Re: possible deadlock in bpf_tcp_close

From: Daniel Borkmann
Date: Mon May 28 2018 - 10:51:59 EST

Next message: Greg Kroah-Hartman: "[PATCH 4.9 009/329] mmc: sdhci-iproc: remove hard coded mmc cap 1.8v"
Previous message: Greg Kroah-Hartman: "[PATCH 4.9 055/329] fm10k: fix "failed to kill vid" message for VF"
In reply to: syzbot: "possible deadlock in bpf_tcp_close"
Next in thread: syzbot: "Re: possible deadlock in bpf_tcp_close"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

[ +John ]

On 05/28/2018 04:47 PM, syzbot wrote:
> Hello,
>
> syzbot found the following crash on:
>
> HEAD commit:ÂÂÂ 7a1a98c171ea Merge branch 'bpf-sendmsg-hook'
> git tree:ÂÂÂÂÂÂ bpf-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=10fd82d7800000
> kernel config:Â https://syzkaller.appspot.com/x/.config?x=e4078980b886800c
> dashboard link: https://syzkaller.appspot.com/bug?extid=47ed903f50684f046b15
> compiler:ÂÂÂÂÂÂ gcc (GCC) 8.0.1 20180413 (experimental)
>
> Unfortunately, I don't have any reproducer for this crash yet.
>
> IMPORTANT: if you fix the bug, please add the following tag to the commit:
> Reported-by: syzbot+47ed903f50684f046b15@xxxxxxxxxxxxxxxxxxxxxxxxx

So this one does have [1] included in the tree. :-( John, can you take a look?

[1] https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git/commit/?id=c1659ee5bb2a9f2023f5065256d5fd742ed660d2

Thanks,
Daniel

> ======================================================
> WARNING: possible circular locking dependency detected
> 4.17.0-rc6+ #25 Not tainted
> ------------------------------------------------------
> syz-executor4/7489 is trying to acquire lock:
> ÂÂÂÂÂÂÂ (ptrval) (&htab->buckets[i].lock#2){+...}, at: bpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285
>
> but task is already holding lock:
> ÂÂÂÂÂÂÂ (ptrval) (clock-AF_INET6){++..}, at: bpf_tcp_close+0x241/0x10b0 kernel/bpf/sockmap.c:260
>
> which lock already depends on the new lock.
>
>
> the existing dependency chain (in reverse order) is:
>
> -> #1 (clock-AF_INET6){++..}:
> ÂÂÂÂÂÂ __raw_write_lock_bh include/linux/rwlock_api_smp.h:203 [inline]
> ÂÂÂÂÂÂ _raw_write_lock_bh+0x31/0x40 kernel/locking/spinlock.c:312
> ÂÂÂÂÂÂ sock_hash_delete_elem+0x7c6/0xaf0 kernel/bpf/sockmap.c:2338
> ÂÂÂÂÂÂ map_delete_elem+0x32e/0x4e0 kernel/bpf/syscall.c:815
> ÂÂÂÂÂÂ __do_sys_bpf kernel/bpf/syscall.c:2349 [inline]
> ÂÂÂÂÂÂ __se_sys_bpf kernel/bpf/syscall.c:2317 [inline]
> ÂÂÂÂÂÂ __x64_sys_bpf+0x342/0x510 kernel/bpf/syscall.c:2317
> ÂÂÂÂÂÂ do_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
> ÂÂÂÂÂÂ entry_SYSCALL_64_after_hwframe+0x49/0xbe
>
> -> #0 (&htab->buckets[i].lock#2){+...}:
> ÂÂÂÂÂÂ lock_acquire+0x1dc/0x520 kernel/locking/lockdep.c:3920
> ÂÂÂÂÂÂ __raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
> ÂÂÂÂÂÂ _raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168
> ÂÂÂÂÂÂ bpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285
> ÂÂÂÂÂÂ inet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
> ÂÂÂÂÂÂ inet6_release+0x50/0x70 net/ipv6/af_inet6.c:459
> ÂÂÂÂÂÂ sock_release+0x96/0x1b0 net/socket.c:594
> ÂÂÂÂÂÂ sock_close+0x16/0x20 net/socket.c:1149
> ÂÂÂÂÂÂ __fput+0x34d/0x890 fs/file_table.c:209
> ÂÂÂÂÂÂ ____fput+0x15/0x20 fs/file_table.c:243
> ÂÂÂÂÂÂ task_work_run+0x1e4/0x290 kernel/task_work.c:113
> ÂÂÂÂÂÂ exit_task_work include/linux/task_work.h:22 [inline]
> ÂÂÂÂÂÂ do_exit+0x1aee/0x2730 kernel/exit.c:865
> ÂÂÂÂÂÂ do_group_exit+0x16f/0x430 kernel/exit.c:968
> ÂÂÂÂÂÂ get_signal+0x886/0x1960 kernel/signal.c:2482
> ÂÂÂÂÂÂ do_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
> ÂÂÂÂÂÂ exit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
> ÂÂÂÂÂÂ prepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
> ÂÂÂÂÂÂ syscall_return_slowpath arch/x86/entry/common.c:265 [inline]
> ÂÂÂÂÂÂ do_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
> ÂÂÂÂÂÂ entry_SYSCALL_64_after_hwframe+0x49/0xbe
>
> other info that might help us debug this:
>
> ÂPossible unsafe locking scenario:
>
> ÂÂÂÂÂÂ CPU0ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ CPU1
> ÂÂÂÂÂÂ ----ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ ----
> Â lock(clock-AF_INET6);
> ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ lock(&htab->buckets[i].lock#2);
> ÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ lock(clock-AF_INET6);
> Â lock(&htab->buckets[i].lock#2);
>
> Â*** DEADLOCK ***
>
> 2 locks held by syz-executor4/7489:
> Â#0:ÂÂÂÂÂÂÂÂ (ptrval) (rcu_read_lock){....}, at: bpf_tcp_close+0x0/0x10b0 kernel/bpf/sockmap.c:2106
> Â#1:ÂÂÂÂÂÂÂÂ (ptrval) (clock-AF_INET6){++..}, at: bpf_tcp_close+0x241/0x10b0 kernel/bpf/sockmap.c:260
>
> stack backtrace:
> CPU: 1 PID: 7489 Comm: syz-executor4 Not tainted 4.17.0-rc6+ #25
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> Â__dump_stack lib/dump_stack.c:77 [inline]
> Âdump_stack+0x1b9/0x294 lib/dump_stack.c:113
> Âprint_circular_bug.isra.36.cold.54+0x1bd/0x27d kernel/locking/lockdep.c:1223
> Âcheck_prev_add kernel/locking/lockdep.c:1863 [inline]
> Âcheck_prevs_add kernel/locking/lockdep.c:1976 [inline]
> Âvalidate_chain kernel/locking/lockdep.c:2417 [inline]
> Â__lock_acquire+0x343e/0x5140 kernel/locking/lockdep.c:3431
> Âlock_acquire+0x1dc/0x520 kernel/locking/lockdep.c:3920
> Â__raw_spin_lock_bh include/linux/spinlock_api_smp.h:135 [inline]
> Â_raw_spin_lock_bh+0x31/0x40 kernel/locking/spinlock.c:168
> Âbpf_tcp_close+0x822/0x10b0 kernel/bpf/sockmap.c:285
> Âinet_release+0x104/0x1f0 net/ipv4/af_inet.c:427
> Âinet6_release+0x50/0x70 net/ipv6/af_inet6.c:459
> Âsock_release+0x96/0x1b0 net/socket.c:594
> Âsock_close+0x16/0x20 net/socket.c:1149
> Â__fput+0x34d/0x890 fs/file_table.c:209
> Â____fput+0x15/0x20 fs/file_table.c:243
> Âtask_work_run+0x1e4/0x290 kernel/task_work.c:113
> Âexit_task_work include/linux/task_work.h:22 [inline]
> Âdo_exit+0x1aee/0x2730 kernel/exit.c:865
> Âdo_group_exit+0x16f/0x430 kernel/exit.c:968
> Âget_signal+0x886/0x1960 kernel/signal.c:2482
> Âdo_signal+0x98/0x2040 arch/x86/kernel/signal.c:810
> Âexit_to_usermode_loop+0x28a/0x310 arch/x86/entry/common.c:162
> Âprepare_exit_to_usermode arch/x86/entry/common.c:196 [inline]
> Âsyscall_return_slowpath arch/x86/entry/common.c:265 [inline]
> Âdo_syscall_64+0x6ac/0x800 arch/x86/entry/common.c:290
> Âentry_SYSCALL_64_after_hwframe+0x49/0xbe
> RIP: 0033:0x455a09
> RSP: 002b:00007f95715e8ce8 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca
> RAX: fffffffffffffe00 RBX: 000000000072c028 RCX: 0000000000455a09
> RDX: 0000000000000000 RSI: 0000000000000000 RDI: 000000000072c028
> RBP: 000000000072c028 R08: 0000000000000000 R09: 000000000072c000
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
> R13: 00007fffdce9240f R14: 00007f95715e99c0 R15: 0000000000000002
> cgroup: cgroup2: unknown option "cgroup2"
> cgroup: cgroup2: unknown option "cgroup2"
> syz-executor4 uses obsolete (PF_INET,SOCK_PACKET)
> FAULT_INJECTION: forcing a failure.
> name failslab, interval 1, probability 0, space 0, times 1
> CPU: 0 PID: 7912 Comm: syz-executor0 Not tainted 4.17.0-rc6+ #25
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> Â__dump_stack lib/dump_stack.c:77 [inline]
> Âdump_stack+0x1b9/0x294 lib/dump_stack.c:113
> Âfail_dump lib/fault-inject.c:51 [inline]
> Âshould_fail.cold.4+0xa/0x1a lib/fault-inject.c:149
> Â__should_failslab+0x124/0x180 mm/failslab.c:32
> Âshould_failslab+0x9/0x14 mm/slab_common.c:1522
> Âslab_pre_alloc_hook mm/slab.h:423 [inline]
> Âslab_alloc_node mm/slab.c:3299 [inline]
> Âkmem_cache_alloc_node+0x272/0x780 mm/slab.c:3642
> Â__alloc_skb+0x111/0x780 net/core/skbuff.c:193
> Âalloc_skb include/linux/skbuff.h:989 [inline]
> Âalloc_skb_with_frags+0x137/0x760 net/core/skbuff.c:5266
> Âsock_alloc_send_pskb+0x87a/0xae0 net/core/sock.c:2095
> Âunix_dgram_sendmsg+0x4f9/0x1730 net/unix/af_unix.c:1672
> Âunix_seqpacket_sendmsg+0x115/0x18f net/unix/af_unix.c:2053
> Âsock_sendmsg_nosec net/socket.c:629 [inline]
> Âsock_sendmsg+0xd5/0x120 net/socket.c:639
> Â___sys_sendmsg+0x805/0x940 net/socket.c:2117
> Â__sys_sendmsg+0x115/0x270 net/socket.c:2155
> Â__do_sys_sendmsg net/socket.c:2164 [inline]
> Â__se_sys_sendmsg net/socket.c:2162 [inline]
> Â__x64_sys_sendmsg+0x78/0xb0 net/socket.c:2162
> Âdo_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
> Âentry_SYSCALL_64_after_hwframe+0x49/0xbe
> RIP: 0033:0x455a09
> RSP: 002b:00007f85737fdc68 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
> RAX: ffffffffffffffda RBX: 00007f85737fe6d4 RCX: 0000000000455a09
> RDX: 0000000000000000 RSI: 00000000200013c0 RDI: 0000000000000013
> RBP: 000000000072bea0 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000015
> R13: 000000000000057f R14: 00000000006fc488 R15: 0000000000000000
> FAULT_INJECTION: forcing a failure.
> name failslab, interval 1, probability 0, space 0, times 0
> CPU: 1 PID: 7947 Comm: syz-executor0 Not tainted 4.17.0-rc6+ #25
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> Â__dump_stack lib/dump_stack.c:77 [inline]
> Âdump_stack+0x1b9/0x294 lib/dump_stack.c:113
> Âfail_dump lib/fault-inject.c:51 [inline]
> Âshould_fail.cold.4+0xa/0x1a lib/fault-inject.c:149
> Â__should_failslab+0x124/0x180 mm/failslab.c:32
> Âshould_failslab+0x9/0x14 mm/slab_common.c:1522
> Âslab_pre_alloc_hook mm/slab.h:423 [inline]
> Âslab_alloc_node mm/slab.c:3299 [inline]
> Âkmem_cache_alloc_node_trace+0x26f/0x770 mm/slab.c:3661
> Â__do_kmalloc_node mm/slab.c:3681 [inline]
> Â__kmalloc_node_track_caller+0x33/0x70 mm/slab.c:3696
> Â__kmalloc_reserve.isra.39+0x3a/0xe0 net/core/skbuff.c:137
> Â__alloc_skb+0x14d/0x780 net/core/skbuff.c:205
> Âalloc_skb include/linux/skbuff.h:989 [inline]
> Âalloc_skb_with_frags+0x137/0x760 net/core/skbuff.c:5266
> Âsock_alloc_send_pskb+0x87a/0xae0 net/core/sock.c:2095
> Âunix_dgram_sendmsg+0x4f9/0x1730 net/unix/af_unix.c:1672
> Âunix_seqpacket_sendmsg+0x115/0x18f net/unix/af_unix.c:2053
> Âsock_sendmsg_nosec net/socket.c:629 [inline]
> Âsock_sendmsg+0xd5/0x120 net/socket.c:639
> Â___sys_sendmsg+0x805/0x940 net/socket.c:2117
> Â__sys_sendmsg+0x115/0x270 net/socket.c:2155
> Â__do_sys_sendmsg net/socket.c:2164 [inline]
> Â__se_sys_sendmsg net/socket.c:2162 [inline]
> Â__x64_sys_sendmsg+0x78/0xb0 net/socket.c:2162
> Âdo_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
> Âentry_SYSCALL_64_after_hwframe+0x49/0xbe
> RIP: 0033:0x455a09
> RSP: 002b:00007f85737fdc68 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
> RAX: ffffffffffffffda RBX: 00007f85737fe6d4 RCX: 0000000000455a09
> RDX: 0000000000000000 RSI: 00000000200013c0 RDI: 0000000000000013
> RBP: 000000000072bea0 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000015
> R13: 000000000000057f R14: 00000000006fc488 R15: 0000000000000001
> FAULT_INJECTION: forcing a failure.
> name failslab, interval 1, probability 0, space 0, times 0
> CPU: 1 PID: 8051 Comm: syz-executor1 Not tainted 4.17.0-rc6+ #25
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> Â__dump_stack lib/dump_stack.c:77 [inline]
> Âdump_stack+0x1b9/0x294 lib/dump_stack.c:113
> Âfail_dump lib/fault-inject.c:51 [inline]
> Âshould_fail.cold.4+0xa/0x1a lib/fault-inject.c:149
> Â__should_failslab+0x124/0x180 mm/failslab.c:32
> Âshould_failslab+0x9/0x14 mm/slab_common.c:1522
> Âslab_pre_alloc_hook mm/slab.h:423 [inline]
> Âslab_alloc mm/slab.c:3378 [inline]
> Â__do_kmalloc mm/slab.c:3716 [inline]
> Â__kmalloc+0x2c8/0x760 mm/slab.c:3727
> Âkmalloc include/linux/slab.h:517 [inline]
> Âmap_get_next_key+0x24a/0x640 kernel/bpf/syscall.c:863
> Â__do_sys_bpf kernel/bpf/syscall.c:2352 [inline]
> Â__se_sys_bpf kernel/bpf/syscall.c:2317 [inline]
> Â__x64_sys_bpf+0x357/0x510 kernel/bpf/syscall.c:2317
> Âdo_syscall_64+0x1b1/0x800 arch/x86/entry/common.c:287
> Âentry_SYSCALL_64_after_hwframe+0x49/0xbe
> RIP: 0033:0x455a09
> RSP: 002b:00007fbd35c1ac68 EFLAGS: 00000246 ORIG_RAX: 0000000000000141
> RAX: ffffffffffffffda RBX: 00007fbd35c1b6d4 RCX: 0000000000455a09
> RDX: 000000000000002c RSI: 0000000020003000 RDI: 0000000000000004
> RBP: 000000000072bea0 R08: 0000000000000000 R09: 0000000000000000
> R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000014
> R13: 000000000000003d R14: 00000000006f4658 R15: 0000000000000000
> FAULT_INJECTION: forcing a failure.
> name fail_page_alloc, interval 1, probability 0, space 0, times 1
> CPU: 1 PID: 8104 Comm: syz-executor1 Not tainted 4.17.0-rc6+ #25
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
> Call Trace:
> Â__dump_stack lib/dump_stack.c:77 [inline]
> Âdump_stack+0x1b9/0x294 lib/dump_stack.c:113
> Âfail_dump lib/fault-inject.c:51 [inline]
> Âshould_fail.cold.4+0xa/0x1a lib/fault-inject.c:149
> Âshould_fail_alloc_page mm/page_alloc.c:3060 [inline]
> Âprepare_alloc_pages mm/page_alloc.c:4319 [inline]
> Â__alloc_pages_nodemask+0x34e/0xd70 mm/page_alloc.c:4358
> Âalloc_pages_vma+0xdd/0x540 mm/mempolicy.c:2057
> Âwp_page_copy+0x24c/0x1440 mm/memory.c:2490
> Âdo_wp_page+0x425/0x1990 mm/memory.c:2776
> Âhandle_pte_fault mm/memory.c:3979 [inline]
> Â__handle_mm_fault+0x2996/0x4310 mm/memory.c:4087
>
>
> ---
> This bug is generated by a bot. It may contain errors.
> See https://goo.gl/tpsmEJ for more information about syzbot.
> syzbot engineers can be reached at syzkaller@xxxxxxxxxxxxxxxxx
>
> syzbot will keep track of this bug report. See:
> https://goo.gl/tpsmEJ#bug-status-tracking for how to communicate with syzbot.

Next message: Greg Kroah-Hartman: "[PATCH 4.9 009/329] mmc: sdhci-iproc: remove hard coded mmc cap 1.8v"
Previous message: Greg Kroah-Hartman: "[PATCH 4.9 055/329] fm10k: fix "failed to kill vid" message for VF"
In reply to: syzbot: "possible deadlock in bpf_tcp_close"
Next in thread: syzbot: "Re: possible deadlock in bpf_tcp_close"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]