diff options
Diffstat (limited to 'drivers')
539 files changed, 11648 insertions, 4307 deletions
diff --git a/drivers/acpi/acpica/exconfig.c b/drivers/acpi/acpica/exconfig.c index a1d177d58254..21932d640a41 100644 --- a/drivers/acpi/acpica/exconfig.c +++ b/drivers/acpi/acpica/exconfig.c @@ -108,7 +108,9 @@ acpi_ex_add_table(u32 table_index, /* Add the table to the namespace */ + acpi_ex_exit_interpreter(); status = acpi_ns_load_table(table_index, parent_node); + acpi_ex_enter_interpreter(); if (ACPI_FAILURE(status)) { acpi_ut_remove_reference(obj_desc); *ddb_handle = NULL; diff --git a/drivers/acpi/acpica/hwregs.c b/drivers/acpi/acpica/hwregs.c index daceb80022b0..3b7fb99362b6 100644 --- a/drivers/acpi/acpica/hwregs.c +++ b/drivers/acpi/acpica/hwregs.c @@ -306,12 +306,6 @@ acpi_status acpi_hw_read(u32 *value, struct acpi_generic_address *reg) acpi_status acpi_hw_write(u32 value, struct acpi_generic_address *reg) { u64 address; - u8 access_width; - u32 bit_width; - u8 bit_offset; - u64 value64; - u32 new_value32, old_value32; - u8 index; acpi_status status; ACPI_FUNCTION_NAME(hw_write); @@ -323,145 +317,23 @@ acpi_status acpi_hw_write(u32 value, struct acpi_generic_address *reg) return (status); } - /* Convert access_width into number of bits based */ - - access_width = acpi_hw_get_access_bit_width(reg, 32); - bit_width = reg->bit_offset + reg->bit_width; - bit_offset = reg->bit_offset; - /* * Two address spaces supported: Memory or IO. PCI_Config is * not supported here because the GAS structure is insufficient */ - index = 0; - while (bit_width) { - /* - * Use offset style bit reads because "Index * AccessWidth" is - * ensured to be less than 32-bits by acpi_hw_validate_register(). - */ - new_value32 = ACPI_GET_BITS(&value, index * access_width, - ACPI_MASK_BITS_ABOVE_32 - (access_width)); - - if (bit_offset >= access_width) { - bit_offset -= access_width; - } else { - /* - * Use offset style bit masks because access_width is ensured - * to be less than 32-bits by acpi_hw_validate_register() and - * bit_offset/bit_width is less than access_width here. - */ - if (bit_offset) { - new_value32 &= ACPI_MASK_BITS_BELOW(bit_offset); - } - if (bit_width < access_width) { - new_value32 &= ACPI_MASK_BITS_ABOVE(bit_width); - } - - if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { - if (bit_offset || bit_width < access_width) { - /* - * Read old values in order not to modify the bits that - * are beyond the register bit_width/bit_offset setting. - */ - status = - acpi_os_read_memory((acpi_physical_address) - address + - index * - ACPI_DIV_8 - (access_width), - &value64, - access_width); - old_value32 = (u32)value64; - - /* - * Use offset style bit masks because access_width is - * ensured to be less than 32-bits by - * acpi_hw_validate_register() and bit_offset/bit_width is - * less than access_width here. - */ - if (bit_offset) { - old_value32 &= - ACPI_MASK_BITS_ABOVE - (bit_offset); - bit_offset = 0; - } - if (bit_width < access_width) { - old_value32 &= - ACPI_MASK_BITS_BELOW - (bit_width); - } - - new_value32 |= old_value32; - } - - value64 = (u64)new_value32; - status = - acpi_os_write_memory((acpi_physical_address) - address + - index * - ACPI_DIV_8 - (access_width), - value64, access_width); - } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ - - if (bit_offset || bit_width < access_width) { - /* - * Read old values in order not to modify the bits that - * are beyond the register bit_width/bit_offset setting. - */ - status = - acpi_hw_read_port((acpi_io_address) - address + - index * - ACPI_DIV_8 - (access_width), - &old_value32, - access_width); - - /* - * Use offset style bit masks because access_width is - * ensured to be less than 32-bits by - * acpi_hw_validate_register() and bit_offset/bit_width is - * less than access_width here. - */ - if (bit_offset) { - old_value32 &= - ACPI_MASK_BITS_ABOVE - (bit_offset); - bit_offset = 0; - } - if (bit_width < access_width) { - old_value32 &= - ACPI_MASK_BITS_BELOW - (bit_width); - } - - new_value32 |= old_value32; - } - - status = acpi_hw_write_port((acpi_io_address) - address + - index * - ACPI_DIV_8 - (access_width), - new_value32, - access_width); - } - } - - /* - * Index * access_width is ensured to be less than 32-bits by - * acpi_hw_validate_register(). - */ - bit_width -= - bit_width > access_width ? access_width : bit_width; - index++; + if (reg->space_id == ACPI_ADR_SPACE_SYSTEM_MEMORY) { + status = acpi_os_write_memory((acpi_physical_address) + address, (u64)value, + reg->bit_width); + } else { /* ACPI_ADR_SPACE_SYSTEM_IO, validated earlier */ + + status = acpi_hw_write_port((acpi_io_address) + address, value, reg->bit_width); } ACPI_DEBUG_PRINT((ACPI_DB_IO, "Wrote: %8.8X width %2d to %8.8X%8.8X (%s)\n", - value, access_width, ACPI_FORMAT_UINT64(address), + value, reg->bit_width, ACPI_FORMAT_UINT64(address), acpi_ut_get_region_name(reg->space_id))); return (status); diff --git a/drivers/acpi/acpica/nsparse.c b/drivers/acpi/acpica/nsparse.c index f631a47724f0..1783cd7e1446 100644 --- a/drivers/acpi/acpica/nsparse.c +++ b/drivers/acpi/acpica/nsparse.c @@ -47,6 +47,7 @@ #include "acparser.h" #include "acdispat.h" #include "actables.h" +#include "acinterp.h" #define _COMPONENT ACPI_NAMESPACE ACPI_MODULE_NAME("nsparse") @@ -170,6 +171,8 @@ acpi_ns_parse_table(u32 table_index, struct acpi_namespace_node *start_node) ACPI_FUNCTION_TRACE(ns_parse_table); + acpi_ex_enter_interpreter(); + /* * AML Parse, pass 1 * @@ -185,7 +188,7 @@ acpi_ns_parse_table(u32 table_index, struct acpi_namespace_node *start_node) status = acpi_ns_one_complete_parse(ACPI_IMODE_LOAD_PASS1, table_index, start_node); if (ACPI_FAILURE(status)) { - return_ACPI_STATUS(status); + goto error_exit; } /* @@ -201,8 +204,10 @@ acpi_ns_parse_table(u32 table_index, struct acpi_namespace_node *start_node) status = acpi_ns_one_complete_parse(ACPI_IMODE_LOAD_PASS2, table_index, start_node); if (ACPI_FAILURE(status)) { - return_ACPI_STATUS(status); + goto error_exit; } +error_exit: + acpi_ex_exit_interpreter(); return_ACPI_STATUS(status); } diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c index 31e8da648fff..262ca31b86d9 100644 --- a/drivers/acpi/bus.c +++ b/drivers/acpi/bus.c @@ -1051,7 +1051,7 @@ static int __init acpi_bus_init(void) * Maybe EC region is required at bus_scan/acpi_get_devices. So it * is necessary to enable it as early as possible. */ - acpi_boot_ec_enable(); + acpi_ec_dsdt_probe(); printk(KERN_INFO PREFIX "Interpreter enabled\n"); diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c index 0e70181f150c..73c76d646064 100644 --- a/drivers/acpi/ec.c +++ b/drivers/acpi/ec.c @@ -1446,10 +1446,30 @@ ec_parse_io_ports(struct acpi_resource *resource, void *context) return AE_OK; } -int __init acpi_boot_ec_enable(void) +static const struct acpi_device_id ec_device_ids[] = { + {"PNP0C09", 0}, + {"", 0}, +}; + +int __init acpi_ec_dsdt_probe(void) { - if (!boot_ec) + acpi_status status; + + if (boot_ec) return 0; + + /* + * Finding EC from DSDT if there is no ECDT EC available. When this + * function is invoked, ACPI tables have been fully loaded, we can + * walk namespace now. + */ + boot_ec = make_acpi_ec(); + if (!boot_ec) + return -ENOMEM; + status = acpi_get_devices(ec_device_ids[0].id, + ec_parse_device, boot_ec, NULL); + if (ACPI_FAILURE(status) || !boot_ec->handle) + return -ENODEV; if (!ec_install_handlers(boot_ec)) { first_ec = boot_ec; return 0; @@ -1457,11 +1477,6 @@ int __init acpi_boot_ec_enable(void) return -EFAULT; } -static const struct acpi_device_id ec_device_ids[] = { - {"PNP0C09", 0}, - {"", 0}, -}; - #if 0 /* * Some EC firmware variations refuses to respond QR_EC when SCI_EVT is not diff --git a/drivers/acpi/internal.h b/drivers/acpi/internal.h index 9bb0773d39bf..27cc7feabfe4 100644 --- a/drivers/acpi/internal.h +++ b/drivers/acpi/internal.h @@ -181,7 +181,7 @@ typedef int (*acpi_ec_query_func) (void *data); int acpi_ec_init(void); int acpi_ec_ecdt_probe(void); -int acpi_boot_ec_enable(void); +int acpi_ec_dsdt_probe(void); void acpi_ec_block_transactions(void); void acpi_ec_unblock_transactions(void); void acpi_ec_unblock_transactions_early(void); diff --git a/drivers/ata/libata-eh.c b/drivers/ata/libata-eh.c index 61dc7a99e89a..c6f017458958 100644 --- a/drivers/ata/libata-eh.c +++ b/drivers/ata/libata-eh.c @@ -606,7 +606,7 @@ void ata_scsi_error(struct Scsi_Host *host) ata_scsi_port_error_handler(host, ap); /* finish or retry handled scmd's and clean up */ - WARN_ON(host->host_failed || !list_empty(&eh_work_q)); + WARN_ON(!list_empty(&eh_work_q)); DPRINTK("EXIT\n"); } diff --git a/drivers/base/Makefile b/drivers/base/Makefile index 6b2a84e7f2be..2609ba20b396 100644 --- a/drivers/base/Makefile +++ b/drivers/base/Makefile @@ -10,7 +10,7 @@ obj-$(CONFIG_DMA_CMA) += dma-contiguous.o obj-y += power/ obj-$(CONFIG_HAS_DMA) += dma-mapping.o obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o -obj-$(CONFIG_ISA) += isa.o +obj-$(CONFIG_ISA_BUS_API) += isa.o obj-$(CONFIG_FW_LOADER) += firmware_class.o obj-$(CONFIG_NUMA) += node.o obj-$(CONFIG_MEMORY_HOTPLUG_SPARSE) += memory.o diff --git a/drivers/base/isa.c b/drivers/base/isa.c index 91dba65d7264..cd6ccdcf9df0 100644 --- a/drivers/base/isa.c +++ b/drivers/base/isa.c @@ -180,4 +180,4 @@ static int __init isa_bus_init(void) return error; } -device_initcall(isa_bus_init); +postcore_initcall(isa_bus_init); diff --git a/drivers/base/module.c b/drivers/base/module.c index db930d3ee312..2a215780eda2 100644 --- a/drivers/base/module.c +++ b/drivers/base/module.c @@ -24,10 +24,12 @@ static char *make_driver_name(struct device_driver *drv) static void module_create_drivers_dir(struct module_kobject *mk) { - if (!mk || mk->drivers_dir) - return; + static DEFINE_MUTEX(drivers_dir_mutex); - mk->drivers_dir = kobject_create_and_add("drivers", &mk->kobj); + mutex_lock(&drivers_dir_mutex); + if (mk && !mk->drivers_dir) + mk->drivers_dir = kobject_create_and_add("drivers", &mk->kobj); + mutex_unlock(&drivers_dir_mutex); } void module_add_driver(struct module *mod, struct device_driver *drv) diff --git a/drivers/base/power/opp/cpu.c b/drivers/base/power/opp/cpu.c index 83d6e7ba1a34..8c3434bdb26d 100644 --- a/drivers/base/power/opp/cpu.c +++ b/drivers/base/power/opp/cpu.c @@ -211,7 +211,7 @@ int dev_pm_opp_set_sharing_cpus(struct device *cpu_dev, } /* Mark opp-table as multiple CPUs are sharing it now */ - opp_table->shared_opp = true; + opp_table->shared_opp = OPP_TABLE_ACCESS_SHARED; } unlock: mutex_unlock(&opp_table_lock); @@ -227,7 +227,8 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_set_sharing_cpus); * * This updates the @cpumask with CPUs that are sharing OPPs with @cpu_dev. * - * Returns -ENODEV if OPP table isn't already present. + * Returns -ENODEV if OPP table isn't already present and -EINVAL if the OPP + * table's status is access-unknown. * * Locking: The internal opp_table and opp structures are RCU protected. * Hence this function internally uses RCU updater strategy with mutex locks @@ -249,9 +250,14 @@ int dev_pm_opp_get_sharing_cpus(struct device *cpu_dev, struct cpumask *cpumask) goto unlock; } + if (opp_table->shared_opp == OPP_TABLE_ACCESS_UNKNOWN) { + ret = -EINVAL; + goto unlock; + } + cpumask_clear(cpumask); - if (opp_table->shared_opp) { + if (opp_table->shared_opp == OPP_TABLE_ACCESS_SHARED) { list_for_each_entry(opp_dev, &opp_table->dev_list, node) cpumask_set_cpu(opp_dev->dev->id, cpumask); } else { diff --git a/drivers/base/power/opp/of.c b/drivers/base/power/opp/of.c index 94d2010558e3..1dfd3dd92624 100644 --- a/drivers/base/power/opp/of.c +++ b/drivers/base/power/opp/of.c @@ -34,7 +34,10 @@ static struct opp_table *_managed_opp(const struct device_node *np) * But the OPPs will be considered as shared only if the * OPP table contains a "opp-shared" property. */ - return opp_table->shared_opp ? opp_table : NULL; + if (opp_table->shared_opp == OPP_TABLE_ACCESS_SHARED) + return opp_table; + + return NULL; } } @@ -353,7 +356,10 @@ static int _of_add_opp_table_v2(struct device *dev, struct device_node *opp_np) } opp_table->np = opp_np; - opp_table->shared_opp = of_property_read_bool(opp_np, "opp-shared"); + if (of_property_read_bool(opp_np, "opp-shared")) + opp_table->shared_opp = OPP_TABLE_ACCESS_SHARED; + else + opp_table->shared_opp = OPP_TABLE_ACCESS_EXCLUSIVE; mutex_unlock(&opp_table_lock); diff --git a/drivers/base/power/opp/opp.h b/drivers/base/power/opp/opp.h index 20f3be22e060..fabd5ca1a083 100644 --- a/drivers/base/power/opp/opp.h +++ b/drivers/base/power/opp/opp.h @@ -119,6 +119,12 @@ struct opp_device { #endif }; +enum opp_table_access { + OPP_TABLE_ACCESS_UNKNOWN = 0, + OPP_TABLE_ACCESS_EXCLUSIVE = 1, + OPP_TABLE_ACCESS_SHARED = 2, +}; + /** * struct opp_table - Device opp structure * @node: table node - contains the devices with OPPs that @@ -166,7 +172,7 @@ struct opp_table { /* For backward compatibility with v1 bindings */ unsigned int voltage_tolerance_v1; - bool shared_opp; + enum opp_table_access shared_opp; struct dev_pm_opp *suspend_opp; unsigned int *supported_hw; diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c index d597e432e195..ab19adb07a12 100644 --- a/drivers/block/aoe/aoecmd.c +++ b/drivers/block/aoe/aoecmd.c @@ -1750,7 +1750,7 @@ aoecmd_init(void) int ret; /* get_zeroed_page returns page with ref count 1 */ - p = (void *) get_zeroed_page(GFP_KERNEL | __GFP_REPEAT); + p = (void *) get_zeroed_page(GFP_KERNEL); if (!p) return -ENOMEM; empty_page = virt_to_page(p); diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 31e73a7a40f2..6a48ed41963f 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -941,7 +941,7 @@ static int nbd_dev_dbg_init(struct nbd_device *nbd) debugfs_create_u64("size_bytes", 0444, dir, &nbd->bytesize); debugfs_create_u32("timeout", 0444, dir, &nbd->xmit_timeout); debugfs_create_u32("blocksize", 0444, dir, &nbd->blksize); - debugfs_create_file("flags", 0444, dir, &nbd, &nbd_dbg_flags_ops); + debugfs_create_file("flags", 0444, dir, nbd, &nbd_dbg_flags_ops); return 0; } diff --git a/drivers/block/xen-blkfront.c b/drivers/block/xen-blkfront.c index ca13df854639..2e6d1e9c3345 100644 --- a/drivers/block/xen-blkfront.c +++ b/drivers/block/xen-blkfront.c @@ -874,8 +874,12 @@ static int blkif_queue_rq(struct blk_mq_hw_ctx *hctx, const struct blk_mq_queue_data *qd) { unsigned long flags; - struct blkfront_ring_info *rinfo = (struct blkfront_ring_info *)hctx->driver_data; + int qid = hctx->queue_num; + struct blkfront_info *info = hctx->queue->queuedata; + struct blkfront_ring_info *rinfo = NULL; + BUG_ON(info->nr_rings <= qid); + rinfo = &info->rinfo[qid]; blk_mq_start_request(qd->rq); spin_lock_irqsave(&rinfo->ring_lock, flags); if (RING_FULL(&rinfo->ring)) @@ -901,20 +905,9 @@ out_busy: return BLK_MQ_RQ_QUEUE_BUSY; } -static int blk_mq_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, - unsigned int index) -{ - struct blkfront_info *info = (struct blkfront_info *)data; - - BUG_ON(info->nr_rings <= index); - hctx->driver_data = &info->rinfo[index]; - return 0; -} - static struct blk_mq_ops blkfront_mq_ops = { .queue_rq = blkif_queue_rq, .map_queue = blk_mq_map_queue, - .init_hctx = blk_mq_init_hctx, }; static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size, @@ -950,6 +943,7 @@ static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size, return PTR_ERR(rq); } + rq->queuedata = info; queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq); if (info->feature_discard) { @@ -2149,6 +2143,8 @@ static int blkfront_resume(struct xenbus_device *dev) return err; err = talk_to_blkback(dev, info); + if (!err) + blk_mq_update_nr_hw_queues(&info->tag_set, info->nr_rings); /* * We have to wait for the backend to switch to @@ -2485,10 +2481,23 @@ static void blkback_changed(struct xenbus_device *dev, break; case XenbusStateConnected: - if (dev->state != XenbusStateInitialised) { + /* + * talk_to_blkback sets state to XenbusStateInitialised + * and blkfront_connect sets it to XenbusStateConnected + * (if connection went OK). + * + * If the backend (or toolstack) decides to poke at backend + * state (and re-trigger the watch by setting the state repeatedly + * to XenbusStateConnected (4)) we need to deal with this. + * This is allowed as this is used to communicate to the guest + * that the size of disk has changed! + */ + if ((dev->state != XenbusStateInitialised) && + (dev->state != XenbusStateConnected)) { if (talk_to_blkback(dev, info)) break; } + blkfront_connect(info); break; diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c index 94fb407d8561..44b1bd6baa38 100644 --- a/drivers/char/ipmi/ipmi_msghandler.c +++ b/drivers/char/ipmi/ipmi_msghandler.c @@ -3820,6 +3820,7 @@ static void handle_new_recv_msgs(ipmi_smi_t intf) while (!list_empty(&intf->waiting_rcv_msgs)) { smi_msg = list_entry(intf->waiting_rcv_msgs.next, struct ipmi_smi_msg, link); + list_del(&smi_msg->link); if (!run_to_completion) spin_unlock_irqrestore(&intf->waiting_rcv_msgs_lock, flags); @@ -3829,11 +3830,14 @@ static void handle_new_recv_msgs(ipmi_smi_t intf) if (rv > 0) { /* * To preserve message order, quit if we - * can't handle a message. + * can't handle a message. Add the message + * back at the head, this is safe because this + * tasklet is the only thing that pulls the + * messages. */ + list_add(&smi_msg->link, &intf->waiting_rcv_msgs); break; } else { - list_del(&smi_msg->link); if (rv == 0) /* Message handled */ ipmi_free_smi_msg(smi_msg); diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig index 53ddba26578c..98efbfcdb503 100644 --- a/drivers/clk/Kconfig +++ b/drivers/clk/Kconfig @@ -175,6 +175,7 @@ config COMMON_CLK_KEYSTONE config COMMON_CLK_NXP def_bool COMMON_CLK && (ARCH_LPC18XX || ARCH_LPC32XX) select REGMAP_MMIO if ARCH_LPC32XX + select MFD_SYSCON if ARCH_LPC18XX ---help--- Support for clock providers on NXP platforms. diff --git a/drivers/clk/microchip/clk-pic32mzda.c b/drivers/clk/microchip/clk-pic32mzda.c index 020a29acc5b0..51f54380474b 100644 --- a/drivers/clk/microchip/clk-pic32mzda.c +++ b/drivers/clk/microchip/clk-pic32mzda.c @@ -180,15 +180,15 @@ static int pic32mzda_clk_probe(struct platform_device *pdev) /* register fixed rate clocks */ clks[POSCCLK] = clk_register_fixed_rate(&pdev->dev, "posc_clk", NULL, - CLK_IS_ROOT, 24000000); + 0, 24000000); clks[FRCCLK] = clk_register_fixed_rate(&pdev->dev, "frc_clk", NULL, - CLK_IS_ROOT, 8000000); + 0, 8000000); clks[BFRCCLK] = clk_register_fixed_rate(&pdev->dev, "bfrc_clk", NULL, - CLK_IS_ROOT, 8000000); + 0, 8000000); clks[LPRCCLK] = clk_register_fixed_rate(&pdev->dev, "lprc_clk", NULL, - CLK_IS_ROOT, 32000); + 0, 32000); clks[UPLLCLK] = clk_register_fixed_rate(&pdev->dev, "usbphy_clk", NULL, - CLK_IS_ROOT, 24000000); + 0, 24000000); /* fixed rate (optional) clock */ if (of_find_property(np, "microchip,pic32mzda-sosc", NULL)) { pr_info("pic32-clk: dt requests SOSC.\n"); diff --git a/drivers/cpufreq/intel_pstate.c b/drivers/cpufreq/intel_pstate.c index 0d159b513469..fe9dc17ea873 100644 --- a/drivers/cpufreq/intel_pstate.c +++ b/drivers/cpufreq/intel_pstate.c @@ -372,26 +372,9 @@ static bool intel_pstate_get_ppc_enable_status(void) return acpi_ppc; } -/* - * The max target pstate ratio is a 8 bit value in both PLATFORM_INFO MSR and - * in TURBO_RATIO_LIMIT MSR, which pstate driver stores in max_pstate and - * max_turbo_pstate fields. The PERF_CTL MSR contains 16 bit value for P state - * ratio, out of it only high 8 bits are used. For example 0x1700 is setting - * target ratio 0x17. The _PSS control value stores in a format which can be - * directly written to PERF_CTL MSR. But in intel_pstate driver this shift - * occurs during write to PERF_CTL (E.g. for cores core_set_pstate()). - * This function converts the _PSS control value to intel pstate driver format - * for comparison and assignment. - */ -static int convert_to_native_pstate_format(struct cpudata *cpu, int index) -{ - return cpu->acpi_perf_data.states[index].control >> 8; -} - static void intel_pstate_init_acpi_perf_limits(struct cpufreq_policy *policy) { struct cpudata *cpu; - int turbo_pss_ctl; int ret; int i; @@ -441,11 +424,10 @@ static void intel_pstate_init_acpi_perf_limits(struct cpufreq_policy *policy) * max frequency, which will cause a reduced performance as * this driver uses real max turbo frequency as the max * frequency. So correct this frequency in _PSS table to - * correct max turbo frequency based on the turbo ratio. + * correct max turbo frequency based on the turbo state. * Also need to convert to MHz as _PSS freq is in MHz. */ - turbo_pss_ctl = convert_to_native_pstate_format(cpu, 0); - if (turbo_pss_ctl > cpu->pstate.max_pstate) + if (!limits->turbo_disabled) cpu->acpi_perf_data.states[0].core_frequency = policy->cpuinfo.max_freq / 1000; cpu->valid_pss_table = true; @@ -1460,6 +1442,9 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) intel_pstate_clear_update_util_hook(policy->cpu); + pr_debug("set_policy cpuinfo.max %u policy->max %u\n", + policy->cpuinfo.max_freq, policy->max); + cpu = all_cpu_data[0]; if (cpu->pstate.max_pstate_physical > cpu->pstate.max_pstate && policy->max < policy->cpuinfo.max_freq && @@ -1495,13 +1480,13 @@ static int intel_pstate_set_policy(struct cpufreq_policy *policy) limits->max_sysfs_pct); limits->max_perf_pct = max(limits->min_policy_pct, limits->max_perf_pct); - limits->max_perf = round_up(limits->max_perf, FRAC_BITS); /* Make sure min_perf_pct <= max_perf_pct */ limits->min_perf_pct = min(limits->max_perf_pct, limits->min_perf_pct); limits->min_perf = div_fp(limits->min_perf_pct, 100); limits->max_perf = div_fp(limits->max_perf_pct, 100); + limits->max_perf = round_up(limits->max_perf, FRAC_BITS); out: intel_pstate_set_update_util_hook(policy->cpu); @@ -1558,8 +1543,11 @@ static int intel_pstate_cpu_init(struct cpufreq_policy *policy) /* cpuinfo and default policy values */ policy->cpuinfo.min_freq = cpu->pstate.min_pstate * cpu->pstate.scaling; - policy->cpuinfo.max_freq = - cpu->pstate.turbo_pstate * cpu->pstate.scaling; + update_turbo_state(); + policy->cpuinfo.max_freq = limits->turbo_disabled ? + cpu->pstate.max_pstate : cpu->pstate.turbo_pstate; + policy->cpuinfo.max_freq *= cpu->pstate.scaling; + intel_pstate_init_acpi_perf_limits(policy); policy->cpuinfo.transition_latency = CPUFREQ_ETERNAL; cpumask_set_cpu(policy->cpu, policy->cpus); diff --git a/drivers/cpufreq/pcc-cpufreq.c b/drivers/cpufreq/pcc-cpufreq.c index 808a320e9d5d..a7ecb9a84c15 100644 --- a/drivers/cpufreq/pcc-cpufreq.c +++ b/drivers/cpufreq/pcc-cpufreq.c @@ -487,7 +487,7 @@ static int __init pcc_cpufreq_probe(void) doorbell.space_id = reg_resource->space_id; doorbell.bit_width = reg_resource->bit_width; doorbell.bit_offset = reg_resource->bit_offset; - doorbell.access_width = 64; + doorbell.access_width = 4; doorbell.address = reg_resource->address; pr_debug("probe: doorbell: space_id is %d, bit_width is %d, " diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c index 1d6c803804d5..e92418facc92 100644 --- a/drivers/devfreq/devfreq.c +++ b/drivers/devfreq/devfreq.c @@ -268,8 +268,11 @@ int update_devfreq(struct devfreq *devfreq) devfreq_notify_transition(devfreq, &freqs, DEVFREQ_PRECHANGE); err = devfreq->profile->target(devfreq->dev.parent, &freq, flags); - if (err) + if (err) { + freqs.new = cur_freq; + devfreq_notify_transition(devfreq, &freqs, DEVFREQ_POSTCHANGE); return err; + } freqs.new = freq; devfreq_notify_transition(devfreq, &freqs, DEVFREQ_POSTCHANGE); @@ -552,6 +555,7 @@ struct devfreq *devfreq_add_device(struct device *dev, devfreq->profile = profile; strncpy(devfreq->governor_name, governor_name, DEVFREQ_NAME_LEN); devfreq->previous_freq = profile->initial_freq; + devfreq->last_status.current_frequency = profile->initial_freq; devfreq->data = data; devfreq->nb.notifier_call = devfreq_notifier_call; @@ -561,23 +565,22 @@ struct devfreq *devfreq_add_device(struct device *dev, mutex_lock(&devfreq->lock); } - devfreq->trans_table = devm_kzalloc(dev, sizeof(unsigned int) * - devfreq->profile->max_state * - devfreq->profile->max_state, - GFP_KERNEL); - devfreq->time_in_state = devm_kzalloc(dev, sizeof(unsigned long) * - devfreq->profile->max_state, - GFP_KERNEL); - devfreq->last_stat_updated = jiffies; - dev_set_name(&devfreq->dev, "%s", dev_name(dev)); err = device_register(&devfreq->dev); if (err) { - put_device(&devfreq->dev); mutex_unlock(&devfreq->lock); goto err_out; } + devfreq->trans_table = devm_kzalloc(&devfreq->dev, sizeof(unsigned int) * + devfreq->profile->max_state * + devfreq->profile->max_state, + GFP_KERNEL); + devfreq->time_in_state = devm_kzalloc(&devfreq->dev, sizeof(unsigned long) * + devfreq->profile->max_state, + GFP_KERNEL); + devfreq->last_stat_updated = jiffies; + srcu_init_notifier_head(&devfreq->transition_notifier_list); mutex_unlock(&devfreq->lock); @@ -603,7 +606,6 @@ struct devfreq *devfreq_add_device(struct device *dev, err_init: list_del(&devfreq->node); device_unregister(&devfreq->dev); - kfree(devfreq); err_out: return ERR_PTR(err); } @@ -621,7 +623,6 @@ int devfreq_remove_device(struct devfreq *devfreq) return -EINVAL; device_unregister(&devfreq->dev); - put_device(&devfreq->dev); return 0; } diff --git a/drivers/devfreq/event/exynos-nocp.c b/drivers/devfreq/event/exynos-nocp.c index 6b6a5f310486..a5841403bde8 100644 --- a/drivers/devfreq/event/exynos-nocp.c +++ b/drivers/devfreq/event/exynos-nocp.c @@ -220,9 +220,6 @@ static int exynos_nocp_parse_dt(struct platform_device *pdev, /* Maps the memory mapped IO to control nocp register */ res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - if (IS_ERR(res)) - return PTR_ERR(res); - base = devm_ioremap_resource(dev, res); if (IS_ERR(base)) return PTR_ERR(base); diff --git a/drivers/dma/at_xdmac.c b/drivers/dma/at_xdmac.c index 8e304b1befc5..75bd6621dc5d 100644 --- a/drivers/dma/at_xdmac.c +++ b/drivers/dma/at_xdmac.c @@ -242,7 +242,7 @@ struct at_xdmac_lld { u32 mbr_dus; /* Destination Microblock Stride Register */ }; - +/* 64-bit alignment needed to update CNDA and CUBC registers in an atomic way. */ struct at_xdmac_desc { struct at_xdmac_lld lld; enum dma_transfer_direction direction; @@ -253,7 +253,7 @@ struct at_xdmac_desc { unsigned int xfer_size; struct list_head descs_list; struct list_head xfer_node; -}; +} __aligned(sizeof(u64)); static inline void __iomem *at_xdmac_chan_reg_base(struct at_xdmac *atxdmac, unsigned int chan_nb) { @@ -1400,6 +1400,7 @@ at_xdmac_tx_status(struct dma_chan *chan, dma_cookie_t cookie, u32 cur_nda, check_nda, cur_ubc, mask, value; u8 dwidth = 0; unsigned long flags; + bool initd; ret = dma_cookie_status(chan, cookie, txstate); if (ret == DMA_COMPLETE) @@ -1424,7 +1425,16 @@ at_xdmac_tx_status(struct dma_chan *chan, dma_cookie_t cookie, residue = desc->xfer_size; /* * Flush FIFO: only relevant when the transfer is source peripheral - * synchronized. + * synchronized. Flush is needed before reading CUBC because data in + * the FIFO are not reported by CUBC. Reporting a residue of the + * transfer length while we have data in FIFO can cause issue. + * Usecase: atmel USART has a timeout which means I have received + * characters but there is no more character received for a while. On + * timeout, it requests the residue. If the data are in the DMA FIFO, + * we will return a residue of the transfer length. It means no data + * received. If an application is waiting for these data, it will hang + * since we won't have another USART timeout without receiving new + * data. */ mask = AT_XDMAC_CC_TYPE | AT_XDMAC_CC_DSYNC; value = AT_XDMAC_CC_TYPE_PER_TRAN | AT_XDMAC_CC_DSYNC_PER2MEM; @@ -1435,34 +1445,43 @@ at_xdmac_tx_status(struct dma_chan *chan, dma_cookie_t cookie, } /* - * When processing the residue, we need to read two registers but we - * can't do it in an atomic way. AT_XDMAC_CNDA is used to find where - * we stand in the descriptor list and AT_XDMAC_CUBC is used - * to know how many data are remaining for the current descriptor. - * Since the dma channel is not paused to not loose data, between the - * AT_XDMAC_CNDA and AT_XDMAC_CUBC read, we may have change of - * descriptor. - * For that reason, after reading AT_XDMAC_CUBC, we check if we are - * still using the same descriptor by reading a second time - * AT_XDMAC_CNDA. If AT_XDMAC_CNDA has changed, it means we have to - * read again AT_XDMAC_CUBC. + * The easiest way to compute the residue should be to pause the DMA + * but doing this can lead to miss some data as some devices don't + * have FIFO. + * We need to read several registers because: + * - DMA is running therefore a descriptor change is possible while + * reading these registers + * - When the block transfer is done, the value of the CUBC register + * is set to its initial value until the fetch of the next descriptor. + * This value will corrupt the residue calculation so we have to skip + * it. + * + * INITD -------- ------------ + * |____________________| + * _______________________ _______________ + * NDA @desc2 \/ @desc3 + * _______________________/\_______________ + * __________ ___________ _______________ + * CUBC 0 \/ MAX desc1 \/ MAX desc2 + * __________/\___________/\_______________ + * + * Since descriptors are aligned on 64 bits, we can assume that + * the update of NDA and CUBC is atomic. * Memory barriers are used to ensure the read order of the registers. - * A max number of retries is set because unlikely it can never ends if - * we are transferring a lot of data with small buffers. + * A max number of retries is set because unlikely it could never ends. */ - cur_nda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA) & 0xfffffffc; - rmb(); - cur_ubc = at_xdmac_chan_read(atchan, AT_XDMAC_CUBC); for (retry = 0; retry < AT_XDMAC_RESIDUE_MAX_RETRIES; retry++) { - rmb(); check_nda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA) & 0xfffffffc; - - if (likely(cur_nda == check_nda)) - break; - - cur_nda = check_nda; + rmb(); + initd = !!(at_xdmac_chan_read(atchan, AT_XDMAC_CC) & AT_XDMAC_CC_INITD); rmb(); cur_ubc = at_xdmac_chan_read(atchan, AT_XDMAC_CUBC); + rmb(); + cur_nda = at_xdmac_chan_read(atchan, AT_XDMAC_CNDA) & 0xfffffffc; + rmb(); + + if ((check_nda == cur_nda) && initd) + break; } if (unlikely(retry >= AT_XDMAC_RESIDUE_MAX_RETRIES)) { @@ -1471,6 +1490,19 @@ at_xdmac_tx_status(struct dma_chan *chan, dma_cookie_t cookie, } /* + * Flush FIFO: only relevant when the transfer is source peripheral + * synchronized. Another flush is needed here because CUBC is updated + * when the controller sends the data write command. It can lead to + * report data that are not written in the memory or the device. The + * FIFO flush ensures that data are really written. + */ + if ((desc->lld.mbr_cfg & mask) == value) { + at_xdmac_write(atxdmac, AT_XDMAC_GSWF, atchan->mask); + while (!(at_xdmac_chan_read(atchan, AT_XDMAC_CIS) & AT_XDMAC_CIS_FIS)) + cpu_relax(); + } + + /* * Remove size of all microblocks already transferred and the current * one. Then add the remaining size to transfer of the current * microblock. diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c index 25d1dadcddd1..d0446a75990a 100644 --- a/drivers/dma/mv_xor.c +++ b/drivers/dma/mv_xor.c @@ -703,8 +703,9 @@ static int mv_chan_memcpy_self_test(struct mv_xor_chan *mv_chan) goto free_resources; } - src_dma = dma_map_page(dma_chan->device->dev, virt_to_page(src), 0, - PAGE_SIZE, DMA_TO_DEVICE); + src_dma = dma_map_page(dma_chan->device->dev, virt_to_page(src), + (size_t)src & ~PAGE_MASK, PAGE_SIZE, + DMA_TO_DEVICE); unmap->addr[0] = src_dma; ret = dma_mapping_error(dma_chan->device->dev, src_dma); @@ -714,8 +715,9 @@ static int mv_chan_memcpy_self_test(struct mv_xor_chan *mv_chan) } unmap->to_cnt = 1; - dest_dma = dma_map_page(dma_chan->device->dev, virt_to_page(dest), 0, - PAGE_SIZE, DMA_FROM_DEVICE); + dest_dma = dma_map_page(dma_chan->device->dev, virt_to_page(dest), + (size_t)dest & ~PAGE_MASK, PAGE_SIZE, + DMA_FROM_DEVICE); unmap->addr[1] = dest_dma; ret = dma_mapping_error(dma_chan->device->dev, dest_dma); diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c index 6aa256b0a1ed..c3ee3ad98a63 100644 --- a/drivers/edac/edac_mc.c +++ b/drivers/edac/edac_mc.c @@ -565,7 +565,8 @@ void edac_mc_reset_delay_period(unsigned long value) list_for_each(item, &mc_devices) { mci = list_entry(item, struct mem_ctl_info, link); - edac_mod_work(&mci->work, value); + if (mci->op_state == OP_RUNNING_POLL) + edac_mod_work(&mci->work, value); } mutex_unlock(&mem_ctls_mutex); } diff --git a/drivers/edac/sb_edac.c b/drivers/edac/sb_edac.c index b4d0bf6534cf..6744d88bdea8 100644 --- a/drivers/edac/sb_edac.c +++ b/drivers/edac/sb_edac.c @@ -239,8 +239,11 @@ static const u32 rir_offset[MAX_RIR_RANGES][MAX_RIR_WAY] = { { 0x1a0, 0x1a4, 0x1a8, 0x1ac, 0x1b0, 0x1b4, 0x1b8, 0x1bc }, }; -#define RIR_RNK_TGT(reg) GET_BITFIELD(reg, 16, 19) -#define RIR_OFFSET(reg) GET_BITFIELD(reg, 2, 14) +#define RIR_RNK_TGT(type, reg) (((type) == BROADWELL) ? \ + GET_BITFIELD(reg, 20, 23) : GET_BITFIELD(reg, 16, 19)) + +#define RIR_OFFSET(type, reg) (((type) == HASWELL || (type) == BROADWELL) ? \ + GET_BITFIELD(reg, 2, 15) : GET_BITFIELD(reg, 2, 14)) /* Device 16, functions 2-7 */ @@ -326,6 +329,7 @@ struct pci_id_descr { struct pci_id_table { const struct pci_id_descr *descr; int n_devs; + enum type type; }; struct sbridge_dev { @@ -394,9 +398,14 @@ static const struct pci_id_descr pci_dev_descr_sbridge[] = { { PCI_DESCR(PCI_DEVICE_ID_INTEL_SBRIDGE_BR, 0) }, }; -#define PCI_ID_TABLE_ENTRY(A) { .descr=A, .n_devs = ARRAY_SIZE(A) } +#define PCI_ID_TABLE_ENTRY(A, T) { \ + .descr = A, \ + .n_devs = ARRAY_SIZE(A), \ + .type = T \ +} + static const struct pci_id_table pci_dev_descr_sbridge_table[] = { - PCI_ID_TABLE_ENTRY(pci_dev_descr_sbridge), + PCI_ID_TABLE_ENTRY(pci_dev_descr_sbridge, SANDY_BRIDGE), {0,} /* 0 terminated list. */ }; @@ -463,7 +472,7 @@ static const struct pci_id_descr pci_dev_descr_ibridge[] = { }; static const struct pci_id_table pci_dev_descr_ibridge_table[] = { - PCI_ID_TABLE_ENTRY(pci_dev_descr_ibridge), + PCI_ID_TABLE_ENTRY(pci_dev_descr_ibridge, IVY_BRIDGE), {0,} /* 0 terminated list. */ }; @@ -536,7 +545,7 @@ static const struct pci_id_descr pci_dev_descr_haswell[] = { }; static const struct pci_id_table pci_dev_descr_haswell_table[] = { - PCI_ID_TABLE_ENTRY(pci_dev_descr_haswell), + PCI_ID_TABLE_ENTRY(pci_dev_descr_haswell, HASWELL), {0,} /* 0 terminated list. */ }; @@ -580,7 +589,7 @@ static const struct pci_id_descr pci_dev_descr_knl[] = { }; static const struct pci_id_table pci_dev_descr_knl_table[] = { - PCI_ID_TABLE_ENTRY(pci_dev_descr_knl), + PCI_ID_TABLE_ENTRY(pci_dev_descr_knl, KNIGHTS_LANDING), {0,} }; @@ -648,7 +657,7 @@ static const struct pci_id_descr pci_dev_descr_broadwell[] = { }; static const struct pci_id_table pci_dev_descr_broadwell_table[] = { - PCI_ID_TABLE_ENTRY(pci_dev_descr_broadwell), + PCI_ID_TABLE_ENTRY(pci_dev_descr_broadwell, BROADWELL), {0,} /* 0 terminated list. */ }; @@ -1894,14 +1903,14 @@ static void get_memory_layout(const struct mem_ctl_info *mci) pci_read_config_dword(pvt->pci_tad[i], rir_offset[j][k], ®); - tmp_mb = RIR_OFFSET(reg) << 6; + tmp_mb = RIR_OFFSET(pvt->info.type, reg) << 6; gb = div_u64_rem(tmp_mb, 1024, &mb); edac_dbg(0, "CH#%d RIR#%d INTL#%d, offset %u.%03u GB (0x%016Lx), tgt: %d, reg=0x%08x\n", i, j, k, gb, (mb*1000)/1024, ((u64)tmp_mb) << 20L, - (u32)RIR_RNK_TGT(reg), + (u32)RIR_RNK_TGT(pvt->info.type, reg), reg); } } @@ -2234,7 +2243,7 @@ static int get_memory_error_data(struct mem_ctl_info *mci, pci_read_config_dword(pvt->pci_tad[ch_add + base_ch], rir_offset[n_rir][idx], ®); - *rank = RIR_RNK_TGT(reg); + *rank = RIR_RNK_TGT(pvt->info.type, reg); edac_dbg(0, "RIR#%d: channel address 0x%08Lx < 0x%08Lx, RIR interleave %d, index %d\n", n_rir, @@ -3357,12 +3366,12 @@ fail0: #define ICPU(model, table) \ { X86_VENDOR_INTEL, 6, model, 0, (unsigned long)&table } -/* Order here must match "enum type" */ static const struct x86_cpu_id sbridge_cpuids[] = { ICPU(0x2d, pci_dev_descr_sbridge_table), /* SANDY_BRIDGE */ ICPU(0x3e, pci_dev_descr_ibridge_table), /* IVY_BRIDGE */ ICPU(0x3f, pci_dev_descr_haswell_table), /* HASWELL */ ICPU(0x4f, pci_dev_descr_broadwell_table), /* BROADWELL */ + ICPU(0x56, pci_dev_descr_broadwell_table), /* BROADWELL-DE */ ICPU(0x57, pci_dev_descr_knl_table), /* KNIGHTS_LANDING */ { } }; @@ -3398,7 +3407,7 @@ static int sbridge_probe(const struct x86_cpu_id *id) mc, mc + 1, num_mc); sbridge_dev->mc = mc++; - rc = sbridge_register_mci(sbridge_dev, id - sbridge_cpuids); + rc = sbridge_register_mci(sbridge_dev, ptable->type); if (unlikely(rc < 0)) goto fail1; } diff --git a/drivers/extcon/extcon-palmas.c b/drivers/extcon/extcon-palmas.c index 8b3226dca1d9..caff46c0e214 100644 --- a/drivers/extcon/extcon-palmas.c +++ b/drivers/extcon/extcon-palmas.c @@ -360,6 +360,8 @@ static int palmas_usb_probe(struct platform_device *pdev) palmas_enable_irq(palmas_usb); /* perform initial detection */ + if (palmas_usb->enable_gpio_vbus_detection) + palmas_vbus_irq_handler(palmas_usb->gpio_vbus_irq, palmas_usb); palmas_gpio_id_detect(&palmas_usb->wq_detectid.work); device_set_wakeup_capable(&pdev->dev, true); return 0; diff --git a/drivers/firmware/efi/arm-init.c b/drivers/firmware/efi/arm-init.c index a850cbc48d8d..c49d50e68aee 100644 --- a/drivers/firmware/efi/arm-init.c +++ b/drivers/firmware/efi/arm-init.c @@ -174,6 +174,7 @@ static __init void reserve_regions(void) { efi_memory_desc_t *md; u64 paddr, npages, size; + int resv; if (efi_enabled(EFI_DBG)) pr_info("Processing EFI memory map:\n"); @@ -190,12 +191,14 @@ static __init void reserve_regions(void) paddr = md->phys_addr; npages = md->num_pages; + resv = is_reserve_region(md); if (efi_enabled(EFI_DBG)) { char buf[64]; - pr_info(" 0x%012llx-0x%012llx %s", + pr_info(" 0x%012llx-0x%012llx %s%s\n", paddr, paddr + (npages << EFI_PAGE_SHIFT) - 1, - efi_md_typeattr_format(buf, sizeof(buf), md)); + efi_md_typeattr_format(buf, sizeof(buf), md), + resv ? "*" : ""); } memrange_efi_to_native(&paddr, &npages); @@ -204,14 +207,9 @@ static __init void reserve_regions(void) if (is_normal_ram(md)) early_init_dt_add_memory_arch(paddr, size); - if (is_reserve_region(md)) { + if (resv) memblock_mark_nomap(paddr, size); - if (efi_enabled(EFI_DBG)) - pr_cont("*"); - } - if (efi_enabled(EFI_DBG)) - pr_cont("\n"); } set_bit(EFI_MEMMAP, &efi.flags); diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index 48da857f4774..cebcb405812e 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -33,6 +33,7 @@ config ARCH_REQUIRE_GPIOLIB menuconfig GPIOLIB bool "GPIO Support" + select ANON_INODES help This enables GPIO support through the generic GPIO library. You only need to enable this, if you also want to enable @@ -530,7 +531,7 @@ menu "Port-mapped I/O GPIO drivers" config GPIO_104_DIO_48E tristate "ACCES 104-DIO-48E GPIO support" - depends on ISA + depends on ISA_BUS_API select GPIOLIB_IRQCHIP help Enables GPIO support for the ACCES 104-DIO-48E series (104-DIO-48E, @@ -540,7 +541,7 @@ config GPIO_104_DIO_48E config GPIO_104_IDIO_16 tristate "ACCES 104-IDIO-16 GPIO support" - depends on ISA + depends on ISA_BUS_API select GPIOLIB_IRQCHIP help Enables GPIO support for the ACCES 104-IDIO-16 family (104-IDIO-16, @@ -551,7 +552,7 @@ config GPIO_104_IDIO_16 config GPIO_104_IDI_48 tristate "ACCES 104-IDI-48 GPIO support" - depends on ISA + depends on ISA_BUS_API select GPIOLIB_IRQCHIP help Enables GPIO support for the ACCES 104-IDI-48 family (104-IDI-48A, @@ -627,7 +628,7 @@ config GPIO_TS5500 config GPIO_WS16C48 tristate "WinSystems WS16C48 GPIO support" - depends on ISA + depends on ISA_BUS_API select GPIOLIB_IRQCHIP help Enables GPIO support for the WinSystems WS16C48. The base port diff --git a/drivers/gpio/gpio-104-dio-48e.c b/drivers/gpio/gpio-104-dio-48e.c index 1a647c07be67..fcf776971ca9 100644 --- a/drivers/gpio/gpio-104-dio-48e.c +++ b/drivers/gpio/gpio-104-dio-48e.c @@ -75,7 +75,7 @@ static int dio48e_gpio_direction_input(struct gpio_chip *chip, unsigned offset) { struct dio48e_gpio *const dio48egpio = gpiochip_get_data(chip); const unsigned io_port = offset / 8; - const unsigned control_port = io_port / 2; + const unsigned int control_port = io_port / 3; const unsigned control_addr = dio48egpio->base + 3 + control_port*4; unsigned long flags; unsigned control; @@ -115,7 +115,7 @@ static int dio48e_gpio_direction_output(struct gpio_chip *chip, unsigned offset, { struct dio48e_gpio *const dio48egpio = gpiochip_get_data(chip); const unsigned io_port = offset / 8; - const unsigned control_port = io_port / 2; + const unsigned int control_port = io_port / 3; const unsigned mask = BIT(offset % 8); const unsigned control_addr = dio48egpio->base + 3 + control_port*4; const unsigned out_port = (io_port > 2) ? io_port + 1 : io_port; diff --git a/drivers/gpio/gpio-104-idi-48.c b/drivers/gpio/gpio-104-idi-48.c index 6c75c83baf5a..2d2763ea1a68 100644 --- a/drivers/gpio/gpio-104-idi-48.c +++ b/drivers/gpio/gpio-104-idi-48.c @@ -247,6 +247,7 @@ static int idi_48_probe(struct device *dev, unsigned int id) idi48gpio->irq = irq[id]; spin_lock_init(&idi48gpio->lock); + spin_lock_init(&idi48gpio->ack_lock); dev_set_drvdata(dev, idi48gpio); diff --git a/drivers/gpio/gpio-bcm-kona.c b/drivers/gpio/gpio-bcm-kona.c index 9aabc48ff5de..953e4b829e32 100644 --- a/drivers/gpio/gpio-bcm-kona.c +++ b/drivers/gpio/gpio-bcm-kona.c @@ -547,11 +547,11 @@ static void bcm_kona_gpio_reset(struct bcm_kona_gpio *kona_gpio) /* disable interrupts and clear status */ for (i = 0; i < kona_gpio->num_bank; i++) { /* Unlock the entire bank first */ - bcm_kona_gpio_write_lock_regs(kona_gpio, i, UNLOCK_CODE); + bcm_kona_gpio_write_lock_regs(reg_base, i, UNLOCK_CODE); writel(0xffffffff, reg_base + GPIO_INT_MASK(i)); writel(0xffffffff, reg_base + GPIO_INT_STATUS(i)); /* Now re-lock the bank */ - bcm_kona_gpio_write_lock_regs(kona_gpio, i, LOCK_CODE); + bcm_kona_gpio_write_lock_regs(reg_base, i, LOCK_CODE); } } diff --git a/drivers/gpio/gpio-zynq.c b/drivers/gpio/gpio-zynq.c index 75c6355b018d..e72794e463aa 100644 --- a/drivers/gpio/gpio-zynq.c +++ b/drivers/gpio/gpio-zynq.c @@ -709,7 +709,13 @@ static int zynq_gpio_probe(struct platform_device *pdev) dev_err(&pdev->dev, "input clock not found.\n"); return PTR_ERR(gpio->clk); } + ret = clk_prepare_enable(gpio->clk); + if (ret) { + dev_err(&pdev->dev, "Unable to enable clock.\n"); + return ret; + } + pm_runtime_set_active(&pdev->dev); pm_runtime_enable(&pdev->dev); ret = pm_runtime_get_sync(&pdev->dev); if (ret < 0) @@ -747,6 +753,7 @@ err_pm_put: pm_runtime_put(&pdev->dev); err_pm_dis: pm_runtime_disable(&pdev->dev); + clk_disable_unprepare(gpio->clk); return ret; } diff --git a/drivers/gpio/gpiolib-of.c b/drivers/gpio/gpiolib-of.c index d22dcc38179d..4aabddb38b59 100644 --- a/drivers/gpio/gpiolib-of.c +++ b/drivers/gpio/gpiolib-of.c @@ -16,6 +16,7 @@ #include <linux/errno.h> #include <linux/module.h> #include <linux/io.h> +#include <linux/io-mapping.h> #include <linux/gpio/consumer.h> #include <linux/of.h> #include <linux/of_address.h> diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c index 24f60d28f0c0..570771ed19e6 100644 --- a/drivers/gpio/gpiolib.c +++ b/drivers/gpio/gpiolib.c @@ -449,7 +449,6 @@ static void gpiodevice_release(struct device *dev) { struct gpio_device *gdev = dev_get_drvdata(dev); - cdev_del(&gdev->chrdev); list_del(&gdev->list); ida_simple_remove(&gpio_ida, gdev->id); kfree(gdev->label); @@ -482,7 +481,6 @@ static int gpiochip_setup_dev(struct gpio_device *gdev) /* From this point, the .release() function cleans up gpio_device */ gdev->dev.release = gpiodevice_release; - get_device(&gdev->dev); pr_debug("%s: registered GPIOs %d to %d on device: %s (%s)\n", __func__, gdev->base, gdev->base + gdev->ngpio - 1, dev_name(&gdev->dev), gdev->chip->label ? : "generic"); @@ -770,6 +768,8 @@ void gpiochip_remove(struct gpio_chip *chip) * be removed, else it will be dangling until the last user is * gone. */ + cdev_del(&gdev->chrdev); + device_del(&gdev->dev); put_device(&gdev->dev); } EXPORT_SYMBOL_GPL(gpiochip_remove); @@ -869,7 +869,7 @@ struct gpio_chip *gpiochip_find(void *data, spin_lock_irqsave(&gpio_lock, flags); list_for_each_entry(gdev, &gpio_devices, list) - if (match(gdev->chip, data)) + if (gdev->chip && match(gdev->chip, data)) break; /* No match? */ @@ -1373,8 +1373,12 @@ done: #define VALIDATE_DESC(desc) do { \ if (!desc) \ return 0; \ + if (IS_ERR(desc)) { \ + pr_warn("%s: invalid GPIO (errorpointer)\n", __func__); \ + return PTR_ERR(desc); \ + } \ if (!desc->gdev) { \ - pr_warn("%s: invalid GPIO\n", __func__); \ + pr_warn("%s: invalid GPIO (no device)\n", __func__); \ return -EINVAL; \ } \ if ( !desc->gdev->chip ) { \ @@ -1386,8 +1390,12 @@ done: #define VALIDATE_DESC_VOID(desc) do { \ if (!desc) \ return; \ + if (IS_ERR(desc)) { \ + pr_warn("%s: invalid GPIO (errorpointer)\n", __func__); \ + return; \ + } \ if (!desc->gdev) { \ - pr_warn("%s: invalid GPIO\n", __func__); \ + pr_warn("%s: invalid GPIO (no device)\n", __func__); \ return; \ } \ if (!desc->gdev->chip) { \ @@ -2056,7 +2064,14 @@ int gpiod_to_irq(const struct gpio_desc *desc) struct gpio_chip *chip; int offset; - VALIDATE_DESC(desc); + /* + * Cannot VALIDATE_DESC() here as gpiod_to_irq() consumer semantics + * requires this function to not return zero on an invalid descriptor + * but rather a negative error number. + */ + if (!desc || IS_ERR(desc) || !desc->gdev || !desc->gdev->chip) + return -EINVAL; + chip = desc->gdev->chip; offset = gpio_chip_hwgpio(desc); if (chip->to_irq) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index da3d02154fa6..eb09037a7161 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -85,8 +85,12 @@ extern int amdgpu_vm_debug; extern int amdgpu_sched_jobs; extern int amdgpu_sched_hw_submission; extern int amdgpu_powerplay; +extern int amdgpu_powercontainment; extern unsigned amdgpu_pcie_gen_cap; extern unsigned amdgpu_pcie_lane_cap; +extern unsigned amdgpu_cg_mask; +extern unsigned amdgpu_pg_mask; +extern char *amdgpu_disable_cu; #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS 3000 #define AMDGPU_MAX_USEC_TIMEOUT 100000 /* 100 ms */ @@ -183,6 +187,10 @@ int amdgpu_set_clockgating_state(struct amdgpu_device *adev, int amdgpu_set_powergating_state(struct amdgpu_device *adev, enum amd_ip_block_type block_type, enum amd_powergating_state state); +int amdgpu_wait_for_idle(struct amdgpu_device *adev, + enum amd_ip_block_type block_type); +bool amdgpu_is_idle(struct amdgpu_device *adev, + enum amd_ip_block_type block_type); struct amdgpu_ip_block_version { enum amd_ip_block_type type; @@ -594,11 +602,9 @@ int amdgpu_sync_resv(struct amdgpu_device *adev, struct amdgpu_sync *sync, struct reservation_object *resv, void *owner); -bool amdgpu_sync_is_idle(struct amdgpu_sync *sync); -int amdgpu_sync_cycle_fences(struct amdgpu_sync *dst, struct amdgpu_sync *src, - struct fence *fence); +struct fence *amdgpu_sync_peek_fence(struct amdgpu_sync *sync, + struct amdgpu_ring *ring); struct fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync); -int amdgpu_sync_wait(struct amdgpu_sync *sync); void amdgpu_sync_free(struct amdgpu_sync *sync); int amdgpu_sync_init(void); void amdgpu_sync_fini(void); @@ -754,12 +760,11 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, struct amdgpu_job **job); +void amdgpu_job_free_resources(struct amdgpu_job *job); void amdgpu_job_free(struct amdgpu_job *job); -void amdgpu_job_free_func(struct kref *refcount); int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring, struct amd_sched_entity *entity, void *owner, struct fence **f); -void amdgpu_job_timeout_func(struct work_struct *work); struct amdgpu_ring { struct amdgpu_device *adev; @@ -771,8 +776,6 @@ struct amdgpu_ring { struct amdgpu_bo *ring_obj; volatile uint32_t *ring; unsigned rptr_offs; - u64 next_rptr_gpu_addr; - volatile u32 *next_rptr_cpu_addr; unsigned wptr; unsigned wptr_old; unsigned ring_size; @@ -791,7 +794,6 @@ struct amdgpu_ring { u32 doorbell_index; bool use_doorbell; unsigned wptr_offs; - unsigned next_rptr_offs; unsigned fence_offs; uint64_t current_ctx; enum amdgpu_ring_type type; @@ -799,6 +801,9 @@ struct amdgpu_ring { unsigned cond_exe_offs; u64 cond_exe_gpu_addr; volatile u32 *cond_exe_cpu_addr; +#if defined(CONFIG_DEBUG_FS) + struct dentry *ent; +#endif }; /* @@ -861,6 +866,7 @@ struct amdgpu_vm { struct amdgpu_bo *page_directory; unsigned max_pde_used; struct fence *page_directory_fence; + uint64_t last_eviction_counter; /* array of page tables, one for each page directory entry */ struct amdgpu_vm_pt *page_tables; @@ -883,13 +889,14 @@ struct amdgpu_vm_id { struct fence *first; struct amdgpu_sync active; struct fence *last_flush; - struct amdgpu_ring *last_user; atomic64_t owner; uint64_t pd_gpu_addr; /* last flushed PD/PT update */ struct fence *flushed_updates; + uint32_t current_gpu_reset_count; + uint32_t gds_base; uint32_t gds_size; uint32_t gws_base; @@ -905,6 +912,10 @@ struct amdgpu_vm_manager { struct list_head ids_lru; struct amdgpu_vm_id ids[AMDGPU_NUM_VM]; + /* Handling of VM fences */ + u64 fence_context; + unsigned seqno[AMDGPU_MAX_RINGS]; + uint32_t max_pfn; /* vram base address for page table entry */ u64 vram_base_offset; @@ -926,17 +937,14 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm, struct list_head *validated, struct amdgpu_bo_list_entry *entry); -void amdgpu_vm_get_pt_bos(struct amdgpu_vm *vm, struct list_head *duplicates); +void amdgpu_vm_get_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct list_head *duplicates); void amdgpu_vm_move_pt_bos_in_lru(struct amdgpu_device *adev, struct amdgpu_vm *vm); int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_sync *sync, struct fence *fence, - unsigned *vm_id, uint64_t *vm_pd_addr); -int amdgpu_vm_flush(struct amdgpu_ring *ring, - unsigned vm_id, uint64_t pd_addr, - uint32_t gds_base, uint32_t gds_size, - uint32_t gws_base, uint32_t gws_size, - uint32_t oa_base, uint32_t oa_size); + struct amdgpu_job *job); +int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job); void amdgpu_vm_reset_id(struct amdgpu_device *adev, unsigned vm_id); uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr); int amdgpu_vm_update_page_directory(struct amdgpu_device *adev, @@ -1142,6 +1150,12 @@ struct amdgpu_cu_info { uint32_t bitmap[4][4]; }; +struct amdgpu_gfx_funcs { + /* get the gpu clock counter */ + uint64_t (*get_gpu_clock_counter)(struct amdgpu_device *adev); + void (*select_se_sh)(struct amdgpu_device *adev, u32 se_num, u32 sh_num, u32 instance); +}; + struct amdgpu_gfx { struct mutex gpu_clock_mutex; struct amdgpu_gca_config config; @@ -1178,6 +1192,7 @@ struct amdgpu_gfx { /* ce ram size*/ unsigned ce_ram_size; struct amdgpu_cu_info cu_info; + const struct amdgpu_gfx_funcs *funcs; }; int amdgpu_ib_get(struct amdgpu_device *adev, struct amdgpu_vm *vm, @@ -1195,10 +1210,6 @@ void amdgpu_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count); void amdgpu_ring_generic_pad_ib(struct amdgpu_ring *ring, struct amdgpu_ib *ib); void amdgpu_ring_commit(struct amdgpu_ring *ring); void amdgpu_ring_undo(struct amdgpu_ring *ring); -unsigned amdgpu_ring_backup(struct amdgpu_ring *ring, - uint32_t **data); -int amdgpu_ring_restore(struct amdgpu_ring *ring, - unsigned size, uint32_t *data); int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, unsigned ring_size, u32 nop, u32 align_mask, struct amdgpu_irq_src *irq_src, unsigned irq_type, @@ -1250,6 +1261,7 @@ struct amdgpu_job { uint32_t num_ibs; void *owner; uint64_t ctx; + bool vm_needs_flush; unsigned vm_id; uint64_t vm_pd_addr; uint32_t gds_base, gds_size; @@ -1257,8 +1269,7 @@ struct amdgpu_job { uint32_t oa_base, oa_size; /* user fence handling */ - struct amdgpu_bo *uf_bo; - uint32_t uf_offset; + uint64_t uf_addr; uint64_t uf_sequence; }; @@ -1560,6 +1571,12 @@ struct amdgpu_dpm_funcs { u32 (*get_fan_control_mode)(struct amdgpu_device *adev); int (*set_fan_speed_percent)(struct amdgpu_device *adev, u32 speed); int (*get_fan_speed_percent)(struct amdgpu_device *adev, u32 *speed); + int (*force_clock_level)(struct amdgpu_device *adev, enum pp_clock_type type, uint32_t mask); + int (*print_clock_levels)(struct amdgpu_device *adev, enum pp_clock_type type, char *buf); + int (*get_sclk_od)(struct amdgpu_device *adev); + int (*set_sclk_od)(struct amdgpu_device *adev, uint32_t value); + int (*get_mclk_od)(struct amdgpu_device *adev); + int (*set_mclk_od)(struct amdgpu_device *adev, uint32_t value); }; struct amdgpu_dpm { @@ -1767,6 +1784,8 @@ int amdgpu_debugfs_init(struct drm_minor *minor); void amdgpu_debugfs_cleanup(struct drm_minor *minor); #endif +int amdgpu_debugfs_firmware_init(struct amdgpu_device *adev); + /* * amdgpu smumgr functions */ @@ -1811,15 +1830,13 @@ struct amdgpu_asic_funcs { u32 sh_num, u32 reg_offset, u32 *value); void (*set_vga_state)(struct amdgpu_device *adev, bool state); int (*reset)(struct amdgpu_device *adev); - /* wait for mc_idle */ - int (*wait_for_mc_idle)(struct amdgpu_device *adev); /* get the reference clock */ u32 (*get_xclk)(struct amdgpu_device *adev); - /* get the gpu clock counter */ - uint64_t (*get_gpu_clock_counter)(struct amdgpu_device *adev); /* MM block clocks */ int (*set_uvd_clocks)(struct amdgpu_device *adev, u32 vclk, u32 dclk); int (*set_vce_clocks)(struct amdgpu_device *adev, u32 evclk, u32 ecclk); + /* query virtual capabilities */ + u32 (*get_virtual_caps)(struct amdgpu_device *adev); }; /* @@ -1914,8 +1931,12 @@ void amdgpu_cgs_destroy_device(struct cgs_device *cgs_device); /* GPU virtualization */ +#define AMDGPU_VIRT_CAPS_SRIOV_EN (1 << 0) +#define AMDGPU_VIRT_CAPS_IS_VF (1 << 1) struct amdgpu_virtualization { bool supports_sr_iov; + bool is_virtual; + u32 caps; }; /* @@ -1997,6 +2018,10 @@ struct amdgpu_device { spinlock_t didt_idx_lock; amdgpu_rreg_t didt_rreg; amdgpu_wreg_t didt_wreg; + /* protects concurrent gc_cac register access */ + spinlock_t gc_cac_idx_lock; + amdgpu_rreg_t gc_cac_rreg; + amdgpu_wreg_t gc_cac_wreg; /* protects concurrent ENDPOINT (audio) register access */ spinlock_t audio_endpt_idx_lock; amdgpu_block_rreg_t audio_endpt_rreg; @@ -2022,6 +2047,7 @@ struct amdgpu_device { atomic64_t vram_vis_usage; atomic64_t gtt_usage; atomic64_t num_bytes_moved; + atomic64_t num_evictions; atomic_t gpu_reset_counter; /* display */ @@ -2125,6 +2151,8 @@ void amdgpu_mm_wdoorbell(struct amdgpu_device *adev, u32 index, u32 v); #define WREG32_UVD_CTX(reg, v) adev->uvd_ctx_wreg(adev, (reg), (v)) #define RREG32_DIDT(reg) adev->didt_rreg(adev, (reg)) #define WREG32_DIDT(reg, v) adev->didt_wreg(adev, (reg), (v)) +#define RREG32_GC_CAC(reg) adev->gc_cac_rreg(adev, (reg)) +#define WREG32_GC_CAC(reg, v) adev->gc_cac_wreg(adev, (reg), (v)) #define RREG32_AUDIO_ENDPT(block, reg) adev->audio_endpt_rreg(adev, (block), (reg)) #define WREG32_AUDIO_ENDPT(block, reg, v) adev->audio_endpt_wreg(adev, (block), (reg), (v)) #define WREG32_P(reg, val, mask) \ @@ -2200,11 +2228,10 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) */ #define amdgpu_asic_set_vga_state(adev, state) (adev)->asic_funcs->set_vga_state((adev), (state)) #define amdgpu_asic_reset(adev) (adev)->asic_funcs->reset((adev)) -#define amdgpu_asic_wait_for_mc_idle(adev) (adev)->asic_funcs->wait_for_mc_idle((adev)) #define amdgpu_asic_get_xclk(adev) (adev)->asic_funcs->get_xclk((adev)) #define amdgpu_asic_set_uvd_clocks(adev, v, d) (adev)->asic_funcs->set_uvd_clocks((adev), (v), (d)) #define amdgpu_asic_set_vce_clocks(adev, ev, ec) (adev)->asic_funcs->set_vce_clocks((adev), (ev), (ec)) -#define amdgpu_asic_get_gpu_clock_counter(adev) (adev)->asic_funcs->get_gpu_clock_counter((adev)) +#define amdgpu_asic_get_virtual_caps(adev) ((adev)->asic_funcs->get_virtual_caps((adev))) #define amdgpu_asic_read_disabled_bios(adev) (adev)->asic_funcs->read_disabled_bios((adev)) #define amdgpu_asic_read_bios_from_rom(adev, b, l) (adev)->asic_funcs->read_bios_from_rom((adev), (b), (l)) #define amdgpu_asic_read_register(adev, se, sh, offset, v)((adev)->asic_funcs->read_register((adev), (se), (sh), (offset), (v))) @@ -2257,6 +2284,8 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_dpm_print_power_state(adev, ps) (adev)->pm.funcs->print_power_state((adev), (ps)) #define amdgpu_dpm_vblank_too_short(adev) (adev)->pm.funcs->vblank_too_short((adev)) #define amdgpu_dpm_enable_bapm(adev, e) (adev)->pm.funcs->enable_bapm((adev), (e)) +#define amdgpu_gfx_get_gpu_clock_counter(adev) (adev)->gfx.funcs->get_gpu_clock_counter((adev)) +#define amdgpu_gfx_select_se_sh(adev, se, sh, instance) (adev)->gfx.funcs->select_se_sh((adev), (se), (sh), (instance)) #define amdgpu_dpm_get_temperature(adev) \ ((adev)->pp_enabled ? \ @@ -2335,6 +2364,18 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring) #define amdgpu_dpm_force_clock_level(adev, type, level) \ (adev)->powerplay.pp_funcs->force_clock_level((adev)->powerplay.pp_handle, type, level) +#define amdgpu_dpm_get_sclk_od(adev) \ + (adev)->powerplay.pp_funcs->get_sclk_od((adev)->powerplay.pp_handle) + +#define amdgpu_dpm_set_sclk_od(adev, value) \ + (adev)->powerplay.pp_funcs->set_sclk_od((adev)->powerplay.pp_handle, value) + +#define amdgpu_dpm_get_mclk_od(adev) \ + ((adev)->powerplay.pp_funcs->get_mclk_od((adev)->powerplay.pp_handle)) + +#define amdgpu_dpm_set_mclk_od(adev, value) \ + ((adev)->powerplay.pp_funcs->set_mclk_od((adev)->powerplay.pp_handle, value)) + #define amdgpu_dpm_dispatch_task(adev, event_id, input, output) \ (adev)->powerplay.pp_funcs->dispatch_tasks((adev)->powerplay.pp_handle, (event_id), (input), (output)) @@ -2376,9 +2417,13 @@ bool amdgpu_device_is_px(struct drm_device *dev); #if defined(CONFIG_VGA_SWITCHEROO) void amdgpu_register_atpx_handler(void); void amdgpu_unregister_atpx_handler(void); +bool amdgpu_has_atpx_dgpu_power_cntl(void); +bool amdgpu_is_atpx_hybrid(void); #else static inline void amdgpu_register_atpx_handler(void) {} static inline void amdgpu_unregister_atpx_handler(void) {} +static inline bool amdgpu_has_atpx_dgpu_power_cntl(void) { return false; } +static inline bool amdgpu_is_atpx_hybrid(void) { return false; } #endif /* diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index 32809f749903..d080d0807a5b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -240,8 +240,8 @@ uint64_t get_gpu_clock_counter(struct kgd_dev *kgd) { struct amdgpu_device *rdev = (struct amdgpu_device *)kgd; - if (rdev->asic_funcs->get_gpu_clock_counter) - return rdev->asic_funcs->get_gpu_clock_counter(rdev); + if (rdev->gfx.funcs->get_gpu_clock_counter) + return rdev->gfx.funcs->get_gpu_clock_counter(rdev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c index 35a1248aaa77..0494fe7b62c0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c @@ -10,6 +10,7 @@ #include <linux/slab.h> #include <linux/acpi.h> #include <linux/pci.h> +#include <linux/delay.h> #include "amd_acpi.h" @@ -27,6 +28,7 @@ struct amdgpu_atpx_functions { struct amdgpu_atpx { acpi_handle handle; struct amdgpu_atpx_functions functions; + bool is_hybrid; }; static struct amdgpu_atpx_priv { @@ -63,6 +65,14 @@ bool amdgpu_has_atpx(void) { return amdgpu_atpx_priv.atpx_detected; } +bool amdgpu_has_atpx_dgpu_power_cntl(void) { + return amdgpu_atpx_priv.atpx.functions.power_cntl; +} + +bool amdgpu_is_atpx_hybrid(void) { + return amdgpu_atpx_priv.atpx.is_hybrid; +} + /** * amdgpu_atpx_call - call an ATPX method * @@ -142,18 +152,12 @@ static void amdgpu_atpx_parse_functions(struct amdgpu_atpx_functions *f, u32 mas */ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) { - /* make sure required functions are enabled */ - /* dGPU power control is required */ - if (atpx->functions.power_cntl == false) { - printk("ATPX dGPU power cntl not present, forcing\n"); - atpx->functions.power_cntl = true; - } + u32 valid_bits = 0; if (atpx->functions.px_params) { union acpi_object *info; struct atpx_px_params output; size_t size; - u32 valid_bits; info = amdgpu_atpx_call(atpx->handle, ATPX_FUNCTION_GET_PX_PARAMETERS, NULL); if (!info) @@ -172,19 +176,43 @@ static int amdgpu_atpx_validate(struct amdgpu_atpx *atpx) memcpy(&output, info->buffer.pointer, size); valid_bits = output.flags & output.valid_flags; - /* if separate mux flag is set, mux controls are required */ - if (valid_bits & ATPX_SEPARATE_MUX_FOR_I2C) { - atpx->functions.i2c_mux_cntl = true; - atpx->functions.disp_mux_cntl = true; - } - /* if any outputs are muxed, mux controls are required */ - if (valid_bits & (ATPX_CRT1_RGB_SIGNAL_MUXED | - ATPX_TV_SIGNAL_MUXED | - ATPX_DFP_SIGNAL_MUXED)) - atpx->functions.disp_mux_cntl = true; kfree(info); } + + /* if separate mux flag is set, mux controls are required */ + if (valid_bits & ATPX_SEPARATE_MUX_FOR_I2C) { + atpx->functions.i2c_mux_cntl = true; + atpx->functions.disp_mux_cntl = true; + } + /* if any outputs are muxed, mux controls are required */ + if (valid_bits & (ATPX_CRT1_RGB_SIGNAL_MUXED | + ATPX_TV_SIGNAL_MUXED | + ATPX_DFP_SIGNAL_MUXED)) + atpx->functions.disp_mux_cntl = true; + + + /* some bioses set these bits rather than flagging power_cntl as supported */ + if (valid_bits & (ATPX_DYNAMIC_PX_SUPPORTED | + ATPX_DYNAMIC_DGPU_POWER_OFF_SUPPORTED)) + atpx->functions.power_cntl = true; + + atpx->is_hybrid = false; + if (valid_bits & ATPX_MS_HYBRID_GFX_SUPPORTED) { + printk("ATPX Hybrid Graphics\n"); +#if 1 + /* This is a temporary hack until the D3 cold support + * makes it upstream. The ATPX power_control method seems + * to still work on even if the system should be using + * the new standardized hybrid D3 cold ACPI interface. + */ + atpx->functions.power_cntl = true; +#else + atpx->functions.power_cntl = false; +#endif + atpx->is_hybrid = true; + } + return 0; } @@ -259,6 +287,10 @@ static int amdgpu_atpx_set_discrete_state(struct amdgpu_atpx *atpx, u8 state) if (!info) return -EIO; kfree(info); + + /* 200ms delay is required after off */ + if (state == 0) + msleep(200); } return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c index 823bf5e0b0c8..651115dcce12 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c @@ -94,6 +94,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, unsigned last_entry = 0, first_userptr = num_entries; unsigned i; int r; + unsigned long total_size = 0; array = drm_malloc_ab(num_entries, sizeof(struct amdgpu_bo_list_entry)); if (!array) @@ -140,6 +141,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, if (entry->robj->prefered_domains == AMDGPU_GEM_DOMAIN_OA) oa_obj = entry->robj; + total_size += amdgpu_bo_size(entry->robj); trace_amdgpu_bo_list_set(list, entry->robj); } @@ -155,6 +157,7 @@ static int amdgpu_bo_list_set(struct amdgpu_device *adev, list->array = array; list->num_entries = num_entries; + trace_amdgpu_cs_bo_status(list->num_entries, total_size); return 0; error_free: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index 199f76baf22c..5556ce979199 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -312,6 +312,8 @@ static uint32_t amdgpu_cgs_read_ind_register(struct cgs_device *cgs_device, return RREG32_UVD_CTX(index); case CGS_IND_REG__DIDT: return RREG32_DIDT(index); + case CGS_IND_REG_GC_CAC: + return RREG32_GC_CAC(index); case CGS_IND_REG__AUDIO_ENDPT: DRM_ERROR("audio endpt register access not implemented.\n"); return 0; @@ -336,6 +338,8 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device, return WREG32_UVD_CTX(index, value); case CGS_IND_REG__DIDT: return WREG32_DIDT(index, value); + case CGS_IND_REG_GC_CAC: + return WREG32_GC_CAC(index, value); case CGS_IND_REG__AUDIO_ENDPT: DRM_ERROR("audio endpt register access not implemented.\n"); return; @@ -696,6 +700,17 @@ static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type) return result; } +static int amdgpu_cgs_rel_firmware(struct cgs_device *cgs_device, enum cgs_ucode_id type) +{ + CGS_FUNC_ADEV; + if ((CGS_UCODE_ID_SMU == type) || (CGS_UCODE_ID_SMU_SK == type)) { + release_firmware(adev->pm.fw); + return 0; + } + /* cannot release other firmware because they are not created by cgs */ + return -EINVAL; +} + static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, enum cgs_ucode_id type, struct cgs_firmware_info *info) @@ -776,6 +791,7 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device, } hdr = (const struct smc_firmware_header_v1_0 *) adev->pm.fw->data; + amdgpu_ucode_print_smc_hdr(&hdr->header); adev->pm.fw_version = le32_to_cpu(hdr->header.ucode_version); ucode_size = le32_to_cpu(hdr->header.ucode_size_bytes); ucode_start_address = le32_to_cpu(hdr->ucode_start_addr); @@ -819,6 +835,9 @@ static int amdgpu_cgs_query_system_info(struct cgs_device *cgs_device, case CGS_SYSTEM_INFO_GFX_CU_INFO: sys_info->value = adev->gfx.cu_info.number; break; + case CGS_SYSTEM_INFO_GFX_SE_INFO: + sys_info->value = adev->gfx.config.max_shader_engines; + break; default: return -ENODEV; } @@ -898,7 +917,7 @@ static int amdgpu_cgs_acpi_eval_object(struct cgs_device *cgs_device, struct cgs_acpi_method_argument *argument = NULL; uint32_t i, count; acpi_status status; - int result; + int result = 0; uint32_t func_no = 0xFFFFFFFF; handle = ACPI_HANDLE(&adev->pdev->dev); @@ -961,11 +980,11 @@ static int amdgpu_cgs_acpi_eval_object(struct cgs_device *cgs_device, params->integer.value = argument->value; break; case ACPI_TYPE_STRING: - params->string.length = argument->method_length; + params->string.length = argument->data_length; params->string.pointer = argument->pointer; break; case ACPI_TYPE_BUFFER: - params->buffer.length = argument->method_length; + params->buffer.length = argument->data_length; params->buffer.pointer = argument->pointer; break; default: @@ -1068,17 +1087,14 @@ int amdgpu_cgs_call_acpi_method(struct cgs_device *cgs_device, struct cgs_acpi_method_info info = {0}; acpi_input[0].type = CGS_ACPI_TYPE_INTEGER; - acpi_input[0].method_length = sizeof(uint32_t); acpi_input[0].data_length = sizeof(uint32_t); acpi_input[0].value = acpi_function; acpi_input[1].type = CGS_ACPI_TYPE_BUFFER; - acpi_input[1].method_length = CGS_ACPI_MAX_BUFFER_SIZE; acpi_input[1].data_length = input_size; acpi_input[1].pointer = pinput; acpi_output.type = CGS_ACPI_TYPE_BUFFER; - acpi_output.method_length = CGS_ACPI_MAX_BUFFER_SIZE; acpi_output.data_length = output_size; acpi_output.pointer = poutput; @@ -1125,6 +1141,7 @@ static const struct cgs_ops amdgpu_cgs_ops = { amdgpu_cgs_pm_query_clock_limits, amdgpu_cgs_set_camera_voltages, amdgpu_cgs_get_firmware_info, + amdgpu_cgs_rel_firmware, amdgpu_cgs_set_powergating_state, amdgpu_cgs_set_clockgating_state, amdgpu_cgs_get_active_displays_info, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c index cb07da41152b..ff0b55a65ca3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c @@ -1690,7 +1690,6 @@ amdgpu_connector_add(struct amdgpu_device *adev, DRM_MODE_SCALE_NONE); /* no HPD on analog connectors */ amdgpu_connector->hpd.hpd = AMDGPU_HPD_NONE; - connector->polled = DRM_CONNECTOR_POLL_CONNECT; connector->interlace_allowed = true; connector->doublescan_allowed = true; break; @@ -1893,8 +1892,10 @@ amdgpu_connector_add(struct amdgpu_device *adev, } if (amdgpu_connector->hpd.hpd == AMDGPU_HPD_NONE) { - if (i2c_bus->valid) - connector->polled = DRM_CONNECTOR_POLL_CONNECT; + if (i2c_bus->valid) { + connector->polled = DRM_CONNECTOR_POLL_CONNECT | + DRM_CONNECTOR_POLL_DISCONNECT; + } } else connector->polled = DRM_CONNECTOR_POLL_HPD; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 9bc8f1d99733..0307ff5887c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -216,11 +216,8 @@ int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) if (ret) goto free_all_kdata; - if (p->uf_entry.robj) { - p->job->uf_bo = amdgpu_bo_ref(p->uf_entry.robj); - p->job->uf_offset = uf_offset; - } - + if (p->uf_entry.robj) + p->job->uf_addr = uf_offset; kfree(chunk_array); return 0; @@ -459,7 +456,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, list_splice(&need_pages, &p->validated); } - amdgpu_vm_get_pt_bos(&fpriv->vm, &duplicates); + amdgpu_vm_get_pt_bos(p->adev, &fpriv->vm, &duplicates); p->bytes_moved_threshold = amdgpu_cs_get_threshold_for_moves(p->adev); p->bytes_moved = 0; @@ -472,6 +469,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, if (r) goto error_validate; + fpriv->vm.last_eviction_counter = + atomic64_read(&p->adev->num_evictions); + if (p->bo_list) { struct amdgpu_bo *gds = p->bo_list->gds_obj; struct amdgpu_bo *gws = p->bo_list->gws_obj; @@ -499,6 +499,9 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, } } + if (p->uf_entry.robj) + p->job->uf_addr += amdgpu_bo_gpu_offset(p->uf_entry.robj); + error_validate: if (r) { amdgpu_vm_move_pt_bos_in_lru(p->adev, &fpriv->vm); @@ -653,18 +656,21 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev, /* Only for UVD/VCE VM emulation */ if (ring->funcs->parse_cs) { + p->job->vm = NULL; for (i = 0; i < p->job->num_ibs; i++) { r = amdgpu_ring_parse_cs(ring, p, i); if (r) return r; } - } + } else { + p->job->vm_pd_addr = amdgpu_bo_gpu_offset(vm->page_directory); - r = amdgpu_bo_vm_update_pte(p, vm); - if (!r) - amdgpu_cs_sync_rings(p); + r = amdgpu_bo_vm_update_pte(p, vm); + if (r) + return r; + } - return r; + return amdgpu_cs_sync_rings(p); } static int amdgpu_cs_handle_lockup(struct amdgpu_device *adev, int r) @@ -761,7 +767,7 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev, } /* UVD & VCE fw doesn't support user fences */ - if (parser->job->uf_bo && ( + if (parser->job->uf_addr && ( parser->job->ring->type == AMDGPU_RING_TYPE_UVD || parser->job->ring->type == AMDGPU_RING_TYPE_VCE)) return -EINVAL; @@ -830,17 +836,13 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, { struct amdgpu_ring *ring = p->job->ring; struct amd_sched_entity *entity = &p->ctx->rings[ring->idx].entity; - struct fence *fence; struct amdgpu_job *job; int r; job = p->job; p->job = NULL; - r = amd_sched_job_init(&job->base, &ring->sched, - entity, amdgpu_job_timeout_func, - amdgpu_job_free_func, - p->filp, &fence); + r = amd_sched_job_init(&job->base, &ring->sched, entity, p->filp); if (r) { amdgpu_job_free(job); return r; @@ -848,9 +850,10 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, job->owner = p->filp; job->ctx = entity->fence_context; - p->fence = fence_get(fence); - cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, fence); + p->fence = fence_get(&job->base.s_fence->finished); + cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence); job->uf_sequence = cs->out.handle; + amdgpu_job_free_resources(job); trace_amdgpu_cs_ioctl(job); amd_sched_entity_push_job(&job->base); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 2ab5e0b184fb..614fb026436d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -25,6 +25,7 @@ * Alex Deucher * Jerome Glisse */ +#include <linux/kthread.h> #include <linux/console.h> #include <linux/slab.h> #include <linux/debugfs.h> @@ -35,6 +36,7 @@ #include <linux/vga_switcheroo.h> #include <linux/efi.h> #include "amdgpu.h" +#include "amdgpu_trace.h" #include "amdgpu_i2c.h" #include "atom.h" #include "amdgpu_atombios.h" @@ -79,24 +81,27 @@ bool amdgpu_device_is_px(struct drm_device *dev) uint32_t amdgpu_mm_rreg(struct amdgpu_device *adev, uint32_t reg, bool always_indirect) { + uint32_t ret; + if ((reg * 4) < adev->rmmio_size && !always_indirect) - return readl(((void __iomem *)adev->rmmio) + (reg * 4)); + ret = readl(((void __iomem *)adev->rmmio) + (reg * 4)); else { unsigned long flags; - uint32_t ret; spin_lock_irqsave(&adev->mmio_idx_lock, flags); writel((reg * 4), ((void __iomem *)adev->rmmio) + (mmMM_INDEX * 4)); ret = readl(((void __iomem *)adev->rmmio) + (mmMM_DATA * 4)); spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); - - return ret; } + trace_amdgpu_mm_rreg(adev->pdev->device, reg, ret); + return ret; } void amdgpu_mm_wreg(struct amdgpu_device *adev, uint32_t reg, uint32_t v, bool always_indirect) { + trace_amdgpu_mm_wreg(adev->pdev->device, reg, v); + if ((reg * 4) < adev->rmmio_size && !always_indirect) writel(v, ((void __iomem *)adev->rmmio) + (reg * 4)); else { @@ -827,8 +832,10 @@ static uint32_t cail_ioreg_read(struct card_info *info, uint32_t reg) */ static void amdgpu_atombios_fini(struct amdgpu_device *adev) { - if (adev->mode_info.atom_context) + if (adev->mode_info.atom_context) { kfree(adev->mode_info.atom_context->scratch); + kfree(adev->mode_info.atom_context->iio); + } kfree(adev->mode_info.atom_context); adev->mode_info.atom_context = NULL; kfree(adev->mode_info.atom_card_info); @@ -1068,11 +1075,14 @@ int amdgpu_set_clockgating_state(struct amdgpu_device *adev, int i, r = 0; for (i = 0; i < adev->num_ip_blocks; i++) { + if (!adev->ip_block_status[i].valid) + continue; if (adev->ip_blocks[i].type == block_type) { r = adev->ip_blocks[i].funcs->set_clockgating_state((void *)adev, state); if (r) return r; + break; } } return r; @@ -1085,16 +1095,53 @@ int amdgpu_set_powergating_state(struct amdgpu_device *adev, int i, r = 0; for (i = 0; i < adev->num_ip_blocks; i++) { + if (!adev->ip_block_status[i].valid) + continue; if (adev->ip_blocks[i].type == block_type) { r = adev->ip_blocks[i].funcs->set_powergating_state((void *)adev, state); if (r) return r; + break; } } return r; } +int amdgpu_wait_for_idle(struct amdgpu_device *adev, + enum amd_ip_block_type block_type) +{ + int i, r; + + for (i = 0; i < adev->num_ip_blocks; i++) { + if (!adev->ip_block_status[i].valid) + continue; + if (adev->ip_blocks[i].type == block_type) { + r = adev->ip_blocks[i].funcs->wait_for_idle((void *)adev); + if (r) + return r; + break; + } + } + return 0; + +} + +bool amdgpu_is_idle(struct amdgpu_device *adev, + enum amd_ip_block_type block_type) +{ + int i; + + for (i = 0; i < adev->num_ip_blocks; i++) { + if (!adev->ip_block_status[i].valid) + continue; + if (adev->ip_blocks[i].type == block_type) + return adev->ip_blocks[i].funcs->is_idle((void *)adev); + } + return true; + +} + const struct amdgpu_ip_block_version * amdgpu_get_ip_block( struct amdgpu_device *adev, enum amd_ip_block_type type) @@ -1207,6 +1254,9 @@ static int amdgpu_early_init(struct amdgpu_device *adev) } } + adev->cg_flags &= amdgpu_cg_mask; + adev->pg_flags &= amdgpu_pg_mask; + return 0; } @@ -1325,6 +1375,11 @@ static int amdgpu_fini(struct amdgpu_device *adev) adev->ip_block_status[i].valid = false; } + for (i = adev->num_ip_blocks - 1; i >= 0; i--) { + if (adev->ip_blocks[i].funcs->late_fini) + adev->ip_blocks[i].funcs->late_fini((void *)adev); + } + return 0; } @@ -1378,6 +1433,15 @@ static int amdgpu_resume(struct amdgpu_device *adev) return 0; } +static bool amdgpu_device_is_virtual(void) +{ +#ifdef CONFIG_X86 + return boot_cpu_has(X86_FEATURE_HYPERVISOR); +#else + return false; +#endif +} + /** * amdgpu_device_init - initialize the driver * @@ -1424,9 +1488,12 @@ int amdgpu_device_init(struct amdgpu_device *adev, adev->uvd_ctx_wreg = &amdgpu_invalid_wreg; adev->didt_rreg = &amdgpu_invalid_rreg; adev->didt_wreg = &amdgpu_invalid_wreg; + adev->gc_cac_rreg = &amdgpu_invalid_rreg; + adev->gc_cac_wreg = &amdgpu_invalid_wreg; adev->audio_endpt_rreg = &amdgpu_block_invalid_rreg; adev->audio_endpt_wreg = &amdgpu_block_invalid_wreg; + DRM_INFO("initializing kernel modesetting (%s 0x%04X:0x%04X 0x%04X:0x%04X 0x%02X).\n", amdgpu_asic_name[adev->asic_type], pdev->vendor, pdev->device, pdev->subsystem_vendor, pdev->subsystem_device, pdev->revision); @@ -1451,6 +1518,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, spin_lock_init(&adev->pcie_idx_lock); spin_lock_init(&adev->uvd_ctx_idx_lock); spin_lock_init(&adev->didt_idx_lock); + spin_lock_init(&adev->gc_cac_idx_lock); spin_lock_init(&adev->audio_endpt_idx_lock); adev->rmmio_base = pci_resource_start(adev->pdev, 5); @@ -1495,29 +1563,38 @@ int amdgpu_device_init(struct amdgpu_device *adev, vga_switcheroo_init_domain_pm_ops(adev->dev, &adev->vga_pm_domain); /* Read BIOS */ - if (!amdgpu_get_bios(adev)) - return -EINVAL; + if (!amdgpu_get_bios(adev)) { + r = -EINVAL; + goto failed; + } /* Must be an ATOMBIOS */ if (!adev->is_atom_bios) { dev_err(adev->dev, "Expecting atombios for GPU\n"); - return -EINVAL; + r = -EINVAL; + goto failed; } r = amdgpu_atombios_init(adev); if (r) { dev_err(adev->dev, "amdgpu_atombios_init failed\n"); - return r; + goto failed; } /* See if the asic supports SR-IOV */ adev->virtualization.supports_sr_iov = amdgpu_atombios_has_gpu_virtualization_table(adev); + /* Check if we are executing in a virtualized environment */ + adev->virtualization.is_virtual = amdgpu_device_is_virtual(); + adev->virtualization.caps = amdgpu_asic_get_virtual_caps(adev); + /* Post card if necessary */ if (!amdgpu_card_posted(adev) || - adev->virtualization.supports_sr_iov) { + (adev->virtualization.is_virtual && + !(adev->virtualization.caps & AMDGPU_VIRT_CAPS_SRIOV_EN))) { if (!adev->bios) { dev_err(adev->dev, "Card not posted and no BIOS - ignoring\n"); - return -EINVAL; + r = -EINVAL; + goto failed; } DRM_INFO("GPU not posted. posting now...\n"); amdgpu_atom_asic_init(adev->mode_info.atom_context); @@ -1527,7 +1604,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_atombios_get_clock_info(adev); if (r) { dev_err(adev->dev, "amdgpu_atombios_get_clock_info failed\n"); - return r; + goto failed; } /* init i2c buses */ amdgpu_atombios_i2c_init(adev); @@ -1536,7 +1613,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_fence_driver_init(adev); if (r) { dev_err(adev->dev, "amdgpu_fence_driver_init failed\n"); - return r; + goto failed; } /* init the mode config */ @@ -1546,7 +1623,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (r) { dev_err(adev->dev, "amdgpu_init failed\n"); amdgpu_fini(adev); - return r; + goto failed; } adev->accel_working = true; @@ -1556,7 +1633,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_ib_pool_init(adev); if (r) { dev_err(adev->dev, "IB initialization failed (%d).\n", r); - return r; + goto failed; } r = amdgpu_ib_ring_tests(adev); @@ -1573,6 +1650,12 @@ int amdgpu_device_init(struct amdgpu_device *adev, DRM_ERROR("registering register debugfs failed (%d).\n", r); } + r = amdgpu_debugfs_firmware_init(adev); + if (r) { + DRM_ERROR("registering firmware debugfs failed (%d).\n", r); + return r; + } + if ((amdgpu_testing & 1)) { if (adev->accel_working) amdgpu_test_moves(adev); @@ -1598,10 +1681,15 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_late_init(adev); if (r) { dev_err(adev->dev, "amdgpu_late_init failed\n"); - return r; + goto failed; } return 0; + +failed: + if (runtime) + vga_switcheroo_fini_domain_pm_ops(adev->dev); + return r; } static void amdgpu_debugfs_remove_files(struct amdgpu_device *adev); @@ -1636,6 +1724,8 @@ void amdgpu_device_fini(struct amdgpu_device *adev) kfree(adev->bios); adev->bios = NULL; vga_switcheroo_unregister_client(adev->pdev); + if (adev->flags & AMD_IS_PX) + vga_switcheroo_fini_domain_pm_ops(adev->dev); vga_client_register(adev->pdev, NULL, NULL, NULL); if (adev->rio_mem) pci_iounmap(adev->pdev, adev->rio_mem); @@ -1841,11 +1931,6 @@ int amdgpu_resume_kms(struct drm_device *dev, bool resume, bool fbcon) */ int amdgpu_gpu_reset(struct amdgpu_device *adev) { - unsigned ring_sizes[AMDGPU_MAX_RINGS]; - uint32_t *ring_data[AMDGPU_MAX_RINGS]; - - bool saved = false; - int i, r; int resched; @@ -1854,22 +1939,30 @@ int amdgpu_gpu_reset(struct amdgpu_device *adev) /* block TTM */ resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev); - r = amdgpu_suspend(adev); - + /* block scheduler */ for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = adev->rings[i]; + if (!ring) continue; - - ring_sizes[i] = amdgpu_ring_backup(ring, &ring_data[i]); - if (ring_sizes[i]) { - saved = true; - dev_info(adev->dev, "Saved %d dwords of commands " - "on ring %d.\n", ring_sizes[i], i); - } + kthread_park(ring->sched.thread); + amd_sched_hw_job_reset(&ring->sched); } + /* after all hw jobs are reset, hw fence is meaningless, so force_completion */ + amdgpu_fence_driver_force_completion(adev); + + /* save scratch */ + amdgpu_atombios_scratch_regs_save(adev); + r = amdgpu_suspend(adev); retry: + /* Disable fb access */ + if (adev->mode_info.num_crtc) { + struct amdgpu_mode_mc_save save; + amdgpu_display_stop_mc_access(adev, &save); + amdgpu_wait_for_idle(adev, AMD_IP_BLOCK_TYPE_GMC); + } + r = amdgpu_asic_reset(adev); /* post card */ amdgpu_atom_asic_init(adev->mode_info.atom_context); @@ -1878,32 +1971,29 @@ retry: dev_info(adev->dev, "GPU reset succeeded, trying to resume\n"); r = amdgpu_resume(adev); } - + /* restore scratch */ + amdgpu_atombios_scratch_regs_restore(adev); if (!r) { + r = amdgpu_ib_ring_tests(adev); + if (r) { + dev_err(adev->dev, "ib ring test failed (%d).\n", r); + r = amdgpu_suspend(adev); + goto retry; + } + for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = adev->rings[i]; if (!ring) continue; - - amdgpu_ring_restore(ring, ring_sizes[i], ring_data[i]); - ring_sizes[i] = 0; - ring_data[i] = NULL; - } - - r = amdgpu_ib_ring_tests(adev); - if (r) { - dev_err(adev->dev, "ib ring test failed (%d).\n", r); - if (saved) { - saved = false; - r = amdgpu_suspend(adev); - goto retry; - } + amd_sched_job_recovery(&ring->sched); + kthread_unpark(ring->sched.thread); } } else { - amdgpu_fence_driver_force_completion(adev); + dev_err(adev->dev, "asic resume failed (%d).\n", r); for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { - if (adev->rings[i]) - kfree(ring_data[i]); + if (adev->rings[i]) { + kthread_unpark(adev->rings[i]->sched.thread); + } } } @@ -1914,13 +2004,11 @@ retry: /* bad news, how to tell it to userspace ? */ dev_info(adev->dev, "GPU reset failed\n"); } + amdgpu_irq_gpu_reset_resume_helper(adev); return r; } -#define AMDGPU_DEFAULT_PCIE_GEN_MASK 0x30007 /* gen: chipset 1/2, asic 1/2/3 */ -#define AMDGPU_DEFAULT_PCIE_MLW_MASK 0x2f0000 /* 1/2/4/8/16 lanes */ - void amdgpu_get_pcie_info(struct amdgpu_device *adev) { u32 mask; @@ -2074,20 +2162,43 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, struct amdgpu_device *adev = f->f_inode->i_private; ssize_t result = 0; int r; + bool use_bank; + unsigned instance_bank, sh_bank, se_bank; if (size & 0x3 || *pos & 0x3) return -EINVAL; + if (*pos & (1ULL << 62)) { + se_bank = (*pos >> 24) & 0x3FF; + sh_bank = (*pos >> 34) & 0x3FF; + instance_bank = (*pos >> 44) & 0x3FF; + use_bank = 1; + *pos &= 0xFFFFFF; + } else { + use_bank = 0; + } + + if (use_bank) { + if (sh_bank >= adev->gfx.config.max_sh_per_se || + se_bank >= adev->gfx.config.max_shader_engines) + return -EINVAL; + mutex_lock(&adev->grbm_idx_mutex); + amdgpu_gfx_select_se_sh(adev, se_bank, + sh_bank, instance_bank); + } + while (size) { uint32_t value; if (*pos > adev->rmmio_size) - return result; + goto end; value = RREG32(*pos >> 2); r = put_user(value, (uint32_t *)buf); - if (r) - return r; + if (r) { + result = r; + goto end; + } result += 4; buf += 4; @@ -2095,6 +2206,12 @@ static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf, size -= 4; } +end: + if (use_bank) { + amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); + mutex_unlock(&adev->grbm_idx_mutex); + } + return result; } @@ -2294,6 +2411,68 @@ static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user * return result; } +static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = f->f_inode->i_private; + ssize_t result = 0; + int r; + uint32_t *config, no_regs = 0; + + if (size & 0x3 || *pos & 0x3) + return -EINVAL; + + config = kmalloc(256 * sizeof(*config), GFP_KERNEL); + if (!config) + return -ENOMEM; + + /* version, increment each time something is added */ + config[no_regs++] = 0; + config[no_regs++] = adev->gfx.config.max_shader_engines; + config[no_regs++] = adev->gfx.config.max_tile_pipes; + config[no_regs++] = adev->gfx.config.max_cu_per_sh; + config[no_regs++] = adev->gfx.config.max_sh_per_se; + config[no_regs++] = adev->gfx.config.max_backends_per_se; + config[no_regs++] = adev->gfx.config.max_texture_channel_caches; + config[no_regs++] = adev->gfx.config.max_gprs; + config[no_regs++] = adev->gfx.config.max_gs_threads; + config[no_regs++] = adev->gfx.config.max_hw_contexts; + config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_frontend; + config[no_regs++] = adev->gfx.config.sc_prim_fifo_size_backend; + config[no_regs++] = adev->gfx.config.sc_hiz_tile_fifo_size; + config[no_regs++] = adev->gfx.config.sc_earlyz_tile_fifo_size; + config[no_regs++] = adev->gfx.config.num_tile_pipes; + config[no_regs++] = adev->gfx.config.backend_enable_mask; + config[no_regs++] = adev->gfx.config.mem_max_burst_length_bytes; + config[no_regs++] = adev->gfx.config.mem_row_size_in_kb; + config[no_regs++] = adev->gfx.config.shader_engine_tile_size; + config[no_regs++] = adev->gfx.config.num_gpus; + config[no_regs++] = adev->gfx.config.multi_gpu_tile_size; + config[no_regs++] = adev->gfx.config.mc_arb_ramcfg; + config[no_regs++] = adev->gfx.config.gb_addr_config; + config[no_regs++] = adev->gfx.config.num_rbs; + + while (size && (*pos < no_regs * 4)) { + uint32_t value; + + value = config[*pos >> 2]; + r = put_user(value, (uint32_t *)buf); + if (r) { + kfree(config); + return r; + } + + result += 4; + buf += 4; + *pos += 4; + size -= 4; + } + + kfree(config); + return result; +} + + static const struct file_operations amdgpu_debugfs_regs_fops = { .owner = THIS_MODULE, .read = amdgpu_debugfs_regs_read, @@ -2319,11 +2498,18 @@ static const struct file_operations amdgpu_debugfs_regs_smc_fops = { .llseek = default_llseek }; +static const struct file_operations amdgpu_debugfs_gca_config_fops = { + .owner = THIS_MODULE, + .read = amdgpu_debugfs_gca_config_read, + .llseek = default_llseek +}; + static const struct file_operations *debugfs_regs[] = { &amdgpu_debugfs_regs_fops, &amdgpu_debugfs_regs_didt_fops, &amdgpu_debugfs_regs_pcie_fops, &amdgpu_debugfs_regs_smc_fops, + &amdgpu_debugfs_gca_config_fops, }; static const char *debugfs_regs_names[] = { @@ -2331,6 +2517,7 @@ static const char *debugfs_regs_names[] = { "amdgpu_regs_didt", "amdgpu_regs_pcie", "amdgpu_regs_smc", + "amdgpu_gca_config", }; static int amdgpu_debugfs_regs_init(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c index a6eecf6f9065..7dbe8d02c5a6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c @@ -220,19 +220,17 @@ int amdgpu_crtc_page_flip(struct drm_crtc *crtc, r = amdgpu_bo_pin_restricted(new_rbo, AMDGPU_GEM_DOMAIN_VRAM, 0, 0, &base); if (unlikely(r != 0)) { - amdgpu_bo_unreserve(new_rbo); r = -EINVAL; DRM_ERROR("failed to pin new rbo buffer before flip\n"); - goto cleanup; + goto unreserve; } r = reservation_object_get_fences_rcu(new_rbo->tbo.resv, &work->excl, &work->shared_count, &work->shared); if (unlikely(r != 0)) { - amdgpu_bo_unreserve(new_rbo); DRM_ERROR("failed to get fences for buffer\n"); - goto cleanup; + goto unpin; } amdgpu_bo_get_tiling_flags(new_rbo, &tiling_flags); @@ -275,9 +273,11 @@ pflip_cleanup: DRM_ERROR("failed to reserve new rbo in error path\n"); goto cleanup; } +unpin: if (unlikely(amdgpu_bo_unpin(new_rbo) != 0)) { DRM_ERROR("failed to unpin new rbo in error path\n"); } +unreserve: amdgpu_bo_unreserve(new_rbo); cleanup: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index f888c015f76c..015f1f4aae53 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -82,8 +82,12 @@ int amdgpu_exp_hw_support = 0; int amdgpu_sched_jobs = 32; int amdgpu_sched_hw_submission = 2; int amdgpu_powerplay = -1; +int amdgpu_powercontainment = 1; unsigned amdgpu_pcie_gen_cap = 0; unsigned amdgpu_pcie_lane_cap = 0; +unsigned amdgpu_cg_mask = 0xffffffff; +unsigned amdgpu_pg_mask = 0xffffffff; +char *amdgpu_disable_cu = NULL; MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes"); module_param_named(vramlimit, amdgpu_vram_limit, int, 0600); @@ -160,6 +164,9 @@ module_param_named(sched_hw_submission, amdgpu_sched_hw_submission, int, 0444); #ifdef CONFIG_DRM_AMD_POWERPLAY MODULE_PARM_DESC(powerplay, "Powerplay component (1 = enable, 0 = disable, -1 = auto (default))"); module_param_named(powerplay, amdgpu_powerplay, int, 0444); + +MODULE_PARM_DESC(powercontainment, "Power Containment (1 = enable (default), 0 = disable)"); +module_param_named(powercontainment, amdgpu_powercontainment, int, 0444); #endif MODULE_PARM_DESC(pcie_gen_cap, "PCIE Gen Caps (0: autodetect (default))"); @@ -168,6 +175,15 @@ module_param_named(pcie_gen_cap, amdgpu_pcie_gen_cap, uint, 0444); MODULE_PARM_DESC(pcie_lane_cap, "PCIE Lane Caps (0: autodetect (default))"); module_param_named(pcie_lane_cap, amdgpu_pcie_lane_cap, uint, 0444); +MODULE_PARM_DESC(cg_mask, "Clockgating flags mask (0 = disable clock gating)"); +module_param_named(cg_mask, amdgpu_cg_mask, uint, 0444); + +MODULE_PARM_DESC(pg_mask, "Powergating flags mask (0 = disable power gating)"); +module_param_named(pg_mask, amdgpu_pg_mask, uint, 0444); + +MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)"); +module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444); + static const struct pci_device_id pciidlist[] = { #ifdef CONFIG_DRM_AMDGPU_CIK /* Kaveri */ @@ -413,7 +429,10 @@ static int amdgpu_pmops_runtime_suspend(struct device *dev) pci_save_state(pdev); pci_disable_device(pdev); pci_ignore_hotplug(pdev); - pci_set_power_state(pdev, PCI_D3cold); + if (amdgpu_is_atpx_hybrid()) + pci_set_power_state(pdev, PCI_D3cold); + else if (!amdgpu_has_atpx_dgpu_power_cntl()) + pci_set_power_state(pdev, PCI_D3hot); drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF; return 0; @@ -430,7 +449,9 @@ static int amdgpu_pmops_runtime_resume(struct device *dev) drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; - pci_set_power_state(pdev, PCI_D0); + if (amdgpu_is_atpx_hybrid() || + !amdgpu_has_atpx_dgpu_power_cntl()) + pci_set_power_state(pdev, PCI_D0); pci_restore_state(pdev); ret = pci_enable_device(pdev); if (ret) @@ -515,7 +536,7 @@ static struct drm_driver kms_driver = { .driver_features = DRIVER_USE_AGP | DRIVER_HAVE_IRQ | DRIVER_IRQ_SHARED | DRIVER_GEM | - DRIVER_PRIME | DRIVER_RENDER, + DRIVER_PRIME | DRIVER_RENDER | DRIVER_MODESET, .dev_priv_size = 0, .load = amdgpu_driver_load_kms, .open = amdgpu_driver_open_kms, @@ -590,7 +611,6 @@ static int __init amdgpu_init(void) DRM_INFO("amdgpu kernel modesetting enabled.\n"); driver = &kms_driver; pdriver = &amdgpu_kms_pci_driver; - driver->driver_features |= DRIVER_MODESET; driver->num_ioctls = amdgpu_max_kms_ioctl; amdgpu_register_atpx_handler(); /* let modprobe override vga console setting */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 8fab6486064f..88fbed2389c0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -503,7 +503,7 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev, if (r) goto error_print; - amdgpu_vm_get_pt_bos(bo_va->vm, &duplicates); + amdgpu_vm_get_pt_bos(adev, bo_va->vm, &duplicates); list_for_each_entry(entry, &list, head) { domain = amdgpu_mem_type_to_domain(entry->bo->mem.mem_type); /* if anything is swapped out don't swap it in here, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c index 9f95da4f0536..a074edd95c70 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c @@ -70,3 +70,47 @@ void amdgpu_gfx_scratch_free(struct amdgpu_device *adev, uint32_t reg) } } } + +/** + * amdgpu_gfx_parse_disable_cu - Parse the disable_cu module parameter + * + * @mask: array in which the per-shader array disable masks will be stored + * @max_se: number of SEs + * @max_sh: number of SHs + * + * The bitmask of CUs to be disabled in the shader array determined by se and + * sh is stored in mask[se * max_sh + sh]. + */ +void amdgpu_gfx_parse_disable_cu(unsigned *mask, unsigned max_se, unsigned max_sh) +{ + unsigned se, sh, cu; + const char *p; + + memset(mask, 0, sizeof(*mask) * max_se * max_sh); + + if (!amdgpu_disable_cu || !*amdgpu_disable_cu) + return; + + p = amdgpu_disable_cu; + for (;;) { + char *next; + int ret = sscanf(p, "%u.%u.%u", &se, &sh, &cu); + if (ret < 3) { + DRM_ERROR("amdgpu: could not parse disable_cu\n"); + return; + } + + if (se < max_se && sh < max_sh && cu < 16) { + DRM_INFO("amdgpu: disabling CU %u.%u.%u\n", se, sh, cu); + mask[se * max_sh + sh] |= 1u << cu; + } else { + DRM_ERROR("amdgpu: disable_cu %u.%u.%u is out of range\n", + se, sh, cu); + } + + next = strchr(p, ','); + if (!next) + break; + p = next + 1; + } +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index dc06cbda7be6..51321e154c09 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -27,4 +27,6 @@ int amdgpu_gfx_scratch_get(struct amdgpu_device *adev, uint32_t *reg); void amdgpu_gfx_scratch_free(struct amdgpu_device *adev, uint32_t reg); +unsigned amdgpu_gfx_parse_disable_cu(unsigned *mask, unsigned max_se, unsigned max_sh); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c index 34e35423b78e..46c3097c5224 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c @@ -160,10 +160,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, patch_offset = amdgpu_ring_init_cond_exec(ring); if (vm) { - r = amdgpu_vm_flush(ring, job->vm_id, job->vm_pd_addr, - job->gds_base, job->gds_size, - job->gws_base, job->gws_size, - job->oa_base, job->oa_size); + r = amdgpu_vm_flush(ring, job); if (r) { amdgpu_ring_undo(ring); return r; @@ -203,11 +200,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs, } /* wrap the last IB with fence */ - if (job && job->uf_bo) { - uint64_t addr = amdgpu_bo_gpu_offset(job->uf_bo); - - addr += job->uf_offset; - amdgpu_ring_emit_fence(ring, addr, job->uf_sequence, + if (job && job->uf_addr) { + amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence, AMDGPU_FENCE_FLAG_64BIT); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 835a3fa8d8df..278708f5a744 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -383,6 +383,18 @@ int amdgpu_irq_update(struct amdgpu_device *adev, return r; } +void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev) +{ + int i, j; + for (i = 0; i < AMDGPU_MAX_IRQ_SRC_ID; i++) { + struct amdgpu_irq_src *src = adev->irq.sources[i]; + if (!src) + continue; + for (j = 0; j < src->num_types; j++) + amdgpu_irq_update(adev, src, j); + } +} + /** * amdgpu_irq_get - enable interrupt * diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h index e124b59f39c1..7ef09352e534 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.h @@ -94,6 +94,7 @@ int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type); bool amdgpu_irq_enabled(struct amdgpu_device *adev, struct amdgpu_irq_src *src, unsigned type); +void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev); int amdgpu_irq_add_domain(struct amdgpu_device *adev); void amdgpu_irq_remove_domain(struct amdgpu_device *adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c index f0dafa514fe4..aaee0c8f6731 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c @@ -28,21 +28,15 @@ #include "amdgpu.h" #include "amdgpu_trace.h" -static void amdgpu_job_free_handler(struct work_struct *ws) +static void amdgpu_job_timedout(struct amd_sched_job *s_job) { - struct amdgpu_job *job = container_of(ws, struct amdgpu_job, base.work_free_job); - amd_sched_job_put(&job->base); -} + struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base); -void amdgpu_job_timeout_func(struct work_struct *work) -{ - struct amdgpu_job *job = container_of(work, struct amdgpu_job, base.work_tdr.work); DRM_ERROR("ring %s timeout, last signaled seq=%u, last emitted seq=%u\n", - job->base.sched->name, - (uint32_t)atomic_read(&job->ring->fence_drv.last_seq), - job->ring->fence_drv.sync_seq); - - amd_sched_job_put(&job->base); + job->base.sched->name, + atomic_read(&job->ring->fence_drv.last_seq), + job->ring->fence_drv.sync_seq); + amdgpu_gpu_reset(job->adev); } int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, @@ -63,7 +57,6 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs, (*job)->vm = vm; (*job)->ibs = (void *)&(*job)[1]; (*job)->num_ibs = num_ibs; - INIT_WORK(&(*job)->base.work_free_job, amdgpu_job_free_handler); amdgpu_sync_create(&(*job)->sync); @@ -86,27 +79,33 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size, return r; } -void amdgpu_job_free(struct amdgpu_job *job) +void amdgpu_job_free_resources(struct amdgpu_job *job) { - unsigned i; struct fence *f; + unsigned i; + /* use sched fence if available */ - f = (job->base.s_fence)? &job->base.s_fence->base : job->fence; + f = job->base.s_fence ? &job->base.s_fence->finished : job->fence; for (i = 0; i < job->num_ibs; ++i) - amdgpu_sa_bo_free(job->adev, &job->ibs[i].sa_bo, f); - fence_put(job->fence); + amdgpu_ib_free(job->adev, &job->ibs[i], f); +} - amdgpu_bo_unref(&job->uf_bo); - amdgpu_sync_free(&job->sync); +void amdgpu_job_free_cb(struct amd_sched_job *s_job) +{ + struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base); - if (!job->base.use_sched) - kfree(job); + fence_put(job->fence); + amdgpu_sync_free(&job->sync); + kfree(job); } -void amdgpu_job_free_func(struct kref *refcount) +void amdgpu_job_free(struct amdgpu_job *job) { - struct amdgpu_job *job = container_of(refcount, struct amdgpu_job, base.refcount); + amdgpu_job_free_resources(job); + + fence_put(job->fence); + amdgpu_sync_free(&job->sync); kfree(job); } @@ -114,22 +113,20 @@ int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring, struct amd_sched_entity *entity, void *owner, struct fence **f) { - struct fence *fence; int r; job->ring = ring; if (!f) return -EINVAL; - r = amd_sched_job_init(&job->base, &ring->sched, - entity, amdgpu_job_timeout_func, - amdgpu_job_free_func, owner, &fence); + r = amd_sched_job_init(&job->base, &ring->sched, entity, owner); if (r) return r; job->owner = owner; job->ctx = entity->fence_context; - *f = fence_get(fence); + *f = fence_get(&job->base.s_fence->finished); + amdgpu_job_free_resources(job); amd_sched_entity_push_job(&job->base); return 0; @@ -147,8 +144,8 @@ static struct fence *amdgpu_job_dependency(struct amd_sched_job *sched_job) int r; r = amdgpu_vm_grab_id(vm, ring, &job->sync, - &job->base.s_fence->base, - &job->vm_id, &job->vm_pd_addr); + &job->base.s_fence->finished, + job); if (r) DRM_ERROR("Error getting VM ID (%d)\n", r); @@ -170,11 +167,7 @@ static struct fence *amdgpu_job_run(struct amd_sched_job *sched_job) } job = to_amdgpu_job(sched_job); - r = amdgpu_sync_wait(&job->sync); - if (r) { - DRM_ERROR("failed to sync wait (%d)\n", r); - return NULL; - } + BUG_ON(amdgpu_sync_peek_fence(&job->sync, NULL)); trace_amdgpu_sched_run_job(job); r = amdgpu_ib_schedule(job->ring, job->num_ibs, job->ibs, @@ -185,14 +178,15 @@ static struct fence *amdgpu_job_run(struct amd_sched_job *sched_job) } err: + /* if gpu reset, hw fence will be replaced here */ + fence_put(job->fence); job->fence = fence; - amdgpu_job_free(job); return fence; } const struct amd_sched_backend_ops amdgpu_sched_ops = { .dependency = amdgpu_job_dependency, .run_job = amdgpu_job_run, - .begin_job = amd_sched_job_begin, - .finish_job = amd_sched_job_finish, + .timedout_job = amdgpu_job_timedout, + .free_job = amdgpu_job_free_cb }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 38a28d137f1d..a8efbb54423f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -148,6 +148,65 @@ out: return r; } +static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, + struct drm_amdgpu_query_fw *query_fw, + struct amdgpu_device *adev) +{ + switch (query_fw->fw_type) { + case AMDGPU_INFO_FW_VCE: + fw_info->ver = adev->vce.fw_version; + fw_info->feature = adev->vce.fb_version; + break; + case AMDGPU_INFO_FW_UVD: + fw_info->ver = adev->uvd.fw_version; + fw_info->feature = 0; + break; + case AMDGPU_INFO_FW_GMC: + fw_info->ver = adev->mc.fw_version; + fw_info->feature = 0; + break; + case AMDGPU_INFO_FW_GFX_ME: + fw_info->ver = adev->gfx.me_fw_version; + fw_info->feature = adev->gfx.me_feature_version; + break; + case AMDGPU_INFO_FW_GFX_PFP: + fw_info->ver = adev->gfx.pfp_fw_version; + fw_info->feature = adev->gfx.pfp_feature_version; + break; + case AMDGPU_INFO_FW_GFX_CE: + fw_info->ver = adev->gfx.ce_fw_version; + fw_info->feature = adev->gfx.ce_feature_version; + break; + case AMDGPU_INFO_FW_GFX_RLC: + fw_info->ver = adev->gfx.rlc_fw_version; + fw_info->feature = adev->gfx.rlc_feature_version; + break; + case AMDGPU_INFO_FW_GFX_MEC: + if (query_fw->index == 0) { + fw_info->ver = adev->gfx.mec_fw_version; + fw_info->feature = adev->gfx.mec_feature_version; + } else if (query_fw->index == 1) { + fw_info->ver = adev->gfx.mec2_fw_version; + fw_info->feature = adev->gfx.mec2_feature_version; + } else + return -EINVAL; + break; + case AMDGPU_INFO_FW_SMC: + fw_info->ver = adev->pm.fw_version; + fw_info->feature = 0; + break; + case AMDGPU_INFO_FW_SDMA: + if (query_fw->index >= adev->sdma.num_instances) + return -EINVAL; + fw_info->ver = adev->sdma.instance[query_fw->index].fw_version; + fw_info->feature = adev->sdma.instance[query_fw->index].feature_version; + break; + default: + return -EINVAL; + } + return 0; +} + /* * Userspace get information ioctl */ @@ -294,67 +353,20 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file return copy_to_user(out, &count, min(size, 4u)) ? -EFAULT : 0; } case AMDGPU_INFO_TIMESTAMP: - ui64 = amdgpu_asic_get_gpu_clock_counter(adev); + ui64 = amdgpu_gfx_get_gpu_clock_counter(adev); return copy_to_user(out, &ui64, min(size, 8u)) ? -EFAULT : 0; case AMDGPU_INFO_FW_VERSION: { struct drm_amdgpu_info_firmware fw_info; + int ret; /* We only support one instance of each IP block right now. */ if (info->query_fw.ip_instance != 0) return -EINVAL; - switch (info->query_fw.fw_type) { - case AMDGPU_INFO_FW_VCE: - fw_info.ver = adev->vce.fw_version; - fw_info.feature = adev->vce.fb_version; - break; - case AMDGPU_INFO_FW_UVD: - fw_info.ver = adev->uvd.fw_version; - fw_info.feature = 0; - break; - case AMDGPU_INFO_FW_GMC: - fw_info.ver = adev->mc.fw_version; - fw_info.feature = 0; - break; - case AMDGPU_INFO_FW_GFX_ME: - fw_info.ver = adev->gfx.me_fw_version; - fw_info.feature = adev->gfx.me_feature_version; - break; - case AMDGPU_INFO_FW_GFX_PFP: - fw_info.ver = adev->gfx.pfp_fw_version; - fw_info.feature = adev->gfx.pfp_feature_version; - break; - case AMDGPU_INFO_FW_GFX_CE: - fw_info.ver = adev->gfx.ce_fw_version; - fw_info.feature = adev->gfx.ce_feature_version; - break; - case AMDGPU_INFO_FW_GFX_RLC: - fw_info.ver = adev->gfx.rlc_fw_version; - fw_info.feature = adev->gfx.rlc_feature_version; - break; - case AMDGPU_INFO_FW_GFX_MEC: - if (info->query_fw.index == 0) { - fw_info.ver = adev->gfx.mec_fw_version; - fw_info.feature = adev->gfx.mec_feature_version; - } else if (info->query_fw.index == 1) { - fw_info.ver = adev->gfx.mec2_fw_version; - fw_info.feature = adev->gfx.mec2_feature_version; - } else - return -EINVAL; - break; - case AMDGPU_INFO_FW_SMC: - fw_info.ver = adev->pm.fw_version; - fw_info.feature = 0; - break; - case AMDGPU_INFO_FW_SDMA: - if (info->query_fw.index >= adev->sdma.num_instances) - return -EINVAL; - fw_info.ver = adev->sdma.instance[info->query_fw.index].fw_version; - fw_info.feature = adev->sdma.instance[info->query_fw.index].feature_version; - break; - default: - return -EINVAL; - } + ret = amdgpu_firmware_info(&fw_info, &info->query_fw, adev); + if (ret) + return ret; + return copy_to_user(out, &fw_info, min((size_t)size, sizeof(fw_info))) ? -EFAULT : 0; } @@ -453,7 +465,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file dev_info.max_memory_clock = adev->pm.default_mclk * 10; } dev_info.enabled_rb_pipes_mask = adev->gfx.config.backend_enable_mask; - dev_info.num_rb_pipes = adev->gfx.config.num_rbs; + dev_info.num_rb_pipes = adev->gfx.config.max_backends_per_se * + adev->gfx.config.max_shader_engines; dev_info.num_hw_gfx_contexts = adev->gfx.config.max_hw_contexts; dev_info._pad = 0; dev_info.ids_flags = 0; @@ -761,3 +774,130 @@ const struct drm_ioctl_desc amdgpu_ioctls_kms[] = { DRM_IOCTL_DEF_DRV(AMDGPU_GEM_USERPTR, amdgpu_gem_userptr_ioctl, DRM_AUTH|DRM_RENDER_ALLOW), }; const int amdgpu_max_kms_ioctl = ARRAY_SIZE(amdgpu_ioctls_kms); + +/* + * Debugfs info + */ +#if defined(CONFIG_DEBUG_FS) + +static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data) +{ + struct drm_info_node *node = (struct drm_info_node *) m->private; + struct drm_device *dev = node->minor->dev; + struct amdgpu_device *adev = dev->dev_private; + struct drm_amdgpu_info_firmware fw_info; + struct drm_amdgpu_query_fw query_fw; + int ret, i; + + /* VCE */ + query_fw.fw_type = AMDGPU_INFO_FW_VCE; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "VCE feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* UVD */ + query_fw.fw_type = AMDGPU_INFO_FW_UVD; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "UVD feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* GMC */ + query_fw.fw_type = AMDGPU_INFO_FW_GMC; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "MC feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* ME */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_ME; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "ME feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* PFP */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_PFP; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "PFP feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* CE */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_CE; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "CE feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* RLC */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "RLC feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* MEC */ + query_fw.fw_type = AMDGPU_INFO_FW_GFX_MEC; + query_fw.index = 0; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "MEC feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* MEC2 */ + if (adev->asic_type == CHIP_KAVERI || + (adev->asic_type > CHIP_TOPAZ && adev->asic_type != CHIP_STONEY)) { + query_fw.index = 1; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "MEC2 feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + } + + /* SMC */ + query_fw.fw_type = AMDGPU_INFO_FW_SMC; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "SMC feature version: %u, firmware version: 0x%08x\n", + fw_info.feature, fw_info.ver); + + /* SDMA */ + query_fw.fw_type = AMDGPU_INFO_FW_SDMA; + for (i = 0; i < adev->sdma.num_instances; i++) { + query_fw.index = i; + ret = amdgpu_firmware_info(&fw_info, &query_fw, adev); + if (ret) + return ret; + seq_printf(m, "SDMA%d feature version: %u, firmware version: 0x%08x\n", + i, fw_info.feature, fw_info.ver); + } + + return 0; +} + +static const struct drm_info_list amdgpu_firmware_info_list[] = { + {"amdgpu_firmware_info", amdgpu_debugfs_firmware_info, 0, NULL}, +}; +#endif + +int amdgpu_debugfs_firmware_init(struct amdgpu_device *adev) +{ +#if defined(CONFIG_DEBUG_FS) + return amdgpu_debugfs_add_files(adev, amdgpu_firmware_info_list, + ARRAY_SIZE(amdgpu_firmware_info_list)); +#else + return 0; +#endif +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 7ecea83ce453..6f0873c75a25 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -589,6 +589,7 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem) { struct amdgpu_bo *rbo; + struct ttm_mem_reg *old_mem = &bo->mem; if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) return; @@ -602,6 +603,8 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo, /* move_notify is called before move happens */ amdgpu_update_memory_usage(rbo->adev, &bo->mem, new_mem); + + trace_amdgpu_ttm_bo_move(rbo, new_mem->mem_type, old_mem->mem_type); } int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 589b36e8c5cf..ff63b88b0ffa 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -270,30 +270,28 @@ static ssize_t amdgpu_set_pp_force_state(struct device *dev, struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = ddev->dev_private; enum amd_pm_state_type state = 0; - long idx; + unsigned long idx; int ret; if (strlen(buf) == 1) adev->pp_force_state_enabled = false; - else { - ret = kstrtol(buf, 0, &idx); + else if (adev->pp_enabled) { + struct pp_states_info data; - if (ret) { + ret = kstrtoul(buf, 0, &idx); + if (ret || idx >= ARRAY_SIZE(data.states)) { count = -EINVAL; goto fail; } - if (adev->pp_enabled) { - struct pp_states_info data; - amdgpu_dpm_get_pp_num_states(adev, &data); - state = data.states[idx]; - /* only set user selected power states */ - if (state != POWER_STATE_TYPE_INTERNAL_BOOT && - state != POWER_STATE_TYPE_DEFAULT) { - amdgpu_dpm_dispatch_task(adev, - AMD_PP_EVENT_ENABLE_USER_STATE, &state, NULL); - adev->pp_force_state_enabled = true; - } + amdgpu_dpm_get_pp_num_states(adev, &data); + state = data.states[idx]; + /* only set user selected power states */ + if (state != POWER_STATE_TYPE_INTERNAL_BOOT && + state != POWER_STATE_TYPE_DEFAULT) { + amdgpu_dpm_dispatch_task(adev, + AMD_PP_EVENT_ENABLE_USER_STATE, &state, NULL); + adev->pp_force_state_enabled = true; } } fail: @@ -349,6 +347,8 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev, if (adev->pp_enabled) size = amdgpu_dpm_print_clock_levels(adev, PP_SCLK, buf); + else if (adev->pm.funcs->print_clock_levels) + size = adev->pm.funcs->print_clock_levels(adev, PP_SCLK, buf); return size; } @@ -365,7 +365,9 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, uint32_t i, mask = 0; char sub_str[2]; - for (i = 0; i < strlen(buf) - 1; i++) { + for (i = 0; i < strlen(buf); i++) { + if (*(buf + i) == '\n') + continue; sub_str[0] = *(buf + i); sub_str[1] = '\0'; ret = kstrtol(sub_str, 0, &level); @@ -379,6 +381,8 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev, if (adev->pp_enabled) amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask); + else if (adev->pm.funcs->force_clock_level) + adev->pm.funcs->force_clock_level(adev, PP_SCLK, mask); fail: return count; } @@ -393,6 +397,8 @@ static ssize_t amdgpu_get_pp_dpm_mclk(struct device *dev, if (adev->pp_enabled) size = amdgpu_dpm_print_clock_levels(adev, PP_MCLK, buf); + else if (adev->pm.funcs->print_clock_levels) + size = adev->pm.funcs->print_clock_levels(adev, PP_MCLK, buf); return size; } @@ -409,7 +415,9 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev, uint32_t i, mask = 0; char sub_str[2]; - for (i = 0; i < strlen(buf) - 1; i++) { + for (i = 0; i < strlen(buf); i++) { + if (*(buf + i) == '\n') + continue; sub_str[0] = *(buf + i); sub_str[1] = '\0'; ret = kstrtol(sub_str, 0, &level); @@ -423,6 +431,8 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev, if (adev->pp_enabled) amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask); + else if (adev->pm.funcs->force_clock_level) + adev->pm.funcs->force_clock_level(adev, PP_MCLK, mask); fail: return count; } @@ -437,6 +447,8 @@ static ssize_t amdgpu_get_pp_dpm_pcie(struct device *dev, if (adev->pp_enabled) size = amdgpu_dpm_print_clock_levels(adev, PP_PCIE, buf); + else if (adev->pm.funcs->print_clock_levels) + size = adev->pm.funcs->print_clock_levels(adev, PP_PCIE, buf); return size; } @@ -453,7 +465,9 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev, uint32_t i, mask = 0; char sub_str[2]; - for (i = 0; i < strlen(buf) - 1; i++) { + for (i = 0; i < strlen(buf); i++) { + if (*(buf + i) == '\n') + continue; sub_str[0] = *(buf + i); sub_str[1] = '\0'; ret = kstrtol(sub_str, 0, &level); @@ -467,6 +481,100 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev, if (adev->pp_enabled) amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask); + else if (adev->pm.funcs->force_clock_level) + adev->pm.funcs->force_clock_level(adev, PP_PCIE, mask); +fail: + return count; +} + +static ssize_t amdgpu_get_pp_sclk_od(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + uint32_t value = 0; + + if (adev->pp_enabled) + value = amdgpu_dpm_get_sclk_od(adev); + else if (adev->pm.funcs->get_sclk_od) + value = adev->pm.funcs->get_sclk_od(adev); + + return snprintf(buf, PAGE_SIZE, "%d\n", value); +} + +static ssize_t amdgpu_set_pp_sclk_od(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + int ret; + long int value; + + ret = kstrtol(buf, 0, &value); + + if (ret) { + count = -EINVAL; + goto fail; + } + + if (adev->pp_enabled) { + amdgpu_dpm_set_sclk_od(adev, (uint32_t)value); + amdgpu_dpm_dispatch_task(adev, AMD_PP_EVENT_READJUST_POWER_STATE, NULL, NULL); + } else if (adev->pm.funcs->set_sclk_od) { + adev->pm.funcs->set_sclk_od(adev, (uint32_t)value); + adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; + amdgpu_pm_compute_clocks(adev); + } + +fail: + return count; +} + +static ssize_t amdgpu_get_pp_mclk_od(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + uint32_t value = 0; + + if (adev->pp_enabled) + value = amdgpu_dpm_get_mclk_od(adev); + else if (adev->pm.funcs->get_mclk_od) + value = adev->pm.funcs->get_mclk_od(adev); + + return snprintf(buf, PAGE_SIZE, "%d\n", value); +} + +static ssize_t amdgpu_set_pp_mclk_od(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t count) +{ + struct drm_device *ddev = dev_get_drvdata(dev); + struct amdgpu_device *adev = ddev->dev_private; + int ret; + long int value; + + ret = kstrtol(buf, 0, &value); + + if (ret) { + count = -EINVAL; + goto fail; + } + + if (adev->pp_enabled) { + amdgpu_dpm_set_mclk_od(adev, (uint32_t)value); + amdgpu_dpm_dispatch_task(adev, AMD_PP_EVENT_READJUST_POWER_STATE, NULL, NULL); + } else if (adev->pm.funcs->set_mclk_od) { + adev->pm.funcs->set_mclk_od(adev, (uint32_t)value); + adev->pm.dpm.current_ps = adev->pm.dpm.boot_ps; + amdgpu_pm_compute_clocks(adev); + } + fail: return count; } @@ -492,6 +600,12 @@ static DEVICE_ATTR(pp_dpm_mclk, S_IRUGO | S_IWUSR, static DEVICE_ATTR(pp_dpm_pcie, S_IRUGO | S_IWUSR, amdgpu_get_pp_dpm_pcie, amdgpu_set_pp_dpm_pcie); +static DEVICE_ATTR(pp_sclk_od, S_IRUGO | S_IWUSR, + amdgpu_get_pp_sclk_od, + amdgpu_set_pp_sclk_od); +static DEVICE_ATTR(pp_mclk_od, S_IRUGO | S_IWUSR, + amdgpu_get_pp_mclk_od, + amdgpu_set_pp_mclk_od); static ssize_t amdgpu_hwmon_show_temp(struct device *dev, struct device_attribute *attr, @@ -1110,22 +1224,34 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev) DRM_ERROR("failed to create device file pp_table\n"); return ret; } - ret = device_create_file(adev->dev, &dev_attr_pp_dpm_sclk); - if (ret) { - DRM_ERROR("failed to create device file pp_dpm_sclk\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_pp_dpm_mclk); - if (ret) { - DRM_ERROR("failed to create device file pp_dpm_mclk\n"); - return ret; - } - ret = device_create_file(adev->dev, &dev_attr_pp_dpm_pcie); - if (ret) { - DRM_ERROR("failed to create device file pp_dpm_pcie\n"); - return ret; - } } + + ret = device_create_file(adev->dev, &dev_attr_pp_dpm_sclk); + if (ret) { + DRM_ERROR("failed to create device file pp_dpm_sclk\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_pp_dpm_mclk); + if (ret) { + DRM_ERROR("failed to create device file pp_dpm_mclk\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_pp_dpm_pcie); + if (ret) { + DRM_ERROR("failed to create device file pp_dpm_pcie\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_pp_sclk_od); + if (ret) { + DRM_ERROR("failed to create device file pp_sclk_od\n"); + return ret; + } + ret = device_create_file(adev->dev, &dev_attr_pp_mclk_od); + if (ret) { + DRM_ERROR("failed to create device file pp_mclk_od\n"); + return ret; + } + ret = amdgpu_debugfs_pm_init(adev); if (ret) { DRM_ERROR("Failed to register debugfs file for dpm!\n"); @@ -1148,10 +1274,12 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev) device_remove_file(adev->dev, &dev_attr_pp_cur_state); device_remove_file(adev->dev, &dev_attr_pp_force_state); device_remove_file(adev->dev, &dev_attr_pp_table); - device_remove_file(adev->dev, &dev_attr_pp_dpm_sclk); - device_remove_file(adev->dev, &dev_attr_pp_dpm_mclk); - device_remove_file(adev->dev, &dev_attr_pp_dpm_pcie); } + device_remove_file(adev->dev, &dev_attr_pp_dpm_sclk); + device_remove_file(adev->dev, &dev_attr_pp_dpm_mclk); + device_remove_file(adev->dev, &dev_attr_pp_dpm_pcie); + device_remove_file(adev->dev, &dev_attr_pp_sclk_od); + device_remove_file(adev->dev, &dev_attr_pp_mclk_od); } void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c index 6bd961fb43dc..c5738a22b690 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c @@ -52,6 +52,7 @@ static int amdgpu_powerplay_init(struct amdgpu_device *adev) pp_init->chip_family = adev->family; pp_init->chip_id = adev->asic_type; pp_init->device = amdgpu_cgs_create_device(adev); + pp_init->powercontainment_enabled = amdgpu_powercontainment; ret = amd_powerplay_init(pp_init, amd_pp); kfree(pp_init); @@ -183,13 +184,6 @@ static int amdgpu_pp_sw_fini(void *handle) if (ret) return ret; -#ifdef CONFIG_DRM_AMD_POWERPLAY - if (adev->pp_enabled) { - amdgpu_pm_sysfs_fini(adev); - amd_powerplay_fini(adev->powerplay.pp_handle); - } -#endif - return ret; } @@ -223,6 +217,22 @@ static int amdgpu_pp_hw_fini(void *handle) return ret; } +static void amdgpu_pp_late_fini(void *handle) +{ +#ifdef CONFIG_DRM_AMD_POWERPLAY + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + if (adev->pp_enabled) { + amdgpu_pm_sysfs_fini(adev); + amd_powerplay_fini(adev->powerplay.pp_handle); + } + + if (adev->powerplay.ip_funcs->late_fini) + adev->powerplay.ip_funcs->late_fini( + adev->powerplay.pp_handle); +#endif +} + static int amdgpu_pp_suspend(void *handle) { int ret = 0; @@ -311,6 +321,7 @@ const struct amd_ip_funcs amdgpu_pp_ip_funcs = { .sw_fini = amdgpu_pp_sw_fini, .hw_init = amdgpu_pp_hw_init, .hw_fini = amdgpu_pp_hw_fini, + .late_fini = amdgpu_pp_late_fini, .suspend = amdgpu_pp_suspend, .resume = amdgpu_pp_resume, .is_idle = amdgpu_pp_is_idle, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c index 3b02272db678..3b885e3e9b56 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c @@ -28,6 +28,7 @@ */ #include <linux/seq_file.h> #include <linux/slab.h> +#include <linux/debugfs.h> #include <drm/drmP.h> #include <drm/amdgpu_drm.h> #include "amdgpu.h" @@ -48,6 +49,7 @@ */ static int amdgpu_debugfs_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring); +static void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring); /** * amdgpu_ring_alloc - allocate space on the ring buffer @@ -140,78 +142,6 @@ void amdgpu_ring_undo(struct amdgpu_ring *ring) } /** - * amdgpu_ring_backup - Back up the content of a ring - * - * @ring: the ring we want to back up - * - * Saves all unprocessed commits from a ring, returns the number of dwords saved. - */ -unsigned amdgpu_ring_backup(struct amdgpu_ring *ring, - uint32_t **data) -{ - unsigned size, ptr, i; - - *data = NULL; - - if (ring->ring_obj == NULL) - return 0; - - /* it doesn't make sense to save anything if all fences are signaled */ - if (!amdgpu_fence_count_emitted(ring)) - return 0; - - ptr = le32_to_cpu(*ring->next_rptr_cpu_addr); - - size = ring->wptr + (ring->ring_size / 4); - size -= ptr; - size &= ring->ptr_mask; - if (size == 0) - return 0; - - /* and then save the content of the ring */ - *data = kmalloc_array(size, sizeof(uint32_t), GFP_KERNEL); - if (!*data) - return 0; - for (i = 0; i < size; ++i) { - (*data)[i] = ring->ring[ptr++]; - ptr &= ring->ptr_mask; - } - - return size; -} - -/** - * amdgpu_ring_restore - append saved commands to the ring again - * - * @ring: ring to append commands to - * @size: number of dwords we want to write - * @data: saved commands - * - * Allocates space on the ring and restore the previously saved commands. - */ -int amdgpu_ring_restore(struct amdgpu_ring *ring, - unsigned size, uint32_t *data) -{ - int i, r; - - if (!size || !data) - return 0; - - /* restore the saved ring content */ - r = amdgpu_ring_alloc(ring, size); - if (r) - return r; - - for (i = 0; i < size; ++i) { - amdgpu_ring_write(ring, data[i]); - } - - amdgpu_ring_commit(ring); - kfree(data); - return 0; -} - -/** * amdgpu_ring_init - init driver ring struct. * * @adev: amdgpu_device pointer @@ -260,14 +190,6 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, return r; } - r = amdgpu_wb_get(adev, &ring->next_rptr_offs); - if (r) { - dev_err(adev->dev, "(%d) ring next_rptr wb alloc failed\n", r); - return r; - } - ring->next_rptr_gpu_addr = adev->wb.gpu_addr + ring->next_rptr_offs * 4; - ring->next_rptr_cpu_addr = &adev->wb.wb[ring->next_rptr_offs]; - r = amdgpu_wb_get(adev, &ring->cond_exe_offs); if (r) { dev_err(adev->dev, "(%d) ring cond_exec_polling wb alloc failed\n", r); @@ -310,6 +232,9 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring, } r = amdgpu_bo_kmap(ring->ring_obj, (void **)&ring->ring); + + memset((void *)ring->ring, 0, ring->ring_size); + amdgpu_bo_unreserve(ring->ring_obj); if (r) { dev_err(adev->dev, "(%d) ring map failed\n", r); @@ -343,10 +268,10 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring) ring->ring = NULL; ring->ring_obj = NULL; + amdgpu_wb_free(ring->adev, ring->cond_exe_offs); amdgpu_wb_free(ring->adev, ring->fence_offs); amdgpu_wb_free(ring->adev, ring->rptr_offs); amdgpu_wb_free(ring->adev, ring->wptr_offs); - amdgpu_wb_free(ring->adev, ring->next_rptr_offs); if (ring_obj) { r = amdgpu_bo_reserve(ring_obj, false); @@ -357,6 +282,7 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring) } amdgpu_bo_unref(&ring_obj); } + amdgpu_debugfs_ring_fini(ring); } /* @@ -364,57 +290,62 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring) */ #if defined(CONFIG_DEBUG_FS) -static int amdgpu_debugfs_ring_info(struct seq_file *m, void *data) +/* Layout of file is 12 bytes consisting of + * - rptr + * - wptr + * - driver's copy of wptr + * + * followed by n-words of ring data + */ +static ssize_t amdgpu_debugfs_ring_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) { - struct drm_info_node *node = (struct drm_info_node *) m->private; - struct drm_device *dev = node->minor->dev; - struct amdgpu_device *adev = dev->dev_private; - int roffset = (unsigned long)node->info_ent->data; - struct amdgpu_ring *ring = (void *)(((uint8_t*)adev) + roffset); - uint32_t rptr, wptr, rptr_next; - unsigned i; - - wptr = amdgpu_ring_get_wptr(ring); - seq_printf(m, "wptr: 0x%08x [%5d]\n", wptr, wptr); - - rptr = amdgpu_ring_get_rptr(ring); - rptr_next = le32_to_cpu(*ring->next_rptr_cpu_addr); - - seq_printf(m, "rptr: 0x%08x [%5d]\n", rptr, rptr); - - seq_printf(m, "driver's copy of the wptr: 0x%08x [%5d]\n", - ring->wptr, ring->wptr); - - if (!ring->ready) - return 0; - - /* print 8 dw before current rptr as often it's the last executed - * packet that is the root issue - */ - i = (rptr + ring->ptr_mask + 1 - 32) & ring->ptr_mask; - while (i != rptr) { - seq_printf(m, "r[%5d]=0x%08x", i, ring->ring[i]); - if (i == rptr) - seq_puts(m, " *"); - if (i == rptr_next) - seq_puts(m, " #"); - seq_puts(m, "\n"); - i = (i + 1) & ring->ptr_mask; + struct amdgpu_ring *ring = (struct amdgpu_ring*)f->f_inode->i_private; + int r, i; + uint32_t value, result, early[3]; + + if (*pos & 3 || size & 3) + return -EINVAL; + + result = 0; + + if (*pos < 12) { + early[0] = amdgpu_ring_get_rptr(ring); + early[1] = amdgpu_ring_get_wptr(ring); + early[2] = ring->wptr; + for (i = *pos / 4; i < 3 && size; i++) { + r = put_user(early[i], (uint32_t *)buf); + if (r) + return r; + buf += 4; + result += 4; + size -= 4; + *pos += 4; + } } - while (i != wptr) { - seq_printf(m, "r[%5d]=0x%08x", i, ring->ring[i]); - if (i == rptr) - seq_puts(m, " *"); - if (i == rptr_next) - seq_puts(m, " #"); - seq_puts(m, "\n"); - i = (i + 1) & ring->ptr_mask; + + while (size) { + if (*pos >= (ring->ring_size + 12)) + return result; + + value = ring->ring[(*pos - 12)/4]; + r = put_user(value, (uint32_t*)buf); + if (r) + return r; + buf += 4; + result += 4; + size -= 4; + *pos += 4; } - return 0; + + return result; } -static struct drm_info_list amdgpu_debugfs_ring_info_list[AMDGPU_MAX_RINGS]; -static char amdgpu_debugfs_ring_names[AMDGPU_MAX_RINGS][32]; +static const struct file_operations amdgpu_debugfs_ring_fops = { + .owner = THIS_MODULE, + .read = amdgpu_debugfs_ring_read, + .llseek = default_llseek +}; #endif @@ -422,28 +353,27 @@ static int amdgpu_debugfs_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring) { #if defined(CONFIG_DEBUG_FS) - unsigned offset = (uint8_t*)ring - (uint8_t*)adev; - unsigned i; - struct drm_info_list *info; - char *name; - - for (i = 0; i < ARRAY_SIZE(amdgpu_debugfs_ring_info_list); ++i) { - info = &amdgpu_debugfs_ring_info_list[i]; - if (!info->data) - break; - } + struct drm_minor *minor = adev->ddev->primary; + struct dentry *ent, *root = minor->debugfs_root; + char name[32]; - if (i == ARRAY_SIZE(amdgpu_debugfs_ring_info_list)) - return -ENOSPC; - - name = &amdgpu_debugfs_ring_names[i][0]; sprintf(name, "amdgpu_ring_%s", ring->name); - info->name = name; - info->show = amdgpu_debugfs_ring_info; - info->driver_features = 0; - info->data = (void*)(uintptr_t)offset; - return amdgpu_debugfs_add_files(adev, info, 1); + ent = debugfs_create_file(name, + S_IFREG | S_IRUGO, root, + ring, &amdgpu_debugfs_ring_fops); + if (IS_ERR(ent)) + return PTR_ERR(ent); + + i_size_write(ent->d_inode, ring->ring_size + 12); + ring->ent = ent; #endif return 0; } + +static void amdgpu_debugfs_ring_fini(struct amdgpu_ring *ring) +{ +#if defined(CONFIG_DEBUG_FS) + debugfs_remove(ring->ent); +#endif +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c index b16366c2b4a0..d8af37a845f4 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sa.c @@ -115,6 +115,7 @@ int amdgpu_sa_bo_manager_start(struct amdgpu_device *adev, return r; } r = amdgpu_bo_kmap(sa_manager->bo, &sa_manager->cpu_ptr); + memset(sa_manager->cpu_ptr, 0, sa_manager->size); amdgpu_bo_unreserve(sa_manager->bo); return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c index 34a92808bbd4..5c8d3022fb87 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_sync.c @@ -223,13 +223,16 @@ int amdgpu_sync_resv(struct amdgpu_device *adev, } /** - * amdgpu_sync_is_idle - test if all fences are signaled + * amdgpu_sync_peek_fence - get the next fence not signaled yet * * @sync: the sync object + * @ring: optional ring to use for test * - * Returns true if all fences in the sync object are signaled. + * Returns the next fence not signaled yet without removing it from the sync + * object. */ -bool amdgpu_sync_is_idle(struct amdgpu_sync *sync) +struct fence *amdgpu_sync_peek_fence(struct amdgpu_sync *sync, + struct amdgpu_ring *ring) { struct amdgpu_sync_entry *e; struct hlist_node *tmp; @@ -237,6 +240,19 @@ bool amdgpu_sync_is_idle(struct amdgpu_sync *sync) hash_for_each_safe(sync->fences, i, tmp, e, node) { struct fence *f = e->fence; + struct amd_sched_fence *s_fence = to_amd_sched_fence(f); + + if (ring && s_fence) { + /* For fences from the same ring it is sufficient + * when they are scheduled. + */ + if (s_fence->sched == &ring->sched) { + if (fence_is_signaled(&s_fence->scheduled)) + continue; + + return &s_fence->scheduled; + } + } if (fence_is_signaled(f)) { hash_del(&e->node); @@ -245,58 +261,19 @@ bool amdgpu_sync_is_idle(struct amdgpu_sync *sync) continue; } - return false; + return f; } - return true; + return NULL; } /** - * amdgpu_sync_cycle_fences - move fences from one sync object into another + * amdgpu_sync_get_fence - get the next fence from the sync object * - * @dst: the destination sync object - * @src: the source sync object - * @fence: fence to add to source + * @sync: sync object to use * - * Remove all fences from source and put them into destination and add - * fence as new one into source. + * Get and removes the next fence from the sync object not signaled yet. */ -int amdgpu_sync_cycle_fences(struct amdgpu_sync *dst, struct amdgpu_sync *src, - struct fence *fence) -{ - struct amdgpu_sync_entry *e, *newone; - struct hlist_node *tmp; - int i; - - /* Allocate the new entry before moving the old ones */ - newone = kmem_cache_alloc(amdgpu_sync_slab, GFP_KERNEL); - if (!newone) - return -ENOMEM; - - hash_for_each_safe(src->fences, i, tmp, e, node) { - struct fence *f = e->fence; - - hash_del(&e->node); - if (fence_is_signaled(f)) { - fence_put(f); - kmem_cache_free(amdgpu_sync_slab, e); - continue; - } - - if (amdgpu_sync_add_later(dst, f)) { - kmem_cache_free(amdgpu_sync_slab, e); - continue; - } - - hash_add(dst->fences, &e->node, f->context); - } - - hash_add(src->fences, &newone->node, fence->context); - newone->fence = fence_get(fence); - - return 0; -} - struct fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync) { struct amdgpu_sync_entry *e; @@ -319,25 +296,6 @@ struct fence *amdgpu_sync_get_fence(struct amdgpu_sync *sync) return NULL; } -int amdgpu_sync_wait(struct amdgpu_sync *sync) -{ - struct amdgpu_sync_entry *e; - struct hlist_node *tmp; - int i, r; - - hash_for_each_safe(sync->fences, i, tmp, e, node) { - r = fence_wait(e->fence, false); - if (r) - return r; - - hash_del(&e->node); - fence_put(e->fence); - kmem_cache_free(amdgpu_sync_slab, e); - } - - return 0; -} - /** * amdgpu_sync_free - free the sync object * diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h index 26a5f4acf584..499803f3ce3a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h @@ -11,19 +11,68 @@ #define TRACE_SYSTEM amdgpu #define TRACE_INCLUDE_FILE amdgpu_trace +TRACE_EVENT(amdgpu_mm_rreg, + TP_PROTO(unsigned did, uint32_t reg, uint32_t value), + TP_ARGS(did, reg, value), + TP_STRUCT__entry( + __field(unsigned, did) + __field(uint32_t, reg) + __field(uint32_t, value) + ), + TP_fast_assign( + __entry->did = did; + __entry->reg = reg; + __entry->value = value; + ), + TP_printk("0x%04lx, 0x%04lx, 0x%08lx", + (unsigned long)__entry->did, + (unsigned long)__entry->reg, + (unsigned long)__entry->value) +); + +TRACE_EVENT(amdgpu_mm_wreg, + TP_PROTO(unsigned did, uint32_t reg, uint32_t value), + TP_ARGS(did, reg, value), + TP_STRUCT__entry( + __field(unsigned, did) + __field(uint32_t, reg) + __field(uint32_t, value) + ), + TP_fast_assign( + __entry->did = did; + __entry->reg = reg; + __entry->value = value; + ), + TP_printk("0x%04lx, 0x%04lx, 0x%08lx", + (unsigned long)__entry->did, + (unsigned long)__entry->reg, + (unsigned long)__entry->value) +); + TRACE_EVENT(amdgpu_bo_create, TP_PROTO(struct amdgpu_bo *bo), TP_ARGS(bo), TP_STRUCT__entry( __field(struct amdgpu_bo *, bo) __field(u32, pages) + __field(u32, type) + __field(u32, prefer) + __field(u32, allow) + __field(u32, visible) ), TP_fast_assign( __entry->bo = bo; __entry->pages = bo->tbo.num_pages; + __entry->type = bo->tbo.mem.mem_type; + __entry->prefer = bo->prefered_domains; + __entry->allow = bo->allowed_domains; + __entry->visible = bo->flags; ), - TP_printk("bo=%p, pages=%u", __entry->bo, __entry->pages) + + TP_printk("bo=%p,pages=%u,type=%d,prefered=%d,allowed=%d,visible=%d", + __entry->bo, __entry->pages, __entry->type, + __entry->prefer, __entry->allow, __entry->visible) ); TRACE_EVENT(amdgpu_cs, @@ -64,7 +113,7 @@ TRACE_EVENT(amdgpu_cs_ioctl, __entry->adev = job->adev; __entry->sched_job = &job->base; __entry->ib = job->ibs; - __entry->fence = &job->base.s_fence->base; + __entry->fence = &job->base.s_fence->finished; __entry->ring_name = job->ring->name; __entry->num_ibs = job->num_ibs; ), @@ -89,7 +138,7 @@ TRACE_EVENT(amdgpu_sched_run_job, __entry->adev = job->adev; __entry->sched_job = &job->base; __entry->ib = job->ibs; - __entry->fence = &job->base.s_fence->base; + __entry->fence = &job->base.s_fence->finished; __entry->ring_name = job->ring->name; __entry->num_ibs = job->num_ibs; ), @@ -244,13 +293,55 @@ TRACE_EVENT(amdgpu_bo_list_set, TP_STRUCT__entry( __field(struct amdgpu_bo_list *, list) __field(struct amdgpu_bo *, bo) + __field(u64, bo_size) ), TP_fast_assign( __entry->list = list; __entry->bo = bo; + __entry->bo_size = amdgpu_bo_size(bo); ), - TP_printk("list=%p, bo=%p", __entry->list, __entry->bo) + TP_printk("list=%p, bo=%p, bo_size = %Ld", + __entry->list, + __entry->bo, + __entry->bo_size) +); + +TRACE_EVENT(amdgpu_cs_bo_status, + TP_PROTO(uint64_t total_bo, uint64_t total_size), + TP_ARGS(total_bo, total_size), + TP_STRUCT__entry( + __field(u64, total_bo) + __field(u64, total_size) + ), + + TP_fast_assign( + __entry->total_bo = total_bo; + __entry->total_size = total_size; + ), + TP_printk("total bo size = %Ld, total bo count = %Ld", + __entry->total_bo, __entry->total_size) +); + +TRACE_EVENT(amdgpu_ttm_bo_move, + TP_PROTO(struct amdgpu_bo* bo, uint32_t new_placement, uint32_t old_placement), + TP_ARGS(bo, new_placement, old_placement), + TP_STRUCT__entry( + __field(struct amdgpu_bo *, bo) + __field(u64, bo_size) + __field(u32, new_placement) + __field(u32, old_placement) + ), + + TP_fast_assign( + __entry->bo = bo; + __entry->bo_size = amdgpu_bo_size(bo); + __entry->new_placement = new_placement; + __entry->old_placement = old_placement; + ), + TP_printk("bo=%p from:%d to %d with size = %Ld", + __entry->bo, __entry->old_placement, + __entry->new_placement, __entry->bo_size) ); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index 3b9053af4762..b7742e62972a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -286,9 +286,10 @@ static int amdgpu_move_blit(struct ttm_buffer_object *bo, r = amdgpu_copy_buffer(ring, old_start, new_start, new_mem->num_pages * PAGE_SIZE, /* bytes */ bo->resv, &fence); - /* FIXME: handle copy error */ - r = ttm_bo_move_accel_cleanup(bo, fence, - evict, no_wait_gpu, new_mem); + if (r) + return r; + + r = ttm_bo_pipeline_move(bo, fence, evict, new_mem); fence_put(fence); return r; } @@ -396,6 +397,11 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, return -EINVAL; adev = amdgpu_get_adev(bo->bdev); + + /* remember the eviction */ + if (evict) + atomic64_inc(&adev->num_evictions); + if (old_mem->mem_type == TTM_PL_SYSTEM && bo->ttm == NULL) { amdgpu_move_null(bo, new_mem); return 0; @@ -429,7 +435,8 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, if (r) { memcpy: - r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); + r = ttm_bo_move_memcpy(bo, evict, interruptible, + no_wait_gpu, new_mem); if (r) { return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c index 01abfc21b4a2..e19520c4b4b6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c @@ -253,19 +253,20 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev) { int r; - if (adev->uvd.vcpu_bo == NULL) - return 0; + kfree(adev->uvd.saved_bo); amd_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity); - r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false); - if (!r) { - amdgpu_bo_kunmap(adev->uvd.vcpu_bo); - amdgpu_bo_unpin(adev->uvd.vcpu_bo); - amdgpu_bo_unreserve(adev->uvd.vcpu_bo); - } + if (adev->uvd.vcpu_bo) { + r = amdgpu_bo_reserve(adev->uvd.vcpu_bo, false); + if (!r) { + amdgpu_bo_kunmap(adev->uvd.vcpu_bo); + amdgpu_bo_unpin(adev->uvd.vcpu_bo); + amdgpu_bo_unreserve(adev->uvd.vcpu_bo); + } - amdgpu_bo_unref(&adev->uvd.vcpu_bo); + amdgpu_bo_unref(&adev->uvd.vcpu_bo); + } amdgpu_ring_fini(&adev->uvd.ring); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 9f36ed30ba11..2f8496d48c94 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -25,6 +25,7 @@ * Alex Deucher * Jerome Glisse */ +#include <linux/fence-array.h> #include <drm/drmP.h> #include <drm/amdgpu_drm.h> #include "amdgpu.h" @@ -114,16 +115,26 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm, /** * amdgpu_vm_get_bos - add the vm BOs to a duplicates list * + * @adev: amdgpu device pointer * @vm: vm providing the BOs * @duplicates: head of duplicates list * * Add the page directory to the BO duplicates list * for command submission. */ -void amdgpu_vm_get_pt_bos(struct amdgpu_vm *vm, struct list_head *duplicates) +void amdgpu_vm_get_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm, + struct list_head *duplicates) { + uint64_t num_evictions; unsigned i; + /* We only need to validate the page tables + * if they aren't already valid. + */ + num_evictions = atomic64_read(&adev->num_evictions); + if (num_evictions == vm->last_eviction_counter) + return; + /* add the vm page table to the list */ for (i = 0; i <= vm->max_pde_used; ++i) { struct amdgpu_bo_list_entry *entry = &vm->page_tables[i].entry; @@ -162,6 +173,13 @@ void amdgpu_vm_move_pt_bos_in_lru(struct amdgpu_device *adev, spin_unlock(&glob->lru_lock); } +static bool amdgpu_vm_is_gpu_reset(struct amdgpu_device *adev, + struct amdgpu_vm_id *id) +{ + return id->current_gpu_reset_count != + atomic_read(&adev->gpu_reset_counter) ? true : false; +} + /** * amdgpu_vm_grab_id - allocate the next free VMID * @@ -174,20 +192,69 @@ void amdgpu_vm_move_pt_bos_in_lru(struct amdgpu_device *adev, */ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, struct amdgpu_sync *sync, struct fence *fence, - unsigned *vm_id, uint64_t *vm_pd_addr) + struct amdgpu_job *job) { - uint64_t pd_addr = amdgpu_bo_gpu_offset(vm->page_directory); struct amdgpu_device *adev = ring->adev; struct fence *updates = sync->last_vm_update; - struct amdgpu_vm_id *id; - unsigned i = ring->idx; - int r; + struct amdgpu_vm_id *id, *idle; + struct fence **fences; + unsigned i; + int r = 0; + + fences = kmalloc_array(sizeof(void *), adev->vm_manager.num_ids, + GFP_KERNEL); + if (!fences) + return -ENOMEM; mutex_lock(&adev->vm_manager.lock); + /* Check if we have an idle VMID */ + i = 0; + list_for_each_entry(idle, &adev->vm_manager.ids_lru, list) { + fences[i] = amdgpu_sync_peek_fence(&idle->active, ring); + if (!fences[i]) + break; + ++i; + } + + /* If we can't find a idle VMID to use, wait till one becomes available */ + if (&idle->list == &adev->vm_manager.ids_lru) { + u64 fence_context = adev->vm_manager.fence_context + ring->idx; + unsigned seqno = ++adev->vm_manager.seqno[ring->idx]; + struct fence_array *array; + unsigned j; + + for (j = 0; j < i; ++j) + fence_get(fences[j]); + + array = fence_array_create(i, fences, fence_context, + seqno, true); + if (!array) { + for (j = 0; j < i; ++j) + fence_put(fences[j]); + kfree(fences); + r = -ENOMEM; + goto error; + } + + + r = amdgpu_sync_fence(ring->adev, sync, &array->base); + fence_put(&array->base); + if (r) + goto error; + + mutex_unlock(&adev->vm_manager.lock); + return 0; + + } + kfree(fences); + + job->vm_needs_flush = true; /* Check if we can use a VMID already assigned to this VM */ + i = ring->idx; do { struct fence *flushed; + bool same_ring = ring->idx == i; id = vm->ids[i++]; if (i == AMDGPU_MAX_RINGS) @@ -196,67 +263,49 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, /* Check all the prerequisites to using this VMID */ if (!id) continue; + if (amdgpu_vm_is_gpu_reset(adev, id)) + continue; if (atomic64_read(&id->owner) != vm->client_id) continue; - if (pd_addr != id->pd_gpu_addr) + if (job->vm_pd_addr != id->pd_gpu_addr) continue; - if (id->last_user != ring && + if (!same_ring && (!id->last_flush || !fence_is_signaled(id->last_flush))) continue; flushed = id->flushed_updates; - if (updates && (!flushed || fence_is_later(updates, flushed))) + if (updates && + (!flushed || fence_is_later(updates, flushed))) continue; - /* Good we can use this VMID */ - if (id->last_user == ring) { - r = amdgpu_sync_fence(ring->adev, sync, - id->first); - if (r) - goto error; - } - - /* And remember this submission as user of the VMID */ + /* Good we can use this VMID. Remember this submission as + * user of the VMID. + */ r = amdgpu_sync_fence(ring->adev, &id->active, fence); if (r) goto error; + id->current_gpu_reset_count = atomic_read(&adev->gpu_reset_counter); list_move_tail(&id->list, &adev->vm_manager.ids_lru); vm->ids[ring->idx] = id; - *vm_id = id - adev->vm_manager.ids; - *vm_pd_addr = AMDGPU_VM_NO_FLUSH; - trace_amdgpu_vm_grab_id(vm, ring->idx, *vm_id, *vm_pd_addr); + job->vm_id = id - adev->vm_manager.ids; + job->vm_needs_flush = false; + trace_amdgpu_vm_grab_id(vm, ring->idx, job->vm_id, job->vm_pd_addr); mutex_unlock(&adev->vm_manager.lock); return 0; } while (i != ring->idx); - id = list_first_entry(&adev->vm_manager.ids_lru, - struct amdgpu_vm_id, - list); + /* Still no ID to use? Then use the idle one found earlier */ + id = idle; - if (!amdgpu_sync_is_idle(&id->active)) { - struct list_head *head = &adev->vm_manager.ids_lru; - struct amdgpu_vm_id *tmp; - - list_for_each_entry_safe(id, tmp, &adev->vm_manager.ids_lru, - list) { - if (amdgpu_sync_is_idle(&id->active)) { - list_move(&id->list, head); - head = &id->list; - } - } - id = list_first_entry(&adev->vm_manager.ids_lru, - struct amdgpu_vm_id, - list); - } - - r = amdgpu_sync_cycle_fences(sync, &id->active, fence); + /* Remember this submission as user of the VMID */ + r = amdgpu_sync_fence(ring->adev, &id->active, fence); if (r) goto error; @@ -269,22 +318,46 @@ int amdgpu_vm_grab_id(struct amdgpu_vm *vm, struct amdgpu_ring *ring, fence_put(id->flushed_updates); id->flushed_updates = fence_get(updates); - id->pd_gpu_addr = pd_addr; - + id->pd_gpu_addr = job->vm_pd_addr; + id->current_gpu_reset_count = atomic_read(&adev->gpu_reset_counter); list_move_tail(&id->list, &adev->vm_manager.ids_lru); - id->last_user = ring; atomic64_set(&id->owner, vm->client_id); vm->ids[ring->idx] = id; - *vm_id = id - adev->vm_manager.ids; - *vm_pd_addr = pd_addr; - trace_amdgpu_vm_grab_id(vm, ring->idx, *vm_id, *vm_pd_addr); + job->vm_id = id - adev->vm_manager.ids; + trace_amdgpu_vm_grab_id(vm, ring->idx, job->vm_id, job->vm_pd_addr); error: mutex_unlock(&adev->vm_manager.lock); return r; } +static bool amdgpu_vm_ring_has_compute_vm_bug(struct amdgpu_ring *ring) +{ + struct amdgpu_device *adev = ring->adev; + const struct amdgpu_ip_block_version *ip_block; + + if (ring->type != AMDGPU_RING_TYPE_COMPUTE) + /* only compute rings */ + return false; + + ip_block = amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_GFX); + if (!ip_block) + return false; + + if (ip_block->major <= 7) { + /* gfx7 has no workaround */ + return true; + } else if (ip_block->major == 8) { + if (adev->gfx.mec_fw_version >= 673) + /* gfx8 is fixed in MEC firmware 673 */ + return false; + else + return true; + } + return false; +} + /** * amdgpu_vm_flush - hardware flush the vm * @@ -294,59 +367,52 @@ error: * * Emit a VM flush when it is necessary. */ -int amdgpu_vm_flush(struct amdgpu_ring *ring, - unsigned vm_id, uint64_t pd_addr, - uint32_t gds_base, uint32_t gds_size, - uint32_t gws_base, uint32_t gws_size, - uint32_t oa_base, uint32_t oa_size) +int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job) { struct amdgpu_device *adev = ring->adev; - struct amdgpu_vm_id *id = &adev->vm_manager.ids[vm_id]; + struct amdgpu_vm_id *id = &adev->vm_manager.ids[job->vm_id]; bool gds_switch_needed = ring->funcs->emit_gds_switch && ( - id->gds_base != gds_base || - id->gds_size != gds_size || - id->gws_base != gws_base || - id->gws_size != gws_size || - id->oa_base != oa_base || - id->oa_size != oa_size); + id->gds_base != job->gds_base || + id->gds_size != job->gds_size || + id->gws_base != job->gws_base || + id->gws_size != job->gws_size || + id->oa_base != job->oa_base || + id->oa_size != job->oa_size); int r; if (ring->funcs->emit_pipeline_sync && ( - pd_addr != AMDGPU_VM_NO_FLUSH || gds_switch_needed || - ring->type == AMDGPU_RING_TYPE_COMPUTE)) + job->vm_needs_flush || gds_switch_needed || + amdgpu_vm_ring_has_compute_vm_bug(ring))) amdgpu_ring_emit_pipeline_sync(ring); - if (ring->funcs->emit_vm_flush && - pd_addr != AMDGPU_VM_NO_FLUSH) { + if (ring->funcs->emit_vm_flush && (job->vm_needs_flush || + amdgpu_vm_is_gpu_reset(adev, id))) { struct fence *fence; - trace_amdgpu_vm_flush(pd_addr, ring->idx, vm_id); - amdgpu_ring_emit_vm_flush(ring, vm_id, pd_addr); + trace_amdgpu_vm_flush(job->vm_pd_addr, ring->idx, job->vm_id); + amdgpu_ring_emit_vm_flush(ring, job->vm_id, job->vm_pd_addr); + + r = amdgpu_fence_emit(ring, &fence); + if (r) + return r; mutex_lock(&adev->vm_manager.lock); - if ((id->pd_gpu_addr == pd_addr) && (id->last_user == ring)) { - r = amdgpu_fence_emit(ring, &fence); - if (r) { - mutex_unlock(&adev->vm_manager.lock); - return r; - } - fence_put(id->last_flush); - id->last_flush = fence; - } + fence_put(id->last_flush); + id->last_flush = fence; mutex_unlock(&adev->vm_manager.lock); } if (gds_switch_needed) { - id->gds_base = gds_base; - id->gds_size = gds_size; - id->gws_base = gws_base; - id->gws_size = gws_size; - id->oa_base = oa_base; - id->oa_size = oa_size; - amdgpu_ring_emit_gds_switch(ring, vm_id, - gds_base, gds_size, - gws_base, gws_size, - oa_base, oa_size); + id->gds_base = job->gds_base; + id->gds_size = job->gds_size; + id->gws_base = job->gws_base; + id->gws_size = job->gws_size; + id->oa_base = job->oa_base; + id->oa_size = job->oa_size; + amdgpu_ring_emit_gds_switch(ring, job->vm_id, + job->gds_base, job->gds_size, + job->gws_base, job->gws_size, + job->oa_base, job->oa_size); } return 0; @@ -723,7 +789,7 @@ static void amdgpu_vm_frag_ptes(struct amdgpu_device *adev, * @vm: requested vm * @start: start of GPU address range * @end: end of GPU address range - * @dst: destination address to map to + * @dst: destination address to map to, the next dst inside the function * @flags: mapping flags * * Update the page tables in the range @start - @end. @@ -737,49 +803,75 @@ static void amdgpu_vm_update_ptes(struct amdgpu_device *adev, { const uint64_t mask = AMDGPU_VM_PTE_COUNT - 1; - uint64_t last_pe_start = ~0, last_pe_end = ~0, last_dst = ~0; - uint64_t addr; + uint64_t cur_pe_start, cur_pe_end, cur_dst; + uint64_t addr; /* next GPU address to be updated */ + uint64_t pt_idx; + struct amdgpu_bo *pt; + unsigned nptes; /* next number of ptes to be updated */ + uint64_t next_pe_start; + + /* initialize the variables */ + addr = start; + pt_idx = addr >> amdgpu_vm_block_size; + pt = vm->page_tables[pt_idx].entry.robj; + + if ((addr & ~mask) == (end & ~mask)) + nptes = end - addr; + else + nptes = AMDGPU_VM_PTE_COUNT - (addr & mask); + + cur_pe_start = amdgpu_bo_gpu_offset(pt); + cur_pe_start += (addr & mask) * 8; + cur_pe_end = cur_pe_start + 8 * nptes; + cur_dst = dst; + + /* for next ptb*/ + addr += nptes; + dst += nptes * AMDGPU_GPU_PAGE_SIZE; /* walk over the address space and update the page tables */ - for (addr = start; addr < end; ) { - uint64_t pt_idx = addr >> amdgpu_vm_block_size; - struct amdgpu_bo *pt = vm->page_tables[pt_idx].entry.robj; - unsigned nptes; - uint64_t pe_start; + while (addr < end) { + pt_idx = addr >> amdgpu_vm_block_size; + pt = vm->page_tables[pt_idx].entry.robj; if ((addr & ~mask) == (end & ~mask)) nptes = end - addr; else nptes = AMDGPU_VM_PTE_COUNT - (addr & mask); - pe_start = amdgpu_bo_gpu_offset(pt); - pe_start += (addr & mask) * 8; - - if (last_pe_end != pe_start) { + next_pe_start = amdgpu_bo_gpu_offset(pt); + next_pe_start += (addr & mask) * 8; + if (cur_pe_end == next_pe_start) { + /* The next ptb is consecutive to current ptb. + * Don't call amdgpu_vm_frag_ptes now. + * Will update two ptbs together in future. + */ + cur_pe_end += 8 * nptes; + } else { amdgpu_vm_frag_ptes(adev, vm_update_params, - last_pe_start, last_pe_end, - last_dst, flags); + cur_pe_start, cur_pe_end, + cur_dst, flags); - last_pe_start = pe_start; - last_pe_end = pe_start + 8 * nptes; - last_dst = dst; - } else { - last_pe_end += 8 * nptes; + cur_pe_start = next_pe_start; + cur_pe_end = next_pe_start + 8 * nptes; + cur_dst = dst; } + /* for next ptb*/ addr += nptes; dst += nptes * AMDGPU_GPU_PAGE_SIZE; } - amdgpu_vm_frag_ptes(adev, vm_update_params, last_pe_start, - last_pe_end, last_dst, flags); + amdgpu_vm_frag_ptes(adev, vm_update_params, cur_pe_start, + cur_pe_end, cur_dst, flags); } /** * amdgpu_vm_bo_update_mapping - update a mapping in the vm page table * * @adev: amdgpu_device pointer + * @exclusive: fence we need to sync to * @src: address where to copy page table entries from * @pages_addr: DMA addresses to use for mapping * @vm: requested vm @@ -793,6 +885,7 @@ static void amdgpu_vm_update_ptes(struct amdgpu_device *adev, * Returns 0 for success, -EINVAL for failure. */ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, + struct fence *exclusive, uint64_t src, dma_addr_t *pages_addr, struct amdgpu_vm *vm, @@ -853,6 +946,10 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev, vm_update_params.ib = &job->ibs[0]; + r = amdgpu_sync_fence(adev, &job->sync, exclusive); + if (r) + goto error_free; + r = amdgpu_sync_resv(adev, &job->sync, vm->page_directory->tbo.resv, owner); if (r) @@ -889,6 +986,7 @@ error_free: * amdgpu_vm_bo_split_mapping - split a mapping into smaller chunks * * @adev: amdgpu_device pointer + * @exclusive: fence we need to sync to * @gtt_flags: flags as they are used for GTT * @pages_addr: DMA addresses to use for mapping * @vm: requested vm @@ -902,6 +1000,7 @@ error_free: * Returns 0 for success, -EINVAL for failure. */ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, + struct fence *exclusive, uint32_t gtt_flags, dma_addr_t *pages_addr, struct amdgpu_vm *vm, @@ -932,7 +1031,8 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, addr += mapping->offset; if (!pages_addr || src) - return amdgpu_vm_bo_update_mapping(adev, src, pages_addr, vm, + return amdgpu_vm_bo_update_mapping(adev, exclusive, + src, pages_addr, vm, start, mapping->it.last, flags, addr, fence); @@ -940,7 +1040,8 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev, uint64_t last; last = min((uint64_t)mapping->it.last, start + max_size - 1); - r = amdgpu_vm_bo_update_mapping(adev, src, pages_addr, vm, + r = amdgpu_vm_bo_update_mapping(adev, exclusive, + src, pages_addr, vm, start, last, flags, addr, fence); if (r) @@ -973,6 +1074,7 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va_mapping *mapping; dma_addr_t *pages_addr = NULL; uint32_t gtt_flags, flags; + struct fence *exclusive; uint64_t addr; int r; @@ -994,8 +1096,11 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, default: break; } + + exclusive = reservation_object_get_excl(bo_va->bo->tbo.resv); } else { addr = 0; + exclusive = NULL; } flags = amdgpu_ttm_tt_pte_flags(adev, bo_va->bo->tbo.ttm, mem); @@ -1007,7 +1112,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, spin_unlock(&vm->status_lock); list_for_each_entry(mapping, &bo_va->invalids, list) { - r = amdgpu_vm_bo_split_mapping(adev, gtt_flags, pages_addr, vm, + r = amdgpu_vm_bo_split_mapping(adev, exclusive, + gtt_flags, pages_addr, vm, mapping, flags, addr, &bo_va->last_pt_update); if (r) @@ -1054,7 +1160,7 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev, struct amdgpu_bo_va_mapping, list); list_del(&mapping->list); - r = amdgpu_vm_bo_split_mapping(adev, 0, NULL, vm, mapping, + r = amdgpu_vm_bo_split_mapping(adev, NULL, 0, NULL, vm, mapping, 0, 0, NULL); kfree(mapping); if (r) @@ -1445,6 +1551,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) amdgpu_bo_unreserve(vm->page_directory); if (r) goto error_free_page_directory; + vm->last_eviction_counter = atomic64_read(&adev->num_evictions); return 0; @@ -1516,6 +1623,10 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) &adev->vm_manager.ids_lru); } + adev->vm_manager.fence_context = fence_context_alloc(AMDGPU_MAX_RINGS); + for (i = 0; i < AMDGPU_MAX_RINGS; ++i) + adev->vm_manager.seqno[i] = 0; + atomic_set(&adev->vm_manager.vm_pte_next_ring, 0); atomic64_set(&adev->vm_manager.client_counter, 0); } diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c index ea407db1fbcf..5c33ed862695 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c @@ -50,7 +50,9 @@ #include "gmc/gmc_7_1_sh_mask.h" MODULE_FIRMWARE("radeon/bonaire_smc.bin"); +MODULE_FIRMWARE("radeon/bonaire_k_smc.bin"); MODULE_FIRMWARE("radeon/hawaii_smc.bin"); +MODULE_FIRMWARE("radeon/hawaii_k_smc.bin"); #define MC_CG_ARB_FREQ_F0 0x0a #define MC_CG_ARB_FREQ_F1 0x0b @@ -736,19 +738,19 @@ static int ci_enable_didt(struct amdgpu_device *adev, bool enable) if (pi->caps_sq_ramping || pi->caps_db_ramping || pi->caps_td_ramping || pi->caps_tcp_ramping) { - gfx_v7_0_enter_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->enter_safe_mode(adev); if (enable) { ret = ci_program_pt_config_registers(adev, didt_config_ci); if (ret) { - gfx_v7_0_exit_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->exit_safe_mode(adev); return ret; } } ci_do_enable_didt(adev, enable); - gfx_v7_0_exit_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->exit_safe_mode(adev); } return 0; @@ -3636,6 +3638,10 @@ static int ci_setup_default_dpm_tables(struct amdgpu_device *adev) ci_setup_default_pcie_tables(adev); + /* save a copy of the default DPM table */ + memcpy(&(pi->golden_dpm_table), &(pi->dpm_table), + sizeof(struct ci_dpm_table)); + return 0; } @@ -5754,10 +5760,18 @@ static int ci_dpm_init_microcode(struct amdgpu_device *adev) switch (adev->asic_type) { case CHIP_BONAIRE: - chip_name = "bonaire"; + if ((adev->pdev->revision == 0x80) || + (adev->pdev->revision == 0x81) || + (adev->pdev->device == 0x665f)) + chip_name = "bonaire_k"; + else + chip_name = "bonaire"; break; case CHIP_HAWAII: - chip_name = "hawaii"; + if (adev->pdev->revision == 0x80) + chip_name = "hawaii_k"; + else + chip_name = "hawaii"; break; case CHIP_KAVERI: case CHIP_KABINI: @@ -6221,6 +6235,9 @@ static int ci_dpm_sw_fini(void *handle) ci_dpm_fini(adev); mutex_unlock(&adev->pm.mutex); + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + return 0; } @@ -6401,6 +6418,186 @@ static int ci_dpm_set_powergating_state(void *handle, return 0; } +static int ci_dpm_print_clock_levels(struct amdgpu_device *adev, + enum pp_clock_type type, char *buf) +{ + struct ci_power_info *pi = ci_get_pi(adev); + struct ci_single_dpm_table *sclk_table = &pi->dpm_table.sclk_table; + struct ci_single_dpm_table *mclk_table = &pi->dpm_table.mclk_table; + struct ci_single_dpm_table *pcie_table = &pi->dpm_table.pcie_speed_table; + + int i, now, size = 0; + uint32_t clock, pcie_speed; + + switch (type) { + case PP_SCLK: + amdgpu_ci_send_msg_to_smc(adev, PPSMC_MSG_API_GetSclkFrequency); + clock = RREG32(mmSMC_MSG_ARG_0); + + for (i = 0; i < sclk_table->count; i++) { + if (clock > sclk_table->dpm_levels[i].value) + continue; + break; + } + now = i; + + for (i = 0; i < sclk_table->count; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, sclk_table->dpm_levels[i].value / 100, + (i == now) ? "*" : ""); + break; + case PP_MCLK: + amdgpu_ci_send_msg_to_smc(adev, PPSMC_MSG_API_GetMclkFrequency); + clock = RREG32(mmSMC_MSG_ARG_0); + + for (i = 0; i < mclk_table->count; i++) { + if (clock > mclk_table->dpm_levels[i].value) + continue; + break; + } + now = i; + + for (i = 0; i < mclk_table->count; i++) + size += sprintf(buf + size, "%d: %uMhz %s\n", + i, mclk_table->dpm_levels[i].value / 100, + (i == now) ? "*" : ""); + break; + case PP_PCIE: + pcie_speed = ci_get_current_pcie_speed(adev); + for (i = 0; i < pcie_table->count; i++) { + if (pcie_speed != pcie_table->dpm_levels[i].value) + continue; + break; + } + now = i; + + for (i = 0; i < pcie_table->count; i++) + size += sprintf(buf + size, "%d: %s %s\n", i, + (pcie_table->dpm_levels[i].value == 0) ? "2.5GB, x1" : + (pcie_table->dpm_levels[i].value == 1) ? "5.0GB, x16" : + (pcie_table->dpm_levels[i].value == 2) ? "8.0GB, x16" : "", + (i == now) ? "*" : ""); + break; + default: + break; + } + + return size; +} + +static int ci_dpm_force_clock_level(struct amdgpu_device *adev, + enum pp_clock_type type, uint32_t mask) +{ + struct ci_power_info *pi = ci_get_pi(adev); + + if (adev->pm.dpm.forced_level + != AMDGPU_DPM_FORCED_LEVEL_MANUAL) + return -EINVAL; + + switch (type) { + case PP_SCLK: + if (!pi->sclk_dpm_key_disabled) + amdgpu_ci_send_msg_to_smc_with_parameter(adev, + PPSMC_MSG_SCLKDPM_SetEnabledMask, + pi->dpm_level_enable_mask.sclk_dpm_enable_mask & mask); + break; + + case PP_MCLK: + if (!pi->mclk_dpm_key_disabled) + amdgpu_ci_send_msg_to_smc_with_parameter(adev, + PPSMC_MSG_MCLKDPM_SetEnabledMask, + pi->dpm_level_enable_mask.mclk_dpm_enable_mask & mask); + break; + + case PP_PCIE: + { + uint32_t tmp = mask & pi->dpm_level_enable_mask.pcie_dpm_enable_mask; + uint32_t level = 0; + + while (tmp >>= 1) + level++; + + if (!pi->pcie_dpm_key_disabled) + amdgpu_ci_send_msg_to_smc_with_parameter(adev, + PPSMC_MSG_PCIeDPM_ForceLevel, + level); + break; + } + default: + break; + } + + return 0; +} + +static int ci_dpm_get_sclk_od(struct amdgpu_device *adev) +{ + struct ci_power_info *pi = ci_get_pi(adev); + struct ci_single_dpm_table *sclk_table = &(pi->dpm_table.sclk_table); + struct ci_single_dpm_table *golden_sclk_table = + &(pi->golden_dpm_table.sclk_table); + int value; + + value = (sclk_table->dpm_levels[sclk_table->count - 1].value - + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value) * + 100 / + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return value; +} + +static int ci_dpm_set_sclk_od(struct amdgpu_device *adev, uint32_t value) +{ + struct ci_power_info *pi = ci_get_pi(adev); + struct ci_ps *ps = ci_get_ps(adev->pm.dpm.requested_ps); + struct ci_single_dpm_table *golden_sclk_table = + &(pi->golden_dpm_table.sclk_table); + + if (value > 20) + value = 20; + + ps->performance_levels[ps->performance_level_count - 1].sclk = + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value * + value / 100 + + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return 0; +} + +static int ci_dpm_get_mclk_od(struct amdgpu_device *adev) +{ + struct ci_power_info *pi = ci_get_pi(adev); + struct ci_single_dpm_table *mclk_table = &(pi->dpm_table.mclk_table); + struct ci_single_dpm_table *golden_mclk_table = + &(pi->golden_dpm_table.mclk_table); + int value; + + value = (mclk_table->dpm_levels[mclk_table->count - 1].value - + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value) * + 100 / + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return value; +} + +static int ci_dpm_set_mclk_od(struct amdgpu_device *adev, uint32_t value) +{ + struct ci_power_info *pi = ci_get_pi(adev); + struct ci_ps *ps = ci_get_ps(adev->pm.dpm.requested_ps); + struct ci_single_dpm_table *golden_mclk_table = + &(pi->golden_dpm_table.mclk_table); + + if (value > 20) + value = 20; + + ps->performance_levels[ps->performance_level_count - 1].mclk = + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value * + value / 100 + + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return 0; +} + const struct amd_ip_funcs ci_dpm_ip_funcs = { .name = "ci_dpm", .early_init = ci_dpm_early_init, @@ -6435,6 +6632,12 @@ static const struct amdgpu_dpm_funcs ci_dpm_funcs = { .get_fan_control_mode = &ci_dpm_get_fan_control_mode, .set_fan_speed_percent = &ci_dpm_set_fan_speed_percent, .get_fan_speed_percent = &ci_dpm_get_fan_speed_percent, + .print_clock_levels = ci_dpm_print_clock_levels, + .force_clock_level = ci_dpm_force_clock_level, + .get_sclk_od = ci_dpm_get_sclk_od, + .set_sclk_od = ci_dpm_set_sclk_od, + .get_mclk_od = ci_dpm_get_mclk_od, + .set_mclk_od = ci_dpm_set_mclk_od, }; static void ci_dpm_set_dpm_funcs(struct amdgpu_device *adev) diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.h b/drivers/gpu/drm/amd/amdgpu/ci_dpm.h index faccc30c93bf..91be2996ae7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.h +++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.h @@ -193,6 +193,7 @@ struct ci_pt_defaults { struct ci_power_info { struct ci_dpm_table dpm_table; + struct ci_dpm_table golden_dpm_table; u32 voltage_control; u32 mvdd_control; u32 vddci_control; diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c index 07bc795a4ca9..a7de4d18ac94 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik.c +++ b/drivers/gpu/drm/amd/amdgpu/cik.c @@ -962,6 +962,12 @@ static bool cik_read_bios_from_rom(struct amdgpu_device *adev, return true; } +static u32 cik_get_virtual_caps(struct amdgpu_device *adev) +{ + /* CIK does not support SR-IOV */ + return 0; +} + static const struct amdgpu_allowed_register_entry cik_allowed_read_registers[] = { {mmGRBM_STATUS, false}, {mmGB_ADDR_CONFIG, false}, @@ -1029,12 +1035,12 @@ static uint32_t cik_read_indexed_register(struct amdgpu_device *adev, mutex_lock(&adev->grbm_idx_mutex); if (se_num != 0xffffffff || sh_num != 0xffffffff) - gfx_v7_0_select_se_sh(adev, se_num, sh_num); + amdgpu_gfx_select_se_sh(adev, se_num, sh_num, 0xffffffff); val = RREG32(reg_offset); if (se_num != 0xffffffff || sh_num != 0xffffffff) - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); return val; } @@ -1152,10 +1158,11 @@ static void kv_restore_regs_for_reset(struct amdgpu_device *adev, WREG32(mmGMCON_RENG_EXECUTE, save->gmcon_reng_execute); } -static void cik_gpu_pci_config_reset(struct amdgpu_device *adev) +static int cik_gpu_pci_config_reset(struct amdgpu_device *adev) { struct kv_reset_save_regs kv_save = { 0 }; u32 i; + int r = -EINVAL; dev_info(adev->dev, "GPU pci config reset\n"); @@ -1171,14 +1178,20 @@ static void cik_gpu_pci_config_reset(struct amdgpu_device *adev) /* wait for asic to come out of reset */ for (i = 0; i < adev->usec_timeout; i++) { - if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) + if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) { + /* enable BM */ + pci_set_master(adev->pdev); + r = 0; break; + } udelay(1); } /* does asic init need to be run first??? */ if (adev->flags & AMD_IS_APU) kv_restore_regs_for_reset(adev, &kv_save); + + return r; } static void cik_set_bios_scratch_engine_hung(struct amdgpu_device *adev, bool hung) @@ -1204,13 +1217,14 @@ static void cik_set_bios_scratch_engine_hung(struct amdgpu_device *adev, bool hu */ static int cik_asic_reset(struct amdgpu_device *adev) { + int r; cik_set_bios_scratch_engine_hung(adev, true); - cik_gpu_pci_config_reset(adev); + r = cik_gpu_pci_config_reset(adev); cik_set_bios_scratch_engine_hung(adev, false); - return 0; + return r; } static int cik_set_uvd_clock(struct amdgpu_device *adev, u32 clock, @@ -2007,9 +2021,7 @@ static const struct amdgpu_asic_funcs cik_asic_funcs = .get_xclk = &cik_get_xclk, .set_uvd_clocks = &cik_set_uvd_clocks, .set_vce_clocks = &cik_set_vce_clocks, - /* these should be moved to their own ip modules */ - .get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter, - .wait_for_mc_idle = &gmc_v7_0_mc_wait_for_idle, + .get_virtual_caps = &cik_get_virtual_caps, }; static int cik_common_early_init(void *handle) diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c index 518dca43b133..46aca16a40aa 100644 --- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c +++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c @@ -66,6 +66,16 @@ MODULE_FIRMWARE("radeon/mullins_sdma1.bin"); u32 amdgpu_cik_gpu_check_soft_reset(struct amdgpu_device *adev); + +static void cik_sdma_free_microcode(struct amdgpu_device *adev) +{ + int i; + for (i = 0; i < adev->sdma.num_instances; i++) { + release_firmware(adev->sdma.instance[i].fw); + adev->sdma.instance[i].fw = NULL; + } +} + /* * sDMA - System DMA * Starting with CIK, the GPU has new asynchronous @@ -214,17 +224,6 @@ static void cik_sdma_ring_emit_ib(struct amdgpu_ring *ring, unsigned vm_id, bool ctx_switch) { u32 extra_bits = vm_id & 0xf; - u32 next_rptr = ring->wptr + 5; - - while ((next_rptr & 7) != 4) - next_rptr++; - - next_rptr += 4; - amdgpu_ring_write(ring, SDMA_PACKET(SDMA_OPCODE_WRITE, SDMA_WRITE_SUB_OPCODE_LINEAR, 0)); - amdgpu_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); - amdgpu_ring_write(ring, 1); /* number of DWs to follow */ - amdgpu_ring_write(ring, next_rptr); /* IB packet must end on a 8 DW boundary */ cik_sdma_ring_insert_nop(ring, (12 - (ring->wptr & 7)) % 8); @@ -419,6 +418,8 @@ static int cik_sdma_gfx_resume(struct amdgpu_device *adev) /* Initialize the ring buffer's read and write pointers */ WREG32(mmSDMA0_GFX_RB_RPTR + sdma_offsets[i], 0); WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_RPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_OFFSET + sdma_offsets[i], 0); /* set the wb address whether it's enabled or not */ WREG32(mmSDMA0_GFX_RB_RPTR_ADDR_HI + sdma_offsets[i], @@ -446,7 +447,12 @@ static int cik_sdma_gfx_resume(struct amdgpu_device *adev) WREG32(mmSDMA0_GFX_IB_CNTL + sdma_offsets[i], ib_cntl); ring->ready = true; + } + + cik_sdma_enable(adev, true); + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; r = amdgpu_ring_test_ring(ring); if (r) { ring->ready = false; @@ -529,8 +535,8 @@ static int cik_sdma_start(struct amdgpu_device *adev) if (r) return r; - /* unhalt the MEs */ - cik_sdma_enable(adev, true); + /* halt the engine before programing */ + cik_sdma_enable(adev, false); /* start the gfx rings and rlc compute queues */ r = cik_sdma_gfx_resume(adev); @@ -998,6 +1004,7 @@ static int cik_sdma_sw_fini(void *handle) for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + cik_sdma_free_microcode(adev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c index 933e425a8154..8ba07e79d4cb 100644 --- a/drivers/gpu/drm/amd/amdgpu/cz_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/cz_dpm.c @@ -2219,6 +2219,7 @@ static void cz_dpm_powergate_vce(struct amdgpu_device *adev, bool gate) } } } else { /*pi->caps_vce_pg*/ + pi->vce_power_gated = gate; cz_update_vce_dpm(adev); cz_enable_vce_dpm(adev, !gate); } diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index c90408bc0fde..d4bf133908b1 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -307,11 +307,10 @@ static void dce_v11_0_page_flip(struct amdgpu_device *adev, struct amdgpu_crtc *amdgpu_crtc = adev->mode_info.crtcs[crtc_id]; u32 tmp; - /* flip at hsync for async, default is vsync */ - /* use UPDATE_IMMEDIATE_EN instead for async? */ + /* flip immediate for async, default is vsync */ tmp = RREG32(mmGRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset); tmp = REG_SET_FIELD(tmp, GRPH_FLIP_CONTROL, - GRPH_SURFACE_UPDATE_H_RETRACE_EN, async ? 1 : 0); + GRPH_SURFACE_UPDATE_IMMEDIATE_EN, async ? 1 : 0); WREG32(mmGRPH_FLIP_CONTROL + amdgpu_crtc->crtc_offset, tmp); /* update the scanout addresses */ WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + amdgpu_crtc->crtc_offset, diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c index 300ff4aab0fd..4fdfab1e9200 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c @@ -526,36 +526,16 @@ static void dce_v8_0_stop_mc_access(struct amdgpu_device *adev, crtc_enabled = REG_GET_FIELD(RREG32(mmCRTC_CONTROL + crtc_offsets[i]), CRTC_CONTROL, CRTC_MASTER_EN); if (crtc_enabled) { -#if 0 - u32 frame_count; - int j; - +#if 1 save->crtc_enabled[i] = true; tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); if (REG_GET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN) == 0) { - amdgpu_display_vblank_wait(adev, i); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); + /*it is correct only for RGB ; black is 0*/ + WREG32(mmCRTC_BLANK_DATA_COLOR + crtc_offsets[i], 0); tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 1); WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - } - /* wait for the next frame */ - frame_count = amdgpu_display_vblank_get_counter(adev, i); - for (j = 0; j < adev->usec_timeout; j++) { - if (amdgpu_display_vblank_get_counter(adev, i) != frame_count) - break; - udelay(1); - } - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK) == 0) { - tmp = REG_SET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK, 1); - WREG32(mmGRPH_UPDATE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK) == 0) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK, 1); - WREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i], tmp); } + mdelay(20); #else /* XXX this is a hack to avoid strange behavior with EFI on certain systems */ WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); @@ -575,55 +555,22 @@ static void dce_v8_0_stop_mc_access(struct amdgpu_device *adev, static void dce_v8_0_resume_mc_access(struct amdgpu_device *adev, struct amdgpu_mode_mc_save *save) { - u32 tmp, frame_count; - int i, j; + u32 tmp; + int i; /* update crtc base addresses */ for (i = 0; i < adev->mode_info.num_crtc; i++) { WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], upper_32_bits(adev->mc.vram_start)); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS_HIGH + crtc_offsets[i], - upper_32_bits(adev->mc.vram_start)); WREG32(mmGRPH_PRIMARY_SURFACE_ADDRESS + crtc_offsets[i], (u32)adev->mc.vram_start); - WREG32(mmGRPH_SECONDARY_SURFACE_ADDRESS + crtc_offsets[i], - (u32)adev->mc.vram_start); if (save->crtc_enabled[i]) { - tmp = RREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE) != 3) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_MODE, MASTER_UPDATE_MODE, 3); - WREG32(mmMASTER_UPDATE_MODE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK)) { - tmp = REG_SET_FIELD(tmp, GRPH_UPDATE, GRPH_UPDATE_LOCK, 0); - WREG32(mmGRPH_UPDATE + crtc_offsets[i], tmp); - } - tmp = RREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK)) { - tmp = REG_SET_FIELD(tmp, MASTER_UPDATE_LOCK, MASTER_UPDATE_LOCK, 0); - WREG32(mmMASTER_UPDATE_LOCK + crtc_offsets[i], tmp); - } - for (j = 0; j < adev->usec_timeout; j++) { - tmp = RREG32(mmGRPH_UPDATE + crtc_offsets[i]); - if (REG_GET_FIELD(tmp, GRPH_UPDATE, GRPH_SURFACE_UPDATE_PENDING) == 0) - break; - udelay(1); - } tmp = RREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i]); tmp = REG_SET_FIELD(tmp, CRTC_BLANK_CONTROL, CRTC_BLANK_DATA_EN, 0); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 1); WREG32(mmCRTC_BLANK_CONTROL + crtc_offsets[i], tmp); - WREG32(mmCRTC_UPDATE_LOCK + crtc_offsets[i], 0); - /* wait for the next frame */ - frame_count = amdgpu_display_vblank_get_counter(adev, i); - for (j = 0; j < adev->usec_timeout; j++) { - if (amdgpu_display_vblank_get_counter(adev, i) != frame_count) - break; - udelay(1); - } } + mdelay(20); } WREG32(mmVGA_MEMORY_BASE_ADDRESS_HIGH, upper_32_bits(adev->mc.vram_start)); diff --git a/drivers/gpu/drm/amd/amdgpu/fiji_dpm.c b/drivers/gpu/drm/amd/amdgpu/fiji_dpm.c index 245cabf06575..ed03b75175d4 100644 --- a/drivers/gpu/drm/amd/amdgpu/fiji_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/fiji_dpm.c @@ -72,6 +72,11 @@ static int fiji_dpm_sw_init(void *handle) static int fiji_dpm_sw_fini(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 7f18a53ab53a..f6bd9465dbdc 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -991,6 +991,22 @@ out: return err; } +static void gfx_v7_0_free_microcode(struct amdgpu_device *adev) +{ + release_firmware(adev->gfx.pfp_fw); + adev->gfx.pfp_fw = NULL; + release_firmware(adev->gfx.me_fw); + adev->gfx.me_fw = NULL; + release_firmware(adev->gfx.ce_fw); + adev->gfx.ce_fw = NULL; + release_firmware(adev->gfx.mec_fw); + adev->gfx.mec_fw = NULL; + release_firmware(adev->gfx.mec2_fw); + adev->gfx.mec2_fw = NULL; + release_firmware(adev->gfx.rlc_fw); + adev->gfx.rlc_fw = NULL; +} + /** * gfx_v7_0_tiling_mode_table_init - init the hw tiling table * @@ -1567,9 +1583,15 @@ static void gfx_v7_0_tiling_mode_table_init(struct amdgpu_device *adev) * registers are instanced per SE or SH. 0xffffffff means * broadcast to all SEs or SHs (CIK). */ -void gfx_v7_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num) +static void gfx_v7_0_select_se_sh(struct amdgpu_device *adev, + u32 se_num, u32 sh_num, u32 instance) { - u32 data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK; + u32 data; + + if (instance == 0xffffffff) + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES, 1); + else + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_INDEX, instance); if ((se_num == 0xffffffff) && (sh_num == 0xffffffff)) data |= GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | @@ -1643,13 +1665,13 @@ static void gfx_v7_0_setup_rb(struct amdgpu_device *adev) mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v7_0_select_se_sh(adev, i, j); + gfx_v7_0_select_se_sh(adev, i, j, 0xffffffff); data = gfx_v7_0_get_rb_active_bitmap(adev); active_rbs |= data << ((i * adev->gfx.config.max_sh_per_se + j) * rb_bitmap_width_per_sh); } } - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); adev->gfx.config.backend_enable_mask = active_rbs; @@ -1730,7 +1752,7 @@ static void gfx_v7_0_gpu_init(struct amdgpu_device *adev) * making sure that the following register writes will be broadcasted * to all the shaders */ - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); /* XXX SH_MEM regs */ /* where to put LDS, scratch, GPUVM in FSA64 space */ @@ -2034,17 +2056,6 @@ static void gfx_v7_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, unsigned vm_id, bool ctx_switch) { u32 header, control = 0; - u32 next_rptr = ring->wptr + 5; - - if (ctx_switch) - next_rptr += 2; - - next_rptr += 4; - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, WRITE_DATA_DST_SEL(5) | WR_CONFIRM); - amdgpu_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); - amdgpu_ring_write(ring, next_rptr); /* insert SWITCH_BUFFER packet before first IB in the ring frame */ if (ctx_switch) { @@ -2073,22 +2084,9 @@ static void gfx_v7_0_ring_emit_ib_compute(struct amdgpu_ring *ring, struct amdgpu_ib *ib, unsigned vm_id, bool ctx_switch) { - u32 header, control = 0; - u32 next_rptr = ring->wptr + 5; - - control |= INDIRECT_BUFFER_VALID; - next_rptr += 4; - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, WRITE_DATA_DST_SEL(5) | WR_CONFIRM); - amdgpu_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); - amdgpu_ring_write(ring, next_rptr); - - header = PACKET3(PACKET3_INDIRECT_BUFFER, 2); - - control |= ib->length_dw | (vm_id << 24); + u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vm_id << 24); - amdgpu_ring_write(ring, header); + amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); amdgpu_ring_write(ring, #ifdef __BIG_ENDIAN (2 << 0) | @@ -3205,7 +3203,8 @@ static int gfx_v7_0_rlc_init(struct amdgpu_device *adev) } } adev->gfx.rlc.cs_data = ci_cs_data; - adev->gfx.rlc.cp_table_size = CP_ME_TABLE_SIZE * 5 * 4; + adev->gfx.rlc.cp_table_size = ALIGN(CP_ME_TABLE_SIZE * 5 * 4, 2048); /* CP JT */ + adev->gfx.rlc.cp_table_size += 64 * 1024; /* GDS */ src_ptr = adev->gfx.rlc.reg_list; dws = adev->gfx.rlc.reg_list_size; @@ -3363,7 +3362,7 @@ static void gfx_v7_0_wait_for_rlc_serdes(struct amdgpu_device *adev) mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v7_0_select_se_sh(adev, i, j); + gfx_v7_0_select_se_sh(adev, i, j, 0xffffffff); for (k = 0; k < adev->usec_timeout; k++) { if (RREG32(mmRLC_SERDES_CU_MASTER_BUSY) == 0) break; @@ -3371,7 +3370,7 @@ static void gfx_v7_0_wait_for_rlc_serdes(struct amdgpu_device *adev) } } } - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); mask = RLC_SERDES_NONCU_MASTER_BUSY__SE_MASTER_BUSY_MASK | @@ -3418,7 +3417,7 @@ static u32 gfx_v7_0_halt_rlc(struct amdgpu_device *adev) return orig; } -void gfx_v7_0_enter_rlc_safe_mode(struct amdgpu_device *adev) +static void gfx_v7_0_enter_rlc_safe_mode(struct amdgpu_device *adev) { u32 tmp, i, mask; @@ -3440,7 +3439,7 @@ void gfx_v7_0_enter_rlc_safe_mode(struct amdgpu_device *adev) } } -void gfx_v7_0_exit_rlc_safe_mode(struct amdgpu_device *adev) +static void gfx_v7_0_exit_rlc_safe_mode(struct amdgpu_device *adev) { u32 tmp; @@ -3455,7 +3454,7 @@ void gfx_v7_0_exit_rlc_safe_mode(struct amdgpu_device *adev) * * Halt the RLC ME (MicroEngine) (CIK). */ -void gfx_v7_0_rlc_stop(struct amdgpu_device *adev) +static void gfx_v7_0_rlc_stop(struct amdgpu_device *adev) { WREG32(mmRLC_CNTL, 0); @@ -3531,7 +3530,7 @@ static int gfx_v7_0_rlc_resume(struct amdgpu_device *adev) WREG32(mmRLC_LB_CNTR_MAX, 0x00008000); mutex_lock(&adev->grbm_idx_mutex); - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmRLC_LB_INIT_CU_MASK, 0xffffffff); WREG32(mmRLC_LB_PARAMS, 0x00600408); WREG32(mmRLC_LB_CNTL, 0x80000004); @@ -3571,7 +3570,7 @@ static void gfx_v7_0_enable_cgcg(struct amdgpu_device *adev, bool enable) tmp = gfx_v7_0_halt_rlc(adev); mutex_lock(&adev->grbm_idx_mutex); - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmRLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff); WREG32(mmRLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff); tmp2 = RLC_SERDES_WR_CTRL__BPM_ADDR_MASK | @@ -3622,7 +3621,7 @@ static void gfx_v7_0_enable_mgcg(struct amdgpu_device *adev, bool enable) tmp = gfx_v7_0_halt_rlc(adev); mutex_lock(&adev->grbm_idx_mutex); - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmRLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff); WREG32(mmRLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff); data = RLC_SERDES_WR_CTRL__BPM_ADDR_MASK | @@ -3673,7 +3672,7 @@ static void gfx_v7_0_enable_mgcg(struct amdgpu_device *adev, bool enable) tmp = gfx_v7_0_halt_rlc(adev); mutex_lock(&adev->grbm_idx_mutex); - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmRLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff); WREG32(mmRLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff); data = RLC_SERDES_WR_CTRL__BPM_ADDR_MASK | RLC_SERDES_WR_CTRL__MGCG_OVERRIDE_1_MASK; @@ -3851,6 +3850,20 @@ static void gfx_v7_0_enable_gfx_cgpg(struct amdgpu_device *adev, } } +static void gfx_v7_0_set_user_cu_inactive_bitmap(struct amdgpu_device *adev, + u32 bitmap) +{ + u32 data; + + if (!bitmap) + return; + + data = bitmap << GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT; + data &= GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK; + + WREG32(mmGC_USER_SHADER_ARRAY_CONFIG, data); +} + static u32 gfx_v7_0_get_cu_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; @@ -4107,7 +4120,7 @@ static void gfx_v7_0_fini_pg(struct amdgpu_device *adev) * Fetches a GPU clock counter snapshot (SI). * Returns the 64 bit clock counter snapshot. */ -uint64_t gfx_v7_0_get_gpu_clock_counter(struct amdgpu_device *adev) +static uint64_t gfx_v7_0_get_gpu_clock_counter(struct amdgpu_device *adev) { uint64_t clock; @@ -4167,12 +4180,24 @@ static void gfx_v7_0_ring_emit_gds_switch(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } +static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = { + .get_gpu_clock_counter = &gfx_v7_0_get_gpu_clock_counter, + .select_se_sh = &gfx_v7_0_select_se_sh, +}; + +static const struct amdgpu_rlc_funcs gfx_v7_0_rlc_funcs = { + .enter_safe_mode = gfx_v7_0_enter_rlc_safe_mode, + .exit_safe_mode = gfx_v7_0_exit_rlc_safe_mode +}; + static int gfx_v7_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; adev->gfx.num_gfx_rings = GFX7_NUM_GFX_RINGS; adev->gfx.num_compute_rings = GFX7_NUM_COMPUTE_RINGS; + adev->gfx.funcs = &gfx_v7_0_gfx_funcs; + adev->gfx.rlc.funcs = &gfx_v7_0_rlc_funcs; gfx_v7_0_set_ring_funcs(adev); gfx_v7_0_set_irq_funcs(adev); gfx_v7_0_set_gds_init(adev); @@ -4489,6 +4514,7 @@ static int gfx_v7_0_sw_fini(void *handle) gfx_v7_0_cp_compute_fini(adev); gfx_v7_0_rlc_fini(adev); gfx_v7_0_mec_fini(adev); + gfx_v7_0_free_microcode(adev); return 0; } @@ -4816,7 +4842,7 @@ static int gfx_v7_0_eop_irq(struct amdgpu_device *adev, case 2: for (i = 0; i < adev->gfx.num_compute_rings; i++) { ring = &adev->gfx.compute_ring[i]; - if ((ring->me == me_id) & (ring->pipe == pipe_id)) + if ((ring->me == me_id) && (ring->pipe == pipe_id)) amdgpu_fence_process(ring); } break; @@ -5015,16 +5041,22 @@ static void gfx_v7_0_get_cu_info(struct amdgpu_device *adev) int i, j, k, counter, active_cu_number = 0; u32 mask, bitmap, ao_bitmap, ao_cu_mask = 0; struct amdgpu_cu_info *cu_info = &adev->gfx.cu_info; + unsigned disable_masks[4 * 2]; memset(cu_info, 0, sizeof(*cu_info)); + amdgpu_gfx_parse_disable_cu(disable_masks, 4, 2); + mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { mask = 1; ao_bitmap = 0; counter = 0; - gfx_v7_0_select_se_sh(adev, i, j); + gfx_v7_0_select_se_sh(adev, i, j, 0xffffffff); + if (i < 4 && j < 2) + gfx_v7_0_set_user_cu_inactive_bitmap( + adev, disable_masks[i * 2 + j]); bitmap = gfx_v7_0_get_cu_active_bitmap(adev); cu_info->bitmap[i][j] = bitmap; @@ -5040,7 +5072,7 @@ static void gfx_v7_0_get_cu_info(struct amdgpu_device *adev) ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); } } - gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); cu_info->number = active_cu_number; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.h b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.h index e747aa935c88..94e3ea147c26 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.h +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.h @@ -26,11 +26,4 @@ extern const struct amd_ip_funcs gfx_v7_0_ip_funcs; -/* XXX these shouldn't be exported */ -void gfx_v7_0_enter_rlc_safe_mode(struct amdgpu_device *adev); -void gfx_v7_0_exit_rlc_safe_mode(struct amdgpu_device *adev); -void gfx_v7_0_rlc_stop(struct amdgpu_device *adev); -uint64_t gfx_v7_0_get_gpu_clock_counter(struct amdgpu_device *adev); -void gfx_v7_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num); - #endif diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index f19bab68fd83..c30b6ac25d89 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -297,7 +297,8 @@ static const u32 polaris11_golden_common_all[] = static const u32 golden_settings_polaris10_a11[] = { mmATC_MISC_CG, 0x000c0fc0, 0x000c0200, - mmCB_HW_CONTROL, 0xfffdf3cf, 0x00006208, + mmCB_HW_CONTROL, 0xfffdf3cf, 0x00007208, + mmCB_HW_CONTROL_2, 0, 0x0f000000, mmCB_HW_CONTROL_3, 0x000001ff, 0x00000040, mmDB_DEBUG2, 0xf00fffff, 0x00000400, mmPA_SC_ENHANCE, 0xffffffff, 0x20000001, @@ -836,6 +837,26 @@ err1: return r; } + +static void gfx_v8_0_free_microcode(struct amdgpu_device *adev) { + release_firmware(adev->gfx.pfp_fw); + adev->gfx.pfp_fw = NULL; + release_firmware(adev->gfx.me_fw); + adev->gfx.me_fw = NULL; + release_firmware(adev->gfx.ce_fw); + adev->gfx.ce_fw = NULL; + release_firmware(adev->gfx.rlc_fw); + adev->gfx.rlc_fw = NULL; + release_firmware(adev->gfx.mec_fw); + adev->gfx.mec_fw = NULL; + if ((adev->asic_type != CHIP_STONEY) && + (adev->asic_type != CHIP_TOPAZ)) + release_firmware(adev->gfx.mec2_fw); + adev->gfx.mec2_fw = NULL; + + kfree(adev->gfx.rlc.register_list_format); +} + static int gfx_v8_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; @@ -1129,6 +1150,71 @@ static void gfx_v8_0_get_csb_buffer(struct amdgpu_device *adev, buffer[count++] = cpu_to_le32(0); } +static void cz_init_cp_jump_table(struct amdgpu_device *adev) +{ + const __le32 *fw_data; + volatile u32 *dst_ptr; + int me, i, max_me = 4; + u32 bo_offset = 0; + u32 table_offset, table_size; + + if (adev->asic_type == CHIP_CARRIZO) + max_me = 5; + + /* write the cp table buffer */ + dst_ptr = adev->gfx.rlc.cp_table_ptr; + for (me = 0; me < max_me; me++) { + if (me == 0) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.ce_fw->data; + fw_data = (const __le32 *) + (adev->gfx.ce_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 1) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.pfp_fw->data; + fw_data = (const __le32 *) + (adev->gfx.pfp_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 2) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.me_fw->data; + fw_data = (const __le32 *) + (adev->gfx.me_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 3) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec_fw->data; + fw_data = (const __le32 *) + (adev->gfx.mec_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } else if (me == 4) { + const struct gfx_firmware_header_v1_0 *hdr = + (const struct gfx_firmware_header_v1_0 *)adev->gfx.mec2_fw->data; + fw_data = (const __le32 *) + (adev->gfx.mec2_fw->data + + le32_to_cpu(hdr->header.ucode_array_offset_bytes)); + table_offset = le32_to_cpu(hdr->jt_offset); + table_size = le32_to_cpu(hdr->jt_size); + } + + for (i = 0; i < table_size; i ++) { + dst_ptr[bo_offset + i] = + cpu_to_le32(le32_to_cpu(fw_data[table_offset + i])); + } + + bo_offset += table_size; + } +} + static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev) { int r; @@ -1144,6 +1230,18 @@ static void gfx_v8_0_rlc_fini(struct amdgpu_device *adev) amdgpu_bo_unref(&adev->gfx.rlc.clear_state_obj); adev->gfx.rlc.clear_state_obj = NULL; } + + /* jump table block */ + if (adev->gfx.rlc.cp_table_obj) { + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + if (unlikely(r != 0)) + dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); + amdgpu_bo_unpin(adev->gfx.rlc.cp_table_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj); + + amdgpu_bo_unref(&adev->gfx.rlc.cp_table_obj); + adev->gfx.rlc.cp_table_obj = NULL; + } } static int gfx_v8_0_rlc_init(struct amdgpu_device *adev) @@ -1200,6 +1298,46 @@ static int gfx_v8_0_rlc_init(struct amdgpu_device *adev) amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj); } + if ((adev->asic_type == CHIP_CARRIZO) || + (adev->asic_type == CHIP_STONEY)) { + adev->gfx.rlc.cp_table_size = ALIGN(96 * 5 * 4, 2048) + (64 * 1024); /* JT + GDS */ + if (adev->gfx.rlc.cp_table_obj == NULL) { + r = amdgpu_bo_create(adev, adev->gfx.rlc.cp_table_size, PAGE_SIZE, true, + AMDGPU_GEM_DOMAIN_VRAM, + AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED, + NULL, NULL, + &adev->gfx.rlc.cp_table_obj); + if (r) { + dev_warn(adev->dev, "(%d) create RLC cp table bo failed\n", r); + return r; + } + } + + r = amdgpu_bo_reserve(adev->gfx.rlc.cp_table_obj, false); + if (unlikely(r != 0)) { + dev_warn(adev->dev, "(%d) reserve RLC cp table bo failed\n", r); + return r; + } + r = amdgpu_bo_pin(adev->gfx.rlc.cp_table_obj, AMDGPU_GEM_DOMAIN_VRAM, + &adev->gfx.rlc.cp_table_gpu_addr); + if (r) { + amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj); + dev_warn(adev->dev, "(%d) pin RLC cp_table bo failed\n", r); + return r; + } + r = amdgpu_bo_kmap(adev->gfx.rlc.cp_table_obj, (void **)&adev->gfx.rlc.cp_table_ptr); + if (r) { + dev_warn(adev->dev, "(%d) map RLC cp table bo failed\n", r); + return r; + } + + cz_init_cp_jump_table(adev); + + amdgpu_bo_kunmap(adev->gfx.rlc.cp_table_obj); + amdgpu_bo_unreserve(adev->gfx.rlc.cp_table_obj); + + } + return 0; } @@ -1983,7 +2121,7 @@ static int gfx_v8_0_sw_fini(void *handle) gfx_v8_0_rlc_fini(adev); - kfree(adev->gfx.rlc.register_list_format); + gfx_v8_0_free_microcode(adev); return 0; } @@ -3308,9 +3446,15 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev) } } -void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num) +static void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, + u32 se_num, u32 sh_num, u32 instance) { - u32 data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES, 1); + u32 data; + + if (instance == 0xffffffff) + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_BROADCAST_WRITES, 1); + else + data = REG_SET_FIELD(0, GRBM_GFX_INDEX, INSTANCE_INDEX, instance); if ((se_num == 0xffffffff) && (sh_num == 0xffffffff)) { data = REG_SET_FIELD(data, GRBM_GFX_INDEX, SH_BROADCAST_WRITES, 1); @@ -3360,13 +3504,13 @@ static void gfx_v8_0_setup_rb(struct amdgpu_device *adev) mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v8_0_select_se_sh(adev, i, j); + gfx_v8_0_select_se_sh(adev, i, j, 0xffffffff); data = gfx_v8_0_get_rb_active_bitmap(adev); active_rbs |= data << ((i * adev->gfx.config.max_sh_per_se + j) * rb_bitmap_width_per_sh); } } - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); adev->gfx.config.backend_enable_mask = active_rbs; @@ -3470,7 +3614,7 @@ static void gfx_v8_0_gpu_init(struct amdgpu_device *adev) * making sure that the following register writes will be broadcasted * to all the shaders */ - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmPA_SC_FIFO_SIZE, (adev->gfx.config.sc_prim_fifo_size_frontend << @@ -3493,7 +3637,7 @@ static void gfx_v8_0_wait_for_rlc_serdes(struct amdgpu_device *adev) mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { - gfx_v8_0_select_se_sh(adev, i, j); + gfx_v8_0_select_se_sh(adev, i, j, 0xffffffff); for (k = 0; k < adev->usec_timeout; k++) { if (RREG32(mmRLC_SERDES_CU_MASTER_BUSY) == 0) break; @@ -3501,7 +3645,7 @@ static void gfx_v8_0_wait_for_rlc_serdes(struct amdgpu_device *adev) } } } - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); mask = RLC_SERDES_NONCU_MASTER_BUSY__SE_MASTER_BUSY_MASK | @@ -3662,13 +3806,13 @@ static void gfx_v8_0_enable_save_restore_machine(struct amdgpu_device *adev) WREG32(mmRLC_SRM_CNTL, data); } -static void polaris11_init_power_gating(struct amdgpu_device *adev) +static void gfx_v8_0_init_power_gating(struct amdgpu_device *adev) { uint32_t data; if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | - AMD_PG_SUPPORT_GFX_SMG | - AMD_PG_SUPPORT_GFX_DMG)) { + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG)) { data = RREG32(mmCP_RB_WPTR_POLL_CNTL); data &= ~CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT_MASK; data |= (0x60 << CP_RB_WPTR_POLL_CNTL__IDLE_POLL_COUNT__SHIFT); @@ -3693,6 +3837,53 @@ static void polaris11_init_power_gating(struct amdgpu_device *adev) } } +static void cz_enable_sck_slow_down_on_power_up(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(mmRLC_PG_CNTL); + + if (enable) + data |= RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PU_ENABLE_MASK; + else + data &= ~RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PU_ENABLE_MASK; + + if (orig != data) + WREG32(mmRLC_PG_CNTL, data); +} + +static void cz_enable_sck_slow_down_on_power_down(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(mmRLC_PG_CNTL); + + if (enable) + data |= RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PD_ENABLE_MASK; + else + data &= ~RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PD_ENABLE_MASK; + + if (orig != data) + WREG32(mmRLC_PG_CNTL, data); +} + +static void cz_enable_cp_power_gating(struct amdgpu_device *adev, bool enable) +{ + u32 data, orig; + + orig = data = RREG32(mmRLC_PG_CNTL); + + if (enable) + data &= ~RLC_PG_CNTL__CP_PG_DISABLE_MASK; + else + data |= RLC_PG_CNTL__CP_PG_DISABLE_MASK; + + if (orig != data) + WREG32(mmRLC_PG_CNTL, data); +} + static void gfx_v8_0_init_pg(struct amdgpu_device *adev) { if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG | @@ -3705,8 +3896,25 @@ static void gfx_v8_0_init_pg(struct amdgpu_device *adev) gfx_v8_0_init_save_restore_list(adev); gfx_v8_0_enable_save_restore_machine(adev); - if (adev->asic_type == CHIP_POLARIS11) - polaris11_init_power_gating(adev); + if ((adev->asic_type == CHIP_CARRIZO) || + (adev->asic_type == CHIP_STONEY)) { + WREG32(mmRLC_JUMP_TABLE_RESTORE, adev->gfx.rlc.cp_table_gpu_addr >> 8); + gfx_v8_0_init_power_gating(adev); + WREG32(mmRLC_PG_ALWAYS_ON_CU_MASK, adev->gfx.cu_info.ao_cu_mask); + if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) { + cz_enable_sck_slow_down_on_power_up(adev, true); + cz_enable_sck_slow_down_on_power_down(adev, true); + } else { + cz_enable_sck_slow_down_on_power_up(adev, false); + cz_enable_sck_slow_down_on_power_down(adev, false); + } + if (adev->pg_flags & AMD_PG_SUPPORT_CP) + cz_enable_cp_power_gating(adev, true); + else + cz_enable_cp_power_gating(adev, false); + } else if (adev->asic_type == CHIP_POLARIS11) { + gfx_v8_0_init_power_gating(adev); + } } } @@ -3974,11 +4182,15 @@ static int gfx_v8_0_cp_gfx_start(struct amdgpu_device *adev) amdgpu_ring_write(ring, 0x3a00161a); amdgpu_ring_write(ring, 0x0000002e); break; - case CHIP_TOPAZ: case CHIP_CARRIZO: amdgpu_ring_write(ring, 0x00000002); amdgpu_ring_write(ring, 0x00000000); break; + case CHIP_TOPAZ: + amdgpu_ring_write(ring, adev->gfx.config.num_rbs == 1 ? + 0x00000000 : 0x00000002); + amdgpu_ring_write(ring, 0x00000000); + break; case CHIP_STONEY: amdgpu_ring_write(ring, 0x00000000); amdgpu_ring_write(ring, 0x00000000); @@ -4941,7 +5153,7 @@ static int gfx_v8_0_soft_reset(void *handle) * Fetches a GPU clock counter snapshot. * Returns the 64 bit clock counter snapshot. */ -uint64_t gfx_v8_0_get_gpu_clock_counter(struct amdgpu_device *adev) +static uint64_t gfx_v8_0_get_gpu_clock_counter(struct amdgpu_device *adev) { uint64_t clock; @@ -5001,12 +5213,18 @@ static void gfx_v8_0_ring_emit_gds_switch(struct amdgpu_ring *ring, amdgpu_ring_write(ring, (1 << (oa_size + oa_base)) - (1 << oa_base)); } +static const struct amdgpu_gfx_funcs gfx_v8_0_gfx_funcs = { + .get_gpu_clock_counter = &gfx_v8_0_get_gpu_clock_counter, + .select_se_sh = &gfx_v8_0_select_se_sh, +}; + static int gfx_v8_0_early_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; adev->gfx.num_gfx_rings = GFX8_NUM_GFX_RINGS; adev->gfx.num_compute_rings = GFX8_NUM_COMPUTE_RINGS; + adev->gfx.funcs = &gfx_v8_0_gfx_funcs; gfx_v8_0_set_ring_funcs(adev); gfx_v8_0_set_irq_funcs(adev); gfx_v8_0_set_gds_init(adev); @@ -5039,51 +5257,43 @@ static int gfx_v8_0_late_init(void *handle) return 0; } -static void polaris11_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, - bool enable) +static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, + bool enable) { uint32_t data, temp; - /* Send msg to SMU via Powerplay */ - amdgpu_set_powergating_state(adev, - AMD_IP_BLOCK_TYPE_SMC, - enable ? AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE); + if (adev->asic_type == CHIP_POLARIS11) + /* Send msg to SMU via Powerplay */ + amdgpu_set_powergating_state(adev, + AMD_IP_BLOCK_TYPE_SMC, + enable ? + AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE); - if (enable) { - /* Enable static MGPG */ - temp = data = RREG32(mmRLC_PG_CNTL); + temp = data = RREG32(mmRLC_PG_CNTL); + /* Enable static MGPG */ + if (enable) data |= RLC_PG_CNTL__STATIC_PER_CU_PG_ENABLE_MASK; - - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); - } else { - temp = data = RREG32(mmRLC_PG_CNTL); + else data &= ~RLC_PG_CNTL__STATIC_PER_CU_PG_ENABLE_MASK; - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); - } + if (temp != data) + WREG32(mmRLC_PG_CNTL, data); } -static void polaris11_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *adev, - bool enable) +static void gfx_v8_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *adev, + bool enable) { uint32_t data, temp; - if (enable) { - /* Enable dynamic MGPG */ - temp = data = RREG32(mmRLC_PG_CNTL); + temp = data = RREG32(mmRLC_PG_CNTL); + /* Enable dynamic MGPG */ + if (enable) data |= RLC_PG_CNTL__DYN_PER_CU_PG_ENABLE_MASK; - - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); - } else { - temp = data = RREG32(mmRLC_PG_CNTL); + else data &= ~RLC_PG_CNTL__DYN_PER_CU_PG_ENABLE_MASK; - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); - } + if (temp != data) + WREG32(mmRLC_PG_CNTL, data); } static void polaris11_enable_gfx_quick_mg_power_gating(struct amdgpu_device *adev, @@ -5091,19 +5301,63 @@ static void polaris11_enable_gfx_quick_mg_power_gating(struct amdgpu_device *ade { uint32_t data, temp; - if (enable) { - /* Enable quick PG */ - temp = data = RREG32(mmRLC_PG_CNTL); - data |= 0x100000; + temp = data = RREG32(mmRLC_PG_CNTL); + /* Enable quick PG */ + if (enable) + data |= RLC_PG_CNTL__QUICK_PG_ENABLE_MASK; + else + data &= ~RLC_PG_CNTL__QUICK_PG_ENABLE_MASK; - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); - } else { - temp = data = RREG32(mmRLC_PG_CNTL); - data &= ~0x100000; + if (temp != data) + WREG32(mmRLC_PG_CNTL, data); +} - if (temp != data) - WREG32(mmRLC_PG_CNTL, data); +static void cz_enable_gfx_cg_power_gating(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(mmRLC_PG_CNTL); + + if (enable) + data |= RLC_PG_CNTL__GFX_POWER_GATING_ENABLE_MASK; + else + data &= ~RLC_PG_CNTL__GFX_POWER_GATING_ENABLE_MASK; + + if (orig != data) + WREG32(mmRLC_PG_CNTL, data); +} + +static void cz_enable_gfx_pipeline_power_gating(struct amdgpu_device *adev, + bool enable) +{ + u32 data, orig; + + orig = data = RREG32(mmRLC_PG_CNTL); + + if (enable) + data |= RLC_PG_CNTL__GFX_PIPELINE_PG_ENABLE_MASK; + else + data &= ~RLC_PG_CNTL__GFX_PIPELINE_PG_ENABLE_MASK; + + if (orig != data) + WREG32(mmRLC_PG_CNTL, data); + + /* Read any GFX register to wake up GFX. */ + if (!enable) + data = RREG32(mmDB_RENDER_CONTROL); +} + +static void cz_update_gfx_cg_power_gating(struct amdgpu_device *adev, + bool enable) +{ + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) && enable) { + cz_enable_gfx_cg_power_gating(adev, true); + if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PIPELINE) + cz_enable_gfx_pipeline_power_gating(adev, true); + } else { + cz_enable_gfx_cg_power_gating(adev, false); + cz_enable_gfx_pipeline_power_gating(adev, false); } } @@ -5111,21 +5365,42 @@ static int gfx_v8_0_set_powergating_state(void *handle, enum amd_powergating_state state) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; + bool enable = (state == AMD_PG_STATE_GATE) ? true : false; if (!(adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)) return 0; switch (adev->asic_type) { + case CHIP_CARRIZO: + case CHIP_STONEY: + if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) + cz_update_gfx_cg_power_gating(adev, enable); + + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) && enable) + gfx_v8_0_enable_gfx_static_mg_power_gating(adev, true); + else + gfx_v8_0_enable_gfx_static_mg_power_gating(adev, false); + + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_DMG) && enable) + gfx_v8_0_enable_gfx_dynamic_mg_power_gating(adev, true); + else + gfx_v8_0_enable_gfx_dynamic_mg_power_gating(adev, false); + break; case CHIP_POLARIS11: - if (adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) - polaris11_enable_gfx_static_mg_power_gating(adev, - state == AMD_PG_STATE_GATE ? true : false); - else if (adev->pg_flags & AMD_PG_SUPPORT_GFX_DMG) - polaris11_enable_gfx_dynamic_mg_power_gating(adev, - state == AMD_PG_STATE_GATE ? true : false); + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) && enable) + gfx_v8_0_enable_gfx_static_mg_power_gating(adev, true); + else + gfx_v8_0_enable_gfx_static_mg_power_gating(adev, false); + + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_DMG) && enable) + gfx_v8_0_enable_gfx_dynamic_mg_power_gating(adev, true); + else + gfx_v8_0_enable_gfx_dynamic_mg_power_gating(adev, false); + + if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_QUICK_MG) && enable) + polaris11_enable_gfx_quick_mg_power_gating(adev, true); else - polaris11_enable_gfx_quick_mg_power_gating(adev, - state == AMD_PG_STATE_GATE ? true : false); + polaris11_enable_gfx_quick_mg_power_gating(adev, false); break; default: break; @@ -5139,7 +5414,7 @@ static void gfx_v8_0_send_serdes_cmd(struct amdgpu_device *adev, { uint32_t data; - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); WREG32(mmRLC_SERDES_WR_CU_MASTER_MASK, 0xffffffff); WREG32(mmRLC_SERDES_WR_NONCU_MASTER_MASK, 0xffffffff); @@ -5527,6 +5802,8 @@ static void gfx_v8_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev WREG32(mmRLC_CGCG_CGLS_CTRL, data); } + gfx_v8_0_wait_for_rlc_serdes(adev); + adev->gfx.rlc.funcs->exit_safe_mode(adev); } static int gfx_v8_0_update_gfx_clock_gating(struct amdgpu_device *adev, @@ -5652,17 +5929,6 @@ static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, unsigned vm_id, bool ctx_switch) { u32 header, control = 0; - u32 next_rptr = ring->wptr + 5; - - if (ctx_switch) - next_rptr += 2; - - next_rptr += 4; - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, WRITE_DATA_DST_SEL(5) | WR_CONFIRM); - amdgpu_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); - amdgpu_ring_write(ring, next_rptr); /* insert SWITCH_BUFFER packet before first IB in the ring frame */ if (ctx_switch) { @@ -5691,23 +5957,9 @@ static void gfx_v8_0_ring_emit_ib_compute(struct amdgpu_ring *ring, struct amdgpu_ib *ib, unsigned vm_id, bool ctx_switch) { - u32 header, control = 0; - u32 next_rptr = ring->wptr + 5; - - control |= INDIRECT_BUFFER_VALID; - - next_rptr += 4; - amdgpu_ring_write(ring, PACKET3(PACKET3_WRITE_DATA, 3)); - amdgpu_ring_write(ring, WRITE_DATA_DST_SEL(5) | WR_CONFIRM); - amdgpu_ring_write(ring, ring->next_rptr_gpu_addr & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr) & 0xffffffff); - amdgpu_ring_write(ring, next_rptr); - - header = PACKET3(PACKET3_INDIRECT_BUFFER, 2); + u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vm_id << 24); - control |= ib->length_dw | (vm_id << 24); - - amdgpu_ring_write(ring, header); + amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); amdgpu_ring_write(ring, #ifdef __BIG_ENDIAN (2 << 0) | @@ -6160,9 +6412,9 @@ static void gfx_v8_0_set_rlc_funcs(struct amdgpu_device *adev) { switch (adev->asic_type) { case CHIP_TOPAZ: - case CHIP_STONEY: adev->gfx.rlc.funcs = &iceland_rlc_funcs; break; + case CHIP_STONEY: case CHIP_CARRIZO: adev->gfx.rlc.funcs = &cz_rlc_funcs; break; @@ -6200,6 +6452,20 @@ static void gfx_v8_0_set_gds_init(struct amdgpu_device *adev) } } +static void gfx_v8_0_set_user_cu_inactive_bitmap(struct amdgpu_device *adev, + u32 bitmap) +{ + u32 data; + + if (!bitmap) + return; + + data = bitmap << GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT; + data &= GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK; + + WREG32(mmGC_USER_SHADER_ARRAY_CONFIG, data); +} + static u32 gfx_v8_0_get_cu_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; @@ -6220,16 +6486,22 @@ static void gfx_v8_0_get_cu_info(struct amdgpu_device *adev) int i, j, k, counter, active_cu_number = 0; u32 mask, bitmap, ao_bitmap, ao_cu_mask = 0; struct amdgpu_cu_info *cu_info = &adev->gfx.cu_info; + unsigned disable_masks[4 * 2]; memset(cu_info, 0, sizeof(*cu_info)); + amdgpu_gfx_parse_disable_cu(disable_masks, 4, 2); + mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { mask = 1; ao_bitmap = 0; counter = 0; - gfx_v8_0_select_se_sh(adev, i, j); + gfx_v8_0_select_se_sh(adev, i, j, 0xffffffff); + if (i < 4 && j < 2) + gfx_v8_0_set_user_cu_inactive_bitmap( + adev, disable_masks[i * 2 + j]); bitmap = gfx_v8_0_get_cu_active_bitmap(adev); cu_info->bitmap[i][j] = bitmap; @@ -6245,7 +6517,7 @@ static void gfx_v8_0_get_cu_info(struct amdgpu_device *adev) ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); } } - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); cu_info->number = active_cu_number; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.h b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.h index 16a49f53a2fa..bc82c794312c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.h +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.h @@ -26,7 +26,6 @@ extern const struct amd_ip_funcs gfx_v8_0_ip_funcs; -uint64_t gfx_v8_0_get_gpu_clock_counter(struct amdgpu_device *adev); void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, u32 se_num, u32 sh_num); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c index 1feb6439cb0b..d24a82bd0c7a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c @@ -39,6 +39,7 @@ static void gmc_v7_0_set_gart_funcs(struct amdgpu_device *adev); static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev); +static int gmc_v7_0_wait_for_idle(void *handle); MODULE_FIRMWARE("radeon/bonaire_mc.bin"); MODULE_FIRMWARE("radeon/hawaii_mc.bin"); @@ -73,39 +74,15 @@ static void gmc_v7_0_init_golden_registers(struct amdgpu_device *adev) } } -/** - * gmc7_mc_wait_for_idle - wait for MC idle callback. - * - * @adev: amdgpu_device pointer - * - * Wait for the MC (memory controller) to be idle. - * (evergreen+). - * Returns 0 if the MC is idle, -1 if not. - */ -int gmc_v7_0_mc_wait_for_idle(struct amdgpu_device *adev) -{ - unsigned i; - u32 tmp; - - for (i = 0; i < adev->usec_timeout; i++) { - /* read MC_STATUS */ - tmp = RREG32(mmSRBM_STATUS) & 0x1F00; - if (!tmp) - return 0; - udelay(1); - } - return -1; -} - -void gmc_v7_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v7_0_mc_stop(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) { u32 blackout; if (adev->mode_info.num_crtc) amdgpu_display_stop_mc_access(adev, save); - amdgpu_asic_wait_for_mc_idle(adev); + gmc_v7_0_wait_for_idle((void *)adev); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); if (REG_GET_FIELD(blackout, MC_SHARED_BLACKOUT_CNTL, BLACKOUT_MODE) != 1) { @@ -120,8 +97,8 @@ void gmc_v7_0_mc_stop(struct amdgpu_device *adev, udelay(100); } -void gmc_v7_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v7_0_mc_resume(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) { u32 tmp; @@ -311,7 +288,7 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) amdgpu_display_set_vga_render_state(adev, false); gmc_v7_0_mc_stop(adev, &save); - if (amdgpu_asic_wait_for_mc_idle(adev)) { + if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } /* Update configuration */ @@ -331,7 +308,7 @@ static void gmc_v7_0_mc_program(struct amdgpu_device *adev) WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, 0x0FFFFFFF); WREG32(mmMC_VM_AGP_BOT, 0x0FFFFFFF); - if (amdgpu_asic_wait_for_mc_idle(adev)) { + if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } gmc_v7_0_mc_resume(adev, &save); @@ -1137,7 +1114,7 @@ static int gmc_v7_0_soft_reset(void *handle) if (srbm_soft_reset) { gmc_v7_0_mc_stop(adev, &save); - if (gmc_v7_0_wait_for_idle(adev)) { + if (gmc_v7_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.h b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.h index 36fcbbc46ada..0b386b5d2f7a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.h +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.h @@ -26,11 +26,4 @@ extern const struct amd_ip_funcs gmc_v7_0_ip_funcs; -/* XXX these shouldn't be exported */ -void gmc_v7_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); -void gmc_v7_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); -int gmc_v7_0_mc_wait_for_idle(struct amdgpu_device *adev); - #endif diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c index 9945d5bbf1fe..717359d3ba8c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c @@ -41,6 +41,7 @@ static void gmc_v8_0_set_gart_funcs(struct amdgpu_device *adev); static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev); +static int gmc_v8_0_wait_for_idle(void *handle); MODULE_FIRMWARE("amdgpu/tonga_mc.bin"); MODULE_FIRMWARE("amdgpu/polaris11_mc.bin"); @@ -147,44 +148,15 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev) } } -/** - * gmc8_mc_wait_for_idle - wait for MC idle callback. - * - * @adev: amdgpu_device pointer - * - * Wait for the MC (memory controller) to be idle. - * (evergreen+). - * Returns 0 if the MC is idle, -1 if not. - */ -int gmc_v8_0_mc_wait_for_idle(struct amdgpu_device *adev) -{ - unsigned i; - u32 tmp; - - for (i = 0; i < adev->usec_timeout; i++) { - /* read MC_STATUS */ - tmp = RREG32(mmSRBM_STATUS) & (SRBM_STATUS__VMC_BUSY_MASK | - SRBM_STATUS__MCB_BUSY_MASK | - SRBM_STATUS__MCB_NON_DISPLAY_BUSY_MASK | - SRBM_STATUS__MCC_BUSY_MASK | - SRBM_STATUS__MCD_BUSY_MASK | - SRBM_STATUS__VMC1_BUSY_MASK); - if (!tmp) - return 0; - udelay(1); - } - return -1; -} - -void gmc_v8_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v8_0_mc_stop(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) { u32 blackout; if (adev->mode_info.num_crtc) amdgpu_display_stop_mc_access(adev, save); - amdgpu_asic_wait_for_mc_idle(adev); + gmc_v8_0_wait_for_idle(adev); blackout = RREG32(mmMC_SHARED_BLACKOUT_CNTL); if (REG_GET_FIELD(blackout, MC_SHARED_BLACKOUT_CNTL, BLACKOUT_MODE) != 1) { @@ -199,8 +171,8 @@ void gmc_v8_0_mc_stop(struct amdgpu_device *adev, udelay(100); } -void gmc_v8_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save) +static void gmc_v8_0_mc_resume(struct amdgpu_device *adev, + struct amdgpu_mode_mc_save *save) { u32 tmp; @@ -393,7 +365,7 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) amdgpu_display_set_vga_render_state(adev, false); gmc_v8_0_mc_stop(adev, &save); - if (amdgpu_asic_wait_for_mc_idle(adev)) { + if (gmc_v8_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } /* Update configuration */ @@ -413,7 +385,7 @@ static void gmc_v8_0_mc_program(struct amdgpu_device *adev) WREG32(mmMC_VM_AGP_BASE, 0); WREG32(mmMC_VM_AGP_TOP, 0x0FFFFFFF); WREG32(mmMC_VM_AGP_BOT, 0x0FFFFFFF); - if (amdgpu_asic_wait_for_mc_idle(adev)) { + if (gmc_v8_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for MC idle timedout !\n"); } gmc_v8_0_mc_resume(adev, &save); @@ -1140,7 +1112,7 @@ static int gmc_v8_0_soft_reset(void *handle) if (srbm_soft_reset) { gmc_v8_0_mc_stop(adev, &save); - if (gmc_v8_0_wait_for_idle(adev)) { + if (gmc_v8_0_wait_for_idle((void *)adev)) { dev_warn(adev->dev, "Wait for GMC idle timed out !\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.h b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.h index 973436086b38..fc5001a8119d 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.h +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.h @@ -26,11 +26,4 @@ extern const struct amd_ip_funcs gmc_v8_0_ip_funcs; -/* XXX these shouldn't be exported */ -void gmc_v8_0_mc_stop(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); -void gmc_v8_0_mc_resume(struct amdgpu_device *adev, - struct amdgpu_mode_mc_save *save); -int gmc_v8_0_mc_wait_for_idle(struct amdgpu_device *adev); - #endif diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c b/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c index 460bc8ad37e6..825ccd63f2dc 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_dpm.c @@ -72,6 +72,11 @@ static int iceland_dpm_sw_init(void *handle) static int iceland_dpm_sw_fini(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c index a789a863d677..5a0e245771ce 100644 --- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c @@ -507,19 +507,19 @@ static int kv_enable_didt(struct amdgpu_device *adev, bool enable) pi->caps_db_ramping || pi->caps_td_ramping || pi->caps_tcp_ramping) { - gfx_v7_0_enter_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->enter_safe_mode(adev); if (enable) { ret = kv_program_pt_config_registers(adev, didt_config_kv); if (ret) { - gfx_v7_0_exit_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->exit_safe_mode(adev); return ret; } } kv_do_enable_didt(adev, enable); - gfx_v7_0_exit_rlc_safe_mode(adev); + adev->gfx.rlc.funcs->exit_safe_mode(adev); } return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c index f4c3130d3fdb..ac3730a6e49f 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c @@ -105,6 +105,15 @@ static void sdma_v2_4_init_golden_registers(struct amdgpu_device *adev) } } +static void sdma_v2_4_free_microcode(struct amdgpu_device *adev) +{ + int i; + for (i = 0; i < adev->sdma.num_instances; i++) { + release_firmware(adev->sdma.instance[i].fw); + adev->sdma.instance[i].fw = NULL; + } +} + /** * sdma_v2_4_init_microcode - load ucode images from disk * @@ -246,19 +255,6 @@ static void sdma_v2_4_ring_emit_ib(struct amdgpu_ring *ring, unsigned vm_id, bool ctx_switch) { u32 vmid = vm_id & 0xf; - u32 next_rptr = ring->wptr + 5; - - while ((next_rptr & 7) != 2) - next_rptr++; - - next_rptr += 6; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) | - SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_WRITE_LINEAR)); - amdgpu_ring_write(ring, lower_32_bits(ring->next_rptr_gpu_addr) & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr)); - amdgpu_ring_write(ring, SDMA_PKT_WRITE_UNTILED_DW_3_COUNT(1)); - amdgpu_ring_write(ring, next_rptr); /* IB packet must end on a 8 DW boundary */ sdma_v2_4_ring_insert_nop(ring, (10 - (ring->wptr & 7)) % 8); @@ -461,6 +457,8 @@ static int sdma_v2_4_gfx_resume(struct amdgpu_device *adev) /* Initialize the ring buffer's read and write pointers */ WREG32(mmSDMA0_GFX_RB_RPTR + sdma_offsets[i], 0); WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_RPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_OFFSET + sdma_offsets[i], 0); /* set the wb address whether it's enabled or not */ WREG32(mmSDMA0_GFX_RB_RPTR_ADDR_HI + sdma_offsets[i], @@ -489,7 +487,11 @@ static int sdma_v2_4_gfx_resume(struct amdgpu_device *adev) WREG32(mmSDMA0_GFX_IB_CNTL + sdma_offsets[i], ib_cntl); ring->ready = true; + } + sdma_v2_4_enable(adev, true); + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; r = amdgpu_ring_test_ring(ring); if (r) { ring->ready = false; @@ -580,8 +582,8 @@ static int sdma_v2_4_start(struct amdgpu_device *adev) return -EINVAL; } - /* unhalt the MEs */ - sdma_v2_4_enable(adev, true); + /* halt the engine before programing */ + sdma_v2_4_enable(adev, false); /* start the gfx rings and rlc compute queues */ r = sdma_v2_4_gfx_resume(adev); @@ -1012,6 +1014,7 @@ static int sdma_v2_4_sw_fini(void *handle) for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + sdma_v2_4_free_microcode(adev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c index 31d99b0010f7..f00db6f4c04c 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c @@ -236,6 +236,15 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev) } } +static void sdma_v3_0_free_microcode(struct amdgpu_device *adev) +{ + int i; + for (i = 0; i < adev->sdma.num_instances; i++) { + release_firmware(adev->sdma.instance[i].fw); + adev->sdma.instance[i].fw = NULL; + } +} + /** * sdma_v3_0_init_microcode - load ucode images from disk * @@ -406,18 +415,6 @@ static void sdma_v3_0_ring_emit_ib(struct amdgpu_ring *ring, unsigned vm_id, bool ctx_switch) { u32 vmid = vm_id & 0xf; - u32 next_rptr = ring->wptr + 5; - - while ((next_rptr & 7) != 2) - next_rptr++; - next_rptr += 6; - - amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_WRITE) | - SDMA_PKT_HEADER_SUB_OP(SDMA_SUBOP_WRITE_LINEAR)); - amdgpu_ring_write(ring, lower_32_bits(ring->next_rptr_gpu_addr) & 0xfffffffc); - amdgpu_ring_write(ring, upper_32_bits(ring->next_rptr_gpu_addr)); - amdgpu_ring_write(ring, SDMA_PKT_WRITE_UNTILED_DW_3_COUNT(1)); - amdgpu_ring_write(ring, next_rptr); /* IB packet must end on a 8 DW boundary */ sdma_v3_0_ring_insert_nop(ring, (10 - (ring->wptr & 7)) % 8); @@ -672,6 +669,8 @@ static int sdma_v3_0_gfx_resume(struct amdgpu_device *adev) /* Initialize the ring buffer's read and write pointers */ WREG32(mmSDMA0_GFX_RB_RPTR + sdma_offsets[i], 0); WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_RPTR + sdma_offsets[i], 0); + WREG32(mmSDMA0_GFX_IB_OFFSET + sdma_offsets[i], 0); /* set the wb address whether it's enabled or not */ WREG32(mmSDMA0_GFX_RB_RPTR_ADDR_HI + sdma_offsets[i], @@ -711,7 +710,15 @@ static int sdma_v3_0_gfx_resume(struct amdgpu_device *adev) WREG32(mmSDMA0_GFX_IB_CNTL + sdma_offsets[i], ib_cntl); ring->ready = true; + } + /* unhalt the MEs */ + sdma_v3_0_enable(adev, true); + /* enable sdma ring preemption */ + sdma_v3_0_ctx_switch_enable(adev, true); + + for (i = 0; i < adev->sdma.num_instances; i++) { + ring = &adev->sdma.instance[i].ring; r = amdgpu_ring_test_ring(ring); if (r) { ring->ready = false; @@ -804,10 +811,9 @@ static int sdma_v3_0_start(struct amdgpu_device *adev) } } - /* unhalt the MEs */ - sdma_v3_0_enable(adev, true); - /* enable sdma ring preemption */ - sdma_v3_0_ctx_switch_enable(adev, true); + /* disble sdma engine before programing it */ + sdma_v3_0_ctx_switch_enable(adev, false); + sdma_v3_0_enable(adev, false); /* start the gfx rings and rlc compute queues */ r = sdma_v3_0_gfx_resume(adev); @@ -1247,6 +1253,7 @@ static int sdma_v3_0_sw_fini(void *handle) for (i = 0; i < adev->sdma.num_instances; i++) amdgpu_ring_fini(&adev->sdma.instance[i].ring); + sdma_v3_0_free_microcode(adev); return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/tonga_dpm.c b/drivers/gpu/drm/amd/amdgpu/tonga_dpm.c index b7615cefcac4..f06f6f4dc3a8 100644 --- a/drivers/gpu/drm/amd/amdgpu/tonga_dpm.c +++ b/drivers/gpu/drm/amd/amdgpu/tonga_dpm.c @@ -71,6 +71,11 @@ static int tonga_dpm_sw_init(void *handle) static int tonga_dpm_sw_fini(void *handle) { + struct amdgpu_device *adev = (struct amdgpu_device *)handle; + + release_firmware(adev->pm.fw); + adev->pm.fw = NULL; + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c index f07551476a70..416c8567d3ed 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c @@ -34,6 +34,8 @@ #include "oss/oss_2_0_d.h" #include "oss/oss_2_0_sh_mask.h" +#include "bif/bif_4_1_d.h" + static void uvd_v4_2_mc_resume(struct amdgpu_device *adev); static void uvd_v4_2_init_cg(struct amdgpu_device *adev); static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev); @@ -439,6 +441,32 @@ static void uvd_v4_2_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq } /** + * uvd_v4_2_ring_emit_hdp_flush - emit an hdp flush + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp flush. + */ +static void uvd_v4_2_ring_emit_hdp_flush(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); + amdgpu_ring_write(ring, 0); +} + +/** + * uvd_v4_2_ring_hdp_invalidate - emit an hdp invalidate + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp invalidate. + */ +static void uvd_v4_2_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); + amdgpu_ring_write(ring, 1); +} + +/** * uvd_v4_2_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -763,6 +791,8 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = { .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_ib = uvd_v4_2_ring_emit_ib, .emit_fence = uvd_v4_2_ring_emit_fence, + .emit_hdp_flush = uvd_v4_2_ring_emit_hdp_flush, + .emit_hdp_invalidate = uvd_v4_2_ring_emit_hdp_invalidate, .test_ring = uvd_v4_2_ring_test_ring, .test_ib = uvd_v4_2_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c index e0a76a883d46..dd636c4c4b08 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c @@ -31,6 +31,7 @@ #include "uvd/uvd_5_0_sh_mask.h" #include "oss/oss_2_0_d.h" #include "oss/oss_2_0_sh_mask.h" +#include "bif/bif_5_0_d.h" #include "vi.h" static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev); @@ -489,6 +490,32 @@ static void uvd_v5_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq } /** + * uvd_v5_0_ring_emit_hdp_flush - emit an hdp flush + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp flush. + */ +static void uvd_v5_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); + amdgpu_ring_write(ring, 0); +} + +/** + * uvd_v5_0_ring_hdp_invalidate - emit an hdp invalidate + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp invalidate. + */ +static void uvd_v5_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); + amdgpu_ring_write(ring, 1); +} + +/** * uvd_v5_0_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -815,6 +842,8 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = { .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_ib = uvd_v5_0_ring_emit_ib, .emit_fence = uvd_v5_0_ring_emit_fence, + .emit_hdp_flush = uvd_v5_0_ring_emit_hdp_flush, + .emit_hdp_invalidate = uvd_v5_0_ring_emit_hdp_invalidate, .test_ring = uvd_v5_0_ring_test_ring, .test_ib = uvd_v5_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c index c9929d665c01..07e9a987fbee 100644 --- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c @@ -33,6 +33,7 @@ #include "oss/oss_2_0_sh_mask.h" #include "smu/smu_7_1_3_d.h" #include "smu/smu_7_1_3_sh_mask.h" +#include "bif/bif_5_1_d.h" #include "vi.h" static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev); @@ -385,8 +386,8 @@ static int uvd_v6_0_start(struct amdgpu_device *adev) uint32_t mp_swap_cntl; int i, j, r; - /*disable DPG */ - WREG32_P(mmUVD_POWER_STATUS, 0, ~(1 << 2)); + /* disable DPG */ + WREG32_P(mmUVD_POWER_STATUS, 0, ~UVD_POWER_STATUS__UVD_PG_MODE_MASK); /* disable byte swapping */ lmi_swap_cntl = 0; @@ -405,17 +406,21 @@ static int uvd_v6_0_start(struct amdgpu_device *adev) } /* disable interupt */ - WREG32_P(mmUVD_MASTINT_EN, 0, ~(1 << 1)); + WREG32_P(mmUVD_MASTINT_EN, 0, ~UVD_MASTINT_EN__VCPU_EN_MASK); /* stall UMC and register bus before resetting VCPU */ - WREG32_P(mmUVD_LMI_CTRL2, 1 << 8, ~(1 << 8)); + WREG32_P(mmUVD_LMI_CTRL2, UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); mdelay(1); /* put LMI, VCPU, RBC etc... into reset */ - WREG32(mmUVD_SOFT_RESET, UVD_SOFT_RESET__LMI_SOFT_RESET_MASK | - UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK | UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK | - UVD_SOFT_RESET__RBC_SOFT_RESET_MASK | UVD_SOFT_RESET__CSM_SOFT_RESET_MASK | - UVD_SOFT_RESET__CXW_SOFT_RESET_MASK | UVD_SOFT_RESET__TAP_SOFT_RESET_MASK | + WREG32(mmUVD_SOFT_RESET, + UVD_SOFT_RESET__LMI_SOFT_RESET_MASK | + UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK | + UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK | + UVD_SOFT_RESET__RBC_SOFT_RESET_MASK | + UVD_SOFT_RESET__CSM_SOFT_RESET_MASK | + UVD_SOFT_RESET__CXW_SOFT_RESET_MASK | + UVD_SOFT_RESET__TAP_SOFT_RESET_MASK | UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK); mdelay(5); @@ -424,8 +429,13 @@ static int uvd_v6_0_start(struct amdgpu_device *adev) mdelay(5); /* initialize UVD memory controller */ - WREG32(mmUVD_LMI_CTRL, 0x40 | (1 << 8) | (1 << 13) | - (1 << 21) | (1 << 9) | (1 << 20)); + WREG32(mmUVD_LMI_CTRL, + (0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) | + UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK | + UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK | + UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK | + UVD_LMI_CTRL__REQ_MODE_MASK | + UVD_LMI_CTRL__DISABLE_ON_FWV_FAIL_MASK); #ifdef __BIG_ENDIAN /* swap (8 in 32) RB and IB */ @@ -447,10 +457,10 @@ static int uvd_v6_0_start(struct amdgpu_device *adev) mdelay(5); /* enable VCPU clock */ - WREG32(mmUVD_VCPU_CNTL, 1 << 9); + WREG32(mmUVD_VCPU_CNTL, UVD_VCPU_CNTL__CLK_EN_MASK); /* enable UMC */ - WREG32_P(mmUVD_LMI_CTRL2, 0, ~(1 << 8)); + WREG32_P(mmUVD_LMI_CTRL2, 0, ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK); /* boot up the VCPU */ WREG32(mmUVD_SOFT_RESET, 0); @@ -484,10 +494,12 @@ static int uvd_v6_0_start(struct amdgpu_device *adev) return r; } /* enable master interrupt */ - WREG32_P(mmUVD_MASTINT_EN, 3 << 1, ~(3 << 1)); + WREG32_P(mmUVD_MASTINT_EN, + (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK), + ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK)); /* clear the bit 4 of UVD_STATUS */ - WREG32_P(mmUVD_STATUS, 0, ~(2 << 1)); + WREG32_P(mmUVD_STATUS, 0, ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT)); rb_bufsz = order_base_2(ring->ring_size); tmp = 0; @@ -581,6 +593,32 @@ static void uvd_v6_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq } /** + * uvd_v6_0_ring_emit_hdp_flush - emit an hdp flush + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp flush. + */ +static void uvd_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_MEM_COHERENCY_FLUSH_CNTL, 0)); + amdgpu_ring_write(ring, 0); +} + +/** + * uvd_v6_0_ring_hdp_invalidate - emit an hdp invalidate + * + * @ring: amdgpu_ring pointer + * + * Emits an hdp invalidate. + */ +static void uvd_v6_0_ring_emit_hdp_invalidate(struct amdgpu_ring *ring) +{ + amdgpu_ring_write(ring, PACKET0(mmHDP_DEBUG0, 0)); + amdgpu_ring_write(ring, 1); +} + +/** * uvd_v6_0_ring_test_ring - register write test * * @ring: amdgpu_ring pointer @@ -847,7 +885,8 @@ static int uvd_v6_0_set_clockgating_state(void *handle, bool enable = (state == AMD_CG_STATE_GATE) ? true : false; static int curstate = -1; - if (adev->asic_type == CHIP_FIJI) + if (adev->asic_type == CHIP_FIJI || + adev->asic_type == CHIP_POLARIS10) uvd_v6_set_bypass_mode(adev, enable); if (!(adev->cg_flags & AMD_CG_SUPPORT_UVD_MGCG)) @@ -919,6 +958,8 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_funcs = { .parse_cs = amdgpu_uvd_ring_parse_cs, .emit_ib = uvd_v6_0_ring_emit_ib, .emit_fence = uvd_v6_0_ring_emit_fence, + .emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush, + .emit_hdp_invalidate = uvd_v6_0_ring_emit_hdp_invalidate, .test_ring = uvd_v6_0_ring_test_ring, .test_ib = uvd_v6_0_ring_test_ib, .insert_nop = amdgpu_ring_insert_nop, diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index 2c88d0b66cf3..cda7def9dc2c 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -203,6 +203,29 @@ static void vi_didt_wreg(struct amdgpu_device *adev, u32 reg, u32 v) spin_unlock_irqrestore(&adev->didt_idx_lock, flags); } +static u32 vi_gc_cac_rreg(struct amdgpu_device *adev, u32 reg) +{ + unsigned long flags; + u32 r; + + spin_lock_irqsave(&adev->gc_cac_idx_lock, flags); + WREG32(mmGC_CAC_IND_INDEX, (reg)); + r = RREG32(mmGC_CAC_IND_DATA); + spin_unlock_irqrestore(&adev->gc_cac_idx_lock, flags); + return r; +} + +static void vi_gc_cac_wreg(struct amdgpu_device *adev, u32 reg, u32 v) +{ + unsigned long flags; + + spin_lock_irqsave(&adev->gc_cac_idx_lock, flags); + WREG32(mmGC_CAC_IND_INDEX, (reg)); + WREG32(mmGC_CAC_IND_DATA, (v)); + spin_unlock_irqrestore(&adev->gc_cac_idx_lock, flags); +} + + static const u32 tonga_mgcg_cgcg_init[] = { mmCGTT_DRM_CLK_CTRL0, 0xffffffff, 0x00600100, @@ -421,6 +444,20 @@ static bool vi_read_bios_from_rom(struct amdgpu_device *adev, return true; } +static u32 vi_get_virtual_caps(struct amdgpu_device *adev) +{ + u32 caps = 0; + u32 reg = RREG32(mmBIF_IOV_FUNC_IDENTIFIER); + + if (REG_GET_FIELD(reg, BIF_IOV_FUNC_IDENTIFIER, IOV_ENABLE)) + caps |= AMDGPU_VIRT_CAPS_SRIOV_EN; + + if (REG_GET_FIELD(reg, BIF_IOV_FUNC_IDENTIFIER, FUNC_IDENTIFIER)) + caps |= AMDGPU_VIRT_CAPS_IS_VF; + + return caps; +} + static const struct amdgpu_allowed_register_entry tonga_allowed_read_registers[] = { {mmGB_MACROTILE_MODE7, true}, }; @@ -519,12 +556,12 @@ static uint32_t vi_read_indexed_register(struct amdgpu_device *adev, u32 se_num, mutex_lock(&adev->grbm_idx_mutex); if (se_num != 0xffffffff || sh_num != 0xffffffff) - gfx_v8_0_select_se_sh(adev, se_num, sh_num); + amdgpu_gfx_select_se_sh(adev, se_num, sh_num, 0xffffffff); val = RREG32(reg_offset); if (se_num != 0xffffffff || sh_num != 0xffffffff) - gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff); + amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); return val; } @@ -583,7 +620,7 @@ static int vi_read_register(struct amdgpu_device *adev, u32 se_num, return -EINVAL; } -static void vi_gpu_pci_config_reset(struct amdgpu_device *adev) +static int vi_gpu_pci_config_reset(struct amdgpu_device *adev) { u32 i; @@ -598,11 +635,14 @@ static void vi_gpu_pci_config_reset(struct amdgpu_device *adev) /* wait for asic to come out of reset */ for (i = 0; i < adev->usec_timeout; i++) { - if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) - break; + if (RREG32(mmCONFIG_MEMSIZE) != 0xffffffff) { + /* enable BM */ + pci_set_master(adev->pdev); + return 0; + } udelay(1); } - + return -EINVAL; } static void vi_set_bios_scratch_engine_hung(struct amdgpu_device *adev, bool hung) @@ -628,13 +668,15 @@ static void vi_set_bios_scratch_engine_hung(struct amdgpu_device *adev, bool hun */ static int vi_asic_reset(struct amdgpu_device *adev) { + int r; + vi_set_bios_scratch_engine_hung(adev, true); - vi_gpu_pci_config_reset(adev); + r = vi_gpu_pci_config_reset(adev); vi_set_bios_scratch_engine_hung(adev, false); - return 0; + return r; } static int vi_set_uvd_clock(struct amdgpu_device *adev, u32 clock, @@ -1118,9 +1160,7 @@ static const struct amdgpu_asic_funcs vi_asic_funcs = .get_xclk = &vi_get_xclk, .set_uvd_clocks = &vi_set_uvd_clocks, .set_vce_clocks = &vi_set_vce_clocks, - /* these should be moved to their own ip modules */ - .get_gpu_clock_counter = &gfx_v8_0_get_gpu_clock_counter, - .wait_for_mc_idle = &gmc_v8_0_mc_wait_for_idle, + .get_virtual_caps = &vi_get_virtual_caps, }; static int vi_common_early_init(void *handle) @@ -1141,6 +1181,8 @@ static int vi_common_early_init(void *handle) adev->uvd_ctx_wreg = &vi_uvd_ctx_wreg; adev->didt_rreg = &vi_didt_rreg; adev->didt_wreg = &vi_didt_wreg; + adev->gc_cac_rreg = &vi_gc_cac_rreg; + adev->gc_cac_wreg = &vi_gc_cac_wreg; adev->asic_funcs = &vi_asic_funcs; @@ -1207,19 +1249,39 @@ static int vi_common_early_init(void *handle) AMD_CG_SUPPORT_HDP_LS | AMD_CG_SUPPORT_SDMA_MGCG | AMD_CG_SUPPORT_SDMA_LS; + /* rev0 hardware doesn't support PG */ adev->pg_flags = 0; + if (adev->rev_id != 0x00) + adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_RLC_SMU_HS | + AMD_PG_SUPPORT_GFX_PIPELINE; adev->external_rev_id = adev->rev_id + 0x1; break; case CHIP_STONEY: adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG | AMD_CG_SUPPORT_GFX_MGCG | AMD_CG_SUPPORT_GFX_MGLS | + AMD_CG_SUPPORT_GFX_RLC_LS | + AMD_CG_SUPPORT_GFX_CP_LS | + AMD_CG_SUPPORT_GFX_CGTS | + AMD_CG_SUPPORT_GFX_MGLS | + AMD_CG_SUPPORT_GFX_CGTS_LS | + AMD_CG_SUPPORT_GFX_CGCG | + AMD_CG_SUPPORT_GFX_CGLS | AMD_CG_SUPPORT_BIF_LS | AMD_CG_SUPPORT_HDP_MGCG | AMD_CG_SUPPORT_HDP_LS | AMD_CG_SUPPORT_SDMA_MGCG | AMD_CG_SUPPORT_SDMA_LS; - adev->pg_flags = 0; + adev->pg_flags = AMD_PG_SUPPORT_GFX_PG | + AMD_PG_SUPPORT_GFX_SMG | + AMD_PG_SUPPORT_GFX_DMG | + AMD_PG_SUPPORT_GFX_PIPELINE | + AMD_PG_SUPPORT_CP | + AMD_PG_SUPPORT_RLC_SMU_HS; adev->external_rev_id = adev->rev_id + 0x1; break; default: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index ac005796b71c..4f3849ac8c07 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -63,13 +63,12 @@ static struct kfd_process *create_process(const struct task_struct *thread); void kfd_process_create_wq(void) { if (!kfd_process_wq) - kfd_process_wq = create_workqueue("kfd_process_wq"); + kfd_process_wq = alloc_workqueue("kfd_process_wq", 0, 0); } void kfd_process_destroy_wq(void) { if (kfd_process_wq) { - flush_workqueue(kfd_process_wq); destroy_workqueue(kfd_process_wq); kfd_process_wq = NULL; } @@ -242,13 +241,19 @@ static void kfd_process_notifier_release(struct mmu_notifier *mn, pqm_uninit(&p->pqm); /* Iterate over all process device data structure and check - * if we should reset all wavefronts */ - list_for_each_entry(pdd, &p->per_device_data, per_device_list) + * if we should delete debug managers and reset all wavefronts + */ + list_for_each_entry(pdd, &p->per_device_data, per_device_list) { + if ((pdd->dev->dbgmgr) && + (pdd->dev->dbgmgr->pasid == p->pasid)) + kfd_dbgmgr_destroy(pdd->dev->dbgmgr); + if (pdd->reset_wavefronts) { pr_warn("amdkfd: Resetting all wave fronts\n"); dbgdev_wave_reset_wavefronts(pdd->dev, p); pdd->reset_wavefronts = false; } + } mutex_unlock(&p->mutex); @@ -324,6 +329,7 @@ err_process_pqm_init: synchronize_rcu(); mmu_notifier_unregister_no_release(&process->mmu_notifier, process->mm); err_mmu_notifier: + mutex_destroy(&process->mutex); kfd_pasid_free(process->pasid); err_alloc_pasid: kfree(process->queues); @@ -404,42 +410,52 @@ void kfd_unbind_process_from_device(struct kfd_dev *dev, unsigned int pasid) idx = srcu_read_lock(&kfd_processes_srcu); + /* + * Look for the process that matches the pasid. If there is no such + * process, we either released it in amdkfd's own notifier, or there + * is a bug. Unfortunately, there is no way to tell... + */ hash_for_each_rcu(kfd_processes_table, i, p, kfd_processes) - if (p->pasid == pasid) - break; + if (p->pasid == pasid) { - srcu_read_unlock(&kfd_processes_srcu, idx); + srcu_read_unlock(&kfd_processes_srcu, idx); - BUG_ON(p->pasid != pasid); + pr_debug("Unbinding process %d from IOMMU\n", pasid); - mutex_lock(&p->mutex); + mutex_lock(&p->mutex); - if ((dev->dbgmgr) && (dev->dbgmgr->pasid == p->pasid)) - kfd_dbgmgr_destroy(dev->dbgmgr); + if ((dev->dbgmgr) && (dev->dbgmgr->pasid == p->pasid)) + kfd_dbgmgr_destroy(dev->dbgmgr); - pqm_uninit(&p->pqm); + pqm_uninit(&p->pqm); - pdd = kfd_get_process_device_data(dev, p); + pdd = kfd_get_process_device_data(dev, p); - if (!pdd) { - mutex_unlock(&p->mutex); - return; - } + if (!pdd) { + mutex_unlock(&p->mutex); + return; + } - if (pdd->reset_wavefronts) { - dbgdev_wave_reset_wavefronts(pdd->dev, p); - pdd->reset_wavefronts = false; - } + if (pdd->reset_wavefronts) { + dbgdev_wave_reset_wavefronts(pdd->dev, p); + pdd->reset_wavefronts = false; + } - /* - * Just mark pdd as unbound, because we still need it to call - * amd_iommu_unbind_pasid() in when the process exits. - * We don't call amd_iommu_unbind_pasid() here - * because the IOMMU called us. - */ - pdd->bound = false; + /* + * Just mark pdd as unbound, because we still need it + * to call amd_iommu_unbind_pasid() in when the + * process exits. + * We don't call amd_iommu_unbind_pasid() here + * because the IOMMU called us. + */ + pdd->bound = false; - mutex_unlock(&p->mutex); + mutex_unlock(&p->mutex); + + return; + } + + srcu_read_unlock(&kfd_processes_srcu, idx); } struct kfd_process_device *kfd_get_first_process_device_data(struct kfd_process *p) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index 74909e72a009..884c96f50c3d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -666,7 +666,7 @@ static ssize_t node_show(struct kobject *kobj, struct attribute *attr, dev->node_props.simd_count); if (dev->mem_bank_count < dev->node_props.mem_banks_count) { - pr_warn("kfd: mem_banks_count truncated from %d to %d\n", + pr_info_once("kfd: mem_banks_count truncated from %d to %d\n", dev->node_props.mem_banks_count, dev->mem_bank_count); sysfs_show_32bit_prop(buffer, "mem_banks_count", diff --git a/drivers/gpu/drm/amd/include/amd_pcie.h b/drivers/gpu/drm/amd/include/amd_pcie.h index 7c2a916c1e63..5eb895fd98bf 100644 --- a/drivers/gpu/drm/amd/include/amd_pcie.h +++ b/drivers/gpu/drm/amd/include/amd_pcie.h @@ -37,6 +37,13 @@ #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_MASK 0x0000FFFF #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_SHIFT 0 +/* gen: chipset 1/2, asic 1/2/3 */ +#define AMDGPU_DEFAULT_PCIE_GEN_MASK (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 \ + | CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 \ + | CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 \ + | CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 \ + | CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3) + /* Following flags shows PCIe lane width switch supported in driver which are decided by chipset and ASIC */ #define CAIL_PCIE_LINK_WIDTH_SUPPORT_X1 0x00010000 #define CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 0x00020000 @@ -47,4 +54,11 @@ #define CAIL_PCIE_LINK_WIDTH_SUPPORT_X32 0x00400000 #define CAIL_PCIE_LINK_WIDTH_SUPPORT_SHIFT 16 +/* 1/2/4/8/16 lanes */ +#define AMDGPU_DEFAULT_PCIE_MLW_MASK (CAIL_PCIE_LINK_WIDTH_SUPPORT_X1 \ + | CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 \ + | CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 \ + | CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 \ + | CAIL_PCIE_LINK_WIDTH_SUPPORT_X16) + #endif diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index 6080951d539d..a74a0d2ff1ca 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -26,15 +26,6 @@ #define AMD_MAX_USEC_TIMEOUT 100000 /* 100 ms */ /* -* Supported GPU families (aligned with amdgpu_drm.h) -*/ -#define AMD_FAMILY_UNKNOWN 0 -#define AMD_FAMILY_CI 120 /* Bonaire, Hawaii */ -#define AMD_FAMILY_KV 125 /* Kaveri, Kabini, Mullins */ -#define AMD_FAMILY_VI 130 /* Iceland, Tonga */ -#define AMD_FAMILY_CZ 135 /* Carrizo */ - -/* * Supported ASIC types */ enum amd_asic_type { @@ -120,6 +111,8 @@ enum amd_powergating_state { #define AMD_PG_SUPPORT_SDMA (1 << 8) #define AMD_PG_SUPPORT_ACP (1 << 9) #define AMD_PG_SUPPORT_SAMU (1 << 10) +#define AMD_PG_SUPPORT_GFX_QUICK_MG (1 << 11) +#define AMD_PG_SUPPORT_GFX_PIPELINE (1 << 12) enum amd_pm_state_type { /* not used for dpm */ @@ -157,6 +150,7 @@ struct amd_ip_funcs { int (*hw_init)(void *handle); /* tears down the hw state */ int (*hw_fini)(void *handle); + void (*late_fini)(void *handle); /* handles IP specific hw/sw changes for suspend */ int (*suspend)(void *handle); /* handles IP specific hw/sw changes for resume */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_d.h b/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_d.h index ebaf67bb1589..90ff7c8a6011 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_d.h +++ b/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_d.h @@ -2823,4 +2823,7 @@ #define mmDC_EDC_CSINVOC_CNT 0x3192 #define mmDC_EDC_RESTORE_CNT 0x3193 +#define mmGC_CAC_IND_INDEX 0x129a +#define mmGC_CAC_IND_DATA 0x129b + #endif /* GFX_8_0_D_H */ diff --git a/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_sh_mask.h index 7d722458d9f5..4070ca3a68eb 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/gca/gfx_8_0_sh_mask.h @@ -8730,8 +8730,6 @@ #define RLC_GPM_STAT__DYN_CU_POWERING_DOWN__SHIFT 0x10 #define RLC_GPM_STAT__ABORTED_PD_SEQUENCE_MASK 0x20000 #define RLC_GPM_STAT__ABORTED_PD_SEQUENCE__SHIFT 0x11 -#define RLC_GPM_STAT__RESERVED_MASK 0xfc0000 -#define RLC_GPM_STAT__RESERVED__SHIFT 0x12 #define RLC_GPM_STAT__PG_ERROR_STATUS_MASK 0xff000000 #define RLC_GPM_STAT__PG_ERROR_STATUS__SHIFT 0x18 #define RLC_GPU_CLOCK_32_RES_SEL__RES_SEL_MASK 0x3f @@ -8764,8 +8762,10 @@ #define RLC_PG_CNTL__SMU_CLK_SLOWDOWN_ON_PD_ENABLE__SHIFT 0x12 #define RLC_PG_CNTL__SMU_HANDSHAKE_ENABLE_MASK 0x80000 #define RLC_PG_CNTL__SMU_HANDSHAKE_ENABLE__SHIFT 0x13 -#define RLC_PG_CNTL__RESERVED1_MASK 0xf00000 -#define RLC_PG_CNTL__RESERVED1__SHIFT 0x14 +#define RLC_PG_CNTL__QUICK_PG_ENABLE_MASK 0x100000 +#define RLC_PG_CNTL__QUICK_PG_ENABLE__SHIFT 0x14 +#define RLC_PG_CNTL__RESERVED1_MASK 0xe00000 +#define RLC_PG_CNTL__RESERVED1__SHIFT 0x15 #define RLC_GPM_THREAD_PRIORITY__THREAD0_PRIORITY_MASK 0xff #define RLC_GPM_THREAD_PRIORITY__THREAD0_PRIORITY__SHIFT 0x0 #define RLC_GPM_THREAD_PRIORITY__THREAD1_PRIORITY_MASK 0xff00 @@ -9102,8 +9102,6 @@ #define RLC_GPM_LOG_CONT__CONT__SHIFT 0x0 #define RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG_MASK 0xff #define RLC_PG_DELAY_3__CGCG_ACTIVE_BEFORE_CGPG__SHIFT 0x0 -#define RLC_PG_DELAY_3__RESERVED_MASK 0xffffff00 -#define RLC_PG_DELAY_3__RESERVED__SHIFT 0x8 #define RLC_GPM_INT_DISABLE_TH0__DISABLE_MASK 0xffffffff #define RLC_GPM_INT_DISABLE_TH0__DISABLE__SHIFT 0x0 #define RLC_GPM_INT_DISABLE_TH1__DISABLE_MASK 0xffffffff @@ -9124,14 +9122,8 @@ #define RLC_SRM_DEBUG_SELECT__RESERVED__SHIFT 0x8 #define RLC_SRM_DEBUG__DATA_MASK 0xffffffff #define RLC_SRM_DEBUG__DATA__SHIFT 0x0 -#define RLC_SRM_ARAM_ADDR__ADDR_MASK 0x3ff -#define RLC_SRM_ARAM_ADDR__ADDR__SHIFT 0x0 -#define RLC_SRM_ARAM_ADDR__RESERVED_MASK 0xfffffc00 -#define RLC_SRM_ARAM_ADDR__RESERVED__SHIFT 0xa #define RLC_SRM_ARAM_DATA__DATA_MASK 0xffffffff #define RLC_SRM_ARAM_DATA__DATA__SHIFT 0x0 -#define RLC_SRM_DRAM_ADDR__ADDR_MASK 0x3ff -#define RLC_SRM_DRAM_ADDR__ADDR__SHIFT 0x0 #define RLC_SRM_DRAM_ADDR__RESERVED_MASK 0xfffffc00 #define RLC_SRM_DRAM_ADDR__RESERVED__SHIFT 0xa #define RLC_SRM_DRAM_DATA__DATA_MASK 0xffffffff @@ -17946,8 +17938,6 @@ #define VGT_TESS_DISTRIBUTION__ACCUM_TRI__SHIFT 0x8 #define VGT_TESS_DISTRIBUTION__ACCUM_QUAD_MASK 0xff0000 #define VGT_TESS_DISTRIBUTION__ACCUM_QUAD__SHIFT 0x10 -#define VGT_TESS_DISTRIBUTION__DONUT_SPLIT_MASK 0xff000000 -#define VGT_TESS_DISTRIBUTION__DONUT_SPLIT__SHIFT 0x18 #define VGT_TF_RING_SIZE__SIZE_MASK 0xffff #define VGT_TF_RING_SIZE__SIZE__SHIFT 0x0 #define VGT_SYS_CONFIG__DUAL_CORE_EN_MASK 0x1 @@ -20502,8 +20492,6 @@ #define DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT 0x4 #define DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK 0x20 #define DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT 0x5 -#define DIDT_SQ_CTRL0__UNUSED_0_MASK 0xffffffc0 -#define DIDT_SQ_CTRL0__UNUSED_0__SHIFT 0x6 #define DIDT_SQ_CTRL1__MIN_POWER_MASK 0xffff #define DIDT_SQ_CTRL1__MIN_POWER__SHIFT 0x0 #define DIDT_SQ_CTRL1__MAX_POWER_MASK 0xffff0000 @@ -20558,8 +20546,6 @@ #define DIDT_DB_CTRL0__DIDT_CTRL_RST__SHIFT 0x4 #define DIDT_DB_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK 0x20 #define DIDT_DB_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT 0x5 -#define DIDT_DB_CTRL0__UNUSED_0_MASK 0xffffffc0 -#define DIDT_DB_CTRL0__UNUSED_0__SHIFT 0x6 #define DIDT_DB_CTRL1__MIN_POWER_MASK 0xffff #define DIDT_DB_CTRL1__MIN_POWER__SHIFT 0x0 #define DIDT_DB_CTRL1__MAX_POWER_MASK 0xffff0000 @@ -20614,8 +20600,6 @@ #define DIDT_TD_CTRL0__DIDT_CTRL_RST__SHIFT 0x4 #define DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK 0x20 #define DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT 0x5 -#define DIDT_TD_CTRL0__UNUSED_0_MASK 0xffffffc0 -#define DIDT_TD_CTRL0__UNUSED_0__SHIFT 0x6 #define DIDT_TD_CTRL1__MIN_POWER_MASK 0xffff #define DIDT_TD_CTRL1__MIN_POWER__SHIFT 0x0 #define DIDT_TD_CTRL1__MAX_POWER_MASK 0xffff0000 @@ -20670,8 +20654,6 @@ #define DIDT_TCP_CTRL0__DIDT_CTRL_RST__SHIFT 0x4 #define DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK 0x20 #define DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT 0x5 -#define DIDT_TCP_CTRL0__UNUSED_0_MASK 0xffffffc0 -#define DIDT_TCP_CTRL0__UNUSED_0__SHIFT 0x6 #define DIDT_TCP_CTRL1__MIN_POWER_MASK 0xffff #define DIDT_TCP_CTRL1__MIN_POWER__SHIFT 0x0 #define DIDT_TCP_CTRL1__MAX_POWER_MASK 0xffff0000 @@ -20726,8 +20708,6 @@ #define DIDT_DBR_CTRL0__DIDT_CTRL_RST__SHIFT 0x4 #define DIDT_DBR_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK 0x20 #define DIDT_DBR_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT 0x5 -#define DIDT_DBR_CTRL0__UNUSED_0_MASK 0xffffffc0 -#define DIDT_DBR_CTRL0__UNUSED_0__SHIFT 0x6 #define DIDT_DBR_CTRL1__MIN_POWER_MASK 0xffff #define DIDT_DBR_CTRL1__MIN_POWER__SHIFT 0x0 #define DIDT_DBR_CTRL1__MAX_POWER_MASK 0xffff0000 @@ -20773,4 +20753,84 @@ #define DIDT_DBR_WEIGHT8_11__WEIGHT11_MASK 0xff000000 #define DIDT_DBR_WEIGHT8_11__WEIGHT11__SHIFT 0x18 +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK 0x00000001 +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT 0x00000000 + +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK 0x0000007e +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK 0x00001f80L +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT 0x00000001 +#define DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT 0x00000007 + +#define DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK 0x1fffe000L +#define DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT 0x0000000d + +#define DIDT_SQ_STALL_CTRL__UNUSED_0_MASK 0xe0000000L +#define DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT 0x0000001d + +#define DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK 0x00000001L +#define DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT 0x00000000 + +#define DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK 0x00007ffeL +#define DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT 0x00000001 +#define DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK 0x1fff8000L +#define DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT 0x0000000f + +#define DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK 0x00000001L +#define DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT 0x00000000 + +#define DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK 0x0000007eL +#define DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK 0x00001f80L +#define DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT 0x00000001 +#define DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT 0x00000007 + +#define DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK 0x1fffe000L +#define DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT 0x0000000d + +#define DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK 0x00000fc0L +#define DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK 0x0003f000L +#define DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT 0x00000006 +#define DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT 0x0000000c + +#define DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK 0x00000001L +#define DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK 0x00007ffeL +#define DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK 0x1fff8000L + +#define DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT 0x00000000 +#define DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT 0x00000001 +#define DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT 0x0000000f + +#define DIDT_TD_STALL_CTRL__UNUSED_0_MASK 0xe0000000L +#define DIDT_TD_STALL_CTRL__UNUSED_0__SHIFT 0x0000001d + +#define DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK 0x00000fc0L +#define DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK 0x0003f000L +#define DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT 0x00000006 +#define DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT 0x0000000c + +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK 0x00000001L +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT 0x00000000 + +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK 0x0000007eL +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK 0x00001f80L +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT 0x00000001 +#define DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT 0x00000007 + +#define DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK 0x1fffe000L +#define DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT 0x0000000d + +#define DIDT_TCP_STALL_CTRL__UNUSED_0_MASK 0xe0000000L +#define DIDT_TCP_STALL_CTRL__UNUSED_0__SHIFT 0x0000001d + +#define DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK 0x00000001L +#define DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK 0x00007ffeL +#define DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK 0x1fff8000L +#define DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT 0x00000000 +#define DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT 0x00000001 +#define DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT 0x0000000f + +#define DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK 0x00000fc0L +#define DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK 0x0003f000L +#define DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT 0x00000006 +#define DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT 0x0000000c + #endif /* GFX_8_0_SH_MASK_H */ diff --git a/drivers/gpu/drm/amd/include/atombios.h b/drivers/gpu/drm/amd/include/atombios.h index 32f3e345de08..3493da5c8f0e 100644 --- a/drivers/gpu/drm/amd/include/atombios.h +++ b/drivers/gpu/drm/amd/include/atombios.h @@ -5538,6 +5538,78 @@ typedef struct _ATOM_ASIC_PROFILING_INFO_V3_5 ULONG ulReserved[12]; }ATOM_ASIC_PROFILING_INFO_V3_5; +/* for Polars10/11 AVFS parameters */ +typedef struct _ATOM_ASIC_PROFILING_INFO_V3_6 +{ + ATOM_COMMON_TABLE_HEADER asHeader; + ULONG ulMaxVddc; + ULONG ulMinVddc; + USHORT usLkgEuseIndex; + UCHAR ucLkgEfuseBitLSB; + UCHAR ucLkgEfuseLength; + ULONG ulLkgEncodeLn_MaxDivMin; + ULONG ulLkgEncodeMax; + ULONG ulLkgEncodeMin; + EFUSE_LINEAR_FUNC_PARAM sRoFuse; + ULONG ulEvvDefaultVddc; + ULONG ulEvvNoCalcVddc; + ULONG ulSpeed_Model; + ULONG ulSM_A0; + ULONG ulSM_A1; + ULONG ulSM_A2; + ULONG ulSM_A3; + ULONG ulSM_A4; + ULONG ulSM_A5; + ULONG ulSM_A6; + ULONG ulSM_A7; + UCHAR ucSM_A0_sign; + UCHAR ucSM_A1_sign; + UCHAR ucSM_A2_sign; + UCHAR ucSM_A3_sign; + UCHAR ucSM_A4_sign; + UCHAR ucSM_A5_sign; + UCHAR ucSM_A6_sign; + UCHAR ucSM_A7_sign; + ULONG ulMargin_RO_a; + ULONG ulMargin_RO_b; + ULONG ulMargin_RO_c; + ULONG ulMargin_fixed; + ULONG ulMargin_Fmax_mean; + ULONG ulMargin_plat_mean; + ULONG ulMargin_Fmax_sigma; + ULONG ulMargin_plat_sigma; + ULONG ulMargin_DC_sigma; + ULONG ulLoadLineSlop; + ULONG ulaTDClimitPerDPM[8]; + ULONG ulaNoCalcVddcPerDPM[8]; + ULONG ulAVFS_meanNsigma_Acontant0; + ULONG ulAVFS_meanNsigma_Acontant1; + ULONG ulAVFS_meanNsigma_Acontant2; + USHORT usAVFS_meanNsigma_DC_tol_sigma; + USHORT usAVFS_meanNsigma_Platform_mean; + USHORT usAVFS_meanNsigma_Platform_sigma; + ULONG ulGB_VDROOP_TABLE_CKSOFF_a0; + ULONG ulGB_VDROOP_TABLE_CKSOFF_a1; + ULONG ulGB_VDROOP_TABLE_CKSOFF_a2; + ULONG ulGB_VDROOP_TABLE_CKSON_a0; + ULONG ulGB_VDROOP_TABLE_CKSON_a1; + ULONG ulGB_VDROOP_TABLE_CKSON_a2; + ULONG ulAVFSGB_FUSE_TABLE_CKSOFF_m1; + USHORT usAVFSGB_FUSE_TABLE_CKSOFF_m2; + ULONG ulAVFSGB_FUSE_TABLE_CKSOFF_b; + ULONG ulAVFSGB_FUSE_TABLE_CKSON_m1; + USHORT usAVFSGB_FUSE_TABLE_CKSON_m2; + ULONG ulAVFSGB_FUSE_TABLE_CKSON_b; + USHORT usMaxVoltage_0_25mv; + UCHAR ucEnableGB_VDROOP_TABLE_CKSOFF; + UCHAR ucEnableGB_VDROOP_TABLE_CKSON; + UCHAR ucEnableGB_FUSE_TABLE_CKSOFF; + UCHAR ucEnableGB_FUSE_TABLE_CKSON; + USHORT usPSM_Age_ComFactor; + UCHAR ucEnableApplyAVFS_CKS_OFF_Voltage; + UCHAR ucReserved; +}ATOM_ASIC_PROFILING_INFO_V3_6; + typedef struct _ATOM_SCLK_FCW_RANGE_ENTRY_V1{ ULONG ulMaxSclkFreq; diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h index a461e155a160..0c8c85d2a2a5 100644 --- a/drivers/gpu/drm/amd/include/cgs_common.h +++ b/drivers/gpu/drm/amd/include/cgs_common.h @@ -49,6 +49,7 @@ enum cgs_ind_reg { CGS_IND_REG__SMC, CGS_IND_REG__UVD_CTX, CGS_IND_REG__DIDT, + CGS_IND_REG_GC_CAC, CGS_IND_REG__AUDIO_ENDPT }; @@ -115,6 +116,7 @@ enum cgs_system_info_id { CGS_SYSTEM_INFO_CG_FLAGS, CGS_SYSTEM_INFO_PG_FLAGS, CGS_SYSTEM_INFO_GFX_CU_INFO, + CGS_SYSTEM_INFO_GFX_SE_INFO, CGS_SYSTEM_INFO_ID_MAXIMUM, }; @@ -189,7 +191,6 @@ typedef unsigned long cgs_handle_t; struct cgs_acpi_method_argument { uint32_t type; - uint32_t method_length; uint32_t data_length; union{ uint32_t value; @@ -581,6 +582,9 @@ typedef int (*cgs_get_firmware_info)(struct cgs_device *cgs_device, enum cgs_ucode_id type, struct cgs_firmware_info *info); +typedef int (*cgs_rel_firmware)(struct cgs_device *cgs_device, + enum cgs_ucode_id type); + typedef int(*cgs_set_powergating_state)(struct cgs_device *cgs_device, enum amd_ip_block_type block_type, enum amd_powergating_state state); @@ -645,6 +649,7 @@ struct cgs_ops { cgs_set_camera_voltages_t set_camera_voltages; /* Firmware Info */ cgs_get_firmware_info get_firmware_info; + cgs_rel_firmware rel_firmware; /* cg pg interface*/ cgs_set_powergating_state set_powergating_state; cgs_set_clockgating_state set_clockgating_state; @@ -738,6 +743,8 @@ struct cgs_device CGS_CALL(set_camera_voltages,dev,mask,voltages) #define cgs_get_firmware_info(dev, type, info) \ CGS_CALL(get_firmware_info, dev, type, info) +#define cgs_rel_firmware(dev, type) \ + CGS_CALL(rel_firmware, dev, type) #define cgs_set_powergating_state(dev, block_type, state) \ CGS_CALL(set_powergating_state, dev, block_type, state) #define cgs_set_clockgating_state(dev, block_type, state) \ diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c index 8e345bfddb69..f9e03ad0baa2 100644 --- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c +++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c @@ -73,11 +73,14 @@ static int pp_sw_init(void *handle) ret = hwmgr->hwmgr_func->backend_init(hwmgr); if (ret) - goto err; + goto err1; pr_info("amdgpu: powerplay initialized\n"); return 0; +err1: + if (hwmgr->pptable_func->pptable_fini) + hwmgr->pptable_func->pptable_fini(hwmgr); err: pr_err("amdgpu: powerplay initialization failed\n"); return ret; @@ -100,6 +103,9 @@ static int pp_sw_fini(void *handle) if (hwmgr->hwmgr_func->backend_fini != NULL) ret = hwmgr->hwmgr_func->backend_fini(hwmgr); + if (hwmgr->pptable_func->pptable_fini) + hwmgr->pptable_func->pptable_fini(hwmgr); + return ret; } @@ -530,6 +536,10 @@ int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_event event_id, void *input, case AMD_PP_EVENT_COMPLETE_INIT: ret = pem_handle_event(pp_handle->eventmgr, event_id, &data); break; + case AMD_PP_EVENT_READJUST_POWER_STATE: + pp_handle->hwmgr->current_ps = pp_handle->hwmgr->boot_ps; + ret = pem_handle_event(pp_handle->eventmgr, event_id, &data); + break; default: break; } @@ -734,12 +744,12 @@ static int pp_dpm_get_pp_table(void *handle, char **table) PP_CHECK_HW(hwmgr); - if (hwmgr->hwmgr_func->get_pp_table == NULL) { - printk(KERN_INFO "%s was not implemented.\n", __func__); - return 0; - } + if (!hwmgr->soft_pp_table) + return -EINVAL; + + *table = (char *)hwmgr->soft_pp_table; - return hwmgr->hwmgr_func->get_pp_table(hwmgr, table); + return hwmgr->soft_pp_table_size; } static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size) @@ -753,12 +763,23 @@ static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size) PP_CHECK_HW(hwmgr); - if (hwmgr->hwmgr_func->set_pp_table == NULL) { - printk(KERN_INFO "%s was not implemented.\n", __func__); - return 0; + if (!hwmgr->hardcode_pp_table) { + hwmgr->hardcode_pp_table = + kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL); + + if (!hwmgr->hardcode_pp_table) + return -ENOMEM; + + /* to avoid powerplay crash when hardcode pptable is empty */ + memcpy(hwmgr->hardcode_pp_table, hwmgr->soft_pp_table, + hwmgr->soft_pp_table_size); } - return hwmgr->hwmgr_func->set_pp_table(hwmgr, buf, size); + memcpy(hwmgr->hardcode_pp_table, buf, size); + + hwmgr->soft_pp_table = hwmgr->hardcode_pp_table; + + return amd_powerplay_reset(handle); } static int pp_dpm_force_clock_level(void *handle, @@ -800,6 +821,82 @@ static int pp_dpm_print_clock_levels(void *handle, return hwmgr->hwmgr_func->print_clock_levels(hwmgr, type, buf); } +static int pp_dpm_get_sclk_od(void *handle) +{ + struct pp_hwmgr *hwmgr; + + if (!handle) + return -EINVAL; + + hwmgr = ((struct pp_instance *)handle)->hwmgr; + + PP_CHECK_HW(hwmgr); + + if (hwmgr->hwmgr_func->get_sclk_od == NULL) { + printk(KERN_INFO "%s was not implemented.\n", __func__); + return 0; + } + + return hwmgr->hwmgr_func->get_sclk_od(hwmgr); +} + +static int pp_dpm_set_sclk_od(void *handle, uint32_t value) +{ + struct pp_hwmgr *hwmgr; + + if (!handle) + return -EINVAL; + + hwmgr = ((struct pp_instance *)handle)->hwmgr; + + PP_CHECK_HW(hwmgr); + + if (hwmgr->hwmgr_func->set_sclk_od == NULL) { + printk(KERN_INFO "%s was not implemented.\n", __func__); + return 0; + } + + return hwmgr->hwmgr_func->set_sclk_od(hwmgr, value); +} + +static int pp_dpm_get_mclk_od(void *handle) +{ + struct pp_hwmgr *hwmgr; + + if (!handle) + return -EINVAL; + + hwmgr = ((struct pp_instance *)handle)->hwmgr; + + PP_CHECK_HW(hwmgr); + + if (hwmgr->hwmgr_func->get_mclk_od == NULL) { + printk(KERN_INFO "%s was not implemented.\n", __func__); + return 0; + } + + return hwmgr->hwmgr_func->get_mclk_od(hwmgr); +} + +static int pp_dpm_set_mclk_od(void *handle, uint32_t value) +{ + struct pp_hwmgr *hwmgr; + + if (!handle) + return -EINVAL; + + hwmgr = ((struct pp_instance *)handle)->hwmgr; + + PP_CHECK_HW(hwmgr); + + if (hwmgr->hwmgr_func->set_mclk_od == NULL) { + printk(KERN_INFO "%s was not implemented.\n", __func__); + return 0; + } + + return hwmgr->hwmgr_func->set_mclk_od(hwmgr, value); +} + const struct amd_powerplay_funcs pp_dpm_funcs = { .get_temperature = pp_dpm_get_temperature, .load_firmware = pp_dpm_load_fw, @@ -822,6 +919,10 @@ const struct amd_powerplay_funcs pp_dpm_funcs = { .set_pp_table = pp_dpm_set_pp_table, .force_clock_level = pp_dpm_force_clock_level, .print_clock_levels = pp_dpm_print_clock_levels, + .get_sclk_od = pp_dpm_get_sclk_od, + .set_sclk_od = pp_dpm_set_sclk_od, + .get_mclk_od = pp_dpm_get_mclk_od, + .set_mclk_od = pp_dpm_set_mclk_od, }; static int amd_pp_instance_init(struct amd_pp_init *pp_init, @@ -903,6 +1004,44 @@ int amd_powerplay_fini(void *handle) return 0; } +int amd_powerplay_reset(void *handle) +{ + struct pp_instance *instance = (struct pp_instance *)handle; + struct pp_eventmgr *eventmgr; + struct pem_event_data event_data = { {0} }; + int ret; + + if (instance == NULL) + return -EINVAL; + + eventmgr = instance->eventmgr; + if (!eventmgr || !eventmgr->pp_eventmgr_fini) + return -EINVAL; + + eventmgr->pp_eventmgr_fini(eventmgr); + + ret = pp_sw_fini(handle); + if (ret) + return ret; + + kfree(instance->hwmgr->ps); + + ret = pp_sw_init(handle); + if (ret) + return ret; + + hw_init_power_state_table(instance->hwmgr); + + if (eventmgr == NULL || eventmgr->pp_eventmgr_init == NULL) + return -EINVAL; + + ret = eventmgr->pp_eventmgr_init(eventmgr); + if (ret) + return ret; + + return pem_handle_event(eventmgr, AMD_PP_EVENT_COMPLETE_INIT, &event_data); +} + /* export this function to DAL */ int amd_powerplay_display_configuration_change(void *handle, diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventmgr.c b/drivers/gpu/drm/amd/powerplay/eventmgr/eventmgr.c index 46410e3c7349..fb88e4e5d625 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventmgr.c +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventmgr.c @@ -58,9 +58,6 @@ static void pem_fini(struct pp_eventmgr *eventmgr) pem_unregister_interrupts(eventmgr); pem_handle_event(eventmgr, AMD_PP_EVENT_UNINITIALIZE, &event_data); - - if (eventmgr != NULL) - kfree(eventmgr); } int eventmgr_init(struct pp_instance *handle) diff --git a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c index 5cd123472db4..b6f45fd01fa6 100644 --- a/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c +++ b/drivers/gpu/drm/amd/powerplay/eventmgr/eventtasks.c @@ -132,8 +132,7 @@ int pem_task_enable_dynamic_state_management(struct pp_eventmgr *eventmgr, struc int pem_task_disable_dynamic_state_management(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data) { - /* TODO */ - return 0; + return phm_disable_dynamic_state_management(eventmgr->hwmgr); } int pem_task_enable_clock_power_gatings_tasks(struct pp_eventmgr *eventmgr, struct pem_event_data *event_data) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c index 436fc16dabb6..2da548f6337e 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_clockpowergating.c @@ -206,7 +206,7 @@ int cz_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate) AMD_IP_BLOCK_TYPE_VCE, AMD_PG_STATE_GATE); cz_enable_disable_vce_dpm(hwmgr, false); - /* TODO: to figure out why vce can't be poweroff*/ + cz_dpm_powerdown_vce(hwmgr); cz_hwmgr->vce_power_gated = true; } else { cz_dpm_powerup_vce(hwmgr); @@ -225,6 +225,7 @@ int cz_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate) } } } else { + cz_hwmgr->vce_power_gated = bgate; cz_dpm_update_vce_dpm(hwmgr); cz_enable_disable_vce_dpm(hwmgr, !bgate); return 0; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c index 1f14c477d15d..9bf622e123b6 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/cz_hwmgr.c @@ -1180,6 +1180,13 @@ static int cz_apply_state_adjust_rules(struct pp_hwmgr *hwmgr, static int cz_hwmgr_backend_init(struct pp_hwmgr *hwmgr) { int result = 0; + struct cz_hwmgr *data; + + data = kzalloc(sizeof(struct cz_hwmgr), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + + hwmgr->backend = data; result = cz_initialize_dpm_defaults(hwmgr); if (result != 0) { @@ -1909,15 +1916,7 @@ static const struct pp_hwmgr_func cz_hwmgr_funcs = { int cz_hwmgr_init(struct pp_hwmgr *hwmgr) { - struct cz_hwmgr *cz_hwmgr; - int ret = 0; - - cz_hwmgr = kzalloc(sizeof(struct cz_hwmgr), GFP_KERNEL); - if (cz_hwmgr == NULL) - return -ENOMEM; - - hwmgr->backend = cz_hwmgr; hwmgr->hwmgr_func = &cz_hwmgr_funcs; hwmgr->pptable_func = &pptable_funcs; - return ret; + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c index 24a16e49b571..744aa886a2be 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.c @@ -581,25 +581,24 @@ static int fiji_patch_boot_state(struct pp_hwmgr *hwmgr, static int fiji_hwmgr_backend_fini(struct pp_hwmgr *hwmgr) { - struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); - - if (data->soft_pp_table) { - kfree(data->soft_pp_table); - data->soft_pp_table = NULL; - } - return phm_hwmgr_backend_fini(hwmgr); } static int fiji_hwmgr_backend_init(struct pp_hwmgr *hwmgr) { - struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_hwmgr *data; uint32_t i; struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); bool stay_in_boot; int result; + data = kzalloc(sizeof(struct fiji_hwmgr), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + + hwmgr->backend = data; + data->dll_default_on = false; data->sram_end = SMC_RAM_END; @@ -633,6 +632,8 @@ static int fiji_hwmgr_backend_init(struct pp_hwmgr *hwmgr) data->vddci_control = FIJI_VOLTAGE_CONTROL_NONE; data->mvdd_control = FIJI_VOLTAGE_CONTROL_NONE; + data->force_pcie_gen = PP_PCIEGenInvalid; + if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr, VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_SVID2)) data->voltage_control = FIJI_VOLTAGE_CONTROL_BY_SVID2; @@ -732,7 +733,7 @@ static int fiji_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_GEN_INFO; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_gen_cap = 0x30007; + data->pcie_gen_cap = AMDGPU_DEFAULT_PCIE_GEN_MASK; else data->pcie_gen_cap = (uint32_t)sys_info.value; if (data->pcie_gen_cap & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) @@ -741,7 +742,7 @@ static int fiji_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_MLW; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_lane_cap = 0x2f0000; + data->pcie_lane_cap = AMDGPU_DEFAULT_PCIE_MLW_MASK; else data->pcie_lane_cap = (uint32_t)sys_info.value; } else { @@ -1234,6 +1235,34 @@ static int fiji_program_voting_clients(struct pp_hwmgr *hwmgr) return 0; } +static int fiji_clear_voting_clients(struct pp_hwmgr *hwmgr) +{ + /* Reset voting clients before disabling DPM */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + SCLK_PWRMGT_CNTL, RESET_SCLK_CNT, 1); + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + SCLK_PWRMGT_CNTL, RESET_BUSY_CNT, 1); + + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_0, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_1, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_2, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_3, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_4, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_5, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_6, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_7, 0); + + return 0; +} + /** * Get the location of various tables inside the FW image. * @@ -1361,6 +1390,17 @@ static int fiji_copy_and_switch_arb_sets(struct pp_hwmgr *hwmgr, } /** +* Call SMC to reset S0/S1 to S1 and Reset SMIO to initial value +* +* @param hwmgr the address of the powerplay hardware manager. +* @return if success then 0; +*/ +static int fiji_reset_to_default(struct pp_hwmgr *hwmgr) +{ + return smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_ResetToDefaults); +} + +/** * Initial switch from ARB F0->F1 * * @param hwmgr the address of the powerplay hardware manager. @@ -1373,6 +1413,21 @@ static int fiji_initial_switch_from_arbf0_to_f1(struct pp_hwmgr *hwmgr) MC_CG_ARB_FREQ_F0, MC_CG_ARB_FREQ_F1); } +static int fiji_force_switch_to_arbf0(struct pp_hwmgr *hwmgr) +{ + uint32_t tmp; + + tmp = (cgs_read_ind_register(hwmgr->device, + CGS_IND_REG__SMC, ixSMC_SCRATCH9) & + 0x0000ff00) >> 8; + + if (tmp == MC_CG_ARB_FREQ_F0) + return 0; + + return fiji_copy_and_switch_arb_sets(hwmgr, + tmp, MC_CG_ARB_FREQ_F0); +} + static int fiji_reset_single_dpm_table(struct pp_hwmgr *hwmgr, struct fiji_single_dpm_table *dpm_table, uint32_t count) { @@ -1830,7 +1885,7 @@ static uint16_t fiji_find_closest_vddci(struct pp_hwmgr *hwmgr, uint16_t vddci) PP_ASSERT_WITH_CODE(false, "VDDCI is larger than max VDDCI in VDDCI Voltage Table!", - return vddci_table->entries[i].value); + return vddci_table->entries[i-1].value); } static int fiji_get_dependency_volt_by_clk(struct pp_hwmgr *hwmgr, @@ -3175,6 +3230,17 @@ static int fiji_enable_ulv(struct pp_hwmgr *hwmgr) return 0; } +static int fiji_disable_ulv(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_ulv_parm *ulv = &(data->ulv); + + if (ulv->ulv_supported) + return smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_DisableULV); + + return 0; +} + static int fiji_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, @@ -3195,6 +3261,21 @@ static int fiji_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) return 0; } +static int fiji_disable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) +{ + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_SclkDeepSleep)) { + if (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_MASTER_DeepSleep_OFF)) { + PP_ASSERT_WITH_CODE(false, + "Attempt to disable Master Deep Sleep switch failed!", + return -1); + } + } + + return 0; +} + static int fiji_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) { struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -3355,6 +3436,70 @@ static int fiji_start_dpm(struct pp_hwmgr *hwmgr) return 0; } +static int fiji_disable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + + /* disable SCLK dpm */ + if (!data->sclk_dpm_key_disabled) + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_DPM_Disable) == 0), + "Failed to disable SCLK DPM!", + return -1); + + /* disable MCLK dpm */ + if (!data->mclk_dpm_key_disabled) { + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, + PPSMC_MSG_MCLKDPM_SetEnabledMask, 1) == 0), + "Failed to force MCLK DPM0!", + return -1); + + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_MCLKDPM_Disable) == 0), + "Failed to disable MCLK DPM!", + return -1); + } + + return 0; +} + +static int fiji_stop_dpm(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + + /* disable general power management */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, + GLOBAL_PWRMGT_EN, 0); + /* disable sclk deep sleep */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SCLK_PWRMGT_CNTL, + DYNAMIC_PM_EN, 0); + + /* disable PCIE dpm */ + if (!data->pcie_dpm_key_disabled) { + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_PCIeDPM_Disable) == 0), + "Failed to disable pcie DPM during DPM Stop Function!", + return -1); + } + + if (fiji_disable_sclk_mclk_dpm(hwmgr)) { + printk(KERN_ERR "Failed to disable Sclk DPM and Mclk DPM!"); + return -1; + } + + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_Voltage_Cntl_Disable) == 0), + "Failed to disable voltage DPM during DPM Stop Function!", + return -1); + + return 0; +} + static void fiji_set_dpm_event_sources(struct pp_hwmgr *hwmgr, uint32_t sources) { @@ -3413,6 +3558,23 @@ static int fiji_enable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) return fiji_enable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); } +static int fiji_disable_auto_throttle_source(struct pp_hwmgr *hwmgr, + PHM_AutoThrottleSource source) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + + if (data->active_auto_throttle_sources & (1 << source)) { + data->active_auto_throttle_sources &= ~(1 << source); + fiji_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources); + } + return 0; +} + +static int fiji_disable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) +{ + return fiji_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); +} + static int fiji_enable_dpm_tasks(struct pp_hwmgr *hwmgr) { int tmp_result, result = 0; @@ -3527,6 +3689,64 @@ static int fiji_enable_dpm_tasks(struct pp_hwmgr *hwmgr) return result; } +static int fiji_disable_dpm_tasks(struct pp_hwmgr *hwmgr) +{ + int tmp_result, result = 0; + + tmp_result = (fiji_is_dpm_running(hwmgr)) ? 0 : -1; + PP_ASSERT_WITH_CODE(tmp_result == 0, + "DPM is not running right now, no need to disable DPM!", + return 0); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_ThermalController)) + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, 1); + + tmp_result = fiji_disable_power_containment(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable power containment!", result = tmp_result); + + tmp_result = fiji_disable_smc_cac(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable SMC CAC!", result = tmp_result); + + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + CG_SPLL_SPREAD_SPECTRUM, SSEN, 0); + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + GENERAL_PWRMGT, DYN_SPREAD_SPECTRUM_EN, 0); + + tmp_result = fiji_disable_thermal_auto_throttle(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable thermal auto throttle!", result = tmp_result); + + tmp_result = fiji_stop_dpm(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to stop DPM!", result = tmp_result); + + tmp_result = fiji_disable_deep_sleep_master_switch(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable deep sleep master switch!", result = tmp_result); + + tmp_result = fiji_disable_ulv(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable ULV!", result = tmp_result); + + tmp_result = fiji_clear_voting_clients(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to clear voting clients!", result = tmp_result); + + tmp_result = fiji_reset_to_default(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to reset to default!", result = tmp_result); + + tmp_result = fiji_force_switch_to_arbf0(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to force to switch arbf0!", result = tmp_result); + + return result; +} + static int fiji_force_dpm_highest(struct pp_hwmgr *hwmgr) { struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -5069,42 +5289,6 @@ static int fiji_get_fan_control_mode(struct pp_hwmgr *hwmgr) CG_FDO_CTRL2, FDO_PWM_MODE); } -static int fiji_get_pp_table(struct pp_hwmgr *hwmgr, char **table) -{ - struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kmemdup(hwmgr->soft_pp_table, - hwmgr->soft_pp_table_size, - GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - *table = (char *)&data->soft_pp_table; - - return hwmgr->soft_pp_table_size; -} - -static int fiji_set_pp_table(struct pp_hwmgr *hwmgr, const char *buf, size_t size) -{ - struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - memcpy(data->soft_pp_table, buf, size); - - hwmgr->soft_pp_table = data->soft_pp_table; - - /* TODO: re-init powerplay to implement modified pptable */ - - return 0; -} - static int fiji_force_clock_level(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask) { @@ -5274,12 +5458,96 @@ bool fiji_check_smc_update_required_for_display_configuration(struct pp_hwmgr *h return is_update_required; } +static int fiji_get_sclk_od(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_single_dpm_table *sclk_table = &(data->dpm_table.sclk_table); + struct fiji_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + int value; + + value = (sclk_table->dpm_levels[sclk_table->count - 1].value - + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value) * + 100 / + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return value; +} + +static int fiji_set_sclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + struct pp_power_state *ps; + struct fiji_power_state *fiji_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + fiji_ps = cast_phw_fiji_power_state(&ps->hardware); + + fiji_ps->performance_levels[fiji_ps->performance_level_count - 1].engine_clock = + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value * + value / 100 + + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return 0; +} + +static int fiji_get_mclk_od(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_single_dpm_table *mclk_table = &(data->dpm_table.mclk_table); + struct fiji_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + int value; + + value = (mclk_table->dpm_levels[mclk_table->count - 1].value - + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value) * + 100 / + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return value; +} + +static int fiji_set_mclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + struct fiji_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + struct pp_power_state *ps; + struct fiji_power_state *fiji_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + fiji_ps = cast_phw_fiji_power_state(&ps->hardware); + + fiji_ps->performance_levels[fiji_ps->performance_level_count - 1].memory_clock = + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value * + value / 100 + + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return 0; +} static const struct pp_hwmgr_func fiji_hwmgr_funcs = { .backend_init = &fiji_hwmgr_backend_init, .backend_fini = &fiji_hwmgr_backend_fini, .asic_setup = &fiji_setup_asic_task, .dynamic_state_management_enable = &fiji_enable_dpm_tasks, + .dynamic_state_management_disable = &fiji_disable_dpm_tasks, .force_dpm_level = &fiji_dpm_force_dpm_level, .get_num_of_pp_table_entries = &tonga_get_number_of_powerplay_table_entries, .get_power_state_size = &fiji_get_power_state_size, @@ -5312,24 +5580,18 @@ static const struct pp_hwmgr_func fiji_hwmgr_funcs = { .get_fan_control_mode = fiji_get_fan_control_mode, .check_states_equal = fiji_check_states_equal, .check_smc_update_required_for_display_configuration = fiji_check_smc_update_required_for_display_configuration, - .get_pp_table = fiji_get_pp_table, - .set_pp_table = fiji_set_pp_table, .force_clock_level = fiji_force_clock_level, .print_clock_levels = fiji_print_clock_levels, + .get_sclk_od = fiji_get_sclk_od, + .set_sclk_od = fiji_set_sclk_od, + .get_mclk_od = fiji_get_mclk_od, + .set_mclk_od = fiji_set_mclk_od, }; int fiji_hwmgr_init(struct pp_hwmgr *hwmgr) { - struct fiji_hwmgr *data; - int ret = 0; - - data = kzalloc(sizeof(struct fiji_hwmgr), GFP_KERNEL); - if (data == NULL) - return -ENOMEM; - - hwmgr->backend = data; hwmgr->hwmgr_func = &fiji_hwmgr_funcs; hwmgr->pptable_func = &tonga_pptable_funcs; pp_fiji_thermal_initialize(hwmgr); - return ret; + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.h index 170edf5a772d..bf67c2a92c68 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_hwmgr.h @@ -302,9 +302,6 @@ struct fiji_hwmgr { bool pg_acp_init; bool frtc_enabled; bool frtc_status_changed; - - /* soft pptable for re-uploading into smu */ - void *soft_pp_table; }; /* To convert to Q8.8 format for firmware */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.c index db23a4068baf..44658451a8d2 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.c @@ -73,17 +73,18 @@ void fiji_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr) if (!tmp) { phm_cap_set(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_PowerContainment); - - phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_CAC); fiji_hwmgr->fast_watermark_threshold = 100; - tmp = 1; - fiji_hwmgr->enable_dte_feature = tmp ? false : true; - fiji_hwmgr->enable_tdc_limit_feature = tmp ? true : false; - fiji_hwmgr->enable_pkg_pwr_tracking_feature = tmp ? true : false; + if (hwmgr->powercontainment_enabled) { + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment); + tmp = 1; + fiji_hwmgr->enable_dte_feature = tmp ? false : true; + fiji_hwmgr->enable_tdc_limit_feature = tmp ? true : false; + fiji_hwmgr->enable_pkg_pwr_tracking_feature = tmp ? true : false; + } } } @@ -459,6 +460,23 @@ int fiji_enable_smc_cac(struct pp_hwmgr *hwmgr) return result; } +int fiji_disable_smc_cac(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + int result = 0; + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_CAC) && data->cac_enabled) { + int smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_DisableCac)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable CAC in SMC.", result = -1); + + data->cac_enabled = false; + } + return result; +} + int fiji_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n) { struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); @@ -528,6 +546,48 @@ int fiji_enable_power_containment(struct pp_hwmgr *hwmgr) return result; } +int fiji_disable_power_containment(struct pp_hwmgr *hwmgr) +{ + struct fiji_hwmgr *data = (struct fiji_hwmgr *)(hwmgr->backend); + int result = 0; + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment) && + data->power_containment_features) { + int smc_result; + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_TDCLimit) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_TDCLimitDisable)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable TDCLimit in SMC.", + result = smc_result); + } + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_DTE) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_DisableDTE)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable DTE in SMC.", + result = smc_result); + } + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_PkgPwrLimit) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_PkgPwrLimitDisable)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable PkgPwrTracking in SMC.", + result = smc_result); + } + data->power_containment_features = 0; + } + + return result; +} + int fiji_power_control_set_level(struct pp_hwmgr *hwmgr) { struct phm_ppt_v1_information *table_info = diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.h b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.h index 55e58200f33a..fec772421733 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/fiji_powertune.h @@ -36,6 +36,19 @@ enum fiji_pt_config_reg_type { #define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002 #define POWERCONTAINMENT_FEATURE_PkgPwrLimit 0x00000004 +#define DIDT_SQ_CTRL0__UNUSED_0_MASK 0xffffffc0 +#define DIDT_SQ_CTRL0__UNUSED_0__SHIFT 0x6 +#define DIDT_TD_CTRL0__UNUSED_0_MASK 0xffffffc0 +#define DIDT_TD_CTRL0__UNUSED_0__SHIFT 0x6 +#define DIDT_TCP_CTRL0__UNUSED_0_MASK 0xffffffc0 +#define DIDT_TCP_CTRL0__UNUSED_0__SHIFT 0x6 +#define DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK 0xe0000000 +#define DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001d +#define DIDT_TD_TUNING_CTRL__UNUSED_0_MASK 0xe0000000 +#define DIDT_TD_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001d +#define DIDT_TCP_TUNING_CTRL__UNUSED_0_MASK 0xe0000000 +#define DIDT_TCP_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001d + struct fiji_pt_config_reg { uint32_t offset; uint32_t mask; @@ -58,7 +71,9 @@ void fiji_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr); int fiji_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr); int fiji_populate_pm_fuses(struct pp_hwmgr *hwmgr); int fiji_enable_smc_cac(struct pp_hwmgr *hwmgr); +int fiji_disable_smc_cac(struct pp_hwmgr *hwmgr); int fiji_enable_power_containment(struct pp_hwmgr *hwmgr); +int fiji_disable_power_containment(struct pp_hwmgr *hwmgr); int fiji_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n); int fiji_power_control_set_level(struct pp_hwmgr *hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c index fa208ada6892..789f98ad2615 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c @@ -154,6 +154,30 @@ int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr) return ret; } +int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr) +{ + int ret = -1; + bool enabled; + + PHM_FUNC_CHECK(hwmgr); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_TablelessHardwareInterface)) { + if (hwmgr->hwmgr_func->dynamic_state_management_disable) + ret = hwmgr->hwmgr_func->dynamic_state_management_disable(hwmgr); + } else { + ret = phm_dispatch_table(hwmgr, + &(hwmgr->disable_dynamic_state_management), + NULL, NULL); + } + + enabled = ret == 0 ? false : true; + + cgs_notify_dpm_enabled(hwmgr->device, enabled); + + return ret; +} + int phm_force_dpm_levels(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level) { PHM_FUNC_CHECK(hwmgr); @@ -306,11 +330,15 @@ int phm_store_dal_configuration_data(struct pp_hwmgr *hwmgr, { PHM_FUNC_CHECK(hwmgr); - if (hwmgr->hwmgr_func->store_cc6_data == NULL) + if (display_config == NULL) return -EINVAL; hwmgr->display_config = *display_config; - /* to do pass other display configuration in furture */ + + if (hwmgr->hwmgr_func->store_cc6_data == NULL) + return -EINVAL; + + /* TODO: pass other display configuration in the future */ if (hwmgr->hwmgr_func->store_cc6_data) hwmgr->hwmgr_func->store_cc6_data(hwmgr, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c index 1c48917da3cf..03b6128ebc20 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c @@ -24,6 +24,7 @@ #include <linux/types.h> #include <linux/kernel.h> #include <linux/slab.h> +#include <drm/amdgpu_drm.h> #include "cgs_common.h" #include "power_state.h" #include "hwmgr.h" @@ -58,12 +59,13 @@ int hwmgr_init(struct amd_pp_init *pp_init, struct pp_instance *handle) hwmgr->hw_revision = pp_init->rev_id; hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT; hwmgr->power_source = PP_PowerSource_AC; + hwmgr->powercontainment_enabled = pp_init->powercontainment_enabled; switch (hwmgr->chip_family) { - case AMD_FAMILY_CZ: + case AMDGPU_FAMILY_CZ: cz_hwmgr_init(hwmgr); break; - case AMD_FAMILY_VI: + case AMDGPU_FAMILY_VI: switch (hwmgr->chip_id) { case CHIP_TONGA: tonga_hwmgr_init(hwmgr); @@ -93,6 +95,15 @@ int hwmgr_fini(struct pp_hwmgr *hwmgr) if (hwmgr == NULL || hwmgr->ps == NULL) return -EINVAL; + /* do hwmgr finish*/ + kfree(hwmgr->hardcode_pp_table); + + kfree(hwmgr->backend); + + kfree(hwmgr->start_thermal_controller.function_list); + + kfree(hwmgr->set_temperature_range.function_list); + kfree(hwmgr->ps); kfree(hwmgr); return 0; @@ -462,7 +473,7 @@ uint16_t phm_find_closest_vddci(struct pp_atomctrl_voltage_table *vddci_table, u PP_ASSERT_WITH_CODE(false, "VDDCI is larger than max VDDCI in VDDCI Voltage Table!", - return vddci_table->entries[i].value); + return vddci_table->entries[i-1].value); } int phm_find_boot_level(void *table, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr_ppt.h b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr_ppt.h index 347fef127ce9..2930a3355948 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr_ppt.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr_ppt.h @@ -39,6 +39,7 @@ struct phm_ppt_v1_clock_voltage_dependency_record { uint8_t phases; uint8_t cks_enable; uint8_t cks_voffset; + uint32_t sclk_offset; }; typedef struct phm_ppt_v1_clock_voltage_dependency_record phm_ppt_v1_clock_voltage_dependency_record; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c index 8f142a74ad08..aeec25c66aa8 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_clockpowergating.c @@ -106,11 +106,17 @@ int polaris10_phm_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate) data->uvd_power_gated = bgate; if (bgate) { + cgs_set_clockgating_state(hwmgr->device, + AMD_IP_BLOCK_TYPE_UVD, + AMD_CG_STATE_GATE); polaris10_update_uvd_dpm(hwmgr, true); polaris10_phm_powerdown_uvd(hwmgr); } else { polaris10_phm_powerup_uvd(hwmgr); polaris10_update_uvd_dpm(hwmgr, false); + cgs_set_clockgating_state(hwmgr->device, + AMD_IP_BLOCK_TYPE_UVD, + AMD_PG_STATE_UNGATE); } return 0; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c index aa6be033f21b..9d764c4d253e 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.c @@ -389,6 +389,34 @@ static int polaris10_program_voting_clients(struct pp_hwmgr *hwmgr) return 0; } +static int polaris10_clear_voting_clients(struct pp_hwmgr *hwmgr) +{ + /* Reset voting clients before disabling DPM */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + SCLK_PWRMGT_CNTL, RESET_SCLK_CNT, 1); + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + SCLK_PWRMGT_CNTL, RESET_BUSY_CNT, 1); + + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_0, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_1, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_2, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_3, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_4, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_5, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_6, 0); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + ixCG_FREQ_TRAN_VOTING_7, 0); + + return 0; +} + /** * Get the location of various tables inside the FW image. * @@ -515,6 +543,11 @@ static int polaris10_copy_and_switch_arb_sets(struct pp_hwmgr *hwmgr, return 0; } +static int polaris10_reset_to_default(struct pp_hwmgr *hwmgr) +{ + return smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_ResetToDefaults); +} + /** * Initial switch from ARB F0->F1 * @@ -528,6 +561,21 @@ static int polaris10_initial_switch_from_arbf0_to_f1(struct pp_hwmgr *hwmgr) MC_CG_ARB_FREQ_F0, MC_CG_ARB_FREQ_F1); } +static int polaris10_force_switch_to_arbf0(struct pp_hwmgr *hwmgr) +{ + uint32_t tmp; + + tmp = (cgs_read_ind_register(hwmgr->device, + CGS_IND_REG__SMC, ixSMC_SCRATCH9) & + 0x0000ff00) >> 8; + + if (tmp == MC_CG_ARB_FREQ_F0) + return 0; + + return polaris10_copy_and_switch_arb_sets(hwmgr, + tmp, MC_CG_ARB_FREQ_F0); +} + static int polaris10_setup_default_pcie_table(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -999,7 +1047,7 @@ static int polaris10_get_dependency_volt_by_clk(struct pp_hwmgr *hwmgr, vddci = phm_find_closest_vddci(&(data->vddci_voltage_table), (dep_table->entries[i].vddc - (uint16_t)data->vddc_vddci_delta)); - *voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT; + *voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT; } if (POLARIS10_VOLTAGE_CONTROL_NONE == data->mvdd_control) @@ -1296,7 +1344,6 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr, } mem_level->MclkFrequency = clock; - mem_level->StutterEnable = 0; mem_level->EnabledForThrottle = 1; mem_level->EnabledForActivity = 0; mem_level->UpHyst = 0; @@ -1304,7 +1351,6 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr, mem_level->VoltageDownHyst = 0; mem_level->ActivityLevel = (uint16_t)data->mclk_activity_target; mem_level->StutterEnable = false; - mem_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW; data->display_timing.num_existing_displays = info.display_count; @@ -1358,12 +1404,12 @@ static int polaris10_populate_all_memory_levels(struct pp_hwmgr *hwmgr) return result; } - /* in order to prevent MC activity from stutter mode to push DPM up. + /* In order to prevent MC activity from stutter mode to push DPM up, * the UVD change complements this by putting the MCLK in - * a higher state by default such that we are not effected by + * a higher state by default such that we are not affected by * up threshold or and MCLK DPM latency. */ - levels[0].ActivityLevel = (uint16_t)data->mclk_dpm0_activity_target; + levels[0].ActivityLevel = 0x1f; CONVERT_FROM_HOST_TO_SMC_US(levels[0].ActivityLevel); data->smc_state_table.MemoryDpmLevelCount = @@ -1761,12 +1807,9 @@ static int polaris10_populate_smc_initailial_state(struct pp_hwmgr *hwmgr) static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr) { - uint32_t ro, efuse, efuse2, clock_freq, volt_without_cks, - volt_with_cks, value; - uint16_t clock_freq_u16; + uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min; struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); - uint8_t type, i, j, cks_setting, stretch_amount, stretch_amount2, - volt_offset = 0; + uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0; struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table = @@ -1778,50 +1821,38 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr) * if the part is SS or FF. if RO >= 1660MHz, part is FF. */ efuse = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, - ixSMU_EFUSE_0 + (146 * 4)); - efuse2 = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, - ixSMU_EFUSE_0 + (148 * 4)); + ixSMU_EFUSE_0 + (67 * 4)); efuse &= 0xFF000000; efuse = efuse >> 24; - efuse2 &= 0xF; - - if (efuse2 == 1) - ro = (2300 - 1350) * efuse / 255 + 1350; - else - ro = (2500 - 1000) * efuse / 255 + 1000; - if (ro >= 1660) - type = 0; - else - type = 1; + if (hwmgr->chip_id == CHIP_POLARIS10) { + min = 1000; + max = 2300; + } else { + min = 1100; + max = 2100; + } - /* Populate Stretch amount */ - data->smc_state_table.ClockStretcherAmount = stretch_amount; + ro = efuse * (max -min)/255 + min; /* Populate Sclk_CKS_masterEn0_7 and Sclk_voltageOffset */ for (i = 0; i < sclk_table->count; i++) { data->smc_state_table.Sclk_CKS_masterEn0_7 |= sclk_table->entries[i].cks_enable << i; - volt_without_cks = (uint32_t)((14041 * - (sclk_table->entries[i].clk/100) / 10000 + 3571 + 75 - ro) * 1000 / - (4026 - (13924 * (sclk_table->entries[i].clk/100) / 10000))); - volt_with_cks = (uint32_t)((13946 * - (sclk_table->entries[i].clk/100) / 10000 + 3320 + 45 - ro) * 1000 / - (3664 - (11454 * (sclk_table->entries[i].clk/100) / 10000))); + + volt_without_cks = (uint32_t)(((ro - 40) * 1000 - 2753594 - sclk_table->entries[i].clk/100 * 136418 /1000) / \ + (sclk_table->entries[i].clk/100 * 1132925 /10000 - 242418)/100); + + volt_with_cks = (uint32_t)((ro * 1000 -2396351 - sclk_table->entries[i].clk/100 * 329021/1000) / \ + (sclk_table->entries[i].clk/10000 * 649434 /1000 - 18005)/10); + if (volt_without_cks >= volt_with_cks) volt_offset = (uint8_t)(((volt_without_cks - volt_with_cks + sclk_table->entries[i].cks_voffset) * 100 / 625) + 1); + data->smc_state_table.Sclk_voltageOffset[i] = volt_offset; } - PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, PWR_CKS_ENABLE, - STRETCH_ENABLE, 0x0); - PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, PWR_CKS_ENABLE, - masterReset, 0x1); - /* PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, PWR_CKS_ENABLE, staticEnable, 0x1); */ - PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, PWR_CKS_ENABLE, - masterReset, 0x0); - /* Populate CKS Lookup Table */ if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5) stretch_amount2 = 0; @@ -1835,69 +1866,6 @@ static int polaris10_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr) return -EINVAL); } - value = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, - ixPWR_CKS_CNTL); - value &= 0xFFC2FF87; - data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].minFreq = - polaris10_clock_stretcher_lookup_table[stretch_amount2][0]; - data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].maxFreq = - polaris10_clock_stretcher_lookup_table[stretch_amount2][1]; - clock_freq_u16 = (uint16_t)(PP_SMC_TO_HOST_UL(data->smc_state_table. - GraphicsLevel[data->smc_state_table.GraphicsDpmLevelCount - 1].SclkSetting.SclkFrequency) / 100); - if (polaris10_clock_stretcher_lookup_table[stretch_amount2][0] < clock_freq_u16 - && polaris10_clock_stretcher_lookup_table[stretch_amount2][1] > clock_freq_u16) { - /* Program PWR_CKS_CNTL. CKS_USE_FOR_LOW_FREQ */ - value |= (polaris10_clock_stretcher_lookup_table[stretch_amount2][3]) << 16; - /* Program PWR_CKS_CNTL. CKS_LDO_REFSEL */ - value |= (polaris10_clock_stretcher_lookup_table[stretch_amount2][2]) << 18; - /* Program PWR_CKS_CNTL. CKS_STRETCH_AMOUNT */ - value |= (polaris10_clock_stretch_amount_conversion - [polaris10_clock_stretcher_lookup_table[stretch_amount2][3]] - [stretch_amount]) << 3; - } - CONVERT_FROM_HOST_TO_SMC_US(data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].minFreq); - CONVERT_FROM_HOST_TO_SMC_US(data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].maxFreq); - data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].setting = - polaris10_clock_stretcher_lookup_table[stretch_amount2][2] & 0x7F; - data->smc_state_table.CKS_LOOKUPTable.CKS_LOOKUPTableEntry[0].setting |= - (polaris10_clock_stretcher_lookup_table[stretch_amount2][3]) << 7; - - cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, - ixPWR_CKS_CNTL, value); - - /* Populate DDT Lookup Table */ - for (i = 0; i < 4; i++) { - /* Assign the minimum and maximum VID stored - * in the last row of Clock Stretcher Voltage Table. - */ - data->smc_state_table.ClockStretcherDataTable.ClockStretcherDataTableEntry[i].minVID = - (uint8_t) polaris10_clock_stretcher_ddt_table[type][i][2]; - data->smc_state_table.ClockStretcherDataTable.ClockStretcherDataTableEntry[i].maxVID = - (uint8_t) polaris10_clock_stretcher_ddt_table[type][i][3]; - /* Loop through each SCLK and check the frequency - * to see if it lies within the frequency for clock stretcher. - */ - for (j = 0; j < data->smc_state_table.GraphicsDpmLevelCount; j++) { - cks_setting = 0; - clock_freq = PP_SMC_TO_HOST_UL( - data->smc_state_table.GraphicsLevel[j].SclkSetting.SclkFrequency); - /* Check the allowed frequency against the sclk level[j]. - * Sclk's endianness has already been converted, - * and it's in 10Khz unit, - * as opposed to Data table, which is in Mhz unit. - */ - if (clock_freq >= (polaris10_clock_stretcher_ddt_table[type][i][0]) * 100) { - cks_setting |= 0x2; - if (clock_freq < (polaris10_clock_stretcher_ddt_table[type][i][1]) * 100) - cks_setting |= 0x1; - } - data->smc_state_table.ClockStretcherDataTable.ClockStretcherDataTableEntry[i].setting - |= cks_setting << (j * 2); - } - CONVERT_FROM_HOST_TO_SMC_US( - data->smc_state_table.ClockStretcherDataTable.ClockStretcherDataTableEntry[i].setting); - } - value = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL); value &= 0xFFFFFFFE; cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL, value); @@ -1945,9 +1913,8 @@ static int polaris10_populate_vr_config(struct pp_hwmgr *hwmgr, if (POLARIS10_VOLTAGE_CONTROL_BY_SVID2 == data->mvdd_control) { config = VR_SVI2_PLANE_2; table->VRConfig |= (config << VRCONF_MVDD_SHIFT); - } else if (POLARIS10_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) { - config = VR_SMIO_PATTERN_2; - table->VRConfig |= (config << VRCONF_MVDD_SHIFT); + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, data->soft_regs_start + + offsetof(SMU74_SoftRegisters, AllowMvddSwitch), 0x1); } else { config = VR_STATIC_VOLTAGE; table->VRConfig |= (config << VRCONF_MVDD_SHIFT); @@ -1956,6 +1923,90 @@ static int polaris10_populate_vr_config(struct pp_hwmgr *hwmgr, return 0; } + +int polaris10_populate_avfs_parameters(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + SMU74_Discrete_DpmTable *table = &(data->smc_state_table); + int result = 0; + struct pp_atom_ctrl__avfs_parameters avfs_params = {0}; + AVFS_meanNsigma_t AVFS_meanNsigma = { {0} }; + AVFS_Sclk_Offset_t AVFS_SclkOffset = { {0} }; + uint32_t tmp, i; + struct pp_smumgr *smumgr = hwmgr->smumgr; + struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); + + struct phm_ppt_v1_information *table_info = + (struct phm_ppt_v1_information *)hwmgr->pptable; + struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table = + table_info->vdd_dep_on_sclk; + + + if (smu_data->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) + return result; + + result = atomctrl_get_avfs_information(hwmgr, &avfs_params); + + if (0 == result) { + table->BTCGB_VDROOP_TABLE[0].a0 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a0); + table->BTCGB_VDROOP_TABLE[0].a1 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a1); + table->BTCGB_VDROOP_TABLE[0].a2 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a2); + table->BTCGB_VDROOP_TABLE[1].a0 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a0); + table->BTCGB_VDROOP_TABLE[1].a1 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a1); + table->BTCGB_VDROOP_TABLE[1].a2 = PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a2); + table->AVFSGB_VDROOP_TABLE[0].m1 = PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_m1); + table->AVFSGB_VDROOP_TABLE[0].m2 = PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSON_m2); + table->AVFSGB_VDROOP_TABLE[0].b = PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_b); + table->AVFSGB_VDROOP_TABLE[0].m1_shift = 24; + table->AVFSGB_VDROOP_TABLE[0].m2_shift = 12; + table->AVFSGB_VDROOP_TABLE[1].m1 = PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_m1); + table->AVFSGB_VDROOP_TABLE[1].m2 = PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSOFF_m2); + table->AVFSGB_VDROOP_TABLE[1].b = PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_b); + table->AVFSGB_VDROOP_TABLE[1].m1_shift = 24; + table->AVFSGB_VDROOP_TABLE[1].m2_shift = 12; + table->MaxVoltage = PP_HOST_TO_SMC_US(avfs_params.usMaxVoltage_0_25mv); + AVFS_meanNsigma.Aconstant[0] = PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant0); + AVFS_meanNsigma.Aconstant[1] = PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant1); + AVFS_meanNsigma.Aconstant[2] = PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant2); + AVFS_meanNsigma.DC_tol_sigma = PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_DC_tol_sigma); + AVFS_meanNsigma.Platform_mean = PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_mean); + AVFS_meanNsigma.PSM_Age_CompFactor = PP_HOST_TO_SMC_US(avfs_params.usPSM_Age_ComFactor); + AVFS_meanNsigma.Platform_sigma = PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_sigma); + + for (i = 0; i < NUM_VFT_COLUMNS; i++) { + AVFS_meanNsigma.Static_Voltage_Offset[i] = (uint8_t)(sclk_table->entries[i].cks_voffset * 100 / 625); + AVFS_SclkOffset.Sclk_Offset[i] = PP_HOST_TO_SMC_US((uint16_t)(sclk_table->entries[i].sclk_offset) / 100); + } + + result = polaris10_read_smc_sram_dword(smumgr, + SMU7_FIRMWARE_HEADER_LOCATION + offsetof(SMU74_Firmware_Header, AvfsMeanNSigma), + &tmp, data->sram_end); + + polaris10_copy_bytes_to_smc(smumgr, + tmp, + (uint8_t *)&AVFS_meanNsigma, + sizeof(AVFS_meanNsigma_t), + data->sram_end); + + result = polaris10_read_smc_sram_dword(smumgr, + SMU7_FIRMWARE_HEADER_LOCATION + offsetof(SMU74_Firmware_Header, AvfsSclkOffsetTable), + &tmp, data->sram_end); + polaris10_copy_bytes_to_smc(smumgr, + tmp, + (uint8_t *)&AVFS_SclkOffset, + sizeof(AVFS_Sclk_Offset_t), + data->sram_end); + + data->avfs_vdroop_override_setting = (avfs_params.ucEnableGB_VDROOP_TABLE_CKSON << BTCGB0_Vdroop_Enable_SHIFT) | + (avfs_params.ucEnableGB_VDROOP_TABLE_CKSOFF << BTCGB1_Vdroop_Enable_SHIFT) | + (avfs_params.ucEnableGB_FUSE_TABLE_CKSON << AVFSGB0_Vdroop_Enable_SHIFT) | + (avfs_params.ucEnableGB_FUSE_TABLE_CKSOFF << AVFSGB1_Vdroop_Enable_SHIFT); + data->apply_avfs_cks_off_voltage = (avfs_params.ucEnableApplyAVFS_CKS_OFF_Voltage == 1) ? true : false; + } + return result; +} + + /** * Initializes the SMC table and uploads it * @@ -2056,6 +2107,10 @@ static int polaris10_init_smc_table(struct pp_hwmgr *hwmgr) "Failed to populate Clock Stretcher Data Table!", return result); } + + result = polaris10_populate_avfs_parameters(hwmgr); + PP_ASSERT_WITH_CODE(0 == result, "Failed to populate AVFS Parameters!", return result;); + table->CurrSclkPllRange = 0xff; table->GraphicsVoltageChangeEnable = 1; table->GraphicsThermThrottleEnable = 1; @@ -2229,6 +2284,17 @@ static int polaris10_enable_ulv(struct pp_hwmgr *hwmgr) return 0; } +static int polaris10_disable_ulv(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_ulv_parm *ulv = &(data->ulv); + + if (ulv->ulv_supported) + return smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_DisableULV); + + return 0; +} + static int polaris10_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) { if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, @@ -2249,9 +2315,27 @@ static int polaris10_enable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) return 0; } +static int polaris10_disable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr) +{ + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_SclkDeepSleep)) { + if (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_MASTER_DeepSleep_OFF)) { + PP_ASSERT_WITH_CODE(false, + "Attempt to disable Master Deep Sleep switch failed!", + return -1); + } + } + + return 0; +} + static int polaris10_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + uint32_t soft_register_value = 0; + uint32_t handshake_disables_offset = data->soft_regs_start + + offsetof(SMU74_SoftRegisters, HandshakeDisables); /* enable SCLK dpm */ if (!data->sclk_dpm_key_disabled) @@ -2262,6 +2346,12 @@ static int polaris10_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) /* enable MCLK dpm */ if (0 == data->mclk_dpm_key_disabled) { +/* Disable UVD - SMU handshake for MCLK. */ + soft_register_value = cgs_read_ind_register(hwmgr->device, + CGS_IND_REG__SMC, handshake_disables_offset); + soft_register_value |= SMU7_UVD_MCLK_HANDSHAKE_DISABLE; + cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, + handshake_disables_offset, soft_register_value); PP_ASSERT_WITH_CODE( (0 == smum_send_msg_to_smc(hwmgr->smumgr, @@ -2269,7 +2359,6 @@ static int polaris10_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) "Failed to enable MCLK DPM during DPM Start Function!", return -1); - PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1); cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x5); @@ -2338,6 +2427,58 @@ static int polaris10_start_dpm(struct pp_hwmgr *hwmgr) return 0; } +static int polaris10_disable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + + /* disable SCLK dpm */ + if (!data->sclk_dpm_key_disabled) + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_DPM_Disable) == 0), + "Failed to disable SCLK DPM!", + return -1); + + /* disable MCLK dpm */ + if (!data->mclk_dpm_key_disabled) { + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_MCLKDPM_Disable) == 0), + "Failed to disable MCLK DPM!", + return -1); + } + + return 0; +} + +static int polaris10_stop_dpm(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + + /* disable general power management */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, GENERAL_PWRMGT, + GLOBAL_PWRMGT_EN, 0); + /* disable sclk deep sleep */ + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SCLK_PWRMGT_CNTL, + DYNAMIC_PM_EN, 0); + + /* disable PCIE dpm */ + if (!data->pcie_dpm_key_disabled) { + PP_ASSERT_WITH_CODE( + (smum_send_msg_to_smc(hwmgr->smumgr, + PPSMC_MSG_PCIeDPM_Disable) == 0), + "Failed to disable pcie DPM during DPM Stop Function!", + return -1); + } + + if (polaris10_disable_sclk_mclk_dpm(hwmgr)) { + printk(KERN_ERR "Failed to disable Sclk DPM and Mclk DPM!"); + return -1; + } + + return 0; +} + static void polaris10_set_dpm_event_sources(struct pp_hwmgr *hwmgr, uint32_t sources) { bool protection; @@ -2395,6 +2536,23 @@ static int polaris10_enable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) return polaris10_enable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); } +static int polaris10_disable_auto_throttle_source(struct pp_hwmgr *hwmgr, + PHM_AutoThrottleSource source) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + + if (data->active_auto_throttle_sources & (1 << source)) { + data->active_auto_throttle_sources &= ~(1 << source); + polaris10_set_dpm_event_sources(hwmgr, data->active_auto_throttle_sources); + } + return 0; +} + +static int polaris10_disable_thermal_auto_throttle(struct pp_hwmgr *hwmgr) +{ + return polaris10_disable_auto_throttle_source(hwmgr, PHM_AutoThrottleSource_Thermal); +} + int polaris10_pcie_performance_request(struct pp_hwmgr *hwmgr) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -2516,8 +2674,60 @@ int polaris10_enable_dpm_tasks(struct pp_hwmgr *hwmgr) int polaris10_disable_dpm_tasks(struct pp_hwmgr *hwmgr) { + int tmp_result, result = 0; - return 0; + tmp_result = (polaris10_is_dpm_running(hwmgr)) ? 0 : -1; + PP_ASSERT_WITH_CODE(tmp_result == 0, + "DPM is not running right now, no need to disable DPM!", + return 0); + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_ThermalController)) + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + GENERAL_PWRMGT, THERMAL_PROTECTION_DIS, 1); + + tmp_result = polaris10_disable_power_containment(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable power containment!", result = tmp_result); + + tmp_result = polaris10_disable_smc_cac(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable SMC CAC!", result = tmp_result); + + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + CG_SPLL_SPREAD_SPECTRUM, SSEN, 0); + PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, + GENERAL_PWRMGT, DYN_SPREAD_SPECTRUM_EN, 0); + + tmp_result = polaris10_disable_thermal_auto_throttle(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable thermal auto throttle!", result = tmp_result); + + tmp_result = polaris10_stop_dpm(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to stop DPM!", result = tmp_result); + + tmp_result = polaris10_disable_deep_sleep_master_switch(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable deep sleep master switch!", result = tmp_result); + + tmp_result = polaris10_disable_ulv(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to disable ULV!", result = tmp_result); + + tmp_result = polaris10_clear_voting_clients(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to clear voting clients!", result = tmp_result); + + tmp_result = polaris10_reset_to_default(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to reset to default!", result = tmp_result); + + tmp_result = polaris10_force_switch_to_arbf0(hwmgr); + PP_ASSERT_WITH_CODE((tmp_result == 0), + "Failed to force to switch arbf0!", result = tmp_result); + + return result; } int polaris10_reset_asic_tasks(struct pp_hwmgr *hwmgr) @@ -2528,13 +2738,6 @@ int polaris10_reset_asic_tasks(struct pp_hwmgr *hwmgr) int polaris10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr) { - struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); - - if (data->soft_pp_table) { - kfree(data->soft_pp_table); - data->soft_pp_table = NULL; - } - return phm_hwmgr_backend_fini(hwmgr); } @@ -2590,8 +2793,13 @@ int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr) phm_cap_unset(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_TCPRamping); - phm_cap_set(hwmgr->platform_descriptor.platformCaps, - PHM_PlatformCaps_PowerContainment); + if (hwmgr->powercontainment_enabled) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment); + else + phm_cap_unset(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment); + phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_CAC); @@ -2606,6 +2814,7 @@ int polaris10_set_features_platform_caps(struct pp_hwmgr *hwmgr) phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_FanSpeedInTableIsRPM); + if (hwmgr->chip_id == CHIP_POLARIS11) phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SPLLShutdownSupport); @@ -2662,12 +2871,12 @@ static int polaris10_get_evv_voltages(struct pp_hwmgr *hwmgr) } } - - PP_ASSERT_WITH_CODE(0 == atomctrl_get_voltage_evv_on_sclk_ai(hwmgr, - VOLTAGE_TYPE_VDDC, sclk, vv_id, &vddc), - "Error retrieving EVV voltage value!", - continue); - + if (atomctrl_get_voltage_evv_on_sclk_ai(hwmgr, + VOLTAGE_TYPE_VDDC, + sclk, vv_id, &vddc) != 0) { + printk(KERN_WARNING "failed to retrieving EVV voltage!\n"); + continue; + } /* need to make sure vddc is less than 2v or else, it could burn the ASIC. */ PP_ASSERT_WITH_CODE((vddc < 2000 && vddc != 0), @@ -2898,13 +3107,19 @@ static int polaris10_set_private_data_based_on_pptable(struct pp_hwmgr *hwmgr) int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) { - struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_hwmgr *data; struct pp_atomctrl_gpio_pin_assignment gpio_pin_assignment; uint32_t temp_reg; int result; struct phm_ppt_v1_information *table_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); + data = kzalloc(sizeof(struct polaris10_hwmgr), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + + hwmgr->backend = data; + data->dll_default_on = false; data->sram_end = SMC_RAM_END; data->mclk_dpm0_activity_target = 0xa; @@ -2938,6 +3153,11 @@ int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) data->vddci_control = POLARIS10_VOLTAGE_CONTROL_NONE; data->mvdd_control = POLARIS10_VOLTAGE_CONTROL_NONE; + data->enable_tdc_limit_feature = true; + data->enable_pkg_pwr_tracking_feature = true; + data->force_pcie_gen = PP_PCIEGenInvalid; + data->mclk_stutter_mode_threshold = 40000; + if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr, VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_SVID2)) data->voltage_control = POLARIS10_VOLTAGE_CONTROL_BY_SVID2; @@ -2962,6 +3182,10 @@ int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) data->vddci_control = POLARIS10_VOLTAGE_CONTROL_BY_SVID2; } + if (table_info->cac_dtp_table->usClockStretchAmount != 0) + phm_cap_set(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_ClockStretcher); + polaris10_set_features_platform_caps(hwmgr); polaris10_init_dpm_defaults(hwmgr); @@ -3068,7 +3292,7 @@ int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_GEN_INFO; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_gen_cap = 0x30007; + data->pcie_gen_cap = AMDGPU_DEFAULT_PCIE_GEN_MASK; else data->pcie_gen_cap = (uint32_t)sys_info.value; if (data->pcie_gen_cap & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) @@ -3077,7 +3301,7 @@ int polaris10_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_MLW; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_lane_cap = 0x2f0000; + data->pcie_lane_cap = AMDGPU_DEFAULT_PCIE_MLW_MASK; else data->pcie_lane_cap = (uint32_t)sys_info.value; @@ -3520,10 +3744,11 @@ static int polaris10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, ATOM_Tonga_State *state_entry = (ATOM_Tonga_State *)state; ATOM_Tonga_POWERPLAYTABLE *powerplay_table = (ATOM_Tonga_POWERPLAYTABLE *)pp_table; - ATOM_Tonga_SCLK_Dependency_Table *sclk_dep_table = - (ATOM_Tonga_SCLK_Dependency_Table *) + PPTable_Generic_SubTable_Header *sclk_dep_table = + (PPTable_Generic_SubTable_Header *) (((unsigned long)powerplay_table) + le16_to_cpu(powerplay_table->usSclkDependencyTableOffset)); + ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table = (ATOM_Tonga_MCLK_Dependency_Table *) (((unsigned long)powerplay_table) + @@ -3575,7 +3800,11 @@ static int polaris10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, /* Performance levels are arranged from low to high. */ performance_level->memory_clock = mclk_dep_table->entries [state_entry->ucMemoryClockIndexLow].ulMclk; - performance_level->engine_clock = sclk_dep_table->entries + if (sclk_dep_table->ucRevId == 0) + performance_level->engine_clock = ((ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table)->entries + [state_entry->ucEngineClockIndexLow].ulSclk; + else if (sclk_dep_table->ucRevId == 1) + performance_level->engine_clock = ((ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table)->entries [state_entry->ucEngineClockIndexLow].ulSclk; performance_level->pcie_gen = get_pcie_gen_support(data->pcie_gen_cap, state_entry->ucPCIEGenLow); @@ -3586,8 +3815,14 @@ static int polaris10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, [polaris10_power_state->performance_level_count++]); performance_level->memory_clock = mclk_dep_table->entries [state_entry->ucMemoryClockIndexHigh].ulMclk; - performance_level->engine_clock = sclk_dep_table->entries + + if (sclk_dep_table->ucRevId == 0) + performance_level->engine_clock = ((ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table)->entries + [state_entry->ucEngineClockIndexHigh].ulSclk; + else if (sclk_dep_table->ucRevId == 1) + performance_level->engine_clock = ((ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table)->entries [state_entry->ucEngineClockIndexHigh].ulSclk; + performance_level->pcie_gen = get_pcie_gen_support(data->pcie_gen_cap, state_entry->ucPCIEGenHigh); performance_level->pcie_lane = get_pcie_lane_support(data->pcie_lane_cap, @@ -3645,7 +3880,6 @@ static int polaris10_get_pp_table_entry(struct pp_hwmgr *hwmgr, switch (state->classification.ui_label) { case PP_StateUILabel_Performance: data->use_pcie_performance_levels = true; - for (i = 0; i < ps->performance_level_count; i++) { if (data->pcie_gen_performance.max < ps->performance_levels[i].pcie_gen) @@ -3661,7 +3895,6 @@ static int polaris10_get_pp_table_entry(struct pp_hwmgr *hwmgr, ps->performance_levels[i].pcie_lane) data->pcie_lane_performance.max = ps->performance_levels[i].pcie_lane; - if (data->pcie_lane_performance.min > ps->performance_levels[i].pcie_lane) data->pcie_lane_performance.min = @@ -4187,12 +4420,9 @@ int polaris10_update_samu_dpm(struct pp_hwmgr *hwmgr, bool bgate) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); uint32_t mm_boot_level_offset, mm_boot_level_value; - struct phm_ppt_v1_information *table_info = - (struct phm_ppt_v1_information *)(hwmgr->pptable); if (!bgate) { - data->smc_state_table.SamuBootLevel = - (uint8_t) (table_info->mm_dep_table->count - 1); + data->smc_state_table.SamuBootLevel = 0; mm_boot_level_offset = data->dpm_table_start + offsetof(SMU74_Discrete_DpmTable, SamuBootLevel); mm_boot_level_offset /= 4; @@ -4721,42 +4951,6 @@ int polaris10_setup_asic_task(struct pp_hwmgr *hwmgr) return result; } -static int polaris10_get_pp_table(struct pp_hwmgr *hwmgr, char **table) -{ - struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kmemdup(hwmgr->soft_pp_table, - hwmgr->soft_pp_table_size, - GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - *table = (char *)&data->soft_pp_table; - - return hwmgr->soft_pp_table_size; -} - -static int polaris10_set_pp_table(struct pp_hwmgr *hwmgr, const char *buf, size_t size) -{ - struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - memcpy(data->soft_pp_table, buf, size); - - hwmgr->soft_pp_table = data->soft_pp_table; - - /* TODO: re-init powerplay to implement modified pptable */ - - return 0; -} - static int polaris10_force_clock_level(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask) { @@ -4899,6 +5093,89 @@ static int polaris10_get_fan_control_mode(struct pp_hwmgr *hwmgr) CG_FDO_CTRL2, FDO_PWM_MODE); } +static int polaris10_get_sclk_od(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_single_dpm_table *sclk_table = &(data->dpm_table.sclk_table); + struct polaris10_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + int value; + + value = (sclk_table->dpm_levels[sclk_table->count - 1].value - + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value) * + 100 / + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return value; +} + +static int polaris10_set_sclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + struct pp_power_state *ps; + struct polaris10_power_state *polaris10_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + polaris10_ps = cast_phw_polaris10_power_state(&ps->hardware); + + polaris10_ps->performance_levels[polaris10_ps->performance_level_count - 1].engine_clock = + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value * + value / 100 + + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return 0; +} + +static int polaris10_get_mclk_od(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_single_dpm_table *mclk_table = &(data->dpm_table.mclk_table); + struct polaris10_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + int value; + + value = (mclk_table->dpm_levels[mclk_table->count - 1].value - + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value) * + 100 / + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return value; +} + +static int polaris10_set_mclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + struct polaris10_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + struct pp_power_state *ps; + struct polaris10_power_state *polaris10_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + polaris10_ps = cast_phw_polaris10_power_state(&ps->hardware); + + polaris10_ps->performance_levels[polaris10_ps->performance_level_count - 1].memory_clock = + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value * + value / 100 + + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return 0; +} static const struct pp_hwmgr_func polaris10_hwmgr_funcs = { .backend_init = &polaris10_hwmgr_backend_init, .backend_fini = &polaris10_hwmgr_backend_fini, @@ -4937,22 +5214,17 @@ static const struct pp_hwmgr_func polaris10_hwmgr_funcs = { .check_states_equal = polaris10_check_states_equal, .set_fan_control_mode = polaris10_set_fan_control_mode, .get_fan_control_mode = polaris10_get_fan_control_mode, - .get_pp_table = polaris10_get_pp_table, - .set_pp_table = polaris10_set_pp_table, .force_clock_level = polaris10_force_clock_level, .print_clock_levels = polaris10_print_clock_levels, .enable_per_cu_power_gating = polaris10_phm_enable_per_cu_power_gating, + .get_sclk_od = polaris10_get_sclk_od, + .set_sclk_od = polaris10_set_sclk_od, + .get_mclk_od = polaris10_get_mclk_od, + .set_mclk_od = polaris10_set_mclk_od, }; int polaris10_hwmgr_init(struct pp_hwmgr *hwmgr) { - struct polaris10_hwmgr *data; - - data = kzalloc (sizeof(struct polaris10_hwmgr), GFP_KERNEL); - if (data == NULL) - return -ENOMEM; - - hwmgr->backend = data; hwmgr->hwmgr_func = &polaris10_hwmgr_funcs; hwmgr->pptable_func = &tonga_pptable_funcs; pp_polaris10_thermal_initialize(hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h index beedf35cbfa6..fd38b0d7a3c2 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_hwmgr.h @@ -309,9 +309,8 @@ struct polaris10_hwmgr { uint32_t up_hyst; uint32_t disable_dpm_mask; bool apply_optimized_settings; - - /* soft pptable for re-uploading into smu */ - void *soft_pp_table; + uint32_t avfs_vdroop_override_setting; + bool apply_avfs_cks_off_voltage; }; /* To convert to Q8.8 format for firmware */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.c index 0b99ab3ba0c5..5620e268b553 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.c @@ -286,7 +286,7 @@ int polaris10_populate_pm_fuses(struct pp_hwmgr *hwmgr) if (polaris10_copy_bytes_to_smc(hwmgr->smumgr, pm_fuse_table_offset, (uint8_t *)&data->power_tune_table, - sizeof(struct SMU74_Discrete_PmFuses), data->sram_end)) + (sizeof(struct SMU74_Discrete_PmFuses) - 92), data->sram_end)) PP_ASSERT_WITH_CODE(false, "Attempt to download PmFuseTable Failed!", return -EINVAL); @@ -312,6 +312,23 @@ int polaris10_enable_smc_cac(struct pp_hwmgr *hwmgr) return result; } +int polaris10_disable_smc_cac(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + int result = 0; + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_CAC) && data->cac_enabled) { + int smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_DisableCac)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable CAC in SMC.", result = -1); + + data->cac_enabled = false; + } + return result; +} + int polaris10_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n) { struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); @@ -373,6 +390,48 @@ int polaris10_enable_power_containment(struct pp_hwmgr *hwmgr) return result; } +int polaris10_disable_power_containment(struct pp_hwmgr *hwmgr) +{ + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); + int result = 0; + + if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, + PHM_PlatformCaps_PowerContainment) && + data->power_containment_features) { + int smc_result; + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_TDCLimit) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_TDCLimitDisable)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable TDCLimit in SMC.", + result = smc_result); + } + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_DTE) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_DisableDTE)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable DTE in SMC.", + result = smc_result); + } + + if (data->power_containment_features & + POWERCONTAINMENT_FEATURE_PkgPwrLimit) { + smc_result = smum_send_msg_to_smc(hwmgr->smumgr, + (uint16_t)(PPSMC_MSG_PkgPwrLimitDisable)); + PP_ASSERT_WITH_CODE((smc_result == 0), + "Failed to disable PkgPwrTracking in SMC.", + result = smc_result); + } + data->power_containment_features = 0; + } + + return result; +} + int polaris10_power_control_set_level(struct pp_hwmgr *hwmgr) { struct phm_ppt_v1_information *table_info = diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h index 68bc1cb6d40c..d492d6d28867 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_powertune.h @@ -31,6 +31,19 @@ enum polaris10_pt_config_reg_type { POLARIS10_CONFIGREG_MAX }; +#define DIDT_SQ_CTRL0__UNUSED_0_MASK 0xfffc0000 +#define DIDT_SQ_CTRL0__UNUSED_0__SHIFT 0x12 +#define DIDT_TD_CTRL0__UNUSED_0_MASK 0xfffc0000 +#define DIDT_TD_CTRL0__UNUSED_0__SHIFT 0x12 +#define DIDT_TCP_CTRL0__UNUSED_0_MASK 0xfffc0000 +#define DIDT_TCP_CTRL0__UNUSED_0__SHIFT 0x12 +#define DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK 0xc0000000 +#define DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001e +#define DIDT_TD_TUNING_CTRL__UNUSED_0_MASK 0xc0000000 +#define DIDT_TD_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001e +#define DIDT_TCP_TUNING_CTRL__UNUSED_0_MASK 0xc0000000 +#define DIDT_TCP_TUNING_CTRL__UNUSED_0__SHIFT 0x0000001e + /* PowerContainment Features */ #define POWERCONTAINMENT_FEATURE_DTE 0x00000001 #define POWERCONTAINMENT_FEATURE_TDCLimit 0x00000002 @@ -62,7 +75,9 @@ void polaris10_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr); int polaris10_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr); int polaris10_populate_pm_fuses(struct pp_hwmgr *hwmgr); int polaris10_enable_smc_cac(struct pp_hwmgr *hwmgr); +int polaris10_disable_smc_cac(struct pp_hwmgr *hwmgr); int polaris10_enable_power_containment(struct pp_hwmgr *hwmgr); +int polaris10_disable_power_containment(struct pp_hwmgr *hwmgr); int polaris10_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n); int polaris10_power_control_set_level(struct pp_hwmgr *hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c index aba167f7d167..b206632d4650 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/polaris10_thermal.c @@ -625,10 +625,14 @@ static int tf_polaris10_thermal_avfs_enable(struct pp_hwmgr *hwmgr, int ret; struct pp_smumgr *smumgr = (struct pp_smumgr *)(hwmgr->smumgr); struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(smumgr->backend); + struct polaris10_hwmgr *data = (struct polaris10_hwmgr *)(hwmgr->backend); - if (smu_data->avfs.avfs_btc_status != AVFS_BTC_ENABLEAVFS) + if (smu_data->avfs.avfs_btc_status == AVFS_BTC_NOTSUPPORTED) return 0; + ret = smum_send_msg_to_smc_with_parameter(hwmgr->smumgr, + PPSMC_MSG_SetGBDroopSettings, data->avfs_vdroop_override_setting); + ret = (smum_send_msg_to_smc(smumgr, PPSMC_MSG_EnableAvfs) == 0) ? 0 : -1; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c index 58742e0d1492..a3c38bbd1e94 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_acpi.c @@ -44,6 +44,20 @@ bool acpi_atcs_functions_supported(void *device, uint32_t index) return result == 0 ? (output_buf.function_bits & (1 << (index - 1))) != 0 : false; } +bool acpi_atcs_notify_pcie_device_ready(void *device) +{ + int32_t temp_buffer = 1; + + return cgs_call_acpi_method(device, CGS_ACPI_METHOD_ATCS, + ATCS_FUNCTION_PCIE_DEVICE_READY_NOTIFICATION, + &temp_buffer, + NULL, + 0, + sizeof(temp_buffer), + 0); +} + + int acpi_pcie_perf_request(void *device, uint8_t perf_req, bool advertise) { struct atcs_pref_req_input atcs_input; @@ -52,7 +66,7 @@ int acpi_pcie_perf_request(void *device, uint8_t perf_req, bool advertise) int result; struct cgs_system_info info = {0}; - if (!acpi_atcs_functions_supported(device, ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST)) + if( 0 != acpi_atcs_notify_pcie_device_ready(device)) return -EINVAL; info.size = sizeof(struct cgs_system_info); @@ -77,7 +91,7 @@ int acpi_pcie_perf_request(void *device, uint8_t perf_req, bool advertise) ATCS_FUNCTION_PCIE_PERFORMANCE_REQUEST, &atcs_input, &atcs_output, - 0, + 1, sizeof(atcs_input), sizeof(atcs_output)); if (result != 0) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c index da9f5f1b6dc2..5d70e2c47faf 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c @@ -552,13 +552,13 @@ static bool atomctrl_lookup_gpio_pin( pin_assignment->ucGpioPinBitShift; gpio_pin_assignment->us_gpio_pin_aindex = le16_to_cpu(pin_assignment->usGpioPin_AIndex); - return false; + return true; } offset += offsetof(ATOM_GPIO_PIN_ASSIGNMENT, ucGPIO_ID) + 1; } - return true; + return false; } /** @@ -1302,3 +1302,46 @@ int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctr return 0; } + +int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__avfs_parameters *param) +{ + ATOM_ASIC_PROFILING_INFO_V3_6 *profile = NULL; + + if (param == NULL) + return -EINVAL; + + profile = (ATOM_ASIC_PROFILING_INFO_V3_6 *) + cgs_atom_get_data_table(hwmgr->device, + GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo), + NULL, NULL, NULL); + if (!profile) + return -1; + + param->ulAVFS_meanNsigma_Acontant0 = profile->ulAVFS_meanNsigma_Acontant0; + param->ulAVFS_meanNsigma_Acontant1 = profile->ulAVFS_meanNsigma_Acontant1; + param->ulAVFS_meanNsigma_Acontant2 = profile->ulAVFS_meanNsigma_Acontant2; + param->usAVFS_meanNsigma_DC_tol_sigma = profile->usAVFS_meanNsigma_DC_tol_sigma; + param->usAVFS_meanNsigma_Platform_mean = profile->usAVFS_meanNsigma_Platform_mean; + param->usAVFS_meanNsigma_Platform_sigma = profile->usAVFS_meanNsigma_Platform_sigma; + param->ulGB_VDROOP_TABLE_CKSOFF_a0 = profile->ulGB_VDROOP_TABLE_CKSOFF_a0; + param->ulGB_VDROOP_TABLE_CKSOFF_a1 = profile->ulGB_VDROOP_TABLE_CKSOFF_a1; + param->ulGB_VDROOP_TABLE_CKSOFF_a2 = profile->ulGB_VDROOP_TABLE_CKSOFF_a2; + param->ulGB_VDROOP_TABLE_CKSON_a0 = profile->ulGB_VDROOP_TABLE_CKSON_a0; + param->ulGB_VDROOP_TABLE_CKSON_a1 = profile->ulGB_VDROOP_TABLE_CKSON_a1; + param->ulGB_VDROOP_TABLE_CKSON_a2 = profile->ulGB_VDROOP_TABLE_CKSON_a2; + param->ulAVFSGB_FUSE_TABLE_CKSOFF_m1 = profile->ulAVFSGB_FUSE_TABLE_CKSOFF_m1; + param->usAVFSGB_FUSE_TABLE_CKSOFF_m2 = profile->usAVFSGB_FUSE_TABLE_CKSOFF_m2; + param->ulAVFSGB_FUSE_TABLE_CKSOFF_b = profile->ulAVFSGB_FUSE_TABLE_CKSOFF_b; + param->ulAVFSGB_FUSE_TABLE_CKSON_m1 = profile->ulAVFSGB_FUSE_TABLE_CKSON_m1; + param->usAVFSGB_FUSE_TABLE_CKSON_m2 = profile->usAVFSGB_FUSE_TABLE_CKSON_m2; + param->ulAVFSGB_FUSE_TABLE_CKSON_b = profile->ulAVFSGB_FUSE_TABLE_CKSON_b; + param->usMaxVoltage_0_25mv = profile->usMaxVoltage_0_25mv; + param->ucEnableGB_VDROOP_TABLE_CKSOFF = profile->ucEnableGB_VDROOP_TABLE_CKSOFF; + param->ucEnableGB_VDROOP_TABLE_CKSON = profile->ucEnableGB_VDROOP_TABLE_CKSON; + param->ucEnableGB_FUSE_TABLE_CKSOFF = profile->ucEnableGB_FUSE_TABLE_CKSOFF; + param->ucEnableGB_FUSE_TABLE_CKSON = profile->ucEnableGB_FUSE_TABLE_CKSON; + param->usPSM_Age_ComFactor = profile->usPSM_Age_ComFactor; + param->ucEnableApplyAVFS_CKS_OFF_Voltage = profile->ucEnableApplyAVFS_CKS_OFF_Voltage; + + return 0; +} diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h index d24ebb566905..248c5db5f380 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h @@ -250,6 +250,35 @@ struct pp_atomctrl_gpio_pin_assignment { }; typedef struct pp_atomctrl_gpio_pin_assignment pp_atomctrl_gpio_pin_assignment; +struct pp_atom_ctrl__avfs_parameters { + uint32_t ulAVFS_meanNsigma_Acontant0; + uint32_t ulAVFS_meanNsigma_Acontant1; + uint32_t ulAVFS_meanNsigma_Acontant2; + uint16_t usAVFS_meanNsigma_DC_tol_sigma; + uint16_t usAVFS_meanNsigma_Platform_mean; + uint16_t usAVFS_meanNsigma_Platform_sigma; + uint32_t ulGB_VDROOP_TABLE_CKSOFF_a0; + uint32_t ulGB_VDROOP_TABLE_CKSOFF_a1; + uint32_t ulGB_VDROOP_TABLE_CKSOFF_a2; + uint32_t ulGB_VDROOP_TABLE_CKSON_a0; + uint32_t ulGB_VDROOP_TABLE_CKSON_a1; + uint32_t ulGB_VDROOP_TABLE_CKSON_a2; + uint32_t ulAVFSGB_FUSE_TABLE_CKSOFF_m1; + uint16_t usAVFSGB_FUSE_TABLE_CKSOFF_m2; + uint32_t ulAVFSGB_FUSE_TABLE_CKSOFF_b; + uint32_t ulAVFSGB_FUSE_TABLE_CKSON_m1; + uint16_t usAVFSGB_FUSE_TABLE_CKSON_m2; + uint32_t ulAVFSGB_FUSE_TABLE_CKSON_b; + uint16_t usMaxVoltage_0_25mv; + uint8_t ucEnableGB_VDROOP_TABLE_CKSOFF; + uint8_t ucEnableGB_VDROOP_TABLE_CKSON; + uint8_t ucEnableGB_FUSE_TABLE_CKSOFF; + uint8_t ucEnableGB_FUSE_TABLE_CKSON; + uint16_t usPSM_Age_ComFactor; + uint8_t ucEnableApplyAVFS_CKS_OFF_Voltage; + uint8_t ucReserved; +}; + extern bool atomctrl_get_pp_assign_pin(struct pp_hwmgr *hwmgr, const uint32_t pinId, pp_atomctrl_gpio_pin_assignment *gpio_pin_assignment); extern int atomctrl_get_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type, uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage); extern uint32_t atomctrl_get_mpll_reference_clock(struct pp_hwmgr *hwmgr); @@ -278,5 +307,8 @@ extern int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clo extern int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_type, uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage); extern int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl_sclk_range_table *table); + +extern int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr, struct pp_atom_ctrl__avfs_parameters *param); + #endif diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c index 2f1a14fe05b1..35bc8a29b773 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c @@ -794,15 +794,18 @@ static const ATOM_PPLIB_STATE_V2 *get_state_entry_v2( static const ATOM_PPLIB_POWERPLAYTABLE *get_powerplay_table( struct pp_hwmgr *hwmgr) { - const void *table_addr = NULL; + const void *table_addr = hwmgr->soft_pp_table; uint8_t frev, crev; uint16_t size; - table_addr = cgs_atom_get_data_table(hwmgr->device, - GetIndexIntoMasterTable(DATA, PowerPlayInfo), - &size, &frev, &crev); + if (!table_addr) { + table_addr = cgs_atom_get_data_table(hwmgr->device, + GetIndexIntoMasterTable(DATA, PowerPlayInfo), + &size, &frev, &crev); - hwmgr->soft_pp_table = table_addr; + hwmgr->soft_pp_table = table_addr; + hwmgr->soft_pp_table_size = size; + } return (const ATOM_PPLIB_POWERPLAYTABLE *)table_addr; } @@ -1499,7 +1502,7 @@ int get_number_of_vce_state_table_entries( const ATOM_PPLIB_VCE_State_Table *vce_table = get_vce_state_table(hwmgr, table); - if (vce_table > 0) + if (vce_table) return vce_table->numEntries; return 0; @@ -1589,11 +1592,6 @@ static int pp_tables_initialize(struct pp_hwmgr *hwmgr) static int pp_tables_uninitialize(struct pp_hwmgr *hwmgr) { - if (NULL != hwmgr->soft_pp_table) { - kfree(hwmgr->soft_pp_table); - hwmgr->soft_pp_table = NULL; - } - if (NULL != hwmgr->dyn_state.vddc_dependency_on_sclk) { kfree(hwmgr->dyn_state.vddc_dependency_on_sclk); hwmgr->dyn_state.vddc_dependency_on_sclk = NULL; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c index 16fed487973b..6c4553cf8023 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.c @@ -2847,27 +2847,6 @@ static int tonga_setup_default_dpm_tables(struct pp_hwmgr *hwmgr) } } - /* Initialize Vddc DPM table based on allow Vddc values. And populate corresponding std values. */ - for (i = 0; i < allowed_vdd_sclk_table->count; i++) { - data->dpm_table.vddc_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].vddc; - /* tonga_hwmgr->dpm_table.VddcTable.dpm_levels[i].param1 = stdVoltageTable->entries[i].Leakage; */ - /* param1 is for corresponding std voltage */ - data->dpm_table.vddc_table.dpm_levels[i].enabled = 1; - } - data->dpm_table.vddc_table.count = allowed_vdd_sclk_table->count; - - if (NULL != allowed_vdd_mclk_table) { - /* Initialize Vddci DPM table based on allow Mclk values */ - for (i = 0; i < allowed_vdd_mclk_table->count; i++) { - data->dpm_table.vdd_ci_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].vddci; - data->dpm_table.vdd_ci_table.dpm_levels[i].enabled = 1; - data->dpm_table.mvdd_table.dpm_levels[i].value = allowed_vdd_mclk_table->entries[i].mvdd; - data->dpm_table.mvdd_table.dpm_levels[i].enabled = 1; - } - data->dpm_table.vdd_ci_table.count = allowed_vdd_mclk_table->count; - data->dpm_table.mvdd_table.count = allowed_vdd_mclk_table->count; - } - /* setup PCIE gen speed levels*/ tonga_setup_default_pcie_tables(hwmgr); @@ -3047,8 +3026,8 @@ int tonga_init_smc_table(struct pp_hwmgr *hwmgr) reg_value = 0; if ((0 == reg_value) && - (0 == atomctrl_get_pp_assign_pin(hwmgr, - VDDC_VRHOT_GPIO_PINID, &gpio_pin_assignment))) { + (atomctrl_get_pp_assign_pin(hwmgr, VDDC_VRHOT_GPIO_PINID, + &gpio_pin_assignment))) { table->VRHotGpio = gpio_pin_assignment.uc_gpio_pin_bit_shift; phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_RegulatorHot); @@ -3061,8 +3040,8 @@ int tonga_init_smc_table(struct pp_hwmgr *hwmgr) /* ACDC Switch GPIO */ reg_value = 0; if ((0 == reg_value) && - (0 == atomctrl_get_pp_assign_pin(hwmgr, - PP_AC_DC_SWITCH_GPIO_PINID, &gpio_pin_assignment))) { + (atomctrl_get_pp_assign_pin(hwmgr, PP_AC_DC_SWITCH_GPIO_PINID, + &gpio_pin_assignment))) { table->AcDcGpio = gpio_pin_assignment.uc_gpio_pin_bit_shift; phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_AutomaticDCTransition); @@ -3084,8 +3063,7 @@ int tonga_init_smc_table(struct pp_hwmgr *hwmgr) } reg_value = 0; - if ((0 == reg_value) && - (0 == atomctrl_get_pp_assign_pin(hwmgr, + if ((0 == reg_value) && (atomctrl_get_pp_assign_pin(hwmgr, THERMAL_INT_OUTPUT_GPIO_PINID, &gpio_pin_assignment))) { phm_cap_set(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_ThermalOutGPIO); @@ -4443,13 +4421,6 @@ int tonga_reset_asic_tasks(struct pp_hwmgr *hwmgr) int tonga_hwmgr_backend_fini(struct pp_hwmgr *hwmgr) { - struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); - - if (data->soft_pp_table) { - kfree(data->soft_pp_table); - data->soft_pp_table = NULL; - } - return phm_hwmgr_backend_fini(hwmgr); } @@ -4463,7 +4434,7 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) { int result = 0; SMU72_Discrete_DpmTable *table = NULL; - tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + tonga_hwmgr *data; pp_atomctrl_gpio_pin_assignment gpio_pin_assignment; struct phm_ppt_v1_information *pptable_info = (struct phm_ppt_v1_information *)(hwmgr->pptable); phw_tonga_ulv_parm *ulv; @@ -4472,6 +4443,12 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((NULL != hwmgr), "Invalid Parameter!", return -1;); + data = kzalloc(sizeof(struct tonga_hwmgr), GFP_KERNEL); + if (data == NULL) + return -ENOMEM; + + hwmgr->backend = data; + data->dll_defaule_on = 0; data->sram_end = SMC_RAM_END; @@ -4510,6 +4487,7 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) data->vdd_ci_control = TONGA_VOLTAGE_CONTROL_NONE; data->vdd_gfx_control = TONGA_VOLTAGE_CONTROL_NONE; data->mvdd_control = TONGA_VOLTAGE_CONTROL_NONE; + data->force_pcie_gen = PP_PCIEGenInvalid; if (atomctrl_is_voltage_controled_by_gpio_v3(hwmgr, VOLTAGE_TYPE_VDDC, VOLTAGE_OBJ_SVID2)) { @@ -4591,7 +4569,7 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) * if ucGPIO_ID=VDDC_PCC_GPIO_PINID in GPIO_LUTable, * Peak Current Control feature is enabled and we should program PCC HW register */ - if (0 == atomctrl_get_pp_assign_pin(hwmgr, VDDC_PCC_GPIO_PINID, &gpio_pin_assignment)) { + if (atomctrl_get_pp_assign_pin(hwmgr, VDDC_PCC_GPIO_PINID, &gpio_pin_assignment)) { uint32_t temp_reg = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCNB_PWRMGT_CNTL); @@ -4659,7 +4637,7 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_GEN_INFO; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_gen_cap = 0x30007; + data->pcie_gen_cap = AMDGPU_DEFAULT_PCIE_GEN_MASK; else data->pcie_gen_cap = (uint32_t)sys_info.value; if (data->pcie_gen_cap & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) @@ -4668,7 +4646,7 @@ int tonga_hwmgr_backend_init(struct pp_hwmgr *hwmgr) sys_info.info_id = CGS_SYSTEM_INFO_PCIE_MLW; result = cgs_query_system_info(hwmgr->device, &sys_info); if (result) - data->pcie_lane_cap = 0x2f0000; + data->pcie_lane_cap = AMDGPU_DEFAULT_PCIE_MLW_MASK; else data->pcie_lane_cap = (uint32_t)sys_info.value; } else { @@ -6051,42 +6029,6 @@ static int tonga_get_fan_control_mode(struct pp_hwmgr *hwmgr) CG_FDO_CTRL2, FDO_PWM_MODE); } -static int tonga_get_pp_table(struct pp_hwmgr *hwmgr, char **table) -{ - struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kmemdup(hwmgr->soft_pp_table, - hwmgr->soft_pp_table_size, - GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - *table = (char *)&data->soft_pp_table; - - return hwmgr->soft_pp_table_size; -} - -static int tonga_set_pp_table(struct pp_hwmgr *hwmgr, const char *buf, size_t size) -{ - struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); - - if (!data->soft_pp_table) { - data->soft_pp_table = kzalloc(hwmgr->soft_pp_table_size, GFP_KERNEL); - if (!data->soft_pp_table) - return -ENOMEM; - } - - memcpy(data->soft_pp_table, buf, size); - - hwmgr->soft_pp_table = data->soft_pp_table; - - /* TODO: re-init powerplay to implement modified pptable */ - - return 0; -} - static int tonga_force_clock_level(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask) { @@ -6194,11 +6136,96 @@ static int tonga_print_clock_levels(struct pp_hwmgr *hwmgr, return size; } +static int tonga_get_sclk_od(struct pp_hwmgr *hwmgr) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + struct tonga_single_dpm_table *sclk_table = &(data->dpm_table.sclk_table); + struct tonga_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + int value; + + value = (sclk_table->dpm_levels[sclk_table->count - 1].value - + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value) * + 100 / + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return value; +} + +static int tonga_set_sclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + struct tonga_single_dpm_table *golden_sclk_table = + &(data->golden_dpm_table.sclk_table); + struct pp_power_state *ps; + struct tonga_power_state *tonga_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + tonga_ps = cast_phw_tonga_power_state(&ps->hardware); + + tonga_ps->performance_levels[tonga_ps->performance_level_count - 1].engine_clock = + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value * + value / 100 + + golden_sclk_table->dpm_levels[golden_sclk_table->count - 1].value; + + return 0; +} + +static int tonga_get_mclk_od(struct pp_hwmgr *hwmgr) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + struct tonga_single_dpm_table *mclk_table = &(data->dpm_table.mclk_table); + struct tonga_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + int value; + + value = (mclk_table->dpm_levels[mclk_table->count - 1].value - + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value) * + 100 / + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return value; +} + +static int tonga_set_mclk_od(struct pp_hwmgr *hwmgr, uint32_t value) +{ + struct tonga_hwmgr *data = (struct tonga_hwmgr *)(hwmgr->backend); + struct tonga_single_dpm_table *golden_mclk_table = + &(data->golden_dpm_table.mclk_table); + struct pp_power_state *ps; + struct tonga_power_state *tonga_ps; + + if (value > 20) + value = 20; + + ps = hwmgr->request_ps; + + if (ps == NULL) + return -EINVAL; + + tonga_ps = cast_phw_tonga_power_state(&ps->hardware); + + tonga_ps->performance_levels[tonga_ps->performance_level_count - 1].memory_clock = + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value * + value / 100 + + golden_mclk_table->dpm_levels[golden_mclk_table->count - 1].value; + + return 0; +} + static const struct pp_hwmgr_func tonga_hwmgr_funcs = { .backend_init = &tonga_hwmgr_backend_init, .backend_fini = &tonga_hwmgr_backend_fini, .asic_setup = &tonga_setup_asic_task, .dynamic_state_management_enable = &tonga_enable_dpm_tasks, + .dynamic_state_management_disable = &tonga_disable_dpm_tasks, .apply_state_adjust_rules = tonga_apply_state_adjust_rules, .force_dpm_level = &tonga_force_dpm_level, .power_state_set = tonga_set_power_state_tasks, @@ -6232,22 +6259,16 @@ static const struct pp_hwmgr_func tonga_hwmgr_funcs = { .check_states_equal = tonga_check_states_equal, .set_fan_control_mode = tonga_set_fan_control_mode, .get_fan_control_mode = tonga_get_fan_control_mode, - .get_pp_table = tonga_get_pp_table, - .set_pp_table = tonga_set_pp_table, .force_clock_level = tonga_force_clock_level, .print_clock_levels = tonga_print_clock_levels, + .get_sclk_od = tonga_get_sclk_od, + .set_sclk_od = tonga_set_sclk_od, + .get_mclk_od = tonga_get_mclk_od, + .set_mclk_od = tonga_set_mclk_od, }; int tonga_hwmgr_init(struct pp_hwmgr *hwmgr) { - tonga_hwmgr *data; - - data = kzalloc (sizeof(tonga_hwmgr), GFP_KERNEL); - if (data == NULL) - return -ENOMEM; - memset(data, 0x00, sizeof(tonga_hwmgr)); - - hwmgr->backend = data; hwmgr->hwmgr_func = &tonga_hwmgr_funcs; hwmgr->pptable_func = &tonga_pptable_funcs; pp_tonga_thermal_initialize(hwmgr); diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.h index 573cd39fe78d..3961884bfa9b 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_hwmgr.h @@ -352,9 +352,6 @@ struct tonga_hwmgr { bool samu_power_gated; /* 1: gated, 0:not gated */ bool acp_power_gated; /* 1: gated, 0:not gated */ bool pg_acp_init; - - /* soft pptable for re-uploading into smu */ - void *soft_pp_table; }; typedef struct tonga_hwmgr tonga_hwmgr; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h index 1b44f4e9b8f5..f127198aafc4 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_pptable.h @@ -197,6 +197,22 @@ typedef struct _ATOM_Tonga_SCLK_Dependency_Table { ATOM_Tonga_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */ } ATOM_Tonga_SCLK_Dependency_Table; +typedef struct _ATOM_Polaris_SCLK_Dependency_Record { + UCHAR ucVddInd; /* Base voltage */ + USHORT usVddcOffset; /* Offset relative to base voltage */ + ULONG ulSclk; + USHORT usEdcCurrent; + UCHAR ucReliabilityTemperature; + UCHAR ucCKSVOffsetandDisable; /* Bits 0~6: Voltage offset for CKS, Bit 7: Disable/enable for the SCLK level. */ + ULONG ulSclkOffset; +} ATOM_Polaris_SCLK_Dependency_Record; + +typedef struct _ATOM_Polaris_SCLK_Dependency_Table { + UCHAR ucRevId; + UCHAR ucNumEntries; /* Number of entries. */ + ATOM_Polaris_SCLK_Dependency_Record entries[1]; /* Dynamically allocate entries. */ +} ATOM_Polaris_SCLK_Dependency_Table; + typedef struct _ATOM_Tonga_PCIE_Record { UCHAR ucPCIEGenSpeed; UCHAR usPCIELaneWidth; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c index 10e3630ee39d..94d6b472e1fe 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/tonga_processpptables.c @@ -408,41 +408,78 @@ static int get_mclk_voltage_dependency_table( static int get_sclk_voltage_dependency_table( struct pp_hwmgr *hwmgr, phm_ppt_v1_clock_voltage_dependency_table **pp_tonga_sclk_dep_table, - const ATOM_Tonga_SCLK_Dependency_Table * sclk_dep_table + const PPTable_Generic_SubTable_Header *sclk_dep_table ) { uint32_t table_size, i; phm_ppt_v1_clock_voltage_dependency_table *sclk_table; - PP_ASSERT_WITH_CODE((0 != sclk_dep_table->ucNumEntries), - "Invalid PowerPlay Table!", return -1); + if (sclk_dep_table->ucRevId < 1) { + const ATOM_Tonga_SCLK_Dependency_Table *tonga_table = + (ATOM_Tonga_SCLK_Dependency_Table *)sclk_dep_table; - table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record) - * sclk_dep_table->ucNumEntries; + PP_ASSERT_WITH_CODE((0 != tonga_table->ucNumEntries), + "Invalid PowerPlay Table!", return -1); - sclk_table = (phm_ppt_v1_clock_voltage_dependency_table *) - kzalloc(table_size, GFP_KERNEL); + table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record) + * tonga_table->ucNumEntries; - if (NULL == sclk_table) - return -ENOMEM; + sclk_table = (phm_ppt_v1_clock_voltage_dependency_table *) + kzalloc(table_size, GFP_KERNEL); - memset(sclk_table, 0x00, table_size); - - sclk_table->count = (uint32_t)sclk_dep_table->ucNumEntries; - - for (i = 0; i < sclk_dep_table->ucNumEntries; i++) { - sclk_table->entries[i].vddInd = - sclk_dep_table->entries[i].ucVddInd; - sclk_table->entries[i].vdd_offset = - sclk_dep_table->entries[i].usVddcOffset; - sclk_table->entries[i].clk = - sclk_dep_table->entries[i].ulSclk; - sclk_table->entries[i].cks_enable = - (((sclk_dep_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; - sclk_table->entries[i].cks_voffset = - (sclk_dep_table->entries[i].ucCKSVOffsetandDisable & 0x7F); - } + if (NULL == sclk_table) + return -ENOMEM; + memset(sclk_table, 0x00, table_size); + + sclk_table->count = (uint32_t)tonga_table->ucNumEntries; + + for (i = 0; i < tonga_table->ucNumEntries; i++) { + sclk_table->entries[i].vddInd = + tonga_table->entries[i].ucVddInd; + sclk_table->entries[i].vdd_offset = + tonga_table->entries[i].usVddcOffset; + sclk_table->entries[i].clk = + tonga_table->entries[i].ulSclk; + sclk_table->entries[i].cks_enable = + (((tonga_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; + sclk_table->entries[i].cks_voffset = + (tonga_table->entries[i].ucCKSVOffsetandDisable & 0x7F); + } + } else { + const ATOM_Polaris_SCLK_Dependency_Table *polaris_table = + (ATOM_Polaris_SCLK_Dependency_Table *)sclk_dep_table; + + PP_ASSERT_WITH_CODE((0 != polaris_table->ucNumEntries), + "Invalid PowerPlay Table!", return -1); + + table_size = sizeof(uint32_t) + sizeof(phm_ppt_v1_clock_voltage_dependency_record) + * polaris_table->ucNumEntries; + + sclk_table = (phm_ppt_v1_clock_voltage_dependency_table *) + kzalloc(table_size, GFP_KERNEL); + + if (NULL == sclk_table) + return -ENOMEM; + + memset(sclk_table, 0x00, table_size); + + sclk_table->count = (uint32_t)polaris_table->ucNumEntries; + + for (i = 0; i < polaris_table->ucNumEntries; i++) { + sclk_table->entries[i].vddInd = + polaris_table->entries[i].ucVddInd; + sclk_table->entries[i].vdd_offset = + polaris_table->entries[i].usVddcOffset; + sclk_table->entries[i].clk = + polaris_table->entries[i].ulSclk; + sclk_table->entries[i].cks_enable = + (((polaris_table->entries[i].ucCKSVOffsetandDisable & 0x80) >> 7) == 0) ? 1 : 0; + sclk_table->entries[i].cks_voffset = + (polaris_table->entries[i].ucCKSVOffsetandDisable & 0x7F); + sclk_table->entries[i].sclk_offset = polaris_table->entries[i].ulSclkOffset; + } + } *pp_tonga_sclk_dep_table = sclk_table; return 0; @@ -708,8 +745,8 @@ static int init_clock_voltage_dependency( const ATOM_Tonga_MCLK_Dependency_Table *mclk_dep_table = (const ATOM_Tonga_MCLK_Dependency_Table *)(((unsigned long) powerplay_table) + le16_to_cpu(powerplay_table->usMclkDependencyTableOffset)); - const ATOM_Tonga_SCLK_Dependency_Table *sclk_dep_table = - (const ATOM_Tonga_SCLK_Dependency_Table *)(((unsigned long) powerplay_table) + + const PPTable_Generic_SubTable_Header *sclk_dep_table = + (const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) + le16_to_cpu(powerplay_table->usSclkDependencyTableOffset)); const ATOM_Tonga_Hard_Limit_Table *pHardLimits = (const ATOM_Tonga_Hard_Limit_Table *)(((unsigned long) powerplay_table) + @@ -1040,48 +1077,41 @@ int tonga_pp_tables_uninitialize(struct pp_hwmgr *hwmgr) struct phm_ppt_v1_information *pp_table_information = (struct phm_ppt_v1_information *)(hwmgr->pptable); - if (NULL != hwmgr->soft_pp_table) { - kfree(hwmgr->soft_pp_table); - hwmgr->soft_pp_table = NULL; - } - - if (NULL != pp_table_information->vdd_dep_on_sclk) - pp_table_information->vdd_dep_on_sclk = NULL; + kfree(pp_table_information->vdd_dep_on_sclk); + pp_table_information->vdd_dep_on_sclk = NULL; - if (NULL != pp_table_information->vdd_dep_on_mclk) - pp_table_information->vdd_dep_on_mclk = NULL; + kfree(pp_table_information->vdd_dep_on_mclk); + pp_table_information->vdd_dep_on_mclk = NULL; - if (NULL != pp_table_information->valid_mclk_values) - pp_table_information->valid_mclk_values = NULL; + kfree(pp_table_information->valid_mclk_values); + pp_table_information->valid_mclk_values = NULL; - if (NULL != pp_table_information->valid_sclk_values) - pp_table_information->valid_sclk_values = NULL; + kfree(pp_table_information->valid_sclk_values); + pp_table_information->valid_sclk_values = NULL; - if (NULL != pp_table_information->vddc_lookup_table) - pp_table_information->vddc_lookup_table = NULL; + kfree(pp_table_information->vddc_lookup_table); + pp_table_information->vddc_lookup_table = NULL; - if (NULL != pp_table_information->vddgfx_lookup_table) - pp_table_information->vddgfx_lookup_table = NULL; + kfree(pp_table_information->vddgfx_lookup_table); + pp_table_information->vddgfx_lookup_table = NULL; - if (NULL != pp_table_information->mm_dep_table) - pp_table_information->mm_dep_table = NULL; + kfree(pp_table_information->mm_dep_table); + pp_table_information->mm_dep_table = NULL; - if (NULL != pp_table_information->cac_dtp_table) - pp_table_information->cac_dtp_table = NULL; + kfree(pp_table_information->cac_dtp_table); + pp_table_information->cac_dtp_table = NULL; - if (NULL != hwmgr->dyn_state.cac_dtp_table) - hwmgr->dyn_state.cac_dtp_table = NULL; + kfree(hwmgr->dyn_state.cac_dtp_table); + hwmgr->dyn_state.cac_dtp_table = NULL; - if (NULL != pp_table_information->ppm_parameter_table) - pp_table_information->ppm_parameter_table = NULL; + kfree(pp_table_information->ppm_parameter_table); + pp_table_information->ppm_parameter_table = NULL; - if (NULL != pp_table_information->pcie_table) - pp_table_information->pcie_table = NULL; + kfree(pp_table_information->pcie_table); + pp_table_information->pcie_table = NULL; - if (NULL != hwmgr->pptable) { - kfree(hwmgr->pptable); - hwmgr->pptable = NULL; - } + kfree(hwmgr->pptable); + hwmgr->pptable = NULL; return result; } diff --git a/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h b/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h index 50b367d44307..b764c8c05ec8 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h +++ b/drivers/gpu/drm/amd/powerplay/inc/amd_powerplay.h @@ -132,6 +132,7 @@ struct amd_pp_init { uint32_t chip_family; uint32_t chip_id; uint32_t rev_id; + bool powercontainment_enabled; }; enum amd_pp_display_config_type{ AMD_PP_DisplayConfigType_None = 0, @@ -342,6 +343,10 @@ struct amd_powerplay_funcs { int (*set_pp_table)(void *handle, const char *buf, size_t size); int (*force_clock_level)(void *handle, enum pp_clock_type type, uint32_t mask); int (*print_clock_levels)(void *handle, enum pp_clock_type type, char *buf); + int (*get_sclk_od)(void *handle); + int (*set_sclk_od)(void *handle, uint32_t value); + int (*get_mclk_od)(void *handle); + int (*set_mclk_od)(void *handle, uint32_t value); }; struct amd_powerplay { @@ -355,6 +360,8 @@ int amd_powerplay_init(struct amd_pp_init *pp_init, int amd_powerplay_fini(void *handle); +int amd_powerplay_reset(void *handle); + int amd_powerplay_display_configuration_change(void *handle, const struct amd_pp_display_configuration *input); diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h index 56f712c7d07a..962cb5385951 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h @@ -340,6 +340,7 @@ extern int phm_powergate_vce(struct pp_hwmgr *hwmgr, bool gate); extern int phm_powerdown_uvd(struct pp_hwmgr *hwmgr); extern int phm_setup_asic(struct pp_hwmgr *hwmgr); extern int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr); +extern int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr); extern void phm_init_dynamic_caps(struct pp_hwmgr *hwmgr); extern bool phm_is_hw_access_blocked(struct pp_hwmgr *hwmgr); extern int phm_block_hw_access(struct pp_hwmgr *hwmgr, bool block); diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h index 28f571449495..0bbc42a224e5 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h @@ -278,6 +278,8 @@ struct pp_hwmgr_func { int (*dynamic_state_management_enable)( struct pp_hwmgr *hw_mgr); + int (*dynamic_state_management_disable)( + struct pp_hwmgr *hw_mgr); int (*patch_boot_state)(struct pp_hwmgr *hwmgr, struct pp_hw_power_state *hw_ps); @@ -333,11 +335,13 @@ struct pp_hwmgr_func { int (*get_clock_by_type)(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type type, struct amd_pp_clocks *clocks); int (*get_max_high_clocks)(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks); int (*power_off_asic)(struct pp_hwmgr *hwmgr); - int (*get_pp_table)(struct pp_hwmgr *hwmgr, char **table); - int (*set_pp_table)(struct pp_hwmgr *hwmgr, const char *buf, size_t size); int (*force_clock_level)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask); int (*print_clock_levels)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf); int (*enable_per_cu_power_gating)(struct pp_hwmgr *hwmgr, bool enable); + int (*get_sclk_od)(struct pp_hwmgr *hwmgr); + int (*set_sclk_od)(struct pp_hwmgr *hwmgr, uint32_t value); + int (*get_mclk_od)(struct pp_hwmgr *hwmgr); + int (*set_mclk_od)(struct pp_hwmgr *hwmgr, uint32_t value); }; struct pp_table_func { @@ -578,6 +582,7 @@ struct pp_hwmgr { struct pp_smumgr *smumgr; const void *soft_pp_table; uint32_t soft_pp_table_size; + void *hardcode_pp_table; bool need_pp_table_upload; enum amd_dpm_forced_level dpm_level; bool block_hw_access; @@ -607,6 +612,7 @@ struct pp_hwmgr { uint32_t num_ps; struct pp_thermal_controller_info thermal_controller; bool fan_ctrl_is_in_default_mode; + bool powercontainment_enabled; uint32_t fan_ctrl_default_mode; uint32_t tmin; struct phm_microcode_version_info microcode_version_info; diff --git a/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h index 0c6a413eaa5b..d41d37ab5b7c 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h +++ b/drivers/gpu/drm/amd/powerplay/inc/polaris10_ppsmc.h @@ -27,6 +27,7 @@ #pragma pack(push, 1) +#define PPSMC_MSG_SetGBDroopSettings ((uint16_t) 0x305) #define PPSMC_SWSTATE_FLAG_DC 0x01 #define PPSMC_SWSTATE_FLAG_UVD 0x02 diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h b/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h index 3bd5e69b9045..3df5de2cdab0 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h +++ b/drivers/gpu/drm/amd/powerplay/inc/pp_acpi.h @@ -26,3 +26,4 @@ extern bool acpi_atcs_functions_supported(void *device, extern int acpi_pcie_perf_request(void *device, uint8_t perf_req, bool advertise); +extern bool acpi_atcs_notify_pcie_device_ready(void *device); diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu74.h b/drivers/gpu/drm/amd/powerplay/inc/smu74.h index 1a12d85b8e97..fd10a9fa843d 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu74.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu74.h @@ -34,6 +34,30 @@ #define SMU__NUM_LCLK_DPM_LEVELS 8 #define SMU__NUM_PCIE_DPM_LEVELS 8 +#define EXP_M1 35 +#define EXP_M2 92821 +#define EXP_B 66629747 + +#define EXP_M1_1 365 +#define EXP_M2_1 658700 +#define EXP_B_1 305506134 + +#define EXP_M1_2 189 +#define EXP_M2_2 379692 +#define EXP_B_2 194609469 + +#define EXP_M1_3 99 +#define EXP_M2_3 217915 +#define EXP_B_3 122255994 + +#define EXP_M1_4 51 +#define EXP_M2_4 122643 +#define EXP_B_4 74893384 + +#define EXP_M1_5 423 +#define EXP_M2_5 1103326 +#define EXP_B_5 728122621 + enum SID_OPTION { SID_OPTION_HI, SID_OPTION_LO, @@ -548,20 +572,20 @@ struct SMU74_Firmware_Header { uint32_t CacConfigTable; uint32_t CacStatusTable; - uint32_t mcRegisterTable; - uint32_t mcArbDramTimingTable; - - - uint32_t PmFuseTable; uint32_t Globals; uint32_t ClockStretcherTable; uint32_t VftTable; - uint32_t Reserved[21]; + uint32_t Reserved1; + uint32_t AvfsTable; + uint32_t AvfsCksOffGbvTable; + uint32_t AvfsMeanNSigma; + uint32_t AvfsSclkOffsetTable; + uint32_t Reserved[16]; uint32_t Signature; }; @@ -701,8 +725,6 @@ VR Config info is contained in dpmTable.VRConfig */ struct SMU_ClockStretcherDataTableEntry { uint8_t minVID; uint8_t maxVID; - - uint16_t setting; }; typedef struct SMU_ClockStretcherDataTableEntry SMU_ClockStretcherDataTableEntry; @@ -769,6 +791,43 @@ struct VFT_TABLE_t { typedef struct VFT_TABLE_t VFT_TABLE_t; +/* Total margin, root mean square of Fmax + DC + Platform */ +struct AVFS_Margin_t { + VFT_CELL_t Cell[NUM_VFT_COLUMNS]; +}; +typedef struct AVFS_Margin_t AVFS_Margin_t; + +#define BTCGB_VDROOP_TABLE_MAX_ENTRIES 2 +#define AVFSGB_VDROOP_TABLE_MAX_ENTRIES 2 + +struct GB_VDROOP_TABLE_t { + int32_t a0; + int32_t a1; + int32_t a2; + uint32_t spare; +}; +typedef struct GB_VDROOP_TABLE_t GB_VDROOP_TABLE_t; + +struct AVFS_CksOff_Gbv_t { + VFT_CELL_t Cell[NUM_VFT_COLUMNS]; +}; +typedef struct AVFS_CksOff_Gbv_t AVFS_CksOff_Gbv_t; + +struct AVFS_meanNsigma_t { + uint32_t Aconstant[3]; + uint16_t DC_tol_sigma; + uint16_t Platform_mean; + uint16_t Platform_sigma; + uint16_t PSM_Age_CompFactor; + uint8_t Static_Voltage_Offset[NUM_VFT_COLUMNS]; +}; +typedef struct AVFS_meanNsigma_t AVFS_meanNsigma_t; + +struct AVFS_Sclk_Offset_t { + uint16_t Sclk_Offset[8]; +}; +typedef struct AVFS_Sclk_Offset_t AVFS_Sclk_Offset_t; + #endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h b/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h index 0dfe82336dc7..b85ff5400e57 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu74_discrete.h @@ -223,6 +223,16 @@ struct SMU74_Discrete_StateInfo { typedef struct SMU74_Discrete_StateInfo SMU74_Discrete_StateInfo; +struct SMU_QuadraticCoeffs { + int32_t m1; + uint32_t b; + + int16_t m2; + uint8_t m1_shift; + uint8_t m2_shift; +}; +typedef struct SMU_QuadraticCoeffs SMU_QuadraticCoeffs; + struct SMU74_Discrete_DpmTable { SMU74_PIDController GraphicsPIDController; @@ -258,7 +268,14 @@ struct SMU74_Discrete_DpmTable { uint8_t ThermOutPolarity; uint8_t ThermOutMode; uint8_t BootPhases; - uint32_t Reserved[4]; + + uint8_t VRHotLevel; + uint8_t Reserved1[3]; + uint16_t FanStartTemperature; + uint16_t FanStopTemperature; + uint16_t MaxVoltage; + uint16_t Reserved2; + uint32_t Reserved[1]; SMU74_Discrete_GraphicsLevel GraphicsLevel[SMU74_MAX_LEVELS_GRAPHICS]; SMU74_Discrete_MemoryLevel MemoryACPILevel; @@ -347,6 +364,8 @@ struct SMU74_Discrete_DpmTable { uint32_t CurrSclkPllRange; sclkFcwRange_t SclkFcwRangeTable[NUM_SCLK_RANGE]; + GB_VDROOP_TABLE_t BTCGB_VDROOP_TABLE[BTCGB_VDROOP_TABLE_MAX_ENTRIES]; + SMU_QuadraticCoeffs AVFSGB_VDROOP_TABLE[AVFSGB_VDROOP_TABLE_MAX_ENTRIES]; }; typedef struct SMU74_Discrete_DpmTable SMU74_Discrete_DpmTable; @@ -550,16 +569,6 @@ struct SMU7_AcpiScoreboard { typedef struct SMU7_AcpiScoreboard SMU7_AcpiScoreboard; -struct SMU_QuadraticCoeffs { - int32_t m1; - uint32_t b; - - int16_t m2; - uint8_t m1_shift; - uint8_t m2_shift; -}; -typedef struct SMU_QuadraticCoeffs SMU_QuadraticCoeffs; - struct SMU74_Discrete_PmFuses { uint8_t BapmVddCVidHiSidd[8]; uint8_t BapmVddCVidLoSidd[8]; @@ -821,6 +830,17 @@ typedef struct SMU7_GfxCuPgScoreboard SMU7_GfxCuPgScoreboard; #define DB_PCC_SHIFT 26 #define DB_EDC_SHIFT 27 +#define BTCGB0_Vdroop_Enable_MASK 0x1 +#define BTCGB1_Vdroop_Enable_MASK 0x2 +#define AVFSGB0_Vdroop_Enable_MASK 0x4 +#define AVFSGB1_Vdroop_Enable_MASK 0x8 + +#define BTCGB0_Vdroop_Enable_SHIFT 0 +#define BTCGB1_Vdroop_Enable_SHIFT 1 +#define AVFSGB0_Vdroop_Enable_SHIFT 2 +#define AVFSGB1_Vdroop_Enable_SHIFT 3 + + #pragma pack(pop) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c index 673a75c74e18..8e52a2e82db5 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c @@ -1006,10 +1006,16 @@ static int fiji_smu_init(struct pp_smumgr *smumgr) static int fiji_smu_fini(struct pp_smumgr *smumgr) { + struct fiji_smumgr *priv = (struct fiji_smumgr *)(smumgr->backend); + + smu_free_memory(smumgr->device, (void *)priv->header_buffer.handle); + if (smumgr->backend) { kfree(smumgr->backend); smumgr->backend = NULL; } + + cgs_rel_firmware(smumgr->device, CGS_UCODE_ID_SMU); return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c index de618ead9db8..5dba7c509710 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c @@ -52,19 +52,18 @@ static const SMU74_Discrete_GraphicsLevel avfs_graphics_level_polaris10[8] = { /* Min pcie DeepSleep Activity CgSpll CgSpll CcPwr CcPwr Sclk Enabled Enabled Voltage Power */ /* Voltage, DpmLevel, DivId, Level, FuncCntl3, FuncCntl4, DynRm, DynRm1 Did, Padding,ForActivity, ForThrottle, UpHyst, DownHyst, DownHyst, Throttle */ - { 0x3c0fd047, 0x00, 0x03, 0x1e00, 0x00200410, 0x87020000, 0, 0, 0x16, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x30750000, 0, 0, 0, 0, 0, 0, 0 } }, - { 0xa00fd047, 0x01, 0x04, 0x1e00, 0x00800510, 0x87020000, 0, 0, 0x16, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x409c0000, 0, 0, 0, 0, 0, 0, 0 } }, - { 0x0410d047, 0x01, 0x00, 0x1e00, 0x00600410, 0x87020000, 0, 0, 0x0e, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x50c30000, 0, 0, 0, 0, 0, 0, 0 } }, - { 0x6810d047, 0x01, 0x00, 0x1e00, 0x00800410, 0x87020000, 0, 0, 0x0c, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x60ea0000, 0, 0, 0, 0, 0, 0, 0 } }, - { 0xcc10d047, 0x01, 0x00, 0x1e00, 0x00e00410, 0x87020000, 0, 0, 0x0c, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0xe8fd0000, 0, 0, 0, 0, 0, 0, 0 } }, - { 0x3011d047, 0x01, 0x00, 0x1e00, 0x00400510, 0x87020000, 0, 0, 0x0c, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x70110100, 0, 0, 0, 0, 0, 0, 0 } }, - { 0x9411d047, 0x01, 0x00, 0x1e00, 0x00a00510, 0x87020000, 0, 0, 0x0c, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0xf8240100, 0, 0, 0, 0, 0, 0, 0 } }, - { 0xf811d047, 0x01, 0x00, 0x1e00, 0x00000610, 0x87020000, 0, 0, 0x0c, 0, 0x01, 0x01, 0x00, 0x00, 0x00, 0x00, { 0x80380100, 0, 0, 0, 0, 0, 0, 0 } } + { 0x100ea446, 0x00, 0x03, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x30750000, 0x3000, 0, 0x2600, 0, 0, 0x0004, 0x8f02, 0xffff, 0x2f00, 0x300e, 0x2700 } }, + { 0x400ea446, 0x01, 0x04, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x409c0000, 0x2000, 0, 0x1e00, 1, 1, 0x0004, 0x8300, 0xffff, 0x1f00, 0xcb5e, 0x1a00 } }, + { 0x740ea446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x50c30000, 0x2800, 0, 0x2000, 1, 1, 0x0004, 0x0c02, 0xffff, 0x2700, 0x6433, 0x2100 } }, + { 0xa40ea446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x60ea0000, 0x3000, 0, 0x2600, 1, 1, 0x0004, 0x8f02, 0xffff, 0x2f00, 0x300e, 0x2700 } }, + { 0xd80ea446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x70110100, 0x3800, 0, 0x2c00, 1, 1, 0x0004, 0x1203, 0xffff, 0x3600, 0xc9e2, 0x2e00 } }, + { 0x3c0fa446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x80380100, 0x2000, 0, 0x1e00, 2, 1, 0x0004, 0x8300, 0xffff, 0x1f00, 0xcb5e, 0x1a00 } }, + { 0x6c0fa446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0x905f0100, 0x2400, 0, 0x1e00, 2, 1, 0x0004, 0x8901, 0xffff, 0x2300, 0x314c, 0x1d00 } }, + { 0xa00fa446, 0x01, 0x00, 0x3200, 0, 0, 0, 0, 0, 0, 0x01, 0x01, 0x0a, 0x00, 0x00, 0x00, { 0xa0860100, 0x2800, 0, 0x2000, 2, 1, 0x0004, 0x0c02, 0xffff, 0x2700, 0x6433, 0x2100 } } }; static const SMU74_Discrete_MemoryLevel avfs_memory_level_polaris10 = - {0x50140000, 0x50140000, 0x00320000, 0x00, 0x00, - 0x00, 0x10, 0x00, 0x00, 0x00, 0x00, 0x0000, 0x00, 0x00}; + {0x100ea446, 0, 0x30750000, 0x01, 0x01, 0x01, 0x00, 0x00, 0x64, 0x00, 0x00, 0x1f00, 0x00, 0x00}; /** * Set the address for reading/writing the SMC SRAM space. @@ -219,6 +218,18 @@ bool polaris10_is_smc_ram_running(struct pp_smumgr *smumgr) && (0x20100 <= cgs_read_ind_register(smumgr->device, CGS_IND_REG__SMC, ixSMC_PC_C))); } +static bool polaris10_is_hw_avfs_present(struct pp_smumgr *smumgr) +{ + uint32_t efuse; + + efuse = cgs_read_ind_register(smumgr->device, CGS_IND_REG__SMC, ixSMU_EFUSE_0 + (49*4)); + efuse &= 0x00000001; + if (efuse) + return true; + + return false; +} + /** * Send a message to the SMC, and wait for its response. * @@ -228,21 +239,27 @@ bool polaris10_is_smc_ram_running(struct pp_smumgr *smumgr) */ int polaris10_send_msg_to_smc(struct pp_smumgr *smumgr, uint16_t msg) { + int ret; + if (!polaris10_is_smc_ram_running(smumgr)) return -1; + SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMC_RESP_0, SMC_RESP, 0); - if (1 != SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP)) - printk("Failed to send Previous Message.\n"); + ret = SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP); + if (ret != 1) + printk("\n failed to send pre message %x ret is %d \n", msg, ret); cgs_write_register(smumgr->device, mmSMC_MESSAGE_0, msg); SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMC_RESP_0, SMC_RESP, 0); - if (1 != SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP)) - printk("Failed to send Message.\n"); + ret = SMUM_READ_FIELD(smumgr->device, SMC_RESP_0, SMC_RESP); + + if (ret != 1) + printk("\n failed to send message %x ret is %d \n", msg, ret); return 0; } @@ -469,6 +486,7 @@ int polaris10_smu_fini(struct pp_smumgr *smumgr) kfree(smumgr->backend); smumgr->backend = NULL; } + cgs_rel_firmware(smumgr->device, CGS_UCODE_ID_SMU); return 0; } @@ -952,6 +970,11 @@ static int polaris10_smu_init(struct pp_smumgr *smumgr) (cgs_handle_t)smu_data->smu_buffer.handle); return -1;); + if (polaris10_is_hw_avfs_present(smumgr)) + smu_data->avfs.avfs_btc_status = AVFS_BTC_BOOT; + else + smu_data->avfs.avfs_btc_status = AVFS_BTC_NOTSUPPORTED; + return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c index c483baf6b4fb..7723473e51a0 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c @@ -23,6 +23,7 @@ #include <linux/types.h> #include <linux/kernel.h> #include <linux/slab.h> +#include <drm/amdgpu_drm.h> #include "pp_instance.h" #include "smumgr.h" #include "cgs_common.h" @@ -52,10 +53,10 @@ int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle) handle->smu_mgr = smumgr; switch (smumgr->chip_family) { - case AMD_FAMILY_CZ: + case AMDGPU_FAMILY_CZ: cz_smum_init(smumgr); break; - case AMD_FAMILY_VI: + case AMDGPU_FAMILY_VI: switch (smumgr->chip_id) { case CHIP_TONGA: tonga_smum_init(smumgr); @@ -81,6 +82,7 @@ int smum_init(struct amd_pp_init *pp_init, struct pp_instance *handle) int smum_fini(struct pp_smumgr *smumgr) { + kfree(smumgr->device); kfree(smumgr); return 0; } diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c index 32820b680d88..b22722eabafc 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c @@ -328,10 +328,17 @@ int tonga_write_smc_sram_dword(struct pp_smumgr *smumgr, static int tonga_smu_fini(struct pp_smumgr *smumgr) { + struct tonga_smumgr *priv = (struct tonga_smumgr *)(smumgr->backend); + + smu_free_memory(smumgr->device, (void *)priv->smu_buffer.handle); + smu_free_memory(smumgr->device, (void *)priv->header_buffer.handle); + if (smumgr->backend != NULL) { kfree(smumgr->backend); smumgr->backend = NULL; } + + cgs_rel_firmware(smumgr->device, CGS_UCODE_ID_SMU); return 0; } diff --git a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h index c89dc777768f..b961a1c6caf3 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_sched_trace.h @@ -26,7 +26,7 @@ TRACE_EVENT(amd_sched_job, TP_fast_assign( __entry->entity = sched_job->s_entity; __entry->sched_job = sched_job; - __entry->fence = &sched_job->s_fence->base; + __entry->fence = &sched_job->s_fence->finished; __entry->name = sched_job->sched->name; __entry->job_count = kfifo_len( &sched_job->s_entity->job_queue) / sizeof(sched_job); @@ -46,7 +46,7 @@ TRACE_EVENT(amd_sched_process_job, ), TP_fast_assign( - __entry->fence = &fence->base; + __entry->fence = &fence->finished; ), TP_printk("fence=%p signaled", __entry->fence) ); diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c index c16248cee779..70ff09d10885 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.c @@ -32,6 +32,7 @@ static bool amd_sched_entity_is_ready(struct amd_sched_entity *entity); static void amd_sched_wakeup(struct amd_gpu_scheduler *sched); +static void amd_sched_process_job(struct fence *f, struct fence_cb *cb); struct kmem_cache *sched_fence_slab; atomic_t sched_fence_slab_ref = ATOMIC_INIT(0); @@ -140,7 +141,7 @@ int amd_sched_entity_init(struct amd_gpu_scheduler *sched, return r; atomic_set(&entity->fence_seq, 0); - entity->fence_context = fence_context_alloc(1); + entity->fence_context = fence_context_alloc(2); return 0; } @@ -251,17 +252,21 @@ static bool amd_sched_entity_add_dependency_cb(struct amd_sched_entity *entity) s_fence = to_amd_sched_fence(fence); if (s_fence && s_fence->sched == sched) { - /* Fence is from the same scheduler */ - if (test_bit(AMD_SCHED_FENCE_SCHEDULED_BIT, &fence->flags)) { - /* Ignore it when it is already scheduled */ - fence_put(entity->dependency); - return false; - } - /* Wait for fence to be scheduled */ - entity->cb.func = amd_sched_entity_clear_dep; - list_add_tail(&entity->cb.node, &s_fence->scheduled_cb); - return true; + /* + * Fence is from the same scheduler, only need to wait for + * it to be scheduled + */ + fence = fence_get(&s_fence->scheduled); + fence_put(entity->dependency); + entity->dependency = fence; + if (!fence_add_callback(fence, &entity->cb, + amd_sched_entity_clear_dep)) + return true; + + /* Ignore it when it is already scheduled */ + fence_put(fence); + return false; } if (!fence_add_callback(entity->dependency, &entity->cb, @@ -319,46 +324,108 @@ static bool amd_sched_entity_in(struct amd_sched_job *sched_job) return added; } -static void amd_sched_free_job(struct fence *f, struct fence_cb *cb) { - struct amd_sched_job *job = container_of(cb, struct amd_sched_job, cb_free_job); - schedule_work(&job->work_free_job); -} - /* job_finish is called after hw fence signaled, and * the job had already been deleted from ring_mirror_list */ -void amd_sched_job_finish(struct amd_sched_job *s_job) +static void amd_sched_job_finish(struct work_struct *work) { - struct amd_sched_job *next; + struct amd_sched_job *s_job = container_of(work, struct amd_sched_job, + finish_work); struct amd_gpu_scheduler *sched = s_job->sched; + /* remove job from ring_mirror_list */ + spin_lock(&sched->job_list_lock); + list_del_init(&s_job->node); if (sched->timeout != MAX_SCHEDULE_TIMEOUT) { - if (cancel_delayed_work(&s_job->work_tdr)) - amd_sched_job_put(s_job); + struct amd_sched_job *next; + + spin_unlock(&sched->job_list_lock); + cancel_delayed_work_sync(&s_job->work_tdr); + spin_lock(&sched->job_list_lock); /* queue TDR for next job */ next = list_first_entry_or_null(&sched->ring_mirror_list, struct amd_sched_job, node); - if (next) { - INIT_DELAYED_WORK(&next->work_tdr, s_job->timeout_callback); - amd_sched_job_get(next); + if (next) schedule_delayed_work(&next->work_tdr, sched->timeout); - } } + spin_unlock(&sched->job_list_lock); + sched->ops->free_job(s_job); } -void amd_sched_job_begin(struct amd_sched_job *s_job) +static void amd_sched_job_finish_cb(struct fence *f, struct fence_cb *cb) +{ + struct amd_sched_job *job = container_of(cb, struct amd_sched_job, + finish_cb); + schedule_work(&job->finish_work); +} + +static void amd_sched_job_begin(struct amd_sched_job *s_job) { struct amd_gpu_scheduler *sched = s_job->sched; + spin_lock(&sched->job_list_lock); + list_add_tail(&s_job->node, &sched->ring_mirror_list); if (sched->timeout != MAX_SCHEDULE_TIMEOUT && - list_first_entry_or_null(&sched->ring_mirror_list, struct amd_sched_job, node) == s_job) - { - INIT_DELAYED_WORK(&s_job->work_tdr, s_job->timeout_callback); - amd_sched_job_get(s_job); + list_first_entry_or_null(&sched->ring_mirror_list, + struct amd_sched_job, node) == s_job) + schedule_delayed_work(&s_job->work_tdr, sched->timeout); + spin_unlock(&sched->job_list_lock); +} + +static void amd_sched_job_timedout(struct work_struct *work) +{ + struct amd_sched_job *job = container_of(work, struct amd_sched_job, + work_tdr.work); + + job->sched->ops->timedout_job(job); +} + +void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched) +{ + struct amd_sched_job *s_job; + + spin_lock(&sched->job_list_lock); + list_for_each_entry_reverse(s_job, &sched->ring_mirror_list, node) { + if (fence_remove_callback(s_job->s_fence->parent, &s_job->s_fence->cb)) { + fence_put(s_job->s_fence->parent); + s_job->s_fence->parent = NULL; + } + } + spin_unlock(&sched->job_list_lock); +} + +void amd_sched_job_recovery(struct amd_gpu_scheduler *sched) +{ + struct amd_sched_job *s_job; + int r; + + spin_lock(&sched->job_list_lock); + s_job = list_first_entry_or_null(&sched->ring_mirror_list, + struct amd_sched_job, node); + if (s_job) schedule_delayed_work(&s_job->work_tdr, sched->timeout); + + list_for_each_entry(s_job, &sched->ring_mirror_list, node) { + struct amd_sched_fence *s_fence = s_job->s_fence; + struct fence *fence = sched->ops->run_job(s_job); + if (fence) { + s_fence->parent = fence_get(fence); + r = fence_add_callback(fence, &s_fence->cb, + amd_sched_process_job); + if (r == -ENOENT) + amd_sched_process_job(fence, &s_fence->cb); + else if (r) + DRM_ERROR("fence add callback failed (%d)\n", + r); + fence_put(fence); + } else { + DRM_ERROR("Failed to run job!\n"); + amd_sched_process_job(NULL, &s_fence->cb); + } } + spin_unlock(&sched->job_list_lock); } /** @@ -372,36 +439,29 @@ void amd_sched_entity_push_job(struct amd_sched_job *sched_job) { struct amd_sched_entity *entity = sched_job->s_entity; - sched_job->use_sched = 1; - fence_add_callback(&sched_job->s_fence->base, - &sched_job->cb_free_job, amd_sched_free_job); trace_amd_sched_job(sched_job); + fence_add_callback(&sched_job->s_fence->finished, &sched_job->finish_cb, + amd_sched_job_finish_cb); wait_event(entity->sched->job_scheduled, amd_sched_entity_in(sched_job)); } /* init a sched_job with basic field */ int amd_sched_job_init(struct amd_sched_job *job, - struct amd_gpu_scheduler *sched, - struct amd_sched_entity *entity, - void (*timeout_cb)(struct work_struct *work), - void (*free_cb)(struct kref *refcount), - void *owner, struct fence **fence) + struct amd_gpu_scheduler *sched, + struct amd_sched_entity *entity, + void *owner) { - INIT_LIST_HEAD(&job->node); - kref_init(&job->refcount); job->sched = sched; job->s_entity = entity; job->s_fence = amd_sched_fence_create(entity, owner); if (!job->s_fence) return -ENOMEM; - job->s_fence->s_job = job; - job->timeout_callback = timeout_cb; - job->free_callback = free_cb; + INIT_WORK(&job->finish_work, amd_sched_job_finish); + INIT_LIST_HEAD(&job->node); + INIT_DELAYED_WORK(&job->work_tdr, amd_sched_job_timedout); - if (fence) - *fence = &job->s_fence->base; return 0; } @@ -450,23 +510,25 @@ static void amd_sched_process_job(struct fence *f, struct fence_cb *cb) struct amd_sched_fence *s_fence = container_of(cb, struct amd_sched_fence, cb); struct amd_gpu_scheduler *sched = s_fence->sched; - unsigned long flags; atomic_dec(&sched->hw_rq_count); - - /* remove job from ring_mirror_list */ - spin_lock_irqsave(&sched->job_list_lock, flags); - list_del_init(&s_fence->s_job->node); - sched->ops->finish_job(s_fence->s_job); - spin_unlock_irqrestore(&sched->job_list_lock, flags); - - amd_sched_fence_signal(s_fence); + amd_sched_fence_finished(s_fence); trace_amd_sched_process_job(s_fence); - fence_put(&s_fence->base); + fence_put(&s_fence->finished); wake_up_interruptible(&sched->wake_up_worker); } +static bool amd_sched_blocked(struct amd_gpu_scheduler *sched) +{ + if (kthread_should_park()) { + kthread_parkme(); + return true; + } + + return false; +} + static int amd_sched_main(void *param) { struct sched_param sparam = {.sched_priority = 1}; @@ -476,14 +538,15 @@ static int amd_sched_main(void *param) sched_setscheduler(current, SCHED_FIFO, &sparam); while (!kthread_should_stop()) { - struct amd_sched_entity *entity; + struct amd_sched_entity *entity = NULL; struct amd_sched_fence *s_fence; struct amd_sched_job *sched_job; struct fence *fence; wait_event_interruptible(sched->wake_up_worker, - (entity = amd_sched_select_entity(sched)) || - kthread_should_stop()); + (!amd_sched_blocked(sched) && + (entity = amd_sched_select_entity(sched))) || + kthread_should_stop()); if (!entity) continue; @@ -495,16 +558,19 @@ static int amd_sched_main(void *param) s_fence = sched_job->s_fence; atomic_inc(&sched->hw_rq_count); - amd_sched_job_pre_schedule(sched, sched_job); + amd_sched_job_begin(sched_job); + fence = sched->ops->run_job(sched_job); amd_sched_fence_scheduled(s_fence); if (fence) { + s_fence->parent = fence_get(fence); r = fence_add_callback(fence, &s_fence->cb, amd_sched_process_job); if (r == -ENOENT) amd_sched_process_job(fence, &s_fence->cb); else if (r) - DRM_ERROR("fence add callback failed (%d)\n", r); + DRM_ERROR("fence add callback failed (%d)\n", + r); fence_put(fence); } else { DRM_ERROR("Failed to run job!\n"); diff --git a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h index 070095a9433c..7cbbbfb502ef 100644 --- a/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h +++ b/drivers/gpu/drm/amd/scheduler/gpu_scheduler.h @@ -27,8 +27,6 @@ #include <linux/kfifo.h> #include <linux/fence.h> -#define AMD_SCHED_FENCE_SCHEDULED_BIT FENCE_FLAG_USER_BITS - struct amd_gpu_scheduler; struct amd_sched_rq; @@ -68,36 +66,34 @@ struct amd_sched_rq { }; struct amd_sched_fence { - struct fence base; + struct fence scheduled; + struct fence finished; struct fence_cb cb; - struct list_head scheduled_cb; + struct fence *parent; struct amd_gpu_scheduler *sched; spinlock_t lock; void *owner; - struct amd_sched_job *s_job; }; struct amd_sched_job { - struct kref refcount; struct amd_gpu_scheduler *sched; struct amd_sched_entity *s_entity; struct amd_sched_fence *s_fence; - bool use_sched; /* true if the job goes to scheduler */ - struct fence_cb cb_free_job; - struct work_struct work_free_job; - struct list_head node; - struct delayed_work work_tdr; - void (*timeout_callback) (struct work_struct *work); - void (*free_callback)(struct kref *refcount); + struct fence_cb finish_cb; + struct work_struct finish_work; + struct list_head node; + struct delayed_work work_tdr; }; -extern const struct fence_ops amd_sched_fence_ops; +extern const struct fence_ops amd_sched_fence_ops_scheduled; +extern const struct fence_ops amd_sched_fence_ops_finished; static inline struct amd_sched_fence *to_amd_sched_fence(struct fence *f) { - struct amd_sched_fence *__f = container_of(f, struct amd_sched_fence, base); + if (f->ops == &amd_sched_fence_ops_scheduled) + return container_of(f, struct amd_sched_fence, scheduled); - if (__f->base.ops == &amd_sched_fence_ops) - return __f; + if (f->ops == &amd_sched_fence_ops_finished) + return container_of(f, struct amd_sched_fence, finished); return NULL; } @@ -109,8 +105,8 @@ static inline struct amd_sched_fence *to_amd_sched_fence(struct fence *f) struct amd_sched_backend_ops { struct fence *(*dependency)(struct amd_sched_job *sched_job); struct fence *(*run_job)(struct amd_sched_job *sched_job); - void (*begin_job)(struct amd_sched_job *sched_job); - void (*finish_job)(struct amd_sched_job *sched_job); + void (*timedout_job)(struct amd_sched_job *sched_job); + void (*free_job)(struct amd_sched_job *sched_job); }; enum amd_sched_priority { @@ -152,25 +148,11 @@ void amd_sched_entity_push_job(struct amd_sched_job *sched_job); struct amd_sched_fence *amd_sched_fence_create( struct amd_sched_entity *s_entity, void *owner); void amd_sched_fence_scheduled(struct amd_sched_fence *fence); -void amd_sched_fence_signal(struct amd_sched_fence *fence); +void amd_sched_fence_finished(struct amd_sched_fence *fence); int amd_sched_job_init(struct amd_sched_job *job, - struct amd_gpu_scheduler *sched, - struct amd_sched_entity *entity, - void (*timeout_cb)(struct work_struct *work), - void (*free_cb)(struct kref* refcount), - void *owner, struct fence **fence); -void amd_sched_job_pre_schedule(struct amd_gpu_scheduler *sched , - struct amd_sched_job *s_job); -void amd_sched_job_finish(struct amd_sched_job *s_job); -void amd_sched_job_begin(struct amd_sched_job *s_job); -static inline void amd_sched_job_get(struct amd_sched_job *job) { - if (job) - kref_get(&job->refcount); -} - -static inline void amd_sched_job_put(struct amd_sched_job *job) { - if (job) - kref_put(&job->refcount, job->free_callback); -} - + struct amd_gpu_scheduler *sched, + struct amd_sched_entity *entity, + void *owner); +void amd_sched_hw_job_reset(struct amd_gpu_scheduler *sched); +void amd_sched_job_recovery(struct amd_gpu_scheduler *sched); #endif diff --git a/drivers/gpu/drm/amd/scheduler/sched_fence.c b/drivers/gpu/drm/amd/scheduler/sched_fence.c index 2a732c490375..6b63beaf7574 100644 --- a/drivers/gpu/drm/amd/scheduler/sched_fence.c +++ b/drivers/gpu/drm/amd/scheduler/sched_fence.c @@ -27,7 +27,8 @@ #include <drm/drmP.h> #include "gpu_scheduler.h" -struct amd_sched_fence *amd_sched_fence_create(struct amd_sched_entity *s_entity, void *owner) +struct amd_sched_fence *amd_sched_fence_create(struct amd_sched_entity *entity, + void *owner) { struct amd_sched_fence *fence = NULL; unsigned seq; @@ -36,46 +37,37 @@ struct amd_sched_fence *amd_sched_fence_create(struct amd_sched_entity *s_entity if (fence == NULL) return NULL; - INIT_LIST_HEAD(&fence->scheduled_cb); fence->owner = owner; - fence->sched = s_entity->sched; + fence->sched = entity->sched; spin_lock_init(&fence->lock); - seq = atomic_inc_return(&s_entity->fence_seq); - fence_init(&fence->base, &amd_sched_fence_ops, &fence->lock, - s_entity->fence_context, seq); + seq = atomic_inc_return(&entity->fence_seq); + fence_init(&fence->scheduled, &amd_sched_fence_ops_scheduled, + &fence->lock, entity->fence_context, seq); + fence_init(&fence->finished, &amd_sched_fence_ops_finished, + &fence->lock, entity->fence_context + 1, seq); return fence; } -void amd_sched_fence_signal(struct amd_sched_fence *fence) +void amd_sched_fence_scheduled(struct amd_sched_fence *fence) { - int ret = fence_signal(&fence->base); + int ret = fence_signal(&fence->scheduled); + if (!ret) - FENCE_TRACE(&fence->base, "signaled from irq context\n"); + FENCE_TRACE(&fence->scheduled, "signaled from irq context\n"); else - FENCE_TRACE(&fence->base, "was already signaled\n"); -} - -void amd_sched_job_pre_schedule(struct amd_gpu_scheduler *sched , - struct amd_sched_job *s_job) -{ - unsigned long flags; - spin_lock_irqsave(&sched->job_list_lock, flags); - list_add_tail(&s_job->node, &sched->ring_mirror_list); - sched->ops->begin_job(s_job); - spin_unlock_irqrestore(&sched->job_list_lock, flags); + FENCE_TRACE(&fence->scheduled, "was already signaled\n"); } -void amd_sched_fence_scheduled(struct amd_sched_fence *s_fence) +void amd_sched_fence_finished(struct amd_sched_fence *fence) { - struct fence_cb *cur, *tmp; + int ret = fence_signal(&fence->finished); - set_bit(AMD_SCHED_FENCE_SCHEDULED_BIT, &s_fence->base.flags); - list_for_each_entry_safe(cur, tmp, &s_fence->scheduled_cb, node) { - list_del_init(&cur->node); - cur->func(&s_fence->base, cur); - } + if (!ret) + FENCE_TRACE(&fence->finished, "signaled from irq context\n"); + else + FENCE_TRACE(&fence->finished, "was already signaled\n"); } static const char *amd_sched_fence_get_driver_name(struct fence *fence) @@ -105,6 +97,8 @@ static void amd_sched_fence_free(struct rcu_head *rcu) { struct fence *f = container_of(rcu, struct fence, rcu); struct amd_sched_fence *fence = to_amd_sched_fence(f); + + fence_put(fence->parent); kmem_cache_free(sched_fence_slab, fence); } @@ -116,16 +110,41 @@ static void amd_sched_fence_free(struct rcu_head *rcu) * This function is called when the reference count becomes zero. * It just RCU schedules freeing up the fence. */ -static void amd_sched_fence_release(struct fence *f) +static void amd_sched_fence_release_scheduled(struct fence *f) { - call_rcu(&f->rcu, amd_sched_fence_free); + struct amd_sched_fence *fence = to_amd_sched_fence(f); + + call_rcu(&fence->finished.rcu, amd_sched_fence_free); } -const struct fence_ops amd_sched_fence_ops = { +/** + * amd_sched_fence_release_scheduled - drop extra reference + * + * @f: fence + * + * Drop the extra reference from the scheduled fence to the base fence. + */ +static void amd_sched_fence_release_finished(struct fence *f) +{ + struct amd_sched_fence *fence = to_amd_sched_fence(f); + + fence_put(&fence->scheduled); +} + +const struct fence_ops amd_sched_fence_ops_scheduled = { + .get_driver_name = amd_sched_fence_get_driver_name, + .get_timeline_name = amd_sched_fence_get_timeline_name, + .enable_signaling = amd_sched_fence_enable_signaling, + .signaled = NULL, + .wait = fence_default_wait, + .release = amd_sched_fence_release_scheduled, +}; + +const struct fence_ops amd_sched_fence_ops_finished = { .get_driver_name = amd_sched_fence_get_driver_name, .get_timeline_name = amd_sched_fence_get_timeline_name, .enable_signaling = amd_sched_fence_enable_signaling, .signaled = NULL, .wait = fence_default_wait, - .release = amd_sched_fence_release, + .release = amd_sched_fence_release_finished, }; diff --git a/drivers/gpu/drm/ast/ast_ttm.c b/drivers/gpu/drm/ast/ast_ttm.c index 59f2f93b6f84..b29a41218fc9 100644 --- a/drivers/gpu/drm/ast/ast_ttm.c +++ b/drivers/gpu/drm/ast/ast_ttm.c @@ -186,17 +186,6 @@ static void ast_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_reg * { } -static int ast_bo_move(struct ttm_buffer_object *bo, - bool evict, bool interruptible, - bool no_wait_gpu, - struct ttm_mem_reg *new_mem) -{ - int r; - r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); - return r; -} - - static void ast_ttm_backend_destroy(struct ttm_tt *tt) { ttm_tt_fini(tt); @@ -241,7 +230,7 @@ struct ttm_bo_driver ast_bo_driver = { .ttm_tt_unpopulate = ast_ttm_tt_unpopulate, .init_mem_type = ast_bo_init_mem_type, .evict_flags = ast_bo_evict_flags, - .move = ast_bo_move, + .move = NULL, .verify_access = ast_bo_verify_access, .io_mem_reserve = &ast_ttm_io_mem_reserve, .io_mem_free = &ast_ttm_io_mem_free, diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c index 473a475f27b1..6119b5085501 100644 --- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c +++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_output.c @@ -254,9 +254,10 @@ int atmel_hlcdc_create_outputs(struct drm_device *dev) if (!ret) ret = atmel_hlcdc_check_endpoint(dev, &ep); - of_node_put(ep_np); - if (ret) + if (ret) { + of_node_put(ep_np); return ret; + } } for_each_endpoint_of_node(dev->dev->of_node, ep_np) { @@ -264,9 +265,10 @@ int atmel_hlcdc_create_outputs(struct drm_device *dev) if (!ret) ret = atmel_hlcdc_attach_endpoint(dev, &ep); - of_node_put(ep_np); - if (ret) + if (ret) { + of_node_put(ep_np); return ret; + } } return 0; diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c index aef3ca8a81fa..016c191221f3 100644 --- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c +++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c @@ -339,6 +339,8 @@ atmel_hlcdc_plane_update_pos_and_size(struct atmel_hlcdc_plane *plane, atmel_hlcdc_layer_update_cfg(&plane->layer, 13, 0xffffffff, factor_reg); + } else { + atmel_hlcdc_layer_update_cfg(&plane->layer, 13, 0xffffffff, 0); } } diff --git a/drivers/gpu/drm/bochs/bochs_mm.c b/drivers/gpu/drm/bochs/bochs_mm.c index 6cf912c45e48..b109fdcaa679 100644 --- a/drivers/gpu/drm/bochs/bochs_mm.c +++ b/drivers/gpu/drm/bochs/bochs_mm.c @@ -165,15 +165,6 @@ static void bochs_ttm_io_mem_free(struct ttm_bo_device *bdev, { } -static int bochs_bo_move(struct ttm_buffer_object *bo, - bool evict, bool interruptible, - bool no_wait_gpu, - struct ttm_mem_reg *new_mem) -{ - return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); -} - - static void bochs_ttm_backend_destroy(struct ttm_tt *tt) { ttm_tt_fini(tt); @@ -208,7 +199,7 @@ struct ttm_bo_driver bochs_bo_driver = { .ttm_tt_unpopulate = ttm_pool_unpopulate, .init_mem_type = bochs_bo_init_mem_type, .evict_flags = bochs_bo_evict_flags, - .move = bochs_bo_move, + .move = NULL, .verify_access = bochs_bo_verify_access, .io_mem_reserve = &bochs_ttm_io_mem_reserve, .io_mem_free = &bochs_ttm_io_mem_free, diff --git a/drivers/gpu/drm/cirrus/cirrus_ttm.c b/drivers/gpu/drm/cirrus/cirrus_ttm.c index 6768b7b1af32..1cc9ee607128 100644 --- a/drivers/gpu/drm/cirrus/cirrus_ttm.c +++ b/drivers/gpu/drm/cirrus/cirrus_ttm.c @@ -186,17 +186,6 @@ static void cirrus_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_re { } -static int cirrus_bo_move(struct ttm_buffer_object *bo, - bool evict, bool interruptible, - bool no_wait_gpu, - struct ttm_mem_reg *new_mem) -{ - int r; - r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); - return r; -} - - static void cirrus_ttm_backend_destroy(struct ttm_tt *tt) { ttm_tt_fini(tt); @@ -241,7 +230,7 @@ struct ttm_bo_driver cirrus_bo_driver = { .ttm_tt_unpopulate = cirrus_ttm_tt_unpopulate, .init_mem_type = cirrus_bo_init_mem_type, .evict_flags = cirrus_bo_evict_flags, - .move = cirrus_bo_move, + .move = NULL, .verify_access = cirrus_bo_verify_access, .io_mem_reserve = &cirrus_ttm_io_mem_reserve, .io_mem_free = &cirrus_ttm_io_mem_free, diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c index dac0875e669c..9359be4a0ca9 100644 --- a/drivers/gpu/drm/drm_atomic.c +++ b/drivers/gpu/drm/drm_atomic.c @@ -1299,14 +1299,39 @@ EXPORT_SYMBOL(drm_atomic_add_affected_planes); */ void drm_atomic_legacy_backoff(struct drm_atomic_state *state) { + struct drm_device *dev = state->dev; + unsigned crtc_mask = 0; + struct drm_crtc *crtc; int ret; + bool global = false; + + drm_for_each_crtc(crtc, dev) { + if (crtc->acquire_ctx != state->acquire_ctx) + continue; + + crtc_mask |= drm_crtc_mask(crtc); + crtc->acquire_ctx = NULL; + } + + if (WARN_ON(dev->mode_config.acquire_ctx == state->acquire_ctx)) { + global = true; + + dev->mode_config.acquire_ctx = NULL; + } retry: drm_modeset_backoff(state->acquire_ctx); - ret = drm_modeset_lock_all_ctx(state->dev, state->acquire_ctx); + ret = drm_modeset_lock_all_ctx(dev, state->acquire_ctx); if (ret) goto retry; + + drm_for_each_crtc(crtc, dev) + if (drm_crtc_mask(crtc) & crtc_mask) + crtc->acquire_ctx = state->acquire_ctx; + + if (global) + dev->mode_config.acquire_ctx = state->acquire_ctx; } EXPORT_SYMBOL(drm_atomic_legacy_backoff); diff --git a/drivers/gpu/drm/drm_crtc_helper.c b/drivers/gpu/drm/drm_crtc_helper.c index d61591274ff6..604d3ef72ffa 100644 --- a/drivers/gpu/drm/drm_crtc_helper.c +++ b/drivers/gpu/drm/drm_crtc_helper.c @@ -531,11 +531,11 @@ drm_crtc_helper_disable(struct drm_crtc *crtc) int drm_crtc_helper_set_config(struct drm_mode_set *set) { struct drm_device *dev; - struct drm_crtc *new_crtc; - struct drm_encoder *save_encoders, *new_encoder, *encoder; + struct drm_crtc **save_encoder_crtcs, *new_crtc; + struct drm_encoder **save_connector_encoders, *new_encoder, *encoder; bool mode_changed = false; /* if true do a full mode set */ bool fb_changed = false; /* if true and !mode_changed just do a flip */ - struct drm_connector *save_connectors, *connector; + struct drm_connector *connector; int count = 0, ro, fail = 0; const struct drm_crtc_helper_funcs *crtc_funcs; struct drm_mode_set save_set; @@ -577,15 +577,15 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) * Allocate space for the backup of all (non-pointer) encoder and * connector data. */ - save_encoders = kzalloc(dev->mode_config.num_encoder * - sizeof(struct drm_encoder), GFP_KERNEL); - if (!save_encoders) + save_encoder_crtcs = kzalloc(dev->mode_config.num_encoder * + sizeof(struct drm_crtc *), GFP_KERNEL); + if (!save_encoder_crtcs) return -ENOMEM; - save_connectors = kzalloc(dev->mode_config.num_connector * - sizeof(struct drm_connector), GFP_KERNEL); - if (!save_connectors) { - kfree(save_encoders); + save_connector_encoders = kzalloc(dev->mode_config.num_connector * + sizeof(struct drm_encoder *), GFP_KERNEL); + if (!save_connector_encoders) { + kfree(save_encoder_crtcs); return -ENOMEM; } @@ -596,12 +596,12 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) */ count = 0; drm_for_each_encoder(encoder, dev) { - save_encoders[count++] = *encoder; + save_encoder_crtcs[count++] = encoder->crtc; } count = 0; drm_for_each_connector(connector, dev) { - save_connectors[count++] = *connector; + save_connector_encoders[count++] = connector->encoder; } save_set.crtc = set->crtc; @@ -634,8 +634,12 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) mode_changed = true; } - /* take a reference on all connectors in set */ + /* take a reference on all unbound connectors in set, reuse the + * already taken reference for bound connectors + */ for (ro = 0; ro < set->num_connectors; ro++) { + if (set->connectors[ro]->encoder) + continue; drm_connector_reference(set->connectors[ro]); } @@ -757,30 +761,28 @@ int drm_crtc_helper_set_config(struct drm_mode_set *set) } } - /* after fail drop reference on all connectors in save set */ - count = 0; - drm_for_each_connector(connector, dev) { - drm_connector_unreference(&save_connectors[count++]); - } - - kfree(save_connectors); - kfree(save_encoders); + kfree(save_connector_encoders); + kfree(save_encoder_crtcs); return 0; fail: /* Restore all previous data. */ count = 0; drm_for_each_encoder(encoder, dev) { - *encoder = save_encoders[count++]; + encoder->crtc = save_encoder_crtcs[count++]; } count = 0; drm_for_each_connector(connector, dev) { - *connector = save_connectors[count++]; + connector->encoder = save_connector_encoders[count++]; } - /* after fail drop reference on all connectors in set */ + /* after fail drop reference on all unbound connectors in set, let + * bound connectors keep their reference + */ for (ro = 0; ro < set->num_connectors; ro++) { + if (set->connectors[ro]->encoder) + continue; drm_connector_unreference(set->connectors[ro]); } @@ -790,8 +792,8 @@ fail: save_set.y, save_set.fb)) DRM_ERROR("failed to restore config after modeset failure\n"); - kfree(save_connectors); - kfree(save_encoders); + kfree(save_connector_encoders); + kfree(save_encoder_crtcs); return ret; } EXPORT_SYMBOL(drm_crtc_helper_set_config); diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c index a13edf5de2d6..6537908050d7 100644 --- a/drivers/gpu/drm/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/drm_dp_mst_topology.c @@ -2927,11 +2927,9 @@ static void drm_dp_destroy_connector_work(struct work_struct *work) drm_dp_port_teardown_pdt(port, port->pdt); if (!port->input && port->vcpi.vcpi > 0) { - if (mgr->mst_state) { - drm_dp_mst_reset_vcpi_slots(mgr, port); - drm_dp_update_payload_part1(mgr); - drm_dp_mst_put_payload_id(mgr, port->vcpi.vcpi); - } + drm_dp_mst_reset_vcpi_slots(mgr, port); + drm_dp_update_payload_part1(mgr); + drm_dp_mst_put_payload_id(mgr, port->vcpi.vcpi); } kref_put(&port->kref, drm_dp_free_mst_port); diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c index c0b0c718994a..1fd6eac1400c 100644 --- a/drivers/gpu/drm/drm_fb_cma_helper.c +++ b/drivers/gpu/drm/drm_fb_cma_helper.c @@ -596,3 +596,18 @@ void drm_fbdev_cma_hotplug_event(struct drm_fbdev_cma *fbdev_cma) drm_fb_helper_hotplug_event(&fbdev_cma->fb_helper); } EXPORT_SYMBOL_GPL(drm_fbdev_cma_hotplug_event); + +/** + * drm_fbdev_cma_set_suspend - wrapper around drm_fb_helper_set_suspend + * @fbdev_cma: The drm_fbdev_cma struct, may be NULL + * @state: desired state, zero to resume, non-zero to suspend + * + * Calls drm_fb_helper_set_suspend, which is a wrapper around + * fb_set_suspend implemented by fbdev core. + */ +void drm_fbdev_cma_set_suspend(struct drm_fbdev_cma *fbdev_cma, int state) +{ + if (fbdev_cma) + drm_fb_helper_set_suspend(&fbdev_cma->fb_helper, state); +} +EXPORT_SYMBOL(drm_fbdev_cma_set_suspend); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c index 340d390306d8..ffd1b32caa8d 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c @@ -91,10 +91,8 @@ static void load_gpu(struct drm_device *dev) int ret; ret = etnaviv_gpu_init(g); - if (ret) { - dev_err(g->dev, "hw init failed: %d\n", ret); + if (ret) priv->gpu[i] = NULL; - } } } } diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index ff6aa5dfb2d7..87ef34150d46 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -487,6 +487,47 @@ static int etnaviv_hw_reset(struct etnaviv_gpu *gpu) return 0; } +static void etnaviv_gpu_enable_mlcg(struct etnaviv_gpu *gpu) +{ + u32 pmc, ppc; + + /* enable clock gating */ + ppc = gpu_read(gpu, VIVS_PM_POWER_CONTROLS); + ppc |= VIVS_PM_POWER_CONTROLS_ENABLE_MODULE_CLOCK_GATING; + + /* Disable stall module clock gating for 4.3.0.1 and 4.3.0.2 revs */ + if (gpu->identity.revision == 0x4301 || + gpu->identity.revision == 0x4302) + ppc |= VIVS_PM_POWER_CONTROLS_DISABLE_STALL_MODULE_CLOCK_GATING; + + gpu_write(gpu, VIVS_PM_POWER_CONTROLS, ppc); + + pmc = gpu_read(gpu, VIVS_PM_MODULE_CONTROLS); + + /* Disable PA clock gating for GC400+ except for GC420 */ + if (gpu->identity.model >= chipModel_GC400 && + gpu->identity.model != chipModel_GC420) + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_PA; + + /* + * Disable PE clock gating on revs < 5.0.0.0 when HZ is + * present without a bug fix. + */ + if (gpu->identity.revision < 0x5000 && + gpu->identity.minor_features0 & chipMinorFeatures0_HZ && + !(gpu->identity.minor_features1 & + chipMinorFeatures1_DISABLE_PE_GATING)) + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_PE; + + if (gpu->identity.revision < 0x5422) + pmc |= BIT(15); /* Unknown bit */ + + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_HZ; + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_EZ; + + gpu_write(gpu, VIVS_PM_MODULE_CONTROLS, pmc); +} + static void etnaviv_gpu_hw_init(struct etnaviv_gpu *gpu) { u16 prefetch; @@ -506,6 +547,9 @@ static void etnaviv_gpu_hw_init(struct etnaviv_gpu *gpu) gpu_write(gpu, VIVS_MC_DEBUG_MEMORY, mc_memory_debug); } + /* enable module-level clock gating */ + etnaviv_gpu_enable_mlcg(gpu); + /* * Update GPU AXI cache atttribute to "cacheable, no allocate". * This is necessary to prevent the iMX6 SoC locking up. @@ -553,8 +597,10 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu) bool mmuv2; ret = pm_runtime_get_sync(gpu->dev); - if (ret < 0) + if (ret < 0) { + dev_err(gpu->dev, "Failed to enable GPU power domain\n"); return ret; + } etnaviv_hw_identify(gpu); @@ -591,8 +637,10 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu) } ret = etnaviv_hw_reset(gpu); - if (ret) + if (ret) { + dev_err(gpu->dev, "GPU reset failed\n"); goto fail; + } /* Setup IOMMU.. eventually we will (I think) do this once per context * and have separate page tables per context. For now, to keep things @@ -610,12 +658,14 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu) } if (!iommu) { + dev_err(gpu->dev, "Failed to allocate GPU IOMMU domain\n"); ret = -ENOMEM; goto fail; } gpu->mmu = etnaviv_iommu_new(gpu, iommu, version); if (!gpu->mmu) { + dev_err(gpu->dev, "Failed to instantiate GPU IOMMU\n"); iommu_domain_free(iommu); ret = -ENOMEM; goto fail; diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c index 522cfd447892..16353ee81651 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c @@ -225,6 +225,7 @@ struct iommu_domain *etnaviv_iommu_domain_alloc(struct etnaviv_gpu *gpu) etnaviv_domain->domain.type = __IOMMU_DOMAIN_PAGING; etnaviv_domain->domain.ops = &etnaviv_iommu_ops.ops; + etnaviv_domain->domain.pgsize_bitmap = SZ_4K; etnaviv_domain->domain.geometry.aperture_start = GPU_MEM_START; etnaviv_domain->domain.geometry.aperture_end = GPU_MEM_START + PT_ENTRIES * SZ_4K - 1; diff --git a/drivers/gpu/drm/etnaviv/state_hi.xml.h b/drivers/gpu/drm/etnaviv/state_hi.xml.h index 6a7de5f1454a..807a3d9e0dd5 100644 --- a/drivers/gpu/drm/etnaviv/state_hi.xml.h +++ b/drivers/gpu/drm/etnaviv/state_hi.xml.h @@ -218,6 +218,13 @@ Copyright (C) 2015 #define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_FE 0x00000001 #define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_DE 0x00000002 #define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_PE 0x00000004 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_SH 0x00000008 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_PA 0x00000010 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_SE 0x00000020 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA 0x00000040 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_TX 0x00000080 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_EZ 0x00010000 +#define VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_HZ 0x00020000 #define VIVS_PM_MODULE_STATUS 0x00000108 #define VIVS_PM_MODULE_STATUS_MODULE_CLOCK_GATED_FE 0x00000001 diff --git a/drivers/gpu/drm/exynos/exynos7_drm_decon.c b/drivers/gpu/drm/exynos/exynos7_drm_decon.c index f6223f907c15..7f9901b7777b 100644 --- a/drivers/gpu/drm/exynos/exynos7_drm_decon.c +++ b/drivers/gpu/drm/exynos/exynos7_drm_decon.c @@ -31,7 +31,6 @@ #include "exynos_drm_plane.h" #include "exynos_drm_drv.h" #include "exynos_drm_fb.h" -#include "exynos_drm_fbdev.h" #include "exynos_drm_iommu.h" /* diff --git a/drivers/gpu/drm/exynos/exynos_dp.c b/drivers/gpu/drm/exynos/exynos_dp.c index 468498e3fec1..4c1fb3f8b5a6 100644 --- a/drivers/gpu/drm/exynos/exynos_dp.c +++ b/drivers/gpu/drm/exynos/exynos_dp.c @@ -34,7 +34,7 @@ struct exynos_dp_device { struct drm_encoder encoder; - struct drm_connector connector; + struct drm_connector *connector; struct drm_bridge *ptn_bridge; struct drm_device *drm_dev; struct device *dev; @@ -70,7 +70,7 @@ static int exynos_dp_poweroff(struct analogix_dp_plat_data *plat_data) static int exynos_dp_get_modes(struct analogix_dp_plat_data *plat_data) { struct exynos_dp_device *dp = to_dp(plat_data); - struct drm_connector *connector = &dp->connector; + struct drm_connector *connector = dp->connector; struct drm_display_mode *mode; int num_modes = 0; @@ -103,6 +103,7 @@ static int exynos_dp_bridge_attach(struct analogix_dp_plat_data *plat_data, int ret; drm_connector_register(connector); + dp->connector = connector; /* Pre-empt DP connector creation if there's a bridge */ if (dp->ptn_bridge) { diff --git a/drivers/gpu/drm/exynos/exynos_drm_core.c b/drivers/gpu/drm/exynos/exynos_drm_core.c index 011211e4167d..edbd98ff293e 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_core.c +++ b/drivers/gpu/drm/exynos/exynos_drm_core.c @@ -15,7 +15,6 @@ #include <drm/drmP.h> #include "exynos_drm_drv.h" #include "exynos_drm_crtc.h" -#include "exynos_drm_fbdev.h" static LIST_HEAD(exynos_drm_subdrv_list); diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c index 3efe1aa89416..d47216488985 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c @@ -30,7 +30,6 @@ #include "exynos_drm_drv.h" #include "exynos_drm_fb.h" -#include "exynos_drm_fbdev.h" #include "exynos_drm_crtc.h" #include "exynos_drm_plane.h" #include "exynos_drm_iommu.h" @@ -120,7 +119,6 @@ static struct fimd_driver_data s3c64xx_fimd_driver_data = { .timing_base = 0x0, .has_clksel = 1, .has_limited_fmt = 1, - .has_hw_trigger = 1, }; static struct fimd_driver_data exynos3_fimd_driver_data = { @@ -171,14 +169,11 @@ static struct fimd_driver_data exynos5420_fimd_driver_data = { .lcdblk_vt_shift = 24, .lcdblk_bypass_shift = 15, .lcdblk_mic_bypass_shift = 11, - .trg_type = I80_HW_TRG, .has_shadowcon = 1, .has_vidoutcon = 1, .has_vtsel = 1, .has_mic_bypass = 1, .has_dp_clk = 1, - .has_hw_trigger = 1, - .has_trigger_per_te = 1, }; struct fimd_context { diff --git a/drivers/gpu/drm/exynos/exynos_drm_g2d.c b/drivers/gpu/drm/exynos/exynos_drm_g2d.c index 493552368295..8564c3da0d22 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_g2d.c +++ b/drivers/gpu/drm/exynos/exynos_drm_g2d.c @@ -48,13 +48,13 @@ /* registers for base address */ #define G2D_SRC_BASE_ADDR 0x0304 -#define G2D_SRC_STRIDE_REG 0x0308 +#define G2D_SRC_STRIDE 0x0308 #define G2D_SRC_COLOR_MODE 0x030C #define G2D_SRC_LEFT_TOP 0x0310 #define G2D_SRC_RIGHT_BOTTOM 0x0314 #define G2D_SRC_PLANE2_BASE_ADDR 0x0318 #define G2D_DST_BASE_ADDR 0x0404 -#define G2D_DST_STRIDE_REG 0x0408 +#define G2D_DST_STRIDE 0x0408 #define G2D_DST_COLOR_MODE 0x040C #define G2D_DST_LEFT_TOP 0x0410 #define G2D_DST_RIGHT_BOTTOM 0x0414 @@ -563,7 +563,7 @@ static enum g2d_reg_type g2d_get_reg_type(int reg_offset) switch (reg_offset) { case G2D_SRC_BASE_ADDR: - case G2D_SRC_STRIDE_REG: + case G2D_SRC_STRIDE: case G2D_SRC_COLOR_MODE: case G2D_SRC_LEFT_TOP: case G2D_SRC_RIGHT_BOTTOM: @@ -573,7 +573,7 @@ static enum g2d_reg_type g2d_get_reg_type(int reg_offset) reg_type = REG_TYPE_SRC_PLANE2; break; case G2D_DST_BASE_ADDR: - case G2D_DST_STRIDE_REG: + case G2D_DST_STRIDE: case G2D_DST_COLOR_MODE: case G2D_DST_LEFT_TOP: case G2D_DST_RIGHT_BOTTOM: @@ -968,8 +968,8 @@ static int g2d_check_reg_offset(struct device *dev, } else buf_info->types[reg_type] = BUF_TYPE_GEM; break; - case G2D_SRC_STRIDE_REG: - case G2D_DST_STRIDE_REG: + case G2D_SRC_STRIDE: + case G2D_DST_STRIDE: if (for_addr) goto err; diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_crtc.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_crtc.c index 706de3278f1c..3371635cd4d7 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_crtc.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_crtc.c @@ -44,6 +44,8 @@ static void fsl_dcu_drm_disable_crtc(struct drm_crtc *crtc) struct drm_device *dev = crtc->dev; struct fsl_dcu_drm_device *fsl_dev = dev->dev_private; + drm_crtc_vblank_off(crtc); + regmap_update_bits(fsl_dev->regmap, DCU_DCU_MODE, DCU_MODE_DCU_MODE_MASK, DCU_MODE_DCU_MODE(DCU_MODE_OFF)); @@ -61,6 +63,8 @@ static void fsl_dcu_drm_crtc_enable(struct drm_crtc *crtc) DCU_MODE_DCU_MODE(DCU_MODE_NORMAL)); regmap_write(fsl_dev->regmap, DCU_UPDATE_MODE, DCU_UPDATE_MODE_READREG); + + drm_crtc_vblank_on(crtc); } static void fsl_dcu_drm_crtc_mode_set_nofb(struct drm_crtc *crtc) @@ -137,9 +141,10 @@ int fsl_dcu_drm_crtc_create(struct fsl_dcu_drm_device *fsl_dev) { struct drm_plane *primary; struct drm_crtc *crtc = &fsl_dev->crtc; - unsigned int i, j, reg_num; int ret; + fsl_dcu_drm_init_planes(fsl_dev->drm); + primary = fsl_dcu_drm_primary_create_plane(fsl_dev->drm); if (!primary) return -ENOMEM; @@ -153,19 +158,5 @@ int fsl_dcu_drm_crtc_create(struct fsl_dcu_drm_device *fsl_dev) drm_crtc_helper_add(crtc, &fsl_dcu_drm_crtc_helper_funcs); - if (!strcmp(fsl_dev->soc->name, "ls1021a")) - reg_num = LS1021A_LAYER_REG_NUM; - else - reg_num = VF610_LAYER_REG_NUM; - for (i = 0; i < fsl_dev->soc->total_layer; i++) { - for (j = 1; j <= reg_num; j++) - regmap_write(fsl_dev->regmap, DCU_CTRLDESCLN(i, j), 0); - } - regmap_update_bits(fsl_dev->regmap, DCU_DCU_MODE, - DCU_MODE_DCU_MODE_MASK, - DCU_MODE_DCU_MODE(DCU_MODE_OFF)); - regmap_write(fsl_dev->regmap, DCU_UPDATE_MODE, - DCU_UPDATE_MODE_READREG); - return 0; } diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c index 33727d5d826a..7882387f9bff 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c @@ -11,6 +11,7 @@ #include <linux/clk.h> #include <linux/clk-provider.h> +#include <linux/console.h> #include <linux/io.h> #include <linux/mfd/syscon.h> #include <linux/mm.h> @@ -22,6 +23,7 @@ #include <linux/regmap.h> #include <drm/drmP.h> +#include <drm/drm_atomic_helper.h> #include <drm/drm_crtc_helper.h> #include <drm/drm_fb_cma_helper.h> #include <drm/drm_gem_cma_helper.h> @@ -42,7 +44,6 @@ static const struct regmap_config fsl_dcu_regmap_config = { .reg_bits = 32, .reg_stride = 4, .val_bits = 32, - .cache_type = REGCACHE_RBTREE, .volatile_reg = fsl_dcu_drm_is_volatile_reg, }; @@ -228,11 +229,26 @@ static int fsl_dcu_drm_pm_suspend(struct device *dev) if (!fsl_dev) return 0; + disable_irq(fsl_dev->irq); drm_kms_helper_poll_disable(fsl_dev->drm); - regcache_cache_only(fsl_dev->regmap, true); - regcache_mark_dirty(fsl_dev->regmap); - clk_disable(fsl_dev->clk); - clk_unprepare(fsl_dev->clk); + + console_lock(); + drm_fbdev_cma_set_suspend(fsl_dev->fbdev, 1); + console_unlock(); + + fsl_dev->state = drm_atomic_helper_suspend(fsl_dev->drm); + if (IS_ERR(fsl_dev->state)) { + console_lock(); + drm_fbdev_cma_set_suspend(fsl_dev->fbdev, 0); + console_unlock(); + + drm_kms_helper_poll_enable(fsl_dev->drm); + enable_irq(fsl_dev->irq); + return PTR_ERR(fsl_dev->state); + } + + clk_disable_unprepare(fsl_dev->pix_clk); + clk_disable_unprepare(fsl_dev->clk); return 0; } @@ -245,21 +261,27 @@ static int fsl_dcu_drm_pm_resume(struct device *dev) if (!fsl_dev) return 0; - ret = clk_enable(fsl_dev->clk); + ret = clk_prepare_enable(fsl_dev->clk); if (ret < 0) { dev_err(dev, "failed to enable dcu clk\n"); - clk_unprepare(fsl_dev->clk); return ret; } - ret = clk_prepare(fsl_dev->clk); + + ret = clk_prepare_enable(fsl_dev->pix_clk); if (ret < 0) { - dev_err(dev, "failed to prepare dcu clk\n"); + dev_err(dev, "failed to enable pix clk\n"); return ret; } + fsl_dcu_drm_init_planes(fsl_dev->drm); + drm_atomic_helper_resume(fsl_dev->drm, fsl_dev->state); + + console_lock(); + drm_fbdev_cma_set_suspend(fsl_dev->fbdev, 0); + console_unlock(); + drm_kms_helper_poll_enable(fsl_dev->drm); - regcache_cache_only(fsl_dev->regmap, false); - regcache_sync(fsl_dev->regmap); + enable_irq(fsl_dev->irq); return 0; } @@ -273,12 +295,14 @@ static const struct fsl_dcu_soc_data fsl_dcu_ls1021a_data = { .name = "ls1021a", .total_layer = 16, .max_layer = 4, + .layer_regs = LS1021A_LAYER_REG_NUM, }; static const struct fsl_dcu_soc_data fsl_dcu_vf610_data = { .name = "vf610", .total_layer = 64, .max_layer = 6, + .layer_regs = VF610_LAYER_REG_NUM, }; static const struct of_device_id fsl_dcu_of_match[] = { diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.h b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.h index c275f900ff23..3b371fe7491e 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.h +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.h @@ -175,6 +175,7 @@ struct fsl_dcu_soc_data { unsigned int total_layer; /*max layer number DCU supported*/ unsigned int max_layer; + unsigned int layer_regs; }; struct fsl_dcu_drm_device { @@ -193,6 +194,7 @@ struct fsl_dcu_drm_device { struct drm_encoder encoder; struct fsl_dcu_drm_connector connector; const struct fsl_dcu_soc_data *soc; + struct drm_atomic_state *state; }; void fsl_dcu_fbdev_init(struct drm_device *dev); diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c index 274558b3b32b..e50467a0deb0 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.c @@ -217,6 +217,22 @@ static const u32 fsl_dcu_drm_plane_formats[] = { DRM_FORMAT_YUV422, }; +void fsl_dcu_drm_init_planes(struct drm_device *dev) +{ + struct fsl_dcu_drm_device *fsl_dev = dev->dev_private; + int i, j; + + for (i = 0; i < fsl_dev->soc->total_layer; i++) { + for (j = 1; j <= fsl_dev->soc->layer_regs; j++) + regmap_write(fsl_dev->regmap, DCU_CTRLDESCLN(i, j), 0); + } + regmap_update_bits(fsl_dev->regmap, DCU_DCU_MODE, + DCU_MODE_DCU_MODE_MASK, + DCU_MODE_DCU_MODE(DCU_MODE_OFF)); + regmap_write(fsl_dev->regmap, DCU_UPDATE_MODE, + DCU_UPDATE_MODE_READREG); +} + struct drm_plane *fsl_dcu_drm_primary_create_plane(struct drm_device *dev) { struct drm_plane *primary; diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.h b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.h index d657f088d859..8ee45f813ee8 100644 --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.h +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_plane.h @@ -12,6 +12,7 @@ #ifndef __FSL_DCU_DRM_PLANE_H__ #define __FSL_DCU_DRM_PLANE_H__ +void fsl_dcu_drm_init_planes(struct drm_device *dev); struct drm_plane *fsl_dcu_drm_primary_create_plane(struct drm_device *dev); #endif /* __FSL_DCU_DRM_PLANE_H__ */ diff --git a/drivers/gpu/drm/hisilicon/kirin/Kconfig b/drivers/gpu/drm/hisilicon/kirin/Kconfig index ea0df6115f7e..499f64405dac 100644 --- a/drivers/gpu/drm/hisilicon/kirin/Kconfig +++ b/drivers/gpu/drm/hisilicon/kirin/Kconfig @@ -4,6 +4,7 @@ config DRM_HISI_KIRIN select DRM_KMS_HELPER select DRM_GEM_CMA_HELPER select DRM_KMS_CMA_HELPER + select HISI_KIRIN_DW_DSI help Choose this option if you have a hisilicon Kirin chipsets(hi6220). If M is selected the module will be called kirin-drm. diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c index 805f4326fafe..c3707d47cd89 100644 --- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c +++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c @@ -967,21 +967,21 @@ static int ade_dts_parse(struct platform_device *pdev, struct ade_hw_ctx *ctx) } ctx->ade_core_clk = devm_clk_get(dev, "clk_ade_core"); - if (!ctx->ade_core_clk) { + if (IS_ERR(ctx->ade_core_clk)) { DRM_ERROR("failed to parse clk ADE_CORE\n"); - return -ENODEV; + return PTR_ERR(ctx->ade_core_clk); } ctx->media_noc_clk = devm_clk_get(dev, "clk_codec_jpeg"); - if (!ctx->media_noc_clk) { + if (IS_ERR(ctx->media_noc_clk)) { DRM_ERROR("failed to parse clk CODEC_JPEG\n"); - return -ENODEV; + return PTR_ERR(ctx->media_noc_clk); } ctx->ade_pix_clk = devm_clk_get(dev, "clk_ade_pix"); - if (!ctx->ade_pix_clk) { + if (IS_ERR(ctx->ade_pix_clk)) { DRM_ERROR("failed to parse clk ADE_PIX\n"); - return -ENODEV; + return PTR_ERR(ctx->ade_pix_clk); } return 0; diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 0b84f8e5df50..ffa43eca14d3 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4622,9 +4622,6 @@ intel_dp_hpd_pulse(struct intel_digital_port *intel_dig_port, bool long_hpd) intel_display_power_get(dev_priv, power_domain); if (long_hpd) { - /* indicate that we need to restart link training */ - intel_dp->train_set_valid = false; - intel_dp_long_pulse(intel_dp->attached_connector); if (intel_dp->is_mst) ret = IRQ_HANDLED; diff --git a/drivers/gpu/drm/i915/intel_dp_link_training.c b/drivers/gpu/drm/i915/intel_dp_link_training.c index 0b8eefc2acc5..60fb39cd220b 100644 --- a/drivers/gpu/drm/i915/intel_dp_link_training.c +++ b/drivers/gpu/drm/i915/intel_dp_link_training.c @@ -85,8 +85,7 @@ static bool intel_dp_reset_link_train(struct intel_dp *intel_dp, uint8_t dp_train_pat) { - if (!intel_dp->train_set_valid) - memset(intel_dp->train_set, 0, sizeof(intel_dp->train_set)); + memset(intel_dp->train_set, 0, sizeof(intel_dp->train_set)); intel_dp_set_signal_levels(intel_dp); return intel_dp_set_link_train(intel_dp, dp_train_pat); } @@ -161,23 +160,6 @@ intel_dp_link_training_clock_recovery(struct intel_dp *intel_dp) break; } - /* - * if we used previously trained voltage and pre-emphasis values - * and we don't get clock recovery, reset link training values - */ - if (intel_dp->train_set_valid) { - DRM_DEBUG_KMS("clock recovery not ok, reset"); - /* clear the flag as we are not reusing train set */ - intel_dp->train_set_valid = false; - if (!intel_dp_reset_link_train(intel_dp, - DP_TRAINING_PATTERN_1 | - DP_LINK_SCRAMBLING_DISABLE)) { - DRM_ERROR("failed to enable link training\n"); - return; - } - continue; - } - /* Check to see if we've tried the max voltage */ for (i = 0; i < intel_dp->lane_count; i++) if ((intel_dp->train_set[i] & DP_TRAIN_MAX_SWING_REACHED) == 0) @@ -284,7 +266,6 @@ intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) /* Make sure clock is still ok */ if (!drm_dp_clock_recovery_ok(link_status, intel_dp->lane_count)) { - intel_dp->train_set_valid = false; intel_dp_link_training_clock_recovery(intel_dp); intel_dp_set_link_train(intel_dp, training_pattern | @@ -301,7 +282,6 @@ intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) /* Try 5 times, then try clock recovery if that fails */ if (tries > 5) { - intel_dp->train_set_valid = false; intel_dp_link_training_clock_recovery(intel_dp); intel_dp_set_link_train(intel_dp, training_pattern | @@ -322,10 +302,8 @@ intel_dp_link_training_channel_equalization(struct intel_dp *intel_dp) intel_dp_set_idle_link_train(intel_dp); - if (channel_eq) { - intel_dp->train_set_valid = true; + if (channel_eq) DRM_DEBUG_KMS("Channel EQ done. DP Training successful\n"); - } } void intel_dp_stop_link_train(struct intel_dp *intel_dp) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 7d0e071fe355..089a42577ea3 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -886,8 +886,6 @@ struct intel_dp { /* This is called before a link training is starterd */ void (*prepare_link_retrain)(struct intel_dp *intel_dp); - bool train_set_valid; - /* Displayport compliance testing */ unsigned long compliance_test_type; unsigned long compliance_test_data; diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c index ecabd59ffbaf..a19944b6dc25 100644 --- a/drivers/gpu/drm/i915/intel_fbc.c +++ b/drivers/gpu/drm/i915/intel_fbc.c @@ -818,8 +818,7 @@ static bool intel_fbc_can_choose(struct intel_crtc *crtc) { struct drm_i915_private *dev_priv = crtc->base.dev->dev_private; struct intel_fbc *fbc = &dev_priv->fbc; - bool enable_by_default = IS_HASWELL(dev_priv) || - IS_BROADWELL(dev_priv); + bool enable_by_default = IS_BROADWELL(dev_priv); if (intel_vgpu_active(dev_priv)) { fbc->no_fbc_reason = "VGPU is active"; diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c index ba812ef2c9d1..334562d06731 100644 --- a/drivers/gpu/drm/mediatek/mtk_hdmi.c +++ b/drivers/gpu/drm/mediatek/mtk_hdmi.c @@ -1535,7 +1535,7 @@ static int mtk_hdmi_dt_parse_pdata(struct mtk_hdmi *hdmi, * HDMI audio codec callbacks */ -static int mtk_hdmi_audio_hw_params(struct device *dev, +static int mtk_hdmi_audio_hw_params(struct device *dev, void *data, struct hdmi_codec_daifmt *daifmt, struct hdmi_codec_params *params) { @@ -1604,7 +1604,7 @@ static int mtk_hdmi_audio_hw_params(struct device *dev, return 0; } -static int mtk_hdmi_audio_startup(struct device *dev) +static int mtk_hdmi_audio_startup(struct device *dev, void *data) { struct mtk_hdmi *hdmi = dev_get_drvdata(dev); @@ -1615,7 +1615,7 @@ static int mtk_hdmi_audio_startup(struct device *dev) return 0; } -static void mtk_hdmi_audio_shutdown(struct device *dev) +static void mtk_hdmi_audio_shutdown(struct device *dev, void *data) { struct mtk_hdmi *hdmi = dev_get_drvdata(dev); @@ -1624,7 +1624,7 @@ static void mtk_hdmi_audio_shutdown(struct device *dev) mtk_hdmi_audio_disable(hdmi); } -int mtk_hdmi_audio_digital_mute(struct device *dev, bool enable) +int mtk_hdmi_audio_digital_mute(struct device *dev, void *data, bool enable) { struct mtk_hdmi *hdmi = dev_get_drvdata(dev); @@ -1638,7 +1638,7 @@ int mtk_hdmi_audio_digital_mute(struct device *dev, bool enable) return 0; } -static int mtk_hdmi_audio_get_eld(struct device *dev, uint8_t *buf, size_t len) +static int mtk_hdmi_audio_get_eld(struct device *dev, void *data, uint8_t *buf, size_t len) { struct mtk_hdmi *hdmi = dev_get_drvdata(dev); diff --git a/drivers/gpu/drm/mgag200/mgag200_ttm.c b/drivers/gpu/drm/mgag200/mgag200_ttm.c index 9d5083d0f1ee..68268e55d595 100644 --- a/drivers/gpu/drm/mgag200/mgag200_ttm.c +++ b/drivers/gpu/drm/mgag200/mgag200_ttm.c @@ -186,17 +186,6 @@ static void mgag200_ttm_io_mem_free(struct ttm_bo_device *bdev, struct ttm_mem_r { } -static int mgag200_bo_move(struct ttm_buffer_object *bo, - bool evict, bool interruptible, - bool no_wait_gpu, - struct ttm_mem_reg *new_mem) -{ - int r; - r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); - return r; -} - - static void mgag200_ttm_backend_destroy(struct ttm_tt *tt) { ttm_tt_fini(tt); @@ -241,7 +230,7 @@ struct ttm_bo_driver mgag200_bo_driver = { .ttm_tt_unpopulate = mgag200_ttm_tt_unpopulate, .init_mem_type = mgag200_bo_init_mem_type, .evict_flags = mgag200_bo_evict_flags, - .move = mgag200_bo_move, + .move = NULL, .verify_access = mgag200_bo_verify_access, .io_mem_reserve = &mgag200_ttm_io_mem_reserve, .io_mem_free = &mgag200_ttm_io_mem_free, diff --git a/drivers/gpu/drm/msm/adreno/adreno_gpu.c b/drivers/gpu/drm/msm/adreno/adreno_gpu.c index fbe304ee6c80..2aec27dbb5bb 100644 --- a/drivers/gpu/drm/msm/adreno/adreno_gpu.c +++ b/drivers/gpu/drm/msm/adreno/adreno_gpu.c @@ -408,7 +408,7 @@ int adreno_gpu_init(struct drm_device *drm, struct platform_device *pdev, } adreno_gpu->memptrs = msm_gem_vaddr(adreno_gpu->memptrs_bo); - if (!adreno_gpu->memptrs) { + if (IS_ERR(adreno_gpu->memptrs)) { dev_err(drm->dev, "could not vmap memptrs\n"); return -ENOMEM; } diff --git a/drivers/gpu/drm/msm/msm_fbdev.c b/drivers/gpu/drm/msm/msm_fbdev.c index 1a061e3e8b9e..a9223bea871b 100644 --- a/drivers/gpu/drm/msm/msm_fbdev.c +++ b/drivers/gpu/drm/msm/msm_fbdev.c @@ -159,6 +159,10 @@ static int msm_fbdev_create(struct drm_fb_helper *helper, dev->mode_config.fb_base = paddr; fbi->screen_base = msm_gem_vaddr_locked(fbdev->bo); + if (IS_ERR(fbi->screen_base)) { + ret = PTR_ERR(fbi->screen_base); + goto fail_unlock; + } fbi->screen_size = fbdev->bo->size; fbi->fix.smem_start = paddr; fbi->fix.smem_len = fbdev->bo->size; diff --git a/drivers/gpu/drm/msm/msm_gem.c b/drivers/gpu/drm/msm/msm_gem.c index 7daf4054dd2b..69836f5685b1 100644 --- a/drivers/gpu/drm/msm/msm_gem.c +++ b/drivers/gpu/drm/msm/msm_gem.c @@ -398,6 +398,8 @@ void *msm_gem_vaddr_locked(struct drm_gem_object *obj) return ERR_CAST(pages); msm_obj->vaddr = vmap(pages, obj->size >> PAGE_SHIFT, VM_MAP, pgprot_writecombine(PAGE_KERNEL)); + if (msm_obj->vaddr == NULL) + return ERR_PTR(-ENOMEM); } return msm_obj->vaddr; } diff --git a/drivers/gpu/drm/msm/msm_gem_submit.c b/drivers/gpu/drm/msm/msm_gem_submit.c index b89ca5174863..eb4bb8b2f3a5 100644 --- a/drivers/gpu/drm/msm/msm_gem_submit.c +++ b/drivers/gpu/drm/msm/msm_gem_submit.c @@ -40,12 +40,14 @@ static struct msm_gem_submit *submit_create(struct drm_device *dev, submit->dev = dev; submit->gpu = gpu; + submit->fence = NULL; submit->pid = get_pid(task_pid(current)); /* initially, until copy_from_user() and bo lookup succeeds: */ submit->nr_bos = 0; submit->nr_cmds = 0; + INIT_LIST_HEAD(&submit->node); INIT_LIST_HEAD(&submit->bo_list); ww_acquire_init(&submit->ticket, &reservation_ww_class); @@ -75,6 +77,11 @@ static int submit_lookup_objects(struct msm_gem_submit *submit, void __user *userptr = u64_to_user_ptr(args->bos + (i * sizeof(submit_bo))); + /* make sure we don't have garbage flags, in case we hit + * error path before flags is initialized: + */ + submit->bos[i].flags = 0; + ret = copy_from_user(&submit_bo, userptr, sizeof(submit_bo)); if (ret) { ret = -EFAULT; diff --git a/drivers/gpu/drm/msm/msm_rd.c b/drivers/gpu/drm/msm/msm_rd.c index b48f73ac6389..0857710c2ff2 100644 --- a/drivers/gpu/drm/msm/msm_rd.c +++ b/drivers/gpu/drm/msm/msm_rd.c @@ -312,6 +312,9 @@ void msm_rd_dump_submit(struct msm_gem_submit *submit) struct msm_gem_object *obj = submit->bos[idx].obj; const char *buf = msm_gem_vaddr_locked(&obj->base); + if (IS_ERR(buf)) + continue; + buf += iova - submit->bos[idx].iova; rd_write_section(rd, RD_GPUADDR, diff --git a/drivers/gpu/drm/msm/msm_ringbuffer.c b/drivers/gpu/drm/msm/msm_ringbuffer.c index 1f14b908b221..42f5359cf988 100644 --- a/drivers/gpu/drm/msm/msm_ringbuffer.c +++ b/drivers/gpu/drm/msm/msm_ringbuffer.c @@ -40,6 +40,10 @@ struct msm_ringbuffer *msm_ringbuffer_new(struct msm_gpu *gpu, int size) } ring->start = msm_gem_vaddr_locked(ring->bo); + if (IS_ERR(ring->start)) { + ret = PTR_ERR(ring->start); + goto fail; + } ring->end = ring->start + (size / 4); ring->cur = ring->start; diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h index 331620a52afa..287a7d6fa480 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h +++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h @@ -29,6 +29,7 @@ struct nv_device_info_v0 { #define NV_DEVICE_INFO_V0_FERMI 0x07 #define NV_DEVICE_INFO_V0_KEPLER 0x08 #define NV_DEVICE_INFO_V0_MAXWELL 0x09 +#define NV_DEVICE_INFO_V0_PASCAL 0x0a __u8 family; __u8 pad06[2]; __u64 ram_size; diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h index 982aad8fa645..e6e9537537cf 100644 --- a/drivers/gpu/drm/nouveau/include/nvif/class.h +++ b/drivers/gpu/drm/nouveau/include/nvif/class.h @@ -39,6 +39,7 @@ #define KEPLER_CHANNEL_GPFIFO_A /* cla06f.h */ 0x0000a06f #define KEPLER_CHANNEL_GPFIFO_B /* cla06f.h */ 0x0000a16f #define MAXWELL_CHANNEL_GPFIFO_A /* cla06f.h */ 0x0000b06f +#define PASCAL_CHANNEL_GPFIFO_A /* cla06f.h */ 0x0000c06f #define NV50_DISP /* cl5070.h */ 0x00005070 #define G82_DISP /* cl5070.h */ 0x00008270 @@ -50,6 +51,8 @@ #define GK110_DISP /* cl5070.h */ 0x00009270 #define GM107_DISP /* cl5070.h */ 0x00009470 #define GM200_DISP /* cl5070.h */ 0x00009570 +#define GP100_DISP /* cl5070.h */ 0x00009770 +#define GP104_DISP /* cl5070.h */ 0x00009870 #define NV31_MPEG 0x00003174 #define G82_MPEG 0x00008274 @@ -86,6 +89,8 @@ #define GK110_DISP_CORE_CHANNEL_DMA /* cl507d.h */ 0x0000927d #define GM107_DISP_CORE_CHANNEL_DMA /* cl507d.h */ 0x0000947d #define GM200_DISP_CORE_CHANNEL_DMA /* cl507d.h */ 0x0000957d +#define GP100_DISP_CORE_CHANNEL_DMA /* cl507d.h */ 0x0000977d +#define GP104_DISP_CORE_CHANNEL_DMA /* cl507d.h */ 0x0000987d #define NV50_DISP_OVERLAY_CHANNEL_DMA /* cl507e.h */ 0x0000507e #define G82_DISP_OVERLAY_CHANNEL_DMA /* cl507e.h */ 0x0000827e @@ -105,6 +110,8 @@ #define MAXWELL_A /* cl9097.h */ 0x0000b097 #define MAXWELL_B /* cl9097.h */ 0x0000b197 +#define PASCAL_A /* cl9097.h */ 0x0000c097 + #define NV74_BSP 0x000074b0 #define GT212_MSVLD 0x000085b1 @@ -128,6 +135,8 @@ #define FERMI_DMA 0x000090b5 #define KEPLER_DMA_COPY_A 0x0000a0b5 #define MAXWELL_DMA_COPY_A 0x0000b0b5 +#define PASCAL_DMA_COPY_A 0x0000c0b5 +#define PASCAL_DMA_COPY_B 0x0000c1b5 #define FERMI_DECOMPRESS 0x000090b8 @@ -137,6 +146,7 @@ #define KEPLER_COMPUTE_B 0x0000a1c0 #define MAXWELL_COMPUTE_A 0x0000b0c0 #define MAXWELL_COMPUTE_B 0x0000b1c0 +#define PASCAL_COMPUTE_A 0x0000c0c0 #define NV74_CIPHER 0x000074c1 #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h index c612dc1f1eb4..7ea8aa7ca408 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h @@ -16,9 +16,9 @@ enum nvkm_devidx { NVKM_SUBDEV_MC, NVKM_SUBDEV_BUS, NVKM_SUBDEV_TIMER, + NVKM_SUBDEV_INSTMEM, NVKM_SUBDEV_FB, NVKM_SUBDEV_LTC, - NVKM_SUBDEV_INSTMEM, NVKM_SUBDEV_MMU, NVKM_SUBDEV_BAR, NVKM_SUBDEV_PMU, @@ -33,7 +33,10 @@ enum nvkm_devidx { NVKM_ENGINE_CE0, NVKM_ENGINE_CE1, NVKM_ENGINE_CE2, - NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE2, + NVKM_ENGINE_CE3, + NVKM_ENGINE_CE4, + NVKM_ENGINE_CE5, + NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE5, NVKM_ENGINE_CIPHER, NVKM_ENGINE_DISP, @@ -50,7 +53,8 @@ enum nvkm_devidx { NVKM_ENGINE_NVENC0, NVKM_ENGINE_NVENC1, - NVKM_ENGINE_NVENC_LAST = NVKM_ENGINE_NVENC1, + NVKM_ENGINE_NVENC2, + NVKM_ENGINE_NVENC_LAST = NVKM_ENGINE_NVENC2, NVKM_ENGINE_NVDEC, NVKM_ENGINE_PM, @@ -102,6 +106,7 @@ struct nvkm_device { NV_C0 = 0xc0, NV_E0 = 0xe0, GM100 = 0x110, + GP100 = 0x130, } card_type; u32 chipset; u8 chiprev; @@ -136,7 +141,7 @@ struct nvkm_device { struct nvkm_volt *volt; struct nvkm_engine *bsp; - struct nvkm_engine *ce[3]; + struct nvkm_engine *ce[6]; struct nvkm_engine *cipher; struct nvkm_disp *disp; struct nvkm_dma *dma; @@ -149,7 +154,7 @@ struct nvkm_device { struct nvkm_engine *mspdec; struct nvkm_engine *msppp; struct nvkm_engine *msvld; - struct nvkm_engine *nvenc[2]; + struct nvkm_engine *nvenc[3]; struct nvkm_engine *nvdec; struct nvkm_pm *pm; struct nvkm_engine *sec; @@ -170,7 +175,6 @@ struct nvkm_device_func { void (*fini)(struct nvkm_device *, bool suspend); resource_size_t (*resource_addr)(struct nvkm_device *, unsigned bar); resource_size_t (*resource_size)(struct nvkm_device *, unsigned bar); - bool cpu_coherent; }; struct nvkm_device_quirk { @@ -206,7 +210,7 @@ struct nvkm_device_chip { int (*volt )(struct nvkm_device *, int idx, struct nvkm_volt **); int (*bsp )(struct nvkm_device *, int idx, struct nvkm_engine **); - int (*ce[3] )(struct nvkm_device *, int idx, struct nvkm_engine **); + int (*ce[6] )(struct nvkm_device *, int idx, struct nvkm_engine **); int (*cipher )(struct nvkm_device *, int idx, struct nvkm_engine **); int (*disp )(struct nvkm_device *, int idx, struct nvkm_disp **); int (*dma )(struct nvkm_device *, int idx, struct nvkm_dma **); @@ -219,7 +223,7 @@ struct nvkm_device_chip { int (*mspdec )(struct nvkm_device *, int idx, struct nvkm_engine **); int (*msppp )(struct nvkm_device *, int idx, struct nvkm_engine **); int (*msvld )(struct nvkm_device *, int idx, struct nvkm_engine **); - int (*nvenc[2])(struct nvkm_device *, int idx, struct nvkm_engine **); + int (*nvenc[3])(struct nvkm_device *, int idx, struct nvkm_engine **); int (*nvdec )(struct nvkm_device *, int idx, struct nvkm_engine **); int (*pm )(struct nvkm_device *, int idx, struct nvkm_pm **); int (*sec )(struct nvkm_device *, int idx, struct nvkm_engine **); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h b/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h index b5370cb56e3c..e5c9b6268dcc 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/core/tegra.h @@ -28,6 +28,7 @@ struct nvkm_device_tegra { } iommu; int gpu_speedo; + int gpu_speedo_id; }; struct nvkm_device_tegra_func { diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h index 594d719ba41e..d3d26a1e215d 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h @@ -7,4 +7,6 @@ int gf100_ce_new(struct nvkm_device *, int, struct nvkm_engine **); int gk104_ce_new(struct nvkm_device *, int, struct nvkm_engine **); int gm107_ce_new(struct nvkm_device *, int, struct nvkm_engine **); int gm200_ce_new(struct nvkm_device *, int, struct nvkm_engine **); +int gp100_ce_new(struct nvkm_device *, int, struct nvkm_engine **); +int gp104_ce_new(struct nvkm_device *, int, struct nvkm_engine **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h index d4fdce27b297..e82049667ce4 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h @@ -32,4 +32,6 @@ int gk104_disp_new(struct nvkm_device *, int, struct nvkm_disp **); int gk110_disp_new(struct nvkm_device *, int, struct nvkm_disp **); int gm107_disp_new(struct nvkm_device *, int, struct nvkm_disp **); int gm200_disp_new(struct nvkm_device *, int, struct nvkm_disp **); +int gp100_disp_new(struct nvkm_device *, int, struct nvkm_disp **); +int gp104_disp_new(struct nvkm_device *, int, struct nvkm_disp **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h index 15ddfcf5e8db..ed92fec5292c 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h @@ -66,4 +66,5 @@ int gk20a_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gm107_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gm200_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); int gm20b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); +int gp100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h index 6515f5810a26..89cf99307828 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h @@ -42,4 +42,5 @@ int gk20a_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gm107_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gm200_gr_new(struct nvkm_device *, int, struct nvkm_gr **); int gm20b_gr_new(struct nvkm_device *, int, struct nvkm_gr **); +int gp100_gr_new(struct nvkm_device *, int, struct nvkm_gr **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios.h index e39a1fea930b..a72f3290528a 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios.h @@ -7,6 +7,9 @@ struct nvkm_bios { u32 size; u8 *data; + u32 image0_size; + u32 imaged_addr; + u32 bmp_offset; u32 bit_offset; @@ -22,10 +25,9 @@ struct nvkm_bios { u8 nvbios_checksum(const u8 *data, int size); u16 nvbios_findstr(const u8 *data, int size, const char *str, int len); int nvbios_memcmp(struct nvkm_bios *, u32 addr, const char *, u32 len); - -#define nvbios_rd08(b,o) (b)->data[(o)] -#define nvbios_rd16(b,o) get_unaligned_le16(&(b)->data[(o)]) -#define nvbios_rd32(b,o) get_unaligned_le32(&(b)->data[(o)]) +u8 nvbios_rd08(struct nvkm_bios *, u32 addr); +u16 nvbios_rd16(struct nvkm_bios *, u32 addr); +u32 nvbios_rd32(struct nvkm_bios *, u32 addr); int nvkm_bios_new(struct nvkm_device *, int, struct nvkm_bios **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/disp.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/disp.h index db10c11f0595..c5a6ebd5a478 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/disp.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/disp.h @@ -25,7 +25,8 @@ u16 nvbios_outp_match(struct nvkm_bios *, u16 type, u16 mask, u8 *ver, u8 *hdr, u8 *cnt, u8 *len, struct nvbios_outp *); struct nvbios_ocfg { - u16 match; + u8 proto; + u8 flags; u16 clkcmp[2]; }; @@ -33,7 +34,7 @@ u16 nvbios_ocfg_entry(struct nvkm_bios *, u16 outp, u8 idx, u8 *ver, u8 *hdr, u8 *cnt, u8 *len); u16 nvbios_ocfg_parse(struct nvkm_bios *, u16 outp, u8 idx, u8 *ver, u8 *hdr, u8 *cnt, u8 *len, struct nvbios_ocfg *); -u16 nvbios_ocfg_match(struct nvkm_bios *, u16 outp, u16 type, +u16 nvbios_ocfg_match(struct nvkm_bios *, u16 outp, u8 proto, u8 flags, u8 *ver, u8 *hdr, u8 *cnt, u8 *len, struct nvbios_ocfg *); u16 nvbios_oclk_match(struct nvkm_bios *, u16 cmp, u32 khz); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h index 0a734fd06acf..3a410275fa71 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h @@ -56,6 +56,8 @@ struct nvkm_fb { int regions; } tile; + u8 page; + struct nvkm_memory *mmu_rd; struct nvkm_memory *mmu_wr; }; @@ -91,6 +93,8 @@ int gk104_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gk20a_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gm107_fb_new(struct nvkm_device *, int, struct nvkm_fb **); int gm200_fb_new(struct nvkm_device *, int, struct nvkm_fb **); +int gp100_fb_new(struct nvkm_device *, int, struct nvkm_fb **); +int gp104_fb_new(struct nvkm_device *, int, struct nvkm_fb **); #include <subdev/bios.h> #include <subdev/bios/ramcfg.h> diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h index c6b90b6543b3..cd755baf9cab 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h @@ -38,4 +38,5 @@ int gk104_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **); int gk20a_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **); int gm107_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **); int gm200_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **); +int gp100_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h index 2e80682b2da1..27d25b18d85c 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mc.h @@ -7,11 +7,14 @@ struct nvkm_mc { struct nvkm_subdev subdev; }; -void nvkm_mc_intr(struct nvkm_mc *, bool *handled); -void nvkm_mc_intr_unarm(struct nvkm_mc *); -void nvkm_mc_intr_rearm(struct nvkm_mc *); -void nvkm_mc_reset(struct nvkm_mc *, enum nvkm_devidx); -void nvkm_mc_unk260(struct nvkm_mc *, u32 data); +void nvkm_mc_enable(struct nvkm_device *, enum nvkm_devidx); +void nvkm_mc_disable(struct nvkm_device *, enum nvkm_devidx); +void nvkm_mc_reset(struct nvkm_device *, enum nvkm_devidx); +void nvkm_mc_intr(struct nvkm_device *, bool *handled); +void nvkm_mc_intr_unarm(struct nvkm_device *); +void nvkm_mc_intr_rearm(struct nvkm_device *); +void nvkm_mc_intr_mask(struct nvkm_device *, enum nvkm_devidx, bool enable); +void nvkm_mc_unk260(struct nvkm_device *, u32 data); int nv04_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int nv11_mc_new(struct nvkm_device *, int, struct nvkm_mc **); @@ -24,4 +27,5 @@ int gt215_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gf100_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gk104_mc_new(struct nvkm_device *, int, struct nvkm_mc **); int gk20a_mc_new(struct nvkm_device *, int, struct nvkm_mc **); +int gp100_mc_new(struct nvkm_device *, int, struct nvkm_mc **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h index ddb913889d7e..e6523e2cea9f 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/pci.h @@ -47,6 +47,7 @@ int g94_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int gf100_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int gf106_pci_new(struct nvkm_device *, int, struct nvkm_pci **); int gk104_pci_new(struct nvkm_device *, int, struct nvkm_pci **); +int gp100_pci_new(struct nvkm_device *, int, struct nvkm_pci **); /* pcie functions */ int nvkm_pcie_set_link(struct nvkm_pci *, enum nvkm_pcie_speed, u8 width); diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h index c6edd95a5b69..b04c38c07761 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/secboot.h @@ -43,9 +43,8 @@ struct nvkm_secboot { const struct nvkm_secboot_func *func; struct nvkm_subdev subdev; + enum nvkm_devidx devidx; u32 base; - u32 irq_mask; - u32 enable_mask; }; #define nvkm_secboot(p) container_of((p), struct nvkm_secboot, subdev) diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h index 8fb575a92c48..71ebbfd4484f 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/top.h @@ -8,10 +8,11 @@ struct nvkm_top { struct list_head device; }; -u32 nvkm_top_reset(struct nvkm_top *, enum nvkm_devidx); -u32 nvkm_top_intr(struct nvkm_top *, u32 intr, u64 *subdevs); -enum nvkm_devidx nvkm_top_fault(struct nvkm_top *, int fault); -enum nvkm_devidx nvkm_top_engine(struct nvkm_top *, int, int *runl, int *engn); +u32 nvkm_top_reset(struct nvkm_device *, enum nvkm_devidx); +u32 nvkm_top_intr(struct nvkm_device *, u32 intr, u64 *subdevs); +u32 nvkm_top_intr_mask(struct nvkm_device *, enum nvkm_devidx); +enum nvkm_devidx nvkm_top_fault(struct nvkm_device *, int fault); +enum nvkm_devidx nvkm_top_engine(struct nvkm_device *, int, int *runl, int *engn); int gk104_top_new(struct nvkm_device *, int, struct nvkm_top **); #endif diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/volt.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/volt.h index feff55cff05b..b765f4ffcde6 100644 --- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/volt.h +++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/volt.h @@ -12,6 +12,9 @@ struct nvkm_volt { u32 uv; u8 vid; } vid[256]; + + u32 max_uv; + u32 min_uv; }; int nvkm_volt_get(struct nvkm_volt *); diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c index eb7de487a2b3..7bd4683216d0 100644 --- a/drivers/gpu/drm/nouveau/nouveau_abi16.c +++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c @@ -100,6 +100,7 @@ nouveau_abi16_swclass(struct nouveau_drm *drm) case NV_DEVICE_INFO_V0_FERMI: case NV_DEVICE_INFO_V0_KEPLER: case NV_DEVICE_INFO_V0_MAXWELL: + case NV_DEVICE_INFO_V0_PASCAL: return NVIF_CLASS_SW_GF100; } diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c index 5e3f3e826476..528bdeffb339 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bo.c +++ b/drivers/gpu/drm/nouveau/nouveau_bo.c @@ -209,8 +209,7 @@ nouveau_bo_new(struct drm_device *dev, int size, int align, nvbo->tile_flags = tile_flags; nvbo->bo.bdev = &drm->ttm.bdev; - if (!nvxx_device(&drm->device)->func->cpu_coherent) - nvbo->force_coherent = flags & TTM_PL_FLAG_UNCACHED; + nvbo->force_coherent = flags & TTM_PL_FLAG_UNCACHED; nvbo->page_shift = 12; if (drm->client.vm) { @@ -424,13 +423,7 @@ nouveau_bo_map(struct nouveau_bo *nvbo) if (ret) return ret; - /* - * TTM buffers allocated using the DMA API already have a mapping, let's - * use it instead. - */ - if (!nvbo->force_coherent) - ret = ttm_bo_kmap(&nvbo->bo, 0, nvbo->bo.mem.num_pages, - &nvbo->kmap); + ret = ttm_bo_kmap(&nvbo->bo, 0, nvbo->bo.mem.num_pages, &nvbo->kmap); ttm_bo_unreserve(&nvbo->bo); return ret; @@ -442,12 +435,7 @@ nouveau_bo_unmap(struct nouveau_bo *nvbo) if (!nvbo) return; - /* - * TTM buffers allocated using the DMA API already had a coherent - * mapping which we used, no need to unmap. - */ - if (!nvbo->force_coherent) - ttm_bo_kunmap(&nvbo->kmap); + ttm_bo_kunmap(&nvbo->kmap); } void @@ -506,35 +494,13 @@ nouveau_bo_validate(struct nouveau_bo *nvbo, bool interruptible, return 0; } -static inline void * -_nouveau_bo_mem_index(struct nouveau_bo *nvbo, unsigned index, void *mem, u8 sz) -{ - struct ttm_dma_tt *dma_tt; - u8 *m = mem; - - index *= sz; - - if (m) { - /* kmap'd address, return the corresponding offset */ - m += index; - } else { - /* DMA-API mapping, lookup the right address */ - dma_tt = (struct ttm_dma_tt *)nvbo->bo.ttm; - m = dma_tt->cpu_address[index / PAGE_SIZE]; - m += index % PAGE_SIZE; - } - - return m; -} -#define nouveau_bo_mem_index(o, i, m) _nouveau_bo_mem_index(o, i, m, sizeof(*m)) - void nouveau_bo_wr16(struct nouveau_bo *nvbo, unsigned index, u16 val) { bool is_iomem; u16 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem); - mem = nouveau_bo_mem_index(nvbo, index, mem); + mem += index; if (is_iomem) iowrite16_native(val, (void __force __iomem *)mem); @@ -548,7 +514,7 @@ nouveau_bo_rd32(struct nouveau_bo *nvbo, unsigned index) bool is_iomem; u32 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem); - mem = nouveau_bo_mem_index(nvbo, index, mem); + mem += index; if (is_iomem) return ioread32_native((void __force __iomem *)mem); @@ -562,7 +528,7 @@ nouveau_bo_wr32(struct nouveau_bo *nvbo, unsigned index, u32 val) bool is_iomem; u32 *mem = ttm_kmap_obj_virtual(&nvbo->kmap, &is_iomem); - mem = nouveau_bo_mem_index(nvbo, index, mem); + mem += index; if (is_iomem) iowrite32_native(val, (void __force __iomem *)mem); @@ -1082,7 +1048,6 @@ nouveau_bo_move_m2mf(struct ttm_buffer_object *bo, int evict, bool intr, ret = ttm_bo_move_accel_cleanup(bo, &fence->base, evict, - no_wait_gpu, new_mem); nouveau_fence_unref(&fence); } @@ -1104,6 +1069,10 @@ nouveau_bo_move_init(struct nouveau_drm *drm) struct ttm_mem_reg *, struct ttm_mem_reg *); int (*init)(struct nouveau_channel *, u32 handle); } _methods[] = { + { "COPY", 4, 0xc1b5, nve0_bo_move_copy, nve0_bo_move_init }, + { "GRCE", 0, 0xc1b5, nve0_bo_move_copy, nvc0_bo_move_init }, + { "COPY", 4, 0xc0b5, nve0_bo_move_copy, nve0_bo_move_init }, + { "GRCE", 0, 0xc0b5, nve0_bo_move_copy, nvc0_bo_move_init }, { "COPY", 4, 0xb0b5, nve0_bo_move_copy, nve0_bo_move_init }, { "GRCE", 0, 0xb0b5, nve0_bo_move_copy, nvc0_bo_move_init }, { "COPY", 4, 0xa0b5, nve0_bo_move_copy, nve0_bo_move_init }, @@ -1289,6 +1258,10 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr, struct nouveau_drm_tile *new_tile = NULL; int ret = 0; + ret = ttm_bo_wait(bo, intr, no_wait_gpu); + if (ret) + return ret; + if (nvbo->pin_refcnt) NV_WARN(drm, "Moving pinned object %p!\n", nvbo); @@ -1324,7 +1297,7 @@ nouveau_bo_move(struct ttm_buffer_object *bo, bool evict, bool intr, /* Fallback to software copy. */ ret = ttm_bo_wait(bo, intr, no_wait_gpu); if (ret == 0) - ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); + ret = ttm_bo_move_memcpy(bo, evict, intr, no_wait_gpu, new_mem); out: if (drm->device.info.family < NV_DEVICE_INFO_V0_TESLA) { @@ -1488,14 +1461,6 @@ nouveau_ttm_tt_populate(struct ttm_tt *ttm) dev = drm->dev; pdev = device->dev; - /* - * Objects matching this condition have been marked as force_coherent, - * so use the DMA API for them. - */ - if (!nvxx_device(&drm->device)->func->cpu_coherent && - ttm->caching_state == tt_uncached) - return ttm_dma_populate(ttm_dma, dev->dev); - #if IS_ENABLED(CONFIG_AGP) if (drm->agp.bridge) { return ttm_agp_tt_populate(ttm); @@ -1553,16 +1518,6 @@ nouveau_ttm_tt_unpopulate(struct ttm_tt *ttm) dev = drm->dev; pdev = device->dev; - /* - * Objects matching this condition have been marked as force_coherent, - * so use the DMA API for them. - */ - if (!nvxx_device(&drm->device)->func->cpu_coherent && - ttm->caching_state == tt_uncached) { - ttm_dma_unpopulate(ttm_dma, dev->dev); - return; - } - #if IS_ENABLED(CONFIG_AGP) if (drm->agp.bridge) { ttm_agp_tt_unpopulate(ttm); diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c index b1d2527c5625..f9b3c811187e 100644 --- a/drivers/gpu/drm/nouveau/nouveau_chan.c +++ b/drivers/gpu/drm/nouveau/nouveau_chan.c @@ -191,7 +191,8 @@ static int nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device, u32 engine, struct nouveau_channel **pchan) { - static const u16 oclasses[] = { MAXWELL_CHANNEL_GPFIFO_A, + static const u16 oclasses[] = { PASCAL_CHANNEL_GPFIFO_A, + MAXWELL_CHANNEL_GPFIFO_A, KEPLER_CHANNEL_GPFIFO_B, KEPLER_CHANNEL_GPFIFO_A, FERMI_CHANNEL_GPFIFO, diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c index 6e97862d31a1..afbf557b23d4 100644 --- a/drivers/gpu/drm/nouveau/nouveau_display.c +++ b/drivers/gpu/drm/nouveau/nouveau_display.c @@ -495,6 +495,8 @@ nouveau_display_create(struct drm_device *dev) if (nouveau_modeset != 2 && drm->vbios.dcb.entries) { static const u16 oclass[] = { + GP104_DISP, + GP100_DISP, GM200_DISP, GM107_DISP, GK110_DISP, diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c index 1161f8b452f2..66c1280c0f1f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drm.c +++ b/drivers/gpu/drm/nouveau/nouveau_drm.c @@ -198,6 +198,7 @@ nouveau_accel_init(struct nouveau_drm *drm) case KEPLER_CHANNEL_GPFIFO_A: case KEPLER_CHANNEL_GPFIFO_B: case MAXWELL_CHANNEL_GPFIFO_A: + case PASCAL_CHANNEL_GPFIFO_A: ret = nvc0_fence_create(drm); break; default: @@ -316,7 +317,16 @@ static int nouveau_drm_probe(struct pci_dev *pdev, if (vga_switcheroo_client_probe_defer(pdev)) return -EPROBE_DEFER; - /* remove conflicting drivers (vesafb, efifb etc) */ + /* We need to check that the chipset is supported before booting + * fbdev off the hardware, as there's no way to put it back. + */ + ret = nvkm_device_pci_new(pdev, NULL, "error", true, false, 0, &device); + if (ret) + return ret; + + nvkm_device_del(&device); + + /* Remove conflicting drivers (vesafb, efifb etc). */ aper = alloc_apertures(3); if (!aper) return -ENOMEM; @@ -430,6 +440,11 @@ nouveau_drm_load(struct drm_device *dev, unsigned long flags) nouveau_vga_init(drm); if (drm->device.info.family >= NV_DEVICE_INFO_V0_TESLA) { + if (!nvxx_device(&drm->device)->mmu) { + ret = -ENOSYS; + goto fail_device; + } + ret = nvkm_vm_new(nvxx_device(&drm->device), 0, (1ULL << 40), 0x1000, NULL, &drm->client.vm); if (ret) diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c index 57aaf98a26f9..d1f248fd3506 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c +++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c @@ -552,6 +552,8 @@ nouveau_fbcon_init(struct drm_device *dev) if (ret) goto fini; + if (fbcon->helper.fbdev) + fbcon->helper.fbdev->pixmap.buf_align = 4; return 0; fini: diff --git a/drivers/gpu/drm/nouveau/nouveau_hwmon.c b/drivers/gpu/drm/nouveau/nouveau_hwmon.c index 1ff4166af26e..71f764bf4cc6 100644 --- a/drivers/gpu/drm/nouveau/nouveau_hwmon.c +++ b/drivers/gpu/drm/nouveau/nouveau_hwmon.c @@ -535,6 +535,40 @@ static SENSOR_DEVICE_ATTR(in0_input, S_IRUGO, nouveau_hwmon_get_in0_input, NULL, 0); static ssize_t +nouveau_hwmon_get_in0_min(struct device *d, + struct device_attribute *a, char *buf) +{ + struct drm_device *dev = dev_get_drvdata(d); + struct nouveau_drm *drm = nouveau_drm(dev); + struct nvkm_volt *volt = nvxx_volt(&drm->device); + + if (!volt || !volt->min_uv) + return -ENODEV; + + return sprintf(buf, "%i\n", volt->min_uv / 1000); +} + +static SENSOR_DEVICE_ATTR(in0_min, S_IRUGO, + nouveau_hwmon_get_in0_min, NULL, 0); + +static ssize_t +nouveau_hwmon_get_in0_max(struct device *d, + struct device_attribute *a, char *buf) +{ + struct drm_device *dev = dev_get_drvdata(d); + struct nouveau_drm *drm = nouveau_drm(dev); + struct nvkm_volt *volt = nvxx_volt(&drm->device); + + if (!volt || !volt->max_uv) + return -ENODEV; + + return sprintf(buf, "%i\n", volt->max_uv / 1000); +} + +static SENSOR_DEVICE_ATTR(in0_max, S_IRUGO, + nouveau_hwmon_get_in0_max, NULL, 0); + +static ssize_t nouveau_hwmon_get_in0_label(struct device *d, struct device_attribute *a, char *buf) { @@ -594,6 +628,8 @@ static struct attribute *hwmon_pwm_fan_attributes[] = { static struct attribute *hwmon_in0_attributes[] = { &sensor_dev_attr_in0_input.dev_attr.attr, + &sensor_dev_attr_in0_min.dev_attr.attr, + &sensor_dev_attr_in0_max.dev_attr.attr, &sensor_dev_attr_in0_label.dev_attr.attr, NULL }; diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c index bcee91497eb9..1825dbc33192 100644 --- a/drivers/gpu/drm/nouveau/nouveau_ttm.c +++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c @@ -164,6 +164,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man, case NV_DEVICE_INFO_V0_FERMI: case NV_DEVICE_INFO_V0_KEPLER: case NV_DEVICE_INFO_V0_MAXWELL: + case NV_DEVICE_INFO_V0_PASCAL: node->memtype = (nvbo->tile_flags & 0xff00) >> 8; break; default: diff --git a/drivers/gpu/drm/nouveau/nv04_fbcon.c b/drivers/gpu/drm/nouveau/nv04_fbcon.c index 0f3e4bb411cc..7d9248b8c664 100644 --- a/drivers/gpu/drm/nouveau/nv04_fbcon.c +++ b/drivers/gpu/drm/nouveau/nv04_fbcon.c @@ -82,7 +82,6 @@ nv04_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) uint32_t fg; uint32_t bg; uint32_t dsize; - uint32_t width; uint32_t *data = (uint32_t *)image->data; int ret; @@ -93,9 +92,6 @@ nv04_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) if (ret) return ret; - width = ALIGN(image->width, 8); - dsize = ALIGN(width * image->height, 32) >> 5; - if (info->fix.visual == FB_VISUAL_TRUECOLOR || info->fix.visual == FB_VISUAL_DIRECTCOLOR) { fg = ((uint32_t *) info->pseudo_palette)[image->fg_color]; @@ -111,10 +107,11 @@ nv04_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) ((image->dx + image->width) & 0xffff)); OUT_RING(chan, bg); OUT_RING(chan, fg); - OUT_RING(chan, (image->height << 16) | width); + OUT_RING(chan, (image->height << 16) | image->width); OUT_RING(chan, (image->height << 16) | image->width); OUT_RING(chan, (image->dy << 16) | (image->dx & 0xffff)); + dsize = ALIGN(image->width * image->height, 32) >> 5; while (dsize) { int iter_len = dsize > 128 ? 128 : dsize; diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c index 7a7788212df7..7d0edcbcfca7 100644 --- a/drivers/gpu/drm/nouveau/nv50_display.c +++ b/drivers/gpu/drm/nouveau/nv50_display.c @@ -297,6 +297,8 @@ nv50_core_create(struct nvif_device *device, struct nvif_object *disp, .pushbuf = 0xb0007d00, }; static const s32 oclass[] = { + GP104_DISP_CORE_CHANNEL_DMA, + GP100_DISP_CORE_CHANNEL_DMA, GM200_DISP_CORE_CHANNEL_DMA, GM107_DISP_CORE_CHANNEL_DMA, GK110_DISP_CORE_CHANNEL_DMA, diff --git a/drivers/gpu/drm/nouveau/nv50_fbcon.c b/drivers/gpu/drm/nouveau/nv50_fbcon.c index 33d9ee0fac40..1aeb698e9707 100644 --- a/drivers/gpu/drm/nouveau/nv50_fbcon.c +++ b/drivers/gpu/drm/nouveau/nv50_fbcon.c @@ -95,7 +95,7 @@ nv50_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) struct nouveau_fbdev *nfbdev = info->par; struct nouveau_drm *drm = nouveau_drm(nfbdev->dev); struct nouveau_channel *chan = drm->channel; - uint32_t width, dwords, *data = (uint32_t *)image->data; + uint32_t dwords, *data = (uint32_t *)image->data; uint32_t mask = ~(~0 >> (32 - info->var.bits_per_pixel)); uint32_t *palette = info->pseudo_palette; int ret; @@ -107,9 +107,6 @@ nv50_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) if (ret) return ret; - width = ALIGN(image->width, 32); - dwords = (width * image->height) >> 5; - BEGIN_NV04(chan, NvSub2D, 0x0814, 2); if (info->fix.visual == FB_VISUAL_TRUECOLOR || info->fix.visual == FB_VISUAL_DIRECTCOLOR) { @@ -128,6 +125,7 @@ nv50_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) OUT_RING(chan, 0); OUT_RING(chan, image->dy); + dwords = ALIGN(image->width * image->height, 32) >> 5; while (dwords) { int push = dwords > 2047 ? 2047 : dwords; diff --git a/drivers/gpu/drm/nouveau/nvc0_fbcon.c b/drivers/gpu/drm/nouveau/nvc0_fbcon.c index a0913359ac05..839f4c8c1805 100644 --- a/drivers/gpu/drm/nouveau/nvc0_fbcon.c +++ b/drivers/gpu/drm/nouveau/nvc0_fbcon.c @@ -95,7 +95,7 @@ nvc0_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) struct nouveau_fbdev *nfbdev = info->par; struct nouveau_drm *drm = nouveau_drm(nfbdev->dev); struct nouveau_channel *chan = drm->channel; - uint32_t width, dwords, *data = (uint32_t *)image->data; + uint32_t dwords, *data = (uint32_t *)image->data; uint32_t mask = ~(~0 >> (32 - info->var.bits_per_pixel)); uint32_t *palette = info->pseudo_palette; int ret; @@ -107,9 +107,6 @@ nvc0_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) if (ret) return ret; - width = ALIGN(image->width, 32); - dwords = (width * image->height) >> 5; - BEGIN_NVC0(chan, NvSub2D, 0x0814, 2); if (info->fix.visual == FB_VISUAL_TRUECOLOR || info->fix.visual == FB_VISUAL_DIRECTCOLOR) { @@ -128,6 +125,7 @@ nvc0_fbcon_imageblit(struct fb_info *info, const struct fb_image *image) OUT_RING (chan, 0); OUT_RING (chan, image->dy); + dwords = ALIGN(image->width * image->height, 32) >> 5; while (dwords) { int push = dwords > 2047 ? 2047 : dwords; diff --git a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c index b18557858f19..19044aba265e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c +++ b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c @@ -57,6 +57,9 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = { [NVKM_ENGINE_CE0 ] = "ce0", [NVKM_ENGINE_CE1 ] = "ce1", [NVKM_ENGINE_CE2 ] = "ce2", + [NVKM_ENGINE_CE3 ] = "ce3", + [NVKM_ENGINE_CE4 ] = "ce4", + [NVKM_ENGINE_CE5 ] = "ce5", [NVKM_ENGINE_CIPHER ] = "cipher", [NVKM_ENGINE_DISP ] = "disp", [NVKM_ENGINE_DMAOBJ ] = "dma", @@ -71,6 +74,7 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = { [NVKM_ENGINE_MSVLD ] = "msvld", [NVKM_ENGINE_NVENC0 ] = "nvenc0", [NVKM_ENGINE_NVENC1 ] = "nvenc1", + [NVKM_ENGINE_NVENC2 ] = "nvenc2", [NVKM_ENGINE_NVDEC ] = "nvdec", [NVKM_ENGINE_PM ] = "pm", [NVKM_ENGINE_SEC ] = "sec", @@ -105,7 +109,7 @@ nvkm_subdev_fini(struct nvkm_subdev *subdev, bool suspend) } } - nvkm_mc_reset(device->mc, subdev->index); + nvkm_mc_reset(device, subdev->index); time = ktime_to_us(ktime_get()) - time; nvkm_trace(subdev, "%s completed in %lldus\n", action, time); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild index 9c19d59b47df..a4458a8eb30a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild @@ -3,3 +3,5 @@ nvkm-y += nvkm/engine/ce/gf100.o nvkm-y += nvkm/engine/ce/gk104.o nvkm-y += nvkm/engine/ce/gm107.o nvkm-y += nvkm/engine/ce/gm200.o +nvkm-y += nvkm/engine/ce/gp100.o +nvkm-y += nvkm/engine/ce/gp104.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp100.c new file mode 100644 index 000000000000..c7710456bc30 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp100.c @@ -0,0 +1,102 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "priv.h" +#include <core/enum.h> + +#include <nvif/class.h> + +static const struct nvkm_enum +gp100_ce_launcherr_report[] = { + { 0x0, "NO_ERR" }, + { 0x1, "2D_LAYER_EXCEEDS_DEPTH" }, + { 0x2, "INVALID_ALIGNMENT" }, + { 0x3, "MEM2MEM_RECT_OUT_OF_BOUNDS" }, + { 0x4, "SRC_LINE_EXCEEDS_PITCH" }, + { 0x5, "SRC_LINE_EXCEEDS_NEG_PITCH" }, + { 0x6, "DST_LINE_EXCEEDS_PITCH" }, + { 0x7, "DST_LINE_EXCEEDS_NEG_PITCH" }, + { 0x8, "BAD_SRC_PIXEL_COMP_REF" }, + { 0x9, "INVALID_VALUE" }, + { 0xa, "UNUSED_FIELD" }, + { 0xb, "INVALID_OPERATION" }, + { 0xc, "NO_RESOURCES" }, + { 0xd, "INVALID_CONFIG" }, + {} +}; + +static void +gp100_ce_intr_launcherr(struct nvkm_engine *ce, const u32 base) +{ + struct nvkm_subdev *subdev = &ce->subdev; + struct nvkm_device *device = subdev->device; + u32 stat = nvkm_rd32(device, 0x104418 + base); + const struct nvkm_enum *en = + nvkm_enum_find(gp100_ce_launcherr_report, stat & 0x0000000f); + nvkm_warn(subdev, "LAUNCHERR %08x [%s]\n", stat, en ? en->name : ""); +} + +void +gp100_ce_intr(struct nvkm_engine *ce) +{ + const u32 base = (ce->subdev.index - NVKM_ENGINE_CE0) * 0x80; + struct nvkm_subdev *subdev = &ce->subdev; + struct nvkm_device *device = subdev->device; + u32 mask = nvkm_rd32(device, 0x10440c + base); + u32 intr = nvkm_rd32(device, 0x104410 + base) & mask; + if (intr & 0x00000001) { //XXX: guess + nvkm_warn(subdev, "BLOCKPIPE\n"); + nvkm_wr32(device, 0x104410 + base, 0x00000001); + intr &= ~0x00000001; + } + if (intr & 0x00000002) { //XXX: guess + nvkm_warn(subdev, "NONBLOCKPIPE\n"); + nvkm_wr32(device, 0x104410 + base, 0x00000002); + intr &= ~0x00000002; + } + if (intr & 0x00000004) { + gp100_ce_intr_launcherr(ce, base); + nvkm_wr32(device, 0x104410 + base, 0x00000004); + intr &= ~0x00000004; + } + if (intr) { + nvkm_warn(subdev, "intr %08x\n", intr); + nvkm_wr32(device, 0x104410 + base, intr); + } +} + +static const struct nvkm_engine_func +gp100_ce = { + .intr = gp100_ce_intr, + .sclass = { + { -1, -1, PASCAL_DMA_COPY_A }, + {} + } +}; + +int +gp100_ce_new(struct nvkm_device *device, int index, + struct nvkm_engine **pengine) +{ + return nvkm_engine_new_(&gp100_ce, device, index, true, pengine); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp104.c new file mode 100644 index 000000000000..20e019788a53 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gp104.c @@ -0,0 +1,44 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "priv.h" +#include <core/enum.h> + +#include <nvif/class.h> + +static const struct nvkm_engine_func +gp104_ce = { + .intr = gp100_ce_intr, + .sclass = { + { -1, -1, PASCAL_DMA_COPY_B }, + { -1, -1, PASCAL_DMA_COPY_A }, + {} + } +}; + +int +gp104_ce_new(struct nvkm_device *device, int index, + struct nvkm_engine **pengine) +{ + return nvkm_engine_new_(&gp104_ce, device, index, true, pengine); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/ce/priv.h index e2fa8b161943..2dce405976ad 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/ce/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/priv.h @@ -4,4 +4,5 @@ void gt215_ce_intr(struct nvkm_falcon *, struct nvkm_fifo_chan *); void gk104_ce_intr(struct nvkm_engine *); +void gp100_ce_intr(struct nvkm_engine *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c index 4572debcb0c9..7218a067a6c5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c @@ -2148,6 +2148,67 @@ nv12b_chipset = { .sw = gf100_sw_new, }; +static const struct nvkm_device_chip +nv130_chipset = { + .name = "GP100", + .bar = gf100_bar_new, + .bios = nvkm_bios_new, + .bus = gf100_bus_new, + .devinit = gm200_devinit_new, + .fb = gp100_fb_new, + .fuse = gm107_fuse_new, + .gpio = gk104_gpio_new, + .i2c = gm200_i2c_new, + .ibus = gm200_ibus_new, + .imem = nv50_instmem_new, + .ltc = gp100_ltc_new, + .mc = gp100_mc_new, + .mmu = gf100_mmu_new, + .secboot = gm200_secboot_new, + .pci = gp100_pci_new, + .timer = gk20a_timer_new, + .top = gk104_top_new, + .ce[0] = gp100_ce_new, + .ce[1] = gp100_ce_new, + .ce[2] = gp100_ce_new, + .ce[3] = gp100_ce_new, + .ce[4] = gp100_ce_new, + .ce[5] = gp100_ce_new, + .dma = gf119_dma_new, + .disp = gp100_disp_new, + .fifo = gp100_fifo_new, + .gr = gp100_gr_new, + .sw = gf100_sw_new, +}; + +static const struct nvkm_device_chip +nv134_chipset = { + .name = "GP104", + .bar = gf100_bar_new, + .bios = nvkm_bios_new, + .bus = gf100_bus_new, + .devinit = gm200_devinit_new, + .fb = gp104_fb_new, + .fuse = gm107_fuse_new, + .gpio = gk104_gpio_new, + .i2c = gm200_i2c_new, + .ibus = gm200_ibus_new, + .imem = nv50_instmem_new, + .ltc = gp100_ltc_new, + .mc = gp100_mc_new, + .mmu = gf100_mmu_new, + .pci = gp100_pci_new, + .timer = gk20a_timer_new, + .top = gk104_top_new, + .ce[0] = gp104_ce_new, + .ce[1] = gp104_ce_new, + .ce[2] = gp104_ce_new, + .ce[3] = gp104_ce_new, + .disp = gp104_disp_new, + .dma = gf119_dma_new, + .fifo = gp100_fifo_new, +}; + static int nvkm_device_event_ctor(struct nvkm_object *object, void *data, u32 size, struct nvkm_notify *notify) @@ -2221,6 +2282,9 @@ nvkm_device_engine(struct nvkm_device *device, int index) _(CE0 , device->ce[0] , device->ce[0]); _(CE1 , device->ce[1] , device->ce[1]); _(CE2 , device->ce[2] , device->ce[2]); + _(CE3 , device->ce[3] , device->ce[3]); + _(CE4 , device->ce[4] , device->ce[4]); + _(CE5 , device->ce[5] , device->ce[5]); _(CIPHER , device->cipher , device->cipher); _(DISP , device->disp , &device->disp->engine); _(DMAOBJ , device->dma , &device->dma->engine); @@ -2235,6 +2299,7 @@ nvkm_device_engine(struct nvkm_device *device, int index) _(MSVLD , device->msvld , device->msvld); _(NVENC0 , device->nvenc[0], device->nvenc[0]); _(NVENC1 , device->nvenc[1], device->nvenc[1]); + _(NVENC2 , device->nvenc[2], device->nvenc[2]); _(NVDEC , device->nvdec , device->nvdec); _(PM , device->pm , &device->pm->engine); _(SEC , device->sec , device->sec); @@ -2492,6 +2557,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func, case 0x100: device->card_type = NV_E0; break; case 0x110: case 0x120: device->card_type = GM100; break; + case 0x130: device->card_type = GP100; break; default: break; } @@ -2576,6 +2642,8 @@ nvkm_device_ctor(const struct nvkm_device_func *func, case 0x124: device->chip = &nv124_chipset; break; case 0x126: device->chip = &nv126_chipset; break; case 0x12b: device->chip = &nv12b_chipset; break; + case 0x130: device->chip = &nv130_chipset; break; + case 0x134: device->chip = &nv134_chipset; break; default: nvdev_error(device, "unknown chipset (%08x)\n", boot0); goto done; @@ -2659,6 +2727,9 @@ nvkm_device_ctor(const struct nvkm_device_func *func, _(NVKM_ENGINE_CE0 , ce[0]); _(NVKM_ENGINE_CE1 , ce[1]); _(NVKM_ENGINE_CE2 , ce[2]); + _(NVKM_ENGINE_CE3 , ce[3]); + _(NVKM_ENGINE_CE4 , ce[4]); + _(NVKM_ENGINE_CE5 , ce[5]); _(NVKM_ENGINE_CIPHER , cipher); _(NVKM_ENGINE_DISP , disp); _(NVKM_ENGINE_DMAOBJ , dma); @@ -2673,6 +2744,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func, _(NVKM_ENGINE_MSVLD , msvld); _(NVKM_ENGINE_NVENC0 , nvenc[0]); _(NVKM_ENGINE_NVENC1 , nvenc[1]); + _(NVKM_ENGINE_NVENC2 , nvenc[2]); _(NVKM_ENGINE_NVDEC , nvdec); _(NVKM_ENGINE_PM , pm); _(NVKM_ENGINE_SEC , sec); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/pci.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/pci.c index 18fab3973ce5..b1b693219db3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/pci.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/pci.c @@ -1614,7 +1614,6 @@ nvkm_device_pci_func = { .fini = nvkm_device_pci_fini, .resource_addr = nvkm_device_pci_resource_addr, .resource_size = nvkm_device_pci_resource_size, - .cpu_coherent = !IS_ENABLED(CONFIG_ARM) && !IS_ENABLED(CONFIG_ARM64), }; int diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c index ec12efb4689a..939682f18788 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c @@ -191,13 +191,11 @@ static irqreturn_t nvkm_device_tegra_intr(int irq, void *arg) { struct nvkm_device_tegra *tdev = arg; - struct nvkm_mc *mc = tdev->device.mc; + struct nvkm_device *device = &tdev->device; bool handled = false; - if (likely(mc)) { - nvkm_mc_intr_unarm(mc); - nvkm_mc_intr(mc, &handled); - nvkm_mc_intr_rearm(mc); - } + nvkm_mc_intr_unarm(device); + nvkm_mc_intr(device, &handled); + nvkm_mc_intr_rearm(device); return handled ? IRQ_HANDLED : IRQ_NONE; } @@ -247,7 +245,6 @@ nvkm_device_tegra_func = { .fini = nvkm_device_tegra_fini, .resource_addr = nvkm_device_tegra_resource_addr, .resource_size = nvkm_device_tegra_resource_size, - .cpu_coherent = false, }; int @@ -313,6 +310,7 @@ nvkm_device_tegra_new(const struct nvkm_device_tegra_func *func, goto remove; tdev->gpu_speedo = tegra_sku_info.gpu_speedo_value; + tdev->gpu_speedo_id = tegra_sku_info.gpu_speedo_id; ret = nvkm_device_ctor(&nvkm_device_tegra_func, NULL, &pdev->dev, NVKM_DEVICE_TEGRA, pdev->id, NULL, cfg, dbg, detect, mmio, subdev_mask, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c index 137066426ed7..79a8f71cf788 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c @@ -102,6 +102,7 @@ nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size) case NV_C0: args->v0.family = NV_DEVICE_INFO_V0_FERMI; break; case NV_E0: args->v0.family = NV_DEVICE_INFO_V0_KEPLER; break; case GM100: args->v0.family = NV_DEVICE_INFO_V0_MAXWELL; break; + case GP100: args->v0.family = NV_DEVICE_INFO_V0_PASCAL; break; default: args->v0.family = 0; break; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild index a74c5dd27dc0..77a52b54a31e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild @@ -10,6 +10,8 @@ nvkm-y += nvkm/engine/disp/gk104.o nvkm-y += nvkm/engine/disp/gk110.o nvkm-y += nvkm/engine/disp/gm107.o nvkm-y += nvkm/engine/disp/gm200.o +nvkm-y += nvkm/engine/disp/gp100.o +nvkm-y += nvkm/engine/disp/gp104.o nvkm-y += nvkm/engine/disp/outp.o nvkm-y += nvkm/engine/disp/outpdp.o @@ -18,6 +20,7 @@ nvkm-y += nvkm/engine/disp/piornv50.o nvkm-y += nvkm/engine/disp/sornv50.o nvkm-y += nvkm/engine/disp/sorg94.o nvkm-y += nvkm/engine/disp/sorgf119.o +nvkm-y += nvkm/engine/disp/sorgm107.o nvkm-y += nvkm/engine/disp/sorgm200.o nvkm-y += nvkm/engine/disp/dport.o @@ -44,12 +47,15 @@ nvkm-y += nvkm/engine/disp/rootgk104.o nvkm-y += nvkm/engine/disp/rootgk110.o nvkm-y += nvkm/engine/disp/rootgm107.o nvkm-y += nvkm/engine/disp/rootgm200.o +nvkm-y += nvkm/engine/disp/rootgp100.o +nvkm-y += nvkm/engine/disp/rootgp104.o nvkm-y += nvkm/engine/disp/channv50.o nvkm-y += nvkm/engine/disp/changf119.o nvkm-y += nvkm/engine/disp/dmacnv50.o nvkm-y += nvkm/engine/disp/dmacgf119.o +nvkm-y += nvkm/engine/disp/dmacgp104.o nvkm-y += nvkm/engine/disp/basenv50.o nvkm-y += nvkm/engine/disp/baseg84.o @@ -58,6 +64,7 @@ nvkm-y += nvkm/engine/disp/basegt215.o nvkm-y += nvkm/engine/disp/basegf119.o nvkm-y += nvkm/engine/disp/basegk104.o nvkm-y += nvkm/engine/disp/basegk110.o +nvkm-y += nvkm/engine/disp/basegp104.o nvkm-y += nvkm/engine/disp/corenv50.o nvkm-y += nvkm/engine/disp/coreg84.o @@ -69,6 +76,8 @@ nvkm-y += nvkm/engine/disp/coregk104.o nvkm-y += nvkm/engine/disp/coregk110.o nvkm-y += nvkm/engine/disp/coregm107.o nvkm-y += nvkm/engine/disp/coregm200.o +nvkm-y += nvkm/engine/disp/coregp100.o +nvkm-y += nvkm/engine/disp/coregp104.o nvkm-y += nvkm/engine/disp/ovlynv50.o nvkm-y += nvkm/engine/disp/ovlyg84.o @@ -76,6 +85,7 @@ nvkm-y += nvkm/engine/disp/ovlygt200.o nvkm-y += nvkm/engine/disp/ovlygt215.o nvkm-y += nvkm/engine/disp/ovlygf119.o nvkm-y += nvkm/engine/disp/ovlygk104.o +nvkm-y += nvkm/engine/disp/ovlygp104.o nvkm-y += nvkm/engine/disp/piocnv50.o nvkm-y += nvkm/engine/disp/piocgf119.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp104.c new file mode 100644 index 000000000000..51688e37c54e --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp104.c @@ -0,0 +1,38 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "dmacnv50.h" +#include "rootnv50.h" + +#include <nvif/class.h> + +const struct nv50_disp_dmac_oclass +gp104_disp_base_oclass = { + .base.oclass = GK110_DISP_BASE_CHANNEL_DMA, + .base.minver = 0, + .base.maxver = 0, + .ctor = nv50_disp_base_new, + .func = &gp104_disp_dmac_func, + .mthd = &gf119_disp_base_chan_mthd, + .chid = 1, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h index aee374884c96..f5f683d9fd20 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h @@ -85,6 +85,7 @@ extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_pior; extern const struct nv50_disp_chan_mthd gf119_disp_base_chan_mthd; extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd; +extern const struct nv50_disp_chan_mthd gk104_disp_ovly_chan_mthd; struct nv50_disp_pioc_oclass { int (*ctor)(const struct nv50_disp_chan_func *, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c index 6b1dc703dac7..21fbf89b6319 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c @@ -171,7 +171,7 @@ gf119_disp_core_chan_mthd = { } }; -static void +void gf119_disp_core_fini(struct nv50_disp_dmac *chan) { struct nv50_disp *disp = chan->base.root->disp; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c new file mode 100644 index 000000000000..d5dff6619d4d --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c @@ -0,0 +1,38 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "dmacnv50.h" +#include "rootnv50.h" + +#include <nvif/class.h> + +const struct nv50_disp_dmac_oclass +gp100_disp_core_oclass = { + .base.oclass = GP100_DISP_CORE_CHANNEL_DMA, + .base.minver = 0, + .base.maxver = 0, + .ctor = nv50_disp_core_new, + .func = &gf119_disp_core_func, + .mthd = &gk104_disp_core_chan_mthd, + .chid = 0, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp104.c new file mode 100644 index 000000000000..6922f4007b61 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp104.c @@ -0,0 +1,78 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "dmacnv50.h" +#include "rootnv50.h" + +#include <subdev/timer.h> + +#include <nvif/class.h> + +static int +gp104_disp_core_init(struct nv50_disp_dmac *chan) +{ + struct nv50_disp *disp = chan->base.root->disp; + struct nvkm_subdev *subdev = &disp->base.engine.subdev; + struct nvkm_device *device = subdev->device; + + /* enable error reporting */ + nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000001); + + /* initialise channel for dma command submission */ + nvkm_wr32(device, 0x611494, chan->push); + nvkm_wr32(device, 0x611498, 0x00010000); + nvkm_wr32(device, 0x61149c, 0x00000001); + nvkm_mask(device, 0x610490, 0x00000010, 0x00000010); + nvkm_wr32(device, 0x640000, 0x00000000); + nvkm_wr32(device, 0x610490, 0x01000013); + + /* wait for it to go inactive */ + if (nvkm_msec(device, 2000, + if (!(nvkm_rd32(device, 0x610490) & 0x80000000)) + break; + ) < 0) { + nvkm_error(subdev, "core init: %08x\n", + nvkm_rd32(device, 0x610490)); + return -EBUSY; + } + + return 0; +} + +const struct nv50_disp_dmac_func +gp104_disp_core_func = { + .init = gp104_disp_core_init, + .fini = gf119_disp_core_fini, + .bind = gf119_disp_dmac_bind, +}; + +const struct nv50_disp_dmac_oclass +gp104_disp_core_oclass = { + .base.oclass = GP104_DISP_CORE_CHANNEL_DMA, + .base.minver = 0, + .base.maxver = 0, + .ctor = nv50_disp_core_new, + .func = &gp104_disp_core_func, + .mthd = &gk104_disp_core_chan_mthd, + .chid = 0, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c index 876b14549a58..a57f7cef307a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c @@ -36,7 +36,7 @@ gf119_disp_dmac_bind(struct nv50_disp_dmac *chan, chan->base.chid << 27 | 0x00000001); } -static void +void gf119_disp_dmac_fini(struct nv50_disp_dmac *chan) { struct nv50_disp *disp = chan->base.root->disp; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp104.c new file mode 100644 index 000000000000..ad24c2c57696 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp104.c @@ -0,0 +1,66 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "dmacnv50.h" +#include "rootnv50.h" + +#include <subdev/timer.h> + +static int +gp104_disp_dmac_init(struct nv50_disp_dmac *chan) +{ + struct nv50_disp *disp = chan->base.root->disp; + struct nvkm_subdev *subdev = &disp->base.engine.subdev; + struct nvkm_device *device = subdev->device; + int chid = chan->base.chid; + + /* enable error reporting */ + nvkm_mask(device, 0x6100a0, 0x00000001 << chid, 0x00000001 << chid); + + /* initialise channel for dma command submission */ + nvkm_wr32(device, 0x611494 + (chid * 0x0010), chan->push); + nvkm_wr32(device, 0x611498 + (chid * 0x0010), 0x00010000); + nvkm_wr32(device, 0x61149c + (chid * 0x0010), 0x00000001); + nvkm_mask(device, 0x610490 + (chid * 0x0010), 0x00000010, 0x00000010); + nvkm_wr32(device, 0x640000 + (chid * 0x1000), 0x00000000); + nvkm_wr32(device, 0x610490 + (chid * 0x0010), 0x00000013); + + /* wait for it to go inactive */ + if (nvkm_msec(device, 2000, + if (!(nvkm_rd32(device, 0x610490 + (chid * 0x10)) & 0x80000000)) + break; + ) < 0) { + nvkm_error(subdev, "ch %d init: %08x\n", chid, + nvkm_rd32(device, 0x610490 + (chid * 0x10))); + return -EBUSY; + } + + return 0; +} + +const struct nv50_disp_dmac_func +gp104_disp_dmac_func = { + .init = gp104_disp_dmac_init, + .fini = gf119_disp_dmac_fini, + .bind = gf119_disp_dmac_bind, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h index fc84eb8b5c45..43ac05857853 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h @@ -25,8 +25,12 @@ int nv50_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32); extern const struct nv50_disp_dmac_func nv50_disp_core_func; extern const struct nv50_disp_dmac_func gf119_disp_dmac_func; +void gf119_disp_dmac_fini(struct nv50_disp_dmac *); int gf119_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32); extern const struct nv50_disp_dmac_func gf119_disp_core_func; +void gf119_disp_core_fini(struct nv50_disp_dmac *); + +extern const struct nv50_disp_dmac_func gp104_disp_dmac_func; struct nv50_disp_dmac_oclass { int (*ctor)(const struct nv50_disp_dmac_func *, @@ -88,4 +92,10 @@ extern const struct nv50_disp_dmac_oclass gk110_disp_base_oclass; extern const struct nv50_disp_dmac_oclass gm107_disp_core_oclass; extern const struct nv50_disp_dmac_oclass gm200_disp_core_oclass; + +extern const struct nv50_disp_dmac_oclass gp100_disp_core_oclass; + +extern const struct nv50_disp_dmac_oclass gp104_disp_core_oclass; +extern const struct nv50_disp_dmac_oclass gp104_disp_base_oclass; +extern const struct nv50_disp_dmac_oclass gp104_disp_ovly_oclass; #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c index f0314664349c..29e84b241cca 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c @@ -79,8 +79,7 @@ exec_lookup(struct nv50_disp *disp, int head, int or, u32 ctrl, list_for_each_entry(outp, &disp->base.outp, head) { if ((outp->info.hasht & 0xff) == type && (outp->info.hashm & mask) == mask) { - *data = nvbios_outp_match(bios, outp->info.hasht, - outp->info.hashm, + *data = nvbios_outp_match(bios, outp->info.hasht, mask, ver, hdr, cnt, len, info); if (!*data) return NULL; @@ -155,25 +154,21 @@ exec_clkcmp(struct nv50_disp *disp, int head, int id, u32 pclk, u32 *conf) if (!outp) return NULL; + *conf = (ctrl & 0x00000f00) >> 8; switch (outp->info.type) { case DCB_OUTPUT_TMDS: - *conf = (ctrl & 0x00000f00) >> 8; if (*conf == 5) *conf |= 0x0100; break; case DCB_OUTPUT_LVDS: - *conf = disp->sor.lvdsconf; + *conf |= disp->sor.lvdsconf; break; - case DCB_OUTPUT_DP: - *conf = (ctrl & 0x00000f00) >> 8; - break; - case DCB_OUTPUT_ANALOG: default: - *conf = 0x00ff; break; } - data = nvbios_ocfg_match(bios, data, *conf, &ver, &hdr, &cnt, &len, &info2); + data = nvbios_ocfg_match(bios, data, *conf & 0xff, *conf >> 8, + &ver, &hdr, &cnt, &len, &info2); if (data && id < 0xff) { data = nvbios_oclk_match(bios, info2.clkcmp[id], pclk); if (data) { @@ -418,7 +413,7 @@ gf119_disp_intr_supervisor(struct work_struct *work) nvkm_wr32(device, 0x6101d0, 0x80000000); } -static void +void gf119_disp_intr_error(struct nv50_disp *disp, int chid) { struct nvkm_subdev *subdev = &disp->base.engine.subdev; @@ -466,7 +461,7 @@ gf119_disp_intr(struct nv50_disp *disp) u32 stat = nvkm_rd32(device, 0x61009c); int chid = ffs(stat) - 1; if (chid >= 0) - gf119_disp_intr_error(disp, chid); + disp->func->intr_error(disp, chid); intr &= ~0x00000002; } @@ -510,6 +505,7 @@ gf119_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device, static const struct nv50_disp_func gf119_disp = { .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, .uevent = &gf119_disp_chan_uevent, .super = gf119_disp_intr_supervisor, .root = &gf119_disp_root_oclass, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c index a86384b8e388..37f145cf30d7 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c @@ -27,6 +27,7 @@ static const struct nv50_disp_func gk104_disp = { .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, .uevent = &gf119_disp_chan_uevent, .super = gf119_disp_intr_supervisor, .root = &gk104_disp_root_oclass, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c index 0d574c7e594a..e14ac946608c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c @@ -27,6 +27,7 @@ static const struct nv50_disp_func gk110_disp = { .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, .uevent = &gf119_disp_chan_uevent, .super = gf119_disp_intr_supervisor, .root = &gk110_disp_root_oclass, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c index b6944142d616..2f2437cc5891 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c @@ -27,6 +27,7 @@ static const struct nv50_disp_func gm107_disp = { .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, .uevent = &gf119_disp_chan_uevent, .super = gf119_disp_intr_supervisor, .root = &gm107_disp_root_oclass, @@ -36,7 +37,7 @@ gm107_disp = { .outp.internal.crt = nv50_dac_output_new, .outp.internal.tmds = nv50_sor_output_new, .outp.internal.lvds = nv50_sor_output_new, - .outp.internal.dp = gf119_sor_dp_new, + .outp.internal.dp = gm107_sor_dp_new, .dac.nr = 3, .dac.power = nv50_dac_power, .dac.sense = nv50_dac_sense, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c index 67eec8620719..9f368d4ee61e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c @@ -27,6 +27,7 @@ static const struct nv50_disp_func gm200_disp = { .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, .uevent = &gf119_disp_chan_uevent, .super = gf119_disp_intr_supervisor, .root = &gm200_disp_root_oclass, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c new file mode 100644 index 000000000000..4f81bf31435e --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c @@ -0,0 +1,55 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "nv50.h" +#include "rootnv50.h" + +static const struct nv50_disp_func +gp100_disp = { + .intr = gf119_disp_intr, + .intr_error = gf119_disp_intr_error, + .uevent = &gf119_disp_chan_uevent, + .super = gf119_disp_intr_supervisor, + .root = &gp100_disp_root_oclass, + .head.vblank_init = gf119_disp_vblank_init, + .head.vblank_fini = gf119_disp_vblank_fini, + .head.scanoutpos = gf119_disp_root_scanoutpos, + .outp.internal.crt = nv50_dac_output_new, + .outp.internal.tmds = nv50_sor_output_new, + .outp.internal.lvds = nv50_sor_output_new, + .outp.internal.dp = gm200_sor_dp_new, + .dac.nr = 3, + .dac.power = nv50_dac_power, + .dac.sense = nv50_dac_sense, + .sor.nr = 4, + .sor.power = nv50_sor_power, + .sor.hda_eld = gf119_hda_eld, + .sor.hdmi = gk104_hdmi_ctrl, + .sor.magic = gm200_sor_magic, +}; + +int +gp100_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp) +{ + return gf119_disp_new_(&gp100_disp, device, index, pdisp); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp104.c new file mode 100644 index 000000000000..3bf3380336e4 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp104.c @@ -0,0 +1,81 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "nv50.h" +#include "rootnv50.h" + +static void +gp104_disp_intr_error(struct nv50_disp *disp, int chid) +{ + struct nvkm_subdev *subdev = &disp->base.engine.subdev; + struct nvkm_device *device = subdev->device; + u32 mthd = nvkm_rd32(device, 0x6111f0 + (chid * 12)); + u32 data = nvkm_rd32(device, 0x6111f4 + (chid * 12)); + u32 unkn = nvkm_rd32(device, 0x6111f8 + (chid * 12)); + + nvkm_error(subdev, "chid %d mthd %04x data %08x %08x %08x\n", + chid, (mthd & 0x0000ffc), data, mthd, unkn); + + if (chid < ARRAY_SIZE(disp->chan)) { + switch (mthd & 0xffc) { + case 0x0080: + nv50_disp_chan_mthd(disp->chan[chid], NV_DBG_ERROR); + break; + default: + break; + } + } + + nvkm_wr32(device, 0x61009c, (1 << chid)); + nvkm_wr32(device, 0x6111f0 + (chid * 12), 0x90000000); +} + +static const struct nv50_disp_func +gp104_disp = { + .intr = gf119_disp_intr, + .intr_error = gp104_disp_intr_error, + .uevent = &gf119_disp_chan_uevent, + .super = gf119_disp_intr_supervisor, + .root = &gp104_disp_root_oclass, + .head.vblank_init = gf119_disp_vblank_init, + .head.vblank_fini = gf119_disp_vblank_fini, + .head.scanoutpos = gf119_disp_root_scanoutpos, + .outp.internal.crt = nv50_dac_output_new, + .outp.internal.tmds = nv50_sor_output_new, + .outp.internal.lvds = nv50_sor_output_new, + .outp.internal.dp = gm200_sor_dp_new, + .dac.nr = 3, + .dac.power = nv50_dac_power, + .dac.sense = nv50_dac_sense, + .sor.nr = 4, + .sor.power = nv50_sor_power, + .sor.hda_eld = gf119_hda_eld, + .sor.hdmi = gk104_hdmi_ctrl, + .sor.magic = gm200_sor_magic, +}; + +int +gp104_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp) +{ + return gf119_disp_new_(&gp104_disp, device, index, pdisp); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c index 4226d2153b9c..fbb8c7dc18fd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c @@ -32,6 +32,7 @@ #include <subdev/bios/init.h> #include <subdev/bios/pll.h> #include <subdev/devinit.h> +#include <subdev/timer.h> static const struct nvkm_disp_oclass * nv50_disp_root_(struct nvkm_disp *base) @@ -269,8 +270,7 @@ exec_lookup(struct nv50_disp *disp, int head, int or, u32 ctrl, list_for_each_entry(outp, &disp->base.outp, head) { if ((outp->info.hasht & 0xff) == type && (outp->info.hashm & mask) == mask) { - *data = nvbios_outp_match(bios, outp->info.hasht, - outp->info.hashm, + *data = nvbios_outp_match(bios, outp->info.hasht, mask, ver, hdr, cnt, len, info); if (!*data) return NULL; @@ -387,22 +387,17 @@ exec_clkcmp(struct nv50_disp *disp, int head, int id, u32 pclk, u32 *conf) if (!outp) return NULL; + *conf = (ctrl & 0x00000f00) >> 8; if (outp->info.location == 0) { switch (outp->info.type) { case DCB_OUTPUT_TMDS: - *conf = (ctrl & 0x00000f00) >> 8; if (*conf == 5) *conf |= 0x0100; break; case DCB_OUTPUT_LVDS: - *conf = disp->sor.lvdsconf; + *conf |= disp->sor.lvdsconf; break; - case DCB_OUTPUT_DP: - *conf = (ctrl & 0x00000f00) >> 8; - break; - case DCB_OUTPUT_ANALOG: default: - *conf = 0x00ff; break; } } else { @@ -410,7 +405,8 @@ exec_clkcmp(struct nv50_disp *disp, int head, int id, u32 pclk, u32 *conf) pclk = pclk / 2; } - data = nvbios_ocfg_match(bios, data, *conf, &ver, &hdr, &cnt, &len, &info2); + data = nvbios_ocfg_match(bios, data, *conf & 0xff, *conf >> 8, + &ver, &hdr, &cnt, &len, &info2); if (data && id < 0xff) { data = nvbios_oclk_match(bios, info2.clkcmp[id], pclk); if (data) { @@ -430,6 +426,134 @@ exec_clkcmp(struct nv50_disp *disp, int head, int id, u32 pclk, u32 *conf) return outp; } +static bool +nv50_disp_dptmds_war(struct nvkm_device *device) +{ + switch (device->chipset) { + case 0x94: + case 0x96: + case 0x98: + case 0xaa: + case 0xac: + return true; + default: + break; + } + return false; +} + +static bool +nv50_disp_dptmds_war_needed(struct nv50_disp *disp, struct dcb_output *outp) +{ + struct nvkm_device *device = disp->base.engine.subdev.device; + const u32 soff = __ffs(outp->or) * 0x800; + if (nv50_disp_dptmds_war(device) && outp->type == DCB_OUTPUT_TMDS) { + switch (nvkm_rd32(device, 0x614300 + soff) & 0x00030000) { + case 0x00000000: + case 0x00030000: + return true; + default: + break; + } + } + return false; + +} + +static void +nv50_disp_dptmds_war_2(struct nv50_disp *disp, struct dcb_output *outp) +{ + struct nvkm_device *device = disp->base.engine.subdev.device; + const u32 soff = __ffs(outp->or) * 0x800; + + if (!nv50_disp_dptmds_war_needed(disp, outp)) + return; + + nvkm_mask(device, 0x00e840, 0x80000000, 0x80000000); + nvkm_mask(device, 0x614300 + soff, 0x03000000, 0x03000000); + nvkm_mask(device, 0x61c10c + soff, 0x00000001, 0x00000001); + + nvkm_mask(device, 0x61c00c + soff, 0x0f000000, 0x00000000); + nvkm_mask(device, 0x61c008 + soff, 0xff000000, 0x14000000); + nvkm_usec(device, 400, NVKM_DELAY); + nvkm_mask(device, 0x61c008 + soff, 0xff000000, 0x00000000); + nvkm_mask(device, 0x61c00c + soff, 0x0f000000, 0x01000000); + + if (nvkm_rd32(device, 0x61c004 + soff) & 0x00000001) { + u32 seqctl = nvkm_rd32(device, 0x61c030 + soff); + u32 pu_pc = seqctl & 0x0000000f; + nvkm_wr32(device, 0x61c040 + soff + pu_pc * 4, 0x1f008000); + } +} + +static void +nv50_disp_dptmds_war_3(struct nv50_disp *disp, struct dcb_output *outp) +{ + struct nvkm_device *device = disp->base.engine.subdev.device; + const u32 soff = __ffs(outp->or) * 0x800; + u32 sorpwr; + + if (!nv50_disp_dptmds_war_needed(disp, outp)) + return; + + sorpwr = nvkm_rd32(device, 0x61c004 + soff); + if (sorpwr & 0x00000001) { + u32 seqctl = nvkm_rd32(device, 0x61c030 + soff); + u32 pd_pc = (seqctl & 0x00000f00) >> 8; + u32 pu_pc = seqctl & 0x0000000f; + + nvkm_wr32(device, 0x61c040 + soff + pd_pc * 4, 0x1f008000); + + nvkm_msec(device, 2000, + if (!(nvkm_rd32(device, 0x61c030 + soff) & 0x10000000)) + break; + ); + nvkm_mask(device, 0x61c004 + soff, 0x80000001, 0x80000000); + nvkm_msec(device, 2000, + if (!(nvkm_rd32(device, 0x61c030 + soff) & 0x10000000)) + break; + ); + + nvkm_wr32(device, 0x61c040 + soff + pd_pc * 4, 0x00002000); + nvkm_wr32(device, 0x61c040 + soff + pu_pc * 4, 0x1f000000); + } + + nvkm_mask(device, 0x61c10c + soff, 0x00000001, 0x00000000); + nvkm_mask(device, 0x614300 + soff, 0x03000000, 0x00000000); + + if (sorpwr & 0x00000001) { + nvkm_mask(device, 0x61c004 + soff, 0x80000001, 0x80000001); + } +} + +static void +nv50_disp_update_sppll1(struct nv50_disp *disp) +{ + struct nvkm_device *device = disp->base.engine.subdev.device; + bool used = false; + int sor; + + if (!nv50_disp_dptmds_war(device)) + return; + + for (sor = 0; sor < disp->func->sor.nr; sor++) { + u32 clksor = nvkm_rd32(device, 0x614300 + (sor * 0x800)); + switch (clksor & 0x03000000) { + case 0x02000000: + case 0x03000000: + used = true; + break; + default: + break; + } + } + + if (used) + return; + + nvkm_mask(device, 0x00e840, 0x80000000, 0x00000000); +} + static void nv50_disp_intr_unk10_0(struct nv50_disp *disp, int head) { @@ -683,6 +807,8 @@ nv50_disp_intr_unk20_2(struct nv50_disp *disp, int head) nvkm_mask(device, hreg, 0x0000000f, hval); nvkm_mask(device, oreg, mask, oval); + + nv50_disp_dptmds_war_2(disp, &outp->info); } /* If programming a TMDS output on a SOR that can also be configured for @@ -724,6 +850,7 @@ nv50_disp_intr_unk40_0(struct nv50_disp *disp, int head) if (outp->info.location == 0 && outp->info.type == DCB_OUTPUT_TMDS) nv50_disp_intr_unk40_0_tmds(disp, &outp->info); + nv50_disp_dptmds_war_3(disp, &outp->info); } void @@ -771,6 +898,7 @@ nv50_disp_intr_supervisor(struct work_struct *work) continue; nv50_disp_intr_unk40_0(disp, head); } + nv50_disp_update_sppll1(disp); } nvkm_wr32(device, 0x610030, 0x80000000); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h index aecebd8717e5..1e1de6bfe85a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h @@ -68,6 +68,7 @@ struct nv50_disp_func_outp { struct nv50_disp_func { void (*intr)(struct nv50_disp *); + void (*intr_error)(struct nv50_disp *, int chid); const struct nvkm_event_func *uevent; void (*super)(struct work_struct *); @@ -114,4 +115,5 @@ void gf119_disp_vblank_init(struct nv50_disp *, int); void gf119_disp_vblank_fini(struct nv50_disp *, int); void gf119_disp_intr(struct nv50_disp *); void gf119_disp_intr_supervisor(struct work_struct *); +void gf119_disp_intr_error(struct nv50_disp *, int); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outpdp.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outpdp.h index e9067ba4e179..4e983f6d7032 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/outpdp.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/outpdp.h @@ -62,7 +62,12 @@ int g94_sor_dp_lnk_pwr(struct nvkm_output_dp *, int); int gf119_sor_dp_new(struct nvkm_disp *, int, struct dcb_output *, struct nvkm_output **); int gf119_sor_dp_lnk_ctl(struct nvkm_output_dp *, int, int, bool); +int gf119_sor_dp_drv_ctl(struct nvkm_output_dp *, int, int, int, int); -int gm200_sor_dp_new(struct nvkm_disp *, int, struct dcb_output *, - struct nvkm_output **); +int gm107_sor_dp_new(struct nvkm_disp *, int, struct dcb_output *, + struct nvkm_output **); +int gm107_sor_dp_pattern(struct nvkm_output_dp *, int); + +int gm200_sor_dp_new(struct nvkm_disp *, int, struct dcb_output *, + struct nvkm_output **); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c index 2e2dc0641ef2..2f0220b39f34 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c @@ -80,7 +80,7 @@ gk104_disp_ovly_mthd_base = { } }; -static const struct nv50_disp_chan_mthd +const struct nv50_disp_chan_mthd gk104_disp_ovly_chan_mthd = { .name = "Overlay", .addr = 0x001000, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp104.c new file mode 100644 index 000000000000..97e2dd2d908e --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp104.c @@ -0,0 +1,38 @@ +/* + * Copyright 2012 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "dmacnv50.h" +#include "rootnv50.h" + +#include <nvif/class.h> + +const struct nv50_disp_dmac_oclass +gp104_disp_ovly_oclass = { + .base.oclass = GK104_DISP_OVERLAY_CONTROL_DMA, + .base.minver = 0, + .base.maxver = 0, + .ctor = nv50_disp_ovly_new, + .func = &gp104_disp_dmac_func, + .mthd = &gk104_disp_ovly_chan_mthd, + .chid = 5, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c new file mode 100644 index 000000000000..ac8fdd728ec6 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c @@ -0,0 +1,58 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "rootnv50.h" +#include "dmacnv50.h" + +#include <nvif/class.h> + +static const struct nv50_disp_root_func +gp100_disp_root = { + .init = gf119_disp_root_init, + .fini = gf119_disp_root_fini, + .dmac = { + &gp100_disp_core_oclass, + &gk110_disp_base_oclass, + &gk104_disp_ovly_oclass, + }, + .pioc = { + &gk104_disp_oimm_oclass, + &gk104_disp_curs_oclass, + }, +}; + +static int +gp100_disp_root_new(struct nvkm_disp *disp, const struct nvkm_oclass *oclass, + void *data, u32 size, struct nvkm_object **pobject) +{ + return nv50_disp_root_new_(&gp100_disp_root, disp, oclass, + data, size, pobject); +} + +const struct nvkm_disp_oclass +gp100_disp_root_oclass = { + .base.oclass = GP100_DISP, + .base.minver = -1, + .base.maxver = -1, + .ctor = gp100_disp_root_new, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp104.c new file mode 100644 index 000000000000..8443e04dc626 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp104.c @@ -0,0 +1,58 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "rootnv50.h" +#include "dmacnv50.h" + +#include <nvif/class.h> + +static const struct nv50_disp_root_func +gp104_disp_root = { + .init = gf119_disp_root_init, + .fini = gf119_disp_root_fini, + .dmac = { + &gp104_disp_core_oclass, + &gp104_disp_base_oclass, + &gp104_disp_ovly_oclass, + }, + .pioc = { + &gk104_disp_oimm_oclass, + &gk104_disp_curs_oclass, + }, +}; + +static int +gp104_disp_root_new(struct nvkm_disp *disp, const struct nvkm_oclass *oclass, + void *data, u32 size, struct nvkm_object **pobject) +{ + return nv50_disp_root_new_(&gp104_disp_root, disp, oclass, + data, size, pobject); +} + +const struct nvkm_disp_oclass +gp104_disp_root_oclass = { + .base.oclass = GP104_DISP, + .base.minver = -1, + .base.maxver = -1, + .ctor = gp104_disp_root_new, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h index cb449ed8d92c..ad00f1724b72 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h @@ -40,4 +40,6 @@ extern const struct nvkm_disp_oclass gk104_disp_root_oclass; extern const struct nvkm_disp_oclass gk110_disp_root_oclass; extern const struct nvkm_disp_oclass gm107_disp_root_oclass; extern const struct nvkm_disp_oclass gm200_disp_root_oclass; +extern const struct nvkm_disp_oclass gp100_disp_root_oclass; +extern const struct nvkm_disp_oclass gp104_disp_root_oclass; #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c index b4b41b135643..22706c0a54b5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c @@ -40,8 +40,7 @@ static int gf119_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern) { struct nvkm_device *device = outp->base.disp->engine.subdev.device; - const u32 loff = gf119_sor_loff(outp); - nvkm_mask(device, 0x61c110 + loff, 0x0f0f0f0f, 0x01010101 * pattern); + nvkm_mask(device, 0x61c110, 0x0f0f0f0f, 0x01010101 * pattern); return 0; } @@ -64,7 +63,7 @@ gf119_sor_dp_lnk_ctl(struct nvkm_output_dp *outp, int nr, int bw, bool ef) return 0; } -static int +int gf119_sor_dp_drv_ctl(struct nvkm_output_dp *outp, int ln, int vs, int pe, int pc) { diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c new file mode 100644 index 000000000000..37790b2617c5 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c @@ -0,0 +1,53 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "nv50.h" +#include "outpdp.h" + +int +gm107_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern) +{ + struct nvkm_device *device = outp->base.disp->engine.subdev.device; + const u32 soff = outp->base.or * 0x800; + const u32 data = 0x01010101 * pattern; + if (outp->base.info.sorconf.link & 1) + nvkm_mask(device, 0x61c110 + soff, 0x0f0f0f0f, data); + else + nvkm_mask(device, 0x61c12c + soff, 0x0f0f0f0f, data); + return 0; +} + +static const struct nvkm_output_dp_func +gm107_sor_dp_func = { + .pattern = gm107_sor_dp_pattern, + .lnk_pwr = g94_sor_dp_lnk_pwr, + .lnk_ctl = gf119_sor_dp_lnk_ctl, + .drv_ctl = gf119_sor_dp_drv_ctl, +}; + +int +gm107_sor_dp_new(struct nvkm_disp *disp, int index, + struct dcb_output *dcbE, struct nvkm_output **poutp) +{ + return nvkm_output_dp_new_(&gm107_sor_dp_func, disp, index, dcbE, poutp); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c index 2cfbef9c344f..c44fa7ea672a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c @@ -57,19 +57,6 @@ gm200_sor_dp_lane_map(struct nvkm_device *device, u8 lane) } static int -gm200_sor_dp_pattern(struct nvkm_output_dp *outp, int pattern) -{ - struct nvkm_device *device = outp->base.disp->engine.subdev.device; - const u32 soff = gm200_sor_soff(outp); - const u32 data = 0x01010101 * pattern; - if (outp->base.info.sorconf.link & 1) - nvkm_mask(device, 0x61c110 + soff, 0x0f0f0f0f, data); - else - nvkm_mask(device, 0x61c12c + soff, 0x0f0f0f0f, data); - return 0; -} - -static int gm200_sor_dp_lnk_pwr(struct nvkm_output_dp *outp, int nr) { struct nvkm_device *device = outp->base.disp->engine.subdev.device; @@ -129,7 +116,7 @@ gm200_sor_dp_drv_ctl(struct nvkm_output_dp *outp, static const struct nvkm_output_dp_func gm200_sor_dp_func = { - .pattern = gm200_sor_dp_pattern, + .pattern = gm107_sor_dp_pattern, .lnk_pwr = gm200_sor_dp_lnk_pwr, .lnk_ctl = gf119_sor_dp_lnk_ctl, .drv_ctl = gm200_sor_dp_drv_ctl, diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild index 65e5d291ecda..98651a43bc12 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild @@ -13,6 +13,7 @@ nvkm-y += nvkm/engine/fifo/gk20a.o nvkm-y += nvkm/engine/fifo/gm107.o nvkm-y += nvkm/engine/fifo/gm200.o nvkm-y += nvkm/engine/fifo/gm20b.o +nvkm-y += nvkm/engine/fifo/gp100.o nvkm-y += nvkm/engine/fifo/chan.o nvkm-y += nvkm/engine/fifo/channv50.o @@ -31,3 +32,4 @@ nvkm-y += nvkm/engine/fifo/gpfifogf100.o nvkm-y += nvkm/engine/fifo/gpfifogk104.o nvkm-y += nvkm/engine/fifo/gpfifogk110.o nvkm-y += nvkm/engine/fifo/gpfifogm200.o +nvkm-y += nvkm/engine/fifo/gpfifogp100.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h index e06f4d46f802..230f64e5f731 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h @@ -27,4 +27,5 @@ int gk104_fifo_gpfifo_new(struct nvkm_fifo *, const struct nvkm_oclass *, extern const struct nvkm_fifo_chan_oclass gk104_fifo_gpfifo_oclass; extern const struct nvkm_fifo_chan_oclass gk110_fifo_gpfifo_oclass; extern const struct nvkm_fifo_chan_oclass gm200_fifo_gpfifo_oclass; +extern const struct nvkm_fifo_chan_oclass gp100_fifo_gpfifo_oclass; #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c index 743f3a189f28..103c0afaaa6d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c @@ -329,7 +329,7 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit) } if (eu == NULL) { - enum nvkm_devidx engidx = nvkm_top_fault(device->top, unit); + enum nvkm_devidx engidx = nvkm_top_fault(device, unit); if (engidx < NVKM_SUBDEV_NR) { const char *src = nvkm_subdev_name[engidx]; char *dst = en; @@ -589,7 +589,6 @@ gk104_fifo_oneinit(struct nvkm_fifo *base) struct gk104_fifo *fifo = gk104_fifo(base); struct nvkm_subdev *subdev = &fifo->base.engine.subdev; struct nvkm_device *device = subdev->device; - struct nvkm_top *top = device->top; int engn, runl, pbid, ret, i, j; enum nvkm_devidx engidx; u32 *map; @@ -608,7 +607,7 @@ gk104_fifo_oneinit(struct nvkm_fifo *base) /* Determine runlist configuration from topology device info. */ i = 0; - while ((int)(engidx = nvkm_top_engine(top, i++, &runl, &engn)) >= 0) { + while ((int)(engidx = nvkm_top_engine(device, i++, &runl, &engn)) >= 0) { /* Determine which PBDMA handles requests for this engine. */ for (j = 0, pbid = -1; j < fifo->pbdma_nr; j++) { if (map[j] & (1 << runl)) { @@ -617,8 +616,8 @@ gk104_fifo_oneinit(struct nvkm_fifo *base) } } - nvkm_debug(subdev, "engine %2d: runlist %2d pbdma %2d\n", - engn, runl, pbid); + nvkm_debug(subdev, "engine %2d: runlist %2d pbdma %2d (%s)\n", + engn, runl, pbid, nvkm_subdev_name[engidx]); fifo->engine[engn].engine = nvkm_device_engine(device, engidx); fifo->engine[engn].runl = runl; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c new file mode 100644 index 000000000000..eff83f7fb705 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c @@ -0,0 +1,67 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "gk104.h" +#include "changk104.h" + +static const struct nvkm_enum +gp100_fifo_fault_engine[] = { + { 0x01, "DISPLAY" }, + { 0x03, "IFB", NULL, NVKM_ENGINE_IFB }, + { 0x04, "BAR1", NULL, NVKM_SUBDEV_BAR }, + { 0x05, "BAR2", NULL, NVKM_SUBDEV_INSTMEM }, + { 0x06, "HOST0" }, + { 0x07, "HOST1" }, + { 0x08, "HOST2" }, + { 0x09, "HOST3" }, + { 0x0a, "HOST4" }, + { 0x0b, "HOST5" }, + { 0x0c, "HOST6" }, + { 0x0d, "HOST7" }, + { 0x0e, "HOST8" }, + { 0x0f, "HOST9" }, + { 0x10, "HOST10" }, + { 0x13, "PERF" }, + { 0x17, "PMU" }, + { 0x18, "PTP" }, + { 0x1f, "PHYSICAL" }, + {} +}; + +static const struct gk104_fifo_func +gp100_fifo = { + .fault.engine = gp100_fifo_fault_engine, + .fault.reason = gk104_fifo_fault_reason, + .fault.hubclient = gk104_fifo_fault_hubclient, + .fault.gpcclient = gk104_fifo_fault_gpcclient, + .chan = { + &gp100_fifo_gpfifo_oclass, + NULL + }, +}; + +int +gp100_fifo_new(struct nvkm_device *device, int index, struct nvkm_fifo **pfifo) +{ + return gk104_fifo_new_(&gp100_fifo, device, index, 4096, pfifo); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c new file mode 100644 index 000000000000..1530a9217aea --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c @@ -0,0 +1,34 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "changk104.h" + +#include <nvif/class.h> + +const struct nvkm_fifo_chan_oclass +gp100_fifo_gpfifo_oclass = { + .base.oclass = PASCAL_CHANNEL_GPFIFO_A, + .base.minver = 0, + .base.maxver = 0, + .ctor = gk104_fifo_gpfifo_new, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild index 290ed0db8047..f1c494182248 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild @@ -31,6 +31,7 @@ nvkm-y += nvkm/engine/gr/gk20a.o nvkm-y += nvkm/engine/gr/gm107.o nvkm-y += nvkm/engine/gr/gm200.o nvkm-y += nvkm/engine/gr/gm20b.o +nvkm-y += nvkm/engine/gr/gp100.o nvkm-y += nvkm/engine/gr/ctxnv40.o nvkm-y += nvkm/engine/gr/ctxnv50.o @@ -48,3 +49,4 @@ nvkm-y += nvkm/engine/gr/ctxgk20a.o nvkm-y += nvkm/engine/gr/ctxgm107.o nvkm-y += nvkm/engine/gr/ctxgm200.o nvkm-y += nvkm/engine/gr/ctxgm20b.o +nvkm-y += nvkm/engine/gr/ctxgp100.o diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c index b02d8f50ea6a..bc77eea351a5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c @@ -1240,7 +1240,7 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) const struct gf100_grctx_func *grctx = gr->func->grctx; u32 idle_timeout; - nvkm_mc_unk260(device->mc, 0); + nvkm_mc_unk260(device, 0); gf100_gr_mmio(gr, grctx->hub); gf100_gr_mmio(gr, grctx->gpc); @@ -1264,7 +1264,7 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) gf100_gr_icmd(gr, grctx->icmd); nvkm_wr32(device, 0x404154, idle_timeout); gf100_gr_mthd(gr, grctx->mthd); - nvkm_mc_unk260(device->mc, 1); + nvkm_mc_unk260(device, 1); } int diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h index ac895edce164..52048b5a5274 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h @@ -101,6 +101,8 @@ void gm200_grctx_generate_405b60(struct gf100_gr *); extern const struct gf100_grctx_func gm20b_grctx; +extern const struct gf100_grctx_func gp100_grctx; + /* context init value lists */ extern const struct gf100_gr_pack gf100_grctx_pack_icmd[]; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c index f521de11a299..c925ade5880e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c @@ -226,7 +226,7 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) u32 idle_timeout; int i; - nvkm_mc_unk260(device->mc, 0); + nvkm_mc_unk260(device, 0); gf100_gr_mmio(gr, grctx->hub); gf100_gr_mmio(gr, grctx->gpc); @@ -253,7 +253,7 @@ gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) gf100_gr_icmd(gr, grctx->icmd); nvkm_wr32(device, 0x404154, idle_timeout); gf100_gr_mthd(gr, grctx->mthd); - nvkm_mc_unk260(device->mc, 1); + nvkm_mc_unk260(device, 1); } const struct gf100_grctx_func diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c index 9ba337778ef5..c46b3fdf7203 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c @@ -950,7 +950,7 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) u32 idle_timeout; int i; - nvkm_mc_unk260(device->mc, 0); + nvkm_mc_unk260(device, 0); gf100_gr_mmio(gr, grctx->hub); gf100_gr_mmio(gr, grctx->gpc); @@ -979,7 +979,7 @@ gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) gf100_gr_icmd(gr, grctx->icmd); nvkm_wr32(device, 0x404154, idle_timeout); gf100_gr_mthd(gr, grctx->mthd); - nvkm_mc_unk260(device->mc, 1); + nvkm_mc_unk260(device, 1); nvkm_mask(device, 0x418800, 0x00200000, 0x00200000); nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c new file mode 100644 index 000000000000..3d1ae7ddf7dd --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c @@ -0,0 +1,179 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "ctxgf100.h" + +#include <subdev/fb.h> + +/******************************************************************************* + * PGRAPH context implementation + ******************************************************************************/ + +static void +gp100_grctx_generate_pagepool(struct gf100_grctx *info) +{ + const struct gf100_grctx_func *grctx = info->gr->func->grctx; + const u32 access = NV_MEM_ACCESS_RW | NV_MEM_ACCESS_SYS; + const int s = 8; + const int b = mmio_vram(info, grctx->pagepool_size, (1 << s), access); + mmio_refn(info, 0x40800c, 0x00000000, s, b); + mmio_wr32(info, 0x408010, 0x80000000); + mmio_refn(info, 0x419004, 0x00000000, s, b); + mmio_wr32(info, 0x419008, 0x00000000); +} + +static void +gp100_grctx_generate_attrib(struct gf100_grctx *info) +{ + struct gf100_gr *gr = info->gr; + const struct gf100_grctx_func *grctx = gr->func->grctx; + const u32 alpha = grctx->alpha_nr; + const u32 attrib = grctx->attrib_nr; + const u32 pertpc = 0x20 * (grctx->attrib_nr_max + grctx->alpha_nr_max); + const u32 size = roundup(gr->tpc_total * pertpc, 0x80); + const u32 access = NV_MEM_ACCESS_RW; + const int s = 12; + const int b = mmio_vram(info, size, (1 << s), access); + const int max_batches = 0xffff; + u32 ao = 0; + u32 bo = ao + grctx->alpha_nr_max * gr->tpc_total; + int gpc, ppc, n = 0; + + mmio_refn(info, 0x418810, 0x80000000, s, b); + mmio_refn(info, 0x419848, 0x10000000, s, b); + mmio_refn(info, 0x419c2c, 0x10000000, s, b); + mmio_refn(info, 0x419b00, 0x00000000, s, b); + mmio_wr32(info, 0x419b04, 0x80000000 | size >> 7); + mmio_wr32(info, 0x405830, attrib); + mmio_wr32(info, 0x40585c, alpha); + mmio_wr32(info, 0x4064c4, ((alpha / 4) << 16) | max_batches); + + for (gpc = 0; gpc < gr->gpc_nr; gpc++) { + for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) { + const u32 as = alpha * gr->ppc_tpc_nr[gpc][ppc]; + const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc]; + const u32 u = 0x418ea0 + (n * 0x04); + const u32 o = PPC_UNIT(gpc, ppc, 0); + if (!(gr->ppc_mask[gpc] & (1 << ppc))) + continue; + mmio_wr32(info, o + 0xc0, bs); + mmio_wr32(info, o + 0xf4, bo); + mmio_wr32(info, o + 0xf0, bs); + bo += grctx->attrib_nr_max * gr->ppc_tpc_nr[gpc][ppc]; + mmio_wr32(info, o + 0xe4, as); + mmio_wr32(info, o + 0xf8, ao); + ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc]; + mmio_wr32(info, u, bs); + } + } + + mmio_wr32(info, 0x418eec, 0x00000000); + mmio_wr32(info, 0x41befc, 0x00000000); +} + +static void +gp100_grctx_generate_405b60(struct gf100_gr *gr) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4); + u32 dist[TPC_MAX / 4] = {}; + u32 gpcs[GPC_MAX * 2] = {}; + u8 tpcnr[GPC_MAX]; + int tpc, gpc, i; + + memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr)); + + /* won't result in the same distribution as the binary driver where + * some of the gpcs have more tpcs than others, but this shall do + * for the moment. the code for earlier gpus has this issue too. + */ + for (gpc = -1, i = 0; i < gr->tpc_total; i++) { + do { + gpc = (gpc + 1) % gr->gpc_nr; + } while(!tpcnr[gpc]); + tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--; + + dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8); + gpcs[gpc + (gr->gpc_nr * (tpc / 4))] |= i << (tpc * 8); + } + + for (i = 0; i < dist_nr; i++) + nvkm_wr32(device, 0x405b60 + (i * 4), dist[i]); + for (i = 0; i < gr->gpc_nr * 2; i++) + nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]); +} + +static void +gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + const struct gf100_grctx_func *grctx = gr->func->grctx; + u32 idle_timeout, tmp; + int i; + + gf100_gr_mmio(gr, gr->fuc_sw_ctx); + + idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000); + + grctx->pagepool(info); + grctx->bundle(info); + grctx->attrib(info); + grctx->unkn(gr); + + gm200_grctx_generate_tpcid(gr); + gf100_grctx_generate_r406028(gr); + gk104_grctx_generate_r418bb8(gr); + + for (i = 0; i < 8; i++) + nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000); + nvkm_wr32(device, 0x406500, 0x00000000); + + nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr); + + for (tmp = 0, i = 0; i < gr->gpc_nr; i++) + tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 5); + nvkm_wr32(device, 0x4041c4, tmp); + + gp100_grctx_generate_405b60(gr); + + gf100_gr_icmd(gr, gr->fuc_bundle); + nvkm_wr32(device, 0x404154, idle_timeout); + gf100_gr_mthd(gr, gr->fuc_method); +} + +const struct gf100_grctx_func +gp100_grctx = { + .main = gp100_grctx_generate_main, + .unkn = gk104_grctx_generate_unkn, + .bundle = gm107_grctx_generate_bundle, + .bundle_size = 0x3000, + .bundle_min_gpm_fifo_depth = 0x180, + .bundle_token_limit = 0x1080, + .pagepool = gp100_grctx_generate_pagepool, + .pagepool_size = 0x20000, + .attrib = gp100_grctx_generate_attrib, + .attrib_nr_max = 0x660, + .attrib_nr = 0x440, + .alpha_nr_max = 0xc00, + .alpha_nr = 0x800, +}; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c index 9513badb8220..157919c788e6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c @@ -949,22 +949,41 @@ gf100_gr_trap_gpc_rop(struct gf100_gr *gr, int gpc) } static const struct nvkm_enum gf100_mp_warp_error[] = { - { 0x00, "NO_ERROR" }, - { 0x01, "STACK_MISMATCH" }, + { 0x01, "STACK_ERROR" }, + { 0x02, "API_STACK_ERROR" }, + { 0x03, "RET_EMPTY_STACK_ERROR" }, + { 0x04, "PC_WRAP" }, { 0x05, "MISALIGNED_PC" }, - { 0x08, "MISALIGNED_GPR" }, - { 0x09, "INVALID_OPCODE" }, - { 0x0d, "GPR_OUT_OF_BOUNDS" }, - { 0x0e, "MEM_OUT_OF_BOUNDS" }, - { 0x0f, "UNALIGNED_MEM_ACCESS" }, + { 0x06, "PC_OVERFLOW" }, + { 0x07, "MISALIGNED_IMMC_ADDR" }, + { 0x08, "MISALIGNED_REG" }, + { 0x09, "ILLEGAL_INSTR_ENCODING" }, + { 0x0a, "ILLEGAL_SPH_INSTR_COMBO" }, + { 0x0b, "ILLEGAL_INSTR_PARAM" }, + { 0x0c, "INVALID_CONST_ADDR" }, + { 0x0d, "OOR_REG" }, + { 0x0e, "OOR_ADDR" }, + { 0x0f, "MISALIGNED_ADDR" }, { 0x10, "INVALID_ADDR_SPACE" }, - { 0x11, "INVALID_PARAM" }, + { 0x11, "ILLEGAL_INSTR_PARAM2" }, + { 0x12, "INVALID_CONST_ADDR_LDC" }, + { 0x13, "GEOMETRY_SM_ERROR" }, + { 0x14, "DIVERGENT" }, + { 0x15, "WARP_EXIT" }, {} }; static const struct nvkm_bitfield gf100_mp_global_error[] = { + { 0x00000001, "SM_TO_SM_FAULT" }, + { 0x00000002, "L1_ERROR" }, { 0x00000004, "MULTIPLE_WARP_ERRORS" }, - { 0x00000008, "OUT_OF_STACK_SPACE" }, + { 0x00000008, "PHYSICAL_STACK_OVERFLOW" }, + { 0x00000010, "BPT_INT" }, + { 0x00000020, "BPT_PAUSE" }, + { 0x00000040, "SINGLE_STEP_COMPLETE" }, + { 0x20000000, "ECC_SEC_ERROR" }, + { 0x40000000, "ECC_DED_ERROR" }, + { 0x80000000, "TIMEOUT" }, {} }; @@ -1438,24 +1457,30 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr) struct nvkm_device *device = subdev->device; struct nvkm_secboot *sb = device->secboot; int i; + int ret = 0; if (gr->firmware) { /* load fuc microcode */ - nvkm_mc_unk260(device->mc, 0); + nvkm_mc_unk260(device, 0); /* securely-managed falcons must be reset using secure boot */ if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_FECS)) - nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_FECS); + ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_FECS); else gf100_gr_init_fw(gr, 0x409000, &gr->fuc409c, &gr->fuc409d); + if (ret) + return ret; + if (nvkm_secboot_is_managed(sb, NVKM_SECBOOT_FALCON_GPCCS)) - nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_GPCCS); + ret = nvkm_secboot_reset(sb, NVKM_SECBOOT_FALCON_GPCCS); else gf100_gr_init_fw(gr, 0x41a000, &gr->fuc41ac, &gr->fuc41ad); + if (ret) + return ret; - nvkm_mc_unk260(device->mc, 1); + nvkm_mc_unk260(device, 1); /* start both of them running */ nvkm_wr32(device, 0x409840, 0xffffffff); @@ -1557,7 +1582,7 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr) } /* load HUB microcode */ - nvkm_mc_unk260(device->mc, 0); + nvkm_mc_unk260(device, 0); nvkm_wr32(device, 0x4091c0, 0x01000000); for (i = 0; i < gr->func->fecs.ucode->data.size / 4; i++) nvkm_wr32(device, 0x4091c4, gr->func->fecs.ucode->data.data[i]); @@ -1580,7 +1605,7 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr) nvkm_wr32(device, 0x41a188, i >> 6); nvkm_wr32(device, 0x41a184, gr->func->gpccs.ucode->code.data[i]); } - nvkm_mc_unk260(device->mc, 1); + nvkm_mc_unk260(device, 1); /* load register lists */ gf100_gr_init_csdata(gr, grctx->hub, 0x409000, 0x000, 0x000000); diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h index 2b98abdb9270..268b8d60ff73 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h @@ -292,4 +292,6 @@ extern const struct gf100_gr_init gm107_gr_init_l1c_0[]; extern const struct gf100_gr_init gm107_gr_init_wwdx_0[]; extern const struct gf100_gr_init gm107_gr_init_cbm_0[]; void gm107_gr_init_bios(struct gf100_gr *); + +void gm200_gr_init_gpc_mmu(struct gf100_gr *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c index 4ca8ed15191c..de8b806b88fd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c @@ -361,6 +361,5 @@ gk20a_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr) if (ret) return ret; - return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c index 4dfa4513bb6c..6435f1257572 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c @@ -38,7 +38,7 @@ gm200_gr_rops(struct gf100_gr *gr) return nvkm_rd32(gr->base.engine.subdev.device, 0x12006c); } -static void +void gm200_gr_init_gpc_mmu(struct gf100_gr *gr) { struct nvkm_device *device = gr->base.engine.subdev.device; diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c new file mode 100644 index 000000000000..26ad79def0ff --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c @@ -0,0 +1,171 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "gf100.h" +#include "ctxgf100.h" + +#include <nvif/class.h> + +/******************************************************************************* + * PGRAPH engine/subdev functions + ******************************************************************************/ + +static void +gp100_gr_init_rop_active_fbps(struct gf100_gr *gr) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + /*XXX: otherwise identical to gm200 aside from mask.. do everywhere? */ + const u32 fbp_count = nvkm_rd32(device, 0x12006c) & 0x0000000f; + nvkm_mask(device, 0x408850, 0x0000000f, fbp_count); /* zrop */ + nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */ +} + +static int +gp100_gr_init(struct gf100_gr *gr) +{ + struct nvkm_device *device = gr->base.engine.subdev.device; + const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total); + u32 data[TPC_MAX / 8] = {}; + u8 tpcnr[GPC_MAX]; + int gpc, tpc, rop; + int i; + + gr->func->init_gpc_mmu(gr); + + gf100_gr_mmio(gr, gr->fuc_sw_nonctx); + + nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001); + + memset(data, 0x00, sizeof(data)); + memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr)); + for (i = 0, gpc = -1; i < gr->tpc_total; i++) { + do { + gpc = (gpc + 1) % gr->gpc_nr; + } while (!tpcnr[gpc]); + tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--; + + data[i / 8] |= tpc << ((i % 8) * 4); + } + + nvkm_wr32(device, GPC_BCAST(0x0980), data[0]); + nvkm_wr32(device, GPC_BCAST(0x0984), data[1]); + nvkm_wr32(device, GPC_BCAST(0x0988), data[2]); + nvkm_wr32(device, GPC_BCAST(0x098c), data[3]); + + for (gpc = 0; gpc < gr->gpc_nr; gpc++) { + nvkm_wr32(device, GPC_UNIT(gpc, 0x0914), + gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]); + nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 | + gr->tpc_total); + nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918); + } + + nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918); + nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800)); + nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804)); + + gr->func->init_rop_active_fbps(gr); + + nvkm_wr32(device, 0x400500, 0x00010001); + nvkm_wr32(device, 0x400100, 0xffffffff); + nvkm_wr32(device, 0x40013c, 0xffffffff); + nvkm_wr32(device, 0x400124, 0x00000002); + nvkm_wr32(device, 0x409c24, 0x000f0002); + nvkm_wr32(device, 0x405848, 0xc0000000); + nvkm_mask(device, 0x40584c, 0x00000000, 0x00000001); + nvkm_wr32(device, 0x404000, 0xc0000000); + nvkm_wr32(device, 0x404600, 0xc0000000); + nvkm_wr32(device, 0x408030, 0xc0000000); + nvkm_wr32(device, 0x404490, 0xc0000000); + nvkm_wr32(device, 0x406018, 0xc0000000); + nvkm_wr32(device, 0x407020, 0x40000000); + nvkm_wr32(device, 0x405840, 0xc0000000); + nvkm_wr32(device, 0x405844, 0x00ffffff); + nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008); + + nvkm_mask(device, 0x419c9c, 0x00010000, 0x00010000); + nvkm_mask(device, 0x419c9c, 0x00020000, 0x00020000); + + gr->func->init_ppc_exceptions(gr); + + for (gpc = 0; gpc < gr->gpc_nr; gpc++) { + nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000); + nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000); + nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000); + nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000); + for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) { + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe); + nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105); + } + nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff); + nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff); + } + + for (rop = 0; rop < gr->rop_nr; rop++) { + nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000); + nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000); + nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff); + nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff); + } + + nvkm_wr32(device, 0x400108, 0xffffffff); + nvkm_wr32(device, 0x400138, 0xffffffff); + nvkm_wr32(device, 0x400118, 0xffffffff); + nvkm_wr32(device, 0x400130, 0xffffffff); + nvkm_wr32(device, 0x40011c, 0xffffffff); + nvkm_wr32(device, 0x400134, 0xffffffff); + + gf100_gr_zbc_init(gr); + + return gf100_gr_init_ctxctl(gr); +} + +static const struct gf100_gr_func +gp100_gr = { + .init = gp100_gr_init, + .init_gpc_mmu = gm200_gr_init_gpc_mmu, + .init_rop_active_fbps = gp100_gr_init_rop_active_fbps, + .init_ppc_exceptions = gk104_gr_init_ppc_exceptions, + .rops = gm200_gr_rops, + .ppc_nr = 2, + .grctx = &gp100_grctx, + .sclass = { + { -1, -1, FERMI_TWOD_A }, + { -1, -1, KEPLER_INLINE_TO_MEMORY_B }, + { -1, -1, PASCAL_A, &gf100_fermi }, + { -1, -1, PASCAL_COMPUTE_A }, + {} + } +}; + +int +gp100_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr) +{ + return gm200_gr_new_(&gp100_gr, device, index, pgr); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c index e15b9627b07e..f3c30b2a788e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/base.c @@ -26,6 +26,49 @@ #include <subdev/bios.h> #include <subdev/bios/bmp.h> #include <subdev/bios/bit.h> +#include <subdev/bios/image.h> + +static bool +nvbios_addr(struct nvkm_bios *bios, u32 *addr, u8 size) +{ + u32 p = *addr; + + if (*addr > bios->image0_size && bios->imaged_addr) { + *addr -= bios->image0_size; + *addr += bios->imaged_addr; + } + + if (unlikely(*addr + size >= bios->size)) { + nvkm_error(&bios->subdev, "OOB %d %08x %08x\n", size, p, *addr); + return false; + } + + return true; +} + +u8 +nvbios_rd08(struct nvkm_bios *bios, u32 addr) +{ + if (likely(nvbios_addr(bios, &addr, 1))) + return bios->data[addr]; + return 0x00; +} + +u16 +nvbios_rd16(struct nvkm_bios *bios, u32 addr) +{ + if (likely(nvbios_addr(bios, &addr, 2))) + return get_unaligned_le16(&bios->data[addr]); + return 0x0000; +} + +u32 +nvbios_rd32(struct nvkm_bios *bios, u32 addr) +{ + if (likely(nvbios_addr(bios, &addr, 4))) + return get_unaligned_le32(&bios->data[addr]); + return 0x00000000; +} u8 nvbios_checksum(const u8 *data, int size) @@ -100,8 +143,9 @@ int nvkm_bios_new(struct nvkm_device *device, int index, struct nvkm_bios **pbios) { struct nvkm_bios *bios; + struct nvbios_image image; struct bit_entry bit_i; - int ret; + int ret, idx = 0; if (!(bios = *pbios = kzalloc(sizeof(*bios), GFP_KERNEL))) return -ENOMEM; @@ -111,6 +155,19 @@ nvkm_bios_new(struct nvkm_device *device, int index, struct nvkm_bios **pbios) if (ret) return ret; + /* Some tables have weird pointers that need adjustment before + * they're dereferenced. I'm not entirely sure why... + */ + if (nvbios_image(bios, idx++, &image)) { + bios->image0_size = image.size; + while (nvbios_image(bios, idx++, &image)) { + if (image.type == 0xe0) { + bios->imaged_addr = image.base; + break; + } + } + } + /* detect type of vbios we're dealing with */ bios->bmp_offset = nvbios_findstr(bios->data, bios->size, "\xff\x7f""NV\0", 5); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/disp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/disp.c index a5e92135cd77..9efb1b48cd54 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/disp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/disp.c @@ -141,7 +141,8 @@ nvbios_ocfg_parse(struct nvkm_bios *bios, u16 outp, u8 idx, { u16 data = nvbios_ocfg_entry(bios, outp, idx, ver, hdr, cnt, len); if (data) { - info->match = nvbios_rd16(bios, data + 0x00); + info->proto = nvbios_rd08(bios, data + 0x00); + info->flags = nvbios_rd16(bios, data + 0x01); info->clkcmp[0] = nvbios_rd16(bios, data + 0x02); info->clkcmp[1] = nvbios_rd16(bios, data + 0x04); } @@ -149,12 +150,13 @@ nvbios_ocfg_parse(struct nvkm_bios *bios, u16 outp, u8 idx, } u16 -nvbios_ocfg_match(struct nvkm_bios *bios, u16 outp, u16 type, +nvbios_ocfg_match(struct nvkm_bios *bios, u16 outp, u8 proto, u8 flags, u8 *ver, u8 *hdr, u8 *cnt, u8 *len, struct nvbios_ocfg *info) { u16 data, idx = 0; while ((data = nvbios_ocfg_parse(bios, outp, idx++, ver, hdr, cnt, len, info))) { - if (info->match == type) + if ((info->proto == proto || info->proto == 0xff) && + (info->flags == flags)) break; } return data; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c index 05332476354a..d89e78c4e689 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c @@ -40,6 +40,7 @@ nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) case 0x30: case 0x40: case 0x41: + case 0x42: *hdr = nvbios_rd08(bios, data + 0x01); *len = nvbios_rd08(bios, data + 0x02); *cnt = nvbios_rd08(bios, data + 0x03); @@ -70,6 +71,7 @@ nvbios_dpout_entry(struct nvkm_bios *bios, u8 idx, break; case 0x40: case 0x41: + case 0x42: *hdr = nvbios_rd08(bios, data + 0x04); *cnt = 0; *len = 0; @@ -109,6 +111,7 @@ nvbios_dpout_parse(struct nvkm_bios *bios, u8 idx, break; case 0x40: case 0x41: + case 0x42: info->flags = nvbios_rd08(bios, data + 0x04); info->script[0] = nvbios_rd16(bios, data + 0x05); info->script[1] = nvbios_rd16(bios, data + 0x07); @@ -180,6 +183,11 @@ nvbios_dpcfg_parse(struct nvkm_bios *bios, u16 outp, u8 idx, info->pe = nvbios_rd08(bios, data + 0x02); info->tx_pu = nvbios_rd08(bios, data + 0x03); break; + case 0x42: + info->dc = nvbios_rd08(bios, data + 0x00); + info->pe = nvbios_rd08(bios, data + 0x01); + info->tx_pu = nvbios_rd08(bios, data + 0x02); + break; default: data = 0x0000; break; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/image.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/image.c index 74b14cf09308..1dbff7aeafec 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/image.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/image.c @@ -68,11 +68,16 @@ nvbios_imagen(struct nvkm_bios *bios, struct nvbios_image *image) bool nvbios_image(struct nvkm_bios *bios, int idx, struct nvbios_image *image) { + u32 imaged_addr = bios->imaged_addr; memset(image, 0x00, sizeof(*image)); + bios->imaged_addr = 0; do { image->base += image->size; - if (image->last || !nvbios_imagen(bios, image)) + if (image->last || !nvbios_imagen(bios, image)) { + bios->imaged_addr = imaged_addr; return false; + } } while(idx--); + bios->imaged_addr = imaged_addr; return true; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c index 91a7dc56e406..2ca23a9157ab 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c @@ -77,15 +77,17 @@ g84_pll_mapping[] = { {} }; -static u16 +static u32 pll_limits_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) { struct bit_entry bit_C; - u16 data = 0x0000; + u32 data = 0x0000; if (!bit_entry(bios, 'C', &bit_C)) { if (bit_C.version == 1 && bit_C.length >= 10) data = nvbios_rd16(bios, bit_C.offset + 8); + if (bit_C.version == 2 && bit_C.length >= 4) + data = nvbios_rd32(bios, bit_C.offset + 0); if (data) { *ver = nvbios_rd08(bios, data + 0); *hdr = nvbios_rd08(bios, data + 1); @@ -137,12 +139,12 @@ pll_map(struct nvkm_bios *bios) } } -static u16 +static u32 pll_map_reg(struct nvkm_bios *bios, u32 reg, u32 *type, u8 *ver, u8 *len) { struct pll_mapping *map; u8 hdr, cnt; - u16 data; + u32 data; data = pll_limits_table(bios, ver, &hdr, &cnt, len); if (data && *ver >= 0x30) { @@ -160,7 +162,7 @@ pll_map_reg(struct nvkm_bios *bios, u32 reg, u32 *type, u8 *ver, u8 *len) map = pll_map(bios); while (map && map->reg) { if (map->reg == reg && *ver >= 0x20) { - u16 addr = (data += hdr); + u32 addr = (data += hdr); *type = map->type; while (cnt--) { if (nvbios_rd32(bios, data) == map->reg) @@ -179,12 +181,12 @@ pll_map_reg(struct nvkm_bios *bios, u32 reg, u32 *type, u8 *ver, u8 *len) return 0x0000; } -static u16 +static u32 pll_map_type(struct nvkm_bios *bios, u8 type, u32 *reg, u8 *ver, u8 *len) { struct pll_mapping *map; u8 hdr, cnt; - u16 data; + u32 data; data = pll_limits_table(bios, ver, &hdr, &cnt, len); if (data && *ver >= 0x30) { @@ -202,7 +204,7 @@ pll_map_type(struct nvkm_bios *bios, u8 type, u32 *reg, u8 *ver, u8 *len) map = pll_map(bios); while (map && map->reg) { if (map->type == type && *ver >= 0x20) { - u16 addr = (data += hdr); + u32 addr = (data += hdr); *reg = map->reg; while (cnt--) { if (nvbios_rd32(bios, data) == map->reg) @@ -228,7 +230,7 @@ nvbios_pll_parse(struct nvkm_bios *bios, u32 type, struct nvbios_pll *info) struct nvkm_device *device = subdev->device; u8 ver, len; u32 reg = type; - u16 data; + u32 data; if (type > PLL_MAX) { reg = type; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pmu.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pmu.c index c268e5afe852..b4a308f3cf7b 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pmu.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pmu.c @@ -26,21 +26,6 @@ #include <subdev/bios/image.h> #include <subdev/bios/pmu.h> -static u32 -weirdo_pointer(struct nvkm_bios *bios, u32 data) -{ - struct nvbios_image image; - int idx = 0; - if (nvbios_image(bios, idx++, &image)) { - data -= image.size; - while (nvbios_image(bios, idx++, &image)) { - if (image.type == 0xe0) - return image.base + data; - } - } - return 0; -} - u32 nvbios_pmuTe(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) { @@ -50,7 +35,7 @@ nvbios_pmuTe(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) if (!bit_entry(bios, 'p', &bit_p)) { if (bit_p.version == 2 && bit_p.length >= 4) data = nvbios_rd32(bios, bit_p.offset + 0x00); - if ((data = weirdo_pointer(bios, data))) { + if (data) { *ver = nvbios_rd08(bios, data + 0x00); /* maybe? */ *hdr = nvbios_rd08(bios, data + 0x01); *len = nvbios_rd08(bios, data + 0x02); @@ -97,8 +82,7 @@ nvbios_pmuRm(struct nvkm_bios *bios, u8 type, struct nvbios_pmuR *info) u32 data; memset(info, 0x00, sizeof(*info)); while ((data = nvbios_pmuEp(bios, idx++, &ver, &hdr, &pmuE))) { - if ( pmuE.type == type && - (data = weirdo_pointer(bios, pmuE.data))) { + if (pmuE.type == type && (data = pmuE.data)) { info->init_addr_pmu = nvbios_rd32(bios, data + 0x08); info->args_addr_pmu = nvbios_rd32(bios, data + 0x0c); info->boot_addr = data + 0x30; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/rammap.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/rammap.c index d0ae7454764e..b57c370c725d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/rammap.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/rammap.c @@ -30,11 +30,11 @@ nvbios_rammapTe(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len, u8 *snr, u8 *ssz) { struct bit_entry bit_P; - u16 rammap = 0x0000; + u32 rammap = 0x0000; if (!bit_entry(bios, 'P', &bit_P)) { if (bit_P.version == 2) - rammap = nvbios_rd16(bios, bit_P.offset + 4); + rammap = nvbios_rd32(bios, bit_P.offset + 4); if (rammap) { *ver = nvbios_rd08(bios, rammap + 0); @@ -61,7 +61,7 @@ nvbios_rammapEe(struct nvkm_bios *bios, int idx, u8 *ver, u8 *hdr, u8 *cnt, u8 *len) { u8 snr, ssz; - u16 rammap = nvbios_rammapTe(bios, ver, hdr, cnt, len, &snr, &ssz); + u32 rammap = nvbios_rammapTe(bios, ver, hdr, cnt, len, &snr, &ssz); if (rammap && idx < *cnt) { rammap = rammap + *hdr + (idx * (*len + (snr * ssz))); *hdr = *len; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c index 78c449b417b7..89d5543118cf 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gf100.c @@ -99,7 +99,7 @@ read_div(struct gf100_clk *clk, int doff, u32 dsrc, u32 dctl) { struct nvkm_device *device = clk->base.subdev.device; u32 ssrc = nvkm_rd32(device, dsrc + (doff * 4)); - u32 sctl = nvkm_rd32(device, dctl + (doff * 4)); + u32 sclk, sctl, sdiv = 2; switch (ssrc & 0x00000003) { case 0: @@ -109,13 +109,21 @@ read_div(struct gf100_clk *clk, int doff, u32 dsrc, u32 dctl) case 2: return 100000; case 3: - if (sctl & 0x80000000) { - u32 sclk = read_vco(clk, dsrc + (doff * 4)); - u32 sdiv = (sctl & 0x0000003f) + 2; - return (sclk * 2) / sdiv; + sclk = read_vco(clk, dsrc + (doff * 4)); + + /* Memclk has doff of 0 despite its alt. location */ + if (doff <= 2) { + sctl = nvkm_rd32(device, dctl + (doff * 4)); + + if (sctl & 0x80000000) { + if (ssrc & 0x100) + sctl >>= 8; + + sdiv = (sctl & 0x3f) + 2; + } } - return read_vco(clk, dsrc + (doff * 4)); + return (sclk * 2) / sdiv; default: return 0; } @@ -366,11 +374,17 @@ gf100_clk_prog_2(struct gf100_clk *clk, int idx) if (info->coef) { nvkm_wr32(device, addr + 0x04, info->coef); nvkm_mask(device, addr + 0x00, 0x00000001, 0x00000001); + + /* Test PLL lock */ + nvkm_mask(device, addr + 0x00, 0x00000010, 0x00000000); nvkm_msec(device, 2000, if (nvkm_rd32(device, addr + 0x00) & 0x00020000) break; ); - nvkm_mask(device, addr + 0x00, 0x00020004, 0x00000004); + nvkm_mask(device, addr + 0x00, 0x00000010, 0x00000010); + + /* Enable sync mode */ + nvkm_mask(device, addr + 0x00, 0x00000004, 0x00000004); } } } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c index 975c401bccab..06bc0d2d6ae1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk104.c @@ -393,11 +393,17 @@ gk104_clk_prog_2(struct gk104_clk *clk, int idx) if (info->coef) { nvkm_wr32(device, addr + 0x04, info->coef); nvkm_mask(device, addr + 0x00, 0x00000001, 0x00000001); + + /* Test PLL lock */ + nvkm_mask(device, addr + 0x00, 0x00000010, 0x00000000); nvkm_msec(device, 2000, if (nvkm_rd32(device, addr + 0x00) & 0x00020000) break; ); - nvkm_mask(device, addr + 0x00, 0x00020004, 0x00000004); + nvkm_mask(device, addr + 0x00, 0x00000010, 0x00000010); + + /* Enable sync mode */ + nvkm_mask(device, addr + 0x00, 0x00000004, 0x00000004); } } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.c index 5f0ee24e31b8..218893e3e5f9 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.c @@ -28,69 +28,6 @@ #include <core/tegra.h> #include <subdev/timer.h> -#define KHZ (1000) -#define MHZ (KHZ * 1000) - -#define MASK(w) ((1 << w) - 1) - -#define GPCPLL_CFG (SYS_GPCPLL_CFG_BASE + 0) -#define GPCPLL_CFG_ENABLE BIT(0) -#define GPCPLL_CFG_IDDQ BIT(1) -#define GPCPLL_CFG_LOCK_DET_OFF BIT(4) -#define GPCPLL_CFG_LOCK BIT(17) - -#define GPCPLL_COEFF (SYS_GPCPLL_CFG_BASE + 4) -#define GPCPLL_COEFF_M_SHIFT 0 -#define GPCPLL_COEFF_M_WIDTH 8 -#define GPCPLL_COEFF_N_SHIFT 8 -#define GPCPLL_COEFF_N_WIDTH 8 -#define GPCPLL_COEFF_P_SHIFT 16 -#define GPCPLL_COEFF_P_WIDTH 6 - -#define GPCPLL_CFG2 (SYS_GPCPLL_CFG_BASE + 0xc) -#define GPCPLL_CFG2_SETUP2_SHIFT 16 -#define GPCPLL_CFG2_PLL_STEPA_SHIFT 24 - -#define GPCPLL_CFG3 (SYS_GPCPLL_CFG_BASE + 0x18) -#define GPCPLL_CFG3_PLL_STEPB_SHIFT 16 - -#define GPC_BCASE_GPCPLL_CFG_BASE 0x00132800 -#define GPCPLL_NDIV_SLOWDOWN (SYS_GPCPLL_CFG_BASE + 0x1c) -#define GPCPLL_NDIV_SLOWDOWN_NDIV_LO_SHIFT 0 -#define GPCPLL_NDIV_SLOWDOWN_NDIV_MID_SHIFT 8 -#define GPCPLL_NDIV_SLOWDOWN_STEP_SIZE_LO2MID_SHIFT 16 -#define GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT 22 -#define GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT 31 - -#define SEL_VCO (SYS_GPCPLL_CFG_BASE + 0x100) -#define SEL_VCO_GPC2CLK_OUT_SHIFT 0 - -#define GPC2CLK_OUT (SYS_GPCPLL_CFG_BASE + 0x250) -#define GPC2CLK_OUT_SDIV14_INDIV4_WIDTH 1 -#define GPC2CLK_OUT_SDIV14_INDIV4_SHIFT 31 -#define GPC2CLK_OUT_SDIV14_INDIV4_MODE 1 -#define GPC2CLK_OUT_VCODIV_WIDTH 6 -#define GPC2CLK_OUT_VCODIV_SHIFT 8 -#define GPC2CLK_OUT_VCODIV1 0 -#define GPC2CLK_OUT_VCODIV_MASK (MASK(GPC2CLK_OUT_VCODIV_WIDTH) << \ - GPC2CLK_OUT_VCODIV_SHIFT) -#define GPC2CLK_OUT_BYPDIV_WIDTH 6 -#define GPC2CLK_OUT_BYPDIV_SHIFT 0 -#define GPC2CLK_OUT_BYPDIV31 0x3c -#define GPC2CLK_OUT_INIT_MASK ((MASK(GPC2CLK_OUT_SDIV14_INDIV4_WIDTH) << \ - GPC2CLK_OUT_SDIV14_INDIV4_SHIFT)\ - | (MASK(GPC2CLK_OUT_VCODIV_WIDTH) << GPC2CLK_OUT_VCODIV_SHIFT)\ - | (MASK(GPC2CLK_OUT_BYPDIV_WIDTH) << GPC2CLK_OUT_BYPDIV_SHIFT)) -#define GPC2CLK_OUT_INIT_VAL ((GPC2CLK_OUT_SDIV14_INDIV4_MODE << \ - GPC2CLK_OUT_SDIV14_INDIV4_SHIFT) \ - | (GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT) \ - | (GPC2CLK_OUT_BYPDIV31 << GPC2CLK_OUT_BYPDIV_SHIFT)) - -#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG (GPC_BCASE_GPCPLL_CFG_BASE + 0xa0) -#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_SHIFT 24 -#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK \ - (0x1 << GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_SHIFT) - static const u8 _pl_to_div[] = { /* PL: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 */ /* p: */ 1, 2, 3, 4, 5, 6, 8, 10, 12, 16, 12, 16, 20, 24, 32, @@ -124,7 +61,7 @@ static const struct gk20a_clk_pllg_params gk20a_pllg_params = { .min_pl = 1, .max_pl = 32, }; -static void +void gk20a_pllg_read_mnp(struct gk20a_clk *clk, struct gk20a_pll *pll) { struct nvkm_device *device = clk->base.subdev.device; @@ -136,20 +73,33 @@ gk20a_pllg_read_mnp(struct gk20a_clk *clk, struct gk20a_pll *pll) pll->pl = (val >> GPCPLL_COEFF_P_SHIFT) & MASK(GPCPLL_COEFF_P_WIDTH); } -static u32 -gk20a_pllg_calc_rate(struct gk20a_clk *clk) +void +gk20a_pllg_write_mnp(struct gk20a_clk *clk, const struct gk20a_pll *pll) +{ + struct nvkm_device *device = clk->base.subdev.device; + u32 val; + + val = (pll->m & MASK(GPCPLL_COEFF_M_WIDTH)) << GPCPLL_COEFF_M_SHIFT; + val |= (pll->n & MASK(GPCPLL_COEFF_N_WIDTH)) << GPCPLL_COEFF_N_SHIFT; + val |= (pll->pl & MASK(GPCPLL_COEFF_P_WIDTH)) << GPCPLL_COEFF_P_SHIFT; + nvkm_wr32(device, GPCPLL_COEFF, val); +} + +u32 +gk20a_pllg_calc_rate(struct gk20a_clk *clk, struct gk20a_pll *pll) { u32 rate; u32 divider; - rate = clk->parent_rate * clk->pll.n; - divider = clk->pll.m * clk->pl_to_div(clk->pll.pl); + rate = clk->parent_rate * pll->n; + divider = pll->m * clk->pl_to_div(pll->pl); return rate / divider / 2; } -static int -gk20a_pllg_calc_mnp(struct gk20a_clk *clk, unsigned long rate) +int +gk20a_pllg_calc_mnp(struct gk20a_clk *clk, unsigned long rate, + struct gk20a_pll *pll) { struct nvkm_subdev *subdev = &clk->base.subdev; u32 target_clk_f, ref_clk_f, target_freq; @@ -163,16 +113,13 @@ gk20a_pllg_calc_mnp(struct gk20a_clk *clk, unsigned long rate) target_clk_f = rate * 2 / KHZ; ref_clk_f = clk->parent_rate / KHZ; - max_vco_f = clk->params->max_vco; + target_vco_f = target_clk_f + target_clk_f / 50; + max_vco_f = max(clk->params->max_vco, target_vco_f); min_vco_f = clk->params->min_vco; best_m = clk->params->max_m; best_n = clk->params->min_n; best_pl = clk->params->min_pl; - target_vco_f = target_clk_f + target_clk_f / 50; - if (max_vco_f < target_vco_f) - max_vco_f = target_vco_f; - /* min_pl <= high_pl <= max_pl */ high_pl = (max_vco_f + target_vco_f - 1) / target_vco_f; high_pl = min(high_pl, clk->params->max_pl); @@ -195,9 +142,7 @@ gk20a_pllg_calc_mnp(struct gk20a_clk *clk, unsigned long rate) target_vco_f = target_clk_f * clk->pl_to_div(pl); for (m = clk->params->min_m; m <= clk->params->max_m; m++) { - u32 u_f, vco_f; - - u_f = ref_clk_f / m; + u32 u_f = ref_clk_f / m; if (u_f < clk->params->min_u) break; @@ -211,6 +156,8 @@ gk20a_pllg_calc_mnp(struct gk20a_clk *clk, unsigned long rate) break; for (; n <= n2; n++) { + u32 vco_f; + if (n < clk->params->min_n) continue; if (n > clk->params->max_n) @@ -247,16 +194,16 @@ found_match: "no best match for target @ %dMHz on gpc_pll", target_clk_f / KHZ); - clk->pll.m = best_m; - clk->pll.n = best_n; - clk->pll.pl = best_pl; + pll->m = best_m; + pll->n = best_n; + pll->pl = best_pl; - target_freq = gk20a_pllg_calc_rate(clk); + target_freq = gk20a_pllg_calc_rate(clk, pll); nvkm_debug(subdev, - "actual target freq %d MHz, M %d, N %d, PL %d(div%d)\n", - target_freq / MHZ, clk->pll.m, clk->pll.n, clk->pll.pl, - clk->pl_to_div(clk->pll.pl)); + "actual target freq %d KHz, M %d, N %d, PL %d(div%d)\n", + target_freq / KHZ, pll->m, pll->n, pll->pl, + clk->pl_to_div(pll->pl)); return 0; } @@ -265,45 +212,36 @@ gk20a_pllg_slide(struct gk20a_clk *clk, u32 n) { struct nvkm_subdev *subdev = &clk->base.subdev; struct nvkm_device *device = subdev->device; - u32 val; - int ramp_timeout; + struct gk20a_pll pll; + int ret = 0; /* get old coefficients */ - val = nvkm_rd32(device, GPCPLL_COEFF); + gk20a_pllg_read_mnp(clk, &pll); /* do nothing if NDIV is the same */ - if (n == ((val >> GPCPLL_COEFF_N_SHIFT) & MASK(GPCPLL_COEFF_N_WIDTH))) + if (n == pll.n) return 0; - /* setup */ - nvkm_mask(device, GPCPLL_CFG2, 0xff << GPCPLL_CFG2_PLL_STEPA_SHIFT, - 0x2b << GPCPLL_CFG2_PLL_STEPA_SHIFT); - nvkm_mask(device, GPCPLL_CFG3, 0xff << GPCPLL_CFG3_PLL_STEPB_SHIFT, - 0xb << GPCPLL_CFG3_PLL_STEPB_SHIFT); - /* pll slowdown mode */ nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, BIT(GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT), BIT(GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT)); /* new ndiv ready for ramp */ - val = nvkm_rd32(device, GPCPLL_COEFF); - val &= ~(MASK(GPCPLL_COEFF_N_WIDTH) << GPCPLL_COEFF_N_SHIFT); - val |= (n & MASK(GPCPLL_COEFF_N_WIDTH)) << GPCPLL_COEFF_N_SHIFT; + pll.n = n; udelay(1); - nvkm_wr32(device, GPCPLL_COEFF, val); + gk20a_pllg_write_mnp(clk, &pll); /* dynamic ramp to new ndiv */ - val = nvkm_rd32(device, GPCPLL_NDIV_SLOWDOWN); - val |= 0x1 << GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT; udelay(1); - nvkm_wr32(device, GPCPLL_NDIV_SLOWDOWN, val); + nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, + BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT), + BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT)); - for (ramp_timeout = 500; ramp_timeout > 0; ramp_timeout--) { - udelay(1); - val = nvkm_rd32(device, GPC_BCAST_NDIV_SLOWDOWN_DEBUG); - if (val & GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK) - break; - } + /* wait for ramping to complete */ + if (nvkm_wait_usec(device, 500, GPC_BCAST_NDIV_SLOWDOWN_DEBUG, + GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK, + GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK) < 0) + ret = -ETIMEDOUT; /* exit slowdown mode */ nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, @@ -311,21 +249,35 @@ gk20a_pllg_slide(struct gk20a_clk *clk, u32 n) BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT), 0); nvkm_rd32(device, GPCPLL_NDIV_SLOWDOWN); - if (ramp_timeout <= 0) { - nvkm_error(subdev, "gpcpll dynamic ramp timeout\n"); - return -ETIMEDOUT; - } - - return 0; + return ret; } -static void +static int gk20a_pllg_enable(struct gk20a_clk *clk) { struct nvkm_device *device = clk->base.subdev.device; + u32 val; nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_ENABLE, GPCPLL_CFG_ENABLE); nvkm_rd32(device, GPCPLL_CFG); + + /* enable lock detection */ + val = nvkm_rd32(device, GPCPLL_CFG); + if (val & GPCPLL_CFG_LOCK_DET_OFF) { + val &= ~GPCPLL_CFG_LOCK_DET_OFF; + nvkm_wr32(device, GPCPLL_CFG, val); + } + + /* wait for lock */ + if (nvkm_wait_usec(device, 300, GPCPLL_CFG, GPCPLL_CFG_LOCK, + GPCPLL_CFG_LOCK) < 0) + return -ETIMEDOUT; + + /* switch to VCO mode */ + nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), + BIT(SEL_VCO_GPC2CLK_OUT_SHIFT)); + + return 0; } static void @@ -333,117 +285,81 @@ gk20a_pllg_disable(struct gk20a_clk *clk) { struct nvkm_device *device = clk->base.subdev.device; + /* put PLL in bypass before disabling it */ + nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), 0); + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_ENABLE, 0); nvkm_rd32(device, GPCPLL_CFG); } static int -_gk20a_pllg_program_mnp(struct gk20a_clk *clk, bool allow_slide) +gk20a_pllg_program_mnp(struct gk20a_clk *clk, const struct gk20a_pll *pll) { struct nvkm_subdev *subdev = &clk->base.subdev; struct nvkm_device *device = subdev->device; - u32 val, cfg; - struct gk20a_pll old_pll; - u32 n_lo; - - /* get old coefficients */ - gk20a_pllg_read_mnp(clk, &old_pll); - - /* do NDIV slide if there is no change in M and PL */ - cfg = nvkm_rd32(device, GPCPLL_CFG); - if (allow_slide && clk->pll.m == old_pll.m && - clk->pll.pl == old_pll.pl && (cfg & GPCPLL_CFG_ENABLE)) { - return gk20a_pllg_slide(clk, clk->pll.n); - } - - /* slide down to NDIV_LO */ - if (allow_slide && (cfg & GPCPLL_CFG_ENABLE)) { - int ret; - - n_lo = DIV_ROUND_UP(old_pll.m * clk->params->min_vco, - clk->parent_rate / KHZ); - ret = gk20a_pllg_slide(clk, n_lo); + struct gk20a_pll cur_pll; + int ret; - if (ret) - return ret; - } + gk20a_pllg_read_mnp(clk, &cur_pll); - /* split FO-to-bypass jump in halfs by setting out divider 1:2 */ + /* split VCO-to-bypass jump in half by setting out divider 1:2 */ nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, - 0x2 << GPC2CLK_OUT_VCODIV_SHIFT); - - /* put PLL in bypass before programming it */ - val = nvkm_rd32(device, SEL_VCO); - val &= ~(BIT(SEL_VCO_GPC2CLK_OUT_SHIFT)); + GPC2CLK_OUT_VCODIV2 << GPC2CLK_OUT_VCODIV_SHIFT); + /* Intentional 2nd write to assure linear divider operation */ + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV2 << GPC2CLK_OUT_VCODIV_SHIFT); + nvkm_rd32(device, GPC2CLK_OUT); udelay(2); - nvkm_wr32(device, SEL_VCO, val); - - /* get out from IDDQ */ - val = nvkm_rd32(device, GPCPLL_CFG); - if (val & GPCPLL_CFG_IDDQ) { - val &= ~GPCPLL_CFG_IDDQ; - nvkm_wr32(device, GPCPLL_CFG, val); - nvkm_rd32(device, GPCPLL_CFG); - udelay(2); - } gk20a_pllg_disable(clk); - nvkm_debug(subdev, "%s: m=%d n=%d pl=%d\n", __func__, - clk->pll.m, clk->pll.n, clk->pll.pl); - - n_lo = DIV_ROUND_UP(clk->pll.m * clk->params->min_vco, - clk->parent_rate / KHZ); - val = clk->pll.m << GPCPLL_COEFF_M_SHIFT; - val |= (allow_slide ? n_lo : clk->pll.n) << GPCPLL_COEFF_N_SHIFT; - val |= clk->pll.pl << GPCPLL_COEFF_P_SHIFT; - nvkm_wr32(device, GPCPLL_COEFF, val); + gk20a_pllg_write_mnp(clk, pll); - gk20a_pllg_enable(clk); - - val = nvkm_rd32(device, GPCPLL_CFG); - if (val & GPCPLL_CFG_LOCK_DET_OFF) { - val &= ~GPCPLL_CFG_LOCK_DET_OFF; - nvkm_wr32(device, GPCPLL_CFG, val); - } - - if (nvkm_usec(device, 300, - if (nvkm_rd32(device, GPCPLL_CFG) & GPCPLL_CFG_LOCK) - break; - ) < 0) - return -ETIMEDOUT; - - /* switch to VCO mode */ - nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), - BIT(SEL_VCO_GPC2CLK_OUT_SHIFT)); + ret = gk20a_pllg_enable(clk); + if (ret) + return ret; /* restore out divider 1:1 */ - val = nvkm_rd32(device, GPC2CLK_OUT); - if ((val & GPC2CLK_OUT_VCODIV_MASK) != - (GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT)) { - val &= ~GPC2CLK_OUT_VCODIV_MASK; - val |= GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT; - udelay(2); - nvkm_wr32(device, GPC2CLK_OUT, val); - /* Intentional 2nd write to assure linear divider operation */ - nvkm_wr32(device, GPC2CLK_OUT, val); - nvkm_rd32(device, GPC2CLK_OUT); - } + udelay(2); + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT); + /* Intentional 2nd write to assure linear divider operation */ + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT); + nvkm_rd32(device, GPC2CLK_OUT); - /* slide up to new NDIV */ - return allow_slide ? gk20a_pllg_slide(clk, clk->pll.n) : 0; + return 0; } static int -gk20a_pllg_program_mnp(struct gk20a_clk *clk) +gk20a_pllg_program_mnp_slide(struct gk20a_clk *clk, const struct gk20a_pll *pll) { - int err; + struct gk20a_pll cur_pll; + int ret; - err = _gk20a_pllg_program_mnp(clk, true); - if (err) - err = _gk20a_pllg_program_mnp(clk, false); + if (gk20a_pllg_is_enabled(clk)) { + gk20a_pllg_read_mnp(clk, &cur_pll); + + /* just do NDIV slide if there is no change to M and PL */ + if (pll->m == cur_pll.m && pll->pl == cur_pll.pl) + return gk20a_pllg_slide(clk, pll->n); + + /* slide down to current NDIV_LO */ + cur_pll.n = gk20a_pllg_n_lo(clk, &cur_pll); + ret = gk20a_pllg_slide(clk, cur_pll.n); + if (ret) + return ret; + } + + /* program MNP with the new clock parameters and new NDIV_LO */ + cur_pll = *pll; + cur_pll.n = gk20a_pllg_n_lo(clk, &cur_pll); + ret = gk20a_pllg_program_mnp(clk, &cur_pll); + if (ret) + return ret; - return err; + /* slide up to new NDIV */ + return gk20a_pllg_slide(clk, pll->n); } static struct nvkm_pstate @@ -546,13 +462,14 @@ gk20a_clk_read(struct nvkm_clk *base, enum nv_clk_src src) struct gk20a_clk *clk = gk20a_clk(base); struct nvkm_subdev *subdev = &clk->base.subdev; struct nvkm_device *device = subdev->device; + struct gk20a_pll pll; switch (src) { case nv_clk_src_crystal: return device->crystal; case nv_clk_src_gpc: - gk20a_pllg_read_mnp(clk, &clk->pll); - return gk20a_pllg_calc_rate(clk) / GK20A_CLK_GPC_MDIV; + gk20a_pllg_read_mnp(clk, &pll); + return gk20a_pllg_calc_rate(clk, &pll) / GK20A_CLK_GPC_MDIV; default: nvkm_error(subdev, "invalid clock source %d\n", src); return -EINVAL; @@ -565,15 +482,20 @@ gk20a_clk_calc(struct nvkm_clk *base, struct nvkm_cstate *cstate) struct gk20a_clk *clk = gk20a_clk(base); return gk20a_pllg_calc_mnp(clk, cstate->domain[nv_clk_src_gpc] * - GK20A_CLK_GPC_MDIV); + GK20A_CLK_GPC_MDIV, &clk->pll); } int gk20a_clk_prog(struct nvkm_clk *base) { struct gk20a_clk *clk = gk20a_clk(base); + int ret; + + ret = gk20a_pllg_program_mnp_slide(clk, &clk->pll); + if (ret) + ret = gk20a_pllg_program_mnp(clk, &clk->pll); - return gk20a_pllg_program_mnp(clk); + return ret; } void @@ -581,29 +503,62 @@ gk20a_clk_tidy(struct nvkm_clk *base) { } +int +gk20a_clk_setup_slide(struct gk20a_clk *clk) +{ + struct nvkm_subdev *subdev = &clk->base.subdev; + struct nvkm_device *device = subdev->device; + u32 step_a, step_b; + + switch (clk->parent_rate) { + case 12000000: + case 12800000: + case 13000000: + step_a = 0x2b; + step_b = 0x0b; + break; + case 19200000: + step_a = 0x12; + step_b = 0x08; + break; + case 38400000: + step_a = 0x04; + step_b = 0x05; + break; + default: + nvkm_error(subdev, "invalid parent clock rate %u KHz", + clk->parent_rate / KHZ); + return -EINVAL; + } + + nvkm_mask(device, GPCPLL_CFG2, 0xff << GPCPLL_CFG2_PLL_STEPA_SHIFT, + step_a << GPCPLL_CFG2_PLL_STEPA_SHIFT); + nvkm_mask(device, GPCPLL_CFG3, 0xff << GPCPLL_CFG3_PLL_STEPB_SHIFT, + step_b << GPCPLL_CFG3_PLL_STEPB_SHIFT); + + return 0; +} + void gk20a_clk_fini(struct nvkm_clk *base) { struct nvkm_device *device = base->subdev.device; struct gk20a_clk *clk = gk20a_clk(base); - u32 val; /* slide to VCO min */ - val = nvkm_rd32(device, GPCPLL_CFG); - if (val & GPCPLL_CFG_ENABLE) { + if (gk20a_pllg_is_enabled(clk)) { struct gk20a_pll pll; u32 n_lo; gk20a_pllg_read_mnp(clk, &pll); - n_lo = DIV_ROUND_UP(pll.m * clk->params->min_vco, - clk->parent_rate / KHZ); + n_lo = gk20a_pllg_n_lo(clk, &pll); gk20a_pllg_slide(clk, n_lo); } - /* put PLL in bypass before disabling it */ - nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), 0); - gk20a_pllg_disable(clk); + + /* set IDDQ */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_IDDQ, 1); } static int @@ -614,9 +569,18 @@ gk20a_clk_init(struct nvkm_clk *base) struct nvkm_device *device = subdev->device; int ret; + /* get out from IDDQ */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_IDDQ, 0); + nvkm_rd32(device, GPCPLL_CFG); + udelay(5); + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_INIT_MASK, GPC2CLK_OUT_INIT_VAL); + ret = gk20a_clk_setup_slide(clk); + if (ret) + return ret; + /* Start with lowest frequency */ base->func->calc(base, &base->func->pstates[0].base); ret = base->func->prog(&clk->base); @@ -646,7 +610,7 @@ gk20a_clk = { }; int -_gk20a_clk_ctor(struct nvkm_device *device, int index, +gk20a_clk_ctor(struct nvkm_device *device, int index, const struct nvkm_clk_func *func, const struct gk20a_clk_pllg_params *params, struct gk20a_clk *clk) @@ -685,7 +649,7 @@ gk20a_clk_new(struct nvkm_device *device, int index, struct nvkm_clk **pclk) return -ENOMEM; *pclk = &clk->base; - ret = _gk20a_clk_ctor(device, index, &gk20a_clk, &gk20a_pllg_params, + ret = gk20a_clk_ctor(device, index, &gk20a_clk, &gk20a_pllg_params, clk); clk->pl_to_div = pl_to_div; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.h b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.h index 13c46740197d..0d1450972162 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gk20a.h @@ -24,9 +24,79 @@ #ifndef __NVKM_CLK_GK20A_H__ #define __NVKM_CLK_GK20A_H__ +#define KHZ (1000) +#define MHZ (KHZ * 1000) + +#define MASK(w) ((1 << (w)) - 1) + #define GK20A_CLK_GPC_MDIV 1000 #define SYS_GPCPLL_CFG_BASE 0x00137000 +#define GPCPLL_CFG (SYS_GPCPLL_CFG_BASE + 0) +#define GPCPLL_CFG_ENABLE BIT(0) +#define GPCPLL_CFG_IDDQ BIT(1) +#define GPCPLL_CFG_LOCK_DET_OFF BIT(4) +#define GPCPLL_CFG_LOCK BIT(17) + +#define GPCPLL_CFG2 (SYS_GPCPLL_CFG_BASE + 0xc) +#define GPCPLL_CFG2_SETUP2_SHIFT 16 +#define GPCPLL_CFG2_PLL_STEPA_SHIFT 24 + +#define GPCPLL_CFG3 (SYS_GPCPLL_CFG_BASE + 0x18) +#define GPCPLL_CFG3_VCO_CTRL_SHIFT 0 +#define GPCPLL_CFG3_VCO_CTRL_WIDTH 9 +#define GPCPLL_CFG3_VCO_CTRL_MASK \ + (MASK(GPCPLL_CFG3_VCO_CTRL_WIDTH) << GPCPLL_CFG3_VCO_CTRL_SHIFT) +#define GPCPLL_CFG3_PLL_STEPB_SHIFT 16 +#define GPCPLL_CFG3_PLL_STEPB_WIDTH 8 + +#define GPCPLL_COEFF (SYS_GPCPLL_CFG_BASE + 4) +#define GPCPLL_COEFF_M_SHIFT 0 +#define GPCPLL_COEFF_M_WIDTH 8 +#define GPCPLL_COEFF_N_SHIFT 8 +#define GPCPLL_COEFF_N_WIDTH 8 +#define GPCPLL_COEFF_N_MASK \ + (MASK(GPCPLL_COEFF_N_WIDTH) << GPCPLL_COEFF_N_SHIFT) +#define GPCPLL_COEFF_P_SHIFT 16 +#define GPCPLL_COEFF_P_WIDTH 6 + +#define GPCPLL_NDIV_SLOWDOWN (SYS_GPCPLL_CFG_BASE + 0x1c) +#define GPCPLL_NDIV_SLOWDOWN_NDIV_LO_SHIFT 0 +#define GPCPLL_NDIV_SLOWDOWN_NDIV_MID_SHIFT 8 +#define GPCPLL_NDIV_SLOWDOWN_STEP_SIZE_LO2MID_SHIFT 16 +#define GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT 22 +#define GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT 31 + +#define GPC_BCAST_GPCPLL_CFG_BASE 0x00132800 +#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG (GPC_BCAST_GPCPLL_CFG_BASE + 0xa0) +#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_SHIFT 24 +#define GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK \ + (0x1 << GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_SHIFT) + +#define SEL_VCO (SYS_GPCPLL_CFG_BASE + 0x100) +#define SEL_VCO_GPC2CLK_OUT_SHIFT 0 + +#define GPC2CLK_OUT (SYS_GPCPLL_CFG_BASE + 0x250) +#define GPC2CLK_OUT_SDIV14_INDIV4_WIDTH 1 +#define GPC2CLK_OUT_SDIV14_INDIV4_SHIFT 31 +#define GPC2CLK_OUT_SDIV14_INDIV4_MODE 1 +#define GPC2CLK_OUT_VCODIV_WIDTH 6 +#define GPC2CLK_OUT_VCODIV_SHIFT 8 +#define GPC2CLK_OUT_VCODIV1 0 +#define GPC2CLK_OUT_VCODIV2 2 +#define GPC2CLK_OUT_VCODIV_MASK (MASK(GPC2CLK_OUT_VCODIV_WIDTH) << \ + GPC2CLK_OUT_VCODIV_SHIFT) +#define GPC2CLK_OUT_BYPDIV_WIDTH 6 +#define GPC2CLK_OUT_BYPDIV_SHIFT 0 +#define GPC2CLK_OUT_BYPDIV31 0x3c +#define GPC2CLK_OUT_INIT_MASK ((MASK(GPC2CLK_OUT_SDIV14_INDIV4_WIDTH) << \ + GPC2CLK_OUT_SDIV14_INDIV4_SHIFT)\ + | (MASK(GPC2CLK_OUT_VCODIV_WIDTH) << GPC2CLK_OUT_VCODIV_SHIFT)\ + | (MASK(GPC2CLK_OUT_BYPDIV_WIDTH) << GPC2CLK_OUT_BYPDIV_SHIFT)) +#define GPC2CLK_OUT_INIT_VAL ((GPC2CLK_OUT_SDIV14_INDIV4_MODE << \ + GPC2CLK_OUT_SDIV14_INDIV4_SHIFT) \ + | (GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT) \ + | (GPC2CLK_OUT_BYPDIV31 << GPC2CLK_OUT_BYPDIV_SHIFT)) /* All frequencies in Khz */ struct gk20a_clk_pllg_params { @@ -54,7 +124,29 @@ struct gk20a_clk { }; #define gk20a_clk(p) container_of((p), struct gk20a_clk, base) -int _gk20a_clk_ctor(struct nvkm_device *, int, const struct nvkm_clk_func *, +u32 gk20a_pllg_calc_rate(struct gk20a_clk *, struct gk20a_pll *); +int gk20a_pllg_calc_mnp(struct gk20a_clk *, unsigned long, struct gk20a_pll *); +void gk20a_pllg_read_mnp(struct gk20a_clk *, struct gk20a_pll *); +void gk20a_pllg_write_mnp(struct gk20a_clk *, const struct gk20a_pll *); + +static inline bool +gk20a_pllg_is_enabled(struct gk20a_clk *clk) +{ + struct nvkm_device *device = clk->base.subdev.device; + u32 val; + + val = nvkm_rd32(device, GPCPLL_CFG); + return val & GPCPLL_CFG_ENABLE; +} + +static inline u32 +gk20a_pllg_n_lo(struct gk20a_clk *clk, struct gk20a_pll *pll) +{ + return DIV_ROUND_UP(pll->m * clk->params->min_vco, + clk->parent_rate / KHZ); +} + +int gk20a_clk_ctor(struct nvkm_device *, int, const struct nvkm_clk_func *, const struct gk20a_clk_pllg_params *, struct gk20a_clk *); void gk20a_clk_fini(struct nvkm_clk *); int gk20a_clk_read(struct nvkm_clk *, enum nv_clk_src); @@ -62,4 +154,6 @@ int gk20a_clk_calc(struct nvkm_clk *, struct nvkm_cstate *); int gk20a_clk_prog(struct nvkm_clk *); void gk20a_clk_tidy(struct nvkm_clk *); +int gk20a_clk_setup_slide(struct gk20a_clk *); + #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gm20b.c index 71b2bbb61973..b284e949f732 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gm20b.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/gm20b.c @@ -21,20 +21,123 @@ */ #include <subdev/clk.h> +#include <subdev/volt.h> +#include <subdev/timer.h> #include <core/device.h> +#include <core/tegra.h> #include "priv.h" #include "gk20a.h" -#define KHZ (1000) -#define MHZ (KHZ * 1000) - -#define MASK(w) ((1 << w) - 1) +#define GPCPLL_CFG_SYNC_MODE BIT(2) #define BYPASSCTRL_SYS (SYS_GPCPLL_CFG_BASE + 0x340) #define BYPASSCTRL_SYS_GPCPLL_SHIFT 0 #define BYPASSCTRL_SYS_GPCPLL_WIDTH 1 +#define GPCPLL_CFG2_SDM_DIN_SHIFT 0 +#define GPCPLL_CFG2_SDM_DIN_WIDTH 8 +#define GPCPLL_CFG2_SDM_DIN_MASK \ + (MASK(GPCPLL_CFG2_SDM_DIN_WIDTH) << GPCPLL_CFG2_SDM_DIN_SHIFT) +#define GPCPLL_CFG2_SDM_DIN_NEW_SHIFT 8 +#define GPCPLL_CFG2_SDM_DIN_NEW_WIDTH 15 +#define GPCPLL_CFG2_SDM_DIN_NEW_MASK \ + (MASK(GPCPLL_CFG2_SDM_DIN_NEW_WIDTH) << GPCPLL_CFG2_SDM_DIN_NEW_SHIFT) +#define GPCPLL_CFG2_SETUP2_SHIFT 16 +#define GPCPLL_CFG2_PLL_STEPA_SHIFT 24 + +#define GPCPLL_DVFS0 (SYS_GPCPLL_CFG_BASE + 0x10) +#define GPCPLL_DVFS0_DFS_COEFF_SHIFT 0 +#define GPCPLL_DVFS0_DFS_COEFF_WIDTH 7 +#define GPCPLL_DVFS0_DFS_COEFF_MASK \ + (MASK(GPCPLL_DVFS0_DFS_COEFF_WIDTH) << GPCPLL_DVFS0_DFS_COEFF_SHIFT) +#define GPCPLL_DVFS0_DFS_DET_MAX_SHIFT 8 +#define GPCPLL_DVFS0_DFS_DET_MAX_WIDTH 7 +#define GPCPLL_DVFS0_DFS_DET_MAX_MASK \ + (MASK(GPCPLL_DVFS0_DFS_DET_MAX_WIDTH) << GPCPLL_DVFS0_DFS_DET_MAX_SHIFT) + +#define GPCPLL_DVFS1 (SYS_GPCPLL_CFG_BASE + 0x14) +#define GPCPLL_DVFS1_DFS_EXT_DET_SHIFT 0 +#define GPCPLL_DVFS1_DFS_EXT_DET_WIDTH 7 +#define GPCPLL_DVFS1_DFS_EXT_STRB_SHIFT 7 +#define GPCPLL_DVFS1_DFS_EXT_STRB_WIDTH 1 +#define GPCPLL_DVFS1_DFS_EXT_CAL_SHIFT 8 +#define GPCPLL_DVFS1_DFS_EXT_CAL_WIDTH 7 +#define GPCPLL_DVFS1_DFS_EXT_SEL_SHIFT 15 +#define GPCPLL_DVFS1_DFS_EXT_SEL_WIDTH 1 +#define GPCPLL_DVFS1_DFS_CTRL_SHIFT 16 +#define GPCPLL_DVFS1_DFS_CTRL_WIDTH 12 +#define GPCPLL_DVFS1_EN_SDM_SHIFT 28 +#define GPCPLL_DVFS1_EN_SDM_WIDTH 1 +#define GPCPLL_DVFS1_EN_SDM_BIT BIT(28) +#define GPCPLL_DVFS1_EN_DFS_SHIFT 29 +#define GPCPLL_DVFS1_EN_DFS_WIDTH 1 +#define GPCPLL_DVFS1_EN_DFS_BIT BIT(29) +#define GPCPLL_DVFS1_EN_DFS_CAL_SHIFT 30 +#define GPCPLL_DVFS1_EN_DFS_CAL_WIDTH 1 +#define GPCPLL_DVFS1_EN_DFS_CAL_BIT BIT(30) +#define GPCPLL_DVFS1_DFS_CAL_DONE_SHIFT 31 +#define GPCPLL_DVFS1_DFS_CAL_DONE_WIDTH 1 +#define GPCPLL_DVFS1_DFS_CAL_DONE_BIT BIT(31) + +#define GPC_BCAST_GPCPLL_DVFS2 (GPC_BCAST_GPCPLL_CFG_BASE + 0x20) +#define GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT BIT(16) + +#define GPCPLL_CFG3_PLL_DFS_TESTOUT_SHIFT 24 +#define GPCPLL_CFG3_PLL_DFS_TESTOUT_WIDTH 7 + +#define DFS_DET_RANGE 6 /* -2^6 ... 2^6-1 */ +#define SDM_DIN_RANGE 12 /* -2^12 ... 2^12-1 */ + +struct gm20b_clk_dvfs_params { + s32 coeff_slope; + s32 coeff_offs; + u32 vco_ctrl; +}; + +static const struct gm20b_clk_dvfs_params gm20b_dvfs_params = { + .coeff_slope = -165230, + .coeff_offs = 214007, + .vco_ctrl = 0x7 << 3, +}; + +/* + * base.n is now the *integer* part of the N factor. + * sdm_din contains n's decimal part. + */ +struct gm20b_pll { + struct gk20a_pll base; + u32 sdm_din; +}; + +struct gm20b_clk_dvfs { + u32 dfs_coeff; + s32 dfs_det_max; + s32 dfs_ext_cal; +}; + +struct gm20b_clk { + /* currently applied parameters */ + struct gk20a_clk base; + struct gm20b_clk_dvfs dvfs; + u32 uv; + + /* new parameters to apply */ + struct gk20a_pll new_pll; + struct gm20b_clk_dvfs new_dvfs; + u32 new_uv; + + const struct gm20b_clk_dvfs_params *dvfs_params; + + /* fused parameters */ + s32 uvdet_slope; + s32 uvdet_offs; + + /* safe frequency we can use at minimum voltage */ + u32 safe_fmax_vmin; +}; +#define gm20b_clk(p) container_of((gk20a_clk(p)), struct gm20b_clk, base) + static u32 pl_to_div(u32 pl) { return pl; @@ -53,6 +156,484 @@ static const struct gk20a_clk_pllg_params gm20b_pllg_params = { .min_pl = 1, .max_pl = 31, }; +static void +gm20b_pllg_read_mnp(struct gm20b_clk *clk, struct gm20b_pll *pll) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + struct nvkm_device *device = subdev->device; + u32 val; + + gk20a_pllg_read_mnp(&clk->base, &pll->base); + val = nvkm_rd32(device, GPCPLL_CFG2); + pll->sdm_din = (val >> GPCPLL_CFG2_SDM_DIN_SHIFT) & + MASK(GPCPLL_CFG2_SDM_DIN_WIDTH); +} + +static void +gm20b_pllg_write_mnp(struct gm20b_clk *clk, const struct gm20b_pll *pll) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + + nvkm_mask(device, GPCPLL_CFG2, GPCPLL_CFG2_SDM_DIN_MASK, + pll->sdm_din << GPCPLL_CFG2_SDM_DIN_SHIFT); + gk20a_pllg_write_mnp(&clk->base, &pll->base); +} + +/* + * Determine DFS_COEFF for the requested voltage. Always select external + * calibration override equal to the voltage, and set maximum detection + * limit "0" (to make sure that PLL output remains under F/V curve when + * voltage increases). + */ +static void +gm20b_dvfs_calc_det_coeff(struct gm20b_clk *clk, s32 uv, + struct gm20b_clk_dvfs *dvfs) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + const struct gm20b_clk_dvfs_params *p = clk->dvfs_params; + u32 coeff; + /* Work with mv as uv would likely trigger an overflow */ + s32 mv = DIV_ROUND_CLOSEST(uv, 1000); + + /* coeff = slope * voltage + offset */ + coeff = DIV_ROUND_CLOSEST(mv * p->coeff_slope, 1000) + p->coeff_offs; + coeff = DIV_ROUND_CLOSEST(coeff, 1000); + dvfs->dfs_coeff = min_t(u32, coeff, MASK(GPCPLL_DVFS0_DFS_COEFF_WIDTH)); + + dvfs->dfs_ext_cal = DIV_ROUND_CLOSEST(uv - clk->uvdet_offs, + clk->uvdet_slope); + /* should never happen */ + if (abs(dvfs->dfs_ext_cal) >= BIT(DFS_DET_RANGE)) + nvkm_error(subdev, "dfs_ext_cal overflow!\n"); + + dvfs->dfs_det_max = 0; + + nvkm_debug(subdev, "%s uv: %d coeff: %x, ext_cal: %d, det_max: %d\n", + __func__, uv, dvfs->dfs_coeff, dvfs->dfs_ext_cal, + dvfs->dfs_det_max); +} + +/* + * Solve equation for integer and fractional part of the effective NDIV: + * + * n_eff = n_int + 1/2 + (SDM_DIN / 2^(SDM_DIN_RANGE + 1)) + + * (DVFS_COEFF * DVFS_DET_DELTA) / 2^DFS_DET_RANGE + * + * The SDM_DIN LSB is finally shifted out, since it is not accessible by sw. + */ +static void +gm20b_dvfs_calc_ndiv(struct gm20b_clk *clk, u32 n_eff, u32 *n_int, u32 *sdm_din) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + const struct gk20a_clk_pllg_params *p = clk->base.params; + u32 n; + s32 det_delta; + u32 rem, rem_range; + + /* calculate current ext_cal and subtract previous one */ + det_delta = DIV_ROUND_CLOSEST(((s32)clk->uv) - clk->uvdet_offs, + clk->uvdet_slope); + det_delta -= clk->dvfs.dfs_ext_cal; + det_delta = min(det_delta, clk->dvfs.dfs_det_max); + det_delta *= clk->dvfs.dfs_coeff; + + /* integer part of n */ + n = (n_eff << DFS_DET_RANGE) - det_delta; + /* should never happen! */ + if (n <= 0) { + nvkm_error(subdev, "ndiv <= 0 - setting to 1...\n"); + n = 1 << DFS_DET_RANGE; + } + if (n >> DFS_DET_RANGE > p->max_n) { + nvkm_error(subdev, "ndiv > max_n - setting to max_n...\n"); + n = p->max_n << DFS_DET_RANGE; + } + *n_int = n >> DFS_DET_RANGE; + + /* fractional part of n */ + rem = ((u32)n) & MASK(DFS_DET_RANGE); + rem_range = SDM_DIN_RANGE + 1 - DFS_DET_RANGE; + /* subtract 2^SDM_DIN_RANGE to account for the 1/2 of the equation */ + rem = (rem << rem_range) - BIT(SDM_DIN_RANGE); + /* lose 8 LSB and clip - sdm_din only keeps the most significant byte */ + *sdm_din = (rem >> BITS_PER_BYTE) & MASK(GPCPLL_CFG2_SDM_DIN_WIDTH); + + nvkm_debug(subdev, "%s n_eff: %d, n_int: %d, sdm_din: %d\n", __func__, + n_eff, *n_int, *sdm_din); +} + +static int +gm20b_pllg_slide(struct gm20b_clk *clk, u32 n) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + struct nvkm_device *device = subdev->device; + struct gm20b_pll pll; + u32 n_int, sdm_din; + int ret = 0; + + /* calculate the new n_int/sdm_din for this n/uv */ + gm20b_dvfs_calc_ndiv(clk, n, &n_int, &sdm_din); + + /* get old coefficients */ + gm20b_pllg_read_mnp(clk, &pll); + /* do nothing if NDIV is the same */ + if (n_int == pll.base.n && sdm_din == pll.sdm_din) + return 0; + + /* pll slowdown mode */ + nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, + BIT(GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT), + BIT(GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT)); + + /* new ndiv ready for ramp */ + /* in DVFS mode SDM is updated via "new" field */ + nvkm_mask(device, GPCPLL_CFG2, GPCPLL_CFG2_SDM_DIN_NEW_MASK, + sdm_din << GPCPLL_CFG2_SDM_DIN_NEW_SHIFT); + pll.base.n = n_int; + udelay(1); + gk20a_pllg_write_mnp(&clk->base, &pll.base); + + /* dynamic ramp to new ndiv */ + udelay(1); + nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, + BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT), + BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT)); + + /* wait for ramping to complete */ + if (nvkm_wait_usec(device, 500, GPC_BCAST_NDIV_SLOWDOWN_DEBUG, + GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK, + GPC_BCAST_NDIV_SLOWDOWN_DEBUG_PLL_DYNRAMP_DONE_SYNCED_MASK) < 0) + ret = -ETIMEDOUT; + + /* in DVFS mode complete SDM update */ + nvkm_mask(device, GPCPLL_CFG2, GPCPLL_CFG2_SDM_DIN_MASK, + sdm_din << GPCPLL_CFG2_SDM_DIN_SHIFT); + + /* exit slowdown mode */ + nvkm_mask(device, GPCPLL_NDIV_SLOWDOWN, + BIT(GPCPLL_NDIV_SLOWDOWN_SLOWDOWN_USING_PLL_SHIFT) | + BIT(GPCPLL_NDIV_SLOWDOWN_EN_DYNRAMP_SHIFT), 0); + nvkm_rd32(device, GPCPLL_NDIV_SLOWDOWN); + + return ret; +} + +static int +gm20b_pllg_enable(struct gm20b_clk *clk) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_ENABLE, GPCPLL_CFG_ENABLE); + nvkm_rd32(device, GPCPLL_CFG); + + /* In DVFS mode lock cannot be used - so just delay */ + udelay(40); + + /* set SYNC_MODE for glitchless switch out of bypass */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_SYNC_MODE, + GPCPLL_CFG_SYNC_MODE); + nvkm_rd32(device, GPCPLL_CFG); + + /* switch to VCO mode */ + nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), + BIT(SEL_VCO_GPC2CLK_OUT_SHIFT)); + + return 0; +} + +static void +gm20b_pllg_disable(struct gm20b_clk *clk) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + + /* put PLL in bypass before disabling it */ + nvkm_mask(device, SEL_VCO, BIT(SEL_VCO_GPC2CLK_OUT_SHIFT), 0); + + /* clear SYNC_MODE before disabling PLL */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_SYNC_MODE, 0); + + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_ENABLE, 0); + nvkm_rd32(device, GPCPLL_CFG); +} + +static int +gm20b_pllg_program_mnp(struct gm20b_clk *clk, const struct gk20a_pll *pll) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + struct nvkm_device *device = subdev->device; + struct gm20b_pll cur_pll; + u32 n_int, sdm_din; + /* if we only change pdiv, we can do a glitchless transition */ + bool pdiv_only; + int ret; + + gm20b_dvfs_calc_ndiv(clk, pll->n, &n_int, &sdm_din); + gm20b_pllg_read_mnp(clk, &cur_pll); + pdiv_only = cur_pll.base.n == n_int && cur_pll.sdm_din == sdm_din && + cur_pll.base.m == pll->m; + + /* need full sequence if clock not enabled yet */ + if (!gk20a_pllg_is_enabled(&clk->base)) + pdiv_only = false; + + /* split VCO-to-bypass jump in half by setting out divider 1:2 */ + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV2 << GPC2CLK_OUT_VCODIV_SHIFT); + /* Intentional 2nd write to assure linear divider operation */ + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV2 << GPC2CLK_OUT_VCODIV_SHIFT); + nvkm_rd32(device, GPC2CLK_OUT); + udelay(2); + + if (pdiv_only) { + u32 old = cur_pll.base.pl; + u32 new = pll->pl; + + /* + * we can do a glitchless transition only if the old and new PL + * parameters share at least one bit set to 1. If this is not + * the case, calculate and program an interim PL that will allow + * us to respect that rule. + */ + if ((old & new) == 0) { + cur_pll.base.pl = min(old | BIT(ffs(new) - 1), + new | BIT(ffs(old) - 1)); + gk20a_pllg_write_mnp(&clk->base, &cur_pll.base); + } + + cur_pll.base.pl = new; + gk20a_pllg_write_mnp(&clk->base, &cur_pll.base); + } else { + /* disable before programming if more than pdiv changes */ + gm20b_pllg_disable(clk); + + cur_pll.base = *pll; + cur_pll.base.n = n_int; + cur_pll.sdm_din = sdm_din; + gm20b_pllg_write_mnp(clk, &cur_pll); + + ret = gm20b_pllg_enable(clk); + if (ret) + return ret; + } + + /* restore out divider 1:1 */ + udelay(2); + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT); + /* Intentional 2nd write to assure linear divider operation */ + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_VCODIV_MASK, + GPC2CLK_OUT_VCODIV1 << GPC2CLK_OUT_VCODIV_SHIFT); + nvkm_rd32(device, GPC2CLK_OUT); + + return 0; +} + +static int +gm20b_pllg_program_mnp_slide(struct gm20b_clk *clk, const struct gk20a_pll *pll) +{ + struct gk20a_pll cur_pll; + int ret; + + if (gk20a_pllg_is_enabled(&clk->base)) { + gk20a_pllg_read_mnp(&clk->base, &cur_pll); + + /* just do NDIV slide if there is no change to M and PL */ + if (pll->m == cur_pll.m && pll->pl == cur_pll.pl) + return gm20b_pllg_slide(clk, pll->n); + + /* slide down to current NDIV_LO */ + cur_pll.n = gk20a_pllg_n_lo(&clk->base, &cur_pll); + ret = gm20b_pllg_slide(clk, cur_pll.n); + if (ret) + return ret; + } + + /* program MNP with the new clock parameters and new NDIV_LO */ + cur_pll = *pll; + cur_pll.n = gk20a_pllg_n_lo(&clk->base, &cur_pll); + ret = gm20b_pllg_program_mnp(clk, &cur_pll); + if (ret) + return ret; + + /* slide up to new NDIV */ + return gm20b_pllg_slide(clk, pll->n); +} + +static int +gm20b_clk_calc(struct nvkm_clk *base, struct nvkm_cstate *cstate) +{ + struct gm20b_clk *clk = gm20b_clk(base); + struct nvkm_subdev *subdev = &base->subdev; + struct nvkm_volt *volt = base->subdev.device->volt; + int ret; + + ret = gk20a_pllg_calc_mnp(&clk->base, cstate->domain[nv_clk_src_gpc] * + GK20A_CLK_GPC_MDIV, &clk->new_pll); + if (ret) + return ret; + + clk->new_uv = volt->vid[cstate->voltage].uv; + gm20b_dvfs_calc_det_coeff(clk, clk->new_uv, &clk->new_dvfs); + + nvkm_debug(subdev, "%s uv: %d uv\n", __func__, clk->new_uv); + + return 0; +} + +/* + * Compute PLL parameters that are always safe for the current voltage + */ +static void +gm20b_dvfs_calc_safe_pll(struct gm20b_clk *clk, struct gk20a_pll *pll) +{ + u32 rate = gk20a_pllg_calc_rate(&clk->base, pll) / KHZ; + u32 parent_rate = clk->base.parent_rate / KHZ; + u32 nmin, nsafe; + + /* remove a safe margin of 10% */ + if (rate > clk->safe_fmax_vmin) + rate = rate * (100 - 10) / 100; + + /* gpc2clk */ + rate *= 2; + + nmin = DIV_ROUND_UP(pll->m * clk->base.params->min_vco, parent_rate); + nsafe = pll->m * rate / (clk->base.parent_rate); + + if (nsafe < nmin) { + pll->pl = DIV_ROUND_UP(nmin * parent_rate, pll->m * rate); + nsafe = nmin; + } + + pll->n = nsafe; +} + +static void +gm20b_dvfs_program_coeff(struct gm20b_clk *clk, u32 coeff) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + + /* strobe to read external DFS coefficient */ + nvkm_mask(device, GPC_BCAST_GPCPLL_DVFS2, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT); + + nvkm_mask(device, GPCPLL_DVFS0, GPCPLL_DVFS0_DFS_COEFF_MASK, + coeff << GPCPLL_DVFS0_DFS_COEFF_SHIFT); + + udelay(1); + nvkm_mask(device, GPC_BCAST_GPCPLL_DVFS2, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT, 0); +} + +static void +gm20b_dvfs_program_ext_cal(struct gm20b_clk *clk, u32 dfs_det_cal) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + u32 val; + + nvkm_mask(device, GPC_BCAST_GPCPLL_DVFS2, MASK(DFS_DET_RANGE + 1), + dfs_det_cal); + udelay(1); + + val = nvkm_rd32(device, GPCPLL_DVFS1); + if (!(val & BIT(25))) { + /* Use external value to overwrite calibration value */ + val |= BIT(25) | BIT(16); + nvkm_wr32(device, GPCPLL_DVFS1, val); + } +} + +static void +gm20b_dvfs_program_dfs_detection(struct gm20b_clk *clk, + struct gm20b_clk_dvfs *dvfs) +{ + struct nvkm_device *device = clk->base.base.subdev.device; + + /* strobe to read external DFS coefficient */ + nvkm_mask(device, GPC_BCAST_GPCPLL_DVFS2, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT); + + nvkm_mask(device, GPCPLL_DVFS0, + GPCPLL_DVFS0_DFS_COEFF_MASK | GPCPLL_DVFS0_DFS_DET_MAX_MASK, + dvfs->dfs_coeff << GPCPLL_DVFS0_DFS_COEFF_SHIFT | + dvfs->dfs_det_max << GPCPLL_DVFS0_DFS_DET_MAX_SHIFT); + + udelay(1); + nvkm_mask(device, GPC_BCAST_GPCPLL_DVFS2, + GPC_BCAST_GPCPLL_DVFS2_DFS_EXT_STROBE_BIT, 0); + + gm20b_dvfs_program_ext_cal(clk, dvfs->dfs_ext_cal); +} + +static int +gm20b_clk_prog(struct nvkm_clk *base) +{ + struct gm20b_clk *clk = gm20b_clk(base); + u32 cur_freq; + int ret; + + /* No change in DVFS settings? */ + if (clk->uv == clk->new_uv) + goto prog; + + /* + * Interim step for changing DVFS detection settings: low enough + * frequency to be safe at at DVFS coeff = 0. + * + * 1. If voltage is increasing: + * - safe frequency target matches the lowest - old - frequency + * - DVFS settings are still old + * - Voltage already increased to new level by volt, but maximum + * detection limit assures PLL output remains under F/V curve + * + * 2. If voltage is decreasing: + * - safe frequency target matches the lowest - new - frequency + * - DVFS settings are still old + * - Voltage is also old, it will be lowered by volt afterwards + * + * Interim step can be skipped if old frequency is below safe minimum, + * i.e., it is low enough to be safe at any voltage in operating range + * with zero DVFS coefficient. + */ + cur_freq = nvkm_clk_read(&clk->base.base, nv_clk_src_gpc); + if (cur_freq > clk->safe_fmax_vmin) { + struct gk20a_pll pll_safe; + + if (clk->uv < clk->new_uv) + /* voltage will raise: safe frequency is current one */ + pll_safe = clk->base.pll; + else + /* voltage will drop: safe frequency is new one */ + pll_safe = clk->new_pll; + + gm20b_dvfs_calc_safe_pll(clk, &pll_safe); + ret = gm20b_pllg_program_mnp_slide(clk, &pll_safe); + if (ret) + return ret; + } + + /* + * DVFS detection settings transition: + * - Set DVFS coefficient zero + * - Set calibration level to new voltage + * - Set DVFS coefficient to match new voltage + */ + gm20b_dvfs_program_coeff(clk, 0); + gm20b_dvfs_program_ext_cal(clk, clk->new_dvfs.dfs_ext_cal); + gm20b_dvfs_program_coeff(clk, clk->new_dvfs.dfs_coeff); + gm20b_dvfs_program_dfs_detection(clk, &clk->new_dvfs); + +prog: + clk->uv = clk->new_uv; + clk->dvfs = clk->new_dvfs; + clk->base.pll = clk->new_pll; + + return gm20b_pllg_program_mnp_slide(clk, &clk->base.pll); +} + static struct nvkm_pstate gm20b_pstates[] = { { @@ -133,9 +714,99 @@ gm20b_pstates[] = { .voltage = 12, }, }, - }; +static void +gm20b_clk_fini(struct nvkm_clk *base) +{ + struct nvkm_device *device = base->subdev.device; + struct gm20b_clk *clk = gm20b_clk(base); + + /* slide to VCO min */ + if (gk20a_pllg_is_enabled(&clk->base)) { + struct gk20a_pll pll; + u32 n_lo; + + gk20a_pllg_read_mnp(&clk->base, &pll); + n_lo = gk20a_pllg_n_lo(&clk->base, &pll); + gm20b_pllg_slide(clk, n_lo); + } + + gm20b_pllg_disable(clk); + + /* set IDDQ */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_IDDQ, 1); +} + +static int +gm20b_clk_init_dvfs(struct gm20b_clk *clk) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + struct nvkm_device *device = subdev->device; + bool fused = clk->uvdet_offs && clk->uvdet_slope; + static const s32 ADC_SLOPE_UV = 10000; /* default ADC detection slope */ + u32 data; + int ret; + + /* Enable NA DVFS */ + nvkm_mask(device, GPCPLL_DVFS1, GPCPLL_DVFS1_EN_DFS_BIT, + GPCPLL_DVFS1_EN_DFS_BIT); + + /* Set VCO_CTRL */ + if (clk->dvfs_params->vco_ctrl) + nvkm_mask(device, GPCPLL_CFG3, GPCPLL_CFG3_VCO_CTRL_MASK, + clk->dvfs_params->vco_ctrl << GPCPLL_CFG3_VCO_CTRL_SHIFT); + + if (fused) { + /* Start internal calibration, but ignore results */ + nvkm_mask(device, GPCPLL_DVFS1, GPCPLL_DVFS1_EN_DFS_CAL_BIT, + GPCPLL_DVFS1_EN_DFS_CAL_BIT); + + /* got uvdev parameters from fuse, skip calibration */ + goto calibrated; + } + + /* + * If calibration parameters are not fused, start internal calibration, + * wait for completion, and use results along with default slope to + * calculate ADC offset during boot. + */ + nvkm_mask(device, GPCPLL_DVFS1, GPCPLL_DVFS1_EN_DFS_CAL_BIT, + GPCPLL_DVFS1_EN_DFS_CAL_BIT); + + /* Wait for internal calibration done (spec < 2us). */ + ret = nvkm_wait_usec(device, 10, GPCPLL_DVFS1, + GPCPLL_DVFS1_DFS_CAL_DONE_BIT, + GPCPLL_DVFS1_DFS_CAL_DONE_BIT); + if (ret < 0) { + nvkm_error(subdev, "GPCPLL calibration timeout\n"); + return -ETIMEDOUT; + } + + data = nvkm_rd32(device, GPCPLL_CFG3) >> + GPCPLL_CFG3_PLL_DFS_TESTOUT_SHIFT; + data &= MASK(GPCPLL_CFG3_PLL_DFS_TESTOUT_WIDTH); + + clk->uvdet_slope = ADC_SLOPE_UV; + clk->uvdet_offs = ((s32)clk->uv) - data * ADC_SLOPE_UV; + + nvkm_debug(subdev, "calibrated DVFS parameters: offs %d, slope %d\n", + clk->uvdet_offs, clk->uvdet_slope); + +calibrated: + /* Compute and apply initial DVFS parameters */ + gm20b_dvfs_calc_det_coeff(clk, clk->uv, &clk->dvfs); + gm20b_dvfs_program_coeff(clk, 0); + gm20b_dvfs_program_ext_cal(clk, clk->dvfs.dfs_ext_cal); + gm20b_dvfs_program_coeff(clk, clk->dvfs.dfs_coeff); + gm20b_dvfs_program_dfs_detection(clk, &clk->new_dvfs); + + return 0; +} + +/* Forward declaration to detect speedo >=1 in gm20b_clk_init() */ +static const struct nvkm_clk_func gm20b_clk; + static int gm20b_clk_init(struct nvkm_clk *base) { @@ -143,15 +814,56 @@ gm20b_clk_init(struct nvkm_clk *base) struct nvkm_subdev *subdev = &clk->base.subdev; struct nvkm_device *device = subdev->device; int ret; + u32 data; + + /* get out from IDDQ */ + nvkm_mask(device, GPCPLL_CFG, GPCPLL_CFG_IDDQ, 0); + nvkm_rd32(device, GPCPLL_CFG); + udelay(5); + + nvkm_mask(device, GPC2CLK_OUT, GPC2CLK_OUT_INIT_MASK, + GPC2CLK_OUT_INIT_VAL); /* Set the global bypass control to VCO */ nvkm_mask(device, BYPASSCTRL_SYS, MASK(BYPASSCTRL_SYS_GPCPLL_WIDTH) << BYPASSCTRL_SYS_GPCPLL_SHIFT, 0); + ret = gk20a_clk_setup_slide(clk); + if (ret) + return ret; + + /* If not fused, set RAM SVOP PDP data 0x2, and enable fuse override */ + data = nvkm_rd32(device, 0x021944); + if (!(data & 0x3)) { + data |= 0x2; + nvkm_wr32(device, 0x021944, data); + + data = nvkm_rd32(device, 0x021948); + data |= 0x1; + nvkm_wr32(device, 0x021948, data); + } + + /* Disable idle slow down */ + nvkm_mask(device, 0x20160, 0x003f0000, 0x0); + + /* speedo >= 1? */ + if (clk->base.func == &gm20b_clk) { + struct gm20b_clk *_clk = gm20b_clk(base); + struct nvkm_volt *volt = device->volt; + + /* Get current voltage */ + _clk->uv = nvkm_volt_get(volt); + + /* Initialize DVFS */ + ret = gm20b_clk_init_dvfs(_clk); + if (ret) + return ret; + } + /* Start with lowest frequency */ base->func->calc(base, &base->func->pstates[0].base); - ret = base->func->prog(&clk->base); + ret = base->func->prog(base); if (ret) { nvkm_error(subdev, "cannot initialize clock\n"); return ret; @@ -169,6 +881,7 @@ gm20b_clk_speedo0 = { .prog = gk20a_clk_prog, .tidy = gk20a_clk_tidy, .pstates = gm20b_pstates, + /* Speedo 0 only supports 12 voltages */ .nr_pstates = ARRAY_SIZE(gm20b_pstates) - 1, .domains = { { nv_clk_src_crystal, 0xff }, @@ -177,8 +890,26 @@ gm20b_clk_speedo0 = { }, }; -int -gm20b_clk_new(struct nvkm_device *device, int index, struct nvkm_clk **pclk) +static const struct nvkm_clk_func +gm20b_clk = { + .init = gm20b_clk_init, + .fini = gm20b_clk_fini, + .read = gk20a_clk_read, + .calc = gm20b_clk_calc, + .prog = gm20b_clk_prog, + .tidy = gk20a_clk_tidy, + .pstates = gm20b_pstates, + .nr_pstates = ARRAY_SIZE(gm20b_pstates), + .domains = { + { nv_clk_src_crystal, 0xff }, + { nv_clk_src_gpc, 0xff, 0, "core", GK20A_CLK_GPC_MDIV }, + { nv_clk_src_max }, + }, +}; + +static int +gm20b_clk_new_speedo0(struct nvkm_device *device, int index, + struct nvkm_clk **pclk) { struct gk20a_clk *clk; int ret; @@ -188,11 +919,156 @@ gm20b_clk_new(struct nvkm_device *device, int index, struct nvkm_clk **pclk) return -ENOMEM; *pclk = &clk->base; - ret = _gk20a_clk_ctor(device, index, &gm20b_clk_speedo0, - &gm20b_pllg_params, clk); + ret = gk20a_clk_ctor(device, index, &gm20b_clk_speedo0, + &gm20b_pllg_params, clk); clk->pl_to_div = pl_to_div; clk->div_to_pl = div_to_pl; return ret; } + +/* FUSE register */ +#define FUSE_RESERVED_CALIB0 0x204 +#define FUSE_RESERVED_CALIB0_INTERCEPT_FRAC_SHIFT 0 +#define FUSE_RESERVED_CALIB0_INTERCEPT_FRAC_WIDTH 4 +#define FUSE_RESERVED_CALIB0_INTERCEPT_INT_SHIFT 4 +#define FUSE_RESERVED_CALIB0_INTERCEPT_INT_WIDTH 10 +#define FUSE_RESERVED_CALIB0_SLOPE_FRAC_SHIFT 14 +#define FUSE_RESERVED_CALIB0_SLOPE_FRAC_WIDTH 10 +#define FUSE_RESERVED_CALIB0_SLOPE_INT_SHIFT 24 +#define FUSE_RESERVED_CALIB0_SLOPE_INT_WIDTH 6 +#define FUSE_RESERVED_CALIB0_FUSE_REV_SHIFT 30 +#define FUSE_RESERVED_CALIB0_FUSE_REV_WIDTH 2 + +static int +gm20b_clk_init_fused_params(struct gm20b_clk *clk) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + u32 val = 0; + u32 rev = 0; + +#if IS_ENABLED(CONFIG_ARCH_TEGRA) + tegra_fuse_readl(FUSE_RESERVED_CALIB0, &val); + rev = (val >> FUSE_RESERVED_CALIB0_FUSE_REV_SHIFT) & + MASK(FUSE_RESERVED_CALIB0_FUSE_REV_WIDTH); +#endif + + /* No fused parameters, we will calibrate later */ + if (rev == 0) + return -EINVAL; + + /* Integer part in mV + fractional part in uV */ + clk->uvdet_slope = ((val >> FUSE_RESERVED_CALIB0_SLOPE_INT_SHIFT) & + MASK(FUSE_RESERVED_CALIB0_SLOPE_INT_WIDTH)) * 1000 + + ((val >> FUSE_RESERVED_CALIB0_SLOPE_FRAC_SHIFT) & + MASK(FUSE_RESERVED_CALIB0_SLOPE_FRAC_WIDTH)); + + /* Integer part in mV + fractional part in 100uV */ + clk->uvdet_offs = ((val >> FUSE_RESERVED_CALIB0_INTERCEPT_INT_SHIFT) & + MASK(FUSE_RESERVED_CALIB0_INTERCEPT_INT_WIDTH)) * 1000 + + ((val >> FUSE_RESERVED_CALIB0_INTERCEPT_FRAC_SHIFT) & + MASK(FUSE_RESERVED_CALIB0_INTERCEPT_FRAC_WIDTH)) * 100; + + nvkm_debug(subdev, "fused calibration data: slope %d, offs %d\n", + clk->uvdet_slope, clk->uvdet_offs); + return 0; +} + +static int +gm20b_clk_init_safe_fmax(struct gm20b_clk *clk) +{ + struct nvkm_subdev *subdev = &clk->base.base.subdev; + struct nvkm_volt *volt = subdev->device->volt; + struct nvkm_pstate *pstates = clk->base.base.func->pstates; + int nr_pstates = clk->base.base.func->nr_pstates; + int vmin, id = 0; + u32 fmax = 0; + int i; + + /* find lowest voltage we can use */ + vmin = volt->vid[0].uv; + for (i = 1; i < volt->vid_nr; i++) { + if (volt->vid[i].uv <= vmin) { + vmin = volt->vid[i].uv; + id = volt->vid[i].vid; + } + } + + /* find max frequency at this voltage */ + for (i = 0; i < nr_pstates; i++) + if (pstates[i].base.voltage == id) + fmax = max(fmax, + pstates[i].base.domain[nv_clk_src_gpc]); + + if (!fmax) { + nvkm_error(subdev, "failed to evaluate safe fmax\n"); + return -EINVAL; + } + + /* we are safe at 90% of the max frequency */ + clk->safe_fmax_vmin = fmax * (100 - 10) / 100; + nvkm_debug(subdev, "safe fmax @ vmin = %u Khz\n", clk->safe_fmax_vmin); + + return 0; +} + +int +gm20b_clk_new(struct nvkm_device *device, int index, struct nvkm_clk **pclk) +{ + struct nvkm_device_tegra *tdev = device->func->tegra(device); + struct gm20b_clk *clk; + struct nvkm_subdev *subdev; + struct gk20a_clk_pllg_params *clk_params; + int ret; + + /* Speedo 0 GPUs cannot use noise-aware PLL */ + if (tdev->gpu_speedo_id == 0) + return gm20b_clk_new_speedo0(device, index, pclk); + + /* Speedo >= 1, use NAPLL */ + clk = kzalloc(sizeof(*clk) + sizeof(*clk_params), GFP_KERNEL); + if (!clk) + return -ENOMEM; + *pclk = &clk->base.base; + subdev = &clk->base.base.subdev; + + /* duplicate the clock parameters since we will patch them below */ + clk_params = (void *) (clk + 1); + *clk_params = gm20b_pllg_params; + ret = gk20a_clk_ctor(device, index, &gm20b_clk, clk_params, + &clk->base); + if (ret) + return ret; + + /* + * NAPLL can only work with max_u, clamp the m range so + * gk20a_pllg_calc_mnp always uses it + */ + clk_params->max_m = clk_params->min_m = DIV_ROUND_UP(clk_params->max_u, + (clk->base.parent_rate / KHZ)); + if (clk_params->max_m == 0) { + nvkm_warn(subdev, "cannot use NAPLL, using legacy clock...\n"); + kfree(clk); + return gm20b_clk_new_speedo0(device, index, pclk); + } + + clk->base.pl_to_div = pl_to_div; + clk->base.div_to_pl = div_to_pl; + + clk->dvfs_params = &gm20b_dvfs_params; + + ret = gm20b_clk_init_fused_params(clk); + /* + * we will calibrate during init - should never happen on + * prod parts + */ + if (ret) + nvkm_warn(subdev, "no fused calibration parameters\n"); + + ret = gm20b_clk_init_safe_fmax(clk); + if (ret) + return ret; + + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild index 842d5de96d73..edcc157e6ac8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild @@ -24,6 +24,8 @@ nvkm-y += nvkm/subdev/fb/gk104.o nvkm-y += nvkm/subdev/fb/gk20a.o nvkm-y += nvkm/subdev/fb/gm107.o nvkm-y += nvkm/subdev/fb/gm200.o +nvkm-y += nvkm/subdev/fb/gp100.o +nvkm-y += nvkm/subdev/fb/gp104.o nvkm-y += nvkm/subdev/fb/ram.o nvkm-y += nvkm/subdev/fb/ramnv04.o @@ -41,6 +43,7 @@ nvkm-y += nvkm/subdev/fb/rammcp77.o nvkm-y += nvkm/subdev/fb/ramgf100.o nvkm-y += nvkm/subdev/fb/ramgk104.o nvkm-y += nvkm/subdev/fb/ramgm107.o +nvkm-y += nvkm/subdev/fb/ramgp100.o nvkm-y += nvkm/subdev/fb/sddr2.o nvkm-y += nvkm/subdev/fb/sddr3.o nvkm-y += nvkm/subdev/fb/gddr3.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c index ce90242b8cce..a7049c041594 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/base.c @@ -25,6 +25,7 @@ #include "ram.h" #include <core/memory.h> +#include <core/option.h> #include <subdev/bios.h> #include <subdev/bios/M0203.h> #include <engine/gr.h> @@ -134,6 +135,10 @@ nvkm_fb_init(struct nvkm_subdev *subdev) if (fb->func->init) fb->func->init(fb); + if (fb->func->init_page) + fb->func->init_page(fb); + if (fb->func->init_unkn) + fb->func->init_unkn(fb); return 0; } @@ -171,6 +176,7 @@ nvkm_fb_ctor(const struct nvkm_fb_func *func, struct nvkm_device *device, nvkm_subdev_ctor(&nvkm_fb, device, index, &fb->subdev); fb->func = func; fb->tile.regions = fb->func->tile.regions; + fb->page = nvkm_longopt(device->cfgopt, "NvFbBigPage", 0); } int diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c index e649ead5ccfc..76433cc66fff 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c @@ -72,6 +72,22 @@ gf100_fb_oneinit(struct nvkm_fb *fb) } void +gf100_fb_init_page(struct nvkm_fb *fb) +{ + struct nvkm_device *device = fb->subdev.device; + switch (fb->page) { + case 16: + nvkm_mask(device, 0x100c80, 0x00000001, 0x00000001); + break; + case 17: + default: + nvkm_mask(device, 0x100c80, 0x00000001, 0x00000000); + fb->page = 17; + break; + } +} + +void gf100_fb_init(struct nvkm_fb *base) { struct gf100_fb *fb = gf100_fb(base); @@ -79,8 +95,6 @@ gf100_fb_init(struct nvkm_fb *base) if (fb->r100c10_page) nvkm_wr32(device, 0x100c10, fb->r100c10 >> 8); - - nvkm_mask(device, 0x100c80, 0x00000001, 0x00000000); /* 128KiB lpg */ } void * @@ -125,6 +139,7 @@ gf100_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gf100_fb_init, + .init_page = gf100_fb_init_page, .intr = gf100_fb_intr, .ram_new = gf100_ram_new, .memtype_valid = gf100_fb_memtype_valid, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.h index 2160e5a39c9a..449f431644b3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.h @@ -14,4 +14,6 @@ int gf100_fb_new_(const struct nvkm_fb_func *, struct nvkm_device *, void *gf100_fb_dtor(struct nvkm_fb *); void gf100_fb_init(struct nvkm_fb *); void gf100_fb_intr(struct nvkm_fb *); + +void gp100_fb_init(struct nvkm_fb *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk104.c index b41f0f70038c..4245e2e6e604 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk104.c @@ -29,6 +29,7 @@ gk104_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gf100_fb_init, + .init_page = gf100_fb_init_page, .intr = gf100_fb_intr, .ram_new = gk104_ram_new, .memtype_valid = gf100_fb_memtype_valid, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk20a.c index 7306f7dfc3b9..f815fe2bbf08 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gk20a.c @@ -27,7 +27,6 @@ static void gk20a_fb_init(struct nvkm_fb *fb) { struct nvkm_device *device = fb->subdev.device; - nvkm_mask(device, 0x100c80, 0x00000001, 0x00000000); /* 128KiB lpg */ nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->mmu_wr) >> 8); nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->mmu_rd) >> 8); } @@ -36,6 +35,7 @@ static const struct nvkm_fb_func gk20a_fb = { .oneinit = gf100_fb_oneinit, .init = gk20a_fb_init, + .init_page = gf100_fb_init_page, .memtype_valid = gf100_fb_memtype_valid, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm107.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm107.c index 4869fdb753c9..db699025f546 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm107.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm107.c @@ -29,6 +29,7 @@ gm107_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gf100_fb_init, + .init_page = gf100_fb_init_page, .intr = gf100_fb_intr, .ram_new = gm107_ram_new, .memtype_valid = gf100_fb_memtype_valid, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c index 44f5716f64d8..62f653240be3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c @@ -26,6 +26,24 @@ #include <core/memory.h> +void +gm200_fb_init_page(struct nvkm_fb *fb) +{ + struct nvkm_device *device = fb->subdev.device; + switch (fb->page) { + case 16: + nvkm_mask(device, 0x100c80, 0x00000801, 0x00000001); + break; + case 17: + nvkm_mask(device, 0x100c80, 0x00000801, 0x00000000); + break; + default: + nvkm_mask(device, 0x100c80, 0x00000800, 0x00000800); + fb->page = 0; + break; + } +} + static void gm200_fb_init(struct nvkm_fb *base) { @@ -48,6 +66,7 @@ gm200_fb = { .dtor = gf100_fb_dtor, .oneinit = gf100_fb_oneinit, .init = gm200_fb_init, + .init_page = gm200_fb_init_page, .intr = gf100_fb_intr, .ram_new = gm107_ram_new, .memtype_valid = gf100_fb_memtype_valid, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c new file mode 100644 index 000000000000..98474aec1921 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c @@ -0,0 +1,69 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "gf100.h" +#include "ram.h" + +#include <core/memory.h> + +static void +gp100_fb_init_unkn(struct nvkm_fb *base) +{ + struct nvkm_device *device = gf100_fb(base)->base.subdev.device; + nvkm_wr32(device, 0x1fac80, nvkm_rd32(device, 0x100c80)); + nvkm_wr32(device, 0x1facc4, nvkm_rd32(device, 0x100cc4)); + nvkm_wr32(device, 0x1facc8, nvkm_rd32(device, 0x100cc8)); + nvkm_wr32(device, 0x1faccc, nvkm_rd32(device, 0x100ccc)); +} + +void +gp100_fb_init(struct nvkm_fb *base) +{ + struct gf100_fb *fb = gf100_fb(base); + struct nvkm_device *device = fb->base.subdev.device; + + if (fb->r100c10_page) + nvkm_wr32(device, 0x100c10, fb->r100c10 >> 8); + + nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->base.mmu_wr) >> 8); + nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->base.mmu_rd) >> 8); + nvkm_mask(device, 0x100cc4, 0x00060000, + max(nvkm_memory_size(fb->base.mmu_rd) >> 16, (u64)2) << 17); +} + +static const struct nvkm_fb_func +gp100_fb = { + .dtor = gf100_fb_dtor, + .oneinit = gf100_fb_oneinit, + .init = gp100_fb_init, + .init_page = gm200_fb_init_page, + .init_unkn = gp100_fb_init_unkn, + .ram_new = gp100_ram_new, + .memtype_valid = gf100_fb_memtype_valid, +}; + +int +gp100_fb_new(struct nvkm_device *device, int index, struct nvkm_fb **pfb) +{ + return gf100_fb_new_(&gp100_fb, device, index, pfb); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp104.c new file mode 100644 index 000000000000..92cb71861bec --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp104.c @@ -0,0 +1,43 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "gf100.h" +#include "ram.h" + +#include <core/memory.h> + +static const struct nvkm_fb_func +gp104_fb = { + .dtor = gf100_fb_dtor, + .oneinit = gf100_fb_oneinit, + .init = gp100_fb_init, + .init_page = gm200_fb_init_page, + .ram_new = gp100_ram_new, + .memtype_valid = gf100_fb_memtype_valid, +}; + +int +gp104_fb_new(struct nvkm_device *device, int index, struct nvkm_fb **pfb) +{ + return gf100_fb_new_(&gp104_fb, device, index, pfb); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h index d97d640e60a0..e905d44fa1d5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h @@ -8,6 +8,8 @@ struct nvkm_fb_func { void *(*dtor)(struct nvkm_fb *); int (*oneinit)(struct nvkm_fb *); void (*init)(struct nvkm_fb *); + void (*init_page)(struct nvkm_fb *); + void (*init_unkn)(struct nvkm_fb *); void (*intr)(struct nvkm_fb *); struct { @@ -60,5 +62,8 @@ void nv46_fb_tile_init(struct nvkm_fb *, int i, u32 addr, u32 size, u32 pitch, u32 flags, struct nvkm_fb_tile *); int gf100_fb_oneinit(struct nvkm_fb *); +void gf100_fb_init_page(struct nvkm_fb *); bool gf100_fb_memtype_valid(struct nvkm_fb *, u32); + +void gm200_fb_init_page(struct nvkm_fb *); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ram.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ram.h index f816cbf2ced3..b9ec0ae6723a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ram.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ram.h @@ -47,4 +47,5 @@ int mcp77_ram_new(struct nvkm_fb *, struct nvkm_ram **); int gf100_ram_new(struct nvkm_fb *, struct nvkm_ram **); int gk104_ram_new(struct nvkm_fb *, struct nvkm_ram **); int gm107_ram_new(struct nvkm_fb *, struct nvkm_ram **); +int gp100_ram_new(struct nvkm_fb *, struct nvkm_ram **); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgp100.c new file mode 100644 index 000000000000..f3be408b5e5e --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgp100.c @@ -0,0 +1,146 @@ +/* + * Copyright 2013 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "ram.h" + +#include <subdev/bios.h> +#include <subdev/bios/init.h> +#include <subdev/bios/rammap.h> + +static int +gp100_ram_init(struct nvkm_ram *ram) +{ + struct nvkm_subdev *subdev = &ram->fb->subdev; + struct nvkm_device *device = subdev->device; + struct nvkm_bios *bios = device->bios; + u8 ver, hdr, cnt, len, snr, ssz; + u32 data; + int i; + + /* run a bunch of tables from rammap table. there's actually + * individual pointers for each rammap entry too, but, nvidia + * seem to just run the last two entries' scripts early on in + * their init, and never again.. we'll just run 'em all once + * for now. + * + * i strongly suspect that each script is for a separate mode + * (likely selected by 0x9a065c's lower bits?), and the + * binary driver skips the one that's already been setup by + * the init tables. + */ + data = nvbios_rammapTe(bios, &ver, &hdr, &cnt, &len, &snr, &ssz); + if (!data || hdr < 0x15) + return -EINVAL; + + cnt = nvbios_rd08(bios, data + 0x14); /* guess at count */ + data = nvbios_rd32(bios, data + 0x10); /* guess u32... */ + if (cnt) { + u32 save = nvkm_rd32(device, 0x9a065c) & 0x000000f0; + for (i = 0; i < cnt; i++, data += 4) { + if (i != save >> 4) { + nvkm_mask(device, 0x9a065c, 0x000000f0, i << 4); + nvbios_exec(&(struct nvbios_init) { + .subdev = subdev, + .bios = bios, + .offset = nvbios_rd32(bios, data), + .execute = 1, + }); + } + } + nvkm_mask(device, 0x9a065c, 0x000000f0, save); + } + + nvkm_mask(device, 0x9a0584, 0x11000000, 0x00000000); + nvkm_wr32(device, 0x10ecc0, 0xffffffff); + nvkm_mask(device, 0x9a0160, 0x00000010, 0x00000010); + return 0; +} + +static const struct nvkm_ram_func +gp100_ram_func = { + .init = gp100_ram_init, + .get = gf100_ram_get, + .put = gf100_ram_put, +}; + +int +gp100_ram_new(struct nvkm_fb *fb, struct nvkm_ram **pram) +{ + struct nvkm_ram *ram; + struct nvkm_subdev *subdev = &fb->subdev; + struct nvkm_device *device = subdev->device; + enum nvkm_ram_type type = nvkm_fb_bios_memtype(device->bios); + const u32 rsvd_head = ( 256 * 1024); /* vga memory */ + const u32 rsvd_tail = (1024 * 1024); /* vbios etc */ + u32 fbpa_num = nvkm_rd32(device, 0x022438), fbpa; + u32 fbio_opt = nvkm_rd32(device, 0x021c14); + u64 part, size = 0, comm = ~0ULL; + bool mixed = false; + int ret; + + nvkm_debug(subdev, "022438: %08x\n", fbpa_num); + nvkm_debug(subdev, "021c14: %08x\n", fbio_opt); + for (fbpa = 0; fbpa < fbpa_num; fbpa++) { + if (!(fbio_opt & (1 << fbpa))) { + part = nvkm_rd32(device, 0x90020c + (fbpa * 0x4000)); + nvkm_debug(subdev, "fbpa %02x: %lld MiB\n", fbpa, part); + part = part << 20; + if (part != comm) { + if (comm != ~0ULL) + mixed = true; + comm = min(comm, part); + } + size = size + part; + } + } + + ret = nvkm_ram_new_(&gp100_ram_func, fb, type, size, 0, &ram); + *pram = ram; + if (ret) + return ret; + + nvkm_mm_fini(&ram->vram); + + if (mixed) { + ret = nvkm_mm_init(&ram->vram, rsvd_head >> NVKM_RAM_MM_SHIFT, + ((comm * fbpa_num) - rsvd_head) >> + NVKM_RAM_MM_SHIFT, 1); + if (ret) + return ret; + + ret = nvkm_mm_init(&ram->vram, (0x1000000000ULL + comm) >> + NVKM_RAM_MM_SHIFT, + (size - (comm * fbpa_num) - rsvd_tail) >> + NVKM_RAM_MM_SHIFT, 1); + if (ret) + return ret; + } else { + ret = nvkm_mm_init(&ram->vram, rsvd_head >> NVKM_RAM_MM_SHIFT, + (size - rsvd_head - rsvd_tail) >> + NVKM_RAM_MM_SHIFT, 1); + if (ret) + return ret; + } + + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c index 323c79abe468..41bd5d0f7692 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/iccsense/base.c @@ -276,6 +276,8 @@ nvkm_iccsense_oneinit(struct nvkm_subdev *subdev) struct pwr_rail_t *r = &stbl.rail[i]; struct nvkm_iccsense_rail *rail; struct nvkm_iccsense_sensor *sensor; + int (*read)(struct nvkm_iccsense *, + struct nvkm_iccsense_rail *); if (!r->mode || r->resistor_mohm == 0) continue; @@ -284,31 +286,31 @@ nvkm_iccsense_oneinit(struct nvkm_subdev *subdev) if (!sensor) continue; - rail = kmalloc(sizeof(*rail), GFP_KERNEL); - if (!rail) - return -ENOMEM; - switch (sensor->type) { case NVBIOS_EXTDEV_INA209: if (r->rail != 0) continue; - rail->read = nvkm_iccsense_ina209_read; + read = nvkm_iccsense_ina209_read; break; case NVBIOS_EXTDEV_INA219: if (r->rail != 0) continue; - rail->read = nvkm_iccsense_ina219_read; + read = nvkm_iccsense_ina219_read; break; case NVBIOS_EXTDEV_INA3221: if (r->rail >= 3) continue; - rail->read = nvkm_iccsense_ina3221_read; + read = nvkm_iccsense_ina3221_read; break; default: continue; } + rail = kmalloc(sizeof(*rail), GFP_KERNEL); + if (!rail) + return -ENOMEM; sensor->rail_mask |= 1 << r->rail; + rail->read = read; rail->sensor = sensor; rail->idx = r->rail; rail->mohm = r->resistor_mohm; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild index 932b366598aa..12d6f4f102cb 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild @@ -3,3 +3,4 @@ nvkm-y += nvkm/subdev/ltc/gf100.o nvkm-y += nvkm/subdev/ltc/gk104.o nvkm-y += nvkm/subdev/ltc/gm107.o nvkm-y += nvkm/subdev/ltc/gm200.o +nvkm-y += nvkm/subdev/ltc/gp100.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gf100.c index c9eb677967a8..4a0fa0a9b802 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gf100.c @@ -23,7 +23,6 @@ */ #include "priv.h" -#include <core/enum.h> #include <subdev/fb.h> #include <subdev/timer.h> @@ -71,7 +70,7 @@ gf100_ltc_zbc_clear_depth(struct nvkm_ltc *ltc, int i, const u32 depth) nvkm_wr32(device, 0x17ea58, depth); } -static const struct nvkm_bitfield +const struct nvkm_bitfield gf100_ltc_lts_intr_name[] = { { 0x00000001, "IDLE_ERROR_IQ" }, { 0x00000002, "IDLE_ERROR_CBC" }, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm107.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm107.c index e292f5679418..ec0a3844b2d1 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm107.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm107.c @@ -68,18 +68,22 @@ gm107_ltc_zbc_clear_depth(struct nvkm_ltc *ltc, int i, const u32 depth) nvkm_wr32(device, 0x17e34c, depth); } -static void -gm107_ltc_lts_isr(struct nvkm_ltc *ltc, int c, int s) +void +gm107_ltc_intr_lts(struct nvkm_ltc *ltc, int c, int s) { struct nvkm_subdev *subdev = <c->subdev; struct nvkm_device *device = subdev->device; - u32 base = 0x140000 + (c * 0x2000) + (s * 0x200); - u32 stat = nvkm_rd32(device, base + 0x00c); + u32 base = 0x140400 + (c * 0x2000) + (s * 0x200); + u32 intr = nvkm_rd32(device, base + 0x00c); + u16 stat = intr & 0x0000ffff; + char msg[128]; if (stat) { - nvkm_error(subdev, "LTC%d_LTS%d: %08x\n", c, s, stat); - nvkm_wr32(device, base + 0x00c, stat); + nvkm_snprintbf(msg, sizeof(msg), gf100_ltc_lts_intr_name, stat); + nvkm_error(subdev, "LTC%d_LTS%d: %08x [%s]\n", c, s, intr, msg); } + + nvkm_wr32(device, base + 0x00c, intr); } void @@ -92,7 +96,7 @@ gm107_ltc_intr(struct nvkm_ltc *ltc) while (mask) { u32 s, c = __ffs(mask); for (s = 0; s < ltc->lts_nr; s++) - gm107_ltc_lts_isr(ltc, c, s); + gm107_ltc_intr_lts(ltc, c, s); mask &= ~(1 << c); } } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm200.c index 2a29bfd5125a..e18e0dc19ec8 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gm200.c @@ -46,7 +46,7 @@ static const struct nvkm_ltc_func gm200_ltc = { .oneinit = gm200_ltc_oneinit, .init = gm200_ltc_init, - .intr = gm107_ltc_intr, /*XXX: not validated */ + .intr = gm107_ltc_intr, .cbc_clear = gm107_ltc_cbc_clear, .cbc_wait = gm107_ltc_cbc_wait, .zbc = 16, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c new file mode 100644 index 000000000000..0bdfb2f40266 --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c @@ -0,0 +1,75 @@ +/* + * Copyright 2016 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#include "priv.h" + +static void +gp100_ltc_intr(struct nvkm_ltc *ltc) +{ + struct nvkm_device *device = ltc->subdev.device; + u32 mask; + + mask = nvkm_rd32(device, 0x0001c0); + while (mask) { + u32 s, c = __ffs(mask); + for (s = 0; s < ltc->lts_nr; s++) + gm107_ltc_intr_lts(ltc, c, s); + mask &= ~(1 << c); + } +} + +static int +gp100_ltc_oneinit(struct nvkm_ltc *ltc) +{ + struct nvkm_device *device = ltc->subdev.device; + ltc->ltc_nr = nvkm_rd32(device, 0x12006c); + ltc->lts_nr = nvkm_rd32(device, 0x17e280) >> 28; + /*XXX: tagram allocation - TBD */ + return nvkm_mm_init(<c->tags, 0, 0, 1); +} + +static void +gp100_ltc_init(struct nvkm_ltc *ltc) +{ + /*XXX: PMU LS call to setup tagram address */ +} + +static const struct nvkm_ltc_func +gp100_ltc = { + .oneinit = gp100_ltc_oneinit, + .init = gp100_ltc_init, + .intr = gp100_ltc_intr, + .cbc_clear = gm107_ltc_cbc_clear, + .cbc_wait = gm107_ltc_cbc_wait, + .zbc = 16, + .zbc_clear_color = gm107_ltc_zbc_clear_color, + .zbc_clear_depth = gm107_ltc_zbc_clear_depth, + .invalidate = gf100_ltc_invalidate, + .flush = gf100_ltc_flush, +}; + +int +gp100_ltc_new(struct nvkm_device *device, int index, struct nvkm_ltc **pltc) +{ + return nvkm_ltc_new_(&gp100_ltc, device, index, pltc); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h index 6d81c695ed0d..8b95f96e3ffa 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h @@ -2,6 +2,7 @@ #define __NVKM_LTC_PRIV_H__ #define nvkm_ltc(p) container_of((p), struct nvkm_ltc, subdev) #include <subdev/ltc.h> +#include <core/enum.h> int nvkm_ltc_new_(const struct nvkm_ltc_func *, struct nvkm_device *, int index, struct nvkm_ltc **); @@ -31,8 +32,10 @@ void gf100_ltc_zbc_clear_color(struct nvkm_ltc *, int, const u32[4]); void gf100_ltc_zbc_clear_depth(struct nvkm_ltc *, int, const u32); void gf100_ltc_invalidate(struct nvkm_ltc *); void gf100_ltc_flush(struct nvkm_ltc *); +extern const struct nvkm_bitfield gf100_ltc_lts_intr_name[]; void gm107_ltc_intr(struct nvkm_ltc *); +void gm107_ltc_intr_lts(struct nvkm_ltc *, int ltc, int lts); void gm107_ltc_cbc_clear(struct nvkm_ltc *, u32, u32); void gm107_ltc_cbc_wait(struct nvkm_ltc *); void gm107_ltc_zbc_clear_color(struct nvkm_ltc *, int, const u32[4]); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild index 49695ac7be2e..12943f92c206 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/Kbuild @@ -10,3 +10,4 @@ nvkm-y += nvkm/subdev/mc/gt215.o nvkm-y += nvkm/subdev/mc/gf100.o nvkm-y += nvkm/subdev/mc/gk104.o nvkm-y += nvkm/subdev/mc/gk20a.o +nvkm-y += nvkm/subdev/mc/gp100.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c index 350a8caa84c8..6b25e25f9eba 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/base.c @@ -27,43 +27,67 @@ #include <subdev/top.h> void -nvkm_mc_unk260(struct nvkm_mc *mc, u32 data) +nvkm_mc_unk260(struct nvkm_device *device, u32 data) { - if (mc->func->unk260) + struct nvkm_mc *mc = device->mc; + if (likely(mc) && mc->func->unk260) mc->func->unk260(mc, data); } void -nvkm_mc_intr_unarm(struct nvkm_mc *mc) +nvkm_mc_intr_mask(struct nvkm_device *device, enum nvkm_devidx devidx, bool en) { - return mc->func->intr_unarm(mc); + struct nvkm_mc *mc = device->mc; + const struct nvkm_mc_map *map; + if (likely(mc) && mc->func->intr_mask) { + u32 mask = nvkm_top_intr_mask(device, devidx); + for (map = mc->func->intr; !mask && map->stat; map++) { + if (map->unit == devidx) + mask = map->stat; + } + mc->func->intr_mask(mc, mask, en ? mask : 0); + } +} + +void +nvkm_mc_intr_unarm(struct nvkm_device *device) +{ + struct nvkm_mc *mc = device->mc; + if (likely(mc)) + mc->func->intr_unarm(mc); } void -nvkm_mc_intr_rearm(struct nvkm_mc *mc) +nvkm_mc_intr_rearm(struct nvkm_device *device) { - return mc->func->intr_rearm(mc); + struct nvkm_mc *mc = device->mc; + if (likely(mc)) + mc->func->intr_rearm(mc); } static u32 -nvkm_mc_intr_mask(struct nvkm_mc *mc) +nvkm_mc_intr_stat(struct nvkm_mc *mc) { - u32 intr = mc->func->intr_mask(mc); + u32 intr = mc->func->intr_stat(mc); if (WARN_ON_ONCE(intr == 0xffffffff)) intr = 0; /* likely fallen off the bus */ return intr; } void -nvkm_mc_intr(struct nvkm_mc *mc, bool *handled) +nvkm_mc_intr(struct nvkm_device *device, bool *handled) { - struct nvkm_device *device = mc->subdev.device; + struct nvkm_mc *mc = device->mc; struct nvkm_subdev *subdev; - const struct nvkm_mc_map *map = mc->func->intr; - u32 stat, intr = nvkm_mc_intr_mask(mc); + const struct nvkm_mc_map *map; + u32 stat, intr; u64 subdevs; - stat = nvkm_top_intr(device->top, intr, &subdevs); + if (unlikely(!mc)) + return; + + intr = nvkm_mc_intr_stat(mc); + stat = nvkm_top_intr(device, intr, &subdevs); while (subdevs) { enum nvkm_devidx subidx = __ffs64(subdevs); subdev = nvkm_device_subdev(device, subidx); @@ -72,14 +96,13 @@ nvkm_mc_intr(struct nvkm_mc *mc, bool *handled) subdevs &= ~BIT_ULL(subidx); } - while (map->stat) { + for (map = mc->func->intr; map->stat; map++) { if (intr & map->stat) { subdev = nvkm_device_subdev(device, map->unit); if (subdev) nvkm_subdev_intr(subdev); stat &= ~map->stat; } - map++; } if (stat) @@ -87,22 +110,32 @@ nvkm_mc_intr(struct nvkm_mc *mc, bool *handled) *handled = intr != 0; } -static void -nvkm_mc_reset_(struct nvkm_mc *mc, enum nvkm_devidx devidx) +static u32 +nvkm_mc_reset_mask(struct nvkm_device *device, bool isauto, + enum nvkm_devidx devidx) { - struct nvkm_device *device = mc->subdev.device; + struct nvkm_mc *mc = device->mc; const struct nvkm_mc_map *map; - u64 pmc_enable; - - if (!(pmc_enable = nvkm_top_reset(device->top, devidx))) { - for (map = mc->func->reset; map && map->stat; map++) { - if (map->unit == devidx) { - pmc_enable = map->stat; - break; + u64 pmc_enable = 0; + if (likely(mc)) { + if (!(pmc_enable = nvkm_top_reset(device, devidx))) { + for (map = mc->func->reset; map && map->stat; map++) { + if (!isauto || !map->noauto) { + if (map->unit == devidx) { + pmc_enable = map->stat; + break; + } + } } } } + return pmc_enable; +} +void +nvkm_mc_reset(struct nvkm_device *device, enum nvkm_devidx devidx) +{ + u64 pmc_enable = nvkm_mc_reset_mask(device, true, devidx); if (pmc_enable) { nvkm_mask(device, 0x000200, pmc_enable, 0x00000000); nvkm_mask(device, 0x000200, pmc_enable, pmc_enable); @@ -111,17 +144,27 @@ nvkm_mc_reset_(struct nvkm_mc *mc, enum nvkm_devidx devidx) } void -nvkm_mc_reset(struct nvkm_mc *mc, enum nvkm_devidx devidx) +nvkm_mc_disable(struct nvkm_device *device, enum nvkm_devidx devidx) { - if (likely(mc)) - nvkm_mc_reset_(mc, devidx); + u64 pmc_enable = nvkm_mc_reset_mask(device, false, devidx); + if (pmc_enable) + nvkm_mask(device, 0x000200, pmc_enable, 0x00000000); +} + +void +nvkm_mc_enable(struct nvkm_device *device, enum nvkm_devidx devidx) +{ + u64 pmc_enable = nvkm_mc_reset_mask(device, false, devidx); + if (pmc_enable) { + nvkm_mask(device, 0x000200, pmc_enable, pmc_enable); + nvkm_rd32(device, 0x000200); + } } static int nvkm_mc_fini(struct nvkm_subdev *subdev, bool suspend) { - struct nvkm_mc *mc = nvkm_mc(subdev); - nvkm_mc_intr_unarm(mc); + nvkm_mc_intr_unarm(subdev->device); return 0; } @@ -131,7 +174,7 @@ nvkm_mc_init(struct nvkm_subdev *subdev) struct nvkm_mc *mc = nvkm_mc(subdev); if (mc->func->init) mc->func->init(mc); - nvkm_mc_intr_rearm(mc); + nvkm_mc_intr_rearm(subdev->device); return 0; } @@ -148,16 +191,21 @@ nvkm_mc = { .fini = nvkm_mc_fini, }; +void +nvkm_mc_ctor(const struct nvkm_mc_func *func, struct nvkm_device *device, + int index, struct nvkm_mc *mc) +{ + nvkm_subdev_ctor(&nvkm_mc, device, index, &mc->subdev); + mc->func = func; +} + int nvkm_mc_new_(const struct nvkm_mc_func *func, struct nvkm_device *device, int index, struct nvkm_mc **pmc) { struct nvkm_mc *mc; - if (!(mc = *pmc = kzalloc(sizeof(*mc), GFP_KERNEL))) return -ENOMEM; - - nvkm_subdev_ctor(&nvkm_mc, device, index, &mc->subdev); - mc->func = func; + nvkm_mc_ctor(func, device, index, *pmc); return 0; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g84.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g84.c index 5c85b47f071d..c3d66ef5dc12 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g84.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g84.c @@ -57,7 +57,7 @@ g84_mc = { .intr = g84_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = g84_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g98.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g98.c index 0280b43cc10c..93ad4982ce5f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g98.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/g98.c @@ -57,7 +57,7 @@ g98_mc = { .intr = g98_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = g98_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gf100.c index 8397e223bd43..d2c4d6033abb 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gf100.c @@ -76,7 +76,7 @@ gf100_mc_intr_rearm(struct nvkm_mc *mc) } u32 -gf100_mc_intr_mask(struct nvkm_mc *mc) +gf100_mc_intr_stat(struct nvkm_mc *mc) { struct nvkm_device *device = mc->subdev.device; u32 intr0 = nvkm_rd32(device, 0x000100); @@ -85,6 +85,14 @@ gf100_mc_intr_mask(struct nvkm_mc *mc) } void +gf100_mc_intr_mask(struct nvkm_mc *mc, u32 mask, u32 stat) +{ + struct nvkm_device *device = mc->subdev.device; + nvkm_mask(device, 0x000640, mask, stat); + nvkm_mask(device, 0x000644, mask, stat); +} + +void gf100_mc_unk260(struct nvkm_mc *mc, u32 data) { nvkm_wr32(mc->subdev.device, 0x000260, data); @@ -97,6 +105,7 @@ gf100_mc = { .intr_unarm = gf100_mc_intr_unarm, .intr_rearm = gf100_mc_intr_rearm, .intr_mask = gf100_mc_intr_mask, + .intr_stat = gf100_mc_intr_stat, .reset = gf100_mc_reset, .unk260 = gf100_mc_unk260, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk104.c index 317464212c7d..7b8c6ecad1a5 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk104.c @@ -26,6 +26,7 @@ const struct nvkm_mc_map gk104_mc_reset[] = { { 0x00000100, NVKM_ENGINE_FIFO }, + { 0x00002000, NVKM_SUBDEV_PMU, true }, {} }; @@ -53,6 +54,7 @@ gk104_mc = { .intr_unarm = gf100_mc_intr_unarm, .intr_rearm = gf100_mc_intr_rearm, .intr_mask = gf100_mc_intr_mask, + .intr_stat = gf100_mc_intr_stat, .reset = gk104_mc_reset, .unk260 = gf100_mc_unk260, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk20a.c index 60b044f517ed..ca1bf3279dbe 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gk20a.c @@ -30,6 +30,7 @@ gk20a_mc = { .intr_unarm = gf100_mc_intr_unarm, .intr_rearm = gf100_mc_intr_rearm, .intr_mask = gf100_mc_intr_mask, + .intr_stat = gf100_mc_intr_stat, .reset = gk104_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c new file mode 100644 index 000000000000..4d22f4abd6de --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c @@ -0,0 +1,103 @@ +/* + * Copyright 2012 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs + */ +#define gp100_mc(p) container_of((p), struct gp100_mc, base) +#include "priv.h" + +struct gp100_mc { + struct nvkm_mc base; + spinlock_t lock; + bool intr; + u32 mask; +}; + +static void +gp100_mc_intr_update(struct gp100_mc *mc) +{ + struct nvkm_device *device = mc->base.subdev.device; + u32 mask = mc->intr ? mc->mask : 0, i; + for (i = 0; i < 2; i++) { + nvkm_wr32(device, 0x000180 + (i * 0x04), ~mask); + nvkm_wr32(device, 0x000160 + (i * 0x04), mask); + } +} + +static void +gp100_mc_intr_unarm(struct nvkm_mc *base) +{ + struct gp100_mc *mc = gp100_mc(base); + unsigned long flags; + spin_lock_irqsave(&mc->lock, flags); + mc->intr = false; + gp100_mc_intr_update(mc); + spin_unlock_irqrestore(&mc->lock, flags); +} + +static void +gp100_mc_intr_rearm(struct nvkm_mc *base) +{ + struct gp100_mc *mc = gp100_mc(base); + unsigned long flags; + spin_lock_irqsave(&mc->lock, flags); + mc->intr = true; + gp100_mc_intr_update(mc); + spin_unlock_irqrestore(&mc->lock, flags); +} + +static void +gp100_mc_intr_mask(struct nvkm_mc *base, u32 mask, u32 intr) +{ + struct gp100_mc *mc = gp100_mc(base); + unsigned long flags; + spin_lock_irqsave(&mc->lock, flags); + mc->mask = (mc->mask & ~mask) | intr; + gp100_mc_intr_update(mc); + spin_unlock_irqrestore(&mc->lock, flags); +} + +static const struct nvkm_mc_func +gp100_mc = { + .init = nv50_mc_init, + .intr = gk104_mc_intr, + .intr_unarm = gp100_mc_intr_unarm, + .intr_rearm = gp100_mc_intr_rearm, + .intr_mask = gp100_mc_intr_mask, + .intr_stat = gf100_mc_intr_stat, + .reset = gk104_mc_reset, +}; + +int +gp100_mc_new(struct nvkm_device *device, int index, struct nvkm_mc **pmc) +{ + struct gp100_mc *mc; + + if (!(mc = kzalloc(sizeof(*mc), GFP_KERNEL))) + return -ENOMEM; + nvkm_mc_ctor(&gp100_mc, device, index, &mc->base); + *pmc = &mc->base; + + spin_lock_init(&mc->lock); + mc->intr = false; + mc->mask = 0x7fffffff; + return 0; +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gt215.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gt215.c index aad0ba95bf18..99d50a3d956f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gt215.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gt215.c @@ -53,13 +53,20 @@ gt215_mc_intr[] = { {}, }; +static void +gt215_mc_intr_mask(struct nvkm_mc *mc, u32 mask, u32 stat) +{ + nvkm_mask(mc->subdev.device, 0x000640, mask, stat); +} + static const struct nvkm_mc_func gt215_mc = { .init = nv50_mc_init, .intr = gt215_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_mask = gt215_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = gt215_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv04.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv04.c index a062624e906b..6509defd1460 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv04.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv04.c @@ -56,7 +56,7 @@ nv04_mc_intr_rearm(struct nvkm_mc *mc) } u32 -nv04_mc_intr_mask(struct nvkm_mc *mc) +nv04_mc_intr_stat(struct nvkm_mc *mc) { return nvkm_rd32(mc->subdev.device, 0x000100); } @@ -75,7 +75,7 @@ nv04_mc = { .intr = nv04_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = nv04_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv11.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv11.c index 55f0b9166b52..9213107901e6 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv11.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv11.c @@ -39,7 +39,7 @@ nv11_mc = { .intr = nv11_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = nv04_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv17.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv17.c index c40fa67f79a5..64bf5bbf8146 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv17.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv17.c @@ -48,7 +48,7 @@ nv17_mc = { .intr = nv17_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = nv17_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv44.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv44.c index cc56271db564..65fa44a64b98 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv44.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv44.c @@ -43,7 +43,7 @@ nv44_mc = { .intr = nv17_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = nv17_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv50.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv50.c index 343b6078580d..fe93b4fd7100 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/nv50.c @@ -50,7 +50,7 @@ nv50_mc = { .intr = nv50_mc_intr, .intr_unarm = nv04_mc_intr_unarm, .intr_rearm = nv04_mc_intr_rearm, - .intr_mask = nv04_mc_intr_mask, + .intr_stat = nv04_mc_intr_stat, .reset = nv17_mc_reset, }; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h index a12038118512..4f0576a06d24 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h @@ -3,12 +3,15 @@ #define nvkm_mc(p) container_of((p), struct nvkm_mc, subdev) #include <subdev/mc.h> +void nvkm_mc_ctor(const struct nvkm_mc_func *, struct nvkm_device *, + int index, struct nvkm_mc *); int nvkm_mc_new_(const struct nvkm_mc_func *, struct nvkm_device *, int index, struct nvkm_mc **); struct nvkm_mc_map { u32 stat; u32 unit; + bool noauto; }; struct nvkm_mc_func { @@ -18,8 +21,10 @@ struct nvkm_mc_func { void (*intr_unarm)(struct nvkm_mc *); /* enable reporting of interrupts to host */ void (*intr_rearm)(struct nvkm_mc *); + /* (un)mask delivery of specific interrupts */ + void (*intr_mask)(struct nvkm_mc *, u32 mask, u32 stat); /* retrieve pending interrupt mask (NV_PMC_INTR) */ - u32 (*intr_mask)(struct nvkm_mc *); + u32 (*intr_stat)(struct nvkm_mc *); const struct nvkm_mc_map *reset; void (*unk260)(struct nvkm_mc *, u32); }; @@ -27,7 +32,7 @@ struct nvkm_mc_func { void nv04_mc_init(struct nvkm_mc *); void nv04_mc_intr_unarm(struct nvkm_mc *); void nv04_mc_intr_rearm(struct nvkm_mc *); -u32 nv04_mc_intr_mask(struct nvkm_mc *); +u32 nv04_mc_intr_stat(struct nvkm_mc *); extern const struct nvkm_mc_map nv04_mc_reset[]; extern const struct nvkm_mc_map nv17_mc_intr[]; @@ -39,7 +44,8 @@ void nv50_mc_init(struct nvkm_mc *); void gf100_mc_intr_unarm(struct nvkm_mc *); void gf100_mc_intr_rearm(struct nvkm_mc *); -u32 gf100_mc_intr_mask(struct nvkm_mc *); +void gf100_mc_intr_mask(struct nvkm_mc *, u32, u32); +u32 gf100_mc_intr_stat(struct nvkm_mc *); void gf100_mc_unk260(struct nvkm_mc *, u32); extern const struct nvkm_mc_map gk104_mc_intr[]; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild index 3c2519fdeb81..2a31b7d66a6d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/Kbuild @@ -10,3 +10,4 @@ nvkm-y += nvkm/subdev/pci/g94.o nvkm-y += nvkm/subdev/pci/gf100.o nvkm-y += nvkm/subdev/pci/gf106.o nvkm-y += nvkm/subdev/pci/gk104.o +nvkm-y += nvkm/subdev/pci/gp100.o diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/base.c index 6b0328bd7eed..eb9b278198b2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/base.c @@ -69,15 +69,13 @@ static irqreturn_t nvkm_pci_intr(int irq, void *arg) { struct nvkm_pci *pci = arg; - struct nvkm_mc *mc = pci->subdev.device->mc; + struct nvkm_device *device = pci->subdev.device; bool handled = false; - if (likely(mc)) { - nvkm_mc_intr_unarm(mc); - if (pci->msi) - pci->func->msi_rearm(pci); - nvkm_mc_intr(mc, &handled); - nvkm_mc_intr_rearm(mc); - } + nvkm_mc_intr_unarm(device); + if (pci->msi) + pci->func->msi_rearm(pci); + nvkm_mc_intr(device, &handled); + nvkm_mc_intr_rearm(device); return handled ? IRQ_HANDLED : IRQ_NONE; } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gp100.c new file mode 100644 index 000000000000..82c5234a06ff --- /dev/null +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/pci/gp100.c @@ -0,0 +1,44 @@ +/* + * Copyright 2015 Red Hat Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: Ben Skeggs <bskeggs@redhat.com> + */ +#include "priv.h" + +static void +gp100_pci_msi_rearm(struct nvkm_pci *pci) +{ + nvkm_pci_wr32(pci, 0x0704, 0x00000000); +} + +static const struct nvkm_pci_func +gp100_pci_func = { + .rd32 = nv40_pci_rd32, + .wr08 = nv40_pci_wr08, + .wr32 = nv40_pci_wr32, + .msi_rearm = gp100_pci_msi_rearm, +}; + +int +gp100_pci_new(struct nvkm_device *device, int index, struct nvkm_pci **ppci) +{ + return nvkm_pci_new_(&gp100_pci_func, device, index, ppci); +} diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c index 213fdba6cfa0..314be2192b7d 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/base.c @@ -19,8 +19,9 @@ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER * DEALINGS IN THE SOFTWARE. */ - #include "priv.h" + +#include <subdev/mc.h> #include <subdev/timer.h> static const char * @@ -70,12 +71,11 @@ nvkm_secboot_falcon_enable(struct nvkm_secboot *sb) int ret; /* enable engine */ - nvkm_mask(device, 0x200, sb->enable_mask, sb->enable_mask); - nvkm_rd32(device, 0x200); + nvkm_mc_enable(device, sb->devidx); ret = nvkm_wait_msec(device, 10, sb->base + 0x10c, 0x6, 0x0); if (ret < 0) { - nvkm_mask(device, 0x200, sb->enable_mask, 0x0); nvkm_error(&sb->subdev, "Falcon mem scrubbing timeout\n"); + nvkm_mc_disable(device, sb->devidx); return ret; } @@ -85,8 +85,7 @@ nvkm_secboot_falcon_enable(struct nvkm_secboot *sb) /* enable IRQs */ nvkm_wr32(device, sb->base + 0x010, 0xff); - nvkm_mask(device, 0x640, sb->irq_mask, sb->irq_mask); - nvkm_mask(device, 0x644, sb->irq_mask, sb->irq_mask); + nvkm_mc_intr_mask(device, sb->devidx, true); return 0; } @@ -97,14 +96,13 @@ nvkm_secboot_falcon_disable(struct nvkm_secboot *sb) struct nvkm_device *device = sb->subdev.device; /* disable IRQs and wait for any previous code to complete */ - nvkm_mask(device, 0x644, sb->irq_mask, 0x0); - nvkm_mask(device, 0x640, sb->irq_mask, 0x0); + nvkm_mc_intr_mask(device, sb->devidx, false); nvkm_wr32(device, sb->base + 0x014, 0xff); falcon_wait_idle(device, sb->base); /* disable engine */ - nvkm_mask(device, 0x200, sb->enable_mask, 0x0); + nvkm_mc_disable(device, sb->devidx); return 0; } @@ -216,14 +214,7 @@ nvkm_secboot_oneinit(struct nvkm_subdev *subdev) return ret; } - /* - * Build all blobs - the same blobs can be used to perform secure boot - * multiple times - */ - if (sb->func->prepare_blobs) - ret = sb->func->prepare_blobs(sb); - - return ret; + return 0; } static int @@ -270,9 +261,8 @@ nvkm_secboot_ctor(const struct nvkm_secboot_func *func, /* setup the performing falcon's base address and masks */ switch (func->boot_falcon) { case NVKM_SECBOOT_FALCON_PMU: + sb->devidx = NVKM_SUBDEV_PMU; sb->base = 0x10a000; - sb->irq_mask = 0x1000000; - sb->enable_mask = 0x2000; break; default: nvkm_error(&sb->subdev, "invalid secure boot falcon\n"); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c index cc100dc940ea..f1e2dc914366 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm200.c @@ -860,6 +860,8 @@ gm200_secboot_prepare_ls_blob(struct gm200_secboot *gsb) /* Write LS blob */ ret = ls_ucode_mgr_write_wpr(gsb, &mgr, gsb->ls_blob); + if (ret) + nvkm_gpuobj_del(&gsb->ls_blob); cleanup: ls_ucode_mgr_cleanup(&mgr); @@ -1023,29 +1025,34 @@ gm20x_secboot_prepare_blobs(struct gm200_secboot *gsb) int ret; /* Load and prepare the managed falcon's firmwares */ - ret = gm200_secboot_prepare_ls_blob(gsb); - if (ret) - return ret; + if (!gsb->ls_blob) { + ret = gm200_secboot_prepare_ls_blob(gsb); + if (ret) + return ret; + } /* Load the HS firmware that will load the LS firmwares */ - ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_load", - &gsb->acr_load_blob, - &gsb->acr_load_bl_desc, true); - if (ret) - return ret; + if (!gsb->acr_load_blob) { + ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_load", + &gsb->acr_load_blob, + &gsb->acr_load_bl_desc, true); + if (ret) + return ret; + } /* Load the HS firmware bootloader */ - ret = gm200_secboot_prepare_hsbl_blob(gsb); - if (ret) - return ret; + if (!gsb->hsbl_blob) { + ret = gm200_secboot_prepare_hsbl_blob(gsb); + if (ret) + return ret; + } return 0; } static int -gm200_secboot_prepare_blobs(struct nvkm_secboot *sb) +gm200_secboot_prepare_blobs(struct gm200_secboot *gsb) { - struct gm200_secboot *gsb = gm200_secboot(sb); int ret; ret = gm20x_secboot_prepare_blobs(gsb); @@ -1053,15 +1060,37 @@ gm200_secboot_prepare_blobs(struct nvkm_secboot *sb) return ret; /* dGPU only: load the HS firmware that unprotects the WPR region */ - ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_unload", - &gsb->acr_unload_blob, - &gsb->acr_unload_bl_desc, false); - if (ret) - return ret; + if (!gsb->acr_unload_blob) { + ret = gm200_secboot_prepare_hs_blob(gsb, "acr/ucode_unload", + &gsb->acr_unload_blob, + &gsb->acr_unload_bl_desc, false); + if (ret) + return ret; + } return 0; } +static int +gm200_secboot_blobs_ready(struct gm200_secboot *gsb) +{ + struct nvkm_subdev *subdev = &gsb->base.subdev; + int ret; + + /* firmware already loaded, nothing to do... */ + if (gsb->firmware_ok) + return 0; + + ret = gsb->func->prepare_blobs(gsb); + if (ret) { + nvkm_error(subdev, "failed to load secure firmware\n"); + return ret; + } + + gsb->firmware_ok = true; + + return 0; +} /* @@ -1234,6 +1263,11 @@ gm200_secboot_reset(struct nvkm_secboot *sb, enum nvkm_secboot_falcon falcon) struct gm200_secboot *gsb = gm200_secboot(sb); int ret; + /* Make sure all blobs are ready */ + ret = gm200_secboot_blobs_ready(gsb); + if (ret) + return ret; + /* * Dummy GM200 implementation: perform secure boot each time we are * called on FECS. Since only FECS and GPCCS are managed and started @@ -1373,7 +1407,6 @@ gm200_secboot = { .dtor = gm200_secboot_dtor, .init = gm200_secboot_init, .fini = gm200_secboot_fini, - .prepare_blobs = gm200_secboot_prepare_blobs, .reset = gm200_secboot_reset, .start = gm200_secboot_start, .managed_falcons = BIT(NVKM_SECBOOT_FALCON_FECS) | @@ -1415,6 +1448,7 @@ gm200_secboot_func = { .bl_desc_size = sizeof(struct gm200_flcn_bl_desc), .fixup_bl_desc = gm200_secboot_fixup_bl_desc, .fixup_hs_desc = gm200_secboot_fixup_hs_desc, + .prepare_blobs = gm200_secboot_prepare_blobs, }; int @@ -1487,3 +1521,19 @@ MODULE_FIRMWARE("nvidia/gm206/gr/sw_ctx.bin"); MODULE_FIRMWARE("nvidia/gm206/gr/sw_nonctx.bin"); MODULE_FIRMWARE("nvidia/gm206/gr/sw_bundle_init.bin"); MODULE_FIRMWARE("nvidia/gm206/gr/sw_method_init.bin"); + +MODULE_FIRMWARE("nvidia/gp100/acr/bl.bin"); +MODULE_FIRMWARE("nvidia/gp100/acr/ucode_load.bin"); +MODULE_FIRMWARE("nvidia/gp100/acr/ucode_unload.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/fecs_bl.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/fecs_inst.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/fecs_data.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/fecs_sig.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/gpccs_bl.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/gpccs_inst.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/gpccs_data.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/gpccs_sig.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/sw_ctx.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/sw_nonctx.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/sw_bundle_init.bin"); +MODULE_FIRMWARE("nvidia/gp100/gr/sw_method_init.bin"); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c index 684320484b70..d5395ebfe8d3 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c @@ -42,6 +42,32 @@ struct gm20b_flcn_bl_desc { u32 data_size; }; +static int +gm20b_secboot_prepare_blobs(struct gm200_secboot *gsb) +{ + struct nvkm_subdev *subdev = &gsb->base.subdev; + int acr_size; + int ret; + + ret = gm20x_secboot_prepare_blobs(gsb); + if (ret) + return ret; + + acr_size = gsb->acr_load_blob->size; + /* + * On Tegra the WPR region is set by the bootloader. It is illegal for + * the HS blob to be larger than this region. + */ + if (acr_size > gsb->wpr_size) { + nvkm_error(subdev, "WPR region too small for FW blob!\n"); + nvkm_error(subdev, "required: %dB\n", acr_size); + nvkm_error(subdev, "WPR size: %dB\n", gsb->wpr_size); + return -ENOSPC; + } + + return 0; +} + /** * gm20b_secboot_fixup_bl_desc - adapt BL descriptor to format used by GM20B FW * @@ -88,6 +114,7 @@ gm20b_secboot_func = { .bl_desc_size = sizeof(struct gm20b_flcn_bl_desc), .fixup_bl_desc = gm20b_secboot_fixup_bl_desc, .fixup_hs_desc = gm20b_secboot_fixup_hs_desc, + .prepare_blobs = gm20b_secboot_prepare_blobs, }; @@ -147,32 +174,6 @@ gm20b_tegra_read_wpr(struct gm200_secboot *gsb) #endif static int -gm20b_secboot_prepare_blobs(struct nvkm_secboot *sb) -{ - struct gm200_secboot *gsb = gm200_secboot(sb); - int acr_size; - int ret; - - ret = gm20x_secboot_prepare_blobs(gsb); - if (ret) - return ret; - - acr_size = gsb->acr_load_blob->size; - /* - * On Tegra the WPR region is set by the bootloader. It is illegal for - * the HS blob to be larger than this region. - */ - if (acr_size > gsb->wpr_size) { - nvkm_error(&sb->subdev, "WPR region too small for FW blob!\n"); - nvkm_error(&sb->subdev, "required: %dB\n", acr_size); - nvkm_error(&sb->subdev, "WPR size: %dB\n", gsb->wpr_size); - return -ENOSPC; - } - - return 0; -} - -static int gm20b_secboot_init(struct nvkm_secboot *sb) { struct gm200_secboot *gsb = gm200_secboot(sb); @@ -189,7 +190,6 @@ static const struct nvkm_secboot_func gm20b_secboot = { .dtor = gm200_secboot_dtor, .init = gm20b_secboot_init, - .prepare_blobs = gm20b_secboot_prepare_blobs, .reset = gm200_secboot_reset, .start = gm200_secboot_start, .managed_falcons = BIT(NVKM_SECBOOT_FALCON_FECS), diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h index f2b09dee7c5d..a9a8a0e1017e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/priv.h @@ -30,7 +30,6 @@ struct nvkm_secboot_func { int (*init)(struct nvkm_secboot *); int (*fini)(struct nvkm_secboot *, bool suspend); void *(*dtor)(struct nvkm_secboot *); - int (*prepare_blobs)(struct nvkm_secboot *); int (*reset)(struct nvkm_secboot *, enum nvkm_secboot_falcon); int (*start)(struct nvkm_secboot *, enum nvkm_secboot_falcon); @@ -147,10 +146,8 @@ struct hsflcn_acr_desc { * @inst: instance block for HS falcon * @pgd: page directory for the HS falcon * @vm: address space used by the HS falcon - * @bl_desc_size: size of the BL descriptor used by this chip. - * @fixup_bl_desc: hook that generates the proper BL descriptor format from - * the generic GM200 format into a data array of size - * bl_desc_size + * @falcon_state: current state of the managed falcons + * @firmware_ok: whether the firmware blobs have been created */ struct gm200_secboot { struct nvkm_secboot base; @@ -196,9 +193,19 @@ struct gm200_secboot { RUNNING, } falcon_state[NVKM_SECBOOT_FALCON_END]; + bool firmware_ok; }; #define gm200_secboot(sb) container_of(sb, struct gm200_secboot, base) +/** + * Contains functions we wish to abstract between GM200-like implementations + * @bl_desc_size: size of the BL descriptor used by this chip. + * @fixup_bl_desc: hook that generates the proper BL descriptor format from + * the generic GM200 format into a data array of size + * bl_desc_size + * @fixup_hs_desc: hook that twiddles the HS descriptor before it is used + * @prepare_blobs: prepares the various blobs needed for secure booting + */ struct gm200_secboot_func { /* * Size of the bootloader descriptor for this chip. A block of this @@ -214,6 +221,7 @@ struct gm200_secboot_func { * we want the HS FW to set up. */ void (*fixup_hs_desc)(struct gm200_secboot *, struct hsflcn_acr_desc *); + int (*prepare_blobs)(struct gm200_secboot *); }; int gm200_secboot_init(struct nvkm_secboot *); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c index a1b264664aad..fe063d5728e2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/top/base.c @@ -41,8 +41,9 @@ nvkm_top_device_new(struct nvkm_top *top) } u32 -nvkm_top_reset(struct nvkm_top *top, enum nvkm_devidx index) +nvkm_top_reset(struct nvkm_device *device, enum nvkm_devidx index) { + struct nvkm_top *top = device->top; struct nvkm_top_device *info; if (top) { @@ -56,8 +57,25 @@ nvkm_top_reset(struct nvkm_top *top, enum nvkm_devidx index) } u32 -nvkm_top_intr(struct nvkm_top *top, u32 intr, u64 *psubdevs) +nvkm_top_intr_mask(struct nvkm_device *device, enum nvkm_devidx devidx) { + struct nvkm_top *top = device->top; + struct nvkm_top_device *info; + + if (top) { + list_for_each_entry(info, &top->device, head) { + if (info->index == devidx && info->intr >= 0) + return BIT(info->intr); + } + } + + return 0; +} + +u32 +nvkm_top_intr(struct nvkm_device *device, u32 intr, u64 *psubdevs) +{ + struct nvkm_top *top = device->top; struct nvkm_top_device *info; u64 subdevs = 0; u32 handled = 0; @@ -78,8 +96,9 @@ nvkm_top_intr(struct nvkm_top *top, u32 intr, u64 *psubdevs) } enum nvkm_devidx -nvkm_top_fault(struct nvkm_top *top, int fault) +nvkm_top_fault(struct nvkm_device *device, int fault) { + struct nvkm_top *top = device->top; struct nvkm_top_device *info; list_for_each_entry(info, &top->device, head) { @@ -91,8 +110,9 @@ nvkm_top_fault(struct nvkm_top *top, int fault) } enum nvkm_devidx -nvkm_top_engine(struct nvkm_top *top, int index, int *runl, int *engn) +nvkm_top_engine(struct nvkm_device *device, int index, int *runl, int *engn) { + struct nvkm_top *top = device->top; struct nvkm_top_device *info; int n = 0; diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c index e06acc340e99..efac3402f9dd 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c @@ -29,7 +29,7 @@ gk104_top_oneinit(struct nvkm_top *top) struct nvkm_subdev *subdev = &top->subdev; struct nvkm_device *device = subdev->device; struct nvkm_top_device *info = NULL; - u32 data, type; + u32 data, type, inst; int i; for (i = 0; i < 64; i++) { @@ -37,6 +37,7 @@ gk104_top_oneinit(struct nvkm_top *top) if (!(info = nvkm_top_device_new(top))) return -ENOMEM; type = ~0; + inst = 0; } data = nvkm_rd32(device, 0x022700 + (i * 0x04)); @@ -45,6 +46,7 @@ gk104_top_oneinit(struct nvkm_top *top) case 0x00000000: /* NOT_VALID */ continue; case 0x00000001: /* DATA */ + inst = (data & 0x3c000000) >> 26; info->addr = (data & 0x00fff000); info->fault = (data & 0x000000f8) >> 3; break; @@ -67,27 +69,32 @@ gk104_top_oneinit(struct nvkm_top *top) continue; /* Translate engine type to NVKM engine identifier. */ +#define A_(A) if (inst == 0) info->index = NVKM_ENGINE_##A +#define B_(A) if (inst + NVKM_ENGINE_##A##0 < NVKM_ENGINE_##A##_LAST + 1) \ + info->index = NVKM_ENGINE_##A##0 + inst switch (type) { - case 0x00000000: info->index = NVKM_ENGINE_GR; break; - case 0x00000001: info->index = NVKM_ENGINE_CE0; break; - case 0x00000002: info->index = NVKM_ENGINE_CE1; break; - case 0x00000003: info->index = NVKM_ENGINE_CE2; break; - case 0x00000008: info->index = NVKM_ENGINE_MSPDEC; break; - case 0x00000009: info->index = NVKM_ENGINE_MSPPP; break; - case 0x0000000a: info->index = NVKM_ENGINE_MSVLD; break; - case 0x0000000b: info->index = NVKM_ENGINE_MSENC; break; - case 0x0000000c: info->index = NVKM_ENGINE_VIC; break; - case 0x0000000d: info->index = NVKM_ENGINE_SEC; break; - case 0x0000000e: info->index = NVKM_ENGINE_NVENC0; break; - case 0x0000000f: info->index = NVKM_ENGINE_NVENC1; break; - case 0x00000010: info->index = NVKM_ENGINE_NVDEC; break; + case 0x00000000: A_(GR ); break; + case 0x00000001: A_(CE0 ); break; + case 0x00000002: A_(CE1 ); break; + case 0x00000003: A_(CE2 ); break; + case 0x00000008: A_(MSPDEC); break; + case 0x00000009: A_(MSPPP ); break; + case 0x0000000a: A_(MSVLD ); break; + case 0x0000000b: A_(MSENC ); break; + case 0x0000000c: A_(VIC ); break; + case 0x0000000d: A_(SEC ); break; + case 0x0000000e: B_(NVENC ); break; + case 0x0000000f: A_(NVENC1); break; + case 0x00000010: A_(NVDEC ); break; + case 0x00000013: B_(CE ); break; break; default: break; } - nvkm_debug(subdev, "%02x (%8s): addr %06x fault %2d engine %2d " - "runlist %2d intr %2d reset %2d\n", type, + nvkm_debug(subdev, "%02x.%d (%8s): addr %06x fault %2d " + "engine %2d runlist %2d intr %2d " + "reset %2d\n", type, inst, info->index == NVKM_SUBDEV_NR ? NULL : nvkm_subdev_name[info->index], info->addr, info->fault, info->engine, info->runlist, diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/base.c index 6b2d7531a7ff..1c3d23b0e84a 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/base.c @@ -120,6 +120,8 @@ nvkm_volt_parse_bios(struct nvkm_bios *bios, struct nvkm_volt *volt) data = nvbios_volt_parse(bios, &ver, &hdr, &cnt, &len, &info); if (data && info.vidmask && info.base && info.step) { + volt->min_uv = info.min; + volt->max_uv = info.max; for (i = 0; i < info.vidmask + 1; i++) { if (info.base >= info.min && info.base <= info.max) { @@ -131,6 +133,8 @@ nvkm_volt_parse_bios(struct nvkm_bios *bios, struct nvkm_volt *volt) } volt->vid_mask = info.vidmask; } else if (data && info.vidmask) { + volt->min_uv = 0xffffffff; + volt->max_uv = 0; for (i = 0; i < cnt; i++) { data = nvbios_volt_entry_parse(bios, i, &ver, &hdr, &ivid); @@ -138,9 +142,14 @@ nvkm_volt_parse_bios(struct nvkm_bios *bios, struct nvkm_volt *volt) volt->vid[volt->vid_nr].uv = ivid.voltage; volt->vid[volt->vid_nr].vid = ivid.vid; volt->vid_nr++; + volt->min_uv = min(volt->min_uv, ivid.voltage); + volt->max_uv = max(volt->max_uv, ivid.voltage); } } volt->vid_mask = info.vidmask; + } else if (data && info.type == NVBIOS_VOLT_PWM) { + volt->min_uv = info.base; + volt->max_uv = info.base + info.pwm_range; } } @@ -181,8 +190,11 @@ nvkm_volt_ctor(const struct nvkm_volt_func *func, struct nvkm_device *device, volt->func = func; /* Assuming the non-bios device should build the voltage table later */ - if (bios) + if (bios) { nvkm_volt_parse_bios(bios, volt); + nvkm_debug(&volt->subdev, "min: %iuv max: %iuv\n", + volt->min_uv, volt->max_uv); + } if (volt->vid_nr) { for (i = 0; i < volt->vid_nr; i++) { diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.c index d554455326da..ce5d83cdc7cf 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.c @@ -77,18 +77,19 @@ gk20a_volt_get_cvb_t_voltage(int speedo, int temp, int s_scale, int t_scale, return mv; } -int +static int gk20a_volt_calc_voltage(const struct cvb_coef *coef, int speedo) { + static const int v_scale = 1000; int mv; mv = gk20a_volt_get_cvb_t_voltage(speedo, -10, 100, 10, coef); - mv = DIV_ROUND_UP(mv, 1000); + mv = DIV_ROUND_UP(mv, v_scale); return mv * 1000; } -int +static int gk20a_volt_vid_get(struct nvkm_volt *base) { struct gk20a_volt *volt = gk20a_volt(base); @@ -103,7 +104,7 @@ gk20a_volt_vid_get(struct nvkm_volt *base) return -EINVAL; } -int +static int gk20a_volt_vid_set(struct nvkm_volt *base, u8 vid) { struct gk20a_volt *volt = gk20a_volt(base); @@ -113,7 +114,7 @@ gk20a_volt_vid_set(struct nvkm_volt *base, u8 vid) return regulator_set_voltage(volt->vdd, volt->base.vid[vid].uv, 1200000); } -int +static int gk20a_volt_set_id(struct nvkm_volt *base, u8 id, int condition) { struct gk20a_volt *volt = gk20a_volt(base); @@ -143,9 +144,9 @@ gk20a_volt = { }; int -_gk20a_volt_ctor(struct nvkm_device *device, int index, - const struct cvb_coef *coefs, int nb_coefs, - struct gk20a_volt *volt) +gk20a_volt_ctor(struct nvkm_device *device, int index, + const struct cvb_coef *coefs, int nb_coefs, + int vmin, struct gk20a_volt *volt) { struct nvkm_device_tegra *tdev = device->func->tegra(device); int i, uv; @@ -160,9 +161,9 @@ _gk20a_volt_ctor(struct nvkm_device *device, int index, volt->base.vid_nr = nb_coefs; for (i = 0; i < volt->base.vid_nr; i++) { volt->base.vid[i].vid = i; - volt->base.vid[i].uv = - gk20a_volt_calc_voltage(&coefs[i], - tdev->gpu_speedo); + volt->base.vid[i].uv = max( + gk20a_volt_calc_voltage(&coefs[i], tdev->gpu_speedo), + vmin); nvkm_debug(&volt->base.subdev, "%2d: vid=%d, uv=%d\n", i, volt->base.vid[i].vid, volt->base.vid[i].uv); } @@ -180,6 +181,6 @@ gk20a_volt_new(struct nvkm_device *device, int index, struct nvkm_volt **pvolt) return -ENOMEM; *pvolt = &volt->base; - return _gk20a_volt_ctor(device, index, gk20a_cvb_coef, - ARRAY_SIZE(gk20a_cvb_coef), volt); + return gk20a_volt_ctor(device, index, gk20a_cvb_coef, + ARRAY_SIZE(gk20a_cvb_coef), 0, volt); } diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.h b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.h index 0fa3b502bcf8..6a6c97f9684e 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.h +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gk20a.h @@ -37,13 +37,8 @@ struct gk20a_volt { struct regulator *vdd; }; -int _gk20a_volt_ctor(struct nvkm_device *device, int index, - const struct cvb_coef *coefs, int nb_coefs, - struct gk20a_volt *volt); - -int gk20a_volt_calc_voltage(const struct cvb_coef *coef, int speedo); -int gk20a_volt_vid_get(struct nvkm_volt *volt); -int gk20a_volt_vid_set(struct nvkm_volt *volt, u8 vid); -int gk20a_volt_set_id(struct nvkm_volt *volt, u8 id, int condition); +int gk20a_volt_ctor(struct nvkm_device *device, int index, + const struct cvb_coef *coefs, int nb_coefs, + int vmin, struct gk20a_volt *volt); #endif diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gm20b.c index 49b5ecb701e4..74db4d28930f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gm20b.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/volt/gm20b.c @@ -41,16 +41,52 @@ const struct cvb_coef gm20b_cvb_coef[] = { /* 921600 */ { 2647676, -106455, 1632 }, }; +static const struct cvb_coef gm20b_na_cvb_coef[] = { + /* KHz, c0, c1, c2, c3, c4, c5 */ + /* 76800 */ { 814294, 8144, -940, 808, -21583, 226 }, + /* 153600 */ { 856185, 8144, -940, 808, -21583, 226 }, + /* 230400 */ { 898077, 8144, -940, 808, -21583, 226 }, + /* 307200 */ { 939968, 8144, -940, 808, -21583, 226 }, + /* 384000 */ { 981860, 8144, -940, 808, -21583, 226 }, + /* 460800 */ { 1023751, 8144, -940, 808, -21583, 226 }, + /* 537600 */ { 1065642, 8144, -940, 808, -21583, 226 }, + /* 614400 */ { 1107534, 8144, -940, 808, -21583, 226 }, + /* 691200 */ { 1149425, 8144, -940, 808, -21583, 226 }, + /* 768000 */ { 1191317, 8144, -940, 808, -21583, 226 }, + /* 844800 */ { 1233208, 8144, -940, 808, -21583, 226 }, + /* 921600 */ { 1275100, 8144, -940, 808, -21583, 226 }, + /* 998400 */ { 1316991, 8144, -940, 808, -21583, 226 }, +}; + +const u32 speedo_to_vmin[] = { + /* 0, 1, 2, 3, 4, */ + 950000, 840000, 818750, 840000, 810000, +}; + int gm20b_volt_new(struct nvkm_device *device, int index, struct nvkm_volt **pvolt) { + struct nvkm_device_tegra *tdev = device->func->tegra(device); struct gk20a_volt *volt; + u32 vmin; + + if (tdev->gpu_speedo_id >= ARRAY_SIZE(speedo_to_vmin)) { + nvdev_error(device, "unsupported speedo %d\n", + tdev->gpu_speedo_id); + return -EINVAL; + } volt = kzalloc(sizeof(*volt), GFP_KERNEL); if (!volt) return -ENOMEM; *pvolt = &volt->base; - return _gk20a_volt_ctor(device, index, gm20b_cvb_coef, - ARRAY_SIZE(gm20b_cvb_coef), volt); + vmin = speedo_to_vmin[tdev->gpu_speedo_id]; + + if (tdev->gpu_speedo_id >= 1) + return gk20a_volt_ctor(device, index, gm20b_na_cvb_coef, + ARRAY_SIZE(gm20b_na_cvb_coef), vmin, volt); + else + return gk20a_volt_ctor(device, index, gm20b_cvb_coef, + ARRAY_SIZE(gm20b_cvb_coef), vmin, volt); } diff --git a/drivers/gpu/drm/omapdrm/dss/dsi.c b/drivers/gpu/drm/omapdrm/dss/dsi.c index 6f45e9d00b41..e1be5e795cd8 100644 --- a/drivers/gpu/drm/omapdrm/dss/dsi.c +++ b/drivers/gpu/drm/omapdrm/dss/dsi.c @@ -1167,7 +1167,6 @@ static int dsi_regulator_init(struct platform_device *dsidev) { struct dsi_data *dsi = dsi_get_dsidrv_data(dsidev); struct regulator *vdds_dsi; - int r; if (dsi->vdds_dsi_reg != NULL) return 0; diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi5.c b/drivers/gpu/drm/omapdrm/dss/hdmi5.c index 061f9bab4c9b..0c0a5139a301 100644 --- a/drivers/gpu/drm/omapdrm/dss/hdmi5.c +++ b/drivers/gpu/drm/omapdrm/dss/hdmi5.c @@ -120,7 +120,6 @@ static irqreturn_t hdmi_irq_handler(int irq, void *data) static int hdmi_init_regulator(void) { - int r; struct regulator *reg; if (hdmi.vdda_reg != NULL) diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c index 0738d74c8d04..d50c9679e631 100644 --- a/drivers/gpu/drm/qxl/qxl_ttm.c +++ b/drivers/gpu/drm/qxl/qxl_ttm.c @@ -350,11 +350,19 @@ static int qxl_bo_move(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem) { struct ttm_mem_reg *old_mem = &bo->mem; + int ret; + + ret = ttm_bo_wait(bo, interruptible, no_wait_gpu); + if (ret) + return ret; + + if (old_mem->mem_type == TTM_PL_SYSTEM && bo->ttm == NULL) { qxl_move_null(bo, new_mem); return 0; } - return ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); + return ttm_bo_move_memcpy(bo, evict, interruptible, + no_wait_gpu, new_mem); } static void qxl_bo_move_notify(struct ttm_buffer_object *bo, diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index e91763d5d800..a97abc8af657 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -589,7 +589,8 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if (ASIC_IS_DCE41(rdev) || ASIC_IS_DCE61(rdev) || ASIC_IS_DCE8(rdev)) radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV; /* use frac fb div on RS780/RS880 */ - if ((rdev->family == CHIP_RS780) || (rdev->family == CHIP_RS880)) + if (((rdev->family == CHIP_RS780) || (rdev->family == CHIP_RS880)) + && !radeon_crtc->ss_enabled) radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV; if (ASIC_IS_DCE32(rdev) && mode->clock > 165000) radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV; @@ -626,7 +627,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if (radeon_crtc->ss.refdiv) { radeon_crtc->pll_flags |= RADEON_PLL_USE_REF_DIV; radeon_crtc->pll_reference_div = radeon_crtc->ss.refdiv; - if (ASIC_IS_AVIVO(rdev)) + if (rdev->family >= CHIP_RV770) radeon_crtc->pll_flags |= RADEON_PLL_USE_FRAC_FB_DIV; } } diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c index ba192a35c607..0c1b9ff433af 100644 --- a/drivers/gpu/drm/radeon/cik.c +++ b/drivers/gpu/drm/radeon/cik.c @@ -53,6 +53,7 @@ MODULE_FIRMWARE("radeon/bonaire_mc.bin"); MODULE_FIRMWARE("radeon/bonaire_rlc.bin"); MODULE_FIRMWARE("radeon/bonaire_sdma.bin"); MODULE_FIRMWARE("radeon/bonaire_smc.bin"); +MODULE_FIRMWARE("radeon/bonaire_k_smc.bin"); MODULE_FIRMWARE("radeon/HAWAII_pfp.bin"); MODULE_FIRMWARE("radeon/HAWAII_me.bin"); @@ -72,6 +73,7 @@ MODULE_FIRMWARE("radeon/hawaii_mc.bin"); MODULE_FIRMWARE("radeon/hawaii_rlc.bin"); MODULE_FIRMWARE("radeon/hawaii_sdma.bin"); MODULE_FIRMWARE("radeon/hawaii_smc.bin"); +MODULE_FIRMWARE("radeon/hawaii_k_smc.bin"); MODULE_FIRMWARE("radeon/KAVERI_pfp.bin"); MODULE_FIRMWARE("radeon/KAVERI_me.bin"); @@ -1990,12 +1992,17 @@ static int cik_init_microcode(struct radeon_device *rdev) int new_fw = 0; int err; int num_fw; + bool new_smc = false; DRM_DEBUG("\n"); switch (rdev->family) { case CHIP_BONAIRE: chip_name = "BONAIRE"; + if ((rdev->pdev->revision == 0x80) || + (rdev->pdev->revision == 0x81) || + (rdev->pdev->device == 0x665f)) + new_smc = true; new_chip_name = "bonaire"; pfp_req_size = CIK_PFP_UCODE_SIZE * 4; me_req_size = CIK_ME_UCODE_SIZE * 4; @@ -2010,6 +2017,8 @@ static int cik_init_microcode(struct radeon_device *rdev) break; case CHIP_HAWAII: chip_name = "HAWAII"; + if (rdev->pdev->revision == 0x80) + new_smc = true; new_chip_name = "hawaii"; pfp_req_size = CIK_PFP_UCODE_SIZE * 4; me_req_size = CIK_ME_UCODE_SIZE * 4; @@ -2259,7 +2268,10 @@ static int cik_init_microcode(struct radeon_device *rdev) } } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name); + if (new_smc) + snprintf(fw_name, sizeof(fw_name), "radeon/%s_k_smc.bin", new_chip_name); + else + snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name); err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev); if (err) { snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name); @@ -8354,7 +8366,8 @@ static int cik_startup(struct radeon_device *rdev) } } rdev->rlc.cs_data = ci_cs_data; - rdev->rlc.cp_table_size = CP_ME_TABLE_SIZE * 5 * 4; + rdev->rlc.cp_table_size = ALIGN(CP_ME_TABLE_SIZE * 5 * 4, 2048); /* CP JT */ + rdev->rlc.cp_table_size += 64 * 1024; /* GDS */ r = sumo_rlc_init(rdev); if (r) { DRM_ERROR("Failed to init rlc BOs!\n"); diff --git a/drivers/gpu/drm/radeon/evergreen_cs.c b/drivers/gpu/drm/radeon/evergreen_cs.c index 0d3f744de35a..d960d3915408 100644 --- a/drivers/gpu/drm/radeon/evergreen_cs.c +++ b/drivers/gpu/drm/radeon/evergreen_cs.c @@ -2209,6 +2209,12 @@ static int evergreen_packet3_check(struct radeon_cs_parser *p, } break; } + case PACKET3_PFP_SYNC_ME: + if (pkt->count) { + DRM_ERROR("bad PFP_SYNC_ME\n"); + return -EINVAL; + } + break; case PACKET3_SURFACE_SYNC: if (pkt->count != 3) { DRM_ERROR("bad SURFACE_SYNC\n"); @@ -3381,6 +3387,7 @@ static int evergreen_vm_packet3_check(struct radeon_device *rdev, case PACKET3_MPEG_INDEX: case PACKET3_WAIT_REG_MEM: case PACKET3_MEM_WRITE: + case PACKET3_PFP_SYNC_ME: case PACKET3_SURFACE_SYNC: case PACKET3_EVENT_WRITE: case PACKET3_EVENT_WRITE_EOP: diff --git a/drivers/gpu/drm/radeon/evergreend.h b/drivers/gpu/drm/radeon/evergreend.h index 0b174e14e9a6..c8e3d394cde7 100644 --- a/drivers/gpu/drm/radeon/evergreend.h +++ b/drivers/gpu/drm/radeon/evergreend.h @@ -1624,6 +1624,7 @@ */ # define PACKET3_CP_DMA_CMD_SAIC (1 << 28) # define PACKET3_CP_DMA_CMD_DAIC (1 << 29) +#define PACKET3_PFP_SYNC_ME 0x42 #define PACKET3_SURFACE_SYNC 0x43 # define PACKET3_CB0_DEST_BASE_ENA (1 << 6) # define PACKET3_CB1_DEST_BASE_ENA (1 << 7) diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c index 95f4fea89302..86dcdf38b732 100644 --- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c +++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c @@ -10,6 +10,7 @@ #include <linux/slab.h> #include <linux/acpi.h> #include <linux/pci.h> +#include <linux/delay.h> #include "radeon_acpi.h" @@ -27,6 +28,7 @@ struct radeon_atpx_functions { struct radeon_atpx { acpi_handle handle; struct radeon_atpx_functions functions; + bool is_hybrid; }; static struct radeon_atpx_priv { @@ -62,6 +64,14 @@ bool radeon_has_atpx(void) { return radeon_atpx_priv.atpx_detected; } +bool radeon_has_atpx_dgpu_power_cntl(void) { + return radeon_atpx_priv.atpx.functions.power_cntl; +} + +bool radeon_is_atpx_hybrid(void) { + return radeon_atpx_priv.atpx.is_hybrid; +} + /** * radeon_atpx_call - call an ATPX method * @@ -141,18 +151,12 @@ static void radeon_atpx_parse_functions(struct radeon_atpx_functions *f, u32 mas */ static int radeon_atpx_validate(struct radeon_atpx *atpx) { - /* make sure required functions are enabled */ - /* dGPU power control is required */ - if (atpx->functions.power_cntl == false) { - printk("ATPX dGPU power cntl not present, forcing\n"); - atpx->functions.power_cntl = true; - } + u32 valid_bits = 0; if (atpx->functions.px_params) { union acpi_object *info; struct atpx_px_params output; size_t size; - u32 valid_bits; info = radeon_atpx_call(atpx->handle, ATPX_FUNCTION_GET_PX_PARAMETERS, NULL); if (!info) @@ -171,19 +175,42 @@ static int radeon_atpx_validate(struct radeon_atpx *atpx) memcpy(&output, info->buffer.pointer, size); valid_bits = output.flags & output.valid_flags; - /* if separate mux flag is set, mux controls are required */ - if (valid_bits & ATPX_SEPARATE_MUX_FOR_I2C) { - atpx->functions.i2c_mux_cntl = true; - atpx->functions.disp_mux_cntl = true; - } - /* if any outputs are muxed, mux controls are required */ - if (valid_bits & (ATPX_CRT1_RGB_SIGNAL_MUXED | - ATPX_TV_SIGNAL_MUXED | - ATPX_DFP_SIGNAL_MUXED)) - atpx->functions.disp_mux_cntl = true; kfree(info); } + + /* if separate mux flag is set, mux controls are required */ + if (valid_bits & ATPX_SEPARATE_MUX_FOR_I2C) { + atpx->functions.i2c_mux_cntl = true; + atpx->functions.disp_mux_cntl = true; + } + /* if any outputs are muxed, mux controls are required */ + if (valid_bits & (ATPX_CRT1_RGB_SIGNAL_MUXED | + ATPX_TV_SIGNAL_MUXED | + ATPX_DFP_SIGNAL_MUXED)) + atpx->functions.disp_mux_cntl = true; + + /* some bioses set these bits rather than flagging power_cntl as supported */ + if (valid_bits & (ATPX_DYNAMIC_PX_SUPPORTED | + ATPX_DYNAMIC_DGPU_POWER_OFF_SUPPORTED)) + atpx->functions.power_cntl = true; + + atpx->is_hybrid = false; + if (valid_bits & ATPX_MS_HYBRID_GFX_SUPPORTED) { + printk("ATPX Hybrid Graphics\n"); +#if 1 + /* This is a temporary hack until the D3 cold support + * makes it upstream. The ATPX power_control method seems + * to still work on even if the system should be using + * the new standardized hybrid D3 cold ACPI interface. + */ + atpx->functions.power_cntl = true; +#else + atpx->functions.power_cntl = false; +#endif + atpx->is_hybrid = true; + } + return 0; } @@ -258,6 +285,10 @@ static int radeon_atpx_set_discrete_state(struct radeon_atpx *atpx, u8 state) if (!info) return -EIO; kfree(info); + + /* 200ms delay is required after off */ + if (state == 0) + msleep(200); } return 0; } diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 81a63d7f5cd9..b79f3b002471 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -2064,7 +2064,6 @@ radeon_add_atom_connector(struct drm_device *dev, RADEON_OUTPUT_CSC_BYPASS); /* no HPD on analog connectors */ radeon_connector->hpd.hpd = RADEON_HPD_NONE; - connector->polled = DRM_CONNECTOR_POLL_CONNECT; connector->interlace_allowed = true; connector->doublescan_allowed = true; break; @@ -2314,8 +2313,10 @@ radeon_add_atom_connector(struct drm_device *dev, } if (radeon_connector->hpd.hpd == RADEON_HPD_NONE) { - if (i2c_bus->valid) - connector->polled = DRM_CONNECTOR_POLL_CONNECT; + if (i2c_bus->valid) { + connector->polled = DRM_CONNECTOR_POLL_CONNECT | + DRM_CONNECTOR_POLL_DISCONNECT; + } } else connector->polled = DRM_CONNECTOR_POLL_HPD; @@ -2391,7 +2392,6 @@ radeon_add_legacy_connector(struct drm_device *dev, 1); /* no HPD on analog connectors */ radeon_connector->hpd.hpd = RADEON_HPD_NONE; - connector->polled = DRM_CONNECTOR_POLL_CONNECT; connector->interlace_allowed = true; connector->doublescan_allowed = true; break; @@ -2476,10 +2476,13 @@ radeon_add_legacy_connector(struct drm_device *dev, } if (radeon_connector->hpd.hpd == RADEON_HPD_NONE) { - if (i2c_bus->valid) - connector->polled = DRM_CONNECTOR_POLL_CONNECT; + if (i2c_bus->valid) { + connector->polled = DRM_CONNECTOR_POLL_CONNECT | + DRM_CONNECTOR_POLL_DISCONNECT; + } } else connector->polled = DRM_CONNECTOR_POLL_HPD; + connector->display_info.subpixel_order = subpixel_order; drm_connector_register(connector); } diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index e0bf77843d79..a00dd2f74527 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -631,6 +631,23 @@ void radeon_gtt_location(struct radeon_device *rdev, struct radeon_mc *mc) /* * GPU helpers function. */ + +/** + * radeon_device_is_virtual - check if we are running is a virtual environment + * + * Check if the asic has been passed through to a VM (all asics). + * Used at driver startup. + * Returns true if virtual or false if not. + */ +static bool radeon_device_is_virtual(void) +{ +#ifdef CONFIG_X86 + return boot_cpu_has(X86_FEATURE_HYPERVISOR); +#else + return false; +#endif +} + /** * radeon_card_posted - check if the hw has already been initialized * @@ -644,6 +661,10 @@ bool radeon_card_posted(struct radeon_device *rdev) { uint32_t reg; + /* for pass through, always force asic_init */ + if (radeon_device_is_virtual()) + return false; + /* required for EFI mode on macbook2,1 which uses an r5xx asic */ if (efi_enabled(EFI_BOOT) && (rdev->pdev->subsystem_vendor == PCI_VENDOR_ID_APPLE) && @@ -1635,7 +1656,7 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend, radeon_agp_suspend(rdev); pci_save_state(dev->pdev); - if (freeze && rdev->family >= CHIP_R600) { + if (freeze && rdev->family >= CHIP_CEDAR) { rdev->asic->asic_reset(rdev, true); pci_restore_state(dev->pdev); } else if (suspend) { diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index a455dc7d4aa1..c01a7c6abb49 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -93,9 +93,10 @@ * 2.43.0 - RADEON_INFO_GPU_RESET_COUNTER * 2.44.0 - SET_APPEND_CNT packet3 support * 2.45.0 - Allow setting shader registers using DMA/COPY packet3 on SI + * 2.46.0 - Add PFP_SYNC_ME support on evergreen */ #define KMS_DRIVER_MAJOR 2 -#define KMS_DRIVER_MINOR 45 +#define KMS_DRIVER_MINOR 46 #define KMS_DRIVER_PATCHLEVEL 0 int radeon_driver_load_kms(struct drm_device *dev, unsigned long flags); int radeon_driver_unload_kms(struct drm_device *dev); @@ -162,9 +163,13 @@ void radeon_debugfs_cleanup(struct drm_minor *minor); #if defined(CONFIG_VGA_SWITCHEROO) void radeon_register_atpx_handler(void); void radeon_unregister_atpx_handler(void); +bool radeon_has_atpx_dgpu_power_cntl(void); +bool radeon_is_atpx_hybrid(void); #else static inline void radeon_register_atpx_handler(void) {} static inline void radeon_unregister_atpx_handler(void) {} +static inline bool radeon_has_atpx_dgpu_power_cntl(void) { return false; } +static inline bool radeon_is_atpx_hybrid(void) { return false; } #endif int radeon_no_wb; @@ -404,7 +409,10 @@ static int radeon_pmops_runtime_suspend(struct device *dev) pci_save_state(pdev); pci_disable_device(pdev); pci_ignore_hotplug(pdev); - pci_set_power_state(pdev, PCI_D3cold); + if (radeon_is_atpx_hybrid()) + pci_set_power_state(pdev, PCI_D3cold); + else if (!radeon_has_atpx_dgpu_power_cntl()) + pci_set_power_state(pdev, PCI_D3hot); drm_dev->switch_power_state = DRM_SWITCH_POWER_DYNAMIC_OFF; return 0; @@ -421,7 +429,9 @@ static int radeon_pmops_runtime_resume(struct device *dev) drm_dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; - pci_set_power_state(pdev, PCI_D0); + if (radeon_is_atpx_hybrid() || + !radeon_has_atpx_dgpu_power_cntl()) + pci_set_power_state(pdev, PCI_D0); pci_restore_state(pdev); ret = pci_enable_device(pdev); if (ret) diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c index 590b0377fbe2..ffdad81ef964 100644 --- a/drivers/gpu/drm/radeon/radeon_ttm.c +++ b/drivers/gpu/drm/radeon/radeon_ttm.c @@ -300,8 +300,7 @@ static int radeon_move_blit(struct ttm_buffer_object *bo, if (IS_ERR(fence)) return PTR_ERR(fence); - r = ttm_bo_move_accel_cleanup(bo, &fence->base, - evict, no_wait_gpu, new_mem); + r = ttm_bo_move_accel_cleanup(bo, &fence->base, evict, new_mem); radeon_fence_unref(&fence); return r; } @@ -403,6 +402,10 @@ static int radeon_bo_move(struct ttm_buffer_object *bo, struct ttm_mem_reg *old_mem = &bo->mem; int r; + r = ttm_bo_wait(bo, interruptible, no_wait_gpu); + if (r) + return r; + /* Can't move a pinned BO */ rbo = container_of(bo, struct radeon_bo, tbo); if (WARN_ON_ONCE(rbo->pin_count > 0)) @@ -441,7 +444,8 @@ static int radeon_bo_move(struct ttm_buffer_object *bo, if (r) { memcpy: - r = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, new_mem); + r = ttm_bo_move_memcpy(bo, evict, interruptible, + no_wait_gpu, new_mem); if (r) { return r; } diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c index b30e719dd56d..2523ca96c6c7 100644 --- a/drivers/gpu/drm/radeon/si.c +++ b/drivers/gpu/drm/radeon/si.c @@ -50,6 +50,7 @@ MODULE_FIRMWARE("radeon/tahiti_ce.bin"); MODULE_FIRMWARE("radeon/tahiti_mc.bin"); MODULE_FIRMWARE("radeon/tahiti_rlc.bin"); MODULE_FIRMWARE("radeon/tahiti_smc.bin"); +MODULE_FIRMWARE("radeon/tahiti_k_smc.bin"); MODULE_FIRMWARE("radeon/PITCAIRN_pfp.bin"); MODULE_FIRMWARE("radeon/PITCAIRN_me.bin"); @@ -65,6 +66,7 @@ MODULE_FIRMWARE("radeon/pitcairn_ce.bin"); MODULE_FIRMWARE("radeon/pitcairn_mc.bin"); MODULE_FIRMWARE("radeon/pitcairn_rlc.bin"); MODULE_FIRMWARE("radeon/pitcairn_smc.bin"); +MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin"); MODULE_FIRMWARE("radeon/VERDE_pfp.bin"); MODULE_FIRMWARE("radeon/VERDE_me.bin"); @@ -80,6 +82,7 @@ MODULE_FIRMWARE("radeon/verde_ce.bin"); MODULE_FIRMWARE("radeon/verde_mc.bin"); MODULE_FIRMWARE("radeon/verde_rlc.bin"); MODULE_FIRMWARE("radeon/verde_smc.bin"); +MODULE_FIRMWARE("radeon/verde_k_smc.bin"); MODULE_FIRMWARE("radeon/OLAND_pfp.bin"); MODULE_FIRMWARE("radeon/OLAND_me.bin"); @@ -95,6 +98,7 @@ MODULE_FIRMWARE("radeon/oland_ce.bin"); MODULE_FIRMWARE("radeon/oland_mc.bin"); MODULE_FIRMWARE("radeon/oland_rlc.bin"); MODULE_FIRMWARE("radeon/oland_smc.bin"); +MODULE_FIRMWARE("radeon/oland_k_smc.bin"); MODULE_FIRMWARE("radeon/HAINAN_pfp.bin"); MODULE_FIRMWARE("radeon/HAINAN_me.bin"); @@ -110,6 +114,7 @@ MODULE_FIRMWARE("radeon/hainan_ce.bin"); MODULE_FIRMWARE("radeon/hainan_mc.bin"); MODULE_FIRMWARE("radeon/hainan_rlc.bin"); MODULE_FIRMWARE("radeon/hainan_smc.bin"); +MODULE_FIRMWARE("radeon/hainan_k_smc.bin"); static u32 si_get_cu_active_bitmap(struct radeon_device *rdev, u32 se, u32 sh); static void si_pcie_gen3_enable(struct radeon_device *rdev); @@ -1653,12 +1658,16 @@ static int si_init_microcode(struct radeon_device *rdev) char fw_name[30]; int err; int new_fw = 0; + bool new_smc = false; DRM_DEBUG("\n"); switch (rdev->family) { case CHIP_TAHITI: chip_name = "TAHITI"; + /* XXX: figure out which Tahitis need the new ucode */ + if (0) + new_smc = true; new_chip_name = "tahiti"; pfp_req_size = SI_PFP_UCODE_SIZE * 4; me_req_size = SI_PM4_UCODE_SIZE * 4; @@ -1670,6 +1679,13 @@ static int si_init_microcode(struct radeon_device *rdev) break; case CHIP_PITCAIRN: chip_name = "PITCAIRN"; + if ((rdev->pdev->revision == 0x81) || + (rdev->pdev->device == 0x6810) || + (rdev->pdev->device == 0x6811) || + (rdev->pdev->device == 0x6816) || + (rdev->pdev->device == 0x6817) || + (rdev->pdev->device == 0x6806)) + new_smc = true; new_chip_name = "pitcairn"; pfp_req_size = SI_PFP_UCODE_SIZE * 4; me_req_size = SI_PM4_UCODE_SIZE * 4; @@ -1681,6 +1697,16 @@ static int si_init_microcode(struct radeon_device *rdev) break; case CHIP_VERDE: chip_name = "VERDE"; + if ((rdev->pdev->revision == 0x81) || + (rdev->pdev->revision == 0x83) || + (rdev->pdev->revision == 0x87) || + (rdev->pdev->device == 0x6820) || + (rdev->pdev->device == 0x6821) || + (rdev->pdev->device == 0x6822) || + (rdev->pdev->device == 0x6823) || + (rdev->pdev->device == 0x682A) || + (rdev->pdev->device == 0x682B)) + new_smc = true; new_chip_name = "verde"; pfp_req_size = SI_PFP_UCODE_SIZE * 4; me_req_size = SI_PM4_UCODE_SIZE * 4; @@ -1692,6 +1718,13 @@ static int si_init_microcode(struct radeon_device *rdev) break; case CHIP_OLAND: chip_name = "OLAND"; + if ((rdev->pdev->revision == 0xC7) || + (rdev->pdev->revision == 0x80) || + (rdev->pdev->revision == 0x81) || + (rdev->pdev->revision == 0x83) || + (rdev->pdev->device == 0x6604) || + (rdev->pdev->device == 0x6605)) + new_smc = true; new_chip_name = "oland"; pfp_req_size = SI_PFP_UCODE_SIZE * 4; me_req_size = SI_PM4_UCODE_SIZE * 4; @@ -1702,6 +1735,13 @@ static int si_init_microcode(struct radeon_device *rdev) break; case CHIP_HAINAN: chip_name = "HAINAN"; + if ((rdev->pdev->revision == 0x81) || + (rdev->pdev->revision == 0x83) || + (rdev->pdev->revision == 0xC3) || + (rdev->pdev->device == 0x6664) || + (rdev->pdev->device == 0x6665) || + (rdev->pdev->device == 0x6667)) + new_smc = true; new_chip_name = "hainan"; pfp_req_size = SI_PFP_UCODE_SIZE * 4; me_req_size = SI_PM4_UCODE_SIZE * 4; @@ -1847,7 +1887,10 @@ static int si_init_microcode(struct radeon_device *rdev) } } - snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name); + if (new_smc) + snprintf(fw_name, sizeof(fw_name), "radeon/%s_k_smc.bin", new_chip_name); + else + snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", new_chip_name); err = request_firmware(&rdev->smc_fw, fw_name, rdev->dev); if (err) { snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name); diff --git a/drivers/gpu/drm/sti/Kconfig b/drivers/gpu/drm/sti/Kconfig index 5ad43a1bb260..494ab257f77c 100644 --- a/drivers/gpu/drm/sti/Kconfig +++ b/drivers/gpu/drm/sti/Kconfig @@ -7,5 +7,6 @@ config DRM_STI select DRM_KMS_CMA_HELPER select DRM_PANEL select FW_LOADER + select SND_SOC_HDMI_CODEC if SND_SOC help Choose this option to enable DRM on STM stiH41x chipset diff --git a/drivers/gpu/drm/sti/sti_awg_utils.c b/drivers/gpu/drm/sti/sti_awg_utils.c index a516eb869f6f..2da7d6866d5d 100644 --- a/drivers/gpu/drm/sti/sti_awg_utils.c +++ b/drivers/gpu/drm/sti/sti_awg_utils.c @@ -6,6 +6,8 @@ #include "sti_awg_utils.h" +#define AWG_DELAY (-5) + #define AWG_OPCODE_OFFSET 10 #define AWG_MAX_ARG 0x3ff @@ -125,7 +127,7 @@ static int awg_generate_line_signal( val = timing->blanking_level; ret |= awg_generate_instr(RPLSET, val, 0, 0, fwparams); - val = timing->trailing_pixels - 1; + val = timing->trailing_pixels - 1 + AWG_DELAY; ret |= awg_generate_instr(SKIP, val, 0, 0, fwparams); } diff --git a/drivers/gpu/drm/sti/sti_crtc.c b/drivers/gpu/drm/sti/sti_crtc.c index 7fab3af7473b..c7d734dc3cf4 100644 --- a/drivers/gpu/drm/sti/sti_crtc.c +++ b/drivers/gpu/drm/sti/sti_crtc.c @@ -23,22 +23,11 @@ static void sti_crtc_enable(struct drm_crtc *crtc) { struct sti_mixer *mixer = to_sti_mixer(crtc); - struct device *dev = mixer->dev; - struct sti_compositor *compo = dev_get_drvdata(dev); DRM_DEBUG_DRIVER("\n"); mixer->status = STI_MIXER_READY; - /* Prepare and enable the compo IP clock */ - if (mixer->id == STI_MIXER_MAIN) { - if (clk_prepare_enable(compo->clk_compo_main)) - DRM_INFO("Failed to prepare/enable compo_main clk\n"); - } else { - if (clk_prepare_enable(compo->clk_compo_aux)) - DRM_INFO("Failed to prepare/enable compo_aux clk\n"); - } - drm_crtc_vblank_on(crtc); } @@ -57,9 +46,8 @@ sti_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode) struct sti_mixer *mixer = to_sti_mixer(crtc); struct device *dev = mixer->dev; struct sti_compositor *compo = dev_get_drvdata(dev); - struct clk *clk; + struct clk *compo_clk, *pix_clk; int rate = mode->clock * 1000; - int res; DRM_DEBUG_KMS("CRTC:%d (%s) mode:%d (%s)\n", crtc->base.id, sti_mixer_to_str(mixer), @@ -74,32 +62,46 @@ sti_crtc_mode_set(struct drm_crtc *crtc, struct drm_display_mode *mode) mode->vsync_start, mode->vsync_end, mode->vtotal, mode->type, mode->flags); - /* Set rate and prepare/enable pixel clock */ - if (mixer->id == STI_MIXER_MAIN) - clk = compo->clk_pix_main; - else - clk = compo->clk_pix_aux; + if (mixer->id == STI_MIXER_MAIN) { + compo_clk = compo->clk_compo_main; + pix_clk = compo->clk_pix_main; + } else { + compo_clk = compo->clk_compo_aux; + pix_clk = compo->clk_pix_aux; + } + + /* Prepare and enable the compo IP clock */ + if (clk_prepare_enable(compo_clk)) { + DRM_INFO("Failed to prepare/enable compositor clk\n"); + goto compo_error; + } - res = clk_set_rate(clk, rate); - if (res < 0) { + /* Set rate and prepare/enable pixel clock */ + if (clk_set_rate(pix_clk, rate) < 0) { DRM_ERROR("Cannot set rate (%dHz) for pix clk\n", rate); - return -EINVAL; + goto pix_error; } - if (clk_prepare_enable(clk)) { + if (clk_prepare_enable(pix_clk)) { DRM_ERROR("Failed to prepare/enable pix clk\n"); - return -EINVAL; + goto pix_error; } sti_vtg_set_config(mixer->id == STI_MIXER_MAIN ? compo->vtg_main : compo->vtg_aux, &crtc->mode); - res = sti_mixer_active_video_area(mixer, &crtc->mode); - if (res) { + if (sti_mixer_active_video_area(mixer, &crtc->mode)) { DRM_ERROR("Can't set active video area\n"); - return -EINVAL; + goto mixer_error; } - return res; + return 0; + +mixer_error: + clk_disable_unprepare(pix_clk); +pix_error: + clk_disable_unprepare(compo_clk); +compo_error: + return -EINVAL; } static void sti_crtc_disable(struct drm_crtc *crtc) @@ -130,7 +132,6 @@ static void sti_crtc_disable(struct drm_crtc *crtc) static void sti_crtc_mode_set_nofb(struct drm_crtc *crtc) { - sti_crtc_enable(crtc); sti_crtc_mode_set(crtc, &crtc->state->adjusted_mode); } @@ -221,9 +222,7 @@ static void sti_crtc_atomic_flush(struct drm_crtc *crtc, static const struct drm_crtc_helper_funcs sti_crtc_helper_funcs = { .enable = sti_crtc_enable, .disable = sti_crtc_disabling, - .mode_set = drm_helper_crtc_mode_set, .mode_set_nofb = sti_crtc_mode_set_nofb, - .mode_set_base = drm_helper_crtc_mode_set_base, .atomic_begin = sti_crtc_atomic_begin, .atomic_flush = sti_crtc_atomic_flush, }; diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c index 8d1402b245bf..fedc17f98d9b 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.c +++ b/drivers/gpu/drm/sti/sti_hdmi.c @@ -18,6 +18,8 @@ #include <drm/drm_crtc_helper.h> #include <drm/drm_edid.h> +#include <sound/hdmi-codec.h> + #include "sti_hdmi.h" #include "sti_hdmi_tx3g4c28phy.h" #include "sti_hdmi_tx3g0c55phy.h" @@ -35,6 +37,8 @@ #define HDMI_DFLT_CHL0_DAT 0x0110 #define HDMI_DFLT_CHL1_DAT 0x0114 #define HDMI_DFLT_CHL2_DAT 0x0118 +#define HDMI_AUDIO_CFG 0x0200 +#define HDMI_SPDIF_FIFO_STATUS 0x0204 #define HDMI_SW_DI_1_HEAD_WORD 0x0210 #define HDMI_SW_DI_1_PKT_WORD0 0x0214 #define HDMI_SW_DI_1_PKT_WORD1 0x0218 @@ -44,6 +48,9 @@ #define HDMI_SW_DI_1_PKT_WORD5 0x0228 #define HDMI_SW_DI_1_PKT_WORD6 0x022C #define HDMI_SW_DI_CFG 0x0230 +#define HDMI_SAMPLE_FLAT_MASK 0x0244 +#define HDMI_AUDN 0x0400 +#define HDMI_AUD_CTS 0x0404 #define HDMI_SW_DI_2_HEAD_WORD 0x0600 #define HDMI_SW_DI_2_PKT_WORD0 0x0604 #define HDMI_SW_DI_2_PKT_WORD1 0x0608 @@ -103,6 +110,7 @@ #define HDMI_INT_DLL_LCK BIT(5) #define HDMI_INT_NEW_FRAME BIT(6) #define HDMI_INT_GENCTRL_PKT BIT(7) +#define HDMI_INT_AUDIO_FIFO_XRUN BIT(8) #define HDMI_INT_SINK_TERM_PRESENT BIT(11) #define HDMI_DEFAULT_INT (HDMI_INT_SINK_TERM_PRESENT \ @@ -111,6 +119,7 @@ | HDMI_INT_GLOBAL) #define HDMI_WORKING_INT (HDMI_INT_SINK_TERM_PRESENT \ + | HDMI_INT_AUDIO_FIFO_XRUN \ | HDMI_INT_GENCTRL_PKT \ | HDMI_INT_NEW_FRAME \ | HDMI_INT_DLL_LCK \ @@ -121,6 +130,27 @@ #define HDMI_STA_SW_RST BIT(1) +#define HDMI_AUD_CFG_8CH BIT(0) +#define HDMI_AUD_CFG_SPDIF_DIV_2 BIT(1) +#define HDMI_AUD_CFG_SPDIF_DIV_3 BIT(2) +#define HDMI_AUD_CFG_SPDIF_CLK_DIV_4 (BIT(1) | BIT(2)) +#define HDMI_AUD_CFG_CTS_CLK_256FS BIT(12) +#define HDMI_AUD_CFG_DTS_INVALID BIT(16) +#define HDMI_AUD_CFG_ONE_BIT_INVALID (BIT(18) | BIT(19) | BIT(20) | BIT(21)) +#define HDMI_AUD_CFG_CH12_VALID BIT(28) +#define HDMI_AUD_CFG_CH34_VALID BIT(29) +#define HDMI_AUD_CFG_CH56_VALID BIT(30) +#define HDMI_AUD_CFG_CH78_VALID BIT(31) + +/* sample flat mask */ +#define HDMI_SAMPLE_FLAT_NO 0 +#define HDMI_SAMPLE_FLAT_SP0 BIT(0) +#define HDMI_SAMPLE_FLAT_SP1 BIT(1) +#define HDMI_SAMPLE_FLAT_SP2 BIT(2) +#define HDMI_SAMPLE_FLAT_SP3 BIT(3) +#define HDMI_SAMPLE_FLAT_ALL (HDMI_SAMPLE_FLAT_SP0 | HDMI_SAMPLE_FLAT_SP1 |\ + HDMI_SAMPLE_FLAT_SP2 | HDMI_SAMPLE_FLAT_SP3) + #define HDMI_INFOFRAME_HEADER_TYPE(x) (((x) & 0xff) << 0) #define HDMI_INFOFRAME_HEADER_VERSION(x) (((x) & 0xff) << 8) #define HDMI_INFOFRAME_HEADER_LEN(x) (((x) & 0x0f) << 16) @@ -171,6 +201,10 @@ static irqreturn_t hdmi_irq_thread(int irq, void *arg) wake_up_interruptible(&hdmi->wait_event); } + /* Audio FIFO underrun IRQ */ + if (hdmi->irq_status & HDMI_INT_AUDIO_FIFO_XRUN) + DRM_INFO("Warning: audio FIFO underrun occurs!"); + return IRQ_HANDLED; } @@ -441,26 +475,29 @@ static int hdmi_avi_infoframe_config(struct sti_hdmi *hdmi) */ static int hdmi_audio_infoframe_config(struct sti_hdmi *hdmi) { - struct hdmi_audio_infoframe infofame; + struct hdmi_audio_params *audio = &hdmi->audio; u8 buffer[HDMI_INFOFRAME_SIZE(AUDIO)]; - int ret; - - ret = hdmi_audio_infoframe_init(&infofame); - if (ret < 0) { - DRM_ERROR("failed to setup audio infoframe: %d\n", ret); - return ret; - } - - infofame.channels = 2; - - ret = hdmi_audio_infoframe_pack(&infofame, buffer, sizeof(buffer)); - if (ret < 0) { - DRM_ERROR("failed to pack audio infoframe: %d\n", ret); - return ret; + int ret, val; + + DRM_DEBUG_DRIVER("enter %s, AIF %s\n", __func__, + audio->enabled ? "enable" : "disable"); + if (audio->enabled) { + /* set audio parameters stored*/ + ret = hdmi_audio_infoframe_pack(&audio->cea, buffer, + sizeof(buffer)); + if (ret < 0) { + DRM_ERROR("failed to pack audio infoframe: %d\n", ret); + return ret; + } + hdmi_infoframe_write_infopack(hdmi, buffer, ret); + } else { + /*disable audio info frame transmission */ + val = hdmi_read(hdmi, HDMI_SW_DI_CFG); + val &= ~HDMI_IFRAME_CFG_DI_N(HDMI_IFRAME_MASK, + HDMI_IFRAME_SLOT_AUDIO); + hdmi_write(hdmi, val, HDMI_SW_DI_CFG); } - hdmi_infoframe_write_infopack(hdmi, buffer, ret); - return 0; } @@ -650,6 +687,10 @@ static int hdmi_dbg_show(struct seq_file *s, void *data) DBGFS_DUMP("", HDMI_SW_DI_CFG); hdmi_dbg_sw_di_cfg(s, hdmi_read(hdmi, HDMI_SW_DI_CFG)); + DBGFS_DUMP("\n", HDMI_AUDIO_CFG); + DBGFS_DUMP("\n", HDMI_SPDIF_FIFO_STATUS); + DBGFS_DUMP("\n", HDMI_AUDN); + seq_printf(s, "\n AVI Infoframe (Data Island slot N=%d):", HDMI_IFRAME_SLOT_AVI); DBGFS_DUMP_DI(HDMI_SW_DI_N_HEAD_WORD, HDMI_IFRAME_SLOT_AVI); @@ -854,6 +895,7 @@ static int sti_hdmi_connector_get_modes(struct drm_connector *connector) count = drm_add_edid_modes(connector, edid); drm_mode_connector_update_edid_property(connector, edid); + drm_edid_to_eld(connector, edid); kfree(edid); return count; @@ -1036,6 +1078,207 @@ static struct drm_encoder *sti_hdmi_find_encoder(struct drm_device *dev) return NULL; } +/** + * sti_hdmi_audio_get_non_coherent_n() - get N parameter for non-coherent + * clocks. None-coherent clocks means that audio and TMDS clocks have not the + * same source (drifts between clocks). In this case assumption is that CTS is + * automatically calculated by hardware. + * + * @audio_fs: audio frame clock frequency in Hz + * + * Values computed are based on table described in HDMI specification 1.4b + * + * Returns n value. + */ +static int sti_hdmi_audio_get_non_coherent_n(unsigned int audio_fs) +{ + unsigned int n; + + switch (audio_fs) { + case 32000: + n = 4096; + break; + case 44100: + n = 6272; + break; + case 48000: + n = 6144; + break; + case 88200: + n = 6272 * 2; + break; + case 96000: + n = 6144 * 2; + break; + case 176400: + n = 6272 * 4; + break; + case 192000: + n = 6144 * 4; + break; + default: + /* Not pre-defined, recommended value: 128 * fs / 1000 */ + n = (audio_fs * 128) / 1000; + } + + return n; +} + +static int hdmi_audio_configure(struct sti_hdmi *hdmi, + struct hdmi_audio_params *params) +{ + int audio_cfg, n; + struct hdmi_audio_infoframe *info = ¶ms->cea; + + DRM_DEBUG_DRIVER("\n"); + + if (!hdmi->enabled) + return 0; + + /* update N parameter */ + n = sti_hdmi_audio_get_non_coherent_n(params->sample_rate); + + DRM_DEBUG_DRIVER("Audio rate = %d Hz, TMDS clock = %d Hz, n = %d\n", + params->sample_rate, hdmi->mode.clock * 1000, n); + hdmi_write(hdmi, n, HDMI_AUDN); + + /* update HDMI registers according to configuration */ + audio_cfg = HDMI_AUD_CFG_SPDIF_DIV_2 | HDMI_AUD_CFG_DTS_INVALID | + HDMI_AUD_CFG_ONE_BIT_INVALID; + + switch (info->channels) { + case 8: + audio_cfg |= HDMI_AUD_CFG_CH78_VALID; + case 6: + audio_cfg |= HDMI_AUD_CFG_CH56_VALID; + case 4: + audio_cfg |= HDMI_AUD_CFG_CH34_VALID | HDMI_AUD_CFG_8CH; + case 2: + audio_cfg |= HDMI_AUD_CFG_CH12_VALID; + break; + default: + DRM_ERROR("ERROR: Unsupported number of channels (%d)!\n", + info->channels); + return -EINVAL; + } + + hdmi_write(hdmi, audio_cfg, HDMI_AUDIO_CFG); + + hdmi->audio = *params; + + return hdmi_audio_infoframe_config(hdmi); +} + +static void hdmi_audio_shutdown(struct device *dev, void *data) +{ + struct sti_hdmi *hdmi = dev_get_drvdata(dev); + int audio_cfg; + + DRM_DEBUG_DRIVER("\n"); + + /* disable audio */ + audio_cfg = HDMI_AUD_CFG_SPDIF_DIV_2 | HDMI_AUD_CFG_DTS_INVALID | + HDMI_AUD_CFG_ONE_BIT_INVALID; + hdmi_write(hdmi, audio_cfg, HDMI_AUDIO_CFG); + + hdmi->audio.enabled = 0; + hdmi_audio_infoframe_config(hdmi); +} + +static int hdmi_audio_hw_params(struct device *dev, + void *data, + struct hdmi_codec_daifmt *daifmt, + struct hdmi_codec_params *params) +{ + struct sti_hdmi *hdmi = dev_get_drvdata(dev); + int ret; + struct hdmi_audio_params audio = { + .sample_width = params->sample_width, + .sample_rate = params->sample_rate, + .cea = params->cea, + }; + + DRM_DEBUG_DRIVER("\n"); + + if (!hdmi->enabled) + return 0; + + if ((daifmt->fmt != HDMI_I2S) || daifmt->bit_clk_inv || + daifmt->frame_clk_inv || daifmt->bit_clk_master || + daifmt->frame_clk_master) { + dev_err(dev, "%s: Bad flags %d %d %d %d\n", __func__, + daifmt->bit_clk_inv, daifmt->frame_clk_inv, + daifmt->bit_clk_master, + daifmt->frame_clk_master); + return -EINVAL; + } + + audio.enabled = 1; + + ret = hdmi_audio_configure(hdmi, &audio); + if (ret < 0) + return ret; + + return 0; +} + +static int hdmi_audio_digital_mute(struct device *dev, void *data, bool enable) +{ + struct sti_hdmi *hdmi = dev_get_drvdata(dev); + + DRM_DEBUG_DRIVER("%s\n", enable ? "enable" : "disable"); + + if (enable) + hdmi_write(hdmi, HDMI_SAMPLE_FLAT_ALL, HDMI_SAMPLE_FLAT_MASK); + else + hdmi_write(hdmi, HDMI_SAMPLE_FLAT_NO, HDMI_SAMPLE_FLAT_MASK); + + return 0; +} + +static int hdmi_audio_get_eld(struct device *dev, void *data, uint8_t *buf, size_t len) +{ + struct sti_hdmi *hdmi = dev_get_drvdata(dev); + struct drm_connector *connector = hdmi->drm_connector; + + DRM_DEBUG_DRIVER("\n"); + memcpy(buf, connector->eld, min(sizeof(connector->eld), len)); + + return 0; +} + +static const struct hdmi_codec_ops audio_codec_ops = { + .hw_params = hdmi_audio_hw_params, + .audio_shutdown = hdmi_audio_shutdown, + .digital_mute = hdmi_audio_digital_mute, + .get_eld = hdmi_audio_get_eld, +}; + +static int sti_hdmi_register_audio_driver(struct device *dev, + struct sti_hdmi *hdmi) +{ + struct hdmi_codec_pdata codec_data = { + .ops = &audio_codec_ops, + .max_i2s_channels = 8, + .i2s = 1, + }; + + DRM_DEBUG_DRIVER("\n"); + + hdmi->audio.enabled = 0; + + hdmi->audio_pdev = platform_device_register_data( + dev, HDMI_CODEC_DRV_NAME, PLATFORM_DEVID_AUTO, + &codec_data, sizeof(codec_data)); + + if (IS_ERR(hdmi->audio_pdev)) + return PTR_ERR(hdmi->audio_pdev); + + DRM_INFO("%s Driver bound %s\n", HDMI_CODEC_DRV_NAME, dev_name(dev)); + + return 0; +} + static int sti_hdmi_bind(struct device *dev, struct device *master, void *data) { struct sti_hdmi *hdmi = dev_get_drvdata(dev); @@ -1082,12 +1325,27 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data) /* initialise property */ sti_hdmi_connector_init_property(drm_dev, drm_connector); + hdmi->drm_connector = drm_connector; + err = drm_mode_connector_attach_encoder(drm_connector, encoder); if (err) { DRM_ERROR("Failed to attach a connector to a encoder\n"); goto err_sysfs; } + err = sti_hdmi_register_audio_driver(dev, hdmi); + if (err) { + DRM_ERROR("Failed to attach an audio codec\n"); + goto err_sysfs; + } + + /* Initialize audio infoframe */ + err = hdmi_audio_infoframe_init(&hdmi->audio.cea); + if (err) { + DRM_ERROR("Failed to init audio infoframe\n"); + goto err_sysfs; + } + /* Enable default interrupts */ hdmi_write(hdmi, HDMI_DEFAULT_INT, HDMI_INT_EN); @@ -1095,6 +1353,7 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data) err_sysfs: drm_bridge_remove(bridge); + hdmi->drm_connector = NULL; return -EINVAL; } @@ -1244,6 +1503,8 @@ static int sti_hdmi_remove(struct platform_device *pdev) struct sti_hdmi *hdmi = dev_get_drvdata(&pdev->dev); i2c_put_adapter(hdmi->ddc_adapt); + if (hdmi->audio_pdev) + platform_device_unregister(hdmi->audio_pdev); component_del(&pdev->dev, &sti_hdmi_ops); return 0; diff --git a/drivers/gpu/drm/sti/sti_hdmi.h b/drivers/gpu/drm/sti/sti_hdmi.h index ef3a94583bbd..119bc3582ac7 100644 --- a/drivers/gpu/drm/sti/sti_hdmi.h +++ b/drivers/gpu/drm/sti/sti_hdmi.h @@ -23,6 +23,13 @@ struct hdmi_phy_ops { void (*stop)(struct sti_hdmi *hdmi); }; +struct hdmi_audio_params { + bool enabled; + unsigned int sample_width; + unsigned int sample_rate; + struct hdmi_audio_infoframe cea; +}; + /* values for the framing mode property */ enum sti_hdmi_modes { HDMI_MODE_HDMI, @@ -67,6 +74,9 @@ static const struct drm_prop_enum_list colorspace_mode_names[] = { * @ddc_adapt: i2c ddc adapter * @colorspace: current colorspace selected * @hdmi_mode: select framing for HDMI or DVI + * @audio_pdev: ASoC hdmi-codec platform device + * @audio: hdmi audio parameters. + * @drm_connector: hdmi connector */ struct sti_hdmi { struct device dev; @@ -89,6 +99,9 @@ struct sti_hdmi { struct i2c_adapter *ddc_adapt; enum hdmi_colorspace colorspace; enum sti_hdmi_modes hdmi_mode; + struct platform_device *audio_pdev; + struct hdmi_audio_params audio; + struct drm_connector *drm_connector; }; u32 hdmi_read(struct sti_hdmi *hdmi, int offset); diff --git a/drivers/gpu/drm/sti/sti_plane.c b/drivers/gpu/drm/sti/sti_plane.c index 85cee9098439..0cf3335ef37c 100644 --- a/drivers/gpu/drm/sti/sti_plane.c +++ b/drivers/gpu/drm/sti/sti_plane.c @@ -45,25 +45,15 @@ const char *sti_plane_to_str(struct sti_plane *plane) #define STI_FPS_INTERVAL_MS 3000 -static int sti_plane_timespec_ms_diff(struct timespec lhs, struct timespec rhs) -{ - struct timespec tmp_ts = timespec_sub(lhs, rhs); - u64 tmp_ns = (u64)timespec_to_ns(&tmp_ts); - - do_div(tmp_ns, NSEC_PER_MSEC); - - return (u32)tmp_ns; -} - void sti_plane_update_fps(struct sti_plane *plane, bool new_frame, bool new_field) { - struct timespec now; + ktime_t now; struct sti_fps_info *fps; int fpks, fipks, ms_since_last, num_frames, num_fields; - getrawmonotonic(&now); + now = ktime_get(); /* Compute number of frame updates */ fps = &plane->fps_info; @@ -76,7 +66,7 @@ void sti_plane_update_fps(struct sti_plane *plane, return; fps->curr_frame_counter++; - ms_since_last = sti_plane_timespec_ms_diff(now, fps->last_timestamp); + ms_since_last = ktime_to_ms(ktime_sub(now, fps->last_timestamp)); num_frames = fps->curr_frame_counter - fps->last_frame_counter; if (num_frames <= 0 || ms_since_last < STI_FPS_INTERVAL_MS) diff --git a/drivers/gpu/drm/sti/sti_plane.h b/drivers/gpu/drm/sti/sti_plane.h index 39d39f5b7dd9..e0ea1dd3bb88 100644 --- a/drivers/gpu/drm/sti/sti_plane.h +++ b/drivers/gpu/drm/sti/sti_plane.h @@ -55,7 +55,7 @@ struct sti_fps_info { unsigned int last_frame_counter; unsigned int curr_field_counter; unsigned int last_field_counter; - struct timespec last_timestamp; + ktime_t last_timestamp; char fps_str[FPS_LENGTH]; char fips_str[FPS_LENGTH]; }; diff --git a/drivers/gpu/drm/sti/sti_vtg.c b/drivers/gpu/drm/sti/sti_vtg.c index 0100c7c668b7..0bdc385eec17 100644 --- a/drivers/gpu/drm/sti/sti_vtg.c +++ b/drivers/gpu/drm/sti/sti_vtg.c @@ -65,7 +65,7 @@ #define HDMI_DELAY (5) /* Delay introduced by the DVO in nb of pixel */ -#define DVO_DELAY (2) +#define DVO_DELAY (7) /* delay introduced by the Arbitrary Waveform Generator in nb of pixels */ #define AWG_DELAY_HD (-9) diff --git a/drivers/gpu/drm/sun4i/Kconfig b/drivers/gpu/drm/sun4i/Kconfig index 99510e64e91a..a4b357db8856 100644 --- a/drivers/gpu/drm/sun4i/Kconfig +++ b/drivers/gpu/drm/sun4i/Kconfig @@ -1,6 +1,6 @@ config DRM_SUN4I tristate "DRM Support for Allwinner A10 Display Engine" - depends on DRM && ARM + depends on DRM && ARM && COMMON_CLK depends on ARCH_SUNXI || COMPILE_TEST select DRM_GEM_CMA_HELPER select DRM_KMS_HELPER diff --git a/drivers/gpu/drm/sun4i/sun4i_backend.c b/drivers/gpu/drm/sun4i/sun4i_backend.c index f7a15c1a93bf..3ab560450a82 100644 --- a/drivers/gpu/drm/sun4i/sun4i_backend.c +++ b/drivers/gpu/drm/sun4i/sun4i_backend.c @@ -190,7 +190,7 @@ int sun4i_backend_update_layer_buffer(struct sun4i_backend *backend, /* Get the physical address of the buffer in memory */ gem = drm_fb_cma_get_gem_obj(fb, 0); - DRM_DEBUG_DRIVER("Using GEM @ 0x%x\n", gem->paddr); + DRM_DEBUG_DRIVER("Using GEM @ %pad\n", &gem->paddr); /* Compute the start of the displayed memory */ bpp = drm_format_plane_cpp(fb->pixel_format, 0); @@ -198,7 +198,7 @@ int sun4i_backend_update_layer_buffer(struct sun4i_backend *backend, paddr += (state->src_x >> 16) * bpp; paddr += (state->src_y >> 16) * fb->pitches[0]; - DRM_DEBUG_DRIVER("Setting buffer address to 0x%x\n", paddr); + DRM_DEBUG_DRIVER("Setting buffer address to %pad\n", &paddr); /* Write the 32 lower bits of the address (in bits) */ lo_paddr = paddr << 3; diff --git a/drivers/gpu/drm/sun4i/sun4i_dotclock.c b/drivers/gpu/drm/sun4i/sun4i_dotclock.c index 3ff668cb463c..5b3463197c48 100644 --- a/drivers/gpu/drm/sun4i/sun4i_dotclock.c +++ b/drivers/gpu/drm/sun4i/sun4i_dotclock.c @@ -72,14 +72,40 @@ static unsigned long sun4i_dclk_recalc_rate(struct clk_hw *hw, static long sun4i_dclk_round_rate(struct clk_hw *hw, unsigned long rate, unsigned long *parent_rate) { - return *parent_rate / DIV_ROUND_CLOSEST(*parent_rate, rate); + unsigned long best_parent = 0; + u8 best_div = 1; + int i; + + for (i = 6; i < 127; i++) { + unsigned long ideal = rate * i; + unsigned long rounded; + + rounded = clk_hw_round_rate(clk_hw_get_parent(hw), + ideal); + + if (rounded == ideal) { + best_parent = rounded; + best_div = i; + goto out; + } + + if ((rounded < ideal) && (rounded > best_parent)) { + best_parent = rounded; + best_div = i; + } + } + +out: + *parent_rate = best_parent; + + return best_parent / best_div; } static int sun4i_dclk_set_rate(struct clk_hw *hw, unsigned long rate, unsigned long parent_rate) { struct sun4i_dclk *dclk = hw_to_dclk(hw); - int div = DIV_ROUND_CLOSEST(parent_rate, rate); + u8 div = parent_rate / rate; return regmap_update_bits(dclk->regmap, SUN4I_TCON0_DCLK_REG, GENMASK(6, 0), div); @@ -127,10 +153,14 @@ int sun4i_dclk_create(struct device *dev, struct sun4i_tcon *tcon) const char *clk_name, *parent_name; struct clk_init_data init; struct sun4i_dclk *dclk; + int ret; parent_name = __clk_get_name(tcon->sclk0); - of_property_read_string_index(dev->of_node, "clock-output-names", 0, - &clk_name); + ret = of_property_read_string_index(dev->of_node, + "clock-output-names", 0, + &clk_name); + if (ret) + return ret; dclk = devm_kzalloc(dev, sizeof(*dclk), GFP_KERNEL); if (!dclk) @@ -140,6 +170,7 @@ int sun4i_dclk_create(struct device *dev, struct sun4i_tcon *tcon) init.ops = &sun4i_dclk_ops; init.parent_names = &parent_name; init.num_parents = 1; + init.flags = CLK_SET_RATE_PARENT; dclk->regmap = tcon->regs; dclk->hw.init = &init; diff --git a/drivers/gpu/drm/sun4i/sun4i_drv.c b/drivers/gpu/drm/sun4i/sun4i_drv.c index 9a67f927a53e..5b89940edcb1 100644 --- a/drivers/gpu/drm/sun4i/sun4i_drv.c +++ b/drivers/gpu/drm/sun4i/sun4i_drv.c @@ -97,6 +97,22 @@ static struct drm_driver sun4i_drv_driver = { .disable_vblank = sun4i_drv_disable_vblank, }; +static void sun4i_remove_framebuffers(void) +{ + struct apertures_struct *ap; + + ap = alloc_apertures(1); + if (!ap) + return; + + /* The framebuffer can be located anywhere in RAM */ + ap->ranges[0].base = 0; + ap->ranges[0].size = ~0; + + remove_conflicting_framebuffers(ap, "sun4i-drm-fb", false); + kfree(ap); +} + static int sun4i_drv_bind(struct device *dev) { struct drm_device *drm; @@ -140,6 +156,9 @@ static int sun4i_drv_bind(struct device *dev) } drm->irq_enabled = true; + /* Remove early framebuffers (ie. simplefb) */ + sun4i_remove_framebuffers(); + /* Create our framebuffer */ drv->fbdev = sun4i_framebuffer_init(drm); if (IS_ERR(drv->fbdev)) { @@ -166,6 +185,7 @@ static void sun4i_drv_unbind(struct device *dev) { struct drm_device *drm = dev_get_drvdata(dev); + drm_connector_unregister_all(drm); drm_dev_unregister(drm); drm_kms_helper_poll_fini(drm); sun4i_framebuffer_free(drm); diff --git a/drivers/gpu/drm/sun4i/sun4i_rgb.c b/drivers/gpu/drm/sun4i/sun4i_rgb.c index 442cfe271688..f5bbac6efb4c 100644 --- a/drivers/gpu/drm/sun4i/sun4i_rgb.c +++ b/drivers/gpu/drm/sun4i/sun4i_rgb.c @@ -54,8 +54,13 @@ static int sun4i_rgb_get_modes(struct drm_connector *connector) static int sun4i_rgb_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { + struct sun4i_rgb *rgb = drm_connector_to_sun4i_rgb(connector); + struct sun4i_drv *drv = rgb->drv; + struct sun4i_tcon *tcon = drv->tcon; u32 hsync = mode->hsync_end - mode->hsync_start; u32 vsync = mode->vsync_end - mode->vsync_start; + unsigned long rate = mode->clock * 1000; + long rounded_rate; DRM_DEBUG_DRIVER("Validating modes...\n"); @@ -87,6 +92,15 @@ static int sun4i_rgb_mode_valid(struct drm_connector *connector, DRM_DEBUG_DRIVER("Vertical parameters OK\n"); + rounded_rate = clk_round_rate(tcon->dclk, rate); + if (rounded_rate < rate) + return MODE_CLOCK_LOW; + + if (rounded_rate > rate) + return MODE_CLOCK_HIGH; + + DRM_DEBUG_DRIVER("Clock rate OK\n"); + return MODE_OK; } @@ -193,7 +207,7 @@ int sun4i_rgb_init(struct drm_device *drm) int ret; /* If we don't have a panel, there's no point in going on */ - if (!tcon->panel) + if (IS_ERR(tcon->panel)) return -ENODEV; rgb = devm_kzalloc(drm->dev, sizeof(*rgb), GFP_KERNEL); diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c index 9f19b0e08560..652385f09735 100644 --- a/drivers/gpu/drm/sun4i/sun4i_tcon.c +++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c @@ -425,11 +425,11 @@ static struct drm_panel *sun4i_tcon_find_panel(struct device_node *node) remote = of_graph_get_remote_port_parent(end_node); if (!remote) { - DRM_DEBUG_DRIVER("Enable to parse remote node\n"); + DRM_DEBUG_DRIVER("Unable to parse remote node\n"); return ERR_PTR(-EINVAL); } - return of_drm_find_panel(remote); + return of_drm_find_panel(remote) ?: ERR_PTR(-EPROBE_DEFER); } static int sun4i_tcon_bind(struct device *dev, struct device *master, @@ -490,7 +490,11 @@ static int sun4i_tcon_bind(struct device *dev, struct device *master, return 0; } - return sun4i_rgb_init(drm); + ret = sun4i_rgb_init(drm); + if (ret < 0) + goto err_free_clocks; + + return 0; err_free_clocks: sun4i_tcon_free_clocks(tcon); @@ -522,12 +526,13 @@ static int sun4i_tcon_probe(struct platform_device *pdev) * Defer the probe. */ panel = sun4i_tcon_find_panel(node); - if (IS_ERR(panel)) { - /* - * If we don't have a panel endpoint, just go on - */ - if (PTR_ERR(panel) != -ENODEV) - return -EPROBE_DEFER; + + /* + * If we don't have a panel endpoint, just go on + */ + if (PTR_ERR(panel) == -EPROBE_DEFER) { + DRM_DEBUG_DRIVER("Still waiting for our panel. Deferring...\n"); + return -EPROBE_DEFER; } return component_add(&pdev->dev, &sun4i_tcon_ops); diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index 39386f50af87..e340d0d66429 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -147,9 +147,9 @@ static void ttm_bo_release_list(struct kref *list_kref) BUG_ON(!list_empty(&bo->lru)); BUG_ON(!list_empty(&bo->ddestroy)); - if (bo->ttm) - ttm_tt_destroy(bo->ttm); + ttm_tt_destroy(bo->ttm); atomic_dec(&bo->glob->bo_count); + fence_put(bo->moving); if (bo->resv == &bo->ttm_resv) reservation_object_fini(&bo->ttm_resv); mutex_destroy(&bo->wu_mutex); @@ -360,7 +360,8 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo, ret = bdev->driver->move(bo, evict, interruptible, no_wait_gpu, mem); else - ret = ttm_bo_move_memcpy(bo, evict, no_wait_gpu, mem); + ret = ttm_bo_move_memcpy(bo, evict, interruptible, + no_wait_gpu, mem); if (ret) { if (bdev->driver->move_notify) { @@ -396,8 +397,7 @@ moved: out_err: new_man = &bdev->man[bo->mem.mem_type]; - if ((new_man->flags & TTM_MEMTYPE_FLAG_FIXED) && bo->ttm) { - ttm_tt_unbind(bo->ttm); + if (new_man->flags & TTM_MEMTYPE_FLAG_FIXED) { ttm_tt_destroy(bo->ttm); bo->ttm = NULL; } @@ -418,11 +418,8 @@ static void ttm_bo_cleanup_memtype_use(struct ttm_buffer_object *bo) if (bo->bdev->driver->move_notify) bo->bdev->driver->move_notify(bo, NULL); - if (bo->ttm) { - ttm_tt_unbind(bo->ttm); - ttm_tt_destroy(bo->ttm); - bo->ttm = NULL; - } + ttm_tt_destroy(bo->ttm); + bo->ttm = NULL; ttm_bo_mem_put(bo, &bo->mem); ww_mutex_unlock (&bo->resv->lock); @@ -688,15 +685,6 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible, struct ttm_placement placement; int ret = 0; - ret = ttm_bo_wait(bo, interruptible, no_wait_gpu); - - if (unlikely(ret != 0)) { - if (ret != -ERESTARTSYS) { - pr_err("Failed to expire sync object before buffer eviction\n"); - } - goto out; - } - lockdep_assert_held(&bo->resv->lock.base); evict_mem = bo->mem; @@ -720,7 +708,7 @@ static int ttm_bo_evict(struct ttm_buffer_object *bo, bool interruptible, ret = ttm_bo_handle_move_mem(bo, &evict_mem, true, interruptible, no_wait_gpu); - if (ret) { + if (unlikely(ret)) { if (ret != -ERESTARTSYS) pr_err("Buffer eviction failed\n"); ttm_bo_mem_put(bo, &evict_mem); @@ -800,6 +788,34 @@ void ttm_bo_mem_put(struct ttm_buffer_object *bo, struct ttm_mem_reg *mem) EXPORT_SYMBOL(ttm_bo_mem_put); /** + * Add the last move fence to the BO and reserve a new shared slot. + */ +static int ttm_bo_add_move_fence(struct ttm_buffer_object *bo, + struct ttm_mem_type_manager *man, + struct ttm_mem_reg *mem) +{ + struct fence *fence; + int ret; + + spin_lock(&man->move_lock); + fence = fence_get(man->move); + spin_unlock(&man->move_lock); + + if (fence) { + reservation_object_add_shared_fence(bo->resv, fence); + + ret = reservation_object_reserve_shared(bo->resv); + if (unlikely(ret)) + return ret; + + fence_put(bo->moving); + bo->moving = fence; + } + + return 0; +} + +/** * Repeatedly evict memory from the LRU for @mem_type until we create enough * space, or we've evicted everything and there isn't enough space. */ @@ -825,10 +841,8 @@ static int ttm_bo_mem_force_space(struct ttm_buffer_object *bo, if (unlikely(ret != 0)) return ret; } while (1); - if (mem->mm_node == NULL) - return -ENOMEM; mem->mem_type = mem_type; - return 0; + return ttm_bo_add_move_fence(bo, man, mem); } static uint32_t ttm_bo_select_caching(struct ttm_mem_type_manager *man, @@ -898,6 +912,10 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo, bool has_erestartsys = false; int i, ret; + ret = reservation_object_reserve_shared(bo->resv); + if (unlikely(ret)) + return ret; + mem->mm_node = NULL; for (i = 0; i < placement->num_placement; ++i) { const struct ttm_place *place = &placement->placement[i]; @@ -931,9 +949,15 @@ int ttm_bo_mem_space(struct ttm_buffer_object *bo, ret = (*man->func->get_node)(man, bo, place, mem); if (unlikely(ret)) return ret; - - if (mem->mm_node) + + if (mem->mm_node) { + ret = ttm_bo_add_move_fence(bo, man, mem); + if (unlikely(ret)) { + (*man->func->put_node)(man, mem); + return ret; + } break; + } } if ((type_ok && (mem_type == TTM_PL_SYSTEM)) || mem->mm_node) { @@ -1000,20 +1024,6 @@ static int ttm_bo_move_buffer(struct ttm_buffer_object *bo, lockdep_assert_held(&bo->resv->lock.base); - /* - * Don't wait for the BO on initial allocation. This is important when - * the BO has an imported reservation object. - */ - if (bo->mem.mem_type != TTM_PL_SYSTEM || bo->ttm != NULL) { - /* - * FIXME: It's possible to pipeline buffer moves. - * Have the driver move function wait for idle when necessary, - * instead of doing it here. - */ - ret = ttm_bo_wait(bo, interruptible, no_wait_gpu); - if (ret) - return ret; - } mem.num_pages = bo->num_pages; mem.size = mem.num_pages << PAGE_SHIFT; mem.page_alignment = bo->mem.page_alignment; @@ -1165,7 +1175,7 @@ int ttm_bo_init(struct ttm_bo_device *bdev, bo->mem.page_alignment = page_alignment; bo->mem.bus.io_reserved_vm = false; bo->mem.bus.io_reserved_count = 0; - bo->priv_flags = 0; + bo->moving = NULL; bo->mem.placement = (TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED); bo->persistent_swap_storage = persistent_swap_storage; bo->acc_size = acc_size; @@ -1277,6 +1287,7 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev, { struct ttm_mem_type_manager *man = &bdev->man[mem_type]; struct ttm_bo_global *glob = bdev->glob; + struct fence *fence; int ret; /* @@ -1297,6 +1308,23 @@ static int ttm_bo_force_list_clean(struct ttm_bo_device *bdev, spin_lock(&glob->lru_lock); } spin_unlock(&glob->lru_lock); + + spin_lock(&man->move_lock); + fence = fence_get(man->move); + spin_unlock(&man->move_lock); + + if (fence) { + ret = fence_wait(fence, false); + fence_put(fence); + if (ret) { + if (allow_errors) { + return ret; + } else { + pr_err("Cleanup eviction failed\n"); + } + } + } + return 0; } @@ -1316,6 +1344,7 @@ int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type) mem_type); return ret; } + fence_put(man->move); man->use_type = false; man->has_type = false; @@ -1361,6 +1390,7 @@ int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, man->io_reserve_fastpath = true; man->use_io_reserve_lru = false; mutex_init(&man->io_reserve_mutex); + spin_lock_init(&man->move_lock); INIT_LIST_HEAD(&man->io_reserve_lru); ret = bdev->driver->init_mem_type(bdev, type, man); @@ -1379,6 +1409,7 @@ int ttm_bo_init_mm(struct ttm_bo_device *bdev, unsigned type, man->size = p_size; INIT_LIST_HEAD(&man->lru); + man->move = NULL; return 0; } @@ -1572,47 +1603,17 @@ EXPORT_SYMBOL(ttm_bo_unmap_virtual); int ttm_bo_wait(struct ttm_buffer_object *bo, bool interruptible, bool no_wait) { - struct reservation_object_list *fobj; - struct reservation_object *resv; - struct fence *excl; - long timeout = 15 * HZ; - int i; - - resv = bo->resv; - fobj = reservation_object_get_list(resv); - excl = reservation_object_get_excl(resv); - if (excl) { - if (!fence_is_signaled(excl)) { - if (no_wait) - return -EBUSY; - - timeout = fence_wait_timeout(excl, - interruptible, timeout); - } - } - - for (i = 0; fobj && timeout > 0 && i < fobj->shared_count; ++i) { - struct fence *fence; - fence = rcu_dereference_protected(fobj->shared[i], - reservation_object_held(resv)); - - if (!fence_is_signaled(fence)) { - if (no_wait) - return -EBUSY; - - timeout = fence_wait_timeout(fence, - interruptible, timeout); - } - } + long timeout = no_wait ? 0 : 15 * HZ; + timeout = reservation_object_wait_timeout_rcu(bo->resv, true, + interruptible, timeout); if (timeout < 0) return timeout; if (timeout == 0) return -EBUSY; - reservation_object_add_excl_fence(resv, NULL); - clear_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags); + reservation_object_add_excl_fence(bo->resv, NULL); return 0; } EXPORT_SYMBOL(ttm_bo_wait); @@ -1682,14 +1683,9 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink) ttm_bo_list_ref_sub(bo, put_count, true); /** - * Wait for GPU, then move to system cached. + * Move to system cached */ - ret = ttm_bo_wait(bo, false, false); - - if (unlikely(ret != 0)) - goto out; - if ((bo->mem.placement & swap_placement) != swap_placement) { struct ttm_mem_reg evict_mem; @@ -1704,6 +1700,14 @@ static int ttm_bo_swapout(struct ttm_mem_shrink *shrink) goto out; } + /** + * Make sure BO is idle. + */ + + ret = ttm_bo_wait(bo, false, false); + if (unlikely(ret != 0)) + goto out; + ttm_bo_unmap_virtual(bo); /** diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c index d9831559706e..4da0e784f9e7 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_util.c +++ b/drivers/gpu/drm/ttm/ttm_bo_util.c @@ -53,7 +53,6 @@ int ttm_bo_move_ttm(struct ttm_buffer_object *bo, int ret; if (old_mem->mem_type != TTM_PL_SYSTEM) { - ttm_tt_unbind(ttm); ttm_bo_free_old_node(bo); ttm_flag_masked(&old_mem->placement, TTM_PL_FLAG_SYSTEM, TTM_PL_MASK_MEM); @@ -321,7 +320,8 @@ static int ttm_copy_ttm_io_page(struct ttm_tt *ttm, void *dst, } int ttm_bo_move_memcpy(struct ttm_buffer_object *bo, - bool evict, bool no_wait_gpu, + bool evict, bool interruptible, + bool no_wait_gpu, struct ttm_mem_reg *new_mem) { struct ttm_bo_device *bdev = bo->bdev; @@ -337,6 +337,10 @@ int ttm_bo_move_memcpy(struct ttm_buffer_object *bo, unsigned long add = 0; int dir; + ret = ttm_bo_wait(bo, interruptible, no_wait_gpu); + if (ret) + return ret; + ret = ttm_mem_reg_ioremap(bdev, old_mem, &old_iomap); if (ret) return ret; @@ -401,8 +405,7 @@ out2: *old_mem = *new_mem; new_mem->mm_node = NULL; - if ((man->flags & TTM_MEMTYPE_FLAG_FIXED) && (ttm != NULL)) { - ttm_tt_unbind(ttm); + if (man->flags & TTM_MEMTYPE_FLAG_FIXED) { ttm_tt_destroy(ttm); bo->ttm = NULL; } @@ -462,6 +465,7 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo, INIT_LIST_HEAD(&fbo->lru); INIT_LIST_HEAD(&fbo->swap); INIT_LIST_HEAD(&fbo->io_reserve_lru); + fbo->moving = NULL; drm_vma_node_reset(&fbo->vma_node); atomic_set(&fbo->cpu_writers, 0); @@ -634,7 +638,6 @@ EXPORT_SYMBOL(ttm_bo_kunmap); int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo, struct fence *fence, bool evict, - bool no_wait_gpu, struct ttm_mem_reg *new_mem) { struct ttm_bo_device *bdev = bo->bdev; @@ -649,9 +652,7 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo, if (ret) return ret; - if ((man->flags & TTM_MEMTYPE_FLAG_FIXED) && - (bo->ttm != NULL)) { - ttm_tt_unbind(bo->ttm); + if (man->flags & TTM_MEMTYPE_FLAG_FIXED) { ttm_tt_destroy(bo->ttm); bo->ttm = NULL; } @@ -665,7 +666,8 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo, * operation has completed. */ - set_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags); + fence_put(bo->moving); + bo->moving = fence_get(fence); ret = ttm_buffer_object_transfer(bo, &ghost_obj); if (ret) @@ -694,3 +696,95 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo, return 0; } EXPORT_SYMBOL(ttm_bo_move_accel_cleanup); + +int ttm_bo_pipeline_move(struct ttm_buffer_object *bo, + struct fence *fence, bool evict, + struct ttm_mem_reg *new_mem) +{ + struct ttm_bo_device *bdev = bo->bdev; + struct ttm_mem_reg *old_mem = &bo->mem; + + struct ttm_mem_type_manager *from = &bdev->man[old_mem->mem_type]; + struct ttm_mem_type_manager *to = &bdev->man[new_mem->mem_type]; + + int ret; + + reservation_object_add_excl_fence(bo->resv, fence); + + if (!evict) { + struct ttm_buffer_object *ghost_obj; + + /** + * This should help pipeline ordinary buffer moves. + * + * Hang old buffer memory on a new buffer object, + * and leave it to be released when the GPU + * operation has completed. + */ + + fence_put(bo->moving); + bo->moving = fence_get(fence); + + ret = ttm_buffer_object_transfer(bo, &ghost_obj); + if (ret) + return ret; + + reservation_object_add_excl_fence(ghost_obj->resv, fence); + + /** + * If we're not moving to fixed memory, the TTM object + * needs to stay alive. Otherwhise hang it on the ghost + * bo to be unbound and destroyed. + */ + + if (!(to->flags & TTM_MEMTYPE_FLAG_FIXED)) + ghost_obj->ttm = NULL; + else + bo->ttm = NULL; + + ttm_bo_unreserve(ghost_obj); + ttm_bo_unref(&ghost_obj); + + } else if (from->flags & TTM_MEMTYPE_FLAG_FIXED) { + + /** + * BO doesn't have a TTM we need to bind/unbind. Just remember + * this eviction and free up the allocation + */ + + spin_lock(&from->move_lock); + if (!from->move || fence_is_later(fence, from->move)) { + fence_put(from->move); + from->move = fence_get(fence); + } + spin_unlock(&from->move_lock); + + ttm_bo_free_old_node(bo); + + fence_put(bo->moving); + bo->moving = fence_get(fence); + + } else { + /** + * Last resort, wait for the move to be completed. + * + * Should never happen in pratice. + */ + + ret = ttm_bo_wait(bo, false, false); + if (ret) + return ret; + + if (to->flags & TTM_MEMTYPE_FLAG_FIXED) { + ttm_tt_destroy(bo->ttm); + bo->ttm = NULL; + } + ttm_bo_free_old_node(bo); + } + + *old_mem = *new_mem; + new_mem->mm_node = NULL; + + return 0; +} +EXPORT_SYMBOL(ttm_bo_pipeline_move); diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c index 3216878bced3..a6ed9d5e5167 100644 --- a/drivers/gpu/drm/ttm/ttm_bo_vm.c +++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c @@ -48,15 +48,14 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo, { int ret = 0; - if (likely(!test_bit(TTM_BO_PRIV_FLAG_MOVING, &bo->priv_flags))) + if (likely(!bo->moving)) goto out_unlock; /* * Quick non-stalling check for idle. */ - ret = ttm_bo_wait(bo, false, true); - if (likely(ret == 0)) - goto out_unlock; + if (fence_is_signaled(bo->moving)) + goto out_clear; /* * If possible, avoid waiting for GPU with mmap_sem @@ -68,17 +67,23 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo, goto out_unlock; up_read(&vma->vm_mm->mmap_sem); - (void) ttm_bo_wait(bo, true, false); + (void) fence_wait(bo->moving, true); goto out_unlock; } /* * Ordinary wait. */ - ret = ttm_bo_wait(bo, true, false); - if (unlikely(ret != 0)) + ret = fence_wait(bo->moving, true); + if (unlikely(ret != 0)) { ret = (ret != -ERESTARTSYS) ? VM_FAULT_SIGBUS : VM_FAULT_NOPAGE; + goto out_unlock; + } + +out_clear: + fence_put(bo->moving); + bo->moving = NULL; out_unlock: return ret; diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c index 077ae9b2865d..d28d4333dcce 100644 --- a/drivers/gpu/drm/ttm/ttm_tt.c +++ b/drivers/gpu/drm/ttm/ttm_tt.c @@ -166,11 +166,15 @@ EXPORT_SYMBOL(ttm_tt_set_placement_caching); void ttm_tt_destroy(struct ttm_tt *ttm) { - if (unlikely(ttm == NULL)) + int ret; + + if (ttm == NULL) return; if (ttm->state == tt_bound) { - ttm_tt_unbind(ttm); + ret = ttm->func->unbind(ttm); + BUG_ON(ret); + ttm->state = tt_unbound; } if (ttm->state == tt_unbound) @@ -251,17 +255,6 @@ void ttm_dma_tt_fini(struct ttm_dma_tt *ttm_dma) } EXPORT_SYMBOL(ttm_dma_tt_fini); -void ttm_tt_unbind(struct ttm_tt *ttm) -{ - int ret; - - if (ttm->state == tt_bound) { - ret = ttm->func->unbind(ttm); - BUG_ON(ret); - ttm->state = tt_unbound; - } -} - int ttm_tt_bind(struct ttm_tt *ttm, struct ttm_mem_reg *bo_mem) { int ret = 0; diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index 4c0f26a644a3..c82d468d178b 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -449,14 +449,6 @@ static void vc4_crtc_atomic_flush(struct drm_crtc *crtc, WARN_ON_ONCE(dlist_next - dlist_start != vc4_state->mm.size); - HVS_WRITE(SCALER_DISPLISTX(vc4_crtc->channel), - vc4_state->mm.start); - - if (debug_dump_regs) { - DRM_INFO("CRTC %d HVS after:\n", drm_crtc_index(crtc)); - vc4_hvs_dump_state(dev); - } - if (crtc->state->event) { unsigned long flags; @@ -466,8 +458,20 @@ static void vc4_crtc_atomic_flush(struct drm_crtc *crtc, spin_lock_irqsave(&dev->event_lock, flags); vc4_crtc->event = crtc->state->event; - spin_unlock_irqrestore(&dev->event_lock, flags); crtc->state->event = NULL; + + HVS_WRITE(SCALER_DISPLISTX(vc4_crtc->channel), + vc4_state->mm.start); + + spin_unlock_irqrestore(&dev->event_lock, flags); + } else { + HVS_WRITE(SCALER_DISPLISTX(vc4_crtc->channel), + vc4_state->mm.start); + } + + if (debug_dump_regs) { + DRM_INFO("CRTC %d HVS after:\n", drm_crtc_index(crtc)); + vc4_hvs_dump_state(dev); } } @@ -493,12 +497,17 @@ static void vc4_crtc_handle_page_flip(struct vc4_crtc *vc4_crtc) { struct drm_crtc *crtc = &vc4_crtc->base; struct drm_device *dev = crtc->dev; + struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(crtc->state); + u32 chan = vc4_crtc->channel; unsigned long flags; spin_lock_irqsave(&dev->event_lock, flags); - if (vc4_crtc->event) { + if (vc4_crtc->event && + (vc4_state->mm.start == HVS_READ(SCALER_DISPLACTX(chan)))) { drm_crtc_send_vblank_event(crtc, vc4_crtc->event); vc4_crtc->event = NULL; + drm_crtc_vblank_put(crtc); } spin_unlock_irqrestore(&dev->event_lock, flags); } @@ -549,6 +558,7 @@ vc4_async_page_flip_complete(struct vc4_seqno_cb *cb) spin_unlock_irqrestore(&dev->event_lock, flags); } + drm_crtc_vblank_put(crtc); drm_framebuffer_unreference(flip_state->fb); kfree(flip_state); @@ -591,6 +601,8 @@ static int vc4_async_page_flip(struct drm_crtc *crtc, return ret; } + WARN_ON(drm_crtc_vblank_get(crtc) != 0); + /* Immediately update the plane's legacy fb pointer, so that later * modeset prep sees the state that will be present when the semaphore * is released. diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c index 9e88231b8906..54d0471243dd 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.c +++ b/drivers/gpu/drm/vc4/vc4_drv.c @@ -66,12 +66,12 @@ static const struct file_operations vc4_drm_fops = { }; static const struct drm_ioctl_desc vc4_drm_ioctls[] = { - DRM_IOCTL_DEF_DRV(VC4_SUBMIT_CL, vc4_submit_cl_ioctl, 0), - DRM_IOCTL_DEF_DRV(VC4_WAIT_SEQNO, vc4_wait_seqno_ioctl, 0), - DRM_IOCTL_DEF_DRV(VC4_WAIT_BO, vc4_wait_bo_ioctl, 0), - DRM_IOCTL_DEF_DRV(VC4_CREATE_BO, vc4_create_bo_ioctl, 0), - DRM_IOCTL_DEF_DRV(VC4_MMAP_BO, vc4_mmap_bo_ioctl, 0), - DRM_IOCTL_DEF_DRV(VC4_CREATE_SHADER_BO, vc4_create_shader_bo_ioctl, 0), + DRM_IOCTL_DEF_DRV(VC4_SUBMIT_CL, vc4_submit_cl_ioctl, DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(VC4_WAIT_SEQNO, vc4_wait_seqno_ioctl, DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(VC4_WAIT_BO, vc4_wait_bo_ioctl, DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(VC4_CREATE_BO, vc4_create_bo_ioctl, DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(VC4_MMAP_BO, vc4_mmap_bo_ioctl, DRM_RENDER_ALLOW), + DRM_IOCTL_DEF_DRV(VC4_CREATE_SHADER_BO, vc4_create_shader_bo_ioctl, DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(VC4_GET_HANG_STATE, vc4_get_hang_state_ioctl, DRM_ROOT_ONLY), }; @@ -91,7 +91,7 @@ static struct drm_driver vc4_drm_driver = { .enable_vblank = vc4_enable_vblank, .disable_vblank = vc4_disable_vblank, - .get_vblank_counter = drm_vblank_count, + .get_vblank_counter = drm_vblank_no_hw_counter, #if defined(CONFIG_DEBUG_FS) .debugfs_init = vc4_debugfs_init, diff --git a/drivers/gpu/drm/vc4/vc4_kms.c b/drivers/gpu/drm/vc4/vc4_kms.c index 8f4d5ffc32be..9a217fd025f3 100644 --- a/drivers/gpu/drm/vc4/vc4_kms.c +++ b/drivers/gpu/drm/vc4/vc4_kms.c @@ -119,10 +119,18 @@ static int vc4_atomic_commit(struct drm_device *dev, return -ENOMEM; /* Make sure that any outstanding modesets have finished. */ - ret = down_interruptible(&vc4->async_modeset); - if (ret) { - kfree(c); - return ret; + if (nonblock) { + ret = down_trylock(&vc4->async_modeset); + if (ret) { + kfree(c); + return -EBUSY; + } + } else { + ret = down_interruptible(&vc4->async_modeset); + if (ret) { + kfree(c); + return ret; + } } ret = drm_atomic_helper_prepare_planes(dev, state); diff --git a/drivers/gpu/drm/vc4/vc4_regs.h b/drivers/gpu/drm/vc4/vc4_regs.h index 6163b95c5411..f99eece4cc97 100644 --- a/drivers/gpu/drm/vc4/vc4_regs.h +++ b/drivers/gpu/drm/vc4/vc4_regs.h @@ -341,6 +341,10 @@ #define SCALER_DISPLACT0 0x00000030 #define SCALER_DISPLACT1 0x00000034 #define SCALER_DISPLACT2 0x00000038 +#define SCALER_DISPLACTX(x) (SCALER_DISPLACT0 + \ + (x) * (SCALER_DISPLACT1 - \ + SCALER_DISPLACT0)) + #define SCALER_DISPCTRL0 0x00000040 # define SCALER_DISPCTRLX_ENABLE BIT(31) # define SCALER_DISPCTRLX_RESET BIT(30) diff --git a/drivers/gpu/drm/virtio/virtgpu_ttm.c b/drivers/gpu/drm/virtio/virtgpu_ttm.c index a0580815629f..80482ac5f95d 100644 --- a/drivers/gpu/drm/virtio/virtgpu_ttm.c +++ b/drivers/gpu/drm/virtio/virtgpu_ttm.c @@ -375,6 +375,12 @@ static int virtio_gpu_bo_move(struct ttm_buffer_object *bo, bool no_wait_gpu, struct ttm_mem_reg *new_mem) { + int ret; + + ret = ttm_bo_wait(bo, interruptible, no_wait_gpu); + if (ret) + return ret; + virtio_gpu_move_null(bo, new_mem); return 0; } diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c index 6de283c8fa3e..f0374f9b56ca 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c @@ -28,6 +28,7 @@ #include <linux/slab.h> #include <linux/module.h> #include <linux/kernel.h> +#include <linux/frame.h> #include <asm/hypervisor.h> #include "drmP.h" #include "vmwgfx_msg.h" @@ -194,7 +195,7 @@ static int vmw_send_msg(struct rpc_channel *channel, const char *msg) return -EINVAL; } - +STACK_FRAME_NON_STANDARD(vmw_send_msg); /** @@ -304,6 +305,7 @@ static int vmw_recv_msg(struct rpc_channel *channel, void **msg, return 0; } +STACK_FRAME_NON_STANDARD(vmw_recv_msg); /** diff --git a/drivers/hid/hid-elo.c b/drivers/hid/hid-elo.c index aad8c162a825..0cd4f7216239 100644 --- a/drivers/hid/hid-elo.c +++ b/drivers/hid/hid-elo.c @@ -261,7 +261,7 @@ static void elo_remove(struct hid_device *hdev) struct elo_priv *priv = hid_get_drvdata(hdev); hid_hw_stop(hdev); - flush_workqueue(wq); + cancel_delayed_work_sync(&priv->work); kfree(priv); } diff --git a/drivers/hid/hid-multitouch.c b/drivers/hid/hid-multitouch.c index c741f5e50a66..95b7d61d9910 100644 --- a/drivers/hid/hid-multitouch.c +++ b/drivers/hid/hid-multitouch.c @@ -1401,6 +1401,11 @@ static const struct hid_device_id mt_devices[] = { MT_USB_DEVICE(USB_VENDOR_ID_NOVATEK, USB_DEVICE_ID_NOVATEK_PCT) }, + /* Ntrig Panel */ + { .driver_data = MT_CLS_NSMU, + HID_DEVICE(BUS_I2C, HID_GROUP_MULTITOUCH_WIN_8, + USB_VENDOR_ID_NTRIG, 0x1b05) }, + /* PixArt optical touch screen */ { .driver_data = MT_CLS_INRANGE_CONTACTNUMBER, MT_USB_DEVICE(USB_VENDOR_ID_PIXART, diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c index 2f1ddca6f2e0..700145b15088 100644 --- a/drivers/hid/usbhid/hiddev.c +++ b/drivers/hid/usbhid/hiddev.c @@ -516,13 +516,13 @@ static noinline int hiddev_ioctl_usage(struct hiddev *hiddev, unsigned int cmd, goto inval; } else if (uref->usage_index >= field->report_count) goto inval; - - else if ((cmd == HIDIOCGUSAGES || cmd == HIDIOCSUSAGES) && - (uref_multi->num_values > HID_MAX_MULTI_USAGES || - uref->usage_index + uref_multi->num_values > field->report_count)) - goto inval; } + if ((cmd == HIDIOCGUSAGES || cmd == HIDIOCSUSAGES) && + (uref_multi->num_values > HID_MAX_MULTI_USAGES || + uref->usage_index + uref_multi->num_values > field->report_count)) + goto inval; + switch (cmd) { case HIDIOCGUSAGE: uref->value = field->value[uref->usage_index]; diff --git a/drivers/hwmon/dell-smm-hwmon.c b/drivers/hwmon/dell-smm-hwmon.c index c43318d3416e..2ac87d553e22 100644 --- a/drivers/hwmon/dell-smm-hwmon.c +++ b/drivers/hwmon/dell-smm-hwmon.c @@ -35,6 +35,7 @@ #include <linux/uaccess.h> #include <linux/io.h> #include <linux/sched.h> +#include <linux/ctype.h> #include <linux/i8k.h> @@ -66,11 +67,13 @@ static DEFINE_MUTEX(i8k_mutex); static char bios_version[4]; +static char bios_machineid[16]; static struct device *i8k_hwmon_dev; static u32 i8k_hwmon_flags; static uint i8k_fan_mult = I8K_FAN_MULT; static uint i8k_pwm_mult; static uint i8k_fan_max = I8K_FAN_HIGH; +static bool disallow_fan_type_call; #define I8K_HWMON_HAVE_TEMP1 (1 << 0) #define I8K_HWMON_HAVE_TEMP2 (1 << 1) @@ -94,13 +97,13 @@ module_param(ignore_dmi, bool, 0); MODULE_PARM_DESC(ignore_dmi, "Continue probing hardware even if DMI data does not match"); #if IS_ENABLED(CONFIG_I8K) -static bool restricted; +static bool restricted = true; module_param(restricted, bool, 0); -MODULE_PARM_DESC(restricted, "Allow fan control if SYS_ADMIN capability set"); +MODULE_PARM_DESC(restricted, "Restrict fan control and serial number to CAP_SYS_ADMIN (default: 1)"); static bool power_status; module_param(power_status, bool, 0600); -MODULE_PARM_DESC(power_status, "Report power status in /proc/i8k"); +MODULE_PARM_DESC(power_status, "Report power status in /proc/i8k (default: 0)"); #endif static uint fan_mult; @@ -235,14 +238,28 @@ static int i8k_get_fan_speed(int fan) /* * Read the fan type. */ -static int i8k_get_fan_type(int fan) +static int _i8k_get_fan_type(int fan) { struct smm_regs regs = { .eax = I8K_SMM_GET_FAN_TYPE, }; + if (disallow_fan_type_call) + return -EINVAL; + regs.ebx = fan & 0xff; return i8k_smm(®s) ? : regs.eax & 0xff; } +static int i8k_get_fan_type(int fan) +{ + /* I8K_SMM_GET_FAN_TYPE SMM call is expensive, so cache values */ + static int types[2] = { INT_MIN, INT_MIN }; + + if (types[fan] == INT_MIN) + types[fan] = _i8k_get_fan_type(fan); + + return types[fan]; +} + /* * Read the fan nominal rpm for specific fan speed. */ @@ -387,14 +404,20 @@ i8k_ioctl_unlocked(struct file *fp, unsigned int cmd, unsigned long arg) switch (cmd) { case I8K_BIOS_VERSION: + if (!isdigit(bios_version[0]) || !isdigit(bios_version[1]) || + !isdigit(bios_version[2])) + return -EINVAL; + val = (bios_version[0] << 16) | (bios_version[1] << 8) | bios_version[2]; break; case I8K_MACHINE_ID: - memset(buff, 0, 16); - strlcpy(buff, i8k_get_dmi_data(DMI_PRODUCT_SERIAL), - sizeof(buff)); + if (restricted && !capable(CAP_SYS_ADMIN)) + return -EPERM; + + memset(buff, 0, sizeof(buff)); + strlcpy(buff, bios_machineid, sizeof(buff)); break; case I8K_FN_STATUS: @@ -511,7 +534,7 @@ static int i8k_proc_show(struct seq_file *seq, void *offset) seq_printf(seq, "%s %s %s %d %d %d %d %d %d %d\n", I8K_PROC_FMT, bios_version, - i8k_get_dmi_data(DMI_PRODUCT_SERIAL), + (restricted && !capable(CAP_SYS_ADMIN)) ? "-1" : bios_machineid, cpu_temp, left_fan, right_fan, left_speed, right_speed, ac_power, fn_key); @@ -718,6 +741,9 @@ static struct attribute *i8k_attrs[] = { static umode_t i8k_is_visible(struct kobject *kobj, struct attribute *attr, int index) { + if (disallow_fan_type_call && + (index == 9 || index == 12)) + return 0; if (index >= 0 && index <= 1 && !(i8k_hwmon_flags & I8K_HWMON_HAVE_TEMP1)) return 0; @@ -767,13 +793,17 @@ static int __init i8k_init_hwmon(void) if (err >= 0) i8k_hwmon_flags |= I8K_HWMON_HAVE_TEMP4; - /* First fan attributes, if fan type is OK */ - err = i8k_get_fan_type(0); + /* First fan attributes, if fan status or type is OK */ + err = i8k_get_fan_status(0); + if (err < 0) + err = i8k_get_fan_type(0); if (err >= 0) i8k_hwmon_flags |= I8K_HWMON_HAVE_FAN1; - /* Second fan attributes, if fan type is OK */ - err = i8k_get_fan_type(1); + /* Second fan attributes, if fan status or type is OK */ + err = i8k_get_fan_status(1); + if (err < 0) + err = i8k_get_fan_type(1); if (err >= 0) i8k_hwmon_flags |= I8K_HWMON_HAVE_FAN2; @@ -929,12 +959,14 @@ static struct dmi_system_id i8k_dmi_table[] __initdata = { MODULE_DEVICE_TABLE(dmi, i8k_dmi_table); -static struct dmi_system_id i8k_blacklist_dmi_table[] __initdata = { +/* + * On some machines once I8K_SMM_GET_FAN_TYPE is issued then CPU fan speed + * randomly going up and down due to bug in Dell SMM or BIOS. Here is blacklist + * of affected Dell machines for which we disallow I8K_SMM_GET_FAN_TYPE call. + * See bug: https://bugzilla.kernel.org/show_bug.cgi?id=100121 + */ +static struct dmi_system_id i8k_blacklist_fan_type_dmi_table[] __initdata = { { - /* - * CPU fan speed going up and down on Dell Studio XPS 8000 - * for unknown reasons. - */ .ident = "Dell Studio XPS 8000", .matches = { DMI_EXACT_MATCH(DMI_SYS_VENDOR, "Dell Inc."), @@ -942,16 +974,19 @@ static struct dmi_system_id i8k_blacklist_dmi_table[] __initdata = { }, }, { - /* - * CPU fan speed going up and down on Dell Studio XPS 8100 - * for unknown reasons. - */ .ident = "Dell Studio XPS 8100", .matches = { DMI_EXACT_MATCH(DMI_SYS_VENDOR, "Dell Inc."), DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "Studio XPS 8100"), }, }, + { + .ident = "Dell Inspiron 580", + .matches = { + DMI_EXACT_MATCH(DMI_SYS_VENDOR, "Dell Inc."), + DMI_EXACT_MATCH(DMI_PRODUCT_NAME, "Inspiron 580 "), + }, + }, { } }; @@ -966,8 +1001,7 @@ static int __init i8k_probe(void) /* * Get DMI information */ - if (!dmi_check_system(i8k_dmi_table) || - dmi_check_system(i8k_blacklist_dmi_table)) { + if (!dmi_check_system(i8k_dmi_table)) { if (!ignore_dmi && !force) return -ENODEV; @@ -978,8 +1012,13 @@ static int __init i8k_probe(void) i8k_get_dmi_data(DMI_BIOS_VERSION)); } + if (dmi_check_system(i8k_blacklist_fan_type_dmi_table)) + disallow_fan_type_call = true; + strlcpy(bios_version, i8k_get_dmi_data(DMI_BIOS_VERSION), sizeof(bios_version)); + strlcpy(bios_machineid, i8k_get_dmi_data(DMI_PRODUCT_SERIAL), + sizeof(bios_machineid)); /* * Get SMM Dell signature diff --git a/drivers/hwmon/fam15h_power.c b/drivers/hwmon/fam15h_power.c index eb97a9241d17..15aa49d082c4 100644 --- a/drivers/hwmon/fam15h_power.c +++ b/drivers/hwmon/fam15h_power.c @@ -172,9 +172,9 @@ static void do_read_registers_on_cu(void *_data) */ static int read_registers(struct fam15h_power_data *data) { - int this_cpu, ret, cpu; int core, this_core; cpumask_var_t mask; + int ret, cpu; ret = zalloc_cpumask_var(&mask, GFP_KERNEL); if (!ret) @@ -183,7 +183,6 @@ static int read_registers(struct fam15h_power_data *data) memset(data->cu_on, 0, sizeof(int) * MAX_CUS); get_online_cpus(); - this_cpu = smp_processor_id(); /* * Choose the first online core of each compute unit, and then @@ -205,12 +204,9 @@ static int read_registers(struct fam15h_power_data *data) cpumask_set_cpu(cpumask_any(topology_sibling_cpumask(cpu)), mask); } - if (cpumask_test_cpu(this_cpu, mask)) - do_read_registers_on_cu(data); + on_each_cpu_mask(mask, do_read_registers_on_cu, data, true); - smp_call_function_many(mask, do_read_registers_on_cu, data, true); put_online_cpus(); - free_cpumask_var(mask); return 0; diff --git a/drivers/hwmon/lm90.c b/drivers/hwmon/lm90.c index c9ff08dbe10c..e30a5939dc0d 100644 --- a/drivers/hwmon/lm90.c +++ b/drivers/hwmon/lm90.c @@ -375,7 +375,7 @@ struct lm90_data { int kind; u32 flags; - int update_interval; /* in milliseconds */ + unsigned int update_interval; /* in milliseconds */ u8 config_orig; /* Original configuration register value */ u8 convrate_orig; /* Original conversion rate register value */ diff --git a/drivers/hwtracing/coresight/coresight-tmc-etr.c b/drivers/hwtracing/coresight/coresight-tmc-etr.c index 847d1b5f2c13..688be9e060fc 100644 --- a/drivers/hwtracing/coresight/coresight-tmc-etr.c +++ b/drivers/hwtracing/coresight/coresight-tmc-etr.c @@ -300,13 +300,10 @@ int tmc_read_unprepare_etr(struct tmc_drvdata *drvdata) if (local_read(&drvdata->mode) == CS_MODE_SYSFS) { /* * The trace run will continue with the same allocated trace - * buffer. As such zero-out the buffer so that we don't end - * up with stale data. - * - * Since the tracer is still enabled drvdata::buf - * can't be NULL. + * buffer. The trace buffer is cleared in tmc_etr_enable_hw(), + * so we don't have to explicitly clear it. Also, since the + * tracer is still enabled drvdata::buf can't be NULL. */ - memset(drvdata->buf, 0, drvdata->size); tmc_etr_enable_hw(drvdata); } else { /* @@ -315,7 +312,7 @@ int tmc_read_unprepare_etr(struct tmc_drvdata *drvdata) */ vaddr = drvdata->vaddr; paddr = drvdata->paddr; - drvdata->buf = NULL; + drvdata->buf = drvdata->vaddr = NULL; } drvdata->reading = false; diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c index 5443d03a1eec..d08d1ab9bba5 100644 --- a/drivers/hwtracing/coresight/coresight.c +++ b/drivers/hwtracing/coresight/coresight.c @@ -385,7 +385,6 @@ static int _coresight_build_path(struct coresight_device *csdev, int i; bool found = false; struct coresight_node *node; - struct coresight_connection *conn; /* An activated sink has been found. Enqueue the element */ if ((csdev->type == CORESIGHT_DEV_TYPE_SINK || @@ -394,8 +393,9 @@ static int _coresight_build_path(struct coresight_device *csdev, /* Not a sink - recursively explore each port found on this element */ for (i = 0; i < csdev->nr_outport; i++) { - conn = &csdev->conns[i]; - if (_coresight_build_path(conn->child_dev, path) == 0) { + struct coresight_device *child_dev = csdev->conns[i].child_dev; + + if (child_dev && _coresight_build_path(child_dev, path) == 0) { found = true; break; } @@ -425,6 +425,7 @@ out: struct list_head *coresight_build_path(struct coresight_device *csdev) { struct list_head *path; + int rc; path = kzalloc(sizeof(struct list_head), GFP_KERNEL); if (!path) @@ -432,9 +433,10 @@ struct list_head *coresight_build_path(struct coresight_device *csdev) INIT_LIST_HEAD(path); - if (_coresight_build_path(csdev, path)) { + rc = _coresight_build_path(csdev, path); + if (rc) { kfree(path); - path = NULL; + return ERR_PTR(rc); } return path; @@ -507,8 +509,9 @@ int coresight_enable(struct coresight_device *csdev) goto out; path = coresight_build_path(csdev); - if (!path) { + if (IS_ERR(path)) { pr_err("building path(s) failed\n"); + ret = PTR_ERR(path); goto out; } diff --git a/drivers/i2c/busses/i2c-i801.c b/drivers/i2c/busses/i2c-i801.c index 64b1208bca5e..4a60ad214747 100644 --- a/drivers/i2c/busses/i2c-i801.c +++ b/drivers/i2c/busses/i2c-i801.c @@ -245,6 +245,13 @@ struct i801_priv { struct platform_device *mux_pdev; #endif struct platform_device *tco_pdev; + + /* + * If set to true the host controller registers are reserved for + * ACPI AML use. Protected by acpi_lock. + */ + bool acpi_reserved; + struct mutex acpi_lock; }; #define FEATURE_SMBUS_PEC (1 << 0) @@ -718,6 +725,12 @@ static s32 i801_access(struct i2c_adapter *adap, u16 addr, int ret = 0, xact = 0; struct i801_priv *priv = i2c_get_adapdata(adap); + mutex_lock(&priv->acpi_lock); + if (priv->acpi_reserved) { + mutex_unlock(&priv->acpi_lock); + return -EBUSY; + } + pm_runtime_get_sync(&priv->pci_dev->dev); hwpec = (priv->features & FEATURE_SMBUS_PEC) && (flags & I2C_CLIENT_PEC) @@ -820,6 +833,7 @@ static s32 i801_access(struct i2c_adapter *adap, u16 addr, out: pm_runtime_mark_last_busy(&priv->pci_dev->dev); pm_runtime_put_autosuspend(&priv->pci_dev->dev); + mutex_unlock(&priv->acpi_lock); return ret; } @@ -1257,6 +1271,83 @@ static void i801_add_tco(struct i801_priv *priv) priv->tco_pdev = pdev; } +#ifdef CONFIG_ACPI +static acpi_status +i801_acpi_io_handler(u32 function, acpi_physical_address address, u32 bits, + u64 *value, void *handler_context, void *region_context) +{ + struct i801_priv *priv = handler_context; + struct pci_dev *pdev = priv->pci_dev; + acpi_status status; + + /* + * Once BIOS AML code touches the OpRegion we warn and inhibit any + * further access from the driver itself. This device is now owned + * by the system firmware. + */ + mutex_lock(&priv->acpi_lock); + + if (!priv->acpi_reserved) { + priv->acpi_reserved = true; + + dev_warn(&pdev->dev, "BIOS is accessing SMBus registers\n"); + dev_warn(&pdev->dev, "Driver SMBus register access inhibited\n"); + + /* + * BIOS is accessing the host controller so prevent it from + * suspending automatically from now on. + */ + pm_runtime_get_sync(&pdev->dev); + } + + if ((function & ACPI_IO_MASK) == ACPI_READ) + status = acpi_os_read_port(address, (u32 *)value, bits); + else + status = acpi_os_write_port(address, (u32)*value, bits); + + mutex_unlock(&priv->acpi_lock); + + return status; +} + +static int i801_acpi_probe(struct i801_priv *priv) +{ + struct acpi_device *adev; + acpi_status status; + + adev = ACPI_COMPANION(&priv->pci_dev->dev); + if (adev) { + status = acpi_install_address_space_handler(adev->handle, + ACPI_ADR_SPACE_SYSTEM_IO, i801_acpi_io_handler, + NULL, priv); + if (ACPI_SUCCESS(status)) + return 0; + } + + return acpi_check_resource_conflict(&priv->pci_dev->resource[SMBBAR]); +} + +static void i801_acpi_remove(struct i801_priv *priv) +{ + struct acpi_device *adev; + + adev = ACPI_COMPANION(&priv->pci_dev->dev); + if (!adev) + return; + + acpi_remove_address_space_handler(adev->handle, + ACPI_ADR_SPACE_SYSTEM_IO, i801_acpi_io_handler); + + mutex_lock(&priv->acpi_lock); + if (priv->acpi_reserved) + pm_runtime_put(&priv->pci_dev->dev); + mutex_unlock(&priv->acpi_lock); +} +#else +static inline int i801_acpi_probe(struct i801_priv *priv) { return 0; } +static inline void i801_acpi_remove(struct i801_priv *priv) { } +#endif + static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id) { unsigned char temp; @@ -1274,6 +1365,7 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id) priv->adapter.dev.parent = &dev->dev; ACPI_COMPANION_SET(&priv->adapter.dev, ACPI_COMPANION(&dev->dev)); priv->adapter.retries = 3; + mutex_init(&priv->acpi_lock); priv->pci_dev = dev; switch (dev->device) { @@ -1336,10 +1428,8 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id) return -ENODEV; } - err = acpi_check_resource_conflict(&dev->resource[SMBBAR]); - if (err) { + if (i801_acpi_probe(priv)) return -ENODEV; - } err = pcim_iomap_regions(dev, 1 << SMBBAR, dev_driver_string(&dev->dev)); @@ -1348,6 +1438,7 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id) "Failed to request SMBus region 0x%lx-0x%Lx\n", priv->smba, (unsigned long long)pci_resource_end(dev, SMBBAR)); + i801_acpi_remove(priv); return err; } @@ -1412,6 +1503,7 @@ static int i801_probe(struct pci_dev *dev, const struct pci_device_id *id) err = i2c_add_adapter(&priv->adapter); if (err) { dev_err(&dev->dev, "Failed to add SMBus adapter\n"); + i801_acpi_remove(priv); return err; } @@ -1438,6 +1530,7 @@ static void i801_remove(struct pci_dev *dev) i801_del_mux(priv); i2c_del_adapter(&priv->adapter); + i801_acpi_remove(priv); pci_write_config_byte(dev, SMBHSTCFG, priv->original_hstcfg); platform_device_unregister(priv->tco_pdev); diff --git a/drivers/i2c/busses/i2c-octeon.c b/drivers/i2c/busses/i2c-octeon.c index aa5f01efd826..30ae35146723 100644 --- a/drivers/i2c/busses/i2c-octeon.c +++ b/drivers/i2c/busses/i2c-octeon.c @@ -934,8 +934,15 @@ static int octeon_i2c_read(struct octeon_i2c *i2c, int target, return result; for (i = 0; i < length; i++) { - /* for the last byte TWSI_CTL_AAK must not be set */ - if (i + 1 == length) + /* + * For the last byte to receive TWSI_CTL_AAK must not be set. + * + * A special case is I2C_M_RECV_LEN where we don't know the + * additional length yet. If recv_len is set we assume we're + * not reading the final byte and therefore need to set + * TWSI_CTL_AAK. + */ + if ((i + 1 == length) && !(recv_len && i == 0)) final_read = true; /* clear iflg to allow next event */ @@ -950,12 +957,8 @@ static int octeon_i2c_read(struct octeon_i2c *i2c, int target, data[i] = octeon_i2c_data_read(i2c); if (recv_len && i == 0) { - if (data[i] > I2C_SMBUS_BLOCK_MAX + 1) { - dev_err(i2c->dev, - "%s: read len > I2C_SMBUS_BLOCK_MAX %d\n", - __func__, data[i]); + if (data[i] > I2C_SMBUS_BLOCK_MAX + 1) return -EPROTO; - } length += data[i]; } diff --git a/drivers/i2c/muxes/i2c-mux-reg.c b/drivers/i2c/muxes/i2c-mux-reg.c index 6773cadf7c9f..26e7c5187a58 100644 --- a/drivers/i2c/muxes/i2c-mux-reg.c +++ b/drivers/i2c/muxes/i2c-mux-reg.c @@ -260,6 +260,7 @@ static struct platform_driver i2c_mux_reg_driver = { .remove = i2c_mux_reg_remove, .driver = { .name = "i2c-mux-reg", + .of_match_table = of_match_ptr(i2c_mux_reg_of_match), }, }; diff --git a/drivers/iio/accel/st_accel_buffer.c b/drivers/iio/accel/st_accel_buffer.c index a1e642ee13d6..7fddc137e91e 100644 --- a/drivers/iio/accel/st_accel_buffer.c +++ b/drivers/iio/accel/st_accel_buffer.c @@ -91,7 +91,7 @@ static const struct iio_buffer_setup_ops st_accel_buffer_setup_ops = { int st_accel_allocate_ring(struct iio_dev *indio_dev) { - return iio_triggered_buffer_setup(indio_dev, &iio_pollfunc_store_time, + return iio_triggered_buffer_setup(indio_dev, NULL, &st_sensors_trigger_handler, &st_accel_buffer_setup_ops); } diff --git a/drivers/iio/accel/st_accel_core.c b/drivers/iio/accel/st_accel_core.c index dc73f2d85e6d..4d95bfc4786c 100644 --- a/drivers/iio/accel/st_accel_core.c +++ b/drivers/iio/accel/st_accel_core.c @@ -741,6 +741,7 @@ static const struct iio_info accel_info = { static const struct iio_trigger_ops st_accel_trigger_ops = { .owner = THIS_MODULE, .set_trigger_state = ST_ACCEL_TRIGGER_SET_STATE, + .validate_device = st_sensors_validate_device, }; #define ST_ACCEL_TRIGGER_OPS (&st_accel_trigger_ops) #else diff --git a/drivers/iio/common/st_sensors/st_sensors_buffer.c b/drivers/iio/common/st_sensors/st_sensors_buffer.c index c55898543a47..f1693dbebb8a 100644 --- a/drivers/iio/common/st_sensors/st_sensors_buffer.c +++ b/drivers/iio/common/st_sensors/st_sensors_buffer.c @@ -57,31 +57,20 @@ irqreturn_t st_sensors_trigger_handler(int irq, void *p) struct iio_poll_func *pf = p; struct iio_dev *indio_dev = pf->indio_dev; struct st_sensor_data *sdata = iio_priv(indio_dev); + s64 timestamp; - /* If we have a status register, check if this IRQ came from us */ - if (sdata->sensor_settings->drdy_irq.addr_stat_drdy) { - u8 status; - - len = sdata->tf->read_byte(&sdata->tb, sdata->dev, - sdata->sensor_settings->drdy_irq.addr_stat_drdy, - &status); - if (len < 0) - dev_err(sdata->dev, "could not read channel status\n"); - - /* - * If this was not caused by any channels on this sensor, - * return IRQ_NONE - */ - if (!(status & (u8)indio_dev->active_scan_mask[0])) - return IRQ_NONE; - } + /* If we do timetamping here, do it before reading the values */ + if (sdata->hw_irq_trigger) + timestamp = sdata->hw_timestamp; + else + timestamp = iio_get_time_ns(); len = st_sensors_get_buffer_element(indio_dev, sdata->buffer_data); if (len < 0) goto st_sensors_get_buffer_element_error; iio_push_to_buffers_with_timestamp(indio_dev, sdata->buffer_data, - pf->timestamp); + timestamp); st_sensors_get_buffer_element_error: iio_trigger_notify_done(indio_dev->trig); diff --git a/drivers/iio/common/st_sensors/st_sensors_core.c b/drivers/iio/common/st_sensors/st_sensors_core.c index dffe00692169..9e59c90f6a8d 100644 --- a/drivers/iio/common/st_sensors/st_sensors_core.c +++ b/drivers/iio/common/st_sensors/st_sensors_core.c @@ -363,6 +363,11 @@ int st_sensors_init_sensor(struct iio_dev *indio_dev, if (err < 0) return err; + /* Disable DRDY, this might be still be enabled after reboot. */ + err = st_sensors_set_dataready_irq(indio_dev, false); + if (err < 0) + return err; + if (sdata->current_fullscale) { err = st_sensors_set_fullscale(indio_dev, sdata->current_fullscale->num); @@ -424,6 +429,9 @@ int st_sensors_set_dataready_irq(struct iio_dev *indio_dev, bool enable) else drdy_mask = sdata->sensor_settings->drdy_irq.mask_int2; + /* Flag to the poll function that the hardware trigger is in use */ + sdata->hw_irq_trigger = enable; + /* Enable/Disable the interrupt generator for data ready. */ err = st_sensors_write_data_with_mask(indio_dev, sdata->sensor_settings->drdy_irq.addr, diff --git a/drivers/iio/common/st_sensors/st_sensors_trigger.c b/drivers/iio/common/st_sensors/st_sensors_trigger.c index da72279fcf99..296e4ff19ae8 100644 --- a/drivers/iio/common/st_sensors/st_sensors_trigger.c +++ b/drivers/iio/common/st_sensors/st_sensors_trigger.c @@ -17,6 +17,73 @@ #include <linux/iio/common/st_sensors.h> #include "st_sensors_core.h" +/** + * st_sensors_irq_handler() - top half of the IRQ-based triggers + * @irq: irq number + * @p: private handler data + */ +irqreturn_t st_sensors_irq_handler(int irq, void *p) +{ + struct iio_trigger *trig = p; + struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig); + struct st_sensor_data *sdata = iio_priv(indio_dev); + + /* Get the time stamp as close in time as possible */ + sdata->hw_timestamp = iio_get_time_ns(); + return IRQ_WAKE_THREAD; +} + +/** + * st_sensors_irq_thread() - bottom half of the IRQ-based triggers + * @irq: irq number + * @p: private handler data + */ +irqreturn_t st_sensors_irq_thread(int irq, void *p) +{ + struct iio_trigger *trig = p; + struct iio_dev *indio_dev = iio_trigger_get_drvdata(trig); + struct st_sensor_data *sdata = iio_priv(indio_dev); + int ret; + + /* + * If this trigger is backed by a hardware interrupt and we have a + * status register, check if this IRQ came from us + */ + if (sdata->sensor_settings->drdy_irq.addr_stat_drdy) { + u8 status; + + ret = sdata->tf->read_byte(&sdata->tb, sdata->dev, + sdata->sensor_settings->drdy_irq.addr_stat_drdy, + &status); + if (ret < 0) { + dev_err(sdata->dev, "could not read channel status\n"); + goto out_poll; + } + /* + * the lower bits of .active_scan_mask[0] is directly mapped + * to the channels on the sensor: either bit 0 for + * one-dimensional sensors, or e.g. x,y,z for accelerometers, + * gyroscopes or magnetometers. No sensor use more than 3 + * channels, so cut the other status bits here. + */ + status &= 0x07; + + /* + * If this was not caused by any channels on this sensor, + * return IRQ_NONE + */ + if (!indio_dev->active_scan_mask) + return IRQ_NONE; + if (!(status & (u8)indio_dev->active_scan_mask[0])) + return IRQ_NONE; + } + +out_poll: + /* It's our IRQ: proceed to handle the register polling */ + iio_trigger_poll_chained(p); + return IRQ_HANDLED; +} + int st_sensors_allocate_trigger(struct iio_dev *indio_dev, const struct iio_trigger_ops *trigger_ops) { @@ -30,6 +97,10 @@ int st_sensors_allocate_trigger(struct iio_dev *indio_dev, return -ENOMEM; } + iio_trigger_set_drvdata(sdata->trig, indio_dev); + sdata->trig->ops = trigger_ops; + sdata->trig->dev.parent = sdata->dev; + irq = sdata->get_irq_data_ready(indio_dev); irq_trig = irqd_get_trigger_type(irq_get_irq_data(irq)); /* @@ -77,9 +148,12 @@ int st_sensors_allocate_trigger(struct iio_dev *indio_dev, sdata->sensor_settings->drdy_irq.addr_stat_drdy) irq_trig |= IRQF_SHARED; - err = request_threaded_irq(irq, - iio_trigger_generic_data_rdy_poll, - NULL, + /* Let's create an interrupt thread masking the hard IRQ here */ + irq_trig |= IRQF_ONESHOT; + + err = request_threaded_irq(sdata->get_irq_data_ready(indio_dev), + st_sensors_irq_handler, + st_sensors_irq_thread, irq_trig, sdata->trig->name, sdata->trig); @@ -88,10 +162,6 @@ int st_sensors_allocate_trigger(struct iio_dev *indio_dev, goto iio_trigger_free; } - iio_trigger_set_drvdata(sdata->trig, indio_dev); - sdata->trig->ops = trigger_ops; - sdata->trig->dev.parent = sdata->dev; - err = iio_trigger_register(sdata->trig); if (err < 0) { dev_err(&indio_dev->dev, "failed to register iio trigger.\n"); @@ -119,6 +189,18 @@ void st_sensors_deallocate_trigger(struct iio_dev *indio_dev) } EXPORT_SYMBOL(st_sensors_deallocate_trigger); +int st_sensors_validate_device(struct iio_trigger *trig, + struct iio_dev *indio_dev) +{ + struct iio_dev *indio = iio_trigger_get_drvdata(trig); + + if (indio != indio_dev) + return -EINVAL; + + return 0; +} +EXPORT_SYMBOL(st_sensors_validate_device); + MODULE_AUTHOR("Denis Ciocca <denis.ciocca@st.com>"); MODULE_DESCRIPTION("STMicroelectronics ST-sensors trigger"); MODULE_LICENSE("GPL v2"); diff --git a/drivers/iio/dac/Kconfig b/drivers/iio/dac/Kconfig index e63b957c985f..f7c71da42f15 100644 --- a/drivers/iio/dac/Kconfig +++ b/drivers/iio/dac/Kconfig @@ -247,7 +247,7 @@ config MCP4922 config STX104 tristate "Apex Embedded Systems STX104 DAC driver" - depends on X86 && ISA + depends on X86 && ISA_BUS_API help Say yes here to build support for the 2-channel DAC on the Apex Embedded Systems STX104 integrated analog PC/104 card. The base port diff --git a/drivers/iio/dac/ad5592r-base.c b/drivers/iio/dac/ad5592r-base.c index 948f600e7059..69bde5909854 100644 --- a/drivers/iio/dac/ad5592r-base.c +++ b/drivers/iio/dac/ad5592r-base.c @@ -525,7 +525,7 @@ static int ad5592r_alloc_channels(struct ad5592r_state *st) device_for_each_child_node(st->dev, child) { ret = fwnode_property_read_u32(child, "reg", ®); - if (ret || reg > ARRAY_SIZE(st->channel_modes)) + if (ret || reg >= ARRAY_SIZE(st->channel_modes)) continue; ret = fwnode_property_read_u32(child, "adi,mode", &tmp); diff --git a/drivers/iio/gyro/st_gyro_buffer.c b/drivers/iio/gyro/st_gyro_buffer.c index d67b17b6a7aa..a5377044e42f 100644 --- a/drivers/iio/gyro/st_gyro_buffer.c +++ b/drivers/iio/gyro/st_gyro_buffer.c @@ -91,7 +91,7 @@ static const struct iio_buffer_setup_ops st_gyro_buffer_setup_ops = { int st_gyro_allocate_ring(struct iio_dev *indio_dev) { - return iio_triggered_buffer_setup(indio_dev, &iio_pollfunc_store_time, + return iio_triggered_buffer_setup(indio_dev, NULL, &st_sensors_trigger_handler, &st_gyro_buffer_setup_ops); } diff --git a/drivers/iio/gyro/st_gyro_core.c b/drivers/iio/gyro/st_gyro_core.c index 52a3c87c375c..a8012955a1f6 100644 --- a/drivers/iio/gyro/st_gyro_core.c +++ b/drivers/iio/gyro/st_gyro_core.c @@ -409,6 +409,7 @@ static const struct iio_info gyro_info = { static const struct iio_trigger_ops st_gyro_trigger_ops = { .owner = THIS_MODULE, .set_trigger_state = ST_GYRO_TRIGGER_SET_STATE, + .validate_device = st_sensors_validate_device, }; #define ST_GYRO_TRIGGER_OPS (&st_gyro_trigger_ops) #else diff --git a/drivers/iio/humidity/am2315.c b/drivers/iio/humidity/am2315.c index 3be6d209a159..11535911a5c6 100644 --- a/drivers/iio/humidity/am2315.c +++ b/drivers/iio/humidity/am2315.c @@ -165,10 +165,8 @@ static irqreturn_t am2315_trigger_handler(int irq, void *p) struct am2315_sensor_data sensor_data; ret = am2315_read_data(data, &sensor_data); - if (ret < 0) { - mutex_unlock(&data->lock); + if (ret < 0) goto err; - } mutex_lock(&data->lock); if (*(indio_dev->active_scan_mask) == AM2315_ALL_CHANNEL_MASK) { diff --git a/drivers/iio/humidity/hdc100x.c b/drivers/iio/humidity/hdc100x.c index fa4767613173..a03832a5fc95 100644 --- a/drivers/iio/humidity/hdc100x.c +++ b/drivers/iio/humidity/hdc100x.c @@ -55,7 +55,7 @@ static const struct { }, { /* IIO_HUMIDITYRELATIVE channel */ .shift = 8, - .mask = 2, + .mask = 3, }, }; @@ -164,14 +164,14 @@ static int hdc100x_get_measurement(struct hdc100x_data *data, dev_err(&client->dev, "cannot read high byte measurement"); return ret; } - val = ret << 6; + val = ret << 8; ret = i2c_smbus_read_byte(client); if (ret < 0) { dev_err(&client->dev, "cannot read low byte measurement"); return ret; } - val |= ret >> 2; + val |= ret; return val; } @@ -211,18 +211,18 @@ static int hdc100x_read_raw(struct iio_dev *indio_dev, return IIO_VAL_INT_PLUS_MICRO; case IIO_CHAN_INFO_SCALE: if (chan->type == IIO_TEMP) { - *val = 165; - *val2 = 65536 >> 2; + *val = 165000; + *val2 = 65536; return IIO_VAL_FRACTIONAL; } else { - *val = 0; - *val2 = 10000; - return IIO_VAL_INT_PLUS_MICRO; + *val = 100; + *val2 = 65536; + return IIO_VAL_FRACTIONAL; } break; case IIO_CHAN_INFO_OFFSET: - *val = -3971; - *val2 = 879096; + *val = -15887; + *val2 = 515151; return IIO_VAL_INT_PLUS_MICRO; default: return -EINVAL; diff --git a/drivers/iio/imu/bmi160/bmi160_core.c b/drivers/iio/imu/bmi160/bmi160_core.c index 0bf92b06d7d8..b8a290ec984e 100644 --- a/drivers/iio/imu/bmi160/bmi160_core.c +++ b/drivers/iio/imu/bmi160/bmi160_core.c @@ -209,11 +209,11 @@ static const struct bmi160_scale_item bmi160_scale_table[] = { }; static const struct bmi160_odr bmi160_accel_odr[] = { - {0x01, 0, 78125}, - {0x02, 1, 5625}, - {0x03, 3, 125}, - {0x04, 6, 25}, - {0x05, 12, 5}, + {0x01, 0, 781250}, + {0x02, 1, 562500}, + {0x03, 3, 125000}, + {0x04, 6, 250000}, + {0x05, 12, 500000}, {0x06, 25, 0}, {0x07, 50, 0}, {0x08, 100, 0}, @@ -229,7 +229,7 @@ static const struct bmi160_odr bmi160_gyro_odr[] = { {0x08, 100, 0}, {0x09, 200, 0}, {0x0A, 400, 0}, - {0x0B, 8000, 0}, + {0x0B, 800, 0}, {0x0C, 1600, 0}, {0x0D, 3200, 0}, }; @@ -364,8 +364,8 @@ int bmi160_set_odr(struct bmi160_data *data, enum bmi160_sensor_type t, return regmap_update_bits(data->regmap, bmi160_regs[t].config, - bmi160_odr_table[t].tbl[i].bits, - bmi160_regs[t].config_odr_mask); + bmi160_regs[t].config_odr_mask, + bmi160_odr_table[t].tbl[i].bits); } static int bmi160_get_odr(struct bmi160_data *data, enum bmi160_sensor_type t, diff --git a/drivers/iio/industrialio-trigger.c b/drivers/iio/industrialio-trigger.c index ae2806aafb72..0c52dfe64977 100644 --- a/drivers/iio/industrialio-trigger.c +++ b/drivers/iio/industrialio-trigger.c @@ -210,22 +210,35 @@ static int iio_trigger_attach_poll_func(struct iio_trigger *trig, /* Prevent the module from being removed whilst attached to a trigger */ __module_get(pf->indio_dev->info->driver_module); + + /* Get irq number */ pf->irq = iio_trigger_get_irq(trig); + if (pf->irq < 0) + goto out_put_module; + + /* Request irq */ ret = request_threaded_irq(pf->irq, pf->h, pf->thread, pf->type, pf->name, pf); - if (ret < 0) { - module_put(pf->indio_dev->info->driver_module); - return ret; - } + if (ret < 0) + goto out_put_irq; + /* Enable trigger in driver */ if (trig->ops && trig->ops->set_trigger_state && notinuse) { ret = trig->ops->set_trigger_state(trig, true); if (ret < 0) - module_put(pf->indio_dev->info->driver_module); + goto out_free_irq; } return ret; + +out_free_irq: + free_irq(pf->irq, pf); +out_put_irq: + iio_trigger_put_irq(trig, pf->irq); +out_put_module: + module_put(pf->indio_dev->info->driver_module); + return ret; } static int iio_trigger_detach_poll_func(struct iio_trigger *trig, diff --git a/drivers/iio/light/apds9960.c b/drivers/iio/light/apds9960.c index b4dbb3912977..651d57b8abbf 100644 --- a/drivers/iio/light/apds9960.c +++ b/drivers/iio/light/apds9960.c @@ -1011,6 +1011,7 @@ static int apds9960_probe(struct i2c_client *client, iio_device_attach_buffer(indio_dev, buffer); + indio_dev->dev.parent = &client->dev; indio_dev->info = &apds9960_info; indio_dev->name = APDS9960_DRV_NAME; indio_dev->channels = apds9960_channels; diff --git a/drivers/iio/light/bh1780.c b/drivers/iio/light/bh1780.c index 72b364e4aa72..b54dcba05a82 100644 --- a/drivers/iio/light/bh1780.c +++ b/drivers/iio/light/bh1780.c @@ -84,7 +84,7 @@ static int bh1780_debugfs_reg_access(struct iio_dev *indio_dev, int ret; if (!readval) - bh1780_write(bh1780, (u8)reg, (u8)writeval); + return bh1780_write(bh1780, (u8)reg, (u8)writeval); ret = bh1780_read(bh1780, (u8)reg); if (ret < 0) @@ -187,7 +187,7 @@ static int bh1780_probe(struct i2c_client *client, indio_dev->dev.parent = &client->dev; indio_dev->info = &bh1780_info; - indio_dev->name = id->name; + indio_dev->name = "bh1780"; indio_dev->channels = bh1780_channels; indio_dev->num_channels = ARRAY_SIZE(bh1780_channels); indio_dev->modes = INDIO_DIRECT_MODE; @@ -226,7 +226,8 @@ static int bh1780_remove(struct i2c_client *client) static int bh1780_runtime_suspend(struct device *dev) { struct i2c_client *client = to_i2c_client(dev); - struct bh1780_data *bh1780 = i2c_get_clientdata(client); + struct iio_dev *indio_dev = i2c_get_clientdata(client); + struct bh1780_data *bh1780 = iio_priv(indio_dev); int ret; ret = bh1780_write(bh1780, BH1780_REG_CONTROL, BH1780_POFF); @@ -241,7 +242,8 @@ static int bh1780_runtime_suspend(struct device *dev) static int bh1780_runtime_resume(struct device *dev) { struct i2c_client *client = to_i2c_client(dev); - struct bh1780_data *bh1780 = i2c_get_clientdata(client); + struct iio_dev *indio_dev = i2c_get_clientdata(client); + struct bh1780_data *bh1780 = iio_priv(indio_dev); int ret; ret = bh1780_write(bh1780, BH1780_REG_CONTROL, BH1780_PON); diff --git a/drivers/iio/light/max44000.c b/drivers/iio/light/max44000.c index e01e58a9bd14..f17cb2ea18f5 100644 --- a/drivers/iio/light/max44000.c +++ b/drivers/iio/light/max44000.c @@ -147,7 +147,6 @@ static const struct iio_chan_spec max44000_channels[] = { { .type = IIO_PROXIMITY, .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), - .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .scan_index = MAX44000_SCAN_INDEX_PRX, .scan_type = { .sign = 'u', diff --git a/drivers/iio/magnetometer/st_magn_buffer.c b/drivers/iio/magnetometer/st_magn_buffer.c index ecd3bd0a9769..0a9e8fadfa9d 100644 --- a/drivers/iio/magnetometer/st_magn_buffer.c +++ b/drivers/iio/magnetometer/st_magn_buffer.c @@ -82,7 +82,7 @@ static const struct iio_buffer_setup_ops st_magn_buffer_setup_ops = { int st_magn_allocate_ring(struct iio_dev *indio_dev) { - return iio_triggered_buffer_setup(indio_dev, &iio_pollfunc_store_time, + return iio_triggered_buffer_setup(indio_dev, NULL, &st_sensors_trigger_handler, &st_magn_buffer_setup_ops); } diff --git a/drivers/iio/magnetometer/st_magn_core.c b/drivers/iio/magnetometer/st_magn_core.c index 62036d2a9956..8250fc322c56 100644 --- a/drivers/iio/magnetometer/st_magn_core.c +++ b/drivers/iio/magnetometer/st_magn_core.c @@ -572,6 +572,7 @@ static const struct iio_info magn_info = { static const struct iio_trigger_ops st_magn_trigger_ops = { .owner = THIS_MODULE, .set_trigger_state = ST_MAGN_TRIGGER_SET_STATE, + .validate_device = st_sensors_validate_device, }; #define ST_MAGN_TRIGGER_OPS (&st_magn_trigger_ops) #else diff --git a/drivers/iio/pressure/bmp280.c b/drivers/iio/pressure/bmp280.c index 2f1498e12bb2..724452d61846 100644 --- a/drivers/iio/pressure/bmp280.c +++ b/drivers/iio/pressure/bmp280.c @@ -879,8 +879,8 @@ static int bmp280_probe(struct i2c_client *client, if (ret < 0) return ret; if (chip_id != id->driver_data) { - dev_err(&client->dev, "bad chip id. expected %x got %x\n", - BMP280_CHIP_ID, chip_id); + dev_err(&client->dev, "bad chip id. expected %lx got %x\n", + id->driver_data, chip_id); return -EINVAL; } diff --git a/drivers/iio/pressure/st_pressure_buffer.c b/drivers/iio/pressure/st_pressure_buffer.c index 2ff53f222352..99468d0a64e7 100644 --- a/drivers/iio/pressure/st_pressure_buffer.c +++ b/drivers/iio/pressure/st_pressure_buffer.c @@ -82,7 +82,7 @@ static const struct iio_buffer_setup_ops st_press_buffer_setup_ops = { int st_press_allocate_ring(struct iio_dev *indio_dev) { - return iio_triggered_buffer_setup(indio_dev, &iio_pollfunc_store_time, + return iio_triggered_buffer_setup(indio_dev, NULL, &st_sensors_trigger_handler, &st_press_buffer_setup_ops); } diff --git a/drivers/iio/pressure/st_pressure_core.c b/drivers/iio/pressure/st_pressure_core.c index 9e9b72a8f18f..92a118c3c4ac 100644 --- a/drivers/iio/pressure/st_pressure_core.c +++ b/drivers/iio/pressure/st_pressure_core.c @@ -28,15 +28,21 @@ #include <linux/iio/common/st_sensors.h> #include "st_pressure.h" +#define MCELSIUS_PER_CELSIUS 1000 + +/* Default pressure sensitivity */ #define ST_PRESS_LSB_PER_MBAR 4096UL #define ST_PRESS_KPASCAL_NANO_SCALE (100000000UL / \ ST_PRESS_LSB_PER_MBAR) + +/* Default temperature sensitivity */ #define ST_PRESS_LSB_PER_CELSIUS 480UL -#define ST_PRESS_CELSIUS_NANO_SCALE (1000000000UL / \ - ST_PRESS_LSB_PER_CELSIUS) +#define ST_PRESS_MILLI_CELSIUS_OFFSET 42500UL + #define ST_PRESS_NUMBER_DATA_CHANNELS 1 /* FULLSCALE */ +#define ST_PRESS_FS_AVL_1100MB 1100 #define ST_PRESS_FS_AVL_1260MB 1260 #define ST_PRESS_1_OUT_XL_ADDR 0x28 @@ -54,9 +60,6 @@ #define ST_PRESS_LPS331AP_PW_MASK 0x80 #define ST_PRESS_LPS331AP_FS_ADDR 0x23 #define ST_PRESS_LPS331AP_FS_MASK 0x30 -#define ST_PRESS_LPS331AP_FS_AVL_1260_VAL 0x00 -#define ST_PRESS_LPS331AP_FS_AVL_1260_GAIN ST_PRESS_KPASCAL_NANO_SCALE -#define ST_PRESS_LPS331AP_FS_AVL_TEMP_GAIN ST_PRESS_CELSIUS_NANO_SCALE #define ST_PRESS_LPS331AP_BDU_ADDR 0x20 #define ST_PRESS_LPS331AP_BDU_MASK 0x04 #define ST_PRESS_LPS331AP_DRDY_IRQ_ADDR 0x22 @@ -67,9 +70,14 @@ #define ST_PRESS_LPS331AP_OD_IRQ_ADDR 0x22 #define ST_PRESS_LPS331AP_OD_IRQ_MASK 0x40 #define ST_PRESS_LPS331AP_MULTIREAD_BIT true -#define ST_PRESS_LPS331AP_TEMP_OFFSET 42500 /* CUSTOM VALUES FOR LPS001WP SENSOR */ + +/* LPS001WP pressure resolution */ +#define ST_PRESS_LPS001WP_LSB_PER_MBAR 16UL +/* LPS001WP temperature resolution */ +#define ST_PRESS_LPS001WP_LSB_PER_CELSIUS 64UL + #define ST_PRESS_LPS001WP_WAI_EXP 0xba #define ST_PRESS_LPS001WP_ODR_ADDR 0x20 #define ST_PRESS_LPS001WP_ODR_MASK 0x30 @@ -78,6 +86,8 @@ #define ST_PRESS_LPS001WP_ODR_AVL_13HZ_VAL 0x03 #define ST_PRESS_LPS001WP_PW_ADDR 0x20 #define ST_PRESS_LPS001WP_PW_MASK 0x40 +#define ST_PRESS_LPS001WP_FS_AVL_PRESS_GAIN \ + (100000000UL / ST_PRESS_LPS001WP_LSB_PER_MBAR) #define ST_PRESS_LPS001WP_BDU_ADDR 0x20 #define ST_PRESS_LPS001WP_BDU_MASK 0x04 #define ST_PRESS_LPS001WP_MULTIREAD_BIT true @@ -94,11 +104,6 @@ #define ST_PRESS_LPS25H_ODR_AVL_25HZ_VAL 0x04 #define ST_PRESS_LPS25H_PW_ADDR 0x20 #define ST_PRESS_LPS25H_PW_MASK 0x80 -#define ST_PRESS_LPS25H_FS_ADDR 0x00 -#define ST_PRESS_LPS25H_FS_MASK 0x00 -#define ST_PRESS_LPS25H_FS_AVL_1260_VAL 0x00 -#define ST_PRESS_LPS25H_FS_AVL_1260_GAIN ST_PRESS_KPASCAL_NANO_SCALE -#define ST_PRESS_LPS25H_FS_AVL_TEMP_GAIN ST_PRESS_CELSIUS_NANO_SCALE #define ST_PRESS_LPS25H_BDU_ADDR 0x20 #define ST_PRESS_LPS25H_BDU_MASK 0x04 #define ST_PRESS_LPS25H_DRDY_IRQ_ADDR 0x23 @@ -109,7 +114,6 @@ #define ST_PRESS_LPS25H_OD_IRQ_ADDR 0x22 #define ST_PRESS_LPS25H_OD_IRQ_MASK 0x40 #define ST_PRESS_LPS25H_MULTIREAD_BIT true -#define ST_PRESS_LPS25H_TEMP_OFFSET 42500 #define ST_PRESS_LPS25H_OUT_XL_ADDR 0x28 #define ST_TEMP_LPS25H_OUT_L_ADDR 0x2b @@ -161,7 +165,9 @@ static const struct iio_chan_spec st_press_lps001wp_channels[] = { .storagebits = 16, .endianness = IIO_LE, }, - .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), + .info_mask_separate = + BIT(IIO_CHAN_INFO_RAW) | + BIT(IIO_CHAN_INFO_SCALE), .modified = 0, }, { @@ -177,7 +183,7 @@ static const struct iio_chan_spec st_press_lps001wp_channels[] = { }, .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | - BIT(IIO_CHAN_INFO_OFFSET), + BIT(IIO_CHAN_INFO_SCALE), .modified = 0, }, IIO_CHAN_SOFT_TIMESTAMP(1) @@ -212,11 +218,14 @@ static const struct st_sensor_settings st_press_sensors_settings[] = { .addr = ST_PRESS_LPS331AP_FS_ADDR, .mask = ST_PRESS_LPS331AP_FS_MASK, .fs_avl = { + /* + * Pressure and temperature sensitivity values + * as defined in table 3 of LPS331AP datasheet. + */ [0] = { .num = ST_PRESS_FS_AVL_1260MB, - .value = ST_PRESS_LPS331AP_FS_AVL_1260_VAL, - .gain = ST_PRESS_LPS331AP_FS_AVL_1260_GAIN, - .gain2 = ST_PRESS_LPS331AP_FS_AVL_TEMP_GAIN, + .gain = ST_PRESS_KPASCAL_NANO_SCALE, + .gain2 = ST_PRESS_LSB_PER_CELSIUS, }, }, }, @@ -261,7 +270,17 @@ static const struct st_sensor_settings st_press_sensors_settings[] = { .value_off = ST_SENSORS_DEFAULT_POWER_OFF_VALUE, }, .fs = { - .addr = 0, + .fs_avl = { + /* + * Pressure and temperature resolution values + * as defined in table 3 of LPS001WP datasheet. + */ + [0] = { + .num = ST_PRESS_FS_AVL_1100MB, + .gain = ST_PRESS_LPS001WP_FS_AVL_PRESS_GAIN, + .gain2 = ST_PRESS_LPS001WP_LSB_PER_CELSIUS, + }, + }, }, .bdu = { .addr = ST_PRESS_LPS001WP_BDU_ADDR, @@ -298,14 +317,15 @@ static const struct st_sensor_settings st_press_sensors_settings[] = { .value_off = ST_SENSORS_DEFAULT_POWER_OFF_VALUE, }, .fs = { - .addr = ST_PRESS_LPS25H_FS_ADDR, - .mask = ST_PRESS_LPS25H_FS_MASK, .fs_avl = { + /* + * Pressure and temperature sensitivity values + * as defined in table 3 of LPS25H datasheet. + */ [0] = { .num = ST_PRESS_FS_AVL_1260MB, - .value = ST_PRESS_LPS25H_FS_AVL_1260_VAL, - .gain = ST_PRESS_LPS25H_FS_AVL_1260_GAIN, - .gain2 = ST_PRESS_LPS25H_FS_AVL_TEMP_GAIN, + .gain = ST_PRESS_KPASCAL_NANO_SCALE, + .gain2 = ST_PRESS_LSB_PER_CELSIUS, }, }, }, @@ -364,26 +384,26 @@ static int st_press_read_raw(struct iio_dev *indio_dev, return IIO_VAL_INT; case IIO_CHAN_INFO_SCALE: - *val = 0; - switch (ch->type) { case IIO_PRESSURE: + *val = 0; *val2 = press_data->current_fullscale->gain; - break; + return IIO_VAL_INT_PLUS_NANO; case IIO_TEMP: + *val = MCELSIUS_PER_CELSIUS; *val2 = press_data->current_fullscale->gain2; - break; + return IIO_VAL_FRACTIONAL; default: err = -EINVAL; goto read_error; } - return IIO_VAL_INT_PLUS_NANO; case IIO_CHAN_INFO_OFFSET: switch (ch->type) { case IIO_TEMP: - *val = 425; - *val2 = 10; + *val = ST_PRESS_MILLI_CELSIUS_OFFSET * + press_data->current_fullscale->gain2; + *val2 = MCELSIUS_PER_CELSIUS; break; default: err = -EINVAL; @@ -425,6 +445,7 @@ static const struct iio_info press_info = { static const struct iio_trigger_ops st_press_trigger_ops = { .owner = THIS_MODULE, .set_trigger_state = ST_PRESS_TRIGGER_SET_STATE, + .validate_device = st_sensors_validate_device, }; #define ST_PRESS_TRIGGER_OPS (&st_press_trigger_ops) #else diff --git a/drivers/iio/proximity/as3935.c b/drivers/iio/proximity/as3935.c index f4d29d5dbd5f..e2f926cdcad2 100644 --- a/drivers/iio/proximity/as3935.c +++ b/drivers/iio/proximity/as3935.c @@ -64,6 +64,7 @@ struct as3935_state { struct delayed_work work; u32 tune_cap; + u8 buffer[16]; /* 8-bit data + 56-bit padding + 64-bit timestamp */ u8 buf[2] ____cacheline_aligned; }; @@ -72,7 +73,8 @@ static const struct iio_chan_spec as3935_channels[] = { .type = IIO_PROXIMITY, .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | - BIT(IIO_CHAN_INFO_PROCESSED), + BIT(IIO_CHAN_INFO_PROCESSED) | + BIT(IIO_CHAN_INFO_SCALE), .scan_index = 0, .scan_type = { .sign = 'u', @@ -181,7 +183,12 @@ static int as3935_read_raw(struct iio_dev *indio_dev, /* storm out of range */ if (*val == AS3935_DATA_MASK) return -EINVAL; - *val *= 1000; + + if (m == IIO_CHAN_INFO_PROCESSED) + *val *= 1000; + break; + case IIO_CHAN_INFO_SCALE: + *val = 1000; break; default: return -EINVAL; @@ -206,10 +213,10 @@ static irqreturn_t as3935_trigger_handler(int irq, void *private) ret = as3935_read(st, AS3935_DATA, &val); if (ret) goto err_read; - val &= AS3935_DATA_MASK; - val *= 1000; - iio_push_to_buffers_with_timestamp(indio_dev, &val, pf->timestamp); + st->buffer[0] = val & AS3935_DATA_MASK; + iio_push_to_buffers_with_timestamp(indio_dev, &st->buffer, + pf->timestamp); err_read: iio_trigger_notify_done(indio_dev->trig); diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index c2e257d97eff..1a2984c28b95 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -178,6 +178,7 @@ static int write_gid(struct ib_device *ib_dev, u8 port, { int ret = 0; struct net_device *old_net_dev; + enum ib_gid_type old_gid_type; /* in rdma_cap_roce_gid_table, this funciton should be protected by a * sleep-able lock. @@ -199,6 +200,7 @@ static int write_gid(struct ib_device *ib_dev, u8 port, } old_net_dev = table->data_vec[ix].attr.ndev; + old_gid_type = table->data_vec[ix].attr.gid_type; if (old_net_dev && old_net_dev != attr->ndev) dev_put(old_net_dev); /* if modify_gid failed, just delete the old gid */ @@ -207,10 +209,14 @@ static int write_gid(struct ib_device *ib_dev, u8 port, attr = &zattr; table->data_vec[ix].context = NULL; } - if (default_gid) - table->data_vec[ix].props |= GID_TABLE_ENTRY_DEFAULT; + memcpy(&table->data_vec[ix].gid, gid, sizeof(*gid)); memcpy(&table->data_vec[ix].attr, attr, sizeof(*attr)); + if (default_gid) { + table->data_vec[ix].props |= GID_TABLE_ENTRY_DEFAULT; + if (action == GID_TABLE_WRITE_ACTION_DEL) + table->data_vec[ix].attr.gid_type = old_gid_type; + } if (table->data_vec[ix].attr.ndev && table->data_vec[ix].attr.ndev != old_net_dev) dev_hold(table->data_vec[ix].attr.ndev); @@ -405,7 +411,9 @@ int ib_cache_gid_del_all_netdev_gids(struct ib_device *ib_dev, u8 port, for (ix = 0; ix < table->sz; ix++) if (table->data_vec[ix].attr.ndev == ndev) - if (!del_gid(ib_dev, port, table, ix, false)) + if (!del_gid(ib_dev, port, table, ix, + !!(table->data_vec[ix].props & + GID_TABLE_ENTRY_DEFAULT))) deleted = true; write_unlock_irq(&table->rwlock); diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 1d92e091e22e..c99525512b34 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -3452,14 +3452,14 @@ static int cm_establish(struct ib_cm_id *cm_id) work->cm_event.event = IB_CM_USER_ESTABLISHED; /* Check if the device started its remove_one */ - spin_lock_irq(&cm.lock); + spin_lock_irqsave(&cm.lock, flags); if (!cm_dev->going_down) { queue_delayed_work(cm.wq, &work->work, 0); } else { kfree(work); ret = -ENODEV; } - spin_unlock_irq(&cm.lock); + spin_unlock_irqrestore(&cm.lock, flags); out: return ret; diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index f0c91ba3178a..ad1b1adcf6f0 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -708,17 +708,6 @@ static void cma_deref_id(struct rdma_id_private *id_priv) complete(&id_priv->comp); } -static int cma_disable_callback(struct rdma_id_private *id_priv, - enum rdma_cm_state state) -{ - mutex_lock(&id_priv->handler_mutex); - if (id_priv->state != state) { - mutex_unlock(&id_priv->handler_mutex); - return -EINVAL; - } - return 0; -} - struct rdma_cm_id *rdma_create_id(struct net *net, rdma_cm_event_handler event_handler, void *context, enum rdma_port_space ps, @@ -1671,11 +1660,12 @@ static int cma_ib_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) struct rdma_cm_event event; int ret = 0; + mutex_lock(&id_priv->handler_mutex); if ((ib_event->event != IB_CM_TIMEWAIT_EXIT && - cma_disable_callback(id_priv, RDMA_CM_CONNECT)) || + id_priv->state != RDMA_CM_CONNECT) || (ib_event->event == IB_CM_TIMEWAIT_EXIT && - cma_disable_callback(id_priv, RDMA_CM_DISCONNECT))) - return 0; + id_priv->state != RDMA_CM_DISCONNECT)) + goto out; memset(&event, 0, sizeof event); switch (ib_event->event) { @@ -1870,7 +1860,7 @@ static int cma_check_req_qp_type(struct rdma_cm_id *id, struct ib_cm_event *ib_e static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) { - struct rdma_id_private *listen_id, *conn_id; + struct rdma_id_private *listen_id, *conn_id = NULL; struct rdma_cm_event event; struct net_device *net_dev; int offset, ret; @@ -1884,9 +1874,10 @@ static int cma_req_handler(struct ib_cm_id *cm_id, struct ib_cm_event *ib_event) goto net_dev_put; } - if (cma_disable_callback(listen_id, RDMA_CM_LISTEN)) { + mutex_lock(&listen_id->handler_mutex); + if (listen_id->state != RDMA_CM_LISTEN) { ret = -ECONNABORTED; - goto net_dev_put; + goto err1; } memset(&event, 0, sizeof event); @@ -1976,8 +1967,9 @@ static int cma_iw_handler(struct iw_cm_id *iw_id, struct iw_cm_event *iw_event) struct sockaddr *laddr = (struct sockaddr *)&iw_event->local_addr; struct sockaddr *raddr = (struct sockaddr *)&iw_event->remote_addr; - if (cma_disable_callback(id_priv, RDMA_CM_CONNECT)) - return 0; + mutex_lock(&id_priv->handler_mutex); + if (id_priv->state != RDMA_CM_CONNECT) + goto out; memset(&event, 0, sizeof event); switch (iw_event->event) { @@ -2029,6 +2021,7 @@ static int cma_iw_handler(struct iw_cm_id *iw_id, struct iw_cm_event *iw_event) return ret; } +out: mutex_unlock(&id_priv->handler_mutex); return ret; } @@ -2039,13 +2032,15 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id, struct rdma_cm_id *new_cm_id; struct rdma_id_private *listen_id, *conn_id; struct rdma_cm_event event; - int ret; + int ret = -ECONNABORTED; struct sockaddr *laddr = (struct sockaddr *)&iw_event->local_addr; struct sockaddr *raddr = (struct sockaddr *)&iw_event->remote_addr; listen_id = cm_id->context; - if (cma_disable_callback(listen_id, RDMA_CM_LISTEN)) - return -ECONNABORTED; + + mutex_lock(&listen_id->handler_mutex); + if (listen_id->state != RDMA_CM_LISTEN) + goto out; /* Create a new RDMA id for the new IW CM ID */ new_cm_id = rdma_create_id(listen_id->id.route.addr.dev_addr.net, @@ -3216,8 +3211,9 @@ static int cma_sidr_rep_handler(struct ib_cm_id *cm_id, struct ib_cm_sidr_rep_event_param *rep = &ib_event->param.sidr_rep_rcvd; int ret = 0; - if (cma_disable_callback(id_priv, RDMA_CM_CONNECT)) - return 0; + mutex_lock(&id_priv->handler_mutex); + if (id_priv->state != RDMA_CM_CONNECT) + goto out; memset(&event, 0, sizeof event); switch (ib_event->event) { @@ -3673,12 +3669,13 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) struct rdma_id_private *id_priv; struct cma_multicast *mc = multicast->context; struct rdma_cm_event event; - int ret; + int ret = 0; id_priv = mc->id_priv; - if (cma_disable_callback(id_priv, RDMA_CM_ADDR_BOUND) && - cma_disable_callback(id_priv, RDMA_CM_ADDR_RESOLVED)) - return 0; + mutex_lock(&id_priv->handler_mutex); + if (id_priv->state != RDMA_CM_ADDR_BOUND && + id_priv->state != RDMA_CM_ADDR_RESOLVED) + goto out; if (!status) status = cma_set_qkey(id_priv, be32_to_cpu(multicast->rec.qkey)); @@ -3720,6 +3717,7 @@ static int cma_ib_mc_handler(int status, struct ib_sa_multicast *multicast) return 0; } +out: mutex_unlock(&id_priv->handler_mutex); return 0; } @@ -3878,12 +3876,12 @@ static int cma_iboe_join_multicast(struct rdma_id_private *id_priv, gid_type = id_priv->cma_dev->default_gid_type[id_priv->id.port_num - rdma_start_port(id_priv->cma_dev->device)]; if (addr->sa_family == AF_INET) { - if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) + if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) { + mc->multicast.ib->rec.hop_limit = IPV6_DEFAULT_HOPLIMIT; err = cma_igmp_send(ndev, &mc->multicast.ib->rec.mgid, true); - if (!err) { - mc->igmp_joined = true; - mc->multicast.ib->rec.hop_limit = IPV6_DEFAULT_HOPLIMIT; + if (!err) + mc->igmp_joined = true; } } else { if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 5516fb070344..5c155fa91eec 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -661,6 +661,9 @@ int ib_query_port(struct ib_device *device, if (err || port_attr->subnet_prefix) return err; + if (rdma_port_get_link_layer(device, port_num) != IB_LINK_LAYER_INFINIBAND) + return 0; + err = ib_query_gid(device, port_num, 0, &gid, NULL); if (err) return err; @@ -1024,7 +1027,8 @@ static int __init ib_core_init(void) goto err_mad; } - if (ib_add_ibnl_clients()) { + ret = ib_add_ibnl_clients(); + if (ret) { pr_warn("Couldn't register ibnl clients\n"); goto err_sa; } diff --git a/drivers/infiniband/core/iwpm_msg.c b/drivers/infiniband/core/iwpm_msg.c index 43e3fa27102b..1c41b95cefec 100644 --- a/drivers/infiniband/core/iwpm_msg.c +++ b/drivers/infiniband/core/iwpm_msg.c @@ -506,7 +506,7 @@ int iwpm_add_and_query_mapping_cb(struct sk_buff *skb, if (!nlmsg_request) { pr_info("%s: Could not find a matching request (seq = %u)\n", __func__, msg_seq); - return -EINVAL; + return -EINVAL; } pm_msg = nlmsg_request->req_buffer; local_sockaddr = (struct sockaddr_storage *) diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index 82fb511112da..2d49228f28b2 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -1638,9 +1638,9 @@ static void remove_mad_reg_req(struct ib_mad_agent_private *agent_priv) /* Now, check to see if there are any methods still in use */ if (!check_method_table(method)) { /* If not, release management method table */ - kfree(method); - class->method_table[mgmt_class] = NULL; - /* Any management classes left ? */ + kfree(method); + class->method_table[mgmt_class] = NULL; + /* Any management classes left ? */ if (!check_class_table(class)) { /* If not, release management class table */ kfree(class); diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index 5e573bb18660..a5793c8f1590 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -889,9 +889,9 @@ static struct attribute *alloc_hsa_lifespan(char *name, u8 port_num) static void setup_hw_stats(struct ib_device *device, struct ib_port *port, u8 port_num) { - struct attribute_group *hsag = NULL; + struct attribute_group *hsag; struct rdma_hw_stats *stats; - int i = 0, ret; + int i, ret; stats = device->alloc_hw_stats(device, port_num); @@ -899,19 +899,22 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port, return; if (!stats->names || stats->num_counters <= 0) - goto err; + goto err_free_stats; + /* + * Two extra attribue elements here, one for the lifespan entry and + * one to NULL terminate the list for the sysfs core code + */ hsag = kzalloc(sizeof(*hsag) + - // 1 extra for the lifespan config entry - sizeof(void *) * (stats->num_counters + 1), + sizeof(void *) * (stats->num_counters + 2), GFP_KERNEL); if (!hsag) - return; + goto err_free_stats; ret = device->get_hw_stats(device, stats, port_num, stats->num_counters); if (ret != stats->num_counters) - goto err; + goto err_free_hsag; stats->timestamp = jiffies; @@ -922,10 +925,13 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port, hsag->attrs[i] = alloc_hsa(i, port_num, stats->names[i]); if (!hsag->attrs[i]) goto err; + sysfs_attr_init(hsag->attrs[i]); } /* treat an error here as non-fatal */ hsag->attrs[i] = alloc_hsa_lifespan("lifespan", port_num); + if (hsag->attrs[i]) + sysfs_attr_init(hsag->attrs[i]); if (port) { struct kobject *kobj = &port->kobj; @@ -946,10 +952,12 @@ static void setup_hw_stats(struct ib_device *device, struct ib_port *port, return; err: - kfree(stats); for (; i >= 0; i--) kfree(hsag->attrs[i]); +err_free_hsag: kfree(hsag); +err_free_stats: + kfree(stats); return; } diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 1a8babb8ee3c..825021d1008b 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -1747,7 +1747,7 @@ static int create_qp(struct ib_uverbs_file *file, struct ib_srq *srq = NULL; struct ib_qp *qp; char *buf; - struct ib_qp_init_attr attr; + struct ib_qp_init_attr attr = {}; struct ib_uverbs_ex_create_qp_resp resp; int ret; diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 1d7d4cf442e3..6298f54b4137 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -511,12 +511,16 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, ah_attr->grh.dgid = sgid; if (!rdma_cap_eth_ah(device, port_num)) { - ret = ib_find_cached_gid_by_port(device, &dgid, - IB_GID_TYPE_IB, - port_num, NULL, - &gid_index); - if (ret) - return ret; + if (dgid.global.interface_id != cpu_to_be64(IB_SA_WELL_KNOWN_GUID)) { + ret = ib_find_cached_gid_by_port(device, &dgid, + IB_GID_TYPE_IB, + port_num, NULL, + &gid_index); + if (ret) + return ret; + } else { + gid_index = 0; + } } ah_attr->grh.sgid_index = (u8) gid_index; diff --git a/drivers/infiniband/hw/hfi1/affinity.c b/drivers/infiniband/hw/hfi1/affinity.c index 6e7050ab9e16..14d7eeb09be6 100644 --- a/drivers/infiniband/hw/hfi1/affinity.c +++ b/drivers/infiniband/hw/hfi1/affinity.c @@ -300,16 +300,15 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node) const struct cpumask *node_mask, *proc_mask = tsk_cpus_allowed(current); struct cpu_mask_set *set = &dd->affinity->proc; - char buf[1024]; /* * check whether process/context affinity has already * been set */ if (cpumask_weight(proc_mask) == 1) { - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(proc_mask)); - hfi1_cdbg(PROC, "PID %u %s affinity set to CPU %s", - current->pid, current->comm, buf); + hfi1_cdbg(PROC, "PID %u %s affinity set to CPU %*pbl", + current->pid, current->comm, + cpumask_pr_args(proc_mask)); /* * Mark the pre-set CPU as used. This is atomic so we don't * need the lock @@ -318,9 +317,9 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node) cpumask_set_cpu(cpu, &set->used); goto done; } else if (cpumask_weight(proc_mask) < cpumask_weight(&set->mask)) { - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(proc_mask)); - hfi1_cdbg(PROC, "PID %u %s affinity set to CPU set(s) %s", - current->pid, current->comm, buf); + hfi1_cdbg(PROC, "PID %u %s affinity set to CPU set(s) %*pbl", + current->pid, current->comm, + cpumask_pr_args(proc_mask)); goto done; } @@ -356,8 +355,8 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node) cpumask_or(intrs, intrs, (dd->affinity->rcv_intr.gen ? &dd->affinity->rcv_intr.mask : &dd->affinity->rcv_intr.used)); - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(intrs)); - hfi1_cdbg(PROC, "CPUs used by interrupts: %s", buf); + hfi1_cdbg(PROC, "CPUs used by interrupts: %*pbl", + cpumask_pr_args(intrs)); /* * If we don't have a NUMA node requested, preference is towards @@ -366,18 +365,16 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node) if (node == -1) node = dd->node; node_mask = cpumask_of_node(node); - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(node_mask)); - hfi1_cdbg(PROC, "device on NUMA %u, CPUs %s", node, buf); + hfi1_cdbg(PROC, "device on NUMA %u, CPUs %*pbl", node, + cpumask_pr_args(node_mask)); /* diff will hold all unused cpus */ cpumask_andnot(diff, &set->mask, &set->used); - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(diff)); - hfi1_cdbg(PROC, "unused CPUs (all) %s", buf); + hfi1_cdbg(PROC, "unused CPUs (all) %*pbl", cpumask_pr_args(diff)); /* get cpumask of available CPUs on preferred NUMA */ cpumask_and(mask, diff, node_mask); - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(mask)); - hfi1_cdbg(PROC, "available cpus on NUMA %s", buf); + hfi1_cdbg(PROC, "available cpus on NUMA %*pbl", cpumask_pr_args(mask)); /* * At first, we don't want to place processes on the same @@ -395,8 +392,8 @@ int hfi1_get_proc_affinity(struct hfi1_devdata *dd, int node) cpumask_andnot(diff, &set->mask, &set->used); cpumask_andnot(mask, diff, node_mask); } - scnprintf(buf, 1024, "%*pbl", cpumask_pr_args(mask)); - hfi1_cdbg(PROC, "possible CPUs for process %s", buf); + hfi1_cdbg(PROC, "possible CPUs for process %*pbl", + cpumask_pr_args(mask)); cpu = cpumask_first(mask); if (cpu >= nr_cpu_ids) /* empty */ diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index 3b876da745a1..f5de85178055 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -1037,7 +1037,7 @@ static void dc_shutdown(struct hfi1_devdata *); static void dc_start(struct hfi1_devdata *); static int qos_rmt_entries(struct hfi1_devdata *dd, unsigned int *mp, unsigned int *np); -static void remove_full_mgmt_pkey(struct hfi1_pportdata *ppd); +static void clear_full_mgmt_pkey(struct hfi1_pportdata *ppd); /* * Error interrupt table entry. This is used as input to the interrupt @@ -6962,8 +6962,6 @@ void handle_link_down(struct work_struct *work) } reset_neighbor_info(ppd); - if (ppd->mgmt_allowed) - remove_full_mgmt_pkey(ppd); /* disable the port */ clear_rcvctrl(ppd->dd, RCV_CTRL_RCV_PORT_ENABLE_SMASK); @@ -7070,12 +7068,16 @@ static void add_full_mgmt_pkey(struct hfi1_pportdata *ppd) __func__, ppd->pkeys[2], FULL_MGMT_P_KEY); ppd->pkeys[2] = FULL_MGMT_P_KEY; (void)hfi1_set_ib_cfg(ppd, HFI1_IB_CFG_PKEYS, 0); + hfi1_event_pkey_change(ppd->dd, ppd->port); } -static void remove_full_mgmt_pkey(struct hfi1_pportdata *ppd) +static void clear_full_mgmt_pkey(struct hfi1_pportdata *ppd) { - ppd->pkeys[2] = 0; - (void)hfi1_set_ib_cfg(ppd, HFI1_IB_CFG_PKEYS, 0); + if (ppd->pkeys[2] != 0) { + ppd->pkeys[2] = 0; + (void)hfi1_set_ib_cfg(ppd, HFI1_IB_CFG_PKEYS, 0); + hfi1_event_pkey_change(ppd->dd, ppd->port); + } } /* @@ -7832,8 +7834,8 @@ static void handle_dcc_err(struct hfi1_devdata *dd, u32 unused, u64 reg) * save first 2 flits in the packet that caused * the error */ - dd->err_info_rcvport.packet_flit1 = hdr0; - dd->err_info_rcvport.packet_flit2 = hdr1; + dd->err_info_rcvport.packet_flit1 = hdr0; + dd->err_info_rcvport.packet_flit2 = hdr1; } switch (info) { case 1: @@ -9168,6 +9170,13 @@ int start_link(struct hfi1_pportdata *ppd) return 0; } + /* + * FULL_MGMT_P_KEY is cleared from the pkey table, so that the + * pkey table can be configured properly if the HFI unit is connected + * to switch port with MgmtAllowed=NO + */ + clear_full_mgmt_pkey(ppd); + return set_link_state(ppd, HLS_DN_POLL); } @@ -9777,7 +9786,7 @@ static void set_send_length(struct hfi1_pportdata *ppd) u64 len1 = 0, len2 = (((dd->vld[15].mtu + max_hb) >> 2) & SEND_LEN_CHECK1_LEN_VL15_MASK) << SEND_LEN_CHECK1_LEN_VL15_SHIFT; - int i; + int i, j; u32 thres; for (i = 0; i < ppd->vls_supported; i++) { @@ -9801,7 +9810,10 @@ static void set_send_length(struct hfi1_pportdata *ppd) sc_mtu_to_threshold(dd->vld[i].sc, dd->vld[i].mtu, dd->rcd[0]->rcvhdrqentsize)); - sc_set_cr_threshold(dd->vld[i].sc, thres); + for (j = 0; j < INIT_SC_PER_VL; j++) + sc_set_cr_threshold( + pio_select_send_context_vl(dd, j, i), + thres); } thres = min(sc_percent_to_threshold(dd->vld[15].sc, 50), sc_mtu_to_threshold(dd->vld[15].sc, @@ -11906,7 +11918,7 @@ static void update_synth_timer(unsigned long opaque) hfi1_cdbg(CNTR, "[%d] No update necessary", dd->unit); } -mod_timer(&dd->synth_stats_timer, jiffies + HZ * SYNTH_CNT_TIME); + mod_timer(&dd->synth_stats_timer, jiffies + HZ * SYNTH_CNT_TIME); } #define C_MAX_NAME 13 /* 12 chars + one for /0 */ diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c index 7a5b0e676cc7..c702a009608f 100644 --- a/drivers/infiniband/hw/hfi1/file_ops.c +++ b/drivers/infiniband/hw/hfi1/file_ops.c @@ -203,6 +203,9 @@ static long hfi1_file_ioctl(struct file *fp, unsigned int cmd, switch (cmd) { case HFI1_IOCTL_ASSIGN_CTXT: + if (uctxt) + return -EINVAL; + if (copy_from_user(&uinfo, (struct hfi1_user_info __user *)arg, sizeof(uinfo))) diff --git a/drivers/infiniband/hw/hfi1/init.c b/drivers/infiniband/hw/hfi1/init.c index 5cc492e5776d..eed971ccd2a1 100644 --- a/drivers/infiniband/hw/hfi1/init.c +++ b/drivers/infiniband/hw/hfi1/init.c @@ -1337,7 +1337,7 @@ static void cleanup_device_data(struct hfi1_devdata *dd) dma_free_coherent(&dd->pcidev->dev, sizeof(u64), (void *)dd->rcvhdrtail_dummy_kvaddr, dd->rcvhdrtail_dummy_physaddr); - dd->rcvhdrtail_dummy_kvaddr = NULL; + dd->rcvhdrtail_dummy_kvaddr = NULL; } for (ctxt = 0; tmp && ctxt < dd->num_rcv_contexts; ctxt++) { @@ -1383,7 +1383,7 @@ static void postinit_cleanup(struct hfi1_devdata *dd) static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent) { int ret = 0, j, pidx, initfail; - struct hfi1_devdata *dd = NULL; + struct hfi1_devdata *dd = ERR_PTR(-EINVAL); struct hfi1_pportdata *ppd; /* First, lock the non-writable module parameters */ diff --git a/drivers/infiniband/hw/hfi1/mad.c b/drivers/infiniband/hw/hfi1/mad.c index 219029576ba0..fca07a1d6c28 100644 --- a/drivers/infiniband/hw/hfi1/mad.c +++ b/drivers/infiniband/hw/hfi1/mad.c @@ -78,6 +78,16 @@ static inline void clear_opa_smp_data(struct opa_smp *smp) memset(data, 0, size); } +void hfi1_event_pkey_change(struct hfi1_devdata *dd, u8 port) +{ + struct ib_event event; + + event.event = IB_EVENT_PKEY_CHANGE; + event.device = &dd->verbs_dev.rdi.ibdev; + event.element.port_num = port; + ib_dispatch_event(&event); +} + static void send_trap(struct hfi1_ibport *ibp, void *data, unsigned len) { struct ib_mad_send_buf *send_buf; @@ -1418,15 +1428,10 @@ static int set_pkeys(struct hfi1_devdata *dd, u8 port, u16 *pkeys) } if (changed) { - struct ib_event event; - (void)hfi1_set_ib_cfg(ppd, HFI1_IB_CFG_PKEYS, 0); - - event.event = IB_EVENT_PKEY_CHANGE; - event.device = &dd->verbs_dev.rdi.ibdev; - event.element.port_num = port; - ib_dispatch_event(&event); + hfi1_event_pkey_change(dd, port); } + return 0; } diff --git a/drivers/infiniband/hw/hfi1/mad.h b/drivers/infiniband/hw/hfi1/mad.h index 55ee08675333..8b734aaae88a 100644 --- a/drivers/infiniband/hw/hfi1/mad.h +++ b/drivers/infiniband/hw/hfi1/mad.h @@ -434,4 +434,6 @@ struct sc2vlnt { COUNTER_MASK(1, 3) | \ COUNTER_MASK(1, 4)) +void hfi1_event_pkey_change(struct hfi1_devdata *dd, u8 port); + #endif /* _HFI1_MAD_H */ diff --git a/drivers/infiniband/hw/hfi1/pio.c b/drivers/infiniband/hw/hfi1/pio.c index d5edb1afbb8f..d4022450b73f 100644 --- a/drivers/infiniband/hw/hfi1/pio.c +++ b/drivers/infiniband/hw/hfi1/pio.c @@ -995,7 +995,7 @@ static void sc_wait_for_packet_egress(struct send_context *sc, int pause) /* counter is reset if occupancy count changes */ if (reg != reg_prev) loop = 0; - if (loop > 500) { + if (loop > 50000) { /* timed out - bounce the link */ dd_dev_err(dd, "%s: context %u(%u) timeout waiting for packets to egress, remaining count %u, bouncing link\n", @@ -1798,6 +1798,21 @@ static void pio_map_rcu_callback(struct rcu_head *list) } /* + * Set credit return threshold for the kernel send context + */ +static void set_threshold(struct hfi1_devdata *dd, int scontext, int i) +{ + u32 thres; + + thres = min(sc_percent_to_threshold(dd->kernel_send_context[scontext], + 50), + sc_mtu_to_threshold(dd->kernel_send_context[scontext], + dd->vld[i].mtu, + dd->rcd[0]->rcvhdrqentsize)); + sc_set_cr_threshold(dd->kernel_send_context[scontext], thres); +} + +/* * pio_map_init - called when #vls change * @dd: hfi1_devdata * @port: port number @@ -1872,11 +1887,16 @@ int pio_map_init(struct hfi1_devdata *dd, u8 port, u8 num_vls, u8 *vl_scontexts) if (!newmap->map[i]) goto bail; newmap->map[i]->mask = (1 << ilog2(sz)) - 1; - /* assign send contexts */ + /* + * assign send contexts and + * adjust credit return threshold + */ for (j = 0; j < sz; j++) { - if (dd->kernel_send_context[scontext]) + if (dd->kernel_send_context[scontext]) { newmap->map[i]->ksc[j] = dd->kernel_send_context[scontext]; + set_threshold(dd, scontext, i); + } if (++scontext >= first_scontext + vl_scontexts[i]) /* wrap back to first send context */ diff --git a/drivers/infiniband/hw/hfi1/qsfp.c b/drivers/infiniband/hw/hfi1/qsfp.c index 2441669f0817..9fb561682c66 100644 --- a/drivers/infiniband/hw/hfi1/qsfp.c +++ b/drivers/infiniband/hw/hfi1/qsfp.c @@ -579,7 +579,8 @@ int qsfp_dump(struct hfi1_pportdata *ppd, char *buf, int len) if (ppd->qsfp_info.cache_valid) { if (QSFP_IS_CU(cache[QSFP_MOD_TECH_OFFS])) - sprintf(lenstr, "%dM ", cache[QSFP_MOD_LEN_OFFS]); + snprintf(lenstr, sizeof(lenstr), "%dM ", + cache[QSFP_MOD_LEN_OFFS]); power_byte = cache[QSFP_MOD_PWR_OFFS]; sofar += scnprintf(buf + sofar, len - sofar, "PWR:%.3sW\n", diff --git a/drivers/infiniband/hw/hfi1/trace.c b/drivers/infiniband/hw/hfi1/trace.c index 79b2952c0dfb..4cfb13771897 100644 --- a/drivers/infiniband/hw/hfi1/trace.c +++ b/drivers/infiniband/hw/hfi1/trace.c @@ -214,19 +214,6 @@ const char *print_u32_array( return ret; } -const char *print_u64_array( - struct trace_seq *p, - u64 *arr, int len) -{ - int i; - const char *ret = trace_seq_buffer_ptr(p); - - for (i = 0; i < len; i++) - trace_seq_printf(p, "%s0x%016llx", i == 0 ? "" : " ", arr[i]); - trace_seq_putc(p, 0); - return ret; -} - __hfi1_trace_fn(PKT); __hfi1_trace_fn(PROC); __hfi1_trace_fn(SDMA); diff --git a/drivers/infiniband/hw/hfi1/user_sdma.c b/drivers/infiniband/hw/hfi1/user_sdma.c index 29f4795f866c..47ffd273ecbd 100644 --- a/drivers/infiniband/hw/hfi1/user_sdma.c +++ b/drivers/infiniband/hw/hfi1/user_sdma.c @@ -183,7 +183,7 @@ struct user_sdma_iovec { struct sdma_mmu_node *node; }; -#define SDMA_CACHE_NODE_EVICT BIT(0) +#define SDMA_CACHE_NODE_EVICT 0 struct sdma_mmu_node { struct mmu_rb_node rb; @@ -1355,11 +1355,11 @@ static int set_txreq_header(struct user_sdma_request *req, */ SDMA_DBG(req, "TID offset %ubytes %uunits om%u", req->tidoffset, req->tidoffset / req->omfactor, - !!(req->omfactor - KDETH_OM_SMALL)); + req->omfactor != KDETH_OM_SMALL); KDETH_SET(hdr->kdeth.ver_tid_offset, OFFSET, req->tidoffset / req->omfactor); KDETH_SET(hdr->kdeth.ver_tid_offset, OM, - !!(req->omfactor - KDETH_OM_SMALL)); + req->omfactor != KDETH_OM_SMALL); } done: trace_hfi1_sdma_user_header(pq->dd, pq->ctxt, pq->subctxt, diff --git a/drivers/infiniband/hw/hfi1/verbs_txreq.c b/drivers/infiniband/hw/hfi1/verbs_txreq.c index bc95c4112c61..d8fb056526f8 100644 --- a/drivers/infiniband/hw/hfi1/verbs_txreq.c +++ b/drivers/infiniband/hw/hfi1/verbs_txreq.c @@ -92,11 +92,10 @@ void hfi1_put_txreq(struct verbs_txreq *tx) struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev, struct rvt_qp *qp) + __must_hold(&qp->s_lock) { struct verbs_txreq *tx = ERR_PTR(-EBUSY); - unsigned long flags; - spin_lock_irqsave(&qp->s_lock, flags); write_seqlock(&dev->iowait_lock); if (ib_rvt_state_ops[qp->state] & RVT_PROCESS_RECV_OK) { struct hfi1_qp_priv *priv; @@ -116,7 +115,6 @@ struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev, } out: write_sequnlock(&dev->iowait_lock); - spin_unlock_irqrestore(&qp->s_lock, flags); return tx; } diff --git a/drivers/infiniband/hw/hfi1/verbs_txreq.h b/drivers/infiniband/hw/hfi1/verbs_txreq.h index 1cf69b2fe4a5..a1d6e0807f97 100644 --- a/drivers/infiniband/hw/hfi1/verbs_txreq.h +++ b/drivers/infiniband/hw/hfi1/verbs_txreq.h @@ -73,6 +73,7 @@ struct verbs_txreq *__get_txreq(struct hfi1_ibdev *dev, static inline struct verbs_txreq *get_txreq(struct hfi1_ibdev *dev, struct rvt_qp *qp) + __must_hold(&qp->slock) { struct verbs_txreq *tx; struct hfi1_qp_priv *priv = qp->priv; diff --git a/drivers/infiniband/hw/i40iw/i40iw.h b/drivers/infiniband/hw/i40iw/i40iw.h index 8b9532034558..b738acdb9b02 100644 --- a/drivers/infiniband/hw/i40iw/i40iw.h +++ b/drivers/infiniband/hw/i40iw/i40iw.h @@ -113,6 +113,8 @@ #define IW_HMC_OBJ_TYPE_NUM ARRAY_SIZE(iw_hmc_obj_types) #define IW_CFG_FPM_QP_COUNT 32768 +#define I40IW_MAX_PAGES_PER_FMR 512 +#define I40IW_MIN_PAGES_PER_FMR 1 #define I40IW_MTU_TO_MSS 40 #define I40IW_DEFAULT_MSS 1460 diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c index 02a735b64208..33959ed14563 100644 --- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c +++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c @@ -79,6 +79,7 @@ static int i40iw_query_device(struct ib_device *ibdev, props->max_qp_init_rd_atom = props->max_qp_rd_atom; props->atomic_cap = IB_ATOMIC_NONE; props->max_map_per_fmr = 1; + props->max_fast_reg_page_list_len = I40IW_MAX_PAGES_PER_FMR; return 0; } @@ -1527,7 +1528,7 @@ static struct ib_mr *i40iw_alloc_mr(struct ib_pd *pd, mutex_lock(&iwdev->pbl_mutex); status = i40iw_get_pble(&iwdev->sc_dev, iwdev->pble_rsrc, palloc, iwmr->page_cnt); mutex_unlock(&iwdev->pbl_mutex); - if (!status) + if (status) goto err1; if (palloc->level != I40IW_LEVEL_1) @@ -2149,6 +2150,7 @@ static int i40iw_post_send(struct ib_qp *ibqp, struct i40iw_sc_dev *dev = &iwqp->iwdev->sc_dev; struct i40iw_fast_reg_stag_info info; + memset(&info, 0, sizeof(info)); info.access_rights = I40IW_ACCESS_FLAGS_LOCALREAD; info.access_rights |= i40iw_get_user_access(flags); info.stag_key = reg_wr(ib_wr)->key & 0xff; @@ -2158,10 +2160,14 @@ static int i40iw_post_send(struct ib_qp *ibqp, info.addr_type = I40IW_ADDR_TYPE_VA_BASED; info.va = (void *)(uintptr_t)iwmr->ibmr.iova; info.total_len = iwmr->ibmr.length; + info.reg_addr_pa = *(u64 *)palloc->level1.addr; info.first_pm_pbl_index = palloc->level1.idx; info.local_fence = ib_wr->send_flags & IB_SEND_FENCE; info.signaled = ib_wr->send_flags & IB_SEND_SIGNALED; + if (iwmr->npages > I40IW_MIN_PAGES_PER_FMR) + info.chunk_size = 1; + if (page_shift == 21) info.page_size = 1; /* 2M page */ @@ -2327,13 +2333,16 @@ static int i40iw_req_notify_cq(struct ib_cq *ibcq, { struct i40iw_cq *iwcq; struct i40iw_cq_uk *ukcq; - enum i40iw_completion_notify cq_notify = IW_CQ_COMPL_SOLICITED; + unsigned long flags; + enum i40iw_completion_notify cq_notify = IW_CQ_COMPL_EVENT; iwcq = (struct i40iw_cq *)ibcq; ukcq = &iwcq->sc_cq.cq_uk; - if (notify_flags == IB_CQ_NEXT_COMP) - cq_notify = IW_CQ_COMPL_EVENT; + if (notify_flags == IB_CQ_SOLICITED) + cq_notify = IW_CQ_COMPL_SOLICITED; + spin_lock_irqsave(&iwcq->lock, flags); ukcq->ops.iw_cq_request_notification(ukcq, cq_notify); + spin_unlock_irqrestore(&iwcq->lock, flags); return 0; } diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c index 105246fba2e7..5fc623362731 100644 --- a/drivers/infiniband/hw/mlx4/ah.c +++ b/drivers/infiniband/hw/mlx4/ah.c @@ -47,6 +47,7 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, ah->av.ib.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24)); ah->av.ib.g_slid = ah_attr->src_path_bits; + ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28); if (ah_attr->ah_flags & IB_AH_GRH) { ah->av.ib.g_slid |= 0x80; ah->av.ib.gid_index = ah_attr->grh.sgid_index; @@ -64,7 +65,6 @@ static struct ib_ah *create_ib_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr, !(1 << ah->av.ib.stat_rate & dev->caps.stat_rate_support)) --ah->av.ib.stat_rate; } - ah->av.ib.sl_tclass_flowlabel = cpu_to_be32(ah_attr->sl << 28); return &ah->ibah; } diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index d68f506c1922..9c2e53d28f98 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -527,7 +527,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, tun_tx_ix = (++tun_qp->tx_ix_head) & (MLX4_NUM_TUNNEL_BUFS - 1); spin_unlock(&tun_qp->tx_lock); if (ret) - goto out; + goto end; tun_mad = (struct mlx4_rcv_tunnel_mad *) (tun_qp->tx_ring[tun_tx_ix].buf.addr); if (tun_qp->tx_ring[tun_tx_ix].ah) @@ -596,9 +596,15 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, wr.wr.send_flags = IB_SEND_SIGNALED; ret = ib_post_send(src_qp, &wr.wr, &bad_wr); -out: - if (ret) - ib_destroy_ah(ah); + if (!ret) + return 0; + out: + spin_lock(&tun_qp->tx_lock); + tun_qp->tx_ix_tail++; + spin_unlock(&tun_qp->tx_lock); + tun_qp->tx_ring[tun_tx_ix].ah = NULL; +end: + ib_destroy_ah(ah); return ret; } @@ -1326,9 +1332,15 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, ret = ib_post_send(send_qp, &wr.wr, &bad_wr); + if (!ret) + return 0; + + spin_lock(&sqp->tx_lock); + sqp->tx_ix_tail++; + spin_unlock(&sqp->tx_lock); + sqp->tx_ring[wire_tx_ix].ah = NULL; out: - if (ret) - ib_destroy_ah(ah); + ib_destroy_ah(ah); return ret; } diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index b01ef6eee6e8..42a46078d7d5 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -505,9 +505,9 @@ static int mlx4_ib_query_device(struct ib_device *ibdev, props->device_cap_flags |= IB_DEVICE_MEM_WINDOW_TYPE_2B; else props->device_cap_flags |= IB_DEVICE_MEM_WINDOW_TYPE_2A; - if (dev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED) - props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; } + if (dev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED) + props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; @@ -1704,6 +1704,9 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp, struct mlx4_dev *dev = (to_mdev(qp->device))->dev; int is_bonded = mlx4_is_bonded(dev); + if (flow_attr->port < 1 || flow_attr->port > qp->device->phys_port_cnt) + return ERR_PTR(-EINVAL); + if ((flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP) && (flow_attr->type != IB_FLOW_ATTR_NORMAL)) return ERR_PTR(-EOPNOTSUPP); diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h index 6c5ac5d8f32f..29acda249612 100644 --- a/drivers/infiniband/hw/mlx4/mlx4_ib.h +++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h @@ -139,7 +139,7 @@ struct mlx4_ib_mr { u32 max_pages; struct mlx4_mr mmr; struct ib_umem *umem; - void *pages_alloc; + size_t page_map_size; }; struct mlx4_ib_mw { diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c index 631272172a0b..5d73989d9771 100644 --- a/drivers/infiniband/hw/mlx4/mr.c +++ b/drivers/infiniband/hw/mlx4/mr.c @@ -277,20 +277,23 @@ mlx4_alloc_priv_pages(struct ib_device *device, struct mlx4_ib_mr *mr, int max_pages) { - int size = max_pages * sizeof(u64); - int add_size; int ret; - add_size = max_t(int, MLX4_MR_PAGES_ALIGN - ARCH_KMALLOC_MINALIGN, 0); + /* Ensure that size is aligned to DMA cacheline + * requirements. + * max_pages is limited to MLX4_MAX_FAST_REG_PAGES + * so page_map_size will never cross PAGE_SIZE. + */ + mr->page_map_size = roundup(max_pages * sizeof(u64), + MLX4_MR_PAGES_ALIGN); - mr->pages_alloc = kzalloc(size + add_size, GFP_KERNEL); - if (!mr->pages_alloc) + /* Prevent cross page boundary allocation. */ + mr->pages = (__be64 *)get_zeroed_page(GFP_KERNEL); + if (!mr->pages) return -ENOMEM; - mr->pages = PTR_ALIGN(mr->pages_alloc, MLX4_MR_PAGES_ALIGN); - mr->page_map = dma_map_single(device->dma_device, mr->pages, - size, DMA_TO_DEVICE); + mr->page_map_size, DMA_TO_DEVICE); if (dma_mapping_error(device->dma_device, mr->page_map)) { ret = -ENOMEM; @@ -298,9 +301,9 @@ mlx4_alloc_priv_pages(struct ib_device *device, } return 0; -err: - kfree(mr->pages_alloc); +err: + free_page((unsigned long)mr->pages); return ret; } @@ -309,11 +312,10 @@ mlx4_free_priv_pages(struct mlx4_ib_mr *mr) { if (mr->pages) { struct ib_device *device = mr->ibmr.device; - int size = mr->max_pages * sizeof(u64); dma_unmap_single(device->dma_device, mr->page_map, - size, DMA_TO_DEVICE); - kfree(mr->pages_alloc); + mr->page_map_size, DMA_TO_DEVICE); + free_page((unsigned long)mr->pages); mr->pages = NULL; } } @@ -537,14 +539,12 @@ int mlx4_ib_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents, mr->npages = 0; ib_dma_sync_single_for_cpu(ibmr->device, mr->page_map, - sizeof(u64) * mr->max_pages, - DMA_TO_DEVICE); + mr->page_map_size, DMA_TO_DEVICE); rc = ib_sg_to_pages(ibmr, sg, sg_nents, sg_offset, mlx4_set_page); ib_dma_sync_single_for_device(ibmr->device, mr->page_map, - sizeof(u64) * mr->max_pages, - DMA_TO_DEVICE); + mr->page_map_size, DMA_TO_DEVICE); return rc; } diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index 81b0e1fbec1d..8db8405c1e99 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -362,7 +362,7 @@ static int send_wqe_overhead(enum mlx4_ib_qp_type type, u32 flags) sizeof (struct mlx4_wqe_raddr_seg); case MLX4_IB_QPT_RC: return sizeof (struct mlx4_wqe_ctrl_seg) + - sizeof (struct mlx4_wqe_atomic_seg) + + sizeof (struct mlx4_wqe_masked_atomic_seg) + sizeof (struct mlx4_wqe_raddr_seg); case MLX4_IB_QPT_SMI: case MLX4_IB_QPT_GSI: @@ -1191,8 +1191,10 @@ static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd, { err = create_qp_common(to_mdev(pd->device), pd, init_attr, udata, 0, &qp, gfp); - if (err) + if (err) { + kfree(qp); return ERR_PTR(err); + } qp->ibqp.qp_num = qp->mqp.qpn; qp->xrcdn = xrcdn; diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c index dabcc65bd65e..9c0e67bd2ba7 100644 --- a/drivers/infiniband/hw/mlx5/cq.c +++ b/drivers/infiniband/hw/mlx5/cq.c @@ -822,7 +822,8 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int eqn; int err; - if (entries < 0) + if (entries < 0 || + (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)))) return ERR_PTR(-EINVAL); if (check_cq_create_flags(attr->flags)) @@ -1168,11 +1169,16 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata) return -ENOSYS; } - if (entries < 1) + if (entries < 1 || + entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz))) { + mlx5_ib_warn(dev, "wrong entries number %d, max %d\n", + entries, + 1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)); return -EINVAL; + } entries = roundup_pow_of_two(entries + 1); - if (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)) + 1) + if (entries > (1 << MLX5_CAP_GEN(dev->mdev, log_max_cq_sz)) + 1) return -EINVAL; if (entries == ibcq->cqe + 1) diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c index 1534af113058..364aab9f3c9e 100644 --- a/drivers/infiniband/hw/mlx5/mad.c +++ b/drivers/infiniband/hw/mlx5/mad.c @@ -121,7 +121,7 @@ static void pma_cnt_ext_assign(struct ib_pma_portcounters_ext *pma_cnt_ext, pma_cnt_ext->port_xmit_data = cpu_to_be64(MLX5_SUM_CNT(out, transmitted_ib_unicast.octets, transmitted_ib_multicast.octets) >> 2); - pma_cnt_ext->port_xmit_data = + pma_cnt_ext->port_rcv_data = cpu_to_be64(MLX5_SUM_CNT(out, received_ib_unicast.octets, received_ib_multicast.octets) >> 2); pma_cnt_ext->port_xmit_packets = diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c index c72797cd9e4f..b48ad85315dc 100644 --- a/drivers/infiniband/hw/mlx5/main.c +++ b/drivers/infiniband/hw/mlx5/main.c @@ -524,6 +524,9 @@ static int mlx5_ib_query_device(struct ib_device *ibdev, MLX5_CAP_ETH(dev->mdev, scatter_fcs)) props->device_cap_flags |= IB_DEVICE_RAW_SCATTER_FCS; + if (mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS)) + props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; + props->vendor_part_id = mdev->pdev->device; props->hw_ver = mdev->pdev->revision; @@ -915,7 +918,8 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE; gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE; resp.qp_tab_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp); - resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size); + if (mlx5_core_is_pf(dev->mdev) && MLX5_CAP_GEN(dev->mdev, bf)) + resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size); resp.cache_line_size = L1_CACHE_BYTES; resp.max_sq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq); resp.max_rq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_rq); @@ -988,7 +992,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, if (field_avail(typeof(resp), cqe_version, udata->outlen)) resp.response_length += sizeof(resp.cqe_version); - if (field_avail(typeof(resp), hca_core_clock_offset, udata->outlen)) { + /* + * We don't want to expose information from the PCI bar that is located + * after 4096 bytes, so if the arch only supports larger pages, let's + * pretend we don't support reading the HCA's core clock. This is also + * forced by mmap function. + */ + if (PAGE_SIZE <= 4096 && + field_avail(typeof(resp), hca_core_clock_offset, udata->outlen)) { resp.comp_mask |= MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET; resp.hca_core_clock_offset = @@ -1798,7 +1809,7 @@ static ssize_t show_fw_ver(struct device *device, struct device_attribute *attr, { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); - return sprintf(buf, "%d.%d.%d\n", fw_rev_maj(dev->mdev), + return sprintf(buf, "%d.%d.%04d\n", fw_rev_maj(dev->mdev), fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev)); } @@ -1866,14 +1877,11 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, break; case MLX5_DEV_EVENT_PORT_DOWN: + case MLX5_DEV_EVENT_PORT_INITIALIZED: ibev.event = IB_EVENT_PORT_ERR; port = (u8)param; break; - case MLX5_DEV_EVENT_PORT_INITIALIZED: - /* not used by ULPs */ - return; - case MLX5_DEV_EVENT_LID_CHANGE: ibev.event = IB_EVENT_LID_CHANGE; port = (u8)param; diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 504117657d41..ce0a7ab35a22 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -235,6 +235,8 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap, qp->rq.max_gs = 0; qp->rq.wqe_cnt = 0; qp->rq.wqe_shift = 0; + cap->max_recv_wr = 0; + cap->max_recv_sge = 0; } else { if (ucmd) { qp->rq.wqe_cnt = ucmd->rq_wqe_count; @@ -1851,13 +1853,15 @@ static int modify_raw_packet_eth_prio(struct mlx5_core_dev *dev, static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, const struct ib_ah_attr *ah, struct mlx5_qp_path *path, u8 port, int attr_mask, - u32 path_flags, const struct ib_qp_attr *attr) + u32 path_flags, const struct ib_qp_attr *attr, + bool alt) { enum rdma_link_layer ll = rdma_port_get_link_layer(&dev->ib_dev, port); int err; if (attr_mask & IB_QP_PKEY_INDEX) - path->pkey_index = attr->pkey_index; + path->pkey_index = cpu_to_be16(alt ? attr->alt_pkey_index : + attr->pkey_index); if (ah->ah_flags & IB_AH_GRH) { if (ah->grh.sgid_index >= @@ -1877,9 +1881,9 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, ah->grh.sgid_index); path->dci_cfi_prio_sl = (ah->sl & 0x7) << 4; } else { - path->fl = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0; - path->free_ar = (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x80 : - 0; + path->fl_free_ar = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0; + path->fl_free_ar |= + (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x40 : 0; path->rlid = cpu_to_be16(ah->dlid); path->grh_mlid = ah->src_path_bits & 0x7f; if (ah->ah_flags & IB_AH_GRH) @@ -1903,7 +1907,7 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, path->port = port; if (attr_mask & IB_QP_TIMEOUT) - path->ackto_lt = attr->timeout << 3; + path->ackto_lt = (alt ? attr->alt_timeout : attr->timeout) << 3; if ((qp->ibqp.qp_type == IB_QPT_RAW_PACKET) && qp->sq.wqe_cnt) return modify_raw_packet_eth_prio(dev->mdev, @@ -2264,7 +2268,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, context->log_pg_sz_remote_qpn = cpu_to_be32(attr->dest_qp_num); if (attr_mask & IB_QP_PKEY_INDEX) - context->pri_path.pkey_index = attr->pkey_index; + context->pri_path.pkey_index = cpu_to_be16(attr->pkey_index); /* todo implement counter_index functionality */ @@ -2277,7 +2281,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, if (attr_mask & IB_QP_AV) { err = mlx5_set_path(dev, qp, &attr->ah_attr, &context->pri_path, attr_mask & IB_QP_PORT ? attr->port_num : qp->port, - attr_mask, 0, attr); + attr_mask, 0, attr, false); if (err) goto out; } @@ -2288,7 +2292,9 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp, if (attr_mask & IB_QP_ALT_PATH) { err = mlx5_set_path(dev, qp, &attr->alt_ah_attr, &context->alt_path, - attr->alt_port_num, attr_mask, 0, attr); + attr->alt_port_num, + attr_mask | IB_QP_PKEY_INDEX | IB_QP_TIMEOUT, + 0, attr, true); if (err) goto out; } @@ -3326,10 +3332,11 @@ static u8 get_fence(u8 fence, struct ib_send_wr *wr) return MLX5_FENCE_MODE_SMALL_AND_FENCE; else return fence; - - } else { - return 0; + } else if (unlikely(wr->send_flags & IB_SEND_FENCE)) { + return MLX5_FENCE_MODE_FENCE; } + + return 0; } static int begin_wqe(struct mlx5_ib_qp *qp, void **seg, @@ -4013,11 +4020,12 @@ static int query_qp_attr(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp, if (qp->ibqp.qp_type == IB_QPT_RC || qp->ibqp.qp_type == IB_QPT_UC) { to_ib_ah_attr(dev, &qp_attr->ah_attr, &context->pri_path); to_ib_ah_attr(dev, &qp_attr->alt_ah_attr, &context->alt_path); - qp_attr->alt_pkey_index = context->alt_path.pkey_index & 0x7f; + qp_attr->alt_pkey_index = + be16_to_cpu(context->alt_path.pkey_index); qp_attr->alt_port_num = qp_attr->alt_ah_attr.port_num; } - qp_attr->pkey_index = context->pri_path.pkey_index & 0x7f; + qp_attr->pkey_index = be16_to_cpu(context->pri_path.pkey_index); qp_attr->port_num = context->pri_path.port; /* qp_attr->en_sqd_async_notify is only applicable in modify qp */ @@ -4079,17 +4087,19 @@ int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, qp_attr->cap.max_recv_sge = qp->rq.max_gs; if (!ibqp->uobject) { - qp_attr->cap.max_send_wr = qp->sq.wqe_cnt; + qp_attr->cap.max_send_wr = qp->sq.max_post; qp_attr->cap.max_send_sge = qp->sq.max_gs; + qp_init_attr->qp_context = ibqp->qp_context; } else { qp_attr->cap.max_send_wr = 0; qp_attr->cap.max_send_sge = 0; } - /* We don't support inline sends for kernel QPs (yet), and we - * don't know what userspace's value should be. - */ - qp_attr->cap.max_inline_data = 0; + qp_init_attr->qp_type = ibqp->qp_type; + qp_init_attr->recv_cq = ibqp->recv_cq; + qp_init_attr->send_cq = ibqp->send_cq; + qp_init_attr->srq = ibqp->srq; + qp_attr->cap.max_inline_data = qp->max_inline_data; qp_init_attr->cap = qp_attr->cap; diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c index ff946d5f59e4..382466a90da7 100644 --- a/drivers/infiniband/hw/qib/qib_file_ops.c +++ b/drivers/infiniband/hw/qib/qib_file_ops.c @@ -2178,6 +2178,11 @@ static ssize_t qib_write(struct file *fp, const char __user *data, switch (cmd.type) { case QIB_CMD_ASSIGN_CTXT: + if (rcd) { + ret = -EINVAL; + goto bail; + } + ret = qib_assign_ctxt(fp, &cmd.cmd.user_info); if (ret) goto bail; diff --git a/drivers/infiniband/hw/usnic/usnic_uiom.c b/drivers/infiniband/hw/usnic/usnic_uiom.c index 7209fbc03ccb..a0b6ebee4d8a 100644 --- a/drivers/infiniband/hw/usnic/usnic_uiom.c +++ b/drivers/infiniband/hw/usnic/usnic_uiom.c @@ -36,7 +36,6 @@ #include <linux/dma-mapping.h> #include <linux/sched.h> #include <linux/hugetlb.h> -#include <linux/dma-attrs.h> #include <linux/iommu.h> #include <linux/workqueue.h> #include <linux/list.h> @@ -112,10 +111,6 @@ static int usnic_uiom_get_pages(unsigned long addr, size_t size, int writable, int i; int flags; dma_addr_t pa; - DEFINE_DMA_ATTRS(attrs); - - if (dmasync) - dma_set_attr(DMA_ATTR_WRITE_BARRIER, &attrs); if (!can_do_mlock()) return -EPERM; diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index 5fa4d4d81ee0..41ba7e9cadaa 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -369,8 +369,8 @@ static int alloc_qpn(struct rvt_dev_info *rdi, struct rvt_qpn_table *qpt, /* wrap to first map page, invert bit 0 */ offset = qpt->incr | ((offset & 1) ^ 1); } - /* there can be no bits at shift and below */ - WARN_ON(offset & (rdi->dparms.qos_shift - 1)); + /* there can be no set bits in low-order QoS bits */ + WARN_ON(offset & (BIT(rdi->dparms.qos_shift) - 1)); qpn = mk_qpn(qpt, map, offset); } @@ -502,6 +502,12 @@ static void rvt_remove_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp) */ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, enum ib_qp_type type) + __releases(&qp->s_lock) + __releases(&qp->s_hlock) + __releases(&qp->r_lock) + __acquires(&qp->r_lock) + __acquires(&qp->s_hlock) + __acquires(&qp->s_lock) { if (qp->state != IB_QPS_RESET) { qp->state = IB_QPS_RESET; @@ -570,12 +576,6 @@ static void rvt_reset_qp(struct rvt_dev_info *rdi, struct rvt_qp *qp, qp->s_ssn = 1; qp->s_lsn = 0; qp->s_mig_state = IB_MIG_MIGRATED; - if (qp->s_ack_queue) - memset( - qp->s_ack_queue, - 0, - rvt_max_atomic(rdi) * - sizeof(*qp->s_ack_queue)); qp->r_head_ack_queue = 0; qp->s_tail_ack_queue = 0; qp->s_num_rd_atomic = 0; @@ -699,8 +699,10 @@ struct ib_qp *rvt_create_qp(struct ib_pd *ibpd, * initialization that is needed. */ priv = rdi->driver_f.qp_priv_alloc(rdi, qp, gfp); - if (!priv) + if (IS_ERR(priv)) { + ret = priv; goto bail_qp; + } qp->priv = priv; qp->timeout_jiffies = usecs_to_jiffies((4096UL * (1UL << qp->timeout)) / diff --git a/drivers/infiniband/sw/rdmavt/vt.c b/drivers/infiniband/sw/rdmavt/vt.c index e1cc2cc42f25..30c4fda7a05a 100644 --- a/drivers/infiniband/sw/rdmavt/vt.c +++ b/drivers/infiniband/sw/rdmavt/vt.c @@ -501,9 +501,7 @@ static noinline int check_support(struct rvt_dev_info *rdi, int verb) !rdi->driver_f.quiesce_qp || !rdi->driver_f.notify_error_qp || !rdi->driver_f.mtu_from_qp || - !rdi->driver_f.mtu_to_path_mtu || - !rdi->driver_f.shut_down_port || - !rdi->driver_f.cap_mask_chg) + !rdi->driver_f.mtu_to_path_mtu) return -EINVAL; break; diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index bab7db6fa9ab..4f7d9b48df64 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -94,6 +94,7 @@ enum { IPOIB_NEIGH_TBL_FLUSH = 12, IPOIB_FLAG_DEV_ADDR_SET = 13, IPOIB_FLAG_DEV_ADDR_CTRL = 14, + IPOIB_FLAG_GOING_DOWN = 15, IPOIB_MAX_BACKOFF_SECONDS = 16, diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index b2f42835d76d..951d9abcca8b 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -1486,6 +1486,10 @@ static ssize_t set_mode(struct device *d, struct device_attribute *attr, { struct net_device *dev = to_net_dev(d); int ret; + struct ipoib_dev_priv *priv = netdev_priv(dev); + + if (test_bit(IPOIB_FLAG_GOING_DOWN, &priv->flags)) + return -EPERM; if (!rtnl_trylock()) return restart_syscall(); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index 45c40a17d6a6..dc6d241b9406 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -1015,7 +1015,7 @@ static bool ipoib_dev_addr_changed_valid(struct ipoib_dev_priv *priv) if (ib_query_gid(priv->ca, priv->port, 0, &gid0, NULL)) return false; - netif_addr_lock(priv->dev); + netif_addr_lock_bh(priv->dev); /* The subnet prefix may have changed, update it now so we won't have * to do it later @@ -1026,12 +1026,12 @@ static bool ipoib_dev_addr_changed_valid(struct ipoib_dev_priv *priv) search_gid.global.interface_id = priv->local_gid.global.interface_id; - netif_addr_unlock(priv->dev); + netif_addr_unlock_bh(priv->dev); err = ib_find_gid(priv->ca, &search_gid, IB_GID_TYPE_IB, priv->dev, &port, &index); - netif_addr_lock(priv->dev); + netif_addr_lock_bh(priv->dev); if (search_gid.global.interface_id != priv->local_gid.global.interface_id) @@ -1092,7 +1092,7 @@ static bool ipoib_dev_addr_changed_valid(struct ipoib_dev_priv *priv) } out: - netif_addr_unlock(priv->dev); + netif_addr_unlock_bh(priv->dev); return ret; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 2d7c16346648..5f58c41ef787 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1206,7 +1206,9 @@ struct ipoib_neigh *ipoib_neigh_get(struct net_device *dev, u8 *daddr) neigh = NULL; goto out_unlock; } - neigh->alive = jiffies; + + if (likely(skb_queue_len(&neigh->queue) < IPOIB_MAX_PATH_REC_QUEUE)) + neigh->alive = jiffies; goto out_unlock; } } @@ -1851,7 +1853,7 @@ static void set_base_guid(struct ipoib_dev_priv *priv, union ib_gid *gid) struct ipoib_dev_priv *child_priv; struct net_device *netdev = priv->dev; - netif_addr_lock(netdev); + netif_addr_lock_bh(netdev); memcpy(&priv->local_gid.global.interface_id, &gid->global.interface_id, @@ -1859,7 +1861,7 @@ static void set_base_guid(struct ipoib_dev_priv *priv, union ib_gid *gid) memcpy(netdev->dev_addr + 4, &priv->local_gid, sizeof(priv->local_gid)); clear_bit(IPOIB_FLAG_DEV_ADDR_SET, &priv->flags); - netif_addr_unlock(netdev); + netif_addr_unlock_bh(netdev); if (!test_bit(IPOIB_FLAG_SUBINTERFACE, &priv->flags)) { down_read(&priv->vlan_rwsem); @@ -1875,7 +1877,7 @@ static int ipoib_check_lladdr(struct net_device *dev, union ib_gid *gid = (union ib_gid *)(ss->__data + 4); int ret = 0; - netif_addr_lock(dev); + netif_addr_lock_bh(dev); /* Make sure the QPN, reserved and subnet prefix match the current * lladdr, it also makes sure the lladdr is unicast. @@ -1885,7 +1887,7 @@ static int ipoib_check_lladdr(struct net_device *dev, gid->global.interface_id == 0) ret = -EINVAL; - netif_addr_unlock(dev); + netif_addr_unlock_bh(dev); return ret; } @@ -2141,6 +2143,9 @@ static void ipoib_remove_one(struct ib_device *device, void *client_data) ib_unregister_event_handler(&priv->event_handler); flush_workqueue(ipoib_workqueue); + /* mark interface in the middle of destruction */ + set_bit(IPOIB_FLAG_GOING_DOWN, &priv->flags); + rtnl_lock(); dev_change_flags(priv->dev, priv->dev->flags & ~IFF_UP); rtnl_unlock(); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 82fbc9442608..d3394b6add24 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -582,13 +582,13 @@ void ipoib_mcast_join_task(struct work_struct *work) return; } priv->local_lid = port_attr.lid; - netif_addr_lock(dev); + netif_addr_lock_bh(dev); if (!test_bit(IPOIB_FLAG_DEV_ADDR_SET, &priv->flags)) { - netif_addr_unlock(dev); + netif_addr_unlock_bh(dev); return; } - netif_addr_unlock(dev); + netif_addr_unlock_bh(dev); spin_lock_irq(&priv->lock); if (!test_bit(IPOIB_FLAG_OPER_UP, &priv->flags)) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index 64a35595eab8..a2f9f29c6ab5 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -131,6 +131,9 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) ppriv = netdev_priv(pdev); + if (test_bit(IPOIB_FLAG_GOING_DOWN, &ppriv->flags)) + return -EPERM; + snprintf(intf_name, sizeof intf_name, "%s.%04x", ppriv->dev->name, pkey); priv = ipoib_intf_alloc(intf_name); @@ -183,6 +186,9 @@ int ipoib_vlan_delete(struct net_device *pdev, unsigned short pkey) ppriv = netdev_priv(pdev); + if (test_bit(IPOIB_FLAG_GOING_DOWN, &ppriv->flags)) + return -EPERM; + if (!rtnl_trylock()) return restart_syscall(); diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index 646de170ec12..3322ed750172 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1457,7 +1457,6 @@ static int srp_map_sg_fr(struct srp_map_state *state, struct srp_rdma_ch *ch, { unsigned int sg_offset = 0; - state->desc = req->indirect_desc; state->fr.next = req->fr_list; state->fr.end = req->fr_list + ch->target->mr_per_cmd; state->sg = scat; @@ -1489,7 +1488,6 @@ static int srp_map_sg_dma(struct srp_map_state *state, struct srp_rdma_ch *ch, struct scatterlist *sg; int i; - state->desc = req->indirect_desc; for_each_sg(scat, sg, count, i) { srp_map_desc(state, ib_sg_dma_address(dev->dev, sg), ib_sg_dma_len(dev->dev, sg), @@ -1655,6 +1653,7 @@ static int srp_map_data(struct scsi_cmnd *scmnd, struct srp_rdma_ch *ch, target->indirect_size, DMA_TO_DEVICE); memset(&state, 0, sizeof(state)); + state.desc = req->indirect_desc; if (dev->use_fast_reg) ret = srp_map_sg_fr(&state, ch, req, scat, count); else if (dev->use_fmr) @@ -3526,7 +3525,7 @@ static void srp_add_one(struct ib_device *device) int mr_page_shift, p; u64 max_pages_per_mr; - srp_dev = kmalloc(sizeof *srp_dev, GFP_KERNEL); + srp_dev = kzalloc(sizeof(*srp_dev), GFP_KERNEL); if (!srp_dev) return; @@ -3586,8 +3585,6 @@ static void srp_add_one(struct ib_device *device) IB_ACCESS_REMOTE_WRITE); if (IS_ERR(srp_dev->global_mr)) goto err_pd; - } else { - srp_dev->global_mr = NULL; } for (p = rdma_start_port(device); p <= rdma_end_port(device); ++p) { diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index e68b20cba70b..4a4155640d51 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -1638,8 +1638,7 @@ retry: */ qp_init->cap.max_send_wr = srp_sq_size / 2; qp_init->cap.max_rdma_ctxs = srp_sq_size / 2; - qp_init->cap.max_send_sge = max(sdev->device->attrs.max_sge_rd, - sdev->device->attrs.max_sge); + qp_init->cap.max_send_sge = SRPT_DEF_SG_PER_WQE; qp_init->port_num = ch->sport->port; ch->qp = ib_create_qp(sdev->pd, qp_init); diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.h b/drivers/infiniband/ulp/srpt/ib_srpt.h index fee6bfd7ca21..389030487da7 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.h +++ b/drivers/infiniband/ulp/srpt/ib_srpt.h @@ -106,6 +106,7 @@ enum { SRP_LOGIN_RSP_MULTICHAN_MAINTAINED = 0x2, SRPT_DEF_SG_TABLESIZE = 128, + SRPT_DEF_SG_PER_WQE = 16, MIN_SRPT_SQ_SIZE = 16, DEF_SRPT_SQ_SIZE = 4096, diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c index 94b68213c50d..5f6b3bcab078 100644 --- a/drivers/iommu/arm-smmu-v3.c +++ b/drivers/iommu/arm-smmu-v3.c @@ -1941,6 +1941,7 @@ static struct iommu_ops arm_smmu_ops = { .attach_dev = arm_smmu_attach_dev, .map = arm_smmu_map, .unmap = arm_smmu_unmap, + .map_sg = default_iommu_map_sg, .iova_to_phys = arm_smmu_iova_to_phys, .add_device = arm_smmu_add_device, .remove_device = arm_smmu_remove_device, diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c index a644d0cec2d8..10700945994e 100644 --- a/drivers/iommu/intel-iommu.c +++ b/drivers/iommu/intel-iommu.c @@ -3222,11 +3222,6 @@ static int __init init_dmars(void) } } - iommu_flush_write_buffer(iommu); - iommu_set_root_entry(iommu); - iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL); - iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH); - if (!ecap_pass_through(iommu->ecap)) hw_pass_through = 0; #ifdef CONFIG_INTEL_IOMMU_SVM @@ -3235,6 +3230,18 @@ static int __init init_dmars(void) #endif } + /* + * Now that qi is enabled on all iommus, set the root entry and flush + * caches. This is required on some Intel X58 chipsets, otherwise the + * flush_context function will loop forever and the boot hangs. + */ + for_each_active_iommu(iommu, drhd) { + iommu_flush_write_buffer(iommu); + iommu_set_root_entry(iommu); + iommu->flush.flush_context(iommu, 0, 0, 0, DMA_CCMD_GLOBAL_INVL); + iommu->flush.flush_iotlb(iommu, 0, 0, 0, DMA_TLB_GLOBAL_FLUSH); + } + if (iommu_pass_through) iommu_identity_mapping |= IDENTMAP_ALL; diff --git a/drivers/iommu/rockchip-iommu.c b/drivers/iommu/rockchip-iommu.c index c7d6156ff536..25b4627cb57f 100644 --- a/drivers/iommu/rockchip-iommu.c +++ b/drivers/iommu/rockchip-iommu.c @@ -815,7 +815,7 @@ static int rk_iommu_attach_device(struct iommu_domain *domain, dte_addr = virt_to_phys(rk_domain->dt); for (i = 0; i < iommu->num_mmu; i++) { rk_iommu_write(iommu->bases[i], RK_MMU_DTE_ADDR, dte_addr); - rk_iommu_command(iommu->bases[i], RK_MMU_CMD_ZAP_CACHE); + rk_iommu_base_command(iommu->bases[i], RK_MMU_CMD_ZAP_CACHE); rk_iommu_write(iommu->bases[i], RK_MMU_INT_MASK, RK_MMU_IRQ_MASK); } diff --git a/drivers/irqchip/irq-mips-gic.c b/drivers/irqchip/irq-mips-gic.c index 3b5e10aa48ab..8a4adbeb2b8c 100644 --- a/drivers/irqchip/irq-mips-gic.c +++ b/drivers/irqchip/irq-mips-gic.c @@ -746,6 +746,12 @@ static int gic_irq_domain_alloc(struct irq_domain *d, unsigned int virq, /* verify that it doesn't conflict with an IPI irq */ if (test_bit(spec->hwirq, ipi_resrv)) return -EBUSY; + + hwirq = GIC_SHARED_TO_HWIRQ(spec->hwirq); + + return irq_domain_set_hwirq_and_chip(d, virq, hwirq, + &gic_level_irq_controller, + NULL); } else { base_hwirq = find_first_bit(ipi_resrv, gic_shared_intrs); if (base_hwirq == gic_shared_intrs) { @@ -867,10 +873,14 @@ static int gic_dev_domain_alloc(struct irq_domain *d, unsigned int virq, &gic_level_irq_controller, NULL); if (ret) - return ret; + goto error; } return 0; + +error: + irq_domain_free_irqs_parent(d, virq, nr_irqs); + return ret; } void gic_dev_domain_free(struct irq_domain *d, unsigned int virq, diff --git a/drivers/leds/led-core.c b/drivers/leds/led-core.c index 3495d5d6547f..3bce44893021 100644 --- a/drivers/leds/led-core.c +++ b/drivers/leds/led-core.c @@ -53,11 +53,12 @@ static void led_timer_function(unsigned long data) if (!led_cdev->blink_delay_on || !led_cdev->blink_delay_off) { led_set_brightness_nosleep(led_cdev, LED_OFF); + led_cdev->flags &= ~LED_BLINK_SW; return; } if (led_cdev->flags & LED_BLINK_ONESHOT_STOP) { - led_cdev->flags &= ~LED_BLINK_ONESHOT_STOP; + led_cdev->flags &= ~(LED_BLINK_ONESHOT_STOP | LED_BLINK_SW); return; } @@ -151,6 +152,7 @@ static void led_set_software_blink(struct led_classdev *led_cdev, return; } + led_cdev->flags |= LED_BLINK_SW; mod_timer(&led_cdev->blink_timer, jiffies + 1); } @@ -219,6 +221,7 @@ void led_stop_software_blink(struct led_classdev *led_cdev) del_timer_sync(&led_cdev->blink_timer); led_cdev->blink_delay_on = 0; led_cdev->blink_delay_off = 0; + led_cdev->flags &= ~LED_BLINK_SW; } EXPORT_SYMBOL_GPL(led_stop_software_blink); @@ -226,10 +229,10 @@ void led_set_brightness(struct led_classdev *led_cdev, enum led_brightness brightness) { /* - * In case blinking is on delay brightness setting + * If software blink is active, delay brightness setting * until the next timer tick. */ - if (led_cdev->blink_delay_on || led_cdev->blink_delay_off) { + if (led_cdev->flags & LED_BLINK_SW) { /* * If we need to disable soft blinking delegate this to the * work queue task to avoid problems in case we are called diff --git a/drivers/leds/trigger/ledtrig-heartbeat.c b/drivers/leds/trigger/ledtrig-heartbeat.c index 410c39c62dc7..c9f386213e9e 100644 --- a/drivers/leds/trigger/ledtrig-heartbeat.c +++ b/drivers/leds/trigger/ledtrig-heartbeat.c @@ -19,6 +19,7 @@ #include <linux/sched.h> #include <linux/leds.h> #include <linux/reboot.h> +#include <linux/suspend.h> #include "../leds.h" static int panic_heartbeats; @@ -154,6 +155,30 @@ static struct led_trigger heartbeat_led_trigger = { .deactivate = heartbeat_trig_deactivate, }; +static int heartbeat_pm_notifier(struct notifier_block *nb, + unsigned long pm_event, void *unused) +{ + int rc; + + switch (pm_event) { + case PM_SUSPEND_PREPARE: + case PM_HIBERNATION_PREPARE: + case PM_RESTORE_PREPARE: + led_trigger_unregister(&heartbeat_led_trigger); + break; + case PM_POST_SUSPEND: + case PM_POST_HIBERNATION: + case PM_POST_RESTORE: + rc = led_trigger_register(&heartbeat_led_trigger); + if (rc) + pr_err("could not re-register heartbeat trigger\n"); + break; + default: + break; + } + return NOTIFY_DONE; +} + static int heartbeat_reboot_notifier(struct notifier_block *nb, unsigned long code, void *unused) { @@ -168,6 +193,10 @@ static int heartbeat_panic_notifier(struct notifier_block *nb, return NOTIFY_DONE; } +static struct notifier_block heartbeat_pm_nb = { + .notifier_call = heartbeat_pm_notifier, +}; + static struct notifier_block heartbeat_reboot_nb = { .notifier_call = heartbeat_reboot_notifier, }; @@ -184,12 +213,14 @@ static int __init heartbeat_trig_init(void) atomic_notifier_chain_register(&panic_notifier_list, &heartbeat_panic_nb); register_reboot_notifier(&heartbeat_reboot_nb); + register_pm_notifier(&heartbeat_pm_nb); } return rc; } static void __exit heartbeat_trig_exit(void) { + unregister_pm_notifier(&heartbeat_pm_nb); unregister_reboot_notifier(&heartbeat_reboot_nb); atomic_notifier_chain_unregister(&panic_notifier_list, &heartbeat_panic_nb); diff --git a/drivers/mcb/mcb-core.c b/drivers/mcb/mcb-core.c index b73c6e7d28e4..6f2c8522e14a 100644 --- a/drivers/mcb/mcb-core.c +++ b/drivers/mcb/mcb-core.c @@ -61,21 +61,36 @@ static int mcb_probe(struct device *dev) struct mcb_driver *mdrv = to_mcb_driver(dev->driver); struct mcb_device *mdev = to_mcb_device(dev); const struct mcb_device_id *found_id; + struct module *carrier_mod; + int ret; found_id = mcb_match_id(mdrv->id_table, mdev); if (!found_id) return -ENODEV; - return mdrv->probe(mdev, found_id); + carrier_mod = mdev->dev.parent->driver->owner; + if (!try_module_get(carrier_mod)) + return -EINVAL; + + get_device(dev); + ret = mdrv->probe(mdev, found_id); + if (ret) + module_put(carrier_mod); + + return ret; } static int mcb_remove(struct device *dev) { struct mcb_driver *mdrv = to_mcb_driver(dev->driver); struct mcb_device *mdev = to_mcb_device(dev); + struct module *carrier_mod; mdrv->remove(mdev); + carrier_mod = mdev->dev.parent->driver->owner; + module_put(carrier_mod); + put_device(&mdev->dev); return 0; diff --git a/drivers/media/usb/uvc/uvc_v4l2.c b/drivers/media/usb/uvc/uvc_v4l2.c index d7723ce772b3..c04bc6afb965 100644 --- a/drivers/media/usb/uvc/uvc_v4l2.c +++ b/drivers/media/usb/uvc/uvc_v4l2.c @@ -1274,8 +1274,6 @@ struct uvc_xu_control_mapping32 { static int uvc_v4l2_get_xu_mapping(struct uvc_xu_control_mapping *kp, const struct uvc_xu_control_mapping32 __user *up) { - struct uvc_menu_info __user *umenus; - struct uvc_menu_info __user *kmenus; compat_caddr_t p; if (!access_ok(VERIFY_READ, up, sizeof(*up)) || @@ -1292,17 +1290,7 @@ static int uvc_v4l2_get_xu_mapping(struct uvc_xu_control_mapping *kp, if (__get_user(p, &up->menu_info)) return -EFAULT; - umenus = compat_ptr(p); - if (!access_ok(VERIFY_READ, umenus, kp->menu_count * sizeof(*umenus))) - return -EFAULT; - - kmenus = compat_alloc_user_space(kp->menu_count * sizeof(*kmenus)); - if (kmenus == NULL) - return -EFAULT; - kp->menu_info = kmenus; - - if (copy_in_user(kmenus, umenus, kp->menu_count * sizeof(*umenus))) - return -EFAULT; + kp->menu_info = compat_ptr(p); return 0; } @@ -1310,10 +1298,6 @@ static int uvc_v4l2_get_xu_mapping(struct uvc_xu_control_mapping *kp, static int uvc_v4l2_put_xu_mapping(const struct uvc_xu_control_mapping *kp, struct uvc_xu_control_mapping32 __user *up) { - struct uvc_menu_info __user *umenus; - struct uvc_menu_info __user *kmenus = kp->menu_info; - compat_caddr_t p; - if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) || __copy_to_user(up, kp, offsetof(typeof(*up), menu_info)) || __put_user(kp->menu_count, &up->menu_count)) @@ -1322,16 +1306,6 @@ static int uvc_v4l2_put_xu_mapping(const struct uvc_xu_control_mapping *kp, if (__clear_user(up->reserved, sizeof(up->reserved))) return -EFAULT; - if (kp->menu_count == 0) - return 0; - - if (get_user(p, &up->menu_info)) - return -EFAULT; - umenus = compat_ptr(p); - - if (copy_in_user(umenus, kmenus, kp->menu_count * sizeof(*umenus))) - return -EFAULT; - return 0; } @@ -1346,8 +1320,6 @@ struct uvc_xu_control_query32 { static int uvc_v4l2_get_xu_query(struct uvc_xu_control_query *kp, const struct uvc_xu_control_query32 __user *up) { - u8 __user *udata; - u8 __user *kdata; compat_caddr_t p; if (!access_ok(VERIFY_READ, up, sizeof(*up)) || @@ -1361,17 +1333,7 @@ static int uvc_v4l2_get_xu_query(struct uvc_xu_control_query *kp, if (__get_user(p, &up->data)) return -EFAULT; - udata = compat_ptr(p); - if (!access_ok(VERIFY_READ, udata, kp->size)) - return -EFAULT; - - kdata = compat_alloc_user_space(kp->size); - if (kdata == NULL) - return -EFAULT; - kp->data = kdata; - - if (copy_in_user(kdata, udata, kp->size)) - return -EFAULT; + kp->data = compat_ptr(p); return 0; } @@ -1379,26 +1341,10 @@ static int uvc_v4l2_get_xu_query(struct uvc_xu_control_query *kp, static int uvc_v4l2_put_xu_query(const struct uvc_xu_control_query *kp, struct uvc_xu_control_query32 __user *up) { - u8 __user *udata; - u8 __user *kdata = kp->data; - compat_caddr_t p; - if (!access_ok(VERIFY_WRITE, up, sizeof(*up)) || __copy_to_user(up, kp, offsetof(typeof(*up), data))) return -EFAULT; - if (kp->size == 0) - return 0; - - if (get_user(p, &up->data)) - return -EFAULT; - udata = compat_ptr(p); - if (!access_ok(VERIFY_READ, udata, kp->size)) - return -EFAULT; - - if (copy_in_user(udata, kdata, kp->size)) - return -EFAULT; - return 0; } @@ -1408,47 +1354,44 @@ static int uvc_v4l2_put_xu_query(const struct uvc_xu_control_query *kp, static long uvc_v4l2_compat_ioctl32(struct file *file, unsigned int cmd, unsigned long arg) { + struct uvc_fh *handle = file->private_data; union { struct uvc_xu_control_mapping xmap; struct uvc_xu_control_query xqry; } karg; void __user *up = compat_ptr(arg); - mm_segment_t old_fs; long ret; switch (cmd) { case UVCIOC_CTRL_MAP32: - cmd = UVCIOC_CTRL_MAP; ret = uvc_v4l2_get_xu_mapping(&karg.xmap, up); + if (ret) + return ret; + ret = uvc_ioctl_ctrl_map(handle->chain, &karg.xmap); + if (ret) + return ret; + ret = uvc_v4l2_put_xu_mapping(&karg.xmap, up); + if (ret) + return ret; + break; case UVCIOC_CTRL_QUERY32: - cmd = UVCIOC_CTRL_QUERY; ret = uvc_v4l2_get_xu_query(&karg.xqry, up); + if (ret) + return ret; + ret = uvc_xu_ctrl_query(handle->chain, &karg.xqry); + if (ret) + return ret; + ret = uvc_v4l2_put_xu_query(&karg.xqry, up); + if (ret) + return ret; break; default: return -ENOIOCTLCMD; } - old_fs = get_fs(); - set_fs(KERNEL_DS); - ret = video_ioctl2(file, cmd, (unsigned long)&karg); - set_fs(old_fs); - - if (ret < 0) - return ret; - - switch (cmd) { - case UVCIOC_CTRL_MAP: - ret = uvc_v4l2_put_xu_mapping(&karg.xmap, up); - break; - - case UVCIOC_CTRL_QUERY: - ret = uvc_v4l2_put_xu_query(&karg.xqry, up); - break; - } - return ret; } #endif diff --git a/drivers/media/v4l2-core/v4l2-mc.c b/drivers/media/v4l2-core/v4l2-mc.c index ca94bded3386..8bef4331bd51 100644 --- a/drivers/media/v4l2-core/v4l2-mc.c +++ b/drivers/media/v4l2-core/v4l2-mc.c @@ -1,7 +1,7 @@ /* * Media Controller ancillary functions * - * Copyright (c) 2016 Mauro Carvalho Chehab <mchehab@osg.samsung.com> + * Copyright (c) 2016 Mauro Carvalho Chehab <mchehab@kernel.org> * Copyright (C) 2016 Shuah Khan <shuahkh@osg.samsung.com> * Copyright (C) 2006-2010 Nokia Corporation * Copyright (c) 2016 Intel Corporation. diff --git a/drivers/memory/omap-gpmc.c b/drivers/memory/omap-gpmc.c index af4884ba6b7c..15508df24e5d 100644 --- a/drivers/memory/omap-gpmc.c +++ b/drivers/memory/omap-gpmc.c @@ -398,7 +398,7 @@ static void gpmc_cs_bool_timings(int cs, const struct gpmc_bool_timings *p) gpmc_cs_modify_reg(cs, GPMC_CS_CONFIG4, GPMC_CONFIG4_OEEXTRADELAY, p->oe_extra_delay); gpmc_cs_modify_reg(cs, GPMC_CS_CONFIG4, - GPMC_CONFIG4_OEEXTRADELAY, p->we_extra_delay); + GPMC_CONFIG4_WEEXTRADELAY, p->we_extra_delay); gpmc_cs_modify_reg(cs, GPMC_CS_CONFIG6, GPMC_CONFIG6_CYCLE2CYCLESAMECSEN, p->cycle2cyclesamecsen); diff --git a/drivers/misc/mei/client.c b/drivers/misc/mei/client.c index eed254da63a8..641c1a566687 100644 --- a/drivers/misc/mei/client.c +++ b/drivers/misc/mei/client.c @@ -730,7 +730,7 @@ static void mei_cl_wake_all(struct mei_cl *cl) /* synchronized under device mutex */ if (waitqueue_active(&cl->wait)) { cl_dbg(dev, cl, "Waking up ctrl write clients!\n"); - wake_up_interruptible(&cl->wait); + wake_up(&cl->wait); } } diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c index 16baeb51b2bd..ef3618299494 100644 --- a/drivers/mtd/ubi/build.c +++ b/drivers/mtd/ubi/build.c @@ -1147,11 +1147,17 @@ int ubi_detach_mtd_dev(int ubi_num, int anyway) */ static struct mtd_info * __init open_mtd_by_chdev(const char *mtd_dev) { - struct kstat stat; int err, minor; + struct path path; + struct kstat stat; /* Probably this is an MTD character device node path */ - err = vfs_stat(mtd_dev, &stat); + err = kern_path(mtd_dev, LOOKUP_FOLLOW, &path); + if (err) + return ERR_PTR(err); + + err = vfs_getattr(&path, &stat); + path_put(&path); if (err) return ERR_PTR(err); @@ -1160,6 +1166,7 @@ static struct mtd_info * __init open_mtd_by_chdev(const char *mtd_dev) return ERR_PTR(-EINVAL); minor = MINOR(stat.rdev); + if (minor & 1) /* * Just do not think the "/dev/mtdrX" devices support is need, diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c index 5780dd1ba79d..ebf517271d29 100644 --- a/drivers/mtd/ubi/eba.c +++ b/drivers/mtd/ubi/eba.c @@ -575,6 +575,7 @@ static int recover_peb(struct ubi_device *ubi, int pnum, int vol_id, int lnum, int err, idx = vol_id2idx(ubi, vol_id), new_pnum, data_size, tries = 0; struct ubi_volume *vol = ubi->volumes[idx]; struct ubi_vid_hdr *vid_hdr; + uint32_t crc; vid_hdr = ubi_zalloc_vid_hdr(ubi, GFP_NOFS); if (!vid_hdr) @@ -599,14 +600,8 @@ retry: goto out_put; } - vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); - err = ubi_io_write_vid_hdr(ubi, new_pnum, vid_hdr); - if (err) { - up_read(&ubi->fm_eba_sem); - goto write_error; - } + ubi_assert(vid_hdr->vol_type == UBI_VID_DYNAMIC); - data_size = offset + len; mutex_lock(&ubi->buf_mutex); memset(ubi->peb_buf + offset, 0xFF, len); @@ -621,6 +616,19 @@ retry: memcpy(ubi->peb_buf + offset, buf, len); + data_size = offset + len; + crc = crc32(UBI_CRC32_INIT, ubi->peb_buf, data_size); + vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); + vid_hdr->copy_flag = 1; + vid_hdr->data_size = cpu_to_be32(data_size); + vid_hdr->data_crc = cpu_to_be32(crc); + err = ubi_io_write_vid_hdr(ubi, new_pnum, vid_hdr); + if (err) { + mutex_unlock(&ubi->buf_mutex); + up_read(&ubi->fm_eba_sem); + goto write_error; + } + err = ubi_io_write_data(ubi, ubi->peb_buf, new_pnum, 0, data_size); if (err) { mutex_unlock(&ubi->buf_mutex); diff --git a/drivers/mtd/ubi/kapi.c b/drivers/mtd/ubi/kapi.c index 348dbbcbedc8..a9e2cef7c95c 100644 --- a/drivers/mtd/ubi/kapi.c +++ b/drivers/mtd/ubi/kapi.c @@ -302,6 +302,7 @@ EXPORT_SYMBOL_GPL(ubi_open_volume_nm); struct ubi_volume_desc *ubi_open_volume_path(const char *pathname, int mode) { int error, ubi_num, vol_id; + struct path path; struct kstat stat; dbg_gen("open volume %s, mode %d", pathname, mode); @@ -309,7 +310,12 @@ struct ubi_volume_desc *ubi_open_volume_path(const char *pathname, int mode) if (!pathname || !*pathname) return ERR_PTR(-EINVAL); - error = vfs_stat(pathname, &stat); + error = kern_path(pathname, LOOKUP_FOLLOW, &path); + if (error) + return ERR_PTR(error); + + error = vfs_getattr(&path, &stat); + path_put(&path); if (error) return ERR_PTR(error); diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index c5fe915870ad..a59d55e25d5f 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -12895,52 +12895,71 @@ static int __bnx2x_vlan_configure_vid(struct bnx2x *bp, u16 vid, bool add) return rc; } -int bnx2x_vlan_reconfigure_vid(struct bnx2x *bp) +static int bnx2x_vlan_configure_vid_list(struct bnx2x *bp) { struct bnx2x_vlan_entry *vlan; int rc = 0; - if (!bp->vlan_cnt) { - DP(NETIF_MSG_IFUP, "No need to re-configure vlan filters\n"); - return 0; - } - + /* Configure all non-configured entries */ list_for_each_entry(vlan, &bp->vlan_reg, link) { - /* Prepare for cleanup in case of errors */ - if (rc) { - vlan->hw = false; - continue; - } - - if (!vlan->hw) + if (vlan->hw) continue; - DP(NETIF_MSG_IFUP, "Re-configuring vlan 0x%04x\n", vlan->vid); + if (bp->vlan_cnt >= bp->vlan_credit) + return -ENOBUFS; rc = __bnx2x_vlan_configure_vid(bp, vlan->vid, true); if (rc) { - BNX2X_ERR("Unable to configure VLAN %d\n", vlan->vid); - vlan->hw = false; - rc = -EINVAL; - continue; + BNX2X_ERR("Unable to config VLAN %d\n", vlan->vid); + return rc; } + + DP(NETIF_MSG_IFUP, "HW configured for VLAN %d\n", vlan->vid); + vlan->hw = true; + bp->vlan_cnt++; } - return rc; + return 0; +} + +static void bnx2x_vlan_configure(struct bnx2x *bp, bool set_rx_mode) +{ + bool need_accept_any_vlan; + + need_accept_any_vlan = !!bnx2x_vlan_configure_vid_list(bp); + + if (bp->accept_any_vlan != need_accept_any_vlan) { + bp->accept_any_vlan = need_accept_any_vlan; + DP(NETIF_MSG_IFUP, "Accept all VLAN %s\n", + bp->accept_any_vlan ? "raised" : "cleared"); + if (set_rx_mode) { + if (IS_PF(bp)) + bnx2x_set_rx_mode_inner(bp); + else + bnx2x_vfpf_storm_rx_mode(bp); + } + } +} + +int bnx2x_vlan_reconfigure_vid(struct bnx2x *bp) +{ + struct bnx2x_vlan_entry *vlan; + + /* The hw forgot all entries after reload */ + list_for_each_entry(vlan, &bp->vlan_reg, link) + vlan->hw = false; + bp->vlan_cnt = 0; + + /* Don't set rx mode here. Our caller will do it. */ + bnx2x_vlan_configure(bp, false); + + return 0; } static int bnx2x_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid) { struct bnx2x *bp = netdev_priv(dev); struct bnx2x_vlan_entry *vlan; - bool hw = false; - int rc = 0; - - if (!netif_running(bp->dev)) { - DP(NETIF_MSG_IFUP, - "Ignoring VLAN configuration the interface is down\n"); - return -EFAULT; - } DP(NETIF_MSG_IFUP, "Adding VLAN %d\n", vid); @@ -12948,93 +12967,47 @@ static int bnx2x_vlan_rx_add_vid(struct net_device *dev, __be16 proto, u16 vid) if (!vlan) return -ENOMEM; - bp->vlan_cnt++; - if (bp->vlan_cnt > bp->vlan_credit && !bp->accept_any_vlan) { - DP(NETIF_MSG_IFUP, "Accept all VLAN raised\n"); - bp->accept_any_vlan = true; - if (IS_PF(bp)) - bnx2x_set_rx_mode_inner(bp); - else - bnx2x_vfpf_storm_rx_mode(bp); - } else if (bp->vlan_cnt <= bp->vlan_credit) { - rc = __bnx2x_vlan_configure_vid(bp, vid, true); - hw = true; - } - vlan->vid = vid; - vlan->hw = hw; + vlan->hw = false; + list_add_tail(&vlan->link, &bp->vlan_reg); - if (!rc) { - list_add(&vlan->link, &bp->vlan_reg); - } else { - bp->vlan_cnt--; - kfree(vlan); - } - - DP(NETIF_MSG_IFUP, "Adding VLAN result %d\n", rc); + if (netif_running(dev)) + bnx2x_vlan_configure(bp, true); - return rc; + return 0; } static int bnx2x_vlan_rx_kill_vid(struct net_device *dev, __be16 proto, u16 vid) { struct bnx2x *bp = netdev_priv(dev); struct bnx2x_vlan_entry *vlan; + bool found = false; int rc = 0; - if (!netif_running(bp->dev)) { - DP(NETIF_MSG_IFUP, - "Ignoring VLAN configuration the interface is down\n"); - return -EFAULT; - } - DP(NETIF_MSG_IFUP, "Removing VLAN %d\n", vid); - if (!bp->vlan_cnt) { - BNX2X_ERR("Unable to kill VLAN %d\n", vid); - return -EINVAL; - } - list_for_each_entry(vlan, &bp->vlan_reg, link) - if (vlan->vid == vid) + if (vlan->vid == vid) { + found = true; break; + } - if (vlan->vid != vid) { + if (!found) { BNX2X_ERR("Unable to kill VLAN %d - not found\n", vid); return -EINVAL; } - if (vlan->hw) + if (netif_running(dev) && vlan->hw) { rc = __bnx2x_vlan_configure_vid(bp, vid, false); + DP(NETIF_MSG_IFUP, "HW deconfigured for VLAN %d\n", vid); + bp->vlan_cnt--; + } list_del(&vlan->link); kfree(vlan); - bp->vlan_cnt--; - - if (bp->vlan_cnt <= bp->vlan_credit && bp->accept_any_vlan) { - /* Configure all non-configured entries */ - list_for_each_entry(vlan, &bp->vlan_reg, link) { - if (vlan->hw) - continue; - - rc = __bnx2x_vlan_configure_vid(bp, vlan->vid, true); - if (rc) { - BNX2X_ERR("Unable to config VLAN %d\n", - vlan->vid); - continue; - } - DP(NETIF_MSG_IFUP, "HW configured for VLAN %d\n", - vlan->vid); - vlan->hw = true; - } - DP(NETIF_MSG_IFUP, "Accept all VLAN Removed\n"); - bp->accept_any_vlan = false; - if (IS_PF(bp)) - bnx2x_set_rx_mode_inner(bp); - else - bnx2x_vfpf_storm_rx_mode(bp); - } + if (netif_running(dev)) + bnx2x_vlan_configure(bp, true); DP(NETIF_MSG_IFUP, "Removing VLAN result %d\n", rc); diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c index 72a2efff8e49..c777cde85ce4 100644 --- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c +++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c @@ -286,7 +286,9 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) cpu_to_le32(DB_KEY_TX_PUSH | DB_LONG_TX_PUSH | prod); txr->tx_prod = prod; + tx_buf->is_push = 1; netdev_tx_sent_queue(txq, skb->len); + wmb(); /* Sync is_push and byte queue before pushing data */ push_len = (length + sizeof(*tx_push) + 7) / 8; if (push_len > 16) { @@ -298,7 +300,6 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) push_len); } - tx_buf->is_push = 1; goto tx_done; } @@ -1112,19 +1113,13 @@ static inline struct sk_buff *bnxt_tpa_end(struct bnxt *bp, if (tpa_info->hash_type != PKT_HASH_TYPE_NONE) skb_set_hash(skb, tpa_info->rss_hash, tpa_info->hash_type); - if (tpa_info->flags2 & RX_CMP_FLAGS2_META_FORMAT_VLAN) { - netdev_features_t features = skb->dev->features; + if ((tpa_info->flags2 & RX_CMP_FLAGS2_META_FORMAT_VLAN) && + (skb->dev->features & NETIF_F_HW_VLAN_CTAG_RX)) { u16 vlan_proto = tpa_info->metadata >> RX_CMP_FLAGS2_METADATA_TPID_SFT; + u16 vtag = tpa_info->metadata & RX_CMP_FLAGS2_METADATA_VID_MASK; - if (((features & NETIF_F_HW_VLAN_CTAG_RX) && - vlan_proto == ETH_P_8021Q) || - ((features & NETIF_F_HW_VLAN_STAG_RX) && - vlan_proto == ETH_P_8021AD)) { - __vlan_hwaccel_put_tag(skb, htons(vlan_proto), - tpa_info->metadata & - RX_CMP_FLAGS2_METADATA_VID_MASK); - } + __vlan_hwaccel_put_tag(skb, htons(vlan_proto), vtag); } skb_checksum_none_assert(skb); @@ -1277,19 +1272,14 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_napi *bnapi, u32 *raw_cons, skb->protocol = eth_type_trans(skb, dev); - if (rxcmp1->rx_cmp_flags2 & - cpu_to_le32(RX_CMP_FLAGS2_META_FORMAT_VLAN)) { - netdev_features_t features = skb->dev->features; + if ((rxcmp1->rx_cmp_flags2 & + cpu_to_le32(RX_CMP_FLAGS2_META_FORMAT_VLAN)) && + (skb->dev->features & NETIF_F_HW_VLAN_CTAG_RX)) { u32 meta_data = le32_to_cpu(rxcmp1->rx_cmp_meta_data); + u16 vtag = meta_data & RX_CMP_FLAGS2_METADATA_VID_MASK; u16 vlan_proto = meta_data >> RX_CMP_FLAGS2_METADATA_TPID_SFT; - if (((features & NETIF_F_HW_VLAN_CTAG_RX) && - vlan_proto == ETH_P_8021Q) || - ((features & NETIF_F_HW_VLAN_STAG_RX) && - vlan_proto == ETH_P_8021AD)) - __vlan_hwaccel_put_tag(skb, htons(vlan_proto), - meta_data & - RX_CMP_FLAGS2_METADATA_VID_MASK); + __vlan_hwaccel_put_tag(skb, htons(vlan_proto), vtag); } skb_checksum_none_assert(skb); @@ -5466,6 +5456,20 @@ static netdev_features_t bnxt_fix_features(struct net_device *dev, if (!bnxt_rfs_capable(bp)) features &= ~NETIF_F_NTUPLE; + + /* Both CTAG and STAG VLAN accelaration on the RX side have to be + * turned on or off together. + */ + if ((features & (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_STAG_RX)) != + (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_STAG_RX)) { + if (dev->features & NETIF_F_HW_VLAN_CTAG_RX) + features &= ~(NETIF_F_HW_VLAN_CTAG_RX | + NETIF_F_HW_VLAN_STAG_RX); + else + features |= NETIF_F_HW_VLAN_CTAG_RX | + NETIF_F_HW_VLAN_STAG_RX; + } + return features; } diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h b/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h index a2cdfc1261dc..50812a1d67bd 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h +++ b/drivers/net/ethernet/chelsio/cxgb4/t4_pci_id_tbl.h @@ -144,6 +144,7 @@ CH_PCI_DEVICE_ID_TABLE_DEFINE_BEGIN CH_PCI_ID_TABLE_FENTRY(0x5015), /* T502-bt */ CH_PCI_ID_TABLE_FENTRY(0x5016), /* T580-OCP-SO */ CH_PCI_ID_TABLE_FENTRY(0x5017), /* T520-OCP-SO */ + CH_PCI_ID_TABLE_FENTRY(0x5018), /* T540-BT */ CH_PCI_ID_TABLE_FENTRY(0x5080), /* Custom T540-cr */ CH_PCI_ID_TABLE_FENTRY(0x5081), /* Custom T540-LL-cr */ CH_PCI_ID_TABLE_FENTRY(0x5082), /* Custom T504-cr */ diff --git a/drivers/net/ethernet/ethoc.c b/drivers/net/ethernet/ethoc.c index 41b010645100..4edb98c3c6c7 100644 --- a/drivers/net/ethernet/ethoc.c +++ b/drivers/net/ethernet/ethoc.c @@ -1195,7 +1195,7 @@ static int ethoc_probe(struct platform_device *pdev) priv->mdio = mdiobus_alloc(); if (!priv->mdio) { ret = -ENOMEM; - goto free; + goto free2; } priv->mdio->name = "ethoc-mdio"; @@ -1208,7 +1208,7 @@ static int ethoc_probe(struct platform_device *pdev) ret = mdiobus_register(priv->mdio); if (ret) { dev_err(&netdev->dev, "failed to register MDIO bus\n"); - goto free; + goto free2; } ret = ethoc_mdio_probe(netdev); @@ -1241,9 +1241,10 @@ error2: error: mdiobus_unregister(priv->mdio); mdiobus_free(priv->mdio); -free: +free2: if (priv->clk) clk_disable_unprepare(priv->clk); +free: free_netdev(netdev); out: return ret; diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index 3c0255e98535..fea0f330ddbd 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -2416,24 +2416,24 @@ fec_enet_set_coalesce(struct net_device *ndev, struct ethtool_coalesce *ec) return -EOPNOTSUPP; if (ec->rx_max_coalesced_frames > 255) { - pr_err("Rx coalesced frames exceed hardware limiation"); + pr_err("Rx coalesced frames exceed hardware limitation\n"); return -EINVAL; } if (ec->tx_max_coalesced_frames > 255) { - pr_err("Tx coalesced frame exceed hardware limiation"); + pr_err("Tx coalesced frame exceed hardware limitation\n"); return -EINVAL; } cycle = fec_enet_us_to_itr_clock(ndev, fep->rx_time_itr); if (cycle > 0xFFFF) { - pr_err("Rx coalesed usec exceeed hardware limiation"); + pr_err("Rx coalesced usec exceed hardware limitation\n"); return -EINVAL; } cycle = fec_enet_us_to_itr_clock(ndev, fep->tx_time_itr); if (cycle > 0xFFFF) { - pr_err("Rx coalesed usec exceeed hardware limiation"); + pr_err("Rx coalesced usec exceed hardware limitation\n"); return -EINVAL; } diff --git a/drivers/net/ethernet/freescale/gianfar.c b/drivers/net/ethernet/freescale/gianfar.c index 7615e0668acb..2e6785b6e8be 100644 --- a/drivers/net/ethernet/freescale/gianfar.c +++ b/drivers/net/ethernet/freescale/gianfar.c @@ -2440,7 +2440,8 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev) tx_queue->tx_ring_size); if (likely(!nr_frags)) { - lstatus |= BD_LFLAG(TXBD_LAST | TXBD_INTERRUPT); + if (likely(!do_tstamp)) + lstatus |= BD_LFLAG(TXBD_LAST | TXBD_INTERRUPT); } else { u32 lstatus_start = lstatus; diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c index c984462fad2a..4763252bbf85 100644 --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c @@ -133,6 +133,8 @@ static int mtk_mdio_read(struct mii_bus *bus, int phy_addr, int phy_reg) static void mtk_phy_link_adjust(struct net_device *dev) { struct mtk_mac *mac = netdev_priv(dev); + u16 lcl_adv = 0, rmt_adv = 0; + u8 flowctrl; u32 mcr = MAC_MCR_MAX_RX_1536 | MAC_MCR_IPG_CFG | MAC_MCR_FORCE_MODE | MAC_MCR_TX_EN | MAC_MCR_RX_EN | MAC_MCR_BACKOFF_EN | @@ -150,11 +152,30 @@ static void mtk_phy_link_adjust(struct net_device *dev) if (mac->phy_dev->link) mcr |= MAC_MCR_FORCE_LINK; - if (mac->phy_dev->duplex) + if (mac->phy_dev->duplex) { mcr |= MAC_MCR_FORCE_DPX; - if (mac->phy_dev->pause) - mcr |= MAC_MCR_FORCE_RX_FC | MAC_MCR_FORCE_TX_FC; + if (mac->phy_dev->pause) + rmt_adv = LPA_PAUSE_CAP; + if (mac->phy_dev->asym_pause) + rmt_adv |= LPA_PAUSE_ASYM; + + if (mac->phy_dev->advertising & ADVERTISED_Pause) + lcl_adv |= ADVERTISE_PAUSE_CAP; + if (mac->phy_dev->advertising & ADVERTISED_Asym_Pause) + lcl_adv |= ADVERTISE_PAUSE_ASYM; + + flowctrl = mii_resolve_flowctrl_fdx(lcl_adv, rmt_adv); + + if (flowctrl & FLOW_CTRL_TX) + mcr |= MAC_MCR_FORCE_TX_FC; + if (flowctrl & FLOW_CTRL_RX) + mcr |= MAC_MCR_FORCE_RX_FC; + + netif_dbg(mac->hw, link, dev, "rx pause %s, tx pause %s\n", + flowctrl & FLOW_CTRL_RX ? "enabled" : "disabled", + flowctrl & FLOW_CTRL_TX ? "enabled" : "disabled"); + } mtk_w32(mac->hw, mcr, MTK_MAC_MCR(mac->id)); @@ -208,10 +229,16 @@ static int mtk_phy_connect(struct mtk_mac *mac) u32 val, ge_mode; np = of_parse_phandle(mac->of_node, "phy-handle", 0); + if (!np && of_phy_is_fixed_link(mac->of_node)) + if (!of_phy_register_fixed_link(mac->of_node)) + np = of_node_get(mac->of_node); if (!np) return -ENODEV; switch (of_get_phy_mode(np)) { + case PHY_INTERFACE_MODE_RGMII_TXID: + case PHY_INTERFACE_MODE_RGMII_RXID: + case PHY_INTERFACE_MODE_RGMII_ID: case PHY_INTERFACE_MODE_RGMII: ge_mode = 0; break; @@ -236,7 +263,8 @@ static int mtk_phy_connect(struct mtk_mac *mac) mac->phy_dev->autoneg = AUTONEG_ENABLE; mac->phy_dev->speed = 0; mac->phy_dev->duplex = 0; - mac->phy_dev->supported &= PHY_BASIC_FEATURES; + mac->phy_dev->supported &= PHY_GBIT_FEATURES | SUPPORTED_Pause | + SUPPORTED_Asym_Pause; mac->phy_dev->advertising = mac->phy_dev->supported | ADVERTISED_Autoneg; phy_start_aneg(mac->phy_dev); @@ -280,7 +308,7 @@ static int mtk_mdio_init(struct mtk_eth *eth) return 0; err_free_bus: - kfree(eth->mii_bus); + mdiobus_free(eth->mii_bus); err_put_node: of_node_put(mii_np); @@ -295,7 +323,7 @@ static void mtk_mdio_cleanup(struct mtk_eth *eth) mdiobus_unregister(eth->mii_bus); of_node_put(eth->mii_bus->dev.of_node); - kfree(eth->mii_bus); + mdiobus_free(eth->mii_bus); } static inline void mtk_irq_disable(struct mtk_eth *eth, u32 mask) diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c index fd4392999eee..f5c8d5db25a8 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_main.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_main.c @@ -3192,10 +3192,7 @@ static void mlx5e_destroy_netdev(struct mlx5_core_dev *mdev, void *vpriv) flush_workqueue(priv->wq); if (test_bit(MLX5_INTERFACE_STATE_SHUTDOWN, &mdev->intf_state)) { netif_device_detach(netdev); - mutex_lock(&priv->state_lock); - if (test_bit(MLX5E_STATE_OPENED, &priv->state)) - mlx5e_close_locked(netdev); - mutex_unlock(&priv->state_lock); + mlx5e_close(netdev); } else { unregister_netdev(netdev); } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c index 229ab16fb8d3..b000ddc29553 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_tx.c @@ -317,7 +317,8 @@ static netdev_tx_t mlx5e_sq_xmit(struct mlx5e_sq *sq, struct sk_buff *skb) while ((sq->pc & wq->sz_m1) > sq->edge) mlx5e_send_nop(sq, false); - sq->bf_budget = bf ? sq->bf_budget - 1 : 0; + if (bf) + sq->bf_budget--; sq->stats.packets++; sq->stats.bytes += num_bytes; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c index b84a6918a700..aebbd6ccb9fe 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/eswitch.c @@ -383,7 +383,7 @@ __esw_fdb_set_vport_rule(struct mlx5_eswitch *esw, u32 vport, bool rx_rule, match_v, MLX5_FLOW_CONTEXT_ACTION_FWD_DEST, 0, &dest); - if (IS_ERR_OR_NULL(flow_rule)) { + if (IS_ERR(flow_rule)) { pr_warn( "FDB: Failed to add flow rule: dmac_v(%pM) dmac_c(%pM) -> vport(%d), err(%ld)\n", dmac_v, dmac_c, vport, PTR_ERR(flow_rule)); @@ -457,7 +457,7 @@ static int esw_create_fdb_table(struct mlx5_eswitch *esw, int nvports) table_size = BIT(MLX5_CAP_ESW_FLOWTABLE_FDB(dev, log_max_ft_size)); fdb = mlx5_create_flow_table(root_ns, 0, table_size, 0); - if (IS_ERR_OR_NULL(fdb)) { + if (IS_ERR(fdb)) { err = PTR_ERR(fdb); esw_warn(dev, "Failed to create FDB Table err %d\n", err); goto out; @@ -474,7 +474,7 @@ static int esw_create_fdb_table(struct mlx5_eswitch *esw, int nvports) MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, table_size - 3); eth_broadcast_addr(dmac); g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create flow group err(%d)\n", err); goto out; @@ -489,7 +489,7 @@ static int esw_create_fdb_table(struct mlx5_eswitch *esw, int nvports) eth_zero_addr(dmac); dmac[0] = 0x01; g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create allmulti flow group err(%d)\n", err); goto out; @@ -506,7 +506,7 @@ static int esw_create_fdb_table(struct mlx5_eswitch *esw, int nvports) MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, table_size - 1); MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, table_size - 1); g = mlx5_create_flow_group(fdb, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create promisc flow group err(%d)\n", err); goto out; @@ -529,7 +529,7 @@ out: } } - kfree(flow_group_in); + kvfree(flow_group_in); return err; } @@ -651,6 +651,7 @@ static void update_allmulti_vports(struct mlx5_eswitch *esw, esw_fdb_set_vport_rule(esw, mac, vport_idx); + iter_vaddr->mc_promisc = true; break; case MLX5_ACTION_DEL: if (!iter_vaddr) @@ -1060,7 +1061,7 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, return; acl = mlx5_create_vport_flow_table(root_ns, 0, table_size, 0, vport->vport); - if (IS_ERR_OR_NULL(acl)) { + if (IS_ERR(acl)) { err = PTR_ERR(acl); esw_warn(dev, "Failed to create E-Switch vport[%d] egress flow Table, err(%d)\n", vport->vport, err); @@ -1075,7 +1076,7 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 0); vlan_grp = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(vlan_grp)) { + if (IS_ERR(vlan_grp)) { err = PTR_ERR(vlan_grp); esw_warn(dev, "Failed to create E-Switch vport[%d] egress allowed vlans flow group, err(%d)\n", vport->vport, err); @@ -1086,7 +1087,7 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, start_flow_index, 1); MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 1); drop_grp = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(drop_grp)) { + if (IS_ERR(drop_grp)) { err = PTR_ERR(drop_grp); esw_warn(dev, "Failed to create E-Switch vport[%d] egress drop flow group, err(%d)\n", vport->vport, err); @@ -1097,7 +1098,7 @@ static void esw_vport_enable_egress_acl(struct mlx5_eswitch *esw, vport->egress.drop_grp = drop_grp; vport->egress.allowed_vlans_grp = vlan_grp; out: - kfree(flow_group_in); + kvfree(flow_group_in); if (err && !IS_ERR_OR_NULL(vlan_grp)) mlx5_destroy_flow_group(vlan_grp); if (err && !IS_ERR_OR_NULL(acl)) @@ -1174,7 +1175,7 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, return; acl = mlx5_create_vport_flow_table(root_ns, 0, table_size, 0, vport->vport); - if (IS_ERR_OR_NULL(acl)) { + if (IS_ERR(acl)) { err = PTR_ERR(acl); esw_warn(dev, "Failed to create E-Switch vport[%d] ingress flow Table, err(%d)\n", vport->vport, err); @@ -1192,7 +1193,7 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 0); g = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create E-Switch vport[%d] ingress untagged spoofchk flow group, err(%d)\n", vport->vport, err); @@ -1207,7 +1208,7 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 1); g = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create E-Switch vport[%d] ingress untagged flow group, err(%d)\n", vport->vport, err); @@ -1223,7 +1224,7 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 2); g = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create E-Switch vport[%d] ingress spoofchk flow group, err(%d)\n", vport->vport, err); @@ -1236,7 +1237,7 @@ static void esw_vport_enable_ingress_acl(struct mlx5_eswitch *esw, MLX5_SET(create_flow_group_in, flow_group_in, end_flow_index, 3); g = mlx5_create_flow_group(acl, flow_group_in); - if (IS_ERR_OR_NULL(g)) { + if (IS_ERR(g)) { err = PTR_ERR(g); esw_warn(dev, "Failed to create E-Switch vport[%d] ingress drop flow group, err(%d)\n", vport->vport, err); @@ -1259,7 +1260,7 @@ out: mlx5_destroy_flow_table(vport->ingress.acl); } - kfree(flow_group_in); + kvfree(flow_group_in); } static void esw_vport_cleanup_ingress_rules(struct mlx5_eswitch *esw, @@ -1363,7 +1364,7 @@ static int esw_vport_ingress_config(struct mlx5_eswitch *esw, match_v, MLX5_FLOW_CONTEXT_ACTION_ALLOW, 0, NULL); - if (IS_ERR_OR_NULL(vport->ingress.allow_rule)) { + if (IS_ERR(vport->ingress.allow_rule)) { err = PTR_ERR(vport->ingress.allow_rule); pr_warn("vport[%d] configure ingress allow rule, err(%d)\n", vport->vport, err); @@ -1380,7 +1381,7 @@ static int esw_vport_ingress_config(struct mlx5_eswitch *esw, match_v, MLX5_FLOW_CONTEXT_ACTION_DROP, 0, NULL); - if (IS_ERR_OR_NULL(vport->ingress.drop_rule)) { + if (IS_ERR(vport->ingress.drop_rule)) { err = PTR_ERR(vport->ingress.drop_rule); pr_warn("vport[%d] configure ingress drop rule, err(%d)\n", vport->vport, err); @@ -1439,7 +1440,7 @@ static int esw_vport_egress_config(struct mlx5_eswitch *esw, match_v, MLX5_FLOW_CONTEXT_ACTION_ALLOW, 0, NULL); - if (IS_ERR_OR_NULL(vport->egress.allowed_vlan)) { + if (IS_ERR(vport->egress.allowed_vlan)) { err = PTR_ERR(vport->egress.allowed_vlan); pr_warn("vport[%d] configure egress allowed vlan rule failed, err(%d)\n", vport->vport, err); @@ -1457,7 +1458,7 @@ static int esw_vport_egress_config(struct mlx5_eswitch *esw, match_v, MLX5_FLOW_CONTEXT_ACTION_DROP, 0, NULL); - if (IS_ERR_OR_NULL(vport->egress.drop_rule)) { + if (IS_ERR(vport->egress.drop_rule)) { err = PTR_ERR(vport->egress.drop_rule); pr_warn("vport[%d] configure egress drop rule failed, err(%d)\n", vport->vport, err); @@ -1491,14 +1492,11 @@ static void esw_enable_vport(struct mlx5_eswitch *esw, int vport_num, /* Sync with current vport context */ vport->enabled_events = enable_events; - esw_vport_change_handle_locked(vport); - vport->enabled = true; /* only PF is trusted by default */ vport->trusted = (vport_num) ? false : true; - - arm_vport_context_events_cmd(esw->dev, vport_num, enable_events); + esw_vport_change_handle_locked(vport); esw->enabled_vports++; esw_debug(esw->dev, "Enabled VPORT(%d)\n", vport_num); @@ -1728,11 +1726,24 @@ void mlx5_eswitch_vport_event(struct mlx5_eswitch *esw, struct mlx5_eqe *eqe) (esw && MLX5_CAP_GEN(esw->dev, vport_group_manager) && mlx5_core_is_pf(esw->dev)) #define LEGAL_VPORT(esw, vport) (vport >= 0 && vport < esw->total_vports) +static void node_guid_gen_from_mac(u64 *node_guid, u8 mac[ETH_ALEN]) +{ + ((u8 *)node_guid)[7] = mac[0]; + ((u8 *)node_guid)[6] = mac[1]; + ((u8 *)node_guid)[5] = mac[2]; + ((u8 *)node_guid)[4] = 0xff; + ((u8 *)node_guid)[3] = 0xfe; + ((u8 *)node_guid)[2] = mac[3]; + ((u8 *)node_guid)[1] = mac[4]; + ((u8 *)node_guid)[0] = mac[5]; +} + int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw, int vport, u8 mac[ETH_ALEN]) { - int err = 0; struct mlx5_vport *evport; + u64 node_guid; + int err = 0; if (!ESW_ALLOWED(esw)) return -EPERM; @@ -1756,11 +1767,17 @@ int mlx5_eswitch_set_vport_mac(struct mlx5_eswitch *esw, return err; } + node_guid_gen_from_mac(&node_guid, mac); + err = mlx5_modify_nic_vport_node_guid(esw->dev, vport, node_guid); + if (err) + mlx5_core_warn(esw->dev, + "Failed to set vport %d node guid, err = %d. RDMA_CM will not function properly for this VF.\n", + vport, err); + mutex_lock(&esw->state_lock); if (evport->enabled) err = esw_vport_ingress_config(esw, evport); mutex_unlock(&esw->state_lock); - return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c index 8b5f0b2c0d5c..e912a3d2505e 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/fs_core.c @@ -1292,8 +1292,8 @@ static int update_root_ft_destroy(struct mlx5_flow_table *ft) ft->id); return err; } - root->root_ft = new_root_ft; } + root->root_ft = new_root_ft; return 0; } @@ -1767,6 +1767,9 @@ static void cleanup_root_ns(struct mlx5_core_dev *dev) void mlx5_cleanup_fs(struct mlx5_core_dev *dev) { + if (MLX5_CAP_GEN(dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) + return; + cleanup_root_ns(dev); cleanup_single_prio_root_ns(dev, dev->priv.fdb_root_ns); cleanup_single_prio_root_ns(dev, dev->priv.esw_egress_root_ns); @@ -1828,29 +1831,36 @@ int mlx5_init_fs(struct mlx5_core_dev *dev) { int err = 0; + if (MLX5_CAP_GEN(dev, port_type) != MLX5_CAP_PORT_TYPE_ETH) + return 0; + err = mlx5_init_fc_stats(dev); if (err) return err; - if (MLX5_CAP_GEN(dev, nic_flow_table)) { + if (MLX5_CAP_GEN(dev, nic_flow_table) && + MLX5_CAP_FLOWTABLE_NIC_RX(dev, ft_support)) { err = init_root_ns(dev); if (err) goto err; } + if (MLX5_CAP_GEN(dev, eswitch_flow_table)) { - err = init_fdb_root_ns(dev); - if (err) - goto err; - } - if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support)) { - err = init_egress_acl_root_ns(dev); - if (err) - goto err; - } - if (MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support)) { - err = init_ingress_acl_root_ns(dev); - if (err) - goto err; + if (MLX5_CAP_ESW_FLOWTABLE_FDB(dev, ft_support)) { + err = init_fdb_root_ns(dev); + if (err) + goto err; + } + if (MLX5_CAP_ESW_EGRESS_ACL(dev, ft_support)) { + err = init_egress_acl_root_ns(dev); + if (err) + goto err; + } + if (MLX5_CAP_ESW_INGRESS_ACL(dev, ft_support)) { + err = init_ingress_acl_root_ns(dev); + if (err) + goto err; + } } return 0; diff --git a/drivers/net/ethernet/mellanox/mlx5/core/qp.c b/drivers/net/ethernet/mellanox/mlx5/core/qp.c index b720a274220d..b82d65802d96 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c @@ -418,7 +418,7 @@ int mlx5_core_xrcd_alloc(struct mlx5_core_dev *dev, u32 *xrcdn) if (out.hdr.status) err = mlx5_cmd_status_to_err(&out.hdr); else - *xrcdn = be32_to_cpu(out.xrcdn); + *xrcdn = be32_to_cpu(out.xrcdn) & 0xffffff; return err; } diff --git a/drivers/net/ethernet/mellanox/mlx5/core/vport.c b/drivers/net/ethernet/mellanox/mlx5/core/vport.c index b69dadcfb897..daf44cd4c566 100644 --- a/drivers/net/ethernet/mellanox/mlx5/core/vport.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/vport.c @@ -508,6 +508,44 @@ int mlx5_query_nic_vport_node_guid(struct mlx5_core_dev *mdev, u64 *node_guid) } EXPORT_SYMBOL_GPL(mlx5_query_nic_vport_node_guid); +int mlx5_modify_nic_vport_node_guid(struct mlx5_core_dev *mdev, + u32 vport, u64 node_guid) +{ + int inlen = MLX5_ST_SZ_BYTES(modify_nic_vport_context_in); + void *nic_vport_context; + u8 *guid; + void *in; + int err; + + if (!vport) + return -EINVAL; + if (!MLX5_CAP_GEN(mdev, vport_group_manager)) + return -EACCES; + if (!MLX5_CAP_ESW(mdev, nic_vport_node_guid_modify)) + return -ENOTSUPP; + + in = mlx5_vzalloc(inlen); + if (!in) + return -ENOMEM; + + MLX5_SET(modify_nic_vport_context_in, in, + field_select.node_guid, 1); + MLX5_SET(modify_nic_vport_context_in, in, vport_number, vport); + MLX5_SET(modify_nic_vport_context_in, in, other_vport, !!vport); + + nic_vport_context = MLX5_ADDR_OF(modify_nic_vport_context_in, + in, nic_vport_context); + guid = MLX5_ADDR_OF(nic_vport_context, nic_vport_context, + node_guid); + MLX5_SET64(nic_vport_context, nic_vport_context, node_guid, node_guid); + + err = mlx5_modify_nic_vport_context(mdev, in, inlen); + + kvfree(in); + + return err; +} + int mlx5_query_nic_vport_qkey_viol_cntr(struct mlx5_core_dev *mdev, u16 *qkey_viol_cntr) { diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c index 4a7273771028..6f9e3ddff4a8 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.c @@ -247,15 +247,23 @@ static int mlxsw_sp_port_mtu_set(struct mlxsw_sp_port *mlxsw_sp_port, u16 mtu) return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pmtu), pmtu_pl); } -static int mlxsw_sp_port_swid_set(struct mlxsw_sp_port *mlxsw_sp_port, u8 swid) +static int __mlxsw_sp_port_swid_set(struct mlxsw_sp *mlxsw_sp, u8 local_port, + u8 swid) { - struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; char pspa_pl[MLXSW_REG_PSPA_LEN]; - mlxsw_reg_pspa_pack(pspa_pl, swid, mlxsw_sp_port->local_port); + mlxsw_reg_pspa_pack(pspa_pl, swid, local_port); return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(pspa), pspa_pl); } +static int mlxsw_sp_port_swid_set(struct mlxsw_sp_port *mlxsw_sp_port, u8 swid) +{ + struct mlxsw_sp *mlxsw_sp = mlxsw_sp_port->mlxsw_sp; + + return __mlxsw_sp_port_swid_set(mlxsw_sp, mlxsw_sp_port->local_port, + swid); +} + static int mlxsw_sp_port_vp_mode_set(struct mlxsw_sp_port *mlxsw_sp_port, bool enable) { @@ -305,9 +313,9 @@ mlxsw_sp_port_system_port_mapping_set(struct mlxsw_sp_port *mlxsw_sp_port) return mlxsw_reg_write(mlxsw_sp->core, MLXSW_REG(sspr), sspr_pl); } -static int __mlxsw_sp_port_module_info_get(struct mlxsw_sp *mlxsw_sp, - u8 local_port, u8 *p_module, - u8 *p_width, u8 *p_lane) +static int mlxsw_sp_port_module_info_get(struct mlxsw_sp *mlxsw_sp, + u8 local_port, u8 *p_module, + u8 *p_width, u8 *p_lane) { char pmlp_pl[MLXSW_REG_PMLP_LEN]; int err; @@ -322,16 +330,6 @@ static int __mlxsw_sp_port_module_info_get(struct mlxsw_sp *mlxsw_sp, return 0; } -static int mlxsw_sp_port_module_info_get(struct mlxsw_sp *mlxsw_sp, - u8 local_port, u8 *p_module, - u8 *p_width) -{ - u8 lane; - - return __mlxsw_sp_port_module_info_get(mlxsw_sp, local_port, p_module, - p_width, &lane); -} - static int mlxsw_sp_port_module_map(struct mlxsw_sp *mlxsw_sp, u8 local_port, u8 module, u8 width, u8 lane) { @@ -949,17 +947,11 @@ static int mlxsw_sp_port_get_phys_port_name(struct net_device *dev, char *name, size_t len) { struct mlxsw_sp_port *mlxsw_sp_port = netdev_priv(dev); - u8 module, width, lane; + u8 module = mlxsw_sp_port->mapping.module; + u8 width = mlxsw_sp_port->mapping.width; + u8 lane = mlxsw_sp_port->mapping.lane; int err; - err = __mlxsw_sp_port_module_info_get(mlxsw_sp_port->mlxsw_sp, - mlxsw_sp_port->local_port, - &module, &width, &lane); - if (err) { - netdev_err(dev, "Failed to retrieve module information\n"); - return err; - } - if (!mlxsw_sp_port->split) err = snprintf(name, len, "p%d", module + 1); else @@ -1681,8 +1673,8 @@ static int mlxsw_sp_port_ets_init(struct mlxsw_sp_port *mlxsw_sp_port) return 0; } -static int __mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, - bool split, u8 module, u8 width) +static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, + bool split, u8 module, u8 width, u8 lane) { struct mlxsw_sp_port *mlxsw_sp_port; struct net_device *dev; @@ -1697,6 +1689,9 @@ static int __mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, mlxsw_sp_port->mlxsw_sp = mlxsw_sp; mlxsw_sp_port->local_port = local_port; mlxsw_sp_port->split = split; + mlxsw_sp_port->mapping.module = module; + mlxsw_sp_port->mapping.width = width; + mlxsw_sp_port->mapping.lane = lane; bytes = DIV_ROUND_UP(VLAN_N_VID, BITS_PER_BYTE); mlxsw_sp_port->active_vlans = kzalloc(bytes, GFP_KERNEL); if (!mlxsw_sp_port->active_vlans) { @@ -1839,28 +1834,6 @@ err_port_active_vlans_alloc: return err; } -static int mlxsw_sp_port_create(struct mlxsw_sp *mlxsw_sp, u8 local_port, - bool split, u8 module, u8 width, u8 lane) -{ - int err; - - err = mlxsw_sp_port_module_map(mlxsw_sp, local_port, module, width, - lane); - if (err) - return err; - - err = __mlxsw_sp_port_create(mlxsw_sp, local_port, split, module, - width); - if (err) - goto err_port_create; - - return 0; - -err_port_create: - mlxsw_sp_port_module_unmap(mlxsw_sp, local_port); - return err; -} - static void mlxsw_sp_port_vports_fini(struct mlxsw_sp_port *mlxsw_sp_port) { struct net_device *dev = mlxsw_sp_port->dev; @@ -1909,8 +1882,8 @@ static void mlxsw_sp_ports_remove(struct mlxsw_sp *mlxsw_sp) static int mlxsw_sp_ports_create(struct mlxsw_sp *mlxsw_sp) { + u8 module, width, lane; size_t alloc_size; - u8 module, width; int i; int err; @@ -1921,13 +1894,14 @@ static int mlxsw_sp_ports_create(struct mlxsw_sp *mlxsw_sp) for (i = 1; i < MLXSW_PORT_MAX_PORTS; i++) { err = mlxsw_sp_port_module_info_get(mlxsw_sp, i, &module, - &width); + &width, &lane); if (err) goto err_port_module_info_get; if (!width) continue; mlxsw_sp->port_to_module[i] = module; - err = __mlxsw_sp_port_create(mlxsw_sp, i, false, module, width); + err = mlxsw_sp_port_create(mlxsw_sp, i, false, module, width, + lane); if (err) goto err_port_create; } @@ -1948,12 +1922,85 @@ static u8 mlxsw_sp_cluster_base_port_get(u8 local_port) return local_port - offset; } +static int mlxsw_sp_port_split_create(struct mlxsw_sp *mlxsw_sp, u8 base_port, + u8 module, unsigned int count) +{ + u8 width = MLXSW_PORT_MODULE_MAX_WIDTH / count; + int err, i; + + for (i = 0; i < count; i++) { + err = mlxsw_sp_port_module_map(mlxsw_sp, base_port + i, module, + width, i * width); + if (err) + goto err_port_module_map; + } + + for (i = 0; i < count; i++) { + err = __mlxsw_sp_port_swid_set(mlxsw_sp, base_port + i, 0); + if (err) + goto err_port_swid_set; + } + + for (i = 0; i < count; i++) { + err = mlxsw_sp_port_create(mlxsw_sp, base_port + i, true, + module, width, i * width); + if (err) + goto err_port_create; + } + + return 0; + +err_port_create: + for (i--; i >= 0; i--) + mlxsw_sp_port_remove(mlxsw_sp, base_port + i); + i = count; +err_port_swid_set: + for (i--; i >= 0; i--) + __mlxsw_sp_port_swid_set(mlxsw_sp, base_port + i, + MLXSW_PORT_SWID_DISABLED_PORT); + i = count; +err_port_module_map: + for (i--; i >= 0; i--) + mlxsw_sp_port_module_unmap(mlxsw_sp, base_port + i); + return err; +} + +static void mlxsw_sp_port_unsplit_create(struct mlxsw_sp *mlxsw_sp, + u8 base_port, unsigned int count) +{ + u8 local_port, module, width = MLXSW_PORT_MODULE_MAX_WIDTH; + int i; + + /* Split by four means we need to re-create two ports, otherwise + * only one. + */ + count = count / 2; + + for (i = 0; i < count; i++) { + local_port = base_port + i * 2; + module = mlxsw_sp->port_to_module[local_port]; + + mlxsw_sp_port_module_map(mlxsw_sp, local_port, module, width, + 0); + } + + for (i = 0; i < count; i++) + __mlxsw_sp_port_swid_set(mlxsw_sp, base_port + i * 2, 0); + + for (i = 0; i < count; i++) { + local_port = base_port + i * 2; + module = mlxsw_sp->port_to_module[local_port]; + + mlxsw_sp_port_create(mlxsw_sp, local_port, false, module, + width, 0); + } +} + static int mlxsw_sp_port_split(struct mlxsw_core *mlxsw_core, u8 local_port, unsigned int count) { struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); struct mlxsw_sp_port *mlxsw_sp_port; - u8 width = MLXSW_PORT_MODULE_MAX_WIDTH / count; u8 module, cur_width, base_port; int i; int err; @@ -1965,18 +2012,14 @@ static int mlxsw_sp_port_split(struct mlxsw_core *mlxsw_core, u8 local_port, return -EINVAL; } + module = mlxsw_sp_port->mapping.module; + cur_width = mlxsw_sp_port->mapping.width; + if (count != 2 && count != 4) { netdev_err(mlxsw_sp_port->dev, "Port can only be split into 2 or 4 ports\n"); return -EINVAL; } - err = mlxsw_sp_port_module_info_get(mlxsw_sp, local_port, &module, - &cur_width); - if (err) { - netdev_err(mlxsw_sp_port->dev, "Failed to get port's width\n"); - return err; - } - if (cur_width != MLXSW_PORT_MODULE_MAX_WIDTH) { netdev_err(mlxsw_sp_port->dev, "Port cannot be split further\n"); return -EINVAL; @@ -2001,25 +2044,16 @@ static int mlxsw_sp_port_split(struct mlxsw_core *mlxsw_core, u8 local_port, for (i = 0; i < count; i++) mlxsw_sp_port_remove(mlxsw_sp, base_port + i); - for (i = 0; i < count; i++) { - err = mlxsw_sp_port_create(mlxsw_sp, base_port + i, true, - module, width, i * width); - if (err) { - dev_err(mlxsw_sp->bus_info->dev, "Failed to create split port\n"); - goto err_port_create; - } + err = mlxsw_sp_port_split_create(mlxsw_sp, base_port, module, count); + if (err) { + dev_err(mlxsw_sp->bus_info->dev, "Failed to create split ports\n"); + goto err_port_split_create; } return 0; -err_port_create: - for (i--; i >= 0; i--) - mlxsw_sp_port_remove(mlxsw_sp, base_port + i); - for (i = 0; i < count / 2; i++) { - module = mlxsw_sp->port_to_module[base_port + i * 2]; - mlxsw_sp_port_create(mlxsw_sp, base_port + i * 2, false, - module, MLXSW_PORT_MODULE_MAX_WIDTH, 0); - } +err_port_split_create: + mlxsw_sp_port_unsplit_create(mlxsw_sp, base_port, count); return err; } @@ -2027,10 +2061,9 @@ static int mlxsw_sp_port_unsplit(struct mlxsw_core *mlxsw_core, u8 local_port) { struct mlxsw_sp *mlxsw_sp = mlxsw_core_driver_priv(mlxsw_core); struct mlxsw_sp_port *mlxsw_sp_port; - u8 module, cur_width, base_port; + u8 cur_width, base_port; unsigned int count; int i; - int err; mlxsw_sp_port = mlxsw_sp->ports[local_port]; if (!mlxsw_sp_port) { @@ -2044,12 +2077,7 @@ static int mlxsw_sp_port_unsplit(struct mlxsw_core *mlxsw_core, u8 local_port) return -EINVAL; } - err = mlxsw_sp_port_module_info_get(mlxsw_sp, local_port, &module, - &cur_width); - if (err) { - netdev_err(mlxsw_sp_port->dev, "Failed to get port's width\n"); - return err; - } + cur_width = mlxsw_sp_port->mapping.width; count = cur_width == 1 ? 4 : 2; base_port = mlxsw_sp_cluster_base_port_get(local_port); @@ -2061,14 +2089,7 @@ static int mlxsw_sp_port_unsplit(struct mlxsw_core *mlxsw_core, u8 local_port) for (i = 0; i < count; i++) mlxsw_sp_port_remove(mlxsw_sp, base_port + i); - for (i = 0; i < count / 2; i++) { - module = mlxsw_sp->port_to_module[base_port + i * 2]; - err = mlxsw_sp_port_create(mlxsw_sp, base_port + i * 2, false, - module, MLXSW_PORT_MODULE_MAX_WIDTH, - 0); - if (err) - dev_err(mlxsw_sp->bus_info->dev, "Failed to reinstantiate port\n"); - } + mlxsw_sp_port_unsplit_create(mlxsw_sp, base_port, count); return 0; } diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h index e2c022d3e2f3..13b30eaa13d4 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum.h +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum.h @@ -229,6 +229,11 @@ struct mlxsw_sp_port { struct ieee_maxrate *maxrate; struct ieee_pfc *pfc; } dcb; + struct { + u8 module; + u8 width; + u8 lane; + } mapping; /* 802.1Q bridge VLANs */ unsigned long *active_vlans; unsigned long *untagged_vlans; diff --git a/drivers/net/ethernet/qlogic/qed/qed_main.c b/drivers/net/ethernet/qlogic/qed/qed_main.c index 753064679bde..61cc6869fa65 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_main.c +++ b/drivers/net/ethernet/qlogic/qed/qed_main.c @@ -1105,6 +1105,39 @@ static int qed_get_port_type(u32 media_type) return port_type; } +static int qed_get_link_data(struct qed_hwfn *hwfn, + struct qed_mcp_link_params *params, + struct qed_mcp_link_state *link, + struct qed_mcp_link_capabilities *link_caps) +{ + void *p; + + if (!IS_PF(hwfn->cdev)) { + qed_vf_get_link_params(hwfn, params); + qed_vf_get_link_state(hwfn, link); + qed_vf_get_link_caps(hwfn, link_caps); + + return 0; + } + + p = qed_mcp_get_link_params(hwfn); + if (!p) + return -ENXIO; + memcpy(params, p, sizeof(*params)); + + p = qed_mcp_get_link_state(hwfn); + if (!p) + return -ENXIO; + memcpy(link, p, sizeof(*link)); + + p = qed_mcp_get_link_capabilities(hwfn); + if (!p) + return -ENXIO; + memcpy(link_caps, p, sizeof(*link_caps)); + + return 0; +} + static void qed_fill_link(struct qed_hwfn *hwfn, struct qed_link_output *if_link) { @@ -1116,15 +1149,9 @@ static void qed_fill_link(struct qed_hwfn *hwfn, memset(if_link, 0, sizeof(*if_link)); /* Prepare source inputs */ - if (IS_PF(hwfn->cdev)) { - memcpy(¶ms, qed_mcp_get_link_params(hwfn), sizeof(params)); - memcpy(&link, qed_mcp_get_link_state(hwfn), sizeof(link)); - memcpy(&link_caps, qed_mcp_get_link_capabilities(hwfn), - sizeof(link_caps)); - } else { - qed_vf_get_link_params(hwfn, ¶ms); - qed_vf_get_link_state(hwfn, &link); - qed_vf_get_link_caps(hwfn, &link_caps); + if (qed_get_link_data(hwfn, ¶ms, &link, &link_caps)) { + dev_warn(&hwfn->cdev->pdev->dev, "no link data available\n"); + return; } /* Set the link parameters to pass to protocol driver */ diff --git a/drivers/net/ethernet/qlogic/qed/qed_sriov.h b/drivers/net/ethernet/qlogic/qed/qed_sriov.h index c8667c65e685..c90b2b6ad969 100644 --- a/drivers/net/ethernet/qlogic/qed/qed_sriov.h +++ b/drivers/net/ethernet/qlogic/qed/qed_sriov.h @@ -12,11 +12,13 @@ #include "qed_vf.h" #define QED_VF_ARRAY_LENGTH (3) +#ifdef CONFIG_QED_SRIOV #define IS_VF(cdev) ((cdev)->b_is_vf) #define IS_PF(cdev) (!((cdev)->b_is_vf)) -#ifdef CONFIG_QED_SRIOV #define IS_PF_SRIOV(p_hwfn) (!!((p_hwfn)->cdev->p_iov_info)) #else +#define IS_VF(cdev) (0) +#define IS_PF(cdev) (1) #define IS_PF_SRIOV(p_hwfn) (0) #endif #define IS_PF_SRIOV_ALLOC(p_hwfn) (!!((p_hwfn)->pf_iov_info)) diff --git a/drivers/net/ethernet/qlogic/qede/qede_main.c b/drivers/net/ethernet/qlogic/qede/qede_main.c index 5d00d1404bfc..5733d1888223 100644 --- a/drivers/net/ethernet/qlogic/qede/qede_main.c +++ b/drivers/net/ethernet/qlogic/qede/qede_main.c @@ -87,7 +87,9 @@ static const struct pci_device_id qede_pci_tbl[] = { {PCI_VDEVICE(QLOGIC, PCI_DEVICE_ID_57980S_100), QEDE_PRIVATE_PF}, {PCI_VDEVICE(QLOGIC, PCI_DEVICE_ID_57980S_50), QEDE_PRIVATE_PF}, {PCI_VDEVICE(QLOGIC, PCI_DEVICE_ID_57980S_25), QEDE_PRIVATE_PF}, +#ifdef CONFIG_QED_SRIOV {PCI_VDEVICE(QLOGIC, PCI_DEVICE_ID_57980S_IOV), QEDE_PRIVATE_VF}, +#endif { 0 } }; diff --git a/drivers/net/ethernet/sfc/mcdi_port.c b/drivers/net/ethernet/sfc/mcdi_port.c index 7f295c4d7b80..2a9228a6e4a0 100644 --- a/drivers/net/ethernet/sfc/mcdi_port.c +++ b/drivers/net/ethernet/sfc/mcdi_port.c @@ -189,11 +189,12 @@ static u32 mcdi_to_ethtool_cap(u32 media, u32 cap) case MC_CMD_MEDIA_XFP: case MC_CMD_MEDIA_SFP_PLUS: - result |= SUPPORTED_FIBRE; - break; - case MC_CMD_MEDIA_QSFP_PLUS: result |= SUPPORTED_FIBRE; + if (cap & (1 << MC_CMD_PHY_CAP_1000FDX_LBN)) + result |= SUPPORTED_1000baseT_Full; + if (cap & (1 << MC_CMD_PHY_CAP_10000FDX_LBN)) + result |= SUPPORTED_10000baseT_Full; if (cap & (1 << MC_CMD_PHY_CAP_40000FDX_LBN)) result |= SUPPORTED_40000baseCR4_Full; break; diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c b/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c index 4f7283d05588..44da877d2483 100644 --- a/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c @@ -156,7 +156,7 @@ static void dwmac4_set_filter(struct mac_device_info *hw, struct netdev_hw_addr *ha; netdev_for_each_uc_addr(ha, dev) { - dwmac4_set_umac_addr(ioaddr, ha->addr, reg); + dwmac4_set_umac_addr(hw, ha->addr, reg); reg++; } } diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c index eac45d0c75e2..a473c182c91d 100644 --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c @@ -3450,8 +3450,6 @@ int stmmac_resume(struct device *dev) if (!netif_running(ndev)) return 0; - spin_lock_irqsave(&priv->lock, flags); - /* Power Down bit, into the PM register, is cleared * automatically as soon as a magic packet or a Wake-up frame * is received. Anyway, it's better to manually clear @@ -3459,7 +3457,9 @@ int stmmac_resume(struct device *dev) * from another devices (e.g. serial console). */ if (device_may_wakeup(priv->device)) { + spin_lock_irqsave(&priv->lock, flags); priv->hw->mac->pmt(priv->hw, 0); + spin_unlock_irqrestore(&priv->lock, flags); priv->irq_wake = 0; } else { pinctrl_pm_select_default_state(priv->device); @@ -3473,6 +3473,8 @@ int stmmac_resume(struct device *dev) netif_device_attach(ndev); + spin_lock_irqsave(&priv->lock, flags); + priv->cur_rx = 0; priv->dirty_rx = 0; priv->dirty_tx = 0; diff --git a/drivers/net/ethernet/ti/cpsw.c b/drivers/net/ethernet/ti/cpsw.c index 4b08a2f52b3e..e6bb0ecb12c7 100644 --- a/drivers/net/ethernet/ti/cpsw.c +++ b/drivers/net/ethernet/ti/cpsw.c @@ -1339,7 +1339,7 @@ static int cpsw_ndo_open(struct net_device *ndev) if (priv->coal_intvl != 0) { struct ethtool_coalesce coal; - coal.rx_coalesce_usecs = (priv->coal_intvl << 4); + coal.rx_coalesce_usecs = priv->coal_intvl; cpsw_set_coalesce(ndev, &coal); } diff --git a/drivers/net/vmxnet3/vmxnet3_drv.c b/drivers/net/vmxnet3/vmxnet3_drv.c index db8022ae415b..08885bc8d6db 100644 --- a/drivers/net/vmxnet3/vmxnet3_drv.c +++ b/drivers/net/vmxnet3/vmxnet3_drv.c @@ -1369,7 +1369,7 @@ vmxnet3_rq_rx_complete(struct vmxnet3_rx_queue *rq, rcdlro = (struct Vmxnet3_RxCompDescExt *)rcd; segCnt = rcdlro->segCnt; - BUG_ON(segCnt <= 1); + WARN_ON_ONCE(segCnt == 0); mss = rcdlro->mss; if (unlikely(segCnt <= 1)) segCnt = 0; diff --git a/drivers/net/vmxnet3/vmxnet3_int.h b/drivers/net/vmxnet3/vmxnet3_int.h index c4825392d64b..3d2b64e63408 100644 --- a/drivers/net/vmxnet3/vmxnet3_int.h +++ b/drivers/net/vmxnet3/vmxnet3_int.h @@ -69,10 +69,10 @@ /* * Version numbers */ -#define VMXNET3_DRIVER_VERSION_STRING "1.4.7.0-k" +#define VMXNET3_DRIVER_VERSION_STRING "1.4.8.0-k" /* a 32-bit int, each byte encode a verion number in VMXNET3_DRIVER_VERSION */ -#define VMXNET3_DRIVER_VERSION_NUM 0x01040700 +#define VMXNET3_DRIVER_VERSION_NUM 0x01040800 #if defined(CONFIG_PCI_MSI) /* RSS only makes sense if MSI-X is supported. */ diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c index d0631b6cfd53..62f475e31077 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/cfg80211.c @@ -2540,12 +2540,14 @@ brcmf_cfg80211_get_station(struct wiphy *wiphy, struct net_device *ndev, const u8 *mac, struct station_info *sinfo) { struct brcmf_if *ifp = netdev_priv(ndev); + struct brcmf_scb_val_le scb_val; s32 err = 0; struct brcmf_sta_info_le sta_info_le; u32 sta_flags; u32 is_tdls_peer; s32 total_rssi; s32 count_rssi; + int rssi; u32 i; brcmf_dbg(TRACE, "Enter, MAC %pM\n", mac); @@ -2629,6 +2631,20 @@ brcmf_cfg80211_get_station(struct wiphy *wiphy, struct net_device *ndev, sinfo->filled |= BIT(NL80211_STA_INFO_SIGNAL); total_rssi /= count_rssi; sinfo->signal = total_rssi; + } else if (test_bit(BRCMF_VIF_STATUS_CONNECTED, + &ifp->vif->sme_state)) { + memset(&scb_val, 0, sizeof(scb_val)); + err = brcmf_fil_cmd_data_get(ifp, BRCMF_C_GET_RSSI, + &scb_val, sizeof(scb_val)); + if (err) { + brcmf_err("Could not get rssi (%d)\n", err); + goto done; + } else { + rssi = le32_to_cpu(scb_val.val); + sinfo->filled |= BIT(NL80211_STA_INFO_SIGNAL); + sinfo->signal = rssi; + brcmf_dbg(CONN, "RSSI %d dBm\n", rssi); + } } } done: diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c index 68f1ce02f4bf..2b9a2bc429d6 100644 --- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c +++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/msgbuf.c @@ -1157,6 +1157,8 @@ brcmf_msgbuf_process_rx_complete(struct brcmf_msgbuf *msgbuf, void *buf) brcmu_pkt_buf_free_skb(skb); return; } + + skb->protocol = eth_type_trans(skb, ifp->ndev); brcmf_netif_rx(ifp, skb); } diff --git a/drivers/net/wireless/mac80211_hwsim.c b/drivers/net/wireless/mac80211_hwsim.c index 9ed0ed1bf514..4dd5adcdd29b 100644 --- a/drivers/net/wireless/mac80211_hwsim.c +++ b/drivers/net/wireless/mac80211_hwsim.c @@ -2776,6 +2776,7 @@ static int hwsim_tx_info_frame_received_nl(struct sk_buff *skb_2, if (!info->attrs[HWSIM_ATTR_ADDR_TRANSMITTER] || !info->attrs[HWSIM_ATTR_FLAGS] || !info->attrs[HWSIM_ATTR_COOKIE] || + !info->attrs[HWSIM_ATTR_SIGNAL] || !info->attrs[HWSIM_ATTR_TX_INFO]) goto out; diff --git a/drivers/net/wireless/realtek/rtlwifi/core.c b/drivers/net/wireless/realtek/rtlwifi/core.c index 0f48048b8654..3a0faa8fe9d4 100644 --- a/drivers/net/wireless/realtek/rtlwifi/core.c +++ b/drivers/net/wireless/realtek/rtlwifi/core.c @@ -54,7 +54,7 @@ EXPORT_SYMBOL(channel5g_80m); void rtl_addr_delay(u32 addr) { if (addr == 0xfe) - msleep(50); + mdelay(50); else if (addr == 0xfd) msleep(5); else if (addr == 0xfc) @@ -75,7 +75,7 @@ void rtl_rfreg_delay(struct ieee80211_hw *hw, enum radio_path rfpath, u32 addr, rtl_addr_delay(addr); } else { rtl_set_rfreg(hw, rfpath, addr, mask, data); - usleep_range(1, 2); + udelay(1); } } EXPORT_SYMBOL(rtl_rfreg_delay); @@ -86,7 +86,7 @@ void rtl_bb_delay(struct ieee80211_hw *hw, u32 addr, u32 data) rtl_addr_delay(addr); } else { rtl_set_bbreg(hw, addr, MASKDWORD, data); - usleep_range(1, 2); + udelay(1); } } EXPORT_SYMBOL(rtl_bb_delay); diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index 78dca3193ca4..befac5b19490 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -1679,9 +1679,14 @@ static int nvme_pci_enable(struct nvme_dev *dev) static void nvme_dev_unmap(struct nvme_dev *dev) { + struct pci_dev *pdev = to_pci_dev(dev->dev); + int bars; + if (dev->bar) iounmap(dev->bar); - pci_release_regions(to_pci_dev(dev->dev)); + + bars = pci_select_bars(pdev, IORESOURCE_MEM); + pci_release_selected_regions(pdev, bars); } static void nvme_pci_disable(struct nvme_dev *dev) @@ -1924,7 +1929,7 @@ static int nvme_dev_map(struct nvme_dev *dev) return 0; release: - pci_release_regions(pdev); + pci_release_selected_regions(pdev, bars); return -ENODEV; } diff --git a/drivers/of/fdt.c b/drivers/of/fdt.c index 14f2f8c7c260..33daffc4392c 100644 --- a/drivers/of/fdt.c +++ b/drivers/of/fdt.c @@ -395,7 +395,7 @@ static int unflatten_dt_nodes(const void *blob, struct device_node **nodepp) { struct device_node *root; - int offset = 0, depth = 0; + int offset = 0, depth = 0, initial_depth = 0; #define FDT_MAX_DEPTH 64 unsigned int fpsizes[FDT_MAX_DEPTH]; struct device_node *nps[FDT_MAX_DEPTH]; @@ -405,11 +405,22 @@ static int unflatten_dt_nodes(const void *blob, if (nodepp) *nodepp = NULL; + /* + * We're unflattening device sub-tree if @dad is valid. There are + * possibly multiple nodes in the first level of depth. We need + * set @depth to 1 to make fdt_next_node() happy as it bails + * immediately when negative @depth is found. Otherwise, the device + * nodes except the first one won't be unflattened successfully. + */ + if (dad) + depth = initial_depth = 1; + root = dad; fpsizes[depth] = dad ? strlen(of_node_full_name(dad)) : 0; nps[depth] = dad; + for (offset = 0; - offset >= 0 && depth >= 0; + offset >= 0 && depth >= initial_depth; offset = fdt_next_node(blob, offset, &depth)) { if (WARN_ON_ONCE(depth >= FDT_MAX_DEPTH)) continue; diff --git a/drivers/of/irq.c b/drivers/of/irq.c index e7bfc175b8e1..6ec743faabe8 100644 --- a/drivers/of/irq.c +++ b/drivers/of/irq.c @@ -386,13 +386,13 @@ int of_irq_to_resource(struct device_node *dev, int index, struct resource *r) EXPORT_SYMBOL_GPL(of_irq_to_resource); /** - * of_irq_get - Decode a node's IRQ and return it as a Linux irq number + * of_irq_get - Decode a node's IRQ and return it as a Linux IRQ number * @dev: pointer to device tree node - * @index: zero-based index of the irq - * - * Returns Linux irq number on success, or -EPROBE_DEFER if the irq domain - * is not yet created. + * @index: zero-based index of the IRQ * + * Returns Linux IRQ number on success, or 0 on the IRQ mapping failure, or + * -EPROBE_DEFER if the IRQ domain is not yet created, or error code in case + * of any other failure. */ int of_irq_get(struct device_node *dev, int index) { @@ -413,12 +413,13 @@ int of_irq_get(struct device_node *dev, int index) EXPORT_SYMBOL_GPL(of_irq_get); /** - * of_irq_get_byname - Decode a node's IRQ and return it as a Linux irq number + * of_irq_get_byname - Decode a node's IRQ and return it as a Linux IRQ number * @dev: pointer to device tree node - * @name: irq name + * @name: IRQ name * - * Returns Linux irq number on success, or -EPROBE_DEFER if the irq domain - * is not yet created, or error code in case of any other failure. + * Returns Linux IRQ number on success, or 0 on the IRQ mapping failure, or + * -EPROBE_DEFER if the IRQ domain is not yet created, or error code in case + * of any other failure. */ int of_irq_get_byname(struct device_node *dev, const char *name) { diff --git a/drivers/of/of_reserved_mem.c b/drivers/of/of_reserved_mem.c index ed01c0172e4a..216648233874 100644 --- a/drivers/of/of_reserved_mem.c +++ b/drivers/of/of_reserved_mem.c @@ -127,8 +127,15 @@ static int __init __reserved_mem_alloc_size(unsigned long node, } /* Need adjust the alignment to satisfy the CMA requirement */ - if (IS_ENABLED(CONFIG_CMA) && of_flat_dt_is_compatible(node, "shared-dma-pool")) - align = max(align, (phys_addr_t)PAGE_SIZE << max(MAX_ORDER - 1, pageblock_order)); + if (IS_ENABLED(CONFIG_CMA) + && of_flat_dt_is_compatible(node, "shared-dma-pool") + && of_get_flat_dt_prop(node, "reusable", NULL) + && !of_get_flat_dt_prop(node, "no-map", NULL)) { + unsigned long order = + max_t(unsigned long, MAX_ORDER - 1, pageblock_order); + + align = max(align, (phys_addr_t)PAGE_SIZE << order); + } prop = of_get_flat_dt_prop(node, "alloc-ranges", &len); if (prop) { diff --git a/drivers/pci/vc.c b/drivers/pci/vc.c index dfbab61a1b47..1fa3a3219c45 100644 --- a/drivers/pci/vc.c +++ b/drivers/pci/vc.c @@ -221,9 +221,9 @@ static int pci_vc_do_save_buffer(struct pci_dev *dev, int pos, else pci_write_config_word(dev, pos + PCI_VC_PORT_CTRL, *(u16 *)buf); - buf += 2; + buf += 4; } - len += 2; + len += 4; /* * If we have any Low Priority VCs and a VC Arbitration Table Offset diff --git a/drivers/perf/arm_pmu.c b/drivers/perf/arm_pmu.c index 1b8304e1efaa..140436a046c0 100644 --- a/drivers/perf/arm_pmu.c +++ b/drivers/perf/arm_pmu.c @@ -1010,8 +1010,8 @@ int arm_pmu_device_probe(struct platform_device *pdev, if (!ret) ret = init_fn(pmu); } else { - ret = probe_current_pmu(pmu, probe_table); cpumask_setall(&pmu->supported_cpus); + ret = probe_current_pmu(pmu, probe_table); } if (ret) { diff --git a/drivers/phy/phy-exynos-mipi-video.c b/drivers/phy/phy-exynos-mipi-video.c index cc093ebfda94..8b851f718123 100644 --- a/drivers/phy/phy-exynos-mipi-video.c +++ b/drivers/phy/phy-exynos-mipi-video.c @@ -233,8 +233,12 @@ static inline int __is_running(const struct exynos_mipi_phy_desc *data, struct exynos_mipi_video_phy *state) { u32 val; + int ret; + + ret = regmap_read(state->regmaps[data->resetn_map], data->resetn_reg, &val); + if (ret) + return 0; - regmap_read(state->regmaps[data->resetn_map], data->resetn_reg, &val); return val & data->resetn_val; } diff --git a/drivers/phy/phy-ti-pipe3.c b/drivers/phy/phy-ti-pipe3.c index 0a477d24cf76..bf46844dc387 100644 --- a/drivers/phy/phy-ti-pipe3.c +++ b/drivers/phy/phy-ti-pipe3.c @@ -293,11 +293,18 @@ static int ti_pipe3_init(struct phy *x) ret = ti_pipe3_dpll_wait_lock(phy); } - /* Program the DPLL only if not locked */ + /* SATA has issues if re-programmed when locked */ val = ti_pipe3_readl(phy->pll_ctrl_base, PLL_STATUS); - if (!(val & PLL_LOCK)) - if (ti_pipe3_dpll_program(phy)) - return -EINVAL; + if ((val & PLL_LOCK) && of_device_is_compatible(phy->dev->of_node, + "ti,phy-pipe3-sata")) + return ret; + + /* Program the DPLL */ + ret = ti_pipe3_dpll_program(phy); + if (ret) { + ti_pipe3_disable_clocks(phy); + return -EINVAL; + } return ret; } diff --git a/drivers/phy/phy-twl4030-usb.c b/drivers/phy/phy-twl4030-usb.c index 6b6af6cba454..d9b10a39a2cf 100644 --- a/drivers/phy/phy-twl4030-usb.c +++ b/drivers/phy/phy-twl4030-usb.c @@ -463,7 +463,8 @@ static int twl4030_phy_power_on(struct phy *phy) twl4030_usb_set_mode(twl, twl->usb_mode); if (twl->usb_mode == T2_USB_MODE_ULPI) twl4030_i2c_access(twl, 0); - schedule_delayed_work(&twl->id_workaround_work, 0); + twl->linkstat = MUSB_UNKNOWN; + schedule_delayed_work(&twl->id_workaround_work, HZ); return 0; } @@ -537,6 +538,7 @@ static irqreturn_t twl4030_usb_irq(int irq, void *_twl) struct twl4030_usb *twl = _twl; enum musb_vbus_id_status status; bool status_changed = false; + int err; status = twl4030_usb_linkstat(twl); @@ -567,7 +569,9 @@ static irqreturn_t twl4030_usb_irq(int irq, void *_twl) pm_runtime_mark_last_busy(twl->dev); pm_runtime_put_autosuspend(twl->dev); } - musb_mailbox(status); + err = musb_mailbox(status); + if (err) + twl->linkstat = MUSB_UNKNOWN; } /* don't schedule during sleep - irq works right then */ @@ -595,7 +599,8 @@ static int twl4030_phy_init(struct phy *phy) struct twl4030_usb *twl = phy_get_drvdata(phy); pm_runtime_get_sync(twl->dev); - schedule_delayed_work(&twl->id_workaround_work, 0); + twl->linkstat = MUSB_UNKNOWN; + schedule_delayed_work(&twl->id_workaround_work, HZ); pm_runtime_mark_last_busy(twl->dev); pm_runtime_put_autosuspend(twl->dev); @@ -763,7 +768,8 @@ static int twl4030_usb_remove(struct platform_device *pdev) if (cable_present(twl->linkstat)) pm_runtime_put_noidle(twl->dev); pm_runtime_mark_last_busy(twl->dev); - pm_runtime_put_sync_suspend(twl->dev); + pm_runtime_dont_use_autosuspend(&pdev->dev); + pm_runtime_put_sync(twl->dev); pm_runtime_disable(twl->dev); /* autogate 60MHz ULPI clock, diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig index c06bb85c2839..3ec0025d19e7 100644 --- a/drivers/platform/x86/Kconfig +++ b/drivers/platform/x86/Kconfig @@ -103,7 +103,6 @@ config DELL_SMBIOS config DELL_LAPTOP tristate "Dell Laptop Extras" - depends on X86 depends on DELL_SMBIOS depends on DMI depends on BACKLIGHT_CLASS_DEVICE @@ -505,7 +504,7 @@ config THINKPAD_ACPI_HOTKEY_POLL config SENSORS_HDAPS tristate "Thinkpad Hard Drive Active Protection System (hdaps)" - depends on INPUT && X86 + depends on INPUT select INPUT_POLLDEV default n help @@ -749,7 +748,7 @@ config TOSHIBA_WMI config ACPI_CMPC tristate "CMPC Laptop Extras" - depends on X86 && ACPI + depends on ACPI depends on RFKILL || RFKILL=n select INPUT select BACKLIGHT_CLASS_DEVICE @@ -848,7 +847,7 @@ config INTEL_IMR config INTEL_PMC_CORE bool "Intel PMC Core driver" - depends on X86 && PCI + depends on PCI ---help--- The Intel Platform Controller Hub for Intel Core SoCs provides access to Power Management Controller registers via a PCI interface. This @@ -860,7 +859,7 @@ config INTEL_PMC_CORE config IBM_RTL tristate "Device driver to enable PRTL support" - depends on X86 && PCI + depends on PCI ---help--- Enable support for IBM Premium Real Time Mode (PRTM). This module will allow you the enter and exit PRTM in the BIOS via @@ -894,7 +893,6 @@ config XO15_EBOOK config SAMSUNG_LAPTOP tristate "Samsung Laptop driver" - depends on X86 depends on RFKILL || RFKILL = n depends on ACPI_VIDEO || ACPI_VIDEO = n depends on BACKLIGHT_CLASS_DEVICE diff --git a/drivers/platform/x86/ideapad-laptop.c b/drivers/platform/x86/ideapad-laptop.c index 4a23fbc66b71..d1a091b93192 100644 --- a/drivers/platform/x86/ideapad-laptop.c +++ b/drivers/platform/x86/ideapad-laptop.c @@ -567,6 +567,7 @@ static void ideapad_sysfs_exit(struct ideapad_private *priv) static const struct key_entry ideapad_keymap[] = { { KE_KEY, 6, { KEY_SWITCHVIDEOMODE } }, { KE_KEY, 7, { KEY_CAMERA } }, + { KE_KEY, 8, { KEY_MICMUTE } }, { KE_KEY, 11, { KEY_F16 } }, { KE_KEY, 13, { KEY_WLAN } }, { KE_KEY, 16, { KEY_PROG1 } }, @@ -809,6 +810,7 @@ static void ideapad_acpi_notify(acpi_handle handle, u32 event, void *data) break; case 13: case 11: + case 8: case 7: case 6: ideapad_input_report(priv, vpc_bit); diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c index c3bfa1fe95bf..b65ce7519411 100644 --- a/drivers/platform/x86/thinkpad_acpi.c +++ b/drivers/platform/x86/thinkpad_acpi.c @@ -2043,6 +2043,7 @@ static int hotkey_autosleep_ack; static u32 hotkey_orig_mask; /* events the BIOS had enabled */ static u32 hotkey_all_mask; /* all events supported in fw */ +static u32 hotkey_adaptive_all_mask; /* all adaptive events supported in fw */ static u32 hotkey_reserved_mask; /* events better left disabled */ static u32 hotkey_driver_mask; /* events needed by the driver */ static u32 hotkey_user_mask; /* events visible to userspace */ @@ -2742,6 +2743,17 @@ static ssize_t hotkey_all_mask_show(struct device *dev, static DEVICE_ATTR_RO(hotkey_all_mask); +/* sysfs hotkey all_mask ----------------------------------------------- */ +static ssize_t hotkey_adaptive_all_mask_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + return snprintf(buf, PAGE_SIZE, "0x%08x\n", + hotkey_adaptive_all_mask | hotkey_source_mask); +} + +static DEVICE_ATTR_RO(hotkey_adaptive_all_mask); + /* sysfs hotkey recommended_mask --------------------------------------- */ static ssize_t hotkey_recommended_mask_show(struct device *dev, struct device_attribute *attr, @@ -2985,6 +2997,7 @@ static struct attribute *hotkey_attributes[] __initdata = { &dev_attr_wakeup_hotunplug_complete.attr, &dev_attr_hotkey_mask.attr, &dev_attr_hotkey_all_mask.attr, + &dev_attr_hotkey_adaptive_all_mask.attr, &dev_attr_hotkey_recommended_mask.attr, #ifdef CONFIG_THINKPAD_ACPI_HOTKEY_POLL &dev_attr_hotkey_source_mask.attr, @@ -3321,20 +3334,6 @@ static int __init hotkey_init(struct ibm_init_struct *iibm) if (!tp_features.hotkey) return 1; - /* - * Check if we have an adaptive keyboard, like on the - * Lenovo Carbon X1 2014 (2nd Gen). - */ - if (acpi_evalf(hkey_handle, &hkeyv, "MHKV", "qd")) { - if ((hkeyv >> 8) == 2) { - tp_features.has_adaptive_kbd = true; - res = sysfs_create_group(&tpacpi_pdev->dev.kobj, - &adaptive_kbd_attr_group); - if (res) - goto err_exit; - } - } - quirks = tpacpi_check_quirks(tpacpi_hotkey_qtable, ARRAY_SIZE(tpacpi_hotkey_qtable)); @@ -3357,30 +3356,70 @@ static int __init hotkey_init(struct ibm_init_struct *iibm) A30, R30, R31, T20-22, X20-21, X22-24. Detected by checking for HKEY interface version 0x100 */ if (acpi_evalf(hkey_handle, &hkeyv, "MHKV", "qd")) { - if ((hkeyv >> 8) != 1) { - pr_err("unknown version of the HKEY interface: 0x%x\n", - hkeyv); - pr_err("please report this to %s\n", TPACPI_MAIL); - } else { + vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_HKEY, + "firmware HKEY interface version: 0x%x\n", + hkeyv); + + switch (hkeyv >> 8) { + case 1: /* * MHKV 0x100 in A31, R40, R40e, * T4x, X31, and later */ - vdbg_printk(TPACPI_DBG_INIT | TPACPI_DBG_HKEY, - "firmware HKEY interface version: 0x%x\n", - hkeyv); /* Paranoia check AND init hotkey_all_mask */ if (!acpi_evalf(hkey_handle, &hotkey_all_mask, "MHKA", "qd")) { - pr_err("missing MHKA handler, " - "please report this to %s\n", + pr_err("missing MHKA handler, please report this to %s\n", TPACPI_MAIL); /* Fallback: pre-init for FN+F3,F4,F12 */ hotkey_all_mask = 0x080cU; } else { tp_features.hotkey_mask = 1; } + break; + + case 2: + /* + * MHKV 0x200 in X1, T460s, X260, T560, X1 Tablet (2016) + */ + + /* Paranoia check AND init hotkey_all_mask */ + if (!acpi_evalf(hkey_handle, &hotkey_all_mask, + "MHKA", "dd", 1)) { + pr_err("missing MHKA handler, please report this to %s\n", + TPACPI_MAIL); + /* Fallback: pre-init for FN+F3,F4,F12 */ + hotkey_all_mask = 0x080cU; + } else { + tp_features.hotkey_mask = 1; + } + + /* + * Check if we have an adaptive keyboard, like on the + * Lenovo Carbon X1 2014 (2nd Gen). + */ + if (acpi_evalf(hkey_handle, &hotkey_adaptive_all_mask, + "MHKA", "dd", 2)) { + if (hotkey_adaptive_all_mask != 0) { + tp_features.has_adaptive_kbd = true; + res = sysfs_create_group( + &tpacpi_pdev->dev.kobj, + &adaptive_kbd_attr_group); + if (res) + goto err_exit; + } + } else { + tp_features.has_adaptive_kbd = false; + hotkey_adaptive_all_mask = 0x0U; + } + break; + + default: + pr_err("unknown version of the HKEY interface: 0x%x\n", + hkeyv); + pr_err("please report this to %s\n", TPACPI_MAIL); + break; } } diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c index dba3843c53b8..ed337a8c34ab 100644 --- a/drivers/pwm/core.c +++ b/drivers/pwm/core.c @@ -457,7 +457,8 @@ int pwm_apply_state(struct pwm_device *pwm, struct pwm_state *state) { int err; - if (!pwm) + if (!pwm || !state || !state->period || + state->duty_cycle > state->period) return -EINVAL; if (!memcmp(state, &pwm->state, sizeof(*state))) diff --git a/drivers/pwm/pwm-atmel-hlcdc.c b/drivers/pwm/pwm-atmel-hlcdc.c index f994c7eaf41c..14fc011faa32 100644 --- a/drivers/pwm/pwm-atmel-hlcdc.c +++ b/drivers/pwm/pwm-atmel-hlcdc.c @@ -272,7 +272,7 @@ static int atmel_hlcdc_pwm_probe(struct platform_device *pdev) chip->chip.of_pwm_n_cells = 3; chip->chip.can_sleep = 1; - ret = pwmchip_add(&chip->chip); + ret = pwmchip_add_with_polarity(&chip->chip, PWM_POLARITY_INVERSED); if (ret) { clk_disable_unprepare(hlcdc->periph_clk); return ret; diff --git a/drivers/pwm/sysfs.c b/drivers/pwm/sysfs.c index d98599249a05..01695d48dd54 100644 --- a/drivers/pwm/sysfs.c +++ b/drivers/pwm/sysfs.c @@ -152,7 +152,7 @@ static ssize_t enable_store(struct device *child, goto unlock; } - pwm_apply_state(pwm, &state); + ret = pwm_apply_state(pwm, &state); unlock: mutex_unlock(&export->lock); diff --git a/drivers/regulator/qcom_smd-regulator.c b/drivers/regulator/qcom_smd-regulator.c index 56a17ec5b5ef..526bf23dcb49 100644 --- a/drivers/regulator/qcom_smd-regulator.c +++ b/drivers/regulator/qcom_smd-regulator.c @@ -140,6 +140,19 @@ static const struct regulator_ops rpm_smps_ldo_ops = { .enable = rpm_reg_enable, .disable = rpm_reg_disable, .is_enabled = rpm_reg_is_enabled, + .list_voltage = regulator_list_voltage_linear_range, + + .get_voltage = rpm_reg_get_voltage, + .set_voltage = rpm_reg_set_voltage, + + .set_load = rpm_reg_set_load, +}; + +static const struct regulator_ops rpm_smps_ldo_ops_fixed = { + .enable = rpm_reg_enable, + .disable = rpm_reg_disable, + .is_enabled = rpm_reg_is_enabled, + .list_voltage = regulator_list_voltage_linear_range, .get_voltage = rpm_reg_get_voltage, .set_voltage = rpm_reg_set_voltage, @@ -247,7 +260,7 @@ static const struct regulator_desc pm8941_nldo = { static const struct regulator_desc pm8941_lnldo = { .fixed_uV = 1740000, .n_voltages = 1, - .ops = &rpm_smps_ldo_ops, + .ops = &rpm_smps_ldo_ops_fixed, }; static const struct regulator_desc pm8941_switch = { diff --git a/drivers/regulator/tps51632-regulator.c b/drivers/regulator/tps51632-regulator.c index 572816e30095..c139890c1514 100644 --- a/drivers/regulator/tps51632-regulator.c +++ b/drivers/regulator/tps51632-regulator.c @@ -94,11 +94,14 @@ static int tps51632_dcdc_set_ramp_delay(struct regulator_dev *rdev, int ramp_delay) { struct tps51632_chip *tps = rdev_get_drvdata(rdev); - int bit = ramp_delay/6000; + int bit; int ret; - if (bit) - bit--; + if (ramp_delay == 0) + bit = 0; + else + bit = DIV_ROUND_UP(ramp_delay, 6000) - 1; + ret = regmap_write(tps->regmap, TPS51632_SLEW_REGS, BIT(bit)); if (ret < 0) dev_err(tps->dev, "SLEW reg write failed, err %d\n", ret); diff --git a/drivers/scsi/53c700.c b/drivers/scsi/53c700.c index d4c285688ce9..3ddc85e6efd6 100644 --- a/drivers/scsi/53c700.c +++ b/drivers/scsi/53c700.c @@ -1122,7 +1122,7 @@ process_script_interrupt(__u32 dsps, __u32 dsp, struct scsi_cmnd *SCp, } else { struct scsi_cmnd *SCp; - SCp = scsi_host_find_tag(SDp->host, SCSI_NO_TAG); + SCp = SDp->current_cmnd; if(unlikely(SCp == NULL)) { sdev_printk(KERN_ERR, SDp, "no saved request for untagged cmd\n"); @@ -1826,7 +1826,7 @@ NCR_700_queuecommand_lck(struct scsi_cmnd *SCp, void (*done)(struct scsi_cmnd *) slot->tag, slot); } else { slot->tag = SCSI_NO_TAG; - /* must populate current_cmnd for scsi_host_find_tag to work */ + /* save current command for reselection */ SCp->device->current_cmnd = SCp; } /* sanity check: some of the commands generated by the mid-layer diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c index 3408578b08d6..ff41c310c900 100644 --- a/drivers/scsi/scsi_devinfo.c +++ b/drivers/scsi/scsi_devinfo.c @@ -230,6 +230,7 @@ static struct { {"PIONEER", "CD-ROM DRM-624X", NULL, BLIST_FORCELUN | BLIST_SINGLELUN}, {"Promise", "VTrak E610f", NULL, BLIST_SPARSELUN | BLIST_NO_RSOC}, {"Promise", "", NULL, BLIST_SPARSELUN}, + {"QEMU", "QEMU CD-ROM", NULL, BLIST_SKIP_VPD_PAGES}, {"QNAP", "iSCSI Storage", NULL, BLIST_MAX_1024}, {"SYNOLOGY", "iSCSI Storage", NULL, BLIST_MAX_1024}, {"QUANTUM", "XP34301", "1071", BLIST_NOTQ}, diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index a8b610eaa0ca..106a6adbd6f1 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -1128,7 +1128,6 @@ static int scsi_eh_action(struct scsi_cmnd *scmd, int rtn) */ void scsi_eh_finish_cmd(struct scsi_cmnd *scmd, struct list_head *done_q) { - scmd->device->host->host_failed--; scmd->eh_eflags = 0; list_move_tail(&scmd->eh_entry, done_q); } @@ -2227,6 +2226,9 @@ int scsi_error_handler(void *data) else scsi_unjam_host(shost); + /* All scmds have been handled */ + shost->host_failed = 0; + /* * Note - if the above fails completely, the action is to take * individual devices offline and flush the queue of any diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index f459dff30512..60bff78e9ead 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -2867,10 +2867,10 @@ static int sd_revalidate_disk(struct gendisk *disk) if (sdkp->opt_xfer_blocks && sdkp->opt_xfer_blocks <= dev_max && sdkp->opt_xfer_blocks <= SD_DEF_XFER_BLOCKS && - sdkp->opt_xfer_blocks * sdp->sector_size >= PAGE_SIZE) - rw_max = q->limits.io_opt = - sdkp->opt_xfer_blocks * sdp->sector_size; - else + logical_to_bytes(sdp, sdkp->opt_xfer_blocks) >= PAGE_SIZE) { + q->limits.io_opt = logical_to_bytes(sdp, sdkp->opt_xfer_blocks); + rw_max = logical_to_sectors(sdp, sdkp->opt_xfer_blocks); + } else rw_max = BLK_DEF_MAX_SECTORS; /* Combine with controller limits */ diff --git a/drivers/scsi/sd.h b/drivers/scsi/sd.h index 654630bb7d0e..765a6f1ac1b7 100644 --- a/drivers/scsi/sd.h +++ b/drivers/scsi/sd.h @@ -151,6 +151,11 @@ static inline sector_t logical_to_sectors(struct scsi_device *sdev, sector_t blo return blocks << (ilog2(sdev->sector_size) - 9); } +static inline unsigned int logical_to_bytes(struct scsi_device *sdev, sector_t blocks) +{ + return blocks * sdev->sector_size; +} + /* * A DIF-capable target device can be formatted with different * protection schemes. Currently 0 through 3 are defined: diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c index bbfee53cfcf5..845e49a52430 100644 --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c @@ -2521,12 +2521,13 @@ kiblnd_passive_connect(struct rdma_cm_id *cmid, void *priv, int priv_nob) return 0; failed: - if (ni) + if (ni) { lnet_ni_decref(ni); + rej.ibr_cp.ibcp_queue_depth = kiblnd_msg_queue_size(version, ni); + rej.ibr_cp.ibcp_max_frags = kiblnd_rdma_frags(version, ni); + } rej.ibr_version = version; - rej.ibr_cp.ibcp_queue_depth = kiblnd_msg_queue_size(version, ni); - rej.ibr_cp.ibcp_max_frags = kiblnd_rdma_frags(version, ni); kiblnd_reject(cmid, &rej); return -ECONNREFUSED; diff --git a/drivers/staging/rtl8188eu/core/rtw_efuse.c b/drivers/staging/rtl8188eu/core/rtw_efuse.c index c17870cddb5b..fbce1f7e68ca 100644 --- a/drivers/staging/rtl8188eu/core/rtw_efuse.c +++ b/drivers/staging/rtl8188eu/core/rtw_efuse.c @@ -102,7 +102,7 @@ efuse_phymap_to_logical(u8 *phymap, u16 _offset, u16 _size_byte, u8 *pbuf) if (!efuseTbl) return; - eFuseWord = (u16 **)rtw_malloc2d(EFUSE_MAX_SECTION_88E, EFUSE_MAX_WORD_UNIT, sizeof(*eFuseWord)); + eFuseWord = (u16 **)rtw_malloc2d(EFUSE_MAX_SECTION_88E, EFUSE_MAX_WORD_UNIT, sizeof(u16)); if (!eFuseWord) { DBG_88E("%s: alloc eFuseWord fail!\n", __func__); goto eFuseWord_failed; diff --git a/drivers/staging/rtl8188eu/hal/usb_halinit.c b/drivers/staging/rtl8188eu/hal/usb_halinit.c index 87ea3b844951..363f3a34ddce 100644 --- a/drivers/staging/rtl8188eu/hal/usb_halinit.c +++ b/drivers/staging/rtl8188eu/hal/usb_halinit.c @@ -2072,7 +2072,8 @@ void rtl8188eu_set_hal_ops(struct adapter *adapt) { struct hal_ops *halfunc = &adapt->HalFunc; - adapt->HalData = kzalloc(sizeof(*adapt->HalData), GFP_KERNEL); + + adapt->HalData = kzalloc(sizeof(struct hal_data_8188e), GFP_KERNEL); if (!adapt->HalData) DBG_88E("cant not alloc memory for HAL DATA\n"); diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c index 6ceac4f2d4b2..5b4b47ed948b 100644 --- a/drivers/thermal/cpu_cooling.c +++ b/drivers/thermal/cpu_cooling.c @@ -857,14 +857,6 @@ __cpufreq_cooling_register(struct device_node *np, goto free_power_table; } - snprintf(dev_name, sizeof(dev_name), "thermal-cpufreq-%d", - cpufreq_dev->id); - - cool_dev = thermal_of_cooling_device_register(np, dev_name, cpufreq_dev, - &cpufreq_cooling_ops); - if (IS_ERR(cool_dev)) - goto remove_idr; - /* Fill freq-table in descending order of frequencies */ for (i = 0, freq = -1; i <= cpufreq_dev->max_level; i++) { freq = find_next_max(table, freq); @@ -877,6 +869,14 @@ __cpufreq_cooling_register(struct device_node *np, pr_debug("%s: freq:%u KHz\n", __func__, freq); } + snprintf(dev_name, sizeof(dev_name), "thermal-cpufreq-%d", + cpufreq_dev->id); + + cool_dev = thermal_of_cooling_device_register(np, dev_name, cpufreq_dev, + &cpufreq_cooling_ops); + if (IS_ERR(cool_dev)) + goto remove_idr; + cpufreq_dev->clipped_freq = cpufreq_dev->freq_table[0]; cpufreq_dev->cool_dev = cool_dev; diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c index 6dc810bce295..944a6dca0fcb 100644 --- a/drivers/usb/core/quirks.c +++ b/drivers/usb/core/quirks.c @@ -44,6 +44,9 @@ static const struct usb_device_id usb_quirk_list[] = { /* Creative SB Audigy 2 NX */ { USB_DEVICE(0x041e, 0x3020), .driver_info = USB_QUIRK_RESET_RESUME }, + /* USB3503 */ + { USB_DEVICE(0x0424, 0x3503), .driver_info = USB_QUIRK_RESET_RESUME }, + /* Microsoft Wireless Laser Mouse 6000 Receiver */ { USB_DEVICE(0x045e, 0x00e1), .driver_info = USB_QUIRK_RESET_RESUME }, @@ -173,6 +176,10 @@ static const struct usb_device_id usb_quirk_list[] = { /* MAYA44USB sound device */ { USB_DEVICE(0x0a92, 0x0091), .driver_info = USB_QUIRK_RESET_RESUME }, + /* ASUS Base Station(T100) */ + { USB_DEVICE(0x0b05, 0x17e0), .driver_info = + USB_QUIRK_IGNORE_REMOTE_WAKEUP }, + /* Action Semiconductor flash disk */ { USB_DEVICE(0x10d6, 0x2200), .driver_info = USB_QUIRK_STRING_FETCH_255 }, @@ -188,26 +195,22 @@ static const struct usb_device_id usb_quirk_list[] = { { USB_DEVICE(0x1908, 0x1315), .driver_info = USB_QUIRK_HONOR_BNUMINTERFACES }, - /* INTEL VALUE SSD */ - { USB_DEVICE(0x8086, 0xf1a5), .driver_info = USB_QUIRK_RESET_RESUME }, - - /* USB3503 */ - { USB_DEVICE(0x0424, 0x3503), .driver_info = USB_QUIRK_RESET_RESUME }, - - /* ASUS Base Station(T100) */ - { USB_DEVICE(0x0b05, 0x17e0), .driver_info = - USB_QUIRK_IGNORE_REMOTE_WAKEUP }, - /* Protocol and OTG Electrical Test Device */ { USB_DEVICE(0x1a0a, 0x0200), .driver_info = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL }, + /* Acer C120 LED Projector */ + { USB_DEVICE(0x1de1, 0xc102), .driver_info = USB_QUIRK_NO_LPM }, + /* Blackmagic Design Intensity Shuttle */ { USB_DEVICE(0x1edb, 0xbd3b), .driver_info = USB_QUIRK_NO_LPM }, /* Blackmagic Design UltraStudio SDI */ { USB_DEVICE(0x1edb, 0xbd4f), .driver_info = USB_QUIRK_NO_LPM }, + /* INTEL VALUE SSD */ + { USB_DEVICE(0x8086, 0xf1a5), .driver_info = USB_QUIRK_RESET_RESUME }, + { } /* terminating entry must be last */ }; diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h index 3c58d633ce80..dec0b21fc626 100644 --- a/drivers/usb/dwc2/core.h +++ b/drivers/usb/dwc2/core.h @@ -64,6 +64,17 @@ DWC2_TRACE_SCHEDULER_VB(pr_fmt("%s: SCH: " fmt), \ dev_name(hsotg->dev), ##__VA_ARGS__) +#ifdef CONFIG_MIPS +/* + * There are some MIPS machines that can run in either big-endian + * or little-endian mode and that use the dwc2 register without + * a byteswap in both ways. + * Unlike other architectures, MIPS apparently does not require a + * barrier before the __raw_writel() to synchronize with DMA but does + * require the barrier after the __raw_writel() to serialize a set of + * writes. This set of operations was added specifically for MIPS and + * should only be used there. + */ static inline u32 dwc2_readl(const void __iomem *addr) { u32 value = __raw_readl(addr); @@ -90,6 +101,22 @@ static inline void dwc2_writel(u32 value, void __iomem *addr) pr_info("INFO:: wrote %08x to %p\n", value, addr); #endif } +#else +/* Normal architectures just use readl/write */ +static inline u32 dwc2_readl(const void __iomem *addr) +{ + return readl(addr); +} + +static inline void dwc2_writel(u32 value, void __iomem *addr) +{ + writel(value, addr); + +#ifdef DWC2_LOG_WRITES + pr_info("info:: wrote %08x to %p\n", value, addr); +#endif +} +#endif /* Maximum number of Endpoints/HostChannels */ #define MAX_EPS_CHANNELS 16 diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c index 4c5e3005e1dc..26cf09d0fe3c 100644 --- a/drivers/usb/dwc2/gadget.c +++ b/drivers/usb/dwc2/gadget.c @@ -1018,7 +1018,7 @@ static int dwc2_hsotg_process_req_status(struct dwc2_hsotg *hsotg, return 1; } -static int dwc2_hsotg_ep_sethalt(struct usb_ep *ep, int value); +static int dwc2_hsotg_ep_sethalt(struct usb_ep *ep, int value, bool now); /** * get_ep_head - return the first request on the endpoint @@ -1094,7 +1094,7 @@ static int dwc2_hsotg_process_req_feature(struct dwc2_hsotg *hsotg, case USB_ENDPOINT_HALT: halted = ep->halted; - dwc2_hsotg_ep_sethalt(&ep->ep, set); + dwc2_hsotg_ep_sethalt(&ep->ep, set, true); ret = dwc2_hsotg_send_reply(hsotg, ep0, NULL, 0); if (ret) { @@ -2948,8 +2948,13 @@ static int dwc2_hsotg_ep_dequeue(struct usb_ep *ep, struct usb_request *req) * dwc2_hsotg_ep_sethalt - set halt on a given endpoint * @ep: The endpoint to set halt. * @value: Set or unset the halt. + * @now: If true, stall the endpoint now. Otherwise return -EAGAIN if + * the endpoint is busy processing requests. + * + * We need to stall the endpoint immediately if request comes from set_feature + * protocol command handler. */ -static int dwc2_hsotg_ep_sethalt(struct usb_ep *ep, int value) +static int dwc2_hsotg_ep_sethalt(struct usb_ep *ep, int value, bool now) { struct dwc2_hsotg_ep *hs_ep = our_ep(ep); struct dwc2_hsotg *hs = hs_ep->parent; @@ -2969,6 +2974,17 @@ static int dwc2_hsotg_ep_sethalt(struct usb_ep *ep, int value) return 0; } + if (hs_ep->isochronous) { + dev_err(hs->dev, "%s is Isochronous Endpoint\n", ep->name); + return -EINVAL; + } + + if (!now && value && !list_empty(&hs_ep->queue)) { + dev_dbg(hs->dev, "%s request is pending, cannot halt\n", + ep->name); + return -EAGAIN; + } + if (hs_ep->dir_in) { epreg = DIEPCTL(index); epctl = dwc2_readl(hs->regs + epreg); @@ -3020,7 +3036,7 @@ static int dwc2_hsotg_ep_sethalt_lock(struct usb_ep *ep, int value) int ret = 0; spin_lock_irqsave(&hs->lock, flags); - ret = dwc2_hsotg_ep_sethalt(ep, value); + ret = dwc2_hsotg_ep_sethalt(ep, value, false); spin_unlock_irqrestore(&hs->lock, flags); return ret; diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h index 7ddf9449a063..654050684f4f 100644 --- a/drivers/usb/dwc3/core.h +++ b/drivers/usb/dwc3/core.h @@ -402,6 +402,7 @@ #define DWC3_DEPCMD_GET_RSC_IDX(x) (((x) >> DWC3_DEPCMD_PARAM_SHIFT) & 0x7f) #define DWC3_DEPCMD_STATUS(x) (((x) >> 12) & 0x0F) #define DWC3_DEPCMD_HIPRI_FORCERM (1 << 11) +#define DWC3_DEPCMD_CLEARPENDIN (1 << 11) #define DWC3_DEPCMD_CMDACT (1 << 10) #define DWC3_DEPCMD_CMDIOC (1 << 8) diff --git a/drivers/usb/dwc3/dwc3-exynos.c b/drivers/usb/dwc3/dwc3-exynos.c index dd5cb5577dca..2f1fb7e7aa54 100644 --- a/drivers/usb/dwc3/dwc3-exynos.c +++ b/drivers/usb/dwc3/dwc3-exynos.c @@ -128,12 +128,6 @@ static int dwc3_exynos_probe(struct platform_device *pdev) platform_set_drvdata(pdev, exynos); - ret = dwc3_exynos_register_phys(exynos); - if (ret) { - dev_err(dev, "couldn't register PHYs\n"); - return ret; - } - exynos->dev = dev; exynos->clk = devm_clk_get(dev, "usbdrd30"); @@ -183,20 +177,29 @@ static int dwc3_exynos_probe(struct platform_device *pdev) goto err3; } + ret = dwc3_exynos_register_phys(exynos); + if (ret) { + dev_err(dev, "couldn't register PHYs\n"); + goto err4; + } + if (node) { ret = of_platform_populate(node, NULL, NULL, dev); if (ret) { dev_err(dev, "failed to add dwc3 core\n"); - goto err4; + goto err5; } } else { dev_err(dev, "no device node, failed to add dwc3 core\n"); ret = -ENODEV; - goto err4; + goto err5; } return 0; +err5: + platform_device_unregister(exynos->usb2_phy); + platform_device_unregister(exynos->usb3_phy); err4: regulator_disable(exynos->vdd10); err3: diff --git a/drivers/usb/dwc3/dwc3-st.c b/drivers/usb/dwc3/dwc3-st.c index 5c0adb9c6fb2..50d6ae6f88bc 100644 --- a/drivers/usb/dwc3/dwc3-st.c +++ b/drivers/usb/dwc3/dwc3-st.c @@ -129,12 +129,18 @@ static int st_dwc3_drd_init(struct st_dwc3 *dwc3_data) switch (dwc3_data->dr_mode) { case USB_DR_MODE_PERIPHERAL: - val &= ~(USB3_FORCE_VBUSVALID | USB3_DELAY_VBUSVALID + val &= ~(USB3_DELAY_VBUSVALID | USB3_SEL_FORCE_OPMODE | USB3_FORCE_OPMODE(0x3) | USB3_SEL_FORCE_DPPULLDOWN2 | USB3_FORCE_DPPULLDOWN2 | USB3_SEL_FORCE_DMPULLDOWN2 | USB3_FORCE_DMPULLDOWN2); - val |= USB3_DEVICE_NOT_HOST; + /* + * USB3_PORT2_FORCE_VBUSVALID When '1' and when + * USB3_PORT2_DEVICE_NOT_HOST = 1, forces VBUSVLDEXT2 input + * of the pico PHY to 1. + */ + + val |= USB3_DEVICE_NOT_HOST | USB3_FORCE_VBUSVALID; break; case USB_DR_MODE_HOST: diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c index 9a7d0bd15dc3..07248ff1be5c 100644 --- a/drivers/usb/dwc3/gadget.c +++ b/drivers/usb/dwc3/gadget.c @@ -347,6 +347,28 @@ int dwc3_send_gadget_ep_cmd(struct dwc3 *dwc, unsigned ep, return ret; } +static int dwc3_send_clear_stall_ep_cmd(struct dwc3_ep *dep) +{ + struct dwc3 *dwc = dep->dwc; + struct dwc3_gadget_ep_cmd_params params; + u32 cmd = DWC3_DEPCMD_CLEARSTALL; + + /* + * As of core revision 2.60a the recommended programming model + * is to set the ClearPendIN bit when issuing a Clear Stall EP + * command for IN endpoints. This is to prevent an issue where + * some (non-compliant) hosts may not send ACK TPs for pending + * IN transfers due to a mishandled error condition. Synopsys + * STAR 9000614252. + */ + if (dep->direction && (dwc->revision >= DWC3_REVISION_260A)) + cmd |= DWC3_DEPCMD_CLEARPENDIN; + + memset(¶ms, 0, sizeof(params)); + + return dwc3_send_gadget_ep_cmd(dwc, dep->number, cmd, ¶ms); +} + static dma_addr_t dwc3_trb_dma_offset(struct dwc3_ep *dep, struct dwc3_trb *trb) { @@ -1314,8 +1336,7 @@ int __dwc3_gadget_ep_set_halt(struct dwc3_ep *dep, int value, int protocol) else dep->flags |= DWC3_EP_STALL; } else { - ret = dwc3_send_gadget_ep_cmd(dwc, dep->number, - DWC3_DEPCMD_CLEARSTALL, ¶ms); + ret = dwc3_send_clear_stall_ep_cmd(dep); if (ret) dev_err(dwc->dev, "failed to clear STALL on %s\n", dep->name); @@ -2247,7 +2268,6 @@ static void dwc3_clear_stall_all_ep(struct dwc3 *dwc) for (epnum = 1; epnum < DWC3_ENDPOINTS_NUM; epnum++) { struct dwc3_ep *dep; - struct dwc3_gadget_ep_cmd_params params; int ret; dep = dwc->eps[epnum]; @@ -2259,9 +2279,7 @@ static void dwc3_clear_stall_all_ep(struct dwc3 *dwc) dep->flags &= ~DWC3_EP_STALL; - memset(¶ms, 0, sizeof(params)); - ret = dwc3_send_gadget_ep_cmd(dwc, dep->number, - DWC3_DEPCMD_CLEARSTALL, ¶ms); + ret = dwc3_send_clear_stall_ep_cmd(dep); WARN_ON_ONCE(ret); } } diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c index d67de0d22a2b..eb648485a58c 100644 --- a/drivers/usb/gadget/composite.c +++ b/drivers/usb/gadget/composite.c @@ -1868,14 +1868,19 @@ unknown: } break; } - req->length = value; - req->context = cdev; - req->zero = value < w_length; - value = composite_ep0_queue(cdev, req, GFP_ATOMIC); - if (value < 0) { - DBG(cdev, "ep_queue --> %d\n", value); - req->status = 0; - composite_setup_complete(gadget->ep0, req); + + if (value >= 0) { + req->length = value; + req->context = cdev; + req->zero = value < w_length; + value = composite_ep0_queue(cdev, req, + GFP_ATOMIC); + if (value < 0) { + DBG(cdev, "ep_queue --> %d\n", value); + req->status = 0; + composite_setup_complete(gadget->ep0, + req); + } } return value; } diff --git a/drivers/usb/gadget/configfs.c b/drivers/usb/gadget/configfs.c index b6f60ca8a035..70cf3477f951 100644 --- a/drivers/usb/gadget/configfs.c +++ b/drivers/usb/gadget/configfs.c @@ -1401,6 +1401,7 @@ static const struct usb_gadget_driver configfs_driver_template = { .owner = THIS_MODULE, .name = "configfs-gadget", }, + .match_existing_only = 1, }; static struct config_group *gadgets_make( diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c index 73515d54e1cc..cc33d2667408 100644 --- a/drivers/usb/gadget/function/f_fs.c +++ b/drivers/usb/gadget/function/f_fs.c @@ -2051,7 +2051,7 @@ static int __ffs_data_do_os_desc(enum ffs_os_desc_type type, if (len < sizeof(*d) || d->bFirstInterfaceNumber >= ffs->interfaces_count || - d->Reserved1) + !d->Reserved1) return -EINVAL; for (i = 0; i < ARRAY_SIZE(d->Reserved2); ++i) if (d->Reserved2[i]) @@ -2729,6 +2729,7 @@ static int _ffs_func_bind(struct usb_configuration *c, func->ffs->ss_descs_count; int fs_len, hs_len, ss_len, ret, i; + struct ffs_ep *eps_ptr; /* Make it a single chunk, less management later on */ vla_group(d); @@ -2777,12 +2778,9 @@ static int _ffs_func_bind(struct usb_configuration *c, ffs->raw_descs_length); memset(vla_ptr(vlabuf, d, inums), 0xff, d_inums__sz); - for (ret = ffs->eps_count; ret; --ret) { - struct ffs_ep *ptr; - - ptr = vla_ptr(vlabuf, d, eps); - ptr[ret].num = -1; - } + eps_ptr = vla_ptr(vlabuf, d, eps); + for (i = 0; i < ffs->eps_count; i++) + eps_ptr[i].num = -1; /* Save pointers * d_eps == vlabuf, func->eps used to kfree vlabuf later @@ -2851,7 +2849,7 @@ static int _ffs_func_bind(struct usb_configuration *c, goto error; func->function.os_desc_table = vla_ptr(vlabuf, d, os_desc_table); - if (c->cdev->use_os_string) + if (c->cdev->use_os_string) { for (i = 0; i < ffs->interfaces_count; ++i) { struct usb_os_desc *desc; @@ -2862,13 +2860,15 @@ static int _ffs_func_bind(struct usb_configuration *c, vla_ptr(vlabuf, d, ext_compat) + i * 16; INIT_LIST_HEAD(&desc->ext_prop); } - ret = ffs_do_os_descs(ffs->ms_os_descs_count, - vla_ptr(vlabuf, d, raw_descs) + - fs_len + hs_len + ss_len, - d_raw_descs__sz - fs_len - hs_len - ss_len, - __ffs_func_bind_do_os_desc, func); - if (unlikely(ret < 0)) - goto error; + ret = ffs_do_os_descs(ffs->ms_os_descs_count, + vla_ptr(vlabuf, d, raw_descs) + + fs_len + hs_len + ss_len, + d_raw_descs__sz - fs_len - hs_len - + ss_len, + __ffs_func_bind_do_os_desc, func); + if (unlikely(ret < 0)) + goto error; + } func->function.os_desc_n = c->cdev->use_os_string ? ffs->interfaces_count : 0; diff --git a/drivers/usb/gadget/function/f_printer.c b/drivers/usb/gadget/function/f_printer.c index c45104e3a64b..64706a789580 100644 --- a/drivers/usb/gadget/function/f_printer.c +++ b/drivers/usb/gadget/function/f_printer.c @@ -161,14 +161,6 @@ static struct usb_endpoint_descriptor hs_ep_out_desc = { .wMaxPacketSize = cpu_to_le16(512) }; -static struct usb_qualifier_descriptor dev_qualifier = { - .bLength = sizeof(dev_qualifier), - .bDescriptorType = USB_DT_DEVICE_QUALIFIER, - .bcdUSB = cpu_to_le16(0x0200), - .bDeviceClass = USB_CLASS_PRINTER, - .bNumConfigurations = 1 -}; - static struct usb_descriptor_header *hs_printer_function[] = { (struct usb_descriptor_header *) &intf_desc, (struct usb_descriptor_header *) &hs_ep_in_desc, diff --git a/drivers/usb/gadget/function/f_tcm.c b/drivers/usb/gadget/function/f_tcm.c index 35fe3c80cfc0..197f73386fac 100644 --- a/drivers/usb/gadget/function/f_tcm.c +++ b/drivers/usb/gadget/function/f_tcm.c @@ -1445,16 +1445,18 @@ static void usbg_drop_tpg(struct se_portal_group *se_tpg) for (i = 0; i < TPG_INSTANCES; ++i) if (tpg_instances[i].tpg == tpg) break; - if (i < TPG_INSTANCES) + if (i < TPG_INSTANCES) { tpg_instances[i].tpg = NULL; - opts = container_of(tpg_instances[i].func_inst, - struct f_tcm_opts, func_inst); - mutex_lock(&opts->dep_lock); - if (opts->has_dep) - module_put(opts->dependent); - else - configfs_undepend_item_unlocked(&opts->func_inst.group.cg_item); - mutex_unlock(&opts->dep_lock); + opts = container_of(tpg_instances[i].func_inst, + struct f_tcm_opts, func_inst); + mutex_lock(&opts->dep_lock); + if (opts->has_dep) + module_put(opts->dependent); + else + configfs_undepend_item_unlocked( + &opts->func_inst.group.cg_item); + mutex_unlock(&opts->dep_lock); + } mutex_unlock(&tpg_instances_lock); kfree(tpg); diff --git a/drivers/usb/gadget/function/f_uac2.c b/drivers/usb/gadget/function/f_uac2.c index 186d4b162524..cd214ec8a601 100644 --- a/drivers/usb/gadget/function/f_uac2.c +++ b/drivers/usb/gadget/function/f_uac2.c @@ -598,18 +598,6 @@ static struct usb_gadget_strings *fn_strings[] = { NULL, }; -static struct usb_qualifier_descriptor devqual_desc = { - .bLength = sizeof devqual_desc, - .bDescriptorType = USB_DT_DEVICE_QUALIFIER, - - .bcdUSB = cpu_to_le16(0x200), - .bDeviceClass = USB_CLASS_MISC, - .bDeviceSubClass = 0x02, - .bDeviceProtocol = 0x01, - .bNumConfigurations = 1, - .bRESERVED = 0, -}; - static struct usb_interface_assoc_descriptor iad_desc = { .bLength = sizeof iad_desc, .bDescriptorType = USB_DT_INTERFACE_ASSOCIATION, @@ -1292,6 +1280,7 @@ in_rq_cur(struct usb_function *fn, const struct usb_ctrlrequest *cr) if (control_selector == UAC2_CS_CONTROL_SAM_FREQ) { struct cntrl_cur_lay3 c; + memset(&c, 0, sizeof(struct cntrl_cur_lay3)); if (entity_id == USB_IN_CLK_ID) c.dCUR = p_srate; diff --git a/drivers/usb/gadget/function/storage_common.c b/drivers/usb/gadget/function/storage_common.c index d62683017cf3..990df221c629 100644 --- a/drivers/usb/gadget/function/storage_common.c +++ b/drivers/usb/gadget/function/storage_common.c @@ -83,9 +83,7 @@ EXPORT_SYMBOL_GPL(fsg_fs_function); * USB 2.0 devices need to expose both high speed and full speed * descriptors, unless they only run at full speed. * - * That means alternate endpoint descriptors (bigger packets) - * and a "device qualifier" ... plus more construction options - * for the configuration descriptor. + * That means alternate endpoint descriptors (bigger packets). */ struct usb_endpoint_descriptor fsg_hs_bulk_in_desc = { .bLength = USB_DT_ENDPOINT_SIZE, diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c index e64479f882a5..aa3707bdebb4 100644 --- a/drivers/usb/gadget/legacy/inode.c +++ b/drivers/usb/gadget/legacy/inode.c @@ -938,8 +938,11 @@ ep0_read (struct file *fd, char __user *buf, size_t len, loff_t *ptr) struct usb_ep *ep = dev->gadget->ep0; struct usb_request *req = dev->req; - if ((retval = setup_req (ep, req, 0)) == 0) - retval = usb_ep_queue (ep, req, GFP_ATOMIC); + if ((retval = setup_req (ep, req, 0)) == 0) { + spin_unlock_irq (&dev->lock); + retval = usb_ep_queue (ep, req, GFP_KERNEL); + spin_lock_irq (&dev->lock); + } dev->state = STATE_DEV_CONNECTED; /* assume that was SET_CONFIGURATION */ @@ -1457,8 +1460,11 @@ delegate: w_length); if (value < 0) break; + + spin_unlock (&dev->lock); value = usb_ep_queue (gadget->ep0, dev->req, - GFP_ATOMIC); + GFP_KERNEL); + spin_lock (&dev->lock); if (value < 0) { clean_req (gadget->ep0, dev->req); break; @@ -1481,11 +1487,14 @@ delegate: if (value >= 0 && dev->state != STATE_DEV_SETUP) { req->length = value; req->zero = value < w_length; - value = usb_ep_queue (gadget->ep0, req, GFP_ATOMIC); + + spin_unlock (&dev->lock); + value = usb_ep_queue (gadget->ep0, req, GFP_KERNEL); if (value < 0) { DBG (dev, "ep_queue --> %d\n", value); req->status = 0; } + return value; } /* device stalls when value < 0 */ diff --git a/drivers/usb/gadget/udc/udc-core.c b/drivers/usb/gadget/udc/udc-core.c index 6e8300d6a737..e1b2dcebdc2e 100644 --- a/drivers/usb/gadget/udc/udc-core.c +++ b/drivers/usb/gadget/udc/udc-core.c @@ -603,11 +603,15 @@ int usb_gadget_probe_driver(struct usb_gadget_driver *driver) } } - list_add_tail(&driver->pending, &gadget_driver_pending_list); - pr_info("udc-core: couldn't find an available UDC - added [%s] to list of pending drivers\n", - driver->function); + if (!driver->match_existing_only) { + list_add_tail(&driver->pending, &gadget_driver_pending_list); + pr_info("udc-core: couldn't find an available UDC - added [%s] to list of pending drivers\n", + driver->function); + ret = 0; + } + mutex_unlock(&udc_lock); - return 0; + return ret; found: ret = udc_bind_to_driver(udc, driver); mutex_unlock(&udc_lock); diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c index ae1b6e69eb96..a962b89b65a6 100644 --- a/drivers/usb/host/ehci-hcd.c +++ b/drivers/usb/host/ehci-hcd.c @@ -368,6 +368,15 @@ static void ehci_shutdown(struct usb_hcd *hcd) { struct ehci_hcd *ehci = hcd_to_ehci(hcd); + /** + * Protect the system from crashing at system shutdown in cases where + * usb host is not added yet from OTG controller driver. + * As ehci_setup() not done yet, so stop accessing registers or + * variables initialized in ehci_setup() + */ + if (!ehci->sbrn) + return; + spin_lock_irq(&ehci->lock); ehci->shutdown = true; ehci->rh_state = EHCI_RH_STOPPING; diff --git a/drivers/usb/host/ehci-hub.c b/drivers/usb/host/ehci-hub.c index ffc90295a95f..74f62d68f013 100644 --- a/drivers/usb/host/ehci-hub.c +++ b/drivers/usb/host/ehci-hub.c @@ -872,15 +872,23 @@ int ehci_hub_control( ) { struct ehci_hcd *ehci = hcd_to_ehci (hcd); int ports = HCS_N_PORTS (ehci->hcs_params); - u32 __iomem *status_reg = &ehci->regs->port_status[ - (wIndex & 0xff) - 1]; - u32 __iomem *hostpc_reg = &ehci->regs->hostpc[(wIndex & 0xff) - 1]; + u32 __iomem *status_reg, *hostpc_reg; u32 temp, temp1, status; unsigned long flags; int retval = 0; unsigned selector; /* + * Avoid underflow while calculating (wIndex & 0xff) - 1. + * The compiler might deduce that wIndex can never be 0 and then + * optimize away the tests for !wIndex below. + */ + temp = wIndex & 0xff; + temp -= (temp > 0); + status_reg = &ehci->regs->port_status[temp]; + hostpc_reg = &ehci->regs->hostpc[temp]; + + /* * FIXME: support SetPortFeatures USB_PORT_FEAT_INDICATOR. * HCS_INDICATOR may say we can change LEDs to off/amber/green. * (track current state ourselves) ... blink for diagnostics, diff --git a/drivers/usb/host/ehci-msm.c b/drivers/usb/host/ehci-msm.c index d3afc89d00f5..2f8d3af811ce 100644 --- a/drivers/usb/host/ehci-msm.c +++ b/drivers/usb/host/ehci-msm.c @@ -179,22 +179,32 @@ static int ehci_msm_remove(struct platform_device *pdev) static int ehci_msm_pm_suspend(struct device *dev) { struct usb_hcd *hcd = dev_get_drvdata(dev); + struct ehci_hcd *ehci = hcd_to_ehci(hcd); bool do_wakeup = device_may_wakeup(dev); dev_dbg(dev, "ehci-msm PM suspend\n"); - return ehci_suspend(hcd, do_wakeup); + /* Only call ehci_suspend if ehci_setup has been done */ + if (ehci->sbrn) + return ehci_suspend(hcd, do_wakeup); + + return 0; } static int ehci_msm_pm_resume(struct device *dev) { struct usb_hcd *hcd = dev_get_drvdata(dev); + struct ehci_hcd *ehci = hcd_to_ehci(hcd); dev_dbg(dev, "ehci-msm PM resume\n"); - ehci_resume(hcd, false); + + /* Only call ehci_resume if ehci_setup has been done */ + if (ehci->sbrn) + ehci_resume(hcd, false); return 0; } + #else #define ehci_msm_pm_suspend NULL #define ehci_msm_pm_resume NULL diff --git a/drivers/usb/host/ehci-tegra.c b/drivers/usb/host/ehci-tegra.c index 4031b372008e..9a3d7db5be57 100644 --- a/drivers/usb/host/ehci-tegra.c +++ b/drivers/usb/host/ehci-tegra.c @@ -81,15 +81,23 @@ static int tegra_reset_usb_controller(struct platform_device *pdev) struct usb_hcd *hcd = platform_get_drvdata(pdev); struct tegra_ehci_hcd *tegra = (struct tegra_ehci_hcd *)hcd_to_ehci(hcd)->priv; + bool has_utmi_pad_registers = false; phy_np = of_parse_phandle(pdev->dev.of_node, "nvidia,phy", 0); if (!phy_np) return -ENOENT; + if (of_property_read_bool(phy_np, "nvidia,has-utmi-pad-registers")) + has_utmi_pad_registers = true; + if (!usb1_reset_attempted) { struct reset_control *usb1_reset; - usb1_reset = of_reset_control_get(phy_np, "usb"); + if (!has_utmi_pad_registers) + usb1_reset = of_reset_control_get(phy_np, "utmi-pads"); + else + usb1_reset = tegra->rst; + if (IS_ERR(usb1_reset)) { dev_warn(&pdev->dev, "can't get utmi-pads reset from the PHY\n"); @@ -99,13 +107,15 @@ static int tegra_reset_usb_controller(struct platform_device *pdev) reset_control_assert(usb1_reset); udelay(1); reset_control_deassert(usb1_reset); + + if (!has_utmi_pad_registers) + reset_control_put(usb1_reset); } - reset_control_put(usb1_reset); usb1_reset_attempted = true; } - if (!of_property_read_bool(phy_np, "nvidia,has-utmi-pad-registers")) { + if (!has_utmi_pad_registers) { reset_control_assert(tegra->rst); udelay(1); reset_control_deassert(tegra->rst); diff --git a/drivers/usb/host/ohci-q.c b/drivers/usb/host/ohci-q.c index d029bbe9eb36..641fed609911 100644 --- a/drivers/usb/host/ohci-q.c +++ b/drivers/usb/host/ohci-q.c @@ -183,7 +183,6 @@ static int ed_schedule (struct ohci_hcd *ohci, struct ed *ed) { int branch; - ed->state = ED_OPER; ed->ed_prev = NULL; ed->ed_next = NULL; ed->hwNextED = 0; @@ -259,6 +258,8 @@ static int ed_schedule (struct ohci_hcd *ohci, struct ed *ed) /* the HC may not see the schedule updates yet, but if it does * then they'll be properly ordered. */ + + ed->state = ED_OPER; return 0; } diff --git a/drivers/usb/host/xhci-pci.c b/drivers/usb/host/xhci-pci.c index 48672fac7ff3..c10972fcc8e4 100644 --- a/drivers/usb/host/xhci-pci.c +++ b/drivers/usb/host/xhci-pci.c @@ -37,6 +37,7 @@ /* Device for a quirk */ #define PCI_VENDOR_ID_FRESCO_LOGIC 0x1b73 #define PCI_DEVICE_ID_FRESCO_LOGIC_PDK 0x1000 +#define PCI_DEVICE_ID_FRESCO_LOGIC_FL1009 0x1009 #define PCI_DEVICE_ID_FRESCO_LOGIC_FL1400 0x1400 #define PCI_VENDOR_ID_ETRON 0x1b6f @@ -114,6 +115,10 @@ static void xhci_pci_quirks(struct device *dev, struct xhci_hcd *xhci) xhci->quirks |= XHCI_TRUST_TX_LENGTH; } + if (pdev->vendor == PCI_VENDOR_ID_FRESCO_LOGIC && + pdev->device == PCI_DEVICE_ID_FRESCO_LOGIC_FL1009) + xhci->quirks |= XHCI_BROKEN_STREAMS; + if (pdev->vendor == PCI_VENDOR_ID_NEC) xhci->quirks |= XHCI_NEC_HOST; diff --git a/drivers/usb/host/xhci-plat.c b/drivers/usb/host/xhci-plat.c index 676ea458148b..1f3f981fe7f8 100644 --- a/drivers/usb/host/xhci-plat.c +++ b/drivers/usb/host/xhci-plat.c @@ -196,6 +196,9 @@ static int xhci_plat_probe(struct platform_device *pdev) ret = clk_prepare_enable(clk); if (ret) goto put_hcd; + } else if (PTR_ERR(clk) == -EPROBE_DEFER) { + ret = -EPROBE_DEFER; + goto put_hcd; } xhci = hcd_to_xhci(hcd); diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 52deae4b7eac..d7d502578d79 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -290,6 +290,14 @@ static int xhci_abort_cmd_ring(struct xhci_hcd *xhci) temp_64 = xhci_read_64(xhci, &xhci->op_regs->cmd_ring); xhci->cmd_ring_state = CMD_RING_STATE_ABORTED; + + /* + * Writing the CMD_RING_ABORT bit should cause a cmd completion event, + * however on some host hw the CMD_RING_RUNNING bit is correctly cleared + * but the completion event in never sent. Use the cmd timeout timer to + * handle those cases. Use twice the time to cover the bit polling retry + */ + mod_timer(&xhci->cmd_timer, jiffies + (2 * XHCI_CMD_DEFAULT_TIMEOUT)); xhci_write_64(xhci, temp_64 | CMD_RING_ABORT, &xhci->op_regs->cmd_ring); @@ -314,6 +322,7 @@ static int xhci_abort_cmd_ring(struct xhci_hcd *xhci) xhci_err(xhci, "Stopped the command ring failed, " "maybe the host is dead\n"); + del_timer(&xhci->cmd_timer); xhci->xhc_state |= XHCI_STATE_DYING; xhci_quiesce(xhci); xhci_halt(xhci); @@ -1246,22 +1255,21 @@ void xhci_handle_command_timeout(unsigned long data) int ret; unsigned long flags; u64 hw_ring_state; - struct xhci_command *cur_cmd = NULL; + bool second_timeout = false; xhci = (struct xhci_hcd *) data; /* mark this command to be cancelled */ spin_lock_irqsave(&xhci->lock, flags); if (xhci->current_cmd) { - cur_cmd = xhci->current_cmd; - cur_cmd->status = COMP_CMD_ABORT; + if (xhci->current_cmd->status == COMP_CMD_ABORT) + second_timeout = true; + xhci->current_cmd->status = COMP_CMD_ABORT; } - /* Make sure command ring is running before aborting it */ hw_ring_state = xhci_read_64(xhci, &xhci->op_regs->cmd_ring); if ((xhci->cmd_ring_state & CMD_RING_STATE_RUNNING) && (hw_ring_state & CMD_RING_RUNNING)) { - spin_unlock_irqrestore(&xhci->lock, flags); xhci_dbg(xhci, "Command timeout\n"); ret = xhci_abort_cmd_ring(xhci); @@ -1273,6 +1281,15 @@ void xhci_handle_command_timeout(unsigned long data) } return; } + + /* command ring failed to restart, or host removed. Bail out */ + if (second_timeout || xhci->xhc_state & XHCI_STATE_REMOVING) { + spin_unlock_irqrestore(&xhci->lock, flags); + xhci_dbg(xhci, "command timed out twice, ring start fail?\n"); + xhci_cleanup_command_queue(xhci); + return; + } + /* command timeout on stopped ring, ring can't be aborted */ xhci_dbg(xhci, "Command timeout on stopped ring\n"); xhci_handle_stopped_cmd_ring(xhci, xhci->current_cmd); @@ -2721,7 +2738,8 @@ hw_died: writel(irq_pending, &xhci->ir_set->irq_pending); } - if (xhci->xhc_state & XHCI_STATE_DYING) { + if (xhci->xhc_state & XHCI_STATE_DYING || + xhci->xhc_state & XHCI_STATE_HALTED) { xhci_dbg(xhci, "xHCI dying, ignoring interrupt. " "Shouldn't IRQs be disabled?\n"); /* Clear the event handler busy flag (RW1C); diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index fa7e1ef36cd9..f2f9518c53ab 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -685,20 +685,23 @@ void xhci_stop(struct usb_hcd *hcd) u32 temp; struct xhci_hcd *xhci = hcd_to_xhci(hcd); - if (xhci->xhc_state & XHCI_STATE_HALTED) - return; - mutex_lock(&xhci->mutex); - spin_lock_irq(&xhci->lock); - xhci->xhc_state |= XHCI_STATE_HALTED; - xhci->cmd_ring_state = CMD_RING_STATE_STOPPED; - /* Make sure the xHC is halted for a USB3 roothub - * (xhci_stop() could be called as part of failed init). - */ - xhci_halt(xhci); - xhci_reset(xhci); - spin_unlock_irq(&xhci->lock); + if (!(xhci->xhc_state & XHCI_STATE_HALTED)) { + spin_lock_irq(&xhci->lock); + + xhci->xhc_state |= XHCI_STATE_HALTED; + xhci->cmd_ring_state = CMD_RING_STATE_STOPPED; + xhci_halt(xhci); + xhci_reset(xhci); + + spin_unlock_irq(&xhci->lock); + } + + if (!usb_hcd_is_primary_hcd(hcd)) { + mutex_unlock(&xhci->mutex); + return; + } xhci_cleanup_msix(xhci); @@ -4886,7 +4889,7 @@ int xhci_gen_setup(struct usb_hcd *hcd, xhci_get_quirks_t get_quirks) xhci->hcc_params2 = readl(&xhci->cap_regs->hcc_params2); xhci_print_registers(xhci); - xhci->quirks = quirks; + xhci->quirks |= quirks; get_quirks(dev, xhci); diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c index 39fd95833eb8..f824336def5c 100644 --- a/drivers/usb/musb/musb_core.c +++ b/drivers/usb/musb/musb_core.c @@ -1090,29 +1090,6 @@ void musb_stop(struct musb *musb) musb_platform_try_idle(musb, 0); } -static void musb_shutdown(struct platform_device *pdev) -{ - struct musb *musb = dev_to_musb(&pdev->dev); - unsigned long flags; - - pm_runtime_get_sync(musb->controller); - - musb_host_cleanup(musb); - musb_gadget_cleanup(musb); - - spin_lock_irqsave(&musb->lock, flags); - musb_platform_disable(musb); - musb_generic_disable(musb); - spin_unlock_irqrestore(&musb->lock, flags); - - musb_writeb(musb->mregs, MUSB_DEVCTL, 0); - musb_platform_exit(musb); - - pm_runtime_put(musb->controller); - /* FIXME power down */ -} - - /*-------------------------------------------------------------------------*/ /* @@ -1702,7 +1679,7 @@ EXPORT_SYMBOL_GPL(musb_dma_completion); #define use_dma 0 #endif -static void (*musb_phy_callback)(enum musb_vbus_id_status status); +static int (*musb_phy_callback)(enum musb_vbus_id_status status); /* * musb_mailbox - optional phy notifier function @@ -1711,11 +1688,12 @@ static void (*musb_phy_callback)(enum musb_vbus_id_status status); * Optionally gets called from the USB PHY. Note that the USB PHY must be * disabled at the point the phy_callback is registered or unregistered. */ -void musb_mailbox(enum musb_vbus_id_status status) +int musb_mailbox(enum musb_vbus_id_status status) { if (musb_phy_callback) - musb_phy_callback(status); + return musb_phy_callback(status); + return -ENODEV; }; EXPORT_SYMBOL_GPL(musb_mailbox); @@ -2028,11 +2006,6 @@ musb_init_controller(struct device *dev, int nIrq, void __iomem *ctrl) musb_readl = musb_default_readl; musb_writel = musb_default_writel; - /* We need musb_read/write functions initialized for PM */ - pm_runtime_use_autosuspend(musb->controller); - pm_runtime_set_autosuspend_delay(musb->controller, 200); - pm_runtime_enable(musb->controller); - /* The musb_platform_init() call: * - adjusts musb->mregs * - sets the musb->isr @@ -2134,6 +2107,16 @@ musb_init_controller(struct device *dev, int nIrq, void __iomem *ctrl) if (musb->ops->phy_callback) musb_phy_callback = musb->ops->phy_callback; + /* + * We need musb_read/write functions initialized for PM. + * Note that at least 2430 glue needs autosuspend delay + * somewhere above 300 ms for the hardware to idle properly + * after disconnecting the cable in host mode. Let's use + * 500 ms for some margin. + */ + pm_runtime_use_autosuspend(musb->controller); + pm_runtime_set_autosuspend_delay(musb->controller, 500); + pm_runtime_enable(musb->controller); pm_runtime_get_sync(musb->controller); status = usb_phy_init(musb->xceiv); @@ -2237,13 +2220,8 @@ musb_init_controller(struct device *dev, int nIrq, void __iomem *ctrl) if (status) goto fail5; - pm_runtime_put(musb->controller); - - /* - * For why this is currently needed, see commit 3e43a0725637 - * ("usb: musb: core: add pm_runtime_irq_safe()") - */ - pm_runtime_irq_safe(musb->controller); + pm_runtime_mark_last_busy(musb->controller); + pm_runtime_put_autosuspend(musb->controller); return 0; @@ -2265,7 +2243,9 @@ fail2_5: usb_phy_shutdown(musb->xceiv); err_usb_phy_init: + pm_runtime_dont_use_autosuspend(musb->controller); pm_runtime_put_sync(musb->controller); + pm_runtime_disable(musb->controller); fail2: if (musb->irq_wake) @@ -2273,7 +2253,6 @@ fail2: musb_platform_exit(musb); fail1: - pm_runtime_disable(musb->controller); dev_err(musb->controller, "musb_init_controller failed with status %d\n", status); @@ -2312,6 +2291,7 @@ static int musb_remove(struct platform_device *pdev) { struct device *dev = &pdev->dev; struct musb *musb = dev_to_musb(dev); + unsigned long flags; /* this gets called on rmmod. * - Host mode: host may still be active @@ -2319,17 +2299,26 @@ static int musb_remove(struct platform_device *pdev) * - OTG mode: both roles are deactivated (or never-activated) */ musb_exit_debugfs(musb); - musb_shutdown(pdev); - musb_phy_callback = NULL; - - if (musb->dma_controller) - musb_dma_controller_destroy(musb->dma_controller); - - usb_phy_shutdown(musb->xceiv); cancel_work_sync(&musb->irq_work); cancel_delayed_work_sync(&musb->finish_resume_work); cancel_delayed_work_sync(&musb->deassert_reset_work); + pm_runtime_get_sync(musb->controller); + musb_host_cleanup(musb); + musb_gadget_cleanup(musb); + spin_lock_irqsave(&musb->lock, flags); + musb_platform_disable(musb); + musb_generic_disable(musb); + spin_unlock_irqrestore(&musb->lock, flags); + musb_writeb(musb->mregs, MUSB_DEVCTL, 0); + pm_runtime_dont_use_autosuspend(musb->controller); + pm_runtime_put_sync(musb->controller); + pm_runtime_disable(musb->controller); + musb_platform_exit(musb); + musb_phy_callback = NULL; + if (musb->dma_controller) + musb_dma_controller_destroy(musb->dma_controller); + usb_phy_shutdown(musb->xceiv); musb_free(musb); device_init_wakeup(dev, 0); return 0; @@ -2429,7 +2418,8 @@ static void musb_restore_context(struct musb *musb) musb_writew(musb_base, MUSB_INTRTXE, musb->intrtxe); musb_writew(musb_base, MUSB_INTRRXE, musb->intrrxe); musb_writeb(musb_base, MUSB_INTRUSBE, musb->context.intrusbe); - musb_writeb(musb_base, MUSB_DEVCTL, musb->context.devctl); + if (musb->context.devctl & MUSB_DEVCTL_SESSION) + musb_writeb(musb_base, MUSB_DEVCTL, musb->context.devctl); for (i = 0; i < musb->config->num_eps; ++i) { struct musb_hw_ep *hw_ep; @@ -2612,7 +2602,6 @@ static struct platform_driver musb_driver = { }, .probe = musb_probe, .remove = musb_remove, - .shutdown = musb_shutdown, }; module_platform_driver(musb_driver); diff --git a/drivers/usb/musb/musb_core.h b/drivers/usb/musb/musb_core.h index b6afe9e43305..b55a776b03eb 100644 --- a/drivers/usb/musb/musb_core.h +++ b/drivers/usb/musb/musb_core.h @@ -215,7 +215,7 @@ struct musb_platform_ops { dma_addr_t *dma_addr, u32 *len); void (*pre_root_reset_end)(struct musb *musb); void (*post_root_reset_end)(struct musb *musb); - void (*phy_callback)(enum musb_vbus_id_status status); + int (*phy_callback)(enum musb_vbus_id_status status); }; /* @@ -312,6 +312,7 @@ struct musb { struct work_struct irq_work; struct delayed_work deassert_reset_work; struct delayed_work finish_resume_work; + struct delayed_work gadget_work; u16 hwvers; u16 intrrxe; diff --git a/drivers/usb/musb/musb_gadget.c b/drivers/usb/musb/musb_gadget.c index 152865b36522..af2a3a7addf9 100644 --- a/drivers/usb/musb/musb_gadget.c +++ b/drivers/usb/musb/musb_gadget.c @@ -1656,6 +1656,20 @@ static int musb_gadget_vbus_draw(struct usb_gadget *gadget, unsigned mA) return usb_phy_set_power(musb->xceiv, mA); } +static void musb_gadget_work(struct work_struct *work) +{ + struct musb *musb; + unsigned long flags; + + musb = container_of(work, struct musb, gadget_work.work); + pm_runtime_get_sync(musb->controller); + spin_lock_irqsave(&musb->lock, flags); + musb_pullup(musb, musb->softconnect); + spin_unlock_irqrestore(&musb->lock, flags); + pm_runtime_mark_last_busy(musb->controller); + pm_runtime_put_autosuspend(musb->controller); +} + static int musb_gadget_pullup(struct usb_gadget *gadget, int is_on) { struct musb *musb = gadget_to_musb(gadget); @@ -1663,20 +1677,16 @@ static int musb_gadget_pullup(struct usb_gadget *gadget, int is_on) is_on = !!is_on; - pm_runtime_get_sync(musb->controller); - /* NOTE: this assumes we are sensing vbus; we'd rather * not pullup unless the B-session is active. */ spin_lock_irqsave(&musb->lock, flags); if (is_on != musb->softconnect) { musb->softconnect = is_on; - musb_pullup(musb, is_on); + schedule_delayed_work(&musb->gadget_work, 0); } spin_unlock_irqrestore(&musb->lock, flags); - pm_runtime_put(musb->controller); - return 0; } @@ -1845,7 +1855,7 @@ int musb_gadget_setup(struct musb *musb) #elif IS_ENABLED(CONFIG_USB_MUSB_GADGET) musb->g.is_otg = 0; #endif - + INIT_DELAYED_WORK(&musb->gadget_work, musb_gadget_work); musb_g_init_endpoints(musb); musb->is_active = 0; @@ -1866,6 +1876,8 @@ void musb_gadget_cleanup(struct musb *musb) { if (musb->port_mode == MUSB_PORT_MODE_HOST) return; + + cancel_delayed_work_sync(&musb->gadget_work); usb_del_gadget_udc(&musb->g); } @@ -1914,8 +1926,8 @@ static int musb_gadget_start(struct usb_gadget *g, if (musb->xceiv->last_event == USB_EVENT_ID) musb_platform_set_vbus(musb, 1); - if (musb->xceiv->last_event == USB_EVENT_NONE) - pm_runtime_put(musb->controller); + pm_runtime_mark_last_busy(musb->controller); + pm_runtime_put_autosuspend(musb->controller); return 0; @@ -1934,8 +1946,7 @@ static int musb_gadget_stop(struct usb_gadget *g) struct musb *musb = gadget_to_musb(g); unsigned long flags; - if (musb->xceiv->last_event == USB_EVENT_NONE) - pm_runtime_get_sync(musb->controller); + pm_runtime_get_sync(musb->controller); /* * REVISIT always use otg_set_peripheral() here too; @@ -1963,7 +1974,8 @@ static int musb_gadget_stop(struct usb_gadget *g) * that currently misbehaves. */ - pm_runtime_put(musb->controller); + pm_runtime_mark_last_busy(musb->controller); + pm_runtime_put_autosuspend(musb->controller); return 0; } diff --git a/drivers/usb/musb/musb_host.c b/drivers/usb/musb/musb_host.c index 2f8ad7f1f482..d227a71d85e1 100644 --- a/drivers/usb/musb/musb_host.c +++ b/drivers/usb/musb/musb_host.c @@ -434,7 +434,13 @@ static void musb_advance_schedule(struct musb *musb, struct urb *urb, } } - if (qh != NULL && qh->is_ready) { + /* + * The pipe must be broken if current urb->status is set, so don't + * start next urb. + * TODO: to minimize the risk of regression, only check urb->status + * for RX, until we have a test case to understand the behavior of TX. + */ + if ((!status || !is_in) && qh && qh->is_ready) { dev_dbg(musb->controller, "... next ep%d %cX urb %p\n", hw_ep->epnum, is_in ? 'R' : 'T', next_urb(qh)); musb_start_urb(musb, is_in, qh); @@ -594,14 +600,13 @@ musb_rx_reinit(struct musb *musb, struct musb_qh *qh, u8 epnum) musb_writew(ep->regs, MUSB_TXCSR, 0); /* scrub all previous state, clearing toggle */ - } else { - csr = musb_readw(ep->regs, MUSB_RXCSR); - if (csr & MUSB_RXCSR_RXPKTRDY) - WARNING("rx%d, packet/%d ready?\n", ep->epnum, - musb_readw(ep->regs, MUSB_RXCOUNT)); - - musb_h_flush_rxfifo(ep, MUSB_RXCSR_CLRDATATOG); } + csr = musb_readw(ep->regs, MUSB_RXCSR); + if (csr & MUSB_RXCSR_RXPKTRDY) + WARNING("rx%d, packet/%d ready?\n", ep->epnum, + musb_readw(ep->regs, MUSB_RXCOUNT)); + + musb_h_flush_rxfifo(ep, MUSB_RXCSR_CLRDATATOG); /* target addr and (for multipoint) hub addr/port */ if (musb->is_multipoint) { @@ -627,7 +632,7 @@ musb_rx_reinit(struct musb *musb, struct musb_qh *qh, u8 epnum) ep->rx_reinit = 0; } -static int musb_tx_dma_set_mode_mentor(struct dma_controller *dma, +static void musb_tx_dma_set_mode_mentor(struct dma_controller *dma, struct musb_hw_ep *hw_ep, struct musb_qh *qh, struct urb *urb, u32 offset, u32 *length, u8 *mode) @@ -664,23 +669,18 @@ static int musb_tx_dma_set_mode_mentor(struct dma_controller *dma, } channel->desired_mode = *mode; musb_writew(epio, MUSB_TXCSR, csr); - - return 0; } -static int musb_tx_dma_set_mode_cppi_tusb(struct dma_controller *dma, - struct musb_hw_ep *hw_ep, - struct musb_qh *qh, - struct urb *urb, - u32 offset, - u32 *length, - u8 *mode) +static void musb_tx_dma_set_mode_cppi_tusb(struct dma_controller *dma, + struct musb_hw_ep *hw_ep, + struct musb_qh *qh, + struct urb *urb, + u32 offset, + u32 *length, + u8 *mode) { struct dma_channel *channel = hw_ep->tx_channel; - if (!is_cppi_enabled(hw_ep->musb) && !tusb_dma_omap(hw_ep->musb)) - return -ENODEV; - channel->actual_len = 0; /* @@ -688,8 +688,6 @@ static int musb_tx_dma_set_mode_cppi_tusb(struct dma_controller *dma, * to identify the zero-length-final-packet case. */ *mode = (urb->transfer_flags & URB_ZERO_PACKET) ? 1 : 0; - - return 0; } static bool musb_tx_dma_program(struct dma_controller *dma, @@ -699,15 +697,14 @@ static bool musb_tx_dma_program(struct dma_controller *dma, struct dma_channel *channel = hw_ep->tx_channel; u16 pkt_size = qh->maxpacket; u8 mode; - int res; if (musb_dma_inventra(hw_ep->musb) || musb_dma_ux500(hw_ep->musb)) - res = musb_tx_dma_set_mode_mentor(dma, hw_ep, qh, urb, - offset, &length, &mode); + musb_tx_dma_set_mode_mentor(dma, hw_ep, qh, urb, offset, + &length, &mode); + else if (is_cppi_enabled(hw_ep->musb) || tusb_dma_omap(hw_ep->musb)) + musb_tx_dma_set_mode_cppi_tusb(dma, hw_ep, qh, urb, offset, + &length, &mode); else - res = musb_tx_dma_set_mode_cppi_tusb(dma, hw_ep, qh, urb, - offset, &length, &mode); - if (res) return false; qh->segsize = length; @@ -995,9 +992,15 @@ static void musb_bulk_nak_timeout(struct musb *musb, struct musb_hw_ep *ep, if (is_in) { dma = is_dma_capable() ? ep->rx_channel : NULL; - /* clear nak timeout bit */ + /* + * Need to stop the transaction by clearing REQPKT first + * then the NAK Timeout bit ref MUSBMHDRC USB 2.0 HIGH-SPEED + * DUAL-ROLE CONTROLLER Programmer's Guide, section 9.2.2 + */ rx_csr = musb_readw(epio, MUSB_RXCSR); rx_csr |= MUSB_RXCSR_H_WZC_BITS; + rx_csr &= ~MUSB_RXCSR_H_REQPKT; + musb_writew(epio, MUSB_RXCSR, rx_csr); rx_csr &= ~MUSB_RXCSR_DATAERROR; musb_writew(epio, MUSB_RXCSR, rx_csr); @@ -1551,7 +1554,7 @@ static int musb_rx_dma_iso_cppi41(struct dma_controller *dma, struct urb *urb, size_t len) { - struct dma_channel *channel = hw_ep->tx_channel; + struct dma_channel *channel = hw_ep->rx_channel; void __iomem *epio = hw_ep->regs; dma_addr_t *buf; u32 length, res; @@ -1870,6 +1873,9 @@ void musb_host_rx(struct musb *musb, u8 epnum) status = -EPROTO; musb_writeb(epio, MUSB_RXINTERVAL, 0); + rx_csr &= ~MUSB_RXCSR_H_ERROR; + musb_writew(epio, MUSB_RXCSR, rx_csr); + } else if (rx_csr & MUSB_RXCSR_DATAERROR) { if (USB_ENDPOINT_XFER_ISOC != qh->type) { diff --git a/drivers/usb/musb/omap2430.c b/drivers/usb/musb/omap2430.c index c84e0322c108..0b4cec940386 100644 --- a/drivers/usb/musb/omap2430.c +++ b/drivers/usb/musb/omap2430.c @@ -49,97 +49,14 @@ struct omap2430_glue { enum musb_vbus_id_status status; struct work_struct omap_musb_mailbox_work; struct device *control_otghs; + bool cable_connected; + bool enabled; + bool powered; }; #define glue_to_musb(g) platform_get_drvdata(g->musb) static struct omap2430_glue *_glue; -static struct timer_list musb_idle_timer; - -static void musb_do_idle(unsigned long _musb) -{ - struct musb *musb = (void *)_musb; - unsigned long flags; - u8 power; - u8 devctl; - - spin_lock_irqsave(&musb->lock, flags); - - switch (musb->xceiv->otg->state) { - case OTG_STATE_A_WAIT_BCON: - - devctl = musb_readb(musb->mregs, MUSB_DEVCTL); - if (devctl & MUSB_DEVCTL_BDEVICE) { - musb->xceiv->otg->state = OTG_STATE_B_IDLE; - MUSB_DEV_MODE(musb); - } else { - musb->xceiv->otg->state = OTG_STATE_A_IDLE; - MUSB_HST_MODE(musb); - } - break; - case OTG_STATE_A_SUSPEND: - /* finish RESUME signaling? */ - if (musb->port1_status & MUSB_PORT_STAT_RESUME) { - power = musb_readb(musb->mregs, MUSB_POWER); - power &= ~MUSB_POWER_RESUME; - dev_dbg(musb->controller, "root port resume stopped, power %02x\n", power); - musb_writeb(musb->mregs, MUSB_POWER, power); - musb->is_active = 1; - musb->port1_status &= ~(USB_PORT_STAT_SUSPEND - | MUSB_PORT_STAT_RESUME); - musb->port1_status |= USB_PORT_STAT_C_SUSPEND << 16; - usb_hcd_poll_rh_status(musb->hcd); - /* NOTE: it might really be A_WAIT_BCON ... */ - musb->xceiv->otg->state = OTG_STATE_A_HOST; - } - break; - case OTG_STATE_A_HOST: - devctl = musb_readb(musb->mregs, MUSB_DEVCTL); - if (devctl & MUSB_DEVCTL_BDEVICE) - musb->xceiv->otg->state = OTG_STATE_B_IDLE; - else - musb->xceiv->otg->state = OTG_STATE_A_WAIT_BCON; - default: - break; - } - spin_unlock_irqrestore(&musb->lock, flags); -} - - -static void omap2430_musb_try_idle(struct musb *musb, unsigned long timeout) -{ - unsigned long default_timeout = jiffies + msecs_to_jiffies(3); - static unsigned long last_timer; - - if (timeout == 0) - timeout = default_timeout; - - /* Never idle if active, or when VBUS timeout is not set as host */ - if (musb->is_active || ((musb->a_wait_bcon == 0) - && (musb->xceiv->otg->state == OTG_STATE_A_WAIT_BCON))) { - dev_dbg(musb->controller, "%s active, deleting timer\n", - usb_otg_state_string(musb->xceiv->otg->state)); - del_timer(&musb_idle_timer); - last_timer = jiffies; - return; - } - - if (time_after(last_timer, timeout)) { - if (!timer_pending(&musb_idle_timer)) - last_timer = timeout; - else { - dev_dbg(musb->controller, "Longer idle timer already pending, ignoring\n"); - return; - } - } - last_timer = timeout; - - dev_dbg(musb->controller, "%s inactive, for idle timer for %lu ms\n", - usb_otg_state_string(musb->xceiv->otg->state), - (unsigned long)jiffies_to_msecs(timeout - jiffies)); - mod_timer(&musb_idle_timer, timeout); -} - static void omap2430_musb_set_vbus(struct musb *musb, int is_on) { struct usb_otg *otg = musb->xceiv->otg; @@ -205,16 +122,6 @@ static void omap2430_musb_set_vbus(struct musb *musb, int is_on) musb_readb(musb->mregs, MUSB_DEVCTL)); } -static int omap2430_musb_set_mode(struct musb *musb, u8 musb_mode) -{ - u8 devctl = musb_readb(musb->mregs, MUSB_DEVCTL); - - devctl |= MUSB_DEVCTL_SESSION; - musb_writeb(musb->mregs, MUSB_DEVCTL, devctl); - - return 0; -} - static inline void omap2430_low_level_exit(struct musb *musb) { u32 l; @@ -234,22 +141,63 @@ static inline void omap2430_low_level_init(struct musb *musb) musb_writel(musb->mregs, OTG_FORCESTDBY, l); } -static void omap2430_musb_mailbox(enum musb_vbus_id_status status) +/* + * We can get multiple cable events so we need to keep track + * of the power state. Only keep power enabled if USB cable is + * connected and a gadget is started. + */ +static void omap2430_set_power(struct musb *musb, bool enabled, bool cable) +{ + struct device *dev = musb->controller; + struct omap2430_glue *glue = dev_get_drvdata(dev->parent); + bool power_up; + int res; + + if (glue->enabled != enabled) + glue->enabled = enabled; + + if (glue->cable_connected != cable) + glue->cable_connected = cable; + + power_up = glue->enabled && glue->cable_connected; + if (power_up == glue->powered) { + dev_warn(musb->controller, "power state already %i\n", + power_up); + return; + } + + glue->powered = power_up; + + if (power_up) { + res = pm_runtime_get_sync(musb->controller); + if (res < 0) { + dev_err(musb->controller, "could not enable: %i", res); + glue->powered = false; + } + } else { + pm_runtime_mark_last_busy(musb->controller); + pm_runtime_put_autosuspend(musb->controller); + } +} + +static int omap2430_musb_mailbox(enum musb_vbus_id_status status) { struct omap2430_glue *glue = _glue; if (!glue) { pr_err("%s: musb core is not yet initialized\n", __func__); - return; + return -EPROBE_DEFER; } glue->status = status; if (!glue_to_musb(glue)) { pr_err("%s: musb core is not yet ready\n", __func__); - return; + return -EPROBE_DEFER; } schedule_work(&glue->omap_musb_mailbox_work); + + return 0; } static void omap_musb_set_mailbox(struct omap2430_glue *glue) @@ -259,6 +207,13 @@ static void omap_musb_set_mailbox(struct omap2430_glue *glue) struct musb_hdrc_platform_data *pdata = dev_get_platdata(dev); struct omap_musb_board_data *data = pdata->board_data; struct usb_otg *otg = musb->xceiv->otg; + bool cable_connected; + + cable_connected = ((glue->status == MUSB_ID_GROUND) || + (glue->status == MUSB_VBUS_VALID)); + + if (cable_connected) + omap2430_set_power(musb, glue->enabled, cable_connected); switch (glue->status) { case MUSB_ID_GROUND: @@ -268,7 +223,6 @@ static void omap_musb_set_mailbox(struct omap2430_glue *glue) musb->xceiv->otg->state = OTG_STATE_A_IDLE; musb->xceiv->last_event = USB_EVENT_ID; if (musb->gadget_driver) { - pm_runtime_get_sync(dev); omap_control_usb_set_mode(glue->control_otghs, USB_MODE_HOST); omap2430_musb_set_vbus(musb, 1); @@ -281,8 +235,6 @@ static void omap_musb_set_mailbox(struct omap2430_glue *glue) otg->default_a = false; musb->xceiv->otg->state = OTG_STATE_B_IDLE; musb->xceiv->last_event = USB_EVENT_VBUS; - if (musb->gadget_driver) - pm_runtime_get_sync(dev); omap_control_usb_set_mode(glue->control_otghs, USB_MODE_DEVICE); break; @@ -291,11 +243,8 @@ static void omap_musb_set_mailbox(struct omap2430_glue *glue) dev_dbg(dev, "VBUS Disconnect\n"); musb->xceiv->last_event = USB_EVENT_NONE; - if (musb->gadget_driver) { + if (musb->gadget_driver) omap2430_musb_set_vbus(musb, 0); - pm_runtime_mark_last_busy(dev); - pm_runtime_put_autosuspend(dev); - } if (data->interface_type == MUSB_INTERFACE_UTMI) otg_set_vbus(musb->xceiv->otg, 0); @@ -307,6 +256,9 @@ static void omap_musb_set_mailbox(struct omap2430_glue *glue) dev_dbg(dev, "ID float\n"); } + if (!cable_connected) + omap2430_set_power(musb, glue->enabled, cable_connected); + atomic_notifier_call_chain(&musb->xceiv->notifier, musb->xceiv->last_event, NULL); } @@ -316,13 +268,8 @@ static void omap_musb_mailbox_work(struct work_struct *mailbox_work) { struct omap2430_glue *glue = container_of(mailbox_work, struct omap2430_glue, omap_musb_mailbox_work); - struct musb *musb = glue_to_musb(glue); - struct device *dev = musb->controller; - pm_runtime_get_sync(dev); omap_musb_set_mailbox(glue); - pm_runtime_mark_last_busy(dev); - pm_runtime_put_autosuspend(dev); } static irqreturn_t omap2430_musb_interrupt(int irq, void *__hci) @@ -389,23 +336,7 @@ static int omap2430_musb_init(struct musb *musb) return PTR_ERR(musb->phy); } musb->isr = omap2430_musb_interrupt; - - /* - * Enable runtime PM for musb parent (this driver). We can't - * do it earlier as struct musb is not yet allocated and we - * need to touch the musb registers for runtime PM. - */ - pm_runtime_enable(glue->dev); - status = pm_runtime_get_sync(glue->dev); - if (status < 0) - goto err1; - - status = pm_runtime_get_sync(dev); - if (status < 0) { - dev_err(dev, "pm_runtime_get_sync FAILED %d\n", status); - pm_runtime_put_sync(glue->dev); - goto err1; - } + phy_init(musb->phy); l = musb_readl(musb->mregs, OTG_INTERFSEL); @@ -427,20 +358,10 @@ static int omap2430_musb_init(struct musb *musb) musb_readl(musb->mregs, OTG_INTERFSEL), musb_readl(musb->mregs, OTG_SIMENABLE)); - setup_timer(&musb_idle_timer, musb_do_idle, (unsigned long) musb); - if (glue->status != MUSB_UNKNOWN) omap_musb_set_mailbox(glue); - phy_init(musb->phy); - phy_power_on(musb->phy); - - pm_runtime_put_noidle(musb->controller); - pm_runtime_put_noidle(glue->dev); return 0; - -err1: - return status; } static void omap2430_musb_enable(struct musb *musb) @@ -452,6 +373,11 @@ static void omap2430_musb_enable(struct musb *musb) struct musb_hdrc_platform_data *pdata = dev_get_platdata(dev); struct omap_musb_board_data *data = pdata->board_data; + if (!WARN_ON(!musb->phy)) + phy_power_on(musb->phy); + + omap2430_set_power(musb, true, glue->cable_connected); + switch (glue->status) { case MUSB_ID_GROUND: @@ -487,18 +413,25 @@ static void omap2430_musb_disable(struct musb *musb) struct device *dev = musb->controller; struct omap2430_glue *glue = dev_get_drvdata(dev->parent); + if (!WARN_ON(!musb->phy)) + phy_power_off(musb->phy); + if (glue->status != MUSB_UNKNOWN) omap_control_usb_set_mode(glue->control_otghs, USB_MODE_DISCONNECT); + + omap2430_set_power(musb, false, glue->cable_connected); } static int omap2430_musb_exit(struct musb *musb) { - del_timer_sync(&musb_idle_timer); + struct device *dev = musb->controller; + struct omap2430_glue *glue = dev_get_drvdata(dev->parent); omap2430_low_level_exit(musb); - phy_power_off(musb->phy); phy_exit(musb->phy); + musb->phy = NULL; + cancel_work_sync(&glue->omap_musb_mailbox_work); return 0; } @@ -512,9 +445,6 @@ static const struct musb_platform_ops omap2430_ops = { .init = omap2430_musb_init, .exit = omap2430_musb_exit, - .set_mode = omap2430_musb_set_mode, - .try_idle = omap2430_musb_try_idle, - .set_vbus = omap2430_musb_set_vbus, .enable = omap2430_musb_enable, @@ -639,11 +569,9 @@ static int omap2430_probe(struct platform_device *pdev) goto err2; } - /* - * Note that we cannot enable PM runtime yet for this - * driver as we need struct musb initialized first. - * See omap2430_musb_init above. - */ + pm_runtime_enable(glue->dev); + pm_runtime_use_autosuspend(glue->dev); + pm_runtime_set_autosuspend_delay(glue->dev, 500); ret = platform_device_add(musb); if (ret) { @@ -662,12 +590,14 @@ err0: static int omap2430_remove(struct platform_device *pdev) { - struct omap2430_glue *glue = platform_get_drvdata(pdev); + struct omap2430_glue *glue = platform_get_drvdata(pdev); + struct musb *musb = glue_to_musb(glue); pm_runtime_get_sync(glue->dev); - cancel_work_sync(&glue->omap_musb_mailbox_work); platform_device_unregister(glue->musb); + omap2430_set_power(musb, false, false); pm_runtime_put_sync(glue->dev); + pm_runtime_dont_use_autosuspend(glue->dev); pm_runtime_disable(glue->dev); return 0; @@ -680,12 +610,13 @@ static int omap2430_runtime_suspend(struct device *dev) struct omap2430_glue *glue = dev_get_drvdata(dev); struct musb *musb = glue_to_musb(glue); - if (musb) { - musb->context.otg_interfsel = musb_readl(musb->mregs, - OTG_INTERFSEL); + if (!musb) + return 0; - omap2430_low_level_exit(musb); - } + musb->context.otg_interfsel = musb_readl(musb->mregs, + OTG_INTERFSEL); + + omap2430_low_level_exit(musb); return 0; } @@ -696,7 +627,7 @@ static int omap2430_runtime_resume(struct device *dev) struct musb *musb = glue_to_musb(glue); if (!musb) - return -EPROBE_DEFER; + return 0; omap2430_low_level_init(musb); musb_writel(musb->mregs, OTG_INTERFSEL, @@ -738,18 +669,8 @@ static struct platform_driver omap2430_driver = { }, }; +module_platform_driver(omap2430_driver); + MODULE_DESCRIPTION("OMAP2PLUS MUSB Glue Layer"); MODULE_AUTHOR("Felipe Balbi <balbi@ti.com>"); MODULE_LICENSE("GPL v2"); - -static int __init omap2430_init(void) -{ - return platform_driver_register(&omap2430_driver); -} -subsys_initcall(omap2430_init); - -static void __exit omap2430_exit(void) -{ - platform_driver_unregister(&omap2430_driver); -} -module_exit(omap2430_exit); diff --git a/drivers/usb/musb/sunxi.c b/drivers/usb/musb/sunxi.c index fdab4232cfbf..76500515dd8b 100644 --- a/drivers/usb/musb/sunxi.c +++ b/drivers/usb/musb/sunxi.c @@ -80,7 +80,8 @@ static struct musb *sunxi_musb; struct sunxi_glue { struct device *dev; - struct platform_device *musb; + struct musb *musb; + struct platform_device *musb_pdev; struct clk *clk; struct reset_control *rst; struct phy *phy; @@ -102,7 +103,7 @@ static void sunxi_musb_work(struct work_struct *work) return; if (test_and_clear_bit(SUNXI_MUSB_FL_HOSTMODE_PEND, &glue->flags)) { - struct musb *musb = platform_get_drvdata(glue->musb); + struct musb *musb = glue->musb; unsigned long flags; u8 devctl; @@ -112,7 +113,7 @@ static void sunxi_musb_work(struct work_struct *work) if (test_bit(SUNXI_MUSB_FL_HOSTMODE, &glue->flags)) { set_bit(SUNXI_MUSB_FL_VBUS_ON, &glue->flags); musb->xceiv->otg->default_a = 1; - musb->xceiv->otg->state = OTG_STATE_A_IDLE; + musb->xceiv->otg->state = OTG_STATE_A_WAIT_VRISE; MUSB_HST_MODE(musb); devctl |= MUSB_DEVCTL_SESSION; } else { @@ -145,10 +146,12 @@ static void sunxi_musb_set_vbus(struct musb *musb, int is_on) { struct sunxi_glue *glue = dev_get_drvdata(musb->controller->parent); - if (is_on) + if (is_on) { set_bit(SUNXI_MUSB_FL_VBUS_ON, &glue->flags); - else + musb->xceiv->otg->state = OTG_STATE_A_WAIT_VRISE; + } else { clear_bit(SUNXI_MUSB_FL_VBUS_ON, &glue->flags); + } schedule_work(&glue->work); } @@ -264,15 +267,6 @@ static int sunxi_musb_init(struct musb *musb) if (ret) goto error_unregister_notifier; - if (musb->port_mode == MUSB_PORT_MODE_HOST) { - ret = phy_power_on(glue->phy); - if (ret) - goto error_phy_exit; - set_bit(SUNXI_MUSB_FL_PHY_ON, &glue->flags); - /* Stop musb work from turning vbus off again */ - set_bit(SUNXI_MUSB_FL_VBUS_ON, &glue->flags); - } - musb->isr = sunxi_musb_interrupt; /* Stop the musb-core from doing runtime pm (not supported on sunxi) */ @@ -280,8 +274,6 @@ static int sunxi_musb_init(struct musb *musb) return 0; -error_phy_exit: - phy_exit(glue->phy); error_unregister_notifier: if (musb->port_mode == MUSB_PORT_MODE_DUAL_ROLE) extcon_unregister_notifier(glue->extcon, EXTCON_USB_HOST, @@ -323,10 +315,31 @@ static int sunxi_musb_exit(struct musb *musb) return 0; } +static int sunxi_set_mode(struct musb *musb, u8 mode) +{ + struct sunxi_glue *glue = dev_get_drvdata(musb->controller->parent); + int ret; + + if (mode == MUSB_HOST) { + ret = phy_power_on(glue->phy); + if (ret) + return ret; + + set_bit(SUNXI_MUSB_FL_PHY_ON, &glue->flags); + /* Stop musb work from turning vbus off again */ + set_bit(SUNXI_MUSB_FL_VBUS_ON, &glue->flags); + musb->xceiv->otg->state = OTG_STATE_A_WAIT_VRISE; + } + + return 0; +} + static void sunxi_musb_enable(struct musb *musb) { struct sunxi_glue *glue = dev_get_drvdata(musb->controller->parent); + glue->musb = musb; + /* musb_core does not call us in a balanced manner */ if (test_and_set_bit(SUNXI_MUSB_FL_ENABLED, &glue->flags)) return; @@ -569,6 +582,7 @@ static const struct musb_platform_ops sunxi_musb_ops = { .exit = sunxi_musb_exit, .enable = sunxi_musb_enable, .disable = sunxi_musb_disable, + .set_mode = sunxi_set_mode, .fifo_offset = sunxi_musb_fifo_offset, .ep_offset = sunxi_musb_ep_offset, .busctl_offset = sunxi_musb_busctl_offset, @@ -721,9 +735,9 @@ static int sunxi_musb_probe(struct platform_device *pdev) pinfo.data = &pdata; pinfo.size_data = sizeof(pdata); - glue->musb = platform_device_register_full(&pinfo); - if (IS_ERR(glue->musb)) { - ret = PTR_ERR(glue->musb); + glue->musb_pdev = platform_device_register_full(&pinfo); + if (IS_ERR(glue->musb_pdev)) { + ret = PTR_ERR(glue->musb_pdev); dev_err(&pdev->dev, "Error registering musb dev: %d\n", ret); goto err_unregister_usb_phy; } @@ -740,7 +754,7 @@ static int sunxi_musb_remove(struct platform_device *pdev) struct sunxi_glue *glue = platform_get_drvdata(pdev); struct platform_device *usb_phy = glue->usb_phy; - platform_device_unregister(glue->musb); /* Frees glue ! */ + platform_device_unregister(glue->musb_pdev); usb_phy_generic_unregister(usb_phy); return 0; diff --git a/drivers/usb/phy/phy-twl6030-usb.c b/drivers/usb/phy/phy-twl6030-usb.c index 24e2b3cf1867..a72e8d670adc 100644 --- a/drivers/usb/phy/phy-twl6030-usb.c +++ b/drivers/usb/phy/phy-twl6030-usb.c @@ -97,6 +97,9 @@ struct twl6030_usb { struct regulator *usb3v3; + /* used to check initial cable status after probe */ + struct delayed_work get_status_work; + /* used to set vbus, in atomic path */ struct work_struct set_vbus_work; @@ -227,12 +230,16 @@ static irqreturn_t twl6030_usb_irq(int irq, void *_twl) twl->asleep = 1; status = MUSB_VBUS_VALID; twl->linkstat = status; - musb_mailbox(status); + ret = musb_mailbox(status); + if (ret) + twl->linkstat = MUSB_UNKNOWN; } else { if (twl->linkstat != MUSB_UNKNOWN) { status = MUSB_VBUS_OFF; twl->linkstat = status; - musb_mailbox(status); + ret = musb_mailbox(status); + if (ret) + twl->linkstat = MUSB_UNKNOWN; if (twl->asleep) { regulator_disable(twl->usb3v3); twl->asleep = 0; @@ -264,7 +271,9 @@ static irqreturn_t twl6030_usbotg_irq(int irq, void *_twl) twl6030_writeb(twl, TWL_MODULE_USB, 0x10, USB_ID_INT_EN_HI_SET); status = MUSB_ID_GROUND; twl->linkstat = status; - musb_mailbox(status); + ret = musb_mailbox(status); + if (ret) + twl->linkstat = MUSB_UNKNOWN; } else { twl6030_writeb(twl, TWL_MODULE_USB, 0x10, USB_ID_INT_EN_HI_CLR); twl6030_writeb(twl, TWL_MODULE_USB, 0x1, USB_ID_INT_EN_HI_SET); @@ -274,6 +283,15 @@ static irqreturn_t twl6030_usbotg_irq(int irq, void *_twl) return IRQ_HANDLED; } +static void twl6030_status_work(struct work_struct *work) +{ + struct twl6030_usb *twl = container_of(work, struct twl6030_usb, + get_status_work.work); + + twl6030_usb_irq(twl->irq2, twl); + twl6030_usbotg_irq(twl->irq1, twl); +} + static int twl6030_enable_irq(struct twl6030_usb *twl) { twl6030_writeb(twl, TWL_MODULE_USB, 0x1, USB_ID_INT_EN_HI_SET); @@ -284,8 +302,6 @@ static int twl6030_enable_irq(struct twl6030_usb *twl) REG_INT_MSK_LINE_C); twl6030_interrupt_unmask(TWL6030_CHARGER_CTRL_INT_MASK, REG_INT_MSK_STS_C); - twl6030_usb_irq(twl->irq2, twl); - twl6030_usbotg_irq(twl->irq1, twl); return 0; } @@ -371,6 +387,7 @@ static int twl6030_usb_probe(struct platform_device *pdev) dev_warn(&pdev->dev, "could not create sysfs file\n"); INIT_WORK(&twl->set_vbus_work, otg_set_vbus_work); + INIT_DELAYED_WORK(&twl->get_status_work, twl6030_status_work); status = request_threaded_irq(twl->irq1, NULL, twl6030_usbotg_irq, IRQF_TRIGGER_FALLING | IRQF_TRIGGER_RISING | IRQF_ONESHOT, @@ -395,6 +412,7 @@ static int twl6030_usb_probe(struct platform_device *pdev) twl->asleep = 0; twl6030_enable_irq(twl); + schedule_delayed_work(&twl->get_status_work, HZ); dev_info(&pdev->dev, "Initialized TWL6030 USB module\n"); return 0; @@ -404,6 +422,7 @@ static int twl6030_usb_remove(struct platform_device *pdev) { struct twl6030_usb *twl = platform_get_drvdata(pdev); + cancel_delayed_work(&twl->get_status_work); twl6030_interrupt_mask(TWL6030_USBOTG_INT_MASK, REG_INT_MSK_LINE_C); twl6030_interrupt_mask(TWL6030_USBOTG_INT_MASK, diff --git a/drivers/usb/serial/mos7720.c b/drivers/usb/serial/mos7720.c index 2eddbe538cda..5608af4a369d 100644 --- a/drivers/usb/serial/mos7720.c +++ b/drivers/usb/serial/mos7720.c @@ -2007,6 +2007,7 @@ static void mos7720_release(struct usb_serial *serial) urblist_entry) usb_unlink_urb(urbtrack->urb); spin_unlock_irqrestore(&mos_parport->listlock, flags); + parport_del_port(mos_parport->pp); kref_put(&mos_parport->ref_count, destroy_mos_parport); } diff --git a/drivers/usb/storage/uas.c b/drivers/usb/storage/uas.c index 4d49fce406e1..5ef014ba6ae8 100644 --- a/drivers/usb/storage/uas.c +++ b/drivers/usb/storage/uas.c @@ -836,6 +836,7 @@ static int uas_slave_configure(struct scsi_device *sdev) if (devinfo->flags & US_FL_BROKEN_FUA) sdev->broken_fua = 1; + scsi_change_queue_depth(sdev, devinfo->qdepth - 2); return 0; } @@ -848,7 +849,6 @@ static struct scsi_host_template uas_host_template = { .slave_configure = uas_slave_configure, .eh_abort_handler = uas_eh_abort_handler, .eh_bus_reset_handler = uas_eh_bus_reset_handler, - .can_queue = MAX_CMNDS, .this_id = -1, .sg_tablesize = SG_NONE, .skip_settle_delay = 1, diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c index fca51105974e..2e0450bec1b1 100644 --- a/drivers/usb/usbip/vhci_hcd.c +++ b/drivers/usb/usbip/vhci_hcd.c @@ -941,7 +941,7 @@ static void vhci_stop(struct usb_hcd *hcd) static int vhci_get_frame_number(struct usb_hcd *hcd) { - pr_err("Not yet implemented\n"); + dev_err_ratelimited(&hcd->self.root_hub->dev, "Not yet implemented\n"); return 0; } diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig index b54f26c55dfd..b4b3e256491b 100644 --- a/drivers/watchdog/Kconfig +++ b/drivers/watchdog/Kconfig @@ -746,7 +746,7 @@ config ALIM7101_WDT config EBC_C384_WDT tristate "WinSystems EBC-C384 Watchdog Timer" - depends on X86 && ISA + depends on X86 && ISA_BUS_API select WATCHDOG_CORE help Enables watchdog timer support for the watchdog timer on the diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c index d46839f51e73..e4db19e88ab1 100644 --- a/drivers/xen/balloon.c +++ b/drivers/xen/balloon.c @@ -151,8 +151,6 @@ static DECLARE_WAIT_QUEUE_HEAD(balloon_wq); static void balloon_process(struct work_struct *work); static DECLARE_DELAYED_WORK(balloon_worker, balloon_process); -static void release_memory_resource(struct resource *resource); - /* When ballooning out (allocating memory to return to Xen) we don't really want the kernel to try too hard since that can trigger the oom killer. */ #define GFP_BALLOON \ @@ -248,6 +246,19 @@ static enum bp_state update_schedule(enum bp_state state) } #ifdef CONFIG_XEN_BALLOON_MEMORY_HOTPLUG +static void release_memory_resource(struct resource *resource) +{ + if (!resource) + return; + + /* + * No need to reset region to identity mapped since we now + * know that no I/O can be in this region + */ + release_resource(resource); + kfree(resource); +} + static struct resource *additional_memory_resource(phys_addr_t size) { struct resource *res; @@ -286,19 +297,6 @@ static struct resource *additional_memory_resource(phys_addr_t size) return res; } -static void release_memory_resource(struct resource *resource) -{ - if (!resource) - return; - - /* - * No need to reset region to identity mapped since we now - * know that no I/O can be in this region - */ - release_resource(resource); - kfree(resource); -} - static enum bp_state reserve_additional_memory(void) { long credit; diff --git a/drivers/xen/xen-pciback/conf_space.c b/drivers/xen/xen-pciback/conf_space.c index 8e67336f8ddd..6a25533da237 100644 --- a/drivers/xen/xen-pciback/conf_space.c +++ b/drivers/xen/xen-pciback/conf_space.c @@ -183,8 +183,7 @@ int xen_pcibk_config_read(struct pci_dev *dev, int offset, int size, field_start = OFFSET(cfg_entry); field_end = OFFSET(cfg_entry) + field->size; - if ((req_start >= field_start && req_start < field_end) - || (req_end > field_start && req_end <= field_end)) { + if (req_end > field_start && field_end > req_start) { err = conf_space_read(dev, cfg_entry, field_start, &tmp_val); if (err) @@ -230,8 +229,7 @@ int xen_pcibk_config_write(struct pci_dev *dev, int offset, int size, u32 value) field_start = OFFSET(cfg_entry); field_end = OFFSET(cfg_entry) + field->size; - if ((req_start >= field_start && req_start < field_end) - || (req_end > field_start && req_end <= field_end)) { + if (req_end > field_start && field_end > req_start) { tmp_val = 0; err = xen_pcibk_config_read(dev, field_start, diff --git a/drivers/xen/xen-pciback/conf_space_header.c b/drivers/xen/xen-pciback/conf_space_header.c index ad3d17d29c81..9ead1c2ff1dd 100644 --- a/drivers/xen/xen-pciback/conf_space_header.c +++ b/drivers/xen/xen-pciback/conf_space_header.c @@ -145,7 +145,7 @@ static int rom_write(struct pci_dev *dev, int offset, u32 value, void *data) /* A write to obtain the length must happen as a 32-bit write. * This does not (yet) support writing individual bytes */ - if (value == ~PCI_ROM_ADDRESS_ENABLE) + if ((value | ~PCI_ROM_ADDRESS_MASK) == ~0U) bar->which = 1; else { u32 tmpval; @@ -225,38 +225,42 @@ static inline void read_dev_bar(struct pci_dev *dev, (PCI_BASE_ADDRESS_SPACE_MEMORY | PCI_BASE_ADDRESS_MEM_TYPE_64))) { bar_info->val = res[pos - 1].start >> 32; - bar_info->len_val = res[pos - 1].end >> 32; + bar_info->len_val = -resource_size(&res[pos - 1]) >> 32; return; } } + if (!res[pos].flags || + (res[pos].flags & (IORESOURCE_DISABLED | IORESOURCE_UNSET | + IORESOURCE_BUSY))) + return; + bar_info->val = res[pos].start | (res[pos].flags & PCI_REGION_FLAG_MASK); - bar_info->len_val = resource_size(&res[pos]); + bar_info->len_val = -resource_size(&res[pos]) | + (res[pos].flags & PCI_REGION_FLAG_MASK); } static void *bar_init(struct pci_dev *dev, int offset) { - struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL); + struct pci_bar_info *bar = kzalloc(sizeof(*bar), GFP_KERNEL); if (!bar) return ERR_PTR(-ENOMEM); read_dev_bar(dev, bar, offset, ~0); - bar->which = 0; return bar; } static void *rom_init(struct pci_dev *dev, int offset) { - struct pci_bar_info *bar = kmalloc(sizeof(*bar), GFP_KERNEL); + struct pci_bar_info *bar = kzalloc(sizeof(*bar), GFP_KERNEL); if (!bar) return ERR_PTR(-ENOMEM); read_dev_bar(dev, bar, offset, ~PCI_ROM_ADDRESS_ENABLE); - bar->which = 0; return bar; } |