Re: WARNING in cleanup_net

[Eric Dumazet]

On Thu, Nov 30, 2023 at 9:24 AM Eric Dumazet <edumazet@xxxxxxxxxx> wrote:
>
> On Thu, Nov 30, 2023 at 9:19 AM xingwei lee <xrivendell7@xxxxxxxxx> wrote:
> >
> > Hi
> > I found a bug with syzkaller titled "WARNING in cleanup_net" in the
> > last upstream.
> > =* BUG DETAILS =*
> > kernel commit: 3b47bc037bd44f142ac09848e8d3ecccc726be99
> > kernel config: https://syzkaller.appspot.com/text?tag=KernelConfig&x=f2a9d08825f82ef3
> > repro.c/repro.txt:
> > https://gist.github.com/xrivendell7/44780af4a9dededc5ff7a7c0583ce3f1
> > the crash report：
> > [ 8584.181281][T11719] bond0 (unregistering): Released all slaves
> > [ 8585.839049][T11719] ref_tracker: net notrefcnt@ffff888021ba0220 has
> > 1/1 users at
> > [ 8585.839049][T11719]   sk_alloc+0xaf0/0xbf0
> > [ 8585.839049][T11719]   inet6_create+0x39b/0x1300
> > [ 8585.839049][T11719]   __sock_create+0x34f/0x850
> > [ 8585.839049][T11719]   rds_tcp_listen_init+0xda/0x4f0
> > [ 8585.839049][T11719]   rds_tcp_init_net+0x147/0x400
> > [ 8585.839049][T11719]   ops_init+0xc4/0x680
> > [ 8585.839049][T11719]   setup_net+0x431/0xa80
> > [ 8585.839049][T11719]   copy_net_ns+0x313/0x6b0
> > [ 8585.839049][T11719]   create_new_namespaces+0x3fb/0xb60
> > [ 8585.839049][T11719]   unshare_nsproxy_namespaces+0xd0/0x200
> > [ 8585.839049][T11719]   ksys_unshare+0x47c/0xa30
> > [ 8585.839049][T11719]   __x64_sys_unshare+0x36/0x50
> > [ 8585.839049][T11719]   do_syscall_64+0x40/0x110
> > [ 8585.839049][T11719]   entry_SYSCALL_64_after_hwframe+0x63/0x6b
> > [ 8585.839049][T11719]
> > [ 8585.858614][T11719] ------------[ cut here ]------------
> > [ 8585.860037][T11719] WARNING: CPU: 3 PID: 11719 at
> > lib/ref_tracker.c:179 ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.862152][T11719] Modules linked in:
> > [ 8585.863038][T11719] CPU: 3 PID: 11719 Comm: kworker/u8:3 Not
> > tainted 6.7.0-rc1-g7475e51b8796-dirty #2
> > [ 8585.865268][T11719] Hardware name: QEMU Standard PC (i440FX + PIIX,
> > 1996), BIOS 1.16.2-1.fc38 04/01/2014
> > [ 8585.867345][T11719] Workqueue: netns cleanup_net
> > [ 8585.868401][T11719] RIP: 0010:ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.869426][T11719] Code: 00 00 4d 39 f5 49 8b 06 4d 89 f7 0f 85 08
> > ff ff ff 48 8b 2c 24 31 ff e8 c4 09 d6 fc 48 8b 74 24 18 48 89 ef e8
> > 67 13 32 06 90 <0f> 0b 90 48 8d 5d 44 31 ff e8 a8 09 d6 fc be 04 00 00
> > 00 48 89 df
> > [ 8585.872786][T11719] RSP: 0018:ffffc9000386fb78 EFLAGS: 00010286
> > [ 8585.873790][T11719] RAX: 0000000080000000 RBX: dffffc0000000000
> > RCX: 0000000000000000
> > [ 8585.875085][T11719] RDX: 0000000000000001 RSI: ffffffff8b2cb900
> > RDI: 0000000000000001
> > [ 8585.876394][T11719] RBP: ffff888021ba0220 R08: 0000000000000001
> > R09: fffffbfff24a13e9
> > [ 8585.877808][T11719] R10: ffffffff92509f4f R11: 0000000000000003
> > R12: ffff888021ba0270
> > [ 8585.879138][T11719] R13: ffff888021ba0270 R14: ffff888021ba0270
> > R15: ffff888021ba0270
> > [ 8585.880481][T11719] FS: 0000000000000000(0000)
> > GS:ffff88823bd00000(0000) knlGS:0000000000000000
> > [ 8585.881965][T11719] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > [ 8585.882719][T11719] CR2: 00007f00c5cd79f0 CR3: 000000000d377000
> > CR4: 0000000000750ef0
> > [ 8585.883536][T11719] PKRU: 55555554
> > [ 8585.883901][T11719] Call Trace:
> > [ 8585.884247][T11719] <TASK>
> > [ 8585.884561][T11719] ? show_regs+0x9a/0xb0
> > [ 8585.885005][T11719] ? __warn+0xf5/0x3c0
> > [ 8585.885423][T11719] ? report_bug+0x506/0x5f0
> > [ 8585.885900][T11719] ? ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.886466][T11719] ? report_bug+0x41c/0x5f0
> > [ 8585.886939][T11719] ? handle_bug+0x3d/0x70
> > [ 8585.887609][T11719] ? exc_invalid_op+0x17/0x40
> > [ 8585.888104][T11719] ? asm_exc_invalid_op+0x1a/0x20
> > [ 8585.888646][T11719] ? ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.889203][T11719] ? ref_tracker_dir_exit+0x3f9/0x6a0
> > [ 8585.889900][T11719] ? ref_tracker_dir_snprint+0xe0/0xe0
> > [ 8585.890465][T11719] ? __kmem_cache_free+0xc0/0x180
> > [ 8585.890989][T11719] cleanup_net+0x927/0xb70
> > [ 8585.891462][T11719] ? unregister_pernet_device+0x80/0x80
> > [ 8585.892038][T11719] process_one_work+0x8ab/0x1730
> > [ 8585.892564][T11719] ? unregister_pernet_device+0x80/0x80
> > [ 8585.893137][T11719] ? workqueue_congested+0x320/0x320
> > [ 8585.893692][T11719] ? assign_work+0x1b7/0x260
> > [ 8585.894177][T11719] worker_thread+0x931/0x1380
> > [ 8585.894683][T11719] ? process_one_work+0x1730/0x1730
> > [ 8585.895224][T11719] kthread+0x2d3/0x3b0
> > [ 8585.895664][T11719] ? _raw_spin_unlock_irq+0x23/0x50
> > [ 8585.896191][T11719] ? kthread_complete_and_exit+0x40/0x40
> > [ 8585.896779][T11719] ret_from_fork+0x4e/0x80
> > [ 8585.897245][T11719] ? kthread_complete_and_exit+0x40/0x40
> > [ 8585.897863][T11719] ret_from_fork_asm+0x11/0x20
> > [ 8585.898371][T11719] </TASK>
> > [ 8585.898702][T11719] Kernel panic - not syncing: kernel: panic_on_warn set ...
> > [ 8585.899428][T11719] CPU: 3 PID: 11719 Comm: kworker/u8:3 Not
> > tainted 6.7.0-rc1-g7475e51b8796-dirty #2
> > [ 8585.900572][T11719] Hardware name: QEMU Standard PC (i440FX + PIIX,
> > 1996), BIOS 1.16.2-1.fc38 04/01/2014
> > [ 8585.901550][T11719] Workqueue: netns cleanup_net
> > [ 8585.902038][T11719] Call Trace:
> > [ 8585.902377][T11719] <TASK>
> > [ 8585.902693][T11719] dump_stack_lvl+0xee/0x1e0
> > [ 8585.903170][T11719] panic+0x754/0x810
> > [ 8585.903805][T11719] ? panic_smp_self_stop+0xa0/0xa0
> > [ 8585.904377][T11719] ? show_trace_log_lvl+0x394/0x540
> > [ 8585.905159][T11719] ? check_panic_on_warn+0xa4/0xc0
> > [ 8585.905742][T11719] ? ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.906532][T11719] check_panic_on_warn+0xb8/0xc0
> > [ 8585.907108][T11719] __warn+0x101/0x3c0
> > [ 8585.907577][T11719] ? report_bug+0x506/0x5f0
> > [ 8585.908317][T11719] ? ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.909080][T11719] report_bug+0x41c/0x5f0
> > [ 8585.909599][T11719] handle_bug+0x3d/0x70
> > [ 8585.910089][T11719] exc_invalid_op+0x17/0x40
> > [ 8585.910571][T11719] asm_exc_invalid_op+0x1a/0x20
> > [ 8585.911115][T11719] RIP: 0010:ref_tracker_dir_exit+0x3fa/0x6a0
> > [ 8585.911798][T11719] Code: 00 00 4d 39 f5 49 8b 06 4d 89 f7 0f 85 08
> > ff ff ff 48 8b 2c 24 31 ff e8 c4 09 d6 fc 48 8b 74 24 18 48 89 ef e8
> > 67 13 32 06 90 <0f> 0b 90 48 8d 5d 44 31 ff e8 a8 09 d6 fc be 04 00 00
> > 00 48 89 df
> > [ 8585.914135][T11719] RSP: 0018:ffffc9000386fb78 EFLAGS: 00010286
> > [ 8585.914813][T11719] RAX: 0000000080000000 RBX: dffffc0000000000
> > RCX: 0000000000000000
> > [ 8585.915654][T11719] RDX: 0000000000000001 RSI: ffffffff8b2cb900
> > RDI: 0000000000000001
> > [ 8585.916556][T11719] RBP: ffff888021ba0220 R08: 0000000000000001
> > R09: fffffbfff24a13e9
> > [ 8585.917647][T11719] R10: ffffffff92509f4f R11: 0000000000000003
> > R12: ffff888021ba0270
> > [ 8585.918692][T11719] R13: ffff888021ba0270 R14: ffff888021ba0270
> > R15: ffff888021ba0270
> > [ 8585.919598][T11719] ? ref_tracker_dir_exit+0x3f9/0x6a0
> > [ 8585.920188][T11719] ? ref_tracker_dir_snprint+0xe0/0xe0
> > [ 8585.920803][T11719] ? __kmem_cache_free+0xc0/0x180
> > [ 8585.921387][T11719] cleanup_net+0x927/0xb70
> > [ 8585.921891][T11719] ? unregister_pernet_device+0x80/0x80
> > [ 8585.922481][T11719] process_one_work+0x8ab/0x1730
> > [ 8585.923093][T11719] ? unregister_pernet_device+0x80/0x80
> > [ 8585.923781][T11719] ? workqueue_congested+0x320/0x320
> > [ 8585.924385][T11719] ? assign_work+0x1b7/0x260
> > [ 8585.924923][T11719] worker_thread+0x931/0x1380
> > [ 8585.925467][T11719] ? process_one_work+0x1730/0x1730
> > [ 8585.926079][T11719] kthread+0x2d3/0x3b0
> > [ 8585.926538][T11719] ? _raw_spin_unlock_irq+0x23/0x50
> > [ 8585.927138][T11719] ? kthread_complete_and_exit+0x40/0x40
> > [ 8585.927763][T11719] ret_from_fork+0x4e/0x80
> > [ 8585.928256][T11719] ? kthread_complete_and_exit+0x40/0x40
> > [ 8585.928903][T11719] ret_from_fork_asm+0x11/0x20
> > [ 8585.929470][T11719] </TASK>
> > [ 8585.930023][T11719] Kernel Offset: disabled
> > [ 8585.930517][T11719] Rebooting in 86400 seconds..
> > =* OTHERS =*
> > I noticed syzbot has two similar bugs named WARNING in cleanup_net：
> > https://syzkaller.appspot.com/bug?extid=7e1e1bdb852961150198
> > https://syzkaller.appspot.com/bug?id=14c45b4081250ebeb4a9000f3774da829f7e43b4
> >
> > However, these two seem fixed and not related to this.
> > Without in-depth analysis, I guess it's maybe a race condition bug and
> > the import part may be the refcnt_tracker in socket$rds not handled
> > properly but I'm not sure.
> >
> > [ 8585.839049][T11719] ref_tracker: net notrefcnt@ffff888021ba0220 has
> > 1/1 users at
> > [ 8585.839049][T11719]   sk_alloc+0xaf0/0xbf0
> > [ 8585.839049][T11719]   inet6_create+0x39b/0x1300
> > [ 8585.839049][T11719]   __sock_create+0x34f/0x850
> > [ 8585.839049][T11719]   rds_tcp_listen_init+0xda/0x4f0
> > [ 8585.839049][T11719]   rds_tcp_init_net+0x147/0x400
> > [ 8585.839049][T11719]   ops_init+0xc4/0x680
> > [ 8585.839049][T11719]   setup_net+0x431/0xa80
> > [ 8585.839049][T11719]   copy_net_ns+0x313/0x6b0
> > [ 8585.839049][T11719]   create_new_namespaces+0x3fb/0xb60
> > [ 8585.839049][T11719]   unshare_nsproxy_namespaces+0xd0/0x200
> > [ 8585.839049][T11719]   ksys_unshare+0x47c/0xa30
> > [ 8585.839049][T11719]   __x64_sys_unshare+0x36/0x50
> > [ 8585.839049][T11719]   do_syscall_64+0x40/0x110
> > [ 8585.839049][T11719]   entry_SYSCALL_64_after_hwframe+0x63/0x6b
>
> I have a similar syzbot  bug in my triage queue I will release right now.
>

This is the "[syzbot] [net?] WARNING in cleanup_net (3)"  thread.

> I already worked on a solution.

But was waiting for a repro...

>
> Thanks.