linux/net/core
Xuan Zhuo 3765996e4f napi: fix race inside napi_enable
The process will cause napi.state to contain NAPI_STATE_SCHED and
not in the poll_list, which will cause napi_disable() to get stuck.

The prefix "NAPI_STATE_" is removed in the figure below, and
NAPI_STATE_HASHED is ignored in napi.state.

                      CPU0       |                   CPU1       | napi.state
===============================================================================
napi_disable()                   |                              | SCHED | NPSVC
napi_enable()                    |                              |
{                                |                              |
    smp_mb__before_atomic();     |                              |
    clear_bit(SCHED, &n->state); |                              | NPSVC
                                 | napi_schedule_prep()         | SCHED | NPSVC
                                 | napi_poll()                  |
                                 |   napi_complete_done()       |
                                 |   {                          |
                                 |      if (n->state & (NPSVC | | (1)
                                 |               _BUSY_POLL)))  |
                                 |           return false;      |
                                 |     ................         |
                                 |   }                          | SCHED | NPSVC
                                 |                              |
    clear_bit(NPSVC, &n->state); |                              | SCHED
}                                |                              |
                                 |                              |
napi_schedule_prep()             |                              | SCHED | MISSED (2)

(1) Here return direct. Because of NAPI_STATE_NPSVC exists.
(2) NAPI_STATE_SCHED exists. So not add napi.poll_list to sd->poll_list

Since NAPI_STATE_SCHED already exists and napi is not in the
sd->poll_list queue, NAPI_STATE_SCHED cannot be cleared and will always
exist.

1. This will cause this queue to no longer receive packets.
2. If you encounter napi_disable under the protection of rtnl_lock, it
   will cause the entire rtnl_lock to be locked, affecting the overall
   system.

This patch uses cmpxchg to implement napi_enable(), which ensures that
there will be no race due to the separation of clear two bits.

Fixes: 2d8bff1269 ("netpoll: Close race condition between poll_one_napi and napi_disable")
Signed-off-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Reviewed-by: Dust Li <dust.li@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2021-09-20 09:41:29 +01:00
..
bpf_sk_storage.c net: in_irq() cleanup 2021-08-13 14:09:19 -07:00
datagram.c udp: fix skb_copy_and_csum_datagram with odd segment sizes 2021-02-04 18:56:56 -08:00
datagram.h
dev_addr_lists.c net-next: When a bond have a massive amount of VLANs with IPv6 addresses, performance of changing link state, attaching a VRF, changing an IPv6 address, etc. go down dramtically. 2021-08-25 10:29:07 +01:00
dev_ioctl.c net: core: don't call SIOCBRADD/DELIF for non-bridge devices 2021-08-05 11:36:59 +01:00
dev.c napi: fix race inside napi_enable 2021-09-20 09:41:29 +01:00
devlink.c devlink: Clear whole devlink_flash_notify struct 2021-08-14 13:59:10 +01:00
drop_monitor.c net: Remove redundant if statements 2021-08-05 13:27:50 +01:00
dst_cache.c treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 152 2019-05-30 11:26:32 -07:00
dst.c net: Remove redundant if statements 2021-08-05 13:27:50 +01:00
failover.c
fib_notifier.c net: fib_notifier: propagate extack down to the notifier block callback 2019-10-04 11:10:56 -07:00
fib_rules.c memcg: enable accounting for IP address and routing-related objects 2021-07-20 06:00:38 -07:00
filter.c bpf: tcp: Allow bpf-tcp-cc to call bpf_(get|set)sockopt 2021-08-25 17:40:35 -07:00
flow_dissector.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2021-07-31 09:14:46 -07:00
flow_offload.c net: Fix offloading indirect devices dependency on qdisc order creation 2021-08-19 13:19:30 +01:00
gen_estimator.c net_sched: gen_estimator: support large ewma log 2021-01-15 18:11:06 -08:00
gen_stats.c docs: networking: convert gen_stats.txt to ReST 2020-04-28 14:39:46 -07:00
gro_cells.c gro_cells: reduce number of synchronize_net() calls 2020-11-25 11:28:12 -08:00
hwbm.c net: hwbm: Make the hwbm_pool lock a mutex 2019-06-09 19:40:10 -07:00
link_watch.c net: linkwatch: fix failure to restore device state across suspend/resume 2021-08-11 14:43:16 -07:00
lwt_bpf.c lwt_bpf: Replace preempt_disable() with migrate_disable() 2020-12-07 11:53:40 -08:00
lwtunnel.c netfilter: add netfilter hooks to SRv6 data plane 2021-08-30 01:51:36 +02:00
Makefile sock_map: Relax config dependency to CONFIG_NET 2021-07-15 18:17:49 -07:00
neighbour.c net: Support filtering interfaces on no master 2021-08-10 16:03:34 -07:00
net_namespace.c net: net_namespace: Optimize the code 2021-08-18 10:34:48 +01:00
net-procfs.c net: procfs: add seq_puts() statement for dev_mcast 2021-08-18 10:13:20 +01:00
net-sysfs.c net-sysfs: remove possible sleep from an RCU read-side critical section 2021-03-22 13:28:13 -07:00
net-sysfs.h net-sysfs: add netdev_change_owner() 2020-02-26 20:07:25 -08:00
net-traces.c tcp: add tracepoint for checksum errors 2021-05-14 15:26:03 -07:00
netclassid_cgroup.c bpf, cgroups: Fix cgroup v2 fallback on v1/v2 mixed mode 2021-09-13 16:35:58 -07:00
netevent.c net: core: Correct function name netevent_unregister_notifier() in the kerneldoc 2021-03-28 17:56:56 -07:00
netpoll.c asm-generic/unaligned: Unify asm/unaligned.h around struct helper 2021-07-02 12:43:40 -07:00
netprio_cgroup.c bpf, cgroups: Fix cgroup v2 fallback on v1/v2 mixed mode 2021-09-13 16:35:58 -07:00
page_pool.c page_pool: use relaxed atomic for release side accounting 2021-08-24 10:46:31 +01:00
pktgen.c pktgen: remove unused variable 2021-09-03 11:48:28 +01:00
ptp_classifier.c bpf: Refactor BPF_PROG_RUN into a function 2021-08-17 00:45:07 +02:00
request_sock.c tcp: add rcu protection around tp->fastopen_rsk 2019-10-13 10:13:08 -07:00
rtnetlink.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net 2021-08-26 17:57:57 -07:00
scm.c memcg: enable accounting for scm_fp_list objects 2021-07-20 06:00:38 -07:00
secure_seq.c crypto: lib/sha1 - remove unnecessary includes of linux/cryptohash.h 2020-05-08 15:32:17 +10:00
selftests.c net: selftests: add MTU test 2021-07-22 00:52:04 -07:00
skbuff.c skbuff: clean up inconsistent indenting 2021-09-03 11:51:26 +01:00
skmsg.c bpf, sockmap: Fix memleak on ingress msg enqueue 2021-07-27 14:55:30 -07:00
sock_diag.c bpf, net: Rework cookie generator as per-cpu one 2020-09-30 11:50:35 -07:00
sock_map.c af_unix: Add unix_stream_proto for sockmap 2021-08-16 18:43:39 -07:00
sock_reuseport.c tcp: Add stats for socket migration. 2021-06-23 12:56:08 -07:00
sock.c net: core: Correct the sock::sk_lock.owned lockdep annotations 2021-09-19 12:48:06 +01:00
stream.c tcp: make sure EPOLLOUT wont be missed 2019-08-19 13:07:43 -07:00
sysctl_net_core.c net: change netdev_unregister_timeout_secs min value to 1 2021-03-25 17:24:06 -07:00
timestamping.c net: Introduce a new MII time stamping interface. 2019-12-25 19:51:33 -08:00
tso.c net: tso: add UDP segmentation support 2020-06-18 20:46:23 -07:00
utils.c net: Fix skb->csum update in inet_proto_csum_replace16(). 2020-01-24 20:54:30 +01:00
xdp.c xdp: Move the rxq_info.mem clearing to unreg_mem_model() 2021-06-28 23:07:59 +02:00