summaryrefslogtreecommitdiffstats
path: root/net/core
diff options
context:
space:
mode:
Diffstat (limited to 'net/core')
-rw-r--r--net/core/datagram.c6
-rw-r--r--net/core/dev.c63
-rw-r--r--net/core/devlink.c32
-rw-r--r--net/core/drop_monitor.c643
-rw-r--r--net/core/filter.c6
-rw-r--r--net/core/flow_offload.c240
-rw-r--r--net/core/neighbour.c2
-rw-r--r--net/core/pktgen.c2
-rw-r--r--net/core/skbuff.c54
-rw-r--r--net/core/skmsg.c4
-rw-r--r--net/core/sock_map.c19
-rw-r--r--net/core/sysctl_net_core.c8
12 files changed, 962 insertions, 117 deletions
diff --git a/net/core/datagram.c b/net/core/datagram.c
index 45a162ef5e02..4cc8dc5db2b7 100644
--- a/net/core/datagram.c
+++ b/net/core/datagram.c
@@ -442,8 +442,8 @@ static int __skb_datagram_iter(const struct sk_buff *skb, int offset,
if (copy > len)
copy = len;
- n = cb(vaddr + frag->page_offset +
- offset - start, copy, data, to);
+ n = cb(vaddr + skb_frag_off(frag) + offset - start,
+ copy, data, to);
kunmap(page);
offset += n;
if (n != copy)
@@ -573,7 +573,7 @@ int skb_copy_datagram_from_iter(struct sk_buff *skb, int offset,
if (copy > len)
copy = len;
copied = copy_page_from_iter(skb_frag_page(frag),
- frag->page_offset + offset - start,
+ skb_frag_off(frag) + offset - start,
copy, from);
if (copied != copy)
goto fault;
diff --git a/net/core/dev.c b/net/core/dev.c
index fc676b2610e3..49589ed2018d 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -3963,6 +3963,8 @@ int dev_weight_rx_bias __read_mostly = 1; /* bias for backlog weight */
int dev_weight_tx_bias __read_mostly = 1; /* bias for output_queue quota */
int dev_rx_weight __read_mostly = 64;
int dev_tx_weight __read_mostly = 64;
+/* Maximum number of GRO_NORMAL skbs to batch up for list-RX */
+int gro_normal_batch __read_mostly = 8;
/* Called with irq disabled */
static inline void ____napi_schedule(struct softnet_data *sd,
@@ -4374,12 +4376,17 @@ static u32 netif_receive_generic_xdp(struct sk_buff *skb,
act = bpf_prog_run_xdp(xdp_prog, xdp);
+ /* check if bpf_xdp_adjust_head was used */
off = xdp->data - orig_data;
- if (off > 0)
- __skb_pull(skb, off);
- else if (off < 0)
- __skb_push(skb, -off);
- skb->mac_header += off;
+ if (off) {
+ if (off > 0)
+ __skb_pull(skb, off);
+ else if (off < 0)
+ __skb_push(skb, -off);
+
+ skb->mac_header += off;
+ skb_reset_network_header(skb);
+ }
/* check if bpf_xdp_adjust_tail was used. it can only "shrink"
* pckt.
@@ -5481,7 +5488,7 @@ static void gro_pull_from_frag0(struct sk_buff *skb, int grow)
skb->data_len -= grow;
skb->tail += grow;
- pinfo->frags[0].page_offset += grow;
+ skb_frag_off_add(&pinfo->frags[0], grow);
skb_frag_size_sub(&pinfo->frags[0], grow);
if (unlikely(!skb_frag_size(&pinfo->frags[0]))) {
@@ -5742,6 +5749,26 @@ struct sk_buff *napi_get_frags(struct napi_struct *napi)
}
EXPORT_SYMBOL(napi_get_frags);
+/* Pass the currently batched GRO_NORMAL SKBs up to the stack. */
+static void gro_normal_list(struct napi_struct *napi)
+{
+ if (!napi->rx_count)
+ return;
+ netif_receive_skb_list_internal(&napi->rx_list);
+ INIT_LIST_HEAD(&napi->rx_list);
+ napi->rx_count = 0;
+}
+
+/* Queue one GRO_NORMAL SKB up for list processing. If batch size exceeded,
+ * pass the whole batch up to the stack.
+ */
+static void gro_normal_one(struct napi_struct *napi, struct sk_buff *skb)
+{
+ list_add_tail(&skb->list, &napi->rx_list);
+ if (++napi->rx_count >= gro_normal_batch)
+ gro_normal_list(napi);
+}
+
static gro_result_t napi_frags_finish(struct napi_struct *napi,
struct sk_buff *skb,
gro_result_t ret)
@@ -5751,8 +5778,8 @@ static gro_result_t napi_frags_finish(struct napi_struct *napi,
case GRO_HELD:
__skb_push(skb, ETH_HLEN);
skb->protocol = eth_type_trans(skb, skb->dev);
- if (ret == GRO_NORMAL && netif_receive_skb_internal(skb))
- ret = GRO_DROP;
+ if (ret == GRO_NORMAL)
+ gro_normal_one(napi, skb);
break;
case GRO_DROP:
@@ -6029,6 +6056,8 @@ bool napi_complete_done(struct napi_struct *n, int work_done)
NAPIF_STATE_IN_BUSY_POLL)))
return false;
+ gro_normal_list(n);
+
if (n->gro_bitmask) {
unsigned long timeout = 0;
@@ -6114,10 +6143,19 @@ static void busy_poll_stop(struct napi_struct *napi, void *have_poll_lock)
* Ideally, a new ndo_busy_poll_stop() could avoid another round.
*/
rc = napi->poll(napi, BUSY_POLL_BUDGET);
+ /* We can't gro_normal_list() here, because napi->poll() might have
+ * rearmed the napi (napi_complete_done()) in which case it could
+ * already be running on another CPU.
+ */
trace_napi_poll(napi, rc, BUSY_POLL_BUDGET);
netpoll_poll_unlock(have_poll_lock);
- if (rc == BUSY_POLL_BUDGET)
+ if (rc == BUSY_POLL_BUDGET) {
+ /* As the whole budget was spent, we still own the napi so can
+ * safely handle the rx_list.
+ */
+ gro_normal_list(napi);
__napi_schedule(napi);
+ }
local_bh_enable();
}
@@ -6162,6 +6200,7 @@ restart:
}
work = napi_poll(napi, BUSY_POLL_BUDGET);
trace_napi_poll(napi, work, BUSY_POLL_BUDGET);
+ gro_normal_list(napi);
count:
if (work > 0)
__NET_ADD_STATS(dev_net(napi->dev),
@@ -6267,6 +6306,8 @@ void netif_napi_add(struct net_device *dev, struct napi_struct *napi,
napi->timer.function = napi_watchdog;
init_gro_hash(napi);
napi->skb = NULL;
+ INIT_LIST_HEAD(&napi->rx_list);
+ napi->rx_count = 0;
napi->poll = poll;
if (weight > NAPI_POLL_WEIGHT)
netdev_err_once(dev, "%s() called with weight %d\n", __func__,
@@ -6363,6 +6404,8 @@ static int napi_poll(struct napi_struct *n, struct list_head *repoll)
goto out_unlock;
}
+ gro_normal_list(n);
+
if (n->gro_bitmask) {
/* flush too old packets
* If HZ < 1000, flush all packets.
@@ -9701,6 +9744,8 @@ static void __net_exit default_device_exit(struct net *net)
/* Push remaining network devices to init_net */
snprintf(fb_name, IFNAMSIZ, "dev%d", dev->ifindex);
+ if (__dev_get_by_name(&init_net, fb_name))
+ snprintf(fb_name, IFNAMSIZ, "dev%%d");
err = dev_change_net_namespace(dev, &init_net, fb_name);
if (err) {
pr_emerg("%s: failed to move %s to init_net: %d\n",
diff --git a/net/core/devlink.c b/net/core/devlink.c
index 4f40aeace902..e3a1ae44f93d 100644
--- a/net/core/devlink.c
+++ b/net/core/devlink.c
@@ -342,7 +342,6 @@ struct devlink_snapshot {
struct list_head list;
struct devlink_region *region;
devlink_snapshot_data_dest_t *data_destructor;
- u64 data_len;
u8 *data;
u32 id;
};
@@ -371,14 +370,6 @@ devlink_region_snapshot_get_by_id(struct devlink_region *region, u32 id)
return NULL;
}
-static void devlink_region_snapshot_del(struct devlink_snapshot *snapshot)
-{
- snapshot->region->cur_snapshots--;
- list_del(&snapshot->list);
- (*snapshot->data_destructor)(snapshot->data);
- kfree(snapshot);
-}
-
#define DEVLINK_NL_FLAG_NEED_DEVLINK BIT(0)
#define DEVLINK_NL_FLAG_NEED_PORT BIT(1)
#define DEVLINK_NL_FLAG_NEED_SB BIT(2)
@@ -3596,6 +3587,16 @@ out_free_msg:
nlmsg_free(msg);
}
+static void devlink_region_snapshot_del(struct devlink_region *region,
+ struct devlink_snapshot *snapshot)
+{
+ devlink_nl_region_notify(region, snapshot, DEVLINK_CMD_REGION_DEL);
+ region->cur_snapshots--;
+ list_del(&snapshot->list);
+ (*snapshot->data_destructor)(snapshot->data);
+ kfree(snapshot);
+}
+
static int devlink_nl_cmd_region_get_doit(struct sk_buff *skb,
struct genl_info *info)
{
@@ -3691,8 +3692,7 @@ static int devlink_nl_cmd_region_del(struct sk_buff *skb,
if (!snapshot)
return -EINVAL;
- devlink_nl_region_notify(region, snapshot, DEVLINK_CMD_REGION_DEL);
- devlink_region_snapshot_del(snapshot);
+ devlink_region_snapshot_del(region, snapshot);
return 0;
}
@@ -3748,8 +3748,8 @@ static int devlink_nl_region_read_snapshot_fill(struct sk_buff *skb,
if (!snapshot)
return -EINVAL;
- if (end_offset > snapshot->data_len || dump)
- end_offset = snapshot->data_len;
+ if (end_offset > region->size || dump)
+ end_offset = region->size;
while (curr_offset < end_offset) {
u32 data_size;
@@ -6744,7 +6744,7 @@ void devlink_region_destroy(struct devlink_region *region)
/* Free all snapshots of region */
list_for_each_entry_safe(snapshot, ts, &region->snapshot_list, list)
- devlink_region_snapshot_del(snapshot);
+ devlink_region_snapshot_del(region, snapshot);
list_del(&region->list);
@@ -6784,12 +6784,11 @@ EXPORT_SYMBOL_GPL(devlink_region_shapshot_id_get);
* The @snapshot_id should be obtained using the getter function.
*
* @region: devlink region of the snapshot
- * @data_len: size of snapshot data
* @data: snapshot data
* @snapshot_id: snapshot id to be created
* @data_destructor: pointer to destructor function to free data
*/
-int devlink_region_snapshot_create(struct devlink_region *region, u64 data_len,
+int devlink_region_snapshot_create(struct devlink_region *region,
u8 *data, u32 snapshot_id,
devlink_snapshot_data_dest_t *data_destructor)
{
@@ -6819,7 +6818,6 @@ int devlink_region_snapshot_create(struct devlink_region *region, u64 data_len,
snapshot->id = snapshot_id;
snapshot->region = region;
snapshot->data = data;
- snapshot->data_len = data_len;
snapshot->data_destructor = data_destructor;
list_add_tail(&snapshot->list, &region->snapshot_list);
diff --git a/net/core/drop_monitor.c b/net/core/drop_monitor.c
index 4ea4347f5062..39e094907391 100644
--- a/net/core/drop_monitor.c
+++ b/net/core/drop_monitor.c
@@ -43,13 +43,26 @@
* netlink alerts
*/
static int trace_state = TRACE_OFF;
-static DEFINE_MUTEX(trace_state_mutex);
+
+/* net_dm_mutex
+ *
+ * An overall lock guarding every operation coming from userspace.
+ * It also guards the global 'hw_stats_list' list.
+ */
+static DEFINE_MUTEX(net_dm_mutex);
+
+struct net_dm_stats {
+ u64 dropped;
+ struct u64_stats_sync syncp;
+};
struct per_cpu_dm_data {
- spinlock_t lock;
+ spinlock_t lock; /* Protects 'skb' and 'send_timer' */
struct sk_buff *skb;
+ struct sk_buff_head drop_queue;
struct work_struct dm_alert_work;
struct timer_list send_timer;
+ struct net_dm_stats stats;
};
struct dm_hw_stat_delta {
@@ -69,6 +82,24 @@ static int dm_delay = 1;
static unsigned long dm_hw_check_delta = 2*HZ;
static LIST_HEAD(hw_stats_list);
+static enum net_dm_alert_mode net_dm_alert_mode = NET_DM_ALERT_MODE_SUMMARY;
+static u32 net_dm_trunc_len;
+static u32 net_dm_queue_len = 1000;
+
+struct net_dm_alert_ops {
+ void (*kfree_skb_probe)(void *ignore, struct sk_buff *skb,
+ void *location);
+ void (*napi_poll_probe)(void *ignore, struct napi_struct *napi,
+ int work, int budget);
+ void (*work_item_func)(struct work_struct *work);
+};
+
+struct net_dm_skb_cb {
+ void *pc;
+};
+
+#define NET_DM_SKB_CB(__skb) ((struct net_dm_skb_cb *)&((__skb)->cb[0]))
+
static struct sk_buff *reset_per_cpu_data(struct per_cpu_dm_data *data)
{
size_t al;
@@ -235,48 +266,330 @@ static void trace_napi_poll_hit(void *ignore, struct napi_struct *napi,
rcu_read_unlock();
}
-static int set_all_monitor_traces(int state)
+static const struct net_dm_alert_ops net_dm_alert_summary_ops = {
+ .kfree_skb_probe = trace_kfree_skb_hit,
+ .napi_poll_probe = trace_napi_poll_hit,
+ .work_item_func = send_dm_alert,
+};
+
+static void net_dm_packet_trace_kfree_skb_hit(void *ignore,
+ struct sk_buff *skb,
+ void *location)
{
- int rc = 0;
- struct dm_hw_stat_delta *new_stat = NULL;
- struct dm_hw_stat_delta *temp;
+ ktime_t tstamp = ktime_get_real();
+ struct per_cpu_dm_data *data;
+ struct sk_buff *nskb;
+ unsigned long flags;
- mutex_lock(&trace_state_mutex);
+ nskb = skb_clone(skb, GFP_ATOMIC);
+ if (!nskb)
+ return;
- if (state == trace_state) {
- rc = -EAGAIN;
- goto out_unlock;
+ NET_DM_SKB_CB(nskb)->pc = location;
+ /* Override the timestamp because we care about the time when the
+ * packet was dropped.
+ */
+ nskb->tstamp = tstamp;
+
+ data = this_cpu_ptr(&dm_cpu_data);
+
+ spin_lock_irqsave(&data->drop_queue.lock, flags);
+ if (skb_queue_len(&data->drop_queue) < net_dm_queue_len)
+ __skb_queue_tail(&data->drop_queue, nskb);
+ else
+ goto unlock_free;
+ spin_unlock_irqrestore(&data->drop_queue.lock, flags);
+
+ schedule_work(&data->dm_alert_work);
+
+ return;
+
+unlock_free:
+ spin_unlock_irqrestore(&data->drop_queue.lock, flags);
+ u64_stats_update_begin(&data->stats.syncp);
+ data->stats.dropped++;
+ u64_stats_update_end(&data->stats.syncp);
+ consume_skb(nskb);
+}
+
+static void net_dm_packet_trace_napi_poll_hit(void *ignore,
+ struct napi_struct *napi,
+ int work, int budget)
+{
+}
+
+static size_t net_dm_in_port_size(void)
+{
+ /* NET_DM_ATTR_IN_PORT nest */
+ return nla_total_size(0) +
+ /* NET_DM_ATTR_PORT_NETDEV_IFINDEX */
+ nla_total_size(sizeof(u32));
+}
+
+#define NET_DM_MAX_SYMBOL_LEN 40
+
+static size_t net_dm_packet_report_size(size_t payload_len)
+{
+ size_t size;
+
+ size = nlmsg_msg_size(GENL_HDRLEN + net_drop_monitor_family.hdrsize);
+
+ return NLMSG_ALIGN(size) +
+ /* NET_DM_ATTR_PC */
+ nla_total_size(sizeof(u64)) +
+ /* NET_DM_ATTR_SYMBOL */
+ nla_total_size(NET_DM_MAX_SYMBOL_LEN + 1) +
+ /* NET_DM_ATTR_IN_PORT */
+ net_dm_in_port_size() +
+ /* NET_DM_ATTR_TIMESTAMP */
+ nla_total_size(sizeof(struct timespec)) +
+ /* NET_DM_ATTR_ORIG_LEN */
+ nla_total_size(sizeof(u32)) +
+ /* NET_DM_ATTR_PROTO */
+ nla_total_size(sizeof(u16)) +
+ /* NET_DM_ATTR_PAYLOAD */
+ nla_total_size(payload_len);
+}
+
+static int net_dm_packet_report_in_port_put(struct sk_buff *msg, int ifindex)
+{
+ struct nlattr *attr;
+
+ attr = nla_nest_start(msg, NET_DM_ATTR_IN_PORT);
+ if (!attr)
+ return -EMSGSIZE;
+
+ if (ifindex &&
+ nla_put_u32(msg, NET_DM_ATTR_PORT_NETDEV_IFINDEX, ifindex))
+ goto nla_put_failure;
+
+ nla_nest_end(msg, attr);
+
+ return 0;
+
+nla_put_failure:
+ nla_nest_cancel(msg, attr);
+ return -EMSGSIZE;
+}
+
+static int net_dm_packet_report_fill(struct sk_buff *msg, struct sk_buff *skb,
+ size_t payload_len)
+{
+ u64 pc = (u64)(uintptr_t) NET_DM_SKB_CB(skb)->pc;
+ char buf[NET_DM_MAX_SYMBOL_LEN];
+ struct nlattr *attr;
+ struct timespec ts;
+ void *hdr;
+ int rc;
+
+ hdr = genlmsg_put(msg, 0, 0, &net_drop_monitor_family, 0,
+ NET_DM_CMD_PACKET_ALERT);
+ if (!hdr)
+ return -EMSGSIZE;
+
+ if (nla_put_u64_64bit(msg, NET_DM_ATTR_PC, pc, NET_DM_ATTR_PAD))
+ goto nla_put_failure;
+
+ snprintf(buf, sizeof(buf), "%pS", NET_DM_SKB_CB(skb)->pc);
+ if (nla_put_string(msg, NET_DM_ATTR_SYMBOL, buf))
+ goto nla_put_failure;
+
+ rc = net_dm_packet_report_in_port_put(msg, skb->skb_iif);
+ if (rc)
+ goto nla_put_failure;
+
+ if (ktime_to_timespec_cond(skb->tstamp, &ts) &&
+ nla_put(msg, NET_DM_ATTR_TIMESTAMP, sizeof(ts), &ts))
+ goto nla_put_failure;
+
+ if (nla_put_u32(msg, NET_DM_ATTR_ORIG_LEN, skb->len))
+ goto nla_put_failure;
+
+ if (!payload_len)
+ goto out;
+
+ if (nla_put_u16(msg, NET_DM_ATTR_PROTO, be16_to_cpu(skb->protocol)))
+ goto nla_put_failure;
+
+ attr = skb_put(msg, nla_total_size(payload_len));
+ attr->nla_type = NET_DM_ATTR_PAYLOAD;
+ attr->nla_len = nla_attr_size(payload_len);
+ if (skb_copy_bits(skb, 0, nla_data(attr), payload_len))
+ goto nla_put_failure;
+
+out:
+ genlmsg_end(msg, hdr);
+
+ return 0;
+
+nla_put_failure:
+ genlmsg_cancel(msg, hdr);
+ return -EMSGSIZE;
+}
+
+#define NET_DM_MAX_PACKET_SIZE (0xffff - NLA_HDRLEN - NLA_ALIGNTO)
+
+static void net_dm_packet_report(struct sk_buff *skb)
+{
+ struct sk_buff *msg;
+ size_t payload_len;
+ int rc;
+
+ /* Make sure we start copying the packet from the MAC header */
+ if (skb->data > skb_mac_header(skb))
+ skb_push(skb, skb->data - skb_mac_header(skb));
+ else
+ skb_pull(skb, skb_mac_header(skb) - skb->data);
+
+ /* Ensure packet fits inside a single netlink attribute */
+ payload_len = min_t(size_t, skb->len, NET_DM_MAX_PACKET_SIZE);
+ if (net_dm_trunc_len)
+ payload_len = min_t(size_t, net_dm_trunc_len, payload_len);
+
+ msg = nlmsg_new(net_dm_packet_report_size(payload_len), GFP_KERNEL);
+ if (!msg)
+ goto out;
+
+ rc = net_dm_packet_report_fill(msg, skb, payload_len);
+ if (rc) {
+ nlmsg_free(msg);
+ goto out;
}
- switch (state) {
- case TRACE_ON:
- if (!try_module_get(THIS_MODULE)) {
- rc = -ENODEV;
- break;
- }
+ genlmsg_multicast(&net_drop_monitor_family, msg, 0, 0, GFP_KERNEL);
- rc |= register_trace_kfree_skb(trace_kfree_skb_hit, NULL);
- rc |= register_trace_napi_poll(trace_napi_poll_hit, NULL);
- break;
+out:
+ consume_skb(skb);
+}
- case TRACE_OFF:
- rc |= unregister_trace_kfree_skb(trace_kfree_skb_hit, NULL);
- rc |= unregister_trace_napi_poll(trace_napi_poll_hit, NULL);
+static void net_dm_packet_work(struct work_struct *work)
+{
+ struct per_cpu_dm_data *data;
+ struct sk_buff_head list;
+ struct sk_buff *skb;
+ unsigned long flags;
+
+ data = container_of(work, struct per_cpu_dm_data, dm_alert_work);
- tracepoint_synchronize_unregister();
+ __skb_queue_head_init(&list);
- /*
- * Clean the device list
+ spin_lock_irqsave(&data->drop_queue.lock, flags);
+ skb_queue_splice_tail_init(&data->drop_queue, &list);
+ spin_unlock_irqrestore(&data->drop_queue.lock, flags);
+
+ while ((skb = __skb_dequeue(&list)))
+ net_dm_packet_report(skb);
+}
+
+static const struct net_dm_alert_ops net_dm_alert_packet_ops = {
+ .kfree_skb_probe = net_dm_packet_trace_kfree_skb_hit,
+ .napi_poll_probe = net_dm_packet_trace_napi_poll_hit,
+ .work_item_func = net_dm_packet_work,
+};
+
+static const struct net_dm_alert_ops *net_dm_alert_ops_arr[] = {
+ [NET_DM_ALERT_MODE_SUMMARY] = &net_dm_alert_summary_ops,
+ [NET_DM_ALERT_MODE_PACKET] = &net_dm_alert_packet_ops,
+};
+
+static int net_dm_trace_on_set(struct netlink_ext_ack *extack)
+{
+ const struct net_dm_alert_ops *ops;
+ int cpu, rc;
+
+ ops = net_dm_alert_ops_arr[net_dm_alert_mode];
+
+ if (!try_module_get(THIS_MODULE)) {
+ NL_SET_ERR_MSG_MOD(extack, "Failed to take reference on module");
+ return -ENODEV;
+ }
+
+ for_each_possible_cpu(cpu) {
+ struct per_cpu_dm_data *data = &per_cpu(dm_cpu_data, cpu);
+ struct sk_buff *skb;
+
+ INIT_WORK(&data->dm_alert_work, ops->work_item_func);
+ timer_setup(&data->send_timer, sched_send_work, 0);
+ /* Allocate a new per-CPU skb for the summary alert message and
+ * free the old one which might contain stale data from
+ * previous tracing.
*/
- list_for_each_entry_safe(new_stat, temp, &hw_stats_list, list) {
- if (new_stat->dev == NULL) {
- list_del_rcu(&new_stat->list);
- kfree_rcu(new_stat, rcu);
- }
+ skb = reset_per_cpu_data(data);
+ consume_skb(skb);
+ }
+
+ rc = register_trace_kfree_skb(ops->kfree_skb_probe, NULL);
+ if (rc) {
+ NL_SET_ERR_MSG_MOD(extack, "Failed to connect probe to kfree_skb() tracepoint");
+ goto err_module_put;
+ }
+
+ rc = register_trace_napi_poll(ops->napi_poll_probe, NULL);
+ if (rc) {
+ NL_SET_ERR_MSG_MOD(extack, "Failed to connect probe to napi_poll() tracepoint");
+ goto err_unregister_trace;
+ }
+
+ return 0;
+
+err_unregister_trace:
+ unregister_trace_kfree_skb(ops->kfree_skb_probe, NULL);
+err_module_put:
+ module_put(THIS_MODULE);
+ return rc;
+}
+
+static void net_dm_trace_off_set(void)
+{
+ struct dm_hw_stat_delta *new_stat, *temp;
+ const struct net_dm_alert_ops *ops;
+ int cpu;
+
+ ops = net_dm_alert_ops_arr[net_dm_alert_mode];
+
+ unregister_trace_napi_poll(ops->napi_poll_probe, NULL);
+ unregister_trace_kfree_skb(ops->kfree_skb_probe, NULL);
+
+ tracepoint_synchronize_unregister();
+
+ /* Make sure we do not send notifications to user space after request
+ * to stop tracing returns.
+ */
+ for_each_possible_cpu(cpu) {
+ struct per_cpu_dm_data *data = &per_cpu(dm_cpu_data, cpu);
+ struct sk_buff *skb;
+
+ del_timer_sync(&data->send_timer);
+ cancel_work_sync(&data->dm_alert_work);
+ while ((skb = __skb_dequeue(&data->drop_queue)))
+ consume_skb(skb);
+ }
+
+ list_for_each_entry_safe(new_stat, temp, &hw_stats_list, list) {
+ if (new_stat->dev == NULL) {
+ list_del_rcu(&new_stat->list);
+ kfree_rcu(new_stat, rcu);
}
+ }
+
+ module_put(THIS_MODULE);
+}
+
+static int set_all_monitor_traces(int state, struct netlink_ext_ack *extack)
+{
+ int rc = 0;
- module_put(THIS_MODULE);
+ if (state == trace_state) {
+ NL_SET_ERR_MSG_MOD(extack, "Trace state already set to requested state");
+ return -EAGAIN;
+ }
+ switch (state) {
+ case TRACE_ON:
+ rc = net_dm_trace_on_set(extack);
+ break;
+ case TRACE_OFF:
+ net_dm_trace_off_set();
break;
default:
rc = 1;
@@ -288,17 +601,84 @@ static int set_all_monitor_traces(int state)
else
rc = -EINPROGRESS;
-out_unlock:
- mutex_unlock(&trace_state_mutex);
-
return rc;
}
+static int net_dm_alert_mode_get_from_info(struct genl_info *info,
+ enum net_dm_alert_mode *p_alert_mode)
+{
+ u8 val;
+
+ val = nla_get_u8(info->attrs[NET_DM_ATTR_ALERT_MODE]);
+
+ switch (val) {
+ case NET_DM_ALERT_MODE_SUMMARY: /* fall-through */
+ case NET_DM_ALERT_MODE_PACKET:
+ *p_alert_mode = val;
+ break;
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int net_dm_alert_mode_set(struct genl_info *info)
+{
+ struct netlink_ext_ack *extack = info->extack;
+ enum net_dm_alert_mode alert_mode;
+ int rc;
+
+ if (!info->attrs[NET_DM_ATTR_ALERT_MODE])
+ return 0;
+
+ rc = net_dm_alert_mode_get_from_info(info, &alert_mode);
+ if (rc) {
+ NL_SET_ERR_MSG_MOD(extack, "Invalid alert mode");
+ return -EINVAL;
+ }
+
+ net_dm_alert_mode = alert_mode;
+
+ return 0;
+}
+
+static void net_dm_trunc_len_set(struct genl_info *info)
+{
+ if (!info->attrs[NET_DM_ATTR_TRUNC_LEN])
+ return;
+
+ net_dm_trunc_len = nla_get_u32(info->attrs[NET_DM_ATTR_TRUNC_LEN]);
+}
+
+static void net_dm_queue_len_set(struct genl_info *info)
+{
+ if (!info->attrs[NET_DM_ATTR_QUEUE_LEN])
+ return;
+
+ net_dm_queue_len = nla_get_u32(info->attrs[NET_DM_ATTR_QUEUE_LEN]);
+}
static int net_dm_cmd_config(struct sk_buff *skb,
struct genl_info *info)
{
- return -ENOTSUPP;
+ struct netlink_ext_ack *extack = info->extack;
+ int rc;
+
+ if (trace_state == TRACE_ON) {
+ NL_SET_ERR_MSG_MOD(extack, "Cannot configure drop monitor while tracing is on");
+ return -EBUSY;
+ }
+
+ rc = net_dm_alert_mode_set(info);
+ if (rc)
+ return rc;
+
+ net_dm_trunc_len_set(info);
+
+ net_dm_queue_len_set(info);
+
+ return 0;
}
static int net_dm_cmd_trace(struct sk_buff *skb,
@@ -306,12 +686,146 @@ static int net_dm_cmd_trace(struct sk_buff *skb,
{
switch (info->genlhdr->cmd) {
case NET_DM_CMD_START:
- return set_all_monitor_traces(TRACE_ON);
+ return set_all_monitor_traces(TRACE_ON, info->extack);
case NET_DM_CMD_STOP:
- return set_all_monitor_traces(TRACE_OFF);
+ return set_all_monitor_traces(TRACE_OFF, info->extack);
+ }
+
+ return -EOPNOTSUPP;
+}
+
+static int net_dm_config_fill(struct sk_buff *msg, struct genl_info *info)
+{
+ void *hdr;
+
+ hdr = genlmsg_put(msg, info->snd_portid, info->snd_seq,
+ &net_drop_monitor_family, 0, NET_DM_CMD_CONFIG_NEW);
+ if (!hdr)
+ return -EMSGSIZE;
+
+ if (nla_put_u8(msg, NET_DM_ATTR_ALERT_MODE, net_dm_alert_mode))
+ goto nla_put_failure;
+
+ if (nla_put_u32(msg, NET_DM_ATTR_TRUNC_LEN, net_dm_trunc_len))
+ goto nla_put_failure;
+
+ if (nla_put_u32(msg, NET_DM_ATTR_QUEUE_LEN, net_dm_queue_len))
+ goto nla_put_failure;
+
+ genlmsg_end(msg, hdr);
+
+ return 0;
+
+nla_put_failure:
+ genlmsg_cancel(msg, hdr);
+ return -EMSGSIZE;
+}
+
+static int net_dm_cmd_config_get(struct sk_buff *skb, struct genl_info *info)
+{
+ struct sk_buff *msg;
+ int rc;
+
+ msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+ if (!msg)
+ return -ENOMEM;
+
+ rc = net_dm_config_fill(msg, info);
+ if (rc)
+ goto free_msg;
+
+ return genlmsg_reply(msg, info);
+
+free_msg:
+ nlmsg_free(msg);
+ return rc;
+}
+
+static void net_dm_stats_read(struct net_dm_stats *stats)
+{
+ int cpu;
+
+ memset(stats, 0, sizeof(*stats));
+ for_each_possible_cpu(cpu) {
+ struct per_cpu_dm_data *data = &per_cpu(dm_cpu_data, cpu);
+ struct net_dm_stats *cpu_stats = &data->stats;
+ unsigned int start;
+ u64 dropped;
+
+ do {
+ start = u64_stats_fetch_begin_irq(&cpu_stats->syncp);
+ dropped = cpu_stats->dropped;
+ } while (u64_stats_fetch_retry_irq(&cpu_stats->syncp, start));
+
+ stats->dropped += dropped;
}
+}
+
+static int net_dm_stats_put(struct sk_buff *msg)
+{
+ struct net_dm_stats stats;
+ struct nlattr *attr;
+
+ net_dm_stats_read(&stats);
+
+ attr = nla_nest_start(msg, NET_DM_ATTR_STATS);
+ if (!attr)
+ return -EMSGSIZE;
+
+ if (nla_put_u64_64bit(msg, NET_DM_ATTR_STATS_DROPPED,
+ stats.dropped, NET_DM_ATTR_PAD))
+ goto nla_put_failure;
+
+ nla_nest_end(msg, attr);
+
+ return 0;
+
+nla_put_failure:
+ nla_nest_cancel(msg, attr);
+ return -EMSGSIZE;
+}
- return -ENOTSUPP;
+static int net_dm_stats_fill(struct sk_buff *msg, struct genl_info *info)
+{
+ void *hdr;
+ int rc;
+
+ hdr = genlmsg_put(msg, info->snd_portid, info->snd_seq,
+ &net_drop_monitor_family, 0, NET_DM_CMD_STATS_NEW);
+ if (!hdr)
+ return -EMSGSIZE;
+
+ rc = net_dm_stats_put(msg);
+ if (rc)
+ goto nla_put_failure;
+
+ genlmsg_end(msg, hdr);
+
+ return 0;
+
+nla_put_failure:
+ genlmsg_cancel(msg, hdr);
+ return -EMSGSIZE;
+}
+
+static int net_dm_cmd_stats_get(struct sk_buff *skb, struct genl_info *info)
+{
+ struct sk_buff *msg;
+ int rc;
+
+ msg = nlmsg_new(NLMSG_DEFAULT_SIZE, GFP_KERNEL);
+ if (!msg)
+ return -ENOMEM;
+
+ rc = net_dm_stats_fill(msg, info);
+ if (rc)
+ goto free_msg;
+
+ return genlmsg_reply(msg, info);
+
+free_msg:
+ nlmsg_free(msg);
+ return rc;
}
static int dropmon_net_event(struct notifier_block *ev_block,
@@ -330,12 +844,12 @@ static int dropmon_net_event(struct notifier_block *ev_block,
new_stat->dev = dev;
new_stat->last_rx = jiffies;
- mutex_lock(&trace_state_mutex);
+ mutex_lock(&net_dm_mutex);
list_add_rcu(&new_stat->list, &hw_stats_list);
- mutex_unlock(&trace_state_mutex);
+ mutex_unlock(&net_dm_mutex);
break;
case NETDEV_UNREGISTER:
- mutex_lock(&trace_state_mutex);
+ mutex_lock(&net_dm_mutex);
list_for_each_entry_safe(new_stat, tmp, &hw_stats_list, list) {
if (new_stat->dev == dev) {
new_stat->dev = NULL;
@@ -346,18 +860,26 @@ static int dropmon_net_event(struct notifier_block *ev_block,
}
}
}
- mutex_unlock(&trace_state_mutex);
+ mutex_unlock(&net_dm_mutex);
break;
}
out:
return NOTIFY_DONE;
}
+static const struct nla_policy net_dm_nl_policy[NET_DM_ATTR_MAX + 1] = {
+ [NET_DM_ATTR_UNSPEC] = { .strict_start_type = NET_DM_ATTR_UNSPEC + 1 },
+ [NET_DM_ATTR_ALERT_MODE] = { .type = NLA_U8 },
+ [NET_DM_ATTR_TRUNC_LEN] = { .type = NLA_U32 },
+ [NET_DM_ATTR_QUEUE_LEN] = { .type = NLA_U32 },
+};
+
static const struct genl_ops dropmon_ops[] = {
{
.cmd = NET_DM_CMD_CONFIG,
.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.doit = net_dm_cmd_config,
+ .flags = GENL_ADMIN_PERM,
},
{
.cmd = NET_DM_CMD_START,
@@ -369,12 +891,38 @@ static const struct genl_ops dropmon_ops[] = {
.validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP,
.doit = net_dm_cmd_trace,
},
+ {
+ .cmd = NET_DM_CMD_CONFIG_GET,
+ .doit = net_dm_cmd_config_get,
+ },
+ {
+ .cmd = NET_DM_CMD_STATS_GET,
+ .doit = net_dm_cmd_stats_get,
+ },
};
+static int net_dm_nl_pre_doit(const struct genl_ops *ops,
+ struct sk_buff *skb, struct genl_info *info)
+{
+ mutex_lock(&net_dm_mutex);
+
+ return 0;
+}
+
+static void net_dm_nl_post_doit(const struct genl_ops *ops,
+ struct sk_buff *skb, struct genl_info *info)
+{
+ mutex_unlock(&net_dm_mutex);
+}
+
static struct genl_family net_drop_monitor_family __ro_after_init = {
.hdrsize = 0,
.name = "NET_DM",
.version = 2,
+ .maxattr = NET_DM_ATTR_MAX,
+ .policy = net_dm_nl_policy,
+ .pre_doit = net_dm_nl_pre_doit,
+ .post_doit = net_dm_nl_post_doit,
.module = THIS_MODULE,
.ops = dropmon_ops,
.n_ops = ARRAY_SIZE(dropmon_ops),
@@ -415,13 +963,11 @@ static int __init init_net_drop_monitor(void)
for_each_possible_cpu(cpu) {
data = &per_cpu(dm_cpu_data, cpu);
- INIT_WORK(&data->dm_alert_work, send_dm_alert);
- timer_setup(&data->send_timer, sched_send_work, 0);
spin_lock_init(&data->lock);
- reset_per_cpu_data(data);
+ skb_queue_head_init(&data->drop_queue);
+ u64_stats_init(&data->stats.syncp);
}
-
goto out;
out_unreg:
@@ -440,19 +986,16 @@ static void exit_net_drop_monitor(void)
/*
* Because of the module_get/put we do in the trace state change path
* we are guarnateed not to have any current users when we get here
- * all we need to do is make sure that we don't have any running timers
- * or pending schedule calls
*/
for_each_possible_cpu(cpu) {
data = &per_cpu(dm_cpu_data, cpu);
- del_timer_sync(&data->send_timer);
- cancel_work_sync(&data->dm_alert_work);
/*
* At this point, we should have exclusive access
* to this struct and can free the skb inside it
*/
kfree_skb(data->skb);
+ WARN_ON(!skb_queue_empty(&data->drop_queue));
}
BUG_ON(genl_unregister_family(&net_drop_monitor_family));
diff --git a/net/core/filter.c b/net/core/filter.c
index 5a2707918629..0c1059cdad3d 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -7539,12 +7539,12 @@ static u32 bpf_convert_ctx_access(enum bpf_access_type type,
case offsetof(struct __sk_buff, gso_segs):
/* si->dst_reg = skb_shinfo(SKB); */
#ifdef NET_SKBUFF_DATA_USES_OFFSET
- *insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, head),
- si->dst_reg, si->src_reg,
- offsetof(struct sk_buff, head));
*insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, end),
BPF_REG_AX, si->src_reg,
offsetof(struct sk_buff, end));
+ *insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, head),
+ si->dst_reg, si->src_reg,
+ offsetof(struct sk_buff, head));
*insn++ = BPF_ALU64_REG(BPF_ADD, si->dst_reg, BPF_REG_AX);
#else
*insn++ = BPF_LDX_MEM(BPF_FIELD_SIZEOF(struct sk_buff, end),
diff --git a/net/core/flow_offload.c b/net/core/flow_offload.c
index d63b970784dc..64c3d4d72b9c 100644
--- a/net/core/flow_offload.c
+++ b/net/core/flow_offload.c
@@ -2,6 +2,8 @@
#include <linux/kernel.h>
#include <linux/slab.h>
#include <net/flow_offload.h>
+#include <linux/rtnetlink.h>
+#include <linux/mutex.h>
struct flow_rule *flow_rule_alloc(unsigned int num_actions)
{
@@ -280,3 +282,241 @@ int flow_block_cb_setup_simple(struct flow_block_offload *f,
}
}
EXPORT_SYMBOL(flow_block_cb_setup_simple);
+
+static LIST_HEAD(block_ing_cb_list);
+
+static struct rhashtable indr_setup_block_ht;
+
+struct flow_indr_block_cb {
+ struct list_head list;
+ void *cb_priv;
+ flow_indr_block_bind_cb_t *cb;
+ void *cb_ident;
+};
+
+struct flow_indr_block_dev {
+ struct rhash_head ht_node;
+ struct net_device *dev;
+ unsigned int refcnt;
+ struct list_head cb_list;
+};
+
+static const struct rhashtable_params flow_indr_setup_block_ht_params = {
+ .key_offset = offsetof(struct flow_indr_block_dev, dev),
+ .head_offset = offsetof(struct flow_indr_block_dev, ht_node),
+ .key_len = sizeof(struct net_device *),
+};
+
+static struct flow_indr_block_dev *
+flow_indr_block_dev_lookup(struct net_device *dev)
+{
+ return rhashtable_lookup_fast(&indr_setup_block_ht, &dev,
+ flow_indr_setup_block_ht_params);
+}
+
+static struct flow_indr_block_dev *
+flow_indr_block_dev_get(struct net_device *dev)
+{
+ struct flow_indr_block_dev *indr_dev;
+
+ indr_dev = flow_indr_block_dev_lookup(dev);
+ if (indr_dev)
+ goto inc_ref;
+
+ indr_dev = kzalloc(sizeof(*indr_dev), GFP_KERNEL);
+ if (!indr_dev)
+ return NULL;
+
+ INIT_LIST_HEAD(&indr_dev->cb_list);
+ indr_dev->dev = dev;
+ if (rhashtable_insert_fast(&indr_setup_block_ht, &indr_dev->ht_node,
+ flow_indr_setup_block_ht_params)) {
+ kfree(indr_dev);
+ return NULL;
+ }
+
+inc_ref:
+ indr_dev->refcnt++;
+ return indr_dev;
+}
+
+static void flow_indr_block_dev_put(struct flow_indr_block_dev *indr_dev)
+{
+ if (--indr_dev->refcnt)
+ return;
+
+ rhashtable_remove_fast(&indr_setup_block_ht, &indr_dev->ht_node,
+ flow_indr_setup_block_ht_params);
+ kfree(indr_dev);
+}
+
+static struct flow_indr_block_cb *
+flow_indr_block_cb_lookup(struct flow_indr_block_dev *indr_dev,
+ flow_indr_block_bind_cb_t *cb, void *cb_ident)
+{
+ struct flow_indr_block_cb *indr_block_cb;
+
+ list_for_each_entry(indr_block_cb, &indr_dev->cb_list, list)
+ if (indr_block_cb->cb == cb &&
+ indr_block_cb->cb_ident == cb_ident)
+ return indr_block_cb;
+ return NULL;
+}
+
+static struct flow_indr_block_cb *
+flow_indr_block_cb_add(struct flow_indr_block_dev *indr_dev, void *cb_priv,
+ flow_indr_block_bind_cb_t *cb, void *cb_ident)
+{
+ struct flow_indr_block_cb *indr_block_cb;
+
+ indr_block_cb = flow_indr_block_cb_lookup(indr_dev, cb, cb_ident);
+ if (indr_block_cb)
+ return ERR_PTR(-EEXIST);
+
+ indr_block_cb = kzalloc(sizeof(*indr_block_cb), GFP_KERNEL);
+ if (!indr_block_cb)
+ return ERR_PTR(-ENOMEM);
+
+ indr_block_cb->cb_priv = cb_priv;
+ indr_block_cb->cb = cb;
+ indr_block_cb->cb_ident = cb_ident;
+ list_add(&indr_block_cb->list, &indr_dev->cb_list);
+
+ return indr_block_cb;
+}
+
+static void flow_indr_block_cb_del(struct flow_indr_block_cb *indr_block_cb)
+{
+ list_del(&indr_block_cb->list);
+ kfree(indr_block_cb);
+}
+
+static void flow_block_ing_cmd(struct net_device *dev,
+ flow_indr_block_bind_cb_t *cb,
+ void *cb_priv,
+ enum flow_block_command command)
+{
+ struct flow_indr_block_ing_entry *entry;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(entry, &block_ing_cb_list, list) {
+ entry->cb(dev, cb, cb_priv, command);
+ }
+ rcu_read_unlock();
+}
+
+int __flow_indr_block_cb_register(struct net_device *dev, void *cb_priv,
+ flow_indr_block_bind_cb_t *cb,
+ void *cb_ident)
+{
+ struct flow_indr_block_cb *indr_block_cb;
+ struct flow_indr_block_dev *indr_dev;
+ int err;
+
+ indr_dev = flow_indr_block_dev_get(dev);
+ if (!indr_dev)
+ return -ENOMEM;
+
+ indr_block_cb = flow_indr_block_cb_add(indr_dev, cb_priv, cb, cb_ident);
+ err = PTR_ERR_OR_ZERO(indr_block_cb);
+ if (err)
+ goto err_dev_put;
+
+ flow_block_ing_cmd(dev, indr_block_cb->cb, indr_block_cb->cb_priv,
+ FLOW_BLOCK_BIND);
+
+ return 0;
+
+err_dev_put:
+ flow_indr_block_dev_put(indr_dev);
+ return err;
+}
+EXPORT_SYMBOL_GPL(__flow_indr_block_cb_register);
+
+int flow_indr_block_cb_register(struct net_device *dev, void *cb_priv,
+ flow_indr_block_bind_cb_t *cb,
+ void *cb_ident)
+{
+ int err;
+
+ rtnl_lock();
+ err = __flow_indr_block_cb_register(dev, cb_priv, cb, cb_ident);
+ rtnl_unlock();
+
+ return err;
+}
+EXPORT_SYMBOL_GPL(flow_indr_block_cb_register);
+
+void __flow_indr_block_cb_unregister(struct net_device *dev,
+ flow_indr_block_bind_cb_t *cb,
+ void *cb_ident)
+{
+ struct flow_indr_block_cb *indr_block_cb;
+ struct flow_indr_block_dev *indr_dev;
+
+ indr_dev = flow_indr_block_dev_lookup(dev);
+ if (!indr_dev)
+ return;
+
+ indr_block_cb = flow_indr_block_cb_lookup(indr_dev, cb, cb_ident);
+ if (!indr_block_cb)
+ return;
+
+ flow_block_ing_cmd(dev, indr_block_cb->cb, indr_block_cb->cb_priv,
+ FLOW_BLOCK_UNBIND);
+
+ flow_indr_block_cb_del(indr_block_cb);
+ flow_indr_block_dev_put(indr_dev);
+}
+EXPORT_SYMBOL_GPL(__flow_indr_block_cb_unregister);
+
+void flow_indr_block_cb_unregister(struct net_device *dev,
+ flow_indr_block_bind_cb_t *cb,
+ void *cb_ident)
+{
+ rtnl_lock();
+ __flow_indr_block_cb_unregister(dev, cb, cb_ident);
+ rtnl_unlock();
+}
+EXPORT_SYMBOL_GPL(flow_indr_block_cb_unregister);
+
+void flow_indr_block_call(struct net_device *dev,
+ struct flow_block_offload *bo,
+ enum flow_block_command command)
+{
+ struct flow_indr_block_cb *indr_block_cb;
+ struct flow_indr_block_dev *indr_dev;
+
+ indr_dev = flow_indr_block_dev_lookup(dev);
+ if (!indr_dev)
+ return;
+
+ list_for_each_entry(indr_block_cb, &indr_dev->cb_list, list)
+ indr_block_cb->cb(dev, indr_block_cb->cb_priv, TC_SETUP_BLOCK,
+ bo);
+}
+EXPORT_SYMBOL_GPL(flow_indr_block_call);
+
+static DEFINE_MUTEX(flow_indr_block_ing_cb_lock);
+void flow_indr_add_block_ing_cb(struct flow_indr_block_ing_entry *entry)
+{
+ mutex_lock(&flow_indr_block_ing_cb_lock);
+ list_add_tail_rcu(&entry->list, &block_ing_cb_list);
+ mutex_unlock(&flow_indr_block_ing_cb_lock);
+}
+EXPORT_SYMBOL_GPL(flow_indr_add_block_ing_cb);
+
+void flow_indr_del_block_ing_cb(struct flow_indr_block_ing_entry *entry)
+{
+ mutex_lock(&flow_indr_block_ing_cb_lock);
+ list_del_rcu(&entry->list);
+ mutex_unlock(&flow_indr_block_ing_cb_lock);
+}
+EXPORT_SYMBOL_GPL(flow_indr_del_block_ing_cb);
+
+static int __init init_flow_indr_rhashtable(void)
+{
+ return rhashtable_init(&indr_setup_block_ht,
+ &flow_indr_setup_block_ht_params);
+}
+subsys_initcall(init_flow_indr_rhashtable);
diff --git a/net/core/neighbour.c b/net/core/neighbour.c
index f79e61c570ea..5480edff0c86 100644
--- a/net/core/neighbour.c
+++ b/net/core/neighbour.c
@@ -3033,7 +3033,7 @@ static struct neighbour *neigh_get_first(struct seq_file *seq)
struct net *net = seq_file_net(seq);
struct neigh_hash_table *nht = state->nht;
struct neighbour *n = NULL;
- int bucket = state->bucket;
+ int bucket;
state->flags &= ~NEIGH_SEQ_IS_PNEIGH;
for (bucket = 0; bucket < (1 << nht->hash_shift); bucket++) {
diff --git a/net/core/pktgen.c b/net/core/pktgen.c
index bb9915291644..c5dbdc87342a 100644
--- a/net/core/pktgen.c
+++ b/net/core/pktgen.c
@@ -2652,7 +2652,7 @@ static void pktgen_finalize_skb(struct pktgen_dev *pkt_dev, struct sk_buff *skb,
}
get_page(pkt_dev->page);
skb_frag_set_page(skb, i, pkt_dev->page);
- skb_shinfo(skb)->frags[i].page_offset = 0;
+ skb_frag_off_set(&skb_shinfo(skb)->frags[i], 0);
/*last fragment, fill rest of data*/
if (i == (frags - 1))
skb_frag_size_set(&skb_shinfo(skb)->frags[i],
diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 0b788df5a75b..ea8e8d332d85 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -785,7 +785,7 @@ void skb_dump(const char *level, const struct sk_buff *skb, bool full_pkt)
struct page *p;
u8 *vaddr;
- skb_frag_foreach_page(frag, frag->page_offset,
+ skb_frag_foreach_page(frag, skb_frag_off(frag),
skb_frag_size(frag), p, p_off, p_len,
copied) {
seg_len = min_t(int, p_len, len);
@@ -1375,7 +1375,7 @@ int skb_copy_ubufs(struct sk_buff *skb, gfp_t gfp_mask)
struct page *p;
u8 *vaddr;
- skb_frag_foreach_page(f, f->page_offset, skb_frag_size(f),
+ skb_frag_foreach_page(f, skb_frag_off(f), skb_frag_size(f),
p, p_off, p_len, copied) {
u32 copy, done = 0;
vaddr = kmap_atomic(p);
@@ -2144,10 +2144,12 @@ pull_pages:
skb_frag_unref(skb, i);
eat -= size;
} else {
- skb_shinfo(skb)->frags[k] = skb_shinfo(skb)->frags[i];
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[k];
+
+ *frag = skb_shinfo(skb)->frags[i];
if (eat) {
- skb_shinfo(skb)->frags[k].page_offset += eat;
- skb_frag_size_sub(&skb_shinfo(skb)->frags[k], eat);
+ skb_frag_off_add(frag, eat);
+ skb_frag_size_sub(frag, eat);
if (!i)
goto end;
eat = 0;
@@ -2219,7 +2221,7 @@ int skb_copy_bits(const struct sk_buff *skb, int offset, void *to, int len)
copy = len;
skb_frag_foreach_page(f,
- f->page_offset + offset - start,
+ skb_frag_off(f) + offset - start,
copy, p, p_off, p_len, copied) {
vaddr = kmap_atomic(p);
memcpy(to + copied, vaddr + p_off, p_len);
@@ -2395,7 +2397,7 @@ static bool __skb_splice_bits(struct sk_buff *skb, struct pipe_inode_info *pipe,
const skb_frag_t *f = &skb_shinfo(skb)->frags[seg];
if (__splice_segment(skb_frag_page(f),
- f->page_offset, skb_frag_size(f),
+ skb_frag_off(f), skb_frag_size(f),
offset, len, spd, false, sk, pipe))
return true;
}
@@ -2498,7 +2500,7 @@ do_frag_list:
while (slen) {
ret = kernel_sendpage_locked(sk, skb_frag_page(frag),
- frag->page_offset + offset,
+ skb_frag_off(frag) + offset,
slen, MSG_DONTWAIT);
if (ret <= 0)
goto error;
@@ -2580,7 +2582,7 @@ int skb_store_bits(struct sk_buff *skb, int offset, const void *from, int len)
copy = len;
skb_frag_foreach_page(frag,
- frag->page_offset + offset - start,
+ skb_frag_off(frag) + offset - start,
copy, p, p_off, p_len, copied) {
vaddr = kmap_atomic(p);
memcpy(vaddr + p_off, from + copied, p_len);
@@ -2660,7 +2662,7 @@ __wsum __skb_checksum(const struct sk_buff *skb, int offset, int len,
copy = len;
skb_frag_foreach_page(frag,
- frag->page_offset + offset - start,
+ skb_frag_off(frag) + offset - start,
copy, p, p_off, p_len, copied) {
vaddr = kmap_atomic(p);
csum2 = INDIRECT_CALL_1(ops->update,
@@ -2759,7 +2761,7 @@ __wsum skb_copy_and_csum_bits(const struct sk_buff *skb, int offset,
copy = len;
skb_frag_foreach_page(frag,
- frag->page_offset + offset - start,
+ skb_frag_off(frag) + offset - start,
copy, p, p_off, p_len, copied) {
vaddr = kmap_atomic(p);
csum2 = csum_partial_copy_nocheck(vaddr + p_off,
@@ -3234,7 +3236,7 @@ static inline void skb_split_no_header(struct sk_buff *skb,
* 2. Split is accurately. We make this.
*/
skb_frag_ref(skb, i);
- skb_shinfo(skb1)->frags[0].page_offset += len - pos;
+ skb_frag_off_add(&skb_shinfo(skb1)->frags[0], len - pos);
skb_frag_size_sub(&skb_shinfo(skb1)->frags[0], len - pos);
skb_frag_size_set(&skb_shinfo(skb)->frags[i], len - pos);
skb_shinfo(skb)->nr_frags++;
@@ -3316,7 +3318,7 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen)
*/
if (!to ||
!skb_can_coalesce(tgt, to, skb_frag_page(fragfrom),
- fragfrom->page_offset)) {
+ skb_frag_off(fragfrom))) {
merge = -1;
} else {
merge = to - 1;
@@ -3333,7 +3335,7 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen)
skb_frag_size_add(fragto, shiftlen);
skb_frag_size_sub(fragfrom, shiftlen);
- fragfrom->page_offset += shiftlen;
+ skb_frag_off_add(fragfrom, shiftlen);
goto onlymerged;
}
@@ -3364,11 +3366,11 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen)
} else {
__skb_frag_ref(fragfrom);
- fragto->bv_page = fragfrom->bv_page;
- fragto->page_offset = fragfrom->page_offset;
+ skb_frag_page_copy(fragto, fragfrom);
+ skb_frag_off_copy(fragto, fragfrom);
skb_frag_size_set(fragto, todo);
- fragfrom->page_offset += todo;
+ skb_frag_off_add(fragfrom, todo);
skb_frag_size_sub(fragfrom, todo);
todo = 0;
@@ -3493,7 +3495,7 @@ next_skb:
if (!st->frag_data)
st->frag_data = kmap_atomic(skb_frag_page(frag));
- *data = (u8 *) st->frag_data + frag->page_offset +
+ *data = (u8 *) st->frag_data + skb_frag_off(frag) +
(abs_offset - st->stepped_offset);
return block_limit - abs_offset;
@@ -3630,8 +3632,8 @@ static inline skb_frag_t skb_head_frag_to_page_desc(struct sk_buff *frag_skb)
page = virt_to_head_page(frag_skb->head);
__skb_frag_set_page(&head_frag, page);
- head_frag.page_offset = frag_skb->data -
- (unsigned char *)page_address(page);
+ skb_frag_off_set(&head_frag, frag_skb->data -
+ (unsigned char *)page_address(page));
skb_frag_size_set(&head_frag, skb_headlen(frag_skb));
return head_frag;
}
@@ -3875,7 +3877,7 @@ normal:
size = skb_frag_size(nskb_frag);
if (pos < offset) {
- nskb_frag->page_offset += offset - pos;
+ skb_frag_off_add(nskb_frag, offset - pos);
skb_frag_size_sub(nskb_frag, offset - pos);
}
@@ -3996,7 +3998,7 @@ int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb)
*--frag = *--frag2;
} while (--i);
- frag->page_offset += offset;
+ skb_frag_off_add(frag, offset);
skb_frag_size_sub(frag, offset);
/* all fragments truesize : remove (head size + sk_buff) */
@@ -4026,7 +4028,7 @@ int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb)
pinfo->nr_frags = nr_frags + 1 + skbinfo->nr_frags;
__skb_frag_set_page(frag, page);
- frag->page_offset = first_offset;
+ skb_frag_off_set(frag, first_offset);
skb_frag_size_set(frag, first_size);
memcpy(frag + 1, skbinfo->frags, sizeof(*frag) * skbinfo->nr_frags);
@@ -4042,7 +4044,7 @@ merge:
if (offset > headlen) {
unsigned int eat = offset - headlen;
- skbinfo->frags[0].page_offset += eat;
+ skb_frag_off_add(&skbinfo->frags[0], eat);
skb_frag_size_sub(&skbinfo->frags[0], eat);
skb->data_len -= eat;
skb->len -= eat;
@@ -4167,7 +4169,7 @@ __skb_to_sgvec(struct sk_buff *skb, struct scatterlist *sg, int offset, int len,
if (copy > len)
copy = len;
sg_set_page(&sg[elt], skb_frag_page(frag), copy,
- frag->page_offset+offset-start);
+ skb_frag_off(frag) + offset - start);
elt++;
if (!(len -= copy))
return elt;
@@ -5838,7 +5840,7 @@ static int pskb_carve_inside_nonlinear(struct sk_buff *skb, const u32 off,
* where splitting is expensive.
* 2. Split is accurately. We make this.
*/
- shinfo->frags[0].page_offset += off - pos;
+ skb_frag_off_add(&shinfo->frags[0], off - pos);
skb_frag_size_sub(&shinfo->frags[0], off - pos);
}
skb_frag_ref(skb, i);
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index 93bffaad2135..6832eeb4b785 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -585,12 +585,12 @@ EXPORT_SYMBOL_GPL(sk_psock_destroy);
void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
{
- rcu_assign_sk_user_data(sk, NULL);
sk_psock_cork_free(psock);
sk_psock_zap_ingress(psock);
- sk_psock_restore_proto(sk, psock);
write_lock_bh(&sk->sk_callback_lock);
+ sk_psock_restore_proto(sk, psock);
+ rcu_assign_sk_user_data(sk, NULL);
if (psock->progs.skb_parser)
sk_psock_stop_strp(sk, psock);
write_unlock_bh(&sk->sk_callback_lock);
diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 52d4faeee18b..1330a7442e5b 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -247,6 +247,8 @@ static void sock_map_free(struct bpf_map *map)
raw_spin_unlock_bh(&stab->lock);
rcu_read_unlock();
+ synchronize_rcu();
+
bpf_map_area_free(stab->sks);
kfree(stab);
}
@@ -276,16 +278,20 @@ static int __sock_map_delete(struct bpf_stab *stab, struct sock *sk_test,
struct sock **psk)
{
struct sock *sk;
+ int err = 0;
raw_spin_lock_bh(&stab->lock);
sk = *psk;
if (!sk_test || sk_test == sk)
- *psk = NULL;
+ sk = xchg(psk, NULL);
+
+ if (likely(sk))
+ sock_map_unref(sk, psk);
+ else
+ err = -EINVAL;
+
raw_spin_unlock_bh(&stab->lock);
- if (unlikely(!sk))
- return -EINVAL;
- sock_map_unref(sk, psk);
- return 0;
+ return err;
}
static void sock_map_delete_from_link(struct bpf_map *map, struct sock *sk,
@@ -328,6 +334,7 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
struct sock *sk, u64 flags)
{
struct bpf_stab *stab = container_of(map, struct bpf_stab, map);
+ struct inet_connection_sock *icsk = inet_csk(sk);
struct sk_psock_link *link;
struct sk_psock *psock;
struct sock *osk;
@@ -338,6 +345,8 @@ static int sock_map_update_common(struct bpf_map *map, u32 idx,
return -EINVAL;
if (unlikely(idx >= map->max_entries))
return -E2BIG;
+ if (unlikely(icsk->icsk_ulp_data))
+ return -EINVAL;
link = sk_psock_init_link();
if (!link)
diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c
index 8da5b3a54dac..eb29e5adc84d 100644
--- a/net/core/sysctl_net_core.c
+++ b/net/core/sysctl_net_core.c
@@ -567,6 +567,14 @@ static struct ctl_table net_core_table[] = {
.mode = 0644,
.proc_handler = proc_do_static_key,
},
+ {
+ .procname = "gro_normal_batch",
+ .data = &gro_normal_batch,
+ .maxlen = sizeof(unsigned int),
+ .mode = 0644,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = SYSCTL_ONE,
+ },
{ }
};
generated by cgit v1.2.3 (git 2.46.0) at 2025-09-19 17:42:34 +0000