kernel/linux.git/include/net/sock.h, branch v5.10.86

tcp: fix page frag corruption on page fault

2021-12-08T08:03:23+00:00

commit dacb5d8875cc6cd3a553363b4d6f06760fcbe70c upstream. Steffen reported a TCP stream corruption for HTTP requests served by the apache web-server using a cifs mount-point and memory mapping the relevant file. The root cause is quite similar to the one addressed by commit 20eb4f29b602 ("net: fix sk_page_frag() recursion from memory reclaim"). Here the nested access to the task page frag is caused by a page fault on the (mmapped) user-space memory buffer coming from the cifs file. The page fault handler performs an smb transaction on a different socket, inside the same process context. Since sk->sk_allaction for such socket does not prevent the usage for the task_frag, the nested allocation modify "under the hood" the page frag in use by the outer sendmsg call, corrupting the stream. The overall relevant stack trace looks like the following: httpd 78268 [001] 3461630.850950: probe:tcp_sendmsg_locked: ffffffff91461d91 tcp_sendmsg_locked+0x1 ffffffff91462b57 tcp_sendmsg+0x27 ffffffff9139814e sock_sendmsg+0x3e ffffffffc06dfe1d smb_send_kvec+0x28 [...] ffffffffc06cfaf8 cifs_readpages+0x213 ffffffff90e83c4b read_pages+0x6b ffffffff90e83f31 __do_page_cache_readahead+0x1c1 ffffffff90e79e98 filemap_fault+0x788 ffffffff90eb0458 __do_fault+0x38 ffffffff90eb5280 do_fault+0x1a0 ffffffff90eb7c84 __handle_mm_fault+0x4d4 ffffffff90eb8093 handle_mm_fault+0xc3 ffffffff90c74f6d __do_page_fault+0x1ed ffffffff90c75277 do_page_fault+0x37 ffffffff9160111e page_fault+0x1e ffffffff9109e7b5 copyin+0x25 ffffffff9109eb40 _copy_from_iter_full+0xe0 ffffffff91462370 tcp_sendmsg_locked+0x5e0 ffffffff91462370 tcp_sendmsg_locked+0x5e0 ffffffff91462b57 tcp_sendmsg+0x27 ffffffff9139815c sock_sendmsg+0x4c ffffffff913981f7 sock_write_iter+0x97 ffffffff90f2cc56 do_iter_readv_writev+0x156 ffffffff90f2dff0 do_iter_write+0x80 ffffffff90f2e1c3 vfs_writev+0xa3 ffffffff90f2e27c do_writev+0x5c ffffffff90c042bb do_syscall_64+0x5b ffffffff916000ad entry_SYSCALL_64_after_hwframe+0x65 The cifs filesystem rightfully sets sk_allocations to GFP_NOFS, we can avoid the nesting using the sk page frag for allocation lacking the __GFP_FS flag. Do not define an additional mm-helper for that, as this is strictly tied to the sk page frag usage. v1 -> v2: - use a stricted sk_page_frag() check instead of reordering the code (Eric) Reported-by: Steffen Froemer Fixes: 5640f7685831 ("net: use a per task frag allocator") Signed-off-by: Paolo Abeni Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

tcp: switch orphan_count to bare per-cpu counters

2021-11-18T13:04:08+00:00

[ Upstream commit 19757cebf0c5016a1f36f7fe9810a9f0b33c0832 ] Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet Reported-by: Stefan Bach Cc: Neal Cardwell Acked-by: Neal Cardwell Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

af_unix: fix races in sk_peer_pid and sk_peer_cred accesses

2021-10-06T13:55:58+00:00

[ Upstream commit 35306eb23814444bd4021f8a1c3047d3cb0c8b2b ] Jann Horn reported that SO_PEERCRED and SO_PEERGROUPS implementations are racy, as af_unix can concurrently change sk_peer_pid and sk_peer_cred. In order to fix this issue, this patch adds a new spinlock that needs to be used whenever these fields are read or written. Jann also pointed out that l2cap_sock_get_peer_pid_cb() is currently reading sk->sk_peer_pid which makes no sense, as this field is only possibly set by AF_UNIX sockets. We will have to clean this in a separate patch. This could be done by reverting b48596d1dc25 "Bluetooth: L2CAP: Add get_peer_pid callback" or implementing what was truly expected. Fixes: 109f6e39fa07 ("af_unix: Allow SO_PEERCRED to work across namespaces.") Signed-off-by: Eric Dumazet Reported-by: Jann Horn Cc: Eric W. Biederman Cc: Luiz Augusto von Dentz Cc: Marcel Holtmann Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

inet: annotate date races around sk->sk_txhash

2021-06-30T12:47:21+00:00

[ Upstream commit b71eaed8c04f72a919a9c44e83e4ee254e69e7f3 ] UDP sendmsg() path can be lockless, it is possible for another thread to re-connect an change sk->sk_txhash under us. There is no serious impact, but we can use READ_ONCE()/WRITE_ONCE() pair to document the race. BUG: KCSAN: data-race in __ip4_datagram_connect / skb_set_owner_w write to 0xffff88813397920c of 4 bytes by task 30997 on cpu 1: sk_set_txhash include/net/sock.h:1937 [inline] __ip4_datagram_connect+0x69e/0x710 net/ipv4/datagram.c:75 __ip6_datagram_connect+0x551/0x840 net/ipv6/datagram.c:189 ip6_datagram_connect+0x2a/0x40 net/ipv6/datagram.c:272 inet_dgram_connect+0xfd/0x180 net/ipv4/af_inet.c:580 __sys_connect_file net/socket.c:1837 [inline] __sys_connect+0x245/0x280 net/socket.c:1854 __do_sys_connect net/socket.c:1864 [inline] __se_sys_connect net/socket.c:1861 [inline] __x64_sys_connect+0x3d/0x50 net/socket.c:1861 do_syscall_64+0x4a/0x90 arch/x86/entry/common.c:47 entry_SYSCALL_64_after_hwframe+0x44/0xae read to 0xffff88813397920c of 4 bytes by task 31039 on cpu 0: skb_set_hash_from_sk include/net/sock.h:2211 [inline] skb_set_owner_w+0x118/0x220 net/core/sock.c:2101 sock_alloc_send_pskb+0x452/0x4e0 net/core/sock.c:2359 sock_alloc_send_skb+0x2d/0x40 net/core/sock.c:2373 __ip6_append_data+0x1743/0x21a0 net/ipv6/ip6_output.c:1621 ip6_make_skb+0x258/0x420 net/ipv6/ip6_output.c:1983 udpv6_sendmsg+0x160a/0x16b0 net/ipv6/udp.c:1527 inet6_sendmsg+0x5f/0x80 net/ipv6/af_inet6.c:642 sock_sendmsg_nosec net/socket.c:654 [inline] sock_sendmsg net/socket.c:674 [inline] ____sys_sendmsg+0x360/0x4d0 net/socket.c:2350 ___sys_sendmsg net/socket.c:2404 [inline] __sys_sendmmsg+0x315/0x4b0 net/socket.c:2490 __do_sys_sendmmsg net/socket.c:2519 [inline] __se_sys_sendmmsg net/socket.c:2516 [inline] __x64_sys_sendmmsg+0x53/0x60 net/socket.c:2516 do_syscall_64+0x4a/0x90 arch/x86/entry/common.c:47 entry_SYSCALL_64_after_hwframe+0x44/0xae value changed: 0xbca3c43d -> 0xfdb309e0 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 31039 Comm: syz-executor.2 Not tainted 5.13.0-rc3-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Signed-off-by: Eric Dumazet Reported-by: syzbot Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

net: annotate data race in sock_error()

2021-06-30T12:47:21+00:00

[ Upstream commit f13ef10059ccf5f4ed201cd050176df62ec25bb8 ] sock_error() is known to be racy. The code avoids an atomic operation is sk_err is zero, and this field could be changed under us, this is fine. Sysbot reported: BUG: KCSAN: data-race in sock_alloc_send_pskb / unix_release_sock write to 0xffff888131855630 of 4 bytes by task 9365 on cpu 1: unix_release_sock+0x2e9/0x6e0 net/unix/af_unix.c:550 unix_release+0x2f/0x50 net/unix/af_unix.c:859 __sock_release net/socket.c:599 [inline] sock_close+0x6c/0x150 net/socket.c:1258 __fput+0x25b/0x4e0 fs/file_table.c:280 ____fput+0x11/0x20 fs/file_table.c:313 task_work_run+0xae/0x130 kernel/task_work.c:164 tracehook_notify_resume include/linux/tracehook.h:189 [inline] exit_to_user_mode_loop kernel/entry/common.c:174 [inline] exit_to_user_mode_prepare+0x156/0x190 kernel/entry/common.c:208 __syscall_exit_to_user_mode_work kernel/entry/common.c:290 [inline] syscall_exit_to_user_mode+0x20/0x40 kernel/entry/common.c:301 do_syscall_64+0x56/0x90 arch/x86/entry/common.c:57 entry_SYSCALL_64_after_hwframe+0x44/0xae read to 0xffff888131855630 of 4 bytes by task 9385 on cpu 0: sock_error include/net/sock.h:2269 [inline] sock_alloc_send_pskb+0xe4/0x4e0 net/core/sock.c:2336 unix_dgram_sendmsg+0x478/0x1610 net/unix/af_unix.c:1671 unix_seqpacket_sendmsg+0xc2/0x100 net/unix/af_unix.c:2055 sock_sendmsg_nosec net/socket.c:654 [inline] sock_sendmsg net/socket.c:674 [inline] ____sys_sendmsg+0x360/0x4d0 net/socket.c:2350 __sys_sendmsg_sock+0x25/0x30 net/socket.c:2416 io_sendmsg fs/io_uring.c:4367 [inline] io_issue_sqe+0x231a/0x6750 fs/io_uring.c:6135 __io_queue_sqe+0xe9/0x360 fs/io_uring.c:6414 __io_req_task_submit fs/io_uring.c:2039 [inline] io_async_task_func+0x312/0x590 fs/io_uring.c:5074 __tctx_task_work fs/io_uring.c:1910 [inline] tctx_task_work+0x1d4/0x3d0 fs/io_uring.c:1924 task_work_run+0xae/0x130 kernel/task_work.c:164 tracehook_notify_signal include/linux/tracehook.h:212 [inline] handle_signal_work kernel/entry/common.c:145 [inline] exit_to_user_mode_loop kernel/entry/common.c:171 [inline] exit_to_user_mode_prepare+0xf8/0x190 kernel/entry/common.c:208 __syscall_exit_to_user_mode_work kernel/entry/common.c:290 [inline] syscall_exit_to_user_mode+0x20/0x40 kernel/entry/common.c:301 do_syscall_64+0x56/0x90 arch/x86/entry/common.c:57 entry_SYSCALL_64_after_hwframe+0x44/0xae value changed: 0x00000000 -> 0x00000068 Reported by Kernel Concurrency Sanitizer on: CPU: 0 PID: 9385 Comm: syz-executor.3 Not tainted 5.13.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Signed-off-by: Eric Dumazet Reported-by: syzbot Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

net: really orphan skbs tied to closing sk

2021-06-03T07:00:46+00:00

[ Upstream commit 098116e7e640ba677d9e345cbee83d253c13d556 ] If the owing socket is shutting down - e.g. the sock reference count already dropped to 0 and only sk_wmem_alloc is keeping the sock alive, skb_orphan_partial() becomes a no-op. When forwarding packets over veth with GRO enabled, the above causes refcount errors. This change addresses the issue with a plain skb_orphan() call in the critical scenario. Fixes: 9adc89af724f ("net: let skb_orphan_partial wake-up waiters.") Signed-off-by: Paolo Abeni Signed-off-by: David S. Miller Signed-off-by: Sasha Levin

net: let skb_orphan_partial wake-up waiters.

2021-04-14T06:42:03+00:00

commit 9adc89af724f12a03b47099cd943ed54e877cd59 upstream. Currently the mentioned helper can end-up freeing the socket wmem without waking-up any processes waiting for more write memory. If the partially orphaned skb is attached to an UDP (or raw) socket, the lack of wake-up can hang the user-space. Even for TCP sockets not calling the sk destructor could have bad effects on TSQ. Address the issue using skb_orphan to release the sk wmem before setting the new sock_efree destructor. Additionally bundle the whole ownership update in a new helper, so that later other potential users could avoid duplicate code. v1 -> v2: - use skb_orphan() instead of sort of open coding it (Eric) - provide an helper for the ownership change (Eric) Fixes: f6ba8d33cfbb ("netem: fix skb_orphan_partial()") Suggested-by: Eric Dumazet Signed-off-by: Paolo Abeni Reviewed-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman

tcp: fix TCP socket rehash stats mis-accounting

2021-01-27T10:55:23+00:00

commit 9c30ae8398b0813e237bde387d67a7f74ab2db2d upstream. The previous commit 32efcc06d2a1 ("tcp: export count for rehash attempts") would mis-account rehashing SNMP and socket stats: a. During handshake of an active open, only counts the first SYN timeout b. After handshake of passive and active open, stop updating after (roughly) TCP_RETRIES1 recurring RTOs c. After the socket aborts, over count timeout_rehash by 1 This patch fixes this by checking the rehash result from sk_rethink_txhash. Fixes: 32efcc06d2a1 ("tcp: export count for rehash attempts") Signed-off-by: Yuchung Cheng Signed-off-by: Eric Dumazet Signed-off-by: Neal Cardwell Link: https://lore.kernel.org/r/20210119192619.1848270-1-ycheng@google.com Signed-off-by: Jakub Kicinski Signed-off-by: Greg Kroah-Hartman

mptcp: add sk_stop_timer_sync helper

2020-09-25T02:58:34+00:00

This patch added a new helper sk_stop_timer_sync, it deactivates a timer like sk_stop_timer, but waits for the handler to finish. Acked-by: Paolo Abeni Signed-off-by: Geliang Tang Reviewed-by: Mat Martineau Signed-off-by: David S. Miller

tcp: remove SOCK_QUEUE_SHRUNK

2020-09-14T20:36:00+00:00

SOCK_QUEUE_SHRUNK is currently used by TCP as a temporary state that remembers if some room has been made in the rtx queue by an incoming ACK packet. This is later used from tcp_check_space() before considering to send EPOLLOUT. Problem is: If we receive SACK packets, and no packet is removed from RTX queue, we can send fresh packets, thus moving them from write queue to rtx queue and eventually empty the write queue. This stall can happen if TCP_NOTSENT_LOWAT is used. With this fix, we no longer risk stalling sends while holes are repaired, and we can fully use socket sndbuf. This also removes a cache line dirtying for typical RPC workloads. Fixes: c9bee3b7fdec ("tcp: TCP_NOTSENT_LOWAT socket option") Signed-off-by: Eric Dumazet Cc: Soheil Hassas Yeganeh Acked-by: Neal Cardwell Acked-by: Soheil Hassas Yeganeh Signed-off-by: David S. Miller