possible deadlock in sctp_addr_wq_timeout_handler

Status: fixed on 2021/11/10 00:50
Subsystems: sctp
[Documentation on labels]
Reported-by: syzbot+959223586843e69a2674@syzkaller.appspotmail.com
Fix commit: 01bfe5e8e428 Revert "net/sctp: fix race condition in sctp_destroy_sock"
First crash: 1918d, last: 1911d
Cause bisection: introduced by (bisect log) :
commit 997acaf6b4b59c6a9c259740312a69ea549cc684
Author: Mark Rutland <mark.rutland@arm.com>
Date: Mon Jan 11 15:37:07 2021 +0000

  lockdep: report broken irq restoration

Crash: WARNING in kvm_wait (log)
Repro: C syz .config
  
▶ ▼ Discussions (10)
Title Replies (including bot) Last reply
[PATCH 4.19 000/425] 4.19.191-rc1 review 438 (438) 2021/06/24 10:04
[PATCH 5.10 000/530] 5.10.37-rc1 review 571 (571) 2021/05/25 00:23
[PATCH 4.9 000/240] 4.9.269-rc1 review 245 (245) 2021/05/21 16:52
[PATCH 4.14 000/323] 4.14.233-rc1 review 326 (326) 2021/05/21 05:55
[PATCH 4.4 000/190] 4.4.269-rc1 review 196 (196) 2021/05/20 21:46
[PATCH 5.12 000/677] 5.12.4-rc1 review 694 (694) 2021/05/13 20:33
[PATCH 5.11 000/601] 5.11.21-rc1 review 607 (607) 2021/05/13 20:32
[PATCH 5.4 000/244] 5.4.119-rc1 review 258 (258) 2021/05/13 20:31
[PATCH net 0/2] sctp: fix the race condition in sctp_destroy_sock in a proper way 6 (6) 2021/05/10 19:18
[syzbot] possible deadlock in sctp_addr_wq_timeout_handler 1 (2) 2021/05/01 21:26
▶ ▼ Similar bugs (1)
Kernel Title Rank 🛈 Repro Cause bisect Fix bisect Count Last Reported Patched Status
linux-4.19 possible deadlock in sctp_addr_wq_timeout_handler 4 1 1913d 1913d 0/1 auto-closed as invalid on 2021/08/26 12:08

Sample crash report:
======================================================
WARNING: possible circular locking dependency detected
5.12.0-rc8-syzkaller #0 Not tainted
------------------------------------------------------
syz-executor044/8536 is trying to acquire lock:
ffff8880183933a0 (slock-AF_INET6){+.-.}-{2:2}, at: spin_lock include/linux/spinlock.h:354 [inline]
ffff8880183933a0 (slock-AF_INET6){+.-.}-{2:2}, at: sctp_addr_wq_timeout_handler+0x1a1/0x550 net/sctp/protocol.c:666

but task is already holding lock:
ffffffff8d659620 (&net->sctp.addr_wq_lock){+.-.}-{2:2}, at: spin_lock_bh include/linux/spinlock.h:359 [inline]
ffffffff8d659620 (&net->sctp.addr_wq_lock){+.-.}-{2:2}, at: sctp_addr_wq_timeout_handler+0x38/0x550 net/sctp/protocol.c:626

which lock already depends on the new lock.


the existing dependency chain (in reverse order) is:

-> #1 (&net->sctp.addr_wq_lock){+.-.}-{2:2}:
       __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
       _raw_spin_lock_bh+0x2f/0x40 kernel/locking/spinlock.c:175
       spin_lock_bh include/linux/spinlock.h:359 [inline]
       sctp_destroy_sock+0x204/0x440 net/sctp/socket.c:5028
       sctp_v6_destroy_sock+0x11/0x20 net/sctp/socket.c:9528
       sk_common_release+0x64/0x390 net/core/sock.c:3264
       sctp_close+0x4da/0x940 net/sctp/socket.c:1531
       inet_release+0x12e/0x280 net/ipv4/af_inet.c:431
       inet6_release+0x4c/0x70 net/ipv6/af_inet6.c:478
       __sock_release+0xcd/0x280 net/socket.c:599
       sock_close+0x18/0x20 net/socket.c:1258
       __fput+0x288/0x920 fs/file_table.c:280
       task_work_run+0xdd/0x1a0 kernel/task_work.c:140
       exit_task_work include/linux/task_work.h:30 [inline]
       do_exit+0xbfc/0x2a60 kernel/exit.c:825
       do_group_exit+0x125/0x310 kernel/exit.c:922
       __do_sys_exit_group kernel/exit.c:933 [inline]
       __se_sys_exit_group kernel/exit.c:931 [inline]
       __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:931
       do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
       entry_SYSCALL_64_after_hwframe+0x44/0xae

-> #0 (slock-AF_INET6){+.-.}-{2:2}:
       check_prev_add kernel/locking/lockdep.c:2937 [inline]
       check_prevs_add kernel/locking/lockdep.c:3060 [inline]
       validate_chain kernel/locking/lockdep.c:3675 [inline]
       __lock_acquire+0x2b14/0x54c0 kernel/locking/lockdep.c:4901
       lock_acquire kernel/locking/lockdep.c:5511 [inline]
       lock_acquire+0x1ab/0x740 kernel/locking/lockdep.c:5476
       __raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
       _raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
       spin_lock include/linux/spinlock.h:354 [inline]
       sctp_addr_wq_timeout_handler+0x1a1/0x550 net/sctp/protocol.c:666
       call_timer_fn+0x1a5/0x6b0 kernel/time/timer.c:1431
       expire_timers kernel/time/timer.c:1476 [inline]
       __run_timers.part.0+0x67c/0xa50 kernel/time/timer.c:1745
       __run_timers kernel/time/timer.c:1726 [inline]
       run_timer_softirq+0xb3/0x1d0 kernel/time/timer.c:1758
       __do_softirq+0x29b/0x9f6 kernel/softirq.c:345
       invoke_softirq kernel/softirq.c:221 [inline]
       __irq_exit_rcu kernel/softirq.c:422 [inline]
       irq_exit_rcu+0x134/0x200 kernel/softirq.c:434
       sysvec_apic_timer_interrupt+0x93/0xc0 arch/x86/kernel/apic/apic.c:1100
       asm_sysvec_apic_timer_interrupt+0x12/0x20 arch/x86/include/asm/idtentry.h:632
       __raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:161 [inline]
       _raw_spin_unlock_irqrestore+0x38/0x70 kernel/locking/spinlock.c:191
       __debug_check_no_obj_freed lib/debugobjects.c:997 [inline]
       debug_check_no_obj_freed+0x20c/0x420 lib/debugobjects.c:1018
       slab_free_hook mm/slub.c:1554 [inline]
       slab_free_freelist_hook+0x147/0x210 mm/slub.c:1600
       slab_free mm/slub.c:3161 [inline]
       kmem_cache_free+0x8a/0x740 mm/slub.c:3177
       free_fs_struct fs/fs_struct.c:92 [inline]
       exit_fs+0x123/0x170 fs/fs_struct.c:108
       do_exit+0xbca/0x2a60 kernel/exit.c:821
       do_group_exit+0x125/0x310 kernel/exit.c:922
       __do_sys_exit_group kernel/exit.c:933 [inline]
       __se_sys_exit_group kernel/exit.c:931 [inline]
       __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:931
       do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
       entry_SYSCALL_64_after_hwframe+0x44/0xae

other info that might help us debug this:

 Possible unsafe locking scenario:

       CPU0                    CPU1
       ----                    ----
  lock(&net->sctp.addr_wq_lock);
                               lock(slock-AF_INET6);
                               lock(&net->sctp.addr_wq_lock);
  lock(slock-AF_INET6);

 *** DEADLOCK ***

2 locks held by syz-executor044/8536:
 #0: ffffc90000007d78 ((&net->sctp.addr_wq_timer)){+.-.}-{0:0}, at: lockdep_copy_map include/linux/lockdep.h:35 [inline]
 #0: ffffc90000007d78 ((&net->sctp.addr_wq_timer)){+.-.}-{0:0}, at: call_timer_fn+0xd5/0x6b0 kernel/time/timer.c:1421
 #1: ffffffff8d659620 (&net->sctp.addr_wq_lock){+.-.}-{2:2}, at: spin_lock_bh include/linux/spinlock.h:359 [inline]
 #1: ffffffff8d659620 (&net->sctp.addr_wq_lock){+.-.}-{2:2}, at: sctp_addr_wq_timeout_handler+0x38/0x550 net/sctp/protocol.c:626

stack backtrace:
CPU: 0 PID: 8536 Comm: syz-executor044 Not tainted 5.12.0-rc8-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 <IRQ>
 __dump_stack lib/dump_stack.c:79 [inline]
 dump_stack+0x141/0x1d7 lib/dump_stack.c:120
 check_noncircular+0x25f/0x2e0 kernel/locking/lockdep.c:2128
 check_prev_add kernel/locking/lockdep.c:2937 [inline]
 check_prevs_add kernel/locking/lockdep.c:3060 [inline]
 validate_chain kernel/locking/lockdep.c:3675 [inline]
 __lock_acquire+0x2b14/0x54c0 kernel/locking/lockdep.c:4901
 lock_acquire kernel/locking/lockdep.c:5511 [inline]
 lock_acquire+0x1ab/0x740 kernel/locking/lockdep.c:5476
 __raw_spin_lock include/linux/spinlock_api_smp.h:142 [inline]
 _raw_spin_lock+0x2a/0x40 kernel/locking/spinlock.c:151
 spin_lock include/linux/spinlock.h:354 [inline]
 sctp_addr_wq_timeout_handler+0x1a1/0x550 net/sctp/protocol.c:666
 call_timer_fn+0x1a5/0x6b0 kernel/time/timer.c:1431
 expire_timers kernel/time/timer.c:1476 [inline]
 __run_timers.part.0+0x67c/0xa50 kernel/time/timer.c:1745
 __run_timers kernel/time/timer.c:1726 [inline]
 run_timer_softirq+0xb3/0x1d0 kernel/time/timer.c:1758
 __do_softirq+0x29b/0x9f6 kernel/softirq.c:345
 invoke_softirq kernel/softirq.c:221 [inline]
 __irq_exit_rcu kernel/softirq.c:422 [inline]
 irq_exit_rcu+0x134/0x200 kernel/softirq.c:434
 sysvec_apic_timer_interrupt+0x93/0xc0 arch/x86/kernel/apic/apic.c:1100
 </IRQ>
 asm_sysvec_apic_timer_interrupt+0x12/0x20 arch/x86/include/asm/idtentry.h:632
RIP: 0010:__raw_spin_unlock_irqrestore include/linux/spinlock_api_smp.h:161 [inline]
RIP: 0010:_raw_spin_unlock_irqrestore+0x38/0x70 kernel/locking/spinlock.c:191
Code: 74 24 10 e8 4a f9 53 f8 48 89 ef e8 82 af 54 f8 81 e3 00 02 00 00 75 25 9c 58 f6 c4 02 75 2d 48 85 db 74 01 fb bf 01 00 00 00 <e8> 63 7d 48 f8 65 8b 05 0c 48 fc 76 85 c0 74 0a 5b 5d c3 e8 d0 3c
RSP: 0018:ffffc9000173fc50 EFLAGS: 00000206
RAX: 0000000000000002 RBX: 0000000000000200 RCX: 1ffffffff1b89e11
RDX: 0000000000000000 RSI: 0000000000000001 RDI: 0000000000000001
RBP: ffffffff9006cf00 R08: 0000000000000001 R09: 0000000000000001
R10: ffffffff8179e4c8 R11: 000000000000003f R12: 1ffffffff200d9df
R13: 0000000000000000 R14: dead000000000100 R15: dffffc0000000000
 __debug_check_no_obj_freed lib/debugobjects.c:997 [inline]
 debug_check_no_obj_freed+0x20c/0x420 lib/debugobjects.c:1018
 slab_free_hook mm/slub.c:1554 [inline]
 slab_free_freelist_hook+0x147/0x210 mm/slub.c:1600
 slab_free mm/slub.c:3161 [inline]
 kmem_cache_free+0x8a/0x740 mm/slub.c:3177
 free_fs_struct fs/fs_struct.c:92 [inline]
 exit_fs+0x123/0x170 fs/fs_struct.c:108
 do_exit+0xbca/0x2a60 kernel/exit.c:821
 do_group_exit+0x125/0x310 kernel/exit.c:922
 __do_sys_exit_group kernel/exit.c:933 [inline]
 __se_sys_exit_group kernel/exit.c:931 [inline]
 __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:931
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x43ea79
Code: Unable to access opcode bytes at RIP 0x43ea4f.
RSP: 002b:00007ffdbc348058 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7
RAX: ffffffffffffffda RBX: 00000000004b0330 RCX: 000000000043ea79
RDX: 000000000000003c RSI: 00000000000000e7 RDI: 0000000000000000
RBP: 0000000000000000 R08: ffffffffffffffc0 R09: 0000000000000001
R10: 0000000000000001 R11: 0000000000000246 R12: 00000000004b0330
R13: 0000000000000001 R14: 0000000000000000 R15: 0000000000000001

Crashes (8):
Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets (help?) Manager Title
2021/04/25 15:53 upstream 2a1d7946fa53 36c88236 .config console log report syz C ci-upstream-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/04/28 03:12 net-next-old 95aafe911db6 805b5003 .config console log report syz ci-upstream-net-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/04/29 21:09 upstream d2b6f8a17919 77e2b668 .config console log report info ci-upstream-kasan-gce-smack-root possible deadlock in sctp_addr_wq_timeout_handler
2021/04/25 13:52 upstream 2a1d7946fa53 36c88236 .config console log report info ci-upstream-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/04/30 03:04 net-old bbd6f0a94813 77e2b668 .config console log report info ci-upstream-net-this-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/04/23 16:35 net-old bb556de79f0a 17f0b706 .config console log report info ci-upstream-net-this-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/05/01 05:43 net-next-old 95aafe911db6 77e2b668 .config console log report info ci-upstream-net-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
2021/04/24 05:14 net-next-old b2f0ca00e6b3 17f0b706 .config console log report info ci-upstream-net-kasan-gce possible deadlock in sctp_addr_wq_timeout_handler
* Struck through repros no longer work on HEAD.