[syzbot] [bpf?] BUG: spinlock recursion in bpf_lru_push_free

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 3d650ab5e7d9 selftests/bpf: Fix a btf_dump selftest failure
git tree: bpf-next
console output: https://syzkaller.appspot.com/x/log.txt?x=13a4c1a1980000
kernel config: https://syzkaller.appspot.com/x/.config?x=5efb917b1462a973
dashboard link: https://syzkaller.appspot.com/bug?extid=d6fb861ed047a275747a
compiler: Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40

Unfortunately, I don't have any reproducer for this issue yet.

Downloadable assets:
disk image: https://storage.googleapis.com/syzbot-assets/630e210de8d9/disk-3d650ab5.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/3576ca35748a/vmlinux-3d650ab5.xz
kernel image: https://storage.googleapis.com/syzbot-assets/5b33f099abfa/bzImage-3d650ab5.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+d6fb86...@syzkaller.appspotmail.com

BUG: spinlock recursion on CPU#1, syz.4.1173/11483
lock: 0xffff888046908300, .magic: dead4ead, .owner: syz.4.1173/11483, .owner_cpu: 1
CPU: 1 UID: 0 PID: 11483 Comm: syz.4.1173 Not tainted 6.10.0-syzkaller-12666-g3d650ab5e7d9 #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/27/2024
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:93 [inline]
dump_stack_lvl+0x241/0x360 lib/dump_stack.c:119
debug_spin_lock_before kernel/locking/spinlock_debug.c:87 [inline]
do_raw_spin_lock+0x227/0x370 kernel/locking/spinlock_debug.c:115
__raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:111 [inline]
_raw_spin_lock_irqsave+0xe1/0x120 kernel/locking/spinlock.c:162
bpf_lru_list_push_free kernel/bpf/bpf_lru_list.c:318 [inline]
bpf_common_lru_push_free kernel/bpf/bpf_lru_list.c:538 [inline]
bpf_lru_push_free+0x1a7/0xb60 kernel/bpf/bpf_lru_list.c:561
htab_lru_map_delete_elem+0x613/0x700 kernel/bpf/hashtab.c:1475
bpf_prog_6f5f05285f674219+0x43/0x4c
bpf_dispatcher_nop_func include/linux/bpf.h:1252 [inline]
__bpf_prog_run include/linux/filter.h:691 [inline]
bpf_prog_run include/linux/filter.h:698 [inline]
__bpf_trace_run kernel/trace/bpf_trace.c:2406 [inline]
bpf_trace_run2+0x2ec/0x540 kernel/trace/bpf_trace.c:2447
trace_contention_begin+0x117/0x140 include/trace/events/lock.h:95
__pv_queued_spin_lock_slowpath+0x114/0xdc0 kernel/locking/qspinlock.c:402
pv_queued_spin_lock_slowpath arch/x86/include/asm/paravirt.h:584 [inline]
queued_spin_lock_slowpath+0x42/0x50 arch/x86/include/asm/qspinlock.h:51
queued_spin_lock include/asm-generic/qspinlock.h:114 [inline]
lockdep_lock+0x1b0/0x2b0 kernel/locking/lockdep.c:143
graph_lock kernel/locking/lockdep.c:169 [inline]
lookup_chain_cache_add kernel/locking/lockdep.c:3803 [inline]
validate_chain+0x21d/0x5900 kernel/locking/lockdep.c:3836
__lock_acquire+0x137a/0x2040 kernel/locking/lockdep.c:5142
lock_acquire+0x1ed/0x550 kernel/locking/lockdep.c:5759
__raw_spin_lock include/linux/spinlock_api_smp.h:133 [inline]
_raw_spin_lock+0x2e/0x40 kernel/locking/spinlock.c:154
htab_lock_bucket+0x1a4/0x370 kernel/bpf/hashtab.c:167
htab_lru_map_delete_node+0x161/0x840 kernel/bpf/hashtab.c:817
__bpf_lru_list_shrink_inactive kernel/bpf/bpf_lru_list.c:225 [inline]
__bpf_lru_list_shrink+0x156/0x9d0 kernel/bpf/bpf_lru_list.c:271
bpf_lru_list_pop_free_to_local kernel/bpf/bpf_lru_list.c:345 [inline]
bpf_common_lru_pop_free kernel/bpf/bpf_lru_list.c:452 [inline]
bpf_lru_pop_free+0xd84/0x1a70 kernel/bpf/bpf_lru_list.c:504
prealloc_lru_pop kernel/bpf/hashtab.c:308 [inline]
__htab_lru_percpu_map_update_elem+0x242/0x9b0 kernel/bpf/hashtab.c:1355
bpf_percpu_hash_update+0x11a/0x200 kernel/bpf/hashtab.c:2421
bpf_map_update_value+0x347/0x540 kernel/bpf/syscall.c:181
generic_map_update_batch+0x60d/0x900 kernel/bpf/syscall.c:1889
bpf_map_do_batch+0x3e0/0x690 kernel/bpf/syscall.c:5218
__sys_bpf+0x377/0x810
__do_sys_bpf kernel/bpf/syscall.c:5817 [inline]
__se_sys_bpf kernel/bpf/syscall.c:5815 [inline]
__x64_sys_bpf+0x7c/0x90 kernel/bpf/syscall.c:5815
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0xf3/0x230 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fed319779f9
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 a8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007fed327ee048 EFLAGS: 00000246 ORIG_RAX: 0000000000000141
RAX: ffffffffffffffda RBX: 00007fed31b05f80 RCX: 00007fed319779f9
RDX: 0000000000000038 RSI: 0000000020000580 RDI: 000000000000001a
RBP: 00007fed319e58ee R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000000b R14: 00007fed31b05f80 R15: 00007ffcf9a1d7f8
</TASK>


---
This report 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 syzk...@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[Yonghong Song]

On 8/5/24 3:36 AM, syzbot wrote:
> Hello,
>
> syzbot found the following issue on:
>
> HEAD commit: 3d650ab5e7d9 selftests/bpf: Fix a btf_dump selftest failure

The failure is not due to this patch.

> git tree: bpf-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=13a4c1a1980000
> kernel config: https://syzkaller.appspot.com/x/.config?x=5efb917b1462a973
> dashboard link: https://syzkaller.appspot.com/bug?extid=d6fb861ed047a275747a
> compiler: Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
>
> Unfortunately, I don't have any reproducer for this issue yet.
>
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/630e210de8d9/disk-3d650ab5.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/3576ca35748a/vmlinux-3d650ab5.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/5b33f099abfa/bzImage-3d650ab5.xz
>
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+d6fb86...@syzkaller.appspotmail.com
>
> BUG: spinlock recursion on CPU#1, syz.4.1173/11483

Actually this is a known issue and has been reported a few times in the past.

> lock: 0xffff888046908300, .magic: dead4ead, .owner: syz.4.1173/11483, .owner_cpu: 1
> CPU: 1 UID: 0 PID: 11483 Comm: syz.4.1173 Not tainted 6.10.0-syzkaller-12666-g3d650ab5e7d9 #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 06/27/2024
> Call Trace:
> <TASK>
> __dump_stack lib/dump_stack.c:93 [inline]
> dump_stack_lvl+0x241/0x360 lib/dump_stack.c:119
> debug_spin_lock_before kernel/locking/spinlock_debug.c:87 [inline]
> do_raw_spin_lock+0x227/0x370 kernel/locking/spinlock_debug.c:115
> __raw_spin_lock_irqsave include/linux/spinlock_api_smp.h:111 [inline]
> _raw_spin_lock_irqsave+0xe1/0x120 kernel/locking/spinlock.c:162
> bpf_lru_list_push_free kernel/bpf/bpf_lru_list.c:318 [inline]
> bpf_common_lru_push_free kernel/bpf/bpf_lru_list.c:538 [inline]
> bpf_lru_push_free+0x1a7/0xb60 kernel/bpf/bpf_lru_list.c:561
> htab_lru_map_delete_elem+0x613/0x700 kernel/bpf/hashtab.c:1475
> bpf_prog_6f5f05285f674219+0x43/0x4c
> bpf_dispatcher_nop_func include/linux/bpf.h:1252 [inline]
> __bpf_prog_run include/linux/filter.h:691 [inline]
> bpf_prog_run include/linux/filter.h:698 [inline]
> __bpf_trace_run kernel/trace/bpf_trace.c:2406 [inline]
> bpf_trace_run2+0x2ec/0x540 kernel/trace/bpf_trace.c:2447
> trace_contention_begin+0x117/0x140 include/trace/events/lock.h:95

The tracepoint trace_contention_begin is the reason for spinlock recursion.
The trace_contention_begin is in
queued_spin_lock_slowpath(...) {
...
trace_contention_begin(lock, LCB_F_SPIN);
...
}

And the bpf prog attached to trace_contention_begin() will go though spin_lock path again
and this may cause dead lock.