_vector *q_vector, void *va, + struct sk_buff *skb) +{ + struct igc_adapter *adapter = q_vector->adapter; + __le64 *regval = (__le64 *)va; + int adjust = 0; + + /* The timestamp is recorded in little endian format. + * DWORD: | 0 | 1 | 2 | 3 + * Field: | Timer0 Low | Timer0 High | Timer1 Low | Timer1 High + */ + igc_ptp_systim_to_hwtstamp(adapter, skb_hwtstamps(skb), + le64_to_cpu(regval[0])); + + /* adjust timestamp for the RX latency based on link speed */ + if (adapter->hw.mac.type == igc_i225) { + switch (adapter->link_speed) { + case SPEED_10: + adjust = IGC_I225_RX_LATENCY_10; + break; + case SPEED_100: + adjust = IGC_I225_RX_LATENCY_100; + break; + case SPEED_1000: + adjust = IGC_I225_RX_LATENCY_1000; + break; + case SPEED_2500: + adjust = IGC_I225_RX_LATENCY_2500; + break; + } + } + skb_hwtstamps(skb)->hwtstamp = + ktime_sub_ns(skb_hwtstamps(skb)->hwtstamp, adjust); +} + +/** + * igc_ptp_rx_rgtstamp - retrieve Rx timestamp stored in register + * @q_vector: Pointer to interrupt specific structure + * @skb: Buffer containing timestamp and packet + * + * This function is meant to retrieve a timestamp from the internal registers + * of the adapter and store it in the skb. + */ +void igc_ptp_rx_rgtstamp(struct igc_q_vector *q_vector, + struct sk_buff *skb) +{ + struct igc_adapter *adapter = q_vector->adapter; + struct igc_hw *hw = &adapter->hw; + u64 regval; + + /* If this bit is set, then the RX registers contain the time + * stamp. No other packet will be time stamped until we read + * these registers, so read the registers to make them + * available again. Because only one packet can be time + * stamped at a time, we know that the register values must + * belong to this one here and therefore we don't need to + * compare any of the additional attributes stored for it. + * + * If nothing went wrong, then it should have a shared + * tx_flags that we can turn into a skb_shared_hwtstamps. + */ + if (!(rd32(IGC_TSYNCRXCTL) & IGC_TSYNCRXCTL_VALID)) + return; + + regval = rd32(IGC_RXSTMPL); + regval |= (u64)rd32(IGC_RXSTMPH) << 32; + + igc_ptp_systim_to_hwtstamp(adapter, skb_hwtstamps(skb), regval); + + /* Update the last_rx_timestamp timer in order to enable watchdog check + * for error case of latched timestamp on a dropped packet. + */ + adapter->last_rx_timestamp = jiffies; +} + +/** + * igc_ptp_enable_tstamp_rxqueue - Enable RX timestamp for a queue + * @rx_ring: Pointer to RX queue + * @timer: Index for timer + * + * This function enables RX timestamping for a queue, and selects + * which 1588 timer will provide the timestamp. + */ +static void igc_ptp_enable_tstamp_rxqueue(struct igc_adapter *adapter, + struct igc_ring *rx_ring, u8 timer) +{ + struct igc_hw *hw = &adapter->hw; + int reg_idx = rx_ring->reg_idx; + u32 srrctl = rd32(IGC_SRRCTL(reg_idx)); + + srrctl |= IGC_SRRCTL_TIMESTAMP; + srrctl |= IGC_SRRCTL_TIMER1SEL(timer); + srrctl |= IGC_SRRCTL_TIMER0SEL(timer); + + wr32(IGC_SRRCTL(reg_idx), srrctl); +} + +static void igc_ptp_enable_tstamp_all_rxqueues(struct igc_adapter *adapter, + u8 timer) +{ + int i; + + for (i = 0; i < adapter->num_rx_queues; i++) { + struct igc_ring *ring = adapter->rx_ring[i]; + + igc_ptp_enable_tstamp_rxqueue(adapter, ring, timer); + } +} + +/** + * igc_ptp_set_timestamp_mode - setup hardware for timestamping + * @adapter: networking device structure + * @config: hwtstamp configuration + * + * Outgoing time stamping can be enabled and disabled. Play nice and + * disable it when requested, although it shouldn't case any overhead + * when no packet needs it. At most one packet in the queue may be + * marked for time stamping, otherwise it would be impossible to tell + * for sure to which packet the hardware time stamp belongs. + * + * Incoming time stamping has to be configured via the hardware + * filters. Not all combinations are supported, in particular event + * type has to be specified. Matching the kind of event packet is + * not supported, with the exception of "all V2 events regardless of + * level 2 or 4". + * + */ +static int igc_ptp_set_timestamp_mode(struct igc_adapter *adapter, + struct hwtstamp_config *config) +{ + u32 tsync_tx_ctl = IGC_TSYNCTXCTL_ENABLED; + u32 tsync_rx_ctl = IGC_TSYNCRXCTL_ENABLED; + struct igc_hw *hw = &adapter->hw; + u32 tsync_rx_cfg = 0; + bool is_l4 = false; + bool is_l2 = false; + u32 regval; + + /* reserved for future extensions */ + if (config->flags) + return -EINVAL; + + switch (config->tx_type) { + case HWTSTAMP_TX_OFF: + tsync_tx_ctl = 0; + case HWTSTAMP_TX_ON: + break; + default: + return -ERANGE; + } + + switch (config->rx_filter) { + case HWTSTAMP_FILTER_NONE: + tsync_rx_ctl = 0; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_SYNC: + tsync_rx_ctl |= IGC_TSYNCRXCTL_TYPE_L4_V1; + tsync_rx_cfg = IGC_TSYNCRXCFG_PTP_V1_SYNC_MESSAGE; + is_l4 = true; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ: + tsync_rx_ctl |= IGC_TSYNCRXCTL_TYPE_L4_V1; + tsync_rx_cfg = IGC_TSYNCRXCFG_PTP_V1_DELAY_REQ_MESSAGE; + is_l4 = true; + break; + case HWTSTAMP_FILTER_PTP_V2_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L2_EVENT: + case HWTSTAMP_FILTER_PTP_V2_L4_EVENT: + case HWTSTAMP_FILTER_PTP_V2_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L2_SYNC: + case HWTSTAMP_FILTER_PTP_V2_L4_SYNC: + case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ: + case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ: + tsync_rx_ctl |= IGC_TSYNCRXCTL_TYPE_EVENT_V2; + config->rx_filter = HWTSTAMP_FILTER_PTP_V2_EVENT; + is_l2 = true; + is_l4 = true; + break; + case HWTSTAMP_FILTER_PTP_V1_L4_EVENT: + case HWTSTAMP_FILTER_NTP_ALL: + case HWTSTAMP_FILTER_ALL: + tsync_rx_ctl |= IGC_TSYNCRXCTL_TYPE_ALL; + config->rx_filter = HWTSTAMP_FILTER_ALL; + break; + /* fall through */ + default: + config->rx_filter = HWTSTAMP_FILTER_NONE; + return -ERANGE; + } + + /* Per-packet timestamping only works if all packets are + * timestamped, so enable timestamping in all packets as long + * as one Rx filter was configured. + */ + if (tsync_rx_ctl) { + tsync_rx_ctl = IGC_TSYNCRXCTL_ENABLED; + tsync_rx_ctl |= IGC_TSYNCRXCTL_TYPE_ALL; + tsync_rx_ctl |= IGC_TSYNCRXCTL_RXSYNSIG; + config->rx_filter = HWTSTAMP_FILTER_ALL; + is_l2 = true; + is_l4 = true; + + if (hw->mac.type == igc_i225) { + regval = rd32(IGC_RXPBS); + regval |= IGC_RXPBS_CFG_TS_EN; + wr32(IGC_RXPBS, regval); + + /* FIXME: For now, only support retrieving RX + * timestamps from timer 0 + */ + igc_ptp_enable_tstamp_all_rxqueues(adapter, 0); + } + } + + if (tsync_tx_ctl) { + tsync_tx_ctl = IGC_TSYNCTXCTL_ENABLED; + tsync_tx_ctl |= IGC_TSYNCTXCTL_TXSYNSIG; + } + + /* enable/disable TX */ + regval = rd32(IGC_TSYNCTXCTL); + regval &= ~IGC_TSYNCTXCTL_ENABLED; + regval |= tsync_tx_ctl; + wr32(IGC_TSYNCTXCTL, regval); + + /* enable/disable RX */ + regval = rd32(IGC_TSYNCRXCTL); + regval &= ~(IGC_TSYNCRXCTL_ENABLED | IGC_TSYNCRXCTL_TYPE_MASK); + regval |= tsync_rx_ctl; + wr32(IGC_TSYNCRXCTL, regval); + + /* define which PTP packets are time stamped */ + wr32(IGC_TSYNCRXCFG, tsync_rx_cfg); + + /* define ethertype filter for timestamped packets */ + if (is_l2) + wr32(IGC_ETQF(3), + (IGC_ETQF_FILTER_ENABLE | /* enable filter */ + IGC_ETQF_1588 | /* enable timestamping */ + ETH_P_1588)); /* 1588 eth protocol type */ + else + wr32(IGC_ETQF(3), 0); + + /* L4 Queue Filter[3]: filter by destination port and protocol */ + if (is_l4) { + u32 ftqf = (IPPROTO_UDP /* UDP */ + | IGC_FTQF_VF_BP /* VF not compared */ + | IGC_FTQF_1588_TIME_STAMP /* Enable Timestamp */ + | IGC_FTQF_MASK); /* mask all inputs */ + ftqf &= ~IGC_FTQF_MASK_PROTO_BP; /* enable protocol check */ + + wr32(IGC_IMIR(3), htons(PTP_EV_PORT)); + wr32(IGC_IMIREXT(3), + (IGC_IMIREXT_SIZE_BP | IGC_IMIREXT_CTRL_BP)); + wr32(IGC_FTQF(3), ftqf); + } else { + wr32(IGC_FTQF(3), IGC_FTQF_MASK); + } + wrfl(); + + /* clear TX/RX time stamp registers, just to be sure */ + regval = rd32(IGC_TXSTMPL); + regval = rd32(IGC_TXSTMPH); + regval = rd32(IGC_RXSTMPL); + regval = rd32(IGC_RXSTMPH); + + return 0; +} + +void igc_ptp_tx_hang(struct igc_adapter *adapter) +{ + bool timeout = time_is_before_jiffies(adapter->ptp_tx_start + + IGC_PTP_TX_TIMEOUT); + struct igc_hw *hw = &adapter->hw; + + if (!adapter->ptp_tx_skb) + return; + + if (!test_bit(__IGC_PTP_TX_IN_PROGRESS, &adapter->state)) + return; + + /* If we haven't received a timestamp within the timeout, it is + * reasonable to assume that it will never occur, so we can unlock the + * timestamp bit when this occurs. + */ + if (timeout) { + cancel_work_sync(&adapter->ptp_tx_work); + dev_kfree_skb_any(adapter->ptp_tx_skb); + adapter->ptp_tx_skb = NULL; + clear_bit_unlock(__IGC_PTP_TX_IN_PROGRESS, &adapter->state); + adapter->tx_hwtstamp_timeouts++; + /* Clear the Tx valid bit in TSYNCTXCTL register to enable + * interrupt + */ + rd32(IGC_TXSTMPH); + dev_warn(&adapter->pdev->dev, "clearing Tx timestamp hang\n"); + } +} + +/** + * igc_ptp_tx_hwtstamp - utility function which checks for TX time stamp + * @adapter: Board private structure + * + * If we were asked to do hardware stamping and such a time stamp is + * available, then it must have been for this skb here because we only + * allow only one such packet into the queue. + */ +static void igc_ptp_tx_hwtstamp(struct igc_adapter *adapter) +{ + struct sk_buff *skb = adapter->ptp_tx_skb; + struct skb_shared_hwtstamps shhwtstamps; + struct igc_hw *hw = &adapter->hw; + u64 regval; + + regval = rd32(IGC_TXSTMPL); + regval |= (u64)rd32(IGC_TXSTMPH) << 32; + igc_ptp_systim_to_hwtstamp(adapter, &shhwtstamps, regval); + + /* Clear the lock early before calling skb_tstamp_tx so that + * applications are not woken up before the lock bit is clear. We use + * a copy of the skb pointer to ensure other threads can't change it + * while we're notifying the stack. + */ + adapter->ptp_tx_skb = NULL; + clear_bit_unlock(__IGC_PTP_TX_IN_PROGRESS, &adapter->state); + + /* Notify the stack and free the skb after we've unlocked */ + skb_tstamp_tx(skb, &shhwtstamps); + dev_kfree_skb_any(skb); +} + +/** + * igc_ptp_tx_work + * @work: pointer to work struct + * + * This work function polls the TSYNCTXCTL valid bit to determine when a + * timestamp has been taken for the current stored skb. + */ +void igc_ptp_tx_work(struct work_struct *work) +{ + struct igc_adapter *adapter = container_of(work, struct igc_adapter, + ptp_tx_work); + struct igc_hw *hw = &adapter->hw; + u32 tsynctxctl; + + if (!adapter->ptp_tx_skb) + return; + + if (time_is_before_jiffies(adapter->ptp_tx_start + + IGC_PTP_TX_TIMEOUT)) { + dev_kfree_skb_any(adapter->ptp_tx_skb); + adapter->ptp_tx_skb = NULL; + clear_bit_unlock(__IGC_PTP_TX_IN_PROGRESS, &adapter->state); + adapter->tx_hwtstamp_timeouts++; + /* Clear the tx valid bit in TSYNCTXCTL register to enable + * interrupt + */ + rd32(IGC_TXSTMPH); + dev_warn(&adapter->pdev->dev, "clearing Tx timestamp hang\n"); + return; + } + + tsynctxctl = rd32(IGC_TSYNCTXCTL); + if (tsynctxctl & IGC_TSYNCTXCTL_VALID) + igc_ptp_tx_hwtstamp(adapter); + else + /* reschedule to check later */ + schedule_work(&adapter->ptp_tx_work); +} + +/** + * igc_ptp_set_ts_config - set hardware time stamping config + * @netdev: network interface device structure + * @ifreq: interface request data + * + **/ +int igc_ptp_set_ts_config(struct net_device *netdev, struct ifreq *ifr) +{ + struct igc_adapter *adapter = netdev_priv(netdev); + struct hwtstamp_config config; + int err; + + if (copy_from_user(&config, ifr->ifr_data, sizeof(config))) + return -EFAULT; + + err = igc_ptp_set_timestamp_mode(adapter, &config); + if (err) + return err; + + /* save these settings for future reference */ + memcpy(&adapter->tstamp_config, &config, + sizeof(adapter->tstamp_config)); + + return copy_to_user(ifr->ifr_data, &config, sizeof(config)) ? + -EFAULT : 0; +} + +/** + * igc_ptp_get_ts_config - get hardware time stamping config + * @netdev: network interface device structure + * @ifreq: interface request data + * + * Get the hwtstamp_config settings to return to the user. Rather than attempt + * to deconstruct the settings from the registers, just return a shadow copy + * of the last known settings. + **/ +int igc_ptp_get_ts_config(struct net_device *netdev, struct ifreq *ifr) +{ + struct igc_adapter *adapter = netdev_priv(netdev); + struct hwtstamp_config *config = &adapter->tstamp_config; + + return copy_to_user(ifr->ifr_data, config, sizeof(*config)) ? + -EFAULT : 0; +} + +/** + * igc_ptp_init - Initialize PTP functionality + * @adapter: Board private structure + * + * This function is called at device probe to initialize the PTP + * functionality. + */ +void igc_ptp_init(struct igc_adapter *adapter) +{ + struct net_device *netdev = adapter->netdev; + struct igc_hw *hw = &adapter->hw; + + switch (hw->mac.type) { + case igc_i225: + snprintf(adapter->ptp_caps.name, 16, "%pm", netdev->dev_addr); + adapter->ptp_caps.owner = THIS_MODULE; + adapter->ptp_caps.max_adj = 62499999; + adapter->ptp_caps.adjfine = igc_ptp_adjfine_i225; + adapter->ptp_caps.adjtime = igc_ptp_adjtime_i225; + adapter->ptp_caps.gettimex64 = igc_ptp_gettimex64_i225; + adapter->ptp_caps.settime64 = igc_ptp_settime_i225; + adapter->ptp_caps.enable = igc_ptp_feature_enable_i225; + break; + default: + adapter->ptp_clock = NULL; + return; + } + + spin_lock_init(&adapter->tmreg_lock); + INIT_WORK(&adapter->ptp_tx_work, igc_ptp_tx_work); + + adapter->tstamp_config.rx_filter = HWTSTAMP_FILTER_NONE; + adapter->tstamp_config.tx_type = HWTSTAMP_TX_OFF; + + igc_ptp_reset(adapter); + + adapter->ptp_clock = ptp_clock_register(&adapter->ptp_caps, + &adapter->pdev->dev); + if (IS_ERR(adapter->ptp_clock)) { + adapter->ptp_clock = NULL; + dev_err(&adapter->pdev->dev, "ptp_clock_register failed\n"); + } else if (adapter->ptp_clock) { + dev_info(&adapter->pdev->dev, "added PHC on %s\n", + adapter->netdev->name); + adapter->ptp_flags |= IGC_PTP_ENABLED; + } +} + +/** + * igc_ptp_suspend - Disable PTP work items and prepare for suspend + * @adapter: Board private structure + * + * This function stops the overflow check work and PTP Tx timestamp work, and + * will prepare the device for OS suspend. + */ +void igc_ptp_suspend(struct igc_adapter *adapter) +{ + if (!(adapter->ptp_flags & IGC_PTP_ENABLED)) + return; + + cancel_work_sync(&adapter->ptp_tx_work); + if (adapter->ptp_tx_skb) { + dev_kfree_skb_any(adapter->ptp_tx_skb); + adapter->ptp_tx_skb = NULL; + clear_bit_unlock(__IGC_PTP_TX_IN_PROGRESS, &adapter->state); + } +} + +/** + * igc_ptp_stop - Disable PTP device and stop the overflow check. + * @adapter: Board private structure. + * + * This function stops the PTP support and cancels the delayed work. + **/ +void igc_ptp_stop(struct igc_adapter *adapter) +{ + igc_ptp_suspend(adapter); + + if (adapter->ptp_clock) { + ptp_clock_unregister(adapter->ptp_clock); + dev_info(&adapter->pdev->dev, "removed PHC on %s\n", + adapter->netdev->name); + adapter->ptp_flags &= ~IGC_PTP_ENABLED; + } +} + +/** + * igc_ptp_reset - Re-enable the adapter for PTP following a reset. + * @adapter: Board private structure. + * + * This function handles the reset work required to re-enable the PTP device. + **/ +void igc_ptp_reset(struct igc_adapter *adapter) +{ + struct igc_hw *hw = &adapter->hw; + unsigned long flags; + + /* reset the tstamp_config */ + igc_ptp_set_timestamp_mode(adapter, &adapter->tstamp_config); + + spin_lock_irqsave(&adapter->tmreg_lock, flags); + + switch (adapter->hw.mac.type) { + case igc_i225: + wr32(IGC_TSAUXC, 0x0); + wr32(IGC_TSSDP, 0x0); + wr32(IGC_TSIM, IGC_TSICR_INTERRUPTS); + wr32(IGC_IMS, IGC_IMS_TS); + break; + default: + /* No work to do. */ + goto out; + } + + /* Re-initialize the timer. */ + if (hw->mac.type == igc_i225) { + struct timespec64 ts64 = ktime_to_timespec64(ktime_get_real()); + + igc_ptp_write_i225(adapter, &ts64); + } else { + timecounter_init(&adapter->tc, &adapter->cc, + ktime_to_ns(ktime_get_real())); + } +out: + spin_unlock_irqrestore(&adapter->tmreg_lock, flags); + + wrfl(); +} diff --git a/drivers/net/ethernet/intel/igc/igc_regs.h b/drivers/net/ethernet/intel/igc/igc_regs.h index 50d7c04dccf5..c9029b549b90 100644 --- a/drivers/net/ethernet/intel/igc/igc_regs.h +++ b/drivers/net/ethernet/intel/igc/igc_regs.h @@ -12,6 +12,7 @@ #define IGC_MDIC 0x00020 /* MDI Control - RW */ #define IGC_MDICNFG 0x00E04 /* MDC/MDIO Configuration - RW */ #define IGC_CONNSW 0x00034 /* Copper/Fiber switch control - RW */ +#define IGC_I225_PHPM 0x00E14 /* I225 PHY Power Management */ /* Internal Packet Buffer Size Registers */ #define IGC_RXPBS 0x02404 /* Rx Packet Buffer Size - RW */ @@ -209,12 +210,48 @@ #define IGC_LENERRS 0x04138 /* Length Errors Count */ #define IGC_HRMPC 0x0A018 /* Header Redirection Missed Packet Count */ +/* Time sync registers */ +#define IGC_TSICR 0x0B66C /* Time Sync Interrupt Cause */ +#define IGC_TSIM 0x0B674 /* Time Sync Interrupt Mask Register */ +#define IGC_TSAUXC 0x0B640 /* Timesync Auxiliary Control register */ +#define IGC_TSYNCRXCTL 0x0B620 /* Rx Time Sync Control register - RW */ +#define IGC_TSYNCTXCTL 0x0B614 /* Tx Time Sync Control register - RW */ +#define IGC_TSYNCRXCFG 0x05F50 /* Time Sync Rx Configuration - RW */ +#define IGC_TSSDP 0x0003C /* Time Sync SDP Configuration Register - RW */ + +#define IGC_IMIR(_i) (0x05A80 + ((_i) * 4)) /* Immediate Interrupt */ +#define IGC_IMIREXT(_i) (0x05AA0 + ((_i) * 4)) /* Immediate INTR Ext*/ + +#define IGC_FTQF(_n) (0x059E0 + (4 * (_n))) /* 5-tuple Queue Fltr */ + +#define IGC_RXPBS 0x02404 /* Rx Packet Buffer Size - RW */ + +/* System Time Registers */ +#define IGC_SYSTIML 0x0B600 /* System time register Low - RO */ +#define IGC_SYSTIMH 0x0B604 /* System time register High - RO */ +#define IGC_SYSTIMR 0x0B6F8 /* System time register Residue */ +#define IGC_TIMINCA 0x0B608 /* Increment attributes register - RW */ + +#define IGC_RXSTMPL 0x0B624 /* Rx timestamp Low - RO */ +#define IGC_RXSTMPH 0x0B628 /* Rx timestamp High - RO */ +#define IGC_TXSTMPL 0x0B618 /* Tx timestamp value Low - RO */ +#define IGC_TXSTMPH 0x0B61C /* Tx timestamp value High - RO */ + /* Management registers */ #define IGC_MANC 0x05820 /* Management Control - RW */ /* Shadow Ram Write Register - RW */ #define IGC_SRWR 0x12018 +/* Wake Up registers */ +#define IGC_WUC 0x05800 /* Wakeup Control - RW */ +#define IGC_WUFC 0x05808 /* Wakeup Filter Control - RW */ +#define IGC_WUS 0x05810 /* Wakeup Status - R/W1C */ +#define IGC_WUPL 0x05900 /* Wakeup Packet Length - RW */ + +/* Wake Up packet memory */ +#define IGC_WUPM_REG(_i) (0x05A00 + ((_i) * 4)) + /* forward declaration */ struct igc_hw; u32 igc_rd32(struct igc_hw *hw, u32 reg); diff --git a/drivers/net/ethernet/intel/ixgb/ixgb_main.c b/drivers/net/ethernet/intel/ixgb/ixgb_main.c index 3d8c051dd327..b64e91ea3465 100644 --- a/drivers/net/ethernet/intel/ixgb/ixgb_main.c +++ b/drivers/net/ethernet/intel/ixgb/ixgb_main.c @@ -70,7 +70,7 @@ static int ixgb_clean(struct napi_struct *, int); static bool ixgb_clean_rx_irq(struct ixgb_adapter *, int *, int); static void ixgb_alloc_rx_buffers(struct ixgb_adapter *, int); -static void ixgb_tx_timeout(struct net_device *dev); +static void ixgb_tx_timeout(struct net_device *dev, unsigned int txqueue); static void ixgb_tx_timeout_task(struct work_struct *work); static void ixgb_vlan_strip_enable(struct ixgb_adapter *adapter); @@ -1538,7 +1538,7 @@ ixgb_xmit_frame(struct sk_buff *skb, struct net_device *netdev) **/ static void -ixgb_tx_timeout(struct net_device *netdev) +ixgb_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ixgb_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_debugfs.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_debugfs.c index 171cdc552961..5b1cf49df3d3 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_debugfs.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_debugfs.c @@ -166,7 +166,9 @@ static ssize_t ixgbe_dbg_netdev_ops_write(struct file *filp, ixgbe_dbg_netdev_ops_buf[len] = '\0'; if (strncmp(ixgbe_dbg_netdev_ops_buf, "tx_timeout", 10) == 0) { - adapter->netdev->netdev_ops->ndo_tx_timeout(adapter->netdev); + /* TX Queue number below is wrong, but ixgbe does not use it */ + adapter->netdev->netdev_ops->ndo_tx_timeout(adapter->netdev, + UINT_MAX); e_dev_info("tx_timeout called\n"); } else { e_dev_info("Unknown command: %s\n", ixgbe_dbg_netdev_ops_buf); diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index a2b2ad1f60b1..718931d951bc 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -6175,7 +6175,7 @@ static void ixgbe_set_eee_capable(struct ixgbe_adapter *adapter) * ixgbe_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void ixgbe_tx_timeout(struct net_device *netdev) +static void ixgbe_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ixgbe_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c index b43be9f14105..74b540ebb3dc 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_xsk.c @@ -277,7 +277,7 @@ static bool ixgbe_alloc_buffer_zc(struct ixgbe_ring *rx_ring, bi->handle = xsk_umem_adjust_offset(umem, handle, umem->headroom); - xsk_umem_discard_addr(umem); + xsk_umem_release_addr(umem); return true; } @@ -304,7 +304,7 @@ static bool ixgbe_alloc_buffer_slow_zc(struct ixgbe_ring *rx_ring, bi->handle = xsk_umem_adjust_offset(umem, handle, umem->headroom); - xsk_umem_discard_addr_rq(umem); + xsk_umem_release_addr_rq(umem); return true; } diff --git a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c index 64ec0e7c64b4..4622c4ea2e46 100644 --- a/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c +++ b/drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c @@ -250,7 +250,7 @@ static void ixgbevf_tx_timeout_reset(struct ixgbevf_adapter *adapter) * ixgbevf_tx_timeout - Respond to a Tx Hang * @netdev: network interface device structure **/ -static void ixgbevf_tx_timeout(struct net_device *netdev) +static void ixgbevf_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ixgbevf_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/jme.c b/drivers/net/ethernet/jme.c index 25aa400e2e3c..2e4975572e9f 100644 --- a/drivers/net/ethernet/jme.c +++ b/drivers/net/ethernet/jme.c @@ -2337,7 +2337,7 @@ jme_change_mtu(struct net_device *netdev, int new_mtu) } static void -jme_tx_timeout(struct net_device *netdev) +jme_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct jme_adapter *jme = netdev_priv(netdev); diff --git a/drivers/net/ethernet/korina.c b/drivers/net/ethernet/korina.c index d3164537b694..f1d84921e42b 100644 --- a/drivers/net/ethernet/korina.c +++ b/drivers/net/ethernet/korina.c @@ -917,7 +917,7 @@ static void korina_restart_task(struct work_struct *work) enable_irq(lp->rx_irq); } -static void korina_tx_timeout(struct net_device *dev) +static void korina_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct korina_private *lp = netdev_priv(dev); diff --git a/drivers/net/ethernet/lantiq_etop.c b/drivers/net/ethernet/lantiq_etop.c index 41f2f5480741..2d0c52f7106b 100644 --- a/drivers/net/ethernet/lantiq_etop.c +++ b/drivers/net/ethernet/lantiq_etop.c @@ -510,13 +510,6 @@ ltq_etop_change_mtu(struct net_device *dev, int new_mtu) } static int -ltq_etop_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - /* TODO: mii-toll reports "No MII transceiver present!." ?!*/ - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - -static int ltq_etop_set_mac_address(struct net_device *dev, void *p) { int ret = eth_mac_addr(dev, p); @@ -594,7 +587,7 @@ err_hw: } static void -ltq_etop_tx_timeout(struct net_device *dev) +ltq_etop_tx_timeout(struct net_device *dev, unsigned int txqueue) { int err; @@ -616,7 +609,7 @@ static const struct net_device_ops ltq_eth_netdev_ops = { .ndo_stop = ltq_etop_stop, .ndo_start_xmit = ltq_etop_tx, .ndo_change_mtu = ltq_etop_change_mtu, - .ndo_do_ioctl = ltq_etop_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_set_mac_address = ltq_etop_set_mac_address, .ndo_validate_addr = eth_validate_addr, .ndo_set_rx_mode = ltq_etop_set_multicast_list, diff --git a/drivers/net/ethernet/marvell/mv643xx_eth.c b/drivers/net/ethernet/marvell/mv643xx_eth.c index 65a093216dac..3c8125cbc84d 100644 --- a/drivers/net/ethernet/marvell/mv643xx_eth.c +++ b/drivers/net/ethernet/marvell/mv643xx_eth.c @@ -2590,7 +2590,7 @@ static void tx_timeout_task(struct work_struct *ugly) } } -static void mv643xx_eth_tx_timeout(struct net_device *dev) +static void mv643xx_eth_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mv643xx_eth_private *mp = netdev_priv(dev); diff --git a/drivers/net/ethernet/marvell/mvneta.c b/drivers/net/ethernet/marvell/mvneta.c index 67ad8b8b127d..2dfbfdff45a8 100644 --- a/drivers/net/ethernet/marvell/mvneta.c +++ b/drivers/net/ethernet/marvell/mvneta.c @@ -3072,7 +3072,7 @@ static int mvneta_create_page_pool(struct mvneta_port *pp, .order = 0, .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV, .pool_size = size, - .nid = cpu_to_node(0), + .nid = NUMA_NO_NODE, .dev = pp->dev->dev.parent, .dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE, .offset = pp->rx_offset_correction, @@ -4226,6 +4226,12 @@ static int mvneta_xdp_setup(struct net_device *dev, struct bpf_prog *prog, return -EOPNOTSUPP; } + if (pp->bm_priv) { + NL_SET_ERR_MSG_MOD(extack, + "Hardware Buffer Management not supported on XDP"); + return -EOPNOTSUPP; + } + need_update = !!pp->xdp_prog != !!prog; if (running && need_update) mvneta_stop(dev); diff --git a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c index 14e372cda7f4..72133cbe55d4 100644 --- a/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c +++ b/drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c @@ -1114,7 +1114,7 @@ mvpp2_shared_interrupt_mask_unmask(struct mvpp2_port *port, bool mask) /* Port configuration routines */ static bool mvpp2_is_xlg(phy_interface_t interface) { - return interface == PHY_INTERFACE_MODE_10GKR || + return interface == PHY_INTERFACE_MODE_10GBASER || interface == PHY_INTERFACE_MODE_XAUI; } @@ -1200,7 +1200,7 @@ static int mvpp22_gop_init(struct mvpp2_port *port) case PHY_INTERFACE_MODE_2500BASEX: mvpp22_gop_init_sgmii(port); break; - case PHY_INTERFACE_MODE_10GKR: + case PHY_INTERFACE_MODE_10GBASER: if (port->gop_id != 0) goto invalid_conf; mvpp22_gop_init_10gkr(port); @@ -1649,7 +1649,7 @@ static void mvpp22_pcs_reset_deassert(struct mvpp2_port *port) xpcs = priv->iface_base + MVPP22_XPCS_BASE(port->gop_id); switch (port->phy_interface) { - case PHY_INTERFACE_MODE_10GKR: + case PHY_INTERFACE_MODE_10GBASER: val = readl(mpcs + MVPP22_MPCS_CLK_RESET); val |= MAC_CLK_RESET_MAC | MAC_CLK_RESET_SD_RX | MAC_CLK_RESET_SD_TX; @@ -4758,7 +4758,7 @@ static void mvpp2_phylink_validate(struct phylink_config *config, /* Invalid combinations */ switch (state->interface) { - case PHY_INTERFACE_MODE_10GKR: + case PHY_INTERFACE_MODE_10GBASER: case PHY_INTERFACE_MODE_XAUI: if (port->gop_id != 0) goto empty_set; @@ -4780,7 +4780,7 @@ static void mvpp2_phylink_validate(struct phylink_config *config, phylink_set(mask, Asym_Pause); switch (state->interface) { - case PHY_INTERFACE_MODE_10GKR: + case PHY_INTERFACE_MODE_10GBASER: case PHY_INTERFACE_MODE_XAUI: case PHY_INTERFACE_MODE_NA: if (port->gop_id == 0) { @@ -4792,6 +4792,8 @@ static void mvpp2_phylink_validate(struct phylink_config *config, phylink_set(mask, 10000baseER_Full); phylink_set(mask, 10000baseKR_Full); } + if (state->interface != PHY_INTERFACE_MODE_NA) + break; /* Fall-through */ case PHY_INTERFACE_MODE_RGMII: case PHY_INTERFACE_MODE_RGMII_ID: @@ -4802,13 +4804,23 @@ static void mvpp2_phylink_validate(struct phylink_config *config, phylink_set(mask, 10baseT_Full); phylink_set(mask, 100baseT_Half); phylink_set(mask, 100baseT_Full); + phylink_set(mask, 1000baseT_Full); + phylink_set(mask, 1000baseX_Full); + if (state->interface != PHY_INTERFACE_MODE_NA) + break; /* Fall-through */ case PHY_INTERFACE_MODE_1000BASEX: case PHY_INTERFACE_MODE_2500BASEX: - phylink_set(mask, 1000baseT_Full); - phylink_set(mask, 1000baseX_Full); - phylink_set(mask, 2500baseT_Full); - phylink_set(mask, 2500baseX_Full); + if (port->comphy || + state->interface != PHY_INTERFACE_MODE_2500BASEX) { + phylink_set(mask, 1000baseT_Full); + phylink_set(mask, 1000baseX_Full); + } + if (port->comphy || + state->interface == PHY_INTERFACE_MODE_2500BASEX) { + phylink_set(mask, 2500baseT_Full); + phylink_set(mask, 2500baseX_Full); + } break; default: goto empty_set; @@ -4817,6 +4829,8 @@ static void mvpp2_phylink_validate(struct phylink_config *config, bitmap_and(supported, supported, mask, __ETHTOOL_LINK_MODE_MASK_NBITS); bitmap_and(state->advertising, state->advertising, mask, __ETHTOOL_LINK_MODE_MASK_NBITS); + + phylink_helper_basex_speed(state); return; empty_set: @@ -5233,6 +5247,15 @@ static int mvpp2_port_probe(struct platform_device *pdev, goto err_free_netdev; } + /* + * Rewrite 10GBASE-KR to 10GBASE-R for compatibility with existing DT. + * Existing usage of 10GBASE-KR is not correct; no backplane + * negotiation is done, and this driver does not actually support + * 10GBASE-KR. + */ + if (phy_mode == PHY_INTERFACE_MODE_10GKR) + phy_mode = PHY_INTERFACE_MODE_10GBASER; + if (port_node) { comphy = devm_of_phy_get(&pdev->dev, port_node, NULL); if (IS_ERR(comphy)) { @@ -5411,6 +5434,16 @@ static int mvpp2_port_probe(struct platform_device *pdev, port->phylink = NULL; } + /* Cycle the comphy to power it down, saving 270mW per port - + * don't worry about an error powering it up. When the comphy + * driver does this, we can remove this code. + */ + if (port->comphy) { + err = mvpp22_comphy_init(port); + if (err == 0) + phy_power_off(port->comphy); + } + err = register_netdev(dev); if (err < 0) { dev_err(&pdev->dev, "failed to register netdev\n"); diff --git a/drivers/net/ethernet/marvell/octeontx2/Kconfig b/drivers/net/ethernet/marvell/octeontx2/Kconfig index fb34fbd62088..ced514c05c97 100644 --- a/drivers/net/ethernet/marvell/octeontx2/Kconfig +++ b/drivers/net/ethernet/marvell/octeontx2/Kconfig @@ -25,3 +25,11 @@ config NDC_DIS_DYNAMIC_CACHING This config option disables caching of dynamic entries such as NIX SQEs , NPA stack pages etc in NDC. Also locks down NIX SQ/CQ/RQ/RSS and NPA Aura/Pool contexts. + +config OCTEONTX2_PF + tristate "Marvell OcteonTX2 NIC Physical Function driver" + select OCTEONTX2_MBOX + depends on (64BIT && COMPILE_TEST) || ARM64 + depends on PCI + help + This driver supports Marvell's OcteonTX2 NIC physical function. diff --git a/drivers/net/ethernet/marvell/octeontx2/Makefile b/drivers/net/ethernet/marvell/octeontx2/Makefile index e579dcd54c97..0064a69e0f72 100644 --- a/drivers/net/ethernet/marvell/octeontx2/Makefile +++ b/drivers/net/ethernet/marvell/octeontx2/Makefile @@ -3,4 +3,6 @@ # Makefile for Marvell OcteonTX2 device drivers. # +obj-$(CONFIG_OCTEONTX2_MBOX) += af/ obj-$(CONFIG_OCTEONTX2_AF) += af/ +obj-$(CONFIG_OCTEONTX2_PF) += nic/ diff --git a/drivers/net/ethernet/marvell/octeontx2/af/common.h b/drivers/net/ethernet/marvell/octeontx2/af/common.h index 784207bae5f8..cd33c2e6ca5f 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/common.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/common.h @@ -143,8 +143,13 @@ enum nix_scheduler { NIX_TXSCH_LVL_CNT = 0x5, }; -#define TXSCH_TL1_DFLT_RR_QTM ((1 << 24) - 1) -#define TXSCH_TL1_DFLT_RR_PRIO (0x1ull) +#define TXSCH_RR_QTM_MAX ((1 << 24) - 1) +#define TXSCH_TL1_DFLT_RR_QTM TXSCH_RR_QTM_MAX +#define TXSCH_TL1_DFLT_RR_PRIO (0x1ull) +#define MAX_SCHED_WEIGHT 0xFF +#define DFLT_RR_WEIGHT 71 +#define DFLT_RR_QTM ((DFLT_RR_WEIGHT * TXSCH_RR_QTM_MAX) \ + / MAX_SCHED_WEIGHT) /* Min/Max packet sizes, excluding FCS */ #define NIC_HW_MIN_FRS 40 diff --git a/drivers/net/ethernet/marvell/octeontx2/af/mbox.h b/drivers/net/ethernet/marvell/octeontx2/af/mbox.h index a589748f1240..8bbc1f1d81f5 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/mbox.h +++ b/drivers/net/ethernet/marvell/octeontx2/af/mbox.h @@ -210,7 +210,8 @@ M(NIX_SET_RX_CFG, 0x8010, nix_set_rx_cfg, nix_rx_cfg, msg_rsp) \ M(NIX_LSO_FORMAT_CFG, 0x8011, nix_lso_format_cfg, \ nix_lso_format_cfg, \ nix_lso_format_cfg_rsp) \ -M(NIX_RXVLAN_ALLOC, 0x8012, nix_rxvlan_alloc, msg_req, msg_rsp) +M(NIX_RXVLAN_ALLOC, 0x8012, nix_rxvlan_alloc, msg_req, msg_rsp) \ +M(NIX_GET_MAC_ADDR, 0x8018, nix_get_mac_addr, msg_req, nix_get_mac_addr_rsp) \ /* Messages initiated by AF (range 0xC00 - 0xDFF) */ #define MBOX_UP_CGX_MESSAGES \ @@ -618,6 +619,11 @@ struct nix_set_mac_addr { u8 mac_addr[ETH_ALEN]; /* MAC address to be set for this pcifunc */ }; +struct nix_get_mac_addr_rsp { + struct mbox_msghdr hdr; + u8 mac_addr[ETH_ALEN]; +}; + struct nix_mark_format_cfg { struct mbox_msghdr hdr; u8 offset; diff --git a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c index 8a59f7d53fbf..eb5e542424e7 100644 --- a/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c +++ b/drivers/net/ethernet/marvell/octeontx2/af/rvu_nix.c @@ -2546,6 +2546,23 @@ int rvu_mbox_handler_nix_set_mac_addr(struct rvu *rvu, return 0; } +int rvu_mbox_handler_nix_get_mac_addr(struct rvu *rvu, + struct msg_req *req, + struct nix_get_mac_addr_rsp *rsp) +{ + u16 pcifunc = req->hdr.pcifunc; + struct rvu_pfvf *pfvf; + + if (!is_nixlf_attached(rvu, pcifunc)) + return NIX_AF_ERR_AF_LF_INVALID; + + pfvf = rvu_get_pfvf(rvu, pcifunc); + + ether_addr_copy(rsp->mac_addr, pfvf->mac_addr); + + return 0; +} + int rvu_mbox_handler_nix_set_rx_mode(struct rvu *rvu, struct nix_rx_mode *req, struct msg_rsp *rsp) { diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/Makefile b/drivers/net/ethernet/marvell/octeontx2/nic/Makefile new file mode 100644 index 000000000000..41bf00cf5b1d --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/Makefile @@ -0,0 +1,10 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Makefile for Marvell's OcteonTX2 ethernet device drivers +# + +obj-$(CONFIG_OCTEONTX2_PF) += octeontx2_nicpf.o + +octeontx2_nicpf-y := otx2_pf.o otx2_common.o otx2_txrx.o otx2_ethtool.o + +ccflags-y += -I$(srctree)/drivers/net/ethernet/marvell/octeontx2/af diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c new file mode 100644 index 000000000000..8247d21d0432 --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c @@ -0,0 +1,1410 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/interrupt.h> +#include <linux/pci.h> +#include <net/tso.h> + +#include "otx2_reg.h" +#include "otx2_common.h" +#include "otx2_struct.h" + +static void otx2_nix_rq_op_stats(struct queue_stats *stats, + struct otx2_nic *pfvf, int qidx) +{ + u64 incr = (u64)qidx << 32; + u64 *ptr; + + ptr = (u64 *)otx2_get_regaddr(pfvf, NIX_LF_RQ_OP_OCTS); + stats->bytes = otx2_atomic64_add(incr, ptr); + + ptr = (u64 *)otx2_get_regaddr(pfvf, NIX_LF_RQ_OP_PKTS); + stats->pkts = otx2_atomic64_add(incr, ptr); +} + +static void otx2_nix_sq_op_stats(struct queue_stats *stats, + struct otx2_nic *pfvf, int qidx) +{ + u64 incr = (u64)qidx << 32; + u64 *ptr; + + ptr = (u64 *)otx2_get_regaddr(pfvf, NIX_LF_SQ_OP_OCTS); + stats->bytes = otx2_atomic64_add(incr, ptr); + + ptr = (u64 *)otx2_get_regaddr(pfvf, NIX_LF_SQ_OP_PKTS); + stats->pkts = otx2_atomic64_add(incr, ptr); +} + +void otx2_update_lmac_stats(struct otx2_nic *pfvf) +{ + struct msg_req *req; + + if (!netif_running(pfvf->netdev)) + return; + + otx2_mbox_lock(&pfvf->mbox); + req = otx2_mbox_alloc_msg_cgx_stats(&pfvf->mbox); + if (!req) { + otx2_mbox_unlock(&pfvf->mbox); + return; + } + + otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); +} + +int otx2_update_rq_stats(struct otx2_nic *pfvf, int qidx) +{ + struct otx2_rcv_queue *rq = &pfvf->qset.rq[qidx]; + + if (!pfvf->qset.rq) + return 0; + + otx2_nix_rq_op_stats(&rq->stats, pfvf, qidx); + return 1; +} + +int otx2_update_sq_stats(struct otx2_nic *pfvf, int qidx) +{ + struct otx2_snd_queue *sq = &pfvf->qset.sq[qidx]; + + if (!pfvf->qset.sq) + return 0; + + otx2_nix_sq_op_stats(&sq->stats, pfvf, qidx); + return 1; +} + +void otx2_get_dev_stats(struct otx2_nic *pfvf) +{ + struct otx2_dev_stats *dev_stats = &pfvf->hw.dev_stats; + +#define OTX2_GET_RX_STATS(reg) \ + otx2_read64(pfvf, NIX_LF_RX_STATX(reg)) +#define OTX2_GET_TX_STATS(reg) \ + otx2_read64(pfvf, NIX_LF_TX_STATX(reg)) + + dev_stats->rx_bytes = OTX2_GET_RX_STATS(RX_OCTS); + dev_stats->rx_drops = OTX2_GET_RX_STATS(RX_DROP); + dev_stats->rx_bcast_frames = OTX2_GET_RX_STATS(RX_BCAST); + dev_stats->rx_mcast_frames = OTX2_GET_RX_STATS(RX_MCAST); + dev_stats->rx_ucast_frames = OTX2_GET_RX_STATS(RX_UCAST); + dev_stats->rx_frames = dev_stats->rx_bcast_frames + + dev_stats->rx_mcast_frames + + dev_stats->rx_ucast_frames; + + dev_stats->tx_bytes = OTX2_GET_TX_STATS(TX_OCTS); + dev_stats->tx_drops = OTX2_GET_TX_STATS(TX_DROP); + dev_stats->tx_bcast_frames = OTX2_GET_TX_STATS(TX_BCAST); + dev_stats->tx_mcast_frames = OTX2_GET_TX_STATS(TX_MCAST); + dev_stats->tx_ucast_frames = OTX2_GET_TX_STATS(TX_UCAST); + dev_stats->tx_frames = dev_stats->tx_bcast_frames + + dev_stats->tx_mcast_frames + + dev_stats->tx_ucast_frames; +} + +void otx2_get_stats64(struct net_device *netdev, + struct rtnl_link_stats64 *stats) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct otx2_dev_stats *dev_stats; + + otx2_get_dev_stats(pfvf); + + dev_stats = &pfvf->hw.dev_stats; + stats->rx_bytes = dev_stats->rx_bytes; + stats->rx_packets = dev_stats->rx_frames; + stats->rx_dropped = dev_stats->rx_drops; + stats->multicast = dev_stats->rx_mcast_frames; + + stats->tx_bytes = dev_stats->tx_bytes; + stats->tx_packets = dev_stats->tx_frames; + stats->tx_dropped = dev_stats->tx_drops; +} + +/* Sync MAC address with RVU AF */ +static int otx2_hw_set_mac_addr(struct otx2_nic *pfvf, u8 *mac) +{ + struct nix_set_mac_addr *req; + int err; + + otx2_mbox_lock(&pfvf->mbox); + req = otx2_mbox_alloc_msg_nix_set_mac_addr(&pfvf->mbox); + if (!req) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + ether_addr_copy(req->mac_addr, mac); + + err = otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); + return err; +} + +static int otx2_hw_get_mac_addr(struct otx2_nic *pfvf, + struct net_device *netdev) +{ + struct nix_get_mac_addr_rsp *rsp; + struct mbox_msghdr *msghdr; + struct msg_req *req; + int err; + + otx2_mbox_lock(&pfvf->mbox); + req = otx2_mbox_alloc_msg_nix_get_mac_addr(&pfvf->mbox); + if (!req) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) { + otx2_mbox_unlock(&pfvf->mbox); + return err; + } + + msghdr = otx2_mbox_get_rsp(&pfvf->mbox.mbox, 0, &req->hdr); + if (!msghdr) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + rsp = (struct nix_get_mac_addr_rsp *)msghdr; + ether_addr_copy(netdev->dev_addr, rsp->mac_addr); + otx2_mbox_unlock(&pfvf->mbox); + + return 0; +} + +int otx2_set_mac_address(struct net_device *netdev, void *p) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct sockaddr *addr = p; + + if (!is_valid_ether_addr(addr->sa_data)) + return -EADDRNOTAVAIL; + + if (!otx2_hw_set_mac_addr(pfvf, addr->sa_data)) + memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len); + else + return -EPERM; + + return 0; +} + +int otx2_hw_set_mtu(struct otx2_nic *pfvf, int mtu) +{ + struct nix_frs_cfg *req; + int err; + + otx2_mbox_lock(&pfvf->mbox); + req = otx2_mbox_alloc_msg_nix_set_hw_frs(&pfvf->mbox); + if (!req) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + /* SMQ config limits maximum pkt size that can be transmitted */ + req->update_smq = true; + pfvf->max_frs = mtu + OTX2_ETH_HLEN; + req->maxlen = pfvf->max_frs; + + err = otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); + return err; +} + +int otx2_set_flowkey_cfg(struct otx2_nic *pfvf) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + struct nix_rss_flowkey_cfg *req; + int err; + + otx2_mbox_lock(&pfvf->mbox); + req = otx2_mbox_alloc_msg_nix_rss_flowkey_cfg(&pfvf->mbox); + if (!req) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + req->mcam_index = -1; /* Default or reserved index */ + req->flowkey_cfg = rss->flowkey_cfg; + req->group = DEFAULT_RSS_CONTEXT_GROUP; + + err = otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); + return err; +} + +int otx2_set_rss_table(struct otx2_nic *pfvf) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + struct mbox *mbox = &pfvf->mbox; + struct nix_aq_enq_req *aq; + int idx, err; + + otx2_mbox_lock(mbox); + /* Get memory to put this msg */ + for (idx = 0; idx < rss->rss_size; idx++) { + aq = otx2_mbox_alloc_msg_nix_aq_enq(mbox); + if (!aq) { + /* The shared memory buffer can be full. + * Flush it and retry + */ + err = otx2_sync_mbox_msg(mbox); + if (err) { + otx2_mbox_unlock(mbox); + return err; + } + aq = otx2_mbox_alloc_msg_nix_aq_enq(mbox); + if (!aq) { + otx2_mbox_unlock(mbox); + return -ENOMEM; + } + } + + aq->rss.rq = rss->ind_tbl[idx]; + + /* Fill AQ info */ + aq->qidx = idx; + aq->ctype = NIX_AQ_CTYPE_RSS; + aq->op = NIX_AQ_INSTOP_INIT; + } + err = otx2_sync_mbox_msg(mbox); + otx2_mbox_unlock(mbox); + return err; +} + +void otx2_set_rss_key(struct otx2_nic *pfvf) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + u64 *key = (u64 *)&rss->key[4]; + int idx; + + /* 352bit or 44byte key needs to be configured as below + * NIX_LF_RX_SECRETX0 = key<351:288> + * NIX_LF_RX_SECRETX1 = key<287:224> + * NIX_LF_RX_SECRETX2 = key<223:160> + * NIX_LF_RX_SECRETX3 = key<159:96> + * NIX_LF_RX_SECRETX4 = key<95:32> + * NIX_LF_RX_SECRETX5<63:32> = key<31:0> + */ + otx2_write64(pfvf, NIX_LF_RX_SECRETX(5), + (u64)(*((u32 *)&rss->key)) << 32); + idx = sizeof(rss->key) / sizeof(u64); + while (idx > 0) { + idx--; + otx2_write64(pfvf, NIX_LF_RX_SECRETX(idx), *key++); + } +} + +int otx2_rss_init(struct otx2_nic *pfvf) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + int idx, ret = 0; + + rss->rss_size = sizeof(rss->ind_tbl); + + /* Init RSS key if it is not setup already */ + if (!rss->enable) + netdev_rss_key_fill(rss->key, sizeof(rss->key)); + otx2_set_rss_key(pfvf); + + if (!netif_is_rxfh_configured(pfvf->netdev)) { + /* Default indirection table */ + for (idx = 0; idx < rss->rss_size; idx++) + rss->ind_tbl[idx] = + ethtool_rxfh_indir_default(idx, + pfvf->hw.rx_queues); + } + ret = otx2_set_rss_table(pfvf); + if (ret) + return ret; + + /* Flowkey or hash config to be used for generating flow tag */ + rss->flowkey_cfg = rss->enable ? rss->flowkey_cfg : + NIX_FLOW_KEY_TYPE_IPV4 | NIX_FLOW_KEY_TYPE_IPV6 | + NIX_FLOW_KEY_TYPE_TCP | NIX_FLOW_KEY_TYPE_UDP | + NIX_FLOW_KEY_TYPE_SCTP; + + ret = otx2_set_flowkey_cfg(pfvf); + if (ret) + return ret; + + rss->enable = true; + return 0; +} + +void otx2_config_irq_coalescing(struct otx2_nic *pfvf, int qidx) +{ + /* Configure CQE interrupt coalescing parameters + * + * HW triggers an irq when ECOUNT > cq_ecount_wait, hence + * set 1 less than cq_ecount_wait. And cq_time_wait is in + * usecs, convert that to 100ns count. + */ + otx2_write64(pfvf, NIX_LF_CINTX_WAIT(qidx), + ((u64)(pfvf->hw.cq_time_wait * 10) << 48) | + ((u64)pfvf->hw.cq_qcount_wait << 32) | + (pfvf->hw.cq_ecount_wait - 1)); +} + +dma_addr_t otx2_alloc_rbuf(struct otx2_nic *pfvf, struct otx2_pool *pool, + gfp_t gfp) +{ + dma_addr_t iova; + + /* Check if request can be accommodated in previous allocated page */ + if (pool->page && ((pool->page_offset + pool->rbsize) <= + (PAGE_SIZE << pool->rbpage_order))) { + pool->pageref++; + goto ret; + } + + otx2_get_page(pool); + + /* Allocate a new page */ + pool->page = alloc_pages(gfp | __GFP_COMP | __GFP_NOWARN, + pool->rbpage_order); + if (unlikely(!pool->page)) + return -ENOMEM; + + pool->page_offset = 0; +ret: + iova = (u64)otx2_dma_map_page(pfvf, pool->page, pool->page_offset, + pool->rbsize, DMA_FROM_DEVICE); + if (!iova) { + if (!pool->page_offset) + __free_pages(pool->page, pool->rbpage_order); + pool->page = NULL; + return -ENOMEM; + } + pool->page_offset += pool->rbsize; + return iova; +} + +void otx2_tx_timeout(struct net_device *netdev, unsigned int txq) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + + schedule_work(&pfvf->reset_task); +} + +void otx2_get_mac_from_af(struct net_device *netdev) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + int err; + + err = otx2_hw_get_mac_addr(pfvf, netdev); + if (err) + dev_warn(pfvf->dev, "Failed to read mac from hardware\n"); + + /* If AF doesn't provide a valid MAC, generate a random one */ + if (!is_valid_ether_addr(netdev->dev_addr)) + eth_hw_addr_random(netdev); +} + +static int otx2_get_link(struct otx2_nic *pfvf) +{ + int link = 0; + u16 map; + + /* cgx lmac link */ + if (pfvf->hw.tx_chan_base >= CGX_CHAN_BASE) { + map = pfvf->hw.tx_chan_base & 0x7FF; + link = 4 * ((map >> 8) & 0xF) + ((map >> 4) & 0xF); + } + /* LBK channel */ + if (pfvf->hw.tx_chan_base < SDP_CHAN_BASE) + link = 12; + + return link; +} + +int otx2_txschq_config(struct otx2_nic *pfvf, int lvl) +{ + struct otx2_hw *hw = &pfvf->hw; + struct nix_txschq_config *req; + u64 schq, parent; + + req = otx2_mbox_alloc_msg_nix_txschq_cfg(&pfvf->mbox); + if (!req) + return -ENOMEM; + + req->lvl = lvl; + req->num_regs = 1; + + schq = hw->txschq_list[lvl][0]; + /* Set topology e.t.c configuration */ + if (lvl == NIX_TXSCH_LVL_SMQ) { + req->reg[0] = NIX_AF_SMQX_CFG(schq); + req->regval[0] = ((pfvf->netdev->mtu + OTX2_ETH_HLEN) << 8) | + OTX2_MIN_MTU; + + req->regval[0] |= (0x20ULL << 51) | (0x80ULL << 39) | + (0x2ULL << 36); + req->num_regs++; + /* MDQ config */ + parent = hw->txschq_list[NIX_TXSCH_LVL_TL4][0]; + req->reg[1] = NIX_AF_MDQX_PARENT(schq); + req->regval[1] = parent << 16; + req->num_regs++; + /* Set DWRR quantum */ + req->reg[2] = NIX_AF_MDQX_SCHEDULE(schq); + req->regval[2] = DFLT_RR_QTM; + } else if (lvl == NIX_TXSCH_LVL_TL4) { + parent = hw->txschq_list[NIX_TXSCH_LVL_TL3][0]; + req->reg[0] = NIX_AF_TL4X_PARENT(schq); + req->regval[0] = parent << 16; + req->num_regs++; + req->reg[1] = NIX_AF_TL4X_SCHEDULE(schq); + req->regval[1] = DFLT_RR_QTM; + } else if (lvl == NIX_TXSCH_LVL_TL3) { + parent = hw->txschq_list[NIX_TXSCH_LVL_TL2][0]; + req->reg[0] = NIX_AF_TL3X_PARENT(schq); + req->regval[0] = parent << 16; + req->num_regs++; + req->reg[1] = NIX_AF_TL3X_SCHEDULE(schq); + req->regval[1] = DFLT_RR_QTM; + } else if (lvl == NIX_TXSCH_LVL_TL2) { + parent = hw->txschq_list[NIX_TXSCH_LVL_TL1][0]; + req->reg[0] = NIX_AF_TL2X_PARENT(schq); + req->regval[0] = parent << 16; + + req->num_regs++; + req->reg[1] = NIX_AF_TL2X_SCHEDULE(schq); + req->regval[1] = TXSCH_TL1_DFLT_RR_PRIO << 24 | DFLT_RR_QTM; + + req->num_regs++; + req->reg[2] = NIX_AF_TL3_TL2X_LINKX_CFG(schq, + otx2_get_link(pfvf)); + /* Enable this queue and backpressure */ + req->regval[2] = BIT_ULL(13) | BIT_ULL(12); + + } else if (lvl == NIX_TXSCH_LVL_TL1) { + /* Default config for TL1. + * For VF this is always ignored. + */ + + /* Set DWRR quantum */ + req->reg[0] = NIX_AF_TL1X_SCHEDULE(schq); + req->regval[0] = TXSCH_TL1_DFLT_RR_QTM; + + req->num_regs++; + req->reg[1] = NIX_AF_TL1X_TOPOLOGY(schq); + req->regval[1] = (TXSCH_TL1_DFLT_RR_PRIO << 1); + + req->num_regs++; + req->reg[2] = NIX_AF_TL1X_CIR(schq); + req->regval[2] = 0; + } + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +int otx2_txsch_alloc(struct otx2_nic *pfvf) +{ + struct nix_txsch_alloc_req *req; + int lvl; + + /* Get memory to put this msg */ + req = otx2_mbox_alloc_msg_nix_txsch_alloc(&pfvf->mbox); + if (!req) + return -ENOMEM; + + /* Request one schq per level */ + for (lvl = 0; lvl < NIX_TXSCH_LVL_CNT; lvl++) + req->schq[lvl] = 1; + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +int otx2_txschq_stop(struct otx2_nic *pfvf) +{ + struct nix_txsch_free_req *free_req; + int lvl, schq, err; + + otx2_mbox_lock(&pfvf->mbox); + /* Free the transmit schedulers */ + free_req = otx2_mbox_alloc_msg_nix_txsch_free(&pfvf->mbox); + if (!free_req) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + free_req->flags = TXSCHQ_FREE_ALL; + err = otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); + + /* Clear the txschq list */ + for (lvl = 0; lvl < NIX_TXSCH_LVL_CNT; lvl++) { + for (schq = 0; schq < MAX_TXSCHQ_PER_FUNC; schq++) + pfvf->hw.txschq_list[lvl][schq] = 0; + } + return err; +} + +void otx2_sqb_flush(struct otx2_nic *pfvf) +{ + int qidx, sqe_tail, sqe_head; + u64 incr, *ptr, val; + + ptr = (u64 *)otx2_get_regaddr(pfvf, NIX_LF_SQ_OP_STATUS); + for (qidx = 0; qidx < pfvf->hw.tx_queues; qidx++) { + incr = (u64)qidx << 32; + while (1) { + val = otx2_atomic64_add(incr, ptr); + sqe_head = (val >> 20) & 0x3F; + sqe_tail = (val >> 28) & 0x3F; + if (sqe_head == sqe_tail) + break; + usleep_range(1, 3); + } + } +} + +/* RED and drop levels of CQ on packet reception. + * For CQ level is measure of emptiness ( 0x0 = full, 255 = empty). + */ +#define RQ_PASS_LVL_CQ(skid, qsize) ((((skid) + 16) * 256) / (qsize)) +#define RQ_DROP_LVL_CQ(skid, qsize) (((skid) * 256) / (qsize)) + +/* RED and drop levels of AURA for packet reception. + * For AURA level is measure of fullness (0x0 = empty, 255 = full). + * Eg: For RQ length 1K, for pass/drop level 204/230. + * RED accepts pkts if free pointers > 102 & <= 205. + * Drops pkts if free pointers < 102. + */ +#define RQ_PASS_LVL_AURA (255 - ((95 * 256) / 100)) /* RED when 95% is full */ +#define RQ_DROP_LVL_AURA (255 - ((99 * 256) / 100)) /* Drop when 99% is full */ + +/* Send skid of 2000 packets required for CQ size of 4K CQEs. */ +#define SEND_CQ_SKID 2000 + +static int otx2_rq_init(struct otx2_nic *pfvf, u16 qidx, u16 lpb_aura) +{ + struct otx2_qset *qset = &pfvf->qset; + struct nix_aq_enq_req *aq; + + /* Get memory to put this msg */ + aq = otx2_mbox_alloc_msg_nix_aq_enq(&pfvf->mbox); + if (!aq) + return -ENOMEM; + + aq->rq.cq = qidx; + aq->rq.ena = 1; + aq->rq.pb_caching = 1; + aq->rq.lpb_aura = lpb_aura; /* Use large packet buffer aura */ + aq->rq.lpb_sizem1 = (DMA_BUFFER_LEN(pfvf->rbsize) / 8) - 1; + aq->rq.xqe_imm_size = 0; /* Copying of packet to CQE not needed */ + aq->rq.flow_tagw = 32; /* Copy full 32bit flow_tag to CQE header */ + aq->rq.qint_idx = 0; + aq->rq.lpb_drop_ena = 1; /* Enable RED dropping for AURA */ + aq->rq.xqe_drop_ena = 1; /* Enable RED dropping for CQ/SSO */ + aq->rq.xqe_pass = RQ_PASS_LVL_CQ(pfvf->hw.rq_skid, qset->rqe_cnt); + aq->rq.xqe_drop = RQ_DROP_LVL_CQ(pfvf->hw.rq_skid, qset->rqe_cnt); + aq->rq.lpb_aura_pass = RQ_PASS_LVL_AURA; + aq->rq.lpb_aura_drop = RQ_DROP_LVL_AURA; + + /* Fill AQ info */ + aq->qidx = qidx; + aq->ctype = NIX_AQ_CTYPE_RQ; + aq->op = NIX_AQ_INSTOP_INIT; + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +static int otx2_sq_init(struct otx2_nic *pfvf, u16 qidx, u16 sqb_aura) +{ + struct otx2_qset *qset = &pfvf->qset; + struct otx2_snd_queue *sq; + struct nix_aq_enq_req *aq; + struct otx2_pool *pool; + int err; + + pool = &pfvf->qset.pool[sqb_aura]; + sq = &qset->sq[qidx]; + sq->sqe_size = NIX_SQESZ_W16 ? 64 : 128; + sq->sqe_cnt = qset->sqe_cnt; + + err = qmem_alloc(pfvf->dev, &sq->sqe, 1, sq->sqe_size); + if (err) + return err; + + err = qmem_alloc(pfvf->dev, &sq->tso_hdrs, qset->sqe_cnt, + TSO_HEADER_SIZE); + if (err) + return err; + + sq->sqe_base = sq->sqe->base; + sq->sg = kcalloc(qset->sqe_cnt, sizeof(struct sg_list), GFP_KERNEL); + if (!sq->sg) + return -ENOMEM; + + sq->head = 0; + sq->sqe_per_sqb = (pfvf->hw.sqb_size / sq->sqe_size) - 1; + sq->num_sqbs = (qset->sqe_cnt + sq->sqe_per_sqb) / sq->sqe_per_sqb; + /* Set SQE threshold to 10% of total SQEs */ + sq->sqe_thresh = ((sq->num_sqbs * sq->sqe_per_sqb) * 10) / 100; + sq->aura_id = sqb_aura; + sq->aura_fc_addr = pool->fc_addr->base; + sq->lmt_addr = (__force u64 *)(pfvf->reg_base + LMT_LF_LMTLINEX(qidx)); + sq->io_addr = (__force u64)otx2_get_regaddr(pfvf, NIX_LF_OP_SENDX(0)); + + sq->stats.bytes = 0; + sq->stats.pkts = 0; + + /* Get memory to put this msg */ + aq = otx2_mbox_alloc_msg_nix_aq_enq(&pfvf->mbox); + if (!aq) + return -ENOMEM; + + aq->sq.cq = pfvf->hw.rx_queues + qidx; + aq->sq.max_sqe_size = NIX_MAXSQESZ_W16; /* 128 byte */ + aq->sq.cq_ena = 1; + aq->sq.ena = 1; + /* Only one SMQ is allocated, map all SQ's to that SMQ */ + aq->sq.smq = pfvf->hw.txschq_list[NIX_TXSCH_LVL_SMQ][0]; + aq->sq.smq_rr_quantum = DFLT_RR_QTM; + aq->sq.default_chan = pfvf->hw.tx_chan_base; + aq->sq.sqe_stype = NIX_STYPE_STF; /* Cache SQB */ + aq->sq.sqb_aura = sqb_aura; + aq->sq.sq_int_ena = NIX_SQINT_BITS; + aq->sq.qint_idx = 0; + /* Due pipelining impact minimum 2000 unused SQ CQE's + * need to maintain to avoid CQ overflow. + */ + aq->sq.cq_limit = ((SEND_CQ_SKID * 256) / (sq->sqe_cnt)); + + /* Fill AQ info */ + aq->qidx = qidx; + aq->ctype = NIX_AQ_CTYPE_SQ; + aq->op = NIX_AQ_INSTOP_INIT; + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +static int otx2_cq_init(struct otx2_nic *pfvf, u16 qidx) +{ + struct otx2_qset *qset = &pfvf->qset; + struct nix_aq_enq_req *aq; + struct otx2_cq_queue *cq; + int err, pool_id; + + cq = &qset->cq[qidx]; + cq->cq_idx = qidx; + if (qidx < pfvf->hw.rx_queues) { + cq->cq_type = CQ_RX; + cq->cint_idx = qidx; + cq->cqe_cnt = qset->rqe_cnt; + } else { + cq->cq_type = CQ_TX; + cq->cint_idx = qidx - pfvf->hw.rx_queues; + cq->cqe_cnt = qset->sqe_cnt; + } + cq->cqe_size = pfvf->qset.xqe_size; + + /* Allocate memory for CQEs */ + err = qmem_alloc(pfvf->dev, &cq->cqe, cq->cqe_cnt, cq->cqe_size); + if (err) + return err; + + /* Save CQE CPU base for faster reference */ + cq->cqe_base = cq->cqe->base; + /* In case where all RQs auras point to single pool, + * all CQs receive buffer pool also point to same pool. + */ + pool_id = ((cq->cq_type == CQ_RX) && + (pfvf->hw.rqpool_cnt != pfvf->hw.rx_queues)) ? 0 : qidx; + cq->rbpool = &qset->pool[pool_id]; + cq->refill_task_sched = false; + + /* Get memory to put this msg */ + aq = otx2_mbox_alloc_msg_nix_aq_enq(&pfvf->mbox); + if (!aq) + return -ENOMEM; + + aq->cq.ena = 1; + aq->cq.qsize = Q_SIZE(cq->cqe_cnt, 4); + aq->cq.caching = 1; + aq->cq.base = cq->cqe->iova; + aq->cq.cint_idx = cq->cint_idx; + aq->cq.cq_err_int_ena = NIX_CQERRINT_BITS; + aq->cq.qint_idx = 0; + aq->cq.avg_level = 255; + + if (qidx < pfvf->hw.rx_queues) { + aq->cq.drop = RQ_DROP_LVL_CQ(pfvf->hw.rq_skid, cq->cqe_cnt); + aq->cq.drop_ena = 1; + } + + /* Fill AQ info */ + aq->qidx = qidx; + aq->ctype = NIX_AQ_CTYPE_CQ; + aq->op = NIX_AQ_INSTOP_INIT; + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +static void otx2_pool_refill_task(struct work_struct *work) +{ + struct otx2_cq_queue *cq; + struct otx2_pool *rbpool; + struct refill_work *wrk; + int qidx, free_ptrs = 0; + struct otx2_nic *pfvf; + s64 bufptr; + + wrk = container_of(work, struct refill_work, pool_refill_work.work); + pfvf = wrk->pf; + qidx = wrk - pfvf->refill_wrk; + cq = &pfvf->qset.cq[qidx]; + rbpool = cq->rbpool; + free_ptrs = cq->pool_ptrs; + + while (cq->pool_ptrs) { + bufptr = otx2_alloc_rbuf(pfvf, rbpool, GFP_KERNEL); + if (bufptr <= 0) { + /* Schedule a WQ if we fails to free atleast half of the + * pointers else enable napi for this RQ. + */ + if (!((free_ptrs - cq->pool_ptrs) > free_ptrs / 2)) { + struct delayed_work *dwork; + + dwork = &wrk->pool_refill_work; + schedule_delayed_work(dwork, + msecs_to_jiffies(100)); + } else { + cq->refill_task_sched = false; + } + return; + } + otx2_aura_freeptr(pfvf, qidx, bufptr + OTX2_HEAD_ROOM); + cq->pool_ptrs--; + } + cq->refill_task_sched = false; +} + +int otx2_config_nix_queues(struct otx2_nic *pfvf) +{ + int qidx, err; + + /* Initialize RX queues */ + for (qidx = 0; qidx < pfvf->hw.rx_queues; qidx++) { + u16 lpb_aura = otx2_get_pool_idx(pfvf, AURA_NIX_RQ, qidx); + + err = otx2_rq_init(pfvf, qidx, lpb_aura); + if (err) + return err; + } + + /* Initialize TX queues */ + for (qidx = 0; qidx < pfvf->hw.tx_queues; qidx++) { + u16 sqb_aura = otx2_get_pool_idx(pfvf, AURA_NIX_SQ, qidx); + + err = otx2_sq_init(pfvf, qidx, sqb_aura); + if (err) + return err; + } + + /* Initialize completion queues */ + for (qidx = 0; qidx < pfvf->qset.cq_cnt; qidx++) { + err = otx2_cq_init(pfvf, qidx); + if (err) + return err; + } + + /* Initialize work queue for receive buffer refill */ + pfvf->refill_wrk = devm_kcalloc(pfvf->dev, pfvf->qset.cq_cnt, + sizeof(struct refill_work), GFP_KERNEL); + if (!pfvf->refill_wrk) + return -ENOMEM; + + for (qidx = 0; qidx < pfvf->qset.cq_cnt; qidx++) { + pfvf->refill_wrk[qidx].pf = pfvf; + INIT_DELAYED_WORK(&pfvf->refill_wrk[qidx].pool_refill_work, + otx2_pool_refill_task); + } + return 0; +} + +int otx2_config_nix(struct otx2_nic *pfvf) +{ + struct nix_lf_alloc_req *nixlf; + struct nix_lf_alloc_rsp *rsp; + int err; + + pfvf->qset.xqe_size = NIX_XQESZ_W16 ? 128 : 512; + + /* Get memory to put this msg */ + nixlf = otx2_mbox_alloc_msg_nix_lf_alloc(&pfvf->mbox); + if (!nixlf) + return -ENOMEM; + + /* Set RQ/SQ/CQ counts */ + nixlf->rq_cnt = pfvf->hw.rx_queues; + nixlf->sq_cnt = pfvf->hw.tx_queues; + nixlf->cq_cnt = pfvf->qset.cq_cnt; + nixlf->rss_sz = MAX_RSS_INDIR_TBL_SIZE; + nixlf->rss_grps = 1; /* Single RSS indir table supported, for now */ + nixlf->xqe_sz = NIX_XQESZ_W16; + /* We don't know absolute NPA LF idx attached. + * AF will replace 'RVU_DEFAULT_PF_FUNC' with + * NPA LF attached to this RVU PF/VF. + */ + nixlf->npa_func = RVU_DEFAULT_PF_FUNC; + /* Disable alignment pad, enable L2 length check, + * enable L4 TCP/UDP checksum verification. + */ + nixlf->rx_cfg = BIT_ULL(33) | BIT_ULL(35) | BIT_ULL(37); + + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) + return err; + + rsp = (struct nix_lf_alloc_rsp *)otx2_mbox_get_rsp(&pfvf->mbox.mbox, 0, + &nixlf->hdr); + if (IS_ERR(rsp)) + return PTR_ERR(rsp); + + if (rsp->qints < 1) + return -ENXIO; + + return rsp->hdr.rc; +} + +void otx2_sq_free_sqbs(struct otx2_nic *pfvf) +{ + struct otx2_qset *qset = &pfvf->qset; + struct otx2_hw *hw = &pfvf->hw; + struct otx2_snd_queue *sq; + int sqb, qidx; + u64 iova, pa; + + for (qidx = 0; qidx < hw->tx_queues; qidx++) { + sq = &qset->sq[qidx]; + if (!sq->sqb_ptrs) + continue; + for (sqb = 0; sqb < sq->sqb_count; sqb++) { + if (!sq->sqb_ptrs[sqb]) + continue; + iova = sq->sqb_ptrs[sqb]; + pa = otx2_iova_to_phys(pfvf->iommu_domain, iova); + dma_unmap_page_attrs(pfvf->dev, iova, hw->sqb_size, + DMA_FROM_DEVICE, + DMA_ATTR_SKIP_CPU_SYNC); + put_page(virt_to_page(phys_to_virt(pa))); + } + sq->sqb_count = 0; + } +} + +void otx2_free_aura_ptr(struct otx2_nic *pfvf, int type) +{ + int pool_id, pool_start = 0, pool_end = 0, size = 0; + u64 iova, pa; + + if (type == AURA_NIX_SQ) { + pool_start = otx2_get_pool_idx(pfvf, type, 0); + pool_end = pool_start + pfvf->hw.sqpool_cnt; + size = pfvf->hw.sqb_size; + } + if (type == AURA_NIX_RQ) { + pool_start = otx2_get_pool_idx(pfvf, type, 0); + pool_end = pfvf->hw.rqpool_cnt; + size = pfvf->rbsize; + } + + /* Free SQB and RQB pointers from the aura pool */ + for (pool_id = pool_start; pool_id < pool_end; pool_id++) { + iova = otx2_aura_allocptr(pfvf, pool_id); + while (iova) { + if (type == AURA_NIX_RQ) + iova -= OTX2_HEAD_ROOM; + + pa = otx2_iova_to_phys(pfvf->iommu_domain, iova); + dma_unmap_page_attrs(pfvf->dev, iova, size, + DMA_FROM_DEVICE, + DMA_ATTR_SKIP_CPU_SYNC); + put_page(virt_to_page(phys_to_virt(pa))); + iova = otx2_aura_allocptr(pfvf, pool_id); + } + } +} + +void otx2_aura_pool_free(struct otx2_nic *pfvf) +{ + struct otx2_pool *pool; + int pool_id; + + if (!pfvf->qset.pool) + return; + + for (pool_id = 0; pool_id < pfvf->hw.pool_cnt; pool_id++) { + pool = &pfvf->qset.pool[pool_id]; + qmem_free(pfvf->dev, pool->stack); + qmem_free(pfvf->dev, pool->fc_addr); + } + devm_kfree(pfvf->dev, pfvf->qset.pool); +} + +static int otx2_aura_init(struct otx2_nic *pfvf, int aura_id, + int pool_id, int numptrs) +{ + struct npa_aq_enq_req *aq; + struct otx2_pool *pool; + int err; + + pool = &pfvf->qset.pool[pool_id]; + + /* Allocate memory for HW to update Aura count. + * Alloc one cache line, so that it fits all FC_STYPE modes. + */ + if (!pool->fc_addr) { + err = qmem_alloc(pfvf->dev, &pool->fc_addr, 1, OTX2_ALIGN); + if (err) + return err; + } + + /* Initialize this aura's context via AF */ + aq = otx2_mbox_alloc_msg_npa_aq_enq(&pfvf->mbox); + if (!aq) { + /* Shared mbox memory buffer is full, flush it and retry */ + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) + return err; + aq = otx2_mbox_alloc_msg_npa_aq_enq(&pfvf->mbox); + if (!aq) + return -ENOMEM; + } + + aq->aura_id = aura_id; + /* Will be filled by AF with correct pool context address */ + aq->aura.pool_addr = pool_id; + aq->aura.pool_caching = 1; + aq->aura.shift = ilog2(numptrs) - 8; + aq->aura.count = numptrs; + aq->aura.limit = numptrs; + aq->aura.avg_level = 255; + aq->aura.ena = 1; + aq->aura.fc_ena = 1; + aq->aura.fc_addr = pool->fc_addr->iova; + aq->aura.fc_hyst_bits = 0; /* Store count on all updates */ + + /* Fill AQ info */ + aq->ctype = NPA_AQ_CTYPE_AURA; + aq->op = NPA_AQ_INSTOP_INIT; + + return 0; +} + +static int otx2_pool_init(struct otx2_nic *pfvf, u16 pool_id, + int stack_pages, int numptrs, int buf_size) +{ + struct npa_aq_enq_req *aq; + struct otx2_pool *pool; + int err; + + pool = &pfvf->qset.pool[pool_id]; + /* Alloc memory for stack which is used to store buffer pointers */ + err = qmem_alloc(pfvf->dev, &pool->stack, + stack_pages, pfvf->hw.stack_pg_bytes); + if (err) + return err; + + pool->rbsize = buf_size; + pool->rbpage_order = get_order(buf_size); + + /* Initialize this pool's context via AF */ + aq = otx2_mbox_alloc_msg_npa_aq_enq(&pfvf->mbox); + if (!aq) { + /* Shared mbox memory buffer is full, flush it and retry */ + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) { + qmem_free(pfvf->dev, pool->stack); + return err; + } + aq = otx2_mbox_alloc_msg_npa_aq_enq(&pfvf->mbox); + if (!aq) { + qmem_free(pfvf->dev, pool->stack); + return -ENOMEM; + } + } + + aq->aura_id = pool_id; + aq->pool.stack_base = pool->stack->iova; + aq->pool.stack_caching = 1; + aq->pool.ena = 1; + aq->pool.buf_size = buf_size / 128; + aq->pool.stack_max_pages = stack_pages; + aq->pool.shift = ilog2(numptrs) - 8; + aq->pool.ptr_start = 0; + aq->pool.ptr_end = ~0ULL; + + /* Fill AQ info */ + aq->ctype = NPA_AQ_CTYPE_POOL; + aq->op = NPA_AQ_INSTOP_INIT; + + return 0; +} + +int otx2_sq_aura_pool_init(struct otx2_nic *pfvf) +{ + int qidx, pool_id, stack_pages, num_sqbs; + struct otx2_qset *qset = &pfvf->qset; + struct otx2_hw *hw = &pfvf->hw; + struct otx2_snd_queue *sq; + struct otx2_pool *pool; + int err, ptr; + s64 bufptr; + + /* Calculate number of SQBs needed. + * + * For a 128byte SQE, and 4K size SQB, 31 SQEs will fit in one SQB. + * Last SQE is used for pointing to next SQB. + */ + num_sqbs = (hw->sqb_size / 128) - 1; + num_sqbs = (qset->sqe_cnt + num_sqbs) / num_sqbs; + + /* Get no of stack pages needed */ + stack_pages = + (num_sqbs + hw->stack_pg_ptrs - 1) / hw->stack_pg_ptrs; + + for (qidx = 0; qidx < hw->tx_queues; qidx++) { + pool_id = otx2_get_pool_idx(pfvf, AURA_NIX_SQ, qidx); + /* Initialize aura context */ + err = otx2_aura_init(pfvf, pool_id, pool_id, num_sqbs); + if (err) + goto fail; + + /* Initialize pool context */ + err = otx2_pool_init(pfvf, pool_id, stack_pages, + num_sqbs, hw->sqb_size); + if (err) + goto fail; + } + + /* Flush accumulated messages */ + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) + goto fail; + + /* Allocate pointers and free them to aura/pool */ + for (qidx = 0; qidx < hw->tx_queues; qidx++) { + pool_id = otx2_get_pool_idx(pfvf, AURA_NIX_SQ, qidx); + pool = &pfvf->qset.pool[pool_id]; + + sq = &qset->sq[qidx]; + sq->sqb_count = 0; + sq->sqb_ptrs = kcalloc(num_sqbs, sizeof(u64 *), GFP_KERNEL); + if (!sq->sqb_ptrs) + return -ENOMEM; + + for (ptr = 0; ptr < num_sqbs; ptr++) { + bufptr = otx2_alloc_rbuf(pfvf, pool, GFP_KERNEL); + if (bufptr <= 0) + return bufptr; + otx2_aura_freeptr(pfvf, pool_id, bufptr); + sq->sqb_ptrs[sq->sqb_count++] = (u64)bufptr; + } + otx2_get_page(pool); + } + + return 0; +fail: + otx2_mbox_reset(&pfvf->mbox.mbox, 0); + otx2_aura_pool_free(pfvf); + return err; +} + +int otx2_rq_aura_pool_init(struct otx2_nic *pfvf) +{ + struct otx2_hw *hw = &pfvf->hw; + int stack_pages, pool_id, rq; + struct otx2_pool *pool; + int err, ptr, num_ptrs; + s64 bufptr; + + num_ptrs = pfvf->qset.rqe_cnt; + + stack_pages = + (num_ptrs + hw->stack_pg_ptrs - 1) / hw->stack_pg_ptrs; + + for (rq = 0; rq < hw->rx_queues; rq++) { + pool_id = otx2_get_pool_idx(pfvf, AURA_NIX_RQ, rq); + /* Initialize aura context */ + err = otx2_aura_init(pfvf, pool_id, pool_id, num_ptrs); + if (err) + goto fail; + } + for (pool_id = 0; pool_id < hw->rqpool_cnt; pool_id++) { + err = otx2_pool_init(pfvf, pool_id, stack_pages, + num_ptrs, pfvf->rbsize); + if (err) + goto fail; + } + + /* Flush accumulated messages */ + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) + goto fail; + + /* Allocate pointers and free them to aura/pool */ + for (pool_id = 0; pool_id < hw->rqpool_cnt; pool_id++) { + pool = &pfvf->qset.pool[pool_id]; + for (ptr = 0; ptr < num_ptrs; ptr++) { + bufptr = otx2_alloc_rbuf(pfvf, pool, GFP_KERNEL); + if (bufptr <= 0) + return bufptr; + otx2_aura_freeptr(pfvf, pool_id, + bufptr + OTX2_HEAD_ROOM); + } + otx2_get_page(pool); + } + + return 0; +fail: + otx2_mbox_reset(&pfvf->mbox.mbox, 0); + otx2_aura_pool_free(pfvf); + return err; +} + +int otx2_config_npa(struct otx2_nic *pfvf) +{ + struct otx2_qset *qset = &pfvf->qset; + struct npa_lf_alloc_req *npalf; + struct otx2_hw *hw = &pfvf->hw; + int aura_cnt; + + /* Pool - Stack of free buffer pointers + * Aura - Alloc/frees pointers from/to pool for NIX DMA. + */ + + if (!hw->pool_cnt) + return -EINVAL; + + qset->pool = devm_kzalloc(pfvf->dev, sizeof(struct otx2_pool) * + hw->pool_cnt, GFP_KERNEL); + if (!qset->pool) + return -ENOMEM; + + /* Get memory to put this msg */ + npalf = otx2_mbox_alloc_msg_npa_lf_alloc(&pfvf->mbox); + if (!npalf) + return -ENOMEM; + + /* Set aura and pool counts */ + npalf->nr_pools = hw->pool_cnt; + aura_cnt = ilog2(roundup_pow_of_two(hw->pool_cnt)); + npalf->aura_sz = (aura_cnt >= ilog2(128)) ? (aura_cnt - 6) : 1; + + return otx2_sync_mbox_msg(&pfvf->mbox); +} + +int otx2_detach_resources(struct mbox *mbox) +{ + struct rsrc_detach *detach; + + otx2_mbox_lock(mbox); + detach = otx2_mbox_alloc_msg_detach_resources(mbox); + if (!detach) { + otx2_mbox_unlock(mbox); + return -ENOMEM; + } + + /* detach all */ + detach->partial = false; + + /* Send detach request to AF */ + otx2_mbox_msg_send(&mbox->mbox, 0); + otx2_mbox_unlock(mbox); + return 0; +} + +int otx2_attach_npa_nix(struct otx2_nic *pfvf) +{ + struct rsrc_attach *attach; + struct msg_req *msix; + int err; + + otx2_mbox_lock(&pfvf->mbox); + /* Get memory to put this msg */ + attach = otx2_mbox_alloc_msg_attach_resources(&pfvf->mbox); + if (!attach) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + attach->npalf = true; + attach->nixlf = true; + + /* Send attach request to AF */ + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) { + otx2_mbox_unlock(&pfvf->mbox); + return err; + } + + pfvf->nix_blkaddr = BLKADDR_NIX0; + + /* If the platform has two NIX blocks then LF may be + * allocated from NIX1. + */ + if (otx2_read64(pfvf, RVU_PF_BLOCK_ADDRX_DISC(BLKADDR_NIX1)) & 0x1FFULL) + pfvf->nix_blkaddr = BLKADDR_NIX1; + + /* Get NPA and NIX MSIX vector offsets */ + msix = otx2_mbox_alloc_msg_msix_offset(&pfvf->mbox); + if (!msix) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + err = otx2_sync_mbox_msg(&pfvf->mbox); + if (err) { + otx2_mbox_unlock(&pfvf->mbox); + return err; + } + otx2_mbox_unlock(&pfvf->mbox); + + if (pfvf->hw.npa_msixoff == MSIX_VECTOR_INVALID || + pfvf->hw.nix_msixoff == MSIX_VECTOR_INVALID) { + dev_err(pfvf->dev, + "RVUPF: Invalid MSIX vector offset for NPA/NIX\n"); + return -EINVAL; + } + + return 0; +} + +void otx2_ctx_disable(struct mbox *mbox, int type, bool npa) +{ + struct hwctx_disable_req *req; + + otx2_mbox_lock(mbox); + /* Request AQ to disable this context */ + if (npa) + req = otx2_mbox_alloc_msg_npa_hwctx_disable(mbox); + else + req = otx2_mbox_alloc_msg_nix_hwctx_disable(mbox); + + if (!req) { + otx2_mbox_unlock(mbox); + return; + } + + req->ctype = type; + + if (otx2_sync_mbox_msg(mbox)) + dev_err(mbox->pfvf->dev, "%s failed to disable context\n", + __func__); + + otx2_mbox_unlock(mbox); +} + +/* Mbox message handlers */ +void mbox_handler_cgx_stats(struct otx2_nic *pfvf, + struct cgx_stats_rsp *rsp) +{ + int id; + + for (id = 0; id < CGX_RX_STATS_COUNT; id++) + pfvf->hw.cgx_rx_stats[id] = rsp->rx_stats[id]; + for (id = 0; id < CGX_TX_STATS_COUNT; id++) + pfvf->hw.cgx_tx_stats[id] = rsp->tx_stats[id]; +} + +void mbox_handler_nix_txsch_alloc(struct otx2_nic *pf, + struct nix_txsch_alloc_rsp *rsp) +{ + int lvl, schq; + + /* Setup transmit scheduler list */ + for (lvl = 0; lvl < NIX_TXSCH_LVL_CNT; lvl++) + for (schq = 0; schq < rsp->schq[lvl]; schq++) + pf->hw.txschq_list[lvl][schq] = + rsp->schq_list[lvl][schq]; +} + +void mbox_handler_npa_lf_alloc(struct otx2_nic *pfvf, + struct npa_lf_alloc_rsp *rsp) +{ + pfvf->hw.stack_pg_ptrs = rsp->stack_pg_ptrs; + pfvf->hw.stack_pg_bytes = rsp->stack_pg_bytes; +} + +void mbox_handler_nix_lf_alloc(struct otx2_nic *pfvf, + struct nix_lf_alloc_rsp *rsp) +{ + pfvf->hw.sqb_size = rsp->sqb_size; + pfvf->hw.rx_chan_base = rsp->rx_chan_base; + pfvf->hw.tx_chan_base = rsp->tx_chan_base; + pfvf->hw.lso_tsov4_idx = rsp->lso_tsov4_idx; + pfvf->hw.lso_tsov6_idx = rsp->lso_tsov6_idx; +} + +void mbox_handler_msix_offset(struct otx2_nic *pfvf, + struct msix_offset_rsp *rsp) +{ + pfvf->hw.npa_msixoff = rsp->npa_msixoff; + pfvf->hw.nix_msixoff = rsp->nix_msixoff; +} + +void otx2_free_cints(struct otx2_nic *pfvf, int n) +{ + struct otx2_qset *qset = &pfvf->qset; + struct otx2_hw *hw = &pfvf->hw; + int irq, qidx; + + for (qidx = 0, irq = hw->nix_msixoff + NIX_LF_CINT_VEC_START; + qidx < n; + qidx++, irq++) { + int vector = pci_irq_vector(pfvf->pdev, irq); + + irq_set_affinity_hint(vector, NULL); + free_cpumask_var(hw->affinity_mask[irq]); + free_irq(vector, &qset->napi[qidx]); + } +} + +void otx2_set_cints_affinity(struct otx2_nic *pfvf) +{ + struct otx2_hw *hw = &pfvf->hw; + int vec, cpu, irq, cint; + + vec = hw->nix_msixoff + NIX_LF_CINT_VEC_START; + cpu = cpumask_first(cpu_online_mask); + + /* CQ interrupts */ + for (cint = 0; cint < pfvf->hw.cint_cnt; cint++, vec++) { + if (!alloc_cpumask_var(&hw->affinity_mask[vec], GFP_KERNEL)) + return; + + cpumask_set_cpu(cpu, hw->affinity_mask[vec]); + + irq = pci_irq_vector(pfvf->pdev, vec); + irq_set_affinity_hint(irq, hw->affinity_mask[vec]); + + cpu = cpumask_next(cpu, cpu_online_mask); + if (unlikely(cpu >= nr_cpu_ids)) + cpu = 0; + } +} + +#define M(_name, _id, _fn_name, _req_type, _rsp_type) \ +int __weak \ +otx2_mbox_up_handler_ ## _fn_name(struct otx2_nic *pfvf, \ + struct _req_type *req, \ + struct _rsp_type *rsp) \ +{ \ + /* Nothing to do here */ \ + return 0; \ +} \ +EXPORT_SYMBOL(otx2_mbox_up_handler_ ## _fn_name); +MBOX_UP_CGX_MESSAGES +#undef M diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h new file mode 100644 index 000000000000..320f3b7bf57f --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h @@ -0,0 +1,615 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef OTX2_COMMON_H +#define OTX2_COMMON_H + +#include <linux/pci.h> +#include <linux/iommu.h> + +#include <mbox.h> +#include "otx2_reg.h" +#include "otx2_txrx.h" + +/* PCI device IDs */ +#define PCI_DEVID_OCTEONTX2_RVU_PF 0xA063 + +#define PCI_SUBSYS_DEVID_96XX_RVU_PFVF 0xB200 + +/* PCI BAR nos */ +#define PCI_CFG_REG_BAR_NUM 2 +#define PCI_MBOX_BAR_NUM 4 + +#define NAME_SIZE 32 + +enum arua_mapped_qtypes { + AURA_NIX_RQ, + AURA_NIX_SQ, +}; + +/* NIX LF interrupts range*/ +#define NIX_LF_QINT_VEC_START 0x00 +#define NIX_LF_CINT_VEC_START 0x40 +#define NIX_LF_GINT_VEC 0x80 +#define NIX_LF_ERR_VEC 0x81 +#define NIX_LF_POISON_VEC 0x82 + +/* RSS configuration */ +struct otx2_rss_info { + u8 enable; + u32 flowkey_cfg; + u16 rss_size; + u8 ind_tbl[MAX_RSS_INDIR_TBL_SIZE]; +#define RSS_HASH_KEY_SIZE 44 /* 352 bit key */ + u8 key[RSS_HASH_KEY_SIZE]; +}; + +/* NIX (or NPC) RX errors */ +enum otx2_errlvl { + NPC_ERRLVL_RE, + NPC_ERRLVL_LID_LA, + NPC_ERRLVL_LID_LB, + NPC_ERRLVL_LID_LC, + NPC_ERRLVL_LID_LD, + NPC_ERRLVL_LID_LE, + NPC_ERRLVL_LID_LF, + NPC_ERRLVL_LID_LG, + NPC_ERRLVL_LID_LH, + NPC_ERRLVL_NIX = 0x0F, +}; + +enum otx2_errcodes_re { + /* NPC_ERRLVL_RE errcodes */ + ERRCODE_FCS = 0x7, + ERRCODE_FCS_RCV = 0x8, + ERRCODE_UNDERSIZE = 0x10, + ERRCODE_OVERSIZE = 0x11, + ERRCODE_OL2_LEN_MISMATCH = 0x12, + /* NPC_ERRLVL_NIX errcodes */ + ERRCODE_OL3_LEN = 0x10, + ERRCODE_OL4_LEN = 0x11, + ERRCODE_OL4_CSUM = 0x12, + ERRCODE_IL3_LEN = 0x20, + ERRCODE_IL4_LEN = 0x21, + ERRCODE_IL4_CSUM = 0x22, +}; + +/* NIX TX stats */ +enum nix_stat_lf_tx { + TX_UCAST = 0x0, + TX_BCAST = 0x1, + TX_MCAST = 0x2, + TX_DROP = 0x3, + TX_OCTS = 0x4, + TX_STATS_ENUM_LAST, +}; + +/* NIX RX stats */ +enum nix_stat_lf_rx { + RX_OCTS = 0x0, + RX_UCAST = 0x1, + RX_BCAST = 0x2, + RX_MCAST = 0x3, + RX_DROP = 0x4, + RX_DROP_OCTS = 0x5, + RX_FCS = 0x6, + RX_ERR = 0x7, + RX_DRP_BCAST = 0x8, + RX_DRP_MCAST = 0x9, + RX_DRP_L3BCAST = 0xa, + RX_DRP_L3MCAST = 0xb, + RX_STATS_ENUM_LAST, +}; + +struct otx2_dev_stats { + u64 rx_bytes; + u64 rx_frames; + u64 rx_ucast_frames; + u64 rx_bcast_frames; + u64 rx_mcast_frames; + u64 rx_drops; + + u64 tx_bytes; + u64 tx_frames; + u64 tx_ucast_frames; + u64 tx_bcast_frames; + u64 tx_mcast_frames; + u64 tx_drops; +}; + +/* Driver counted stats */ +struct otx2_drv_stats { + atomic_t rx_fcs_errs; + atomic_t rx_oversize_errs; + atomic_t rx_undersize_errs; + atomic_t rx_csum_errs; + atomic_t rx_len_errs; + atomic_t rx_other_errs; +}; + +struct mbox { + struct otx2_mbox mbox; + struct work_struct mbox_wrk; + struct otx2_mbox mbox_up; + struct work_struct mbox_up_wrk; + struct otx2_nic *pfvf; + void *bbuf_base; /* Bounce buffer for mbox memory */ + struct mutex lock; /* serialize mailbox access */ + int num_msgs; /* mbox number of messages */ + int up_num_msgs; /* mbox_up number of messages */ +}; + +struct otx2_hw { + struct pci_dev *pdev; + struct otx2_rss_info rss_info; + u16 rx_queues; + u16 tx_queues; + u16 max_queues; + u16 pool_cnt; + u16 rqpool_cnt; + u16 sqpool_cnt; + + /* NPA */ + u32 stack_pg_ptrs; /* No of ptrs per stack page */ + u32 stack_pg_bytes; /* Size of stack page */ + u16 sqb_size; + + /* NIX */ + u16 txschq_list[NIX_TXSCH_LVL_CNT][MAX_TXSCHQ_PER_FUNC]; + + /* HW settings, coalescing etc */ + u16 rx_chan_base; + u16 tx_chan_base; + u16 cq_qcount_wait; + u16 cq_ecount_wait; + u16 rq_skid; + u8 cq_time_wait; + + /* For TSO segmentation */ + u8 lso_tsov4_idx; + u8 lso_tsov6_idx; + u8 hw_tso; + + /* MSI-X */ + u8 cint_cnt; /* CQ interrupt count */ + u16 npa_msixoff; /* Offset of NPA vectors */ + u16 nix_msixoff; /* Offset of NIX vectors */ + char *irq_name; + cpumask_var_t *affinity_mask; + + /* Stats */ + struct otx2_dev_stats dev_stats; + struct otx2_drv_stats drv_stats; + u64 cgx_rx_stats[CGX_RX_STATS_COUNT]; + u64 cgx_tx_stats[CGX_TX_STATS_COUNT]; +}; + +struct refill_work { + struct delayed_work pool_refill_work; + struct otx2_nic *pf; +}; + +struct otx2_nic { + void __iomem *reg_base; + struct net_device *netdev; + void *iommu_domain; + u16 max_frs; + u16 rbsize; /* Receive buffer size */ + +#define OTX2_FLAG_INTF_DOWN BIT_ULL(2) + u64 flags; + + struct otx2_qset qset; + struct otx2_hw hw; + struct pci_dev *pdev; + struct device *dev; + + /* Mbox */ + struct mbox mbox; + struct workqueue_struct *mbox_wq; + + u16 pcifunc; /* RVU PF_FUNC */ + struct cgx_link_user_info linfo; + + u64 reset_count; + struct work_struct reset_task; + struct refill_work *refill_wrk; + + /* Ethtool stuff */ + u32 msg_enable; + + /* Block address of NIX either BLKADDR_NIX0 or BLKADDR_NIX1 */ + int nix_blkaddr; +}; + +static inline bool is_96xx_A0(struct pci_dev *pdev) +{ + return (pdev->revision == 0x00) && + (pdev->subsystem_device == PCI_SUBSYS_DEVID_96XX_RVU_PFVF); +} + +static inline bool is_96xx_B0(struct pci_dev *pdev) +{ + return (pdev->revision == 0x01) && + (pdev->subsystem_device == PCI_SUBSYS_DEVID_96XX_RVU_PFVF); +} + +static inline void otx2_setup_dev_hw_settings(struct otx2_nic *pfvf) +{ + struct otx2_hw *hw = &pfvf->hw; + + pfvf->hw.cq_time_wait = CQ_TIMER_THRESH_DEFAULT; + pfvf->hw.cq_ecount_wait = CQ_CQE_THRESH_DEFAULT; + pfvf->hw.cq_qcount_wait = CQ_QCOUNT_DEFAULT; + + hw->hw_tso = true; + + if (is_96xx_A0(pfvf->pdev)) { + hw->hw_tso = false; + + /* Time based irq coalescing is not supported */ + pfvf->hw.cq_qcount_wait = 0x0; + + /* Due to HW issue previous silicons required minimum + * 600 unused CQE to avoid CQ overflow. + */ + pfvf->hw.rq_skid = 600; + pfvf->qset.rqe_cnt = Q_COUNT(Q_SIZE_1K); + } +} + +/* Register read/write APIs */ +static inline void __iomem *otx2_get_regaddr(struct otx2_nic *nic, u64 offset) +{ + u64 blkaddr; + + switch ((offset >> RVU_FUNC_BLKADDR_SHIFT) & RVU_FUNC_BLKADDR_MASK) { + case BLKTYPE_NIX: + blkaddr = nic->nix_blkaddr; + break; + case BLKTYPE_NPA: + blkaddr = BLKADDR_NPA; + break; + default: + blkaddr = BLKADDR_RVUM; + break; + }; + + offset &= ~(RVU_FUNC_BLKADDR_MASK << RVU_FUNC_BLKADDR_SHIFT); + offset |= (blkaddr << RVU_FUNC_BLKADDR_SHIFT); + + return nic->reg_base + offset; +} + +static inline void otx2_write64(struct otx2_nic *nic, u64 offset, u64 val) +{ + void __iomem *addr = otx2_get_regaddr(nic, offset); + + writeq(val, addr); +} + +static inline u64 otx2_read64(struct otx2_nic *nic, u64 offset) +{ + void __iomem *addr = otx2_get_regaddr(nic, offset); + + return readq(addr); +} + +/* Mbox bounce buffer APIs */ +static inline int otx2_mbox_bbuf_init(struct mbox *mbox, struct pci_dev *pdev) +{ + struct otx2_mbox *otx2_mbox; + struct otx2_mbox_dev *mdev; + + mbox->bbuf_base = devm_kmalloc(&pdev->dev, MBOX_SIZE, GFP_KERNEL); + if (!mbox->bbuf_base) + return -ENOMEM; + + /* Overwrite mbox mbase to point to bounce buffer, so that PF/VF + * prepare all mbox messages in bounce buffer instead of directly + * in hw mbox memory. + */ + otx2_mbox = &mbox->mbox; + mdev = &otx2_mbox->dev[0]; + mdev->mbase = mbox->bbuf_base; + + otx2_mbox = &mbox->mbox_up; + mdev = &otx2_mbox->dev[0]; + mdev->mbase = mbox->bbuf_base; + return 0; +} + +static inline void otx2_sync_mbox_bbuf(struct otx2_mbox *mbox, int devid) +{ + u16 msgs_offset = ALIGN(sizeof(struct mbox_hdr), MBOX_MSG_ALIGN); + void *hw_mbase = mbox->hwbase + (devid * MBOX_SIZE); + struct otx2_mbox_dev *mdev = &mbox->dev[devid]; + struct mbox_hdr *hdr; + u64 msg_size; + + if (mdev->mbase == hw_mbase) + return; + + hdr = hw_mbase + mbox->rx_start; + msg_size = hdr->msg_size; + + if (msg_size > mbox->rx_size - msgs_offset) + msg_size = mbox->rx_size - msgs_offset; + + /* Copy mbox messages from mbox memory to bounce buffer */ + memcpy(mdev->mbase + mbox->rx_start, + hw_mbase + mbox->rx_start, msg_size + msgs_offset); +} + +static inline void otx2_mbox_lock_init(struct mbox *mbox) +{ + mutex_init(&mbox->lock); +} + +static inline void otx2_mbox_lock(struct mbox *mbox) +{ + mutex_lock(&mbox->lock); +} + +static inline void otx2_mbox_unlock(struct mbox *mbox) +{ + mutex_unlock(&mbox->lock); +} + +/* With the absence of API for 128-bit IO memory access for arm64, + * implement required operations at place. + */ +#if defined(CONFIG_ARM64) +static inline void otx2_write128(u64 lo, u64 hi, void __iomem *addr) +{ + __asm__ volatile("stp %x[x0], %x[x1], [%x[p1],#0]!" + ::[x0]"r"(lo), [x1]"r"(hi), [p1]"r"(addr)); +} + +static inline u64 otx2_atomic64_add(u64 incr, u64 *ptr) +{ + u64 result; + + __asm__ volatile(".cpu generic+lse\n" + "ldadd %x[i], %x[r], [%[b]]" + : [r]"=r"(result), "+m"(*ptr) + : [i]"r"(incr), [b]"r"(ptr) + : "memory"); + return result; +} + +static inline u64 otx2_lmt_flush(uint64_t addr) +{ + u64 result = 0; + + __asm__ volatile(".cpu generic+lse\n" + "ldeor xzr,%x[rf],[%[rs]]" + : [rf]"=r"(result) + : [rs]"r"(addr)); + return result; +} + +#else +#define otx2_write128(lo, hi, addr) +#define otx2_atomic64_add(incr, ptr) ({ *ptr += incr; }) +#define otx2_lmt_flush(addr) ({ 0; }) +#endif + +/* Alloc pointer from pool/aura */ +static inline u64 otx2_aura_allocptr(struct otx2_nic *pfvf, int aura) +{ + u64 *ptr = (u64 *)otx2_get_regaddr(pfvf, + NPA_LF_AURA_OP_ALLOCX(0)); + u64 incr = (u64)aura | BIT_ULL(63); + + return otx2_atomic64_add(incr, ptr); +} + +/* Free pointer to a pool/aura */ +static inline void otx2_aura_freeptr(struct otx2_nic *pfvf, + int aura, s64 buf) +{ + otx2_write128((u64)buf, (u64)aura | BIT_ULL(63), + otx2_get_regaddr(pfvf, NPA_LF_AURA_OP_FREE0)); +} + +/* Update page ref count */ +static inline void otx2_get_page(struct otx2_pool *pool) +{ + if (!pool->page) + return; + + if (pool->pageref) + page_ref_add(pool->page, pool->pageref); + pool->pageref = 0; + pool->page = NULL; +} + +static inline int otx2_get_pool_idx(struct otx2_nic *pfvf, int type, int idx) +{ + if (type == AURA_NIX_SQ) + return pfvf->hw.rqpool_cnt + idx; + + /* AURA_NIX_RQ */ + return idx; +} + +/* Mbox APIs */ +static inline int otx2_sync_mbox_msg(struct mbox *mbox) +{ + int err; + + if (!otx2_mbox_nonempty(&mbox->mbox, 0)) + return 0; + otx2_mbox_msg_send(&mbox->mbox, 0); + err = otx2_mbox_wait_for_rsp(&mbox->mbox, 0); + if (err) + return err; + + return otx2_mbox_check_rsp_msgs(&mbox->mbox, 0); +} + +static inline int otx2_sync_mbox_up_msg(struct mbox *mbox, int devid) +{ + int err; + + if (!otx2_mbox_nonempty(&mbox->mbox_up, devid)) + return 0; + otx2_mbox_msg_send(&mbox->mbox_up, devid); + err = otx2_mbox_wait_for_rsp(&mbox->mbox_up, devid); + if (err) + return err; + + return otx2_mbox_check_rsp_msgs(&mbox->mbox_up, devid); +} + +/* Use this API to send mbox msgs in atomic context + * where sleeping is not allowed + */ +static inline int otx2_sync_mbox_msg_busy_poll(struct mbox *mbox) +{ + int err; + + if (!otx2_mbox_nonempty(&mbox->mbox, 0)) + return 0; + otx2_mbox_msg_send(&mbox->mbox, 0); + err = otx2_mbox_busy_poll_for_rsp(&mbox->mbox, 0); + if (err) + return err; + + return otx2_mbox_check_rsp_msgs(&mbox->mbox, 0); +} + +#define M(_name, _id, _fn_name, _req_type, _rsp_type) \ +static struct _req_type __maybe_unused \ +*otx2_mbox_alloc_msg_ ## _fn_name(struct mbox *mbox) \ +{ \ + struct _req_type *req; \ + \ + req = (struct _req_type *)otx2_mbox_alloc_msg_rsp( \ + &mbox->mbox, 0, sizeof(struct _req_type), \ + sizeof(struct _rsp_type)); \ + if (!req) \ + return NULL; \ + req->hdr.sig = OTX2_MBOX_REQ_SIG; \ + req->hdr.id = _id; \ + return req; \ +} + +MBOX_MESSAGES +#undef M + +#define M(_name, _id, _fn_name, _req_type, _rsp_type) \ +int \ +otx2_mbox_up_handler_ ## _fn_name(struct otx2_nic *pfvf, \ + struct _req_type *req, \ + struct _rsp_type *rsp); \ + +MBOX_UP_CGX_MESSAGES +#undef M + +/* Time to wait before watchdog kicks off */ +#define OTX2_TX_TIMEOUT (100 * HZ) + +#define RVU_PFVF_PF_SHIFT 10 +#define RVU_PFVF_PF_MASK 0x3F +#define RVU_PFVF_FUNC_SHIFT 0 +#define RVU_PFVF_FUNC_MASK 0x3FF + +static inline int rvu_get_pf(u16 pcifunc) +{ + return (pcifunc >> RVU_PFVF_PF_SHIFT) & RVU_PFVF_PF_MASK; +} + +static inline dma_addr_t otx2_dma_map_page(struct otx2_nic *pfvf, + struct page *page, + size_t offset, size_t size, + enum dma_data_direction dir) +{ + dma_addr_t iova; + + iova = dma_map_page_attrs(pfvf->dev, page, + offset, size, dir, DMA_ATTR_SKIP_CPU_SYNC); + if (unlikely(dma_mapping_error(pfvf->dev, iova))) + return (dma_addr_t)NULL; + return iova; +} + +static inline void otx2_dma_unmap_page(struct otx2_nic *pfvf, + dma_addr_t addr, size_t size, + enum dma_data_direction dir) +{ + dma_unmap_page_attrs(pfvf->dev, addr, size, + dir, DMA_ATTR_SKIP_CPU_SYNC); +} + +/* MSI-X APIs */ +void otx2_free_cints(struct otx2_nic *pfvf, int n); +void otx2_set_cints_affinity(struct otx2_nic *pfvf); +int otx2_set_mac_address(struct net_device *netdev, void *p); +int otx2_hw_set_mtu(struct otx2_nic *pfvf, int mtu); +void otx2_tx_timeout(struct net_device *netdev, unsigned int txq); +void otx2_get_mac_from_af(struct net_device *netdev); +void otx2_config_irq_coalescing(struct otx2_nic *pfvf, int qidx); + +/* RVU block related APIs */ +int otx2_attach_npa_nix(struct otx2_nic *pfvf); +int otx2_detach_resources(struct mbox *mbox); +int otx2_config_npa(struct otx2_nic *pfvf); +int otx2_sq_aura_pool_init(struct otx2_nic *pfvf); +int otx2_rq_aura_pool_init(struct otx2_nic *pfvf); +void otx2_aura_pool_free(struct otx2_nic *pfvf); +void otx2_free_aura_ptr(struct otx2_nic *pfvf, int type); +void otx2_sq_free_sqbs(struct otx2_nic *pfvf); +int otx2_config_nix(struct otx2_nic *pfvf); +int otx2_config_nix_queues(struct otx2_nic *pfvf); +int otx2_txschq_config(struct otx2_nic *pfvf, int lvl); +int otx2_txsch_alloc(struct otx2_nic *pfvf); +int otx2_txschq_stop(struct otx2_nic *pfvf); +void otx2_sqb_flush(struct otx2_nic *pfvf); +dma_addr_t otx2_alloc_rbuf(struct otx2_nic *pfvf, struct otx2_pool *pool, + gfp_t gfp); +int otx2_rxtx_enable(struct otx2_nic *pfvf, bool enable); +void otx2_ctx_disable(struct mbox *mbox, int type, bool npa); +void otx2_cleanup_rx_cqes(struct otx2_nic *pfvf, struct otx2_cq_queue *cq); +void otx2_cleanup_tx_cqes(struct otx2_nic *pfvf, struct otx2_cq_queue *cq); + +/* RSS configuration APIs*/ +int otx2_rss_init(struct otx2_nic *pfvf); +int otx2_set_flowkey_cfg(struct otx2_nic *pfvf); +void otx2_set_rss_key(struct otx2_nic *pfvf); +int otx2_set_rss_table(struct otx2_nic *pfvf); + +/* Mbox handlers */ +void mbox_handler_msix_offset(struct otx2_nic *pfvf, + struct msix_offset_rsp *rsp); +void mbox_handler_npa_lf_alloc(struct otx2_nic *pfvf, + struct npa_lf_alloc_rsp *rsp); +void mbox_handler_nix_lf_alloc(struct otx2_nic *pfvf, + struct nix_lf_alloc_rsp *rsp); +void mbox_handler_nix_txsch_alloc(struct otx2_nic *pf, + struct nix_txsch_alloc_rsp *rsp); +void mbox_handler_cgx_stats(struct otx2_nic *pfvf, + struct cgx_stats_rsp *rsp); + +/* Device stats APIs */ +void otx2_get_dev_stats(struct otx2_nic *pfvf); +void otx2_get_stats64(struct net_device *netdev, + struct rtnl_link_stats64 *stats); +void otx2_update_lmac_stats(struct otx2_nic *pfvf); +int otx2_update_rq_stats(struct otx2_nic *pfvf, int qidx); +int otx2_update_sq_stats(struct otx2_nic *pfvf, int qidx); +void otx2_set_ethtool_ops(struct net_device *netdev); + +int otx2_open(struct net_device *netdev); +int otx2_stop(struct net_device *netdev); +int otx2_set_real_num_queues(struct net_device *netdev, + int tx_queues, int rx_queues); +#endif /* OTX2_COMMON_H */ diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c new file mode 100644 index 000000000000..60fcf82dd8cb --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c @@ -0,0 +1,662 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/pci.h> +#include <linux/ethtool.h> +#include <linux/stddef.h> +#include <linux/etherdevice.h> +#include <linux/log2.h> + +#include "otx2_common.h" + +#define DRV_NAME "octeontx2-nicpf" + +struct otx2_stat { + char name[ETH_GSTRING_LEN]; + unsigned int index; +}; + +/* HW device stats */ +#define OTX2_DEV_STAT(stat) { \ + .name = #stat, \ + .index = offsetof(struct otx2_dev_stats, stat) / sizeof(u64), \ +} + +static const struct otx2_stat otx2_dev_stats[] = { + OTX2_DEV_STAT(rx_ucast_frames), + OTX2_DEV_STAT(rx_bcast_frames), + OTX2_DEV_STAT(rx_mcast_frames), + + OTX2_DEV_STAT(tx_ucast_frames), + OTX2_DEV_STAT(tx_bcast_frames), + OTX2_DEV_STAT(tx_mcast_frames), +}; + +/* Driver level stats */ +#define OTX2_DRV_STAT(stat) { \ + .name = #stat, \ + .index = offsetof(struct otx2_drv_stats, stat) / sizeof(atomic_t), \ +} + +static const struct otx2_stat otx2_drv_stats[] = { + OTX2_DRV_STAT(rx_fcs_errs), + OTX2_DRV_STAT(rx_oversize_errs), + OTX2_DRV_STAT(rx_undersize_errs), + OTX2_DRV_STAT(rx_csum_errs), + OTX2_DRV_STAT(rx_len_errs), + OTX2_DRV_STAT(rx_other_errs), +}; + +static const struct otx2_stat otx2_queue_stats[] = { + { "bytes", 0 }, + { "frames", 1 }, +}; + +static const unsigned int otx2_n_dev_stats = ARRAY_SIZE(otx2_dev_stats); +static const unsigned int otx2_n_drv_stats = ARRAY_SIZE(otx2_drv_stats); +static const unsigned int otx2_n_queue_stats = ARRAY_SIZE(otx2_queue_stats); + +static void otx2_dev_open(struct net_device *netdev) +{ + otx2_open(netdev); +} + +static void otx2_dev_stop(struct net_device *netdev) +{ + otx2_stop(netdev); +} + +static void otx2_get_drvinfo(struct net_device *netdev, + struct ethtool_drvinfo *info) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + + strlcpy(info->driver, DRV_NAME, sizeof(info->driver)); + strlcpy(info->bus_info, pci_name(pfvf->pdev), sizeof(info->bus_info)); +} + +static void otx2_get_qset_strings(struct otx2_nic *pfvf, u8 **data, int qset) +{ + int start_qidx = qset * pfvf->hw.rx_queues; + int qidx, stats; + + for (qidx = 0; qidx < pfvf->hw.rx_queues; qidx++) { + for (stats = 0; stats < otx2_n_queue_stats; stats++) { + sprintf(*data, "rxq%d: %s", qidx + start_qidx, + otx2_queue_stats[stats].name); + *data += ETH_GSTRING_LEN; + } + } + for (qidx = 0; qidx < pfvf->hw.tx_queues; qidx++) { + for (stats = 0; stats < otx2_n_queue_stats; stats++) { + sprintf(*data, "txq%d: %s", qidx + start_qidx, + otx2_queue_stats[stats].name); + *data += ETH_GSTRING_LEN; + } + } +} + +static void otx2_get_strings(struct net_device *netdev, u32 sset, u8 *data) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + int stats; + + if (sset != ETH_SS_STATS) + return; + + for (stats = 0; stats < otx2_n_dev_stats; stats++) { + memcpy(data, otx2_dev_stats[stats].name, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + + for (stats = 0; stats < otx2_n_drv_stats; stats++) { + memcpy(data, otx2_drv_stats[stats].name, ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } + + otx2_get_qset_strings(pfvf, &data, 0); + + for (stats = 0; stats < CGX_RX_STATS_COUNT; stats++) { + sprintf(data, "cgx_rxstat%d: ", stats); + data += ETH_GSTRING_LEN; + } + + for (stats = 0; stats < CGX_TX_STATS_COUNT; stats++) { + sprintf(data, "cgx_txstat%d: ", stats); + data += ETH_GSTRING_LEN; + } + + strcpy(data, "reset_count"); + data += ETH_GSTRING_LEN; +} + +static void otx2_get_qset_stats(struct otx2_nic *pfvf, + struct ethtool_stats *stats, u64 **data) +{ + int stat, qidx; + + if (!pfvf) + return; + for (qidx = 0; qidx < pfvf->hw.rx_queues; qidx++) { + if (!otx2_update_rq_stats(pfvf, qidx)) { + for (stat = 0; stat < otx2_n_queue_stats; stat++) + *((*data)++) = 0; + continue; + } + for (stat = 0; stat < otx2_n_queue_stats; stat++) + *((*data)++) = ((u64 *)&pfvf->qset.rq[qidx].stats) + [otx2_queue_stats[stat].index]; + } + + for (qidx = 0; qidx < pfvf->hw.tx_queues; qidx++) { + if (!otx2_update_sq_stats(pfvf, qidx)) { + for (stat = 0; stat < otx2_n_queue_stats; stat++) + *((*data)++) = 0; + continue; + } + for (stat = 0; stat < otx2_n_queue_stats; stat++) + *((*data)++) = ((u64 *)&pfvf->qset.sq[qidx].stats) + [otx2_queue_stats[stat].index]; + } +} + +/* Get device and per queue statistics */ +static void otx2_get_ethtool_stats(struct net_device *netdev, + struct ethtool_stats *stats, u64 *data) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + int stat; + + otx2_get_dev_stats(pfvf); + for (stat = 0; stat < otx2_n_dev_stats; stat++) + *(data++) = ((u64 *)&pfvf->hw.dev_stats) + [otx2_dev_stats[stat].index]; + + for (stat = 0; stat < otx2_n_drv_stats; stat++) + *(data++) = atomic_read(&((atomic_t *)&pfvf->hw.drv_stats) + [otx2_drv_stats[stat].index]); + + otx2_get_qset_stats(pfvf, stats, &data); + otx2_update_lmac_stats(pfvf); + for (stat = 0; stat < CGX_RX_STATS_COUNT; stat++) + *(data++) = pfvf->hw.cgx_rx_stats[stat]; + for (stat = 0; stat < CGX_TX_STATS_COUNT; stat++) + *(data++) = pfvf->hw.cgx_tx_stats[stat]; + *(data++) = pfvf->reset_count; +} + +static int otx2_get_sset_count(struct net_device *netdev, int sset) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + int qstats_count; + + if (sset != ETH_SS_STATS) + return -EINVAL; + + qstats_count = otx2_n_queue_stats * + (pfvf->hw.rx_queues + pfvf->hw.tx_queues); + + return otx2_n_dev_stats + otx2_n_drv_stats + qstats_count + + CGX_RX_STATS_COUNT + CGX_TX_STATS_COUNT + 1; +} + +/* Get no of queues device supports and current queue count */ +static void otx2_get_channels(struct net_device *dev, + struct ethtool_channels *channel) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + + channel->max_rx = pfvf->hw.max_queues; + channel->max_tx = pfvf->hw.max_queues; + + channel->rx_count = pfvf->hw.rx_queues; + channel->tx_count = pfvf->hw.tx_queues; +} + +/* Set no of Tx, Rx queues to be used */ +static int otx2_set_channels(struct net_device *dev, + struct ethtool_channels *channel) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + bool if_up = netif_running(dev); + int err = 0; + + if (!channel->rx_count || !channel->tx_count) + return -EINVAL; + + if (if_up) + otx2_dev_stop(dev); + + err = otx2_set_real_num_queues(dev, channel->tx_count, + channel->rx_count); + if (err) + goto fail; + + pfvf->hw.rx_queues = channel->rx_count; + pfvf->hw.tx_queues = channel->tx_count; + pfvf->qset.cq_cnt = pfvf->hw.tx_queues + pfvf->hw.rx_queues; + +fail: + if (if_up) + otx2_dev_open(dev); + + netdev_info(dev, "Setting num Tx rings to %d, Rx rings to %d success\n", + pfvf->hw.tx_queues, pfvf->hw.rx_queues); + + return err; +} + +static void otx2_get_ringparam(struct net_device *netdev, + struct ethtool_ringparam *ring) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct otx2_qset *qs = &pfvf->qset; + + ring->rx_max_pending = Q_COUNT(Q_SIZE_MAX); + ring->rx_pending = qs->rqe_cnt ? qs->rqe_cnt : Q_COUNT(Q_SIZE_256); + ring->tx_max_pending = Q_COUNT(Q_SIZE_MAX); + ring->tx_pending = qs->sqe_cnt ? qs->sqe_cnt : Q_COUNT(Q_SIZE_4K); +} + +static int otx2_set_ringparam(struct net_device *netdev, + struct ethtool_ringparam *ring) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + bool if_up = netif_running(netdev); + struct otx2_qset *qs = &pfvf->qset; + u32 rx_count, tx_count; + + if (ring->rx_mini_pending || ring->rx_jumbo_pending) + return -EINVAL; + + /* Permitted lengths are 16 64 256 1K 4K 16K 64K 256K 1M */ + rx_count = ring->rx_pending; + /* On some silicon variants a skid or reserved CQEs are + * needed to avoid CQ overflow. + */ + if (rx_count < pfvf->hw.rq_skid) + rx_count = pfvf->hw.rq_skid; + rx_count = Q_COUNT(Q_SIZE(rx_count, 3)); + + /* Due pipelining impact minimum 2000 unused SQ CQE's + * need to be maintained to avoid CQ overflow, hence the + * minimum 4K size. + */ + tx_count = clamp_t(u32, ring->tx_pending, + Q_COUNT(Q_SIZE_4K), Q_COUNT(Q_SIZE_MAX)); + tx_count = Q_COUNT(Q_SIZE(tx_count, 3)); + + if (tx_count == qs->sqe_cnt && rx_count == qs->rqe_cnt) + return 0; + + if (if_up) + otx2_dev_stop(netdev); + + /* Assigned to the nearest possible exponent. */ + qs->sqe_cnt = tx_count; + qs->rqe_cnt = rx_count; + + if (if_up) + otx2_dev_open(netdev); + return 0; +} + +static int otx2_get_coalesce(struct net_device *netdev, + struct ethtool_coalesce *cmd) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct otx2_hw *hw = &pfvf->hw; + + cmd->rx_coalesce_usecs = hw->cq_time_wait; + cmd->rx_max_coalesced_frames = hw->cq_ecount_wait; + cmd->tx_coalesce_usecs = hw->cq_time_wait; + cmd->tx_max_coalesced_frames = hw->cq_ecount_wait; + + return 0; +} + +static int otx2_set_coalesce(struct net_device *netdev, + struct ethtool_coalesce *ec) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct otx2_hw *hw = &pfvf->hw; + int qidx; + + if (ec->use_adaptive_rx_coalesce || ec->use_adaptive_tx_coalesce || + ec->rx_coalesce_usecs_irq || ec->rx_max_coalesced_frames_irq || + ec->tx_coalesce_usecs_irq || ec->tx_max_coalesced_frames_irq || + ec->stats_block_coalesce_usecs || ec->pkt_rate_low || + ec->rx_coalesce_usecs_low || ec->rx_max_coalesced_frames_low || + ec->tx_coalesce_usecs_low || ec->tx_max_coalesced_frames_low || + ec->pkt_rate_high || ec->rx_coalesce_usecs_high || + ec->rx_max_coalesced_frames_high || ec->tx_coalesce_usecs_high || + ec->tx_max_coalesced_frames_high || ec->rate_sample_interval) + return -EOPNOTSUPP; + + if (!ec->rx_max_coalesced_frames || !ec->tx_max_coalesced_frames) + return 0; + + /* 'cq_time_wait' is 8bit and is in multiple of 100ns, + * so clamp the user given value to the range of 1 to 25usec. + */ + ec->rx_coalesce_usecs = clamp_t(u32, ec->rx_coalesce_usecs, + 1, CQ_TIMER_THRESH_MAX); + ec->tx_coalesce_usecs = clamp_t(u32, ec->tx_coalesce_usecs, + 1, CQ_TIMER_THRESH_MAX); + + /* Rx and Tx are mapped to same CQ, check which one + * is changed, if both then choose the min. + */ + if (hw->cq_time_wait == ec->rx_coalesce_usecs) + hw->cq_time_wait = ec->tx_coalesce_usecs; + else if (hw->cq_time_wait == ec->tx_coalesce_usecs) + hw->cq_time_wait = ec->rx_coalesce_usecs; + else + hw->cq_time_wait = min_t(u8, ec->rx_coalesce_usecs, + ec->tx_coalesce_usecs); + + /* Max ecount_wait supported is 16bit, + * so clamp the user given value to the range of 1 to 64k. + */ + ec->rx_max_coalesced_frames = clamp_t(u32, ec->rx_max_coalesced_frames, + 1, U16_MAX); + ec->tx_max_coalesced_frames = clamp_t(u32, ec->tx_max_coalesced_frames, + 1, U16_MAX); + + /* Rx and Tx are mapped to same CQ, check which one + * is changed, if both then choose the min. + */ + if (hw->cq_ecount_wait == ec->rx_max_coalesced_frames) + hw->cq_ecount_wait = ec->tx_max_coalesced_frames; + else if (hw->cq_ecount_wait == ec->tx_max_coalesced_frames) + hw->cq_ecount_wait = ec->rx_max_coalesced_frames; + else + hw->cq_ecount_wait = min_t(u16, ec->rx_max_coalesced_frames, + ec->tx_max_coalesced_frames); + + if (netif_running(netdev)) { + for (qidx = 0; qidx < pfvf->hw.cint_cnt; qidx++) + otx2_config_irq_coalescing(pfvf, qidx); + } + + return 0; +} + +static int otx2_get_rss_hash_opts(struct otx2_nic *pfvf, + struct ethtool_rxnfc *nfc) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + + if (!(rss->flowkey_cfg & + (NIX_FLOW_KEY_TYPE_IPV4 | NIX_FLOW_KEY_TYPE_IPV6))) + return 0; + + /* Mimimum is IPv4 and IPv6, SIP/DIP */ + nfc->data = RXH_IP_SRC | RXH_IP_DST; + + switch (nfc->flow_type) { + case TCP_V4_FLOW: + case TCP_V6_FLOW: + if (rss->flowkey_cfg & NIX_FLOW_KEY_TYPE_TCP) + nfc->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3; + break; + case UDP_V4_FLOW: + case UDP_V6_FLOW: + if (rss->flowkey_cfg & NIX_FLOW_KEY_TYPE_UDP) + nfc->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3; + break; + case SCTP_V4_FLOW: + case SCTP_V6_FLOW: + if (rss->flowkey_cfg & NIX_FLOW_KEY_TYPE_SCTP) + nfc->data |= RXH_L4_B_0_1 | RXH_L4_B_2_3; + break; + case AH_ESP_V4_FLOW: + case AH_V4_FLOW: + case ESP_V4_FLOW: + case IPV4_FLOW: + case AH_ESP_V6_FLOW: + case AH_V6_FLOW: + case ESP_V6_FLOW: + case IPV6_FLOW: + break; + default: + return -EINVAL; + } + return 0; +} + +static int otx2_set_rss_hash_opts(struct otx2_nic *pfvf, + struct ethtool_rxnfc *nfc) +{ + struct otx2_rss_info *rss = &pfvf->hw.rss_info; + u32 rxh_l4 = RXH_L4_B_0_1 | RXH_L4_B_2_3; + u32 rss_cfg = rss->flowkey_cfg; + + if (!rss->enable) { + netdev_err(pfvf->netdev, + "RSS is disabled, cannot change settings\n"); + return -EIO; + } + + /* Mimimum is IPv4 and IPv6, SIP/DIP */ + if (!(nfc->data & RXH_IP_SRC) || !(nfc->data & RXH_IP_DST)) + return -EINVAL; + + switch (nfc->flow_type) { + case TCP_V4_FLOW: + case TCP_V6_FLOW: + /* Different config for v4 and v6 is not supported. + * Both of them have to be either 4-tuple or 2-tuple. + */ + switch (nfc->data & rxh_l4) { + case 0: + rss_cfg &= ~NIX_FLOW_KEY_TYPE_TCP; + break; + case (RXH_L4_B_0_1 | RXH_L4_B_2_3): + rss_cfg |= NIX_FLOW_KEY_TYPE_TCP; + break; + default: + return -EINVAL; + } + break; + case UDP_V4_FLOW: + case UDP_V6_FLOW: + switch (nfc->data & rxh_l4) { + case 0: + rss_cfg &= ~NIX_FLOW_KEY_TYPE_UDP; + break; + case (RXH_L4_B_0_1 | RXH_L4_B_2_3): + rss_cfg |= NIX_FLOW_KEY_TYPE_UDP; + break; + default: + return -EINVAL; + } + break; + case SCTP_V4_FLOW: + case SCTP_V6_FLOW: + switch (nfc->data & rxh_l4) { + case 0: + rss_cfg &= ~NIX_FLOW_KEY_TYPE_SCTP; + break; + case (RXH_L4_B_0_1 | RXH_L4_B_2_3): + rss_cfg |= NIX_FLOW_KEY_TYPE_SCTP; + break; + default: + return -EINVAL; + } + break; + case IPV4_FLOW: + case IPV6_FLOW: + rss_cfg = NIX_FLOW_KEY_TYPE_IPV4 | NIX_FLOW_KEY_TYPE_IPV6; + break; + default: + return -EINVAL; + } + + rss->flowkey_cfg = rss_cfg; + otx2_set_flowkey_cfg(pfvf); + return 0; +} + +static int otx2_get_rxnfc(struct net_device *dev, + struct ethtool_rxnfc *nfc, u32 *rules) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + int ret = -EOPNOTSUPP; + + switch (nfc->cmd) { + case ETHTOOL_GRXRINGS: + nfc->data = pfvf->hw.rx_queues; + ret = 0; + break; + case ETHTOOL_GRXFH: + return otx2_get_rss_hash_opts(pfvf, nfc); + default: + break; + } + return ret; +} + +static int otx2_set_rxnfc(struct net_device *dev, struct ethtool_rxnfc *nfc) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + int ret = -EOPNOTSUPP; + + switch (nfc->cmd) { + case ETHTOOL_SRXFH: + ret = otx2_set_rss_hash_opts(pfvf, nfc); + break; + default: + break; + } + + return ret; +} + +static u32 otx2_get_rxfh_key_size(struct net_device *netdev) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + struct otx2_rss_info *rss; + + rss = &pfvf->hw.rss_info; + + return sizeof(rss->key); +} + +static u32 otx2_get_rxfh_indir_size(struct net_device *dev) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + + return pfvf->hw.rss_info.rss_size; +} + +/* Get RSS configuration */ +static int otx2_get_rxfh(struct net_device *dev, u32 *indir, + u8 *hkey, u8 *hfunc) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + struct otx2_rss_info *rss; + int idx; + + rss = &pfvf->hw.rss_info; + + if (indir) { + for (idx = 0; idx < rss->rss_size; idx++) + indir[idx] = rss->ind_tbl[idx]; + } + + if (hkey) + memcpy(hkey, rss->key, sizeof(rss->key)); + + if (hfunc) + *hfunc = ETH_RSS_HASH_TOP; + + return 0; +} + +/* Configure RSS table and hash key */ +static int otx2_set_rxfh(struct net_device *dev, const u32 *indir, + const u8 *hkey, const u8 hfunc) +{ + struct otx2_nic *pfvf = netdev_priv(dev); + struct otx2_rss_info *rss; + int idx; + + if (hfunc != ETH_RSS_HASH_NO_CHANGE && hfunc != ETH_RSS_HASH_TOP) + return -EOPNOTSUPP; + + rss = &pfvf->hw.rss_info; + + if (!rss->enable) { + netdev_err(dev, "RSS is disabled, cannot change settings\n"); + return -EIO; + } + + if (indir) { + for (idx = 0; idx < rss->rss_size; idx++) + rss->ind_tbl[idx] = indir[idx]; + } + + if (hkey) { + memcpy(rss->key, hkey, sizeof(rss->key)); + otx2_set_rss_key(pfvf); + } + + otx2_set_rss_table(pfvf); + return 0; +} + +static u32 otx2_get_msglevel(struct net_device *netdev) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + + return pfvf->msg_enable; +} + +static void otx2_set_msglevel(struct net_device *netdev, u32 val) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + + pfvf->msg_enable = val; +} + +static u32 otx2_get_link(struct net_device *netdev) +{ + struct otx2_nic *pfvf = netdev_priv(netdev); + + return pfvf->linfo.link_up; +} + +static const struct ethtool_ops otx2_ethtool_ops = { + .get_link = otx2_get_link, + .get_drvinfo = otx2_get_drvinfo, + .get_strings = otx2_get_strings, + .get_ethtool_stats = otx2_get_ethtool_stats, + .get_sset_count = otx2_get_sset_count, + .set_channels = otx2_set_channels, + .get_channels = otx2_get_channels, + .get_ringparam = otx2_get_ringparam, + .set_ringparam = otx2_set_ringparam, + .get_coalesce = otx2_get_coalesce, + .set_coalesce = otx2_set_coalesce, + .get_rxnfc = otx2_get_rxnfc, + .set_rxnfc = otx2_set_rxnfc, + .get_rxfh_key_size = otx2_get_rxfh_key_size, + .get_rxfh_indir_size = otx2_get_rxfh_indir_size, + .get_rxfh = otx2_get_rxfh, + .set_rxfh = otx2_set_rxfh, + .get_msglevel = otx2_get_msglevel, + .set_msglevel = otx2_set_msglevel, +}; + +void otx2_set_ethtool_ops(struct net_device *netdev) +{ + netdev->ethtool_ops = &otx2_ethtool_ops; +} diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c new file mode 100644 index 000000000000..85f9b9ba6bd5 --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c @@ -0,0 +1,1349 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Marvell OcteonTx2 RVU Physcial Function ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/module.h> +#include <linux/interrupt.h> +#include <linux/pci.h> +#include <linux/etherdevice.h> +#include <linux/of.h> +#include <linux/if_vlan.h> +#include <linux/iommu.h> +#include <net/ip.h> + +#include "otx2_reg.h" +#include "otx2_common.h" +#include "otx2_txrx.h" +#include "otx2_struct.h" + +#define DRV_NAME "octeontx2-nicpf" +#define DRV_STRING "Marvell OcteonTX2 NIC Physical Function Driver" +#define DRV_VERSION "1.0" + +/* Supported devices */ +static const struct pci_device_id otx2_pf_id_table[] = { + { PCI_DEVICE(PCI_VENDOR_ID_CAVIUM, PCI_DEVID_OCTEONTX2_RVU_PF) }, + { 0, } /* end of table */ +}; + +MODULE_AUTHOR("Marvell International Ltd."); +MODULE_DESCRIPTION(DRV_STRING); +MODULE_LICENSE("GPL v2"); +MODULE_VERSION(DRV_VERSION); +MODULE_DEVICE_TABLE(pci, otx2_pf_id_table); + +enum { + TYPE_PFAF, + TYPE_PFVF, +}; + +static int otx2_change_mtu(struct net_device *netdev, int new_mtu) +{ + bool if_up = netif_running(netdev); + int err = 0; + + if (if_up) + otx2_stop(netdev); + + netdev_info(netdev, "Changing MTU from %d to %d\n", + netdev->mtu, new_mtu); + netdev->mtu = new_mtu; + + if (if_up) + err = otx2_open(netdev); + + return err; +} + +static void otx2_queue_work(struct mbox *mw, struct workqueue_struct *mbox_wq, + int first, int mdevs, u64 intr, int type) +{ + struct otx2_mbox_dev *mdev; + struct otx2_mbox *mbox; + struct mbox_hdr *hdr; + int i; + + for (i = first; i < mdevs; i++) { + /* start from 0 */ + if (!(intr & BIT_ULL(i - first))) + continue; + + mbox = &mw->mbox; + mdev = &mbox->dev[i]; + if (type == TYPE_PFAF) + otx2_sync_mbox_bbuf(mbox, i); + hdr = mdev->mbase + mbox->rx_start; + /* The hdr->num_msgs is set to zero immediately in the interrupt + * handler to ensure that it holds a correct value next time + * when the interrupt handler is called. + * pf->mbox.num_msgs holds the data for use in pfaf_mbox_handler + * pf>mbox.up_num_msgs holds the data for use in + * pfaf_mbox_up_handler. + */ + if (hdr->num_msgs) { + mw[i].num_msgs = hdr->num_msgs; + hdr->num_msgs = 0; + if (type == TYPE_PFAF) + memset(mbox->hwbase + mbox->rx_start, 0, + ALIGN(sizeof(struct mbox_hdr), + sizeof(u64))); + + queue_work(mbox_wq, &mw[i].mbox_wrk); + } + + mbox = &mw->mbox_up; + mdev = &mbox->dev[i]; + if (type == TYPE_PFAF) + otx2_sync_mbox_bbuf(mbox, i); + hdr = mdev->mbase + mbox->rx_start; + if (hdr->num_msgs) { + mw[i].up_num_msgs = hdr->num_msgs; + hdr->num_msgs = 0; + if (type == TYPE_PFAF) + memset(mbox->hwbase + mbox->rx_start, 0, + ALIGN(sizeof(struct mbox_hdr), + sizeof(u64))); + + queue_work(mbox_wq, &mw[i].mbox_up_wrk); + } + } +} + +static void otx2_process_pfaf_mbox_msg(struct otx2_nic *pf, + struct mbox_msghdr *msg) +{ + if (msg->id >= MBOX_MSG_MAX) { + dev_err(pf->dev, + "Mbox msg with unknown ID 0x%x\n", msg->id); + return; + } + + if (msg->sig != OTX2_MBOX_RSP_SIG) { + dev_err(pf->dev, + "Mbox msg with wrong signature %x, ID 0x%x\n", + msg->sig, msg->id); + return; + } + + switch (msg->id) { + case MBOX_MSG_READY: + pf->pcifunc = msg->pcifunc; + break; + case MBOX_MSG_MSIX_OFFSET: + mbox_handler_msix_offset(pf, (struct msix_offset_rsp *)msg); + break; + case MBOX_MSG_NPA_LF_ALLOC: + mbox_handler_npa_lf_alloc(pf, (struct npa_lf_alloc_rsp *)msg); + break; + case MBOX_MSG_NIX_LF_ALLOC: + mbox_handler_nix_lf_alloc(pf, (struct nix_lf_alloc_rsp *)msg); + break; + case MBOX_MSG_NIX_TXSCH_ALLOC: + mbox_handler_nix_txsch_alloc(pf, + (struct nix_txsch_alloc_rsp *)msg); + break; + case MBOX_MSG_CGX_STATS: + mbox_handler_cgx_stats(pf, (struct cgx_stats_rsp *)msg); + break; + default: + if (msg->rc) + dev_err(pf->dev, + "Mbox msg response has err %d, ID 0x%x\n", + msg->rc, msg->id); + break; + } +} + +static void otx2_pfaf_mbox_handler(struct work_struct *work) +{ + struct otx2_mbox_dev *mdev; + struct mbox_hdr *rsp_hdr; + struct mbox_msghdr *msg; + struct otx2_mbox *mbox; + struct mbox *af_mbox; + struct otx2_nic *pf; + int offset, id; + + af_mbox = container_of(work, struct mbox, mbox_wrk); + mbox = &af_mbox->mbox; + mdev = &mbox->dev[0]; + rsp_hdr = (struct mbox_hdr *)(mdev->mbase + mbox->rx_start); + + offset = mbox->rx_start + ALIGN(sizeof(*rsp_hdr), MBOX_MSG_ALIGN); + pf = af_mbox->pfvf; + + for (id = 0; id < af_mbox->num_msgs; id++) { + msg = (struct mbox_msghdr *)(mdev->mbase + offset); + otx2_process_pfaf_mbox_msg(pf, msg); + offset = mbox->rx_start + msg->next_msgoff; + mdev->msgs_acked++; + } + + otx2_mbox_reset(mbox, 0); +} + +static void otx2_handle_link_event(struct otx2_nic *pf) +{ + struct cgx_link_user_info *linfo = &pf->linfo; + struct net_device *netdev = pf->netdev; + + pr_info("%s NIC Link is %s %d Mbps %s duplex\n", netdev->name, + linfo->link_up ? "UP" : "DOWN", linfo->speed, + linfo->full_duplex ? "Full" : "Half"); + if (linfo->link_up) { + netif_carrier_on(netdev); + netif_tx_start_all_queues(netdev); + } else { + netif_tx_stop_all_queues(netdev); + netif_carrier_off(netdev); + } +} + +int otx2_mbox_up_handler_cgx_link_event(struct otx2_nic *pf, + struct cgx_link_info_msg *msg, + struct msg_rsp *rsp) +{ + /* Copy the link info sent by AF */ + pf->linfo = msg->link_info; + + /* interface has not been fully configured yet */ + if (pf->flags & OTX2_FLAG_INTF_DOWN) + return 0; + + otx2_handle_link_event(pf); + return 0; +} + +static int otx2_process_mbox_msg_up(struct otx2_nic *pf, + struct mbox_msghdr *req) +{ + /* Check if valid, if not reply with a invalid msg */ + if (req->sig != OTX2_MBOX_REQ_SIG) { + otx2_reply_invalid_msg(&pf->mbox.mbox_up, 0, 0, req->id); + return -ENODEV; + } + + switch (req->id) { +#define M(_name, _id, _fn_name, _req_type, _rsp_type) \ + case _id: { \ + struct _rsp_type *rsp; \ + int err; \ + \ + rsp = (struct _rsp_type *)otx2_mbox_alloc_msg( \ + &pf->mbox.mbox_up, 0, \ + sizeof(struct _rsp_type)); \ + if (!rsp) \ + return -ENOMEM; \ + \ + rsp->hdr.id = _id; \ + rsp->hdr.sig = OTX2_MBOX_RSP_SIG; \ + rsp->hdr.pcifunc = 0; \ + rsp->hdr.rc = 0; \ + \ + err = otx2_mbox_up_handler_ ## _fn_name( \ + pf, (struct _req_type *)req, rsp); \ + return err; \ + } +MBOX_UP_CGX_MESSAGES +#undef M + break; + default: + otx2_reply_invalid_msg(&pf->mbox.mbox_up, 0, 0, req->id); + return -ENODEV; + } + return 0; +} + +static void otx2_pfaf_mbox_up_handler(struct work_struct *work) +{ + struct mbox *af_mbox = container_of(work, struct mbox, mbox_up_wrk); + struct otx2_mbox *mbox = &af_mbox->mbox_up; + struct otx2_mbox_dev *mdev = &mbox->dev[0]; + struct otx2_nic *pf = af_mbox->pfvf; + int offset, id, devid = 0; + struct mbox_hdr *rsp_hdr; + struct mbox_msghdr *msg; + + rsp_hdr = (struct mbox_hdr *)(mdev->mbase + mbox->rx_start); + + offset = mbox->rx_start + ALIGN(sizeof(*rsp_hdr), MBOX_MSG_ALIGN); + + for (id = 0; id < af_mbox->up_num_msgs; id++) { + msg = (struct mbox_msghdr *)(mdev->mbase + offset); + + devid = msg->pcifunc & RVU_PFVF_FUNC_MASK; + /* Skip processing VF's messages */ + if (!devid) + otx2_process_mbox_msg_up(pf, msg); + offset = mbox->rx_start + msg->next_msgoff; + } + + otx2_mbox_msg_send(mbox, 0); +} + +static irqreturn_t otx2_pfaf_mbox_intr_handler(int irq, void *pf_irq) +{ + struct otx2_nic *pf = (struct otx2_nic *)pf_irq; + struct mbox *mbox; + + /* Clear the IRQ */ + otx2_write64(pf, RVU_PF_INT, BIT_ULL(0)); + + mbox = &pf->mbox; + otx2_queue_work(mbox, pf->mbox_wq, 0, 1, 1, TYPE_PFAF); + + return IRQ_HANDLED; +} + +static void otx2_disable_mbox_intr(struct otx2_nic *pf) +{ + int vector = pci_irq_vector(pf->pdev, RVU_PF_INT_VEC_AFPF_MBOX); + + /* Disable AF => PF mailbox IRQ */ + otx2_write64(pf, RVU_PF_INT_ENA_W1C, BIT_ULL(0)); + free_irq(vector, pf); +} + +static int otx2_register_mbox_intr(struct otx2_nic *pf, bool probe_af) +{ + struct otx2_hw *hw = &pf->hw; + struct msg_req *req; + char *irq_name; + int err; + + /* Register mailbox interrupt handler */ + irq_name = &hw->irq_name[RVU_PF_INT_VEC_AFPF_MBOX * NAME_SIZE]; + snprintf(irq_name, NAME_SIZE, "RVUPFAF Mbox"); + err = request_irq(pci_irq_vector(pf->pdev, RVU_PF_INT_VEC_AFPF_MBOX), + otx2_pfaf_mbox_intr_handler, 0, irq_name, pf); + if (err) { + dev_err(pf->dev, + "RVUPF: IRQ registration failed for PFAF mbox irq\n"); + return err; + } + + /* Enable mailbox interrupt for msgs coming from AF. + * First clear to avoid spurious interrupts, if any. + */ + otx2_write64(pf, RVU_PF_INT, BIT_ULL(0)); + otx2_write64(pf, RVU_PF_INT_ENA_W1S, BIT_ULL(0)); + + if (!probe_af) + return 0; + + /* Check mailbox communication with AF */ + req = otx2_mbox_alloc_msg_ready(&pf->mbox); + if (!req) { + otx2_disable_mbox_intr(pf); + return -ENOMEM; + } + err = otx2_sync_mbox_msg(&pf->mbox); + if (err) { + dev_warn(pf->dev, + "AF not responding to mailbox, deferring probe\n"); + otx2_disable_mbox_intr(pf); + return -EPROBE_DEFER; + } + + return 0; +} + +static void otx2_pfaf_mbox_destroy(struct otx2_nic *pf) +{ + struct mbox *mbox = &pf->mbox; + + if (pf->mbox_wq) { + flush_workqueue(pf->mbox_wq); + destroy_workqueue(pf->mbox_wq); + pf->mbox_wq = NULL; + } + + if (mbox->mbox.hwbase) + iounmap((void __iomem *)mbox->mbox.hwbase); + + otx2_mbox_destroy(&mbox->mbox); + otx2_mbox_destroy(&mbox->mbox_up); +} + +static int otx2_pfaf_mbox_init(struct otx2_nic *pf) +{ + struct mbox *mbox = &pf->mbox; + void __iomem *hwbase; + int err; + + mbox->pfvf = pf; + pf->mbox_wq = alloc_workqueue("otx2_pfaf_mailbox", + WQ_UNBOUND | WQ_HIGHPRI | + WQ_MEM_RECLAIM, 1); + if (!pf->mbox_wq) + return -ENOMEM; + + /* Mailbox is a reserved memory (in RAM) region shared between + * admin function (i.e AF) and this PF, shouldn't be mapped as + * device memory to allow unaligned accesses. + */ + hwbase = ioremap_wc(pci_resource_start(pf->pdev, PCI_MBOX_BAR_NUM), + pci_resource_len(pf->pdev, PCI_MBOX_BAR_NUM)); + if (!hwbase) { + dev_err(pf->dev, "Unable to map PFAF mailbox region\n"); + err = -ENOMEM; + goto exit; + } + + err = otx2_mbox_init(&mbox->mbox, hwbase, pf->pdev, pf->reg_base, + MBOX_DIR_PFAF, 1); + if (err) + goto exit; + + err = otx2_mbox_init(&mbox->mbox_up, hwbase, pf->pdev, pf->reg_base, + MBOX_DIR_PFAF_UP, 1); + if (err) + goto exit; + + err = otx2_mbox_bbuf_init(mbox, pf->pdev); + if (err) + goto exit; + + INIT_WORK(&mbox->mbox_wrk, otx2_pfaf_mbox_handler); + INIT_WORK(&mbox->mbox_up_wrk, otx2_pfaf_mbox_up_handler); + otx2_mbox_lock_init(&pf->mbox); + + return 0; +exit: + otx2_pfaf_mbox_destroy(pf); + return err; +} + +static int otx2_cgx_config_linkevents(struct otx2_nic *pf, bool enable) +{ + struct msg_req *msg; + int err; + + otx2_mbox_lock(&pf->mbox); + if (enable) + msg = otx2_mbox_alloc_msg_cgx_start_linkevents(&pf->mbox); + else + msg = otx2_mbox_alloc_msg_cgx_stop_linkevents(&pf->mbox); + + if (!msg) { + otx2_mbox_unlock(&pf->mbox); + return -ENOMEM; + } + + err = otx2_sync_mbox_msg(&pf->mbox); + otx2_mbox_unlock(&pf->mbox); + return err; +} + +static int otx2_cgx_config_loopback(struct otx2_nic *pf, bool enable) +{ + struct msg_req *msg; + int err; + + otx2_mbox_lock(&pf->mbox); + if (enable) + msg = otx2_mbox_alloc_msg_cgx_intlbk_enable(&pf->mbox); + else + msg = otx2_mbox_alloc_msg_cgx_intlbk_disable(&pf->mbox); + + if (!msg) { + otx2_mbox_unlock(&pf->mbox); + return -ENOMEM; + } + + err = otx2_sync_mbox_msg(&pf->mbox); + otx2_mbox_unlock(&pf->mbox); + return err; +} + +int otx2_set_real_num_queues(struct net_device *netdev, + int tx_queues, int rx_queues) +{ + int err; + + err = netif_set_real_num_tx_queues(netdev, tx_queues); + if (err) { + netdev_err(netdev, + "Failed to set no of Tx queues: %d\n", tx_queues); + return err; + } + + err = netif_set_real_num_rx_queues(netdev, rx_queues); + if (err) + netdev_err(netdev, + "Failed to set no of Rx queues: %d\n", rx_queues); + return err; +} + +static irqreturn_t otx2_q_intr_handler(int irq, void *data) +{ + struct otx2_nic *pf = data; + u64 val, *ptr; + u64 qidx = 0; + + /* CQ */ + for (qidx = 0; qidx < pf->qset.cq_cnt; qidx++) { + ptr = otx2_get_regaddr(pf, NIX_LF_CQ_OP_INT); + val = otx2_atomic64_add((qidx << 44), ptr); + + otx2_write64(pf, NIX_LF_CQ_OP_INT, (qidx << 44) | + (val & NIX_CQERRINT_BITS)); + if (!(val & (NIX_CQERRINT_BITS | BIT_ULL(42)))) + continue; + + if (val & BIT_ULL(42)) { + netdev_err(pf->netdev, "CQ%lld: error reading NIX_LF_CQ_OP_INT, NIX_LF_ERR_INT 0x%llx\n", + qidx, otx2_read64(pf, NIX_LF_ERR_INT)); + } else { + if (val & BIT_ULL(NIX_CQERRINT_DOOR_ERR)) + netdev_err(pf->netdev, "CQ%lld: Doorbell error", + qidx); + if (val & BIT_ULL(NIX_CQERRINT_CQE_FAULT)) + netdev_err(pf->netdev, "CQ%lld: Memory fault on CQE write to LLC/DRAM", + qidx); + } + + schedule_work(&pf->reset_task); + } + + /* SQ */ + for (qidx = 0; qidx < pf->hw.tx_queues; qidx++) { + ptr = otx2_get_regaddr(pf, NIX_LF_SQ_OP_INT); + val = otx2_atomic64_add((qidx << 44), ptr); + otx2_write64(pf, NIX_LF_SQ_OP_INT, (qidx << 44) | + (val & NIX_SQINT_BITS)); + + if (!(val & (NIX_SQINT_BITS | BIT_ULL(42)))) + continue; + + if (val & BIT_ULL(42)) { + netdev_err(pf->netdev, "SQ%lld: error reading NIX_LF_SQ_OP_INT, NIX_LF_ERR_INT 0x%llx\n", + qidx, otx2_read64(pf, NIX_LF_ERR_INT)); + } else { + if (val & BIT_ULL(NIX_SQINT_LMT_ERR)) { + netdev_err(pf->netdev, "SQ%lld: LMT store error NIX_LF_SQ_OP_ERR_DBG:0x%llx", + qidx, + otx2_read64(pf, + NIX_LF_SQ_OP_ERR_DBG)); + otx2_write64(pf, NIX_LF_SQ_OP_ERR_DBG, + BIT_ULL(44)); + } + if (val & BIT_ULL(NIX_SQINT_MNQ_ERR)) { + netdev_err(pf->netdev, "SQ%lld: Meta-descriptor enqueue error NIX_LF_MNQ_ERR_DGB:0x%llx\n", + qidx, + otx2_read64(pf, NIX_LF_MNQ_ERR_DBG)); + otx2_write64(pf, NIX_LF_MNQ_ERR_DBG, + BIT_ULL(44)); + } + if (val & BIT_ULL(NIX_SQINT_SEND_ERR)) { + netdev_err(pf->netdev, "SQ%lld: Send error, NIX_LF_SEND_ERR_DBG 0x%llx", + qidx, + otx2_read64(pf, + NIX_LF_SEND_ERR_DBG)); + otx2_write64(pf, NIX_LF_SEND_ERR_DBG, + BIT_ULL(44)); + } + if (val & BIT_ULL(NIX_SQINT_SQB_ALLOC_FAIL)) + netdev_err(pf->netdev, "SQ%lld: SQB allocation failed", + qidx); + } + + schedule_work(&pf->reset_task); + } + + return IRQ_HANDLED; +} + +static irqreturn_t otx2_cq_intr_handler(int irq, void *cq_irq) +{ + struct otx2_cq_poll *cq_poll = (struct otx2_cq_poll *)cq_irq; + struct otx2_nic *pf = (struct otx2_nic *)cq_poll->dev; + int qidx = cq_poll->cint_idx; + + /* Disable interrupts. + * + * Completion interrupts behave in a level-triggered interrupt + * fashion, and hence have to be cleared only after it is serviced. + */ + otx2_write64(pf, NIX_LF_CINTX_ENA_W1C(qidx), BIT_ULL(0)); + + /* Schedule NAPI */ + napi_schedule_irqoff(&cq_poll->napi); + + return IRQ_HANDLED; +} + +static void otx2_disable_napi(struct otx2_nic *pf) +{ + struct otx2_qset *qset = &pf->qset; + struct otx2_cq_poll *cq_poll; + int qidx; + + for (qidx = 0; qidx < pf->hw.cint_cnt; qidx++) { + cq_poll = &qset->napi[qidx]; + napi_disable(&cq_poll->napi); + netif_napi_del(&cq_poll->napi); + } +} + +static void otx2_free_cq_res(struct otx2_nic *pf) +{ + struct otx2_qset *qset = &pf->qset; + struct otx2_cq_queue *cq; + int qidx; + + /* Disable CQs */ + otx2_ctx_disable(&pf->mbox, NIX_AQ_CTYPE_CQ, false); + for (qidx = 0; qidx < qset->cq_cnt; qidx++) { + cq = &qset->cq[qidx]; + qmem_free(pf->dev, cq->cqe); + } +} + +static void otx2_free_sq_res(struct otx2_nic *pf) +{ + struct otx2_qset *qset = &pf->qset; + struct otx2_snd_queue *sq; + int qidx; + + /* Disable SQs */ + otx2_ctx_disable(&pf->mbox, NIX_AQ_CTYPE_SQ, false); + /* Free SQB pointers */ + otx2_sq_free_sqbs(pf); + for (qidx = 0; qidx < pf->hw.tx_queues; qidx++) { + sq = &qset->sq[qidx]; + qmem_free(pf->dev, sq->sqe); + qmem_free(pf->dev, sq->tso_hdrs); + kfree(sq->sg); + kfree(sq->sqb_ptrs); + } +} + +static int otx2_init_hw_resources(struct otx2_nic *pf) +{ + struct mbox *mbox = &pf->mbox; + struct otx2_hw *hw = &pf->hw; + struct msg_req *req; + int err = 0, lvl; + + /* Set required NPA LF's pool counts + * Auras and Pools are used in a 1:1 mapping, + * so, aura count = pool count. + */ + hw->rqpool_cnt = hw->rx_queues; + hw->sqpool_cnt = hw->tx_queues; + hw->pool_cnt = hw->rqpool_cnt + hw->sqpool_cnt; + + /* Get the size of receive buffers to allocate */ + pf->rbsize = RCV_FRAG_LEN(pf->netdev->mtu + OTX2_ETH_HLEN); + + otx2_mbox_lock(mbox); + /* NPA init */ + err = otx2_config_npa(pf); + if (err) + goto exit; + + /* NIX init */ + err = otx2_config_nix(pf); + if (err) + goto err_free_npa_lf; + + /* Init Auras and pools used by NIX RQ, for free buffer ptrs */ + err = otx2_rq_aura_pool_init(pf); + if (err) { + otx2_mbox_unlock(mbox); + goto err_free_nix_lf; + } + /* Init Auras and pools used by NIX SQ, for queueing SQEs */ + err = otx2_sq_aura_pool_init(pf); + if (err) { + otx2_mbox_unlock(mbox); + goto err_free_rq_ptrs; + } + + err = otx2_txsch_alloc(pf); + if (err) { + otx2_mbox_unlock(mbox); + goto err_free_sq_ptrs; + } + + err = otx2_config_nix_queues(pf); + if (err) { + otx2_mbox_unlock(mbox); + goto err_free_txsch; + } + for (lvl = 0; lvl < NIX_TXSCH_LVL_CNT; lvl++) { + err = otx2_txschq_config(pf, lvl); + if (err) { + otx2_mbox_unlock(mbox); + goto err_free_nix_queues; + } + } + otx2_mbox_unlock(mbox); + return err; + +err_free_nix_queues: + otx2_free_sq_res(pf); + otx2_free_cq_res(pf); + otx2_ctx_disable(mbox, NIX_AQ_CTYPE_RQ, false); +err_free_txsch: + if (otx2_txschq_stop(pf)) + dev_err(pf->dev, "%s failed to stop TX schedulers\n", __func__); +err_free_sq_ptrs: + otx2_sq_free_sqbs(pf); +err_free_rq_ptrs: + otx2_free_aura_ptr(pf, AURA_NIX_RQ); + otx2_ctx_disable(mbox, NPA_AQ_CTYPE_POOL, true); + otx2_ctx_disable(mbox, NPA_AQ_CTYPE_AURA, true); + otx2_aura_pool_free(pf); +err_free_nix_lf: + otx2_mbox_lock(mbox); + req = otx2_mbox_alloc_msg_nix_lf_free(mbox); + if (req) { + if (otx2_sync_mbox_msg(mbox)) + dev_err(pf->dev, "%s failed to free nixlf\n", __func__); + } +err_free_npa_lf: + /* Reset NPA LF */ + req = otx2_mbox_alloc_msg_npa_lf_free(mbox); + if (req) { + if (otx2_sync_mbox_msg(mbox)) + dev_err(pf->dev, "%s failed to free npalf\n", __func__); + } +exit: + otx2_mbox_unlock(mbox); + return err; +} + +static void otx2_free_hw_resources(struct otx2_nic *pf) +{ + struct otx2_qset *qset = &pf->qset; + struct mbox *mbox = &pf->mbox; + struct otx2_cq_queue *cq; + struct msg_req *req; + int qidx, err; + + /* Ensure all SQE are processed */ + otx2_sqb_flush(pf); + + /* Stop transmission */ + err = otx2_txschq_stop(pf); + if (err) + dev_err(pf->dev, "RVUPF: Failed to stop/free TX schedulers\n"); + + /* Disable RQs */ + otx2_ctx_disable(mbox, NIX_AQ_CTYPE_RQ, false); + + /*Dequeue all CQEs */ + for (qidx = 0; qidx < qset->cq_cnt; qidx++) { + cq = &qset->cq[qidx]; + if (cq->cq_type == CQ_RX) + otx2_cleanup_rx_cqes(pf, cq); + else + otx2_cleanup_tx_cqes(pf, cq); + } + + otx2_free_sq_res(pf); + + /* Free RQ buffer pointers*/ + otx2_free_aura_ptr(pf, AURA_NIX_RQ); + + otx2_free_cq_res(pf); + + otx2_mbox_lock(mbox); + /* Reset NIX LF */ + req = otx2_mbox_alloc_msg_nix_lf_free(mbox); + if (req) { + if (otx2_sync_mbox_msg(mbox)) + dev_err(pf->dev, "%s failed to free nixlf\n", __func__); + } + otx2_mbox_unlock(mbox); + + /* Disable NPA Pool and Aura hw context */ + otx2_ctx_disable(mbox, NPA_AQ_CTYPE_POOL, true); + otx2_ctx_disable(mbox, NPA_AQ_CTYPE_AURA, true); + otx2_aura_pool_free(pf); + + otx2_mbox_lock(mbox); + /* Reset NPA LF */ + req = otx2_mbox_alloc_msg_npa_lf_free(mbox); + if (req) { + if (otx2_sync_mbox_msg(mbox)) + dev_err(pf->dev, "%s failed to free npalf\n", __func__); + } + otx2_mbox_unlock(mbox); +} + +int otx2_open(struct net_device *netdev) +{ + struct otx2_nic *pf = netdev_priv(netdev); + struct otx2_cq_poll *cq_poll = NULL; + struct otx2_qset *qset = &pf->qset; + int err = 0, qidx, vec; + char *irq_name; + + netif_carrier_off(netdev); + + pf->qset.cq_cnt = pf->hw.rx_queues + pf->hw.tx_queues; + /* RQ and SQs are mapped to different CQs, + * so find out max CQ IRQs (i.e CINTs) needed. + */ + pf->hw.cint_cnt = max(pf->hw.rx_queues, pf->hw.tx_queues); + qset->napi = kcalloc(pf->hw.cint_cnt, sizeof(*cq_poll), GFP_KERNEL); + if (!qset->napi) + return -ENOMEM; + + /* CQ size of RQ */ + qset->rqe_cnt = qset->rqe_cnt ? qset->rqe_cnt : Q_COUNT(Q_SIZE_256); + /* CQ size of SQ */ + qset->sqe_cnt = qset->sqe_cnt ? qset->sqe_cnt : Q_COUNT(Q_SIZE_4K); + + err = -ENOMEM; + qset->cq = kcalloc(pf->qset.cq_cnt, + sizeof(struct otx2_cq_queue), GFP_KERNEL); + if (!qset->cq) + goto err_free_mem; + + qset->sq = kcalloc(pf->hw.tx_queues, + sizeof(struct otx2_snd_queue), GFP_KERNEL); + if (!qset->sq) + goto err_free_mem; + + qset->rq = kcalloc(pf->hw.rx_queues, + sizeof(struct otx2_rcv_queue), GFP_KERNEL); + if (!qset->rq) + goto err_free_mem; + + err = otx2_init_hw_resources(pf); + if (err) + goto err_free_mem; + + /* Register NAPI handler */ + for (qidx = 0; qidx < pf->hw.cint_cnt; qidx++) { + cq_poll = &qset->napi[qidx]; + cq_poll->cint_idx = qidx; + /* RQ0 & SQ0 are mapped to CINT0 and so on.. + * 'cq_ids[0]' points to RQ's CQ and + * 'cq_ids[1]' points to SQ's CQ and + */ + cq_poll->cq_ids[CQ_RX] = + (qidx < pf->hw.rx_queues) ? qidx : CINT_INVALID_CQ; + cq_poll->cq_ids[CQ_TX] = (qidx < pf->hw.tx_queues) ? + qidx + pf->hw.rx_queues : CINT_INVALID_CQ; + cq_poll->dev = (void *)pf; + netif_napi_add(netdev, &cq_poll->napi, + otx2_napi_handler, NAPI_POLL_WEIGHT); + napi_enable(&cq_poll->napi); + } + + /* Set maximum frame size allowed in HW */ + err = otx2_hw_set_mtu(pf, netdev->mtu); + if (err) + goto err_disable_napi; + + /* Initialize RSS */ + err = otx2_rss_init(pf); + if (err) + goto err_disable_napi; + + /* Register Queue IRQ handlers */ + vec = pf->hw.nix_msixoff + NIX_LF_QINT_VEC_START; + irq_name = &pf->hw.irq_name[vec * NAME_SIZE]; + + snprintf(irq_name, NAME_SIZE, "%s-qerr", pf->netdev->name); + + err = request_irq(pci_irq_vector(pf->pdev, vec), + otx2_q_intr_handler, 0, irq_name, pf); + if (err) { + dev_err(pf->dev, + "RVUPF%d: IRQ registration failed for QERR\n", + rvu_get_pf(pf->pcifunc)); + goto err_disable_napi; + } + + /* Enable QINT IRQ */ + otx2_write64(pf, NIX_LF_QINTX_ENA_W1S(0), BIT_ULL(0)); + + /* Register CQ IRQ handlers */ + vec = pf->hw.nix_msixoff + NIX_LF_CINT_VEC_START; + for (qidx = 0; qidx < pf->hw.cint_cnt; qidx++) { + irq_name = &pf->hw.irq_name[vec * NAME_SIZE]; + + snprintf(irq_name, NAME_SIZE, "%s-rxtx-%d", pf->netdev->name, + qidx); + + err = request_irq(pci_irq_vector(pf->pdev, vec), + otx2_cq_intr_handler, 0, irq_name, + &qset->napi[qidx]); + if (err) { + dev_err(pf->dev, + "RVUPF%d: IRQ registration failed for CQ%d\n", + rvu_get_pf(pf->pcifunc), qidx); + goto err_free_cints; + } + vec++; + + otx2_config_irq_coalescing(pf, qidx); + + /* Enable CQ IRQ */ + otx2_write64(pf, NIX_LF_CINTX_INT(qidx), BIT_ULL(0)); + otx2_write64(pf, NIX_LF_CINTX_ENA_W1S(qidx), BIT_ULL(0)); + } + + otx2_set_cints_affinity(pf); + + pf->flags &= ~OTX2_FLAG_INTF_DOWN; + /* 'intf_down' may be checked on any cpu */ + smp_wmb(); + + /* we have already received link status notification */ + if (pf->linfo.link_up && !(pf->pcifunc & RVU_PFVF_FUNC_MASK)) + otx2_handle_link_event(pf); + + err = otx2_rxtx_enable(pf, true); + if (err) + goto err_free_cints; + + return 0; + +err_free_cints: + otx2_free_cints(pf, qidx); + vec = pci_irq_vector(pf->pdev, + pf->hw.nix_msixoff + NIX_LF_QINT_VEC_START); + otx2_write64(pf, NIX_LF_QINTX_ENA_W1C(0), BIT_ULL(0)); + synchronize_irq(vec); + free_irq(vec, pf); +err_disable_napi: + otx2_disable_napi(pf); + otx2_free_hw_resources(pf); +err_free_mem: + kfree(qset->sq); + kfree(qset->cq); + kfree(qset->rq); + kfree(qset->napi); + return err; +} + +int otx2_stop(struct net_device *netdev) +{ + struct otx2_nic *pf = netdev_priv(netdev); + struct otx2_cq_poll *cq_poll = NULL; + struct otx2_qset *qset = &pf->qset; + int qidx, vec, wrk; + + netif_carrier_off(netdev); + netif_tx_stop_all_queues(netdev); + + pf->flags |= OTX2_FLAG_INTF_DOWN; + /* 'intf_down' may be checked on any cpu */ + smp_wmb(); + + /* First stop packet Rx/Tx */ + otx2_rxtx_enable(pf, false); + + /* Cleanup Queue IRQ */ + vec = pci_irq_vector(pf->pdev, + pf->hw.nix_msixoff + NIX_LF_QINT_VEC_START); + otx2_write64(pf, NIX_LF_QINTX_ENA_W1C(0), BIT_ULL(0)); + synchronize_irq(vec); + free_irq(vec, pf); + + /* Cleanup CQ NAPI and IRQ */ + vec = pf->hw.nix_msixoff + NIX_LF_CINT_VEC_START; + for (qidx = 0; qidx < pf->hw.cint_cnt; qidx++) { + /* Disable interrupt */ + otx2_write64(pf, NIX_LF_CINTX_ENA_W1C(qidx), BIT_ULL(0)); + + synchronize_irq(pci_irq_vector(pf->pdev, vec)); + + cq_poll = &qset->napi[qidx]; + napi_synchronize(&cq_poll->napi); + vec++; + } + + netif_tx_disable(netdev); + + otx2_free_hw_resources(pf); + otx2_free_cints(pf, pf->hw.cint_cnt); + otx2_disable_napi(pf); + + for (qidx = 0; qidx < netdev->num_tx_queues; qidx++) + netdev_tx_reset_queue(netdev_get_tx_queue(netdev, qidx)); + + for (wrk = 0; wrk < pf->qset.cq_cnt; wrk++) + cancel_delayed_work_sync(&pf->refill_wrk[wrk].pool_refill_work); + devm_kfree(pf->dev, pf->refill_wrk); + + kfree(qset->sq); + kfree(qset->cq); + kfree(qset->rq); + kfree(qset->napi); + /* Do not clear RQ/SQ ringsize settings */ + memset((void *)qset + offsetof(struct otx2_qset, sqe_cnt), 0, + sizeof(*qset) - offsetof(struct otx2_qset, sqe_cnt)); + return 0; +} + +static netdev_tx_t otx2_xmit(struct sk_buff *skb, struct net_device *netdev) +{ + struct otx2_nic *pf = netdev_priv(netdev); + int qidx = skb_get_queue_mapping(skb); + struct otx2_snd_queue *sq; + struct netdev_queue *txq; + + /* Check for minimum and maximum packet length */ + if (skb->len <= ETH_HLEN || + (!skb_shinfo(skb)->gso_size && skb->len > pf->max_frs)) { + dev_kfree_skb(skb); + return NETDEV_TX_OK; + } + + sq = &pf->qset.sq[qidx]; + txq = netdev_get_tx_queue(netdev, qidx); + + if (!otx2_sq_append_skb(netdev, sq, skb, qidx)) { + netif_tx_stop_queue(txq); + + /* Check again, incase SQBs got freed up */ + smp_mb(); + if (((sq->num_sqbs - *sq->aura_fc_addr) * sq->sqe_per_sqb) + > sq->sqe_thresh) + netif_tx_wake_queue(txq); + + return NETDEV_TX_BUSY; + } + + return NETDEV_TX_OK; +} + +static void otx2_set_rx_mode(struct net_device *netdev) +{ + struct otx2_nic *pf = netdev_priv(netdev); + struct nix_rx_mode *req; + + if (!(netdev->flags & IFF_UP)) + return; + + otx2_mbox_lock(&pf->mbox); + req = otx2_mbox_alloc_msg_nix_set_rx_mode(&pf->mbox); + if (!req) { + otx2_mbox_unlock(&pf->mbox); + return; + } + + req->mode = NIX_RX_MODE_UCAST; + + /* We don't support MAC address filtering yet */ + if (netdev->flags & IFF_PROMISC) + req->mode |= NIX_RX_MODE_PROMISC; + else if (netdev->flags & (IFF_ALLMULTI | IFF_MULTICAST)) + req->mode |= NIX_RX_MODE_ALLMULTI; + + otx2_sync_mbox_msg(&pf->mbox); + otx2_mbox_unlock(&pf->mbox); +} + +static int otx2_set_features(struct net_device *netdev, + netdev_features_t features) +{ + netdev_features_t changed = features ^ netdev->features; + struct otx2_nic *pf = netdev_priv(netdev); + + if ((changed & NETIF_F_LOOPBACK) && netif_running(netdev)) + return otx2_cgx_config_loopback(pf, + features & NETIF_F_LOOPBACK); + return 0; +} + +static void otx2_reset_task(struct work_struct *work) +{ + struct otx2_nic *pf = container_of(work, struct otx2_nic, reset_task); + + if (!netif_running(pf->netdev)) + return; + + otx2_stop(pf->netdev); + pf->reset_count++; + otx2_open(pf->netdev); + netif_trans_update(pf->netdev); +} + +static const struct net_device_ops otx2_netdev_ops = { + .ndo_open = otx2_open, + .ndo_stop = otx2_stop, + .ndo_start_xmit = otx2_xmit, + .ndo_set_mac_address = otx2_set_mac_address, + .ndo_change_mtu = otx2_change_mtu, + .ndo_set_rx_mode = otx2_set_rx_mode, + .ndo_set_features = otx2_set_features, + .ndo_tx_timeout = otx2_tx_timeout, + .ndo_get_stats64 = otx2_get_stats64, +}; + +static int otx2_check_pf_usable(struct otx2_nic *nic) +{ + u64 rev; + + rev = otx2_read64(nic, RVU_PF_BLOCK_ADDRX_DISC(BLKADDR_RVUM)); + rev = (rev >> 12) & 0xFF; + /* Check if AF has setup revision for RVUM block, + * otherwise this driver probe should be deferred + * until AF driver comes up. + */ + if (!rev) { + dev_warn(nic->dev, + "AF is not initialized, deferring probe\n"); + return -EPROBE_DEFER; + } + return 0; +} + +static int otx2_realloc_msix_vectors(struct otx2_nic *pf) +{ + struct otx2_hw *hw = &pf->hw; + int num_vec, err; + + /* NPA interrupts are inot registered, so alloc only + * upto NIX vector offset. + */ + num_vec = hw->nix_msixoff; + num_vec += NIX_LF_CINT_VEC_START + hw->max_queues; + + otx2_disable_mbox_intr(pf); + pci_free_irq_vectors(hw->pdev); + pci_free_irq_vectors(hw->pdev); + err = pci_alloc_irq_vectors(hw->pdev, num_vec, num_vec, PCI_IRQ_MSIX); + if (err < 0) { + dev_err(pf->dev, "%s: Failed to realloc %d IRQ vectors\n", + __func__, num_vec); + return err; + } + + return otx2_register_mbox_intr(pf, false); +} + +static int otx2_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct device *dev = &pdev->dev; + struct net_device *netdev; + struct otx2_nic *pf; + struct otx2_hw *hw; + int err, qcount; + int num_vec; + + err = pcim_enable_device(pdev); + if (err) { + dev_err(dev, "Failed to enable PCI device\n"); + return err; + } + + err = pci_request_regions(pdev, DRV_NAME); + if (err) { + dev_err(dev, "PCI request regions failed 0x%x\n", err); + return err; + } + + err = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(48)); + if (err) { + dev_err(dev, "DMA mask config failed, abort\n"); + goto err_release_regions; + } + + pci_set_master(pdev); + + /* Set number of queues */ + qcount = min_t(int, num_online_cpus(), OTX2_MAX_CQ_CNT); + + netdev = alloc_etherdev_mqs(sizeof(*pf), qcount, qcount); + if (!netdev) { + err = -ENOMEM; + goto err_release_regions; + } + + pci_set_drvdata(pdev, netdev); + SET_NETDEV_DEV(netdev, &pdev->dev); + pf = netdev_priv(netdev); + pf->netdev = netdev; + pf->pdev = pdev; + pf->dev = dev; + pf->flags |= OTX2_FLAG_INTF_DOWN; + + hw = &pf->hw; + hw->pdev = pdev; + hw->rx_queues = qcount; + hw->tx_queues = qcount; + hw->max_queues = qcount; + + num_vec = pci_msix_vec_count(pdev); + hw->irq_name = devm_kmalloc_array(&hw->pdev->dev, num_vec, NAME_SIZE, + GFP_KERNEL); + if (!hw->irq_name) + goto err_free_netdev; + + hw->affinity_mask = devm_kcalloc(&hw->pdev->dev, num_vec, + sizeof(cpumask_var_t), GFP_KERNEL); + if (!hw->affinity_mask) + goto err_free_netdev; + + /* Map CSRs */ + pf->reg_base = pcim_iomap(pdev, PCI_CFG_REG_BAR_NUM, 0); + if (!pf->reg_base) { + dev_err(dev, "Unable to map physical function CSRs, aborting\n"); + err = -ENOMEM; + goto err_free_netdev; + } + + err = otx2_check_pf_usable(pf); + if (err) + goto err_free_netdev; + + err = pci_alloc_irq_vectors(hw->pdev, RVU_PF_INT_VEC_CNT, + RVU_PF_INT_VEC_CNT, PCI_IRQ_MSIX); + if (err < 0) { + dev_err(dev, "%s: Failed to alloc %d IRQ vectors\n", + __func__, num_vec); + goto err_free_netdev; + } + + /* Init PF <=> AF mailbox stuff */ + err = otx2_pfaf_mbox_init(pf); + if (err) + goto err_free_irq_vectors; + + /* Register mailbox interrupt */ + err = otx2_register_mbox_intr(pf, true); + if (err) + goto err_mbox_destroy; + + /* Request AF to attach NPA and NIX LFs to this PF. + * NIX and NPA LFs are needed for this PF to function as a NIC. + */ + err = otx2_attach_npa_nix(pf); + if (err) + goto err_disable_mbox_intr; + + err = otx2_realloc_msix_vectors(pf); + if (err) + goto err_detach_rsrc; + + err = otx2_set_real_num_queues(netdev, hw->tx_queues, hw->rx_queues); + if (err) + goto err_detach_rsrc; + + otx2_setup_dev_hw_settings(pf); + + /* Assign default mac address */ + otx2_get_mac_from_af(netdev); + + /* NPA's pool is a stack to which SW frees buffer pointers via Aura. + * HW allocates buffer pointer from stack and uses it for DMA'ing + * ingress packet. In some scenarios HW can free back allocated buffer + * pointers to pool. This makes it impossible for SW to maintain a + * parallel list where physical addresses of buffer pointers (IOVAs) + * given to HW can be saved for later reference. + * + * So the only way to convert Rx packet's buffer address is to use + * IOMMU's iova_to_phys() handler which translates the address by + * walking through the translation tables. + */ + pf->iommu_domain = iommu_get_domain_for_dev(dev); + + netdev->hw_features = (NETIF_F_RXCSUM | NETIF_F_IP_CSUM | + NETIF_F_IPV6_CSUM | NETIF_F_RXHASH | + NETIF_F_SG | NETIF_F_TSO | NETIF_F_TSO6); + netdev->features |= netdev->hw_features; + + netdev->hw_features |= NETIF_F_LOOPBACK | NETIF_F_RXALL; + + netdev->gso_max_segs = OTX2_MAX_GSO_SEGS; + netdev->watchdog_timeo = OTX2_TX_TIMEOUT; + + netdev->netdev_ops = &otx2_netdev_ops; + + /* MTU range: 64 - 9190 */ + netdev->min_mtu = OTX2_MIN_MTU; + netdev->max_mtu = OTX2_MAX_MTU; + + INIT_WORK(&pf->reset_task, otx2_reset_task); + + err = register_netdev(netdev); + if (err) { + dev_err(dev, "Failed to register netdevice\n"); + goto err_detach_rsrc; + } + + otx2_set_ethtool_ops(netdev); + + /* Enable link notifications */ + otx2_cgx_config_linkevents(pf, true); + + return 0; + +err_detach_rsrc: + otx2_detach_resources(&pf->mbox); +err_disable_mbox_intr: + otx2_disable_mbox_intr(pf); +err_mbox_destroy: + otx2_pfaf_mbox_destroy(pf); +err_free_irq_vectors: + pci_free_irq_vectors(hw->pdev); +err_free_netdev: + pci_set_drvdata(pdev, NULL); + free_netdev(netdev); +err_release_regions: + pci_release_regions(pdev); + return err; +} + +static void otx2_remove(struct pci_dev *pdev) +{ + struct net_device *netdev = pci_get_drvdata(pdev); + struct otx2_nic *pf; + + if (!netdev) + return; + + pf = netdev_priv(netdev); + + /* Disable link notifications */ + otx2_cgx_config_linkevents(pf, false); + + unregister_netdev(netdev); + otx2_detach_resources(&pf->mbox); + otx2_disable_mbox_intr(pf); + otx2_pfaf_mbox_destroy(pf); + pci_free_irq_vectors(pf->pdev); + pci_set_drvdata(pdev, NULL); + free_netdev(netdev); + + pci_release_regions(pdev); +} + +static struct pci_driver otx2_pf_driver = { + .name = DRV_NAME, + .id_table = otx2_pf_id_table, + .probe = otx2_probe, + .shutdown = otx2_remove, + .remove = otx2_remove, +}; + +static int __init otx2_rvupf_init_module(void) +{ + pr_info("%s: %s\n", DRV_NAME, DRV_STRING); + + return pci_register_driver(&otx2_pf_driver); +} + +static void __exit otx2_rvupf_cleanup_module(void) +{ + pci_unregister_driver(&otx2_pf_driver); +} + +module_init(otx2_rvupf_init_module); +module_exit(otx2_rvupf_cleanup_module); diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_reg.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_reg.h new file mode 100644 index 000000000000..7963d418886a --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_reg.h @@ -0,0 +1,147 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef OTX2_REG_H +#define OTX2_REG_H + +#include <rvu_struct.h> + +/* RVU PF registers */ +#define RVU_PF_VFX_PFVF_MBOX0 (0x00000) +#define RVU_PF_VFX_PFVF_MBOX1 (0x00008) +#define RVU_PF_VFX_PFVF_MBOXX(a, b) (0x0 | (a) << 12 | (b) << 3) +#define RVU_PF_VF_BAR4_ADDR (0x10) +#define RVU_PF_BLOCK_ADDRX_DISC(a) (0x200 | (a) << 3) +#define RVU_PF_VFME_STATUSX(a) (0x800 | (a) << 3) +#define RVU_PF_VFTRPENDX(a) (0x820 | (a) << 3) +#define RVU_PF_VFTRPEND_W1SX(a) (0x840 | (a) << 3) +#define RVU_PF_VFPF_MBOX_INTX(a) (0x880 | (a) << 3) +#define RVU_PF_VFPF_MBOX_INT_W1SX(a) (0x8A0 | (a) << 3) +#define RVU_PF_VFPF_MBOX_INT_ENA_W1SX(a) (0x8C0 | (a) << 3) +#define RVU_PF_VFPF_MBOX_INT_ENA_W1CX(a) (0x8E0 | (a) << 3) +#define RVU_PF_VFFLR_INTX(a) (0x900 | (a) << 3) +#define RVU_PF_VFFLR_INT_W1SX(a) (0x920 | (a) << 3) +#define RVU_PF_VFFLR_INT_ENA_W1SX(a) (0x940 | (a) << 3) +#define RVU_PF_VFFLR_INT_ENA_W1CX(a) (0x960 | (a) << 3) +#define RVU_PF_VFME_INTX(a) (0x980 | (a) << 3) +#define RVU_PF_VFME_INT_W1SX(a) (0x9A0 | (a) << 3) +#define RVU_PF_VFME_INT_ENA_W1SX(a) (0x9C0 | (a) << 3) +#define RVU_PF_VFME_INT_ENA_W1CX(a) (0x9E0 | (a) << 3) +#define RVU_PF_PFAF_MBOX0 (0xC00) +#define RVU_PF_PFAF_MBOX1 (0xC08) +#define RVU_PF_PFAF_MBOXX(a) (0xC00 | (a) << 3) +#define RVU_PF_INT (0xc20) +#define RVU_PF_INT_W1S (0xc28) +#define RVU_PF_INT_ENA_W1S (0xc30) +#define RVU_PF_INT_ENA_W1C (0xc38) +#define RVU_PF_MSIX_VECX_ADDR(a) (0x000 | (a) << 4) +#define RVU_PF_MSIX_VECX_CTL(a) (0x008 | (a) << 4) +#define RVU_PF_MSIX_PBAX(a) (0xF0000 | (a) << 3) + +#define RVU_FUNC_BLKADDR_SHIFT 20 +#define RVU_FUNC_BLKADDR_MASK 0x1FULL + +/* NPA LF registers */ +#define NPA_LFBASE (BLKTYPE_NPA << RVU_FUNC_BLKADDR_SHIFT) +#define NPA_LF_AURA_OP_ALLOCX(a) (NPA_LFBASE | 0x10 | (a) << 3) +#define NPA_LF_AURA_OP_FREE0 (NPA_LFBASE | 0x20) +#define NPA_LF_AURA_OP_FREE1 (NPA_LFBASE | 0x28) +#define NPA_LF_AURA_OP_CNT (NPA_LFBASE | 0x30) +#define NPA_LF_AURA_OP_LIMIT (NPA_LFBASE | 0x50) +#define NPA_LF_AURA_OP_INT (NPA_LFBASE | 0x60) +#define NPA_LF_AURA_OP_THRESH (NPA_LFBASE | 0x70) +#define NPA_LF_POOL_OP_PC (NPA_LFBASE | 0x100) +#define NPA_LF_POOL_OP_AVAILABLE (NPA_LFBASE | 0x110) +#define NPA_LF_POOL_OP_PTR_START0 (NPA_LFBASE | 0x120) +#define NPA_LF_POOL_OP_PTR_START1 (NPA_LFBASE | 0x128) +#define NPA_LF_POOL_OP_PTR_END0 (NPA_LFBASE | 0x130) +#define NPA_LF_POOL_OP_PTR_END1 (NPA_LFBASE | 0x138) +#define NPA_LF_POOL_OP_INT (NPA_LFBASE | 0x160) +#define NPA_LF_POOL_OP_THRESH (NPA_LFBASE | 0x170) +#define NPA_LF_ERR_INT (NPA_LFBASE | 0x200) +#define NPA_LF_ERR_INT_W1S (NPA_LFBASE | 0x208) +#define NPA_LF_ERR_INT_ENA_W1C (NPA_LFBASE | 0x210) +#define NPA_LF_ERR_INT_ENA_W1S (NPA_LFBASE | 0x218) +#define NPA_LF_RAS (NPA_LFBASE | 0x220) +#define NPA_LF_RAS_W1S (NPA_LFBASE | 0x228) +#define NPA_LF_RAS_ENA_W1C (NPA_LFBASE | 0x230) +#define NPA_LF_RAS_ENA_W1S (NPA_LFBASE | 0x238) +#define NPA_LF_QINTX_CNT(a) (NPA_LFBASE | 0x300 | (a) << 12) +#define NPA_LF_QINTX_INT(a) (NPA_LFBASE | 0x310 | (a) << 12) +#define NPA_LF_QINTX_INT_W1S(a) (NPA_LFBASE | 0x318 | (a) << 12) +#define NPA_LF_QINTX_ENA_W1S(a) (NPA_LFBASE | 0x320 | (a) << 12) +#define NPA_LF_QINTX_ENA_W1C(a) (NPA_LFBASE | 0x330 | (a) << 12) + +/* NIX LF registers */ +#define NIX_LFBASE (BLKTYPE_NIX << RVU_FUNC_BLKADDR_SHIFT) +#define NIX_LF_RX_SECRETX(a) (NIX_LFBASE | 0x0 | (a) << 3) +#define NIX_LF_CFG (NIX_LFBASE | 0x100) +#define NIX_LF_GINT (NIX_LFBASE | 0x200) +#define NIX_LF_GINT_W1S (NIX_LFBASE | 0x208) +#define NIX_LF_GINT_ENA_W1C (NIX_LFBASE | 0x210) +#define NIX_LF_GINT_ENA_W1S (NIX_LFBASE | 0x218) +#define NIX_LF_ERR_INT (NIX_LFBASE | 0x220) +#define NIX_LF_ERR_INT_W1S (NIX_LFBASE | 0x228) +#define NIX_LF_ERR_INT_ENA_W1C (NIX_LFBASE | 0x230) +#define NIX_LF_ERR_INT_ENA_W1S (NIX_LFBASE | 0x238) +#define NIX_LF_RAS (NIX_LFBASE | 0x240) +#define NIX_LF_RAS_W1S (NIX_LFBASE | 0x248) +#define NIX_LF_RAS_ENA_W1C (NIX_LFBASE | 0x250) +#define NIX_LF_RAS_ENA_W1S (NIX_LFBASE | 0x258) +#define NIX_LF_SQ_OP_ERR_DBG (NIX_LFBASE | 0x260) +#define NIX_LF_MNQ_ERR_DBG (NIX_LFBASE | 0x270) +#define NIX_LF_SEND_ERR_DBG (NIX_LFBASE | 0x280) +#define NIX_LF_TX_STATX(a) (NIX_LFBASE | 0x300 | (a) << 3) +#define NIX_LF_RX_STATX(a) (NIX_LFBASE | 0x400 | (a) << 3) +#define NIX_LF_OP_SENDX(a) (NIX_LFBASE | 0x800 | (a) << 3) +#define NIX_LF_RQ_OP_INT (NIX_LFBASE | 0x900) +#define NIX_LF_RQ_OP_OCTS (NIX_LFBASE | 0x910) +#define NIX_LF_RQ_OP_PKTS (NIX_LFBASE | 0x920) +#define NIX_LF_OP_IPSEC_DYNO_CN (NIX_LFBASE | 0x980) +#define NIX_LF_SQ_OP_INT (NIX_LFBASE | 0xa00) +#define NIX_LF_SQ_OP_OCTS (NIX_LFBASE | 0xa10) +#define NIX_LF_SQ_OP_PKTS (NIX_LFBASE | 0xa20) +#define NIX_LF_SQ_OP_STATUS (NIX_LFBASE | 0xa30) +#define NIX_LF_CQ_OP_INT (NIX_LFBASE | 0xb00) +#define NIX_LF_CQ_OP_DOOR (NIX_LFBASE | 0xb30) +#define NIX_LF_CQ_OP_STATUS (NIX_LFBASE | 0xb40) +#define NIX_LF_QINTX_CNT(a) (NIX_LFBASE | 0xC00 | (a) << 12) +#define NIX_LF_QINTX_INT(a) (NIX_LFBASE | 0xC10 | (a) << 12) +#define NIX_LF_QINTX_INT_W1S(a) (NIX_LFBASE | 0xC18 | (a) << 12) +#define NIX_LF_QINTX_ENA_W1S(a) (NIX_LFBASE | 0xC20 | (a) << 12) +#define NIX_LF_QINTX_ENA_W1C(a) (NIX_LFBASE | 0xC30 | (a) << 12) +#define NIX_LF_CINTX_CNT(a) (NIX_LFBASE | 0xD00 | (a) << 12) +#define NIX_LF_CINTX_WAIT(a) (NIX_LFBASE | 0xD10 | (a) << 12) +#define NIX_LF_CINTX_INT(a) (NIX_LFBASE | 0xD20 | (a) << 12) +#define NIX_LF_CINTX_INT_W1S(a) (NIX_LFBASE | 0xD30 | (a) << 12) +#define NIX_LF_CINTX_ENA_W1S(a) (NIX_LFBASE | 0xD40 | (a) << 12) +#define NIX_LF_CINTX_ENA_W1C(a) (NIX_LFBASE | 0xD50 | (a) << 12) + +/* NIX AF transmit scheduler registers */ +#define NIX_AF_SMQX_CFG(a) (0x700 | (a) << 16) +#define NIX_AF_TL1X_SCHEDULE(a) (0xC00 | (a) << 16) +#define NIX_AF_TL1X_CIR(a) (0xC20 | (a) << 16) +#define NIX_AF_TL1X_TOPOLOGY(a) (0xC80 | (a) << 16) +#define NIX_AF_TL2X_PARENT(a) (0xE88 | (a) << 16) +#define NIX_AF_TL2X_SCHEDULE(a) (0xE00 | (a) << 16) +#define NIX_AF_TL3X_PARENT(a) (0x1088 | (a) << 16) +#define NIX_AF_TL3X_SCHEDULE(a) (0x1000 | (a) << 16) +#define NIX_AF_TL4X_PARENT(a) (0x1288 | (a) << 16) +#define NIX_AF_TL4X_SCHEDULE(a) (0x1200 | (a) << 16) +#define NIX_AF_MDQX_SCHEDULE(a) (0x1400 | (a) << 16) +#define NIX_AF_MDQX_PARENT(a) (0x1480 | (a) << 16) +#define NIX_AF_TL3_TL2X_LINKX_CFG(a, b) (0x1700 | (a) << 16 | (b) << 3) + +/* LMT LF registers */ +#define LMT_LFBASE BIT_ULL(RVU_FUNC_BLKADDR_SHIFT) +#define LMT_LF_LMTLINEX(a) (LMT_LFBASE | 0x000 | (a) << 12) +#define LMT_LF_LMTCANCEL (LMT_LFBASE | 0x400) + +#endif /* OTX2_REG_H */ diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_struct.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_struct.h new file mode 100644 index 000000000000..cba59ddf71bb --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_struct.h @@ -0,0 +1,276 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef OTX2_STRUCT_H +#define OTX2_STRUCT_H + +/* NIX WQE/CQE size 128 byte or 512 byte */ +enum nix_cqesz_e { + NIX_XQESZ_W64 = 0x0, + NIX_XQESZ_W16 = 0x1, +}; + +enum nix_sqes_e { + NIX_SQESZ_W16 = 0x0, + NIX_SQESZ_W8 = 0x1, +}; + +enum nix_send_ldtype { + NIX_SEND_LDTYPE_LDD = 0x0, + NIX_SEND_LDTYPE_LDT = 0x1, + NIX_SEND_LDTYPE_LDWB = 0x2, +}; + +/* CSUM offload */ +enum nix_sendl3type { + NIX_SENDL3TYPE_NONE = 0x0, + NIX_SENDL3TYPE_IP4 = 0x2, + NIX_SENDL3TYPE_IP4_CKSUM = 0x3, + NIX_SENDL3TYPE_IP6 = 0x4, +}; + +enum nix_sendl4type { + NIX_SENDL4TYPE_NONE, + NIX_SENDL4TYPE_TCP_CKSUM, + NIX_SENDL4TYPE_SCTP_CKSUM, + NIX_SENDL4TYPE_UDP_CKSUM, +}; + +/* NIX wqe/cqe types */ +enum nix_xqe_type { + NIX_XQE_TYPE_INVALID = 0x0, + NIX_XQE_TYPE_RX = 0x1, + NIX_XQE_TYPE_RX_IPSECS = 0x2, + NIX_XQE_TYPE_RX_IPSECH = 0x3, + NIX_XQE_TYPE_RX_IPSECD = 0x4, + NIX_XQE_TYPE_SEND = 0x8, +}; + +/* NIX CQE/SQE subdescriptor types */ +enum nix_subdc { + NIX_SUBDC_NOP = 0x0, + NIX_SUBDC_EXT = 0x1, + NIX_SUBDC_CRC = 0x2, + NIX_SUBDC_IMM = 0x3, + NIX_SUBDC_SG = 0x4, + NIX_SUBDC_MEM = 0x5, + NIX_SUBDC_JUMP = 0x6, + NIX_SUBDC_WORK = 0x7, + NIX_SUBDC_SOD = 0xf, +}; + +/* Algorithm for nix_sqe_mem_s header (value of the `alg` field) */ +enum nix_sendmemalg { + NIX_SENDMEMALG_E_SET = 0x0, + NIX_SENDMEMALG_E_SETTSTMP = 0x1, + NIX_SENDMEMALG_E_SETRSLT = 0x2, + NIX_SENDMEMALG_E_ADD = 0x8, + NIX_SENDMEMALG_E_SUB = 0x9, + NIX_SENDMEMALG_E_ADDLEN = 0xa, + NIX_SENDMEMALG_E_SUBLEN = 0xb, + NIX_SENDMEMALG_E_ADDMBUF = 0xc, + NIX_SENDMEMALG_E_SUBMBUF = 0xd, + NIX_SENDMEMALG_E_ENUM_LAST = 0xe, +}; + +/* NIX CQE header structure */ +struct nix_cqe_hdr_s { + u64 flow_tag : 32; + u64 q : 20; + u64 reserved_52_57 : 6; + u64 node : 2; + u64 cqe_type : 4; +}; + +/* NIX CQE RX parse structure */ +struct nix_rx_parse_s { + u64 chan : 12; + u64 desc_sizem1 : 5; + u64 rsvd_17 : 1; + u64 express : 1; + u64 wqwd : 1; + u64 errlev : 4; + u64 errcode : 8; + u64 latype : 4; + u64 lbtype : 4; + u64 lctype : 4; + u64 ldtype : 4; + u64 letype : 4; + u64 lftype : 4; + u64 lgtype : 4; + u64 lhtype : 4; + u64 pkt_lenm1 : 16; /* W1 */ + u64 l2m : 1; + u64 l2b : 1; + u64 l3m : 1; + u64 l3b : 1; + u64 vtag0_valid : 1; + u64 vtag0_gone : 1; + u64 vtag1_valid : 1; + u64 vtag1_gone : 1; + u64 pkind : 6; + u64 rsvd_95_94 : 2; + u64 vtag0_tci : 16; + u64 vtag1_tci : 16; + u64 laflags : 8; /* W2 */ + u64 lbflags : 8; + u64 lcflags : 8; + u64 ldflags : 8; + u64 leflags : 8; + u64 lfflags : 8; + u64 lgflags : 8; + u64 lhflags : 8; + u64 eoh_ptr : 8; /* W3 */ + u64 wqe_aura : 20; + u64 pb_aura : 20; + u64 match_id : 16; + u64 laptr : 8; /* W4 */ + u64 lbptr : 8; + u64 lcptr : 8; + u64 ldptr : 8; + u64 leptr : 8; + u64 lfptr : 8; + u64 lgptr : 8; + u64 lhptr : 8; + u64 vtag0_ptr : 8; /* W5 */ + u64 vtag1_ptr : 8; + u64 flow_key_alg : 5; + u64 rsvd_383_341 : 43; + u64 rsvd_447_384; /* W6 */ +}; + +/* NIX CQE RX scatter/gather subdescriptor structure */ +struct nix_rx_sg_s { + u64 seg_size : 16; /* W0 */ + u64 seg2_size : 16; + u64 seg3_size : 16; + u64 segs : 2; + u64 rsvd_59_50 : 10; + u64 subdc : 4; + u64 seg_addr; + u64 seg2_addr; + u64 seg3_addr; +}; + +struct nix_send_comp_s { + u64 status : 8; + u64 sqe_id : 16; + u64 rsvd_24_63 : 40; +}; + +struct nix_cqe_rx_s { + struct nix_cqe_hdr_s hdr; + struct nix_rx_parse_s parse; + struct nix_rx_sg_s sg; +}; + +struct nix_cqe_tx_s { + struct nix_cqe_hdr_s hdr; + struct nix_send_comp_s comp; +}; + +/* NIX SQE header structure */ +struct nix_sqe_hdr_s { + u64 total : 18; /* W0 */ + u64 reserved_18 : 1; + u64 df : 1; + u64 aura : 20; + u64 sizem1 : 3; + u64 pnc : 1; + u64 sq : 20; + u64 ol3ptr : 8; /* W1 */ + u64 ol4ptr : 8; + u64 il3ptr : 8; + u64 il4ptr : 8; + u64 ol3type : 4; + u64 ol4type : 4; + u64 il3type : 4; + u64 il4type : 4; + u64 sqe_id : 16; + +}; + +/* NIX send extended header subdescriptor structure */ +struct nix_sqe_ext_s { + u64 lso_mps : 14; /* W0 */ + u64 lso : 1; + u64 tstmp : 1; + u64 lso_sb : 8; + u64 lso_format : 5; + u64 rsvd_31_29 : 3; + u64 shp_chg : 9; + u64 shp_dis : 1; + u64 shp_ra : 2; + u64 markptr : 8; + u64 markform : 7; + u64 mark_en : 1; + u64 subdc : 4; + u64 vlan0_ins_ptr : 8; /* W1 */ + u64 vlan0_ins_tci : 16; + u64 vlan1_ins_ptr : 8; + u64 vlan1_ins_tci : 16; + u64 vlan0_ins_ena : 1; + u64 vlan1_ins_ena : 1; + u64 rsvd_127_114 : 14; +}; + +struct nix_sqe_sg_s { + u64 seg1_size : 16; + u64 seg2_size : 16; + u64 seg3_size : 16; + u64 segs : 2; + u64 rsvd_54_50 : 5; + u64 i1 : 1; + u64 i2 : 1; + u64 i3 : 1; + u64 ld_type : 2; + u64 subdc : 4; +}; + +/* NIX send memory subdescriptor structure */ +struct nix_sqe_mem_s { + u64 offset : 16; /* W0 */ + u64 rsvd_52_16 : 37; + u64 wmem : 1; + u64 dsz : 2; + u64 alg : 4; + u64 subdc : 4; + u64 addr; /* W1 */ +}; + +enum nix_cqerrint_e { + NIX_CQERRINT_DOOR_ERR = 0, + NIX_CQERRINT_WR_FULL = 1, + NIX_CQERRINT_CQE_FAULT = 2, +}; + +#define NIX_CQERRINT_BITS (BIT_ULL(NIX_CQERRINT_DOOR_ERR) | \ + BIT_ULL(NIX_CQERRINT_CQE_FAULT)) + +enum nix_rqint_e { + NIX_RQINT_DROP = 0, + NIX_RQINT_RED = 1, +}; + +#define NIX_RQINT_BITS (BIT_ULL(NIX_RQINT_DROP) | BIT_ULL(NIX_RQINT_RED)) + +enum nix_sqint_e { + NIX_SQINT_LMT_ERR = 0, + NIX_SQINT_MNQ_ERR = 1, + NIX_SQINT_SEND_ERR = 2, + NIX_SQINT_SQB_ALLOC_FAIL = 3, +}; + +#define NIX_SQINT_BITS (BIT_ULL(NIX_SQINT_LMT_ERR) | \ + BIT_ULL(NIX_SQINT_MNQ_ERR) | \ + BIT_ULL(NIX_SQINT_SEND_ERR) | \ + BIT_ULL(NIX_SQINT_SQB_ALLOC_FAIL)) + +#endif /* OTX2_STRUCT_H */ diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c new file mode 100644 index 000000000000..bef4c20fe314 --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c @@ -0,0 +1,848 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include <linux/etherdevice.h> +#include <net/ip.h> +#include <net/tso.h> + +#include "otx2_reg.h" +#include "otx2_common.h" +#include "otx2_struct.h" +#include "otx2_txrx.h" + +#define CQE_ADDR(CQ, idx) ((CQ)->cqe_base + ((CQ)->cqe_size * (idx))) + +static struct nix_cqe_hdr_s *otx2_get_next_cqe(struct otx2_cq_queue *cq) +{ + struct nix_cqe_hdr_s *cqe_hdr; + + cqe_hdr = (struct nix_cqe_hdr_s *)CQE_ADDR(cq, cq->cq_head); + if (cqe_hdr->cqe_type == NIX_XQE_TYPE_INVALID) + return NULL; + + cq->cq_head++; + cq->cq_head &= (cq->cqe_cnt - 1); + + return cqe_hdr; +} + +static unsigned int frag_num(unsigned int i) +{ +#ifdef __BIG_ENDIAN + return (i & ~3) + 3 - (i & 3); +#else + return i; +#endif +} + +static dma_addr_t otx2_dma_map_skb_frag(struct otx2_nic *pfvf, + struct sk_buff *skb, int seg, int *len) +{ + const skb_frag_t *frag; + struct page *page; + int offset; + + /* First segment is always skb->data */ + if (!seg) { + page = virt_to_page(skb->data); + offset = offset_in_page(skb->data); + *len = skb_headlen(skb); + } else { + frag = &skb_shinfo(skb)->frags[seg - 1]; + page = skb_frag_page(frag); + offset = skb_frag_off(frag); + *len = skb_frag_size(frag); + } + return otx2_dma_map_page(pfvf, page, offset, *len, DMA_TO_DEVICE); +} + +static void otx2_dma_unmap_skb_frags(struct otx2_nic *pfvf, struct sg_list *sg) +{ + int seg; + + for (seg = 0; seg < sg->num_segs; seg++) { + otx2_dma_unmap_page(pfvf, sg->dma_addr[seg], + sg->size[seg], DMA_TO_DEVICE); + } + sg->num_segs = 0; +} + +static void otx2_snd_pkt_handler(struct otx2_nic *pfvf, + struct otx2_cq_queue *cq, + struct otx2_snd_queue *sq, + struct nix_cqe_tx_s *cqe, + int budget, int *tx_pkts, int *tx_bytes) +{ + struct nix_send_comp_s *snd_comp = &cqe->comp; + struct sk_buff *skb = NULL; + struct sg_list *sg; + + if (unlikely(snd_comp->status) && netif_msg_tx_err(pfvf)) + net_err_ratelimited("%s: TX%d: Error in send CQ status:%x\n", + pfvf->netdev->name, cq->cint_idx, + snd_comp->status); + + sg = &sq->sg[snd_comp->sqe_id]; + skb = (struct sk_buff *)sg->skb; + if (unlikely(!skb)) + return; + + *tx_bytes += skb->len; + (*tx_pkts)++; + otx2_dma_unmap_skb_frags(pfvf, sg); + napi_consume_skb(skb, budget); + sg->skb = (u64)NULL; +} + +static void otx2_skb_add_frag(struct otx2_nic *pfvf, struct sk_buff *skb, + u64 iova, int len) +{ + struct page *page; + void *va; + + va = phys_to_virt(otx2_iova_to_phys(pfvf->iommu_domain, iova)); + page = virt_to_page(va); + skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags, page, + va - page_address(page), len, pfvf->rbsize); + + otx2_dma_unmap_page(pfvf, iova - OTX2_HEAD_ROOM, + pfvf->rbsize, DMA_FROM_DEVICE); +} + +static void otx2_set_rxhash(struct otx2_nic *pfvf, + struct nix_cqe_rx_s *cqe, struct sk_buff *skb) +{ + enum pkt_hash_types hash_type = PKT_HASH_TYPE_NONE; + struct otx2_rss_info *rss; + u32 hash = 0; + + if (!(pfvf->netdev->features & NETIF_F_RXHASH)) + return; + + rss = &pfvf->hw.rss_info; + if (rss->flowkey_cfg) { + if (rss->flowkey_cfg & + ~(NIX_FLOW_KEY_TYPE_IPV4 | NIX_FLOW_KEY_TYPE_IPV6)) + hash_type = PKT_HASH_TYPE_L4; + else + hash_type = PKT_HASH_TYPE_L3; + hash = cqe->hdr.flow_tag; + } + skb_set_hash(skb, hash, hash_type); +} + +static bool otx2_check_rcv_errors(struct otx2_nic *pfvf, + struct nix_cqe_rx_s *cqe, int qidx) +{ + struct otx2_drv_stats *stats = &pfvf->hw.drv_stats; + struct nix_rx_parse_s *parse = &cqe->parse; + + if (netif_msg_rx_err(pfvf)) + netdev_err(pfvf->netdev, + "RQ%d: Error pkt with errlev:0x%x errcode:0x%x\n", + qidx, parse->errlev, parse->errcode); + + if (parse->errlev == NPC_ERRLVL_RE) { + switch (parse->errcode) { + case ERRCODE_FCS: + case ERRCODE_FCS_RCV: + atomic_inc(&stats->rx_fcs_errs); + break; + case ERRCODE_UNDERSIZE: + atomic_inc(&stats->rx_undersize_errs); + break; + case ERRCODE_OVERSIZE: + atomic_inc(&stats->rx_oversize_errs); + break; + case ERRCODE_OL2_LEN_MISMATCH: + atomic_inc(&stats->rx_len_errs); + break; + default: + atomic_inc(&stats->rx_other_errs); + break; + } + } else if (parse->errlev == NPC_ERRLVL_NIX) { + switch (parse->errcode) { + case ERRCODE_OL3_LEN: + case ERRCODE_OL4_LEN: + case ERRCODE_IL3_LEN: + case ERRCODE_IL4_LEN: + atomic_inc(&stats->rx_len_errs); + break; + case ERRCODE_OL4_CSUM: + case ERRCODE_IL4_CSUM: + atomic_inc(&stats->rx_csum_errs); + break; + default: + atomic_inc(&stats->rx_other_errs); + break; + } + } else { + atomic_inc(&stats->rx_other_errs); + /* For now ignore all the NPC parser errors and + * pass the packets to stack. + */ + return false; + } + + /* If RXALL is enabled pass on packets to stack. */ + if (cqe->sg.segs && (pfvf->netdev->features & NETIF_F_RXALL)) + return false; + + /* Free buffer back to pool */ + if (cqe->sg.segs) + otx2_aura_freeptr(pfvf, qidx, cqe->sg.seg_addr & ~0x07ULL); + return true; +} + +static void otx2_rcv_pkt_handler(struct otx2_nic *pfvf, + struct napi_struct *napi, + struct otx2_cq_queue *cq, + struct nix_cqe_rx_s *cqe) +{ + struct nix_rx_parse_s *parse = &cqe->parse; + struct sk_buff *skb = NULL; + + if (unlikely(parse->errlev || parse->errcode)) { + if (otx2_check_rcv_errors(pfvf, cqe, cq->cq_idx)) + return; + } + + skb = napi_get_frags(napi); + if (unlikely(!skb)) + return; + + otx2_skb_add_frag(pfvf, skb, cqe->sg.seg_addr, cqe->sg.seg_size); + cq->pool_ptrs++; + + otx2_set_rxhash(pfvf, cqe, skb); + + skb_record_rx_queue(skb, cq->cq_idx); + if (pfvf->netdev->features & NETIF_F_RXCSUM) + skb->ip_summed = CHECKSUM_UNNECESSARY; + + napi_gro_frags(napi); +} + +static int otx2_rx_napi_handler(struct otx2_nic *pfvf, + struct napi_struct *napi, + struct otx2_cq_queue *cq, int budget) +{ + struct nix_cqe_rx_s *cqe; + int processed_cqe = 0; + s64 bufptr; + + while (likely(processed_cqe < budget)) { + cqe = (struct nix_cqe_rx_s *)CQE_ADDR(cq, cq->cq_head); + if (cqe->hdr.cqe_type == NIX_XQE_TYPE_INVALID || + !cqe->sg.seg_addr) { + if (!processed_cqe) + return 0; + break; + } + cq->cq_head++; + cq->cq_head &= (cq->cqe_cnt - 1); + + otx2_rcv_pkt_handler(pfvf, napi, cq, cqe); + + cqe->hdr.cqe_type = NIX_XQE_TYPE_INVALID; + cqe->sg.seg_addr = 0x00; + processed_cqe++; + } + + /* Free CQEs to HW */ + otx2_write64(pfvf, NIX_LF_CQ_OP_DOOR, + ((u64)cq->cq_idx << 32) | processed_cqe); + + if (unlikely(!cq->pool_ptrs)) + return 0; + + /* Refill pool with new buffers */ + while (cq->pool_ptrs) { + bufptr = otx2_alloc_rbuf(pfvf, cq->rbpool, GFP_ATOMIC); + if (unlikely(bufptr <= 0)) { + struct refill_work *work; + struct delayed_work *dwork; + + work = &pfvf->refill_wrk[cq->cq_idx]; + dwork = &work->pool_refill_work; + /* Schedule a task if no other task is running */ + if (!cq->refill_task_sched) { + cq->refill_task_sched = true; + schedule_delayed_work(dwork, + msecs_to_jiffies(100)); + } + break; + } + otx2_aura_freeptr(pfvf, cq->cq_idx, bufptr + OTX2_HEAD_ROOM); + cq->pool_ptrs--; + } + + return processed_cqe; +} + +static int otx2_tx_napi_handler(struct otx2_nic *pfvf, + struct otx2_cq_queue *cq, int budget) +{ + int tx_pkts = 0, tx_bytes = 0; + struct nix_cqe_tx_s *cqe; + int processed_cqe = 0; + + while (likely(processed_cqe < budget)) { + cqe = (struct nix_cqe_tx_s *)otx2_get_next_cqe(cq); + if (unlikely(!cqe)) { + if (!processed_cqe) + return 0; + break; + } + otx2_snd_pkt_handler(pfvf, cq, &pfvf->qset.sq[cq->cint_idx], + cqe, budget, &tx_pkts, &tx_bytes); + + cqe->hdr.cqe_type = NIX_XQE_TYPE_INVALID; + processed_cqe++; + } + + /* Free CQEs to HW */ + otx2_write64(pfvf, NIX_LF_CQ_OP_DOOR, + ((u64)cq->cq_idx << 32) | processed_cqe); + + if (likely(tx_pkts)) { + struct netdev_queue *txq; + + txq = netdev_get_tx_queue(pfvf->netdev, cq->cint_idx); + netdev_tx_completed_queue(txq, tx_pkts, tx_bytes); + /* Check if queue was stopped earlier due to ring full */ + smp_mb(); + if (netif_tx_queue_stopped(txq) && + netif_carrier_ok(pfvf->netdev)) + netif_tx_wake_queue(txq); + } + return 0; +} + +int otx2_napi_handler(struct napi_struct *napi, int budget) +{ + struct otx2_cq_poll *cq_poll; + int workdone = 0, cq_idx, i; + struct otx2_cq_queue *cq; + struct otx2_qset *qset; + struct otx2_nic *pfvf; + + cq_poll = container_of(napi, struct otx2_cq_poll, napi); + pfvf = (struct otx2_nic *)cq_poll->dev; + qset = &pfvf->qset; + + for (i = CQS_PER_CINT - 1; i >= 0; i--) { + cq_idx = cq_poll->cq_ids[i]; + if (unlikely(cq_idx == CINT_INVALID_CQ)) + continue; + cq = &qset->cq[cq_idx]; + if (cq->cq_type == CQ_RX) { + /* If the RQ refill WQ task is running, skip napi + * scheduler for this queue. + */ + if (cq->refill_task_sched) + continue; + workdone += otx2_rx_napi_handler(pfvf, napi, + cq, budget); + } else { + workdone += otx2_tx_napi_handler(pfvf, cq, budget); + } + } + + /* Clear the IRQ */ + otx2_write64(pfvf, NIX_LF_CINTX_INT(cq_poll->cint_idx), BIT_ULL(0)); + + if (workdone < budget && napi_complete_done(napi, workdone)) { + /* If interface is going down, don't re-enable IRQ */ + if (pfvf->flags & OTX2_FLAG_INTF_DOWN) + return workdone; + + /* Re-enable interrupts */ + otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx), + BIT_ULL(0)); + } + return workdone; +} + +static void otx2_sqe_flush(struct otx2_snd_queue *sq, int size) +{ + u64 status; + + /* Packet data stores should finish before SQE is flushed to HW */ + dma_wmb(); + + do { + memcpy(sq->lmt_addr, sq->sqe_base, size); + status = otx2_lmt_flush(sq->io_addr); + } while (status == 0); + + sq->head++; + sq->head &= (sq->sqe_cnt - 1); +} + +#define MAX_SEGS_PER_SG 3 +/* Add SQE scatter/gather subdescriptor structure */ +static bool otx2_sqe_add_sg(struct otx2_nic *pfvf, struct otx2_snd_queue *sq, + struct sk_buff *skb, int num_segs, int *offset) +{ + struct nix_sqe_sg_s *sg = NULL; + u64 dma_addr, *iova = NULL; + u16 *sg_lens = NULL; + int seg, len; + + sq->sg[sq->head].num_segs = 0; + + for (seg = 0; seg < num_segs; seg++) { + if ((seg % MAX_SEGS_PER_SG) == 0) { + sg = (struct nix_sqe_sg_s *)(sq->sqe_base + *offset); + sg->ld_type = NIX_SEND_LDTYPE_LDD; + sg->subdc = NIX_SUBDC_SG; + sg->segs = 0; + sg_lens = (void *)sg; + iova = (void *)sg + sizeof(*sg); + /* Next subdc always starts at a 16byte boundary. + * So if sg->segs is whether 2 or 3, offset += 16bytes. + */ + if ((num_segs - seg) >= (MAX_SEGS_PER_SG - 1)) + *offset += sizeof(*sg) + (3 * sizeof(u64)); + else + *offset += sizeof(*sg) + sizeof(u64); + } + dma_addr = otx2_dma_map_skb_frag(pfvf, skb, seg, &len); + if (dma_mapping_error(pfvf->dev, dma_addr)) + return false; + + sg_lens[frag_num(seg % MAX_SEGS_PER_SG)] = len; + sg->segs++; + *iova++ = dma_addr; + + /* Save DMA mapping info for later unmapping */ + sq->sg[sq->head].dma_addr[seg] = dma_addr; + sq->sg[sq->head].size[seg] = len; + sq->sg[sq->head].num_segs++; + } + + sq->sg[sq->head].skb = (u64)skb; + return true; +} + +/* Add SQE extended header subdescriptor */ +static void otx2_sqe_add_ext(struct otx2_nic *pfvf, struct otx2_snd_queue *sq, + struct sk_buff *skb, int *offset) +{ + struct nix_sqe_ext_s *ext; + + ext = (struct nix_sqe_ext_s *)(sq->sqe_base + *offset); + ext->subdc = NIX_SUBDC_EXT; + if (skb_shinfo(skb)->gso_size) { + ext->lso = 1; + ext->lso_sb = skb_transport_offset(skb) + tcp_hdrlen(skb); + ext->lso_mps = skb_shinfo(skb)->gso_size; + + /* Only TSOv4 and TSOv6 GSO offloads are supported */ + if (skb_shinfo(skb)->gso_type & SKB_GSO_TCPV4) { + ext->lso_format = pfvf->hw.lso_tsov4_idx; + + /* HW adds payload size to 'ip_hdr->tot_len' while + * sending TSO segment, hence set payload length + * in IP header of the packet to just header length. + */ + ip_hdr(skb)->tot_len = + htons(ext->lso_sb - skb_network_offset(skb)); + } else { + ext->lso_format = pfvf->hw.lso_tsov6_idx; + ipv6_hdr(skb)->payload_len = + htons(ext->lso_sb - skb_network_offset(skb)); + } + } + *offset += sizeof(*ext); +} + +/* Add SQE header subdescriptor structure */ +static void otx2_sqe_add_hdr(struct otx2_nic *pfvf, struct otx2_snd_queue *sq, + struct nix_sqe_hdr_s *sqe_hdr, + struct sk_buff *skb, u16 qidx) +{ + int proto = 0; + + /* Check if SQE was framed before, if yes then no need to + * set these constants again and again. + */ + if (!sqe_hdr->total) { + /* Don't free Tx buffers to Aura */ + sqe_hdr->df = 1; + sqe_hdr->aura = sq->aura_id; + /* Post a CQE Tx after pkt transmission */ + sqe_hdr->pnc = 1; + sqe_hdr->sq = qidx; + } + sqe_hdr->total = skb->len; + /* Set SQE identifier which will be used later for freeing SKB */ + sqe_hdr->sqe_id = sq->head; + + /* Offload TCP/UDP checksum to HW */ + if (skb->ip_summed == CHECKSUM_PARTIAL) { + sqe_hdr->ol3ptr = skb_network_offset(skb); + sqe_hdr->ol4ptr = skb_transport_offset(skb); + /* get vlan protocol Ethertype */ + if (eth_type_vlan(skb->protocol)) + skb->protocol = vlan_get_protocol(skb); + + if (skb->protocol == htons(ETH_P_IP)) { + proto = ip_hdr(skb)->protocol; + /* In case of TSO, HW needs this to be explicitly set. + * So set this always, instead of adding a check. + */ + sqe_hdr->ol3type = NIX_SENDL3TYPE_IP4_CKSUM; + } else if (skb->protocol == htons(ETH_P_IPV6)) { + proto = ipv6_hdr(skb)->nexthdr; + } + + if (proto == IPPROTO_TCP) + sqe_hdr->ol4type = NIX_SENDL4TYPE_TCP_CKSUM; + else if (proto == IPPROTO_UDP) + sqe_hdr->ol4type = NIX_SENDL4TYPE_UDP_CKSUM; + } +} + +static int otx2_dma_map_tso_skb(struct otx2_nic *pfvf, + struct otx2_snd_queue *sq, + struct sk_buff *skb, int sqe, int hdr_len) +{ + int num_segs = skb_shinfo(skb)->nr_frags + 1; + struct sg_list *sg = &sq->sg[sqe]; + u64 dma_addr; + int seg, len; + + sg->num_segs = 0; + + /* Get payload length at skb->data */ + len = skb_headlen(skb) - hdr_len; + + for (seg = 0; seg < num_segs; seg++) { + /* Skip skb->data, if there is no payload */ + if (!seg && !len) + continue; + dma_addr = otx2_dma_map_skb_frag(pfvf, skb, seg, &len); + if (dma_mapping_error(pfvf->dev, dma_addr)) + goto unmap; + + /* Save DMA mapping info for later unmapping */ + sg->dma_addr[sg->num_segs] = dma_addr; + sg->size[sg->num_segs] = len; + sg->num_segs++; + } + return 0; +unmap: + otx2_dma_unmap_skb_frags(pfvf, sg); + return -EINVAL; +} + +static u64 otx2_tso_frag_dma_addr(struct otx2_snd_queue *sq, + struct sk_buff *skb, int seg, + u64 seg_addr, int hdr_len, int sqe) +{ + struct sg_list *sg = &sq->sg[sqe]; + const skb_frag_t *frag; + int offset; + + if (seg < 0) + return sg->dma_addr[0] + (seg_addr - (u64)skb->data); + + frag = &skb_shinfo(skb)->frags[seg]; + offset = seg_addr - (u64)skb_frag_address(frag); + if (skb_headlen(skb) - hdr_len) + seg++; + return sg->dma_addr[seg] + offset; +} + +static void otx2_sqe_tso_add_sg(struct otx2_snd_queue *sq, + struct sg_list *list, int *offset) +{ + struct nix_sqe_sg_s *sg = NULL; + u16 *sg_lens = NULL; + u64 *iova = NULL; + int seg; + + /* Add SG descriptors with buffer addresses */ + for (seg = 0; seg < list->num_segs; seg++) { + if ((seg % MAX_SEGS_PER_SG) == 0) { + sg = (struct nix_sqe_sg_s *)(sq->sqe_base + *offset); + sg->ld_type = NIX_SEND_LDTYPE_LDD; + sg->subdc = NIX_SUBDC_SG; + sg->segs = 0; + sg_lens = (void *)sg; + iova = (void *)sg + sizeof(*sg); + /* Next subdc always starts at a 16byte boundary. + * So if sg->segs is whether 2 or 3, offset += 16bytes. + */ + if ((list->num_segs - seg) >= (MAX_SEGS_PER_SG - 1)) + *offset += sizeof(*sg) + (3 * sizeof(u64)); + else + *offset += sizeof(*sg) + sizeof(u64); + } + sg_lens[frag_num(seg % MAX_SEGS_PER_SG)] = list->size[seg]; + *iova++ = list->dma_addr[seg]; + sg->segs++; + } +} + +static void otx2_sq_append_tso(struct otx2_nic *pfvf, struct otx2_snd_queue *sq, + struct sk_buff *skb, u16 qidx) +{ + struct netdev_queue *txq = netdev_get_tx_queue(pfvf->netdev, qidx); + int hdr_len = skb_transport_offset(skb) + tcp_hdrlen(skb); + int tcp_data, seg_len, pkt_len, offset; + struct nix_sqe_hdr_s *sqe_hdr; + int first_sqe = sq->head; + struct sg_list list; + struct tso_t tso; + + /* Map SKB's fragments to DMA. + * It's done here to avoid mapping for every TSO segment's packet. + */ + if (otx2_dma_map_tso_skb(pfvf, sq, skb, first_sqe, hdr_len)) { + dev_kfree_skb_any(skb); + return; + } + + netdev_tx_sent_queue(txq, skb->len); + + tso_start(skb, &tso); + tcp_data = skb->len - hdr_len; + while (tcp_data > 0) { + char *hdr; + + seg_len = min_t(int, skb_shinfo(skb)->gso_size, tcp_data); + tcp_data -= seg_len; + + /* Set SQE's SEND_HDR */ + memset(sq->sqe_base, 0, sq->sqe_size); + sqe_hdr = (struct nix_sqe_hdr_s *)(sq->sqe_base); + otx2_sqe_add_hdr(pfvf, sq, sqe_hdr, skb, qidx); + offset = sizeof(*sqe_hdr); + + /* Add TSO segment's pkt header */ + hdr = sq->tso_hdrs->base + (sq->head * TSO_HEADER_SIZE); + tso_build_hdr(skb, hdr, &tso, seg_len, tcp_data == 0); + list.dma_addr[0] = + sq->tso_hdrs->iova + (sq->head * TSO_HEADER_SIZE); + list.size[0] = hdr_len; + list.num_segs = 1; + + /* Add TSO segment's payload data fragments */ + pkt_len = hdr_len; + while (seg_len > 0) { + int size; + + size = min_t(int, tso.size, seg_len); + + list.size[list.num_segs] = size; + list.dma_addr[list.num_segs] = + otx2_tso_frag_dma_addr(sq, skb, + tso.next_frag_idx - 1, + (u64)tso.data, hdr_len, + first_sqe); + list.num_segs++; + pkt_len += size; + seg_len -= size; + tso_build_data(skb, &tso, size); + } + sqe_hdr->total = pkt_len; + otx2_sqe_tso_add_sg(sq, &list, &offset); + + /* DMA mappings and skb needs to be freed only after last + * TSO segment is transmitted out. So set 'PNC' only for + * last segment. Also point last segment's sqe_id to first + * segment's SQE index where skb address and DMA mappings + * are saved. + */ + if (!tcp_data) { + sqe_hdr->pnc = 1; + sqe_hdr->sqe_id = first_sqe; + sq->sg[first_sqe].skb = (u64)skb; + } else { + sqe_hdr->pnc = 0; + } + + sqe_hdr->sizem1 = (offset / 16) - 1; + + /* Flush SQE to HW */ + otx2_sqe_flush(sq, offset); + } +} + +static bool is_hw_tso_supported(struct otx2_nic *pfvf, + struct sk_buff *skb) +{ + int payload_len, last_seg_size; + + if (!pfvf->hw.hw_tso) + return false; + + /* HW has an issue due to which when the payload of the last LSO + * segment is shorter than 16 bytes, some header fields may not + * be correctly modified, hence don't offload such TSO segments. + */ + if (!is_96xx_B0(pfvf->pdev)) + return true; + + payload_len = skb->len - (skb_transport_offset(skb) + tcp_hdrlen(skb)); + last_seg_size = payload_len % skb_shinfo(skb)->gso_size; + if (last_seg_size && last_seg_size < 16) + return false; + + return true; +} + +static int otx2_get_sqe_count(struct otx2_nic *pfvf, struct sk_buff *skb) +{ + if (!skb_shinfo(skb)->gso_size) + return 1; + + /* HW TSO */ + if (is_hw_tso_supported(pfvf, skb)) + return 1; + + /* SW TSO */ + return skb_shinfo(skb)->gso_segs; +} + +bool otx2_sq_append_skb(struct net_device *netdev, struct otx2_snd_queue *sq, + struct sk_buff *skb, u16 qidx) +{ + struct netdev_queue *txq = netdev_get_tx_queue(netdev, qidx); + struct otx2_nic *pfvf = netdev_priv(netdev); + int offset, num_segs, free_sqe; + struct nix_sqe_hdr_s *sqe_hdr; + + /* Check if there is room for new SQE. + * 'Num of SQBs freed to SQ's pool - SQ's Aura count' + * will give free SQE count. + */ + free_sqe = (sq->num_sqbs - *sq->aura_fc_addr) * sq->sqe_per_sqb; + + if (free_sqe < sq->sqe_thresh || + free_sqe < otx2_get_sqe_count(pfvf, skb)) + return false; + + num_segs = skb_shinfo(skb)->nr_frags + 1; + + /* If SKB doesn't fit in a single SQE, linearize it. + * TODO: Consider adding JUMP descriptor instead. + */ + if (unlikely(num_segs > OTX2_MAX_FRAGS_IN_SQE)) { + if (__skb_linearize(skb)) { + dev_kfree_skb_any(skb); + return true; + } + num_segs = skb_shinfo(skb)->nr_frags + 1; + } + + if (skb_shinfo(skb)->gso_size && !is_hw_tso_supported(pfvf, skb)) { + otx2_sq_append_tso(pfvf, sq, skb, qidx); + return true; + } + + /* Set SQE's SEND_HDR. + * Do not clear the first 64bit as it contains constant info. + */ + memset(sq->sqe_base + 8, 0, sq->sqe_size - 8); + sqe_hdr = (struct nix_sqe_hdr_s *)(sq->sqe_base); + otx2_sqe_add_hdr(pfvf, sq, sqe_hdr, skb, qidx); + offset = sizeof(*sqe_hdr); + + /* Add extended header if needed */ + otx2_sqe_add_ext(pfvf, sq, skb, &offset); + + /* Add SG subdesc with data frags */ + if (!otx2_sqe_add_sg(pfvf, sq, skb, num_segs, &offset)) { + otx2_dma_unmap_skb_frags(pfvf, &sq->sg[sq->head]); + return false; + } + + sqe_hdr->sizem1 = (offset / 16) - 1; + + netdev_tx_sent_queue(txq, skb->len); + + /* Flush SQE to HW */ + otx2_sqe_flush(sq, offset); + + return true; +} + +void otx2_cleanup_rx_cqes(struct otx2_nic *pfvf, struct otx2_cq_queue *cq) +{ + struct nix_cqe_rx_s *cqe; + int processed_cqe = 0; + u64 iova, pa; + + while ((cqe = (struct nix_cqe_rx_s *)otx2_get_next_cqe(cq))) { + if (!cqe->sg.subdc) + continue; + iova = cqe->sg.seg_addr - OTX2_HEAD_ROOM; + pa = otx2_iova_to_phys(pfvf->iommu_domain, iova); + otx2_dma_unmap_page(pfvf, iova, pfvf->rbsize, DMA_FROM_DEVICE); + put_page(virt_to_page(phys_to_virt(pa))); + processed_cqe++; + } + + /* Free CQEs to HW */ + otx2_write64(pfvf, NIX_LF_CQ_OP_DOOR, + ((u64)cq->cq_idx << 32) | processed_cqe); +} + +void otx2_cleanup_tx_cqes(struct otx2_nic *pfvf, struct otx2_cq_queue *cq) +{ + struct sk_buff *skb = NULL; + struct otx2_snd_queue *sq; + struct nix_cqe_tx_s *cqe; + int processed_cqe = 0; + struct sg_list *sg; + + sq = &pfvf->qset.sq[cq->cint_idx]; + + while ((cqe = (struct nix_cqe_tx_s *)otx2_get_next_cqe(cq))) { + sg = &sq->sg[cqe->comp.sqe_id]; + skb = (struct sk_buff *)sg->skb; + if (skb) { + otx2_dma_unmap_skb_frags(pfvf, sg); + dev_kfree_skb_any(skb); + sg->skb = (u64)NULL; + } + processed_cqe++; + } + + /* Free CQEs to HW */ + otx2_write64(pfvf, NIX_LF_CQ_OP_DOOR, + ((u64)cq->cq_idx << 32) | processed_cqe); +} + +int otx2_rxtx_enable(struct otx2_nic *pfvf, bool enable) +{ + struct msg_req *msg; + int err; + + otx2_mbox_lock(&pfvf->mbox); + if (enable) + msg = otx2_mbox_alloc_msg_nix_lf_start_rx(&pfvf->mbox); + else + msg = otx2_mbox_alloc_msg_nix_lf_stop_rx(&pfvf->mbox); + + if (!msg) { + otx2_mbox_unlock(&pfvf->mbox); + return -ENOMEM; + } + + err = otx2_sync_mbox_msg(&pfvf->mbox); + otx2_mbox_unlock(&pfvf->mbox); + return err; +} diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h new file mode 100644 index 000000000000..4ab32d3adb78 --- /dev/null +++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h @@ -0,0 +1,162 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* Marvell OcteonTx2 RVU Ethernet driver + * + * Copyright (C) 2020 Marvell International Ltd. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef OTX2_TXRX_H +#define OTX2_TXRX_H + +#include <linux/etherdevice.h> +#include <linux/iommu.h> +#include <linux/if_vlan.h> + +#define LBK_CHAN_BASE 0x000 +#define SDP_CHAN_BASE 0x700 +#define CGX_CHAN_BASE 0x800 + +#define OTX2_DATA_ALIGN(X) ALIGN(X, OTX2_ALIGN) +#define OTX2_HEAD_ROOM OTX2_ALIGN + +#define OTX2_ETH_HLEN (VLAN_ETH_HLEN + VLAN_HLEN) +#define OTX2_MIN_MTU 64 +#define OTX2_MAX_MTU (9212 - OTX2_ETH_HLEN) + +#define OTX2_MAX_GSO_SEGS 255 +#define OTX2_MAX_FRAGS_IN_SQE 9 + +/* Rx buffer size should be in multiples of 128bytes */ +#define RCV_FRAG_LEN1(x) \ + ((OTX2_HEAD_ROOM + OTX2_DATA_ALIGN(x)) + \ + OTX2_DATA_ALIGN(sizeof(struct skb_shared_info))) + +/* Prefer 2048 byte buffers for better last level cache + * utilization or data distribution across regions. + */ +#define RCV_FRAG_LEN(x) \ + ((RCV_FRAG_LEN1(x) < 2048) ? 2048 : RCV_FRAG_LEN1(x)) + +#define DMA_BUFFER_LEN(x) \ + ((x) - OTX2_HEAD_ROOM - \ + OTX2_DATA_ALIGN(sizeof(struct skb_shared_info))) + +/* IRQ triggered when NIX_LF_CINTX_CNT[ECOUNT] + * is equal to this value. + */ +#define CQ_CQE_THRESH_DEFAULT 10 + +/* IRQ triggered when NIX_LF_CINTX_CNT[ECOUNT] + * is nonzero and this much time elapses after that. + */ +#define CQ_TIMER_THRESH_DEFAULT 1 /* 1 usec */ +#define CQ_TIMER_THRESH_MAX 25 /* 25 usec */ + +/* Min number of CQs (of the ones mapped to this CINT) + * with valid CQEs. + */ +#define CQ_QCOUNT_DEFAULT 1 + +struct queue_stats { + u64 bytes; + u64 pkts; +}; + +struct otx2_rcv_queue { + struct queue_stats stats; +}; + +struct sg_list { + u16 num_segs; + u64 skb; + u64 size[OTX2_MAX_FRAGS_IN_SQE]; + u64 dma_addr[OTX2_MAX_FRAGS_IN_SQE]; +}; + +struct otx2_snd_queue { + u8 aura_id; + u16 head; + u16 sqe_size; + u32 sqe_cnt; + u16 num_sqbs; + u16 sqe_thresh; + u8 sqe_per_sqb; + u64 io_addr; + u64 *aura_fc_addr; + u64 *lmt_addr; + void *sqe_base; + struct qmem *sqe; + struct qmem *tso_hdrs; + struct sg_list *sg; + struct queue_stats stats; + u16 sqb_count; + u64 *sqb_ptrs; +} ____cacheline_aligned_in_smp; + +enum cq_type { + CQ_RX, + CQ_TX, + CQS_PER_CINT = 2, /* RQ + SQ */ +}; + +struct otx2_cq_poll { + void *dev; +#define CINT_INVALID_CQ 255 + u8 cint_idx; + u8 cq_ids[CQS_PER_CINT]; + struct napi_struct napi; +}; + +struct otx2_pool { + struct qmem *stack; + struct qmem *fc_addr; + u8 rbpage_order; + u16 rbsize; + u32 page_offset; + u16 pageref; + struct page *page; +}; + +struct otx2_cq_queue { + u8 cq_idx; + u8 cq_type; + u8 cint_idx; /* CQ interrupt id */ + u8 refill_task_sched; + u16 cqe_size; + u16 pool_ptrs; + u32 cqe_cnt; + u32 cq_head; + void *cqe_base; + struct qmem *cqe; + struct otx2_pool *rbpool; +} ____cacheline_aligned_in_smp; + +struct otx2_qset { + u32 rqe_cnt; + u32 sqe_cnt; /* Keep these two at top */ +#define OTX2_MAX_CQ_CNT 64 + u16 cq_cnt; + u16 xqe_size; + struct otx2_pool *pool; + struct otx2_cq_poll *napi; + struct otx2_cq_queue *cq; + struct otx2_snd_queue *sq; + struct otx2_rcv_queue *rq; +}; + +/* Translate IOVA to physical address */ +static inline u64 otx2_iova_to_phys(void *iommu_domain, dma_addr_t dma_addr) +{ + /* Translation is installed only when IOMMU is present */ + if (likely(iommu_domain)) + return iommu_iova_to_phys(iommu_domain, dma_addr); + return dma_addr; +} + +int otx2_napi_handler(struct napi_struct *napi, int budget); +bool otx2_sq_append_skb(struct net_device *netdev, struct otx2_snd_queue *sq, + struct sk_buff *skb, u16 qidx); +#endif /* OTX2_TXRX_H */ diff --git a/drivers/net/ethernet/marvell/pxa168_eth.c b/drivers/net/ethernet/marvell/pxa168_eth.c index 3fb7ee3d4d13..7a0d785b826c 100644 --- a/drivers/net/ethernet/marvell/pxa168_eth.c +++ b/drivers/net/ethernet/marvell/pxa168_eth.c @@ -742,7 +742,7 @@ txq_reclaim_end: return released; } -static void pxa168_eth_tx_timeout(struct net_device *dev) +static void pxa168_eth_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct pxa168_eth_private *pep = netdev_priv(dev); @@ -1344,15 +1344,6 @@ static int pxa168_smi_write(struct mii_bus *bus, int phy_addr, int regnum, return 0; } -static int pxa168_eth_do_ioctl(struct net_device *dev, struct ifreq *ifr, - int cmd) -{ - if (dev->phydev) - return phy_mii_ioctl(dev->phydev, ifr, cmd); - - return -EOPNOTSUPP; -} - #ifdef CONFIG_NET_POLL_CONTROLLER static void pxa168_eth_netpoll(struct net_device *dev) { @@ -1387,7 +1378,7 @@ static const struct net_device_ops pxa168_eth_netdev_ops = { .ndo_set_rx_mode = pxa168_eth_set_rx_mode, .ndo_set_mac_address = pxa168_eth_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = pxa168_eth_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_change_mtu = pxa168_eth_change_mtu, .ndo_tx_timeout = pxa168_eth_tx_timeout, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/marvell/skge.c b/drivers/net/ethernet/marvell/skge.c index 7515d079c600..97f270d30cce 100644 --- a/drivers/net/ethernet/marvell/skge.c +++ b/drivers/net/ethernet/marvell/skge.c @@ -2884,7 +2884,7 @@ static void skge_tx_clean(struct net_device *dev) skge->tx_ring.to_clean = e; } -static void skge_tx_timeout(struct net_device *dev) +static void skge_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct skge_port *skge = netdev_priv(dev); diff --git a/drivers/net/ethernet/marvell/sky2.c b/drivers/net/ethernet/marvell/sky2.c index 535dee35e04e..ebfd0ceac884 100644 --- a/drivers/net/ethernet/marvell/sky2.c +++ b/drivers/net/ethernet/marvell/sky2.c @@ -2358,7 +2358,7 @@ static void sky2_qlink_intr(struct sky2_hw *hw) /* Transmit timeout is only called if we are running, carrier is up * and tx queue is full (stopped). */ -static void sky2_tx_timeout(struct net_device *dev) +static void sky2_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sky2_port *sky2 = netdev_priv(dev); struct sky2_hw *hw = sky2->hw; diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index 527ad2aadcca..8c6cfd15481c 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -2081,7 +2081,7 @@ static void mtk_dma_free(struct mtk_eth *eth) kfree(eth->scratch_head); } -static void mtk_tx_timeout(struct net_device *dev) +static void mtk_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mtk_mac *mac = netdev_priv(dev); struct mtk_eth *eth = mac->hw; diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 7af75b63245f..43dcbd8214c6 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -1363,24 +1363,18 @@ static void mlx4_en_delete_rss_steer_rules(struct mlx4_en_priv *priv) } } -static void mlx4_en_tx_timeout(struct net_device *dev) +static void mlx4_en_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mlx4_en_priv *priv = netdev_priv(dev); struct mlx4_en_dev *mdev = priv->mdev; - int i; + struct mlx4_en_tx_ring *tx_ring = priv->tx_ring[TX][txqueue]; if (netif_msg_timer(priv)) en_warn(priv, "Tx timeout called on port:%d\n", priv->port); - for (i = 0; i < priv->tx_ring_num[TX]; i++) { - struct mlx4_en_tx_ring *tx_ring = priv->tx_ring[TX][i]; - - if (!netif_tx_queue_stopped(netdev_get_tx_queue(dev, i))) - continue; - en_warn(priv, "TX timeout on queue: %d, QP: 0x%x, CQ: 0x%x, Cons: 0x%x, Prod: 0x%x\n", - i, tx_ring->qpn, tx_ring->sp_cqn, - tx_ring->cons, tx_ring->prod); - } + en_warn(priv, "TX timeout on queue: %d, QP: 0x%x, CQ: 0x%x, Cons: 0x%x, Prod: 0x%x\n", + txqueue, tx_ring->qpn, tx_ring->sp_cqn, + tx_ring->cons, tx_ring->prod); priv->port_stats.tx_timeout++; en_dbg(DRV, priv, "Scheduling watchdog\n"); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/Makefile b/drivers/net/ethernet/mellanox/mlx5/core/Makefile index a6f390fdb971..d3e06cec8317 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/Makefile +++ b/drivers/net/ethernet/mellanox/mlx5/core/Makefile @@ -42,7 +42,7 @@ mlx5_core-$(CONFIG_PCI_HYPERV_INTERFACE) += en/hv_vhca_stats.o # Core extra # mlx5_core-$(CONFIG_MLX5_ESWITCH) += eswitch.o eswitch_offloads.o eswitch_offloads_termtbl.o \ - ecpf.o rdma.o + ecpf.o rdma.o eswitch_offloads_chains.o mlx5_core-$(CONFIG_MLX5_MPFS) += lib/mpfs.o mlx5_core-$(CONFIG_VXLAN) += lib/vxlan.o mlx5_core-$(CONFIG_PTP_1588_CLOCK) += lib/clock.o diff --git a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c index 549f962cd86e..42198e64a7f4 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/alloc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/alloc.c @@ -71,8 +71,8 @@ static void *mlx5_dma_zalloc_coherent_node(struct mlx5_core_dev *dev, return cpu_handle; } -int mlx5_buf_alloc_node(struct mlx5_core_dev *dev, int size, - struct mlx5_frag_buf *buf, int node) +static int mlx5_buf_alloc_node(struct mlx5_core_dev *dev, int size, + struct mlx5_frag_buf *buf, int node) { dma_addr_t t; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en.h b/drivers/net/ethernet/mellanox/mlx5/core/en.h index 9c8427698238..220ef9f06f84 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en.h @@ -135,7 +135,7 @@ struct page_pool; #define MLX5E_LOG_INDIR_RQT_SIZE 0x7 #define MLX5E_INDIR_RQT_SIZE BIT(MLX5E_LOG_INDIR_RQT_SIZE) #define MLX5E_MIN_NUM_CHANNELS 0x1 -#define MLX5E_MAX_NUM_CHANNELS (MLX5E_INDIR_RQT_SIZE >> 1) +#define MLX5E_MAX_NUM_CHANNELS MLX5E_INDIR_RQT_SIZE #define MLX5E_MAX_NUM_SQS (MLX5E_MAX_NUM_CHANNELS * MLX5E_MAX_NUM_TC) #define MLX5E_TX_CQ_POLL_BUDGET 128 #define MLX5E_TX_XSK_POLL_BUDGET 64 @@ -892,6 +892,8 @@ struct mlx5e_profile { int (*update_rx)(struct mlx5e_priv *priv); void (*update_stats)(struct mlx5e_priv *priv); void (*update_carrier)(struct mlx5e_priv *priv); + unsigned int (*stats_grps_num)(struct mlx5e_priv *priv); + mlx5e_stats_grp_t *stats_grps; struct { mlx5e_fp_handle_rx_cqe handle_rx_cqe; mlx5e_fp_handle_rx_cqe handle_rx_cqe_mpwqe; @@ -964,7 +966,6 @@ struct sk_buff * mlx5e_skb_from_cqe_nonlinear(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe, struct mlx5e_wqe_frag_info *wi, u32 cqe_bcnt); -void mlx5e_update_stats(struct mlx5e_priv *priv); void mlx5e_get_stats(struct net_device *dev, struct rtnl_link_stats64 *stats); void mlx5e_fold_sw_stats64(struct mlx5e_priv *priv, struct rtnl_link_stats64 *s); @@ -1175,11 +1176,11 @@ int mlx5e_attach_netdev(struct mlx5e_priv *priv); void mlx5e_detach_netdev(struct mlx5e_priv *priv); void mlx5e_destroy_netdev(struct mlx5e_priv *priv); void mlx5e_set_netdev_mtu_boundaries(struct mlx5e_priv *priv); -void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, +void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, struct mlx5e_rss_params *rss_params, struct mlx5e_params *params, - u16 max_channels, u16 mtu); + u16 mtu); void mlx5e_build_rq_params(struct mlx5_core_dev *mdev, struct mlx5e_params *params); void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h index d48292ccda29..0416f7712109 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/fs.h @@ -21,6 +21,7 @@ struct mlx5e_tc_table { DECLARE_HASHTABLE(hairpin_tbl, 8); struct notifier_block netdevice_nb; + struct netdev_net_notifier netdevice_nn; }; struct mlx5e_flow_table { diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c index 475b6bd5d29b..62fc8a128a8d 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.c @@ -35,7 +35,7 @@ int mlx5e_xsk_page_alloc_umem(struct mlx5e_rq *rq, */ dma_info->addr = xdp_umem_get_dma(umem, handle); - xsk_umem_discard_addr_rq(umem); + xsk_umem_release_addr_rq(umem); dma_sync_single_for_device(rq->pdev, dma_info->addr, PAGE_SIZE, DMA_BIDIRECTIONAL); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c index c6776f308d5e..d674cb679895 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_ethtool.c @@ -218,13 +218,9 @@ static const struct pflag_desc mlx5e_priv_flags[MLX5E_NUM_PFLAGS]; int mlx5e_ethtool_get_sset_count(struct mlx5e_priv *priv, int sset) { - int i, num_stats = 0; - switch (sset) { case ETH_SS_STATS: - for (i = 0; i < mlx5e_num_stats_grps; i++) - num_stats += mlx5e_stats_grps[i].get_num_stats(priv); - return num_stats; + return mlx5e_stats_total_num(priv); case ETH_SS_PRIV_FLAGS: return MLX5E_NUM_PFLAGS; case ETH_SS_TEST: @@ -242,14 +238,6 @@ static int mlx5e_get_sset_count(struct net_device *dev, int sset) return mlx5e_ethtool_get_sset_count(priv, sset); } -static void mlx5e_fill_stats_strings(struct mlx5e_priv *priv, u8 *data) -{ - int i, idx = 0; - - for (i = 0; i < mlx5e_num_stats_grps; i++) - idx = mlx5e_stats_grps[i].fill_strings(priv, data, idx); -} - void mlx5e_ethtool_get_strings(struct mlx5e_priv *priv, u32 stringset, u8 *data) { int i; @@ -268,7 +256,7 @@ void mlx5e_ethtool_get_strings(struct mlx5e_priv *priv, u32 stringset, u8 *data) break; case ETH_SS_STATS: - mlx5e_fill_stats_strings(priv, data); + mlx5e_stats_fill_strings(priv, data); break; } } @@ -283,14 +271,13 @@ static void mlx5e_get_strings(struct net_device *dev, u32 stringset, u8 *data) void mlx5e_ethtool_get_ethtool_stats(struct mlx5e_priv *priv, struct ethtool_stats *stats, u64 *data) { - int i, idx = 0; + int idx = 0; mutex_lock(&priv->state_lock); - mlx5e_update_stats(priv); + mlx5e_stats_update(priv); mutex_unlock(&priv->state_lock); - for (i = 0; i < mlx5e_num_stats_grps; i++) - idx = mlx5e_stats_grps[i].fill_stats(priv, data, idx); + mlx5e_stats_fill(priv, data, idx); } static void mlx5e_get_ethtool_stats(struct net_device *dev, diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c index acd946f2ddbe..3bc2ac3d53fc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_fs_ethtool.c @@ -58,6 +58,7 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, struct ethtool_rx_flow_spec *fs, int num_tuples) { + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5e_ethtool_table *eth_ft; struct mlx5_flow_namespace *ns; struct mlx5_flow_table *ft; @@ -102,9 +103,11 @@ static struct mlx5e_ethtool_table *get_flow_table(struct mlx5e_priv *priv, table_size = min_t(u32, BIT(MLX5_CAP_FLOWTABLE(priv->mdev, flow_table_properties_nic_receive.log_max_ft_size)), MLX5E_ETHTOOL_NUM_ENTRIES); - ft = mlx5_create_auto_grouped_flow_table(ns, prio, - table_size, - MLX5E_ETHTOOL_NUM_GROUPS, 0, 0); + + ft_attr.prio = prio; + ft_attr.max_fte = table_size; + ft_attr.autogroup.max_num_groups = MLX5E_ETHTOOL_NUM_GROUPS; + ft = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); if (IS_ERR(ft)) return (void *)ft; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index 4997b8a51994..454d3459bd8b 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -159,23 +159,14 @@ static void mlx5e_update_carrier_work(struct work_struct *work) mutex_unlock(&priv->state_lock); } -void mlx5e_update_stats(struct mlx5e_priv *priv) -{ - int i; - - for (i = mlx5e_num_stats_grps - 1; i >= 0; i--) - if (mlx5e_stats_grps[i].update_stats) - mlx5e_stats_grps[i].update_stats(priv); -} - void mlx5e_update_ndo_stats(struct mlx5e_priv *priv) { int i; - for (i = mlx5e_num_stats_grps - 1; i >= 0; i--) - if (mlx5e_stats_grps[i].update_stats_mask & + for (i = mlx5e_nic_stats_grps_num(priv) - 1; i >= 0; i--) + if (mlx5e_nic_stats_grps[i]->update_stats_mask & MLX5E_NDO_UPDATE_STATS) - mlx5e_stats_grps[i].update_stats(priv); + mlx5e_nic_stats_grps[i]->update_stats(priv); } static void mlx5e_update_stats_work(struct work_struct *work) @@ -4325,7 +4316,7 @@ unlock: rtnl_unlock(); } -static void mlx5e_tx_timeout(struct net_device *dev) +static void mlx5e_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct mlx5e_priv *priv = netdev_priv(dev); @@ -4739,17 +4730,19 @@ void mlx5e_build_rss_params(struct mlx5e_rss_params *rss_params, tirc_default_config[tt].rx_hash_fields; } -void mlx5e_build_nic_params(struct mlx5_core_dev *mdev, +void mlx5e_build_nic_params(struct mlx5e_priv *priv, struct mlx5e_xsk *xsk, struct mlx5e_rss_params *rss_params, struct mlx5e_params *params, - u16 max_channels, u16 mtu) + u16 mtu) { + struct mlx5_core_dev *mdev = priv->mdev; u8 rx_cq_period_mode; params->sw_mtu = mtu; params->hard_mtu = MLX5E_ETH_HARD_MTU; - params->num_channels = max_channels; + params->num_channels = min_t(unsigned int, MLX5E_MAX_NUM_CHANNELS / 2, + priv->max_nch); params->num_tc = 1; /* SQ */ @@ -4876,6 +4869,8 @@ static void mlx5e_build_nic_netdev(struct net_device *netdev) netdev->hw_enc_features |= NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_UDP_TUNNEL_CSUM; netdev->gso_partial_features = NETIF_F_GSO_UDP_TUNNEL_CSUM; + netdev->vlan_features |= NETIF_F_GSO_UDP_TUNNEL | + NETIF_F_GSO_UDP_TUNNEL_CSUM; } if (mlx5e_tunnel_proto_supported(mdev, IPPROTO_GRE)) { @@ -4986,8 +4981,8 @@ static int mlx5e_nic_init(struct mlx5_core_dev *mdev, if (err) return err; - mlx5e_build_nic_params(mdev, &priv->xsk, rss, &priv->channels.params, - priv->max_nch, netdev->mtu); + mlx5e_build_nic_params(priv, &priv->xsk, rss, &priv->channels.params, + netdev->mtu); mlx5e_timestamp_init(priv); @@ -5149,6 +5144,7 @@ static void mlx5e_nic_enable(struct mlx5e_priv *priv) static void mlx5e_nic_disable(struct mlx5e_priv *priv) { + struct net_device *netdev = priv->netdev; struct mlx5_core_dev *mdev = priv->mdev; #ifdef CONFIG_MLX5_CORE_EN_DCB @@ -5169,7 +5165,7 @@ static void mlx5e_nic_disable(struct mlx5e_priv *priv) mlx5e_monitor_counter_cleanup(priv); mlx5e_disable_async_events(priv); - mlx5_lag_remove(mdev); + mlx5_lag_remove(mdev, netdev); } int mlx5e_update_nic_rx(struct mlx5e_priv *priv) @@ -5193,6 +5189,8 @@ static const struct mlx5e_profile mlx5e_nic_profile = { .rx_handlers.handle_rx_cqe_mpwqe = mlx5e_handle_rx_cqe_mpwrq, .max_tc = MLX5E_MAX_NUM_TC, .rq_groups = MLX5E_NUM_RQ_GROUPS(XSK), + .stats_grps = mlx5e_nic_stats_grps, + .stats_grps_num = mlx5e_nic_stats_grps_num, }; /* mlx5e generic netdev management API (move to en_common.c) */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c index f175cb24bb67..7b48ccacebe2 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.c @@ -41,6 +41,7 @@ #include <net/ipv6_stubs.h> #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "en.h" #include "en_rep.h" #include "en_tc.h" @@ -116,24 +117,71 @@ static const struct counter_desc vport_rep_stats_desc[] = { #define NUM_VPORT_REP_SW_COUNTERS ARRAY_SIZE(sw_rep_stats_desc) #define NUM_VPORT_REP_HW_COUNTERS ARRAY_SIZE(vport_rep_stats_desc) -static void mlx5e_rep_get_strings(struct net_device *dev, - u32 stringset, uint8_t *data) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(sw_rep) { - int i, j; + return NUM_VPORT_REP_SW_COUNTERS; +} - switch (stringset) { - case ETH_SS_STATS: - for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++) - strcpy(data + (i * ETH_GSTRING_LEN), - sw_rep_stats_desc[i].format); - for (j = 0; j < NUM_VPORT_REP_HW_COUNTERS; j++, i++) - strcpy(data + (i * ETH_GSTRING_LEN), - vport_rep_stats_desc[j].format); - break; - } +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(sw_rep) +{ + int i; + + for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++) + strcpy(data + (idx++) * ETH_GSTRING_LEN, + sw_rep_stats_desc[i].format); + return idx; +} + +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(sw_rep) +{ + int i; + + for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++) + data[idx++] = MLX5E_READ_CTR64_CPU(&priv->stats.sw, + sw_rep_stats_desc, i); + return idx; +} + +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(sw_rep) +{ + struct mlx5e_sw_stats *s = &priv->stats.sw; + struct rtnl_link_stats64 stats64 = {}; + + memset(s, 0, sizeof(*s)); + mlx5e_fold_sw_stats64(priv, &stats64); + + s->rx_packets = stats64.rx_packets; + s->rx_bytes = stats64.rx_bytes; + s->tx_packets = stats64.tx_packets; + s->tx_bytes = stats64.tx_bytes; + s->tx_queue_dropped = stats64.tx_dropped; +} + +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(vport_rep) +{ + return NUM_VPORT_REP_HW_COUNTERS; +} + +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(vport_rep) +{ + int i; + + for (i = 0; i < NUM_VPORT_REP_HW_COUNTERS; i++) + strcpy(data + (idx++) * ETH_GSTRING_LEN, vport_rep_stats_desc[i].format); + return idx; } -static void mlx5e_rep_update_hw_counters(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(vport_rep) +{ + int i; + + for (i = 0; i < NUM_VPORT_REP_HW_COUNTERS; i++) + data[idx++] = MLX5E_READ_CTR64_CPU(&priv->stats.vf_vport, + vport_rep_stats_desc, i); + return idx; +} + +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(vport_rep) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; struct mlx5e_rep_priv *rpriv = priv->ppriv; @@ -156,64 +204,33 @@ static void mlx5e_rep_update_hw_counters(struct mlx5e_priv *priv) vport_stats->tx_bytes = vf_stats.rx_bytes; } -static void mlx5e_uplink_rep_update_hw_counters(struct mlx5e_priv *priv) -{ - struct mlx5e_pport_stats *pstats = &priv->stats.pport; - struct rtnl_link_stats64 *vport_stats; - - mlx5e_grp_802_3_update_stats(priv); - - vport_stats = &priv->stats.vf_vport; - - vport_stats->rx_packets = PPORT_802_3_GET(pstats, a_frames_received_ok); - vport_stats->rx_bytes = PPORT_802_3_GET(pstats, a_octets_received_ok); - vport_stats->tx_packets = PPORT_802_3_GET(pstats, a_frames_transmitted_ok); - vport_stats->tx_bytes = PPORT_802_3_GET(pstats, a_octets_transmitted_ok); -} - -static void mlx5e_rep_update_sw_counters(struct mlx5e_priv *priv) +static void mlx5e_rep_get_strings(struct net_device *dev, + u32 stringset, uint8_t *data) { - struct mlx5e_sw_stats *s = &priv->stats.sw; - struct rtnl_link_stats64 stats64 = {}; - - memset(s, 0, sizeof(*s)); - mlx5e_fold_sw_stats64(priv, &stats64); + struct mlx5e_priv *priv = netdev_priv(dev); - s->rx_packets = stats64.rx_packets; - s->rx_bytes = stats64.rx_bytes; - s->tx_packets = stats64.tx_packets; - s->tx_bytes = stats64.tx_bytes; - s->tx_queue_dropped = stats64.tx_dropped; + switch (stringset) { + case ETH_SS_STATS: + mlx5e_stats_fill_strings(priv, data); + break; + } } static void mlx5e_rep_get_ethtool_stats(struct net_device *dev, struct ethtool_stats *stats, u64 *data) { struct mlx5e_priv *priv = netdev_priv(dev); - int i, j; - - if (!data) - return; - - mutex_lock(&priv->state_lock); - mlx5e_rep_update_sw_counters(priv); - priv->profile->update_stats(priv); - mutex_unlock(&priv->state_lock); - for (i = 0; i < NUM_VPORT_REP_SW_COUNTERS; i++) - data[i] = MLX5E_READ_CTR64_CPU(&priv->stats.sw, - sw_rep_stats_desc, i); - - for (j = 0; j < NUM_VPORT_REP_HW_COUNTERS; j++, i++) - data[i] = MLX5E_READ_CTR64_CPU(&priv->stats.vf_vport, - vport_rep_stats_desc, j); + mlx5e_ethtool_get_ethtool_stats(priv, stats, data); } static int mlx5e_rep_get_sset_count(struct net_device *dev, int sset) { + struct mlx5e_priv *priv = netdev_priv(dev); + switch (sset) { case ETH_SS_STATS: - return NUM_VPORT_REP_SW_COUNTERS + NUM_VPORT_REP_HW_COUNTERS; + return mlx5e_stats_total_num(priv); default: return -EOPNOTSUPP; } @@ -1247,8 +1264,7 @@ static int mlx5e_rep_setup_tc_cb(enum tc_setup_type type, void *type_data, static int mlx5e_rep_setup_ft_cb(enum tc_setup_type type, void *type_data, void *cb_priv) { - struct flow_cls_offload *f = type_data; - struct flow_cls_offload cls_flower; + struct flow_cls_offload tmp, *f = type_data; struct mlx5e_priv *priv = cb_priv; struct mlx5_eswitch *esw; unsigned long flags; @@ -1261,16 +1277,30 @@ static int mlx5e_rep_setup_ft_cb(enum tc_setup_type type, void *type_data, switch (type) { case TC_SETUP_CLSFLOWER: - if (!mlx5_eswitch_prios_supported(esw) || f->common.chain_index) + memcpy(&tmp, f, sizeof(*f)); + + if (!mlx5_esw_chains_prios_supported(esw) || + tmp.common.chain_index) return -EOPNOTSUPP; /* Re-use tc offload path by moving the ft flow to the * reserved ft chain. + * + * FT offload can use prio range [0, INT_MAX], so we normalize + * it to range [1, mlx5_esw_chains_get_prio_range(esw)] + * as with tc, where prio 0 isn't supported. + * + * We only support chain 0 of FT offload. */ - memcpy(&cls_flower, f, sizeof(*f)); - cls_flower.common.chain_index = FDB_FT_CHAIN; - err = mlx5e_rep_setup_tc_cls_flower(priv, &cls_flower, flags); - memcpy(&f->stats, &cls_flower.stats, sizeof(f->stats)); + if (tmp.common.prio >= mlx5_esw_chains_get_prio_range(esw)) + return -EOPNOTSUPP; + if (tmp.common.chain_index != 0) + return -EOPNOTSUPP; + + tmp.common.chain_index = mlx5_esw_chains_get_ft_chain(esw); + tmp.common.prio++; + err = mlx5e_rep_setup_tc_cls_flower(priv, &tmp, flags); + memcpy(&f->stats, &tmp.stats, sizeof(f->stats)); return err; default: return -EOPNOTSUPP; @@ -1660,10 +1690,65 @@ static void mlx5e_cleanup_rep_rx(struct mlx5e_priv *priv) mlx5e_close_drop_rq(&priv->drop_rq); } +static int mlx5e_init_ul_rep_rx(struct mlx5e_priv *priv) +{ + int err = mlx5e_init_rep_rx(priv); + + if (err) + return err; + + mlx5e_create_q_counters(priv); + return 0; +} + +static void mlx5e_cleanup_ul_rep_rx(struct mlx5e_priv *priv) +{ + mlx5e_destroy_q_counters(priv); + mlx5e_cleanup_rep_rx(priv); +} + +static int mlx5e_init_uplink_rep_tx(struct mlx5e_rep_priv *rpriv) +{ + struct mlx5_rep_uplink_priv *uplink_priv; + struct net_device *netdev; + struct mlx5e_priv *priv; + int err; + + netdev = rpriv->netdev; + priv = netdev_priv(netdev); + uplink_priv = &rpriv->uplink_priv; + + mutex_init(&uplink_priv->unready_flows_lock); + INIT_LIST_HEAD(&uplink_priv->unready_flows); + + /* init shared tc flow table */ + err = mlx5e_tc_esw_init(&uplink_priv->tc_ht); + if (err) + return err; + + mlx5_init_port_tun_entropy(&uplink_priv->tun_entropy, priv->mdev); + + /* init indirect block notifications */ + INIT_LIST_HEAD(&uplink_priv->tc_indr_block_priv_list); + uplink_priv->netdevice_nb.notifier_call = mlx5e_nic_rep_netdevice_event; + err = register_netdevice_notifier_dev_net(rpriv->netdev, + &uplink_priv->netdevice_nb, + &uplink_priv->netdevice_nn); + if (err) { + mlx5_core_err(priv->mdev, "Failed to register netdev notifier\n"); + goto tc_esw_cleanup; + } + + return 0; + +tc_esw_cleanup: + mlx5e_tc_esw_cleanup(&uplink_priv->tc_ht); + return err; +} + static int mlx5e_init_rep_tx(struct mlx5e_priv *priv) { struct mlx5e_rep_priv *rpriv = priv->ppriv; - struct mlx5_rep_uplink_priv *uplink_priv; int err; err = mlx5e_create_tises(priv); @@ -1673,52 +1758,41 @@ static int mlx5e_init_rep_tx(struct mlx5e_priv *priv) } if (rpriv->rep->vport == MLX5_VPORT_UPLINK) { - uplink_priv = &rpriv->uplink_priv; - - mutex_init(&uplink_priv->unready_flows_lock); - INIT_LIST_HEAD(&uplink_priv->unready_flows); - - /* init shared tc flow table */ - err = mlx5e_tc_esw_init(&uplink_priv->tc_ht); + err = mlx5e_init_uplink_rep_tx(rpriv); if (err) goto destroy_tises; - - mlx5_init_port_tun_entropy(&uplink_priv->tun_entropy, priv->mdev); - - /* init indirect block notifications */ - INIT_LIST_HEAD(&uplink_priv->tc_indr_block_priv_list); - uplink_priv->netdevice_nb.notifier_call = mlx5e_nic_rep_netdevice_event; - err = register_netdevice_notifier(&uplink_priv->netdevice_nb); - if (err) { - mlx5_core_err(priv->mdev, "Failed to register netdev notifier\n"); - goto tc_esw_cleanup; - } } return 0; -tc_esw_cleanup: - mlx5e_tc_esw_cleanup(&uplink_priv->tc_ht); destroy_tises: mlx5e_destroy_tises(priv); return err; } +static void mlx5e_cleanup_uplink_rep_tx(struct mlx5e_rep_priv *rpriv) +{ + struct mlx5_rep_uplink_priv *uplink_priv = &rpriv->uplink_priv; + + /* clean indirect TC block notifications */ + unregister_netdevice_notifier_dev_net(rpriv->netdev, + &uplink_priv->netdevice_nb, + &uplink_priv->netdevice_nn); + mlx5e_rep_indr_clean_block_privs(rpriv); + + /* delete shared tc flow table */ + mlx5e_tc_esw_cleanup(&rpriv->uplink_priv.tc_ht); + mutex_destroy(&rpriv->uplink_priv.unready_flows_lock); +} + static void mlx5e_cleanup_rep_tx(struct mlx5e_priv *priv) { struct mlx5e_rep_priv *rpriv = priv->ppriv; mlx5e_destroy_tises(priv); - if (rpriv->rep->vport == MLX5_VPORT_UPLINK) { - /* clean indirect TC block notifications */ - unregister_netdevice_notifier(&rpriv->uplink_priv.netdevice_nb); - mlx5e_rep_indr_clean_block_privs(rpriv); - - /* delete shared tc flow table */ - mlx5e_tc_esw_cleanup(&rpriv->uplink_priv.tc_ht); - mutex_destroy(&rpriv->uplink_priv.unready_flows_lock); - } + if (rpriv->rep->vport == MLX5_VPORT_UPLINK) + mlx5e_cleanup_uplink_rep_tx(rpriv); } static void mlx5e_rep_enable(struct mlx5e_priv *priv) @@ -1787,6 +1861,7 @@ static void mlx5e_uplink_rep_enable(struct mlx5e_priv *priv) static void mlx5e_uplink_rep_disable(struct mlx5e_priv *priv) { + struct net_device *netdev = priv->netdev; struct mlx5_core_dev *mdev = priv->mdev; struct mlx5e_rep_priv *rpriv = priv->ppriv; @@ -1795,7 +1870,44 @@ static void mlx5e_uplink_rep_disable(struct mlx5e_priv *priv) #endif mlx5_notifier_unregister(mdev, &priv->events_nb); cancel_work_sync(&rpriv->uplink_priv.reoffload_flows_work); - mlx5_lag_remove(mdev); + mlx5_lag_remove(mdev, netdev); +} + +static MLX5E_DEFINE_STATS_GRP(sw_rep, 0); +static MLX5E_DEFINE_STATS_GRP(vport_rep, MLX5E_NDO_UPDATE_STATS); + +/* The stats groups order is opposite to the update_stats() order calls */ +static mlx5e_stats_grp_t mlx5e_rep_stats_grps[] = { + &MLX5E_STATS_GRP(sw_rep), + &MLX5E_STATS_GRP(vport_rep), +}; + +static unsigned int mlx5e_rep_stats_grps_num(struct mlx5e_priv *priv) +{ + return ARRAY_SIZE(mlx5e_rep_stats_grps); +} + +/* The stats groups order is opposite to the update_stats() order calls */ +static mlx5e_stats_grp_t mlx5e_ul_rep_stats_grps[] = { + &MLX5E_STATS_GRP(sw), + &MLX5E_STATS_GRP(qcnt), + &MLX5E_STATS_GRP(vnic_env), + &MLX5E_STATS_GRP(vport), + &MLX5E_STATS_GRP(802_3), + &MLX5E_STATS_GRP(2863), + &MLX5E_STATS_GRP(2819), + &MLX5E_STATS_GRP(phy), + &MLX5E_STATS_GRP(eth_ext), + &MLX5E_STATS_GRP(pcie), + &MLX5E_STATS_GRP(per_prio), + &MLX5E_STATS_GRP(pme), + &MLX5E_STATS_GRP(channels), + &MLX5E_STATS_GRP(per_port_buff_congest), +}; + +static unsigned int mlx5e_ul_rep_stats_grps_num(struct mlx5e_priv *priv) +{ + return ARRAY_SIZE(mlx5e_ul_rep_stats_grps); } static const struct mlx5e_profile mlx5e_rep_profile = { @@ -1807,29 +1919,33 @@ static const struct mlx5e_profile mlx5e_rep_profile = { .cleanup_tx = mlx5e_cleanup_rep_tx, .enable = mlx5e_rep_enable, .update_rx = mlx5e_update_rep_rx, - .update_stats = mlx5e_rep_update_hw_counters, + .update_stats = mlx5e_update_ndo_stats, .rx_handlers.handle_rx_cqe = mlx5e_handle_rx_cqe_rep, .rx_handlers.handle_rx_cqe_mpwqe = mlx5e_handle_rx_cqe_mpwrq, .max_tc = 1, .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), + .stats_grps = mlx5e_rep_stats_grps, + .stats_grps_num = mlx5e_rep_stats_grps_num, }; static const struct mlx5e_profile mlx5e_uplink_rep_profile = { .init = mlx5e_init_rep, .cleanup = mlx5e_cleanup_rep, - .init_rx = mlx5e_init_rep_rx, - .cleanup_rx = mlx5e_cleanup_rep_rx, + .init_rx = mlx5e_init_ul_rep_rx, + .cleanup_rx = mlx5e_cleanup_ul_rep_rx, .init_tx = mlx5e_init_rep_tx, .cleanup_tx = mlx5e_cleanup_rep_tx, .enable = mlx5e_uplink_rep_enable, .disable = mlx5e_uplink_rep_disable, .update_rx = mlx5e_update_rep_rx, - .update_stats = mlx5e_uplink_rep_update_hw_counters, + .update_stats = mlx5e_update_ndo_stats, .update_carrier = mlx5e_update_carrier, .rx_handlers.handle_rx_cqe = mlx5e_handle_rx_cqe_rep, .rx_handlers.handle_rx_cqe_mpwqe = mlx5e_handle_rx_cqe_mpwrq, .max_tc = MLX5E_MAX_NUM_TC, .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), + .stats_grps = mlx5e_ul_rep_stats_grps, + .stats_grps_num = mlx5e_ul_rep_stats_grps_num, }; static bool diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h index 31f83c8adcc9..3f756d51435f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rep.h @@ -73,6 +73,7 @@ struct mlx5_rep_uplink_priv { */ struct list_head tc_indr_block_priv_list; struct notifier_block netdevice_nb; + struct netdev_net_notifier netdevice_nn; struct mlx5_tun_entropy tun_entropy; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c index 9f09253f9f46..30b216d9284c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.c @@ -35,6 +35,58 @@ #include "en_accel/ipsec.h" #include "en_accel/tls.h" +static unsigned int stats_grps_num(struct mlx5e_priv *priv) +{ + return !priv->profile->stats_grps_num ? 0 : + priv->profile->stats_grps_num(priv); +} + +unsigned int mlx5e_stats_total_num(struct mlx5e_priv *priv) +{ + mlx5e_stats_grp_t *stats_grps = priv->profile->stats_grps; + const unsigned int num_stats_grps = stats_grps_num(priv); + unsigned int total = 0; + int i; + + for (i = 0; i < num_stats_grps; i++) + total += stats_grps[i]->get_num_stats(priv); + + return total; +} + +void mlx5e_stats_update(struct mlx5e_priv *priv) +{ + mlx5e_stats_grp_t *stats_grps = priv->profile->stats_grps; + const unsigned int num_stats_grps = stats_grps_num(priv); + int i; + + for (i = num_stats_grps - 1; i >= 0; i--) + if (stats_grps[i]->update_stats) + stats_grps[i]->update_stats(priv); +} + +void mlx5e_stats_fill(struct mlx5e_priv *priv, u64 *data, int idx) +{ + mlx5e_stats_grp_t *stats_grps = priv->profile->stats_grps; + const unsigned int num_stats_grps = stats_grps_num(priv); + int i; + + for (i = 0; i < num_stats_grps; i++) + idx = stats_grps[i]->fill_stats(priv, data, idx); +} + +void mlx5e_stats_fill_strings(struct mlx5e_priv *priv, u8 *data) +{ + mlx5e_stats_grp_t *stats_grps = priv->profile->stats_grps; + const unsigned int num_stats_grps = stats_grps_num(priv); + int i, idx = 0; + + for (i = 0; i < num_stats_grps; i++) + idx = stats_grps[i]->fill_strings(priv, data, idx); +} + +/* Concrete NIC Stats */ + static const struct counter_desc sw_stats_desc[] = { { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_packets) }, { MLX5E_DECLARE_STAT(struct mlx5e_sw_stats, rx_bytes) }, @@ -146,12 +198,12 @@ static const struct counter_desc sw_stats_desc[] = { #define NUM_SW_COUNTERS ARRAY_SIZE(sw_stats_desc) -static int mlx5e_grp_sw_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(sw) { return NUM_SW_COUNTERS; } -static int mlx5e_grp_sw_fill_strings(struct mlx5e_priv *priv, u8 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(sw) { int i; @@ -160,7 +212,7 @@ static int mlx5e_grp_sw_fill_strings(struct mlx5e_priv *priv, u8 *data, int idx) return idx; } -static int mlx5e_grp_sw_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(sw) { int i; @@ -169,7 +221,7 @@ static int mlx5e_grp_sw_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } -static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(sw) { struct mlx5e_sw_stats *s = &priv->stats.sw; int i; @@ -297,6 +349,9 @@ static void mlx5e_grp_sw_update_stats(struct mlx5e_priv *priv) s->tx_tls_drop_bypass_req += sq_stats->tls_drop_bypass_req; #endif s->tx_cqes += sq_stats->cqes; + + /* https://gcc.gnu.org/bugzilla/show_bug.cgi?id=92657 */ + barrier(); } } } @@ -312,7 +367,7 @@ static const struct counter_desc drop_rq_stats_desc[] = { #define NUM_Q_COUNTERS ARRAY_SIZE(q_stats_desc) #define NUM_DROP_RQ_COUNTERS ARRAY_SIZE(drop_rq_stats_desc) -static int mlx5e_grp_q_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(qcnt) { int num_stats = 0; @@ -325,7 +380,7 @@ static int mlx5e_grp_q_get_num_stats(struct mlx5e_priv *priv) return num_stats; } -static int mlx5e_grp_q_fill_strings(struct mlx5e_priv *priv, u8 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(qcnt) { int i; @@ -340,7 +395,7 @@ static int mlx5e_grp_q_fill_strings(struct mlx5e_priv *priv, u8 *data, int idx) return idx; } -static int mlx5e_grp_q_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(qcnt) { int i; @@ -353,7 +408,7 @@ static int mlx5e_grp_q_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } -static void mlx5e_grp_q_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(qcnt) { struct mlx5e_qcounter_stats *qcnt = &priv->stats.qcnt; u32 out[MLX5_ST_SZ_DW(query_q_counter_out)]; @@ -388,14 +443,13 @@ static const struct counter_desc vnic_env_stats_dev_oob_desc[] = { (MLX5_CAP_GEN(dev, vnic_env_int_rq_oob) ? \ ARRAY_SIZE(vnic_env_stats_dev_oob_desc) : 0) -static int mlx5e_grp_vnic_env_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(vnic_env) { return NUM_VNIC_ENV_STEER_COUNTERS(priv->mdev) + NUM_VNIC_ENV_DEV_OOB_COUNTERS(priv->mdev); } -static int mlx5e_grp_vnic_env_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(vnic_env) { int i; @@ -409,8 +463,7 @@ static int mlx5e_grp_vnic_env_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_vnic_env_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(vnic_env) { int i; @@ -424,7 +477,7 @@ static int mlx5e_grp_vnic_env_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_vnic_env_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(vnic_env) { u32 *out = (u32 *)priv->stats.vnic.query_vnic_env_out; int outlen = MLX5_ST_SZ_BYTES(query_vnic_env_out); @@ -487,13 +540,12 @@ static const struct counter_desc vport_stats_desc[] = { #define NUM_VPORT_COUNTERS ARRAY_SIZE(vport_stats_desc) -static int mlx5e_grp_vport_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(vport) { return NUM_VPORT_COUNTERS; } -static int mlx5e_grp_vport_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(vport) { int i; @@ -502,8 +554,7 @@ static int mlx5e_grp_vport_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_vport_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(vport) { int i; @@ -513,7 +564,7 @@ static int mlx5e_grp_vport_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_vport_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(vport) { int outlen = MLX5_ST_SZ_BYTES(query_vport_counter_out); u32 *out = (u32 *)priv->stats.vport.query_vport_out; @@ -552,13 +603,12 @@ static const struct counter_desc pport_802_3_stats_desc[] = { #define NUM_PPORT_802_3_COUNTERS ARRAY_SIZE(pport_802_3_stats_desc) -static int mlx5e_grp_802_3_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(802_3) { return NUM_PPORT_802_3_COUNTERS; } -static int mlx5e_grp_802_3_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(802_3) { int i; @@ -567,8 +617,7 @@ static int mlx5e_grp_802_3_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_802_3_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(802_3) { int i; @@ -581,7 +630,7 @@ static int mlx5e_grp_802_3_fill_stats(struct mlx5e_priv *priv, u64 *data, #define MLX5_BASIC_PPCNT_SUPPORTED(mdev) \ (MLX5_CAP_GEN(mdev, pcam_reg) ? MLX5_CAP_PCAM_REG(mdev, ppcnt) : 1) -void mlx5e_grp_802_3_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(802_3) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -609,13 +658,12 @@ static const struct counter_desc pport_2863_stats_desc[] = { #define NUM_PPORT_2863_COUNTERS ARRAY_SIZE(pport_2863_stats_desc) -static int mlx5e_grp_2863_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(2863) { return NUM_PPORT_2863_COUNTERS; } -static int mlx5e_grp_2863_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(2863) { int i; @@ -624,8 +672,7 @@ static int mlx5e_grp_2863_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_2863_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(2863) { int i; @@ -635,7 +682,7 @@ static int mlx5e_grp_2863_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_2863_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(2863) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -670,13 +717,12 @@ static const struct counter_desc pport_2819_stats_desc[] = { #define NUM_PPORT_2819_COUNTERS ARRAY_SIZE(pport_2819_stats_desc) -static int mlx5e_grp_2819_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(2819) { return NUM_PPORT_2819_COUNTERS; } -static int mlx5e_grp_2819_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(2819) { int i; @@ -685,8 +731,7 @@ static int mlx5e_grp_2819_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_2819_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(2819) { int i; @@ -696,7 +741,7 @@ static int mlx5e_grp_2819_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_2819_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(2819) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -734,7 +779,7 @@ pport_phy_statistical_err_lanes_stats_desc[] = { #define NUM_PPORT_PHY_STATISTICAL_PER_LANE_COUNTERS \ ARRAY_SIZE(pport_phy_statistical_err_lanes_stats_desc) -static int mlx5e_grp_phy_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(phy) { struct mlx5_core_dev *mdev = priv->mdev; int num_stats; @@ -751,8 +796,7 @@ static int mlx5e_grp_phy_get_num_stats(struct mlx5e_priv *priv) return num_stats; } -static int mlx5e_grp_phy_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(phy) { struct mlx5_core_dev *mdev = priv->mdev; int i; @@ -774,7 +818,7 @@ static int mlx5e_grp_phy_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_phy_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(phy) { struct mlx5_core_dev *mdev = priv->mdev; int i; @@ -800,7 +844,7 @@ static int mlx5e_grp_phy_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) return idx; } -static void mlx5e_grp_phy_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(phy) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -830,7 +874,7 @@ static const struct counter_desc pport_eth_ext_stats_desc[] = { #define NUM_PPORT_ETH_EXT_COUNTERS ARRAY_SIZE(pport_eth_ext_stats_desc) -static int mlx5e_grp_eth_ext_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(eth_ext) { if (MLX5_CAP_PCAM_FEATURE((priv)->mdev, rx_buffer_fullness_counters)) return NUM_PPORT_ETH_EXT_COUNTERS; @@ -838,8 +882,7 @@ static int mlx5e_grp_eth_ext_get_num_stats(struct mlx5e_priv *priv) return 0; } -static int mlx5e_grp_eth_ext_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(eth_ext) { int i; @@ -850,8 +893,7 @@ static int mlx5e_grp_eth_ext_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_eth_ext_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(eth_ext) { int i; @@ -863,7 +905,7 @@ static int mlx5e_grp_eth_ext_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_eth_ext_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(eth_ext) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -904,7 +946,7 @@ static const struct counter_desc pcie_perf_stall_stats_desc[] = { #define NUM_PCIE_PERF_COUNTERS64 ARRAY_SIZE(pcie_perf_stats_desc64) #define NUM_PCIE_PERF_STALL_COUNTERS ARRAY_SIZE(pcie_perf_stall_stats_desc) -static int mlx5e_grp_pcie_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(pcie) { int num_stats = 0; @@ -920,8 +962,7 @@ static int mlx5e_grp_pcie_get_num_stats(struct mlx5e_priv *priv) return num_stats; } -static int mlx5e_grp_pcie_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(pcie) { int i; @@ -942,8 +983,7 @@ static int mlx5e_grp_pcie_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_pcie_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(pcie) { int i; @@ -967,7 +1007,7 @@ static int mlx5e_grp_pcie_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static void mlx5e_grp_pcie_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(pcie) { struct mlx5e_pcie_stats *pcie_stats = &priv->stats.pcie; struct mlx5_core_dev *mdev = priv->mdev; @@ -1015,8 +1055,7 @@ static int mlx5e_grp_per_tc_prio_get_num_stats(struct mlx5e_priv *priv) return NUM_PPORT_PER_TC_PRIO_COUNTERS * NUM_PPORT_PRIO; } -static int mlx5e_grp_per_port_buffer_congest_fill_strings(struct mlx5e_priv *priv, - u8 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(per_port_buff_congest) { struct mlx5_core_dev *mdev = priv->mdev; int i, prio; @@ -1036,8 +1075,7 @@ static int mlx5e_grp_per_port_buffer_congest_fill_strings(struct mlx5e_priv *pri return idx; } -static int mlx5e_grp_per_port_buffer_congest_fill_stats(struct mlx5e_priv *priv, - u64 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(per_port_buff_congest) { struct mlx5e_pport_stats *pport = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -1112,13 +1150,13 @@ static void mlx5e_grp_per_tc_congest_prio_update_stats(struct mlx5e_priv *priv) } } -static int mlx5e_grp_per_port_buffer_congest_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(per_port_buff_congest) { return mlx5e_grp_per_tc_prio_get_num_stats(priv) + mlx5e_grp_per_tc_congest_prio_get_num_stats(priv); } -static void mlx5e_grp_per_port_buffer_congest_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(per_port_buff_congest) { mlx5e_grp_per_tc_prio_update_stats(priv); mlx5e_grp_per_tc_congest_prio_update_stats(priv); @@ -1130,6 +1168,7 @@ static void mlx5e_grp_per_port_buffer_congest_update_stats(struct mlx5e_priv *pr static const struct counter_desc pport_per_prio_traffic_stats_desc[] = { { "rx_prio%d_bytes", PPORT_PER_PRIO_OFF(rx_octets) }, { "rx_prio%d_packets", PPORT_PER_PRIO_OFF(rx_frames) }, + { "rx_prio%d_discards", PPORT_PER_PRIO_OFF(rx_discards) }, { "tx_prio%d_bytes", PPORT_PER_PRIO_OFF(tx_octets) }, { "tx_prio%d_packets", PPORT_PER_PRIO_OFF(tx_frames) }, }; @@ -1292,29 +1331,27 @@ static int mlx5e_grp_per_prio_pfc_fill_stats(struct mlx5e_priv *priv, return idx; } -static int mlx5e_grp_per_prio_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(per_prio) { return mlx5e_grp_per_prio_traffic_get_num_stats() + mlx5e_grp_per_prio_pfc_get_num_stats(priv); } -static int mlx5e_grp_per_prio_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(per_prio) { idx = mlx5e_grp_per_prio_traffic_fill_strings(priv, data, idx); idx = mlx5e_grp_per_prio_pfc_fill_strings(priv, data, idx); return idx; } -static int mlx5e_grp_per_prio_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(per_prio) { idx = mlx5e_grp_per_prio_traffic_fill_stats(priv, data, idx); idx = mlx5e_grp_per_prio_pfc_fill_stats(priv, data, idx); return idx; } -static void mlx5e_grp_per_prio_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(per_prio) { struct mlx5e_pport_stats *pstats = &priv->stats.pport; struct mlx5_core_dev *mdev = priv->mdev; @@ -1349,13 +1386,12 @@ static const struct counter_desc mlx5e_pme_error_desc[] = { #define NUM_PME_STATUS_STATS ARRAY_SIZE(mlx5e_pme_status_desc) #define NUM_PME_ERR_STATS ARRAY_SIZE(mlx5e_pme_error_desc) -static int mlx5e_grp_pme_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(pme) { return NUM_PME_STATUS_STATS + NUM_PME_ERR_STATS; } -static int mlx5e_grp_pme_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(pme) { int i; @@ -1368,8 +1404,7 @@ static int mlx5e_grp_pme_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_pme_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(pme) { struct mlx5_pme_stats pme_stats; int i; @@ -1387,45 +1422,46 @@ static int mlx5e_grp_pme_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } -static int mlx5e_grp_ipsec_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(pme) { return; } + +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(ipsec) { return mlx5e_ipsec_get_count(priv); } -static int mlx5e_grp_ipsec_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(ipsec) { return idx + mlx5e_ipsec_get_strings(priv, data + idx * ETH_GSTRING_LEN); } -static int mlx5e_grp_ipsec_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(ipsec) { return idx + mlx5e_ipsec_get_stats(priv, data + idx); } -static void mlx5e_grp_ipsec_update_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(ipsec) { mlx5e_ipsec_update_stats(priv); } -static int mlx5e_grp_tls_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(tls) { return mlx5e_tls_get_count(priv); } -static int mlx5e_grp_tls_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(tls) { return idx + mlx5e_tls_get_strings(priv, data + idx * ETH_GSTRING_LEN); } -static int mlx5e_grp_tls_fill_stats(struct mlx5e_priv *priv, u64 *data, int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(tls) { return idx + mlx5e_tls_get_stats(priv, data + idx); } +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(tls) { return; } + static const struct counter_desc rq_stats_desc[] = { { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, packets) }, { MLX5E_DECLARE_RX_STAT(struct mlx5e_rq_stats, bytes) }, @@ -1559,7 +1595,7 @@ static const struct counter_desc ch_stats_desc[] = { #define NUM_XSKSQ_STATS ARRAY_SIZE(xsksq_stats_desc) #define NUM_CH_STATS ARRAY_SIZE(ch_stats_desc) -static int mlx5e_grp_channels_get_num_stats(struct mlx5e_priv *priv) +static MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(channels) { int max_nch = priv->max_nch; @@ -1572,8 +1608,7 @@ static int mlx5e_grp_channels_get_num_stats(struct mlx5e_priv *priv) (NUM_XSKSQ_STATS * max_nch * priv->xsk.ever_used); } -static int mlx5e_grp_channels_fill_strings(struct mlx5e_priv *priv, u8 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(channels) { bool is_xsk = priv->xsk.ever_used; int max_nch = priv->max_nch; @@ -1615,8 +1650,7 @@ static int mlx5e_grp_channels_fill_strings(struct mlx5e_priv *priv, u8 *data, return idx; } -static int mlx5e_grp_channels_fill_stats(struct mlx5e_priv *priv, u64 *data, - int idx) +static MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(channels) { bool is_xsk = priv->xsk.ever_used; int max_nch = priv->max_nch; @@ -1664,104 +1698,46 @@ static int mlx5e_grp_channels_fill_stats(struct mlx5e_priv *priv, u64 *data, return idx; } +static MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(channels) { return; } + +MLX5E_DEFINE_STATS_GRP(sw, 0); +MLX5E_DEFINE_STATS_GRP(qcnt, MLX5E_NDO_UPDATE_STATS); +MLX5E_DEFINE_STATS_GRP(vnic_env, 0); +MLX5E_DEFINE_STATS_GRP(vport, MLX5E_NDO_UPDATE_STATS); +MLX5E_DEFINE_STATS_GRP(802_3, MLX5E_NDO_UPDATE_STATS); +MLX5E_DEFINE_STATS_GRP(2863, 0); +MLX5E_DEFINE_STATS_GRP(2819, 0); +MLX5E_DEFINE_STATS_GRP(phy, 0); +MLX5E_DEFINE_STATS_GRP(pcie, 0); +MLX5E_DEFINE_STATS_GRP(per_prio, 0); +MLX5E_DEFINE_STATS_GRP(pme, 0); +MLX5E_DEFINE_STATS_GRP(channels, 0); +MLX5E_DEFINE_STATS_GRP(per_port_buff_congest, 0); +MLX5E_DEFINE_STATS_GRP(eth_ext, 0); +static MLX5E_DEFINE_STATS_GRP(ipsec, 0); +static MLX5E_DEFINE_STATS_GRP(tls, 0); + /* The stats groups order is opposite to the update_stats() order calls */ -const struct mlx5e_stats_grp mlx5e_stats_grps[] = { - { - .get_num_stats = mlx5e_grp_sw_get_num_stats, - .fill_strings = mlx5e_grp_sw_fill_strings, - .fill_stats = mlx5e_grp_sw_fill_stats, - .update_stats = mlx5e_grp_sw_update_stats, - }, - { - .get_num_stats = mlx5e_grp_q_get_num_stats, - .fill_strings = mlx5e_grp_q_fill_strings, - .fill_stats = mlx5e_grp_q_fill_stats, - .update_stats_mask = MLX5E_NDO_UPDATE_STATS, - .update_stats = mlx5e_grp_q_update_stats, - }, - { - .get_num_stats = mlx5e_grp_vnic_env_get_num_stats, - .fill_strings = mlx5e_grp_vnic_env_fill_strings, - .fill_stats = mlx5e_grp_vnic_env_fill_stats, - .update_stats = mlx5e_grp_vnic_env_update_stats, - }, - { - .get_num_stats = mlx5e_grp_vport_get_num_stats, - .fill_strings = mlx5e_grp_vport_fill_strings, - .fill_stats = mlx5e_grp_vport_fill_stats, - .update_stats_mask = MLX5E_NDO_UPDATE_STATS, - .update_stats = mlx5e_grp_vport_update_stats, - }, - { - .get_num_stats = mlx5e_grp_802_3_get_num_stats, - .fill_strings = mlx5e_grp_802_3_fill_strings, - .fill_stats = mlx5e_grp_802_3_fill_stats, - .update_stats_mask = MLX5E_NDO_UPDATE_STATS, - .update_stats = mlx5e_grp_802_3_update_stats, - }, - { - .get_num_stats = mlx5e_grp_2863_get_num_stats, - .fill_strings = mlx5e_grp_2863_fill_strings, - .fill_stats = mlx5e_grp_2863_fill_stats, - .update_stats = mlx5e_grp_2863_update_stats, - }, - { - .get_num_stats = mlx5e_grp_2819_get_num_stats, - .fill_strings = mlx5e_grp_2819_fill_strings, - .fill_stats = mlx5e_grp_2819_fill_stats, - .update_stats = mlx5e_grp_2819_update_stats, - }, - { - .get_num_stats = mlx5e_grp_phy_get_num_stats, - .fill_strings = mlx5e_grp_phy_fill_strings, - .fill_stats = mlx5e_grp_phy_fill_stats, - .update_stats = mlx5e_grp_phy_update_stats, - }, - { - .get_num_stats = mlx5e_grp_eth_ext_get_num_stats, - .fill_strings = mlx5e_grp_eth_ext_fill_strings, - .fill_stats = mlx5e_grp_eth_ext_fill_stats, - .update_stats = mlx5e_grp_eth_ext_update_stats, - }, - { - .get_num_stats = mlx5e_grp_pcie_get_num_stats, - .fill_strings = mlx5e_grp_pcie_fill_strings, - .fill_stats = mlx5e_grp_pcie_fill_stats, - .update_stats = mlx5e_grp_pcie_update_stats, - }, - { - .get_num_stats = mlx5e_grp_per_prio_get_num_stats, - .fill_strings = mlx5e_grp_per_prio_fill_strings, - .fill_stats = mlx5e_grp_per_prio_fill_stats, - .update_stats = mlx5e_grp_per_prio_update_stats, - }, - { - .get_num_stats = mlx5e_grp_pme_get_num_stats, - .fill_strings = mlx5e_grp_pme_fill_strings, - .fill_stats = mlx5e_grp_pme_fill_stats, - }, - { - .get_num_stats = mlx5e_grp_ipsec_get_num_stats, - .fill_strings = mlx5e_grp_ipsec_fill_strings, - .fill_stats = mlx5e_grp_ipsec_fill_stats, - .update_stats = mlx5e_grp_ipsec_update_stats, - }, - { - .get_num_stats = mlx5e_grp_tls_get_num_stats, - .fill_strings = mlx5e_grp_tls_fill_strings, - .fill_stats = mlx5e_grp_tls_fill_stats, - }, - { - .get_num_stats = mlx5e_grp_channels_get_num_stats, - .fill_strings = mlx5e_grp_channels_fill_strings, - .fill_stats = mlx5e_grp_channels_fill_stats, - }, - { - .get_num_stats = mlx5e_grp_per_port_buffer_congest_get_num_stats, - .fill_strings = mlx5e_grp_per_port_buffer_congest_fill_strings, - .fill_stats = mlx5e_grp_per_port_buffer_congest_fill_stats, - .update_stats = mlx5e_grp_per_port_buffer_congest_update_stats, - }, +mlx5e_stats_grp_t mlx5e_nic_stats_grps[] = { + &MLX5E_STATS_GRP(sw), + &MLX5E_STATS_GRP(qcnt), + &MLX5E_STATS_GRP(vnic_env), + &MLX5E_STATS_GRP(vport), + &MLX5E_STATS_GRP(802_3), + &MLX5E_STATS_GRP(2863), + &MLX5E_STATS_GRP(2819), + &MLX5E_STATS_GRP(phy), + &MLX5E_STATS_GRP(eth_ext), + &MLX5E_STATS_GRP(pcie), + &MLX5E_STATS_GRP(per_prio), + &MLX5E_STATS_GRP(pme), + &MLX5E_STATS_GRP(ipsec), + &MLX5E_STATS_GRP(tls), + &MLX5E_STATS_GRP(channels), + &MLX5E_STATS_GRP(per_port_buff_congest), }; -const int mlx5e_num_stats_grps = ARRAY_SIZE(mlx5e_stats_grps); +unsigned int mlx5e_nic_stats_grps_num(struct mlx5e_priv *priv) +{ + return ARRAY_SIZE(mlx5e_nic_stats_grps); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h index 869f3502f631..092b39ffa32a 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_stats.h @@ -29,6 +29,7 @@ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. */ + #ifndef __MLX5_EN_STATS_H__ #define __MLX5_EN_STATS_H__ @@ -55,6 +56,56 @@ struct counter_desc { size_t offset; /* Byte offset */ }; +enum { + MLX5E_NDO_UPDATE_STATS = BIT(0x1), +}; + +struct mlx5e_priv; +struct mlx5e_stats_grp { + u16 update_stats_mask; + int (*get_num_stats)(struct mlx5e_priv *priv); + int (*fill_strings)(struct mlx5e_priv *priv, u8 *data, int idx); + int (*fill_stats)(struct mlx5e_priv *priv, u64 *data, int idx); + void (*update_stats)(struct mlx5e_priv *priv); +}; + +typedef const struct mlx5e_stats_grp *const mlx5e_stats_grp_t; + +#define MLX5E_STATS_GRP_OP(grp, name) mlx5e_stats_grp_ ## grp ## _ ## name + +#define MLX5E_DECLARE_STATS_GRP_OP_NUM_STATS(grp) \ + int MLX5E_STATS_GRP_OP(grp, num_stats)(struct mlx5e_priv *priv) + +#define MLX5E_DECLARE_STATS_GRP_OP_UPDATE_STATS(grp) \ + void MLX5E_STATS_GRP_OP(grp, update_stats)(struct mlx5e_priv *priv) + +#define MLX5E_DECLARE_STATS_GRP_OP_FILL_STRS(grp) \ + int MLX5E_STATS_GRP_OP(grp, fill_strings)(struct mlx5e_priv *priv, u8 *data, int idx) + +#define MLX5E_DECLARE_STATS_GRP_OP_FILL_STATS(grp) \ + int MLX5E_STATS_GRP_OP(grp, fill_stats)(struct mlx5e_priv *priv, u64 *data, int idx) + +#define MLX5E_STATS_GRP(grp) mlx5e_stats_grp_ ## grp + +#define MLX5E_DECLARE_STATS_GRP(grp) \ + const struct mlx5e_stats_grp MLX5E_STATS_GRP(grp) + +#define MLX5E_DEFINE_STATS_GRP(grp, mask) \ +MLX5E_DECLARE_STATS_GRP(grp) = { \ + .get_num_stats = MLX5E_STATS_GRP_OP(grp, num_stats), \ + .fill_stats = MLX5E_STATS_GRP_OP(grp, fill_stats), \ + .fill_strings = MLX5E_STATS_GRP_OP(grp, fill_strings), \ + .update_stats = MLX5E_STATS_GRP_OP(grp, update_stats), \ + .update_stats_mask = mask, \ +} + +unsigned int mlx5e_stats_total_num(struct mlx5e_priv *priv); +void mlx5e_stats_update(struct mlx5e_priv *priv); +void mlx5e_stats_fill(struct mlx5e_priv *priv, u64 *data, int idx); +void mlx5e_stats_fill_strings(struct mlx5e_priv *priv, u8 *data); + +/* Concrete NIC Stats */ + struct mlx5e_sw_stats { u64 rx_packets; u64 rx_bytes; @@ -322,22 +373,22 @@ struct mlx5e_stats { struct mlx5e_pcie_stats pcie; }; -enum { - MLX5E_NDO_UPDATE_STATS = BIT(0x1), -}; - -struct mlx5e_priv; -struct mlx5e_stats_grp { - u16 update_stats_mask; - int (*get_num_stats)(struct mlx5e_priv *priv); - int (*fill_strings)(struct mlx5e_priv *priv, u8 *data, int idx); - int (*fill_stats)(struct mlx5e_priv *priv, u64 *data, int idx); - void (*update_stats)(struct mlx5e_priv *priv); -}; - -extern const struct mlx5e_stats_grp mlx5e_stats_grps[]; -extern const int mlx5e_num_stats_grps; +extern mlx5e_stats_grp_t mlx5e_nic_stats_grps[]; +unsigned int mlx5e_nic_stats_grps_num(struct mlx5e_priv *priv); -void mlx5e_grp_802_3_update_stats(struct mlx5e_priv *priv); +extern MLX5E_DECLARE_STATS_GRP(sw); +extern MLX5E_DECLARE_STATS_GRP(qcnt); +extern MLX5E_DECLARE_STATS_GRP(vnic_env); +extern MLX5E_DECLARE_STATS_GRP(vport); +extern MLX5E_DECLARE_STATS_GRP(802_3); +extern MLX5E_DECLARE_STATS_GRP(2863); +extern MLX5E_DECLARE_STATS_GRP(2819); +extern MLX5E_DECLARE_STATS_GRP(phy); +extern MLX5E_DECLARE_STATS_GRP(eth_ext); +extern MLX5E_DECLARE_STATS_GRP(pcie); +extern MLX5E_DECLARE_STATS_GRP(per_prio); +extern MLX5E_DECLARE_STATS_GRP(pme); +extern MLX5E_DECLARE_STATS_GRP(channels); +extern MLX5E_DECLARE_STATS_GRP(per_port_buff_congest); #endif /* __MLX5_EN_STATS_H__ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c index 7e32b9e3667c..74091f72c9a8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tc.c @@ -51,6 +51,7 @@ #include "en_rep.h" #include "en_tc.h" #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "fs_core.h" #include "en/port.h" #include "en/tc_tun.h" @@ -960,7 +961,8 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, mutex_lock(&priv->fs.tc.t_lock); if (IS_ERR_OR_NULL(priv->fs.tc.t)) { - int tc_grp_size, tc_tbl_size; + struct mlx5_flow_table_attr ft_attr = {}; + int tc_grp_size, tc_tbl_size, tc_num_grps; u32 max_flow_counter; max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | @@ -970,13 +972,15 @@ mlx5e_tc_add_nic_flow(struct mlx5e_priv *priv, tc_tbl_size = min_t(int, tc_grp_size * MLX5E_TC_TABLE_NUM_GROUPS, BIT(MLX5_CAP_FLOWTABLE_NIC_RX(dev, log_max_ft_size))); + tc_num_grps = MLX5E_TC_TABLE_NUM_GROUPS; + ft_attr.prio = MLX5E_TC_PRIO; + ft_attr.max_fte = tc_tbl_size; + ft_attr.level = MLX5E_TC_FT_LEVEL; + ft_attr.autogroup.max_num_groups = tc_num_grps; priv->fs.tc.t = mlx5_create_auto_grouped_flow_table(priv->fs.ns, - MLX5E_TC_PRIO, - tc_tbl_size, - MLX5E_TC_TABLE_NUM_GROUPS, - MLX5E_TC_FT_LEVEL, 0); + &ft_attr); if (IS_ERR(priv->fs.tc.t)) { mutex_unlock(&priv->fs.tc.t_lock); NL_SET_ERR_MSG_MOD(extack, @@ -1080,7 +1084,7 @@ mlx5e_tc_offload_to_slow_path(struct mlx5_eswitch *esw, memcpy(slow_attr, flow->esw_attr, sizeof(*slow_attr)); slow_attr->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; slow_attr->split_count = 0; - slow_attr->dest_chain = FDB_TC_SLOW_PATH_CHAIN; + slow_attr->flags |= MLX5_ESW_ATTR_FLAG_SLOW_PATH; rule = mlx5e_tc_offload_fdb_rules(esw, flow, spec, slow_attr); if (!IS_ERR(rule)) @@ -1097,7 +1101,7 @@ mlx5e_tc_unoffload_from_slow_path(struct mlx5_eswitch *esw, memcpy(slow_attr, flow->esw_attr, sizeof(*slow_attr)); slow_attr->action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; slow_attr->split_count = 0; - slow_attr->dest_chain = FDB_TC_SLOW_PATH_CHAIN; + slow_attr->flags |= MLX5_ESW_ATTR_FLAG_SLOW_PATH; mlx5e_tc_unoffload_fdb_rules(esw, flow, slow_attr); flow_flag_clear(flow, SLOW); } @@ -1157,19 +1161,18 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv, struct netlink_ext_ack *extack) { struct mlx5_eswitch *esw = priv->mdev->priv.eswitch; - u32 max_chain = mlx5_eswitch_get_chain_range(esw); struct mlx5_esw_flow_attr *attr = flow->esw_attr; struct mlx5e_tc_flow_parse_attr *parse_attr = attr->parse_attr; - u16 max_prio = mlx5_eswitch_get_prio_range(esw); struct net_device *out_dev, *encap_dev = NULL; struct mlx5_fc *counter = NULL; struct mlx5e_rep_priv *rpriv; struct mlx5e_priv *out_priv; bool encap_valid = true; + u32 max_prio, max_chain; int err = 0; int out_index; - if (!mlx5_eswitch_prios_supported(esw) && attr->prio != 1) { + if (!mlx5_esw_chains_prios_supported(esw) && attr->prio != 1) { NL_SET_ERR_MSG(extack, "E-switch priorities unsupported, upgrade FW"); return -EOPNOTSUPP; } @@ -1179,11 +1182,13 @@ mlx5e_tc_add_fdb_flow(struct mlx5e_priv *priv, * FDB_FT_CHAIN which is outside tc range. * See mlx5e_rep_setup_ft_cb(). */ + max_chain = mlx5_esw_chains_get_chain_range(esw); if (!mlx5e_is_ft_flow(flow) && attr->chain > max_chain) { NL_SET_ERR_MSG(extack, "Requested chain is out of supported range"); return -EOPNOTSUPP; } + max_prio = mlx5_esw_chains_get_prio_range(esw); if (attr->prio > max_prio) { NL_SET_ERR_MSG(extack, "Requested priority is out of supported range"); return -EOPNOTSUPP; @@ -1805,6 +1810,40 @@ static void *get_match_headers_value(u32 flags, outer_headers); } +static int mlx5e_flower_parse_meta(struct net_device *filter_dev, + struct flow_cls_offload *f) +{ + struct flow_rule *rule = flow_cls_offload_flow_rule(f); + struct netlink_ext_ack *extack = f->common.extack; + struct net_device *ingress_dev; + struct flow_match_meta match; + + if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_META)) + return 0; + + flow_rule_match_meta(rule, &match); + if (match.mask->ingress_ifindex != 0xFFFFFFFF) { + NL_SET_ERR_MSG_MOD(extack, "Unsupported ingress ifindex mask"); + return -EINVAL; + } + + ingress_dev = __dev_get_by_index(dev_net(filter_dev), + match.key->ingress_ifindex); + if (!ingress_dev) { + NL_SET_ERR_MSG_MOD(extack, + "Can't find the ingress port to match on"); + return -EINVAL; + } + + if (ingress_dev != filter_dev) { + NL_SET_ERR_MSG_MOD(extack, + "Can't match on the ingress filter port"); + return -EINVAL; + } + + return 0; +} + static int __parse_cls_flower(struct mlx5e_priv *priv, struct mlx5_flow_spec *spec, struct flow_cls_offload *f, @@ -1825,6 +1864,7 @@ static int __parse_cls_flower(struct mlx5e_priv *priv, u16 addr_type = 0; u8 ip_proto = 0; u8 *match_level; + int err; match_level = outer_match_level; @@ -1868,6 +1908,10 @@ static int __parse_cls_flower(struct mlx5e_priv *priv, spec); } + err = mlx5e_flower_parse_meta(filter_dev, f); + if (err) + return err; + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { struct flow_match_basic match; @@ -2842,6 +2886,10 @@ static int parse_tc_nic_actions(struct mlx5e_priv *priv, flow_action_for_each(i, act, flow_action) { switch (act->id) { + case FLOW_ACTION_ACCEPT: + action |= MLX5_FLOW_CONTEXT_ACTION_FWD_DEST | + MLX5_FLOW_CONTEXT_ACTION_COUNT; + break; case FLOW_ACTION_DROP: action |= MLX5_FLOW_CONTEXT_ACTION_DROP; if (MLX5_CAP_FLOWTABLE(priv->mdev, @@ -3462,7 +3510,7 @@ static int parse_tc_fdb_actions(struct mlx5e_priv *priv, break; case FLOW_ACTION_GOTO: { u32 dest_chain = act->chain_index; - u32 max_chain = mlx5_eswitch_get_chain_range(esw); + u32 max_chain = mlx5_esw_chains_get_chain_range(esw); if (ft_flow) { NL_SET_ERR_MSG_MOD(extack, "Goto action is not supported"); @@ -4203,7 +4251,10 @@ int mlx5e_tc_nic_init(struct mlx5e_priv *priv) return err; tc->netdevice_nb.notifier_call = mlx5e_tc_netdev_event; - if (register_netdevice_notifier(&tc->netdevice_nb)) { + err = register_netdevice_notifier_dev_net(priv->netdev, + &tc->netdevice_nb, + &tc->netdevice_nn); + if (err) { tc->netdevice_nb.notifier_call = NULL; mlx5_core_warn(priv->mdev, "Failed to register netdev notifier\n"); } @@ -4225,7 +4276,9 @@ void mlx5e_tc_nic_cleanup(struct mlx5e_priv *priv) struct mlx5e_tc_table *tc = &priv->fs.tc; if (tc->netdevice_nb.notifier_call) - unregister_netdevice_notifier(&tc->netdevice_nb); + unregister_netdevice_notifier_dev_net(priv->netdev, + &tc->netdevice_nb, + &tc->netdevice_nn); mutex_destroy(&tc->mod_hdr.lock); mutex_destroy(&tc->hairpin_tbl_lock); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c index 580c71cb9dfa..cccea3a8eddd 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c @@ -156,7 +156,8 @@ static int mlx5_eq_comp_int(struct notifier_block *nb, cq->comp(cq, eqe); mlx5_cq_put(cq); } else { - mlx5_core_warn(eq->dev, "Completion event for bogus CQ 0x%x\n", cqn); + dev_dbg_ratelimited(eq->dev->device, + "Completion event for bogus CQ 0x%x\n", cqn); } ++eq->cons_index; @@ -563,6 +564,39 @@ static void gather_async_events_mask(struct mlx5_core_dev *dev, u64 mask[4]) gather_user_async_events(dev, mask); } +static int +setup_async_eq(struct mlx5_core_dev *dev, struct mlx5_eq_async *eq, + struct mlx5_eq_param *param, const char *name) +{ + int err; + + eq->irq_nb.notifier_call = mlx5_eq_async_int; + + err = create_async_eq(dev, &eq->core, param); + if (err) { + mlx5_core_warn(dev, "failed to create %s EQ %d\n", name, err); + return err; + } + err = mlx5_eq_enable(dev, &eq->core, &eq->irq_nb); + if (err) { + mlx5_core_warn(dev, "failed to enable %s EQ %d\n", name, err); + destroy_async_eq(dev, &eq->core); + } + return err; +} + +static void cleanup_async_eq(struct mlx5_core_dev *dev, + struct mlx5_eq_async *eq, const char *name) +{ + int err; + + mlx5_eq_disable(dev, &eq->core, &eq->irq_nb); + err = destroy_async_eq(dev, &eq->core); + if (err) + mlx5_core_err(dev, "failed to destroy %s eq, err(%d)\n", + name, err); +} + static int create_async_eqs(struct mlx5_core_dev *dev) { struct mlx5_eq_table *table = dev->priv.eq_table; @@ -572,77 +606,45 @@ static int create_async_eqs(struct mlx5_core_dev *dev) MLX5_NB_INIT(&table->cq_err_nb, cq_err_event_notifier, CQ_ERROR); mlx5_eq_notifier_register(dev, &table->cq_err_nb); - table->cmd_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = MLX5_NUM_CMD_EQE, + .mask[0] = 1ull << MLX5_EVENT_TYPE_CMD, }; - - param.mask[0] = 1ull << MLX5_EVENT_TYPE_CMD; - err = create_async_eq(dev, &table->cmd_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create cmd EQ %d\n", err); - goto err0; - } - err = mlx5_eq_enable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable cmd EQ %d\n", err); + err = setup_async_eq(dev, &table->cmd_eq, ¶m, "cmd"); + if (err) goto err1; - } + mlx5_cmd_use_events(dev); - table->async_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = MLX5_NUM_ASYNC_EQE, }; gather_async_events_mask(dev, param.mask); - err = create_async_eq(dev, &table->async_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create async EQ %d\n", err); + err = setup_async_eq(dev, &table->async_eq, ¶m, "async"); + if (err) goto err2; - } - err = mlx5_eq_enable(dev, &table->async_eq.core, - &table->async_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable async EQ %d\n", err); - goto err3; - } - table->pages_eq.irq_nb.notifier_call = mlx5_eq_async_int; param = (struct mlx5_eq_param) { .irq_index = 0, .nent = /* TODO: sriov max_vf + */ 1, + .mask[0] = 1ull << MLX5_EVENT_TYPE_PAGE_REQUEST, }; - param.mask[0] = 1ull << MLX5_EVENT_TYPE_PAGE_REQUEST; - err = create_async_eq(dev, &table->pages_eq.core, ¶m); - if (err) { - mlx5_core_warn(dev, "failed to create pages EQ %d\n", err); - goto err4; - } - err = mlx5_eq_enable(dev, &table->pages_eq.core, - &table->pages_eq.irq_nb); - if (err) { - mlx5_core_warn(dev, "failed to enable pages EQ %d\n", err); - goto err5; - } + err = setup_async_eq(dev, &table->pages_eq, ¶m, "pages"); + if (err) + goto err3; - return err; + return 0; -err5: - destroy_async_eq(dev, &table->pages_eq.core); -err4: - mlx5_eq_disable(dev, &table->async_eq.core, &table->async_eq.irq_nb); err3: - destroy_async_eq(dev, &table->async_eq.core); + cleanup_async_eq(dev, &table->async_eq, "async"); err2: mlx5_cmd_use_polling(dev); - mlx5_eq_disable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); + cleanup_async_eq(dev, &table->cmd_eq, "cmd"); err1: - destroy_async_eq(dev, &table->cmd_eq.core); -err0: mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); return err; } @@ -650,28 +652,11 @@ err0: static void destroy_async_eqs(struct mlx5_core_dev *dev) { struct mlx5_eq_table *table = dev->priv.eq_table; - int err; - - mlx5_eq_disable(dev, &table->pages_eq.core, &table->pages_eq.irq_nb); - err = destroy_async_eq(dev, &table->pages_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy pages eq, err(%d)\n", - err); - - mlx5_eq_disable(dev, &table->async_eq.core, &table->async_eq.irq_nb); - err = destroy_async_eq(dev, &table->async_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy async eq, err(%d)\n", - err); + cleanup_async_eq(dev, &table->pages_eq, "pages"); + cleanup_async_eq(dev, &table->async_eq, "async"); mlx5_cmd_use_polling(dev); - - mlx5_eq_disable(dev, &table->cmd_eq.core, &table->cmd_eq.irq_nb); - err = destroy_async_eq(dev, &table->cmd_eq.core); - if (err) - mlx5_core_err(dev, "failed to destroy command eq, err(%d)\n", - err); - + cleanup_async_eq(dev, &table->cmd_eq, "cmd"); mlx5_eq_notifier_unregister(dev, &table->cq_err_nb); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index 3df3604e8929..5acf60b1bbfe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -277,6 +277,7 @@ enum { static int esw_create_legacy_vepa_table(struct mlx5_eswitch *esw) { + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_core_dev *dev = esw->dev; struct mlx5_flow_namespace *root_ns; struct mlx5_flow_table *fdb; @@ -289,8 +290,10 @@ static int esw_create_legacy_vepa_table(struct mlx5_eswitch *esw) } /* num FTE 2, num FG 2 */ - fdb = mlx5_create_auto_grouped_flow_table(root_ns, LEGACY_VEPA_PRIO, - 2, 2, 0, 0); + ft_attr.prio = LEGACY_VEPA_PRIO; + ft_attr.max_fte = 2; + ft_attr.autogroup.max_num_groups = 2; + fdb = mlx5_create_auto_grouped_flow_table(root_ns, &ft_attr); if (IS_ERR(fdb)) { err = PTR_ERR(fdb); esw_warn(dev, "Failed to create VEPA FDB err %d\n", err); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h index ffcff3ba3701..4472710ccc9c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.h @@ -157,7 +157,7 @@ enum offloads_fdb_flags { ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED = BIT(0), }; -extern const unsigned int ESW_POOLS[4]; +struct mlx5_esw_chains_priv; struct mlx5_eswitch_fdb { union { @@ -182,14 +182,7 @@ struct mlx5_eswitch_fdb { struct mlx5_flow_handle *miss_rule_multi; int vlan_push_pop_refcount; - struct { - struct mlx5_flow_table *fdb; - u32 num_rules; - } fdb_prio[FDB_NUM_CHAINS][FDB_TC_MAX_PRIO + 1][FDB_TC_LEVELS_PER_PRIO]; - /* Protects fdb_prio table */ - struct mutex fdb_prio_lock; - - int fdb_left[ARRAY_SIZE(ESW_POOLS)]; + struct mlx5_esw_chains_priv *esw_chains_priv; } offloads; }; u32 flags; @@ -355,15 +348,6 @@ mlx5_eswitch_del_fwd_rule(struct mlx5_eswitch *esw, struct mlx5_flow_handle *rule, struct mlx5_esw_flow_attr *attr); -bool -mlx5_eswitch_prios_supported(struct mlx5_eswitch *esw); - -u16 -mlx5_eswitch_get_prio_range(struct mlx5_eswitch *esw); - -u32 -mlx5_eswitch_get_chain_range(struct mlx5_eswitch *esw); - struct mlx5_flow_handle * mlx5_eswitch_create_vport_rx_rule(struct mlx5_eswitch *esw, u16 vport, struct mlx5_flow_destination *dest); @@ -388,6 +372,11 @@ enum { MLX5_ESW_DEST_ENCAP_VALID = BIT(1), }; +enum { + MLX5_ESW_ATTR_FLAG_VLAN_HANDLED = BIT(0), + MLX5_ESW_ATTR_FLAG_SLOW_PATH = BIT(1), +}; + struct mlx5_esw_flow_attr { struct mlx5_eswitch_rep *in_rep; struct mlx5_core_dev *in_mdev; @@ -401,7 +390,6 @@ struct mlx5_esw_flow_attr { u16 vlan_vid[MLX5_FS_VLAN_DEPTH]; u8 vlan_prio[MLX5_FS_VLAN_DEPTH]; u8 total_vlan; - bool vlan_handled; struct { u32 flags; struct mlx5_eswitch_rep *rep; @@ -416,6 +404,7 @@ struct mlx5_esw_flow_attr { u32 chain; u16 prio; u32 dest_chain; + u32 flags; struct mlx5e_tc_flow_parse_attr *parse_attr; }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c index 3e6412783078..979f13bdc203 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads.c @@ -37,6 +37,7 @@ #include <linux/mlx5/fs.h> #include "mlx5_core.h" #include "eswitch.h" +#include "eswitch_offloads_chains.h" #include "rdma.h" #include "en.h" #include "fs_core.h" @@ -47,10 +48,6 @@ * one for multicast. */ #define MLX5_ESW_MISS_FLOWS (2) - -#define fdb_prio_table(esw, chain, prio, level) \ - (esw)->fdb_table.offloads.fdb_prio[(chain)][(prio)][(level)] - #define UPLINK_REP_INDEX 0 static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, @@ -62,32 +59,6 @@ static struct mlx5_eswitch_rep *mlx5_eswitch_get_rep(struct mlx5_eswitch *esw, return &esw->offloads.vport_reps[idx]; } -static struct mlx5_flow_table * -esw_get_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level); -static void -esw_put_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level); - -bool mlx5_eswitch_prios_supported(struct mlx5_eswitch *esw) -{ - return (!!(esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED)); -} - -u32 mlx5_eswitch_get_chain_range(struct mlx5_eswitch *esw) -{ - if (esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED) - return FDB_TC_MAX_CHAIN; - - return 0; -} - -u16 mlx5_eswitch_get_prio_range(struct mlx5_eswitch *esw) -{ - if (esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED) - return FDB_TC_MAX_PRIO; - - return 1; -} - static bool esw_check_ingress_prio_tag_enabled(const struct mlx5_eswitch *esw, const struct mlx5_vport *vport) @@ -175,10 +146,17 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, } if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { - if (attr->dest_chain) { - struct mlx5_flow_table *ft; + struct mlx5_flow_table *ft; - ft = esw_get_prio_table(esw, attr->dest_chain, 1, 0); + if (attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) { + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; + dest[i].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest[i].ft = mlx5_esw_chains_get_tc_end_ft(esw); + i++; + } else if (attr->dest_chain) { + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; + ft = mlx5_esw_chains_get_table(esw, attr->dest_chain, + 1, 0); if (IS_ERR(ft)) { rule = ERR_CAST(ft); goto err_create_goto_table; @@ -223,7 +201,8 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, if (flow_act.action & MLX5_FLOW_CONTEXT_ACTION_MOD_HDR) flow_act.modify_hdr = attr->modify_hdr; - fdb = esw_get_prio_table(esw, attr->chain, attr->prio, !!split); + fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, + !!split); if (IS_ERR(fdb)) { rule = ERR_CAST(fdb); goto err_esw_get; @@ -242,10 +221,10 @@ mlx5_eswitch_add_offloaded_rule(struct mlx5_eswitch *esw, return rule; err_add_rule: - esw_put_prio_table(esw, attr->chain, attr->prio, !!split); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, !!split); err_esw_get: - if (attr->dest_chain) - esw_put_prio_table(esw, attr->dest_chain, 1, 0); + if (!(attr->flags & MLX5_ESW_ATTR_FLAG_SLOW_PATH) && attr->dest_chain) + mlx5_esw_chains_put_table(esw, attr->dest_chain, 1, 0); err_create_goto_table: return rule; } @@ -262,13 +241,13 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, struct mlx5_flow_handle *rule; int i; - fast_fdb = esw_get_prio_table(esw, attr->chain, attr->prio, 0); + fast_fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, 0); if (IS_ERR(fast_fdb)) { rule = ERR_CAST(fast_fdb); goto err_get_fast; } - fwd_fdb = esw_get_prio_table(esw, attr->chain, attr->prio, 1); + fwd_fdb = mlx5_esw_chains_get_table(esw, attr->chain, attr->prio, 1); if (IS_ERR(fwd_fdb)) { rule = ERR_CAST(fwd_fdb); goto err_get_fwd; @@ -296,6 +275,7 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, if (attr->outer_match_level != MLX5_MATCH_NONE) spec->match_criteria_enable |= MLX5_MATCH_OUTER_HEADERS; + flow_act.flags |= FLOW_ACT_IGNORE_FLOW_LEVEL; rule = mlx5_add_flow_rules(fast_fdb, spec, &flow_act, dest, i); if (IS_ERR(rule)) @@ -305,9 +285,9 @@ mlx5_eswitch_add_fwd_rule(struct mlx5_eswitch *esw, return rule; add_err: - esw_put_prio_table(esw, attr->chain, attr->prio, 1); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 1); err_get_fwd: - esw_put_prio_table(esw, attr->chain, attr->prio, 0); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 0); err_get_fast: return rule; } @@ -332,12 +312,13 @@ __mlx5_eswitch_del_rule(struct mlx5_eswitch *esw, atomic64_dec(&esw->offloads.num_flows); if (fwd_rule) { - esw_put_prio_table(esw, attr->chain, attr->prio, 1); - esw_put_prio_table(esw, attr->chain, attr->prio, 0); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 1); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, 0); } else { - esw_put_prio_table(esw, attr->chain, attr->prio, !!split); + mlx5_esw_chains_put_table(esw, attr->chain, attr->prio, + !!split); if (attr->dest_chain) - esw_put_prio_table(esw, attr->dest_chain, 1, 0); + mlx5_esw_chains_put_table(esw, attr->dest_chain, 1, 0); } } @@ -451,7 +432,7 @@ int mlx5_eswitch_add_vlan_action(struct mlx5_eswitch *esw, if (err) goto unlock; - attr->vlan_handled = false; + attr->flags &= ~MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; vport = esw_vlan_action_get_vport(attr, push, pop); @@ -459,7 +440,7 @@ int mlx5_eswitch_add_vlan_action(struct mlx5_eswitch *esw, /* tracks VF --> wire rules without vlan push action */ if (attr->dests[0].rep->vport == MLX5_VPORT_UPLINK) { vport->vlan_refcount++; - attr->vlan_handled = true; + attr->flags |= MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; } goto unlock; @@ -490,7 +471,7 @@ skip_set_push: } out: if (!err) - attr->vlan_handled = true; + attr->flags |= MLX5_ESW_ATTR_FLAG_VLAN_HANDLED; unlock: mutex_unlock(&esw->state_lock); return err; @@ -508,7 +489,7 @@ int mlx5_eswitch_del_vlan_action(struct mlx5_eswitch *esw, if (mlx5_eswitch_vlan_actions_supported(esw->dev, 1)) return 0; - if (!attr->vlan_handled) + if (!(attr->flags & MLX5_ESW_ATTR_FLAG_VLAN_HANDLED)) return 0; push = !!(attr->action & MLX5_FLOW_CONTEXT_ACTION_VLAN_PUSH); @@ -582,8 +563,8 @@ mlx5_eswitch_add_send_to_vport_rule(struct mlx5_eswitch *esw, u16 vport, dest.vport.num = vport; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) esw_warn(esw->dev, "FDB: Failed to add send to vport rule err %ld\n", PTR_ERR(flow_rule)); out: @@ -824,8 +805,8 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw) dest.vport.num = esw->manager_vport; flow_act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) { err = PTR_ERR(flow_rule); esw_warn(esw->dev, "FDB: Failed to add unicast miss flow rule err %d\n", err); @@ -839,8 +820,8 @@ static int esw_add_fdb_miss_rule(struct mlx5_eswitch *esw) dmac_v = MLX5_ADDR_OF(fte_match_param, headers_v, outer_headers.dmac_47_16); dmac_v[0] = 0x01; - flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, spec, - &flow_act, &dest, 1); + flow_rule = mlx5_add_flow_rules(esw->fdb_table.offloads.slow_fdb, + spec, &flow_act, &dest, 1); if (IS_ERR(flow_rule)) { err = PTR_ERR(flow_rule); esw_warn(esw->dev, "FDB: Failed to add multicast miss flow rule err %d\n", err); @@ -855,174 +836,6 @@ out: return err; } -#define ESW_OFFLOADS_NUM_GROUPS 4 - -/* Firmware currently has 4 pool of 4 sizes that it supports (ESW_POOLS), - * and a virtual memory region of 16M (ESW_SIZE), this region is duplicated - * for each flow table pool. We can allocate up to 16M of each pool, - * and we keep track of how much we used via put/get_sz_to_pool. - * Firmware doesn't report any of this for now. - * ESW_POOL is expected to be sorted from large to small - */ -#define ESW_SIZE (16 * 1024 * 1024) -const unsigned int ESW_POOLS[4] = { 4 * 1024 * 1024, 1 * 1024 * 1024, - 64 * 1024, 128 }; - -static int -get_sz_from_pool(struct mlx5_eswitch *esw) -{ - int sz = 0, i; - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) { - if (esw->fdb_table.offloads.fdb_left[i]) { - --esw->fdb_table.offloads.fdb_left[i]; - sz = ESW_POOLS[i]; - break; - } - } - - return sz; -} - -static void -put_sz_to_pool(struct mlx5_eswitch *esw, int sz) -{ - int i; - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) { - if (sz >= ESW_POOLS[i]) { - ++esw->fdb_table.offloads.fdb_left[i]; - break; - } - } -} - -static struct mlx5_flow_table * -create_next_size_table(struct mlx5_eswitch *esw, - struct mlx5_flow_namespace *ns, - u16 table_prio, - int level, - u32 flags) -{ - struct mlx5_flow_table *fdb; - int sz; - - sz = get_sz_from_pool(esw); - if (!sz) - return ERR_PTR(-ENOSPC); - - fdb = mlx5_create_auto_grouped_flow_table(ns, - table_prio, - sz, - ESW_OFFLOADS_NUM_GROUPS, - level, - flags); - if (IS_ERR(fdb)) { - esw_warn(esw->dev, "Failed to create FDB Table err %d (table prio: %d, level: %d, size: %d)\n", - (int)PTR_ERR(fdb), table_prio, level, sz); - put_sz_to_pool(esw, sz); - } - - return fdb; -} - -static struct mlx5_flow_table * -esw_get_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level) -{ - struct mlx5_core_dev *dev = esw->dev; - struct mlx5_flow_table *fdb = NULL; - struct mlx5_flow_namespace *ns; - int table_prio, l = 0; - u32 flags = 0; - - if (chain == FDB_TC_SLOW_PATH_CHAIN) - return esw->fdb_table.offloads.slow_fdb; - - mutex_lock(&esw->fdb_table.offloads.fdb_prio_lock); - - fdb = fdb_prio_table(esw, chain, prio, level).fdb; - if (fdb) { - /* take ref on earlier levels as well */ - while (level >= 0) - fdb_prio_table(esw, chain, prio, level--).num_rules++; - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return fdb; - } - - ns = mlx5_get_fdb_sub_ns(dev, chain); - if (!ns) { - esw_warn(dev, "Failed to get FDB sub namespace\n"); - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return ERR_PTR(-EOPNOTSUPP); - } - - if (esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) - flags |= (MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT | - MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); - - table_prio = prio - 1; - - /* create earlier levels for correct fs_core lookup when - * connecting tables - */ - for (l = 0; l <= level; l++) { - if (fdb_prio_table(esw, chain, prio, l).fdb) { - fdb_prio_table(esw, chain, prio, l).num_rules++; - continue; - } - - fdb = create_next_size_table(esw, ns, table_prio, l, flags); - if (IS_ERR(fdb)) { - l--; - goto err_create_fdb; - } - - fdb_prio_table(esw, chain, prio, l).fdb = fdb; - fdb_prio_table(esw, chain, prio, l).num_rules = 1; - } - - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - return fdb; - -err_create_fdb: - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); - if (l >= 0) - esw_put_prio_table(esw, chain, prio, l); - - return fdb; -} - -static void -esw_put_prio_table(struct mlx5_eswitch *esw, u32 chain, u16 prio, int level) -{ - int l; - - if (chain == FDB_TC_SLOW_PATH_CHAIN) - return; - - mutex_lock(&esw->fdb_table.offloads.fdb_prio_lock); - - for (l = level; l >= 0; l--) { - if (--(fdb_prio_table(esw, chain, prio, l).num_rules) > 0) - continue; - - put_sz_to_pool(esw, fdb_prio_table(esw, chain, prio, l).fdb->max_fte); - mlx5_destroy_flow_table(fdb_prio_table(esw, chain, prio, l).fdb); - fdb_prio_table(esw, chain, prio, l).fdb = NULL; - } - - mutex_unlock(&esw->fdb_table.offloads.fdb_prio_lock); -} - -static void esw_destroy_offloads_fast_fdb_tables(struct mlx5_eswitch *esw) -{ - /* If lazy creation isn't supported, deref the fast path tables */ - if (!(esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED)) { - esw_put_prio_table(esw, 0, 1, 1); - esw_put_prio_table(esw, 0, 1, 0); - } -} - #define MAX_PF_SQ 256 #define MAX_SQ_NVPORTS 32 @@ -1055,16 +868,16 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_core_dev *dev = esw->dev; - u32 *flow_group_in, max_flow_counter; struct mlx5_flow_namespace *root_ns; struct mlx5_flow_table *fdb = NULL; - int table_size, ix, err = 0, i; + u32 flags = 0, *flow_group_in; + int table_size, ix, err = 0; struct mlx5_flow_group *g; - u32 flags = 0, fdb_max; void *match_criteria; u8 *dmac; esw_debug(esw->dev, "Create offloads FDB Tables\n"); + flow_group_in = kvzalloc(inlen, GFP_KERNEL); if (!flow_group_in) return -ENOMEM; @@ -1083,19 +896,6 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) goto ns_err; } - max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | - MLX5_CAP_GEN(dev, max_flow_counter_15_0); - fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size); - - esw_debug(dev, "Create offloads FDB table, min (max esw size(2^%d), max counters(%d), groups(%d), max flow table size(%d))\n", - MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size), - max_flow_counter, ESW_OFFLOADS_NUM_GROUPS, - fdb_max); - - for (i = 0; i < ARRAY_SIZE(ESW_POOLS); i++) - esw->fdb_table.offloads.fdb_left[i] = - ESW_POOLS[i] <= fdb_max ? ESW_SIZE / ESW_POOLS[i] : 0; - table_size = nvports * MAX_SQ_NVPORTS + MAX_PF_SQ + MLX5_ESW_MISS_FLOWS + esw->total_vports; @@ -1118,16 +918,10 @@ static int esw_create_offloads_fdb_tables(struct mlx5_eswitch *esw, int nvports) } esw->fdb_table.offloads.slow_fdb = fdb; - /* If lazy creation isn't supported, open the fast path tables now */ - if (!MLX5_CAP_ESW_FLOWTABLE(esw->dev, multi_fdb_encap) && - esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) { - esw->fdb_table.flags &= ~ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; - esw_warn(dev, "Lazy creation of flow tables isn't supported, ignoring priorities\n"); - esw_get_prio_table(esw, 0, 1, 0); - esw_get_prio_table(esw, 0, 1, 1); - } else { - esw_debug(dev, "Lazy creation of flow tables supported, deferring table opening\n"); - esw->fdb_table.flags |= ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + err = mlx5_esw_chains_create(esw); + if (err) { + esw_warn(dev, "Failed to create fdb chains err(%d)\n", err); + goto fdb_chains_err; } /* create send-to-vport group */ @@ -1218,7 +1012,8 @@ miss_err: peer_miss_err: mlx5_destroy_flow_group(esw->fdb_table.offloads.send_to_vport_grp); send_vport_err: - esw_destroy_offloads_fast_fdb_tables(esw); + mlx5_esw_chains_destroy(esw); +fdb_chains_err: mlx5_destroy_flow_table(esw->fdb_table.offloads.slow_fdb); slow_fdb_err: /* Holds true only as long as DMFS is the default */ @@ -1240,8 +1035,8 @@ static void esw_destroy_offloads_fdb_tables(struct mlx5_eswitch *esw) mlx5_destroy_flow_group(esw->fdb_table.offloads.peer_miss_grp); mlx5_destroy_flow_group(esw->fdb_table.offloads.miss_grp); + mlx5_esw_chains_destroy(esw); mlx5_destroy_flow_table(esw->fdb_table.offloads.slow_fdb); - esw_destroy_offloads_fast_fdb_tables(esw); /* Holds true only as long as DMFS is the default */ mlx5_flow_namespace_set_mode(esw->fdb_table.offloads.ns, MLX5_FLOW_STEERING_MODE_DMFS); @@ -2111,7 +1906,6 @@ static int esw_offloads_steering_init(struct mlx5_eswitch *esw) total_vports = num_vfs + MLX5_SPECIAL_VPORTS(esw->dev); memset(&esw->fdb_table.offloads, 0, sizeof(struct offloads_fdb)); - mutex_init(&esw->fdb_table.offloads.fdb_prio_lock); err = esw_create_uplink_offloads_acl_tables(esw); if (err) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c new file mode 100644 index 000000000000..c5a446e295aa --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.c @@ -0,0 +1,758 @@ +// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB +// Copyright (c) 2020 Mellanox Technologies. + +#include <linux/mlx5/driver.h> +#include <linux/mlx5/mlx5_ifc.h> +#include <linux/mlx5/fs.h> + +#include "eswitch_offloads_chains.h" +#include "mlx5_core.h" +#include "fs_core.h" +#include "eswitch.h" +#include "en.h" + +#define esw_chains_priv(esw) ((esw)->fdb_table.offloads.esw_chains_priv) +#define esw_chains_lock(esw) (esw_chains_priv(esw)->lock) +#define esw_chains_ht(esw) (esw_chains_priv(esw)->chains_ht) +#define esw_prios_ht(esw) (esw_chains_priv(esw)->prios_ht) +#define fdb_pool_left(esw) (esw_chains_priv(esw)->fdb_left) +#define tc_slow_fdb(esw) ((esw)->fdb_table.offloads.slow_fdb) +#define tc_end_fdb(esw) (esw_chains_priv(esw)->tc_end_fdb) +#define fdb_ignore_flow_level_supported(esw) \ + (MLX5_CAP_ESW_FLOWTABLE_FDB((esw)->dev, ignore_flow_level)) + +#define ESW_OFFLOADS_NUM_GROUPS 4 + +/* Firmware currently has 4 pool of 4 sizes that it supports (ESW_POOLS), + * and a virtual memory region of 16M (ESW_SIZE), this region is duplicated + * for each flow table pool. We can allocate up to 16M of each pool, + * and we keep track of how much we used via get_next_avail_sz_from_pool. + * Firmware doesn't report any of this for now. + * ESW_POOL is expected to be sorted from large to small and match firmware + * pools. + */ +#define ESW_SIZE (16 * 1024 * 1024) +static const unsigned int ESW_POOLS[] = { 4 * 1024 * 1024, + 1 * 1024 * 1024, + 64 * 1024, + 4 * 1024, }; + +struct mlx5_esw_chains_priv { + struct rhashtable chains_ht; + struct rhashtable prios_ht; + /* Protects above chains_ht and prios_ht */ + struct mutex lock; + + struct mlx5_flow_table *tc_end_fdb; + + int fdb_left[ARRAY_SIZE(ESW_POOLS)]; +}; + +struct fdb_chain { + struct rhash_head node; + + u32 chain; + + int ref; + + struct mlx5_eswitch *esw; + struct list_head prios_list; +}; + +struct fdb_prio_key { + u32 chain; + u32 prio; + u32 level; +}; + +struct fdb_prio { + struct rhash_head node; + struct list_head list; + + struct fdb_prio_key key; + + int ref; + + struct fdb_chain *fdb_chain; + struct mlx5_flow_table *fdb; + struct mlx5_flow_table *next_fdb; + struct mlx5_flow_group *miss_group; + struct mlx5_flow_handle *miss_rule; +}; + +static const struct rhashtable_params chain_params = { + .head_offset = offsetof(struct fdb_chain, node), + .key_offset = offsetof(struct fdb_chain, chain), + .key_len = sizeof_field(struct fdb_chain, chain), + .automatic_shrinking = true, +}; + +static const struct rhashtable_params prio_params = { + .head_offset = offsetof(struct fdb_prio, node), + .key_offset = offsetof(struct fdb_prio, key), + .key_len = sizeof_field(struct fdb_prio, key), + .automatic_shrinking = true, +}; + +bool mlx5_esw_chains_prios_supported(struct mlx5_eswitch *esw) +{ + return esw->fdb_table.flags & ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; +} + +u32 mlx5_esw_chains_get_chain_range(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + return 1; + + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX - 1; + + return FDB_TC_MAX_CHAIN; +} + +u32 mlx5_esw_chains_get_ft_chain(struct mlx5_eswitch *esw) +{ + return mlx5_esw_chains_get_chain_range(esw) + 1; +} + +u32 mlx5_esw_chains_get_prio_range(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + return 1; + + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX; + + return FDB_TC_MAX_PRIO; +} + +static unsigned int mlx5_esw_chains_get_level_range(struct mlx5_eswitch *esw) +{ + if (fdb_ignore_flow_level_supported(esw)) + return UINT_MAX; + + return FDB_TC_LEVELS_PER_PRIO; +} + +#define POOL_NEXT_SIZE 0 +static int +mlx5_esw_chains_get_avail_sz_from_pool(struct mlx5_eswitch *esw, + int desired_size) +{ + int i, found_i = -1; + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) { + if (fdb_pool_left(esw)[i] && ESW_POOLS[i] > desired_size) { + found_i = i; + if (desired_size != POOL_NEXT_SIZE) + break; + } + } + + if (found_i != -1) { + --fdb_pool_left(esw)[found_i]; + return ESW_POOLS[found_i]; + } + + return 0; +} + +static void +mlx5_esw_chains_put_sz_to_pool(struct mlx5_eswitch *esw, int sz) +{ + int i; + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) { + if (sz == ESW_POOLS[i]) { + ++fdb_pool_left(esw)[i]; + return; + } + } + + WARN_ONCE(1, "Couldn't find size %d in fdb size pool", sz); +} + +static void +mlx5_esw_chains_init_sz_pool(struct mlx5_eswitch *esw) +{ + u32 fdb_max; + int i; + + fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(esw->dev, log_max_ft_size); + + for (i = ARRAY_SIZE(ESW_POOLS) - 1; i >= 0; i--) + fdb_pool_left(esw)[i] = + ESW_POOLS[i] <= fdb_max ? ESW_SIZE / ESW_POOLS[i] : 0; +} + +static struct mlx5_flow_table * +mlx5_esw_chains_create_fdb_table(struct mlx5_eswitch *esw, + u32 chain, u32 prio, u32 level) +{ + struct mlx5_flow_table_attr ft_attr = {}; + struct mlx5_flow_namespace *ns; + struct mlx5_flow_table *fdb; + int sz; + + if (esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) + ft_attr.flags |= (MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT | + MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); + + sz = mlx5_esw_chains_get_avail_sz_from_pool(esw, POOL_NEXT_SIZE); + if (!sz) + return ERR_PTR(-ENOSPC); + ft_attr.max_fte = sz; + + /* We use tc_slow_fdb(esw) as the table's next_ft till + * ignore_flow_level is allowed on FT creation and not just for FTEs. + * Instead caller should add an explicit miss rule if needed. + */ + ft_attr.next_ft = tc_slow_fdb(esw); + + /* The root table(chain 0, prio 1, level 0) is required to be + * connected to the previous prio (FDB_BYPASS_PATH if exists). + * We always create it, as a managed table, in order to align with + * fs_core logic. + */ + if (!fdb_ignore_flow_level_supported(esw) || + (chain == 0 && prio == 1 && level == 0)) { + ft_attr.level = level; + ft_attr.prio = prio - 1; + ns = mlx5_get_fdb_sub_ns(esw->dev, chain); + } else { + ft_attr.flags |= MLX5_FLOW_TABLE_UNMANAGED; + ft_attr.prio = FDB_TC_OFFLOAD; + /* Firmware doesn't allow us to create another level 0 table, + * so we create all unmanaged tables as level 1. + * + * To connect them, we use explicit miss rules with + * ignore_flow_level. Caller is responsible to create + * these rules (if needed). + */ + ft_attr.level = 1; + ns = mlx5_get_flow_namespace(esw->dev, MLX5_FLOW_NAMESPACE_FDB); + } + + ft_attr.autogroup.num_reserved_entries = 2; + ft_attr.autogroup.max_num_groups = ESW_OFFLOADS_NUM_GROUPS; + fdb = mlx5_create_auto_grouped_flow_table(ns, &ft_attr); + if (IS_ERR(fdb)) { + esw_warn(esw->dev, + "Failed to create FDB table err %d (chain: %d, prio: %d, level: %d, size: %d)\n", + (int)PTR_ERR(fdb), chain, prio, level, sz); + mlx5_esw_chains_put_sz_to_pool(esw, sz); + return fdb; + } + + return fdb; +} + +static void +mlx5_esw_chains_destroy_fdb_table(struct mlx5_eswitch *esw, + struct mlx5_flow_table *fdb) +{ + mlx5_esw_chains_put_sz_to_pool(esw, fdb->max_fte); + mlx5_destroy_flow_table(fdb); +} + +static struct fdb_chain * +mlx5_esw_chains_create_fdb_chain(struct mlx5_eswitch *esw, u32 chain) +{ + struct fdb_chain *fdb_chain = NULL; + int err; + + fdb_chain = kvzalloc(sizeof(*fdb_chain), GFP_KERNEL); + if (!fdb_chain) + return ERR_PTR(-ENOMEM); + + fdb_chain->esw = esw; + fdb_chain->chain = chain; + INIT_LIST_HEAD(&fdb_chain->prios_list); + + err = rhashtable_insert_fast(&esw_chains_ht(esw), &fdb_chain->node, + chain_params); + if (err) + goto err_insert; + + return fdb_chain; + +err_insert: + kvfree(fdb_chain); + return ERR_PTR(err); +} + +static void +mlx5_esw_chains_destroy_fdb_chain(struct fdb_chain *fdb_chain) +{ + struct mlx5_eswitch *esw = fdb_chain->esw; + + rhashtable_remove_fast(&esw_chains_ht(esw), &fdb_chain->node, + chain_params); + kvfree(fdb_chain); +} + +static struct fdb_chain * +mlx5_esw_chains_get_fdb_chain(struct mlx5_eswitch *esw, u32 chain) +{ + struct fdb_chain *fdb_chain; + + fdb_chain = rhashtable_lookup_fast(&esw_chains_ht(esw), &chain, + chain_params); + if (!fdb_chain) { + fdb_chain = mlx5_esw_chains_create_fdb_chain(esw, chain); + if (IS_ERR(fdb_chain)) + return fdb_chain; + } + + fdb_chain->ref++; + + return fdb_chain; +} + +static struct mlx5_flow_handle * +mlx5_esw_chains_add_miss_rule(struct mlx5_flow_table *fdb, + struct mlx5_flow_table *next_fdb) +{ + static const struct mlx5_flow_spec spec = {}; + struct mlx5_flow_destination dest = {}; + struct mlx5_flow_act act = {}; + + act.flags = FLOW_ACT_IGNORE_FLOW_LEVEL | FLOW_ACT_NO_APPEND; + act.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; + dest.type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + dest.ft = next_fdb; + + return mlx5_add_flow_rules(fdb, &spec, &act, &dest, 1); +} + +static int +mlx5_esw_chains_update_prio_prevs(struct fdb_prio *fdb_prio, + struct mlx5_flow_table *next_fdb) +{ + struct mlx5_flow_handle *miss_rules[FDB_TC_LEVELS_PER_PRIO + 1] = {}; + struct fdb_chain *fdb_chain = fdb_prio->fdb_chain; + struct fdb_prio *pos; + int n = 0, err; + + if (fdb_prio->key.level) + return 0; + + /* Iterate in reverse order until reaching the level 0 rule of + * the previous priority, adding all the miss rules first, so we can + * revert them if any of them fails. + */ + pos = fdb_prio; + list_for_each_entry_continue_reverse(pos, + &fdb_chain->prios_list, + list) { + miss_rules[n] = mlx5_esw_chains_add_miss_rule(pos->fdb, + next_fdb); + if (IS_ERR(miss_rules[n])) { + err = PTR_ERR(miss_rules[n]); + goto err_prev_rule; + } + + n++; + if (!pos->key.level) + break; + } + + /* Success, delete old miss rules, and update the pointers. */ + n = 0; + pos = fdb_prio; + list_for_each_entry_continue_reverse(pos, + &fdb_chain->prios_list, + list) { + mlx5_del_flow_rules(pos->miss_rule); + + pos->miss_rule = miss_rules[n]; + pos->next_fdb = next_fdb; + + n++; + if (!pos->key.level) + break; + } + + return 0; + +err_prev_rule: + while (--n >= 0) + mlx5_del_flow_rules(miss_rules[n]); + + return err; +} + +static void +mlx5_esw_chains_put_fdb_chain(struct fdb_chain *fdb_chain) +{ + if (--fdb_chain->ref == 0) + mlx5_esw_chains_destroy_fdb_chain(fdb_chain); +} + +static struct fdb_prio * +mlx5_esw_chains_create_fdb_prio(struct mlx5_eswitch *esw, + u32 chain, u32 prio, u32 level) +{ + int inlen = MLX5_ST_SZ_BYTES(create_flow_group_in); + struct mlx5_flow_handle *miss_rule = NULL; + struct mlx5_flow_group *miss_group; + struct fdb_prio *fdb_prio = NULL; + struct mlx5_flow_table *next_fdb; + struct fdb_chain *fdb_chain; + struct mlx5_flow_table *fdb; + struct list_head *pos; + u32 *flow_group_in; + int err; + + fdb_chain = mlx5_esw_chains_get_fdb_chain(esw, chain); + if (IS_ERR(fdb_chain)) + return ERR_CAST(fdb_chain); + + fdb_prio = kvzalloc(sizeof(*fdb_prio), GFP_KERNEL); + flow_group_in = kvzalloc(inlen, GFP_KERNEL); + if (!fdb_prio || !flow_group_in) { + err = -ENOMEM; + goto err_alloc; + } + + /* Chain's prio list is sorted by prio and level. + * And all levels of some prio point to the next prio's level 0. + * Example list (prio, level): + * (3,0)->(3,1)->(5,0)->(5,1)->(6,1)->(7,0) + * In hardware, we will we have the following pointers: + * (3,0) -> (5,0) -> (7,0) -> Slow path + * (3,1) -> (5,0) + * (5,1) -> (7,0) + * (6,1) -> (7,0) + */ + + /* Default miss for each chain: */ + next_fdb = (chain == mlx5_esw_chains_get_ft_chain(esw)) ? + tc_slow_fdb(esw) : + tc_end_fdb(esw); + list_for_each(pos, &fdb_chain->prios_list) { + struct fdb_prio *p = list_entry(pos, struct fdb_prio, list); + + /* exit on first pos that is larger */ + if (prio < p->key.prio || (prio == p->key.prio && + level < p->key.level)) { + /* Get next level 0 table */ + next_fdb = p->key.level == 0 ? p->fdb : p->next_fdb; + break; + } + } + + fdb = mlx5_esw_chains_create_fdb_table(esw, chain, prio, level); + if (IS_ERR(fdb)) { + err = PTR_ERR(fdb); + goto err_create; + } + + MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, + fdb->max_fte - 2); + MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, + fdb->max_fte - 1); + miss_group = mlx5_create_flow_group(fdb, flow_group_in); + if (IS_ERR(miss_group)) { + err = PTR_ERR(miss_group); + goto err_group; + } + + /* Add miss rule to next_fdb */ + miss_rule = mlx5_esw_chains_add_miss_rule(fdb, next_fdb); + if (IS_ERR(miss_rule)) { + err = PTR_ERR(miss_rule); + goto err_miss_rule; + } + + fdb_prio->miss_group = miss_group; + fdb_prio->miss_rule = miss_rule; + fdb_prio->next_fdb = next_fdb; + fdb_prio->fdb_chain = fdb_chain; + fdb_prio->key.chain = chain; + fdb_prio->key.prio = prio; + fdb_prio->key.level = level; + fdb_prio->fdb = fdb; + + err = rhashtable_insert_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); + if (err) + goto err_insert; + + list_add(&fdb_prio->list, pos->prev); + + /* Table is ready, connect it */ + err = mlx5_esw_chains_update_prio_prevs(fdb_prio, fdb); + if (err) + goto err_update; + + kvfree(flow_group_in); + return fdb_prio; + +err_update: + list_del(&fdb_prio->list); + rhashtable_remove_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); +err_insert: + mlx5_del_flow_rules(miss_rule); +err_miss_rule: + mlx5_destroy_flow_group(miss_group); +err_group: + mlx5_esw_chains_destroy_fdb_table(esw, fdb); +err_create: +err_alloc: + kvfree(fdb_prio); + kvfree(flow_group_in); + mlx5_esw_chains_put_fdb_chain(fdb_chain); + return ERR_PTR(err); +} + +static void +mlx5_esw_chains_destroy_fdb_prio(struct mlx5_eswitch *esw, + struct fdb_prio *fdb_prio) +{ + struct fdb_chain *fdb_chain = fdb_prio->fdb_chain; + + WARN_ON(mlx5_esw_chains_update_prio_prevs(fdb_prio, + fdb_prio->next_fdb)); + + list_del(&fdb_prio->list); + rhashtable_remove_fast(&esw_prios_ht(esw), &fdb_prio->node, + prio_params); + mlx5_del_flow_rules(fdb_prio->miss_rule); + mlx5_destroy_flow_group(fdb_prio->miss_group); + mlx5_esw_chains_destroy_fdb_table(esw, fdb_prio->fdb); + mlx5_esw_chains_put_fdb_chain(fdb_chain); + kvfree(fdb_prio); +} + +struct mlx5_flow_table * +mlx5_esw_chains_get_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level) +{ + struct mlx5_flow_table *prev_fts; + struct fdb_prio *fdb_prio; + struct fdb_prio_key key; + int l = 0; + + if ((chain > mlx5_esw_chains_get_chain_range(esw) && + chain != mlx5_esw_chains_get_ft_chain(esw)) || + prio > mlx5_esw_chains_get_prio_range(esw) || + level > mlx5_esw_chains_get_level_range(esw)) + return ERR_PTR(-EOPNOTSUPP); + + /* create earlier levels for correct fs_core lookup when + * connecting tables. + */ + for (l = 0; l < level; l++) { + prev_fts = mlx5_esw_chains_get_table(esw, chain, prio, l); + if (IS_ERR(prev_fts)) { + fdb_prio = ERR_CAST(prev_fts); + goto err_get_prevs; + } + } + + key.chain = chain; + key.prio = prio; + key.level = level; + + mutex_lock(&esw_chains_lock(esw)); + fdb_prio = rhashtable_lookup_fast(&esw_prios_ht(esw), &key, + prio_params); + if (!fdb_prio) { + fdb_prio = mlx5_esw_chains_create_fdb_prio(esw, chain, + prio, level); + if (IS_ERR(fdb_prio)) + goto err_create_prio; + } + + ++fdb_prio->ref; + mutex_unlock(&esw_chains_lock(esw)); + + return fdb_prio->fdb; + +err_create_prio: + mutex_unlock(&esw_chains_lock(esw)); +err_get_prevs: + while (--l >= 0) + mlx5_esw_chains_put_table(esw, chain, prio, l); + return ERR_CAST(fdb_prio); +} + +void +mlx5_esw_chains_put_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level) +{ + struct fdb_prio *fdb_prio; + struct fdb_prio_key key; + + key.chain = chain; + key.prio = prio; + key.level = level; + + mutex_lock(&esw_chains_lock(esw)); + fdb_prio = rhashtable_lookup_fast(&esw_prios_ht(esw), &key, + prio_params); + if (!fdb_prio) + goto err_get_prio; + + if (--fdb_prio->ref == 0) + mlx5_esw_chains_destroy_fdb_prio(esw, fdb_prio); + mutex_unlock(&esw_chains_lock(esw)); + + while (level-- > 0) + mlx5_esw_chains_put_table(esw, chain, prio, level); + + return; + +err_get_prio: + mutex_unlock(&esw_chains_lock(esw)); + WARN_ONCE(1, + "Couldn't find table: (chain: %d prio: %d level: %d)", + chain, prio, level); +} + +struct mlx5_flow_table * +mlx5_esw_chains_get_tc_end_ft(struct mlx5_eswitch *esw) +{ + return tc_end_fdb(esw); +} + +static int +mlx5_esw_chains_init(struct mlx5_eswitch *esw) +{ + struct mlx5_esw_chains_priv *chains_priv; + struct mlx5_core_dev *dev = esw->dev; + u32 max_flow_counter, fdb_max; + int err; + + chains_priv = kzalloc(sizeof(*chains_priv), GFP_KERNEL); + if (!chains_priv) + return -ENOMEM; + esw_chains_priv(esw) = chains_priv; + + max_flow_counter = (MLX5_CAP_GEN(dev, max_flow_counter_31_16) << 16) | + MLX5_CAP_GEN(dev, max_flow_counter_15_0); + fdb_max = 1 << MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size); + + esw_debug(dev, + "Init esw offloads chains, max counters(%d), groups(%d), max flow table size(%d)\n", + max_flow_counter, ESW_OFFLOADS_NUM_GROUPS, fdb_max); + + mlx5_esw_chains_init_sz_pool(esw); + + if (!MLX5_CAP_ESW_FLOWTABLE(esw->dev, multi_fdb_encap) && + esw->offloads.encap != DEVLINK_ESWITCH_ENCAP_MODE_NONE) { + esw->fdb_table.flags &= ~ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + esw_warn(dev, "Tc chains and priorities offload aren't supported, update firmware if needed\n"); + } else { + esw->fdb_table.flags |= ESW_FDB_CHAINS_AND_PRIOS_SUPPORTED; + esw_info(dev, "Supported tc offload range - chains: %u, prios: %u\n", + mlx5_esw_chains_get_chain_range(esw), + mlx5_esw_chains_get_prio_range(esw)); + } + + err = rhashtable_init(&esw_chains_ht(esw), &chain_params); + if (err) + goto init_chains_ht_err; + + err = rhashtable_init(&esw_prios_ht(esw), &prio_params); + if (err) + goto init_prios_ht_err; + + mutex_init(&esw_chains_lock(esw)); + + return 0; + +init_prios_ht_err: + rhashtable_destroy(&esw_chains_ht(esw)); +init_chains_ht_err: + kfree(chains_priv); + return err; +} + +static void +mlx5_esw_chains_cleanup(struct mlx5_eswitch *esw) +{ + mutex_destroy(&esw_chains_lock(esw)); + rhashtable_destroy(&esw_prios_ht(esw)); + rhashtable_destroy(&esw_chains_ht(esw)); + + kfree(esw_chains_priv(esw)); +} + +static int +mlx5_esw_chains_open(struct mlx5_eswitch *esw) +{ + struct mlx5_flow_table *ft; + int err; + + /* Create tc_end_fdb(esw) which is the always created ft chain */ + ft = mlx5_esw_chains_get_table(esw, mlx5_esw_chains_get_ft_chain(esw), + 1, 0); + if (IS_ERR(ft)) + return PTR_ERR(ft); + + tc_end_fdb(esw) = ft; + + /* Always open the root for fast path */ + ft = mlx5_esw_chains_get_table(esw, 0, 1, 0); + if (IS_ERR(ft)) { + err = PTR_ERR(ft); + goto level_0_err; + } + + /* Open level 1 for split rules now if prios isn't supported */ + if (!mlx5_esw_chains_prios_supported(esw)) { + ft = mlx5_esw_chains_get_table(esw, 0, 1, 1); + + if (IS_ERR(ft)) { + err = PTR_ERR(ft); + goto level_1_err; + } + } + + return 0; + +level_1_err: + mlx5_esw_chains_put_table(esw, 0, 1, 0); +level_0_err: + mlx5_esw_chains_put_table(esw, mlx5_esw_chains_get_ft_chain(esw), 1, 0); + return err; +} + +static void +mlx5_esw_chains_close(struct mlx5_eswitch *esw) +{ + if (!mlx5_esw_chains_prios_supported(esw)) + mlx5_esw_chains_put_table(esw, 0, 1, 1); + mlx5_esw_chains_put_table(esw, 0, 1, 0); + mlx5_esw_chains_put_table(esw, mlx5_esw_chains_get_ft_chain(esw), 1, 0); +} + +int +mlx5_esw_chains_create(struct mlx5_eswitch *esw) +{ + int err; + + err = mlx5_esw_chains_init(esw); + if (err) + return err; + + err = mlx5_esw_chains_open(esw); + if (err) + goto err_open; + + return 0; + +err_open: + mlx5_esw_chains_cleanup(esw); + return err; +} + +void +mlx5_esw_chains_destroy(struct mlx5_eswitch *esw) +{ + mlx5_esw_chains_close(esw); + mlx5_esw_chains_cleanup(esw); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h new file mode 100644 index 000000000000..2e13097fe348 --- /dev/null +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_chains.h @@ -0,0 +1,30 @@ +/* SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB */ +/* Copyright (c) 2020 Mellanox Technologies. */ + +#ifndef __ML5_ESW_CHAINS_H__ +#define __ML5_ESW_CHAINS_H__ + +bool +mlx5_esw_chains_prios_supported(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_prio_range(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_chain_range(struct mlx5_eswitch *esw); +u32 +mlx5_esw_chains_get_ft_chain(struct mlx5_eswitch *esw); + +struct mlx5_flow_table * +mlx5_esw_chains_get_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level); +void +mlx5_esw_chains_put_table(struct mlx5_eswitch *esw, u32 chain, u32 prio, + u32 level); + +struct mlx5_flow_table * +mlx5_esw_chains_get_tc_end_ft(struct mlx5_eswitch *esw); + +int mlx5_esw_chains_create(struct mlx5_eswitch *esw); +void mlx5_esw_chains_destroy(struct mlx5_eswitch *esw); + +#endif /* __ML5_ESW_CHAINS_H__ */ + diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c index 366bda1bb1c3..dc08ed9339ab 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch_offloads_termtbl.c @@ -50,8 +50,8 @@ mlx5_eswitch_termtbl_create(struct mlx5_core_dev *dev, struct mlx5_flow_act *flow_act) { static const struct mlx5_flow_spec spec = {}; + struct mlx5_flow_table_attr ft_attr = {}; struct mlx5_flow_namespace *root_ns; - int prio, flags; int err; root_ns = mlx5_get_flow_namespace(dev, MLX5_FLOW_NAMESPACE_FDB); @@ -63,10 +63,11 @@ mlx5_eswitch_termtbl_create(struct mlx5_core_dev *dev, /* As this is the terminating action then the termination table is the * same prio as the slow path */ - prio = FDB_SLOW_PATH; - flags = MLX5_FLOW_TABLE_TERMINATION; - tt->termtbl = mlx5_create_auto_grouped_flow_table(root_ns, prio, 1, 1, - 0, flags); + ft_attr.flags = MLX5_FLOW_TABLE_TERMINATION; + ft_attr.prio = FDB_SLOW_PATH; + ft_attr.max_fte = 1; + ft_attr.autogroup.max_num_groups = 1; + tt->termtbl = mlx5_create_auto_grouped_flow_table(root_ns, &ft_attr); if (IS_ERR(tt->termtbl)) { esw_warn(dev, "Failed to create termination table\n"); return -EOPNOTSUPP; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c index 3c816e81f8d9..b25465d9e030 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_cmd.c @@ -432,6 +432,9 @@ static int mlx5_cmd_set_fte(struct mlx5_core_dev *dev, MLX5_SET(set_fte_in, in, table_type, ft->type); MLX5_SET(set_fte_in, in, table_id, ft->id); MLX5_SET(set_fte_in, in, flow_index, fte->index); + MLX5_SET(set_fte_in, in, ignore_flow_level, + !!(fte->action.flags & FLOW_ACT_IGNORE_FLOW_LEVEL)); + if (ft->vport) { MLX5_SET(set_fte_in, in, vport_number, ft->vport); MLX5_SET(set_fte_in, in, other_vport, 1); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 8c5df6c7d7b6..c7a16ae05fa8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -579,7 +579,9 @@ static void del_sw_flow_group(struct fs_node *node) rhashtable_destroy(&fg->ftes_hash); ida_destroy(&fg->fte_allocator); - if (ft->autogroup.active && fg->max_ftes == ft->autogroup.group_size) + if (ft->autogroup.active && + fg->max_ftes == ft->autogroup.group_size && + fg->start_index < ft->autogroup.max_fte) ft->autogroup.num_groups--; err = rhltable_remove(&ft->fgs_hash, &fg->hash, @@ -1006,7 +1008,8 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa u16 vport) { struct mlx5_flow_root_namespace *root = find_root(&ns->node); - struct mlx5_flow_table *next_ft = NULL; + bool unmanaged = ft_attr->flags & MLX5_FLOW_TABLE_UNMANAGED; + struct mlx5_flow_table *next_ft; struct fs_prio *fs_prio = NULL; struct mlx5_flow_table *ft; int log_table_sz; @@ -1023,14 +1026,21 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa err = -EINVAL; goto unlock_root; } - if (ft_attr->level >= fs_prio->num_levels) { - err = -ENOSPC; - goto unlock_root; + if (!unmanaged) { + /* The level is related to the + * priority level range. + */ + if (ft_attr->level >= fs_prio->num_levels) { + err = -ENOSPC; + goto unlock_root; + } + + ft_attr->level += fs_prio->start_level; } + /* The level is related to the * priority level range. */ - ft_attr->level += fs_prio->start_level; ft = alloc_flow_table(ft_attr->level, vport, ft_attr->max_fte ? roundup_pow_of_two(ft_attr->max_fte) : 0, @@ -1043,19 +1053,27 @@ static struct mlx5_flow_table *__mlx5_create_flow_table(struct mlx5_flow_namespa tree_init_node(&ft->node, del_hw_flow_table, del_sw_flow_table); log_table_sz = ft->max_fte ? ilog2(ft->max_fte) : 0; - next_ft = find_next_chained_ft(fs_prio); + next_ft = unmanaged ? ft_attr->next_ft : + find_next_chained_ft(fs_prio); ft->def_miss_action = ns->def_miss_action; err = root->cmds->create_flow_table(root, ft, log_table_sz, next_ft); if (err) goto free_ft; - err = connect_flow_table(root->dev, ft, fs_prio); - if (err) - goto destroy_ft; + if (!unmanaged) { + err = connect_flow_table(root->dev, ft, fs_prio); + if (err) + goto destroy_ft; + } + ft->node.active = true; down_write_ref_node(&fs_prio->node, false); - tree_add_node(&ft->node, &fs_prio->node); - list_add_flow_table(ft, fs_prio); + if (!unmanaged) { + tree_add_node(&ft->node, &fs_prio->node); + list_add_flow_table(ft, fs_prio); + } else { + ft->node.root = fs_prio->node.root; + } fs_prio->num_ft++; up_write_ref_node(&fs_prio->node, false); mutex_unlock(&root->chain_lock); @@ -1103,31 +1121,27 @@ EXPORT_SYMBOL(mlx5_create_lag_demux_flow_table); struct mlx5_flow_table* mlx5_create_auto_grouped_flow_table(struct mlx5_flow_namespace *ns, - int prio, - int num_flow_table_entries, - int max_num_groups, - u32 level, - u32 flags) + struct mlx5_flow_table_attr *ft_attr) { - struct mlx5_flow_table_attr ft_attr = {}; + int num_reserved_entries = ft_attr->autogroup.num_reserved_entries; + int autogroups_max_fte = ft_attr->max_fte - num_reserved_entries; + int max_num_groups = ft_attr->autogroup.max_num_groups; struct mlx5_flow_table *ft; - if (max_num_groups > num_flow_table_entries) + if (max_num_groups > autogroups_max_fte) + return ERR_PTR(-EINVAL); + if (num_reserved_entries > ft_attr->max_fte) return ERR_PTR(-EINVAL); - ft_attr.max_fte = num_flow_table_entries; - ft_attr.prio = prio; - ft_attr.level = level; - ft_attr.flags = flags; - - ft = mlx5_create_flow_table(ns, &ft_attr); + ft = mlx5_create_flow_table(ns, ft_attr); if (IS_ERR(ft)) return ft; ft->autogroup.active = true; ft->autogroup.required_groups = max_num_groups; + ft->autogroup.max_fte = autogroups_max_fte; /* We save place for flow groups in addition to max types */ - ft->autogroup.group_size = ft->max_fte / (max_num_groups + 1); + ft->autogroup.group_size = autogroups_max_fte / (max_num_groups + 1); return ft; } @@ -1149,7 +1163,7 @@ struct mlx5_flow_group *mlx5_create_flow_group(struct mlx5_flow_table *ft, struct mlx5_flow_group *fg; int err; - if (ft->autogroup.active) + if (ft->autogroup.active && start_index < ft->autogroup.max_fte) return ERR_PTR(-EPERM); down_write_ref_node(&ft->node, false); @@ -1322,9 +1336,10 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft const struct mlx5_flow_spec *spec) { struct list_head *prev = &ft->node.children; - struct mlx5_flow_group *fg; + u32 max_fte = ft->autogroup.max_fte; unsigned int candidate_index = 0; unsigned int group_size = 0; + struct mlx5_flow_group *fg; if (!ft->autogroup.active) return ERR_PTR(-ENOENT); @@ -1332,7 +1347,7 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft if (ft->autogroup.num_groups < ft->autogroup.required_groups) group_size = ft->autogroup.group_size; - /* ft->max_fte == ft->autogroup.max_types */ + /* max_fte == ft->autogroup.max_types */ if (group_size == 0) group_size = 1; @@ -1345,7 +1360,7 @@ static struct mlx5_flow_group *alloc_auto_flow_group(struct mlx5_flow_table *ft prev = &fg->node.list; } - if (candidate_index + group_size > ft->max_fte) + if (candidate_index + group_size > max_fte) return ERR_PTR(-ENOSPC); fg = alloc_insert_flow_group(ft, @@ -1529,18 +1544,30 @@ static bool counter_is_valid(u32 action) } static bool dest_is_valid(struct mlx5_flow_destination *dest, - u32 action, + struct mlx5_flow_act *flow_act, struct mlx5_flow_table *ft) { + bool ignore_level = flow_act->flags & FLOW_ACT_IGNORE_FLOW_LEVEL; + u32 action = flow_act->action; + if (dest && (dest->type == MLX5_FLOW_DESTINATION_TYPE_COUNTER)) return counter_is_valid(action); if (!(action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST)) return true; + if (ignore_level) { + if (ft->type != FS_FT_FDB) + return false; + + if (dest->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE && + dest->ft->type != FS_FT_FDB) + return false; + } + if (!dest || ((dest->type == MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE) && - (dest->ft->level <= ft->level))) + (dest->ft->level <= ft->level && !ignore_level))) return false; return true; } @@ -1770,7 +1797,7 @@ _mlx5_add_flow_rules(struct mlx5_flow_table *ft, return ERR_PTR(-EINVAL); for (i = 0; i < dest_num; i++) { - if (!dest_is_valid(&dest[i], flow_act->action, ft)) + if (!dest_is_valid(&dest[i], flow_act, ft)) return ERR_PTR(-EINVAL); } nested_down_read_ref_node(&ft->node, FS_LOCK_GRANDPARENT); @@ -2033,7 +2060,8 @@ int mlx5_destroy_flow_table(struct mlx5_flow_table *ft) int err = 0; mutex_lock(&root->chain_lock); - err = disconnect_flow_table(ft); + if (!(ft->flags & MLX5_FLOW_TABLE_UNMANAGED)) + err = disconnect_flow_table(ft); if (err) { mutex_unlock(&root->chain_lock); return err; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h index c2621b911563..be5f5e32c1e8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.h @@ -164,6 +164,7 @@ struct mlx5_flow_table { unsigned int required_groups; unsigned int group_size; unsigned int num_groups; + unsigned int max_fte; } autogroup; /* Protect fwd_rules */ struct mutex lock; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c index a19790dee7b2..d89ff1d09119 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c @@ -131,11 +131,11 @@ static int mlx5_get_pcam_reg(struct mlx5_core_dev *dev) MLX5_PCAM_REGS_5000_TO_507F); } -static int mlx5_get_mcam_reg(struct mlx5_core_dev *dev) +static int mlx5_get_mcam_access_reg_group(struct mlx5_core_dev *dev, + enum mlx5_mcam_reg_groups group) { - return mlx5_query_mcam_reg(dev, dev->caps.mcam, - MLX5_MCAM_FEATURE_ENHANCED_FEATURES, - MLX5_MCAM_REGS_FIRST_128); + return mlx5_query_mcam_reg(dev, dev->caps.mcam[group], + MLX5_MCAM_FEATURE_ENHANCED_FEATURES, group); } static int mlx5_get_qcam_reg(struct mlx5_core_dev *dev) @@ -221,8 +221,11 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) if (MLX5_CAP_GEN(dev, pcam_reg)) mlx5_get_pcam_reg(dev); - if (MLX5_CAP_GEN(dev, mcam_reg)) - mlx5_get_mcam_reg(dev); + if (MLX5_CAP_GEN(dev, mcam_reg)) { + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_FIRST_128); + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_0x9080_0x90FF); + mlx5_get_mcam_access_reg_group(dev, MLX5_MCAM_REGS_0x9100_0x917F); + } if (MLX5_CAP_GEN(dev, qcam_reg)) mlx5_get_qcam_reg(dev); @@ -245,6 +248,13 @@ int mlx5_query_hca_caps(struct mlx5_core_dev *dev) return err; } + if (MLX5_CAP_GEN_64(dev, general_obj_types) & + MLX5_GENERAL_OBJ_TYPES_CAP_VIRTIO_NET_Q) { + err = mlx5_core_get_caps(dev, MLX5_CAP_VDPA_EMULATION); + if (err) + return err; + } + return 0; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c index 3ed8ab2d703d..56078b23f1a0 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.c @@ -87,8 +87,8 @@ int mlx5i_init(struct mlx5_core_dev *mdev, mlx5e_set_netdev_mtu_boundaries(priv); netdev->mtu = netdev->max_mtu; - mlx5e_build_nic_params(mdev, NULL, &priv->rss_params, &priv->channels.params, - priv->max_nch, netdev->mtu); + mlx5e_build_nic_params(priv, NULL, &priv->rss_params, &priv->channels.params, + netdev->mtu); mlx5i_build_nic_params(mdev, &priv->channels.params); mlx5e_timestamp_init(priv); @@ -419,6 +419,28 @@ static void mlx5i_cleanup_rx(struct mlx5e_priv *priv) mlx5e_destroy_q_counters(priv); } +/* The stats groups order is opposite to the update_stats() order calls */ +static mlx5e_stats_grp_t mlx5i_stats_grps[] = { + &MLX5E_STATS_GRP(sw), + &MLX5E_STATS_GRP(qcnt), + &MLX5E_STATS_GRP(vnic_env), + &MLX5E_STATS_GRP(vport), + &MLX5E_STATS_GRP(802_3), + &MLX5E_STATS_GRP(2863), + &MLX5E_STATS_GRP(2819), + &MLX5E_STATS_GRP(phy), + &MLX5E_STATS_GRP(pcie), + &MLX5E_STATS_GRP(per_prio), + &MLX5E_STATS_GRP(pme), + &MLX5E_STATS_GRP(channels), + &MLX5E_STATS_GRP(per_port_buff_congest), +}; + +static unsigned int mlx5i_stats_grps_num(struct mlx5e_priv *priv) +{ + return ARRAY_SIZE(mlx5i_stats_grps); +} + static const struct mlx5e_profile mlx5i_nic_profile = { .init = mlx5i_init, .cleanup = mlx5i_cleanup, @@ -435,6 +457,8 @@ static const struct mlx5e_profile mlx5i_nic_profile = { .rx_handlers.handle_rx_cqe_mpwqe = NULL, /* Not supported */ .max_tc = MLX5I_MAX_NUM_TC, .rq_groups = MLX5E_NUM_RQ_GROUPS(REGULAR), + .stats_grps = mlx5i_stats_grps, + .stats_grps_num = mlx5i_stats_grps_num, }; /* mlx5i netdev NDos */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag.c b/drivers/net/ethernet/mellanox/mlx5/core/lag.c index fc0d9583475d..b91eabc09fbc 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag.c @@ -586,7 +586,8 @@ void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev) if (!ldev->nb.notifier_call) { ldev->nb.notifier_call = mlx5_lag_netdev_event; - if (register_netdevice_notifier(&ldev->nb)) { + if (register_netdevice_notifier_dev_net(netdev, &ldev->nb, + &ldev->nn)) { ldev->nb.notifier_call = NULL; mlx5_core_err(dev, "Failed to register LAG netdev notifier\n"); } @@ -599,7 +600,7 @@ void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev) } /* Must be called with intf_mutex held */ -void mlx5_lag_remove(struct mlx5_core_dev *dev) +void mlx5_lag_remove(struct mlx5_core_dev *dev, struct net_device *netdev) { struct mlx5_lag *ldev; int i; @@ -619,7 +620,8 @@ void mlx5_lag_remove(struct mlx5_core_dev *dev) if (i == MLX5_MAX_PORTS) { if (ldev->nb.notifier_call) - unregister_netdevice_notifier(&ldev->nb); + unregister_netdevice_notifier_dev_net(netdev, &ldev->nb, + &ldev->nn); mlx5_lag_mp_cleanup(ldev); cancel_delayed_work_sync(&ldev->bond_work); mlx5_lag_dev_free(ldev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag.h b/drivers/net/ethernet/mellanox/mlx5/core/lag.h index f1068aac6406..316ab09e2664 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag.h @@ -44,6 +44,7 @@ struct mlx5_lag { struct workqueue_struct *wq; struct delayed_work bond_work; struct notifier_block nb; + struct netdev_net_notifier nn; struct lag_mp lag_mp; }; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/lag_mp.c b/drivers/net/ethernet/mellanox/mlx5/core/lag_mp.c index b70afa310ad2..416676c35b1f 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/lag_mp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lag_mp.c @@ -200,8 +200,6 @@ static void mlx5_lag_fib_update(struct work_struct *work) rtnl_lock(); switch (fib_work->event) { case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ - case FIB_EVENT_ENTRY_ADD: /* fall through */ case FIB_EVENT_ENTRY_DEL: mlx5_lag_fib_route_event(ldev, fib_work->event, fib_work->fen_info.fi); @@ -259,8 +257,6 @@ static int mlx5_lag_fib_event(struct notifier_block *nb, switch (event) { case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ - case FIB_EVENT_ENTRY_ADD: /* fall through */ case FIB_EVENT_ENTRY_DEL: fen_info = container_of(info, struct fib_entry_notifier_info, info); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h index da67b28d6e23..fcce9e0fc82c 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/mlx5_core.h @@ -157,7 +157,7 @@ int mlx5_query_qcam_reg(struct mlx5_core_dev *mdev, u32 *qcam, u8 feature_group, u8 access_reg_group); void mlx5_lag_add(struct mlx5_core_dev *dev, struct net_device *netdev); -void mlx5_lag_remove(struct mlx5_core_dev *dev); +void mlx5_lag_remove(struct mlx5_core_dev *dev, struct net_device *netdev); int mlx5_irq_table_init(struct mlx5_core_dev *dev); void mlx5_irq_table_cleanup(struct mlx5_core_dev *dev); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c index 004c56c2fc0c..6dec2a550a10 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_action.c @@ -677,9 +677,12 @@ int mlx5dr_actions_build_ste_arr(struct mlx5dr_matcher *matcher, goto out_invalid_arg; } if (action->dest_tbl.tbl->level <= matcher->tbl->level) { + mlx5_core_warn_once(dmn->mdev, + "Connecting table to a lower/same level destination table\n"); mlx5dr_dbg(dmn, - "Destination table level should be higher than source table\n"); - goto out_invalid_arg; + "Connecting table at level %d to a destination table at level %d\n", + matcher->tbl->level, + action->dest_tbl.tbl->level); } attr.final_icm_addr = rx_rule ? action->dest_tbl.tbl->rx.s_anchor->chunk->icm_addr : @@ -690,9 +693,9 @@ int mlx5dr_actions_build_ste_arr(struct mlx5dr_matcher *matcher, /* get the relevant addresses */ if (!action->dest_tbl.fw_tbl.rx_icm_addr) { - ret = mlx5dr_cmd_query_flow_table(action->dest_tbl.fw_tbl.mdev, - action->dest_tbl.fw_tbl.ft->type, - action->dest_tbl.fw_tbl.ft->id, + ret = mlx5dr_cmd_query_flow_table(dmn->mdev, + action->dest_tbl.fw_tbl.type, + action->dest_tbl.fw_tbl.id, &output); if (!ret) { action->dest_tbl.fw_tbl.tx_icm_addr = @@ -982,8 +985,106 @@ dec_ref: } struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev) +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests) +{ + struct mlx5dr_cmd_flow_destination_hw_info *hw_dests; + struct mlx5dr_action **ref_actions; + struct mlx5dr_action *action; + bool reformat_req = false; + u32 num_of_ref = 0; + int ret; + int i; + + if (dmn->type != MLX5DR_DOMAIN_TYPE_FDB) { + mlx5dr_err(dmn, "Multiple destination support is for FDB only\n"); + return NULL; + } + + hw_dests = kzalloc(sizeof(*hw_dests) * num_of_dests, GFP_KERNEL); + if (!hw_dests) + return NULL; + + ref_actions = kzalloc(sizeof(*ref_actions) * num_of_dests * 2, GFP_KERNEL); + if (!ref_actions) + goto free_hw_dests; + + for (i = 0; i < num_of_dests; i++) { + struct mlx5dr_action *reformat_action = dests[i].reformat; + struct mlx5dr_action *dest_action = dests[i].dest; + + ref_actions[num_of_ref++] = dest_action; + + switch (dest_action->action_type) { + case DR_ACTION_TYP_VPORT: + hw_dests[i].vport.flags = MLX5_FLOW_DEST_VPORT_VHCA_ID; + hw_dests[i].type = MLX5_FLOW_DESTINATION_TYPE_VPORT; + hw_dests[i].vport.num = dest_action->vport.caps->num; + hw_dests[i].vport.vhca_id = dest_action->vport.caps->vhca_gvmi; + if (reformat_action) { + reformat_req = true; + hw_dests[i].vport.reformat_id = + reformat_action->reformat.reformat_id; + ref_actions[num_of_ref++] = reformat_action; + hw_dests[i].vport.flags |= MLX5_FLOW_DEST_VPORT_REFORMAT_ID; + } + break; + + case DR_ACTION_TYP_FT: + hw_dests[i].type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + if (dest_action->dest_tbl.is_fw_tbl) + hw_dests[i].ft_id = dest_action->dest_tbl.fw_tbl.id; + else + hw_dests[i].ft_id = dest_action->dest_tbl.tbl->table_id; + break; + + default: + mlx5dr_dbg(dmn, "Invalid multiple destinations action\n"); + goto free_ref_actions; + } + } + + action = dr_action_create_generic(DR_ACTION_TYP_FT); + if (!action) + goto free_ref_actions; + + ret = mlx5dr_fw_create_md_tbl(dmn, + hw_dests, + num_of_dests, + reformat_req, + &action->dest_tbl.fw_tbl.id, + &action->dest_tbl.fw_tbl.group_id); + if (ret) + goto free_action; + + refcount_inc(&dmn->refcount); + + for (i = 0; i < num_of_ref; i++) + refcount_inc(&ref_actions[i]->refcount); + + action->dest_tbl.is_fw_tbl = true; + action->dest_tbl.fw_tbl.dmn = dmn; + action->dest_tbl.fw_tbl.type = FS_FT_FDB; + action->dest_tbl.fw_tbl.ref_actions = ref_actions; + action->dest_tbl.fw_tbl.num_of_ref_actions = num_of_ref; + + kfree(hw_dests); + + return action; + +free_action: + kfree(action); +free_ref_actions: + kfree(ref_actions); +free_hw_dests: + kfree(hw_dests); + return NULL; +} + +struct mlx5dr_action * +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *dmn, + struct mlx5_flow_table *ft) { struct mlx5dr_action *action; @@ -992,8 +1093,11 @@ mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, return NULL; action->dest_tbl.is_fw_tbl = 1; - action->dest_tbl.fw_tbl.ft = ft; - action->dest_tbl.fw_tbl.mdev = mdev; + action->dest_tbl.fw_tbl.type = ft->type; + action->dest_tbl.fw_tbl.id = ft->id; + action->dest_tbl.fw_tbl.dmn = dmn; + + refcount_inc(&dmn->refcount); return action; } @@ -1213,58 +1317,85 @@ not_found: } static int -dr_action_modify_sw_to_hw(struct mlx5dr_domain *dmn, - __be64 *sw_action, - __be64 *hw_action, - const struct dr_action_modify_field_conv **ret_hw_info) +dr_action_modify_sw_to_hw_add(struct mlx5dr_domain *dmn, + __be64 *sw_action, + __be64 *hw_action, + const struct dr_action_modify_field_conv **ret_hw_info) { const struct dr_action_modify_field_conv *hw_action_info; - u8 offset, length, max_length, action; + u8 max_length; u16 sw_field; - u8 hw_opcode; u32 data; /* Get SW modify action data */ - action = MLX5_GET(set_action_in, sw_action, action_type); - length = MLX5_GET(set_action_in, sw_action, length); - offset = MLX5_GET(set_action_in, sw_action, offset); sw_field = MLX5_GET(set_action_in, sw_action, field); data = MLX5_GET(set_action_in, sw_action, data); /* Convert SW data to HW modify action format */ hw_action_info = dr_action_modify_get_hw_info(sw_field); if (!hw_action_info) { - mlx5dr_dbg(dmn, "Modify action invalid field given\n"); + mlx5dr_dbg(dmn, "Modify add action invalid field given\n"); return -EINVAL; } max_length = hw_action_info->end - hw_action_info->start + 1; - switch (action) { - case MLX5_ACTION_TYPE_SET: - hw_opcode = MLX5DR_ACTION_MDFY_HW_OP_SET; - /* PRM defines that length zero specific length of 32bits */ - if (!length) - length = 32; + MLX5_SET(dr_action_hw_set, hw_action, + opcode, MLX5DR_ACTION_MDFY_HW_OP_ADD); - if (length + offset > max_length) { - mlx5dr_dbg(dmn, "Modify action length + offset exceeds limit\n"); - return -EINVAL; - } - break; + MLX5_SET(dr_action_hw_set, hw_action, destination_field_code, + hw_action_info->hw_field); - case MLX5_ACTION_TYPE_ADD: - hw_opcode = MLX5DR_ACTION_MDFY_HW_OP_ADD; - offset = 0; - length = max_length; - break; + MLX5_SET(dr_action_hw_set, hw_action, destination_left_shifter, + hw_action_info->start); - default: - mlx5dr_info(dmn, "Unsupported action_type for modify action\n"); - return -EOPNOTSUPP; + /* PRM defines that length zero specific length of 32bits */ + MLX5_SET(dr_action_hw_set, hw_action, destination_length, + max_length == 32 ? 0 : max_length); + + MLX5_SET(dr_action_hw_set, hw_action, inline_data, data); + + *ret_hw_info = hw_action_info; + + return 0; +} + +static int +dr_action_modify_sw_to_hw_set(struct mlx5dr_domain *dmn, + __be64 *sw_action, + __be64 *hw_action, + const struct dr_action_modify_field_conv **ret_hw_info) +{ + const struct dr_action_modify_field_conv *hw_action_info; + u8 offset, length, max_length; + u16 sw_field; + u32 data; + + /* Get SW modify action data */ + length = MLX5_GET(set_action_in, sw_action, length); + offset = MLX5_GET(set_action_in, sw_action, offset); + sw_field = MLX5_GET(set_action_in, sw_action, field); + data = MLX5_GET(set_action_in, sw_action, data); + + /* Convert SW data to HW modify action format */ + hw_action_info = dr_action_modify_get_hw_info(sw_field); + if (!hw_action_info) { + mlx5dr_dbg(dmn, "Modify set action invalid field given\n"); + return -EINVAL; + } + + /* PRM defines that length zero specific length of 32bits */ + length = length ? length : 32; + + max_length = hw_action_info->end - hw_action_info->start + 1; + + if (length + offset > max_length) { + mlx5dr_dbg(dmn, "Modify action length + offset exceeds limit\n"); + return -EINVAL; } - MLX5_SET(dr_action_hw_set, hw_action, opcode, hw_opcode); + MLX5_SET(dr_action_hw_set, hw_action, + opcode, MLX5DR_ACTION_MDFY_HW_OP_SET); MLX5_SET(dr_action_hw_set, hw_action, destination_field_code, hw_action_info->hw_field); @@ -1283,48 +1414,236 @@ dr_action_modify_sw_to_hw(struct mlx5dr_domain *dmn, } static int -dr_action_modify_check_field_limitation(struct mlx5dr_domain *dmn, - const __be64 *sw_action) +dr_action_modify_sw_to_hw_copy(struct mlx5dr_domain *dmn, + __be64 *sw_action, + __be64 *hw_action, + const struct dr_action_modify_field_conv **ret_dst_hw_info, + const struct dr_action_modify_field_conv **ret_src_hw_info) +{ + u8 src_offset, dst_offset, src_max_length, dst_max_length, length; + const struct dr_action_modify_field_conv *hw_dst_action_info; + const struct dr_action_modify_field_conv *hw_src_action_info; + u16 src_field, dst_field; + + /* Get SW modify action data */ + src_field = MLX5_GET(copy_action_in, sw_action, src_field); + dst_field = MLX5_GET(copy_action_in, sw_action, dst_field); + src_offset = MLX5_GET(copy_action_in, sw_action, src_offset); + dst_offset = MLX5_GET(copy_action_in, sw_action, dst_offset); + length = MLX5_GET(copy_action_in, sw_action, length); + + /* Convert SW data to HW modify action format */ + hw_src_action_info = dr_action_modify_get_hw_info(src_field); + hw_dst_action_info = dr_action_modify_get_hw_info(dst_field); + if (!hw_src_action_info || !hw_dst_action_info) { + mlx5dr_dbg(dmn, "Modify copy action invalid field given\n"); + return -EINVAL; + } + + /* PRM defines that length zero specific length of 32bits */ + length = length ? length : 32; + + src_max_length = hw_src_action_info->end - + hw_src_action_info->start + 1; + dst_max_length = hw_dst_action_info->end - + hw_dst_action_info->start + 1; + + if (length + src_offset > src_max_length || + length + dst_offset > dst_max_length) { + mlx5dr_dbg(dmn, "Modify action length + offset exceeds limit\n"); + return -EINVAL; + } + + MLX5_SET(dr_action_hw_copy, hw_action, + opcode, MLX5DR_ACTION_MDFY_HW_OP_COPY); + + MLX5_SET(dr_action_hw_copy, hw_action, destination_field_code, + hw_dst_action_info->hw_field); + + MLX5_SET(dr_action_hw_copy, hw_action, destination_left_shifter, + hw_dst_action_info->start + dst_offset); + + MLX5_SET(dr_action_hw_copy, hw_action, destination_length, + length == 32 ? 0 : length); + + MLX5_SET(dr_action_hw_copy, hw_action, source_field_code, + hw_src_action_info->hw_field); + + MLX5_SET(dr_action_hw_copy, hw_action, source_left_shifter, + hw_src_action_info->start + dst_offset); + + *ret_dst_hw_info = hw_dst_action_info; + *ret_src_hw_info = hw_src_action_info; + + return 0; +} + +static int +dr_action_modify_sw_to_hw(struct mlx5dr_domain *dmn, + __be64 *sw_action, + __be64 *hw_action, + const struct dr_action_modify_field_conv **ret_dst_hw_info, + const struct dr_action_modify_field_conv **ret_src_hw_info) { - u16 sw_field; u8 action; + int ret; - sw_field = MLX5_GET(set_action_in, sw_action, field); + *hw_action = 0; + *ret_src_hw_info = NULL; + + /* Get SW modify action type */ action = MLX5_GET(set_action_in, sw_action, action_type); - /* Check if SW field is supported in current domain (RX/TX) */ - if (action == MLX5_ACTION_TYPE_SET) { - if (sw_field == MLX5_ACTION_IN_FIELD_METADATA_REG_A) { + switch (action) { + case MLX5_ACTION_TYPE_SET: + ret = dr_action_modify_sw_to_hw_set(dmn, sw_action, + hw_action, + ret_dst_hw_info); + break; + + case MLX5_ACTION_TYPE_ADD: + ret = dr_action_modify_sw_to_hw_add(dmn, sw_action, + hw_action, + ret_dst_hw_info); + break; + + case MLX5_ACTION_TYPE_COPY: + ret = dr_action_modify_sw_to_hw_copy(dmn, sw_action, + hw_action, + ret_dst_hw_info, + ret_src_hw_info); + break; + + default: + mlx5dr_info(dmn, "Unsupported action_type for modify action\n"); + ret = -EOPNOTSUPP; + } + + return ret; +} + +static int +dr_action_modify_check_set_field_limitation(struct mlx5dr_action *action, + const __be64 *sw_action) +{ + u16 sw_field = MLX5_GET(set_action_in, sw_action, field); + struct mlx5dr_domain *dmn = action->rewrite.dmn; + + if (sw_field == MLX5_ACTION_IN_FIELD_METADATA_REG_A) { + action->rewrite.allow_rx = 0; + if (dmn->type != MLX5DR_DOMAIN_TYPE_NIC_TX) { + mlx5dr_dbg(dmn, "Unsupported field %d for RX/FDB set action\n", + sw_field); + return -EINVAL; + } + } else if (sw_field == MLX5_ACTION_IN_FIELD_METADATA_REG_B) { + action->rewrite.allow_tx = 0; + if (dmn->type != MLX5DR_DOMAIN_TYPE_NIC_RX) { + mlx5dr_dbg(dmn, "Unsupported field %d for TX/FDB set action\n", + sw_field); + return -EINVAL; + } + } + + if (!action->rewrite.allow_rx && !action->rewrite.allow_tx) { + mlx5dr_dbg(dmn, "Modify SET actions not supported on both RX and TX\n"); + return -EINVAL; + } + + return 0; +} + +static int +dr_action_modify_check_add_field_limitation(struct mlx5dr_action *action, + const __be64 *sw_action) +{ + u16 sw_field = MLX5_GET(set_action_in, sw_action, field); + struct mlx5dr_domain *dmn = action->rewrite.dmn; + + if (sw_field != MLX5_ACTION_IN_FIELD_OUT_IP_TTL && + sw_field != MLX5_ACTION_IN_FIELD_OUT_IPV6_HOPLIMIT && + sw_field != MLX5_ACTION_IN_FIELD_OUT_TCP_SEQ_NUM && + sw_field != MLX5_ACTION_IN_FIELD_OUT_TCP_ACK_NUM) { + mlx5dr_dbg(dmn, "Unsupported field %d for add action\n", + sw_field); + return -EINVAL; + } + + return 0; +} + +static int +dr_action_modify_check_copy_field_limitation(struct mlx5dr_action *action, + const __be64 *sw_action) +{ + struct mlx5dr_domain *dmn = action->rewrite.dmn; + u16 sw_fields[2]; + int i; + + sw_fields[0] = MLX5_GET(copy_action_in, sw_action, src_field); + sw_fields[1] = MLX5_GET(copy_action_in, sw_action, dst_field); + + for (i = 0; i < 2; i++) { + if (sw_fields[i] == MLX5_ACTION_IN_FIELD_METADATA_REG_A) { + action->rewrite.allow_rx = 0; if (dmn->type != MLX5DR_DOMAIN_TYPE_NIC_TX) { mlx5dr_dbg(dmn, "Unsupported field %d for RX/FDB set action\n", - sw_field); + sw_fields[i]); return -EINVAL; } - } - - if (sw_field == MLX5_ACTION_IN_FIELD_METADATA_REG_B) { + } else if (sw_fields[i] == MLX5_ACTION_IN_FIELD_METADATA_REG_B) { + action->rewrite.allow_tx = 0; if (dmn->type != MLX5DR_DOMAIN_TYPE_NIC_RX) { mlx5dr_dbg(dmn, "Unsupported field %d for TX/FDB set action\n", - sw_field); + sw_fields[i]); return -EINVAL; } } - } else if (action == MLX5_ACTION_TYPE_ADD) { - if (sw_field != MLX5_ACTION_IN_FIELD_OUT_IP_TTL && - sw_field != MLX5_ACTION_IN_FIELD_OUT_IPV6_HOPLIMIT && - sw_field != MLX5_ACTION_IN_FIELD_OUT_TCP_SEQ_NUM && - sw_field != MLX5_ACTION_IN_FIELD_OUT_TCP_ACK_NUM) { - mlx5dr_dbg(dmn, "Unsupported field %d for add action\n", sw_field); - return -EINVAL; - } - } else { - mlx5dr_info(dmn, "Unsupported action %d modify action\n", action); - return -EOPNOTSUPP; + } + + if (!action->rewrite.allow_rx && !action->rewrite.allow_tx) { + mlx5dr_dbg(dmn, "Modify copy actions not supported on both RX and TX\n"); + return -EINVAL; } return 0; } +static int +dr_action_modify_check_field_limitation(struct mlx5dr_action *action, + const __be64 *sw_action) +{ + struct mlx5dr_domain *dmn = action->rewrite.dmn; + u8 action_type; + int ret; + + action_type = MLX5_GET(set_action_in, sw_action, action_type); + + switch (action_type) { + case MLX5_ACTION_TYPE_SET: + ret = dr_action_modify_check_set_field_limitation(action, + sw_action); + break; + + case MLX5_ACTION_TYPE_ADD: + ret = dr_action_modify_check_add_field_limitation(action, + sw_action); + break; + + case MLX5_ACTION_TYPE_COPY: + ret = dr_action_modify_check_copy_field_limitation(action, + sw_action); + break; + + default: + mlx5dr_info(dmn, "Unsupported action %d modify action\n", + action_type); + ret = -EOPNOTSUPP; + } + + return ret; +} + static bool dr_action_modify_check_is_ttl_modify(const u64 *sw_action) { @@ -1333,7 +1652,7 @@ dr_action_modify_check_is_ttl_modify(const u64 *sw_action) return sw_field == MLX5_ACTION_IN_FIELD_OUT_IP_TTL; } -static int dr_actions_convert_modify_header(struct mlx5dr_domain *dmn, +static int dr_actions_convert_modify_header(struct mlx5dr_action *action, u32 max_hw_actions, u32 num_sw_actions, __be64 sw_actions[], @@ -1341,20 +1660,26 @@ static int dr_actions_convert_modify_header(struct mlx5dr_domain *dmn, u32 *num_hw_actions, bool *modify_ttl) { - const struct dr_action_modify_field_conv *hw_action_info; + const struct dr_action_modify_field_conv *hw_dst_action_info; + const struct dr_action_modify_field_conv *hw_src_action_info; u16 hw_field = MLX5DR_ACTION_MDFY_HW_FLD_RESERVED; u32 l3_type = MLX5DR_ACTION_MDFY_HW_HDR_L3_NONE; u32 l4_type = MLX5DR_ACTION_MDFY_HW_HDR_L4_NONE; + struct mlx5dr_domain *dmn = action->rewrite.dmn; int ret, i, hw_idx = 0; __be64 *sw_action; __be64 hw_action; *modify_ttl = false; + action->rewrite.allow_rx = 1; + action->rewrite.allow_tx = 1; + for (i = 0; i < num_sw_actions; i++) { sw_action = &sw_actions[i]; - ret = dr_action_modify_check_field_limitation(dmn, sw_action); + ret = dr_action_modify_check_field_limitation(action, + sw_action); if (ret) return ret; @@ -1365,32 +1690,35 @@ static int dr_actions_convert_modify_header(struct mlx5dr_domain *dmn, ret = dr_action_modify_sw_to_hw(dmn, sw_action, &hw_action, - &hw_action_info); + &hw_dst_action_info, + &hw_src_action_info); if (ret) return ret; /* Due to a HW limitation we cannot modify 2 different L3 types */ - if (l3_type && hw_action_info->l3_type && - hw_action_info->l3_type != l3_type) { + if (l3_type && hw_dst_action_info->l3_type && + hw_dst_action_info->l3_type != l3_type) { mlx5dr_dbg(dmn, "Action list can't support two different L3 types\n"); return -EINVAL; } - if (hw_action_info->l3_type) - l3_type = hw_action_info->l3_type; + if (hw_dst_action_info->l3_type) + l3_type = hw_dst_action_info->l3_type; /* Due to a HW limitation we cannot modify two different L4 types */ - if (l4_type && hw_action_info->l4_type && - hw_action_info->l4_type != l4_type) { + if (l4_type && hw_dst_action_info->l4_type && + hw_dst_action_info->l4_type != l4_type) { mlx5dr_dbg(dmn, "Action list can't support two different L4 types\n"); return -EINVAL; } - if (hw_action_info->l4_type) - l4_type = hw_action_info->l4_type; + if (hw_dst_action_info->l4_type) + l4_type = hw_dst_action_info->l4_type; /* HW reads and executes two actions at once this means we * need to create a gap if two actions access the same field */ - if ((hw_idx % 2) && hw_field == hw_action_info->hw_field) { + if ((hw_idx % 2) && (hw_field == hw_dst_action_info->hw_field || + (hw_src_action_info && + hw_field == hw_src_action_info->hw_field))) { /* Check if after gap insertion the total number of HW * modify actions doesn't exceeds the limit */ @@ -1400,7 +1728,7 @@ static int dr_actions_convert_modify_header(struct mlx5dr_domain *dmn, return -EINVAL; } } - hw_field = hw_action_info->hw_field; + hw_field = hw_dst_action_info->hw_field; hw_actions[hw_idx] = hw_action; hw_idx++; @@ -1443,7 +1771,7 @@ static int dr_action_create_modify_action(struct mlx5dr_domain *dmn, goto free_chunk; } - ret = dr_actions_convert_modify_header(dmn, + ret = dr_actions_convert_modify_header(action, max_hw_actions, num_sw_actions, actions, @@ -1559,8 +1887,26 @@ int mlx5dr_action_destroy(struct mlx5dr_action *action) switch (action->action_type) { case DR_ACTION_TYP_FT: - if (!action->dest_tbl.is_fw_tbl) + if (action->dest_tbl.is_fw_tbl) + refcount_dec(&action->dest_tbl.fw_tbl.dmn->refcount); + else refcount_dec(&action->dest_tbl.tbl->refcount); + + if (action->dest_tbl.is_fw_tbl && + action->dest_tbl.fw_tbl.num_of_ref_actions) { + struct mlx5dr_action **ref_actions; + int i; + + ref_actions = action->dest_tbl.fw_tbl.ref_actions; + for (i = 0; i < action->dest_tbl.fw_tbl.num_of_ref_actions; i++) + refcount_dec(&ref_actions[i]->refcount); + + kfree(ref_actions); + + mlx5dr_fw_destroy_md_tbl(action->dest_tbl.fw_tbl.dmn, + action->dest_tbl.fw_tbl.id, + action->dest_tbl.fw_tbl.group_id); + } break; case DR_ACTION_TYP_TNL_L2_TO_L2: refcount_dec(&action->reformat.dmn->refcount); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c index 41662c4e2664..461b39376daf 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_cmd.c @@ -320,12 +320,7 @@ int mlx5dr_cmd_destroy_flow_group(struct mlx5_core_dev *mdev, } int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, - u32 table_type, - u64 icm_addr_rx, - u64 icm_addr_tx, - u8 level, - bool sw_owner, - bool term_tbl, + struct mlx5dr_cmd_create_flow_table_attr *attr, u64 *fdb_rx_icm_addr, u32 *table_id) { @@ -335,37 +330,43 @@ int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, int err; MLX5_SET(create_flow_table_in, in, opcode, MLX5_CMD_OP_CREATE_FLOW_TABLE); - MLX5_SET(create_flow_table_in, in, table_type, table_type); + MLX5_SET(create_flow_table_in, in, table_type, attr->table_type); ft_mdev = MLX5_ADDR_OF(create_flow_table_in, in, flow_table_context); - MLX5_SET(flow_table_context, ft_mdev, termination_table, term_tbl); - MLX5_SET(flow_table_context, ft_mdev, sw_owner, sw_owner); - MLX5_SET(flow_table_context, ft_mdev, level, level); + MLX5_SET(flow_table_context, ft_mdev, termination_table, attr->term_tbl); + MLX5_SET(flow_table_context, ft_mdev, sw_owner, attr->sw_owner); + MLX5_SET(flow_table_context, ft_mdev, level, attr->level); - if (sw_owner) { + if (attr->sw_owner) { /* icm_addr_0 used for FDB RX / NIC TX / NIC_RX * icm_addr_1 used for FDB TX */ - if (table_type == MLX5_FLOW_TABLE_TYPE_NIC_RX) { + if (attr->table_type == MLX5_FLOW_TABLE_TYPE_NIC_RX) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_rx); - } else if (table_type == MLX5_FLOW_TABLE_TYPE_NIC_TX) { + sw_owner_icm_root_0, attr->icm_addr_rx); + } else if (attr->table_type == MLX5_FLOW_TABLE_TYPE_NIC_TX) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_tx); - } else if (table_type == MLX5_FLOW_TABLE_TYPE_FDB) { + sw_owner_icm_root_0, attr->icm_addr_tx); + } else if (attr->table_type == MLX5_FLOW_TABLE_TYPE_FDB) { MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_0, icm_addr_rx); + sw_owner_icm_root_0, attr->icm_addr_rx); MLX5_SET64(flow_table_context, ft_mdev, - sw_owner_icm_root_1, icm_addr_tx); + sw_owner_icm_root_1, attr->icm_addr_tx); } } + MLX5_SET(create_flow_table_in, in, flow_table_context.decap_en, + attr->decap_en); + MLX5_SET(create_flow_table_in, in, flow_table_context.reformat_en, + attr->reformat_en); + err = mlx5_cmd_exec(mdev, in, sizeof(in), out, sizeof(out)); if (err) return err; *table_id = MLX5_GET(create_flow_table_out, out, table_id); - if (!sw_owner && table_type == MLX5_FLOW_TABLE_TYPE_FDB) + if (!attr->sw_owner && attr->table_type == MLX5_FLOW_TABLE_TYPE_FDB && + fdb_rx_icm_addr) *fdb_rx_icm_addr = (u64)MLX5_GET(create_flow_table_out, out, icm_address_31_0) | (u64)MLX5_GET(create_flow_table_out, out, icm_address_39_32) << 32 | @@ -478,3 +479,208 @@ int mlx5dr_cmd_query_gid(struct mlx5_core_dev *mdev, u8 vhca_port_num, return 0; } + +static int mlx5dr_cmd_set_extended_dest(struct mlx5_core_dev *dev, + struct mlx5dr_cmd_fte_info *fte, + bool *extended_dest) +{ + int fw_log_max_fdb_encap_uplink = MLX5_CAP_ESW(dev, log_max_fdb_encap_uplink); + int num_fwd_destinations = 0; + int num_encap = 0; + int i; + + *extended_dest = false; + if (!(fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST)) + return 0; + for (i = 0; i < fte->dests_size; i++) { + if (fte->dest_arr[i].type == MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + if (fte->dest_arr[i].type == MLX5_FLOW_DESTINATION_TYPE_VPORT && + fte->dest_arr[i].vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID) + num_encap++; + num_fwd_destinations++; + } + + if (num_fwd_destinations > 1 && num_encap > 0) + *extended_dest = true; + + if (*extended_dest && !fw_log_max_fdb_encap_uplink) { + mlx5_core_warn(dev, "FW does not support extended destination"); + return -EOPNOTSUPP; + } + if (num_encap > (1 << fw_log_max_fdb_encap_uplink)) { + mlx5_core_warn(dev, "FW does not support more than %d encaps", + 1 << fw_log_max_fdb_encap_uplink); + return -EOPNOTSUPP; + } + + return 0; +} + +int mlx5dr_cmd_set_fte(struct mlx5_core_dev *dev, + int opmod, int modify_mask, + struct mlx5dr_cmd_ft_info *ft, + u32 group_id, + struct mlx5dr_cmd_fte_info *fte) +{ + u32 out[MLX5_ST_SZ_DW(set_fte_out)] = {}; + void *in_flow_context, *vlan; + bool extended_dest = false; + void *in_match_value; + unsigned int inlen; + int dst_cnt_size; + void *in_dests; + u32 *in; + int err; + int i; + + if (mlx5dr_cmd_set_extended_dest(dev, fte, &extended_dest)) + return -EOPNOTSUPP; + + if (!extended_dest) + dst_cnt_size = MLX5_ST_SZ_BYTES(dest_format_struct); + else + dst_cnt_size = MLX5_ST_SZ_BYTES(extended_dest_format); + + inlen = MLX5_ST_SZ_BYTES(set_fte_in) + fte->dests_size * dst_cnt_size; + in = kvzalloc(inlen, GFP_KERNEL); + if (!in) + return -ENOMEM; + + MLX5_SET(set_fte_in, in, opcode, MLX5_CMD_OP_SET_FLOW_TABLE_ENTRY); + MLX5_SET(set_fte_in, in, op_mod, opmod); + MLX5_SET(set_fte_in, in, modify_enable_mask, modify_mask); + MLX5_SET(set_fte_in, in, table_type, ft->type); + MLX5_SET(set_fte_in, in, table_id, ft->id); + MLX5_SET(set_fte_in, in, flow_index, fte->index); + if (ft->vport) { + MLX5_SET(set_fte_in, in, vport_number, ft->vport); + MLX5_SET(set_fte_in, in, other_vport, 1); + } + + in_flow_context = MLX5_ADDR_OF(set_fte_in, in, flow_context); + MLX5_SET(flow_context, in_flow_context, group_id, group_id); + + MLX5_SET(flow_context, in_flow_context, flow_tag, + fte->flow_context.flow_tag); + MLX5_SET(flow_context, in_flow_context, flow_source, + fte->flow_context.flow_source); + + MLX5_SET(flow_context, in_flow_context, extended_destination, + extended_dest); + if (extended_dest) { + u32 action; + + action = fte->action.action & + ~MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT; + MLX5_SET(flow_context, in_flow_context, action, action); + } else { + MLX5_SET(flow_context, in_flow_context, action, + fte->action.action); + if (fte->action.pkt_reformat) + MLX5_SET(flow_context, in_flow_context, packet_reformat_id, + fte->action.pkt_reformat->id); + } + if (fte->action.modify_hdr) + MLX5_SET(flow_context, in_flow_context, modify_header_id, + fte->action.modify_hdr->id); + + vlan = MLX5_ADDR_OF(flow_context, in_flow_context, push_vlan); + + MLX5_SET(vlan, vlan, ethtype, fte->action.vlan[0].ethtype); + MLX5_SET(vlan, vlan, vid, fte->action.vlan[0].vid); + MLX5_SET(vlan, vlan, prio, fte->action.vlan[0].prio); + + vlan = MLX5_ADDR_OF(flow_context, in_flow_context, push_vlan_2); + + MLX5_SET(vlan, vlan, ethtype, fte->action.vlan[1].ethtype); + MLX5_SET(vlan, vlan, vid, fte->action.vlan[1].vid); + MLX5_SET(vlan, vlan, prio, fte->action.vlan[1].prio); + + in_match_value = MLX5_ADDR_OF(flow_context, in_flow_context, + match_value); + memcpy(in_match_value, fte->val, sizeof(u32) * MLX5_ST_SZ_DW_MATCH_PARAM); + + in_dests = MLX5_ADDR_OF(flow_context, in_flow_context, destination); + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { + int list_size = 0; + + for (i = 0; i < fte->dests_size; i++) { + unsigned int id, type = fte->dest_arr[i].type; + + if (type == MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + + switch (type) { + case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE_NUM: + id = fte->dest_arr[i].ft_num; + type = MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE; + break; + case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE: + id = fte->dest_arr[i].ft_id; + break; + case MLX5_FLOW_DESTINATION_TYPE_VPORT: + id = fte->dest_arr[i].vport.num; + MLX5_SET(dest_format_struct, in_dests, + destination_eswitch_owner_vhca_id_valid, + !!(fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_VHCA_ID)); + MLX5_SET(dest_format_struct, in_dests, + destination_eswitch_owner_vhca_id, + fte->dest_arr[i].vport.vhca_id); + if (extended_dest && (fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID)) { + MLX5_SET(dest_format_struct, in_dests, + packet_reformat, + !!(fte->dest_arr[i].vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID)); + MLX5_SET(extended_dest_format, in_dests, + packet_reformat_id, + fte->dest_arr[i].vport.reformat_id); + } + break; + default: + id = fte->dest_arr[i].tir_num; + } + + MLX5_SET(dest_format_struct, in_dests, destination_type, + type); + MLX5_SET(dest_format_struct, in_dests, destination_id, id); + in_dests += dst_cnt_size; + list_size++; + } + + MLX5_SET(flow_context, in_flow_context, destination_list_size, + list_size); + } + + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_COUNT) { + int max_list_size = BIT(MLX5_CAP_FLOWTABLE_TYPE(dev, + log_max_flow_counter, + ft->type)); + int list_size = 0; + + for (i = 0; i < fte->dests_size; i++) { + if (fte->dest_arr[i].type != + MLX5_FLOW_DESTINATION_TYPE_COUNTER) + continue; + + MLX5_SET(flow_counter_list, in_dests, flow_counter_id, + fte->dest_arr[i].counter_id); + in_dests += dst_cnt_size; + list_size++; + } + if (list_size > max_list_size) { + err = -EINVAL; + goto err_out; + } + + MLX5_SET(flow_context, in_flow_context, flow_counter_list_size, + list_size); + } + + err = mlx5_cmd_exec(dev, in, inlen, out, sizeof(out)); +err_out: + kvfree(in); + return err; +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c index 60ef6e6171e3..1fbcd012bb85 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_fw.c @@ -7,6 +7,7 @@ struct mlx5dr_fw_recalc_cs_ft * mlx5dr_fw_create_recalc_cs_ft(struct mlx5dr_domain *dmn, u32 vport_num) { + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; struct mlx5dr_fw_recalc_cs_ft *recalc_cs_ft; u32 table_id, group_id, modify_hdr_id; u64 rx_icm_addr, modify_ttl_action; @@ -16,9 +17,14 @@ mlx5dr_fw_create_recalc_cs_ft(struct mlx5dr_domain *dmn, u32 vport_num) if (!recalc_cs_ft) return NULL; - ret = mlx5dr_cmd_create_flow_table(dmn->mdev, MLX5_FLOW_TABLE_TYPE_FDB, - 0, 0, dmn->info.caps.max_ft_level - 1, - false, true, &rx_icm_addr, &table_id); + ft_attr.table_type = MLX5_FLOW_TABLE_TYPE_FDB; + ft_attr.level = dmn->info.caps.max_ft_level - 1; + ft_attr.term_tbl = true; + + ret = mlx5dr_cmd_create_flow_table(dmn->mdev, + &ft_attr, + &rx_icm_addr, + &table_id); if (ret) { mlx5dr_err(dmn, "Failed creating TTL W/A FW flow table %d\n", ret); goto free_ttl_tbl; @@ -91,3 +97,70 @@ void mlx5dr_fw_destroy_recalc_cs_ft(struct mlx5dr_domain *dmn, kfree(recalc_cs_ft); } + +int mlx5dr_fw_create_md_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_cmd_flow_destination_hw_info *dest, + int num_dest, + bool reformat_req, + u32 *tbl_id, + u32 *group_id) +{ + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; + struct mlx5dr_cmd_fte_info fte_info = {}; + u32 val[MLX5_ST_SZ_DW_MATCH_PARAM] = {}; + struct mlx5dr_cmd_ft_info ft_info = {}; + int ret; + + ft_attr.table_type = MLX5_FLOW_TABLE_TYPE_FDB; + ft_attr.level = dmn->info.caps.max_ft_level - 2; + ft_attr.reformat_en = reformat_req; + ft_attr.decap_en = reformat_req; + + ret = mlx5dr_cmd_create_flow_table(dmn->mdev, &ft_attr, NULL, tbl_id); + if (ret) { + mlx5dr_err(dmn, "Failed creating multi dest FW flow table %d\n", ret); + return ret; + } + + ret = mlx5dr_cmd_create_empty_flow_group(dmn->mdev, + MLX5_FLOW_TABLE_TYPE_FDB, + *tbl_id, group_id); + if (ret) { + mlx5dr_err(dmn, "Failed creating multi dest FW flow group %d\n", ret); + goto free_flow_table; + } + + ft_info.id = *tbl_id; + ft_info.type = FS_FT_FDB; + fte_info.action.action = MLX5_FLOW_CONTEXT_ACTION_FWD_DEST; + fte_info.dests_size = num_dest; + fte_info.val = val; + fte_info.dest_arr = dest; + + ret = mlx5dr_cmd_set_fte(dmn->mdev, 0, 0, &ft_info, *group_id, &fte_info); + if (ret) { + mlx5dr_err(dmn, "Failed setting fte into table %d\n", ret); + goto free_flow_group; + } + + return 0; + +free_flow_group: + mlx5dr_cmd_destroy_flow_group(dmn->mdev, MLX5_FLOW_TABLE_TYPE_FDB, + *tbl_id, *group_id); +free_flow_table: + mlx5dr_cmd_destroy_flow_table(dmn->mdev, *tbl_id, + MLX5_FLOW_TABLE_TYPE_FDB); + return ret; +} + +void mlx5dr_fw_destroy_md_tbl(struct mlx5dr_domain *dmn, + u32 tbl_id, u32 group_id) +{ + mlx5dr_cmd_del_flow_table_entry(dmn->mdev, FS_FT_FDB, tbl_id); + mlx5dr_cmd_destroy_flow_group(dmn->mdev, + MLX5_FLOW_TABLE_TYPE_FDB, + tbl_id, group_id); + mlx5dr_cmd_destroy_flow_table(dmn->mdev, tbl_id, + MLX5_FLOW_TABLE_TYPE_FDB); +} diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c index e178d8d3dbc9..14ce2d7dbb66 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_table.c @@ -211,6 +211,9 @@ static int dr_table_destroy_sw_owned_tbl(struct mlx5dr_table *tbl) static int dr_table_create_sw_owned_tbl(struct mlx5dr_table *tbl) { + bool en_encap = !!(tbl->flags & MLX5_FLOW_TABLE_TUNNEL_EN_REFORMAT); + bool en_decap = !!(tbl->flags & MLX5_FLOW_TABLE_TUNNEL_EN_DECAP); + struct mlx5dr_cmd_create_flow_table_attr ft_attr = {}; u64 icm_addr_rx = 0; u64 icm_addr_tx = 0; int ret; @@ -221,18 +224,21 @@ static int dr_table_create_sw_owned_tbl(struct mlx5dr_table *tbl) if (tbl->tx.s_anchor) icm_addr_tx = tbl->tx.s_anchor->chunk->icm_addr; - ret = mlx5dr_cmd_create_flow_table(tbl->dmn->mdev, - tbl->table_type, - icm_addr_rx, - icm_addr_tx, - tbl->dmn->info.caps.max_ft_level - 1, - true, false, NULL, - &tbl->table_id); + ft_attr.table_type = tbl->table_type; + ft_attr.icm_addr_rx = icm_addr_rx; + ft_attr.icm_addr_tx = icm_addr_tx; + ft_attr.level = tbl->dmn->info.caps.max_ft_level - 1; + ft_attr.sw_owner = true; + ft_attr.decap_en = en_decap; + ft_attr.reformat_en = en_encap; + + ret = mlx5dr_cmd_create_flow_table(tbl->dmn->mdev, &ft_attr, + NULL, &tbl->table_id); return ret; } -struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level) +struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level, u32 flags) { struct mlx5dr_table *tbl; int ret; @@ -245,6 +251,7 @@ struct mlx5dr_table *mlx5dr_table_create(struct mlx5dr_domain *dmn, u32 level) tbl->dmn = dmn; tbl->level = level; + tbl->flags = flags; refcount_set(&tbl->refcount, 1); ret = dr_table_init(tbl); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h index 3fdf4a5eb031..dffe35145d19 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_types.h @@ -681,6 +681,7 @@ struct mlx5dr_table { u32 level; u32 table_type; u32 table_id; + u32 flags; struct list_head matcher_list; struct mlx5dr_action *miss_action; refcount_t refcount; @@ -744,10 +745,14 @@ struct mlx5dr_action { union { struct mlx5dr_table *tbl; struct { - struct mlx5_flow_table *ft; + struct mlx5dr_domain *dmn; + u32 id; + u32 group_id; + enum fs_flow_table_type type; u64 rx_icm_addr; u64 tx_icm_addr; - struct mlx5_core_dev *mdev; + struct mlx5dr_action **ref_actions; + u32 num_of_ref_actions; } fw_tbl; }; } dest_tbl; @@ -869,6 +874,17 @@ struct mlx5dr_cmd_query_flow_table_details { u64 sw_owner_icm_root_0; }; +struct mlx5dr_cmd_create_flow_table_attr { + u32 table_type; + u64 icm_addr_rx; + u64 icm_addr_tx; + u8 level; + bool sw_owner; + bool term_tbl; + bool decap_en; + bool reformat_en; +}; + /* internal API functions */ int mlx5dr_cmd_query_device(struct mlx5_core_dev *mdev, struct mlx5dr_cmd_caps *caps); @@ -906,12 +922,7 @@ int mlx5dr_cmd_destroy_flow_group(struct mlx5_core_dev *mdev, u32 table_id, u32 group_id); int mlx5dr_cmd_create_flow_table(struct mlx5_core_dev *mdev, - u32 table_type, - u64 icm_addr_rx, - u64 icm_addr_tx, - u8 level, - bool sw_owner, - bool term_tbl, + struct mlx5dr_cmd_create_flow_table_attr *attr, u64 *fdb_rx_icm_addr, u32 *table_id); int mlx5dr_cmd_destroy_flow_table(struct mlx5_core_dev *mdev, @@ -1053,6 +1064,43 @@ int mlx5dr_send_postsend_formatted_htbl(struct mlx5dr_domain *dmn, int mlx5dr_send_postsend_action(struct mlx5dr_domain *dmn, struct mlx5dr_action *action); +struct mlx5dr_cmd_ft_info { + u32 id; + u16 vport; + enum fs_flow_table_type type; +}; + +struct mlx5dr_cmd_flow_destination_hw_info { + enum mlx5_flow_destination_type type; + union { + u32 tir_num; + u32 ft_num; + u32 ft_id; + u32 counter_id; + struct { + u16 num; + u16 vhca_id; + u32 reformat_id; + u8 flags; + } vport; + }; +}; + +struct mlx5dr_cmd_fte_info { + u32 dests_size; + u32 index; + struct mlx5_flow_context flow_context; + u32 *val; + struct mlx5_flow_act action; + struct mlx5dr_cmd_flow_destination_hw_info *dest_arr; +}; + +int mlx5dr_cmd_set_fte(struct mlx5_core_dev *dev, + int opmod, int modify_mask, + struct mlx5dr_cmd_ft_info *ft, + u32 group_id, + struct mlx5dr_cmd_fte_info *fte); + struct mlx5dr_fw_recalc_cs_ft { u64 rx_icm_addr; u32 table_id; @@ -1067,4 +1115,12 @@ void mlx5dr_fw_destroy_recalc_cs_ft(struct mlx5dr_domain *dmn, int mlx5dr_domain_cache_get_recalc_cs_ft_addr(struct mlx5dr_domain *dmn, u32 vport_num, u64 *rx_icm_addr); +int mlx5dr_fw_create_md_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_cmd_flow_destination_hw_info *dest, + int num_dest, + bool reformat_req, + u32 *tbl_id, + u32 *group_id); +void mlx5dr_fw_destroy_md_tbl(struct mlx5dr_domain *dmn, u32 tbl_id, + u32 group_id); #endif /* _DR_TYPES_H_ */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c index 1e32e2443f73..3abfc8125926 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/fs_dr.c @@ -74,7 +74,7 @@ static int mlx5_cmd_dr_create_flow_table(struct mlx5_flow_root_namespace *ns, next_ft); tbl = mlx5dr_table_create(ns->fs_dr_domain.dr_domain, - ft->level); + ft->level, ft->flags); if (!tbl) { mlx5_core_err(ns->dev, "Failed creating dr flow_table\n"); return -EINVAL; @@ -184,13 +184,13 @@ static struct mlx5dr_action *create_vport_action(struct mlx5dr_domain *domain, dest_attr->vport.vhca_id); } -static struct mlx5dr_action *create_ft_action(struct mlx5_core_dev *dev, +static struct mlx5dr_action *create_ft_action(struct mlx5dr_domain *domain, struct mlx5_flow_rule *dst) { struct mlx5_flow_table *dest_ft = dst->dest_attr.ft; if (mlx5_dr_is_fw_table(dest_ft->flags)) - return mlx5dr_create_action_dest_flow_fw_table(dest_ft, dev); + return mlx5dr_action_create_dest_flow_fw_table(domain, dest_ft); return mlx5dr_action_create_dest_table(dest_ft->fs_dr_table.dr_table); } @@ -206,6 +206,12 @@ static struct mlx5dr_action *create_action_push_vlan(struct mlx5dr_domain *domai return mlx5dr_action_create_push_vlan(domain, htonl(vlan_hdr)); } +static bool contain_vport_reformat_action(struct mlx5_flow_rule *dst) +{ + return dst->dest_attr.type == MLX5_FLOW_DESTINATION_TYPE_VPORT && + dst->dest_attr.vport.flags & MLX5_FLOW_DEST_VPORT_REFORMAT_ID; +} + #define MLX5_FLOW_CONTEXT_ACTION_MAX 20 static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct mlx5_flow_table *ft, @@ -213,7 +219,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct fs_fte *fte) { struct mlx5dr_domain *domain = ns->fs_dr_domain.dr_domain; - struct mlx5dr_action *term_action = NULL; + struct mlx5dr_action_dest *term_actions; struct mlx5dr_match_parameters params; struct mlx5_core_dev *dev = ns->dev; struct mlx5dr_action **fs_dr_actions; @@ -223,6 +229,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, struct mlx5dr_rule *rule; struct mlx5_flow_rule *dst; int fs_dr_num_actions = 0; + int num_term_actions = 0; int num_actions = 0; size_t match_sz; int err = 0; @@ -233,18 +240,38 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, sizeof(*actions), GFP_KERNEL); - if (!actions) - return -ENOMEM; + if (!actions) { + err = -ENOMEM; + goto out_err; + } fs_dr_actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, sizeof(*fs_dr_actions), GFP_KERNEL); if (!fs_dr_actions) { - kfree(actions); - return -ENOMEM; + err = -ENOMEM; + goto free_actions_alloc; + } + + term_actions = kcalloc(MLX5_FLOW_CONTEXT_ACTION_MAX, + sizeof(*term_actions), GFP_KERNEL); + if (!term_actions) { + err = -ENOMEM; + goto free_fs_dr_actions_alloc; } match_sz = sizeof(fte->val); + /* Drop reformat action bit if destination vport set with reformat */ + if (fte->action.action & MLX5_FLOW_CONTEXT_ACTION_FWD_DEST) { + list_for_each_entry(dst, &fte->node.children, node.list) { + if (!contain_vport_reformat_action(dst)) + continue; + + fte->action.action &= ~MLX5_FLOW_CONTEXT_ACTION_PACKET_REFORMAT; + break; + } + } + /* The order of the actions are must to be keep, only the following * order is supported by SW steering: * TX: push vlan -> modify header -> encap @@ -335,7 +362,7 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions++].dest = tmp_action; } if (fte->flow_context.flow_tag) { @@ -353,7 +380,8 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, list_for_each_entry(dst, &fte->node.children, node.list) { enum mlx5_flow_destination_type type = dst->dest_attr.type; - if (num_actions == MLX5_FLOW_CONTEXT_ACTION_MAX) { + if (num_actions == MLX5_FLOW_CONTEXT_ACTION_MAX || + num_term_actions >= MLX5_FLOW_CONTEXT_ACTION_MAX) { err = -ENOSPC; goto free_actions; } @@ -363,13 +391,13 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, switch (type) { case MLX5_FLOW_DESTINATION_TYPE_FLOW_TABLE: - tmp_action = create_ft_action(dev, dst); + tmp_action = create_ft_action(domain, dst); if (!tmp_action) { err = -ENOMEM; goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions++].dest = tmp_action; break; case MLX5_FLOW_DESTINATION_TYPE_VPORT: tmp_action = create_vport_action(domain, dst); @@ -378,7 +406,14 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } fs_dr_actions[fs_dr_num_actions++] = tmp_action; - term_action = tmp_action; + term_actions[num_term_actions].dest = tmp_action; + + if (dst->dest_attr.vport.flags & + MLX5_FLOW_DEST_VPORT_REFORMAT_ID) + term_actions[num_term_actions].reformat = + dst->dest_attr.vport.pkt_reformat->action.dr_action; + + num_term_actions++; break; default: err = -EOPNOTSUPP; @@ -415,9 +450,22 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, params.match_sz = match_sz; params.match_buf = (u64 *)fte->val; - - if (term_action) - actions[num_actions++] = term_action; + if (num_term_actions == 1) { + if (term_actions->reformat) + actions[num_actions++] = term_actions->reformat; + + actions[num_actions++] = term_actions->dest; + } else if (num_term_actions > 1) { + tmp_action = mlx5dr_action_create_mult_dest_tbl(domain, + term_actions, + num_term_actions); + if (!tmp_action) { + err = -EOPNOTSUPP; + goto free_actions; + } + fs_dr_actions[fs_dr_num_actions++] = tmp_action; + actions[num_actions++] = tmp_action; + } rule = mlx5dr_rule_create(group->fs_dr_matcher.dr_matcher, ¶ms, @@ -428,7 +476,9 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, goto free_actions; } + kfree(term_actions); kfree(actions); + fte->fs_dr_rule.dr_rule = rule; fte->fs_dr_rule.num_actions = fs_dr_num_actions; fte->fs_dr_rule.dr_actions = fs_dr_actions; @@ -436,13 +486,18 @@ static int mlx5_cmd_dr_create_fte(struct mlx5_flow_root_namespace *ns, return 0; free_actions: - for (i = 0; i < fs_dr_num_actions; i++) + /* Free in reverse order to handle action dependencies */ + for (i = fs_dr_num_actions - 1; i >= 0; i--) if (!IS_ERR_OR_NULL(fs_dr_actions[i])) mlx5dr_action_destroy(fs_dr_actions[i]); - mlx5_core_err(dev, "Failed to create dr rule err(%d)\n", err); - kfree(actions); + kfree(term_actions); +free_fs_dr_actions_alloc: kfree(fs_dr_actions); +free_actions_alloc: + kfree(actions); +out_err: + mlx5_core_err(dev, "Failed to create dr rule err(%d)\n", err); return err; } @@ -549,7 +604,8 @@ static int mlx5_cmd_dr_delete_fte(struct mlx5_flow_root_namespace *ns, if (err) return err; - for (i = 0; i < rule->num_actions; i++) + /* Free in reverse order to handle action dependencies */ + for (i = rule->num_actions - 1; i >= 0; i--) if (!IS_ERR_OR_NULL(rule->dr_actions[i])) mlx5dr_action_destroy(rule->dr_actions[i]); diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5_ifc_dr.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5_ifc_dr.h index 1722f4668269..e01c3766c7de 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5_ifc_dr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5_ifc_dr.h @@ -32,6 +32,7 @@ enum { }; enum { + MLX5DR_ACTION_MDFY_HW_OP_COPY = 0x1, MLX5DR_ACTION_MDFY_HW_OP_SET = 0x2, MLX5DR_ACTION_MDFY_HW_OP_ADD = 0x3, }; @@ -625,4 +626,19 @@ struct mlx5_ifc_dr_action_hw_set_bits { u8 inline_data[0x20]; }; +struct mlx5_ifc_dr_action_hw_copy_bits { + u8 opcode[0x8]; + u8 destination_field_code[0x8]; + u8 reserved_at_10[0x2]; + u8 destination_left_shifter[0x6]; + u8 reserved_at_18[0x2]; + u8 destination_length[0x6]; + + u8 reserved_at_20[0x8]; + u8 source_field_code[0x8]; + u8 reserved_at_30[0x2]; + u8 source_left_shifter[0x6]; + u8 reserved_at_38[0x8]; +}; + #endif /* MLX5_IFC_DR_H */ diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h index adda9cbfba45..e1edc9c247b7 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/mlx5dr.h @@ -33,6 +33,11 @@ struct mlx5dr_match_parameters { u64 *match_buf; /* Device spec format */ }; +struct mlx5dr_action_dest { + struct mlx5dr_action *dest; + struct mlx5dr_action *reformat; +}; + #ifdef CONFIG_MLX5_SW_STEERING struct mlx5dr_domain * @@ -46,7 +51,7 @@ void mlx5dr_domain_set_peer(struct mlx5dr_domain *dmn, struct mlx5dr_domain *peer_dmn); struct mlx5dr_table * -mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level); +mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level, u32 flags); int mlx5dr_table_destroy(struct mlx5dr_table *table); @@ -75,14 +80,19 @@ struct mlx5dr_action * mlx5dr_action_create_dest_table(struct mlx5dr_table *table); struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev); +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *domain, + struct mlx5_flow_table *ft); struct mlx5dr_action * mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, u32 vport, u8 vhca_id_valid, u16 vhca_id); +struct mlx5dr_action * +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests); + struct mlx5dr_action *mlx5dr_action_create_drop(void); struct mlx5dr_action *mlx5dr_action_create_tag(u32 tag_value); @@ -131,7 +141,7 @@ mlx5dr_domain_set_peer(struct mlx5dr_domain *dmn, struct mlx5dr_domain *peer_dmn) { } static inline struct mlx5dr_table * -mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level) { return NULL; } +mlx5dr_table_create(struct mlx5dr_domain *domain, u32 level, u32 flags) { return NULL; } static inline int mlx5dr_table_destroy(struct mlx5dr_table *table) { return 0; } @@ -165,8 +175,8 @@ static inline struct mlx5dr_action * mlx5dr_action_create_dest_table(struct mlx5dr_table *table) { return NULL; } static inline struct mlx5dr_action * -mlx5dr_create_action_dest_flow_fw_table(struct mlx5_flow_table *ft, - struct mlx5_core_dev *mdev) { return NULL; } +mlx5dr_action_create_dest_flow_fw_table(struct mlx5dr_domain *domain, + struct mlx5_flow_table *ft) { return NULL; } static inline struct mlx5dr_action * mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, @@ -174,6 +184,11 @@ mlx5dr_action_create_dest_vport(struct mlx5dr_domain *domain, u16 vhca_id) { return NULL; } static inline struct mlx5dr_action * +mlx5dr_action_create_mult_dest_tbl(struct mlx5dr_domain *dmn, + struct mlx5dr_action_dest *dests, + u32 num_of_dests) { return NULL; } + +static inline struct mlx5dr_action * mlx5dr_action_create_drop(void) { return NULL; } static inline struct mlx5dr_action * diff --git a/drivers/net/ethernet/mellanox/mlx5/core/wq.c b/drivers/net/ethernet/mellanox/mlx5/core/wq.c index f2a0e72285ba..02f7e4a39578 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/wq.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/wq.c @@ -89,7 +89,7 @@ void mlx5_wq_cyc_wqe_dump(struct mlx5_wq_cyc *wq, u16 ix, u8 nstrides) len = nstrides << wq->fbc.log_stride; wqe = mlx5_wq_cyc_get_wqe(wq, ix); - pr_info("WQE DUMP: WQ size %d WQ cur size %d, WQE index 0x%x, len: %ld\n", + pr_info("WQE DUMP: WQ size %d WQ cur size %d, WQE index 0x%x, len: %zu\n", mlx5_wq_cyc_get_size(wq), wq->cur_sz, ix, len); print_hex_dump(KERN_WARNING, "", DUMP_PREFIX_OFFSET, 16, 1, wqe, len, false); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/minimal.c b/drivers/net/ethernet/mellanox/mlxsw/minimal.c index 2b543911ae00..c4caeeadcba9 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/minimal.c +++ b/drivers/net/ethernet/mellanox/mlxsw/minimal.c @@ -213,8 +213,8 @@ mlxsw_m_port_create(struct mlxsw_m *mlxsw_m, u8 local_port, u8 module) err_register_netdev: mlxsw_m->ports[local_port] = NULL; - free_netdev(dev); err_dev_addr_get: + free_netdev(dev); err_alloc_etherdev: mlxsw_core_port_fini(mlxsw_m->core, local_port); return err; diff --git a/drivers/net/ethernet/mellanox/mlxsw/reg.h b/drivers/net/ethernet/mellanox/mlxsw/reg.h index af30e8a76682..dd6685156396 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/reg.h +++ b/drivers/net/ethernet/mellanox/mlxsw/reg.h @@ -3477,10 +3477,10 @@ MLXSW_REG_DEFINE(qeec, MLXSW_REG_QEEC_ID, MLXSW_REG_QEEC_LEN); MLXSW_ITEM32(reg, qeec, local_port, 0x00, 16, 8); enum mlxsw_reg_qeec_hr { - MLXSW_REG_QEEC_HIERARCY_PORT, - MLXSW_REG_QEEC_HIERARCY_GROUP, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, - MLXSW_REG_QEEC_HIERARCY_TC, + MLXSW_REG_QEEC_HR_PORT, + MLXSW_REG_QEEC_HR_GROUP, + MLXSW_REG_QEEC_HR_SUBGROUP, + MLXSW_REG_QEEC_HR_TC, }; /* reg_qeec_element_hierarchy @@ -3563,8 +3563,8 @@ MLXSW_ITEM32(reg, qeec, min_shaper_rate, 0x0C, 0, 28); */ MLXSW_ITEM32(reg, qeec, mase, 0x10, 31, 1); -/* A large max rate will disable the max shaper. */ -#define MLXSW_REG_QEEC_MAS_DIS 200000000 /* Kbps */ +/* The largest max shaper value possible to disable the shaper. */ +#define MLXSW_REG_QEEC_MAS_DIS ((1u << 31) - 1) /* Kbps */ /* reg_qeec_max_shaper_rate * Max shaper information rate. @@ -3602,6 +3602,21 @@ MLXSW_ITEM32(reg, qeec, dwrr, 0x18, 15, 1); */ MLXSW_ITEM32(reg, qeec, dwrr_weight, 0x18, 0, 8); +/* reg_qeec_max_shaper_bs + * Max shaper burst size + * Burst size is 2^max_shaper_bs * 512 bits + * For Spectrum-1: Range is: 5..25 + * For Spectrum-2: Range is: 11..25 + * Reserved when ptps = 1 + * Access: RW + */ +MLXSW_ITEM32(reg, qeec, max_shaper_bs, 0x1C, 0, 6); + +#define MLXSW_REG_QEEC_HIGHEST_SHAPER_BS 25 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP1 5 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP2 11 +#define MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP3 5 + static inline void mlxsw_reg_qeec_pack(char *payload, u8 local_port, enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index) @@ -3618,8 +3633,7 @@ static inline void mlxsw_reg_qeec_ptps_pack(char *payload, u8 local_port, { MLXSW_REG_ZERO(qeec, payload); mlxsw_reg_qeec_local_port_set(payload, local_port); - mlxsw_reg_qeec_element_hierarchy_set(payload, - MLXSW_REG_QEEC_HIERARCY_PORT); + mlxsw_reg_qeec_element_hierarchy_set(payload, MLXSW_REG_QEEC_HR_PORT); mlxsw_reg_qeec_ptps_set(payload, ptps); } @@ -3749,6 +3763,38 @@ mlxsw_reg_qpdsm_prio_pack(char *payload, unsigned short prio, u8 dscp) mlxsw_reg_qpdsm_prio_entry_color2_dscp_set(payload, prio, dscp); } +/* QPDP - QoS Port DSCP to Priority Mapping Register + * ------------------------------------------------- + * This register controls the port default Switch Priority and Color. The + * default Switch Priority and Color are used for frames where the trust state + * uses default values. All member ports of a LAG should be configured with the + * same default values. + */ +#define MLXSW_REG_QPDP_ID 0x4007 +#define MLXSW_REG_QPDP_LEN 0x8 + +MLXSW_REG_DEFINE(qpdp, MLXSW_REG_QPDP_ID, MLXSW_REG_QPDP_LEN); + +/* reg_qpdp_local_port + * Local Port. Supported for data packets from CPU port. + * Access: Index + */ +MLXSW_ITEM32(reg, qpdp, local_port, 0x00, 16, 8); + +/* reg_qpdp_switch_prio + * Default port Switch Priority (default 0) + * Access: RW + */ +MLXSW_ITEM32(reg, qpdp, switch_prio, 0x04, 0, 4); + +static inline void mlxsw_reg_qpdp_pack(char *payload, u8 local_port, + u8 switch_prio) +{ + MLXSW_REG_ZERO(qpdp, payload); + mlxsw_reg_qpdp_local_port_set(payload, local_port); + mlxsw_reg_qpdp_switch_prio_set(payload, switch_prio); +} + /* QPDPM - QoS Port DSCP to Priority Mapping Register * -------------------------------------------------- * This register controls the mapping from DSCP field to @@ -5482,6 +5528,7 @@ enum mlxsw_reg_htgt_discard_trap_group { MLXSW_REG_HTGT_DISCARD_TRAP_GROUP_BASE = MLXSW_REG_HTGT_TRAP_GROUP_MAX, MLXSW_REG_HTGT_TRAP_GROUP_SP_L2_DISCARDS, MLXSW_REG_HTGT_TRAP_GROUP_SP_L3_DISCARDS, + MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS, }; /* reg_htgt_trap_group @@ -10109,6 +10156,92 @@ static inline void mlxsw_reg_tigcr_pack(char *payload, bool ttlc, u8 ttl_uc) mlxsw_reg_tigcr_ttl_uc_set(payload, ttl_uc); } +/* TIEEM - Tunneling IPinIP Encapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIEEM register maps ECN of the IP header at the ingress to the + * encapsulation to the ECN of the underlay network. + */ +#define MLXSW_REG_TIEEM_ID 0xA812 +#define MLXSW_REG_TIEEM_LEN 0x0C + +MLXSW_REG_DEFINE(tieem, MLXSW_REG_TIEEM_ID, MLXSW_REG_TIEEM_LEN); + +/* reg_tieem_overlay_ecn + * ECN of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tieem, overlay_ecn, 0x04, 24, 2); + +/* reg_tineem_underlay_ecn + * ECN of the IP header in the underlay network. + * Access: RW + */ +MLXSW_ITEM32(reg, tieem, underlay_ecn, 0x04, 16, 2); + +static inline void mlxsw_reg_tieem_pack(char *payload, u8 overlay_ecn, + u8 underlay_ecn) +{ + MLXSW_REG_ZERO(tieem, payload); + mlxsw_reg_tieem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tieem_underlay_ecn_set(payload, underlay_ecn); +} + +/* TIDEM - Tunneling IPinIP Decapsulation ECN Mapping Register + * ----------------------------------------------------------- + * The TIDEM register configures the actions that are done in the + * decapsulation. + */ +#define MLXSW_REG_TIDEM_ID 0xA813 +#define MLXSW_REG_TIDEM_LEN 0x0C + +MLXSW_REG_DEFINE(tidem, MLXSW_REG_TIDEM_ID, MLXSW_REG_TIDEM_LEN); + +/* reg_tidem_underlay_ecn + * ECN field of the IP header in the underlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, underlay_ecn, 0x04, 24, 2); + +/* reg_tidem_overlay_ecn + * ECN field of the IP header in the overlay network. + * Access: Index + */ +MLXSW_ITEM32(reg, tidem, overlay_ecn, 0x04, 16, 2); + +/* reg_tidem_eip_ecn + * Egress IP ECN. ECN field of the IP header of the packet which goes out + * from the decapsulation. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, eip_ecn, 0x04, 8, 2); + +/* reg_tidem_trap_en + * Trap enable: + * 0 - No trap due to decap ECN + * 1 - Trap enable with trap_id + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_en, 0x08, 28, 4); + +/* reg_tidem_trap_id + * Trap ID. Either DECAP_ECN0 or DECAP_ECN1. + * Reserved when trap_en is '0'. + * Access: RW + */ +MLXSW_ITEM32(reg, tidem, trap_id, 0x08, 0, 9); + +static inline void mlxsw_reg_tidem_pack(char *payload, u8 underlay_ecn, + u8 overlay_ecn, u8 eip_ecn, + bool trap_en, u16 trap_id) +{ + MLXSW_REG_ZERO(tidem, payload); + mlxsw_reg_tidem_underlay_ecn_set(payload, underlay_ecn); + mlxsw_reg_tidem_overlay_ecn_set(payload, overlay_ecn); + mlxsw_reg_tidem_eip_ecn_set(payload, eip_ecn); + mlxsw_reg_tidem_trap_en_set(payload, trap_en); + mlxsw_reg_tidem_trap_id_set(payload, trap_id); +} + /* SBPR - Shared Buffer Pools Register * ----------------------------------- * The SBPR configures and retrieves the shared buffer pools and configuration. @@ -10581,6 +10714,7 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = { MLXSW_REG(qeec), MLXSW_REG(qrwe), MLXSW_REG(qpdsm), + MLXSW_REG(qpdp), MLXSW_REG(qpdpm), MLXSW_REG(qtctm), MLXSW_REG(qpsc), @@ -10652,6 +10786,8 @@ static const struct mlxsw_reg_info *mlxsw_reg_infos[] = { MLXSW_REG(tndem), MLXSW_REG(tnpc), MLXSW_REG(tigcr), + MLXSW_REG(tieem), + MLXSW_REG(tidem), MLXSW_REG(sbpr), MLXSW_REG(sbcm), MLXSW_REG(sbpm), diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 8ed15199eb4f..7358b5bc7eb6 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -45,11 +45,9 @@ #include "spectrum_ptp.h" #include "../mlxfw/mlxfw.h" -#define MLXSW_SP_FWREV_MINOR_TO_BRANCH(minor) ((minor) / 100) - #define MLXSW_SP1_FWREV_MAJOR 13 #define MLXSW_SP1_FWREV_MINOR 2000 -#define MLXSW_SP1_FWREV_SUBMINOR 2308 +#define MLXSW_SP1_FWREV_SUBMINOR 2714 #define MLXSW_SP1_FWREV_CAN_RESET_MINOR 1702 static const struct mlxsw_fw_rev mlxsw_sp1_fw_rev = { @@ -66,7 +64,7 @@ static const struct mlxsw_fw_rev mlxsw_sp1_fw_rev = { #define MLXSW_SP2_FWREV_MAJOR 29 #define MLXSW_SP2_FWREV_MINOR 2000 -#define MLXSW_SP2_FWREV_SUBMINOR 2308 +#define MLXSW_SP2_FWREV_SUBMINOR 2714 static const struct mlxsw_fw_rev mlxsw_sp2_fw_rev = { .major = MLXSW_SP2_FWREV_MAJOR, @@ -197,6 +195,10 @@ struct mlxsw_sp_ptp_ops { u64 *data, int data_index); }; +struct mlxsw_sp_span_ops { + u32 (*buffsize_get)(int mtu, u32 speed); +}; + static int mlxsw_sp_component_query(struct mlxfw_dev *mlxfw_dev, u16 component_index, u32 *p_max_size, u8 *p_align_bits, u16 *p_max_write_size) @@ -423,13 +425,12 @@ static int mlxsw_sp_fw_rev_validate(struct mlxsw_sp *mlxsw_sp) rev->major, req_rev->major); return -EINVAL; } - if (MLXSW_SP_FWREV_MINOR_TO_BRANCH(rev->minor) == - MLXSW_SP_FWREV_MINOR_TO_BRANCH(req_rev->minor) && - mlxsw_core_fw_rev_minor_subminor_validate(rev, req_rev)) + if (mlxsw_core_fw_rev_minor_subminor_validate(rev, req_rev)) return 0; - dev_info(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver\n", - rev->major, rev->minor, rev->subminor); + dev_err(mlxsw_sp->bus_info->dev, "The firmware version %d.%d.%d is incompatible with the driver (required >= %d.%d.%d)\n", + rev->major, rev->minor, rev->subminor, req_rev->major, + req_rev->minor, req_rev->subminor); dev_info(mlxsw_sp->bus_info->dev, "Flashing firmware using file %s\n", fw_filename); @@ -1793,6 +1794,10 @@ static int mlxsw_sp_setup_tc(struct net_device *dev, enum tc_setup_type type, return mlxsw_sp_setup_tc_red(mlxsw_sp_port, type_data); case TC_SETUP_QDISC_PRIO: return mlxsw_sp_setup_tc_prio(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_ETS: + return mlxsw_sp_setup_tc_ets(mlxsw_sp_port, type_data); + case TC_SETUP_QDISC_TBF: + return mlxsw_sp_setup_tc_tbf(mlxsw_sp_port, type_data); default: return -EOPNOTSUPP; } @@ -3536,6 +3541,27 @@ mlxsw_sp_port_speed_by_width_set(struct mlxsw_sp_port *mlxsw_sp_port) return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); } +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed) +{ + const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char ptys_pl[MLXSW_REG_PTYS_LEN]; + u32 eth_proto_oper; + int err; + + port_type_speed_ops = mlxsw_sp->port_type_speed_ops; + port_type_speed_ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, + mlxsw_sp_port->local_port, 0, + false); + err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + if (err) + return err; + port_type_speed_ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, NULL, NULL, + ð_proto_oper); + *speed = port_type_speed_ops->from_ptys_speed(mlxsw_sp, eth_proto_oper); + return 0; +} + int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, bool dwrr, u8 dwrr_weight) @@ -3553,7 +3579,7 @@ int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, int mlxsw_sp_port_ets_maxrate_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, - u8 next_index, u32 maxrate) + u8 next_index, u32 maxrate, u8 burst_size) { struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; char qeec_pl[MLXSW_REG_QEEC_LEN]; @@ -3562,6 +3588,7 @@ int mlxsw_sp_port_ets_maxrate_set(struct mlxsw_sp_port *mlxsw_sp_port, next_index); mlxsw_reg_qeec_mase_set(qeec_pl, true); mlxsw_reg_qeec_max_shaper_rate_set(qeec_pl, maxrate); + mlxsw_reg_qeec_max_shaper_bs_set(qeec_pl, burst_size); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qeec), qeec_pl); } @@ -3599,26 +3626,25 @@ static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) * one subgroup, which are all member in the same group. */ err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_GROUP, 0, 0, false, - 0); + MLXSW_REG_QEEC_HR_GROUP, 0, 0, false, 0); if (err) return err; for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, i, + MLXSW_REG_QEEC_HR_SUBGROUP, i, 0, false, 0); if (err) return err; } for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_TC, i, i, + MLXSW_REG_QEEC_HR_TC, i, i, false, 0); if (err) return err; err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_TC, + MLXSW_REG_QEEC_HR_TC, i + 8, i, true, 100); if (err) @@ -3630,30 +3656,30 @@ static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) * for the initial configuration. */ err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_PORT, 0, 0, - MLXSW_REG_QEEC_MAS_DIS); + MLXSW_REG_QEEC_HR_PORT, 0, 0, + MLXSW_REG_QEEC_MAS_DIS, 0); if (err) return err; for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, + MLXSW_REG_QEEC_HR_SUBGROUP, i, 0, - MLXSW_REG_QEEC_MAS_DIS); + MLXSW_REG_QEEC_MAS_DIS, 0); if (err) return err; } for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_TC, + MLXSW_REG_QEEC_HR_TC, i, i, - MLXSW_REG_QEEC_MAS_DIS); + MLXSW_REG_QEEC_MAS_DIS, 0); if (err) return err; err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_TC, + MLXSW_REG_QEEC_HR_TC, i + 8, i, - MLXSW_REG_QEEC_MAS_DIS); + MLXSW_REG_QEEC_MAS_DIS, 0); if (err) return err; } @@ -3661,7 +3687,7 @@ static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) /* Configure the min shaper for multicast TCs. */ for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_min_bw_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_TC, + MLXSW_REG_QEEC_HR_TC, i + 8, i, MLXSW_REG_QEEC_MIS_MIN); if (err) @@ -3885,6 +3911,8 @@ static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, INIT_DELAYED_WORK(&mlxsw_sp_port->ptp.shaper_dw, mlxsw_sp->ptp_ops->shaper_work); + INIT_DELAYED_WORK(&mlxsw_sp_port->span.speed_update_dw, + mlxsw_sp_span_speed_update_work); mlxsw_sp->ports[local_port] = mlxsw_sp_port; err = register_netdev(dev); @@ -3941,6 +3969,7 @@ static void mlxsw_sp_port_remove(struct mlxsw_sp *mlxsw_sp, u8 local_port) struct mlxsw_sp_port *mlxsw_sp_port = mlxsw_sp->ports[local_port]; cancel_delayed_work_sync(&mlxsw_sp_port->periodic_hw_stats.update_dw); + cancel_delayed_work_sync(&mlxsw_sp_port->span.speed_update_dw); cancel_delayed_work_sync(&mlxsw_sp_port->ptp.shaper_dw); mlxsw_sp_port_ptp_clear(mlxsw_sp_port); mlxsw_core_port_clear(mlxsw_sp->core, local_port, mlxsw_sp); @@ -4348,6 +4377,7 @@ static void mlxsw_sp_pude_event_func(const struct mlxsw_reg_info *reg, netdev_info(mlxsw_sp_port->dev, "link up\n"); netif_carrier_on(mlxsw_sp_port->dev); mlxsw_core_schedule_dw(&mlxsw_sp_port->ptp.shaper_dw, 0); + mlxsw_core_schedule_dw(&mlxsw_sp_port->span.speed_update_dw, 0); } else { netdev_info(mlxsw_sp_port->dev, "link down\n"); netif_carrier_off(mlxsw_sp_port->dev); @@ -4547,10 +4577,16 @@ static const struct mlxsw_listener mlxsw_sp_listener[] = { false), MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV4, TRAP_TO_CPU, ROUTER_EXP, false), MLXSW_SP_RXL_MARK(ROUTER_ALERT_IPV6, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_MARK(IPIP_DECAP_ERROR, TRAP_TO_CPU, ROUTER_EXP, false), - MLXSW_SP_RXL_MARK(DECAP_ECN0, TRAP_TO_CPU, ROUTER_EXP, false), MLXSW_SP_RXL_MARK(IPV4_VRRP, TRAP_TO_CPU, VRRP, false), MLXSW_SP_RXL_MARK(IPV6_VRRP, TRAP_TO_CPU, VRRP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_SIP_CLASS_E, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_MC_DMAC, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_SIP_DIP, FORWARD, + ROUTER_EXP, false), + MLXSW_SP_RXL_NO_MARK(DISCARD_ING_ROUTER_DIP_LINK_LOCAL, FORWARD, + ROUTER_EXP, false), /* PKT Sample trap */ MLXSW_RXL(mlxsw_sp_rx_listener_sample_func, PKT_SAMPLE, MIRROR_TO_CPU, false, SP_IP2ME, DISCARD), @@ -4889,6 +4925,33 @@ static const struct mlxsw_sp_ptp_ops mlxsw_sp2_ptp_ops = { .get_stats = mlxsw_sp2_get_stats, }; +static u32 mlxsw_sp1_span_buffsize_get(int mtu, u32 speed) +{ + return mtu * 5 / 2; +} + +static const struct mlxsw_sp_span_ops mlxsw_sp1_span_ops = { + .buffsize_get = mlxsw_sp1_span_buffsize_get, +}; + +#define MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR 38 + +static u32 mlxsw_sp2_span_buffsize_get(int mtu, u32 speed) +{ + return 3 * mtu + MLXSW_SP2_SPAN_EG_MIRROR_BUFFER_FACTOR * speed / 1000; +} + +static const struct mlxsw_sp_span_ops mlxsw_sp2_span_ops = { + .buffsize_get = mlxsw_sp2_span_buffsize_get, +}; + +u32 mlxsw_sp_span_buffsize_get(struct mlxsw_sp *mlxsw_sp, int mtu, u32 speed) +{ + u32 buffsize = mlxsw_sp->span_ops->buffsize_get(speed, mtu); + + return mlxsw_sp_bytes_cells(mlxsw_sp, buffsize) + 1; +} + static int mlxsw_sp_netdevice_event(struct notifier_block *unused, unsigned long event, void *ptr); @@ -5110,8 +5173,10 @@ static int mlxsw_sp1_init(struct mlxsw_core *mlxsw_core, mlxsw_sp->sb_vals = &mlxsw_sp1_sb_vals; mlxsw_sp->port_type_speed_ops = &mlxsw_sp1_port_type_speed_ops; mlxsw_sp->ptp_ops = &mlxsw_sp1_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp1_span_ops; mlxsw_sp->listeners = mlxsw_sp1_listener; mlxsw_sp->listeners_count = ARRAY_SIZE(mlxsw_sp1_listener); + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP1; return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); } @@ -5135,6 +5200,8 @@ static int mlxsw_sp2_init(struct mlxsw_core *mlxsw_core, mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp2_span_ops; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP2; return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); } @@ -5156,6 +5223,8 @@ static int mlxsw_sp3_init(struct mlxsw_core *mlxsw_core, mlxsw_sp->sb_vals = &mlxsw_sp2_sb_vals; mlxsw_sp->port_type_speed_ops = &mlxsw_sp2_port_type_speed_ops; mlxsw_sp->ptp_ops = &mlxsw_sp2_ptp_ops; + mlxsw_sp->span_ops = &mlxsw_sp2_span_ops; + mlxsw_sp->lowest_shaper_bs = MLXSW_REG_QEEC_LOWEST_SHAPER_BS_SP3; return mlxsw_sp_init(mlxsw_core, mlxsw_bus_info, extack); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index 347bec9d1ecf..a0f1f9dceec5 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -140,6 +140,7 @@ struct mlxsw_sp_sb_vals; struct mlxsw_sp_port_type_speed_ops; struct mlxsw_sp_ptp_state; struct mlxsw_sp_ptp_ops; +struct mlxsw_sp_span_ops; struct mlxsw_sp_port_mapping { u8 module; @@ -185,8 +186,10 @@ struct mlxsw_sp { const struct mlxsw_sp_sb_vals *sb_vals; const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; const struct mlxsw_sp_ptp_ops *ptp_ops; + const struct mlxsw_sp_span_ops *span_ops; const struct mlxsw_listener *listeners; size_t listeners_count; + u32 lowest_shaper_bs; }; static inline struct mlxsw_sp_upper * @@ -292,6 +295,9 @@ struct mlxsw_sp_port { struct mlxsw_sp_ptp_port_stats stats; } ptp; u8 split_base_local_port; + struct { + struct delayed_work speed_update_dw; + } span; }; struct mlxsw_sp_port_type_speed_ops { @@ -471,6 +477,7 @@ extern struct notifier_block mlxsw_sp_switchdev_notifier; /* spectrum.c */ void mlxsw_sp_rx_listener_no_mark_func(struct sk_buff *skb, u8 local_port, void *priv); +int mlxsw_sp_port_speed_get(struct mlxsw_sp_port *mlxsw_sp_port, u32 *speed); int mlxsw_sp_port_ets_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, u8 next_index, bool dwrr, u8 dwrr_weight); @@ -481,7 +488,7 @@ int __mlxsw_sp_port_headroom_set(struct mlxsw_sp_port *mlxsw_sp_port, int mtu, struct ieee_pfc *my_pfc); int mlxsw_sp_port_ets_maxrate_set(struct mlxsw_sp_port *mlxsw_sp_port, enum mlxsw_reg_qeec_hr hr, u8 index, - u8 next_index, u32 maxrate); + u8 next_index, u32 maxrate, u8 burst_size); enum mlxsw_reg_spms_state mlxsw_sp_stp_spms_state(u8 stp_state); int mlxsw_sp_port_vid_stp_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 vid, u8 state); @@ -501,6 +508,7 @@ int mlxsw_sp_flow_counter_alloc(struct mlxsw_sp *mlxsw_sp, unsigned int *p_counter_index); void mlxsw_sp_flow_counter_free(struct mlxsw_sp *mlxsw_sp, unsigned int counter_index); +u32 mlxsw_sp_span_buffsize_get(struct mlxsw_sp *mlxsw_sp, int mtu, u32 speed); bool mlxsw_sp_port_dev_check(const struct net_device *dev); struct mlxsw_sp *mlxsw_sp_lower_get(struct net_device *dev); struct mlxsw_sp_port *mlxsw_sp_port_dev_lower_find(struct net_device *dev); @@ -852,6 +860,10 @@ int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_red_qopt_offload *p); int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_prio_qopt_offload *p); +int mlxsw_sp_setup_tc_ets(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_ets_qopt_offload *p); +int mlxsw_sp_setup_tc_tbf(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_tbf_qopt_offload *p); /* spectrum_fid.c */ bool mlxsw_sp_fid_is_dummy(struct mlxsw_sp *mlxsw_sp, u16 fid_index); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c index 21296fa7f7fb..49a72a8f1f57 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c @@ -160,7 +160,7 @@ static int __mlxsw_sp_dcbnl_ieee_setets(struct mlxsw_sp_port *mlxsw_sp_port, u8 weight = ets->tc_tx_bw[i]; err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, i, + MLXSW_REG_QEEC_HR_SUBGROUP, i, 0, dwrr, weight); if (err) { netdev_err(dev, "Failed to link subgroup ETS element %d to group\n", @@ -198,7 +198,7 @@ err_port_ets_set: u8 weight = my_ets->tc_tx_bw[i]; err = mlxsw_sp_port_ets_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, i, + MLXSW_REG_QEEC_HR_SUBGROUP, i, 0, dwrr, weight); } return err; @@ -369,6 +369,17 @@ err_update_qrwe: } static int +mlxsw_sp_port_dcb_app_update_qpdp(struct mlxsw_sp_port *mlxsw_sp_port, + u8 default_prio) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char qpdp_pl[MLXSW_REG_QPDP_LEN]; + + mlxsw_reg_qpdp_pack(qpdp_pl, mlxsw_sp_port->local_port, default_prio); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(qpdp), qpdp_pl); +} + +static int mlxsw_sp_port_dcb_app_update_qpdpm(struct mlxsw_sp_port *mlxsw_sp_port, struct dcb_ieee_app_dscp_map *map) { @@ -405,6 +416,12 @@ static int mlxsw_sp_port_dcb_app_update(struct mlxsw_sp_port *mlxsw_sp_port) int err; default_prio = mlxsw_sp_port_dcb_app_default_prio(mlxsw_sp_port); + err = mlxsw_sp_port_dcb_app_update_qpdp(mlxsw_sp_port, default_prio); + if (err) { + netdev_err(mlxsw_sp_port->dev, "Couldn't configure port default priority\n"); + return err; + } + have_dscp = mlxsw_sp_port_dcb_app_prio_dscp_map(mlxsw_sp_port, &prio_map); @@ -507,9 +524,9 @@ static int mlxsw_sp_dcbnl_ieee_setmaxrate(struct net_device *dev, for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { err = mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, + MLXSW_REG_QEEC_HR_SUBGROUP, i, 0, - maxrate->tc_maxrate[i]); + maxrate->tc_maxrate[i], 0); if (err) { netdev_err(dev, "Failed to set maxrate for TC %d\n", i); goto err_port_ets_maxrate_set; @@ -523,8 +540,9 @@ static int mlxsw_sp_dcbnl_ieee_setmaxrate(struct net_device *dev, err_port_ets_maxrate_set: for (i--; i >= 0; i--) mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, - MLXSW_REG_QEEC_HIERARCY_SUBGROUP, - i, 0, my_maxrate->tc_maxrate[i]); + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, + my_maxrate->tc_maxrate[i], 0); return err; } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c index 6400cd644b7a..a8525992528f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ipip.c @@ -3,8 +3,10 @@ #include <net/ip_tunnels.h> #include <net/ip6_tunnel.h> +#include <net/inet_ecn.h> #include "spectrum_ipip.h" +#include "reg.h" struct ip_tunnel_parm mlxsw_sp_ipip_netdev_parms4(const struct net_device *ol_dev) @@ -338,3 +340,61 @@ static const struct mlxsw_sp_ipip_ops mlxsw_sp_ipip_gre4_ops = { const struct mlxsw_sp_ipip_ops *mlxsw_sp_ipip_ops_arr[] = { [MLXSW_SP_IPIP_TYPE_GRE4] = &mlxsw_sp_ipip_gre4_ops, }; + +static int mlxsw_sp_ipip_ecn_encap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tieem_pl[MLXSW_REG_TIEEM_LEN]; + + mlxsw_reg_tieem_pack(tieem_pl, inner_ecn, outer_ecn); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tieem), tieem_pl); +} + +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + u8 outer_ecn = INET_ECN_encapsulate(0, i); + int err; + + err = mlxsw_sp_ipip_ecn_encap_init_one(mlxsw_sp, i, outer_ecn); + if (err) + return err; + } + + return 0; +} + +static int mlxsw_sp_ipip_ecn_decap_init_one(struct mlxsw_sp *mlxsw_sp, + u8 inner_ecn, u8 outer_ecn) +{ + char tidem_pl[MLXSW_REG_TIDEM_LEN]; + bool trap_en, set_ce = false; + u8 new_inner_ecn; + + trap_en = __INET_ECN_decapsulate(outer_ecn, inner_ecn, &set_ce); + new_inner_ecn = set_ce ? INET_ECN_CE : inner_ecn; + + mlxsw_reg_tidem_pack(tidem_pl, outer_ecn, inner_ecn, new_inner_ecn, + trap_en, trap_en ? MLXSW_TRAP_ID_DECAP_ECN0 : 0); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(tidem), tidem_pl); +} + +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp) +{ + int i, j, err; + + /* Iterate over inner ECN values */ + for (i = INET_ECN_NOT_ECT; i <= INET_ECN_CE; i++) { + /* Iterate over outer ECN values */ + for (j = INET_ECN_NOT_ECT; j <= INET_ECN_CE; j++) { + err = mlxsw_sp_ipip_ecn_decap_init_one(mlxsw_sp, i, j); + if (err) + return err; + } + } + + return 0; +} diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c index ec2ff3d7f41c..34f7c3501b08 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_ptp.c @@ -920,6 +920,7 @@ static int mlxsw_sp_ptp_get_message_types(const struct hwtstamp_config *config, egr_types = 0xff; break; case HWTSTAMP_TX_ONESTEP_SYNC: + case HWTSTAMP_TX_ONESTEP_P2P: return -ERANGE; } @@ -1015,27 +1016,17 @@ mlxsw_sp1_ptp_port_shaper_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) static int mlxsw_sp1_ptp_port_shaper_check(struct mlxsw_sp_port *mlxsw_sp_port) { - const struct mlxsw_sp_port_type_speed_ops *port_type_speed_ops; - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; - char ptys_pl[MLXSW_REG_PTYS_LEN]; - u32 eth_proto_oper, speed; bool ptps = false; int err, i; + u32 speed; if (!mlxsw_sp1_ptp_hwtstamp_enabled(mlxsw_sp_port)) return mlxsw_sp1_ptp_port_shaper_set(mlxsw_sp_port, false); - port_type_speed_ops = mlxsw_sp->port_type_speed_ops; - port_type_speed_ops->reg_ptys_eth_pack(mlxsw_sp, ptys_pl, - mlxsw_sp_port->local_port, 0, - false); - err = mlxsw_reg_query(mlxsw_sp->core, MLXSW_REG(ptys), ptys_pl); + err = mlxsw_sp_port_speed_get(mlxsw_sp_port, &speed); if (err) return err; - port_type_speed_ops->reg_ptys_eth_unpack(mlxsw_sp, ptys_pl, NULL, NULL, - ð_proto_oper); - speed = port_type_speed_ops->from_ptys_speed(mlxsw_sp, eth_proto_oper); for (i = 0; i < MLXSW_SP1_PTP_SHAPER_PARAMS_LEN; i++) { if (mlxsw_sp1_ptp_shaper_params[i].ethtool_speed == speed) { ptps = true; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c index 0124bfe1963b..79a2801d59f6 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_qdisc.c @@ -18,6 +18,8 @@ enum mlxsw_sp_qdisc_type { MLXSW_SP_QDISC_NO_QDISC, MLXSW_SP_QDISC_RED, MLXSW_SP_QDISC_PRIO, + MLXSW_SP_QDISC_ETS, + MLXSW_SP_QDISC_TBF, }; struct mlxsw_sp_qdisc_ops { @@ -226,6 +228,70 @@ mlxsw_sp_qdisc_bstats_per_priority_get(struct mlxsw_sp_port_xstats *xstats, } } +static void +mlxsw_sp_qdisc_collect_tc_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u64 *p_tx_bytes, u64 *p_tx_packets, + u64 *p_drops, u64 *p_backlog) +{ + u8 tclass_num = mlxsw_sp_qdisc->tclass_num; + struct mlxsw_sp_port_xstats *xstats; + u64 tx_bytes, tx_packets; + + xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; + mlxsw_sp_qdisc_bstats_per_priority_get(xstats, + mlxsw_sp_qdisc->prio_bitmap, + &tx_packets, &tx_bytes); + + *p_tx_packets += tx_packets; + *p_tx_bytes += tx_bytes; + *p_drops += xstats->wred_drop[tclass_num] + + mlxsw_sp_xstats_tail_drop(xstats, tclass_num); + *p_backlog += mlxsw_sp_xstats_backlog(xstats, tclass_num); +} + +static void +mlxsw_sp_qdisc_update_stats(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u64 tx_bytes, u64 tx_packets, + u64 drops, u64 backlog, + struct tc_qopt_offload_stats *stats_ptr) +{ + struct mlxsw_sp_qdisc_stats *stats_base = &mlxsw_sp_qdisc->stats_base; + + tx_bytes -= stats_base->tx_bytes; + tx_packets -= stats_base->tx_packets; + drops -= stats_base->drops; + backlog -= stats_base->backlog; + + _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); + stats_ptr->qstats->drops += drops; + stats_ptr->qstats->backlog += mlxsw_sp_cells_bytes(mlxsw_sp, backlog); + + stats_base->backlog += backlog; + stats_base->drops += drops; + stats_base->tx_bytes += tx_bytes; + stats_base->tx_packets += tx_packets; +} + +static void +mlxsw_sp_qdisc_get_tc_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 backlog = 0; + u64 drops = 0; + + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog); + mlxsw_sp_qdisc_update_stats(mlxsw_sp_port->mlxsw_sp, mlxsw_sp_qdisc, + tx_bytes, tx_packets, drops, backlog, + stats_ptr); +} + static int mlxsw_sp_tclass_congestion_enable(struct mlxsw_sp_port *mlxsw_sp_port, int tclass_num, u32 min, u32 max, @@ -356,19 +422,28 @@ mlxsw_sp_qdisc_red_replace(struct mlxsw_sp_port *mlxsw_sp_port, } static void -mlxsw_sp_qdisc_red_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - void *params) +mlxsw_sp_qdisc_leaf_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct gnet_stats_queue *qstats) { - struct tc_red_qopt_offload_params *p = params; u64 backlog; backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, mlxsw_sp_qdisc->stats_base.backlog); - p->qstats->backlog -= backlog; + qstats->backlog -= backlog; mlxsw_sp_qdisc->stats_base.backlog = 0; } +static void +mlxsw_sp_qdisc_red_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_red_qopt_offload_params *p = params; + + mlxsw_sp_qdisc_leaf_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, p->qstats); +} + static int mlxsw_sp_qdisc_get_red_xstats(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, @@ -402,41 +477,21 @@ mlxsw_sp_qdisc_get_red_stats(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, struct tc_qopt_offload_stats *stats_ptr) { - u64 tx_bytes, tx_packets, overlimits, drops, backlog; u8 tclass_num = mlxsw_sp_qdisc->tclass_num; struct mlxsw_sp_qdisc_stats *stats_base; struct mlxsw_sp_port_xstats *xstats; + u64 overlimits; xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; stats_base = &mlxsw_sp_qdisc->stats_base; - mlxsw_sp_qdisc_bstats_per_priority_get(xstats, - mlxsw_sp_qdisc->prio_bitmap, - &tx_packets, &tx_bytes); - tx_bytes = tx_bytes - stats_base->tx_bytes; - tx_packets = tx_packets - stats_base->tx_packets; - + mlxsw_sp_qdisc_get_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, stats_ptr); overlimits = xstats->wred_drop[tclass_num] + xstats->ecn - stats_base->overlimits; - drops = xstats->wred_drop[tclass_num] + - mlxsw_sp_xstats_tail_drop(xstats, tclass_num) - - stats_base->drops; - backlog = mlxsw_sp_xstats_backlog(xstats, tclass_num); - _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); stats_ptr->qstats->overlimits += overlimits; - stats_ptr->qstats->drops += drops; - stats_ptr->qstats->backlog += - mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - backlog) - - mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - stats_base->backlog); - - stats_base->backlog = backlog; - stats_base->drops += drops; stats_base->overlimits += overlimits; - stats_base->tx_bytes += tx_bytes; - stats_base->tx_packets += tx_packets; + return 0; } @@ -486,15 +541,215 @@ int mlxsw_sp_setup_tc_red(struct mlxsw_sp_port *mlxsw_sp_port, } } +static void +mlxsw_sp_setup_tc_qdisc_leaf_clean_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + u64 backlog_cells = 0; + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 drops = 0; + + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog_cells); + + mlxsw_sp_qdisc->stats_base.tx_packets = tx_packets; + mlxsw_sp_qdisc->stats_base.tx_bytes = tx_bytes; + mlxsw_sp_qdisc->stats_base.drops = drops; + mlxsw_sp_qdisc->stats_base.backlog = 0; +} + static int -mlxsw_sp_qdisc_prio_destroy(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +mlxsw_sp_qdisc_tbf_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + struct mlxsw_sp_qdisc *root_qdisc = mlxsw_sp_port->root_qdisc; + + if (root_qdisc != mlxsw_sp_qdisc) + root_qdisc->stats_base.backlog -= + mlxsw_sp_qdisc->stats_base.backlog; + + return mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + mlxsw_sp_qdisc->tclass_num, 0, + MLXSW_REG_QEEC_MAS_DIS, 0); +} + +static int +mlxsw_sp_qdisc_tbf_bs(struct mlxsw_sp_port *mlxsw_sp_port, + u32 max_size, u8 *p_burst_size) +{ + /* TBF burst size is configured in bytes. The ASIC burst size value is + * ((2 ^ bs) * 512 bits. Convert the TBF bytes to 512-bit units. + */ + u32 bs512 = max_size / 64; + u8 bs = fls(bs512); + + if (!bs) + return -EINVAL; + --bs; + + /* Demand a power of two. */ + if ((1 << bs) != bs512) + return -EINVAL; + + if (bs < mlxsw_sp_port->mlxsw_sp->lowest_shaper_bs || + bs > MLXSW_REG_QEEC_HIGHEST_SHAPER_BS) + return -EINVAL; + + *p_burst_size = bs; + return 0; +} + +static u32 +mlxsw_sp_qdisc_tbf_max_size(u8 bs) +{ + return (1U << bs) * 64; +} + +static u64 +mlxsw_sp_qdisc_tbf_rate_kbps(struct tc_tbf_qopt_offload_replace_params *p) +{ + /* TBF interface is in bytes/s, whereas Spectrum ASIC is configured in + * Kbits/s. + */ + return p->rate.rate_bytes_ps / 1000 * 8; +} + +static int +mlxsw_sp_qdisc_tbf_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_tbf_qopt_offload_replace_params *p = params; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + u64 rate_kbps = mlxsw_sp_qdisc_tbf_rate_kbps(p); + u8 burst_size; + int err; + + if (rate_kbps >= MLXSW_REG_QEEC_MAS_DIS) { + dev_err(mlxsw_sp_port->mlxsw_sp->bus_info->dev, + "spectrum: TBF: rate of %lluKbps must be below %u\n", + rate_kbps, MLXSW_REG_QEEC_MAS_DIS); + return -EINVAL; + } + + err = mlxsw_sp_qdisc_tbf_bs(mlxsw_sp_port, p->max_size, &burst_size); + if (err) { + u8 highest_shaper_bs = MLXSW_REG_QEEC_HIGHEST_SHAPER_BS; + + dev_err(mlxsw_sp->bus_info->dev, + "spectrum: TBF: invalid burst size of %u, must be a power of two between %u and %u", + p->max_size, + mlxsw_sp_qdisc_tbf_max_size(mlxsw_sp->lowest_shaper_bs), + mlxsw_sp_qdisc_tbf_max_size(highest_shaper_bs)); + return -EINVAL; + } + + return 0; +} + +static int +mlxsw_sp_qdisc_tbf_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_tbf_qopt_offload_replace_params *p = params; + u64 rate_kbps = mlxsw_sp_qdisc_tbf_rate_kbps(p); + u8 burst_size; + int err; + + err = mlxsw_sp_qdisc_tbf_bs(mlxsw_sp_port, p->max_size, &burst_size); + if (WARN_ON_ONCE(err)) + /* check_params above was supposed to reject this value. */ + return -EINVAL; + + /* Configure subgroup shaper, so that both UC and MC traffic is subject + * to shaping. That is unlike RED, however UC queue lengths are going to + * be different than MC ones due to different pool and quota + * configurations, so the configuration is not applicable. For shaper on + * the other hand, subjecting the overall stream to the configured + * shaper makes sense. Also note that that is what we do for + * ieee_setmaxrate(). + */ + return mlxsw_sp_port_ets_maxrate_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + mlxsw_sp_qdisc->tclass_num, 0, + rate_kbps, burst_size); +} + +static void +mlxsw_sp_qdisc_tbf_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_tbf_qopt_offload_replace_params *p = params; + + mlxsw_sp_qdisc_leaf_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, p->qstats); +} + +static int +mlxsw_sp_qdisc_get_tbf_stats(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_qopt_offload_stats *stats_ptr) +{ + mlxsw_sp_qdisc_get_tc_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + stats_ptr); + return 0; +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_tbf = { + .type = MLXSW_SP_QDISC_TBF, + .check_params = mlxsw_sp_qdisc_tbf_check_params, + .replace = mlxsw_sp_qdisc_tbf_replace, + .unoffload = mlxsw_sp_qdisc_tbf_unoffload, + .destroy = mlxsw_sp_qdisc_tbf_destroy, + .get_stats = mlxsw_sp_qdisc_get_tbf_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_leaf_clean_stats, +}; + +int mlxsw_sp_setup_tc_tbf(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_tbf_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent, false); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_TBF_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_tbf, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle, + MLXSW_SP_QDISC_TBF)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_TBF_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_TBF_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + default: + return -EOPNOTSUPP; + } +} + +static int +__mlxsw_sp_qdisc_ets_destroy(struct mlxsw_sp_port *mlxsw_sp_port) { int i; for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { mlxsw_sp_port_prio_tc_set(mlxsw_sp_port, i, MLXSW_SP_PORT_DEFAULT_TCLASS); + mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + i, 0, false, 0); mlxsw_sp_qdisc_destroy(mlxsw_sp_port, &mlxsw_sp_port->tclass_qdiscs[i]); mlxsw_sp_port->tclass_qdiscs[i].prio_bitmap = 0; @@ -504,36 +759,58 @@ mlxsw_sp_qdisc_prio_destroy(struct mlxsw_sp_port *mlxsw_sp_port, } static int -mlxsw_sp_qdisc_prio_check_params(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - void *params) +mlxsw_sp_qdisc_prio_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) { - struct tc_prio_qopt_offload_params *p = params; + return __mlxsw_sp_qdisc_ets_destroy(mlxsw_sp_port); +} - if (p->bands > IEEE_8021QAZ_MAX_TCS) +static int +__mlxsw_sp_qdisc_ets_check_params(unsigned int nbands) +{ + if (nbands > IEEE_8021QAZ_MAX_TCS) return -EOPNOTSUPP; return 0; } static int -mlxsw_sp_qdisc_prio_replace(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - void *params) +mlxsw_sp_qdisc_prio_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) { struct tc_prio_qopt_offload_params *p = params; + + return __mlxsw_sp_qdisc_ets_check_params(p->bands); +} + +static int +__mlxsw_sp_qdisc_ets_replace(struct mlxsw_sp_port *mlxsw_sp_port, + unsigned int nbands, + const unsigned int *quanta, + const unsigned int *weights, + const u8 *priomap) +{ struct mlxsw_sp_qdisc *child_qdisc; int tclass, i, band, backlog; u8 old_priomap; int err; - for (band = 0; band < p->bands; band++) { + for (band = 0; band < nbands; band++) { tclass = MLXSW_SP_PRIO_BAND_TO_TCLASS(band); child_qdisc = &mlxsw_sp_port->tclass_qdiscs[tclass]; old_priomap = child_qdisc->prio_bitmap; child_qdisc->prio_bitmap = 0; + + err = mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + tclass, 0, !!quanta[band], + weights[band]); + if (err) + return err; + for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { - if (p->priomap[i] == band) { + if (priomap[i] == band) { child_qdisc->prio_bitmap |= BIT(i); if (BIT(i) & old_priomap) continue; @@ -556,21 +833,46 @@ mlxsw_sp_qdisc_prio_replace(struct mlxsw_sp_port *mlxsw_sp_port, child_qdisc = &mlxsw_sp_port->tclass_qdiscs[tclass]; child_qdisc->prio_bitmap = 0; mlxsw_sp_qdisc_destroy(mlxsw_sp_port, child_qdisc); + mlxsw_sp_port_ets_set(mlxsw_sp_port, + MLXSW_REG_QEEC_HR_SUBGROUP, + tclass, 0, false, 0); } return 0; } +static int +mlxsw_sp_qdisc_prio_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_prio_qopt_offload_params *p = params; + unsigned int zeroes[TCQ_ETS_MAX_BANDS] = {0}; + + return __mlxsw_sp_qdisc_ets_replace(mlxsw_sp_port, p->bands, + zeroes, zeroes, p->priomap); +} + +static void +__mlxsw_sp_qdisc_ets_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct gnet_stats_queue *qstats) +{ + u64 backlog; + + backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, + mlxsw_sp_qdisc->stats_base.backlog); + qstats->backlog -= backlog; +} + static void mlxsw_sp_qdisc_prio_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, void *params) { struct tc_prio_qopt_offload_params *p = params; - u64 backlog; - backlog = mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - mlxsw_sp_qdisc->stats_base.backlog); - p->qstats->backlog -= backlog; + __mlxsw_sp_qdisc_ets_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, + p->qstats); } static int @@ -578,37 +880,23 @@ mlxsw_sp_qdisc_get_prio_stats(struct mlxsw_sp_port *mlxsw_sp_port, struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, struct tc_qopt_offload_stats *stats_ptr) { - u64 tx_bytes, tx_packets, drops = 0, backlog = 0; - struct mlxsw_sp_qdisc_stats *stats_base; - struct mlxsw_sp_port_xstats *xstats; - struct rtnl_link_stats64 *stats; + struct mlxsw_sp_qdisc *tc_qdisc; + u64 tx_packets = 0; + u64 tx_bytes = 0; + u64 backlog = 0; + u64 drops = 0; int i; - xstats = &mlxsw_sp_port->periodic_hw_stats.xstats; - stats = &mlxsw_sp_port->periodic_hw_stats.stats; - stats_base = &mlxsw_sp_qdisc->stats_base; - - tx_bytes = stats->tx_bytes - stats_base->tx_bytes; - tx_packets = stats->tx_packets - stats_base->tx_packets; - for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) { - drops += mlxsw_sp_xstats_tail_drop(xstats, i); - drops += xstats->wred_drop[i]; - backlog += mlxsw_sp_xstats_backlog(xstats, i); + tc_qdisc = &mlxsw_sp_port->tclass_qdiscs[i]; + mlxsw_sp_qdisc_collect_tc_stats(mlxsw_sp_port, tc_qdisc, + &tx_bytes, &tx_packets, + &drops, &backlog); } - drops = drops - stats_base->drops; - _bstats_update(stats_ptr->bstats, tx_bytes, tx_packets); - stats_ptr->qstats->drops += drops; - stats_ptr->qstats->backlog += - mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - backlog) - - mlxsw_sp_cells_bytes(mlxsw_sp_port->mlxsw_sp, - stats_base->backlog); - stats_base->backlog = backlog; - stats_base->drops += drops; - stats_base->tx_bytes += tx_bytes; - stats_base->tx_packets += tx_packets; + mlxsw_sp_qdisc_update_stats(mlxsw_sp_port->mlxsw_sp, mlxsw_sp_qdisc, + tx_bytes, tx_packets, drops, backlog, + stats_ptr); return 0; } @@ -647,27 +935,93 @@ static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_prio = { .clean_stats = mlxsw_sp_setup_tc_qdisc_prio_clean_stats, }; -/* Grafting is not supported in mlxsw. It will result in un-offloading of the - * grafted qdisc as well as the qdisc in the qdisc new location. - * (However, if the graft is to the location where the qdisc is already at, it - * will be ignored completely and won't cause un-offloading). +static int +mlxsw_sp_qdisc_ets_check_params(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + return __mlxsw_sp_qdisc_ets_check_params(p->bands); +} + +static int +mlxsw_sp_qdisc_ets_replace(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + return __mlxsw_sp_qdisc_ets_replace(mlxsw_sp_port, p->bands, + p->quanta, p->weights, p->priomap); +} + +static void +mlxsw_sp_qdisc_ets_unoffload(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + void *params) +{ + struct tc_ets_qopt_offload_replace_params *p = params; + + __mlxsw_sp_qdisc_ets_unoffload(mlxsw_sp_port, mlxsw_sp_qdisc, + p->qstats); +} + +static int +mlxsw_sp_qdisc_ets_destroy(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc) +{ + return __mlxsw_sp_qdisc_ets_destroy(mlxsw_sp_port); +} + +static struct mlxsw_sp_qdisc_ops mlxsw_sp_qdisc_ops_ets = { + .type = MLXSW_SP_QDISC_ETS, + .check_params = mlxsw_sp_qdisc_ets_check_params, + .replace = mlxsw_sp_qdisc_ets_replace, + .unoffload = mlxsw_sp_qdisc_ets_unoffload, + .destroy = mlxsw_sp_qdisc_ets_destroy, + .get_stats = mlxsw_sp_qdisc_get_prio_stats, + .clean_stats = mlxsw_sp_setup_tc_qdisc_prio_clean_stats, +}; + +/* Linux allows linking of Qdiscs to arbitrary classes (so long as the resulting + * graph is free of cycles). These operations do not change the parent handle + * though, which means it can be incomplete (if there is more than one class + * where the Qdisc in question is grafted) or outright wrong (if the Qdisc was + * linked to a different class and then removed from the original class). + * + * E.g. consider this sequence of operations: + * + * # tc qdisc add dev swp1 root handle 1: prio + * # tc qdisc add dev swp1 parent 1:3 handle 13: red limit 1000000 avpkt 10000 + * RED: set bandwidth to 10Mbit + * # tc qdisc link dev swp1 handle 13: parent 1:2 + * + * At this point, both 1:2 and 1:3 have the same RED Qdisc instance as their + * child. But RED will still only claim that 1:3 is its parent. If it's removed + * from that band, its only parent will be 1:2, but it will continue to claim + * that it is in fact 1:3. + * + * The notification for child Qdisc replace (e.g. TC_RED_REPLACE) comes before + * the notification for parent graft (e.g. TC_PRIO_GRAFT). We take the replace + * notification to offload the child Qdisc, based on its parent handle, and use + * the graft operation to validate that the class where the child is actually + * grafted corresponds to the parent handle. If the two don't match, we + * unoffload the child. */ static int -mlxsw_sp_qdisc_prio_graft(struct mlxsw_sp_port *mlxsw_sp_port, - struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, - struct tc_prio_qopt_offload_graft_params *p) +__mlxsw_sp_qdisc_ets_graft(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + u8 band, u32 child_handle) { - int tclass_num = MLXSW_SP_PRIO_BAND_TO_TCLASS(p->band); + int tclass_num = MLXSW_SP_PRIO_BAND_TO_TCLASS(band); struct mlxsw_sp_qdisc *old_qdisc; - /* Check if the grafted qdisc is already in its "new" location. If so - - * nothing needs to be done. - */ - if (p->band < IEEE_8021QAZ_MAX_TCS && - mlxsw_sp_port->tclass_qdiscs[tclass_num].handle == p->child_handle) + if (band < IEEE_8021QAZ_MAX_TCS && + mlxsw_sp_port->tclass_qdiscs[tclass_num].handle == child_handle) return 0; - if (!p->child_handle) { + if (!child_handle) { /* This is an invisible FIFO replacing the original Qdisc. * Ignore it--the original Qdisc's destroy will follow. */ @@ -678,7 +1032,7 @@ mlxsw_sp_qdisc_prio_graft(struct mlxsw_sp_port *mlxsw_sp_port, * unoffload it. */ old_qdisc = mlxsw_sp_qdisc_find_by_handle(mlxsw_sp_port, - p->child_handle); + child_handle); if (old_qdisc) mlxsw_sp_qdisc_destroy(mlxsw_sp_port, old_qdisc); @@ -687,6 +1041,15 @@ mlxsw_sp_qdisc_prio_graft(struct mlxsw_sp_port *mlxsw_sp_port, return -EOPNOTSUPP; } +static int +mlxsw_sp_qdisc_prio_graft(struct mlxsw_sp_port *mlxsw_sp_port, + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc, + struct tc_prio_qopt_offload_graft_params *p) +{ + return __mlxsw_sp_qdisc_ets_graft(mlxsw_sp_port, mlxsw_sp_qdisc, + p->band, p->child_handle); +} + int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, struct tc_prio_qopt_offload *p) { @@ -720,6 +1083,40 @@ int mlxsw_sp_setup_tc_prio(struct mlxsw_sp_port *mlxsw_sp_port, } } +int mlxsw_sp_setup_tc_ets(struct mlxsw_sp_port *mlxsw_sp_port, + struct tc_ets_qopt_offload *p) +{ + struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; + + mlxsw_sp_qdisc = mlxsw_sp_qdisc_find(mlxsw_sp_port, p->parent, true); + if (!mlxsw_sp_qdisc) + return -EOPNOTSUPP; + + if (p->command == TC_ETS_REPLACE) + return mlxsw_sp_qdisc_replace(mlxsw_sp_port, p->handle, + mlxsw_sp_qdisc, + &mlxsw_sp_qdisc_ops_ets, + &p->replace_params); + + if (!mlxsw_sp_qdisc_compare(mlxsw_sp_qdisc, p->handle, + MLXSW_SP_QDISC_ETS)) + return -EOPNOTSUPP; + + switch (p->command) { + case TC_ETS_DESTROY: + return mlxsw_sp_qdisc_destroy(mlxsw_sp_port, mlxsw_sp_qdisc); + case TC_ETS_STATS: + return mlxsw_sp_qdisc_get_stats(mlxsw_sp_port, mlxsw_sp_qdisc, + &p->stats); + case TC_ETS_GRAFT: + return __mlxsw_sp_qdisc_ets_graft(mlxsw_sp_port, mlxsw_sp_qdisc, + p->graft_params.band, + p->graft_params.child_handle); + default: + return -EOPNOTSUPP; + } +} + int mlxsw_sp_tc_qdisc_init(struct mlxsw_sp_port *mlxsw_sp_port) { struct mlxsw_sp_qdisc *mlxsw_sp_qdisc; diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c index 8290e82240fc..ce707723f8cf 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c @@ -382,9 +382,10 @@ enum mlxsw_sp_fib_entry_type { }; struct mlxsw_sp_nexthop_group; +struct mlxsw_sp_fib_entry; struct mlxsw_sp_fib_node { - struct list_head entry_list; + struct mlxsw_sp_fib_entry *fib_entry; struct list_head list; struct rhash_head ht_node; struct mlxsw_sp_fib *fib; @@ -397,7 +398,6 @@ struct mlxsw_sp_fib_entry_decap { }; struct mlxsw_sp_fib_entry { - struct list_head list; struct mlxsw_sp_fib_node *fib_node; enum mlxsw_sp_fib_entry_type type; struct list_head nexthop_group_node; @@ -1162,7 +1162,6 @@ mlxsw_sp_router_ip2me_fib_entry_find(struct mlxsw_sp *mlxsw_sp, u32 tb_id, const union mlxsw_sp_l3addr *addr, enum mlxsw_sp_fib_entry_type type) { - struct mlxsw_sp_fib_entry *fib_entry; struct mlxsw_sp_fib_node *fib_node; unsigned char addr_prefix_len; struct mlxsw_sp_fib *fib; @@ -1191,15 +1190,10 @@ mlxsw_sp_router_ip2me_fib_entry_find(struct mlxsw_sp *mlxsw_sp, u32 tb_id, fib_node = mlxsw_sp_fib_node_lookup(fib, addrp, addr_len, addr_prefix_len); - if (!fib_node || list_empty(&fib_node->entry_list)) + if (!fib_node || fib_node->fib_entry->type != type) return NULL; - fib_entry = list_first_entry(&fib_node->entry_list, - struct mlxsw_sp_fib_entry, list); - if (fib_entry->type != type) - return NULL; - - return fib_entry; + return fib_node->fib_entry; } /* Given an IPIP entry, find the corresponding decap route. */ @@ -1209,7 +1203,6 @@ mlxsw_sp_ipip_entry_find_decap(struct mlxsw_sp *mlxsw_sp, { static struct mlxsw_sp_fib_node *fib_node; const struct mlxsw_sp_ipip_ops *ipip_ops; - struct mlxsw_sp_fib_entry *fib_entry; unsigned char saddr_prefix_len; union mlxsw_sp_l3addr saddr; struct mlxsw_sp_fib *ul_fib; @@ -1244,15 +1237,11 @@ mlxsw_sp_ipip_entry_find_decap(struct mlxsw_sp *mlxsw_sp, fib_node = mlxsw_sp_fib_node_lookup(ul_fib, saddrp, saddr_len, saddr_prefix_len); - if (!fib_node || list_empty(&fib_node->entry_list)) + if (!fib_node || + fib_node->fib_entry->type != MLXSW_SP_FIB_ENTRY_TYPE_TRAP) return NULL; - fib_entry = list_first_entry(&fib_node->entry_list, - struct mlxsw_sp_fib_entry, list); - if (fib_entry->type != MLXSW_SP_FIB_ENTRY_TYPE_TRAP) - return NULL; - - return fib_entry; + return fib_node->fib_entry; } static struct mlxsw_sp_ipip_entry * @@ -3231,10 +3220,6 @@ mlxsw_sp_nexthop_group_update(struct mlxsw_sp *mlxsw_sp, return 0; } -static bool -mlxsw_sp_fib_node_entry_is_first(const struct mlxsw_sp_fib_node *fib_node, - const struct mlxsw_sp_fib_entry *fib_entry); - static int mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_nexthop_group *nh_grp) @@ -3243,9 +3228,6 @@ mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, int err; list_for_each_entry(fib_entry, &nh_grp->fib_list, nexthop_group_node) { - if (!mlxsw_sp_fib_node_entry_is_first(fib_entry->fib_node, - fib_entry)) - continue; err = mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); if (err) return err; @@ -3253,24 +3235,6 @@ mlxsw_sp_nexthop_fib_entries_update(struct mlxsw_sp *mlxsw_sp, return 0; } -static void -mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op, int err); - -static void -mlxsw_sp_nexthop_fib_entries_refresh(struct mlxsw_sp_nexthop_group *nh_grp) -{ - enum mlxsw_reg_ralue_op op = MLXSW_REG_RALUE_OP_WRITE_WRITE; - struct mlxsw_sp_fib_entry *fib_entry; - - list_for_each_entry(fib_entry, &nh_grp->fib_list, nexthop_group_node) { - if (!mlxsw_sp_fib_node_entry_is_first(fib_entry->fib_node, - fib_entry)) - continue; - mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, 0); - } -} - static void mlxsw_sp_adj_grp_size_round_up(u16 *p_adj_grp_size) { /* Valid sizes for an adjacency group are: @@ -3374,6 +3338,73 @@ mlxsw_sp_nexthop_group_rebalance(struct mlxsw_sp_nexthop_group *nh_grp) } } +static struct mlxsw_sp_nexthop * +mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, + const struct mlxsw_sp_rt6 *mlxsw_sp_rt6); + +static void +mlxsw_sp_nexthop4_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + int i; + + for (i = 0; i < nh_grp->count; i++) { + struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + + if (nh->offloaded) + nh->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +__mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp_nexthop_group *nh_grp, + struct mlxsw_sp_fib6_entry *fib6_entry) +{ + struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { + struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; + struct mlxsw_sp_nexthop *nh; + + nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); + if (nh && nh->offloaded) + fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; + else + fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; + } +} + +static void +mlxsw_sp_nexthop6_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + + /* Unfortunately, in IPv6 the route and the nexthop are described by + * the same struct, so we need to iterate over all the routes using the + * nexthop group and set / clear the offload indication for them. + */ + list_for_each_entry(fib6_entry, &nh_grp->fib_list, + common.nexthop_group_node) + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); +} + +static void +mlxsw_sp_nexthop_group_offload_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_nexthop_group *nh_grp) +{ + switch (mlxsw_sp_nexthop_group_type(nh_grp)) { + case AF_INET: + mlxsw_sp_nexthop4_group_offload_refresh(mlxsw_sp, nh_grp); + break; + case AF_INET6: + mlxsw_sp_nexthop6_group_offload_refresh(mlxsw_sp, nh_grp); + break; + } +} + static void mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_nexthop_group *nh_grp) @@ -3447,6 +3478,8 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, goto set_trap; } + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); + if (!old_adj_index_valid) { /* The trap was set for fib entries, so we have to call * fib entry update to unset it and use adjacency index. @@ -3468,9 +3501,6 @@ mlxsw_sp_nexthop_group_refresh(struct mlxsw_sp *mlxsw_sp, goto set_trap; } - /* Offload state within the group changed, so update the flags. */ - mlxsw_sp_nexthop_fib_entries_refresh(nh_grp); - return; set_trap: @@ -3483,6 +3513,7 @@ set_trap: err = mlxsw_sp_nexthop_fib_entries_update(mlxsw_sp, nh_grp); if (err) dev_warn(mlxsw_sp->bus_info->dev, "Failed to set traps for fib entries.\n"); + mlxsw_sp_nexthop_group_offload_refresh(mlxsw_sp, nh_grp); if (old_adj_index_valid) mlxsw_sp_kvdl_free(mlxsw_sp, MLXSW_SP_KVDL_ENTRY_TYPE_ADJ, nh_grp->ecmp_size, nh_grp->adj_index); @@ -3845,7 +3876,7 @@ static void mlxsw_sp_nexthop4_event(struct mlxsw_sp *mlxsw_sp, key.fib_nh = fib_nh; nh = mlxsw_sp_nexthop_lookup(mlxsw_sp, key); - if (WARN_ON_ONCE(!nh)) + if (!nh) return; switch (event) { @@ -4065,131 +4096,128 @@ mlxsw_sp_rt6_nexthop(struct mlxsw_sp_nexthop_group *nh_grp, } static void -mlxsw_sp_fib4_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib4_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - int i; - - if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_NVE_DECAP) { - nh_grp->nexthops->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - return; - } - - for (i = 0; i < nh_grp->count; i++) { - struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + struct fib_info *fi = mlxsw_sp_nexthop4_group_fi(fib_entry->nh_group); + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; + bool should_offload; - if (nh->offloaded) - nh->key.fib_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - else - nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.tos = fib4_entry->tos; + fri.type = fib4_entry->type; + fri.offload = should_offload; + fri.trap = !should_offload; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); } static void -mlxsw_sp_fib4_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib4_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - int i; - - if (!list_is_singular(&nh_grp->fib_list)) - return; - - for (i = 0; i < nh_grp->count; i++) { - struct mlxsw_sp_nexthop *nh = &nh_grp->nexthops[i]; + struct fib_info *fi = mlxsw_sp_nexthop4_group_fi(fib_entry->nh_group); + u32 *p_dst = (u32 *) fib_entry->fib_node->key.addr; + int dst_len = fib_entry->fib_node->key.prefix_len; + struct mlxsw_sp_fib4_entry *fib4_entry; + struct fib_rt_info fri; - nh->key.fib_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + fib4_entry = container_of(fib_entry, struct mlxsw_sp_fib4_entry, + common); + fri.fi = fi; + fri.tb_id = fib4_entry->tb_id; + fri.dst = cpu_to_be32(*p_dst); + fri.dst_len = dst_len; + fri.tos = fib4_entry->tos; + fri.type = fib4_entry->type; + fri.offload = false; + fri.trap = false; + fib_alias_hw_flags_set(mlxsw_sp_net(mlxsw_sp), &fri); } static void -mlxsw_sp_fib6_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib6_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_rt6 *mlxsw_sp_rt6; + bool should_offload; + + should_offload = mlxsw_sp_fib_entry_should_offload(fib_entry); + /* In IPv6 a multipath route is represented using multiple routes, so + * we need to set the flags on all of them. + */ fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, common); - - if (fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_LOCAL || - fib_entry->type == MLXSW_SP_FIB_ENTRY_TYPE_BLACKHOLE) { - list_first_entry(&fib6_entry->rt6_list, struct mlxsw_sp_rt6, - list)->rt->fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - return; - } - - list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { - struct mlxsw_sp_nexthop_group *nh_grp = fib_entry->nh_group; - struct fib6_nh *fib6_nh = mlxsw_sp_rt6->rt->fib6_nh; - struct mlxsw_sp_nexthop *nh; - - nh = mlxsw_sp_rt6_nexthop(nh_grp, mlxsw_sp_rt6); - if (nh && nh->offloaded) - fib6_nh->fib_nh_flags |= RTNH_F_OFFLOAD; - else - fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_rt6->rt, should_offload, + !should_offload); } static void -mlxsw_sp_fib6_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib6_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_rt6 *mlxsw_sp_rt6; fib6_entry = container_of(fib_entry, struct mlxsw_sp_fib6_entry, common); - list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) { - struct fib6_info *rt = mlxsw_sp_rt6->rt; - - rt->fib6_nh->fib_nh_flags &= ~RTNH_F_OFFLOAD; - } + list_for_each_entry(mlxsw_sp_rt6, &fib6_entry->rt6_list, list) + fib6_info_hw_flags_set(mlxsw_sp_rt6->rt, false, false); } -static void mlxsw_sp_fib_entry_offload_set(struct mlxsw_sp_fib_entry *fib_entry) +static void +mlxsw_sp_fib_entry_hw_flags_set(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - mlxsw_sp_fib4_entry_offload_set(fib_entry); + mlxsw_sp_fib4_entry_hw_flags_set(mlxsw_sp, fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - mlxsw_sp_fib6_entry_offload_set(fib_entry); + mlxsw_sp_fib6_entry_hw_flags_set(mlxsw_sp, fib_entry); break; } } static void -mlxsw_sp_fib_entry_offload_unset(struct mlxsw_sp_fib_entry *fib_entry) +mlxsw_sp_fib_entry_hw_flags_clear(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { switch (fib_entry->fib_node->fib->proto) { case MLXSW_SP_L3_PROTO_IPV4: - mlxsw_sp_fib4_entry_offload_unset(fib_entry); + mlxsw_sp_fib4_entry_hw_flags_clear(mlxsw_sp, fib_entry); break; case MLXSW_SP_L3_PROTO_IPV6: - mlxsw_sp_fib6_entry_offload_unset(fib_entry); + mlxsw_sp_fib6_entry_hw_flags_clear(mlxsw_sp, fib_entry); break; } } static void -mlxsw_sp_fib_entry_offload_refresh(struct mlxsw_sp_fib_entry *fib_entry, - enum mlxsw_reg_ralue_op op, int err) +mlxsw_sp_fib_entry_hw_flags_refresh(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry, + enum mlxsw_reg_ralue_op op) { switch (op) { - case MLXSW_REG_RALUE_OP_WRITE_DELETE: - return mlxsw_sp_fib_entry_offload_unset(fib_entry); case MLXSW_REG_RALUE_OP_WRITE_WRITE: - if (err) - return; - if (mlxsw_sp_fib_entry_should_offload(fib_entry)) - mlxsw_sp_fib_entry_offload_set(fib_entry); - else - mlxsw_sp_fib_entry_offload_unset(fib_entry); - return; + mlxsw_sp_fib_entry_hw_flags_set(mlxsw_sp, fib_entry); + break; + case MLXSW_REG_RALUE_OP_WRITE_DELETE: + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, fib_entry); + break; default: - return; + break; } } @@ -4416,7 +4444,10 @@ static int mlxsw_sp_fib_entry_op(struct mlxsw_sp *mlxsw_sp, { int err = __mlxsw_sp_fib_entry_op(mlxsw_sp, fib_entry, op); - mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, err); + if (err) + return err; + + mlxsw_sp_fib_entry_hw_flags_refresh(mlxsw_sp, fib_entry, op); return err; } @@ -4491,6 +4522,19 @@ mlxsw_sp_fib4_entry_type_set(struct mlxsw_sp *mlxsw_sp, } } +static void +mlxsw_sp_fib4_entry_type_unset(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) +{ + switch (fib_entry->type) { + case MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP: + mlxsw_sp_fib_entry_decap_fini(mlxsw_sp, fib_entry); + break; + default: + break; + } +} + static struct mlxsw_sp_fib4_entry * mlxsw_sp_fib4_entry_create(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node, @@ -4523,6 +4567,7 @@ mlxsw_sp_fib4_entry_create(struct mlxsw_sp *mlxsw_sp, return fib4_entry; err_nexthop4_group_get: + mlxsw_sp_fib4_entry_type_unset(mlxsw_sp, fib_entry); err_fib4_entry_type_set: kfree(fib4_entry); return ERR_PTR(err); @@ -4532,6 +4577,7 @@ static void mlxsw_sp_fib4_entry_destroy(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib4_entry *fib4_entry) { mlxsw_sp_nexthop4_group_put(mlxsw_sp, &fib4_entry->common); + mlxsw_sp_fib4_entry_type_unset(mlxsw_sp, &fib4_entry->common); kfree(fib4_entry); } @@ -4555,15 +4601,14 @@ mlxsw_sp_fib4_entry_lookup(struct mlxsw_sp *mlxsw_sp, if (!fib_node) return NULL; - list_for_each_entry(fib4_entry, &fib_node->entry_list, common.list) { - if (fib4_entry->tb_id == fen_info->tb_id && - fib4_entry->tos == fen_info->tos && - fib4_entry->type == fen_info->type && - mlxsw_sp_nexthop4_group_fi(fib4_entry->common.nh_group) == - fen_info->fi) { - return fib4_entry; - } - } + fib4_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib4_entry, common); + if (fib4_entry->tb_id == fen_info->tb_id && + fib4_entry->tos == fen_info->tos && + fib4_entry->type == fen_info->type && + mlxsw_sp_nexthop4_group_fi(fib4_entry->common.nh_group) == + fen_info->fi) + return fib4_entry; return NULL; } @@ -4611,7 +4656,6 @@ mlxsw_sp_fib_node_create(struct mlxsw_sp_fib *fib, const void *addr, if (!fib_node) return NULL; - INIT_LIST_HEAD(&fib_node->entry_list); list_add(&fib_node->list, &fib->node_list); memcpy(fib_node->key.addr, addr, addr_len); fib_node->key.prefix_len = prefix_len; @@ -4622,18 +4666,9 @@ mlxsw_sp_fib_node_create(struct mlxsw_sp_fib *fib, const void *addr, static void mlxsw_sp_fib_node_destroy(struct mlxsw_sp_fib_node *fib_node) { list_del(&fib_node->list); - WARN_ON(!list_empty(&fib_node->entry_list)); kfree(fib_node); } -static bool -mlxsw_sp_fib_node_entry_is_first(const struct mlxsw_sp_fib_node *fib_node, - const struct mlxsw_sp_fib_entry *fib_entry) -{ - return list_first_entry(&fib_node->entry_list, - struct mlxsw_sp_fib_entry, list) == fib_entry; -} - static int mlxsw_sp_fib_lpm_tree_link(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node) { @@ -4773,200 +4808,48 @@ static void mlxsw_sp_fib_node_put(struct mlxsw_sp *mlxsw_sp, { struct mlxsw_sp_vr *vr = fib_node->fib->vr; - if (!list_empty(&fib_node->entry_list)) + if (fib_node->fib_entry) return; mlxsw_sp_fib_node_fini(mlxsw_sp, fib_node); mlxsw_sp_fib_node_destroy(fib_node); mlxsw_sp_vr_put(mlxsw_sp, vr); } -static struct mlxsw_sp_fib4_entry * -mlxsw_sp_fib4_node_entry_find(const struct mlxsw_sp_fib_node *fib_node, - const struct mlxsw_sp_fib4_entry *new4_entry) -{ - struct mlxsw_sp_fib4_entry *fib4_entry; - - list_for_each_entry(fib4_entry, &fib_node->entry_list, common.list) { - if (fib4_entry->tb_id > new4_entry->tb_id) - continue; - if (fib4_entry->tb_id != new4_entry->tb_id) - break; - if (fib4_entry->tos > new4_entry->tos) - continue; - if (fib4_entry->prio >= new4_entry->prio || - fib4_entry->tos < new4_entry->tos) - return fib4_entry; - } - - return NULL; -} - -static int -mlxsw_sp_fib4_node_list_append(struct mlxsw_sp_fib4_entry *fib4_entry, - struct mlxsw_sp_fib4_entry *new4_entry) -{ - struct mlxsw_sp_fib_node *fib_node; - - if (WARN_ON(!fib4_entry)) - return -EINVAL; - - fib_node = fib4_entry->common.fib_node; - list_for_each_entry_from(fib4_entry, &fib_node->entry_list, - common.list) { - if (fib4_entry->tb_id != new4_entry->tb_id || - fib4_entry->tos != new4_entry->tos || - fib4_entry->prio != new4_entry->prio) - break; - } - - list_add_tail(&new4_entry->common.list, &fib4_entry->common.list); - return 0; -} - -static int -mlxsw_sp_fib4_node_list_insert(struct mlxsw_sp_fib4_entry *new4_entry, - bool replace, bool append) -{ - struct mlxsw_sp_fib_node *fib_node = new4_entry->common.fib_node; - struct mlxsw_sp_fib4_entry *fib4_entry; - - fib4_entry = mlxsw_sp_fib4_node_entry_find(fib_node, new4_entry); - - if (append) - return mlxsw_sp_fib4_node_list_append(fib4_entry, new4_entry); - if (replace && WARN_ON(!fib4_entry)) - return -EINVAL; - - /* Insert new entry before replaced one, so that we can later - * remove the second. - */ - if (fib4_entry) { - list_add_tail(&new4_entry->common.list, - &fib4_entry->common.list); - } else { - struct mlxsw_sp_fib4_entry *last; - - list_for_each_entry(last, &fib_node->entry_list, common.list) { - if (new4_entry->tb_id > last->tb_id) - break; - fib4_entry = last; - } - - if (fib4_entry) - list_add(&new4_entry->common.list, - &fib4_entry->common.list); - else - list_add(&new4_entry->common.list, - &fib_node->entry_list); - } - - return 0; -} - -static void -mlxsw_sp_fib4_node_list_remove(struct mlxsw_sp_fib4_entry *fib4_entry) -{ - list_del(&fib4_entry->common.list); -} - -static int mlxsw_sp_fib_node_entry_add(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib_entry *fib_entry) -{ - struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; - - if (!mlxsw_sp_fib_node_entry_is_first(fib_node, fib_entry)) - return 0; - - /* To prevent packet loss, overwrite the previously offloaded - * entry. - */ - if (!list_is_singular(&fib_node->entry_list)) { - enum mlxsw_reg_ralue_op op = MLXSW_REG_RALUE_OP_WRITE_DELETE; - struct mlxsw_sp_fib_entry *n = list_next_entry(fib_entry, list); - - mlxsw_sp_fib_entry_offload_refresh(n, op, 0); - } - - return mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); -} - -static void mlxsw_sp_fib_node_entry_del(struct mlxsw_sp *mlxsw_sp, +static int mlxsw_sp_fib_node_entry_link(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_entry *fib_entry) { struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; - - if (!mlxsw_sp_fib_node_entry_is_first(fib_node, fib_entry)) - return; - - /* Promote the next entry by overwriting the deleted entry */ - if (!list_is_singular(&fib_node->entry_list)) { - struct mlxsw_sp_fib_entry *n = list_next_entry(fib_entry, list); - enum mlxsw_reg_ralue_op op = MLXSW_REG_RALUE_OP_WRITE_DELETE; - - mlxsw_sp_fib_entry_update(mlxsw_sp, n); - mlxsw_sp_fib_entry_offload_refresh(fib_entry, op, 0); - return; - } - - mlxsw_sp_fib_entry_del(mlxsw_sp, fib_entry); -} - -static int mlxsw_sp_fib4_node_entry_link(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib4_entry *fib4_entry, - bool replace, bool append) -{ int err; - err = mlxsw_sp_fib4_node_list_insert(fib4_entry, replace, append); - if (err) - return err; + fib_node->fib_entry = fib_entry; - err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib4_entry->common); + err = mlxsw_sp_fib_entry_update(mlxsw_sp, fib_entry); if (err) - goto err_fib_node_entry_add; + goto err_fib_entry_update; return 0; -err_fib_node_entry_add: - mlxsw_sp_fib4_node_list_remove(fib4_entry); +err_fib_entry_update: + fib_node->fib_entry = NULL; return err; } static void -mlxsw_sp_fib4_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib4_entry *fib4_entry) -{ - mlxsw_sp_fib_node_entry_del(mlxsw_sp, &fib4_entry->common); - mlxsw_sp_fib4_node_list_remove(fib4_entry); - - if (fib4_entry->common.type == MLXSW_SP_FIB_ENTRY_TYPE_IPIP_DECAP) - mlxsw_sp_fib_entry_decap_fini(mlxsw_sp, &fib4_entry->common); -} - -static void mlxsw_sp_fib4_entry_replace(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib4_entry *fib4_entry, - bool replace) +mlxsw_sp_fib_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_fib_entry *fib_entry) { - struct mlxsw_sp_fib_node *fib_node = fib4_entry->common.fib_node; - struct mlxsw_sp_fib4_entry *replaced; - - if (!replace) - return; - - /* We inserted the new entry before replaced one */ - replaced = list_next_entry(fib4_entry, common.list); + struct mlxsw_sp_fib_node *fib_node = fib_entry->fib_node; - mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, replaced); - mlxsw_sp_fib4_entry_destroy(mlxsw_sp, replaced); - mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + mlxsw_sp_fib_entry_del(mlxsw_sp, fib_entry); + fib_node->fib_entry = NULL; } static int -mlxsw_sp_router_fib4_add(struct mlxsw_sp *mlxsw_sp, - const struct fib_entry_notifier_info *fen_info, - bool replace, bool append) +mlxsw_sp_router_fib4_replace(struct mlxsw_sp *mlxsw_sp, + const struct fib_entry_notifier_info *fen_info) { - struct mlxsw_sp_fib4_entry *fib4_entry; + struct mlxsw_sp_fib4_entry *fib4_entry, *fib4_replaced; + struct mlxsw_sp_fib_entry *replaced; struct mlxsw_sp_fib_node *fib_node; int err; @@ -4989,18 +4872,26 @@ mlxsw_sp_router_fib4_add(struct mlxsw_sp *mlxsw_sp, goto err_fib4_entry_create; } - err = mlxsw_sp_fib4_node_entry_link(mlxsw_sp, fib4_entry, replace, - append); + replaced = fib_node->fib_entry; + err = mlxsw_sp_fib_node_entry_link(mlxsw_sp, &fib4_entry->common); if (err) { dev_warn(mlxsw_sp->bus_info->dev, "Failed to link FIB entry to node\n"); - goto err_fib4_node_entry_link; + goto err_fib_node_entry_link; } - mlxsw_sp_fib4_entry_replace(mlxsw_sp, fib4_entry, replace); + /* Nothing to replace */ + if (!replaced) + return 0; + + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); + fib4_replaced = container_of(replaced, struct mlxsw_sp_fib4_entry, + common); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_replaced); return 0; -err_fib4_node_entry_link: +err_fib_node_entry_link: + fib_node->fib_entry = replaced; mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); err_fib4_entry_create: mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); @@ -5021,7 +4912,7 @@ static void mlxsw_sp_router_fib4_del(struct mlxsw_sp *mlxsw_sp, return; fib_node = fib4_entry->common.fib_node; - mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, &fib4_entry->common); mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } @@ -5083,13 +4974,6 @@ static void mlxsw_sp_rt6_destroy(struct mlxsw_sp_rt6 *mlxsw_sp_rt6) kfree(mlxsw_sp_rt6); } -static bool mlxsw_sp_fib6_rt_can_mp(const struct fib6_info *rt) -{ - /* RTF_CACHE routes are ignored */ - return !(rt->fib6_flags & RTF_ADDRCONF) && - rt->fib6_nh->fib_nh_gw_family; -} - static struct fib6_info * mlxsw_sp_fib6_entry_rt(const struct mlxsw_sp_fib6_entry *fib6_entry) { @@ -5097,37 +4981,6 @@ mlxsw_sp_fib6_entry_rt(const struct mlxsw_sp_fib6_entry *fib6_entry) list)->rt; } -static struct mlxsw_sp_fib6_entry * -mlxsw_sp_fib6_node_mp_entry_find(const struct mlxsw_sp_fib_node *fib_node, - const struct fib6_info *nrt, bool replace) -{ - struct mlxsw_sp_fib6_entry *fib6_entry; - - if (!mlxsw_sp_fib6_rt_can_mp(nrt) || replace) - return NULL; - - list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { - struct fib6_info *rt = mlxsw_sp_fib6_entry_rt(fib6_entry); - - /* RT6_TABLE_LOCAL and RT6_TABLE_MAIN share the same - * virtual router. - */ - if (rt->fib6_table->tb6_id > nrt->fib6_table->tb6_id) - continue; - if (rt->fib6_table->tb6_id != nrt->fib6_table->tb6_id) - break; - if (rt->fib6_metric < nrt->fib6_metric) - continue; - if (rt->fib6_metric == nrt->fib6_metric && - mlxsw_sp_fib6_rt_can_mp(rt)) - return fib6_entry; - if (rt->fib6_metric > nrt->fib6_metric) - break; - } - - return NULL; -} - static struct mlxsw_sp_rt6 * mlxsw_sp_fib6_entry_rt_find(const struct mlxsw_sp_fib6_entry *fib6_entry, const struct fib6_info *rt) @@ -5313,6 +5166,11 @@ static int mlxsw_sp_nexthop6_group_get(struct mlxsw_sp *mlxsw_sp, &nh_grp->fib_list); fib6_entry->common.nh_group = nh_grp; + /* The route and the nexthop are described by the same struct, so we + * need to the update the nexthop offload indication for the new route. + */ + __mlxsw_sp_nexthop6_group_offload_refresh(nh_grp, fib6_entry); + return 0; } @@ -5345,16 +5203,16 @@ mlxsw_sp_nexthop6_group_update(struct mlxsw_sp *mlxsw_sp, * currently associated with it in the device's table is that * of the old group. Start using the new one instead. */ - err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib6_entry->common); + err = mlxsw_sp_fib_entry_update(mlxsw_sp, &fib6_entry->common); if (err) - goto err_fib_node_entry_add; + goto err_fib_entry_update; if (list_empty(&old_nh_grp->fib_list)) mlxsw_sp_nexthop6_group_destroy(mlxsw_sp, old_nh_grp); return 0; -err_fib_node_entry_add: +err_fib_entry_update: mlxsw_sp_nexthop6_group_put(mlxsw_sp, &fib6_entry->common); err_nexthop6_group_get: list_add_tail(&fib6_entry->common.nexthop_group_node, @@ -5519,112 +5377,13 @@ static void mlxsw_sp_fib6_entry_destroy(struct mlxsw_sp *mlxsw_sp, } static struct mlxsw_sp_fib6_entry * -mlxsw_sp_fib6_node_entry_find(const struct mlxsw_sp_fib_node *fib_node, - const struct fib6_info *nrt, bool replace) -{ - struct mlxsw_sp_fib6_entry *fib6_entry, *fallback = NULL; - - list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { - struct fib6_info *rt = mlxsw_sp_fib6_entry_rt(fib6_entry); - - if (rt->fib6_table->tb6_id > nrt->fib6_table->tb6_id) - continue; - if (rt->fib6_table->tb6_id != nrt->fib6_table->tb6_id) - break; - if (replace && rt->fib6_metric == nrt->fib6_metric) { - if (mlxsw_sp_fib6_rt_can_mp(rt) == - mlxsw_sp_fib6_rt_can_mp(nrt)) - return fib6_entry; - if (mlxsw_sp_fib6_rt_can_mp(nrt)) - fallback = fallback ?: fib6_entry; - } - if (rt->fib6_metric > nrt->fib6_metric) - return fallback ?: fib6_entry; - } - - return fallback; -} - -static int -mlxsw_sp_fib6_node_list_insert(struct mlxsw_sp_fib6_entry *new6_entry, - bool *p_replace) -{ - struct mlxsw_sp_fib_node *fib_node = new6_entry->common.fib_node; - struct fib6_info *nrt = mlxsw_sp_fib6_entry_rt(new6_entry); - struct mlxsw_sp_fib6_entry *fib6_entry; - - fib6_entry = mlxsw_sp_fib6_node_entry_find(fib_node, nrt, *p_replace); - - if (*p_replace && !fib6_entry) - *p_replace = false; - - if (fib6_entry) { - list_add_tail(&new6_entry->common.list, - &fib6_entry->common.list); - } else { - struct mlxsw_sp_fib6_entry *last; - - list_for_each_entry(last, &fib_node->entry_list, common.list) { - struct fib6_info *rt = mlxsw_sp_fib6_entry_rt(last); - - if (nrt->fib6_table->tb6_id > rt->fib6_table->tb6_id) - break; - fib6_entry = last; - } - - if (fib6_entry) - list_add(&new6_entry->common.list, - &fib6_entry->common.list); - else - list_add(&new6_entry->common.list, - &fib_node->entry_list); - } - - return 0; -} - -static void -mlxsw_sp_fib6_node_list_remove(struct mlxsw_sp_fib6_entry *fib6_entry) -{ - list_del(&fib6_entry->common.list); -} - -static int mlxsw_sp_fib6_node_entry_link(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib6_entry *fib6_entry, - bool *p_replace) -{ - int err; - - err = mlxsw_sp_fib6_node_list_insert(fib6_entry, p_replace); - if (err) - return err; - - err = mlxsw_sp_fib_node_entry_add(mlxsw_sp, &fib6_entry->common); - if (err) - goto err_fib_node_entry_add; - - return 0; - -err_fib_node_entry_add: - mlxsw_sp_fib6_node_list_remove(fib6_entry); - return err; -} - -static void -mlxsw_sp_fib6_node_entry_unlink(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib6_entry *fib6_entry) -{ - mlxsw_sp_fib_node_entry_del(mlxsw_sp, &fib6_entry->common); - mlxsw_sp_fib6_node_list_remove(fib6_entry); -} - -static struct mlxsw_sp_fib6_entry * mlxsw_sp_fib6_entry_lookup(struct mlxsw_sp *mlxsw_sp, const struct fib6_info *rt) { struct mlxsw_sp_fib6_entry *fib6_entry; struct mlxsw_sp_fib_node *fib_node; struct mlxsw_sp_fib *fib; + struct fib6_info *cmp_rt; struct mlxsw_sp_vr *vr; vr = mlxsw_sp_vr_find(mlxsw_sp, rt->fib6_table->tb6_id); @@ -5638,40 +5397,23 @@ mlxsw_sp_fib6_entry_lookup(struct mlxsw_sp *mlxsw_sp, if (!fib_node) return NULL; - list_for_each_entry(fib6_entry, &fib_node->entry_list, common.list) { - struct fib6_info *iter_rt = mlxsw_sp_fib6_entry_rt(fib6_entry); - - if (rt->fib6_table->tb6_id == iter_rt->fib6_table->tb6_id && - rt->fib6_metric == iter_rt->fib6_metric && - mlxsw_sp_fib6_entry_rt_find(fib6_entry, rt)) - return fib6_entry; - } + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + cmp_rt = mlxsw_sp_fib6_entry_rt(fib6_entry); + if (rt->fib6_table->tb6_id == cmp_rt->fib6_table->tb6_id && + rt->fib6_metric == cmp_rt->fib6_metric && + mlxsw_sp_fib6_entry_rt_find(fib6_entry, rt)) + return fib6_entry; return NULL; } -static void mlxsw_sp_fib6_entry_replace(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_fib6_entry *fib6_entry, - bool replace) -{ - struct mlxsw_sp_fib_node *fib_node = fib6_entry->common.fib_node; - struct mlxsw_sp_fib6_entry *replaced; - - if (!replace) - return; - - replaced = list_next_entry(fib6_entry, common.list); - - mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, replaced); - mlxsw_sp_fib6_entry_destroy(mlxsw_sp, replaced); - mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); -} - -static int mlxsw_sp_router_fib6_add(struct mlxsw_sp *mlxsw_sp, - struct fib6_info **rt_arr, - unsigned int nrt6, bool replace) +static int mlxsw_sp_router_fib6_replace(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) { - struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib6_entry *fib6_entry, *fib6_replaced; + struct mlxsw_sp_fib_entry *replaced; struct mlxsw_sp_fib_node *fib_node; struct fib6_info *rt = rt_arr[0]; int err; @@ -5693,18 +5435,6 @@ static int mlxsw_sp_router_fib6_add(struct mlxsw_sp *mlxsw_sp, if (IS_ERR(fib_node)) return PTR_ERR(fib_node); - /* Before creating a new entry, try to append route to an existing - * multipath entry. - */ - fib6_entry = mlxsw_sp_fib6_node_mp_entry_find(fib_node, rt, replace); - if (fib6_entry) { - err = mlxsw_sp_fib6_entry_nexthop_add(mlxsw_sp, fib6_entry, - rt_arr, nrt6); - if (err) - goto err_fib6_entry_nexthop_add; - return 0; - } - fib6_entry = mlxsw_sp_fib6_entry_create(mlxsw_sp, fib_node, rt_arr, nrt6); if (IS_ERR(fib6_entry)) { @@ -5712,17 +5442,70 @@ static int mlxsw_sp_router_fib6_add(struct mlxsw_sp *mlxsw_sp, goto err_fib6_entry_create; } - err = mlxsw_sp_fib6_node_entry_link(mlxsw_sp, fib6_entry, &replace); + replaced = fib_node->fib_entry; + err = mlxsw_sp_fib_node_entry_link(mlxsw_sp, &fib6_entry->common); if (err) - goto err_fib6_node_entry_link; + goto err_fib_node_entry_link; + + /* Nothing to replace */ + if (!replaced) + return 0; - mlxsw_sp_fib6_entry_replace(mlxsw_sp, fib6_entry, replace); + mlxsw_sp_fib_entry_hw_flags_clear(mlxsw_sp, replaced); + fib6_replaced = container_of(replaced, struct mlxsw_sp_fib6_entry, + common); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_replaced); return 0; -err_fib6_node_entry_link: +err_fib_node_entry_link: + fib_node->fib_entry = replaced; mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); err_fib6_entry_create: + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return err; +} + +static int mlxsw_sp_router_fib6_append(struct mlxsw_sp *mlxsw_sp, + struct fib6_info **rt_arr, + unsigned int nrt6) +{ + struct mlxsw_sp_fib6_entry *fib6_entry; + struct mlxsw_sp_fib_node *fib_node; + struct fib6_info *rt = rt_arr[0]; + int err; + + if (mlxsw_sp->router->aborted) + return 0; + + if (rt->fib6_src.plen) + return -EINVAL; + + if (mlxsw_sp_fib6_rt_should_ignore(rt)) + return 0; + + fib_node = mlxsw_sp_fib_node_get(mlxsw_sp, rt->fib6_table->tb6_id, + &rt->fib6_dst.addr, + sizeof(rt->fib6_dst.addr), + rt->fib6_dst.plen, + MLXSW_SP_L3_PROTO_IPV6); + if (IS_ERR(fib_node)) + return PTR_ERR(fib_node); + + if (WARN_ON_ONCE(!fib_node->fib_entry)) { + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); + return -EINVAL; + } + + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + err = mlxsw_sp_fib6_entry_nexthop_add(mlxsw_sp, fib6_entry, rt_arr, + nrt6); + if (err) + goto err_fib6_entry_nexthop_add; + + return 0; + err_fib6_entry_nexthop_add: mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); return err; @@ -5762,7 +5545,7 @@ static void mlxsw_sp_router_fib6_del(struct mlxsw_sp *mlxsw_sp, fib_node = fib6_entry->common.fib_node; - mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, &fib6_entry->common); mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } @@ -5916,39 +5699,25 @@ static int mlxsw_sp_router_set_abort_trap(struct mlxsw_sp *mlxsw_sp) static void mlxsw_sp_fib4_node_flush(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node) { - struct mlxsw_sp_fib4_entry *fib4_entry, *tmp; - - list_for_each_entry_safe(fib4_entry, tmp, &fib_node->entry_list, - common.list) { - bool do_break = &tmp->common.list == &fib_node->entry_list; + struct mlxsw_sp_fib4_entry *fib4_entry; - mlxsw_sp_fib4_node_entry_unlink(mlxsw_sp, fib4_entry); - mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); - mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); - /* Break when entry list is empty and node was freed. - * Otherwise, we'll access freed memory in the next - * iteration. - */ - if (do_break) - break; - } + fib4_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib4_entry, common); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, fib_node->fib_entry); + mlxsw_sp_fib4_entry_destroy(mlxsw_sp, fib4_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } static void mlxsw_sp_fib6_node_flush(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_fib_node *fib_node) { - struct mlxsw_sp_fib6_entry *fib6_entry, *tmp; - - list_for_each_entry_safe(fib6_entry, tmp, &fib_node->entry_list, - common.list) { - bool do_break = &tmp->common.list == &fib_node->entry_list; + struct mlxsw_sp_fib6_entry *fib6_entry; - mlxsw_sp_fib6_node_entry_unlink(mlxsw_sp, fib6_entry); - mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); - mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); - if (do_break) - break; - } + fib6_entry = container_of(fib_node->fib_entry, + struct mlxsw_sp_fib6_entry, common); + mlxsw_sp_fib_node_entry_unlink(mlxsw_sp, fib_node->fib_entry); + mlxsw_sp_fib6_entry_destroy(mlxsw_sp, fib6_entry); + mlxsw_sp_fib_node_put(mlxsw_sp, fib_node); } static void mlxsw_sp_fib_node_flush(struct mlxsw_sp *mlxsw_sp, @@ -6099,7 +5868,6 @@ static void mlxsw_sp_router_fib4_event_work(struct work_struct *work) struct mlxsw_sp_fib_event_work *fib_work = container_of(work, struct mlxsw_sp_fib_event_work, work); struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; - bool replace, append; int err; /* Protect internal structures from changes */ @@ -6107,13 +5875,9 @@ static void mlxsw_sp_router_fib4_event_work(struct work_struct *work) mlxsw_sp_span_respin(mlxsw_sp); switch (fib_work->event) { - case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ - case FIB_EVENT_ENTRY_ADD: - replace = fib_work->event == FIB_EVENT_ENTRY_REPLACE; - append = fib_work->event == FIB_EVENT_ENTRY_APPEND; - err = mlxsw_sp_router_fib4_add(mlxsw_sp, &fib_work->fen_info, - replace, append); + case FIB_EVENT_ENTRY_REPLACE: + err = mlxsw_sp_router_fib4_replace(mlxsw_sp, + &fib_work->fen_info); if (err) mlxsw_sp_router_fib_abort(mlxsw_sp); fib_info_put(fib_work->fen_info.fi); @@ -6138,20 +5902,24 @@ static void mlxsw_sp_router_fib6_event_work(struct work_struct *work) struct mlxsw_sp_fib_event_work *fib_work = container_of(work, struct mlxsw_sp_fib_event_work, work); struct mlxsw_sp *mlxsw_sp = fib_work->mlxsw_sp; - bool replace; int err; rtnl_lock(); mlxsw_sp_span_respin(mlxsw_sp); switch (fib_work->event) { - case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_ADD: - replace = fib_work->event == FIB_EVENT_ENTRY_REPLACE; - err = mlxsw_sp_router_fib6_add(mlxsw_sp, - fib_work->fib6_work.rt_arr, - fib_work->fib6_work.nrt6, - replace); + case FIB_EVENT_ENTRY_REPLACE: + err = mlxsw_sp_router_fib6_replace(mlxsw_sp, + fib_work->fib6_work.rt_arr, + fib_work->fib6_work.nrt6); + if (err) + mlxsw_sp_router_fib_abort(mlxsw_sp); + mlxsw_sp_router_fib6_work_fini(&fib_work->fib6_work); + break; + case FIB_EVENT_ENTRY_APPEND: + err = mlxsw_sp_router_fib6_append(mlxsw_sp, + fib_work->fib6_work.rt_arr, + fib_work->fib6_work.nrt6); if (err) mlxsw_sp_router_fib_abort(mlxsw_sp); mlxsw_sp_router_fib6_work_fini(&fib_work->fib6_work); @@ -6216,8 +5984,6 @@ static void mlxsw_sp_router_fib4_event(struct mlxsw_sp_fib_event_work *fib_work, switch (fib_work->event) { case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ - case FIB_EVENT_ENTRY_ADD: /* fall through */ case FIB_EVENT_ENTRY_DEL: fen_info = container_of(info, struct fib_entry_notifier_info, info); @@ -6245,7 +6011,7 @@ static int mlxsw_sp_router_fib6_event(struct mlxsw_sp_fib_event_work *fib_work, switch (fib_work->event) { case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_ADD: /* fall through */ + case FIB_EVENT_ENTRY_APPEND: /* fall through */ case FIB_EVENT_ENTRY_DEL: fen6_info = container_of(info, struct fib6_entry_notifier_info, info); @@ -6348,9 +6114,9 @@ static int mlxsw_sp_router_fib_event(struct notifier_block *nb, err = mlxsw_sp_router_fib_rule_event(event, info, router->mlxsw_sp); return notifier_from_errno(err); - case FIB_EVENT_ENTRY_ADD: + case FIB_EVENT_ENTRY_ADD: /* fall through */ case FIB_EVENT_ENTRY_REPLACE: /* fall through */ - case FIB_EVENT_ENTRY_APPEND: /* fall through */ + case FIB_EVENT_ENTRY_APPEND: if (router->aborted) { NL_SET_ERR_MSG_MOD(info->extack, "FIB offload was aborted. Not configuring route"); return notifier_from_errno(-EINVAL); @@ -8025,8 +7791,18 @@ mlxsw_sp_ipip_config_tigcr(struct mlxsw_sp *mlxsw_sp) static int mlxsw_sp_ipips_init(struct mlxsw_sp *mlxsw_sp) { + int err; + mlxsw_sp->router->ipip_ops_arr = mlxsw_sp_ipip_ops_arr; INIT_LIST_HEAD(&mlxsw_sp->router->ipip_list); + + err = mlxsw_sp_ipip_ecn_encap_init(mlxsw_sp); + if (err) + return err; + err = mlxsw_sp_ipip_ecn_decap_init(mlxsw_sp); + if (err) + return err; + return mlxsw_sp_ipip_config_tigcr(mlxsw_sp); } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h index cc1de91e8217..c9b94f435cdd 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_router.h @@ -104,4 +104,7 @@ static inline bool mlxsw_sp_l3addr_eq(const union mlxsw_sp_l3addr *addr1, return !memcmp(addr1, addr2, sizeof(*addr1)); } +int mlxsw_sp_ipip_ecn_encap_init(struct mlxsw_sp *mlxsw_sp); +int mlxsw_sp_ipip_ecn_decap_init(struct mlxsw_sp *mlxsw_sp); + #endif /* _MLXSW_ROUTER_H_*/ diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c index 200d324e6d99..0cdd7954a085 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.c @@ -748,33 +748,50 @@ static bool mlxsw_sp_span_is_egress_mirror(struct mlxsw_sp_port *port) return false; } -static int mlxsw_sp_span_mtu_to_buffsize(const struct mlxsw_sp *mlxsw_sp, - int mtu) +static int +mlxsw_sp_span_port_buffsize_update(struct mlxsw_sp_port *mlxsw_sp_port, u16 mtu) { - return mlxsw_sp_bytes_cells(mlxsw_sp, mtu * 5 / 2) + 1; + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + char sbib_pl[MLXSW_REG_SBIB_LEN]; + u32 buffsize; + u32 speed; + int err; + + err = mlxsw_sp_port_speed_get(mlxsw_sp_port, &speed); + if (err) + return err; + if (speed == SPEED_UNKNOWN) + speed = 0; + + buffsize = mlxsw_sp_span_buffsize_get(mlxsw_sp, speed, mtu); + mlxsw_reg_sbib_pack(sbib_pl, mlxsw_sp_port->local_port, buffsize); + return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); } int mlxsw_sp_span_port_mtu_update(struct mlxsw_sp_port *port, u16 mtu) { - struct mlxsw_sp *mlxsw_sp = port->mlxsw_sp; - char sbib_pl[MLXSW_REG_SBIB_LEN]; - int err; - /* If port is egress mirrored, the shared buffer size should be * updated according to the mtu value */ - if (mlxsw_sp_span_is_egress_mirror(port)) { - u32 buffsize = mlxsw_sp_span_mtu_to_buffsize(mlxsw_sp, mtu); + if (mlxsw_sp_span_is_egress_mirror(port)) + return mlxsw_sp_span_port_buffsize_update(port, mtu); + return 0; +} - mlxsw_reg_sbib_pack(sbib_pl, port->local_port, buffsize); - err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); - if (err) { - netdev_err(port->dev, "Could not update shared buffer for mirroring\n"); - return err; - } - } +void mlxsw_sp_span_speed_update_work(struct work_struct *work) +{ + struct delayed_work *dwork = to_delayed_work(work); + struct mlxsw_sp_port *mlxsw_sp_port; - return 0; + mlxsw_sp_port = container_of(dwork, struct mlxsw_sp_port, + span.speed_update_dw); + + /* If port is egress mirrored, the shared buffer size should be + * updated according to the speed value. + */ + if (mlxsw_sp_span_is_egress_mirror(mlxsw_sp_port)) + mlxsw_sp_span_port_buffsize_update(mlxsw_sp_port, + mlxsw_sp_port->dev->mtu); } static struct mlxsw_sp_span_inspected_port * @@ -836,15 +853,9 @@ mlxsw_sp_span_inspected_port_add(struct mlxsw_sp_port *port, /* if it is an egress SPAN, bind a shared buffer to it */ if (type == MLXSW_SP_SPAN_EGRESS) { - u32 buffsize = mlxsw_sp_span_mtu_to_buffsize(mlxsw_sp, - port->dev->mtu); - - mlxsw_reg_sbib_pack(sbib_pl, port->local_port, buffsize); - err = mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sbib), sbib_pl); - if (err) { - netdev_err(port->dev, "Could not create shared buffer for mirroring\n"); + err = mlxsw_sp_span_port_buffsize_update(port, port->dev->mtu); + if (err) return err; - } } if (bind) { diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h index 5e04252f2a11..59724335525f 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_span.h @@ -74,5 +74,6 @@ void mlxsw_sp_span_entry_invalidate(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_span_entry *span_entry); int mlxsw_sp_span_port_mtu_update(struct mlxsw_sp_port *port, u16 mtu); +void mlxsw_sp_span_speed_update_work(struct work_struct *work); #endif diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c index e0d7c49ffae0..60205aa3f6a5 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_trap.c @@ -9,6 +9,20 @@ #include "reg.h" #include "spectrum.h" +/* All driver-specific traps must be documented in + * Documentation/networking/devlink/mlxsw.rst + */ +enum { + DEVLINK_MLXSW_TRAP_ID_BASE = DEVLINK_TRAP_GENERIC_ID_MAX, + DEVLINK_MLXSW_TRAP_ID_IRIF_DISABLED, + DEVLINK_MLXSW_TRAP_ID_ERIF_DISABLED, +}; + +#define DEVLINK_MLXSW_TRAP_NAME_IRIF_DISABLED \ + "irif_disabled" +#define DEVLINK_MLXSW_TRAP_NAME_ERIF_DISABLED \ + "erif_disabled" + #define MLXSW_SP_TRAP_METADATA DEVLINK_TRAP_METADATA_TYPE_F_IN_PORT static void mlxsw_sp_rx_drop_listener(struct sk_buff *skb, u8 local_port, @@ -21,6 +35,12 @@ static void mlxsw_sp_rx_exception_listener(struct sk_buff *skb, u8 local_port, DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ MLXSW_SP_TRAP_METADATA) +#define MLXSW_SP_TRAP_DRIVER_DROP(_id, _group_id) \ + DEVLINK_TRAP_DRIVER(DROP, DROP, DEVLINK_MLXSW_TRAP_ID_##_id, \ + DEVLINK_MLXSW_TRAP_NAME_##_id, \ + DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ + MLXSW_SP_TRAP_METADATA) + #define MLXSW_SP_TRAP_EXCEPTION(_id, _group_id) \ DEVLINK_TRAP_GENERIC(EXCEPTION, TRAP, _id, \ DEVLINK_TRAP_GROUP_GENERIC(_group_id), \ @@ -58,6 +78,11 @@ static struct devlink_trap mlxsw_sp_traps_arr[] = { MLXSW_SP_TRAP_EXCEPTION(UNRESOLVED_NEIGH, L3_DROPS), MLXSW_SP_TRAP_EXCEPTION(IPV4_LPM_UNICAST_MISS, L3_DROPS), MLXSW_SP_TRAP_EXCEPTION(IPV6_LPM_UNICAST_MISS, L3_DROPS), + MLXSW_SP_TRAP_DRIVER_DROP(IRIF_DISABLED, L3_DROPS), + MLXSW_SP_TRAP_DRIVER_DROP(ERIF_DISABLED, L3_DROPS), + MLXSW_SP_TRAP_DROP(NON_ROUTABLE, L3_DROPS), + MLXSW_SP_TRAP_EXCEPTION(DECAP_ERROR, TUNNEL_DROPS), + MLXSW_SP_TRAP_DROP(OVERLAY_SMAC_MC, TUNNEL_DROPS), }; static struct mlxsw_listener mlxsw_sp_listeners_arr[] = { @@ -90,6 +115,15 @@ static struct mlxsw_listener mlxsw_sp_listeners_arr[] = { TRAP_EXCEPTION_TO_CPU), MLXSW_SP_RXL_EXCEPTION(DISCARD_ROUTER_LPM6, ROUTER_EXP, TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_DISCARD(ROUTER_IRIF_EN, L3_DISCARDS), + MLXSW_SP_RXL_DISCARD(ROUTER_ERIF_EN, L3_DISCARDS), + MLXSW_SP_RXL_DISCARD(NON_ROUTABLE, L3_DISCARDS), + MLXSW_SP_RXL_EXCEPTION(DECAP_ECN0, ROUTER_EXP, TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(IPIP_DECAP_ERROR, ROUTER_EXP, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_EXCEPTION(DISCARD_DEC_PKT, TUNNEL_DISCARDS, + TRAP_EXCEPTION_TO_CPU), + MLXSW_SP_RXL_DISCARD(OVERLAY_SMAC_MC, TUNNEL_DISCARDS), }; /* Mapping between hardware trap and devlink trap. Multiple hardware traps can @@ -123,6 +157,13 @@ static u16 mlxsw_sp_listener_devlink_map[] = { DEVLINK_TRAP_GENERIC_ID_UNRESOLVED_NEIGH, DEVLINK_TRAP_GENERIC_ID_IPV4_LPM_UNICAST_MISS, DEVLINK_TRAP_GENERIC_ID_IPV6_LPM_UNICAST_MISS, + DEVLINK_MLXSW_TRAP_ID_IRIF_DISABLED, + DEVLINK_MLXSW_TRAP_ID_ERIF_DISABLED, + DEVLINK_TRAP_GENERIC_ID_NON_ROUTABLE, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_DECAP_ERROR, + DEVLINK_TRAP_GENERIC_ID_OVERLAY_SMAC_MC, }; static int mlxsw_sp_rx_listener(struct mlxsw_sp *mlxsw_sp, struct sk_buff *skb, @@ -304,8 +345,9 @@ mlxsw_sp_trap_group_policer_init(struct mlxsw_sp *mlxsw_sp, u32 rate; switch (group->id) { - case DEVLINK_TRAP_GROUP_GENERIC_ID_L3_DROPS:/* fall through */ - case DEVLINK_TRAP_GROUP_GENERIC_ID_L2_DROPS: + case DEVLINK_TRAP_GROUP_GENERIC_ID_L2_DROPS: /* fall through */ + case DEVLINK_TRAP_GROUP_GENERIC_ID_L3_DROPS: /* fall through */ + case DEVLINK_TRAP_GROUP_GENERIC_ID_TUNNEL_DROPS: policer_id = MLXSW_SP_DISCARD_POLICER_ID; ir_units = MLXSW_REG_QPCR_IR_UNITS_M; is_bytes = false; @@ -342,6 +384,12 @@ __mlxsw_sp_trap_group_init(struct mlxsw_sp *mlxsw_sp, priority = 0; tc = 1; break; + case DEVLINK_TRAP_GROUP_GENERIC_ID_TUNNEL_DROPS: + group_id = MLXSW_REG_HTGT_TRAP_GROUP_SP_TUNNEL_DISCARDS; + policer_id = MLXSW_SP_DISCARD_POLICER_ID; + priority = 0; + tc = 1; + break; default: return -EINVAL; } diff --git a/drivers/net/ethernet/mellanox/mlxsw/trap.h b/drivers/net/ethernet/mellanox/mlxsw/trap.h index 0c1c142bb6b0..12e1fa998d42 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/trap.h +++ b/drivers/net/ethernet/mellanox/mlxsw/trap.h @@ -67,6 +67,7 @@ enum { MLXSW_TRAP_ID_NVE_ENCAP_ARP = 0xBD, MLXSW_TRAP_ID_ROUTER_ALERT_IPV4 = 0xD6, MLXSW_TRAP_ID_ROUTER_ALERT_IPV6 = 0xD7, + MLXSW_TRAP_ID_DISCARD_NON_ROUTABLE = 0x11A, MLXSW_TRAP_ID_DISCARD_ROUTER2 = 0x130, MLXSW_TRAP_ID_DISCARD_ROUTER3 = 0x131, MLXSW_TRAP_ID_DISCARD_ING_PACKET_SMAC_MC = 0x140, @@ -80,12 +81,20 @@ enum { MLXSW_TRAP_ID_DISCARD_ING_ROUTER_UC_DIP_MC_DMAC = 0x161, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_DIP_LB = 0x162, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_MC = 0x163, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_CLASS_E = 0x164, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_LB = 0x165, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_CORRUPTED_IP_HDR = 0x167, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_MC_DMAC = 0x168, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_SIP_DIP = 0x169, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_SIP_BC = 0x16A, MLXSW_TRAP_ID_DISCARD_ING_ROUTER_IPV4_DIP_LOCAL_NET = 0x16B, + MLXSW_TRAP_ID_DISCARD_ING_ROUTER_DIP_LINK_LOCAL = 0x16C, + MLXSW_TRAP_ID_DISCARD_ROUTER_IRIF_EN = 0x178, + MLXSW_TRAP_ID_DISCARD_ROUTER_ERIF_EN = 0x179, MLXSW_TRAP_ID_DISCARD_ROUTER_LPM4 = 0x17B, MLXSW_TRAP_ID_DISCARD_ROUTER_LPM6 = 0x17C, + MLXSW_TRAP_ID_DISCARD_DEC_PKT = 0x188, + MLXSW_TRAP_ID_DISCARD_OVERLAY_SMAC_MC = 0x190, MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_RESERVED_SCOPE = 0x1B0, MLXSW_TRAP_ID_DISCARD_IPV6_MC_DIP_INTERFACE_LOCAL_SCOPE = 0x1B1, MLXSW_TRAP_ID_ACL0 = 0x1C0, diff --git a/drivers/net/ethernet/micrel/ks8842.c b/drivers/net/ethernet/micrel/ks8842.c index da329ca115cc..f3f6dfe3eddc 100644 --- a/drivers/net/ethernet/micrel/ks8842.c +++ b/drivers/net/ethernet/micrel/ks8842.c @@ -1103,7 +1103,7 @@ static void ks8842_tx_timeout_work(struct work_struct *work) __ks8842_start_new_rx_dma(netdev); } -static void ks8842_tx_timeout(struct net_device *netdev) +static void ks8842_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ks8842_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/micrel/ksz884x.c b/drivers/net/ethernet/micrel/ksz884x.c index e102e1560ac7..d1444ba36e10 100644 --- a/drivers/net/ethernet/micrel/ksz884x.c +++ b/drivers/net/ethernet/micrel/ksz884x.c @@ -4896,7 +4896,7 @@ unlock: * triggered to free up resources so that the transmit routine can continue * sending out packets. The hardware is reset to correct the problem. */ -static void netdev_tx_timeout(struct net_device *dev) +static void netdev_tx_timeout(struct net_device *dev, unsigned int txqueue) { static unsigned long last_reset; diff --git a/drivers/net/ethernet/microchip/enc28j60.c b/drivers/net/ethernet/microchip/enc28j60.c index 0567e4f387a5..09cdc2f2e7ff 100644 --- a/drivers/net/ethernet/microchip/enc28j60.c +++ b/drivers/net/ethernet/microchip/enc28j60.c @@ -1325,7 +1325,7 @@ static irqreturn_t enc28j60_irq(int irq, void *dev_id) return IRQ_HANDLED; } -static void enc28j60_tx_timeout(struct net_device *ndev) +static void enc28j60_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct enc28j60_net *priv = netdev_priv(ndev); diff --git a/drivers/net/ethernet/microchip/encx24j600.c b/drivers/net/ethernet/microchip/encx24j600.c index 52c41d11f565..39925e4bf2ec 100644 --- a/drivers/net/ethernet/microchip/encx24j600.c +++ b/drivers/net/ethernet/microchip/encx24j600.c @@ -892,7 +892,7 @@ static netdev_tx_t encx24j600_tx(struct sk_buff *skb, struct net_device *dev) } /* Deal with a transmit timeout */ -static void encx24j600_tx_timeout(struct net_device *dev) +static void encx24j600_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct encx24j600_priv *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/microchip/lan743x_ptp.c b/drivers/net/ethernet/microchip/lan743x_ptp.c index afe52463dc57..9399f6a98748 100644 --- a/drivers/net/ethernet/microchip/lan743x_ptp.c +++ b/drivers/net/ethernet/microchip/lan743x_ptp.c @@ -1265,6 +1265,9 @@ int lan743x_ptp_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) lan743x_ptp_set_sync_ts_insert(adapter, true); break; + case HWTSTAMP_TX_ONESTEP_P2P: + ret = -ERANGE; + break; default: netif_warn(adapter, drv, adapter->netdev, " tx_type = %d, UNKNOWN\n", config.tx_type); diff --git a/drivers/net/ethernet/mscc/ocelot.c b/drivers/net/ethernet/mscc/ocelot.c index 985b46d7e3d1..86d543ab1ab9 100644 --- a/drivers/net/ethernet/mscc/ocelot.c +++ b/drivers/net/ethernet/mscc/ocelot.c @@ -500,13 +500,14 @@ EXPORT_SYMBOL(ocelot_port_enable); static int ocelot_port_open(struct net_device *dev) { struct ocelot_port_private *priv = netdev_priv(dev); - struct ocelot *ocelot = priv->port.ocelot; + struct ocelot_port *ocelot_port = &priv->port; + struct ocelot *ocelot = ocelot_port->ocelot; int port = priv->chip_port; int err; if (priv->serdes) { err = phy_set_mode_ext(priv->serdes, PHY_MODE_ETHERNET, - priv->phy_mode); + ocelot_port->phy_mode); if (err) { netdev_err(dev, "Could not set mode of SerDes\n"); return err; @@ -514,7 +515,7 @@ static int ocelot_port_open(struct net_device *dev) } err = phy_connect_direct(dev, priv->phy, &ocelot_port_adjust_link, - priv->phy_mode); + ocelot_port->phy_mode); if (err) { netdev_err(dev, "Could not attach to PHY\n"); return err; diff --git a/drivers/net/ethernet/mscc/ocelot.h b/drivers/net/ethernet/mscc/ocelot.h index c259114c48fd..04372ba72fec 100644 --- a/drivers/net/ethernet/mscc/ocelot.h +++ b/drivers/net/ethernet/mscc/ocelot.h @@ -18,11 +18,11 @@ #include <linux/ptp_clock_kernel.h> #include <linux/regmap.h> +#include <soc/mscc/ocelot_qsys.h> #include <soc/mscc/ocelot_sys.h> +#include <soc/mscc/ocelot_dev.h> +#include <soc/mscc/ocelot_ana.h> #include <soc/mscc/ocelot.h> -#include "ocelot_ana.h" -#include "ocelot_dev.h" -#include "ocelot_qsys.h" #include "ocelot_rew.h" #include "ocelot_qs.h" #include "ocelot_tc.h" @@ -68,7 +68,6 @@ struct ocelot_port_private { u8 vlan_aware; - phy_interface_t phy_mode; struct phy *serdes; struct ocelot_port_tc tc; diff --git a/drivers/net/ethernet/mscc/ocelot_ana.h b/drivers/net/ethernet/mscc/ocelot_ana.h deleted file mode 100644 index 841c6ec22b64..000000000000 --- a/drivers/net/ethernet/mscc/ocelot_ana.h +++ /dev/null @@ -1,625 +0,0 @@ -/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ -/* - * Microsemi Ocelot Switch driver - * - * Copyright (c) 2017 Microsemi Corporation - */ - -#ifndef _MSCC_OCELOT_ANA_H_ -#define _MSCC_OCELOT_ANA_H_ - -#define ANA_ANAGEFIL_B_DOM_EN BIT(22) -#define ANA_ANAGEFIL_B_DOM_VAL BIT(21) -#define ANA_ANAGEFIL_AGE_LOCKED BIT(20) -#define ANA_ANAGEFIL_PID_EN BIT(19) -#define ANA_ANAGEFIL_PID_VAL(x) (((x) << 14) & GENMASK(18, 14)) -#define ANA_ANAGEFIL_PID_VAL_M GENMASK(18, 14) -#define ANA_ANAGEFIL_PID_VAL_X(x) (((x) & GENMASK(18, 14)) >> 14) -#define ANA_ANAGEFIL_VID_EN BIT(13) -#define ANA_ANAGEFIL_VID_VAL(x) ((x) & GENMASK(12, 0)) -#define ANA_ANAGEFIL_VID_VAL_M GENMASK(12, 0) - -#define ANA_STORMLIMIT_CFG_RSZ 0x4 - -#define ANA_STORMLIMIT_CFG_STORM_RATE(x) (((x) << 3) & GENMASK(6, 3)) -#define ANA_STORMLIMIT_CFG_STORM_RATE_M GENMASK(6, 3) -#define ANA_STORMLIMIT_CFG_STORM_RATE_X(x) (((x) & GENMASK(6, 3)) >> 3) -#define ANA_STORMLIMIT_CFG_STORM_UNIT BIT(2) -#define ANA_STORMLIMIT_CFG_STORM_MODE(x) ((x) & GENMASK(1, 0)) -#define ANA_STORMLIMIT_CFG_STORM_MODE_M GENMASK(1, 0) - -#define ANA_AUTOAGE_AGE_FAST BIT(21) -#define ANA_AUTOAGE_AGE_PERIOD(x) (((x) << 1) & GENMASK(20, 1)) -#define ANA_AUTOAGE_AGE_PERIOD_M GENMASK(20, 1) -#define ANA_AUTOAGE_AGE_PERIOD_X(x) (((x) & GENMASK(20, 1)) >> 1) -#define ANA_AUTOAGE_AUTOAGE_LOCKED BIT(0) - -#define ANA_MACTOPTIONS_REDUCED_TABLE BIT(1) -#define ANA_MACTOPTIONS_SHADOW BIT(0) - -#define ANA_AGENCTRL_FID_MASK(x) (((x) << 12) & GENMASK(23, 12)) -#define ANA_AGENCTRL_FID_MASK_M GENMASK(23, 12) -#define ANA_AGENCTRL_FID_MASK_X(x) (((x) & GENMASK(23, 12)) >> 12) -#define ANA_AGENCTRL_IGNORE_DMAC_FLAGS BIT(11) -#define ANA_AGENCTRL_IGNORE_SMAC_FLAGS BIT(10) -#define ANA_AGENCTRL_FLOOD_SPECIAL BIT(9) -#define ANA_AGENCTRL_FLOOD_IGNORE_VLAN BIT(8) -#define ANA_AGENCTRL_MIRROR_CPU BIT(7) -#define ANA_AGENCTRL_LEARN_CPU_COPY BIT(6) -#define ANA_AGENCTRL_LEARN_FWD_KILL BIT(5) -#define ANA_AGENCTRL_LEARN_IGNORE_VLAN BIT(4) -#define ANA_AGENCTRL_CPU_CPU_KILL_ENA BIT(3) -#define ANA_AGENCTRL_GREEN_COUNT_MODE BIT(2) -#define ANA_AGENCTRL_YELLOW_COUNT_MODE BIT(1) -#define ANA_AGENCTRL_RED_COUNT_MODE BIT(0) - -#define ANA_FLOODING_RSZ 0x4 - -#define ANA_FLOODING_FLD_UNICAST(x) (((x) << 12) & GENMASK(17, 12)) -#define ANA_FLOODING_FLD_UNICAST_M GENMASK(17, 12) -#define ANA_FLOODING_FLD_UNICAST_X(x) (((x) & GENMASK(17, 12)) >> 12) -#define ANA_FLOODING_FLD_BROADCAST(x) (((x) << 6) & GENMASK(11, 6)) -#define ANA_FLOODING_FLD_BROADCAST_M GENMASK(11, 6) -#define ANA_FLOODING_FLD_BROADCAST_X(x) (((x) & GENMASK(11, 6)) >> 6) -#define ANA_FLOODING_FLD_MULTICAST(x) ((x) & GENMASK(5, 0)) -#define ANA_FLOODING_FLD_MULTICAST_M GENMASK(5, 0) - -#define ANA_FLOODING_IPMC_FLD_MC4_CTRL(x) (((x) << 18) & GENMASK(23, 18)) -#define ANA_FLOODING_IPMC_FLD_MC4_CTRL_M GENMASK(23, 18) -#define ANA_FLOODING_IPMC_FLD_MC4_CTRL_X(x) (((x) & GENMASK(23, 18)) >> 18) -#define ANA_FLOODING_IPMC_FLD_MC4_DATA(x) (((x) << 12) & GENMASK(17, 12)) -#define ANA_FLOODING_IPMC_FLD_MC4_DATA_M GENMASK(17, 12) -#define ANA_FLOODING_IPMC_FLD_MC4_DATA_X(x) (((x) & GENMASK(17, 12)) >> 12) -#define ANA_FLOODING_IPMC_FLD_MC6_CTRL(x) (((x) << 6) & GENMASK(11, 6)) -#define ANA_FLOODING_IPMC_FLD_MC6_CTRL_M GENMASK(11, 6) -#define ANA_FLOODING_IPMC_FLD_MC6_CTRL_X(x) (((x) & GENMASK(11, 6)) >> 6) -#define ANA_FLOODING_IPMC_FLD_MC6_DATA(x) ((x) & GENMASK(5, 0)) -#define ANA_FLOODING_IPMC_FLD_MC6_DATA_M GENMASK(5, 0) - -#define ANA_SFLOW_CFG_RSZ 0x4 - -#define ANA_SFLOW_CFG_SF_RATE(x) (((x) << 2) & GENMASK(13, 2)) -#define ANA_SFLOW_CFG_SF_RATE_M GENMASK(13, 2) -#define ANA_SFLOW_CFG_SF_RATE_X(x) (((x) & GENMASK(13, 2)) >> 2) -#define ANA_SFLOW_CFG_SF_SAMPLE_RX BIT(1) -#define ANA_SFLOW_CFG_SF_SAMPLE_TX BIT(0) - -#define ANA_PORT_MODE_RSZ 0x4 - -#define ANA_PORT_MODE_REDTAG_PARSE_CFG BIT(3) -#define ANA_PORT_MODE_VLAN_PARSE_CFG(x) (((x) << 1) & GENMASK(2, 1)) -#define ANA_PORT_MODE_VLAN_PARSE_CFG_M GENMASK(2, 1) -#define ANA_PORT_MODE_VLAN_PARSE_CFG_X(x) (((x) & GENMASK(2, 1)) >> 1) -#define ANA_PORT_MODE_L3_PARSE_CFG BIT(0) - -#define ANA_CUT_THRU_CFG_RSZ 0x4 - -#define ANA_PGID_PGID_RSZ 0x4 - -#define ANA_PGID_PGID_PGID(x) ((x) & GENMASK(11, 0)) -#define ANA_PGID_PGID_PGID_M GENMASK(11, 0) -#define ANA_PGID_PGID_CPUQ_DST_PGID(x) (((x) << 27) & GENMASK(29, 27)) -#define ANA_PGID_PGID_CPUQ_DST_PGID_M GENMASK(29, 27) -#define ANA_PGID_PGID_CPUQ_DST_PGID_X(x) (((x) & GENMASK(29, 27)) >> 27) - -#define ANA_TABLES_MACHDATA_VID(x) (((x) << 16) & GENMASK(28, 16)) -#define ANA_TABLES_MACHDATA_VID_M GENMASK(28, 16) -#define ANA_TABLES_MACHDATA_VID_X(x) (((x) & GENMASK(28, 16)) >> 16) -#define ANA_TABLES_MACHDATA_MACHDATA(x) ((x) & GENMASK(15, 0)) -#define ANA_TABLES_MACHDATA_MACHDATA_M GENMASK(15, 0) - -#define ANA_TABLES_STREAMDATA_SSID_VALID BIT(16) -#define ANA_TABLES_STREAMDATA_SSID(x) (((x) << 9) & GENMASK(15, 9)) -#define ANA_TABLES_STREAMDATA_SSID_M GENMASK(15, 9) -#define ANA_TABLES_STREAMDATA_SSID_X(x) (((x) & GENMASK(15, 9)) >> 9) -#define ANA_TABLES_STREAMDATA_SFID_VALID BIT(8) -#define ANA_TABLES_STREAMDATA_SFID(x) ((x) & GENMASK(7, 0)) -#define ANA_TABLES_STREAMDATA_SFID_M GENMASK(7, 0) - -#define ANA_TABLES_MACACCESS_MAC_CPU_COPY BIT(15) -#define ANA_TABLES_MACACCESS_SRC_KILL BIT(14) -#define ANA_TABLES_MACACCESS_IGNORE_VLAN BIT(13) -#define ANA_TABLES_MACACCESS_AGED_FLAG BIT(12) -#define ANA_TABLES_MACACCESS_VALID BIT(11) -#define ANA_TABLES_MACACCESS_ENTRYTYPE(x) (((x) << 9) & GENMASK(10, 9)) -#define ANA_TABLES_MACACCESS_ENTRYTYPE_M GENMASK(10, 9) -#define ANA_TABLES_MACACCESS_ENTRYTYPE_X(x) (((x) & GENMASK(10, 9)) >> 9) -#define ANA_TABLES_MACACCESS_DEST_IDX(x) (((x) << 3) & GENMASK(8, 3)) -#define ANA_TABLES_MACACCESS_DEST_IDX_M GENMASK(8, 3) -#define ANA_TABLES_MACACCESS_DEST_IDX_X(x) (((x) & GENMASK(8, 3)) >> 3) -#define ANA_TABLES_MACACCESS_MAC_TABLE_CMD(x) ((x) & GENMASK(2, 0)) -#define ANA_TABLES_MACACCESS_MAC_TABLE_CMD_M GENMASK(2, 0) -#define MACACCESS_CMD_IDLE 0 -#define MACACCESS_CMD_LEARN 1 -#define MACACCESS_CMD_FORGET 2 -#define MACACCESS_CMD_AGE 3 -#define MACACCESS_CMD_GET_NEXT 4 -#define MACACCESS_CMD_INIT 5 -#define MACACCESS_CMD_READ 6 -#define MACACCESS_CMD_WRITE 7 - -#define ANA_TABLES_VLANACCESS_VLAN_PORT_MASK(x) (((x) << 2) & GENMASK(13, 2)) -#define ANA_TABLES_VLANACCESS_VLAN_PORT_MASK_M GENMASK(13, 2) -#define ANA_TABLES_VLANACCESS_VLAN_PORT_MASK_X(x) (((x) & GENMASK(13, 2)) >> 2) -#define ANA_TABLES_VLANACCESS_VLAN_TBL_CMD(x) ((x) & GENMASK(1, 0)) -#define ANA_TABLES_VLANACCESS_VLAN_TBL_CMD_M GENMASK(1, 0) -#define ANA_TABLES_VLANACCESS_CMD_IDLE 0x0 -#define ANA_TABLES_VLANACCESS_CMD_WRITE 0x2 -#define ANA_TABLES_VLANACCESS_CMD_INIT 0x3 - -#define ANA_TABLES_VLANTIDX_VLAN_SEC_FWD_ENA BIT(17) -#define ANA_TABLES_VLANTIDX_VLAN_FLOOD_DIS BIT(16) -#define ANA_TABLES_VLANTIDX_VLAN_PRIV_VLAN BIT(15) -#define ANA_TABLES_VLANTIDX_VLAN_LEARN_DISABLED BIT(14) -#define ANA_TABLES_VLANTIDX_VLAN_MIRROR BIT(13) -#define ANA_TABLES_VLANTIDX_VLAN_SRC_CHK BIT(12) -#define ANA_TABLES_VLANTIDX_V_INDEX(x) ((x) & GENMASK(11, 0)) -#define ANA_TABLES_VLANTIDX_V_INDEX_M GENMASK(11, 0) - -#define ANA_TABLES_ISDXACCESS_ISDX_PORT_MASK(x) (((x) << 2) & GENMASK(8, 2)) -#define ANA_TABLES_ISDXACCESS_ISDX_PORT_MASK_M GENMASK(8, 2) -#define ANA_TABLES_ISDXACCESS_ISDX_PORT_MASK_X(x) (((x) & GENMASK(8, 2)) >> 2) -#define ANA_TABLES_ISDXACCESS_ISDX_TBL_CMD(x) ((x) & GENMASK(1, 0)) -#define ANA_TABLES_ISDXACCESS_ISDX_TBL_CMD_M GENMASK(1, 0) - -#define ANA_TABLES_ISDXTIDX_ISDX_SDLBI(x) (((x) << 21) & GENMASK(28, 21)) -#define ANA_TABLES_ISDXTIDX_ISDX_SDLBI_M GENMASK(28, 21) -#define ANA_TABLES_ISDXTIDX_ISDX_SDLBI_X(x) (((x) & GENMASK(28, 21)) >> 21) -#define ANA_TABLES_ISDXTIDX_ISDX_MSTI(x) (((x) << 15) & GENMASK(20, 15)) -#define ANA_TABLES_ISDXTIDX_ISDX_MSTI_M GENMASK(20, 15) -#define ANA_TABLES_ISDXTIDX_ISDX_MSTI_X(x) (((x) & GENMASK(20, 15)) >> 15) -#define ANA_TABLES_ISDXTIDX_ISDX_ES0_KEY_ENA BIT(14) -#define ANA_TABLES_ISDXTIDX_ISDX_FORCE_ENA BIT(10) -#define ANA_TABLES_ISDXTIDX_ISDX_INDEX(x) ((x) & GENMASK(7, 0)) -#define ANA_TABLES_ISDXTIDX_ISDX_INDEX_M GENMASK(7, 0) - -#define ANA_TABLES_ENTRYLIM_RSZ 0x4 - -#define ANA_TABLES_ENTRYLIM_ENTRYLIM(x) (((x) << 14) & GENMASK(17, 14)) -#define ANA_TABLES_ENTRYLIM_ENTRYLIM_M GENMASK(17, 14) -#define ANA_TABLES_ENTRYLIM_ENTRYLIM_X(x) (((x) & GENMASK(17, 14)) >> 14) -#define ANA_TABLES_ENTRYLIM_ENTRYSTAT(x) ((x) & GENMASK(13, 0)) -#define ANA_TABLES_ENTRYLIM_ENTRYSTAT_M GENMASK(13, 0) - -#define ANA_TABLES_STREAMACCESS_GEN_REC_SEQ_NUM(x) (((x) << 4) & GENMASK(31, 4)) -#define ANA_TABLES_STREAMACCESS_GEN_REC_SEQ_NUM_M GENMASK(31, 4) -#define ANA_TABLES_STREAMACCESS_GEN_REC_SEQ_NUM_X(x) (((x) & GENMASK(31, 4)) >> 4) -#define ANA_TABLES_STREAMACCESS_SEQ_GEN_REC_ENA BIT(3) -#define ANA_TABLES_STREAMACCESS_GEN_REC_TYPE BIT(2) -#define ANA_TABLES_STREAMACCESS_STREAM_TBL_CMD(x) ((x) & GENMASK(1, 0)) -#define ANA_TABLES_STREAMACCESS_STREAM_TBL_CMD_M GENMASK(1, 0) - -#define ANA_TABLES_STREAMTIDX_SEQ_GEN_ERR_STATUS(x) (((x) << 30) & GENMASK(31, 30)) -#define ANA_TABLES_STREAMTIDX_SEQ_GEN_ERR_STATUS_M GENMASK(31, 30) -#define ANA_TABLES_STREAMTIDX_SEQ_GEN_ERR_STATUS_X(x) (((x) & GENMASK(31, 30)) >> 30) -#define ANA_TABLES_STREAMTIDX_S_INDEX(x) (((x) << 16) & GENMASK(22, 16)) -#define ANA_TABLES_STREAMTIDX_S_INDEX_M GENMASK(22, 16) -#define ANA_TABLES_STREAMTIDX_S_INDEX_X(x) (((x) & GENMASK(22, 16)) >> 16) -#define ANA_TABLES_STREAMTIDX_FORCE_SF_BEHAVIOUR BIT(14) -#define ANA_TABLES_STREAMTIDX_SEQ_HISTORY_LEN(x) (((x) << 8) & GENMASK(13, 8)) -#define ANA_TABLES_STREAMTIDX_SEQ_HISTORY_LEN_M GENMASK(13, 8) -#define ANA_TABLES_STREAMTIDX_SEQ_HISTORY_LEN_X(x) (((x) & GENMASK(13, 8)) >> 8) -#define ANA_TABLES_STREAMTIDX_RESET_ON_ROGUE BIT(7) -#define ANA_TABLES_STREAMTIDX_REDTAG_POP BIT(6) -#define ANA_TABLES_STREAMTIDX_STREAM_SPLIT BIT(5) -#define ANA_TABLES_STREAMTIDX_SEQ_SPACE_LOG2(x) ((x) & GENMASK(4, 0)) -#define ANA_TABLES_STREAMTIDX_SEQ_SPACE_LOG2_M GENMASK(4, 0) - -#define ANA_TABLES_SEQ_MASK_SPLIT_MASK(x) (((x) << 16) & GENMASK(22, 16)) -#define ANA_TABLES_SEQ_MASK_SPLIT_MASK_M GENMASK(22, 16) -#define ANA_TABLES_SEQ_MASK_SPLIT_MASK_X(x) (((x) & GENMASK(22, 16)) >> 16) -#define ANA_TABLES_SEQ_MASK_INPUT_PORT_MASK(x) ((x) & GENMASK(6, 0)) -#define ANA_TABLES_SEQ_MASK_INPUT_PORT_MASK_M GENMASK(6, 0) - -#define ANA_TABLES_SFID_MASK_IGR_PORT_MASK(x) (((x) << 1) & GENMASK(7, 1)) -#define ANA_TABLES_SFID_MASK_IGR_PORT_MASK_M GENMASK(7, 1) -#define ANA_TABLES_SFID_MASK_IGR_PORT_MASK_X(x) (((x) & GENMASK(7, 1)) >> 1) -#define ANA_TABLES_SFID_MASK_IGR_SRCPORT_MATCH_ENA BIT(0) - -#define ANA_TABLES_SFIDACCESS_IGR_PRIO_MATCH_ENA BIT(22) -#define ANA_TABLES_SFIDACCESS_IGR_PRIO(x) (((x) << 19) & GENMASK(21, 19)) -#define ANA_TABLES_SFIDACCESS_IGR_PRIO_M GENMASK(21, 19) -#define ANA_TABLES_SFIDACCESS_IGR_PRIO_X(x) (((x) & GENMASK(21, 19)) >> 19) -#define ANA_TABLES_SFIDACCESS_FORCE_BLOCK BIT(18) -#define ANA_TABLES_SFIDACCESS_MAX_SDU_LEN(x) (((x) << 2) & GENMASK(17, 2)) -#define ANA_TABLES_SFIDACCESS_MAX_SDU_LEN_M GENMASK(17, 2) -#define ANA_TABLES_SFIDACCESS_MAX_SDU_LEN_X(x) (((x) & GENMASK(17, 2)) >> 2) -#define ANA_TABLES_SFIDACCESS_SFID_TBL_CMD(x) ((x) & GENMASK(1, 0)) -#define ANA_TABLES_SFIDACCESS_SFID_TBL_CMD_M GENMASK(1, 0) - -#define ANA_TABLES_SFIDTIDX_SGID_VALID BIT(26) -#define ANA_TABLES_SFIDTIDX_SGID(x) (((x) << 18) & GENMASK(25, 18)) -#define ANA_TABLES_SFIDTIDX_SGID_M GENMASK(25, 18) -#define ANA_TABLES_SFIDTIDX_SGID_X(x) (((x) & GENMASK(25, 18)) >> 18) -#define ANA_TABLES_SFIDTIDX_POL_ENA BIT(17) -#define ANA_TABLES_SFIDTIDX_POL_IDX(x) (((x) << 8) & GENMASK(16, 8)) -#define ANA_TABLES_SFIDTIDX_POL_IDX_M GENMASK(16, 8) -#define ANA_TABLES_SFIDTIDX_POL_IDX_X(x) (((x) & GENMASK(16, 8)) >> 8) -#define ANA_TABLES_SFIDTIDX_SFID_INDEX(x) ((x) & GENMASK(7, 0)) -#define ANA_TABLES_SFIDTIDX_SFID_INDEX_M GENMASK(7, 0) - -#define ANA_MSTI_STATE_RSZ 0x4 - -#define ANA_OAM_UPM_LM_CNT_RSZ 0x4 - -#define ANA_SG_ACCESS_CTRL_SGID(x) ((x) & GENMASK(7, 0)) -#define ANA_SG_ACCESS_CTRL_SGID_M GENMASK(7, 0) -#define ANA_SG_ACCESS_CTRL_CONFIG_CHANGE BIT(28) - -#define ANA_SG_CONFIG_REG_3_BASE_TIME_SEC_MSB(x) ((x) & GENMASK(15, 0)) -#define ANA_SG_CONFIG_REG_3_BASE_TIME_SEC_MSB_M GENMASK(15, 0) -#define ANA_SG_CONFIG_REG_3_LIST_LENGTH(x) (((x) << 16) & GENMASK(18, 16)) -#define ANA_SG_CONFIG_REG_3_LIST_LENGTH_M GENMASK(18, 16) -#define ANA_SG_CONFIG_REG_3_LIST_LENGTH_X(x) (((x) & GENMASK(18, 16)) >> 16) -#define ANA_SG_CONFIG_REG_3_GATE_ENABLE BIT(20) -#define ANA_SG_CONFIG_REG_3_INIT_IPS(x) (((x) << 24) & GENMASK(27, 24)) -#define ANA_SG_CONFIG_REG_3_INIT_IPS_M GENMASK(27, 24) -#define ANA_SG_CONFIG_REG_3_INIT_IPS_X(x) (((x) & GENMASK(27, 24)) >> 24) -#define ANA_SG_CONFIG_REG_3_INIT_GATE_STATE BIT(28) - -#define ANA_SG_GCL_GS_CONFIG_RSZ 0x4 - -#define ANA_SG_GCL_GS_CONFIG_IPS(x) ((x) & GENMASK(3, 0)) -#define ANA_SG_GCL_GS_CONFIG_IPS_M GENMASK(3, 0) -#define ANA_SG_GCL_GS_CONFIG_GATE_STATE BIT(4) - -#define ANA_SG_GCL_TI_CONFIG_RSZ 0x4 - -#define ANA_SG_STATUS_REG_3_CFG_CHG_TIME_SEC_MSB(x) ((x) & GENMASK(15, 0)) -#define ANA_SG_STATUS_REG_3_CFG_CHG_TIME_SEC_MSB_M GENMASK(15, 0) -#define ANA_SG_STATUS_REG_3_GATE_STATE BIT(16) -#define ANA_SG_STATUS_REG_3_IPS(x) (((x) << 20) & GENMASK(23, 20)) -#define ANA_SG_STATUS_REG_3_IPS_M GENMASK(23, 20) -#define ANA_SG_STATUS_REG_3_IPS_X(x) (((x) & GENMASK(23, 20)) >> 20) -#define ANA_SG_STATUS_REG_3_CONFIG_PENDING BIT(24) - -#define ANA_PORT_VLAN_CFG_GSZ 0x100 - -#define ANA_PORT_VLAN_CFG_VLAN_VID_AS_ISDX BIT(21) -#define ANA_PORT_VLAN_CFG_VLAN_AWARE_ENA BIT(20) -#define ANA_PORT_VLAN_CFG_VLAN_POP_CNT(x) (((x) << 18) & GENMASK(19, 18)) -#define ANA_PORT_VLAN_CFG_VLAN_POP_CNT_M GENMASK(19, 18) -#define ANA_PORT_VLAN_CFG_VLAN_POP_CNT_X(x) (((x) & GENMASK(19, 18)) >> 18) -#define ANA_PORT_VLAN_CFG_VLAN_INNER_TAG_ENA BIT(17) -#define ANA_PORT_VLAN_CFG_VLAN_TAG_TYPE BIT(16) -#define ANA_PORT_VLAN_CFG_VLAN_DEI BIT(15) -#define ANA_PORT_VLAN_CFG_VLAN_PCP(x) (((x) << 12) & GENMASK(14, 12)) -#define ANA_PORT_VLAN_CFG_VLAN_PCP_M GENMASK(14, 12) -#define ANA_PORT_VLAN_CFG_VLAN_PCP_X(x) (((x) & GENMASK(14, 12)) >> 12) -#define ANA_PORT_VLAN_CFG_VLAN_VID(x) ((x) & GENMASK(11, 0)) -#define ANA_PORT_VLAN_CFG_VLAN_VID_M GENMASK(11, 0) - -#define ANA_PORT_DROP_CFG_GSZ 0x100 - -#define ANA_PORT_DROP_CFG_DROP_UNTAGGED_ENA BIT(6) -#define ANA_PORT_DROP_CFG_DROP_S_TAGGED_ENA BIT(5) -#define ANA_PORT_DROP_CFG_DROP_C_TAGGED_ENA BIT(4) -#define ANA_PORT_DROP_CFG_DROP_PRIO_S_TAGGED_ENA BIT(3) -#define ANA_PORT_DROP_CFG_DROP_PRIO_C_TAGGED_ENA BIT(2) -#define ANA_PORT_DROP_CFG_DROP_NULL_MAC_ENA BIT(1) -#define ANA_PORT_DROP_CFG_DROP_MC_SMAC_ENA BIT(0) - -#define ANA_PORT_QOS_CFG_GSZ 0x100 - -#define ANA_PORT_QOS_CFG_DP_DEFAULT_VAL BIT(8) -#define ANA_PORT_QOS_CFG_QOS_DEFAULT_VAL(x) (((x) << 5) & GENMASK(7, 5)) -#define ANA_PORT_QOS_CFG_QOS_DEFAULT_VAL_M GENMASK(7, 5) -#define ANA_PORT_QOS_CFG_QOS_DEFAULT_VAL_X(x) (((x) & GENMASK(7, 5)) >> 5) -#define ANA_PORT_QOS_CFG_QOS_DSCP_ENA BIT(4) -#define ANA_PORT_QOS_CFG_QOS_PCP_ENA BIT(3) -#define ANA_PORT_QOS_CFG_DSCP_TRANSLATE_ENA BIT(2) -#define ANA_PORT_QOS_CFG_DSCP_REWR_CFG(x) ((x) & GENMASK(1, 0)) -#define ANA_PORT_QOS_CFG_DSCP_REWR_CFG_M GENMASK(1, 0) - -#define ANA_PORT_VCAP_CFG_GSZ 0x100 - -#define ANA_PORT_VCAP_CFG_S1_ENA BIT(14) -#define ANA_PORT_VCAP_CFG_S1_DMAC_DIP_ENA(x) (((x) << 11) & GENMASK(13, 11)) -#define ANA_PORT_VCAP_CFG_S1_DMAC_DIP_ENA_M GENMASK(13, 11) -#define ANA_PORT_VCAP_CFG_S1_DMAC_DIP_ENA_X(x) (((x) & GENMASK(13, 11)) >> 11) -#define ANA_PORT_VCAP_CFG_S1_VLAN_INNER_TAG_ENA(x) (((x) << 8) & GENMASK(10, 8)) -#define ANA_PORT_VCAP_CFG_S1_VLAN_INNER_TAG_ENA_M GENMASK(10, 8) -#define ANA_PORT_VCAP_CFG_S1_VLAN_INNER_TAG_ENA_X(x) (((x) & GENMASK(10, 8)) >> 8) -#define ANA_PORT_VCAP_CFG_PAG_VAL(x) ((x) & GENMASK(7, 0)) -#define ANA_PORT_VCAP_CFG_PAG_VAL_M GENMASK(7, 0) - -#define ANA_PORT_VCAP_S1_KEY_CFG_GSZ 0x100 -#define ANA_PORT_VCAP_S1_KEY_CFG_RSZ 0x4 - -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP6_CFG(x) (((x) << 4) & GENMASK(6, 4)) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP6_CFG_M GENMASK(6, 4) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP6_CFG_X(x) (((x) & GENMASK(6, 4)) >> 4) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP4_CFG(x) (((x) << 2) & GENMASK(3, 2)) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP4_CFG_M GENMASK(3, 2) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_IP4_CFG_X(x) (((x) & GENMASK(3, 2)) >> 2) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_OTHER_CFG(x) ((x) & GENMASK(1, 0)) -#define ANA_PORT_VCAP_S1_KEY_CFG_S1_KEY_OTHER_CFG_M GENMASK(1, 0) - -#define ANA_PORT_VCAP_S2_CFG_GSZ 0x100 - -#define ANA_PORT_VCAP_S2_CFG_S2_UDP_PAYLOAD_ENA(x) (((x) << 17) & GENMASK(18, 17)) -#define ANA_PORT_VCAP_S2_CFG_S2_UDP_PAYLOAD_ENA_M GENMASK(18, 17) -#define ANA_PORT_VCAP_S2_CFG_S2_UDP_PAYLOAD_ENA_X(x) (((x) & GENMASK(18, 17)) >> 17) -#define ANA_PORT_VCAP_S2_CFG_S2_ETYPE_PAYLOAD_ENA(x) (((x) << 15) & GENMASK(16, 15)) -#define ANA_PORT_VCAP_S2_CFG_S2_ETYPE_PAYLOAD_ENA_M GENMASK(16, 15) -#define ANA_PORT_VCAP_S2_CFG_S2_ETYPE_PAYLOAD_ENA_X(x) (((x) & GENMASK(16, 15)) >> 15) -#define ANA_PORT_VCAP_S2_CFG_S2_ENA BIT(14) -#define ANA_PORT_VCAP_S2_CFG_S2_SNAP_DIS(x) (((x) << 12) & GENMASK(13, 12)) -#define ANA_PORT_VCAP_S2_CFG_S2_SNAP_DIS_M GENMASK(13, 12) -#define ANA_PORT_VCAP_S2_CFG_S2_SNAP_DIS_X(x) (((x) & GENMASK(13, 12)) >> 12) -#define ANA_PORT_VCAP_S2_CFG_S2_ARP_DIS(x) (((x) << 10) & GENMASK(11, 10)) -#define ANA_PORT_VCAP_S2_CFG_S2_ARP_DIS_M GENMASK(11, 10) -#define ANA_PORT_VCAP_S2_CFG_S2_ARP_DIS_X(x) (((x) & GENMASK(11, 10)) >> 10) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_TCPUDP_DIS(x) (((x) << 8) & GENMASK(9, 8)) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_TCPUDP_DIS_M GENMASK(9, 8) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_TCPUDP_DIS_X(x) (((x) & GENMASK(9, 8)) >> 8) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_OTHER_DIS(x) (((x) << 6) & GENMASK(7, 6)) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_OTHER_DIS_M GENMASK(7, 6) -#define ANA_PORT_VCAP_S2_CFG_S2_IP_OTHER_DIS_X(x) (((x) & GENMASK(7, 6)) >> 6) -#define ANA_PORT_VCAP_S2_CFG_S2_IP6_CFG(x) (((x) << 2) & GENMASK(5, 2)) -#define ANA_PORT_VCAP_S2_CFG_S2_IP6_CFG_M GENMASK(5, 2) -#define ANA_PORT_VCAP_S2_CFG_S2_IP6_CFG_X(x) (((x) & GENMASK(5, 2)) >> 2) -#define ANA_PORT_VCAP_S2_CFG_S2_OAM_DIS(x) ((x) & GENMASK(1, 0)) -#define ANA_PORT_VCAP_S2_CFG_S2_OAM_DIS_M GENMASK(1, 0) - -#define ANA_PORT_PCP_DEI_MAP_GSZ 0x100 -#define ANA_PORT_PCP_DEI_MAP_RSZ 0x4 - -#define ANA_PORT_PCP_DEI_MAP_DP_PCP_DEI_VAL BIT(3) -#define ANA_PORT_PCP_DEI_MAP_QOS_PCP_DEI_VAL(x) ((x) & GENMASK(2, 0)) -#define ANA_PORT_PCP_DEI_MAP_QOS_PCP_DEI_VAL_M GENMASK(2, 0) - -#define ANA_PORT_CPU_FWD_CFG_GSZ 0x100 - -#define ANA_PORT_CPU_FWD_CFG_CPU_VRAP_REDIR_ENA BIT(7) -#define ANA_PORT_CPU_FWD_CFG_CPU_MLD_REDIR_ENA BIT(6) -#define ANA_PORT_CPU_FWD_CFG_CPU_IGMP_REDIR_ENA BIT(5) -#define ANA_PORT_CPU_FWD_CFG_CPU_IPMC_CTRL_COPY_ENA BIT(4) -#define ANA_PORT_CPU_FWD_CFG_CPU_SRC_COPY_ENA BIT(3) -#define ANA_PORT_CPU_FWD_CFG_CPU_ALLBRIDGE_DROP_ENA BIT(2) -#define ANA_PORT_CPU_FWD_CFG_CPU_ALLBRIDGE_REDIR_ENA BIT(1) -#define ANA_PORT_CPU_FWD_CFG_CPU_OAM_ENA BIT(0) - -#define ANA_PORT_CPU_FWD_BPDU_CFG_GSZ 0x100 - -#define ANA_PORT_CPU_FWD_BPDU_CFG_BPDU_DROP_ENA(x) (((x) << 16) & GENMASK(31, 16)) -#define ANA_PORT_CPU_FWD_BPDU_CFG_BPDU_DROP_ENA_M GENMASK(31, 16) -#define ANA_PORT_CPU_FWD_BPDU_CFG_BPDU_DROP_ENA_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define ANA_PORT_CPU_FWD_BPDU_CFG_BPDU_REDIR_ENA(x) ((x) & GENMASK(15, 0)) -#define ANA_PORT_CPU_FWD_BPDU_CFG_BPDU_REDIR_ENA_M GENMASK(15, 0) - -#define ANA_PORT_CPU_FWD_GARP_CFG_GSZ 0x100 - -#define ANA_PORT_CPU_FWD_GARP_CFG_GARP_DROP_ENA(x) (((x) << 16) & GENMASK(31, 16)) -#define ANA_PORT_CPU_FWD_GARP_CFG_GARP_DROP_ENA_M GENMASK(31, 16) -#define ANA_PORT_CPU_FWD_GARP_CFG_GARP_DROP_ENA_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define ANA_PORT_CPU_FWD_GARP_CFG_GARP_REDIR_ENA(x) ((x) & GENMASK(15, 0)) -#define ANA_PORT_CPU_FWD_GARP_CFG_GARP_REDIR_ENA_M GENMASK(15, 0) - -#define ANA_PORT_CPU_FWD_CCM_CFG_GSZ 0x100 - -#define ANA_PORT_CPU_FWD_CCM_CFG_CCM_DROP_ENA(x) (((x) << 16) & GENMASK(31, 16)) -#define ANA_PORT_CPU_FWD_CCM_CFG_CCM_DROP_ENA_M GENMASK(31, 16) -#define ANA_PORT_CPU_FWD_CCM_CFG_CCM_DROP_ENA_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define ANA_PORT_CPU_FWD_CCM_CFG_CCM_REDIR_ENA(x) ((x) & GENMASK(15, 0)) -#define ANA_PORT_CPU_FWD_CCM_CFG_CCM_REDIR_ENA_M GENMASK(15, 0) - -#define ANA_PORT_PORT_CFG_GSZ 0x100 - -#define ANA_PORT_PORT_CFG_SRC_MIRROR_ENA BIT(15) -#define ANA_PORT_PORT_CFG_LIMIT_DROP BIT(14) -#define ANA_PORT_PORT_CFG_LIMIT_CPU BIT(13) -#define ANA_PORT_PORT_CFG_LOCKED_PORTMOVE_DROP BIT(12) -#define ANA_PORT_PORT_CFG_LOCKED_PORTMOVE_CPU BIT(11) -#define ANA_PORT_PORT_CFG_LEARNDROP BIT(10) -#define ANA_PORT_PORT_CFG_LEARNCPU BIT(9) -#define ANA_PORT_PORT_CFG_LEARNAUTO BIT(8) -#define ANA_PORT_PORT_CFG_LEARN_ENA BIT(7) -#define ANA_PORT_PORT_CFG_RECV_ENA BIT(6) -#define ANA_PORT_PORT_CFG_PORTID_VAL(x) (((x) << 2) & GENMASK(5, 2)) -#define ANA_PORT_PORT_CFG_PORTID_VAL_M GENMASK(5, 2) -#define ANA_PORT_PORT_CFG_PORTID_VAL_X(x) (((x) & GENMASK(5, 2)) >> 2) -#define ANA_PORT_PORT_CFG_USE_B_DOM_TBL BIT(1) -#define ANA_PORT_PORT_CFG_LSR_MODE BIT(0) - -#define ANA_PORT_POL_CFG_GSZ 0x100 - -#define ANA_PORT_POL_CFG_POL_CPU_REDIR_8021 BIT(19) -#define ANA_PORT_POL_CFG_POL_CPU_REDIR_IP BIT(18) -#define ANA_PORT_POL_CFG_PORT_POL_ENA BIT(17) -#define ANA_PORT_POL_CFG_QUEUE_POL_ENA(x) (((x) << 9) & GENMASK(16, 9)) -#define ANA_PORT_POL_CFG_QUEUE_POL_ENA_M GENMASK(16, 9) -#define ANA_PORT_POL_CFG_QUEUE_POL_ENA_X(x) (((x) & GENMASK(16, 9)) >> 9) -#define ANA_PORT_POL_CFG_POL_ORDER(x) ((x) & GENMASK(8, 0)) -#define ANA_PORT_POL_CFG_POL_ORDER_M GENMASK(8, 0) - -#define ANA_PORT_PTP_CFG_GSZ 0x100 - -#define ANA_PORT_PTP_CFG_PTP_BACKPLANE_MODE BIT(0) - -#define ANA_PORT_PTP_DLY1_CFG_GSZ 0x100 - -#define ANA_PORT_PTP_DLY2_CFG_GSZ 0x100 - -#define ANA_PORT_SFID_CFG_GSZ 0x100 -#define ANA_PORT_SFID_CFG_RSZ 0x4 - -#define ANA_PORT_SFID_CFG_SFID_VALID BIT(8) -#define ANA_PORT_SFID_CFG_SFID(x) ((x) & GENMASK(7, 0)) -#define ANA_PORT_SFID_CFG_SFID_M GENMASK(7, 0) - -#define ANA_PFC_PFC_CFG_GSZ 0x40 - -#define ANA_PFC_PFC_CFG_RX_PFC_ENA(x) (((x) << 2) & GENMASK(9, 2)) -#define ANA_PFC_PFC_CFG_RX_PFC_ENA_M GENMASK(9, 2) -#define ANA_PFC_PFC_CFG_RX_PFC_ENA_X(x) (((x) & GENMASK(9, 2)) >> 2) -#define ANA_PFC_PFC_CFG_FC_LINK_SPEED(x) ((x) & GENMASK(1, 0)) -#define ANA_PFC_PFC_CFG_FC_LINK_SPEED_M GENMASK(1, 0) - -#define ANA_PFC_PFC_TIMER_GSZ 0x40 -#define ANA_PFC_PFC_TIMER_RSZ 0x4 - -#define ANA_IPT_OAM_MEP_CFG_GSZ 0x8 - -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_P(x) (((x) << 6) & GENMASK(10, 6)) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_P_M GENMASK(10, 6) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_P_X(x) (((x) & GENMASK(10, 6)) >> 6) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX(x) (((x) << 1) & GENMASK(5, 1)) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_M GENMASK(5, 1) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_X(x) (((x) & GENMASK(5, 1)) >> 1) -#define ANA_IPT_OAM_MEP_CFG_MEP_IDX_ENA BIT(0) - -#define ANA_IPT_IPT_GSZ 0x8 - -#define ANA_IPT_IPT_IPT_CFG(x) (((x) << 15) & GENMASK(16, 15)) -#define ANA_IPT_IPT_IPT_CFG_M GENMASK(16, 15) -#define ANA_IPT_IPT_IPT_CFG_X(x) (((x) & GENMASK(16, 15)) >> 15) -#define ANA_IPT_IPT_ISDX_P(x) (((x) << 7) & GENMASK(14, 7)) -#define ANA_IPT_IPT_ISDX_P_M GENMASK(14, 7) -#define ANA_IPT_IPT_ISDX_P_X(x) (((x) & GENMASK(14, 7)) >> 7) -#define ANA_IPT_IPT_PPT_IDX(x) ((x) & GENMASK(6, 0)) -#define ANA_IPT_IPT_PPT_IDX_M GENMASK(6, 0) - -#define ANA_PPT_PPT_RSZ 0x4 - -#define ANA_FID_MAP_FID_MAP_RSZ 0x4 - -#define ANA_FID_MAP_FID_MAP_FID_C_VAL(x) (((x) << 6) & GENMASK(11, 6)) -#define ANA_FID_MAP_FID_MAP_FID_C_VAL_M GENMASK(11, 6) -#define ANA_FID_MAP_FID_MAP_FID_C_VAL_X(x) (((x) & GENMASK(11, 6)) >> 6) -#define ANA_FID_MAP_FID_MAP_FID_B_VAL(x) ((x) & GENMASK(5, 0)) -#define ANA_FID_MAP_FID_MAP_FID_B_VAL_M GENMASK(5, 0) - -#define ANA_AGGR_CFG_AC_RND_ENA BIT(7) -#define ANA_AGGR_CFG_AC_DMAC_ENA BIT(6) -#define ANA_AGGR_CFG_AC_SMAC_ENA BIT(5) -#define ANA_AGGR_CFG_AC_IP6_FLOW_LBL_ENA BIT(4) -#define ANA_AGGR_CFG_AC_IP6_TCPUDP_ENA BIT(3) -#define ANA_AGGR_CFG_AC_IP4_SIPDIP_ENA BIT(2) -#define ANA_AGGR_CFG_AC_IP4_TCPUDP_ENA BIT(1) -#define ANA_AGGR_CFG_AC_ISDX_ENA BIT(0) - -#define ANA_CPUQ_CFG_CPUQ_MLD(x) (((x) << 27) & GENMASK(29, 27)) -#define ANA_CPUQ_CFG_CPUQ_MLD_M GENMASK(29, 27) -#define ANA_CPUQ_CFG_CPUQ_MLD_X(x) (((x) & GENMASK(29, 27)) >> 27) -#define ANA_CPUQ_CFG_CPUQ_IGMP(x) (((x) << 24) & GENMASK(26, 24)) -#define ANA_CPUQ_CFG_CPUQ_IGMP_M GENMASK(26, 24) -#define ANA_CPUQ_CFG_CPUQ_IGMP_X(x) (((x) & GENMASK(26, 24)) >> 24) -#define ANA_CPUQ_CFG_CPUQ_IPMC_CTRL(x) (((x) << 21) & GENMASK(23, 21)) -#define ANA_CPUQ_CFG_CPUQ_IPMC_CTRL_M GENMASK(23, 21) -#define ANA_CPUQ_CFG_CPUQ_IPMC_CTRL_X(x) (((x) & GENMASK(23, 21)) >> 21) -#define ANA_CPUQ_CFG_CPUQ_ALLBRIDGE(x) (((x) << 18) & GENMASK(20, 18)) -#define ANA_CPUQ_CFG_CPUQ_ALLBRIDGE_M GENMASK(20, 18) -#define ANA_CPUQ_CFG_CPUQ_ALLBRIDGE_X(x) (((x) & GENMASK(20, 18)) >> 18) -#define ANA_CPUQ_CFG_CPUQ_LOCKED_PORTMOVE(x) (((x) << 15) & GENMASK(17, 15)) -#define ANA_CPUQ_CFG_CPUQ_LOCKED_PORTMOVE_M GENMASK(17, 15) -#define ANA_CPUQ_CFG_CPUQ_LOCKED_PORTMOVE_X(x) (((x) & GENMASK(17, 15)) >> 15) -#define ANA_CPUQ_CFG_CPUQ_SRC_COPY(x) (((x) << 12) & GENMASK(14, 12)) -#define ANA_CPUQ_CFG_CPUQ_SRC_COPY_M GENMASK(14, 12) -#define ANA_CPUQ_CFG_CPUQ_SRC_COPY_X(x) (((x) & GENMASK(14, 12)) >> 12) -#define ANA_CPUQ_CFG_CPUQ_MAC_COPY(x) (((x) << 9) & GENMASK(11, 9)) -#define ANA_CPUQ_CFG_CPUQ_MAC_COPY_M GENMASK(11, 9) -#define ANA_CPUQ_CFG_CPUQ_MAC_COPY_X(x) (((x) & GENMASK(11, 9)) >> 9) -#define ANA_CPUQ_CFG_CPUQ_LRN(x) (((x) << 6) & GENMASK(8, 6)) -#define ANA_CPUQ_CFG_CPUQ_LRN_M GENMASK(8, 6) -#define ANA_CPUQ_CFG_CPUQ_LRN_X(x) (((x) & GENMASK(8, 6)) >> 6) -#define ANA_CPUQ_CFG_CPUQ_MIRROR(x) (((x) << 3) & GENMASK(5, 3)) -#define ANA_CPUQ_CFG_CPUQ_MIRROR_M GENMASK(5, 3) -#define ANA_CPUQ_CFG_CPUQ_MIRROR_X(x) (((x) & GENMASK(5, 3)) >> 3) -#define ANA_CPUQ_CFG_CPUQ_SFLOW(x) ((x) & GENMASK(2, 0)) -#define ANA_CPUQ_CFG_CPUQ_SFLOW_M GENMASK(2, 0) - -#define ANA_CPUQ_8021_CFG_RSZ 0x4 - -#define ANA_CPUQ_8021_CFG_CPUQ_BPDU_VAL(x) (((x) << 6) & GENMASK(8, 6)) -#define ANA_CPUQ_8021_CFG_CPUQ_BPDU_VAL_M GENMASK(8, 6) -#define ANA_CPUQ_8021_CFG_CPUQ_BPDU_VAL_X(x) (((x) & GENMASK(8, 6)) >> 6) -#define ANA_CPUQ_8021_CFG_CPUQ_GARP_VAL(x) (((x) << 3) & GENMASK(5, 3)) -#define ANA_CPUQ_8021_CFG_CPUQ_GARP_VAL_M GENMASK(5, 3) -#define ANA_CPUQ_8021_CFG_CPUQ_GARP_VAL_X(x) (((x) & GENMASK(5, 3)) >> 3) -#define ANA_CPUQ_8021_CFG_CPUQ_CCM_VAL(x) ((x) & GENMASK(2, 0)) -#define ANA_CPUQ_8021_CFG_CPUQ_CCM_VAL_M GENMASK(2, 0) - -#define ANA_DSCP_CFG_RSZ 0x4 - -#define ANA_DSCP_CFG_DP_DSCP_VAL BIT(11) -#define ANA_DSCP_CFG_QOS_DSCP_VAL(x) (((x) << 8) & GENMASK(10, 8)) -#define ANA_DSCP_CFG_QOS_DSCP_VAL_M GENMASK(10, 8) -#define ANA_DSCP_CFG_QOS_DSCP_VAL_X(x) (((x) & GENMASK(10, 8)) >> 8) -#define ANA_DSCP_CFG_DSCP_TRANSLATE_VAL(x) (((x) << 2) & GENMASK(7, 2)) -#define ANA_DSCP_CFG_DSCP_TRANSLATE_VAL_M GENMASK(7, 2) -#define ANA_DSCP_CFG_DSCP_TRANSLATE_VAL_X(x) (((x) & GENMASK(7, 2)) >> 2) -#define ANA_DSCP_CFG_DSCP_TRUST_ENA BIT(1) -#define ANA_DSCP_CFG_DSCP_REWR_ENA BIT(0) - -#define ANA_DSCP_REWR_CFG_RSZ 0x4 - -#define ANA_VCAP_RNG_TYPE_CFG_RSZ 0x4 - -#define ANA_VCAP_RNG_VAL_CFG_RSZ 0x4 - -#define ANA_VCAP_RNG_VAL_CFG_VCAP_RNG_MIN_VAL(x) (((x) << 16) & GENMASK(31, 16)) -#define ANA_VCAP_RNG_VAL_CFG_VCAP_RNG_MIN_VAL_M GENMASK(31, 16) -#define ANA_VCAP_RNG_VAL_CFG_VCAP_RNG_MIN_VAL_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define ANA_VCAP_RNG_VAL_CFG_VCAP_RNG_MAX_VAL(x) ((x) & GENMASK(15, 0)) -#define ANA_VCAP_RNG_VAL_CFG_VCAP_RNG_MAX_VAL_M GENMASK(15, 0) - -#define ANA_VRAP_CFG_VRAP_VLAN_AWARE_ENA BIT(12) -#define ANA_VRAP_CFG_VRAP_VID(x) ((x) & GENMASK(11, 0)) -#define ANA_VRAP_CFG_VRAP_VID_M GENMASK(11, 0) - -#define ANA_DISCARD_CFG_DROP_TAGGING_ISDX0 BIT(3) -#define ANA_DISCARD_CFG_DROP_CTRLPROT_ISDX0 BIT(2) -#define ANA_DISCARD_CFG_DROP_TAGGING_S2_ENA BIT(1) -#define ANA_DISCARD_CFG_DROP_CTRLPROT_S2_ENA BIT(0) - -#define ANA_FID_CFG_VID_MC_ENA BIT(0) - -#define ANA_POL_PIR_CFG_GSZ 0x20 - -#define ANA_POL_PIR_CFG_PIR_RATE(x) (((x) << 6) & GENMASK(20, 6)) -#define ANA_POL_PIR_CFG_PIR_RATE_M GENMASK(20, 6) -#define ANA_POL_PIR_CFG_PIR_RATE_X(x) (((x) & GENMASK(20, 6)) >> 6) -#define ANA_POL_PIR_CFG_PIR_BURST(x) ((x) & GENMASK(5, 0)) -#define ANA_POL_PIR_CFG_PIR_BURST_M GENMASK(5, 0) - -#define ANA_POL_CIR_CFG_GSZ 0x20 - -#define ANA_POL_CIR_CFG_CIR_RATE(x) (((x) << 6) & GENMASK(20, 6)) -#define ANA_POL_CIR_CFG_CIR_RATE_M GENMASK(20, 6) -#define ANA_POL_CIR_CFG_CIR_RATE_X(x) (((x) & GENMASK(20, 6)) >> 6) -#define ANA_POL_CIR_CFG_CIR_BURST(x) ((x) & GENMASK(5, 0)) -#define ANA_POL_CIR_CFG_CIR_BURST_M GENMASK(5, 0) - -#define ANA_POL_MODE_CFG_GSZ 0x20 - -#define ANA_POL_MODE_CFG_IPG_SIZE(x) (((x) << 5) & GENMASK(9, 5)) -#define ANA_POL_MODE_CFG_IPG_SIZE_M GENMASK(9, 5) -#define ANA_POL_MODE_CFG_IPG_SIZE_X(x) (((x) & GENMASK(9, 5)) >> 5) -#define ANA_POL_MODE_CFG_FRM_MODE(x) (((x) << 3) & GENMASK(4, 3)) -#define ANA_POL_MODE_CFG_FRM_MODE_M GENMASK(4, 3) -#define ANA_POL_MODE_CFG_FRM_MODE_X(x) (((x) & GENMASK(4, 3)) >> 3) -#define ANA_POL_MODE_CFG_DLB_COUPLED BIT(2) -#define ANA_POL_MODE_CFG_CIR_ENA BIT(1) -#define ANA_POL_MODE_CFG_OVERSHOOT_ENA BIT(0) - -#define ANA_POL_PIR_STATE_GSZ 0x20 - -#define ANA_POL_CIR_STATE_GSZ 0x20 - -#define ANA_POL_STATE_GSZ 0x20 - -#define ANA_POL_FLOWC_RSZ 0x4 - -#define ANA_POL_FLOWC_POL_FLOWC BIT(0) - -#define ANA_POL_HYST_POL_FC_HYST(x) (((x) << 4) & GENMASK(9, 4)) -#define ANA_POL_HYST_POL_FC_HYST_M GENMASK(9, 4) -#define ANA_POL_HYST_POL_FC_HYST_X(x) (((x) & GENMASK(9, 4)) >> 4) -#define ANA_POL_HYST_POL_STOP_HYST(x) ((x) & GENMASK(3, 0)) -#define ANA_POL_HYST_POL_STOP_HYST_M GENMASK(3, 0) - -#define ANA_POL_MISC_CFG_POL_CLOSE_ALL BIT(1) -#define ANA_POL_MISC_CFG_POL_LEAK_DIS BIT(0) - -#endif diff --git a/drivers/net/ethernet/mscc/ocelot_board.c b/drivers/net/ethernet/mscc/ocelot_board.c index 2da8eee27e98..b38820849faa 100644 --- a/drivers/net/ethernet/mscc/ocelot_board.c +++ b/drivers/net/ethernet/mscc/ocelot_board.c @@ -402,9 +402,9 @@ static int mscc_ocelot_probe(struct platform_device *pdev) of_get_phy_mode(portnp, &phy_mode); - priv->phy_mode = phy_mode; + ocelot_port->phy_mode = phy_mode; - switch (priv->phy_mode) { + switch (ocelot_port->phy_mode) { case PHY_INTERFACE_MODE_NA: continue; case PHY_INTERFACE_MODE_SGMII: diff --git a/drivers/net/ethernet/mscc/ocelot_dev.h b/drivers/net/ethernet/mscc/ocelot_dev.h deleted file mode 100644 index 0a50d53bbd3f..000000000000 --- a/drivers/net/ethernet/mscc/ocelot_dev.h +++ /dev/null @@ -1,275 +0,0 @@ -/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ -/* - * Microsemi Ocelot Switch driver - * - * Copyright (c) 2017 Microsemi Corporation - */ - -#ifndef _MSCC_OCELOT_DEV_H_ -#define _MSCC_OCELOT_DEV_H_ - -#define DEV_CLOCK_CFG 0x0 - -#define DEV_CLOCK_CFG_MAC_TX_RST BIT(7) -#define DEV_CLOCK_CFG_MAC_RX_RST BIT(6) -#define DEV_CLOCK_CFG_PCS_TX_RST BIT(5) -#define DEV_CLOCK_CFG_PCS_RX_RST BIT(4) -#define DEV_CLOCK_CFG_PORT_RST BIT(3) -#define DEV_CLOCK_CFG_PHY_RST BIT(2) -#define DEV_CLOCK_CFG_LINK_SPEED(x) ((x) & GENMASK(1, 0)) -#define DEV_CLOCK_CFG_LINK_SPEED_M GENMASK(1, 0) - -#define DEV_PORT_MISC 0x4 - -#define DEV_PORT_MISC_FWD_ERROR_ENA BIT(4) -#define DEV_PORT_MISC_FWD_PAUSE_ENA BIT(3) -#define DEV_PORT_MISC_FWD_CTRL_ENA BIT(2) -#define DEV_PORT_MISC_DEV_LOOP_ENA BIT(1) -#define DEV_PORT_MISC_HDX_FAST_DIS BIT(0) - -#define DEV_EVENTS 0x8 - -#define DEV_EEE_CFG 0xc - -#define DEV_EEE_CFG_EEE_ENA BIT(22) -#define DEV_EEE_CFG_EEE_TIMER_AGE(x) (((x) << 15) & GENMASK(21, 15)) -#define DEV_EEE_CFG_EEE_TIMER_AGE_M GENMASK(21, 15) -#define DEV_EEE_CFG_EEE_TIMER_AGE_X(x) (((x) & GENMASK(21, 15)) >> 15) -#define DEV_EEE_CFG_EEE_TIMER_WAKEUP(x) (((x) << 8) & GENMASK(14, 8)) -#define DEV_EEE_CFG_EEE_TIMER_WAKEUP_M GENMASK(14, 8) -#define DEV_EEE_CFG_EEE_TIMER_WAKEUP_X(x) (((x) & GENMASK(14, 8)) >> 8) -#define DEV_EEE_CFG_EEE_TIMER_HOLDOFF(x) (((x) << 1) & GENMASK(7, 1)) -#define DEV_EEE_CFG_EEE_TIMER_HOLDOFF_M GENMASK(7, 1) -#define DEV_EEE_CFG_EEE_TIMER_HOLDOFF_X(x) (((x) & GENMASK(7, 1)) >> 1) -#define DEV_EEE_CFG_PORT_LPI BIT(0) - -#define DEV_RX_PATH_DELAY 0x10 - -#define DEV_TX_PATH_DELAY 0x14 - -#define DEV_PTP_PREDICT_CFG 0x18 - -#define DEV_PTP_PREDICT_CFG_PTP_PHY_PREDICT_CFG(x) (((x) << 4) & GENMASK(11, 4)) -#define DEV_PTP_PREDICT_CFG_PTP_PHY_PREDICT_CFG_M GENMASK(11, 4) -#define DEV_PTP_PREDICT_CFG_PTP_PHY_PREDICT_CFG_X(x) (((x) & GENMASK(11, 4)) >> 4) -#define DEV_PTP_PREDICT_CFG_PTP_PHASE_PREDICT_CFG(x) ((x) & GENMASK(3, 0)) -#define DEV_PTP_PREDICT_CFG_PTP_PHASE_PREDICT_CFG_M GENMASK(3, 0) - -#define DEV_MAC_ENA_CFG 0x1c - -#define DEV_MAC_ENA_CFG_RX_ENA BIT(4) -#define DEV_MAC_ENA_CFG_TX_ENA BIT(0) - -#define DEV_MAC_MODE_CFG 0x20 - -#define DEV_MAC_MODE_CFG_FC_WORD_SYNC_ENA BIT(8) -#define DEV_MAC_MODE_CFG_GIGA_MODE_ENA BIT(4) -#define DEV_MAC_MODE_CFG_FDX_ENA BIT(0) - -#define DEV_MAC_MAXLEN_CFG 0x24 - -#define DEV_MAC_TAGS_CFG 0x28 - -#define DEV_MAC_TAGS_CFG_TAG_ID(x) (((x) << 16) & GENMASK(31, 16)) -#define DEV_MAC_TAGS_CFG_TAG_ID_M GENMASK(31, 16) -#define DEV_MAC_TAGS_CFG_TAG_ID_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define DEV_MAC_TAGS_CFG_VLAN_LEN_AWR_ENA BIT(2) -#define DEV_MAC_TAGS_CFG_PB_ENA BIT(1) -#define DEV_MAC_TAGS_CFG_VLAN_AWR_ENA BIT(0) - -#define DEV_MAC_ADV_CHK_CFG 0x2c - -#define DEV_MAC_ADV_CHK_CFG_LEN_DROP_ENA BIT(0) - -#define DEV_MAC_IFG_CFG 0x30 - -#define DEV_MAC_IFG_CFG_RESTORE_OLD_IPG_CHECK BIT(17) -#define DEV_MAC_IFG_CFG_REDUCED_TX_IFG BIT(16) -#define DEV_MAC_IFG_CFG_TX_IFG(x) (((x) << 8) & GENMASK(12, 8)) -#define DEV_MAC_IFG_CFG_TX_IFG_M GENMASK(12, 8) -#define DEV_MAC_IFG_CFG_TX_IFG_X(x) (((x) & GENMASK(12, 8)) >> 8) -#define DEV_MAC_IFG_CFG_RX_IFG2(x) (((x) << 4) & GENMASK(7, 4)) -#define DEV_MAC_IFG_CFG_RX_IFG2_M GENMASK(7, 4) -#define DEV_MAC_IFG_CFG_RX_IFG2_X(x) (((x) & GENMASK(7, 4)) >> 4) -#define DEV_MAC_IFG_CFG_RX_IFG1(x) ((x) & GENMASK(3, 0)) -#define DEV_MAC_IFG_CFG_RX_IFG1_M GENMASK(3, 0) - -#define DEV_MAC_HDX_CFG 0x34 - -#define DEV_MAC_HDX_CFG_BYPASS_COL_SYNC BIT(26) -#define DEV_MAC_HDX_CFG_OB_ENA BIT(25) -#define DEV_MAC_HDX_CFG_WEXC_DIS BIT(24) -#define DEV_MAC_HDX_CFG_SEED(x) (((x) << 16) & GENMASK(23, 16)) -#define DEV_MAC_HDX_CFG_SEED_M GENMASK(23, 16) -#define DEV_MAC_HDX_CFG_SEED_X(x) (((x) & GENMASK(23, 16)) >> 16) -#define DEV_MAC_HDX_CFG_SEED_LOAD BIT(12) -#define DEV_MAC_HDX_CFG_RETRY_AFTER_EXC_COL_ENA BIT(8) -#define DEV_MAC_HDX_CFG_LATE_COL_POS(x) ((x) & GENMASK(6, 0)) -#define DEV_MAC_HDX_CFG_LATE_COL_POS_M GENMASK(6, 0) - -#define DEV_MAC_DBG_CFG 0x38 - -#define DEV_MAC_DBG_CFG_TBI_MODE BIT(4) -#define DEV_MAC_DBG_CFG_IFG_CRS_EXT_CHK_ENA BIT(0) - -#define DEV_MAC_FC_MAC_LOW_CFG 0x3c - -#define DEV_MAC_FC_MAC_HIGH_CFG 0x40 - -#define DEV_MAC_STICKY 0x44 - -#define DEV_MAC_STICKY_RX_IPG_SHRINK_STICKY BIT(9) -#define DEV_MAC_STICKY_RX_PREAM_SHRINK_STICKY BIT(8) -#define DEV_MAC_STICKY_RX_CARRIER_EXT_STICKY BIT(7) -#define DEV_MAC_STICKY_RX_CARRIER_EXT_ERR_STICKY BIT(6) -#define DEV_MAC_STICKY_RX_JUNK_STICKY BIT(5) -#define DEV_MAC_STICKY_TX_RETRANSMIT_STICKY BIT(4) -#define DEV_MAC_STICKY_TX_JAM_STICKY BIT(3) -#define DEV_MAC_STICKY_TX_FIFO_OFLW_STICKY BIT(2) -#define DEV_MAC_STICKY_TX_FRM_LEN_OVR_STICKY BIT(1) -#define DEV_MAC_STICKY_TX_ABORT_STICKY BIT(0) - -#define PCS1G_CFG 0x48 - -#define PCS1G_CFG_LINK_STATUS_TYPE BIT(4) -#define PCS1G_CFG_AN_LINK_CTRL_ENA BIT(1) -#define PCS1G_CFG_PCS_ENA BIT(0) - -#define PCS1G_MODE_CFG 0x4c - -#define PCS1G_MODE_CFG_UNIDIR_MODE_ENA BIT(4) -#define PCS1G_MODE_CFG_SGMII_MODE_ENA BIT(0) - -#define PCS1G_SD_CFG 0x50 - -#define PCS1G_SD_CFG_SD_SEL BIT(8) -#define PCS1G_SD_CFG_SD_POL BIT(4) -#define PCS1G_SD_CFG_SD_ENA BIT(0) - -#define PCS1G_ANEG_CFG 0x54 - -#define PCS1G_ANEG_CFG_ADV_ABILITY(x) (((x) << 16) & GENMASK(31, 16)) -#define PCS1G_ANEG_CFG_ADV_ABILITY_M GENMASK(31, 16) -#define PCS1G_ANEG_CFG_ADV_ABILITY_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define PCS1G_ANEG_CFG_SW_RESOLVE_ENA BIT(8) -#define PCS1G_ANEG_CFG_ANEG_RESTART_ONE_SHOT BIT(1) -#define PCS1G_ANEG_CFG_ANEG_ENA BIT(0) - -#define PCS1G_ANEG_NP_CFG 0x58 - -#define PCS1G_ANEG_NP_CFG_NP_TX(x) (((x) << 16) & GENMASK(31, 16)) -#define PCS1G_ANEG_NP_CFG_NP_TX_M GENMASK(31, 16) -#define PCS1G_ANEG_NP_CFG_NP_TX_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define PCS1G_ANEG_NP_CFG_NP_LOADED_ONE_SHOT BIT(0) - -#define PCS1G_LB_CFG 0x5c - -#define PCS1G_LB_CFG_RA_ENA BIT(4) -#define PCS1G_LB_CFG_GMII_PHY_LB_ENA BIT(1) -#define PCS1G_LB_CFG_TBI_HOST_LB_ENA BIT(0) - -#define PCS1G_DBG_CFG 0x60 - -#define PCS1G_DBG_CFG_UDLT BIT(0) - -#define PCS1G_CDET_CFG 0x64 - -#define PCS1G_CDET_CFG_CDET_ENA BIT(0) - -#define PCS1G_ANEG_STATUS 0x68 - -#define PCS1G_ANEG_STATUS_LP_ADV_ABILITY(x) (((x) << 16) & GENMASK(31, 16)) -#define PCS1G_ANEG_STATUS_LP_ADV_ABILITY_M GENMASK(31, 16) -#define PCS1G_ANEG_STATUS_LP_ADV_ABILITY_X(x) (((x) & GENMASK(31, 16)) >> 16) -#define PCS1G_ANEG_STATUS_PR BIT(4) -#define PCS1G_ANEG_STATUS_PAGE_RX_STICKY BIT(3) -#define PCS1G_ANEG_STATUS_ANEG_COMPLETE BIT(0) - -#define PCS1G_ANEG_NP_STATUS 0x6c - -#define PCS1G_LINK_STATUS 0x70 - -#define PCS1G_LINK_STATUS_DELAY_VAR(x) (((x) << 12) & GENMASK(15, 12)) -#define PCS1G_LINK_STATUS_DELAY_VAR_M GENMASK(15, 12) -#define PCS1G_LINK_STATUS_DELAY_VAR_X(x) (((x) & GENMASK(15, 12)) >> 12) -#define PCS1G_LINK_STATUS_SIGNAL_DETECT BIT(8) -#define PCS1G_LINK_STATUS_LINK_STATUS BIT(4) -#define PCS1G_LINK_STATUS_SYNC_STATUS BIT(0) - -#define PCS1G_LINK_DOWN_CNT 0x74 - -#define PCS1G_STICKY 0x78 - -#define PCS1G_STICKY_LINK_DOWN_STICKY BIT(4) -#define PCS1G_STICKY_OUT_OF_SYNC_STICKY BIT(0) - -#define PCS1G_DEBUG_STATUS 0x7c - -#define PCS1G_LPI_CFG 0x80 - -#define PCS1G_LPI_CFG_QSGMII_MS_SEL BIT(20) -#define PCS1G_LPI_CFG_RX_LPI_OUT_DIS BIT(17) -#define PCS1G_LPI_CFG_LPI_TESTMODE BIT(16) -#define PCS1G_LPI_CFG_LPI_RX_WTIM(x) (((x) << 4) & GENMASK(5, 4)) -#define PCS1G_LPI_CFG_LPI_RX_WTIM_M GENMASK(5, 4) -#define PCS1G_LPI_CFG_LPI_RX_WTIM_X(x) (((x) & GENMASK(5, 4)) >> 4) -#define PCS1G_LPI_CFG_TX_ASSERT_LPIDLE BIT(0) - -#define PCS1G_LPI_WAKE_ERROR_CNT 0x84 - -#define PCS1G_LPI_STATUS 0x88 - -#define PCS1G_LPI_STATUS_RX_LPI_FAIL BIT(16) -#define PCS1G_LPI_STATUS_RX_LPI_EVENT_STICKY BIT(12) -#define PCS1G_LPI_STATUS_RX_QUIET BIT(9) -#define PCS1G_LPI_STATUS_RX_LPI_MODE BIT(8) -#define PCS1G_LPI_STATUS_TX_LPI_EVENT_STICKY BIT(4) -#define PCS1G_LPI_STATUS_TX_QUIET BIT(1) -#define PCS1G_LPI_STATUS_TX_LPI_MODE BIT(0) - -#define PCS1G_TSTPAT_MODE_CFG 0x8c - -#define PCS1G_TSTPAT_STATUS 0x90 - -#define PCS1G_TSTPAT_STATUS_JTP_ERR_CNT(x) (((x) << 8) & GENMASK(15, 8)) -#define PCS1G_TSTPAT_STATUS_JTP_ERR_CNT_M GENMASK(15, 8) -#define PCS1G_TSTPAT_STATUS_JTP_ERR_CNT_X(x) (((x) & GENMASK(15, 8)) >> 8) -#define PCS1G_TSTPAT_STATUS_JTP_ERR BIT(4) -#define PCS1G_TSTPAT_STATUS_JTP_LOCK BIT(0) - -#define DEV_PCS_FX100_CFG 0x94 - -#define DEV_PCS_FX100_CFG_SD_SEL BIT(26) -#define DEV_PCS_FX100_CFG_SD_POL BIT(25) -#define DEV_PCS_FX100_CFG_SD_ENA BIT(24) -#define DEV_PCS_FX100_CFG_LOOPBACK_ENA BIT(20) -#define DEV_PCS_FX100_CFG_SWAP_MII_ENA BIT(16) -#define DEV_PCS_FX100_CFG_RXBITSEL(x) (((x) << 12) & GENMASK(15, 12)) -#define DEV_PCS_FX100_CFG_RXBITSEL_M GENMASK(15, 12) -#define DEV_PCS_FX100_CFG_RXBITSEL_X(x) (((x) & GENMASK(15, 12)) >> 12) -#define DEV_PCS_FX100_CFG_SIGDET_CFG(x) (((x) << 9) & GENMASK(10, 9)) -#define DEV_PCS_FX100_CFG_SIGDET_CFG_M GENMASK(10, 9) -#define DEV_PCS_FX100_CFG_SIGDET_CFG_X(x) (((x) & GENMASK(10, 9)) >> 9) -#define DEV_PCS_FX100_CFG_LINKHYST_TM_ENA BIT(8) -#define DEV_PCS_FX100_CFG_LINKHYSTTIMER(x) (((x) << 4) & GENMASK(7, 4)) -#define DEV_PCS_FX100_CFG_LINKHYSTTIMER_M GENMASK(7, 4) -#define DEV_PCS_FX100_CFG_LINKHYSTTIMER_X(x) (((x) & GENMASK(7, 4)) >> 4) -#define DEV_PCS_FX100_CFG_UNIDIR_MODE_ENA BIT(3) -#define DEV_PCS_FX100_CFG_FEFCHK_ENA BIT(2) -#define DEV_PCS_FX100_CFG_FEFGEN_ENA BIT(1) -#define DEV_PCS_FX100_CFG_PCS_ENA BIT(0) - -#define DEV_PCS_FX100_STATUS 0x98 - -#define DEV_PCS_FX100_STATUS_EDGE_POS_PTP(x) (((x) << 8) & GENMASK(11, 8)) -#define DEV_PCS_FX100_STATUS_EDGE_POS_PTP_M GENMASK(11, 8) -#define DEV_PCS_FX100_STATUS_EDGE_POS_PTP_X(x) (((x) & GENMASK(11, 8)) >> 8) -#define DEV_PCS_FX100_STATUS_PCS_ERROR_STICKY BIT(7) -#define DEV_PCS_FX100_STATUS_FEF_FOUND_STICKY BIT(6) -#define DEV_PCS_FX100_STATUS_SSD_ERROR_STICKY BIT(5) -#define DEV_PCS_FX100_STATUS_SYNC_LOST_STICKY BIT(4) -#define DEV_PCS_FX100_STATUS_FEF_STATUS BIT(2) -#define DEV_PCS_FX100_STATUS_SIGNAL_DETECT BIT(1) -#define DEV_PCS_FX100_STATUS_SYNC_STATUS BIT(0) - -#endif diff --git a/drivers/net/ethernet/mscc/ocelot_qsys.h b/drivers/net/ethernet/mscc/ocelot_qsys.h deleted file mode 100644 index d8c63aa761be..000000000000 --- a/drivers/net/ethernet/mscc/ocelot_qsys.h +++ /dev/null @@ -1,270 +0,0 @@ -/* SPDX-License-Identifier: (GPL-2.0 OR MIT) */ -/* - * Microsemi Ocelot Switch driver - * - * Copyright (c) 2017 Microsemi Corporation - */ - -#ifndef _MSCC_OCELOT_QSYS_H_ -#define _MSCC_OCELOT_QSYS_H_ - -#define QSYS_PORT_MODE_RSZ 0x4 - -#define QSYS_PORT_MODE_DEQUEUE_DIS BIT(1) -#define QSYS_PORT_MODE_DEQUEUE_LATE BIT(0) - -#define QSYS_SWITCH_PORT_MODE_RSZ 0x4 - -#define QSYS_SWITCH_PORT_MODE_PORT_ENA BIT(14) -#define QSYS_SWITCH_PORT_MODE_SCH_NEXT_CFG(x) (((x) << 11) & GENMASK(13, 11)) -#define QSYS_SWITCH_PORT_MODE_SCH_NEXT_CFG_M GENMASK(13, 11) -#define QSYS_SWITCH_PORT_MODE_SCH_NEXT_CFG_X(x) (((x) & GENMASK(13, 11)) >> 11) -#define QSYS_SWITCH_PORT_MODE_YEL_RSRVD BIT(10) -#define QSYS_SWITCH_PORT_MODE_INGRESS_DROP_MODE BIT(9) -#define QSYS_SWITCH_PORT_MODE_TX_PFC_ENA(x) (((x) << 1) & GENMASK(8, 1)) -#define QSYS_SWITCH_PORT_MODE_TX_PFC_ENA_M GENMASK(8, 1) -#define QSYS_SWITCH_PORT_MODE_TX_PFC_ENA_X(x) (((x) & GENMASK(8, 1)) >> 1) -#define QSYS_SWITCH_PORT_MODE_TX_PFC_MODE BIT(0) - -#define QSYS_STAT_CNT_CFG_TX_GREEN_CNT_MODE BIT(5) -#define QSYS_STAT_CNT_CFG_TX_YELLOW_CNT_MODE BIT(4) -#define QSYS_STAT_CNT_CFG_DROP_GREEN_CNT_MODE BIT(3) -#define QSYS_STAT_CNT_CFG_DROP_YELLOW_CNT_MODE BIT(2) -#define QSYS_STAT_CNT_CFG_DROP_COUNT_ONCE BIT(1) -#define QSYS_STAT_CNT_CFG_DROP_COUNT_EGRESS BIT(0) - -#define QSYS_EEE_CFG_RSZ 0x4 - -#define QSYS_EEE_THRES_EEE_HIGH_BYTES(x) (((x) << 8) & GENMASK(15, 8)) -#define QSYS_EEE_THRES_EEE_HIGH_BYTES_M GENMASK(15, 8) -#define QSYS_EEE_THRES_EEE_HIGH_BYTES_X(x) (((x) & GENMASK(15, 8)) >> 8) -#define QSYS_EEE_THRES_EEE_HIGH_FRAMES(x) ((x) & GENMASK(7, 0)) -#define QSYS_EEE_THRES_EEE_HIGH_FRAMES_M GENMASK(7, 0) - -#define QSYS_SW_STATUS_RSZ 0x4 - -#define QSYS_EXT_CPU_CFG_EXT_CPU_PORT(x) (((x) << 8) & GENMASK(12, 8)) -#define QSYS_EXT_CPU_CFG_EXT_CPU_PORT_M GENMASK(12, 8) -#define QSYS_EXT_CPU_CFG_EXT_CPU_PORT_X(x) (((x) & GENMASK(12, 8)) >> 8) -#define QSYS_EXT_CPU_CFG_EXT_CPUQ_MSK(x) ((x) & GENMASK(7, 0)) -#define QSYS_EXT_CPU_CFG_EXT_CPUQ_MSK_M GENMASK(7, 0) - -#define QSYS_QMAP_GSZ 0x4 - -#define QSYS_QMAP_SE_BASE(x) (((x) << 5) & GENMASK(12, 5)) -#define QSYS_QMAP_SE_BASE_M GENMASK(12, 5) -#define QSYS_QMAP_SE_BASE_X(x) (((x) & GENMASK(12, 5)) >> 5) -#define QSYS_QMAP_SE_IDX_SEL(x) (((x) << 2) & GENMASK(4, 2)) -#define QSYS_QMAP_SE_IDX_SEL_M GENMASK(4, 2) -#define QSYS_QMAP_SE_IDX_SEL_X(x) (((x) & GENMASK(4, 2)) >> 2) -#define QSYS_QMAP_SE_INP_SEL(x) ((x) & GENMASK(1, 0)) -#define QSYS_QMAP_SE_INP_SEL_M GENMASK(1, 0) - -#define QSYS_ISDX_SGRP_GSZ 0x4 - -#define QSYS_TIMED_FRAME_ENTRY_GSZ 0x4 - -#define QSYS_TFRM_MISC_TIMED_CANCEL_SLOT(x) (((x) << 9) & GENMASK(18, 9)) -#define QSYS_TFRM_MISC_TIMED_CANCEL_SLOT_M GENMASK(18, 9) -#define QSYS_TFRM_MISC_TIMED_CANCEL_SLOT_X(x) (((x) & GENMASK(18, 9)) >> 9) -#define QSYS_TFRM_MISC_TIMED_CANCEL_1SHOT BIT(8) -#define QSYS_TFRM_MISC_TIMED_SLOT_MODE_MC BIT(7) -#define QSYS_TFRM_MISC_TIMED_ENTRY_FAST_CNT(x) ((x) & GENMASK(6, 0)) -#define QSYS_TFRM_MISC_TIMED_ENTRY_FAST_CNT_M GENMASK(6, 0) - -#define QSYS_RED_PROFILE_RSZ 0x4 - -#define QSYS_RED_PROFILE_WM_RED_LOW(x) (((x) << 8) & GENMASK(15, 8)) -#define QSYS_RED_PROFILE_WM_RED_LOW_M GENMASK(15, 8) -#define QSYS_RED_PROFILE_WM_RED_LOW_X(x) (((x) & GENMASK(15, 8)) >> 8) -#define QSYS_RED_PROFILE_WM_RED_HIGH(x) ((x) & GENMASK(7, 0)) -#define QSYS_RED_PROFILE_WM_RED_HIGH_M GENMASK(7, 0) - -#define QSYS_RES_CFG_GSZ 0x8 - -#define QSYS_RES_STAT_GSZ 0x8 - -#define QSYS_RES_STAT_INUSE(x) (((x) << 12) & GENMASK(23, 12)) -#define QSYS_RES_STAT_INUSE_M GENMASK(23, 12) -#define QSYS_RES_STAT_INUSE_X(x) (((x) & GENMASK(23, 12)) >> 12) -#define QSYS_RES_STAT_MAXUSE(x) ((x) & GENMASK(11, 0)) -#define QSYS_RES_STAT_MAXUSE_M GENMASK(11, 0) - -#define QSYS_EVENTS_CORE_EV_FDC(x) (((x) << 2) & GENMASK(4, 2)) -#define QSYS_EVENTS_CORE_EV_FDC_M GENMASK(4, 2) -#define QSYS_EVENTS_CORE_EV_FDC_X(x) (((x) & GENMASK(4, 2)) >> 2) -#define QSYS_EVENTS_CORE_EV_FRD(x) ((x) & GENMASK(1, 0)) -#define QSYS_EVENTS_CORE_EV_FRD_M GENMASK(1, 0) - -#define QSYS_QMAXSDU_CFG_0_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_1_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_2_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_3_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_4_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_5_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_6_RSZ 0x4 - -#define QSYS_QMAXSDU_CFG_7_RSZ 0x4 - -#define QSYS_PREEMPTION_CFG_RSZ 0x4 - -#define QSYS_PREEMPTION_CFG_P_QUEUES(x) ((x) & GENMASK(7, 0)) -#define QSYS_PREEMPTION_CFG_P_QUEUES_M GENMASK(7, 0) -#define QSYS_PREEMPTION_CFG_MM_ADD_FRAG_SIZE(x) (((x) << 8) & GENMASK(9, 8)) -#define QSYS_PREEMPTION_CFG_MM_ADD_FRAG_SIZE_M GENMASK(9, 8) -#define QSYS_PREEMPTION_CFG_MM_ADD_FRAG_SIZE_X(x) (((x) & GENMASK(9, 8)) >> 8) -#define QSYS_PREEMPTION_CFG_STRICT_IPG(x) (((x) << 12) & GENMASK(13, 12)) -#define QSYS_PREEMPTION_CFG_STRICT_IPG_M GENMASK(13, 12) -#define QSYS_PREEMPTION_CFG_STRICT_IPG_X(x) (((x) & GENMASK(13, 12)) >> 12) -#define QSYS_PREEMPTION_CFG_HOLD_ADVANCE(x) (((x) << 16) & GENMASK(31, 16)) -#define QSYS_PREEMPTION_CFG_HOLD_ADVANCE_M GENMASK(31, 16) -#define QSYS_PREEMPTION_CFG_HOLD_ADVANCE_X(x) (((x) & GENMASK(31, 16)) >> 16) - -#define QSYS_CIR_CFG_GSZ 0x80 - -#define QSYS_CIR_CFG_CIR_RATE(x) (((x) << 6) & GENMASK(20, 6)) -#define QSYS_CIR_CFG_CIR_RATE_M GENMASK(20, 6) -#define QSYS_CIR_CFG_CIR_RATE_X(x) (((x) & GENMASK(20, 6)) >> 6) -#define QSYS_CIR_CFG_CIR_BURST(x) ((x) & GENMASK(5, 0)) -#define QSYS_CIR_CFG_CIR_BURST_M GENMASK(5, 0) - -#define QSYS_EIR_CFG_GSZ 0x80 - -#define QSYS_EIR_CFG_EIR_RATE(x) (((x) << 7) & GENMASK(21, 7)) -#define QSYS_EIR_CFG_EIR_RATE_M GENMASK(21, 7) -#define QSYS_EIR_CFG_EIR_RATE_X(x) (((x) & GENMASK(21, 7)) >> 7) -#define QSYS_EIR_CFG_EIR_BURST(x) (((x) << 1) & GENMASK(6, 1)) -#define QSYS_EIR_CFG_EIR_BURST_M GENMASK(6, 1) -#define QSYS_EIR_CFG_EIR_BURST_X(x) (((x) & GENMASK(6, 1)) >> 1) -#define QSYS_EIR_CFG_EIR_MARK_ENA BIT(0) - -#define QSYS_SE_CFG_GSZ 0x80 - -#define QSYS_SE_CFG_SE_DWRR_CNT(x) (((x) << 6) & GENMASK(9, 6)) -#define QSYS_SE_CFG_SE_DWRR_CNT_M GENMASK(9, 6) -#define QSYS_SE_CFG_SE_DWRR_CNT_X(x) (((x) & GENMASK(9, 6)) >> 6) -#define QSYS_SE_CFG_SE_RR_ENA BIT(5) -#define QSYS_SE_CFG_SE_AVB_ENA BIT(4) -#define QSYS_SE_CFG_SE_FRM_MODE(x) (((x) << 2) & GENMASK(3, 2)) -#define QSYS_SE_CFG_SE_FRM_MODE_M GENMASK(3, 2) -#define QSYS_SE_CFG_SE_FRM_MODE_X(x) (((x) & GENMASK(3, 2)) >> 2) -#define QSYS_SE_CFG_SE_EXC_ENA BIT(1) -#define QSYS_SE_CFG_SE_EXC_FWD BIT(0) - -#define QSYS_SE_DWRR_CFG_GSZ 0x80 -#define QSYS_SE_DWRR_CFG_RSZ 0x4 - -#define QSYS_SE_CONNECT_GSZ 0x80 - -#define QSYS_SE_CONNECT_SE_OUTP_IDX(x) (((x) << 17) & GENMASK(24, 17)) -#define QSYS_SE_CONNECT_SE_OUTP_IDX_M GENMASK(24, 17) -#define QSYS_SE_CONNECT_SE_OUTP_IDX_X(x) (((x) & GENMASK(24, 17)) >> 17) -#define QSYS_SE_CONNECT_SE_INP_IDX(x) (((x) << 9) & GENMASK(16, 9)) -#define QSYS_SE_CONNECT_SE_INP_IDX_M GENMASK(16, 9) -#define QSYS_SE_CONNECT_SE_INP_IDX_X(x) (((x) & GENMASK(16, 9)) >> 9) -#define QSYS_SE_CONNECT_SE_OUTP_CON(x) (((x) << 5) & GENMASK(8, 5)) -#define QSYS_SE_CONNECT_SE_OUTP_CON_M GENMASK(8, 5) -#define QSYS_SE_CONNECT_SE_OUTP_CON_X(x) (((x) & GENMASK(8, 5)) >> 5) -#define QSYS_SE_CONNECT_SE_INP_CNT(x) (((x) << 1) & GENMASK(4, 1)) -#define QSYS_SE_CONNECT_SE_INP_CNT_M GENMASK(4, 1) -#define QSYS_SE_CONNECT_SE_INP_CNT_X(x) (((x) & GENMASK(4, 1)) >> 1) -#define QSYS_SE_CONNECT_SE_TERMINAL BIT(0) - -#define QSYS_SE_DLB_SENSE_GSZ 0x80 - -#define QSYS_SE_DLB_SENSE_SE_DLB_PRIO(x) (((x) << 11) & GENMASK(13, 11)) -#define QSYS_SE_DLB_SENSE_SE_DLB_PRIO_M GENMASK(13, 11) -#define QSYS_SE_DLB_SENSE_SE_DLB_PRIO_X(x) (((x) & GENMASK(13, 11)) >> 11) -#define QSYS_SE_DLB_SENSE_SE_DLB_SPORT(x) (((x) << 7) & GENMASK(10, 7)) -#define QSYS_SE_DLB_SENSE_SE_DLB_SPORT_M GENMASK(10, 7) -#define QSYS_SE_DLB_SENSE_SE_DLB_SPORT_X(x) (((x) & GENMASK(10, 7)) >> 7) -#define QSYS_SE_DLB_SENSE_SE_DLB_DPORT(x) (((x) << 3) & GENMASK(6, 3)) -#define QSYS_SE_DLB_SENSE_SE_DLB_DPORT_M GENMASK(6, 3) -#define QSYS_SE_DLB_SENSE_SE_DLB_DPORT_X(x) (((x) & GENMASK(6, 3)) >> 3) -#define QSYS_SE_DLB_SENSE_SE_DLB_PRIO_ENA BIT(2) -#define QSYS_SE_DLB_SENSE_SE_DLB_SPORT_ENA BIT(1) -#define QSYS_SE_DLB_SENSE_SE_DLB_DPORT_ENA BIT(0) - -#define QSYS_CIR_STATE_GSZ 0x80 - -#define QSYS_CIR_STATE_CIR_LVL(x) (((x) << 4) & GENMASK(25, 4)) -#define QSYS_CIR_STATE_CIR_LVL_M GENMASK(25, 4) -#define QSYS_CIR_STATE_CIR_LVL_X(x) (((x) & GENMASK(25, 4)) >> 4) -#define QSYS_CIR_STATE_SHP_TIME(x) ((x) & GENMASK(3, 0)) -#define QSYS_CIR_STATE_SHP_TIME_M GENMASK(3, 0) - -#define QSYS_EIR_STATE_GSZ 0x80 - -#define QSYS_SE_STATE_GSZ 0x80 - -#define QSYS_SE_STATE_SE_OUTP_LVL(x) (((x) << 1) & GENMASK(2, 1)) -#define QSYS_SE_STATE_SE_OUTP_LVL_M GENMASK(2, 1) -#define QSYS_SE_STATE_SE_OUTP_LVL_X(x) (((x) & GENMASK(2, 1)) >> 1) -#define QSYS_SE_STATE_SE_WAS_YEL BIT(0) - -#define QSYS_HSCH_MISC_CFG_SE_CONNECT_VLD BIT(8) -#define QSYS_HSCH_MISC_CFG_FRM_ADJ(x) (((x) << 3) & GENMASK(7, 3)) -#define QSYS_HSCH_MISC_CFG_FRM_ADJ_M GENMASK(7, 3) -#define QSYS_HSCH_MISC_CFG_FRM_ADJ_X(x) (((x) & GENMASK(7, 3)) >> 3) -#define QSYS_HSCH_MISC_CFG_LEAK_DIS BIT(2) -#define QSYS_HSCH_MISC_CFG_QSHP_EXC_ENA BIT(1) -#define QSYS_HSCH_MISC_CFG_PFC_BYP_UPD BIT(0) - -#define QSYS_TAG_CONFIG_RSZ 0x4 - -#define QSYS_TAG_CONFIG_ENABLE BIT(0) -#define QSYS_TAG_CONFIG_LINK_SPEED(x) (((x) << 4) & GENMASK(5, 4)) -#define QSYS_TAG_CONFIG_LINK_SPEED_M GENMASK(5, 4) -#define QSYS_TAG_CONFIG_LINK_SPEED_X(x) (((x) & GENMASK(5, 4)) >> 4) -#define QSYS_TAG_CONFIG_INIT_GATE_STATE(x) (((x) << 8) & GENMASK(15, 8)) -#define QSYS_TAG_CONFIG_INIT_GATE_STATE_M GENMASK(15, 8) -#define QSYS_TAG_CONFIG_INIT_GATE_STATE_X(x) (((x) & GENMASK(15, 8)) >> 8) -#define QSYS_TAG_CONFIG_SCH_TRAFFIC_QUEUES(x) (((x) << 16) & GENMASK(23, 16)) -#define QSYS_TAG_CONFIG_SCH_TRAFFIC_QUEUES_M GENMASK(23, 16) -#define QSYS_TAG_CONFIG_SCH_TRAFFIC_QUEUES_X(x) (((x) & GENMASK(23, 16)) >> 16) - -#define QSYS_TAS_PARAM_CFG_CTRL_PORT_NUM(x) ((x) & GENMASK(7, 0)) -#define QSYS_TAS_PARAM_CFG_CTRL_PORT_NUM_M GENMASK(7, 0) -#define QSYS_TAS_PARAM_CFG_CTRL_ALWAYS_GUARD_BAND_SCH_Q BIT(8) -#define QSYS_TAS_PARAM_CFG_CTRL_CONFIG_CHANGE BIT(16) - -#define QSYS_PORT_MAX_SDU_RSZ 0x4 - -#define QSYS_PARAM_CFG_REG_3_BASE_TIME_SEC_MSB(x) ((x) & GENMASK(15, 0)) -#define QSYS_PARAM_CFG_REG_3_BASE_TIME_SEC_MSB_M GENMASK(15, 0) -#define QSYS_PARAM_CFG_REG_3_LIST_LENGTH(x) (((x) << 16) & GENMASK(31, 16)) -#define QSYS_PARAM_CFG_REG_3_LIST_LENGTH_M GENMASK(31, 16) -#define QSYS_PARAM_CFG_REG_3_LIST_LENGTH_X(x) (((x) & GENMASK(31, 16)) >> 16) - -#define QSYS_GCL_CFG_REG_1_GCL_ENTRY_NUM(x) ((x) & GENMASK(5, 0)) -#define QSYS_GCL_CFG_REG_1_GCL_ENTRY_NUM_M GENMASK(5, 0) -#define QSYS_GCL_CFG_REG_1_GATE_STATE(x) (((x) << 8) & GENMASK(15, 8)) -#define QSYS_GCL_CFG_REG_1_GATE_STATE_M GENMASK(15, 8) -#define QSYS_GCL_CFG_REG_1_GATE_STATE_X(x) (((x) & GENMASK(15, 8)) >> 8) - -#define QSYS_PARAM_STATUS_REG_3_BASE_TIME_SEC_MSB(x) ((x) & GENMASK(15, 0)) -#define QSYS_PARAM_STATUS_REG_3_BASE_TIME_SEC_MSB_M GENMASK(15, 0) -#define QSYS_PARAM_STATUS_REG_3_LIST_LENGTH(x) (((x) << 16) & GENMASK(31, 16)) -#define QSYS_PARAM_STATUS_REG_3_LIST_LENGTH_M GENMASK(31, 16) -#define QSYS_PARAM_STATUS_REG_3_LIST_LENGTH_X(x) (((x) & GENMASK(31, 16)) >> 16) - -#define QSYS_PARAM_STATUS_REG_8_CFG_CHG_TIME_SEC_MSB(x) ((x) & GENMASK(15, 0)) -#define QSYS_PARAM_STATUS_REG_8_CFG_CHG_TIME_SEC_MSB_M GENMASK(15, 0) -#define QSYS_PARAM_STATUS_REG_8_OPER_GATE_STATE(x) (((x) << 16) & GENMASK(23, 16)) -#define QSYS_PARAM_STATUS_REG_8_OPER_GATE_STATE_M GENMASK(23, 16) -#define QSYS_PARAM_STATUS_REG_8_OPER_GATE_STATE_X(x) (((x) & GENMASK(23, 16)) >> 16) -#define QSYS_PARAM_STATUS_REG_8_CONFIG_PENDING BIT(24) - -#define QSYS_GCL_STATUS_REG_1_GCL_ENTRY_NUM(x) ((x) & GENMASK(5, 0)) -#define QSYS_GCL_STATUS_REG_1_GCL_ENTRY_NUM_M GENMASK(5, 0) -#define QSYS_GCL_STATUS_REG_1_GATE_STATE(x) (((x) << 8) & GENMASK(15, 8)) -#define QSYS_GCL_STATUS_REG_1_GATE_STATE_M GENMASK(15, 8) -#define QSYS_GCL_STATUS_REG_1_GATE_STATE_X(x) (((x) & GENMASK(15, 8)) >> 8) - -#endif diff --git a/drivers/net/ethernet/myricom/myri10ge/myri10ge.c b/drivers/net/ethernet/myricom/myri10ge/myri10ge.c index c979f38a2e0c..2ee0d0be113a 100644 --- a/drivers/net/ethernet/myricom/myri10ge/myri10ge.c +++ b/drivers/net/ethernet/myricom/myri10ge/myri10ge.c @@ -2892,7 +2892,7 @@ drop: static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, struct net_device *dev) { - struct sk_buff *segs, *curr; + struct sk_buff *segs, *curr, *next; struct myri10ge_priv *mgp = netdev_priv(dev); struct myri10ge_slice_state *ss; netdev_tx_t status; @@ -2901,10 +2901,8 @@ static netdev_tx_t myri10ge_sw_tso(struct sk_buff *skb, if (IS_ERR(segs)) goto drop; - while (segs) { - curr = segs; - segs = segs->next; - curr->next = NULL; + skb_list_walk_safe(segs, curr, next) { + skb_mark_not_on_list(curr); status = myri10ge_xmit(curr, dev); if (status != 0) { dev_kfree_skb_any(curr); diff --git a/drivers/net/ethernet/natsemi/natsemi.c b/drivers/net/ethernet/natsemi/natsemi.c index 1a2634cbbb69..d21d706b83a7 100644 --- a/drivers/net/ethernet/natsemi/natsemi.c +++ b/drivers/net/ethernet/natsemi/natsemi.c @@ -612,7 +612,7 @@ static void undo_cable_magic(struct net_device *dev); static void check_link(struct net_device *dev); static void netdev_timer(struct timer_list *t); static void dump_ring(struct net_device *dev); -static void ns_tx_timeout(struct net_device *dev); +static void ns_tx_timeout(struct net_device *dev, unsigned int txqueue); static int alloc_ring(struct net_device *dev); static void refill_rx(struct net_device *dev); static void init_ring(struct net_device *dev); @@ -1881,7 +1881,7 @@ static void dump_ring(struct net_device *dev) } } -static void ns_tx_timeout(struct net_device *dev) +static void ns_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct netdev_private *np = netdev_priv(dev); void __iomem * ioaddr = ns_ioaddr(dev); diff --git a/drivers/net/ethernet/natsemi/ns83820.c b/drivers/net/ethernet/natsemi/ns83820.c index 091254061052..8e24c7acf79b 100644 --- a/drivers/net/ethernet/natsemi/ns83820.c +++ b/drivers/net/ethernet/natsemi/ns83820.c @@ -1549,7 +1549,7 @@ static int ns83820_stop(struct net_device *ndev) return 0; } -static void ns83820_tx_timeout(struct net_device *ndev) +static void ns83820_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct ns83820 *dev = PRIV(ndev); u32 tx_done_idx; @@ -1603,7 +1603,7 @@ static void ns83820_tx_watch(struct timer_list *t) ndev->name, dev->tx_done_idx, dev->tx_free_idx, atomic_read(&dev->nr_tx_skbs)); - ns83820_tx_timeout(ndev); + ns83820_tx_timeout(ndev, UINT_MAX); } mod_timer(&dev->tx_watchdog, jiffies + 2*HZ); diff --git a/drivers/net/ethernet/natsemi/sonic.c b/drivers/net/ethernet/natsemi/sonic.c index 05e760444a92..31be3ba66877 100644 --- a/drivers/net/ethernet/natsemi/sonic.c +++ b/drivers/net/ethernet/natsemi/sonic.c @@ -184,7 +184,7 @@ static int sonic_close(struct net_device *dev) return 0; } -static void sonic_tx_timeout(struct net_device *dev) +static void sonic_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sonic_local *lp = netdev_priv(dev); int i; diff --git a/drivers/net/ethernet/natsemi/sonic.h b/drivers/net/ethernet/natsemi/sonic.h index 1df6d2f06cc4..e0e4cba6f6f6 100644 --- a/drivers/net/ethernet/natsemi/sonic.h +++ b/drivers/net/ethernet/natsemi/sonic.h @@ -340,7 +340,7 @@ static int sonic_close(struct net_device *dev); static struct net_device_stats *sonic_get_stats(struct net_device *dev); static void sonic_multicast_list(struct net_device *dev); static int sonic_init(struct net_device *dev); -static void sonic_tx_timeout(struct net_device *dev); +static void sonic_tx_timeout(struct net_device *dev, unsigned int txqueue); static void sonic_msg_init(struct net_device *dev); /* Internal inlines for reading/writing DMA buffers. Note that bus diff --git a/drivers/net/ethernet/neterion/s2io.c b/drivers/net/ethernet/neterion/s2io.c index e0b2bf327905..0ec6b8e8b549 100644 --- a/drivers/net/ethernet/neterion/s2io.c +++ b/drivers/net/ethernet/neterion/s2io.c @@ -7238,7 +7238,7 @@ out_unlock: * void */ -static void s2io_tx_watchdog(struct net_device *dev) +static void s2io_tx_watchdog(struct net_device *dev, unsigned int txqueue) { struct s2io_nic *sp = netdev_priv(dev); struct swStat *swstats = &sp->mac_control.stats_info->sw_stat; diff --git a/drivers/net/ethernet/neterion/s2io.h b/drivers/net/ethernet/neterion/s2io.h index 0a921f30f98f..6fa3159a977f 100644 --- a/drivers/net/ethernet/neterion/s2io.h +++ b/drivers/net/ethernet/neterion/s2io.h @@ -1065,7 +1065,7 @@ static void s2io_txpic_intr_handle(struct s2io_nic *sp); static void tx_intr_handler(struct fifo_info *fifo_data); static void s2io_handle_errors(void * dev_id); -static void s2io_tx_watchdog(struct net_device *dev); +static void s2io_tx_watchdog(struct net_device *dev, unsigned int txqueue); static void s2io_set_multicast(struct net_device *dev); static int rx_osm_handler(struct ring_info *ring_data, struct RxD_t * rxdp); static void s2io_link(struct s2io_nic * sp, int link); diff --git a/drivers/net/ethernet/neterion/vxge/vxge-main.c b/drivers/net/ethernet/neterion/vxge/vxge-main.c index 1d334f2e0a56..9b63574b6202 100644 --- a/drivers/net/ethernet/neterion/vxge/vxge-main.c +++ b/drivers/net/ethernet/neterion/vxge/vxge-main.c @@ -3273,7 +3273,7 @@ static int vxge_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) * This function is triggered if the Tx Queue is stopped * for a pre-defined amount of time when the Interface is still up. */ -static void vxge_tx_watchdog(struct net_device *dev) +static void vxge_tx_watchdog(struct net_device *dev, unsigned int txqueue) { struct vxgedev *vdev; diff --git a/drivers/net/ethernet/netronome/Kconfig b/drivers/net/ethernet/netronome/Kconfig index bac5be4d4f43..a3f68a718813 100644 --- a/drivers/net/ethernet/netronome/Kconfig +++ b/drivers/net/ethernet/netronome/Kconfig @@ -31,6 +31,7 @@ config NFP_APP_FLOWER bool "NFP4000/NFP6000 TC Flower offload support" depends on NFP depends on NET_SWITCHDEV + depends on IPV6!=m || NFP=m default y ---help--- Enable driver support for TC Flower offload on NFP4000 and NFP6000. diff --git a/drivers/net/ethernet/netronome/nfp/abm/cls.c b/drivers/net/ethernet/netronome/nfp/abm/cls.c index 9f8a1f69c0c4..23ebddfb9532 100644 --- a/drivers/net/ethernet/netronome/nfp/abm/cls.c +++ b/drivers/net/ethernet/netronome/nfp/abm/cls.c @@ -176,10 +176,8 @@ nfp_abm_u32_knode_replace(struct nfp_abm_link *alink, u8 mask, val; int err; - if (!nfp_abm_u32_check_knode(alink->abm, knode, proto, extack)) { - err = -EOPNOTSUPP; + if (!nfp_abm_u32_check_knode(alink->abm, knode, proto, extack)) goto err_delete; - } tos_off = proto == htons(ETH_P_IP) ? 16 : 20; @@ -200,18 +198,14 @@ nfp_abm_u32_knode_replace(struct nfp_abm_link *alink, if ((iter->val & cmask) == (val & cmask) && iter->band != knode->res->classid) { NL_SET_ERR_MSG_MOD(extack, "conflict with already offloaded filter"); - err = -EOPNOTSUPP; goto err_delete; } } if (!match) { match = kzalloc(sizeof(*match), GFP_KERNEL); - if (!match) { - err = -ENOMEM; - goto err_delete; - } - + if (!match) + return -ENOMEM; list_add(&match->list, &alink->dscp_map); } match->handle = knode->handle; @@ -227,7 +221,7 @@ nfp_abm_u32_knode_replace(struct nfp_abm_link *alink, err_delete: nfp_abm_u32_knode_delete(alink, knode); - return err; + return -EOPNOTSUPP; } static int nfp_abm_setup_tc_block_cb(enum tc_setup_type type, diff --git a/drivers/net/ethernet/netronome/nfp/ccm.h b/drivers/net/ethernet/netronome/nfp/ccm.h index a460c75522be..d81d450be50e 100644 --- a/drivers/net/ethernet/netronome/nfp/ccm.h +++ b/drivers/net/ethernet/netronome/nfp/ccm.h @@ -26,6 +26,7 @@ enum nfp_ccm_type { NFP_CCM_TYPE_CRYPTO_ADD = 10, NFP_CCM_TYPE_CRYPTO_DEL = 11, NFP_CCM_TYPE_CRYPTO_UPDATE = 12, + NFP_CCM_TYPE_CRYPTO_RESYNC = 13, __NFP_CCM_TYPE_MAX, }; diff --git a/drivers/net/ethernet/netronome/nfp/crypto/crypto.h b/drivers/net/ethernet/netronome/nfp/crypto/crypto.h index 60372ddf69f0..bffe58bb2f27 100644 --- a/drivers/net/ethernet/netronome/nfp/crypto/crypto.h +++ b/drivers/net/ethernet/netronome/nfp/crypto/crypto.h @@ -4,6 +4,10 @@ #ifndef NFP_CRYPTO_H #define NFP_CRYPTO_H 1 +struct net_device; +struct nfp_net; +struct nfp_net_tls_resync_req; + struct nfp_net_tls_offload_ctx { __be32 fw_handle[2]; @@ -17,11 +21,22 @@ struct nfp_net_tls_offload_ctx { #ifdef CONFIG_TLS_DEVICE int nfp_net_tls_init(struct nfp_net *nn); +int nfp_net_tls_rx_resync_req(struct net_device *netdev, + struct nfp_net_tls_resync_req *req, + void *pkt, unsigned int pkt_len); #else static inline int nfp_net_tls_init(struct nfp_net *nn) { return 0; } + +static inline int +nfp_net_tls_rx_resync_req(struct net_device *netdev, + struct nfp_net_tls_resync_req *req, + void *pkt, unsigned int pkt_len) +{ + return -EOPNOTSUPP; +} #endif #endif diff --git a/drivers/net/ethernet/netronome/nfp/crypto/fw.h b/drivers/net/ethernet/netronome/nfp/crypto/fw.h index 67413d946c4a..8d1458896bcb 100644 --- a/drivers/net/ethernet/netronome/nfp/crypto/fw.h +++ b/drivers/net/ethernet/netronome/nfp/crypto/fw.h @@ -9,6 +9,14 @@ #define NFP_NET_CRYPTO_OP_TLS_1_2_AES_GCM_128_ENC 0 #define NFP_NET_CRYPTO_OP_TLS_1_2_AES_GCM_128_DEC 1 +struct nfp_net_tls_resync_req { + __be32 fw_handle[2]; + __be32 tcp_seq; + u8 l3_offset; + u8 l4_offset; + u8 resv[2]; +}; + struct nfp_crypto_reply_simple { struct nfp_ccm_hdr hdr; __be32 error; diff --git a/drivers/net/ethernet/netronome/nfp/crypto/tls.c b/drivers/net/ethernet/netronome/nfp/crypto/tls.c index 96a96b35c0ca..7c50e3dfb9d5 100644 --- a/drivers/net/ethernet/netronome/nfp/crypto/tls.c +++ b/drivers/net/ethernet/netronome/nfp/crypto/tls.c @@ -5,6 +5,7 @@ #include <linux/ipv6.h> #include <linux/skbuff.h> #include <linux/string.h> +#include <net/inet6_hashtables.h> #include <net/tls.h> #include "../ccm.h" @@ -391,8 +392,9 @@ nfp_net_tls_add(struct net_device *netdev, struct sock *sk, if (direction == TLS_OFFLOAD_CTX_DIR_TX) return 0; - tls_offload_rx_resync_set_type(sk, - TLS_OFFLOAD_SYNC_TYPE_CORE_NEXT_HINT); + if (!nn->tlv_caps.tls_resync_ss) + tls_offload_rx_resync_set_type(sk, TLS_OFFLOAD_SYNC_TYPE_CORE_NEXT_HINT); + return 0; err_fw_remove: @@ -424,6 +426,7 @@ nfp_net_tls_resync(struct net_device *netdev, struct sock *sk, u32 seq, struct nfp_net *nn = netdev_priv(netdev); struct nfp_net_tls_offload_ctx *ntls; struct nfp_crypto_req_update *req; + enum nfp_ccm_type type; struct sk_buff *skb; gfp_t flags; int err; @@ -442,15 +445,18 @@ nfp_net_tls_resync(struct net_device *netdev, struct sock *sk, u32 seq, req->tcp_seq = cpu_to_be32(seq); memcpy(req->rec_no, rcd_sn, sizeof(req->rec_no)); + type = NFP_CCM_TYPE_CRYPTO_UPDATE; if (direction == TLS_OFFLOAD_CTX_DIR_TX) { - err = nfp_net_tls_communicate_simple(nn, skb, "sync", - NFP_CCM_TYPE_CRYPTO_UPDATE); + err = nfp_net_tls_communicate_simple(nn, skb, "sync", type); if (err) return err; ntls->next_seq = seq; } else { - nfp_ccm_mbox_post(nn, skb, NFP_CCM_TYPE_CRYPTO_UPDATE, + if (nn->tlv_caps.tls_resync_ss) + type = NFP_CCM_TYPE_CRYPTO_RESYNC; + nfp_ccm_mbox_post(nn, skb, type, sizeof(struct nfp_crypto_reply_simple)); + atomic_inc(&nn->ktls_rx_resync_sent); } return 0; @@ -462,6 +468,79 @@ static const struct tlsdev_ops nfp_net_tls_ops = { .tls_dev_resync = nfp_net_tls_resync, }; +int nfp_net_tls_rx_resync_req(struct net_device *netdev, + struct nfp_net_tls_resync_req *req, + void *pkt, unsigned int pkt_len) +{ + struct nfp_net *nn = netdev_priv(netdev); + struct nfp_net_tls_offload_ctx *ntls; + struct ipv6hdr *ipv6h; + struct tcphdr *th; + struct iphdr *iph; + struct sock *sk; + __be32 tcp_seq; + int err; + + iph = pkt + req->l3_offset; + ipv6h = pkt + req->l3_offset; + th = pkt + req->l4_offset; + + if ((u8 *)&th[1] > (u8 *)pkt + pkt_len) { + netdev_warn_once(netdev, "invalid TLS RX resync request (l3_off: %hhu l4_off: %hhu pkt_len: %u)\n", + req->l3_offset, req->l4_offset, pkt_len); + err = -EINVAL; + goto err_cnt_ign; + } + + switch (iph->version) { + case 4: + sk = inet_lookup_established(dev_net(netdev), &tcp_hashinfo, + iph->saddr, th->source, iph->daddr, + th->dest, netdev->ifindex); + break; +#if IS_ENABLED(CONFIG_IPV6) + case 6: + sk = __inet6_lookup_established(dev_net(netdev), &tcp_hashinfo, + &ipv6h->saddr, th->source, + &ipv6h->daddr, ntohs(th->dest), + netdev->ifindex, 0); + break; +#endif + default: + netdev_warn_once(netdev, "invalid TLS RX resync request (l3_off: %hhu l4_off: %hhu ipver: %u)\n", + req->l3_offset, req->l4_offset, iph->version); + err = -EINVAL; + goto err_cnt_ign; + } + + err = 0; + if (!sk) + goto err_cnt_ign; + if (!tls_is_sk_rx_device_offloaded(sk) || + sk->sk_shutdown & RCV_SHUTDOWN) + goto err_put_sock; + + ntls = tls_driver_ctx(sk, TLS_OFFLOAD_CTX_DIR_RX); + /* some FW versions can't report the handle and report 0s */ + if (memchr_inv(&req->fw_handle, 0, sizeof(req->fw_handle)) && + memcmp(&req->fw_handle, &ntls->fw_handle, sizeof(ntls->fw_handle))) + goto err_put_sock; + + /* copy to ensure alignment */ + memcpy(&tcp_seq, &req->tcp_seq, sizeof(tcp_seq)); + tls_offload_rx_resync_request(sk, tcp_seq); + atomic_inc(&nn->ktls_rx_resync_req); + + sock_gen_put(sk); + return 0; + +err_put_sock: + sock_gen_put(sk); +err_cnt_ign: + atomic_inc(&nn->ktls_rx_resync_ign); + return err; +} + static int nfp_net_tls_reset(struct nfp_net *nn) { struct nfp_crypto_req_reset *req; diff --git a/drivers/net/ethernet/netronome/nfp/flower/action.c b/drivers/net/ethernet/netronome/nfp/flower/action.c index 1b019fdfcd97..c06600fb47ff 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/action.c +++ b/drivers/net/ethernet/netronome/nfp/flower/action.c @@ -22,8 +22,9 @@ #define NFP_FL_TUNNEL_CSUM cpu_to_be16(0x01) #define NFP_FL_TUNNEL_KEY cpu_to_be16(0x04) #define NFP_FL_TUNNEL_GENEVE_OPT cpu_to_be16(0x0800) -#define NFP_FL_SUPPORTED_TUNNEL_INFO_FLAGS IP_TUNNEL_INFO_TX -#define NFP_FL_SUPPORTED_IPV4_UDP_TUN_FLAGS (NFP_FL_TUNNEL_CSUM | \ +#define NFP_FL_SUPPORTED_TUNNEL_INFO_FLAGS (IP_TUNNEL_INFO_TX | \ + IP_TUNNEL_INFO_IPV6) +#define NFP_FL_SUPPORTED_UDP_TUN_FLAGS (NFP_FL_TUNNEL_CSUM | \ NFP_FL_TUNNEL_KEY | \ NFP_FL_TUNNEL_GENEVE_OPT) @@ -394,19 +395,26 @@ nfp_fl_push_geneve_options(struct nfp_fl_payload *nfp_fl, int *list_len, } static int -nfp_fl_set_ipv4_tun(struct nfp_app *app, struct nfp_fl_set_ipv4_tun *set_tun, - const struct flow_action_entry *act, - struct nfp_fl_pre_tunnel *pre_tun, - enum nfp_flower_tun_type tun_type, - struct net_device *netdev, struct netlink_ext_ack *extack) +nfp_fl_set_tun(struct nfp_app *app, struct nfp_fl_set_tun *set_tun, + const struct flow_action_entry *act, + struct nfp_fl_pre_tunnel *pre_tun, + enum nfp_flower_tun_type tun_type, + struct net_device *netdev, struct netlink_ext_ack *extack) { - size_t act_size = sizeof(struct nfp_fl_set_ipv4_tun); const struct ip_tunnel_info *ip_tun = act->tunnel; + bool ipv6 = ip_tunnel_info_af(ip_tun) == AF_INET6; + size_t act_size = sizeof(struct nfp_fl_set_tun); struct nfp_flower_priv *priv = app->priv; u32 tmp_set_ip_tun_type_index = 0; /* Currently support one pre-tunnel so index is always 0. */ int pretun_idx = 0; + if (!IS_ENABLED(CONFIG_IPV6) && ipv6) + return -EOPNOTSUPP; + + if (ipv6 && !(priv->flower_ext_feats & NFP_FL_FEATS_IPV6_TUN)) + return -EOPNOTSUPP; + BUILD_BUG_ON(NFP_FL_TUNNEL_CSUM != TUNNEL_CSUM || NFP_FL_TUNNEL_KEY != TUNNEL_KEY || NFP_FL_TUNNEL_GENEVE_OPT != TUNNEL_GENEVE_OPT); @@ -417,19 +425,35 @@ nfp_fl_set_ipv4_tun(struct nfp_app *app, struct nfp_fl_set_ipv4_tun *set_tun, return -EOPNOTSUPP; } - set_tun->head.jump_id = NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL; + set_tun->head.jump_id = NFP_FL_ACTION_OPCODE_SET_TUNNEL; set_tun->head.len_lw = act_size >> NFP_FL_LW_SIZ; /* Set tunnel type and pre-tunnel index. */ tmp_set_ip_tun_type_index |= - FIELD_PREP(NFP_FL_IPV4_TUNNEL_TYPE, tun_type) | - FIELD_PREP(NFP_FL_IPV4_PRE_TUN_INDEX, pretun_idx); + FIELD_PREP(NFP_FL_TUNNEL_TYPE, tun_type) | + FIELD_PREP(NFP_FL_PRE_TUN_INDEX, pretun_idx); set_tun->tun_type_index = cpu_to_be32(tmp_set_ip_tun_type_index); set_tun->tun_id = ip_tun->key.tun_id; if (ip_tun->key.ttl) { set_tun->ttl = ip_tun->key.ttl; +#ifdef CONFIG_IPV6 + } else if (ipv6) { + struct net *net = dev_net(netdev); + struct flowi6 flow = {}; + struct dst_entry *dst; + + flow.daddr = ip_tun->key.u.ipv6.dst; + flow.flowi4_proto = IPPROTO_UDP; + dst = ipv6_stub->ipv6_dst_lookup_flow(net, NULL, &flow, NULL); + if (!IS_ERR(dst)) { + set_tun->ttl = ip6_dst_hoplimit(dst); + dst_release(dst); + } else { + set_tun->ttl = net->ipv6.devconf_all->hop_limit; + } +#endif } else { struct net *net = dev_net(netdev); struct flowi4 flow = {}; @@ -455,7 +479,7 @@ nfp_fl_set_ipv4_tun(struct nfp_app *app, struct nfp_fl_set_ipv4_tun *set_tun, set_tun->tos = ip_tun->key.tos; if (!(ip_tun->key.tun_flags & NFP_FL_TUNNEL_KEY) || - ip_tun->key.tun_flags & ~NFP_FL_SUPPORTED_IPV4_UDP_TUN_FLAGS) { + ip_tun->key.tun_flags & ~NFP_FL_SUPPORTED_UDP_TUN_FLAGS) { NL_SET_ERR_MSG_MOD(extack, "unsupported offload: loaded firmware does not support tunnel flag offload"); return -EOPNOTSUPP; } @@ -467,7 +491,12 @@ nfp_fl_set_ipv4_tun(struct nfp_app *app, struct nfp_fl_set_ipv4_tun *set_tun, } /* Complete pre_tunnel action. */ - pre_tun->ipv4_dst = ip_tun->key.u.ipv4.dst; + if (ipv6) { + pre_tun->flags |= cpu_to_be16(NFP_FL_PRE_TUN_IPV6); + pre_tun->ipv6_dst = ip_tun->key.u.ipv6.dst; + } else { + pre_tun->ipv4_dst = ip_tun->key.u.ipv4.dst; + } return 0; } @@ -956,8 +985,8 @@ nfp_flower_loop_action(struct nfp_app *app, const struct flow_action_entry *act, struct nfp_flower_pedit_acts *set_act, bool *pkt_host, struct netlink_ext_ack *extack, int act_idx) { - struct nfp_fl_set_ipv4_tun *set_tun; struct nfp_fl_pre_tunnel *pre_tun; + struct nfp_fl_set_tun *set_tun; struct nfp_fl_push_vlan *psh_v; struct nfp_fl_push_mpls *psh_m; struct nfp_fl_pop_vlan *pop_v; @@ -1032,7 +1061,7 @@ nfp_flower_loop_action(struct nfp_app *app, const struct flow_action_entry *act, * If none, the packet falls back before applying other actions. */ if (*a_len + sizeof(struct nfp_fl_pre_tunnel) + - sizeof(struct nfp_fl_set_ipv4_tun) > NFP_FL_MAX_A_SIZ) { + sizeof(struct nfp_fl_set_tun) > NFP_FL_MAX_A_SIZ) { NL_SET_ERR_MSG_MOD(extack, "unsupported offload: maximum allowed action list size exceeded at tunnel encap"); return -EOPNOTSUPP; } @@ -1046,11 +1075,11 @@ nfp_flower_loop_action(struct nfp_app *app, const struct flow_action_entry *act, return err; set_tun = (void *)&nfp_fl->action_data[*a_len]; - err = nfp_fl_set_ipv4_tun(app, set_tun, act, pre_tun, - *tun_type, netdev, extack); + err = nfp_fl_set_tun(app, set_tun, act, pre_tun, *tun_type, + netdev, extack); if (err) return err; - *a_len += sizeof(struct nfp_fl_set_ipv4_tun); + *a_len += sizeof(struct nfp_fl_set_tun); } break; case FLOW_ACTION_TUNNEL_DECAP: diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c index 05981b54eaab..a595ddb92bff 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.c +++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.c @@ -270,11 +270,17 @@ nfp_flower_cmsg_process_one_rx(struct nfp_app *app, struct sk_buff *skb) } goto err_default; case NFP_FLOWER_CMSG_TYPE_NO_NEIGH: - nfp_tunnel_request_route(app, skb); + nfp_tunnel_request_route_v4(app, skb); + break; + case NFP_FLOWER_CMSG_TYPE_NO_NEIGH_V6: + nfp_tunnel_request_route_v6(app, skb); break; case NFP_FLOWER_CMSG_TYPE_ACTIVE_TUNS: nfp_tunnel_keep_alive(app, skb); break; + case NFP_FLOWER_CMSG_TYPE_ACTIVE_TUNS_V6: + nfp_tunnel_keep_alive_v6(app, skb); + break; case NFP_FLOWER_CMSG_TYPE_QOS_STATS: nfp_flower_stats_rlim_reply(app, skb); break; @@ -361,7 +367,8 @@ void nfp_flower_cmsg_rx(struct nfp_app *app, struct sk_buff *skb) nfp_flower_process_mtu_ack(app, skb)) { /* Handle MTU acks outside wq to prevent RTNL conflict. */ dev_consume_skb_any(skb); - } else if (cmsg_hdr->type == NFP_FLOWER_CMSG_TYPE_TUN_NEIGH) { + } else if (cmsg_hdr->type == NFP_FLOWER_CMSG_TYPE_TUN_NEIGH || + cmsg_hdr->type == NFP_FLOWER_CMSG_TYPE_TUN_NEIGH_V6) { /* Acks from the NFP that the route is added - ignore. */ dev_consume_skb_any(skb); } else if (cmsg_hdr->type == NFP_FLOWER_CMSG_TYPE_PORT_REIFY) { diff --git a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h index 7eb2ec8969c3..9b50d76bbc09 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/cmsg.h +++ b/drivers/net/ethernet/netronome/nfp/flower/cmsg.h @@ -26,6 +26,7 @@ #define NFP_FLOWER_LAYER2_GRE BIT(0) #define NFP_FLOWER_LAYER2_GENEVE BIT(5) #define NFP_FLOWER_LAYER2_GENEVE_OP BIT(6) +#define NFP_FLOWER_LAYER2_TUN_IPV6 BIT(7) #define NFP_FLOWER_MASK_VLAN_PRIO GENMASK(15, 13) #define NFP_FLOWER_MASK_VLAN_PRESENT BIT(12) @@ -63,6 +64,7 @@ #define NFP_FL_MAX_GENEVE_OPT_ACT 32 #define NFP_FL_MAX_GENEVE_OPT_CNT 64 #define NFP_FL_MAX_GENEVE_OPT_KEY 32 +#define NFP_FL_MAX_GENEVE_OPT_KEY_V6 8 /* Action opcodes */ #define NFP_FL_ACTION_OPCODE_OUTPUT 0 @@ -70,7 +72,7 @@ #define NFP_FL_ACTION_OPCODE_POP_VLAN 2 #define NFP_FL_ACTION_OPCODE_PUSH_MPLS 3 #define NFP_FL_ACTION_OPCODE_POP_MPLS 4 -#define NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL 6 +#define NFP_FL_ACTION_OPCODE_SET_TUNNEL 6 #define NFP_FL_ACTION_OPCODE_SET_ETHERNET 7 #define NFP_FL_ACTION_OPCODE_SET_MPLS 8 #define NFP_FL_ACTION_OPCODE_SET_IPV4_ADDRS 9 @@ -99,8 +101,8 @@ /* Tunnel ports */ #define NFP_FL_PORT_TYPE_TUN 0x50000000 -#define NFP_FL_IPV4_TUNNEL_TYPE GENMASK(7, 4) -#define NFP_FL_IPV4_PRE_TUN_INDEX GENMASK(2, 0) +#define NFP_FL_TUNNEL_TYPE GENMASK(7, 4) +#define NFP_FL_PRE_TUN_INDEX GENMASK(2, 0) #define NFP_FLOWER_WORKQ_MAX_SKBS 30000 @@ -206,13 +208,16 @@ struct nfp_fl_pre_lag { struct nfp_fl_pre_tunnel { struct nfp_fl_act_head head; - __be16 reserved; - __be32 ipv4_dst; - /* reserved for use with IPv6 addresses */ - __be32 extra[3]; + __be16 flags; + union { + __be32 ipv4_dst; + struct in6_addr ipv6_dst; + }; }; -struct nfp_fl_set_ipv4_tun { +#define NFP_FL_PRE_TUN_IPV6 BIT(0) + +struct nfp_fl_set_tun { struct nfp_fl_act_head head; __be16 reserved; __be64 tun_id __packed; @@ -387,6 +392,11 @@ struct nfp_flower_tun_ipv4 { __be32 dst; }; +struct nfp_flower_tun_ipv6 { + struct in6_addr src; + struct in6_addr dst; +}; + struct nfp_flower_tun_ip_ext { u8 tos; u8 ttl; @@ -416,6 +426,42 @@ struct nfp_flower_ipv4_udp_tun { __be32 tun_id; }; +/* Flow Frame IPv6 UDP TUNNEL --> Tunnel details (11W/44B) + * ----------------------------------------------------------------- + * 3 2 1 + * 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 31 - 0 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 63 - 32 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 95 - 64 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 127 - 96 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 31 - 0 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 63 - 32 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 95 - 64 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 127 - 96 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Reserved | tos | ttl | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Reserved | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | VNI | Reserved | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + */ +struct nfp_flower_ipv6_udp_tun { + struct nfp_flower_tun_ipv6 ipv6; + __be16 reserved1; + struct nfp_flower_tun_ip_ext ip_ext; + __be32 reserved2; + __be32 tun_id; +}; + /* Flow Frame GRE TUNNEL --> Tunnel details (6W/24B) * ----------------------------------------------------------------- * 3 2 1 @@ -445,6 +491,46 @@ struct nfp_flower_ipv4_gre_tun { __be32 reserved2; }; +/* Flow Frame GRE TUNNEL V6 --> Tunnel details (12W/48B) + * ----------------------------------------------------------------- + * 3 2 1 + * 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 9 8 7 6 5 4 3 2 1 0 + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 31 - 0 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 63 - 32 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 95 - 64 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_src, 127 - 96 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 31 - 0 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 63 - 32 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 95 - 64 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | ipv6_addr_dst, 127 - 96 | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | tun_flags | tos | ttl | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Reserved | Ethertype | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Key | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + * | Reserved | + * +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ + */ +struct nfp_flower_ipv6_gre_tun { + struct nfp_flower_tun_ipv6 ipv6; + __be16 tun_flags; + struct nfp_flower_tun_ip_ext ip_ext; + __be16 reserved1; + __be16 ethertype; + __be32 tun_key; + __be32 reserved2; +}; + struct nfp_flower_geneve_options { u8 data[NFP_FL_MAX_GENEVE_OPT_KEY]; }; @@ -485,6 +571,10 @@ enum nfp_flower_cmsg_type_port { NFP_FLOWER_CMSG_TYPE_QOS_DEL = 19, NFP_FLOWER_CMSG_TYPE_QOS_STATS = 20, NFP_FLOWER_CMSG_TYPE_PRE_TUN_RULE = 21, + NFP_FLOWER_CMSG_TYPE_TUN_IPS_V6 = 22, + NFP_FLOWER_CMSG_TYPE_NO_NEIGH_V6 = 23, + NFP_FLOWER_CMSG_TYPE_TUN_NEIGH_V6 = 24, + NFP_FLOWER_CMSG_TYPE_ACTIVE_TUNS_V6 = 25, NFP_FLOWER_CMSG_TYPE_MAX = 32, }; diff --git a/drivers/net/ethernet/netronome/nfp/flower/main.h b/drivers/net/ethernet/netronome/nfp/flower/main.h index e0c985fcaec1..d55d0d33bc45 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/main.h +++ b/drivers/net/ethernet/netronome/nfp/flower/main.h @@ -43,6 +43,7 @@ struct nfp_app; #define NFP_FL_FEATS_VF_RLIM BIT(4) #define NFP_FL_FEATS_FLOW_MOD BIT(5) #define NFP_FL_FEATS_PRE_TUN_RULES BIT(6) +#define NFP_FL_FEATS_IPV6_TUN BIT(7) #define NFP_FL_FEATS_FLOW_MERGE BIT(30) #define NFP_FL_FEATS_LAG BIT(31) @@ -62,18 +63,26 @@ struct nfp_fl_stats_id { * struct nfp_fl_tunnel_offloads - priv data for tunnel offloads * @offloaded_macs: Hashtable of the offloaded MAC addresses * @ipv4_off_list: List of IPv4 addresses to offload - * @neigh_off_list: List of neighbour offloads + * @ipv6_off_list: List of IPv6 addresses to offload + * @neigh_off_list_v4: List of IPv4 neighbour offloads + * @neigh_off_list_v6: List of IPv6 neighbour offloads * @ipv4_off_lock: Lock for the IPv4 address list - * @neigh_off_lock: Lock for the neighbour address list + * @ipv6_off_lock: Lock for the IPv6 address list + * @neigh_off_lock_v4: Lock for the IPv4 neighbour address list + * @neigh_off_lock_v6: Lock for the IPv6 neighbour address list * @mac_off_ids: IDA to manage id assignment for offloaded MACs * @neigh_nb: Notifier to monitor neighbour state */ struct nfp_fl_tunnel_offloads { struct rhashtable offloaded_macs; struct list_head ipv4_off_list; - struct list_head neigh_off_list; + struct list_head ipv6_off_list; + struct list_head neigh_off_list_v4; + struct list_head neigh_off_list_v6; struct mutex ipv4_off_lock; - spinlock_t neigh_off_lock; + struct mutex ipv6_off_lock; + spinlock_t neigh_off_lock_v4; + spinlock_t neigh_off_lock_v6; struct ida mac_off_ids; struct notifier_block neigh_nb; }; @@ -273,12 +282,25 @@ struct nfp_fl_stats { u64 used; }; +/** + * struct nfp_ipv6_addr_entry - cached IPv6 addresses + * @ipv6_addr: IP address + * @ref_count: number of rules currently using this IP + * @list: list pointer + */ +struct nfp_ipv6_addr_entry { + struct in6_addr ipv6_addr; + int ref_count; + struct list_head list; +}; + struct nfp_fl_payload { struct nfp_fl_rule_metadata meta; unsigned long tc_flower_cookie; struct rhash_head fl_node; struct rcu_head rcu; __be32 nfp_tun_ipv4_addr; + struct nfp_ipv6_addr_entry *nfp_tun_ipv6; struct net_device *ingress_dev; char *unmasked_data; char *mask_data; @@ -396,8 +418,14 @@ int nfp_tunnel_mac_event_handler(struct nfp_app *app, unsigned long event, void *ptr); void nfp_tunnel_del_ipv4_off(struct nfp_app *app, __be32 ipv4); void nfp_tunnel_add_ipv4_off(struct nfp_app *app, __be32 ipv4); -void nfp_tunnel_request_route(struct nfp_app *app, struct sk_buff *skb); +void +nfp_tunnel_put_ipv6_off(struct nfp_app *app, struct nfp_ipv6_addr_entry *entry); +struct nfp_ipv6_addr_entry * +nfp_tunnel_add_ipv6_off(struct nfp_app *app, struct in6_addr *ipv6); +void nfp_tunnel_request_route_v4(struct nfp_app *app, struct sk_buff *skb); +void nfp_tunnel_request_route_v6(struct nfp_app *app, struct sk_buff *skb); void nfp_tunnel_keep_alive(struct nfp_app *app, struct sk_buff *skb); +void nfp_tunnel_keep_alive_v6(struct nfp_app *app, struct sk_buff *skb); void nfp_flower_lag_init(struct nfp_fl_lag *lag); void nfp_flower_lag_cleanup(struct nfp_fl_lag *lag); int nfp_flower_lag_reset(struct nfp_fl_lag *lag); diff --git a/drivers/net/ethernet/netronome/nfp/flower/match.c b/drivers/net/ethernet/netronome/nfp/flower/match.c index 9cc3ba17ff69..546bc01d507d 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/match.c +++ b/drivers/net/ethernet/netronome/nfp/flower/match.c @@ -10,9 +10,8 @@ static void nfp_flower_compile_meta_tci(struct nfp_flower_meta_tci *ext, struct nfp_flower_meta_tci *msk, - struct flow_cls_offload *flow, u8 key_type) + struct flow_rule *rule, u8 key_type) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); u16 tmp_tci; memset(ext, 0, sizeof(struct nfp_flower_meta_tci)); @@ -77,11 +76,8 @@ nfp_flower_compile_port(struct nfp_flower_in_port *frame, u32 cmsg_port, static void nfp_flower_compile_mac(struct nfp_flower_mac_mpls *ext, - struct nfp_flower_mac_mpls *msk, - struct flow_cls_offload *flow) + struct nfp_flower_mac_mpls *msk, struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - memset(ext, 0, sizeof(struct nfp_flower_mac_mpls)); memset(msk, 0, sizeof(struct nfp_flower_mac_mpls)); @@ -130,10 +126,8 @@ nfp_flower_compile_mac(struct nfp_flower_mac_mpls *ext, static void nfp_flower_compile_tport(struct nfp_flower_tp_ports *ext, struct nfp_flower_tp_ports *msk, - struct flow_cls_offload *flow) + struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - memset(ext, 0, sizeof(struct nfp_flower_tp_ports)); memset(msk, 0, sizeof(struct nfp_flower_tp_ports)); @@ -150,11 +144,8 @@ nfp_flower_compile_tport(struct nfp_flower_tp_ports *ext, static void nfp_flower_compile_ip_ext(struct nfp_flower_ip_ext *ext, - struct nfp_flower_ip_ext *msk, - struct flow_cls_offload *flow) + struct nfp_flower_ip_ext *msk, struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_BASIC)) { struct flow_match_basic match; @@ -224,10 +215,8 @@ nfp_flower_compile_ip_ext(struct nfp_flower_ip_ext *ext, static void nfp_flower_compile_ipv4(struct nfp_flower_ipv4 *ext, - struct nfp_flower_ipv4 *msk, - struct flow_cls_offload *flow) + struct nfp_flower_ipv4 *msk, struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); struct flow_match_ipv4_addrs match; memset(ext, 0, sizeof(struct nfp_flower_ipv4)); @@ -241,16 +230,13 @@ nfp_flower_compile_ipv4(struct nfp_flower_ipv4 *ext, msk->ipv4_dst = match.mask->dst; } - nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); + nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); } static void nfp_flower_compile_ipv6(struct nfp_flower_ipv6 *ext, - struct nfp_flower_ipv6 *msk, - struct flow_cls_offload *flow) + struct nfp_flower_ipv6 *msk, struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - memset(ext, 0, sizeof(struct nfp_flower_ipv6)); memset(msk, 0, sizeof(struct nfp_flower_ipv6)); @@ -264,16 +250,15 @@ nfp_flower_compile_ipv6(struct nfp_flower_ipv6 *ext, msk->ipv6_dst = match.mask->dst; } - nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); + nfp_flower_compile_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); } static int -nfp_flower_compile_geneve_opt(void *ext, void *msk, - struct flow_cls_offload *flow) +nfp_flower_compile_geneve_opt(void *ext, void *msk, struct flow_rule *rule) { struct flow_match_enc_opts match; - flow_rule_match_enc_opts(flow->rule, &match); + flow_rule_match_enc_opts(rule, &match); memcpy(ext, match.key->data, match.key->len); memcpy(msk, match.mask->data, match.mask->len); @@ -283,10 +268,8 @@ nfp_flower_compile_geneve_opt(void *ext, void *msk, static void nfp_flower_compile_tun_ipv4_addrs(struct nfp_flower_tun_ipv4 *ext, struct nfp_flower_tun_ipv4 *msk, - struct flow_cls_offload *flow) + struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS)) { struct flow_match_ipv4_addrs match; @@ -299,12 +282,26 @@ nfp_flower_compile_tun_ipv4_addrs(struct nfp_flower_tun_ipv4 *ext, } static void +nfp_flower_compile_tun_ipv6_addrs(struct nfp_flower_tun_ipv6 *ext, + struct nfp_flower_tun_ipv6 *msk, + struct flow_rule *rule) +{ + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_IPV6_ADDRS)) { + struct flow_match_ipv6_addrs match; + + flow_rule_match_enc_ipv6_addrs(rule, &match); + ext->src = match.key->src; + ext->dst = match.key->dst; + msk->src = match.mask->src; + msk->dst = match.mask->dst; + } +} + +static void nfp_flower_compile_tun_ip_ext(struct nfp_flower_tun_ip_ext *ext, struct nfp_flower_tun_ip_ext *msk, - struct flow_cls_offload *flow) + struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_IP)) { struct flow_match_ip match; @@ -317,57 +314,97 @@ nfp_flower_compile_tun_ip_ext(struct nfp_flower_tun_ip_ext *ext, } static void -nfp_flower_compile_ipv4_gre_tun(struct nfp_flower_ipv4_gre_tun *ext, - struct nfp_flower_ipv4_gre_tun *msk, - struct flow_cls_offload *flow) +nfp_flower_compile_tun_udp_key(__be32 *key, __be32 *key_msk, + struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - - memset(ext, 0, sizeof(struct nfp_flower_ipv4_gre_tun)); - memset(msk, 0, sizeof(struct nfp_flower_ipv4_gre_tun)); + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_KEYID)) { + struct flow_match_enc_keyid match; + u32 vni; - /* NVGRE is the only supported GRE tunnel type */ - ext->ethertype = cpu_to_be16(ETH_P_TEB); - msk->ethertype = cpu_to_be16(~0); + flow_rule_match_enc_keyid(rule, &match); + vni = be32_to_cpu(match.key->keyid) << NFP_FL_TUN_VNI_OFFSET; + *key = cpu_to_be32(vni); + vni = be32_to_cpu(match.mask->keyid) << NFP_FL_TUN_VNI_OFFSET; + *key_msk = cpu_to_be32(vni); + } +} +static void +nfp_flower_compile_tun_gre_key(__be32 *key, __be32 *key_msk, __be16 *flags, + __be16 *flags_msk, struct flow_rule *rule) +{ if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_KEYID)) { struct flow_match_enc_keyid match; flow_rule_match_enc_keyid(rule, &match); - ext->tun_key = match.key->keyid; - msk->tun_key = match.mask->keyid; + *key = match.key->keyid; + *key_msk = match.mask->keyid; - ext->tun_flags = cpu_to_be16(NFP_FL_GRE_FLAG_KEY); - msk->tun_flags = cpu_to_be16(NFP_FL_GRE_FLAG_KEY); + *flags = cpu_to_be16(NFP_FL_GRE_FLAG_KEY); + *flags_msk = cpu_to_be16(NFP_FL_GRE_FLAG_KEY); } +} + +static void +nfp_flower_compile_ipv4_gre_tun(struct nfp_flower_ipv4_gre_tun *ext, + struct nfp_flower_ipv4_gre_tun *msk, + struct flow_rule *rule) +{ + memset(ext, 0, sizeof(struct nfp_flower_ipv4_gre_tun)); + memset(msk, 0, sizeof(struct nfp_flower_ipv4_gre_tun)); + + /* NVGRE is the only supported GRE tunnel type */ + ext->ethertype = cpu_to_be16(ETH_P_TEB); + msk->ethertype = cpu_to_be16(~0); - nfp_flower_compile_tun_ipv4_addrs(&ext->ipv4, &msk->ipv4, flow); - nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); + nfp_flower_compile_tun_ipv4_addrs(&ext->ipv4, &msk->ipv4, rule); + nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); + nfp_flower_compile_tun_gre_key(&ext->tun_key, &msk->tun_key, + &ext->tun_flags, &msk->tun_flags, rule); } static void nfp_flower_compile_ipv4_udp_tun(struct nfp_flower_ipv4_udp_tun *ext, struct nfp_flower_ipv4_udp_tun *msk, - struct flow_cls_offload *flow) + struct flow_rule *rule) { - struct flow_rule *rule = flow_cls_offload_flow_rule(flow); - memset(ext, 0, sizeof(struct nfp_flower_ipv4_udp_tun)); memset(msk, 0, sizeof(struct nfp_flower_ipv4_udp_tun)); - if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_KEYID)) { - struct flow_match_enc_keyid match; - u32 temp_vni; + nfp_flower_compile_tun_ipv4_addrs(&ext->ipv4, &msk->ipv4, rule); + nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); + nfp_flower_compile_tun_udp_key(&ext->tun_id, &msk->tun_id, rule); +} - flow_rule_match_enc_keyid(rule, &match); - temp_vni = be32_to_cpu(match.key->keyid) << NFP_FL_TUN_VNI_OFFSET; - ext->tun_id = cpu_to_be32(temp_vni); - temp_vni = be32_to_cpu(match.mask->keyid) << NFP_FL_TUN_VNI_OFFSET; - msk->tun_id = cpu_to_be32(temp_vni); - } +static void +nfp_flower_compile_ipv6_udp_tun(struct nfp_flower_ipv6_udp_tun *ext, + struct nfp_flower_ipv6_udp_tun *msk, + struct flow_rule *rule) +{ + memset(ext, 0, sizeof(struct nfp_flower_ipv6_udp_tun)); + memset(msk, 0, sizeof(struct nfp_flower_ipv6_udp_tun)); + + nfp_flower_compile_tun_ipv6_addrs(&ext->ipv6, &msk->ipv6, rule); + nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); + nfp_flower_compile_tun_udp_key(&ext->tun_id, &msk->tun_id, rule); +} + +static void +nfp_flower_compile_ipv6_gre_tun(struct nfp_flower_ipv6_gre_tun *ext, + struct nfp_flower_ipv6_gre_tun *msk, + struct flow_rule *rule) +{ + memset(ext, 0, sizeof(struct nfp_flower_ipv6_gre_tun)); + memset(msk, 0, sizeof(struct nfp_flower_ipv6_gre_tun)); + + /* NVGRE is the only supported GRE tunnel type */ + ext->ethertype = cpu_to_be16(ETH_P_TEB); + msk->ethertype = cpu_to_be16(~0); - nfp_flower_compile_tun_ipv4_addrs(&ext->ipv4, &msk->ipv4, flow); - nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, flow); + nfp_flower_compile_tun_ipv6_addrs(&ext->ipv6, &msk->ipv6, rule); + nfp_flower_compile_tun_ip_ext(&ext->ip_ext, &msk->ip_ext, rule); + nfp_flower_compile_tun_gre_key(&ext->tun_key, &msk->tun_key, + &ext->tun_flags, &msk->tun_flags, rule); } int nfp_flower_compile_flow_match(struct nfp_app *app, @@ -378,6 +415,7 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, enum nfp_flower_tun_type tun_type, struct netlink_ext_ack *extack) { + struct flow_rule *rule = flow_cls_offload_flow_rule(flow); u32 port_id; int err; u8 *ext; @@ -393,7 +431,7 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, nfp_flower_compile_meta_tci((struct nfp_flower_meta_tci *)ext, (struct nfp_flower_meta_tci *)msk, - flow, key_ls->key_layer); + rule, key_ls->key_layer); ext += sizeof(struct nfp_flower_meta_tci); msk += sizeof(struct nfp_flower_meta_tci); @@ -425,7 +463,7 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, if (NFP_FLOWER_LAYER_MAC & key_ls->key_layer) { nfp_flower_compile_mac((struct nfp_flower_mac_mpls *)ext, (struct nfp_flower_mac_mpls *)msk, - flow); + rule); ext += sizeof(struct nfp_flower_mac_mpls); msk += sizeof(struct nfp_flower_mac_mpls); } @@ -433,7 +471,7 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, if (NFP_FLOWER_LAYER_TP & key_ls->key_layer) { nfp_flower_compile_tport((struct nfp_flower_tp_ports *)ext, (struct nfp_flower_tp_ports *)msk, - flow); + rule); ext += sizeof(struct nfp_flower_tp_ports); msk += sizeof(struct nfp_flower_tp_ports); } @@ -441,7 +479,7 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, if (NFP_FLOWER_LAYER_IPV4 & key_ls->key_layer) { nfp_flower_compile_ipv4((struct nfp_flower_ipv4 *)ext, (struct nfp_flower_ipv4 *)msk, - flow); + rule); ext += sizeof(struct nfp_flower_ipv4); msk += sizeof(struct nfp_flower_ipv4); } @@ -449,43 +487,83 @@ int nfp_flower_compile_flow_match(struct nfp_app *app, if (NFP_FLOWER_LAYER_IPV6 & key_ls->key_layer) { nfp_flower_compile_ipv6((struct nfp_flower_ipv6 *)ext, (struct nfp_flower_ipv6 *)msk, - flow); + rule); ext += sizeof(struct nfp_flower_ipv6); msk += sizeof(struct nfp_flower_ipv6); } if (key_ls->key_layer_two & NFP_FLOWER_LAYER2_GRE) { - __be32 tun_dst; - - nfp_flower_compile_ipv4_gre_tun((void *)ext, (void *)msk, flow); - tun_dst = ((struct nfp_flower_ipv4_gre_tun *)ext)->ipv4.dst; - ext += sizeof(struct nfp_flower_ipv4_gre_tun); - msk += sizeof(struct nfp_flower_ipv4_gre_tun); - - /* Store the tunnel destination in the rule data. - * This must be present and be an exact match. - */ - nfp_flow->nfp_tun_ipv4_addr = tun_dst; - nfp_tunnel_add_ipv4_off(app, tun_dst); + if (key_ls->key_layer_two & NFP_FLOWER_LAYER2_TUN_IPV6) { + struct nfp_flower_ipv6_gre_tun *gre_match; + struct nfp_ipv6_addr_entry *entry; + struct in6_addr *dst; + + nfp_flower_compile_ipv6_gre_tun((void *)ext, + (void *)msk, rule); + gre_match = (struct nfp_flower_ipv6_gre_tun *)ext; + dst = &gre_match->ipv6.dst; + ext += sizeof(struct nfp_flower_ipv6_gre_tun); + msk += sizeof(struct nfp_flower_ipv6_gre_tun); + + entry = nfp_tunnel_add_ipv6_off(app, dst); + if (!entry) + return -EOPNOTSUPP; + + nfp_flow->nfp_tun_ipv6 = entry; + } else { + __be32 dst; + + nfp_flower_compile_ipv4_gre_tun((void *)ext, + (void *)msk, rule); + dst = ((struct nfp_flower_ipv4_gre_tun *)ext)->ipv4.dst; + ext += sizeof(struct nfp_flower_ipv4_gre_tun); + msk += sizeof(struct nfp_flower_ipv4_gre_tun); + + /* Store the tunnel destination in the rule data. + * This must be present and be an exact match. + */ + nfp_flow->nfp_tun_ipv4_addr = dst; + nfp_tunnel_add_ipv4_off(app, dst); + } } if (key_ls->key_layer & NFP_FLOWER_LAYER_VXLAN || key_ls->key_layer_two & NFP_FLOWER_LAYER2_GENEVE) { - __be32 tun_dst; - - nfp_flower_compile_ipv4_udp_tun((void *)ext, (void *)msk, flow); - tun_dst = ((struct nfp_flower_ipv4_udp_tun *)ext)->ipv4.dst; - ext += sizeof(struct nfp_flower_ipv4_udp_tun); - msk += sizeof(struct nfp_flower_ipv4_udp_tun); - - /* Store the tunnel destination in the rule data. - * This must be present and be an exact match. - */ - nfp_flow->nfp_tun_ipv4_addr = tun_dst; - nfp_tunnel_add_ipv4_off(app, tun_dst); + if (key_ls->key_layer_two & NFP_FLOWER_LAYER2_TUN_IPV6) { + struct nfp_flower_ipv6_udp_tun *udp_match; + struct nfp_ipv6_addr_entry *entry; + struct in6_addr *dst; + + nfp_flower_compile_ipv6_udp_tun((void *)ext, + (void *)msk, rule); + udp_match = (struct nfp_flower_ipv6_udp_tun *)ext; + dst = &udp_match->ipv6.dst; + ext += sizeof(struct nfp_flower_ipv6_udp_tun); + msk += sizeof(struct nfp_flower_ipv6_udp_tun); + + entry = nfp_tunnel_add_ipv6_off(app, dst); + if (!entry) + return -EOPNOTSUPP; + + nfp_flow->nfp_tun_ipv6 = entry; + } else { + __be32 dst; + + nfp_flower_compile_ipv4_udp_tun((void *)ext, + (void *)msk, rule); + dst = ((struct nfp_flower_ipv4_udp_tun *)ext)->ipv4.dst; + ext += sizeof(struct nfp_flower_ipv4_udp_tun); + msk += sizeof(struct nfp_flower_ipv4_udp_tun); + + /* Store the tunnel destination in the rule data. + * This must be present and be an exact match. + */ + nfp_flow->nfp_tun_ipv4_addr = dst; + nfp_tunnel_add_ipv4_off(app, dst); + } if (key_ls->key_layer_two & NFP_FLOWER_LAYER2_GENEVE_OP) { - err = nfp_flower_compile_geneve_opt(ext, msk, flow); + err = nfp_flower_compile_geneve_opt(ext, msk, rule); if (err) return err; } diff --git a/drivers/net/ethernet/netronome/nfp/flower/offload.c b/drivers/net/ethernet/netronome/nfp/flower/offload.c index 987ae221f6be..7ca5c1becfcf 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/offload.c +++ b/drivers/net/ethernet/netronome/nfp/flower/offload.c @@ -54,6 +54,10 @@ (BIT(FLOW_DISSECTOR_KEY_ENC_CONTROL) | \ BIT(FLOW_DISSECTOR_KEY_ENC_IPV4_ADDRS)) +#define NFP_FLOWER_WHITELIST_TUN_DISSECTOR_V6_R \ + (BIT(FLOW_DISSECTOR_KEY_ENC_CONTROL) | \ + BIT(FLOW_DISSECTOR_KEY_ENC_IPV6_ADDRS)) + #define NFP_FLOWER_MERGE_FIELDS \ (NFP_FLOWER_LAYER_PORT | \ NFP_FLOWER_LAYER_MAC | \ @@ -64,7 +68,8 @@ #define NFP_FLOWER_PRE_TUN_RULE_FIELDS \ (NFP_FLOWER_LAYER_PORT | \ NFP_FLOWER_LAYER_MAC | \ - NFP_FLOWER_LAYER_IPV4) + NFP_FLOWER_LAYER_IPV4 | \ + NFP_FLOWER_LAYER_IPV6) struct nfp_flower_merge_check { union { @@ -146,10 +151,11 @@ static bool nfp_flower_check_higher_than_l3(struct flow_cls_offload *f) static int nfp_flower_calc_opt_layer(struct flow_dissector_key_enc_opts *enc_opts, - u32 *key_layer_two, int *key_size, + u32 *key_layer_two, int *key_size, bool ipv6, struct netlink_ext_ack *extack) { - if (enc_opts->len > NFP_FL_MAX_GENEVE_OPT_KEY) { + if (enc_opts->len > NFP_FL_MAX_GENEVE_OPT_KEY || + (ipv6 && enc_opts->len > NFP_FL_MAX_GENEVE_OPT_KEY_V6)) { NL_SET_ERR_MSG_MOD(extack, "unsupported offload: geneve options exceed maximum length"); return -EOPNOTSUPP; } @@ -167,7 +173,7 @@ nfp_flower_calc_udp_tun_layer(struct flow_dissector_key_ports *enc_ports, struct flow_dissector_key_enc_opts *enc_op, u32 *key_layer_two, u8 *key_layer, int *key_size, struct nfp_flower_priv *priv, - enum nfp_flower_tun_type *tun_type, + enum nfp_flower_tun_type *tun_type, bool ipv6, struct netlink_ext_ack *extack) { int err; @@ -176,7 +182,15 @@ nfp_flower_calc_udp_tun_layer(struct flow_dissector_key_ports *enc_ports, case htons(IANA_VXLAN_UDP_PORT): *tun_type = NFP_FL_TUNNEL_VXLAN; *key_layer |= NFP_FLOWER_LAYER_VXLAN; - *key_size += sizeof(struct nfp_flower_ipv4_udp_tun); + + if (ipv6) { + *key_layer |= NFP_FLOWER_LAYER_EXT_META; + *key_size += sizeof(struct nfp_flower_ext_meta); + *key_layer_two |= NFP_FLOWER_LAYER2_TUN_IPV6; + *key_size += sizeof(struct nfp_flower_ipv6_udp_tun); + } else { + *key_size += sizeof(struct nfp_flower_ipv4_udp_tun); + } if (enc_op) { NL_SET_ERR_MSG_MOD(extack, "unsupported offload: encap options not supported on vxlan tunnels"); @@ -192,7 +206,13 @@ nfp_flower_calc_udp_tun_layer(struct flow_dissector_key_ports *enc_ports, *key_layer |= NFP_FLOWER_LAYER_EXT_META; *key_size += sizeof(struct nfp_flower_ext_meta); *key_layer_two |= NFP_FLOWER_LAYER2_GENEVE; - *key_size += sizeof(struct nfp_flower_ipv4_udp_tun); + + if (ipv6) { + *key_layer_two |= NFP_FLOWER_LAYER2_TUN_IPV6; + *key_size += sizeof(struct nfp_flower_ipv6_udp_tun); + } else { + *key_size += sizeof(struct nfp_flower_ipv4_udp_tun); + } if (!enc_op) break; @@ -200,8 +220,8 @@ nfp_flower_calc_udp_tun_layer(struct flow_dissector_key_ports *enc_ports, NL_SET_ERR_MSG_MOD(extack, "unsupported offload: loaded firmware does not support geneve option offload"); return -EOPNOTSUPP; } - err = nfp_flower_calc_opt_layer(enc_op, key_layer_two, - key_size, extack); + err = nfp_flower_calc_opt_layer(enc_op, key_layer_two, key_size, + ipv6, extack); if (err) return err; break; @@ -237,6 +257,8 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, /* If any tun dissector is used then the required set must be used. */ if (dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR && + (dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR_V6_R) + != NFP_FLOWER_WHITELIST_TUN_DISSECTOR_V6_R && (dissector->used_keys & NFP_FLOWER_WHITELIST_TUN_DISSECTOR_R) != NFP_FLOWER_WHITELIST_TUN_DISSECTOR_R) { NL_SET_ERR_MSG_MOD(extack, "unsupported offload: tunnel match not supported"); @@ -268,8 +290,10 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_CONTROL)) { struct flow_match_enc_opts enc_op = { NULL, NULL }; struct flow_match_ipv4_addrs ipv4_addrs; + struct flow_match_ipv6_addrs ipv6_addrs; struct flow_match_control enc_ctl; struct flow_match_ports enc_ports; + bool ipv6_tun = false; flow_rule_match_enc_control(rule, &enc_ctl); @@ -277,38 +301,62 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, NL_SET_ERR_MSG_MOD(extack, "unsupported offload: wildcarded protocols on tunnels are not supported"); return -EOPNOTSUPP; } - if (enc_ctl.key->addr_type != FLOW_DISSECTOR_KEY_IPV4_ADDRS) { - NL_SET_ERR_MSG_MOD(extack, "unsupported offload: only IPv4 tunnels are supported"); + + ipv6_tun = enc_ctl.key->addr_type == + FLOW_DISSECTOR_KEY_IPV6_ADDRS; + if (ipv6_tun && + !(priv->flower_ext_feats & NFP_FL_FEATS_IPV6_TUN)) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: firmware does not support IPv6 tunnels"); return -EOPNOTSUPP; } - /* These fields are already verified as used. */ - flow_rule_match_enc_ipv4_addrs(rule, &ipv4_addrs); - if (ipv4_addrs.mask->dst != cpu_to_be32(~0)) { - NL_SET_ERR_MSG_MOD(extack, "unsupported offload: only an exact match IPv4 destination address is supported"); + if (!ipv6_tun && + enc_ctl.key->addr_type != FLOW_DISSECTOR_KEY_IPV4_ADDRS) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: tunnel address type not IPv4 or IPv6"); return -EOPNOTSUPP; } + if (ipv6_tun) { + flow_rule_match_enc_ipv6_addrs(rule, &ipv6_addrs); + if (memchr_inv(&ipv6_addrs.mask->dst, 0xff, + sizeof(ipv6_addrs.mask->dst))) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: only an exact match IPv6 destination address is supported"); + return -EOPNOTSUPP; + } + } else { + flow_rule_match_enc_ipv4_addrs(rule, &ipv4_addrs); + if (ipv4_addrs.mask->dst != cpu_to_be32(~0)) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: only an exact match IPv4 destination address is supported"); + return -EOPNOTSUPP; + } + } + if (flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_OPTS)) flow_rule_match_enc_opts(rule, &enc_op); - if (!flow_rule_match_key(rule, FLOW_DISSECTOR_KEY_ENC_PORTS)) { /* check if GRE, which has no enc_ports */ - if (netif_is_gretap(netdev)) { - *tun_type = NFP_FL_TUNNEL_GRE; - key_layer |= NFP_FLOWER_LAYER_EXT_META; - key_size += sizeof(struct nfp_flower_ext_meta); - key_layer_two |= NFP_FLOWER_LAYER2_GRE; - key_size += - sizeof(struct nfp_flower_ipv4_gre_tun); + if (!netif_is_gretap(netdev)) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: an exact match on L4 destination port is required for non-GRE tunnels"); + return -EOPNOTSUPP; + } - if (enc_op.key) { - NL_SET_ERR_MSG_MOD(extack, "unsupported offload: encap options not supported on GRE tunnels"); - return -EOPNOTSUPP; - } + *tun_type = NFP_FL_TUNNEL_GRE; + key_layer |= NFP_FLOWER_LAYER_EXT_META; + key_size += sizeof(struct nfp_flower_ext_meta); + key_layer_two |= NFP_FLOWER_LAYER2_GRE; + + if (ipv6_tun) { + key_layer_two |= NFP_FLOWER_LAYER2_TUN_IPV6; + key_size += + sizeof(struct nfp_flower_ipv6_udp_tun); } else { - NL_SET_ERR_MSG_MOD(extack, "unsupported offload: an exact match on L4 destination port is required for non-GRE tunnels"); + key_size += + sizeof(struct nfp_flower_ipv4_udp_tun); + } + + if (enc_op.key) { + NL_SET_ERR_MSG_MOD(extack, "unsupported offload: encap options not supported on GRE tunnels"); return -EOPNOTSUPP; } } else { @@ -323,7 +371,8 @@ nfp_flower_calculate_key_layers(struct nfp_app *app, &key_layer_two, &key_layer, &key_size, priv, - tun_type, extack); + tun_type, ipv6_tun, + extack); if (err) return err; @@ -491,6 +540,7 @@ nfp_flower_allocate_new(struct nfp_fl_key_ls *key_layer) goto err_free_mask; flow_pay->nfp_tun_ipv4_addr = 0; + flow_pay->nfp_tun_ipv6 = NULL; flow_pay->meta.flags = 0; INIT_LIST_HEAD(&flow_pay->linked_flows); flow_pay->in_hw = false; @@ -517,10 +567,12 @@ nfp_flower_update_merge_with_actions(struct nfp_fl_payload *flow, struct nfp_fl_set_ip4_addrs *ipv4_add; struct nfp_fl_set_ipv6_addr *ipv6_add; struct nfp_fl_push_vlan *push_vlan; + struct nfp_fl_pre_tunnel *pre_tun; struct nfp_fl_set_tport *tport; struct nfp_fl_set_eth *eth; struct nfp_fl_act_head *a; unsigned int act_off = 0; + bool ipv6_tun = false; u8 act_id = 0; u8 *ports; int i; @@ -542,14 +594,18 @@ nfp_flower_update_merge_with_actions(struct nfp_fl_payload *flow, case NFP_FL_ACTION_OPCODE_POP_VLAN: merge->tci = cpu_to_be16(0); break; - case NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL: + case NFP_FL_ACTION_OPCODE_SET_TUNNEL: /* New tunnel header means l2 to l4 can be matched. */ eth_broadcast_addr(&merge->l2.mac_dst[0]); eth_broadcast_addr(&merge->l2.mac_src[0]); memset(&merge->l4, 0xff, sizeof(struct nfp_flower_tp_ports)); - memset(&merge->ipv4, 0xff, - sizeof(struct nfp_flower_ipv4)); + if (ipv6_tun) + memset(&merge->ipv6, 0xff, + sizeof(struct nfp_flower_ipv6)); + else + memset(&merge->ipv4, 0xff, + sizeof(struct nfp_flower_ipv4)); break; case NFP_FL_ACTION_OPCODE_SET_ETHERNET: eth = (struct nfp_fl_set_eth *)a; @@ -597,6 +653,10 @@ nfp_flower_update_merge_with_actions(struct nfp_fl_payload *flow, ports[i] |= tport->tp_port_mask[i]; break; case NFP_FL_ACTION_OPCODE_PRE_TUNNEL: + pre_tun = (struct nfp_fl_pre_tunnel *)a; + ipv6_tun = be16_to_cpu(pre_tun->flags) & + NFP_FL_PRE_TUN_IPV6; + break; case NFP_FL_ACTION_OPCODE_PRE_LAG: case NFP_FL_ACTION_OPCODE_PUSH_GENEVE: break; @@ -765,15 +825,15 @@ nfp_fl_verify_post_tun_acts(char *acts, int len, struct nfp_fl_push_vlan **vlan) static int nfp_fl_push_vlan_after_tun(char *acts, int len, struct nfp_fl_push_vlan *vlan) { - struct nfp_fl_set_ipv4_tun *tun; + struct nfp_fl_set_tun *tun; struct nfp_fl_act_head *a; unsigned int act_off = 0; while (act_off < len) { a = (struct nfp_fl_act_head *)&acts[act_off]; - if (a->jump_id == NFP_FL_ACTION_OPCODE_SET_IPV4_TUNNEL) { - tun = (struct nfp_fl_set_ipv4_tun *)a; + if (a->jump_id == NFP_FL_ACTION_OPCODE_SET_TUNNEL) { + tun = (struct nfp_fl_set_tun *)a; tun->outer_vlan_tpid = vlan->vlan_tpid; tun->outer_vlan_tci = vlan->vlan_tci; @@ -1058,15 +1118,22 @@ nfp_flower_validate_pre_tun_rule(struct nfp_app *app, return -EOPNOTSUPP; } - if (key_layer & NFP_FLOWER_LAYER_IPV4) { + if (key_layer & NFP_FLOWER_LAYER_IPV4 || + key_layer & NFP_FLOWER_LAYER_IPV6) { + /* Flags and proto fields have same offset in IPv4 and IPv6. */ int ip_flags = offsetof(struct nfp_flower_ipv4, ip_ext.flags); int ip_proto = offsetof(struct nfp_flower_ipv4, ip_ext.proto); + int size; int i; + size = key_layer & NFP_FLOWER_LAYER_IPV4 ? + sizeof(struct nfp_flower_ipv4) : + sizeof(struct nfp_flower_ipv6); + mask += sizeof(struct nfp_flower_mac_mpls); /* Ensure proto and flags are the only IP layer fields. */ - for (i = 0; i < sizeof(struct nfp_flower_ipv4); i++) + for (i = 0; i < size; i++) if (mask[i] && i != ip_flags && i != ip_proto) { NL_SET_ERR_MSG_MOD(extack, "unsupported pre-tunnel rule: only flags and proto can be matched in ip header"); return -EOPNOTSUPP; @@ -1195,6 +1262,8 @@ err_remove_rhash: err_release_metadata: nfp_modify_flow_metadata(app, flow_pay); err_destroy_flow: + if (flow_pay->nfp_tun_ipv6) + nfp_tunnel_put_ipv6_off(app, flow_pay->nfp_tun_ipv6); kfree(flow_pay->action_data); kfree(flow_pay->mask_data); kfree(flow_pay->unmasked_data); @@ -1311,6 +1380,9 @@ nfp_flower_del_offload(struct nfp_app *app, struct net_device *netdev, if (nfp_flow->nfp_tun_ipv4_addr) nfp_tunnel_del_ipv4_off(app, nfp_flow->nfp_tun_ipv4_addr); + if (nfp_flow->nfp_tun_ipv6) + nfp_tunnel_put_ipv6_off(app, nfp_flow->nfp_tun_ipv6); + if (!nfp_flow->in_hw) { err = 0; goto err_free_merge_flow; diff --git a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c index 2600ce476d6b..2df3deedf9fd 100644 --- a/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c +++ b/drivers/net/ethernet/netronome/nfp/flower/tunnel_conf.c @@ -55,6 +55,25 @@ struct nfp_tun_active_tuns { }; /** + * struct nfp_tun_active_tuns_v6 - periodic message of active IPv6 tunnels + * @seq: sequence number of the message + * @count: number of tunnels report in message + * @flags: options part of the request + * @tun_info.ipv6: dest IPv6 address of active route + * @tun_info.egress_port: port the encapsulated packet egressed + * @tun_info: tunnels that have sent traffic in reported period + */ +struct nfp_tun_active_tuns_v6 { + __be32 seq; + __be32 count; + __be32 flags; + struct route_ip_info_v6 { + struct in6_addr ipv6; + __be32 egress_port; + } tun_info[]; +}; + +/** * struct nfp_tun_neigh - neighbour/route entry on the NFP * @dst_ipv4: destination IPv4 address * @src_ipv4: source IPv4 address @@ -71,6 +90,22 @@ struct nfp_tun_neigh { }; /** + * struct nfp_tun_neigh_v6 - neighbour/route entry on the NFP + * @dst_ipv6: destination IPv6 address + * @src_ipv6: source IPv6 address + * @dst_addr: destination MAC address + * @src_addr: source MAC address + * @port_id: NFP port to output packet on - associated with source IPv6 + */ +struct nfp_tun_neigh_v6 { + struct in6_addr dst_ipv6; + struct in6_addr src_ipv6; + u8 dst_addr[ETH_ALEN]; + u8 src_addr[ETH_ALEN]; + __be32 port_id; +}; + +/** * struct nfp_tun_req_route_ipv4 - NFP requests a route/neighbour lookup * @ingress_port: ingress port of packet that signalled request * @ipv4_addr: destination ipv4 address for route @@ -83,13 +118,23 @@ struct nfp_tun_req_route_ipv4 { }; /** - * struct nfp_ipv4_route_entry - routes that are offloaded to the NFP - * @ipv4_addr: destination of route + * struct nfp_tun_req_route_ipv6 - NFP requests an IPv6 route/neighbour lookup + * @ingress_port: ingress port of packet that signalled request + * @ipv6_addr: destination ipv6 address for route + */ +struct nfp_tun_req_route_ipv6 { + __be32 ingress_port; + struct in6_addr ipv6_addr; +}; + +/** + * struct nfp_offloaded_route - routes that are offloaded to the NFP * @list: list pointer + * @ip_add: destination of route - can be IPv4 or IPv6 */ -struct nfp_ipv4_route_entry { - __be32 ipv4_addr; +struct nfp_offloaded_route { struct list_head list; + u8 ip_add[]; }; #define NFP_FL_IPV4_ADDRS_MAX 32 @@ -116,6 +161,18 @@ struct nfp_ipv4_addr_entry { struct list_head list; }; +#define NFP_FL_IPV6_ADDRS_MAX 4 + +/** + * struct nfp_tun_ipv6_addr - set the IP address list on the NFP + * @count: number of IPs populated in the array + * @ipv6_addr: array of IPV6_ADDRS_MAX 128 bit IPv6 addresses + */ +struct nfp_tun_ipv6_addr { + __be32 count; + struct in6_addr ipv6_addr[NFP_FL_IPV6_ADDRS_MAX]; +}; + #define NFP_TUN_MAC_OFFLOAD_DEL_FLAG 0x2 /** @@ -206,6 +263,49 @@ void nfp_tunnel_keep_alive(struct nfp_app *app, struct sk_buff *skb) rcu_read_unlock(); } +void nfp_tunnel_keep_alive_v6(struct nfp_app *app, struct sk_buff *skb) +{ +#if IS_ENABLED(CONFIG_IPV6) + struct nfp_tun_active_tuns_v6 *payload; + struct net_device *netdev; + int count, i, pay_len; + struct neighbour *n; + void *ipv6_add; + u32 port; + + payload = nfp_flower_cmsg_get_data(skb); + count = be32_to_cpu(payload->count); + if (count > NFP_FL_IPV6_ADDRS_MAX) { + nfp_flower_cmsg_warn(app, "IPv6 tunnel keep-alive request exceeds max routes.\n"); + return; + } + + pay_len = nfp_flower_cmsg_get_data_len(skb); + if (pay_len != struct_size(payload, tun_info, count)) { + nfp_flower_cmsg_warn(app, "Corruption in tunnel keep-alive message.\n"); + return; + } + + rcu_read_lock(); + for (i = 0; i < count; i++) { + ipv6_add = &payload->tun_info[i].ipv6; + port = be32_to_cpu(payload->tun_info[i].egress_port); + netdev = nfp_app_dev_get(app, port, NULL); + if (!netdev) + continue; + + n = neigh_lookup(&nd_tbl, ipv6_add, netdev); + if (!n) + continue; + + /* Update the used timestamp of neighbour */ + neigh_event_send(n, NULL); + neigh_release(n); + } + rcu_read_unlock(); +#endif +} + static int nfp_flower_xmit_tun_conf(struct nfp_app *app, u8 mtype, u16 plen, void *pdata, gfp_t flag) @@ -224,71 +324,126 @@ nfp_flower_xmit_tun_conf(struct nfp_app *app, u8 mtype, u16 plen, void *pdata, return 0; } -static bool nfp_tun_has_route(struct nfp_app *app, __be32 ipv4_addr) +static bool +__nfp_tun_has_route(struct list_head *route_list, spinlock_t *list_lock, + void *add, int add_len) { - struct nfp_flower_priv *priv = app->priv; - struct nfp_ipv4_route_entry *entry; - struct list_head *ptr, *storage; + struct nfp_offloaded_route *entry; - spin_lock_bh(&priv->tun.neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { - entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); - if (entry->ipv4_addr == ipv4_addr) { - spin_unlock_bh(&priv->tun.neigh_off_lock); + spin_lock_bh(list_lock); + list_for_each_entry(entry, route_list, list) + if (!memcmp(entry->ip_add, add, add_len)) { + spin_unlock_bh(list_lock); return true; } - } - spin_unlock_bh(&priv->tun.neigh_off_lock); + spin_unlock_bh(list_lock); return false; } -static void nfp_tun_add_route_to_cache(struct nfp_app *app, __be32 ipv4_addr) +static int +__nfp_tun_add_route_to_cache(struct list_head *route_list, + spinlock_t *list_lock, void *add, int add_len) { - struct nfp_flower_priv *priv = app->priv; - struct nfp_ipv4_route_entry *entry; - struct list_head *ptr, *storage; + struct nfp_offloaded_route *entry; - spin_lock_bh(&priv->tun.neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { - entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); - if (entry->ipv4_addr == ipv4_addr) { - spin_unlock_bh(&priv->tun.neigh_off_lock); - return; + spin_lock_bh(list_lock); + list_for_each_entry(entry, route_list, list) + if (!memcmp(entry->ip_add, add, add_len)) { + spin_unlock_bh(list_lock); + return 0; } - } - entry = kmalloc(sizeof(*entry), GFP_ATOMIC); + + entry = kmalloc(sizeof(*entry) + add_len, GFP_ATOMIC); if (!entry) { - spin_unlock_bh(&priv->tun.neigh_off_lock); - nfp_flower_cmsg_warn(app, "Mem error when storing new route.\n"); - return; + spin_unlock_bh(list_lock); + return -ENOMEM; } - entry->ipv4_addr = ipv4_addr; - list_add_tail(&entry->list, &priv->tun.neigh_off_list); - spin_unlock_bh(&priv->tun.neigh_off_lock); + memcpy(entry->ip_add, add, add_len); + list_add_tail(&entry->list, route_list); + spin_unlock_bh(list_lock); + + return 0; } -static void nfp_tun_del_route_from_cache(struct nfp_app *app, __be32 ipv4_addr) +static void +__nfp_tun_del_route_from_cache(struct list_head *route_list, + spinlock_t *list_lock, void *add, int add_len) { - struct nfp_flower_priv *priv = app->priv; - struct nfp_ipv4_route_entry *entry; - struct list_head *ptr, *storage; + struct nfp_offloaded_route *entry; - spin_lock_bh(&priv->tun.neigh_off_lock); - list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { - entry = list_entry(ptr, struct nfp_ipv4_route_entry, list); - if (entry->ipv4_addr == ipv4_addr) { + spin_lock_bh(list_lock); + list_for_each_entry(entry, route_list, list) + if (!memcmp(entry->ip_add, add, add_len)) { list_del(&entry->list); kfree(entry); break; } - } - spin_unlock_bh(&priv->tun.neigh_off_lock); + spin_unlock_bh(list_lock); +} + +static bool nfp_tun_has_route_v4(struct nfp_app *app, __be32 *ipv4_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + return __nfp_tun_has_route(&priv->tun.neigh_off_list_v4, + &priv->tun.neigh_off_lock_v4, ipv4_addr, + sizeof(*ipv4_addr)); +} + +static bool +nfp_tun_has_route_v6(struct nfp_app *app, struct in6_addr *ipv6_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + return __nfp_tun_has_route(&priv->tun.neigh_off_list_v6, + &priv->tun.neigh_off_lock_v6, ipv6_addr, + sizeof(*ipv6_addr)); +} + +static void +nfp_tun_add_route_to_cache_v4(struct nfp_app *app, __be32 *ipv4_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + __nfp_tun_add_route_to_cache(&priv->tun.neigh_off_list_v4, + &priv->tun.neigh_off_lock_v4, ipv4_addr, + sizeof(*ipv4_addr)); +} + +static void +nfp_tun_add_route_to_cache_v6(struct nfp_app *app, struct in6_addr *ipv6_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + __nfp_tun_add_route_to_cache(&priv->tun.neigh_off_list_v6, + &priv->tun.neigh_off_lock_v6, ipv6_addr, + sizeof(*ipv6_addr)); } static void -nfp_tun_write_neigh(struct net_device *netdev, struct nfp_app *app, - struct flowi4 *flow, struct neighbour *neigh, gfp_t flag) +nfp_tun_del_route_from_cache_v4(struct nfp_app *app, __be32 *ipv4_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + __nfp_tun_del_route_from_cache(&priv->tun.neigh_off_list_v4, + &priv->tun.neigh_off_lock_v4, ipv4_addr, + sizeof(*ipv4_addr)); +} + +static void +nfp_tun_del_route_from_cache_v6(struct nfp_app *app, struct in6_addr *ipv6_addr) +{ + struct nfp_flower_priv *priv = app->priv; + + __nfp_tun_del_route_from_cache(&priv->tun.neigh_off_list_v6, + &priv->tun.neigh_off_lock_v6, ipv6_addr, + sizeof(*ipv6_addr)); +} + +static void +nfp_tun_write_neigh_v4(struct net_device *netdev, struct nfp_app *app, + struct flowi4 *flow, struct neighbour *neigh, gfp_t flag) { struct nfp_tun_neigh payload; u32 port_id; @@ -302,7 +457,7 @@ nfp_tun_write_neigh(struct net_device *netdev, struct nfp_app *app, /* If entry has expired send dst IP with all other fields 0. */ if (!(neigh->nud_state & NUD_VALID) || neigh->dead) { - nfp_tun_del_route_from_cache(app, payload.dst_ipv4); + nfp_tun_del_route_from_cache_v4(app, &payload.dst_ipv4); /* Trigger ARP to verify invalid neighbour state. */ neigh_event_send(neigh, NULL); goto send_msg; @@ -314,7 +469,7 @@ nfp_tun_write_neigh(struct net_device *netdev, struct nfp_app *app, neigh_ha_snapshot(payload.dst_addr, neigh, netdev); payload.port_id = cpu_to_be32(port_id); /* Add destination of new route to NFP cache. */ - nfp_tun_add_route_to_cache(app, payload.dst_ipv4); + nfp_tun_add_route_to_cache_v4(app, &payload.dst_ipv4); send_msg: nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_NEIGH, @@ -322,16 +477,54 @@ send_msg: (unsigned char *)&payload, flag); } +static void +nfp_tun_write_neigh_v6(struct net_device *netdev, struct nfp_app *app, + struct flowi6 *flow, struct neighbour *neigh, gfp_t flag) +{ + struct nfp_tun_neigh_v6 payload; + u32 port_id; + + port_id = nfp_flower_get_port_id_from_netdev(app, netdev); + if (!port_id) + return; + + memset(&payload, 0, sizeof(struct nfp_tun_neigh_v6)); + payload.dst_ipv6 = flow->daddr; + + /* If entry has expired send dst IP with all other fields 0. */ + if (!(neigh->nud_state & NUD_VALID) || neigh->dead) { + nfp_tun_del_route_from_cache_v6(app, &payload.dst_ipv6); + /* Trigger probe to verify invalid neighbour state. */ + neigh_event_send(neigh, NULL); + goto send_msg; + } + + /* Have a valid neighbour so populate rest of entry. */ + payload.src_ipv6 = flow->saddr; + ether_addr_copy(payload.src_addr, netdev->dev_addr); + neigh_ha_snapshot(payload.dst_addr, neigh, netdev); + payload.port_id = cpu_to_be32(port_id); + /* Add destination of new route to NFP cache. */ + nfp_tun_add_route_to_cache_v6(app, &payload.dst_ipv6); + +send_msg: + nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_NEIGH_V6, + sizeof(struct nfp_tun_neigh_v6), + (unsigned char *)&payload, flag); +} + static int nfp_tun_neigh_event_handler(struct notifier_block *nb, unsigned long event, void *ptr) { struct nfp_flower_priv *app_priv; struct netevent_redirect *redir; - struct flowi4 flow = {}; + struct flowi4 flow4 = {}; + struct flowi6 flow6 = {}; struct neighbour *n; struct nfp_app *app; struct rtable *rt; + bool ipv6 = false; int err; switch (event) { @@ -346,7 +539,13 @@ nfp_tun_neigh_event_handler(struct notifier_block *nb, unsigned long event, return NOTIFY_DONE; } - flow.daddr = *(__be32 *)n->primary_key; + if (n->tbl->family == AF_INET6) + ipv6 = true; + + if (ipv6) + flow6.daddr = *(struct in6_addr *)n->primary_key; + else + flow4.daddr = *(__be32 *)n->primary_key; app_priv = container_of(nb, struct nfp_flower_priv, tun.neigh_nb); app = app_priv->app; @@ -356,28 +555,46 @@ nfp_tun_neigh_event_handler(struct notifier_block *nb, unsigned long event, return NOTIFY_DONE; /* Only concerned with changes to routes already added to NFP. */ - if (!nfp_tun_has_route(app, flow.daddr)) + if ((ipv6 && !nfp_tun_has_route_v6(app, &flow6.daddr)) || + (!ipv6 && !nfp_tun_has_route_v4(app, &flow4.daddr))) return NOTIFY_DONE; #if IS_ENABLED(CONFIG_INET) - /* Do a route lookup to populate flow data. */ - rt = ip_route_output_key(dev_net(n->dev), &flow); - err = PTR_ERR_OR_ZERO(rt); - if (err) + if (ipv6) { +#if IS_ENABLED(CONFIG_IPV6) + struct dst_entry *dst; + + dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(n->dev), NULL, + &flow6, NULL); + if (IS_ERR(dst)) + return NOTIFY_DONE; + + dst_release(dst); + flow6.flowi6_proto = IPPROTO_UDP; + nfp_tun_write_neigh_v6(n->dev, app, &flow6, n, GFP_ATOMIC); +#else return NOTIFY_DONE; +#endif /* CONFIG_IPV6 */ + } else { + /* Do a route lookup to populate flow data. */ + rt = ip_route_output_key(dev_net(n->dev), &flow4); + err = PTR_ERR_OR_ZERO(rt); + if (err) + return NOTIFY_DONE; - ip_rt_put(rt); + ip_rt_put(rt); + + flow4.flowi4_proto = IPPROTO_UDP; + nfp_tun_write_neigh_v4(n->dev, app, &flow4, n, GFP_ATOMIC); + } #else return NOTIFY_DONE; -#endif - - flow.flowi4_proto = IPPROTO_UDP; - nfp_tun_write_neigh(n->dev, app, &flow, n, GFP_ATOMIC); +#endif /* CONFIG_INET */ return NOTIFY_OK; } -void nfp_tunnel_request_route(struct nfp_app *app, struct sk_buff *skb) +void nfp_tunnel_request_route_v4(struct nfp_app *app, struct sk_buff *skb) { struct nfp_tun_req_route_ipv4 *payload; struct net_device *netdev; @@ -411,7 +628,7 @@ void nfp_tunnel_request_route(struct nfp_app *app, struct sk_buff *skb) ip_rt_put(rt); if (!n) goto fail_rcu_unlock; - nfp_tun_write_neigh(n->dev, app, &flow, n, GFP_ATOMIC); + nfp_tun_write_neigh_v4(n->dev, app, &flow, n, GFP_ATOMIC); neigh_release(n); rcu_read_unlock(); return; @@ -421,6 +638,48 @@ fail_rcu_unlock: nfp_flower_cmsg_warn(app, "Requested route not found.\n"); } +void nfp_tunnel_request_route_v6(struct nfp_app *app, struct sk_buff *skb) +{ + struct nfp_tun_req_route_ipv6 *payload; + struct net_device *netdev; + struct flowi6 flow = {}; + struct dst_entry *dst; + struct neighbour *n; + + payload = nfp_flower_cmsg_get_data(skb); + + rcu_read_lock(); + netdev = nfp_app_dev_get(app, be32_to_cpu(payload->ingress_port), NULL); + if (!netdev) + goto fail_rcu_unlock; + + flow.daddr = payload->ipv6_addr; + flow.flowi6_proto = IPPROTO_UDP; + +#if IS_ENABLED(CONFIG_INET) && IS_ENABLED(CONFIG_IPV6) + dst = ipv6_stub->ipv6_dst_lookup_flow(dev_net(netdev), NULL, &flow, + NULL); + if (IS_ERR(dst)) + goto fail_rcu_unlock; +#else + goto fail_rcu_unlock; +#endif + + n = dst_neigh_lookup(dst, &flow.daddr); + dst_release(dst); + if (!n) + goto fail_rcu_unlock; + + nfp_tun_write_neigh_v6(n->dev, app, &flow, n, GFP_ATOMIC); + neigh_release(n); + rcu_read_unlock(); + return; + +fail_rcu_unlock: + rcu_read_unlock(); + nfp_flower_cmsg_warn(app, "Requested IPv6 route not found.\n"); +} + static void nfp_tun_write_ipv4_list(struct nfp_app *app) { struct nfp_flower_priv *priv = app->priv; @@ -502,6 +761,78 @@ void nfp_tunnel_del_ipv4_off(struct nfp_app *app, __be32 ipv4) nfp_tun_write_ipv4_list(app); } +static void nfp_tun_write_ipv6_list(struct nfp_app *app) +{ + struct nfp_flower_priv *priv = app->priv; + struct nfp_ipv6_addr_entry *entry; + struct nfp_tun_ipv6_addr payload; + int count = 0; + + memset(&payload, 0, sizeof(struct nfp_tun_ipv6_addr)); + mutex_lock(&priv->tun.ipv6_off_lock); + list_for_each_entry(entry, &priv->tun.ipv6_off_list, list) { + if (count >= NFP_FL_IPV6_ADDRS_MAX) { + nfp_flower_cmsg_warn(app, "Too many IPv6 tunnel endpoint addresses, some cannot be offloaded.\n"); + break; + } + payload.ipv6_addr[count++] = entry->ipv6_addr; + } + mutex_unlock(&priv->tun.ipv6_off_lock); + payload.count = cpu_to_be32(count); + + nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_IPS_V6, + sizeof(struct nfp_tun_ipv6_addr), + &payload, GFP_KERNEL); +} + +struct nfp_ipv6_addr_entry * +nfp_tunnel_add_ipv6_off(struct nfp_app *app, struct in6_addr *ipv6) +{ + struct nfp_flower_priv *priv = app->priv; + struct nfp_ipv6_addr_entry *entry; + + mutex_lock(&priv->tun.ipv6_off_lock); + list_for_each_entry(entry, &priv->tun.ipv6_off_list, list) + if (!memcmp(&entry->ipv6_addr, ipv6, sizeof(*ipv6))) { + entry->ref_count++; + mutex_unlock(&priv->tun.ipv6_off_lock); + return entry; + } + + entry = kmalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) { + mutex_unlock(&priv->tun.ipv6_off_lock); + nfp_flower_cmsg_warn(app, "Mem error when offloading IP address.\n"); + return NULL; + } + entry->ipv6_addr = *ipv6; + entry->ref_count = 1; + list_add_tail(&entry->list, &priv->tun.ipv6_off_list); + mutex_unlock(&priv->tun.ipv6_off_lock); + + nfp_tun_write_ipv6_list(app); + + return entry; +} + +void +nfp_tunnel_put_ipv6_off(struct nfp_app *app, struct nfp_ipv6_addr_entry *entry) +{ + struct nfp_flower_priv *priv = app->priv; + bool freed = false; + + mutex_lock(&priv->tun.ipv6_off_lock); + if (!--entry->ref_count) { + list_del(&entry->list); + kfree(entry); + freed = true; + } + mutex_unlock(&priv->tun.ipv6_off_lock); + + if (freed) + nfp_tun_write_ipv6_list(app); +} + static int __nfp_tunnel_offload_mac(struct nfp_app *app, u8 *mac, u16 idx, bool del) { @@ -1013,13 +1344,17 @@ int nfp_tunnel_config_start(struct nfp_app *app) ida_init(&priv->tun.mac_off_ids); - /* Initialise priv data for IPv4 offloading. */ + /* Initialise priv data for IPv4/v6 offloading. */ mutex_init(&priv->tun.ipv4_off_lock); INIT_LIST_HEAD(&priv->tun.ipv4_off_list); + mutex_init(&priv->tun.ipv6_off_lock); + INIT_LIST_HEAD(&priv->tun.ipv6_off_list); /* Initialise priv data for neighbour offloading. */ - spin_lock_init(&priv->tun.neigh_off_lock); - INIT_LIST_HEAD(&priv->tun.neigh_off_list); + spin_lock_init(&priv->tun.neigh_off_lock_v4); + INIT_LIST_HEAD(&priv->tun.neigh_off_list_v4); + spin_lock_init(&priv->tun.neigh_off_lock_v6); + INIT_LIST_HEAD(&priv->tun.neigh_off_list_v6); priv->tun.neigh_nb.notifier_call = nfp_tun_neigh_event_handler; err = register_netevent_notifier(&priv->tun.neigh_nb); @@ -1034,9 +1369,11 @@ int nfp_tunnel_config_start(struct nfp_app *app) void nfp_tunnel_config_stop(struct nfp_app *app) { + struct nfp_offloaded_route *route_entry, *temp; struct nfp_flower_priv *priv = app->priv; - struct nfp_ipv4_route_entry *route_entry; struct nfp_ipv4_addr_entry *ip_entry; + struct nfp_tun_neigh_v6 ipv6_route; + struct nfp_tun_neigh ipv4_route; struct list_head *ptr, *storage; unregister_netevent_notifier(&priv->tun.neigh_nb); @@ -1050,12 +1387,35 @@ void nfp_tunnel_config_stop(struct nfp_app *app) kfree(ip_entry); } - /* Free any memory that may be occupied by the route list. */ - list_for_each_safe(ptr, storage, &priv->tun.neigh_off_list) { - route_entry = list_entry(ptr, struct nfp_ipv4_route_entry, - list); + mutex_destroy(&priv->tun.ipv6_off_lock); + + /* Free memory in the route list and remove entries from fw cache. */ + list_for_each_entry_safe(route_entry, temp, + &priv->tun.neigh_off_list_v4, list) { + memset(&ipv4_route, 0, sizeof(ipv4_route)); + memcpy(&ipv4_route.dst_ipv4, &route_entry->ip_add, + sizeof(ipv4_route.dst_ipv4)); list_del(&route_entry->list); kfree(route_entry); + + nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_NEIGH, + sizeof(struct nfp_tun_neigh), + (unsigned char *)&ipv4_route, + GFP_KERNEL); + } + + list_for_each_entry_safe(route_entry, temp, + &priv->tun.neigh_off_list_v6, list) { + memset(&ipv6_route, 0, sizeof(ipv6_route)); + memcpy(&ipv6_route.dst_ipv6, &route_entry->ip_add, + sizeof(ipv6_route.dst_ipv6)); + list_del(&route_entry->list); + kfree(route_entry); + + nfp_flower_xmit_tun_conf(app, NFP_FLOWER_CMSG_TYPE_TUN_NEIGH_V6, + sizeof(struct nfp_tun_neigh), + (unsigned char *)&ipv6_route, + GFP_KERNEL); } /* Destroy rhash. Entries should be cleaned on netdev notifier unreg. */ diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net.h b/drivers/net/ethernet/netronome/nfp/nfp_net.h index 250f510b1d21..ff4438478ea9 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_net.h @@ -586,6 +586,9 @@ struct nfp_net_dp { * @ktls_conn_id_gen: Trivial generator for kTLS connection ids (for TX) * @ktls_no_space: Counter of firmware rejecting kTLS connection due to * lack of space + * @ktls_rx_resync_req: Counter of TLS RX resync requested + * @ktls_rx_resync_ign: Counter of TLS RX resync requests ignored + * @ktls_rx_resync_sent: Counter of TLS RX resync completed * @mbox_cmsg: Common Control Message via vNIC mailbox state * @mbox_cmsg.queue: CCM mbox queue of pending messages * @mbox_cmsg.wq: CCM mbox wait queue of waiting processes @@ -674,6 +677,9 @@ struct nfp_net { atomic64_t ktls_conn_id_gen; atomic_t ktls_no_space; + atomic_t ktls_rx_resync_req; + atomic_t ktls_rx_resync_ign; + atomic_t ktls_rx_resync_sent; struct { struct sk_buff_head queue; diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c index bcdcd6de7dea..9bfb3b077bc1 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_common.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_common.c @@ -47,6 +47,7 @@ #include "nfp_net_sriov.h" #include "nfp_port.h" #include "crypto/crypto.h" +#include "crypto/fw.h" /** * nfp_net_get_fw_version() - Read and parse the FW version @@ -1321,17 +1322,11 @@ nfp_net_tx_ring_reset(struct nfp_net_dp *dp, struct nfp_net_tx_ring *tx_ring) netdev_tx_reset_queue(nd_q); } -static void nfp_net_tx_timeout(struct net_device *netdev) +static void nfp_net_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct nfp_net *nn = netdev_priv(netdev); - int i; - for (i = 0; i < nn->dp.netdev->real_num_tx_queues; i++) { - if (!netif_tx_queue_stopped(netdev_get_tx_queue(netdev, i))) - continue; - nn_warn(nn, "TX timeout on ring: %d\n", i); - } - nn_warn(nn, "TX watchdog timeout\n"); + nn_warn(nn, "TX watchdog timeout on ring: %u\n", txqueue); } /* Receive processing @@ -1667,9 +1662,9 @@ nfp_net_set_hash_desc(struct net_device *netdev, struct nfp_meta_parsed *meta, &rx_hash->hash); } -static void * +static bool nfp_net_parse_meta(struct net_device *netdev, struct nfp_meta_parsed *meta, - void *data, int meta_len) + void *data, void *pkt, unsigned int pkt_len, int meta_len) { u32 meta_info; @@ -1699,14 +1694,20 @@ nfp_net_parse_meta(struct net_device *netdev, struct nfp_meta_parsed *meta, (__force __wsum)__get_unaligned_cpu32(data); data += 4; break; + case NFP_NET_META_RESYNC_INFO: + if (nfp_net_tls_rx_resync_req(netdev, data, pkt, + pkt_len)) + return NULL; + data += sizeof(struct nfp_net_tls_resync_req); + break; default: - return NULL; + return true; } meta_info >>= NFP_NET_META_FIELD_SIZE; } - return data; + return data != pkt; } static void @@ -1891,12 +1892,10 @@ static int nfp_net_rx(struct nfp_net_rx_ring *rx_ring, int budget) nfp_net_set_hash_desc(dp->netdev, &meta, rxbuf->frag + meta_off, rxd); } else if (meta_len) { - void *end; - - end = nfp_net_parse_meta(dp->netdev, &meta, - rxbuf->frag + meta_off, - meta_len); - if (unlikely(end != rxbuf->frag + pkt_off)) { + if (unlikely(nfp_net_parse_meta(dp->netdev, &meta, + rxbuf->frag + meta_off, + rxbuf->frag + pkt_off, + pkt_len, meta_len))) { nn_dp_warn(dp, "invalid RX packet metadata\n"); nfp_net_rx_drop(dp, r_vec, rx_ring, rxbuf, NULL); diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c index d835c14b7257..c3a763134e79 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.c @@ -17,6 +17,30 @@ static void nfp_net_tlv_caps_reset(struct nfp_net_tlv_caps *caps) caps->mbox_len = NFP_NET_CFG_MBOX_VAL_MAX_SZ; } +static bool +nfp_net_tls_parse_crypto_ops(struct device *dev, struct nfp_net_tlv_caps *caps, + u8 __iomem *ctrl_mem, u8 __iomem *data, + unsigned int length, unsigned int offset, + bool rx_stream_scan) +{ + /* Ignore the legacy TLV if new one was already parsed */ + if (caps->tls_resync_ss && !rx_stream_scan) + return true; + + if (length < 32) { + dev_err(dev, + "CRYPTO OPS TLV should be at least 32B, is %dB offset:%u\n", + length, offset); + return false; + } + + caps->crypto_ops = readl(data); + caps->crypto_enable_off = data - ctrl_mem + 16; + caps->tls_resync_ss = rx_stream_scan; + + return true; +} + int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, struct nfp_net_tlv_caps *caps) { @@ -104,15 +128,25 @@ int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, caps->mbox_cmsg_types = readl(data); break; case NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS: - if (length < 32) { - dev_err(dev, - "CRYPTO OPS TLV should be at least 32B, is %dB offset:%u\n", - length, offset); + if (!nfp_net_tls_parse_crypto_ops(dev, caps, ctrl_mem, + data, length, offset, + false)) return -EINVAL; + break; + case NFP_NET_CFG_TLV_TYPE_VNIC_STATS: + if ((data - ctrl_mem) % 8) { + dev_warn(dev, "VNIC STATS TLV misaligned, ignoring offset:%u len:%u\n", + offset, length); + break; } - - caps->crypto_ops = readl(data); - caps->crypto_enable_off = data - ctrl_mem + 16; + caps->vnic_stats_off = data - ctrl_mem; + caps->vnic_stats_cnt = length / 10; + break; + case NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS_RX_SCAN: + if (!nfp_net_tls_parse_crypto_ops(dev, caps, ctrl_mem, + data, length, offset, + true)) + return -EINVAL; break; default: if (!FIELD_GET(NFP_NET_CFG_TLV_HEADER_REQUIRED, hdr)) diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h index ee6b24e4eacd..3d61a8cb60b0 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ctrl.h @@ -45,6 +45,7 @@ #define NFP_NET_META_PORTID 5 #define NFP_NET_META_CSUM 6 /* checksum complete type */ #define NFP_NET_META_CONN_HANDLE 7 +#define NFP_NET_META_RESYNC_INFO 8 /* RX resync info request */ #define NFP_META_PORT_ID_CTRL ~0U @@ -479,6 +480,22 @@ * 8 words, bitmaps of supported and enabled crypto operations. * First 16B (4 words) contains a bitmap of supported crypto operations, * and next 16B contain the enabled operations. + * This capability is made obsolete by ones with better sync methods. + * + * %NFP_NET_CFG_TLV_TYPE_VNIC_STATS: + * Variable, per-vNIC statistics, data should be 8B aligned (FW should insert + * zero-length RESERVED TLV to pad). + * TLV data has two sections. First is an array of statistics' IDs (2B each). + * Second 8B statistics themselves. Statistics are 8B aligned, meaning there + * may be a padding between sections. + * Number of statistics can be determined as floor(tlv.length / (2 + 8)). + * This TLV overwrites %NFP_NET_CFG_STATS_* values (statistics in this TLV + * duplicate the old ones, so driver should be careful not to unnecessarily + * render both). + * + * %NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS_RX_SCAN: + * Same as %NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS, but crypto TLS does stream scan + * RX sync, rather than kernel-assisted sync. */ #define NFP_NET_CFG_TLV_TYPE_UNKNOWN 0 #define NFP_NET_CFG_TLV_TYPE_RESERVED 1 @@ -490,6 +507,8 @@ #define NFP_NET_CFG_TLV_TYPE_REPR_CAP 7 #define NFP_NET_CFG_TLV_TYPE_MBOX_CMSG_TYPES 10 #define NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS 11 /* see crypto/fw.h */ +#define NFP_NET_CFG_TLV_TYPE_VNIC_STATS 12 +#define NFP_NET_CFG_TLV_TYPE_CRYPTO_OPS_RX_SCAN 13 struct device; @@ -502,6 +521,9 @@ struct device; * @mbox_cmsg_types: cmsgs which can be passed through the mailbox * @crypto_ops: supported crypto operations * @crypto_enable_off: offset of crypto ops enable region + * @vnic_stats_off: offset of vNIC stats area + * @vnic_stats_cnt: number of vNIC stats + * @tls_resync_ss: TLS resync will be performed via stream scan */ struct nfp_net_tlv_caps { u32 me_freq_mhz; @@ -511,6 +533,9 @@ struct nfp_net_tlv_caps { u32 mbox_cmsg_types; u32 crypto_ops; unsigned int crypto_enable_off; + unsigned int vnic_stats_off; + unsigned int vnic_stats_cnt; + unsigned int tls_resync_ss:1; }; int nfp_net_tlv_caps_parse(struct device *dev, u8 __iomem *ctrl_mem, diff --git a/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c b/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c index 1b840ee47339..d648e32c0520 100644 --- a/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c +++ b/drivers/net/ethernet/netronome/nfp/nfp_net_ethtool.c @@ -148,11 +148,33 @@ static const struct nfp_et_stat nfp_mac_et_stats[] = { { "tx_pause_frames_class7", NFP_MAC_STATS_TX_PAUSE_FRAMES_CLASS7, }, }; +static const char nfp_tlv_stat_names[][ETH_GSTRING_LEN] = { + [1] = "dev_rx_discards", + [2] = "dev_rx_errors", + [3] = "dev_rx_bytes", + [4] = "dev_rx_uc_bytes", + [5] = "dev_rx_mc_bytes", + [6] = "dev_rx_bc_bytes", + [7] = "dev_rx_pkts", + [8] = "dev_rx_mc_pkts", + [9] = "dev_rx_bc_pkts", + + [10] = "dev_tx_discards", + [11] = "dev_tx_errors", + [12] = "dev_tx_bytes", + [13] = "dev_tx_uc_bytes", + [14] = "dev_tx_mc_bytes", + [15] = "dev_tx_bc_bytes", + [16] = "dev_tx_pkts", + [17] = "dev_tx_mc_pkts", + [18] = "dev_tx_bc_pkts", +}; + #define NN_ET_GLOBAL_STATS_LEN ARRAY_SIZE(nfp_net_et_stats) #define NN_ET_SWITCH_STATS_LEN 9 #define NN_RVEC_GATHER_STATS 13 #define NN_RVEC_PER_Q_STATS 3 -#define NN_CTRL_PATH_STATS 1 +#define NN_CTRL_PATH_STATS 4 #define SFP_SFF_REV_COMPLIANCE 1 @@ -454,6 +476,9 @@ static u8 *nfp_vnic_get_sw_stats_strings(struct net_device *netdev, u8 *data) data = nfp_pr_et(data, "tx_tls_drop_no_sync_data"); data = nfp_pr_et(data, "hw_tls_no_space"); + data = nfp_pr_et(data, "rx_tls_resync_req_ok"); + data = nfp_pr_et(data, "rx_tls_resync_req_ign"); + data = nfp_pr_et(data, "rx_tls_resync_sent"); return data; } @@ -502,6 +527,9 @@ static u64 *nfp_vnic_get_sw_stats(struct net_device *netdev, u64 *data) *data++ = gathered_stats[j]; *data++ = atomic_read(&nn->ktls_no_space); + *data++ = atomic_read(&nn->ktls_rx_resync_req); + *data++ = atomic_read(&nn->ktls_rx_resync_ign); + *data++ = atomic_read(&nn->ktls_rx_resync_sent); return data; } @@ -560,6 +588,65 @@ nfp_vnic_get_hw_stats(u64 *data, u8 __iomem *mem, unsigned int num_vecs) return data; } +static unsigned int nfp_vnic_get_tlv_stats_count(struct nfp_net *nn) +{ + return nn->tlv_caps.vnic_stats_cnt + nn->max_r_vecs * 4; +} + +static u8 *nfp_vnic_get_tlv_stats_strings(struct nfp_net *nn, u8 *data) +{ + unsigned int i, id; + u8 __iomem *mem; + u64 id_word = 0; + + mem = nn->dp.ctrl_bar + nn->tlv_caps.vnic_stats_off; + for (i = 0; i < nn->tlv_caps.vnic_stats_cnt; i++) { + if (!(i % 4)) + id_word = readq(mem + i * 2); + + id = (u16)id_word; + id_word >>= 16; + + if (id < ARRAY_SIZE(nfp_tlv_stat_names) && + nfp_tlv_stat_names[id][0]) { + memcpy(data, nfp_tlv_stat_names[id], ETH_GSTRING_LEN); + data += ETH_GSTRING_LEN; + } else { + data = nfp_pr_et(data, "dev_unknown_stat%u", id); + } + } + + for (i = 0; i < nn->max_r_vecs; i++) { + data = nfp_pr_et(data, "rxq_%u_pkts", i); + data = nfp_pr_et(data, "rxq_%u_bytes", i); + data = nfp_pr_et(data, "txq_%u_pkts", i); + data = nfp_pr_et(data, "txq_%u_bytes", i); + } + + return data; +} + +static u64 *nfp_vnic_get_tlv_stats(struct nfp_net *nn, u64 *data) +{ + u8 __iomem *mem; + unsigned int i; + + mem = nn->dp.ctrl_bar + nn->tlv_caps.vnic_stats_off; + mem += roundup(2 * nn->tlv_caps.vnic_stats_cnt, 8); + for (i = 0; i < nn->tlv_caps.vnic_stats_cnt; i++) + *data++ = readq(mem + i * 8); + + mem = nn->dp.ctrl_bar; + for (i = 0; i < nn->max_r_vecs; i++) { + *data++ = readq(mem + NFP_NET_CFG_RXR_STATS(i)); + *data++ = readq(mem + NFP_NET_CFG_RXR_STATS(i) + 8); + *data++ = readq(mem + NFP_NET_CFG_TXR_STATS(i)); + *data++ = readq(mem + NFP_NET_CFG_TXR_STATS(i) + 8); + } + + return data; +} + static unsigned int nfp_mac_get_stats_count(struct net_device *netdev) { struct nfp_port *port; @@ -609,8 +696,12 @@ static void nfp_net_get_strings(struct net_device *netdev, switch (stringset) { case ETH_SS_STATS: data = nfp_vnic_get_sw_stats_strings(netdev, data); - data = nfp_vnic_get_hw_stats_strings(data, nn->max_r_vecs, - false); + if (!nn->tlv_caps.vnic_stats_off) + data = nfp_vnic_get_hw_stats_strings(data, + nn->max_r_vecs, + false); + else + data = nfp_vnic_get_tlv_stats_strings(nn, data); data = nfp_mac_get_stats_strings(netdev, data); data = nfp_app_port_get_stats_strings(nn->port, data); break; @@ -624,7 +715,11 @@ nfp_net_get_stats(struct net_device *netdev, struct ethtool_stats *stats, struct nfp_net *nn = netdev_priv(netdev); data = nfp_vnic_get_sw_stats(netdev, data); - data = nfp_vnic_get_hw_stats(data, nn->dp.ctrl_bar, nn->max_r_vecs); + if (!nn->tlv_caps.vnic_stats_off) + data = nfp_vnic_get_hw_stats(data, nn->dp.ctrl_bar, + nn->max_r_vecs); + else + data = nfp_vnic_get_tlv_stats(nn, data); data = nfp_mac_get_stats(netdev, data); data = nfp_app_port_get_stats(nn->port, data); } @@ -632,13 +727,18 @@ nfp_net_get_stats(struct net_device *netdev, struct ethtool_stats *stats, static int nfp_net_get_sset_count(struct net_device *netdev, int sset) { struct nfp_net *nn = netdev_priv(netdev); + unsigned int cnt; switch (sset) { case ETH_SS_STATS: - return nfp_vnic_get_sw_stats_count(netdev) + - nfp_vnic_get_hw_stats_count(nn->max_r_vecs) + - nfp_mac_get_stats_count(netdev) + - nfp_app_port_get_stats_count(nn->port); + cnt = nfp_vnic_get_sw_stats_count(netdev); + if (!nn->tlv_caps.vnic_stats_off) + cnt += nfp_vnic_get_hw_stats_count(nn->max_r_vecs); + else + cnt += nfp_vnic_get_tlv_stats_count(nn); + cnt += nfp_mac_get_stats_count(netdev); + cnt += nfp_app_port_get_stats_count(nn->port); + return cnt; default: return -EOPNOTSUPP; } diff --git a/drivers/net/ethernet/nvidia/forcedeth.c b/drivers/net/ethernet/nvidia/forcedeth.c index 6b54cb3b681d..2fc10a36afa4 100644 --- a/drivers/net/ethernet/nvidia/forcedeth.c +++ b/drivers/net/ethernet/nvidia/forcedeth.c @@ -2739,7 +2739,7 @@ static int nv_tx_done_optimized(struct net_device *dev, int limit) * nv_tx_timeout: dev->tx_timeout function * Called with netif_tx_lock held. */ -static void nv_tx_timeout(struct net_device *dev) +static void nv_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct fe_priv *np = netdev_priv(dev); u8 __iomem *base = get_hwbase(dev); diff --git a/drivers/net/ethernet/nxp/lpc_eth.c b/drivers/net/ethernet/nxp/lpc_eth.c index 656169214cdb..d20cf03a3ea0 100644 --- a/drivers/net/ethernet/nxp/lpc_eth.c +++ b/drivers/net/ethernet/nxp/lpc_eth.c @@ -1149,19 +1149,6 @@ static void lpc_eth_set_multicast_list(struct net_device *ndev) spin_unlock_irqrestore(&pldat->lock, flags); } -static int lpc_eth_ioctl(struct net_device *ndev, struct ifreq *req, int cmd) -{ - struct phy_device *phydev = ndev->phydev; - - if (!netif_running(ndev)) - return -EINVAL; - - if (!phydev) - return -ENODEV; - - return phy_mii_ioctl(phydev, req, cmd); -} - static int lpc_eth_open(struct net_device *ndev) { struct netdata_local *pldat = netdev_priv(ndev); @@ -1229,7 +1216,7 @@ static const struct net_device_ops lpc_netdev_ops = { .ndo_stop = lpc_eth_close, .ndo_start_xmit = lpc_eth_hard_start_xmit, .ndo_set_rx_mode = lpc_eth_set_multicast_list, - .ndo_do_ioctl = lpc_eth_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_mac_address = lpc_set_mac_address, .ndo_validate_addr = eth_validate_addr, }; diff --git a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c index 18e6d87c607b..73ec195fbc30 100644 --- a/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c +++ b/drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c @@ -2271,7 +2271,7 @@ static int pch_gbe_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) * pch_gbe_tx_timeout - Respond to a Tx Hang * @netdev: Network interface device structure */ -static void pch_gbe_tx_timeout(struct net_device *netdev) +static void pch_gbe_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct pch_gbe_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/packetengines/hamachi.c b/drivers/net/ethernet/packetengines/hamachi.c index eee883a2aa8d..70816d2e2990 100644 --- a/drivers/net/ethernet/packetengines/hamachi.c +++ b/drivers/net/ethernet/packetengines/hamachi.c @@ -548,7 +548,7 @@ static void mdio_write(struct net_device *dev, int phy_id, int location, int val static int hamachi_open(struct net_device *dev); static int netdev_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static void hamachi_timer(struct timer_list *t); -static void hamachi_tx_timeout(struct net_device *dev); +static void hamachi_tx_timeout(struct net_device *dev, unsigned int txqueue); static void hamachi_init_ring(struct net_device *dev); static netdev_tx_t hamachi_start_xmit(struct sk_buff *skb, struct net_device *dev); @@ -1042,7 +1042,7 @@ static void hamachi_timer(struct timer_list *t) add_timer(&hmp->timer); } -static void hamachi_tx_timeout(struct net_device *dev) +static void hamachi_tx_timeout(struct net_device *dev, unsigned int txqueue) { int i; struct hamachi_private *hmp = netdev_priv(dev); diff --git a/drivers/net/ethernet/packetengines/yellowfin.c b/drivers/net/ethernet/packetengines/yellowfin.c index 5113ee647090..520779f05e1a 100644 --- a/drivers/net/ethernet/packetengines/yellowfin.c +++ b/drivers/net/ethernet/packetengines/yellowfin.c @@ -344,7 +344,7 @@ static void mdio_write(void __iomem *ioaddr, int phy_id, int location, int value static int netdev_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static int yellowfin_open(struct net_device *dev); static void yellowfin_timer(struct timer_list *t); -static void yellowfin_tx_timeout(struct net_device *dev); +static void yellowfin_tx_timeout(struct net_device *dev, unsigned int txqueue); static int yellowfin_init_ring(struct net_device *dev); static netdev_tx_t yellowfin_start_xmit(struct sk_buff *skb, struct net_device *dev); @@ -677,7 +677,7 @@ static void yellowfin_timer(struct timer_list *t) add_timer(&yp->timer); } -static void yellowfin_tx_timeout(struct net_device *dev) +static void yellowfin_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct yellowfin_private *yp = netdev_priv(dev); void __iomem *ioaddr = yp->base; diff --git a/drivers/net/ethernet/pensando/ionic/ionic.h b/drivers/net/ethernet/pensando/ionic/ionic.h index 98e102af7756..bb106a32f416 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic.h +++ b/drivers/net/ethernet/pensando/ionic/ionic.h @@ -12,19 +12,27 @@ struct ionic_lif; #define IONIC_DRV_NAME "ionic" #define IONIC_DRV_DESCRIPTION "Pensando Ethernet NIC Driver" -#define IONIC_DRV_VERSION "0.18.0-k" +#define IONIC_DRV_VERSION "0.20.0-k" #define PCI_VENDOR_ID_PENSANDO 0x1dd8 #define PCI_DEVICE_ID_PENSANDO_IONIC_ETH_PF 0x1002 #define PCI_DEVICE_ID_PENSANDO_IONIC_ETH_VF 0x1003 -#define IONIC_SUBDEV_ID_NAPLES_25 0x4000 -#define IONIC_SUBDEV_ID_NAPLES_100_4 0x4001 -#define IONIC_SUBDEV_ID_NAPLES_100_8 0x4002 - #define DEVCMD_TIMEOUT 10 +struct ionic_vf { + u16 index; + u8 macaddr[6]; + __le32 maxrate; + __le16 vlanid; + u8 spoofchk; + u8 trusted; + u8 linkstate; + dma_addr_t stats_pa; + struct ionic_lif_stats stats; +}; + struct ionic { struct pci_dev *pdev; struct device *dev; @@ -46,6 +54,9 @@ struct ionic { DECLARE_BITMAP(intrs, IONIC_INTR_CTRL_REGS_MAX); struct work_struct nb_work; struct notifier_block nb; + struct rw_semaphore vf_op_lock; /* lock for VF operations */ + struct ionic_vf *vfs; + int num_vfs; struct timer_list watchdog_timer; int watchdog_period; }; diff --git a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c index 9a9ab8cb2cb3..448d7b23b2f7 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_bus_pci.c @@ -104,10 +104,112 @@ void ionic_bus_unmap_dbpage(struct ionic *ionic, void __iomem *page) iounmap(page); } +static void ionic_vf_dealloc_locked(struct ionic *ionic) +{ + struct ionic_vf *v; + dma_addr_t dma = 0; + int i; + + if (!ionic->vfs) + return; + + for (i = ionic->num_vfs - 1; i >= 0; i--) { + v = &ionic->vfs[i]; + + if (v->stats_pa) { + (void)ionic_set_vf_config(ionic, i, + IONIC_VF_ATTR_STATSADDR, + (u8 *)&dma); + dma_unmap_single(ionic->dev, v->stats_pa, + sizeof(v->stats), DMA_FROM_DEVICE); + v->stats_pa = 0; + } + } + + kfree(ionic->vfs); + ionic->vfs = NULL; + ionic->num_vfs = 0; +} + +static void ionic_vf_dealloc(struct ionic *ionic) +{ + down_write(&ionic->vf_op_lock); + ionic_vf_dealloc_locked(ionic); + up_write(&ionic->vf_op_lock); +} + +static int ionic_vf_alloc(struct ionic *ionic, int num_vfs) +{ + struct ionic_vf *v; + int err = 0; + int i; + + down_write(&ionic->vf_op_lock); + + ionic->vfs = kcalloc(num_vfs, sizeof(struct ionic_vf), GFP_KERNEL); + if (!ionic->vfs) { + err = -ENOMEM; + goto out; + } + + for (i = 0; i < num_vfs; i++) { + v = &ionic->vfs[i]; + v->stats_pa = dma_map_single(ionic->dev, &v->stats, + sizeof(v->stats), DMA_FROM_DEVICE); + if (dma_mapping_error(ionic->dev, v->stats_pa)) { + v->stats_pa = 0; + err = -ENODEV; + goto out; + } + + /* ignore failures from older FW, we just won't get stats */ + (void)ionic_set_vf_config(ionic, i, IONIC_VF_ATTR_STATSADDR, + (u8 *)&v->stats_pa); + ionic->num_vfs++; + } + +out: + if (err) + ionic_vf_dealloc_locked(ionic); + up_write(&ionic->vf_op_lock); + return err; +} + +static int ionic_sriov_configure(struct pci_dev *pdev, int num_vfs) +{ + struct ionic *ionic = pci_get_drvdata(pdev); + struct device *dev = ionic->dev; + int ret = 0; + + if (num_vfs > 0) { + ret = pci_enable_sriov(pdev, num_vfs); + if (ret) { + dev_err(dev, "Cannot enable SRIOV: %d\n", ret); + goto out; + } + + ret = ionic_vf_alloc(ionic, num_vfs); + if (ret) { + dev_err(dev, "Cannot alloc VFs: %d\n", ret); + pci_disable_sriov(pdev); + goto out; + } + + ret = num_vfs; + } else { + pci_disable_sriov(pdev); + ionic_vf_dealloc(ionic); + } + +out: + return ret; +} + static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) { struct device *dev = &pdev->dev; struct ionic *ionic; + int num_vfs; int err; ionic = ionic_devlink_alloc(dev); @@ -206,6 +308,15 @@ static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) goto err_out_free_lifs; } + init_rwsem(&ionic->vf_op_lock); + num_vfs = pci_num_vf(pdev); + if (num_vfs) { + dev_info(dev, "%d VFs found already enabled\n", num_vfs); + err = ionic_vf_alloc(ionic, num_vfs); + if (err) + dev_err(dev, "Cannot enable existing VFs: %d\n", err); + } + err = ionic_lifs_register(ionic); if (err) { dev_err(dev, "Cannot register LIFs: %d, aborting\n", err); @@ -223,6 +334,7 @@ static int ionic_probe(struct pci_dev *pdev, const struct pci_device_id *ent) err_out_deregister_lifs: ionic_lifs_unregister(ionic); err_out_deinit_lifs: + ionic_vf_dealloc(ionic); ionic_lifs_deinit(ionic); err_out_free_lifs: ionic_lifs_free(ionic); @@ -279,6 +391,7 @@ static struct pci_driver ionic_driver = { .id_table = ionic_id_table, .probe = ionic_probe, .remove = ionic_remove, + .sriov_configure = ionic_sriov_configure, }; int ionic_bus_register_driver(void) diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.c b/drivers/net/ethernet/pensando/ionic/ionic_dev.c index 5f9d2ec70446..87f82f36812f 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_dev.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.c @@ -286,6 +286,64 @@ void ionic_dev_cmd_port_pause(struct ionic_dev *idev, u8 pause_type) ionic_dev_cmd_go(idev, &cmd); } +/* VF commands */ +int ionic_set_vf_config(struct ionic *ionic, int vf, u8 attr, u8 *data) +{ + union ionic_dev_cmd cmd = { + .vf_setattr.opcode = IONIC_CMD_VF_SETATTR, + .vf_setattr.attr = attr, + .vf_setattr.vf_index = vf, + }; + int err; + + switch (attr) { + case IONIC_VF_ATTR_SPOOFCHK: + cmd.vf_setattr.spoofchk = *data; + dev_dbg(ionic->dev, "%s: vf %d spoof %d\n", + __func__, vf, *data); + break; + case IONIC_VF_ATTR_TRUST: + cmd.vf_setattr.trust = *data; + dev_dbg(ionic->dev, "%s: vf %d trust %d\n", + __func__, vf, *data); + break; + case IONIC_VF_ATTR_LINKSTATE: + cmd.vf_setattr.linkstate = *data; + dev_dbg(ionic->dev, "%s: vf %d linkstate %d\n", + __func__, vf, *data); + break; + case IONIC_VF_ATTR_MAC: + ether_addr_copy(cmd.vf_setattr.macaddr, data); + dev_dbg(ionic->dev, "%s: vf %d macaddr %pM\n", + __func__, vf, data); + break; + case IONIC_VF_ATTR_VLAN: + cmd.vf_setattr.vlanid = cpu_to_le16(*(u16 *)data); + dev_dbg(ionic->dev, "%s: vf %d vlan %d\n", + __func__, vf, *(u16 *)data); + break; + case IONIC_VF_ATTR_RATE: + cmd.vf_setattr.maxrate = cpu_to_le32(*(u32 *)data); + dev_dbg(ionic->dev, "%s: vf %d maxrate %d\n", + __func__, vf, *(u32 *)data); + break; + case IONIC_VF_ATTR_STATSADDR: + cmd.vf_setattr.stats_pa = cpu_to_le64(*(u64 *)data); + dev_dbg(ionic->dev, "%s: vf %d stats_pa 0x%08llx\n", + __func__, vf, *(u64 *)data); + break; + default: + return -EINVAL; + } + + mutex_lock(&ionic->dev_cmd_lock); + ionic_dev_cmd_go(&ionic->idev, &cmd); + err = ionic_dev_cmd_wait(ionic, DEVCMD_TIMEOUT); + mutex_unlock(&ionic->dev_cmd_lock); + + return err; +} + /* LIF commands */ void ionic_dev_cmd_lif_identify(struct ionic_dev *idev, u8 type, u8 ver) { diff --git a/drivers/net/ethernet/pensando/ionic/ionic_dev.h b/drivers/net/ethernet/pensando/ionic/ionic_dev.h index 4665c5dc5324..7838e342c4fd 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_dev.h +++ b/drivers/net/ethernet/pensando/ionic/ionic_dev.h @@ -113,6 +113,12 @@ static_assert(sizeof(struct ionic_rxq_desc) == 16); static_assert(sizeof(struct ionic_rxq_sg_desc) == 128); static_assert(sizeof(struct ionic_rxq_comp) == 16); +/* SR/IOV */ +static_assert(sizeof(struct ionic_vf_setattr_cmd) == 64); +static_assert(sizeof(struct ionic_vf_setattr_comp) == 16); +static_assert(sizeof(struct ionic_vf_getattr_cmd) == 64); +static_assert(sizeof(struct ionic_vf_getattr_comp) == 16); + struct ionic_devinfo { u8 asic_type; u8 asic_rev; @@ -275,6 +281,7 @@ void ionic_dev_cmd_port_autoneg(struct ionic_dev *idev, u8 an_enable); void ionic_dev_cmd_port_fec(struct ionic_dev *idev, u8 fec_type); void ionic_dev_cmd_port_pause(struct ionic_dev *idev, u8 pause_type); +int ionic_set_vf_config(struct ionic *ionic, int vf, u8 attr, u8 *data); void ionic_dev_cmd_lif_identify(struct ionic_dev *idev, u8 type, u8 ver); void ionic_dev_cmd_lif_init(struct ionic_dev *idev, u16 lif_index, dma_addr_t addr); diff --git a/drivers/net/ethernet/pensando/ionic/ionic_if.h b/drivers/net/ethernet/pensando/ionic/ionic_if.h index 39317cdfa6cf..f131adad96e3 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_if.h +++ b/drivers/net/ethernet/pensando/ionic/ionic_if.h @@ -51,6 +51,10 @@ enum ionic_cmd_opcode { IONIC_CMD_RDMA_CREATE_CQ = 52, IONIC_CMD_RDMA_CREATE_ADMINQ = 53, + /* SR/IOV commands */ + IONIC_CMD_VF_GETATTR = 60, + IONIC_CMD_VF_SETATTR = 61, + /* QoS commands */ IONIC_CMD_QOS_CLASS_IDENTIFY = 240, IONIC_CMD_QOS_CLASS_INIT = 241, @@ -1639,6 +1643,93 @@ enum ionic_qos_sched_type { IONIC_QOS_SCHED_TYPE_DWRR = 1, /* Deficit weighted round-robin */ }; +enum ionic_vf_attr { + IONIC_VF_ATTR_SPOOFCHK = 1, + IONIC_VF_ATTR_TRUST = 2, + IONIC_VF_ATTR_MAC = 3, + IONIC_VF_ATTR_LINKSTATE = 4, + IONIC_VF_ATTR_VLAN = 5, + IONIC_VF_ATTR_RATE = 6, + IONIC_VF_ATTR_STATSADDR = 7, +}; + +/** + * VF link status + */ +enum ionic_vf_link_status { + IONIC_VF_LINK_STATUS_AUTO = 0, /* link state of the uplink */ + IONIC_VF_LINK_STATUS_UP = 1, /* link is always up */ + IONIC_VF_LINK_STATUS_DOWN = 2, /* link is always down */ +}; + +/** + * struct ionic_vf_setattr_cmd - Set VF attributes on the NIC + * @opcode: Opcode + * @index: VF index + * @attr: Attribute type (enum ionic_vf_attr) + * macaddr mac address + * vlanid vlan ID + * maxrate max Tx rate in Mbps + * spoofchk enable address spoof checking + * trust enable VF trust + * linkstate set link up or down + * stats_pa set DMA address for VF stats + */ +struct ionic_vf_setattr_cmd { + u8 opcode; + u8 attr; + __le16 vf_index; + union { + u8 macaddr[6]; + __le16 vlanid; + __le32 maxrate; + u8 spoofchk; + u8 trust; + u8 linkstate; + __le64 stats_pa; + u8 pad[60]; + }; +}; + +struct ionic_vf_setattr_comp { + u8 status; + u8 attr; + __le16 vf_index; + __le16 comp_index; + u8 rsvd[9]; + u8 color; +}; + +/** + * struct ionic_vf_getattr_cmd - Get VF attributes from the NIC + * @opcode: Opcode + * @index: VF index + * @attr: Attribute type (enum ionic_vf_attr) + */ +struct ionic_vf_getattr_cmd { + u8 opcode; + u8 attr; + __le16 vf_index; + u8 rsvd[60]; +}; + +struct ionic_vf_getattr_comp { + u8 status; + u8 attr; + __le16 vf_index; + union { + u8 macaddr[6]; + __le16 vlanid; + __le32 maxrate; + u8 spoofchk; + u8 trust; + u8 linkstate; + __le64 stats_pa; + u8 pad[11]; + }; + u8 color; +}; + /** * union ionic_qos_config - Qos configuration structure * @flags: Configuration flags @@ -2289,6 +2380,9 @@ union ionic_dev_cmd { struct ionic_port_getattr_cmd port_getattr; struct ionic_port_setattr_cmd port_setattr; + struct ionic_vf_setattr_cmd vf_setattr; + struct ionic_vf_getattr_cmd vf_getattr; + struct ionic_lif_identify_cmd lif_identify; struct ionic_lif_init_cmd lif_init; struct ionic_lif_reset_cmd lif_reset; @@ -2318,6 +2412,9 @@ union ionic_dev_cmd_comp { struct ionic_port_getattr_comp port_getattr; struct ionic_port_setattr_comp port_setattr; + struct ionic_vf_setattr_comp vf_setattr; + struct ionic_vf_getattr_comp vf_getattr; + struct ionic_lif_identify_comp lif_identify; struct ionic_lif_init_comp lif_init; ionic_lif_reset_comp lif_reset; diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.c b/drivers/net/ethernet/pensando/ionic/ionic_lif.c index ef8258713369..191271f6260d 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_lif.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.c @@ -1285,7 +1285,7 @@ static void ionic_tx_timeout_work(struct work_struct *ws) rtnl_unlock(); } -static void ionic_tx_timeout(struct net_device *netdev) +static void ionic_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct ionic_lif *lif = netdev_priv(netdev); @@ -1619,6 +1619,227 @@ int ionic_stop(struct net_device *netdev) return err; } +static int ionic_get_vf_config(struct net_device *netdev, + int vf, struct ifla_vf_info *ivf) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + int ret = 0; + + down_read(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ivf->vf = vf; + ivf->vlan = ionic->vfs[vf].vlanid; + ivf->qos = 0; + ivf->spoofchk = ionic->vfs[vf].spoofchk; + ivf->linkstate = ionic->vfs[vf].linkstate; + ivf->max_tx_rate = ionic->vfs[vf].maxrate; + ivf->trusted = ionic->vfs[vf].trusted; + ether_addr_copy(ivf->mac, ionic->vfs[vf].macaddr); + } + + up_read(&ionic->vf_op_lock); + return ret; +} + +static int ionic_get_vf_stats(struct net_device *netdev, int vf, + struct ifla_vf_stats *vf_stats) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + struct ionic_lif_stats *vs; + int ret = 0; + + down_read(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + memset(vf_stats, 0, sizeof(*vf_stats)); + vs = &ionic->vfs[vf].stats; + + vf_stats->rx_packets = le64_to_cpu(vs->rx_ucast_packets); + vf_stats->tx_packets = le64_to_cpu(vs->tx_ucast_packets); + vf_stats->rx_bytes = le64_to_cpu(vs->rx_ucast_bytes); + vf_stats->tx_bytes = le64_to_cpu(vs->tx_ucast_bytes); + vf_stats->broadcast = le64_to_cpu(vs->rx_bcast_packets); + vf_stats->multicast = le64_to_cpu(vs->rx_mcast_packets); + vf_stats->rx_dropped = le64_to_cpu(vs->rx_ucast_drop_packets) + + le64_to_cpu(vs->rx_mcast_drop_packets) + + le64_to_cpu(vs->rx_bcast_drop_packets); + vf_stats->tx_dropped = le64_to_cpu(vs->tx_ucast_drop_packets) + + le64_to_cpu(vs->tx_mcast_drop_packets) + + le64_to_cpu(vs->tx_bcast_drop_packets); + } + + up_read(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_mac(struct net_device *netdev, int vf, u8 *mac) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + int ret; + + if (!(is_zero_ether_addr(mac) || is_valid_ether_addr(mac))) + return -EINVAL; + + down_read(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, IONIC_VF_ATTR_MAC, mac); + if (!ret) + ether_addr_copy(ionic->vfs[vf].macaddr, mac); + } + + up_read(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_vlan(struct net_device *netdev, int vf, u16 vlan, + u8 qos, __be16 proto) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + int ret; + + /* until someday when we support qos */ + if (qos) + return -EINVAL; + + if (vlan > 4095) + return -EINVAL; + + if (proto != htons(ETH_P_8021Q)) + return -EPROTONOSUPPORT; + + down_read(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, + IONIC_VF_ATTR_VLAN, (u8 *)&vlan); + if (!ret) + ionic->vfs[vf].vlanid = vlan; + } + + up_read(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_rate(struct net_device *netdev, int vf, + int tx_min, int tx_max) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + int ret; + + /* setting the min just seems silly */ + if (tx_min) + return -EINVAL; + + down_write(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, + IONIC_VF_ATTR_RATE, (u8 *)&tx_max); + if (!ret) + lif->ionic->vfs[vf].maxrate = tx_max; + } + + up_write(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_spoofchk(struct net_device *netdev, int vf, bool set) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + u8 data = set; /* convert to u8 for config */ + int ret; + + down_write(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, + IONIC_VF_ATTR_SPOOFCHK, &data); + if (!ret) + ionic->vfs[vf].spoofchk = data; + } + + up_write(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_trust(struct net_device *netdev, int vf, bool set) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + u8 data = set; /* convert to u8 for config */ + int ret; + + down_write(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, + IONIC_VF_ATTR_TRUST, &data); + if (!ret) + ionic->vfs[vf].trusted = data; + } + + up_write(&ionic->vf_op_lock); + return ret; +} + +static int ionic_set_vf_link_state(struct net_device *netdev, int vf, int set) +{ + struct ionic_lif *lif = netdev_priv(netdev); + struct ionic *ionic = lif->ionic; + u8 data; + int ret; + + switch (set) { + case IFLA_VF_LINK_STATE_ENABLE: + data = IONIC_VF_LINK_STATUS_UP; + break; + case IFLA_VF_LINK_STATE_DISABLE: + data = IONIC_VF_LINK_STATUS_DOWN; + break; + case IFLA_VF_LINK_STATE_AUTO: + data = IONIC_VF_LINK_STATUS_AUTO; + break; + default: + return -EINVAL; + } + + down_write(&ionic->vf_op_lock); + + if (vf >= pci_num_vf(ionic->pdev) || !ionic->vfs) { + ret = -EINVAL; + } else { + ret = ionic_set_vf_config(ionic, vf, + IONIC_VF_ATTR_LINKSTATE, &data); + if (!ret) + ionic->vfs[vf].linkstate = set; + } + + up_write(&ionic->vf_op_lock); + return ret; +} + static const struct net_device_ops ionic_netdev_ops = { .ndo_open = ionic_open, .ndo_stop = ionic_stop, @@ -1632,6 +1853,14 @@ static const struct net_device_ops ionic_netdev_ops = { .ndo_change_mtu = ionic_change_mtu, .ndo_vlan_rx_add_vid = ionic_vlan_rx_add_vid, .ndo_vlan_rx_kill_vid = ionic_vlan_rx_kill_vid, + .ndo_set_vf_vlan = ionic_set_vf_vlan, + .ndo_set_vf_trust = ionic_set_vf_trust, + .ndo_set_vf_mac = ionic_set_vf_mac, + .ndo_set_vf_rate = ionic_set_vf_rate, + .ndo_set_vf_spoofchk = ionic_set_vf_spoofchk, + .ndo_get_vf_config = ionic_get_vf_config, + .ndo_set_vf_link_state = ionic_set_vf_link_state, + .ndo_get_vf_stats = ionic_get_vf_stats, }; int ionic_reset_queues(struct ionic_lif *lif) @@ -1965,18 +2194,22 @@ static int ionic_station_set(struct ionic_lif *lif) if (err) return err; + if (is_zero_ether_addr(ctx.comp.lif_getattr.mac)) + return 0; + memcpy(addr.sa_data, ctx.comp.lif_getattr.mac, netdev->addr_len); addr.sa_family = AF_INET; err = eth_prepare_mac_addr_change(netdev, &addr); - if (err) - return err; - - if (!is_zero_ether_addr(netdev->dev_addr)) { - netdev_dbg(lif->netdev, "deleting station MAC addr %pM\n", - netdev->dev_addr); - ionic_lif_addr(lif, netdev->dev_addr, false); + if (err) { + netdev_warn(lif->netdev, "ignoring bad MAC addr from NIC %pM\n", + addr.sa_data); + return 0; } + netdev_dbg(lif->netdev, "deleting station MAC addr %pM\n", + netdev->dev_addr); + ionic_lif_addr(lif, netdev->dev_addr, false); + eth_commit_mac_addr_change(netdev, &addr); netdev_dbg(lif->netdev, "adding station MAC addr %pM\n", netdev->dev_addr); diff --git a/drivers/net/ethernet/pensando/ionic/ionic_lif.h b/drivers/net/ethernet/pensando/ionic/ionic_lif.h index a55fd1f8c31b..9c5a7dd45f9d 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_lif.h +++ b/drivers/net/ethernet/pensando/ionic/ionic_lif.h @@ -37,6 +37,7 @@ struct ionic_rx_stats { u64 csum_complete; u64 csum_error; u64 buffers_posted; + u64 dropped; }; #define IONIC_QCQ_F_INITED BIT(0) diff --git a/drivers/net/ethernet/pensando/ionic/ionic_main.c b/drivers/net/ethernet/pensando/ionic/ionic_main.c index 3590ea7fd88a..a8e3fb73b465 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_main.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_main.c @@ -165,6 +165,10 @@ static const char *ionic_opcode_to_str(enum ionic_cmd_opcode opcode) return "IONIC_CMD_FW_DOWNLOAD"; case IONIC_CMD_FW_CONTROL: return "IONIC_CMD_FW_CONTROL"; + case IONIC_CMD_VF_GETATTR: + return "IONIC_CMD_VF_GETATTR"; + case IONIC_CMD_VF_SETATTR: + return "IONIC_CMD_VF_SETATTR"; default: return "DEVCMD_UNKNOWN"; } @@ -326,9 +330,9 @@ int ionic_dev_cmd_wait(struct ionic *ionic, unsigned long max_seconds) unsigned long max_wait; unsigned long duration; int opcode; + int hb = 0; int done; int err; - int hb; WARN_ON(in_interrupt()); diff --git a/drivers/net/ethernet/pensando/ionic/ionic_stats.c b/drivers/net/ethernet/pensando/ionic/ionic_stats.c index 03916b6d47f2..a1e9796a660a 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_stats.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_stats.c @@ -39,6 +39,7 @@ static const struct ionic_stat_desc ionic_rx_stats_desc[] = { IONIC_RX_STAT_DESC(csum_none), IONIC_RX_STAT_DESC(csum_complete), IONIC_RX_STAT_DESC(csum_error), + IONIC_RX_STAT_DESC(dropped), }; static const struct ionic_stat_desc ionic_txq_stats_desc[] = { diff --git a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c index 97e79949b359..e452f4242ba0 100644 --- a/drivers/net/ethernet/pensando/ionic/ionic_txrx.c +++ b/drivers/net/ethernet/pensando/ionic/ionic_txrx.c @@ -152,12 +152,16 @@ static void ionic_rx_clean(struct ionic_queue *q, struct ionic_desc_info *desc_i stats = q_to_rx_stats(q); netdev = q->lif->netdev; - if (comp->status) + if (comp->status) { + stats->dropped++; return; + } /* no packet processing while resetting */ - if (unlikely(test_bit(IONIC_LIF_QUEUE_RESET, q->lif->state))) + if (unlikely(test_bit(IONIC_LIF_QUEUE_RESET, q->lif->state))) { + stats->dropped++; return; + } stats->pkts++; stats->bytes += le16_to_cpu(comp->len); @@ -167,8 +171,10 @@ static void ionic_rx_clean(struct ionic_queue *q, struct ionic_desc_info *desc_i else skb = ionic_rx_frags(q, desc_info, cq_info); - if (unlikely(!skb)) + if (unlikely(!skb)) { + stats->dropped++; return; + } skb_record_rx_queue(skb, q->index); @@ -337,6 +343,8 @@ void ionic_rx_fill(struct ionic_queue *q) struct ionic_rxq_sg_desc *sg_desc; struct ionic_rxq_sg_elem *sg_elem; struct ionic_rxq_desc *desc; + unsigned int remain_len; + unsigned int seg_len; unsigned int nfrags; bool ring_doorbell; unsigned int i, j; @@ -346,6 +354,7 @@ void ionic_rx_fill(struct ionic_queue *q) nfrags = round_up(len, PAGE_SIZE) / PAGE_SIZE; for (i = ionic_q_space_avail(q); i; i--) { + remain_len = len; desc_info = q->head; desc = desc_info->desc; sg_desc = desc_info->sg_desc; @@ -369,7 +378,9 @@ void ionic_rx_fill(struct ionic_queue *q) return; } desc->addr = cpu_to_le64(page_info->dma_addr); - desc->len = cpu_to_le16(PAGE_SIZE); + seg_len = min_t(unsigned int, PAGE_SIZE, len); + desc->len = cpu_to_le16(seg_len); + remain_len -= seg_len; page_info++; /* fill sg descriptors - pages[1..n] */ @@ -385,7 +396,9 @@ void ionic_rx_fill(struct ionic_queue *q) return; } sg_elem->addr = cpu_to_le64(page_info->dma_addr); - sg_elem->len = cpu_to_le16(PAGE_SIZE); + seg_len = min_t(unsigned int, PAGE_SIZE, remain_len); + sg_elem->len = cpu_to_le16(seg_len); + remain_len -= seg_len; page_info++; } diff --git a/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c b/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c index c692a41e4548..8067ea04d455 100644 --- a/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c +++ b/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c @@ -49,7 +49,7 @@ static int netxen_nic_open(struct net_device *netdev); static int netxen_nic_close(struct net_device *netdev); static netdev_tx_t netxen_nic_xmit_frame(struct sk_buff *, struct net_device *); -static void netxen_tx_timeout(struct net_device *netdev); +static void netxen_tx_timeout(struct net_device *netdev, unsigned int txqueue); static void netxen_tx_timeout_task(struct work_struct *work); static void netxen_fw_poll_work(struct work_struct *work); static void netxen_schedule_work(struct netxen_adapter *adapter, @@ -2222,7 +2222,7 @@ static void netxen_nic_handle_phy_intr(struct netxen_adapter *adapter) netxen_advert_link_change(adapter, linkup); } -static void netxen_tx_timeout(struct net_device *netdev) +static void netxen_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct netxen_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/qlogic/qed/qed.h b/drivers/net/ethernet/qlogic/qed/qed.h index 89fe091c958d..fa41bf08a589 100644 --- a/drivers/net/ethernet/qlogic/qed/qed.h +++ b/drivers/net/ethernet/qlogic/qed/qed.h @@ -253,7 +253,8 @@ enum qed_resources { QED_VLAN, QED_RDMA_CNQ_RAM, QED_ILT, - QED_LL2_QUEUE, + QED_LL2_RAM_QUEUE, + QED_LL2_CTX_QUEUE, QED_CMDQS_CQS, QED_RDMA_STATS_QUEUE, QED_BDQ, @@ -461,6 +462,8 @@ struct qed_fw_data { const u8 *modes_tree_buf; union init_op *init_ops; const u32 *arr_data; + const u32 *fw_overlays; + u32 fw_overlays_len; u32 init_ops_size; }; @@ -531,6 +534,23 @@ struct qed_nvm_image_info { bool valid; }; +enum qed_hsi_def_type { + QED_HSI_DEF_MAX_NUM_VFS, + QED_HSI_DEF_MAX_NUM_L2_QUEUES, + QED_HSI_DEF_MAX_NUM_PORTS, + QED_HSI_DEF_MAX_SB_PER_PATH, + QED_HSI_DEF_MAX_NUM_PFS, + QED_HSI_DEF_MAX_NUM_VPORTS, + QED_HSI_DEF_NUM_ETH_RSS_ENGINE, + QED_HSI_DEF_MAX_QM_TX_QUEUES, + QED_HSI_DEF_NUM_PXP_ILT_RECORDS, + QED_HSI_DEF_NUM_RDMA_STATISTIC_COUNTERS, + QED_HSI_DEF_MAX_QM_GLOBAL_RLS, + QED_HSI_DEF_MAX_PBF_CMD_LINES, + QED_HSI_DEF_MAX_BTB_BLOCKS, + QED_NUM_HSI_DEFS +}; + #define DRV_MODULE_VERSION \ __stringify(QED_MAJOR_VERSION) "." \ __stringify(QED_MINOR_VERSION) "." \ @@ -646,6 +666,7 @@ struct qed_hwfn { struct dbg_tools_data dbg_info; void *dbg_user_info; + struct virt_mem_desc dbg_arrays[MAX_BIN_DBG_BUFFER_TYPE]; /* PWM region specific data */ u16 wid_count; @@ -668,6 +689,7 @@ struct qed_hwfn { /* Nvm images number and attributes */ struct qed_nvm_image_info nvm_info; + struct phys_mem_desc *fw_overlay_mem; struct qed_ptt *p_arfs_ptt; struct qed_simd_fp_handler simd_proto_handler[64]; @@ -796,8 +818,8 @@ struct qed_dev { u8 cache_shift; /* Init */ - const struct iro *iro_arr; -#define IRO (p_hwfn->cdev->iro_arr) + const u32 *iro_arr; +#define IRO ((const struct iro *)p_hwfn->cdev->iro_arr) /* HW functions */ u8 num_hwfns; @@ -856,6 +878,8 @@ struct qed_dev { struct qed_cb_ll2_info *ll2; u8 ll2_mac_address[ETH_ALEN]; #endif + struct qed_dbg_feature dbg_features[DBG_FEATURE_NUM]; + bool disable_ilt_dump; DECLARE_HASHTABLE(connections, 10); const struct firmware *firmware; @@ -868,16 +892,35 @@ struct qed_dev { bool iwarp_cmt; }; -#define NUM_OF_VFS(dev) (QED_IS_BB(dev) ? MAX_NUM_VFS_BB \ - : MAX_NUM_VFS_K2) -#define NUM_OF_L2_QUEUES(dev) (QED_IS_BB(dev) ? MAX_NUM_L2_QUEUES_BB \ - : MAX_NUM_L2_QUEUES_K2) -#define NUM_OF_PORTS(dev) (QED_IS_BB(dev) ? MAX_NUM_PORTS_BB \ - : MAX_NUM_PORTS_K2) -#define NUM_OF_SBS(dev) (QED_IS_BB(dev) ? MAX_SB_PER_PATH_BB \ - : MAX_SB_PER_PATH_K2) -#define NUM_OF_ENG_PFS(dev) (QED_IS_BB(dev) ? MAX_NUM_PFS_BB \ - : MAX_NUM_PFS_K2) +u32 qed_get_hsi_def_val(struct qed_dev *cdev, enum qed_hsi_def_type type); + +#define NUM_OF_VFS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_NUM_VFS) +#define NUM_OF_L2_QUEUES(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_NUM_L2_QUEUES) +#define NUM_OF_PORTS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_NUM_PORTS) +#define NUM_OF_SBS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_SB_PER_PATH) +#define NUM_OF_ENG_PFS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_NUM_PFS) +#define NUM_OF_VPORTS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_NUM_VPORTS) +#define NUM_OF_RSS_ENGINES(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_NUM_ETH_RSS_ENGINE) +#define NUM_OF_QM_TX_QUEUES(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_QM_TX_QUEUES) +#define NUM_OF_PXP_ILT_RECORDS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_NUM_PXP_ILT_RECORDS) +#define NUM_OF_RDMA_STATISTIC_COUNTERS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_NUM_RDMA_STATISTIC_COUNTERS) +#define NUM_OF_QM_GLOBAL_RLS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_QM_GLOBAL_RLS) +#define NUM_OF_PBF_CMD_LINES(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_PBF_CMD_LINES) +#define NUM_OF_BTB_BLOCKS(dev) \ + qed_get_hsi_def_val(dev, QED_HSI_DEF_MAX_BTB_BLOCKS) + /** * @brief qed_concrete_to_sw_fid - get the sw function id from diff --git a/drivers/net/ethernet/qlogic/qed/qed_cxt.c b/drivers/net/ethernet/qlogic/qed/qed_cxt.c index 8e1bdf58b9e7..fbfff2b1dc93 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_cxt.c +++ b/drivers/net/ethernet/qlogic/qed/qed_cxt.c @@ -50,12 +50,6 @@ #include "qed_reg_addr.h" #include "qed_sriov.h" -/* Max number of connection types in HW (DQ/CDU etc.) */ -#define MAX_CONN_TYPES PROTOCOLID_COMMON -#define NUM_TASK_TYPES 2 -#define NUM_TASK_PF_SEGMENTS 4 -#define NUM_TASK_VF_SEGMENTS 1 - /* QM constants */ #define QM_PQ_ELEMENT_SIZE 4 /* in bytes */ @@ -123,126 +117,6 @@ struct src_ent { /* Alignment is inherent to the type1_task_context structure */ #define TYPE1_TASK_CXT_SIZE(p_hwfn) sizeof(union type1_task_context) -/* PF per protocl configuration object */ -#define TASK_SEGMENTS (NUM_TASK_PF_SEGMENTS + NUM_TASK_VF_SEGMENTS) -#define TASK_SEGMENT_VF (NUM_TASK_PF_SEGMENTS) - -struct qed_tid_seg { - u32 count; - u8 type; - bool has_fl_mem; -}; - -struct qed_conn_type_cfg { - u32 cid_count; - u32 cids_per_vf; - struct qed_tid_seg tid_seg[TASK_SEGMENTS]; -}; - -/* ILT Client configuration, Per connection type (protocol) resources. */ -#define ILT_CLI_PF_BLOCKS (1 + NUM_TASK_PF_SEGMENTS * 2) -#define ILT_CLI_VF_BLOCKS (1 + NUM_TASK_VF_SEGMENTS * 2) -#define CDUC_BLK (0) -#define SRQ_BLK (0) -#define CDUT_SEG_BLK(n) (1 + (u8)(n)) -#define CDUT_FL_SEG_BLK(n, X) (1 + (n) + NUM_TASK_ ## X ## _SEGMENTS) - -enum ilt_clients { - ILT_CLI_CDUC, - ILT_CLI_CDUT, - ILT_CLI_QM, - ILT_CLI_TM, - ILT_CLI_SRC, - ILT_CLI_TSDM, - ILT_CLI_MAX -}; - -struct ilt_cfg_pair { - u32 reg; - u32 val; -}; - -struct qed_ilt_cli_blk { - u32 total_size; /* 0 means not active */ - u32 real_size_in_page; - u32 start_line; - u32 dynamic_line_cnt; -}; - -struct qed_ilt_client_cfg { - bool active; - - /* ILT boundaries */ - struct ilt_cfg_pair first; - struct ilt_cfg_pair last; - struct ilt_cfg_pair p_size; - - /* ILT client blocks for PF */ - struct qed_ilt_cli_blk pf_blks[ILT_CLI_PF_BLOCKS]; - u32 pf_total_lines; - - /* ILT client blocks for VFs */ - struct qed_ilt_cli_blk vf_blks[ILT_CLI_VF_BLOCKS]; - u32 vf_total_lines; -}; - -/* Per Path - - * ILT shadow table - * Protocol acquired CID lists - * PF start line in ILT - */ -struct qed_dma_mem { - dma_addr_t p_phys; - void *p_virt; - size_t size; -}; - -struct qed_cid_acquired_map { - u32 start_cid; - u32 max_count; - unsigned long *cid_map; -}; - -struct qed_cxt_mngr { - /* Per protocl configuration */ - struct qed_conn_type_cfg conn_cfg[MAX_CONN_TYPES]; - - /* computed ILT structure */ - struct qed_ilt_client_cfg clients[ILT_CLI_MAX]; - - /* Task type sizes */ - u32 task_type_size[NUM_TASK_TYPES]; - - /* total number of VFs for this hwfn - - * ALL VFs are symmetric in terms of HW resources - */ - u32 vf_count; - - /* Acquired CIDs */ - struct qed_cid_acquired_map acquired[MAX_CONN_TYPES]; - - struct qed_cid_acquired_map - acquired_vf[MAX_CONN_TYPES][MAX_NUM_VFS]; - - /* ILT shadow table */ - struct qed_dma_mem *ilt_shadow; - u32 pf_start_line; - - /* Mutex for a dynamic ILT allocation */ - struct mutex mutex; - - /* SRC T2 */ - struct qed_dma_mem *t2; - u32 t2_num_pages; - u64 first_free; - u64 last_free; - - /* total number of SRQ's for this hwfn */ - u32 srq_count; - - /* Maximal number of L2 steering filters */ - u32 arfs_count; -}; static bool src_proto(enum protocol_type type) { return type == PROTOCOLID_ISCSI || @@ -880,30 +754,60 @@ u32 qed_cxt_cfg_ilt_compute_excess(struct qed_hwfn *p_hwfn, u32 used_lines) static void qed_cxt_src_t2_free(struct qed_hwfn *p_hwfn) { - struct qed_cxt_mngr *p_mngr = p_hwfn->p_cxt_mngr; + struct qed_src_t2 *p_t2 = &p_hwfn->p_cxt_mngr->src_t2; u32 i; - if (!p_mngr->t2) + if (!p_t2 || !p_t2->dma_mem) return; - for (i = 0; i < p_mngr->t2_num_pages; i++) - if (p_mngr->t2[i].p_virt) + for (i = 0; i < p_t2->num_pages; i++) + if (p_t2->dma_mem[i].virt_addr) dma_free_coherent(&p_hwfn->cdev->pdev->dev, - p_mngr->t2[i].size, - p_mngr->t2[i].p_virt, - p_mngr->t2[i].p_phys); + p_t2->dma_mem[i].size, + p_t2->dma_mem[i].virt_addr, + p_t2->dma_mem[i].phys_addr); + + kfree(p_t2->dma_mem); + p_t2->dma_mem = NULL; +} + +static int +qed_cxt_t2_alloc_pages(struct qed_hwfn *p_hwfn, + struct qed_src_t2 *p_t2, u32 total_size, u32 page_size) +{ + void **p_virt; + u32 size, i; + + if (!p_t2 || !p_t2->dma_mem) + return -EINVAL; - kfree(p_mngr->t2); - p_mngr->t2 = NULL; + for (i = 0; i < p_t2->num_pages; i++) { + size = min_t(u32, total_size, page_size); + p_virt = &p_t2->dma_mem[i].virt_addr; + + *p_virt = dma_alloc_coherent(&p_hwfn->cdev->pdev->dev, + size, + &p_t2->dma_mem[i].phys_addr, + GFP_KERNEL); + if (!p_t2->dma_mem[i].virt_addr) + return -ENOMEM; + + memset(*p_virt, 0, size); + p_t2->dma_mem[i].size = size; + total_size -= size; + } + + return 0; } static int qed_cxt_src_t2_alloc(struct qed_hwfn *p_hwfn) { struct qed_cxt_mngr *p_mngr = p_hwfn->p_cxt_mngr; u32 conn_num, total_size, ent_per_page, psz, i; + struct phys_mem_desc *p_t2_last_page; struct qed_ilt_client_cfg *p_src; struct qed_src_iids src_iids; - struct qed_dma_mem *p_t2; + struct qed_src_t2 *p_t2; int rc; memset(&src_iids, 0, sizeof(src_iids)); @@ -921,49 +825,39 @@ static int qed_cxt_src_t2_alloc(struct qed_hwfn *p_hwfn) /* use the same page size as the SRC ILT client */ psz = ILT_PAGE_IN_BYTES(p_src->p_size.val); - p_mngr->t2_num_pages = DIV_ROUND_UP(total_size, psz); + p_t2 = &p_mngr->src_t2; + p_t2->num_pages = DIV_ROUND_UP(total_size, psz); /* allocate t2 */ - p_mngr->t2 = kcalloc(p_mngr->t2_num_pages, sizeof(struct qed_dma_mem), - GFP_KERNEL); - if (!p_mngr->t2) { + p_t2->dma_mem = kcalloc(p_t2->num_pages, sizeof(struct phys_mem_desc), + GFP_KERNEL); + if (!p_t2->dma_mem) { + DP_NOTICE(p_hwfn, "Failed to allocate t2 table\n"); rc = -ENOMEM; goto t2_fail; } - /* allocate t2 pages */ - for (i = 0; i < p_mngr->t2_num_pages; i++) { - u32 size = min_t(u32, total_size, psz); - void **p_virt = &p_mngr->t2[i].p_virt; - - *p_virt = dma_alloc_coherent(&p_hwfn->cdev->pdev->dev, size, - &p_mngr->t2[i].p_phys, - GFP_KERNEL); - if (!p_mngr->t2[i].p_virt) { - rc = -ENOMEM; - goto t2_fail; - } - p_mngr->t2[i].size = size; - total_size -= size; - } + rc = qed_cxt_t2_alloc_pages(p_hwfn, p_t2, total_size, psz); + if (rc) + goto t2_fail; /* Set the t2 pointers */ /* entries per page - must be a power of two */ ent_per_page = psz / sizeof(struct src_ent); - p_mngr->first_free = (u64) p_mngr->t2[0].p_phys; + p_t2->first_free = (u64)p_t2->dma_mem[0].phys_addr; - p_t2 = &p_mngr->t2[(conn_num - 1) / ent_per_page]; - p_mngr->last_free = (u64) p_t2->p_phys + + p_t2_last_page = &p_t2->dma_mem[(conn_num - 1) / ent_per_page]; + p_t2->last_free = (u64)p_t2_last_page->phys_addr + ((conn_num - 1) & (ent_per_page - 1)) * sizeof(struct src_ent); - for (i = 0; i < p_mngr->t2_num_pages; i++) { + for (i = 0; i < p_t2->num_pages; i++) { u32 ent_num = min_t(u32, ent_per_page, conn_num); - struct src_ent *entries = p_mngr->t2[i].p_virt; - u64 p_ent_phys = (u64) p_mngr->t2[i].p_phys, val; + struct src_ent *entries = p_t2->dma_mem[i].virt_addr; + u64 p_ent_phys = (u64)p_t2->dma_mem[i].phys_addr, val; u32 j; for (j = 0; j < ent_num - 1; j++) { @@ -971,8 +865,8 @@ static int qed_cxt_src_t2_alloc(struct qed_hwfn *p_hwfn) entries[j].next = cpu_to_be64(val); } - if (i < p_mngr->t2_num_pages - 1) - val = (u64) p_mngr->t2[i + 1].p_phys; + if (i < p_t2->num_pages - 1) + val = (u64)p_t2->dma_mem[i + 1].phys_addr; else val = 0; entries[j].next = cpu_to_be64(val); @@ -988,7 +882,7 @@ t2_fail: } #define for_each_ilt_valid_client(pos, clients) \ - for (pos = 0; pos < ILT_CLI_MAX; pos++) \ + for (pos = 0; pos < MAX_ILT_CLIENTS; pos++) \ if (!clients[pos].active) { \ continue; \ } else \ @@ -1014,13 +908,13 @@ static void qed_ilt_shadow_free(struct qed_hwfn *p_hwfn) ilt_size = qed_cxt_ilt_shadow_size(p_cli); for (i = 0; p_mngr->ilt_shadow && i < ilt_size; i++) { - struct qed_dma_mem *p_dma = &p_mngr->ilt_shadow[i]; + struct phys_mem_desc *p_dma = &p_mngr->ilt_shadow[i]; - if (p_dma->p_virt) + if (p_dma->virt_addr) dma_free_coherent(&p_hwfn->cdev->pdev->dev, - p_dma->size, p_dma->p_virt, - p_dma->p_phys); - p_dma->p_virt = NULL; + p_dma->size, p_dma->virt_addr, + p_dma->phys_addr); + p_dma->virt_addr = NULL; } kfree(p_mngr->ilt_shadow); } @@ -1030,7 +924,7 @@ static int qed_ilt_blk_alloc(struct qed_hwfn *p_hwfn, enum ilt_clients ilt_client, u32 start_line_offset) { - struct qed_dma_mem *ilt_shadow = p_hwfn->p_cxt_mngr->ilt_shadow; + struct phys_mem_desc *ilt_shadow = p_hwfn->p_cxt_mngr->ilt_shadow; u32 lines, line, sz_left, lines_to_skip = 0; /* Special handling for RoCE that supports dynamic allocation */ @@ -1059,8 +953,8 @@ static int qed_ilt_blk_alloc(struct qed_hwfn *p_hwfn, if (!p_virt) return -ENOMEM; - ilt_shadow[line].p_phys = p_phys; - ilt_shadow[line].p_virt = p_virt; + ilt_shadow[line].phys_addr = p_phys; + ilt_shadow[line].virt_addr = p_virt; ilt_shadow[line].size = size; DP_VERBOSE(p_hwfn, QED_MSG_ILT, @@ -1083,7 +977,7 @@ static int qed_ilt_shadow_alloc(struct qed_hwfn *p_hwfn) int rc; size = qed_cxt_ilt_shadow_size(clients); - p_mngr->ilt_shadow = kcalloc(size, sizeof(struct qed_dma_mem), + p_mngr->ilt_shadow = kcalloc(size, sizeof(struct phys_mem_desc), GFP_KERNEL); if (!p_mngr->ilt_shadow) { rc = -ENOMEM; @@ -1092,7 +986,7 @@ static int qed_ilt_shadow_alloc(struct qed_hwfn *p_hwfn) DP_VERBOSE(p_hwfn, QED_MSG_ILT, "Allocated 0x%x bytes for ilt shadow\n", - (u32)(size * sizeof(struct qed_dma_mem))); + (u32)(size * sizeof(struct phys_mem_desc))); for_each_ilt_valid_client(i, clients) { for (j = 0; j < ILT_CLI_PF_BLOCKS; j++) { @@ -1238,15 +1132,20 @@ int qed_cxt_mngr_alloc(struct qed_hwfn *p_hwfn) clients[ILT_CLI_TSDM].last.reg = ILT_CFG_REG(TSDM, LAST_ILT); clients[ILT_CLI_TSDM].p_size.reg = ILT_CFG_REG(TSDM, P_SIZE); /* default ILT page size for all clients is 64K */ - for (i = 0; i < ILT_CLI_MAX; i++) + for (i = 0; i < MAX_ILT_CLIENTS; i++) p_mngr->clients[i].p_size.val = ILT_DEFAULT_HW_P_SIZE; + p_mngr->conn_ctx_size = CONN_CXT_SIZE(p_hwfn); + /* Initialize task sizes */ p_mngr->task_type_size[0] = TYPE0_TASK_CXT_SIZE(p_hwfn); p_mngr->task_type_size[1] = TYPE1_TASK_CXT_SIZE(p_hwfn); - if (p_hwfn->cdev->p_iov_info) + if (p_hwfn->cdev->p_iov_info) { p_mngr->vf_count = p_hwfn->cdev->p_iov_info->total_vfs; + p_mngr->first_vf_in_pf = + p_hwfn->cdev->p_iov_info->first_vf_in_pf; + } /* Initialize the dynamic ILT allocation mutex */ mutex_init(&p_mngr->mutex); @@ -1522,7 +1421,6 @@ void qed_qm_init_pf(struct qed_hwfn *p_hwfn, params.num_vports = qm_info->num_vports; params.pf_wfq = qm_info->pf_wfq; params.pf_rl = qm_info->pf_rl; - params.link_speed = p_link->speed; params.pq_params = qm_info->qm_pq_params; params.vport_params = qm_info->qm_vport_params; @@ -1674,7 +1572,7 @@ static void qed_ilt_init_pf(struct qed_hwfn *p_hwfn) { struct qed_ilt_client_cfg *clients; struct qed_cxt_mngr *p_mngr; - struct qed_dma_mem *p_shdw; + struct phys_mem_desc *p_shdw; u32 line, rt_offst, i; qed_ilt_bounds_init(p_hwfn); @@ -1699,15 +1597,15 @@ static void qed_ilt_init_pf(struct qed_hwfn *p_hwfn) /** p_virt could be NULL incase of dynamic * allocation */ - if (p_shdw[line].p_virt) { + if (p_shdw[line].virt_addr) { SET_FIELD(ilt_hw_entry, ILT_ENTRY_VALID, 1ULL); SET_FIELD(ilt_hw_entry, ILT_ENTRY_PHY_ADDR, - (p_shdw[line].p_phys >> 12)); + (p_shdw[line].phys_addr >> 12)); DP_VERBOSE(p_hwfn, QED_MSG_ILT, "Setting RT[0x%08x] from ILT[0x%08x] [Client is %d] to Physical addr: 0x%llx\n", rt_offst, line, i, - (u64)(p_shdw[line].p_phys >> 12)); + (u64)(p_shdw[line].phys_addr >> 12)); } STORE_RT_REG_AGG(p_hwfn, rt_offst, ilt_hw_entry); @@ -2050,10 +1948,10 @@ int qed_cxt_get_cid_info(struct qed_hwfn *p_hwfn, struct qed_cxt_info *p_info) line = p_info->iid / cxts_per_p; /* Make sure context is allocated (dynamic allocation) */ - if (!p_mngr->ilt_shadow[line].p_virt) + if (!p_mngr->ilt_shadow[line].virt_addr) return -EINVAL; - p_info->p_cxt = p_mngr->ilt_shadow[line].p_virt + + p_info->p_cxt = p_mngr->ilt_shadow[line].virt_addr + p_info->iid % cxts_per_p * conn_cxt_size; DP_VERBOSE(p_hwfn, (QED_MSG_ILT | QED_MSG_CXT), @@ -2234,7 +2132,7 @@ int qed_cxt_get_tid_mem_info(struct qed_hwfn *p_hwfn, for (i = 0; i < total_lines; i++) { shadow_line = i + p_fl_seg->start_line - p_hwfn->p_cxt_mngr->pf_start_line; - p_info->blocks[i] = p_mngr->ilt_shadow[shadow_line].p_virt; + p_info->blocks[i] = p_mngr->ilt_shadow[shadow_line].virt_addr; } p_info->waste = ILT_PAGE_IN_BYTES(p_cli->p_size.val) - p_fl_seg->real_size_in_page; @@ -2296,7 +2194,7 @@ qed_cxt_dynamic_ilt_alloc(struct qed_hwfn *p_hwfn, mutex_lock(&p_hwfn->p_cxt_mngr->mutex); - if (p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].p_virt) + if (p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].virt_addr) goto out0; p_ptt = qed_ptt_acquire(p_hwfn); @@ -2334,8 +2232,8 @@ qed_cxt_dynamic_ilt_alloc(struct qed_hwfn *p_hwfn, } } - p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].p_virt = p_virt; - p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].p_phys = p_phys; + p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].virt_addr = p_virt; + p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].phys_addr = p_phys; p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].size = p_blk->real_size_in_page; @@ -2345,9 +2243,9 @@ qed_cxt_dynamic_ilt_alloc(struct qed_hwfn *p_hwfn, ilt_hw_entry = 0; SET_FIELD(ilt_hw_entry, ILT_ENTRY_VALID, 1ULL); - SET_FIELD(ilt_hw_entry, - ILT_ENTRY_PHY_ADDR, - (p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].p_phys >> 12)); + SET_FIELD(ilt_hw_entry, ILT_ENTRY_PHY_ADDR, + (p_hwfn->p_cxt_mngr->ilt_shadow[shadow_line].phys_addr + >> 12)); /* Write via DMAE since the PSWRQ2_REG_ILT_MEMORY line is a wide-bus */ qed_dmae_host2grc(p_hwfn, p_ptt, (u64) (uintptr_t)&ilt_hw_entry, @@ -2434,16 +2332,16 @@ qed_cxt_free_ilt_range(struct qed_hwfn *p_hwfn, } for (i = shadow_start_line; i < shadow_end_line; i++) { - if (!p_hwfn->p_cxt_mngr->ilt_shadow[i].p_virt) + if (!p_hwfn->p_cxt_mngr->ilt_shadow[i].virt_addr) continue; dma_free_coherent(&p_hwfn->cdev->pdev->dev, p_hwfn->p_cxt_mngr->ilt_shadow[i].size, - p_hwfn->p_cxt_mngr->ilt_shadow[i].p_virt, - p_hwfn->p_cxt_mngr->ilt_shadow[i].p_phys); + p_hwfn->p_cxt_mngr->ilt_shadow[i].virt_addr, + p_hwfn->p_cxt_mngr->ilt_shadow[i].phys_addr); - p_hwfn->p_cxt_mngr->ilt_shadow[i].p_virt = NULL; - p_hwfn->p_cxt_mngr->ilt_shadow[i].p_phys = 0; + p_hwfn->p_cxt_mngr->ilt_shadow[i].virt_addr = NULL; + p_hwfn->p_cxt_mngr->ilt_shadow[i].phys_addr = 0; p_hwfn->p_cxt_mngr->ilt_shadow[i].size = 0; /* compute absolute offset */ @@ -2547,8 +2445,76 @@ int qed_cxt_get_task_ctx(struct qed_hwfn *p_hwfn, ilt_idx = tid / num_tids_per_block + p_seg->start_line - p_mngr->pf_start_line; - *pp_task_ctx = (u8 *)p_mngr->ilt_shadow[ilt_idx].p_virt + + *pp_task_ctx = (u8 *)p_mngr->ilt_shadow[ilt_idx].virt_addr + (tid % num_tids_per_block) * tid_size; return 0; } + +static u16 qed_blk_calculate_pages(struct qed_ilt_cli_blk *p_blk) +{ + if (p_blk->real_size_in_page == 0) + return 0; + + return DIV_ROUND_UP(p_blk->total_size, p_blk->real_size_in_page); +} + +u16 qed_get_cdut_num_pf_init_pages(struct qed_hwfn *p_hwfn) +{ + struct qed_ilt_client_cfg *p_cli; + struct qed_ilt_cli_blk *p_blk; + u16 i, pages = 0; + + p_cli = &p_hwfn->p_cxt_mngr->clients[ILT_CLI_CDUT]; + for (i = 0; i < NUM_TASK_PF_SEGMENTS; i++) { + p_blk = &p_cli->pf_blks[CDUT_FL_SEG_BLK(i, PF)]; + pages += qed_blk_calculate_pages(p_blk); + } + + return pages; +} + +u16 qed_get_cdut_num_vf_init_pages(struct qed_hwfn *p_hwfn) +{ + struct qed_ilt_client_cfg *p_cli; + struct qed_ilt_cli_blk *p_blk; + u16 i, pages = 0; + + p_cli = &p_hwfn->p_cxt_mngr->clients[ILT_CLI_CDUT]; + for (i = 0; i < NUM_TASK_VF_SEGMENTS; i++) { + p_blk = &p_cli->vf_blks[CDUT_FL_SEG_BLK(i, VF)]; + pages += qed_blk_calculate_pages(p_blk); + } + + return pages; +} + +u16 qed_get_cdut_num_pf_work_pages(struct qed_hwfn *p_hwfn) +{ + struct qed_ilt_client_cfg *p_cli; + struct qed_ilt_cli_blk *p_blk; + u16 i, pages = 0; + + p_cli = &p_hwfn->p_cxt_mngr->clients[ILT_CLI_CDUT]; + for (i = 0; i < NUM_TASK_PF_SEGMENTS; i++) { + p_blk = &p_cli->pf_blks[CDUT_SEG_BLK(i)]; + pages += qed_blk_calculate_pages(p_blk); + } + + return pages; +} + +u16 qed_get_cdut_num_vf_work_pages(struct qed_hwfn *p_hwfn) +{ + struct qed_ilt_client_cfg *p_cli; + struct qed_ilt_cli_blk *p_blk; + u16 pages = 0, i; + + p_cli = &p_hwfn->p_cxt_mngr->clients[ILT_CLI_CDUT]; + for (i = 0; i < NUM_TASK_VF_SEGMENTS; i++) { + p_blk = &p_cli->vf_blks[CDUT_SEG_BLK(i)]; + pages += qed_blk_calculate_pages(p_blk); + } + + return pages; +} diff --git a/drivers/net/ethernet/qlogic/qed/qed_cxt.h b/drivers/net/ethernet/qlogic/qed/qed_cxt.h index 758a8b4c0de8..c4e815f6cabd 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_cxt.h +++ b/drivers/net/ethernet/qlogic/qed/qed_cxt.h @@ -242,4 +242,134 @@ int qed_cxt_free_proto_ilt(struct qed_hwfn *p_hwfn, enum protocol_type proto); #define QED_CTX_FL_MEM 1 int qed_cxt_get_task_ctx(struct qed_hwfn *p_hwfn, u32 tid, u8 ctx_type, void **task_ctx); + +/* Max number of connection types in HW (DQ/CDU etc.) */ +#define MAX_CONN_TYPES PROTOCOLID_COMMON +#define NUM_TASK_TYPES 2 +#define NUM_TASK_PF_SEGMENTS 4 +#define NUM_TASK_VF_SEGMENTS 1 + +/* PF per protocl configuration object */ +#define TASK_SEGMENTS (NUM_TASK_PF_SEGMENTS + NUM_TASK_VF_SEGMENTS) +#define TASK_SEGMENT_VF (NUM_TASK_PF_SEGMENTS) + +struct qed_tid_seg { + u32 count; + u8 type; + bool has_fl_mem; +}; + +struct qed_conn_type_cfg { + u32 cid_count; + u32 cids_per_vf; + struct qed_tid_seg tid_seg[TASK_SEGMENTS]; +}; + +/* ILT Client configuration, + * Per connection type (protocol) resources (cids, tis, vf cids etc.) + * 1 - for connection context (CDUC) and for each task context we need two + * values, for regular task context and for force load memory + */ +#define ILT_CLI_PF_BLOCKS (1 + NUM_TASK_PF_SEGMENTS * 2) +#define ILT_CLI_VF_BLOCKS (1 + NUM_TASK_VF_SEGMENTS * 2) +#define CDUC_BLK (0) +#define SRQ_BLK (0) +#define CDUT_SEG_BLK(n) (1 + (u8)(n)) +#define CDUT_FL_SEG_BLK(n, X) (1 + (n) + NUM_TASK_ ## X ## _SEGMENTS) + +struct ilt_cfg_pair { + u32 reg; + u32 val; +}; + +struct qed_ilt_cli_blk { + u32 total_size; /* 0 means not active */ + u32 real_size_in_page; + u32 start_line; + u32 dynamic_line_offset; + u32 dynamic_line_cnt; +}; + +struct qed_ilt_client_cfg { + bool active; + + /* ILT boundaries */ + struct ilt_cfg_pair first; + struct ilt_cfg_pair last; + struct ilt_cfg_pair p_size; + + /* ILT client blocks for PF */ + struct qed_ilt_cli_blk pf_blks[ILT_CLI_PF_BLOCKS]; + u32 pf_total_lines; + + /* ILT client blocks for VFs */ + struct qed_ilt_cli_blk vf_blks[ILT_CLI_VF_BLOCKS]; + u32 vf_total_lines; +}; + +struct qed_cid_acquired_map { + u32 start_cid; + u32 max_count; + unsigned long *cid_map; +}; + +struct qed_src_t2 { + struct phys_mem_desc *dma_mem; + u32 num_pages; + u64 first_free; + u64 last_free; +}; + +struct qed_cxt_mngr { + /* Per protocl configuration */ + struct qed_conn_type_cfg conn_cfg[MAX_CONN_TYPES]; + + /* computed ILT structure */ + struct qed_ilt_client_cfg clients[MAX_ILT_CLIENTS]; + + /* Task type sizes */ + u32 task_type_size[NUM_TASK_TYPES]; + + /* total number of VFs for this hwfn - + * ALL VFs are symmetric in terms of HW resources + */ + u32 vf_count; + u32 first_vf_in_pf; + + /* Acquired CIDs */ + struct qed_cid_acquired_map acquired[MAX_CONN_TYPES]; + + struct qed_cid_acquired_map + acquired_vf[MAX_CONN_TYPES][MAX_NUM_VFS]; + + /* ILT shadow table */ + struct phys_mem_desc *ilt_shadow; + u32 ilt_shadow_size; + u32 pf_start_line; + + /* Mutex for a dynamic ILT allocation */ + struct mutex mutex; + + /* SRC T2 */ + struct qed_src_t2 src_t2; + u32 t2_num_pages; + u64 first_free; + u64 last_free; + + /* total number of SRQ's for this hwfn */ + u32 srq_count; + + /* Maximal number of L2 steering filters */ + u32 arfs_count; + + u8 task_type_id; + u16 task_ctx_size; + u16 conn_ctx_size; +}; + +u16 qed_get_cdut_num_pf_init_pages(struct qed_hwfn *p_hwfn); +u16 qed_get_cdut_num_vf_init_pages(struct qed_hwfn *p_hwfn); +u16 qed_get_cdut_num_pf_work_pages(struct qed_hwfn *p_hwfn); +u16 qed_get_cdut_num_vf_work_pages(struct qed_hwfn *p_hwfn); + #endif diff --git a/drivers/net/ethernet/qlogic/qed/qed_debug.c b/drivers/net/ethernet/qlogic/qed/qed_debug.c index 859caa6c1a1f..f4eebaabb6d0 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_debug.c +++ b/drivers/net/ethernet/qlogic/qed/qed_debug.c @@ -7,6 +7,7 @@ #include <linux/vmalloc.h> #include <linux/crc32.h> #include "qed.h" +#include "qed_cxt.h" #include "qed_hsi.h" #include "qed_hw.h" #include "qed_mcp.h" @@ -22,27 +23,28 @@ enum mem_groups { MEM_GROUP_BRB_RAM, MEM_GROUP_BRB_MEM, MEM_GROUP_PRS_MEM, + MEM_GROUP_SDM_MEM, + MEM_GROUP_PBUF, MEM_GROUP_IOR, + MEM_GROUP_RAM, MEM_GROUP_BTB_RAM, + MEM_GROUP_RDIF_CTX, + MEM_GROUP_TDIF_CTX, + MEM_GROUP_CFC_MEM, MEM_GROUP_CONN_CFC_MEM, - MEM_GROUP_TASK_CFC_MEM, MEM_GROUP_CAU_PI, MEM_GROUP_CAU_MEM, + MEM_GROUP_CAU_MEM_EXT, MEM_GROUP_PXP_ILT, - MEM_GROUP_TM_MEM, - MEM_GROUP_SDM_MEM, - MEM_GROUP_PBUF, - MEM_GROUP_RAM, MEM_GROUP_MULD_MEM, MEM_GROUP_BTB_MEM, - MEM_GROUP_RDIF_CTX, - MEM_GROUP_TDIF_CTX, - MEM_GROUP_CFC_MEM, MEM_GROUP_IGU_MEM, MEM_GROUP_IGU_MSIX, MEM_GROUP_CAU_SB, MEM_GROUP_BMB_RAM, MEM_GROUP_BMB_MEM, + MEM_GROUP_TM_MEM, + MEM_GROUP_TASK_CFC_MEM, MEM_GROUPS_NUM }; @@ -56,27 +58,28 @@ static const char * const s_mem_group_names[] = { "BRB_RAM", "BRB_MEM", "PRS_MEM", + "SDM_MEM", + "PBUF", "IOR", + "RAM", "BTB_RAM", + "RDIF_CTX", + "TDIF_CTX", + "CFC_MEM", "CONN_CFC_MEM", - "TASK_CFC_MEM", "CAU_PI", "CAU_MEM", + "CAU_MEM_EXT", "PXP_ILT", - "TM_MEM", - "SDM_MEM", - "PBUF", - "RAM", "MULD_MEM", "BTB_MEM", - "RDIF_CTX", - "TDIF_CTX", - "CFC_MEM", "IGU_MEM", "IGU_MSIX", "CAU_SB", "BMB_RAM", "BMB_MEM", + "TM_MEM", + "TASK_CFC_MEM", }; /* Idle check conditions */ @@ -170,35 +173,66 @@ static u32(*cond_arr[]) (const u32 *r, const u32 *imm) = { cond13, }; +#define NUM_PHYS_BLOCKS 84 + +#define NUM_DBG_RESET_REGS 8 + /******************************* Data Types **********************************/ -enum platform_ids { - PLATFORM_ASIC, +enum hw_types { + HW_TYPE_ASIC, PLATFORM_RESERVED, PLATFORM_RESERVED2, PLATFORM_RESERVED3, - MAX_PLATFORM_IDS + PLATFORM_RESERVED4, + MAX_HW_TYPES +}; + +/* CM context types */ +enum cm_ctx_types { + CM_CTX_CONN_AG, + CM_CTX_CONN_ST, + CM_CTX_TASK_AG, + CM_CTX_TASK_ST, + NUM_CM_CTX_TYPES +}; + +/* Debug bus frame modes */ +enum dbg_bus_frame_modes { + DBG_BUS_FRAME_MODE_4ST = 0, /* 4 Storm dwords (no HW) */ + DBG_BUS_FRAME_MODE_2ST_2HW = 1, /* 2 Storm dwords, 2 HW dwords */ + DBG_BUS_FRAME_MODE_1ST_3HW = 2, /* 1 Storm dwords, 3 HW dwords */ + DBG_BUS_FRAME_MODE_4HW = 3, /* 4 HW dwords (no Storms) */ + DBG_BUS_FRAME_MODE_8HW = 4, /* 8 HW dwords (no Storms) */ + DBG_BUS_NUM_FRAME_MODES }; /* Chip constant definitions */ struct chip_defs { const char *name; + u32 num_ilt_pages; }; -/* Platform constant definitions */ -struct platform_defs { +/* HW type constant definitions */ +struct hw_type_defs { const char *name; u32 delay_factor; u32 dmae_thresh; u32 log_thresh; }; +/* RBC reset definitions */ +struct rbc_reset_defs { + u32 reset_reg_addr; + u32 reset_val[MAX_CHIP_IDS]; +}; + /* Storm constant definitions. * Addresses are in bytes, sizes are in quad-regs. */ struct storm_defs { char letter; - enum block_id block_id; + enum block_id sem_block_id; enum dbg_bus_clients dbg_client_id[MAX_CHIP_IDS]; bool has_vfc; u32 sem_fast_mem_addr; @@ -207,47 +241,26 @@ struct storm_defs { u32 sem_slow_mode_addr; u32 sem_slow_mode1_conf_addr; u32 sem_sync_dbg_empty_addr; - u32 sem_slow_dbg_empty_addr; + u32 sem_gpre_vect_addr; u32 cm_ctx_wr_addr; - u32 cm_conn_ag_ctx_lid_size; - u32 cm_conn_ag_ctx_rd_addr; - u32 cm_conn_st_ctx_lid_size; - u32 cm_conn_st_ctx_rd_addr; - u32 cm_task_ag_ctx_lid_size; - u32 cm_task_ag_ctx_rd_addr; - u32 cm_task_st_ctx_lid_size; - u32 cm_task_st_ctx_rd_addr; + u32 cm_ctx_rd_addr[NUM_CM_CTX_TYPES]; + u32 cm_ctx_lid_sizes[MAX_CHIP_IDS][NUM_CM_CTX_TYPES]; }; -/* Block constant definitions */ -struct block_defs { - const char *name; - bool exists[MAX_CHIP_IDS]; - bool associated_to_storm; - - /* Valid only if associated_to_storm is true */ - u32 storm_id; - enum dbg_bus_clients dbg_client_id[MAX_CHIP_IDS]; - u32 dbg_select_addr; - u32 dbg_enable_addr; - u32 dbg_shift_addr; - u32 dbg_force_valid_addr; - u32 dbg_force_frame_addr; - bool has_reset_bit; - - /* If true, block is taken out of reset before dump */ - bool unreset; - enum dbg_reset_regs reset_reg; - - /* Bit offset in reset register */ - u8 reset_bit_offset; +/* Debug Bus Constraint operation constant definitions */ +struct dbg_bus_constraint_op_defs { + u8 hw_op_val; + bool is_cyclic; }; -/* Reset register definitions */ -struct reset_reg_defs { - u32 addr; +/* Storm Mode definitions */ +struct storm_mode_defs { + const char *name; + bool is_fast_dbg; + u8 id_in_hw; + u32 src_disable_reg_addr; + u32 src_enable_val; bool exists[MAX_CHIP_IDS]; - u32 unreset_val[MAX_CHIP_IDS]; }; struct grc_param_defs { @@ -257,7 +270,7 @@ struct grc_param_defs { bool is_preset; bool is_persistent; u32 exclude_all_preset_val; - u32 crash_preset_val; + u32 crash_preset_val[MAX_CHIP_IDS]; }; /* Address is in 128b units. Width is in bits. */ @@ -314,15 +327,7 @@ struct split_type_defs { /******************************** Constants **********************************/ -#define MAX_LCIDS 320 -#define MAX_LTIDS 320 - -#define NUM_IOR_SETS 2 -#define IORS_PER_SET 176 -#define IOR_SET_OFFSET(set_id) ((set_id) * 256) - #define BYTES_IN_DWORD sizeof(u32) - /* In the macros below, size and offset are specified in bits */ #define CEIL_DWORDS(size) DIV_ROUND_UP(size, 32) #define FIELD_BIT_OFFSET(type, field) type ## _ ## field ## _ ## OFFSET @@ -348,20 +353,17 @@ struct split_type_defs { qed_wr(dev, ptt, addr, (arr)[i]); \ } while (0) -#define ARR_REG_RD(dev, ptt, addr, arr, arr_size) \ - do { \ - for (i = 0; i < (arr_size); i++) \ - (arr)[i] = qed_rd(dev, ptt, addr); \ - } while (0) - #define DWORDS_TO_BYTES(dwords) ((dwords) * BYTES_IN_DWORD) #define BYTES_TO_DWORDS(bytes) ((bytes) / BYTES_IN_DWORD) -/* Extra lines include a signature line + optional latency events line */ -#define NUM_EXTRA_DBG_LINES(block_desc) \ - (1 + ((block_desc)->has_latency_events ? 1 : 0)) -#define NUM_DBG_LINES(block_desc) \ - ((block_desc)->num_of_lines + NUM_EXTRA_DBG_LINES(block_desc)) +/* extra lines include a signature line + optional latency events line */ +#define NUM_EXTRA_DBG_LINES(block) \ + (GET_FIELD((block)->flags, DBG_BLOCK_CHIP_HAS_LATENCY_EVENTS) ? 2 : 1) +#define NUM_DBG_LINES(block) \ + ((block)->num_of_dbg_bus_lines + NUM_EXTRA_DBG_LINES(block)) + +#define USE_DMAE true +#define PROTECT_WIDE_BUS true #define RAM_LINES_TO_DWORDS(lines) ((lines) * 2) #define RAM_LINES_TO_BYTES(lines) \ @@ -380,6 +382,9 @@ struct split_type_defs { #define IDLE_CHK_RESULT_REG_HDR_DWORDS \ BYTES_TO_DWORDS(sizeof(struct dbg_idle_chk_result_reg_hdr)) +#define PAGE_MEM_DESC_SIZE_DWORDS \ + BYTES_TO_DWORDS(sizeof(struct phys_mem_desc)) + #define IDLE_CHK_MAX_ENTRIES_SIZE 32 /* The sizes and offsets below are specified in bits */ @@ -425,7 +430,9 @@ struct split_type_defs { #define STATIC_DEBUG_LINE_DWORDS 9 -#define NUM_COMMON_GLOBAL_PARAMS 8 +#define NUM_COMMON_GLOBAL_PARAMS 9 + +#define MAX_RECURSION_DEPTH 10 #define FW_IMG_MAIN 1 @@ -449,1054 +456,121 @@ struct split_type_defs { (MCP_REG_SCRATCH + \ offsetof(struct static_init, sections[SPAD_SECTION_TRACE])) +#define MAX_SW_PLTAFORM_STR_SIZE 64 + #define EMPTY_FW_VERSION_STR "???_???_???_???" #define EMPTY_FW_IMAGE_STR "???????????????" /***************************** Constant Arrays *******************************/ -struct dbg_array { - const u32 *ptr; - u32 size_in_dwords; -}; - -/* Debug arrays */ -static struct dbg_array s_dbg_arrays[MAX_BIN_DBG_BUFFER_TYPE] = { {NULL} }; - /* Chip constant definitions array */ static struct chip_defs s_chip_defs[MAX_CHIP_IDS] = { - {"bb"}, - {"ah"}, - {"reserved"}, + {"bb", PSWRQ2_REG_ILT_MEMORY_SIZE_BB / 2}, + {"ah", PSWRQ2_REG_ILT_MEMORY_SIZE_K2 / 2} }; /* Storm constant definitions array */ static struct storm_defs s_storm_defs[] = { /* Tstorm */ {'T', BLOCK_TSEM, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT, - DBG_BUS_CLIENT_RBCT}, true, - TSEM_REG_FAST_MEMORY, - TSEM_REG_DBG_FRAME_MODE_BB_K2, TSEM_REG_SLOW_DBG_ACTIVE_BB_K2, - TSEM_REG_SLOW_DBG_MODE_BB_K2, TSEM_REG_DBG_MODE1_CFG_BB_K2, - TSEM_REG_SYNC_DBG_EMPTY, TSEM_REG_SLOW_DBG_EMPTY_BB_K2, - TCM_REG_CTX_RBC_ACCS, - 4, TCM_REG_AGG_CON_CTX, - 16, TCM_REG_SM_CON_CTX, - 2, TCM_REG_AGG_TASK_CTX, - 4, TCM_REG_SM_TASK_CTX}, + {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT}, + true, + TSEM_REG_FAST_MEMORY, + TSEM_REG_DBG_FRAME_MODE_BB_K2, TSEM_REG_SLOW_DBG_ACTIVE_BB_K2, + TSEM_REG_SLOW_DBG_MODE_BB_K2, TSEM_REG_DBG_MODE1_CFG_BB_K2, + TSEM_REG_SYNC_DBG_EMPTY, TSEM_REG_DBG_GPRE_VECT, + TCM_REG_CTX_RBC_ACCS, + {TCM_REG_AGG_CON_CTX, TCM_REG_SM_CON_CTX, TCM_REG_AGG_TASK_CTX, + TCM_REG_SM_TASK_CTX}, + {{4, 16, 2, 4}, {4, 16, 2, 4}} /* {bb} {k2} */ + }, /* Mstorm */ {'M', BLOCK_MSEM, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, - DBG_BUS_CLIENT_RBCM}, false, - MSEM_REG_FAST_MEMORY, - MSEM_REG_DBG_FRAME_MODE_BB_K2, MSEM_REG_SLOW_DBG_ACTIVE_BB_K2, - MSEM_REG_SLOW_DBG_MODE_BB_K2, MSEM_REG_DBG_MODE1_CFG_BB_K2, - MSEM_REG_SYNC_DBG_EMPTY, MSEM_REG_SLOW_DBG_EMPTY_BB_K2, - MCM_REG_CTX_RBC_ACCS, - 1, MCM_REG_AGG_CON_CTX, - 10, MCM_REG_SM_CON_CTX, - 2, MCM_REG_AGG_TASK_CTX, - 7, MCM_REG_SM_TASK_CTX}, + {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM}, + false, + MSEM_REG_FAST_MEMORY, + MSEM_REG_DBG_FRAME_MODE_BB_K2, + MSEM_REG_SLOW_DBG_ACTIVE_BB_K2, + MSEM_REG_SLOW_DBG_MODE_BB_K2, + MSEM_REG_DBG_MODE1_CFG_BB_K2, + MSEM_REG_SYNC_DBG_EMPTY, + MSEM_REG_DBG_GPRE_VECT, + MCM_REG_CTX_RBC_ACCS, + {MCM_REG_AGG_CON_CTX, MCM_REG_SM_CON_CTX, MCM_REG_AGG_TASK_CTX, + MCM_REG_SM_TASK_CTX }, + {{1, 10, 2, 7}, {1, 10, 2, 7}} /* {bb} {k2}*/ + }, /* Ustorm */ {'U', BLOCK_USEM, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, - DBG_BUS_CLIENT_RBCU}, false, - USEM_REG_FAST_MEMORY, - USEM_REG_DBG_FRAME_MODE_BB_K2, USEM_REG_SLOW_DBG_ACTIVE_BB_K2, - USEM_REG_SLOW_DBG_MODE_BB_K2, USEM_REG_DBG_MODE1_CFG_BB_K2, - USEM_REG_SYNC_DBG_EMPTY, USEM_REG_SLOW_DBG_EMPTY_BB_K2, - UCM_REG_CTX_RBC_ACCS, - 2, UCM_REG_AGG_CON_CTX, - 13, UCM_REG_SM_CON_CTX, - 3, UCM_REG_AGG_TASK_CTX, - 3, UCM_REG_SM_TASK_CTX}, + {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU}, + false, + USEM_REG_FAST_MEMORY, + USEM_REG_DBG_FRAME_MODE_BB_K2, + USEM_REG_SLOW_DBG_ACTIVE_BB_K2, + USEM_REG_SLOW_DBG_MODE_BB_K2, + USEM_REG_DBG_MODE1_CFG_BB_K2, + USEM_REG_SYNC_DBG_EMPTY, + USEM_REG_DBG_GPRE_VECT, + UCM_REG_CTX_RBC_ACCS, + {UCM_REG_AGG_CON_CTX, UCM_REG_SM_CON_CTX, UCM_REG_AGG_TASK_CTX, + UCM_REG_SM_TASK_CTX}, + {{2, 13, 3, 3}, {2, 13, 3, 3}} /* {bb} {k2} */ + }, /* Xstorm */ {'X', BLOCK_XSEM, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX, - DBG_BUS_CLIENT_RBCX}, false, - XSEM_REG_FAST_MEMORY, - XSEM_REG_DBG_FRAME_MODE_BB_K2, XSEM_REG_SLOW_DBG_ACTIVE_BB_K2, - XSEM_REG_SLOW_DBG_MODE_BB_K2, XSEM_REG_DBG_MODE1_CFG_BB_K2, - XSEM_REG_SYNC_DBG_EMPTY, XSEM_REG_SLOW_DBG_EMPTY_BB_K2, - XCM_REG_CTX_RBC_ACCS, - 9, XCM_REG_AGG_CON_CTX, - 15, XCM_REG_SM_CON_CTX, - 0, 0, - 0, 0}, + {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX}, + false, + XSEM_REG_FAST_MEMORY, + XSEM_REG_DBG_FRAME_MODE_BB_K2, + XSEM_REG_SLOW_DBG_ACTIVE_BB_K2, + XSEM_REG_SLOW_DBG_MODE_BB_K2, + XSEM_REG_DBG_MODE1_CFG_BB_K2, + XSEM_REG_SYNC_DBG_EMPTY, + XSEM_REG_DBG_GPRE_VECT, + XCM_REG_CTX_RBC_ACCS, + {XCM_REG_AGG_CON_CTX, XCM_REG_SM_CON_CTX, 0, 0}, + {{9, 15, 0, 0}, {9, 15, 0, 0}} /* {bb} {k2} */ + }, /* Ystorm */ {'Y', BLOCK_YSEM, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY, - DBG_BUS_CLIENT_RBCY}, false, - YSEM_REG_FAST_MEMORY, - YSEM_REG_DBG_FRAME_MODE_BB_K2, YSEM_REG_SLOW_DBG_ACTIVE_BB_K2, - YSEM_REG_SLOW_DBG_MODE_BB_K2, YSEM_REG_DBG_MODE1_CFG_BB_K2, - YSEM_REG_SYNC_DBG_EMPTY, TSEM_REG_SLOW_DBG_EMPTY_BB_K2, - YCM_REG_CTX_RBC_ACCS, - 2, YCM_REG_AGG_CON_CTX, - 3, YCM_REG_SM_CON_CTX, - 2, YCM_REG_AGG_TASK_CTX, - 12, YCM_REG_SM_TASK_CTX}, + {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY}, + false, + YSEM_REG_FAST_MEMORY, + YSEM_REG_DBG_FRAME_MODE_BB_K2, + YSEM_REG_SLOW_DBG_ACTIVE_BB_K2, + YSEM_REG_SLOW_DBG_MODE_BB_K2, + YSEM_REG_DBG_MODE1_CFG_BB_K2, + YSEM_REG_SYNC_DBG_EMPTY, + YSEM_REG_DBG_GPRE_VECT, + YCM_REG_CTX_RBC_ACCS, + {YCM_REG_AGG_CON_CTX, YCM_REG_SM_CON_CTX, YCM_REG_AGG_TASK_CTX, + YCM_REG_SM_TASK_CTX}, + {{2, 3, 2, 12}, {2, 3, 2, 12}} /* {bb} {k2} */ + }, /* Pstorm */ {'P', BLOCK_PSEM, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, - DBG_BUS_CLIENT_RBCS}, true, - PSEM_REG_FAST_MEMORY, - PSEM_REG_DBG_FRAME_MODE_BB_K2, PSEM_REG_SLOW_DBG_ACTIVE_BB_K2, - PSEM_REG_SLOW_DBG_MODE_BB_K2, PSEM_REG_DBG_MODE1_CFG_BB_K2, - PSEM_REG_SYNC_DBG_EMPTY, PSEM_REG_SLOW_DBG_EMPTY_BB_K2, - PCM_REG_CTX_RBC_ACCS, - 0, 0, - 10, PCM_REG_SM_CON_CTX, - 0, 0, - 0, 0} -}; - -/* Block definitions array */ - -static struct block_defs block_grc_defs = { - "grc", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCN, DBG_BUS_CLIENT_RBCN, DBG_BUS_CLIENT_RBCN}, - GRC_REG_DBG_SELECT, GRC_REG_DBG_DWORD_ENABLE, - GRC_REG_DBG_SHIFT, GRC_REG_DBG_FORCE_VALID, - GRC_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_UA, 1 -}; - -static struct block_defs block_miscs_defs = { - "miscs", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_misc_defs = { - "misc", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_dbu_defs = { - "dbu", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_pglue_b_defs = { - "pglue_b", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCH, DBG_BUS_CLIENT_RBCH, DBG_BUS_CLIENT_RBCH}, - PGLUE_B_REG_DBG_SELECT, PGLUE_B_REG_DBG_DWORD_ENABLE, - PGLUE_B_REG_DBG_SHIFT, PGLUE_B_REG_DBG_FORCE_VALID, - PGLUE_B_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISCS_PL_HV, 1 -}; - -static struct block_defs block_cnig_defs = { - "cnig", - {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCW, - DBG_BUS_CLIENT_RBCW}, - CNIG_REG_DBG_SELECT_K2_E5, CNIG_REG_DBG_DWORD_ENABLE_K2_E5, - CNIG_REG_DBG_SHIFT_K2_E5, CNIG_REG_DBG_FORCE_VALID_K2_E5, - CNIG_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV, 0 -}; - -static struct block_defs block_cpmu_defs = { - "cpmu", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISCS_PL_HV, 8 -}; - -static struct block_defs block_ncsi_defs = { - "ncsi", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ}, - NCSI_REG_DBG_SELECT, NCSI_REG_DBG_DWORD_ENABLE, - NCSI_REG_DBG_SHIFT, NCSI_REG_DBG_FORCE_VALID, - NCSI_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISCS_PL_HV, 5 -}; - -static struct block_defs block_opte_defs = { - "opte", {true, true, false}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISCS_PL_HV, 4 -}; - -static struct block_defs block_bmb_defs = { - "bmb", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCB, DBG_BUS_CLIENT_RBCB}, - BMB_REG_DBG_SELECT, BMB_REG_DBG_DWORD_ENABLE, - BMB_REG_DBG_SHIFT, BMB_REG_DBG_FORCE_VALID, - BMB_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISCS_PL_UA, 7 -}; - -static struct block_defs block_pcie_defs = { - "pcie", - {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCH, - DBG_BUS_CLIENT_RBCH}, - PCIE_REG_DBG_COMMON_SELECT_K2_E5, - PCIE_REG_DBG_COMMON_DWORD_ENABLE_K2_E5, - PCIE_REG_DBG_COMMON_SHIFT_K2_E5, - PCIE_REG_DBG_COMMON_FORCE_VALID_K2_E5, - PCIE_REG_DBG_COMMON_FORCE_FRAME_K2_E5, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_mcp_defs = { - "mcp", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_mcp2_defs = { - "mcp2", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ}, - MCP2_REG_DBG_SELECT, MCP2_REG_DBG_DWORD_ENABLE, - MCP2_REG_DBG_SHIFT, MCP2_REG_DBG_FORCE_VALID, - MCP2_REG_DBG_FORCE_FRAME, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_pswhst_defs = { - "pswhst", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWHST_REG_DBG_SELECT, PSWHST_REG_DBG_DWORD_ENABLE, - PSWHST_REG_DBG_SHIFT, PSWHST_REG_DBG_FORCE_VALID, - PSWHST_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 0 -}; - -static struct block_defs block_pswhst2_defs = { - "pswhst2", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWHST2_REG_DBG_SELECT, PSWHST2_REG_DBG_DWORD_ENABLE, - PSWHST2_REG_DBG_SHIFT, PSWHST2_REG_DBG_FORCE_VALID, - PSWHST2_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 0 -}; - -static struct block_defs block_pswrd_defs = { - "pswrd", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWRD_REG_DBG_SELECT, PSWRD_REG_DBG_DWORD_ENABLE, - PSWRD_REG_DBG_SHIFT, PSWRD_REG_DBG_FORCE_VALID, - PSWRD_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 2 -}; - -static struct block_defs block_pswrd2_defs = { - "pswrd2", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWRD2_REG_DBG_SELECT, PSWRD2_REG_DBG_DWORD_ENABLE, - PSWRD2_REG_DBG_SHIFT, PSWRD2_REG_DBG_FORCE_VALID, - PSWRD2_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 2 -}; - -static struct block_defs block_pswwr_defs = { - "pswwr", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWWR_REG_DBG_SELECT, PSWWR_REG_DBG_DWORD_ENABLE, - PSWWR_REG_DBG_SHIFT, PSWWR_REG_DBG_FORCE_VALID, - PSWWR_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 3 -}; - -static struct block_defs block_pswwr2_defs = { - "pswwr2", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISC_PL_HV, 3 -}; - -static struct block_defs block_pswrq_defs = { - "pswrq", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWRQ_REG_DBG_SELECT, PSWRQ_REG_DBG_DWORD_ENABLE, - PSWRQ_REG_DBG_SHIFT, PSWRQ_REG_DBG_FORCE_VALID, - PSWRQ_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 1 -}; - -static struct block_defs block_pswrq2_defs = { - "pswrq2", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PSWRQ2_REG_DBG_SELECT, PSWRQ2_REG_DBG_DWORD_ENABLE, - PSWRQ2_REG_DBG_SHIFT, PSWRQ2_REG_DBG_FORCE_VALID, - PSWRQ2_REG_DBG_FORCE_FRAME, - true, false, DBG_RESET_REG_MISC_PL_HV, 1 -}; - -static struct block_defs block_pglcs_defs = { - "pglcs", - {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCH, - DBG_BUS_CLIENT_RBCH}, - PGLCS_REG_DBG_SELECT_K2_E5, PGLCS_REG_DBG_DWORD_ENABLE_K2_E5, - PGLCS_REG_DBG_SHIFT_K2_E5, PGLCS_REG_DBG_FORCE_VALID_K2_E5, - PGLCS_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV, 2 -}; - -static struct block_defs block_ptu_defs = { - "ptu", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - PTU_REG_DBG_SELECT, PTU_REG_DBG_DWORD_ENABLE, - PTU_REG_DBG_SHIFT, PTU_REG_DBG_FORCE_VALID, - PTU_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 20 -}; - -static struct block_defs block_dmae_defs = { - "dmae", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - DMAE_REG_DBG_SELECT, DMAE_REG_DBG_DWORD_ENABLE, - DMAE_REG_DBG_SHIFT, DMAE_REG_DBG_FORCE_VALID, - DMAE_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 28 -}; - -static struct block_defs block_tcm_defs = { - "tcm", - {true, true, true}, true, DBG_TSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT}, - TCM_REG_DBG_SELECT, TCM_REG_DBG_DWORD_ENABLE, - TCM_REG_DBG_SHIFT, TCM_REG_DBG_FORCE_VALID, - TCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 5 -}; - -static struct block_defs block_mcm_defs = { - "mcm", - {true, true, true}, true, DBG_MSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - MCM_REG_DBG_SELECT, MCM_REG_DBG_DWORD_ENABLE, - MCM_REG_DBG_SHIFT, MCM_REG_DBG_FORCE_VALID, - MCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 3 -}; - -static struct block_defs block_ucm_defs = { - "ucm", - {true, true, true}, true, DBG_USTORM_ID, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU}, - UCM_REG_DBG_SELECT, UCM_REG_DBG_DWORD_ENABLE, - UCM_REG_DBG_SHIFT, UCM_REG_DBG_FORCE_VALID, - UCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 8 -}; - -static struct block_defs block_xcm_defs = { - "xcm", - {true, true, true}, true, DBG_XSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX}, - XCM_REG_DBG_SELECT, XCM_REG_DBG_DWORD_ENABLE, - XCM_REG_DBG_SHIFT, XCM_REG_DBG_FORCE_VALID, - XCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 19 -}; - -static struct block_defs block_ycm_defs = { - "ycm", - {true, true, true}, true, DBG_YSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY, DBG_BUS_CLIENT_RBCY}, - YCM_REG_DBG_SELECT, YCM_REG_DBG_DWORD_ENABLE, - YCM_REG_DBG_SHIFT, YCM_REG_DBG_FORCE_VALID, - YCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 5 -}; - -static struct block_defs block_pcm_defs = { - "pcm", - {true, true, true}, true, DBG_PSTORM_ID, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, - PCM_REG_DBG_SELECT, PCM_REG_DBG_DWORD_ENABLE, - PCM_REG_DBG_SHIFT, PCM_REG_DBG_FORCE_VALID, - PCM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 4 -}; - -static struct block_defs block_qm_defs = { - "qm", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCQ, DBG_BUS_CLIENT_RBCQ}, - QM_REG_DBG_SELECT, QM_REG_DBG_DWORD_ENABLE, - QM_REG_DBG_SHIFT, QM_REG_DBG_FORCE_VALID, - QM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 16 -}; - -static struct block_defs block_tm_defs = { - "tm", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, - TM_REG_DBG_SELECT, TM_REG_DBG_DWORD_ENABLE, - TM_REG_DBG_SHIFT, TM_REG_DBG_FORCE_VALID, - TM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 17 -}; - -static struct block_defs block_dorq_defs = { - "dorq", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY, DBG_BUS_CLIENT_RBCY}, - DORQ_REG_DBG_SELECT, DORQ_REG_DBG_DWORD_ENABLE, - DORQ_REG_DBG_SHIFT, DORQ_REG_DBG_FORCE_VALID, - DORQ_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 18 -}; - -static struct block_defs block_brb_defs = { - "brb", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCR, DBG_BUS_CLIENT_RBCR, DBG_BUS_CLIENT_RBCR}, - BRB_REG_DBG_SELECT, BRB_REG_DBG_DWORD_ENABLE, - BRB_REG_DBG_SHIFT, BRB_REG_DBG_FORCE_VALID, - BRB_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 0 -}; - -static struct block_defs block_src_defs = { - "src", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF}, - SRC_REG_DBG_SELECT, SRC_REG_DBG_DWORD_ENABLE, - SRC_REG_DBG_SHIFT, SRC_REG_DBG_FORCE_VALID, - SRC_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 2 -}; - -static struct block_defs block_prs_defs = { - "prs", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCR, DBG_BUS_CLIENT_RBCR, DBG_BUS_CLIENT_RBCR}, - PRS_REG_DBG_SELECT, PRS_REG_DBG_DWORD_ENABLE, - PRS_REG_DBG_SHIFT, PRS_REG_DBG_FORCE_VALID, - PRS_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 1 -}; - -static struct block_defs block_tsdm_defs = { - "tsdm", - {true, true, true}, true, DBG_TSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT}, - TSDM_REG_DBG_SELECT, TSDM_REG_DBG_DWORD_ENABLE, - TSDM_REG_DBG_SHIFT, TSDM_REG_DBG_FORCE_VALID, - TSDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 3 -}; - -static struct block_defs block_msdm_defs = { - "msdm", - {true, true, true}, true, DBG_MSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - MSDM_REG_DBG_SELECT, MSDM_REG_DBG_DWORD_ENABLE, - MSDM_REG_DBG_SHIFT, MSDM_REG_DBG_FORCE_VALID, - MSDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 6 -}; - -static struct block_defs block_usdm_defs = { - "usdm", - {true, true, true}, true, DBG_USTORM_ID, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU}, - USDM_REG_DBG_SELECT, USDM_REG_DBG_DWORD_ENABLE, - USDM_REG_DBG_SHIFT, USDM_REG_DBG_FORCE_VALID, - USDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 7 -}; - -static struct block_defs block_xsdm_defs = { - "xsdm", - {true, true, true}, true, DBG_XSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX}, - XSDM_REG_DBG_SELECT, XSDM_REG_DBG_DWORD_ENABLE, - XSDM_REG_DBG_SHIFT, XSDM_REG_DBG_FORCE_VALID, - XSDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 20 -}; - -static struct block_defs block_ysdm_defs = { - "ysdm", - {true, true, true}, true, DBG_YSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY, DBG_BUS_CLIENT_RBCY}, - YSDM_REG_DBG_SELECT, YSDM_REG_DBG_DWORD_ENABLE, - YSDM_REG_DBG_SHIFT, YSDM_REG_DBG_FORCE_VALID, - YSDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 8 -}; - -static struct block_defs block_psdm_defs = { - "psdm", - {true, true, true}, true, DBG_PSTORM_ID, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, - PSDM_REG_DBG_SELECT, PSDM_REG_DBG_DWORD_ENABLE, - PSDM_REG_DBG_SHIFT, PSDM_REG_DBG_FORCE_VALID, - PSDM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 7 -}; - -static struct block_defs block_tsem_defs = { - "tsem", - {true, true, true}, true, DBG_TSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT}, - TSEM_REG_DBG_SELECT, TSEM_REG_DBG_DWORD_ENABLE, - TSEM_REG_DBG_SHIFT, TSEM_REG_DBG_FORCE_VALID, - TSEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 4 -}; - -static struct block_defs block_msem_defs = { - "msem", - {true, true, true}, true, DBG_MSTORM_ID, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - MSEM_REG_DBG_SELECT, MSEM_REG_DBG_DWORD_ENABLE, - MSEM_REG_DBG_SHIFT, MSEM_REG_DBG_FORCE_VALID, - MSEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 9 -}; - -static struct block_defs block_usem_defs = { - "usem", - {true, true, true}, true, DBG_USTORM_ID, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU}, - USEM_REG_DBG_SELECT, USEM_REG_DBG_DWORD_ENABLE, - USEM_REG_DBG_SHIFT, USEM_REG_DBG_FORCE_VALID, - USEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 9 -}; - -static struct block_defs block_xsem_defs = { - "xsem", - {true, true, true}, true, DBG_XSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX}, - XSEM_REG_DBG_SELECT, XSEM_REG_DBG_DWORD_ENABLE, - XSEM_REG_DBG_SHIFT, XSEM_REG_DBG_FORCE_VALID, - XSEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 21 -}; - -static struct block_defs block_ysem_defs = { - "ysem", - {true, true, true}, true, DBG_YSTORM_ID, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCY, DBG_BUS_CLIENT_RBCY}, - YSEM_REG_DBG_SELECT, YSEM_REG_DBG_DWORD_ENABLE, - YSEM_REG_DBG_SHIFT, YSEM_REG_DBG_FORCE_VALID, - YSEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 11 -}; - -static struct block_defs block_psem_defs = { - "psem", - {true, true, true}, true, DBG_PSTORM_ID, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, - PSEM_REG_DBG_SELECT, PSEM_REG_DBG_DWORD_ENABLE, - PSEM_REG_DBG_SHIFT, PSEM_REG_DBG_FORCE_VALID, - PSEM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 10 -}; - -static struct block_defs block_rss_defs = { - "rss", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCT}, - RSS_REG_DBG_SELECT, RSS_REG_DBG_DWORD_ENABLE, - RSS_REG_DBG_SHIFT, RSS_REG_DBG_FORCE_VALID, - RSS_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 18 -}; - -static struct block_defs block_tmld_defs = { - "tmld", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - TMLD_REG_DBG_SELECT, TMLD_REG_DBG_DWORD_ENABLE, - TMLD_REG_DBG_SHIFT, TMLD_REG_DBG_FORCE_VALID, - TMLD_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 13 -}; - -static struct block_defs block_muld_defs = { - "muld", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU}, - MULD_REG_DBG_SELECT, MULD_REG_DBG_DWORD_ENABLE, - MULD_REG_DBG_SHIFT, MULD_REG_DBG_FORCE_VALID, - MULD_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 14 -}; - -static struct block_defs block_yuld_defs = { - "yuld", - {true, true, false}, false, 0, - {DBG_BUS_CLIENT_RBCU, DBG_BUS_CLIENT_RBCU, - MAX_DBG_BUS_CLIENTS}, - YULD_REG_DBG_SELECT_BB_K2, YULD_REG_DBG_DWORD_ENABLE_BB_K2, - YULD_REG_DBG_SHIFT_BB_K2, YULD_REG_DBG_FORCE_VALID_BB_K2, - YULD_REG_DBG_FORCE_FRAME_BB_K2, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, - 15 -}; - -static struct block_defs block_xyld_defs = { - "xyld", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX, DBG_BUS_CLIENT_RBCX}, - XYLD_REG_DBG_SELECT, XYLD_REG_DBG_DWORD_ENABLE, - XYLD_REG_DBG_SHIFT, XYLD_REG_DBG_FORCE_VALID, - XYLD_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 12 -}; - -static struct block_defs block_ptld_defs = { - "ptld", - {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCT}, - PTLD_REG_DBG_SELECT_E5, PTLD_REG_DBG_DWORD_ENABLE_E5, - PTLD_REG_DBG_SHIFT_E5, PTLD_REG_DBG_FORCE_VALID_E5, - PTLD_REG_DBG_FORCE_FRAME_E5, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, - 28 -}; - -static struct block_defs block_ypld_defs = { - "ypld", - {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCS}, - YPLD_REG_DBG_SELECT_E5, YPLD_REG_DBG_DWORD_ENABLE_E5, - YPLD_REG_DBG_SHIFT_E5, YPLD_REG_DBG_FORCE_VALID_E5, - YPLD_REG_DBG_FORCE_FRAME_E5, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, - 27 -}; - -static struct block_defs block_prm_defs = { - "prm", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - PRM_REG_DBG_SELECT, PRM_REG_DBG_DWORD_ENABLE, - PRM_REG_DBG_SHIFT, PRM_REG_DBG_FORCE_VALID, - PRM_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 21 -}; - -static struct block_defs block_pbf_pb1_defs = { - "pbf_pb1", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCV, DBG_BUS_CLIENT_RBCV}, - PBF_PB1_REG_DBG_SELECT, PBF_PB1_REG_DBG_DWORD_ENABLE, - PBF_PB1_REG_DBG_SHIFT, PBF_PB1_REG_DBG_FORCE_VALID, - PBF_PB1_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, - 11 -}; - -static struct block_defs block_pbf_pb2_defs = { - "pbf_pb2", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCV, DBG_BUS_CLIENT_RBCV}, - PBF_PB2_REG_DBG_SELECT, PBF_PB2_REG_DBG_DWORD_ENABLE, - PBF_PB2_REG_DBG_SHIFT, PBF_PB2_REG_DBG_FORCE_VALID, - PBF_PB2_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, - 12 -}; - -static struct block_defs block_rpb_defs = { - "rpb", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - RPB_REG_DBG_SELECT, RPB_REG_DBG_DWORD_ENABLE, - RPB_REG_DBG_SHIFT, RPB_REG_DBG_FORCE_VALID, - RPB_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 13 -}; - -static struct block_defs block_btb_defs = { - "btb", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCR, DBG_BUS_CLIENT_RBCV, DBG_BUS_CLIENT_RBCV}, - BTB_REG_DBG_SELECT, BTB_REG_DBG_DWORD_ENABLE, - BTB_REG_DBG_SHIFT, BTB_REG_DBG_FORCE_VALID, - BTB_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 10 -}; - -static struct block_defs block_pbf_defs = { - "pbf", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCV, DBG_BUS_CLIENT_RBCV}, - PBF_REG_DBG_SELECT, PBF_REG_DBG_DWORD_ENABLE, - PBF_REG_DBG_SHIFT, PBF_REG_DBG_FORCE_VALID, - PBF_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 15 -}; - -static struct block_defs block_rdif_defs = { - "rdif", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCT, DBG_BUS_CLIENT_RBCM, DBG_BUS_CLIENT_RBCM}, - RDIF_REG_DBG_SELECT, RDIF_REG_DBG_DWORD_ENABLE, - RDIF_REG_DBG_SHIFT, RDIF_REG_DBG_FORCE_VALID, - RDIF_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 16 -}; - -static struct block_defs block_tdif_defs = { - "tdif", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, - TDIF_REG_DBG_SELECT, TDIF_REG_DBG_DWORD_ENABLE, - TDIF_REG_DBG_SHIFT, TDIF_REG_DBG_FORCE_VALID, - TDIF_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 17 -}; - -static struct block_defs block_cdu_defs = { - "cdu", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF}, - CDU_REG_DBG_SELECT, CDU_REG_DBG_DWORD_ENABLE, - CDU_REG_DBG_SHIFT, CDU_REG_DBG_FORCE_VALID, - CDU_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 23 -}; - -static struct block_defs block_ccfc_defs = { - "ccfc", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF}, - CCFC_REG_DBG_SELECT, CCFC_REG_DBG_DWORD_ENABLE, - CCFC_REG_DBG_SHIFT, CCFC_REG_DBG_FORCE_VALID, - CCFC_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 24 -}; - -static struct block_defs block_tcfc_defs = { - "tcfc", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF, DBG_BUS_CLIENT_RBCF}, - TCFC_REG_DBG_SELECT, TCFC_REG_DBG_DWORD_ENABLE, - TCFC_REG_DBG_SHIFT, TCFC_REG_DBG_FORCE_VALID, - TCFC_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 25 -}; - -static struct block_defs block_igu_defs = { - "igu", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - IGU_REG_DBG_SELECT, IGU_REG_DBG_DWORD_ENABLE, - IGU_REG_DBG_SHIFT, IGU_REG_DBG_FORCE_VALID, - IGU_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, 27 -}; - -static struct block_defs block_cau_defs = { - "cau", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP, DBG_BUS_CLIENT_RBCP}, - CAU_REG_DBG_SELECT, CAU_REG_DBG_DWORD_ENABLE, - CAU_REG_DBG_SHIFT, CAU_REG_DBG_FORCE_VALID, - CAU_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 19 -}; - -static struct block_defs block_rgfs_defs = { - "rgfs", {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 29 -}; - -static struct block_defs block_rgsrc_defs = { - "rgsrc", - {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCH}, - RGSRC_REG_DBG_SELECT_E5, RGSRC_REG_DBG_DWORD_ENABLE_E5, - RGSRC_REG_DBG_SHIFT_E5, RGSRC_REG_DBG_FORCE_VALID_E5, - RGSRC_REG_DBG_FORCE_FRAME_E5, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, - 30 -}; - -static struct block_defs block_tgfs_defs = { - "tgfs", {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, 30 -}; - -static struct block_defs block_tgsrc_defs = { - "tgsrc", - {false, false, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCV}, - TGSRC_REG_DBG_SELECT_E5, TGSRC_REG_DBG_DWORD_ENABLE_E5, - TGSRC_REG_DBG_SHIFT_E5, TGSRC_REG_DBG_FORCE_VALID_E5, - TGSRC_REG_DBG_FORCE_FRAME_E5, - true, true, DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, - 31 -}; - -static struct block_defs block_umac_defs = { - "umac", - {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCZ, - DBG_BUS_CLIENT_RBCZ}, - UMAC_REG_DBG_SELECT_K2_E5, UMAC_REG_DBG_DWORD_ENABLE_K2_E5, - UMAC_REG_DBG_SHIFT_K2_E5, UMAC_REG_DBG_FORCE_VALID_K2_E5, - UMAC_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV, 6 -}; - -static struct block_defs block_xmac_defs = { - "xmac", {true, false, false}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_dbg_defs = { - "dbg", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, true, DBG_RESET_REG_MISC_PL_PDA_VAUX, 3 -}; - -static struct block_defs block_nig_defs = { - "nig", - {true, true, true}, false, 0, - {DBG_BUS_CLIENT_RBCN, DBG_BUS_CLIENT_RBCN, DBG_BUS_CLIENT_RBCN}, - NIG_REG_DBG_SELECT, NIG_REG_DBG_DWORD_ENABLE, - NIG_REG_DBG_SHIFT, NIG_REG_DBG_FORCE_VALID, - NIG_REG_DBG_FORCE_FRAME, - true, true, DBG_RESET_REG_MISC_PL_PDA_VAUX, 0 -}; - -static struct block_defs block_wol_defs = { - "wol", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ}, - WOL_REG_DBG_SELECT_K2_E5, WOL_REG_DBG_DWORD_ENABLE_K2_E5, - WOL_REG_DBG_SHIFT_K2_E5, WOL_REG_DBG_FORCE_VALID_K2_E5, - WOL_REG_DBG_FORCE_FRAME_K2_E5, - true, true, DBG_RESET_REG_MISC_PL_PDA_VAUX, 7 -}; - -static struct block_defs block_bmbn_defs = { - "bmbn", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCB, - DBG_BUS_CLIENT_RBCB}, - BMBN_REG_DBG_SELECT_K2_E5, BMBN_REG_DBG_DWORD_ENABLE_K2_E5, - BMBN_REG_DBG_SHIFT_K2_E5, BMBN_REG_DBG_FORCE_VALID_K2_E5, - BMBN_REG_DBG_FORCE_FRAME_K2_E5, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_ipc_defs = { - "ipc", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISCS_PL_UA, 8 -}; - -static struct block_defs block_nwm_defs = { - "nwm", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCW, DBG_BUS_CLIENT_RBCW}, - NWM_REG_DBG_SELECT_K2_E5, NWM_REG_DBG_DWORD_ENABLE_K2_E5, - NWM_REG_DBG_SHIFT_K2_E5, NWM_REG_DBG_FORCE_VALID_K2_E5, - NWM_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV_2, 0 -}; - -static struct block_defs block_nws_defs = { - "nws", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCW, DBG_BUS_CLIENT_RBCW}, - NWS_REG_DBG_SELECT_K2_E5, NWS_REG_DBG_DWORD_ENABLE_K2_E5, - NWS_REG_DBG_SHIFT_K2_E5, NWS_REG_DBG_FORCE_VALID_K2_E5, - NWS_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV, 12 -}; - -static struct block_defs block_ms_defs = { - "ms", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCZ, DBG_BUS_CLIENT_RBCZ}, - MS_REG_DBG_SELECT_K2_E5, MS_REG_DBG_DWORD_ENABLE_K2_E5, - MS_REG_DBG_SHIFT_K2_E5, MS_REG_DBG_FORCE_VALID_K2_E5, - MS_REG_DBG_FORCE_FRAME_K2_E5, - true, false, DBG_RESET_REG_MISCS_PL_HV, 13 -}; - -static struct block_defs block_phy_pcie_defs = { - "phy_pcie", - {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, DBG_BUS_CLIENT_RBCH, - DBG_BUS_CLIENT_RBCH}, - PCIE_REG_DBG_COMMON_SELECT_K2_E5, - PCIE_REG_DBG_COMMON_DWORD_ENABLE_K2_E5, - PCIE_REG_DBG_COMMON_SHIFT_K2_E5, - PCIE_REG_DBG_COMMON_FORCE_VALID_K2_E5, - PCIE_REG_DBG_COMMON_FORCE_FRAME_K2_E5, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_led_defs = { - "led", {false, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISCS_PL_HV, 14 -}; - -static struct block_defs block_avs_wrap_defs = { - "avs_wrap", {false, true, false}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - true, false, DBG_RESET_REG_MISCS_PL_UA, 11 -}; - -static struct block_defs block_pxpreqbus_defs = { - "pxpreqbus", {false, false, false}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_misc_aeu_defs = { - "misc_aeu", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs block_bar0_map_defs = { - "bar0_map", {true, true, true}, false, 0, - {MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS, MAX_DBG_BUS_CLIENTS}, - 0, 0, 0, 0, 0, - false, false, MAX_DBG_RESET_REGS, 0 -}; - -static struct block_defs *s_block_defs[MAX_BLOCK_ID] = { - &block_grc_defs, - &block_miscs_defs, - &block_misc_defs, - &block_dbu_defs, - &block_pglue_b_defs, - &block_cnig_defs, - &block_cpmu_defs, - &block_ncsi_defs, - &block_opte_defs, - &block_bmb_defs, - &block_pcie_defs, - &block_mcp_defs, - &block_mcp2_defs, - &block_pswhst_defs, - &block_pswhst2_defs, - &block_pswrd_defs, - &block_pswrd2_defs, - &block_pswwr_defs, - &block_pswwr2_defs, - &block_pswrq_defs, - &block_pswrq2_defs, - &block_pglcs_defs, - &block_dmae_defs, - &block_ptu_defs, - &block_tcm_defs, - &block_mcm_defs, - &block_ucm_defs, - &block_xcm_defs, - &block_ycm_defs, - &block_pcm_defs, - &block_qm_defs, - &block_tm_defs, - &block_dorq_defs, - &block_brb_defs, - &block_src_defs, - &block_prs_defs, - &block_tsdm_defs, - &block_msdm_defs, - &block_usdm_defs, - &block_xsdm_defs, - &block_ysdm_defs, - &block_psdm_defs, - &block_tsem_defs, - &block_msem_defs, - &block_usem_defs, - &block_xsem_defs, - &block_ysem_defs, - &block_psem_defs, - &block_rss_defs, - &block_tmld_defs, - &block_muld_defs, - &block_yuld_defs, - &block_xyld_defs, - &block_ptld_defs, - &block_ypld_defs, - &block_prm_defs, - &block_pbf_pb1_defs, - &block_pbf_pb2_defs, - &block_rpb_defs, - &block_btb_defs, - &block_pbf_defs, - &block_rdif_defs, - &block_tdif_defs, - &block_cdu_defs, - &block_ccfc_defs, - &block_tcfc_defs, - &block_igu_defs, - &block_cau_defs, - &block_rgfs_defs, - &block_rgsrc_defs, - &block_tgfs_defs, - &block_tgsrc_defs, - &block_umac_defs, - &block_xmac_defs, - &block_dbg_defs, - &block_nig_defs, - &block_wol_defs, - &block_bmbn_defs, - &block_ipc_defs, - &block_nwm_defs, - &block_nws_defs, - &block_ms_defs, - &block_phy_pcie_defs, - &block_led_defs, - &block_avs_wrap_defs, - &block_pxpreqbus_defs, - &block_misc_aeu_defs, - &block_bar0_map_defs, -}; - -static struct platform_defs s_platform_defs[] = { + {DBG_BUS_CLIENT_RBCS, DBG_BUS_CLIENT_RBCS}, + true, + PSEM_REG_FAST_MEMORY, + PSEM_REG_DBG_FRAME_MODE_BB_K2, + PSEM_REG_SLOW_DBG_ACTIVE_BB_K2, + PSEM_REG_SLOW_DBG_MODE_BB_K2, + PSEM_REG_DBG_MODE1_CFG_BB_K2, + PSEM_REG_SYNC_DBG_EMPTY, + PSEM_REG_DBG_GPRE_VECT, + PCM_REG_CTX_RBC_ACCS, + {0, PCM_REG_SM_CON_CTX, 0, 0}, + {{0, 10, 0, 0}, {0, 10, 0, 0}} /* {bb} {k2} */ + }, +}; + +static struct hw_type_defs s_hw_type_defs[] = { + /* HW_TYPE_ASIC */ {"asic", 1, 256, 32768}, {"reserved", 0, 0, 0}, {"reserved2", 0, 0, 0}, @@ -1505,146 +579,159 @@ static struct platform_defs s_platform_defs[] = { static struct grc_param_defs s_grc_param_defs[] = { /* DBG_GRC_PARAM_DUMP_TSTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_MSTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_USTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_XSTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_YSTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_PSTORM */ - {{1, 1, 1}, 0, 1, false, false, 1, 1}, + {{1, 1}, 0, 1, false, false, 1, {1, 1}}, /* DBG_GRC_PARAM_DUMP_REGS */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_RAM */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_PBUF */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_IOR */ - {{0, 0, 0}, 0, 1, false, false, 0, 1}, + {{0, 0}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_VFC */ - {{0, 0, 0}, 0, 1, false, false, 0, 1}, + {{0, 0}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_CM_CTX */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_ILT */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_RSS */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_CAU */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_QM */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_MCP */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, - /* DBG_GRC_PARAM_MCP_TRACE_META_SIZE */ - {{1, 1, 1}, 1, 0xffffffff, false, true, 0, 1}, + /* DBG_GRC_PARAM_DUMP_DORQ */ + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_CFC */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_IGU */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_BRB */ - {{0, 0, 0}, 0, 1, false, false, 0, 1}, + {{0, 0}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_BTB */ - {{0, 0, 0}, 0, 1, false, false, 0, 1}, + {{0, 0}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_BMB */ - {{0, 0, 0}, 0, 1, false, false, 0, 0}, + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, - /* DBG_GRC_PARAM_DUMP_NIG */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + /* DBG_GRC_PARAM_RESERVED1 */ + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, /* DBG_GRC_PARAM_DUMP_MULD */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_PRS */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_DMAE */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_TM */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_SDM */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_DIF */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_STATIC */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_UNSTALL */ - {{0, 0, 0}, 0, 1, false, false, 0, 0}, + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, - /* DBG_GRC_PARAM_NUM_LCIDS */ - {{MAX_LCIDS, MAX_LCIDS, MAX_LCIDS}, 1, MAX_LCIDS, false, false, - MAX_LCIDS, MAX_LCIDS}, + /* DBG_GRC_PARAM_RESERVED2 */ + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, - /* DBG_GRC_PARAM_NUM_LTIDS */ - {{MAX_LTIDS, MAX_LTIDS, MAX_LTIDS}, 1, MAX_LTIDS, false, false, - MAX_LTIDS, MAX_LTIDS}, + /* DBG_GRC_PARAM_MCP_TRACE_META_SIZE */ + {{0, 0}, 1, 0xffffffff, false, true, 0, {0, 0}}, /* DBG_GRC_PARAM_EXCLUDE_ALL */ - {{0, 0, 0}, 0, 1, true, false, 0, 0}, + {{0, 0}, 0, 1, true, false, 0, {0, 0}}, /* DBG_GRC_PARAM_CRASH */ - {{0, 0, 0}, 0, 1, true, false, 0, 0}, + {{0, 0}, 0, 1, true, false, 0, {0, 0}}, /* DBG_GRC_PARAM_PARITY_SAFE */ - {{0, 0, 0}, 0, 1, false, false, 1, 0}, + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, /* DBG_GRC_PARAM_DUMP_CM */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{1, 1}, 0, 1, false, false, 0, {1, 1}}, /* DBG_GRC_PARAM_DUMP_PHY */ - {{1, 1, 1}, 0, 1, false, false, 0, 1}, + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, /* DBG_GRC_PARAM_NO_MCP */ - {{0, 0, 0}, 0, 1, false, false, 0, 0}, + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, /* DBG_GRC_PARAM_NO_FW_VER */ - {{0, 0, 0}, 0, 1, false, false, 0, 0} + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, + + /* DBG_GRC_PARAM_RESERVED3 */ + {{0, 0}, 0, 1, false, false, 0, {0, 0}}, + + /* DBG_GRC_PARAM_DUMP_MCP_HW_DUMP */ + {{0, 1}, 0, 1, false, false, 0, {0, 1}}, + + /* DBG_GRC_PARAM_DUMP_ILT_CDUC */ + {{1, 1}, 0, 1, false, false, 0, {0, 0}}, + + /* DBG_GRC_PARAM_DUMP_ILT_CDUT */ + {{1, 1}, 0, 1, false, false, 0, {0, 0}}, + + /* DBG_GRC_PARAM_DUMP_CAU_EXT */ + {{0, 0}, 0, 1, false, false, 0, {1, 1}} }; static struct rss_mem_defs s_rss_mem_defs[] = { - { "rss_mem_cid", "rss_cid", 0, 32, - {256, 320, 512} }, + {"rss_mem_cid", "rss_cid", 0, 32, + {256, 320}}, - { "rss_mem_key_msb", "rss_key", 1024, 256, - {128, 208, 257} }, + {"rss_mem_key_msb", "rss_key", 1024, 256, + {128, 208}}, - { "rss_mem_key_lsb", "rss_key", 2048, 64, - {128, 208, 257} }, + {"rss_mem_key_lsb", "rss_key", 2048, 64, + {128, 208}}, - { "rss_mem_info", "rss_info", 3072, 16, - {128, 208, 256} }, + {"rss_mem_info", "rss_info", 3072, 16, + {128, 208}}, - { "rss_mem_ind", "rss_ind", 4096, 16, - {16384, 26624, 32768} } + {"rss_mem_ind", "rss_ind", 4096, 16, + {16384, 26624}} }; static struct vfc_ram_defs s_vfc_ram_defs[] = { @@ -1655,54 +742,31 @@ static struct vfc_ram_defs s_vfc_ram_defs[] = { }; static struct big_ram_defs s_big_ram_defs[] = { - { "BRB", MEM_GROUP_BRB_MEM, MEM_GROUP_BRB_RAM, DBG_GRC_PARAM_DUMP_BRB, - BRB_REG_BIG_RAM_ADDRESS, BRB_REG_BIG_RAM_DATA, - MISC_REG_BLOCK_256B_EN, {0, 0, 0}, - {153600, 180224, 282624} }, - - { "BTB", MEM_GROUP_BTB_MEM, MEM_GROUP_BTB_RAM, DBG_GRC_PARAM_DUMP_BTB, - BTB_REG_BIG_RAM_ADDRESS, BTB_REG_BIG_RAM_DATA, - MISC_REG_BLOCK_256B_EN, {0, 1, 1}, - {92160, 117760, 168960} }, - - { "BMB", MEM_GROUP_BMB_MEM, MEM_GROUP_BMB_RAM, DBG_GRC_PARAM_DUMP_BMB, - BMB_REG_BIG_RAM_ADDRESS, BMB_REG_BIG_RAM_DATA, - MISCS_REG_BLOCK_256B_EN, {0, 0, 0}, - {36864, 36864, 36864} } -}; - -static struct reset_reg_defs s_reset_regs_defs[] = { - /* DBG_RESET_REG_MISCS_PL_UA */ - { MISCS_REG_RESET_PL_UA, - {true, true, true}, {0x0, 0x0, 0x0} }, - - /* DBG_RESET_REG_MISCS_PL_HV */ - { MISCS_REG_RESET_PL_HV, - {true, true, true}, {0x0, 0x400, 0x600} }, - - /* DBG_RESET_REG_MISCS_PL_HV_2 */ - { MISCS_REG_RESET_PL_HV_2_K2_E5, - {false, true, true}, {0x0, 0x0, 0x0} }, - - /* DBG_RESET_REG_MISC_PL_UA */ - { MISC_REG_RESET_PL_UA, - {true, true, true}, {0x0, 0x0, 0x0} }, - - /* DBG_RESET_REG_MISC_PL_HV */ - { MISC_REG_RESET_PL_HV, - {true, true, true}, {0x0, 0x0, 0x0} }, + {"BRB", MEM_GROUP_BRB_MEM, MEM_GROUP_BRB_RAM, DBG_GRC_PARAM_DUMP_BRB, + BRB_REG_BIG_RAM_ADDRESS, BRB_REG_BIG_RAM_DATA, + MISC_REG_BLOCK_256B_EN, {0, 0}, + {153600, 180224}}, - /* DBG_RESET_REG_MISC_PL_PDA_VMAIN_1 */ - { MISC_REG_RESET_PL_PDA_VMAIN_1, - {true, true, true}, {0x4404040, 0x4404040, 0x404040} }, + {"BTB", MEM_GROUP_BTB_MEM, MEM_GROUP_BTB_RAM, DBG_GRC_PARAM_DUMP_BTB, + BTB_REG_BIG_RAM_ADDRESS, BTB_REG_BIG_RAM_DATA, + MISC_REG_BLOCK_256B_EN, {0, 1}, + {92160, 117760}}, - /* DBG_RESET_REG_MISC_PL_PDA_VMAIN_2 */ - { MISC_REG_RESET_PL_PDA_VMAIN_2, - {true, true, true}, {0x7, 0x7c00007, 0x5c08007} }, + {"BMB", MEM_GROUP_BMB_MEM, MEM_GROUP_BMB_RAM, DBG_GRC_PARAM_DUMP_BMB, + BMB_REG_BIG_RAM_ADDRESS, BMB_REG_BIG_RAM_DATA, + MISCS_REG_BLOCK_256B_EN, {0, 0}, + {36864, 36864}} +}; - /* DBG_RESET_REG_MISC_PL_PDA_VAUX */ - { MISC_REG_RESET_PL_PDA_VAUX, - {true, true, true}, {0x2, 0x2, 0x2} }, +static struct rbc_reset_defs s_rbc_reset_defs[] = { + {MISCS_REG_RESET_PL_HV, + {0x0, 0x400}}, + {MISC_REG_RESET_PL_PDA_VMAIN_1, + {0x4404040, 0x4404040}}, + {MISC_REG_RESET_PL_PDA_VMAIN_2, + {0x7, 0x7c00007}}, + {MISC_REG_RESET_PL_PDA_VAUX, + {0x2, 0x2}}, }; static struct phy_defs s_phy_defs[] = { @@ -1785,9 +849,19 @@ static void qed_dbg_grc_init_params(struct qed_hwfn *p_hwfn) } } +/* Sets pointer and size for the specified binary buffer type */ +static void qed_set_dbg_bin_buf(struct qed_hwfn *p_hwfn, + enum bin_dbg_buffer_type buf_type, + const u32 *ptr, u32 size) +{ + struct virt_mem_desc *buf = &p_hwfn->dbg_arrays[buf_type]; + + buf->ptr = (void *)ptr; + buf->size = size; +} + /* Initializes debug data for the specified device */ -static enum dbg_status qed_dbg_dev_init(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt) +static enum dbg_status qed_dbg_dev_init(struct qed_hwfn *p_hwfn) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; u8 num_pfs = 0, max_pfs_per_port = 0; @@ -1812,26 +886,25 @@ static enum dbg_status qed_dbg_dev_init(struct qed_hwfn *p_hwfn, return DBG_STATUS_UNKNOWN_CHIP; } - /* Set platofrm */ - dev_data->platform_id = PLATFORM_ASIC; + /* Set HW type */ + dev_data->hw_type = HW_TYPE_ASIC; dev_data->mode_enable[MODE_ASIC] = 1; /* Set port mode */ - switch (qed_rd(p_hwfn, p_ptt, MISC_REG_PORT_MODE)) { - case 0: + switch (p_hwfn->cdev->num_ports_in_engine) { + case 1: dev_data->mode_enable[MODE_PORTS_PER_ENG_1] = 1; break; - case 1: + case 2: dev_data->mode_enable[MODE_PORTS_PER_ENG_2] = 1; break; - case 2: + case 4: dev_data->mode_enable[MODE_PORTS_PER_ENG_4] = 1; break; } /* Set 100G mode */ - if (dev_data->chip_id == CHIP_BB && - qed_rd(p_hwfn, p_ptt, CNIG_REG_NW_PORT_MODE_BB) == 2) + if (QED_IS_CMT(p_hwfn->cdev)) dev_data->mode_enable[MODE_100G] = 1; /* Set number of ports */ @@ -1857,14 +930,36 @@ static enum dbg_status qed_dbg_dev_init(struct qed_hwfn *p_hwfn, return DBG_STATUS_OK; } -static struct dbg_bus_block *get_dbg_bus_block_desc(struct qed_hwfn *p_hwfn, - enum block_id block_id) +static const struct dbg_block *get_dbg_block(struct qed_hwfn *p_hwfn, + enum block_id block_id) +{ + const struct dbg_block *dbg_block; + + dbg_block = p_hwfn->dbg_arrays[BIN_BUF_DBG_BLOCKS].ptr; + return dbg_block + block_id; +} + +static const struct dbg_block_chip *qed_get_dbg_block_per_chip(struct qed_hwfn + *p_hwfn, + enum block_id + block_id) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - return (struct dbg_bus_block *)&dbg_bus_blocks[block_id * - MAX_CHIP_IDS + - dev_data->chip_id]; + return (const struct dbg_block_chip *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_BLOCKS_CHIP_DATA].ptr + + block_id * MAX_CHIP_IDS + dev_data->chip_id; +} + +static const struct dbg_reset_reg *qed_get_dbg_reset_reg(struct qed_hwfn + *p_hwfn, + u8 reset_reg_id) +{ + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; + + return (const struct dbg_reset_reg *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_RESET_REGS].ptr + + reset_reg_id * MAX_CHIP_IDS + dev_data->chip_id; } /* Reads the FW info structure for the specified Storm from the chip, @@ -1885,8 +980,9 @@ static void qed_read_storm_fw_info(struct qed_hwfn *p_hwfn, * The address is located in the last line of the Storm RAM. */ addr = storm->sem_fast_mem_addr + SEM_FAST_REG_INT_RAM + - DWORDS_TO_BYTES(SEM_FAST_REG_INT_RAM_SIZE_BB_K2) - - sizeof(fw_info_location); + DWORDS_TO_BYTES(SEM_FAST_REG_INT_RAM_SIZE) - + sizeof(fw_info_location); + dest = (u32 *)&fw_info_location; for (i = 0; i < BYTES_TO_DWORDS(sizeof(fw_info_location)); @@ -2081,6 +1177,29 @@ static u32 qed_dump_mfw_ver_param(struct qed_hwfn *p_hwfn, return qed_dump_str_param(dump_buf, dump, "mfw-version", mfw_ver_str); } +/* Reads the chip revision from the chip and writes it as a param to the + * specified buffer. Returns the dumped size in dwords. + */ +static u32 qed_dump_chip_revision_param(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 *dump_buf, bool dump) +{ + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; + char param_str[3] = "??"; + + if (dev_data->hw_type == HW_TYPE_ASIC) { + u32 chip_rev, chip_metal; + + chip_rev = qed_rd(p_hwfn, p_ptt, MISCS_REG_CHIP_REV); + chip_metal = qed_rd(p_hwfn, p_ptt, MISCS_REG_CHIP_METAL); + + param_str[0] = 'a' + (u8)chip_rev; + param_str[1] = '0' + (u8)chip_metal; + } + + return qed_dump_str_param(dump_buf, dump, "chip-revision", param_str); +} + /* Writes a section header to the specified buffer. * Returns the dumped size in dwords. */ @@ -2104,7 +1223,8 @@ static u32 qed_dump_common_global_params(struct qed_hwfn *p_hwfn, u8 num_params; /* Dump global params section header */ - num_params = NUM_COMMON_GLOBAL_PARAMS + num_specific_global_params; + num_params = NUM_COMMON_GLOBAL_PARAMS + num_specific_global_params + + (dev_data->chip_id == CHIP_BB ? 1 : 0); offset += qed_dump_section_hdr(dump_buf + offset, dump, "global_params", num_params); @@ -2112,6 +1232,8 @@ static u32 qed_dump_common_global_params(struct qed_hwfn *p_hwfn, offset += qed_dump_fw_ver_param(p_hwfn, p_ptt, dump_buf + offset, dump); offset += qed_dump_mfw_ver_param(p_hwfn, p_ptt, dump_buf + offset, dump); + offset += qed_dump_chip_revision_param(p_hwfn, + p_ptt, dump_buf + offset, dump); offset += qed_dump_num_param(dump_buf + offset, dump, "tools-version", TOOLS_VERSION); offset += qed_dump_str_param(dump_buf + offset, @@ -2121,11 +1243,12 @@ static u32 qed_dump_common_global_params(struct qed_hwfn *p_hwfn, offset += qed_dump_str_param(dump_buf + offset, dump, "platform", - s_platform_defs[dev_data->platform_id]. - name); - offset += - qed_dump_num_param(dump_buf + offset, dump, "pci-func", - p_hwfn->abs_pf_id); + s_hw_type_defs[dev_data->hw_type].name); + offset += qed_dump_num_param(dump_buf + offset, + dump, "pci-func", p_hwfn->abs_pf_id); + if (dev_data->chip_id == CHIP_BB) + offset += qed_dump_num_param(dump_buf + offset, + dump, "path", QED_PATH_ID(p_hwfn)); return offset; } @@ -2156,24 +1279,87 @@ static void qed_update_blocks_reset_state(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - u32 reg_val[MAX_DBG_RESET_REGS] = { 0 }; - u32 i; + u32 reg_val[NUM_DBG_RESET_REGS] = { 0 }; + u8 rst_reg_id; + u32 blk_id; /* Read reset registers */ - for (i = 0; i < MAX_DBG_RESET_REGS; i++) - if (s_reset_regs_defs[i].exists[dev_data->chip_id]) - reg_val[i] = qed_rd(p_hwfn, - p_ptt, s_reset_regs_defs[i].addr); + for (rst_reg_id = 0; rst_reg_id < NUM_DBG_RESET_REGS; rst_reg_id++) { + const struct dbg_reset_reg *rst_reg; + bool rst_reg_removed; + u32 rst_reg_addr; + + rst_reg = qed_get_dbg_reset_reg(p_hwfn, rst_reg_id); + rst_reg_removed = GET_FIELD(rst_reg->data, + DBG_RESET_REG_IS_REMOVED); + rst_reg_addr = DWORDS_TO_BYTES(GET_FIELD(rst_reg->data, + DBG_RESET_REG_ADDR)); + + if (!rst_reg_removed) + reg_val[rst_reg_id] = qed_rd(p_hwfn, p_ptt, + rst_reg_addr); + } /* Check if blocks are in reset */ - for (i = 0; i < MAX_BLOCK_ID; i++) { - struct block_defs *block = s_block_defs[i]; + for (blk_id = 0; blk_id < NUM_PHYS_BLOCKS; blk_id++) { + const struct dbg_block_chip *blk; + bool has_rst_reg; + bool is_removed; + + blk = qed_get_dbg_block_per_chip(p_hwfn, (enum block_id)blk_id); + is_removed = GET_FIELD(blk->flags, DBG_BLOCK_CHIP_IS_REMOVED); + has_rst_reg = GET_FIELD(blk->flags, + DBG_BLOCK_CHIP_HAS_RESET_REG); + + if (!is_removed && has_rst_reg) + dev_data->block_in_reset[blk_id] = + !(reg_val[blk->reset_reg_id] & + BIT(blk->reset_reg_bit_offset)); + } +} + +/* is_mode_match recursive function */ +static bool qed_is_mode_match_rec(struct qed_hwfn *p_hwfn, + u16 *modes_buf_offset, u8 rec_depth) +{ + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; + u8 *dbg_array; + bool arg1, arg2; + u8 tree_val; + + if (rec_depth > MAX_RECURSION_DEPTH) { + DP_NOTICE(p_hwfn, + "Unexpected error: is_mode_match_rec exceeded the max recursion depth. This is probably due to a corrupt init/debug buffer.\n"); + return false; + } - dev_data->block_in_reset[i] = block->has_reset_bit && - !(reg_val[block->reset_reg] & BIT(block->reset_bit_offset)); + /* Get next element from modes tree buffer */ + dbg_array = p_hwfn->dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr; + tree_val = dbg_array[(*modes_buf_offset)++]; + + switch (tree_val) { + case INIT_MODE_OP_NOT: + return !qed_is_mode_match_rec(p_hwfn, + modes_buf_offset, rec_depth + 1); + case INIT_MODE_OP_OR: + case INIT_MODE_OP_AND: + arg1 = qed_is_mode_match_rec(p_hwfn, + modes_buf_offset, rec_depth + 1); + arg2 = qed_is_mode_match_rec(p_hwfn, + modes_buf_offset, rec_depth + 1); + return (tree_val == INIT_MODE_OP_OR) ? (arg1 || + arg2) : (arg1 && arg2); + default: + return dev_data->mode_enable[tree_val - MAX_INIT_MODE_OPS] > 0; } } +/* Returns true if the mode (specified using modes_buf_offset) is enabled */ +static bool qed_is_mode_match(struct qed_hwfn *p_hwfn, u16 *modes_buf_offset) +{ + return qed_is_mode_match_rec(p_hwfn, modes_buf_offset, 0); +} + /* Enable / disable the Debug block */ static void qed_bus_enable_dbg_block(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, bool enable) @@ -2185,23 +1371,21 @@ static void qed_bus_enable_dbg_block(struct qed_hwfn *p_hwfn, static void qed_bus_reset_dbg_block(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) { - u32 dbg_reset_reg_addr, old_reset_reg_val, new_reset_reg_val; - struct block_defs *dbg_block = s_block_defs[BLOCK_DBG]; + u32 reset_reg_addr, old_reset_reg_val, new_reset_reg_val; + const struct dbg_reset_reg *reset_reg; + const struct dbg_block_chip *block; - dbg_reset_reg_addr = s_reset_regs_defs[dbg_block->reset_reg].addr; - old_reset_reg_val = qed_rd(p_hwfn, p_ptt, dbg_reset_reg_addr); - new_reset_reg_val = - old_reset_reg_val & ~BIT(dbg_block->reset_bit_offset); + block = qed_get_dbg_block_per_chip(p_hwfn, BLOCK_DBG); + reset_reg = qed_get_dbg_reset_reg(p_hwfn, block->reset_reg_id); + reset_reg_addr = + DWORDS_TO_BYTES(GET_FIELD(reset_reg->data, DBG_RESET_REG_ADDR)); - qed_wr(p_hwfn, p_ptt, dbg_reset_reg_addr, new_reset_reg_val); - qed_wr(p_hwfn, p_ptt, dbg_reset_reg_addr, old_reset_reg_val); -} + old_reset_reg_val = qed_rd(p_hwfn, p_ptt, reset_reg_addr); + new_reset_reg_val = + old_reset_reg_val & ~BIT(block->reset_reg_bit_offset); -static void qed_bus_set_framing_mode(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, - enum dbg_bus_frame_modes mode) -{ - qed_wr(p_hwfn, p_ptt, DBG_REG_FRAMING_MODE, (u8)mode); + qed_wr(p_hwfn, p_ptt, reset_reg_addr, new_reset_reg_val); + qed_wr(p_hwfn, p_ptt, reset_reg_addr, old_reset_reg_val); } /* Enable / disable Debug Bus clients according to the specified mask @@ -2213,28 +1397,65 @@ static void qed_bus_enable_clients(struct qed_hwfn *p_hwfn, qed_wr(p_hwfn, p_ptt, DBG_REG_CLIENT_ENABLE, client_mask); } -static bool qed_is_mode_match(struct qed_hwfn *p_hwfn, u16 *modes_buf_offset) +static void qed_bus_config_dbg_line(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + enum block_id block_id, + u8 line_id, + u8 enable_mask, + u8 right_shift, + u8 force_valid_mask, u8 force_frame_mask) +{ + const struct dbg_block_chip *block = + qed_get_dbg_block_per_chip(p_hwfn, block_id); + + qed_wr(p_hwfn, p_ptt, DWORDS_TO_BYTES(block->dbg_select_reg_addr), + line_id); + qed_wr(p_hwfn, p_ptt, DWORDS_TO_BYTES(block->dbg_dword_enable_reg_addr), + enable_mask); + qed_wr(p_hwfn, p_ptt, DWORDS_TO_BYTES(block->dbg_shift_reg_addr), + right_shift); + qed_wr(p_hwfn, p_ptt, DWORDS_TO_BYTES(block->dbg_force_valid_reg_addr), + force_valid_mask); + qed_wr(p_hwfn, p_ptt, DWORDS_TO_BYTES(block->dbg_force_frame_reg_addr), + force_frame_mask); +} + +/* Disable debug bus in all blocks */ +static void qed_bus_disable_blocks(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - bool arg1, arg2; - const u32 *ptr; - u8 tree_val; + u32 block_id; - /* Get next element from modes tree buffer */ - ptr = s_dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr; - tree_val = ((u8 *)ptr)[(*modes_buf_offset)++]; + /* Disable all blocks */ + for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { + const struct dbg_block_chip *block_per_chip = + qed_get_dbg_block_per_chip(p_hwfn, + (enum block_id)block_id); - switch (tree_val) { - case INIT_MODE_OP_NOT: - return !qed_is_mode_match(p_hwfn, modes_buf_offset); - case INIT_MODE_OP_OR: - case INIT_MODE_OP_AND: - arg1 = qed_is_mode_match(p_hwfn, modes_buf_offset); - arg2 = qed_is_mode_match(p_hwfn, modes_buf_offset); - return (tree_val == INIT_MODE_OP_OR) ? (arg1 || - arg2) : (arg1 && arg2); - default: - return dev_data->mode_enable[tree_val - MAX_INIT_MODE_OPS] > 0; + if (GET_FIELD(block_per_chip->flags, + DBG_BLOCK_CHIP_IS_REMOVED) || + dev_data->block_in_reset[block_id]) + continue; + + /* Disable debug bus */ + if (GET_FIELD(block_per_chip->flags, + DBG_BLOCK_CHIP_HAS_DBG_BUS)) { + u32 dbg_en_addr = + block_per_chip->dbg_dword_enable_reg_addr; + u16 modes_buf_offset = + GET_FIELD(block_per_chip->dbg_bus_mode.data, + DBG_MODE_HDR_MODES_BUF_OFFSET); + bool eval_mode = + GET_FIELD(block_per_chip->dbg_bus_mode.data, + DBG_MODE_HDR_EVAL_MODE) > 0; + + if (!eval_mode || + qed_is_mode_match(p_hwfn, &modes_buf_offset)) + qed_wr(p_hwfn, p_ptt, + DWORDS_TO_BYTES(dbg_en_addr), + 0); + } } } @@ -2247,6 +1468,20 @@ static bool qed_grc_is_included(struct qed_hwfn *p_hwfn, return qed_grc_get_param(p_hwfn, grc_param) > 0; } +/* Returns the storm_id that matches the specified Storm letter, + * or MAX_DBG_STORMS if invalid storm letter. + */ +static enum dbg_storms qed_get_id_from_letter(char storm_letter) +{ + u8 storm_id; + + for (storm_id = 0; storm_id < MAX_DBG_STORMS; storm_id++) + if (s_storm_defs[storm_id].letter == storm_letter) + return (enum dbg_storms)storm_id; + + return MAX_DBG_STORMS; +} + /* Returns true of the specified Storm should be included in the dump, false * otherwise. */ @@ -2262,14 +1497,20 @@ static bool qed_grc_is_storm_included(struct qed_hwfn *p_hwfn, static bool qed_grc_is_mem_included(struct qed_hwfn *p_hwfn, enum block_id block_id, u8 mem_group_id) { - struct block_defs *block = s_block_defs[block_id]; + const struct dbg_block *block; u8 i; - /* Check Storm match */ - if (block->associated_to_storm && - !qed_grc_is_storm_included(p_hwfn, - (enum dbg_storms)block->storm_id)) - return false; + block = get_dbg_block(p_hwfn, block_id); + + /* If the block is associated with a Storm, check Storm match */ + if (block->associated_storm_letter) { + enum dbg_storms associated_storm_id = + qed_get_id_from_letter(block->associated_storm_letter); + + if (associated_storm_id == MAX_DBG_STORMS || + !qed_grc_is_storm_included(p_hwfn, associated_storm_id)) + return false; + } for (i = 0; i < NUM_BIG_RAM_TYPES; i++) { struct big_ram_defs *big_ram = &s_big_ram_defs[i]; @@ -2291,6 +1532,8 @@ static bool qed_grc_is_mem_included(struct qed_hwfn *p_hwfn, case MEM_GROUP_CAU_SB: case MEM_GROUP_CAU_PI: return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_CAU); + case MEM_GROUP_CAU_MEM_EXT: + return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_CAU_EXT); case MEM_GROUP_QM_MEM: return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_QM); case MEM_GROUP_CFC_MEM: @@ -2298,6 +1541,8 @@ static bool qed_grc_is_mem_included(struct qed_hwfn *p_hwfn, case MEM_GROUP_TASK_CFC_MEM: return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_CFC) || qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_CM_CTX); + case MEM_GROUP_DORQ_MEM: + return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_DORQ); case MEM_GROUP_IGU_MEM: case MEM_GROUP_IGU_MSIX: return qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_IGU); @@ -2343,64 +1588,104 @@ static void qed_grc_stall_storms(struct qed_hwfn *p_hwfn, msleep(STALL_DELAY_MS); } -/* Takes all blocks out of reset */ +/* Takes all blocks out of reset. If rbc_only is true, only RBC clients are + * taken out of reset. + */ static void qed_grc_unreset_blocks(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt) + struct qed_ptt *p_ptt, bool rbc_only) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - u32 reg_val[MAX_DBG_RESET_REGS] = { 0 }; - u32 block_id, i; + u8 chip_id = dev_data->chip_id; + u32 i; - /* Fill reset regs values */ - for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { - struct block_defs *block = s_block_defs[block_id]; + /* Take RBCs out of reset */ + for (i = 0; i < ARRAY_SIZE(s_rbc_reset_defs); i++) + if (s_rbc_reset_defs[i].reset_val[dev_data->chip_id]) + qed_wr(p_hwfn, + p_ptt, + s_rbc_reset_defs[i].reset_reg_addr + + RESET_REG_UNRESET_OFFSET, + s_rbc_reset_defs[i].reset_val[chip_id]); - if (block->exists[dev_data->chip_id] && block->has_reset_bit && - block->unreset) - reg_val[block->reset_reg] |= - BIT(block->reset_bit_offset); - } + if (!rbc_only) { + u32 reg_val[NUM_DBG_RESET_REGS] = { 0 }; + u8 reset_reg_id; + u32 block_id; - /* Write reset registers */ - for (i = 0; i < MAX_DBG_RESET_REGS; i++) { - if (!s_reset_regs_defs[i].exists[dev_data->chip_id]) - continue; + /* Fill reset regs values */ + for (block_id = 0; block_id < NUM_PHYS_BLOCKS; block_id++) { + bool is_removed, has_reset_reg, unreset_before_dump; + const struct dbg_block_chip *block; + + block = qed_get_dbg_block_per_chip(p_hwfn, + (enum block_id) + block_id); + is_removed = + GET_FIELD(block->flags, DBG_BLOCK_CHIP_IS_REMOVED); + has_reset_reg = + GET_FIELD(block->flags, + DBG_BLOCK_CHIP_HAS_RESET_REG); + unreset_before_dump = + GET_FIELD(block->flags, + DBG_BLOCK_CHIP_UNRESET_BEFORE_DUMP); + + if (!is_removed && has_reset_reg && unreset_before_dump) + reg_val[block->reset_reg_id] |= + BIT(block->reset_reg_bit_offset); + } - reg_val[i] |= - s_reset_regs_defs[i].unreset_val[dev_data->chip_id]; + /* Write reset registers */ + for (reset_reg_id = 0; reset_reg_id < NUM_DBG_RESET_REGS; + reset_reg_id++) { + const struct dbg_reset_reg *reset_reg; + u32 reset_reg_addr; - if (reg_val[i]) - qed_wr(p_hwfn, - p_ptt, - s_reset_regs_defs[i].addr + - RESET_REG_UNRESET_OFFSET, reg_val[i]); + reset_reg = qed_get_dbg_reset_reg(p_hwfn, reset_reg_id); + + if (GET_FIELD + (reset_reg->data, DBG_RESET_REG_IS_REMOVED)) + continue; + + if (reg_val[reset_reg_id]) { + reset_reg_addr = + GET_FIELD(reset_reg->data, + DBG_RESET_REG_ADDR); + qed_wr(p_hwfn, + p_ptt, + DWORDS_TO_BYTES(reset_reg_addr) + + RESET_REG_UNRESET_OFFSET, + reg_val[reset_reg_id]); + } + } } } /* Returns the attention block data of the specified block */ static const struct dbg_attn_block_type_data * -qed_get_block_attn_data(enum block_id block_id, enum dbg_attn_type attn_type) +qed_get_block_attn_data(struct qed_hwfn *p_hwfn, + enum block_id block_id, enum dbg_attn_type attn_type) { const struct dbg_attn_block *base_attn_block_arr = - (const struct dbg_attn_block *) - s_dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr; + (const struct dbg_attn_block *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr; return &base_attn_block_arr[block_id].per_type_data[attn_type]; } /* Returns the attention registers of the specified block */ static const struct dbg_attn_reg * -qed_get_block_attn_regs(enum block_id block_id, enum dbg_attn_type attn_type, +qed_get_block_attn_regs(struct qed_hwfn *p_hwfn, + enum block_id block_id, enum dbg_attn_type attn_type, u8 *num_attn_regs) { const struct dbg_attn_block_type_data *block_type_data = - qed_get_block_attn_data(block_id, attn_type); + qed_get_block_attn_data(p_hwfn, block_id, attn_type); *num_attn_regs = block_type_data->num_regs; - return &((const struct dbg_attn_reg *) - s_dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr)[block_type_data-> - regs_offset]; + return (const struct dbg_attn_reg *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr + + block_type_data->regs_offset; } /* For each block, clear the status of all parities */ @@ -2412,11 +1697,12 @@ static void qed_grc_clear_all_prty(struct qed_hwfn *p_hwfn, u8 reg_idx, num_attn_regs; u32 block_id; - for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { + for (block_id = 0; block_id < NUM_PHYS_BLOCKS; block_id++) { if (dev_data->block_in_reset[block_id]) continue; - attn_reg_arr = qed_get_block_attn_regs((enum block_id)block_id, + attn_reg_arr = qed_get_block_attn_regs(p_hwfn, + (enum block_id)block_id, ATTN_TYPE_PARITY, &num_attn_regs); @@ -2444,22 +1730,20 @@ static void qed_grc_clear_all_prty(struct qed_hwfn *p_hwfn, } /* Dumps GRC registers section header. Returns the dumped size in dwords. - * The following parameters are dumped: + * the following parameters are dumped: * - count: no. of dumped entries * - split_type: split type * - split_id: split ID (dumped only if split_id != SPLIT_TYPE_NONE) - * - param_name: user parameter value (dumped only if param_name != NULL - * and param_val != NULL). + * - reg_type_name: register type name (dumped only if reg_type_name != NULL) */ static u32 qed_grc_dump_regs_hdr(u32 *dump_buf, bool dump, u32 num_reg_entries, enum init_split_types split_type, - u8 split_id, - const char *param_name, const char *param_val) + u8 split_id, const char *reg_type_name) { u8 num_params = 2 + - (split_type != SPLIT_TYPE_NONE ? 1 : 0) + (param_name ? 1 : 0); + (split_type != SPLIT_TYPE_NONE ? 1 : 0) + (reg_type_name ? 1 : 0); u32 offset = 0; offset += qed_dump_section_hdr(dump_buf + offset, @@ -2472,9 +1756,9 @@ static u32 qed_grc_dump_regs_hdr(u32 *dump_buf, if (split_type != SPLIT_TYPE_NONE) offset += qed_dump_num_param(dump_buf + offset, dump, "id", split_id); - if (param_name && param_val) + if (reg_type_name) offset += qed_dump_str_param(dump_buf + offset, - dump, param_name, param_val); + dump, "type", reg_type_name); return offset; } @@ -2504,21 +1788,12 @@ static u32 qed_grc_dump_addr_range(struct qed_hwfn *p_hwfn, { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; u8 port_id = 0, pf_id = 0, vf_id = 0, fid = 0; + bool read_using_dmae = false; + u32 thresh; if (!dump) return len; - /* Print log if needed */ - dev_data->num_regs_read += len; - if (dev_data->num_regs_read >= - s_platform_defs[dev_data->platform_id].log_thresh) { - DP_VERBOSE(p_hwfn, - QED_MSG_DEBUG, - "Dumping %d registers...\n", - dev_data->num_regs_read); - dev_data->num_regs_read = 0; - } - switch (split_type) { case SPLIT_TYPE_PORT: port_id = split_id; @@ -2539,38 +1814,77 @@ static u32 qed_grc_dump_addr_range(struct qed_hwfn *p_hwfn, } /* Try reading using DMAE */ - if (dev_data->use_dmae && split_type == SPLIT_TYPE_NONE && - (len >= s_platform_defs[dev_data->platform_id].dmae_thresh || - wide_bus)) { - if (!qed_dmae_grc2host(p_hwfn, p_ptt, DWORDS_TO_BYTES(addr), - (u64)(uintptr_t)(dump_buf), len, NULL)) - return len; - dev_data->use_dmae = 0; - DP_VERBOSE(p_hwfn, - QED_MSG_DEBUG, - "Failed reading from chip using DMAE, using GRC instead\n"); + if (dev_data->use_dmae && split_type != SPLIT_TYPE_VF && + (len >= s_hw_type_defs[dev_data->hw_type].dmae_thresh || + (PROTECT_WIDE_BUS && wide_bus))) { + struct qed_dmae_params dmae_params; + + /* Set DMAE params */ + memset(&dmae_params, 0, sizeof(dmae_params)); + SET_FIELD(dmae_params.flags, QED_DMAE_PARAMS_COMPLETION_DST, 1); + switch (split_type) { + case SPLIT_TYPE_PORT: + SET_FIELD(dmae_params.flags, QED_DMAE_PARAMS_PORT_VALID, + 1); + dmae_params.port_id = port_id; + break; + case SPLIT_TYPE_PF: + SET_FIELD(dmae_params.flags, + QED_DMAE_PARAMS_SRC_PF_VALID, 1); + dmae_params.src_pfid = pf_id; + break; + case SPLIT_TYPE_PORT_PF: + SET_FIELD(dmae_params.flags, QED_DMAE_PARAMS_PORT_VALID, + 1); + SET_FIELD(dmae_params.flags, + QED_DMAE_PARAMS_SRC_PF_VALID, 1); + dmae_params.port_id = port_id; + dmae_params.src_pfid = pf_id; + break; + default: + break; + } + + /* Execute DMAE command */ + read_using_dmae = !qed_dmae_grc2host(p_hwfn, + p_ptt, + DWORDS_TO_BYTES(addr), + (u64)(uintptr_t)(dump_buf), + len, &dmae_params); + if (!read_using_dmae) { + dev_data->use_dmae = 0; + DP_VERBOSE(p_hwfn, + QED_MSG_DEBUG, + "Failed reading from chip using DMAE, using GRC instead\n"); + } } + if (read_using_dmae) + goto print_log; + /* If not read using DMAE, read using GRC */ /* Set pretend */ - if (split_type != dev_data->pretend.split_type || split_id != - dev_data->pretend.split_id) { + if (split_type != dev_data->pretend.split_type || + split_id != dev_data->pretend.split_id) { switch (split_type) { case SPLIT_TYPE_PORT: qed_port_pretend(p_hwfn, p_ptt, port_id); break; case SPLIT_TYPE_PF: - fid = pf_id << PXP_PRETEND_CONCRETE_FID_PFID_SHIFT; + fid = FIELD_VALUE(PXP_PRETEND_CONCRETE_FID_PFID, + pf_id); qed_fid_pretend(p_hwfn, p_ptt, fid); break; case SPLIT_TYPE_PORT_PF: - fid = pf_id << PXP_PRETEND_CONCRETE_FID_PFID_SHIFT; + fid = FIELD_VALUE(PXP_PRETEND_CONCRETE_FID_PFID, + pf_id); qed_port_fid_pretend(p_hwfn, p_ptt, port_id, fid); break; case SPLIT_TYPE_VF: - fid = BIT(PXP_PRETEND_CONCRETE_FID_VFVALID_SHIFT) | - (vf_id << PXP_PRETEND_CONCRETE_FID_VFID_SHIFT); + fid = FIELD_VALUE(PXP_PRETEND_CONCRETE_FID_VFVALID, 1) + | FIELD_VALUE(PXP_PRETEND_CONCRETE_FID_VFID, + vf_id); qed_fid_pretend(p_hwfn, p_ptt, fid); break; default: @@ -2584,6 +1898,16 @@ static u32 qed_grc_dump_addr_range(struct qed_hwfn *p_hwfn, /* Read registers using GRC */ qed_read_regs(p_hwfn, p_ptt, dump_buf, addr, len); +print_log: + /* Print log */ + dev_data->num_regs_read += len; + thresh = s_hw_type_defs[dev_data->hw_type].log_thresh; + if ((dev_data->num_regs_read / thresh) > + ((dev_data->num_regs_read - len) / thresh)) + DP_VERBOSE(p_hwfn, + QED_MSG_DEBUG, + "Dumped %d registers...\n", dev_data->num_regs_read); + return len; } @@ -2668,7 +1992,7 @@ static u32 qed_grc_dump_reg_entry_skip(struct qed_hwfn *p_hwfn, /* Dumps GRC registers entries. Returns the dumped size in dwords. */ static u32 qed_grc_dump_regs_entries(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, - struct dbg_array input_regs_arr, + struct virt_mem_desc input_regs_arr, u32 *dump_buf, bool dump, enum init_split_types split_type, @@ -2681,10 +2005,10 @@ static u32 qed_grc_dump_regs_entries(struct qed_hwfn *p_hwfn, *num_dumped_reg_entries = 0; - while (input_offset < input_regs_arr.size_in_dwords) { + while (input_offset < BYTES_TO_DWORDS(input_regs_arr.size)) { const struct dbg_dump_cond_hdr *cond_hdr = (const struct dbg_dump_cond_hdr *) - &input_regs_arr.ptr[input_offset++]; + input_regs_arr.ptr + input_offset++; u16 modes_buf_offset; bool eval_mode; @@ -2707,7 +2031,7 @@ static u32 qed_grc_dump_regs_entries(struct qed_hwfn *p_hwfn, for (i = 0; i < cond_hdr->data_size; i++, input_offset++) { const struct dbg_dump_reg *reg = (const struct dbg_dump_reg *) - &input_regs_arr.ptr[input_offset]; + input_regs_arr.ptr + input_offset; u32 addr, len; bool wide_bus; @@ -2732,14 +2056,12 @@ static u32 qed_grc_dump_regs_entries(struct qed_hwfn *p_hwfn, /* Dumps GRC registers entries. Returns the dumped size in dwords. */ static u32 qed_grc_dump_split_data(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, - struct dbg_array input_regs_arr, + struct virt_mem_desc input_regs_arr, u32 *dump_buf, bool dump, bool block_enable[MAX_BLOCK_ID], enum init_split_types split_type, - u8 split_id, - const char *param_name, - const char *param_val) + u8 split_id, const char *reg_type_name) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; enum init_split_types hdr_split_type = split_type; @@ -2757,7 +2079,7 @@ static u32 qed_grc_dump_split_data(struct qed_hwfn *p_hwfn, false, 0, hdr_split_type, - hdr_split_id, param_name, param_val); + hdr_split_id, reg_type_name); /* Dump registers */ offset += qed_grc_dump_regs_entries(p_hwfn, @@ -2776,7 +2098,7 @@ static u32 qed_grc_dump_split_data(struct qed_hwfn *p_hwfn, dump, num_dumped_reg_entries, hdr_split_type, - hdr_split_id, param_name, param_val); + hdr_split_id, reg_type_name); return num_dumped_reg_entries > 0 ? offset : 0; } @@ -2789,32 +2111,33 @@ static u32 qed_grc_dump_registers(struct qed_hwfn *p_hwfn, u32 *dump_buf, bool dump, bool block_enable[MAX_BLOCK_ID], - const char *param_name, const char *param_val) + const char *reg_type_name) { + struct virt_mem_desc *dbg_buf = + &p_hwfn->dbg_arrays[BIN_BUF_DBG_DUMP_REG]; struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; u32 offset = 0, input_offset = 0; - u16 fid; - while (input_offset < - s_dbg_arrays[BIN_BUF_DBG_DUMP_REG].size_in_dwords) { + + while (input_offset < BYTES_TO_DWORDS(dbg_buf->size)) { const struct dbg_dump_split_hdr *split_hdr; - struct dbg_array curr_input_regs_arr; + struct virt_mem_desc curr_input_regs_arr; enum init_split_types split_type; u16 split_count = 0; u32 split_data_size; u8 split_id; split_hdr = - (const struct dbg_dump_split_hdr *) - &s_dbg_arrays[BIN_BUF_DBG_DUMP_REG].ptr[input_offset++]; + (const struct dbg_dump_split_hdr *) + dbg_buf->ptr + input_offset++; split_type = - GET_FIELD(split_hdr->hdr, - DBG_DUMP_SPLIT_HDR_SPLIT_TYPE_ID); - split_data_size = - GET_FIELD(split_hdr->hdr, - DBG_DUMP_SPLIT_HDR_DATA_SIZE); + GET_FIELD(split_hdr->hdr, + DBG_DUMP_SPLIT_HDR_SPLIT_TYPE_ID); + split_data_size = GET_FIELD(split_hdr->hdr, + DBG_DUMP_SPLIT_HDR_DATA_SIZE); curr_input_regs_arr.ptr = - &s_dbg_arrays[BIN_BUF_DBG_DUMP_REG].ptr[input_offset]; - curr_input_regs_arr.size_in_dwords = split_data_size; + (u32 *)p_hwfn->dbg_arrays[BIN_BUF_DBG_DUMP_REG].ptr + + input_offset; + curr_input_regs_arr.size = DWORDS_TO_BYTES(split_data_size); switch (split_type) { case SPLIT_TYPE_NONE: @@ -2842,16 +2165,16 @@ static u32 qed_grc_dump_registers(struct qed_hwfn *p_hwfn, dump, block_enable, split_type, split_id, - param_name, - param_val); + reg_type_name); input_offset += split_data_size; } /* Cancel pretends (pretend to original PF) */ if (dump) { - fid = p_hwfn->rel_pf_id << PXP_PRETEND_CONCRETE_FID_PFID_SHIFT; - qed_fid_pretend(p_hwfn, p_ptt, fid); + qed_fid_pretend(p_hwfn, p_ptt, + FIELD_VALUE(PXP_PRETEND_CONCRETE_FID_PFID, + p_hwfn->rel_pf_id)); dev_data->pretend.split_type = SPLIT_TYPE_NONE; dev_data->pretend.split_id = 0; } @@ -2864,26 +2187,32 @@ static u32 qed_grc_dump_reset_regs(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { - struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - u32 i, offset = 0, num_regs = 0; + u32 offset = 0, num_regs = 0; + u8 reset_reg_id; /* Calculate header size */ offset += qed_grc_dump_regs_hdr(dump_buf, - false, 0, - SPLIT_TYPE_NONE, 0, NULL, NULL); + false, + 0, SPLIT_TYPE_NONE, 0, "RESET_REGS"); /* Write reset registers */ - for (i = 0; i < MAX_DBG_RESET_REGS; i++) { - if (!s_reset_regs_defs[i].exists[dev_data->chip_id]) + for (reset_reg_id = 0; reset_reg_id < NUM_DBG_RESET_REGS; + reset_reg_id++) { + const struct dbg_reset_reg *reset_reg; + u32 reset_reg_addr; + + reset_reg = qed_get_dbg_reset_reg(p_hwfn, reset_reg_id); + + if (GET_FIELD(reset_reg->data, DBG_RESET_REG_IS_REMOVED)) continue; + reset_reg_addr = GET_FIELD(reset_reg->data, DBG_RESET_REG_ADDR); offset += qed_grc_dump_reg_entry(p_hwfn, p_ptt, dump_buf + offset, dump, - BYTES_TO_DWORDS - (s_reset_regs_defs[i].addr), 1, - false, SPLIT_TYPE_NONE, 0); + reset_reg_addr, + 1, false, SPLIT_TYPE_NONE, 0); num_regs++; } @@ -2891,7 +2220,7 @@ static u32 qed_grc_dump_reset_regs(struct qed_hwfn *p_hwfn, if (dump) qed_grc_dump_regs_hdr(dump_buf, true, num_regs, SPLIT_TYPE_NONE, - 0, NULL, NULL); + 0, "RESET_REGS"); return offset; } @@ -2904,21 +2233,23 @@ static u32 qed_grc_dump_modified_regs(struct qed_hwfn *p_hwfn, u32 *dump_buf, bool dump) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - u32 block_id, offset = 0, num_reg_entries = 0; + u32 block_id, offset = 0, stall_regs_offset; const struct dbg_attn_reg *attn_reg_arr; u8 storm_id, reg_idx, num_attn_regs; + u32 num_reg_entries = 0; - /* Calculate header size */ + /* Write empty header for attention registers */ offset += qed_grc_dump_regs_hdr(dump_buf, - false, 0, SPLIT_TYPE_NONE, - 0, NULL, NULL); + false, + 0, SPLIT_TYPE_NONE, 0, "ATTN_REGS"); /* Write parity registers */ - for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { + for (block_id = 0; block_id < NUM_PHYS_BLOCKS; block_id++) { if (dev_data->block_in_reset[block_id] && dump) continue; - attn_reg_arr = qed_get_block_attn_regs((enum block_id)block_id, + attn_reg_arr = qed_get_block_attn_regs(p_hwfn, + (enum block_id)block_id, ATTN_TYPE_PARITY, &num_attn_regs); @@ -2961,16 +2292,29 @@ static u32 qed_grc_dump_modified_regs(struct qed_hwfn *p_hwfn, } } + /* Overwrite header for attention registers */ + if (dump) + qed_grc_dump_regs_hdr(dump_buf, + true, + num_reg_entries, + SPLIT_TYPE_NONE, 0, "ATTN_REGS"); + + /* Write empty header for stall registers */ + stall_regs_offset = offset; + offset += qed_grc_dump_regs_hdr(dump_buf, + false, 0, SPLIT_TYPE_NONE, 0, "REGS"); + /* Write Storm stall status registers */ - for (storm_id = 0; storm_id < MAX_DBG_STORMS; storm_id++) { + for (storm_id = 0, num_reg_entries = 0; storm_id < MAX_DBG_STORMS; + storm_id++) { struct storm_defs *storm = &s_storm_defs[storm_id]; u32 addr; - if (dev_data->block_in_reset[storm->block_id] && dump) + if (dev_data->block_in_reset[storm->sem_block_id] && dump) continue; addr = - BYTES_TO_DWORDS(s_storm_defs[storm_id].sem_fast_mem_addr + + BYTES_TO_DWORDS(storm->sem_fast_mem_addr + SEM_FAST_REG_STALLED); offset += qed_grc_dump_reg_entry(p_hwfn, p_ptt, @@ -2982,12 +2326,12 @@ static u32 qed_grc_dump_modified_regs(struct qed_hwfn *p_hwfn, num_reg_entries++; } - /* Write header */ + /* Overwrite header for stall registers */ if (dump) - qed_grc_dump_regs_hdr(dump_buf, + qed_grc_dump_regs_hdr(dump_buf + stall_regs_offset, true, - num_reg_entries, SPLIT_TYPE_NONE, - 0, NULL, NULL); + num_reg_entries, + SPLIT_TYPE_NONE, 0, "REGS"); return offset; } @@ -3000,8 +2344,7 @@ static u32 qed_grc_dump_special_regs(struct qed_hwfn *p_hwfn, u32 offset = 0, addr; offset += qed_grc_dump_regs_hdr(dump_buf, - dump, 2, SPLIT_TYPE_NONE, 0, - NULL, NULL); + dump, 2, SPLIT_TYPE_NONE, 0, "REGS"); /* Dump R/TDIF_REG_DEBUG_ERROR_INFO_SIZE (every 8'th register should be * skipped). @@ -3049,8 +2392,7 @@ static u32 qed_grc_dump_mem_hdr(struct qed_hwfn *p_hwfn, u32 len, u32 bit_width, bool packed, - const char *mem_group, - bool is_storm, char storm_letter) + const char *mem_group, char storm_letter) { u8 num_params = 3; u32 offset = 0; @@ -3071,7 +2413,7 @@ static u32 qed_grc_dump_mem_hdr(struct qed_hwfn *p_hwfn, if (name) { /* Dump name */ - if (is_storm) { + if (storm_letter) { strcpy(buf, "?STORM_"); buf[0] = storm_letter; strcpy(buf + strlen(buf), name); @@ -3103,7 +2445,7 @@ static u32 qed_grc_dump_mem_hdr(struct qed_hwfn *p_hwfn, dump, "packed", 1); /* Dump reg type */ - if (is_storm) { + if (storm_letter) { strcpy(buf, "?STORM_"); buf[0] = storm_letter; strcpy(buf + strlen(buf), mem_group); @@ -3130,8 +2472,7 @@ static u32 qed_grc_dump_mem(struct qed_hwfn *p_hwfn, bool wide_bus, u32 bit_width, bool packed, - const char *mem_group, - bool is_storm, char storm_letter) + const char *mem_group, char storm_letter) { u32 offset = 0; @@ -3142,8 +2483,7 @@ static u32 qed_grc_dump_mem(struct qed_hwfn *p_hwfn, addr, len, bit_width, - packed, - mem_group, is_storm, storm_letter); + packed, mem_group, storm_letter); offset += qed_grc_dump_addr_range(p_hwfn, p_ptt, dump_buf + offset, @@ -3156,20 +2496,21 @@ static u32 qed_grc_dump_mem(struct qed_hwfn *p_hwfn, /* Dumps GRC memories entries. Returns the dumped size in dwords. */ static u32 qed_grc_dump_mem_entries(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, - struct dbg_array input_mems_arr, + struct virt_mem_desc input_mems_arr, u32 *dump_buf, bool dump) { u32 i, offset = 0, input_offset = 0; bool mode_match = true; - while (input_offset < input_mems_arr.size_in_dwords) { + while (input_offset < BYTES_TO_DWORDS(input_mems_arr.size)) { const struct dbg_dump_cond_hdr *cond_hdr; u16 modes_buf_offset; u32 num_entries; bool eval_mode; - cond_hdr = (const struct dbg_dump_cond_hdr *) - &input_mems_arr.ptr[input_offset++]; + cond_hdr = + (const struct dbg_dump_cond_hdr *)input_mems_arr.ptr + + input_offset++; num_entries = cond_hdr->data_size / MEM_DUMP_ENTRY_SIZE_DWORDS; /* Check required mode */ @@ -3191,24 +2532,25 @@ static u32 qed_grc_dump_mem_entries(struct qed_hwfn *p_hwfn, for (i = 0; i < num_entries; i++, input_offset += MEM_DUMP_ENTRY_SIZE_DWORDS) { const struct dbg_dump_mem *mem = - (const struct dbg_dump_mem *) - &input_mems_arr.ptr[input_offset]; - u8 mem_group_id = GET_FIELD(mem->dword0, - DBG_DUMP_MEM_MEM_GROUP_ID); - bool is_storm = false, mem_wide_bus; - enum dbg_grc_params grc_param; - char storm_letter = 'a'; - enum block_id block_id; + (const struct dbg_dump_mem *)((u32 *) + input_mems_arr.ptr + + input_offset); + const struct dbg_block *block; + char storm_letter = 0; u32 mem_addr, mem_len; + bool mem_wide_bus; + u8 mem_group_id; + mem_group_id = GET_FIELD(mem->dword0, + DBG_DUMP_MEM_MEM_GROUP_ID); if (mem_group_id >= MEM_GROUPS_NUM) { DP_NOTICE(p_hwfn, "Invalid mem_group_id\n"); return 0; } - block_id = (enum block_id)cond_hdr->block_id; if (!qed_grc_is_mem_included(p_hwfn, - block_id, + (enum block_id) + cond_hdr->block_id, mem_group_id)) continue; @@ -3217,42 +2559,14 @@ static u32 qed_grc_dump_mem_entries(struct qed_hwfn *p_hwfn, mem_wide_bus = GET_FIELD(mem->dword1, DBG_DUMP_MEM_WIDE_BUS); - /* Update memory length for CCFC/TCFC memories - * according to number of LCIDs/LTIDs. - */ - if (mem_group_id == MEM_GROUP_CONN_CFC_MEM) { - if (mem_len % MAX_LCIDS) { - DP_NOTICE(p_hwfn, - "Invalid CCFC connection memory size\n"); - return 0; - } - - grc_param = DBG_GRC_PARAM_NUM_LCIDS; - mem_len = qed_grc_get_param(p_hwfn, grc_param) * - (mem_len / MAX_LCIDS); - } else if (mem_group_id == MEM_GROUP_TASK_CFC_MEM) { - if (mem_len % MAX_LTIDS) { - DP_NOTICE(p_hwfn, - "Invalid TCFC task memory size\n"); - return 0; - } - - grc_param = DBG_GRC_PARAM_NUM_LTIDS; - mem_len = qed_grc_get_param(p_hwfn, grc_param) * - (mem_len / MAX_LTIDS); - } + block = get_dbg_block(p_hwfn, + cond_hdr->block_id); - /* If memory is associated with Storm, update Storm - * details. + /* If memory is associated with Storm, + * update storm details */ - if (s_block_defs - [cond_hdr->block_id]->associated_to_storm) { - is_storm = true; - storm_letter = - s_storm_defs[s_block_defs - [cond_hdr->block_id]-> - storm_id].letter; - } + if (block->associated_storm_letter) + storm_letter = block->associated_storm_letter; /* Dump memory */ offset += qed_grc_dump_mem(p_hwfn, @@ -3266,7 +2580,6 @@ static u32 qed_grc_dump_mem_entries(struct qed_hwfn *p_hwfn, 0, false, s_mem_group_names[mem_group_id], - is_storm, storm_letter); } } @@ -3281,26 +2594,25 @@ static u32 qed_grc_dump_memories(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { + struct virt_mem_desc *dbg_buf = + &p_hwfn->dbg_arrays[BIN_BUF_DBG_DUMP_MEM]; u32 offset = 0, input_offset = 0; - while (input_offset < - s_dbg_arrays[BIN_BUF_DBG_DUMP_MEM].size_in_dwords) { + while (input_offset < BYTES_TO_DWORDS(dbg_buf->size)) { const struct dbg_dump_split_hdr *split_hdr; - struct dbg_array curr_input_mems_arr; + struct virt_mem_desc curr_input_mems_arr; enum init_split_types split_type; u32 split_data_size; - split_hdr = (const struct dbg_dump_split_hdr *) - &s_dbg_arrays[BIN_BUF_DBG_DUMP_MEM].ptr[input_offset++]; - split_type = - GET_FIELD(split_hdr->hdr, - DBG_DUMP_SPLIT_HDR_SPLIT_TYPE_ID); - split_data_size = - GET_FIELD(split_hdr->hdr, - DBG_DUMP_SPLIT_HDR_DATA_SIZE); - curr_input_mems_arr.ptr = - &s_dbg_arrays[BIN_BUF_DBG_DUMP_MEM].ptr[input_offset]; - curr_input_mems_arr.size_in_dwords = split_data_size; + split_hdr = + (const struct dbg_dump_split_hdr *)dbg_buf->ptr + + input_offset++; + split_type = GET_FIELD(split_hdr->hdr, + DBG_DUMP_SPLIT_HDR_SPLIT_TYPE_ID); + split_data_size = GET_FIELD(split_hdr->hdr, + DBG_DUMP_SPLIT_HDR_DATA_SIZE); + curr_input_mems_arr.ptr = (u32 *)dbg_buf->ptr + input_offset; + curr_input_mems_arr.size = DWORDS_TO_BYTES(split_data_size); if (split_type == SPLIT_TYPE_NONE) offset += qed_grc_dump_mem_entries(p_hwfn, @@ -3328,17 +2640,19 @@ static u32 qed_grc_dump_ctx_data(struct qed_hwfn *p_hwfn, bool dump, const char *name, u32 num_lids, - u32 lid_size, - u32 rd_reg_addr, - u8 storm_id) + enum cm_ctx_types ctx_type, u8 storm_id) { + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; struct storm_defs *storm = &s_storm_defs[storm_id]; - u32 i, lid, total_size, offset = 0; + u32 i, lid, lid_size, total_size; + u32 rd_reg_addr, offset = 0; + + /* Convert quad-regs to dwords */ + lid_size = storm->cm_ctx_lid_sizes[dev_data->chip_id][ctx_type] * 4; if (!lid_size) return 0; - lid_size *= BYTES_IN_DWORD; total_size = num_lids * lid_size; offset += qed_grc_dump_mem_hdr(p_hwfn, @@ -3348,18 +2662,26 @@ static u32 qed_grc_dump_ctx_data(struct qed_hwfn *p_hwfn, 0, total_size, lid_size * 32, - false, name, true, storm->letter); + false, name, storm->letter); if (!dump) return offset + total_size; + rd_reg_addr = BYTES_TO_DWORDS(storm->cm_ctx_rd_addr[ctx_type]); + /* Dump context data */ for (lid = 0; lid < num_lids; lid++) { - for (i = 0; i < lid_size; i++, offset++) { + for (i = 0; i < lid_size; i++) { qed_wr(p_hwfn, p_ptt, storm->cm_ctx_wr_addr, (i << 9) | lid); - *(dump_buf + offset) = qed_rd(p_hwfn, - p_ptt, rd_reg_addr); + offset += qed_grc_dump_addr_range(p_hwfn, + p_ptt, + dump_buf + offset, + dump, + rd_reg_addr, + 1, + false, + SPLIT_TYPE_NONE, 0); } } @@ -3370,115 +2692,126 @@ static u32 qed_grc_dump_ctx_data(struct qed_hwfn *p_hwfn, static u32 qed_grc_dump_ctx(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { - enum dbg_grc_params grc_param; u32 offset = 0; u8 storm_id; for (storm_id = 0; storm_id < MAX_DBG_STORMS; storm_id++) { - struct storm_defs *storm = &s_storm_defs[storm_id]; - if (!qed_grc_is_storm_included(p_hwfn, (enum dbg_storms)storm_id)) continue; /* Dump Conn AG context size */ - grc_param = DBG_GRC_PARAM_NUM_LCIDS; - offset += - qed_grc_dump_ctx_data(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - "CONN_AG_CTX", - qed_grc_get_param(p_hwfn, - grc_param), - storm->cm_conn_ag_ctx_lid_size, - storm->cm_conn_ag_ctx_rd_addr, - storm_id); + offset += qed_grc_dump_ctx_data(p_hwfn, + p_ptt, + dump_buf + offset, + dump, + "CONN_AG_CTX", + NUM_OF_LCIDS, + CM_CTX_CONN_AG, storm_id); /* Dump Conn ST context size */ - grc_param = DBG_GRC_PARAM_NUM_LCIDS; - offset += - qed_grc_dump_ctx_data(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - "CONN_ST_CTX", - qed_grc_get_param(p_hwfn, - grc_param), - storm->cm_conn_st_ctx_lid_size, - storm->cm_conn_st_ctx_rd_addr, - storm_id); + offset += qed_grc_dump_ctx_data(p_hwfn, + p_ptt, + dump_buf + offset, + dump, + "CONN_ST_CTX", + NUM_OF_LCIDS, + CM_CTX_CONN_ST, storm_id); /* Dump Task AG context size */ - grc_param = DBG_GRC_PARAM_NUM_LTIDS; - offset += - qed_grc_dump_ctx_data(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - "TASK_AG_CTX", - qed_grc_get_param(p_hwfn, - grc_param), - storm->cm_task_ag_ctx_lid_size, - storm->cm_task_ag_ctx_rd_addr, - storm_id); + offset += qed_grc_dump_ctx_data(p_hwfn, + p_ptt, + dump_buf + offset, + dump, + "TASK_AG_CTX", + NUM_OF_LTIDS, + CM_CTX_TASK_AG, storm_id); /* Dump Task ST context size */ - grc_param = DBG_GRC_PARAM_NUM_LTIDS; - offset += - qed_grc_dump_ctx_data(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - "TASK_ST_CTX", - qed_grc_get_param(p_hwfn, - grc_param), - storm->cm_task_st_ctx_lid_size, - storm->cm_task_st_ctx_rd_addr, - storm_id); + offset += qed_grc_dump_ctx_data(p_hwfn, + p_ptt, + dump_buf + offset, + dump, + "TASK_ST_CTX", + NUM_OF_LTIDS, + CM_CTX_TASK_ST, storm_id); } return offset; } -/* Dumps GRC IORs data. Returns the dumped size in dwords. */ -static u32 qed_grc_dump_iors(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) -{ - char buf[10] = "IOR_SET_?"; - u32 addr, offset = 0; - u8 storm_id, set_id; +#define VFC_STATUS_RESP_READY_BIT 0 +#define VFC_STATUS_BUSY_BIT 1 +#define VFC_STATUS_SENDING_CMD_BIT 2 - for (storm_id = 0; storm_id < MAX_DBG_STORMS; storm_id++) { - struct storm_defs *storm = &s_storm_defs[storm_id]; +#define VFC_POLLING_DELAY_MS 1 +#define VFC_POLLING_COUNT 20 - if (!qed_grc_is_storm_included(p_hwfn, - (enum dbg_storms)storm_id)) - continue; +/* Reads data from VFC. Returns the number of dwords read (0 on error). + * Sizes are specified in dwords. + */ +static u32 qed_grc_dump_read_from_vfc(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + struct storm_defs *storm, + u32 *cmd_data, + u32 cmd_size, + u32 *addr_data, + u32 addr_size, + u32 resp_size, u32 *dump_buf) +{ + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; + u32 vfc_status, polling_ms, polling_count = 0, i; + u32 reg_addr, sem_base; + bool is_ready = false; + + sem_base = storm->sem_fast_mem_addr; + polling_ms = VFC_POLLING_DELAY_MS * + s_hw_type_defs[dev_data->hw_type].delay_factor; + + /* Write VFC command */ + ARR_REG_WR(p_hwfn, + p_ptt, + sem_base + SEM_FAST_REG_VFC_DATA_WR, + cmd_data, cmd_size); + + /* Write VFC address */ + ARR_REG_WR(p_hwfn, + p_ptt, + sem_base + SEM_FAST_REG_VFC_ADDR, + addr_data, addr_size); + + /* Read response */ + for (i = 0; i < resp_size; i++) { + /* Poll until ready */ + do { + reg_addr = sem_base + SEM_FAST_REG_VFC_STATUS; + qed_grc_dump_addr_range(p_hwfn, + p_ptt, + &vfc_status, + true, + BYTES_TO_DWORDS(reg_addr), + 1, + false, SPLIT_TYPE_NONE, 0); + is_ready = vfc_status & BIT(VFC_STATUS_RESP_READY_BIT); + + if (!is_ready) { + if (polling_count++ == VFC_POLLING_COUNT) + return 0; - for (set_id = 0; set_id < NUM_IOR_SETS; set_id++) { - addr = BYTES_TO_DWORDS(storm->sem_fast_mem_addr + - SEM_FAST_REG_STORM_REG_FILE) + - IOR_SET_OFFSET(set_id); - if (strlen(buf) > 0) - buf[strlen(buf) - 1] = '0' + set_id; - offset += qed_grc_dump_mem(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - buf, - addr, - IORS_PER_SET, - false, - 32, - false, - "ior", - true, - storm->letter); - } + msleep(polling_ms); + } + } while (!is_ready); + + reg_addr = sem_base + SEM_FAST_REG_VFC_DATA_RD; + qed_grc_dump_addr_range(p_hwfn, + p_ptt, + dump_buf + i, + true, + BYTES_TO_DWORDS(reg_addr), + 1, false, SPLIT_TYPE_NONE, 0); } - return offset; + return resp_size; } /* Dump VFC CAM. Returns the dumped size in dwords. */ @@ -3490,7 +2823,7 @@ static u32 qed_grc_dump_vfc_cam(struct qed_hwfn *p_hwfn, struct storm_defs *storm = &s_storm_defs[storm_id]; u32 cam_addr[VFC_CAM_ADDR_DWORDS] = { 0 }; u32 cam_cmd[VFC_CAM_CMD_DWORDS] = { 0 }; - u32 row, i, offset = 0; + u32 row, offset = 0; offset += qed_grc_dump_mem_hdr(p_hwfn, dump_buf + offset, @@ -3499,7 +2832,7 @@ static u32 qed_grc_dump_vfc_cam(struct qed_hwfn *p_hwfn, 0, total_size, 256, - false, "vfc_cam", true, storm->letter); + false, "vfc_cam", storm->letter); if (!dump) return offset + total_size; @@ -3507,26 +2840,18 @@ static u32 qed_grc_dump_vfc_cam(struct qed_hwfn *p_hwfn, /* Prepare CAM address */ SET_VAR_FIELD(cam_addr, VFC_CAM_ADDR, OP, VFC_OPCODE_CAM_RD); - for (row = 0; row < VFC_CAM_NUM_ROWS; - row++, offset += VFC_CAM_RESP_DWORDS) { - /* Write VFC CAM command */ + /* Read VFC CAM data */ + for (row = 0; row < VFC_CAM_NUM_ROWS; row++) { SET_VAR_FIELD(cam_cmd, VFC_CAM_CMD, ROW, row); - ARR_REG_WR(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_DATA_WR, - cam_cmd, VFC_CAM_CMD_DWORDS); - - /* Write VFC CAM address */ - ARR_REG_WR(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_ADDR, - cam_addr, VFC_CAM_ADDR_DWORDS); - - /* Read VFC CAM read response */ - ARR_REG_RD(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_DATA_RD, - dump_buf + offset, VFC_CAM_RESP_DWORDS); + offset += qed_grc_dump_read_from_vfc(p_hwfn, + p_ptt, + storm, + cam_cmd, + VFC_CAM_CMD_DWORDS, + cam_addr, + VFC_CAM_ADDR_DWORDS, + VFC_CAM_RESP_DWORDS, + dump_buf + offset); } return offset; @@ -3543,7 +2868,7 @@ static u32 qed_grc_dump_vfc_ram(struct qed_hwfn *p_hwfn, struct storm_defs *storm = &s_storm_defs[storm_id]; u32 ram_addr[VFC_RAM_ADDR_DWORDS] = { 0 }; u32 ram_cmd[VFC_RAM_CMD_DWORDS] = { 0 }; - u32 row, i, offset = 0; + u32 row, offset = 0; offset += qed_grc_dump_mem_hdr(p_hwfn, dump_buf + offset, @@ -3554,35 +2879,27 @@ static u32 qed_grc_dump_vfc_ram(struct qed_hwfn *p_hwfn, 256, false, ram_defs->type_name, - true, storm->letter); - - /* Prepare RAM address */ - SET_VAR_FIELD(ram_addr, VFC_RAM_ADDR, OP, VFC_OPCODE_RAM_RD); + storm->letter); if (!dump) return offset + total_size; + /* Prepare RAM address */ + SET_VAR_FIELD(ram_addr, VFC_RAM_ADDR, OP, VFC_OPCODE_RAM_RD); + + /* Read VFC RAM data */ for (row = ram_defs->base_row; - row < ram_defs->base_row + ram_defs->num_rows; - row++, offset += VFC_RAM_RESP_DWORDS) { - /* Write VFC RAM command */ - ARR_REG_WR(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_DATA_WR, - ram_cmd, VFC_RAM_CMD_DWORDS); - - /* Write VFC RAM address */ + row < ram_defs->base_row + ram_defs->num_rows; row++) { SET_VAR_FIELD(ram_addr, VFC_RAM_ADDR, ROW, row); - ARR_REG_WR(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_ADDR, - ram_addr, VFC_RAM_ADDR_DWORDS); - - /* Read VFC RAM read response */ - ARR_REG_RD(p_hwfn, - p_ptt, - storm->sem_fast_mem_addr + SEM_FAST_REG_VFC_DATA_RD, - dump_buf + offset, VFC_RAM_RESP_DWORDS); + offset += qed_grc_dump_read_from_vfc(p_hwfn, + p_ptt, + storm, + ram_cmd, + VFC_RAM_CMD_DWORDS, + ram_addr, + VFC_RAM_ADDR_DWORDS, + VFC_RAM_RESP_DWORDS, + dump_buf + offset); } return offset; @@ -3592,16 +2909,13 @@ static u32 qed_grc_dump_vfc_ram(struct qed_hwfn *p_hwfn, static u32 qed_grc_dump_vfc(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { - struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; u8 storm_id, i; u32 offset = 0; for (storm_id = 0; storm_id < MAX_DBG_STORMS; storm_id++) { if (!qed_grc_is_storm_included(p_hwfn, (enum dbg_storms)storm_id) || - !s_storm_defs[storm_id].has_vfc || - (storm_id == DBG_PSTORM_ID && dev_data->platform_id != - PLATFORM_ASIC)) + !s_storm_defs[storm_id].has_vfc) continue; /* Read CAM */ @@ -3651,7 +2965,7 @@ static u32 qed_grc_dump_rss(struct qed_hwfn *p_hwfn, total_dwords, rss_defs->entry_width, packed, - rss_defs->type_name, false, 0); + rss_defs->type_name, 0); /* Dump RSS data */ if (!dump) { @@ -3711,7 +3025,7 @@ static u32 qed_grc_dump_big_ram(struct qed_hwfn *p_hwfn, 0, ram_size, block_size * 8, - false, type_name, false, 0); + false, type_name, 0); /* Read and dump Big RAM data */ if (!dump) @@ -3737,6 +3051,7 @@ static u32 qed_grc_dump_big_ram(struct qed_hwfn *p_hwfn, return offset; } +/* Dumps MCP scratchpad. Returns the dumped size in dwords. */ static u32 qed_grc_dump_mcp(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { @@ -3758,8 +3073,8 @@ static u32 qed_grc_dump_mcp(struct qed_hwfn *p_hwfn, dump, NULL, BYTES_TO_DWORDS(MCP_REG_SCRATCH), - MCP_REG_SCRATCH_SIZE_BB_K2, - false, 0, false, "MCP", false, 0); + MCP_REG_SCRATCH_SIZE, + false, 0, false, "MCP", 0); /* Dump MCP cpu_reg_file */ offset += qed_grc_dump_mem(p_hwfn, @@ -3769,19 +3084,19 @@ static u32 qed_grc_dump_mcp(struct qed_hwfn *p_hwfn, NULL, BYTES_TO_DWORDS(MCP_REG_CPU_REG_FILE), MCP_REG_CPU_REG_FILE_SIZE, - false, 0, false, "MCP", false, 0); + false, 0, false, "MCP", 0); /* Dump MCP registers */ block_enable[BLOCK_MCP] = true; offset += qed_grc_dump_registers(p_hwfn, p_ptt, dump_buf + offset, - dump, block_enable, "block", "MCP"); + dump, block_enable, "MCP"); /* Dump required non-MCP registers */ offset += qed_grc_dump_regs_hdr(dump_buf + offset, dump, 1, SPLIT_TYPE_NONE, 0, - "block", "MCP"); + "MCP"); addr = BYTES_TO_DWORDS(MISC_REG_SHARED_MEM_ADDR); offset += qed_grc_dump_reg_entry(p_hwfn, p_ptt, @@ -3798,7 +3113,9 @@ static u32 qed_grc_dump_mcp(struct qed_hwfn *p_hwfn, return offset; } -/* Dumps the tbus indirect memory for all PHYs. */ +/* Dumps the tbus indirect memory for all PHYs. + * Returns the dumped size in dwords. + */ static u32 qed_grc_dump_phy(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { @@ -3832,7 +3149,7 @@ static u32 qed_grc_dump_phy(struct qed_hwfn *p_hwfn, mem_name, 0, PHY_DUMP_SIZE_DWORDS, - 16, true, mem_name, false, 0); + 16, true, mem_name, 0); if (!dump) { offset += PHY_DUMP_SIZE_DWORDS; @@ -3863,21 +3180,58 @@ static u32 qed_grc_dump_phy(struct qed_hwfn *p_hwfn, return offset; } -static void qed_config_dbg_line(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, - enum block_id block_id, - u8 line_id, - u8 enable_mask, - u8 right_shift, - u8 force_valid_mask, u8 force_frame_mask) +static enum dbg_status qed_find_nvram_image(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 image_type, + u32 *nvram_offset_bytes, + u32 *nvram_size_bytes); + +static enum dbg_status qed_nvram_read(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 nvram_offset_bytes, + u32 nvram_size_bytes, u32 *ret_buf); + +/* Dumps the MCP HW dump from NVRAM. Returns the dumped size in dwords. */ +static u32 qed_grc_dump_mcp_hw_dump(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 *dump_buf, bool dump) { - struct block_defs *block = s_block_defs[block_id]; + u32 hw_dump_offset_bytes = 0, hw_dump_size_bytes = 0; + u32 hw_dump_size_dwords = 0, offset = 0; + enum dbg_status status; - qed_wr(p_hwfn, p_ptt, block->dbg_select_addr, line_id); - qed_wr(p_hwfn, p_ptt, block->dbg_enable_addr, enable_mask); - qed_wr(p_hwfn, p_ptt, block->dbg_shift_addr, right_shift); - qed_wr(p_hwfn, p_ptt, block->dbg_force_valid_addr, force_valid_mask); - qed_wr(p_hwfn, p_ptt, block->dbg_force_frame_addr, force_frame_mask); + /* Read HW dump image from NVRAM */ + status = qed_find_nvram_image(p_hwfn, + p_ptt, + NVM_TYPE_HW_DUMP_OUT, + &hw_dump_offset_bytes, + &hw_dump_size_bytes); + if (status != DBG_STATUS_OK) + return 0; + + hw_dump_size_dwords = BYTES_TO_DWORDS(hw_dump_size_bytes); + + /* Dump HW dump image section */ + offset += qed_dump_section_hdr(dump_buf + offset, + dump, "mcp_hw_dump", 1); + offset += qed_dump_num_param(dump_buf + offset, + dump, "size", hw_dump_size_dwords); + + /* Read MCP HW dump image into dump buffer */ + if (dump && hw_dump_size_dwords) { + status = qed_nvram_read(p_hwfn, + p_ptt, + hw_dump_offset_bytes, + hw_dump_size_bytes, dump_buf + offset); + if (status != DBG_STATUS_OK) { + DP_NOTICE(p_hwfn, + "Failed to read MCP HW Dump image from NVRAM\n"); + return 0; + } + } + offset += hw_dump_size_dwords; + + return offset; } /* Dumps Static Debug data. Returns the dumped size in dwords. */ @@ -3886,26 +3240,19 @@ static u32 qed_grc_dump_static_debug(struct qed_hwfn *p_hwfn, u32 *dump_buf, bool dump) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - u32 block_id, line_id, offset = 0; + u32 block_id, line_id, offset = 0, addr, len; /* Don't dump static debug if a debug bus recording is in progress */ if (dump && qed_rd(p_hwfn, p_ptt, DBG_REG_DBG_BLOCK_ON)) return 0; if (dump) { - /* Disable all blocks debug output */ - for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { - struct block_defs *block = s_block_defs[block_id]; - - if (block->dbg_client_id[dev_data->chip_id] != - MAX_DBG_BUS_CLIENTS) - qed_wr(p_hwfn, p_ptt, block->dbg_enable_addr, - 0); - } + /* Disable debug bus in all blocks */ + qed_bus_disable_blocks(p_hwfn, p_ptt); qed_bus_reset_dbg_block(p_hwfn, p_ptt); - qed_bus_set_framing_mode(p_hwfn, - p_ptt, DBG_BUS_FRAME_MODE_8HW_0ST); + qed_wr(p_hwfn, + p_ptt, DBG_REG_FRAMING_MODE, DBG_BUS_FRAME_MODE_8HW); qed_wr(p_hwfn, p_ptt, DBG_REG_DEBUG_TARGET, DBG_BUS_TARGET_ID_INT_BUF); qed_wr(p_hwfn, p_ptt, DBG_REG_FULL_MODE, 1); @@ -3914,28 +3261,48 @@ static u32 qed_grc_dump_static_debug(struct qed_hwfn *p_hwfn, /* Dump all static debug lines for each relevant block */ for (block_id = 0; block_id < MAX_BLOCK_ID; block_id++) { - struct block_defs *block = s_block_defs[block_id]; - struct dbg_bus_block *block_desc; - u32 block_dwords, addr, len; - u8 dbg_client_id; + const struct dbg_block_chip *block_per_chip; + const struct dbg_block *block; + bool is_removed, has_dbg_bus; + u16 modes_buf_offset; + u32 block_dwords; + + block_per_chip = + qed_get_dbg_block_per_chip(p_hwfn, (enum block_id)block_id); + is_removed = GET_FIELD(block_per_chip->flags, + DBG_BLOCK_CHIP_IS_REMOVED); + has_dbg_bus = GET_FIELD(block_per_chip->flags, + DBG_BLOCK_CHIP_HAS_DBG_BUS); - if (block->dbg_client_id[dev_data->chip_id] == - MAX_DBG_BUS_CLIENTS) + /* read+clear for NWS parity is not working, skip NWS block */ + if (block_id == BLOCK_NWS) continue; - block_desc = get_dbg_bus_block_desc(p_hwfn, - (enum block_id)block_id); - block_dwords = NUM_DBG_LINES(block_desc) * + if (!is_removed && has_dbg_bus && + GET_FIELD(block_per_chip->dbg_bus_mode.data, + DBG_MODE_HDR_EVAL_MODE) > 0) { + modes_buf_offset = + GET_FIELD(block_per_chip->dbg_bus_mode.data, + DBG_MODE_HDR_MODES_BUF_OFFSET); + if (!qed_is_mode_match(p_hwfn, &modes_buf_offset)) + has_dbg_bus = false; + } + + if (is_removed || !has_dbg_bus) + continue; + + block_dwords = NUM_DBG_LINES(block_per_chip) * STATIC_DEBUG_LINE_DWORDS; /* Dump static section params */ + block = get_dbg_block(p_hwfn, (enum block_id)block_id); offset += qed_grc_dump_mem_hdr(p_hwfn, dump_buf + offset, dump, block->name, 0, block_dwords, - 32, false, "STATIC", false, 0); + 32, false, "STATIC", 0); if (!dump) { offset += block_dwords; @@ -3951,20 +3318,19 @@ static u32 qed_grc_dump_static_debug(struct qed_hwfn *p_hwfn, } /* Enable block's client */ - dbg_client_id = block->dbg_client_id[dev_data->chip_id]; qed_bus_enable_clients(p_hwfn, p_ptt, - BIT(dbg_client_id)); + BIT(block_per_chip->dbg_client_id)); addr = BYTES_TO_DWORDS(DBG_REG_CALENDAR_OUT_DATA); len = STATIC_DEBUG_LINE_DWORDS; - for (line_id = 0; line_id < (u32)NUM_DBG_LINES(block_desc); + for (line_id = 0; line_id < (u32)NUM_DBG_LINES(block_per_chip); line_id++) { /* Configure debug line ID */ - qed_config_dbg_line(p_hwfn, - p_ptt, - (enum block_id)block_id, - (u8)line_id, 0xf, 0, 0, 0); + qed_bus_config_dbg_line(p_hwfn, + p_ptt, + (enum block_id)block_id, + (u8)line_id, 0xf, 0, 0, 0); /* Read debug line info */ offset += qed_grc_dump_addr_range(p_hwfn, @@ -3979,7 +3345,8 @@ static u32 qed_grc_dump_static_debug(struct qed_hwfn *p_hwfn, /* Disable block's client and debug output */ qed_bus_enable_clients(p_hwfn, p_ptt, 0); - qed_wr(p_hwfn, p_ptt, block->dbg_enable_addr, 0); + qed_bus_config_dbg_line(p_hwfn, p_ptt, + (enum block_id)block_id, 0, 0, 0, 0, 0); } if (dump) { @@ -3999,8 +3366,8 @@ static enum dbg_status qed_grc_dump(struct qed_hwfn *p_hwfn, bool dump, u32 *num_dumped_dwords) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; + u32 dwords_read, offset = 0; bool parities_masked = false; - u32 offset = 0; u8 i; *num_dumped_dwords = 0; @@ -4019,13 +3386,11 @@ static enum dbg_status qed_grc_dump(struct qed_hwfn *p_hwfn, offset += qed_dump_num_param(dump_buf + offset, dump, "num-lcids", - qed_grc_get_param(p_hwfn, - DBG_GRC_PARAM_NUM_LCIDS)); + NUM_OF_LCIDS); offset += qed_dump_num_param(dump_buf + offset, dump, "num-ltids", - qed_grc_get_param(p_hwfn, - DBG_GRC_PARAM_NUM_LTIDS)); + NUM_OF_LTIDS); offset += qed_dump_num_param(dump_buf + offset, dump, "num-ports", dev_data->num_ports); @@ -4037,7 +3402,7 @@ static enum dbg_status qed_grc_dump(struct qed_hwfn *p_hwfn, /* Take all blocks out of reset (using reset registers) */ if (dump) { - qed_grc_unreset_blocks(p_hwfn, p_ptt); + qed_grc_unreset_blocks(p_hwfn, p_ptt, false); qed_update_blocks_reset_state(p_hwfn, p_ptt); } @@ -4080,7 +3445,7 @@ static enum dbg_status qed_grc_dump(struct qed_hwfn *p_hwfn, dump_buf + offset, dump, - block_enable, NULL, NULL); + block_enable, NULL); /* Dump special registers */ offset += qed_grc_dump_special_regs(p_hwfn, @@ -4114,23 +3479,29 @@ static enum dbg_status qed_grc_dump(struct qed_hwfn *p_hwfn, dump_buf + offset, dump, i); - /* Dump IORs */ - if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_IOR)) - offset += qed_grc_dump_iors(p_hwfn, - p_ptt, dump_buf + offset, dump); - /* Dump VFC */ - if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_VFC)) - offset += qed_grc_dump_vfc(p_hwfn, - p_ptt, dump_buf + offset, dump); + if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_VFC)) { + dwords_read = qed_grc_dump_vfc(p_hwfn, + p_ptt, dump_buf + offset, dump); + offset += dwords_read; + if (!dwords_read) + return DBG_STATUS_VFC_READ_ERROR; + } /* Dump PHY tbus */ if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_PHY) && dev_data->chip_id == - CHIP_K2 && dev_data->platform_id == PLATFORM_ASIC) + CHIP_K2 && dev_data->hw_type == HW_TYPE_ASIC) offset += qed_grc_dump_phy(p_hwfn, p_ptt, dump_buf + offset, dump); + /* Dump MCP HW Dump */ + if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_MCP_HW_DUMP) && + !qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_NO_MCP) && 1) + offset += qed_grc_dump_mcp_hw_dump(p_hwfn, + p_ptt, + dump_buf + offset, dump); + /* Dump static debug data (only if not during debug bus recording) */ if (qed_grc_is_included(p_hwfn, DBG_GRC_PARAM_DUMP_STATIC) && @@ -4181,8 +3552,9 @@ static u32 qed_idle_chk_dump_failure(struct qed_hwfn *p_hwfn, u8 reg_id; hdr = (struct dbg_idle_chk_result_hdr *)dump_buf; - regs = &((const union dbg_idle_chk_reg *) - s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr)[rule->reg_offset]; + regs = (const union dbg_idle_chk_reg *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr + + rule->reg_offset; cond_regs = ®s[0].cond_reg; info_regs = ®s[rule->num_cond_regs].info_reg; @@ -4202,8 +3574,8 @@ static u32 qed_idle_chk_dump_failure(struct qed_hwfn *p_hwfn, const struct dbg_idle_chk_cond_reg *reg = &cond_regs[reg_id]; struct dbg_idle_chk_result_reg_hdr *reg_hdr; - reg_hdr = (struct dbg_idle_chk_result_reg_hdr *) - (dump_buf + offset); + reg_hdr = + (struct dbg_idle_chk_result_reg_hdr *)(dump_buf + offset); /* Write register header */ if (!dump) { @@ -4320,12 +3692,13 @@ qed_idle_chk_dump_rule_entries(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, const u32 *imm_values; rule = &input_rules[i]; - regs = &((const union dbg_idle_chk_reg *) - s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr) - [rule->reg_offset]; + regs = (const union dbg_idle_chk_reg *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr + + rule->reg_offset; cond_regs = ®s[0].cond_reg; - imm_values = &s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_IMMS].ptr - [rule->imm_offset]; + imm_values = + (u32 *)p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_IMMS].ptr + + rule->imm_offset; /* Check if all condition register blocks are out of reset, and * find maximal number of entries (all condition registers that @@ -4443,10 +3816,12 @@ qed_idle_chk_dump_rule_entries(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, static u32 qed_idle_chk_dump(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) { - u32 num_failing_rules_offset, offset = 0, input_offset = 0; - u32 num_failing_rules = 0; + struct virt_mem_desc *dbg_buf = + &p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES]; + u32 num_failing_rules_offset, offset = 0, + input_offset = 0, num_failing_rules = 0; - /* Dump global params */ + /* Dump global params - 1 must match below amount of params */ offset += qed_dump_common_global_params(p_hwfn, p_ptt, dump_buf + offset, dump, 1); @@ -4458,12 +3833,10 @@ static u32 qed_idle_chk_dump(struct qed_hwfn *p_hwfn, num_failing_rules_offset = offset; offset += qed_dump_num_param(dump_buf + offset, dump, "num_rules", 0); - while (input_offset < - s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES].size_in_dwords) { + while (input_offset < BYTES_TO_DWORDS(dbg_buf->size)) { const struct dbg_idle_chk_cond_hdr *cond_hdr = - (const struct dbg_idle_chk_cond_hdr *) - &s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES].ptr - [input_offset++]; + (const struct dbg_idle_chk_cond_hdr *)dbg_buf->ptr + + input_offset++; bool eval_mode, mode_match = true; u32 curr_failing_rules; u16 modes_buf_offset; @@ -4480,16 +3853,21 @@ static u32 qed_idle_chk_dump(struct qed_hwfn *p_hwfn, } if (mode_match) { + const struct dbg_idle_chk_rule *rule = + (const struct dbg_idle_chk_rule *)((u32 *) + dbg_buf->ptr + + input_offset); + u32 num_input_rules = + cond_hdr->data_size / IDLE_CHK_RULE_SIZE_DWORDS; offset += qed_idle_chk_dump_rule_entries(p_hwfn, - p_ptt, - dump_buf + offset, - dump, - (const struct dbg_idle_chk_rule *) - &s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES]. - ptr[input_offset], - cond_hdr->data_size / IDLE_CHK_RULE_SIZE_DWORDS, - &curr_failing_rules); + p_ptt, + dump_buf + + offset, + dump, + rule, + num_input_rules, + &curr_failing_rules); num_failing_rules += curr_failing_rules; } @@ -4556,7 +3934,7 @@ static enum dbg_status qed_nvram_read(struct qed_hwfn *p_hwfn, { u32 ret_mcp_resp, ret_mcp_param, ret_read_size, bytes_to_copy; s32 bytes_left = nvram_size_bytes; - u32 read_offset = 0; + u32 read_offset = 0, param = 0; DP_VERBOSE(p_hwfn, QED_MSG_DEBUG, @@ -4569,14 +3947,14 @@ static enum dbg_status qed_nvram_read(struct qed_hwfn *p_hwfn, MCP_DRV_NVM_BUF_LEN) ? MCP_DRV_NVM_BUF_LEN : bytes_left; /* Call NVRAM read command */ + SET_MFW_FIELD(param, + DRV_MB_PARAM_NVM_OFFSET, + nvram_offset_bytes + read_offset); + SET_MFW_FIELD(param, DRV_MB_PARAM_NVM_LEN, bytes_to_copy); if (qed_mcp_nvm_rd_cmd(p_hwfn, p_ptt, - DRV_MSG_CODE_NVM_READ_NVRAM, - (nvram_offset_bytes + - read_offset) | - (bytes_to_copy << - DRV_MB_PARAM_NVM_LEN_OFFSET), - &ret_mcp_resp, &ret_mcp_param, - &ret_read_size, + DRV_MSG_CODE_NVM_READ_NVRAM, param, + &ret_mcp_resp, + &ret_mcp_param, &ret_read_size, (u32 *)((u8 *)ret_buf + read_offset))) return DBG_STATUS_NVRAM_READ_FAILED; @@ -4714,12 +4092,12 @@ static enum dbg_status qed_mcp_trace_dump(struct qed_hwfn *p_hwfn, u32 trace_meta_size_dwords = 0, running_bundle_id, offset = 0; u32 trace_meta_offset_bytes = 0, trace_meta_size_bytes = 0; enum dbg_status status; - bool mcp_access; int halted = 0; + bool use_mfw; *num_dumped_dwords = 0; - mcp_access = !qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_NO_MCP); + use_mfw = !qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_NO_MCP); /* Get trace data info */ status = qed_mcp_trace_get_data_info(p_hwfn, @@ -4740,7 +4118,7 @@ static enum dbg_status qed_mcp_trace_dump(struct qed_hwfn *p_hwfn, * consistent. if halt fails, MCP trace is taken anyway, with a small * risk that it may be corrupt. */ - if (dump && mcp_access) { + if (dump && use_mfw) { halted = !qed_mcp_halt(p_hwfn, p_ptt); if (!halted) DP_NOTICE(p_hwfn, "MCP halt failed!\n"); @@ -4780,17 +4158,15 @@ static enum dbg_status qed_mcp_trace_dump(struct qed_hwfn *p_hwfn, */ trace_meta_size_bytes = qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_MCP_TRACE_META_SIZE); - if ((!trace_meta_size_bytes || dump) && mcp_access) { + if ((!trace_meta_size_bytes || dump) && use_mfw) status = qed_mcp_trace_get_meta_info(p_hwfn, p_ptt, trace_data_size_bytes, &running_bundle_id, &trace_meta_offset_bytes, &trace_meta_size_bytes); - if (status == DBG_STATUS_OK) - trace_meta_size_dwords = - BYTES_TO_DWORDS(trace_meta_size_bytes); - } + if (status == DBG_STATUS_OK) + trace_meta_size_dwords = BYTES_TO_DWORDS(trace_meta_size_bytes); /* Dump trace meta size param */ offset += qed_dump_num_param(dump_buf + offset, @@ -4814,7 +4190,7 @@ static enum dbg_status qed_mcp_trace_dump(struct qed_hwfn *p_hwfn, /* If no mcp access, indicate that the dump doesn't contain the meta * data from NVRAM. */ - return mcp_access ? status : DBG_STATUS_NVRAM_GET_IMAGE_FAILED; + return use_mfw ? status : DBG_STATUS_NVRAM_GET_IMAGE_FAILED; } /* Dump GRC FIFO */ @@ -4992,16 +4368,18 @@ static enum dbg_status qed_protection_override_dump(struct qed_hwfn *p_hwfn, override_window_dwords = qed_rd(p_hwfn, p_ptt, GRC_REG_NUMBER_VALID_OVERRIDE_WINDOW) * PROTECTION_OVERRIDE_ELEMENT_DWORDS; - addr = BYTES_TO_DWORDS(GRC_REG_PROTECTION_OVERRIDE_WINDOW); - offset += qed_grc_dump_addr_range(p_hwfn, - p_ptt, - dump_buf + offset, - true, - addr, - override_window_dwords, - true, SPLIT_TYPE_NONE, 0); - qed_dump_num_param(dump_buf + size_param_offset, dump, "size", - override_window_dwords); + if (override_window_dwords) { + addr = BYTES_TO_DWORDS(GRC_REG_PROTECTION_OVERRIDE_WINDOW); + offset += qed_grc_dump_addr_range(p_hwfn, + p_ptt, + dump_buf + offset, + true, + addr, + override_window_dwords, + true, SPLIT_TYPE_NONE, 0); + qed_dump_num_param(dump_buf + size_param_offset, dump, "size", + override_window_dwords); + } out: /* Dump last section */ offset += qed_dump_last_section(dump_buf, offset, dump); @@ -5037,7 +4415,7 @@ static u32 qed_fw_asserts_dump(struct qed_hwfn *p_hwfn, struct storm_defs *storm = &s_storm_defs[storm_id]; u32 last_list_idx, addr; - if (dev_data->block_in_reset[storm->block_id]) + if (dev_data->block_in_reset[storm->sem_block_id]) continue; /* Read FW info for the current Storm */ @@ -5088,20 +4466,362 @@ static u32 qed_fw_asserts_dump(struct qed_hwfn *p_hwfn, return offset; } +/* Dumps the specified ILT pages to the specified buffer. + * Returns the dumped size in dwords. + */ +static u32 qed_ilt_dump_pages_range(u32 *dump_buf, + bool dump, + u32 start_page_id, + u32 num_pages, + struct phys_mem_desc *ilt_pages, + bool dump_page_ids) +{ + u32 page_id, end_page_id, offset = 0; + + if (num_pages == 0) + return offset; + + end_page_id = start_page_id + num_pages - 1; + + for (page_id = start_page_id; page_id <= end_page_id; page_id++) { + struct phys_mem_desc *mem_desc = &ilt_pages[page_id]; + + /** + * + * if (page_id >= ->p_cxt_mngr->ilt_shadow_size) + * break; + */ + + if (!ilt_pages[page_id].virt_addr) + continue; + + if (dump_page_ids) { + /* Copy page ID to dump buffer */ + if (dump) + *(dump_buf + offset) = page_id; + offset++; + } else { + /* Copy page memory to dump buffer */ + if (dump) + memcpy(dump_buf + offset, + mem_desc->virt_addr, mem_desc->size); + offset += BYTES_TO_DWORDS(mem_desc->size); + } + } + + return offset; +} + +/* Dumps a section containing the dumped ILT pages. + * Returns the dumped size in dwords. + */ +static u32 qed_ilt_dump_pages_section(struct qed_hwfn *p_hwfn, + u32 *dump_buf, + bool dump, + u32 valid_conn_pf_pages, + u32 valid_conn_vf_pages, + struct phys_mem_desc *ilt_pages, + bool dump_page_ids) +{ + struct qed_ilt_client_cfg *clients = p_hwfn->p_cxt_mngr->clients; + u32 pf_start_line, start_page_id, offset = 0; + u32 cdut_pf_init_pages, cdut_vf_init_pages; + u32 cdut_pf_work_pages, cdut_vf_work_pages; + u32 base_data_offset, size_param_offset; + u32 cdut_pf_pages, cdut_vf_pages; + const char *section_name; + u8 i; + + section_name = dump_page_ids ? "ilt_page_ids" : "ilt_page_mem"; + cdut_pf_init_pages = qed_get_cdut_num_pf_init_pages(p_hwfn); + cdut_vf_init_pages = qed_get_cdut_num_vf_init_pages(p_hwfn); + cdut_pf_work_pages = qed_get_cdut_num_pf_work_pages(p_hwfn); + cdut_vf_work_pages = qed_get_cdut_num_vf_work_pages(p_hwfn); + cdut_pf_pages = cdut_pf_init_pages + cdut_pf_work_pages; + cdut_vf_pages = cdut_vf_init_pages + cdut_vf_work_pages; + pf_start_line = p_hwfn->p_cxt_mngr->pf_start_line; + + offset += + qed_dump_section_hdr(dump_buf + offset, dump, section_name, 1); + + /* Dump size parameter (0 for now, overwritten with real size later) */ + size_param_offset = offset; + offset += qed_dump_num_param(dump_buf + offset, dump, "size", 0); + base_data_offset = offset; + + /* CDUC pages are ordered as follows: + * - PF pages - valid section (included in PF connection type mapping) + * - PF pages - invalid section (not dumped) + * - For each VF in the PF: + * - VF pages - valid section (included in VF connection type mapping) + * - VF pages - invalid section (not dumped) + */ + if (qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_DUMP_ILT_CDUC)) { + /* Dump connection PF pages */ + start_page_id = clients[ILT_CLI_CDUC].first.val - pf_start_line; + offset += qed_ilt_dump_pages_range(dump_buf + offset, + dump, + start_page_id, + valid_conn_pf_pages, + ilt_pages, dump_page_ids); + + /* Dump connection VF pages */ + start_page_id += clients[ILT_CLI_CDUC].pf_total_lines; + for (i = 0; i < p_hwfn->p_cxt_mngr->vf_count; + i++, start_page_id += clients[ILT_CLI_CDUC].vf_total_lines) + offset += qed_ilt_dump_pages_range(dump_buf + offset, + dump, + start_page_id, + valid_conn_vf_pages, + ilt_pages, + dump_page_ids); + } + + /* CDUT pages are ordered as follows: + * - PF init pages (not dumped) + * - PF work pages + * - For each VF in the PF: + * - VF init pages (not dumped) + * - VF work pages + */ + if (qed_grc_get_param(p_hwfn, DBG_GRC_PARAM_DUMP_ILT_CDUT)) { + /* Dump task PF pages */ + start_page_id = clients[ILT_CLI_CDUT].first.val + + cdut_pf_init_pages - pf_start_line; + offset += qed_ilt_dump_pages_range(dump_buf + offset, + dump, + start_page_id, + cdut_pf_work_pages, + ilt_pages, dump_page_ids); + + /* Dump task VF pages */ + start_page_id = clients[ILT_CLI_CDUT].first.val + + cdut_pf_pages + cdut_vf_init_pages - pf_start_line; + for (i = 0; i < p_hwfn->p_cxt_mngr->vf_count; + i++, start_page_id += cdut_vf_pages) + offset += qed_ilt_dump_pages_range(dump_buf + offset, + dump, + start_page_id, + cdut_vf_work_pages, + ilt_pages, + dump_page_ids); + } + + /* Overwrite size param */ + if (dump) + qed_dump_num_param(dump_buf + size_param_offset, + dump, "size", offset - base_data_offset); + + return offset; +} + +/* Performs ILT Dump to the specified buffer. + * Returns the dumped size in dwords. + */ +static u32 qed_ilt_dump(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, u32 *dump_buf, bool dump) +{ + struct qed_ilt_client_cfg *clients = p_hwfn->p_cxt_mngr->clients; + u32 valid_conn_vf_cids, valid_conn_vf_pages, offset = 0; + u32 valid_conn_pf_cids, valid_conn_pf_pages, num_pages; + u32 num_cids_per_page, conn_ctx_size; + u32 cduc_page_size, cdut_page_size; + struct phys_mem_desc *ilt_pages; + u8 conn_type; + + cduc_page_size = 1 << + (clients[ILT_CLI_CDUC].p_size.val + PXP_ILT_PAGE_SIZE_NUM_BITS_MIN); + cdut_page_size = 1 << + (clients[ILT_CLI_CDUT].p_size.val + PXP_ILT_PAGE_SIZE_NUM_BITS_MIN); + conn_ctx_size = p_hwfn->p_cxt_mngr->conn_ctx_size; + num_cids_per_page = (int)(cduc_page_size / conn_ctx_size); + ilt_pages = p_hwfn->p_cxt_mngr->ilt_shadow; + + /* Dump global params - 22 must match number of params below */ + offset += qed_dump_common_global_params(p_hwfn, p_ptt, + dump_buf + offset, dump, 22); + offset += qed_dump_str_param(dump_buf + offset, + dump, "dump-type", "ilt-dump"); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cduc-page-size", cduc_page_size); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cduc-first-page-id", + clients[ILT_CLI_CDUC].first.val); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cduc-last-page-id", + clients[ILT_CLI_CDUC].last.val); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cduc-num-pf-pages", + clients + [ILT_CLI_CDUC].pf_total_lines); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cduc-num-vf-pages", + clients + [ILT_CLI_CDUC].vf_total_lines); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "max-conn-ctx-size", + conn_ctx_size); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-page-size", cdut_page_size); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-first-page-id", + clients[ILT_CLI_CDUT].first.val); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-last-page-id", + clients[ILT_CLI_CDUT].last.val); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-num-pf-init-pages", + qed_get_cdut_num_pf_init_pages(p_hwfn)); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-num-vf-init-pages", + qed_get_cdut_num_vf_init_pages(p_hwfn)); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-num-pf-work-pages", + qed_get_cdut_num_pf_work_pages(p_hwfn)); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "cdut-num-vf-work-pages", + qed_get_cdut_num_vf_work_pages(p_hwfn)); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "max-task-ctx-size", + p_hwfn->p_cxt_mngr->task_ctx_size); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "task-type-id", + p_hwfn->p_cxt_mngr->task_type_id); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "first-vf-id-in-pf", + p_hwfn->p_cxt_mngr->first_vf_in_pf); + offset += /* 18 */ qed_dump_num_param(dump_buf + offset, + dump, + "num-vfs-in-pf", + p_hwfn->p_cxt_mngr->vf_count); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "ptr-size-bytes", sizeof(void *)); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "pf-start-line", + p_hwfn->p_cxt_mngr->pf_start_line); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "page-mem-desc-size-dwords", + PAGE_MEM_DESC_SIZE_DWORDS); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "ilt-shadow-size", + p_hwfn->p_cxt_mngr->ilt_shadow_size); + /* Additional/Less parameters require matching of number in call to + * dump_common_global_params() + */ + + /* Dump section containing number of PF CIDs per connection type */ + offset += qed_dump_section_hdr(dump_buf + offset, + dump, "num_pf_cids_per_conn_type", 1); + offset += qed_dump_num_param(dump_buf + offset, + dump, "size", NUM_OF_CONNECTION_TYPES_E4); + for (conn_type = 0, valid_conn_pf_cids = 0; + conn_type < NUM_OF_CONNECTION_TYPES_E4; conn_type++, offset++) { + u32 num_pf_cids = + p_hwfn->p_cxt_mngr->conn_cfg[conn_type].cid_count; + + if (dump) + *(dump_buf + offset) = num_pf_cids; + valid_conn_pf_cids += num_pf_cids; + } + + /* Dump section containing number of VF CIDs per connection type */ + offset += qed_dump_section_hdr(dump_buf + offset, + dump, "num_vf_cids_per_conn_type", 1); + offset += qed_dump_num_param(dump_buf + offset, + dump, "size", NUM_OF_CONNECTION_TYPES_E4); + for (conn_type = 0, valid_conn_vf_cids = 0; + conn_type < NUM_OF_CONNECTION_TYPES_E4; conn_type++, offset++) { + u32 num_vf_cids = + p_hwfn->p_cxt_mngr->conn_cfg[conn_type].cids_per_vf; + + if (dump) + *(dump_buf + offset) = num_vf_cids; + valid_conn_vf_cids += num_vf_cids; + } + + /* Dump section containing physical memory descs for each ILT page */ + num_pages = p_hwfn->p_cxt_mngr->ilt_shadow_size; + offset += qed_dump_section_hdr(dump_buf + offset, + dump, "ilt_page_desc", 1); + offset += qed_dump_num_param(dump_buf + offset, + dump, + "size", + num_pages * PAGE_MEM_DESC_SIZE_DWORDS); + + /* Copy memory descriptors to dump buffer */ + if (dump) { + u32 page_id; + + for (page_id = 0; page_id < num_pages; + page_id++, offset += PAGE_MEM_DESC_SIZE_DWORDS) + memcpy(dump_buf + offset, + &ilt_pages[page_id], + DWORDS_TO_BYTES(PAGE_MEM_DESC_SIZE_DWORDS)); + } else { + offset += num_pages * PAGE_MEM_DESC_SIZE_DWORDS; + } + + valid_conn_pf_pages = DIV_ROUND_UP(valid_conn_pf_cids, + num_cids_per_page); + valid_conn_vf_pages = DIV_ROUND_UP(valid_conn_vf_cids, + num_cids_per_page); + + /* Dump ILT pages IDs */ + offset += qed_ilt_dump_pages_section(p_hwfn, + dump_buf + offset, + dump, + valid_conn_pf_pages, + valid_conn_vf_pages, + ilt_pages, true); + + /* Dump ILT pages memory */ + offset += qed_ilt_dump_pages_section(p_hwfn, + dump_buf + offset, + dump, + valid_conn_pf_pages, + valid_conn_vf_pages, + ilt_pages, false); + + /* Dump last section */ + offset += qed_dump_last_section(dump_buf, offset, dump); + + return offset; +} + /***************************** Public Functions *******************************/ -enum dbg_status qed_dbg_set_bin_ptr(const u8 * const bin_ptr) +enum dbg_status qed_dbg_set_bin_ptr(struct qed_hwfn *p_hwfn, + const u8 * const bin_ptr) { - struct bin_buffer_hdr *buf_array = (struct bin_buffer_hdr *)bin_ptr; + struct bin_buffer_hdr *buf_hdrs = (struct bin_buffer_hdr *)bin_ptr; u8 buf_id; - /* convert binary data to debug arrays */ - for (buf_id = 0; buf_id < MAX_BIN_DBG_BUFFER_TYPE; buf_id++) { - s_dbg_arrays[buf_id].ptr = - (u32 *)(bin_ptr + buf_array[buf_id].offset); - s_dbg_arrays[buf_id].size_in_dwords = - BYTES_TO_DWORDS(buf_array[buf_id].length); - } + /* Convert binary data to debug arrays */ + for (buf_id = 0; buf_id < MAX_BIN_DBG_BUFFER_TYPE; buf_id++) + qed_set_dbg_bin_buf(p_hwfn, + buf_id, + (u32 *)(bin_ptr + buf_hdrs[buf_id].offset), + buf_hdrs[buf_id].length); return DBG_STATUS_OK; } @@ -5116,7 +4836,7 @@ bool qed_read_fw_info(struct qed_hwfn *p_hwfn, struct storm_defs *storm = &s_storm_defs[storm_id]; /* Skip Storm if it's in reset */ - if (dev_data->block_in_reset[storm->block_id]) + if (dev_data->block_in_reset[storm->sem_block_id]) continue; /* Read FW info for the current Storm */ @@ -5129,16 +4849,17 @@ bool qed_read_fw_info(struct qed_hwfn *p_hwfn, } enum dbg_status qed_dbg_grc_config(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, enum dbg_grc_params grc_param, u32 val) { + struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; enum dbg_status status; int i; - DP_VERBOSE(p_hwfn, QED_MSG_DEBUG, + DP_VERBOSE(p_hwfn, + QED_MSG_DEBUG, "dbg_grc_config: paramId = %d, val = %d\n", grc_param, val); - status = qed_dbg_dev_init(p_hwfn, p_ptt); + status = qed_dbg_dev_init(p_hwfn); if (status != DBG_STATUS_OK) return status; @@ -5164,24 +4885,23 @@ enum dbg_status qed_dbg_grc_config(struct qed_hwfn *p_hwfn, /* Update all params with the preset values */ for (i = 0; i < MAX_DBG_GRC_PARAMS; i++) { + struct grc_param_defs *defs = &s_grc_param_defs[i]; u32 preset_val; - /* Skip persistent params */ - if (s_grc_param_defs[i].is_persistent) + if (defs->is_persistent) continue; /* Find preset value */ if (grc_param == DBG_GRC_PARAM_EXCLUDE_ALL) preset_val = - s_grc_param_defs[i].exclude_all_preset_val; + defs->exclude_all_preset_val; else if (grc_param == DBG_GRC_PARAM_CRASH) preset_val = - s_grc_param_defs[i].crash_preset_val; + defs->crash_preset_val[dev_data->chip_id]; else return DBG_STATUS_INVALID_ARGS; - qed_grc_set_param(p_hwfn, - (enum dbg_grc_params)i, preset_val); + qed_grc_set_param(p_hwfn, i, preset_val); } } else { /* Regular param - set its value */ @@ -5207,18 +4927,18 @@ enum dbg_status qed_dbg_grc_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; if (status != DBG_STATUS_OK) return status; - if (!s_dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || - !s_dbg_arrays[BIN_BUF_DBG_DUMP_REG].ptr || - !s_dbg_arrays[BIN_BUF_DBG_DUMP_MEM].ptr || - !s_dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr || - !s_dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr) + if (!p_hwfn->dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_DUMP_REG].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_DUMP_MEM].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr) return DBG_STATUS_DBG_ARRAY_NOT_SET; return qed_grc_dump(p_hwfn, p_ptt, NULL, false, buf_size); @@ -5258,20 +4978,19 @@ enum dbg_status qed_dbg_idle_chk_get_dump_buf_size(struct qed_hwfn *p_hwfn, u32 *buf_size) { struct dbg_tools_data *dev_data = &p_hwfn->dbg_info; - struct idle_chk_data *idle_chk; + struct idle_chk_data *idle_chk = &dev_data->idle_chk; enum dbg_status status; - idle_chk = &dev_data->idle_chk; *buf_size = 0; - status = qed_dbg_dev_init(p_hwfn, p_ptt); + status = qed_dbg_dev_init(p_hwfn); if (status != DBG_STATUS_OK) return status; - if (!s_dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || - !s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr || - !s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_IMMS].ptr || - !s_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES].ptr) + if (!p_hwfn->dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_REGS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_IMMS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_RULES].ptr) return DBG_STATUS_DBG_ARRAY_NOT_SET; if (!idle_chk->buf_size_set) { @@ -5306,6 +5025,7 @@ enum dbg_status qed_dbg_idle_chk_dump(struct qed_hwfn *p_hwfn, return DBG_STATUS_DUMP_BUF_TOO_SMALL; /* Update reset state */ + qed_grc_unreset_blocks(p_hwfn, p_ptt, true); qed_update_blocks_reset_state(p_hwfn, p_ptt); /* Idle Check Dump */ @@ -5321,7 +5041,7 @@ enum dbg_status qed_dbg_mcp_trace_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; @@ -5368,7 +5088,7 @@ enum dbg_status qed_dbg_reg_fifo_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; @@ -5414,7 +5134,7 @@ enum dbg_status qed_dbg_igu_fifo_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; @@ -5460,7 +5180,7 @@ qed_dbg_protection_override_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; @@ -5510,7 +5230,7 @@ enum dbg_status qed_dbg_fw_asserts_get_dump_buf_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u32 *buf_size) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); *buf_size = 0; @@ -5554,6 +5274,50 @@ enum dbg_status qed_dbg_fw_asserts_dump(struct qed_hwfn *p_hwfn, return DBG_STATUS_OK; } +static enum dbg_status qed_dbg_ilt_get_dump_buf_size(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 *buf_size) +{ + enum dbg_status status = qed_dbg_dev_init(p_hwfn); + + *buf_size = 0; + + if (status != DBG_STATUS_OK) + return status; + + *buf_size = qed_ilt_dump(p_hwfn, p_ptt, NULL, false); + + return DBG_STATUS_OK; +} + +static enum dbg_status qed_dbg_ilt_dump(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 *dump_buf, + u32 buf_size_in_dwords, + u32 *num_dumped_dwords) +{ + u32 needed_buf_size_in_dwords; + enum dbg_status status; + + *num_dumped_dwords = 0; + + status = qed_dbg_ilt_get_dump_buf_size(p_hwfn, + p_ptt, + &needed_buf_size_in_dwords); + if (status != DBG_STATUS_OK) + return status; + + if (buf_size_in_dwords < needed_buf_size_in_dwords) + return DBG_STATUS_DUMP_BUF_TOO_SMALL; + + *num_dumped_dwords = qed_ilt_dump(p_hwfn, p_ptt, dump_buf, true); + + /* Reveret GRC params to their default */ + qed_dbg_grc_set_params_default(p_hwfn); + + return DBG_STATUS_OK; +} + enum dbg_status qed_dbg_read_attn(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, enum block_id block_id, @@ -5561,19 +5325,20 @@ enum dbg_status qed_dbg_read_attn(struct qed_hwfn *p_hwfn, bool clear_status, struct dbg_attn_block_result *results) { - enum dbg_status status = qed_dbg_dev_init(p_hwfn, p_ptt); + enum dbg_status status = qed_dbg_dev_init(p_hwfn); u8 reg_idx, num_attn_regs, num_result_regs = 0; const struct dbg_attn_reg *attn_reg_arr; if (status != DBG_STATUS_OK) return status; - if (!s_dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || - !s_dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr || - !s_dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr) + if (!p_hwfn->dbg_arrays[BIN_BUF_DBG_MODE_TREE].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_BLOCKS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_REGS].ptr) return DBG_STATUS_DBG_ARRAY_NOT_SET; - attn_reg_arr = qed_get_block_attn_regs(block_id, + attn_reg_arr = qed_get_block_attn_regs(p_hwfn, + block_id, attn_type, &num_attn_regs); for (reg_idx = 0; reg_idx < num_attn_regs; reg_idx++) { @@ -5618,7 +5383,7 @@ enum dbg_status qed_dbg_read_attn(struct qed_hwfn *p_hwfn, results->block_id = (u8)block_id; results->names_offset = - qed_get_block_attn_data(block_id, attn_type)->names_offset; + qed_get_block_attn_data(p_hwfn, block_id, attn_type)->names_offset; SET_FIELD(results->data, DBG_ATTN_BLOCK_RESULT_ATTN_TYPE, attn_type); SET_FIELD(results->data, DBG_ATTN_BLOCK_RESULT_NUM_REGS, num_result_regs); @@ -5628,11 +5393,6 @@ enum dbg_status qed_dbg_read_attn(struct qed_hwfn *p_hwfn, /******************************* Data Types **********************************/ -struct block_info { - const char *name; - enum block_id id; -}; - /* REG fifo element */ struct reg_fifo_element { u64 data; @@ -5656,6 +5416,12 @@ struct reg_fifo_element { #define REG_FIFO_ELEMENT_ERROR_MASK 0x1f }; +/* REG fifo error element */ +struct reg_fifo_err { + u32 err_code; + const char *err_msg; +}; + /* IGU fifo element */ struct igu_fifo_element { u32 dword0; @@ -5755,20 +5521,6 @@ struct igu_fifo_addr_data { enum igu_fifo_addr_types type; }; -struct mcp_trace_meta { - u32 modules_num; - char **modules; - u32 formats_num; - struct mcp_trace_format *formats; - bool is_allocated; -}; - -/* Debug Tools user data */ -struct dbg_tools_user_data { - struct mcp_trace_meta mcp_trace_meta; - const u32 *mcp_trace_user_meta_buf; -}; - /******************************** Constants **********************************/ #define MAX_MSG_LEN 1024 @@ -5776,7 +5528,7 @@ struct dbg_tools_user_data { #define MCP_TRACE_MAX_MODULE_LEN 8 #define MCP_TRACE_FORMAT_MAX_PARAMS 3 #define MCP_TRACE_FORMAT_PARAM_WIDTH \ - (MCP_TRACE_FORMAT_P2_SIZE_SHIFT - MCP_TRACE_FORMAT_P1_SIZE_SHIFT) + (MCP_TRACE_FORMAT_P2_SIZE_OFFSET - MCP_TRACE_FORMAT_P1_SIZE_OFFSET) #define REG_FIFO_ELEMENT_ADDR_FACTOR 4 #define REG_FIFO_ELEMENT_IS_PF_VF_VAL 127 @@ -5785,107 +5537,6 @@ struct dbg_tools_user_data { /***************************** Constant Arrays *******************************/ -struct user_dbg_array { - const u32 *ptr; - u32 size_in_dwords; -}; - -/* Debug arrays */ -static struct user_dbg_array -s_user_dbg_arrays[MAX_BIN_DBG_BUFFER_TYPE] = { {NULL} }; - -/* Block names array */ -static struct block_info s_block_info_arr[] = { - {"grc", BLOCK_GRC}, - {"miscs", BLOCK_MISCS}, - {"misc", BLOCK_MISC}, - {"dbu", BLOCK_DBU}, - {"pglue_b", BLOCK_PGLUE_B}, - {"cnig", BLOCK_CNIG}, - {"cpmu", BLOCK_CPMU}, - {"ncsi", BLOCK_NCSI}, - {"opte", BLOCK_OPTE}, - {"bmb", BLOCK_BMB}, - {"pcie", BLOCK_PCIE}, - {"mcp", BLOCK_MCP}, - {"mcp2", BLOCK_MCP2}, - {"pswhst", BLOCK_PSWHST}, - {"pswhst2", BLOCK_PSWHST2}, - {"pswrd", BLOCK_PSWRD}, - {"pswrd2", BLOCK_PSWRD2}, - {"pswwr", BLOCK_PSWWR}, - {"pswwr2", BLOCK_PSWWR2}, - {"pswrq", BLOCK_PSWRQ}, - {"pswrq2", BLOCK_PSWRQ2}, - {"pglcs", BLOCK_PGLCS}, - {"ptu", BLOCK_PTU}, - {"dmae", BLOCK_DMAE}, - {"tcm", BLOCK_TCM}, - {"mcm", BLOCK_MCM}, - {"ucm", BLOCK_UCM}, - {"xcm", BLOCK_XCM}, - {"ycm", BLOCK_YCM}, - {"pcm", BLOCK_PCM}, - {"qm", BLOCK_QM}, - {"tm", BLOCK_TM}, - {"dorq", BLOCK_DORQ}, - {"brb", BLOCK_BRB}, - {"src", BLOCK_SRC}, - {"prs", BLOCK_PRS}, - {"tsdm", BLOCK_TSDM}, - {"msdm", BLOCK_MSDM}, - {"usdm", BLOCK_USDM}, - {"xsdm", BLOCK_XSDM}, - {"ysdm", BLOCK_YSDM}, - {"psdm", BLOCK_PSDM}, - {"tsem", BLOCK_TSEM}, - {"msem", BLOCK_MSEM}, - {"usem", BLOCK_USEM}, - {"xsem", BLOCK_XSEM}, - {"ysem", BLOCK_YSEM}, - {"psem", BLOCK_PSEM}, - {"rss", BLOCK_RSS}, - {"tmld", BLOCK_TMLD}, - {"muld", BLOCK_MULD}, - {"yuld", BLOCK_YULD}, - {"xyld", BLOCK_XYLD}, - {"ptld", BLOCK_PTLD}, - {"ypld", BLOCK_YPLD}, - {"prm", BLOCK_PRM}, - {"pbf_pb1", BLOCK_PBF_PB1}, - {"pbf_pb2", BLOCK_PBF_PB2}, - {"rpb", BLOCK_RPB}, - {"btb", BLOCK_BTB}, - {"pbf", BLOCK_PBF}, - {"rdif", BLOCK_RDIF}, - {"tdif", BLOCK_TDIF}, - {"cdu", BLOCK_CDU}, - {"ccfc", BLOCK_CCFC}, - {"tcfc", BLOCK_TCFC}, - {"igu", BLOCK_IGU}, - {"cau", BLOCK_CAU}, - {"rgfs", BLOCK_RGFS}, - {"rgsrc", BLOCK_RGSRC}, - {"tgfs", BLOCK_TGFS}, - {"tgsrc", BLOCK_TGSRC}, - {"umac", BLOCK_UMAC}, - {"xmac", BLOCK_XMAC}, - {"dbg", BLOCK_DBG}, - {"nig", BLOCK_NIG}, - {"wol", BLOCK_WOL}, - {"bmbn", BLOCK_BMBN}, - {"ipc", BLOCK_IPC}, - {"nwm", BLOCK_NWM}, - {"nws", BLOCK_NWS}, - {"ms", BLOCK_MS}, - {"phy_pcie", BLOCK_PHY_PCIE}, - {"led", BLOCK_LED}, - {"avs_wrap", BLOCK_AVS_WRAP}, - {"pxpreqbus", BLOCK_PXPREQBUS}, - {"misc_aeu", BLOCK_MISC_AEU}, - {"bar0_map", BLOCK_BAR0_MAP} -}; - /* Status string array */ static const char * const s_status_str[] = { /* DBG_STATUS_OK */ @@ -5915,14 +5566,12 @@ static const char * const s_status_str[] = { /* DBG_STATUS_PCI_BUF_NOT_ALLOCATED */ "A PCI buffer wasn't allocated", - /* DBG_STATUS_TOO_MANY_INPUTS */ - "Too many inputs were enabled. Enabled less inputs, or set 'unifyInputs' to true", + /* DBG_STATUS_INVALID_FILTER_TRIGGER_DWORDS */ + "The filter/trigger constraint dword offsets are not enabled for recording", - /* DBG_STATUS_INPUT_OVERLAP */ - "Overlapping debug bus inputs", - /* DBG_STATUS_HW_ONLY_RECORDING */ - "Cannot record Storm data since the entire recording cycle is used by HW", + /* DBG_STATUS_VFC_READ_ERROR */ + "Error reading from VFC", /* DBG_STATUS_STORM_ALREADY_ENABLED */ "The Storm was already enabled", @@ -5939,8 +5588,8 @@ static const char * const s_status_str[] = { /* DBG_STATUS_NO_INPUT_ENABLED */ "No input was enabled for recording", - /* DBG_STATUS_NO_FILTER_TRIGGER_64B */ - "Filters and triggers are not allowed when recording in 64b units", + /* DBG_STATUS_NO_FILTER_TRIGGER_256B */ + "Filters and triggers are not allowed in E4 256-bit mode", /* DBG_STATUS_FILTER_ALREADY_ENABLED */ "The filter was already enabled", @@ -6014,8 +5663,8 @@ static const char * const s_status_str[] = { /* DBG_STATUS_MCP_COULD_NOT_RESUME */ "Failed to resume MCP after halt", - /* DBG_STATUS_RESERVED2 */ - "Reserved debug status - shouldn't be returned", + /* DBG_STATUS_RESERVED0 */ + "", /* DBG_STATUS_SEMI_FIFO_NOT_EMPTY */ "Failed to empty SEMI sync FIFO", @@ -6038,17 +5687,32 @@ static const char * const s_status_str[] = { /* DBG_STATUS_DBG_ARRAY_NOT_SET */ "Debug arrays were not set (when using binary files, dbg_set_bin_ptr must be called)", - /* DBG_STATUS_FILTER_BUG */ - "Debug Bus filtering requires the -unifyInputs option (due to a HW bug)", + /* DBG_STATUS_RESERVED1 */ + "", /* DBG_STATUS_NON_MATCHING_LINES */ - "Non-matching debug lines - all lines must be of the same type (either 128b or 256b)", + "Non-matching debug lines - in E4, all lines must be of the same type (either 128b or 256b)", - /* DBG_STATUS_INVALID_TRIGGER_DWORD_OFFSET */ - "The selected trigger dword offset wasn't enabled in the recorded HW block", + /* DBG_STATUS_INSUFFICIENT_HW_IDS */ + "Insufficient HW IDs. Try to record less Storms/blocks", /* DBG_STATUS_DBG_BUS_IN_USE */ - "The debug bus is in use" + "The debug bus is in use", + + /* DBG_STATUS_INVALID_STORM_DBG_MODE */ + "The storm debug mode is not supported in the current chip", + + /* DBG_STATUS_OTHER_ENGINE_BB_ONLY */ + "Other engine is supported only in BB", + + /* DBG_STATUS_FILTER_SINGLE_HW_ID */ + "The configured filter mode requires a single Storm/block input", + + /* DBG_STATUS_TRIGGER_SINGLE_HW_ID */ + "The configured filter mode requires that all the constraints of a single trigger state will be defined on a single Storm/block input", + + /* DBG_STATUS_MISSING_TRIGGER_STATE_STORM */ + "When triggering on Storm data, the Storm to trigger on must be specified" }; /* Idle check severity names array */ @@ -6104,7 +5768,7 @@ static const char * const s_master_strs[] = { "xsdm", "dbu", "dmae", - "???", + "jdap", "???", "???", "???", @@ -6112,12 +5776,13 @@ static const char * const s_master_strs[] = { }; /* REG FIFO error messages array */ -static const char * const s_reg_fifo_error_strs[] = { - "grc timeout", - "address doesn't belong to any block", - "reserved address in block or write to read-only address", - "privilege/protection mismatch", - "path isolation error" +static struct reg_fifo_err s_reg_fifo_errors[] = { + {1, "grc timeout"}, + {2, "address doesn't belong to any block"}, + {4, "reserved address in block or write to read-only address"}, + {8, "privilege/protection mismatch"}, + {16, "path isolation error"}, + {17, "RSL error"} }; /* IGU FIFO sources array */ @@ -6357,8 +6022,21 @@ static u32 qed_print_section_params(u32 *dump_buf, return dump_offset; } -static struct dbg_tools_user_data * -qed_dbg_get_user_data(struct qed_hwfn *p_hwfn) +/* Returns the block name that matches the specified block ID, + * or NULL if not found. + */ +static const char *qed_dbg_get_block_name(struct qed_hwfn *p_hwfn, + enum block_id block_id) +{ + const struct dbg_block_user *block = + (const struct dbg_block_user *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_BLOCKS_USER_DATA].ptr + block_id; + + return (const char *)block->name; +} + +static struct dbg_tools_user_data *qed_dbg_get_user_data(struct qed_hwfn + *p_hwfn) { return (struct dbg_tools_user_data *)p_hwfn->dbg_user_info; } @@ -6366,7 +6044,8 @@ qed_dbg_get_user_data(struct qed_hwfn *p_hwfn) /* Parses the idle check rules and returns the number of characters printed. * In case of parsing error, returns 0. */ -static u32 qed_parse_idle_chk_dump_rules(u32 *dump_buf, +static u32 qed_parse_idle_chk_dump_rules(struct qed_hwfn *p_hwfn, + u32 *dump_buf, u32 *dump_buf_end, u32 num_rules, bool print_fw_idle_chk, @@ -6394,19 +6073,18 @@ static u32 qed_parse_idle_chk_dump_rules(u32 *dump_buf, hdr = (struct dbg_idle_chk_result_hdr *)dump_buf; rule_parsing_data = - (const struct dbg_idle_chk_rule_parsing_data *) - &s_user_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_PARSING_DATA]. - ptr[hdr->rule_id]; + (const struct dbg_idle_chk_rule_parsing_data *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_PARSING_DATA].ptr + + hdr->rule_id; parsing_str_offset = - GET_FIELD(rule_parsing_data->data, - DBG_IDLE_CHK_RULE_PARSING_DATA_STR_OFFSET); + GET_FIELD(rule_parsing_data->data, + DBG_IDLE_CHK_RULE_PARSING_DATA_STR_OFFSET); has_fw_msg = - GET_FIELD(rule_parsing_data->data, - DBG_IDLE_CHK_RULE_PARSING_DATA_HAS_FW_MSG) > 0; - parsing_str = - &((const char *) - s_user_dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr) - [parsing_str_offset]; + GET_FIELD(rule_parsing_data->data, + DBG_IDLE_CHK_RULE_PARSING_DATA_HAS_FW_MSG) > 0; + parsing_str = (const char *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr + + parsing_str_offset; lsi_msg = parsing_str; curr_reg_id = 0; @@ -6510,7 +6188,8 @@ static u32 qed_parse_idle_chk_dump_rules(u32 *dump_buf, * parsed_results_bytes. * The parsing status is returned. */ -static enum dbg_status qed_parse_idle_chk_dump(u32 *dump_buf, +static enum dbg_status qed_parse_idle_chk_dump(struct qed_hwfn *p_hwfn, + u32 *dump_buf, u32 num_dumped_dwords, char *results_buf, u32 *parsed_results_bytes, @@ -6528,8 +6207,8 @@ static enum dbg_status qed_parse_idle_chk_dump(u32 *dump_buf, *num_errors = 0; *num_warnings = 0; - if (!s_user_dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr || - !s_user_dbg_arrays[BIN_BUF_DBG_IDLE_CHK_PARSING_DATA].ptr) + if (!p_hwfn->dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_IDLE_CHK_PARSING_DATA].ptr) return DBG_STATUS_DBG_ARRAY_NOT_SET; /* Read global_params section */ @@ -6562,7 +6241,8 @@ static enum dbg_status qed_parse_idle_chk_dump(u32 *dump_buf, results_offset), "FW_IDLE_CHECK:\n"); rules_print_size = - qed_parse_idle_chk_dump_rules(dump_buf, + qed_parse_idle_chk_dump_rules(p_hwfn, + dump_buf, dump_buf_end, num_rules, true, @@ -6582,7 +6262,8 @@ static enum dbg_status qed_parse_idle_chk_dump(u32 *dump_buf, results_offset), "\nLSI_IDLE_CHECK:\n"); rules_print_size = - qed_parse_idle_chk_dump_rules(dump_buf, + qed_parse_idle_chk_dump_rules(p_hwfn, + dump_buf, dump_buf_end, num_rules, false, @@ -6694,9 +6375,8 @@ qed_mcp_trace_alloc_meta_data(struct qed_hwfn *p_hwfn, format_ptr->data = qed_read_dword_from_buf(meta_buf_bytes, &offset); - format_len = - (format_ptr->data & - MCP_TRACE_FORMAT_LEN_MASK) >> MCP_TRACE_FORMAT_LEN_SHIFT; + format_len = GET_MFW_FIELD(format_ptr->data, + MCP_TRACE_FORMAT_LEN); format_ptr->format_str = kzalloc(format_len, GFP_KERNEL); if (!format_ptr->format_str) { /* Update number of modules to be released */ @@ -6719,7 +6399,7 @@ qed_mcp_trace_alloc_meta_data(struct qed_hwfn *p_hwfn, * trace_buf - MCP trace cyclic buffer * trace_buf_size - MCP trace cyclic buffer size in bytes * data_offset - offset in bytes of the data to parse in the MCP trace cyclic - * buffer. + * buffer. * data_size - size in bytes of data to parse. * parsed_buf - destination buffer for parsed data. * parsed_results_bytes - size of parsed data in bytes. @@ -6764,9 +6444,8 @@ static enum dbg_status qed_parse_mcp_trace_buf(struct qed_hwfn *p_hwfn, /* Skip message if its index doesn't exist in the meta data */ if (format_idx >= meta->formats_num) { - u8 format_size = - (u8)((header & MFW_TRACE_PRM_SIZE_MASK) >> - MFW_TRACE_PRM_SIZE_SHIFT); + u8 format_size = (u8)GET_MFW_FIELD(header, + MFW_TRACE_PRM_SIZE); if (data_size < format_size) return DBG_STATUS_MCP_TRACE_BAD_DATA; @@ -6781,11 +6460,10 @@ static enum dbg_status qed_parse_mcp_trace_buf(struct qed_hwfn *p_hwfn, format_ptr = &meta->formats[format_idx]; for (i = 0, - param_mask = MCP_TRACE_FORMAT_P1_SIZE_MASK, - param_shift = MCP_TRACE_FORMAT_P1_SIZE_SHIFT; + param_mask = MCP_TRACE_FORMAT_P1_SIZE_MASK, param_shift = + MCP_TRACE_FORMAT_P1_SIZE_OFFSET; i < MCP_TRACE_FORMAT_MAX_PARAMS; - i++, - param_mask <<= MCP_TRACE_FORMAT_PARAM_WIDTH, + i++, param_mask <<= MCP_TRACE_FORMAT_PARAM_WIDTH, param_shift += MCP_TRACE_FORMAT_PARAM_WIDTH) { /* Extract param size (0..3) */ u8 param_size = (u8)((format_ptr->data & param_mask) >> @@ -6813,12 +6491,10 @@ static enum dbg_status qed_parse_mcp_trace_buf(struct qed_hwfn *p_hwfn, data_size -= param_size; } - format_level = (u8)((format_ptr->data & - MCP_TRACE_FORMAT_LEVEL_MASK) >> - MCP_TRACE_FORMAT_LEVEL_SHIFT); - format_module = (u8)((format_ptr->data & - MCP_TRACE_FORMAT_MODULE_MASK) >> - MCP_TRACE_FORMAT_MODULE_SHIFT); + format_level = (u8)GET_MFW_FIELD(format_ptr->data, + MCP_TRACE_FORMAT_LEVEL); + format_module = (u8)GET_MFW_FIELD(format_ptr->data, + MCP_TRACE_FORMAT_MODULE); if (format_level >= ARRAY_SIZE(s_mcp_trace_level_str)) return DBG_STATUS_MCP_TRACE_BAD_DATA; @@ -6960,7 +6636,7 @@ static enum dbg_status qed_parse_reg_fifo_dump(u32 *dump_buf, const char *section_name, *param_name, *param_str_val; u32 param_num_val, num_section_params, num_elements; struct reg_fifo_element *elements; - u8 i, j, err_val, vf_val; + u8 i, j, err_code, vf_val; u32 results_offset = 0; char vf_str[4]; @@ -6991,7 +6667,7 @@ static enum dbg_status qed_parse_reg_fifo_dump(u32 *dump_buf, /* Decode elements */ for (i = 0; i < num_elements; i++) { - bool err_printed = false; + const char *err_msg = NULL; /* Discover if element belongs to a VF or a PF */ vf_val = GET_FIELD(elements[i].data, REG_FIFO_ELEMENT_VF); @@ -7000,11 +6676,17 @@ static enum dbg_status qed_parse_reg_fifo_dump(u32 *dump_buf, else sprintf(vf_str, "%d", vf_val); + /* Find error message */ + err_code = GET_FIELD(elements[i].data, REG_FIFO_ELEMENT_ERROR); + for (j = 0; j < ARRAY_SIZE(s_reg_fifo_errors) && !err_msg; j++) + if (err_code == s_reg_fifo_errors[j].err_code) + err_msg = s_reg_fifo_errors[j].err_msg; + /* Add parsed element to parsed buffer */ results_offset += sprintf(qed_get_buf_ptr(results_buf, results_offset), - "raw: 0x%016llx, address: 0x%07x, access: %-5s, pf: %2d, vf: %s, port: %d, privilege: %-3s, protection: %-12s, master: %-4s, errors: ", + "raw: 0x%016llx, address: 0x%07x, access: %-5s, pf: %2d, vf: %s, port: %d, privilege: %-3s, protection: %-12s, master: %-4s, error: %s\n", elements[i].data, (u32)GET_FIELD(elements[i].data, REG_FIFO_ELEMENT_ADDRESS) * @@ -7021,30 +6703,8 @@ static enum dbg_status qed_parse_reg_fifo_dump(u32 *dump_buf, s_protection_strs[GET_FIELD(elements[i].data, REG_FIFO_ELEMENT_PROTECTION)], s_master_strs[GET_FIELD(elements[i].data, - REG_FIFO_ELEMENT_MASTER)]); - - /* Print errors */ - for (j = 0, - err_val = GET_FIELD(elements[i].data, - REG_FIFO_ELEMENT_ERROR); - j < ARRAY_SIZE(s_reg_fifo_error_strs); - j++, err_val >>= 1) { - if (err_val & 0x1) { - if (err_printed) - results_offset += - sprintf(qed_get_buf_ptr - (results_buf, - results_offset), ", "); - results_offset += - sprintf(qed_get_buf_ptr - (results_buf, results_offset), "%s", - s_reg_fifo_error_strs[j]); - err_printed = true; - } - } - - results_offset += - sprintf(qed_get_buf_ptr(results_buf, results_offset), "\n"); + REG_FIFO_ELEMENT_MASTER)], + err_msg ? err_msg : "unknown error code"); } results_offset += sprintf(qed_get_buf_ptr(results_buf, @@ -7398,27 +7058,28 @@ static enum dbg_status qed_parse_fw_asserts_dump(u32 *dump_buf, /***************************** Public Functions *******************************/ -enum dbg_status qed_dbg_user_set_bin_ptr(const u8 * const bin_ptr) +enum dbg_status qed_dbg_user_set_bin_ptr(struct qed_hwfn *p_hwfn, + const u8 * const bin_ptr) { - struct bin_buffer_hdr *buf_array = (struct bin_buffer_hdr *)bin_ptr; + struct bin_buffer_hdr *buf_hdrs = (struct bin_buffer_hdr *)bin_ptr; u8 buf_id; /* Convert binary data to debug arrays */ - for (buf_id = 0; buf_id < MAX_BIN_DBG_BUFFER_TYPE; buf_id++) { - s_user_dbg_arrays[buf_id].ptr = - (u32 *)(bin_ptr + buf_array[buf_id].offset); - s_user_dbg_arrays[buf_id].size_in_dwords = - BYTES_TO_DWORDS(buf_array[buf_id].length); - } + for (buf_id = 0; buf_id < MAX_BIN_DBG_BUFFER_TYPE; buf_id++) + qed_set_dbg_bin_buf(p_hwfn, + (enum bin_dbg_buffer_type)buf_id, + (u32 *)(bin_ptr + buf_hdrs[buf_id].offset), + buf_hdrs[buf_id].length); return DBG_STATUS_OK; } -enum dbg_status qed_dbg_alloc_user_data(struct qed_hwfn *p_hwfn) +enum dbg_status qed_dbg_alloc_user_data(struct qed_hwfn *p_hwfn, + void **user_data_ptr) { - p_hwfn->dbg_user_info = kzalloc(sizeof(struct dbg_tools_user_data), - GFP_KERNEL); - if (!p_hwfn->dbg_user_info) + *user_data_ptr = kzalloc(sizeof(struct dbg_tools_user_data), + GFP_KERNEL); + if (!(*user_data_ptr)) return DBG_STATUS_VIRT_MEM_ALLOC_FAILED; return DBG_STATUS_OK; @@ -7437,7 +7098,8 @@ enum dbg_status qed_get_idle_chk_results_buf_size(struct qed_hwfn *p_hwfn, { u32 num_errors, num_warnings; - return qed_parse_idle_chk_dump(dump_buf, + return qed_parse_idle_chk_dump(p_hwfn, + dump_buf, num_dumped_dwords, NULL, results_buf_size, @@ -7453,7 +7115,8 @@ enum dbg_status qed_print_idle_chk_results(struct qed_hwfn *p_hwfn, { u32 parsed_buf_size; - return qed_parse_idle_chk_dump(dump_buf, + return qed_parse_idle_chk_dump(p_hwfn, + dump_buf, num_dumped_dwords, results_buf, &parsed_buf_size, @@ -7624,25 +7287,28 @@ enum dbg_status qed_print_fw_asserts_results(struct qed_hwfn *p_hwfn, enum dbg_status qed_dbg_parse_attn(struct qed_hwfn *p_hwfn, struct dbg_attn_block_result *results) { - struct user_dbg_array *block_attn, *pstrings; const u32 *block_attn_name_offsets; - enum dbg_attn_type attn_type; + const char *attn_name_base; const char *block_name; + enum dbg_attn_type attn_type; u8 num_regs, i, j; num_regs = GET_FIELD(results->data, DBG_ATTN_BLOCK_RESULT_NUM_REGS); - attn_type = (enum dbg_attn_type) - GET_FIELD(results->data, - DBG_ATTN_BLOCK_RESULT_ATTN_TYPE); - block_name = s_block_info_arr[results->block_id].name; - - if (!s_user_dbg_arrays[BIN_BUF_DBG_ATTN_INDEXES].ptr || - !s_user_dbg_arrays[BIN_BUF_DBG_ATTN_NAME_OFFSETS].ptr || - !s_user_dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr) + attn_type = GET_FIELD(results->data, DBG_ATTN_BLOCK_RESULT_ATTN_TYPE); + block_name = qed_dbg_get_block_name(p_hwfn, results->block_id); + if (!block_name) + return DBG_STATUS_INVALID_ARGS; + + if (!p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_INDEXES].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_NAME_OFFSETS].ptr || + !p_hwfn->dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr) return DBG_STATUS_DBG_ARRAY_NOT_SET; - block_attn = &s_user_dbg_arrays[BIN_BUF_DBG_ATTN_NAME_OFFSETS]; - block_attn_name_offsets = &block_attn->ptr[results->names_offset]; + block_attn_name_offsets = + (u32 *)p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_NAME_OFFSETS].ptr + + results->names_offset; + + attn_name_base = p_hwfn->dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS].ptr; /* Go over registers with a non-zero attention status */ for (i = 0; i < num_regs; i++) { @@ -7653,18 +7319,17 @@ enum dbg_status qed_dbg_parse_attn(struct qed_hwfn *p_hwfn, reg_result = &results->reg_results[i]; num_reg_attn = GET_FIELD(reg_result->data, DBG_ATTN_REG_RESULT_NUM_REG_ATTN); - block_attn = &s_user_dbg_arrays[BIN_BUF_DBG_ATTN_INDEXES]; - bit_mapping = &((struct dbg_attn_bit_mapping *) - block_attn->ptr)[reg_result->block_attn_offset]; - - pstrings = &s_user_dbg_arrays[BIN_BUF_DBG_PARSING_STRINGS]; + bit_mapping = (struct dbg_attn_bit_mapping *) + p_hwfn->dbg_arrays[BIN_BUF_DBG_ATTN_INDEXES].ptr + + reg_result->block_attn_offset; /* Go over attention status bits */ - for (j = 0; j < num_reg_attn; j++) { + for (j = 0; j < num_reg_attn; j++, bit_idx++) { u16 attn_idx_val = GET_FIELD(bit_mapping[j].data, DBG_ATTN_BIT_MAPPING_VAL); const char *attn_name, *attn_type_str, *masked_str; - u32 attn_name_offset, sts_addr; + u32 attn_name_offset; + u32 sts_addr; /* Check if bit mask should be advanced (due to unused * bits). @@ -7676,18 +7341,19 @@ enum dbg_status qed_dbg_parse_attn(struct qed_hwfn *p_hwfn, } /* Check current bit index */ - if (!(reg_result->sts_val & BIT(bit_idx))) { - bit_idx++; + if (!(reg_result->sts_val & BIT(bit_idx))) continue; - } - /* Find attention name */ + /* An attention bit with value=1 was found + * Find attention name + */ attn_name_offset = block_attn_name_offsets[attn_idx_val]; - attn_name = &((const char *) - pstrings->ptr)[attn_name_offset]; - attn_type_str = attn_type == ATTN_TYPE_INTERRUPT ? - "Interrupt" : "Parity"; + attn_name = attn_name_base + attn_name_offset; + attn_type_str = + (attn_type == + ATTN_TYPE_INTERRUPT ? "Interrupt" : + "Parity"); masked_str = reg_result->mask_val & BIT(bit_idx) ? " [masked]" : ""; sts_addr = GET_FIELD(reg_result->data, @@ -7695,15 +7361,15 @@ enum dbg_status qed_dbg_parse_attn(struct qed_hwfn *p_hwfn, DP_NOTICE(p_hwfn, "%s (%s) : %s [address 0x%08x, bit %d]%s\n", block_name, attn_type_str, attn_name, - sts_addr, bit_idx, masked_str); - - bit_idx++; + sts_addr * 4, bit_idx, masked_str); } } return DBG_STATUS_OK; } +static DEFINE_MUTEX(qed_dbg_lock); + /* Wrapper for unifying the idle_chk and mcp_trace api */ static enum dbg_status qed_print_idle_chk_results_wrapper(struct qed_hwfn *p_hwfn, @@ -7763,7 +7429,10 @@ static struct { qed_dbg_fw_asserts_get_dump_buf_size, qed_dbg_fw_asserts_dump, qed_print_fw_asserts_results, - qed_get_fw_asserts_results_buf_size},}; + qed_get_fw_asserts_results_buf_size}, { + "ilt", + qed_dbg_ilt_get_dump_buf_size, + qed_dbg_ilt_dump, NULL, NULL},}; static void qed_dbg_print_feature(u8 *p_text_buf, u32 text_size) { @@ -7846,6 +7515,8 @@ static enum dbg_status format_feature(struct qed_hwfn *p_hwfn, return rc; } +#define MAX_DBG_FEATURE_SIZE_DWORDS 0x3FFFFFFF + /* Generic function for performing the dump of a debug feature. */ static enum dbg_status qed_dbg_dump(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, @@ -7875,6 +7546,17 @@ static enum dbg_status qed_dbg_dump(struct qed_hwfn *p_hwfn, &buf_size_dwords); if (rc != DBG_STATUS_OK && rc != DBG_STATUS_NVRAM_GET_IMAGE_FAILED) return rc; + + if (buf_size_dwords > MAX_DBG_FEATURE_SIZE_DWORDS) { + feature->buf_size = 0; + DP_NOTICE(p_hwfn->cdev, + "Debug feature [\"%s\"] size (0x%x dwords) exceeds maximum size (0x%x dwords)\n", + qed_features_lookup[feature_idx].name, + buf_size_dwords, MAX_DBG_FEATURE_SIZE_DWORDS); + + return DBG_STATUS_OK; + } + feature->buf_size = buf_size_dwords * sizeof(u32); feature->dump_buf = vmalloc(feature->buf_size); if (!feature->dump_buf) @@ -8021,6 +7703,16 @@ int qed_dbg_fw_asserts_size(struct qed_dev *cdev) return qed_dbg_feature_size(cdev, DBG_FEATURE_FW_ASSERTS); } +int qed_dbg_ilt(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes) +{ + return qed_dbg_feature(cdev, buffer, DBG_FEATURE_ILT, num_dumped_bytes); +} + +int qed_dbg_ilt_size(struct qed_dev *cdev) +{ + return qed_dbg_feature_size(cdev, DBG_FEATURE_ILT); +} + int qed_dbg_mcp_trace(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes) { @@ -8037,9 +7729,17 @@ int qed_dbg_mcp_trace_size(struct qed_dev *cdev) * feature buffer. */ #define REGDUMP_HEADER_SIZE sizeof(u32) +#define REGDUMP_HEADER_SIZE_SHIFT 0 +#define REGDUMP_HEADER_SIZE_MASK 0xffffff #define REGDUMP_HEADER_FEATURE_SHIFT 24 -#define REGDUMP_HEADER_ENGINE_SHIFT 31 +#define REGDUMP_HEADER_FEATURE_MASK 0x3f #define REGDUMP_HEADER_OMIT_ENGINE_SHIFT 30 +#define REGDUMP_HEADER_OMIT_ENGINE_MASK 0x1 +#define REGDUMP_HEADER_ENGINE_SHIFT 31 +#define REGDUMP_HEADER_ENGINE_MASK 0x1 +#define REGDUMP_MAX_SIZE 0x1000000 +#define ILT_DUMP_MAX_SIZE (1024 * 1024 * 15) + enum debug_print_features { OLD_MODE = 0, IDLE_CHK = 1, @@ -8053,17 +7753,27 @@ enum debug_print_features { NVM_CFG1 = 9, DEFAULT_CFG = 10, NVM_META = 11, + MDUMP = 12, + ILT_DUMP = 13, }; -static u32 qed_calc_regdump_header(enum debug_print_features feature, +static u32 qed_calc_regdump_header(struct qed_dev *cdev, + enum debug_print_features feature, int engine, u32 feature_size, u8 omit_engine) { - /* Insert the engine, feature and mode inside the header and combine it - * with feature size. - */ - return feature_size | (feature << REGDUMP_HEADER_FEATURE_SHIFT) | - (omit_engine << REGDUMP_HEADER_OMIT_ENGINE_SHIFT) | - (engine << REGDUMP_HEADER_ENGINE_SHIFT); + u32 res = 0; + + SET_FIELD(res, REGDUMP_HEADER_SIZE, feature_size); + if (res != feature_size) + DP_NOTICE(cdev, + "Feature %d is too large (size 0x%x) and will corrupt the dump\n", + feature, feature_size); + + SET_FIELD(res, REGDUMP_HEADER_FEATURE, feature); + SET_FIELD(res, REGDUMP_HEADER_OMIT_ENGINE, omit_engine); + SET_FIELD(res, REGDUMP_HEADER_ENGINE, engine); + + return res; } int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) @@ -8079,9 +7789,11 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) for (i = 0; i < MAX_DBG_GRC_PARAMS; i++) grc_params[i] = dev_data->grc.param_val[i]; - if (cdev->num_hwfns == 1) + if (!QED_IS_CMT(cdev)) omit_engine = 1; + mutex_lock(&qed_dbg_lock); + org_engine = qed_get_debug_engine(cdev); for (cur_engine = 0; cur_engine < cdev->num_hwfns; cur_engine++) { /* Collect idle_chks and grcDump for each hw function */ @@ -8094,7 +7806,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(IDLE_CHK, cur_engine, + qed_calc_regdump_header(cdev, IDLE_CHK, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8106,7 +7818,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(IDLE_CHK, cur_engine, + qed_calc_regdump_header(cdev, IDLE_CHK, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8118,7 +7830,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(REG_FIFO, cur_engine, + qed_calc_regdump_header(cdev, REG_FIFO, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8130,7 +7842,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(IGU_FIFO, cur_engine, + qed_calc_regdump_header(cdev, IGU_FIFO, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8143,7 +7855,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(PROTECTION_OVERRIDE, + qed_calc_regdump_header(cdev, PROTECTION_OVERRIDE, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); @@ -8158,25 +7870,45 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(FW_ASSERTS, cur_engine, - feature_size, omit_engine); + qed_calc_regdump_header(cdev, FW_ASSERTS, + cur_engine, feature_size, + omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { DP_ERR(cdev, "qed_dbg_fw_asserts failed. rc = %d\n", rc); } - for (i = 0; i < MAX_DBG_GRC_PARAMS; i++) - dev_data->grc.param_val[i] = grc_params[i]; + feature_size = qed_dbg_ilt_size(cdev); + if (!cdev->disable_ilt_dump && + feature_size < ILT_DUMP_MAX_SIZE) { + rc = qed_dbg_ilt(cdev, (u8 *)buffer + offset + + REGDUMP_HEADER_SIZE, &feature_size); + if (!rc) { + *(u32 *)((u8 *)buffer + offset) = + qed_calc_regdump_header(cdev, ILT_DUMP, + cur_engine, + feature_size, + omit_engine); + offset += feature_size + REGDUMP_HEADER_SIZE; + } else { + DP_ERR(cdev, "qed_dbg_ilt failed. rc = %d\n", + rc); + } + } /* GRC dump - must be last because when mcp stuck it will * clutter idle_chk, reg_fifo, ... */ + for (i = 0; i < MAX_DBG_GRC_PARAMS; i++) + dev_data->grc.param_val[i] = grc_params[i]; + rc = qed_dbg_grc(cdev, (u8 *)buffer + offset + REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(GRC_DUMP, cur_engine, + qed_calc_regdump_header(cdev, GRC_DUMP, + cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8185,12 +7917,13 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) } qed_set_debug_engine(cdev, org_engine); + /* mcp_trace */ rc = qed_dbg_mcp_trace(cdev, (u8 *)buffer + offset + REGDUMP_HEADER_SIZE, &feature_size); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(MCP_TRACE, cur_engine, + qed_calc_regdump_header(cdev, MCP_TRACE, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else { @@ -8199,11 +7932,12 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) /* nvm cfg1 */ rc = qed_dbg_nvm_image(cdev, - (u8 *)buffer + offset + REGDUMP_HEADER_SIZE, - &feature_size, QED_NVM_IMAGE_NVM_CFG1); + (u8 *)buffer + offset + + REGDUMP_HEADER_SIZE, &feature_size, + QED_NVM_IMAGE_NVM_CFG1); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(NVM_CFG1, cur_engine, + qed_calc_regdump_header(cdev, NVM_CFG1, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else if (rc != -ENOENT) { @@ -8218,7 +7952,7 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) &feature_size, QED_NVM_IMAGE_DEFAULT_CFG); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(DEFAULT_CFG, cur_engine, + qed_calc_regdump_header(cdev, DEFAULT_CFG, cur_engine, feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else if (rc != -ENOENT) { @@ -8234,8 +7968,8 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) &feature_size, QED_NVM_IMAGE_NVM_META); if (!rc) { *(u32 *)((u8 *)buffer + offset) = - qed_calc_regdump_header(NVM_META, cur_engine, - feature_size, omit_engine); + qed_calc_regdump_header(cdev, NVM_META, cur_engine, + feature_size, omit_engine); offset += (feature_size + REGDUMP_HEADER_SIZE); } else if (rc != -ENOENT) { DP_ERR(cdev, @@ -8243,6 +7977,23 @@ int qed_dbg_all_data(struct qed_dev *cdev, void *buffer) QED_NVM_IMAGE_NVM_META, "QED_NVM_IMAGE_NVM_META", rc); } + /* nvm mdump */ + rc = qed_dbg_nvm_image(cdev, (u8 *)buffer + offset + + REGDUMP_HEADER_SIZE, &feature_size, + QED_NVM_IMAGE_MDUMP); + if (!rc) { + *(u32 *)((u8 *)buffer + offset) = + qed_calc_regdump_header(cdev, MDUMP, cur_engine, + feature_size, omit_engine); + offset += (feature_size + REGDUMP_HEADER_SIZE); + } else if (rc != -ENOENT) { + DP_ERR(cdev, + "qed_dbg_nvm_image failed for image %d (%s), rc = %d\n", + QED_NVM_IMAGE_MDUMP, "QED_NVM_IMAGE_MDUMP", rc); + } + + mutex_unlock(&qed_dbg_lock); + return 0; } @@ -8250,9 +8001,10 @@ int qed_dbg_all_data_size(struct qed_dev *cdev) { struct qed_hwfn *p_hwfn = &cdev->hwfns[cdev->dbg_params.engine_for_debug]; - u32 regs_len = 0, image_len = 0; + u32 regs_len = 0, image_len = 0, ilt_len = 0, total_ilt_len = 0; u8 cur_engine, org_engine; + cdev->disable_ilt_dump = false; org_engine = qed_get_debug_engine(cdev); for (cur_engine = 0; cur_engine < cdev->num_hwfns; cur_engine++) { /* Engine specific */ @@ -8267,6 +8019,12 @@ int qed_dbg_all_data_size(struct qed_dev *cdev) REGDUMP_HEADER_SIZE + qed_dbg_protection_override_size(cdev) + REGDUMP_HEADER_SIZE + qed_dbg_fw_asserts_size(cdev); + + ilt_len = REGDUMP_HEADER_SIZE + qed_dbg_ilt_size(cdev); + if (ilt_len < ILT_DUMP_MAX_SIZE) { + total_ilt_len += ilt_len; + regs_len += ilt_len; + } } qed_set_debug_engine(cdev, org_engine); @@ -8282,6 +8040,17 @@ int qed_dbg_all_data_size(struct qed_dev *cdev) qed_dbg_nvm_image_length(p_hwfn, QED_NVM_IMAGE_NVM_META, &image_len); if (image_len) regs_len += REGDUMP_HEADER_SIZE + image_len; + qed_dbg_nvm_image_length(p_hwfn, QED_NVM_IMAGE_MDUMP, &image_len); + if (image_len) + regs_len += REGDUMP_HEADER_SIZE + image_len; + + if (regs_len > REGDUMP_MAX_SIZE) { + DP_VERBOSE(cdev, QED_MSG_DEBUG, + "Dump exceeds max size 0x%x, disable ILT dump\n", + REGDUMP_MAX_SIZE); + cdev->disable_ilt_dump = true; + regs_len -= total_ilt_len; + } return regs_len; } @@ -8327,9 +8096,8 @@ int qed_dbg_feature_size(struct qed_dev *cdev, enum qed_dbg_features feature) { struct qed_hwfn *p_hwfn = &cdev->hwfns[cdev->dbg_params.engine_for_debug]; + struct qed_dbg_feature *qed_feature = &cdev->dbg_features[feature]; struct qed_ptt *p_ptt = qed_ptt_acquire(p_hwfn); - struct qed_dbg_feature *qed_feature = - &cdev->dbg_params.features[feature]; u32 buf_size_dwords; enum dbg_status rc; @@ -8341,6 +8109,10 @@ int qed_dbg_feature_size(struct qed_dev *cdev, enum qed_dbg_features feature) if (rc != DBG_STATUS_OK) buf_size_dwords = 0; + /* Feature will not be dumped if it exceeds maximum size */ + if (buf_size_dwords > MAX_DBG_FEATURE_SIZE_DWORDS) + buf_size_dwords = 0; + qed_ptt_release(p_hwfn, p_ptt); qed_feature->buf_size = buf_size_dwords * sizeof(u32); return qed_feature->buf_size; @@ -8360,14 +8132,21 @@ void qed_set_debug_engine(struct qed_dev *cdev, int engine_number) void qed_dbg_pf_init(struct qed_dev *cdev) { - const u8 *dbg_values; + const u8 *dbg_values = NULL; + int i; /* Debug values are after init values. * The offset is the first dword of the file. */ dbg_values = cdev->firmware->data + *(u32 *)cdev->firmware->data; - qed_dbg_set_bin_ptr((u8 *)dbg_values); - qed_dbg_user_set_bin_ptr((u8 *)dbg_values); + + for_each_hwfn(cdev, i) { + qed_dbg_set_bin_ptr(&cdev->hwfns[i], dbg_values); + qed_dbg_user_set_bin_ptr(&cdev->hwfns[i], dbg_values); + } + + /* Set the hwfn to be 0 as default */ + cdev->dbg_params.engine_for_debug = 0; } void qed_dbg_pf_exit(struct qed_dev *cdev) @@ -8375,11 +8154,11 @@ void qed_dbg_pf_exit(struct qed_dev *cdev) struct qed_dbg_feature *feature = NULL; enum qed_dbg_features feature_idx; - /* Debug features' buffers may be allocated if debug feature was used - * but dump wasn't called. + /* debug features' buffers may be allocated if debug feature was used + * but dump wasn't called */ for (feature_idx = 0; feature_idx < DBG_FEATURE_NUM; feature_idx++) { - feature = &cdev->dbg_params.features[feature_idx]; + feature = &cdev->dbg_features[feature_idx]; if (feature->dump_buf) { vfree(feature->dump_buf); feature->dump_buf = NULL; diff --git a/drivers/net/ethernet/qlogic/qed/qed_debug.h b/drivers/net/ethernet/qlogic/qed/qed_debug.h index e47e0e8d75b0..edf99d296bd1 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_debug.h +++ b/drivers/net/ethernet/qlogic/qed/qed_debug.h @@ -14,11 +14,13 @@ enum qed_dbg_features { DBG_FEATURE_IGU_FIFO, DBG_FEATURE_PROTECTION_OVERRIDE, DBG_FEATURE_FW_ASSERTS, + DBG_FEATURE_ILT, DBG_FEATURE_NUM }; /* Forward Declaration */ struct qed_dev; +struct qed_hwfn; int qed_dbg_grc(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes); int qed_dbg_grc_size(struct qed_dev *cdev); @@ -37,6 +39,8 @@ int qed_dbg_protection_override_size(struct qed_dev *cdev); int qed_dbg_fw_asserts(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes); int qed_dbg_fw_asserts_size(struct qed_dev *cdev); +int qed_dbg_ilt(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes); +int qed_dbg_ilt_size(struct qed_dev *cdev); int qed_dbg_mcp_trace(struct qed_dev *cdev, void *buffer, u32 *num_dumped_bytes); int qed_dbg_mcp_trace_size(struct qed_dev *cdev); diff --git a/drivers/net/ethernet/qlogic/qed/qed_dev.c b/drivers/net/ethernet/qlogic/qed/qed_dev.c index a1ebc2b1ca0b..7912911337d4 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_dev.c +++ b/drivers/net/ethernet/qlogic/qed/qed_dev.c @@ -907,7 +907,7 @@ qed_llh_access_filter(struct qed_hwfn *p_hwfn, /* Filter value */ addr = NIG_REG_LLH_FUNC_FILTER_VALUE + 2 * filter_idx * 0x4; - params.flags = QED_DMAE_FLAG_PF_DST; + SET_FIELD(params.flags, QED_DMAE_PARAMS_DST_PF_VALID, 0x1); params.dst_pfid = pfid; rc = qed_dmae_host2grc(p_hwfn, p_ptt, @@ -1412,6 +1412,7 @@ void qed_resc_free(struct qed_dev *cdev) qed_dmae_info_free(p_hwfn); qed_dcbx_info_free(p_hwfn); qed_dbg_user_data_free(p_hwfn); + qed_fw_overlay_mem_free(p_hwfn, p_hwfn->fw_overlay_mem); /* Destroy doorbell recovery mechanism */ qed_db_recovery_teardown(p_hwfn); @@ -1571,7 +1572,7 @@ static void qed_init_qm_vport_params(struct qed_hwfn *p_hwfn) /* all vports participate in weighted fair queueing */ for (i = 0; i < qed_init_qm_get_num_vports(p_hwfn); i++) - qm_info->qm_vport_params[i].vport_wfq = 1; + qm_info->qm_vport_params[i].wfq = 1; } /* initialize qm port params */ @@ -1579,6 +1580,7 @@ static void qed_init_qm_port_params(struct qed_hwfn *p_hwfn) { /* Initialize qm port parameters */ u8 i, active_phys_tcs, num_ports = p_hwfn->cdev->num_ports_in_engine; + struct qed_dev *cdev = p_hwfn->cdev; /* indicate how ooo and high pri traffic is dealt with */ active_phys_tcs = num_ports == MAX_NUM_PORTS_K2 ? @@ -1588,11 +1590,13 @@ static void qed_init_qm_port_params(struct qed_hwfn *p_hwfn) for (i = 0; i < num_ports; i++) { struct init_qm_port_params *p_qm_port = &p_hwfn->qm_info.qm_port_params[i]; + u16 pbf_max_cmd_lines; p_qm_port->active = 1; p_qm_port->active_phys_tcs = active_phys_tcs; - p_qm_port->num_pbf_cmd_lines = PBF_MAX_CMD_LINES / num_ports; - p_qm_port->num_btb_blocks = BTB_MAX_BLOCKS / num_ports; + pbf_max_cmd_lines = (u16)NUM_OF_PBF_CMD_LINES(cdev); + p_qm_port->num_pbf_cmd_lines = pbf_max_cmd_lines / num_ports; + p_qm_port->num_btb_blocks = NUM_OF_BTB_BLOCKS(cdev) / num_ports; } } @@ -2034,9 +2038,8 @@ static void qed_dp_init_qm_params(struct qed_hwfn *p_hwfn) vport = &(qm_info->qm_vport_params[i]); DP_VERBOSE(p_hwfn, NETIF_MSG_HW, - "vport idx %d, vport_rl %d, wfq %d, first_tx_pq_id [ ", - qm_info->start_vport + i, - vport->vport_rl, vport->vport_wfq); + "vport idx %d, wfq %d, first_tx_pq_id [ ", + qm_info->start_vport + i, vport->wfq); for (tc = 0; tc < NUM_OF_TCS; tc++) DP_VERBOSE(p_hwfn, NETIF_MSG_HW, @@ -2049,11 +2052,11 @@ static void qed_dp_init_qm_params(struct qed_hwfn *p_hwfn) pq = &(qm_info->qm_pq_params[i]); DP_VERBOSE(p_hwfn, NETIF_MSG_HW, - "pq idx %d, port %d, vport_id %d, tc %d, wrr_grp %d, rl_valid %d\n", + "pq idx %d, port %d, vport_id %d, tc %d, wrr_grp %d, rl_valid %d rl_id %d\n", qm_info->start_pq + i, pq->port_id, pq->vport_id, - pq->tc_id, pq->wrr_group, pq->rl_valid); + pq->tc_id, pq->wrr_group, pq->rl_valid, pq->rl_id); } } @@ -2103,9 +2106,6 @@ int qed_qm_reconf(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) if (!b_rc) return -EINVAL; - /* clear the QM_PF runtime phase leftovers from previous init */ - qed_init_clear_rt_data(p_hwfn); - /* prepare QM portion of runtime array */ qed_qm_init_pf(p_hwfn, p_ptt, false); @@ -2346,7 +2346,7 @@ int qed_resc_alloc(struct qed_dev *cdev) if (rc) goto alloc_err; - rc = qed_dbg_alloc_user_data(p_hwfn); + rc = qed_dbg_alloc_user_data(p_hwfn, &p_hwfn->dbg_user_info); if (rc) goto alloc_err; } @@ -2623,7 +2623,7 @@ static int qed_hw_init_common(struct qed_hwfn *p_hwfn, params.max_phys_tcs_per_port = qm_info->max_phys_tcs_per_port; params.pf_rl_en = qm_info->pf_rl_en; params.pf_wfq_en = qm_info->pf_wfq_en; - params.vport_rl_en = qm_info->vport_rl_en; + params.global_rl_en = qm_info->vport_rl_en; params.vport_wfq_en = qm_info->vport_wfq_en; params.port_params = qm_info->qm_port_params; @@ -2891,6 +2891,8 @@ static int qed_hw_init_pf(struct qed_hwfn *p_hwfn, if (rc) return rc; + qed_fw_overlay_init_ram(p_hwfn, p_ptt, p_hwfn->fw_overlay_mem); + /* Pure runtime initializations - directly to the HW */ qed_int_igu_init_pure_rt(p_hwfn, p_ptt, true, true); @@ -3000,8 +3002,10 @@ int qed_hw_init(struct qed_dev *cdev, struct qed_hw_init_params *p_params) u32 load_code, resp, param, drv_mb_param; bool b_default_mtu = true; struct qed_hwfn *p_hwfn; - int rc = 0, i; + const u32 *fw_overlays; + u32 fw_overlays_len; u16 ether_type; + int rc = 0, i; if ((p_params->int_mode == QED_INT_MODE_MSI) && (cdev->num_hwfns > 1)) { DP_NOTICE(cdev, "MSI mode is not supported for CMT devices\n"); @@ -3102,6 +3106,17 @@ int qed_hw_init(struct qed_dev *cdev, struct qed_hw_init_params *p_params) */ qed_pglueb_clear_err(p_hwfn, p_hwfn->p_main_ptt); + fw_overlays = cdev->fw_data->fw_overlays; + fw_overlays_len = cdev->fw_data->fw_overlays_len; + p_hwfn->fw_overlay_mem = + qed_fw_overlay_mem_alloc(p_hwfn, fw_overlays, + fw_overlays_len); + if (!p_hwfn->fw_overlay_mem) { + DP_NOTICE(p_hwfn, + "Failed to allocate fw overlay memory\n"); + goto load_err; + } + switch (load_code) { case FW_MSG_CODE_DRV_LOAD_ENGINE: rc = qed_hw_init_common(p_hwfn, p_hwfn->p_main_ptt, @@ -3566,8 +3581,10 @@ const char *qed_hw_get_resc_name(enum qed_resources res_id) return "RDMA_CNQ_RAM"; case QED_ILT: return "ILT"; - case QED_LL2_QUEUE: - return "LL2_QUEUE"; + case QED_LL2_RAM_QUEUE: + return "LL2_RAM_QUEUE"; + case QED_LL2_CTX_QUEUE: + return "LL2_CTX_QUEUE"; case QED_CMDQS_CQS: return "CMDQS_CQS"; case QED_RDMA_STATS_QUEUE: @@ -3606,18 +3623,46 @@ __qed_hw_set_soft_resc_size(struct qed_hwfn *p_hwfn, return 0; } +static u32 qed_hsi_def_val[][MAX_CHIP_IDS] = { + {MAX_NUM_VFS_BB, MAX_NUM_VFS_K2}, + {MAX_NUM_L2_QUEUES_BB, MAX_NUM_L2_QUEUES_K2}, + {MAX_NUM_PORTS_BB, MAX_NUM_PORTS_K2}, + {MAX_SB_PER_PATH_BB, MAX_SB_PER_PATH_K2,}, + {MAX_NUM_PFS_BB, MAX_NUM_PFS_K2}, + {MAX_NUM_VPORTS_BB, MAX_NUM_VPORTS_K2}, + {ETH_RSS_ENGINE_NUM_BB, ETH_RSS_ENGINE_NUM_K2}, + {MAX_QM_TX_QUEUES_BB, MAX_QM_TX_QUEUES_K2}, + {PXP_NUM_ILT_RECORDS_BB, PXP_NUM_ILT_RECORDS_K2}, + {RDMA_NUM_STATISTIC_COUNTERS_BB, RDMA_NUM_STATISTIC_COUNTERS_K2}, + {MAX_QM_GLOBAL_RLS, MAX_QM_GLOBAL_RLS}, + {PBF_MAX_CMD_LINES, PBF_MAX_CMD_LINES}, + {BTB_MAX_BLOCKS_BB, BTB_MAX_BLOCKS_K2}, +}; + +u32 qed_get_hsi_def_val(struct qed_dev *cdev, enum qed_hsi_def_type type) +{ + enum chip_ids chip_id = QED_IS_BB(cdev) ? CHIP_BB : CHIP_K2; + + if (type >= QED_NUM_HSI_DEFS) { + DP_ERR(cdev, "Unexpected HSI definition type [%d]\n", type); + return 0; + } + + return qed_hsi_def_val[type][chip_id]; +} static int qed_hw_set_soft_resc_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) { - bool b_ah = QED_IS_AH(p_hwfn->cdev); u32 resc_max_val, mcp_resp; u8 res_id; int rc; - for (res_id = 0; res_id < QED_MAX_RESC; res_id++) { switch (res_id) { - case QED_LL2_QUEUE: - resc_max_val = MAX_NUM_LL2_RX_QUEUES; + case QED_LL2_RAM_QUEUE: + resc_max_val = MAX_NUM_LL2_RX_RAM_QUEUES; + break; + case QED_LL2_CTX_QUEUE: + resc_max_val = MAX_NUM_LL2_RX_CTX_QUEUES; break; case QED_RDMA_CNQ_RAM: /* No need for a case for QED_CMDQS_CQS since @@ -3626,8 +3671,8 @@ qed_hw_set_soft_resc_size(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) resc_max_val = NUM_OF_GLOBAL_QUEUES; break; case QED_RDMA_STATS_QUEUE: - resc_max_val = b_ah ? RDMA_NUM_STATISTIC_COUNTERS_K2 - : RDMA_NUM_STATISTIC_COUNTERS_BB; + resc_max_val = + NUM_OF_RDMA_STATISTIC_COUNTERS(p_hwfn->cdev); break; case QED_BDQ: resc_max_val = BDQ_NUM_RESOURCES; @@ -3660,28 +3705,24 @@ int qed_hw_get_dflt_resc(struct qed_hwfn *p_hwfn, u32 *p_resc_num, u32 *p_resc_start) { u8 num_funcs = p_hwfn->num_funcs_on_engine; - bool b_ah = QED_IS_AH(p_hwfn->cdev); + struct qed_dev *cdev = p_hwfn->cdev; switch (res_id) { case QED_L2_QUEUE: - *p_resc_num = (b_ah ? MAX_NUM_L2_QUEUES_K2 : - MAX_NUM_L2_QUEUES_BB) / num_funcs; + *p_resc_num = NUM_OF_L2_QUEUES(cdev) / num_funcs; break; case QED_VPORT: - *p_resc_num = (b_ah ? MAX_NUM_VPORTS_K2 : - MAX_NUM_VPORTS_BB) / num_funcs; + *p_resc_num = NUM_OF_VPORTS(cdev) / num_funcs; break; case QED_RSS_ENG: - *p_resc_num = (b_ah ? ETH_RSS_ENGINE_NUM_K2 : - ETH_RSS_ENGINE_NUM_BB) / num_funcs; + *p_resc_num = NUM_OF_RSS_ENGINES(cdev) / num_funcs; break; case QED_PQ: - *p_resc_num = (b_ah ? MAX_QM_TX_QUEUES_K2 : - MAX_QM_TX_QUEUES_BB) / num_funcs; + *p_resc_num = NUM_OF_QM_TX_QUEUES(cdev) / num_funcs; *p_resc_num &= ~0x7; /* The granularity of the PQs is 8 */ break; case QED_RL: - *p_resc_num = MAX_QM_GLOBAL_RLS / num_funcs; + *p_resc_num = NUM_OF_QM_GLOBAL_RLS(cdev) / num_funcs; break; case QED_MAC: case QED_VLAN: @@ -3689,11 +3730,13 @@ int qed_hw_get_dflt_resc(struct qed_hwfn *p_hwfn, *p_resc_num = ETH_NUM_MAC_FILTERS / num_funcs; break; case QED_ILT: - *p_resc_num = (b_ah ? PXP_NUM_ILT_RECORDS_K2 : - PXP_NUM_ILT_RECORDS_BB) / num_funcs; + *p_resc_num = NUM_OF_PXP_ILT_RECORDS(cdev) / num_funcs; + break; + case QED_LL2_RAM_QUEUE: + *p_resc_num = MAX_NUM_LL2_RX_RAM_QUEUES / num_funcs; break; - case QED_LL2_QUEUE: - *p_resc_num = MAX_NUM_LL2_RX_QUEUES / num_funcs; + case QED_LL2_CTX_QUEUE: + *p_resc_num = MAX_NUM_LL2_RX_CTX_QUEUES / num_funcs; break; case QED_RDMA_CNQ_RAM: case QED_CMDQS_CQS: @@ -3701,8 +3744,7 @@ int qed_hw_get_dflt_resc(struct qed_hwfn *p_hwfn, *p_resc_num = NUM_OF_GLOBAL_QUEUES / num_funcs; break; case QED_RDMA_STATS_QUEUE: - *p_resc_num = (b_ah ? RDMA_NUM_STATISTIC_COUNTERS_K2 : - RDMA_NUM_STATISTIC_COUNTERS_BB) / num_funcs; + *p_resc_num = NUM_OF_RDMA_STATISTIC_COUNTERS(cdev) / num_funcs; break; case QED_BDQ: if (p_hwfn->hw_info.personality != QED_PCI_ISCSI && @@ -5087,11 +5129,11 @@ static void qed_configure_wfq_for_all_vports(struct qed_hwfn *p_hwfn, for (i = 0; i < p_hwfn->qm_info.num_vports; i++) { u32 wfq_speed = p_hwfn->qm_info.wfq_data[i].min_speed; - vport_params[i].vport_wfq = (wfq_speed * QED_WFQ_UNIT) / + vport_params[i].wfq = (wfq_speed * QED_WFQ_UNIT) / min_pf_rate; qed_init_vport_wfq(p_hwfn, p_ptt, vport_params[i].first_tx_pq_id, - vport_params[i].vport_wfq); + vport_params[i].wfq); } } @@ -5102,7 +5144,7 @@ static void qed_init_wfq_default_param(struct qed_hwfn *p_hwfn, int i; for (i = 0; i < p_hwfn->qm_info.num_vports; i++) - p_hwfn->qm_info.qm_vport_params[i].vport_wfq = 1; + p_hwfn->qm_info.qm_vport_params[i].wfq = 1; } static void qed_disable_wfq_for_all_vports(struct qed_hwfn *p_hwfn, @@ -5118,7 +5160,7 @@ static void qed_disable_wfq_for_all_vports(struct qed_hwfn *p_hwfn, qed_init_wfq_default_param(p_hwfn, min_pf_rate); qed_init_vport_wfq(p_hwfn, p_ptt, vport_params[i].first_tx_pq_id, - vport_params[i].vport_wfq); + vport_params[i].wfq); } } diff --git a/drivers/net/ethernet/qlogic/qed/qed_dev_api.h b/drivers/net/ethernet/qlogic/qed/qed_dev_api.h index 47376d4d071f..eb4808b3bf67 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_dev_api.h +++ b/drivers/net/ethernet/qlogic/qed/qed_dev_api.h @@ -230,30 +230,6 @@ enum qed_dmae_address_type_t { QED_DMAE_ADDRESS_GRC }; -/* value of flags If QED_DMAE_FLAG_RW_REPL_SRC flag is set and the - * source is a block of length DMAE_MAX_RW_SIZE and the - * destination is larger, the source block will be duplicated as - * many times as required to fill the destination block. This is - * used mostly to write a zeroed buffer to destination address - * using DMA - */ -#define QED_DMAE_FLAG_RW_REPL_SRC 0x00000001 -#define QED_DMAE_FLAG_VF_SRC 0x00000002 -#define QED_DMAE_FLAG_VF_DST 0x00000004 -#define QED_DMAE_FLAG_COMPLETION_DST 0x00000008 -#define QED_DMAE_FLAG_PORT 0x00000010 -#define QED_DMAE_FLAG_PF_SRC 0x00000020 -#define QED_DMAE_FLAG_PF_DST 0x00000040 - -struct qed_dmae_params { - u32 flags; /* consists of QED_DMAE_FLAG_* values */ - u8 src_vfid; - u8 dst_vfid; - u8 port_id; - u8 src_pfid; - u8 dst_pfid; -}; - /** * @brief qed_dmae_host2grc - copy data from source addr to * dmae registers using the given ptt diff --git a/drivers/net/ethernet/qlogic/qed/qed_fcoe.c b/drivers/net/ethernet/qlogic/qed/qed_fcoe.c index de31a382f58e..4c7fa391fd33 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_fcoe.c +++ b/drivers/net/ethernet/qlogic/qed/qed_fcoe.c @@ -167,6 +167,8 @@ qed_sp_fcoe_func_start(struct qed_hwfn *p_hwfn, goto err; } p_cxt = cxt_info.p_cxt; + memset(p_cxt, 0, sizeof(*p_cxt)); + SET_FIELD(p_cxt->tstorm_ag_context.flags3, E4_TSTORM_FCOE_CONN_AG_CTX_DUMMY_TIMER_CF_EN, 1); diff --git a/drivers/net/ethernet/qlogic/qed/qed_hsi.h b/drivers/net/ethernet/qlogic/qed/qed_hsi.h index cf3ceb62e397..4597015b8bff 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_hsi.h +++ b/drivers/net/ethernet/qlogic/qed/qed_hsi.h @@ -98,6 +98,7 @@ enum core_event_opcode { CORE_EVENT_RX_QUEUE_STOP, CORE_EVENT_RX_QUEUE_FLUSH, CORE_EVENT_TX_QUEUE_UPDATE, + CORE_EVENT_QUEUE_STATS_QUERY, MAX_CORE_EVENT_OPCODE }; @@ -116,7 +117,7 @@ struct core_ll2_port_stats { struct regpair gsi_crcchksm_error; }; -/* Ethernet TX Per Queue Stats */ +/* LL2 TX Per Queue Stats */ struct core_ll2_pstorm_per_queue_stat { struct regpair sent_ucast_bytes; struct regpair sent_mcast_bytes; @@ -124,13 +125,13 @@ struct core_ll2_pstorm_per_queue_stat { struct regpair sent_ucast_pkts; struct regpair sent_mcast_pkts; struct regpair sent_bcast_pkts; + struct regpair error_drop_pkts; }; /* Light-L2 RX Producers in Tstorm RAM */ struct core_ll2_rx_prod { __le16 bd_prod; __le16 cqe_prod; - __le32 reserved; }; struct core_ll2_tstorm_per_queue_stat { @@ -147,6 +148,18 @@ struct core_ll2_ustorm_per_queue_stat { struct regpair rcv_bcast_pkts; }; +/* Structure for doorbell data, in PWM mode, for RX producers update. */ +struct core_pwm_prod_update_data { + __le16 icid; /* internal CID */ + u8 reserved0; + u8 params; +#define CORE_PWM_PROD_UPDATE_DATA_AGG_CMD_MASK 0x3 +#define CORE_PWM_PROD_UPDATE_DATA_AGG_CMD_SHIFT 0 +#define CORE_PWM_PROD_UPDATE_DATA_RESERVED1_MASK 0x3F /* Set 0 */ +#define CORE_PWM_PROD_UPDATE_DATA_RESERVED1_SHIFT 2 + struct core_ll2_rx_prod prod; /* Producers */ +}; + /* Core Ramrod Command IDs (light L2) */ enum core_ramrod_cmd_id { CORE_RAMROD_UNUSED, @@ -156,6 +169,7 @@ enum core_ramrod_cmd_id { CORE_RAMROD_TX_QUEUE_STOP, CORE_RAMROD_RX_QUEUE_FLUSH, CORE_RAMROD_TX_QUEUE_UPDATE, + CORE_RAMROD_QUEUE_STATS_QUERY, MAX_CORE_RAMROD_CMD_ID }; @@ -236,7 +250,8 @@ struct core_rx_gsi_offload_cqe { __le16 src_mac_addrlo; __le16 qp_id; __le32 src_qp; - __le32 reserved[3]; + struct core_rx_cqe_opaque_data opaque_data; + __le32 reserved; }; /* Core RX CQE for Light L2 */ @@ -274,8 +289,11 @@ struct core_rx_start_ramrod_data { u8 mf_si_mcast_accept_all; struct core_rx_action_on_error action_on_error; u8 gsi_offload_flag; + u8 vport_id_valid; + u8 vport_id; + u8 zero_prod_flg; u8 wipe_inner_vlan_pri_en; - u8 reserved[5]; + u8 reserved[2]; }; /* Ramrod data for rx queue stop ramrod */ @@ -352,8 +370,11 @@ struct core_tx_start_ramrod_data { __le16 pbl_size; __le16 qm_pq_id; u8 gsi_offload_flag; + u8 ctx_stats_en; + u8 vport_id_valid; u8 vport_id; - u8 resrved[2]; + u8 enforce_security_flag; + u8 reserved[7]; }; /* Ramrod data for tx queue stop ramrod */ @@ -385,7 +406,7 @@ struct ystorm_core_conn_st_ctx { /* The core storm context for the Pstorm */ struct pstorm_core_conn_st_ctx { - __le32 reserved[4]; + __le32 reserved[20]; }; /* Core Slowpath Connection storm context of Xstorm */ @@ -761,7 +782,7 @@ struct e4_tstorm_core_conn_ag_ctx { __le16 word1; __le16 word2; __le16 word3; - __le32 reg9; + __le32 ll2_rx_prod; __le32 reg10; }; @@ -836,11 +857,16 @@ struct e4_ustorm_core_conn_ag_ctx { /* The core storm context for the Mstorm */ struct mstorm_core_conn_st_ctx { - __le32 reserved[24]; + __le32 reserved[40]; }; /* The core storm context for the Ustorm */ struct ustorm_core_conn_st_ctx { + __le32 reserved[20]; +}; + +/* The core storm context for the Tstorm */ +struct tstorm_core_conn_st_ctx { __le32 reserved[4]; }; @@ -857,6 +883,8 @@ struct e4_core_conn_context { struct mstorm_core_conn_st_ctx mstorm_st_context; struct ustorm_core_conn_st_ctx ustorm_st_context; struct regpair ustorm_st_padding[2]; + struct tstorm_core_conn_st_ctx tstorm_st_context; + struct regpair tstorm_st_padding[2]; }; struct eth_mstorm_per_pf_stat { @@ -888,12 +916,21 @@ struct eth_pstorm_per_pf_stat { struct regpair sent_gre_bytes; struct regpair sent_vxlan_bytes; struct regpair sent_geneve_bytes; + struct regpair sent_mpls_bytes; + struct regpair sent_gre_mpls_bytes; + struct regpair sent_udp_mpls_bytes; struct regpair sent_gre_pkts; struct regpair sent_vxlan_pkts; struct regpair sent_geneve_pkts; + struct regpair sent_mpls_pkts; + struct regpair sent_gre_mpls_pkts; + struct regpair sent_udp_mpls_pkts; struct regpair gre_drop_pkts; struct regpair vxlan_drop_pkts; struct regpair geneve_drop_pkts; + struct regpair mpls_drop_pkts; + struct regpair gre_mpls_drop_pkts; + struct regpair udp_mpls_drop_pkts; }; /* Ethernet TX Per Queue Stats */ @@ -983,7 +1020,8 @@ union event_ring_data { struct event_ring_entry { u8 protocol_id; u8 opcode; - __le16 reserved0; + u8 reserved0; + u8 vf_id; __le16 echo; u8 fw_return_code; u8 flags; @@ -1061,7 +1099,20 @@ enum malicious_vf_error_id { ETH_CONTROL_PACKET_VIOLATION, ETH_ANTI_SPOOFING_ERR, ETH_PACKET_SIZE_TOO_LARGE, - MAX_MALICIOUS_VF_ERROR_ID + CORE_ILLEGAL_VLAN_MODE, + CORE_ILLEGAL_NBDS, + CORE_FIRST_BD_WO_SOP, + CORE_INSUFFICIENT_BDS, + CORE_PACKET_TOO_SMALL, + CORE_ILLEGAL_INBAND_TAGS, + CORE_VLAN_INSERT_AND_INBAND_VLAN, + CORE_MTU_VIOLATION, + CORE_CONTROL_PACKET_VIOLATION, + CORE_ANTI_SPOOFING_ERR, + CORE_PACKET_SIZE_TOO_LARGE, + CORE_ILLEGAL_BD_FLAGS, + CORE_GSI_PACKET_VIOLATION, + MAX_MALICIOUS_VF_ERROR_ID, }; /* Mstorm non-triggering VF zone */ @@ -1367,6 +1418,16 @@ enum vf_zone_size_mode { MAX_VF_ZONE_SIZE_MODE }; +/* Xstorm non-triggering VF zone */ +struct xstorm_non_trigger_vf_zone { + struct regpair non_edpm_ack_pkts; +}; + +/* Tstorm VF zone */ +struct xstorm_vf_zone { + struct xstorm_non_trigger_vf_zone non_trigger; +}; + /* Attentions status block */ struct atten_status_block { __le32 atten_bits; @@ -1435,7 +1496,11 @@ struct dmae_cmd { __le16 crc16; __le16 crc16_c; __le16 crc10; - __le16 reserved; + __le16 error_bit_reserved; +#define DMAE_CMD_ERROR_BIT_MASK 0x1 +#define DMAE_CMD_ERROR_BIT_SHIFT 0 +#define DMAE_CMD_RESERVED_MASK 0x7FFF +#define DMAE_CMD_RESERVED_SHIFT 1 __le16 xsum16; __le16 xsum8; }; @@ -1566,6 +1631,41 @@ struct e4_ystorm_core_conn_ag_ctx { __le32 reg3; }; +/* DMAE parameters */ +struct qed_dmae_params { + u32 flags; +/* If QED_DMAE_PARAMS_RW_REPL_SRC flag is set and the + * source is a block of length DMAE_MAX_RW_SIZE and the + * destination is larger, the source block will be duplicated as + * many times as required to fill the destination block. This is + * used mostly to write a zeroed buffer to destination address + * using DMA + */ +#define QED_DMAE_PARAMS_RW_REPL_SRC_MASK 0x1 +#define QED_DMAE_PARAMS_RW_REPL_SRC_SHIFT 0 +#define QED_DMAE_PARAMS_SRC_VF_VALID_MASK 0x1 +#define QED_DMAE_PARAMS_SRC_VF_VALID_SHIFT 1 +#define QED_DMAE_PARAMS_DST_VF_VALID_MASK 0x1 +#define QED_DMAE_PARAMS_DST_VF_VALID_SHIFT 2 +#define QED_DMAE_PARAMS_COMPLETION_DST_MASK 0x1 +#define QED_DMAE_PARAMS_COMPLETION_DST_SHIFT 3 +#define QED_DMAE_PARAMS_PORT_VALID_MASK 0x1 +#define QED_DMAE_PARAMS_PORT_VALID_SHIFT 4 +#define QED_DMAE_PARAMS_SRC_PF_VALID_MASK 0x1 +#define QED_DMAE_PARAMS_SRC_PF_VALID_SHIFT 5 +#define QED_DMAE_PARAMS_DST_PF_VALID_MASK 0x1 +#define QED_DMAE_PARAMS_DST_PF_VALID_SHIFT 6 +#define QED_DMAE_PARAMS_RESERVED_MASK 0x1FFFFFF +#define QED_DMAE_PARAMS_RESERVED_SHIFT 7 + u8 src_vfid; + u8 dst_vfid; + u8 port_id; + u8 src_pfid; + u8 dst_pfid; + u8 reserved1; + __le16 reserved2; +}; + /* IGU cleanup command */ struct igu_cleanup { __le32 sb_id_and_flags; @@ -1743,102 +1843,23 @@ struct sdm_op_gen { #define SDM_OP_GEN_RESERVED_SHIFT 20 }; +/* Physical memory descriptor */ +struct phys_mem_desc { + dma_addr_t phys_addr; + void *virt_addr; + u32 size; /* In bytes */ +}; + +/* Virtual memory descriptor */ +struct virt_mem_desc { + void *ptr; + u32 size; /* In bytes */ +}; + /****************************************/ /* Debug Tools HSI constants and macros */ /****************************************/ -enum block_addr { - GRCBASE_GRC = 0x50000, - GRCBASE_MISCS = 0x9000, - GRCBASE_MISC = 0x8000, - GRCBASE_DBU = 0xa000, - GRCBASE_PGLUE_B = 0x2a8000, - GRCBASE_CNIG = 0x218000, - GRCBASE_CPMU = 0x30000, - GRCBASE_NCSI = 0x40000, - GRCBASE_OPTE = 0x53000, - GRCBASE_BMB = 0x540000, - GRCBASE_PCIE = 0x54000, - GRCBASE_MCP = 0xe00000, - GRCBASE_MCP2 = 0x52000, - GRCBASE_PSWHST = 0x2a0000, - GRCBASE_PSWHST2 = 0x29e000, - GRCBASE_PSWRD = 0x29c000, - GRCBASE_PSWRD2 = 0x29d000, - GRCBASE_PSWWR = 0x29a000, - GRCBASE_PSWWR2 = 0x29b000, - GRCBASE_PSWRQ = 0x280000, - GRCBASE_PSWRQ2 = 0x240000, - GRCBASE_PGLCS = 0x0, - GRCBASE_DMAE = 0xc000, - GRCBASE_PTU = 0x560000, - GRCBASE_TCM = 0x1180000, - GRCBASE_MCM = 0x1200000, - GRCBASE_UCM = 0x1280000, - GRCBASE_XCM = 0x1000000, - GRCBASE_YCM = 0x1080000, - GRCBASE_PCM = 0x1100000, - GRCBASE_QM = 0x2f0000, - GRCBASE_TM = 0x2c0000, - GRCBASE_DORQ = 0x100000, - GRCBASE_BRB = 0x340000, - GRCBASE_SRC = 0x238000, - GRCBASE_PRS = 0x1f0000, - GRCBASE_TSDM = 0xfb0000, - GRCBASE_MSDM = 0xfc0000, - GRCBASE_USDM = 0xfd0000, - GRCBASE_XSDM = 0xf80000, - GRCBASE_YSDM = 0xf90000, - GRCBASE_PSDM = 0xfa0000, - GRCBASE_TSEM = 0x1700000, - GRCBASE_MSEM = 0x1800000, - GRCBASE_USEM = 0x1900000, - GRCBASE_XSEM = 0x1400000, - GRCBASE_YSEM = 0x1500000, - GRCBASE_PSEM = 0x1600000, - GRCBASE_RSS = 0x238800, - GRCBASE_TMLD = 0x4d0000, - GRCBASE_MULD = 0x4e0000, - GRCBASE_YULD = 0x4c8000, - GRCBASE_XYLD = 0x4c0000, - GRCBASE_PTLD = 0x5a0000, - GRCBASE_YPLD = 0x5c0000, - GRCBASE_PRM = 0x230000, - GRCBASE_PBF_PB1 = 0xda0000, - GRCBASE_PBF_PB2 = 0xda4000, - GRCBASE_RPB = 0x23c000, - GRCBASE_BTB = 0xdb0000, - GRCBASE_PBF = 0xd80000, - GRCBASE_RDIF = 0x300000, - GRCBASE_TDIF = 0x310000, - GRCBASE_CDU = 0x580000, - GRCBASE_CCFC = 0x2e0000, - GRCBASE_TCFC = 0x2d0000, - GRCBASE_IGU = 0x180000, - GRCBASE_CAU = 0x1c0000, - GRCBASE_RGFS = 0xf00000, - GRCBASE_RGSRC = 0x320000, - GRCBASE_TGFS = 0xd00000, - GRCBASE_TGSRC = 0x322000, - GRCBASE_UMAC = 0x51000, - GRCBASE_XMAC = 0x210000, - GRCBASE_DBG = 0x10000, - GRCBASE_NIG = 0x500000, - GRCBASE_WOL = 0x600000, - GRCBASE_BMBN = 0x610000, - GRCBASE_IPC = 0x20000, - GRCBASE_NWM = 0x800000, - GRCBASE_NWS = 0x700000, - GRCBASE_MS = 0x6a0000, - GRCBASE_PHY_PCIE = 0x620000, - GRCBASE_LED = 0x6b8000, - GRCBASE_AVS_WRAP = 0x6b0000, - GRCBASE_PXPREQBUS = 0x56000, - GRCBASE_MISC_AEU = 0x8000, - GRCBASE_BAR0_MAP = 0x1c00000, - MAX_BLOCK_ADDR -}; - enum block_id { BLOCK_GRC, BLOCK_MISCS, @@ -1893,8 +1914,6 @@ enum block_id { BLOCK_MULD, BLOCK_YULD, BLOCK_XYLD, - BLOCK_PTLD, - BLOCK_YPLD, BLOCK_PRM, BLOCK_PBF_PB1, BLOCK_PBF_PB2, @@ -1908,12 +1927,9 @@ enum block_id { BLOCK_TCFC, BLOCK_IGU, BLOCK_CAU, - BLOCK_RGFS, - BLOCK_RGSRC, - BLOCK_TGFS, - BLOCK_TGSRC, BLOCK_UMAC, BLOCK_XMAC, + BLOCK_MSTAT, BLOCK_DBG, BLOCK_NIG, BLOCK_WOL, @@ -1926,8 +1942,17 @@ enum block_id { BLOCK_LED, BLOCK_AVS_WRAP, BLOCK_PXPREQBUS, - BLOCK_MISC_AEU, BLOCK_BAR0_MAP, + BLOCK_MCP_FIO, + BLOCK_LAST_INIT, + BLOCK_PRS_FC, + BLOCK_PBF_FC, + BLOCK_NIG_LB_FC, + BLOCK_NIG_LB_FC_PLLH, + BLOCK_NIG_TX_FC_PLLH, + BLOCK_NIG_TX_FC, + BLOCK_NIG_RX_FC_PLLH, + BLOCK_NIG_RX_FC, MAX_BLOCK_ID }; @@ -1944,10 +1969,13 @@ enum bin_dbg_buffer_type { BIN_BUF_DBG_ATTN_REGS, BIN_BUF_DBG_ATTN_INDEXES, BIN_BUF_DBG_ATTN_NAME_OFFSETS, - BIN_BUF_DBG_BUS_BLOCKS, + BIN_BUF_DBG_BLOCKS, + BIN_BUF_DBG_BLOCKS_CHIP_DATA, BIN_BUF_DBG_BUS_LINES, - BIN_BUF_DBG_BUS_BLOCKS_USER_DATA, + BIN_BUF_DBG_BLOCKS_USER_DATA, + BIN_BUF_DBG_BLOCKS_CHIP_USER_DATA, BIN_BUF_DBG_BUS_LINE_NAME_OFFSETS, + BIN_BUF_DBG_RESET_REGS, BIN_BUF_DBG_PARSING_STRINGS, MAX_BIN_DBG_BUFFER_TYPE }; @@ -2031,20 +2059,54 @@ enum dbg_attn_type { MAX_DBG_ATTN_TYPE }; -/* Debug Bus block data */ -struct dbg_bus_block { - u8 num_of_lines; - u8 has_latency_events; - u16 lines_offset; +/* Block debug data */ +struct dbg_block { + u8 name[15]; + u8 associated_storm_letter; }; -/* Debug Bus block user data */ -struct dbg_bus_block_user_data { - u8 num_of_lines; +/* Chip-specific block debug data */ +struct dbg_block_chip { + u8 flags; +#define DBG_BLOCK_CHIP_IS_REMOVED_MASK 0x1 +#define DBG_BLOCK_CHIP_IS_REMOVED_SHIFT 0 +#define DBG_BLOCK_CHIP_HAS_RESET_REG_MASK 0x1 +#define DBG_BLOCK_CHIP_HAS_RESET_REG_SHIFT 1 +#define DBG_BLOCK_CHIP_UNRESET_BEFORE_DUMP_MASK 0x1 +#define DBG_BLOCK_CHIP_UNRESET_BEFORE_DUMP_SHIFT 2 +#define DBG_BLOCK_CHIP_HAS_DBG_BUS_MASK 0x1 +#define DBG_BLOCK_CHIP_HAS_DBG_BUS_SHIFT 3 +#define DBG_BLOCK_CHIP_HAS_LATENCY_EVENTS_MASK 0x1 +#define DBG_BLOCK_CHIP_HAS_LATENCY_EVENTS_SHIFT 4 +#define DBG_BLOCK_CHIP_RESERVED0_MASK 0x7 +#define DBG_BLOCK_CHIP_RESERVED0_SHIFT 5 + u8 dbg_client_id; + u8 reset_reg_id; + u8 reset_reg_bit_offset; + struct dbg_mode_hdr dbg_bus_mode; + u16 reserved1; + u8 reserved2; + u8 num_of_dbg_bus_lines; + u16 dbg_bus_lines_offset; + u32 dbg_select_reg_addr; + u32 dbg_dword_enable_reg_addr; + u32 dbg_shift_reg_addr; + u32 dbg_force_valid_reg_addr; + u32 dbg_force_frame_reg_addr; +}; + +/* Chip-specific block user debug data */ +struct dbg_block_chip_user { + u8 num_of_dbg_bus_lines; u8 has_latency_events; u16 names_offset; }; +/* Block user debug data */ +struct dbg_block_user { + u8 name[16]; +}; + /* Block Debug line data */ struct dbg_bus_line { u8 data; @@ -2197,22 +2259,33 @@ enum dbg_idle_chk_severity_types { MAX_DBG_IDLE_CHK_SEVERITY_TYPES }; +/* Reset register */ +struct dbg_reset_reg { + u32 data; +#define DBG_RESET_REG_ADDR_MASK 0xFFFFFF +#define DBG_RESET_REG_ADDR_SHIFT 0 +#define DBG_RESET_REG_IS_REMOVED_MASK 0x1 +#define DBG_RESET_REG_IS_REMOVED_SHIFT 24 +#define DBG_RESET_REG_RESERVED_MASK 0x7F +#define DBG_RESET_REG_RESERVED_SHIFT 25 +}; + /* Debug Bus block data */ struct dbg_bus_block_data { - u16 data; -#define DBG_BUS_BLOCK_DATA_ENABLE_MASK_MASK 0xF -#define DBG_BUS_BLOCK_DATA_ENABLE_MASK_SHIFT 0 -#define DBG_BUS_BLOCK_DATA_RIGHT_SHIFT_MASK 0xF -#define DBG_BUS_BLOCK_DATA_RIGHT_SHIFT_SHIFT 4 -#define DBG_BUS_BLOCK_DATA_FORCE_VALID_MASK_MASK 0xF -#define DBG_BUS_BLOCK_DATA_FORCE_VALID_MASK_SHIFT 8 -#define DBG_BUS_BLOCK_DATA_FORCE_FRAME_MASK_MASK 0xF -#define DBG_BUS_BLOCK_DATA_FORCE_FRAME_MASK_SHIFT 12 + u8 enable_mask; + u8 right_shift; + u8 force_valid_mask; + u8 force_frame_mask; + u8 dword_mask; u8 line_num; u8 hw_id; + u8 flags; +#define DBG_BUS_BLOCK_DATA_IS_256B_LINE_MASK 0x1 +#define DBG_BUS_BLOCK_DATA_IS_256B_LINE_SHIFT 0 +#define DBG_BUS_BLOCK_DATA_RESERVED_MASK 0x7F +#define DBG_BUS_BLOCK_DATA_RESERVED_SHIFT 1 }; -/* Debug Bus Clients */ enum dbg_bus_clients { DBG_BUS_CLIENT_RBCN, DBG_BUS_CLIENT_RBCP, @@ -2253,11 +2326,10 @@ enum dbg_bus_constraint_ops { /* Debug Bus trigger state data */ struct dbg_bus_trigger_state_data { - u8 data; -#define DBG_BUS_TRIGGER_STATE_DATA_BLOCK_SHIFTED_ENABLE_MASK_MASK 0xF -#define DBG_BUS_TRIGGER_STATE_DATA_BLOCK_SHIFTED_ENABLE_MASK_SHIFT 0 -#define DBG_BUS_TRIGGER_STATE_DATA_CONSTRAINT_DWORD_MASK_MASK 0xF -#define DBG_BUS_TRIGGER_STATE_DATA_CONSTRAINT_DWORD_MASK_SHIFT 4 + u8 msg_len; + u8 constraint_dword_mask; + u8 storm_id; + u8 reserved; }; /* Debug Bus memory address */ @@ -2307,8 +2379,7 @@ struct dbg_bus_storm_data { struct dbg_bus_data { u32 app_version; u8 state; - u8 hw_dwords; - u16 hw_id_mask; + u8 mode_256b_en; u8 num_enabled_blocks; u8 num_enabled_storms; u8 target; @@ -2319,67 +2390,21 @@ struct dbg_bus_data { u8 adding_filter; u8 filter_pre_trigger; u8 filter_post_trigger; - u16 reserved; u8 trigger_en; - struct dbg_bus_trigger_state_data trigger_states[3]; + u8 filter_constraint_dword_mask; u8 next_trigger_state; u8 next_constraint_id; - u8 unify_inputs; + struct dbg_bus_trigger_state_data trigger_states[3]; + u8 filter_msg_len; u8 rcv_from_other_engine; + u8 blocks_dword_mask; + u8 blocks_dword_overlap; + u32 hw_id_mask; struct dbg_bus_pci_buf_data pci_buf; - struct dbg_bus_block_data blocks[88]; + struct dbg_bus_block_data blocks[132]; struct dbg_bus_storm_data storms[6]; }; -/* Debug bus filter types */ -enum dbg_bus_filter_types { - DBG_BUS_FILTER_TYPE_OFF, - DBG_BUS_FILTER_TYPE_PRE, - DBG_BUS_FILTER_TYPE_POST, - DBG_BUS_FILTER_TYPE_ON, - MAX_DBG_BUS_FILTER_TYPES -}; - -/* Debug bus frame modes */ -enum dbg_bus_frame_modes { - DBG_BUS_FRAME_MODE_0HW_4ST = 0, /* 0 HW dwords, 4 Storm dwords */ - DBG_BUS_FRAME_MODE_4HW_0ST = 3, /* 4 HW dwords, 0 Storm dwords */ - DBG_BUS_FRAME_MODE_8HW_0ST = 4, /* 8 HW dwords, 0 Storm dwords */ - MAX_DBG_BUS_FRAME_MODES -}; - -/* Debug bus other engine mode */ -enum dbg_bus_other_engine_modes { - DBG_BUS_OTHER_ENGINE_MODE_NONE, - DBG_BUS_OTHER_ENGINE_MODE_DOUBLE_BW_TX, - DBG_BUS_OTHER_ENGINE_MODE_DOUBLE_BW_RX, - DBG_BUS_OTHER_ENGINE_MODE_CROSS_ENGINE_TX, - DBG_BUS_OTHER_ENGINE_MODE_CROSS_ENGINE_RX, - MAX_DBG_BUS_OTHER_ENGINE_MODES -}; - -/* Debug bus post-trigger recording types */ -enum dbg_bus_post_trigger_types { - DBG_BUS_POST_TRIGGER_RECORD, - DBG_BUS_POST_TRIGGER_DROP, - MAX_DBG_BUS_POST_TRIGGER_TYPES -}; - -/* Debug bus pre-trigger recording types */ -enum dbg_bus_pre_trigger_types { - DBG_BUS_PRE_TRIGGER_START_FROM_ZERO, - DBG_BUS_PRE_TRIGGER_NUM_CHUNKS, - DBG_BUS_PRE_TRIGGER_DROP, - MAX_DBG_BUS_PRE_TRIGGER_TYPES -}; - -/* Debug bus SEMI frame modes */ -enum dbg_bus_semi_frame_modes { - DBG_BUS_SEMI_FRAME_MODE_0SLOW_4FAST = 0, - DBG_BUS_SEMI_FRAME_MODE_4SLOW_0FAST = 3, - MAX_DBG_BUS_SEMI_FRAME_MODES -}; - /* Debug bus states */ enum dbg_bus_states { DBG_BUS_STATE_IDLE, @@ -2397,7 +2422,9 @@ enum dbg_bus_storm_modes { DBG_BUS_STORM_MODE_DRA_W, DBG_BUS_STORM_MODE_LD_ST_ADDR, DBG_BUS_STORM_MODE_DRA_FSM, + DBG_BUS_STORM_MODE_FAST_DBGMUX, DBG_BUS_STORM_MODE_RH, + DBG_BUS_STORM_MODE_RH_WITH_STORE, DBG_BUS_STORM_MODE_FOC, DBG_BUS_STORM_MODE_EXT_STORE, MAX_DBG_BUS_STORM_MODES @@ -2438,13 +2465,13 @@ enum dbg_grc_params { DBG_GRC_PARAM_DUMP_CAU, DBG_GRC_PARAM_DUMP_QM, DBG_GRC_PARAM_DUMP_MCP, - DBG_GRC_PARAM_MCP_TRACE_META_SIZE, + DBG_GRC_PARAM_DUMP_DORQ, DBG_GRC_PARAM_DUMP_CFC, DBG_GRC_PARAM_DUMP_IGU, DBG_GRC_PARAM_DUMP_BRB, DBG_GRC_PARAM_DUMP_BTB, DBG_GRC_PARAM_DUMP_BMB, - DBG_GRC_PARAM_DUMP_NIG, + DBG_GRC_PARAM_RESERVD1, DBG_GRC_PARAM_DUMP_MULD, DBG_GRC_PARAM_DUMP_PRS, DBG_GRC_PARAM_DUMP_DMAE, @@ -2453,8 +2480,8 @@ enum dbg_grc_params { DBG_GRC_PARAM_DUMP_DIF, DBG_GRC_PARAM_DUMP_STATIC, DBG_GRC_PARAM_UNSTALL, - DBG_GRC_PARAM_NUM_LCIDS, - DBG_GRC_PARAM_NUM_LTIDS, + DBG_GRC_PARAM_RESERVED2, + DBG_GRC_PARAM_MCP_TRACE_META_SIZE, DBG_GRC_PARAM_EXCLUDE_ALL, DBG_GRC_PARAM_CRASH, DBG_GRC_PARAM_PARITY_SAFE, @@ -2462,22 +2489,14 @@ enum dbg_grc_params { DBG_GRC_PARAM_DUMP_PHY, DBG_GRC_PARAM_NO_MCP, DBG_GRC_PARAM_NO_FW_VER, + DBG_GRC_PARAM_RESERVED3, + DBG_GRC_PARAM_DUMP_MCP_HW_DUMP, + DBG_GRC_PARAM_DUMP_ILT_CDUC, + DBG_GRC_PARAM_DUMP_ILT_CDUT, + DBG_GRC_PARAM_DUMP_CAU_EXT, MAX_DBG_GRC_PARAMS }; -/* Debug reset registers */ -enum dbg_reset_regs { - DBG_RESET_REG_MISCS_PL_UA, - DBG_RESET_REG_MISCS_PL_HV, - DBG_RESET_REG_MISCS_PL_HV_2, - DBG_RESET_REG_MISC_PL_UA, - DBG_RESET_REG_MISC_PL_HV, - DBG_RESET_REG_MISC_PL_PDA_VMAIN_1, - DBG_RESET_REG_MISC_PL_PDA_VMAIN_2, - DBG_RESET_REG_MISC_PL_PDA_VAUX, - MAX_DBG_RESET_REGS -}; - /* Debug status codes */ enum dbg_status { DBG_STATUS_OK, @@ -2489,15 +2508,15 @@ enum dbg_status { DBG_STATUS_INVALID_PCI_BUF_SIZE, DBG_STATUS_PCI_BUF_ALLOC_FAILED, DBG_STATUS_PCI_BUF_NOT_ALLOCATED, - DBG_STATUS_TOO_MANY_INPUTS, - DBG_STATUS_INPUT_OVERLAP, - DBG_STATUS_HW_ONLY_RECORDING, + DBG_STATUS_INVALID_FILTER_TRIGGER_DWORDS, + DBG_STATUS_NO_MATCHING_FRAMING_MODE, + DBG_STATUS_VFC_READ_ERROR, DBG_STATUS_STORM_ALREADY_ENABLED, DBG_STATUS_STORM_NOT_ENABLED, DBG_STATUS_BLOCK_ALREADY_ENABLED, DBG_STATUS_BLOCK_NOT_ENABLED, DBG_STATUS_NO_INPUT_ENABLED, - DBG_STATUS_NO_FILTER_TRIGGER_64B, + DBG_STATUS_NO_FILTER_TRIGGER_256B, DBG_STATUS_FILTER_ALREADY_ENABLED, DBG_STATUS_TRIGGER_ALREADY_ENABLED, DBG_STATUS_TRIGGER_NOT_ENABLED, @@ -2522,7 +2541,7 @@ enum dbg_status { DBG_STATUS_MCP_TRACE_NO_META, DBG_STATUS_MCP_COULD_NOT_HALT, DBG_STATUS_MCP_COULD_NOT_RESUME, - DBG_STATUS_RESERVED2, + DBG_STATUS_RESERVED0, DBG_STATUS_SEMI_FIFO_NOT_EMPTY, DBG_STATUS_IGU_FIFO_BAD_DATA, DBG_STATUS_MCP_COULD_NOT_MASK_PRTY, @@ -2530,10 +2549,15 @@ enum dbg_status { DBG_STATUS_REG_FIFO_BAD_DATA, DBG_STATUS_PROTECTION_OVERRIDE_BAD_DATA, DBG_STATUS_DBG_ARRAY_NOT_SET, - DBG_STATUS_FILTER_BUG, + DBG_STATUS_RESERVED1, DBG_STATUS_NON_MATCHING_LINES, - DBG_STATUS_INVALID_TRIGGER_DWORD_OFFSET, + DBG_STATUS_INSUFFICIENT_HW_IDS, DBG_STATUS_DBG_BUS_IN_USE, + DBG_STATUS_INVALID_STORM_DBG_MODE, + DBG_STATUS_OTHER_ENGINE_BB_ONLY, + DBG_STATUS_FILTER_SINGLE_HW_ID, + DBG_STATUS_TRIGGER_SINGLE_HW_ID, + DBG_STATUS_MISSING_TRIGGER_STATE_STORM, MAX_DBG_STATUS }; @@ -2569,9 +2593,9 @@ struct dbg_tools_data { struct dbg_bus_data bus; struct idle_chk_data idle_chk; u8 mode_enable[40]; - u8 block_in_reset[88]; + u8 block_in_reset[132]; u8 chip_id; - u8 platform_id; + u8 hw_type; u8 num_ports; u8 num_pfs_per_port; u8 num_vfs; @@ -2582,6 +2606,19 @@ struct dbg_tools_data { u32 num_regs_read; }; +/* ILT Clients */ +enum ilt_clients { + ILT_CLI_CDUC, + ILT_CLI_CDUT, + ILT_CLI_QM, + ILT_CLI_TM, + ILT_CLI_SRC, + ILT_CLI_TSDM, + ILT_CLI_RGFS, + ILT_CLI_TGFS, + MAX_ILT_CLIENTS +}; + /********************************/ /* HSI Init Functions constants */ /********************************/ @@ -2630,13 +2667,18 @@ struct init_nig_pri_tc_map_req { struct init_nig_pri_tc_map_entry pri[NUM_OF_VLAN_PRIORITIES]; }; +/* QM per global RL init parameters */ +struct init_qm_global_rl_params { + u32 rate_limit; +}; + /* QM per-port init parameters */ struct init_qm_port_params { - u8 active; - u8 active_phys_tcs; + u16 active_phys_tcs; u16 num_pbf_cmd_lines; u16 num_btb_blocks; - u16 reserved; + u8 active; + u8 reserved; }; /* QM per-PQ init parameters */ @@ -2645,15 +2687,14 @@ struct init_qm_pq_params { u8 tc_id; u8 wrr_group; u8 rl_valid; + u16 rl_id; u8 port_id; - u8 reserved0; - u16 reserved1; + u8 reserved; }; /* QM per-vport init parameters */ struct init_qm_vport_params { - u32 vport_rl; - u16 vport_wfq; + u16 wfq; u16 first_tx_pq_id[NUM_OF_TCS]; }; @@ -2673,13 +2714,12 @@ struct init_qm_vport_params { enum chip_ids { CHIP_BB, CHIP_K2, - CHIP_RESERVED, MAX_CHIP_IDS }; struct fw_asserts_ram_section { - u16 section_ram_line_offset; - u16 section_ram_line_size; + __le16 section_ram_line_offset; + __le16 section_ram_line_size; u8 list_dword_offset; u8 list_element_dword_size; u8 list_num_elements; @@ -2729,6 +2769,7 @@ enum init_modes { MODE_PORTS_PER_ENG_4, MODE_100G, MODE_RESERVED6, + MODE_RESERVED7, MAX_INIT_MODES }; @@ -2763,9 +2804,19 @@ enum bin_init_buffer_type { BIN_BUF_INIT_VAL, BIN_BUF_INIT_MODE_TREE, BIN_BUF_INIT_IRO, + BIN_BUF_INIT_OVERLAYS, MAX_BIN_INIT_BUFFER_TYPE }; +/* FW overlay buffer header */ +struct fw_overlay_buf_hdr { + u32 data; +#define FW_OVERLAY_BUF_HDR_STORM_ID_MASK 0xFF +#define FW_OVERLAY_BUF_HDR_STORM_ID_SHIFT 0 +#define FW_OVERLAY_BUF_HDR_BUF_SIZE_MASK 0xFFFFFF +#define FW_OVERLAY_BUF_HDR_BUF_SIZE_SHIFT 8 +}; + /* init array header: raw */ struct init_array_raw_hdr { u32 data; @@ -2859,10 +2910,8 @@ struct init_if_phase_op { u32 op_data; #define INIT_IF_PHASE_OP_OP_MASK 0xF #define INIT_IF_PHASE_OP_OP_SHIFT 0 -#define INIT_IF_PHASE_OP_DMAE_ENABLE_MASK 0x1 -#define INIT_IF_PHASE_OP_DMAE_ENABLE_SHIFT 4 -#define INIT_IF_PHASE_OP_RESERVED1_MASK 0x7FF -#define INIT_IF_PHASE_OP_RESERVED1_SHIFT 5 +#define INIT_IF_PHASE_OP_RESERVED1_MASK 0xFFF +#define INIT_IF_PHASE_OP_RESERVED1_SHIFT 4 #define INIT_IF_PHASE_OP_CMD_OFFSET_MASK 0xFFFF #define INIT_IF_PHASE_OP_CMD_OFFSET_SHIFT 16 u32 phase_data; @@ -2991,9 +3040,11 @@ struct iro { * @brief qed_dbg_set_bin_ptr - Sets a pointer to the binary data with debug * arrays. * + * @param p_hwfn - HW device data * @param bin_ptr - a pointer to the binary data with debug arrays. */ -enum dbg_status qed_dbg_set_bin_ptr(const u8 * const bin_ptr); +enum dbg_status qed_dbg_set_bin_ptr(struct qed_hwfn *p_hwfn, + const u8 * const bin_ptr); /** * @brief qed_read_regs - Reads registers into a buffer (using GRC). @@ -3037,7 +3088,6 @@ bool qed_read_fw_info(struct qed_hwfn *p_hwfn, * - val is outside the allowed boundaries */ enum dbg_status qed_dbg_grc_config(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, enum dbg_grc_params grc_param, u32 val); /** @@ -3358,20 +3408,36 @@ enum dbg_status qed_dbg_print_attn(struct qed_hwfn *p_hwfn, struct mcp_trace_format { u32 data; #define MCP_TRACE_FORMAT_MODULE_MASK 0x0000ffff -#define MCP_TRACE_FORMAT_MODULE_SHIFT 0 +#define MCP_TRACE_FORMAT_MODULE_OFFSET 0 #define MCP_TRACE_FORMAT_LEVEL_MASK 0x00030000 -#define MCP_TRACE_FORMAT_LEVEL_SHIFT 16 +#define MCP_TRACE_FORMAT_LEVEL_OFFSET 16 #define MCP_TRACE_FORMAT_P1_SIZE_MASK 0x000c0000 -#define MCP_TRACE_FORMAT_P1_SIZE_SHIFT 18 +#define MCP_TRACE_FORMAT_P1_SIZE_OFFSET 18 #define MCP_TRACE_FORMAT_P2_SIZE_MASK 0x00300000 -#define MCP_TRACE_FORMAT_P2_SIZE_SHIFT 20 +#define MCP_TRACE_FORMAT_P2_SIZE_OFFSET 20 #define MCP_TRACE_FORMAT_P3_SIZE_MASK 0x00c00000 -#define MCP_TRACE_FORMAT_P3_SIZE_SHIFT 22 +#define MCP_TRACE_FORMAT_P3_SIZE_OFFSET 22 #define MCP_TRACE_FORMAT_LEN_MASK 0xff000000 -#define MCP_TRACE_FORMAT_LEN_SHIFT 24 +#define MCP_TRACE_FORMAT_LEN_OFFSET 24 + char *format_str; }; +/* MCP Trace Meta data structure */ +struct mcp_trace_meta { + u32 modules_num; + char **modules; + u32 formats_num; + struct mcp_trace_format *formats; + bool is_allocated; +}; + +/* Debug Tools user data */ +struct dbg_tools_user_data { + struct mcp_trace_meta mcp_trace_meta; + const u32 *mcp_trace_user_meta_buf; +}; + /******************************** Constants **********************************/ #define MAX_NAME_LEN 16 @@ -3382,16 +3448,20 @@ struct mcp_trace_format { * @brief qed_dbg_user_set_bin_ptr - Sets a pointer to the binary data with * debug arrays. * + * @param p_hwfn - HW device data * @param bin_ptr - a pointer to the binary data with debug arrays. */ -enum dbg_status qed_dbg_user_set_bin_ptr(const u8 * const bin_ptr); +enum dbg_status qed_dbg_user_set_bin_ptr(struct qed_hwfn *p_hwfn, + const u8 * const bin_ptr); /** * @brief qed_dbg_alloc_user_data - Allocates user debug data. * * @param p_hwfn - HW device data + * @param user_data_ptr - OUT: a pointer to the allocated memory. */ -enum dbg_status qed_dbg_alloc_user_data(struct qed_hwfn *p_hwfn); +enum dbg_status qed_dbg_alloc_user_data(struct qed_hwfn *p_hwfn, + void **user_data_ptr); /** * @brief qed_dbg_get_status_str - Returns a string for the specified status. @@ -3664,271 +3734,6 @@ enum dbg_status qed_print_fw_asserts_results(struct qed_hwfn *p_hwfn, enum dbg_status qed_dbg_parse_attn(struct qed_hwfn *p_hwfn, struct dbg_attn_block_result *results); -/* Debug Bus blocks */ -static const u32 dbg_bus_blocks[] = { - 0x0000000f, /* grc, bb, 15 lines */ - 0x0000000f, /* grc, k2, 15 lines */ - 0x00000000, - 0x00000000, /* miscs, bb, 0 lines */ - 0x00000000, /* miscs, k2, 0 lines */ - 0x00000000, - 0x00000000, /* misc, bb, 0 lines */ - 0x00000000, /* misc, k2, 0 lines */ - 0x00000000, - 0x00000000, /* dbu, bb, 0 lines */ - 0x00000000, /* dbu, k2, 0 lines */ - 0x00000000, - 0x000f0127, /* pglue_b, bb, 39 lines */ - 0x0036012a, /* pglue_b, k2, 42 lines */ - 0x00000000, - 0x00000000, /* cnig, bb, 0 lines */ - 0x00120102, /* cnig, k2, 2 lines */ - 0x00000000, - 0x00000000, /* cpmu, bb, 0 lines */ - 0x00000000, /* cpmu, k2, 0 lines */ - 0x00000000, - 0x00000001, /* ncsi, bb, 1 lines */ - 0x00000001, /* ncsi, k2, 1 lines */ - 0x00000000, - 0x00000000, /* opte, bb, 0 lines */ - 0x00000000, /* opte, k2, 0 lines */ - 0x00000000, - 0x00600085, /* bmb, bb, 133 lines */ - 0x00600085, /* bmb, k2, 133 lines */ - 0x00000000, - 0x00000000, /* pcie, bb, 0 lines */ - 0x00e50033, /* pcie, k2, 51 lines */ - 0x00000000, - 0x00000000, /* mcp, bb, 0 lines */ - 0x00000000, /* mcp, k2, 0 lines */ - 0x00000000, - 0x01180009, /* mcp2, bb, 9 lines */ - 0x01180009, /* mcp2, k2, 9 lines */ - 0x00000000, - 0x01210104, /* pswhst, bb, 4 lines */ - 0x01210104, /* pswhst, k2, 4 lines */ - 0x00000000, - 0x01250103, /* pswhst2, bb, 3 lines */ - 0x01250103, /* pswhst2, k2, 3 lines */ - 0x00000000, - 0x00340101, /* pswrd, bb, 1 lines */ - 0x00340101, /* pswrd, k2, 1 lines */ - 0x00000000, - 0x01280119, /* pswrd2, bb, 25 lines */ - 0x01280119, /* pswrd2, k2, 25 lines */ - 0x00000000, - 0x01410109, /* pswwr, bb, 9 lines */ - 0x01410109, /* pswwr, k2, 9 lines */ - 0x00000000, - 0x00000000, /* pswwr2, bb, 0 lines */ - 0x00000000, /* pswwr2, k2, 0 lines */ - 0x00000000, - 0x001c0001, /* pswrq, bb, 1 lines */ - 0x001c0001, /* pswrq, k2, 1 lines */ - 0x00000000, - 0x014a0015, /* pswrq2, bb, 21 lines */ - 0x014a0015, /* pswrq2, k2, 21 lines */ - 0x00000000, - 0x00000000, /* pglcs, bb, 0 lines */ - 0x00120006, /* pglcs, k2, 6 lines */ - 0x00000000, - 0x00100001, /* dmae, bb, 1 lines */ - 0x00100001, /* dmae, k2, 1 lines */ - 0x00000000, - 0x015f0105, /* ptu, bb, 5 lines */ - 0x015f0105, /* ptu, k2, 5 lines */ - 0x00000000, - 0x01640120, /* tcm, bb, 32 lines */ - 0x01640120, /* tcm, k2, 32 lines */ - 0x00000000, - 0x01640120, /* mcm, bb, 32 lines */ - 0x01640120, /* mcm, k2, 32 lines */ - 0x00000000, - 0x01640120, /* ucm, bb, 32 lines */ - 0x01640120, /* ucm, k2, 32 lines */ - 0x00000000, - 0x01640120, /* xcm, bb, 32 lines */ - 0x01640120, /* xcm, k2, 32 lines */ - 0x00000000, - 0x01640120, /* ycm, bb, 32 lines */ - 0x01640120, /* ycm, k2, 32 lines */ - 0x00000000, - 0x01640120, /* pcm, bb, 32 lines */ - 0x01640120, /* pcm, k2, 32 lines */ - 0x00000000, - 0x01840062, /* qm, bb, 98 lines */ - 0x01840062, /* qm, k2, 98 lines */ - 0x00000000, - 0x01e60021, /* tm, bb, 33 lines */ - 0x01e60021, /* tm, k2, 33 lines */ - 0x00000000, - 0x02070107, /* dorq, bb, 7 lines */ - 0x02070107, /* dorq, k2, 7 lines */ - 0x00000000, - 0x00600185, /* brb, bb, 133 lines */ - 0x00600185, /* brb, k2, 133 lines */ - 0x00000000, - 0x020e0019, /* src, bb, 25 lines */ - 0x020c001a, /* src, k2, 26 lines */ - 0x00000000, - 0x02270104, /* prs, bb, 4 lines */ - 0x02270104, /* prs, k2, 4 lines */ - 0x00000000, - 0x022b0133, /* tsdm, bb, 51 lines */ - 0x022b0133, /* tsdm, k2, 51 lines */ - 0x00000000, - 0x022b0133, /* msdm, bb, 51 lines */ - 0x022b0133, /* msdm, k2, 51 lines */ - 0x00000000, - 0x022b0133, /* usdm, bb, 51 lines */ - 0x022b0133, /* usdm, k2, 51 lines */ - 0x00000000, - 0x022b0133, /* xsdm, bb, 51 lines */ - 0x022b0133, /* xsdm, k2, 51 lines */ - 0x00000000, - 0x022b0133, /* ysdm, bb, 51 lines */ - 0x022b0133, /* ysdm, k2, 51 lines */ - 0x00000000, - 0x022b0133, /* psdm, bb, 51 lines */ - 0x022b0133, /* psdm, k2, 51 lines */ - 0x00000000, - 0x025e010c, /* tsem, bb, 12 lines */ - 0x025e010c, /* tsem, k2, 12 lines */ - 0x00000000, - 0x025e010c, /* msem, bb, 12 lines */ - 0x025e010c, /* msem, k2, 12 lines */ - 0x00000000, - 0x025e010c, /* usem, bb, 12 lines */ - 0x025e010c, /* usem, k2, 12 lines */ - 0x00000000, - 0x025e010c, /* xsem, bb, 12 lines */ - 0x025e010c, /* xsem, k2, 12 lines */ - 0x00000000, - 0x025e010c, /* ysem, bb, 12 lines */ - 0x025e010c, /* ysem, k2, 12 lines */ - 0x00000000, - 0x025e010c, /* psem, bb, 12 lines */ - 0x025e010c, /* psem, k2, 12 lines */ - 0x00000000, - 0x026a000d, /* rss, bb, 13 lines */ - 0x026a000d, /* rss, k2, 13 lines */ - 0x00000000, - 0x02770106, /* tmld, bb, 6 lines */ - 0x02770106, /* tmld, k2, 6 lines */ - 0x00000000, - 0x027d0106, /* muld, bb, 6 lines */ - 0x027d0106, /* muld, k2, 6 lines */ - 0x00000000, - 0x02770005, /* yuld, bb, 5 lines */ - 0x02770005, /* yuld, k2, 5 lines */ - 0x00000000, - 0x02830107, /* xyld, bb, 7 lines */ - 0x027d0107, /* xyld, k2, 7 lines */ - 0x00000000, - 0x00000000, /* ptld, bb, 0 lines */ - 0x00000000, /* ptld, k2, 0 lines */ - 0x00000000, - 0x00000000, /* ypld, bb, 0 lines */ - 0x00000000, /* ypld, k2, 0 lines */ - 0x00000000, - 0x028a010e, /* prm, bb, 14 lines */ - 0x02980110, /* prm, k2, 16 lines */ - 0x00000000, - 0x02a8000d, /* pbf_pb1, bb, 13 lines */ - 0x02a8000d, /* pbf_pb1, k2, 13 lines */ - 0x00000000, - 0x02a8000d, /* pbf_pb2, bb, 13 lines */ - 0x02a8000d, /* pbf_pb2, k2, 13 lines */ - 0x00000000, - 0x02a8000d, /* rpb, bb, 13 lines */ - 0x02a8000d, /* rpb, k2, 13 lines */ - 0x00000000, - 0x00600185, /* btb, bb, 133 lines */ - 0x00600185, /* btb, k2, 133 lines */ - 0x00000000, - 0x02b50117, /* pbf, bb, 23 lines */ - 0x02b50117, /* pbf, k2, 23 lines */ - 0x00000000, - 0x02cc0006, /* rdif, bb, 6 lines */ - 0x02cc0006, /* rdif, k2, 6 lines */ - 0x00000000, - 0x02d20006, /* tdif, bb, 6 lines */ - 0x02d20006, /* tdif, k2, 6 lines */ - 0x00000000, - 0x02d80003, /* cdu, bb, 3 lines */ - 0x02db000e, /* cdu, k2, 14 lines */ - 0x00000000, - 0x02e9010d, /* ccfc, bb, 13 lines */ - 0x02f60117, /* ccfc, k2, 23 lines */ - 0x00000000, - 0x02e9010d, /* tcfc, bb, 13 lines */ - 0x02f60117, /* tcfc, k2, 23 lines */ - 0x00000000, - 0x030d0133, /* igu, bb, 51 lines */ - 0x030d0133, /* igu, k2, 51 lines */ - 0x00000000, - 0x03400106, /* cau, bb, 6 lines */ - 0x03400106, /* cau, k2, 6 lines */ - 0x00000000, - 0x00000000, /* rgfs, bb, 0 lines */ - 0x00000000, /* rgfs, k2, 0 lines */ - 0x00000000, - 0x00000000, /* rgsrc, bb, 0 lines */ - 0x00000000, /* rgsrc, k2, 0 lines */ - 0x00000000, - 0x00000000, /* tgfs, bb, 0 lines */ - 0x00000000, /* tgfs, k2, 0 lines */ - 0x00000000, - 0x00000000, /* tgsrc, bb, 0 lines */ - 0x00000000, /* tgsrc, k2, 0 lines */ - 0x00000000, - 0x00000000, /* umac, bb, 0 lines */ - 0x00120006, /* umac, k2, 6 lines */ - 0x00000000, - 0x00000000, /* xmac, bb, 0 lines */ - 0x00000000, /* xmac, k2, 0 lines */ - 0x00000000, - 0x00000000, /* dbg, bb, 0 lines */ - 0x00000000, /* dbg, k2, 0 lines */ - 0x00000000, - 0x0346012b, /* nig, bb, 43 lines */ - 0x0346011d, /* nig, k2, 29 lines */ - 0x00000000, - 0x00000000, /* wol, bb, 0 lines */ - 0x001c0002, /* wol, k2, 2 lines */ - 0x00000000, - 0x00000000, /* bmbn, bb, 0 lines */ - 0x00210008, /* bmbn, k2, 8 lines */ - 0x00000000, - 0x00000000, /* ipc, bb, 0 lines */ - 0x00000000, /* ipc, k2, 0 lines */ - 0x00000000, - 0x00000000, /* nwm, bb, 0 lines */ - 0x0371000b, /* nwm, k2, 11 lines */ - 0x00000000, - 0x00000000, /* nws, bb, 0 lines */ - 0x037c0009, /* nws, k2, 9 lines */ - 0x00000000, - 0x00000000, /* ms, bb, 0 lines */ - 0x00120004, /* ms, k2, 4 lines */ - 0x00000000, - 0x00000000, /* phy_pcie, bb, 0 lines */ - 0x00e5001a, /* phy_pcie, k2, 26 lines */ - 0x00000000, - 0x00000000, /* led, bb, 0 lines */ - 0x00000000, /* led, k2, 0 lines */ - 0x00000000, - 0x00000000, /* avs_wrap, bb, 0 lines */ - 0x00000000, /* avs_wrap, k2, 0 lines */ - 0x00000000, - 0x00000000, /* bar0_map, bb, 0 lines */ - 0x00000000, /* bar0_map, k2, 0 lines */ - 0x00000000, - 0x00000000, /* bar0_map, bb, 0 lines */ - 0x00000000, /* bar0_map, k2, 0 lines */ - 0x00000000, -}; - /* Win 2 */ #define GTT_BAR0_MAP_REG_IGU_CMD 0x00f000UL @@ -3942,22 +3747,28 @@ static const u32 dbg_bus_blocks[] = { #define GTT_BAR0_MAP_REG_MSDM_RAM_1024 0x012000UL /* Win 6 */ -#define GTT_BAR0_MAP_REG_USDM_RAM 0x013000UL +#define GTT_BAR0_MAP_REG_MSDM_RAM_2048 0x013000UL /* Win 7 */ -#define GTT_BAR0_MAP_REG_USDM_RAM_1024 0x014000UL +#define GTT_BAR0_MAP_REG_USDM_RAM 0x014000UL /* Win 8 */ -#define GTT_BAR0_MAP_REG_USDM_RAM_2048 0x015000UL +#define GTT_BAR0_MAP_REG_USDM_RAM_1024 0x015000UL /* Win 9 */ -#define GTT_BAR0_MAP_REG_XSDM_RAM 0x016000UL +#define GTT_BAR0_MAP_REG_USDM_RAM_2048 0x016000UL /* Win 10 */ -#define GTT_BAR0_MAP_REG_YSDM_RAM 0x017000UL +#define GTT_BAR0_MAP_REG_XSDM_RAM 0x017000UL /* Win 11 */ -#define GTT_BAR0_MAP_REG_PSDM_RAM 0x018000UL +#define GTT_BAR0_MAP_REG_XSDM_RAM_1024 0x018000UL + +/* Win 12 */ +#define GTT_BAR0_MAP_REG_YSDM_RAM 0x019000UL + +/* Win 13 */ +#define GTT_BAR0_MAP_REG_PSDM_RAM 0x01a000UL /** * @brief qed_qm_pf_mem_size - prepare QM ILT sizes @@ -3982,7 +3793,7 @@ struct qed_qm_common_rt_init_params { u8 max_phys_tcs_per_port; bool pf_rl_en; bool pf_wfq_en; - bool vport_rl_en; + bool global_rl_en; bool vport_wfq_en; struct init_qm_port_params *port_params; }; @@ -4001,11 +3812,10 @@ struct qed_qm_pf_rt_init_params { u16 start_pq; u16 num_pf_pqs; u16 num_vf_pqs; - u8 start_vport; - u8 num_vports; + u16 start_vport; + u16 num_vports; u16 pf_wfq; u32 pf_rl; - u32 link_speed; struct init_qm_pq_params *pq_params; struct init_qm_vport_params *vport_params; }; @@ -4054,22 +3864,22 @@ int qed_init_pf_rl(struct qed_hwfn *p_hwfn, */ int qed_init_vport_wfq(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, - u16 first_tx_pq_id[NUM_OF_TCS], u16 vport_wfq); + u16 first_tx_pq_id[NUM_OF_TCS], u16 wfq); /** - * @brief qed_init_vport_rl - Initializes the rate limit of the specified VPORT + * @brief qed_init_global_rl - Initializes the rate limit of the specified + * rate limiter * * @param p_hwfn * @param p_ptt - ptt window used for writing the registers - * @param vport_id - VPORT ID - * @param vport_rl - rate limit in Mb/sec units - * @param link_speed - link speed in Mbps. + * @param rl_id - RL ID + * @param rate_limit - rate limit in Mb/sec units * * @return 0 on success, -1 on error. */ -int qed_init_vport_rl(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, - u8 vport_id, u32 vport_rl, u32 link_speed); +int qed_init_global_rl(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u16 rl_id, u32 rate_limit); /** * @brief qed_send_qm_stop_cmd Sends a stop command to the QM @@ -4157,7 +3967,7 @@ void qed_gft_disable(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u16 pf_id); /** * @brief qed_gft_config - Enable and configure HW for GFT * - * @param p_hwfn + * @param p_hwfn - HW device data * @param p_ptt - ptt window used for writing the registers. * @param pf_id - pf on which to enable GFT. * @param tcp - set profile tcp packets. @@ -4242,6 +4052,42 @@ void qed_memset_task_ctx(void *p_ctx_mem, u32 ctx_size, u8 ctx_type); void qed_set_rdma_error_level(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u8 assert_level[NUM_STORMS]); +/** + * @brief qed_fw_overlay_mem_alloc - Allocates and fills the FW overlay memory. + * + * @param p_hwfn - HW device data + * @param fw_overlay_in_buf - the input FW overlay buffer. + * @param buf_size - the size of the input FW overlay buffer in bytes. + * must be aligned to dwords. + * @param fw_overlay_out_mem - OUT: a pointer to the allocated overlays memory. + * + * @return a pointer to the allocated overlays memory, + * or NULL in case of failures. + */ +struct phys_mem_desc * +qed_fw_overlay_mem_alloc(struct qed_hwfn *p_hwfn, + const u32 * const fw_overlay_in_buf, + u32 buf_size_in_bytes); + +/** + * @brief qed_fw_overlay_init_ram - Initializes the FW overlay RAM. + * + * @param p_hwfn - HW device data. + * @param p_ptt - ptt window used for writing the registers. + * @param fw_overlay_mem - the allocated FW overlay memory. + */ +void qed_fw_overlay_init_ram(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + struct phys_mem_desc *fw_overlay_mem); + +/** + * @brief qed_fw_overlay_mem_free - Frees the FW overlay memory. + * + * @param p_hwfn - HW device data. + * @param fw_overlay_mem - the allocated FW overlay memory to free. + */ +void qed_fw_overlay_mem_free(struct qed_hwfn *p_hwfn, + struct phys_mem_desc *fw_overlay_mem); /* Ystorm flow control mode. Use enum fw_flow_ctrl_mode */ #define YSTORM_FLOW_CONTROL_MODE_OFFSET (IRO[0].base) @@ -4282,851 +4128,807 @@ void qed_set_rdma_error_level(struct qed_hwfn *p_hwfn, (IRO[7].base + ((queue_zone_id) * IRO[7].m1)) #define USTORM_COMMON_QUEUE_CONS_SIZE (IRO[7].size) +/* Xstorm common PQ info */ +#define XSTORM_PQ_INFO_OFFSET(pq_id) \ + (IRO[8].base + ((pq_id) * IRO[8].m1)) +#define XSTORM_PQ_INFO_SIZE (IRO[8].size) + /* Xstorm Integration Test Data */ -#define XSTORM_INTEG_TEST_DATA_OFFSET (IRO[8].base) -#define XSTORM_INTEG_TEST_DATA_SIZE (IRO[8].size) +#define XSTORM_INTEG_TEST_DATA_OFFSET (IRO[9].base) +#define XSTORM_INTEG_TEST_DATA_SIZE (IRO[9].size) /* Ystorm Integration Test Data */ -#define YSTORM_INTEG_TEST_DATA_OFFSET (IRO[9].base) -#define YSTORM_INTEG_TEST_DATA_SIZE (IRO[9].size) +#define YSTORM_INTEG_TEST_DATA_OFFSET (IRO[10].base) +#define YSTORM_INTEG_TEST_DATA_SIZE (IRO[10].size) /* Pstorm Integration Test Data */ -#define PSTORM_INTEG_TEST_DATA_OFFSET (IRO[10].base) -#define PSTORM_INTEG_TEST_DATA_SIZE (IRO[10].size) +#define PSTORM_INTEG_TEST_DATA_OFFSET (IRO[11].base) +#define PSTORM_INTEG_TEST_DATA_SIZE (IRO[11].size) /* Tstorm Integration Test Data */ -#define TSTORM_INTEG_TEST_DATA_OFFSET (IRO[11].base) -#define TSTORM_INTEG_TEST_DATA_SIZE (IRO[11].size) +#define TSTORM_INTEG_TEST_DATA_OFFSET (IRO[12].base) +#define TSTORM_INTEG_TEST_DATA_SIZE (IRO[12].size) /* Mstorm Integration Test Data */ -#define MSTORM_INTEG_TEST_DATA_OFFSET (IRO[12].base) -#define MSTORM_INTEG_TEST_DATA_SIZE (IRO[12].size) +#define MSTORM_INTEG_TEST_DATA_OFFSET (IRO[13].base) +#define MSTORM_INTEG_TEST_DATA_SIZE (IRO[13].size) /* Ustorm Integration Test Data */ -#define USTORM_INTEG_TEST_DATA_OFFSET (IRO[13].base) -#define USTORM_INTEG_TEST_DATA_SIZE (IRO[13].size) +#define USTORM_INTEG_TEST_DATA_OFFSET (IRO[14].base) +#define USTORM_INTEG_TEST_DATA_SIZE (IRO[14].size) + +/* Xstorm overlay buffer host address */ +#define XSTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[15].base) +#define XSTORM_OVERLAY_BUF_ADDR_SIZE (IRO[15].size) + +/* Ystorm overlay buffer host address */ +#define YSTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[16].base) +#define YSTORM_OVERLAY_BUF_ADDR_SIZE (IRO[16].size) + +/* Pstorm overlay buffer host address */ +#define PSTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[17].base) +#define PSTORM_OVERLAY_BUF_ADDR_SIZE (IRO[17].size) + +/* Tstorm overlay buffer host address */ +#define TSTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[18].base) +#define TSTORM_OVERLAY_BUF_ADDR_SIZE (IRO[18].size) + +/* Mstorm overlay buffer host address */ +#define MSTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[19].base) +#define MSTORM_OVERLAY_BUF_ADDR_SIZE (IRO[19].size) + +/* Ustorm overlay buffer host address */ +#define USTORM_OVERLAY_BUF_ADDR_OFFSET (IRO[20].base) +#define USTORM_OVERLAY_BUF_ADDR_SIZE (IRO[20].size) /* Tstorm producers */ #define TSTORM_LL2_RX_PRODS_OFFSET(core_rx_queue_id) \ - (IRO[14].base + ((core_rx_queue_id) * IRO[14].m1)) -#define TSTORM_LL2_RX_PRODS_SIZE (IRO[14].size) + (IRO[21].base + ((core_rx_queue_id) * IRO[21].m1)) +#define TSTORM_LL2_RX_PRODS_SIZE (IRO[21].size) /* Tstorm LightL2 queue statistics */ #define CORE_LL2_TSTORM_PER_QUEUE_STAT_OFFSET(core_rx_queue_id) \ - (IRO[15].base + ((core_rx_queue_id) * IRO[15].m1)) -#define CORE_LL2_TSTORM_PER_QUEUE_STAT_SIZE (IRO[15].size) + (IRO[22].base + ((core_rx_queue_id) * IRO[22].m1)) +#define CORE_LL2_TSTORM_PER_QUEUE_STAT_SIZE (IRO[22].size) /* Ustorm LiteL2 queue statistics */ #define CORE_LL2_USTORM_PER_QUEUE_STAT_OFFSET(core_rx_queue_id) \ - (IRO[16].base + ((core_rx_queue_id) * IRO[16].m1)) -#define CORE_LL2_USTORM_PER_QUEUE_STAT_SIZE (IRO[16].size) + (IRO[23].base + ((core_rx_queue_id) * IRO[23].m1)) +#define CORE_LL2_USTORM_PER_QUEUE_STAT_SIZE (IRO[23].size) /* Pstorm LiteL2 queue statistics */ #define CORE_LL2_PSTORM_PER_QUEUE_STAT_OFFSET(core_tx_stats_id) \ - (IRO[17].base + ((core_tx_stats_id) * IRO[17].m1)) -#define CORE_LL2_PSTORM_PER_QUEUE_STAT_SIZE (IRO[17].size) + (IRO[24].base + ((core_tx_stats_id) * IRO[24].m1)) +#define CORE_LL2_PSTORM_PER_QUEUE_STAT_SIZE (IRO[24].size) /* Mstorm queue statistics */ #define MSTORM_QUEUE_STAT_OFFSET(stat_counter_id) \ - (IRO[18].base + ((stat_counter_id) * IRO[18].m1)) -#define MSTORM_QUEUE_STAT_SIZE (IRO[18].size) + (IRO[25].base + ((stat_counter_id) * IRO[25].m1)) +#define MSTORM_QUEUE_STAT_SIZE (IRO[25].size) -/* Mstorm ETH PF queues producers */ -#define MSTORM_ETH_PF_PRODS_OFFSET(queue_id) \ - (IRO[19].base + ((queue_id) * IRO[19].m1)) -#define MSTORM_ETH_PF_PRODS_SIZE (IRO[19].size) +/* TPA agregation timeout in us resolution (on ASIC) */ +#define MSTORM_TPA_TIMEOUT_US_OFFSET (IRO[26].base) +#define MSTORM_TPA_TIMEOUT_US_SIZE (IRO[26].size) /* Mstorm ETH VF queues producers offset in RAM. Used in default VF zone size - * mode. + * mode */ #define MSTORM_ETH_VF_PRODS_OFFSET(vf_id, vf_queue_id) \ - (IRO[20].base + ((vf_id) * IRO[20].m1) + ((vf_queue_id) * IRO[20].m2)) -#define MSTORM_ETH_VF_PRODS_SIZE (IRO[20].size) + (IRO[27].base + ((vf_id) * IRO[27].m1) + ((vf_queue_id) * IRO[27].m2)) +#define MSTORM_ETH_VF_PRODS_SIZE (IRO[27].size) -/* TPA agregation timeout in us resolution (on ASIC) */ -#define MSTORM_TPA_TIMEOUT_US_OFFSET (IRO[21].base) -#define MSTORM_TPA_TIMEOUT_US_SIZE (IRO[21].size) +/* Mstorm ETH PF queues producers */ +#define MSTORM_ETH_PF_PRODS_OFFSET(queue_id) \ + (IRO[28].base + ((queue_id) * IRO[28].m1)) +#define MSTORM_ETH_PF_PRODS_SIZE (IRO[28].size) /* Mstorm pf statistics */ #define MSTORM_ETH_PF_STAT_OFFSET(pf_id) \ - (IRO[22].base + ((pf_id) * IRO[22].m1)) -#define MSTORM_ETH_PF_STAT_SIZE (IRO[22].size) + (IRO[29].base + ((pf_id) * IRO[29].m1)) +#define MSTORM_ETH_PF_STAT_SIZE (IRO[29].size) /* Ustorm queue statistics */ #define USTORM_QUEUE_STAT_OFFSET(stat_counter_id) \ - (IRO[23].base + ((stat_counter_id) * IRO[23].m1)) -#define USTORM_QUEUE_STAT_SIZE (IRO[23].size) + (IRO[30].base + ((stat_counter_id) * IRO[30].m1)) +#define USTORM_QUEUE_STAT_SIZE (IRO[30].size) /* Ustorm pf statistics */ -#define USTORM_ETH_PF_STAT_OFFSET(pf_id)\ - (IRO[24].base + ((pf_id) * IRO[24].m1)) -#define USTORM_ETH_PF_STAT_SIZE (IRO[24].size) +#define USTORM_ETH_PF_STAT_OFFSET(pf_id) \ + (IRO[31].base + ((pf_id) * IRO[31].m1)) +#define USTORM_ETH_PF_STAT_SIZE (IRO[31].size) /* Pstorm queue statistics */ -#define PSTORM_QUEUE_STAT_OFFSET(stat_counter_id) \ - (IRO[25].base + ((stat_counter_id) * IRO[25].m1)) -#define PSTORM_QUEUE_STAT_SIZE (IRO[25].size) +#define PSTORM_QUEUE_STAT_OFFSET(stat_counter_id) \ + (IRO[32].base + ((stat_counter_id) * IRO[32].m1)) +#define PSTORM_QUEUE_STAT_SIZE (IRO[32].size) /* Pstorm pf statistics */ #define PSTORM_ETH_PF_STAT_OFFSET(pf_id) \ - (IRO[26].base + ((pf_id) * IRO[26].m1)) -#define PSTORM_ETH_PF_STAT_SIZE (IRO[26].size) + (IRO[33].base + ((pf_id) * IRO[33].m1)) +#define PSTORM_ETH_PF_STAT_SIZE (IRO[33].size) /* Control frame's EthType configuration for TX control frame security */ -#define PSTORM_CTL_FRAME_ETHTYPE_OFFSET(eth_type_id) \ - (IRO[27].base + ((eth_type_id) * IRO[27].m1)) -#define PSTORM_CTL_FRAME_ETHTYPE_SIZE (IRO[27].size) +#define PSTORM_CTL_FRAME_ETHTYPE_OFFSET(eth_type_id) \ + (IRO[34].base + ((eth_type_id) * IRO[34].m1)) +#define PSTORM_CTL_FRAME_ETHTYPE_SIZE (IRO[34].size) /* Tstorm last parser message */ -#define TSTORM_ETH_PRS_INPUT_OFFSET (IRO[28].base) -#define TSTORM_ETH_PRS_INPUT_SIZE (IRO[28].size) +#define TSTORM_ETH_PRS_INPUT_OFFSET (IRO[35].base) +#define TSTORM_ETH_PRS_INPUT_SIZE (IRO[35].size) /* Tstorm Eth limit Rx rate */ -#define ETH_RX_RATE_LIMIT_OFFSET(pf_id) \ - (IRO[29].base + ((pf_id) * IRO[29].m1)) -#define ETH_RX_RATE_LIMIT_SIZE (IRO[29].size) +#define ETH_RX_RATE_LIMIT_OFFSET(pf_id) \ + (IRO[36].base + ((pf_id) * IRO[36].m1)) +#define ETH_RX_RATE_LIMIT_SIZE (IRO[36].size) /* RSS indirection table entry update command per PF offset in TSTORM PF BAR0. - * Use eth_tstorm_rss_update_data for update. + * Use eth_tstorm_rss_update_data for update */ #define TSTORM_ETH_RSS_UPDATE_OFFSET(pf_id) \ - (IRO[30].base + ((pf_id) * IRO[30].m1)) -#define TSTORM_ETH_RSS_UPDATE_SIZE (IRO[30].size) + (IRO[37].base + ((pf_id) * IRO[37].m1)) +#define TSTORM_ETH_RSS_UPDATE_SIZE (IRO[37].size) /* Xstorm queue zone */ #define XSTORM_ETH_QUEUE_ZONE_OFFSET(queue_id) \ - (IRO[31].base + ((queue_id) * IRO[31].m1)) -#define XSTORM_ETH_QUEUE_ZONE_SIZE (IRO[31].size) + (IRO[38].base + ((queue_id) * IRO[38].m1)) +#define XSTORM_ETH_QUEUE_ZONE_SIZE (IRO[38].size) /* Ystorm cqe producer */ #define YSTORM_TOE_CQ_PROD_OFFSET(rss_id) \ - (IRO[32].base + ((rss_id) * IRO[32].m1)) -#define YSTORM_TOE_CQ_PROD_SIZE (IRO[32].size) + (IRO[39].base + ((rss_id) * IRO[39].m1)) +#define YSTORM_TOE_CQ_PROD_SIZE (IRO[39].size) /* Ustorm cqe producer */ #define USTORM_TOE_CQ_PROD_OFFSET(rss_id) \ - (IRO[33].base + ((rss_id) * IRO[33].m1)) -#define USTORM_TOE_CQ_PROD_SIZE (IRO[33].size) + (IRO[40].base + ((rss_id) * IRO[40].m1)) +#define USTORM_TOE_CQ_PROD_SIZE (IRO[40].size) /* Ustorm grq producer */ #define USTORM_TOE_GRQ_PROD_OFFSET(pf_id) \ - (IRO[34].base + ((pf_id) * IRO[34].m1)) -#define USTORM_TOE_GRQ_PROD_SIZE (IRO[34].size) + (IRO[41].base + ((pf_id) * IRO[41].m1)) +#define USTORM_TOE_GRQ_PROD_SIZE (IRO[41].size) /* Tstorm cmdq-cons of given command queue-id */ #define TSTORM_SCSI_CMDQ_CONS_OFFSET(cmdq_queue_id) \ - (IRO[35].base + ((cmdq_queue_id) * IRO[35].m1)) -#define TSTORM_SCSI_CMDQ_CONS_SIZE (IRO[35].size) + (IRO[42].base + ((cmdq_queue_id) * IRO[42].m1)) +#define TSTORM_SCSI_CMDQ_CONS_SIZE (IRO[42].size) /* Tstorm (reflects M-Storm) bdq-external-producer of given function ID, - * BDqueue-id. + * BDqueue-id */ -#define TSTORM_SCSI_BDQ_EXT_PROD_OFFSET(func_id, bdq_id) \ - (IRO[36].base + ((func_id) * IRO[36].m1) + ((bdq_id) * IRO[36].m2)) -#define TSTORM_SCSI_BDQ_EXT_PROD_SIZE (IRO[36].size) +#define TSTORM_SCSI_BDQ_EXT_PROD_OFFSET(storage_func_id, bdq_id) \ + (IRO[43].base + ((storage_func_id) * IRO[43].m1) + \ + ((bdq_id) * IRO[43].m2)) +#define TSTORM_SCSI_BDQ_EXT_PROD_SIZE (IRO[43].size) /* Mstorm bdq-external-producer of given BDQ resource ID, BDqueue-id */ -#define MSTORM_SCSI_BDQ_EXT_PROD_OFFSET(func_id, bdq_id) \ - (IRO[37].base + ((func_id) * IRO[37].m1) + ((bdq_id) * IRO[37].m2)) -#define MSTORM_SCSI_BDQ_EXT_PROD_SIZE (IRO[37].size) +#define MSTORM_SCSI_BDQ_EXT_PROD_OFFSET(storage_func_id, bdq_id) \ + (IRO[44].base + ((storage_func_id) * IRO[44].m1) + \ + ((bdq_id) * IRO[44].m2)) +#define MSTORM_SCSI_BDQ_EXT_PROD_SIZE (IRO[44].size) /* Tstorm iSCSI RX stats */ -#define TSTORM_ISCSI_RX_STATS_OFFSET(pf_id) \ - (IRO[38].base + ((pf_id) * IRO[38].m1)) -#define TSTORM_ISCSI_RX_STATS_SIZE (IRO[38].size) +#define TSTORM_ISCSI_RX_STATS_OFFSET(storage_func_id) \ + (IRO[45].base + ((storage_func_id) * IRO[45].m1)) +#define TSTORM_ISCSI_RX_STATS_SIZE (IRO[45].size) /* Mstorm iSCSI RX stats */ -#define MSTORM_ISCSI_RX_STATS_OFFSET(pf_id) \ - (IRO[39].base + ((pf_id) * IRO[39].m1)) -#define MSTORM_ISCSI_RX_STATS_SIZE (IRO[39].size) +#define MSTORM_ISCSI_RX_STATS_OFFSET(storage_func_id) \ + (IRO[46].base + ((storage_func_id) * IRO[46].m1)) +#define MSTORM_ISCSI_RX_STATS_SIZE (IRO[46].size) /* Ustorm iSCSI RX stats */ -#define USTORM_ISCSI_RX_STATS_OFFSET(pf_id) \ - (IRO[40].base + ((pf_id) * IRO[40].m1)) -#define USTORM_ISCSI_RX_STATS_SIZE (IRO[40].size) +#define USTORM_ISCSI_RX_STATS_OFFSET(storage_func_id) \ + (IRO[47].base + ((storage_func_id) * IRO[47].m1)) +#define USTORM_ISCSI_RX_STATS_SIZE (IRO[47].size) /* Xstorm iSCSI TX stats */ -#define XSTORM_ISCSI_TX_STATS_OFFSET(pf_id) \ - (IRO[41].base + ((pf_id) * IRO[41].m1)) -#define XSTORM_ISCSI_TX_STATS_SIZE (IRO[41].size) +#define XSTORM_ISCSI_TX_STATS_OFFSET(storage_func_id) \ + (IRO[48].base + ((storage_func_id) * IRO[48].m1)) +#define XSTORM_ISCSI_TX_STATS_SIZE (IRO[48].size) /* Ystorm iSCSI TX stats */ -#define YSTORM_ISCSI_TX_STATS_OFFSET(pf_id) \ - (IRO[42].base + ((pf_id) * IRO[42].m1)) -#define YSTORM_ISCSI_TX_STATS_SIZE (IRO[42].size) +#define YSTORM_ISCSI_TX_STATS_OFFSET(storage_func_id) \ + (IRO[49].base + ((storage_func_id) * IRO[49].m1)) +#define YSTORM_ISCSI_TX_STATS_SIZE (IRO[49].size) /* Pstorm iSCSI TX stats */ -#define PSTORM_ISCSI_TX_STATS_OFFSET(pf_id) \ - (IRO[43].base + ((pf_id) * IRO[43].m1)) -#define PSTORM_ISCSI_TX_STATS_SIZE (IRO[43].size) +#define PSTORM_ISCSI_TX_STATS_OFFSET(storage_func_id) \ + (IRO[50].base + ((storage_func_id) * IRO[50].m1)) +#define PSTORM_ISCSI_TX_STATS_SIZE (IRO[50].size) /* Tstorm FCoE RX stats */ #define TSTORM_FCOE_RX_STATS_OFFSET(pf_id) \ - (IRO[44].base + ((pf_id) * IRO[44].m1)) -#define TSTORM_FCOE_RX_STATS_SIZE (IRO[44].size) + (IRO[51].base + ((pf_id) * IRO[51].m1)) +#define TSTORM_FCOE_RX_STATS_SIZE (IRO[51].size) /* Pstorm FCoE TX stats */ #define PSTORM_FCOE_TX_STATS_OFFSET(pf_id) \ - (IRO[45].base + ((pf_id) * IRO[45].m1)) -#define PSTORM_FCOE_TX_STATS_SIZE (IRO[45].size) + (IRO[52].base + ((pf_id) * IRO[52].m1)) +#define PSTORM_FCOE_TX_STATS_SIZE (IRO[52].size) /* Pstorm RDMA queue statistics */ #define PSTORM_RDMA_QUEUE_STAT_OFFSET(rdma_stat_counter_id) \ - (IRO[46].base + ((rdma_stat_counter_id) * IRO[46].m1)) -#define PSTORM_RDMA_QUEUE_STAT_SIZE (IRO[46].size) + (IRO[53].base + ((rdma_stat_counter_id) * IRO[53].m1)) +#define PSTORM_RDMA_QUEUE_STAT_SIZE (IRO[53].size) /* Tstorm RDMA queue statistics */ #define TSTORM_RDMA_QUEUE_STAT_OFFSET(rdma_stat_counter_id) \ - (IRO[47].base + ((rdma_stat_counter_id) * IRO[47].m1)) -#define TSTORM_RDMA_QUEUE_STAT_SIZE (IRO[47].size) + (IRO[54].base + ((rdma_stat_counter_id) * IRO[54].m1)) +#define TSTORM_RDMA_QUEUE_STAT_SIZE (IRO[54].size) /* Xstorm error level for assert */ #define XSTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[48].base + ((pf_id) * IRO[48].m1)) -#define XSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[48].size) + (IRO[55].base + ((pf_id) * IRO[55].m1)) +#define XSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[55].size) /* Ystorm error level for assert */ #define YSTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[49].base + ((pf_id) * IRO[49].m1)) -#define YSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[49].size) + (IRO[56].base + ((pf_id) * IRO[56].m1)) +#define YSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[56].size) /* Pstorm error level for assert */ #define PSTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[50].base + ((pf_id) * IRO[50].m1)) -#define PSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[50].size) + (IRO[57].base + ((pf_id) * IRO[57].m1)) +#define PSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[57].size) /* Tstorm error level for assert */ #define TSTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[51].base + ((pf_id) * IRO[51].m1)) -#define TSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[51].size) + (IRO[58].base + ((pf_id) * IRO[58].m1)) +#define TSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[58].size) /* Mstorm error level for assert */ #define MSTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[52].base + ((pf_id) * IRO[52].m1)) -#define MSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[52].size) + (IRO[59].base + ((pf_id) * IRO[59].m1)) +#define MSTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[59].size) /* Ustorm error level for assert */ #define USTORM_RDMA_ASSERT_LEVEL_OFFSET(pf_id) \ - (IRO[53].base + ((pf_id) * IRO[53].m1)) -#define USTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[53].size) + (IRO[60].base + ((pf_id) * IRO[60].m1)) +#define USTORM_RDMA_ASSERT_LEVEL_SIZE (IRO[60].size) /* Xstorm iWARP rxmit stats */ #define XSTORM_IWARP_RXMIT_STATS_OFFSET(pf_id) \ - (IRO[54].base + ((pf_id) * IRO[54].m1)) -#define XSTORM_IWARP_RXMIT_STATS_SIZE (IRO[54].size) + (IRO[61].base + ((pf_id) * IRO[61].m1)) +#define XSTORM_IWARP_RXMIT_STATS_SIZE (IRO[61].size) /* Tstorm RoCE Event Statistics */ -#define TSTORM_ROCE_EVENTS_STAT_OFFSET(roce_pf_id) \ - (IRO[55].base + ((roce_pf_id) * IRO[55].m1)) -#define TSTORM_ROCE_EVENTS_STAT_SIZE (IRO[55].size) +#define TSTORM_ROCE_EVENTS_STAT_OFFSET(roce_pf_id) \ + (IRO[62].base + ((roce_pf_id) * IRO[62].m1)) +#define TSTORM_ROCE_EVENTS_STAT_SIZE (IRO[62].size) /* DCQCN Received Statistics */ -#define YSTORM_ROCE_DCQCN_RECEIVED_STATS_OFFSET(roce_pf_id) \ - (IRO[56].base + ((roce_pf_id) * IRO[56].m1)) -#define YSTORM_ROCE_DCQCN_RECEIVED_STATS_SIZE (IRO[56].size) +#define YSTORM_ROCE_DCQCN_RECEIVED_STATS_OFFSET(roce_pf_id)\ + (IRO[63].base + ((roce_pf_id) * IRO[63].m1)) +#define YSTORM_ROCE_DCQCN_RECEIVED_STATS_SIZE (IRO[63].size) /* RoCE Error Statistics */ -#define YSTORM_ROCE_ERROR_STATS_OFFSET(roce_pf_id) \ - (IRO[57].base + ((roce_pf_id) * IRO[57].m1)) -#define YSTORM_ROCE_ERROR_STATS_SIZE (IRO[57].size) +#define YSTORM_ROCE_ERROR_STATS_OFFSET(roce_pf_id) \ + (IRO[64].base + ((roce_pf_id) * IRO[64].m1)) +#define YSTORM_ROCE_ERROR_STATS_SIZE (IRO[64].size) /* DCQCN Sent Statistics */ -#define PSTORM_ROCE_DCQCN_SENT_STATS_OFFSET(roce_pf_id) \ - (IRO[58].base + ((roce_pf_id) * IRO[58].m1)) -#define PSTORM_ROCE_DCQCN_SENT_STATS_SIZE (IRO[58].size) +#define PSTORM_ROCE_DCQCN_SENT_STATS_OFFSET(roce_pf_id) \ + (IRO[65].base + ((roce_pf_id) * IRO[65].m1)) +#define PSTORM_ROCE_DCQCN_SENT_STATS_SIZE (IRO[65].size) /* RoCE CQEs Statistics */ -#define USTORM_ROCE_CQE_STATS_OFFSET(roce_pf_id) \ - (IRO[59].base + ((roce_pf_id) * IRO[59].m1)) -#define USTORM_ROCE_CQE_STATS_SIZE (IRO[59].size) - -static const struct iro iro_arr[60] = { - {0x0, 0x0, 0x0, 0x0, 0x8}, - {0x4cb8, 0x88, 0x0, 0x0, 0x88}, - {0x6530, 0x20, 0x0, 0x0, 0x20}, - {0xb00, 0x8, 0x0, 0x0, 0x4}, - {0xa80, 0x8, 0x0, 0x0, 0x4}, - {0x0, 0x8, 0x0, 0x0, 0x2}, - {0x80, 0x8, 0x0, 0x0, 0x4}, - {0x84, 0x8, 0x0, 0x0, 0x2}, - {0x4c48, 0x0, 0x0, 0x0, 0x78}, - {0x3e38, 0x0, 0x0, 0x0, 0x78}, - {0x3ef8, 0x0, 0x0, 0x0, 0x78}, - {0x4c40, 0x0, 0x0, 0x0, 0x78}, - {0x4998, 0x0, 0x0, 0x0, 0x78}, - {0x7f50, 0x0, 0x0, 0x0, 0x78}, - {0xa28, 0x8, 0x0, 0x0, 0x8}, - {0x6210, 0x10, 0x0, 0x0, 0x10}, - {0xb820, 0x30, 0x0, 0x0, 0x30}, - {0xa990, 0x30, 0x0, 0x0, 0x30}, - {0x4b68, 0x80, 0x0, 0x0, 0x40}, - {0x1f8, 0x4, 0x0, 0x0, 0x4}, - {0x53a8, 0x80, 0x4, 0x0, 0x4}, - {0xc7d0, 0x0, 0x0, 0x0, 0x4}, - {0x4ba8, 0x80, 0x0, 0x0, 0x20}, - {0x8158, 0x40, 0x0, 0x0, 0x30}, - {0xe770, 0x60, 0x0, 0x0, 0x60}, - {0x4090, 0x80, 0x0, 0x0, 0x38}, - {0xfea8, 0x78, 0x0, 0x0, 0x78}, - {0x1f8, 0x4, 0x0, 0x0, 0x4}, - {0xaf20, 0x0, 0x0, 0x0, 0xf0}, - {0xb010, 0x8, 0x0, 0x0, 0x8}, - {0xc00, 0x8, 0x0, 0x0, 0x8}, - {0x1f8, 0x8, 0x0, 0x0, 0x8}, - {0xac0, 0x8, 0x0, 0x0, 0x8}, - {0x2578, 0x8, 0x0, 0x0, 0x8}, - {0x24f8, 0x8, 0x0, 0x0, 0x8}, - {0x0, 0x8, 0x0, 0x0, 0x8}, - {0x400, 0x18, 0x8, 0x0, 0x8}, - {0xb78, 0x18, 0x8, 0x0, 0x2}, - {0xd898, 0x50, 0x0, 0x0, 0x3c}, - {0x12908, 0x18, 0x0, 0x0, 0x10}, - {0x11aa8, 0x40, 0x0, 0x0, 0x18}, - {0xa588, 0x50, 0x0, 0x0, 0x20}, - {0x8f00, 0x40, 0x0, 0x0, 0x28}, - {0x10e30, 0x18, 0x0, 0x0, 0x10}, - {0xde48, 0x48, 0x0, 0x0, 0x38}, - {0x11298, 0x20, 0x0, 0x0, 0x20}, - {0x40c8, 0x80, 0x0, 0x0, 0x10}, - {0x5048, 0x10, 0x0, 0x0, 0x10}, - {0xc748, 0x8, 0x0, 0x0, 0x1}, - {0xa928, 0x8, 0x0, 0x0, 0x1}, - {0x11a30, 0x8, 0x0, 0x0, 0x1}, - {0xf030, 0x8, 0x0, 0x0, 0x1}, - {0x13028, 0x8, 0x0, 0x0, 0x1}, - {0x12c58, 0x8, 0x0, 0x0, 0x1}, - {0xc9b8, 0x30, 0x0, 0x0, 0x10}, - {0xed90, 0x28, 0x0, 0x0, 0x28}, - {0xad20, 0x18, 0x0, 0x0, 0x18}, - {0xaea0, 0x8, 0x0, 0x0, 0x8}, - {0x13c38, 0x8, 0x0, 0x0, 0x8}, - {0x13c50, 0x18, 0x0, 0x0, 0x18}, +#define USTORM_ROCE_CQE_STATS_OFFSET(roce_pf_id) \ + (IRO[66].base + ((roce_pf_id) * IRO[66].m1)) +#define USTORM_ROCE_CQE_STATS_SIZE (IRO[66].size) + +/* IRO Array */ +static const u32 iro_arr[] = { + 0x00000000, 0x00000000, 0x00080000, + 0x00003288, 0x00000088, 0x00880000, + 0x000058e8, 0x00000020, 0x00200000, + 0x00000b00, 0x00000008, 0x00040000, + 0x00000a80, 0x00000008, 0x00040000, + 0x00000000, 0x00000008, 0x00020000, + 0x00000080, 0x00000008, 0x00040000, + 0x00000084, 0x00000008, 0x00020000, + 0x00005718, 0x00000004, 0x00040000, + 0x00004dd0, 0x00000000, 0x00780000, + 0x00003e40, 0x00000000, 0x00780000, + 0x00004480, 0x00000000, 0x00780000, + 0x00003210, 0x00000000, 0x00780000, + 0x00003b50, 0x00000000, 0x00780000, + 0x00007f58, 0x00000000, 0x00780000, + 0x00005f58, 0x00000000, 0x00080000, + 0x00007100, 0x00000000, 0x00080000, + 0x0000aea0, 0x00000000, 0x00080000, + 0x00004398, 0x00000000, 0x00080000, + 0x0000a5a0, 0x00000000, 0x00080000, + 0x0000bde8, 0x00000000, 0x00080000, + 0x00000020, 0x00000004, 0x00040000, + 0x000056c8, 0x00000010, 0x00100000, + 0x0000c210, 0x00000030, 0x00300000, + 0x0000b088, 0x00000038, 0x00380000, + 0x00003d20, 0x00000080, 0x00400000, + 0x0000bf60, 0x00000000, 0x00040000, + 0x00004560, 0x00040080, 0x00040000, + 0x000001f8, 0x00000004, 0x00040000, + 0x00003d60, 0x00000080, 0x00200000, + 0x00008960, 0x00000040, 0x00300000, + 0x0000e840, 0x00000060, 0x00600000, + 0x00004618, 0x00000080, 0x00380000, + 0x00010738, 0x000000c0, 0x00c00000, + 0x000001f8, 0x00000002, 0x00020000, + 0x0000a2a0, 0x00000000, 0x01080000, + 0x0000a3a8, 0x00000008, 0x00080000, + 0x000001c0, 0x00000008, 0x00080000, + 0x000001f8, 0x00000008, 0x00080000, + 0x00000ac0, 0x00000008, 0x00080000, + 0x00002578, 0x00000008, 0x00080000, + 0x000024f8, 0x00000008, 0x00080000, + 0x00000280, 0x00000008, 0x00080000, + 0x00000680, 0x00080018, 0x00080000, + 0x00000b78, 0x00080018, 0x00020000, + 0x0000c640, 0x00000050, 0x003c0000, + 0x00012038, 0x00000018, 0x00100000, + 0x00011b00, 0x00000040, 0x00180000, + 0x000095d0, 0x00000050, 0x00200000, + 0x00008b10, 0x00000040, 0x00280000, + 0x00011640, 0x00000018, 0x00100000, + 0x0000c828, 0x00000048, 0x00380000, + 0x00011710, 0x00000020, 0x00200000, + 0x00004650, 0x00000080, 0x00100000, + 0x00003618, 0x00000010, 0x00100000, + 0x0000a968, 0x00000008, 0x00010000, + 0x000097a0, 0x00000008, 0x00010000, + 0x00011990, 0x00000008, 0x00010000, + 0x0000f018, 0x00000008, 0x00010000, + 0x00012628, 0x00000008, 0x00010000, + 0x00011da8, 0x00000008, 0x00010000, + 0x0000aa78, 0x00000030, 0x00100000, + 0x0000d768, 0x00000028, 0x00280000, + 0x00009a58, 0x00000018, 0x00180000, + 0x00009bd8, 0x00000008, 0x00080000, + 0x00013a18, 0x00000008, 0x00080000, + 0x000126e8, 0x00000018, 0x00180000, + 0x0000e608, 0x00500288, 0x00100000, + 0x00012970, 0x00000138, 0x00280000, }; /* Runtime array offsets */ -#define DORQ_REG_PF_MAX_ICID_0_RT_OFFSET 0 -#define DORQ_REG_PF_MAX_ICID_1_RT_OFFSET 1 -#define DORQ_REG_PF_MAX_ICID_2_RT_OFFSET 2 -#define DORQ_REG_PF_MAX_ICID_3_RT_OFFSET 3 -#define DORQ_REG_PF_MAX_ICID_4_RT_OFFSET 4 -#define DORQ_REG_PF_MAX_ICID_5_RT_OFFSET 5 -#define DORQ_REG_PF_MAX_ICID_6_RT_OFFSET 6 -#define DORQ_REG_PF_MAX_ICID_7_RT_OFFSET 7 -#define DORQ_REG_VF_MAX_ICID_0_RT_OFFSET 8 -#define DORQ_REG_VF_MAX_ICID_1_RT_OFFSET 9 -#define DORQ_REG_VF_MAX_ICID_2_RT_OFFSET 10 -#define DORQ_REG_VF_MAX_ICID_3_RT_OFFSET 11 -#define DORQ_REG_VF_MAX_ICID_4_RT_OFFSET 12 -#define DORQ_REG_VF_MAX_ICID_5_RT_OFFSET 13 -#define DORQ_REG_VF_MAX_ICID_6_RT_OFFSET 14 -#define DORQ_REG_VF_MAX_ICID_7_RT_OFFSET 15 -#define DORQ_REG_PF_WAKE_ALL_RT_OFFSET 16 -#define DORQ_REG_TAG1_ETHERTYPE_RT_OFFSET 17 -#define DORQ_REG_GLB_MAX_ICID_0_RT_OFFSET 18 -#define DORQ_REG_GLB_MAX_ICID_1_RT_OFFSET 19 -#define DORQ_REG_GLB_RANGE2CONN_TYPE_0_RT_OFFSET 20 -#define DORQ_REG_GLB_RANGE2CONN_TYPE_1_RT_OFFSET 21 -#define DORQ_REG_PRV_PF_MAX_ICID_2_RT_OFFSET 22 -#define DORQ_REG_PRV_PF_MAX_ICID_3_RT_OFFSET 23 -#define DORQ_REG_PRV_PF_MAX_ICID_4_RT_OFFSET 24 -#define DORQ_REG_PRV_PF_MAX_ICID_5_RT_OFFSET 25 -#define DORQ_REG_PRV_VF_MAX_ICID_2_RT_OFFSET 26 -#define DORQ_REG_PRV_VF_MAX_ICID_3_RT_OFFSET 27 -#define DORQ_REG_PRV_VF_MAX_ICID_4_RT_OFFSET 28 -#define DORQ_REG_PRV_VF_MAX_ICID_5_RT_OFFSET 29 -#define DORQ_REG_PRV_PF_RANGE2CONN_TYPE_2_RT_OFFSET 30 -#define DORQ_REG_PRV_PF_RANGE2CONN_TYPE_3_RT_OFFSET 31 -#define DORQ_REG_PRV_PF_RANGE2CONN_TYPE_4_RT_OFFSET 32 -#define DORQ_REG_PRV_PF_RANGE2CONN_TYPE_5_RT_OFFSET 33 -#define DORQ_REG_PRV_VF_RANGE2CONN_TYPE_2_RT_OFFSET 34 -#define DORQ_REG_PRV_VF_RANGE2CONN_TYPE_3_RT_OFFSET 35 -#define DORQ_REG_PRV_VF_RANGE2CONN_TYPE_4_RT_OFFSET 36 -#define DORQ_REG_PRV_VF_RANGE2CONN_TYPE_5_RT_OFFSET 37 -#define IGU_REG_PF_CONFIGURATION_RT_OFFSET 38 -#define IGU_REG_VF_CONFIGURATION_RT_OFFSET 39 -#define IGU_REG_ATTN_MSG_ADDR_L_RT_OFFSET 40 -#define IGU_REG_ATTN_MSG_ADDR_H_RT_OFFSET 41 -#define IGU_REG_LEADING_EDGE_LATCH_RT_OFFSET 42 -#define IGU_REG_TRAILING_EDGE_LATCH_RT_OFFSET 43 -#define CAU_REG_CQE_AGG_UNIT_SIZE_RT_OFFSET 44 -#define CAU_REG_SB_VAR_MEMORY_RT_OFFSET 45 -#define CAU_REG_SB_VAR_MEMORY_RT_SIZE 1024 -#define CAU_REG_SB_ADDR_MEMORY_RT_OFFSET 1069 -#define CAU_REG_SB_ADDR_MEMORY_RT_SIZE 1024 -#define CAU_REG_PI_MEMORY_RT_OFFSET 2093 -#define CAU_REG_PI_MEMORY_RT_SIZE 4416 -#define PRS_REG_SEARCH_RESP_INITIATOR_TYPE_RT_OFFSET 6509 -#define PRS_REG_TASK_ID_MAX_INITIATOR_PF_RT_OFFSET 6510 -#define PRS_REG_TASK_ID_MAX_INITIATOR_VF_RT_OFFSET 6511 -#define PRS_REG_TASK_ID_MAX_TARGET_PF_RT_OFFSET 6512 -#define PRS_REG_TASK_ID_MAX_TARGET_VF_RT_OFFSET 6513 -#define PRS_REG_SEARCH_TCP_RT_OFFSET 6514 -#define PRS_REG_SEARCH_FCOE_RT_OFFSET 6515 -#define PRS_REG_SEARCH_ROCE_RT_OFFSET 6516 -#define PRS_REG_ROCE_DEST_QP_MAX_VF_RT_OFFSET 6517 -#define PRS_REG_ROCE_DEST_QP_MAX_PF_RT_OFFSET 6518 -#define PRS_REG_SEARCH_OPENFLOW_RT_OFFSET 6519 -#define PRS_REG_SEARCH_NON_IP_AS_OPENFLOW_RT_OFFSET 6520 -#define PRS_REG_OPENFLOW_SUPPORT_ONLY_KNOWN_OVER_IP_RT_OFFSET 6521 -#define PRS_REG_OPENFLOW_SEARCH_KEY_MASK_RT_OFFSET 6522 -#define PRS_REG_TAG_ETHERTYPE_0_RT_OFFSET 6523 -#define PRS_REG_LIGHT_L2_ETHERTYPE_EN_RT_OFFSET 6524 -#define SRC_REG_FIRSTFREE_RT_OFFSET 6525 -#define SRC_REG_FIRSTFREE_RT_SIZE 2 -#define SRC_REG_LASTFREE_RT_OFFSET 6527 -#define SRC_REG_LASTFREE_RT_SIZE 2 -#define SRC_REG_COUNTFREE_RT_OFFSET 6529 -#define SRC_REG_NUMBER_HASH_BITS_RT_OFFSET 6530 -#define PSWRQ2_REG_CDUT_P_SIZE_RT_OFFSET 6531 -#define PSWRQ2_REG_CDUC_P_SIZE_RT_OFFSET 6532 -#define PSWRQ2_REG_TM_P_SIZE_RT_OFFSET 6533 -#define PSWRQ2_REG_QM_P_SIZE_RT_OFFSET 6534 -#define PSWRQ2_REG_SRC_P_SIZE_RT_OFFSET 6535 -#define PSWRQ2_REG_TSDM_P_SIZE_RT_OFFSET 6536 -#define PSWRQ2_REG_TM_FIRST_ILT_RT_OFFSET 6537 -#define PSWRQ2_REG_TM_LAST_ILT_RT_OFFSET 6538 -#define PSWRQ2_REG_QM_FIRST_ILT_RT_OFFSET 6539 -#define PSWRQ2_REG_QM_LAST_ILT_RT_OFFSET 6540 -#define PSWRQ2_REG_SRC_FIRST_ILT_RT_OFFSET 6541 -#define PSWRQ2_REG_SRC_LAST_ILT_RT_OFFSET 6542 -#define PSWRQ2_REG_CDUC_FIRST_ILT_RT_OFFSET 6543 -#define PSWRQ2_REG_CDUC_LAST_ILT_RT_OFFSET 6544 -#define PSWRQ2_REG_CDUT_FIRST_ILT_RT_OFFSET 6545 -#define PSWRQ2_REG_CDUT_LAST_ILT_RT_OFFSET 6546 -#define PSWRQ2_REG_TSDM_FIRST_ILT_RT_OFFSET 6547 -#define PSWRQ2_REG_TSDM_LAST_ILT_RT_OFFSET 6548 -#define PSWRQ2_REG_TM_NUMBER_OF_PF_BLOCKS_RT_OFFSET 6549 -#define PSWRQ2_REG_CDUT_NUMBER_OF_PF_BLOCKS_RT_OFFSET 6550 -#define PSWRQ2_REG_CDUC_NUMBER_OF_PF_BLOCKS_RT_OFFSET 6551 -#define PSWRQ2_REG_TM_VF_BLOCKS_RT_OFFSET 6552 -#define PSWRQ2_REG_CDUT_VF_BLOCKS_RT_OFFSET 6553 -#define PSWRQ2_REG_CDUC_VF_BLOCKS_RT_OFFSET 6554 -#define PSWRQ2_REG_TM_BLOCKS_FACTOR_RT_OFFSET 6555 -#define PSWRQ2_REG_CDUT_BLOCKS_FACTOR_RT_OFFSET 6556 -#define PSWRQ2_REG_CDUC_BLOCKS_FACTOR_RT_OFFSET 6557 -#define PSWRQ2_REG_VF_BASE_RT_OFFSET 6558 -#define PSWRQ2_REG_VF_LAST_ILT_RT_OFFSET 6559 -#define PSWRQ2_REG_DRAM_ALIGN_WR_RT_OFFSET 6560 -#define PSWRQ2_REG_DRAM_ALIGN_RD_RT_OFFSET 6561 -#define PSWRQ2_REG_TGSRC_FIRST_ILT_RT_OFFSET 6562 -#define PSWRQ2_REG_RGSRC_FIRST_ILT_RT_OFFSET 6563 -#define PSWRQ2_REG_TGSRC_LAST_ILT_RT_OFFSET 6564 -#define PSWRQ2_REG_RGSRC_LAST_ILT_RT_OFFSET 6565 -#define PSWRQ2_REG_ILT_MEMORY_RT_OFFSET 6566 -#define PSWRQ2_REG_ILT_MEMORY_RT_SIZE 26414 -#define PGLUE_REG_B_VF_BASE_RT_OFFSET 32980 -#define PGLUE_REG_B_MSDM_OFFSET_MASK_B_RT_OFFSET 32981 -#define PGLUE_REG_B_MSDM_VF_SHIFT_B_RT_OFFSET 32982 -#define PGLUE_REG_B_CACHE_LINE_SIZE_RT_OFFSET 32983 -#define PGLUE_REG_B_PF_BAR0_SIZE_RT_OFFSET 32984 -#define PGLUE_REG_B_PF_BAR1_SIZE_RT_OFFSET 32985 -#define PGLUE_REG_B_VF_BAR1_SIZE_RT_OFFSET 32986 -#define TM_REG_VF_ENABLE_CONN_RT_OFFSET 32987 -#define TM_REG_PF_ENABLE_CONN_RT_OFFSET 32988 -#define TM_REG_PF_ENABLE_TASK_RT_OFFSET 32989 -#define TM_REG_GROUP_SIZE_RESOLUTION_CONN_RT_OFFSET 32990 -#define TM_REG_GROUP_SIZE_RESOLUTION_TASK_RT_OFFSET 32991 -#define TM_REG_CONFIG_CONN_MEM_RT_OFFSET 32992 -#define TM_REG_CONFIG_CONN_MEM_RT_SIZE 416 -#define TM_REG_CONFIG_TASK_MEM_RT_OFFSET 33408 -#define TM_REG_CONFIG_TASK_MEM_RT_SIZE 608 -#define QM_REG_MAXPQSIZE_0_RT_OFFSET 34016 -#define QM_REG_MAXPQSIZE_1_RT_OFFSET 34017 -#define QM_REG_MAXPQSIZE_2_RT_OFFSET 34018 -#define QM_REG_MAXPQSIZETXSEL_0_RT_OFFSET 34019 -#define QM_REG_MAXPQSIZETXSEL_1_RT_OFFSET 34020 -#define QM_REG_MAXPQSIZETXSEL_2_RT_OFFSET 34021 -#define QM_REG_MAXPQSIZETXSEL_3_RT_OFFSET 34022 -#define QM_REG_MAXPQSIZETXSEL_4_RT_OFFSET 34023 -#define QM_REG_MAXPQSIZETXSEL_5_RT_OFFSET 34024 -#define QM_REG_MAXPQSIZETXSEL_6_RT_OFFSET 34025 -#define QM_REG_MAXPQSIZETXSEL_7_RT_OFFSET 34026 -#define QM_REG_MAXPQSIZETXSEL_8_RT_OFFSET 34027 -#define QM_REG_MAXPQSIZETXSEL_9_RT_OFFSET 34028 -#define QM_REG_MAXPQSIZETXSEL_10_RT_OFFSET 34029 -#define QM_REG_MAXPQSIZETXSEL_11_RT_OFFSET 34030 -#define QM_REG_MAXPQSIZETXSEL_12_RT_OFFSET 34031 -#define QM_REG_MAXPQSIZETXSEL_13_RT_OFFSET 34032 -#define QM_REG_MAXPQSIZETXSEL_14_RT_OFFSET 34033 -#define QM_REG_MAXPQSIZETXSEL_15_RT_OFFSET 34034 -#define QM_REG_MAXPQSIZETXSEL_16_RT_OFFSET 34035 -#define QM_REG_MAXPQSIZETXSEL_17_RT_OFFSET 34036 -#define QM_REG_MAXPQSIZETXSEL_18_RT_OFFSET 34037 -#define QM_REG_MAXPQSIZETXSEL_19_RT_OFFSET 34038 -#define QM_REG_MAXPQSIZETXSEL_20_RT_OFFSET 34039 -#define QM_REG_MAXPQSIZETXSEL_21_RT_OFFSET 34040 -#define QM_REG_MAXPQSIZETXSEL_22_RT_OFFSET 34041 -#define QM_REG_MAXPQSIZETXSEL_23_RT_OFFSET 34042 -#define QM_REG_MAXPQSIZETXSEL_24_RT_OFFSET 34043 -#define QM_REG_MAXPQSIZETXSEL_25_RT_OFFSET 34044 -#define QM_REG_MAXPQSIZETXSEL_26_RT_OFFSET 34045 -#define QM_REG_MAXPQSIZETXSEL_27_RT_OFFSET 34046 -#define QM_REG_MAXPQSIZETXSEL_28_RT_OFFSET 34047 -#define QM_REG_MAXPQSIZETXSEL_29_RT_OFFSET 34048 -#define QM_REG_MAXPQSIZETXSEL_30_RT_OFFSET 34049 -#define QM_REG_MAXPQSIZETXSEL_31_RT_OFFSET 34050 -#define QM_REG_MAXPQSIZETXSEL_32_RT_OFFSET 34051 -#define QM_REG_MAXPQSIZETXSEL_33_RT_OFFSET 34052 -#define QM_REG_MAXPQSIZETXSEL_34_RT_OFFSET 34053 -#define QM_REG_MAXPQSIZETXSEL_35_RT_OFFSET 34054 -#define QM_REG_MAXPQSIZETXSEL_36_RT_OFFSET 34055 -#define QM_REG_MAXPQSIZETXSEL_37_RT_OFFSET 34056 -#define QM_REG_MAXPQSIZETXSEL_38_RT_OFFSET 34057 -#define QM_REG_MAXPQSIZETXSEL_39_RT_OFFSET 34058 -#define QM_REG_MAXPQSIZETXSEL_40_RT_OFFSET 34059 -#define QM_REG_MAXPQSIZETXSEL_41_RT_OFFSET 34060 -#define QM_REG_MAXPQSIZETXSEL_42_RT_OFFSET 34061 -#define QM_REG_MAXPQSIZETXSEL_43_RT_OFFSET 34062 -#define QM_REG_MAXPQSIZETXSEL_44_RT_OFFSET 34063 -#define QM_REG_MAXPQSIZETXSEL_45_RT_OFFSET 34064 -#define QM_REG_MAXPQSIZETXSEL_46_RT_OFFSET 34065 -#define QM_REG_MAXPQSIZETXSEL_47_RT_OFFSET 34066 -#define QM_REG_MAXPQSIZETXSEL_48_RT_OFFSET 34067 -#define QM_REG_MAXPQSIZETXSEL_49_RT_OFFSET 34068 -#define QM_REG_MAXPQSIZETXSEL_50_RT_OFFSET 34069 -#define QM_REG_MAXPQSIZETXSEL_51_RT_OFFSET 34070 -#define QM_REG_MAXPQSIZETXSEL_52_RT_OFFSET 34071 -#define QM_REG_MAXPQSIZETXSEL_53_RT_OFFSET 34072 -#define QM_REG_MAXPQSIZETXSEL_54_RT_OFFSET 34073 -#define QM_REG_MAXPQSIZETXSEL_55_RT_OFFSET 34074 -#define QM_REG_MAXPQSIZETXSEL_56_RT_OFFSET 34075 -#define QM_REG_MAXPQSIZETXSEL_57_RT_OFFSET 34076 -#define QM_REG_MAXPQSIZETXSEL_58_RT_OFFSET 34077 -#define QM_REG_MAXPQSIZETXSEL_59_RT_OFFSET 34078 -#define QM_REG_MAXPQSIZETXSEL_60_RT_OFFSET 34079 -#define QM_REG_MAXPQSIZETXSEL_61_RT_OFFSET 34080 -#define QM_REG_MAXPQSIZETXSEL_62_RT_OFFSET 34081 -#define QM_REG_MAXPQSIZETXSEL_63_RT_OFFSET 34082 -#define QM_REG_BASEADDROTHERPQ_RT_OFFSET 34083 -#define QM_REG_BASEADDROTHERPQ_RT_SIZE 128 -#define QM_REG_PTRTBLOTHER_RT_OFFSET 34211 -#define QM_REG_PTRTBLOTHER_RT_SIZE 256 -#define QM_REG_AFULLQMBYPTHRPFWFQ_RT_OFFSET 34467 -#define QM_REG_AFULLQMBYPTHRVPWFQ_RT_OFFSET 34468 -#define QM_REG_AFULLQMBYPTHRPFRL_RT_OFFSET 34469 -#define QM_REG_AFULLQMBYPTHRGLBLRL_RT_OFFSET 34470 -#define QM_REG_AFULLOPRTNSTCCRDMASK_RT_OFFSET 34471 -#define QM_REG_WRROTHERPQGRP_0_RT_OFFSET 34472 -#define QM_REG_WRROTHERPQGRP_1_RT_OFFSET 34473 -#define QM_REG_WRROTHERPQGRP_2_RT_OFFSET 34474 -#define QM_REG_WRROTHERPQGRP_3_RT_OFFSET 34475 -#define QM_REG_WRROTHERPQGRP_4_RT_OFFSET 34476 -#define QM_REG_WRROTHERPQGRP_5_RT_OFFSET 34477 -#define QM_REG_WRROTHERPQGRP_6_RT_OFFSET 34478 -#define QM_REG_WRROTHERPQGRP_7_RT_OFFSET 34479 -#define QM_REG_WRROTHERPQGRP_8_RT_OFFSET 34480 -#define QM_REG_WRROTHERPQGRP_9_RT_OFFSET 34481 -#define QM_REG_WRROTHERPQGRP_10_RT_OFFSET 34482 -#define QM_REG_WRROTHERPQGRP_11_RT_OFFSET 34483 -#define QM_REG_WRROTHERPQGRP_12_RT_OFFSET 34484 -#define QM_REG_WRROTHERPQGRP_13_RT_OFFSET 34485 -#define QM_REG_WRROTHERPQGRP_14_RT_OFFSET 34486 -#define QM_REG_WRROTHERPQGRP_15_RT_OFFSET 34487 -#define QM_REG_WRROTHERGRPWEIGHT_0_RT_OFFSET 34488 -#define QM_REG_WRROTHERGRPWEIGHT_1_RT_OFFSET 34489 -#define QM_REG_WRROTHERGRPWEIGHT_2_RT_OFFSET 34490 -#define QM_REG_WRROTHERGRPWEIGHT_3_RT_OFFSET 34491 -#define QM_REG_WRRTXGRPWEIGHT_0_RT_OFFSET 34492 -#define QM_REG_WRRTXGRPWEIGHT_1_RT_OFFSET 34493 -#define QM_REG_PQTX2PF_0_RT_OFFSET 34494 -#define QM_REG_PQTX2PF_1_RT_OFFSET 34495 -#define QM_REG_PQTX2PF_2_RT_OFFSET 34496 -#define QM_REG_PQTX2PF_3_RT_OFFSET 34497 -#define QM_REG_PQTX2PF_4_RT_OFFSET 34498 -#define QM_REG_PQTX2PF_5_RT_OFFSET 34499 -#define QM_REG_PQTX2PF_6_RT_OFFSET 34500 -#define QM_REG_PQTX2PF_7_RT_OFFSET 34501 -#define QM_REG_PQTX2PF_8_RT_OFFSET 34502 -#define QM_REG_PQTX2PF_9_RT_OFFSET 34503 -#define QM_REG_PQTX2PF_10_RT_OFFSET 34504 -#define QM_REG_PQTX2PF_11_RT_OFFSET 34505 -#define QM_REG_PQTX2PF_12_RT_OFFSET 34506 -#define QM_REG_PQTX2PF_13_RT_OFFSET 34507 -#define QM_REG_PQTX2PF_14_RT_OFFSET 34508 -#define QM_REG_PQTX2PF_15_RT_OFFSET 34509 -#define QM_REG_PQTX2PF_16_RT_OFFSET 34510 -#define QM_REG_PQTX2PF_17_RT_OFFSET 34511 -#define QM_REG_PQTX2PF_18_RT_OFFSET 34512 -#define QM_REG_PQTX2PF_19_RT_OFFSET 34513 -#define QM_REG_PQTX2PF_20_RT_OFFSET 34514 -#define QM_REG_PQTX2PF_21_RT_OFFSET 34515 -#define QM_REG_PQTX2PF_22_RT_OFFSET 34516 -#define QM_REG_PQTX2PF_23_RT_OFFSET 34517 -#define QM_REG_PQTX2PF_24_RT_OFFSET 34518 -#define QM_REG_PQTX2PF_25_RT_OFFSET 34519 -#define QM_REG_PQTX2PF_26_RT_OFFSET 34520 -#define QM_REG_PQTX2PF_27_RT_OFFSET 34521 -#define QM_REG_PQTX2PF_28_RT_OFFSET 34522 -#define QM_REG_PQTX2PF_29_RT_OFFSET 34523 -#define QM_REG_PQTX2PF_30_RT_OFFSET 34524 -#define QM_REG_PQTX2PF_31_RT_OFFSET 34525 -#define QM_REG_PQTX2PF_32_RT_OFFSET 34526 -#define QM_REG_PQTX2PF_33_RT_OFFSET 34527 -#define QM_REG_PQTX2PF_34_RT_OFFSET 34528 -#define QM_REG_PQTX2PF_35_RT_OFFSET 34529 -#define QM_REG_PQTX2PF_36_RT_OFFSET 34530 -#define QM_REG_PQTX2PF_37_RT_OFFSET 34531 -#define QM_REG_PQTX2PF_38_RT_OFFSET 34532 -#define QM_REG_PQTX2PF_39_RT_OFFSET 34533 -#define QM_REG_PQTX2PF_40_RT_OFFSET 34534 -#define QM_REG_PQTX2PF_41_RT_OFFSET 34535 -#define QM_REG_PQTX2PF_42_RT_OFFSET 34536 -#define QM_REG_PQTX2PF_43_RT_OFFSET 34537 -#define QM_REG_PQTX2PF_44_RT_OFFSET 34538 -#define QM_REG_PQTX2PF_45_RT_OFFSET 34539 -#define QM_REG_PQTX2PF_46_RT_OFFSET 34540 -#define QM_REG_PQTX2PF_47_RT_OFFSET 34541 -#define QM_REG_PQTX2PF_48_RT_OFFSET 34542 -#define QM_REG_PQTX2PF_49_RT_OFFSET 34543 -#define QM_REG_PQTX2PF_50_RT_OFFSET 34544 -#define QM_REG_PQTX2PF_51_RT_OFFSET 34545 -#define QM_REG_PQTX2PF_52_RT_OFFSET 34546 -#define QM_REG_PQTX2PF_53_RT_OFFSET 34547 -#define QM_REG_PQTX2PF_54_RT_OFFSET 34548 -#define QM_REG_PQTX2PF_55_RT_OFFSET 34549 -#define QM_REG_PQTX2PF_56_RT_OFFSET 34550 -#define QM_REG_PQTX2PF_57_RT_OFFSET 34551 -#define QM_REG_PQTX2PF_58_RT_OFFSET 34552 -#define QM_REG_PQTX2PF_59_RT_OFFSET 34553 -#define QM_REG_PQTX2PF_60_RT_OFFSET 34554 -#define QM_REG_PQTX2PF_61_RT_OFFSET 34555 -#define QM_REG_PQTX2PF_62_RT_OFFSET 34556 -#define QM_REG_PQTX2PF_63_RT_OFFSET 34557 -#define QM_REG_PQOTHER2PF_0_RT_OFFSET 34558 -#define QM_REG_PQOTHER2PF_1_RT_OFFSET 34559 -#define QM_REG_PQOTHER2PF_2_RT_OFFSET 34560 -#define QM_REG_PQOTHER2PF_3_RT_OFFSET 34561 -#define QM_REG_PQOTHER2PF_4_RT_OFFSET 34562 -#define QM_REG_PQOTHER2PF_5_RT_OFFSET 34563 -#define QM_REG_PQOTHER2PF_6_RT_OFFSET 34564 -#define QM_REG_PQOTHER2PF_7_RT_OFFSET 34565 -#define QM_REG_PQOTHER2PF_8_RT_OFFSET 34566 -#define QM_REG_PQOTHER2PF_9_RT_OFFSET 34567 -#define QM_REG_PQOTHER2PF_10_RT_OFFSET 34568 -#define QM_REG_PQOTHER2PF_11_RT_OFFSET 34569 -#define QM_REG_PQOTHER2PF_12_RT_OFFSET 34570 -#define QM_REG_PQOTHER2PF_13_RT_OFFSET 34571 -#define QM_REG_PQOTHER2PF_14_RT_OFFSET 34572 -#define QM_REG_PQOTHER2PF_15_RT_OFFSET 34573 -#define QM_REG_RLGLBLPERIOD_0_RT_OFFSET 34574 -#define QM_REG_RLGLBLPERIOD_1_RT_OFFSET 34575 -#define QM_REG_RLGLBLPERIODTIMER_0_RT_OFFSET 34576 -#define QM_REG_RLGLBLPERIODTIMER_1_RT_OFFSET 34577 -#define QM_REG_RLGLBLPERIODSEL_0_RT_OFFSET 34578 -#define QM_REG_RLGLBLPERIODSEL_1_RT_OFFSET 34579 -#define QM_REG_RLGLBLPERIODSEL_2_RT_OFFSET 34580 -#define QM_REG_RLGLBLPERIODSEL_3_RT_OFFSET 34581 -#define QM_REG_RLGLBLPERIODSEL_4_RT_OFFSET 34582 -#define QM_REG_RLGLBLPERIODSEL_5_RT_OFFSET 34583 -#define QM_REG_RLGLBLPERIODSEL_6_RT_OFFSET 34584 -#define QM_REG_RLGLBLPERIODSEL_7_RT_OFFSET 34585 -#define QM_REG_RLGLBLINCVAL_RT_OFFSET 34586 -#define QM_REG_RLGLBLINCVAL_RT_SIZE 256 -#define QM_REG_RLGLBLUPPERBOUND_RT_OFFSET 34842 -#define QM_REG_RLGLBLUPPERBOUND_RT_SIZE 256 -#define QM_REG_RLGLBLCRD_RT_OFFSET 35098 -#define QM_REG_RLGLBLCRD_RT_SIZE 256 -#define QM_REG_RLGLBLENABLE_RT_OFFSET 35354 -#define QM_REG_RLPFPERIOD_RT_OFFSET 35355 -#define QM_REG_RLPFPERIODTIMER_RT_OFFSET 35356 -#define QM_REG_RLPFINCVAL_RT_OFFSET 35357 -#define QM_REG_RLPFINCVAL_RT_SIZE 16 -#define QM_REG_RLPFUPPERBOUND_RT_OFFSET 35373 -#define QM_REG_RLPFUPPERBOUND_RT_SIZE 16 -#define QM_REG_RLPFCRD_RT_OFFSET 35389 -#define QM_REG_RLPFCRD_RT_SIZE 16 -#define QM_REG_RLPFENABLE_RT_OFFSET 35405 -#define QM_REG_RLPFVOQENABLE_RT_OFFSET 35406 -#define QM_REG_WFQPFWEIGHT_RT_OFFSET 35407 -#define QM_REG_WFQPFWEIGHT_RT_SIZE 16 -#define QM_REG_WFQPFUPPERBOUND_RT_OFFSET 35423 -#define QM_REG_WFQPFUPPERBOUND_RT_SIZE 16 -#define QM_REG_WFQPFCRD_RT_OFFSET 35439 -#define QM_REG_WFQPFCRD_RT_SIZE 256 -#define QM_REG_WFQPFENABLE_RT_OFFSET 35695 -#define QM_REG_WFQVPENABLE_RT_OFFSET 35696 -#define QM_REG_BASEADDRTXPQ_RT_OFFSET 35697 -#define QM_REG_BASEADDRTXPQ_RT_SIZE 512 -#define QM_REG_TXPQMAP_RT_OFFSET 36209 -#define QM_REG_TXPQMAP_RT_SIZE 512 -#define QM_REG_WFQVPWEIGHT_RT_OFFSET 36721 -#define QM_REG_WFQVPWEIGHT_RT_SIZE 512 -#define QM_REG_WFQVPCRD_RT_OFFSET 37233 -#define QM_REG_WFQVPCRD_RT_SIZE 512 -#define QM_REG_WFQVPMAP_RT_OFFSET 37745 -#define QM_REG_WFQVPMAP_RT_SIZE 512 -#define QM_REG_PTRTBLTX_RT_OFFSET 38257 -#define QM_REG_PTRTBLTX_RT_SIZE 1024 -#define QM_REG_WFQPFCRD_MSB_RT_OFFSET 39281 -#define QM_REG_WFQPFCRD_MSB_RT_SIZE 320 -#define QM_REG_VOQCRDLINE_RT_OFFSET 39601 -#define QM_REG_VOQCRDLINE_RT_SIZE 36 -#define QM_REG_VOQINITCRDLINE_RT_OFFSET 39637 -#define QM_REG_VOQINITCRDLINE_RT_SIZE 36 -#define QM_REG_RLPFVOQENABLE_MSB_RT_OFFSET 39673 -#define NIG_REG_TAG_ETHERTYPE_0_RT_OFFSET 39674 -#define NIG_REG_BRB_GATE_DNTFWD_PORT_RT_OFFSET 39675 -#define NIG_REG_OUTER_TAG_VALUE_LIST0_RT_OFFSET 39676 -#define NIG_REG_OUTER_TAG_VALUE_LIST1_RT_OFFSET 39677 -#define NIG_REG_OUTER_TAG_VALUE_LIST2_RT_OFFSET 39678 -#define NIG_REG_OUTER_TAG_VALUE_LIST3_RT_OFFSET 39679 -#define NIG_REG_LLH_FUNC_TAGMAC_CLS_TYPE_RT_OFFSET 39680 -#define NIG_REG_LLH_FUNC_TAG_EN_RT_OFFSET 39681 -#define NIG_REG_LLH_FUNC_TAG_EN_RT_SIZE 4 -#define NIG_REG_LLH_FUNC_TAG_VALUE_RT_OFFSET 39685 -#define NIG_REG_LLH_FUNC_TAG_VALUE_RT_SIZE 4 -#define NIG_REG_LLH_FUNC_FILTER_VALUE_RT_OFFSET 39689 -#define NIG_REG_LLH_FUNC_FILTER_VALUE_RT_SIZE 32 -#define NIG_REG_LLH_FUNC_FILTER_EN_RT_OFFSET 39721 -#define NIG_REG_LLH_FUNC_FILTER_EN_RT_SIZE 16 -#define NIG_REG_LLH_FUNC_FILTER_MODE_RT_OFFSET 39737 -#define NIG_REG_LLH_FUNC_FILTER_MODE_RT_SIZE 16 -#define NIG_REG_LLH_FUNC_FILTER_PROTOCOL_TYPE_RT_OFFSET 39753 -#define NIG_REG_LLH_FUNC_FILTER_PROTOCOL_TYPE_RT_SIZE 16 -#define NIG_REG_LLH_FUNC_FILTER_HDR_SEL_RT_OFFSET 39769 -#define NIG_REG_LLH_FUNC_FILTER_HDR_SEL_RT_SIZE 16 -#define NIG_REG_TX_EDPM_CTRL_RT_OFFSET 39785 -#define NIG_REG_PPF_TO_ENGINE_SEL_RT_OFFSET 39786 -#define NIG_REG_PPF_TO_ENGINE_SEL_RT_SIZE 8 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_VALUE_RT_OFFSET 39794 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_VALUE_RT_SIZE 1024 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_EN_RT_OFFSET 40818 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_EN_RT_SIZE 512 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_MODE_RT_OFFSET 41330 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_MODE_RT_SIZE 512 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_PROTOCOL_TYPE_RT_OFFSET 41842 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_PROTOCOL_TYPE_RT_SIZE 512 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_HDR_SEL_RT_OFFSET 42354 -#define NIG_REG_LLH_PF_CLS_FUNC_FILTER_HDR_SEL_RT_SIZE 512 -#define NIG_REG_LLH_PF_CLS_FILTERS_MAP_RT_OFFSET 42866 -#define NIG_REG_LLH_PF_CLS_FILTERS_MAP_RT_SIZE 32 -#define CDU_REG_CID_ADDR_PARAMS_RT_OFFSET 42898 -#define CDU_REG_SEGMENT0_PARAMS_RT_OFFSET 42899 -#define CDU_REG_SEGMENT1_PARAMS_RT_OFFSET 42900 -#define CDU_REG_PF_SEG0_TYPE_OFFSET_RT_OFFSET 42901 -#define CDU_REG_PF_SEG1_TYPE_OFFSET_RT_OFFSET 42902 -#define CDU_REG_PF_SEG2_TYPE_OFFSET_RT_OFFSET 42903 -#define CDU_REG_PF_SEG3_TYPE_OFFSET_RT_OFFSET 42904 -#define CDU_REG_PF_FL_SEG0_TYPE_OFFSET_RT_OFFSET 42905 -#define CDU_REG_PF_FL_SEG1_TYPE_OFFSET_RT_OFFSET 42906 -#define CDU_REG_PF_FL_SEG2_TYPE_OFFSET_RT_OFFSET 42907 -#define CDU_REG_PF_FL_SEG3_TYPE_OFFSET_RT_OFFSET 42908 -#define CDU_REG_VF_SEG_TYPE_OFFSET_RT_OFFSET 42909 -#define CDU_REG_VF_FL_SEG_TYPE_OFFSET_RT_OFFSET 42910 -#define PBF_REG_TAG_ETHERTYPE_0_RT_OFFSET 42911 -#define PBF_REG_BTB_SHARED_AREA_SIZE_RT_OFFSET 42912 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ0_RT_OFFSET 42913 -#define PBF_REG_BTB_GUARANTEED_VOQ0_RT_OFFSET 42914 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ0_RT_OFFSET 42915 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ1_RT_OFFSET 42916 -#define PBF_REG_BTB_GUARANTEED_VOQ1_RT_OFFSET 42917 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ1_RT_OFFSET 42918 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ2_RT_OFFSET 42919 -#define PBF_REG_BTB_GUARANTEED_VOQ2_RT_OFFSET 42920 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ2_RT_OFFSET 42921 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ3_RT_OFFSET 42922 -#define PBF_REG_BTB_GUARANTEED_VOQ3_RT_OFFSET 42923 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ3_RT_OFFSET 42924 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ4_RT_OFFSET 42925 -#define PBF_REG_BTB_GUARANTEED_VOQ4_RT_OFFSET 42926 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ4_RT_OFFSET 42927 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ5_RT_OFFSET 42928 -#define PBF_REG_BTB_GUARANTEED_VOQ5_RT_OFFSET 42929 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ5_RT_OFFSET 42930 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ6_RT_OFFSET 42931 -#define PBF_REG_BTB_GUARANTEED_VOQ6_RT_OFFSET 42932 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ6_RT_OFFSET 42933 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ7_RT_OFFSET 42934 -#define PBF_REG_BTB_GUARANTEED_VOQ7_RT_OFFSET 42935 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ7_RT_OFFSET 42936 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ8_RT_OFFSET 42937 -#define PBF_REG_BTB_GUARANTEED_VOQ8_RT_OFFSET 42938 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ8_RT_OFFSET 42939 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ9_RT_OFFSET 42940 -#define PBF_REG_BTB_GUARANTEED_VOQ9_RT_OFFSET 42941 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ9_RT_OFFSET 42942 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ10_RT_OFFSET 42943 -#define PBF_REG_BTB_GUARANTEED_VOQ10_RT_OFFSET 42944 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ10_RT_OFFSET 42945 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ11_RT_OFFSET 42946 -#define PBF_REG_BTB_GUARANTEED_VOQ11_RT_OFFSET 42947 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ11_RT_OFFSET 42948 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ12_RT_OFFSET 42949 -#define PBF_REG_BTB_GUARANTEED_VOQ12_RT_OFFSET 42950 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ12_RT_OFFSET 42951 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ13_RT_OFFSET 42952 -#define PBF_REG_BTB_GUARANTEED_VOQ13_RT_OFFSET 42953 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ13_RT_OFFSET 42954 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ14_RT_OFFSET 42955 -#define PBF_REG_BTB_GUARANTEED_VOQ14_RT_OFFSET 42956 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ14_RT_OFFSET 42957 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ15_RT_OFFSET 42958 -#define PBF_REG_BTB_GUARANTEED_VOQ15_RT_OFFSET 42959 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ15_RT_OFFSET 42960 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ16_RT_OFFSET 42961 -#define PBF_REG_BTB_GUARANTEED_VOQ16_RT_OFFSET 42962 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ16_RT_OFFSET 42963 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ17_RT_OFFSET 42964 -#define PBF_REG_BTB_GUARANTEED_VOQ17_RT_OFFSET 42965 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ17_RT_OFFSET 42966 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ18_RT_OFFSET 42967 -#define PBF_REG_BTB_GUARANTEED_VOQ18_RT_OFFSET 42968 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ18_RT_OFFSET 42969 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ19_RT_OFFSET 42970 -#define PBF_REG_BTB_GUARANTEED_VOQ19_RT_OFFSET 42971 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ19_RT_OFFSET 42972 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ20_RT_OFFSET 42973 -#define PBF_REG_BTB_GUARANTEED_VOQ20_RT_OFFSET 42974 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ20_RT_OFFSET 42975 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ21_RT_OFFSET 42976 -#define PBF_REG_BTB_GUARANTEED_VOQ21_RT_OFFSET 42977 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ21_RT_OFFSET 42978 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ22_RT_OFFSET 42979 -#define PBF_REG_BTB_GUARANTEED_VOQ22_RT_OFFSET 42980 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ22_RT_OFFSET 42981 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ23_RT_OFFSET 42982 -#define PBF_REG_BTB_GUARANTEED_VOQ23_RT_OFFSET 42983 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ23_RT_OFFSET 42984 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ24_RT_OFFSET 42985 -#define PBF_REG_BTB_GUARANTEED_VOQ24_RT_OFFSET 42986 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ24_RT_OFFSET 42987 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ25_RT_OFFSET 42988 -#define PBF_REG_BTB_GUARANTEED_VOQ25_RT_OFFSET 42989 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ25_RT_OFFSET 42990 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ26_RT_OFFSET 42991 -#define PBF_REG_BTB_GUARANTEED_VOQ26_RT_OFFSET 42992 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ26_RT_OFFSET 42993 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ27_RT_OFFSET 42994 -#define PBF_REG_BTB_GUARANTEED_VOQ27_RT_OFFSET 42995 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ27_RT_OFFSET 42996 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ28_RT_OFFSET 42997 -#define PBF_REG_BTB_GUARANTEED_VOQ28_RT_OFFSET 42998 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ28_RT_OFFSET 42999 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ29_RT_OFFSET 43000 -#define PBF_REG_BTB_GUARANTEED_VOQ29_RT_OFFSET 43001 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ29_RT_OFFSET 43002 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ30_RT_OFFSET 43003 -#define PBF_REG_BTB_GUARANTEED_VOQ30_RT_OFFSET 43004 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ30_RT_OFFSET 43005 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ31_RT_OFFSET 43006 -#define PBF_REG_BTB_GUARANTEED_VOQ31_RT_OFFSET 43007 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ31_RT_OFFSET 43008 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ32_RT_OFFSET 43009 -#define PBF_REG_BTB_GUARANTEED_VOQ32_RT_OFFSET 43010 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ32_RT_OFFSET 43011 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ33_RT_OFFSET 43012 -#define PBF_REG_BTB_GUARANTEED_VOQ33_RT_OFFSET 43013 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ33_RT_OFFSET 43014 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ34_RT_OFFSET 43015 -#define PBF_REG_BTB_GUARANTEED_VOQ34_RT_OFFSET 43016 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ34_RT_OFFSET 43017 -#define PBF_REG_YCMD_QS_NUM_LINES_VOQ35_RT_OFFSET 43018 -#define PBF_REG_BTB_GUARANTEED_VOQ35_RT_OFFSET 43019 -#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ35_RT_OFFSET 43020 -#define XCM_REG_CON_PHY_Q3_RT_OFFSET 43021 - -#define RUNTIME_ARRAY_SIZE 43022 - +#define DORQ_REG_PF_MAX_ICID_0_RT_OFFSET 0 +#define DORQ_REG_PF_MAX_ICID_1_RT_OFFSET 1 +#define DORQ_REG_PF_MAX_ICID_2_RT_OFFSET 2 +#define DORQ_REG_PF_MAX_ICID_3_RT_OFFSET 3 +#define DORQ_REG_PF_MAX_ICID_4_RT_OFFSET 4 +#define DORQ_REG_PF_MAX_ICID_5_RT_OFFSET 5 +#define DORQ_REG_PF_MAX_ICID_6_RT_OFFSET 6 +#define DORQ_REG_PF_MAX_ICID_7_RT_OFFSET 7 +#define DORQ_REG_VF_MAX_ICID_0_RT_OFFSET 8 +#define DORQ_REG_VF_MAX_ICID_1_RT_OFFSET 9 +#define DORQ_REG_VF_MAX_ICID_2_RT_OFFSET 10 +#define DORQ_REG_VF_MAX_ICID_3_RT_OFFSET 11 +#define DORQ_REG_VF_MAX_ICID_4_RT_OFFSET 12 +#define DORQ_REG_VF_MAX_ICID_5_RT_OFFSET 13 +#define DORQ_REG_VF_MAX_ICID_6_RT_OFFSET 14 +#define DORQ_REG_VF_MAX_ICID_7_RT_OFFSET 15 +#define DORQ_REG_VF_ICID_BIT_SHIFT_NORM_RT_OFFSET 16 +#define DORQ_REG_PF_WAKE_ALL_RT_OFFSET 17 +#define DORQ_REG_TAG1_ETHERTYPE_RT_OFFSET 18 +#define IGU_REG_PF_CONFIGURATION_RT_OFFSET 19 +#define IGU_REG_VF_CONFIGURATION_RT_OFFSET 20 +#define IGU_REG_ATTN_MSG_ADDR_L_RT_OFFSET 21 +#define IGU_REG_ATTN_MSG_ADDR_H_RT_OFFSET 22 +#define IGU_REG_LEADING_EDGE_LATCH_RT_OFFSET 23 +#define IGU_REG_TRAILING_EDGE_LATCH_RT_OFFSET 24 +#define CAU_REG_CQE_AGG_UNIT_SIZE_RT_OFFSET 25 +#define CAU_REG_SB_VAR_MEMORY_RT_OFFSET 26 +#define CAU_REG_SB_VAR_MEMORY_RT_SIZE 736 +#define CAU_REG_SB_ADDR_MEMORY_RT_OFFSET 762 +#define CAU_REG_SB_ADDR_MEMORY_RT_SIZE 736 +#define CAU_REG_PI_MEMORY_RT_OFFSET 1498 +#define CAU_REG_PI_MEMORY_RT_SIZE 4416 +#define PRS_REG_SEARCH_RESP_INITIATOR_TYPE_RT_OFFSET 5914 +#define PRS_REG_TASK_ID_MAX_INITIATOR_PF_RT_OFFSET 5915 +#define PRS_REG_TASK_ID_MAX_INITIATOR_VF_RT_OFFSET 5916 +#define PRS_REG_TASK_ID_MAX_TARGET_PF_RT_OFFSET 5917 +#define PRS_REG_TASK_ID_MAX_TARGET_VF_RT_OFFSET 5918 +#define PRS_REG_SEARCH_TCP_RT_OFFSET 5919 +#define PRS_REG_SEARCH_FCOE_RT_OFFSET 5920 +#define PRS_REG_SEARCH_ROCE_RT_OFFSET 5921 +#define PRS_REG_ROCE_DEST_QP_MAX_VF_RT_OFFSET 5922 +#define PRS_REG_ROCE_DEST_QP_MAX_PF_RT_OFFSET 5923 +#define PRS_REG_SEARCH_OPENFLOW_RT_OFFSET 5924 +#define PRS_REG_SEARCH_NON_IP_AS_OPENFLOW_RT_OFFSET 5925 +#define PRS_REG_OPENFLOW_SUPPORT_ONLY_KNOWN_OVER_IP_RT_OFFSET 5926 +#define PRS_REG_OPENFLOW_SEARCH_KEY_MASK_RT_OFFSET 5927 +#define PRS_REG_TAG_ETHERTYPE_0_RT_OFFSET 5928 +#define PRS_REG_LIGHT_L2_ETHERTYPE_EN_RT_OFFSET 5929 +#define SRC_REG_FIRSTFREE_RT_OFFSET 5930 +#define SRC_REG_FIRSTFREE_RT_SIZE 2 +#define SRC_REG_LASTFREE_RT_OFFSET 5932 +#define SRC_REG_LASTFREE_RT_SIZE 2 +#define SRC_REG_COUNTFREE_RT_OFFSET 5934 +#define SRC_REG_NUMBER_HASH_BITS_RT_OFFSET 5935 +#define PSWRQ2_REG_CDUT_P_SIZE_RT_OFFSET 5936 +#define PSWRQ2_REG_CDUC_P_SIZE_RT_OFFSET 5937 +#define PSWRQ2_REG_TM_P_SIZE_RT_OFFSET 5938 +#define PSWRQ2_REG_QM_P_SIZE_RT_OFFSET 5939 +#define PSWRQ2_REG_SRC_P_SIZE_RT_OFFSET 5940 +#define PSWRQ2_REG_TSDM_P_SIZE_RT_OFFSET 5941 +#define PSWRQ2_REG_TM_FIRST_ILT_RT_OFFSET 5942 +#define PSWRQ2_REG_TM_LAST_ILT_RT_OFFSET 5943 +#define PSWRQ2_REG_QM_FIRST_ILT_RT_OFFSET 5944 +#define PSWRQ2_REG_QM_LAST_ILT_RT_OFFSET 5945 +#define PSWRQ2_REG_SRC_FIRST_ILT_RT_OFFSET 5946 +#define PSWRQ2_REG_SRC_LAST_ILT_RT_OFFSET 5947 +#define PSWRQ2_REG_CDUC_FIRST_ILT_RT_OFFSET 5948 +#define PSWRQ2_REG_CDUC_LAST_ILT_RT_OFFSET 5949 +#define PSWRQ2_REG_CDUT_FIRST_ILT_RT_OFFSET 5950 +#define PSWRQ2_REG_CDUT_LAST_ILT_RT_OFFSET 5951 +#define PSWRQ2_REG_TSDM_FIRST_ILT_RT_OFFSET 5952 +#define PSWRQ2_REG_TSDM_LAST_ILT_RT_OFFSET 5953 +#define PSWRQ2_REG_TM_NUMBER_OF_PF_BLOCKS_RT_OFFSET 5954 +#define PSWRQ2_REG_CDUT_NUMBER_OF_PF_BLOCKS_RT_OFFSET 5955 +#define PSWRQ2_REG_CDUC_NUMBER_OF_PF_BLOCKS_RT_OFFSET 5956 +#define PSWRQ2_REG_TM_VF_BLOCKS_RT_OFFSET 5957 +#define PSWRQ2_REG_CDUT_VF_BLOCKS_RT_OFFSET 5958 +#define PSWRQ2_REG_CDUC_VF_BLOCKS_RT_OFFSET 5959 +#define PSWRQ2_REG_TM_BLOCKS_FACTOR_RT_OFFSET 5960 +#define PSWRQ2_REG_CDUT_BLOCKS_FACTOR_RT_OFFSET 5961 +#define PSWRQ2_REG_CDUC_BLOCKS_FACTOR_RT_OFFSET 5962 +#define PSWRQ2_REG_VF_BASE_RT_OFFSET 5963 +#define PSWRQ2_REG_VF_LAST_ILT_RT_OFFSET 5964 +#define PSWRQ2_REG_DRAM_ALIGN_WR_RT_OFFSET 5965 +#define PSWRQ2_REG_DRAM_ALIGN_RD_RT_OFFSET 5966 +#define PSWRQ2_REG_ILT_MEMORY_RT_OFFSET 5967 +#define PSWRQ2_REG_ILT_MEMORY_RT_SIZE 22000 +#define PGLUE_REG_B_VF_BASE_RT_OFFSET 27967 +#define PGLUE_REG_B_MSDM_OFFSET_MASK_B_RT_OFFSET 27968 +#define PGLUE_REG_B_MSDM_VF_SHIFT_B_RT_OFFSET 27969 +#define PGLUE_REG_B_CACHE_LINE_SIZE_RT_OFFSET 27970 +#define PGLUE_REG_B_PF_BAR0_SIZE_RT_OFFSET 27971 +#define PGLUE_REG_B_PF_BAR1_SIZE_RT_OFFSET 27972 +#define PGLUE_REG_B_VF_BAR1_SIZE_RT_OFFSET 27973 +#define TM_REG_VF_ENABLE_CONN_RT_OFFSET 27974 +#define TM_REG_PF_ENABLE_CONN_RT_OFFSET 27975 +#define TM_REG_PF_ENABLE_TASK_RT_OFFSET 27976 +#define TM_REG_GROUP_SIZE_RESOLUTION_CONN_RT_OFFSET 27977 +#define TM_REG_GROUP_SIZE_RESOLUTION_TASK_RT_OFFSET 27978 +#define TM_REG_CONFIG_CONN_MEM_RT_OFFSET 27979 +#define TM_REG_CONFIG_CONN_MEM_RT_SIZE 416 +#define TM_REG_CONFIG_TASK_MEM_RT_OFFSET 28395 +#define TM_REG_CONFIG_TASK_MEM_RT_SIZE 512 +#define QM_REG_MAXPQSIZE_0_RT_OFFSET 28907 +#define QM_REG_MAXPQSIZE_1_RT_OFFSET 28908 +#define QM_REG_MAXPQSIZE_2_RT_OFFSET 28909 +#define QM_REG_MAXPQSIZETXSEL_0_RT_OFFSET 28910 +#define QM_REG_MAXPQSIZETXSEL_1_RT_OFFSET 28911 +#define QM_REG_MAXPQSIZETXSEL_2_RT_OFFSET 28912 +#define QM_REG_MAXPQSIZETXSEL_3_RT_OFFSET 28913 +#define QM_REG_MAXPQSIZETXSEL_4_RT_OFFSET 28914 +#define QM_REG_MAXPQSIZETXSEL_5_RT_OFFSET 28915 +#define QM_REG_MAXPQSIZETXSEL_6_RT_OFFSET 28916 +#define QM_REG_MAXPQSIZETXSEL_7_RT_OFFSET 28917 +#define QM_REG_MAXPQSIZETXSEL_8_RT_OFFSET 28918 +#define QM_REG_MAXPQSIZETXSEL_9_RT_OFFSET 28919 +#define QM_REG_MAXPQSIZETXSEL_10_RT_OFFSET 28920 +#define QM_REG_MAXPQSIZETXSEL_11_RT_OFFSET 28921 +#define QM_REG_MAXPQSIZETXSEL_12_RT_OFFSET 28922 +#define QM_REG_MAXPQSIZETXSEL_13_RT_OFFSET 28923 +#define QM_REG_MAXPQSIZETXSEL_14_RT_OFFSET 28924 +#define QM_REG_MAXPQSIZETXSEL_15_RT_OFFSET 28925 +#define QM_REG_MAXPQSIZETXSEL_16_RT_OFFSET 28926 +#define QM_REG_MAXPQSIZETXSEL_17_RT_OFFSET 28927 +#define QM_REG_MAXPQSIZETXSEL_18_RT_OFFSET 28928 +#define QM_REG_MAXPQSIZETXSEL_19_RT_OFFSET 28929 +#define QM_REG_MAXPQSIZETXSEL_20_RT_OFFSET 28930 +#define QM_REG_MAXPQSIZETXSEL_21_RT_OFFSET 28931 +#define QM_REG_MAXPQSIZETXSEL_22_RT_OFFSET 28932 +#define QM_REG_MAXPQSIZETXSEL_23_RT_OFFSET 28933 +#define QM_REG_MAXPQSIZETXSEL_24_RT_OFFSET 28934 +#define QM_REG_MAXPQSIZETXSEL_25_RT_OFFSET 28935 +#define QM_REG_MAXPQSIZETXSEL_26_RT_OFFSET 28936 +#define QM_REG_MAXPQSIZETXSEL_27_RT_OFFSET 28937 +#define QM_REG_MAXPQSIZETXSEL_28_RT_OFFSET 28938 +#define QM_REG_MAXPQSIZETXSEL_29_RT_OFFSET 28939 +#define QM_REG_MAXPQSIZETXSEL_30_RT_OFFSET 28940 +#define QM_REG_MAXPQSIZETXSEL_31_RT_OFFSET 28941 +#define QM_REG_MAXPQSIZETXSEL_32_RT_OFFSET 28942 +#define QM_REG_MAXPQSIZETXSEL_33_RT_OFFSET 28943 +#define QM_REG_MAXPQSIZETXSEL_34_RT_OFFSET 28944 +#define QM_REG_MAXPQSIZETXSEL_35_RT_OFFSET 28945 +#define QM_REG_MAXPQSIZETXSEL_36_RT_OFFSET 28946 +#define QM_REG_MAXPQSIZETXSEL_37_RT_OFFSET 28947 +#define QM_REG_MAXPQSIZETXSEL_38_RT_OFFSET 28948 +#define QM_REG_MAXPQSIZETXSEL_39_RT_OFFSET 28949 +#define QM_REG_MAXPQSIZETXSEL_40_RT_OFFSET 28950 +#define QM_REG_MAXPQSIZETXSEL_41_RT_OFFSET 28951 +#define QM_REG_MAXPQSIZETXSEL_42_RT_OFFSET 28952 +#define QM_REG_MAXPQSIZETXSEL_43_RT_OFFSET 28953 +#define QM_REG_MAXPQSIZETXSEL_44_RT_OFFSET 28954 +#define QM_REG_MAXPQSIZETXSEL_45_RT_OFFSET 28955 +#define QM_REG_MAXPQSIZETXSEL_46_RT_OFFSET 28956 +#define QM_REG_MAXPQSIZETXSEL_47_RT_OFFSET 28957 +#define QM_REG_MAXPQSIZETXSEL_48_RT_OFFSET 28958 +#define QM_REG_MAXPQSIZETXSEL_49_RT_OFFSET 28959 +#define QM_REG_MAXPQSIZETXSEL_50_RT_OFFSET 28960 +#define QM_REG_MAXPQSIZETXSEL_51_RT_OFFSET 28961 +#define QM_REG_MAXPQSIZETXSEL_52_RT_OFFSET 28962 +#define QM_REG_MAXPQSIZETXSEL_53_RT_OFFSET 28963 +#define QM_REG_MAXPQSIZETXSEL_54_RT_OFFSET 28964 +#define QM_REG_MAXPQSIZETXSEL_55_RT_OFFSET 28965 +#define QM_REG_MAXPQSIZETXSEL_56_RT_OFFSET 28966 +#define QM_REG_MAXPQSIZETXSEL_57_RT_OFFSET 28967 +#define QM_REG_MAXPQSIZETXSEL_58_RT_OFFSET 28968 +#define QM_REG_MAXPQSIZETXSEL_59_RT_OFFSET 28969 +#define QM_REG_MAXPQSIZETXSEL_60_RT_OFFSET 28970 +#define QM_REG_MAXPQSIZETXSEL_61_RT_OFFSET 28971 +#define QM_REG_MAXPQSIZETXSEL_62_RT_OFFSET 28972 +#define QM_REG_MAXPQSIZETXSEL_63_RT_OFFSET 28973 +#define QM_REG_BASEADDROTHERPQ_RT_OFFSET 28974 +#define QM_REG_BASEADDROTHERPQ_RT_SIZE 128 +#define QM_REG_PTRTBLOTHER_RT_OFFSET 29102 +#define QM_REG_PTRTBLOTHER_RT_SIZE 256 +#define QM_REG_VOQCRDLINE_RT_OFFSET 29358 +#define QM_REG_VOQCRDLINE_RT_SIZE 20 +#define QM_REG_VOQINITCRDLINE_RT_OFFSET 29378 +#define QM_REG_VOQINITCRDLINE_RT_SIZE 20 +#define QM_REG_AFULLQMBYPTHRPFWFQ_RT_OFFSET 29398 +#define QM_REG_AFULLQMBYPTHRVPWFQ_RT_OFFSET 29399 +#define QM_REG_AFULLQMBYPTHRPFRL_RT_OFFSET 29400 +#define QM_REG_AFULLQMBYPTHRGLBLRL_RT_OFFSET 29401 +#define QM_REG_AFULLOPRTNSTCCRDMASK_RT_OFFSET 29402 +#define QM_REG_WRROTHERPQGRP_0_RT_OFFSET 29403 +#define QM_REG_WRROTHERPQGRP_1_RT_OFFSET 29404 +#define QM_REG_WRROTHERPQGRP_2_RT_OFFSET 29405 +#define QM_REG_WRROTHERPQGRP_3_RT_OFFSET 29406 +#define QM_REG_WRROTHERPQGRP_4_RT_OFFSET 29407 +#define QM_REG_WRROTHERPQGRP_5_RT_OFFSET 29408 +#define QM_REG_WRROTHERPQGRP_6_RT_OFFSET 29409 +#define QM_REG_WRROTHERPQGRP_7_RT_OFFSET 29410 +#define QM_REG_WRROTHERPQGRP_8_RT_OFFSET 29411 +#define QM_REG_WRROTHERPQGRP_9_RT_OFFSET 29412 +#define QM_REG_WRROTHERPQGRP_10_RT_OFFSET 29413 +#define QM_REG_WRROTHERPQGRP_11_RT_OFFSET 29414 +#define QM_REG_WRROTHERPQGRP_12_RT_OFFSET 29415 +#define QM_REG_WRROTHERPQGRP_13_RT_OFFSET 29416 +#define QM_REG_WRROTHERPQGRP_14_RT_OFFSET 29417 +#define QM_REG_WRROTHERPQGRP_15_RT_OFFSET 29418 +#define QM_REG_WRROTHERGRPWEIGHT_0_RT_OFFSET 29419 +#define QM_REG_WRROTHERGRPWEIGHT_1_RT_OFFSET 29420 +#define QM_REG_WRROTHERGRPWEIGHT_2_RT_OFFSET 29421 +#define QM_REG_WRROTHERGRPWEIGHT_3_RT_OFFSET 29422 +#define QM_REG_WRRTXGRPWEIGHT_0_RT_OFFSET 29423 +#define QM_REG_WRRTXGRPWEIGHT_1_RT_OFFSET 29424 +#define QM_REG_PQTX2PF_0_RT_OFFSET 29425 +#define QM_REG_PQTX2PF_1_RT_OFFSET 29426 +#define QM_REG_PQTX2PF_2_RT_OFFSET 29427 +#define QM_REG_PQTX2PF_3_RT_OFFSET 29428 +#define QM_REG_PQTX2PF_4_RT_OFFSET 29429 +#define QM_REG_PQTX2PF_5_RT_OFFSET 29430 +#define QM_REG_PQTX2PF_6_RT_OFFSET 29431 +#define QM_REG_PQTX2PF_7_RT_OFFSET 29432 +#define QM_REG_PQTX2PF_8_RT_OFFSET 29433 +#define QM_REG_PQTX2PF_9_RT_OFFSET 29434 +#define QM_REG_PQTX2PF_10_RT_OFFSET 29435 +#define QM_REG_PQTX2PF_11_RT_OFFSET 29436 +#define QM_REG_PQTX2PF_12_RT_OFFSET 29437 +#define QM_REG_PQTX2PF_13_RT_OFFSET 29438 +#define QM_REG_PQTX2PF_14_RT_OFFSET 29439 +#define QM_REG_PQTX2PF_15_RT_OFFSET 29440 +#define QM_REG_PQTX2PF_16_RT_OFFSET 29441 +#define QM_REG_PQTX2PF_17_RT_OFFSET 29442 +#define QM_REG_PQTX2PF_18_RT_OFFSET 29443 +#define QM_REG_PQTX2PF_19_RT_OFFSET 29444 +#define QM_REG_PQTX2PF_20_RT_OFFSET 29445 +#define QM_REG_PQTX2PF_21_RT_OFFSET 29446 +#define QM_REG_PQTX2PF_22_RT_OFFSET 29447 +#define QM_REG_PQTX2PF_23_RT_OFFSET 29448 +#define QM_REG_PQTX2PF_24_RT_OFFSET 29449 +#define QM_REG_PQTX2PF_25_RT_OFFSET 29450 +#define QM_REG_PQTX2PF_26_RT_OFFSET 29451 +#define QM_REG_PQTX2PF_27_RT_OFFSET 29452 +#define QM_REG_PQTX2PF_28_RT_OFFSET 29453 +#define QM_REG_PQTX2PF_29_RT_OFFSET 29454 +#define QM_REG_PQTX2PF_30_RT_OFFSET 29455 +#define QM_REG_PQTX2PF_31_RT_OFFSET 29456 +#define QM_REG_PQTX2PF_32_RT_OFFSET 29457 +#define QM_REG_PQTX2PF_33_RT_OFFSET 29458 +#define QM_REG_PQTX2PF_34_RT_OFFSET 29459 +#define QM_REG_PQTX2PF_35_RT_OFFSET 29460 +#define QM_REG_PQTX2PF_36_RT_OFFSET 29461 +#define QM_REG_PQTX2PF_37_RT_OFFSET 29462 +#define QM_REG_PQTX2PF_38_RT_OFFSET 29463 +#define QM_REG_PQTX2PF_39_RT_OFFSET 29464 +#define QM_REG_PQTX2PF_40_RT_OFFSET 29465 +#define QM_REG_PQTX2PF_41_RT_OFFSET 29466 +#define QM_REG_PQTX2PF_42_RT_OFFSET 29467 +#define QM_REG_PQTX2PF_43_RT_OFFSET 29468 +#define QM_REG_PQTX2PF_44_RT_OFFSET 29469 +#define QM_REG_PQTX2PF_45_RT_OFFSET 29470 +#define QM_REG_PQTX2PF_46_RT_OFFSET 29471 +#define QM_REG_PQTX2PF_47_RT_OFFSET 29472 +#define QM_REG_PQTX2PF_48_RT_OFFSET 29473 +#define QM_REG_PQTX2PF_49_RT_OFFSET 29474 +#define QM_REG_PQTX2PF_50_RT_OFFSET 29475 +#define QM_REG_PQTX2PF_51_RT_OFFSET 29476 +#define QM_REG_PQTX2PF_52_RT_OFFSET 29477 +#define QM_REG_PQTX2PF_53_RT_OFFSET 29478 +#define QM_REG_PQTX2PF_54_RT_OFFSET 29479 +#define QM_REG_PQTX2PF_55_RT_OFFSET 29480 +#define QM_REG_PQTX2PF_56_RT_OFFSET 29481 +#define QM_REG_PQTX2PF_57_RT_OFFSET 29482 +#define QM_REG_PQTX2PF_58_RT_OFFSET 29483 +#define QM_REG_PQTX2PF_59_RT_OFFSET 29484 +#define QM_REG_PQTX2PF_60_RT_OFFSET 29485 +#define QM_REG_PQTX2PF_61_RT_OFFSET 29486 +#define QM_REG_PQTX2PF_62_RT_OFFSET 29487 +#define QM_REG_PQTX2PF_63_RT_OFFSET 29488 +#define QM_REG_PQOTHER2PF_0_RT_OFFSET 29489 +#define QM_REG_PQOTHER2PF_1_RT_OFFSET 29490 +#define QM_REG_PQOTHER2PF_2_RT_OFFSET 29491 +#define QM_REG_PQOTHER2PF_3_RT_OFFSET 29492 +#define QM_REG_PQOTHER2PF_4_RT_OFFSET 29493 +#define QM_REG_PQOTHER2PF_5_RT_OFFSET 29494 +#define QM_REG_PQOTHER2PF_6_RT_OFFSET 29495 +#define QM_REG_PQOTHER2PF_7_RT_OFFSET 29496 +#define QM_REG_PQOTHER2PF_8_RT_OFFSET 29497 +#define QM_REG_PQOTHER2PF_9_RT_OFFSET 29498 +#define QM_REG_PQOTHER2PF_10_RT_OFFSET 29499 +#define QM_REG_PQOTHER2PF_11_RT_OFFSET 29500 +#define QM_REG_PQOTHER2PF_12_RT_OFFSET 29501 +#define QM_REG_PQOTHER2PF_13_RT_OFFSET 29502 +#define QM_REG_PQOTHER2PF_14_RT_OFFSET 29503 +#define QM_REG_PQOTHER2PF_15_RT_OFFSET 29504 +#define QM_REG_RLGLBLPERIOD_0_RT_OFFSET 29505 +#define QM_REG_RLGLBLPERIOD_1_RT_OFFSET 29506 +#define QM_REG_RLGLBLPERIODTIMER_0_RT_OFFSET 29507 +#define QM_REG_RLGLBLPERIODTIMER_1_RT_OFFSET 29508 +#define QM_REG_RLGLBLPERIODSEL_0_RT_OFFSET 29509 +#define QM_REG_RLGLBLPERIODSEL_1_RT_OFFSET 29510 +#define QM_REG_RLGLBLPERIODSEL_2_RT_OFFSET 29511 +#define QM_REG_RLGLBLPERIODSEL_3_RT_OFFSET 29512 +#define QM_REG_RLGLBLPERIODSEL_4_RT_OFFSET 29513 +#define QM_REG_RLGLBLPERIODSEL_5_RT_OFFSET 29514 +#define QM_REG_RLGLBLPERIODSEL_6_RT_OFFSET 29515 +#define QM_REG_RLGLBLPERIODSEL_7_RT_OFFSET 29516 +#define QM_REG_RLGLBLINCVAL_RT_OFFSET 29517 +#define QM_REG_RLGLBLINCVAL_RT_SIZE 256 +#define QM_REG_RLGLBLUPPERBOUND_RT_OFFSET 29773 +#define QM_REG_RLGLBLUPPERBOUND_RT_SIZE 256 +#define QM_REG_RLGLBLCRD_RT_OFFSET 30029 +#define QM_REG_RLGLBLCRD_RT_SIZE 256 +#define QM_REG_RLGLBLENABLE_RT_OFFSET 30285 +#define QM_REG_RLPFPERIOD_RT_OFFSET 30286 +#define QM_REG_RLPFPERIODTIMER_RT_OFFSET 30287 +#define QM_REG_RLPFINCVAL_RT_OFFSET 30288 +#define QM_REG_RLPFINCVAL_RT_SIZE 16 +#define QM_REG_RLPFUPPERBOUND_RT_OFFSET 30304 +#define QM_REG_RLPFUPPERBOUND_RT_SIZE 16 +#define QM_REG_RLPFCRD_RT_OFFSET 30320 +#define QM_REG_RLPFCRD_RT_SIZE 16 +#define QM_REG_RLPFENABLE_RT_OFFSET 30336 +#define QM_REG_RLPFVOQENABLE_RT_OFFSET 30337 +#define QM_REG_WFQPFWEIGHT_RT_OFFSET 30338 +#define QM_REG_WFQPFWEIGHT_RT_SIZE 16 +#define QM_REG_WFQPFUPPERBOUND_RT_OFFSET 30354 +#define QM_REG_WFQPFUPPERBOUND_RT_SIZE 16 +#define QM_REG_WFQPFCRD_RT_OFFSET 30370 +#define QM_REG_WFQPFCRD_RT_SIZE 160 +#define QM_REG_WFQPFENABLE_RT_OFFSET 30530 +#define QM_REG_WFQVPENABLE_RT_OFFSET 30531 +#define QM_REG_BASEADDRTXPQ_RT_OFFSET 30532 +#define QM_REG_BASEADDRTXPQ_RT_SIZE 512 +#define QM_REG_TXPQMAP_RT_OFFSET 31044 +#define QM_REG_TXPQMAP_RT_SIZE 512 +#define QM_REG_WFQVPWEIGHT_RT_OFFSET 31556 +#define QM_REG_WFQVPWEIGHT_RT_SIZE 512 +#define QM_REG_WFQVPCRD_RT_OFFSET 32068 +#define QM_REG_WFQVPCRD_RT_SIZE 512 +#define QM_REG_WFQVPMAP_RT_OFFSET 32580 +#define QM_REG_WFQVPMAP_RT_SIZE 512 +#define QM_REG_PTRTBLTX_RT_OFFSET 33092 +#define QM_REG_PTRTBLTX_RT_SIZE 1024 +#define QM_REG_WFQPFCRD_MSB_RT_OFFSET 34116 +#define QM_REG_WFQPFCRD_MSB_RT_SIZE 160 +#define NIG_REG_TAG_ETHERTYPE_0_RT_OFFSET 34276 +#define NIG_REG_BRB_GATE_DNTFWD_PORT_RT_OFFSET 34277 +#define NIG_REG_OUTER_TAG_VALUE_LIST0_RT_OFFSET 34278 +#define NIG_REG_OUTER_TAG_VALUE_LIST1_RT_OFFSET 34279 +#define NIG_REG_OUTER_TAG_VALUE_LIST2_RT_OFFSET 34280 +#define NIG_REG_OUTER_TAG_VALUE_LIST3_RT_OFFSET 34281 +#define NIG_REG_LLH_FUNC_TAGMAC_CLS_TYPE_RT_OFFSET 34282 +#define NIG_REG_LLH_FUNC_TAG_EN_RT_OFFSET 34283 +#define NIG_REG_LLH_FUNC_TAG_EN_RT_SIZE 4 +#define NIG_REG_LLH_FUNC_TAG_VALUE_RT_OFFSET 34287 +#define NIG_REG_LLH_FUNC_TAG_VALUE_RT_SIZE 4 +#define NIG_REG_LLH_FUNC_FILTER_VALUE_RT_OFFSET 34291 +#define NIG_REG_LLH_FUNC_FILTER_VALUE_RT_SIZE 32 +#define NIG_REG_LLH_FUNC_FILTER_EN_RT_OFFSET 34323 +#define NIG_REG_LLH_FUNC_FILTER_EN_RT_SIZE 16 +#define NIG_REG_LLH_FUNC_FILTER_MODE_RT_OFFSET 34339 +#define NIG_REG_LLH_FUNC_FILTER_MODE_RT_SIZE 16 +#define NIG_REG_LLH_FUNC_FILTER_PROTOCOL_TYPE_RT_OFFSET 34355 +#define NIG_REG_LLH_FUNC_FILTER_PROTOCOL_TYPE_RT_SIZE 16 +#define NIG_REG_LLH_FUNC_FILTER_HDR_SEL_RT_OFFSET 34371 +#define NIG_REG_LLH_FUNC_FILTER_HDR_SEL_RT_SIZE 16 +#define NIG_REG_TX_EDPM_CTRL_RT_OFFSET 34387 +#define NIG_REG_PPF_TO_ENGINE_SEL_RT_OFFSET 34388 +#define NIG_REG_PPF_TO_ENGINE_SEL_RT_SIZE 8 +#define CDU_REG_CID_ADDR_PARAMS_RT_OFFSET 34396 +#define CDU_REG_SEGMENT0_PARAMS_RT_OFFSET 34397 +#define CDU_REG_SEGMENT1_PARAMS_RT_OFFSET 34398 +#define CDU_REG_PF_SEG0_TYPE_OFFSET_RT_OFFSET 34399 +#define CDU_REG_PF_SEG1_TYPE_OFFSET_RT_OFFSET 34400 +#define CDU_REG_PF_SEG2_TYPE_OFFSET_RT_OFFSET 34401 +#define CDU_REG_PF_SEG3_TYPE_OFFSET_RT_OFFSET 34402 +#define CDU_REG_PF_FL_SEG0_TYPE_OFFSET_RT_OFFSET 34403 +#define CDU_REG_PF_FL_SEG1_TYPE_OFFSET_RT_OFFSET 34404 +#define CDU_REG_PF_FL_SEG2_TYPE_OFFSET_RT_OFFSET 34405 +#define CDU_REG_PF_FL_SEG3_TYPE_OFFSET_RT_OFFSET 34406 +#define CDU_REG_VF_SEG_TYPE_OFFSET_RT_OFFSET 34407 +#define CDU_REG_VF_FL_SEG_TYPE_OFFSET_RT_OFFSET 34408 +#define PBF_REG_TAG_ETHERTYPE_0_RT_OFFSET 34409 +#define PBF_REG_BTB_SHARED_AREA_SIZE_RT_OFFSET 34410 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ0_RT_OFFSET 34411 +#define PBF_REG_BTB_GUARANTEED_VOQ0_RT_OFFSET 34412 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ0_RT_OFFSET 34413 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ1_RT_OFFSET 34414 +#define PBF_REG_BTB_GUARANTEED_VOQ1_RT_OFFSET 34415 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ1_RT_OFFSET 34416 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ2_RT_OFFSET 34417 +#define PBF_REG_BTB_GUARANTEED_VOQ2_RT_OFFSET 34418 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ2_RT_OFFSET 34419 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ3_RT_OFFSET 34420 +#define PBF_REG_BTB_GUARANTEED_VOQ3_RT_OFFSET 34421 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ3_RT_OFFSET 34422 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ4_RT_OFFSET 34423 +#define PBF_REG_BTB_GUARANTEED_VOQ4_RT_OFFSET 34424 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ4_RT_OFFSET 34425 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ5_RT_OFFSET 34426 +#define PBF_REG_BTB_GUARANTEED_VOQ5_RT_OFFSET 34427 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ5_RT_OFFSET 34428 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ6_RT_OFFSET 34429 +#define PBF_REG_BTB_GUARANTEED_VOQ6_RT_OFFSET 34430 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ6_RT_OFFSET 34431 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ7_RT_OFFSET 34432 +#define PBF_REG_BTB_GUARANTEED_VOQ7_RT_OFFSET 34433 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ7_RT_OFFSET 34434 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ8_RT_OFFSET 34435 +#define PBF_REG_BTB_GUARANTEED_VOQ8_RT_OFFSET 34436 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ8_RT_OFFSET 34437 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ9_RT_OFFSET 34438 +#define PBF_REG_BTB_GUARANTEED_VOQ9_RT_OFFSET 34439 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ9_RT_OFFSET 34440 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ10_RT_OFFSET 34441 +#define PBF_REG_BTB_GUARANTEED_VOQ10_RT_OFFSET 34442 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ10_RT_OFFSET 34443 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ11_RT_OFFSET 34444 +#define PBF_REG_BTB_GUARANTEED_VOQ11_RT_OFFSET 34445 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ11_RT_OFFSET 34446 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ12_RT_OFFSET 34447 +#define PBF_REG_BTB_GUARANTEED_VOQ12_RT_OFFSET 34448 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ12_RT_OFFSET 34449 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ13_RT_OFFSET 34450 +#define PBF_REG_BTB_GUARANTEED_VOQ13_RT_OFFSET 34451 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ13_RT_OFFSET 34452 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ14_RT_OFFSET 34453 +#define PBF_REG_BTB_GUARANTEED_VOQ14_RT_OFFSET 34454 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ14_RT_OFFSET 34455 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ15_RT_OFFSET 34456 +#define PBF_REG_BTB_GUARANTEED_VOQ15_RT_OFFSET 34457 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ15_RT_OFFSET 34458 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ16_RT_OFFSET 34459 +#define PBF_REG_BTB_GUARANTEED_VOQ16_RT_OFFSET 34460 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ16_RT_OFFSET 34461 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ17_RT_OFFSET 34462 +#define PBF_REG_BTB_GUARANTEED_VOQ17_RT_OFFSET 34463 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ17_RT_OFFSET 34464 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ18_RT_OFFSET 34465 +#define PBF_REG_BTB_GUARANTEED_VOQ18_RT_OFFSET 34466 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ18_RT_OFFSET 34467 +#define PBF_REG_YCMD_QS_NUM_LINES_VOQ19_RT_OFFSET 34468 +#define PBF_REG_BTB_GUARANTEED_VOQ19_RT_OFFSET 34469 +#define PBF_REG_BTB_SHARED_AREA_SETUP_VOQ19_RT_OFFSET 34470 +#define XCM_REG_CON_PHY_Q3_RT_OFFSET 34471 + +#define RUNTIME_ARRAY_SIZE 34472 /* Init Callbacks */ #define DMAE_READY_CB 0 @@ -5648,9 +5450,9 @@ struct e4_eth_conn_context { struct pstorm_eth_conn_st_ctx pstorm_st_context; struct xstorm_eth_conn_st_ctx xstorm_st_context; struct e4_xstorm_eth_conn_ag_ctx xstorm_ag_context; + struct e4_tstorm_eth_conn_ag_ctx tstorm_ag_context; struct ystorm_eth_conn_st_ctx ystorm_st_context; struct e4_ystorm_eth_conn_ag_ctx ystorm_ag_context; - struct e4_tstorm_eth_conn_ag_ctx tstorm_ag_context; struct e4_ustorm_eth_conn_ag_ctx ustorm_ag_context; struct ustorm_eth_conn_st_ctx ustorm_st_context; struct mstorm_eth_conn_st_ctx mstorm_st_context; @@ -5680,6 +5482,16 @@ enum eth_error_code { ETH_FILTERS_VNI_ADD_FAIL_FULL, ETH_FILTERS_VNI_ADD_FAIL_DUP, ETH_FILTERS_GFT_UPDATE_FAIL, + ETH_RX_QUEUE_FAIL_LOAD_VF_DATA, + ETH_FILTERS_GFS_ADD_FILTER_FAIL_MAX_HOPS, + ETH_FILTERS_GFS_ADD_FILTER_FAIL_NO_FREE_ENRTY, + ETH_FILTERS_GFS_ADD_FILTER_FAIL_ALREADY_EXISTS, + ETH_FILTERS_GFS_ADD_FILTER_FAIL_PCI_ERROR, + ETH_FILTERS_GFS_ADD_FINLER_FAIL_MAGIC_NUM_ERROR, + ETH_FILTERS_GFS_DEL_FILTER_FAIL_MAX_HOPS, + ETH_FILTERS_GFS_DEL_FILTER_FAIL_NO_MATCH_ENRTY, + ETH_FILTERS_GFS_DEL_FILTER_FAIL_PCI_ERROR, + ETH_FILTERS_GFS_DEL_FILTER_FAIL_MAGIC_NUM_ERROR, MAX_ETH_ERROR_CODE }; @@ -5703,6 +5515,11 @@ enum eth_event_opcode { ETH_EVENT_RX_CREATE_GFT_ACTION, ETH_EVENT_RX_GFT_UPDATE_FILTER, ETH_EVENT_TX_QUEUE_UPDATE, + ETH_EVENT_RGFS_ADD_FILTER, + ETH_EVENT_RGFS_DEL_FILTER, + ETH_EVENT_TGFS_ADD_FILTER, + ETH_EVENT_TGFS_DEL_FILTER, + ETH_EVENT_GFS_COUNTERS_REPORT_REQUEST, MAX_ETH_EVENT_OPCODE }; @@ -5795,18 +5612,31 @@ enum eth_ramrod_cmd_id { ETH_RAMROD_RX_CREATE_GFT_ACTION, ETH_RAMROD_GFT_UPDATE_FILTER, ETH_RAMROD_TX_QUEUE_UPDATE, + ETH_RAMROD_RGFS_FILTER_ADD, + ETH_RAMROD_RGFS_FILTER_DEL, + ETH_RAMROD_TGFS_FILTER_ADD, + ETH_RAMROD_TGFS_FILTER_DEL, + ETH_RAMROD_GFS_COUNTERS_REPORT_REQUEST, MAX_ETH_RAMROD_CMD_ID }; /* Return code from eth sp ramrods */ struct eth_return_code { u8 value; -#define ETH_RETURN_CODE_ERR_CODE_MASK 0x1F -#define ETH_RETURN_CODE_ERR_CODE_SHIFT 0 -#define ETH_RETURN_CODE_RESERVED_MASK 0x3 -#define ETH_RETURN_CODE_RESERVED_SHIFT 5 -#define ETH_RETURN_CODE_RX_TX_MASK 0x1 -#define ETH_RETURN_CODE_RX_TX_SHIFT 7 +#define ETH_RETURN_CODE_ERR_CODE_MASK 0x3F +#define ETH_RETURN_CODE_ERR_CODE_SHIFT 0 +#define ETH_RETURN_CODE_RESERVED_MASK 0x1 +#define ETH_RETURN_CODE_RESERVED_SHIFT 6 +#define ETH_RETURN_CODE_RX_TX_MASK 0x1 +#define ETH_RETURN_CODE_RX_TX_SHIFT 7 +}; + +/* tx destination enum */ +enum eth_tx_dst_mode_config_enum { + ETH_TX_DST_MODE_CONFIG_DISABLE, + ETH_TX_DST_MODE_CONFIG_FORWARD_DATA_IN_BD, + ETH_TX_DST_MODE_CONFIG_FORWARD_DATA_IN_VPORT, + MAX_ETH_TX_DST_MODE_CONFIG_ENUM }; /* What to do in case an error occurs */ @@ -5833,8 +5663,10 @@ struct eth_tx_err_vals { #define ETH_TX_ERR_VALS_MTU_VIOLATION_SHIFT 5 #define ETH_TX_ERR_VALS_ILLEGAL_CONTROL_FRAME_MASK 0x1 #define ETH_TX_ERR_VALS_ILLEGAL_CONTROL_FRAME_SHIFT 6 -#define ETH_TX_ERR_VALS_RESERVED_MASK 0x1FF -#define ETH_TX_ERR_VALS_RESERVED_SHIFT 7 +#define ETH_TX_ERR_VALS_ILLEGAL_BD_FLAGS_MASK 0x1 +#define ETH_TX_ERR_VALS_ILLEGAL_BD_FLAGS_SHIFT 7 +#define ETH_TX_ERR_VALS_RESERVED_MASK 0xFF +#define ETH_TX_ERR_VALS_RESERVED_SHIFT 8 }; /* vport rss configuration data */ @@ -5864,7 +5696,6 @@ struct eth_vport_rss_config { u8 tbl_size; __le32 reserved2[2]; __le16 indirection_table[ETH_RSS_IND_TABLE_ENTRIES_NUM]; - __le32 rss_key[ETH_RSS_KEY_SIZE_REGS]; __le32 reserved3[2]; }; @@ -6066,7 +5897,7 @@ struct rx_update_gft_filter_data { u8 inner_vlan_removal_en; }; -/* Ramrod data for rx queue start ramrod */ +/* Ramrod data for tx queue start ramrod */ struct tx_queue_start_ramrod_data { __le16 sb_id; u8 sb_index; @@ -6079,16 +5910,14 @@ struct tx_queue_start_ramrod_data { #define TX_QUEUE_START_RAMROD_DATA_DISABLE_OPPORTUNISTIC_SHIFT 0 #define TX_QUEUE_START_RAMROD_DATA_TEST_MODE_PKT_DUP_MASK 0x1 #define TX_QUEUE_START_RAMROD_DATA_TEST_MODE_PKT_DUP_SHIFT 1 -#define TX_QUEUE_START_RAMROD_DATA_TEST_MODE_TX_DEST_MASK 0x1 -#define TX_QUEUE_START_RAMROD_DATA_TEST_MODE_TX_DEST_SHIFT 2 #define TX_QUEUE_START_RAMROD_DATA_PMD_MODE_MASK 0x1 -#define TX_QUEUE_START_RAMROD_DATA_PMD_MODE_SHIFT 3 +#define TX_QUEUE_START_RAMROD_DATA_PMD_MODE_SHIFT 2 #define TX_QUEUE_START_RAMROD_DATA_NOTIFY_EN_MASK 0x1 -#define TX_QUEUE_START_RAMROD_DATA_NOTIFY_EN_SHIFT 4 +#define TX_QUEUE_START_RAMROD_DATA_NOTIFY_EN_SHIFT 3 #define TX_QUEUE_START_RAMROD_DATA_PIN_CONTEXT_MASK 0x1 -#define TX_QUEUE_START_RAMROD_DATA_PIN_CONTEXT_SHIFT 5 -#define TX_QUEUE_START_RAMROD_DATA_RESERVED1_MASK 0x3 -#define TX_QUEUE_START_RAMROD_DATA_RESERVED1_SHIFT 6 +#define TX_QUEUE_START_RAMROD_DATA_PIN_CONTEXT_SHIFT 4 +#define TX_QUEUE_START_RAMROD_DATA_RESERVED1_MASK 0x7 +#define TX_QUEUE_START_RAMROD_DATA_RESERVED1_SHIFT 5 u8 pxp_st_hint; u8 pxp_tph_valid_bd; u8 pxp_tph_valid_pkt; @@ -6144,18 +5973,22 @@ struct vport_start_ramrod_data { __le16 default_vlan; u8 tx_switching_en; u8 anti_spoofing_en; - u8 default_vlan_en; - u8 handle_ptp_pkts; u8 silent_vlan_removal_en; u8 untagged; struct eth_tx_err_vals tx_err_behav; - u8 zero_placement_offset; u8 ctl_frame_mac_check_en; u8 ctl_frame_ethtype_check_en; + u8 reserved0; + u8 reserved1; + u8 tx_dst_port_mode_config; + u8 dst_vport_id; + u8 tx_dst_port_mode; + u8 dst_vport_id_valid; u8 wipe_inner_vlan_pri_en; + u8 reserved2[2]; struct eth_in_to_in_pri_map_cfg in_to_in_vlan_pri_map_cfg; }; @@ -6715,19 +6548,6 @@ struct e4_xstorm_eth_hw_conn_ag_ctx { __le16 conn_dpi; }; -/* GFT CAM line struct */ -struct gft_cam_line { - __le32 camline; -#define GFT_CAM_LINE_VALID_MASK 0x1 -#define GFT_CAM_LINE_VALID_SHIFT 0 -#define GFT_CAM_LINE_DATA_MASK 0x3FFF -#define GFT_CAM_LINE_DATA_SHIFT 1 -#define GFT_CAM_LINE_MASK_BITS_MASK 0x3FFF -#define GFT_CAM_LINE_MASK_BITS_SHIFT 15 -#define GFT_CAM_LINE_RESERVED1_MASK 0x7 -#define GFT_CAM_LINE_RESERVED1_SHIFT 29 -}; - /* GFT CAM line struct with fields breakout */ struct gft_cam_line_mapped { __le32 camline; @@ -6757,10 +6577,6 @@ struct gft_cam_line_mapped { #define GFT_CAM_LINE_MAPPED_RESERVED1_SHIFT 29 }; -union gft_cam_line_union { - struct gft_cam_line cam_line; - struct gft_cam_line_mapped cam_line_mapped; -}; /* Used in gft_profile_key: Indication for ip version */ enum gft_profile_ip_version { @@ -7039,6 +6855,11 @@ struct mstorm_rdma_task_st_ctx { struct regpair temp[4]; }; +/* The roce task context of Ustorm */ +struct ustorm_rdma_task_st_ctx { + struct regpair temp[6]; +}; + struct e4_ustorm_rdma_task_ag_ctx { u8 reserved; u8 state; @@ -7048,8 +6869,8 @@ struct e4_ustorm_rdma_task_ag_ctx { #define E4_USTORM_RDMA_TASK_AG_CTX_CONNECTION_TYPE_SHIFT 0 #define E4_USTORM_RDMA_TASK_AG_CTX_EXIST_IN_QM0_MASK 0x1 #define E4_USTORM_RDMA_TASK_AG_CTX_EXIST_IN_QM0_SHIFT 4 -#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_RUNT_VALID_MASK 0x1 -#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_RUNT_VALID_SHIFT 5 +#define E4_USTORM_RDMA_TASK_AG_CTX_BIT1_MASK 0x1 +#define E4_USTORM_RDMA_TASK_AG_CTX_BIT1_SHIFT 5 #define E4_USTORM_RDMA_TASK_AG_CTX_DIF_WRITE_RESULT_CF_MASK 0x3 #define E4_USTORM_RDMA_TASK_AG_CTX_DIF_WRITE_RESULT_CF_SHIFT 6 u8 flags1; @@ -7079,29 +6900,29 @@ struct e4_ustorm_rdma_task_ag_ctx { #define E4_USTORM_RDMA_TASK_AG_CTX_RULE2EN_MASK 0x1 #define E4_USTORM_RDMA_TASK_AG_CTX_RULE2EN_SHIFT 7 u8 flags3; -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE3EN_MASK 0x1 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE3EN_SHIFT 0 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE4EN_MASK 0x1 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE4EN_SHIFT 1 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE5EN_MASK 0x1 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE5EN_SHIFT 2 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE6EN_MASK 0x1 -#define E4_USTORM_RDMA_TASK_AG_CTX_RULE6EN_SHIFT 3 -#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_ERROR_TYPE_MASK 0xF -#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_ERROR_TYPE_SHIFT 4 +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_RXMIT_PROD_CONS_EN_MASK 0x1 +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_RXMIT_PROD_CONS_EN_SHIFT 0 +#define E4_USTORM_RDMA_TASK_AG_CTX_RULE4EN_MASK 0x1 +#define E4_USTORM_RDMA_TASK_AG_CTX_RULE4EN_SHIFT 1 +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_WRITE_PROD_CONS_EN_MASK 0x1 +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_WRITE_PROD_CONS_EN_SHIFT 2 +#define E4_USTORM_RDMA_TASK_AG_CTX_RULE6EN_MASK 0x1 +#define E4_USTORM_RDMA_TASK_AG_CTX_RULE6EN_SHIFT 3 +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_ERROR_TYPE_MASK 0xF +#define E4_USTORM_RDMA_TASK_AG_CTX_DIF_ERROR_TYPE_SHIFT 4 __le32 dif_err_intervals; __le32 dif_error_1st_interval; - __le32 sq_cons; - __le32 dif_runt_value; + __le32 dif_rxmit_cons; + __le32 dif_rxmit_prod; __le32 sge_index; - __le32 reg5; + __le32 sq_cons; u8 byte2; u8 byte3; - __le16 word1; - __le16 word2; + __le16 dif_write_cons; + __le16 dif_write_prod; __le16 word3; - __le32 reg6; - __le32 reg7; + __le32 dif_error_buffer_address_lo; + __le32 dif_error_buffer_address_hi; }; /* RDMA task context */ @@ -7112,6 +6933,8 @@ struct e4_rdma_task_context { struct e4_mstorm_rdma_task_ag_ctx mstorm_ag_context; struct mstorm_rdma_task_st_ctx mstorm_st_context; struct rdif_task_context rdif_context; + struct ustorm_rdma_task_st_ctx ustorm_st_context; + struct regpair ustorm_st_padding[2]; struct e4_ustorm_rdma_task_ag_ctx ustorm_ag_context; }; @@ -7147,7 +6970,12 @@ struct rdma_create_cq_ramrod_data { u8 pbl_log_page_size; u8 toggle_bit; __le16 int_timeout; - __le16 reserved1; + u8 vf_id; + u8 flags; +#define RDMA_CREATE_CQ_RAMROD_DATA_VF_ID_VALID_MASK 0x1 +#define RDMA_CREATE_CQ_RAMROD_DATA_VF_ID_VALID_SHIFT 0 +#define RDMA_CREATE_CQ_RAMROD_DATA_RESERVED1_MASK 0x7F +#define RDMA_CREATE_CQ_RAMROD_DATA_RESERVED1_SHIFT 1 }; /* rdma deregister tid ramrod data */ @@ -7191,6 +7019,7 @@ enum rdma_fw_return_code { RDMA_RETURN_DEREGISTER_MR_BAD_STATE_ERR, RDMA_RETURN_RESIZE_CQ_ERR, RDMA_RETURN_NIG_DRAIN_REQ, + RDMA_RETURN_GENERAL_ERR, MAX_RDMA_FW_RETURN_CODE }; @@ -7204,7 +7033,10 @@ struct rdma_init_func_hdr { u8 relaxed_ordering; __le16 first_reg_srq_id; __le32 reg_srq_base_addr; - __le32 reserved; + u8 searcher_mode; + u8 pvrdma_mode; + u8 max_num_ns_log; + u8 reserved; }; /* rdma function init ramrod data */ @@ -7294,16 +7126,20 @@ struct rdma_resize_cq_ramrod_data { #define RDMA_RESIZE_CQ_RAMROD_DATA_TOGGLE_BIT_SHIFT 0 #define RDMA_RESIZE_CQ_RAMROD_DATA_IS_TWO_LEVEL_PBL_MASK 0x1 #define RDMA_RESIZE_CQ_RAMROD_DATA_IS_TWO_LEVEL_PBL_SHIFT 1 -#define RDMA_RESIZE_CQ_RAMROD_DATA_RESERVED_MASK 0x3F -#define RDMA_RESIZE_CQ_RAMROD_DATA_RESERVED_SHIFT 2 +#define RDMA_RESIZE_CQ_RAMROD_DATA_VF_ID_VALID_MASK 0x1 +#define RDMA_RESIZE_CQ_RAMROD_DATA_VF_ID_VALID_SHIFT 2 +#define RDMA_RESIZE_CQ_RAMROD_DATA_RESERVED_MASK 0x1F +#define RDMA_RESIZE_CQ_RAMROD_DATA_RESERVED_SHIFT 3 u8 pbl_log_page_size; __le16 pbl_num_pages; __le32 max_cqes; struct regpair pbl_addr; struct regpair output_params_addr; + u8 vf_id; + u8 reserved1[7]; }; -/* The rdma storm context of Mstorm */ +/* The rdma SRQ context */ struct rdma_srq_context { struct regpair temp[8]; }; @@ -7350,6 +7186,7 @@ enum rdma_tid_type { MAX_RDMA_TID_TYPE }; +/* The rdma XRC SRQ context */ struct rdma_xrc_srq_context { struct regpair temp[9]; }; @@ -7531,12 +7368,12 @@ struct e4_xstorm_roce_conn_ag_ctx { #define E4_XSTORM_ROCE_CONN_AG_CTX_BIT10_SHIFT 2 #define E4_XSTORM_ROCE_CONN_AG_CTX_BIT11_MASK 0x1 #define E4_XSTORM_ROCE_CONN_AG_CTX_BIT11_SHIFT 3 -#define E4_XSTORM_ROCE_CONN_AG_CTX_BIT12_MASK 0x1 -#define E4_XSTORM_ROCE_CONN_AG_CTX_BIT12_SHIFT 4 +#define E4_XSTORM_ROCE_CONN_AG_CTX_MSDM_FLUSH_MASK 0x1 +#define E4_XSTORM_ROCE_CONN_AG_CTX_MSDM_FLUSH_SHIFT 4 #define E4_XSTORM_ROCE_CONN_AG_CTX_MSEM_FLUSH_MASK 0x1 #define E4_XSTORM_ROCE_CONN_AG_CTX_MSEM_FLUSH_SHIFT 5 -#define E4_XSTORM_ROCE_CONN_AG_CTX_MSDM_FLUSH_MASK 0x1 -#define E4_XSTORM_ROCE_CONN_AG_CTX_MSDM_FLUSH_SHIFT 6 +#define E4_XSTORM_ROCE_CONN_AG_CTX_BIT14_MASK 0x1 +#define E4_XSTORM_ROCE_CONN_AG_CTX_BIT14_SHIFT 6 #define E4_XSTORM_ROCE_CONN_AG_CTX_YSTORM_FLUSH_MASK 0x1 #define E4_XSTORM_ROCE_CONN_AG_CTX_YSTORM_FLUSH_SHIFT 7 u8 flags2; @@ -7860,9 +7697,9 @@ struct mstorm_roce_conn_st_ctx { struct regpair temp[6]; }; -/* The roce storm context of Ystorm */ +/* The roce storm context of Ustorm */ struct ustorm_roce_conn_st_ctx { - struct regpair temp[12]; + struct regpair temp[14]; }; /* roce connection context */ @@ -7880,6 +7717,7 @@ struct e4_roce_conn_context { struct mstorm_roce_conn_st_ctx mstorm_st_context; struct regpair mstorm_st_padding[2]; struct ustorm_roce_conn_st_ctx ustorm_st_context; + struct regpair ustorm_st_padding[2]; }; /* roce cqes statistics */ @@ -7934,12 +7772,17 @@ struct roce_create_qp_req_ramrod_data { struct regpair qp_handle_for_cqe; struct regpair qp_handle_for_async; u8 stats_counter_id; - u8 reserved3[6]; + u8 vf_id; + u8 vport_id; u8 flags2; #define ROCE_CREATE_QP_REQ_RAMROD_DATA_EDPM_MODE_MASK 0x1 #define ROCE_CREATE_QP_REQ_RAMROD_DATA_EDPM_MODE_SHIFT 0 -#define ROCE_CREATE_QP_REQ_RAMROD_DATA_RESERVED_MASK 0x7F -#define ROCE_CREATE_QP_REQ_RAMROD_DATA_RESERVED_SHIFT 1 +#define ROCE_CREATE_QP_REQ_RAMROD_DATA_VF_ID_VALID_MASK 0x1 +#define ROCE_CREATE_QP_REQ_RAMROD_DATA_VF_ID_VALID_SHIFT 1 +#define ROCE_CREATE_QP_REQ_RAMROD_DATA_RESERVED_MASK 0x3F +#define ROCE_CREATE_QP_REQ_RAMROD_DATA_RESERVED_SHIFT 2 + u8 name_space; + u8 reserved3[3]; __le16 regular_latency_phy_queue; __le16 dpi; }; @@ -7967,8 +7810,10 @@ struct roce_create_qp_resp_ramrod_data { #define ROCE_CREATE_QP_RESP_RAMROD_DATA_MIN_RNR_NAK_TIMER_SHIFT 11 #define ROCE_CREATE_QP_RESP_RAMROD_DATA_XRC_FLAG_MASK 0x1 #define ROCE_CREATE_QP_RESP_RAMROD_DATA_XRC_FLAG_SHIFT 16 -#define ROCE_CREATE_QP_RESP_RAMROD_DATA_RESERVED_MASK 0x7FFF -#define ROCE_CREATE_QP_RESP_RAMROD_DATA_RESERVED_SHIFT 17 +#define ROCE_CREATE_QP_RESP_RAMROD_DATA_VF_ID_VALID_MASK 0x1 +#define ROCE_CREATE_QP_RESP_RAMROD_DATA_VF_ID_VALID_SHIFT 17 +#define ROCE_CREATE_QP_RESP_RAMROD_DATA_RESERVED_MASK 0x3FFF +#define ROCE_CREATE_QP_RESP_RAMROD_DATA_RESERVED_SHIFT 18 __le16 xrc_domain; u8 max_ird; u8 traffic_class; @@ -7995,10 +7840,14 @@ struct roce_create_qp_resp_ramrod_data { struct regpair qp_handle_for_cqe; struct regpair qp_handle_for_async; __le16 low_latency_phy_queue; - u8 reserved2[2]; + u8 vf_id; + u8 vport_id; __le32 cq_cid; __le16 regular_latency_phy_queue; __le16 dpi; + __le32 src_qp_id; + u8 name_space; + u8 reserved3[3]; }; /* roce DCQCN received statistics */ @@ -8032,6 +7881,8 @@ struct roce_destroy_qp_resp_output_params { /* RoCE destroy qp responder ramrod data */ struct roce_destroy_qp_resp_ramrod_data { struct regpair output_params_addr; + __le32 src_qp_id; + __le32 reserved; }; /* roce error statistics */ @@ -8115,8 +7966,8 @@ struct roce_modify_qp_req_ramrod_data { #define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PRI_FLG_SHIFT 9 #define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PRI_MASK 0x7 #define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PRI_SHIFT 10 -#define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PHYSICAL_QUEUES_FLG_MASK 0x1 -#define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PHYSICAL_QUEUES_FLG_SHIFT 13 +#define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PHYSICAL_QUEUE_FLG_MASK 0x1 +#define ROCE_MODIFY_QP_REQ_RAMROD_DATA_PHYSICAL_QUEUE_FLG_SHIFT 13 #define ROCE_MODIFY_QP_REQ_RAMROD_DATA_RESERVED1_MASK 0x3 #define ROCE_MODIFY_QP_REQ_RAMROD_DATA_RESERVED1_SHIFT 14 u8 fields; @@ -8162,8 +8013,8 @@ struct roce_modify_qp_resp_ramrod_data { #define ROCE_MODIFY_QP_RESP_RAMROD_DATA_MIN_RNR_NAK_TIMER_FLG_SHIFT 8 #define ROCE_MODIFY_QP_RESP_RAMROD_DATA_RDMA_OPS_EN_FLG_MASK 0x1 #define ROCE_MODIFY_QP_RESP_RAMROD_DATA_RDMA_OPS_EN_FLG_SHIFT 9 -#define ROCE_MODIFY_QP_RESP_RAMROD_DATA_PHYSICAL_QUEUES_FLG_MASK 0x1 -#define ROCE_MODIFY_QP_RESP_RAMROD_DATA_PHYSICAL_QUEUES_FLG_SHIFT 10 +#define ROCE_MODIFY_QP_RESP_RAMROD_DATA_PHYSICAL_QUEUE_FLG_MASK 0x1 +#define ROCE_MODIFY_QP_RESP_RAMROD_DATA_PHYSICAL_QUEUE_FLG_SHIFT 10 #define ROCE_MODIFY_QP_RESP_RAMROD_DATA_RESERVED1_MASK 0x1F #define ROCE_MODIFY_QP_RESP_RAMROD_DATA_RESERVED1_SHIFT 11 u8 fields; @@ -8204,7 +8055,7 @@ struct roce_query_qp_req_ramrod_data { /* RoCE query qp responder output params */ struct roce_query_qp_resp_output_params { __le32 psn; - __le32 err_flag; + __le32 flags; #define ROCE_QUERY_QP_RESP_OUTPUT_PARAMS_ERROR_FLG_MASK 0x1 #define ROCE_QUERY_QP_RESP_OUTPUT_PARAMS_ERROR_FLG_SHIFT 0 #define ROCE_QUERY_QP_RESP_OUTPUT_PARAMS_RESERVED0_MASK 0x7FFFFFFF @@ -8271,12 +8122,12 @@ struct e4_xstorm_roce_conn_ag_ctx_dq_ext_ld_part { #define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT10_SHIFT 2 #define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT11_MASK 0x1 #define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT11_SHIFT 3 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT12_MASK 0x1 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT12_SHIFT 4 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSEM_FLUSH_MASK 0x1 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSEM_FLUSH_SHIFT 5 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSDM_FLUSH_MASK 0x1 -#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSDM_FLUSH_SHIFT 6 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSDM_FLUSH_MASK 0x1 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSDM_FLUSH_SHIFT 4 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSEM_FLUSH_MASK 0x1 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_MSEM_FLUSH_SHIFT 5 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT14_MASK 0x1 +#define E4XSTORMROCECONNAGCTXDQEXTLDPART_BIT14_SHIFT 6 #define E4XSTORMROCECONNAGCTXDQEXTLDPART_YSTORM_FLUSH_MASK 0x1 #define E4XSTORMROCECONNAGCTXDQEXTLDPART_YSTORM_FLUSH_SHIFT 7 u8 flags2; @@ -8649,8 +8500,8 @@ struct e4_tstorm_roce_req_conn_ag_ctx { u8 flags5; #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE1EN_MASK 0x1 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE1EN_SHIFT 0 -#define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE2EN_MASK 0x1 -#define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE2EN_SHIFT 1 +#define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_DIF_CNT_EN_MASK 0x1 +#define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_DIF_CNT_EN_SHIFT 1 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE3EN_MASK 0x1 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE3EN_SHIFT 2 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE4EN_MASK 0x1 @@ -8663,13 +8514,13 @@ struct e4_tstorm_roce_req_conn_ag_ctx { #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE7EN_SHIFT 6 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE8EN_MASK 0x1 #define E4_TSTORM_ROCE_REQ_CONN_AG_CTX_RULE8EN_SHIFT 7 - __le32 reg0; + __le32 dif_rxmit_cnt; __le32 snd_nxt_psn; __le32 snd_max_psn; __le32 orq_prod; __le32 reg4; - __le32 reg5; - __le32 reg6; + __le32 dif_acked_cnt; + __le32 dif_cnt; __le32 reg7; __le32 reg8; u8 tx_cqe_error_type; @@ -8680,7 +8531,7 @@ struct e4_tstorm_roce_req_conn_ag_ctx { __le16 snd_sq_cons; __le16 conn_dpi; __le16 force_comp_cons; - __le32 reg9; + __le32 dif_rxmit_acked_cnt; __le32 reg10; }; @@ -8955,10 +8806,10 @@ struct e4_xstorm_roce_req_conn_ag_ctx { #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT10_SHIFT 2 #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT11_MASK 0x1 #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT11_SHIFT 3 -#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT12_MASK 0x1 -#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT12_SHIFT 4 -#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT13_MASK 0x1 -#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_BIT13_SHIFT 5 +#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_MSDM_FLUSH_MASK 0x1 +#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_MSDM_FLUSH_SHIFT 4 +#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_MSEM_FLUSH_MASK 0x1 +#define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_MSEM_FLUSH_SHIFT 5 #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_ERROR_STATE_MASK 0x1 #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_ERROR_STATE_SHIFT 6 #define E4_XSTORM_ROCE_REQ_CONN_AG_CTX_YSTORM_FLUSH_MASK 0x1 @@ -9184,10 +9035,10 @@ struct e4_xstorm_roce_resp_conn_ag_ctx { #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT10_SHIFT 2 #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT11_MASK 0x1 #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT11_SHIFT 3 -#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT12_MASK 0x1 -#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT12_SHIFT 4 -#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT13_MASK 0x1 -#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_BIT13_SHIFT 5 +#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_MSDM_FLUSH_MASK 0x1 +#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_MSDM_FLUSH_SHIFT 4 +#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_MSEM_FLUSH_MASK 0x1 +#define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_MSEM_FLUSH_SHIFT 5 #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_ERROR_STATE_MASK 0x1 #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_ERROR_STATE_SHIFT 6 #define E4_XSTORM_ROCE_RESP_CONN_AG_CTX_YSTORM_FLUSH_MASK 0x1 @@ -9914,7 +9765,7 @@ struct mstorm_iwarp_conn_st_ctx { /* The iwarp storm context of Ustorm */ struct ustorm_iwarp_conn_st_ctx { - __le32 reserved[24]; + struct regpair reserved[14]; }; /* iwarp connection context */ @@ -9932,6 +9783,7 @@ struct e4_iwarp_conn_context { struct regpair tstorm_st_padding[2]; struct mstorm_iwarp_conn_st_ctx mstorm_st_context; struct ustorm_iwarp_conn_st_ctx ustorm_st_context; + struct regpair ustorm_st_padding[2]; }; /* iWARP create QP params passed by driver to FW in CreateQP Request Ramrod */ @@ -9984,7 +9836,8 @@ enum iwarp_eqe_async_opcode { struct iwarp_eqe_data_mpa_async_completion { __le16 ulp_data_len; - u8 reserved[6]; + u8 rtr_type_sent; + u8 reserved[5]; }; struct iwarp_eqe_data_tcp_async_completion { @@ -10009,7 +9862,7 @@ enum iwarp_eqe_sync_opcode { /* iWARP EQE completion status */ enum iwarp_fw_return_code { - IWARP_CONN_ERROR_TCP_CONNECT_INVALID_PACKET = 5, + IWARP_CONN_ERROR_TCP_CONNECT_INVALID_PACKET = 6, IWARP_CONN_ERROR_TCP_CONNECTION_RST, IWARP_CONN_ERROR_TCP_CONNECT_TIMEOUT, IWARP_CONN_ERROR_MPA_ERROR_REJECT, @@ -10178,8 +10031,8 @@ struct iwarp_rxmit_stats_drv { * offload ramrod. */ struct iwarp_tcp_offload_ramrod_data { - struct iwarp_offload_params iwarp; struct tcp_offload_params_opt2 tcp; + struct iwarp_offload_params iwarp; }; /* iWARP MPA negotiation types */ @@ -11471,8 +11324,8 @@ struct e4_tstorm_iscsi_conn_ag_ctx { u8 flags3; #define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_Q0_MASK 0x3 #define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_Q0_SHIFT 0 -#define E4_TSTORM_ISCSI_CONN_AG_CTX_CF10_MASK 0x3 -#define E4_TSTORM_ISCSI_CONN_AG_CTX_CF10_SHIFT 2 +#define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_OOO_ISLES_CF_MASK 0x3 +#define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_OOO_ISLES_CF_SHIFT 2 #define E4_TSTORM_ISCSI_CONN_AG_CTX_CF0EN_MASK 0x1 #define E4_TSTORM_ISCSI_CONN_AG_CTX_CF0EN_SHIFT 4 #define E4_TSTORM_ISCSI_CONN_AG_CTX_P2T_FLUSH_CF_EN_MASK 0x1 @@ -11494,8 +11347,8 @@ struct e4_tstorm_iscsi_conn_ag_ctx { #define E4_TSTORM_ISCSI_CONN_AG_CTX_CF8EN_SHIFT 4 #define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_Q0_EN_MASK 0x1 #define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_Q0_EN_SHIFT 5 -#define E4_TSTORM_ISCSI_CONN_AG_CTX_CF10EN_MASK 0x1 -#define E4_TSTORM_ISCSI_CONN_AG_CTX_CF10EN_SHIFT 6 +#define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_OOO_ISLES_CF_EN_MASK 0x1 +#define E4_TSTORM_ISCSI_CONN_AG_CTX_FLUSH_OOO_ISLES_CF_EN_SHIFT 6 #define E4_TSTORM_ISCSI_CONN_AG_CTX_RULE0EN_MASK 0x1 #define E4_TSTORM_ISCSI_CONN_AG_CTX_RULE0EN_SHIFT 7 u8 flags5; @@ -11727,7 +11580,7 @@ struct e4_ystorm_iscsi_conn_ag_ctx { /* The trace in the buffer */ #define MFW_TRACE_EVENTID_MASK 0x00ffff #define MFW_TRACE_PRM_SIZE_MASK 0x0f0000 -#define MFW_TRACE_PRM_SIZE_SHIFT 16 +#define MFW_TRACE_PRM_SIZE_OFFSET 16 #define MFW_TRACE_ENTRY_SIZE 3 struct mcp_trace { @@ -12485,6 +12338,11 @@ enum resource_id_enum { RESOURCE_LL2_QUEUE_E = 15, RESOURCE_RDMA_STATS_QUEUE_E = 16, RESOURCE_BDQ_E = 17, + RESOURCE_QCN_E = 18, + RESOURCE_LLH_FILTER_E = 19, + RESOURCE_VF_MAC_ADDR = 20, + RESOURCE_LL2_CQS_E = 21, + RESOURCE_VF_CNQS = 22, RESOURCE_MAX_NUM, RESOURCE_NUM_INVALID = 0xFFFFFFFF }; @@ -12675,7 +12533,10 @@ struct public_drv_mb { #define DRV_MB_PARAM_DCBX_NOTIFY_SHIFT 3 #define DRV_MB_PARAM_NVM_PUT_FILE_BEGIN_MBI 0x3 +#define DRV_MB_PARAM_NVM_OFFSET_OFFSET 0 +#define DRV_MB_PARAM_NVM_OFFSET_MASK 0x00FFFFFF #define DRV_MB_PARAM_NVM_LEN_OFFSET 24 +#define DRV_MB_PARAM_NVM_LEN_MASK 0xFF000000 #define DRV_MB_PARAM_CFG_VF_MSIX_VF_ID_SHIFT 0 #define DRV_MB_PARAM_CFG_VF_MSIX_VF_ID_MASK 0x000000FF @@ -13436,6 +13297,21 @@ enum nvm_image_type { NVM_TYPE_FCOE_CFG = 0x1f, NVM_TYPE_ETH_PHY_FW1 = 0x20, NVM_TYPE_ETH_PHY_FW2 = 0x21, + NVM_TYPE_BDN = 0x22, + NVM_TYPE_8485X_PHY_FW = 0x23, + NVM_TYPE_PUB_KEY = 0x24, + NVM_TYPE_RECOVERY = 0x25, + NVM_TYPE_PLDM = 0x26, + NVM_TYPE_UPK1 = 0x27, + NVM_TYPE_UPK2 = 0x28, + NVM_TYPE_MASTER_KC = 0x29, + NVM_TYPE_BACKUP_KC = 0x2a, + NVM_TYPE_HW_DUMP = 0x2b, + NVM_TYPE_HW_DUMP_OUT = 0x2c, + NVM_TYPE_BIN_NVM_META = 0x30, + NVM_TYPE_ROM_TEST = 0xf0, + NVM_TYPE_88X33X0_PHY_FW = 0x31, + NVM_TYPE_88X33X0_PHY_SLAVE_FW = 0x32, NVM_TYPE_MAX, }; diff --git a/drivers/net/ethernet/qlogic/qed/qed_hw.c b/drivers/net/ethernet/qlogic/qed/qed_hw.c index a4de9e3ef72c..4ab8cfaf63d1 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_hw.c +++ b/drivers/net/ethernet/qlogic/qed/qed_hw.c @@ -393,7 +393,7 @@ u32 qed_vfid_to_concrete(struct qed_hwfn *p_hwfn, u8 vfid) /* DMAE */ #define QED_DMAE_FLAGS_IS_SET(params, flag) \ - ((params) != NULL && ((params)->flags & QED_DMAE_FLAG_##flag)) + ((params) != NULL && GET_FIELD((params)->flags, QED_DMAE_PARAMS_##flag)) static void qed_dmae_opcode(struct qed_hwfn *p_hwfn, const u8 is_src_type_grc, @@ -408,62 +408,55 @@ static void qed_dmae_opcode(struct qed_hwfn *p_hwfn, * 0- The source is the PCIe * 1- The source is the GRC. */ - opcode |= (is_src_type_grc ? DMAE_CMD_SRC_MASK_GRC - : DMAE_CMD_SRC_MASK_PCIE) << - DMAE_CMD_SRC_SHIFT; - src_pfid = QED_DMAE_FLAGS_IS_SET(p_params, PF_SRC) ? - p_params->src_pfid : p_hwfn->rel_pf_id; - opcode |= ((src_pfid & DMAE_CMD_SRC_PF_ID_MASK) << - DMAE_CMD_SRC_PF_ID_SHIFT); + SET_FIELD(opcode, DMAE_CMD_SRC, + (is_src_type_grc ? dmae_cmd_src_grc : dmae_cmd_src_pcie)); + src_pfid = QED_DMAE_FLAGS_IS_SET(p_params, SRC_PF_VALID) ? + p_params->src_pfid : p_hwfn->rel_pf_id; + SET_FIELD(opcode, DMAE_CMD_SRC_PF_ID, src_pfid); /* The destination of the DMA can be: 0-None 1-PCIe 2-GRC 3-None */ - opcode |= (is_dst_type_grc ? DMAE_CMD_DST_MASK_GRC - : DMAE_CMD_DST_MASK_PCIE) << - DMAE_CMD_DST_SHIFT; - dst_pfid = QED_DMAE_FLAGS_IS_SET(p_params, PF_DST) ? - p_params->dst_pfid : p_hwfn->rel_pf_id; - opcode |= ((dst_pfid & DMAE_CMD_DST_PF_ID_MASK) << - DMAE_CMD_DST_PF_ID_SHIFT); + SET_FIELD(opcode, DMAE_CMD_DST, + (is_dst_type_grc ? dmae_cmd_dst_grc : dmae_cmd_dst_pcie)); + dst_pfid = QED_DMAE_FLAGS_IS_SET(p_params, DST_PF_VALID) ? + p_params->dst_pfid : p_hwfn->rel_pf_id; + SET_FIELD(opcode, DMAE_CMD_DST_PF_ID, dst_pfid); + /* Whether to write a completion word to the completion destination: * 0-Do not write a completion word * 1-Write the completion word */ - opcode |= (DMAE_CMD_COMP_WORD_EN_MASK << DMAE_CMD_COMP_WORD_EN_SHIFT); - opcode |= (DMAE_CMD_SRC_ADDR_RESET_MASK << - DMAE_CMD_SRC_ADDR_RESET_SHIFT); + SET_FIELD(opcode, DMAE_CMD_COMP_WORD_EN, 1); + SET_FIELD(opcode, DMAE_CMD_SRC_ADDR_RESET, 1); if (QED_DMAE_FLAGS_IS_SET(p_params, COMPLETION_DST)) - opcode |= (1 << DMAE_CMD_COMP_FUNC_SHIFT); + SET_FIELD(opcode, DMAE_CMD_COMP_FUNC, 1); - opcode |= (DMAE_CMD_ENDIANITY << DMAE_CMD_ENDIANITY_MODE_SHIFT); + /* swapping mode 3 - big endian */ + SET_FIELD(opcode, DMAE_CMD_ENDIANITY_MODE, DMAE_CMD_ENDIANITY); - port_id = (QED_DMAE_FLAGS_IS_SET(p_params, PORT)) ? - p_params->port_id : p_hwfn->port_id; - opcode |= (port_id << DMAE_CMD_PORT_ID_SHIFT); + port_id = (QED_DMAE_FLAGS_IS_SET(p_params, PORT_VALID)) ? + p_params->port_id : p_hwfn->port_id; + SET_FIELD(opcode, DMAE_CMD_PORT_ID, port_id); /* reset source address in next go */ - opcode |= (DMAE_CMD_SRC_ADDR_RESET_MASK << - DMAE_CMD_SRC_ADDR_RESET_SHIFT); + SET_FIELD(opcode, DMAE_CMD_SRC_ADDR_RESET, 1); /* reset dest address in next go */ - opcode |= (DMAE_CMD_DST_ADDR_RESET_MASK << - DMAE_CMD_DST_ADDR_RESET_SHIFT); + SET_FIELD(opcode, DMAE_CMD_DST_ADDR_RESET, 1); /* SRC/DST VFID: all 1's - pf, otherwise VF id */ - if (QED_DMAE_FLAGS_IS_SET(p_params, VF_SRC)) { - opcode |= 1 << DMAE_CMD_SRC_VF_ID_VALID_SHIFT; - opcode_b |= p_params->src_vfid << DMAE_CMD_SRC_VF_ID_SHIFT; + if (QED_DMAE_FLAGS_IS_SET(p_params, SRC_VF_VALID)) { + SET_FIELD(opcode, DMAE_CMD_SRC_VF_ID_VALID, 1); + SET_FIELD(opcode_b, DMAE_CMD_SRC_VF_ID, p_params->src_vfid); } else { - opcode_b |= DMAE_CMD_SRC_VF_ID_MASK << - DMAE_CMD_SRC_VF_ID_SHIFT; + SET_FIELD(opcode_b, DMAE_CMD_SRC_VF_ID, 0xFF); } - - if (QED_DMAE_FLAGS_IS_SET(p_params, VF_DST)) { - opcode |= 1 << DMAE_CMD_DST_VF_ID_VALID_SHIFT; - opcode_b |= p_params->dst_vfid << DMAE_CMD_DST_VF_ID_SHIFT; + if (QED_DMAE_FLAGS_IS_SET(p_params, DST_VF_VALID)) { + SET_FIELD(opcode, DMAE_CMD_DST_VF_ID_VALID, 1); + SET_FIELD(opcode_b, DMAE_CMD_DST_VF_ID, p_params->dst_vfid); } else { - opcode_b |= DMAE_CMD_DST_VF_ID_MASK << DMAE_CMD_DST_VF_ID_SHIFT; + SET_FIELD(opcode_b, DMAE_CMD_DST_VF_ID, 0xFF); } p_hwfn->dmae_info.p_dmae_cmd->opcode = cpu_to_le32(opcode); diff --git a/drivers/net/ethernet/qlogic/qed/qed_init_fw_funcs.c b/drivers/net/ethernet/qlogic/qed/qed_init_fw_funcs.c index d6430dfebd83..2f1049b0b93a 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_init_fw_funcs.c +++ b/drivers/net/ethernet/qlogic/qed/qed_init_fw_funcs.c @@ -44,9 +44,9 @@ #define CDU_VALIDATION_DEFAULT_CFG 61 static u16 con_region_offsets[3][NUM_OF_CONNECTION_TYPES_E4] = { - {400, 336, 352, 304, 304, 384, 416, 352}, /* region 3 offsets */ - {528, 496, 416, 448, 448, 512, 544, 480}, /* region 4 offsets */ - {608, 544, 496, 512, 576, 592, 624, 560} /* region 5 offsets */ + {400, 336, 352, 368, 304, 384, 416, 352}, /* region 3 offsets */ + {528, 496, 416, 512, 448, 512, 544, 480}, /* region 4 offsets */ + {608, 544, 496, 576, 576, 592, 624, 560} /* region 5 offsets */ }; static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { @@ -61,6 +61,9 @@ static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { 0x100) - 1 : 0) #define QM_INVALID_PQ_ID 0xffff +/* Max link speed (in Mbps) */ +#define QM_MAX_LINK_SPEED 100000 + /* Feature enable */ #define QM_BYPASS_EN 1 #define QM_BYTE_CRD_EN 1 @@ -128,8 +131,6 @@ static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { /* Pure LB CmdQ lines (+spare) */ #define PBF_CMDQ_PURE_LB_LINES 150 -#define PBF_CMDQ_LINES_E5_RSVD_RATIO 8 - #define PBF_CMDQ_LINES_RT_OFFSET(ext_voq) \ (PBF_REG_YCMD_QS_NUM_LINES_VOQ0_RT_OFFSET + \ (ext_voq) * (PBF_REG_YCMD_QS_NUM_LINES_VOQ1_RT_OFFSET - \ @@ -140,6 +141,9 @@ static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { (ext_voq) * (PBF_REG_BTB_GUARANTEED_VOQ1_RT_OFFSET - \ PBF_REG_BTB_GUARANTEED_VOQ0_RT_OFFSET)) +/* Returns the VOQ line credit for the specified number of PBF command lines. + * PBF lines are specified in 256b units. + */ #define QM_VOQ_LINE_CRD(pbf_cmd_lines) \ ((((pbf_cmd_lines) - 4) * 2) | QM_LINE_CRD_REG_SIGN_BIT) @@ -178,14 +182,14 @@ static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { cmd ## _ ## field, \ value) -#define QM_INIT_TX_PQ_MAP(p_hwfn, map, chip, pq_id, rl_valid, vp_pq_id, rl_id, \ +#define QM_INIT_TX_PQ_MAP(p_hwfn, map, chip, pq_id, vp_pq_id, rl_valid, rl_id, \ ext_voq, wrr) \ do { \ typeof(map) __map; \ memset(&__map, 0, sizeof(__map)); \ SET_FIELD(__map.reg, QM_RF_PQ_MAP_ ## chip ## _PQ_VALID, 1); \ SET_FIELD(__map.reg, QM_RF_PQ_MAP_ ## chip ## _RL_VALID, \ - rl_valid); \ + rl_valid ? 1 : 0);\ SET_FIELD(__map.reg, QM_RF_PQ_MAP_ ## chip ## _VP_PQ_ID, \ vp_pq_id); \ SET_FIELD(__map.reg, QM_RF_PQ_MAP_ ## chip ## _RL_ID, rl_id); \ @@ -200,9 +204,12 @@ static u16 task_region_offsets[1][NUM_OF_CONNECTION_TYPES_E4] = { #define WRITE_PQ_INFO_TO_RAM 1 #define PQ_INFO_ELEMENT(vp, pf, tc, port, rl_valid, rl) \ (((vp) << 0) | ((pf) << 12) | ((tc) << 16) | ((port) << 20) | \ - ((rl_valid) << 22) | ((rl) << 24)) + ((rl_valid ? 1 : 0) << 22) | (((rl) & 255) << 24) | \ + (((rl) >> 8) << 9)) + #define PQ_INFO_RAM_GRC_ADDRESS(pq_id) \ - (XSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + 21776 + (pq_id) * 4) + XSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + \ + XSTORM_PQ_INFO_OFFSET(pq_id) /******************** INTERNAL IMPLEMENTATION *********************/ @@ -228,9 +235,6 @@ static void qed_enable_pf_rl(struct qed_hwfn *p_hwfn, bool pf_rl_en) STORE_RT_REG(p_hwfn, QM_REG_RLPFVOQENABLE_RT_OFFSET, (u32)voq_bit_mask); - if (num_ext_voqs >= 32) - STORE_RT_REG(p_hwfn, QM_REG_RLPFVOQENABLE_MSB_RT_OFFSET, - (u32)(voq_bit_mask >> 32)); /* Write RL period */ STORE_RT_REG(p_hwfn, @@ -259,12 +263,12 @@ static void qed_enable_pf_wfq(struct qed_hwfn *p_hwfn, bool pf_wfq_en) QM_WFQ_UPPER_BOUND); } -/* Prepare VPORT RL enable/disable runtime init values */ -static void qed_enable_vport_rl(struct qed_hwfn *p_hwfn, bool vport_rl_en) +/* Prepare global RL enable/disable runtime init values */ +static void qed_enable_global_rl(struct qed_hwfn *p_hwfn, bool global_rl_en) { STORE_RT_REG(p_hwfn, QM_REG_RLGLBLENABLE_RT_OFFSET, - vport_rl_en ? 1 : 0); - if (vport_rl_en) { + global_rl_en ? 1 : 0); + if (global_rl_en) { /* Write RL period (use timer 0 only) */ STORE_RT_REG(p_hwfn, QM_REG_RLGLBLPERIOD_0_RT_OFFSET, @@ -331,8 +335,7 @@ static void qed_cmdq_lines_rt_init( continue; /* Find number of command queue lines to divide between the - * active physical TCs. In E5, 1/8 of the lines are reserved. - * the lines for pure LB TC are subtracted. + * active physical TCs. */ phys_lines = port_params[port_id].num_pbf_cmd_lines; phys_lines -= PBF_CMDQ_PURE_LB_LINES; @@ -361,11 +364,30 @@ static void qed_cmdq_lines_rt_init( ext_voq = qed_get_ext_voq(p_hwfn, port_id, PURE_LB_TC, max_phys_tcs_per_port); - qed_cmdq_lines_voq_rt_init(p_hwfn, - ext_voq, PBF_CMDQ_PURE_LB_LINES); + qed_cmdq_lines_voq_rt_init(p_hwfn, ext_voq, + PBF_CMDQ_PURE_LB_LINES); } } +/* Prepare runtime init values to allocate guaranteed BTB blocks for the + * specified port. The guaranteed BTB space is divided between the TCs as + * follows (shared space Is currently not used): + * 1. Parameters: + * B - BTB blocks for this port + * C - Number of physical TCs for this port + * 2. Calculation: + * a. 38 blocks (9700B jumbo frame) are allocated for global per port + * headroom. + * b. B = B - 38 (remainder after global headroom allocation). + * c. MAX(38,B/(C+0.7)) blocks are allocated for the pure LB VOQ. + * d. B = B - MAX(38, B/(C+0.7)) (remainder after pure LB allocation). + * e. B/C blocks are allocated for each physical TC. + * Assumptions: + * - MTU is up to 9700 bytes (38 blocks) + * - All TCs are considered symmetrical (same rate and packet size) + * - No optimization for lossy TC (all are considered lossless). Shared space + * is not enabled and allocated for each TC. + */ static void qed_btb_blocks_rt_init( struct qed_hwfn *p_hwfn, u8 max_ports_per_engine, @@ -424,6 +446,34 @@ static void qed_btb_blocks_rt_init( } } +/* Prepare runtime init values for the specified RL. + * Set max link speed (100Gbps) per rate limiter. + * Return -1 on error. + */ +static int qed_global_rl_rt_init(struct qed_hwfn *p_hwfn) +{ + u32 upper_bound = QM_VP_RL_UPPER_BOUND(QM_MAX_LINK_SPEED) | + (u32)QM_RL_CRD_REG_SIGN_BIT; + u32 inc_val; + u16 rl_id; + + /* Go over all global RLs */ + for (rl_id = 0; rl_id < MAX_QM_GLOBAL_RLS; rl_id++) { + inc_val = QM_RL_INC_VAL(QM_MAX_LINK_SPEED); + + STORE_RT_REG(p_hwfn, + QM_REG_RLGLBLCRD_RT_OFFSET + rl_id, + (u32)QM_RL_CRD_REG_SIGN_BIT); + STORE_RT_REG(p_hwfn, + QM_REG_RLGLBLUPPERBOUND_RT_OFFSET + rl_id, + upper_bound); + STORE_RT_REG(p_hwfn, + QM_REG_RLGLBLINCVAL_RT_OFFSET + rl_id, inc_val); + } + + return 0; +} + /* Prepare Tx PQ mapping runtime init values for the specified PF */ static void qed_tx_pq_map_rt_init(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, @@ -460,18 +510,17 @@ static void qed_tx_pq_map_rt_init(struct qed_hwfn *p_hwfn, /* Go over all Tx PQs */ for (i = 0, pq_id = p_params->start_pq; i < num_pqs; i++, pq_id++) { - u8 ext_voq, vport_id_in_pf, tc_id = pq_params[i].tc_id; - u32 max_qm_global_rls = MAX_QM_GLOBAL_RLS; + u16 *p_first_tx_pq_id, vport_id_in_pf; struct qm_rf_pq_map_e4 tx_pq_map; - bool is_vf_pq, rl_valid; - u16 *p_first_tx_pq_id; + u8 tc_id = pq_params[i].tc_id; + bool is_vf_pq; + u8 ext_voq; ext_voq = qed_get_ext_voq(p_hwfn, pq_params[i].port_id, tc_id, p_params->max_phys_tcs_per_port); is_vf_pq = (i >= p_params->num_pf_pqs); - rl_valid = pq_params[i].rl_valid > 0; /* Update first Tx PQ of VPORT/TC */ vport_id_in_pf = pq_params[i].vport_id - p_params->start_vport; @@ -492,21 +541,14 @@ static void qed_tx_pq_map_rt_init(struct qed_hwfn *p_hwfn, map_val); } - /* Check RL ID */ - if (rl_valid && pq_params[i].vport_id >= max_qm_global_rls) { - DP_NOTICE(p_hwfn, - "Invalid VPORT ID for rate limiter configuration\n"); - rl_valid = false; - } - /* Prepare PQ map entry */ QM_INIT_TX_PQ_MAP(p_hwfn, tx_pq_map, E4, pq_id, - rl_valid ? 1 : 0, *p_first_tx_pq_id, - rl_valid ? pq_params[i].vport_id : 0, + pq_params[i].rl_valid, + pq_params[i].rl_id, ext_voq, pq_params[i].wrr_group); /* Set PQ base address */ @@ -529,9 +571,8 @@ static void qed_tx_pq_map_rt_init(struct qed_hwfn *p_hwfn, p_params->pf_id, tc_id, pq_params[i].port_id, - rl_valid ? 1 : 0, - rl_valid ? - pq_params[i].vport_id : 0); + pq_params[i].rl_valid, + pq_params[i].rl_id); qed_wr(p_hwfn, p_ptt, PQ_INFO_RAM_GRC_ADDRESS(pq_id), pq_info); } @@ -669,19 +710,19 @@ static int qed_pf_rl_rt_init(struct qed_hwfn *p_hwfn, u8 pf_id, u32 pf_rl) * Return -1 on error. */ static int qed_vp_wfq_rt_init(struct qed_hwfn *p_hwfn, - u8 num_vports, + u16 num_vports, struct init_qm_vport_params *vport_params) { - u16 vport_pq_id; + u16 vport_pq_id, i; u32 inc_val; - u8 tc, i; + u8 tc; /* Go over all PF VPORTs */ for (i = 0; i < num_vports; i++) { - if (!vport_params[i].vport_wfq) + if (!vport_params[i].wfq) continue; - inc_val = QM_WFQ_INC_VAL(vport_params[i].vport_wfq); + inc_val = QM_WFQ_INC_VAL(vport_params[i].wfq); if (inc_val > QM_WFQ_MAX_INC_VAL) { DP_NOTICE(p_hwfn, "Invalid VPORT WFQ weight configuration\n"); @@ -706,48 +747,6 @@ static int qed_vp_wfq_rt_init(struct qed_hwfn *p_hwfn, return 0; } -/* Prepare VPORT RL runtime init values for the specified VPORTs. - * Return -1 on error. - */ -static int qed_vport_rl_rt_init(struct qed_hwfn *p_hwfn, - u8 start_vport, - u8 num_vports, - u32 link_speed, - struct init_qm_vport_params *vport_params) -{ - u8 i, vport_id; - u32 inc_val; - - if (start_vport + num_vports >= MAX_QM_GLOBAL_RLS) { - DP_NOTICE(p_hwfn, - "Invalid VPORT ID for rate limiter configuration\n"); - return -1; - } - - /* Go over all PF VPORTs */ - for (i = 0, vport_id = start_vport; i < num_vports; i++, vport_id++) { - inc_val = QM_RL_INC_VAL(vport_params[i].vport_rl ? - vport_params[i].vport_rl : - link_speed); - if (inc_val > QM_VP_RL_MAX_INC_VAL(link_speed)) { - DP_NOTICE(p_hwfn, - "Invalid VPORT rate-limit configuration\n"); - return -1; - } - - STORE_RT_REG(p_hwfn, QM_REG_RLGLBLCRD_RT_OFFSET + vport_id, - (u32)QM_RL_CRD_REG_SIGN_BIT); - STORE_RT_REG(p_hwfn, - QM_REG_RLGLBLUPPERBOUND_RT_OFFSET + vport_id, - QM_VP_RL_UPPER_BOUND(link_speed) | - (u32)QM_RL_CRD_REG_SIGN_BIT); - STORE_RT_REG(p_hwfn, QM_REG_RLGLBLINCVAL_RT_OFFSET + vport_id, - inc_val); - } - - return 0; -} - static bool qed_poll_on_qm_cmd_ready(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt) { @@ -799,23 +798,20 @@ u32 qed_qm_pf_mem_size(u32 num_pf_cids, int qed_qm_common_rt_init(struct qed_hwfn *p_hwfn, struct qed_qm_common_rt_init_params *p_params) { - /* Init AFullOprtnstcCrdMask */ - u32 mask = (QM_OPPOR_LINE_VOQ_DEF << - QM_RF_OPPORTUNISTIC_MASK_LINEVOQ_SHIFT) | - (QM_BYTE_CRD_EN << QM_RF_OPPORTUNISTIC_MASK_BYTEVOQ_SHIFT) | - (p_params->pf_wfq_en << - QM_RF_OPPORTUNISTIC_MASK_PFWFQ_SHIFT) | - (p_params->vport_wfq_en << - QM_RF_OPPORTUNISTIC_MASK_VPWFQ_SHIFT) | - (p_params->pf_rl_en << - QM_RF_OPPORTUNISTIC_MASK_PFRL_SHIFT) | - (p_params->vport_rl_en << - QM_RF_OPPORTUNISTIC_MASK_VPQCNRL_SHIFT) | - (QM_OPPOR_FW_STOP_DEF << - QM_RF_OPPORTUNISTIC_MASK_FWPAUSE_SHIFT) | - (QM_OPPOR_PQ_EMPTY_DEF << - QM_RF_OPPORTUNISTIC_MASK_QUEUEEMPTY_SHIFT); + u32 mask = 0; + /* Init AFullOprtnstcCrdMask */ + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_LINEVOQ, + QM_OPPOR_LINE_VOQ_DEF); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_BYTEVOQ, QM_BYTE_CRD_EN); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_PFWFQ, p_params->pf_wfq_en); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_VPWFQ, p_params->vport_wfq_en); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_PFRL, p_params->pf_rl_en); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_VPQCNRL, + p_params->global_rl_en); + SET_FIELD(mask, QM_RF_OPPORTUNISTIC_MASK_FWPAUSE, QM_OPPOR_FW_STOP_DEF); + SET_FIELD(mask, + QM_RF_OPPORTUNISTIC_MASK_QUEUEEMPTY, QM_OPPOR_PQ_EMPTY_DEF); STORE_RT_REG(p_hwfn, QM_REG_AFULLOPRTNSTCCRDMASK_RT_OFFSET, mask); /* Enable/disable PF RL */ @@ -824,8 +820,8 @@ int qed_qm_common_rt_init(struct qed_hwfn *p_hwfn, /* Enable/disable PF WFQ */ qed_enable_pf_wfq(p_hwfn, p_params->pf_wfq_en); - /* Enable/disable VPORT RL */ - qed_enable_vport_rl(p_hwfn, p_params->vport_rl_en); + /* Enable/disable global RL */ + qed_enable_global_rl(p_hwfn, p_params->global_rl_en); /* Enable/disable VPORT WFQ */ qed_enable_vport_wfq(p_hwfn, p_params->vport_wfq_en); @@ -842,6 +838,8 @@ int qed_qm_common_rt_init(struct qed_hwfn *p_hwfn, p_params->max_phys_tcs_per_port, p_params->port_params); + qed_global_rl_rt_init(p_hwfn); + return 0; } @@ -853,7 +851,9 @@ int qed_qm_pf_rt_init(struct qed_hwfn *p_hwfn, u32 other_mem_size_4kb = QM_PQ_MEM_4KB(p_params->num_pf_cids + p_params->num_tids) * QM_OTHER_PQS_PER_PF; - u8 tc, i; + u16 i; + u8 tc; + /* Clear first Tx PQ ID array for each VPORT */ for (i = 0; i < p_params->num_vports; i++) @@ -878,16 +878,10 @@ int qed_qm_pf_rt_init(struct qed_hwfn *p_hwfn, if (qed_pf_rl_rt_init(p_hwfn, p_params->pf_id, p_params->pf_rl)) return -1; - /* Set VPORT WFQ */ + /* Init VPORT WFQ */ if (qed_vp_wfq_rt_init(p_hwfn, p_params->num_vports, vport_params)) return -1; - /* Set VPORT RL */ - if (qed_vport_rl_rt_init(p_hwfn, p_params->start_vport, - p_params->num_vports, p_params->link_speed, - vport_params)) - return -1; - return 0; } @@ -925,18 +919,19 @@ int qed_init_pf_rl(struct qed_hwfn *p_hwfn, int qed_init_vport_wfq(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, - u16 first_tx_pq_id[NUM_OF_TCS], u16 vport_wfq) + u16 first_tx_pq_id[NUM_OF_TCS], u16 wfq) { u16 vport_pq_id; u32 inc_val; u8 tc; - inc_val = QM_WFQ_INC_VAL(vport_wfq); + inc_val = QM_WFQ_INC_VAL(wfq); if (!inc_val || inc_val > QM_WFQ_MAX_INC_VAL) { - DP_NOTICE(p_hwfn, "Invalid VPORT WFQ weight configuration\n"); + DP_NOTICE(p_hwfn, "Invalid VPORT WFQ configuration.\n"); return -1; } + /* A VPORT can have several VPORT PQ IDs for various TCs */ for (tc = 0; tc < NUM_OF_TCS; tc++) { vport_pq_id = first_tx_pq_id[tc]; if (vport_pq_id != QM_INVALID_PQ_ID) @@ -948,28 +943,20 @@ int qed_init_vport_wfq(struct qed_hwfn *p_hwfn, return 0; } -int qed_init_vport_rl(struct qed_hwfn *p_hwfn, - struct qed_ptt *p_ptt, - u8 vport_id, u32 vport_rl, u32 link_speed) +int qed_init_global_rl(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, u16 rl_id, u32 rate_limit) { - u32 inc_val, max_qm_global_rls = MAX_QM_GLOBAL_RLS; + u32 inc_val; - if (vport_id >= max_qm_global_rls) { - DP_NOTICE(p_hwfn, - "Invalid VPORT ID for rate limiter configuration\n"); + inc_val = QM_RL_INC_VAL(rate_limit); + if (inc_val > QM_VP_RL_MAX_INC_VAL(rate_limit)) { + DP_NOTICE(p_hwfn, "Invalid rate limit configuration.\n"); return -1; } - inc_val = QM_RL_INC_VAL(vport_rl ? vport_rl : link_speed); - if (inc_val > QM_VP_RL_MAX_INC_VAL(link_speed)) { - DP_NOTICE(p_hwfn, "Invalid VPORT rate-limit configuration\n"); - return -1; - } - - qed_wr(p_hwfn, - p_ptt, - QM_REG_RLGLBLCRD + vport_id * 4, (u32)QM_RL_CRD_REG_SIGN_BIT); - qed_wr(p_hwfn, p_ptt, QM_REG_RLGLBLINCVAL + vport_id * 4, inc_val); + qed_wr(p_hwfn, p_ptt, + QM_REG_RLGLBLCRD + rl_id * 4, (u32)QM_RL_CRD_REG_SIGN_BIT); + qed_wr(p_hwfn, p_ptt, QM_REG_RLGLBLINCVAL + rl_id * 4, inc_val); return 0; } @@ -1013,7 +1000,6 @@ bool qed_send_qm_stop_cmd(struct qed_hwfn *p_hwfn, return true; } - #define SET_TUNNEL_TYPE_ENABLE_BIT(var, offset, enable) \ do { \ typeof(var) *__p_var = &(var); \ @@ -1021,8 +1007,59 @@ bool qed_send_qm_stop_cmd(struct qed_hwfn *p_hwfn, *__p_var = (*__p_var & ~BIT(__offset)) | \ ((enable) ? BIT(__offset) : 0); \ } while (0) -#define PRS_ETH_TUNN_OUTPUT_FORMAT -188897008 -#define PRS_ETH_OUTPUT_FORMAT -46832 + +#define PRS_ETH_TUNN_OUTPUT_FORMAT 0xF4DAB910 +#define PRS_ETH_OUTPUT_FORMAT 0xFFFF4910 + +#define ARR_REG_WR(dev, ptt, addr, arr, arr_size) \ + do { \ + u32 i; \ + \ + for (i = 0; i < (arr_size); i++) \ + qed_wr(dev, ptt, \ + ((addr) + (4 * i)), \ + ((u32 *)&(arr))[i]); \ + } while (0) + +/** + * @brief qed_dmae_to_grc - is an internal function - writes from host to + * wide-bus registers (split registers are not supported yet) + * + * @param p_hwfn - HW device data + * @param p_ptt - ptt window used for writing the registers. + * @param p_data - pointer to source data. + * @param addr - Destination register address. + * @param len_in_dwords - data length in DWARDS (u32) + */ +static int qed_dmae_to_grc(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + u32 *p_data, u32 addr, u32 len_in_dwords) +{ + struct qed_dmae_params params = {}; + int rc; + + if (!p_data) + return -1; + + /* Set DMAE params */ + SET_FIELD(params.flags, QED_DMAE_PARAMS_COMPLETION_DST, 1); + + /* Execute DMAE command */ + rc = qed_dmae_host2grc(p_hwfn, p_ptt, + (u64)(uintptr_t)(p_data), + addr, len_in_dwords, ¶ms); + + /* If not read using DMAE, read using GRC */ + if (rc) { + DP_VERBOSE(p_hwfn, + QED_MSG_DEBUG, + "Failed writing to chip using DMAE, using GRC instead\n"); + /* write to registers using GRC */ + ARR_REG_WR(p_hwfn, p_ptt, addr, p_data, len_in_dwords); + } + + return len_in_dwords; +} void qed_set_vxlan_dest_port(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u16 dest_port) @@ -1166,8 +1203,8 @@ void qed_set_geneve_enable(struct qed_hwfn *p_hwfn, ip_geneve_enable ? 1 : 0); } -#define PRS_ETH_VXLAN_NO_L2_ENABLE_OFFSET 4 -#define PRS_ETH_VXLAN_NO_L2_OUTPUT_FORMAT -927094512 +#define PRS_ETH_VXLAN_NO_L2_ENABLE_OFFSET 3 +#define PRS_ETH_VXLAN_NO_L2_OUTPUT_FORMAT -925189872 void qed_set_vxlan_no_l2_enable(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, bool enable) @@ -1208,6 +1245,8 @@ void qed_set_vxlan_no_l2_enable(struct qed_hwfn *p_hwfn, void qed_gft_disable(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u16 pf_id) { + struct regpair ram_line = { }; + /* Disable gft search for PF */ qed_wr(p_hwfn, p_ptt, PRS_REG_SEARCH_GFT, 0); @@ -1217,12 +1256,9 @@ void qed_gft_disable(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, u16 pf_id) qed_wr(p_hwfn, p_ptt, PRS_REG_GFT_CAM + CAM_LINE_SIZE * pf_id, 0); /* Zero ramline */ - qed_wr(p_hwfn, - p_ptt, PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id, 0); - qed_wr(p_hwfn, - p_ptt, - PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id + REG_SIZE, - 0); + qed_dmae_to_grc(p_hwfn, p_ptt, (u32 *)&ram_line, + PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id, + sizeof(ram_line) / REG_SIZE); } void qed_gft_config(struct qed_hwfn *p_hwfn, @@ -1232,7 +1268,8 @@ void qed_gft_config(struct qed_hwfn *p_hwfn, bool udp, bool ipv4, bool ipv6, enum gft_profile_type profile_type) { - u32 reg_val, cam_line, ram_line_lo, ram_line_hi, search_non_ip_as_gft; + u32 reg_val, cam_line, search_non_ip_as_gft; + struct regpair ram_line = { }; if (!ipv6 && !ipv4) DP_NOTICE(p_hwfn, @@ -1298,35 +1335,33 @@ void qed_gft_config(struct qed_hwfn *p_hwfn, qed_rd(p_hwfn, p_ptt, PRS_REG_GFT_CAM + CAM_LINE_SIZE * pf_id); /* Write line to RAM - compare to filter 4 tuple */ - ram_line_lo = 0; - ram_line_hi = 0; /* Search no IP as GFT */ search_non_ip_as_gft = 0; /* Tunnel type */ - SET_FIELD(ram_line_lo, GFT_RAM_LINE_TUNNEL_DST_PORT, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_TUNNEL_OVER_IP_PROTOCOL, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_TUNNEL_DST_PORT, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_TUNNEL_OVER_IP_PROTOCOL, 1); if (profile_type == GFT_PROFILE_TYPE_4_TUPLE) { - SET_FIELD(ram_line_hi, GFT_RAM_LINE_DST_IP, 1); - SET_FIELD(ram_line_hi, GFT_RAM_LINE_SRC_IP, 1); - SET_FIELD(ram_line_hi, GFT_RAM_LINE_OVER_IP_PROTOCOL, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_ETHERTYPE, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_SRC_PORT, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_DST_PORT, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_DST_IP, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_SRC_IP, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_OVER_IP_PROTOCOL, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_ETHERTYPE, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_SRC_PORT, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_DST_PORT, 1); } else if (profile_type == GFT_PROFILE_TYPE_L4_DST_PORT) { - SET_FIELD(ram_line_hi, GFT_RAM_LINE_OVER_IP_PROTOCOL, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_ETHERTYPE, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_DST_PORT, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_OVER_IP_PROTOCOL, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_ETHERTYPE, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_DST_PORT, 1); } else if (profile_type == GFT_PROFILE_TYPE_IP_DST_ADDR) { - SET_FIELD(ram_line_hi, GFT_RAM_LINE_DST_IP, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_ETHERTYPE, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_DST_IP, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_ETHERTYPE, 1); } else if (profile_type == GFT_PROFILE_TYPE_IP_SRC_ADDR) { - SET_FIELD(ram_line_hi, GFT_RAM_LINE_SRC_IP, 1); - SET_FIELD(ram_line_lo, GFT_RAM_LINE_ETHERTYPE, 1); + SET_FIELD(ram_line.hi, GFT_RAM_LINE_SRC_IP, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_ETHERTYPE, 1); } else if (profile_type == GFT_PROFILE_TYPE_TUNNEL_TYPE) { - SET_FIELD(ram_line_lo, GFT_RAM_LINE_TUNNEL_ETHERTYPE, 1); + SET_FIELD(ram_line.lo, GFT_RAM_LINE_TUNNEL_ETHERTYPE, 1); /* Allow tunneled traffic without inner IP */ search_non_ip_as_gft = 1; @@ -1334,24 +1369,17 @@ void qed_gft_config(struct qed_hwfn *p_hwfn, qed_wr(p_hwfn, p_ptt, PRS_REG_SEARCH_NON_IP_AS_GFT, search_non_ip_as_gft); - qed_wr(p_hwfn, - p_ptt, - PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id, - ram_line_lo); - qed_wr(p_hwfn, - p_ptt, - PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id + REG_SIZE, - ram_line_hi); + qed_dmae_to_grc(p_hwfn, p_ptt, (u32 *)&ram_line, + PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * pf_id, + sizeof(ram_line) / REG_SIZE); /* Set default profile so that no filter match will happen */ - qed_wr(p_hwfn, - p_ptt, - PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * - PRS_GFT_CAM_LINES_NO_MATCH, 0xffffffff); - qed_wr(p_hwfn, - p_ptt, - PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * - PRS_GFT_CAM_LINES_NO_MATCH + REG_SIZE, 0x3ff); + ram_line.lo = 0xffffffff; + ram_line.hi = 0x3ff; + qed_dmae_to_grc(p_hwfn, p_ptt, (u32 *)&ram_line, + PRS_REG_GFT_PROFILE_MASK_RAM + RAM_LINE_SIZE * + PRS_GFT_CAM_LINES_NO_MATCH, + sizeof(ram_line) / REG_SIZE); /* Enable gft search */ qed_wr(p_hwfn, p_ptt, PRS_REG_SEARCH_GFT, 1); @@ -1544,3 +1572,144 @@ void qed_set_rdma_error_level(struct qed_hwfn *p_hwfn, qed_wr(p_hwfn, p_ptt, ram_addr, assert_level[storm_id]); } } + +#define PHYS_ADDR_DWORDS DIV_ROUND_UP(sizeof(dma_addr_t), 4) +#define OVERLAY_HDR_SIZE_DWORDS (sizeof(struct fw_overlay_buf_hdr) / 4) + +static u32 qed_get_overlay_addr_ram_addr(struct qed_hwfn *p_hwfn, u8 storm_id) +{ + switch (storm_id) { + case 0: + return TSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + TSTORM_OVERLAY_BUF_ADDR_OFFSET; + case 1: + return MSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + MSTORM_OVERLAY_BUF_ADDR_OFFSET; + case 2: + return USEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + USTORM_OVERLAY_BUF_ADDR_OFFSET; + case 3: + return XSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + XSTORM_OVERLAY_BUF_ADDR_OFFSET; + case 4: + return YSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + YSTORM_OVERLAY_BUF_ADDR_OFFSET; + case 5: + return PSEM_REG_FAST_MEMORY + SEM_FAST_REG_INT_RAM + + PSTORM_OVERLAY_BUF_ADDR_OFFSET; + + default: + return 0; + } +} + +struct phys_mem_desc *qed_fw_overlay_mem_alloc(struct qed_hwfn *p_hwfn, + const u32 * const + fw_overlay_in_buf, + u32 buf_size_in_bytes) +{ + u32 buf_size = buf_size_in_bytes / sizeof(u32), buf_offset = 0; + struct phys_mem_desc *allocated_mem; + + if (!buf_size) + return NULL; + + allocated_mem = kcalloc(NUM_STORMS, sizeof(struct phys_mem_desc), + GFP_KERNEL); + if (!allocated_mem) + return NULL; + + memset(allocated_mem, 0, NUM_STORMS * sizeof(struct phys_mem_desc)); + + /* For each Storm, set physical address in RAM */ + while (buf_offset < buf_size) { + struct phys_mem_desc *storm_mem_desc; + struct fw_overlay_buf_hdr *hdr; + u32 storm_buf_size; + u8 storm_id; + + hdr = + (struct fw_overlay_buf_hdr *)&fw_overlay_in_buf[buf_offset]; + storm_buf_size = GET_FIELD(hdr->data, + FW_OVERLAY_BUF_HDR_BUF_SIZE); + storm_id = GET_FIELD(hdr->data, FW_OVERLAY_BUF_HDR_STORM_ID); + storm_mem_desc = allocated_mem + storm_id; + storm_mem_desc->size = storm_buf_size * sizeof(u32); + + /* Allocate physical memory for Storm's overlays buffer */ + storm_mem_desc->virt_addr = + dma_alloc_coherent(&p_hwfn->cdev->pdev->dev, + storm_mem_desc->size, + &storm_mem_desc->phys_addr, GFP_KERNEL); + if (!storm_mem_desc->virt_addr) + break; + + /* Skip overlays buffer header */ + buf_offset += OVERLAY_HDR_SIZE_DWORDS; + + /* Copy Storm's overlays buffer to allocated memory */ + memcpy(storm_mem_desc->virt_addr, + &fw_overlay_in_buf[buf_offset], storm_mem_desc->size); + + /* Advance to next Storm */ + buf_offset += storm_buf_size; + } + + /* If memory allocation has failed, free all allocated memory */ + if (buf_offset < buf_size) { + qed_fw_overlay_mem_free(p_hwfn, allocated_mem); + return NULL; + } + + return allocated_mem; +} + +void qed_fw_overlay_init_ram(struct qed_hwfn *p_hwfn, + struct qed_ptt *p_ptt, + struct phys_mem_desc *fw_overlay_mem) +{ + u8 storm_id; + + for (storm_id = 0; storm_id < NUM_STORMS; storm_id++) { + struct phys_mem_desc *storm_mem_desc = + (struct phys_mem_desc *)fw_overlay_mem + storm_id; + u32 ram_addr, i; + + /* Skip Storms with no FW overlays */ + if (!storm_mem_desc->virt_addr) + continue; + + /* Calculate overlay RAM GRC address of current PF */ + ram_addr = qed_get_overlay_addr_ram_addr(p_hwfn, storm_id) + + sizeof(dma_addr_t) * p_hwfn->rel_pf_id; + + /* Write Storm's overlay physical address to RAM */ + for (i = 0; i < PHYS_ADDR_DWORDS; i++, ram_addr += sizeof(u32)) + qed_wr(p_hwfn, p_ptt, ram_addr, + ((u32 *)&storm_mem_desc->phys_addr)[i]); + } +} + +void qed_fw_overlay_mem_free(struct qed_hwfn *p_hwfn, + struct phys_mem_desc *fw_overlay_mem) +{ + u8 storm_id; + + if (!fw_overlay_mem) + return; + + for (storm_id = 0; storm_id < NUM_STORMS; storm_id++) { + struct phys_mem_desc *storm_mem_desc = + (struct phys_mem_desc *)fw_overlay_mem + storm_id; + + /* Free Storm's physical memory */ + if (storm_mem_desc->virt_addr) + dma_free_coherent(&p_hwfn->cdev->pdev->dev, + storm_mem_desc->size, + storm_mem_desc->virt_addr, + storm_mem_desc->phys_addr); + } + + /* Free allocated virtual memory */ + kfree(fw_overlay_mem); +} diff --git a/drivers/net/ethernet/qlogic/qed/qed_init_ops.c b/drivers/net/ethernet/qlogic/qed/qed_init_ops.c index a868d7f88601..5a6e4ac4fef4 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_init_ops.c +++ b/drivers/net/ethernet/qlogic/qed/qed_init_ops.c @@ -54,15 +54,15 @@ static u32 pxp_global_win[] = { 0x1c80, /* win 3: addr=0x1c80000, size=4096 bytes */ 0x1d00, /* win 4: addr=0x1d00000, size=4096 bytes */ 0x1d01, /* win 5: addr=0x1d01000, size=4096 bytes */ - 0x1d80, /* win 6: addr=0x1d80000, size=4096 bytes */ - 0x1d81, /* win 7: addr=0x1d81000, size=4096 bytes */ - 0x1d82, /* win 8: addr=0x1d82000, size=4096 bytes */ - 0x1e00, /* win 9: addr=0x1e00000, size=4096 bytes */ - 0x1e80, /* win 10: addr=0x1e80000, size=4096 bytes */ - 0x1f00, /* win 11: addr=0x1f00000, size=4096 bytes */ - 0, - 0, - 0, + 0x1d02, /* win 6: addr=0x1d02000, size=4096 bytes */ + 0x1d80, /* win 7: addr=0x1d80000, size=4096 bytes */ + 0x1d81, /* win 8: addr=0x1d81000, size=4096 bytes */ + 0x1d82, /* win 9: addr=0x1d82000, size=4096 bytes */ + 0x1e00, /* win 10: addr=0x1e00000, size=4096 bytes */ + 0x1e01, /* win 11: addr=0x1e01000, size=4096 bytes */ + 0x1e80, /* win 12: addr=0x1e80000, size=4096 bytes */ + 0x1f00, /* win 13: addr=0x1f00000, size=4096 bytes */ + 0x1c08, /* win 14: addr=0x1c08000, size=4096 bytes */ 0, 0, 0, @@ -74,15 +74,6 @@ void qed_init_iro_array(struct qed_dev *cdev) cdev->iro_arr = iro_arr; } -/* Runtime configuration helpers */ -void qed_init_clear_rt_data(struct qed_hwfn *p_hwfn) -{ - int i; - - for (i = 0; i < RUNTIME_ARRAY_SIZE; i++) - p_hwfn->rt_data.b_valid[i] = false; -} - void qed_init_store_rt_reg(struct qed_hwfn *p_hwfn, u32 rt_offset, u32 val) { p_hwfn->rt_data.init_val[rt_offset] = val; @@ -106,7 +97,7 @@ static int qed_init_rt(struct qed_hwfn *p_hwfn, { u32 *p_init_val = &p_hwfn->rt_data.init_val[rt_offset]; bool *p_valid = &p_hwfn->rt_data.b_valid[rt_offset]; - u16 i, segment; + u16 i, j, segment; int rc = 0; /* Since not all RT entries are initialized, go over the RT and @@ -121,6 +112,7 @@ static int qed_init_rt(struct qed_hwfn *p_hwfn, */ if (!b_must_dmae) { qed_wr(p_hwfn, p_ptt, addr + (i << 2), p_init_val[i]); + p_valid[i] = false; continue; } @@ -135,6 +127,10 @@ static int qed_init_rt(struct qed_hwfn *p_hwfn, if (rc) return rc; + /* invalidate after writing */ + for (j = i; j < i + segment; j++) + p_valid[j] = false; + /* Jump over the entire segment, including invalid entry */ i += segment; } @@ -215,7 +211,7 @@ static int qed_init_fill_dmae(struct qed_hwfn *p_hwfn, * 3. p_hwfb->temp_data, * 4. fill_count */ - params.flags = QED_DMAE_FLAG_RW_REPL_SRC; + SET_FIELD(params.flags, QED_DMAE_PARAMS_RW_REPL_SRC, 0x1); return qed_dmae_host2grc(p_hwfn, p_ptt, (uintptr_t)(&zero_buffer[0]), addr, fill_count, ¶ms); @@ -490,10 +486,10 @@ static u32 qed_init_cmd_phase(struct qed_hwfn *p_hwfn, int qed_init_run(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, int phase, int phase_id, int modes) { + bool b_dmae = (phase != PHASE_ENGINE); struct qed_dev *cdev = p_hwfn->cdev; u32 cmd_num, num_init_ops; union init_op *init_ops; - bool b_dmae = false; int rc = 0; num_init_ops = cdev->fw_data->init_ops_size; @@ -522,7 +518,6 @@ int qed_init_run(struct qed_hwfn *p_hwfn, case INIT_OP_IF_PHASE: cmd_num += qed_init_cmd_phase(p_hwfn, &cmd->if_phase, phase, phase_id); - b_dmae = GET_FIELD(data, INIT_IF_PHASE_OP_DMAE_ENABLE); break; case INIT_OP_DELAY: /* qed_init_run is always invoked from @@ -533,6 +528,9 @@ int qed_init_run(struct qed_hwfn *p_hwfn, case INIT_OP_CALLBACK: rc = qed_init_cmd_cb(p_hwfn, p_ptt, &cmd->callback); + if (phase == PHASE_ENGINE && + cmd->callback.callback_id == DMAE_READY_CB) + b_dmae = true; break; } @@ -587,5 +585,10 @@ int qed_init_fw_data(struct qed_dev *cdev, const u8 *data) len = buf_hdr[BIN_BUF_INIT_CMD].length; fw->init_ops_size = len / sizeof(struct init_raw_op); + offset = buf_hdr[BIN_BUF_INIT_OVERLAYS].offset; + fw->fw_overlays = (u32 *)(data + offset); + len = buf_hdr[BIN_BUF_INIT_OVERLAYS].length; + fw->fw_overlays_len = len; + return 0; } diff --git a/drivers/net/ethernet/qlogic/qed/qed_init_ops.h b/drivers/net/ethernet/qlogic/qed/qed_init_ops.h index 555dd086796d..e9e8ade50ed3 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_init_ops.h +++ b/drivers/net/ethernet/qlogic/qed/qed_init_ops.h @@ -81,14 +81,6 @@ int qed_init_alloc(struct qed_hwfn *p_hwfn); void qed_init_free(struct qed_hwfn *p_hwfn); /** - * @brief qed_init_clear_rt_data - Clears the runtime init array. - * - * - * @param p_hwfn - */ -void qed_init_clear_rt_data(struct qed_hwfn *p_hwfn); - -/** * @brief qed_init_store_rt_reg - Store a configuration value in the RT array. * * diff --git a/drivers/net/ethernet/qlogic/qed/qed_iscsi.c b/drivers/net/ethernet/qlogic/qed/qed_iscsi.c index 5585c18053ec..7245a615517a 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_iscsi.c +++ b/drivers/net/ethernet/qlogic/qed/qed_iscsi.c @@ -204,17 +204,14 @@ qed_sp_iscsi_func_start(struct qed_hwfn *p_hwfn, return -EINVAL; } - SET_FIELD(p_init->hdr.flags, - ISCSI_SLOW_PATH_HDR_LAYER_CODE, ISCSI_SLOW_PATH_LAYER_CODE); - p_init->hdr.op_code = ISCSI_RAMROD_CMD_ID_INIT_FUNC; - val = p_params->half_way_close_timeout; p_init->half_way_close_timeout = cpu_to_le16(val); p_init->num_sq_pages_in_ring = p_params->num_sq_pages_in_ring; p_init->num_r2tq_pages_in_ring = p_params->num_r2tq_pages_in_ring; p_init->num_uhq_pages_in_ring = p_params->num_uhq_pages_in_ring; - p_init->ll2_rx_queue_id = p_hwfn->hw_info.resc_start[QED_LL2_QUEUE] + - p_params->ll2_ooo_queue_id; + p_init->ll2_rx_queue_id = + p_hwfn->hw_info.resc_start[QED_LL2_RAM_QUEUE] + + p_params->ll2_ooo_queue_id; p_init->func_params.log_page_size = p_params->log_page_size; val = p_params->num_tasks; @@ -331,12 +328,7 @@ static int qed_sp_iscsi_conn_offload(struct qed_hwfn *p_hwfn, p_conn->physical_q1 = cpu_to_le16(physical_q); p_ramrod->iscsi.physical_q1 = cpu_to_le16(physical_q); - p_ramrod->hdr.op_code = ISCSI_RAMROD_CMD_ID_OFFLOAD_CONN; - SET_FIELD(p_ramrod->hdr.flags, ISCSI_SLOW_PATH_HDR_LAYER_CODE, - p_conn->layer_code); - p_ramrod->conn_id = cpu_to_le16(p_conn->conn_id); - p_ramrod->fw_cid = cpu_to_le32(p_conn->icid); DMA_REGPAIR_LE(p_ramrod->iscsi.sq_pbl_addr, p_conn->sq_pbl_addr); @@ -492,12 +484,8 @@ static int qed_sp_iscsi_conn_update(struct qed_hwfn *p_hwfn, return rc; p_ramrod = &p_ent->ramrod.iscsi_conn_update; - p_ramrod->hdr.op_code = ISCSI_RAMROD_CMD_ID_UPDATE_CONN; - SET_FIELD(p_ramrod->hdr.flags, - ISCSI_SLOW_PATH_HDR_LAYER_CODE, p_conn->layer_code); p_ramrod->conn_id = cpu_to_le16(p_conn->conn_id); - p_ramrod->fw_cid = cpu_to_le32(p_conn->icid); p_ramrod->flags = p_conn->update_flag; p_ramrod->max_seq_size = cpu_to_le32(p_conn->max_seq_size); dval = p_conn->max_recv_pdu_length; @@ -537,12 +525,8 @@ qed_sp_iscsi_mac_update(struct qed_hwfn *p_hwfn, return rc; p_ramrod = &p_ent->ramrod.iscsi_conn_mac_update; - p_ramrod->hdr.op_code = ISCSI_RAMROD_CMD_ID_MAC_UPDATE; - SET_FIELD(p_ramrod->hdr.flags, - ISCSI_SLOW_PATH_HDR_LAYER_CODE, p_conn->layer_code); p_ramrod->conn_id = cpu_to_le16(p_conn->conn_id); - p_ramrod->fw_cid = cpu_to_le32(p_conn->icid); ucval = p_conn->remote_mac[1]; ((u8 *)(&p_ramrod->remote_mac_addr_hi))[0] = ucval; ucval = p_conn->remote_mac[0]; @@ -583,12 +567,8 @@ static int qed_sp_iscsi_conn_terminate(struct qed_hwfn *p_hwfn, return rc; p_ramrod = &p_ent->ramrod.iscsi_conn_terminate; - p_ramrod->hdr.op_code = ISCSI_RAMROD_CMD_ID_TERMINATION_CONN; - SET_FIELD(p_ramrod->hdr.flags, - ISCSI_SLOW_PATH_HDR_LAYER_CODE, p_conn->layer_code); p_ramrod->conn_id = cpu_to_le16(p_conn->conn_id); - p_ramrod->fw_cid = cpu_to_le32(p_conn->icid); p_ramrod->abortive = p_conn->abortive_dsconnect; DMA_REGPAIR_LE(p_ramrod->query_params_addr, @@ -603,7 +583,6 @@ static int qed_sp_iscsi_conn_clear_sq(struct qed_hwfn *p_hwfn, enum spq_mode comp_mode, struct qed_spq_comp_cb *p_comp_addr) { - struct iscsi_slow_path_hdr *p_ramrod = NULL; struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; int rc = -EINVAL; @@ -621,11 +600,6 @@ static int qed_sp_iscsi_conn_clear_sq(struct qed_hwfn *p_hwfn, if (rc) return rc; - p_ramrod = &p_ent->ramrod.iscsi_empty; - p_ramrod->op_code = ISCSI_RAMROD_CMD_ID_CLEAR_SQ; - SET_FIELD(p_ramrod->flags, - ISCSI_SLOW_PATH_HDR_LAYER_CODE, p_conn->layer_code); - return qed_spq_post(p_hwfn, p_ent, NULL); } @@ -633,7 +607,6 @@ static int qed_sp_iscsi_func_stop(struct qed_hwfn *p_hwfn, enum spq_mode comp_mode, struct qed_spq_comp_cb *p_comp_addr) { - struct iscsi_spe_func_dstry *p_ramrod = NULL; struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; int rc = 0; @@ -651,9 +624,6 @@ static int qed_sp_iscsi_func_stop(struct qed_hwfn *p_hwfn, if (rc) return rc; - p_ramrod = &p_ent->ramrod.iscsi_destroy; - p_ramrod->hdr.op_code = ISCSI_RAMROD_CMD_ID_DESTROY_FUNC; - rc = qed_spq_post(p_hwfn, p_ent, NULL); qed_spq_unregister_async_cb(p_hwfn, PROTOCOLID_ISCSI); diff --git a/drivers/net/ethernet/qlogic/qed/qed_iwarp.c b/drivers/net/ethernet/qlogic/qed/qed_iwarp.c index 65ec16a31658..d2fe61a5cf56 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_iwarp.c +++ b/drivers/net/ethernet/qlogic/qed/qed_iwarp.c @@ -137,8 +137,8 @@ qed_iwarp_init_fw_ramrod(struct qed_hwfn *p_hwfn, struct iwarp_init_func_ramrod_data *p_ramrod) { p_ramrod->iwarp.ll2_ooo_q_index = - RESC_START(p_hwfn, QED_LL2_QUEUE) + - p_hwfn->p_rdma_info->iwarp.ll2_ooo_handle; + RESC_START(p_hwfn, QED_LL2_RAM_QUEUE) + + p_hwfn->p_rdma_info->iwarp.ll2_ooo_handle; p_ramrod->tcp.max_fin_rt = QED_IWARP_MAX_FIN_RT_DEFAULT; @@ -2651,6 +2651,8 @@ qed_iwarp_ll2_start(struct qed_hwfn *p_hwfn, memset(&data, 0, sizeof(data)); data.input.conn_type = QED_LL2_TYPE_IWARP; + /* SYN will use ctx based queues */ + data.input.rx_conn_type = QED_LL2_RX_TYPE_CTX; data.input.mtu = params->max_mtu; data.input.rx_num_desc = QED_IWARP_LL2_SYN_RX_SIZE; data.input.tx_num_desc = QED_IWARP_LL2_SYN_TX_SIZE; @@ -2683,6 +2685,8 @@ qed_iwarp_ll2_start(struct qed_hwfn *p_hwfn, /* Start OOO connection */ data.input.conn_type = QED_LL2_TYPE_OOO; + /* OOO/unaligned will use legacy ll2 queues (ram based) */ + data.input.rx_conn_type = QED_LL2_RX_TYPE_LEGACY; data.input.mtu = params->max_mtu; n_ooo_bufs = (QED_IWARP_MAX_OOO * rcv_wnd_size) / diff --git a/drivers/net/ethernet/qlogic/qed/qed_ll2.c b/drivers/net/ethernet/qlogic/qed/qed_ll2.c index 19a1a58d60f8..037e5978787e 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_ll2.c +++ b/drivers/net/ethernet/qlogic/qed/qed_ll2.c @@ -962,7 +962,7 @@ static int qed_sp_ll2_rx_queue_start(struct qed_hwfn *p_hwfn, return rc; p_ramrod = &p_ent->ramrod.core_rx_queue_start; - + memset(p_ramrod, 0, sizeof(*p_ramrod)); p_ramrod->sb_id = cpu_to_le16(qed_int_get_sp_sb_id(p_hwfn)); p_ramrod->sb_index = p_rx->rx_sb_index; p_ramrod->complete_event_flg = 1; @@ -996,6 +996,8 @@ static int qed_sp_ll2_rx_queue_start(struct qed_hwfn *p_hwfn, p_ramrod->action_on_error.error_type = action_on_error; p_ramrod->gsi_offload_flag = p_ll2_conn->input.gsi_enable; + p_ramrod->zero_prod_flg = 1; + return qed_spq_post(p_hwfn, p_ent, NULL); } @@ -1317,6 +1319,25 @@ qed_ll2_set_cbs(struct qed_ll2_info *p_ll2_info, const struct qed_ll2_cbs *cbs) return 0; } +static void _qed_ll2_calc_allowed_conns(struct qed_hwfn *p_hwfn, + struct qed_ll2_acquire_data *data, + u8 *start_idx, u8 *last_idx) +{ + /* LL2 queues handles will be split as follows: + * First will be the legacy queues, and then the ctx based. + */ + if (data->input.rx_conn_type == QED_LL2_RX_TYPE_LEGACY) { + *start_idx = QED_LL2_LEGACY_CONN_BASE_PF; + *last_idx = *start_idx + + QED_MAX_NUM_OF_LEGACY_LL2_CONNS_PF; + } else { + /* QED_LL2_RX_TYPE_CTX */ + *start_idx = QED_LL2_CTX_CONN_BASE_PF; + *last_idx = *start_idx + + QED_MAX_NUM_OF_CTX_LL2_CONNS_PF; + } +} + static enum core_error_handle qed_ll2_get_error_choice(enum qed_ll2_error_handle err) { @@ -1337,14 +1358,16 @@ int qed_ll2_acquire_connection(void *cxt, struct qed_ll2_acquire_data *data) struct qed_hwfn *p_hwfn = cxt; qed_int_comp_cb_t comp_rx_cb, comp_tx_cb; struct qed_ll2_info *p_ll2_info = NULL; - u8 i, *p_tx_max; + u8 i, first_idx, last_idx, *p_tx_max; int rc; if (!data->p_connection_handle || !p_hwfn->p_ll2_info) return -EINVAL; + _qed_ll2_calc_allowed_conns(p_hwfn, data, &first_idx, &last_idx); + /* Find a free connection to be used */ - for (i = 0; (i < QED_MAX_NUM_OF_LL2_CONNECTIONS); i++) { + for (i = first_idx; i < last_idx; i++) { mutex_lock(&p_hwfn->p_ll2_info[i].mutex); if (p_hwfn->p_ll2_info[i].b_active) { mutex_unlock(&p_hwfn->p_ll2_info[i].mutex); @@ -1448,6 +1471,7 @@ static int qed_ll2_establish_connection_rx(struct qed_hwfn *p_hwfn, enum qed_ll2_error_handle error_input; enum core_error_handle error_mode; u8 action_on_error = 0; + int rc; if (!QED_LL2_RX_REGISTERED(p_ll2_conn)) return 0; @@ -1461,7 +1485,18 @@ static int qed_ll2_establish_connection_rx(struct qed_hwfn *p_hwfn, error_mode = qed_ll2_get_error_choice(error_input); SET_FIELD(action_on_error, CORE_RX_ACTION_ON_ERROR_NO_BUFF, error_mode); - return qed_sp_ll2_rx_queue_start(p_hwfn, p_ll2_conn, action_on_error); + rc = qed_sp_ll2_rx_queue_start(p_hwfn, p_ll2_conn, action_on_error); + if (rc) + return rc; + + if (p_ll2_conn->rx_queue.ctx_based) { + rc = qed_db_recovery_add(p_hwfn->cdev, + p_ll2_conn->rx_queue.set_prod_addr, + &p_ll2_conn->rx_queue.db_data, + DB_REC_WIDTH_64B, DB_REC_KERNEL); + } + + return rc; } static void @@ -1475,13 +1510,41 @@ qed_ll2_establish_connection_ooo(struct qed_hwfn *p_hwfn, qed_ooo_submit_rx_buffers(p_hwfn, p_ll2_conn); } +static inline u8 qed_ll2_handle_to_queue_id(struct qed_hwfn *p_hwfn, + u8 handle, + u8 ll2_queue_type) +{ + u8 qid; + + if (ll2_queue_type == QED_LL2_RX_TYPE_LEGACY) + return p_hwfn->hw_info.resc_start[QED_LL2_RAM_QUEUE] + handle; + + /* QED_LL2_RX_TYPE_CTX + * FW distinguishes between the legacy queues (ram based) and the + * ctx based queues by the queue_id. + * The first MAX_NUM_LL2_RX_RAM_QUEUES queues are legacy + * and the queue ids above that are ctx base. + */ + qid = p_hwfn->hw_info.resc_start[QED_LL2_CTX_QUEUE] + + MAX_NUM_LL2_RX_RAM_QUEUES; + + /* See comment on the acquire connection for how the ll2 + * queues handles are divided. + */ + qid += (handle - QED_MAX_NUM_OF_LEGACY_LL2_CONNS_PF); + + return qid; +} + int qed_ll2_establish_connection(void *cxt, u8 connection_handle) { - struct qed_hwfn *p_hwfn = cxt; - struct qed_ll2_info *p_ll2_conn; + struct e4_core_conn_context *p_cxt; struct qed_ll2_tx_packet *p_pkt; + struct qed_ll2_info *p_ll2_conn; + struct qed_hwfn *p_hwfn = cxt; struct qed_ll2_rx_queue *p_rx; struct qed_ll2_tx_queue *p_tx; + struct qed_cxt_info cxt_info; struct qed_ptt *p_ptt; int rc = -EINVAL; u32 i, capacity; @@ -1539,13 +1602,46 @@ int qed_ll2_establish_connection(void *cxt, u8 connection_handle) rc = qed_cxt_acquire_cid(p_hwfn, PROTOCOLID_CORE, &p_ll2_conn->cid); if (rc) goto out; + cxt_info.iid = p_ll2_conn->cid; + rc = qed_cxt_get_cid_info(p_hwfn, &cxt_info); + if (rc) { + DP_NOTICE(p_hwfn, "Cannot find context info for cid=%d\n", + p_ll2_conn->cid); + goto out; + } + + p_cxt = cxt_info.p_cxt; + + memset(p_cxt, 0, sizeof(*p_cxt)); - qid = p_hwfn->hw_info.resc_start[QED_LL2_QUEUE] + connection_handle; + qid = qed_ll2_handle_to_queue_id(p_hwfn, connection_handle, + p_ll2_conn->input.rx_conn_type); p_ll2_conn->queue_id = qid; p_ll2_conn->tx_stats_id = qid; - p_rx->set_prod_addr = (u8 __iomem *)p_hwfn->regview + - GTT_BAR0_MAP_REG_TSDM_RAM + - TSTORM_LL2_RX_PRODS_OFFSET(qid); + + DP_VERBOSE(p_hwfn, QED_MSG_LL2, + "Establishing ll2 queue. PF %d ctx_based=%d abs qid=%d\n", + p_hwfn->rel_pf_id, p_ll2_conn->input.rx_conn_type, qid); + + if (p_ll2_conn->input.rx_conn_type == QED_LL2_RX_TYPE_LEGACY) { + p_rx->set_prod_addr = p_hwfn->regview + + GTT_BAR0_MAP_REG_TSDM_RAM + TSTORM_LL2_RX_PRODS_OFFSET(qid); + } else { + /* QED_LL2_RX_TYPE_CTX - using doorbell */ + p_rx->ctx_based = 1; + + p_rx->set_prod_addr = p_hwfn->doorbells + + p_hwfn->dpi_start_offset + + DB_ADDR_SHIFT(DQ_PWM_OFFSET_TCM_LL2_PROD_UPDATE); + + /* prepare db data */ + p_rx->db_data.icid = cpu_to_le16((u16)p_ll2_conn->cid); + SET_FIELD(p_rx->db_data.params, + CORE_PWM_PROD_UPDATE_DATA_AGG_CMD, DB_AGG_CMD_SET); + SET_FIELD(p_rx->db_data.params, + CORE_PWM_PROD_UPDATE_DATA_RESERVED1, 0); + } + p_tx->doorbell_addr = (u8 __iomem *)p_hwfn->doorbells + qed_db_addr(p_ll2_conn->cid, DQ_DEMS_LEGACY); @@ -1556,7 +1652,6 @@ int qed_ll2_establish_connection(void *cxt, u8 connection_handle) DQ_XCM_CORE_TX_BD_PROD_CMD); p_tx->db_msg.agg_flags = DQ_XCM_CORE_DQ_CF_CMD; - rc = qed_ll2_establish_connection_rx(p_hwfn, p_ll2_conn); if (rc) goto out; @@ -1590,7 +1685,7 @@ static void qed_ll2_post_rx_buffer_notify_fw(struct qed_hwfn *p_hwfn, struct qed_ll2_rx_packet *p_curp) { struct qed_ll2_rx_packet *p_posting_packet = NULL; - struct core_ll2_rx_prod rx_prod = { 0, 0, 0 }; + struct core_ll2_rx_prod rx_prod = { 0, 0 }; bool b_notify_fw = false; u16 bd_prod, cq_prod; @@ -1615,13 +1710,27 @@ static void qed_ll2_post_rx_buffer_notify_fw(struct qed_hwfn *p_hwfn, bd_prod = qed_chain_get_prod_idx(&p_rx->rxq_chain); cq_prod = qed_chain_get_prod_idx(&p_rx->rcq_chain); - rx_prod.bd_prod = cpu_to_le16(bd_prod); - rx_prod.cqe_prod = cpu_to_le16(cq_prod); + if (p_rx->ctx_based) { + /* update producer by giving a doorbell */ + p_rx->db_data.prod.bd_prod = cpu_to_le16(bd_prod); + p_rx->db_data.prod.cqe_prod = cpu_to_le16(cq_prod); + /* Make sure chain element is updated before ringing the + * doorbell + */ + dma_wmb(); + DIRECT_REG_WR64(p_rx->set_prod_addr, + *((u64 *)&p_rx->db_data)); + } else { + rx_prod.bd_prod = cpu_to_le16(bd_prod); + rx_prod.cqe_prod = cpu_to_le16(cq_prod); - /* Make sure chain element is updated before ringing the doorbell */ - dma_wmb(); + /* Make sure chain element is updated before ringing the + * doorbell + */ + dma_wmb(); - DIRECT_REG_WR(p_rx->set_prod_addr, *((u32 *)&rx_prod)); + DIRECT_REG_WR(p_rx->set_prod_addr, *((u32 *)&rx_prod)); + } } int qed_ll2_post_rx_buffer(void *cxt, @@ -1965,6 +2074,12 @@ int qed_ll2_terminate_connection(void *cxt, u8 connection_handle) if (QED_LL2_RX_REGISTERED(p_ll2_conn)) { p_ll2_conn->rx_queue.b_cb_registered = false; smp_wmb(); /* Make sure this is seen by ll2_lb_rxq_completion */ + + if (p_ll2_conn->rx_queue.ctx_based) + qed_db_recovery_del(p_hwfn->cdev, + p_ll2_conn->rx_queue.set_prod_addr, + &p_ll2_conn->rx_queue.db_data); + rc = qed_sp_ll2_rx_queue_stop(p_hwfn, p_ll2_conn); if (rc) goto out; diff --git a/drivers/net/ethernet/qlogic/qed/qed_ll2.h b/drivers/net/ethernet/qlogic/qed/qed_ll2.h index 5f01fbd3c073..288642d526b7 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_ll2.h +++ b/drivers/net/ethernet/qlogic/qed/qed_ll2.h @@ -46,6 +46,18 @@ #include "qed_sp.h" #define QED_MAX_NUM_OF_LL2_CONNECTIONS (4) +/* LL2 queues handles will be split as follows: + * first will be legacy queues, and then the ctx based queues. + */ +#define QED_MAX_NUM_OF_LL2_CONNS_PF (4) +#define QED_MAX_NUM_OF_LEGACY_LL2_CONNS_PF (3) + +#define QED_MAX_NUM_OF_CTX_LL2_CONNS_PF \ + (QED_MAX_NUM_OF_LL2_CONNS_PF - QED_MAX_NUM_OF_LEGACY_LL2_CONNS_PF) + +#define QED_LL2_LEGACY_CONN_BASE_PF 0 +#define QED_LL2_CTX_CONN_BASE_PF QED_MAX_NUM_OF_LEGACY_LL2_CONNS_PF + struct qed_ll2_rx_packet { struct list_head list_entry; @@ -79,6 +91,7 @@ struct qed_ll2_rx_queue { struct qed_chain rxq_chain; struct qed_chain rcq_chain; u8 rx_sb_index; + u8 ctx_based; bool b_cb_registered; __le16 *p_fw_cons; struct list_head active_descq; @@ -86,6 +99,7 @@ struct qed_ll2_rx_queue { struct list_head posting_descq; struct qed_ll2_rx_packet *descq_array; void __iomem *set_prod_addr; + struct core_pwm_prod_update_data db_data; }; struct qed_ll2_tx_queue { diff --git a/drivers/net/ethernet/qlogic/qed/qed_main.c b/drivers/net/ethernet/qlogic/qed/qed_main.c index 38f7f40b3a4d..2c189c637cca 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_main.c +++ b/drivers/net/ethernet/qlogic/qed/qed_main.c @@ -2637,7 +2637,7 @@ static int qed_set_grc_config(struct qed_dev *cdev, u32 cfg_id, u32 val) if (!ptt) return -EAGAIN; - rc = qed_dbg_grc_config(hwfn, ptt, cfg_id, val); + rc = qed_dbg_grc_config(hwfn, cfg_id, val); qed_ptt_release(hwfn, ptt); diff --git a/drivers/net/ethernet/qlogic/qed/qed_mcp.c b/drivers/net/ethernet/qlogic/qed/qed_mcp.c index 36ddb89856a8..280527cc0578 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_mcp.c +++ b/drivers/net/ethernet/qlogic/qed/qed_mcp.c @@ -48,6 +48,8 @@ #include "qed_reg_addr.h" #include "qed_sriov.h" +#define GRCBASE_MCP 0xe00000 + #define QED_MCP_RESP_ITER_US 10 #define QED_DRV_MB_MAX_RETRIES (500 * 1000) /* Account for 5 sec */ @@ -3165,6 +3167,9 @@ qed_mcp_get_nvm_image_att(struct qed_hwfn *p_hwfn, case QED_NVM_IMAGE_FCOE_CFG: type = NVM_TYPE_FCOE_CFG; break; + case QED_NVM_IMAGE_MDUMP: + type = NVM_TYPE_MDUMP; + break; case QED_NVM_IMAGE_NVM_CFG1: type = NVM_TYPE_NVM_CFG1; break; @@ -3261,9 +3266,12 @@ static enum resource_id_enum qed_mcp_get_mfw_res_id(enum qed_resources res_id) case QED_ILT: mfw_res_id = RESOURCE_ILT_E; break; - case QED_LL2_QUEUE: + case QED_LL2_RAM_QUEUE: mfw_res_id = RESOURCE_LL2_QUEUE_E; break; + case QED_LL2_CTX_QUEUE: + mfw_res_id = RESOURCE_LL2_CQS_E; + break; case QED_RDMA_CNQ_RAM: case QED_CMDQS_CQS: /* CNQ/CMDQS are the same resource */ diff --git a/drivers/net/ethernet/qlogic/qed/qed_reg_addr.h b/drivers/net/ethernet/qlogic/qed/qed_reg_addr.h index 60f850c3bdd6..3dcb6ff58e73 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_reg_addr.h +++ b/drivers/net/ethernet/qlogic/qed/qed_reg_addr.h @@ -178,6 +178,8 @@ 0x008c80UL #define MCP_REG_SCRATCH \ 0xe20000UL +#define MCP_REG_SCRATCH_SIZE \ + 57344 #define CNIG_REG_NW_PORT_MODE_BB \ 0x218200UL #define MISCS_REG_CHIP_NUM \ @@ -212,6 +214,8 @@ 0x580900UL #define DBG_REG_CLIENT_ENABLE \ 0x010004UL +#define DBG_REG_TIMESTAMP_VALID_EN \ + 0x010b58UL #define DMAE_REG_INIT \ 0x00c000UL #define DORQ_REG_IFEN \ @@ -350,6 +354,10 @@ 0x24000cUL #define PSWRQ2_REG_ILT_MEMORY \ 0x260000UL +#define PSWRQ2_REG_ILT_MEMORY_SIZE_BB \ + 15200 +#define PSWRQ2_REG_ILT_MEMORY_SIZE_K2 \ + 22000 #define PSWHST_REG_DISCARD_INTERNAL_WRITES \ 0x2a0040UL #define PSWHST2_REG_DBGSYN_ALMOST_FULL_THR \ @@ -1453,6 +1461,8 @@ 0x1401404UL #define XSEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1401408UL +#define XSEM_REG_DBG_GPRE_VECT \ + 0x1401410UL #define XSEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1401420UL #define XSEM_REG_FAST_MEMORY \ @@ -1465,6 +1475,8 @@ 0x1501404UL #define YSEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1501408UL +#define YSEM_REG_DBG_GPRE_VECT \ + 0x1501410UL #define YSEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1501420UL #define YSEM_REG_FAST_MEMORY \ @@ -1479,6 +1491,8 @@ 0x1601404UL #define PSEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1601408UL +#define PSEM_REG_DBG_GPRE_VECT \ + 0x1601410UL #define PSEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1601420UL #define PSEM_REG_FAST_MEMORY \ @@ -1493,6 +1507,8 @@ 0x1701404UL #define TSEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1701408UL +#define TSEM_REG_DBG_GPRE_VECT \ + 0x1701410UL #define TSEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1701420UL #define TSEM_REG_FAST_MEMORY \ @@ -1507,12 +1523,16 @@ 0x1801404UL #define MSEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1801408UL +#define MSEM_REG_DBG_GPRE_VECT \ + 0x1801410UL #define MSEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1801420UL #define MSEM_REG_FAST_MEMORY \ 0x1840000UL #define USEM_REG_SLOW_DBG_EMPTY_BB_K2 \ 0x1901140UL +#define SEM_FAST_REG_INT_RAM_SIZE \ + 20480 #define USEM_REG_SYNC_DBG_EMPTY \ 0x1901160UL #define USEM_REG_SLOW_DBG_ACTIVE_BB_K2 \ @@ -1521,14 +1541,26 @@ 0x1901404UL #define USEM_REG_DBG_FRAME_MODE_BB_K2 \ 0x1901408UL +#define USEM_REG_DBG_GPRE_VECT \ + 0x1901410UL #define USEM_REG_DBG_MODE1_CFG_BB_K2 \ 0x1901420UL #define USEM_REG_FAST_MEMORY \ 0x1940000UL +#define SEM_FAST_REG_DBG_MODE23_SRC_DISABLE \ + 0x000748UL +#define SEM_FAST_REG_DBG_MODE4_SRC_DISABLE \ + 0x00074cUL +#define SEM_FAST_REG_DBG_MODE6_SRC_DISABLE \ + 0x000750UL +#define SEM_FAST_REG_DEBUG_ACTIVE \ + 0x000740UL #define SEM_FAST_REG_INT_RAM \ 0x020000UL #define SEM_FAST_REG_INT_RAM_SIZE_BB_K2 \ 20480 +#define SEM_FAST_REG_RECORD_FILTER_ENABLE \ + 0x000768UL #define GRC_REG_TRACE_FIFO_VALID_DATA \ 0x050064UL #define GRC_REG_NUMBER_VALID_OVERRIDE_WINDOW \ @@ -1583,14 +1615,20 @@ 0x181530UL #define DBG_REG_DBG_BLOCK_ON \ 0x010454UL +#define DBG_REG_FILTER_ENABLE \ + 0x0109d0UL #define DBG_REG_FRAMING_MODE \ 0x010058UL +#define DBG_REG_TRIGGER_ENABLE \ + 0x01054cUL #define SEM_FAST_REG_VFC_DATA_WR \ 0x000b40UL #define SEM_FAST_REG_VFC_ADDR \ 0x000b44UL #define SEM_FAST_REG_VFC_DATA_RD \ 0x000b48UL +#define SEM_FAST_REG_VFC_STATUS \ + 0x000b4cUL #define RSS_REG_RSS_RAM_DATA \ 0x238c20UL #define RSS_REG_RSS_RAM_DATA_SIZE \ diff --git a/drivers/net/ethernet/qlogic/qed/qed_roce.c b/drivers/net/ethernet/qlogic/qed/qed_roce.c index e49fada85410..37e70562a964 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_roce.c +++ b/drivers/net/ethernet/qlogic/qed/qed_roce.c @@ -900,7 +900,7 @@ int qed_roce_query_qp(struct qed_hwfn *p_hwfn, goto err_resp; out_params->rq_psn = le32_to_cpu(p_resp_ramrod_res->psn); - rq_err_state = GET_FIELD(le32_to_cpu(p_resp_ramrod_res->err_flag), + rq_err_state = GET_FIELD(le32_to_cpu(p_resp_ramrod_res->flags), ROCE_QUERY_QP_RESP_OUTPUT_PARAMS_ERROR_FLG); dma_free_coherent(&p_hwfn->cdev->pdev->dev, sizeof(*p_resp_ramrod_res), diff --git a/drivers/net/ethernet/qlogic/qed/qed_sp.h b/drivers/net/ethernet/qlogic/qed/qed_sp.h index 96ab77ae6af5..b7b4fbbbccfe 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_sp.h +++ b/drivers/net/ethernet/qlogic/qed/qed_sp.h @@ -120,9 +120,7 @@ union ramrod_data { struct fcoe_conn_terminate_ramrod_params fcoe_conn_terminate; struct fcoe_stat_ramrod_params fcoe_stat; - struct iscsi_slow_path_hdr iscsi_empty; struct iscsi_init_ramrod_params iscsi_init; - struct iscsi_spe_func_dstry iscsi_destroy; struct iscsi_spe_conn_offload iscsi_conn_offload; struct iscsi_conn_update_ramrod_params iscsi_conn_update; struct iscsi_spe_conn_mac_update iscsi_conn_mac_update; diff --git a/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c b/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c index 7e0b795230b2..900bc603e30a 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c +++ b/drivers/net/ethernet/qlogic/qed/qed_sp_commands.c @@ -331,8 +331,8 @@ int qed_sp_pf_start(struct qed_hwfn *p_hwfn, u8 sb_index = p_hwfn->p_eq->eq_sb_index; struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; - int rc = -EINVAL; u8 page_cnt, i; + int rc; /* update initial eq producer */ qed_eq_prod_update(p_hwfn, @@ -447,7 +447,7 @@ int qed_sp_pf_update(struct qed_hwfn *p_hwfn) { struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; - int rc = -EINVAL; + int rc; /* Get SPQ entry */ memset(&init_data, 0, sizeof(init_data)); @@ -471,7 +471,7 @@ int qed_sp_pf_update_ufp(struct qed_hwfn *p_hwfn) { struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; - int rc = -EOPNOTSUPP; + int rc; if (p_hwfn->ufp_info.pri_type == QED_UFP_PRI_UNKNOWN) { DP_INFO(p_hwfn, "Invalid priority type %d\n", @@ -509,7 +509,7 @@ int qed_sp_pf_update_tunn_cfg(struct qed_hwfn *p_hwfn, { struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; - int rc = -EINVAL; + int rc; if (IS_VF(p_hwfn->cdev)) return qed_vf_pf_tunnel_param_update(p_hwfn, p_tunn); @@ -546,7 +546,7 @@ int qed_sp_pf_stop(struct qed_hwfn *p_hwfn) { struct qed_spq_entry *p_ent = NULL; struct qed_sp_init_data init_data; - int rc = -EINVAL; + int rc; /* Get SPQ entry */ memset(&init_data, 0, sizeof(init_data)); diff --git a/drivers/net/ethernet/qlogic/qed/qed_sriov.c b/drivers/net/ethernet/qlogic/qed/qed_sriov.c index dcb5c917f373..66876af814c4 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_sriov.c +++ b/drivers/net/ethernet/qlogic/qed/qed_sriov.c @@ -352,7 +352,7 @@ static int qed_iov_post_vf_bulletin(struct qed_hwfn *p_hwfn, /* propagate bulletin board via dmae to vm memory */ memset(¶ms, 0, sizeof(params)); - params.flags = QED_DMAE_FLAG_VF_DST; + SET_FIELD(params.flags, QED_DMAE_PARAMS_DST_VF_VALID, 0x1); params.dst_vfid = p_vf->abs_vf_id; return qed_dmae_host2host(p_hwfn, p_ptt, p_vf->bulletin.phys, p_vf->vf_bulletin, p_vf->bulletin.size / 4, @@ -1225,8 +1225,8 @@ static void qed_iov_send_response(struct qed_hwfn *p_hwfn, eng_vf_id = p_vf->abs_vf_id; - memset(¶ms, 0, sizeof(struct qed_dmae_params)); - params.flags = QED_DMAE_FLAG_VF_DST; + memset(¶ms, 0, sizeof(params)); + SET_FIELD(params.flags, QED_DMAE_PARAMS_DST_VF_VALID, 0x1); params.dst_vfid = eng_vf_id; qed_dmae_host2host(p_hwfn, p_ptt, mbx->reply_phys + sizeof(u64), @@ -4103,8 +4103,9 @@ static int qed_iov_copy_vf_msg(struct qed_hwfn *p_hwfn, struct qed_ptt *ptt, if (!vf_info) return -EINVAL; - memset(¶ms, 0, sizeof(struct qed_dmae_params)); - params.flags = QED_DMAE_FLAG_VF_SRC | QED_DMAE_FLAG_COMPLETION_DST; + memset(¶ms, 0, sizeof(params)); + SET_FIELD(params.flags, QED_DMAE_PARAMS_SRC_VF_VALID, 0x1); + SET_FIELD(params.flags, QED_DMAE_PARAMS_COMPLETION_DST, 0x1); params.src_vfid = vf_info->abs_vf_id; if (qed_dmae_host2host(p_hwfn, ptt, @@ -4354,9 +4355,9 @@ qed_iov_bulletin_get_forced_vlan(struct qed_hwfn *p_hwfn, u16 rel_vf_id) static int qed_iov_configure_tx_rate(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt, int vfid, int val) { - struct qed_mcp_link_state *p_link; struct qed_vf_info *vf; u8 abs_vp_id = 0; + u16 rl_id; int rc; vf = qed_iov_get_vf_info(p_hwfn, (u16)vfid, true); @@ -4367,10 +4368,8 @@ static int qed_iov_configure_tx_rate(struct qed_hwfn *p_hwfn, if (rc) return rc; - p_link = &QED_LEADING_HWFN(p_hwfn->cdev)->mcp_info->link_output; - - return qed_init_vport_rl(p_hwfn, p_ptt, abs_vp_id, (u32)val, - p_link->speed); + rl_id = abs_vp_id; /* The "rl_id" is set as the "vport_id" */ + return qed_init_global_rl(p_hwfn, p_ptt, rl_id, (u32)val); } static int diff --git a/drivers/net/ethernet/qlogic/qede/qede_fp.c b/drivers/net/ethernet/qlogic/qede/qede_fp.c index 004c0bfec41d..c6c20776b474 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_fp.c +++ b/drivers/net/ethernet/qlogic/qede/qede_fp.c @@ -848,13 +848,13 @@ static void qede_tpa_start(struct qede_dev *edev, qede_set_gro_params(edev, tpa_info->skb, cqe); cons_buf: /* We still need to handle bd_len_list to consume buffers */ - if (likely(cqe->ext_bd_len_list[0])) + if (likely(cqe->bw_ext_bd_len_list[0])) qede_fill_frag_skb(edev, rxq, cqe->tpa_agg_index, - le16_to_cpu(cqe->ext_bd_len_list[0])); + le16_to_cpu(cqe->bw_ext_bd_len_list[0])); - if (unlikely(cqe->ext_bd_len_list[1])) { + if (unlikely(cqe->bw_ext_bd_len_list[1])) { DP_ERR(edev, - "Unlikely - got a TPA aggregation with more than one ext_bd_len_list entry in the TPA start\n"); + "Unlikely - got a TPA aggregation with more than one bw_ext_bd_len_list entry in the TPA start\n"); tpa_info->state = QEDE_AGG_STATE_ERROR; } } diff --git a/drivers/net/ethernet/qlogic/qede/qede_ptp.c b/drivers/net/ethernet/qlogic/qede/qede_ptp.c index f815435cf106..4c7f7a7fc151 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_ptp.c +++ b/drivers/net/ethernet/qlogic/qede/qede_ptp.c @@ -247,6 +247,7 @@ static int qede_ptp_cfg_filters(struct qede_dev *edev) break; case HWTSTAMP_TX_ONESTEP_SYNC: + case HWTSTAMP_TX_ONESTEP_P2P: DP_ERR(edev, "One-step timestamping is not supported\n"); return -ERANGE; } diff --git a/drivers/net/ethernet/qlogic/qla3xxx.c b/drivers/net/ethernet/qlogic/qla3xxx.c index 986f26578d34..0fade19e00d4 100644 --- a/drivers/net/ethernet/qlogic/qla3xxx.c +++ b/drivers/net/ethernet/qlogic/qla3xxx.c @@ -3602,7 +3602,7 @@ static int ql3xxx_set_mac_address(struct net_device *ndev, void *p) return 0; } -static void ql3xxx_tx_timeout(struct net_device *ndev) +static void ql3xxx_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct ql3_adapter *qdev = netdev_priv(ndev); diff --git a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c index c07438db30ba..9dd6cb36f366 100644 --- a/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c +++ b/drivers/net/ethernet/qlogic/qlcnic/qlcnic_main.c @@ -56,7 +56,7 @@ static int qlcnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent); static void qlcnic_remove(struct pci_dev *pdev); static int qlcnic_open(struct net_device *netdev); static int qlcnic_close(struct net_device *netdev); -static void qlcnic_tx_timeout(struct net_device *netdev); +static void qlcnic_tx_timeout(struct net_device *netdev, unsigned int txqueue); static void qlcnic_attach_work(struct work_struct *work); static void qlcnic_fwinit_work(struct work_struct *work); @@ -3068,7 +3068,7 @@ static void qlcnic_dump_rings(struct qlcnic_adapter *adapter) } -static void qlcnic_tx_timeout(struct net_device *netdev) +static void qlcnic_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct qlcnic_adapter *adapter = netdev_priv(netdev); diff --git a/drivers/net/ethernet/qualcomm/emac/emac.c b/drivers/net/ethernet/qualcomm/emac/emac.c index 98f92268cbaa..18b0c7a2d6dc 100644 --- a/drivers/net/ethernet/qualcomm/emac/emac.c +++ b/drivers/net/ethernet/qualcomm/emac/emac.c @@ -282,25 +282,13 @@ static int emac_close(struct net_device *netdev) } /* Respond to a TX hang */ -static void emac_tx_timeout(struct net_device *netdev) +static void emac_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct emac_adapter *adpt = netdev_priv(netdev); schedule_work(&adpt->work_thread); } -/* IOCTL support for the interface */ -static int emac_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd) -{ - if (!netif_running(netdev)) - return -EINVAL; - - if (!netdev->phydev) - return -ENODEV; - - return phy_mii_ioctl(netdev->phydev, ifr, cmd); -} - /** * emac_update_hw_stats - read the EMAC stat registers * @@ -387,7 +375,7 @@ static const struct net_device_ops emac_netdev_ops = { .ndo_start_xmit = emac_start_xmit, .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = emac_change_mtu, - .ndo_do_ioctl = emac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_tx_timeout = emac_tx_timeout, .ndo_get_stats64 = emac_get_stats64, .ndo_set_features = emac_set_features, diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c index baac016f3ec0..5a3b65a6eb4f 100644 --- a/drivers/net/ethernet/qualcomm/qca_spi.c +++ b/drivers/net/ethernet/qualcomm/qca_spi.c @@ -785,7 +785,7 @@ qcaspi_netdev_xmit(struct sk_buff *skb, struct net_device *dev) } static void -qcaspi_netdev_tx_timeout(struct net_device *dev) +qcaspi_netdev_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct qcaspi *qca = netdev_priv(dev); diff --git a/drivers/net/ethernet/qualcomm/qca_uart.c b/drivers/net/ethernet/qualcomm/qca_uart.c index 0981068504fa..375a844cd27c 100644 --- a/drivers/net/ethernet/qualcomm/qca_uart.c +++ b/drivers/net/ethernet/qualcomm/qca_uart.c @@ -248,7 +248,7 @@ out: return NETDEV_TX_OK; } -static void qcauart_netdev_tx_timeout(struct net_device *dev) +static void qcauart_netdev_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct qcauart *qca = netdev_priv(dev); diff --git a/drivers/net/ethernet/rdc/r6040.c b/drivers/net/ethernet/rdc/r6040.c index 274e5b4bc4ac..f5ecc410ff85 100644 --- a/drivers/net/ethernet/rdc/r6040.c +++ b/drivers/net/ethernet/rdc/r6040.c @@ -410,7 +410,7 @@ static void r6040_init_mac_regs(struct net_device *dev) iowrite16(TM2TX, ioaddr + MTPR); } -static void r6040_tx_timeout(struct net_device *dev) +static void r6040_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct r6040_private *priv = netdev_priv(dev); void __iomem *ioaddr = priv->base; @@ -498,14 +498,6 @@ static int r6040_close(struct net_device *dev) return 0; } -static int r6040_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static int r6040_rx(struct net_device *dev, int limit) { struct r6040_private *priv = netdev_priv(dev); @@ -957,7 +949,7 @@ static const struct net_device_ops r6040_netdev_ops = { .ndo_set_rx_mode = r6040_multicast_list, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, - .ndo_do_ioctl = r6040_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_tx_timeout = r6040_tx_timeout, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = r6040_poll_controller, diff --git a/drivers/net/ethernet/realtek/8139cp.c b/drivers/net/ethernet/realtek/8139cp.c index 4f910c4f67b0..60d342f82fb3 100644 --- a/drivers/net/ethernet/realtek/8139cp.c +++ b/drivers/net/ethernet/realtek/8139cp.c @@ -1235,7 +1235,7 @@ static int cp_close (struct net_device *dev) return 0; } -static void cp_tx_timeout(struct net_device *dev) +static void cp_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct cp_private *cp = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/ethernet/realtek/8139too.c b/drivers/net/ethernet/realtek/8139too.c index 55d01266e615..5caeb8368eab 100644 --- a/drivers/net/ethernet/realtek/8139too.c +++ b/drivers/net/ethernet/realtek/8139too.c @@ -642,7 +642,7 @@ static int mdio_read (struct net_device *dev, int phy_id, int location); static void mdio_write (struct net_device *dev, int phy_id, int location, int val); static void rtl8139_start_thread(struct rtl8139_private *tp); -static void rtl8139_tx_timeout (struct net_device *dev); +static void rtl8139_tx_timeout (struct net_device *dev, unsigned int txqueue); static void rtl8139_init_ring (struct net_device *dev); static netdev_tx_t rtl8139_start_xmit (struct sk_buff *skb, struct net_device *dev); @@ -1700,7 +1700,7 @@ static void rtl8139_tx_timeout_task (struct work_struct *work) spin_unlock_bh(&tp->rx_lock); } -static void rtl8139_tx_timeout (struct net_device *dev) +static void rtl8139_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct rtl8139_private *tp = netdev_priv(dev); diff --git a/drivers/net/ethernet/realtek/Makefile b/drivers/net/ethernet/realtek/Makefile index d5304bad2372..2e1d78b106b0 100644 --- a/drivers/net/ethernet/realtek/Makefile +++ b/drivers/net/ethernet/realtek/Makefile @@ -6,5 +6,5 @@ obj-$(CONFIG_8139CP) += 8139cp.o obj-$(CONFIG_8139TOO) += 8139too.o obj-$(CONFIG_ATP) += atp.o -r8169-objs += r8169_main.o r8169_firmware.o +r8169-objs += r8169_main.o r8169_firmware.o r8169_phy_config.o obj-$(CONFIG_R8169) += r8169.o diff --git a/drivers/net/ethernet/realtek/atp.c b/drivers/net/ethernet/realtek/atp.c index 58e0ca9093d3..9e3b35c97e63 100644 --- a/drivers/net/ethernet/realtek/atp.c +++ b/drivers/net/ethernet/realtek/atp.c @@ -204,7 +204,7 @@ static void net_rx(struct net_device *dev); static void read_block(long ioaddr, int length, unsigned char *buffer, int data_mode); static int net_close(struct net_device *dev); static void set_rx_mode(struct net_device *dev); -static void tx_timeout(struct net_device *dev); +static void tx_timeout(struct net_device *dev, unsigned int txqueue); /* A list of all installed ATP devices, for removing the driver module. */ @@ -533,7 +533,7 @@ static void write_packet(long ioaddr, int length, unsigned char *packet, int pad outb(Ctrl_HNibWrite | Ctrl_SelData | Ctrl_IRQEN, ioaddr + PAR_CONTROL); } -static void tx_timeout(struct net_device *dev) +static void tx_timeout(struct net_device *dev, unsigned int txqueue) { long ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/realtek/r8169.h b/drivers/net/ethernet/realtek/r8169.h new file mode 100644 index 000000000000..22a6a057b11e --- /dev/null +++ b/drivers/net/ethernet/realtek/r8169.h @@ -0,0 +1,78 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* r8169.h: RealTek 8169/8168/8101 ethernet driver. + * + * Copyright (c) 2002 ShuChen <shuchen@realtek.com.tw> + * Copyright (c) 2003 - 2007 Francois Romieu <romieu@fr.zoreil.com> + * Copyright (c) a lot of people too. Please respect their work. + * + * See MAINTAINERS file for support contact information. + */ + +#include <linux/types.h> +#include <linux/phy.h> + +enum mac_version { + /* support for ancient RTL_GIGA_MAC_VER_01 has been removed */ + RTL_GIGA_MAC_VER_02, + RTL_GIGA_MAC_VER_03, + RTL_GIGA_MAC_VER_04, + RTL_GIGA_MAC_VER_05, + RTL_GIGA_MAC_VER_06, + RTL_GIGA_MAC_VER_07, + RTL_GIGA_MAC_VER_08, + RTL_GIGA_MAC_VER_09, + RTL_GIGA_MAC_VER_10, + RTL_GIGA_MAC_VER_11, + RTL_GIGA_MAC_VER_12, + RTL_GIGA_MAC_VER_13, + RTL_GIGA_MAC_VER_14, + RTL_GIGA_MAC_VER_15, + RTL_GIGA_MAC_VER_16, + RTL_GIGA_MAC_VER_17, + RTL_GIGA_MAC_VER_18, + RTL_GIGA_MAC_VER_19, + RTL_GIGA_MAC_VER_20, + RTL_GIGA_MAC_VER_21, + RTL_GIGA_MAC_VER_22, + RTL_GIGA_MAC_VER_23, + RTL_GIGA_MAC_VER_24, + RTL_GIGA_MAC_VER_25, + RTL_GIGA_MAC_VER_26, + RTL_GIGA_MAC_VER_27, + RTL_GIGA_MAC_VER_28, + RTL_GIGA_MAC_VER_29, + RTL_GIGA_MAC_VER_30, + RTL_GIGA_MAC_VER_31, + RTL_GIGA_MAC_VER_32, + RTL_GIGA_MAC_VER_33, + RTL_GIGA_MAC_VER_34, + RTL_GIGA_MAC_VER_35, + RTL_GIGA_MAC_VER_36, + RTL_GIGA_MAC_VER_37, + RTL_GIGA_MAC_VER_38, + RTL_GIGA_MAC_VER_39, + RTL_GIGA_MAC_VER_40, + RTL_GIGA_MAC_VER_41, + RTL_GIGA_MAC_VER_42, + RTL_GIGA_MAC_VER_43, + RTL_GIGA_MAC_VER_44, + RTL_GIGA_MAC_VER_45, + RTL_GIGA_MAC_VER_46, + RTL_GIGA_MAC_VER_47, + RTL_GIGA_MAC_VER_48, + RTL_GIGA_MAC_VER_49, + RTL_GIGA_MAC_VER_50, + RTL_GIGA_MAC_VER_51, + RTL_GIGA_MAC_VER_52, + RTL_GIGA_MAC_VER_60, + RTL_GIGA_MAC_VER_61, + RTL_GIGA_MAC_NONE +}; + +struct rtl8169_private; + +void r8169_apply_firmware(struct rtl8169_private *tp); +u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp); +u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr); +void r8169_hw_phy_config(struct rtl8169_private *tp, struct phy_device *phydev, + enum mac_version ver); diff --git a/drivers/net/ethernet/realtek/r8169_main.c b/drivers/net/ethernet/realtek/r8169_main.c index 67a4d5d45e3a..aaa316be6183 100644 --- a/drivers/net/ethernet/realtek/r8169_main.c +++ b/drivers/net/ethernet/realtek/r8169_main.c @@ -31,6 +31,7 @@ #include <linux/ipv6.h> #include <net/ip6_checksum.h> +#include "r8169.h" #include "r8169_firmware.h" #define MODULENAME "r8169" @@ -84,65 +85,6 @@ #define RTL_R16(tp, reg) readw(tp->mmio_addr + (reg)) #define RTL_R32(tp, reg) readl(tp->mmio_addr + (reg)) -enum mac_version { - /* support for ancient RTL_GIGA_MAC_VER_01 has been removed */ - RTL_GIGA_MAC_VER_02, - RTL_GIGA_MAC_VER_03, - RTL_GIGA_MAC_VER_04, - RTL_GIGA_MAC_VER_05, - RTL_GIGA_MAC_VER_06, - RTL_GIGA_MAC_VER_07, - RTL_GIGA_MAC_VER_08, - RTL_GIGA_MAC_VER_09, - RTL_GIGA_MAC_VER_10, - RTL_GIGA_MAC_VER_11, - RTL_GIGA_MAC_VER_12, - RTL_GIGA_MAC_VER_13, - RTL_GIGA_MAC_VER_14, - RTL_GIGA_MAC_VER_15, - RTL_GIGA_MAC_VER_16, - RTL_GIGA_MAC_VER_17, - RTL_GIGA_MAC_VER_18, - RTL_GIGA_MAC_VER_19, - RTL_GIGA_MAC_VER_20, - RTL_GIGA_MAC_VER_21, - RTL_GIGA_MAC_VER_22, - RTL_GIGA_MAC_VER_23, - RTL_GIGA_MAC_VER_24, - RTL_GIGA_MAC_VER_25, - RTL_GIGA_MAC_VER_26, - RTL_GIGA_MAC_VER_27, - RTL_GIGA_MAC_VER_28, - RTL_GIGA_MAC_VER_29, - RTL_GIGA_MAC_VER_30, - RTL_GIGA_MAC_VER_31, - RTL_GIGA_MAC_VER_32, - RTL_GIGA_MAC_VER_33, - RTL_GIGA_MAC_VER_34, - RTL_GIGA_MAC_VER_35, - RTL_GIGA_MAC_VER_36, - RTL_GIGA_MAC_VER_37, - RTL_GIGA_MAC_VER_38, - RTL_GIGA_MAC_VER_39, - RTL_GIGA_MAC_VER_40, - RTL_GIGA_MAC_VER_41, - RTL_GIGA_MAC_VER_42, - RTL_GIGA_MAC_VER_43, - RTL_GIGA_MAC_VER_44, - RTL_GIGA_MAC_VER_45, - RTL_GIGA_MAC_VER_46, - RTL_GIGA_MAC_VER_47, - RTL_GIGA_MAC_VER_48, - RTL_GIGA_MAC_VER_49, - RTL_GIGA_MAC_VER_50, - RTL_GIGA_MAC_VER_51, - RTL_GIGA_MAC_VER_52, - RTL_GIGA_MAC_VER_60, - RTL_GIGA_MAC_VER_61, - RTL_GIGA_MAC_NONE -}; - -#define JUMBO_1K ETH_DATA_LEN #define JUMBO_4K (4*1024 - ETH_HLEN - 2) #define JUMBO_6K (6*1024 - ETH_HLEN - 2) #define JUMBO_7K (7*1024 - ETH_HLEN - 2) @@ -492,6 +434,7 @@ enum rtl_register_content { /* CPlusCmd p.31 */ EnableBist = (1 << 15), // 8168 8101 Mac_dbgo_oe = (1 << 14), // 8168 8101 + EnAnaPLL = (1 << 14), // 8169 Normal_mode = (1 << 13), // unused Force_half_dup = (1 << 12), // 8168 8101 Force_rxflow_en = (1 << 11), // 8168 8101 @@ -1078,52 +1021,6 @@ static int rtl_readphy(struct rtl8169_private *tp, int location) } } -static void rtl_patchphy(struct rtl8169_private *tp, int reg_addr, int value) -{ - rtl_writephy(tp, reg_addr, rtl_readphy(tp, reg_addr) | value); -} - -static void rtl_w0w1_phy(struct rtl8169_private *tp, int reg_addr, int p, int m) -{ - int val; - - val = rtl_readphy(tp, reg_addr); - rtl_writephy(tp, reg_addr, (val & ~m) | p); -} - -static void r8168d_modify_extpage(struct phy_device *phydev, int extpage, - int reg, u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0007); - - __phy_write(phydev, 0x1e, extpage); - __phy_modify(phydev, reg, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - -static void r8168d_phy_param(struct phy_device *phydev, u16 parm, - u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0005); - - __phy_write(phydev, 0x05, parm); - __phy_modify(phydev, 0x06, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - -static void r8168g_phy_param(struct phy_device *phydev, u16 parm, - u16 mask, u16 val) -{ - int oldpage = phy_select_page(phydev, 0x0a43); - - __phy_write(phydev, 0x13, parm); - __phy_modify(phydev, 0x14, mask, val); - - phy_restore_page(phydev, oldpage, 0); -} - DECLARE_RTL_COND(rtl_ephyar_cond) { return RTL_R32(tp, EPHYAR) & EPHYAR_FLAG; @@ -1372,7 +1269,7 @@ DECLARE_RTL_COND(rtl_efusear_cond) return RTL_R32(tp, EFUSEAR) & EFUSEAR_FLAG; } -static u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr) +u8 rtl8168d_efuse_read(struct rtl8169_private *tp, int reg_addr) { RTL_W32(tp, EFUSEAR, (reg_addr & EFUSEAR_REG_MASK) << EFUSEAR_REG_SHIFT); @@ -1596,7 +1493,7 @@ static netdev_features_t rtl8169_fix_features(struct net_device *dev, if (dev->mtu > TD_MSS_MAX) features &= ~NETIF_F_ALL_TSO; - if (dev->mtu > JUMBO_1K && + if (dev->mtu > ETH_DATA_LEN && tp->mac_version > RTL_GIGA_MAC_VER_06) features &= ~(NETIF_F_CSUM_MASK | NETIF_F_ALL_TSO); @@ -2268,22 +2165,6 @@ static void rtl8169_get_mac_version(struct rtl8169_private *tp) } } -struct phy_reg { - u16 reg; - u16 val; -}; - -static void __rtl_writephy_batch(struct rtl8169_private *tp, - const struct phy_reg *regs, int len) -{ - while (len-- > 0) { - rtl_writephy(tp, regs->reg, regs->val); - regs++; - } -} - -#define rtl_writephy_batch(tp, a) __rtl_writephy_batch(tp, a, ARRAY_SIZE(a)) - static void rtl_release_firmware(struct rtl8169_private *tp) { if (tp->rtl_fw) { @@ -2293,7 +2174,7 @@ static void rtl_release_firmware(struct rtl8169_private *tp) } } -static void rtl_apply_firmware(struct rtl8169_private *tp) +void r8169_apply_firmware(struct rtl8169_private *tp) { /* TODO: release firmware if rtl_fw_write_firmware signals failure. */ if (tp->rtl_fw) @@ -2315,594 +2196,6 @@ static void rtl8125_config_eee_mac(struct rtl8169_private *tp) r8168_mac_ocp_modify(tp, 0xeb62, 0, BIT(2) | BIT(1)); } -static void rtl8168f_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - r8168d_modify_extpage(phydev, 0x0020, 0x15, 0, BIT(8)); - r8168d_phy_param(phydev, 0x8b85, 0, BIT(13)); -} - -static void rtl8168g_config_eee_phy(struct rtl8169_private *tp) -{ - phy_modify_paged(tp->phydev, 0x0a43, 0x11, 0, BIT(4)); -} - -static void rtl8168h_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168g_config_eee_phy(tp); - - phy_modify_paged(phydev, 0xa4a, 0x11, 0x0000, 0x0200); - phy_modify_paged(phydev, 0xa42, 0x14, 0x0000, 0x0080); -} - -static void rtl8125_config_eee_phy(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168h_config_eee_phy(tp); - - phy_modify_paged(phydev, 0xa6d, 0x12, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa6d, 0x14, 0x0010, 0x0000); -} - -static void rtl8169s_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x06, 0x006e }, - { 0x08, 0x0708 }, - { 0x15, 0x4000 }, - { 0x18, 0x65c7 }, - - { 0x1f, 0x0001 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x0000 }, - - { 0x03, 0xff41 }, - { 0x02, 0xdf60 }, - { 0x01, 0x0140 }, - { 0x00, 0x0077 }, - { 0x04, 0x7800 }, - { 0x04, 0x7000 }, - - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf0f9 }, - { 0x04, 0x9800 }, - { 0x04, 0x9000 }, - - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xa000 }, - - { 0x03, 0xff41 }, - { 0x02, 0xdf20 }, - { 0x01, 0x0140 }, - { 0x00, 0x00bb }, - { 0x04, 0xb800 }, - { 0x04, 0xb000 }, - - { 0x03, 0xdf41 }, - { 0x02, 0xdc60 }, - { 0x01, 0x6340 }, - { 0x00, 0x007d }, - { 0x04, 0xd800 }, - { 0x04, 0xd000 }, - - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x100a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0xf000 }, - - { 0x1f, 0x0000 }, - { 0x0b, 0x0000 }, - { 0x00, 0x9200 } - }; - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8169sb_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0002, 0x01, 0x90d0); -} - -static void rtl8169scd_hw_phy_config_quirk(struct rtl8169_private *tp) -{ - struct pci_dev *pdev = tp->pci_dev; - - if ((pdev->subsystem_vendor != PCI_VENDOR_ID_GIGABYTE) || - (pdev->subsystem_device != 0xe000)) - return; - - phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf01b); -} - -static void rtl8169scd_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x04, 0x0000 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x9000 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0xa000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0x0000 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x10, 0xf41b }, - { 0x14, 0xfb54 }, - { 0x18, 0xf5c7 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl8169scd_hw_phy_config_quirk(tp); -} - -static void rtl8169sce_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x04, 0x0000 }, - { 0x03, 0x00a1 }, - { 0x02, 0x0008 }, - { 0x01, 0x0120 }, - { 0x00, 0x1000 }, - { 0x04, 0x0800 }, - { 0x04, 0x9000 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0xa000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0xff95 }, - { 0x00, 0xba00 }, - { 0x04, 0xa800 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0x0000 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x0b, 0x8480 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x18, 0x67c7 }, - { 0x04, 0x2000 }, - { 0x03, 0x002f }, - { 0x02, 0x4360 }, - { 0x01, 0x0109 }, - { 0x00, 0x3022 }, - { 0x04, 0x2800 }, - { 0x1f, 0x0000 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8168bb_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy(tp, 0x1f, 0x0001); - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_writephy(tp, 0x10, 0xf41b); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168bef_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf41b); -} - -static void rtl8168cp_1_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write(tp->phydev, 0x1d, 0x0f00); - phy_write_paged(tp->phydev, 0x0002, 0x0c, 0x1ec8); -} - -static void rtl8168cp_2_hw_phy_config(struct rtl8169_private *tp) -{ - phy_set_bits(tp->phydev, 0x14, BIT(5)); - phy_set_bits(tp->phydev, 0x0d, BIT(5)); - phy_write_paged(tp->phydev, 0x0001, 0x1d, 0x3d98); -} - -static void rtl8168c_1_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x1f, 0x0002 }, - { 0x00, 0x88d4 }, - { 0x01, 0x82b1 }, - { 0x03, 0x7002 }, - { 0x08, 0x9e30 }, - { 0x09, 0x01f0 }, - { 0x0a, 0x5500 }, - { 0x0c, 0x00c8 }, - { 0x1f, 0x0003 }, - { 0x12, 0xc096 }, - { 0x16, 0x000a }, - { 0x1f, 0x0000 }, - { 0x1f, 0x0000 }, - { 0x09, 0x2000 }, - { 0x09, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168c_2_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x03, 0x802f }, - { 0x02, 0x4f02 }, - { 0x01, 0x0409 }, - { 0x00, 0xf099 }, - { 0x04, 0x9800 }, - { 0x04, 0x9000 }, - { 0x1d, 0x3d98 }, - { 0x1f, 0x0002 }, - { 0x0c, 0x7eb8 }, - { 0x06, 0x0761 }, - { 0x1f, 0x0003 }, - { 0x16, 0x0f0a }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168c_3_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0001 }, - { 0x12, 0x2300 }, - { 0x1d, 0x3d98 }, - { 0x1f, 0x0002 }, - { 0x0c, 0x7eb8 }, - { 0x06, 0x5461 }, - { 0x1f, 0x0003 }, - { 0x16, 0x0f0a }, - { 0x1f, 0x0000 } - }; - - rtl_writephy_batch(tp, phy_reg_init); - - rtl_patchphy(tp, 0x16, 1 << 0); - rtl_patchphy(tp, 0x14, 1 << 5); - rtl_patchphy(tp, 0x0d, 1 << 5); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static const struct phy_reg rtl8168d_1_phy_reg_init_0[] = { - /* Channel Estimation */ - { 0x1f, 0x0001 }, - { 0x06, 0x4064 }, - { 0x07, 0x2863 }, - { 0x08, 0x059c }, - { 0x09, 0x26b4 }, - { 0x0a, 0x6a19 }, - { 0x0b, 0xdcc8 }, - { 0x10, 0xf06d }, - { 0x14, 0x7f68 }, - { 0x18, 0x7fd9 }, - { 0x1c, 0xf0ff }, - { 0x1d, 0x3d9c }, - { 0x1f, 0x0003 }, - { 0x12, 0xf49f }, - { 0x13, 0x070b }, - { 0x1a, 0x05ad }, - { 0x14, 0x94c0 }, - - /* - * Tx Error Issue - * Enhance line driver power - */ - { 0x1f, 0x0002 }, - { 0x06, 0x5561 }, - { 0x1f, 0x0005 }, - { 0x05, 0x8332 }, - { 0x06, 0x5561 }, - - /* - * Can not link to 1Gbps with bad cable - * Decrease SNR threshold form 21.07dB to 19.04dB - */ - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - - { 0x1f, 0x0000 }, - { 0x0d, 0xf880 } -}; - -static const struct phy_reg rtl8168d_1_phy_reg_init_1[] = { - { 0x1f, 0x0002 }, - { 0x05, 0x669a }, - { 0x1f, 0x0005 }, - { 0x05, 0x8330 }, - { 0x06, 0x669a }, - { 0x1f, 0x0002 } -}; - -static void rtl8168d_apply_firmware_cond(struct rtl8169_private *tp, u16 val) -{ - u16 reg_val; - - rtl_writephy(tp, 0x1f, 0x0005); - rtl_writephy(tp, 0x05, 0x001b); - reg_val = rtl_readphy(tp, 0x06); - rtl_writephy(tp, 0x1f, 0x0000); - - if (reg_val != val) - netif_warn(tp, hw, tp->dev, "chipset not ready for firmware\n"); - else - rtl_apply_firmware(tp); -} - -static void rtl8168d_1_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_0); - - /* - * Rx Error Issue - * Fine Tune Switching regulator parameter - */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x0b, 0x0010, 0x00ef); - rtl_w0w1_phy(tp, 0x0c, 0xa200, 0x5d00); - - if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { - int val; - - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_1); - - val = rtl_readphy(tp, 0x0d); - - if ((val & 0x00ff) != 0x006c) { - static const u32 set[] = { - 0x0065, 0x0066, 0x0067, 0x0068, - 0x0069, 0x006a, 0x006b, 0x006c - }; - int i; - - rtl_writephy(tp, 0x1f, 0x0002); - - val &= 0xff00; - for (i = 0; i < ARRAY_SIZE(set); i++) - rtl_writephy(tp, 0x0d, val | set[i]); - } - } else { - phy_write_paged(tp->phydev, 0x0002, 0x05, 0x6662); - r8168d_phy_param(tp->phydev, 0x8330, 0xffff, 0x6662); - } - - /* RSET couple improve */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_patchphy(tp, 0x0d, 0x0300); - rtl_patchphy(tp, 0x0f, 0x0010); - - /* Fine tune PLL performance */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x02, 0x0100, 0x0600); - rtl_w0w1_phy(tp, 0x03, 0x0000, 0xe000); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168d_apply_firmware_cond(tp, 0xbf00); -} - -static void rtl8168d_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_0); - - if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { - int val; - - rtl_writephy_batch(tp, rtl8168d_1_phy_reg_init_1); - - val = rtl_readphy(tp, 0x0d); - if ((val & 0x00ff) != 0x006c) { - static const u32 set[] = { - 0x0065, 0x0066, 0x0067, 0x0068, - 0x0069, 0x006a, 0x006b, 0x006c - }; - int i; - - rtl_writephy(tp, 0x1f, 0x0002); - - val &= 0xff00; - for (i = 0; i < ARRAY_SIZE(set); i++) - rtl_writephy(tp, 0x0d, val | set[i]); - } - } else { - phy_write_paged(tp->phydev, 0x0002, 0x05, 0x2642); - r8168d_phy_param(tp->phydev, 0x8330, 0xffff, 0x2642); - } - - /* Fine tune PLL performance */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_w0w1_phy(tp, 0x02, 0x0100, 0x0600); - rtl_w0w1_phy(tp, 0x03, 0x0000, 0xe000); - - /* Switching regulator Slew rate */ - rtl_writephy(tp, 0x1f, 0x0002); - rtl_patchphy(tp, 0x0f, 0x0017); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168d_apply_firmware_cond(tp, 0xb300); -} - -static void rtl8168d_3_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0002 }, - { 0x10, 0x0008 }, - { 0x0d, 0x006c }, - - { 0x1f, 0x0000 }, - { 0x0d, 0xf880 }, - - { 0x1f, 0x0001 }, - { 0x17, 0x0cc0 }, - - { 0x1f, 0x0001 }, - { 0x0b, 0xa4d8 }, - { 0x09, 0x281c }, - { 0x07, 0x2883 }, - { 0x0a, 0x6b35 }, - { 0x1d, 0x3da4 }, - { 0x1c, 0xeffd }, - { 0x14, 0x7f52 }, - { 0x18, 0x7fc6 }, - { 0x08, 0x0601 }, - { 0x06, 0x4063 }, - { 0x10, 0xf074 }, - { 0x1f, 0x0003 }, - { 0x13, 0x0789 }, - { 0x12, 0xf4bd }, - { 0x1a, 0x04fd }, - { 0x14, 0x84b0 }, - { 0x1f, 0x0000 }, - { 0x00, 0x9200 }, - - { 0x1f, 0x0005 }, - { 0x01, 0x0340 }, - { 0x1f, 0x0001 }, - { 0x04, 0x4000 }, - { 0x03, 0x1d21 }, - { 0x02, 0x0c32 }, - { 0x01, 0x0200 }, - { 0x00, 0x5554 }, - { 0x04, 0x4800 }, - { 0x04, 0x4000 }, - { 0x04, 0xf000 }, - { 0x03, 0xdf01 }, - { 0x02, 0xdf20 }, - { 0x01, 0x101a }, - { 0x00, 0xa0ff }, - { 0x04, 0xf800 }, - { 0x04, 0xf000 }, - { 0x1f, 0x0000 }, - }; - - rtl_writephy_batch(tp, phy_reg_init); - - r8168d_modify_extpage(tp->phydev, 0x0023, 0x16, 0xffff, 0x0000); -} - -static void rtl8168d_4_hw_phy_config(struct rtl8169_private *tp) -{ - phy_write_paged(tp->phydev, 0x0001, 0x17, 0x0cc0); - r8168d_modify_extpage(tp->phydev, 0x002d, 0x18, 0xffff, 0x0040); - phy_set_bits(tp->phydev, 0x0d, BIT(5)); -} - -static void rtl8168e_1_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - /* Channel estimation fine tune */ - { 0x1f, 0x0001 }, - { 0x0b, 0x6c20 }, - { 0x07, 0x2872 }, - { 0x1c, 0xefff }, - { 0x1f, 0x0003 }, - { 0x14, 0x6420 }, - { 0x1f, 0x0000 }, - }; - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Enable Delay cap */ - r8168d_phy_param(phydev, 0x8b80, 0xffff, 0xc896); - - rtl_writephy_batch(tp, phy_reg_init); - - /* Update PFM & 10M TX idle timer */ - r8168d_modify_extpage(phydev, 0x002f, 0x15, 0xffff, 0x1919); - - r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); - - /* DCO enable for 10M IDLE Power */ - r8168d_modify_extpage(phydev, 0x0023, 0x17, 0x0000, 0x0006); - - /* For impedance matching */ - phy_modify_paged(phydev, 0x0002, 0x08, 0x7f00, 0x8000); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0050); - phy_set_bits(phydev, 0x14, BIT(15)); - - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - r8168d_phy_param(phydev, 0x8b85, 0x2000, 0x0000); - - r8168d_modify_extpage(phydev, 0x0020, 0x15, 0x1100, 0x0000); - phy_write_paged(phydev, 0x0006, 0x00, 0x5a00); - - phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0x0000); -} - static void rtl_rar_exgmac_set(struct rtl8169_private *tp, u8 *addr) { const u16 w[] = { @@ -2917,698 +2210,20 @@ static void rtl_rar_exgmac_set(struct rtl8169_private *tp, u8 *addr) rtl_eri_write(tp, 0xf4, ERIAR_MASK_1111, w[1] | (w[2] << 16)); } -static void rtl8168e_2_hw_phy_config(struct rtl8169_private *tp) +u16 rtl8168h_2_get_adc_bias_ioffset(struct rtl8169_private *tp) { - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Enable Delay cap */ - r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Green Setting */ - r8168d_phy_param(phydev, 0x8b5b, 0xffff, 0x9222); - r8168d_phy_param(phydev, 0x8b6d, 0xffff, 0x8000); - r8168d_phy_param(phydev, 0x8b76, 0xffff, 0x8000); - - /* For 4-corner performance improve */ - rtl_writephy(tp, 0x1f, 0x0005); - rtl_writephy(tp, 0x05, 0x8b80); - rtl_w0w1_phy(tp, 0x17, 0x0006, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); - phy_set_bits(phydev, 0x14, BIT(15)); - - /* improve 10M EEE waveform */ - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); - - rtl8168f_config_eee_phy(tp); - rtl_enable_eee(tp); - - /* Green feature */ - rtl_writephy(tp, 0x1f, 0x0003); - rtl_w0w1_phy(tp, 0x19, 0x0001, 0x0000); - rtl_w0w1_phy(tp, 0x10, 0x0400, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); - rtl_writephy(tp, 0x1f, 0x0005); - rtl_w0w1_phy(tp, 0x01, 0x0100, 0x0000); - rtl_writephy(tp, 0x1f, 0x0000); - - /* Broken BIOS workaround: feed GigaMAC registers with MAC address. */ - rtl_rar_exgmac_set(tp, tp->dev->dev_addr); -} - -static void rtl8168f_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* For 4-corner performance improve */ - r8168d_phy_param(phydev, 0x8b80, 0x0000, 0x0006); - - /* PHY auto speed down */ - r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); - phy_set_bits(phydev, 0x14, BIT(15)); - - /* Improve 10M EEE waveform */ - r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); - - rtl8168f_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168f_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Modify green table for giga & fnet */ - r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00fb); - - /* Modify green table for 10M */ - r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); - - /* Disable hiimpedance detection (RTCT) */ - phy_write_paged(phydev, 0x0003, 0x01, 0x328a); - - rtl8168f_hw_phy_config(tp); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); -} - -static void rtl8168f_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_apply_firmware(tp); - - rtl8168f_hw_phy_config(tp); -} - -static void rtl8411_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl_apply_firmware(tp); - - rtl8168f_hw_phy_config(tp); - - /* Improve 2-pair detection performance */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); - - /* Channel estimation fine tune */ - phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); - - /* Modify green table for giga & fnet */ - r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); - r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); - r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00aa); - - /* Modify green table for 10M */ - r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); - - /* Disable hiimpedance detection (RTCT) */ - phy_write_paged(phydev, 0x0003, 0x01, 0x328a); - - /* Modify green table for giga */ - r8168d_phy_param(phydev, 0x8b54, 0x0800, 0x0000); - r8168d_phy_param(phydev, 0x8b5d, 0x0800, 0x0000); - r8168d_phy_param(phydev, 0x8a7c, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a7f, 0x0000, 0x0100); - r8168d_phy_param(phydev, 0x8a82, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a85, 0x0100, 0x0000); - r8168d_phy_param(phydev, 0x8a88, 0x0100, 0x0000); - - /* uc same-seed solution */ - r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x8000); - - /* Green feature */ - rtl_writephy(tp, 0x1f, 0x0003); - rtl_w0w1_phy(tp, 0x19, 0x0000, 0x0001); - rtl_w0w1_phy(tp, 0x10, 0x0000, 0x0400); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8168g_disable_aldps(struct rtl8169_private *tp) -{ - phy_modify_paged(tp->phydev, 0x0a43, 0x10, BIT(2), 0); -} - -static void rtl8168g_phy_adjust_10m_aldps(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - phy_modify_paged(phydev, 0x0bcc, 0x14, BIT(8), 0); - phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(7) | BIT(6)); - r8168g_phy_param(phydev, 0x8084, 0x6000, 0x0000); - phy_modify_paged(phydev, 0x0a43, 0x10, 0x0000, 0x1003); -} - -static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp) -{ - int ret; - - rtl_apply_firmware(tp); - - ret = phy_read_paged(tp->phydev, 0x0a46, 0x10); - if (ret & BIT(8)) - phy_modify_paged(tp->phydev, 0x0bcc, 0x12, BIT(15), 0); - else - phy_modify_paged(tp->phydev, 0x0bcc, 0x12, 0, BIT(15)); - - ret = phy_read_paged(tp->phydev, 0x0a46, 0x13); - if (ret & BIT(8)) - phy_modify_paged(tp->phydev, 0x0c41, 0x15, 0, BIT(1)); - else - phy_modify_paged(tp->phydev, 0x0c41, 0x15, BIT(1), 0); - - /* Enable PHY auto speed down */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* EEE auto-fallback function */ - phy_modify_paged(tp->phydev, 0x0a4b, 0x11, 0, BIT(2)); - - /* Enable UC LPF tune function */ - r8168g_phy_param(tp->phydev, 0x8012, 0x0000, 0x8000); - - phy_modify_paged(tp->phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - /* Improve SWR Efficiency */ - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x5065); - rtl_writephy(tp, 0x14, 0xd065); - rtl_writephy(tp, 0x1f, 0x0bc8); - rtl_writephy(tp, 0x11, 0x5655); - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x14, 0x9065); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168g_2_hw_phy_config(struct rtl8169_private *tp) -{ - rtl_apply_firmware(tp); - rtl8168g_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - u16 dout_tapbin; - u32 data; - - rtl_apply_firmware(tp); - - /* CHN EST parameters adjust - giga master */ - r8168g_phy_param(phydev, 0x809b, 0xf800, 0x8000); - r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x8000); - r8168g_phy_param(phydev, 0x80a4, 0xff00, 0x8500); - r8168g_phy_param(phydev, 0x809c, 0xff00, 0xbd00); - - /* CHN EST parameters adjust - giga slave */ - r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x7000); - r8168g_phy_param(phydev, 0x80b4, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x80ac, 0xff00, 0x4000); - - /* CHN EST parameters adjust - fnet */ - r8168g_phy_param(phydev, 0x808e, 0xff00, 0x1200); - r8168g_phy_param(phydev, 0x8090, 0xff00, 0xe500); - r8168g_phy_param(phydev, 0x8092, 0xff00, 0x9f00); - - /* enable R-tune & PGA-retune function */ - dout_tapbin = 0; - data = phy_read_paged(phydev, 0x0a46, 0x13); - data &= 3; - data <<= 2; - dout_tapbin |= data; - data = phy_read_paged(phydev, 0x0a46, 0x12); - data &= 0xc000; - data >>= 14; - dout_tapbin |= data; - dout_tapbin = ~(dout_tapbin^0x08); - dout_tapbin <<= 12; - dout_tapbin &= 0xf000; - - r8168g_phy_param(phydev, 0x827a, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827b, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827c, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x827d, 0xf000, dout_tapbin); - r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); - phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); - - /* SAR ADC performance */ - phy_modify_paged(tp->phydev, 0x0bca, 0x17, BIT(12) | BIT(13), BIT(14)); - - r8168g_phy_param(phydev, 0x803f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8047, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x804f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8057, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x805f, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x8067, 0x3000, 0x0000); - r8168g_phy_param(phydev, 0x806f, 0x3000, 0x0000); - - /* disable phy pfm mode */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, BIT(7), 0); - - rtl8168g_disable_aldps(tp); - rtl8168h_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168h_2_hw_phy_config(struct rtl8169_private *tp) -{ - u16 ioffset_p3, ioffset_p2, ioffset_p1, ioffset_p0; - struct phy_device *phydev = tp->phydev; - u16 rlen; - u32 data; - - rtl_apply_firmware(tp); - - /* CHIN EST parameter update */ - r8168g_phy_param(phydev, 0x808a, 0x003f, 0x000a); - - /* enable R-tune & PGA-retune function */ - r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); - phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); + u16 data1, data2, ioffset; r8168_mac_ocp_write(tp, 0xdd02, 0x807d); - data = r8168_mac_ocp_read(tp, 0xdd02); - ioffset_p3 = ((data & 0x80)>>7); - ioffset_p3 <<= 3; - - data = r8168_mac_ocp_read(tp, 0xdd00); - ioffset_p3 |= ((data & (0xe000))>>13); - ioffset_p2 = ((data & (0x1e00))>>9); - ioffset_p1 = ((data & (0x01e0))>>5); - ioffset_p0 = ((data & 0x0010)>>4); - ioffset_p0 <<= 3; - ioffset_p0 |= (data & (0x07)); - data = (ioffset_p3<<12)|(ioffset_p2<<8)|(ioffset_p1<<4)|(ioffset_p0); - - if ((ioffset_p3 != 0x0f) || (ioffset_p2 != 0x0f) || - (ioffset_p1 != 0x0f) || (ioffset_p0 != 0x0f)) - phy_write_paged(phydev, 0x0bcf, 0x16, data); - - /* Modify rlen (TX LPF corner frequency) level */ - data = phy_read_paged(phydev, 0x0bcd, 0x16); - data &= 0x000f; - rlen = 0; - if (data > 3) - rlen = data - 3; - data = rlen | (rlen<<4) | (rlen<<8) | (rlen<<12); - phy_write_paged(phydev, 0x0bcd, 0x17, data); - - /* disable phy pfm mode */ - phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168ep_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* Enable PHY auto speed down */ - phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* Enable EEE auto-fallback function */ - phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); - - /* Enable UC LPF tune function */ - r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); - - /* set rg_sel_sdm_rate */ - phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8168ep_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - rtl8168g_phy_adjust_10m_aldps(tp); - - /* Enable UC LPF tune function */ - r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); - - /* Set rg_sel_sdm_rate */ - phy_modify_paged(tp->phydev, 0x0c42, 0x11, BIT(13), BIT(14)); - - /* Channel estimation parameters */ - r8168g_phy_param(phydev, 0x80f3, 0xff00, 0x8b00); - r8168g_phy_param(phydev, 0x80f0, 0xff00, 0x3a00); - r8168g_phy_param(phydev, 0x80ef, 0xff00, 0x0500); - r8168g_phy_param(phydev, 0x80f6, 0xff00, 0x6e00); - r8168g_phy_param(phydev, 0x80ec, 0xff00, 0x6800); - r8168g_phy_param(phydev, 0x80ed, 0xff00, 0x7c00); - r8168g_phy_param(phydev, 0x80f2, 0xff00, 0xf400); - r8168g_phy_param(phydev, 0x80f4, 0xff00, 0x8500); - r8168g_phy_param(phydev, 0x8110, 0xff00, 0xa800); - r8168g_phy_param(phydev, 0x810f, 0xff00, 0x1d00); - r8168g_phy_param(phydev, 0x8111, 0xff00, 0xf500); - r8168g_phy_param(phydev, 0x8113, 0xff00, 0x6100); - r8168g_phy_param(phydev, 0x8115, 0xff00, 0x9200); - r8168g_phy_param(phydev, 0x810e, 0xff00, 0x0400); - r8168g_phy_param(phydev, 0x810c, 0xff00, 0x7c00); - r8168g_phy_param(phydev, 0x810b, 0xff00, 0x5a00); - r8168g_phy_param(phydev, 0x80d1, 0xff00, 0xff00); - r8168g_phy_param(phydev, 0x80cd, 0xff00, 0x9e00); - r8168g_phy_param(phydev, 0x80d3, 0xff00, 0x0e00); - r8168g_phy_param(phydev, 0x80d5, 0xff00, 0xca00); - r8168g_phy_param(phydev, 0x80d7, 0xff00, 0x8400); - - /* Force PWM-mode */ - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x5065); - rtl_writephy(tp, 0x14, 0xd065); - rtl_writephy(tp, 0x1f, 0x0bc8); - rtl_writephy(tp, 0x12, 0x00ed); - rtl_writephy(tp, 0x1f, 0x0bcd); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x14, 0x9065); - rtl_writephy(tp, 0x14, 0x1065); - rtl_writephy(tp, 0x1f, 0x0000); - - rtl8168g_disable_aldps(tp); - rtl8168g_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8117_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - /* CHN EST parameters adjust - fnet */ - r8168g_phy_param(phydev, 0x808e, 0xff00, 0x4800); - r8168g_phy_param(phydev, 0x8090, 0xff00, 0xcc00); - r8168g_phy_param(phydev, 0x8092, 0xff00, 0xb000); - - r8168g_phy_param(phydev, 0x8088, 0xff00, 0x6000); - r8168g_phy_param(phydev, 0x808b, 0x3f00, 0x0b00); - r8168g_phy_param(phydev, 0x808d, 0x1f00, 0x0600); - r8168g_phy_param(phydev, 0x808c, 0xff00, 0xb000); - r8168g_phy_param(phydev, 0x80a0, 0xff00, 0x2800); - r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x809b, 0xf800, 0xb000); - r8168g_phy_param(phydev, 0x809a, 0xff00, 0x4b00); - r8168g_phy_param(phydev, 0x809d, 0x3f00, 0x0800); - r8168g_phy_param(phydev, 0x80a1, 0xff00, 0x7000); - r8168g_phy_param(phydev, 0x809f, 0x1f00, 0x0300); - r8168g_phy_param(phydev, 0x809e, 0xff00, 0x8800); - r8168g_phy_param(phydev, 0x80b2, 0xff00, 0x2200); - r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x9800); - r8168g_phy_param(phydev, 0x80af, 0x3f00, 0x0800); - r8168g_phy_param(phydev, 0x80b3, 0xff00, 0x6f00); - r8168g_phy_param(phydev, 0x80b1, 0x1f00, 0x0300); - r8168g_phy_param(phydev, 0x80b0, 0xff00, 0x9300); - - r8168g_phy_param(phydev, 0x8011, 0x0000, 0x0800); - - /* enable GPHY 10M */ - phy_modify_paged(tp->phydev, 0x0a44, 0x11, 0, BIT(11)); - - r8168g_phy_param(phydev, 0x8016, 0x0000, 0x0400); - - rtl8168g_disable_aldps(tp); - rtl8168h_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8102e_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0003 }, - { 0x08, 0x441d }, - { 0x01, 0x9100 }, - { 0x1f, 0x0000 } - }; - - rtl_writephy(tp, 0x1f, 0x0000); - rtl_patchphy(tp, 0x11, 1 << 12); - rtl_patchphy(tp, 0x19, 1 << 13); - rtl_patchphy(tp, 0x10, 1 << 15); - - rtl_writephy_batch(tp, phy_reg_init); -} - -static void rtl8105e_hw_phy_config(struct rtl8169_private *tp) -{ - /* Disable ALDPS before ram code */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(100); - - rtl_apply_firmware(tp); - - phy_write_paged(tp->phydev, 0x0005, 0x1a, 0x0000); - phy_write_paged(tp->phydev, 0x0004, 0x1c, 0x0000); - phy_write_paged(tp->phydev, 0x0001, 0x15, 0x7701); -} - -static void rtl8402_hw_phy_config(struct rtl8169_private *tp) -{ - /* Disable ALDPS before setting firmware */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(20); - - rtl_apply_firmware(tp); - - /* EEE setting */ - rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); - rtl_writephy(tp, 0x1f, 0x0004); - rtl_writephy(tp, 0x10, 0x401f); - rtl_writephy(tp, 0x19, 0x7030); - rtl_writephy(tp, 0x1f, 0x0000); -} - -static void rtl8106e_hw_phy_config(struct rtl8169_private *tp) -{ - static const struct phy_reg phy_reg_init[] = { - { 0x1f, 0x0004 }, - { 0x10, 0xc07f }, - { 0x19, 0x7030 }, - { 0x1f, 0x0000 } - }; - - /* Disable ALDPS before ram code */ - phy_write(tp->phydev, 0x18, 0x0310); - msleep(100); - - rtl_apply_firmware(tp); - - rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); - rtl_writephy_batch(tp, phy_reg_init); - - rtl_eri_write(tp, 0x1d0, ERIAR_MASK_0011, 0x0000); -} - -static void rtl8125_1_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - - phy_modify_paged(phydev, 0xad4, 0x10, 0x03ff, 0x0084); - phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); - phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x0006); - phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); - phy_modify_paged(phydev, 0xac0, 0x14, 0x0000, 0x1100); - phy_modify_paged(phydev, 0xac8, 0x15, 0xf000, 0x7000); - phy_modify_paged(phydev, 0xad1, 0x14, 0x0000, 0x0400); - phy_modify_paged(phydev, 0xad1, 0x15, 0x0000, 0x03ff); - phy_modify_paged(phydev, 0xad1, 0x16, 0x0000, 0x03ff); - - r8168g_phy_param(phydev, 0x80ea, 0xff00, 0xc400); - r8168g_phy_param(phydev, 0x80eb, 0x0700, 0x0300); - r8168g_phy_param(phydev, 0x80f8, 0xff00, 0x1c00); - r8168g_phy_param(phydev, 0x80f1, 0xff00, 0x3000); - r8168g_phy_param(phydev, 0x80fe, 0xff00, 0xa500); - r8168g_phy_param(phydev, 0x8102, 0xff00, 0x5000); - r8168g_phy_param(phydev, 0x8105, 0xff00, 0x3300); - r8168g_phy_param(phydev, 0x8100, 0xff00, 0x7000); - r8168g_phy_param(phydev, 0x8104, 0xff00, 0xf000); - r8168g_phy_param(phydev, 0x8106, 0xff00, 0x6500); - r8168g_phy_param(phydev, 0x80dc, 0xff00, 0xed00); - r8168g_phy_param(phydev, 0x80df, 0x0000, 0x0100); - r8168g_phy_param(phydev, 0x80e1, 0x0100, 0x0000); - - phy_modify_paged(phydev, 0xbf0, 0x13, 0x003f, 0x0038); - r8168g_phy_param(phydev, 0x819f, 0xffff, 0xd0b6); - - phy_write_paged(phydev, 0xbc3, 0x12, 0x5555); - phy_modify_paged(phydev, 0xbf0, 0x15, 0x0e00, 0x0a00); - phy_modify_paged(phydev, 0xa5c, 0x10, 0x0400, 0x0000); - phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); - - rtl8125_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl8125_2_hw_phy_config(struct rtl8169_private *tp) -{ - struct phy_device *phydev = tp->phydev; - int i; + data1 = r8168_mac_ocp_read(tp, 0xdd02); + data2 = r8168_mac_ocp_read(tp, 0xdd00); - phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); - phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x03ff); - phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); - phy_modify_paged(phydev, 0xac0, 0x14, 0x1100, 0x0000); - phy_modify_paged(phydev, 0xacc, 0x10, 0x0003, 0x0002); - phy_modify_paged(phydev, 0xad4, 0x10, 0x00e7, 0x0044); - phy_modify_paged(phydev, 0xac1, 0x12, 0x0080, 0x0000); - phy_modify_paged(phydev, 0xac8, 0x10, 0x0300, 0x0000); - phy_modify_paged(phydev, 0xac5, 0x17, 0x0007, 0x0002); - phy_write_paged(phydev, 0xad4, 0x16, 0x00a8); - phy_write_paged(phydev, 0xac5, 0x16, 0x01ff); - phy_modify_paged(phydev, 0xac8, 0x15, 0x00f0, 0x0030); - - phy_write(phydev, 0x1f, 0x0b87); - phy_write(phydev, 0x16, 0x80a2); - phy_write(phydev, 0x17, 0x0153); - phy_write(phydev, 0x16, 0x809c); - phy_write(phydev, 0x17, 0x0153); - phy_write(phydev, 0x1f, 0x0000); - - phy_write(phydev, 0x1f, 0x0a43); - phy_write(phydev, 0x13, 0x81B3); - phy_write(phydev, 0x14, 0x0043); - phy_write(phydev, 0x14, 0x00A7); - phy_write(phydev, 0x14, 0x00D6); - phy_write(phydev, 0x14, 0x00EC); - phy_write(phydev, 0x14, 0x00F6); - phy_write(phydev, 0x14, 0x00FB); - phy_write(phydev, 0x14, 0x00FD); - phy_write(phydev, 0x14, 0x00FF); - phy_write(phydev, 0x14, 0x00BB); - phy_write(phydev, 0x14, 0x0058); - phy_write(phydev, 0x14, 0x0029); - phy_write(phydev, 0x14, 0x0013); - phy_write(phydev, 0x14, 0x0009); - phy_write(phydev, 0x14, 0x0004); - phy_write(phydev, 0x14, 0x0002); - for (i = 0; i < 25; i++) - phy_write(phydev, 0x14, 0x0000); - phy_write(phydev, 0x1f, 0x0000); - - r8168g_phy_param(phydev, 0x8257, 0xffff, 0x020F); - r8168g_phy_param(phydev, 0x80ea, 0xffff, 0x7843); - - rtl_apply_firmware(tp); - - phy_modify_paged(phydev, 0xd06, 0x14, 0x0000, 0x2000); - - r8168g_phy_param(phydev, 0x81a2, 0x0000, 0x0100); - - phy_modify_paged(phydev, 0xb54, 0x16, 0xff00, 0xdb00); - phy_modify_paged(phydev, 0xa45, 0x12, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa5d, 0x12, 0x0000, 0x0020); - phy_modify_paged(phydev, 0xad4, 0x17, 0x0010, 0x0000); - phy_modify_paged(phydev, 0xa86, 0x15, 0x0001, 0x0000); - phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); - - rtl8125_config_eee_phy(tp); - rtl_enable_eee(tp); -} - -static void rtl_hw_phy_config(struct net_device *dev) -{ - static const rtl_generic_fct phy_configs[] = { - /* PCI devices. */ - [RTL_GIGA_MAC_VER_02] = rtl8169s_hw_phy_config, - [RTL_GIGA_MAC_VER_03] = rtl8169s_hw_phy_config, - [RTL_GIGA_MAC_VER_04] = rtl8169sb_hw_phy_config, - [RTL_GIGA_MAC_VER_05] = rtl8169scd_hw_phy_config, - [RTL_GIGA_MAC_VER_06] = rtl8169sce_hw_phy_config, - /* PCI-E devices. */ - [RTL_GIGA_MAC_VER_07] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_08] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_09] = rtl8102e_hw_phy_config, - [RTL_GIGA_MAC_VER_10] = NULL, - [RTL_GIGA_MAC_VER_11] = rtl8168bb_hw_phy_config, - [RTL_GIGA_MAC_VER_12] = rtl8168bef_hw_phy_config, - [RTL_GIGA_MAC_VER_13] = NULL, - [RTL_GIGA_MAC_VER_14] = NULL, - [RTL_GIGA_MAC_VER_15] = NULL, - [RTL_GIGA_MAC_VER_16] = NULL, - [RTL_GIGA_MAC_VER_17] = rtl8168bef_hw_phy_config, - [RTL_GIGA_MAC_VER_18] = rtl8168cp_1_hw_phy_config, - [RTL_GIGA_MAC_VER_19] = rtl8168c_1_hw_phy_config, - [RTL_GIGA_MAC_VER_20] = rtl8168c_2_hw_phy_config, - [RTL_GIGA_MAC_VER_21] = rtl8168c_3_hw_phy_config, - [RTL_GIGA_MAC_VER_22] = rtl8168c_3_hw_phy_config, - [RTL_GIGA_MAC_VER_23] = rtl8168cp_2_hw_phy_config, - [RTL_GIGA_MAC_VER_24] = rtl8168cp_2_hw_phy_config, - [RTL_GIGA_MAC_VER_25] = rtl8168d_1_hw_phy_config, - [RTL_GIGA_MAC_VER_26] = rtl8168d_2_hw_phy_config, - [RTL_GIGA_MAC_VER_27] = rtl8168d_3_hw_phy_config, - [RTL_GIGA_MAC_VER_28] = rtl8168d_4_hw_phy_config, - [RTL_GIGA_MAC_VER_29] = rtl8105e_hw_phy_config, - [RTL_GIGA_MAC_VER_30] = rtl8105e_hw_phy_config, - [RTL_GIGA_MAC_VER_31] = NULL, - [RTL_GIGA_MAC_VER_32] = rtl8168e_1_hw_phy_config, - [RTL_GIGA_MAC_VER_33] = rtl8168e_1_hw_phy_config, - [RTL_GIGA_MAC_VER_34] = rtl8168e_2_hw_phy_config, - [RTL_GIGA_MAC_VER_35] = rtl8168f_1_hw_phy_config, - [RTL_GIGA_MAC_VER_36] = rtl8168f_2_hw_phy_config, - [RTL_GIGA_MAC_VER_37] = rtl8402_hw_phy_config, - [RTL_GIGA_MAC_VER_38] = rtl8411_hw_phy_config, - [RTL_GIGA_MAC_VER_39] = rtl8106e_hw_phy_config, - [RTL_GIGA_MAC_VER_40] = rtl8168g_1_hw_phy_config, - [RTL_GIGA_MAC_VER_41] = NULL, - [RTL_GIGA_MAC_VER_42] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_43] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_44] = rtl8168g_2_hw_phy_config, - [RTL_GIGA_MAC_VER_45] = rtl8168h_1_hw_phy_config, - [RTL_GIGA_MAC_VER_46] = rtl8168h_2_hw_phy_config, - [RTL_GIGA_MAC_VER_47] = rtl8168h_1_hw_phy_config, - [RTL_GIGA_MAC_VER_48] = rtl8168h_2_hw_phy_config, - [RTL_GIGA_MAC_VER_49] = rtl8168ep_1_hw_phy_config, - [RTL_GIGA_MAC_VER_50] = rtl8168ep_2_hw_phy_config, - [RTL_GIGA_MAC_VER_51] = rtl8168ep_2_hw_phy_config, - [RTL_GIGA_MAC_VER_52] = rtl8117_hw_phy_config, - [RTL_GIGA_MAC_VER_60] = rtl8125_1_hw_phy_config, - [RTL_GIGA_MAC_VER_61] = rtl8125_2_hw_phy_config, - }; - struct rtl8169_private *tp = netdev_priv(dev); + ioffset = (data2 >> 1) & 0x7ff8; + ioffset |= data2 & 0x0007; + if (data1 & BIT(7)) + ioffset |= BIT(15); - if (phy_configs[tp->mac_version]) - phy_configs[tp->mac_version](tp); + return ioffset; } static void rtl_schedule_task(struct rtl8169_private *tp, enum rtl_flag flag) @@ -3617,21 +2232,28 @@ static void rtl_schedule_task(struct rtl8169_private *tp, enum rtl_flag flag) schedule_work(&tp->wk.work); } -static void rtl8169_init_phy(struct net_device *dev, struct rtl8169_private *tp) +static void rtl8169_init_phy(struct rtl8169_private *tp) { - rtl_hw_phy_config(dev); + r8169_hw_phy_config(tp, tp->phydev, tp->mac_version); if (tp->mac_version <= RTL_GIGA_MAC_VER_06) { pci_write_config_byte(tp->pci_dev, PCI_LATENCY_TIMER, 0x40); pci_write_config_byte(tp->pci_dev, PCI_CACHE_LINE_SIZE, 0x08); - netif_dbg(tp, drv, dev, - "Set MAC Reg C+CR Offset 0x82h = 0x01h\n"); + /* set undocumented MAC Reg C+CR Offset 0x82h */ RTL_W8(tp, 0x82, 0x01); } + if (tp->mac_version == RTL_GIGA_MAC_VER_05 && + tp->pci_dev->subsystem_vendor == PCI_VENDOR_ID_GIGABYTE && + tp->pci_dev->subsystem_device == 0xe000) + phy_write_paged(tp->phydev, 0x0001, 0x10, 0xf01b); + /* We may have called phy_speed_down before */ phy_speed_up(tp->phydev); + if (rtl_supports_eee(tp)) + rtl_enable_eee(tp); + genphy_soft_reset(tp->phydev); } @@ -3675,16 +2297,6 @@ static int rtl_set_mac_address(struct net_device *dev, void *p) return 0; } -static int rtl8169_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - struct rtl8169_private *tp = netdev_priv(dev); - - if (!netif_running(dev)) - return -ENODEV; - - return phy_mii_ioctl(tp->phydev, ifr, cmd); -} - static void rtl_wol_suspend_quirk(struct rtl8169_private *tp) { switch (tp->mac_version) { @@ -4710,9 +3322,7 @@ static void rtl_hw_start_8168h_1(struct rtl8169_private *tp) rtl_pcie_state_l2l3_disable(tp); - rtl_writephy(tp, 0x1f, 0x0c42); - rg_saw_cnt = (rtl_readphy(tp, 0x13) & 0x3fff); - rtl_writephy(tp, 0x1f, 0x0000); + rg_saw_cnt = phy_read_paged(tp->phydev, 0x0c42, 0x13) & 0x3fff; if (rg_saw_cnt > 0) { u16 sw_cnt_1ms_ini; @@ -4887,7 +3497,7 @@ static void rtl_hw_start_8117(struct rtl8169_private *tp) r8168_mac_ocp_write(tp, 0xc09e, 0x0000); /* firmware is for MAC only */ - rtl_apply_firmware(tp); + r8169_apply_firmware(tp); rtl_hw_aspm_clkreq_enable(tp, true); } @@ -4991,6 +3601,9 @@ static void rtl_hw_start_8402(struct rtl8169_private *tp) rtl_eri_write(tp, 0xb8, ERIAR_MASK_0011, 0x0000); rtl_w0w1_eri(tp, 0x0d4, ERIAR_MASK_0011, 0x0e00, 0xff00); + /* disable EEE */ + rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); + rtl_pcie_state_l2l3_disable(tp); } @@ -5005,6 +3618,11 @@ static void rtl_hw_start_8106(struct rtl8169_private *tp) RTL_W8(tp, MCU, RTL_R8(tp, MCU) | EN_NDP | EN_OOB_RESET); RTL_W8(tp, DLLPR, RTL_R8(tp, DLLPR) & ~PFM_EN); + rtl_eri_write(tp, 0x1d0, ERIAR_MASK_0011, 0x0000); + + /* disable EEE */ + rtl_eri_write(tp, 0x1b0, ERIAR_MASK_0011, 0x0000); + rtl_pcie_state_l2l3_disable(tp); rtl_hw_aspm_clkreq_enable(tp, true); } @@ -5222,11 +3840,8 @@ static void rtl_hw_start_8169(struct rtl8169_private *tp) tp->cp_cmd |= PCIMulRW; if (tp->mac_version == RTL_GIGA_MAC_VER_02 || - tp->mac_version == RTL_GIGA_MAC_VER_03) { - netif_dbg(tp, drv, tp->dev, - "Set MAC Reg C+CR Offset 0xe0. Bit 3 and Bit 14 MUST be 1\n"); - tp->cp_cmd |= (1 << 14); - } + tp->mac_version == RTL_GIGA_MAC_VER_03) + tp->cp_cmd |= EnAnaPLL; RTL_W16(tp, CPlusCmd, tp->cp_cmd); @@ -5435,7 +4050,7 @@ static void rtl_reset_work(struct rtl8169_private *tp) netif_wake_queue(dev); } -static void rtl8169_tx_timeout(struct net_device *dev) +static void rtl8169_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct rtl8169_private *tp = netdev_priv(dev); @@ -6240,7 +4855,7 @@ static int rtl_open(struct net_device *dev) napi_enable(&tp->napi); - rtl8169_init_phy(dev, tp); + rtl8169_init_phy(tp); rtl_pll_power_up(tp); @@ -6371,7 +4986,7 @@ static void __rtl8169_resume(struct net_device *dev) netif_device_attach(dev); rtl_pll_power_up(tp); - rtl8169_init_phy(dev, tp); + rtl8169_init_phy(tp); phy_start(tp->phydev); @@ -6542,7 +5157,7 @@ static const struct net_device_ops rtl_netdev_ops = { .ndo_fix_features = rtl8169_fix_features, .ndo_set_features = rtl8169_set_features, .ndo_set_mac_address = rtl_set_mac_address, - .ndo_do_ioctl = rtl8169_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_set_rx_mode = rtl_set_rx_mode, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = rtl8169_netpoll, @@ -6744,7 +5359,7 @@ static int rtl_jumbo_max(struct rtl8169_private *tp) { /* Non-GBit versions don't support jumbo frames */ if (!tp->supports_gmii) - return JUMBO_1K; + return 0; switch (tp->mac_version) { /* RTL8169 */ @@ -6825,6 +5440,15 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) int chipset, region; int jumbo_max, rc; + /* Some tools for creating an initramfs don't consider softdeps, then + * r8169.ko may be in initramfs, but realtek.ko not. Then the generic + * PHY driver is used that doesn't work with most chip versions. + */ + if (!driver_find("RTL8201CP Ethernet", &mdio_bus_type)) { + dev_err(&pdev->dev, "realtek.ko not loaded, maybe it needs to be added to initramfs?\n"); + return -ENOENT; + } + dev = devm_alloc_etherdev(&pdev->dev, sizeof (*tp)); if (!dev) return -ENOMEM; @@ -6966,10 +5590,9 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) dev->hw_features |= NETIF_F_RXALL; dev->hw_features |= NETIF_F_RXFCS; - /* MTU range: 60 - hw-specific max */ - dev->min_mtu = ETH_ZLEN; jumbo_max = rtl_jumbo_max(tp); - dev->max_mtu = jumbo_max; + if (jumbo_max) + dev->max_mtu = jumbo_max; rtl_set_irq_mask(tp); @@ -6999,7 +5622,7 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent) (RTL_R32(tp, TxConfig) >> 20) & 0xfcf, pci_irq_vector(pdev, 0)); - if (jumbo_max > JUMBO_1K) + if (jumbo_max) netif_info(tp, probe, dev, "jumbo features [frames: %d bytes, tx checksumming: %s]\n", jumbo_max, tp->mac_version <= RTL_GIGA_MAC_VER_06 ? diff --git a/drivers/net/ethernet/realtek/r8169_phy_config.c b/drivers/net/ethernet/realtek/r8169_phy_config.c new file mode 100644 index 000000000000..e367e77c773b --- /dev/null +++ b/drivers/net/ethernet/realtek/r8169_phy_config.c @@ -0,0 +1,1307 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * r8169_phy_config.c: RealTek 8169/8168/8101 ethernet driver. + * + * Copyright (c) 2002 ShuChen <shuchen@realtek.com.tw> + * Copyright (c) 2003 - 2007 Francois Romieu <romieu@fr.zoreil.com> + * Copyright (c) a lot of people too. Please respect their work. + * + * See MAINTAINERS file for support contact information. + */ + +#include <linux/delay.h> +#include <linux/phy.h> + +#include "r8169.h" + +typedef void (*rtl_phy_cfg_fct)(struct rtl8169_private *tp, + struct phy_device *phydev); + +static void r8168d_modify_extpage(struct phy_device *phydev, int extpage, + int reg, u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0007); + + __phy_write(phydev, 0x1e, extpage); + __phy_modify(phydev, reg, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +static void r8168d_phy_param(struct phy_device *phydev, u16 parm, + u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0005); + + __phy_write(phydev, 0x05, parm); + __phy_modify(phydev, 0x06, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +static void r8168g_phy_param(struct phy_device *phydev, u16 parm, + u16 mask, u16 val) +{ + int oldpage = phy_select_page(phydev, 0x0a43); + + __phy_write(phydev, 0x13, parm); + __phy_modify(phydev, 0x14, mask, val); + + phy_restore_page(phydev, oldpage, 0); +} + +struct phy_reg { + u16 reg; + u16 val; +}; + +static void __rtl_writephy_batch(struct phy_device *phydev, + const struct phy_reg *regs, int len) +{ + phy_lock_mdio_bus(phydev); + + while (len-- > 0) { + __phy_write(phydev, regs->reg, regs->val); + regs++; + } + + phy_unlock_mdio_bus(phydev); +} + +#define rtl_writephy_batch(p, a) __rtl_writephy_batch(p, a, ARRAY_SIZE(a)) + +static void rtl8168f_config_eee_phy(struct phy_device *phydev) +{ + r8168d_modify_extpage(phydev, 0x0020, 0x15, 0, BIT(8)); + r8168d_phy_param(phydev, 0x8b85, 0, BIT(13)); +} + +static void rtl8168g_config_eee_phy(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0a43, 0x11, 0, BIT(4)); +} + +static void rtl8168h_config_eee_phy(struct phy_device *phydev) +{ + rtl8168g_config_eee_phy(phydev); + + phy_modify_paged(phydev, 0xa4a, 0x11, 0x0000, 0x0200); + phy_modify_paged(phydev, 0xa42, 0x14, 0x0000, 0x0080); +} + +static void rtl8125_config_eee_phy(struct phy_device *phydev) +{ + rtl8168h_config_eee_phy(phydev); + + phy_modify_paged(phydev, 0xa6d, 0x12, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa6d, 0x14, 0x0010, 0x0000); +} + +static void rtl8169s_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x06, 0x006e }, + { 0x08, 0x0708 }, + { 0x15, 0x4000 }, + { 0x18, 0x65c7 }, + + { 0x1f, 0x0001 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x0000 }, + + { 0x03, 0xff41 }, + { 0x02, 0xdf60 }, + { 0x01, 0x0140 }, + { 0x00, 0x0077 }, + { 0x04, 0x7800 }, + { 0x04, 0x7000 }, + + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf0f9 }, + { 0x04, 0x9800 }, + { 0x04, 0x9000 }, + + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xa000 }, + + { 0x03, 0xff41 }, + { 0x02, 0xdf20 }, + { 0x01, 0x0140 }, + { 0x00, 0x00bb }, + { 0x04, 0xb800 }, + { 0x04, 0xb000 }, + + { 0x03, 0xdf41 }, + { 0x02, 0xdc60 }, + { 0x01, 0x6340 }, + { 0x00, 0x007d }, + { 0x04, 0xd800 }, + { 0x04, 0xd000 }, + + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x100a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0xf000 }, + + { 0x1f, 0x0000 }, + { 0x0b, 0x0000 }, + { 0x00, 0x9200 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8169sb_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0002, 0x01, 0x90d0); +} + +static void rtl8169scd_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x04, 0x0000 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x9000 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0xa000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0x0000 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x10, 0xf41b }, + { 0x14, 0xfb54 }, + { 0x18, 0xf5c7 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8169sce_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x04, 0x0000 }, + { 0x03, 0x00a1 }, + { 0x02, 0x0008 }, + { 0x01, 0x0120 }, + { 0x00, 0x1000 }, + { 0x04, 0x0800 }, + { 0x04, 0x9000 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0xa000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0xff95 }, + { 0x00, 0xba00 }, + { 0x04, 0xa800 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0x0000 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x0b, 0x8480 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x18, 0x67c7 }, + { 0x04, 0x2000 }, + { 0x03, 0x002f }, + { 0x02, 0x4360 }, + { 0x01, 0x0109 }, + { 0x00, 0x3022 }, + { 0x04, 0x2800 }, + { 0x1f, 0x0000 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8168bb_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write(phydev, 0x1f, 0x0001); + phy_set_bits(phydev, 0x16, BIT(0)); + phy_write(phydev, 0x10, 0xf41b); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8168bef_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0001, 0x10, 0xf41b); +} + +static void rtl8168cp_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write(phydev, 0x1d, 0x0f00); + phy_write_paged(phydev, 0x0002, 0x0c, 0x1ec8); +} + +static void rtl8168cp_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); + phy_write_paged(phydev, 0x0001, 0x1d, 0x3d98); +} + +static void rtl8168c_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x1f, 0x0002 }, + { 0x00, 0x88d4 }, + { 0x01, 0x82b1 }, + { 0x03, 0x7002 }, + { 0x08, 0x9e30 }, + { 0x09, 0x01f0 }, + { 0x0a, 0x5500 }, + { 0x0c, 0x00c8 }, + { 0x1f, 0x0003 }, + { 0x12, 0xc096 }, + { 0x16, 0x000a }, + { 0x1f, 0x0000 }, + { 0x1f, 0x0000 }, + { 0x09, 0x2000 }, + { 0x09, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168c_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x03, 0x802f }, + { 0x02, 0x4f02 }, + { 0x01, 0x0409 }, + { 0x00, 0xf099 }, + { 0x04, 0x9800 }, + { 0x04, 0x9000 }, + { 0x1d, 0x3d98 }, + { 0x1f, 0x0002 }, + { 0x0c, 0x7eb8 }, + { 0x06, 0x0761 }, + { 0x1f, 0x0003 }, + { 0x16, 0x0f0a }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x16, BIT(0)); + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168c_3_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0001 }, + { 0x12, 0x2300 }, + { 0x1d, 0x3d98 }, + { 0x1f, 0x0002 }, + { 0x0c, 0x7eb8 }, + { 0x06, 0x5461 }, + { 0x1f, 0x0003 }, + { 0x16, 0x0f0a }, + { 0x1f, 0x0000 } + }; + + rtl_writephy_batch(phydev, phy_reg_init); + + phy_set_bits(phydev, 0x16, BIT(0)); + phy_set_bits(phydev, 0x14, BIT(5)); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static const struct phy_reg rtl8168d_1_phy_reg_init_0[] = { + /* Channel Estimation */ + { 0x1f, 0x0001 }, + { 0x06, 0x4064 }, + { 0x07, 0x2863 }, + { 0x08, 0x059c }, + { 0x09, 0x26b4 }, + { 0x0a, 0x6a19 }, + { 0x0b, 0xdcc8 }, + { 0x10, 0xf06d }, + { 0x14, 0x7f68 }, + { 0x18, 0x7fd9 }, + { 0x1c, 0xf0ff }, + { 0x1d, 0x3d9c }, + { 0x1f, 0x0003 }, + { 0x12, 0xf49f }, + { 0x13, 0x070b }, + { 0x1a, 0x05ad }, + { 0x14, 0x94c0 }, + + /* + * Tx Error Issue + * Enhance line driver power + */ + { 0x1f, 0x0002 }, + { 0x06, 0x5561 }, + { 0x1f, 0x0005 }, + { 0x05, 0x8332 }, + { 0x06, 0x5561 }, + + /* + * Can not link to 1Gbps with bad cable + * Decrease SNR threshold form 21.07dB to 19.04dB + */ + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + + { 0x1f, 0x0000 }, + { 0x0d, 0xf880 } +}; + +static const struct phy_reg rtl8168d_1_phy_reg_init_1[] = { + { 0x1f, 0x0002 }, + { 0x05, 0x669a }, + { 0x1f, 0x0005 }, + { 0x05, 0x8330 }, + { 0x06, 0x669a }, + { 0x1f, 0x0002 } +}; + +static void rtl8168d_apply_firmware_cond(struct rtl8169_private *tp, + struct phy_device *phydev, + u16 val) +{ + u16 reg_val; + + phy_write(phydev, 0x1f, 0x0005); + phy_write(phydev, 0x05, 0x001b); + reg_val = phy_read(phydev, 0x06); + phy_write(phydev, 0x1f, 0x0000); + + if (reg_val != val) + phydev_warn(phydev, "chipset not ready for firmware\n"); + else + r8169_apply_firmware(tp); +} + +static void rtl8168d_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_0); + + /* + * Rx Error Issue + * Fine Tune Switching regulator parameter + */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x0b, 0x00ef, 0x0010); + phy_modify(phydev, 0x0c, 0x5d00, 0xa200); + + if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { + int val; + + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_1); + + val = phy_read(phydev, 0x0d); + + if ((val & 0x00ff) != 0x006c) { + static const u32 set[] = { + 0x0065, 0x0066, 0x0067, 0x0068, + 0x0069, 0x006a, 0x006b, 0x006c + }; + int i; + + phy_write(phydev, 0x1f, 0x0002); + + val &= 0xff00; + for (i = 0; i < ARRAY_SIZE(set); i++) + phy_write(phydev, 0x0d, val | set[i]); + } + } else { + phy_write_paged(phydev, 0x0002, 0x05, 0x6662); + r8168d_phy_param(phydev, 0x8330, 0xffff, 0x6662); + } + + /* RSET couple improve */ + phy_write(phydev, 0x1f, 0x0002); + phy_set_bits(phydev, 0x0d, 0x0300); + phy_set_bits(phydev, 0x0f, 0x0010); + + /* Fine tune PLL performance */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x02, 0x0600, 0x0100); + phy_clear_bits(phydev, 0x03, 0xe000); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168d_apply_firmware_cond(tp, phydev, 0xbf00); +} + +static void rtl8168d_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_0); + + if (rtl8168d_efuse_read(tp, 0x01) == 0xb1) { + int val; + + rtl_writephy_batch(phydev, rtl8168d_1_phy_reg_init_1); + + val = phy_read(phydev, 0x0d); + if ((val & 0x00ff) != 0x006c) { + static const u32 set[] = { + 0x0065, 0x0066, 0x0067, 0x0068, + 0x0069, 0x006a, 0x006b, 0x006c + }; + int i; + + phy_write(phydev, 0x1f, 0x0002); + + val &= 0xff00; + for (i = 0; i < ARRAY_SIZE(set); i++) + phy_write(phydev, 0x0d, val | set[i]); + } + } else { + phy_write_paged(phydev, 0x0002, 0x05, 0x2642); + r8168d_phy_param(phydev, 0x8330, 0xffff, 0x2642); + } + + /* Fine tune PLL performance */ + phy_write(phydev, 0x1f, 0x0002); + phy_modify(phydev, 0x02, 0x0600, 0x0100); + phy_clear_bits(phydev, 0x03, 0xe000); + phy_write(phydev, 0x1f, 0x0000); + + /* Switching regulator Slew rate */ + phy_modify_paged(phydev, 0x0002, 0x0f, 0x0000, 0x0017); + + rtl8168d_apply_firmware_cond(tp, phydev, 0xb300); +} + +static void rtl8168d_3_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0002 }, + { 0x10, 0x0008 }, + { 0x0d, 0x006c }, + + { 0x1f, 0x0000 }, + { 0x0d, 0xf880 }, + + { 0x1f, 0x0001 }, + { 0x17, 0x0cc0 }, + + { 0x1f, 0x0001 }, + { 0x0b, 0xa4d8 }, + { 0x09, 0x281c }, + { 0x07, 0x2883 }, + { 0x0a, 0x6b35 }, + { 0x1d, 0x3da4 }, + { 0x1c, 0xeffd }, + { 0x14, 0x7f52 }, + { 0x18, 0x7fc6 }, + { 0x08, 0x0601 }, + { 0x06, 0x4063 }, + { 0x10, 0xf074 }, + { 0x1f, 0x0003 }, + { 0x13, 0x0789 }, + { 0x12, 0xf4bd }, + { 0x1a, 0x04fd }, + { 0x14, 0x84b0 }, + { 0x1f, 0x0000 }, + { 0x00, 0x9200 }, + + { 0x1f, 0x0005 }, + { 0x01, 0x0340 }, + { 0x1f, 0x0001 }, + { 0x04, 0x4000 }, + { 0x03, 0x1d21 }, + { 0x02, 0x0c32 }, + { 0x01, 0x0200 }, + { 0x00, 0x5554 }, + { 0x04, 0x4800 }, + { 0x04, 0x4000 }, + { 0x04, 0xf000 }, + { 0x03, 0xdf01 }, + { 0x02, 0xdf20 }, + { 0x01, 0x101a }, + { 0x00, 0xa0ff }, + { 0x04, 0xf800 }, + { 0x04, 0xf000 }, + { 0x1f, 0x0000 }, + }; + + rtl_writephy_batch(phydev, phy_reg_init); + r8168d_modify_extpage(phydev, 0x0023, 0x16, 0xffff, 0x0000); +} + +static void rtl8168d_4_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_write_paged(phydev, 0x0001, 0x17, 0x0cc0); + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0xffff, 0x0040); + phy_set_bits(phydev, 0x0d, BIT(5)); +} + +static void rtl8168e_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + /* Channel estimation fine tune */ + { 0x1f, 0x0001 }, + { 0x0b, 0x6c20 }, + { 0x07, 0x2872 }, + { 0x1c, 0xefff }, + { 0x1f, 0x0003 }, + { 0x14, 0x6420 }, + { 0x1f, 0x0000 }, + }; + + r8169_apply_firmware(tp); + + /* Enable Delay cap */ + r8168d_phy_param(phydev, 0x8b80, 0xffff, 0xc896); + + rtl_writephy_batch(phydev, phy_reg_init); + + /* Update PFM & 10M TX idle timer */ + r8168d_modify_extpage(phydev, 0x002f, 0x15, 0xffff, 0x1919); + + r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); + + /* DCO enable for 10M IDLE Power */ + r8168d_modify_extpage(phydev, 0x0023, 0x17, 0x0000, 0x0006); + + /* For impedance matching */ + phy_modify_paged(phydev, 0x0002, 0x08, 0x7f00, 0x8000); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0050); + phy_set_bits(phydev, 0x14, BIT(15)); + + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + r8168d_phy_param(phydev, 0x8b85, 0x2000, 0x0000); + + r8168d_modify_extpage(phydev, 0x0020, 0x15, 0x1100, 0x0000); + phy_write_paged(phydev, 0x0006, 0x00, 0x5a00); + + phy_write_mmd(phydev, MDIO_MMD_AN, MDIO_AN_EEE_ADV, 0x0000); +} + +static void rtl8168e_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + /* Enable Delay cap */ + r8168d_modify_extpage(phydev, 0x00ac, 0x18, 0xffff, 0x0006); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Green Setting */ + r8168d_phy_param(phydev, 0x8b5b, 0xffff, 0x9222); + r8168d_phy_param(phydev, 0x8b6d, 0xffff, 0x8000); + r8168d_phy_param(phydev, 0x8b76, 0xffff, 0x8000); + + /* For 4-corner performance improve */ + phy_write(phydev, 0x1f, 0x0005); + phy_write(phydev, 0x05, 0x8b80); + phy_set_bits(phydev, 0x17, 0x0006); + phy_write(phydev, 0x1f, 0x0000); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); + phy_set_bits(phydev, 0x14, BIT(15)); + + /* improve 10M EEE waveform */ + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); + + rtl8168f_config_eee_phy(phydev); + + /* Green feature */ + phy_write(phydev, 0x1f, 0x0003); + phy_set_bits(phydev, 0x19, BIT(0)); + phy_set_bits(phydev, 0x10, BIT(10)); + phy_write(phydev, 0x1f, 0x0000); + phy_modify_paged(phydev, 0x0005, 0x01, 0, BIT(8)); +} + +static void rtl8168f_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* For 4-corner performance improve */ + r8168d_phy_param(phydev, 0x8b80, 0x0000, 0x0006); + + /* PHY auto speed down */ + r8168d_modify_extpage(phydev, 0x002d, 0x18, 0x0000, 0x0010); + phy_set_bits(phydev, 0x14, BIT(15)); + + /* Improve 10M EEE waveform */ + r8168d_phy_param(phydev, 0x8b86, 0x0000, 0x0001); + + rtl8168f_config_eee_phy(phydev); +} + +static void rtl8168f_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Modify green table for giga & fnet */ + r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00fb); + + /* Modify green table for 10M */ + r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); + + /* Disable hiimpedance detection (RTCT) */ + phy_write_paged(phydev, 0x0003, 0x01, 0x328a); + + rtl8168f_hw_phy_config(tp, phydev); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); +} + +static void rtl8168f_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + rtl8168f_hw_phy_config(tp, phydev); +} + +static void rtl8411_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + + rtl8168f_hw_phy_config(tp, phydev); + + /* Improve 2-pair detection performance */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x4000); + + /* Channel estimation fine tune */ + phy_write_paged(phydev, 0x0003, 0x09, 0xa20f); + + /* Modify green table for giga & fnet */ + r8168d_phy_param(phydev, 0x8b55, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b5e, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b67, 0xffff, 0x0000); + r8168d_phy_param(phydev, 0x8b70, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x17, 0xffff, 0x0000); + r8168d_modify_extpage(phydev, 0x0078, 0x19, 0xffff, 0x00aa); + + /* Modify green table for 10M */ + r8168d_phy_param(phydev, 0x8b79, 0xffff, 0xaa00); + + /* Disable hiimpedance detection (RTCT) */ + phy_write_paged(phydev, 0x0003, 0x01, 0x328a); + + /* Modify green table for giga */ + r8168d_phy_param(phydev, 0x8b54, 0x0800, 0x0000); + r8168d_phy_param(phydev, 0x8b5d, 0x0800, 0x0000); + r8168d_phy_param(phydev, 0x8a7c, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a7f, 0x0000, 0x0100); + r8168d_phy_param(phydev, 0x8a82, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a85, 0x0100, 0x0000); + r8168d_phy_param(phydev, 0x8a88, 0x0100, 0x0000); + + /* uc same-seed solution */ + r8168d_phy_param(phydev, 0x8b85, 0x0000, 0x8000); + + /* Green feature */ + phy_write(phydev, 0x1f, 0x0003); + phy_clear_bits(phydev, 0x19, BIT(0)); + phy_clear_bits(phydev, 0x10, BIT(10)); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8168g_disable_aldps(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0a43, 0x10, BIT(2), 0); +} + +static void rtl8168g_phy_adjust_10m_aldps(struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0x0bcc, 0x14, BIT(8), 0); + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(7) | BIT(6)); + r8168g_phy_param(phydev, 0x8084, 0x6000, 0x0000); + phy_modify_paged(phydev, 0x0a43, 0x10, 0x0000, 0x1003); +} + +static void rtl8168g_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + int ret; + + r8169_apply_firmware(tp); + + ret = phy_read_paged(phydev, 0x0a46, 0x10); + if (ret & BIT(8)) + phy_modify_paged(phydev, 0x0bcc, 0x12, BIT(15), 0); + else + phy_modify_paged(phydev, 0x0bcc, 0x12, 0, BIT(15)); + + ret = phy_read_paged(phydev, 0x0a46, 0x13); + if (ret & BIT(8)) + phy_modify_paged(phydev, 0x0c41, 0x15, 0, BIT(1)); + else + phy_modify_paged(phydev, 0x0c41, 0x15, BIT(1), 0); + + /* Enable PHY auto speed down */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); + + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* EEE auto-fallback function */ + phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + /* Improve SWR Efficiency */ + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x5065); + phy_write(phydev, 0x14, 0xd065); + phy_write(phydev, 0x1f, 0x0bc8); + phy_write(phydev, 0x11, 0x5655); + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x14, 0x9065); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168g_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + r8169_apply_firmware(tp); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168h_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + u16 dout_tapbin; + u32 data; + + r8169_apply_firmware(tp); + + /* CHN EST parameters adjust - giga master */ + r8168g_phy_param(phydev, 0x809b, 0xf800, 0x8000); + r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x8000); + r8168g_phy_param(phydev, 0x80a4, 0xff00, 0x8500); + r8168g_phy_param(phydev, 0x809c, 0xff00, 0xbd00); + + /* CHN EST parameters adjust - giga slave */ + r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x7000); + r8168g_phy_param(phydev, 0x80b4, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x80ac, 0xff00, 0x4000); + + /* CHN EST parameters adjust - fnet */ + r8168g_phy_param(phydev, 0x808e, 0xff00, 0x1200); + r8168g_phy_param(phydev, 0x8090, 0xff00, 0xe500); + r8168g_phy_param(phydev, 0x8092, 0xff00, 0x9f00); + + /* enable R-tune & PGA-retune function */ + dout_tapbin = 0; + data = phy_read_paged(phydev, 0x0a46, 0x13); + data &= 3; + data <<= 2; + dout_tapbin |= data; + data = phy_read_paged(phydev, 0x0a46, 0x12); + data &= 0xc000; + data >>= 14; + dout_tapbin |= data; + dout_tapbin = ~(dout_tapbin ^ 0x08); + dout_tapbin <<= 12; + dout_tapbin &= 0xf000; + + r8168g_phy_param(phydev, 0x827a, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827b, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827c, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x827d, 0xf000, dout_tapbin); + r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); + phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + /* SAR ADC performance */ + phy_modify_paged(phydev, 0x0bca, 0x17, BIT(12) | BIT(13), BIT(14)); + + r8168g_phy_param(phydev, 0x803f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8047, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x804f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8057, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x805f, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x8067, 0x3000, 0x0000); + r8168g_phy_param(phydev, 0x806f, 0x3000, 0x0000); + + /* disable phy pfm mode */ + phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); + + rtl8168g_disable_aldps(phydev); + rtl8168h_config_eee_phy(phydev); +} + +static void rtl8168h_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + u16 ioffset, rlen; + u32 data; + + r8169_apply_firmware(tp); + + /* CHIN EST parameter update */ + r8168g_phy_param(phydev, 0x808a, 0x003f, 0x000a); + + /* enable R-tune & PGA-retune function */ + r8168g_phy_param(phydev, 0x0811, 0x0000, 0x0800); + phy_modify_paged(phydev, 0x0a42, 0x16, 0x0000, 0x0002); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + ioffset = rtl8168h_2_get_adc_bias_ioffset(tp); + if (ioffset != 0xffff) + phy_write_paged(phydev, 0x0bcf, 0x16, ioffset); + + /* Modify rlen (TX LPF corner frequency) level */ + data = phy_read_paged(phydev, 0x0bcd, 0x16); + data &= 0x000f; + rlen = 0; + if (data > 3) + rlen = data - 3; + data = rlen | (rlen << 4) | (rlen << 8) | (rlen << 12); + phy_write_paged(phydev, 0x0bcd, 0x17, data); + + /* disable phy pfm mode */ + phy_modify_paged(phydev, 0x0a44, 0x11, BIT(7), 0); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168ep_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Enable PHY auto speed down */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(3) | BIT(2)); + + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* Enable EEE auto-fallback function */ + phy_modify_paged(phydev, 0x0a4b, 0x11, 0, BIT(2)); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + /* set rg_sel_sdm_rate */ + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8168ep_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + rtl8168g_phy_adjust_10m_aldps(phydev); + + /* Enable UC LPF tune function */ + r8168g_phy_param(phydev, 0x8012, 0x0000, 0x8000); + + /* Set rg_sel_sdm_rate */ + phy_modify_paged(phydev, 0x0c42, 0x11, BIT(13), BIT(14)); + + /* Channel estimation parameters */ + r8168g_phy_param(phydev, 0x80f3, 0xff00, 0x8b00); + r8168g_phy_param(phydev, 0x80f0, 0xff00, 0x3a00); + r8168g_phy_param(phydev, 0x80ef, 0xff00, 0x0500); + r8168g_phy_param(phydev, 0x80f6, 0xff00, 0x6e00); + r8168g_phy_param(phydev, 0x80ec, 0xff00, 0x6800); + r8168g_phy_param(phydev, 0x80ed, 0xff00, 0x7c00); + r8168g_phy_param(phydev, 0x80f2, 0xff00, 0xf400); + r8168g_phy_param(phydev, 0x80f4, 0xff00, 0x8500); + r8168g_phy_param(phydev, 0x8110, 0xff00, 0xa800); + r8168g_phy_param(phydev, 0x810f, 0xff00, 0x1d00); + r8168g_phy_param(phydev, 0x8111, 0xff00, 0xf500); + r8168g_phy_param(phydev, 0x8113, 0xff00, 0x6100); + r8168g_phy_param(phydev, 0x8115, 0xff00, 0x9200); + r8168g_phy_param(phydev, 0x810e, 0xff00, 0x0400); + r8168g_phy_param(phydev, 0x810c, 0xff00, 0x7c00); + r8168g_phy_param(phydev, 0x810b, 0xff00, 0x5a00); + r8168g_phy_param(phydev, 0x80d1, 0xff00, 0xff00); + r8168g_phy_param(phydev, 0x80cd, 0xff00, 0x9e00); + r8168g_phy_param(phydev, 0x80d3, 0xff00, 0x0e00); + r8168g_phy_param(phydev, 0x80d5, 0xff00, 0xca00); + r8168g_phy_param(phydev, 0x80d7, 0xff00, 0x8400); + + /* Force PWM-mode */ + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x5065); + phy_write(phydev, 0x14, 0xd065); + phy_write(phydev, 0x1f, 0x0bc8); + phy_write(phydev, 0x12, 0x00ed); + phy_write(phydev, 0x1f, 0x0bcd); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x14, 0x9065); + phy_write(phydev, 0x14, 0x1065); + phy_write(phydev, 0x1f, 0x0000); + + rtl8168g_disable_aldps(phydev); + rtl8168g_config_eee_phy(phydev); +} + +static void rtl8117_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* CHN EST parameters adjust - fnet */ + r8168g_phy_param(phydev, 0x808e, 0xff00, 0x4800); + r8168g_phy_param(phydev, 0x8090, 0xff00, 0xcc00); + r8168g_phy_param(phydev, 0x8092, 0xff00, 0xb000); + + r8168g_phy_param(phydev, 0x8088, 0xff00, 0x6000); + r8168g_phy_param(phydev, 0x808b, 0x3f00, 0x0b00); + r8168g_phy_param(phydev, 0x808d, 0x1f00, 0x0600); + r8168g_phy_param(phydev, 0x808c, 0xff00, 0xb000); + r8168g_phy_param(phydev, 0x80a0, 0xff00, 0x2800); + r8168g_phy_param(phydev, 0x80a2, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x809b, 0xf800, 0xb000); + r8168g_phy_param(phydev, 0x809a, 0xff00, 0x4b00); + r8168g_phy_param(phydev, 0x809d, 0x3f00, 0x0800); + r8168g_phy_param(phydev, 0x80a1, 0xff00, 0x7000); + r8168g_phy_param(phydev, 0x809f, 0x1f00, 0x0300); + r8168g_phy_param(phydev, 0x809e, 0xff00, 0x8800); + r8168g_phy_param(phydev, 0x80b2, 0xff00, 0x2200); + r8168g_phy_param(phydev, 0x80ad, 0xf800, 0x9800); + r8168g_phy_param(phydev, 0x80af, 0x3f00, 0x0800); + r8168g_phy_param(phydev, 0x80b3, 0xff00, 0x6f00); + r8168g_phy_param(phydev, 0x80b1, 0x1f00, 0x0300); + r8168g_phy_param(phydev, 0x80b0, 0xff00, 0x9300); + + r8168g_phy_param(phydev, 0x8011, 0x0000, 0x0800); + + /* enable GPHY 10M */ + phy_modify_paged(phydev, 0x0a44, 0x11, 0, BIT(11)); + + r8168g_phy_param(phydev, 0x8016, 0x0000, 0x0400); + + rtl8168g_disable_aldps(phydev); + rtl8168h_config_eee_phy(phydev); +} + +static void rtl8102e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0003 }, + { 0x08, 0x441d }, + { 0x01, 0x9100 }, + { 0x1f, 0x0000 } + }; + + phy_set_bits(phydev, 0x11, BIT(12)); + phy_set_bits(phydev, 0x19, BIT(13)); + phy_set_bits(phydev, 0x10, BIT(15)); + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8105e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Disable ALDPS before ram code */ + phy_write(phydev, 0x18, 0x0310); + msleep(100); + + r8169_apply_firmware(tp); + + phy_write_paged(phydev, 0x0005, 0x1a, 0x0000); + phy_write_paged(phydev, 0x0004, 0x1c, 0x0000); + phy_write_paged(phydev, 0x0001, 0x15, 0x7701); +} + +static void rtl8402_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + /* Disable ALDPS before setting firmware */ + phy_write(phydev, 0x18, 0x0310); + msleep(20); + + r8169_apply_firmware(tp); + + /* EEE setting */ + phy_write(phydev, 0x1f, 0x0004); + phy_write(phydev, 0x10, 0x401f); + phy_write(phydev, 0x19, 0x7030); + phy_write(phydev, 0x1f, 0x0000); +} + +static void rtl8106e_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + static const struct phy_reg phy_reg_init[] = { + { 0x1f, 0x0004 }, + { 0x10, 0xc07f }, + { 0x19, 0x7030 }, + { 0x1f, 0x0000 } + }; + + /* Disable ALDPS before ram code */ + phy_write(phydev, 0x18, 0x0310); + msleep(100); + + r8169_apply_firmware(tp); + + rtl_writephy_batch(phydev, phy_reg_init); +} + +static void rtl8125_1_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + phy_modify_paged(phydev, 0xad4, 0x10, 0x03ff, 0x0084); + phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); + phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x0006); + phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); + phy_modify_paged(phydev, 0xac0, 0x14, 0x0000, 0x1100); + phy_modify_paged(phydev, 0xac8, 0x15, 0xf000, 0x7000); + phy_modify_paged(phydev, 0xad1, 0x14, 0x0000, 0x0400); + phy_modify_paged(phydev, 0xad1, 0x15, 0x0000, 0x03ff); + phy_modify_paged(phydev, 0xad1, 0x16, 0x0000, 0x03ff); + + r8168g_phy_param(phydev, 0x80ea, 0xff00, 0xc400); + r8168g_phy_param(phydev, 0x80eb, 0x0700, 0x0300); + r8168g_phy_param(phydev, 0x80f8, 0xff00, 0x1c00); + r8168g_phy_param(phydev, 0x80f1, 0xff00, 0x3000); + r8168g_phy_param(phydev, 0x80fe, 0xff00, 0xa500); + r8168g_phy_param(phydev, 0x8102, 0xff00, 0x5000); + r8168g_phy_param(phydev, 0x8105, 0xff00, 0x3300); + r8168g_phy_param(phydev, 0x8100, 0xff00, 0x7000); + r8168g_phy_param(phydev, 0x8104, 0xff00, 0xf000); + r8168g_phy_param(phydev, 0x8106, 0xff00, 0x6500); + r8168g_phy_param(phydev, 0x80dc, 0xff00, 0xed00); + r8168g_phy_param(phydev, 0x80df, 0x0000, 0x0100); + r8168g_phy_param(phydev, 0x80e1, 0x0100, 0x0000); + + phy_modify_paged(phydev, 0xbf0, 0x13, 0x003f, 0x0038); + r8168g_phy_param(phydev, 0x819f, 0xffff, 0xd0b6); + + phy_write_paged(phydev, 0xbc3, 0x12, 0x5555); + phy_modify_paged(phydev, 0xbf0, 0x15, 0x0e00, 0x0a00); + phy_modify_paged(phydev, 0xa5c, 0x10, 0x0400, 0x0000); + phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); + + rtl8125_config_eee_phy(phydev); +} + +static void rtl8125_2_hw_phy_config(struct rtl8169_private *tp, + struct phy_device *phydev) +{ + int i; + + phy_modify_paged(phydev, 0xad4, 0x17, 0x0000, 0x0010); + phy_modify_paged(phydev, 0xad1, 0x13, 0x03ff, 0x03ff); + phy_modify_paged(phydev, 0xad3, 0x11, 0x003f, 0x0006); + phy_modify_paged(phydev, 0xac0, 0x14, 0x1100, 0x0000); + phy_modify_paged(phydev, 0xacc, 0x10, 0x0003, 0x0002); + phy_modify_paged(phydev, 0xad4, 0x10, 0x00e7, 0x0044); + phy_modify_paged(phydev, 0xac1, 0x12, 0x0080, 0x0000); + phy_modify_paged(phydev, 0xac8, 0x10, 0x0300, 0x0000); + phy_modify_paged(phydev, 0xac5, 0x17, 0x0007, 0x0002); + phy_write_paged(phydev, 0xad4, 0x16, 0x00a8); + phy_write_paged(phydev, 0xac5, 0x16, 0x01ff); + phy_modify_paged(phydev, 0xac8, 0x15, 0x00f0, 0x0030); + + phy_write(phydev, 0x1f, 0x0b87); + phy_write(phydev, 0x16, 0x80a2); + phy_write(phydev, 0x17, 0x0153); + phy_write(phydev, 0x16, 0x809c); + phy_write(phydev, 0x17, 0x0153); + phy_write(phydev, 0x1f, 0x0000); + + phy_write(phydev, 0x1f, 0x0a43); + phy_write(phydev, 0x13, 0x81B3); + phy_write(phydev, 0x14, 0x0043); + phy_write(phydev, 0x14, 0x00A7); + phy_write(phydev, 0x14, 0x00D6); + phy_write(phydev, 0x14, 0x00EC); + phy_write(phydev, 0x14, 0x00F6); + phy_write(phydev, 0x14, 0x00FB); + phy_write(phydev, 0x14, 0x00FD); + phy_write(phydev, 0x14, 0x00FF); + phy_write(phydev, 0x14, 0x00BB); + phy_write(phydev, 0x14, 0x0058); + phy_write(phydev, 0x14, 0x0029); + phy_write(phydev, 0x14, 0x0013); + phy_write(phydev, 0x14, 0x0009); + phy_write(phydev, 0x14, 0x0004); + phy_write(phydev, 0x14, 0x0002); + for (i = 0; i < 25; i++) + phy_write(phydev, 0x14, 0x0000); + phy_write(phydev, 0x1f, 0x0000); + + r8168g_phy_param(phydev, 0x8257, 0xffff, 0x020F); + r8168g_phy_param(phydev, 0x80ea, 0xffff, 0x7843); + + r8169_apply_firmware(tp); + + phy_modify_paged(phydev, 0xd06, 0x14, 0x0000, 0x2000); + + r8168g_phy_param(phydev, 0x81a2, 0x0000, 0x0100); + + phy_modify_paged(phydev, 0xb54, 0x16, 0xff00, 0xdb00); + phy_modify_paged(phydev, 0xa45, 0x12, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa5d, 0x12, 0x0000, 0x0020); + phy_modify_paged(phydev, 0xad4, 0x17, 0x0010, 0x0000); + phy_modify_paged(phydev, 0xa86, 0x15, 0x0001, 0x0000); + phy_modify_paged(phydev, 0xa44, 0x11, 0x0000, 0x0800); + + rtl8125_config_eee_phy(phydev); +} + +void r8169_hw_phy_config(struct rtl8169_private *tp, struct phy_device *phydev, + enum mac_version ver) +{ + static const rtl_phy_cfg_fct phy_configs[] = { + /* PCI devices. */ + [RTL_GIGA_MAC_VER_02] = rtl8169s_hw_phy_config, + [RTL_GIGA_MAC_VER_03] = rtl8169s_hw_phy_config, + [RTL_GIGA_MAC_VER_04] = rtl8169sb_hw_phy_config, + [RTL_GIGA_MAC_VER_05] = rtl8169scd_hw_phy_config, + [RTL_GIGA_MAC_VER_06] = rtl8169sce_hw_phy_config, + /* PCI-E devices. */ + [RTL_GIGA_MAC_VER_07] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_08] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_09] = rtl8102e_hw_phy_config, + [RTL_GIGA_MAC_VER_10] = NULL, + [RTL_GIGA_MAC_VER_11] = rtl8168bb_hw_phy_config, + [RTL_GIGA_MAC_VER_12] = rtl8168bef_hw_phy_config, + [RTL_GIGA_MAC_VER_13] = NULL, + [RTL_GIGA_MAC_VER_14] = NULL, + [RTL_GIGA_MAC_VER_15] = NULL, + [RTL_GIGA_MAC_VER_16] = NULL, + [RTL_GIGA_MAC_VER_17] = rtl8168bef_hw_phy_config, + [RTL_GIGA_MAC_VER_18] = rtl8168cp_1_hw_phy_config, + [RTL_GIGA_MAC_VER_19] = rtl8168c_1_hw_phy_config, + [RTL_GIGA_MAC_VER_20] = rtl8168c_2_hw_phy_config, + [RTL_GIGA_MAC_VER_21] = rtl8168c_3_hw_phy_config, + [RTL_GIGA_MAC_VER_22] = rtl8168c_3_hw_phy_config, + [RTL_GIGA_MAC_VER_23] = rtl8168cp_2_hw_phy_config, + [RTL_GIGA_MAC_VER_24] = rtl8168cp_2_hw_phy_config, + [RTL_GIGA_MAC_VER_25] = rtl8168d_1_hw_phy_config, + [RTL_GIGA_MAC_VER_26] = rtl8168d_2_hw_phy_config, + [RTL_GIGA_MAC_VER_27] = rtl8168d_3_hw_phy_config, + [RTL_GIGA_MAC_VER_28] = rtl8168d_4_hw_phy_config, + [RTL_GIGA_MAC_VER_29] = rtl8105e_hw_phy_config, + [RTL_GIGA_MAC_VER_30] = rtl8105e_hw_phy_config, + [RTL_GIGA_MAC_VER_31] = NULL, + [RTL_GIGA_MAC_VER_32] = rtl8168e_1_hw_phy_config, + [RTL_GIGA_MAC_VER_33] = rtl8168e_1_hw_phy_config, + [RTL_GIGA_MAC_VER_34] = rtl8168e_2_hw_phy_config, + [RTL_GIGA_MAC_VER_35] = rtl8168f_1_hw_phy_config, + [RTL_GIGA_MAC_VER_36] = rtl8168f_2_hw_phy_config, + [RTL_GIGA_MAC_VER_37] = rtl8402_hw_phy_config, + [RTL_GIGA_MAC_VER_38] = rtl8411_hw_phy_config, + [RTL_GIGA_MAC_VER_39] = rtl8106e_hw_phy_config, + [RTL_GIGA_MAC_VER_40] = rtl8168g_1_hw_phy_config, + [RTL_GIGA_MAC_VER_41] = NULL, + [RTL_GIGA_MAC_VER_42] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_43] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_44] = rtl8168g_2_hw_phy_config, + [RTL_GIGA_MAC_VER_45] = rtl8168h_1_hw_phy_config, + [RTL_GIGA_MAC_VER_46] = rtl8168h_2_hw_phy_config, + [RTL_GIGA_MAC_VER_47] = rtl8168h_1_hw_phy_config, + [RTL_GIGA_MAC_VER_48] = rtl8168h_2_hw_phy_config, + [RTL_GIGA_MAC_VER_49] = rtl8168ep_1_hw_phy_config, + [RTL_GIGA_MAC_VER_50] = rtl8168ep_2_hw_phy_config, + [RTL_GIGA_MAC_VER_51] = rtl8168ep_2_hw_phy_config, + [RTL_GIGA_MAC_VER_52] = rtl8117_hw_phy_config, + [RTL_GIGA_MAC_VER_60] = rtl8125_1_hw_phy_config, + [RTL_GIGA_MAC_VER_61] = rtl8125_2_hw_phy_config, + }; + + if (phy_configs[ver]) + phy_configs[ver](tp, phydev); +} diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c index 4b13a184bfc7..067ad25553b9 100644 --- a/drivers/net/ethernet/renesas/ravb_main.c +++ b/drivers/net/ethernet/renesas/ravb_main.c @@ -1425,7 +1425,7 @@ out_napi_off: } /* Timeout function for Ethernet AVB */ -static void ravb_tx_timeout(struct net_device *ndev) +static void ravb_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct ravb_private *priv = netdev_priv(ndev); diff --git a/drivers/net/ethernet/renesas/sh_eth.c b/drivers/net/ethernet/renesas/sh_eth.c index 3591285250e1..58ca126518a2 100644 --- a/drivers/net/ethernet/renesas/sh_eth.c +++ b/drivers/net/ethernet/renesas/sh_eth.c @@ -2482,7 +2482,7 @@ out_napi_off: } /* Timeout function */ -static void sh_eth_tx_timeout(struct net_device *ndev) +static void sh_eth_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct sh_eth_private *mdp = netdev_priv(ndev); struct sh_eth_rxdesc *rxdesc; @@ -2647,20 +2647,6 @@ static int sh_eth_close(struct net_device *ndev) return 0; } -/* ioctl to device function */ -static int sh_eth_do_ioctl(struct net_device *ndev, struct ifreq *rq, int cmd) -{ - struct phy_device *phydev = ndev->phydev; - - if (!netif_running(ndev)) - return -EINVAL; - - if (!phydev) - return -ENODEV; - - return phy_mii_ioctl(phydev, rq, cmd); -} - static int sh_eth_change_mtu(struct net_device *ndev, int new_mtu) { if (netif_running(ndev)) @@ -3159,7 +3145,7 @@ static const struct net_device_ops sh_eth_netdev_ops = { .ndo_get_stats = sh_eth_get_stats, .ndo_set_rx_mode = sh_eth_set_rx_mode, .ndo_tx_timeout = sh_eth_tx_timeout, - .ndo_do_ioctl = sh_eth_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_change_mtu = sh_eth_change_mtu, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, @@ -3175,7 +3161,7 @@ static const struct net_device_ops sh_eth_netdev_ops_tsu = { .ndo_vlan_rx_add_vid = sh_eth_vlan_rx_add_vid, .ndo_vlan_rx_kill_vid = sh_eth_vlan_rx_kill_vid, .ndo_tx_timeout = sh_eth_tx_timeout, - .ndo_do_ioctl = sh_eth_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_change_mtu = sh_eth_change_mtu, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, diff --git a/drivers/net/ethernet/rocker/rocker_main.c b/drivers/net/ethernet/rocker/rocker_main.c index bc4f951315da..7585cd2270ba 100644 --- a/drivers/net/ethernet/rocker/rocker_main.c +++ b/drivers/net/ethernet/rocker/rocker_main.c @@ -2159,7 +2159,7 @@ static void rocker_router_fib_event_work(struct work_struct *work) /* Protect internal structures from changes */ rtnl_lock(); switch (fib_work->event) { - case FIB_EVENT_ENTRY_ADD: + case FIB_EVENT_ENTRY_REPLACE: err = rocker_world_fib4_add(rocker, &fib_work->fen_info); if (err) rocker_world_fib4_abort(rocker); @@ -2201,7 +2201,7 @@ static int rocker_router_fib_event(struct notifier_block *nb, fib_work->event = event; switch (event) { - case FIB_EVENT_ENTRY_ADD: /* fall through */ + case FIB_EVENT_ENTRY_REPLACE: /* fall through */ case FIB_EVENT_ENTRY_DEL: if (info->family == AF_INET) { struct fib_entry_notifier_info *fen_info = ptr; diff --git a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c index 52ed111d98f4..c705743d69f7 100644 --- a/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c +++ b/drivers/net/ethernet/samsung/sxgbe/sxgbe_main.c @@ -1572,7 +1572,7 @@ static int sxgbe_poll(struct napi_struct *napi, int budget) * netdev structure and arrange for the device to be reset to a sane state * in order to transmit a new packet. */ -static void sxgbe_tx_timeout(struct net_device *dev) +static void sxgbe_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sxgbe_priv_data *priv = netdev_priv(dev); @@ -1939,9 +1939,7 @@ static int sxgbe_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) case SIOCGMIIPHY: case SIOCGMIIREG: case SIOCSMIIREG: - if (!dev->phydev) - return -EINVAL; - ret = phy_mii_ioctl(dev->phydev, rq, cmd); + ret = phy_do_ioctl(dev, rq, cmd); break; default: break; diff --git a/drivers/net/ethernet/seeq/ether3.c b/drivers/net/ethernet/seeq/ether3.c index 632a7c85964d..128ee7cda1ed 100644 --- a/drivers/net/ethernet/seeq/ether3.c +++ b/drivers/net/ethernet/seeq/ether3.c @@ -79,7 +79,7 @@ static netdev_tx_t ether3_sendpacket(struct sk_buff *skb, static irqreturn_t ether3_interrupt (int irq, void *dev_id); static int ether3_close (struct net_device *dev); static void ether3_setmulticastlist (struct net_device *dev); -static void ether3_timeout(struct net_device *dev); +static void ether3_timeout(struct net_device *dev, unsigned int txqueue); #define BUS_16 2 #define BUS_8 1 @@ -450,7 +450,7 @@ static void ether3_setmulticastlist(struct net_device *dev) ether3_outw(priv(dev)->regs.config1 | CFG1_LOCBUFMEM, REG_CONFIG1); } -static void ether3_timeout(struct net_device *dev) +static void ether3_timeout(struct net_device *dev, unsigned int txqueue) { unsigned long flags; diff --git a/drivers/net/ethernet/seeq/sgiseeq.c b/drivers/net/ethernet/seeq/sgiseeq.c index 276c7cae7cee..8507ff242014 100644 --- a/drivers/net/ethernet/seeq/sgiseeq.c +++ b/drivers/net/ethernet/seeq/sgiseeq.c @@ -645,7 +645,7 @@ sgiseeq_start_xmit(struct sk_buff *skb, struct net_device *dev) return NETDEV_TX_OK; } -static void timeout(struct net_device *dev) +static void timeout(struct net_device *dev, unsigned int txqueue) { printk(KERN_NOTICE "%s: transmit timed out, resetting\n", dev->name); sgiseeq_reset(dev); diff --git a/drivers/net/ethernet/sfc/Kconfig b/drivers/net/ethernet/sfc/Kconfig index 5f36774bf4b8..ea5a9220196c 100644 --- a/drivers/net/ethernet/sfc/Kconfig +++ b/drivers/net/ethernet/sfc/Kconfig @@ -21,8 +21,6 @@ config SFC depends on PCI select MDIO select CRC32 - select I2C - select I2C_ALGOBIT imply PTP_1588_CLOCK ---help--- This driver supports 10/40-gigabit Ethernet cards based on diff --git a/drivers/net/ethernet/sfc/Makefile b/drivers/net/ethernet/sfc/Makefile index c5c297e78d06..87d093da22ca 100644 --- a/drivers/net/ethernet/sfc/Makefile +++ b/drivers/net/ethernet/sfc/Makefile @@ -1,7 +1,10 @@ # SPDX-License-Identifier: GPL-2.0 -sfc-y += efx.o nic.o farch.o siena.o ef10.o tx.o rx.o \ - selftest.o ethtool.o ptp.o tx_tso.o \ - mcdi.o mcdi_port.o mcdi_mon.o +sfc-y += efx.o efx_common.o efx_channels.o nic.o \ + farch.o siena.o ef10.o \ + tx.o tx_common.o tx_tso.o rx.o rx_common.o \ + selftest.o ethtool.o ethtool_common.o ptp.o \ + mcdi.o mcdi_port.o mcdi_port_common.o \ + mcdi_functions.o mcdi_filters.o mcdi_mon.o sfc-$(CONFIG_SFC_MTD) += mtd.o sfc-$(CONFIG_SFC_SRIOV) += sriov.o siena_sriov.o ef10_sriov.o diff --git a/drivers/net/ethernet/sfc/ef10.c b/drivers/net/ethernet/sfc/ef10.c index a6ae2cdc1986..52113b7529d6 100644 --- a/drivers/net/ethernet/sfc/ef10.c +++ b/drivers/net/ethernet/sfc/ef10.c @@ -5,11 +5,15 @@ */ #include "net_driver.h" +#include "rx_common.h" #include "ef10_regs.h" #include "io.h" #include "mcdi.h" #include "mcdi_pcol.h" +#include "mcdi_port_common.h" +#include "mcdi_functions.h" #include "nic.h" +#include "mcdi_filters.h" #include "workarounds.h" #include "selftest.h" #include "ef10_sriov.h" @@ -25,28 +29,6 @@ enum { EFX_EF10_TEST = 1, EFX_EF10_REFILL, }; -/* The maximum size of a shared RSS context */ -/* TODO: this should really be from the mcdi protocol export */ -#define EFX_EF10_MAX_SHARED_RSS_CONTEXT_SIZE 64UL - -/* The filter table(s) are managed by firmware and we have write-only - * access. When removing filters we must identify them to the - * firmware by a 64-bit handle, but this is too wide for Linux kernel - * interfaces (32-bit for RX NFC, 16-bit for RFS). Also, we need to - * be able to tell in advance whether a requested insertion will - * replace an existing filter. Therefore we maintain a software hash - * table, which should be at least as large as the hardware hash - * table. - * - * Huntington has a single 8K filter table shared between all filter - * types and both ports. - */ -#define HUNT_FILTER_TBL_ROWS 8192 - -#define EFX_EF10_FILTER_ID_INVALID 0xffff - -#define EFX_EF10_FILTER_DEV_UC_MAX 32 -#define EFX_EF10_FILTER_DEV_MC_MAX 256 /* VLAN list entry */ struct efx_ef10_vlan { @@ -54,95 +36,8 @@ struct efx_ef10_vlan { u16 vid; }; -enum efx_ef10_default_filters { - EFX_EF10_BCAST, - EFX_EF10_UCDEF, - EFX_EF10_MCDEF, - EFX_EF10_VXLAN4_UCDEF, - EFX_EF10_VXLAN4_MCDEF, - EFX_EF10_VXLAN6_UCDEF, - EFX_EF10_VXLAN6_MCDEF, - EFX_EF10_NVGRE4_UCDEF, - EFX_EF10_NVGRE4_MCDEF, - EFX_EF10_NVGRE6_UCDEF, - EFX_EF10_NVGRE6_MCDEF, - EFX_EF10_GENEVE4_UCDEF, - EFX_EF10_GENEVE4_MCDEF, - EFX_EF10_GENEVE6_UCDEF, - EFX_EF10_GENEVE6_MCDEF, - - EFX_EF10_NUM_DEFAULT_FILTERS -}; - -/* Per-VLAN filters information */ -struct efx_ef10_filter_vlan { - struct list_head list; - u16 vid; - u16 uc[EFX_EF10_FILTER_DEV_UC_MAX]; - u16 mc[EFX_EF10_FILTER_DEV_MC_MAX]; - u16 default_filters[EFX_EF10_NUM_DEFAULT_FILTERS]; -}; - -struct efx_ef10_dev_addr { - u8 addr[ETH_ALEN]; -}; - -struct efx_ef10_filter_table { -/* The MCDI match masks supported by this fw & hw, in order of priority */ - u32 rx_match_mcdi_flags[ - MC_CMD_GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES_MAXNUM * 2]; - unsigned int rx_match_count; - - struct rw_semaphore lock; /* Protects entries */ - struct { - unsigned long spec; /* pointer to spec plus flag bits */ -/* AUTO_OLD is used to mark and sweep MAC filters for the device address lists. */ -/* unused flag 1UL */ -#define EFX_EF10_FILTER_FLAG_AUTO_OLD 2UL -#define EFX_EF10_FILTER_FLAGS 3UL - u64 handle; /* firmware handle */ - } *entry; -/* Shadow of net_device address lists, guarded by mac_lock */ - struct efx_ef10_dev_addr dev_uc_list[EFX_EF10_FILTER_DEV_UC_MAX]; - struct efx_ef10_dev_addr dev_mc_list[EFX_EF10_FILTER_DEV_MC_MAX]; - int dev_uc_count; - int dev_mc_count; - bool uc_promisc; - bool mc_promisc; -/* Whether in multicast promiscuous mode when last changed */ - bool mc_promisc_last; - bool mc_overflow; /* Too many MC addrs; should always imply mc_promisc */ - bool vlan_filter; - struct list_head vlan_list; -}; - -/* An arbitrary search limit for the software hash table */ -#define EFX_EF10_FILTER_SEARCH_LIMIT 200 - -static void efx_ef10_rx_free_indir_table(struct efx_nic *efx); -static void efx_ef10_filter_table_remove(struct efx_nic *efx); -static int efx_ef10_filter_add_vlan(struct efx_nic *efx, u16 vid); -static void efx_ef10_filter_del_vlan_internal(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan); -static void efx_ef10_filter_del_vlan(struct efx_nic *efx, u16 vid); static int efx_ef10_set_udp_tnl_ports(struct efx_nic *efx, bool unloading); -static u32 efx_ef10_filter_get_unsafe_id(u32 filter_id) -{ - WARN_ON_ONCE(filter_id == EFX_EF10_FILTER_ID_INVALID); - return filter_id & (HUNT_FILTER_TBL_ROWS - 1); -} - -static unsigned int efx_ef10_filter_get_unsafe_pri(u32 filter_id) -{ - return filter_id / (HUNT_FILTER_TBL_ROWS * 2); -} - -static u32 efx_ef10_make_filter_id(unsigned int pri, u16 idx) -{ - return pri * HUNT_FILTER_TBL_ROWS * 2 + idx; -} - static int efx_ef10_get_warm_boot_count(struct efx_nic *efx) { efx_dword_t reg; @@ -185,24 +80,6 @@ static bool efx_ef10_is_vf(struct efx_nic *efx) return efx->type->is_vf; } -static int efx_ef10_get_pf_index(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_FUNCTION_INFO_OUT_LEN); - struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t outlen; - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_FUNCTION_INFO, NULL, 0, outbuf, - sizeof(outbuf), &outlen); - if (rc) - return rc; - if (outlen < sizeof(outbuf)) - return -EIO; - - nic_data->pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); - return 0; -} - #ifdef CONFIG_SFC_SRIOV static int efx_ef10_get_vf_index(struct efx_nic *efx) { @@ -273,24 +150,9 @@ static int efx_ef10_init_datapath_caps(struct efx_nic *efx) u8 vi_window_mode = MCDI_BYTE(outbuf, GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE); - switch (vi_window_mode) { - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_8K: - efx->vi_stride = 8192; - break; - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_16K: - efx->vi_stride = 16384; - break; - case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_64K: - efx->vi_stride = 65536; - break; - default: - netif_err(efx, probe, efx->net_dev, - "Unrecognised VI window mode %d\n", - vi_window_mode); - return -EIO; - } - netif_dbg(efx, probe, efx->net_dev, "vi_stride = %u\n", - efx->vi_stride); + rc = efx_mcdi_window_mode_to_stride(efx, vi_window_mode); + if (rc) + return rc; } else { /* keep default VI stride */ netif_dbg(efx, probe, efx->net_dev, @@ -576,7 +438,7 @@ static int efx_ef10_add_vlan(struct efx_nic *efx, u16 vid) if (efx->filter_state) { mutex_lock(&efx->mac_lock); down_write(&efx->filter_sem); - rc = efx_ef10_filter_add_vlan(efx, vlan->vid); + rc = efx_mcdi_filter_add_vlan(efx, vlan->vid); up_write(&efx->filter_sem); mutex_unlock(&efx->mac_lock); if (rc) @@ -605,7 +467,7 @@ static void efx_ef10_del_vlan_internal(struct efx_nic *efx, if (efx->filter_state) { down_write(&efx->filter_sem); - efx_ef10_filter_del_vlan(efx, vlan->vid); + efx_mcdi_filter_del_vlan(efx, vlan->vid); up_write(&efx->filter_sem); } @@ -689,7 +551,7 @@ static int efx_ef10_probe(struct efx_nic *efx) } nic_data->warm_boot_count = rc; - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; nic_data->vport_id = EVB_PORT_ID_ASSIGNED; @@ -725,7 +587,7 @@ static int efx_ef10_probe(struct efx_nic *efx) if (rc) goto fail4; - rc = efx_ef10_get_pf_index(efx); + rc = efx_get_pf_index(efx, &nic_data->pf_index); if (rc) goto fail5; @@ -831,22 +693,6 @@ fail1: return rc; } -static int efx_ef10_free_vis(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF_ERR(outbuf); - size_t outlen; - int rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FREE_VIS, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - - /* -EALREADY means nothing to free, so ignore */ - if (rc == -EALREADY) - rc = 0; - if (rc) - efx_mcdi_display_error(efx, MC_CMD_FREE_VIS, 0, outbuf, outlen, - rc); - return rc; -} - #ifdef EFX_USE_PIO static void efx_ef10_free_piobufs(struct efx_nic *efx) @@ -1084,12 +930,12 @@ static void efx_ef10_remove(struct efx_nic *efx) efx_mcdi_mon_remove(efx); - efx_ef10_rx_free_indir_table(efx); + efx_mcdi_rx_free_indir_table(efx); if (nic_data->wc_membase) iounmap(nic_data->wc_membase); - rc = efx_ef10_free_vis(efx); + rc = efx_mcdi_free_vis(efx); WARN_ON(rc != 0); if (!nic_data->must_restore_piobufs) @@ -1260,28 +1106,10 @@ static int efx_ef10_probe_vf(struct efx_nic *efx __attribute__ ((unused))) static int efx_ef10_alloc_vis(struct efx_nic *efx, unsigned int min_vis, unsigned int max_vis) { - MCDI_DECLARE_BUF(inbuf, MC_CMD_ALLOC_VIS_IN_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_ALLOC_VIS_OUT_LEN); struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MIN_VI_COUNT, min_vis); - MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MAX_VI_COUNT, max_vis); - rc = efx_mcdi_rpc(efx, MC_CMD_ALLOC_VIS, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - if (rc != 0) - return rc; - - if (outlen < MC_CMD_ALLOC_VIS_OUT_LEN) - return -EIO; - netif_dbg(efx, drv, efx->net_dev, "base VI is A0x%03x\n", - MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE)); - - nic_data->vi_base = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE); - nic_data->n_allocated_vis = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_COUNT); - return 0; + return efx_mcdi_alloc_vis(efx, min_vis, max_vis, &nic_data->vi_base, + &nic_data->n_allocated_vis); } /* Note that the failure path of this function does not free @@ -1363,7 +1191,7 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx) } /* In case the last attached driver failed to free VIs, do it now */ - rc = efx_ef10_free_vis(efx); + rc = efx_mcdi_free_vis(efx); if (rc != 0) return rc; @@ -1384,7 +1212,7 @@ static int efx_ef10_dimension_resources(struct efx_nic *efx) efx->max_tx_channels = nic_data->n_allocated_vis / EFX_TXQ_TYPES; - efx_ef10_free_vis(efx); + efx_mcdi_free_vis(efx); return -EAGAIN; } @@ -1490,7 +1318,7 @@ static int efx_ef10_init_nic(struct efx_nic *efx) return 0; } -static void efx_ef10_reset_mc_allocations(struct efx_nic *efx) +static void efx_ef10_table_reset_mc_allocations(struct efx_nic *efx) { struct efx_ef10_nic_data *nic_data = efx->nic_data; #ifdef CONFIG_SFC_SRIOV @@ -1503,7 +1331,7 @@ static void efx_ef10_reset_mc_allocations(struct efx_nic *efx) nic_data->must_restore_filters = true; nic_data->must_restore_piobufs = true; efx_ef10_forget_old_piobufs(efx); - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; /* Driver-created vswitches and vports must be re-created */ nic_data->must_probe_vswitching = true; @@ -1571,7 +1399,7 @@ static int efx_ef10_reset(struct efx_nic *efx, enum reset_type reset_type) */ if ((reset_type == RESET_TYPE_ALL || reset_type == RESET_TYPE_MCDI_TIMEOUT) && !rc) - efx_ef10_reset_mc_allocations(efx); + efx_ef10_table_reset_mc_allocations(efx); return rc; } @@ -2187,7 +2015,7 @@ static void efx_ef10_mcdi_reboot_detected(struct efx_nic *efx) struct efx_ef10_nic_data *nic_data = efx->nic_data; /* All our allocations have been reset */ - efx_ef10_reset_mc_allocations(efx); + efx_ef10_table_reset_mc_allocations(efx); /* The datapath firmware might have been changed */ nic_data->must_check_datapath_caps = true; @@ -2408,20 +2236,15 @@ static u32 efx_ef10_tso_versions(struct efx_nic *efx) static void efx_ef10_tx_init(struct efx_tx_queue *tx_queue) { - MCDI_DECLARE_BUF(inbuf, MC_CMD_INIT_TXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / - EFX_BUF_SIZE)); bool csum_offload = tx_queue->queue & EFX_TXQ_TYPE_OFFLOAD; - size_t entries = tx_queue->txd.buf.len / EFX_BUF_SIZE; struct efx_channel *channel = tx_queue->channel; struct efx_nic *efx = tx_queue->efx; - struct efx_ef10_nic_data *nic_data = efx->nic_data; + struct efx_ef10_nic_data *nic_data; bool tso_v2 = false; - size_t inlen; - dma_addr_t dma_addr; efx_qword_t *txd; int rc; - int i; - BUILD_BUG_ON(MC_CMD_INIT_TXQ_OUT_LEN != 0); + + nic_data = efx->nic_data; /* Only attempt to enable TX timestamping if we have the license for it, * otherwise TXQ init will fail @@ -2448,51 +2271,9 @@ static void efx_ef10_tx_init(struct efx_tx_queue *tx_queue) channel->channel); } - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_SIZE, tx_queue->ptr_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_TARGET_EVQ, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_LABEL, tx_queue->queue); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_INSTANCE, tx_queue->queue); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_OWNER_ID, 0); - MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_PORT_ID, nic_data->vport_id); - - dma_addr = tx_queue->txd.buf.dma_addr; - - netif_dbg(efx, hw, efx->net_dev, "pushing TXQ %d. %zu entries (%llx)\n", - tx_queue->queue, entries, (u64)dma_addr); - - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_TXQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_TXQ_IN_LEN(entries); - - do { - MCDI_POPULATE_DWORD_4(inbuf, INIT_TXQ_IN_FLAGS, - /* This flag was removed from mcdi_pcol.h for - * the non-_EXT version of INIT_TXQ. However, - * firmware still honours it. - */ - INIT_TXQ_EXT_IN_FLAG_TSOV2_EN, tso_v2, - INIT_TXQ_IN_FLAG_IP_CSUM_DIS, !csum_offload, - INIT_TXQ_IN_FLAG_TCP_CSUM_DIS, !csum_offload, - INIT_TXQ_EXT_IN_FLAG_TIMESTAMP, - tx_queue->timestamping); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_INIT_TXQ, inbuf, inlen, - NULL, 0, NULL); - if (rc == -ENOSPC && tso_v2) { - /* Retry without TSOv2 if we're short on contexts. */ - tso_v2 = false; - netif_warn(efx, probe, efx->net_dev, - "TSOv2 context not available to segment in hardware. TCP performance may be reduced.\n"); - } else if (rc) { - efx_mcdi_display_error(efx, MC_CMD_INIT_TXQ, - MC_CMD_INIT_TXQ_EXT_IN_LEN, - NULL, 0, rc); - goto fail; - } - } while (rc); + rc = efx_mcdi_tx_init(tx_queue, tso_v2); + if (rc) + goto fail; /* A previous user of this TX queue might have set us up the * bomb by writing a descriptor to the TX push collector but @@ -2530,35 +2311,6 @@ fail: tx_queue->queue); } -static void efx_ef10_tx_fini(struct efx_tx_queue *tx_queue) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_TXQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = tx_queue->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_TXQ_IN_INSTANCE, - tx_queue->queue); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_TXQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_TXQ, MC_CMD_FINI_TXQ_IN_LEN, - outbuf, outlen, rc); -} - -static void efx_ef10_tx_remove(struct efx_tx_queue *tx_queue) -{ - efx_nic_free_buffer(tx_queue->efx, &tx_queue->txd.buf); -} - /* This writes to the TX_DESC_WPTR; write pointer for TX descriptor ring */ static inline void efx_ef10_notify_tx_desc(struct efx_tx_queue *tx_queue) { @@ -2637,527 +2389,6 @@ static void efx_ef10_tx_write(struct efx_tx_queue *tx_queue) } } -#define RSS_MODE_HASH_ADDRS (1 << RSS_MODE_HASH_SRC_ADDR_LBN |\ - 1 << RSS_MODE_HASH_DST_ADDR_LBN) -#define RSS_MODE_HASH_PORTS (1 << RSS_MODE_HASH_SRC_PORT_LBN |\ - 1 << RSS_MODE_HASH_DST_PORT_LBN) -#define RSS_CONTEXT_FLAGS_DEFAULT (1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_IPV4_EN_LBN |\ - 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_TCPV4_EN_LBN |\ - 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_IPV6_EN_LBN |\ - 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_TCPV6_EN_LBN |\ - (RSS_MODE_HASH_ADDRS | RSS_MODE_HASH_PORTS) << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TCP_IPV4_RSS_MODE_LBN |\ - RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV4_RSS_MODE_LBN |\ - RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_OTHER_IPV4_RSS_MODE_LBN |\ - (RSS_MODE_HASH_ADDRS | RSS_MODE_HASH_PORTS) << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TCP_IPV6_RSS_MODE_LBN |\ - RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV6_RSS_MODE_LBN |\ - RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_OTHER_IPV6_RSS_MODE_LBN) - -static int efx_ef10_get_rss_flags(struct efx_nic *efx, u32 context, u32 *flags) -{ - /* Firmware had a bug (sfc bug 61952) where it would not actually - * fill in the flags field in the response to MC_CMD_RSS_CONTEXT_GET_FLAGS. - * This meant that it would always contain whatever was previously - * in the MCDI buffer. Fortunately, all firmware versions with - * this bug have the same default flags value for a newly-allocated - * RSS context, and the only time we want to get the flags is just - * after allocating. Moreover, the response has a 32-bit hole - * where the context ID would be in the request, so we can use an - * overlength buffer in the request and pre-fill the flags field - * with what we believe the default to be. Thus if the firmware - * has the bug, it will leave our pre-filled value in the flags - * field of the response, and we will get the right answer. - * - * However, this does mean that this function should NOT be used if - * the RSS context flags might not be their defaults - it is ONLY - * reliably correct for a newly-allocated RSS context. - */ - MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN); - size_t outlen; - int rc; - - /* Check we have a hole for the context ID */ - BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_GET_FLAGS_IN_LEN != MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_FLAGS_OFST); - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_FLAGS_IN_RSS_CONTEXT_ID, context); - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_FLAGS_OUT_FLAGS, - RSS_CONTEXT_FLAGS_DEFAULT); - rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_FLAGS, inbuf, - sizeof(inbuf), outbuf, sizeof(outbuf), &outlen); - if (rc == 0) { - if (outlen < MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN) - rc = -EIO; - else - *flags = MCDI_DWORD(outbuf, RSS_CONTEXT_GET_FLAGS_OUT_FLAGS); - } - return rc; -} - -/* Attempt to enable 4-tuple UDP hashing on the specified RSS context. - * If we fail, we just leave the RSS context at its default hash settings, - * which is safe but may slightly reduce performance. - * Defaults are 4-tuple for TCP and 2-tuple for UDP and other-IP, so we - * just need to set the UDP ports flags (for both IP versions). - */ -static void efx_ef10_set_rss_flags(struct efx_nic *efx, - struct efx_rss_context *ctx) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_SET_FLAGS_IN_LEN); - u32 flags; - - BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_SET_FLAGS_OUT_LEN != 0); - - if (efx_ef10_get_rss_flags(efx, ctx->context_id, &flags) != 0) - return; - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_SET_FLAGS_IN_RSS_CONTEXT_ID, - ctx->context_id); - flags |= RSS_MODE_HASH_PORTS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV4_RSS_MODE_LBN; - flags |= RSS_MODE_HASH_PORTS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV6_RSS_MODE_LBN; - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_SET_FLAGS_IN_FLAGS, flags); - if (!efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_FLAGS, inbuf, sizeof(inbuf), - NULL, 0, NULL)) - /* Succeeded, so UDP 4-tuple is now enabled */ - ctx->rx_hash_udp_4tuple = true; -} - -static int efx_ef10_alloc_rss_context(struct efx_nic *efx, bool exclusive, - struct efx_rss_context *ctx, - unsigned *context_size) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_ALLOC_IN_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_RSS_CONTEXT_ALLOC_OUT_LEN); - struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t outlen; - int rc; - u32 alloc_type = exclusive ? - MC_CMD_RSS_CONTEXT_ALLOC_IN_TYPE_EXCLUSIVE : - MC_CMD_RSS_CONTEXT_ALLOC_IN_TYPE_SHARED; - unsigned rss_spread = exclusive ? - efx->rss_spread : - min(rounddown_pow_of_two(efx->rss_spread), - EFX_EF10_MAX_SHARED_RSS_CONTEXT_SIZE); - - if (!exclusive && rss_spread == 1) { - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; - if (context_size) - *context_size = 1; - return 0; - } - - if (nic_data->datapath_caps & - 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_RSS_LIMITED_LBN) - return -EOPNOTSUPP; - - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_UPSTREAM_PORT_ID, - nic_data->vport_id); - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_TYPE, alloc_type); - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_NUM_QUEUES, rss_spread); - - rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_ALLOC, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - if (rc != 0) - return rc; - - if (outlen < MC_CMD_RSS_CONTEXT_ALLOC_OUT_LEN) - return -EIO; - - ctx->context_id = MCDI_DWORD(outbuf, RSS_CONTEXT_ALLOC_OUT_RSS_CONTEXT_ID); - - if (context_size) - *context_size = rss_spread; - - if (nic_data->datapath_caps & - 1 << MC_CMD_GET_CAPABILITIES_OUT_ADDITIONAL_RSS_MODES_LBN) - efx_ef10_set_rss_flags(efx, ctx); - - return 0; -} - -static int efx_ef10_free_rss_context(struct efx_nic *efx, u32 context) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_FREE_IN_LEN); - - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_FREE_IN_RSS_CONTEXT_ID, - context); - return efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_FREE, inbuf, sizeof(inbuf), - NULL, 0, NULL); -} - -static int efx_ef10_populate_rss_table(struct efx_nic *efx, u32 context, - const u32 *rx_indir_table, const u8 *key) -{ - MCDI_DECLARE_BUF(tablebuf, MC_CMD_RSS_CONTEXT_SET_TABLE_IN_LEN); - MCDI_DECLARE_BUF(keybuf, MC_CMD_RSS_CONTEXT_SET_KEY_IN_LEN); - int i, rc; - - MCDI_SET_DWORD(tablebuf, RSS_CONTEXT_SET_TABLE_IN_RSS_CONTEXT_ID, - context); - BUILD_BUG_ON(ARRAY_SIZE(efx->rss_context.rx_indir_table) != - MC_CMD_RSS_CONTEXT_SET_TABLE_IN_INDIRECTION_TABLE_LEN); - - /* This iterates over the length of efx->rss_context.rx_indir_table, but - * copies bytes from rx_indir_table. That's because the latter is a - * pointer rather than an array, but should have the same length. - * The efx->rss_context.rx_hash_key loop below is similar. - */ - for (i = 0; i < ARRAY_SIZE(efx->rss_context.rx_indir_table); ++i) - MCDI_PTR(tablebuf, - RSS_CONTEXT_SET_TABLE_IN_INDIRECTION_TABLE)[i] = - (u8) rx_indir_table[i]; - - rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_TABLE, tablebuf, - sizeof(tablebuf), NULL, 0, NULL); - if (rc != 0) - return rc; - - MCDI_SET_DWORD(keybuf, RSS_CONTEXT_SET_KEY_IN_RSS_CONTEXT_ID, - context); - BUILD_BUG_ON(ARRAY_SIZE(efx->rss_context.rx_hash_key) != - MC_CMD_RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY_LEN); - for (i = 0; i < ARRAY_SIZE(efx->rss_context.rx_hash_key); ++i) - MCDI_PTR(keybuf, RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY)[i] = key[i]; - - return efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_KEY, keybuf, - sizeof(keybuf), NULL, 0, NULL); -} - -static void efx_ef10_rx_free_indir_table(struct efx_nic *efx) -{ - int rc; - - if (efx->rss_context.context_id != EFX_EF10_RSS_CONTEXT_INVALID) { - rc = efx_ef10_free_rss_context(efx, efx->rss_context.context_id); - WARN_ON(rc != 0); - } - efx->rss_context.context_id = EFX_EF10_RSS_CONTEXT_INVALID; -} - -static int efx_ef10_rx_push_shared_rss_config(struct efx_nic *efx, - unsigned *context_size) -{ - struct efx_ef10_nic_data *nic_data = efx->nic_data; - int rc = efx_ef10_alloc_rss_context(efx, false, &efx->rss_context, - context_size); - - if (rc != 0) - return rc; - - nic_data->rx_rss_context_exclusive = false; - efx_set_default_rx_indir_table(efx, &efx->rss_context); - return 0; -} - -static int efx_ef10_rx_push_exclusive_rss_config(struct efx_nic *efx, - const u32 *rx_indir_table, - const u8 *key) -{ - u32 old_rx_rss_context = efx->rss_context.context_id; - struct efx_ef10_nic_data *nic_data = efx->nic_data; - int rc; - - if (efx->rss_context.context_id == EFX_EF10_RSS_CONTEXT_INVALID || - !nic_data->rx_rss_context_exclusive) { - rc = efx_ef10_alloc_rss_context(efx, true, &efx->rss_context, - NULL); - if (rc == -EOPNOTSUPP) - return rc; - else if (rc != 0) - goto fail1; - } - - rc = efx_ef10_populate_rss_table(efx, efx->rss_context.context_id, - rx_indir_table, key); - if (rc != 0) - goto fail2; - - if (efx->rss_context.context_id != old_rx_rss_context && - old_rx_rss_context != EFX_EF10_RSS_CONTEXT_INVALID) - WARN_ON(efx_ef10_free_rss_context(efx, old_rx_rss_context) != 0); - nic_data->rx_rss_context_exclusive = true; - if (rx_indir_table != efx->rss_context.rx_indir_table) - memcpy(efx->rss_context.rx_indir_table, rx_indir_table, - sizeof(efx->rss_context.rx_indir_table)); - if (key != efx->rss_context.rx_hash_key) - memcpy(efx->rss_context.rx_hash_key, key, - efx->type->rx_hash_key_size); - - return 0; - -fail2: - if (old_rx_rss_context != efx->rss_context.context_id) { - WARN_ON(efx_ef10_free_rss_context(efx, efx->rss_context.context_id) != 0); - efx->rss_context.context_id = old_rx_rss_context; - } -fail1: - netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); - return rc; -} - -static int efx_ef10_rx_push_rss_context_config(struct efx_nic *efx, - struct efx_rss_context *ctx, - const u32 *rx_indir_table, - const u8 *key) -{ - int rc; - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { - rc = efx_ef10_alloc_rss_context(efx, true, ctx, NULL); - if (rc) - return rc; - } - - if (!rx_indir_table) /* Delete this context */ - return efx_ef10_free_rss_context(efx, ctx->context_id); - - rc = efx_ef10_populate_rss_table(efx, ctx->context_id, - rx_indir_table, key); - if (rc) - return rc; - - memcpy(ctx->rx_indir_table, rx_indir_table, - sizeof(efx->rss_context.rx_indir_table)); - memcpy(ctx->rx_hash_key, key, efx->type->rx_hash_key_size); - - return 0; -} - -static int efx_ef10_rx_pull_rss_context_config(struct efx_nic *efx, - struct efx_rss_context *ctx) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_GET_TABLE_IN_LEN); - MCDI_DECLARE_BUF(tablebuf, MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_LEN); - MCDI_DECLARE_BUF(keybuf, MC_CMD_RSS_CONTEXT_GET_KEY_OUT_LEN); - size_t outlen; - int rc, i; - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_GET_TABLE_IN_LEN != - MC_CMD_RSS_CONTEXT_GET_KEY_IN_LEN); - - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) - return -ENOENT; - - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_TABLE_IN_RSS_CONTEXT_ID, - ctx->context_id); - BUILD_BUG_ON(ARRAY_SIZE(ctx->rx_indir_table) != - MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_INDIRECTION_TABLE_LEN); - rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_TABLE, inbuf, sizeof(inbuf), - tablebuf, sizeof(tablebuf), &outlen); - if (rc != 0) - return rc; - - if (WARN_ON(outlen != MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_LEN)) - return -EIO; - - for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) - ctx->rx_indir_table[i] = MCDI_PTR(tablebuf, - RSS_CONTEXT_GET_TABLE_OUT_INDIRECTION_TABLE)[i]; - - MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_KEY_IN_RSS_CONTEXT_ID, - ctx->context_id); - BUILD_BUG_ON(ARRAY_SIZE(ctx->rx_hash_key) != - MC_CMD_RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY_LEN); - rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_KEY, inbuf, sizeof(inbuf), - keybuf, sizeof(keybuf), &outlen); - if (rc != 0) - return rc; - - if (WARN_ON(outlen != MC_CMD_RSS_CONTEXT_GET_KEY_OUT_LEN)) - return -EIO; - - for (i = 0; i < ARRAY_SIZE(ctx->rx_hash_key); ++i) - ctx->rx_hash_key[i] = MCDI_PTR( - keybuf, RSS_CONTEXT_GET_KEY_OUT_TOEPLITZ_KEY)[i]; - - return 0; -} - -static int efx_ef10_rx_pull_rss_config(struct efx_nic *efx) -{ - int rc; - - mutex_lock(&efx->rss_lock); - rc = efx_ef10_rx_pull_rss_context_config(efx, &efx->rss_context); - mutex_unlock(&efx->rss_lock); - return rc; -} - -static void efx_ef10_rx_restore_rss_contexts(struct efx_nic *efx) -{ - struct efx_ef10_nic_data *nic_data = efx->nic_data; - struct efx_rss_context *ctx; - int rc; - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - if (!nic_data->must_restore_rss_contexts) - return; - - list_for_each_entry(ctx, &efx->rss_context.list, list) { - /* previous NIC RSS context is gone */ - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; - /* so try to allocate a new one */ - rc = efx_ef10_rx_push_rss_context_config(efx, ctx, - ctx->rx_indir_table, - ctx->rx_hash_key); - if (rc) - netif_warn(efx, probe, efx->net_dev, - "failed to restore RSS context %u, rc=%d" - "; RSS filters may fail to be applied\n", - ctx->user_id, rc); - } - nic_data->must_restore_rss_contexts = false; -} - -static int efx_ef10_pf_rx_push_rss_config(struct efx_nic *efx, bool user, - const u32 *rx_indir_table, - const u8 *key) -{ - int rc; - - if (efx->rss_spread == 1) - return 0; - - if (!key) - key = efx->rss_context.rx_hash_key; - - rc = efx_ef10_rx_push_exclusive_rss_config(efx, rx_indir_table, key); - - if (rc == -ENOBUFS && !user) { - unsigned context_size; - bool mismatch = false; - size_t i; - - for (i = 0; - i < ARRAY_SIZE(efx->rss_context.rx_indir_table) && !mismatch; - i++) - mismatch = rx_indir_table[i] != - ethtool_rxfh_indir_default(i, efx->rss_spread); - - rc = efx_ef10_rx_push_shared_rss_config(efx, &context_size); - if (rc == 0) { - if (context_size != efx->rss_spread) - netif_warn(efx, probe, efx->net_dev, - "Could not allocate an exclusive RSS" - " context; allocated a shared one of" - " different size." - " Wanted %u, got %u.\n", - efx->rss_spread, context_size); - else if (mismatch) - netif_warn(efx, probe, efx->net_dev, - "Could not allocate an exclusive RSS" - " context; allocated a shared one but" - " could not apply custom" - " indirection.\n"); - else - netif_info(efx, probe, efx->net_dev, - "Could not allocate an exclusive RSS" - " context; allocated a shared one.\n"); - } - } - return rc; -} - -static int efx_ef10_vf_rx_push_rss_config(struct efx_nic *efx, bool user, - const u32 *rx_indir_table - __attribute__ ((unused)), - const u8 *key - __attribute__ ((unused))) -{ - if (user) - return -EOPNOTSUPP; - if (efx->rss_context.context_id != EFX_EF10_RSS_CONTEXT_INVALID) - return 0; - return efx_ef10_rx_push_shared_rss_config(efx, NULL); -} - -static int efx_ef10_rx_probe(struct efx_rx_queue *rx_queue) -{ - return efx_nic_alloc_buffer(rx_queue->efx, &rx_queue->rxd.buf, - (rx_queue->ptr_mask + 1) * - sizeof(efx_qword_t), - GFP_KERNEL); -} - -static void efx_ef10_rx_init(struct efx_rx_queue *rx_queue) -{ - MCDI_DECLARE_BUF(inbuf, - MC_CMD_INIT_RXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / - EFX_BUF_SIZE)); - struct efx_channel *channel = efx_rx_queue_channel(rx_queue); - size_t entries = rx_queue->rxd.buf.len / EFX_BUF_SIZE; - struct efx_nic *efx = rx_queue->efx; - struct efx_ef10_nic_data *nic_data = efx->nic_data; - size_t inlen; - dma_addr_t dma_addr; - int rc; - int i; - BUILD_BUG_ON(MC_CMD_INIT_RXQ_OUT_LEN != 0); - - rx_queue->scatter_n = 0; - rx_queue->scatter_len = 0; - - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_SIZE, rx_queue->ptr_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_TARGET_EVQ, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_LABEL, efx_rx_queue_index(rx_queue)); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_INSTANCE, - efx_rx_queue_index(rx_queue)); - MCDI_POPULATE_DWORD_2(inbuf, INIT_RXQ_IN_FLAGS, - INIT_RXQ_IN_FLAG_PREFIX, 1, - INIT_RXQ_IN_FLAG_TIMESTAMP, 1); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_OWNER_ID, 0); - MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_PORT_ID, nic_data->vport_id); - - dma_addr = rx_queue->rxd.buf.dma_addr; - - netif_dbg(efx, hw, efx->net_dev, "pushing RXQ %d. %zu entries (%llx)\n", - efx_rx_queue_index(rx_queue), entries, (u64)dma_addr); - - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_RXQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_RXQ_IN_LEN(entries); - - rc = efx_mcdi_rpc(efx, MC_CMD_INIT_RXQ, inbuf, inlen, - NULL, 0, NULL); - if (rc) - netdev_WARN(efx->net_dev, "failed to initialise RXQ %d\n", - efx_rx_queue_index(rx_queue)); -} - -static void efx_ef10_rx_fini(struct efx_rx_queue *rx_queue) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_RXQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = rx_queue->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_RXQ_IN_INSTANCE, - efx_rx_queue_index(rx_queue)); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_RXQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_RXQ, MC_CMD_FINI_RXQ_IN_LEN, - outbuf, outlen, rc); -} - -static void efx_ef10_rx_remove(struct efx_rx_queue *rx_queue) -{ - efx_nic_free_buffer(rx_queue->efx, &rx_queue->rxd.buf); -} - /* This creates an entry in the RX descriptor queue */ static inline void efx_ef10_build_rx_desc(struct efx_rx_queue *rx_queue, unsigned int index) @@ -3229,106 +2460,20 @@ efx_ef10_rx_defer_refill_complete(struct efx_nic *efx, unsigned long cookie, /* nothing to do */ } -static int efx_ef10_ev_probe(struct efx_channel *channel) -{ - return efx_nic_alloc_buffer(channel->efx, &channel->eventq.buf, - (channel->eventq_mask + 1) * - sizeof(efx_qword_t), - GFP_KERNEL); -} - -static void efx_ef10_ev_fini(struct efx_channel *channel) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_EVQ_IN_LEN); - MCDI_DECLARE_BUF_ERR(outbuf); - struct efx_nic *efx = channel->efx; - size_t outlen; - int rc; - - MCDI_SET_DWORD(inbuf, FINI_EVQ_IN_INSTANCE, channel->channel); - - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_EVQ, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - - if (rc && rc != -EALREADY) - goto fail; - - return; - -fail: - efx_mcdi_display_error(efx, MC_CMD_FINI_EVQ, MC_CMD_FINI_EVQ_IN_LEN, - outbuf, outlen, rc); -} - static int efx_ef10_ev_init(struct efx_channel *channel) { - MCDI_DECLARE_BUF(inbuf, - MC_CMD_INIT_EVQ_V2_IN_LEN(EFX_MAX_EVQ_SIZE * 8 / - EFX_BUF_SIZE)); - MCDI_DECLARE_BUF(outbuf, MC_CMD_INIT_EVQ_V2_OUT_LEN); - size_t entries = channel->eventq.buf.len / EFX_BUF_SIZE; struct efx_nic *efx = channel->efx; struct efx_ef10_nic_data *nic_data; - size_t inlen, outlen; unsigned int enabled, implemented; - dma_addr_t dma_addr; + bool use_v2, cut_thru; int rc; - int i; nic_data = efx->nic_data; - - /* Fill event queue with all ones (i.e. empty events) */ - memset(channel->eventq.buf.addr, 0xff, channel->eventq.buf.len); - - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_SIZE, channel->eventq_mask + 1); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_INSTANCE, channel->channel); - /* INIT_EVQ expects index in vector table, not absolute */ - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_IRQ_NUM, channel->channel); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_MODE, - MC_CMD_INIT_EVQ_IN_TMR_MODE_DIS); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_LOAD, 0); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_RELOAD, 0); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_MODE, - MC_CMD_INIT_EVQ_IN_COUNT_MODE_DIS); - MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_THRSHLD, 0); - - if (nic_data->datapath_caps2 & - 1 << MC_CMD_GET_CAPABILITIES_V2_OUT_INIT_EVQ_V2_LBN) { - /* Use the new generic approach to specifying event queue - * configuration, requesting lower latency or higher throughput. - * The options that actually get used appear in the output. - */ - MCDI_POPULATE_DWORD_2(inbuf, INIT_EVQ_V2_IN_FLAGS, - INIT_EVQ_V2_IN_FLAG_INTERRUPTING, 1, - INIT_EVQ_V2_IN_FLAG_TYPE, - MC_CMD_INIT_EVQ_V2_IN_FLAG_TYPE_AUTO); - } else { - bool cut_thru = !(nic_data->datapath_caps & - 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_BATCHING_LBN); - - MCDI_POPULATE_DWORD_4(inbuf, INIT_EVQ_IN_FLAGS, - INIT_EVQ_IN_FLAG_INTERRUPTING, 1, - INIT_EVQ_IN_FLAG_RX_MERGE, 1, - INIT_EVQ_IN_FLAG_TX_MERGE, 1, - INIT_EVQ_IN_FLAG_CUT_THRU, cut_thru); - } - - dma_addr = channel->eventq.buf.dma_addr; - for (i = 0; i < entries; ++i) { - MCDI_SET_ARRAY_QWORD(inbuf, INIT_EVQ_IN_DMA_ADDR, i, dma_addr); - dma_addr += EFX_BUF_SIZE; - } - - inlen = MC_CMD_INIT_EVQ_IN_LEN(entries); - - rc = efx_mcdi_rpc(efx, MC_CMD_INIT_EVQ, inbuf, inlen, - outbuf, sizeof(outbuf), &outlen); - - if (outlen >= MC_CMD_INIT_EVQ_V2_OUT_LEN) - netif_dbg(efx, drv, efx->net_dev, - "Channel %d using event queue flags %08x\n", - channel->channel, - MCDI_DWORD(outbuf, INIT_EVQ_V2_OUT_FLAGS)); + use_v2 = nic_data->datapath_caps2 & + 1 << MC_CMD_GET_CAPABILITIES_V2_OUT_INIT_EVQ_V2_LBN; + cut_thru = !(nic_data->datapath_caps & + 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_BATCHING_LBN); + rc = efx_mcdi_ev_init(channel, cut_thru, use_v2); /* IRQ return is ignored */ if (channel->channel || rc) @@ -3386,15 +2531,10 @@ static int efx_ef10_ev_init(struct efx_channel *channel) return 0; fail: - efx_ef10_ev_fini(channel); + efx_mcdi_ev_fini(channel); return rc; } -static void efx_ef10_ev_remove(struct efx_channel *channel) -{ - efx_nic_free_buffer(channel->efx, &channel->eventq.buf); -} - static void efx_ef10_handle_rx_wrong_queue(struct efx_rx_queue *rx_queue, unsigned int rx_queue_label) { @@ -3976,9 +3116,9 @@ static int efx_ef10_fini_dmaq(struct efx_nic *efx) if (efx->state != STATE_RECOVERY) { efx_for_each_channel(channel, efx) { efx_for_each_channel_rx_queue(rx_queue, channel) - efx_ef10_rx_fini(rx_queue); + efx_mcdi_rx_fini(rx_queue); efx_for_each_channel_tx_queue(tx_queue, channel) - efx_ef10_tx_fini(tx_queue); + efx_mcdi_tx_fini(tx_queue); } wait_event_timeout(efx->flush_wq, @@ -4000,1538 +3140,6 @@ static void efx_ef10_prepare_flr(struct efx_nic *efx) atomic_set(&efx->active_queues, 0); } -/* Decide whether a filter should be exclusive or else should allow - * delivery to additional recipients. Currently we decide that - * filters for specific local unicast MAC and IP addresses are - * exclusive. - */ -static bool efx_ef10_filter_is_exclusive(const struct efx_filter_spec *spec) -{ - if (spec->match_flags & EFX_FILTER_MATCH_LOC_MAC && - !is_multicast_ether_addr(spec->loc_mac)) - return true; - - if ((spec->match_flags & - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { - if (spec->ether_type == htons(ETH_P_IP) && - !ipv4_is_multicast(spec->loc_host[0])) - return true; - if (spec->ether_type == htons(ETH_P_IPV6) && - ((const u8 *)spec->loc_host)[0] != 0xff) - return true; - } - - return false; -} - -static struct efx_filter_spec * -efx_ef10_filter_entry_spec(const struct efx_ef10_filter_table *table, - unsigned int filter_idx) -{ - return (struct efx_filter_spec *)(table->entry[filter_idx].spec & - ~EFX_EF10_FILTER_FLAGS); -} - -static unsigned int -efx_ef10_filter_entry_flags(const struct efx_ef10_filter_table *table, - unsigned int filter_idx) -{ - return table->entry[filter_idx].spec & EFX_EF10_FILTER_FLAGS; -} - -static void -efx_ef10_filter_set_entry(struct efx_ef10_filter_table *table, - unsigned int filter_idx, - const struct efx_filter_spec *spec, - unsigned int flags) -{ - table->entry[filter_idx].spec = (unsigned long)spec | flags; -} - -static void -efx_ef10_filter_push_prep_set_match_fields(struct efx_nic *efx, - const struct efx_filter_spec *spec, - efx_dword_t *inbuf) -{ - enum efx_encap_type encap_type = efx_filter_get_encap_type(spec); - u32 match_fields = 0, uc_match, mc_match; - - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, - efx_ef10_filter_is_exclusive(spec) ? - MC_CMD_FILTER_OP_IN_OP_INSERT : - MC_CMD_FILTER_OP_IN_OP_SUBSCRIBE); - - /* Convert match flags and values. Unlike almost - * everything else in MCDI, these fields are in - * network byte order. - */ -#define COPY_VALUE(value, mcdi_field) \ - do { \ - match_fields |= \ - 1 << MC_CMD_FILTER_OP_IN_MATCH_ ## \ - mcdi_field ## _LBN; \ - BUILD_BUG_ON( \ - MC_CMD_FILTER_OP_IN_ ## mcdi_field ## _LEN < \ - sizeof(value)); \ - memcpy(MCDI_PTR(inbuf, FILTER_OP_IN_ ## mcdi_field), \ - &value, sizeof(value)); \ - } while (0) -#define COPY_FIELD(gen_flag, gen_field, mcdi_field) \ - if (spec->match_flags & EFX_FILTER_MATCH_ ## gen_flag) { \ - COPY_VALUE(spec->gen_field, mcdi_field); \ - } - /* Handle encap filters first. They will always be mismatch - * (unknown UC or MC) filters - */ - if (encap_type) { - /* ether_type and outer_ip_proto need to be variables - * because COPY_VALUE wants to memcpy them - */ - __be16 ether_type = - htons(encap_type & EFX_ENCAP_FLAG_IPV6 ? - ETH_P_IPV6 : ETH_P_IP); - u8 vni_type = MC_CMD_FILTER_OP_EXT_IN_VNI_TYPE_GENEVE; - u8 outer_ip_proto; - - switch (encap_type & EFX_ENCAP_TYPES_MASK) { - case EFX_ENCAP_TYPE_VXLAN: - vni_type = MC_CMD_FILTER_OP_EXT_IN_VNI_TYPE_VXLAN; - /* fallthrough */ - case EFX_ENCAP_TYPE_GENEVE: - COPY_VALUE(ether_type, ETHER_TYPE); - outer_ip_proto = IPPROTO_UDP; - COPY_VALUE(outer_ip_proto, IP_PROTO); - /* We always need to set the type field, even - * though we're not matching on the TNI. - */ - MCDI_POPULATE_DWORD_1(inbuf, - FILTER_OP_EXT_IN_VNI_OR_VSID, - FILTER_OP_EXT_IN_VNI_TYPE, - vni_type); - break; - case EFX_ENCAP_TYPE_NVGRE: - COPY_VALUE(ether_type, ETHER_TYPE); - outer_ip_proto = IPPROTO_GRE; - COPY_VALUE(outer_ip_proto, IP_PROTO); - break; - default: - WARN_ON(1); - } - - uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_UCAST_DST_LBN; - mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_MCAST_DST_LBN; - } else { - uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_UCAST_DST_LBN; - mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_MCAST_DST_LBN; - } - - if (spec->match_flags & EFX_FILTER_MATCH_LOC_MAC_IG) - match_fields |= - is_multicast_ether_addr(spec->loc_mac) ? - 1 << mc_match : - 1 << uc_match; - COPY_FIELD(REM_HOST, rem_host, SRC_IP); - COPY_FIELD(LOC_HOST, loc_host, DST_IP); - COPY_FIELD(REM_MAC, rem_mac, SRC_MAC); - COPY_FIELD(REM_PORT, rem_port, SRC_PORT); - COPY_FIELD(LOC_MAC, loc_mac, DST_MAC); - COPY_FIELD(LOC_PORT, loc_port, DST_PORT); - COPY_FIELD(ETHER_TYPE, ether_type, ETHER_TYPE); - COPY_FIELD(INNER_VID, inner_vid, INNER_VLAN); - COPY_FIELD(OUTER_VID, outer_vid, OUTER_VLAN); - COPY_FIELD(IP_PROTO, ip_proto, IP_PROTO); -#undef COPY_FIELD -#undef COPY_VALUE - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_MATCH_FIELDS, - match_fields); -} - -static void efx_ef10_filter_push_prep(struct efx_nic *efx, - const struct efx_filter_spec *spec, - efx_dword_t *inbuf, u64 handle, - struct efx_rss_context *ctx, - bool replacing) -{ - struct efx_ef10_nic_data *nic_data = efx->nic_data; - u32 flags = spec->flags; - - memset(inbuf, 0, MC_CMD_FILTER_OP_EXT_IN_LEN); - - /* If RSS filter, caller better have given us an RSS context */ - if (flags & EFX_FILTER_FLAG_RX_RSS) { - /* We don't have the ability to return an error, so we'll just - * log a warning and disable RSS for the filter. - */ - if (WARN_ON_ONCE(!ctx)) - flags &= ~EFX_FILTER_FLAG_RX_RSS; - else if (WARN_ON_ONCE(ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID)) - flags &= ~EFX_FILTER_FLAG_RX_RSS; - } - - if (replacing) { - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, - MC_CMD_FILTER_OP_IN_OP_REPLACE); - MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, handle); - } else { - efx_ef10_filter_push_prep_set_match_fields(efx, spec, inbuf); - } - - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_PORT_ID, nic_data->vport_id); - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_DEST, - spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP ? - MC_CMD_FILTER_OP_IN_RX_DEST_DROP : - MC_CMD_FILTER_OP_IN_RX_DEST_HOST); - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_TX_DOMAIN, 0); - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_TX_DEST, - MC_CMD_FILTER_OP_IN_TX_DEST_DEFAULT); - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_QUEUE, - spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP ? - 0 : spec->dmaq_id); - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_MODE, - (flags & EFX_FILTER_FLAG_RX_RSS) ? - MC_CMD_FILTER_OP_IN_RX_MODE_RSS : - MC_CMD_FILTER_OP_IN_RX_MODE_SIMPLE); - if (flags & EFX_FILTER_FLAG_RX_RSS) - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_CONTEXT, ctx->context_id); -} - -static int efx_ef10_filter_push(struct efx_nic *efx, - const struct efx_filter_spec *spec, u64 *handle, - struct efx_rss_context *ctx, bool replacing) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_FILTER_OP_EXT_OUT_LEN); - size_t outlen; - int rc; - - efx_ef10_filter_push_prep(efx, spec, inbuf, *handle, ctx, replacing); - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, inbuf, sizeof(inbuf), - outbuf, sizeof(outbuf), &outlen); - if (rc && spec->priority != EFX_FILTER_PRI_HINT) - efx_mcdi_display_error(efx, MC_CMD_FILTER_OP, sizeof(inbuf), - outbuf, outlen, rc); - if (rc == 0) - *handle = MCDI_QWORD(outbuf, FILTER_OP_OUT_HANDLE); - if (rc == -ENOSPC) - rc = -EBUSY; /* to match efx_farch_filter_insert() */ - return rc; -} - -static u32 efx_ef10_filter_mcdi_flags_from_spec(const struct efx_filter_spec *spec) -{ - enum efx_encap_type encap_type = efx_filter_get_encap_type(spec); - unsigned int match_flags = spec->match_flags; - unsigned int uc_match, mc_match; - u32 mcdi_flags = 0; - -#define MAP_FILTER_TO_MCDI_FLAG(gen_flag, mcdi_field, encap) { \ - unsigned int old_match_flags = match_flags; \ - match_flags &= ~EFX_FILTER_MATCH_ ## gen_flag; \ - if (match_flags != old_match_flags) \ - mcdi_flags |= \ - (1 << ((encap) ? \ - MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_ ## \ - mcdi_field ## _LBN : \ - MC_CMD_FILTER_OP_EXT_IN_MATCH_ ##\ - mcdi_field ## _LBN)); \ - } - /* inner or outer based on encap type */ - MAP_FILTER_TO_MCDI_FLAG(REM_HOST, SRC_IP, encap_type); - MAP_FILTER_TO_MCDI_FLAG(LOC_HOST, DST_IP, encap_type); - MAP_FILTER_TO_MCDI_FLAG(REM_MAC, SRC_MAC, encap_type); - MAP_FILTER_TO_MCDI_FLAG(REM_PORT, SRC_PORT, encap_type); - MAP_FILTER_TO_MCDI_FLAG(LOC_MAC, DST_MAC, encap_type); - MAP_FILTER_TO_MCDI_FLAG(LOC_PORT, DST_PORT, encap_type); - MAP_FILTER_TO_MCDI_FLAG(ETHER_TYPE, ETHER_TYPE, encap_type); - MAP_FILTER_TO_MCDI_FLAG(IP_PROTO, IP_PROTO, encap_type); - /* always outer */ - MAP_FILTER_TO_MCDI_FLAG(INNER_VID, INNER_VLAN, false); - MAP_FILTER_TO_MCDI_FLAG(OUTER_VID, OUTER_VLAN, false); -#undef MAP_FILTER_TO_MCDI_FLAG - - /* special handling for encap type, and mismatch */ - if (encap_type) { - match_flags &= ~EFX_FILTER_MATCH_ENCAP_TYPE; - mcdi_flags |= - (1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ETHER_TYPE_LBN); - mcdi_flags |= (1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_IP_PROTO_LBN); - - uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_UCAST_DST_LBN; - mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_MCAST_DST_LBN; - } else { - uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_UCAST_DST_LBN; - mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_MCAST_DST_LBN; - } - - if (match_flags & EFX_FILTER_MATCH_LOC_MAC_IG) { - match_flags &= ~EFX_FILTER_MATCH_LOC_MAC_IG; - mcdi_flags |= - is_multicast_ether_addr(spec->loc_mac) ? - 1 << mc_match : - 1 << uc_match; - } - - /* Did we map them all? */ - WARN_ON_ONCE(match_flags); - - return mcdi_flags; -} - -static int efx_ef10_filter_pri(struct efx_ef10_filter_table *table, - const struct efx_filter_spec *spec) -{ - u32 mcdi_flags = efx_ef10_filter_mcdi_flags_from_spec(spec); - unsigned int match_pri; - - for (match_pri = 0; - match_pri < table->rx_match_count; - match_pri++) - if (table->rx_match_mcdi_flags[match_pri] == mcdi_flags) - return match_pri; - - return -EPROTONOSUPPORT; -} - -static s32 efx_ef10_filter_insert_locked(struct efx_nic *efx, - struct efx_filter_spec *spec, - bool replace_equal) -{ - DECLARE_BITMAP(mc_rem_map, EFX_EF10_FILTER_SEARCH_LIMIT); - struct efx_ef10_nic_data *nic_data = efx->nic_data; - struct efx_ef10_filter_table *table; - struct efx_filter_spec *saved_spec; - struct efx_rss_context *ctx = NULL; - unsigned int match_pri, hash; - unsigned int priv_flags; - bool rss_locked = false; - bool replacing = false; - unsigned int depth, i; - int ins_index = -1; - DEFINE_WAIT(wait); - bool is_mc_recip; - s32 rc; - - WARN_ON(!rwsem_is_locked(&efx->filter_sem)); - table = efx->filter_state; - down_write(&table->lock); - - /* For now, only support RX filters */ - if ((spec->flags & (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX)) != - EFX_FILTER_FLAG_RX) { - rc = -EINVAL; - goto out_unlock; - } - - rc = efx_ef10_filter_pri(table, spec); - if (rc < 0) - goto out_unlock; - match_pri = rc; - - hash = efx_filter_spec_hash(spec); - is_mc_recip = efx_filter_is_mc_recipient(spec); - if (is_mc_recip) - bitmap_zero(mc_rem_map, EFX_EF10_FILTER_SEARCH_LIMIT); - - if (spec->flags & EFX_FILTER_FLAG_RX_RSS) { - mutex_lock(&efx->rss_lock); - rss_locked = true; - if (spec->rss_context) - ctx = efx_find_rss_context_entry(efx, spec->rss_context); - else - ctx = &efx->rss_context; - if (!ctx) { - rc = -ENOENT; - goto out_unlock; - } - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { - rc = -EOPNOTSUPP; - goto out_unlock; - } - } - - /* Find any existing filters with the same match tuple or - * else a free slot to insert at. - */ - for (depth = 1; depth < EFX_EF10_FILTER_SEARCH_LIMIT; depth++) { - i = (hash + depth) & (HUNT_FILTER_TBL_ROWS - 1); - saved_spec = efx_ef10_filter_entry_spec(table, i); - - if (!saved_spec) { - if (ins_index < 0) - ins_index = i; - } else if (efx_filter_spec_equal(spec, saved_spec)) { - if (spec->priority < saved_spec->priority && - spec->priority != EFX_FILTER_PRI_AUTO) { - rc = -EPERM; - goto out_unlock; - } - if (!is_mc_recip) { - /* This is the only one */ - if (spec->priority == - saved_spec->priority && - !replace_equal) { - rc = -EEXIST; - goto out_unlock; - } - ins_index = i; - break; - } else if (spec->priority > - saved_spec->priority || - (spec->priority == - saved_spec->priority && - replace_equal)) { - if (ins_index < 0) - ins_index = i; - else - __set_bit(depth, mc_rem_map); - } - } - } - - /* Once we reach the maximum search depth, use the first suitable - * slot, or return -EBUSY if there was none - */ - if (ins_index < 0) { - rc = -EBUSY; - goto out_unlock; - } - - /* Create a software table entry if necessary. */ - saved_spec = efx_ef10_filter_entry_spec(table, ins_index); - if (saved_spec) { - if (spec->priority == EFX_FILTER_PRI_AUTO && - saved_spec->priority >= EFX_FILTER_PRI_AUTO) { - /* Just make sure it won't be removed */ - if (saved_spec->priority > EFX_FILTER_PRI_AUTO) - saved_spec->flags |= EFX_FILTER_FLAG_RX_OVER_AUTO; - table->entry[ins_index].spec &= - ~EFX_EF10_FILTER_FLAG_AUTO_OLD; - rc = ins_index; - goto out_unlock; - } - replacing = true; - priv_flags = efx_ef10_filter_entry_flags(table, ins_index); - } else { - saved_spec = kmalloc(sizeof(*spec), GFP_ATOMIC); - if (!saved_spec) { - rc = -ENOMEM; - goto out_unlock; - } - *saved_spec = *spec; - priv_flags = 0; - } - efx_ef10_filter_set_entry(table, ins_index, saved_spec, priv_flags); - - /* Actually insert the filter on the HW */ - rc = efx_ef10_filter_push(efx, spec, &table->entry[ins_index].handle, - ctx, replacing); - - if (rc == -EINVAL && nic_data->must_realloc_vis) - /* The MC rebooted under us, causing it to reject our filter - * insertion as pointing to an invalid VI (spec->dmaq_id). - */ - rc = -EAGAIN; - - /* Finalise the software table entry */ - if (rc == 0) { - if (replacing) { - /* Update the fields that may differ */ - if (saved_spec->priority == EFX_FILTER_PRI_AUTO) - saved_spec->flags |= - EFX_FILTER_FLAG_RX_OVER_AUTO; - saved_spec->priority = spec->priority; - saved_spec->flags &= EFX_FILTER_FLAG_RX_OVER_AUTO; - saved_spec->flags |= spec->flags; - saved_spec->rss_context = spec->rss_context; - saved_spec->dmaq_id = spec->dmaq_id; - } - } else if (!replacing) { - kfree(saved_spec); - saved_spec = NULL; - } else { - /* We failed to replace, so the old filter is still present. - * Roll back the software table to reflect this. In fact the - * efx_ef10_filter_set_entry() call below will do the right - * thing, so nothing extra is needed here. - */ - } - efx_ef10_filter_set_entry(table, ins_index, saved_spec, priv_flags); - - /* Remove and finalise entries for lower-priority multicast - * recipients - */ - if (is_mc_recip) { - MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); - unsigned int depth, i; - - memset(inbuf, 0, sizeof(inbuf)); - - for (depth = 0; depth < EFX_EF10_FILTER_SEARCH_LIMIT; depth++) { - if (!test_bit(depth, mc_rem_map)) - continue; - - i = (hash + depth) & (HUNT_FILTER_TBL_ROWS - 1); - saved_spec = efx_ef10_filter_entry_spec(table, i); - priv_flags = efx_ef10_filter_entry_flags(table, i); - - if (rc == 0) { - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, - MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); - MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, - table->entry[i].handle); - rc = efx_mcdi_rpc(efx, MC_CMD_FILTER_OP, - inbuf, sizeof(inbuf), - NULL, 0, NULL); - } - - if (rc == 0) { - kfree(saved_spec); - saved_spec = NULL; - priv_flags = 0; - } - efx_ef10_filter_set_entry(table, i, saved_spec, - priv_flags); - } - } - - /* If successful, return the inserted filter ID */ - if (rc == 0) - rc = efx_ef10_make_filter_id(match_pri, ins_index); - -out_unlock: - if (rss_locked) - mutex_unlock(&efx->rss_lock); - up_write(&table->lock); - return rc; -} - -static s32 efx_ef10_filter_insert(struct efx_nic *efx, - struct efx_filter_spec *spec, - bool replace_equal) -{ - s32 ret; - - down_read(&efx->filter_sem); - ret = efx_ef10_filter_insert_locked(efx, spec, replace_equal); - up_read(&efx->filter_sem); - - return ret; -} - -static void efx_ef10_filter_update_rx_scatter(struct efx_nic *efx) -{ - /* no need to do anything here on EF10 */ -} - -/* Remove a filter. - * If !by_index, remove by ID - * If by_index, remove by index - * Filter ID may come from userland and must be range-checked. - * Caller must hold efx->filter_sem for read, and efx->filter_state->lock - * for write. - */ -static int efx_ef10_filter_remove_internal(struct efx_nic *efx, - unsigned int priority_mask, - u32 filter_id, bool by_index) -{ - unsigned int filter_idx = efx_ef10_filter_get_unsafe_id(filter_id); - struct efx_ef10_filter_table *table = efx->filter_state; - MCDI_DECLARE_BUF(inbuf, - MC_CMD_FILTER_OP_IN_HANDLE_OFST + - MC_CMD_FILTER_OP_IN_HANDLE_LEN); - struct efx_filter_spec *spec; - DEFINE_WAIT(wait); - int rc; - - spec = efx_ef10_filter_entry_spec(table, filter_idx); - if (!spec || - (!by_index && - efx_ef10_filter_pri(table, spec) != - efx_ef10_filter_get_unsafe_pri(filter_id))) - return -ENOENT; - - if (spec->flags & EFX_FILTER_FLAG_RX_OVER_AUTO && - priority_mask == (1U << EFX_FILTER_PRI_AUTO)) { - /* Just remove flags */ - spec->flags &= ~EFX_FILTER_FLAG_RX_OVER_AUTO; - table->entry[filter_idx].spec &= ~EFX_EF10_FILTER_FLAG_AUTO_OLD; - return 0; - } - - if (!(priority_mask & (1U << spec->priority))) - return -ENOENT; - - if (spec->flags & EFX_FILTER_FLAG_RX_OVER_AUTO) { - /* Reset to an automatic filter */ - - struct efx_filter_spec new_spec = *spec; - - new_spec.priority = EFX_FILTER_PRI_AUTO; - new_spec.flags = (EFX_FILTER_FLAG_RX | - (efx_rss_active(&efx->rss_context) ? - EFX_FILTER_FLAG_RX_RSS : 0)); - new_spec.dmaq_id = 0; - new_spec.rss_context = 0; - rc = efx_ef10_filter_push(efx, &new_spec, - &table->entry[filter_idx].handle, - &efx->rss_context, - true); - - if (rc == 0) - *spec = new_spec; - } else { - /* Really remove the filter */ - - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, - efx_ef10_filter_is_exclusive(spec) ? - MC_CMD_FILTER_OP_IN_OP_REMOVE : - MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); - MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, - table->entry[filter_idx].handle); - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, - inbuf, sizeof(inbuf), NULL, 0, NULL); - - if ((rc == 0) || (rc == -ENOENT)) { - /* Filter removed OK or didn't actually exist */ - kfree(spec); - efx_ef10_filter_set_entry(table, filter_idx, NULL, 0); - } else { - efx_mcdi_display_error(efx, MC_CMD_FILTER_OP, - MC_CMD_FILTER_OP_EXT_IN_LEN, - NULL, 0, rc); - } - } - - return rc; -} - -static int efx_ef10_filter_remove_safe(struct efx_nic *efx, - enum efx_filter_priority priority, - u32 filter_id) -{ - struct efx_ef10_filter_table *table; - int rc; - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_write(&table->lock); - rc = efx_ef10_filter_remove_internal(efx, 1U << priority, filter_id, - false); - up_write(&table->lock); - up_read(&efx->filter_sem); - return rc; -} - -/* Caller must hold efx->filter_sem for read */ -static void efx_ef10_filter_remove_unsafe(struct efx_nic *efx, - enum efx_filter_priority priority, - u32 filter_id) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - - if (filter_id == EFX_EF10_FILTER_ID_INVALID) - return; - - down_write(&table->lock); - efx_ef10_filter_remove_internal(efx, 1U << priority, filter_id, - true); - up_write(&table->lock); -} - -static int efx_ef10_filter_get_safe(struct efx_nic *efx, - enum efx_filter_priority priority, - u32 filter_id, struct efx_filter_spec *spec) -{ - unsigned int filter_idx = efx_ef10_filter_get_unsafe_id(filter_id); - const struct efx_filter_spec *saved_spec; - struct efx_ef10_filter_table *table; - int rc; - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_read(&table->lock); - saved_spec = efx_ef10_filter_entry_spec(table, filter_idx); - if (saved_spec && saved_spec->priority == priority && - efx_ef10_filter_pri(table, saved_spec) == - efx_ef10_filter_get_unsafe_pri(filter_id)) { - *spec = *saved_spec; - rc = 0; - } else { - rc = -ENOENT; - } - up_read(&table->lock); - up_read(&efx->filter_sem); - return rc; -} - -static int efx_ef10_filter_clear_rx(struct efx_nic *efx, - enum efx_filter_priority priority) -{ - struct efx_ef10_filter_table *table; - unsigned int priority_mask; - unsigned int i; - int rc; - - priority_mask = (((1U << (priority + 1)) - 1) & - ~(1U << EFX_FILTER_PRI_AUTO)); - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_write(&table->lock); - for (i = 0; i < HUNT_FILTER_TBL_ROWS; i++) { - rc = efx_ef10_filter_remove_internal(efx, priority_mask, - i, true); - if (rc && rc != -ENOENT) - break; - rc = 0; - } - - up_write(&table->lock); - up_read(&efx->filter_sem); - return rc; -} - -static u32 efx_ef10_filter_count_rx_used(struct efx_nic *efx, - enum efx_filter_priority priority) -{ - struct efx_ef10_filter_table *table; - unsigned int filter_idx; - s32 count = 0; - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_read(&table->lock); - for (filter_idx = 0; filter_idx < HUNT_FILTER_TBL_ROWS; filter_idx++) { - if (table->entry[filter_idx].spec && - efx_ef10_filter_entry_spec(table, filter_idx)->priority == - priority) - ++count; - } - up_read(&table->lock); - up_read(&efx->filter_sem); - return count; -} - -static u32 efx_ef10_filter_get_rx_id_limit(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - - return table->rx_match_count * HUNT_FILTER_TBL_ROWS * 2; -} - -static s32 efx_ef10_filter_get_rx_ids(struct efx_nic *efx, - enum efx_filter_priority priority, - u32 *buf, u32 size) -{ - struct efx_ef10_filter_table *table; - struct efx_filter_spec *spec; - unsigned int filter_idx; - s32 count = 0; - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_read(&table->lock); - - for (filter_idx = 0; filter_idx < HUNT_FILTER_TBL_ROWS; filter_idx++) { - spec = efx_ef10_filter_entry_spec(table, filter_idx); - if (spec && spec->priority == priority) { - if (count == size) { - count = -EMSGSIZE; - break; - } - buf[count++] = - efx_ef10_make_filter_id( - efx_ef10_filter_pri(table, spec), - filter_idx); - } - } - up_read(&table->lock); - up_read(&efx->filter_sem); - return count; -} - -#ifdef CONFIG_RFS_ACCEL - -static bool efx_ef10_filter_rfs_expire_one(struct efx_nic *efx, u32 flow_id, - unsigned int filter_idx) -{ - struct efx_filter_spec *spec, saved_spec; - struct efx_ef10_filter_table *table; - struct efx_arfs_rule *rule = NULL; - bool ret = true, force = false; - u16 arfs_id; - - down_read(&efx->filter_sem); - table = efx->filter_state; - down_write(&table->lock); - spec = efx_ef10_filter_entry_spec(table, filter_idx); - - if (!spec || spec->priority != EFX_FILTER_PRI_HINT) - goto out_unlock; - - spin_lock_bh(&efx->rps_hash_lock); - if (!efx->rps_hash_table) { - /* In the absence of the table, we always return 0 to ARFS. */ - arfs_id = 0; - } else { - rule = efx_rps_hash_find(efx, spec); - if (!rule) - /* ARFS table doesn't know of this filter, so remove it */ - goto expire; - arfs_id = rule->arfs_id; - ret = efx_rps_check_rule(rule, filter_idx, &force); - if (force) - goto expire; - if (!ret) { - spin_unlock_bh(&efx->rps_hash_lock); - goto out_unlock; - } - } - if (!rps_may_expire_flow(efx->net_dev, spec->dmaq_id, flow_id, arfs_id)) - ret = false; - else if (rule) - rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; -expire: - saved_spec = *spec; /* remove operation will kfree spec */ - spin_unlock_bh(&efx->rps_hash_lock); - /* At this point (since we dropped the lock), another thread might queue - * up a fresh insertion request (but the actual insertion will be held - * up by our possession of the filter table lock). In that case, it - * will set rule->filter_id to EFX_ARFS_FILTER_ID_PENDING, meaning that - * the rule is not removed by efx_rps_hash_del() below. - */ - if (ret) - ret = efx_ef10_filter_remove_internal(efx, 1U << spec->priority, - filter_idx, true) == 0; - /* While we can't safely dereference rule (we dropped the lock), we can - * still test it for NULL. - */ - if (ret && rule) { - /* Expiring, so remove entry from ARFS table */ - spin_lock_bh(&efx->rps_hash_lock); - efx_rps_hash_del(efx, &saved_spec); - spin_unlock_bh(&efx->rps_hash_lock); - } -out_unlock: - up_write(&table->lock); - up_read(&efx->filter_sem); - return ret; -} - -#endif /* CONFIG_RFS_ACCEL */ - -static int efx_ef10_filter_match_flags_from_mcdi(bool encap, u32 mcdi_flags) -{ - int match_flags = 0; - -#define MAP_FLAG(gen_flag, mcdi_field) do { \ - u32 old_mcdi_flags = mcdi_flags; \ - mcdi_flags &= ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ ## \ - mcdi_field ## _LBN); \ - if (mcdi_flags != old_mcdi_flags) \ - match_flags |= EFX_FILTER_MATCH_ ## gen_flag; \ - } while (0) - - if (encap) { - /* encap filters must specify encap type */ - match_flags |= EFX_FILTER_MATCH_ENCAP_TYPE; - /* and imply ethertype and ip proto */ - mcdi_flags &= - ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_IP_PROTO_LBN); - mcdi_flags &= - ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ETHER_TYPE_LBN); - /* VLAN tags refer to the outer packet */ - MAP_FLAG(INNER_VID, INNER_VLAN); - MAP_FLAG(OUTER_VID, OUTER_VLAN); - /* everything else refers to the inner packet */ - MAP_FLAG(LOC_MAC_IG, IFRM_UNKNOWN_UCAST_DST); - MAP_FLAG(LOC_MAC_IG, IFRM_UNKNOWN_MCAST_DST); - MAP_FLAG(REM_HOST, IFRM_SRC_IP); - MAP_FLAG(LOC_HOST, IFRM_DST_IP); - MAP_FLAG(REM_MAC, IFRM_SRC_MAC); - MAP_FLAG(REM_PORT, IFRM_SRC_PORT); - MAP_FLAG(LOC_MAC, IFRM_DST_MAC); - MAP_FLAG(LOC_PORT, IFRM_DST_PORT); - MAP_FLAG(ETHER_TYPE, IFRM_ETHER_TYPE); - MAP_FLAG(IP_PROTO, IFRM_IP_PROTO); - } else { - MAP_FLAG(LOC_MAC_IG, UNKNOWN_UCAST_DST); - MAP_FLAG(LOC_MAC_IG, UNKNOWN_MCAST_DST); - MAP_FLAG(REM_HOST, SRC_IP); - MAP_FLAG(LOC_HOST, DST_IP); - MAP_FLAG(REM_MAC, SRC_MAC); - MAP_FLAG(REM_PORT, SRC_PORT); - MAP_FLAG(LOC_MAC, DST_MAC); - MAP_FLAG(LOC_PORT, DST_PORT); - MAP_FLAG(ETHER_TYPE, ETHER_TYPE); - MAP_FLAG(INNER_VID, INNER_VLAN); - MAP_FLAG(OUTER_VID, OUTER_VLAN); - MAP_FLAG(IP_PROTO, IP_PROTO); - } -#undef MAP_FLAG - - /* Did we map them all? */ - if (mcdi_flags) - return -EINVAL; - - return match_flags; -} - -static void efx_ef10_filter_cleanup_vlans(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_filter_vlan *vlan, *next_vlan; - - /* See comment in efx_ef10_filter_table_remove() */ - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return; - - if (!table) - return; - - list_for_each_entry_safe(vlan, next_vlan, &table->vlan_list, list) - efx_ef10_filter_del_vlan_internal(efx, vlan); -} - -static bool efx_ef10_filter_match_supported(struct efx_ef10_filter_table *table, - bool encap, - enum efx_filter_match_flags match_flags) -{ - unsigned int match_pri; - int mf; - - for (match_pri = 0; - match_pri < table->rx_match_count; - match_pri++) { - mf = efx_ef10_filter_match_flags_from_mcdi(encap, - table->rx_match_mcdi_flags[match_pri]); - if (mf == match_flags) - return true; - } - - return false; -} - -static int -efx_ef10_filter_table_probe_matches(struct efx_nic *efx, - struct efx_ef10_filter_table *table, - bool encap) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_GET_PARSER_DISP_INFO_IN_LEN); - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PARSER_DISP_INFO_OUT_LENMAX); - unsigned int pd_match_pri, pd_match_count; - size_t outlen; - int rc; - - /* Find out which RX filter types are supported, and their priorities */ - MCDI_SET_DWORD(inbuf, GET_PARSER_DISP_INFO_IN_OP, - encap ? - MC_CMD_GET_PARSER_DISP_INFO_IN_OP_GET_SUPPORTED_ENCAP_RX_MATCHES : - MC_CMD_GET_PARSER_DISP_INFO_IN_OP_GET_SUPPORTED_RX_MATCHES); - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PARSER_DISP_INFO, - inbuf, sizeof(inbuf), outbuf, sizeof(outbuf), - &outlen); - if (rc) - return rc; - - pd_match_count = MCDI_VAR_ARRAY_LEN( - outlen, GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES); - - for (pd_match_pri = 0; pd_match_pri < pd_match_count; pd_match_pri++) { - u32 mcdi_flags = - MCDI_ARRAY_DWORD( - outbuf, - GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES, - pd_match_pri); - rc = efx_ef10_filter_match_flags_from_mcdi(encap, mcdi_flags); - if (rc < 0) { - netif_dbg(efx, probe, efx->net_dev, - "%s: fw flags %#x pri %u not supported in driver\n", - __func__, mcdi_flags, pd_match_pri); - } else { - netif_dbg(efx, probe, efx->net_dev, - "%s: fw flags %#x pri %u supported as driver flags %#x pri %u\n", - __func__, mcdi_flags, pd_match_pri, - rc, table->rx_match_count); - table->rx_match_mcdi_flags[table->rx_match_count] = mcdi_flags; - table->rx_match_count++; - } - } - - return 0; -} - -static int efx_ef10_filter_table_probe(struct efx_nic *efx) -{ - struct efx_ef10_nic_data *nic_data = efx->nic_data; - struct net_device *net_dev = efx->net_dev; - struct efx_ef10_filter_table *table; - struct efx_ef10_vlan *vlan; - int rc; - - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return -EINVAL; - - if (efx->filter_state) /* already probed */ - return 0; - - table = kzalloc(sizeof(*table), GFP_KERNEL); - if (!table) - return -ENOMEM; - - table->rx_match_count = 0; - rc = efx_ef10_filter_table_probe_matches(efx, table, false); - if (rc) - goto fail; - if (nic_data->datapath_caps & - (1 << MC_CMD_GET_CAPABILITIES_OUT_VXLAN_NVGRE_LBN)) - rc = efx_ef10_filter_table_probe_matches(efx, table, true); - if (rc) - goto fail; - if ((efx_supported_features(efx) & NETIF_F_HW_VLAN_CTAG_FILTER) && - !(efx_ef10_filter_match_supported(table, false, - (EFX_FILTER_MATCH_OUTER_VID | EFX_FILTER_MATCH_LOC_MAC)) && - efx_ef10_filter_match_supported(table, false, - (EFX_FILTER_MATCH_OUTER_VID | EFX_FILTER_MATCH_LOC_MAC_IG)))) { - netif_info(efx, probe, net_dev, - "VLAN filters are not supported in this firmware variant\n"); - net_dev->features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; - efx->fixed_features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; - net_dev->hw_features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; - } - - table->entry = vzalloc(array_size(HUNT_FILTER_TBL_ROWS, - sizeof(*table->entry))); - if (!table->entry) { - rc = -ENOMEM; - goto fail; - } - - table->mc_promisc_last = false; - table->vlan_filter = - !!(efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_FILTER); - INIT_LIST_HEAD(&table->vlan_list); - init_rwsem(&table->lock); - - efx->filter_state = table; - - list_for_each_entry(vlan, &nic_data->vlan_list, list) { - rc = efx_ef10_filter_add_vlan(efx, vlan->vid); - if (rc) - goto fail_add_vlan; - } - - return 0; - -fail_add_vlan: - efx_ef10_filter_cleanup_vlans(efx); - efx->filter_state = NULL; -fail: - kfree(table); - return rc; -} - -/* Caller must hold efx->filter_sem for read if race against - * efx_ef10_filter_table_remove() is possible - */ -static void efx_ef10_filter_table_restore(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_nic_data *nic_data = efx->nic_data; - unsigned int invalid_filters = 0, failed = 0; - struct efx_ef10_filter_vlan *vlan; - struct efx_filter_spec *spec; - struct efx_rss_context *ctx; - unsigned int filter_idx; - u32 mcdi_flags; - int match_pri; - int rc, i; - - WARN_ON(!rwsem_is_locked(&efx->filter_sem)); - - if (!nic_data->must_restore_filters) - return; - - if (!table) - return; - - down_write(&table->lock); - mutex_lock(&efx->rss_lock); - - for (filter_idx = 0; filter_idx < HUNT_FILTER_TBL_ROWS; filter_idx++) { - spec = efx_ef10_filter_entry_spec(table, filter_idx); - if (!spec) - continue; - - mcdi_flags = efx_ef10_filter_mcdi_flags_from_spec(spec); - match_pri = 0; - while (match_pri < table->rx_match_count && - table->rx_match_mcdi_flags[match_pri] != mcdi_flags) - ++match_pri; - if (match_pri >= table->rx_match_count) { - invalid_filters++; - goto not_restored; - } - if (spec->rss_context) - ctx = efx_find_rss_context_entry(efx, spec->rss_context); - else - ctx = &efx->rss_context; - if (spec->flags & EFX_FILTER_FLAG_RX_RSS) { - if (!ctx) { - netif_warn(efx, drv, efx->net_dev, - "Warning: unable to restore a filter with nonexistent RSS context %u.\n", - spec->rss_context); - invalid_filters++; - goto not_restored; - } - if (ctx->context_id == EFX_EF10_RSS_CONTEXT_INVALID) { - netif_warn(efx, drv, efx->net_dev, - "Warning: unable to restore a filter with RSS context %u as it was not created.\n", - spec->rss_context); - invalid_filters++; - goto not_restored; - } - } - - rc = efx_ef10_filter_push(efx, spec, - &table->entry[filter_idx].handle, - ctx, false); - if (rc) - failed++; - - if (rc) { -not_restored: - list_for_each_entry(vlan, &table->vlan_list, list) - for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; ++i) - if (vlan->default_filters[i] == filter_idx) - vlan->default_filters[i] = - EFX_EF10_FILTER_ID_INVALID; - - kfree(spec); - efx_ef10_filter_set_entry(table, filter_idx, NULL, 0); - } - } - - mutex_unlock(&efx->rss_lock); - up_write(&table->lock); - - /* This can happen validly if the MC's capabilities have changed, so - * is not an error. - */ - if (invalid_filters) - netif_dbg(efx, drv, efx->net_dev, - "Did not restore %u filters that are now unsupported.\n", - invalid_filters); - - if (failed) - netif_err(efx, hw, efx->net_dev, - "unable to restore %u filters\n", failed); - else - nic_data->must_restore_filters = false; -} - -static void efx_ef10_filter_table_remove(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); - struct efx_filter_spec *spec; - unsigned int filter_idx; - int rc; - - efx_ef10_filter_cleanup_vlans(efx); - efx->filter_state = NULL; - /* If we were called without locking, then it's not safe to free - * the table as others might be using it. So we just WARN, leak - * the memory, and potentially get an inconsistent filter table - * state. - * This should never actually happen. - */ - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return; - - if (!table) - return; - - for (filter_idx = 0; filter_idx < HUNT_FILTER_TBL_ROWS; filter_idx++) { - spec = efx_ef10_filter_entry_spec(table, filter_idx); - if (!spec) - continue; - - MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, - efx_ef10_filter_is_exclusive(spec) ? - MC_CMD_FILTER_OP_IN_OP_REMOVE : - MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); - MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, - table->entry[filter_idx].handle); - rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, inbuf, - sizeof(inbuf), NULL, 0, NULL); - if (rc) - netif_info(efx, drv, efx->net_dev, - "%s: filter %04x remove failed\n", - __func__, filter_idx); - kfree(spec); - } - - vfree(table->entry); - kfree(table); -} - -static void efx_ef10_filter_mark_one_old(struct efx_nic *efx, uint16_t *id) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - unsigned int filter_idx; - - efx_rwsem_assert_write_locked(&table->lock); - - if (*id != EFX_EF10_FILTER_ID_INVALID) { - filter_idx = efx_ef10_filter_get_unsafe_id(*id); - if (!table->entry[filter_idx].spec) - netif_dbg(efx, drv, efx->net_dev, - "marked null spec old %04x:%04x\n", *id, - filter_idx); - table->entry[filter_idx].spec |= EFX_EF10_FILTER_FLAG_AUTO_OLD; - *id = EFX_EF10_FILTER_ID_INVALID; - } -} - -/* Mark old per-VLAN filters that may need to be removed */ -static void _efx_ef10_filter_vlan_mark_old(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - unsigned int i; - - for (i = 0; i < table->dev_uc_count; i++) - efx_ef10_filter_mark_one_old(efx, &vlan->uc[i]); - for (i = 0; i < table->dev_mc_count; i++) - efx_ef10_filter_mark_one_old(efx, &vlan->mc[i]); - for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) - efx_ef10_filter_mark_one_old(efx, &vlan->default_filters[i]); -} - -/* Mark old filters that may need to be removed. - * Caller must hold efx->filter_sem for read if race against - * efx_ef10_filter_table_remove() is possible - */ -static void efx_ef10_filter_mark_old(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_filter_vlan *vlan; - - down_write(&table->lock); - list_for_each_entry(vlan, &table->vlan_list, list) - _efx_ef10_filter_vlan_mark_old(efx, vlan); - up_write(&table->lock); -} - -static void efx_ef10_filter_uc_addr_list(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct net_device *net_dev = efx->net_dev; - struct netdev_hw_addr *uc; - unsigned int i; - - table->uc_promisc = !!(net_dev->flags & IFF_PROMISC); - ether_addr_copy(table->dev_uc_list[0].addr, net_dev->dev_addr); - i = 1; - netdev_for_each_uc_addr(uc, net_dev) { - if (i >= EFX_EF10_FILTER_DEV_UC_MAX) { - table->uc_promisc = true; - break; - } - ether_addr_copy(table->dev_uc_list[i].addr, uc->addr); - i++; - } - - table->dev_uc_count = i; -} - -static void efx_ef10_filter_mc_addr_list(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct net_device *net_dev = efx->net_dev; - struct netdev_hw_addr *mc; - unsigned int i; - - table->mc_overflow = false; - table->mc_promisc = !!(net_dev->flags & (IFF_PROMISC | IFF_ALLMULTI)); - - i = 0; - netdev_for_each_mc_addr(mc, net_dev) { - if (i >= EFX_EF10_FILTER_DEV_MC_MAX) { - table->mc_promisc = true; - table->mc_overflow = true; - break; - } - ether_addr_copy(table->dev_mc_list[i].addr, mc->addr); - i++; - } - - table->dev_mc_count = i; -} - -static int efx_ef10_filter_insert_addr_list(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan, - bool multicast, bool rollback) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_dev_addr *addr_list; - enum efx_filter_flags filter_flags; - struct efx_filter_spec spec; - u8 baddr[ETH_ALEN]; - unsigned int i, j; - int addr_count; - u16 *ids; - int rc; - - if (multicast) { - addr_list = table->dev_mc_list; - addr_count = table->dev_mc_count; - ids = vlan->mc; - } else { - addr_list = table->dev_uc_list; - addr_count = table->dev_uc_count; - ids = vlan->uc; - } - - filter_flags = efx_rss_active(&efx->rss_context) ? EFX_FILTER_FLAG_RX_RSS : 0; - - /* Insert/renew filters */ - for (i = 0; i < addr_count; i++) { - EFX_WARN_ON_PARANOID(ids[i] != EFX_EF10_FILTER_ID_INVALID); - efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); - efx_filter_set_eth_local(&spec, vlan->vid, addr_list[i].addr); - rc = efx_ef10_filter_insert_locked(efx, &spec, true); - if (rc < 0) { - if (rollback) { - netif_info(efx, drv, efx->net_dev, - "efx_ef10_filter_insert failed rc=%d\n", - rc); - /* Fall back to promiscuous */ - for (j = 0; j < i; j++) { - efx_ef10_filter_remove_unsafe( - efx, EFX_FILTER_PRI_AUTO, - ids[j]); - ids[j] = EFX_EF10_FILTER_ID_INVALID; - } - return rc; - } else { - /* keep invalid ID, and carry on */ - } - } else { - ids[i] = efx_ef10_filter_get_unsafe_id(rc); - } - } - - if (multicast && rollback) { - /* Also need an Ethernet broadcast filter */ - EFX_WARN_ON_PARANOID(vlan->default_filters[EFX_EF10_BCAST] != - EFX_EF10_FILTER_ID_INVALID); - efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); - eth_broadcast_addr(baddr); - efx_filter_set_eth_local(&spec, vlan->vid, baddr); - rc = efx_ef10_filter_insert_locked(efx, &spec, true); - if (rc < 0) { - netif_warn(efx, drv, efx->net_dev, - "Broadcast filter insert failed rc=%d\n", rc); - /* Fall back to promiscuous */ - for (j = 0; j < i; j++) { - efx_ef10_filter_remove_unsafe( - efx, EFX_FILTER_PRI_AUTO, - ids[j]); - ids[j] = EFX_EF10_FILTER_ID_INVALID; - } - return rc; - } else { - vlan->default_filters[EFX_EF10_BCAST] = - efx_ef10_filter_get_unsafe_id(rc); - } - } - - return 0; -} - -static int efx_ef10_filter_insert_def(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan, - enum efx_encap_type encap_type, - bool multicast, bool rollback) -{ - struct efx_ef10_nic_data *nic_data = efx->nic_data; - enum efx_filter_flags filter_flags; - struct efx_filter_spec spec; - u8 baddr[ETH_ALEN]; - int rc; - u16 *id; - - filter_flags = efx_rss_active(&efx->rss_context) ? EFX_FILTER_FLAG_RX_RSS : 0; - - efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); - - if (multicast) - efx_filter_set_mc_def(&spec); - else - efx_filter_set_uc_def(&spec); - - if (encap_type) { - if (nic_data->datapath_caps & - (1 << MC_CMD_GET_CAPABILITIES_OUT_VXLAN_NVGRE_LBN)) - efx_filter_set_encap_type(&spec, encap_type); - else - /* don't insert encap filters on non-supporting - * platforms. ID will be left as INVALID. - */ - return 0; - } - - if (vlan->vid != EFX_FILTER_VID_UNSPEC) - efx_filter_set_eth_local(&spec, vlan->vid, NULL); - - rc = efx_ef10_filter_insert_locked(efx, &spec, true); - if (rc < 0) { - const char *um = multicast ? "Multicast" : "Unicast"; - const char *encap_name = ""; - const char *encap_ipv = ""; - - if ((encap_type & EFX_ENCAP_TYPES_MASK) == - EFX_ENCAP_TYPE_VXLAN) - encap_name = "VXLAN "; - else if ((encap_type & EFX_ENCAP_TYPES_MASK) == - EFX_ENCAP_TYPE_NVGRE) - encap_name = "NVGRE "; - else if ((encap_type & EFX_ENCAP_TYPES_MASK) == - EFX_ENCAP_TYPE_GENEVE) - encap_name = "GENEVE "; - if (encap_type & EFX_ENCAP_FLAG_IPV6) - encap_ipv = "IPv6 "; - else if (encap_type) - encap_ipv = "IPv4 "; - - /* unprivileged functions can't insert mismatch filters - * for encapsulated or unicast traffic, so downgrade - * those warnings to debug. - */ - netif_cond_dbg(efx, drv, efx->net_dev, - rc == -EPERM && (encap_type || !multicast), warn, - "%s%s%s mismatch filter insert failed rc=%d\n", - encap_name, encap_ipv, um, rc); - } else if (multicast) { - /* mapping from encap types to default filter IDs (multicast) */ - static enum efx_ef10_default_filters map[] = { - [EFX_ENCAP_TYPE_NONE] = EFX_EF10_MCDEF, - [EFX_ENCAP_TYPE_VXLAN] = EFX_EF10_VXLAN4_MCDEF, - [EFX_ENCAP_TYPE_NVGRE] = EFX_EF10_NVGRE4_MCDEF, - [EFX_ENCAP_TYPE_GENEVE] = EFX_EF10_GENEVE4_MCDEF, - [EFX_ENCAP_TYPE_VXLAN | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_VXLAN6_MCDEF, - [EFX_ENCAP_TYPE_NVGRE | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_NVGRE6_MCDEF, - [EFX_ENCAP_TYPE_GENEVE | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_GENEVE6_MCDEF, - }; - - /* quick bounds check (BCAST result impossible) */ - BUILD_BUG_ON(EFX_EF10_BCAST != 0); - if (encap_type >= ARRAY_SIZE(map) || map[encap_type] == 0) { - WARN_ON(1); - return -EINVAL; - } - /* then follow map */ - id = &vlan->default_filters[map[encap_type]]; - - EFX_WARN_ON_PARANOID(*id != EFX_EF10_FILTER_ID_INVALID); - *id = efx_ef10_filter_get_unsafe_id(rc); - if (!nic_data->workaround_26807 && !encap_type) { - /* Also need an Ethernet broadcast filter */ - efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, - filter_flags, 0); - eth_broadcast_addr(baddr); - efx_filter_set_eth_local(&spec, vlan->vid, baddr); - rc = efx_ef10_filter_insert_locked(efx, &spec, true); - if (rc < 0) { - netif_warn(efx, drv, efx->net_dev, - "Broadcast filter insert failed rc=%d\n", - rc); - if (rollback) { - /* Roll back the mc_def filter */ - efx_ef10_filter_remove_unsafe( - efx, EFX_FILTER_PRI_AUTO, - *id); - *id = EFX_EF10_FILTER_ID_INVALID; - return rc; - } - } else { - EFX_WARN_ON_PARANOID( - vlan->default_filters[EFX_EF10_BCAST] != - EFX_EF10_FILTER_ID_INVALID); - vlan->default_filters[EFX_EF10_BCAST] = - efx_ef10_filter_get_unsafe_id(rc); - } - } - rc = 0; - } else { - /* mapping from encap types to default filter IDs (unicast) */ - static enum efx_ef10_default_filters map[] = { - [EFX_ENCAP_TYPE_NONE] = EFX_EF10_UCDEF, - [EFX_ENCAP_TYPE_VXLAN] = EFX_EF10_VXLAN4_UCDEF, - [EFX_ENCAP_TYPE_NVGRE] = EFX_EF10_NVGRE4_UCDEF, - [EFX_ENCAP_TYPE_GENEVE] = EFX_EF10_GENEVE4_UCDEF, - [EFX_ENCAP_TYPE_VXLAN | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_VXLAN6_UCDEF, - [EFX_ENCAP_TYPE_NVGRE | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_NVGRE6_UCDEF, - [EFX_ENCAP_TYPE_GENEVE | EFX_ENCAP_FLAG_IPV6] = - EFX_EF10_GENEVE6_UCDEF, - }; - - /* quick bounds check (BCAST result impossible) */ - BUILD_BUG_ON(EFX_EF10_BCAST != 0); - if (encap_type >= ARRAY_SIZE(map) || map[encap_type] == 0) { - WARN_ON(1); - return -EINVAL; - } - /* then follow map */ - id = &vlan->default_filters[map[encap_type]]; - EFX_WARN_ON_PARANOID(*id != EFX_EF10_FILTER_ID_INVALID); - *id = rc; - rc = 0; - } - return rc; -} - -/* Remove filters that weren't renewed. */ -static void efx_ef10_filter_remove_old(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - int remove_failed = 0; - int remove_noent = 0; - int rc; - int i; - - down_write(&table->lock); - for (i = 0; i < HUNT_FILTER_TBL_ROWS; i++) { - if (READ_ONCE(table->entry[i].spec) & - EFX_EF10_FILTER_FLAG_AUTO_OLD) { - rc = efx_ef10_filter_remove_internal(efx, - 1U << EFX_FILTER_PRI_AUTO, i, true); - if (rc == -ENOENT) - remove_noent++; - else if (rc) - remove_failed++; - } - } - up_write(&table->lock); - - if (remove_failed) - netif_info(efx, drv, efx->net_dev, - "%s: failed to remove %d filters\n", - __func__, remove_failed); - if (remove_noent) - netif_info(efx, drv, efx->net_dev, - "%s: failed to remove %d non-existent filters\n", - __func__, remove_noent); -} - static int efx_ef10_vport_set_mac_address(struct efx_nic *efx) { struct efx_ef10_nic_data *nic_data = efx->nic_data; @@ -5545,7 +3153,7 @@ static int efx_ef10_vport_set_mac_address(struct efx_nic *efx) efx_device_detach_sync(efx); efx_net_stop(efx->net_dev); down_write(&efx->filter_sem); - efx_ef10_filter_table_remove(efx); + efx_mcdi_filter_table_remove(efx); up_write(&efx->filter_sem); rc = efx_ef10_vadaptor_free(efx, nic_data->vport_id); @@ -5577,7 +3185,7 @@ restore_vadaptor: goto reset_nic; restore_filters: down_write(&efx->filter_sem); - rc2 = efx_ef10_filter_table_probe(efx); + rc2 = efx_mcdi_filter_table_probe(efx); up_write(&efx->filter_sem); if (rc2) goto reset_nic; @@ -5598,256 +3206,6 @@ reset_nic: return rc ? rc : rc2; } -/* Caller must hold efx->filter_sem for read if race against - * efx_ef10_filter_table_remove() is possible - */ -static void efx_ef10_filter_vlan_sync_rx_mode(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_nic_data *nic_data = efx->nic_data; - - /* Do not install unspecified VID if VLAN filtering is enabled. - * Do not install all specified VIDs if VLAN filtering is disabled. - */ - if ((vlan->vid == EFX_FILTER_VID_UNSPEC) == table->vlan_filter) - return; - - /* Insert/renew unicast filters */ - if (table->uc_promisc) { - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NONE, - false, false); - efx_ef10_filter_insert_addr_list(efx, vlan, false, false); - } else { - /* If any of the filters failed to insert, fall back to - * promiscuous mode - add in the uc_def filter. But keep - * our individual unicast filters. - */ - if (efx_ef10_filter_insert_addr_list(efx, vlan, false, false)) - efx_ef10_filter_insert_def(efx, vlan, - EFX_ENCAP_TYPE_NONE, - false, false); - } - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN, - false, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN | - EFX_ENCAP_FLAG_IPV6, - false, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE, - false, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE | - EFX_ENCAP_FLAG_IPV6, - false, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE, - false, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE | - EFX_ENCAP_FLAG_IPV6, - false, false); - - /* Insert/renew multicast filters */ - /* If changing promiscuous state with cascaded multicast filters, remove - * old filters first, so that packets are dropped rather than duplicated - */ - if (nic_data->workaround_26807 && - table->mc_promisc_last != table->mc_promisc) - efx_ef10_filter_remove_old(efx); - if (table->mc_promisc) { - if (nic_data->workaround_26807) { - /* If we failed to insert promiscuous filters, rollback - * and fall back to individual multicast filters - */ - if (efx_ef10_filter_insert_def(efx, vlan, - EFX_ENCAP_TYPE_NONE, - true, true)) { - /* Changing promisc state, so remove old filters */ - efx_ef10_filter_remove_old(efx); - efx_ef10_filter_insert_addr_list(efx, vlan, - true, false); - } - } else { - /* If we failed to insert promiscuous filters, don't - * rollback. Regardless, also insert the mc_list, - * unless it's incomplete due to overflow - */ - efx_ef10_filter_insert_def(efx, vlan, - EFX_ENCAP_TYPE_NONE, - true, false); - if (!table->mc_overflow) - efx_ef10_filter_insert_addr_list(efx, vlan, - true, false); - } - } else { - /* If any filters failed to insert, rollback and fall back to - * promiscuous mode - mc_def filter and maybe broadcast. If - * that fails, roll back again and insert as many of our - * individual multicast filters as we can. - */ - if (efx_ef10_filter_insert_addr_list(efx, vlan, true, true)) { - /* Changing promisc state, so remove old filters */ - if (nic_data->workaround_26807) - efx_ef10_filter_remove_old(efx); - if (efx_ef10_filter_insert_def(efx, vlan, - EFX_ENCAP_TYPE_NONE, - true, true)) - efx_ef10_filter_insert_addr_list(efx, vlan, - true, false); - } - } - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN, - true, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN | - EFX_ENCAP_FLAG_IPV6, - true, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE, - true, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE | - EFX_ENCAP_FLAG_IPV6, - true, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE, - true, false); - efx_ef10_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE | - EFX_ENCAP_FLAG_IPV6, - true, false); -} - -/* Caller must hold efx->filter_sem for read if race against - * efx_ef10_filter_table_remove() is possible - */ -static void efx_ef10_filter_sync_rx_mode(struct efx_nic *efx) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct net_device *net_dev = efx->net_dev; - struct efx_ef10_filter_vlan *vlan; - bool vlan_filter; - - if (!efx_dev_registered(efx)) - return; - - if (!table) - return; - - efx_ef10_filter_mark_old(efx); - - /* Copy/convert the address lists; add the primary station - * address and broadcast address - */ - netif_addr_lock_bh(net_dev); - efx_ef10_filter_uc_addr_list(efx); - efx_ef10_filter_mc_addr_list(efx); - netif_addr_unlock_bh(net_dev); - - /* If VLAN filtering changes, all old filters are finally removed. - * Do it in advance to avoid conflicts for unicast untagged and - * VLAN 0 tagged filters. - */ - vlan_filter = !!(net_dev->features & NETIF_F_HW_VLAN_CTAG_FILTER); - if (table->vlan_filter != vlan_filter) { - table->vlan_filter = vlan_filter; - efx_ef10_filter_remove_old(efx); - } - - list_for_each_entry(vlan, &table->vlan_list, list) - efx_ef10_filter_vlan_sync_rx_mode(efx, vlan); - - efx_ef10_filter_remove_old(efx); - table->mc_promisc_last = table->mc_promisc; -} - -static struct efx_ef10_filter_vlan *efx_ef10_filter_find_vlan(struct efx_nic *efx, u16 vid) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_filter_vlan *vlan; - - WARN_ON(!rwsem_is_locked(&efx->filter_sem)); - - list_for_each_entry(vlan, &table->vlan_list, list) { - if (vlan->vid == vid) - return vlan; - } - - return NULL; -} - -static int efx_ef10_filter_add_vlan(struct efx_nic *efx, u16 vid) -{ - struct efx_ef10_filter_table *table = efx->filter_state; - struct efx_ef10_filter_vlan *vlan; - unsigned int i; - - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return -EINVAL; - - vlan = efx_ef10_filter_find_vlan(efx, vid); - if (WARN_ON(vlan)) { - netif_err(efx, drv, efx->net_dev, - "VLAN %u already added\n", vid); - return -EALREADY; - } - - vlan = kzalloc(sizeof(*vlan), GFP_KERNEL); - if (!vlan) - return -ENOMEM; - - vlan->vid = vid; - - for (i = 0; i < ARRAY_SIZE(vlan->uc); i++) - vlan->uc[i] = EFX_EF10_FILTER_ID_INVALID; - for (i = 0; i < ARRAY_SIZE(vlan->mc); i++) - vlan->mc[i] = EFX_EF10_FILTER_ID_INVALID; - for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) - vlan->default_filters[i] = EFX_EF10_FILTER_ID_INVALID; - - list_add_tail(&vlan->list, &table->vlan_list); - - if (efx_dev_registered(efx)) - efx_ef10_filter_vlan_sync_rx_mode(efx, vlan); - - return 0; -} - -static void efx_ef10_filter_del_vlan_internal(struct efx_nic *efx, - struct efx_ef10_filter_vlan *vlan) -{ - unsigned int i; - - /* See comment in efx_ef10_filter_table_remove() */ - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return; - - list_del(&vlan->list); - - for (i = 0; i < ARRAY_SIZE(vlan->uc); i++) - efx_ef10_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, - vlan->uc[i]); - for (i = 0; i < ARRAY_SIZE(vlan->mc); i++) - efx_ef10_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, - vlan->mc[i]); - for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) - if (vlan->default_filters[i] != EFX_EF10_FILTER_ID_INVALID) - efx_ef10_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, - vlan->default_filters[i]); - - kfree(vlan); -} - -static void efx_ef10_filter_del_vlan(struct efx_nic *efx, u16 vid) -{ - struct efx_ef10_filter_vlan *vlan; - - /* See comment in efx_ef10_filter_table_remove() */ - if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) - return; - - vlan = efx_ef10_filter_find_vlan(efx, vid); - if (!vlan) { - netif_err(efx, drv, efx->net_dev, - "VLAN %u not found in filter state\n", vid); - return; - } - - efx_ef10_filter_del_vlan_internal(efx, vlan); -} - static int efx_ef10_set_mac_address(struct efx_nic *efx) { MCDI_DECLARE_BUF(inbuf, MC_CMD_VADAPTOR_SET_MAC_IN_LEN); @@ -5860,7 +3218,7 @@ static int efx_ef10_set_mac_address(struct efx_nic *efx) mutex_lock(&efx->mac_lock); down_write(&efx->filter_sem); - efx_ef10_filter_table_remove(efx); + efx_mcdi_filter_table_remove(efx); ether_addr_copy(MCDI_PTR(inbuf, VADAPTOR_SET_MAC_IN_MACADDR), efx->net_dev->dev_addr); @@ -5869,7 +3227,7 @@ static int efx_ef10_set_mac_address(struct efx_nic *efx) rc = efx_mcdi_rpc_quiet(efx, MC_CMD_VADAPTOR_SET_MAC, inbuf, sizeof(inbuf), NULL, 0, NULL); - efx_ef10_filter_table_probe(efx); + efx_mcdi_filter_table_probe(efx); up_write(&efx->filter_sem); mutex_unlock(&efx->mac_lock); @@ -5931,14 +3289,14 @@ static int efx_ef10_set_mac_address(struct efx_nic *efx) static int efx_ef10_mac_reconfigure(struct efx_nic *efx) { - efx_ef10_filter_sync_rx_mode(efx); + efx_mcdi_filter_sync_rx_mode(efx); return efx_mcdi_set_mac(efx); } static int efx_ef10_mac_reconfigure_vf(struct efx_nic *efx) { - efx_ef10_filter_sync_rx_mode(efx); + efx_mcdi_filter_sync_rx_mode(efx); return 0; } @@ -6650,36 +4008,36 @@ const struct efx_nic_type efx_hunt_a0_vf_nic_type = { .irq_handle_legacy = efx_ef10_legacy_interrupt, .tx_probe = efx_ef10_tx_probe, .tx_init = efx_ef10_tx_init, - .tx_remove = efx_ef10_tx_remove, + .tx_remove = efx_mcdi_tx_remove, .tx_write = efx_ef10_tx_write, .tx_limit_len = efx_ef10_tx_limit_len, - .rx_push_rss_config = efx_ef10_vf_rx_push_rss_config, - .rx_pull_rss_config = efx_ef10_rx_pull_rss_config, - .rx_probe = efx_ef10_rx_probe, - .rx_init = efx_ef10_rx_init, - .rx_remove = efx_ef10_rx_remove, + .rx_push_rss_config = efx_mcdi_vf_rx_push_rss_config, + .rx_pull_rss_config = efx_mcdi_rx_pull_rss_config, + .rx_probe = efx_mcdi_rx_probe, + .rx_init = efx_mcdi_rx_init, + .rx_remove = efx_mcdi_rx_remove, .rx_write = efx_ef10_rx_write, .rx_defer_refill = efx_ef10_rx_defer_refill, - .ev_probe = efx_ef10_ev_probe, + .ev_probe = efx_mcdi_ev_probe, .ev_init = efx_ef10_ev_init, - .ev_fini = efx_ef10_ev_fini, - .ev_remove = efx_ef10_ev_remove, + .ev_fini = efx_mcdi_ev_fini, + .ev_remove = efx_mcdi_ev_remove, .ev_process = efx_ef10_ev_process, .ev_read_ack = efx_ef10_ev_read_ack, .ev_test_generate = efx_ef10_ev_test_generate, - .filter_table_probe = efx_ef10_filter_table_probe, - .filter_table_restore = efx_ef10_filter_table_restore, - .filter_table_remove = efx_ef10_filter_table_remove, - .filter_update_rx_scatter = efx_ef10_filter_update_rx_scatter, - .filter_insert = efx_ef10_filter_insert, - .filter_remove_safe = efx_ef10_filter_remove_safe, - .filter_get_safe = efx_ef10_filter_get_safe, - .filter_clear_rx = efx_ef10_filter_clear_rx, - .filter_count_rx_used = efx_ef10_filter_count_rx_used, - .filter_get_rx_id_limit = efx_ef10_filter_get_rx_id_limit, - .filter_get_rx_ids = efx_ef10_filter_get_rx_ids, + .filter_table_probe = efx_mcdi_filter_table_probe, + .filter_table_restore = efx_mcdi_filter_table_restore, + .filter_table_remove = efx_mcdi_filter_table_remove, + .filter_update_rx_scatter = efx_mcdi_update_rx_scatter, + .filter_insert = efx_mcdi_filter_insert, + .filter_remove_safe = efx_mcdi_filter_remove_safe, + .filter_get_safe = efx_mcdi_filter_get_safe, + .filter_clear_rx = efx_mcdi_filter_clear_rx, + .filter_count_rx_used = efx_mcdi_filter_count_rx_used, + .filter_get_rx_id_limit = efx_mcdi_filter_get_rx_id_limit, + .filter_get_rx_ids = efx_mcdi_filter_get_rx_ids, #ifdef CONFIG_RFS_ACCEL - .filter_rfs_expire_one = efx_ef10_filter_rfs_expire_one, + .filter_rfs_expire_one = efx_mcdi_filter_rfs_expire_one, #endif #ifdef CONFIG_SFC_MTD .mtd_probe = efx_port_dummy_op_int, @@ -6709,7 +4067,7 @@ const struct efx_nic_type efx_hunt_a0_vf_nic_type = { .timer_period_max = 1 << ERF_DD_EVQ_IND_TIMER_VAL_WIDTH, .offload_features = EF10_OFFLOAD_FEATURES, .mcdi_max_ver = 2, - .max_rx_ip_filters = HUNT_FILTER_TBL_ROWS, + .max_rx_ip_filters = EFX_MCDI_FILTER_TBL_ROWS, .hwtstamp_filters = 1 << HWTSTAMP_FILTER_NONE | 1 << HWTSTAMP_FILTER_ALL, .rx_hash_key_size = 40, @@ -6759,39 +4117,39 @@ const struct efx_nic_type efx_hunt_a0_nic_type = { .irq_handle_legacy = efx_ef10_legacy_interrupt, .tx_probe = efx_ef10_tx_probe, .tx_init = efx_ef10_tx_init, - .tx_remove = efx_ef10_tx_remove, + .tx_remove = efx_mcdi_tx_remove, .tx_write = efx_ef10_tx_write, .tx_limit_len = efx_ef10_tx_limit_len, - .rx_push_rss_config = efx_ef10_pf_rx_push_rss_config, - .rx_pull_rss_config = efx_ef10_rx_pull_rss_config, - .rx_push_rss_context_config = efx_ef10_rx_push_rss_context_config, - .rx_pull_rss_context_config = efx_ef10_rx_pull_rss_context_config, - .rx_restore_rss_contexts = efx_ef10_rx_restore_rss_contexts, - .rx_probe = efx_ef10_rx_probe, - .rx_init = efx_ef10_rx_init, - .rx_remove = efx_ef10_rx_remove, + .rx_push_rss_config = efx_mcdi_pf_rx_push_rss_config, + .rx_pull_rss_config = efx_mcdi_rx_pull_rss_config, + .rx_push_rss_context_config = efx_mcdi_rx_push_rss_context_config, + .rx_pull_rss_context_config = efx_mcdi_rx_pull_rss_context_config, + .rx_restore_rss_contexts = efx_mcdi_rx_restore_rss_contexts, + .rx_probe = efx_mcdi_rx_probe, + .rx_init = efx_mcdi_rx_init, + .rx_remove = efx_mcdi_rx_remove, .rx_write = efx_ef10_rx_write, .rx_defer_refill = efx_ef10_rx_defer_refill, - .ev_probe = efx_ef10_ev_probe, + .ev_probe = efx_mcdi_ev_probe, .ev_init = efx_ef10_ev_init, - .ev_fini = efx_ef10_ev_fini, - .ev_remove = efx_ef10_ev_remove, + .ev_fini = efx_mcdi_ev_fini, + .ev_remove = efx_mcdi_ev_remove, .ev_process = efx_ef10_ev_process, .ev_read_ack = efx_ef10_ev_read_ack, .ev_test_generate = efx_ef10_ev_test_generate, - .filter_table_probe = efx_ef10_filter_table_probe, - .filter_table_restore = efx_ef10_filter_table_restore, - .filter_table_remove = efx_ef10_filter_table_remove, - .filter_update_rx_scatter = efx_ef10_filter_update_rx_scatter, - .filter_insert = efx_ef10_filter_insert, - .filter_remove_safe = efx_ef10_filter_remove_safe, - .filter_get_safe = efx_ef10_filter_get_safe, - .filter_clear_rx = efx_ef10_filter_clear_rx, - .filter_count_rx_used = efx_ef10_filter_count_rx_used, - .filter_get_rx_id_limit = efx_ef10_filter_get_rx_id_limit, - .filter_get_rx_ids = efx_ef10_filter_get_rx_ids, + .filter_table_probe = efx_mcdi_filter_table_probe, + .filter_table_restore = efx_mcdi_filter_table_restore, + .filter_table_remove = efx_mcdi_filter_table_remove, + .filter_update_rx_scatter = efx_mcdi_update_rx_scatter, + .filter_insert = efx_mcdi_filter_insert, + .filter_remove_safe = efx_mcdi_filter_remove_safe, + .filter_get_safe = efx_mcdi_filter_get_safe, + .filter_clear_rx = efx_mcdi_filter_clear_rx, + .filter_count_rx_used = efx_mcdi_filter_count_rx_used, + .filter_get_rx_id_limit = efx_mcdi_filter_get_rx_id_limit, + .filter_get_rx_ids = efx_mcdi_filter_get_rx_ids, #ifdef CONFIG_RFS_ACCEL - .filter_rfs_expire_one = efx_ef10_filter_rfs_expire_one, + .filter_rfs_expire_one = efx_mcdi_filter_rfs_expire_one, #endif #ifdef CONFIG_SFC_MTD .mtd_probe = efx_ef10_mtd_probe, @@ -6844,7 +4202,7 @@ const struct efx_nic_type efx_hunt_a0_nic_type = { .timer_period_max = 1 << ERF_DD_EVQ_IND_TIMER_VAL_WIDTH, .offload_features = EF10_OFFLOAD_FEATURES, .mcdi_max_ver = 2, - .max_rx_ip_filters = HUNT_FILTER_TBL_ROWS, + .max_rx_ip_filters = EFX_MCDI_FILTER_TBL_ROWS, .hwtstamp_filters = 1 << HWTSTAMP_FILTER_NONE | 1 << HWTSTAMP_FILTER_ALL, .rx_hash_key_size = 40, diff --git a/drivers/net/ethernet/sfc/ef10_sriov.c b/drivers/net/ethernet/sfc/ef10_sriov.c index 52bd43f45761..14393767ef9f 100644 --- a/drivers/net/ethernet/sfc/ef10_sriov.c +++ b/drivers/net/ethernet/sfc/ef10_sriov.c @@ -522,10 +522,9 @@ int efx_ef10_sriov_set_vf_mac(struct efx_nic *efx, int vf_i, u8 *mac) if (!is_zero_ether_addr(mac)) { rc = efx_ef10_vport_add_mac(efx, vf->vport_id, mac); - if (rc) { - eth_zero_addr(vf->mac); + if (rc) goto fail; - } + if (vf->efx) ether_addr_copy(vf->efx->net_dev->dev_addr, mac); } diff --git a/drivers/net/ethernet/sfc/efx.c b/drivers/net/ethernet/sfc/efx.c index 6891df471538..4481f21a1f43 100644 --- a/drivers/net/ethernet/sfc/efx.c +++ b/drivers/net/ethernet/sfc/efx.c @@ -23,6 +23,10 @@ #include <net/gre.h> #include <net/udp_tunnel.h> #include "efx.h" +#include "efx_common.h" +#include "efx_channels.h" +#include "rx_common.h" +#include "tx_common.h" #include "nic.h" #include "io.h" #include "selftest.h" @@ -39,56 +43,6 @@ ************************************************************************** */ -/* Loopback mode names (see LOOPBACK_MODE()) */ -const unsigned int efx_loopback_mode_max = LOOPBACK_MAX; -const char *const efx_loopback_mode_names[] = { - [LOOPBACK_NONE] = "NONE", - [LOOPBACK_DATA] = "DATAPATH", - [LOOPBACK_GMAC] = "GMAC", - [LOOPBACK_XGMII] = "XGMII", - [LOOPBACK_XGXS] = "XGXS", - [LOOPBACK_XAUI] = "XAUI", - [LOOPBACK_GMII] = "GMII", - [LOOPBACK_SGMII] = "SGMII", - [LOOPBACK_XGBR] = "XGBR", - [LOOPBACK_XFI] = "XFI", - [LOOPBACK_XAUI_FAR] = "XAUI_FAR", - [LOOPBACK_GMII_FAR] = "GMII_FAR", - [LOOPBACK_SGMII_FAR] = "SGMII_FAR", - [LOOPBACK_XFI_FAR] = "XFI_FAR", - [LOOPBACK_GPHY] = "GPHY", - [LOOPBACK_PHYXS] = "PHYXS", - [LOOPBACK_PCS] = "PCS", - [LOOPBACK_PMAPMD] = "PMA/PMD", - [LOOPBACK_XPORT] = "XPORT", - [LOOPBACK_XGMII_WS] = "XGMII_WS", - [LOOPBACK_XAUI_WS] = "XAUI_WS", - [LOOPBACK_XAUI_WS_FAR] = "XAUI_WS_FAR", - [LOOPBACK_XAUI_WS_NEAR] = "XAUI_WS_NEAR", - [LOOPBACK_GMII_WS] = "GMII_WS", - [LOOPBACK_XFI_WS] = "XFI_WS", - [LOOPBACK_XFI_WS_FAR] = "XFI_WS_FAR", - [LOOPBACK_PHYXS_WS] = "PHYXS_WS", -}; - -const unsigned int efx_reset_type_max = RESET_TYPE_MAX; -const char *const efx_reset_type_names[] = { - [RESET_TYPE_INVISIBLE] = "INVISIBLE", - [RESET_TYPE_ALL] = "ALL", - [RESET_TYPE_RECOVER_OR_ALL] = "RECOVER_OR_ALL", - [RESET_TYPE_WORLD] = "WORLD", - [RESET_TYPE_RECOVER_OR_DISABLE] = "RECOVER_OR_DISABLE", - [RESET_TYPE_DATAPATH] = "DATAPATH", - [RESET_TYPE_MC_BIST] = "MC_BIST", - [RESET_TYPE_DISABLE] = "DISABLE", - [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", - [RESET_TYPE_INT_ERROR] = "INT_ERROR", - [RESET_TYPE_DMA_ERROR] = "DMA_ERROR", - [RESET_TYPE_TX_SKIP] = "TX_SKIP", - [RESET_TYPE_MC_FAILURE] = "MC_FAILURE", - [RESET_TYPE_MCDI_TIMEOUT] = "MCDI_TIMEOUT (FLR)", -}; - /* UDP tunnel type names */ static const char *const efx_udp_tunnel_type_names[] = { [TUNNEL_ENCAP_UDP_PORT_ENTRY_VXLAN] = "vxlan", @@ -104,18 +58,6 @@ void efx_get_udp_tunnel_type_name(u16 type, char *buf, size_t buflen) snprintf(buf, buflen, "type %d", type); } -/* Reset workqueue. If any NIC has a hardware failure then a reset will be - * queued onto this work queue. This is not a per-nic work queue, because - * efx_reset_work() acquires the rtnl lock, so resets are naturally serialised. - */ -static struct workqueue_struct *reset_workqueue; - -/* How often and how many times to poll for a reset while waiting for a - * BIST that another function started to complete. - */ -#define BIST_WAIT_DELAY_MS 100 -#define BIST_WAIT_DELAY_COUNT 100 - /************************************************************************** * * Configurable values @@ -135,21 +77,6 @@ module_param(efx_separate_tx_channels, bool, 0444); MODULE_PARM_DESC(efx_separate_tx_channels, "Use separate channels for TX and RX"); -/* This is the weight assigned to each of the (per-channel) virtual - * NAPI devices. - */ -static int napi_weight = 64; - -/* This is the time (in jiffies) between invocations of the hardware - * monitor. - * On Falcon-based NICs, this will: - * - Check the on-board hardware monitor; - * - Poll the link state and reconfigure the hardware as necessary. - * On Siena-based NICs for power systems with EEH support, this will give EEH a - * chance to start. - */ -static unsigned int efx_monitor_interval = 1 * HZ; - /* Initial interrupt moderation settings. They can be modified after * module load with ethtool. * @@ -169,38 +96,10 @@ static unsigned int rx_irq_mod_usec = 60; */ static unsigned int tx_irq_mod_usec = 150; -/* This is the first interrupt mode to try out of: - * 0 => MSI-X - * 1 => MSI - * 2 => legacy - */ -static unsigned int interrupt_mode; - -/* This is the requested number of CPUs to use for Receive-Side Scaling (RSS), - * i.e. the number of CPUs among which we may distribute simultaneous - * interrupt handling. - * - * Cards without MSI-X will only target one CPU via legacy or MSI interrupt. - * The default (0) means to assign an interrupt to each core. - */ -static unsigned int rss_cpus; -module_param(rss_cpus, uint, 0444); -MODULE_PARM_DESC(rss_cpus, "Number of CPUs to use for Receive-Side Scaling"); - static bool phy_flash_cfg; module_param(phy_flash_cfg, bool, 0644); MODULE_PARM_DESC(phy_flash_cfg, "Set PHYs into reflash mode initially"); -static unsigned irq_adapt_low_thresh = 8000; -module_param(irq_adapt_low_thresh, uint, 0644); -MODULE_PARM_DESC(irq_adapt_low_thresh, - "Threshold score for reducing IRQ moderation"); - -static unsigned irq_adapt_high_thresh = 16000; -module_param(irq_adapt_high_thresh, uint, 0644); -MODULE_PARM_DESC(irq_adapt_high_thresh, - "Threshold score for increasing IRQ moderation"); - static unsigned debug = (NETIF_MSG_DRV | NETIF_MSG_PROBE | NETIF_MSG_LINK | NETIF_MSG_IFDOWN | NETIF_MSG_IFUP | NETIF_MSG_RX_ERR | @@ -214,18 +113,8 @@ MODULE_PARM_DESC(debug, "Bitmapped debugging message enable value"); * *************************************************************************/ -static int efx_soft_enable_interrupts(struct efx_nic *efx); -static void efx_soft_disable_interrupts(struct efx_nic *efx); -static void efx_remove_channel(struct efx_channel *channel); -static void efx_remove_channels(struct efx_nic *efx); static const struct efx_channel_type efx_default_channel_type; static void efx_remove_port(struct efx_nic *efx); -static void efx_init_napi_channel(struct efx_channel *channel); -static void efx_fini_napi(struct efx_nic *efx); -static void efx_fini_napi_channel(struct efx_channel *channel); -static void efx_fini_struct(struct efx_nic *efx); -static void efx_start_all(struct efx_nic *efx); -static void efx_stop_all(struct efx_nic *efx); static int efx_xdp_setup_prog(struct efx_nic *efx, struct bpf_prog *prog); static int efx_xdp(struct net_device *dev, struct netdev_bpf *xdp); static int efx_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **xdpfs, @@ -239,776 +128,12 @@ static int efx_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **xdpfs, ASSERT_RTNL(); \ } while (0) -static int efx_check_disabled(struct efx_nic *efx) -{ - if (efx->state == STATE_DISABLED || efx->state == STATE_RECOVERY) { - netif_err(efx, drv, efx->net_dev, - "device is disabled due to earlier errors\n"); - return -EIO; - } - return 0; -} - -/************************************************************************** - * - * Event queue processing - * - *************************************************************************/ - -/* Process channel's event queue - * - * This function is responsible for processing the event queue of a - * single channel. The caller must guarantee that this function will - * never be concurrently called more than once on the same channel, - * though different channels may be being processed concurrently. - */ -static int efx_process_channel(struct efx_channel *channel, int budget) -{ - struct efx_tx_queue *tx_queue; - struct list_head rx_list; - int spent; - - if (unlikely(!channel->enabled)) - return 0; - - /* Prepare the batch receive list */ - EFX_WARN_ON_PARANOID(channel->rx_list != NULL); - INIT_LIST_HEAD(&rx_list); - channel->rx_list = &rx_list; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - tx_queue->pkts_compl = 0; - tx_queue->bytes_compl = 0; - } - - spent = efx_nic_process_eventq(channel, budget); - if (spent && efx_channel_has_rx_queue(channel)) { - struct efx_rx_queue *rx_queue = - efx_channel_get_rx_queue(channel); - - efx_rx_flush_packet(channel); - efx_fast_push_rx_descriptors(rx_queue, true); - } - - /* Update BQL */ - efx_for_each_channel_tx_queue(tx_queue, channel) { - if (tx_queue->bytes_compl) { - netdev_tx_completed_queue(tx_queue->core_txq, - tx_queue->pkts_compl, tx_queue->bytes_compl); - } - } - - /* Receive any packets we queued up */ - netif_receive_skb_list(channel->rx_list); - channel->rx_list = NULL; - - return spent; -} - -/* NAPI poll handler - * - * NAPI guarantees serialisation of polls of the same device, which - * provides the guarantee required by efx_process_channel(). - */ -static void efx_update_irq_mod(struct efx_nic *efx, struct efx_channel *channel) -{ - int step = efx->irq_mod_step_us; - - if (channel->irq_mod_score < irq_adapt_low_thresh) { - if (channel->irq_moderation_us > step) { - channel->irq_moderation_us -= step; - efx->type->push_irq_moderation(channel); - } - } else if (channel->irq_mod_score > irq_adapt_high_thresh) { - if (channel->irq_moderation_us < - efx->irq_rx_moderation_us) { - channel->irq_moderation_us += step; - efx->type->push_irq_moderation(channel); - } - } - - channel->irq_count = 0; - channel->irq_mod_score = 0; -} - -static int efx_poll(struct napi_struct *napi, int budget) -{ - struct efx_channel *channel = - container_of(napi, struct efx_channel, napi_str); - struct efx_nic *efx = channel->efx; - int spent; - - netif_vdbg(efx, intr, efx->net_dev, - "channel %d NAPI poll executing on CPU %d\n", - channel->channel, raw_smp_processor_id()); - - spent = efx_process_channel(channel, budget); - - xdp_do_flush_map(); - - if (spent < budget) { - if (efx_channel_has_rx_queue(channel) && - efx->irq_rx_adaptive && - unlikely(++channel->irq_count == 1000)) { - efx_update_irq_mod(efx, channel); - } - -#ifdef CONFIG_RFS_ACCEL - /* Perhaps expire some ARFS filters */ - mod_delayed_work(system_wq, &channel->filter_work, 0); -#endif - - /* There is no race here; although napi_disable() will - * only wait for napi_complete(), this isn't a problem - * since efx_nic_eventq_read_ack() will have no effect if - * interrupts have already been disabled. - */ - if (napi_complete_done(napi, spent)) - efx_nic_eventq_read_ack(channel); - } - - return spent; -} - -/* Create event queue - * Event queue memory allocations are done only once. If the channel - * is reset, the memory buffer will be reused; this guards against - * errors during channel reset and also simplifies interrupt handling. - */ -static int efx_probe_eventq(struct efx_channel *channel) -{ - struct efx_nic *efx = channel->efx; - unsigned long entries; - - netif_dbg(efx, probe, efx->net_dev, - "chan %d create event queue\n", channel->channel); - - /* Build an event queue with room for one event per tx and rx buffer, - * plus some extra for link state events and MCDI completions. */ - entries = roundup_pow_of_two(efx->rxq_entries + efx->txq_entries + 128); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_EVQ_SIZE); - channel->eventq_mask = max(entries, EFX_MIN_EVQ_SIZE) - 1; - - return efx_nic_probe_eventq(channel); -} - -/* Prepare channel's event queue */ -static int efx_init_eventq(struct efx_channel *channel) -{ - struct efx_nic *efx = channel->efx; - int rc; - - EFX_WARN_ON_PARANOID(channel->eventq_init); - - netif_dbg(efx, drv, efx->net_dev, - "chan %d init event queue\n", channel->channel); - - rc = efx_nic_init_eventq(channel); - if (rc == 0) { - efx->type->push_irq_moderation(channel); - channel->eventq_read_ptr = 0; - channel->eventq_init = true; - } - return rc; -} - -/* Enable event queue processing and NAPI */ -void efx_start_eventq(struct efx_channel *channel) -{ - netif_dbg(channel->efx, ifup, channel->efx->net_dev, - "chan %d start event queue\n", channel->channel); - - /* Make sure the NAPI handler sees the enabled flag set */ - channel->enabled = true; - smp_wmb(); - - napi_enable(&channel->napi_str); - efx_nic_eventq_read_ack(channel); -} - -/* Disable event queue processing and NAPI */ -void efx_stop_eventq(struct efx_channel *channel) -{ - if (!channel->enabled) - return; - - napi_disable(&channel->napi_str); - channel->enabled = false; -} - -static void efx_fini_eventq(struct efx_channel *channel) -{ - if (!channel->eventq_init) - return; - - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "chan %d fini event queue\n", channel->channel); - - efx_nic_fini_eventq(channel); - channel->eventq_init = false; -} - -static void efx_remove_eventq(struct efx_channel *channel) -{ - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "chan %d remove event queue\n", channel->channel); - - efx_nic_remove_eventq(channel); -} - -/************************************************************************** - * - * Channel handling - * - *************************************************************************/ - -/* Allocate and initialise a channel structure. */ -static struct efx_channel * -efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) -{ - struct efx_channel *channel; - struct efx_rx_queue *rx_queue; - struct efx_tx_queue *tx_queue; - int j; - - channel = kzalloc(sizeof(*channel), GFP_KERNEL); - if (!channel) - return NULL; - - channel->efx = efx; - channel->channel = i; - channel->type = &efx_default_channel_type; - - for (j = 0; j < EFX_TXQ_TYPES; j++) { - tx_queue = &channel->tx_queue[j]; - tx_queue->efx = efx; - tx_queue->queue = i * EFX_TXQ_TYPES + j; - tx_queue->channel = channel; - } - -#ifdef CONFIG_RFS_ACCEL - INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); -#endif - - rx_queue = &channel->rx_queue; - rx_queue->efx = efx; - timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); - - return channel; -} - -/* Allocate and initialise a channel structure, copying parameters - * (but not resources) from an old channel structure. - */ -static struct efx_channel * -efx_copy_channel(const struct efx_channel *old_channel) -{ - struct efx_channel *channel; - struct efx_rx_queue *rx_queue; - struct efx_tx_queue *tx_queue; - int j; - - channel = kmalloc(sizeof(*channel), GFP_KERNEL); - if (!channel) - return NULL; - - *channel = *old_channel; - - channel->napi_dev = NULL; - INIT_HLIST_NODE(&channel->napi_str.napi_hash_node); - channel->napi_str.napi_id = 0; - channel->napi_str.state = 0; - memset(&channel->eventq, 0, sizeof(channel->eventq)); - - for (j = 0; j < EFX_TXQ_TYPES; j++) { - tx_queue = &channel->tx_queue[j]; - if (tx_queue->channel) - tx_queue->channel = channel; - tx_queue->buffer = NULL; - memset(&tx_queue->txd, 0, sizeof(tx_queue->txd)); - } - - rx_queue = &channel->rx_queue; - rx_queue->buffer = NULL; - memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); - timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); -#ifdef CONFIG_RFS_ACCEL - INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); -#endif - - return channel; -} - -static int efx_probe_channel(struct efx_channel *channel) -{ - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int rc; - - netif_dbg(channel->efx, probe, channel->efx->net_dev, - "creating channel %d\n", channel->channel); - - rc = channel->type->pre_probe(channel); - if (rc) - goto fail; - - rc = efx_probe_eventq(channel); - if (rc) - goto fail; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - rc = efx_probe_tx_queue(tx_queue); - if (rc) - goto fail; - } - - efx_for_each_channel_rx_queue(rx_queue, channel) { - rc = efx_probe_rx_queue(rx_queue); - if (rc) - goto fail; - } - - channel->rx_list = NULL; - - return 0; - -fail: - efx_remove_channel(channel); - return rc; -} - -static void -efx_get_channel_name(struct efx_channel *channel, char *buf, size_t len) -{ - struct efx_nic *efx = channel->efx; - const char *type; - int number; - - number = channel->channel; - - if (number >= efx->xdp_channel_offset && - !WARN_ON_ONCE(!efx->n_xdp_channels)) { - type = "-xdp"; - number -= efx->xdp_channel_offset; - } else if (efx->tx_channel_offset == 0) { - type = ""; - } else if (number < efx->tx_channel_offset) { - type = "-rx"; - } else { - type = "-tx"; - number -= efx->tx_channel_offset; - } - snprintf(buf, len, "%s%s-%d", efx->name, type, number); -} - -static void efx_set_channel_names(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - channel->type->get_name(channel, - efx->msi_context[channel->channel].name, - sizeof(efx->msi_context[0].name)); -} - -static int efx_probe_channels(struct efx_nic *efx) -{ - struct efx_channel *channel; - int rc; - - /* Restart special buffer allocation */ - efx->next_buffer_table = 0; - - /* Probe channels in reverse, so that any 'extra' channels - * use the start of the buffer table. This allows the traffic - * channels to be resized without moving them or wasting the - * entries before them. - */ - efx_for_each_channel_rev(channel, efx) { - rc = efx_probe_channel(channel); - if (rc) { - netif_err(efx, probe, efx->net_dev, - "failed to create channel %d\n", - channel->channel); - goto fail; - } - } - efx_set_channel_names(efx); - - return 0; - -fail: - efx_remove_channels(efx); - return rc; -} - -/* Channels are shutdown and reinitialised whilst the NIC is running - * to propagate configuration changes (mtu, checksum offload), or - * to clear hardware error conditions - */ -static void efx_start_datapath(struct efx_nic *efx) -{ - netdev_features_t old_features = efx->net_dev->features; - bool old_rx_scatter = efx->rx_scatter; - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - struct efx_channel *channel; - size_t rx_buf_len; - - /* Calculate the rx buffer allocation parameters required to - * support the current MTU, including padding for header - * alignment and overruns. - */ - efx->rx_dma_len = (efx->rx_prefix_size + - EFX_MAX_FRAME_LEN(efx->net_dev->mtu) + - efx->type->rx_buffer_padding); - rx_buf_len = (sizeof(struct efx_rx_page_state) + XDP_PACKET_HEADROOM + - efx->rx_ip_align + efx->rx_dma_len); - if (rx_buf_len <= PAGE_SIZE) { - efx->rx_scatter = efx->type->always_rx_scatter; - efx->rx_buffer_order = 0; - } else if (efx->type->can_rx_scatter) { - BUILD_BUG_ON(EFX_RX_USR_BUF_SIZE % L1_CACHE_BYTES); - BUILD_BUG_ON(sizeof(struct efx_rx_page_state) + - 2 * ALIGN(NET_IP_ALIGN + EFX_RX_USR_BUF_SIZE, - EFX_RX_BUF_ALIGNMENT) > - PAGE_SIZE); - efx->rx_scatter = true; - efx->rx_dma_len = EFX_RX_USR_BUF_SIZE; - efx->rx_buffer_order = 0; - } else { - efx->rx_scatter = false; - efx->rx_buffer_order = get_order(rx_buf_len); - } - - efx_rx_config_page_split(efx); - if (efx->rx_buffer_order) - netif_dbg(efx, drv, efx->net_dev, - "RX buf len=%u; page order=%u batch=%u\n", - efx->rx_dma_len, efx->rx_buffer_order, - efx->rx_pages_per_batch); - else - netif_dbg(efx, drv, efx->net_dev, - "RX buf len=%u step=%u bpp=%u; page batch=%u\n", - efx->rx_dma_len, efx->rx_page_buf_step, - efx->rx_bufs_per_page, efx->rx_pages_per_batch); - - /* Restore previously fixed features in hw_features and remove - * features which are fixed now - */ - efx->net_dev->hw_features |= efx->net_dev->features; - efx->net_dev->hw_features &= ~efx->fixed_features; - efx->net_dev->features |= efx->fixed_features; - if (efx->net_dev->features != old_features) - netdev_features_change(efx->net_dev); - - /* RX filters may also have scatter-enabled flags */ - if (efx->rx_scatter != old_rx_scatter) - efx->type->filter_update_rx_scatter(efx); - - /* We must keep at least one descriptor in a TX ring empty. - * We could avoid this when the queue size does not exactly - * match the hardware ring size, but it's not that important. - * Therefore we stop the queue when one more skb might fill - * the ring completely. We wake it when half way back to - * empty. - */ - efx->txq_stop_thresh = efx->txq_entries - efx_tx_max_skb_descs(efx); - efx->txq_wake_thresh = efx->txq_stop_thresh / 2; - - /* Initialise the channels */ - efx_for_each_channel(channel, efx) { - efx_for_each_channel_tx_queue(tx_queue, channel) { - efx_init_tx_queue(tx_queue); - atomic_inc(&efx->active_queues); - } - - efx_for_each_channel_rx_queue(rx_queue, channel) { - efx_init_rx_queue(rx_queue); - atomic_inc(&efx->active_queues); - efx_stop_eventq(channel); - efx_fast_push_rx_descriptors(rx_queue, false); - efx_start_eventq(channel); - } - - WARN_ON(channel->rx_pkt_n_frags); - } - - efx_ptp_start_datapath(efx); - - if (netif_device_present(efx->net_dev)) - netif_tx_wake_all_queues(efx->net_dev); -} - -static void efx_stop_datapath(struct efx_nic *efx) -{ - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int rc; - - EFX_ASSERT_RESET_SERIALISED(efx); - BUG_ON(efx->port_enabled); - - efx_ptp_stop_datapath(efx); - - /* Stop RX refill */ - efx_for_each_channel(channel, efx) { - efx_for_each_channel_rx_queue(rx_queue, channel) - rx_queue->refill_enabled = false; - } - - efx_for_each_channel(channel, efx) { - /* RX packet processing is pipelined, so wait for the - * NAPI handler to complete. At least event queue 0 - * might be kept active by non-data events, so don't - * use napi_synchronize() but actually disable NAPI - * temporarily. - */ - if (efx_channel_has_rx_queue(channel)) { - efx_stop_eventq(channel); - efx_start_eventq(channel); - } - } - - rc = efx->type->fini_dmaq(efx); - if (rc) { - netif_err(efx, drv, efx->net_dev, "failed to flush queues\n"); - } else { - netif_dbg(efx, drv, efx->net_dev, - "successfully flushed all queues\n"); - } - - efx_for_each_channel(channel, efx) { - efx_for_each_channel_rx_queue(rx_queue, channel) - efx_fini_rx_queue(rx_queue); - efx_for_each_possible_channel_tx_queue(tx_queue, channel) - efx_fini_tx_queue(tx_queue); - } - efx->xdp_rxq_info_failed = false; -} - -static void efx_remove_channel(struct efx_channel *channel) -{ - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - - netif_dbg(channel->efx, drv, channel->efx->net_dev, - "destroy chan %d\n", channel->channel); - - efx_for_each_channel_rx_queue(rx_queue, channel) - efx_remove_rx_queue(rx_queue); - efx_for_each_possible_channel_tx_queue(tx_queue, channel) - efx_remove_tx_queue(tx_queue); - efx_remove_eventq(channel); - channel->type->post_remove(channel); -} - -static void efx_remove_channels(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - efx_remove_channel(channel); - - kfree(efx->xdp_tx_queues); -} - -int -efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries) -{ - struct efx_channel *other_channel[EFX_MAX_CHANNELS], *channel; - u32 old_rxq_entries, old_txq_entries; - unsigned i, next_buffer_table = 0; - int rc, rc2; - - rc = efx_check_disabled(efx); - if (rc) - return rc; - - /* Not all channels should be reallocated. We must avoid - * reallocating their buffer table entries. - */ - efx_for_each_channel(channel, efx) { - struct efx_rx_queue *rx_queue; - struct efx_tx_queue *tx_queue; - - if (channel->type->copy) - continue; - next_buffer_table = max(next_buffer_table, - channel->eventq.index + - channel->eventq.entries); - efx_for_each_channel_rx_queue(rx_queue, channel) - next_buffer_table = max(next_buffer_table, - rx_queue->rxd.index + - rx_queue->rxd.entries); - efx_for_each_channel_tx_queue(tx_queue, channel) - next_buffer_table = max(next_buffer_table, - tx_queue->txd.index + - tx_queue->txd.entries); - } - - efx_device_detach_sync(efx); - efx_stop_all(efx); - efx_soft_disable_interrupts(efx); - - /* Clone channels (where possible) */ - memset(other_channel, 0, sizeof(other_channel)); - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - if (channel->type->copy) - channel = channel->type->copy(channel); - if (!channel) { - rc = -ENOMEM; - goto out; - } - other_channel[i] = channel; - } - - /* Swap entry counts and channel pointers */ - old_rxq_entries = efx->rxq_entries; - old_txq_entries = efx->txq_entries; - efx->rxq_entries = rxq_entries; - efx->txq_entries = txq_entries; - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - efx->channel[i] = other_channel[i]; - other_channel[i] = channel; - } - - /* Restart buffer table allocation */ - efx->next_buffer_table = next_buffer_table; - - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - if (!channel->type->copy) - continue; - rc = efx_probe_channel(channel); - if (rc) - goto rollback; - efx_init_napi_channel(efx->channel[i]); - } - -out: - /* Destroy unused channel structures */ - for (i = 0; i < efx->n_channels; i++) { - channel = other_channel[i]; - if (channel && channel->type->copy) { - efx_fini_napi_channel(channel); - efx_remove_channel(channel); - kfree(channel); - } - } - - rc2 = efx_soft_enable_interrupts(efx); - if (rc2) { - rc = rc ? rc : rc2; - netif_err(efx, drv, efx->net_dev, - "unable to restart interrupts on channel reallocation\n"); - efx_schedule_reset(efx, RESET_TYPE_DISABLE); - } else { - efx_start_all(efx); - efx_device_attach_if_not_resetting(efx); - } - return rc; - -rollback: - /* Swap back */ - efx->rxq_entries = old_rxq_entries; - efx->txq_entries = old_txq_entries; - for (i = 0; i < efx->n_channels; i++) { - channel = efx->channel[i]; - efx->channel[i] = other_channel[i]; - other_channel[i] = channel; - } - goto out; -} - -void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue) -{ - mod_timer(&rx_queue->slow_fill, jiffies + msecs_to_jiffies(10)); -} - -static bool efx_default_channel_want_txqs(struct efx_channel *channel) -{ - return channel->channel - channel->efx->tx_channel_offset < - channel->efx->n_tx_channels; -} - -static const struct efx_channel_type efx_default_channel_type = { - .pre_probe = efx_channel_dummy_op_int, - .post_remove = efx_channel_dummy_op_void, - .get_name = efx_get_channel_name, - .copy = efx_copy_channel, - .want_txqs = efx_default_channel_want_txqs, - .keep_eventq = false, - .want_pio = true, -}; - -int efx_channel_dummy_op_int(struct efx_channel *channel) -{ - return 0; -} - -void efx_channel_dummy_op_void(struct efx_channel *channel) -{ -} - /************************************************************************** * * Port handling * **************************************************************************/ -/* This ensures that the kernel is kept informed (via - * netif_carrier_on/off) of the link status, and also maintains the - * link status's stop on the port's TX queue. - */ -void efx_link_status_changed(struct efx_nic *efx) -{ - struct efx_link_state *link_state = &efx->link_state; - - /* SFC Bug 5356: A net_dev notifier is registered, so we must ensure - * that no events are triggered between unregister_netdev() and the - * driver unloading. A more general condition is that NETDEV_CHANGE - * can only be generated between NETDEV_UP and NETDEV_DOWN */ - if (!netif_running(efx->net_dev)) - return; - - if (link_state->up != netif_carrier_ok(efx->net_dev)) { - efx->n_link_state_changes++; - - if (link_state->up) - netif_carrier_on(efx->net_dev); - else - netif_carrier_off(efx->net_dev); - } - - /* Status message for kernel log */ - if (link_state->up) - netif_info(efx, link, efx->net_dev, - "link up at %uMbps %s-duplex (MTU %d)\n", - link_state->speed, link_state->fd ? "full" : "half", - efx->net_dev->mtu); - else - netif_info(efx, link, efx->net_dev, "link down\n"); -} - -void efx_link_set_advertising(struct efx_nic *efx, - const unsigned long *advertising) -{ - memcpy(efx->link_advertising, advertising, - sizeof(__ETHTOOL_DECLARE_LINK_MODE_MASK())); - - efx->link_advertising[0] |= ADVERTISED_Autoneg; - if (advertising[0] & ADVERTISED_Pause) - efx->wanted_fc |= (EFX_FC_TX | EFX_FC_RX); - else - efx->wanted_fc &= ~(EFX_FC_TX | EFX_FC_RX); - if (advertising[0] & ADVERTISED_Asym_Pause) - efx->wanted_fc ^= EFX_FC_TX; -} - /* Equivalent to efx_link_set_advertising with all-zeroes, except does not * force the Autoneg bit on. */ @@ -1035,73 +160,6 @@ void efx_link_set_wanted_fc(struct efx_nic *efx, u8 wanted_fc) static void efx_fini_port(struct efx_nic *efx); -/* We assume that efx->type->reconfigure_mac will always try to sync RX - * filters and therefore needs to read-lock the filter table against freeing - */ -void efx_mac_reconfigure(struct efx_nic *efx) -{ - down_read(&efx->filter_sem); - efx->type->reconfigure_mac(efx); - up_read(&efx->filter_sem); -} - -/* Push loopback/power/transmit disable settings to the PHY, and reconfigure - * the MAC appropriately. All other PHY configuration changes are pushed - * through phy_op->set_settings(), and pushed asynchronously to the MAC - * through efx_monitor(). - * - * Callers must hold the mac_lock - */ -int __efx_reconfigure_port(struct efx_nic *efx) -{ - enum efx_phy_mode phy_mode; - int rc; - - WARN_ON(!mutex_is_locked(&efx->mac_lock)); - - /* Disable PHY transmit in mac level loopbacks */ - phy_mode = efx->phy_mode; - if (LOOPBACK_INTERNAL(efx)) - efx->phy_mode |= PHY_MODE_TX_DISABLED; - else - efx->phy_mode &= ~PHY_MODE_TX_DISABLED; - - rc = efx->type->reconfigure_port(efx); - - if (rc) - efx->phy_mode = phy_mode; - - return rc; -} - -/* Reinitialise the MAC to pick up new PHY settings, even if the port is - * disabled. */ -int efx_reconfigure_port(struct efx_nic *efx) -{ - int rc; - - EFX_ASSERT_RESET_SERIALISED(efx); - - mutex_lock(&efx->mac_lock); - rc = __efx_reconfigure_port(efx); - mutex_unlock(&efx->mac_lock); - - return rc; -} - -/* Asynchronous work item for changing MAC promiscuity and multicast - * hash. Avoid a drain/rx_ingress enable by reconfiguring the current - * MAC directly. */ -static void efx_mac_work(struct work_struct *data) -{ - struct efx_nic *efx = container_of(data, struct efx_nic, mac_work); - - mutex_lock(&efx->mac_lock); - if (efx->port_enabled) - efx_mac_reconfigure(efx); - mutex_unlock(&efx->mac_lock); -} - static int efx_probe_port(struct efx_nic *efx) { int rc; @@ -1155,44 +213,6 @@ fail1: return rc; } -static void efx_start_port(struct efx_nic *efx) -{ - netif_dbg(efx, ifup, efx->net_dev, "start port\n"); - BUG_ON(efx->port_enabled); - - mutex_lock(&efx->mac_lock); - efx->port_enabled = true; - - /* Ensure MAC ingress/egress is enabled */ - efx_mac_reconfigure(efx); - - mutex_unlock(&efx->mac_lock); -} - -/* Cancel work for MAC reconfiguration, periodic hardware monitoring - * and the async self-test, wait for them to finish and prevent them - * being scheduled again. This doesn't cover online resets, which - * should only be cancelled when removing the device. - */ -static void efx_stop_port(struct efx_nic *efx) -{ - netif_dbg(efx, ifdown, efx->net_dev, "stop port\n"); - - EFX_ASSERT_RESET_SERIALISED(efx); - - mutex_lock(&efx->mac_lock); - efx->port_enabled = false; - mutex_unlock(&efx->mac_lock); - - /* Serialise against efx_set_multicast_list() */ - netif_addr_lock_bh(efx->net_dev); - netif_addr_unlock_bh(efx->net_dev); - - cancel_delayed_work_sync(&efx->monitor_work); - efx_selftest_async_cancel(efx); - cancel_work_sync(&efx->mac_work); -} - static void efx_fini_port(struct efx_nic *efx) { netif_dbg(efx, drv, efx->net_dev, "shut down port\n"); @@ -1291,582 +311,6 @@ static void efx_dissociate(struct efx_nic *efx) } } -/* This configures the PCI device to enable I/O and DMA. */ -static int efx_init_io(struct efx_nic *efx) -{ - struct pci_dev *pci_dev = efx->pci_dev; - dma_addr_t dma_mask = efx->type->max_dma_mask; - unsigned int mem_map_size = efx->type->mem_map_size(efx); - int rc, bar; - - netif_dbg(efx, probe, efx->net_dev, "initialising I/O\n"); - - bar = efx->type->mem_bar(efx); - - rc = pci_enable_device(pci_dev); - if (rc) { - netif_err(efx, probe, efx->net_dev, - "failed to enable PCI device\n"); - goto fail1; - } - - pci_set_master(pci_dev); - - /* Set the PCI DMA mask. Try all possibilities from our genuine mask - * down to 32 bits, because some architectures will allow 40 bit - * masks event though they reject 46 bit masks. - */ - while (dma_mask > 0x7fffffffUL) { - rc = dma_set_mask_and_coherent(&pci_dev->dev, dma_mask); - if (rc == 0) - break; - dma_mask >>= 1; - } - if (rc) { - netif_err(efx, probe, efx->net_dev, - "could not find a suitable DMA mask\n"); - goto fail2; - } - netif_dbg(efx, probe, efx->net_dev, - "using DMA mask %llx\n", (unsigned long long) dma_mask); - - efx->membase_phys = pci_resource_start(efx->pci_dev, bar); - rc = pci_request_region(pci_dev, bar, "sfc"); - if (rc) { - netif_err(efx, probe, efx->net_dev, - "request for memory BAR failed\n"); - rc = -EIO; - goto fail3; - } - efx->membase = ioremap(efx->membase_phys, mem_map_size); - if (!efx->membase) { - netif_err(efx, probe, efx->net_dev, - "could not map memory BAR at %llx+%x\n", - (unsigned long long)efx->membase_phys, mem_map_size); - rc = -ENOMEM; - goto fail4; - } - netif_dbg(efx, probe, efx->net_dev, - "memory BAR at %llx+%x (virtual %p)\n", - (unsigned long long)efx->membase_phys, mem_map_size, - efx->membase); - - return 0; - - fail4: - pci_release_region(efx->pci_dev, bar); - fail3: - efx->membase_phys = 0; - fail2: - pci_disable_device(efx->pci_dev); - fail1: - return rc; -} - -static void efx_fini_io(struct efx_nic *efx) -{ - int bar; - - netif_dbg(efx, drv, efx->net_dev, "shutting down I/O\n"); - - if (efx->membase) { - iounmap(efx->membase); - efx->membase = NULL; - } - - if (efx->membase_phys) { - bar = efx->type->mem_bar(efx); - pci_release_region(efx->pci_dev, bar); - efx->membase_phys = 0; - } - - /* Don't disable bus-mastering if VFs are assigned */ - if (!pci_vfs_assigned(efx->pci_dev)) - pci_disable_device(efx->pci_dev); -} - -void efx_set_default_rx_indir_table(struct efx_nic *efx, - struct efx_rss_context *ctx) -{ - size_t i; - - for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) - ctx->rx_indir_table[i] = - ethtool_rxfh_indir_default(i, efx->rss_spread); -} - -static unsigned int efx_wanted_parallelism(struct efx_nic *efx) -{ - cpumask_var_t thread_mask; - unsigned int count; - int cpu; - - if (rss_cpus) { - count = rss_cpus; - } else { - if (unlikely(!zalloc_cpumask_var(&thread_mask, GFP_KERNEL))) { - netif_warn(efx, probe, efx->net_dev, - "RSS disabled due to allocation failure\n"); - return 1; - } - - count = 0; - for_each_online_cpu(cpu) { - if (!cpumask_test_cpu(cpu, thread_mask)) { - ++count; - cpumask_or(thread_mask, thread_mask, - topology_sibling_cpumask(cpu)); - } - } - - free_cpumask_var(thread_mask); - } - - if (count > EFX_MAX_RX_QUEUES) { - netif_cond_dbg(efx, probe, efx->net_dev, !rss_cpus, warn, - "Reducing number of rx queues from %u to %u.\n", - count, EFX_MAX_RX_QUEUES); - count = EFX_MAX_RX_QUEUES; - } - - /* If RSS is requested for the PF *and* VFs then we can't write RSS - * table entries that are inaccessible to VFs - */ -#ifdef CONFIG_SFC_SRIOV - if (efx->type->sriov_wanted) { - if (efx->type->sriov_wanted(efx) && efx_vf_size(efx) > 1 && - count > efx_vf_size(efx)) { - netif_warn(efx, probe, efx->net_dev, - "Reducing number of RSS channels from %u to %u for " - "VF support. Increase vf-msix-limit to use more " - "channels on the PF.\n", - count, efx_vf_size(efx)); - count = efx_vf_size(efx); - } - } -#endif - - return count; -} - -static int efx_allocate_msix_channels(struct efx_nic *efx, - unsigned int max_channels, - unsigned int extra_channels, - unsigned int parallelism) -{ - unsigned int n_channels = parallelism; - int vec_count; - int n_xdp_tx; - int n_xdp_ev; - - if (efx_separate_tx_channels) - n_channels *= 2; - n_channels += extra_channels; - - /* To allow XDP transmit to happen from arbitrary NAPI contexts - * we allocate a TX queue per CPU. We share event queues across - * multiple tx queues, assuming tx and ev queues are both - * maximum size. - */ - - n_xdp_tx = num_possible_cpus(); - n_xdp_ev = DIV_ROUND_UP(n_xdp_tx, EFX_TXQ_TYPES); - - vec_count = pci_msix_vec_count(efx->pci_dev); - if (vec_count < 0) - return vec_count; - - max_channels = min_t(unsigned int, vec_count, max_channels); - - /* Check resources. - * We need a channel per event queue, plus a VI per tx queue. - * This may be more pessimistic than it needs to be. - */ - if (n_channels + n_xdp_ev > max_channels) { - netif_err(efx, drv, efx->net_dev, - "Insufficient resources for %d XDP event queues (%d other channels, max %d)\n", - n_xdp_ev, n_channels, max_channels); - efx->n_xdp_channels = 0; - efx->xdp_tx_per_channel = 0; - efx->xdp_tx_queue_count = 0; - } else { - efx->n_xdp_channels = n_xdp_ev; - efx->xdp_tx_per_channel = EFX_TXQ_TYPES; - efx->xdp_tx_queue_count = n_xdp_tx; - n_channels += n_xdp_ev; - netif_dbg(efx, drv, efx->net_dev, - "Allocating %d TX and %d event queues for XDP\n", - n_xdp_tx, n_xdp_ev); - } - - if (vec_count < n_channels) { - netif_err(efx, drv, efx->net_dev, - "WARNING: Insufficient MSI-X vectors available (%d < %u).\n", - vec_count, n_channels); - netif_err(efx, drv, efx->net_dev, - "WARNING: Performance may be reduced.\n"); - n_channels = vec_count; - } - - n_channels = min(n_channels, max_channels); - - efx->n_channels = n_channels; - - /* Ignore XDP tx channels when creating rx channels. */ - n_channels -= efx->n_xdp_channels; - - if (efx_separate_tx_channels) { - efx->n_tx_channels = - min(max(n_channels / 2, 1U), - efx->max_tx_channels); - efx->tx_channel_offset = - n_channels - efx->n_tx_channels; - efx->n_rx_channels = - max(n_channels - - efx->n_tx_channels, 1U); - } else { - efx->n_tx_channels = min(n_channels, efx->max_tx_channels); - efx->tx_channel_offset = 0; - efx->n_rx_channels = n_channels; - } - - efx->n_rx_channels = min(efx->n_rx_channels, parallelism); - efx->n_tx_channels = min(efx->n_tx_channels, parallelism); - - efx->xdp_channel_offset = n_channels; - - netif_dbg(efx, drv, efx->net_dev, - "Allocating %u RX channels\n", - efx->n_rx_channels); - - return efx->n_channels; -} - -/* Probe the number and type of interrupts we are able to obtain, and - * the resulting numbers of channels and RX queues. - */ -static int efx_probe_interrupts(struct efx_nic *efx) -{ - unsigned int extra_channels = 0; - unsigned int rss_spread; - unsigned int i, j; - int rc; - - for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) - if (efx->extra_channel_type[i]) - ++extra_channels; - - if (efx->interrupt_mode == EFX_INT_MODE_MSIX) { - unsigned int parallelism = efx_wanted_parallelism(efx); - struct msix_entry xentries[EFX_MAX_CHANNELS]; - unsigned int n_channels; - - rc = efx_allocate_msix_channels(efx, efx->max_channels, - extra_channels, parallelism); - if (rc >= 0) { - n_channels = rc; - for (i = 0; i < n_channels; i++) - xentries[i].entry = i; - rc = pci_enable_msix_range(efx->pci_dev, xentries, 1, - n_channels); - } - if (rc < 0) { - /* Fall back to single channel MSI */ - netif_err(efx, drv, efx->net_dev, - "could not enable MSI-X\n"); - if (efx->type->min_interrupt_mode >= EFX_INT_MODE_MSI) - efx->interrupt_mode = EFX_INT_MODE_MSI; - else - return rc; - } else if (rc < n_channels) { - netif_err(efx, drv, efx->net_dev, - "WARNING: Insufficient MSI-X vectors" - " available (%d < %u).\n", rc, n_channels); - netif_err(efx, drv, efx->net_dev, - "WARNING: Performance may be reduced.\n"); - n_channels = rc; - } - - if (rc > 0) { - for (i = 0; i < efx->n_channels; i++) - efx_get_channel(efx, i)->irq = - xentries[i].vector; - } - } - - /* Try single interrupt MSI */ - if (efx->interrupt_mode == EFX_INT_MODE_MSI) { - efx->n_channels = 1; - efx->n_rx_channels = 1; - efx->n_tx_channels = 1; - efx->n_xdp_channels = 0; - efx->xdp_channel_offset = efx->n_channels; - rc = pci_enable_msi(efx->pci_dev); - if (rc == 0) { - efx_get_channel(efx, 0)->irq = efx->pci_dev->irq; - } else { - netif_err(efx, drv, efx->net_dev, - "could not enable MSI\n"); - if (efx->type->min_interrupt_mode >= EFX_INT_MODE_LEGACY) - efx->interrupt_mode = EFX_INT_MODE_LEGACY; - else - return rc; - } - } - - /* Assume legacy interrupts */ - if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) { - efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0); - efx->n_rx_channels = 1; - efx->n_tx_channels = 1; - efx->n_xdp_channels = 0; - efx->xdp_channel_offset = efx->n_channels; - efx->legacy_irq = efx->pci_dev->irq; - } - - /* Assign extra channels if possible, before XDP channels */ - efx->n_extra_tx_channels = 0; - j = efx->xdp_channel_offset; - for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) { - if (!efx->extra_channel_type[i]) - continue; - if (j <= efx->tx_channel_offset + efx->n_tx_channels) { - efx->extra_channel_type[i]->handle_no_channel(efx); - } else { - --j; - efx_get_channel(efx, j)->type = - efx->extra_channel_type[i]; - if (efx_channel_has_tx_queues(efx_get_channel(efx, j))) - efx->n_extra_tx_channels++; - } - } - - rss_spread = efx->n_rx_channels; - /* RSS might be usable on VFs even if it is disabled on the PF */ -#ifdef CONFIG_SFC_SRIOV - if (efx->type->sriov_wanted) { - efx->rss_spread = ((rss_spread > 1 || - !efx->type->sriov_wanted(efx)) ? - rss_spread : efx_vf_size(efx)); - return 0; - } -#endif - efx->rss_spread = rss_spread; - - return 0; -} - -#if defined(CONFIG_SMP) -static void efx_set_interrupt_affinity(struct efx_nic *efx) -{ - struct efx_channel *channel; - unsigned int cpu; - - efx_for_each_channel(channel, efx) { - cpu = cpumask_local_spread(channel->channel, - pcibus_to_node(efx->pci_dev->bus)); - irq_set_affinity_hint(channel->irq, cpumask_of(cpu)); - } -} - -static void efx_clear_interrupt_affinity(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - irq_set_affinity_hint(channel->irq, NULL); -} -#else -static void -efx_set_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) -{ -} - -static void -efx_clear_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) -{ -} -#endif /* CONFIG_SMP */ - -static int efx_soft_enable_interrupts(struct efx_nic *efx) -{ - struct efx_channel *channel, *end_channel; - int rc; - - BUG_ON(efx->state == STATE_DISABLED); - - efx->irq_soft_enabled = true; - smp_wmb(); - - efx_for_each_channel(channel, efx) { - if (!channel->type->keep_eventq) { - rc = efx_init_eventq(channel); - if (rc) - goto fail; - } - efx_start_eventq(channel); - } - - efx_mcdi_mode_event(efx); - - return 0; -fail: - end_channel = channel; - efx_for_each_channel(channel, efx) { - if (channel == end_channel) - break; - efx_stop_eventq(channel); - if (!channel->type->keep_eventq) - efx_fini_eventq(channel); - } - - return rc; -} - -static void efx_soft_disable_interrupts(struct efx_nic *efx) -{ - struct efx_channel *channel; - - if (efx->state == STATE_DISABLED) - return; - - efx_mcdi_mode_poll(efx); - - efx->irq_soft_enabled = false; - smp_wmb(); - - if (efx->legacy_irq) - synchronize_irq(efx->legacy_irq); - - efx_for_each_channel(channel, efx) { - if (channel->irq) - synchronize_irq(channel->irq); - - efx_stop_eventq(channel); - if (!channel->type->keep_eventq) - efx_fini_eventq(channel); - } - - /* Flush the asynchronous MCDI request queue */ - efx_mcdi_flush_async(efx); -} - -static int efx_enable_interrupts(struct efx_nic *efx) -{ - struct efx_channel *channel, *end_channel; - int rc; - - BUG_ON(efx->state == STATE_DISABLED); - - if (efx->eeh_disabled_legacy_irq) { - enable_irq(efx->legacy_irq); - efx->eeh_disabled_legacy_irq = false; - } - - efx->type->irq_enable_master(efx); - - efx_for_each_channel(channel, efx) { - if (channel->type->keep_eventq) { - rc = efx_init_eventq(channel); - if (rc) - goto fail; - } - } - - rc = efx_soft_enable_interrupts(efx); - if (rc) - goto fail; - - return 0; - -fail: - end_channel = channel; - efx_for_each_channel(channel, efx) { - if (channel == end_channel) - break; - if (channel->type->keep_eventq) - efx_fini_eventq(channel); - } - - efx->type->irq_disable_non_ev(efx); - - return rc; -} - -static void efx_disable_interrupts(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_soft_disable_interrupts(efx); - - efx_for_each_channel(channel, efx) { - if (channel->type->keep_eventq) - efx_fini_eventq(channel); - } - - efx->type->irq_disable_non_ev(efx); -} - -static void efx_remove_interrupts(struct efx_nic *efx) -{ - struct efx_channel *channel; - - /* Remove MSI/MSI-X interrupts */ - efx_for_each_channel(channel, efx) - channel->irq = 0; - pci_disable_msi(efx->pci_dev); - pci_disable_msix(efx->pci_dev); - - /* Remove legacy interrupt */ - efx->legacy_irq = 0; -} - -static int efx_set_channels(struct efx_nic *efx) -{ - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - int xdp_queue_number; - - efx->tx_channel_offset = - efx_separate_tx_channels ? - efx->n_channels - efx->n_tx_channels : 0; - - if (efx->xdp_tx_queue_count) { - EFX_WARN_ON_PARANOID(efx->xdp_tx_queues); - - /* Allocate array for XDP TX queue lookup. */ - efx->xdp_tx_queues = kcalloc(efx->xdp_tx_queue_count, - sizeof(*efx->xdp_tx_queues), - GFP_KERNEL); - if (!efx->xdp_tx_queues) - return -ENOMEM; - } - - /* We need to mark which channels really have RX and TX - * queues, and adjust the TX queue numbers if we have separate - * RX-only and TX-only channels. - */ - xdp_queue_number = 0; - efx_for_each_channel(channel, efx) { - if (channel->channel < efx->n_rx_channels) - channel->rx_queue.core_index = channel->channel; - else - channel->rx_queue.core_index = -1; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - tx_queue->queue -= (efx->tx_channel_offset * - EFX_TXQ_TYPES); - - if (efx_channel_is_xdp_tx(channel) && - xdp_queue_number < efx->xdp_tx_queue_count) { - efx->xdp_tx_queues[xdp_queue_number] = tx_queue; - xdp_queue_number++; - } - } - } - return 0; -} - static int efx_probe_nic(struct efx_nic *efx) { int rc; @@ -1939,70 +383,6 @@ static void efx_remove_nic(struct efx_nic *efx) efx->type->remove(efx); } -static int efx_probe_filters(struct efx_nic *efx) -{ - int rc; - - init_rwsem(&efx->filter_sem); - mutex_lock(&efx->mac_lock); - down_write(&efx->filter_sem); - rc = efx->type->filter_table_probe(efx); - if (rc) - goto out_unlock; - -#ifdef CONFIG_RFS_ACCEL - if (efx->type->offload_features & NETIF_F_NTUPLE) { - struct efx_channel *channel; - int i, success = 1; - - efx_for_each_channel(channel, efx) { - channel->rps_flow_id = - kcalloc(efx->type->max_rx_ip_filters, - sizeof(*channel->rps_flow_id), - GFP_KERNEL); - if (!channel->rps_flow_id) - success = 0; - else - for (i = 0; - i < efx->type->max_rx_ip_filters; - ++i) - channel->rps_flow_id[i] = - RPS_FLOW_ID_INVALID; - channel->rfs_expire_index = 0; - channel->rfs_filter_count = 0; - } - - if (!success) { - efx_for_each_channel(channel, efx) - kfree(channel->rps_flow_id); - efx->type->filter_table_remove(efx); - rc = -ENOMEM; - goto out_unlock; - } - } -#endif -out_unlock: - up_write(&efx->filter_sem); - mutex_unlock(&efx->mac_lock); - return rc; -} - -static void efx_remove_filters(struct efx_nic *efx) -{ -#ifdef CONFIG_RFS_ACCEL - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) { - cancel_delayed_work_sync(&channel->filter_work); - kfree(channel->rps_flow_id); - } -#endif - down_write(&efx->filter_sem); - efx->type->filter_table_remove(efx); - up_write(&efx->filter_sem); -} - - /************************************************************************** * * NIC startup/shutdown @@ -2067,81 +447,6 @@ static int efx_probe_all(struct efx_nic *efx) return rc; } -/* If the interface is supposed to be running but is not, start - * the hardware and software data path, regular activity for the port - * (MAC statistics, link polling, etc.) and schedule the port to be - * reconfigured. Interrupts must already be enabled. This function - * is safe to call multiple times, so long as the NIC is not disabled. - * Requires the RTNL lock. - */ -static void efx_start_all(struct efx_nic *efx) -{ - EFX_ASSERT_RESET_SERIALISED(efx); - BUG_ON(efx->state == STATE_DISABLED); - - /* Check that it is appropriate to restart the interface. All - * of these flags are safe to read under just the rtnl lock */ - if (efx->port_enabled || !netif_running(efx->net_dev) || - efx->reset_pending) - return; - - efx_start_port(efx); - efx_start_datapath(efx); - - /* Start the hardware monitor if there is one */ - if (efx->type->monitor != NULL) - queue_delayed_work(efx->workqueue, &efx->monitor_work, - efx_monitor_interval); - - /* Link state detection is normally event-driven; we have - * to poll now because we could have missed a change - */ - mutex_lock(&efx->mac_lock); - if (efx->phy_op->poll(efx)) - efx_link_status_changed(efx); - mutex_unlock(&efx->mac_lock); - - efx->type->start_stats(efx); - efx->type->pull_stats(efx); - spin_lock_bh(&efx->stats_lock); - efx->type->update_stats(efx, NULL, NULL); - spin_unlock_bh(&efx->stats_lock); -} - -/* Quiesce the hardware and software data path, and regular activity - * for the port without bringing the link down. Safe to call multiple - * times with the NIC in almost any state, but interrupts should be - * enabled. Requires the RTNL lock. - */ -static void efx_stop_all(struct efx_nic *efx) -{ - EFX_ASSERT_RESET_SERIALISED(efx); - - /* port_enabled can be read safely under the rtnl lock */ - if (!efx->port_enabled) - return; - - /* update stats before we go down so we can accurately count - * rx_nodesc_drops - */ - efx->type->pull_stats(efx); - spin_lock_bh(&efx->stats_lock); - efx->type->update_stats(efx, NULL, NULL); - spin_unlock_bh(&efx->stats_lock); - efx->type->stop_stats(efx); - efx_stop_port(efx); - - /* Stop the kernel transmit interface. This is only valid if - * the device is stopped or detached; otherwise the watchdog - * may fire immediately. - */ - WARN_ON(netif_running(efx->net_dev) && - netif_device_present(efx->net_dev)); - netif_tx_disable(efx->net_dev); - - efx_stop_datapath(efx); -} - static void efx_remove_all(struct efx_nic *efx) { rtnl_lock(); @@ -2237,36 +542,6 @@ void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, /************************************************************************** * - * Hardware monitor - * - **************************************************************************/ - -/* Run periodically off the general workqueue */ -static void efx_monitor(struct work_struct *data) -{ - struct efx_nic *efx = container_of(data, struct efx_nic, - monitor_work.work); - - netif_vdbg(efx, timer, efx->net_dev, - "hardware monitor executing on CPU %d\n", - raw_smp_processor_id()); - BUG_ON(efx->type->monitor == NULL); - - /* If the mac_lock is already held then it is likely a port - * reconfiguration is already in place, which will likely do - * most of the work of monitor() anyway. */ - if (mutex_trylock(&efx->mac_lock)) { - if (efx->port_enabled) - efx->type->monitor(efx); - mutex_unlock(&efx->mac_lock); - } - - queue_delayed_work(efx->workqueue, &efx->monitor_work, - efx_monitor_interval); -} - -/************************************************************************** - * * ioctls * *************************************************************************/ @@ -2294,45 +569,6 @@ static int efx_ioctl(struct net_device *net_dev, struct ifreq *ifr, int cmd) /************************************************************************** * - * NAPI interface - * - **************************************************************************/ - -static void efx_init_napi_channel(struct efx_channel *channel) -{ - struct efx_nic *efx = channel->efx; - - channel->napi_dev = efx->net_dev; - netif_napi_add(channel->napi_dev, &channel->napi_str, - efx_poll, napi_weight); -} - -static void efx_init_napi(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - efx_init_napi_channel(channel); -} - -static void efx_fini_napi_channel(struct efx_channel *channel) -{ - if (channel->napi_dev) - netif_napi_del(&channel->napi_str); - - channel->napi_dev = NULL; -} - -static void efx_fini_napi(struct efx_nic *efx) -{ - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) - efx_fini_napi_channel(channel); -} - -/************************************************************************** - * * Kernel net device interface * *************************************************************************/ @@ -2382,19 +618,8 @@ int efx_net_stop(struct net_device *net_dev) return 0; } -/* Context: process, dev_base_lock or RTNL held, non-blocking. */ -static void efx_net_stats(struct net_device *net_dev, - struct rtnl_link_stats64 *stats) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - spin_lock_bh(&efx->stats_lock); - efx->type->update_stats(efx, NULL, stats); - spin_unlock_bh(&efx->stats_lock); -} - /* Context: netif_tx_lock held, BHs disabled. */ -static void efx_watchdog(struct net_device *net_dev) +static void efx_watchdog(struct net_device *net_dev, unsigned int txqueue) { struct efx_nic *efx = netdev_priv(net_dev); @@ -2405,51 +630,6 @@ static void efx_watchdog(struct net_device *net_dev) efx_schedule_reset(efx, RESET_TYPE_TX_WATCHDOG); } -static unsigned int efx_xdp_max_mtu(struct efx_nic *efx) -{ - /* The maximum MTU that we can fit in a single page, allowing for - * framing, overhead and XDP headroom. - */ - int overhead = EFX_MAX_FRAME_LEN(0) + sizeof(struct efx_rx_page_state) + - efx->rx_prefix_size + efx->type->rx_buffer_padding + - efx->rx_ip_align + XDP_PACKET_HEADROOM; - - return PAGE_SIZE - overhead; -} - -/* Context: process, rtnl_lock() held. */ -static int efx_change_mtu(struct net_device *net_dev, int new_mtu) -{ - struct efx_nic *efx = netdev_priv(net_dev); - int rc; - - rc = efx_check_disabled(efx); - if (rc) - return rc; - - if (rtnl_dereference(efx->xdp_prog) && - new_mtu > efx_xdp_max_mtu(efx)) { - netif_err(efx, drv, efx->net_dev, - "Requested MTU of %d too big for XDP (max: %d)\n", - new_mtu, efx_xdp_max_mtu(efx)); - return -EINVAL; - } - - netif_dbg(efx, drv, efx->net_dev, "changing MTU to %d\n", new_mtu); - - efx_device_detach_sync(efx); - efx_stop_all(efx); - - mutex_lock(&efx->mac_lock); - net_dev->mtu = new_mtu; - efx_mac_reconfigure(efx); - mutex_unlock(&efx->mac_lock); - - efx_start_all(efx); - efx_device_attach_if_not_resetting(efx); - return 0; -} - static int efx_set_mac_address(struct net_device *net_dev, void *data) { struct efx_nic *efx = netdev_priv(net_dev); @@ -2726,28 +906,6 @@ show_phy_type(struct device *dev, struct device_attribute *attr, char *buf) } static DEVICE_ATTR(phy_type, 0444, show_phy_type, NULL); -#ifdef CONFIG_SFC_MCDI_LOGGING -static ssize_t show_mcdi_log(struct device *dev, struct device_attribute *attr, - char *buf) -{ - struct efx_nic *efx = dev_get_drvdata(dev); - struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - - return scnprintf(buf, PAGE_SIZE, "%d\n", mcdi->logging_enabled); -} -static ssize_t set_mcdi_log(struct device *dev, struct device_attribute *attr, - const char *buf, size_t count) -{ - struct efx_nic *efx = dev_get_drvdata(dev); - struct efx_mcdi_iface *mcdi = efx_mcdi(efx); - bool enable = count > 0 && *buf != '0'; - - mcdi->logging_enabled = enable; - return count; -} -static DEVICE_ATTR(mcdi_logging, 0644, show_mcdi_log, set_mcdi_log); -#endif - static int efx_register_netdev(struct efx_nic *efx) { struct net_device *net_dev = efx->net_dev; @@ -2807,21 +965,11 @@ static int efx_register_netdev(struct efx_nic *efx) "failed to init net dev attributes\n"); goto fail_registered; } -#ifdef CONFIG_SFC_MCDI_LOGGING - rc = device_create_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); - if (rc) { - netif_err(efx, drv, efx->net_dev, - "failed to init net dev attributes\n"); - goto fail_attr_mcdi_logging; - } -#endif + + efx_init_mcdi_logging(efx); return 0; -#ifdef CONFIG_SFC_MCDI_LOGGING -fail_attr_mcdi_logging: - device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); -#endif fail_registered: rtnl_lock(); efx_dissociate(efx); @@ -2842,9 +990,7 @@ static void efx_unregister_netdev(struct efx_nic *efx) if (efx_dev_registered(efx)) { strlcpy(efx->name, pci_name(efx->pci_dev), sizeof(efx->name)); -#ifdef CONFIG_SFC_MCDI_LOGGING - device_remove_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); -#endif + efx_fini_mcdi_logging(efx); device_remove_file(&efx->pci_dev->dev, &dev_attr_phy_type); unregister_netdev(efx->net_dev); } @@ -2852,292 +998,6 @@ static void efx_unregister_netdev(struct efx_nic *efx) /************************************************************************** * - * Device reset and suspend - * - **************************************************************************/ - -/* Tears down the entire software state and most of the hardware state - * before reset. */ -void efx_reset_down(struct efx_nic *efx, enum reset_type method) -{ - EFX_ASSERT_RESET_SERIALISED(efx); - - if (method == RESET_TYPE_MCDI_TIMEOUT) - efx->type->prepare_flr(efx); - - efx_stop_all(efx); - efx_disable_interrupts(efx); - - mutex_lock(&efx->mac_lock); - down_write(&efx->filter_sem); - mutex_lock(&efx->rss_lock); - if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && - method != RESET_TYPE_DATAPATH) - efx->phy_op->fini(efx); - efx->type->fini(efx); -} - -/* This function will always ensure that the locks acquired in - * efx_reset_down() are released. A failure return code indicates - * that we were unable to reinitialise the hardware, and the - * driver should be disabled. If ok is false, then the rx and tx - * engines are not restarted, pending a RESET_DISABLE. */ -int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) -{ - int rc; - - EFX_ASSERT_RESET_SERIALISED(efx); - - if (method == RESET_TYPE_MCDI_TIMEOUT) - efx->type->finish_flr(efx); - - /* Ensure that SRAM is initialised even if we're disabling the device */ - rc = efx->type->init(efx); - if (rc) { - netif_err(efx, drv, efx->net_dev, "failed to initialise NIC\n"); - goto fail; - } - - if (!ok) - goto fail; - - if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && - method != RESET_TYPE_DATAPATH) { - rc = efx->phy_op->init(efx); - if (rc) - goto fail; - rc = efx->phy_op->reconfigure(efx); - if (rc && rc != -EPERM) - netif_err(efx, drv, efx->net_dev, - "could not restore PHY settings\n"); - } - - rc = efx_enable_interrupts(efx); - if (rc) - goto fail; - -#ifdef CONFIG_SFC_SRIOV - rc = efx->type->vswitching_restore(efx); - if (rc) /* not fatal; the PF will still work fine */ - netif_warn(efx, probe, efx->net_dev, - "failed to restore vswitching rc=%d;" - " VFs may not function\n", rc); -#endif - - if (efx->type->rx_restore_rss_contexts) - efx->type->rx_restore_rss_contexts(efx); - mutex_unlock(&efx->rss_lock); - efx->type->filter_table_restore(efx); - up_write(&efx->filter_sem); - if (efx->type->sriov_reset) - efx->type->sriov_reset(efx); - - mutex_unlock(&efx->mac_lock); - - efx_start_all(efx); - - if (efx->type->udp_tnl_push_ports) - efx->type->udp_tnl_push_ports(efx); - - return 0; - -fail: - efx->port_initialized = false; - - mutex_unlock(&efx->rss_lock); - up_write(&efx->filter_sem); - mutex_unlock(&efx->mac_lock); - - return rc; -} - -/* Reset the NIC using the specified method. Note that the reset may - * fail, in which case the card will be left in an unusable state. - * - * Caller must hold the rtnl_lock. - */ -int efx_reset(struct efx_nic *efx, enum reset_type method) -{ - int rc, rc2; - bool disabled; - - netif_info(efx, drv, efx->net_dev, "resetting (%s)\n", - RESET_TYPE(method)); - - efx_device_detach_sync(efx); - efx_reset_down(efx, method); - - rc = efx->type->reset(efx, method); - if (rc) { - netif_err(efx, drv, efx->net_dev, "failed to reset hardware\n"); - goto out; - } - - /* Clear flags for the scopes we covered. We assume the NIC and - * driver are now quiescent so that there is no race here. - */ - if (method < RESET_TYPE_MAX_METHOD) - efx->reset_pending &= -(1 << (method + 1)); - else /* it doesn't fit into the well-ordered scope hierarchy */ - __clear_bit(method, &efx->reset_pending); - - /* Reinitialise bus-mastering, which may have been turned off before - * the reset was scheduled. This is still appropriate, even in the - * RESET_TYPE_DISABLE since this driver generally assumes the hardware - * can respond to requests. */ - pci_set_master(efx->pci_dev); - -out: - /* Leave device stopped if necessary */ - disabled = rc || - method == RESET_TYPE_DISABLE || - method == RESET_TYPE_RECOVER_OR_DISABLE; - rc2 = efx_reset_up(efx, method, !disabled); - if (rc2) { - disabled = true; - if (!rc) - rc = rc2; - } - - if (disabled) { - dev_close(efx->net_dev); - netif_err(efx, drv, efx->net_dev, "has been disabled\n"); - efx->state = STATE_DISABLED; - } else { - netif_dbg(efx, drv, efx->net_dev, "reset complete\n"); - efx_device_attach_if_not_resetting(efx); - } - return rc; -} - -/* Try recovery mechanisms. - * For now only EEH is supported. - * Returns 0 if the recovery mechanisms are unsuccessful. - * Returns a non-zero value otherwise. - */ -int efx_try_recovery(struct efx_nic *efx) -{ -#ifdef CONFIG_EEH - /* A PCI error can occur and not be seen by EEH because nothing - * happens on the PCI bus. In this case the driver may fail and - * schedule a 'recover or reset', leading to this recovery handler. - * Manually call the eeh failure check function. - */ - struct eeh_dev *eehdev = pci_dev_to_eeh_dev(efx->pci_dev); - if (eeh_dev_check_failure(eehdev)) { - /* The EEH mechanisms will handle the error and reset the - * device if necessary. - */ - return 1; - } -#endif - return 0; -} - -static void efx_wait_for_bist_end(struct efx_nic *efx) -{ - int i; - - for (i = 0; i < BIST_WAIT_DELAY_COUNT; ++i) { - if (efx_mcdi_poll_reboot(efx)) - goto out; - msleep(BIST_WAIT_DELAY_MS); - } - - netif_err(efx, drv, efx->net_dev, "Warning: No MC reboot after BIST mode\n"); -out: - /* Either way unset the BIST flag. If we found no reboot we probably - * won't recover, but we should try. - */ - efx->mc_bist_for_other_fn = false; -} - -/* The worker thread exists so that code that cannot sleep can - * schedule a reset for later. - */ -static void efx_reset_work(struct work_struct *data) -{ - struct efx_nic *efx = container_of(data, struct efx_nic, reset_work); - unsigned long pending; - enum reset_type method; - - pending = READ_ONCE(efx->reset_pending); - method = fls(pending) - 1; - - if (method == RESET_TYPE_MC_BIST) - efx_wait_for_bist_end(efx); - - if ((method == RESET_TYPE_RECOVER_OR_DISABLE || - method == RESET_TYPE_RECOVER_OR_ALL) && - efx_try_recovery(efx)) - return; - - if (!pending) - return; - - rtnl_lock(); - - /* We checked the state in efx_schedule_reset() but it may - * have changed by now. Now that we have the RTNL lock, - * it cannot change again. - */ - if (efx->state == STATE_READY) - (void)efx_reset(efx, method); - - rtnl_unlock(); -} - -void efx_schedule_reset(struct efx_nic *efx, enum reset_type type) -{ - enum reset_type method; - - if (efx->state == STATE_RECOVERY) { - netif_dbg(efx, drv, efx->net_dev, - "recovering: skip scheduling %s reset\n", - RESET_TYPE(type)); - return; - } - - switch (type) { - case RESET_TYPE_INVISIBLE: - case RESET_TYPE_ALL: - case RESET_TYPE_RECOVER_OR_ALL: - case RESET_TYPE_WORLD: - case RESET_TYPE_DISABLE: - case RESET_TYPE_RECOVER_OR_DISABLE: - case RESET_TYPE_DATAPATH: - case RESET_TYPE_MC_BIST: - case RESET_TYPE_MCDI_TIMEOUT: - method = type; - netif_dbg(efx, drv, efx->net_dev, "scheduling %s reset\n", - RESET_TYPE(method)); - break; - default: - method = efx->type->map_reset_reason(type); - netif_dbg(efx, drv, efx->net_dev, - "scheduling %s reset for %s\n", - RESET_TYPE(method), RESET_TYPE(type)); - break; - } - - set_bit(method, &efx->reset_pending); - smp_mb(); /* ensure we change reset_pending before checking state */ - - /* If we're not READY then just leave the flags set as the cue - * to abort probing or reschedule the reset later. - */ - if (READ_ONCE(efx->state) != STATE_READY) - return; - - /* efx_process_channel() will no longer read events once a - * reset is scheduled. So switch back to poll'd MCDI completions. */ - efx_mcdi_mode_poll(efx); - - queue_work(reset_workqueue, &efx->reset_work); -} - -/************************************************************************** - * * List of NICs we support * **************************************************************************/ @@ -3169,139 +1029,10 @@ static const struct pci_device_id efx_pci_table[] = { /************************************************************************** * - * Dummy PHY/MAC operations - * - * Can be used for some unimplemented operations - * Needed so all function pointers are valid and do not have to be tested - * before use - * - **************************************************************************/ -int efx_port_dummy_op_int(struct efx_nic *efx) -{ - return 0; -} -void efx_port_dummy_op_void(struct efx_nic *efx) {} - -static bool efx_port_dummy_op_poll(struct efx_nic *efx) -{ - return false; -} - -static const struct efx_phy_operations efx_dummy_phy_operations = { - .init = efx_port_dummy_op_int, - .reconfigure = efx_port_dummy_op_int, - .poll = efx_port_dummy_op_poll, - .fini = efx_port_dummy_op_void, -}; - -/************************************************************************** - * * Data housekeeping * **************************************************************************/ -/* This zeroes out and then fills in the invariants in a struct - * efx_nic (including all sub-structures). - */ -static int efx_init_struct(struct efx_nic *efx, - struct pci_dev *pci_dev, struct net_device *net_dev) -{ - int rc = -ENOMEM, i; - - /* Initialise common structures */ - INIT_LIST_HEAD(&efx->node); - INIT_LIST_HEAD(&efx->secondary_list); - spin_lock_init(&efx->biu_lock); -#ifdef CONFIG_SFC_MTD - INIT_LIST_HEAD(&efx->mtd_list); -#endif - INIT_WORK(&efx->reset_work, efx_reset_work); - INIT_DELAYED_WORK(&efx->monitor_work, efx_monitor); - INIT_DELAYED_WORK(&efx->selftest_work, efx_selftest_async_work); - efx->pci_dev = pci_dev; - efx->msg_enable = debug; - efx->state = STATE_UNINIT; - strlcpy(efx->name, pci_name(pci_dev), sizeof(efx->name)); - - efx->net_dev = net_dev; - efx->rx_prefix_size = efx->type->rx_prefix_size; - efx->rx_ip_align = - NET_IP_ALIGN ? (efx->rx_prefix_size + NET_IP_ALIGN) % 4 : 0; - efx->rx_packet_hash_offset = - efx->type->rx_hash_offset - efx->type->rx_prefix_size; - efx->rx_packet_ts_offset = - efx->type->rx_ts_offset - efx->type->rx_prefix_size; - INIT_LIST_HEAD(&efx->rss_context.list); - mutex_init(&efx->rss_lock); - spin_lock_init(&efx->stats_lock); - efx->vi_stride = EFX_DEFAULT_VI_STRIDE; - efx->num_mac_stats = MC_CMD_MAC_NSTATS; - BUILD_BUG_ON(MC_CMD_MAC_NSTATS - 1 != MC_CMD_MAC_GENERATION_END); - mutex_init(&efx->mac_lock); -#ifdef CONFIG_RFS_ACCEL - mutex_init(&efx->rps_mutex); - spin_lock_init(&efx->rps_hash_lock); - /* Failure to allocate is not fatal, but may degrade ARFS performance */ - efx->rps_hash_table = kcalloc(EFX_ARFS_HASH_TABLE_SIZE, - sizeof(*efx->rps_hash_table), GFP_KERNEL); -#endif - efx->phy_op = &efx_dummy_phy_operations; - efx->mdio.dev = net_dev; - INIT_WORK(&efx->mac_work, efx_mac_work); - init_waitqueue_head(&efx->flush_wq); - - for (i = 0; i < EFX_MAX_CHANNELS; i++) { - efx->channel[i] = efx_alloc_channel(efx, i, NULL); - if (!efx->channel[i]) - goto fail; - efx->msi_context[i].efx = efx; - efx->msi_context[i].index = i; - } - - /* Higher numbered interrupt modes are less capable! */ - if (WARN_ON_ONCE(efx->type->max_interrupt_mode > - efx->type->min_interrupt_mode)) { - rc = -EIO; - goto fail; - } - efx->interrupt_mode = max(efx->type->max_interrupt_mode, - interrupt_mode); - efx->interrupt_mode = min(efx->type->min_interrupt_mode, - interrupt_mode); - - /* Would be good to use the net_dev name, but we're too early */ - snprintf(efx->workqueue_name, sizeof(efx->workqueue_name), "sfc%s", - pci_name(pci_dev)); - efx->workqueue = create_singlethread_workqueue(efx->workqueue_name); - if (!efx->workqueue) - goto fail; - - return 0; - -fail: - efx_fini_struct(efx); - return rc; -} - -static void efx_fini_struct(struct efx_nic *efx) -{ - int i; - -#ifdef CONFIG_RFS_ACCEL - kfree(efx->rps_hash_table); -#endif - - for (i = 0; i < EFX_MAX_CHANNELS; i++) - kfree(efx->channel[i]); - - kfree(efx->vpd_sn); - - if (efx->workqueue) { - destroy_workqueue(efx->workqueue); - efx->workqueue = NULL; - } -} - void efx_update_sw_stats(struct efx_nic *efx, u64 *stats) { u64 n_rx_nodesc_trunc = 0; @@ -3313,197 +1044,6 @@ void efx_update_sw_stats(struct efx_nic *efx, u64 *stats) stats[GENERIC_STAT_rx_noskb_drops] = atomic_read(&efx->n_rx_noskb_drops); } -bool efx_filter_spec_equal(const struct efx_filter_spec *left, - const struct efx_filter_spec *right) -{ - if ((left->match_flags ^ right->match_flags) | - ((left->flags ^ right->flags) & - (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX))) - return false; - - return memcmp(&left->outer_vid, &right->outer_vid, - sizeof(struct efx_filter_spec) - - offsetof(struct efx_filter_spec, outer_vid)) == 0; -} - -u32 efx_filter_spec_hash(const struct efx_filter_spec *spec) -{ - BUILD_BUG_ON(offsetof(struct efx_filter_spec, outer_vid) & 3); - return jhash2((const u32 *)&spec->outer_vid, - (sizeof(struct efx_filter_spec) - - offsetof(struct efx_filter_spec, outer_vid)) / 4, - 0); -} - -#ifdef CONFIG_RFS_ACCEL -bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, - bool *force) -{ - if (rule->filter_id == EFX_ARFS_FILTER_ID_PENDING) { - /* ARFS is currently updating this entry, leave it */ - return false; - } - if (rule->filter_id == EFX_ARFS_FILTER_ID_ERROR) { - /* ARFS tried and failed to update this, so it's probably out - * of date. Remove the filter and the ARFS rule entry. - */ - rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; - *force = true; - return true; - } else if (WARN_ON(rule->filter_id != filter_idx)) { /* can't happen */ - /* ARFS has moved on, so old filter is not needed. Since we did - * not mark the rule with EFX_ARFS_FILTER_ID_REMOVING, it will - * not be removed by efx_rps_hash_del() subsequently. - */ - *force = true; - return true; - } - /* Remove it iff ARFS wants to. */ - return true; -} - -static -struct hlist_head *efx_rps_hash_bucket(struct efx_nic *efx, - const struct efx_filter_spec *spec) -{ - u32 hash = efx_filter_spec_hash(spec); - - lockdep_assert_held(&efx->rps_hash_lock); - if (!efx->rps_hash_table) - return NULL; - return &efx->rps_hash_table[hash % EFX_ARFS_HASH_TABLE_SIZE]; -} - -struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, - const struct efx_filter_spec *spec) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (!head) - return NULL; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) - return rule; - } - return NULL; -} - -struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, - const struct efx_filter_spec *spec, - bool *new) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (!head) - return NULL; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) { - *new = false; - return rule; - } - } - rule = kmalloc(sizeof(*rule), GFP_ATOMIC); - *new = true; - if (rule) { - memcpy(&rule->spec, spec, sizeof(rule->spec)); - hlist_add_head(&rule->node, head); - } - return rule; -} - -void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec) -{ - struct efx_arfs_rule *rule; - struct hlist_head *head; - struct hlist_node *node; - - head = efx_rps_hash_bucket(efx, spec); - if (WARN_ON(!head)) - return; - hlist_for_each(node, head) { - rule = container_of(node, struct efx_arfs_rule, node); - if (efx_filter_spec_equal(spec, &rule->spec)) { - /* Someone already reused the entry. We know that if - * this check doesn't fire (i.e. filter_id == REMOVING) - * then the REMOVING mark was put there by our caller, - * because caller is holding a lock on filter table and - * only holders of that lock set REMOVING. - */ - if (rule->filter_id != EFX_ARFS_FILTER_ID_REMOVING) - return; - hlist_del(node); - kfree(rule); - return; - } - } - /* We didn't find it. */ - WARN_ON(1); -} -#endif - -/* RSS contexts. We're using linked lists and crappy O(n) algorithms, because - * (a) this is an infrequent control-plane operation and (b) n is small (max 64) - */ -struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx) -{ - struct list_head *head = &efx->rss_context.list; - struct efx_rss_context *ctx, *new; - u32 id = 1; /* Don't use zero, that refers to the master RSS context */ - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - /* Search for first gap in the numbering */ - list_for_each_entry(ctx, head, list) { - if (ctx->user_id != id) - break; - id++; - /* Check for wrap. If this happens, we have nearly 2^32 - * allocated RSS contexts, which seems unlikely. - */ - if (WARN_ON_ONCE(!id)) - return NULL; - } - - /* Create the new entry */ - new = kmalloc(sizeof(struct efx_rss_context), GFP_KERNEL); - if (!new) - return NULL; - new->context_id = EFX_EF10_RSS_CONTEXT_INVALID; - new->rx_hash_udp_4tuple = false; - - /* Insert the new entry into the gap */ - new->user_id = id; - list_add_tail(&new->list, &ctx->list); - return new; -} - -struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id) -{ - struct list_head *head = &efx->rss_context.list; - struct efx_rss_context *ctx; - - WARN_ON(!mutex_is_locked(&efx->rss_lock)); - - list_for_each_entry(ctx, head, list) - if (ctx->user_id == id) - return ctx; - return NULL; -} - -void efx_free_rss_context_entry(struct efx_rss_context *ctx) -{ - list_del(&ctx->list); - kfree(ctx); -} - /************************************************************************** * * PCI interface @@ -3519,7 +1059,7 @@ static void efx_pci_remove_main(struct efx_nic *efx) * are not READY. */ BUG_ON(efx->state == STATE_READY); - cancel_work_sync(&efx->reset_work); + efx_flush_reset_workqueue(efx); efx_disable_interrupts(efx); efx_clear_interrupt_affinity(efx); @@ -3559,7 +1099,7 @@ static void efx_pci_remove(struct pci_dev *pci_dev) efx_pci_remove_main(efx); - efx_fini_io(efx); + efx_fini_io(efx, efx->type->mem_bar(efx)); netif_dbg(efx, drv, efx->net_dev, "shutdown successful\n"); efx_fini_struct(efx); @@ -3782,7 +1322,8 @@ static int efx_pci_probe(struct pci_dev *pci_dev, efx_probe_vpd_strings(efx); /* Set up basic I/O (BAR mappings etc) */ - rc = efx_init_io(efx); + rc = efx_init_io(efx, efx->type->mem_bar(efx), efx->type->max_dma_mask, + efx->type->mem_map_size(efx)); if (rc) goto fail2; @@ -3826,7 +1367,7 @@ static int efx_pci_probe(struct pci_dev *pci_dev, return 0; fail3: - efx_fini_io(efx); + efx_fini_io(efx, efx->type->mem_bar(efx)); fail2: efx_fini_struct(efx); fail1: @@ -3904,7 +1445,7 @@ static int efx_pm_thaw(struct device *dev) rtnl_unlock(); /* Reschedule any quenched resets scheduled during efx_pm_freeze() */ - queue_work(reset_workqueue, &efx->reset_work); + efx_queue_reset_work(efx); return 0; @@ -4083,10 +1624,6 @@ static struct pci_driver efx_pci_driver = { * *************************************************************************/ -module_param(interrupt_mode, uint, 0444); -MODULE_PARM_DESC(interrupt_mode, - "Interrupt mode (0=>MSIX 1=>MSI 2=>legacy)"); - static int __init efx_init_module(void) { int rc; @@ -4103,11 +1640,9 @@ static int __init efx_init_module(void) goto err_sriov; #endif - reset_workqueue = create_singlethread_workqueue("sfc_reset"); - if (!reset_workqueue) { - rc = -ENOMEM; + rc = efx_create_reset_workqueue(); + if (rc) goto err_reset; - } rc = pci_register_driver(&efx_pci_driver); if (rc < 0) @@ -4116,7 +1651,7 @@ static int __init efx_init_module(void) return 0; err_pci: - destroy_workqueue(reset_workqueue); + efx_destroy_reset_workqueue(); err_reset: #ifdef CONFIG_SFC_SRIOV efx_fini_sriov(); @@ -4132,7 +1667,7 @@ static void __exit efx_exit_module(void) printk(KERN_INFO "Solarflare NET driver unloading\n"); pci_unregister_driver(&efx_pci_driver); - destroy_workqueue(reset_workqueue); + efx_destroy_reset_workqueue(); #ifdef CONFIG_SFC_SRIOV efx_fini_sriov(); #endif diff --git a/drivers/net/ethernet/sfc/efx.h b/drivers/net/ethernet/sfc/efx.h index 2dd8d5002315..f1bdb04efbe4 100644 --- a/drivers/net/ethernet/sfc/efx.h +++ b/drivers/net/ethernet/sfc/efx.h @@ -15,31 +15,17 @@ int efx_net_open(struct net_device *net_dev); int efx_net_stop(struct net_device *net_dev); /* TX */ -int efx_probe_tx_queue(struct efx_tx_queue *tx_queue); -void efx_remove_tx_queue(struct efx_tx_queue *tx_queue); -void efx_init_tx_queue(struct efx_tx_queue *tx_queue); void efx_init_tx_queue_core_txq(struct efx_tx_queue *tx_queue); -void efx_fini_tx_queue(struct efx_tx_queue *tx_queue); netdev_tx_t efx_hard_start_xmit(struct sk_buff *skb, struct net_device *net_dev); netdev_tx_t efx_enqueue_skb(struct efx_tx_queue *tx_queue, struct sk_buff *skb); void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index); int efx_setup_tc(struct net_device *net_dev, enum tc_setup_type type, void *type_data); -unsigned int efx_tx_max_skb_descs(struct efx_nic *efx); extern unsigned int efx_piobuf_size; extern bool efx_separate_tx_channels; /* RX */ -void efx_set_default_rx_indir_table(struct efx_nic *efx, - struct efx_rss_context *ctx); -void efx_rx_config_page_split(struct efx_nic *efx); -int efx_probe_rx_queue(struct efx_rx_queue *rx_queue); -void efx_remove_rx_queue(struct efx_rx_queue *rx_queue); -void efx_init_rx_queue(struct efx_rx_queue *rx_queue); -void efx_fini_rx_queue(struct efx_rx_queue *rx_queue); -void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic); -void efx_rx_slow_fill(struct timer_list *t); void __efx_rx_packet(struct efx_channel *channel); void efx_rx_packet(struct efx_rx_queue *rx_queue, unsigned int index, unsigned int n_frags, unsigned int len, u16 flags); @@ -48,7 +34,6 @@ static inline void efx_rx_flush_packet(struct efx_channel *channel) if (channel->rx_pkt_n_frags) __efx_rx_packet(channel); } -void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue); #define EFX_MAX_DMAQ_SIZE 4096UL #define EFX_DEFAULT_DMAQ_SIZE 1024UL @@ -80,8 +65,6 @@ static inline bool efx_rss_enabled(struct efx_nic *efx) /* Filters */ -void efx_mac_reconfigure(struct efx_nic *efx); - /** * efx_filter_insert_filter - add or replace a filter * @efx: NIC in which to insert the filter @@ -186,58 +169,17 @@ static inline void efx_filter_rfs_expire(struct work_struct *data) static inline void efx_filter_rfs_expire(struct work_struct *data) {} #define efx_filter_rfs_enabled() 0 #endif -bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec); - -bool efx_filter_spec_equal(const struct efx_filter_spec *left, - const struct efx_filter_spec *right); -u32 efx_filter_spec_hash(const struct efx_filter_spec *spec); - -#ifdef CONFIG_RFS_ACCEL -bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, - bool *force); - -struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, - const struct efx_filter_spec *spec); - -/* @new is written to indicate if entry was newly added (true) or if an old - * entry was found and returned (false). - */ -struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, - const struct efx_filter_spec *spec, - bool *new); - -void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec); -#endif /* RSS contexts */ -struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx); -struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id); -void efx_free_rss_context_entry(struct efx_rss_context *ctx); static inline bool efx_rss_active(struct efx_rss_context *ctx) { - return ctx->context_id != EFX_EF10_RSS_CONTEXT_INVALID; + return ctx->context_id != EFX_MCDI_RSS_CONTEXT_INVALID; } -/* Channels */ -int efx_channel_dummy_op_int(struct efx_channel *channel); -void efx_channel_dummy_op_void(struct efx_channel *channel); -int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries); - -/* Ports */ -int efx_reconfigure_port(struct efx_nic *efx); -int __efx_reconfigure_port(struct efx_nic *efx); - /* Ethtool support */ extern const struct ethtool_ops efx_ethtool_ops; -/* Reset handling */ -int efx_reset(struct efx_nic *efx, enum reset_type method); -void efx_reset_down(struct efx_nic *efx, enum reset_type method); -int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok); -int efx_try_recovery(struct efx_nic *efx); - /* Global */ -void efx_schedule_reset(struct efx_nic *efx, enum reset_type type); unsigned int efx_usecs_to_ticks(struct efx_nic *efx, unsigned int usecs); unsigned int efx_ticks_to_usecs(struct efx_nic *efx, unsigned int ticks); int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, @@ -245,8 +187,6 @@ int efx_init_irq_moderation(struct efx_nic *efx, unsigned int tx_usecs, bool rx_may_override_tx); void efx_get_irq_moderation(struct efx_nic *efx, unsigned int *tx_usecs, unsigned int *rx_usecs, bool *rx_adaptive); -void efx_stop_eventq(struct efx_channel *channel); -void efx_start_eventq(struct efx_channel *channel); /* Dummy PHY ops for PHY drivers */ int efx_port_dummy_op_int(struct efx_nic *efx); @@ -293,9 +233,6 @@ static inline void efx_schedule_channel_irq(struct efx_channel *channel) efx_schedule_channel(channel); } -void efx_link_status_changed(struct efx_nic *efx); -void efx_link_set_advertising(struct efx_nic *efx, - const unsigned long *advertising); void efx_link_clear_advertising(struct efx_nic *efx); void efx_link_set_wanted_fc(struct efx_nic *efx, u8); diff --git a/drivers/net/ethernet/sfc/efx_channels.c b/drivers/net/ethernet/sfc/efx_channels.c new file mode 100644 index 000000000000..aeb5e8aa2f2a --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_channels.c @@ -0,0 +1,1234 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include "efx_channels.h" +#include "efx.h" +#include "efx_common.h" +#include "tx_common.h" +#include "rx_common.h" +#include "nic.h" +#include "sriov.h" + +/* This is the first interrupt mode to try out of: + * 0 => MSI-X + * 1 => MSI + * 2 => legacy + */ +static unsigned int interrupt_mode; +module_param(interrupt_mode, uint, 0444); +MODULE_PARM_DESC(interrupt_mode, + "Interrupt mode (0=>MSIX 1=>MSI 2=>legacy)"); + +/* This is the requested number of CPUs to use for Receive-Side Scaling (RSS), + * i.e. the number of CPUs among which we may distribute simultaneous + * interrupt handling. + * + * Cards without MSI-X will only target one CPU via legacy or MSI interrupt. + * The default (0) means to assign an interrupt to each core. + */ +static unsigned int rss_cpus; +module_param(rss_cpus, uint, 0444); +MODULE_PARM_DESC(rss_cpus, "Number of CPUs to use for Receive-Side Scaling"); + +static unsigned int irq_adapt_low_thresh = 8000; +module_param(irq_adapt_low_thresh, uint, 0644); +MODULE_PARM_DESC(irq_adapt_low_thresh, + "Threshold score for reducing IRQ moderation"); + +static unsigned int irq_adapt_high_thresh = 16000; +module_param(irq_adapt_high_thresh, uint, 0644); +MODULE_PARM_DESC(irq_adapt_high_thresh, + "Threshold score for increasing IRQ moderation"); + +/* This is the weight assigned to each of the (per-channel) virtual + * NAPI devices. + */ +static int napi_weight = 64; + +/*************** + * Housekeeping + ***************/ + +int efx_channel_dummy_op_int(struct efx_channel *channel) +{ + return 0; +} + +void efx_channel_dummy_op_void(struct efx_channel *channel) +{ +} + +static const struct efx_channel_type efx_default_channel_type = { + .pre_probe = efx_channel_dummy_op_int, + .post_remove = efx_channel_dummy_op_void, + .get_name = efx_get_channel_name, + .copy = efx_copy_channel, + .want_txqs = efx_default_channel_want_txqs, + .keep_eventq = false, + .want_pio = true, +}; + +/************* + * INTERRUPTS + *************/ + +static unsigned int efx_wanted_parallelism(struct efx_nic *efx) +{ + cpumask_var_t thread_mask; + unsigned int count; + int cpu; + + if (rss_cpus) { + count = rss_cpus; + } else { + if (unlikely(!zalloc_cpumask_var(&thread_mask, GFP_KERNEL))) { + netif_warn(efx, probe, efx->net_dev, + "RSS disabled due to allocation failure\n"); + return 1; + } + + count = 0; + for_each_online_cpu(cpu) { + if (!cpumask_test_cpu(cpu, thread_mask)) { + ++count; + cpumask_or(thread_mask, thread_mask, + topology_sibling_cpumask(cpu)); + } + } + + free_cpumask_var(thread_mask); + } + + if (count > EFX_MAX_RX_QUEUES) { + netif_cond_dbg(efx, probe, efx->net_dev, !rss_cpus, warn, + "Reducing number of rx queues from %u to %u.\n", + count, EFX_MAX_RX_QUEUES); + count = EFX_MAX_RX_QUEUES; + } + + /* If RSS is requested for the PF *and* VFs then we can't write RSS + * table entries that are inaccessible to VFs + */ +#ifdef CONFIG_SFC_SRIOV + if (efx->type->sriov_wanted) { + if (efx->type->sriov_wanted(efx) && efx_vf_size(efx) > 1 && + count > efx_vf_size(efx)) { + netif_warn(efx, probe, efx->net_dev, + "Reducing number of RSS channels from %u to %u for " + "VF support. Increase vf-msix-limit to use more " + "channels on the PF.\n", + count, efx_vf_size(efx)); + count = efx_vf_size(efx); + } + } +#endif + + return count; +} + +static int efx_allocate_msix_channels(struct efx_nic *efx, + unsigned int max_channels, + unsigned int extra_channels, + unsigned int parallelism) +{ + unsigned int n_channels = parallelism; + int vec_count; + int n_xdp_tx; + int n_xdp_ev; + + if (efx_separate_tx_channels) + n_channels *= 2; + n_channels += extra_channels; + + /* To allow XDP transmit to happen from arbitrary NAPI contexts + * we allocate a TX queue per CPU. We share event queues across + * multiple tx queues, assuming tx and ev queues are both + * maximum size. + */ + + n_xdp_tx = num_possible_cpus(); + n_xdp_ev = DIV_ROUND_UP(n_xdp_tx, EFX_TXQ_TYPES); + + vec_count = pci_msix_vec_count(efx->pci_dev); + if (vec_count < 0) + return vec_count; + + max_channels = min_t(unsigned int, vec_count, max_channels); + + /* Check resources. + * We need a channel per event queue, plus a VI per tx queue. + * This may be more pessimistic than it needs to be. + */ + if (n_channels + n_xdp_ev > max_channels) { + netif_err(efx, drv, efx->net_dev, + "Insufficient resources for %d XDP event queues (%d other channels, max %d)\n", + n_xdp_ev, n_channels, max_channels); + efx->n_xdp_channels = 0; + efx->xdp_tx_per_channel = 0; + efx->xdp_tx_queue_count = 0; + } else { + efx->n_xdp_channels = n_xdp_ev; + efx->xdp_tx_per_channel = EFX_TXQ_TYPES; + efx->xdp_tx_queue_count = n_xdp_tx; + n_channels += n_xdp_ev; + netif_dbg(efx, drv, efx->net_dev, + "Allocating %d TX and %d event queues for XDP\n", + n_xdp_tx, n_xdp_ev); + } + + if (vec_count < n_channels) { + netif_err(efx, drv, efx->net_dev, + "WARNING: Insufficient MSI-X vectors available (%d < %u).\n", + vec_count, n_channels); + netif_err(efx, drv, efx->net_dev, + "WARNING: Performance may be reduced.\n"); + n_channels = vec_count; + } + + n_channels = min(n_channels, max_channels); + + efx->n_channels = n_channels; + + /* Ignore XDP tx channels when creating rx channels. */ + n_channels -= efx->n_xdp_channels; + + if (efx_separate_tx_channels) { + efx->n_tx_channels = + min(max(n_channels / 2, 1U), + efx->max_tx_channels); + efx->tx_channel_offset = + n_channels - efx->n_tx_channels; + efx->n_rx_channels = + max(n_channels - + efx->n_tx_channels, 1U); + } else { + efx->n_tx_channels = min(n_channels, efx->max_tx_channels); + efx->tx_channel_offset = 0; + efx->n_rx_channels = n_channels; + } + + efx->n_rx_channels = min(efx->n_rx_channels, parallelism); + efx->n_tx_channels = min(efx->n_tx_channels, parallelism); + + efx->xdp_channel_offset = n_channels; + + netif_dbg(efx, drv, efx->net_dev, + "Allocating %u RX channels\n", + efx->n_rx_channels); + + return efx->n_channels; +} + +/* Probe the number and type of interrupts we are able to obtain, and + * the resulting numbers of channels and RX queues. + */ +int efx_probe_interrupts(struct efx_nic *efx) +{ + unsigned int extra_channels = 0; + unsigned int rss_spread; + unsigned int i, j; + int rc; + + for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) + if (efx->extra_channel_type[i]) + ++extra_channels; + + if (efx->interrupt_mode == EFX_INT_MODE_MSIX) { + unsigned int parallelism = efx_wanted_parallelism(efx); + struct msix_entry xentries[EFX_MAX_CHANNELS]; + unsigned int n_channels; + + rc = efx_allocate_msix_channels(efx, efx->max_channels, + extra_channels, parallelism); + if (rc >= 0) { + n_channels = rc; + for (i = 0; i < n_channels; i++) + xentries[i].entry = i; + rc = pci_enable_msix_range(efx->pci_dev, xentries, 1, + n_channels); + } + if (rc < 0) { + /* Fall back to single channel MSI */ + netif_err(efx, drv, efx->net_dev, + "could not enable MSI-X\n"); + if (efx->type->min_interrupt_mode >= EFX_INT_MODE_MSI) + efx->interrupt_mode = EFX_INT_MODE_MSI; + else + return rc; + } else if (rc < n_channels) { + netif_err(efx, drv, efx->net_dev, + "WARNING: Insufficient MSI-X vectors" + " available (%d < %u).\n", rc, n_channels); + netif_err(efx, drv, efx->net_dev, + "WARNING: Performance may be reduced.\n"); + n_channels = rc; + } + + if (rc > 0) { + for (i = 0; i < efx->n_channels; i++) + efx_get_channel(efx, i)->irq = + xentries[i].vector; + } + } + + /* Try single interrupt MSI */ + if (efx->interrupt_mode == EFX_INT_MODE_MSI) { + efx->n_channels = 1; + efx->n_rx_channels = 1; + efx->n_tx_channels = 1; + efx->n_xdp_channels = 0; + efx->xdp_channel_offset = efx->n_channels; + rc = pci_enable_msi(efx->pci_dev); + if (rc == 0) { + efx_get_channel(efx, 0)->irq = efx->pci_dev->irq; + } else { + netif_err(efx, drv, efx->net_dev, + "could not enable MSI\n"); + if (efx->type->min_interrupt_mode >= EFX_INT_MODE_LEGACY) + efx->interrupt_mode = EFX_INT_MODE_LEGACY; + else + return rc; + } + } + + /* Assume legacy interrupts */ + if (efx->interrupt_mode == EFX_INT_MODE_LEGACY) { + efx->n_channels = 1 + (efx_separate_tx_channels ? 1 : 0); + efx->n_rx_channels = 1; + efx->n_tx_channels = 1; + efx->n_xdp_channels = 0; + efx->xdp_channel_offset = efx->n_channels; + efx->legacy_irq = efx->pci_dev->irq; + } + + /* Assign extra channels if possible, before XDP channels */ + efx->n_extra_tx_channels = 0; + j = efx->xdp_channel_offset; + for (i = 0; i < EFX_MAX_EXTRA_CHANNELS; i++) { + if (!efx->extra_channel_type[i]) + continue; + if (j <= efx->tx_channel_offset + efx->n_tx_channels) { + efx->extra_channel_type[i]->handle_no_channel(efx); + } else { + --j; + efx_get_channel(efx, j)->type = + efx->extra_channel_type[i]; + if (efx_channel_has_tx_queues(efx_get_channel(efx, j))) + efx->n_extra_tx_channels++; + } + } + + rss_spread = efx->n_rx_channels; + /* RSS might be usable on VFs even if it is disabled on the PF */ +#ifdef CONFIG_SFC_SRIOV + if (efx->type->sriov_wanted) { + efx->rss_spread = ((rss_spread > 1 || + !efx->type->sriov_wanted(efx)) ? + rss_spread : efx_vf_size(efx)); + return 0; + } +#endif + efx->rss_spread = rss_spread; + + return 0; +} + +#if defined(CONFIG_SMP) +void efx_set_interrupt_affinity(struct efx_nic *efx) +{ + struct efx_channel *channel; + unsigned int cpu; + + efx_for_each_channel(channel, efx) { + cpu = cpumask_local_spread(channel->channel, + pcibus_to_node(efx->pci_dev->bus)); + irq_set_affinity_hint(channel->irq, cpumask_of(cpu)); + } +} + +void efx_clear_interrupt_affinity(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + irq_set_affinity_hint(channel->irq, NULL); +} +#else +void +efx_set_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) +{ +} + +void +efx_clear_interrupt_affinity(struct efx_nic *efx __attribute__ ((unused))) +{ +} +#endif /* CONFIG_SMP */ + +void efx_remove_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + /* Remove MSI/MSI-X interrupts */ + efx_for_each_channel(channel, efx) + channel->irq = 0; + pci_disable_msi(efx->pci_dev); + pci_disable_msix(efx->pci_dev); + + /* Remove legacy interrupt */ + efx->legacy_irq = 0; +} + +/*************** + * EVENT QUEUES + ***************/ + +/* Create event queue + * Event queue memory allocations are done only once. If the channel + * is reset, the memory buffer will be reused; this guards against + * errors during channel reset and also simplifies interrupt handling. + */ +int efx_probe_eventq(struct efx_channel *channel) +{ + struct efx_nic *efx = channel->efx; + unsigned long entries; + + netif_dbg(efx, probe, efx->net_dev, + "chan %d create event queue\n", channel->channel); + + /* Build an event queue with room for one event per tx and rx buffer, + * plus some extra for link state events and MCDI completions. + */ + entries = roundup_pow_of_two(efx->rxq_entries + efx->txq_entries + 128); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_EVQ_SIZE); + channel->eventq_mask = max(entries, EFX_MIN_EVQ_SIZE) - 1; + + return efx_nic_probe_eventq(channel); +} + +/* Prepare channel's event queue */ +int efx_init_eventq(struct efx_channel *channel) +{ + struct efx_nic *efx = channel->efx; + int rc; + + EFX_WARN_ON_PARANOID(channel->eventq_init); + + netif_dbg(efx, drv, efx->net_dev, + "chan %d init event queue\n", channel->channel); + + rc = efx_nic_init_eventq(channel); + if (rc == 0) { + efx->type->push_irq_moderation(channel); + channel->eventq_read_ptr = 0; + channel->eventq_init = true; + } + return rc; +} + +/* Enable event queue processing and NAPI */ +void efx_start_eventq(struct efx_channel *channel) +{ + netif_dbg(channel->efx, ifup, channel->efx->net_dev, + "chan %d start event queue\n", channel->channel); + + /* Make sure the NAPI handler sees the enabled flag set */ + channel->enabled = true; + smp_wmb(); + + napi_enable(&channel->napi_str); + efx_nic_eventq_read_ack(channel); +} + +/* Disable event queue processing and NAPI */ +void efx_stop_eventq(struct efx_channel *channel) +{ + if (!channel->enabled) + return; + + napi_disable(&channel->napi_str); + channel->enabled = false; +} + +void efx_fini_eventq(struct efx_channel *channel) +{ + if (!channel->eventq_init) + return; + + netif_dbg(channel->efx, drv, channel->efx->net_dev, + "chan %d fini event queue\n", channel->channel); + + efx_nic_fini_eventq(channel); + channel->eventq_init = false; +} + +void efx_remove_eventq(struct efx_channel *channel) +{ + netif_dbg(channel->efx, drv, channel->efx->net_dev, + "chan %d remove event queue\n", channel->channel); + + efx_nic_remove_eventq(channel); +} + +/************************************************************************** + * + * Channel handling + * + *************************************************************************/ + +/* Allocate and initialise a channel structure. */ +struct efx_channel * +efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel) +{ + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + struct efx_channel *channel; + int j; + + channel = kzalloc(sizeof(*channel), GFP_KERNEL); + if (!channel) + return NULL; + + channel->efx = efx; + channel->channel = i; + channel->type = &efx_default_channel_type; + + for (j = 0; j < EFX_TXQ_TYPES; j++) { + tx_queue = &channel->tx_queue[j]; + tx_queue->efx = efx; + tx_queue->queue = i * EFX_TXQ_TYPES + j; + tx_queue->channel = channel; + } + +#ifdef CONFIG_RFS_ACCEL + INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); +#endif + + rx_queue = &channel->rx_queue; + rx_queue->efx = efx; + timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); + + return channel; +} + +int efx_init_channels(struct efx_nic *efx) +{ + unsigned int i; + + for (i = 0; i < EFX_MAX_CHANNELS; i++) { + efx->channel[i] = efx_alloc_channel(efx, i, NULL); + if (!efx->channel[i]) + return -ENOMEM; + efx->msi_context[i].efx = efx; + efx->msi_context[i].index = i; + } + + /* Higher numbered interrupt modes are less capable! */ + if (WARN_ON_ONCE(efx->type->max_interrupt_mode > + efx->type->min_interrupt_mode)) { + return -EIO; + } + efx->interrupt_mode = max(efx->type->max_interrupt_mode, + interrupt_mode); + efx->interrupt_mode = min(efx->type->min_interrupt_mode, + interrupt_mode); + + return 0; +} + +void efx_fini_channels(struct efx_nic *efx) +{ + unsigned int i; + + for (i = 0; i < EFX_MAX_CHANNELS; i++) + if (efx->channel[i]) { + kfree(efx->channel[i]); + efx->channel[i] = NULL; + } +} + +/* Allocate and initialise a channel structure, copying parameters + * (but not resources) from an old channel structure. + */ +struct efx_channel *efx_copy_channel(const struct efx_channel *old_channel) +{ + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + struct efx_channel *channel; + int j; + + channel = kmalloc(sizeof(*channel), GFP_KERNEL); + if (!channel) + return NULL; + + *channel = *old_channel; + + channel->napi_dev = NULL; + INIT_HLIST_NODE(&channel->napi_str.napi_hash_node); + channel->napi_str.napi_id = 0; + channel->napi_str.state = 0; + memset(&channel->eventq, 0, sizeof(channel->eventq)); + + for (j = 0; j < EFX_TXQ_TYPES; j++) { + tx_queue = &channel->tx_queue[j]; + if (tx_queue->channel) + tx_queue->channel = channel; + tx_queue->buffer = NULL; + memset(&tx_queue->txd, 0, sizeof(tx_queue->txd)); + } + + rx_queue = &channel->rx_queue; + rx_queue->buffer = NULL; + memset(&rx_queue->rxd, 0, sizeof(rx_queue->rxd)); + timer_setup(&rx_queue->slow_fill, efx_rx_slow_fill, 0); +#ifdef CONFIG_RFS_ACCEL + INIT_DELAYED_WORK(&channel->filter_work, efx_filter_rfs_expire); +#endif + + return channel; +} + +static int efx_probe_channel(struct efx_channel *channel) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + int rc; + + netif_dbg(channel->efx, probe, channel->efx->net_dev, + "creating channel %d\n", channel->channel); + + rc = channel->type->pre_probe(channel); + if (rc) + goto fail; + + rc = efx_probe_eventq(channel); + if (rc) + goto fail; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + rc = efx_probe_tx_queue(tx_queue); + if (rc) + goto fail; + } + + efx_for_each_channel_rx_queue(rx_queue, channel) { + rc = efx_probe_rx_queue(rx_queue); + if (rc) + goto fail; + } + + channel->rx_list = NULL; + + return 0; + +fail: + efx_remove_channel(channel); + return rc; +} + +void efx_get_channel_name(struct efx_channel *channel, char *buf, size_t len) +{ + struct efx_nic *efx = channel->efx; + const char *type; + int number; + + number = channel->channel; + + if (number >= efx->xdp_channel_offset && + !WARN_ON_ONCE(!efx->n_xdp_channels)) { + type = "-xdp"; + number -= efx->xdp_channel_offset; + } else if (efx->tx_channel_offset == 0) { + type = ""; + } else if (number < efx->tx_channel_offset) { + type = "-rx"; + } else { + type = "-tx"; + number -= efx->tx_channel_offset; + } + snprintf(buf, len, "%s%s-%d", efx->name, type, number); +} + +void efx_set_channel_names(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + channel->type->get_name(channel, + efx->msi_context[channel->channel].name, + sizeof(efx->msi_context[0].name)); +} + +int efx_probe_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + int rc; + + /* Restart special buffer allocation */ + efx->next_buffer_table = 0; + + /* Probe channels in reverse, so that any 'extra' channels + * use the start of the buffer table. This allows the traffic + * channels to be resized without moving them or wasting the + * entries before them. + */ + efx_for_each_channel_rev(channel, efx) { + rc = efx_probe_channel(channel); + if (rc) { + netif_err(efx, probe, efx->net_dev, + "failed to create channel %d\n", + channel->channel); + goto fail; + } + } + efx_set_channel_names(efx); + + return 0; + +fail: + efx_remove_channels(efx); + return rc; +} + +void efx_remove_channel(struct efx_channel *channel) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + + netif_dbg(channel->efx, drv, channel->efx->net_dev, + "destroy chan %d\n", channel->channel); + + efx_for_each_channel_rx_queue(rx_queue, channel) + efx_remove_rx_queue(rx_queue); + efx_for_each_possible_channel_tx_queue(tx_queue, channel) + efx_remove_tx_queue(tx_queue); + efx_remove_eventq(channel); + channel->type->post_remove(channel); +} + +void efx_remove_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_remove_channel(channel); + + kfree(efx->xdp_tx_queues); +} + +int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries) +{ + struct efx_channel *other_channel[EFX_MAX_CHANNELS], *channel; + unsigned int i, next_buffer_table = 0; + u32 old_rxq_entries, old_txq_entries; + int rc, rc2; + + rc = efx_check_disabled(efx); + if (rc) + return rc; + + /* Not all channels should be reallocated. We must avoid + * reallocating their buffer table entries. + */ + efx_for_each_channel(channel, efx) { + struct efx_rx_queue *rx_queue; + struct efx_tx_queue *tx_queue; + + if (channel->type->copy) + continue; + next_buffer_table = max(next_buffer_table, + channel->eventq.index + + channel->eventq.entries); + efx_for_each_channel_rx_queue(rx_queue, channel) + next_buffer_table = max(next_buffer_table, + rx_queue->rxd.index + + rx_queue->rxd.entries); + efx_for_each_channel_tx_queue(tx_queue, channel) + next_buffer_table = max(next_buffer_table, + tx_queue->txd.index + + tx_queue->txd.entries); + } + + efx_device_detach_sync(efx); + efx_stop_all(efx); + efx_soft_disable_interrupts(efx); + + /* Clone channels (where possible) */ + memset(other_channel, 0, sizeof(other_channel)); + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + if (channel->type->copy) + channel = channel->type->copy(channel); + if (!channel) { + rc = -ENOMEM; + goto out; + } + other_channel[i] = channel; + } + + /* Swap entry counts and channel pointers */ + old_rxq_entries = efx->rxq_entries; + old_txq_entries = efx->txq_entries; + efx->rxq_entries = rxq_entries; + efx->txq_entries = txq_entries; + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + efx->channel[i] = other_channel[i]; + other_channel[i] = channel; + } + + /* Restart buffer table allocation */ + efx->next_buffer_table = next_buffer_table; + + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + if (!channel->type->copy) + continue; + rc = efx_probe_channel(channel); + if (rc) + goto rollback; + efx_init_napi_channel(efx->channel[i]); + } + +out: + /* Destroy unused channel structures */ + for (i = 0; i < efx->n_channels; i++) { + channel = other_channel[i]; + if (channel && channel->type->copy) { + efx_fini_napi_channel(channel); + efx_remove_channel(channel); + kfree(channel); + } + } + + rc2 = efx_soft_enable_interrupts(efx); + if (rc2) { + rc = rc ? rc : rc2; + netif_err(efx, drv, efx->net_dev, + "unable to restart interrupts on channel reallocation\n"); + efx_schedule_reset(efx, RESET_TYPE_DISABLE); + } else { + efx_start_all(efx); + efx_device_attach_if_not_resetting(efx); + } + return rc; + +rollback: + /* Swap back */ + efx->rxq_entries = old_rxq_entries; + efx->txq_entries = old_txq_entries; + for (i = 0; i < efx->n_channels; i++) { + channel = efx->channel[i]; + efx->channel[i] = other_channel[i]; + other_channel[i] = channel; + } + goto out; +} + +int efx_set_channels(struct efx_nic *efx) +{ + struct efx_channel *channel; + struct efx_tx_queue *tx_queue; + int xdp_queue_number; + + efx->tx_channel_offset = + efx_separate_tx_channels ? + efx->n_channels - efx->n_tx_channels : 0; + + if (efx->xdp_tx_queue_count) { + EFX_WARN_ON_PARANOID(efx->xdp_tx_queues); + + /* Allocate array for XDP TX queue lookup. */ + efx->xdp_tx_queues = kcalloc(efx->xdp_tx_queue_count, + sizeof(*efx->xdp_tx_queues), + GFP_KERNEL); + if (!efx->xdp_tx_queues) + return -ENOMEM; + } + + /* We need to mark which channels really have RX and TX + * queues, and adjust the TX queue numbers if we have separate + * RX-only and TX-only channels. + */ + xdp_queue_number = 0; + efx_for_each_channel(channel, efx) { + if (channel->channel < efx->n_rx_channels) + channel->rx_queue.core_index = channel->channel; + else + channel->rx_queue.core_index = -1; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + tx_queue->queue -= (efx->tx_channel_offset * + EFX_TXQ_TYPES); + + if (efx_channel_is_xdp_tx(channel) && + xdp_queue_number < efx->xdp_tx_queue_count) { + efx->xdp_tx_queues[xdp_queue_number] = tx_queue; + xdp_queue_number++; + } + } + } + return 0; +} + +bool efx_default_channel_want_txqs(struct efx_channel *channel) +{ + return channel->channel - channel->efx->tx_channel_offset < + channel->efx->n_tx_channels; +} + +/************* + * START/STOP + *************/ + +int efx_soft_enable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel, *end_channel; + int rc; + + BUG_ON(efx->state == STATE_DISABLED); + + efx->irq_soft_enabled = true; + smp_wmb(); + + efx_for_each_channel(channel, efx) { + if (!channel->type->keep_eventq) { + rc = efx_init_eventq(channel); + if (rc) + goto fail; + } + efx_start_eventq(channel); + } + + efx_mcdi_mode_event(efx); + + return 0; +fail: + end_channel = channel; + efx_for_each_channel(channel, efx) { + if (channel == end_channel) + break; + efx_stop_eventq(channel); + if (!channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + return rc; +} + +void efx_soft_disable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + if (efx->state == STATE_DISABLED) + return; + + efx_mcdi_mode_poll(efx); + + efx->irq_soft_enabled = false; + smp_wmb(); + + if (efx->legacy_irq) + synchronize_irq(efx->legacy_irq); + + efx_for_each_channel(channel, efx) { + if (channel->irq) + synchronize_irq(channel->irq); + + efx_stop_eventq(channel); + if (!channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + /* Flush the asynchronous MCDI request queue */ + efx_mcdi_flush_async(efx); +} + +int efx_enable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel, *end_channel; + int rc; + + /* TODO: Is this really a bug? */ + BUG_ON(efx->state == STATE_DISABLED); + + if (efx->eeh_disabled_legacy_irq) { + enable_irq(efx->legacy_irq); + efx->eeh_disabled_legacy_irq = false; + } + + efx->type->irq_enable_master(efx); + + efx_for_each_channel(channel, efx) { + if (channel->type->keep_eventq) { + rc = efx_init_eventq(channel); + if (rc) + goto fail; + } + } + + rc = efx_soft_enable_interrupts(efx); + if (rc) + goto fail; + + return 0; + +fail: + end_channel = channel; + efx_for_each_channel(channel, efx) { + if (channel == end_channel) + break; + if (channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + efx->type->irq_disable_non_ev(efx); + + return rc; +} + +void efx_disable_interrupts(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_soft_disable_interrupts(efx); + + efx_for_each_channel(channel, efx) { + if (channel->type->keep_eventq) + efx_fini_eventq(channel); + } + + efx->type->irq_disable_non_ev(efx); +} + +void efx_start_channels(struct efx_nic *efx) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + efx_for_each_channel_tx_queue(tx_queue, channel) { + efx_init_tx_queue(tx_queue); + atomic_inc(&efx->active_queues); + } + + efx_for_each_channel_rx_queue(rx_queue, channel) { + efx_init_rx_queue(rx_queue); + atomic_inc(&efx->active_queues); + efx_stop_eventq(channel); + efx_fast_push_rx_descriptors(rx_queue, false); + efx_start_eventq(channel); + } + + WARN_ON(channel->rx_pkt_n_frags); + } +} + +void efx_stop_channels(struct efx_nic *efx) +{ + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + struct efx_channel *channel; + int rc = 0; + + /* Stop RX refill */ + efx_for_each_channel(channel, efx) { + efx_for_each_channel_rx_queue(rx_queue, channel) + rx_queue->refill_enabled = false; + } + + efx_for_each_channel(channel, efx) { + /* RX packet processing is pipelined, so wait for the + * NAPI handler to complete. At least event queue 0 + * might be kept active by non-data events, so don't + * use napi_synchronize() but actually disable NAPI + * temporarily. + */ + if (efx_channel_has_rx_queue(channel)) { + efx_stop_eventq(channel); + efx_start_eventq(channel); + } + } + + if (efx->type->fini_dmaq) + rc = efx->type->fini_dmaq(efx); + + if (rc) { + netif_err(efx, drv, efx->net_dev, "failed to flush queues\n"); + } else { + netif_dbg(efx, drv, efx->net_dev, + "successfully flushed all queues\n"); + } + + efx_for_each_channel(channel, efx) { + efx_for_each_channel_rx_queue(rx_queue, channel) + efx_fini_rx_queue(rx_queue); + efx_for_each_possible_channel_tx_queue(tx_queue, channel) + efx_fini_tx_queue(tx_queue); + } +} + +/************************************************************************** + * + * NAPI interface + * + *************************************************************************/ + +/* Process channel's event queue + * + * This function is responsible for processing the event queue of a + * single channel. The caller must guarantee that this function will + * never be concurrently called more than once on the same channel, + * though different channels may be being processed concurrently. + */ +static int efx_process_channel(struct efx_channel *channel, int budget) +{ + struct efx_tx_queue *tx_queue; + struct list_head rx_list; + int spent; + + if (unlikely(!channel->enabled)) + return 0; + + /* Prepare the batch receive list */ + EFX_WARN_ON_PARANOID(channel->rx_list != NULL); + INIT_LIST_HEAD(&rx_list); + channel->rx_list = &rx_list; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + tx_queue->pkts_compl = 0; + tx_queue->bytes_compl = 0; + } + + spent = efx_nic_process_eventq(channel, budget); + if (spent && efx_channel_has_rx_queue(channel)) { + struct efx_rx_queue *rx_queue = + efx_channel_get_rx_queue(channel); + + efx_rx_flush_packet(channel); + efx_fast_push_rx_descriptors(rx_queue, true); + } + + /* Update BQL */ + efx_for_each_channel_tx_queue(tx_queue, channel) { + if (tx_queue->bytes_compl) { + netdev_tx_completed_queue(tx_queue->core_txq, + tx_queue->pkts_compl, + tx_queue->bytes_compl); + } + } + + /* Receive any packets we queued up */ + netif_receive_skb_list(channel->rx_list); + channel->rx_list = NULL; + + return spent; +} + +static void efx_update_irq_mod(struct efx_nic *efx, struct efx_channel *channel) +{ + int step = efx->irq_mod_step_us; + + if (channel->irq_mod_score < irq_adapt_low_thresh) { + if (channel->irq_moderation_us > step) { + channel->irq_moderation_us -= step; + efx->type->push_irq_moderation(channel); + } + } else if (channel->irq_mod_score > irq_adapt_high_thresh) { + if (channel->irq_moderation_us < + efx->irq_rx_moderation_us) { + channel->irq_moderation_us += step; + efx->type->push_irq_moderation(channel); + } + } + + channel->irq_count = 0; + channel->irq_mod_score = 0; +} + +/* NAPI poll handler + * + * NAPI guarantees serialisation of polls of the same device, which + * provides the guarantee required by efx_process_channel(). + */ +static int efx_poll(struct napi_struct *napi, int budget) +{ + struct efx_channel *channel = + container_of(napi, struct efx_channel, napi_str); + struct efx_nic *efx = channel->efx; + int spent; + + netif_vdbg(efx, intr, efx->net_dev, + "channel %d NAPI poll executing on CPU %d\n", + channel->channel, raw_smp_processor_id()); + + spent = efx_process_channel(channel, budget); + + xdp_do_flush_map(); + + if (spent < budget) { + if (efx_channel_has_rx_queue(channel) && + efx->irq_rx_adaptive && + unlikely(++channel->irq_count == 1000)) { + efx_update_irq_mod(efx, channel); + } + +#ifdef CONFIG_RFS_ACCEL + /* Perhaps expire some ARFS filters */ + mod_delayed_work(system_wq, &channel->filter_work, 0); +#endif + + /* There is no race here; although napi_disable() will + * only wait for napi_complete(), this isn't a problem + * since efx_nic_eventq_read_ack() will have no effect if + * interrupts have already been disabled. + */ + if (napi_complete_done(napi, spent)) + efx_nic_eventq_read_ack(channel); + } + + return spent; +} + +void efx_init_napi_channel(struct efx_channel *channel) +{ + struct efx_nic *efx = channel->efx; + + channel->napi_dev = efx->net_dev; + netif_napi_add(channel->napi_dev, &channel->napi_str, + efx_poll, napi_weight); +} + +void efx_init_napi(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_init_napi_channel(channel); +} + +void efx_fini_napi_channel(struct efx_channel *channel) +{ + if (channel->napi_dev) + netif_napi_del(&channel->napi_str); + + channel->napi_dev = NULL; +} + +void efx_fini_napi(struct efx_nic *efx) +{ + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) + efx_fini_napi_channel(channel); +} diff --git a/drivers/net/ethernet/sfc/efx_channels.h b/drivers/net/ethernet/sfc/efx_channels.h new file mode 100644 index 000000000000..8d7b8c4142d7 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_channels.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_CHANNELS_H +#define EFX_CHANNELS_H + +int efx_probe_interrupts(struct efx_nic *efx); +void efx_remove_interrupts(struct efx_nic *efx); +int efx_soft_enable_interrupts(struct efx_nic *efx); +void efx_soft_disable_interrupts(struct efx_nic *efx); +int efx_enable_interrupts(struct efx_nic *efx); +void efx_disable_interrupts(struct efx_nic *efx); + +void efx_set_interrupt_affinity(struct efx_nic *efx); +void efx_clear_interrupt_affinity(struct efx_nic *efx); + +int efx_probe_eventq(struct efx_channel *channel); +int efx_init_eventq(struct efx_channel *channel); +void efx_start_eventq(struct efx_channel *channel); +void efx_stop_eventq(struct efx_channel *channel); +void efx_fini_eventq(struct efx_channel *channel); +void efx_remove_eventq(struct efx_channel *channel); + +struct efx_channel * +efx_alloc_channel(struct efx_nic *efx, int i, struct efx_channel *old_channel); +int efx_realloc_channels(struct efx_nic *efx, u32 rxq_entries, u32 txq_entries); +void efx_get_channel_name(struct efx_channel *channel, char *buf, size_t len); +void efx_set_channel_names(struct efx_nic *efx); +int efx_init_channels(struct efx_nic *efx); +int efx_probe_channels(struct efx_nic *efx); +int efx_set_channels(struct efx_nic *efx); +bool efx_default_channel_want_txqs(struct efx_channel *channel); +void efx_remove_channel(struct efx_channel *channel); +void efx_remove_channels(struct efx_nic *efx); +void efx_fini_channels(struct efx_nic *efx); +struct efx_channel *efx_copy_channel(const struct efx_channel *old_channel); +void efx_start_channels(struct efx_nic *efx); +void efx_stop_channels(struct efx_nic *efx); + +void efx_init_napi_channel(struct efx_channel *channel); +void efx_init_napi(struct efx_nic *efx); +void efx_fini_napi_channel(struct efx_channel *channel); +void efx_fini_napi(struct efx_nic *efx); + +int efx_channel_dummy_op_int(struct efx_channel *channel); +void efx_channel_dummy_op_void(struct efx_channel *channel); + +#endif diff --git a/drivers/net/ethernet/sfc/efx_common.c b/drivers/net/ethernet/sfc/efx_common.c new file mode 100644 index 000000000000..b0d76bc19673 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_common.c @@ -0,0 +1,1102 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include <linux/netdevice.h> +#include "efx_common.h" +#include "efx_channels.h" +#include "efx.h" +#include "mcdi.h" +#include "selftest.h" +#include "rx_common.h" +#include "tx_common.h" +#include "nic.h" +#include "io.h" +#include "mcdi_pcol.h" + +static unsigned int debug = (NETIF_MSG_DRV | NETIF_MSG_PROBE | + NETIF_MSG_LINK | NETIF_MSG_IFDOWN | + NETIF_MSG_IFUP | NETIF_MSG_RX_ERR | + NETIF_MSG_TX_ERR | NETIF_MSG_HW); +module_param(debug, uint, 0); +MODULE_PARM_DESC(debug, "Bitmapped debugging message enable value"); + +/* This is the time (in jiffies) between invocations of the hardware + * monitor. + * On Falcon-based NICs, this will: + * - Check the on-board hardware monitor; + * - Poll the link state and reconfigure the hardware as necessary. + * On Siena-based NICs for power systems with EEH support, this will give EEH a + * chance to start. + */ +static unsigned int efx_monitor_interval = 1 * HZ; + +/* How often and how many times to poll for a reset while waiting for a + * BIST that another function started to complete. + */ +#define BIST_WAIT_DELAY_MS 100 +#define BIST_WAIT_DELAY_COUNT 100 + +/* Default stats update time */ +#define STATS_PERIOD_MS_DEFAULT 1000 + +const unsigned int efx_reset_type_max = RESET_TYPE_MAX; +const char *const efx_reset_type_names[] = { + [RESET_TYPE_INVISIBLE] = "INVISIBLE", + [RESET_TYPE_ALL] = "ALL", + [RESET_TYPE_RECOVER_OR_ALL] = "RECOVER_OR_ALL", + [RESET_TYPE_WORLD] = "WORLD", + [RESET_TYPE_RECOVER_OR_DISABLE] = "RECOVER_OR_DISABLE", + [RESET_TYPE_DATAPATH] = "DATAPATH", + [RESET_TYPE_MC_BIST] = "MC_BIST", + [RESET_TYPE_DISABLE] = "DISABLE", + [RESET_TYPE_TX_WATCHDOG] = "TX_WATCHDOG", + [RESET_TYPE_INT_ERROR] = "INT_ERROR", + [RESET_TYPE_DMA_ERROR] = "DMA_ERROR", + [RESET_TYPE_TX_SKIP] = "TX_SKIP", + [RESET_TYPE_MC_FAILURE] = "MC_FAILURE", + [RESET_TYPE_MCDI_TIMEOUT] = "MCDI_TIMEOUT (FLR)", +}; + +#define RESET_TYPE(type) \ + STRING_TABLE_LOOKUP(type, efx_reset_type) + +/* Loopback mode names (see LOOPBACK_MODE()) */ +const unsigned int efx_loopback_mode_max = LOOPBACK_MAX; +const char *const efx_loopback_mode_names[] = { + [LOOPBACK_NONE] = "NONE", + [LOOPBACK_DATA] = "DATAPATH", + [LOOPBACK_GMAC] = "GMAC", + [LOOPBACK_XGMII] = "XGMII", + [LOOPBACK_XGXS] = "XGXS", + [LOOPBACK_XAUI] = "XAUI", + [LOOPBACK_GMII] = "GMII", + [LOOPBACK_SGMII] = "SGMII", + [LOOPBACK_XGBR] = "XGBR", + [LOOPBACK_XFI] = "XFI", + [LOOPBACK_XAUI_FAR] = "XAUI_FAR", + [LOOPBACK_GMII_FAR] = "GMII_FAR", + [LOOPBACK_SGMII_FAR] = "SGMII_FAR", + [LOOPBACK_XFI_FAR] = "XFI_FAR", + [LOOPBACK_GPHY] = "GPHY", + [LOOPBACK_PHYXS] = "PHYXS", + [LOOPBACK_PCS] = "PCS", + [LOOPBACK_PMAPMD] = "PMA/PMD", + [LOOPBACK_XPORT] = "XPORT", + [LOOPBACK_XGMII_WS] = "XGMII_WS", + [LOOPBACK_XAUI_WS] = "XAUI_WS", + [LOOPBACK_XAUI_WS_FAR] = "XAUI_WS_FAR", + [LOOPBACK_XAUI_WS_NEAR] = "XAUI_WS_NEAR", + [LOOPBACK_GMII_WS] = "GMII_WS", + [LOOPBACK_XFI_WS] = "XFI_WS", + [LOOPBACK_XFI_WS_FAR] = "XFI_WS_FAR", + [LOOPBACK_PHYXS_WS] = "PHYXS_WS", +}; + +/* Reset workqueue. If any NIC has a hardware failure then a reset will be + * queued onto this work queue. This is not a per-nic work queue, because + * efx_reset_work() acquires the rtnl lock, so resets are naturally serialised. + */ +static struct workqueue_struct *reset_workqueue; + +int efx_create_reset_workqueue(void) +{ + reset_workqueue = create_singlethread_workqueue("sfc_reset"); + if (!reset_workqueue) { + printk(KERN_ERR "Failed to create reset workqueue\n"); + return -ENOMEM; + } + + return 0; +} + +void efx_queue_reset_work(struct efx_nic *efx) +{ + queue_work(reset_workqueue, &efx->reset_work); +} + +void efx_flush_reset_workqueue(struct efx_nic *efx) +{ + cancel_work_sync(&efx->reset_work); +} + +void efx_destroy_reset_workqueue(void) +{ + if (reset_workqueue) { + destroy_workqueue(reset_workqueue); + reset_workqueue = NULL; + } +} + +/* We assume that efx->type->reconfigure_mac will always try to sync RX + * filters and therefore needs to read-lock the filter table against freeing + */ +void efx_mac_reconfigure(struct efx_nic *efx) +{ + if (efx->type->reconfigure_mac) { + down_read(&efx->filter_sem); + efx->type->reconfigure_mac(efx); + up_read(&efx->filter_sem); + } +} + +/* Asynchronous work item for changing MAC promiscuity and multicast + * hash. Avoid a drain/rx_ingress enable by reconfiguring the current + * MAC directly. + */ +static void efx_mac_work(struct work_struct *data) +{ + struct efx_nic *efx = container_of(data, struct efx_nic, mac_work); + + mutex_lock(&efx->mac_lock); + if (efx->port_enabled) + efx_mac_reconfigure(efx); + mutex_unlock(&efx->mac_lock); +} + +/* This ensures that the kernel is kept informed (via + * netif_carrier_on/off) of the link status, and also maintains the + * link status's stop on the port's TX queue. + */ +void efx_link_status_changed(struct efx_nic *efx) +{ + struct efx_link_state *link_state = &efx->link_state; + + /* SFC Bug 5356: A net_dev notifier is registered, so we must ensure + * that no events are triggered between unregister_netdev() and the + * driver unloading. A more general condition is that NETDEV_CHANGE + * can only be generated between NETDEV_UP and NETDEV_DOWN + */ + if (!netif_running(efx->net_dev)) + return; + + if (link_state->up != netif_carrier_ok(efx->net_dev)) { + efx->n_link_state_changes++; + + if (link_state->up) + netif_carrier_on(efx->net_dev); + else + netif_carrier_off(efx->net_dev); + } + + /* Status message for kernel log */ + if (link_state->up) + netif_info(efx, link, efx->net_dev, + "link up at %uMbps %s-duplex (MTU %d)\n", + link_state->speed, link_state->fd ? "full" : "half", + efx->net_dev->mtu); + else + netif_info(efx, link, efx->net_dev, "link down\n"); +} + +unsigned int efx_xdp_max_mtu(struct efx_nic *efx) +{ + /* The maximum MTU that we can fit in a single page, allowing for + * framing, overhead and XDP headroom. + */ + int overhead = EFX_MAX_FRAME_LEN(0) + sizeof(struct efx_rx_page_state) + + efx->rx_prefix_size + efx->type->rx_buffer_padding + + efx->rx_ip_align + XDP_PACKET_HEADROOM; + + return PAGE_SIZE - overhead; +} + +/* Context: process, rtnl_lock() held. */ +int efx_change_mtu(struct net_device *net_dev, int new_mtu) +{ + struct efx_nic *efx = netdev_priv(net_dev); + int rc; + + rc = efx_check_disabled(efx); + if (rc) + return rc; + + if (rtnl_dereference(efx->xdp_prog) && + new_mtu > efx_xdp_max_mtu(efx)) { + netif_err(efx, drv, efx->net_dev, + "Requested MTU of %d too big for XDP (max: %d)\n", + new_mtu, efx_xdp_max_mtu(efx)); + return -EINVAL; + } + + netif_dbg(efx, drv, efx->net_dev, "changing MTU to %d\n", new_mtu); + + efx_device_detach_sync(efx); + efx_stop_all(efx); + + mutex_lock(&efx->mac_lock); + net_dev->mtu = new_mtu; + efx_mac_reconfigure(efx); + mutex_unlock(&efx->mac_lock); + + efx_start_all(efx); + efx_device_attach_if_not_resetting(efx); + return 0; +} + +/************************************************************************** + * + * Hardware monitor + * + **************************************************************************/ + +/* Run periodically off the general workqueue */ +static void efx_monitor(struct work_struct *data) +{ + struct efx_nic *efx = container_of(data, struct efx_nic, + monitor_work.work); + + netif_vdbg(efx, timer, efx->net_dev, + "hardware monitor executing on CPU %d\n", + raw_smp_processor_id()); + BUG_ON(efx->type->monitor == NULL); + + /* If the mac_lock is already held then it is likely a port + * reconfiguration is already in place, which will likely do + * most of the work of monitor() anyway. + */ + if (mutex_trylock(&efx->mac_lock)) { + if (efx->port_enabled && efx->type->monitor) + efx->type->monitor(efx); + mutex_unlock(&efx->mac_lock); + } + + efx_start_monitor(efx); +} + +void efx_start_monitor(struct efx_nic *efx) +{ + if (efx->type->monitor) + queue_delayed_work(efx->workqueue, &efx->monitor_work, + efx_monitor_interval); +} + +/************************************************************************** + * + * Event queue processing + * + *************************************************************************/ + +/* Channels are shutdown and reinitialised whilst the NIC is running + * to propagate configuration changes (mtu, checksum offload), or + * to clear hardware error conditions + */ +static void efx_start_datapath(struct efx_nic *efx) +{ + netdev_features_t old_features = efx->net_dev->features; + bool old_rx_scatter = efx->rx_scatter; + size_t rx_buf_len; + + /* Calculate the rx buffer allocation parameters required to + * support the current MTU, including padding for header + * alignment and overruns. + */ + efx->rx_dma_len = (efx->rx_prefix_size + + EFX_MAX_FRAME_LEN(efx->net_dev->mtu) + + efx->type->rx_buffer_padding); + rx_buf_len = (sizeof(struct efx_rx_page_state) + XDP_PACKET_HEADROOM + + efx->rx_ip_align + efx->rx_dma_len); + if (rx_buf_len <= PAGE_SIZE) { + efx->rx_scatter = efx->type->always_rx_scatter; + efx->rx_buffer_order = 0; + } else if (efx->type->can_rx_scatter) { + BUILD_BUG_ON(EFX_RX_USR_BUF_SIZE % L1_CACHE_BYTES); + BUILD_BUG_ON(sizeof(struct efx_rx_page_state) + + 2 * ALIGN(NET_IP_ALIGN + EFX_RX_USR_BUF_SIZE, + EFX_RX_BUF_ALIGNMENT) > + PAGE_SIZE); + efx->rx_scatter = true; + efx->rx_dma_len = EFX_RX_USR_BUF_SIZE; + efx->rx_buffer_order = 0; + } else { + efx->rx_scatter = false; + efx->rx_buffer_order = get_order(rx_buf_len); + } + + efx_rx_config_page_split(efx); + if (efx->rx_buffer_order) + netif_dbg(efx, drv, efx->net_dev, + "RX buf len=%u; page order=%u batch=%u\n", + efx->rx_dma_len, efx->rx_buffer_order, + efx->rx_pages_per_batch); + else + netif_dbg(efx, drv, efx->net_dev, + "RX buf len=%u step=%u bpp=%u; page batch=%u\n", + efx->rx_dma_len, efx->rx_page_buf_step, + efx->rx_bufs_per_page, efx->rx_pages_per_batch); + + /* Restore previously fixed features in hw_features and remove + * features which are fixed now + */ + efx->net_dev->hw_features |= efx->net_dev->features; + efx->net_dev->hw_features &= ~efx->fixed_features; + efx->net_dev->features |= efx->fixed_features; + if (efx->net_dev->features != old_features) + netdev_features_change(efx->net_dev); + + /* RX filters may also have scatter-enabled flags */ + if ((efx->rx_scatter != old_rx_scatter) && + efx->type->filter_update_rx_scatter) + efx->type->filter_update_rx_scatter(efx); + + /* We must keep at least one descriptor in a TX ring empty. + * We could avoid this when the queue size does not exactly + * match the hardware ring size, but it's not that important. + * Therefore we stop the queue when one more skb might fill + * the ring completely. We wake it when half way back to + * empty. + */ + efx->txq_stop_thresh = efx->txq_entries - efx_tx_max_skb_descs(efx); + efx->txq_wake_thresh = efx->txq_stop_thresh / 2; + + /* Initialise the channels */ + efx_start_channels(efx); + + efx_ptp_start_datapath(efx); + + if (netif_device_present(efx->net_dev)) + netif_tx_wake_all_queues(efx->net_dev); +} + +static void efx_stop_datapath(struct efx_nic *efx) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + BUG_ON(efx->port_enabled); + + efx_ptp_stop_datapath(efx); + + efx_stop_channels(efx); +} + +/************************************************************************** + * + * Port handling + * + **************************************************************************/ + +static void efx_start_port(struct efx_nic *efx) +{ + netif_dbg(efx, ifup, efx->net_dev, "start port\n"); + BUG_ON(efx->port_enabled); + + mutex_lock(&efx->mac_lock); + efx->port_enabled = true; + + /* Ensure MAC ingress/egress is enabled */ + efx_mac_reconfigure(efx); + + mutex_unlock(&efx->mac_lock); +} + +/* Cancel work for MAC reconfiguration, periodic hardware monitoring + * and the async self-test, wait for them to finish and prevent them + * being scheduled again. This doesn't cover online resets, which + * should only be cancelled when removing the device. + */ +static void efx_stop_port(struct efx_nic *efx) +{ + netif_dbg(efx, ifdown, efx->net_dev, "stop port\n"); + + EFX_ASSERT_RESET_SERIALISED(efx); + + mutex_lock(&efx->mac_lock); + efx->port_enabled = false; + mutex_unlock(&efx->mac_lock); + + /* Serialise against efx_set_multicast_list() */ + netif_addr_lock_bh(efx->net_dev); + netif_addr_unlock_bh(efx->net_dev); + + cancel_delayed_work_sync(&efx->monitor_work); + efx_selftest_async_cancel(efx); + cancel_work_sync(&efx->mac_work); +} + +/* If the interface is supposed to be running but is not, start + * the hardware and software data path, regular activity for the port + * (MAC statistics, link polling, etc.) and schedule the port to be + * reconfigured. Interrupts must already be enabled. This function + * is safe to call multiple times, so long as the NIC is not disabled. + * Requires the RTNL lock. + */ +void efx_start_all(struct efx_nic *efx) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + BUG_ON(efx->state == STATE_DISABLED); + + /* Check that it is appropriate to restart the interface. All + * of these flags are safe to read under just the rtnl lock + */ + if (efx->port_enabled || !netif_running(efx->net_dev) || + efx->reset_pending) + return; + + efx_start_port(efx); + efx_start_datapath(efx); + + /* Start the hardware monitor if there is one */ + efx_start_monitor(efx); + + /* Link state detection is normally event-driven; we have + * to poll now because we could have missed a change + */ + mutex_lock(&efx->mac_lock); + if (efx->phy_op->poll(efx)) + efx_link_status_changed(efx); + mutex_unlock(&efx->mac_lock); + + if (efx->type->start_stats) { + efx->type->start_stats(efx); + efx->type->pull_stats(efx); + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, NULL); + spin_unlock_bh(&efx->stats_lock); + } +} + +/* Quiesce the hardware and software data path, and regular activity + * for the port without bringing the link down. Safe to call multiple + * times with the NIC in almost any state, but interrupts should be + * enabled. Requires the RTNL lock. + */ +void efx_stop_all(struct efx_nic *efx) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + + /* port_enabled can be read safely under the rtnl lock */ + if (!efx->port_enabled) + return; + + if (efx->type->update_stats) { + /* update stats before we go down so we can accurately count + * rx_nodesc_drops + */ + efx->type->pull_stats(efx); + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, NULL); + spin_unlock_bh(&efx->stats_lock); + efx->type->stop_stats(efx); + } + + efx_stop_port(efx); + + /* Stop the kernel transmit interface. This is only valid if + * the device is stopped or detached; otherwise the watchdog + * may fire immediately. + */ + WARN_ON(netif_running(efx->net_dev) && + netif_device_present(efx->net_dev)); + netif_tx_disable(efx->net_dev); + + efx_stop_datapath(efx); +} + +/* Context: process, dev_base_lock or RTNL held, non-blocking. */ +void efx_net_stats(struct net_device *net_dev, struct rtnl_link_stats64 *stats) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + spin_lock_bh(&efx->stats_lock); + efx->type->update_stats(efx, NULL, stats); + spin_unlock_bh(&efx->stats_lock); +} + +/* Push loopback/power/transmit disable settings to the PHY, and reconfigure + * the MAC appropriately. All other PHY configuration changes are pushed + * through phy_op->set_settings(), and pushed asynchronously to the MAC + * through efx_monitor(). + * + * Callers must hold the mac_lock + */ +int __efx_reconfigure_port(struct efx_nic *efx) +{ + enum efx_phy_mode phy_mode; + int rc = 0; + + WARN_ON(!mutex_is_locked(&efx->mac_lock)); + + /* Disable PHY transmit in mac level loopbacks */ + phy_mode = efx->phy_mode; + if (LOOPBACK_INTERNAL(efx)) + efx->phy_mode |= PHY_MODE_TX_DISABLED; + else + efx->phy_mode &= ~PHY_MODE_TX_DISABLED; + + if (efx->type->reconfigure_port) + rc = efx->type->reconfigure_port(efx); + + if (rc) + efx->phy_mode = phy_mode; + + return rc; +} + +/* Reinitialise the MAC to pick up new PHY settings, even if the port is + * disabled. + */ +int efx_reconfigure_port(struct efx_nic *efx) +{ + int rc; + + EFX_ASSERT_RESET_SERIALISED(efx); + + mutex_lock(&efx->mac_lock); + rc = __efx_reconfigure_port(efx); + mutex_unlock(&efx->mac_lock); + + return rc; +} + +/************************************************************************** + * + * Device reset and suspend + * + **************************************************************************/ + +static void efx_wait_for_bist_end(struct efx_nic *efx) +{ + int i; + + for (i = 0; i < BIST_WAIT_DELAY_COUNT; ++i) { + if (efx_mcdi_poll_reboot(efx)) + goto out; + msleep(BIST_WAIT_DELAY_MS); + } + + netif_err(efx, drv, efx->net_dev, "Warning: No MC reboot after BIST mode\n"); +out: + /* Either way unset the BIST flag. If we found no reboot we probably + * won't recover, but we should try. + */ + efx->mc_bist_for_other_fn = false; +} + +/* Try recovery mechanisms. + * For now only EEH is supported. + * Returns 0 if the recovery mechanisms are unsuccessful. + * Returns a non-zero value otherwise. + */ +int efx_try_recovery(struct efx_nic *efx) +{ +#ifdef CONFIG_EEH + /* A PCI error can occur and not be seen by EEH because nothing + * happens on the PCI bus. In this case the driver may fail and + * schedule a 'recover or reset', leading to this recovery handler. + * Manually call the eeh failure check function. + */ + struct eeh_dev *eehdev = pci_dev_to_eeh_dev(efx->pci_dev); + if (eeh_dev_check_failure(eehdev)) { + /* The EEH mechanisms will handle the error and reset the + * device if necessary. + */ + return 1; + } +#endif + return 0; +} + +/* Tears down the entire software state and most of the hardware state + * before reset. + */ +void efx_reset_down(struct efx_nic *efx, enum reset_type method) +{ + EFX_ASSERT_RESET_SERIALISED(efx); + + if (method == RESET_TYPE_MCDI_TIMEOUT) + efx->type->prepare_flr(efx); + + efx_stop_all(efx); + efx_disable_interrupts(efx); + + mutex_lock(&efx->mac_lock); + down_write(&efx->filter_sem); + mutex_lock(&efx->rss_lock); + if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && + method != RESET_TYPE_DATAPATH) + efx->phy_op->fini(efx); + efx->type->fini(efx); +} + +/* This function will always ensure that the locks acquired in + * efx_reset_down() are released. A failure return code indicates + * that we were unable to reinitialise the hardware, and the + * driver should be disabled. If ok is false, then the rx and tx + * engines are not restarted, pending a RESET_DISABLE. + */ +int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok) +{ + int rc; + + EFX_ASSERT_RESET_SERIALISED(efx); + + if (method == RESET_TYPE_MCDI_TIMEOUT) + efx->type->finish_flr(efx); + + /* Ensure that SRAM is initialised even if we're disabling the device */ + rc = efx->type->init(efx); + if (rc) { + netif_err(efx, drv, efx->net_dev, "failed to initialise NIC\n"); + goto fail; + } + + if (!ok) + goto fail; + + if (efx->port_initialized && method != RESET_TYPE_INVISIBLE && + method != RESET_TYPE_DATAPATH) { + rc = efx->phy_op->init(efx); + if (rc) + goto fail; + rc = efx->phy_op->reconfigure(efx); + if (rc && rc != -EPERM) + netif_err(efx, drv, efx->net_dev, + "could not restore PHY settings\n"); + } + + rc = efx_enable_interrupts(efx); + if (rc) + goto fail; + +#ifdef CONFIG_SFC_SRIOV + rc = efx->type->vswitching_restore(efx); + if (rc) /* not fatal; the PF will still work fine */ + netif_warn(efx, probe, efx->net_dev, + "failed to restore vswitching rc=%d;" + " VFs may not function\n", rc); +#endif + + if (efx->type->rx_restore_rss_contexts) + efx->type->rx_restore_rss_contexts(efx); + mutex_unlock(&efx->rss_lock); + efx->type->filter_table_restore(efx); + up_write(&efx->filter_sem); + if (efx->type->sriov_reset) + efx->type->sriov_reset(efx); + + mutex_unlock(&efx->mac_lock); + + efx_start_all(efx); + + if (efx->type->udp_tnl_push_ports) + efx->type->udp_tnl_push_ports(efx); + + return 0; + +fail: + efx->port_initialized = false; + + mutex_unlock(&efx->rss_lock); + up_write(&efx->filter_sem); + mutex_unlock(&efx->mac_lock); + + return rc; +} + +/* Reset the NIC using the specified method. Note that the reset may + * fail, in which case the card will be left in an unusable state. + * + * Caller must hold the rtnl_lock. + */ +int efx_reset(struct efx_nic *efx, enum reset_type method) +{ + bool disabled; + int rc, rc2; + + netif_info(efx, drv, efx->net_dev, "resetting (%s)\n", + RESET_TYPE(method)); + + efx_device_detach_sync(efx); + efx_reset_down(efx, method); + + rc = efx->type->reset(efx, method); + if (rc) { + netif_err(efx, drv, efx->net_dev, "failed to reset hardware\n"); + goto out; + } + + /* Clear flags for the scopes we covered. We assume the NIC and + * driver are now quiescent so that there is no race here. + */ + if (method < RESET_TYPE_MAX_METHOD) + efx->reset_pending &= -(1 << (method + 1)); + else /* it doesn't fit into the well-ordered scope hierarchy */ + __clear_bit(method, &efx->reset_pending); + + /* Reinitialise bus-mastering, which may have been turned off before + * the reset was scheduled. This is still appropriate, even in the + * RESET_TYPE_DISABLE since this driver generally assumes the hardware + * can respond to requests. + */ + pci_set_master(efx->pci_dev); + +out: + /* Leave device stopped if necessary */ + disabled = rc || + method == RESET_TYPE_DISABLE || + method == RESET_TYPE_RECOVER_OR_DISABLE; + rc2 = efx_reset_up(efx, method, !disabled); + if (rc2) { + disabled = true; + if (!rc) + rc = rc2; + } + + if (disabled) { + dev_close(efx->net_dev); + netif_err(efx, drv, efx->net_dev, "has been disabled\n"); + efx->state = STATE_DISABLED; + } else { + netif_dbg(efx, drv, efx->net_dev, "reset complete\n"); + efx_device_attach_if_not_resetting(efx); + } + return rc; +} + +/* The worker thread exists so that code that cannot sleep can + * schedule a reset for later. + */ +static void efx_reset_work(struct work_struct *data) +{ + struct efx_nic *efx = container_of(data, struct efx_nic, reset_work); + unsigned long pending; + enum reset_type method; + + pending = READ_ONCE(efx->reset_pending); + method = fls(pending) - 1; + + if (method == RESET_TYPE_MC_BIST) + efx_wait_for_bist_end(efx); + + if ((method == RESET_TYPE_RECOVER_OR_DISABLE || + method == RESET_TYPE_RECOVER_OR_ALL) && + efx_try_recovery(efx)) + return; + + if (!pending) + return; + + rtnl_lock(); + + /* We checked the state in efx_schedule_reset() but it may + * have changed by now. Now that we have the RTNL lock, + * it cannot change again. + */ + if (efx->state == STATE_READY) + (void)efx_reset(efx, method); + + rtnl_unlock(); +} + +void efx_schedule_reset(struct efx_nic *efx, enum reset_type type) +{ + enum reset_type method; + + if (efx->state == STATE_RECOVERY) { + netif_dbg(efx, drv, efx->net_dev, + "recovering: skip scheduling %s reset\n", + RESET_TYPE(type)); + return; + } + + switch (type) { + case RESET_TYPE_INVISIBLE: + case RESET_TYPE_ALL: + case RESET_TYPE_RECOVER_OR_ALL: + case RESET_TYPE_WORLD: + case RESET_TYPE_DISABLE: + case RESET_TYPE_RECOVER_OR_DISABLE: + case RESET_TYPE_DATAPATH: + case RESET_TYPE_MC_BIST: + case RESET_TYPE_MCDI_TIMEOUT: + method = type; + netif_dbg(efx, drv, efx->net_dev, "scheduling %s reset\n", + RESET_TYPE(method)); + break; + default: + method = efx->type->map_reset_reason(type); + netif_dbg(efx, drv, efx->net_dev, + "scheduling %s reset for %s\n", + RESET_TYPE(method), RESET_TYPE(type)); + break; + } + + set_bit(method, &efx->reset_pending); + smp_mb(); /* ensure we change reset_pending before checking state */ + + /* If we're not READY then just leave the flags set as the cue + * to abort probing or reschedule the reset later. + */ + if (READ_ONCE(efx->state) != STATE_READY) + return; + + /* efx_process_channel() will no longer read events once a + * reset is scheduled. So switch back to poll'd MCDI completions. + */ + efx_mcdi_mode_poll(efx); + + efx_queue_reset_work(efx); +} + +/************************************************************************** + * + * Dummy PHY/MAC operations + * + * Can be used for some unimplemented operations + * Needed so all function pointers are valid and do not have to be tested + * before use + * + **************************************************************************/ +int efx_port_dummy_op_int(struct efx_nic *efx) +{ + return 0; +} +void efx_port_dummy_op_void(struct efx_nic *efx) {} + +static bool efx_port_dummy_op_poll(struct efx_nic *efx) +{ + return false; +} + +static const struct efx_phy_operations efx_dummy_phy_operations = { + .init = efx_port_dummy_op_int, + .reconfigure = efx_port_dummy_op_int, + .poll = efx_port_dummy_op_poll, + .fini = efx_port_dummy_op_void, +}; + +/************************************************************************** + * + * Data housekeeping + * + **************************************************************************/ + +/* This zeroes out and then fills in the invariants in a struct + * efx_nic (including all sub-structures). + */ +int efx_init_struct(struct efx_nic *efx, + struct pci_dev *pci_dev, struct net_device *net_dev) +{ + int rc = -ENOMEM; + + /* Initialise common structures */ + INIT_LIST_HEAD(&efx->node); + INIT_LIST_HEAD(&efx->secondary_list); + spin_lock_init(&efx->biu_lock); +#ifdef CONFIG_SFC_MTD + INIT_LIST_HEAD(&efx->mtd_list); +#endif + INIT_WORK(&efx->reset_work, efx_reset_work); + INIT_DELAYED_WORK(&efx->monitor_work, efx_monitor); + efx_selftest_async_init(efx); + efx->pci_dev = pci_dev; + efx->msg_enable = debug; + efx->state = STATE_UNINIT; + strlcpy(efx->name, pci_name(pci_dev), sizeof(efx->name)); + + efx->net_dev = net_dev; + efx->rx_prefix_size = efx->type->rx_prefix_size; + efx->rx_ip_align = + NET_IP_ALIGN ? (efx->rx_prefix_size + NET_IP_ALIGN) % 4 : 0; + efx->rx_packet_hash_offset = + efx->type->rx_hash_offset - efx->type->rx_prefix_size; + efx->rx_packet_ts_offset = + efx->type->rx_ts_offset - efx->type->rx_prefix_size; + INIT_LIST_HEAD(&efx->rss_context.list); + mutex_init(&efx->rss_lock); + spin_lock_init(&efx->stats_lock); + efx->vi_stride = EFX_DEFAULT_VI_STRIDE; + efx->num_mac_stats = MC_CMD_MAC_NSTATS; + BUILD_BUG_ON(MC_CMD_MAC_NSTATS - 1 != MC_CMD_MAC_GENERATION_END); + mutex_init(&efx->mac_lock); +#ifdef CONFIG_RFS_ACCEL + mutex_init(&efx->rps_mutex); + spin_lock_init(&efx->rps_hash_lock); + /* Failure to allocate is not fatal, but may degrade ARFS performance */ + efx->rps_hash_table = kcalloc(EFX_ARFS_HASH_TABLE_SIZE, + sizeof(*efx->rps_hash_table), GFP_KERNEL); +#endif + efx->phy_op = &efx_dummy_phy_operations; + efx->mdio.dev = net_dev; + INIT_WORK(&efx->mac_work, efx_mac_work); + init_waitqueue_head(&efx->flush_wq); + + rc = efx_init_channels(efx); + if (rc) + goto fail; + + /* Would be good to use the net_dev name, but we're too early */ + snprintf(efx->workqueue_name, sizeof(efx->workqueue_name), "sfc%s", + pci_name(pci_dev)); + efx->workqueue = create_singlethread_workqueue(efx->workqueue_name); + if (!efx->workqueue) { + rc = -ENOMEM; + goto fail; + } + + return 0; + +fail: + efx_fini_struct(efx); + return rc; +} + +void efx_fini_struct(struct efx_nic *efx) +{ +#ifdef CONFIG_RFS_ACCEL + kfree(efx->rps_hash_table); +#endif + + efx_fini_channels(efx); + + kfree(efx->vpd_sn); + + if (efx->workqueue) { + destroy_workqueue(efx->workqueue); + efx->workqueue = NULL; + } +} + +/* This configures the PCI device to enable I/O and DMA. */ +int efx_init_io(struct efx_nic *efx, int bar, dma_addr_t dma_mask, + unsigned int mem_map_size) +{ + struct pci_dev *pci_dev = efx->pci_dev; + int rc; + + netif_dbg(efx, probe, efx->net_dev, "initialising I/O\n"); + + rc = pci_enable_device(pci_dev); + if (rc) { + netif_err(efx, probe, efx->net_dev, + "failed to enable PCI device\n"); + goto fail1; + } + + pci_set_master(pci_dev); + + /* Set the PCI DMA mask. Try all possibilities from our + * genuine mask down to 32 bits, because some architectures + * (e.g. x86_64 with iommu_sac_force set) will allow 40 bit + * masks event though they reject 46 bit masks. + */ + while (dma_mask > 0x7fffffffUL) { + rc = dma_set_mask_and_coherent(&pci_dev->dev, dma_mask); + if (rc == 0) + break; + dma_mask >>= 1; + } + if (rc) { + netif_err(efx, probe, efx->net_dev, + "could not find a suitable DMA mask\n"); + goto fail2; + } + netif_dbg(efx, probe, efx->net_dev, + "using DMA mask %llx\n", (unsigned long long)dma_mask); + + efx->membase_phys = pci_resource_start(efx->pci_dev, bar); + if (!efx->membase_phys) { + netif_err(efx, probe, efx->net_dev, + "ERROR: No BAR%d mapping from the BIOS. " + "Try pci=realloc on the kernel command line\n", bar); + rc = -ENODEV; + goto fail3; + } + + rc = pci_request_region(pci_dev, bar, "sfc"); + if (rc) { + netif_err(efx, probe, efx->net_dev, + "request for memory BAR failed\n"); + rc = -EIO; + goto fail3; + } + + efx->membase = ioremap(efx->membase_phys, mem_map_size); + if (!efx->membase) { + netif_err(efx, probe, efx->net_dev, + "could not map memory BAR at %llx+%x\n", + (unsigned long long)efx->membase_phys, mem_map_size); + rc = -ENOMEM; + goto fail4; + } + netif_dbg(efx, probe, efx->net_dev, + "memory BAR at %llx+%x (virtual %p)\n", + (unsigned long long)efx->membase_phys, mem_map_size, + efx->membase); + + return 0; + +fail4: + pci_release_region(efx->pci_dev, bar); +fail3: + efx->membase_phys = 0; +fail2: + pci_disable_device(efx->pci_dev); +fail1: + return rc; +} + +void efx_fini_io(struct efx_nic *efx, int bar) +{ + netif_dbg(efx, drv, efx->net_dev, "shutting down I/O\n"); + + if (efx->membase) { + iounmap(efx->membase); + efx->membase = NULL; + } + + if (efx->membase_phys) { + pci_release_region(efx->pci_dev, bar); + efx->membase_phys = 0; + } + + /* Don't disable bus-mastering if VFs are assigned */ + if (!pci_vfs_assigned(efx->pci_dev)) + pci_disable_device(efx->pci_dev); +} + +#ifdef CONFIG_SFC_MCDI_LOGGING +static ssize_t show_mcdi_log(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct efx_nic *efx = dev_get_drvdata(dev); + struct efx_mcdi_iface *mcdi = efx_mcdi(efx); + + return scnprintf(buf, PAGE_SIZE, "%d\n", mcdi->logging_enabled); +} + +static ssize_t set_mcdi_log(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct efx_nic *efx = dev_get_drvdata(dev); + struct efx_mcdi_iface *mcdi = efx_mcdi(efx); + bool enable = count > 0 && *buf != '0'; + + mcdi->logging_enabled = enable; + return count; +} + +static DEVICE_ATTR(mcdi_logging, 0644, show_mcdi_log, set_mcdi_log); + +void efx_init_mcdi_logging(struct efx_nic *efx) +{ + int rc = device_create_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); + + if (rc) { + netif_warn(efx, drv, efx->net_dev, + "failed to init net dev attributes\n"); + } +} + +void efx_fini_mcdi_logging(struct efx_nic *efx) +{ + device_remove_file(&efx->pci_dev->dev, &dev_attr_mcdi_logging); +} +#endif diff --git a/drivers/net/ethernet/sfc/efx_common.h b/drivers/net/ethernet/sfc/efx_common.h new file mode 100644 index 000000000000..fa2fc681e7f9 --- /dev/null +++ b/drivers/net/ethernet/sfc/efx_common.h @@ -0,0 +1,73 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_COMMON_H +#define EFX_COMMON_H + +int efx_init_io(struct efx_nic *efx, int bar, dma_addr_t dma_mask, + unsigned int mem_map_size); +void efx_fini_io(struct efx_nic *efx, int bar); +int efx_init_struct(struct efx_nic *efx, struct pci_dev *pci_dev, + struct net_device *net_dev); +void efx_fini_struct(struct efx_nic *efx); + +void efx_start_all(struct efx_nic *efx); +void efx_stop_all(struct efx_nic *efx); + +void efx_net_stats(struct net_device *net_dev, struct rtnl_link_stats64 *stats); + +int efx_create_reset_workqueue(void); +void efx_queue_reset_work(struct efx_nic *efx); +void efx_flush_reset_workqueue(struct efx_nic *efx); +void efx_destroy_reset_workqueue(void); + +void efx_start_monitor(struct efx_nic *efx); + +int __efx_reconfigure_port(struct efx_nic *efx); +int efx_reconfigure_port(struct efx_nic *efx); + +#define EFX_ASSERT_RESET_SERIALISED(efx) \ + do { \ + if ((efx->state == STATE_READY) || \ + (efx->state == STATE_RECOVERY) || \ + (efx->state == STATE_DISABLED)) \ + ASSERT_RTNL(); \ + } while (0) + +int efx_try_recovery(struct efx_nic *efx); +void efx_reset_down(struct efx_nic *efx, enum reset_type method); +int efx_reset_up(struct efx_nic *efx, enum reset_type method, bool ok); +int efx_reset(struct efx_nic *efx, enum reset_type method); +void efx_schedule_reset(struct efx_nic *efx, enum reset_type type); + +static inline int efx_check_disabled(struct efx_nic *efx) +{ + if (efx->state == STATE_DISABLED || efx->state == STATE_RECOVERY) { + netif_err(efx, drv, efx->net_dev, + "device is disabled due to earlier errors\n"); + return -EIO; + } + return 0; +} + +#ifdef CONFIG_SFC_MCDI_LOGGING +void efx_init_mcdi_logging(struct efx_nic *efx); +void efx_fini_mcdi_logging(struct efx_nic *efx); +#else +static inline void efx_init_mcdi_logging(struct efx_nic *efx) {} +static inline void efx_fini_mcdi_logging(struct efx_nic *efx) {} +#endif + +void efx_mac_reconfigure(struct efx_nic *efx); +void efx_link_status_changed(struct efx_nic *efx); +unsigned int efx_xdp_max_mtu(struct efx_nic *efx); +int efx_change_mtu(struct net_device *net_dev, int new_mtu); + +#endif diff --git a/drivers/net/ethernet/sfc/ethtool.c b/drivers/net/ethernet/sfc/ethtool.c index b31032da4bcb..993b5769525b 100644 --- a/drivers/net/ethernet/sfc/ethtool.c +++ b/drivers/net/ethernet/sfc/ethtool.c @@ -13,92 +13,13 @@ #include "workarounds.h" #include "selftest.h" #include "efx.h" +#include "efx_channels.h" +#include "rx_common.h" +#include "tx_common.h" +#include "ethtool_common.h" #include "filter.h" #include "nic.h" -struct efx_sw_stat_desc { - const char *name; - enum { - EFX_ETHTOOL_STAT_SOURCE_nic, - EFX_ETHTOOL_STAT_SOURCE_channel, - EFX_ETHTOOL_STAT_SOURCE_tx_queue - } source; - unsigned offset; - u64(*get_stat) (void *field); /* Reader function */ -}; - -/* Initialiser for a struct efx_sw_stat_desc with type-checking */ -#define EFX_ETHTOOL_STAT(stat_name, source_name, field, field_type, \ - get_stat_function) { \ - .name = #stat_name, \ - .source = EFX_ETHTOOL_STAT_SOURCE_##source_name, \ - .offset = ((((field_type *) 0) == \ - &((struct efx_##source_name *)0)->field) ? \ - offsetof(struct efx_##source_name, field) : \ - offsetof(struct efx_##source_name, field)), \ - .get_stat = get_stat_function, \ -} - -static u64 efx_get_uint_stat(void *field) -{ - return *(unsigned int *)field; -} - -static u64 efx_get_atomic_stat(void *field) -{ - return atomic_read((atomic_t *) field); -} - -#define EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(field) \ - EFX_ETHTOOL_STAT(field, nic, field, \ - atomic_t, efx_get_atomic_stat) - -#define EFX_ETHTOOL_UINT_CHANNEL_STAT(field) \ - EFX_ETHTOOL_STAT(field, channel, n_##field, \ - unsigned int, efx_get_uint_stat) -#define EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(field) \ - EFX_ETHTOOL_STAT(field, channel, field, \ - unsigned int, efx_get_uint_stat) - -#define EFX_ETHTOOL_UINT_TXQ_STAT(field) \ - EFX_ETHTOOL_STAT(tx_##field, tx_queue, field, \ - unsigned int, efx_get_uint_stat) - -static const struct efx_sw_stat_desc efx_sw_stat_desc[] = { - EFX_ETHTOOL_UINT_TXQ_STAT(merge_events), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_bursts), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_long_headers), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_packets), - EFX_ETHTOOL_UINT_TXQ_STAT(tso_fallbacks), - EFX_ETHTOOL_UINT_TXQ_STAT(pushes), - EFX_ETHTOOL_UINT_TXQ_STAT(pio_packets), - EFX_ETHTOOL_UINT_TXQ_STAT(cb_packets), - EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(rx_reset), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tobe_disc), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_ip_hdr_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_tcp_udp_chksum_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_eth_crc_err), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_mcast_mismatch), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_frm_trunc), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_events), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_packets), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_drops), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_bad_drops), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_tx), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_redirect), -#ifdef CONFIG_RFS_ACCEL - EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(rfs_filter_count), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_succeeded), - EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_failed), -#endif -}; - -#define EFX_ETHTOOL_SW_STAT_COUNT ARRAY_SIZE(efx_sw_stat_desc) - #define EFX_ETHTOOL_EEPROM_MAGIC 0xEFAB /************************************************************************** @@ -185,18 +106,6 @@ efx_ethtool_set_link_ksettings(struct net_device *net_dev, return rc; } -static void efx_ethtool_get_drvinfo(struct net_device *net_dev, - struct ethtool_drvinfo *info) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); - strlcpy(info->version, EFX_DRIVER_VERSION, sizeof(info->version)); - efx_mcdi_print_fwver(efx, info->fw_version, - sizeof(info->fw_version)); - strlcpy(info->bus_info, pci_name(efx->pci_dev), sizeof(info->bus_info)); -} - static int efx_ethtool_get_regs_len(struct net_device *net_dev) { return efx_nic_get_regs_len(netdev_priv(net_dev)); @@ -211,341 +120,6 @@ static void efx_ethtool_get_regs(struct net_device *net_dev, efx_nic_get_regs(efx, buf); } -static u32 efx_ethtool_get_msglevel(struct net_device *net_dev) -{ - struct efx_nic *efx = netdev_priv(net_dev); - return efx->msg_enable; -} - -static void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable) -{ - struct efx_nic *efx = netdev_priv(net_dev); - efx->msg_enable = msg_enable; -} - -/** - * efx_fill_test - fill in an individual self-test entry - * @test_index: Index of the test - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * @test: Pointer to test result (used only if data != %NULL) - * @unit_format: Unit name format (e.g. "chan\%d") - * @unit_id: Unit id (e.g. 0 for "chan0") - * @test_format: Test name format (e.g. "loopback.\%s.tx.sent") - * @test_id: Test id (e.g. "PHYXS" for "loopback.PHYXS.tx_sent") - * - * Fill in an individual self-test entry. - */ -static void efx_fill_test(unsigned int test_index, u8 *strings, u64 *data, - int *test, const char *unit_format, int unit_id, - const char *test_format, const char *test_id) -{ - char unit_str[ETH_GSTRING_LEN], test_str[ETH_GSTRING_LEN]; - - /* Fill data value, if applicable */ - if (data) - data[test_index] = *test; - - /* Fill string, if applicable */ - if (strings) { - if (strchr(unit_format, '%')) - snprintf(unit_str, sizeof(unit_str), - unit_format, unit_id); - else - strcpy(unit_str, unit_format); - snprintf(test_str, sizeof(test_str), test_format, test_id); - snprintf(strings + test_index * ETH_GSTRING_LEN, - ETH_GSTRING_LEN, - "%-6s %-24s", unit_str, test_str); - } -} - -#define EFX_CHANNEL_NAME(_channel) "chan%d", _channel->channel -#define EFX_TX_QUEUE_NAME(_tx_queue) "txq%d", _tx_queue->queue -#define EFX_RX_QUEUE_NAME(_rx_queue) "rxq%d", _rx_queue->queue -#define EFX_LOOPBACK_NAME(_mode, _counter) \ - "loopback.%s." _counter, STRING_TABLE_LOOKUP(_mode, efx_loopback_mode) - -/** - * efx_fill_loopback_test - fill in a block of loopback self-test entries - * @efx: Efx NIC - * @lb_tests: Efx loopback self-test results structure - * @mode: Loopback test mode - * @test_index: Starting index of the test - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * - * Fill in a block of loopback self-test entries. Return new test - * index. - */ -static int efx_fill_loopback_test(struct efx_nic *efx, - struct efx_loopback_self_tests *lb_tests, - enum efx_loopback_mode mode, - unsigned int test_index, - u8 *strings, u64 *data) -{ - struct efx_channel *channel = - efx_get_channel(efx, efx->tx_channel_offset); - struct efx_tx_queue *tx_queue; - - efx_for_each_channel_tx_queue(tx_queue, channel) { - efx_fill_test(test_index++, strings, data, - &lb_tests->tx_sent[tx_queue->queue], - EFX_TX_QUEUE_NAME(tx_queue), - EFX_LOOPBACK_NAME(mode, "tx_sent")); - efx_fill_test(test_index++, strings, data, - &lb_tests->tx_done[tx_queue->queue], - EFX_TX_QUEUE_NAME(tx_queue), - EFX_LOOPBACK_NAME(mode, "tx_done")); - } - efx_fill_test(test_index++, strings, data, - &lb_tests->rx_good, - "rx", 0, - EFX_LOOPBACK_NAME(mode, "rx_good")); - efx_fill_test(test_index++, strings, data, - &lb_tests->rx_bad, - "rx", 0, - EFX_LOOPBACK_NAME(mode, "rx_bad")); - - return test_index; -} - -/** - * efx_ethtool_fill_self_tests - get self-test details - * @efx: Efx NIC - * @tests: Efx self-test results structure, or %NULL - * @strings: Ethtool strings, or %NULL - * @data: Ethtool test results, or %NULL - * - * Get self-test number of strings, strings, and/or test results. - * Return number of strings (== number of test results). - * - * The reason for merging these three functions is to make sure that - * they can never be inconsistent. - */ -static int efx_ethtool_fill_self_tests(struct efx_nic *efx, - struct efx_self_tests *tests, - u8 *strings, u64 *data) -{ - struct efx_channel *channel; - unsigned int n = 0, i; - enum efx_loopback_mode mode; - - efx_fill_test(n++, strings, data, &tests->phy_alive, - "phy", 0, "alive", NULL); - efx_fill_test(n++, strings, data, &tests->nvram, - "core", 0, "nvram", NULL); - efx_fill_test(n++, strings, data, &tests->interrupt, - "core", 0, "interrupt", NULL); - - /* Event queues */ - efx_for_each_channel(channel, efx) { - efx_fill_test(n++, strings, data, - &tests->eventq_dma[channel->channel], - EFX_CHANNEL_NAME(channel), - "eventq.dma", NULL); - efx_fill_test(n++, strings, data, - &tests->eventq_int[channel->channel], - EFX_CHANNEL_NAME(channel), - "eventq.int", NULL); - } - - efx_fill_test(n++, strings, data, &tests->memory, - "core", 0, "memory", NULL); - efx_fill_test(n++, strings, data, &tests->registers, - "core", 0, "registers", NULL); - - if (efx->phy_op->run_tests != NULL) { - EFX_WARN_ON_PARANOID(efx->phy_op->test_name == NULL); - - for (i = 0; true; ++i) { - const char *name; - - EFX_WARN_ON_PARANOID(i >= EFX_MAX_PHY_TESTS); - name = efx->phy_op->test_name(efx, i); - if (name == NULL) - break; - - efx_fill_test(n++, strings, data, &tests->phy_ext[i], - "phy", 0, name, NULL); - } - } - - /* Loopback tests */ - for (mode = LOOPBACK_NONE; mode <= LOOPBACK_TEST_MAX; mode++) { - if (!(efx->loopback_modes & (1 << mode))) - continue; - n = efx_fill_loopback_test(efx, - &tests->loopback[mode], mode, n, - strings, data); - } - - return n; -} - -static size_t efx_describe_per_queue_stats(struct efx_nic *efx, u8 *strings) -{ - size_t n_stats = 0; - struct efx_channel *channel; - - efx_for_each_channel(channel, efx) { - if (efx_channel_has_tx_queues(channel)) { - n_stats++; - if (strings != NULL) { - snprintf(strings, ETH_GSTRING_LEN, - "tx-%u.tx_packets", - channel->tx_queue[0].queue / - EFX_TXQ_TYPES); - - strings += ETH_GSTRING_LEN; - } - } - } - efx_for_each_channel(channel, efx) { - if (efx_channel_has_rx_queue(channel)) { - n_stats++; - if (strings != NULL) { - snprintf(strings, ETH_GSTRING_LEN, - "rx-%d.rx_packets", channel->channel); - strings += ETH_GSTRING_LEN; - } - } - } - if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { - unsigned short xdp; - - for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { - n_stats++; - if (strings) { - snprintf(strings, ETH_GSTRING_LEN, - "tx-xdp-cpu-%hu.tx_packets", xdp); - strings += ETH_GSTRING_LEN; - } - } - } - - return n_stats; -} - -static int efx_ethtool_get_sset_count(struct net_device *net_dev, - int string_set) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - switch (string_set) { - case ETH_SS_STATS: - return efx->type->describe_stats(efx, NULL) + - EFX_ETHTOOL_SW_STAT_COUNT + - efx_describe_per_queue_stats(efx, NULL) + - efx_ptp_describe_stats(efx, NULL); - case ETH_SS_TEST: - return efx_ethtool_fill_self_tests(efx, NULL, NULL, NULL); - default: - return -EINVAL; - } -} - -static void efx_ethtool_get_strings(struct net_device *net_dev, - u32 string_set, u8 *strings) -{ - struct efx_nic *efx = netdev_priv(net_dev); - int i; - - switch (string_set) { - case ETH_SS_STATS: - strings += (efx->type->describe_stats(efx, strings) * - ETH_GSTRING_LEN); - for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) - strlcpy(strings + i * ETH_GSTRING_LEN, - efx_sw_stat_desc[i].name, ETH_GSTRING_LEN); - strings += EFX_ETHTOOL_SW_STAT_COUNT * ETH_GSTRING_LEN; - strings += (efx_describe_per_queue_stats(efx, strings) * - ETH_GSTRING_LEN); - efx_ptp_describe_stats(efx, strings); - break; - case ETH_SS_TEST: - efx_ethtool_fill_self_tests(efx, NULL, strings, NULL); - break; - default: - /* No other string sets */ - break; - } -} - -static void efx_ethtool_get_stats(struct net_device *net_dev, - struct ethtool_stats *stats, - u64 *data) -{ - struct efx_nic *efx = netdev_priv(net_dev); - const struct efx_sw_stat_desc *stat; - struct efx_channel *channel; - struct efx_tx_queue *tx_queue; - struct efx_rx_queue *rx_queue; - int i; - - spin_lock_bh(&efx->stats_lock); - - /* Get NIC statistics */ - data += efx->type->update_stats(efx, data, NULL); - - /* Get software statistics */ - for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) { - stat = &efx_sw_stat_desc[i]; - switch (stat->source) { - case EFX_ETHTOOL_STAT_SOURCE_nic: - data[i] = stat->get_stat((void *)efx + stat->offset); - break; - case EFX_ETHTOOL_STAT_SOURCE_channel: - data[i] = 0; - efx_for_each_channel(channel, efx) - data[i] += stat->get_stat((void *)channel + - stat->offset); - break; - case EFX_ETHTOOL_STAT_SOURCE_tx_queue: - data[i] = 0; - efx_for_each_channel(channel, efx) { - efx_for_each_channel_tx_queue(tx_queue, channel) - data[i] += - stat->get_stat((void *)tx_queue - + stat->offset); - } - break; - } - } - data += EFX_ETHTOOL_SW_STAT_COUNT; - - spin_unlock_bh(&efx->stats_lock); - - efx_for_each_channel(channel, efx) { - if (efx_channel_has_tx_queues(channel)) { - *data = 0; - efx_for_each_channel_tx_queue(tx_queue, channel) { - *data += tx_queue->tx_packets; - } - data++; - } - } - efx_for_each_channel(channel, efx) { - if (efx_channel_has_rx_queue(channel)) { - *data = 0; - efx_for_each_channel_rx_queue(rx_queue, channel) { - *data += rx_queue->rx_packets; - } - data++; - } - } - if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { - int xdp; - - for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { - data[0] = efx->xdp_tx_queues[xdp]->tx_packets; - data++; - } - } - - efx_ptp_update_stats(efx, data); -} - static void efx_ethtool_self_test(struct net_device *net_dev, struct ethtool_test *test, u64 *data) { @@ -787,16 +361,6 @@ out: return rc; } -static void efx_ethtool_get_pauseparam(struct net_device *net_dev, - struct ethtool_pauseparam *pause) -{ - struct efx_nic *efx = netdev_priv(net_dev); - - pause->rx_pause = !!(efx->wanted_fc & EFX_FC_RX); - pause->tx_pause = !!(efx->wanted_fc & EFX_FC_TX); - pause->autoneg = !!(efx->wanted_fc & EFX_FC_AUTO); -} - static void efx_ethtool_get_wol(struct net_device *net_dev, struct ethtool_wolinfo *wol) { @@ -1456,7 +1020,7 @@ static int efx_ethtool_set_rxfh_context(struct net_device *net_dev, rc = -ENOMEM; goto out_unlock; } - ctx->context_id = EFX_EF10_RSS_CONTEXT_INVALID; + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; /* Initialise indir table and key to defaults */ efx_set_default_rx_indir_table(efx, ctx); netdev_rss_key_fill(ctx->rx_hash_key, sizeof(ctx->rx_hash_key)); diff --git a/drivers/net/ethernet/sfc/ethtool_common.c b/drivers/net/ethernet/sfc/ethtool_common.c new file mode 100644 index 000000000000..b8d281ab6c7a --- /dev/null +++ b/drivers/net/ethernet/sfc/ethtool_common.c @@ -0,0 +1,457 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ +#include <linux/module.h> +#include <linux/netdevice.h> +#include "net_driver.h" +#include "mcdi.h" +#include "nic.h" +#include "selftest.h" +#include "ethtool_common.h" + +struct efx_sw_stat_desc { + const char *name; + enum { + EFX_ETHTOOL_STAT_SOURCE_nic, + EFX_ETHTOOL_STAT_SOURCE_channel, + EFX_ETHTOOL_STAT_SOURCE_tx_queue + } source; + unsigned int offset; + u64 (*get_stat)(void *field); /* Reader function */ +}; + +/* Initialiser for a struct efx_sw_stat_desc with type-checking */ +#define EFX_ETHTOOL_STAT(stat_name, source_name, field, field_type, \ + get_stat_function) { \ + .name = #stat_name, \ + .source = EFX_ETHTOOL_STAT_SOURCE_##source_name, \ + .offset = ((((field_type *) 0) == \ + &((struct efx_##source_name *)0)->field) ? \ + offsetof(struct efx_##source_name, field) : \ + offsetof(struct efx_##source_name, field)), \ + .get_stat = get_stat_function, \ +} + +static u64 efx_get_uint_stat(void *field) +{ + return *(unsigned int *)field; +} + +static u64 efx_get_atomic_stat(void *field) +{ + return atomic_read((atomic_t *) field); +} + +#define EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(field) \ + EFX_ETHTOOL_STAT(field, nic, field, \ + atomic_t, efx_get_atomic_stat) + +#define EFX_ETHTOOL_UINT_CHANNEL_STAT(field) \ + EFX_ETHTOOL_STAT(field, channel, n_##field, \ + unsigned int, efx_get_uint_stat) +#define EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(field) \ + EFX_ETHTOOL_STAT(field, channel, field, \ + unsigned int, efx_get_uint_stat) + +#define EFX_ETHTOOL_UINT_TXQ_STAT(field) \ + EFX_ETHTOOL_STAT(tx_##field, tx_queue, field, \ + unsigned int, efx_get_uint_stat) + +static const struct efx_sw_stat_desc efx_sw_stat_desc[] = { + EFX_ETHTOOL_UINT_TXQ_STAT(merge_events), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_bursts), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_long_headers), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_packets), + EFX_ETHTOOL_UINT_TXQ_STAT(tso_fallbacks), + EFX_ETHTOOL_UINT_TXQ_STAT(pushes), + EFX_ETHTOOL_UINT_TXQ_STAT(pio_packets), + EFX_ETHTOOL_UINT_TXQ_STAT(cb_packets), + EFX_ETHTOOL_ATOMIC_NIC_ERROR_STAT(rx_reset), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tobe_disc), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_inner_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_ip_hdr_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_outer_tcp_udp_chksum_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_eth_crc_err), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_mcast_mismatch), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_frm_trunc), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_events), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_merge_packets), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_drops), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_bad_drops), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_tx), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rx_xdp_redirect), +#ifdef CONFIG_RFS_ACCEL + EFX_ETHTOOL_UINT_CHANNEL_STAT_NO_N(rfs_filter_count), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_succeeded), + EFX_ETHTOOL_UINT_CHANNEL_STAT(rfs_failed), +#endif +}; + +#define EFX_ETHTOOL_SW_STAT_COUNT ARRAY_SIZE(efx_sw_stat_desc) + +void efx_ethtool_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + strlcpy(info->driver, KBUILD_MODNAME, sizeof(info->driver)); + strlcpy(info->version, EFX_DRIVER_VERSION, sizeof(info->version)); + efx_mcdi_print_fwver(efx, info->fw_version, + sizeof(info->fw_version)); + strlcpy(info->bus_info, pci_name(efx->pci_dev), sizeof(info->bus_info)); +} + +u32 efx_ethtool_get_msglevel(struct net_device *net_dev) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + return efx->msg_enable; +} + +void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + efx->msg_enable = msg_enable; +} + +void efx_ethtool_get_pauseparam(struct net_device *net_dev, + struct ethtool_pauseparam *pause) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + pause->rx_pause = !!(efx->wanted_fc & EFX_FC_RX); + pause->tx_pause = !!(efx->wanted_fc & EFX_FC_TX); + pause->autoneg = !!(efx->wanted_fc & EFX_FC_AUTO); +} + +/** + * efx_fill_test - fill in an individual self-test entry + * @test_index: Index of the test + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * @test: Pointer to test result (used only if data != %NULL) + * @unit_format: Unit name format (e.g. "chan\%d") + * @unit_id: Unit id (e.g. 0 for "chan0") + * @test_format: Test name format (e.g. "loopback.\%s.tx.sent") + * @test_id: Test id (e.g. "PHYXS" for "loopback.PHYXS.tx_sent") + * + * Fill in an individual self-test entry. + */ +static void efx_fill_test(unsigned int test_index, u8 *strings, u64 *data, + int *test, const char *unit_format, int unit_id, + const char *test_format, const char *test_id) +{ + char unit_str[ETH_GSTRING_LEN], test_str[ETH_GSTRING_LEN]; + + /* Fill data value, if applicable */ + if (data) + data[test_index] = *test; + + /* Fill string, if applicable */ + if (strings) { + if (strchr(unit_format, '%')) + snprintf(unit_str, sizeof(unit_str), + unit_format, unit_id); + else + strcpy(unit_str, unit_format); + snprintf(test_str, sizeof(test_str), test_format, test_id); + snprintf(strings + test_index * ETH_GSTRING_LEN, + ETH_GSTRING_LEN, + "%-6s %-24s", unit_str, test_str); + } +} + +#define EFX_CHANNEL_NAME(_channel) "chan%d", _channel->channel +#define EFX_TX_QUEUE_NAME(_tx_queue) "txq%d", _tx_queue->queue +#define EFX_RX_QUEUE_NAME(_rx_queue) "rxq%d", _rx_queue->queue +#define EFX_LOOPBACK_NAME(_mode, _counter) \ + "loopback.%s." _counter, STRING_TABLE_LOOKUP(_mode, efx_loopback_mode) + +/** + * efx_fill_loopback_test - fill in a block of loopback self-test entries + * @efx: Efx NIC + * @lb_tests: Efx loopback self-test results structure + * @mode: Loopback test mode + * @test_index: Starting index of the test + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * + * Fill in a block of loopback self-test entries. Return new test + * index. + */ +static int efx_fill_loopback_test(struct efx_nic *efx, + struct efx_loopback_self_tests *lb_tests, + enum efx_loopback_mode mode, + unsigned int test_index, + u8 *strings, u64 *data) +{ + struct efx_channel *channel = + efx_get_channel(efx, efx->tx_channel_offset); + struct efx_tx_queue *tx_queue; + + efx_for_each_channel_tx_queue(tx_queue, channel) { + efx_fill_test(test_index++, strings, data, + &lb_tests->tx_sent[tx_queue->queue], + EFX_TX_QUEUE_NAME(tx_queue), + EFX_LOOPBACK_NAME(mode, "tx_sent")); + efx_fill_test(test_index++, strings, data, + &lb_tests->tx_done[tx_queue->queue], + EFX_TX_QUEUE_NAME(tx_queue), + EFX_LOOPBACK_NAME(mode, "tx_done")); + } + efx_fill_test(test_index++, strings, data, + &lb_tests->rx_good, + "rx", 0, + EFX_LOOPBACK_NAME(mode, "rx_good")); + efx_fill_test(test_index++, strings, data, + &lb_tests->rx_bad, + "rx", 0, + EFX_LOOPBACK_NAME(mode, "rx_bad")); + + return test_index; +} + +/** + * efx_ethtool_fill_self_tests - get self-test details + * @efx: Efx NIC + * @tests: Efx self-test results structure, or %NULL + * @strings: Ethtool strings, or %NULL + * @data: Ethtool test results, or %NULL + * + * Get self-test number of strings, strings, and/or test results. + * Return number of strings (== number of test results). + * + * The reason for merging these three functions is to make sure that + * they can never be inconsistent. + */ +int efx_ethtool_fill_self_tests(struct efx_nic *efx, + struct efx_self_tests *tests, + u8 *strings, u64 *data) +{ + struct efx_channel *channel; + unsigned int n = 0, i; + enum efx_loopback_mode mode; + + efx_fill_test(n++, strings, data, &tests->phy_alive, + "phy", 0, "alive", NULL); + efx_fill_test(n++, strings, data, &tests->nvram, + "core", 0, "nvram", NULL); + efx_fill_test(n++, strings, data, &tests->interrupt, + "core", 0, "interrupt", NULL); + + /* Event queues */ + efx_for_each_channel(channel, efx) { + efx_fill_test(n++, strings, data, + &tests->eventq_dma[channel->channel], + EFX_CHANNEL_NAME(channel), + "eventq.dma", NULL); + efx_fill_test(n++, strings, data, + &tests->eventq_int[channel->channel], + EFX_CHANNEL_NAME(channel), + "eventq.int", NULL); + } + + efx_fill_test(n++, strings, data, &tests->memory, + "core", 0, "memory", NULL); + efx_fill_test(n++, strings, data, &tests->registers, + "core", 0, "registers", NULL); + + if (efx->phy_op->run_tests != NULL) { + EFX_WARN_ON_PARANOID(efx->phy_op->test_name == NULL); + + for (i = 0; true; ++i) { + const char *name; + + EFX_WARN_ON_PARANOID(i >= EFX_MAX_PHY_TESTS); + name = efx->phy_op->test_name(efx, i); + if (name == NULL) + break; + + efx_fill_test(n++, strings, data, &tests->phy_ext[i], + "phy", 0, name, NULL); + } + } + + /* Loopback tests */ + for (mode = LOOPBACK_NONE; mode <= LOOPBACK_TEST_MAX; mode++) { + if (!(efx->loopback_modes & (1 << mode))) + continue; + n = efx_fill_loopback_test(efx, + &tests->loopback[mode], mode, n, + strings, data); + } + + return n; +} + +static size_t efx_describe_per_queue_stats(struct efx_nic *efx, u8 *strings) +{ + size_t n_stats = 0; + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + if (efx_channel_has_tx_queues(channel)) { + n_stats++; + if (strings != NULL) { + snprintf(strings, ETH_GSTRING_LEN, + "tx-%u.tx_packets", + channel->tx_queue[0].queue / + EFX_TXQ_TYPES); + + strings += ETH_GSTRING_LEN; + } + } + } + efx_for_each_channel(channel, efx) { + if (efx_channel_has_rx_queue(channel)) { + n_stats++; + if (strings != NULL) { + snprintf(strings, ETH_GSTRING_LEN, + "rx-%d.rx_packets", channel->channel); + strings += ETH_GSTRING_LEN; + } + } + } + if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { + unsigned short xdp; + + for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { + n_stats++; + if (strings) { + snprintf(strings, ETH_GSTRING_LEN, + "tx-xdp-cpu-%hu.tx_packets", xdp); + strings += ETH_GSTRING_LEN; + } + } + } + + return n_stats; +} + +int efx_ethtool_get_sset_count(struct net_device *net_dev, int string_set) +{ + struct efx_nic *efx = netdev_priv(net_dev); + + switch (string_set) { + case ETH_SS_STATS: + return efx->type->describe_stats(efx, NULL) + + EFX_ETHTOOL_SW_STAT_COUNT + + efx_describe_per_queue_stats(efx, NULL) + + efx_ptp_describe_stats(efx, NULL); + case ETH_SS_TEST: + return efx_ethtool_fill_self_tests(efx, NULL, NULL, NULL); + default: + return -EINVAL; + } +} + +void efx_ethtool_get_strings(struct net_device *net_dev, + u32 string_set, u8 *strings) +{ + struct efx_nic *efx = netdev_priv(net_dev); + int i; + + switch (string_set) { + case ETH_SS_STATS: + strings += (efx->type->describe_stats(efx, strings) * + ETH_GSTRING_LEN); + for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) + strlcpy(strings + i * ETH_GSTRING_LEN, + efx_sw_stat_desc[i].name, ETH_GSTRING_LEN); + strings += EFX_ETHTOOL_SW_STAT_COUNT * ETH_GSTRING_LEN; + strings += (efx_describe_per_queue_stats(efx, strings) * + ETH_GSTRING_LEN); + efx_ptp_describe_stats(efx, strings); + break; + case ETH_SS_TEST: + efx_ethtool_fill_self_tests(efx, NULL, strings, NULL); + break; + default: + /* No other string sets */ + break; + } +} + +void efx_ethtool_get_stats(struct net_device *net_dev, + struct ethtool_stats *stats, + u64 *data) +{ + struct efx_nic *efx = netdev_priv(net_dev); + const struct efx_sw_stat_desc *stat; + struct efx_channel *channel; + struct efx_tx_queue *tx_queue; + struct efx_rx_queue *rx_queue; + int i; + + spin_lock_bh(&efx->stats_lock); + + /* Get NIC statistics */ + data += efx->type->update_stats(efx, data, NULL); + + /* Get software statistics */ + for (i = 0; i < EFX_ETHTOOL_SW_STAT_COUNT; i++) { + stat = &efx_sw_stat_desc[i]; + switch (stat->source) { + case EFX_ETHTOOL_STAT_SOURCE_nic: + data[i] = stat->get_stat((void *)efx + stat->offset); + break; + case EFX_ETHTOOL_STAT_SOURCE_channel: + data[i] = 0; + efx_for_each_channel(channel, efx) + data[i] += stat->get_stat((void *)channel + + stat->offset); + break; + case EFX_ETHTOOL_STAT_SOURCE_tx_queue: + data[i] = 0; + efx_for_each_channel(channel, efx) { + efx_for_each_channel_tx_queue(tx_queue, channel) + data[i] += + stat->get_stat((void *)tx_queue + + stat->offset); + } + break; + } + } + data += EFX_ETHTOOL_SW_STAT_COUNT; + + spin_unlock_bh(&efx->stats_lock); + + efx_for_each_channel(channel, efx) { + if (efx_channel_has_tx_queues(channel)) { + *data = 0; + efx_for_each_channel_tx_queue(tx_queue, channel) { + *data += tx_queue->tx_packets; + } + data++; + } + } + efx_for_each_channel(channel, efx) { + if (efx_channel_has_rx_queue(channel)) { + *data = 0; + efx_for_each_channel_rx_queue(rx_queue, channel) { + *data += rx_queue->rx_packets; + } + data++; + } + } + if (efx->xdp_tx_queue_count && efx->xdp_tx_queues) { + int xdp; + + for (xdp = 0; xdp < efx->xdp_tx_queue_count; xdp++) { + data[0] = efx->xdp_tx_queues[xdp]->tx_packets; + data++; + } + } + + efx_ptp_update_stats(efx, data); +} diff --git a/drivers/net/ethernet/sfc/ethtool_common.h b/drivers/net/ethernet/sfc/ethtool_common.h new file mode 100644 index 000000000000..fa624313f330 --- /dev/null +++ b/drivers/net/ethernet/sfc/ethtool_common.h @@ -0,0 +1,30 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_ETHTOOL_COMMON_H +#define EFX_ETHTOOL_COMMON_H + +void efx_ethtool_get_drvinfo(struct net_device *net_dev, + struct ethtool_drvinfo *info); +u32 efx_ethtool_get_msglevel(struct net_device *net_dev); +void efx_ethtool_set_msglevel(struct net_device *net_dev, u32 msg_enable); +void efx_ethtool_get_pauseparam(struct net_device *net_dev, + struct ethtool_pauseparam *pause); +int efx_ethtool_fill_self_tests(struct efx_nic *efx, + struct efx_self_tests *tests, + u8 *strings, u64 *data); +int efx_ethtool_get_sset_count(struct net_device *net_dev, int string_set); +void efx_ethtool_get_strings(struct net_device *net_dev, u32 string_set, + u8 *strings); +void efx_ethtool_get_stats(struct net_device *net_dev, + struct ethtool_stats *stats __attribute__ ((unused)), + u64 *data); + +#endif diff --git a/drivers/net/ethernet/sfc/falcon/efx.c b/drivers/net/ethernet/sfc/falcon/efx.c index 53ae9faeb4c3..42bcd34fc508 100644 --- a/drivers/net/ethernet/sfc/falcon/efx.c +++ b/drivers/net/ethernet/sfc/falcon/efx.c @@ -2108,7 +2108,7 @@ static void ef4_net_stats(struct net_device *net_dev, } /* Context: netif_tx_lock held, BHs disabled. */ -static void ef4_watchdog(struct net_device *net_dev) +static void ef4_watchdog(struct net_device *net_dev, unsigned int txqueue) { struct ef4_nic *efx = netdev_priv(net_dev); diff --git a/drivers/net/ethernet/sfc/farch.c b/drivers/net/ethernet/sfc/farch.c index eedd32e2bfcb..dbbb898adddb 100644 --- a/drivers/net/ethernet/sfc/farch.c +++ b/drivers/net/ethernet/sfc/farch.c @@ -15,6 +15,7 @@ #include "net_driver.h" #include "bitfield.h" #include "efx.h" +#include "rx_common.h" #include "nic.h" #include "farch_regs.h" #include "sriov.h" diff --git a/drivers/net/ethernet/sfc/mcdi.h b/drivers/net/ethernet/sfc/mcdi.h index 9081f84a2604..54a45010b576 100644 --- a/drivers/net/ethernet/sfc/mcdi.h +++ b/drivers/net/ethernet/sfc/mcdi.h @@ -346,11 +346,8 @@ int efx_mcdi_flush_rxqs(struct efx_nic *efx); int efx_mcdi_port_probe(struct efx_nic *efx); void efx_mcdi_port_remove(struct efx_nic *efx); int efx_mcdi_port_reconfigure(struct efx_nic *efx); -int efx_mcdi_port_get_number(struct efx_nic *efx); u32 efx_mcdi_phy_get_caps(struct efx_nic *efx); void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev); -int efx_mcdi_set_mac(struct efx_nic *efx); -#define EFX_MC_STATS_GENERATION_INVALID ((__force __le64)(-1)) void efx_mcdi_mac_start_stats(struct efx_nic *efx); void efx_mcdi_mac_stop_stats(struct efx_nic *efx); void efx_mcdi_mac_pull_stats(struct efx_nic *efx); diff --git a/drivers/net/ethernet/sfc/mcdi_filters.c b/drivers/net/ethernet/sfc/mcdi_filters.c new file mode 100644 index 000000000000..4310ae5bd898 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_filters.c @@ -0,0 +1,2270 @@ +#include "mcdi_filters.h" +#include "mcdi.h" +#include "nic.h" +#include "rx_common.h" + +/* The maximum size of a shared RSS context */ +/* TODO: this should really be from the mcdi protocol export */ +#define EFX_EF10_MAX_SHARED_RSS_CONTEXT_SIZE 64UL + +#define EFX_EF10_FILTER_ID_INVALID 0xffff + +/* An arbitrary search limit for the software hash table */ +#define EFX_EF10_FILTER_SEARCH_LIMIT 200 + +static struct efx_filter_spec * +efx_mcdi_filter_entry_spec(const struct efx_mcdi_filter_table *table, + unsigned int filter_idx) +{ + return (struct efx_filter_spec *)(table->entry[filter_idx].spec & + ~EFX_EF10_FILTER_FLAGS); +} + +static unsigned int +efx_mcdi_filter_entry_flags(const struct efx_mcdi_filter_table *table, + unsigned int filter_idx) +{ + return table->entry[filter_idx].spec & EFX_EF10_FILTER_FLAGS; +} + +static u32 efx_mcdi_filter_get_unsafe_id(u32 filter_id) +{ + WARN_ON_ONCE(filter_id == EFX_EF10_FILTER_ID_INVALID); + return filter_id & (EFX_MCDI_FILTER_TBL_ROWS - 1); +} + +static unsigned int efx_mcdi_filter_get_unsafe_pri(u32 filter_id) +{ + return filter_id / (EFX_MCDI_FILTER_TBL_ROWS * 2); +} + +static u32 efx_mcdi_filter_make_filter_id(unsigned int pri, u16 idx) +{ + return pri * EFX_MCDI_FILTER_TBL_ROWS * 2 + idx; +} + +/* + * Decide whether a filter should be exclusive or else should allow + * delivery to additional recipients. Currently we decide that + * filters for specific local unicast MAC and IP addresses are + * exclusive. + */ +static bool efx_mcdi_filter_is_exclusive(const struct efx_filter_spec *spec) +{ + if (spec->match_flags & EFX_FILTER_MATCH_LOC_MAC && + !is_multicast_ether_addr(spec->loc_mac)) + return true; + + if ((spec->match_flags & + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { + if (spec->ether_type == htons(ETH_P_IP) && + !ipv4_is_multicast(spec->loc_host[0])) + return true; + if (spec->ether_type == htons(ETH_P_IPV6) && + ((const u8 *)spec->loc_host)[0] != 0xff) + return true; + } + + return false; +} + +static void +efx_mcdi_filter_set_entry(struct efx_mcdi_filter_table *table, + unsigned int filter_idx, + const struct efx_filter_spec *spec, + unsigned int flags) +{ + table->entry[filter_idx].spec = (unsigned long)spec | flags; +} + +static void +efx_mcdi_filter_push_prep_set_match_fields(struct efx_nic *efx, + const struct efx_filter_spec *spec, + efx_dword_t *inbuf) +{ + enum efx_encap_type encap_type = efx_filter_get_encap_type(spec); + u32 match_fields = 0, uc_match, mc_match; + + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, + efx_mcdi_filter_is_exclusive(spec) ? + MC_CMD_FILTER_OP_IN_OP_INSERT : + MC_CMD_FILTER_OP_IN_OP_SUBSCRIBE); + + /* + * Convert match flags and values. Unlike almost + * everything else in MCDI, these fields are in + * network byte order. + */ +#define COPY_VALUE(value, mcdi_field) \ + do { \ + match_fields |= \ + 1 << MC_CMD_FILTER_OP_IN_MATCH_ ## \ + mcdi_field ## _LBN; \ + BUILD_BUG_ON( \ + MC_CMD_FILTER_OP_IN_ ## mcdi_field ## _LEN < \ + sizeof(value)); \ + memcpy(MCDI_PTR(inbuf, FILTER_OP_IN_ ## mcdi_field), \ + &value, sizeof(value)); \ + } while (0) +#define COPY_FIELD(gen_flag, gen_field, mcdi_field) \ + if (spec->match_flags & EFX_FILTER_MATCH_ ## gen_flag) { \ + COPY_VALUE(spec->gen_field, mcdi_field); \ + } + /* + * Handle encap filters first. They will always be mismatch + * (unknown UC or MC) filters + */ + if (encap_type) { + /* + * ether_type and outer_ip_proto need to be variables + * because COPY_VALUE wants to memcpy them + */ + __be16 ether_type = + htons(encap_type & EFX_ENCAP_FLAG_IPV6 ? + ETH_P_IPV6 : ETH_P_IP); + u8 vni_type = MC_CMD_FILTER_OP_EXT_IN_VNI_TYPE_GENEVE; + u8 outer_ip_proto; + + switch (encap_type & EFX_ENCAP_TYPES_MASK) { + case EFX_ENCAP_TYPE_VXLAN: + vni_type = MC_CMD_FILTER_OP_EXT_IN_VNI_TYPE_VXLAN; + /* fallthrough */ + case EFX_ENCAP_TYPE_GENEVE: + COPY_VALUE(ether_type, ETHER_TYPE); + outer_ip_proto = IPPROTO_UDP; + COPY_VALUE(outer_ip_proto, IP_PROTO); + /* + * We always need to set the type field, even + * though we're not matching on the TNI. + */ + MCDI_POPULATE_DWORD_1(inbuf, + FILTER_OP_EXT_IN_VNI_OR_VSID, + FILTER_OP_EXT_IN_VNI_TYPE, + vni_type); + break; + case EFX_ENCAP_TYPE_NVGRE: + COPY_VALUE(ether_type, ETHER_TYPE); + outer_ip_proto = IPPROTO_GRE; + COPY_VALUE(outer_ip_proto, IP_PROTO); + break; + default: + WARN_ON(1); + } + + uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_UCAST_DST_LBN; + mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_MCAST_DST_LBN; + } else { + uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_UCAST_DST_LBN; + mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_MCAST_DST_LBN; + } + + if (spec->match_flags & EFX_FILTER_MATCH_LOC_MAC_IG) + match_fields |= + is_multicast_ether_addr(spec->loc_mac) ? + 1 << mc_match : + 1 << uc_match; + COPY_FIELD(REM_HOST, rem_host, SRC_IP); + COPY_FIELD(LOC_HOST, loc_host, DST_IP); + COPY_FIELD(REM_MAC, rem_mac, SRC_MAC); + COPY_FIELD(REM_PORT, rem_port, SRC_PORT); + COPY_FIELD(LOC_MAC, loc_mac, DST_MAC); + COPY_FIELD(LOC_PORT, loc_port, DST_PORT); + COPY_FIELD(ETHER_TYPE, ether_type, ETHER_TYPE); + COPY_FIELD(INNER_VID, inner_vid, INNER_VLAN); + COPY_FIELD(OUTER_VID, outer_vid, OUTER_VLAN); + COPY_FIELD(IP_PROTO, ip_proto, IP_PROTO); +#undef COPY_FIELD +#undef COPY_VALUE + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_MATCH_FIELDS, + match_fields); +} + +static void efx_mcdi_filter_push_prep(struct efx_nic *efx, + const struct efx_filter_spec *spec, + efx_dword_t *inbuf, u64 handle, + struct efx_rss_context *ctx, + bool replacing) +{ + struct efx_ef10_nic_data *nic_data = efx->nic_data; + u32 flags = spec->flags; + + memset(inbuf, 0, MC_CMD_FILTER_OP_EXT_IN_LEN); + + /* If RSS filter, caller better have given us an RSS context */ + if (flags & EFX_FILTER_FLAG_RX_RSS) { + /* + * We don't have the ability to return an error, so we'll just + * log a warning and disable RSS for the filter. + */ + if (WARN_ON_ONCE(!ctx)) + flags &= ~EFX_FILTER_FLAG_RX_RSS; + else if (WARN_ON_ONCE(ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID)) + flags &= ~EFX_FILTER_FLAG_RX_RSS; + } + + if (replacing) { + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, + MC_CMD_FILTER_OP_IN_OP_REPLACE); + MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, handle); + } else { + efx_mcdi_filter_push_prep_set_match_fields(efx, spec, inbuf); + } + + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_PORT_ID, nic_data->vport_id); + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_DEST, + spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP ? + MC_CMD_FILTER_OP_IN_RX_DEST_DROP : + MC_CMD_FILTER_OP_IN_RX_DEST_HOST); + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_TX_DOMAIN, 0); + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_TX_DEST, + MC_CMD_FILTER_OP_IN_TX_DEST_DEFAULT); + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_QUEUE, + spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP ? + 0 : spec->dmaq_id); + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_MODE, + (flags & EFX_FILTER_FLAG_RX_RSS) ? + MC_CMD_FILTER_OP_IN_RX_MODE_RSS : + MC_CMD_FILTER_OP_IN_RX_MODE_SIMPLE); + if (flags & EFX_FILTER_FLAG_RX_RSS) + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_RX_CONTEXT, ctx->context_id); +} + +static int efx_mcdi_filter_push(struct efx_nic *efx, + const struct efx_filter_spec *spec, u64 *handle, + struct efx_rss_context *ctx, bool replacing) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); + MCDI_DECLARE_BUF(outbuf, MC_CMD_FILTER_OP_EXT_OUT_LEN); + size_t outlen; + int rc; + + efx_mcdi_filter_push_prep(efx, spec, inbuf, *handle, ctx, replacing); + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + if (rc && spec->priority != EFX_FILTER_PRI_HINT) + efx_mcdi_display_error(efx, MC_CMD_FILTER_OP, sizeof(inbuf), + outbuf, outlen, rc); + if (rc == 0) + *handle = MCDI_QWORD(outbuf, FILTER_OP_OUT_HANDLE); + if (rc == -ENOSPC) + rc = -EBUSY; /* to match efx_farch_filter_insert() */ + return rc; +} + +static u32 efx_mcdi_filter_mcdi_flags_from_spec(const struct efx_filter_spec *spec) +{ + enum efx_encap_type encap_type = efx_filter_get_encap_type(spec); + unsigned int match_flags = spec->match_flags; + unsigned int uc_match, mc_match; + u32 mcdi_flags = 0; + +#define MAP_FILTER_TO_MCDI_FLAG(gen_flag, mcdi_field, encap) { \ + unsigned int old_match_flags = match_flags; \ + match_flags &= ~EFX_FILTER_MATCH_ ## gen_flag; \ + if (match_flags != old_match_flags) \ + mcdi_flags |= \ + (1 << ((encap) ? \ + MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_ ## \ + mcdi_field ## _LBN : \ + MC_CMD_FILTER_OP_EXT_IN_MATCH_ ##\ + mcdi_field ## _LBN)); \ + } + /* inner or outer based on encap type */ + MAP_FILTER_TO_MCDI_FLAG(REM_HOST, SRC_IP, encap_type); + MAP_FILTER_TO_MCDI_FLAG(LOC_HOST, DST_IP, encap_type); + MAP_FILTER_TO_MCDI_FLAG(REM_MAC, SRC_MAC, encap_type); + MAP_FILTER_TO_MCDI_FLAG(REM_PORT, SRC_PORT, encap_type); + MAP_FILTER_TO_MCDI_FLAG(LOC_MAC, DST_MAC, encap_type); + MAP_FILTER_TO_MCDI_FLAG(LOC_PORT, DST_PORT, encap_type); + MAP_FILTER_TO_MCDI_FLAG(ETHER_TYPE, ETHER_TYPE, encap_type); + MAP_FILTER_TO_MCDI_FLAG(IP_PROTO, IP_PROTO, encap_type); + /* always outer */ + MAP_FILTER_TO_MCDI_FLAG(INNER_VID, INNER_VLAN, false); + MAP_FILTER_TO_MCDI_FLAG(OUTER_VID, OUTER_VLAN, false); +#undef MAP_FILTER_TO_MCDI_FLAG + + /* special handling for encap type, and mismatch */ + if (encap_type) { + match_flags &= ~EFX_FILTER_MATCH_ENCAP_TYPE; + mcdi_flags |= + (1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ETHER_TYPE_LBN); + mcdi_flags |= (1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_IP_PROTO_LBN); + + uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_UCAST_DST_LBN; + mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_IFRM_UNKNOWN_MCAST_DST_LBN; + } else { + uc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_UCAST_DST_LBN; + mc_match = MC_CMD_FILTER_OP_EXT_IN_MATCH_UNKNOWN_MCAST_DST_LBN; + } + + if (match_flags & EFX_FILTER_MATCH_LOC_MAC_IG) { + match_flags &= ~EFX_FILTER_MATCH_LOC_MAC_IG; + mcdi_flags |= + is_multicast_ether_addr(spec->loc_mac) ? + 1 << mc_match : + 1 << uc_match; + } + + /* Did we map them all? */ + WARN_ON_ONCE(match_flags); + + return mcdi_flags; +} + +static int efx_mcdi_filter_pri(struct efx_mcdi_filter_table *table, + const struct efx_filter_spec *spec) +{ + u32 mcdi_flags = efx_mcdi_filter_mcdi_flags_from_spec(spec); + unsigned int match_pri; + + for (match_pri = 0; + match_pri < table->rx_match_count; + match_pri++) + if (table->rx_match_mcdi_flags[match_pri] == mcdi_flags) + return match_pri; + + return -EPROTONOSUPPORT; +} + +static s32 efx_mcdi_filter_insert_locked(struct efx_nic *efx, + struct efx_filter_spec *spec, + bool replace_equal) +{ + DECLARE_BITMAP(mc_rem_map, EFX_EF10_FILTER_SEARCH_LIMIT); + struct efx_ef10_nic_data *nic_data = efx->nic_data; + struct efx_mcdi_filter_table *table; + struct efx_filter_spec *saved_spec; + struct efx_rss_context *ctx = NULL; + unsigned int match_pri, hash; + unsigned int priv_flags; + bool rss_locked = false; + bool replacing = false; + unsigned int depth, i; + int ins_index = -1; + DEFINE_WAIT(wait); + bool is_mc_recip; + s32 rc; + + WARN_ON(!rwsem_is_locked(&efx->filter_sem)); + table = efx->filter_state; + down_write(&table->lock); + + /* For now, only support RX filters */ + if ((spec->flags & (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX)) != + EFX_FILTER_FLAG_RX) { + rc = -EINVAL; + goto out_unlock; + } + + rc = efx_mcdi_filter_pri(table, spec); + if (rc < 0) + goto out_unlock; + match_pri = rc; + + hash = efx_filter_spec_hash(spec); + is_mc_recip = efx_filter_is_mc_recipient(spec); + if (is_mc_recip) + bitmap_zero(mc_rem_map, EFX_EF10_FILTER_SEARCH_LIMIT); + + if (spec->flags & EFX_FILTER_FLAG_RX_RSS) { + mutex_lock(&efx->rss_lock); + rss_locked = true; + if (spec->rss_context) + ctx = efx_find_rss_context_entry(efx, spec->rss_context); + else + ctx = &efx->rss_context; + if (!ctx) { + rc = -ENOENT; + goto out_unlock; + } + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { + rc = -EOPNOTSUPP; + goto out_unlock; + } + } + + /* Find any existing filters with the same match tuple or + * else a free slot to insert at. + */ + for (depth = 1; depth < EFX_EF10_FILTER_SEARCH_LIMIT; depth++) { + i = (hash + depth) & (EFX_MCDI_FILTER_TBL_ROWS - 1); + saved_spec = efx_mcdi_filter_entry_spec(table, i); + + if (!saved_spec) { + if (ins_index < 0) + ins_index = i; + } else if (efx_filter_spec_equal(spec, saved_spec)) { + if (spec->priority < saved_spec->priority && + spec->priority != EFX_FILTER_PRI_AUTO) { + rc = -EPERM; + goto out_unlock; + } + if (!is_mc_recip) { + /* This is the only one */ + if (spec->priority == + saved_spec->priority && + !replace_equal) { + rc = -EEXIST; + goto out_unlock; + } + ins_index = i; + break; + } else if (spec->priority > + saved_spec->priority || + (spec->priority == + saved_spec->priority && + replace_equal)) { + if (ins_index < 0) + ins_index = i; + else + __set_bit(depth, mc_rem_map); + } + } + } + + /* Once we reach the maximum search depth, use the first suitable + * slot, or return -EBUSY if there was none + */ + if (ins_index < 0) { + rc = -EBUSY; + goto out_unlock; + } + + /* Create a software table entry if necessary. */ + saved_spec = efx_mcdi_filter_entry_spec(table, ins_index); + if (saved_spec) { + if (spec->priority == EFX_FILTER_PRI_AUTO && + saved_spec->priority >= EFX_FILTER_PRI_AUTO) { + /* Just make sure it won't be removed */ + if (saved_spec->priority > EFX_FILTER_PRI_AUTO) + saved_spec->flags |= EFX_FILTER_FLAG_RX_OVER_AUTO; + table->entry[ins_index].spec &= + ~EFX_EF10_FILTER_FLAG_AUTO_OLD; + rc = ins_index; + goto out_unlock; + } + replacing = true; + priv_flags = efx_mcdi_filter_entry_flags(table, ins_index); + } else { + saved_spec = kmalloc(sizeof(*spec), GFP_ATOMIC); + if (!saved_spec) { + rc = -ENOMEM; + goto out_unlock; + } + *saved_spec = *spec; + priv_flags = 0; + } + efx_mcdi_filter_set_entry(table, ins_index, saved_spec, priv_flags); + + /* Actually insert the filter on the HW */ + rc = efx_mcdi_filter_push(efx, spec, &table->entry[ins_index].handle, + ctx, replacing); + + if (rc == -EINVAL && nic_data->must_realloc_vis) + /* The MC rebooted under us, causing it to reject our filter + * insertion as pointing to an invalid VI (spec->dmaq_id). + */ + rc = -EAGAIN; + + /* Finalise the software table entry */ + if (rc == 0) { + if (replacing) { + /* Update the fields that may differ */ + if (saved_spec->priority == EFX_FILTER_PRI_AUTO) + saved_spec->flags |= + EFX_FILTER_FLAG_RX_OVER_AUTO; + saved_spec->priority = spec->priority; + saved_spec->flags &= EFX_FILTER_FLAG_RX_OVER_AUTO; + saved_spec->flags |= spec->flags; + saved_spec->rss_context = spec->rss_context; + saved_spec->dmaq_id = spec->dmaq_id; + } + } else if (!replacing) { + kfree(saved_spec); + saved_spec = NULL; + } else { + /* We failed to replace, so the old filter is still present. + * Roll back the software table to reflect this. In fact the + * efx_mcdi_filter_set_entry() call below will do the right + * thing, so nothing extra is needed here. + */ + } + efx_mcdi_filter_set_entry(table, ins_index, saved_spec, priv_flags); + + /* Remove and finalise entries for lower-priority multicast + * recipients + */ + if (is_mc_recip) { + MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); + unsigned int depth, i; + + memset(inbuf, 0, sizeof(inbuf)); + + for (depth = 0; depth < EFX_EF10_FILTER_SEARCH_LIMIT; depth++) { + if (!test_bit(depth, mc_rem_map)) + continue; + + i = (hash + depth) & (EFX_MCDI_FILTER_TBL_ROWS - 1); + saved_spec = efx_mcdi_filter_entry_spec(table, i); + priv_flags = efx_mcdi_filter_entry_flags(table, i); + + if (rc == 0) { + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, + MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); + MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, + table->entry[i].handle); + rc = efx_mcdi_rpc(efx, MC_CMD_FILTER_OP, + inbuf, sizeof(inbuf), + NULL, 0, NULL); + } + + if (rc == 0) { + kfree(saved_spec); + saved_spec = NULL; + priv_flags = 0; + } + efx_mcdi_filter_set_entry(table, i, saved_spec, + priv_flags); + } + } + + /* If successful, return the inserted filter ID */ + if (rc == 0) + rc = efx_mcdi_filter_make_filter_id(match_pri, ins_index); + +out_unlock: + if (rss_locked) + mutex_unlock(&efx->rss_lock); + up_write(&table->lock); + return rc; +} + +s32 efx_mcdi_filter_insert(struct efx_nic *efx, struct efx_filter_spec *spec, + bool replace_equal) +{ + s32 ret; + + down_read(&efx->filter_sem); + ret = efx_mcdi_filter_insert_locked(efx, spec, replace_equal); + up_read(&efx->filter_sem); + + return ret; +} + +/* + * Remove a filter. + * If !by_index, remove by ID + * If by_index, remove by index + * Filter ID may come from userland and must be range-checked. + * Caller must hold efx->filter_sem for read, and efx->filter_state->lock + * for write. + */ +static int efx_mcdi_filter_remove_internal(struct efx_nic *efx, + unsigned int priority_mask, + u32 filter_id, bool by_index) +{ + unsigned int filter_idx = efx_mcdi_filter_get_unsafe_id(filter_id); + struct efx_mcdi_filter_table *table = efx->filter_state; + MCDI_DECLARE_BUF(inbuf, + MC_CMD_FILTER_OP_IN_HANDLE_OFST + + MC_CMD_FILTER_OP_IN_HANDLE_LEN); + struct efx_filter_spec *spec; + DEFINE_WAIT(wait); + int rc; + + spec = efx_mcdi_filter_entry_spec(table, filter_idx); + if (!spec || + (!by_index && + efx_mcdi_filter_pri(table, spec) != + efx_mcdi_filter_get_unsafe_pri(filter_id))) + return -ENOENT; + + if (spec->flags & EFX_FILTER_FLAG_RX_OVER_AUTO && + priority_mask == (1U << EFX_FILTER_PRI_AUTO)) { + /* Just remove flags */ + spec->flags &= ~EFX_FILTER_FLAG_RX_OVER_AUTO; + table->entry[filter_idx].spec &= ~EFX_EF10_FILTER_FLAG_AUTO_OLD; + return 0; + } + + if (!(priority_mask & (1U << spec->priority))) + return -ENOENT; + + if (spec->flags & EFX_FILTER_FLAG_RX_OVER_AUTO) { + /* Reset to an automatic filter */ + + struct efx_filter_spec new_spec = *spec; + + new_spec.priority = EFX_FILTER_PRI_AUTO; + new_spec.flags = (EFX_FILTER_FLAG_RX | + (efx_rss_active(&efx->rss_context) ? + EFX_FILTER_FLAG_RX_RSS : 0)); + new_spec.dmaq_id = 0; + new_spec.rss_context = 0; + rc = efx_mcdi_filter_push(efx, &new_spec, + &table->entry[filter_idx].handle, + &efx->rss_context, + true); + + if (rc == 0) + *spec = new_spec; + } else { + /* Really remove the filter */ + + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, + efx_mcdi_filter_is_exclusive(spec) ? + MC_CMD_FILTER_OP_IN_OP_REMOVE : + MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); + MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, + table->entry[filter_idx].handle); + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, + inbuf, sizeof(inbuf), NULL, 0, NULL); + + if ((rc == 0) || (rc == -ENOENT)) { + /* Filter removed OK or didn't actually exist */ + kfree(spec); + efx_mcdi_filter_set_entry(table, filter_idx, NULL, 0); + } else { + efx_mcdi_display_error(efx, MC_CMD_FILTER_OP, + MC_CMD_FILTER_OP_EXT_IN_LEN, + NULL, 0, rc); + } + } + + return rc; +} + +/* Remove filters that weren't renewed. */ +static void efx_mcdi_filter_remove_old(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + int remove_failed = 0; + int remove_noent = 0; + int rc; + int i; + + down_write(&table->lock); + for (i = 0; i < EFX_MCDI_FILTER_TBL_ROWS; i++) { + if (READ_ONCE(table->entry[i].spec) & + EFX_EF10_FILTER_FLAG_AUTO_OLD) { + rc = efx_mcdi_filter_remove_internal(efx, + 1U << EFX_FILTER_PRI_AUTO, i, true); + if (rc == -ENOENT) + remove_noent++; + else if (rc) + remove_failed++; + } + } + up_write(&table->lock); + + if (remove_failed) + netif_info(efx, drv, efx->net_dev, + "%s: failed to remove %d filters\n", + __func__, remove_failed); + if (remove_noent) + netif_info(efx, drv, efx->net_dev, + "%s: failed to remove %d non-existent filters\n", + __func__, remove_noent); +} + +int efx_mcdi_filter_remove_safe(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 filter_id) +{ + struct efx_mcdi_filter_table *table; + int rc; + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_write(&table->lock); + rc = efx_mcdi_filter_remove_internal(efx, 1U << priority, filter_id, + false); + up_write(&table->lock); + up_read(&efx->filter_sem); + return rc; +} + +/* Caller must hold efx->filter_sem for read */ +static void efx_mcdi_filter_remove_unsafe(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 filter_id) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + + if (filter_id == EFX_EF10_FILTER_ID_INVALID) + return; + + down_write(&table->lock); + efx_mcdi_filter_remove_internal(efx, 1U << priority, filter_id, + true); + up_write(&table->lock); +} + +int efx_mcdi_filter_get_safe(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 filter_id, struct efx_filter_spec *spec) +{ + unsigned int filter_idx = efx_mcdi_filter_get_unsafe_id(filter_id); + const struct efx_filter_spec *saved_spec; + struct efx_mcdi_filter_table *table; + int rc; + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_read(&table->lock); + saved_spec = efx_mcdi_filter_entry_spec(table, filter_idx); + if (saved_spec && saved_spec->priority == priority && + efx_mcdi_filter_pri(table, saved_spec) == + efx_mcdi_filter_get_unsafe_pri(filter_id)) { + *spec = *saved_spec; + rc = 0; + } else { + rc = -ENOENT; + } + up_read(&table->lock); + up_read(&efx->filter_sem); + return rc; +} + +static int efx_mcdi_filter_insert_addr_list(struct efx_nic *efx, + struct efx_mcdi_filter_vlan *vlan, + bool multicast, bool rollback) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_mcdi_dev_addr *addr_list; + enum efx_filter_flags filter_flags; + struct efx_filter_spec spec; + u8 baddr[ETH_ALEN]; + unsigned int i, j; + int addr_count; + u16 *ids; + int rc; + + if (multicast) { + addr_list = table->dev_mc_list; + addr_count = table->dev_mc_count; + ids = vlan->mc; + } else { + addr_list = table->dev_uc_list; + addr_count = table->dev_uc_count; + ids = vlan->uc; + } + + filter_flags = efx_rss_active(&efx->rss_context) ? EFX_FILTER_FLAG_RX_RSS : 0; + + /* Insert/renew filters */ + for (i = 0; i < addr_count; i++) { + EFX_WARN_ON_PARANOID(ids[i] != EFX_EF10_FILTER_ID_INVALID); + efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); + efx_filter_set_eth_local(&spec, vlan->vid, addr_list[i].addr); + rc = efx_mcdi_filter_insert_locked(efx, &spec, true); + if (rc < 0) { + if (rollback) { + netif_info(efx, drv, efx->net_dev, + "efx_mcdi_filter_insert failed rc=%d\n", + rc); + /* Fall back to promiscuous */ + for (j = 0; j < i; j++) { + efx_mcdi_filter_remove_unsafe( + efx, EFX_FILTER_PRI_AUTO, + ids[j]); + ids[j] = EFX_EF10_FILTER_ID_INVALID; + } + return rc; + } else { + /* keep invalid ID, and carry on */ + } + } else { + ids[i] = efx_mcdi_filter_get_unsafe_id(rc); + } + } + + if (multicast && rollback) { + /* Also need an Ethernet broadcast filter */ + EFX_WARN_ON_PARANOID(vlan->default_filters[EFX_EF10_BCAST] != + EFX_EF10_FILTER_ID_INVALID); + efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); + eth_broadcast_addr(baddr); + efx_filter_set_eth_local(&spec, vlan->vid, baddr); + rc = efx_mcdi_filter_insert_locked(efx, &spec, true); + if (rc < 0) { + netif_warn(efx, drv, efx->net_dev, + "Broadcast filter insert failed rc=%d\n", rc); + /* Fall back to promiscuous */ + for (j = 0; j < i; j++) { + efx_mcdi_filter_remove_unsafe( + efx, EFX_FILTER_PRI_AUTO, + ids[j]); + ids[j] = EFX_EF10_FILTER_ID_INVALID; + } + return rc; + } else { + vlan->default_filters[EFX_EF10_BCAST] = + efx_mcdi_filter_get_unsafe_id(rc); + } + } + + return 0; +} + +static int efx_mcdi_filter_insert_def(struct efx_nic *efx, + struct efx_mcdi_filter_vlan *vlan, + enum efx_encap_type encap_type, + bool multicast, bool rollback) +{ + struct efx_ef10_nic_data *nic_data = efx->nic_data; + enum efx_filter_flags filter_flags; + struct efx_filter_spec spec; + u8 baddr[ETH_ALEN]; + int rc; + u16 *id; + + filter_flags = efx_rss_active(&efx->rss_context) ? EFX_FILTER_FLAG_RX_RSS : 0; + + efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, filter_flags, 0); + + if (multicast) + efx_filter_set_mc_def(&spec); + else + efx_filter_set_uc_def(&spec); + + if (encap_type) { + if (nic_data->datapath_caps & + (1 << MC_CMD_GET_CAPABILITIES_OUT_VXLAN_NVGRE_LBN)) + efx_filter_set_encap_type(&spec, encap_type); + else + /* + * don't insert encap filters on non-supporting + * platforms. ID will be left as INVALID. + */ + return 0; + } + + if (vlan->vid != EFX_FILTER_VID_UNSPEC) + efx_filter_set_eth_local(&spec, vlan->vid, NULL); + + rc = efx_mcdi_filter_insert_locked(efx, &spec, true); + if (rc < 0) { + const char *um = multicast ? "Multicast" : "Unicast"; + const char *encap_name = ""; + const char *encap_ipv = ""; + + if ((encap_type & EFX_ENCAP_TYPES_MASK) == + EFX_ENCAP_TYPE_VXLAN) + encap_name = "VXLAN "; + else if ((encap_type & EFX_ENCAP_TYPES_MASK) == + EFX_ENCAP_TYPE_NVGRE) + encap_name = "NVGRE "; + else if ((encap_type & EFX_ENCAP_TYPES_MASK) == + EFX_ENCAP_TYPE_GENEVE) + encap_name = "GENEVE "; + if (encap_type & EFX_ENCAP_FLAG_IPV6) + encap_ipv = "IPv6 "; + else if (encap_type) + encap_ipv = "IPv4 "; + + /* + * unprivileged functions can't insert mismatch filters + * for encapsulated or unicast traffic, so downgrade + * those warnings to debug. + */ + netif_cond_dbg(efx, drv, efx->net_dev, + rc == -EPERM && (encap_type || !multicast), warn, + "%s%s%s mismatch filter insert failed rc=%d\n", + encap_name, encap_ipv, um, rc); + } else if (multicast) { + /* mapping from encap types to default filter IDs (multicast) */ + static enum efx_mcdi_filter_default_filters map[] = { + [EFX_ENCAP_TYPE_NONE] = EFX_EF10_MCDEF, + [EFX_ENCAP_TYPE_VXLAN] = EFX_EF10_VXLAN4_MCDEF, + [EFX_ENCAP_TYPE_NVGRE] = EFX_EF10_NVGRE4_MCDEF, + [EFX_ENCAP_TYPE_GENEVE] = EFX_EF10_GENEVE4_MCDEF, + [EFX_ENCAP_TYPE_VXLAN | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_VXLAN6_MCDEF, + [EFX_ENCAP_TYPE_NVGRE | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_NVGRE6_MCDEF, + [EFX_ENCAP_TYPE_GENEVE | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_GENEVE6_MCDEF, + }; + + /* quick bounds check (BCAST result impossible) */ + BUILD_BUG_ON(EFX_EF10_BCAST != 0); + if (encap_type >= ARRAY_SIZE(map) || map[encap_type] == 0) { + WARN_ON(1); + return -EINVAL; + } + /* then follow map */ + id = &vlan->default_filters[map[encap_type]]; + + EFX_WARN_ON_PARANOID(*id != EFX_EF10_FILTER_ID_INVALID); + *id = efx_mcdi_filter_get_unsafe_id(rc); + if (!nic_data->workaround_26807 && !encap_type) { + /* Also need an Ethernet broadcast filter */ + efx_filter_init_rx(&spec, EFX_FILTER_PRI_AUTO, + filter_flags, 0); + eth_broadcast_addr(baddr); + efx_filter_set_eth_local(&spec, vlan->vid, baddr); + rc = efx_mcdi_filter_insert_locked(efx, &spec, true); + if (rc < 0) { + netif_warn(efx, drv, efx->net_dev, + "Broadcast filter insert failed rc=%d\n", + rc); + if (rollback) { + /* Roll back the mc_def filter */ + efx_mcdi_filter_remove_unsafe( + efx, EFX_FILTER_PRI_AUTO, + *id); + *id = EFX_EF10_FILTER_ID_INVALID; + return rc; + } + } else { + EFX_WARN_ON_PARANOID( + vlan->default_filters[EFX_EF10_BCAST] != + EFX_EF10_FILTER_ID_INVALID); + vlan->default_filters[EFX_EF10_BCAST] = + efx_mcdi_filter_get_unsafe_id(rc); + } + } + rc = 0; + } else { + /* mapping from encap types to default filter IDs (unicast) */ + static enum efx_mcdi_filter_default_filters map[] = { + [EFX_ENCAP_TYPE_NONE] = EFX_EF10_UCDEF, + [EFX_ENCAP_TYPE_VXLAN] = EFX_EF10_VXLAN4_UCDEF, + [EFX_ENCAP_TYPE_NVGRE] = EFX_EF10_NVGRE4_UCDEF, + [EFX_ENCAP_TYPE_GENEVE] = EFX_EF10_GENEVE4_UCDEF, + [EFX_ENCAP_TYPE_VXLAN | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_VXLAN6_UCDEF, + [EFX_ENCAP_TYPE_NVGRE | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_NVGRE6_UCDEF, + [EFX_ENCAP_TYPE_GENEVE | EFX_ENCAP_FLAG_IPV6] = + EFX_EF10_GENEVE6_UCDEF, + }; + + /* quick bounds check (BCAST result impossible) */ + BUILD_BUG_ON(EFX_EF10_BCAST != 0); + if (encap_type >= ARRAY_SIZE(map) || map[encap_type] == 0) { + WARN_ON(1); + return -EINVAL; + } + /* then follow map */ + id = &vlan->default_filters[map[encap_type]]; + EFX_WARN_ON_PARANOID(*id != EFX_EF10_FILTER_ID_INVALID); + *id = rc; + rc = 0; + } + return rc; +} + +/* + * Caller must hold efx->filter_sem for read if race against + * efx_mcdi_filter_table_remove() is possible + */ +static void efx_mcdi_filter_vlan_sync_rx_mode(struct efx_nic *efx, + struct efx_mcdi_filter_vlan *vlan) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_ef10_nic_data *nic_data = efx->nic_data; + + /* + * Do not install unspecified VID if VLAN filtering is enabled. + * Do not install all specified VIDs if VLAN filtering is disabled. + */ + if ((vlan->vid == EFX_FILTER_VID_UNSPEC) == table->vlan_filter) + return; + + /* Insert/renew unicast filters */ + if (table->uc_promisc) { + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NONE, + false, false); + efx_mcdi_filter_insert_addr_list(efx, vlan, false, false); + } else { + /* + * If any of the filters failed to insert, fall back to + * promiscuous mode - add in the uc_def filter. But keep + * our individual unicast filters. + */ + if (efx_mcdi_filter_insert_addr_list(efx, vlan, false, false)) + efx_mcdi_filter_insert_def(efx, vlan, + EFX_ENCAP_TYPE_NONE, + false, false); + } + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN, + false, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN | + EFX_ENCAP_FLAG_IPV6, + false, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE, + false, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE | + EFX_ENCAP_FLAG_IPV6, + false, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE, + false, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE | + EFX_ENCAP_FLAG_IPV6, + false, false); + + /* + * Insert/renew multicast filters + * + * If changing promiscuous state with cascaded multicast filters, remove + * old filters first, so that packets are dropped rather than duplicated + */ + if (nic_data->workaround_26807 && + table->mc_promisc_last != table->mc_promisc) + efx_mcdi_filter_remove_old(efx); + if (table->mc_promisc) { + if (nic_data->workaround_26807) { + /* + * If we failed to insert promiscuous filters, rollback + * and fall back to individual multicast filters + */ + if (efx_mcdi_filter_insert_def(efx, vlan, + EFX_ENCAP_TYPE_NONE, + true, true)) { + /* Changing promisc state, so remove old filters */ + efx_mcdi_filter_remove_old(efx); + efx_mcdi_filter_insert_addr_list(efx, vlan, + true, false); + } + } else { + /* + * If we failed to insert promiscuous filters, don't + * rollback. Regardless, also insert the mc_list, + * unless it's incomplete due to overflow + */ + efx_mcdi_filter_insert_def(efx, vlan, + EFX_ENCAP_TYPE_NONE, + true, false); + if (!table->mc_overflow) + efx_mcdi_filter_insert_addr_list(efx, vlan, + true, false); + } + } else { + /* + * If any filters failed to insert, rollback and fall back to + * promiscuous mode - mc_def filter and maybe broadcast. If + * that fails, roll back again and insert as many of our + * individual multicast filters as we can. + */ + if (efx_mcdi_filter_insert_addr_list(efx, vlan, true, true)) { + /* Changing promisc state, so remove old filters */ + if (nic_data->workaround_26807) + efx_mcdi_filter_remove_old(efx); + if (efx_mcdi_filter_insert_def(efx, vlan, + EFX_ENCAP_TYPE_NONE, + true, true)) + efx_mcdi_filter_insert_addr_list(efx, vlan, + true, false); + } + } + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN, + true, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_VXLAN | + EFX_ENCAP_FLAG_IPV6, + true, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE, + true, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_NVGRE | + EFX_ENCAP_FLAG_IPV6, + true, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE, + true, false); + efx_mcdi_filter_insert_def(efx, vlan, EFX_ENCAP_TYPE_GENEVE | + EFX_ENCAP_FLAG_IPV6, + true, false); +} + +int efx_mcdi_filter_clear_rx(struct efx_nic *efx, + enum efx_filter_priority priority) +{ + struct efx_mcdi_filter_table *table; + unsigned int priority_mask; + unsigned int i; + int rc; + + priority_mask = (((1U << (priority + 1)) - 1) & + ~(1U << EFX_FILTER_PRI_AUTO)); + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_write(&table->lock); + for (i = 0; i < EFX_MCDI_FILTER_TBL_ROWS; i++) { + rc = efx_mcdi_filter_remove_internal(efx, priority_mask, + i, true); + if (rc && rc != -ENOENT) + break; + rc = 0; + } + + up_write(&table->lock); + up_read(&efx->filter_sem); + return rc; +} + +u32 efx_mcdi_filter_count_rx_used(struct efx_nic *efx, + enum efx_filter_priority priority) +{ + struct efx_mcdi_filter_table *table; + unsigned int filter_idx; + s32 count = 0; + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_read(&table->lock); + for (filter_idx = 0; filter_idx < EFX_MCDI_FILTER_TBL_ROWS; filter_idx++) { + if (table->entry[filter_idx].spec && + efx_mcdi_filter_entry_spec(table, filter_idx)->priority == + priority) + ++count; + } + up_read(&table->lock); + up_read(&efx->filter_sem); + return count; +} + +u32 efx_mcdi_filter_get_rx_id_limit(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + + return table->rx_match_count * EFX_MCDI_FILTER_TBL_ROWS * 2; +} + +s32 efx_mcdi_filter_get_rx_ids(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 *buf, u32 size) +{ + struct efx_mcdi_filter_table *table; + struct efx_filter_spec *spec; + unsigned int filter_idx; + s32 count = 0; + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_read(&table->lock); + + for (filter_idx = 0; filter_idx < EFX_MCDI_FILTER_TBL_ROWS; filter_idx++) { + spec = efx_mcdi_filter_entry_spec(table, filter_idx); + if (spec && spec->priority == priority) { + if (count == size) { + count = -EMSGSIZE; + break; + } + buf[count++] = + efx_mcdi_filter_make_filter_id( + efx_mcdi_filter_pri(table, spec), + filter_idx); + } + } + up_read(&table->lock); + up_read(&efx->filter_sem); + return count; +} + +static int efx_mcdi_filter_match_flags_from_mcdi(bool encap, u32 mcdi_flags) +{ + int match_flags = 0; + +#define MAP_FLAG(gen_flag, mcdi_field) do { \ + u32 old_mcdi_flags = mcdi_flags; \ + mcdi_flags &= ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ ## \ + mcdi_field ## _LBN); \ + if (mcdi_flags != old_mcdi_flags) \ + match_flags |= EFX_FILTER_MATCH_ ## gen_flag; \ + } while (0) + + if (encap) { + /* encap filters must specify encap type */ + match_flags |= EFX_FILTER_MATCH_ENCAP_TYPE; + /* and imply ethertype and ip proto */ + mcdi_flags &= + ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_IP_PROTO_LBN); + mcdi_flags &= + ~(1 << MC_CMD_FILTER_OP_EXT_IN_MATCH_ETHER_TYPE_LBN); + /* VLAN tags refer to the outer packet */ + MAP_FLAG(INNER_VID, INNER_VLAN); + MAP_FLAG(OUTER_VID, OUTER_VLAN); + /* everything else refers to the inner packet */ + MAP_FLAG(LOC_MAC_IG, IFRM_UNKNOWN_UCAST_DST); + MAP_FLAG(LOC_MAC_IG, IFRM_UNKNOWN_MCAST_DST); + MAP_FLAG(REM_HOST, IFRM_SRC_IP); + MAP_FLAG(LOC_HOST, IFRM_DST_IP); + MAP_FLAG(REM_MAC, IFRM_SRC_MAC); + MAP_FLAG(REM_PORT, IFRM_SRC_PORT); + MAP_FLAG(LOC_MAC, IFRM_DST_MAC); + MAP_FLAG(LOC_PORT, IFRM_DST_PORT); + MAP_FLAG(ETHER_TYPE, IFRM_ETHER_TYPE); + MAP_FLAG(IP_PROTO, IFRM_IP_PROTO); + } else { + MAP_FLAG(LOC_MAC_IG, UNKNOWN_UCAST_DST); + MAP_FLAG(LOC_MAC_IG, UNKNOWN_MCAST_DST); + MAP_FLAG(REM_HOST, SRC_IP); + MAP_FLAG(LOC_HOST, DST_IP); + MAP_FLAG(REM_MAC, SRC_MAC); + MAP_FLAG(REM_PORT, SRC_PORT); + MAP_FLAG(LOC_MAC, DST_MAC); + MAP_FLAG(LOC_PORT, DST_PORT); + MAP_FLAG(ETHER_TYPE, ETHER_TYPE); + MAP_FLAG(INNER_VID, INNER_VLAN); + MAP_FLAG(OUTER_VID, OUTER_VLAN); + MAP_FLAG(IP_PROTO, IP_PROTO); + } +#undef MAP_FLAG + + /* Did we map them all? */ + if (mcdi_flags) + return -EINVAL; + + return match_flags; +} + +bool efx_mcdi_filter_match_supported(struct efx_mcdi_filter_table *table, + bool encap, + enum efx_filter_match_flags match_flags) +{ + unsigned int match_pri; + int mf; + + for (match_pri = 0; + match_pri < table->rx_match_count; + match_pri++) { + mf = efx_mcdi_filter_match_flags_from_mcdi(encap, + table->rx_match_mcdi_flags[match_pri]); + if (mf == match_flags) + return true; + } + + return false; +} + +static int +efx_mcdi_filter_table_probe_matches(struct efx_nic *efx, + struct efx_mcdi_filter_table *table, + bool encap) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_GET_PARSER_DISP_INFO_IN_LEN); + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PARSER_DISP_INFO_OUT_LENMAX); + unsigned int pd_match_pri, pd_match_count; + size_t outlen; + int rc; + + /* Find out which RX filter types are supported, and their priorities */ + MCDI_SET_DWORD(inbuf, GET_PARSER_DISP_INFO_IN_OP, + encap ? + MC_CMD_GET_PARSER_DISP_INFO_IN_OP_GET_SUPPORTED_ENCAP_RX_MATCHES : + MC_CMD_GET_PARSER_DISP_INFO_IN_OP_GET_SUPPORTED_RX_MATCHES); + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PARSER_DISP_INFO, + inbuf, sizeof(inbuf), outbuf, sizeof(outbuf), + &outlen); + if (rc) + return rc; + + pd_match_count = MCDI_VAR_ARRAY_LEN( + outlen, GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES); + + for (pd_match_pri = 0; pd_match_pri < pd_match_count; pd_match_pri++) { + u32 mcdi_flags = + MCDI_ARRAY_DWORD( + outbuf, + GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES, + pd_match_pri); + rc = efx_mcdi_filter_match_flags_from_mcdi(encap, mcdi_flags); + if (rc < 0) { + netif_dbg(efx, probe, efx->net_dev, + "%s: fw flags %#x pri %u not supported in driver\n", + __func__, mcdi_flags, pd_match_pri); + } else { + netif_dbg(efx, probe, efx->net_dev, + "%s: fw flags %#x pri %u supported as driver flags %#x pri %u\n", + __func__, mcdi_flags, pd_match_pri, + rc, table->rx_match_count); + table->rx_match_mcdi_flags[table->rx_match_count] = mcdi_flags; + table->rx_match_count++; + } + } + + return 0; +} + +int efx_mcdi_filter_table_probe(struct efx_nic *efx) +{ + struct efx_ef10_nic_data *nic_data = efx->nic_data; + struct net_device *net_dev = efx->net_dev; + struct efx_mcdi_filter_table *table; + struct efx_mcdi_filter_vlan *vlan; + int rc; + + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return -EINVAL; + + if (efx->filter_state) /* already probed */ + return 0; + + table = kzalloc(sizeof(*table), GFP_KERNEL); + if (!table) + return -ENOMEM; + + table->rx_match_count = 0; + rc = efx_mcdi_filter_table_probe_matches(efx, table, false); + if (rc) + goto fail; + if (nic_data->datapath_caps & + (1 << MC_CMD_GET_CAPABILITIES_OUT_VXLAN_NVGRE_LBN)) + rc = efx_mcdi_filter_table_probe_matches(efx, table, true); + if (rc) + goto fail; + if ((efx_supported_features(efx) & NETIF_F_HW_VLAN_CTAG_FILTER) && + !(efx_mcdi_filter_match_supported(table, false, + (EFX_FILTER_MATCH_OUTER_VID | EFX_FILTER_MATCH_LOC_MAC)) && + efx_mcdi_filter_match_supported(table, false, + (EFX_FILTER_MATCH_OUTER_VID | EFX_FILTER_MATCH_LOC_MAC_IG)))) { + netif_info(efx, probe, net_dev, + "VLAN filters are not supported in this firmware variant\n"); + net_dev->features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; + efx->fixed_features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; + net_dev->hw_features &= ~NETIF_F_HW_VLAN_CTAG_FILTER; + } + + table->entry = vzalloc(array_size(EFX_MCDI_FILTER_TBL_ROWS, + sizeof(*table->entry))); + if (!table->entry) { + rc = -ENOMEM; + goto fail; + } + + table->mc_promisc_last = false; + table->vlan_filter = + !!(efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_FILTER); + INIT_LIST_HEAD(&table->vlan_list); + init_rwsem(&table->lock); + + efx->filter_state = table; + + list_for_each_entry(vlan, &nic_data->vlan_list, list) { + rc = efx_mcdi_filter_add_vlan(efx, vlan->vid); + if (rc) + goto fail_add_vlan; + } + + return 0; + +fail_add_vlan: + efx_mcdi_filter_cleanup_vlans(efx); + efx->filter_state = NULL; +fail: + kfree(table); + return rc; +} + +/* + * Caller must hold efx->filter_sem for read if race against + * efx_mcdi_filter_table_remove() is possible + */ +void efx_mcdi_filter_table_restore(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_ef10_nic_data *nic_data = efx->nic_data; + unsigned int invalid_filters = 0, failed = 0; + struct efx_mcdi_filter_vlan *vlan; + struct efx_filter_spec *spec; + struct efx_rss_context *ctx; + unsigned int filter_idx; + u32 mcdi_flags; + int match_pri; + int rc, i; + + WARN_ON(!rwsem_is_locked(&efx->filter_sem)); + + if (!nic_data->must_restore_filters) + return; + + if (!table) + return; + + down_write(&table->lock); + mutex_lock(&efx->rss_lock); + + for (filter_idx = 0; filter_idx < EFX_MCDI_FILTER_TBL_ROWS; filter_idx++) { + spec = efx_mcdi_filter_entry_spec(table, filter_idx); + if (!spec) + continue; + + mcdi_flags = efx_mcdi_filter_mcdi_flags_from_spec(spec); + match_pri = 0; + while (match_pri < table->rx_match_count && + table->rx_match_mcdi_flags[match_pri] != mcdi_flags) + ++match_pri; + if (match_pri >= table->rx_match_count) { + invalid_filters++; + goto not_restored; + } + if (spec->rss_context) + ctx = efx_find_rss_context_entry(efx, spec->rss_context); + else + ctx = &efx->rss_context; + if (spec->flags & EFX_FILTER_FLAG_RX_RSS) { + if (!ctx) { + netif_warn(efx, drv, efx->net_dev, + "Warning: unable to restore a filter with nonexistent RSS context %u.\n", + spec->rss_context); + invalid_filters++; + goto not_restored; + } + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { + netif_warn(efx, drv, efx->net_dev, + "Warning: unable to restore a filter with RSS context %u as it was not created.\n", + spec->rss_context); + invalid_filters++; + goto not_restored; + } + } + + rc = efx_mcdi_filter_push(efx, spec, + &table->entry[filter_idx].handle, + ctx, false); + if (rc) + failed++; + + if (rc) { +not_restored: + list_for_each_entry(vlan, &table->vlan_list, list) + for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; ++i) + if (vlan->default_filters[i] == filter_idx) + vlan->default_filters[i] = + EFX_EF10_FILTER_ID_INVALID; + + kfree(spec); + efx_mcdi_filter_set_entry(table, filter_idx, NULL, 0); + } + } + + mutex_unlock(&efx->rss_lock); + up_write(&table->lock); + + /* + * This can happen validly if the MC's capabilities have changed, so + * is not an error. + */ + if (invalid_filters) + netif_dbg(efx, drv, efx->net_dev, + "Did not restore %u filters that are now unsupported.\n", + invalid_filters); + + if (failed) + netif_err(efx, hw, efx->net_dev, + "unable to restore %u filters\n", failed); + else + nic_data->must_restore_filters = false; +} + +void efx_mcdi_filter_table_remove(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + MCDI_DECLARE_BUF(inbuf, MC_CMD_FILTER_OP_EXT_IN_LEN); + struct efx_filter_spec *spec; + unsigned int filter_idx; + int rc; + + efx_mcdi_filter_cleanup_vlans(efx); + efx->filter_state = NULL; + /* + * If we were called without locking, then it's not safe to free + * the table as others might be using it. So we just WARN, leak + * the memory, and potentially get an inconsistent filter table + * state. + * This should never actually happen. + */ + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return; + + if (!table) + return; + + for (filter_idx = 0; filter_idx < EFX_MCDI_FILTER_TBL_ROWS; filter_idx++) { + spec = efx_mcdi_filter_entry_spec(table, filter_idx); + if (!spec) + continue; + + MCDI_SET_DWORD(inbuf, FILTER_OP_IN_OP, + efx_mcdi_filter_is_exclusive(spec) ? + MC_CMD_FILTER_OP_IN_OP_REMOVE : + MC_CMD_FILTER_OP_IN_OP_UNSUBSCRIBE); + MCDI_SET_QWORD(inbuf, FILTER_OP_IN_HANDLE, + table->entry[filter_idx].handle); + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FILTER_OP, inbuf, + sizeof(inbuf), NULL, 0, NULL); + if (rc) + netif_info(efx, drv, efx->net_dev, + "%s: filter %04x remove failed\n", + __func__, filter_idx); + kfree(spec); + } + + vfree(table->entry); + kfree(table); +} + +static void efx_mcdi_filter_mark_one_old(struct efx_nic *efx, uint16_t *id) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + unsigned int filter_idx; + + efx_rwsem_assert_write_locked(&table->lock); + + if (*id != EFX_EF10_FILTER_ID_INVALID) { + filter_idx = efx_mcdi_filter_get_unsafe_id(*id); + if (!table->entry[filter_idx].spec) + netif_dbg(efx, drv, efx->net_dev, + "marked null spec old %04x:%04x\n", *id, + filter_idx); + table->entry[filter_idx].spec |= EFX_EF10_FILTER_FLAG_AUTO_OLD; + *id = EFX_EF10_FILTER_ID_INVALID; + } +} + +/* Mark old per-VLAN filters that may need to be removed */ +static void _efx_mcdi_filter_vlan_mark_old(struct efx_nic *efx, + struct efx_mcdi_filter_vlan *vlan) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + unsigned int i; + + for (i = 0; i < table->dev_uc_count; i++) + efx_mcdi_filter_mark_one_old(efx, &vlan->uc[i]); + for (i = 0; i < table->dev_mc_count; i++) + efx_mcdi_filter_mark_one_old(efx, &vlan->mc[i]); + for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) + efx_mcdi_filter_mark_one_old(efx, &vlan->default_filters[i]); +} + +/* + * Mark old filters that may need to be removed. + * Caller must hold efx->filter_sem for read if race against + * efx_mcdi_filter_table_remove() is possible + */ +static void efx_mcdi_filter_mark_old(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_mcdi_filter_vlan *vlan; + + down_write(&table->lock); + list_for_each_entry(vlan, &table->vlan_list, list) + _efx_mcdi_filter_vlan_mark_old(efx, vlan); + up_write(&table->lock); +} + +int efx_mcdi_filter_add_vlan(struct efx_nic *efx, u16 vid) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_mcdi_filter_vlan *vlan; + unsigned int i; + + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return -EINVAL; + + vlan = efx_mcdi_filter_find_vlan(efx, vid); + if (WARN_ON(vlan)) { + netif_err(efx, drv, efx->net_dev, + "VLAN %u already added\n", vid); + return -EALREADY; + } + + vlan = kzalloc(sizeof(*vlan), GFP_KERNEL); + if (!vlan) + return -ENOMEM; + + vlan->vid = vid; + + for (i = 0; i < ARRAY_SIZE(vlan->uc); i++) + vlan->uc[i] = EFX_EF10_FILTER_ID_INVALID; + for (i = 0; i < ARRAY_SIZE(vlan->mc); i++) + vlan->mc[i] = EFX_EF10_FILTER_ID_INVALID; + for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) + vlan->default_filters[i] = EFX_EF10_FILTER_ID_INVALID; + + list_add_tail(&vlan->list, &table->vlan_list); + + if (efx_dev_registered(efx)) + efx_mcdi_filter_vlan_sync_rx_mode(efx, vlan); + + return 0; +} + +static void efx_mcdi_filter_del_vlan_internal(struct efx_nic *efx, + struct efx_mcdi_filter_vlan *vlan) +{ + unsigned int i; + + /* See comment in efx_mcdi_filter_table_remove() */ + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return; + + list_del(&vlan->list); + + for (i = 0; i < ARRAY_SIZE(vlan->uc); i++) + efx_mcdi_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, + vlan->uc[i]); + for (i = 0; i < ARRAY_SIZE(vlan->mc); i++) + efx_mcdi_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, + vlan->mc[i]); + for (i = 0; i < EFX_EF10_NUM_DEFAULT_FILTERS; i++) + if (vlan->default_filters[i] != EFX_EF10_FILTER_ID_INVALID) + efx_mcdi_filter_remove_unsafe(efx, EFX_FILTER_PRI_AUTO, + vlan->default_filters[i]); + + kfree(vlan); +} + +void efx_mcdi_filter_del_vlan(struct efx_nic *efx, u16 vid) +{ + struct efx_mcdi_filter_vlan *vlan; + + /* See comment in efx_mcdi_filter_table_remove() */ + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return; + + vlan = efx_mcdi_filter_find_vlan(efx, vid); + if (!vlan) { + netif_err(efx, drv, efx->net_dev, + "VLAN %u not found in filter state\n", vid); + return; + } + + efx_mcdi_filter_del_vlan_internal(efx, vlan); +} + +struct efx_mcdi_filter_vlan *efx_mcdi_filter_find_vlan(struct efx_nic *efx, + u16 vid) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_mcdi_filter_vlan *vlan; + + WARN_ON(!rwsem_is_locked(&efx->filter_sem)); + + list_for_each_entry(vlan, &table->vlan_list, list) { + if (vlan->vid == vid) + return vlan; + } + + return NULL; +} + +void efx_mcdi_filter_cleanup_vlans(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct efx_mcdi_filter_vlan *vlan, *next_vlan; + + /* See comment in efx_mcdi_filter_table_remove() */ + if (!efx_rwsem_assert_write_locked(&efx->filter_sem)) + return; + + if (!table) + return; + + list_for_each_entry_safe(vlan, next_vlan, &table->vlan_list, list) + efx_mcdi_filter_del_vlan_internal(efx, vlan); +} + +static void efx_mcdi_filter_uc_addr_list(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct net_device *net_dev = efx->net_dev; + struct netdev_hw_addr *uc; + unsigned int i; + + table->uc_promisc = !!(net_dev->flags & IFF_PROMISC); + ether_addr_copy(table->dev_uc_list[0].addr, net_dev->dev_addr); + i = 1; + netdev_for_each_uc_addr(uc, net_dev) { + if (i >= EFX_EF10_FILTER_DEV_UC_MAX) { + table->uc_promisc = true; + break; + } + ether_addr_copy(table->dev_uc_list[i].addr, uc->addr); + i++; + } + + table->dev_uc_count = i; +} + +static void efx_mcdi_filter_mc_addr_list(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct net_device *net_dev = efx->net_dev; + struct netdev_hw_addr *mc; + unsigned int i; + + table->mc_overflow = false; + table->mc_promisc = !!(net_dev->flags & (IFF_PROMISC | IFF_ALLMULTI)); + + i = 0; + netdev_for_each_mc_addr(mc, net_dev) { + if (i >= EFX_EF10_FILTER_DEV_MC_MAX) { + table->mc_promisc = true; + table->mc_overflow = true; + break; + } + ether_addr_copy(table->dev_mc_list[i].addr, mc->addr); + i++; + } + + table->dev_mc_count = i; +} + +/* + * Caller must hold efx->filter_sem for read if race against + * efx_mcdi_filter_table_remove() is possible + */ +void efx_mcdi_filter_sync_rx_mode(struct efx_nic *efx) +{ + struct efx_mcdi_filter_table *table = efx->filter_state; + struct net_device *net_dev = efx->net_dev; + struct efx_mcdi_filter_vlan *vlan; + bool vlan_filter; + + if (!efx_dev_registered(efx)) + return; + + if (!table) + return; + + efx_mcdi_filter_mark_old(efx); + + /* + * Copy/convert the address lists; add the primary station + * address and broadcast address + */ + netif_addr_lock_bh(net_dev); + efx_mcdi_filter_uc_addr_list(efx); + efx_mcdi_filter_mc_addr_list(efx); + netif_addr_unlock_bh(net_dev); + + /* + * If VLAN filtering changes, all old filters are finally removed. + * Do it in advance to avoid conflicts for unicast untagged and + * VLAN 0 tagged filters. + */ + vlan_filter = !!(net_dev->features & NETIF_F_HW_VLAN_CTAG_FILTER); + if (table->vlan_filter != vlan_filter) { + table->vlan_filter = vlan_filter; + efx_mcdi_filter_remove_old(efx); + } + + list_for_each_entry(vlan, &table->vlan_list, list) + efx_mcdi_filter_vlan_sync_rx_mode(efx, vlan); + + efx_mcdi_filter_remove_old(efx); + table->mc_promisc_last = table->mc_promisc; +} + +#ifdef CONFIG_RFS_ACCEL + +bool efx_mcdi_filter_rfs_expire_one(struct efx_nic *efx, u32 flow_id, + unsigned int filter_idx) +{ + struct efx_filter_spec *spec, saved_spec; + struct efx_mcdi_filter_table *table; + struct efx_arfs_rule *rule = NULL; + bool ret = true, force = false; + u16 arfs_id; + + down_read(&efx->filter_sem); + table = efx->filter_state; + down_write(&table->lock); + spec = efx_mcdi_filter_entry_spec(table, filter_idx); + + if (!spec || spec->priority != EFX_FILTER_PRI_HINT) + goto out_unlock; + + spin_lock_bh(&efx->rps_hash_lock); + if (!efx->rps_hash_table) { + /* In the absence of the table, we always return 0 to ARFS. */ + arfs_id = 0; + } else { + rule = efx_rps_hash_find(efx, spec); + if (!rule) + /* ARFS table doesn't know of this filter, so remove it */ + goto expire; + arfs_id = rule->arfs_id; + ret = efx_rps_check_rule(rule, filter_idx, &force); + if (force) + goto expire; + if (!ret) { + spin_unlock_bh(&efx->rps_hash_lock); + goto out_unlock; + } + } + if (!rps_may_expire_flow(efx->net_dev, spec->dmaq_id, flow_id, arfs_id)) + ret = false; + else if (rule) + rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; +expire: + saved_spec = *spec; /* remove operation will kfree spec */ + spin_unlock_bh(&efx->rps_hash_lock); + /* + * At this point (since we dropped the lock), another thread might queue + * up a fresh insertion request (but the actual insertion will be held + * up by our possession of the filter table lock). In that case, it + * will set rule->filter_id to EFX_ARFS_FILTER_ID_PENDING, meaning that + * the rule is not removed by efx_rps_hash_del() below. + */ + if (ret) + ret = efx_mcdi_filter_remove_internal(efx, 1U << spec->priority, + filter_idx, true) == 0; + /* + * While we can't safely dereference rule (we dropped the lock), we can + * still test it for NULL. + */ + if (ret && rule) { + /* Expiring, so remove entry from ARFS table */ + spin_lock_bh(&efx->rps_hash_lock); + efx_rps_hash_del(efx, &saved_spec); + spin_unlock_bh(&efx->rps_hash_lock); + } +out_unlock: + up_write(&table->lock); + up_read(&efx->filter_sem); + return ret; +} + +#endif /* CONFIG_RFS_ACCEL */ + +#define RSS_MODE_HASH_ADDRS (1 << RSS_MODE_HASH_SRC_ADDR_LBN |\ + 1 << RSS_MODE_HASH_DST_ADDR_LBN) +#define RSS_MODE_HASH_PORTS (1 << RSS_MODE_HASH_SRC_PORT_LBN |\ + 1 << RSS_MODE_HASH_DST_PORT_LBN) +#define RSS_CONTEXT_FLAGS_DEFAULT (1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_IPV4_EN_LBN |\ + 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_TCPV4_EN_LBN |\ + 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_IPV6_EN_LBN |\ + 1 << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TOEPLITZ_TCPV6_EN_LBN |\ + (RSS_MODE_HASH_ADDRS | RSS_MODE_HASH_PORTS) << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TCP_IPV4_RSS_MODE_LBN |\ + RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV4_RSS_MODE_LBN |\ + RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_OTHER_IPV4_RSS_MODE_LBN |\ + (RSS_MODE_HASH_ADDRS | RSS_MODE_HASH_PORTS) << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_TCP_IPV6_RSS_MODE_LBN |\ + RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV6_RSS_MODE_LBN |\ + RSS_MODE_HASH_ADDRS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_OTHER_IPV6_RSS_MODE_LBN) + +int efx_mcdi_get_rss_context_flags(struct efx_nic *efx, u32 context, u32 *flags) +{ + /* + * Firmware had a bug (sfc bug 61952) where it would not actually + * fill in the flags field in the response to MC_CMD_RSS_CONTEXT_GET_FLAGS. + * This meant that it would always contain whatever was previously + * in the MCDI buffer. Fortunately, all firmware versions with + * this bug have the same default flags value for a newly-allocated + * RSS context, and the only time we want to get the flags is just + * after allocating. Moreover, the response has a 32-bit hole + * where the context ID would be in the request, so we can use an + * overlength buffer in the request and pre-fill the flags field + * with what we believe the default to be. Thus if the firmware + * has the bug, it will leave our pre-filled value in the flags + * field of the response, and we will get the right answer. + * + * However, this does mean that this function should NOT be used if + * the RSS context flags might not be their defaults - it is ONLY + * reliably correct for a newly-allocated RSS context. + */ + MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN); + MCDI_DECLARE_BUF(outbuf, MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN); + size_t outlen; + int rc; + + /* Check we have a hole for the context ID */ + BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_GET_FLAGS_IN_LEN != MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_FLAGS_OFST); + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_FLAGS_IN_RSS_CONTEXT_ID, context); + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_FLAGS_OUT_FLAGS, + RSS_CONTEXT_FLAGS_DEFAULT); + rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_FLAGS, inbuf, + sizeof(inbuf), outbuf, sizeof(outbuf), &outlen); + if (rc == 0) { + if (outlen < MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_LEN) + rc = -EIO; + else + *flags = MCDI_DWORD(outbuf, RSS_CONTEXT_GET_FLAGS_OUT_FLAGS); + } + return rc; +} + +/* + * Attempt to enable 4-tuple UDP hashing on the specified RSS context. + * If we fail, we just leave the RSS context at its default hash settings, + * which is safe but may slightly reduce performance. + * Defaults are 4-tuple for TCP and 2-tuple for UDP and other-IP, so we + * just need to set the UDP ports flags (for both IP versions). + */ +void efx_mcdi_set_rss_context_flags(struct efx_nic *efx, + struct efx_rss_context *ctx) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_SET_FLAGS_IN_LEN); + u32 flags; + + BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_SET_FLAGS_OUT_LEN != 0); + + if (efx_mcdi_get_rss_context_flags(efx, ctx->context_id, &flags) != 0) + return; + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_SET_FLAGS_IN_RSS_CONTEXT_ID, + ctx->context_id); + flags |= RSS_MODE_HASH_PORTS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV4_RSS_MODE_LBN; + flags |= RSS_MODE_HASH_PORTS << MC_CMD_RSS_CONTEXT_GET_FLAGS_OUT_UDP_IPV6_RSS_MODE_LBN; + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_SET_FLAGS_IN_FLAGS, flags); + if (!efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_FLAGS, inbuf, sizeof(inbuf), + NULL, 0, NULL)) + /* Succeeded, so UDP 4-tuple is now enabled */ + ctx->rx_hash_udp_4tuple = true; +} + +static int efx_mcdi_filter_alloc_rss_context(struct efx_nic *efx, bool exclusive, + struct efx_rss_context *ctx, + unsigned *context_size) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_ALLOC_IN_LEN); + MCDI_DECLARE_BUF(outbuf, MC_CMD_RSS_CONTEXT_ALLOC_OUT_LEN); + struct efx_ef10_nic_data *nic_data = efx->nic_data; + size_t outlen; + int rc; + u32 alloc_type = exclusive ? + MC_CMD_RSS_CONTEXT_ALLOC_IN_TYPE_EXCLUSIVE : + MC_CMD_RSS_CONTEXT_ALLOC_IN_TYPE_SHARED; + unsigned rss_spread = exclusive ? + efx->rss_spread : + min(rounddown_pow_of_two(efx->rss_spread), + EFX_EF10_MAX_SHARED_RSS_CONTEXT_SIZE); + + if (!exclusive && rss_spread == 1) { + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; + if (context_size) + *context_size = 1; + return 0; + } + + if (nic_data->datapath_caps & + 1 << MC_CMD_GET_CAPABILITIES_OUT_RX_RSS_LIMITED_LBN) + return -EOPNOTSUPP; + + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_UPSTREAM_PORT_ID, + nic_data->vport_id); + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_TYPE, alloc_type); + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_ALLOC_IN_NUM_QUEUES, rss_spread); + + rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_ALLOC, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + if (rc != 0) + return rc; + + if (outlen < MC_CMD_RSS_CONTEXT_ALLOC_OUT_LEN) + return -EIO; + + ctx->context_id = MCDI_DWORD(outbuf, RSS_CONTEXT_ALLOC_OUT_RSS_CONTEXT_ID); + + if (context_size) + *context_size = rss_spread; + + if (nic_data->datapath_caps & + 1 << MC_CMD_GET_CAPABILITIES_OUT_ADDITIONAL_RSS_MODES_LBN) + efx_mcdi_set_rss_context_flags(efx, ctx); + + return 0; +} + +static int efx_mcdi_filter_free_rss_context(struct efx_nic *efx, u32 context) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_FREE_IN_LEN); + + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_FREE_IN_RSS_CONTEXT_ID, + context); + return efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_FREE, inbuf, sizeof(inbuf), + NULL, 0, NULL); +} + +static int efx_mcdi_filter_populate_rss_table(struct efx_nic *efx, u32 context, + const u32 *rx_indir_table, const u8 *key) +{ + MCDI_DECLARE_BUF(tablebuf, MC_CMD_RSS_CONTEXT_SET_TABLE_IN_LEN); + MCDI_DECLARE_BUF(keybuf, MC_CMD_RSS_CONTEXT_SET_KEY_IN_LEN); + int i, rc; + + MCDI_SET_DWORD(tablebuf, RSS_CONTEXT_SET_TABLE_IN_RSS_CONTEXT_ID, + context); + BUILD_BUG_ON(ARRAY_SIZE(efx->rss_context.rx_indir_table) != + MC_CMD_RSS_CONTEXT_SET_TABLE_IN_INDIRECTION_TABLE_LEN); + + /* This iterates over the length of efx->rss_context.rx_indir_table, but + * copies bytes from rx_indir_table. That's because the latter is a + * pointer rather than an array, but should have the same length. + * The efx->rss_context.rx_hash_key loop below is similar. + */ + for (i = 0; i < ARRAY_SIZE(efx->rss_context.rx_indir_table); ++i) + MCDI_PTR(tablebuf, + RSS_CONTEXT_SET_TABLE_IN_INDIRECTION_TABLE)[i] = + (u8) rx_indir_table[i]; + + rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_TABLE, tablebuf, + sizeof(tablebuf), NULL, 0, NULL); + if (rc != 0) + return rc; + + MCDI_SET_DWORD(keybuf, RSS_CONTEXT_SET_KEY_IN_RSS_CONTEXT_ID, + context); + BUILD_BUG_ON(ARRAY_SIZE(efx->rss_context.rx_hash_key) != + MC_CMD_RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY_LEN); + for (i = 0; i < ARRAY_SIZE(efx->rss_context.rx_hash_key); ++i) + MCDI_PTR(keybuf, RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY)[i] = key[i]; + + return efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_SET_KEY, keybuf, + sizeof(keybuf), NULL, 0, NULL); +} + +void efx_mcdi_rx_free_indir_table(struct efx_nic *efx) +{ + int rc; + + if (efx->rss_context.context_id != EFX_MCDI_RSS_CONTEXT_INVALID) { + rc = efx_mcdi_filter_free_rss_context(efx, efx->rss_context.context_id); + WARN_ON(rc != 0); + } + efx->rss_context.context_id = EFX_MCDI_RSS_CONTEXT_INVALID; +} + +static int efx_mcdi_filter_rx_push_shared_rss_config(struct efx_nic *efx, + unsigned *context_size) +{ + struct efx_ef10_nic_data *nic_data = efx->nic_data; + int rc = efx_mcdi_filter_alloc_rss_context(efx, false, &efx->rss_context, + context_size); + + if (rc != 0) + return rc; + + nic_data->rx_rss_context_exclusive = false; + efx_set_default_rx_indir_table(efx, &efx->rss_context); + return 0; +} + +static int efx_mcdi_filter_rx_push_exclusive_rss_config(struct efx_nic *efx, + const u32 *rx_indir_table, + const u8 *key) +{ + u32 old_rx_rss_context = efx->rss_context.context_id; + struct efx_ef10_nic_data *nic_data = efx->nic_data; + int rc; + + if (efx->rss_context.context_id == EFX_MCDI_RSS_CONTEXT_INVALID || + !nic_data->rx_rss_context_exclusive) { + rc = efx_mcdi_filter_alloc_rss_context(efx, true, &efx->rss_context, + NULL); + if (rc == -EOPNOTSUPP) + return rc; + else if (rc != 0) + goto fail1; + } + + rc = efx_mcdi_filter_populate_rss_table(efx, efx->rss_context.context_id, + rx_indir_table, key); + if (rc != 0) + goto fail2; + + if (efx->rss_context.context_id != old_rx_rss_context && + old_rx_rss_context != EFX_MCDI_RSS_CONTEXT_INVALID) + WARN_ON(efx_mcdi_filter_free_rss_context(efx, old_rx_rss_context) != 0); + nic_data->rx_rss_context_exclusive = true; + if (rx_indir_table != efx->rss_context.rx_indir_table) + memcpy(efx->rss_context.rx_indir_table, rx_indir_table, + sizeof(efx->rss_context.rx_indir_table)); + if (key != efx->rss_context.rx_hash_key) + memcpy(efx->rss_context.rx_hash_key, key, + efx->type->rx_hash_key_size); + + return 0; + +fail2: + if (old_rx_rss_context != efx->rss_context.context_id) { + WARN_ON(efx_mcdi_filter_free_rss_context(efx, efx->rss_context.context_id) != 0); + efx->rss_context.context_id = old_rx_rss_context; + } +fail1: + netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); + return rc; +} + +int efx_mcdi_rx_push_rss_context_config(struct efx_nic *efx, + struct efx_rss_context *ctx, + const u32 *rx_indir_table, + const u8 *key) +{ + int rc; + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) { + rc = efx_mcdi_filter_alloc_rss_context(efx, true, ctx, NULL); + if (rc) + return rc; + } + + if (!rx_indir_table) /* Delete this context */ + return efx_mcdi_filter_free_rss_context(efx, ctx->context_id); + + rc = efx_mcdi_filter_populate_rss_table(efx, ctx->context_id, + rx_indir_table, key); + if (rc) + return rc; + + memcpy(ctx->rx_indir_table, rx_indir_table, + sizeof(efx->rss_context.rx_indir_table)); + memcpy(ctx->rx_hash_key, key, efx->type->rx_hash_key_size); + + return 0; +} + +int efx_mcdi_rx_pull_rss_context_config(struct efx_nic *efx, + struct efx_rss_context *ctx) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_RSS_CONTEXT_GET_TABLE_IN_LEN); + MCDI_DECLARE_BUF(tablebuf, MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_LEN); + MCDI_DECLARE_BUF(keybuf, MC_CMD_RSS_CONTEXT_GET_KEY_OUT_LEN); + size_t outlen; + int rc, i; + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + BUILD_BUG_ON(MC_CMD_RSS_CONTEXT_GET_TABLE_IN_LEN != + MC_CMD_RSS_CONTEXT_GET_KEY_IN_LEN); + + if (ctx->context_id == EFX_MCDI_RSS_CONTEXT_INVALID) + return -ENOENT; + + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_TABLE_IN_RSS_CONTEXT_ID, + ctx->context_id); + BUILD_BUG_ON(ARRAY_SIZE(ctx->rx_indir_table) != + MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_INDIRECTION_TABLE_LEN); + rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_TABLE, inbuf, sizeof(inbuf), + tablebuf, sizeof(tablebuf), &outlen); + if (rc != 0) + return rc; + + if (WARN_ON(outlen != MC_CMD_RSS_CONTEXT_GET_TABLE_OUT_LEN)) + return -EIO; + + for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) + ctx->rx_indir_table[i] = MCDI_PTR(tablebuf, + RSS_CONTEXT_GET_TABLE_OUT_INDIRECTION_TABLE)[i]; + + MCDI_SET_DWORD(inbuf, RSS_CONTEXT_GET_KEY_IN_RSS_CONTEXT_ID, + ctx->context_id); + BUILD_BUG_ON(ARRAY_SIZE(ctx->rx_hash_key) != + MC_CMD_RSS_CONTEXT_SET_KEY_IN_TOEPLITZ_KEY_LEN); + rc = efx_mcdi_rpc(efx, MC_CMD_RSS_CONTEXT_GET_KEY, inbuf, sizeof(inbuf), + keybuf, sizeof(keybuf), &outlen); + if (rc != 0) + return rc; + + if (WARN_ON(outlen != MC_CMD_RSS_CONTEXT_GET_KEY_OUT_LEN)) + return -EIO; + + for (i = 0; i < ARRAY_SIZE(ctx->rx_hash_key); ++i) + ctx->rx_hash_key[i] = MCDI_PTR( + keybuf, RSS_CONTEXT_GET_KEY_OUT_TOEPLITZ_KEY)[i]; + + return 0; +} + +int efx_mcdi_rx_pull_rss_config(struct efx_nic *efx) +{ + int rc; + + mutex_lock(&efx->rss_lock); + rc = efx_mcdi_rx_pull_rss_context_config(efx, &efx->rss_context); + mutex_unlock(&efx->rss_lock); + return rc; +} + +void efx_mcdi_rx_restore_rss_contexts(struct efx_nic *efx) +{ + struct efx_ef10_nic_data *nic_data = efx->nic_data; + struct efx_rss_context *ctx; + int rc; + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + if (!nic_data->must_restore_rss_contexts) + return; + + list_for_each_entry(ctx, &efx->rss_context.list, list) { + /* previous NIC RSS context is gone */ + ctx->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; + /* so try to allocate a new one */ + rc = efx_mcdi_rx_push_rss_context_config(efx, ctx, + ctx->rx_indir_table, + ctx->rx_hash_key); + if (rc) + netif_warn(efx, probe, efx->net_dev, + "failed to restore RSS context %u, rc=%d" + "; RSS filters may fail to be applied\n", + ctx->user_id, rc); + } + nic_data->must_restore_rss_contexts = false; +} + +int efx_mcdi_pf_rx_push_rss_config(struct efx_nic *efx, bool user, + const u32 *rx_indir_table, + const u8 *key) +{ + int rc; + + if (efx->rss_spread == 1) + return 0; + + if (!key) + key = efx->rss_context.rx_hash_key; + + rc = efx_mcdi_filter_rx_push_exclusive_rss_config(efx, rx_indir_table, key); + + if (rc == -ENOBUFS && !user) { + unsigned context_size; + bool mismatch = false; + size_t i; + + for (i = 0; + i < ARRAY_SIZE(efx->rss_context.rx_indir_table) && !mismatch; + i++) + mismatch = rx_indir_table[i] != + ethtool_rxfh_indir_default(i, efx->rss_spread); + + rc = efx_mcdi_filter_rx_push_shared_rss_config(efx, &context_size); + if (rc == 0) { + if (context_size != efx->rss_spread) + netif_warn(efx, probe, efx->net_dev, + "Could not allocate an exclusive RSS" + " context; allocated a shared one of" + " different size." + " Wanted %u, got %u.\n", + efx->rss_spread, context_size); + else if (mismatch) + netif_warn(efx, probe, efx->net_dev, + "Could not allocate an exclusive RSS" + " context; allocated a shared one but" + " could not apply custom" + " indirection.\n"); + else + netif_info(efx, probe, efx->net_dev, + "Could not allocate an exclusive RSS" + " context; allocated a shared one.\n"); + } + } + return rc; +} + +int efx_mcdi_vf_rx_push_rss_config(struct efx_nic *efx, bool user, + const u32 *rx_indir_table + __attribute__ ((unused)), + const u8 *key + __attribute__ ((unused))) +{ + if (user) + return -EOPNOTSUPP; + if (efx->rss_context.context_id != EFX_MCDI_RSS_CONTEXT_INVALID) + return 0; + return efx_mcdi_filter_rx_push_shared_rss_config(efx, NULL); +} diff --git a/drivers/net/ethernet/sfc/mcdi_filters.h b/drivers/net/ethernet/sfc/mcdi_filters.h new file mode 100644 index 000000000000..1837f4f5d661 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_filters.h @@ -0,0 +1,159 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ +#ifndef EFX_MCDI_FILTERS_H +#define EFX_MCDI_FILTERS_H + +#include "net_driver.h" +#include "filter.h" +#include "mcdi_pcol.h" + +#define EFX_EF10_FILTER_DEV_UC_MAX 32 +#define EFX_EF10_FILTER_DEV_MC_MAX 256 + +enum efx_mcdi_filter_default_filters { + EFX_EF10_BCAST, + EFX_EF10_UCDEF, + EFX_EF10_MCDEF, + EFX_EF10_VXLAN4_UCDEF, + EFX_EF10_VXLAN4_MCDEF, + EFX_EF10_VXLAN6_UCDEF, + EFX_EF10_VXLAN6_MCDEF, + EFX_EF10_NVGRE4_UCDEF, + EFX_EF10_NVGRE4_MCDEF, + EFX_EF10_NVGRE6_UCDEF, + EFX_EF10_NVGRE6_MCDEF, + EFX_EF10_GENEVE4_UCDEF, + EFX_EF10_GENEVE4_MCDEF, + EFX_EF10_GENEVE6_UCDEF, + EFX_EF10_GENEVE6_MCDEF, + + EFX_EF10_NUM_DEFAULT_FILTERS +}; + +/* Per-VLAN filters information */ +struct efx_mcdi_filter_vlan { + struct list_head list; + u16 vid; + u16 uc[EFX_EF10_FILTER_DEV_UC_MAX]; + u16 mc[EFX_EF10_FILTER_DEV_MC_MAX]; + u16 default_filters[EFX_EF10_NUM_DEFAULT_FILTERS]; +}; + +struct efx_mcdi_dev_addr { + u8 addr[ETH_ALEN]; +}; + +struct efx_mcdi_filter_table { +/* The MCDI match masks supported by this fw & hw, in order of priority */ + u32 rx_match_mcdi_flags[ + MC_CMD_GET_PARSER_DISP_INFO_OUT_SUPPORTED_MATCHES_MAXNUM * 2]; + unsigned int rx_match_count; + + struct rw_semaphore lock; /* Protects entries */ + struct { + unsigned long spec; /* pointer to spec plus flag bits */ +/* AUTO_OLD is used to mark and sweep MAC filters for the device address lists. */ +/* unused flag 1UL */ +#define EFX_EF10_FILTER_FLAG_AUTO_OLD 2UL +#define EFX_EF10_FILTER_FLAGS 3UL + u64 handle; /* firmware handle */ + } *entry; +/* Shadow of net_device address lists, guarded by mac_lock */ + struct efx_mcdi_dev_addr dev_uc_list[EFX_EF10_FILTER_DEV_UC_MAX]; + struct efx_mcdi_dev_addr dev_mc_list[EFX_EF10_FILTER_DEV_MC_MAX]; + int dev_uc_count; + int dev_mc_count; + bool uc_promisc; + bool mc_promisc; +/* Whether in multicast promiscuous mode when last changed */ + bool mc_promisc_last; + bool mc_overflow; /* Too many MC addrs; should always imply mc_promisc */ + bool vlan_filter; + struct list_head vlan_list; +}; + +int efx_mcdi_filter_table_probe(struct efx_nic *efx); +void efx_mcdi_filter_table_remove(struct efx_nic *efx); +void efx_mcdi_filter_table_restore(struct efx_nic *efx); + +/* + * The filter table(s) are managed by firmware and we have write-only + * access. When removing filters we must identify them to the + * firmware by a 64-bit handle, but this is too wide for Linux kernel + * interfaces (32-bit for RX NFC, 16-bit for RFS). Also, we need to + * be able to tell in advance whether a requested insertion will + * replace an existing filter. Therefore we maintain a software hash + * table, which should be at least as large as the hardware hash + * table. + * + * Huntington has a single 8K filter table shared between all filter + * types and both ports. + */ +#define EFX_MCDI_FILTER_TBL_ROWS 8192 + +bool efx_mcdi_filter_match_supported(struct efx_mcdi_filter_table *table, + bool encap, + enum efx_filter_match_flags match_flags); + +void efx_mcdi_filter_sync_rx_mode(struct efx_nic *efx); +s32 efx_mcdi_filter_insert(struct efx_nic *efx, struct efx_filter_spec *spec, + bool replace_equal); +int efx_mcdi_filter_remove_safe(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 filter_id); +int efx_mcdi_filter_get_safe(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 filter_id, struct efx_filter_spec *spec); + +u32 efx_mcdi_filter_count_rx_used(struct efx_nic *efx, + enum efx_filter_priority priority); +int efx_mcdi_filter_clear_rx(struct efx_nic *efx, + enum efx_filter_priority priority); +u32 efx_mcdi_filter_get_rx_id_limit(struct efx_nic *efx); +s32 efx_mcdi_filter_get_rx_ids(struct efx_nic *efx, + enum efx_filter_priority priority, + u32 *buf, u32 size); + +void efx_mcdi_filter_cleanup_vlans(struct efx_nic *efx); +int efx_mcdi_filter_add_vlan(struct efx_nic *efx, u16 vid); +struct efx_mcdi_filter_vlan *efx_mcdi_filter_find_vlan(struct efx_nic *efx, u16 vid); +void efx_mcdi_filter_del_vlan(struct efx_nic *efx, u16 vid); + +void efx_mcdi_rx_free_indir_table(struct efx_nic *efx); +int efx_mcdi_rx_push_rss_context_config(struct efx_nic *efx, + struct efx_rss_context *ctx, + const u32 *rx_indir_table, + const u8 *key); +int efx_mcdi_pf_rx_push_rss_config(struct efx_nic *efx, bool user, + const u32 *rx_indir_table, + const u8 *key); +int efx_mcdi_vf_rx_push_rss_config(struct efx_nic *efx, bool user, + const u32 *rx_indir_table + __attribute__ ((unused)), + const u8 *key + __attribute__ ((unused))); +int efx_mcdi_rx_pull_rss_config(struct efx_nic *efx); +int efx_mcdi_rx_pull_rss_context_config(struct efx_nic *efx, + struct efx_rss_context *ctx); +int efx_mcdi_get_rss_context_flags(struct efx_nic *efx, u32 context, + u32 *flags); +void efx_mcdi_set_rss_context_flags(struct efx_nic *efx, + struct efx_rss_context *ctx); +void efx_mcdi_rx_restore_rss_contexts(struct efx_nic *efx); + +static inline void efx_mcdi_update_rx_scatter(struct efx_nic *efx) +{ + /* no need to do anything here */ +} + +bool efx_mcdi_filter_rfs_expire_one(struct efx_nic *efx, u32 flow_id, + unsigned int filter_idx); + +#endif diff --git a/drivers/net/ethernet/sfc/mcdi_functions.c b/drivers/net/ethernet/sfc/mcdi_functions.c new file mode 100644 index 000000000000..dcfe78b0fa5a --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_functions.c @@ -0,0 +1,386 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2019 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include "efx.h" +#include "nic.h" +#include "mcdi_functions.h" +#include "mcdi.h" +#include "mcdi_pcol.h" + +int efx_mcdi_free_vis(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF_ERR(outbuf); + size_t outlen; + int rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FREE_VIS, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + + /* -EALREADY means nothing to free, so ignore */ + if (rc == -EALREADY) + rc = 0; + if (rc) + efx_mcdi_display_error(efx, MC_CMD_FREE_VIS, 0, outbuf, outlen, + rc); + return rc; +} + +int efx_mcdi_alloc_vis(struct efx_nic *efx, unsigned int min_vis, + unsigned int max_vis, unsigned int *vi_base, + unsigned int *allocated_vis) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_ALLOC_VIS_OUT_LEN); + MCDI_DECLARE_BUF(inbuf, MC_CMD_ALLOC_VIS_IN_LEN); + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MIN_VI_COUNT, min_vis); + MCDI_SET_DWORD(inbuf, ALLOC_VIS_IN_MAX_VI_COUNT, max_vis); + rc = efx_mcdi_rpc(efx, MC_CMD_ALLOC_VIS, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + if (rc != 0) + return rc; + + if (outlen < MC_CMD_ALLOC_VIS_OUT_LEN) + return -EIO; + + netif_dbg(efx, drv, efx->net_dev, "base VI is A0x%03x\n", + MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE)); + + if (vi_base) + *vi_base = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_BASE); + if (allocated_vis) + *allocated_vis = MCDI_DWORD(outbuf, ALLOC_VIS_OUT_VI_COUNT); + return 0; +} + +int efx_mcdi_ev_probe(struct efx_channel *channel) +{ + return efx_nic_alloc_buffer(channel->efx, &channel->eventq.buf, + (channel->eventq_mask + 1) * + sizeof(efx_qword_t), + GFP_KERNEL); +} + +int efx_mcdi_ev_init(struct efx_channel *channel, bool v1_cut_thru, bool v2) +{ + MCDI_DECLARE_BUF(inbuf, + MC_CMD_INIT_EVQ_V2_IN_LEN(EFX_MAX_EVQ_SIZE * 8 / + EFX_BUF_SIZE)); + MCDI_DECLARE_BUF(outbuf, MC_CMD_INIT_EVQ_V2_OUT_LEN); + size_t entries = channel->eventq.buf.len / EFX_BUF_SIZE; + struct efx_nic *efx = channel->efx; + size_t inlen, outlen; + dma_addr_t dma_addr; + int rc, i; + + /* Fill event queue with all ones (i.e. empty events) */ + memset(channel->eventq.buf.addr, 0xff, channel->eventq.buf.len); + + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_SIZE, channel->eventq_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_INSTANCE, channel->channel); + /* INIT_EVQ expects index in vector table, not absolute */ + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_IRQ_NUM, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_MODE, + MC_CMD_INIT_EVQ_IN_TMR_MODE_DIS); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_LOAD, 0); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_TMR_RELOAD, 0); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_MODE, + MC_CMD_INIT_EVQ_IN_COUNT_MODE_DIS); + MCDI_SET_DWORD(inbuf, INIT_EVQ_IN_COUNT_THRSHLD, 0); + + if (v2) { + /* Use the new generic approach to specifying event queue + * configuration, requesting lower latency or higher throughput. + * The options that actually get used appear in the output. + */ + MCDI_POPULATE_DWORD_2(inbuf, INIT_EVQ_V2_IN_FLAGS, + INIT_EVQ_V2_IN_FLAG_INTERRUPTING, 1, + INIT_EVQ_V2_IN_FLAG_TYPE, + MC_CMD_INIT_EVQ_V2_IN_FLAG_TYPE_AUTO); + } else { + MCDI_POPULATE_DWORD_4(inbuf, INIT_EVQ_IN_FLAGS, + INIT_EVQ_IN_FLAG_INTERRUPTING, 1, + INIT_EVQ_IN_FLAG_RX_MERGE, 1, + INIT_EVQ_IN_FLAG_TX_MERGE, 1, + INIT_EVQ_IN_FLAG_CUT_THRU, v1_cut_thru); + } + + dma_addr = channel->eventq.buf.dma_addr; + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_EVQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_EVQ_IN_LEN(entries); + + rc = efx_mcdi_rpc(efx, MC_CMD_INIT_EVQ, inbuf, inlen, + outbuf, sizeof(outbuf), &outlen); + + if (outlen >= MC_CMD_INIT_EVQ_V2_OUT_LEN) + netif_dbg(efx, drv, efx->net_dev, + "Channel %d using event queue flags %08x\n", + channel->channel, + MCDI_DWORD(outbuf, INIT_EVQ_V2_OUT_FLAGS)); + + return rc; +} + +void efx_mcdi_ev_remove(struct efx_channel *channel) +{ + efx_nic_free_buffer(channel->efx, &channel->eventq.buf); +} + +void efx_mcdi_ev_fini(struct efx_channel *channel) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_EVQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = channel->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_EVQ_IN_INSTANCE, channel->channel); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_EVQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_EVQ, MC_CMD_FINI_EVQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_tx_init(struct efx_tx_queue *tx_queue, bool tso_v2) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_INIT_TXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / + EFX_BUF_SIZE)); + bool csum_offload = tx_queue->queue & EFX_TXQ_TYPE_OFFLOAD; + size_t entries = tx_queue->txd.buf.len / EFX_BUF_SIZE; + struct efx_channel *channel = tx_queue->channel; + struct efx_nic *efx = tx_queue->efx; + struct efx_ef10_nic_data *nic_data; + dma_addr_t dma_addr; + size_t inlen; + int rc, i; + + BUILD_BUG_ON(MC_CMD_INIT_TXQ_OUT_LEN != 0); + + nic_data = efx->nic_data; + + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_SIZE, tx_queue->ptr_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_TARGET_EVQ, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_LABEL, tx_queue->queue); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_INSTANCE, tx_queue->queue); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_OWNER_ID, 0); + MCDI_SET_DWORD(inbuf, INIT_TXQ_IN_PORT_ID, nic_data->vport_id); + + dma_addr = tx_queue->txd.buf.dma_addr; + + netif_dbg(efx, hw, efx->net_dev, "pushing TXQ %d. %zu entries (%llx)\n", + tx_queue->queue, entries, (u64)dma_addr); + + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_TXQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_TXQ_IN_LEN(entries); + + do { + MCDI_POPULATE_DWORD_4(inbuf, INIT_TXQ_IN_FLAGS, + /* This flag was removed from mcdi_pcol.h for + * the non-_EXT version of INIT_TXQ. However, + * firmware still honours it. + */ + INIT_TXQ_EXT_IN_FLAG_TSOV2_EN, tso_v2, + INIT_TXQ_IN_FLAG_IP_CSUM_DIS, !csum_offload, + INIT_TXQ_IN_FLAG_TCP_CSUM_DIS, !csum_offload, + INIT_TXQ_EXT_IN_FLAG_TIMESTAMP, + tx_queue->timestamping); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_INIT_TXQ, inbuf, inlen, + NULL, 0, NULL); + if (rc == -ENOSPC && tso_v2) { + /* Retry without TSOv2 if we're short on contexts. */ + tso_v2 = false; + netif_warn(efx, probe, efx->net_dev, + "TSOv2 context not available to segment in " + "hardware. TCP performance may be reduced.\n" + ); + } else if (rc) { + efx_mcdi_display_error(efx, MC_CMD_INIT_TXQ, + MC_CMD_INIT_TXQ_EXT_IN_LEN, + NULL, 0, rc); + goto fail; + } + } while (rc); + + return 0; + +fail: + return rc; +} + +void efx_mcdi_tx_remove(struct efx_tx_queue *tx_queue) +{ + efx_nic_free_buffer(tx_queue->efx, &tx_queue->txd.buf); +} + +void efx_mcdi_tx_fini(struct efx_tx_queue *tx_queue) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_TXQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = tx_queue->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_TXQ_IN_INSTANCE, + tx_queue->queue); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_TXQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_TXQ, MC_CMD_FINI_TXQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_rx_probe(struct efx_rx_queue *rx_queue) +{ + return efx_nic_alloc_buffer(rx_queue->efx, &rx_queue->rxd.buf, + (rx_queue->ptr_mask + 1) * + sizeof(efx_qword_t), + GFP_KERNEL); +} + +void efx_mcdi_rx_init(struct efx_rx_queue *rx_queue) +{ + MCDI_DECLARE_BUF(inbuf, + MC_CMD_INIT_RXQ_IN_LEN(EFX_MAX_DMAQ_SIZE * 8 / + EFX_BUF_SIZE)); + struct efx_channel *channel = efx_rx_queue_channel(rx_queue); + size_t entries = rx_queue->rxd.buf.len / EFX_BUF_SIZE; + struct efx_nic *efx = rx_queue->efx; + struct efx_ef10_nic_data *nic_data = efx->nic_data; + dma_addr_t dma_addr; + size_t inlen; + int rc; + int i; + BUILD_BUG_ON(MC_CMD_INIT_RXQ_OUT_LEN != 0); + + rx_queue->scatter_n = 0; + rx_queue->scatter_len = 0; + + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_SIZE, rx_queue->ptr_mask + 1); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_TARGET_EVQ, channel->channel); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_LABEL, efx_rx_queue_index(rx_queue)); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_INSTANCE, + efx_rx_queue_index(rx_queue)); + MCDI_POPULATE_DWORD_2(inbuf, INIT_RXQ_IN_FLAGS, + INIT_RXQ_IN_FLAG_PREFIX, 1, + INIT_RXQ_IN_FLAG_TIMESTAMP, 1); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_OWNER_ID, 0); + MCDI_SET_DWORD(inbuf, INIT_RXQ_IN_PORT_ID, nic_data->vport_id); + + dma_addr = rx_queue->rxd.buf.dma_addr; + + netif_dbg(efx, hw, efx->net_dev, "pushing RXQ %d. %zu entries (%llx)\n", + efx_rx_queue_index(rx_queue), entries, (u64)dma_addr); + + for (i = 0; i < entries; ++i) { + MCDI_SET_ARRAY_QWORD(inbuf, INIT_RXQ_IN_DMA_ADDR, i, dma_addr); + dma_addr += EFX_BUF_SIZE; + } + + inlen = MC_CMD_INIT_RXQ_IN_LEN(entries); + + rc = efx_mcdi_rpc(efx, MC_CMD_INIT_RXQ, inbuf, inlen, + NULL, 0, NULL); + if (rc) + netdev_WARN(efx->net_dev, "failed to initialise RXQ %d\n", + efx_rx_queue_index(rx_queue)); +} + +void efx_mcdi_rx_remove(struct efx_rx_queue *rx_queue) +{ + efx_nic_free_buffer(rx_queue->efx, &rx_queue->rxd.buf); +} + +void efx_mcdi_rx_fini(struct efx_rx_queue *rx_queue) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_FINI_RXQ_IN_LEN); + MCDI_DECLARE_BUF_ERR(outbuf); + struct efx_nic *efx = rx_queue->efx; + size_t outlen; + int rc; + + MCDI_SET_DWORD(inbuf, FINI_RXQ_IN_INSTANCE, + efx_rx_queue_index(rx_queue)); + + rc = efx_mcdi_rpc_quiet(efx, MC_CMD_FINI_RXQ, inbuf, sizeof(inbuf), + outbuf, sizeof(outbuf), &outlen); + + if (rc && rc != -EALREADY) + goto fail; + + return; + +fail: + efx_mcdi_display_error(efx, MC_CMD_FINI_RXQ, MC_CMD_FINI_RXQ_IN_LEN, + outbuf, outlen, rc); +} + +int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode) +{ + switch (vi_window_mode) { + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_8K: + efx->vi_stride = 8192; + break; + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_16K: + efx->vi_stride = 16384; + break; + case MC_CMD_GET_CAPABILITIES_V3_OUT_VI_WINDOW_MODE_64K: + efx->vi_stride = 65536; + break; + default: + netif_err(efx, probe, efx->net_dev, + "Unrecognised VI window mode %d\n", + vi_window_mode); + return -EIO; + } + netif_dbg(efx, probe, efx->net_dev, "vi_stride = %u\n", + efx->vi_stride); + return 0; +} + +int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_FUNCTION_INFO_OUT_LEN); + size_t outlen; + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_FUNCTION_INFO, NULL, 0, outbuf, + sizeof(outbuf), &outlen); + if (rc) + return rc; + if (outlen < sizeof(outbuf)) + return -EIO; + + *pf_index = MCDI_DWORD(outbuf, GET_FUNCTION_INFO_OUT_PF); + return 0; +} diff --git a/drivers/net/ethernet/sfc/mcdi_functions.h b/drivers/net/ethernet/sfc/mcdi_functions.h new file mode 100644 index 000000000000..ca4a5ac1a66b --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_functions.h @@ -0,0 +1,32 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ +#ifndef EFX_MCDI_FUNCTIONS_H +#define EFX_MCDI_FUNCTIONS_H + +int efx_mcdi_alloc_vis(struct efx_nic *efx, unsigned int min_vis, + unsigned int max_vis, unsigned int *vi_base, + unsigned int *allocated_vis); +int efx_mcdi_free_vis(struct efx_nic *efx); + +int efx_mcdi_ev_probe(struct efx_channel *channel); +int efx_mcdi_ev_init(struct efx_channel *channel, bool v1_cut_thru, bool v2); +void efx_mcdi_ev_remove(struct efx_channel *channel); +void efx_mcdi_ev_fini(struct efx_channel *channel); +int efx_mcdi_tx_init(struct efx_tx_queue *tx_queue, bool tso_v2); +void efx_mcdi_tx_remove(struct efx_tx_queue *tx_queue); +void efx_mcdi_tx_fini(struct efx_tx_queue *tx_queue); +int efx_mcdi_rx_probe(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_init(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_remove(struct efx_rx_queue *rx_queue); +void efx_mcdi_rx_fini(struct efx_rx_queue *rx_queue); +int efx_mcdi_window_mode_to_stride(struct efx_nic *efx, u8 vi_window_mode); +int efx_get_pf_index(struct efx_nic *efx, unsigned int *pf_index); + +#endif diff --git a/drivers/net/ethernet/sfc/mcdi_port.c b/drivers/net/ethernet/sfc/mcdi_port.c index fb7cde4980ed..ab5227b13ae6 100644 --- a/drivers/net/ethernet/sfc/mcdi_port.c +++ b/drivers/net/ethernet/sfc/mcdi_port.c @@ -14,106 +14,7 @@ #include "mcdi_pcol.h" #include "nic.h" #include "selftest.h" - -struct efx_mcdi_phy_data { - u32 flags; - u32 type; - u32 supported_cap; - u32 channel; - u32 port; - u32 stats_mask; - u8 name[20]; - u32 media; - u32 mmd_mask; - u8 revision[20]; - u32 forced_cap; -}; - -static int -efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_CFG_OUT_LEN); - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_IN_LEN != 0); - BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_OUT_NAME_LEN != sizeof(cfg->name)); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_CFG, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - goto fail; - - if (outlen < MC_CMD_GET_PHY_CFG_OUT_LEN) { - rc = -EIO; - goto fail; - } - - cfg->flags = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_FLAGS); - cfg->type = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_TYPE); - cfg->supported_cap = - MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_SUPPORTED_CAP); - cfg->channel = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_CHANNEL); - cfg->port = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_PRT); - cfg->stats_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_STATS_MASK); - memcpy(cfg->name, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_NAME), - sizeof(cfg->name)); - cfg->media = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MEDIA_TYPE); - cfg->mmd_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MMD_MASK); - memcpy(cfg->revision, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_REVISION), - sizeof(cfg->revision)); - - return 0; - -fail: - netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); - return rc; -} - -static int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, - u32 flags, u32 loopback_mode, - u32 loopback_speed) -{ - MCDI_DECLARE_BUF(inbuf, MC_CMD_SET_LINK_IN_LEN); - int rc; - - BUILD_BUG_ON(MC_CMD_SET_LINK_OUT_LEN != 0); - - MCDI_SET_DWORD(inbuf, SET_LINK_IN_CAP, capabilities); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_FLAGS, flags); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_MODE, loopback_mode); - MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_SPEED, loopback_speed); - - rc = efx_mcdi_rpc(efx, MC_CMD_SET_LINK, inbuf, sizeof(inbuf), - NULL, 0, NULL); - return rc; -} - -static int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LOOPBACK_MODES_OUT_LEN); - size_t outlen; - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LOOPBACK_MODES, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - goto fail; - - if (outlen < (MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_OFST + - MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_LEN)) { - rc = -EIO; - goto fail; - } - - *loopback_modes = MCDI_QWORD(outbuf, GET_LOOPBACK_MODES_OUT_SUGGESTED); - - return 0; - -fail: - netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); - return rc; -} +#include "mcdi_port_common.h" static int efx_mcdi_mdio_read(struct net_device *net_dev, int prtad, int devad, u16 addr) @@ -168,246 +69,6 @@ static int efx_mcdi_mdio_write(struct net_device *net_dev, return 0; } -static void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset) -{ - #define SET_BIT(name) __set_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ - linkset) - - bitmap_zero(linkset, __ETHTOOL_LINK_MODE_MASK_NBITS); - switch (media) { - case MC_CMD_MEDIA_KX4: - SET_BIT(Backplane); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseKX_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseKX4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) - SET_BIT(40000baseKR4_Full); - break; - - case MC_CMD_MEDIA_XFP: - case MC_CMD_MEDIA_SFP_PLUS: - case MC_CMD_MEDIA_QSFP_PLUS: - SET_BIT(FIBRE); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) - SET_BIT(40000baseCR4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_100000FDX_LBN)) - SET_BIT(100000baseCR4_Full); - if (cap & (1 << MC_CMD_PHY_CAP_25000FDX_LBN)) - SET_BIT(25000baseCR_Full); - if (cap & (1 << MC_CMD_PHY_CAP_50000FDX_LBN)) - SET_BIT(50000baseCR2_Full); - break; - - case MC_CMD_MEDIA_BASE_T: - SET_BIT(TP); - if (cap & (1 << MC_CMD_PHY_CAP_10HDX_LBN)) - SET_BIT(10baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_10FDX_LBN)) - SET_BIT(10baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_100HDX_LBN)) - SET_BIT(100baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_100FDX_LBN)) - SET_BIT(100baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_1000HDX_LBN)) - SET_BIT(1000baseT_Half); - if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) - SET_BIT(1000baseT_Full); - if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) - SET_BIT(10000baseT_Full); - break; - } - - if (cap & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) - SET_BIT(Pause); - if (cap & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) - SET_BIT(Asym_Pause); - if (cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) - SET_BIT(Autoneg); - - #undef SET_BIT -} - -static u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset) -{ - u32 result = 0; - - #define TEST_BIT(name) test_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ - linkset) - - if (TEST_BIT(10baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_10HDX_LBN); - if (TEST_BIT(10baseT_Full)) - result |= (1 << MC_CMD_PHY_CAP_10FDX_LBN); - if (TEST_BIT(100baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_100HDX_LBN); - if (TEST_BIT(100baseT_Full)) - result |= (1 << MC_CMD_PHY_CAP_100FDX_LBN); - if (TEST_BIT(1000baseT_Half)) - result |= (1 << MC_CMD_PHY_CAP_1000HDX_LBN); - if (TEST_BIT(1000baseT_Full) || TEST_BIT(1000baseKX_Full)) - result |= (1 << MC_CMD_PHY_CAP_1000FDX_LBN); - if (TEST_BIT(10000baseT_Full) || TEST_BIT(10000baseKX4_Full)) - result |= (1 << MC_CMD_PHY_CAP_10000FDX_LBN); - if (TEST_BIT(40000baseCR4_Full) || TEST_BIT(40000baseKR4_Full)) - result |= (1 << MC_CMD_PHY_CAP_40000FDX_LBN); - if (TEST_BIT(100000baseCR4_Full)) - result |= (1 << MC_CMD_PHY_CAP_100000FDX_LBN); - if (TEST_BIT(25000baseCR_Full)) - result |= (1 << MC_CMD_PHY_CAP_25000FDX_LBN); - if (TEST_BIT(50000baseCR2_Full)) - result |= (1 << MC_CMD_PHY_CAP_50000FDX_LBN); - if (TEST_BIT(Pause)) - result |= (1 << MC_CMD_PHY_CAP_PAUSE_LBN); - if (TEST_BIT(Asym_Pause)) - result |= (1 << MC_CMD_PHY_CAP_ASYM_LBN); - if (TEST_BIT(Autoneg)) - result |= (1 << MC_CMD_PHY_CAP_AN_LBN); - - #undef TEST_BIT - - return result; -} - -static u32 efx_get_mcdi_phy_flags(struct efx_nic *efx) -{ - struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; - enum efx_phy_mode mode, supported; - u32 flags; - - /* TODO: Advertise the capabilities supported by this PHY */ - supported = 0; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_TXDIS_LBN)) - supported |= PHY_MODE_TX_DISABLED; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_LOWPOWER_LBN)) - supported |= PHY_MODE_LOW_POWER; - if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_POWEROFF_LBN)) - supported |= PHY_MODE_OFF; - - mode = efx->phy_mode & supported; - - flags = 0; - if (mode & PHY_MODE_TX_DISABLED) - flags |= (1 << MC_CMD_SET_LINK_IN_TXDIS_LBN); - if (mode & PHY_MODE_LOW_POWER) - flags |= (1 << MC_CMD_SET_LINK_IN_LOWPOWER_LBN); - if (mode & PHY_MODE_OFF) - flags |= (1 << MC_CMD_SET_LINK_IN_POWEROFF_LBN); - - return flags; -} - -static u8 mcdi_to_ethtool_media(u32 media) -{ - switch (media) { - case MC_CMD_MEDIA_XAUI: - case MC_CMD_MEDIA_CX4: - case MC_CMD_MEDIA_KX4: - return PORT_OTHER; - - case MC_CMD_MEDIA_XFP: - case MC_CMD_MEDIA_SFP_PLUS: - case MC_CMD_MEDIA_QSFP_PLUS: - return PORT_FIBRE; - - case MC_CMD_MEDIA_BASE_T: - return PORT_TP; - - default: - return PORT_OTHER; - } -} - -static void efx_mcdi_phy_decode_link(struct efx_nic *efx, - struct efx_link_state *link_state, - u32 speed, u32 flags, u32 fcntl) -{ - switch (fcntl) { - case MC_CMD_FCNTL_AUTO: - WARN_ON(1); /* This is not a link mode */ - link_state->fc = EFX_FC_AUTO | EFX_FC_TX | EFX_FC_RX; - break; - case MC_CMD_FCNTL_BIDIR: - link_state->fc = EFX_FC_TX | EFX_FC_RX; - break; - case MC_CMD_FCNTL_RESPOND: - link_state->fc = EFX_FC_RX; - break; - default: - WARN_ON(1); - /* Fall through */ - case MC_CMD_FCNTL_OFF: - link_state->fc = 0; - break; - } - - link_state->up = !!(flags & (1 << MC_CMD_GET_LINK_OUT_LINK_UP_LBN)); - link_state->fd = !!(flags & (1 << MC_CMD_GET_LINK_OUT_FULL_DUPLEX_LBN)); - link_state->speed = speed; -} - -/* The semantics of the ethtool FEC mode bitmask are not well defined, - * particularly the meaning of combinations of bits. Which means we get to - * define our own semantics, as follows: - * OFF overrides any other bits, and means "disable all FEC" (with the - * exception of 25G KR4/CR4, where it is not possible to reject it if AN - * partner requests it). - * AUTO on its own means use cable requirements and link partner autoneg with - * fw-default preferences for the cable type. - * AUTO and either RS or BASER means use the specified FEC type if cable and - * link partner support it, otherwise autoneg/fw-default. - * RS or BASER alone means use the specified FEC type if cable and link partner - * support it and either requests it, otherwise no FEC. - * Both RS and BASER (whether AUTO or not) means use FEC if cable and link - * partner support it, preferring RS to BASER. - */ -static u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap) -{ - u32 ret = 0; - - if (ethtool_cap & ETHTOOL_FEC_OFF) - return 0; - - if (ethtool_cap & ETHTOOL_FEC_AUTO) - ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_RS_FEC_LBN); - if (ethtool_cap & ETHTOOL_FEC_RS) - ret |= (1 << MC_CMD_PHY_CAP_RS_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN); - if (ethtool_cap & ETHTOOL_FEC_BASER) - ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | - (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN) | - (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN); - return ret; -} - -/* Invert ethtool_fec_caps_to_mcdi. There are two combinations that function - * can never produce, (baser xor rs) and neither req; the implementation below - * maps both of those to AUTO. This should never matter, and it's not clear - * what a better mapping would be anyway. - */ -static u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g) -{ - bool rs = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_LBN), - rs_req = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN), - baser = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) - : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN), - baser_req = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN) - : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN); - - if (!baser && !rs) - return ETHTOOL_FEC_OFF; - return (rs_req ? ETHTOOL_FEC_RS : 0) | - (baser_req ? ETHTOOL_FEC_BASER : 0) | - (baser == baser_req && rs == rs_req ? 0 : ETHTOOL_FEC_AUTO); -} - static int efx_mcdi_phy_probe(struct efx_nic *efx) { struct efx_mcdi_phy_data *phy_data; @@ -527,58 +188,6 @@ int efx_mcdi_port_reconfigure(struct efx_nic *efx) efx->loopback_mode, 0); } -/* Verify that the forced flow control settings (!EFX_FC_AUTO) are - * supported by the link partner. Warn the user if this isn't the case - */ -static void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa) -{ - struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; - u32 rmtadv; - - /* The link partner capabilities are only relevant if the - * link supports flow control autonegotiation */ - if (~phy_cfg->supported_cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) - return; - - /* If flow control autoneg is supported and enabled, then fine */ - if (efx->wanted_fc & EFX_FC_AUTO) - return; - - rmtadv = 0; - if (lpa & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) - rmtadv |= ADVERTISED_Pause; - if (lpa & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) - rmtadv |= ADVERTISED_Asym_Pause; - - if ((efx->wanted_fc & EFX_FC_TX) && rmtadv == ADVERTISED_Asym_Pause) - netif_err(efx, link, efx->net_dev, - "warning: link partner doesn't support pause frames"); -} - -static bool efx_mcdi_phy_poll(struct efx_nic *efx) -{ - struct efx_link_state old_state = efx->link_state; - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); - int rc; - - WARN_ON(!mutex_is_locked(&efx->mac_lock)); - - BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, - outbuf, sizeof(outbuf), NULL); - if (rc) - efx->link_state.up = false; - else - efx_mcdi_phy_decode_link( - efx, &efx->link_state, - MCDI_DWORD(outbuf, GET_LINK_OUT_LINK_SPEED), - MCDI_DWORD(outbuf, GET_LINK_OUT_FLAGS), - MCDI_DWORD(outbuf, GET_LINK_OUT_FCNTL)); - - return !efx_link_state_equal(&efx->link_state, &old_state); -} - static void efx_mcdi_phy_remove(struct efx_nic *efx) { struct efx_mcdi_phy_data *phy_data = efx->phy_data; @@ -666,58 +275,6 @@ efx_mcdi_phy_set_link_ksettings(struct efx_nic *efx, return 0; } -static int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, - struct ethtool_fecparam *fec) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_V2_LEN); - u32 caps, active, speed; /* MCDI format */ - bool is_25g = false; - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); - rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - return rc; - if (outlen < MC_CMD_GET_LINK_OUT_V2_LEN) - return -EOPNOTSUPP; - - /* behaviour for 25G/50G links depends on 25G BASER bit */ - speed = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_LINK_SPEED); - is_25g = speed == 25000 || speed == 50000; - - caps = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_CAP); - fec->fec = mcdi_fec_caps_to_ethtool(caps, is_25g); - /* BASER is never supported on 100G */ - if (speed == 100000) - fec->fec &= ~ETHTOOL_FEC_BASER; - - active = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_FEC_TYPE); - switch (active) { - case MC_CMD_FEC_NONE: - fec->active_fec = ETHTOOL_FEC_OFF; - break; - case MC_CMD_FEC_BASER: - fec->active_fec = ETHTOOL_FEC_BASER; - break; - case MC_CMD_FEC_RS: - fec->active_fec = ETHTOOL_FEC_RS; - break; - default: - netif_warn(efx, hw, efx->net_dev, - "Firmware reports unrecognised FEC_TYPE %u\n", - active); - /* We don't know what firmware has picked. AUTO is as good a - * "can't happen" value as any other. - */ - fec->active_fec = ETHTOOL_FEC_AUTO; - break; - } - - return 0; -} - static int efx_mcdi_phy_set_fecparam(struct efx_nic *efx, const struct ethtool_fecparam *fec) { @@ -745,27 +302,6 @@ static int efx_mcdi_phy_set_fecparam(struct efx_nic *efx, return 0; } -static int efx_mcdi_phy_test_alive(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_STATE_OUT_LEN); - size_t outlen; - int rc; - - BUILD_BUG_ON(MC_CMD_GET_PHY_STATE_IN_LEN != 0); - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_STATE, NULL, 0, - outbuf, sizeof(outbuf), &outlen); - if (rc) - return rc; - - if (outlen < MC_CMD_GET_PHY_STATE_OUT_LEN) - return -EIO; - if (MCDI_DWORD(outbuf, GET_PHY_STATE_OUT_STATE) != MC_CMD_PHY_STATE_OK) - return -EINVAL; - - return 0; -} - static const char *const mcdi_sft9001_cable_diag_names[] = { "cable.pairA.length", "cable.pairB.length", @@ -1139,84 +675,6 @@ u32 efx_mcdi_phy_get_caps(struct efx_nic *efx) return phy_data->supported_cap; } -static unsigned int efx_mcdi_event_link_speed[] = { - [MCDI_EVENT_LINKCHANGE_SPEED_100M] = 100, - [MCDI_EVENT_LINKCHANGE_SPEED_1G] = 1000, - [MCDI_EVENT_LINKCHANGE_SPEED_10G] = 10000, - [MCDI_EVENT_LINKCHANGE_SPEED_40G] = 40000, - [MCDI_EVENT_LINKCHANGE_SPEED_25G] = 25000, - [MCDI_EVENT_LINKCHANGE_SPEED_50G] = 50000, - [MCDI_EVENT_LINKCHANGE_SPEED_100G] = 100000, -}; - -void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev) -{ - u32 flags, fcntl, speed, lpa; - - speed = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_SPEED); - EFX_WARN_ON_PARANOID(speed >= ARRAY_SIZE(efx_mcdi_event_link_speed)); - speed = efx_mcdi_event_link_speed[speed]; - - flags = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LINK_FLAGS); - fcntl = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_FCNTL); - lpa = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LP_CAP); - - /* efx->link_state is only modified by efx_mcdi_phy_get_link(), - * which is only run after flushing the event queues. Therefore, it - * is safe to modify the link state outside of the mac_lock here. - */ - efx_mcdi_phy_decode_link(efx, &efx->link_state, speed, flags, fcntl); - - efx_mcdi_phy_check_fcntl(efx, lpa); - - efx_link_status_changed(efx); -} - -int efx_mcdi_set_mac(struct efx_nic *efx) -{ - u32 fcntl; - MCDI_DECLARE_BUF(cmdbytes, MC_CMD_SET_MAC_IN_LEN); - - BUILD_BUG_ON(MC_CMD_SET_MAC_OUT_LEN != 0); - - /* This has no effect on EF10 */ - ether_addr_copy(MCDI_PTR(cmdbytes, SET_MAC_IN_ADDR), - efx->net_dev->dev_addr); - - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_MTU, - EFX_MAX_FRAME_LEN(efx->net_dev->mtu)); - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_DRAIN, 0); - - /* Set simple MAC filter for Siena */ - MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_REJECT, - SET_MAC_IN_REJECT_UNCST, efx->unicast_filter); - - MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_FLAGS, - SET_MAC_IN_FLAG_INCLUDE_FCS, - !!(efx->net_dev->features & NETIF_F_RXFCS)); - - switch (efx->wanted_fc) { - case EFX_FC_RX | EFX_FC_TX: - fcntl = MC_CMD_FCNTL_BIDIR; - break; - case EFX_FC_RX: - fcntl = MC_CMD_FCNTL_RESPOND; - break; - default: - fcntl = MC_CMD_FCNTL_OFF; - break; - } - if (efx->wanted_fc & EFX_FC_AUTO) - fcntl = MC_CMD_FCNTL_AUTO; - if (efx->fc_disable) - fcntl = MC_CMD_FCNTL_OFF; - - MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_FCNTL, fcntl); - - return efx_mcdi_rpc(efx, MC_CMD_SET_MAC, cmdbytes, sizeof(cmdbytes), - NULL, 0, NULL); -} - bool efx_mcdi_mac_check_fault(struct efx_nic *efx) { MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); @@ -1348,17 +806,3 @@ void efx_mcdi_port_remove(struct efx_nic *efx) efx->phy_op->remove(efx); efx_nic_free_buffer(efx, &efx->stats_buffer); } - -/* Get physical port number (EF10 only; on Siena it is same as PF number) */ -int efx_mcdi_port_get_number(struct efx_nic *efx) -{ - MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PORT_ASSIGNMENT_OUT_LEN); - int rc; - - rc = efx_mcdi_rpc(efx, MC_CMD_GET_PORT_ASSIGNMENT, NULL, 0, - outbuf, sizeof(outbuf), NULL); - if (rc) - return rc; - - return MCDI_DWORD(outbuf, GET_PORT_ASSIGNMENT_OUT_PORT); -} diff --git a/drivers/net/ethernet/sfc/mcdi_port_common.c b/drivers/net/ethernet/sfc/mcdi_port_common.c new file mode 100644 index 000000000000..a6a072ba46d3 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_port_common.c @@ -0,0 +1,568 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "mcdi_port_common.h" +#include "efx_common.h" + +int efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_CFG_OUT_LEN); + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_IN_LEN != 0); + BUILD_BUG_ON(MC_CMD_GET_PHY_CFG_OUT_NAME_LEN != sizeof(cfg->name)); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_CFG, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + goto fail; + + if (outlen < MC_CMD_GET_PHY_CFG_OUT_LEN) { + rc = -EIO; + goto fail; + } + + cfg->flags = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_FLAGS); + cfg->type = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_TYPE); + cfg->supported_cap = + MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_SUPPORTED_CAP); + cfg->channel = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_CHANNEL); + cfg->port = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_PRT); + cfg->stats_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_STATS_MASK); + memcpy(cfg->name, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_NAME), + sizeof(cfg->name)); + cfg->media = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MEDIA_TYPE); + cfg->mmd_mask = MCDI_DWORD(outbuf, GET_PHY_CFG_OUT_MMD_MASK); + memcpy(cfg->revision, MCDI_PTR(outbuf, GET_PHY_CFG_OUT_REVISION), + sizeof(cfg->revision)); + + return 0; + +fail: + netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); + return rc; +} + +void efx_link_set_advertising(struct efx_nic *efx, + const unsigned long *advertising) +{ + memcpy(efx->link_advertising, advertising, + sizeof(__ETHTOOL_DECLARE_LINK_MODE_MASK())); + + efx->link_advertising[0] |= ADVERTISED_Autoneg; + if (advertising[0] & ADVERTISED_Pause) + efx->wanted_fc |= (EFX_FC_TX | EFX_FC_RX); + else + efx->wanted_fc &= ~(EFX_FC_TX | EFX_FC_RX); + if (advertising[0] & ADVERTISED_Asym_Pause) + efx->wanted_fc ^= EFX_FC_TX; +} + +int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, + u32 flags, u32 loopback_mode, u32 loopback_speed) +{ + MCDI_DECLARE_BUF(inbuf, MC_CMD_SET_LINK_IN_LEN); + int rc; + + BUILD_BUG_ON(MC_CMD_SET_LINK_OUT_LEN != 0); + + MCDI_SET_DWORD(inbuf, SET_LINK_IN_CAP, capabilities); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_FLAGS, flags); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_MODE, loopback_mode); + MCDI_SET_DWORD(inbuf, SET_LINK_IN_LOOPBACK_SPEED, loopback_speed); + + rc = efx_mcdi_rpc(efx, MC_CMD_SET_LINK, inbuf, sizeof(inbuf), + NULL, 0, NULL); + return rc; +} + +int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LOOPBACK_MODES_OUT_LEN); + size_t outlen; + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LOOPBACK_MODES, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + goto fail; + + if (outlen < (MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_OFST + + MC_CMD_GET_LOOPBACK_MODES_OUT_SUGGESTED_LEN)) { + rc = -EIO; + goto fail; + } + + *loopback_modes = MCDI_QWORD(outbuf, GET_LOOPBACK_MODES_OUT_SUGGESTED); + + return 0; + +fail: + netif_err(efx, hw, efx->net_dev, "%s: failed rc=%d\n", __func__, rc); + return rc; +} + +void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset) +{ + #define SET_BIT(name) __set_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ + linkset) + + bitmap_zero(linkset, __ETHTOOL_LINK_MODE_MASK_NBITS); + switch (media) { + case MC_CMD_MEDIA_KX4: + SET_BIT(Backplane); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseKX_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseKX4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) + SET_BIT(40000baseKR4_Full); + break; + + case MC_CMD_MEDIA_XFP: + case MC_CMD_MEDIA_SFP_PLUS: + case MC_CMD_MEDIA_QSFP_PLUS: + SET_BIT(FIBRE); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) + SET_BIT(40000baseCR4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_100000FDX_LBN)) + SET_BIT(100000baseCR4_Full); + if (cap & (1 << MC_CMD_PHY_CAP_25000FDX_LBN)) + SET_BIT(25000baseCR_Full); + if (cap & (1 << MC_CMD_PHY_CAP_50000FDX_LBN)) + SET_BIT(50000baseCR2_Full); + break; + + case MC_CMD_MEDIA_BASE_T: + SET_BIT(TP); + if (cap & (1 << MC_CMD_PHY_CAP_10HDX_LBN)) + SET_BIT(10baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_10FDX_LBN)) + SET_BIT(10baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_100HDX_LBN)) + SET_BIT(100baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_100FDX_LBN)) + SET_BIT(100baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_1000HDX_LBN)) + SET_BIT(1000baseT_Half); + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + SET_BIT(1000baseT_Full); + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + SET_BIT(10000baseT_Full); + break; + } + + if (cap & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) + SET_BIT(Pause); + if (cap & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) + SET_BIT(Asym_Pause); + if (cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) + SET_BIT(Autoneg); + + #undef SET_BIT +} + +u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset) +{ + u32 result = 0; + + #define TEST_BIT(name) test_bit(ETHTOOL_LINK_MODE_ ## name ## _BIT, \ + linkset) + + if (TEST_BIT(10baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_10HDX_LBN); + if (TEST_BIT(10baseT_Full)) + result |= (1 << MC_CMD_PHY_CAP_10FDX_LBN); + if (TEST_BIT(100baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_100HDX_LBN); + if (TEST_BIT(100baseT_Full)) + result |= (1 << MC_CMD_PHY_CAP_100FDX_LBN); + if (TEST_BIT(1000baseT_Half)) + result |= (1 << MC_CMD_PHY_CAP_1000HDX_LBN); + if (TEST_BIT(1000baseT_Full) || TEST_BIT(1000baseKX_Full)) + result |= (1 << MC_CMD_PHY_CAP_1000FDX_LBN); + if (TEST_BIT(10000baseT_Full) || TEST_BIT(10000baseKX4_Full)) + result |= (1 << MC_CMD_PHY_CAP_10000FDX_LBN); + if (TEST_BIT(40000baseCR4_Full) || TEST_BIT(40000baseKR4_Full)) + result |= (1 << MC_CMD_PHY_CAP_40000FDX_LBN); + if (TEST_BIT(100000baseCR4_Full)) + result |= (1 << MC_CMD_PHY_CAP_100000FDX_LBN); + if (TEST_BIT(25000baseCR_Full)) + result |= (1 << MC_CMD_PHY_CAP_25000FDX_LBN); + if (TEST_BIT(50000baseCR2_Full)) + result |= (1 << MC_CMD_PHY_CAP_50000FDX_LBN); + if (TEST_BIT(Pause)) + result |= (1 << MC_CMD_PHY_CAP_PAUSE_LBN); + if (TEST_BIT(Asym_Pause)) + result |= (1 << MC_CMD_PHY_CAP_ASYM_LBN); + if (TEST_BIT(Autoneg)) + result |= (1 << MC_CMD_PHY_CAP_AN_LBN); + + #undef TEST_BIT + + return result; +} + +u32 efx_get_mcdi_phy_flags(struct efx_nic *efx) +{ + struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; + enum efx_phy_mode mode, supported; + u32 flags; + + /* TODO: Advertise the capabilities supported by this PHY */ + supported = 0; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_TXDIS_LBN)) + supported |= PHY_MODE_TX_DISABLED; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_LOWPOWER_LBN)) + supported |= PHY_MODE_LOW_POWER; + if (phy_cfg->flags & (1 << MC_CMD_GET_PHY_CFG_OUT_POWEROFF_LBN)) + supported |= PHY_MODE_OFF; + + mode = efx->phy_mode & supported; + + flags = 0; + if (mode & PHY_MODE_TX_DISABLED) + flags |= (1 << MC_CMD_SET_LINK_IN_TXDIS_LBN); + if (mode & PHY_MODE_LOW_POWER) + flags |= (1 << MC_CMD_SET_LINK_IN_LOWPOWER_LBN); + if (mode & PHY_MODE_OFF) + flags |= (1 << MC_CMD_SET_LINK_IN_POWEROFF_LBN); + + return flags; +} + +u8 mcdi_to_ethtool_media(u32 media) +{ + switch (media) { + case MC_CMD_MEDIA_XAUI: + case MC_CMD_MEDIA_CX4: + case MC_CMD_MEDIA_KX4: + return PORT_OTHER; + + case MC_CMD_MEDIA_XFP: + case MC_CMD_MEDIA_SFP_PLUS: + case MC_CMD_MEDIA_QSFP_PLUS: + return PORT_FIBRE; + + case MC_CMD_MEDIA_BASE_T: + return PORT_TP; + + default: + return PORT_OTHER; + } +} + +void efx_mcdi_phy_decode_link(struct efx_nic *efx, + struct efx_link_state *link_state, + u32 speed, u32 flags, u32 fcntl) +{ + switch (fcntl) { + case MC_CMD_FCNTL_AUTO: + WARN_ON(1); /* This is not a link mode */ + link_state->fc = EFX_FC_AUTO | EFX_FC_TX | EFX_FC_RX; + break; + case MC_CMD_FCNTL_BIDIR: + link_state->fc = EFX_FC_TX | EFX_FC_RX; + break; + case MC_CMD_FCNTL_RESPOND: + link_state->fc = EFX_FC_RX; + break; + default: + WARN_ON(1); + /* Fall through */ + case MC_CMD_FCNTL_OFF: + link_state->fc = 0; + break; + } + + link_state->up = !!(flags & (1 << MC_CMD_GET_LINK_OUT_LINK_UP_LBN)); + link_state->fd = !!(flags & (1 << MC_CMD_GET_LINK_OUT_FULL_DUPLEX_LBN)); + link_state->speed = speed; +} + +/* The semantics of the ethtool FEC mode bitmask are not well defined, + * particularly the meaning of combinations of bits. Which means we get to + * define our own semantics, as follows: + * OFF overrides any other bits, and means "disable all FEC" (with the + * exception of 25G KR4/CR4, where it is not possible to reject it if AN + * partner requests it). + * AUTO on its own means use cable requirements and link partner autoneg with + * fw-default preferences for the cable type. + * AUTO and either RS or BASER means use the specified FEC type if cable and + * link partner support it, otherwise autoneg/fw-default. + * RS or BASER alone means use the specified FEC type if cable and link partner + * support it and either requests it, otherwise no FEC. + * Both RS and BASER (whether AUTO or not) means use FEC if cable and link + * partner support it, preferring RS to BASER. + */ +u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap) +{ + u32 ret = 0; + + if (ethtool_cap & ETHTOOL_FEC_OFF) + return 0; + + if (ethtool_cap & ETHTOOL_FEC_AUTO) + ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_RS_FEC_LBN); + if (ethtool_cap & ETHTOOL_FEC_RS) + ret |= (1 << MC_CMD_PHY_CAP_RS_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN); + if (ethtool_cap & ETHTOOL_FEC_BASER) + ret |= (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) | + (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN) | + (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN); + return ret; +} + +/* Invert ethtool_fec_caps_to_mcdi. There are two combinations that function + * can never produce, (baser xor rs) and neither req; the implementation below + * maps both of those to AUTO. This should never matter, and it's not clear + * what a better mapping would be anyway. + */ +u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g) +{ + bool rs = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_LBN), + rs_req = caps & (1 << MC_CMD_PHY_CAP_RS_FEC_REQUESTED_LBN), + baser = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_LBN) + : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_LBN), + baser_req = is_25g ? caps & (1 << MC_CMD_PHY_CAP_25G_BASER_FEC_REQUESTED_LBN) + : caps & (1 << MC_CMD_PHY_CAP_BASER_FEC_REQUESTED_LBN); + + if (!baser && !rs) + return ETHTOOL_FEC_OFF; + return (rs_req ? ETHTOOL_FEC_RS : 0) | + (baser_req ? ETHTOOL_FEC_BASER : 0) | + (baser == baser_req && rs == rs_req ? 0 : ETHTOOL_FEC_AUTO); +} + +/* Verify that the forced flow control settings (!EFX_FC_AUTO) are + * supported by the link partner. Warn the user if this isn't the case + */ +void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa) +{ + struct efx_mcdi_phy_data *phy_cfg = efx->phy_data; + u32 rmtadv; + + /* The link partner capabilities are only relevant if the + * link supports flow control autonegotiation + */ + if (~phy_cfg->supported_cap & (1 << MC_CMD_PHY_CAP_AN_LBN)) + return; + + /* If flow control autoneg is supported and enabled, then fine */ + if (efx->wanted_fc & EFX_FC_AUTO) + return; + + rmtadv = 0; + if (lpa & (1 << MC_CMD_PHY_CAP_PAUSE_LBN)) + rmtadv |= ADVERTISED_Pause; + if (lpa & (1 << MC_CMD_PHY_CAP_ASYM_LBN)) + rmtadv |= ADVERTISED_Asym_Pause; + + if ((efx->wanted_fc & EFX_FC_TX) && rmtadv == ADVERTISED_Asym_Pause) + netif_err(efx, link, efx->net_dev, + "warning: link partner doesn't support pause frames"); +} + +bool efx_mcdi_phy_poll(struct efx_nic *efx) +{ + struct efx_link_state old_state = efx->link_state; + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_LEN); + int rc; + + WARN_ON(!mutex_is_locked(&efx->mac_lock)); + + BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, + outbuf, sizeof(outbuf), NULL); + if (rc) + efx->link_state.up = false; + else + efx_mcdi_phy_decode_link( + efx, &efx->link_state, + MCDI_DWORD(outbuf, GET_LINK_OUT_LINK_SPEED), + MCDI_DWORD(outbuf, GET_LINK_OUT_FLAGS), + MCDI_DWORD(outbuf, GET_LINK_OUT_FCNTL)); + + return !efx_link_state_equal(&efx->link_state, &old_state); +} + +int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, struct ethtool_fecparam *fec) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_LINK_OUT_V2_LEN); + u32 caps, active, speed; /* MCDI format */ + bool is_25g = false; + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_LINK_IN_LEN != 0); + rc = efx_mcdi_rpc(efx, MC_CMD_GET_LINK, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + return rc; + if (outlen < MC_CMD_GET_LINK_OUT_V2_LEN) + return -EOPNOTSUPP; + + /* behaviour for 25G/50G links depends on 25G BASER bit */ + speed = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_LINK_SPEED); + is_25g = speed == 25000 || speed == 50000; + + caps = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_CAP); + fec->fec = mcdi_fec_caps_to_ethtool(caps, is_25g); + /* BASER is never supported on 100G */ + if (speed == 100000) + fec->fec &= ~ETHTOOL_FEC_BASER; + + active = MCDI_DWORD(outbuf, GET_LINK_OUT_V2_FEC_TYPE); + switch (active) { + case MC_CMD_FEC_NONE: + fec->active_fec = ETHTOOL_FEC_OFF; + break; + case MC_CMD_FEC_BASER: + fec->active_fec = ETHTOOL_FEC_BASER; + break; + case MC_CMD_FEC_RS: + fec->active_fec = ETHTOOL_FEC_RS; + break; + default: + netif_warn(efx, hw, efx->net_dev, + "Firmware reports unrecognised FEC_TYPE %u\n", + active); + /* We don't know what firmware has picked. AUTO is as good a + * "can't happen" value as any other. + */ + fec->active_fec = ETHTOOL_FEC_AUTO; + break; + } + + return 0; +} + +int efx_mcdi_phy_test_alive(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PHY_STATE_OUT_LEN); + size_t outlen; + int rc; + + BUILD_BUG_ON(MC_CMD_GET_PHY_STATE_IN_LEN != 0); + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PHY_STATE, NULL, 0, + outbuf, sizeof(outbuf), &outlen); + if (rc) + return rc; + + if (outlen < MC_CMD_GET_PHY_STATE_OUT_LEN) + return -EIO; + if (MCDI_DWORD(outbuf, GET_PHY_STATE_OUT_STATE) != MC_CMD_PHY_STATE_OK) + return -EINVAL; + + return 0; +} + +int efx_mcdi_set_mac(struct efx_nic *efx) +{ + u32 fcntl; + MCDI_DECLARE_BUF(cmdbytes, MC_CMD_SET_MAC_IN_LEN); + + BUILD_BUG_ON(MC_CMD_SET_MAC_OUT_LEN != 0); + + /* This has no effect on EF10 */ + ether_addr_copy(MCDI_PTR(cmdbytes, SET_MAC_IN_ADDR), + efx->net_dev->dev_addr); + + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_MTU, + EFX_MAX_FRAME_LEN(efx->net_dev->mtu)); + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_DRAIN, 0); + + /* Set simple MAC filter for Siena */ + MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_REJECT, + SET_MAC_IN_REJECT_UNCST, efx->unicast_filter); + + MCDI_POPULATE_DWORD_1(cmdbytes, SET_MAC_IN_FLAGS, + SET_MAC_IN_FLAG_INCLUDE_FCS, + !!(efx->net_dev->features & NETIF_F_RXFCS)); + + switch (efx->wanted_fc) { + case EFX_FC_RX | EFX_FC_TX: + fcntl = MC_CMD_FCNTL_BIDIR; + break; + case EFX_FC_RX: + fcntl = MC_CMD_FCNTL_RESPOND; + break; + default: + fcntl = MC_CMD_FCNTL_OFF; + break; + } + if (efx->wanted_fc & EFX_FC_AUTO) + fcntl = MC_CMD_FCNTL_AUTO; + if (efx->fc_disable) + fcntl = MC_CMD_FCNTL_OFF; + + MCDI_SET_DWORD(cmdbytes, SET_MAC_IN_FCNTL, fcntl); + + return efx_mcdi_rpc(efx, MC_CMD_SET_MAC, cmdbytes, sizeof(cmdbytes), + NULL, 0, NULL); +} + +/* Get physical port number (EF10 only; on Siena it is same as PF number) */ +int efx_mcdi_port_get_number(struct efx_nic *efx) +{ + MCDI_DECLARE_BUF(outbuf, MC_CMD_GET_PORT_ASSIGNMENT_OUT_LEN); + int rc; + + rc = efx_mcdi_rpc(efx, MC_CMD_GET_PORT_ASSIGNMENT, NULL, 0, + outbuf, sizeof(outbuf), NULL); + if (rc) + return rc; + + return MCDI_DWORD(outbuf, GET_PORT_ASSIGNMENT_OUT_PORT); +} + +static unsigned int efx_mcdi_event_link_speed[] = { + [MCDI_EVENT_LINKCHANGE_SPEED_100M] = 100, + [MCDI_EVENT_LINKCHANGE_SPEED_1G] = 1000, + [MCDI_EVENT_LINKCHANGE_SPEED_10G] = 10000, + [MCDI_EVENT_LINKCHANGE_SPEED_40G] = 40000, + [MCDI_EVENT_LINKCHANGE_SPEED_25G] = 25000, + [MCDI_EVENT_LINKCHANGE_SPEED_50G] = 50000, + [MCDI_EVENT_LINKCHANGE_SPEED_100G] = 100000, +}; + +void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev) +{ + u32 flags, fcntl, speed, lpa; + + speed = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_SPEED); + EFX_WARN_ON_PARANOID(speed >= ARRAY_SIZE(efx_mcdi_event_link_speed)); + speed = efx_mcdi_event_link_speed[speed]; + + flags = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LINK_FLAGS); + fcntl = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_FCNTL); + lpa = EFX_QWORD_FIELD(*ev, MCDI_EVENT_LINKCHANGE_LP_CAP); + + /* efx->link_state is only modified by efx_mcdi_phy_get_link(), + * which is only run after flushing the event queues. Therefore, it + * is safe to modify the link state outside of the mac_lock here. + */ + efx_mcdi_phy_decode_link(efx, &efx->link_state, speed, flags, fcntl); + + efx_mcdi_phy_check_fcntl(efx, lpa); + + efx_link_status_changed(efx); +} diff --git a/drivers/net/ethernet/sfc/mcdi_port_common.h b/drivers/net/ethernet/sfc/mcdi_port_common.h new file mode 100644 index 000000000000..b16f11265269 --- /dev/null +++ b/drivers/net/ethernet/sfc/mcdi_port_common.h @@ -0,0 +1,57 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ +#ifndef EFX_MCDI_PORT_COMMON_H +#define EFX_MCDI_PORT_COMMON_H + +#include "net_driver.h" +#include "mcdi.h" +#include "mcdi_pcol.h" + +struct efx_mcdi_phy_data { + u32 flags; + u32 type; + u32 supported_cap; + u32 channel; + u32 port; + u32 stats_mask; + u8 name[20]; + u32 media; + u32 mmd_mask; + u8 revision[20]; + u32 forced_cap; +}; + +#define EFX_MC_STATS_GENERATION_INVALID ((__force __le64)(-1)) + +int efx_mcdi_get_phy_cfg(struct efx_nic *efx, struct efx_mcdi_phy_data *cfg); +void efx_link_set_advertising(struct efx_nic *efx, + const unsigned long *advertising); +int efx_mcdi_set_link(struct efx_nic *efx, u32 capabilities, + u32 flags, u32 loopback_mode, u32 loopback_speed); +int efx_mcdi_loopback_modes(struct efx_nic *efx, u64 *loopback_modes); +void mcdi_to_ethtool_linkset(u32 media, u32 cap, unsigned long *linkset); +u32 ethtool_linkset_to_mcdi_cap(const unsigned long *linkset); +u32 efx_get_mcdi_phy_flags(struct efx_nic *efx); +u8 mcdi_to_ethtool_media(u32 media); +void efx_mcdi_phy_decode_link(struct efx_nic *efx, + struct efx_link_state *link_state, + u32 speed, u32 flags, u32 fcntl); +u32 ethtool_fec_caps_to_mcdi(u32 ethtool_cap); +u32 mcdi_fec_caps_to_ethtool(u32 caps, bool is_25g); +void efx_mcdi_phy_check_fcntl(struct efx_nic *efx, u32 lpa); +bool efx_mcdi_phy_poll(struct efx_nic *efx); +int efx_mcdi_phy_get_fecparam(struct efx_nic *efx, + struct ethtool_fecparam *fec); +int efx_mcdi_phy_test_alive(struct efx_nic *efx); +int efx_mcdi_set_mac(struct efx_nic *efx); +int efx_mcdi_port_get_number(struct efx_nic *efx); +void efx_mcdi_process_link_change(struct efx_nic *efx, efx_qword_t *ev); + +#endif diff --git a/drivers/net/ethernet/sfc/net_driver.h b/drivers/net/ethernet/sfc/net_driver.h index dfd5182d9e47..9f9886f222c8 100644 --- a/drivers/net/ethernet/sfc/net_driver.h +++ b/drivers/net/ethernet/sfc/net_driver.h @@ -24,7 +24,6 @@ #include <linux/mutex.h> #include <linux/rwsem.h> #include <linux/vmalloc.h> -#include <linux/i2c.h> #include <linux/mtd/mtd.h> #include <net/busy_poll.h> #include <net/xdp.h> @@ -139,6 +138,8 @@ struct efx_special_buffer { * freed when descriptor completes * @xdpf: When @flags & %EFX_TX_BUF_XDP, the XDP frame information; its @data * member is the associated buffer to drop a page reference on. + * @option: When @flags & %EFX_TX_BUF_OPTION, an EF10-specific option + * descriptor. * @dma_addr: DMA address of the fragment. * @flags: Flags for allocation and DMA mapping type * @len: Length of this fragment. @@ -153,7 +154,7 @@ struct efx_tx_buffer { struct xdp_frame *xdpf; }; union { - efx_qword_t option; + efx_qword_t option; /* EF10 */ dma_addr_t dma_addr; }; unsigned short flags; @@ -743,13 +744,13 @@ union efx_multicast_hash { struct vfdi_status; /* The reserved RSS context value */ -#define EFX_EF10_RSS_CONTEXT_INVALID 0xffffffff +#define EFX_MCDI_RSS_CONTEXT_INVALID 0xffffffff /** * struct efx_rss_context - A user-defined RSS context for filtering * @list: node of linked list on which this struct is stored * @context_id: the RSS_CONTEXT_ID returned by MC firmware, or - * %EFX_EF10_RSS_CONTEXT_INVALID if this context is not present on the NIC. - * For Siena, 0 if RSS is active, else %EFX_EF10_RSS_CONTEXT_INVALID. + * %EFX_MCDI_RSS_CONTEXT_INVALID if this context is not present on the NIC. + * For Siena, 0 if RSS is active, else %EFX_MCDI_RSS_CONTEXT_INVALID. * @user_id: the rss_context ID exposed to userspace over ethtool. * @rx_hash_udp_4tuple: UDP 4-tuple hashing enabled * @rx_hash_key: Toeplitz hash key for this RSS context @@ -1611,6 +1612,15 @@ static inline struct efx_rx_buffer *efx_rx_buffer(struct efx_rx_queue *rx_queue, return &rx_queue->buffer[index]; } +static inline struct efx_rx_buffer * +efx_rx_buf_next(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf) +{ + if (unlikely(rx_buf == efx_rx_buffer(rx_queue, rx_queue->ptr_mask))) + return efx_rx_buffer(rx_queue, 0); + else + return rx_buf + 1; +} + /** * EFX_MAX_FRAME_LEN - calculate maximum frame length * diff --git a/drivers/net/ethernet/sfc/nic.h b/drivers/net/ethernet/sfc/nic.h index 1f7c5717de75..6670fda8f35a 100644 --- a/drivers/net/ethernet/sfc/nic.h +++ b/drivers/net/ethernet/sfc/nic.h @@ -9,9 +9,9 @@ #define EFX_NIC_H #include <linux/net_tstamp.h> -#include <linux/i2c-algo-bit.h> #include "net_driver.h" #include "efx.h" +#include "efx_common.h" #include "mcdi.h" enum { @@ -506,6 +506,9 @@ static inline void efx_nic_push_buffers(struct efx_tx_queue *tx_queue) tx_queue->efx->type->tx_write(tx_queue); } +int efx_enqueue_skb_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + bool *data_mapped); + /* RX data path */ static inline int efx_nic_probe_rx(struct efx_rx_queue *rx_queue) { @@ -554,6 +557,7 @@ static inline void efx_nic_eventq_read_ack(struct efx_channel *channel) { channel->efx->type->ev_read_ack(channel); } + void efx_nic_event_test_start(struct efx_channel *channel); /* Falcon/Siena queue operations */ @@ -671,6 +675,7 @@ struct efx_farch_register_test { unsigned address; efx_oword_t mask; }; + int efx_farch_test_registers(struct efx_nic *efx, const struct efx_farch_register_test *regs, size_t n_regs); diff --git a/drivers/net/ethernet/sfc/rx.c b/drivers/net/ethernet/sfc/rx.c index c29bf862a94c..a2042f16babc 100644 --- a/drivers/net/ethernet/sfc/rx.c +++ b/drivers/net/ethernet/sfc/rx.c @@ -21,6 +21,7 @@ #include <linux/bpf_trace.h> #include "net_driver.h" #include "efx.h" +#include "rx_common.h" #include "filter.h" #include "nic.h" #include "selftest.h" @@ -32,60 +33,13 @@ /* Maximum rx prefix used by any architecture. */ #define EFX_MAX_RX_PREFIX_SIZE 16 -/* Number of RX buffers to recycle pages for. When creating the RX page recycle - * ring, this number is divided by the number of buffers per page to calculate - * the number of pages to store in the RX page recycle ring. - */ -#define EFX_RECYCLE_RING_SIZE_IOMMU 4096 -#define EFX_RECYCLE_RING_SIZE_NOIOMMU (2 * EFX_RX_PREFERRED_BATCH) - /* Size of buffer allocated for skb header area. */ #define EFX_SKB_HEADERS 128u -/* This is the percentage fill level below which new RX descriptors - * will be added to the RX descriptor ring. - */ -static unsigned int rx_refill_threshold; - /* Each packet can consume up to ceil(max_frame_len / buffer_size) buffers */ #define EFX_RX_MAX_FRAGS DIV_ROUND_UP(EFX_MAX_FRAME_LEN(EFX_MAX_MTU), \ EFX_RX_USR_BUF_SIZE) -/* - * RX maximum head room required. - * - * This must be at least 1 to prevent overflow, plus one packet-worth - * to allow pipelined receives. - */ -#define EFX_RXD_HEAD_ROOM (1 + EFX_RX_MAX_FRAGS) - -static inline u8 *efx_rx_buf_va(struct efx_rx_buffer *buf) -{ - return page_address(buf->page) + buf->page_offset; -} - -static inline u32 efx_rx_buf_hash(struct efx_nic *efx, const u8 *eh) -{ -#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) - return __le32_to_cpup((const __le32 *)(eh + efx->rx_packet_hash_offset)); -#else - const u8 *data = eh + efx->rx_packet_hash_offset; - return (u32)data[0] | - (u32)data[1] << 8 | - (u32)data[2] << 16 | - (u32)data[3] << 24; -#endif -} - -static inline struct efx_rx_buffer * -efx_rx_buf_next(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf) -{ - if (unlikely(rx_buf == efx_rx_buffer(rx_queue, rx_queue->ptr_mask))) - return efx_rx_buffer(rx_queue, 0); - else - return rx_buf + 1; -} - static inline void efx_sync_rx_buffer(struct efx_nic *efx, struct efx_rx_buffer *rx_buf, unsigned int len) @@ -94,301 +48,6 @@ static inline void efx_sync_rx_buffer(struct efx_nic *efx, DMA_FROM_DEVICE); } -void efx_rx_config_page_split(struct efx_nic *efx) -{ - efx->rx_page_buf_step = ALIGN(efx->rx_dma_len + efx->rx_ip_align + - XDP_PACKET_HEADROOM, - EFX_RX_BUF_ALIGNMENT); - efx->rx_bufs_per_page = efx->rx_buffer_order ? 1 : - ((PAGE_SIZE - sizeof(struct efx_rx_page_state)) / - efx->rx_page_buf_step); - efx->rx_buffer_truesize = (PAGE_SIZE << efx->rx_buffer_order) / - efx->rx_bufs_per_page; - efx->rx_pages_per_batch = DIV_ROUND_UP(EFX_RX_PREFERRED_BATCH, - efx->rx_bufs_per_page); -} - -/* Check the RX page recycle ring for a page that can be reused. */ -static struct page *efx_reuse_page(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - struct page *page; - struct efx_rx_page_state *state; - unsigned index; - - index = rx_queue->page_remove & rx_queue->page_ptr_mask; - page = rx_queue->page_ring[index]; - if (page == NULL) - return NULL; - - rx_queue->page_ring[index] = NULL; - /* page_remove cannot exceed page_add. */ - if (rx_queue->page_remove != rx_queue->page_add) - ++rx_queue->page_remove; - - /* If page_count is 1 then we hold the only reference to this page. */ - if (page_count(page) == 1) { - ++rx_queue->page_recycle_count; - return page; - } else { - state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - put_page(page); - ++rx_queue->page_recycle_failed; - } - - return NULL; -} - -/** - * efx_init_rx_buffers - create EFX_RX_BATCH page-based RX buffers - * - * @rx_queue: Efx RX queue - * - * This allocates a batch of pages, maps them for DMA, and populates - * struct efx_rx_buffers for each one. Return a negative error code or - * 0 on success. If a single page can be used for multiple buffers, - * then the page will either be inserted fully, or not at all. - */ -static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic) -{ - struct efx_nic *efx = rx_queue->efx; - struct efx_rx_buffer *rx_buf; - struct page *page; - unsigned int page_offset; - struct efx_rx_page_state *state; - dma_addr_t dma_addr; - unsigned index, count; - - count = 0; - do { - page = efx_reuse_page(rx_queue); - if (page == NULL) { - page = alloc_pages(__GFP_COMP | - (atomic ? GFP_ATOMIC : GFP_KERNEL), - efx->rx_buffer_order); - if (unlikely(page == NULL)) - return -ENOMEM; - dma_addr = - dma_map_page(&efx->pci_dev->dev, page, 0, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - if (unlikely(dma_mapping_error(&efx->pci_dev->dev, - dma_addr))) { - __free_pages(page, efx->rx_buffer_order); - return -EIO; - } - state = page_address(page); - state->dma_addr = dma_addr; - } else { - state = page_address(page); - dma_addr = state->dma_addr; - } - - dma_addr += sizeof(struct efx_rx_page_state); - page_offset = sizeof(struct efx_rx_page_state); - - do { - index = rx_queue->added_count & rx_queue->ptr_mask; - rx_buf = efx_rx_buffer(rx_queue, index); - rx_buf->dma_addr = dma_addr + efx->rx_ip_align + - XDP_PACKET_HEADROOM; - rx_buf->page = page; - rx_buf->page_offset = page_offset + efx->rx_ip_align + - XDP_PACKET_HEADROOM; - rx_buf->len = efx->rx_dma_len; - rx_buf->flags = 0; - ++rx_queue->added_count; - get_page(page); - dma_addr += efx->rx_page_buf_step; - page_offset += efx->rx_page_buf_step; - } while (page_offset + efx->rx_page_buf_step <= PAGE_SIZE); - - rx_buf->flags = EFX_RX_BUF_LAST_IN_PAGE; - } while (++count < efx->rx_pages_per_batch); - - return 0; -} - -/* Unmap a DMA-mapped page. This function is only called for the final RX - * buffer in a page. - */ -static void efx_unmap_rx_buffer(struct efx_nic *efx, - struct efx_rx_buffer *rx_buf) -{ - struct page *page = rx_buf->page; - - if (page) { - struct efx_rx_page_state *state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, - state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - } -} - -static void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, - struct efx_rx_buffer *rx_buf, - unsigned int num_bufs) -{ - do { - if (rx_buf->page) { - put_page(rx_buf->page); - rx_buf->page = NULL; - } - rx_buf = efx_rx_buf_next(rx_queue, rx_buf); - } while (--num_bufs); -} - -/* Attempt to recycle the page if there is an RX recycle ring; the page can - * only be added if this is the final RX buffer, to prevent pages being used in - * the descriptor ring and appearing in the recycle ring simultaneously. - */ -static void efx_recycle_rx_page(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf) -{ - struct page *page = rx_buf->page; - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - struct efx_nic *efx = rx_queue->efx; - unsigned index; - - /* Only recycle the page after processing the final buffer. */ - if (!(rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE)) - return; - - index = rx_queue->page_add & rx_queue->page_ptr_mask; - if (rx_queue->page_ring[index] == NULL) { - unsigned read_index = rx_queue->page_remove & - rx_queue->page_ptr_mask; - - /* The next slot in the recycle ring is available, but - * increment page_remove if the read pointer currently - * points here. - */ - if (read_index == index) - ++rx_queue->page_remove; - rx_queue->page_ring[index] = page; - ++rx_queue->page_add; - return; - } - ++rx_queue->page_recycle_full; - efx_unmap_rx_buffer(efx, rx_buf); - put_page(rx_buf->page); -} - -static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue, - struct efx_rx_buffer *rx_buf) -{ - /* Release the page reference we hold for the buffer. */ - if (rx_buf->page) - put_page(rx_buf->page); - - /* If this is the last buffer in a page, unmap and free it. */ - if (rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE) { - efx_unmap_rx_buffer(rx_queue->efx, rx_buf); - efx_free_rx_buffers(rx_queue, rx_buf, 1); - } - rx_buf->page = NULL; -} - -/* Recycle the pages that are used by buffers that have just been received. */ -static void efx_recycle_rx_pages(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf, - unsigned int n_frags) -{ - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - - do { - efx_recycle_rx_page(channel, rx_buf); - rx_buf = efx_rx_buf_next(rx_queue, rx_buf); - } while (--n_frags); -} - -static void efx_discard_rx_packet(struct efx_channel *channel, - struct efx_rx_buffer *rx_buf, - unsigned int n_frags) -{ - struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); - - efx_recycle_rx_pages(channel, rx_buf, n_frags); - - efx_free_rx_buffers(rx_queue, rx_buf, n_frags); -} - -/** - * efx_fast_push_rx_descriptors - push new RX descriptors quickly - * @rx_queue: RX descriptor queue - * - * This will aim to fill the RX descriptor queue up to - * @rx_queue->@max_fill. If there is insufficient atomic - * memory to do so, a slow fill will be scheduled. - * - * The caller must provide serialisation (none is used here). In practise, - * this means this function must run from the NAPI handler, or be called - * when NAPI is disabled. - */ -void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int fill_level, batch_size; - int space, rc = 0; - - if (!rx_queue->refill_enabled) - return; - - /* Calculate current fill level, and exit if we don't need to fill */ - fill_level = (rx_queue->added_count - rx_queue->removed_count); - EFX_WARN_ON_ONCE_PARANOID(fill_level > rx_queue->efx->rxq_entries); - if (fill_level >= rx_queue->fast_fill_trigger) - goto out; - - /* Record minimum fill level */ - if (unlikely(fill_level < rx_queue->min_fill)) { - if (fill_level) - rx_queue->min_fill = fill_level; - } - - batch_size = efx->rx_pages_per_batch * efx->rx_bufs_per_page; - space = rx_queue->max_fill - fill_level; - EFX_WARN_ON_ONCE_PARANOID(space < batch_size); - - netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, - "RX queue %d fast-filling descriptor ring from" - " level %d to level %d\n", - efx_rx_queue_index(rx_queue), fill_level, - rx_queue->max_fill); - - - do { - rc = efx_init_rx_buffers(rx_queue, atomic); - if (unlikely(rc)) { - /* Ensure that we don't leave the rx queue empty */ - efx_schedule_slow_fill(rx_queue); - goto out; - } - } while ((space -= batch_size) >= batch_size); - - netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, - "RX queue %d fast-filled descriptor ring " - "to level %d\n", efx_rx_queue_index(rx_queue), - rx_queue->added_count - rx_queue->removed_count); - - out: - if (rx_queue->notified_count != rx_queue->added_count) - efx_nic_notify_rx_desc(rx_queue); -} - -void efx_rx_slow_fill(struct timer_list *t) -{ - struct efx_rx_queue *rx_queue = from_timer(rx_queue, t, slow_fill); - - /* Post an event to cause NAPI to run and refill the queue */ - efx_nic_generate_fill_event(rx_queue); - ++rx_queue->slow_fill_count; -} - static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue, struct efx_rx_buffer *rx_buf, int len) @@ -412,53 +71,6 @@ static void efx_rx_packet__check_len(struct efx_rx_queue *rx_queue, efx_rx_queue_channel(rx_queue)->n_rx_overlength++; } -/* Pass a received packet up through GRO. GRO can handle pages - * regardless of checksum state and skbs with a good checksum. - */ -static void -efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, - unsigned int n_frags, u8 *eh) -{ - struct napi_struct *napi = &channel->napi_str; - struct efx_nic *efx = channel->efx; - struct sk_buff *skb; - - skb = napi_get_frags(napi); - if (unlikely(!skb)) { - struct efx_rx_queue *rx_queue; - - rx_queue = efx_channel_get_rx_queue(channel); - efx_free_rx_buffers(rx_queue, rx_buf, n_frags); - return; - } - - if (efx->net_dev->features & NETIF_F_RXHASH) - skb_set_hash(skb, efx_rx_buf_hash(efx, eh), - PKT_HASH_TYPE_L3); - skb->ip_summed = ((rx_buf->flags & EFX_RX_PKT_CSUMMED) ? - CHECKSUM_UNNECESSARY : CHECKSUM_NONE); - skb->csum_level = !!(rx_buf->flags & EFX_RX_PKT_CSUM_LEVEL); - - for (;;) { - skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, - rx_buf->page, rx_buf->page_offset, - rx_buf->len); - rx_buf->page = NULL; - skb->len += rx_buf->len; - if (skb_shinfo(skb)->nr_frags == n_frags) - break; - - rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf); - } - - skb->data_len = skb->len; - skb->truesize += n_frags * efx->rx_buffer_truesize; - - skb_record_rx_queue(skb, channel->rx_queue.core_index); - - napi_gro_frags(napi); -} - /* Allocate and construct an SKB around page fragments */ static struct sk_buff *efx_rx_mk_skb(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, @@ -805,174 +417,6 @@ out: channel->rx_pkt_n_frags = 0; } -int efx_probe_rx_queue(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int entries; - int rc; - - /* Create the smallest power-of-two aligned ring */ - entries = max(roundup_pow_of_two(efx->rxq_entries), EFX_MIN_DMAQ_SIZE); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); - rx_queue->ptr_mask = entries - 1; - - netif_dbg(efx, probe, efx->net_dev, - "creating RX queue %d size %#x mask %#x\n", - efx_rx_queue_index(rx_queue), efx->rxq_entries, - rx_queue->ptr_mask); - - /* Allocate RX buffers */ - rx_queue->buffer = kcalloc(entries, sizeof(*rx_queue->buffer), - GFP_KERNEL); - if (!rx_queue->buffer) - return -ENOMEM; - - rc = efx_nic_probe_rx(rx_queue); - if (rc) { - kfree(rx_queue->buffer); - rx_queue->buffer = NULL; - } - - return rc; -} - -static void efx_init_rx_recycle_ring(struct efx_nic *efx, - struct efx_rx_queue *rx_queue) -{ - unsigned int bufs_in_recycle_ring, page_ring_size; - - /* Set the RX recycle ring size */ -#ifdef CONFIG_PPC64 - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; -#else - if (iommu_present(&pci_bus_type)) - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; - else - bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_NOIOMMU; -#endif /* CONFIG_PPC64 */ - - page_ring_size = roundup_pow_of_two(bufs_in_recycle_ring / - efx->rx_bufs_per_page); - rx_queue->page_ring = kcalloc(page_ring_size, - sizeof(*rx_queue->page_ring), GFP_KERNEL); - rx_queue->page_ptr_mask = page_ring_size - 1; -} - -void efx_init_rx_queue(struct efx_rx_queue *rx_queue) -{ - struct efx_nic *efx = rx_queue->efx; - unsigned int max_fill, trigger, max_trigger; - int rc = 0; - - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "initialising RX queue %d\n", efx_rx_queue_index(rx_queue)); - - /* Initialise ptr fields */ - rx_queue->added_count = 0; - rx_queue->notified_count = 0; - rx_queue->removed_count = 0; - rx_queue->min_fill = -1U; - efx_init_rx_recycle_ring(efx, rx_queue); - - rx_queue->page_remove = 0; - rx_queue->page_add = rx_queue->page_ptr_mask + 1; - rx_queue->page_recycle_count = 0; - rx_queue->page_recycle_failed = 0; - rx_queue->page_recycle_full = 0; - - /* Initialise limit fields */ - max_fill = efx->rxq_entries - EFX_RXD_HEAD_ROOM; - max_trigger = - max_fill - efx->rx_pages_per_batch * efx->rx_bufs_per_page; - if (rx_refill_threshold != 0) { - trigger = max_fill * min(rx_refill_threshold, 100U) / 100U; - if (trigger > max_trigger) - trigger = max_trigger; - } else { - trigger = max_trigger; - } - - rx_queue->max_fill = max_fill; - rx_queue->fast_fill_trigger = trigger; - rx_queue->refill_enabled = true; - - /* Initialise XDP queue information */ - rc = xdp_rxq_info_reg(&rx_queue->xdp_rxq_info, efx->net_dev, - rx_queue->core_index); - - if (rc) { - netif_err(efx, rx_err, efx->net_dev, - "Failure to initialise XDP queue information rc=%d\n", - rc); - efx->xdp_rxq_info_failed = true; - } else { - rx_queue->xdp_rxq_info_valid = true; - } - - /* Set up RX descriptor ring */ - efx_nic_init_rx(rx_queue); -} - -void efx_fini_rx_queue(struct efx_rx_queue *rx_queue) -{ - int i; - struct efx_nic *efx = rx_queue->efx; - struct efx_rx_buffer *rx_buf; - - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "shutting down RX queue %d\n", efx_rx_queue_index(rx_queue)); - - del_timer_sync(&rx_queue->slow_fill); - - /* Release RX buffers from the current read ptr to the write ptr */ - if (rx_queue->buffer) { - for (i = rx_queue->removed_count; i < rx_queue->added_count; - i++) { - unsigned index = i & rx_queue->ptr_mask; - rx_buf = efx_rx_buffer(rx_queue, index); - efx_fini_rx_buffer(rx_queue, rx_buf); - } - } - - /* Unmap and release the pages in the recycle ring. Remove the ring. */ - for (i = 0; i <= rx_queue->page_ptr_mask; i++) { - struct page *page = rx_queue->page_ring[i]; - struct efx_rx_page_state *state; - - if (page == NULL) - continue; - - state = page_address(page); - dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, - PAGE_SIZE << efx->rx_buffer_order, - DMA_FROM_DEVICE); - put_page(page); - } - kfree(rx_queue->page_ring); - rx_queue->page_ring = NULL; - - if (rx_queue->xdp_rxq_info_valid) - xdp_rxq_info_unreg(&rx_queue->xdp_rxq_info); - - rx_queue->xdp_rxq_info_valid = false; -} - -void efx_remove_rx_queue(struct efx_rx_queue *rx_queue) -{ - netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, - "destroying RX queue %d\n", efx_rx_queue_index(rx_queue)); - - efx_nic_remove_rx(rx_queue); - - kfree(rx_queue->buffer); - rx_queue->buffer = NULL; -} - - -module_param(rx_refill_threshold, uint, 0444); -MODULE_PARM_DESC(rx_refill_threshold, - "RX descriptor ring refill threshold (%)"); - #ifdef CONFIG_RFS_ACCEL static void efx_filter_rfs_work(struct work_struct *data) @@ -1206,37 +650,3 @@ bool __efx_filter_rfs_expire(struct efx_channel *channel, unsigned int quota) } #endif /* CONFIG_RFS_ACCEL */ - -/** - * efx_filter_is_mc_recipient - test whether spec is a multicast recipient - * @spec: Specification to test - * - * Return: %true if the specification is a non-drop RX filter that - * matches a local MAC address I/G bit value of 1 or matches a local - * IPv4 or IPv6 address value in the respective multicast address - * range. Otherwise %false. - */ -bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec) -{ - if (!(spec->flags & EFX_FILTER_FLAG_RX) || - spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP) - return false; - - if (spec->match_flags & - (EFX_FILTER_MATCH_LOC_MAC | EFX_FILTER_MATCH_LOC_MAC_IG) && - is_multicast_ether_addr(spec->loc_mac)) - return true; - - if ((spec->match_flags & - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == - (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { - if (spec->ether_type == htons(ETH_P_IP) && - ipv4_is_multicast(spec->loc_host[0])) - return true; - if (spec->ether_type == htons(ETH_P_IPV6) && - ((const u8 *)spec->loc_host)[0] == 0xff) - return true; - } - - return false; -} diff --git a/drivers/net/ethernet/sfc/rx_common.c b/drivers/net/ethernet/sfc/rx_common.c new file mode 100644 index 000000000000..ee8beb87bdc1 --- /dev/null +++ b/drivers/net/ethernet/sfc/rx_common.c @@ -0,0 +1,851 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include <linux/module.h> +#include <linux/iommu.h> +#include "efx.h" +#include "nic.h" +#include "rx_common.h" + +/* This is the percentage fill level below which new RX descriptors + * will be added to the RX descriptor ring. + */ +static unsigned int rx_refill_threshold; +module_param(rx_refill_threshold, uint, 0444); +MODULE_PARM_DESC(rx_refill_threshold, + "RX descriptor ring refill threshold (%)"); + +/* Number of RX buffers to recycle pages for. When creating the RX page recycle + * ring, this number is divided by the number of buffers per page to calculate + * the number of pages to store in the RX page recycle ring. + */ +#define EFX_RECYCLE_RING_SIZE_IOMMU 4096 +#define EFX_RECYCLE_RING_SIZE_NOIOMMU (2 * EFX_RX_PREFERRED_BATCH) + +/* RX maximum head room required. + * + * This must be at least 1 to prevent overflow, plus one packet-worth + * to allow pipelined receives. + */ +#define EFX_RXD_HEAD_ROOM (1 + EFX_RX_MAX_FRAGS) + +/* Check the RX page recycle ring for a page that can be reused. */ +static struct page *efx_reuse_page(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + struct efx_rx_page_state *state; + unsigned int index; + struct page *page; + + index = rx_queue->page_remove & rx_queue->page_ptr_mask; + page = rx_queue->page_ring[index]; + if (page == NULL) + return NULL; + + rx_queue->page_ring[index] = NULL; + /* page_remove cannot exceed page_add. */ + if (rx_queue->page_remove != rx_queue->page_add) + ++rx_queue->page_remove; + + /* If page_count is 1 then we hold the only reference to this page. */ + if (page_count(page) == 1) { + ++rx_queue->page_recycle_count; + return page; + } else { + state = page_address(page); + dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + put_page(page); + ++rx_queue->page_recycle_failed; + } + + return NULL; +} + +/* Attempt to recycle the page if there is an RX recycle ring; the page can + * only be added if this is the final RX buffer, to prevent pages being used in + * the descriptor ring and appearing in the recycle ring simultaneously. + */ +static void efx_recycle_rx_page(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + struct efx_nic *efx = rx_queue->efx; + struct page *page = rx_buf->page; + unsigned int index; + + /* Only recycle the page after processing the final buffer. */ + if (!(rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE)) + return; + + index = rx_queue->page_add & rx_queue->page_ptr_mask; + if (rx_queue->page_ring[index] == NULL) { + unsigned int read_index = rx_queue->page_remove & + rx_queue->page_ptr_mask; + + /* The next slot in the recycle ring is available, but + * increment page_remove if the read pointer currently + * points here. + */ + if (read_index == index) + ++rx_queue->page_remove; + rx_queue->page_ring[index] = page; + ++rx_queue->page_add; + return; + } + ++rx_queue->page_recycle_full; + efx_unmap_rx_buffer(efx, rx_buf); + put_page(rx_buf->page); +} + +/* Recycle the pages that are used by buffers that have just been received. */ +void efx_recycle_rx_pages(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + + do { + efx_recycle_rx_page(channel, rx_buf); + rx_buf = efx_rx_buf_next(rx_queue, rx_buf); + } while (--n_frags); +} + +void efx_discard_rx_packet(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags) +{ + struct efx_rx_queue *rx_queue = efx_channel_get_rx_queue(channel); + + efx_recycle_rx_pages(channel, rx_buf, n_frags); + + efx_free_rx_buffers(rx_queue, rx_buf, n_frags); +} + +static void efx_init_rx_recycle_ring(struct efx_rx_queue *rx_queue) +{ + unsigned int bufs_in_recycle_ring, page_ring_size; + struct efx_nic *efx = rx_queue->efx; + + /* Set the RX recycle ring size */ +#ifdef CONFIG_PPC64 + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; +#else + if (iommu_present(&pci_bus_type)) + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_IOMMU; + else + bufs_in_recycle_ring = EFX_RECYCLE_RING_SIZE_NOIOMMU; +#endif /* CONFIG_PPC64 */ + + page_ring_size = roundup_pow_of_two(bufs_in_recycle_ring / + efx->rx_bufs_per_page); + rx_queue->page_ring = kcalloc(page_ring_size, + sizeof(*rx_queue->page_ring), GFP_KERNEL); + rx_queue->page_ptr_mask = page_ring_size - 1; +} + +static void efx_fini_rx_recycle_ring(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + int i; + + /* Unmap and release the pages in the recycle ring. Remove the ring. */ + for (i = 0; i <= rx_queue->page_ptr_mask; i++) { + struct page *page = rx_queue->page_ring[i]; + struct efx_rx_page_state *state; + + if (page == NULL) + continue; + + state = page_address(page); + dma_unmap_page(&efx->pci_dev->dev, state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + put_page(page); + } + kfree(rx_queue->page_ring); + rx_queue->page_ring = NULL; +} + +static void efx_fini_rx_buffer(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf) +{ + /* Release the page reference we hold for the buffer. */ + if (rx_buf->page) + put_page(rx_buf->page); + + /* If this is the last buffer in a page, unmap and free it. */ + if (rx_buf->flags & EFX_RX_BUF_LAST_IN_PAGE) { + efx_unmap_rx_buffer(rx_queue->efx, rx_buf); + efx_free_rx_buffers(rx_queue, rx_buf, 1); + } + rx_buf->page = NULL; +} + +int efx_probe_rx_queue(struct efx_rx_queue *rx_queue) +{ + struct efx_nic *efx = rx_queue->efx; + unsigned int entries; + int rc; + + /* Create the smallest power-of-two aligned ring */ + entries = max(roundup_pow_of_two(efx->rxq_entries), EFX_MIN_DMAQ_SIZE); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); + rx_queue->ptr_mask = entries - 1; + + netif_dbg(efx, probe, efx->net_dev, + "creating RX queue %d size %#x mask %#x\n", + efx_rx_queue_index(rx_queue), efx->rxq_entries, + rx_queue->ptr_mask); + + /* Allocate RX buffers */ + rx_queue->buffer = kcalloc(entries, sizeof(*rx_queue->buffer), + GFP_KERNEL); + if (!rx_queue->buffer) + return -ENOMEM; + + rc = efx_nic_probe_rx(rx_queue); + if (rc) { + kfree(rx_queue->buffer); + rx_queue->buffer = NULL; + } + + return rc; +} + +void efx_init_rx_queue(struct efx_rx_queue *rx_queue) +{ + unsigned int max_fill, trigger, max_trigger; + struct efx_nic *efx = rx_queue->efx; + int rc = 0; + + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "initialising RX queue %d\n", efx_rx_queue_index(rx_queue)); + + /* Initialise ptr fields */ + rx_queue->added_count = 0; + rx_queue->notified_count = 0; + rx_queue->removed_count = 0; + rx_queue->min_fill = -1U; + efx_init_rx_recycle_ring(rx_queue); + + rx_queue->page_remove = 0; + rx_queue->page_add = rx_queue->page_ptr_mask + 1; + rx_queue->page_recycle_count = 0; + rx_queue->page_recycle_failed = 0; + rx_queue->page_recycle_full = 0; + + /* Initialise limit fields */ + max_fill = efx->rxq_entries - EFX_RXD_HEAD_ROOM; + max_trigger = + max_fill - efx->rx_pages_per_batch * efx->rx_bufs_per_page; + if (rx_refill_threshold != 0) { + trigger = max_fill * min(rx_refill_threshold, 100U) / 100U; + if (trigger > max_trigger) + trigger = max_trigger; + } else { + trigger = max_trigger; + } + + rx_queue->max_fill = max_fill; + rx_queue->fast_fill_trigger = trigger; + rx_queue->refill_enabled = true; + + /* Initialise XDP queue information */ + rc = xdp_rxq_info_reg(&rx_queue->xdp_rxq_info, efx->net_dev, + rx_queue->core_index); + + if (rc) { + netif_err(efx, rx_err, efx->net_dev, + "Failure to initialise XDP queue information rc=%d\n", + rc); + efx->xdp_rxq_info_failed = true; + } else { + rx_queue->xdp_rxq_info_valid = true; + } + + /* Set up RX descriptor ring */ + efx_nic_init_rx(rx_queue); +} + +void efx_fini_rx_queue(struct efx_rx_queue *rx_queue) +{ + struct efx_rx_buffer *rx_buf; + int i; + + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "shutting down RX queue %d\n", efx_rx_queue_index(rx_queue)); + + del_timer_sync(&rx_queue->slow_fill); + + /* Release RX buffers from the current read ptr to the write ptr */ + if (rx_queue->buffer) { + for (i = rx_queue->removed_count; i < rx_queue->added_count; + i++) { + unsigned int index = i & rx_queue->ptr_mask; + + rx_buf = efx_rx_buffer(rx_queue, index); + efx_fini_rx_buffer(rx_queue, rx_buf); + } + } + + efx_fini_rx_recycle_ring(rx_queue); + + if (rx_queue->xdp_rxq_info_valid) + xdp_rxq_info_unreg(&rx_queue->xdp_rxq_info); + + rx_queue->xdp_rxq_info_valid = false; +} + +void efx_remove_rx_queue(struct efx_rx_queue *rx_queue) +{ + netif_dbg(rx_queue->efx, drv, rx_queue->efx->net_dev, + "destroying RX queue %d\n", efx_rx_queue_index(rx_queue)); + + efx_nic_remove_rx(rx_queue); + + kfree(rx_queue->buffer); + rx_queue->buffer = NULL; +} + +/* Unmap a DMA-mapped page. This function is only called for the final RX + * buffer in a page. + */ +void efx_unmap_rx_buffer(struct efx_nic *efx, + struct efx_rx_buffer *rx_buf) +{ + struct page *page = rx_buf->page; + + if (page) { + struct efx_rx_page_state *state = page_address(page); + + dma_unmap_page(&efx->pci_dev->dev, + state->dma_addr, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + } +} + +void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf, + unsigned int num_bufs) +{ + do { + if (rx_buf->page) { + put_page(rx_buf->page); + rx_buf->page = NULL; + } + rx_buf = efx_rx_buf_next(rx_queue, rx_buf); + } while (--num_bufs); +} + +void efx_rx_slow_fill(struct timer_list *t) +{ + struct efx_rx_queue *rx_queue = from_timer(rx_queue, t, slow_fill); + + /* Post an event to cause NAPI to run and refill the queue */ + efx_nic_generate_fill_event(rx_queue); + ++rx_queue->slow_fill_count; +} + +void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue) +{ + mod_timer(&rx_queue->slow_fill, jiffies + msecs_to_jiffies(10)); +} + +/* efx_init_rx_buffers - create EFX_RX_BATCH page-based RX buffers + * + * @rx_queue: Efx RX queue + * + * This allocates a batch of pages, maps them for DMA, and populates + * struct efx_rx_buffers for each one. Return a negative error code or + * 0 on success. If a single page can be used for multiple buffers, + * then the page will either be inserted fully, or not at all. + */ +static int efx_init_rx_buffers(struct efx_rx_queue *rx_queue, bool atomic) +{ + unsigned int page_offset, index, count; + struct efx_nic *efx = rx_queue->efx; + struct efx_rx_page_state *state; + struct efx_rx_buffer *rx_buf; + dma_addr_t dma_addr; + struct page *page; + + count = 0; + do { + page = efx_reuse_page(rx_queue); + if (page == NULL) { + page = alloc_pages(__GFP_COMP | + (atomic ? GFP_ATOMIC : GFP_KERNEL), + efx->rx_buffer_order); + if (unlikely(page == NULL)) + return -ENOMEM; + dma_addr = + dma_map_page(&efx->pci_dev->dev, page, 0, + PAGE_SIZE << efx->rx_buffer_order, + DMA_FROM_DEVICE); + if (unlikely(dma_mapping_error(&efx->pci_dev->dev, + dma_addr))) { + __free_pages(page, efx->rx_buffer_order); + return -EIO; + } + state = page_address(page); + state->dma_addr = dma_addr; + } else { + state = page_address(page); + dma_addr = state->dma_addr; + } + + dma_addr += sizeof(struct efx_rx_page_state); + page_offset = sizeof(struct efx_rx_page_state); + + do { + index = rx_queue->added_count & rx_queue->ptr_mask; + rx_buf = efx_rx_buffer(rx_queue, index); + rx_buf->dma_addr = dma_addr + efx->rx_ip_align + + XDP_PACKET_HEADROOM; + rx_buf->page = page; + rx_buf->page_offset = page_offset + efx->rx_ip_align + + XDP_PACKET_HEADROOM; + rx_buf->len = efx->rx_dma_len; + rx_buf->flags = 0; + ++rx_queue->added_count; + get_page(page); + dma_addr += efx->rx_page_buf_step; + page_offset += efx->rx_page_buf_step; + } while (page_offset + efx->rx_page_buf_step <= PAGE_SIZE); + + rx_buf->flags = EFX_RX_BUF_LAST_IN_PAGE; + } while (++count < efx->rx_pages_per_batch); + + return 0; +} + +void efx_rx_config_page_split(struct efx_nic *efx) +{ + efx->rx_page_buf_step = ALIGN(efx->rx_dma_len + efx->rx_ip_align + + XDP_PACKET_HEADROOM, + EFX_RX_BUF_ALIGNMENT); + efx->rx_bufs_per_page = efx->rx_buffer_order ? 1 : + ((PAGE_SIZE - sizeof(struct efx_rx_page_state)) / + efx->rx_page_buf_step); + efx->rx_buffer_truesize = (PAGE_SIZE << efx->rx_buffer_order) / + efx->rx_bufs_per_page; + efx->rx_pages_per_batch = DIV_ROUND_UP(EFX_RX_PREFERRED_BATCH, + efx->rx_bufs_per_page); +} + +/* efx_fast_push_rx_descriptors - push new RX descriptors quickly + * @rx_queue: RX descriptor queue + * + * This will aim to fill the RX descriptor queue up to + * @rx_queue->@max_fill. If there is insufficient atomic + * memory to do so, a slow fill will be scheduled. + * + * The caller must provide serialisation (none is used here). In practise, + * this means this function must run from the NAPI handler, or be called + * when NAPI is disabled. + */ +void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic) +{ + struct efx_nic *efx = rx_queue->efx; + unsigned int fill_level, batch_size; + int space, rc = 0; + + if (!rx_queue->refill_enabled) + return; + + /* Calculate current fill level, and exit if we don't need to fill */ + fill_level = (rx_queue->added_count - rx_queue->removed_count); + EFX_WARN_ON_ONCE_PARANOID(fill_level > rx_queue->efx->rxq_entries); + if (fill_level >= rx_queue->fast_fill_trigger) + goto out; + + /* Record minimum fill level */ + if (unlikely(fill_level < rx_queue->min_fill)) { + if (fill_level) + rx_queue->min_fill = fill_level; + } + + batch_size = efx->rx_pages_per_batch * efx->rx_bufs_per_page; + space = rx_queue->max_fill - fill_level; + EFX_WARN_ON_ONCE_PARANOID(space < batch_size); + + netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, + "RX queue %d fast-filling descriptor ring from" + " level %d to level %d\n", + efx_rx_queue_index(rx_queue), fill_level, + rx_queue->max_fill); + + do { + rc = efx_init_rx_buffers(rx_queue, atomic); + if (unlikely(rc)) { + /* Ensure that we don't leave the rx queue empty */ + efx_schedule_slow_fill(rx_queue); + goto out; + } + } while ((space -= batch_size) >= batch_size); + + netif_vdbg(rx_queue->efx, rx_status, rx_queue->efx->net_dev, + "RX queue %d fast-filled descriptor ring " + "to level %d\n", efx_rx_queue_index(rx_queue), + rx_queue->added_count - rx_queue->removed_count); + + out: + if (rx_queue->notified_count != rx_queue->added_count) + efx_nic_notify_rx_desc(rx_queue); +} + +/* Pass a received packet up through GRO. GRO can handle pages + * regardless of checksum state and skbs with a good checksum. + */ +void +efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, + unsigned int n_frags, u8 *eh) +{ + struct napi_struct *napi = &channel->napi_str; + struct efx_nic *efx = channel->efx; + struct sk_buff *skb; + + skb = napi_get_frags(napi); + if (unlikely(!skb)) { + struct efx_rx_queue *rx_queue; + + rx_queue = efx_channel_get_rx_queue(channel); + efx_free_rx_buffers(rx_queue, rx_buf, n_frags); + return; + } + + if (efx->net_dev->features & NETIF_F_RXHASH) + skb_set_hash(skb, efx_rx_buf_hash(efx, eh), + PKT_HASH_TYPE_L3); + skb->ip_summed = ((rx_buf->flags & EFX_RX_PKT_CSUMMED) ? + CHECKSUM_UNNECESSARY : CHECKSUM_NONE); + skb->csum_level = !!(rx_buf->flags & EFX_RX_PKT_CSUM_LEVEL); + + for (;;) { + skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, + rx_buf->page, rx_buf->page_offset, + rx_buf->len); + rx_buf->page = NULL; + skb->len += rx_buf->len; + if (skb_shinfo(skb)->nr_frags == n_frags) + break; + + rx_buf = efx_rx_buf_next(&channel->rx_queue, rx_buf); + } + + skb->data_len = skb->len; + skb->truesize += n_frags * efx->rx_buffer_truesize; + + skb_record_rx_queue(skb, channel->rx_queue.core_index); + + napi_gro_frags(napi); +} + +/* RSS contexts. We're using linked lists and crappy O(n) algorithms, because + * (a) this is an infrequent control-plane operation and (b) n is small (max 64) + */ +struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx) +{ + struct list_head *head = &efx->rss_context.list; + struct efx_rss_context *ctx, *new; + u32 id = 1; /* Don't use zero, that refers to the master RSS context */ + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + /* Search for first gap in the numbering */ + list_for_each_entry(ctx, head, list) { + if (ctx->user_id != id) + break; + id++; + /* Check for wrap. If this happens, we have nearly 2^32 + * allocated RSS contexts, which seems unlikely. + */ + if (WARN_ON_ONCE(!id)) + return NULL; + } + + /* Create the new entry */ + new = kmalloc(sizeof(*new), GFP_KERNEL); + if (!new) + return NULL; + new->context_id = EFX_MCDI_RSS_CONTEXT_INVALID; + new->rx_hash_udp_4tuple = false; + + /* Insert the new entry into the gap */ + new->user_id = id; + list_add_tail(&new->list, &ctx->list); + return new; +} + +struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id) +{ + struct list_head *head = &efx->rss_context.list; + struct efx_rss_context *ctx; + + WARN_ON(!mutex_is_locked(&efx->rss_lock)); + + list_for_each_entry(ctx, head, list) + if (ctx->user_id == id) + return ctx; + return NULL; +} + +void efx_free_rss_context_entry(struct efx_rss_context *ctx) +{ + list_del(&ctx->list); + kfree(ctx); +} + +void efx_set_default_rx_indir_table(struct efx_nic *efx, + struct efx_rss_context *ctx) +{ + size_t i; + + for (i = 0; i < ARRAY_SIZE(ctx->rx_indir_table); i++) + ctx->rx_indir_table[i] = + ethtool_rxfh_indir_default(i, efx->rss_spread); +} + +/** + * efx_filter_is_mc_recipient - test whether spec is a multicast recipient + * @spec: Specification to test + * + * Return: %true if the specification is a non-drop RX filter that + * matches a local MAC address I/G bit value of 1 or matches a local + * IPv4 or IPv6 address value in the respective multicast address + * range. Otherwise %false. + */ +bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec) +{ + if (!(spec->flags & EFX_FILTER_FLAG_RX) || + spec->dmaq_id == EFX_FILTER_RX_DMAQ_ID_DROP) + return false; + + if (spec->match_flags & + (EFX_FILTER_MATCH_LOC_MAC | EFX_FILTER_MATCH_LOC_MAC_IG) && + is_multicast_ether_addr(spec->loc_mac)) + return true; + + if ((spec->match_flags & + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) == + (EFX_FILTER_MATCH_ETHER_TYPE | EFX_FILTER_MATCH_LOC_HOST)) { + if (spec->ether_type == htons(ETH_P_IP) && + ipv4_is_multicast(spec->loc_host[0])) + return true; + if (spec->ether_type == htons(ETH_P_IPV6) && + ((const u8 *)spec->loc_host)[0] == 0xff) + return true; + } + + return false; +} + +bool efx_filter_spec_equal(const struct efx_filter_spec *left, + const struct efx_filter_spec *right) +{ + if ((left->match_flags ^ right->match_flags) | + ((left->flags ^ right->flags) & + (EFX_FILTER_FLAG_RX | EFX_FILTER_FLAG_TX))) + return false; + + return memcmp(&left->outer_vid, &right->outer_vid, + sizeof(struct efx_filter_spec) - + offsetof(struct efx_filter_spec, outer_vid)) == 0; +} + +u32 efx_filter_spec_hash(const struct efx_filter_spec *spec) +{ + BUILD_BUG_ON(offsetof(struct efx_filter_spec, outer_vid) & 3); + return jhash2((const u32 *)&spec->outer_vid, + (sizeof(struct efx_filter_spec) - + offsetof(struct efx_filter_spec, outer_vid)) / 4, + 0); +} + +#ifdef CONFIG_RFS_ACCEL +bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, + bool *force) +{ + if (rule->filter_id == EFX_ARFS_FILTER_ID_PENDING) { + /* ARFS is currently updating this entry, leave it */ + return false; + } + if (rule->filter_id == EFX_ARFS_FILTER_ID_ERROR) { + /* ARFS tried and failed to update this, so it's probably out + * of date. Remove the filter and the ARFS rule entry. + */ + rule->filter_id = EFX_ARFS_FILTER_ID_REMOVING; + *force = true; + return true; + } else if (WARN_ON(rule->filter_id != filter_idx)) { /* can't happen */ + /* ARFS has moved on, so old filter is not needed. Since we did + * not mark the rule with EFX_ARFS_FILTER_ID_REMOVING, it will + * not be removed by efx_rps_hash_del() subsequently. + */ + *force = true; + return true; + } + /* Remove it iff ARFS wants to. */ + return true; +} + +static +struct hlist_head *efx_rps_hash_bucket(struct efx_nic *efx, + const struct efx_filter_spec *spec) +{ + u32 hash = efx_filter_spec_hash(spec); + + lockdep_assert_held(&efx->rps_hash_lock); + if (!efx->rps_hash_table) + return NULL; + return &efx->rps_hash_table[hash % EFX_ARFS_HASH_TABLE_SIZE]; +} + +struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, + const struct efx_filter_spec *spec) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (!head) + return NULL; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) + return rule; + } + return NULL; +} + +struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, + const struct efx_filter_spec *spec, + bool *new) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (!head) + return NULL; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) { + *new = false; + return rule; + } + } + rule = kmalloc(sizeof(*rule), GFP_ATOMIC); + *new = true; + if (rule) { + memcpy(&rule->spec, spec, sizeof(rule->spec)); + hlist_add_head(&rule->node, head); + } + return rule; +} + +void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec) +{ + struct efx_arfs_rule *rule; + struct hlist_head *head; + struct hlist_node *node; + + head = efx_rps_hash_bucket(efx, spec); + if (WARN_ON(!head)) + return; + hlist_for_each(node, head) { + rule = container_of(node, struct efx_arfs_rule, node); + if (efx_filter_spec_equal(spec, &rule->spec)) { + /* Someone already reused the entry. We know that if + * this check doesn't fire (i.e. filter_id == REMOVING) + * then the REMOVING mark was put there by our caller, + * because caller is holding a lock on filter table and + * only holders of that lock set REMOVING. + */ + if (rule->filter_id != EFX_ARFS_FILTER_ID_REMOVING) + return; + hlist_del(node); + kfree(rule); + return; + } + } + /* We didn't find it. */ + WARN_ON(1); +} +#endif + +int efx_probe_filters(struct efx_nic *efx) +{ + int rc; + + init_rwsem(&efx->filter_sem); + mutex_lock(&efx->mac_lock); + down_write(&efx->filter_sem); + rc = efx->type->filter_table_probe(efx); + if (rc) + goto out_unlock; + +#ifdef CONFIG_RFS_ACCEL + if (efx->type->offload_features & NETIF_F_NTUPLE) { + struct efx_channel *channel; + int i, success = 1; + + efx_for_each_channel(channel, efx) { + channel->rps_flow_id = + kcalloc(efx->type->max_rx_ip_filters, + sizeof(*channel->rps_flow_id), + GFP_KERNEL); + if (!channel->rps_flow_id) + success = 0; + else + for (i = 0; + i < efx->type->max_rx_ip_filters; + ++i) + channel->rps_flow_id[i] = + RPS_FLOW_ID_INVALID; + channel->rfs_expire_index = 0; + channel->rfs_filter_count = 0; + } + + if (!success) { + efx_for_each_channel(channel, efx) + kfree(channel->rps_flow_id); + efx->type->filter_table_remove(efx); + rc = -ENOMEM; + goto out_unlock; + } + } +#endif +out_unlock: + up_write(&efx->filter_sem); + mutex_unlock(&efx->mac_lock); + return rc; +} + +void efx_remove_filters(struct efx_nic *efx) +{ +#ifdef CONFIG_RFS_ACCEL + struct efx_channel *channel; + + efx_for_each_channel(channel, efx) { + cancel_delayed_work_sync(&channel->filter_work); + kfree(channel->rps_flow_id); + } +#endif + down_write(&efx->filter_sem); + efx->type->filter_table_remove(efx); + up_write(&efx->filter_sem); +} diff --git a/drivers/net/ethernet/sfc/rx_common.h b/drivers/net/ethernet/sfc/rx_common.h new file mode 100644 index 000000000000..c41f12a89477 --- /dev/null +++ b/drivers/net/ethernet/sfc/rx_common.h @@ -0,0 +1,97 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_RX_COMMON_H +#define EFX_RX_COMMON_H + +/* Preferred number of descriptors to fill at once */ +#define EFX_RX_PREFERRED_BATCH 8U + +/* Each packet can consume up to ceil(max_frame_len / buffer_size) buffers */ +#define EFX_RX_MAX_FRAGS DIV_ROUND_UP(EFX_MAX_FRAME_LEN(EFX_MAX_MTU), \ + EFX_RX_USR_BUF_SIZE) + +static inline u8 *efx_rx_buf_va(struct efx_rx_buffer *buf) +{ + return page_address(buf->page) + buf->page_offset; +} + +static inline u32 efx_rx_buf_hash(struct efx_nic *efx, const u8 *eh) +{ +#if defined(CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS) + return __le32_to_cpup((const __le32 *)(eh + efx->rx_packet_hash_offset)); +#else + const u8 *data = eh + efx->rx_packet_hash_offset; + + return (u32)data[0] | + (u32)data[1] << 8 | + (u32)data[2] << 16 | + (u32)data[3] << 24; +#endif +} + +void efx_rx_slow_fill(struct timer_list *t); + +void efx_recycle_rx_pages(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags); +void efx_discard_rx_packet(struct efx_channel *channel, + struct efx_rx_buffer *rx_buf, + unsigned int n_frags); + +int efx_probe_rx_queue(struct efx_rx_queue *rx_queue); +void efx_init_rx_queue(struct efx_rx_queue *rx_queue); +void efx_fini_rx_queue(struct efx_rx_queue *rx_queue); +void efx_remove_rx_queue(struct efx_rx_queue *rx_queue); +void efx_destroy_rx_queue(struct efx_rx_queue *rx_queue); + +void efx_init_rx_buffer(struct efx_rx_queue *rx_queue, + struct page *page, + unsigned int page_offset, + u16 flags); +void efx_unmap_rx_buffer(struct efx_nic *efx, struct efx_rx_buffer *rx_buf); +void efx_free_rx_buffers(struct efx_rx_queue *rx_queue, + struct efx_rx_buffer *rx_buf, + unsigned int num_bufs); + +void efx_schedule_slow_fill(struct efx_rx_queue *rx_queue); +void efx_rx_config_page_split(struct efx_nic *efx); +void efx_fast_push_rx_descriptors(struct efx_rx_queue *rx_queue, bool atomic); + +void +efx_rx_packet_gro(struct efx_channel *channel, struct efx_rx_buffer *rx_buf, + unsigned int n_frags, u8 *eh); + +struct efx_rss_context *efx_alloc_rss_context_entry(struct efx_nic *efx); +struct efx_rss_context *efx_find_rss_context_entry(struct efx_nic *efx, u32 id); +void efx_free_rss_context_entry(struct efx_rss_context *ctx); +void efx_set_default_rx_indir_table(struct efx_nic *efx, + struct efx_rss_context *ctx); + +bool efx_filter_is_mc_recipient(const struct efx_filter_spec *spec); +bool efx_filter_spec_equal(const struct efx_filter_spec *left, + const struct efx_filter_spec *right); +u32 efx_filter_spec_hash(const struct efx_filter_spec *spec); + +#ifdef CONFIG_RFS_ACCEL +bool efx_rps_check_rule(struct efx_arfs_rule *rule, unsigned int filter_idx, + bool *force); +struct efx_arfs_rule *efx_rps_hash_find(struct efx_nic *efx, + const struct efx_filter_spec *spec); +struct efx_arfs_rule *efx_rps_hash_add(struct efx_nic *efx, + const struct efx_filter_spec *spec, + bool *new); +void efx_rps_hash_del(struct efx_nic *efx, const struct efx_filter_spec *spec); +#endif + +int efx_probe_filters(struct efx_nic *efx); +void efx_remove_filters(struct efx_nic *efx); + +#endif diff --git a/drivers/net/ethernet/sfc/selftest.c b/drivers/net/ethernet/sfc/selftest.c index 8474cf8ea7d3..1ae369022d7d 100644 --- a/drivers/net/ethernet/sfc/selftest.c +++ b/drivers/net/ethernet/sfc/selftest.c @@ -18,6 +18,8 @@ #include <linux/slab.h> #include "net_driver.h" #include "efx.h" +#include "efx_common.h" +#include "efx_channels.h" #include "nic.h" #include "selftest.h" #include "workarounds.h" @@ -783,7 +785,7 @@ void efx_selftest_async_cancel(struct efx_nic *efx) cancel_delayed_work_sync(&efx->selftest_work); } -void efx_selftest_async_work(struct work_struct *data) +static void efx_selftest_async_work(struct work_struct *data) { struct efx_nic *efx = container_of(data, struct efx_nic, selftest_work.work); @@ -802,3 +804,8 @@ void efx_selftest_async_work(struct work_struct *data) channel->channel, cpu); } } + +void efx_selftest_async_init(struct efx_nic *efx) +{ + INIT_DELAYED_WORK(&efx->selftest_work, efx_selftest_async_work); +} diff --git a/drivers/net/ethernet/sfc/selftest.h b/drivers/net/ethernet/sfc/selftest.h index a3553816d92c..ca88ebb4f6b1 100644 --- a/drivers/net/ethernet/sfc/selftest.h +++ b/drivers/net/ethernet/sfc/selftest.h @@ -45,8 +45,8 @@ void efx_loopback_rx_packet(struct efx_nic *efx, const char *buf_ptr, int pkt_len); int efx_selftest(struct efx_nic *efx, struct efx_self_tests *tests, unsigned flags); +void efx_selftest_async_init(struct efx_nic *efx); void efx_selftest_async_start(struct efx_nic *efx); void efx_selftest_async_cancel(struct efx_nic *efx); -void efx_selftest_async_work(struct work_struct *data); #endif /* EFX_SELFTEST_H */ diff --git a/drivers/net/ethernet/sfc/siena.c b/drivers/net/ethernet/sfc/siena.c index 81499244a4b4..baa464161626 100644 --- a/drivers/net/ethernet/sfc/siena.c +++ b/drivers/net/ethernet/sfc/siena.c @@ -14,12 +14,14 @@ #include "net_driver.h" #include "bitfield.h" #include "efx.h" +#include "efx_common.h" #include "nic.h" #include "farch_regs.h" #include "io.h" #include "workarounds.h" #include "mcdi.h" #include "mcdi_pcol.h" +#include "mcdi_port_common.h" #include "selftest.h" #include "siena_sriov.h" diff --git a/drivers/net/ethernet/sfc/siena_sriov.c b/drivers/net/ethernet/sfc/siena_sriov.c index dfbdf05dcf79..83dcfcae3d4b 100644 --- a/drivers/net/ethernet/sfc/siena_sriov.c +++ b/drivers/net/ethernet/sfc/siena_sriov.c @@ -7,6 +7,7 @@ #include <linux/module.h> #include "net_driver.h" #include "efx.h" +#include "efx_channels.h" #include "nic.h" #include "io.h" #include "mcdi.h" diff --git a/drivers/net/ethernet/sfc/tx.c b/drivers/net/ethernet/sfc/tx.c index 00c1c4402451..04d7f41d7ed9 100644 --- a/drivers/net/ethernet/sfc/tx.c +++ b/drivers/net/ethernet/sfc/tx.c @@ -20,6 +20,7 @@ #include "io.h" #include "nic.h" #include "tx.h" +#include "tx_common.h" #include "workarounds.h" #include "ef10_regs.h" @@ -56,72 +57,6 @@ u8 *efx_tx_get_copy_buffer_limited(struct efx_tx_queue *tx_queue, return efx_tx_get_copy_buffer(tx_queue, buffer); } -static void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, - struct efx_tx_buffer *buffer, - unsigned int *pkts_compl, - unsigned int *bytes_compl) -{ - if (buffer->unmap_len) { - struct device *dma_dev = &tx_queue->efx->pci_dev->dev; - dma_addr_t unmap_addr = buffer->dma_addr - buffer->dma_offset; - if (buffer->flags & EFX_TX_BUF_MAP_SINGLE) - dma_unmap_single(dma_dev, unmap_addr, buffer->unmap_len, - DMA_TO_DEVICE); - else - dma_unmap_page(dma_dev, unmap_addr, buffer->unmap_len, - DMA_TO_DEVICE); - buffer->unmap_len = 0; - } - - if (buffer->flags & EFX_TX_BUF_SKB) { - struct sk_buff *skb = (struct sk_buff *)buffer->skb; - - EFX_WARN_ON_PARANOID(!pkts_compl || !bytes_compl); - (*pkts_compl)++; - (*bytes_compl) += skb->len; - if (tx_queue->timestamping && - (tx_queue->completed_timestamp_major || - tx_queue->completed_timestamp_minor)) { - struct skb_shared_hwtstamps hwtstamp; - - hwtstamp.hwtstamp = - efx_ptp_nic_to_kernel_time(tx_queue); - skb_tstamp_tx(skb, &hwtstamp); - - tx_queue->completed_timestamp_major = 0; - tx_queue->completed_timestamp_minor = 0; - } - dev_consume_skb_any((struct sk_buff *)buffer->skb); - netif_vdbg(tx_queue->efx, tx_done, tx_queue->efx->net_dev, - "TX queue %d transmission id %x complete\n", - tx_queue->queue, tx_queue->read_count); - } else if (buffer->flags & EFX_TX_BUF_XDP) { - xdp_return_frame_rx_napi(buffer->xdpf); - } - - buffer->len = 0; - buffer->flags = 0; -} - -unsigned int efx_tx_max_skb_descs(struct efx_nic *efx) -{ - /* Header and payload descriptor for each output segment, plus - * one for every input fragment boundary within a segment - */ - unsigned int max_descs = EFX_TSO_MAX_SEGS * 2 + MAX_SKB_FRAGS; - - /* Possibly one more per segment for option descriptors */ - if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) - max_descs += EFX_TSO_MAX_SEGS; - - /* Possibly more for PCIe page boundaries within input fragments */ - if (PAGE_SIZE > EFX_PAGE_SIZE) - max_descs += max_t(unsigned int, MAX_SKB_FRAGS, - DIV_ROUND_UP(GSO_MAX_SIZE, EFX_PAGE_SIZE)); - - return max_descs; -} - static void efx_tx_maybe_stop_queue(struct efx_tx_queue *txq1) { /* We need to consider both queues that the net core sees as one */ @@ -333,125 +268,6 @@ static int efx_enqueue_skb_pio(struct efx_tx_queue *tx_queue, } #endif /* EFX_USE_PIO */ -static struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, - dma_addr_t dma_addr, - size_t len) -{ - const struct efx_nic_type *nic_type = tx_queue->efx->type; - struct efx_tx_buffer *buffer; - unsigned int dma_len; - - /* Map the fragment taking account of NIC-dependent DMA limits. */ - do { - buffer = efx_tx_queue_get_insert_buffer(tx_queue); - dma_len = nic_type->tx_limit_len(tx_queue, dma_addr, len); - - buffer->len = dma_len; - buffer->dma_addr = dma_addr; - buffer->flags = EFX_TX_BUF_CONT; - len -= dma_len; - dma_addr += dma_len; - ++tx_queue->insert_count; - } while (len); - - return buffer; -} - -/* Map all data from an SKB for DMA and create descriptors on the queue. - */ -static int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, - unsigned int segment_count) -{ - struct efx_nic *efx = tx_queue->efx; - struct device *dma_dev = &efx->pci_dev->dev; - unsigned int frag_index, nr_frags; - dma_addr_t dma_addr, unmap_addr; - unsigned short dma_flags; - size_t len, unmap_len; - - nr_frags = skb_shinfo(skb)->nr_frags; - frag_index = 0; - - /* Map header data. */ - len = skb_headlen(skb); - dma_addr = dma_map_single(dma_dev, skb->data, len, DMA_TO_DEVICE); - dma_flags = EFX_TX_BUF_MAP_SINGLE; - unmap_len = len; - unmap_addr = dma_addr; - - if (unlikely(dma_mapping_error(dma_dev, dma_addr))) - return -EIO; - - if (segment_count) { - /* For TSO we need to put the header in to a separate - * descriptor. Map this separately if necessary. - */ - size_t header_len = skb_transport_header(skb) - skb->data + - (tcp_hdr(skb)->doff << 2u); - - if (header_len != len) { - tx_queue->tso_long_headers++; - efx_tx_map_chunk(tx_queue, dma_addr, header_len); - len -= header_len; - dma_addr += header_len; - } - } - - /* Add descriptors for each fragment. */ - do { - struct efx_tx_buffer *buffer; - skb_frag_t *fragment; - - buffer = efx_tx_map_chunk(tx_queue, dma_addr, len); - - /* The final descriptor for a fragment is responsible for - * unmapping the whole fragment. - */ - buffer->flags = EFX_TX_BUF_CONT | dma_flags; - buffer->unmap_len = unmap_len; - buffer->dma_offset = buffer->dma_addr - unmap_addr; - - if (frag_index >= nr_frags) { - /* Store SKB details with the final buffer for - * the completion. - */ - buffer->skb = skb; - buffer->flags = EFX_TX_BUF_SKB | dma_flags; - return 0; - } - - /* Move on to the next fragment. */ - fragment = &skb_shinfo(skb)->frags[frag_index++]; - len = skb_frag_size(fragment); - dma_addr = skb_frag_dma_map(dma_dev, fragment, - 0, len, DMA_TO_DEVICE); - dma_flags = 0; - unmap_len = len; - unmap_addr = dma_addr; - - if (unlikely(dma_mapping_error(dma_dev, dma_addr))) - return -EIO; - } while (1); -} - -/* Remove buffers put into a tx_queue for the current packet. - * None of the buffers must have an skb attached. - */ -static void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, - unsigned int insert_count) -{ - struct efx_tx_buffer *buffer; - unsigned int bytes_compl = 0; - unsigned int pkts_compl = 0; - - /* Work backwards until we hit the original insert pointer value */ - while (tx_queue->insert_count != insert_count) { - --tx_queue->insert_count; - buffer = __efx_tx_queue_get_insert_buffer(tx_queue); - efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); - } -} - /* * Fallback to software TSO. * @@ -473,12 +289,9 @@ static int efx_tx_tso_fallback(struct efx_tx_queue *tx_queue, dev_consume_skb_any(skb); skb = segments; - while (skb) { - next = skb->next; - skb->next = NULL; - + skb_list_walk_safe(skb, skb, next) { + skb_mark_not_on_list(skb); efx_enqueue_skb(tx_queue, skb); - skb = next; } return 0; @@ -687,41 +500,6 @@ int efx_xdp_tx_buffers(struct efx_nic *efx, int n, struct xdp_frame **xdpfs, return i; } -/* Remove packets from the TX queue - * - * This removes packets from the TX queue, up to and including the - * specified index. - */ -static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue, - unsigned int index, - unsigned int *pkts_compl, - unsigned int *bytes_compl) -{ - struct efx_nic *efx = tx_queue->efx; - unsigned int stop_index, read_ptr; - - stop_index = (index + 1) & tx_queue->ptr_mask; - read_ptr = tx_queue->read_count & tx_queue->ptr_mask; - - while (read_ptr != stop_index) { - struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr]; - - if (!(buffer->flags & EFX_TX_BUF_OPTION) && - unlikely(buffer->len == 0)) { - netif_err(efx, tx_err, efx->net_dev, - "TX queue %d spurious TX completion id %x\n", - tx_queue->queue, read_ptr); - efx_schedule_reset(efx, RESET_TYPE_TX_SKIP); - return; - } - - efx_dequeue_buffer(tx_queue, buffer, pkts_compl, bytes_compl); - - ++tx_queue->read_count; - read_ptr = tx_queue->read_count & tx_queue->ptr_mask; - } -} - /* Initiate a packet transmission. We use one channel per CPU * (sharing when we have more CPUs than channels). On Falcon, the TX * completion events will be directed back to the CPU that transmitted @@ -834,173 +612,3 @@ int efx_setup_tc(struct net_device *net_dev, enum tc_setup_type type, net_dev->num_tc = num_tc; return 0; } - -void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index) -{ - unsigned fill_level; - struct efx_nic *efx = tx_queue->efx; - struct efx_tx_queue *txq2; - unsigned int pkts_compl = 0, bytes_compl = 0; - - EFX_WARN_ON_ONCE_PARANOID(index > tx_queue->ptr_mask); - - efx_dequeue_buffers(tx_queue, index, &pkts_compl, &bytes_compl); - tx_queue->pkts_compl += pkts_compl; - tx_queue->bytes_compl += bytes_compl; - - if (pkts_compl > 1) - ++tx_queue->merge_events; - - /* See if we need to restart the netif queue. This memory - * barrier ensures that we write read_count (inside - * efx_dequeue_buffers()) before reading the queue status. - */ - smp_mb(); - if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) && - likely(efx->port_enabled) && - likely(netif_device_present(efx->net_dev))) { - txq2 = efx_tx_queue_partner(tx_queue); - fill_level = max(tx_queue->insert_count - tx_queue->read_count, - txq2->insert_count - txq2->read_count); - if (fill_level <= efx->txq_wake_thresh) - netif_tx_wake_queue(tx_queue->core_txq); - } - - /* Check whether the hardware queue is now empty */ - if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) { - tx_queue->old_write_count = READ_ONCE(tx_queue->write_count); - if (tx_queue->read_count == tx_queue->old_write_count) { - smp_mb(); - tx_queue->empty_read_count = - tx_queue->read_count | EFX_EMPTY_COUNT_VALID; - } - } -} - -static unsigned int efx_tx_cb_page_count(struct efx_tx_queue *tx_queue) -{ - return DIV_ROUND_UP(tx_queue->ptr_mask + 1, PAGE_SIZE >> EFX_TX_CB_ORDER); -} - -int efx_probe_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_nic *efx = tx_queue->efx; - unsigned int entries; - int rc; - - /* Create the smallest power-of-two aligned ring */ - entries = max(roundup_pow_of_two(efx->txq_entries), EFX_MIN_DMAQ_SIZE); - EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); - tx_queue->ptr_mask = entries - 1; - - netif_dbg(efx, probe, efx->net_dev, - "creating TX queue %d size %#x mask %#x\n", - tx_queue->queue, efx->txq_entries, tx_queue->ptr_mask); - - /* Allocate software ring */ - tx_queue->buffer = kcalloc(entries, sizeof(*tx_queue->buffer), - GFP_KERNEL); - if (!tx_queue->buffer) - return -ENOMEM; - - tx_queue->cb_page = kcalloc(efx_tx_cb_page_count(tx_queue), - sizeof(tx_queue->cb_page[0]), GFP_KERNEL); - if (!tx_queue->cb_page) { - rc = -ENOMEM; - goto fail1; - } - - /* Allocate hardware ring */ - rc = efx_nic_probe_tx(tx_queue); - if (rc) - goto fail2; - - return 0; - -fail2: - kfree(tx_queue->cb_page); - tx_queue->cb_page = NULL; -fail1: - kfree(tx_queue->buffer); - tx_queue->buffer = NULL; - return rc; -} - -void efx_init_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_nic *efx = tx_queue->efx; - - netif_dbg(efx, drv, efx->net_dev, - "initialising TX queue %d\n", tx_queue->queue); - - tx_queue->insert_count = 0; - tx_queue->write_count = 0; - tx_queue->packet_write_count = 0; - tx_queue->old_write_count = 0; - tx_queue->read_count = 0; - tx_queue->old_read_count = 0; - tx_queue->empty_read_count = 0 | EFX_EMPTY_COUNT_VALID; - tx_queue->xmit_more_available = false; - tx_queue->timestamping = (efx_ptp_use_mac_tx_timestamps(efx) && - tx_queue->channel == efx_ptp_channel(efx)); - tx_queue->completed_desc_ptr = tx_queue->ptr_mask; - tx_queue->completed_timestamp_major = 0; - tx_queue->completed_timestamp_minor = 0; - - tx_queue->xdp_tx = efx_channel_is_xdp_tx(tx_queue->channel); - - /* Set up default function pointers. These may get replaced by - * efx_nic_init_tx() based off NIC/queue capabilities. - */ - tx_queue->handle_tso = efx_enqueue_skb_tso; - - /* Set up TX descriptor ring */ - efx_nic_init_tx(tx_queue); - - tx_queue->initialised = true; -} - -void efx_fini_tx_queue(struct efx_tx_queue *tx_queue) -{ - struct efx_tx_buffer *buffer; - - netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, - "shutting down TX queue %d\n", tx_queue->queue); - - if (!tx_queue->buffer) - return; - - /* Free any buffers left in the ring */ - while (tx_queue->read_count != tx_queue->write_count) { - unsigned int pkts_compl = 0, bytes_compl = 0; - buffer = &tx_queue->buffer[tx_queue->read_count & tx_queue->ptr_mask]; - efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); - - ++tx_queue->read_count; - } - tx_queue->xmit_more_available = false; - netdev_tx_reset_queue(tx_queue->core_txq); -} - -void efx_remove_tx_queue(struct efx_tx_queue *tx_queue) -{ - int i; - - if (!tx_queue->buffer) - return; - - netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, - "destroying TX queue %d\n", tx_queue->queue); - efx_nic_remove_tx(tx_queue); - - if (tx_queue->cb_page) { - for (i = 0; i < efx_tx_cb_page_count(tx_queue); i++) - efx_nic_free_buffer(tx_queue->efx, - &tx_queue->cb_page[i]); - kfree(tx_queue->cb_page); - tx_queue->cb_page = NULL; - } - - kfree(tx_queue->buffer); - tx_queue->buffer = NULL; -} diff --git a/drivers/net/ethernet/sfc/tx_common.c b/drivers/net/ethernet/sfc/tx_common.c new file mode 100644 index 000000000000..b1571e9789d0 --- /dev/null +++ b/drivers/net/ethernet/sfc/tx_common.c @@ -0,0 +1,404 @@ +// SPDX-License-Identifier: GPL-2.0-only +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#include "net_driver.h" +#include "efx.h" +#include "nic.h" +#include "tx_common.h" + +static unsigned int efx_tx_cb_page_count(struct efx_tx_queue *tx_queue) +{ + return DIV_ROUND_UP(tx_queue->ptr_mask + 1, + PAGE_SIZE >> EFX_TX_CB_ORDER); +} + +int efx_probe_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_nic *efx = tx_queue->efx; + unsigned int entries; + int rc; + + /* Create the smallest power-of-two aligned ring */ + entries = max(roundup_pow_of_two(efx->txq_entries), EFX_MIN_DMAQ_SIZE); + EFX_WARN_ON_PARANOID(entries > EFX_MAX_DMAQ_SIZE); + tx_queue->ptr_mask = entries - 1; + + netif_dbg(efx, probe, efx->net_dev, + "creating TX queue %d size %#x mask %#x\n", + tx_queue->queue, efx->txq_entries, tx_queue->ptr_mask); + + /* Allocate software ring */ + tx_queue->buffer = kcalloc(entries, sizeof(*tx_queue->buffer), + GFP_KERNEL); + if (!tx_queue->buffer) + return -ENOMEM; + + tx_queue->cb_page = kcalloc(efx_tx_cb_page_count(tx_queue), + sizeof(tx_queue->cb_page[0]), GFP_KERNEL); + if (!tx_queue->cb_page) { + rc = -ENOMEM; + goto fail1; + } + + /* Allocate hardware ring */ + rc = efx_nic_probe_tx(tx_queue); + if (rc) + goto fail2; + + return 0; + +fail2: + kfree(tx_queue->cb_page); + tx_queue->cb_page = NULL; +fail1: + kfree(tx_queue->buffer); + tx_queue->buffer = NULL; + return rc; +} + +void efx_init_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_nic *efx = tx_queue->efx; + + netif_dbg(efx, drv, efx->net_dev, + "initialising TX queue %d\n", tx_queue->queue); + + tx_queue->insert_count = 0; + tx_queue->write_count = 0; + tx_queue->packet_write_count = 0; + tx_queue->old_write_count = 0; + tx_queue->read_count = 0; + tx_queue->old_read_count = 0; + tx_queue->empty_read_count = 0 | EFX_EMPTY_COUNT_VALID; + tx_queue->xmit_more_available = false; + tx_queue->timestamping = (efx_ptp_use_mac_tx_timestamps(efx) && + tx_queue->channel == efx_ptp_channel(efx)); + tx_queue->completed_desc_ptr = tx_queue->ptr_mask; + tx_queue->completed_timestamp_major = 0; + tx_queue->completed_timestamp_minor = 0; + + tx_queue->xdp_tx = efx_channel_is_xdp_tx(tx_queue->channel); + + /* Set up default function pointers. These may get replaced by + * efx_nic_init_tx() based off NIC/queue capabilities. + */ + tx_queue->handle_tso = efx_enqueue_skb_tso; + + /* Set up TX descriptor ring */ + efx_nic_init_tx(tx_queue); + + tx_queue->initialised = true; +} + +void efx_fini_tx_queue(struct efx_tx_queue *tx_queue) +{ + struct efx_tx_buffer *buffer; + + netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, + "shutting down TX queue %d\n", tx_queue->queue); + + if (!tx_queue->buffer) + return; + + /* Free any buffers left in the ring */ + while (tx_queue->read_count != tx_queue->write_count) { + unsigned int pkts_compl = 0, bytes_compl = 0; + + buffer = &tx_queue->buffer[tx_queue->read_count & tx_queue->ptr_mask]; + efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); + + ++tx_queue->read_count; + } + tx_queue->xmit_more_available = false; + netdev_tx_reset_queue(tx_queue->core_txq); +} + +void efx_remove_tx_queue(struct efx_tx_queue *tx_queue) +{ + int i; + + if (!tx_queue->buffer) + return; + + netif_dbg(tx_queue->efx, drv, tx_queue->efx->net_dev, + "destroying TX queue %d\n", tx_queue->queue); + efx_nic_remove_tx(tx_queue); + + if (tx_queue->cb_page) { + for (i = 0; i < efx_tx_cb_page_count(tx_queue); i++) + efx_nic_free_buffer(tx_queue->efx, + &tx_queue->cb_page[i]); + kfree(tx_queue->cb_page); + tx_queue->cb_page = NULL; + } + + kfree(tx_queue->buffer); + tx_queue->buffer = NULL; +} + +void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, + struct efx_tx_buffer *buffer, + unsigned int *pkts_compl, + unsigned int *bytes_compl) +{ + if (buffer->unmap_len) { + struct device *dma_dev = &tx_queue->efx->pci_dev->dev; + dma_addr_t unmap_addr = buffer->dma_addr - buffer->dma_offset; + + if (buffer->flags & EFX_TX_BUF_MAP_SINGLE) + dma_unmap_single(dma_dev, unmap_addr, buffer->unmap_len, + DMA_TO_DEVICE); + else + dma_unmap_page(dma_dev, unmap_addr, buffer->unmap_len, + DMA_TO_DEVICE); + buffer->unmap_len = 0; + } + + if (buffer->flags & EFX_TX_BUF_SKB) { + struct sk_buff *skb = (struct sk_buff *)buffer->skb; + + EFX_WARN_ON_PARANOID(!pkts_compl || !bytes_compl); + (*pkts_compl)++; + (*bytes_compl) += skb->len; + if (tx_queue->timestamping && + (tx_queue->completed_timestamp_major || + tx_queue->completed_timestamp_minor)) { + struct skb_shared_hwtstamps hwtstamp; + + hwtstamp.hwtstamp = + efx_ptp_nic_to_kernel_time(tx_queue); + skb_tstamp_tx(skb, &hwtstamp); + + tx_queue->completed_timestamp_major = 0; + tx_queue->completed_timestamp_minor = 0; + } + dev_consume_skb_any((struct sk_buff *)buffer->skb); + netif_vdbg(tx_queue->efx, tx_done, tx_queue->efx->net_dev, + "TX queue %d transmission id %x complete\n", + tx_queue->queue, tx_queue->read_count); + } else if (buffer->flags & EFX_TX_BUF_XDP) { + xdp_return_frame_rx_napi(buffer->xdpf); + } + + buffer->len = 0; + buffer->flags = 0; +} + +/* Remove packets from the TX queue + * + * This removes packets from the TX queue, up to and including the + * specified index. + */ +static void efx_dequeue_buffers(struct efx_tx_queue *tx_queue, + unsigned int index, + unsigned int *pkts_compl, + unsigned int *bytes_compl) +{ + struct efx_nic *efx = tx_queue->efx; + unsigned int stop_index, read_ptr; + + stop_index = (index + 1) & tx_queue->ptr_mask; + read_ptr = tx_queue->read_count & tx_queue->ptr_mask; + + while (read_ptr != stop_index) { + struct efx_tx_buffer *buffer = &tx_queue->buffer[read_ptr]; + + if (!(buffer->flags & EFX_TX_BUF_OPTION) && + unlikely(buffer->len == 0)) { + netif_err(efx, tx_err, efx->net_dev, + "TX queue %d spurious TX completion id %x\n", + tx_queue->queue, read_ptr); + efx_schedule_reset(efx, RESET_TYPE_TX_SKIP); + return; + } + + efx_dequeue_buffer(tx_queue, buffer, pkts_compl, bytes_compl); + + ++tx_queue->read_count; + read_ptr = tx_queue->read_count & tx_queue->ptr_mask; + } +} + +void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index) +{ + unsigned int fill_level, pkts_compl = 0, bytes_compl = 0; + struct efx_nic *efx = tx_queue->efx; + struct efx_tx_queue *txq2; + + EFX_WARN_ON_ONCE_PARANOID(index > tx_queue->ptr_mask); + + efx_dequeue_buffers(tx_queue, index, &pkts_compl, &bytes_compl); + tx_queue->pkts_compl += pkts_compl; + tx_queue->bytes_compl += bytes_compl; + + if (pkts_compl > 1) + ++tx_queue->merge_events; + + /* See if we need to restart the netif queue. This memory + * barrier ensures that we write read_count (inside + * efx_dequeue_buffers()) before reading the queue status. + */ + smp_mb(); + if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) && + likely(efx->port_enabled) && + likely(netif_device_present(efx->net_dev))) { + txq2 = efx_tx_queue_partner(tx_queue); + fill_level = max(tx_queue->insert_count - tx_queue->read_count, + txq2->insert_count - txq2->read_count); + if (fill_level <= efx->txq_wake_thresh) + netif_tx_wake_queue(tx_queue->core_txq); + } + + /* Check whether the hardware queue is now empty */ + if ((int)(tx_queue->read_count - tx_queue->old_write_count) >= 0) { + tx_queue->old_write_count = READ_ONCE(tx_queue->write_count); + if (tx_queue->read_count == tx_queue->old_write_count) { + smp_mb(); + tx_queue->empty_read_count = + tx_queue->read_count | EFX_EMPTY_COUNT_VALID; + } + } +} + +/* Remove buffers put into a tx_queue for the current packet. + * None of the buffers must have an skb attached. + */ +void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, + unsigned int insert_count) +{ + struct efx_tx_buffer *buffer; + unsigned int bytes_compl = 0; + unsigned int pkts_compl = 0; + + /* Work backwards until we hit the original insert pointer value */ + while (tx_queue->insert_count != insert_count) { + --tx_queue->insert_count; + buffer = __efx_tx_queue_get_insert_buffer(tx_queue); + efx_dequeue_buffer(tx_queue, buffer, &pkts_compl, &bytes_compl); + } +} + +struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, + dma_addr_t dma_addr, size_t len) +{ + const struct efx_nic_type *nic_type = tx_queue->efx->type; + struct efx_tx_buffer *buffer; + unsigned int dma_len; + + /* Map the fragment taking account of NIC-dependent DMA limits. */ + do { + buffer = efx_tx_queue_get_insert_buffer(tx_queue); + dma_len = nic_type->tx_limit_len(tx_queue, dma_addr, len); + + buffer->len = dma_len; + buffer->dma_addr = dma_addr; + buffer->flags = EFX_TX_BUF_CONT; + len -= dma_len; + dma_addr += dma_len; + ++tx_queue->insert_count; + } while (len); + + return buffer; +} + +/* Map all data from an SKB for DMA and create descriptors on the queue. */ +int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + unsigned int segment_count) +{ + struct efx_nic *efx = tx_queue->efx; + struct device *dma_dev = &efx->pci_dev->dev; + unsigned int frag_index, nr_frags; + dma_addr_t dma_addr, unmap_addr; + unsigned short dma_flags; + size_t len, unmap_len; + + nr_frags = skb_shinfo(skb)->nr_frags; + frag_index = 0; + + /* Map header data. */ + len = skb_headlen(skb); + dma_addr = dma_map_single(dma_dev, skb->data, len, DMA_TO_DEVICE); + dma_flags = EFX_TX_BUF_MAP_SINGLE; + unmap_len = len; + unmap_addr = dma_addr; + + if (unlikely(dma_mapping_error(dma_dev, dma_addr))) + return -EIO; + + if (segment_count) { + /* For TSO we need to put the header in to a separate + * descriptor. Map this separately if necessary. + */ + size_t header_len = skb_transport_header(skb) - skb->data + + (tcp_hdr(skb)->doff << 2u); + + if (header_len != len) { + tx_queue->tso_long_headers++; + efx_tx_map_chunk(tx_queue, dma_addr, header_len); + len -= header_len; + dma_addr += header_len; + } + } + + /* Add descriptors for each fragment. */ + do { + struct efx_tx_buffer *buffer; + skb_frag_t *fragment; + + buffer = efx_tx_map_chunk(tx_queue, dma_addr, len); + + /* The final descriptor for a fragment is responsible for + * unmapping the whole fragment. + */ + buffer->flags = EFX_TX_BUF_CONT | dma_flags; + buffer->unmap_len = unmap_len; + buffer->dma_offset = buffer->dma_addr - unmap_addr; + + if (frag_index >= nr_frags) { + /* Store SKB details with the final buffer for + * the completion. + */ + buffer->skb = skb; + buffer->flags = EFX_TX_BUF_SKB | dma_flags; + return 0; + } + + /* Move on to the next fragment. */ + fragment = &skb_shinfo(skb)->frags[frag_index++]; + len = skb_frag_size(fragment); + dma_addr = skb_frag_dma_map(dma_dev, fragment, 0, len, + DMA_TO_DEVICE); + dma_flags = 0; + unmap_len = len; + unmap_addr = dma_addr; + + if (unlikely(dma_mapping_error(dma_dev, dma_addr))) + return -EIO; + } while (1); +} + +unsigned int efx_tx_max_skb_descs(struct efx_nic *efx) +{ + /* Header and payload descriptor for each output segment, plus + * one for every input fragment boundary within a segment + */ + unsigned int max_descs = EFX_TSO_MAX_SEGS * 2 + MAX_SKB_FRAGS; + + /* Possibly one more per segment for option descriptors */ + if (efx_nic_rev(efx) >= EFX_REV_HUNT_A0) + max_descs += EFX_TSO_MAX_SEGS; + + /* Possibly more for PCIe page boundaries within input fragments */ + if (PAGE_SIZE > EFX_PAGE_SIZE) + max_descs += max_t(unsigned int, MAX_SKB_FRAGS, + DIV_ROUND_UP(GSO_MAX_SIZE, EFX_PAGE_SIZE)); + + return max_descs; +} diff --git a/drivers/net/ethernet/sfc/tx_common.h b/drivers/net/ethernet/sfc/tx_common.h new file mode 100644 index 000000000000..f92f1fe3a87f --- /dev/null +++ b/drivers/net/ethernet/sfc/tx_common.h @@ -0,0 +1,36 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/**************************************************************************** + * Driver for Solarflare network controllers and boards + * Copyright 2018 Solarflare Communications Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation, incorporated herein by reference. + */ + +#ifndef EFX_TX_COMMON_H +#define EFX_TX_COMMON_H + +int efx_probe_tx_queue(struct efx_tx_queue *tx_queue); +void efx_init_tx_queue(struct efx_tx_queue *tx_queue); +void efx_fini_tx_queue(struct efx_tx_queue *tx_queue); +void efx_remove_tx_queue(struct efx_tx_queue *tx_queue); + +void efx_dequeue_buffer(struct efx_tx_queue *tx_queue, + struct efx_tx_buffer *buffer, + unsigned int *pkts_compl, + unsigned int *bytes_compl); + +void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index); + +void efx_enqueue_unwind(struct efx_tx_queue *tx_queue, + unsigned int insert_count); + +struct efx_tx_buffer *efx_tx_map_chunk(struct efx_tx_queue *tx_queue, + dma_addr_t dma_addr, size_t len); +int efx_tx_map_data(struct efx_tx_queue *tx_queue, struct sk_buff *skb, + unsigned int segment_count); + +unsigned int efx_tx_max_skb_descs(struct efx_nic *efx); + +#endif diff --git a/drivers/net/ethernet/sgi/ioc3-eth.c b/drivers/net/ethernet/sgi/ioc3-eth.c index d242906ae233..06637b03deed 100644 --- a/drivers/net/ethernet/sgi/ioc3-eth.c +++ b/drivers/net/ethernet/sgi/ioc3-eth.c @@ -114,7 +114,7 @@ struct ioc3_private { static int ioc3_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static void ioc3_set_multicast_list(struct net_device *dev); static netdev_tx_t ioc3_start_xmit(struct sk_buff *skb, struct net_device *dev); -static void ioc3_timeout(struct net_device *dev); +static void ioc3_timeout(struct net_device *dev, unsigned int txqueue); static inline unsigned int ioc3_hash(const unsigned char *addr); static void ioc3_start(struct ioc3_private *ip); static inline void ioc3_stop(struct ioc3_private *ip); @@ -1479,7 +1479,7 @@ drop_packet: return NETDEV_TX_OK; } -static void ioc3_timeout(struct net_device *dev) +static void ioc3_timeout(struct net_device *dev, unsigned int txqueue) { struct ioc3_private *ip = netdev_priv(dev); diff --git a/drivers/net/ethernet/sgi/meth.c b/drivers/net/ethernet/sgi/meth.c index 539bc5db989c..0c396ecd3389 100644 --- a/drivers/net/ethernet/sgi/meth.c +++ b/drivers/net/ethernet/sgi/meth.c @@ -90,7 +90,7 @@ struct meth_private { spinlock_t meth_lock; }; -static void meth_tx_timeout(struct net_device *dev); +static void meth_tx_timeout(struct net_device *dev, unsigned int txqueue); static irqreturn_t meth_interrupt(int irq, void *dev_id); /* global, initialized in ip32-setup.c */ @@ -727,7 +727,7 @@ static netdev_tx_t meth_tx(struct sk_buff *skb, struct net_device *dev) /* * Deal with a transmit timeout. */ -static void meth_tx_timeout(struct net_device *dev) +static void meth_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct meth_private *priv = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/ethernet/silan/sc92031.c b/drivers/net/ethernet/silan/sc92031.c index c7641a236eb8..cb043eb1bdc1 100644 --- a/drivers/net/ethernet/silan/sc92031.c +++ b/drivers/net/ethernet/silan/sc92031.c @@ -1078,7 +1078,7 @@ static void sc92031_set_multicast_list(struct net_device *dev) spin_unlock_bh(&priv->lock); } -static void sc92031_tx_timeout(struct net_device *dev) +static void sc92031_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sc92031_priv *priv = netdev_priv(dev); diff --git a/drivers/net/ethernet/sis/sis190.c b/drivers/net/ethernet/sis/sis190.c index 5b351beb78cb..5a4b6e3ab38f 100644 --- a/drivers/net/ethernet/sis/sis190.c +++ b/drivers/net/ethernet/sis/sis190.c @@ -1538,7 +1538,7 @@ err_out_0: goto out; } -static void sis190_tx_timeout(struct net_device *dev) +static void sis190_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sis190_private *tp = netdev_priv(dev); void __iomem *ioaddr = tp->mmio_addr; diff --git a/drivers/net/ethernet/sis/sis900.c b/drivers/net/ethernet/sis/sis900.c index 85eaccbbbac1..81ed7589e33c 100644 --- a/drivers/net/ethernet/sis/sis900.c +++ b/drivers/net/ethernet/sis/sis900.c @@ -222,7 +222,7 @@ static int mdio_read(struct net_device *net_dev, int phy_id, int location); static void mdio_write(struct net_device *net_dev, int phy_id, int location, int val); static void sis900_timer(struct timer_list *t); static void sis900_check_mode (struct net_device *net_dev, struct mii_phy *mii_phy); -static void sis900_tx_timeout(struct net_device *net_dev); +static void sis900_tx_timeout(struct net_device *net_dev, unsigned int txqueue); static void sis900_init_tx_ring(struct net_device *net_dev); static void sis900_init_rx_ring(struct net_device *net_dev); static netdev_tx_t sis900_start_xmit(struct sk_buff *skb, @@ -1537,7 +1537,7 @@ static void sis900_read_mode(struct net_device *net_dev, int *speed, int *duplex * disable interrupts and do some tasks */ -static void sis900_tx_timeout(struct net_device *net_dev) +static void sis900_tx_timeout(struct net_device *net_dev, unsigned int txqueue) { struct sis900_private *sis_priv = netdev_priv(net_dev); void __iomem *ioaddr = sis_priv->ioaddr; diff --git a/drivers/net/ethernet/smsc/epic100.c b/drivers/net/ethernet/smsc/epic100.c index be47d864f8b9..61ddee0c2a2e 100644 --- a/drivers/net/ethernet/smsc/epic100.c +++ b/drivers/net/ethernet/smsc/epic100.c @@ -280,6 +280,7 @@ struct epic_private { signed char phys[4]; /* MII device addresses. */ u16 advertising; /* NWay media advertisement */ int mii_phy_cnt; + u32 ethtool_ops_nesting; struct mii_if_info mii; unsigned int tx_full:1; /* The Tx queue is full. */ unsigned int default_port:4; /* Last dev->if_port value. */ @@ -291,7 +292,7 @@ static int mdio_read(struct net_device *dev, int phy_id, int location); static void mdio_write(struct net_device *dev, int phy_id, int loc, int val); static void epic_restart(struct net_device *dev); static void epic_timer(struct timer_list *t); -static void epic_tx_timeout(struct net_device *dev); +static void epic_tx_timeout(struct net_device *dev, unsigned int txqueue); static void epic_init_ring(struct net_device *dev); static netdev_tx_t epic_start_xmit(struct sk_buff *skb, struct net_device *dev); @@ -861,7 +862,7 @@ static void epic_timer(struct timer_list *t) add_timer(&ep->timer); } -static void epic_tx_timeout(struct net_device *dev) +static void epic_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct epic_private *ep = netdev_priv(dev); void __iomem *ioaddr = ep->ioaddr; @@ -1435,8 +1436,10 @@ static int ethtool_begin(struct net_device *dev) struct epic_private *ep = netdev_priv(dev); void __iomem *ioaddr = ep->ioaddr; + if (ep->ethtool_ops_nesting == U32_MAX) + return -EBUSY; /* power-up, if interface is down */ - if (!netif_running(dev)) { + if (!ep->ethtool_ops_nesting++ && !netif_running(dev)) { ew32(GENCTL, 0x0200); ew32(NVCTL, (er32(NVCTL) & ~0x003c) | 0x4800); } @@ -1449,7 +1452,7 @@ static void ethtool_complete(struct net_device *dev) void __iomem *ioaddr = ep->ioaddr; /* power-down, if interface is down */ - if (!netif_running(dev)) { + if (!--ep->ethtool_ops_nesting && !netif_running(dev)) { ew32(GENCTL, 0x0008); ew32(NVCTL, (er32(NVCTL) & ~0x483c) | 0x0000); } diff --git a/drivers/net/ethernet/smsc/smc911x.c b/drivers/net/ethernet/smsc/smc911x.c index 8d88e4083456..186c0bddbe5f 100644 --- a/drivers/net/ethernet/smsc/smc911x.c +++ b/drivers/net/ethernet/smsc/smc911x.c @@ -936,7 +936,7 @@ static void smc911x_phy_configure(struct work_struct *work) if (lp->ctl_rspeed != 100) my_ad_caps &= ~(ADVERTISE_100BASE4|ADVERTISE_100FULL|ADVERTISE_100HALF); - if (!lp->ctl_rfduplx) + if (!lp->ctl_rfduplx) my_ad_caps &= ~(ADVERTISE_100FULL|ADVERTISE_10FULL); /* Update our Auto-Neg Advertisement Register */ @@ -1245,7 +1245,7 @@ static void smc911x_poll_controller(struct net_device *dev) #endif /* Our watchdog timed out. Called by the networking layer */ -static void smc911x_timeout(struct net_device *dev) +static void smc911x_timeout(struct net_device *dev, unsigned int txqueue) { struct smc911x_local *lp = netdev_priv(dev); int status, mask; diff --git a/drivers/net/ethernet/smsc/smc9194.c b/drivers/net/ethernet/smsc/smc9194.c index d3bb2ba51f40..4b2330deed47 100644 --- a/drivers/net/ethernet/smsc/smc9194.c +++ b/drivers/net/ethernet/smsc/smc9194.c @@ -216,7 +216,7 @@ static int smc_open(struct net_device *dev); /* . Our watchdog timed out. Called by the networking layer */ -static void smc_timeout(struct net_device *dev); +static void smc_timeout(struct net_device *dev, unsigned int txqueue); /* . This is called by the kernel in response to 'ifconfig ethX down'. It @@ -1094,7 +1094,7 @@ static int smc_open(struct net_device *dev) .-------------------------------------------------------- */ -static void smc_timeout(struct net_device *dev) +static void smc_timeout(struct net_device *dev, unsigned int txqueue) { /* If we get here, some higher level has decided we are broken. There should really be a "kick me" function call instead. */ diff --git a/drivers/net/ethernet/smsc/smc91c92_cs.c b/drivers/net/ethernet/smsc/smc91c92_cs.c index a55f430f6a7b..f2a50eb3c1e0 100644 --- a/drivers/net/ethernet/smsc/smc91c92_cs.c +++ b/drivers/net/ethernet/smsc/smc91c92_cs.c @@ -271,7 +271,7 @@ static void smc91c92_release(struct pcmcia_device *link); static int smc_open(struct net_device *dev); static int smc_close(struct net_device *dev); static int smc_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); -static void smc_tx_timeout(struct net_device *dev); +static void smc_tx_timeout(struct net_device *dev, unsigned int txqueue); static netdev_tx_t smc_start_xmit(struct sk_buff *skb, struct net_device *dev); static irqreturn_t smc_interrupt(int irq, void *dev_id); @@ -1178,7 +1178,7 @@ static void smc_hardware_send_packet(struct net_device * dev) /*====================================================================*/ -static void smc_tx_timeout(struct net_device *dev) +static void smc_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct smc_private *smc = netdev_priv(dev); unsigned int ioaddr = dev->base_addr; diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c index 3a6761131f4c..90410f9d3b1a 100644 --- a/drivers/net/ethernet/smsc/smc91x.c +++ b/drivers/net/ethernet/smsc/smc91x.c @@ -1321,7 +1321,7 @@ static void smc_poll_controller(struct net_device *dev) #endif /* Our watchdog timed out. Called by the networking layer */ -static void smc_timeout(struct net_device *dev) +static void smc_timeout(struct net_device *dev, unsigned int txqueue) { struct smc_local *lp = netdev_priv(dev); void __iomem *ioaddr = lp->base; diff --git a/drivers/net/ethernet/smsc/smsc911x.c b/drivers/net/ethernet/smsc/smsc911x.c index 6d90a097ce4e..49a6a9167af4 100644 --- a/drivers/net/ethernet/smsc/smsc911x.c +++ b/drivers/net/ethernet/smsc/smsc911x.c @@ -1943,15 +1943,6 @@ static int smsc911x_set_mac_address(struct net_device *dev, void *p) return 0; } -/* Standard ioctls for mii-tool */ -static int smsc911x_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - if (!netif_running(dev) || !dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, ifr, cmd); -} - static void smsc911x_ethtool_getdrvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { @@ -2151,7 +2142,7 @@ static const struct net_device_ops smsc911x_netdev_ops = { .ndo_start_xmit = smsc911x_hard_start_xmit, .ndo_get_stats = smsc911x_get_stats, .ndo_set_rx_mode = smsc911x_set_multicast_list, - .ndo_do_ioctl = smsc911x_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = smsc911x_set_mac_address, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/smsc/smsc9420.c b/drivers/net/ethernet/smsc/smsc9420.c index a6962a41c3d2..7312e522c022 100644 --- a/drivers/net/ethernet/smsc/smsc9420.c +++ b/drivers/net/ethernet/smsc/smsc9420.c @@ -210,15 +210,6 @@ static int smsc9420_eeprom_reload(struct smsc9420_pdata *pd) return -EIO; } -/* Standard ioctls for mii-tool */ -static int smsc9420_do_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - if (!netif_running(dev) || !dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, ifr, cmd); -} - static void smsc9420_ethtool_get_drvinfo(struct net_device *netdev, struct ethtool_drvinfo *drvinfo) { @@ -1504,7 +1495,7 @@ static const struct net_device_ops smsc9420_netdev_ops = { .ndo_start_xmit = smsc9420_hard_start_xmit, .ndo_get_stats = smsc9420_get_stats, .ndo_set_rx_mode = smsc9420_set_multicast_list, - .ndo_do_ioctl = smsc9420_do_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, #ifdef CONFIG_NET_POLL_CONTROLLER diff --git a/drivers/net/ethernet/socionext/netsec.c b/drivers/net/ethernet/socionext/netsec.c index 869a498e3b5e..e8224b543dfc 100644 --- a/drivers/net/ethernet/socionext/netsec.c +++ b/drivers/net/ethernet/socionext/netsec.c @@ -243,6 +243,7 @@ NET_IP_ALIGN) #define NETSEC_RX_BUF_NON_DATA (NETSEC_RXBUF_HEADROOM + \ SKB_DATA_ALIGN(sizeof(struct skb_shared_info))) +#define NETSEC_RX_BUF_SIZE (PAGE_SIZE - NETSEC_RX_BUF_NON_DATA) #define DESC_SZ sizeof(struct netsec_de) @@ -719,7 +720,6 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv, { struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; - enum dma_data_direction dma_dir; struct page *page; page = page_pool_dev_alloc_pages(dring->page_pool); @@ -734,9 +734,7 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv, /* Make sure the incoming payload fits in the page for XDP and non-XDP * cases and reserve enough space for headroom + skb_shared_info */ - *desc_len = PAGE_SIZE - NETSEC_RX_BUF_NON_DATA; - dma_dir = page_pool_get_dma_dir(dring->page_pool); - dma_sync_single_for_device(priv->dev, *dma_handle, *desc_len, dma_dir); + *desc_len = NETSEC_RX_BUF_SIZE; return page_address(page); } @@ -883,6 +881,8 @@ static u32 netsec_xdp_xmit_back(struct netsec_priv *priv, struct xdp_buff *xdp) static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, struct xdp_buff *xdp) { + struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; + unsigned int len = xdp->data_end - xdp->data; u32 ret = NETSEC_XDP_PASS; int err; u32 act; @@ -896,7 +896,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, case XDP_TX: ret = netsec_xdp_xmit_back(priv, xdp); if (ret != NETSEC_XDP_TX) - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); break; case XDP_REDIRECT: err = xdp_do_redirect(priv->ndev, xdp, prog); @@ -904,7 +906,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, ret = NETSEC_XDP_REDIR; } else { ret = NETSEC_XDP_CONSUMED; - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); } break; default: @@ -915,7 +919,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog, /* fall through -- handle aborts by dropping packet */ case XDP_DROP: ret = NETSEC_XDP_CONSUMED; - xdp_return_buff(xdp); + __page_pool_put_page(dring->page_pool, + virt_to_head_page(xdp->data), + len, true); break; } @@ -929,7 +935,6 @@ static int netsec_process_rx(struct netsec_priv *priv, int budget) struct netsec_rx_pkt_info rx_info; enum dma_data_direction dma_dir; struct bpf_prog *xdp_prog; - struct sk_buff *skb = NULL; u16 xdp_xmit = 0; u32 xdp_act = 0; int done = 0; @@ -943,7 +948,8 @@ static int netsec_process_rx(struct netsec_priv *priv, int budget) struct netsec_de *de = dring->vaddr + (DESC_SZ * idx); struct netsec_desc *desc = &dring->desc[idx]; struct page *page = virt_to_page(desc->addr); - u32 xdp_result = XDP_PASS; + u32 xdp_result = NETSEC_XDP_PASS; + struct sk_buff *skb = NULL; u16 pkt_len, desc_len; dma_addr_t dma_handle; struct xdp_buff xdp; @@ -1014,7 +1020,8 @@ static int netsec_process_rx(struct netsec_priv *priv, int budget) * cache state. Since we paid the allocation cost if * building an skb fails try to put the page into cache */ - page_pool_recycle_direct(dring->page_pool, page); + __page_pool_put_page(dring->page_pool, page, + pkt_len, true); netif_err(priv, drv, priv->ndev, "rx failed to build skb\n"); break; @@ -1272,17 +1279,19 @@ static int netsec_setup_rx_dring(struct netsec_priv *priv) { struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX]; struct bpf_prog *xdp_prog = READ_ONCE(priv->xdp_prog); - struct page_pool_params pp_params = { 0 }; + struct page_pool_params pp_params = { + .order = 0, + /* internal DMA mapping in page_pool */ + .flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV, + .pool_size = DESC_NUM, + .nid = NUMA_NO_NODE, + .dev = priv->dev, + .dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE, + .offset = NETSEC_RXBUF_HEADROOM, + .max_len = NETSEC_RX_BUF_SIZE, + }; int i, err; - pp_params.order = 0; - /* internal DMA mapping in page_pool */ - pp_params.flags = PP_FLAG_DMA_MAP; - pp_params.pool_size = DESC_NUM; - pp_params.nid = cpu_to_node(0); - pp_params.dev = priv->dev; - pp_params.dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE; - dring->page_pool = page_pool_create(&pp_params); if (IS_ERR(dring->page_pool)) { err = PTR_ERR(dring->page_pool); @@ -1731,12 +1740,6 @@ static int netsec_netdev_set_features(struct net_device *ndev, return 0; } -static int netsec_netdev_ioctl(struct net_device *ndev, struct ifreq *ifr, - int cmd) -{ - return phy_mii_ioctl(ndev->phydev, ifr, cmd); -} - static int netsec_xdp_xmit(struct net_device *ndev, int n, struct xdp_frame **frames, u32 flags) { @@ -1821,7 +1824,7 @@ static const struct net_device_ops netsec_netdev_ops = { .ndo_set_features = netsec_netdev_set_features, .ndo_set_mac_address = eth_mac_addr, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = netsec_netdev_ioctl, + .ndo_do_ioctl = phy_do_ioctl, .ndo_xdp_xmit = netsec_xdp_xmit, .ndo_bpf = netsec_xdp, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/common.h b/drivers/net/ethernet/stmicro/stmmac/common.h index 94f94686cf7d..487099092693 100644 --- a/drivers/net/ethernet/stmicro/stmmac/common.h +++ b/drivers/net/ethernet/stmicro/stmmac/common.h @@ -363,6 +363,12 @@ struct dma_features { unsigned int dvlan; unsigned int l3l4fnum; unsigned int arpoffsel; + /* TSN Features */ + unsigned int estwid; + unsigned int estdep; + unsigned int estsel; + unsigned int fpesel; + unsigned int tbssel; }; /* RX Buffer size must be multiple of 4/8/16 bytes */ diff --git a/drivers/net/ethernet/stmicro/stmmac/descs.h b/drivers/net/ethernet/stmicro/stmmac/descs.h index 9f0b9a9e63b3..49d6a866244f 100644 --- a/drivers/net/ethernet/stmicro/stmmac/descs.h +++ b/drivers/net/ethernet/stmicro/stmmac/descs.h @@ -171,6 +171,15 @@ struct dma_extended_desc { __le32 des7; /* Tx/Rx Timestamp High */ }; +/* Enhanced descriptor for TBS */ +struct dma_edesc { + __le32 des4; + __le32 des5; + __le32 des6; + __le32 des7; + struct dma_desc basic; +}; + /* Transmit checksum insertion control */ #define TX_CIC_FULL 3 /* Include IP header and pseudoheader */ diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-dwc-qos-eth.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-dwc-qos-eth.c index dd9967aeda22..2342d497348e 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-dwc-qos-eth.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-dwc-qos-eth.c @@ -40,7 +40,7 @@ struct tegra_eqos { static int dwc_eth_dwmac_config_dt(struct platform_device *pdev, struct plat_stmmacenet_data *plat_dat) { - struct device_node *np = pdev->dev.of_node; + struct device *dev = &pdev->dev; u32 burst_map = 0; u32 bit_index = 0; u32 a_index = 0; @@ -52,9 +52,10 @@ static int dwc_eth_dwmac_config_dt(struct platform_device *pdev, return -ENOMEM; } - plat_dat->axi->axi_lpi_en = of_property_read_bool(np, "snps,en-lpi"); - if (of_property_read_u32(np, "snps,write-requests", - &plat_dat->axi->axi_wr_osr_lmt)) { + plat_dat->axi->axi_lpi_en = device_property_read_bool(dev, + "snps,en-lpi"); + if (device_property_read_u32(dev, "snps,write-requests", + &plat_dat->axi->axi_wr_osr_lmt)) { /** * Since the register has a reset value of 1, if property * is missing, default to 1. @@ -68,8 +69,8 @@ static int dwc_eth_dwmac_config_dt(struct platform_device *pdev, plat_dat->axi->axi_wr_osr_lmt--; } - if (of_property_read_u32(np, "snps,read-requests", - &plat_dat->axi->axi_rd_osr_lmt)) { + if (device_property_read_u32(dev, "snps,read-requests", + &plat_dat->axi->axi_rd_osr_lmt)) { /** * Since the register has a reset value of 1, if property * is missing, default to 1. @@ -82,7 +83,7 @@ static int dwc_eth_dwmac_config_dt(struct platform_device *pdev, */ plat_dat->axi->axi_rd_osr_lmt--; } - of_property_read_u32(np, "snps,burst-map", &burst_map); + device_property_read_u32(dev, "snps,burst-map", &burst_map); /* converts burst-map bitmask to burst array */ for (bit_index = 0; bit_index < 7; bit_index++) { @@ -270,6 +271,7 @@ static void *tegra_eqos_probe(struct platform_device *pdev, struct plat_stmmacenet_data *data, struct stmmac_resources *res) { + struct device *dev = &pdev->dev; struct tegra_eqos *eqos; int err; @@ -282,6 +284,9 @@ static void *tegra_eqos_probe(struct platform_device *pdev, eqos->dev = &pdev->dev; eqos->regs = res->addr; + if (!is_of_node(dev->fwnode)) + goto bypass_clk_reset_gpio; + eqos->clk_master = devm_clk_get(&pdev->dev, "master_bus"); if (IS_ERR(eqos->clk_master)) { err = PTR_ERR(eqos->clk_master); @@ -354,6 +359,7 @@ static void *tegra_eqos_probe(struct platform_device *pdev, usleep_range(2000, 4000); +bypass_clk_reset_gpio: data->fix_mac_speed = tegra_eqos_fix_speed; data->init = tegra_eqos_init; data->bsp_priv = eqos; @@ -421,7 +427,7 @@ static int dwc_eth_dwmac_probe(struct platform_device *pdev) void *priv; int ret; - data = of_device_get_match_data(&pdev->dev); + data = device_get_match_data(&pdev->dev); memset(&stmmac_res, 0, sizeof(struct stmmac_resources)); @@ -478,7 +484,7 @@ static int dwc_eth_dwmac_remove(struct platform_device *pdev) const struct dwc_eth_dwmac_data *data; int err; - data = of_device_get_match_data(&pdev->dev); + data = device_get_match_data(&pdev->dev); err = stmmac_dvr_remove(&pdev->dev); if (err < 0) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-mediatek.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-mediatek.c index bdb80421acac..9e4b83832938 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-mediatek.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-mediatek.c @@ -55,6 +55,8 @@ struct mediatek_dwmac_plat_data { struct regmap *peri_regmap; struct device *dev; phy_interface_t phy_mode; + int num_clks_to_config; + bool rmii_clk_from_mac; bool rmii_rxc; }; @@ -73,21 +75,33 @@ struct mediatek_dwmac_variant { /* list of clocks required for mac */ static const char * const mt2712_dwmac_clk_l[] = { - "axi", "apb", "mac_main", "ptp_ref" + "axi", "apb", "mac_main", "ptp_ref", "rmii_internal" }; static int mt2712_set_interface(struct mediatek_dwmac_plat_data *plat) { + int rmii_clk_from_mac = plat->rmii_clk_from_mac ? RMII_CLK_SRC_INTERNAL : 0; int rmii_rxc = plat->rmii_rxc ? RMII_CLK_SRC_RXC : 0; u32 intf_val = 0; + /* The clock labeled as "rmii_internal" in mt2712_dwmac_clk_l is needed + * only in RMII(when MAC provides the reference clock), and useless for + * RGMII/MII/RMII(when PHY provides the reference clock). + * num_clks_to_config indicates the real number of clocks should be + * configured, equals to (plat->variant->num_clks - 1) in default for all the case, + * then +1 for rmii_clk_from_mac case. + */ + plat->num_clks_to_config = plat->variant->num_clks - 1; + /* select phy interface in top control domain */ switch (plat->phy_mode) { case PHY_INTERFACE_MODE_MII: intf_val |= PHY_INTF_MII; break; case PHY_INTERFACE_MODE_RMII: - intf_val |= (PHY_INTF_RMII | rmii_rxc); + if (plat->rmii_clk_from_mac) + plat->num_clks_to_config++; + intf_val |= (PHY_INTF_RMII | rmii_rxc | rmii_clk_from_mac); break; case PHY_INTERFACE_MODE_RGMII: case PHY_INTERFACE_MODE_RGMII_TXID: @@ -173,35 +187,50 @@ static int mt2712_set_delay(struct mediatek_dwmac_plat_data *plat) delay_val |= FIELD_PREP(ETH_DLY_RXC_INV, mac_delay->rx_inv); break; case PHY_INTERFACE_MODE_RMII: - /* the rmii reference clock is from external phy, - * and the property "rmii_rxc" indicates which pin(TXC/RXC) - * the reference clk is connected to. The reference clock is a - * received signal, so rx_delay/rx_inv are used to indicate - * the reference clock timing adjustment - */ - if (plat->rmii_rxc) { - /* the rmii reference clock from outside is connected - * to RXC pin, the reference clock will be adjusted - * by RXC delay macro circuit. - */ - delay_val |= FIELD_PREP(ETH_DLY_RXC_ENABLE, !!mac_delay->rx_delay); - delay_val |= FIELD_PREP(ETH_DLY_RXC_STAGES, mac_delay->rx_delay); - delay_val |= FIELD_PREP(ETH_DLY_RXC_INV, mac_delay->rx_inv); - } else { - /* the rmii reference clock from outside is connected - * to TXC pin, the reference clock will be adjusted - * by TXC delay macro circuit. + if (plat->rmii_clk_from_mac) { + /* case 1: mac provides the rmii reference clock, + * and the clock output to TXC pin. + * The egress timing can be adjusted by GTXC delay macro circuit. + * The ingress timing can be adjusted by TXC delay macro circuit. */ delay_val |= FIELD_PREP(ETH_DLY_TXC_ENABLE, !!mac_delay->rx_delay); delay_val |= FIELD_PREP(ETH_DLY_TXC_STAGES, mac_delay->rx_delay); delay_val |= FIELD_PREP(ETH_DLY_TXC_INV, mac_delay->rx_inv); + + delay_val |= FIELD_PREP(ETH_DLY_GTXC_ENABLE, !!mac_delay->tx_delay); + delay_val |= FIELD_PREP(ETH_DLY_GTXC_STAGES, mac_delay->tx_delay); + delay_val |= FIELD_PREP(ETH_DLY_GTXC_INV, mac_delay->tx_inv); + } else { + /* case 2: the rmii reference clock is from external phy, + * and the property "rmii_rxc" indicates which pin(TXC/RXC) + * the reference clk is connected to. The reference clock is a + * received signal, so rx_delay/rx_inv are used to indicate + * the reference clock timing adjustment + */ + if (plat->rmii_rxc) { + /* the rmii reference clock from outside is connected + * to RXC pin, the reference clock will be adjusted + * by RXC delay macro circuit. + */ + delay_val |= FIELD_PREP(ETH_DLY_RXC_ENABLE, !!mac_delay->rx_delay); + delay_val |= FIELD_PREP(ETH_DLY_RXC_STAGES, mac_delay->rx_delay); + delay_val |= FIELD_PREP(ETH_DLY_RXC_INV, mac_delay->rx_inv); + } else { + /* the rmii reference clock from outside is connected + * to TXC pin, the reference clock will be adjusted + * by TXC delay macro circuit. + */ + delay_val |= FIELD_PREP(ETH_DLY_TXC_ENABLE, !!mac_delay->rx_delay); + delay_val |= FIELD_PREP(ETH_DLY_TXC_STAGES, mac_delay->rx_delay); + delay_val |= FIELD_PREP(ETH_DLY_TXC_INV, mac_delay->rx_inv); + } + /* tx_inv will inverse the tx clock inside mac relateive to + * reference clock from external phy, + * and this bit is located in the same register with fine-tune + */ + if (mac_delay->tx_inv) + fine_val = ETH_RMII_DLY_TX_INV; } - /* tx_inv will inverse the tx clock inside mac relateive to - * reference clock from external phy, - * and this bit is located in the same register with fine-tune - */ - if (mac_delay->tx_inv) - fine_val = ETH_RMII_DLY_TX_INV; break; case PHY_INTERFACE_MODE_RGMII: case PHY_INTERFACE_MODE_RGMII_TXID: @@ -278,6 +307,7 @@ static int mediatek_dwmac_config_dt(struct mediatek_dwmac_plat_data *plat) mac_delay->tx_inv = of_property_read_bool(plat->np, "mediatek,txc-inverse"); mac_delay->rx_inv = of_property_read_bool(plat->np, "mediatek,rxc-inverse"); plat->rmii_rxc = of_property_read_bool(plat->np, "mediatek,rmii-rxc"); + plat->rmii_clk_from_mac = of_property_read_bool(plat->np, "mediatek,rmii-clk-from-mac"); return 0; } @@ -294,6 +324,8 @@ static int mediatek_dwmac_clk_init(struct mediatek_dwmac_plat_data *plat) for (i = 0; i < num; i++) plat->clks[i].id = variant->clk_list[i]; + plat->num_clks_to_config = variant->num_clks; + return devm_clk_bulk_get(plat->dev, num, plat->clks); } @@ -321,7 +353,7 @@ static int mediatek_dwmac_init(struct platform_device *pdev, void *priv) return ret; } - ret = clk_bulk_prepare_enable(variant->num_clks, plat->clks); + ret = clk_bulk_prepare_enable(plat->num_clks_to_config, plat->clks); if (ret) { dev_err(plat->dev, "failed to enable clks, err = %d\n", ret); return ret; @@ -336,9 +368,8 @@ static int mediatek_dwmac_init(struct platform_device *pdev, void *priv) static void mediatek_dwmac_exit(struct platform_device *pdev, void *priv) { struct mediatek_dwmac_plat_data *plat = priv; - const struct mediatek_dwmac_variant *variant = plat->variant; - clk_bulk_disable_unprepare(variant->num_clks, plat->clks); + clk_bulk_disable_unprepare(plat->num_clks_to_config, plat->clks); pm_runtime_put_sync(&pdev->dev); pm_runtime_disable(&pdev->dev); diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c index 01b484cb177e..58e0511badba 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-sun8i.c @@ -335,14 +335,30 @@ static void sun8i_dwmac_dump_mac_regs(struct mac_device_info *hw, } } -static void sun8i_dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan) +static void sun8i_dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan, + bool rx, bool tx) { - writel(EMAC_RX_INT | EMAC_TX_INT, ioaddr + EMAC_INT_EN); + u32 value = readl(ioaddr + EMAC_INT_EN); + + if (rx) + value |= EMAC_RX_INT; + if (tx) + value |= EMAC_TX_INT; + + writel(value, ioaddr + EMAC_INT_EN); } -static void sun8i_dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan) +static void sun8i_dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan, + bool rx, bool tx) { - writel(0, ioaddr + EMAC_INT_EN); + u32 value = readl(ioaddr + EMAC_INT_EN); + + if (rx) + value &= ~EMAC_RX_INT; + if (tx) + value &= ~EMAC_TX_INT; + + writel(value, ioaddr + EMAC_INT_EN); } static void sun8i_dwmac_dma_start_tx(void __iomem *ioaddr, u32 chan) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4.h index 2dc70d104161..af50af27550b 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4.h @@ -64,6 +64,8 @@ #define GMAC_RXQCTRL_MCBCQEN_SHIFT 20 #define GMAC_RXQCTRL_TACPQE BIT(21) #define GMAC_RXQCTRL_TACPQE_SHIFT 21 +#define GMAC_RXQCTRL_FPRQ GENMASK(26, 24) +#define GMAC_RXQCTRL_FPRQ_SHIFT 24 /* MAC Packet Filtering */ #define GMAC_PACKET_FILTER_PR BIT(0) @@ -176,6 +178,8 @@ enum power_event { #define GMAC_CONFIG_SARC GENMASK(30, 28) #define GMAC_CONFIG_SARC_SHIFT 28 #define GMAC_CONFIG_IPC BIT(27) +#define GMAC_CONFIG_IPG GENMASK(26, 24) +#define GMAC_CONFIG_IPG_SHIFT 24 #define GMAC_CONFIG_2K BIT(22) #define GMAC_CONFIG_ACS BIT(20) #define GMAC_CONFIG_BE BIT(18) @@ -183,6 +187,7 @@ enum power_event { #define GMAC_CONFIG_JE BIT(16) #define GMAC_CONFIG_PS BIT(15) #define GMAC_CONFIG_FES BIT(14) +#define GMAC_CONFIG_FES_SHIFT 14 #define GMAC_CONFIG_DM BIT(13) #define GMAC_CONFIG_LM BIT(12) #define GMAC_CONFIG_DCRS BIT(9) @@ -190,6 +195,9 @@ enum power_event { #define GMAC_CONFIG_RE BIT(0) /* MAC extended config */ +#define GMAC_CONFIG_EIPG GENMASK(29, 25) +#define GMAC_CONFIG_EIPG_SHIFT 25 +#define GMAC_CONFIG_EIPG_EN BIT(24) #define GMAC_CONFIG_HDSMS GENMASK(22, 20) #define GMAC_CONFIG_HDSMS_SHIFT 20 #define GMAC_CONFIG_HDSMS_256 (0x2 << GMAC_CONFIG_HDSMS_SHIFT) @@ -231,6 +239,11 @@ enum power_event { /* MAC HW features3 bitmap */ #define GMAC_HW_FEAT_ASP GENMASK(29, 28) +#define GMAC_HW_FEAT_TBSSEL BIT(27) +#define GMAC_HW_FEAT_FPESEL BIT(26) +#define GMAC_HW_FEAT_ESTWID GENMASK(21, 20) +#define GMAC_HW_FEAT_ESTDEP GENMASK(19, 17) +#define GMAC_HW_FEAT_ESTSEL BIT(16) #define GMAC_HW_FEAT_FRPES GENMASK(14, 13) #define GMAC_HW_FEAT_FRPBS GENMASK(12, 11) #define GMAC_HW_FEAT_FRPSEL BIT(10) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c index 40ca00e596dd..f0c0ea616032 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c @@ -984,6 +984,8 @@ const struct stmmac_ops dwmac410_ops = { .set_arp_offload = dwmac4_set_arp_offload, .config_l3_filter = dwmac4_config_l3_filter, .config_l4_filter = dwmac4_config_l4_filter, + .est_configure = dwmac5_est_configure, + .fpe_configure = dwmac5_fpe_configure, }; const struct stmmac_ops dwmac510_ops = { @@ -1027,6 +1029,8 @@ const struct stmmac_ops dwmac510_ops = { .set_arp_offload = dwmac4_set_arp_offload, .config_l3_filter = dwmac4_config_l3_filter, .config_l4_filter = dwmac4_config_l4_filter, + .est_configure = dwmac5_est_configure, + .fpe_configure = dwmac5_fpe_configure, }; int dwmac4_setup(struct stmmac_priv *priv) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c index 3e14da69f378..eff82065a501 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.c @@ -10,6 +10,7 @@ #include <linux/stmmac.h> #include "common.h" +#include "dwmac4.h" #include "dwmac4_descs.h" static int dwmac4_wrback_get_tx_status(void *data, struct stmmac_extra_stats *x, @@ -505,6 +506,14 @@ static void dwmac4_set_sec_addr(struct dma_desc *p, dma_addr_t addr) p->des3 = cpu_to_le32(upper_32_bits(addr) | RDES3_BUFFER2_VALID_ADDR); } +static void dwmac4_set_tbs(struct dma_edesc *p, u32 sec, u32 nsec) +{ + p->des4 = cpu_to_le32((sec & TDES4_LT) | TDES4_LTV); + p->des5 = cpu_to_le32(nsec & TDES5_LT); + p->des6 = 0; + p->des7 = 0; +} + const struct stmmac_desc_ops dwmac4_desc_ops = { .tx_status = dwmac4_wrback_get_tx_status, .rx_status = dwmac4_wrback_get_rx_status, @@ -534,6 +543,7 @@ const struct stmmac_desc_ops dwmac4_desc_ops = { .set_vlan = dwmac4_set_vlan, .get_rx_header_len = dwmac4_get_rx_header_len, .set_sec_addr = dwmac4_set_sec_addr, + .set_tbs = dwmac4_set_tbs, }; const struct stmmac_mode_ops dwmac4_ring_mode_ops = { diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h index 6d92109dc9aa..6da070ccd737 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_descs.h @@ -73,6 +73,13 @@ #define TDES3_CONTEXT_TYPE BIT(30) #define TDES3_CONTEXT_TYPE_SHIFT 30 +/* TDES4 */ +#define TDES4_LTV BIT(31) +#define TDES4_LT GENMASK(7, 0) + +/* TDES5 */ +#define TDES5_LT GENMASK(31, 8) + /* TDS3 use for both format (read and write back) */ #define TDES3_OWN BIT(31) #define TDES3_OWN_SHIFT 31 diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c index c15409030710..bb29bfcd62c3 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.c @@ -404,6 +404,11 @@ static void dwmac4_get_hw_feature(void __iomem *ioaddr, /* 5.10 Features */ dma_cap->asp = (hw_cap & GMAC_HW_FEAT_ASP) >> 28; + dma_cap->tbssel = (hw_cap & GMAC_HW_FEAT_TBSSEL) >> 27; + dma_cap->fpesel = (hw_cap & GMAC_HW_FEAT_FPESEL) >> 26; + dma_cap->estwid = (hw_cap & GMAC_HW_FEAT_ESTWID) >> 20; + dma_cap->estdep = (hw_cap & GMAC_HW_FEAT_ESTDEP) >> 17; + dma_cap->estsel = (hw_cap & GMAC_HW_FEAT_ESTSEL) >> 16; dma_cap->frpes = (hw_cap & GMAC_HW_FEAT_FRPES) >> 13; dma_cap->frpbs = (hw_cap & GMAC_HW_FEAT_FRPBS) >> 11; dma_cap->frpsel = (hw_cap & GMAC_HW_FEAT_FRPSEL) >> 10; @@ -467,6 +472,25 @@ static void dwmac4_enable_sph(void __iomem *ioaddr, bool en, u32 chan) writel(value, ioaddr + DMA_CHAN_CONTROL(chan)); } +static int dwmac4_enable_tbs(void __iomem *ioaddr, bool en, u32 chan) +{ + u32 value = readl(ioaddr + DMA_CHAN_TX_CONTROL(chan)); + + if (en) + value |= DMA_CONTROL_EDSE; + else + value &= ~DMA_CONTROL_EDSE; + + writel(value, ioaddr + DMA_CHAN_TX_CONTROL(chan)); + + value = readl(ioaddr + DMA_CHAN_TX_CONTROL(chan)) & DMA_CONTROL_EDSE; + if (en && !value) + return -EIO; + + writel(DMA_TBS_DEF_FTOS, ioaddr + DMA_TBS_CTRL); + return 0; +} + const struct stmmac_dma_ops dwmac4_dma_ops = { .reset = dwmac4_dma_reset, .init = dwmac4_dma_init, @@ -523,4 +547,5 @@ const struct stmmac_dma_ops dwmac410_dma_ops = { .qmode = dwmac4_qmode, .set_bfsize = dwmac4_set_bfsize, .enable_sph = dwmac4_enable_sph, + .enable_tbs = dwmac4_enable_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h index 589931795847..8391ca63d943 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_dma.h @@ -22,6 +22,7 @@ #define DMA_DEBUG_STATUS_1 0x00001010 #define DMA_DEBUG_STATUS_2 0x00001014 #define DMA_AXI_BUS_MODE 0x00001028 +#define DMA_TBS_CTRL 0x00001050 /* DMA Bus Mode bitmap */ #define DMA_BUS_MODE_SFT_RESET BIT(0) @@ -82,6 +83,11 @@ #define DMA_AXI_BURST_LEN_MASK 0x000000FE +/* DMA TBS Control */ +#define DMA_TBS_FTOS GENMASK(31, 8) +#define DMA_TBS_FTOV BIT(0) +#define DMA_TBS_DEF_FTOS (DMA_TBS_FTOS | DMA_TBS_FTOV) + /* Following DMA defines are chanels oriented */ #define DMA_CHAN_BASE_ADDR 0x00001100 #define DMA_CHAN_BASE_OFFSET 0x80 @@ -114,6 +120,7 @@ #define DMA_CONTROL_MSS_MASK GENMASK(13, 0) /* DMA Tx Channel X Control register defines */ +#define DMA_CONTROL_EDSE BIT(28) #define DMA_CONTROL_TSE BIT(12) #define DMA_CONTROL_OSP BIT(4) #define DMA_CONTROL_ST BIT(0) @@ -168,6 +175,8 @@ /* DMA default interrupt mask for 4.00 */ #define DMA_CHAN_INTR_DEFAULT_MASK (DMA_CHAN_INTR_NORMAL | \ DMA_CHAN_INTR_ABNORMAL) +#define DMA_CHAN_INTR_DEFAULT_RX (DMA_CHAN_INTR_ENA_RIE) +#define DMA_CHAN_INTR_DEFAULT_TX (DMA_CHAN_INTR_ENA_TIE) #define DMA_CHAN_INTR_NORMAL_4_10 (DMA_CHAN_INTR_ENA_NIE_4_10 | \ DMA_CHAN_INTR_ENA_RIE | \ @@ -178,6 +187,8 @@ /* DMA default interrupt mask for 4.10a */ #define DMA_CHAN_INTR_DEFAULT_MASK_4_10 (DMA_CHAN_INTR_NORMAL_4_10 | \ DMA_CHAN_INTR_ABNORMAL_4_10) +#define DMA_CHAN_INTR_DEFAULT_RX_4_10 (DMA_CHAN_INTR_ENA_RIE) +#define DMA_CHAN_INTR_DEFAULT_TX_4_10 (DMA_CHAN_INTR_ENA_TIE) /* channel 0 specific fields */ #define DMA_CHAN0_DBG_STAT_TPS GENMASK(15, 12) @@ -186,9 +197,10 @@ #define DMA_CHAN0_DBG_STAT_RPS_SHIFT 8 int dwmac4_dma_reset(void __iomem *ioaddr); -void dwmac4_enable_dma_irq(void __iomem *ioaddr, u32 chan); -void dwmac410_enable_dma_irq(void __iomem *ioaddr, u32 chan); -void dwmac4_disable_dma_irq(void __iomem *ioaddr, u32 chan); +void dwmac4_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); +void dwmac410_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); +void dwmac4_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); +void dwmac410_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); void dwmac4_dma_start_tx(void __iomem *ioaddr, u32 chan); void dwmac4_dma_stop_tx(void __iomem *ioaddr, u32 chan); void dwmac4_dma_start_rx(void __iomem *ioaddr, u32 chan); diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_lib.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_lib.c index f2a29a90e085..9becca280074 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_lib.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_lib.c @@ -97,21 +97,52 @@ void dwmac4_set_rx_ring_len(void __iomem *ioaddr, u32 len, u32 chan) writel(len, ioaddr + DMA_CHAN_RX_RING_LEN(chan)); } -void dwmac4_enable_dma_irq(void __iomem *ioaddr, u32 chan) +void dwmac4_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) { - writel(DMA_CHAN_INTR_DEFAULT_MASK, ioaddr + - DMA_CHAN_INTR_ENA(chan)); + u32 value = readl(ioaddr + DMA_CHAN_INTR_ENA(chan)); + + if (rx) + value |= DMA_CHAN_INTR_DEFAULT_RX; + if (tx) + value |= DMA_CHAN_INTR_DEFAULT_TX; + + writel(value, ioaddr + DMA_CHAN_INTR_ENA(chan)); } -void dwmac410_enable_dma_irq(void __iomem *ioaddr, u32 chan) +void dwmac410_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) { - writel(DMA_CHAN_INTR_DEFAULT_MASK_4_10, - ioaddr + DMA_CHAN_INTR_ENA(chan)); + u32 value = readl(ioaddr + DMA_CHAN_INTR_ENA(chan)); + + if (rx) + value |= DMA_CHAN_INTR_DEFAULT_RX_4_10; + if (tx) + value |= DMA_CHAN_INTR_DEFAULT_TX_4_10; + + writel(value, ioaddr + DMA_CHAN_INTR_ENA(chan)); } -void dwmac4_disable_dma_irq(void __iomem *ioaddr, u32 chan) +void dwmac4_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) { - writel(0, ioaddr + DMA_CHAN_INTR_ENA(chan)); + u32 value = readl(ioaddr + DMA_CHAN_INTR_ENA(chan)); + + if (rx) + value &= ~DMA_CHAN_INTR_DEFAULT_RX; + if (tx) + value &= ~DMA_CHAN_INTR_DEFAULT_TX; + + writel(value, ioaddr + DMA_CHAN_INTR_ENA(chan)); +} + +void dwmac410_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) +{ + u32 value = readl(ioaddr + DMA_CHAN_INTR_ENA(chan)); + + if (rx) + value &= ~DMA_CHAN_INTR_DEFAULT_RX_4_10; + if (tx) + value &= ~DMA_CHAN_INTR_DEFAULT_TX_4_10; + + writel(value, ioaddr + DMA_CHAN_INTR_ENA(chan)); } int dwmac4_dma_interrupt(void __iomem *ioaddr, diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac5.c b/drivers/net/ethernet/stmicro/stmmac/dwmac5.c index e436fa160c7d..494c859b4ade 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac5.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac5.c @@ -550,3 +550,122 @@ int dwmac5_flex_pps_config(void __iomem *ioaddr, int index, writel(val, ioaddr + MAC_PPS_CONTROL); return 0; } + +static int dwmac5_est_write(void __iomem *ioaddr, u32 reg, u32 val, bool gcl) +{ + u32 ctrl; + + writel(val, ioaddr + MTL_EST_GCL_DATA); + + ctrl = (reg << ADDR_SHIFT); + ctrl |= gcl ? 0 : GCRR; + + writel(ctrl, ioaddr + MTL_EST_GCL_CONTROL); + + ctrl |= SRWO; + writel(ctrl, ioaddr + MTL_EST_GCL_CONTROL); + + return readl_poll_timeout(ioaddr + MTL_EST_GCL_CONTROL, + ctrl, !(ctrl & SRWO), 100, 5000); +} + +int dwmac5_est_configure(void __iomem *ioaddr, struct stmmac_est *cfg, + unsigned int ptp_rate) +{ + u32 speed, total_offset, offset, ctrl, ctr_low; + u32 extcfg = readl(ioaddr + GMAC_EXT_CONFIG); + u32 mac_cfg = readl(ioaddr + GMAC_CONFIG); + int i, ret = 0x0; + u64 total_ctr; + + if (extcfg & GMAC_CONFIG_EIPG_EN) { + offset = (extcfg & GMAC_CONFIG_EIPG) >> GMAC_CONFIG_EIPG_SHIFT; + offset = 104 + (offset * 8); + } else { + offset = (mac_cfg & GMAC_CONFIG_IPG) >> GMAC_CONFIG_IPG_SHIFT; + offset = 96 - (offset * 8); + } + + speed = mac_cfg & (GMAC_CONFIG_PS | GMAC_CONFIG_FES); + speed = speed >> GMAC_CONFIG_FES_SHIFT; + + switch (speed) { + case 0x0: + offset = offset * 1000; /* 1G */ + break; + case 0x1: + offset = offset * 400; /* 2.5G */ + break; + case 0x2: + offset = offset * 100000; /* 10M */ + break; + case 0x3: + offset = offset * 10000; /* 100M */ + break; + default: + return -EINVAL; + } + + offset = offset / 1000; + + ret |= dwmac5_est_write(ioaddr, BTR_LOW, cfg->btr[0], false); + ret |= dwmac5_est_write(ioaddr, BTR_HIGH, cfg->btr[1], false); + ret |= dwmac5_est_write(ioaddr, TER, cfg->ter, false); + ret |= dwmac5_est_write(ioaddr, LLR, cfg->gcl_size, false); + if (ret) + return ret; + + total_offset = 0; + for (i = 0; i < cfg->gcl_size; i++) { + ret = dwmac5_est_write(ioaddr, i, cfg->gcl[i] + offset, true); + if (ret) + return ret; + + total_offset += offset; + } + + total_ctr = cfg->ctr[0] + cfg->ctr[1] * 1000000000; + total_ctr += total_offset; + + ctr_low = do_div(total_ctr, 1000000000); + + ret |= dwmac5_est_write(ioaddr, CTR_LOW, ctr_low, false); + ret |= dwmac5_est_write(ioaddr, CTR_HIGH, total_ctr, false); + if (ret) + return ret; + + ctrl = readl(ioaddr + MTL_EST_CONTROL); + ctrl &= ~PTOV; + ctrl |= ((1000000000 / ptp_rate) * 6) << PTOV_SHIFT; + if (cfg->enable) + ctrl |= EEST | SSWL; + else + ctrl &= ~EEST; + + writel(ctrl, ioaddr + MTL_EST_CONTROL); + return 0; +} + +void dwmac5_fpe_configure(void __iomem *ioaddr, u32 num_txq, u32 num_rxq, + bool enable) +{ + u32 value; + + if (!enable) { + value = readl(ioaddr + MAC_FPE_CTRL_STS); + + value &= ~EFPE; + + writel(value, ioaddr + MAC_FPE_CTRL_STS); + return; + } + + value = readl(ioaddr + GMAC_RXQ_CTRL1); + value &= ~GMAC_RXQCTRL_FPRQ; + value |= (num_rxq - 1) << GMAC_RXQCTRL_FPRQ_SHIFT; + writel(value, ioaddr + GMAC_RXQ_CTRL1); + + value = readl(ioaddr + MAC_FPE_CTRL_STS); + value |= EFPE; + writel(value, ioaddr + MAC_FPE_CTRL_STS); +} diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac5.h b/drivers/net/ethernet/stmicro/stmmac/dwmac5.h index 23fecf68f781..3e8faa96b4d4 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac5.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac5.h @@ -11,6 +11,9 @@ #define PRTYEN BIT(1) #define TMOUTEN BIT(0) +#define MAC_FPE_CTRL_STS 0x00000234 +#define EFPE BIT(0) + #define MAC_PPS_CONTROL 0x00000b70 #define PPS_MAXIDX(x) ((((x) + 1) * 8) - 1) #define PPS_MINIDX(x) ((x) * 8) @@ -30,6 +33,23 @@ #define MAC_PPSx_INTERVAL(x) (0x00000b88 + ((x) * 0x10)) #define MAC_PPSx_WIDTH(x) (0x00000b8c + ((x) * 0x10)) +#define MTL_EST_CONTROL 0x00000c50 +#define PTOV GENMASK(31, 24) +#define PTOV_SHIFT 24 +#define SSWL BIT(1) +#define EEST BIT(0) +#define MTL_EST_GCL_CONTROL 0x00000c80 +#define BTR_LOW 0x0 +#define BTR_HIGH 0x1 +#define CTR_LOW 0x2 +#define CTR_HIGH 0x3 +#define TER 0x4 +#define LLR 0x5 +#define ADDR_SHIFT 8 +#define GCRR BIT(2) +#define SRWO BIT(0) +#define MTL_EST_GCL_DATA 0x00000c84 + #define MTL_RXP_CONTROL_STATUS 0x00000ca0 #define RXPI BIT(31) #define NPE GENMASK(23, 16) @@ -83,5 +103,9 @@ int dwmac5_rxp_config(void __iomem *ioaddr, struct stmmac_tc_entry *entries, int dwmac5_flex_pps_config(void __iomem *ioaddr, int index, struct stmmac_pps_cfg *cfg, bool enable, u32 sub_second_inc, u32 systime_flags); +int dwmac5_est_configure(void __iomem *ioaddr, struct stmmac_est *cfg, + unsigned int ptp_rate); +void dwmac5_fpe_configure(void __iomem *ioaddr, u32 num_txq, u32 num_rxq, + bool enable); #endif /* __DWMAC5_H__ */ diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h index 292b880f3f9f..e5dbd0bc257e 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_dma.h @@ -96,6 +96,8 @@ /* DMA default interrupt mask */ #define DMA_INTR_DEFAULT_MASK (DMA_INTR_NORMAL | DMA_INTR_ABNORMAL) +#define DMA_INTR_DEFAULT_RX (DMA_INTR_ENA_RIE) +#define DMA_INTR_DEFAULT_TX (DMA_INTR_ENA_TIE) /* DMA Status register defines */ #define DMA_STATUS_GLPII 0x40000000 /* GMAC LPI interrupt */ @@ -130,8 +132,8 @@ #define NUM_DWMAC1000_DMA_REGS 23 void dwmac_enable_dma_transmission(void __iomem *ioaddr); -void dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan); -void dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan); +void dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); +void dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx); void dwmac_dma_start_tx(void __iomem *ioaddr, u32 chan); void dwmac_dma_stop_tx(void __iomem *ioaddr, u32 chan); void dwmac_dma_start_rx(void __iomem *ioaddr, u32 chan); diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c index 1bc25aa86dbd..688d36095333 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac_lib.c @@ -37,14 +37,28 @@ void dwmac_enable_dma_transmission(void __iomem *ioaddr) writel(1, ioaddr + DMA_XMT_POLL_DEMAND); } -void dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan) +void dwmac_enable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) { - writel(DMA_INTR_DEFAULT_MASK, ioaddr + DMA_INTR_ENA); + u32 value = readl(ioaddr + DMA_INTR_ENA); + + if (rx) + value |= DMA_INTR_DEFAULT_RX; + if (tx) + value |= DMA_INTR_DEFAULT_TX; + + writel(value, ioaddr + DMA_INTR_ENA); } -void dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan) +void dwmac_disable_dma_irq(void __iomem *ioaddr, u32 chan, bool rx, bool tx) { - writel(0, ioaddr + DMA_INTR_ENA); + u32 value = readl(ioaddr + DMA_INTR_ENA); + + if (rx) + value &= ~DMA_INTR_DEFAULT_RX; + if (tx) + value &= ~DMA_INTR_DEFAULT_TX; + + writel(value, ioaddr + DMA_INTR_ENA); } void dwmac_dma_start_tx(void __iomem *ioaddr, u32 chan) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h index ef8a07c68ca7..6c3b8a950f58 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2.h @@ -73,6 +73,9 @@ #define XGMAC_RXQ_CTRL0 0x000000a0 #define XGMAC_RXQEN(x) GENMASK((x) * 2 + 1, (x) * 2) #define XGMAC_RXQEN_SHIFT(x) ((x) * 2) +#define XGMAC_RXQ_CTRL1 0x000000a4 +#define XGMAC_RQ GENMASK(7, 4) +#define XGMAC_RQ_SHIFT 4 #define XGMAC_RXQ_CTRL2 0x000000a8 #define XGMAC_RXQ_CTRL3 0x000000ac #define XGMAC_PSRQ(x) GENMASK((x) * 8 + 7, (x) * 8) @@ -136,6 +139,11 @@ #define XGMAC_HWFEAT_TXQCNT GENMASK(9, 6) #define XGMAC_HWFEAT_RXQCNT GENMASK(3, 0) #define XGMAC_HW_FEATURE3 0x00000128 +#define XGMAC_HWFEAT_TBSSEL BIT(27) +#define XGMAC_HWFEAT_FPESEL BIT(26) +#define XGMAC_HWFEAT_ESTWID GENMASK(24, 23) +#define XGMAC_HWFEAT_ESTDEP GENMASK(22, 20) +#define XGMAC_HWFEAT_ESTSEL BIT(19) #define XGMAC_HWFEAT_ASP GENMASK(15, 14) #define XGMAC_HWFEAT_DVLAN BIT(13) #define XGMAC_HWFEAT_FRPES GENMASK(12, 11) @@ -148,6 +156,8 @@ #define XGMAC_MDIO_ADDR 0x00000200 #define XGMAC_MDIO_DATA 0x00000204 #define XGMAC_MDIO_C22P 0x00000220 +#define XGMAC_FPE_CTRL_STS 0x00000280 +#define XGMAC_EFPE BIT(0) #define XGMAC_ADDRx_HIGH(x) (0x00000300 + (x) * 0x8) #define XGMAC_ADDR_MAX 32 #define XGMAC_AE BIT(31) @@ -237,6 +247,22 @@ #define XGMAC_TC_PRTY_MAP1 0x00001044 #define XGMAC_PSTC(x) GENMASK((x) * 8 + 7, (x) * 8) #define XGMAC_PSTC_SHIFT(x) ((x) * 8) +#define XGMAC_MTL_EST_CONTROL 0x00001050 +#define XGMAC_PTOV GENMASK(31, 23) +#define XGMAC_PTOV_SHIFT 23 +#define XGMAC_SSWL BIT(1) +#define XGMAC_EEST BIT(0) +#define XGMAC_MTL_EST_GCL_CONTROL 0x00001080 +#define XGMAC_BTR_LOW 0x0 +#define XGMAC_BTR_HIGH 0x1 +#define XGMAC_CTR_LOW 0x2 +#define XGMAC_CTR_HIGH 0x3 +#define XGMAC_TER 0x4 +#define XGMAC_LLR 0x5 +#define XGMAC_ADDR_SHIFT 8 +#define XGMAC_GCRR BIT(2) +#define XGMAC_SRWO BIT(0) +#define XGMAC_MTL_EST_GCL_DATA 0x00001084 #define XGMAC_MTL_RXP_CONTROL_STATUS 0x000010a0 #define XGMAC_RXPI BIT(31) #define XGMAC_NPE GENMASK(23, 16) @@ -321,6 +347,13 @@ #define XGMAC_TDPS GENMASK(29, 0) #define XGMAC_RX_EDMA_CTRL 0x00003044 #define XGMAC_RDPS GENMASK(29, 0) +#define XGMAC_DMA_TBS_CTRL0 0x00003054 +#define XGMAC_DMA_TBS_CTRL1 0x00003058 +#define XGMAC_DMA_TBS_CTRL2 0x0000305c +#define XGMAC_DMA_TBS_CTRL3 0x00003060 +#define XGMAC_FTOS GENMASK(31, 8) +#define XGMAC_FTOV BIT(0) +#define XGMAC_DEF_FTOS (XGMAC_FTOS | XGMAC_FTOV) #define XGMAC_DMA_SAFETY_INT_STATUS 0x00003064 #define XGMAC_MCSIS BIT(31) #define XGMAC_MSUIS BIT(29) @@ -335,6 +368,7 @@ #define XGMAC_SPH BIT(24) #define XGMAC_PBLx8 BIT(16) #define XGMAC_DMA_CH_TX_CONTROL(x) (0x00003104 + (0x80 * (x))) +#define XGMAC_EDSE BIT(28) #define XGMAC_TxPBL GENMASK(21, 16) #define XGMAC_TxPBL_SHIFT 16 #define XGMAC_TSE BIT(12) @@ -363,6 +397,8 @@ #define XGMAC_TIE BIT(0) #define XGMAC_DMA_INT_DEFAULT_EN (XGMAC_NIE | XGMAC_AIE | XGMAC_RBUE | \ XGMAC_RIE | XGMAC_TIE) +#define XGMAC_DMA_INT_DEFAULT_RX (XGMAC_RBUE | XGMAC_RIE) +#define XGMAC_DMA_INT_DEFAULT_TX (XGMAC_TIE) #define XGMAC_DMA_CH_Rx_WATCHDOG(x) (0x0000313c + (0x80 * (x))) #define XGMAC_RWT GENMASK(7, 0) #define XGMAC_DMA_CH_STATUS(x) (0x00003160 + (0x80 * (x))) @@ -377,6 +413,9 @@ #define XGMAC_REGSIZE ((0x0000317c + (0x80 * 15)) / 4) /* Descriptors */ +#define XGMAC_TDES0_LTV BIT(31) +#define XGMAC_TDES0_LT GENMASK(7, 0) +#define XGMAC_TDES1_LT GENMASK(31, 8) #define XGMAC_TDES2_IVT GENMASK(31, 16) #define XGMAC_TDES2_IVT_SHIFT 16 #define XGMAC_TDES2_IOC BIT(31) @@ -395,6 +434,7 @@ #define XGMAC_TDES3_TCMSSV BIT(26) #define XGMAC_TDES3_SAIC GENMASK(25, 23) #define XGMAC_TDES3_SAIC_SHIFT 23 +#define XGMAC_TDES3_TBSV BIT(24) #define XGMAC_TDES3_THL GENMASK(22, 19) #define XGMAC_TDES3_THL_SHIFT 19 #define XGMAC_TDES3_IVTIR GENMASK(19, 18) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c index 082f5ee9e525..2af3ac5409b7 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_core.c @@ -1359,6 +1359,81 @@ static void dwxgmac2_set_arp_offload(struct mac_device_info *hw, bool en, writel(value, ioaddr + XGMAC_RX_CONFIG); } +static int dwxgmac3_est_write(void __iomem *ioaddr, u32 reg, u32 val, bool gcl) +{ + u32 ctrl; + + writel(val, ioaddr + XGMAC_MTL_EST_GCL_DATA); + + ctrl = (reg << XGMAC_ADDR_SHIFT); + ctrl |= gcl ? 0 : XGMAC_GCRR; + + writel(ctrl, ioaddr + XGMAC_MTL_EST_GCL_CONTROL); + + ctrl |= XGMAC_SRWO; + writel(ctrl, ioaddr + XGMAC_MTL_EST_GCL_CONTROL); + + return readl_poll_timeout_atomic(ioaddr + XGMAC_MTL_EST_GCL_CONTROL, + ctrl, !(ctrl & XGMAC_SRWO), 100, 5000); +} + +static int dwxgmac3_est_configure(void __iomem *ioaddr, struct stmmac_est *cfg, + unsigned int ptp_rate) +{ + int i, ret = 0x0; + u32 ctrl; + + ret |= dwxgmac3_est_write(ioaddr, XGMAC_BTR_LOW, cfg->btr[0], false); + ret |= dwxgmac3_est_write(ioaddr, XGMAC_BTR_HIGH, cfg->btr[1], false); + ret |= dwxgmac3_est_write(ioaddr, XGMAC_TER, cfg->ter, false); + ret |= dwxgmac3_est_write(ioaddr, XGMAC_LLR, cfg->gcl_size, false); + ret |= dwxgmac3_est_write(ioaddr, XGMAC_CTR_LOW, cfg->ctr[0], false); + ret |= dwxgmac3_est_write(ioaddr, XGMAC_CTR_HIGH, cfg->ctr[1], false); + if (ret) + return ret; + + for (i = 0; i < cfg->gcl_size; i++) { + ret = dwxgmac3_est_write(ioaddr, i, cfg->gcl[i], true); + if (ret) + return ret; + } + + ctrl = readl(ioaddr + XGMAC_MTL_EST_CONTROL); + ctrl &= ~XGMAC_PTOV; + ctrl |= ((1000000000 / ptp_rate) * 9) << XGMAC_PTOV_SHIFT; + if (cfg->enable) + ctrl |= XGMAC_EEST | XGMAC_SSWL; + else + ctrl &= ~XGMAC_EEST; + + writel(ctrl, ioaddr + XGMAC_MTL_EST_CONTROL); + return 0; +} + +static void dwxgmac3_fpe_configure(void __iomem *ioaddr, u32 num_txq, + u32 num_rxq, bool enable) +{ + u32 value; + + if (!enable) { + value = readl(ioaddr + XGMAC_FPE_CTRL_STS); + + value &= ~XGMAC_EFPE; + + writel(value, ioaddr + XGMAC_FPE_CTRL_STS); + return; + } + + value = readl(ioaddr + XGMAC_RXQ_CTRL1); + value &= ~XGMAC_RQ; + value |= (num_rxq - 1) << XGMAC_RQ_SHIFT; + writel(value, ioaddr + XGMAC_RXQ_CTRL1); + + value = readl(ioaddr + XGMAC_FPE_CTRL_STS); + value |= XGMAC_EFPE; + writel(value, ioaddr + XGMAC_FPE_CTRL_STS); +} + const struct stmmac_ops dwxgmac210_ops = { .core_init = dwxgmac2_core_init, .set_mac = dwxgmac2_set_mac, @@ -1402,6 +1477,8 @@ const struct stmmac_ops dwxgmac210_ops = { .config_l3_filter = dwxgmac2_config_l3_filter, .config_l4_filter = dwxgmac2_config_l4_filter, .set_arp_offload = dwxgmac2_set_arp_offload, + .est_configure = dwxgmac3_est_configure, + .fpe_configure = dwxgmac3_fpe_configure, }; int dwxgmac2_setup(struct stmmac_priv *priv) diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c index bd5838ce1e8a..c3d654cfa9ef 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_descs.c @@ -339,6 +339,14 @@ static void dwxgmac2_set_vlan(struct dma_desc *p, u32 type) p->des2 |= cpu_to_le32(type & XGMAC_TDES2_VTIR); } +static void dwxgmac2_set_tbs(struct dma_edesc *p, u32 sec, u32 nsec) +{ + p->des4 = cpu_to_le32((sec & XGMAC_TDES0_LT) | XGMAC_TDES0_LTV); + p->des5 = cpu_to_le32(nsec & XGMAC_TDES1_LT); + p->des6 = 0; + p->des7 = 0; +} + const struct stmmac_desc_ops dwxgmac210_desc_ops = { .tx_status = dwxgmac2_get_tx_status, .rx_status = dwxgmac2_get_rx_status, @@ -368,4 +376,5 @@ const struct stmmac_desc_ops dwxgmac210_desc_ops = { .set_sarc = dwxgmac2_set_sarc, .set_vlan_tag = dwxgmac2_set_vlan_tag, .set_vlan = dwxgmac2_set_vlan, + .set_tbs = dwxgmac2_set_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c index f3f08ccc379b..77308c5c5d29 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwxgmac2_dma.c @@ -248,14 +248,30 @@ static void dwxgmac2_dma_tx_mode(void __iomem *ioaddr, int mode, writel(value, ioaddr + XGMAC_MTL_TXQ_OPMODE(channel)); } -static void dwxgmac2_enable_dma_irq(void __iomem *ioaddr, u32 chan) +static void dwxgmac2_enable_dma_irq(void __iomem *ioaddr, u32 chan, + bool rx, bool tx) { - writel(XGMAC_DMA_INT_DEFAULT_EN, ioaddr + XGMAC_DMA_CH_INT_EN(chan)); + u32 value = readl(ioaddr + XGMAC_DMA_CH_INT_EN(chan)); + + if (rx) + value |= XGMAC_DMA_INT_DEFAULT_RX; + if (tx) + value |= XGMAC_DMA_INT_DEFAULT_TX; + + writel(value, ioaddr + XGMAC_DMA_CH_INT_EN(chan)); } -static void dwxgmac2_disable_dma_irq(void __iomem *ioaddr, u32 chan) +static void dwxgmac2_disable_dma_irq(void __iomem *ioaddr, u32 chan, + bool rx, bool tx) { - writel(0, ioaddr + XGMAC_DMA_CH_INT_EN(chan)); + u32 value = readl(ioaddr + XGMAC_DMA_CH_INT_EN(chan)); + + if (rx) + value &= ~XGMAC_DMA_INT_DEFAULT_RX; + if (tx) + value &= ~XGMAC_DMA_INT_DEFAULT_TX; + + writel(value, ioaddr + XGMAC_DMA_CH_INT_EN(chan)); } static void dwxgmac2_dma_start_tx(void __iomem *ioaddr, u32 chan) @@ -413,6 +429,11 @@ static void dwxgmac2_get_hw_feature(void __iomem *ioaddr, /* MAC HW feature 3 */ hw_cap = readl(ioaddr + XGMAC_HW_FEATURE3); + dma_cap->tbssel = (hw_cap & XGMAC_HWFEAT_TBSSEL) >> 27; + dma_cap->fpesel = (hw_cap & XGMAC_HWFEAT_FPESEL) >> 26; + dma_cap->estwid = (hw_cap & XGMAC_HWFEAT_ESTWID) >> 23; + dma_cap->estdep = (hw_cap & XGMAC_HWFEAT_ESTDEP) >> 20; + dma_cap->estsel = (hw_cap & XGMAC_HWFEAT_ESTSEL) >> 19; dma_cap->asp = (hw_cap & XGMAC_HWFEAT_ASP) >> 14; dma_cap->dvlan = (hw_cap & XGMAC_HWFEAT_DVLAN) >> 13; dma_cap->frpes = (hw_cap & XGMAC_HWFEAT_FRPES) >> 11; @@ -503,6 +524,28 @@ static void dwxgmac2_enable_sph(void __iomem *ioaddr, bool en, u32 chan) writel(value, ioaddr + XGMAC_DMA_CH_CONTROL(chan)); } +static int dwxgmac2_enable_tbs(void __iomem *ioaddr, bool en, u32 chan) +{ + u32 value = readl(ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)); + + if (en) + value |= XGMAC_EDSE; + else + value &= ~XGMAC_EDSE; + + writel(value, ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)); + + value = readl(ioaddr + XGMAC_DMA_CH_TX_CONTROL(chan)) & XGMAC_EDSE; + if (en && !value) + return -EIO; + + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL0); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL1); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL2); + writel(XGMAC_DEF_FTOS, ioaddr + XGMAC_DMA_TBS_CTRL3); + return 0; +} + const struct stmmac_dma_ops dwxgmac210_dma_ops = { .reset = dwxgmac2_dma_reset, .init = dwxgmac2_dma_init, @@ -530,4 +573,5 @@ const struct stmmac_dma_ops dwxgmac210_dma_ops = { .qmode = dwxgmac2_qmode, .set_bfsize = dwxgmac2_set_bfsize, .enable_sph = dwxgmac2_enable_sph, + .enable_tbs = dwxgmac2_enable_tbs, }; diff --git a/drivers/net/ethernet/stmicro/stmmac/hwif.h b/drivers/net/ethernet/stmicro/stmmac/hwif.h index aa5b917398fe..df63b0367aff 100644 --- a/drivers/net/ethernet/stmicro/stmmac/hwif.h +++ b/drivers/net/ethernet/stmicro/stmmac/hwif.h @@ -29,6 +29,7 @@ struct stmmac_extra_stats; struct stmmac_safety_stats; struct dma_desc; struct dma_extended_desc; +struct dma_edesc; /* Descriptors helpers */ struct stmmac_desc_ops { @@ -95,6 +96,7 @@ struct stmmac_desc_ops { void (*set_vlan_tag)(struct dma_desc *p, u16 tag, u16 inner_tag, u32 inner_type); void (*set_vlan)(struct dma_desc *p, u32 type); + void (*set_tbs)(struct dma_edesc *p, u32 sec, u32 nsec); }; #define stmmac_init_rx_desc(__priv, __args...) \ @@ -157,6 +159,8 @@ struct stmmac_desc_ops { stmmac_do_void_callback(__priv, desc, set_vlan_tag, __args) #define stmmac_set_desc_vlan(__priv, __args...) \ stmmac_do_void_callback(__priv, desc, set_vlan, __args) +#define stmmac_set_desc_tbs(__priv, __args...) \ + stmmac_do_void_callback(__priv, desc, set_tbs, __args) struct stmmac_dma_cfg; struct dma_features; @@ -187,8 +191,10 @@ struct stmmac_dma_ops { void (*dma_diagnostic_fr) (void *data, struct stmmac_extra_stats *x, void __iomem *ioaddr); void (*enable_dma_transmission) (void __iomem *ioaddr); - void (*enable_dma_irq)(void __iomem *ioaddr, u32 chan); - void (*disable_dma_irq)(void __iomem *ioaddr, u32 chan); + void (*enable_dma_irq)(void __iomem *ioaddr, u32 chan, + bool rx, bool tx); + void (*disable_dma_irq)(void __iomem *ioaddr, u32 chan, + bool rx, bool tx); void (*start_tx)(void __iomem *ioaddr, u32 chan); void (*stop_tx)(void __iomem *ioaddr, u32 chan); void (*start_rx)(void __iomem *ioaddr, u32 chan); @@ -208,6 +214,7 @@ struct stmmac_dma_ops { void (*qmode)(void __iomem *ioaddr, u32 channel, u8 qmode); void (*set_bfsize)(void __iomem *ioaddr, int bfsize, u32 chan); void (*enable_sph)(void __iomem *ioaddr, bool en, u32 chan); + int (*enable_tbs)(void __iomem *ioaddr, bool en, u32 chan); }; #define stmmac_reset(__priv, __args...) \ @@ -266,6 +273,8 @@ struct stmmac_dma_ops { stmmac_do_void_callback(__priv, dma, set_bfsize, __args) #define stmmac_enable_sph(__priv, __args...) \ stmmac_do_void_callback(__priv, dma, enable_sph, __args) +#define stmmac_enable_tbs(__priv, __args...) \ + stmmac_do_callback(__priv, dma, enable_tbs, __args) struct mac_device_info; struct net_device; @@ -274,6 +283,7 @@ struct stmmac_safety_stats; struct stmmac_tc_entry; struct stmmac_pps_cfg; struct stmmac_rss; +struct stmmac_est; /* Helpers to program the MAC core */ struct stmmac_ops { @@ -371,6 +381,10 @@ struct stmmac_ops { bool en, bool udp, bool sa, bool inv, u32 match); void (*set_arp_offload)(struct mac_device_info *hw, bool en, u32 addr); + int (*est_configure)(void __iomem *ioaddr, struct stmmac_est *cfg, + unsigned int ptp_rate); + void (*fpe_configure)(void __iomem *ioaddr, u32 num_txq, u32 num_rxq, + bool enable); }; #define stmmac_core_init(__priv, __args...) \ @@ -457,6 +471,10 @@ struct stmmac_ops { stmmac_do_callback(__priv, mac, config_l4_filter, __args) #define stmmac_set_arp_offload(__priv, __args...) \ stmmac_do_void_callback(__priv, mac, set_arp_offload, __args) +#define stmmac_est_configure(__priv, __args...) \ + stmmac_do_callback(__priv, mac, est_configure, __args) +#define stmmac_fpe_configure(__priv, __args...) \ + stmmac_do_void_callback(__priv, mac, fpe_configure, __args) /* PTP and HW Timer helpers */ struct stmmac_hwtimestamp { @@ -514,6 +532,8 @@ struct stmmac_priv; struct tc_cls_u32_offload; struct tc_cbs_qopt_offload; struct flow_cls_offload; +struct tc_taprio_qopt_offload; +struct tc_etf_qopt_offload; struct stmmac_tc_ops { int (*init)(struct stmmac_priv *priv); @@ -523,6 +543,10 @@ struct stmmac_tc_ops { struct tc_cbs_qopt_offload *qopt); int (*setup_cls)(struct stmmac_priv *priv, struct flow_cls_offload *cls); + int (*setup_taprio)(struct stmmac_priv *priv, + struct tc_taprio_qopt_offload *qopt); + int (*setup_etf)(struct stmmac_priv *priv, + struct tc_etf_qopt_offload *qopt); }; #define stmmac_tc_init(__priv, __args...) \ @@ -533,6 +557,10 @@ struct stmmac_tc_ops { stmmac_do_callback(__priv, tc, setup_cbs, __args) #define stmmac_tc_setup_cls(__priv, __args...) \ stmmac_do_callback(__priv, tc, setup_cls, __args) +#define stmmac_tc_setup_taprio(__priv, __args...) \ + stmmac_do_callback(__priv, tc, setup_taprio, __args) +#define stmmac_tc_setup_etf(__priv, __args...) \ + stmmac_do_callback(__priv, tc, setup_etf, __args) struct stmmac_counters; diff --git a/drivers/net/ethernet/stmicro/stmmac/mmc_core.c b/drivers/net/ethernet/stmicro/stmmac/mmc_core.c index 252cf48c5816..a57b0fa815ab 100644 --- a/drivers/net/ethernet/stmicro/stmmac/mmc_core.c +++ b/drivers/net/ethernet/stmicro/stmmac/mmc_core.c @@ -119,6 +119,13 @@ #define MMC_RX_ICMP_GD_OCTETS 0x180 #define MMC_RX_ICMP_ERR_OCTETS 0x184 +#define MMC_TX_FPE_FRAG 0x1a8 +#define MMC_TX_HOLD_REQ 0x1ac +#define MMC_RX_PKT_ASSEMBLY_ERR 0x1c8 +#define MMC_RX_PKT_SMD_ERR 0x1cc +#define MMC_RX_PKT_ASSEMBLY_OK 0x1d0 +#define MMC_RX_FPE_FRAG 0x1d4 + /* XGMAC MMC Registers */ #define MMC_XGMAC_TX_OCTET_GB 0x14 #define MMC_XGMAC_TX_PKT_GB 0x1c @@ -315,6 +322,15 @@ static void dwmac_mmc_read(void __iomem *mmcaddr, struct stmmac_counters *mmc) mmc->mmc_rx_tcp_err_octets += readl(mmcaddr + MMC_RX_TCP_ERR_OCTETS); mmc->mmc_rx_icmp_gd_octets += readl(mmcaddr + MMC_RX_ICMP_GD_OCTETS); mmc->mmc_rx_icmp_err_octets += readl(mmcaddr + MMC_RX_ICMP_ERR_OCTETS); + + mmc->mmc_tx_fpe_fragment_cntr += readl(mmcaddr + MMC_TX_FPE_FRAG); + mmc->mmc_tx_hold_req_cntr += readl(mmcaddr + MMC_TX_HOLD_REQ); + mmc->mmc_rx_packet_assembly_err_cntr += + readl(mmcaddr + MMC_RX_PKT_ASSEMBLY_ERR); + mmc->mmc_rx_packet_smd_err_cntr += readl(mmcaddr + MMC_RX_PKT_SMD_ERR); + mmc->mmc_rx_packet_assembly_ok_cntr += + readl(mmcaddr + MMC_RX_PKT_ASSEMBLY_OK); + mmc->mmc_rx_fpe_fragment_cntr += readl(mmcaddr + MMC_RX_FPE_FRAG); } const struct stmmac_mmc_ops dwmac_mmc_ops = { diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac.h b/drivers/net/ethernet/stmicro/stmmac/stmmac.h index d993fc7e82c3..9c02fc754bf1 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac.h +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac.h @@ -39,13 +39,18 @@ struct stmmac_tx_info { bool is_jumbo; }; +#define STMMAC_TBS_AVAIL BIT(0) +#define STMMAC_TBS_EN BIT(1) + /* Frequently used values are kept adjacent for cache effect */ struct stmmac_tx_queue { u32 tx_count_frames; + int tbs; struct timer_list txtimer; u32 queue_index; struct stmmac_priv *priv_data; struct dma_extended_desc *dma_etx ____cacheline_aligned_in_smp; + struct dma_edesc *dma_entx; struct dma_desc *dma_tx; struct sk_buff **tx_skbuff; struct stmmac_tx_info *tx_skbuff_dma; @@ -88,6 +93,7 @@ struct stmmac_channel { struct napi_struct rx_napi ____cacheline_aligned_in_smp; struct napi_struct tx_napi ____cacheline_aligned_in_smp; struct stmmac_priv *priv_data; + spinlock_t lock; u32 index; }; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index 80d59b775907..ff1cbfc834b0 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -388,9 +388,8 @@ bool stmmac_eee_init(struct stmmac_priv *priv) /* Using PCS we cannot dial with the phy registers at this stage * so we do not support extra feature like EEE. */ - if ((priv->hw->pcs == STMMAC_PCS_RGMII) || - (priv->hw->pcs == STMMAC_PCS_TBI) || - (priv->hw->pcs == STMMAC_PCS_RTBI)) + if (priv->hw->pcs == STMMAC_PCS_TBI || + priv->hw->pcs == STMMAC_PCS_RTBI) return false; /* Check if MAC core supports the EEE feature. */ @@ -1090,6 +1089,8 @@ static void stmmac_display_tx_rings(struct stmmac_priv *priv) if (priv->extend_desc) head_tx = (void *)tx_q->dma_etx; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + head_tx = (void *)tx_q->dma_entx; else head_tx = (void *)tx_q->dma_tx; @@ -1163,13 +1164,19 @@ static void stmmac_clear_tx_descriptors(struct stmmac_priv *priv, u32 queue) int i; /* Clear the TX descriptors */ - for (i = 0; i < DMA_TX_SIZE; i++) + for (i = 0; i < DMA_TX_SIZE; i++) { + int last = (i == (DMA_TX_SIZE - 1)); + struct dma_desc *p; + if (priv->extend_desc) - stmmac_init_tx_desc(priv, &tx_q->dma_etx[i].basic, - priv->mode, (i == DMA_TX_SIZE - 1)); + p = &tx_q->dma_etx[i].basic; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[i].basic; else - stmmac_init_tx_desc(priv, &tx_q->dma_tx[i], - priv->mode, (i == DMA_TX_SIZE - 1)); + p = &tx_q->dma_tx[i]; + + stmmac_init_tx_desc(priv, p, priv->mode, last); + } } /** @@ -1383,7 +1390,7 @@ static int init_dma_tx_desc_rings(struct net_device *dev) if (priv->extend_desc) stmmac_mode_init(priv, tx_q->dma_etx, tx_q->dma_tx_phy, DMA_TX_SIZE, 1); - else + else if (!(tx_q->tbs & STMMAC_TBS_AVAIL)) stmmac_mode_init(priv, tx_q->dma_tx, tx_q->dma_tx_phy, DMA_TX_SIZE, 0); } @@ -1392,6 +1399,8 @@ static int init_dma_tx_desc_rings(struct net_device *dev) struct dma_desc *p; if (priv->extend_desc) p = &((tx_q->dma_etx + i)->basic); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &((tx_q->dma_entx + i)->basic); else p = tx_q->dma_tx + i; @@ -1511,19 +1520,26 @@ static void free_dma_tx_desc_resources(struct stmmac_priv *priv) /* Free TX queue resources */ for (queue = 0; queue < tx_count; queue++) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[queue]; + size_t size; + void *addr; /* Release the DMA TX socket buffers */ dma_free_tx_skbufs(priv, queue); - /* Free DMA regions of consistent memory previously allocated */ - if (!priv->extend_desc) - dma_free_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_desc), - tx_q->dma_tx, tx_q->dma_tx_phy); - else - dma_free_coherent(priv->device, DMA_TX_SIZE * - sizeof(struct dma_extended_desc), - tx_q->dma_etx, tx_q->dma_tx_phy); + if (priv->extend_desc) { + size = sizeof(struct dma_extended_desc); + addr = tx_q->dma_etx; + } else if (tx_q->tbs & STMMAC_TBS_AVAIL) { + size = sizeof(struct dma_edesc); + addr = tx_q->dma_entx; + } else { + size = sizeof(struct dma_desc); + addr = tx_q->dma_tx; + } + + size *= DMA_TX_SIZE; + + dma_free_coherent(priv->device, size, addr, tx_q->dma_tx_phy); kfree(tx_q->tx_skbuff_dma); kfree(tx_q->tx_skbuff); @@ -1616,6 +1632,8 @@ static int alloc_dma_tx_desc_resources(struct stmmac_priv *priv) /* TX queues buffers and DMA */ for (queue = 0; queue < tx_count; queue++) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[queue]; + size_t size; + void *addr; tx_q->queue_index = queue; tx_q->priv_data = priv; @@ -1632,28 +1650,32 @@ static int alloc_dma_tx_desc_resources(struct stmmac_priv *priv) if (!tx_q->tx_skbuff) goto err_dma; - if (priv->extend_desc) { - tx_q->dma_etx = dma_alloc_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_extended_desc), - &tx_q->dma_tx_phy, - GFP_KERNEL); - if (!tx_q->dma_etx) - goto err_dma; - } else { - tx_q->dma_tx = dma_alloc_coherent(priv->device, - DMA_TX_SIZE * sizeof(struct dma_desc), - &tx_q->dma_tx_phy, - GFP_KERNEL); - if (!tx_q->dma_tx) - goto err_dma; - } + if (priv->extend_desc) + size = sizeof(struct dma_extended_desc); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + size = sizeof(struct dma_edesc); + else + size = sizeof(struct dma_desc); + + size *= DMA_TX_SIZE; + + addr = dma_alloc_coherent(priv->device, size, + &tx_q->dma_tx_phy, GFP_KERNEL); + if (!addr) + goto err_dma; + + if (priv->extend_desc) + tx_q->dma_etx = addr; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + tx_q->dma_entx = addr; + else + tx_q->dma_tx = addr; } return 0; err_dma: free_dma_tx_desc_resources(priv); - return ret; } @@ -1885,6 +1907,8 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) if (priv->extend_desc) p = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[entry].basic; else p = tx_q->dma_tx + entry; @@ -1966,7 +1990,7 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) /* We still have pending packets, let's call for a new scheduling */ if (tx_q->dirty_tx != tx_q->cur_tx) - mod_timer(&tx_q->txtimer, STMMAC_COAL_TIMER(10)); + mod_timer(&tx_q->txtimer, STMMAC_COAL_TIMER(priv->tx_coal_timer)); __netif_tx_unlock_bh(netdev_get_tx_queue(priv->dev, queue)); @@ -1983,19 +2007,12 @@ static int stmmac_tx_clean(struct stmmac_priv *priv, int budget, u32 queue) static void stmmac_tx_err(struct stmmac_priv *priv, u32 chan) { struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; - int i; netif_tx_stop_queue(netdev_get_tx_queue(priv->dev, chan)); stmmac_stop_tx_dma(priv, chan); dma_free_tx_skbufs(priv, chan); - for (i = 0; i < DMA_TX_SIZE; i++) - if (priv->extend_desc) - stmmac_init_tx_desc(priv, &tx_q->dma_etx[i].basic, - priv->mode, (i == DMA_TX_SIZE - 1)); - else - stmmac_init_tx_desc(priv, &tx_q->dma_tx[i], - priv->mode, (i == DMA_TX_SIZE - 1)); + stmmac_clear_tx_descriptors(priv, chan); tx_q->dirty_tx = 0; tx_q->cur_tx = 0; tx_q->mss = 0; @@ -2060,17 +2077,25 @@ static int stmmac_napi_check(struct stmmac_priv *priv, u32 chan) int status = stmmac_dma_interrupt_status(priv, priv->ioaddr, &priv->xstats, chan); struct stmmac_channel *ch = &priv->channel[chan]; + unsigned long flags; if ((status & handle_rx) && (chan < priv->plat->rx_queues_to_use)) { if (napi_schedule_prep(&ch->rx_napi)) { - stmmac_disable_dma_irq(priv, priv->ioaddr, chan); + spin_lock_irqsave(&ch->lock, flags); + stmmac_disable_dma_irq(priv, priv->ioaddr, chan, 1, 0); + spin_unlock_irqrestore(&ch->lock, flags); __napi_schedule_irqoff(&ch->rx_napi); - status |= handle_tx; } } - if ((status & handle_tx) && (chan < priv->plat->tx_queues_to_use)) - napi_schedule_irqoff(&ch->tx_napi); + if ((status & handle_tx) && (chan < priv->plat->tx_queues_to_use)) { + if (napi_schedule_prep(&ch->tx_napi)) { + spin_lock_irqsave(&ch->lock, flags); + stmmac_disable_dma_irq(priv, priv->ioaddr, chan, 0, 1); + spin_unlock_irqrestore(&ch->lock, flags); + __napi_schedule_irqoff(&ch->tx_napi); + } + } return status; } @@ -2265,14 +2290,14 @@ static void stmmac_tx_timer(struct timer_list *t) ch = &priv->channel[tx_q->queue_index]; - /* - * If NAPI is already running we can miss some events. Let's rearm - * the timer and try again. - */ - if (likely(napi_schedule_prep(&ch->tx_napi))) + if (likely(napi_schedule_prep(&ch->tx_napi))) { + unsigned long flags; + + spin_lock_irqsave(&ch->lock, flags); + stmmac_disable_dma_irq(priv, priv->ioaddr, ch->index, 0, 1); + spin_unlock_irqrestore(&ch->lock, flags); __napi_schedule(&ch->tx_napi); - else - mod_timer(&tx_q->txtimer, STMMAC_COAL_TIMER(10)); + } } /** @@ -2624,6 +2649,14 @@ static int stmmac_hw_setup(struct net_device *dev, bool init_ptp) if (priv->dma_cap.vlins) stmmac_enable_vlan(priv, priv->hw, STMMAC_VLAN_INSERT); + /* TBS */ + for (chan = 0; chan < tx_cnt; chan++) { + struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; + int enable = tx_q->tbs & STMMAC_TBS_AVAIL; + + stmmac_enable_tbs(priv, priv->ioaddr, enable, chan); + } + /* Start the ball rolling... */ stmmac_start_all_dma(priv); @@ -2653,8 +2686,7 @@ static int stmmac_open(struct net_device *dev) u32 chan; int ret; - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) { ret = stmmac_init_phy(dev); if (ret) { @@ -2681,6 +2713,16 @@ static int stmmac_open(struct net_device *dev) priv->rx_copybreak = STMMAC_RX_COPYBREAK; + /* Earlier check for TBS */ + for (chan = 0; chan < priv->plat->tx_queues_to_use; chan++) { + struct stmmac_tx_queue *tx_q = &priv->tx_queue[chan]; + int tbs_en = priv->plat->tx_queues_cfg[chan].tbs_en; + + tx_q->tbs |= tbs_en ? STMMAC_TBS_AVAIL : 0; + if (stmmac_enable_tbs(priv, priv->ioaddr, tbs_en, chan)) + tx_q->tbs &= ~STMMAC_TBS_AVAIL; + } + ret = alloc_dma_desc_resources(priv); if (ret < 0) { netdev_err(priv->dev, "%s: DMA descriptors allocation failed\n", @@ -2829,7 +2871,11 @@ static bool stmmac_vlan_insert(struct stmmac_priv *priv, struct sk_buff *skb, tag = skb_vlan_tag_get(skb); - p = tx_q->dma_tx + tx_q->cur_tx; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + p = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + p = &tx_q->dma_tx[tx_q->cur_tx]; + if (stmmac_set_desc_vlan_tag(priv, p, tag, inner_tag, inner_type)) return false; @@ -2864,7 +2910,11 @@ static void stmmac_tso_allocator(struct stmmac_priv *priv, dma_addr_t des, tx_q->cur_tx = STMMAC_GET_ENTRY(tx_q->cur_tx, DMA_TX_SIZE); WARN_ON(tx_q->tx_skbuff[tx_q->cur_tx]); - desc = tx_q->dma_tx + tx_q->cur_tx; + + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + desc = &tx_q->dma_tx[tx_q->cur_tx]; curr_addr = des + (total_len - tmp_len); if (priv->dma_cap.addr64 <= 32) @@ -2915,13 +2965,13 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) { struct dma_desc *desc, *first, *mss_desc = NULL; struct stmmac_priv *priv = netdev_priv(dev); + int desc_size, tmp_pay_len = 0, first_tx; int nfrags = skb_shinfo(skb)->nr_frags; u32 queue = skb_get_queue_mapping(skb); unsigned int first_entry, tx_packets; - int tmp_pay_len = 0, first_tx; struct stmmac_tx_queue *tx_q; - u8 proto_hdr_len, hdr; bool has_vlan, set_ic; + u8 proto_hdr_len, hdr; u32 pay_len, mss; dma_addr_t des; int i; @@ -2958,7 +3008,11 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) /* set new MSS value if needed */ if (mss != tx_q->mss) { - mss_desc = tx_q->dma_tx + tx_q->cur_tx; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + mss_desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + mss_desc = &tx_q->dma_tx[tx_q->cur_tx]; + stmmac_set_mss(priv, mss_desc, mss); tx_q->mss = mss; tx_q->cur_tx = STMMAC_GET_ENTRY(tx_q->cur_tx, DMA_TX_SIZE); @@ -2978,7 +3032,10 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) first_entry = tx_q->cur_tx; WARN_ON(tx_q->tx_skbuff[first_entry]); - desc = tx_q->dma_tx + first_entry; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[first_entry].basic; + else + desc = &tx_q->dma_tx[first_entry]; first = desc; if (has_vlan) @@ -3050,7 +3107,11 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) set_ic = false; if (set_ic) { - desc = &tx_q->dma_tx[tx_q->cur_tx]; + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[tx_q->cur_tx].basic; + else + desc = &tx_q->dma_tx[tx_q->cur_tx]; + tx_q->tx_count_frames = 0; stmmac_set_tx_ic(priv, desc); priv->xstats.tx_set_ic_bit++; @@ -3113,16 +3174,18 @@ static netdev_tx_t stmmac_tso_xmit(struct sk_buff *skb, struct net_device *dev) pr_info("%s: curr=%d dirty=%d f=%d, e=%d, f_p=%p, nfrags %d\n", __func__, tx_q->cur_tx, tx_q->dirty_tx, first_entry, tx_q->cur_tx, first, nfrags); - - stmmac_display_ring(priv, (void *)tx_q->dma_tx, DMA_TX_SIZE, 0); - pr_info(">>> frame to be transmitted: "); print_pkt(skb->data, skb_headlen(skb)); } netdev_tx_sent_queue(netdev_get_tx_queue(dev, queue), skb->len); - tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc)); + if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc_size = sizeof(struct dma_edesc); + else + desc_size = sizeof(struct dma_desc); + + tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * desc_size); stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue); stmmac_tx_timer_arm(priv, queue); @@ -3152,10 +3215,11 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) u32 queue = skb_get_queue_mapping(skb); int nfrags = skb_shinfo(skb)->nr_frags; int gso = skb_shinfo(skb)->gso_type; + struct dma_edesc *tbs_desc = NULL; + int entry, desc_size, first_tx; struct dma_desc *desc, *first; struct stmmac_tx_queue *tx_q; bool has_vlan, set_ic; - int entry, first_tx; dma_addr_t des; tx_q = &priv->tx_queue[queue]; @@ -3195,6 +3259,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (likely(priv->extend_desc)) desc = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = tx_q->dma_tx + entry; @@ -3224,6 +3290,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (likely(priv->extend_desc)) desc = (struct dma_desc *)(tx_q->dma_etx + entry); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = tx_q->dma_tx + entry; @@ -3270,6 +3338,8 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) if (set_ic) { if (likely(priv->extend_desc)) desc = &tx_q->dma_etx[entry].basic; + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc = &tx_q->dma_entx[entry].basic; else desc = &tx_q->dma_tx[entry]; @@ -3287,20 +3357,11 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) tx_q->cur_tx = entry; if (netif_msg_pktdata(priv)) { - void *tx_head; - netdev_dbg(priv->dev, "%s: curr=%d dirty=%d f=%d, e=%d, first=%p, nfrags=%d", __func__, tx_q->cur_tx, tx_q->dirty_tx, first_entry, entry, first, nfrags); - if (priv->extend_desc) - tx_head = (void *)tx_q->dma_etx; - else - tx_head = (void *)tx_q->dma_tx; - - stmmac_display_ring(priv, tx_head, DMA_TX_SIZE, false); - netdev_dbg(priv->dev, ">>> frame to be transmitted: "); print_pkt(skb->data, skb->len); } @@ -3346,12 +3407,19 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) /* Prepare the first descriptor setting the OWN bit too */ stmmac_prepare_tx_desc(priv, first, 1, nopaged_len, - csum_insertion, priv->mode, 1, last_segment, + csum_insertion, priv->mode, 0, last_segment, skb->len); - } else { - stmmac_set_tx_owner(priv, first); } + if (tx_q->tbs & STMMAC_TBS_EN) { + struct timespec64 ts = ns_to_timespec64(skb->tstamp); + + tbs_desc = &tx_q->dma_entx[first_entry]; + stmmac_set_desc_tbs(priv, tbs_desc, ts.tv_sec, ts.tv_nsec); + } + + stmmac_set_tx_owner(priv, first); + /* The own bit must be the latest setting done when prepare the * descriptor and then barrier is needed to make sure that * all is coherent before granting the DMA engine. @@ -3362,7 +3430,14 @@ static netdev_tx_t stmmac_xmit(struct sk_buff *skb, struct net_device *dev) stmmac_enable_dma_transmission(priv, priv->ioaddr); - tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * sizeof(*desc)); + if (likely(priv->extend_desc)) + desc_size = sizeof(struct dma_extended_desc); + else if (tx_q->tbs & STMMAC_TBS_AVAIL) + desc_size = sizeof(struct dma_edesc); + else + desc_size = sizeof(struct dma_desc); + + tx_q->tx_tail_addr = tx_q->dma_tx_phy + (tx_q->cur_tx * desc_size); stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, queue); stmmac_tx_timer_arm(priv, queue); @@ -3751,8 +3826,14 @@ static int stmmac_napi_poll_rx(struct napi_struct *napi, int budget) priv->xstats.napi_poll++; work_done = stmmac_rx(priv, budget, chan); - if (work_done < budget && napi_complete_done(napi, work_done)) - stmmac_enable_dma_irq(priv, priv->ioaddr, chan); + if (work_done < budget && napi_complete_done(napi, work_done)) { + unsigned long flags; + + spin_lock_irqsave(&ch->lock, flags); + stmmac_enable_dma_irq(priv, priv->ioaddr, chan, 1, 0); + spin_unlock_irqrestore(&ch->lock, flags); + } + return work_done; } @@ -3761,7 +3842,6 @@ static int stmmac_napi_poll_tx(struct napi_struct *napi, int budget) struct stmmac_channel *ch = container_of(napi, struct stmmac_channel, tx_napi); struct stmmac_priv *priv = ch->priv_data; - struct stmmac_tx_queue *tx_q; u32 chan = ch->index; int work_done; @@ -3770,15 +3850,12 @@ static int stmmac_napi_poll_tx(struct napi_struct *napi, int budget) work_done = stmmac_tx_clean(priv, DMA_TX_SIZE, chan); work_done = min(work_done, budget); - if (work_done < budget) - napi_complete_done(napi, work_done); + if (work_done < budget && napi_complete_done(napi, work_done)) { + unsigned long flags; - /* Force transmission restart */ - tx_q = &priv->tx_queue[chan]; - if (tx_q->cur_tx != tx_q->dirty_tx) { - stmmac_enable_dma_transmission(priv, priv->ioaddr); - stmmac_set_tx_tail_ptr(priv, priv->ioaddr, tx_q->tx_tail_addr, - chan); + spin_lock_irqsave(&ch->lock, flags); + stmmac_enable_dma_irq(priv, priv->ioaddr, chan, 0, 1); + spin_unlock_irqrestore(&ch->lock, flags); } return work_done; @@ -3792,7 +3869,7 @@ static int stmmac_napi_poll_tx(struct napi_struct *napi, int budget) * netdev structure and arrange for the device to be reset to a sane state * in order to transmit a new packet. */ -static void stmmac_tx_timeout(struct net_device *dev) +static void stmmac_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct stmmac_priv *priv = netdev_priv(dev); @@ -4078,6 +4155,10 @@ static int stmmac_setup_tc(struct net_device *ndev, enum tc_setup_type type, priv, priv, true); case TC_SETUP_QDISC_CBS: return stmmac_tc_setup_cbs(priv, priv, type_data); + case TC_SETUP_QDISC_TAPRIO: + return stmmac_tc_setup_taprio(priv, priv, type_data); + case TC_SETUP_QDISC_ETF: + return stmmac_tc_setup_etf(priv, priv, type_data); default: return -EOPNOTSUPP; } @@ -4181,7 +4262,7 @@ static int stmmac_rings_status_show(struct seq_file *seq, void *v) seq_printf(seq, "Extended descriptor ring:\n"); sysfs_display_ring((void *)tx_q->dma_etx, DMA_TX_SIZE, 1, seq); - } else { + } else if (!(tx_q->tbs & STMMAC_TBS_AVAIL)) { seq_printf(seq, "Descriptor ring:\n"); sysfs_display_ring((void *)tx_q->dma_tx, DMA_TX_SIZE, 0, seq); @@ -4250,9 +4331,44 @@ static int stmmac_dma_cap_show(struct seq_file *seq, void *v) priv->dma_cap.number_rx_channel); seq_printf(seq, "\tNumber of Additional TX channel: %d\n", priv->dma_cap.number_tx_channel); + seq_printf(seq, "\tNumber of Additional RX queues: %d\n", + priv->dma_cap.number_rx_queues); + seq_printf(seq, "\tNumber of Additional TX queues: %d\n", + priv->dma_cap.number_tx_queues); seq_printf(seq, "\tEnhanced descriptors: %s\n", (priv->dma_cap.enh_desc) ? "Y" : "N"); - + seq_printf(seq, "\tTX Fifo Size: %d\n", priv->dma_cap.tx_fifo_size); + seq_printf(seq, "\tRX Fifo Size: %d\n", priv->dma_cap.rx_fifo_size); + seq_printf(seq, "\tHash Table Size: %d\n", priv->dma_cap.hash_tb_sz); + seq_printf(seq, "\tTSO: %s\n", priv->dma_cap.tsoen ? "Y" : "N"); + seq_printf(seq, "\tNumber of PPS Outputs: %d\n", + priv->dma_cap.pps_out_num); + seq_printf(seq, "\tSafety Features: %s\n", + priv->dma_cap.asp ? "Y" : "N"); + seq_printf(seq, "\tFlexible RX Parser: %s\n", + priv->dma_cap.frpsel ? "Y" : "N"); + seq_printf(seq, "\tEnhanced Addressing: %d\n", + priv->dma_cap.addr64); + seq_printf(seq, "\tReceive Side Scaling: %s\n", + priv->dma_cap.rssen ? "Y" : "N"); + seq_printf(seq, "\tVLAN Hash Filtering: %s\n", + priv->dma_cap.vlhash ? "Y" : "N"); + seq_printf(seq, "\tSplit Header: %s\n", + priv->dma_cap.sphen ? "Y" : "N"); + seq_printf(seq, "\tVLAN TX Insertion: %s\n", + priv->dma_cap.vlins ? "Y" : "N"); + seq_printf(seq, "\tDouble VLAN: %s\n", + priv->dma_cap.dvlan ? "Y" : "N"); + seq_printf(seq, "\tNumber of L3/L4 Filters: %d\n", + priv->dma_cap.l3l4fnum); + seq_printf(seq, "\tARP Offloading: %s\n", + priv->dma_cap.arpoffsel ? "Y" : "N"); + seq_printf(seq, "\tEnhancements to Scheduled Traffic (EST): %s\n", + priv->dma_cap.estsel ? "Y" : "N"); + seq_printf(seq, "\tFrame Preemption (FPE): %s\n", + priv->dma_cap.fpesel ? "Y" : "N"); + seq_printf(seq, "\tTime-Based Scheduling (TBS): %s\n", + priv->dma_cap.tbssel ? "Y" : "N"); return 0; } DEFINE_SHOW_ATTRIBUTE(stmmac_dma_cap); @@ -4728,6 +4844,7 @@ int stmmac_dvr_probe(struct device *device, for (queue = 0; queue < maxq; queue++) { struct stmmac_channel *ch = &priv->channel[queue]; + spin_lock_init(&ch->lock); ch->priv_data = priv; ch->index = queue; @@ -4757,8 +4874,7 @@ int stmmac_dvr_probe(struct device *device, stmmac_check_pcs_mode(priv); - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) { /* MDIO bus Registration */ ret = stmmac_mdio_register(ndev); @@ -4792,8 +4908,7 @@ int stmmac_dvr_probe(struct device *device, error_netdev_register: phylink_destroy(priv->phylink); error_phy_setup: - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) stmmac_mdio_unregister(ndev); error_mdio_register: @@ -4838,8 +4953,7 @@ int stmmac_dvr_remove(struct device *dev) reset_control_assert(priv->plat->stmmac_rst); clk_disable_unprepare(priv->plat->pclk); clk_disable_unprepare(priv->plat->stmmac_clk); - if (priv->hw->pcs != STMMAC_PCS_RGMII && - priv->hw->pcs != STMMAC_PCS_TBI && + if (priv->hw->pcs != STMMAC_PCS_TBI && priv->hw->pcs != STMMAC_PCS_RTBI) stmmac_mdio_unregister(ndev); destroy_workqueue(priv->wq); diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c index 8237dbc3e991..623521052152 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c @@ -65,7 +65,6 @@ static void common_default_data(struct plat_stmmacenet_data *plat) plat->force_sf_dma_mode = 1; plat->mdio_bus_data->needs_reset = true; - plat->mdio_bus_data->phy_mask = 0; /* Set default value for multicast hash bins */ plat->multicast_filter_bins = HASH_TABLE_SIZE; @@ -154,8 +153,6 @@ static int intel_mgbe_common_data(struct pci_dev *pdev, plat->tx_queues_cfg[6].weight = 0x0F; plat->tx_queues_cfg[7].weight = 0x10; - plat->mdio_bus_data->phy_mask = 0; - plat->dma_cfg->pbl = 32; plat->dma_cfg->pblx8 = true; plat->dma_cfg->fixed_burst = 0; @@ -386,8 +383,6 @@ static int snps_gmac5_default_data(struct pci_dev *pdev, plat->tso_en = 1; plat->pmt = 1; - plat->mdio_bus_data->phy_mask = 0; - /* Set default value for multicast hash bins */ plat->multicast_filter_bins = HASH_TABLE_SIZE; @@ -406,6 +401,8 @@ static int snps_gmac5_default_data(struct pci_dev *pdev, plat->tx_queues_cfg[i].use_prio = false; plat->tx_queues_cfg[i].mode_to_use = MTL_QUEUE_DCB; plat->tx_queues_cfg[i].weight = 25; + if (i > 0) + plat->tx_queues_cfg[i].tbs_en = 1; } plat->rx_sched_algorithm = MTL_RX_ALGORITHM_SP; diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c index 450d7dac3ea6..2aba2673d6c3 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_selftests.c @@ -14,6 +14,7 @@ #include <linux/phy.h> #include <linux/udp.h> #include <net/pkt_cls.h> +#include <net/pkt_sched.h> #include <net/tcp.h> #include <net/udp.h> #include <net/tc_act/tc_gact.h> @@ -50,6 +51,7 @@ struct stmmac_packet_attrs { u8 id; int sarc; u16 queue_mapping; + u64 timestamp; }; static u8 stmmac_test_next_id; @@ -208,6 +210,9 @@ static struct sk_buff *stmmac_test_get_udp_skb(struct stmmac_priv *priv, skb->pkt_type = PACKET_HOST; skb->dev = priv->dev; + if (attr->timestamp) + skb->tstamp = ns_to_ktime(attr->timestamp); + return skb; } @@ -341,8 +346,7 @@ static int __stmmac_test_loopback(struct stmmac_priv *priv, goto cleanup; } - skb_set_queue_mapping(skb, attr->queue_mapping); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, attr->queue_mapping); if (ret) goto cleanup; @@ -932,8 +936,7 @@ static int __stmmac_test_vlanfilt(struct stmmac_priv *priv) goto vlan_del; } - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -1027,8 +1030,7 @@ static int __stmmac_test_dvlanfilt(struct stmmac_priv *priv) goto vlan_del; } - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -1298,8 +1300,7 @@ static int stmmac_test_vlanoff_common(struct stmmac_priv *priv, bool svlan) __vlan_hwaccel_put_tag(skb, htons(proto), tpriv->vlan_id); skb->protocol = htons(proto); - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto vlan_del; @@ -1659,8 +1660,7 @@ static int stmmac_test_arpoffload(struct stmmac_priv *priv) if (ret) goto cleanup; - skb_set_queue_mapping(skb, 0); - ret = dev_queue_xmit(skb); + ret = dev_direct_xmit(skb, 0); if (ret) goto cleanup_promisc; @@ -1748,6 +1748,68 @@ static int stmmac_test_sph(struct stmmac_priv *priv) return 0; } +static int stmmac_test_tbs(struct stmmac_priv *priv) +{ +#define STMMAC_TBS_LT_OFFSET (500 * 1000 * 1000) /* 500 ms*/ + struct stmmac_packet_attrs attr = { }; + struct tc_etf_qopt_offload qopt; + u64 start_time, curr_time = 0; + unsigned long flags; + int ret, i; + + if (!priv->hwts_tx_en) + return -EOPNOTSUPP; + + /* Find first TBS enabled Queue, if any */ + for (i = 0; i < priv->plat->tx_queues_to_use; i++) + if (priv->tx_queue[i].tbs & STMMAC_TBS_AVAIL) + break; + + if (i >= priv->plat->tx_queues_to_use) + return -EOPNOTSUPP; + + qopt.enable = true; + qopt.queue = i; + + ret = stmmac_tc_setup_etf(priv, priv, &qopt); + if (ret) + return ret; + + spin_lock_irqsave(&priv->ptp_lock, flags); + stmmac_get_systime(priv, priv->ptpaddr, &curr_time); + spin_unlock_irqrestore(&priv->ptp_lock, flags); + + if (!curr_time) { + ret = -EOPNOTSUPP; + goto fail_disable; + } + + start_time = curr_time; + curr_time += STMMAC_TBS_LT_OFFSET; + + attr.dst = priv->dev->dev_addr; + attr.timestamp = curr_time; + attr.timeout = nsecs_to_jiffies(2 * STMMAC_TBS_LT_OFFSET); + attr.queue_mapping = i; + + ret = __stmmac_test_loopback(priv, &attr); + if (ret) + goto fail_disable; + + /* Check if expected time has elapsed */ + spin_lock_irqsave(&priv->ptp_lock, flags); + stmmac_get_systime(priv, priv->ptpaddr, &curr_time); + spin_unlock_irqrestore(&priv->ptp_lock, flags); + + if ((curr_time - start_time) < STMMAC_TBS_LT_OFFSET) + ret = -EINVAL; + +fail_disable: + qopt.enable = false; + stmmac_tc_setup_etf(priv, priv, &qopt); + return ret; +} + #define STMMAC_LOOPBACK_NONE 0 #define STMMAC_LOOPBACK_MAC 1 #define STMMAC_LOOPBACK_PHY 2 @@ -1881,6 +1943,10 @@ static const struct stmmac_test { .name = "Split Header ", .lb = STMMAC_LOOPBACK_PHY, .fn = stmmac_test_sph, + }, { + .name = "TBS (ETF Scheduler) ", + .lb = STMMAC_LOOPBACK_PHY, + .fn = stmmac_test_tbs, }, }; @@ -1889,7 +1955,6 @@ void stmmac_selftest_run(struct net_device *dev, { struct stmmac_priv *priv = netdev_priv(dev); int count = stmmac_selftest_get_count(priv); - int carrier = netif_carrier_ok(dev); int i, ret; memset(buf, 0, sizeof(*buf) * count); @@ -1899,15 +1964,12 @@ void stmmac_selftest_run(struct net_device *dev, netdev_err(priv->dev, "Only offline tests are supported\n"); etest->flags |= ETH_TEST_FL_FAILED; return; - } else if (!carrier) { + } else if (!netif_carrier_ok(dev)) { netdev_err(priv->dev, "You need valid Link to execute tests\n"); etest->flags |= ETH_TEST_FL_FAILED; return; } - /* We don't want extra traffic */ - netif_carrier_off(dev); - /* Wait for queues drain */ msleep(200); @@ -1962,10 +2024,6 @@ void stmmac_selftest_run(struct net_device *dev, break; } } - - /* Restart everything */ - if (carrier) - netif_carrier_on(dev); } void stmmac_selftest_get_strings(struct stmmac_priv *priv, u8 *data) diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c index 9ffae12a2122..7a01dee2f9a8 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_tc.c @@ -595,9 +595,167 @@ static int tc_setup_cls(struct stmmac_priv *priv, return ret; } +static int tc_setup_taprio(struct stmmac_priv *priv, + struct tc_taprio_qopt_offload *qopt) +{ + u32 size, wid = priv->dma_cap.estwid, dep = priv->dma_cap.estdep; + struct plat_stmmacenet_data *plat = priv->plat; + struct timespec64 time; + bool fpe = false; + int i, ret = 0; + u64 ctr; + + if (!priv->dma_cap.estsel) + return -EOPNOTSUPP; + + switch (wid) { + case 0x1: + wid = 16; + break; + case 0x2: + wid = 20; + break; + case 0x3: + wid = 24; + break; + default: + return -EOPNOTSUPP; + } + + switch (dep) { + case 0x1: + dep = 64; + break; + case 0x2: + dep = 128; + break; + case 0x3: + dep = 256; + break; + case 0x4: + dep = 512; + break; + case 0x5: + dep = 1024; + break; + default: + return -EOPNOTSUPP; + } + + if (!qopt->enable) + goto disable; + if (qopt->num_entries >= dep) + return -EINVAL; + if (!qopt->base_time) + return -ERANGE; + if (!qopt->cycle_time) + return -ERANGE; + + if (!plat->est) { + plat->est = devm_kzalloc(priv->device, sizeof(*plat->est), + GFP_KERNEL); + if (!plat->est) + return -ENOMEM; + } else { + memset(plat->est, 0, sizeof(*plat->est)); + } + + size = qopt->num_entries; + + priv->plat->est->gcl_size = size; + priv->plat->est->enable = qopt->enable; + + for (i = 0; i < size; i++) { + s64 delta_ns = qopt->entries[i].interval; + u32 gates = qopt->entries[i].gate_mask; + + if (delta_ns > GENMASK(wid, 0)) + return -ERANGE; + if (gates > GENMASK(31 - wid, 0)) + return -ERANGE; + + switch (qopt->entries[i].command) { + case TC_TAPRIO_CMD_SET_GATES: + if (fpe) + return -EINVAL; + break; + case TC_TAPRIO_CMD_SET_AND_HOLD: + gates |= BIT(0); + fpe = true; + break; + case TC_TAPRIO_CMD_SET_AND_RELEASE: + gates &= ~BIT(0); + fpe = true; + break; + default: + return -EOPNOTSUPP; + } + + priv->plat->est->gcl[i] = delta_ns | (gates << wid); + } + + /* Adjust for real system time */ + time = ktime_to_timespec64(qopt->base_time); + priv->plat->est->btr[0] = (u32)time.tv_nsec; + priv->plat->est->btr[1] = (u32)time.tv_sec; + + ctr = qopt->cycle_time; + priv->plat->est->ctr[0] = do_div(ctr, NSEC_PER_SEC); + priv->plat->est->ctr[1] = (u32)ctr; + + if (fpe && !priv->dma_cap.fpesel) + return -EOPNOTSUPP; + + ret = stmmac_fpe_configure(priv, priv->ioaddr, + priv->plat->tx_queues_to_use, + priv->plat->rx_queues_to_use, fpe); + if (ret && fpe) { + netdev_err(priv->dev, "failed to enable Frame Preemption\n"); + return ret; + } + + ret = stmmac_est_configure(priv, priv->ioaddr, priv->plat->est, + priv->plat->clk_ptp_rate); + if (ret) { + netdev_err(priv->dev, "failed to configure EST\n"); + goto disable; + } + + netdev_info(priv->dev, "configured EST\n"); + return 0; + +disable: + priv->plat->est->enable = false; + stmmac_est_configure(priv, priv->ioaddr, priv->plat->est, + priv->plat->clk_ptp_rate); + return ret; +} + +static int tc_setup_etf(struct stmmac_priv *priv, + struct tc_etf_qopt_offload *qopt) +{ + if (!priv->dma_cap.tbssel) + return -EOPNOTSUPP; + if (qopt->queue >= priv->plat->tx_queues_to_use) + return -EINVAL; + if (!(priv->tx_queue[qopt->queue].tbs & STMMAC_TBS_AVAIL)) + return -EINVAL; + + if (qopt->enable) + priv->tx_queue[qopt->queue].tbs |= STMMAC_TBS_EN; + else + priv->tx_queue[qopt->queue].tbs &= ~STMMAC_TBS_EN; + + netdev_info(priv->dev, "%s ETF for Queue %d\n", + qopt->enable ? "enabled" : "disabled", qopt->queue); + return 0; +} + const struct stmmac_tc_ops dwmac510_tc_ops = { .init = tc_init, .setup_cls_u32 = tc_setup_cls_u32, .setup_cbs = tc_setup_cbs, .setup_cls = tc_setup_cls, + .setup_taprio = tc_setup_taprio, + .setup_etf = tc_setup_etf, }; diff --git a/drivers/net/ethernet/sun/cassini.c b/drivers/net/ethernet/sun/cassini.c index c91876f8c536..6ec9163e232c 100644 --- a/drivers/net/ethernet/sun/cassini.c +++ b/drivers/net/ethernet/sun/cassini.c @@ -2666,7 +2666,7 @@ static void cas_netpoll(struct net_device *dev) } #endif -static void cas_tx_timeout(struct net_device *dev) +static void cas_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct cas *cp = netdev_priv(dev); diff --git a/drivers/net/ethernet/sun/niu.c b/drivers/net/ethernet/sun/niu.c index f5fd1f3c07cc..9a5004f674c7 100644 --- a/drivers/net/ethernet/sun/niu.c +++ b/drivers/net/ethernet/sun/niu.c @@ -6517,7 +6517,7 @@ static void niu_reset_task(struct work_struct *work) spin_unlock_irqrestore(&np->lock, flags); } -static void niu_tx_timeout(struct net_device *dev) +static void niu_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct niu *np = netdev_priv(dev); diff --git a/drivers/net/ethernet/sun/sunbmac.c b/drivers/net/ethernet/sun/sunbmac.c index e9b757b03b56..c5add0b45eed 100644 --- a/drivers/net/ethernet/sun/sunbmac.c +++ b/drivers/net/ethernet/sun/sunbmac.c @@ -941,7 +941,7 @@ static int bigmac_close(struct net_device *dev) return 0; } -static void bigmac_tx_timeout(struct net_device *dev) +static void bigmac_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct bigmac *bp = netdev_priv(dev); diff --git a/drivers/net/ethernet/sun/sungem.c b/drivers/net/ethernet/sun/sungem.c index 3e7631160384..8358064fbd48 100644 --- a/drivers/net/ethernet/sun/sungem.c +++ b/drivers/net/ethernet/sun/sungem.c @@ -970,7 +970,7 @@ static void gem_poll_controller(struct net_device *dev) } #endif -static void gem_tx_timeout(struct net_device *dev) +static void gem_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct gem *gp = netdev_priv(dev); diff --git a/drivers/net/ethernet/sun/sunhme.c b/drivers/net/ethernet/sun/sunhme.c index d007dfeba5c3..f0fe7bb2a750 100644 --- a/drivers/net/ethernet/sun/sunhme.c +++ b/drivers/net/ethernet/sun/sunhme.c @@ -2246,7 +2246,7 @@ static int happy_meal_close(struct net_device *dev) #define SXD(x) #endif -static void happy_meal_tx_timeout(struct net_device *dev) +static void happy_meal_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct happy_meal *hp = netdev_priv(dev); diff --git a/drivers/net/ethernet/sun/sunqe.c b/drivers/net/ethernet/sun/sunqe.c index 1468fa0a54e9..2102b95ec347 100644 --- a/drivers/net/ethernet/sun/sunqe.c +++ b/drivers/net/ethernet/sun/sunqe.c @@ -544,7 +544,7 @@ static void qe_tx_reclaim(struct sunqe *qep) qep->tx_old = elem; } -static void qe_tx_timeout(struct net_device *dev) +static void qe_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct sunqe *qep = netdev_priv(dev); int tx_full; diff --git a/drivers/net/ethernet/sun/sunvnet_common.c b/drivers/net/ethernet/sun/sunvnet_common.c index 8b94d9ad9e2b..c23ce838ff63 100644 --- a/drivers/net/ethernet/sun/sunvnet_common.c +++ b/drivers/net/ethernet/sun/sunvnet_common.c @@ -1223,7 +1223,7 @@ vnet_handle_offloads(struct vnet_port *port, struct sk_buff *skb, { struct net_device *dev = VNET_PORT_TO_NET_DEVICE(port); struct vio_dring_state *dr = &port->vio.drings[VIO_DRIVER_TX_RING]; - struct sk_buff *segs; + struct sk_buff *segs, *curr, *next; int maclen, datalen; int status; int gso_size, gso_type, gso_segs; @@ -1282,11 +1282,8 @@ vnet_handle_offloads(struct vnet_port *port, struct sk_buff *skb, skb_reset_mac_header(skb); status = 0; - while (segs) { - struct sk_buff *curr = segs; - - segs = segs->next; - curr->next = NULL; + skb_list_walk_safe(segs, curr, next) { + skb_mark_not_on_list(curr); if (port->tso && curr->len > dev->mtu) { skb_shinfo(curr)->gso_size = gso_size; skb_shinfo(curr)->gso_type = gso_type; @@ -1539,7 +1536,7 @@ out_dropped: } EXPORT_SYMBOL_GPL(sunvnet_start_xmit_common); -void sunvnet_tx_timeout_common(struct net_device *dev) +void sunvnet_tx_timeout_common(struct net_device *dev, unsigned int txqueue) { /* XXX Implement me XXX */ } diff --git a/drivers/net/ethernet/sun/sunvnet_common.h b/drivers/net/ethernet/sun/sunvnet_common.h index 2b808d2482d6..5416a3cb9e7d 100644 --- a/drivers/net/ethernet/sun/sunvnet_common.h +++ b/drivers/net/ethernet/sun/sunvnet_common.h @@ -135,7 +135,7 @@ int sunvnet_open_common(struct net_device *dev); int sunvnet_close_common(struct net_device *dev); void sunvnet_set_rx_mode_common(struct net_device *dev, struct vnet *vp); int sunvnet_set_mac_addr_common(struct net_device *dev, void *p); -void sunvnet_tx_timeout_common(struct net_device *dev); +void sunvnet_tx_timeout_common(struct net_device *dev, unsigned int txqueue); netdev_tx_t sunvnet_start_xmit_common(struct sk_buff *skb, struct net_device *dev, struct vnet_port *(*vnet_tx_port) diff --git a/drivers/net/ethernet/synopsys/dwc-xlgmac-net.c b/drivers/net/ethernet/synopsys/dwc-xlgmac-net.c index a1f5a1e61040..07046a2370b3 100644 --- a/drivers/net/ethernet/synopsys/dwc-xlgmac-net.c +++ b/drivers/net/ethernet/synopsys/dwc-xlgmac-net.c @@ -689,7 +689,7 @@ static int xlgmac_close(struct net_device *netdev) return 0; } -static void xlgmac_tx_timeout(struct net_device *netdev) +static void xlgmac_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct xlgmac_pdata *pdata = netdev_priv(netdev); diff --git a/drivers/net/ethernet/ti/cpmac.c b/drivers/net/ethernet/ti/cpmac.c index 3a655a4dc10e..a530afe3ce12 100644 --- a/drivers/net/ethernet/ti/cpmac.c +++ b/drivers/net/ethernet/ti/cpmac.c @@ -797,7 +797,7 @@ static irqreturn_t cpmac_irq(int irq, void *dev_id) return IRQ_HANDLED; } -static void cpmac_tx_timeout(struct net_device *dev) +static void cpmac_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct cpmac_priv *priv = netdev_priv(dev); @@ -816,16 +816,6 @@ static void cpmac_tx_timeout(struct net_device *dev) netif_tx_wake_all_queues(priv->dev); } -static int cpmac_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) -{ - if (!(netif_running(dev))) - return -EINVAL; - if (!dev->phydev) - return -EINVAL; - - return phy_mii_ioctl(dev->phydev, ifr, cmd); -} - static void cpmac_get_ringparam(struct net_device *dev, struct ethtool_ringparam *ring) { @@ -1054,7 +1044,7 @@ static const struct net_device_ops cpmac_netdev_ops = { .ndo_start_xmit = cpmac_start_xmit, .ndo_tx_timeout = cpmac_tx_timeout, .ndo_set_rx_mode = cpmac_set_multicast_list, - .ndo_do_ioctl = cpmac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, }; diff --git a/drivers/net/ethernet/ti/cpsw_priv.c b/drivers/net/ethernet/ti/cpsw_priv.c index 707d5eb480ce..97a058ca60ac 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.c +++ b/drivers/net/ethernet/ti/cpsw_priv.c @@ -272,7 +272,7 @@ void soft_reset(const char *module, void __iomem *reg) WARN(readl_relaxed(reg) & 1, "failed to soft-reset %s\n", module); } -void cpsw_ndo_tx_timeout(struct net_device *ndev) +void cpsw_ndo_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct cpsw_priv *priv = netdev_priv(ndev); struct cpsw_common *cpsw = priv->cpsw; diff --git a/drivers/net/ethernet/ti/cpsw_priv.h b/drivers/net/ethernet/ti/cpsw_priv.h index bc726356a72c..b8d7b924ee3d 100644 --- a/drivers/net/ethernet/ti/cpsw_priv.h +++ b/drivers/net/ethernet/ti/cpsw_priv.h @@ -449,7 +449,7 @@ int cpsw_rx_poll(struct napi_struct *napi_rx, int budget); void cpsw_rx_vlan_encap(struct sk_buff *skb); void soft_reset(const char *module, void __iomem *reg); void cpsw_set_slave_mac(struct cpsw_slave *slave, struct cpsw_priv *priv); -void cpsw_ndo_tx_timeout(struct net_device *ndev); +void cpsw_ndo_tx_timeout(struct net_device *ndev, unsigned int txqueue); int cpsw_need_resplit(struct cpsw_common *cpsw); int cpsw_ndo_ioctl(struct net_device *dev, struct ifreq *req, int cmd); int cpsw_ndo_set_tx_maxrate(struct net_device *ndev, int queue, u32 rate); diff --git a/drivers/net/ethernet/ti/davinci_emac.c b/drivers/net/ethernet/ti/davinci_emac.c index ae27be85e363..75d4e16c692b 100644 --- a/drivers/net/ethernet/ti/davinci_emac.c +++ b/drivers/net/ethernet/ti/davinci_emac.c @@ -983,7 +983,7 @@ fail_tx: * error and re-initialize the TX channel for hardware operation * */ -static void emac_dev_tx_timeout(struct net_device *ndev) +static void emac_dev_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct emac_priv *priv = netdev_priv(ndev); struct device *emac_dev = &ndev->dev; diff --git a/drivers/net/ethernet/ti/netcp_core.c b/drivers/net/ethernet/ti/netcp_core.c index 675f31de59dd..d7a144b4a09f 100644 --- a/drivers/net/ethernet/ti/netcp_core.c +++ b/drivers/net/ethernet/ti/netcp_core.c @@ -1811,7 +1811,7 @@ out: return (ret == 0) ? 0 : err; } -static void netcp_ndo_tx_timeout(struct net_device *ndev) +static void netcp_ndo_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct netcp_intf *netcp = netdev_priv(ndev); unsigned int descs = knav_pool_count(netcp->tx_pool); diff --git a/drivers/net/ethernet/ti/netcp_ethss.c b/drivers/net/ethernet/ti/netcp_ethss.c index d6a192c1f337..fb36115e9c51 100644 --- a/drivers/net/ethernet/ti/netcp_ethss.c +++ b/drivers/net/ethernet/ti/netcp_ethss.c @@ -2533,8 +2533,6 @@ static int gbe_del_vid(void *intf_priv, int vid) } #if IS_ENABLED(CONFIG_TI_CPTS) -#define HAS_PHY_TXTSTAMP(p) ((p)->drv && (p)->drv->txtstamp) -#define HAS_PHY_RXTSTAMP(p) ((p)->drv && (p)->drv->rxtstamp) static void gbe_txtstamp(void *context, struct sk_buff *skb) { @@ -2566,7 +2564,7 @@ static int gbe_txtstamp_mark_pkt(struct gbe_intf *gbe_intf, * We mark it here because skb_tx_timestamp() is called * after all the txhooks are called. */ - if (phydev && HAS_PHY_TXTSTAMP(phydev)) { + if (phy_has_txtstamp(phydev)) { skb_shinfo(p_info->skb)->tx_flags |= SKBTX_IN_PROGRESS; return 0; } @@ -2588,7 +2586,7 @@ static int gbe_rxtstamp(struct gbe_intf *gbe_intf, struct netcp_packet *p_info) if (p_info->rxtstamp_complete) return 0; - if (phydev && HAS_PHY_RXTSTAMP(phydev)) { + if (phy_has_rxtstamp(phydev)) { p_info->rxtstamp_complete = true; return 0; } @@ -2830,7 +2828,7 @@ static int gbe_ioctl(void *intf_priv, struct ifreq *req, int cmd) struct gbe_intf *gbe_intf = intf_priv; struct phy_device *phy = gbe_intf->slave->phy; - if (!phy || !phy->drv->hwtstamp) { + if (!phy_has_hwtstamp(phy)) { switch (cmd) { case SIOCGHWTSTAMP: return gbe_hwtstamp_get(gbe_intf, req); diff --git a/drivers/net/ethernet/ti/tlan.c b/drivers/net/ethernet/ti/tlan.c index 78f0f2d59e22..ad465202980a 100644 --- a/drivers/net/ethernet/ti/tlan.c +++ b/drivers/net/ethernet/ti/tlan.c @@ -161,7 +161,7 @@ static void tlan_set_multicast_list(struct net_device *); static int tlan_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); static int tlan_probe1(struct pci_dev *pdev, long ioaddr, int irq, int rev, const struct pci_device_id *ent); -static void tlan_tx_timeout(struct net_device *dev); +static void tlan_tx_timeout(struct net_device *dev, unsigned int txqueue); static void tlan_tx_timeout_work(struct work_struct *work); static int tlan_init_one(struct pci_dev *pdev, const struct pci_device_id *ent); @@ -997,7 +997,7 @@ static int tlan_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) * **************************************************************/ -static void tlan_tx_timeout(struct net_device *dev) +static void tlan_tx_timeout(struct net_device *dev, unsigned int txqueue) { TLAN_DBG(TLAN_DEBUG_GNRL, "%s: Transmit timed out.\n", dev->name); @@ -1028,7 +1028,7 @@ static void tlan_tx_timeout_work(struct work_struct *work) struct tlan_priv *priv = container_of(work, struct tlan_priv, tlan_tqueue); - tlan_tx_timeout(priv->dev); + tlan_tx_timeout(priv->dev, UINT_MAX); } diff --git a/drivers/net/ethernet/toshiba/ps3_gelic_net.c b/drivers/net/ethernet/toshiba/ps3_gelic_net.c index 9d9f8acb7ee3..070dd6fa9401 100644 --- a/drivers/net/ethernet/toshiba/ps3_gelic_net.c +++ b/drivers/net/ethernet/toshiba/ps3_gelic_net.c @@ -1405,7 +1405,7 @@ out: * * called, if tx hangs. Schedules a task that resets the interface */ -void gelic_net_tx_timeout(struct net_device *netdev) +void gelic_net_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct gelic_card *card; diff --git a/drivers/net/ethernet/toshiba/ps3_gelic_net.h b/drivers/net/ethernet/toshiba/ps3_gelic_net.h index 051033580f0a..805903dbddcc 100644 --- a/drivers/net/ethernet/toshiba/ps3_gelic_net.h +++ b/drivers/net/ethernet/toshiba/ps3_gelic_net.h @@ -359,7 +359,7 @@ int gelic_net_open(struct net_device *netdev); int gelic_net_stop(struct net_device *netdev); netdev_tx_t gelic_net_xmit(struct sk_buff *skb, struct net_device *netdev); void gelic_net_set_multi(struct net_device *netdev); -void gelic_net_tx_timeout(struct net_device *netdev); +void gelic_net_tx_timeout(struct net_device *netdev, unsigned int txqueue); int gelic_net_setup_netdev(struct net_device *netdev, struct gelic_card *card); /* shared ethtool ops */ diff --git a/drivers/net/ethernet/toshiba/spider_net.c b/drivers/net/ethernet/toshiba/spider_net.c index 538e70810d3d..6576271642c1 100644 --- a/drivers/net/ethernet/toshiba/spider_net.c +++ b/drivers/net/ethernet/toshiba/spider_net.c @@ -2180,7 +2180,7 @@ out: * called, if tx hangs. Schedules a task that resets the interface */ static void -spider_net_tx_timeout(struct net_device *netdev) +spider_net_tx_timeout(struct net_device *netdev, unsigned int txqueue) { struct spider_net_card *card; diff --git a/drivers/net/ethernet/toshiba/tc35815.c b/drivers/net/ethernet/toshiba/tc35815.c index 12466a72cefc..3fd43d30b20d 100644 --- a/drivers/net/ethernet/toshiba/tc35815.c +++ b/drivers/net/ethernet/toshiba/tc35815.c @@ -483,8 +483,7 @@ static void tc35815_txdone(struct net_device *dev); static int tc35815_close(struct net_device *dev); static struct net_device_stats *tc35815_get_stats(struct net_device *dev); static void tc35815_set_multicast_list(struct net_device *dev); -static void tc35815_tx_timeout(struct net_device *dev); -static int tc35815_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); +static void tc35815_tx_timeout(struct net_device *dev, unsigned int txqueue); #ifdef CONFIG_NET_POLL_CONTROLLER static void tc35815_poll_controller(struct net_device *dev); #endif @@ -751,7 +750,7 @@ static const struct net_device_ops tc35815_netdev_ops = { .ndo_get_stats = tc35815_get_stats, .ndo_set_rx_mode = tc35815_set_multicast_list, .ndo_tx_timeout = tc35815_tx_timeout, - .ndo_do_ioctl = tc35815_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, #ifdef CONFIG_NET_POLL_CONTROLLER @@ -1189,7 +1188,7 @@ static void tc35815_schedule_restart(struct net_device *dev) spin_unlock_irqrestore(&lp->lock, flags); } -static void tc35815_tx_timeout(struct net_device *dev) +static void tc35815_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct tc35815_regs __iomem *tr = (struct tc35815_regs __iomem *)dev->base_addr; @@ -2009,15 +2008,6 @@ static const struct ethtool_ops tc35815_ethtool_ops = { .set_link_ksettings = phy_ethtool_set_link_ksettings, }; -static int tc35815_ioctl(struct net_device *dev, struct ifreq *rq, int cmd) -{ - if (!netif_running(dev)) - return -EINVAL; - if (!dev->phydev) - return -ENODEV; - return phy_mii_ioctl(dev->phydev, rq, cmd); -} - static void tc35815_chip_reset(struct net_device *dev) { struct tc35815_regs __iomem *tr = diff --git a/drivers/net/ethernet/via/via-rhine.c b/drivers/net/ethernet/via/via-rhine.c index ed12dbd156f0..803247d51fe9 100644 --- a/drivers/net/ethernet/via/via-rhine.c +++ b/drivers/net/ethernet/via/via-rhine.c @@ -506,7 +506,7 @@ static void mdio_write(struct net_device *dev, int phy_id, int location, int val static int rhine_open(struct net_device *dev); static void rhine_reset_task(struct work_struct *work); static void rhine_slow_event_task(struct work_struct *work); -static void rhine_tx_timeout(struct net_device *dev); +static void rhine_tx_timeout(struct net_device *dev, unsigned int txqueue); static netdev_tx_t rhine_start_tx(struct sk_buff *skb, struct net_device *dev); static irqreturn_t rhine_interrupt(int irq, void *dev_instance); @@ -1761,7 +1761,7 @@ out_unlock: mutex_unlock(&rp->task_lock); } -static void rhine_tx_timeout(struct net_device *dev) +static void rhine_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct rhine_private *rp = netdev_priv(dev); void __iomem *ioaddr = rp->base; diff --git a/drivers/net/ethernet/via/via-velocity.c b/drivers/net/ethernet/via/via-velocity.c index 346e44115c4e..4b556b74541a 100644 --- a/drivers/net/ethernet/via/via-velocity.c +++ b/drivers/net/ethernet/via/via-velocity.c @@ -3257,12 +3257,16 @@ static struct platform_driver velocity_platform_driver = { * @dev: network device * * Called before an ethtool operation. We need to make sure the - * chip is out of D3 state before we poke at it. + * chip is out of D3 state before we poke at it. In case of ethtool + * ops nesting, only wake the device up in the outermost block. */ static int velocity_ethtool_up(struct net_device *dev) { struct velocity_info *vptr = netdev_priv(dev); - if (!netif_running(dev)) + + if (vptr->ethtool_ops_nesting == U32_MAX) + return -EBUSY; + if (!vptr->ethtool_ops_nesting++ && !netif_running(dev)) velocity_set_power_state(vptr, PCI_D0); return 0; } @@ -3272,12 +3276,14 @@ static int velocity_ethtool_up(struct net_device *dev) * @dev: network device * * Called after an ethtool operation. Restore the chip back to D3 - * state if it isn't running. + * state if it isn't running. In case of ethtool ops nesting, only + * put the device to sleep in the outermost block. */ static void velocity_ethtool_down(struct net_device *dev) { struct velocity_info *vptr = netdev_priv(dev); - if (!netif_running(dev)) + + if (!--vptr->ethtool_ops_nesting && !netif_running(dev)) velocity_set_power_state(vptr, PCI_D3hot); } diff --git a/drivers/net/ethernet/via/via-velocity.h b/drivers/net/ethernet/via/via-velocity.h index cdfe7809e3c1..f196e71d2c04 100644 --- a/drivers/net/ethernet/via/via-velocity.h +++ b/drivers/net/ethernet/via/via-velocity.h @@ -1483,6 +1483,7 @@ struct velocity_info { struct velocity_context context; u32 ticks; + u32 ethtool_ops_nesting; u8 rev_id; diff --git a/drivers/net/ethernet/wiznet/w5100.c b/drivers/net/ethernet/wiznet/w5100.c index bede1ff289c5..c0d181a7f83a 100644 --- a/drivers/net/ethernet/wiznet/w5100.c +++ b/drivers/net/ethernet/wiznet/w5100.c @@ -790,7 +790,7 @@ static void w5100_restart_work(struct work_struct *work) w5100_restart(priv->ndev); } -static void w5100_tx_timeout(struct net_device *ndev) +static void w5100_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct w5100_priv *priv = netdev_priv(ndev); diff --git a/drivers/net/ethernet/wiznet/w5300.c b/drivers/net/ethernet/wiznet/w5300.c index 6ba2747779ce..46aae30c4636 100644 --- a/drivers/net/ethernet/wiznet/w5300.c +++ b/drivers/net/ethernet/wiznet/w5300.c @@ -341,7 +341,7 @@ static void w5300_get_regs(struct net_device *ndev, } } -static void w5300_tx_timeout(struct net_device *ndev) +static void w5300_tx_timeout(struct net_device *ndev, unsigned int txqueue) { struct w5300_priv *priv = netdev_priv(ndev); diff --git a/drivers/net/ethernet/xilinx/ll_temac_main.c b/drivers/net/ethernet/xilinx/ll_temac_main.c index c66aab78dcac..6f11f52c9a9e 100644 --- a/drivers/net/ethernet/xilinx/ll_temac_main.c +++ b/drivers/net/ethernet/xilinx/ll_temac_main.c @@ -1080,17 +1080,6 @@ temac_poll_controller(struct net_device *ndev) } #endif -static int temac_ioctl(struct net_device *ndev, struct ifreq *rq, int cmd) -{ - if (!netif_running(ndev)) - return -EINVAL; - - if (!ndev->phydev) - return -EINVAL; - - return phy_mii_ioctl(ndev->phydev, rq, cmd); -} - static const struct net_device_ops temac_netdev_ops = { .ndo_open = temac_open, .ndo_stop = temac_stop, @@ -1098,7 +1087,7 @@ static const struct net_device_ops temac_netdev_ops = { .ndo_set_rx_mode = temac_set_multicast_list, .ndo_set_mac_address = temac_set_mac_address, .ndo_validate_addr = eth_validate_addr, - .ndo_do_ioctl = temac_ioctl, + .ndo_do_ioctl = phy_do_ioctl_running, #ifdef CONFIG_NET_POLL_CONTROLLER .ndo_poll_controller = temac_poll_controller, #endif diff --git a/drivers/net/ethernet/xilinx/xilinx_emaclite.c b/drivers/net/ethernet/xilinx/xilinx_emaclite.c index 0de52e70abcc..0c26f5bcc523 100644 --- a/drivers/net/ethernet/xilinx/xilinx_emaclite.c +++ b/drivers/net/ethernet/xilinx/xilinx_emaclite.c @@ -521,7 +521,7 @@ static int xemaclite_set_mac_address(struct net_device *dev, void *address) * * This function is called when Tx time out occurs for Emaclite device. */ -static void xemaclite_tx_timeout(struct net_device *dev) +static void xemaclite_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct net_local *lp = netdev_priv(dev); unsigned long flags; diff --git a/drivers/net/ethernet/xircom/xirc2ps_cs.c b/drivers/net/ethernet/xircom/xirc2ps_cs.c index fd5288ff53b5..480ab7251515 100644 --- a/drivers/net/ethernet/xircom/xirc2ps_cs.c +++ b/drivers/net/ethernet/xircom/xirc2ps_cs.c @@ -288,7 +288,7 @@ struct local_info { */ static netdev_tx_t do_start_xmit(struct sk_buff *skb, struct net_device *dev); -static void xirc_tx_timeout(struct net_device *dev); +static void xirc_tx_timeout(struct net_device *dev, unsigned int txqueue); static void xirc2ps_tx_timeout_task(struct work_struct *work); static void set_addresses(struct net_device *dev); static void set_multicast_list(struct net_device *dev); @@ -1203,7 +1203,7 @@ xirc2ps_tx_timeout_task(struct work_struct *work) } static void -xirc_tx_timeout(struct net_device *dev) +xirc_tx_timeout(struct net_device *dev, unsigned int txqueue) { struct local_info *lp = netdev_priv(dev); dev->stats.tx_errors++; diff --git a/drivers/net/ethernet/xscale/Kconfig b/drivers/net/ethernet/xscale/Kconfig index cd0a8f46e7c6..98aa7b8ddb06 100644 --- a/drivers/net/ethernet/xscale/Kconfig +++ b/drivers/net/ethernet/xscale/Kconfig @@ -27,4 +27,18 @@ config IXP4XX_ETH Say Y here if you want to use built-in Ethernet ports on IXP4xx processor. +config PTP_1588_CLOCK_IXP46X + tristate "Intel IXP46x as PTP clock" + depends on IXP4XX_ETH + depends on PTP_1588_CLOCK + default y + help + This driver adds support for using the IXP46X as a PTP + clock. This clock is only useful if your PTP programs are + getting hardware time stamps on the PTP Ethernet packets + using the SO_TIMESTAMPING API. + + To compile this driver as a module, choose M here: the module + will be called ptp_ixp46x. + endif # NET_VENDOR_XSCALE diff --git a/drivers/net/ethernet/xscale/Makefile b/drivers/net/ethernet/xscale/Makefile index 794a519d07b3..607f91b1e878 100644 --- a/drivers/net/ethernet/xscale/Makefile +++ b/drivers/net/ethernet/xscale/Makefile @@ -3,4 +3,5 @@ # Makefile for the Intel XScale IXP device drivers. # -obj-$(CONFIG_IXP4XX_ETH) += ixp4xx_eth.o +obj-$(CONFIG_IXP4XX_ETH) += ixp4xx_eth.o +obj-$(CONFIG_PTP_1588_CLOCK_IXP46X) += ptp_ixp46x.o diff --git a/drivers/net/ethernet/xscale/ixp46x_ts.h b/drivers/net/ethernet/xscale/ixp46x_ts.h new file mode 100644 index 000000000000..d792130e27b0 --- /dev/null +++ b/drivers/net/ethernet/xscale/ixp46x_ts.h @@ -0,0 +1,68 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * PTP 1588 clock using the IXP46X + * + * Copyright (C) 2010 OMICRON electronics GmbH + */ + +#ifndef _IXP46X_TS_H_ +#define _IXP46X_TS_H_ + +#define DEFAULT_ADDEND 0xF0000029 +#define TICKS_NS_SHIFT 4 + +struct ixp46x_channel_ctl { + u32 ch_control; /* 0x40 Time Synchronization Channel Control */ + u32 ch_event; /* 0x44 Time Synchronization Channel Event */ + u32 tx_snap_lo; /* 0x48 Transmit Snapshot Low Register */ + u32 tx_snap_hi; /* 0x4C Transmit Snapshot High Register */ + u32 rx_snap_lo; /* 0x50 Receive Snapshot Low Register */ + u32 rx_snap_hi; /* 0x54 Receive Snapshot High Register */ + u32 src_uuid_lo; /* 0x58 Source UUID0 Low Register */ + u32 src_uuid_hi; /* 0x5C Sequence Identifier/Source UUID0 High */ +}; + +struct ixp46x_ts_regs { + u32 control; /* 0x00 Time Sync Control Register */ + u32 event; /* 0x04 Time Sync Event Register */ + u32 addend; /* 0x08 Time Sync Addend Register */ + u32 accum; /* 0x0C Time Sync Accumulator Register */ + u32 test; /* 0x10 Time Sync Test Register */ + u32 unused; /* 0x14 */ + u32 rsystime_lo; /* 0x18 RawSystemTime_Low Register */ + u32 rsystime_hi; /* 0x1C RawSystemTime_High Register */ + u32 systime_lo; /* 0x20 SystemTime_Low Register */ + u32 systime_hi; /* 0x24 SystemTime_High Register */ + u32 trgt_lo; /* 0x28 TargetTime_Low Register */ + u32 trgt_hi; /* 0x2C TargetTime_High Register */ + u32 asms_lo; /* 0x30 Auxiliary Slave Mode Snapshot Low */ + u32 asms_hi; /* 0x34 Auxiliary Slave Mode Snapshot High */ + u32 amms_lo; /* 0x38 Auxiliary Master Mode Snapshot Low */ + u32 amms_hi; /* 0x3C Auxiliary Master Mode Snapshot High */ + + struct ixp46x_channel_ctl channel[3]; +}; + +/* 0x00 Time Sync Control Register Bits */ +#define TSCR_AMM (1<<3) +#define TSCR_ASM (1<<2) +#define TSCR_TTM (1<<1) +#define TSCR_RST (1<<0) + +/* 0x04 Time Sync Event Register Bits */ +#define TSER_SNM (1<<3) +#define TSER_SNS (1<<2) +#define TTIPEND (1<<1) + +/* 0x40 Time Synchronization Channel Control Register Bits */ +#define MASTER_MODE (1<<0) +#define TIMESTAMP_ALL (1<<1) + +/* 0x44 Time Synchronization Channel Event Register Bits */ +#define TX_SNAPSHOT_LOCKED (1<<0) +#define RX_SNAPSHOT_LOCKED (1<<1) + +/* The ptp_ixp46x module will set this variable */ +extern int ixp46x_phc_index; + +#endif diff --git a/drivers/net/ethernet/xscale/ixp4xx_eth.c b/drivers/net/ethernet/xscale/ixp4xx_eth.c index 6fc04ffb22c2..269596c15133 100644 --- a/drivers/net/ethernet/xscale/ixp4xx_eth.c +++ b/drivers/net/ethernet/xscale/ixp4xx_eth.c @@ -29,14 +29,16 @@ #include <linux/net_tstamp.h> #include <linux/of.h> #include <linux/phy.h> +#include <linux/platform_data/eth_ixp4xx.h> #include <linux/platform_device.h> #include <linux/ptp_classify.h> #include <linux/slab.h> #include <linux/module.h> -#include <mach/ixp46x_ts.h> #include <linux/soc/ixp4xx/npe.h> #include <linux/soc/ixp4xx/qmgr.h> +#include "ixp46x_ts.h" + #define DEBUG_DESC 0 #define DEBUG_RX 0 #define DEBUG_TX 0 @@ -517,25 +519,14 @@ static int ixp4xx_mdio_write(struct mii_bus *bus, int phy_id, int location, return ret; } -static int ixp4xx_mdio_register(void) +static int ixp4xx_mdio_register(struct eth_regs __iomem *regs) { int err; if (!(mdio_bus = mdiobus_alloc())) return -ENOMEM; - if (cpu_is_ixp43x()) { - /* IXP43x lacks NPE-B and uses NPE-C for MII PHY access */ - if (!(ixp4xx_read_feature_bits() & IXP4XX_FEATURE_NPEC_ETH)) - return -ENODEV; - mdio_regs = (struct eth_regs __iomem *)IXP4XX_EthC_BASE_VIRT; - } else { - /* All MII PHY accesses use NPE-B Ethernet registers */ - if (!(ixp4xx_read_feature_bits() & IXP4XX_FEATURE_NPEB_ETH0)) - return -ENODEV; - mdio_regs = (struct eth_regs __iomem *)IXP4XX_EthB_BASE_VIRT; - } - + mdio_regs = regs; __raw_writel(DEFAULT_CORE_CNTRL, &mdio_regs->core_control); spin_lock_init(&mdio_lock); mdio_bus->name = "IXP4xx MII Bus"; @@ -581,8 +572,8 @@ static void ixp4xx_adjust_link(struct net_device *dev) __raw_writel(DEFAULT_TX_CNTRL0 | TX_CNTRL0_HALFDUPLEX, &port->regs->tx_control[0]); - printk(KERN_INFO "%s: link up, speed %u Mb/s, %s duplex\n", - dev->name, port->speed, port->duplex ? "full" : "half"); + netdev_info(dev, "%s: link up, speed %u Mb/s, %s duplex\n", + dev->name, port->speed, port->duplex ? "full" : "half"); } @@ -592,7 +583,7 @@ static inline void debug_pkt(struct net_device *dev, const char *func, #if DEBUG_PKT_BYTES int i; - printk(KERN_DEBUG "%s: %s(%i) ", dev->name, func, len); + netdev_debug(dev, "%s(%i) ", func, len); for (i = 0; i < len; i++) { if (i >= DEBUG_PKT_BYTES) break; @@ -683,7 +674,7 @@ static int eth_poll(struct napi_struct *napi, int budget) int received = 0; #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll\n", dev->name); + netdev_debug(dev, "eth_poll\n"); #endif while (received < budget) { @@ -697,23 +688,20 @@ static int eth_poll(struct napi_struct *napi, int budget) if ((n = queue_get_desc(rxq, port, 0)) < 0) { #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll napi_complete\n", - dev->name); + netdev_debug(dev, "eth_poll napi_complete\n"); #endif napi_complete(napi); qmgr_enable_irq(rxq); if (!qmgr_stat_below_low_watermark(rxq) && napi_reschedule(napi)) { /* not empty again */ #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll napi_reschedule succeeded\n", - dev->name); + netdev_debug(dev, "eth_poll napi_reschedule succeeded\n"); #endif qmgr_disable_irq(rxq); continue; } #if DEBUG_RX - printk(KERN_DEBUG "%s: eth_poll all done\n", - dev->name); + netdev_debug(dev, "eth_poll all done\n"); #endif return received; /* all work done */ } @@ -778,7 +766,7 @@ static int eth_poll(struct napi_struct *napi, int budget) } #if DEBUG_RX - printk(KERN_DEBUG "eth_poll(): end, not all work done\n"); + netdev_debug(dev, "eth_poll(): end, not all work done\n"); #endif return received; /* not all work done */ } @@ -842,7 +830,7 @@ static int eth_xmit(struct sk_buff *skb, struct net_device *dev) struct desc *desc; #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit\n", dev->name); + netdev_debug(dev, "eth_xmit\n"); #endif if (unlikely(skb->len > MAX_MRU)) { @@ -897,22 +885,21 @@ static int eth_xmit(struct sk_buff *skb, struct net_device *dev) if (qmgr_stat_below_low_watermark(txreadyq)) { /* empty */ #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit queue full\n", dev->name); + netdev_debug(dev, "eth_xmit queue full\n"); #endif netif_stop_queue(dev); /* we could miss TX ready interrupt */ /* really empty in fact */ if (!qmgr_stat_below_low_watermark(txreadyq)) { #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit ready again\n", - dev->name); + netdev_debug(dev, "eth_xmit ready again\n"); #endif netif_wake_queue(dev); } } #if DEBUG_TX - printk(KERN_DEBUG "%s: eth_xmit end\n", dev->name); + netdev_debug(dev, "eth_xmit end\n"); #endif ixp_tx_timestamp(port, skb); @@ -1099,7 +1086,7 @@ static int init_queues(struct port *port) int i; if (!ports_open) { - dma_pool = dma_pool_create(DRV_NAME, &port->netdev->dev, + dma_pool = dma_pool_create(DRV_NAME, port->netdev->dev.parent, POOL_ALLOC_SIZE, 32, 0); if (!dma_pool) return -ENOMEM; @@ -1186,8 +1173,7 @@ static int eth_open(struct net_device *dev) return err; if (npe_recv_message(npe, &msg, "ETH_GET_STATUS")) { - printk(KERN_ERR "%s: %s not responding\n", dev->name, - npe_name(npe)); + netdev_err(dev, "%s not responding\n", npe_name(npe)); return -EIO; } port->firmware[0] = msg.byte4; @@ -1299,7 +1285,7 @@ static int eth_close(struct net_device *dev) msg.eth_id = port->id; msg.byte3 = 1; if (npe_send_recv_message(port->npe, &msg, "ETH_ENABLE_LOOPBACK")) - printk(KERN_CRIT "%s: unable to enable loopback\n", dev->name); + netdev_crit(dev, "unable to enable loopback\n"); i = 0; do { /* drain RX buffers */ @@ -1323,11 +1309,11 @@ static int eth_close(struct net_device *dev) } while (++i < MAX_CLOSE_WAIT); if (buffs) - printk(KERN_CRIT "%s: unable to drain RX queue, %i buffer(s)" - " left in NPE\n", dev->name, buffs); + netdev_crit(dev, "unable to drain RX queue, %i buffer(s)" + " left in NPE\n", buffs); #if DEBUG_CLOSE if (!buffs) - printk(KERN_DEBUG "Draining RX queue took %i cycles\n", i); + netdev_debug(dev, "draining RX queue took %i cycles\n", i); #endif buffs = TX_DESCS; @@ -1343,17 +1329,16 @@ static int eth_close(struct net_device *dev) } while (++i < MAX_CLOSE_WAIT); if (buffs) - printk(KERN_CRIT "%s: unable to drain TX queue, %i buffer(s) " - "left in NPE\n", dev->name, buffs); + netdev_crit(dev, "unable to drain TX queue, %i buffer(s) " + "left in NPE\n", buffs); #if DEBUG_CLOSE if (!buffs) - printk(KERN_DEBUG "Draining TX queues took %i cycles\n", i); + netdev_debug(dev, "draining TX queues took %i cycles\n", i); #endif msg.byte3 = 0; if (npe_send_recv_message(port->npe, &msg, "ETH_DISABLE_LOOPBACK")) - printk(KERN_CRIT "%s: unable to disable loopback\n", - dev->name); + netdev_crit(dev, "unable to disable loopback\n"); phy_stop(dev->phydev); @@ -1374,54 +1359,88 @@ static const struct net_device_ops ixp4xx_netdev_ops = { .ndo_validate_addr = eth_validate_addr, }; -static int eth_init_one(struct platform_device *pdev) +static int ixp4xx_eth_probe(struct platform_device *pdev) { - struct port *port; - struct net_device *dev; - struct eth_plat_info *plat = dev_get_platdata(&pdev->dev); - struct phy_device *phydev = NULL; - u32 regs_phys; char phy_id[MII_BUS_ID_SIZE + 3]; + struct phy_device *phydev = NULL; + struct device *dev = &pdev->dev; + struct eth_plat_info *plat; + resource_size_t regs_phys; + struct net_device *ndev; + struct resource *res; + struct port *port; int err; - if (!(dev = alloc_etherdev(sizeof(struct port)))) + plat = dev_get_platdata(dev); + + if (!(ndev = devm_alloc_etherdev(dev, sizeof(struct port)))) return -ENOMEM; - SET_NETDEV_DEV(dev, &pdev->dev); - port = netdev_priv(dev); - port->netdev = dev; + SET_NETDEV_DEV(ndev, dev); + port = netdev_priv(ndev); + port->netdev = ndev; port->id = pdev->id; + /* Get the port resource and remap */ + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) + return -ENODEV; + regs_phys = res->start; + port->regs = devm_ioremap_resource(dev, res); + switch (port->id) { case IXP4XX_ETH_NPEA: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthA_BASE_VIRT; - regs_phys = IXP4XX_EthA_BASE_PHYS; + /* If the MDIO bus is not up yet, defer probe */ + if (!mdio_bus) + return -EPROBE_DEFER; break; case IXP4XX_ETH_NPEB: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthB_BASE_VIRT; - regs_phys = IXP4XX_EthB_BASE_PHYS; + /* + * On all except IXP43x, NPE-B is used for the MDIO bus. + * If there is no NPE-B in the feature set, bail out, else + * register the MDIO bus. + */ + if (!cpu_is_ixp43x()) { + if (!(ixp4xx_read_feature_bits() & + IXP4XX_FEATURE_NPEB_ETH0)) + return -ENODEV; + /* Else register the MDIO bus on NPE-B */ + if ((err = ixp4xx_mdio_register(port->regs))) + return err; + } + if (!mdio_bus) + return -EPROBE_DEFER; break; case IXP4XX_ETH_NPEC: - port->regs = (struct eth_regs __iomem *)IXP4XX_EthC_BASE_VIRT; - regs_phys = IXP4XX_EthC_BASE_PHYS; + /* + * IXP43x lacks NPE-B and uses NPE-C for the MDIO bus access, + * of there is no NPE-C, no bus, nothing works, so bail out. + */ + if (cpu_is_ixp43x()) { + if (!(ixp4xx_read_feature_bits() & + IXP4XX_FEATURE_NPEC_ETH)) + return -ENODEV; + /* Else register the MDIO bus on NPE-C */ + if ((err = ixp4xx_mdio_register(port->regs))) + return err; + } + if (!mdio_bus) + return -EPROBE_DEFER; break; default: - err = -ENODEV; - goto err_free; + return -ENODEV; } - dev->netdev_ops = &ixp4xx_netdev_ops; - dev->ethtool_ops = &ixp4xx_ethtool_ops; - dev->tx_queue_len = 100; + ndev->netdev_ops = &ixp4xx_netdev_ops; + ndev->ethtool_ops = &ixp4xx_ethtool_ops; + ndev->tx_queue_len = 100; - netif_napi_add(dev, &port->napi, eth_poll, NAPI_WEIGHT); + netif_napi_add(ndev, &port->napi, eth_poll, NAPI_WEIGHT); - if (!(port->npe = npe_request(NPE_ID(port->id)))) { - err = -EIO; - goto err_free; - } + if (!(port->npe = npe_request(NPE_ID(port->id)))) + return -EIO; - port->mem_res = request_mem_region(regs_phys, REGS_SIZE, dev->name); + port->mem_res = request_mem_region(regs_phys, REGS_SIZE, ndev->name); if (!port->mem_res) { err = -EBUSY; goto err_npe_rel; @@ -1429,9 +1448,9 @@ static int eth_init_one(struct platform_device *pdev) port->plat = plat; npe_port_tab[NPE_ID(port->id)] = port; - memcpy(dev->dev_addr, plat->hwaddr, ETH_ALEN); + memcpy(ndev->dev_addr, plat->hwaddr, ETH_ALEN); - platform_set_drvdata(pdev, dev); + platform_set_drvdata(pdev, ndev); __raw_writel(DEFAULT_CORE_CNTRL | CORE_RESET, &port->regs->core_control); @@ -1441,7 +1460,7 @@ static int eth_init_one(struct platform_device *pdev) snprintf(phy_id, MII_BUS_ID_SIZE + 3, PHY_ID_FMT, mdio_bus->id, plat->phy); - phydev = phy_connect(dev, phy_id, &ixp4xx_adjust_link, + phydev = phy_connect(ndev, phy_id, &ixp4xx_adjust_link, PHY_INTERFACE_MODE_MII); if (IS_ERR(phydev)) { err = PTR_ERR(phydev); @@ -1450,11 +1469,11 @@ static int eth_init_one(struct platform_device *pdev) phydev->irq = PHY_POLL; - if ((err = register_netdev(dev))) + if ((err = register_netdev(ndev))) goto err_phy_dis; - printk(KERN_INFO "%s: MII PHY %i on %s\n", dev->name, plat->phy, - npe_name(port->npe)); + netdev_info(ndev, "%s: MII PHY %i on %s\n", ndev->name, plat->phy, + npe_name(port->npe)); return 0; @@ -1465,58 +1484,32 @@ err_free_mem: release_resource(port->mem_res); err_npe_rel: npe_release(port->npe); -err_free: - free_netdev(dev); return err; } -static int eth_remove_one(struct platform_device *pdev) +static int ixp4xx_eth_remove(struct platform_device *pdev) { - struct net_device *dev = platform_get_drvdata(pdev); - struct phy_device *phydev = dev->phydev; - struct port *port = netdev_priv(dev); + struct net_device *ndev = platform_get_drvdata(pdev); + struct phy_device *phydev = ndev->phydev; + struct port *port = netdev_priv(ndev); - unregister_netdev(dev); + unregister_netdev(ndev); phy_disconnect(phydev); + ixp4xx_mdio_remove(); npe_port_tab[NPE_ID(port->id)] = NULL; npe_release(port->npe); release_resource(port->mem_res); - free_netdev(dev); return 0; } static struct platform_driver ixp4xx_eth_driver = { .driver.name = DRV_NAME, - .probe = eth_init_one, - .remove = eth_remove_one, + .probe = ixp4xx_eth_probe, + .remove = ixp4xx_eth_remove, }; - -static int __init eth_init_module(void) -{ - int err; - - /* - * FIXME: we bail out on device tree boot but this really needs - * to be fixed in a nicer way: this registers the MDIO bus before - * even matching the driver infrastructure, we should only probe - * detected hardware. - */ - if (of_have_populated_dt()) - return -ENODEV; - if ((err = ixp4xx_mdio_register())) - return err; - return platform_driver_register(&ixp4xx_eth_driver); -} - -static void __exit eth_cleanup_module(void) -{ - platform_driver_unregister(&ixp4xx_eth_driver); - ixp4xx_mdio_remove(); -} +module_platform_driver(ixp4xx_eth_driver); MODULE_AUTHOR("Krzysztof Halasa"); MODULE_DESCRIPTION("Intel IXP4xx Ethernet driver"); MODULE_LICENSE("GPL v2"); MODULE_ALIAS("platform:ixp4xx_eth"); -module_init(eth_init_module); -module_exit(eth_cleanup_module); diff --git a/drivers/net/ethernet/xscale/ptp_ixp46x.c b/drivers/net/ethernet/xscale/ptp_ixp46x.c new file mode 100644 index 000000000000..9ecc395239e9 --- /dev/null +++ b/drivers/net/ethernet/xscale/ptp_ixp46x.c @@ -0,0 +1,329 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * PTP 1588 clock using the IXP46X + * + * Copyright (C) 2010 OMICRON electronics GmbH + */ +#include <linux/device.h> +#include <linux/err.h> +#include <linux/gpio.h> +#include <linux/init.h> +#include <linux/interrupt.h> +#include <linux/io.h> +#include <linux/irq.h> +#include <linux/kernel.h> +#include <linux/module.h> + +#include <linux/ptp_clock_kernel.h> + +#include "ixp46x_ts.h" + +#define DRIVER "ptp_ixp46x" +#define N_EXT_TS 2 +#define MASTER_GPIO 8 +#define MASTER_IRQ 25 +#define SLAVE_GPIO 7 +#define SLAVE_IRQ 24 + +struct ixp_clock { + struct ixp46x_ts_regs *regs; + struct ptp_clock *ptp_clock; + struct ptp_clock_info caps; + int exts0_enabled; + int exts1_enabled; +}; + +DEFINE_SPINLOCK(register_lock); + +/* + * Register access functions + */ + +static u64 ixp_systime_read(struct ixp46x_ts_regs *regs) +{ + u64 ns; + u32 lo, hi; + + lo = __raw_readl(®s->systime_lo); + hi = __raw_readl(®s->systime_hi); + + ns = ((u64) hi) << 32; + ns |= lo; + ns <<= TICKS_NS_SHIFT; + + return ns; +} + +static void ixp_systime_write(struct ixp46x_ts_regs *regs, u64 ns) +{ + u32 hi, lo; + + ns >>= TICKS_NS_SHIFT; + hi = ns >> 32; + lo = ns & 0xffffffff; + + __raw_writel(lo, ®s->systime_lo); + __raw_writel(hi, ®s->systime_hi); +} + +/* + * Interrupt service routine + */ + +static irqreturn_t isr(int irq, void *priv) +{ + struct ixp_clock *ixp_clock = priv; + struct ixp46x_ts_regs *regs = ixp_clock->regs; + struct ptp_clock_event event; + u32 ack = 0, lo, hi, val; + + val = __raw_readl(®s->event); + + if (val & TSER_SNS) { + ack |= TSER_SNS; + if (ixp_clock->exts0_enabled) { + hi = __raw_readl(®s->asms_hi); + lo = __raw_readl(®s->asms_lo); + event.type = PTP_CLOCK_EXTTS; + event.index = 0; + event.timestamp = ((u64) hi) << 32; + event.timestamp |= lo; + event.timestamp <<= TICKS_NS_SHIFT; + ptp_clock_event(ixp_clock->ptp_clock, &event); + } + } + + if (val & TSER_SNM) { + ack |= TSER_SNM; + if (ixp_clock->exts1_enabled) { + hi = __raw_readl(®s->amms_hi); + lo = __raw_readl(®s->amms_lo); + event.type = PTP_CLOCK_EXTTS; + event.index = 1; + event.timestamp = ((u64) hi) << 32; + event.timestamp |= lo; + event.timestamp <<= TICKS_NS_SHIFT; + ptp_clock_event(ixp_clock->ptp_clock, &event); + } + } + + if (val & TTIPEND) + ack |= TTIPEND; /* this bit seems to be always set */ + + if (ack) { + __raw_writel(ack, ®s->event); + return IRQ_HANDLED; + } else + return IRQ_NONE; +} + +/* + * PTP clock operations + */ + +static int ptp_ixp_adjfreq(struct ptp_clock_info *ptp, s32 ppb) +{ + u64 adj; + u32 diff, addend; + int neg_adj = 0; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + if (ppb < 0) { + neg_adj = 1; + ppb = -ppb; + } + addend = DEFAULT_ADDEND; + adj = addend; + adj *= ppb; + diff = div_u64(adj, 1000000000ULL); + + addend = neg_adj ? addend - diff : addend + diff; + + __raw_writel(addend, ®s->addend); + + return 0; +} + +static int ptp_ixp_adjtime(struct ptp_clock_info *ptp, s64 delta) +{ + s64 now; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + spin_lock_irqsave(®ister_lock, flags); + + now = ixp_systime_read(regs); + now += delta; + ixp_systime_write(regs, now); + + spin_unlock_irqrestore(®ister_lock, flags); + + return 0; +} + +static int ptp_ixp_gettime(struct ptp_clock_info *ptp, struct timespec64 *ts) +{ + u64 ns; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + spin_lock_irqsave(®ister_lock, flags); + + ns = ixp_systime_read(regs); + + spin_unlock_irqrestore(®ister_lock, flags); + + *ts = ns_to_timespec64(ns); + return 0; +} + +static int ptp_ixp_settime(struct ptp_clock_info *ptp, + const struct timespec64 *ts) +{ + u64 ns; + unsigned long flags; + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + struct ixp46x_ts_regs *regs = ixp_clock->regs; + + ns = timespec64_to_ns(ts); + + spin_lock_irqsave(®ister_lock, flags); + + ixp_systime_write(regs, ns); + + spin_unlock_irqrestore(®ister_lock, flags); + + return 0; +} + +static int ptp_ixp_enable(struct ptp_clock_info *ptp, + struct ptp_clock_request *rq, int on) +{ + struct ixp_clock *ixp_clock = container_of(ptp, struct ixp_clock, caps); + + switch (rq->type) { + case PTP_CLK_REQ_EXTTS: + switch (rq->extts.index) { + case 0: + ixp_clock->exts0_enabled = on ? 1 : 0; + break; + case 1: + ixp_clock->exts1_enabled = on ? 1 : 0; + break; + default: + return -EINVAL; + } + return 0; + default: + break; + } + + return -EOPNOTSUPP; +} + +static const struct ptp_clock_info ptp_ixp_caps = { + .owner = THIS_MODULE, + .name = "IXP46X timer", + .max_adj = 66666655, + .n_ext_ts = N_EXT_TS, + .n_pins = 0, + .pps = 0, + .adjfreq = ptp_ixp_adjfreq, + .adjtime = ptp_ixp_adjtime, + .gettime64 = ptp_ixp_gettime, + .settime64 = ptp_ixp_settime, + .enable = ptp_ixp_enable, +}; + +/* module operations */ + +static struct ixp_clock ixp_clock; + +static int setup_interrupt(int gpio) +{ + int irq; + int err; + + err = gpio_request(gpio, "ixp4-ptp"); + if (err) + return err; + + err = gpio_direction_input(gpio); + if (err) + return err; + + irq = gpio_to_irq(gpio); + if (irq < 0) + return irq; + + err = irq_set_irq_type(irq, IRQF_TRIGGER_FALLING); + if (err) { + pr_err("cannot set trigger type for irq %d\n", irq); + return err; + } + + err = request_irq(irq, isr, 0, DRIVER, &ixp_clock); + if (err) { + pr_err("request_irq failed for irq %d\n", irq); + return err; + } + + return irq; +} + +static void __exit ptp_ixp_exit(void) +{ + free_irq(MASTER_IRQ, &ixp_clock); + free_irq(SLAVE_IRQ, &ixp_clock); + ixp46x_phc_index = -1; + ptp_clock_unregister(ixp_clock.ptp_clock); +} + +static int __init ptp_ixp_init(void) +{ + if (!cpu_is_ixp46x()) + return -ENODEV; + + ixp_clock.regs = + (struct ixp46x_ts_regs __iomem *) IXP4XX_TIMESYNC_BASE_VIRT; + + ixp_clock.caps = ptp_ixp_caps; + + ixp_clock.ptp_clock = ptp_clock_register(&ixp_clock.caps, NULL); + + if (IS_ERR(ixp_clock.ptp_clock)) + return PTR_ERR(ixp_clock.ptp_clock); + + ixp46x_phc_index = ptp_clock_index(ixp_clock.ptp_clock); + + __raw_writel(DEFAULT_ADDEND, &ixp_clock.regs->addend); + __raw_writel(1, &ixp_clock.regs->trgt_lo); + __raw_writel(0, &ixp_clock.regs->trgt_hi); + __raw_writel(TTIPEND, &ixp_clock.regs->event); + + if (MASTER_IRQ != setup_interrupt(MASTER_GPIO)) { + pr_err("failed to setup gpio %d as irq\n", MASTER_GPIO); + goto no_master; + } + if (SLAVE_IRQ != setup_interrupt(SLAVE_GPIO)) { + pr_err("failed to setup gpio %d as irq\n", SLAVE_GPIO); + goto no_slave; + } + + return 0; +no_slave: + free_irq(MASTER_IRQ, &ixp_clock); +no_master: + ptp_clock_unregister(ixp_clock.ptp_clock); + return -ENODEV; +} + +module_init(ptp_ixp_init); +module_exit(ptp_ixp_exit); + +MODULE_AUTHOR("Richard Cochran <richardcochran@gmail.com>"); +MODULE_DESCRIPTION("PTP clock using the IXP46X timer"); +MODULE_LICENSE("GPL"); |