Re: [syzbot] [block?] possible deadlock in blk_mq_alloc_request

[Ming Lei <ming.lei@redhat.com>]

On Tue, Nov 05, 2024 at 06:40:22PM -0800, syzbot wrote:
> Hello,
> 
> syzbot found the following issue on:
> 
> HEAD commit:    c88416ba074a Add linux-next specific files for 20241101
> git tree:       linux-next
> console output: https://syzkaller.appspot.com/x/log.txt?x=17e59aa7980000
> kernel config:  https://syzkaller.appspot.com/x/.config?x=704b6be2ac2f205f
> dashboard link: https://syzkaller.appspot.com/bug?extid=ca7d7c797fee31d2b474
> compiler:       Debian clang version 15.0.6, GNU ld (GNU Binutils for Debian) 2.40
> syz repro:      https://syzkaller.appspot.com/x/repro.syz?x=1250b630580000
> 
> Downloadable assets:
> disk image: https://storage.googleapis.com/syzbot-assets/760a8c88d0c3/disk-c88416ba.raw.xz
> vmlinux: https://storage.googleapis.com/syzbot-assets/46e4b0a851a2/vmlinux-c88416ba.xz
> kernel image: https://storage.googleapis.com/syzbot-assets/428e2c784b75/bzImage-c88416ba.xz
> 
> IMPORTANT: if you fix the issue, please add the following tag to the commit:
> Reported-by: syzbot+ca7d7c797fee31d2b474@syzkaller.appspotmail.com
> 
> ============================================
> WARNING: possible recursive locking detected
> 6.12.0-rc5-next-20241101-syzkaller #0 Not tainted
> --------------------------------------------
> udevd/6086 is trying to acquire lock:
> ffff8880288261c0 (&q->q_usage_counter(queue)#67){++++}-{0:0}, at: blk_mq_alloc_request+0x26b/0xab0 block/blk-mq.c:626
> 
> but task is already holding lock:
> ffff8880288261c0 (&q->q_usage_counter(queue)#67){++++}-{0:0}, at: blk_freeze_queue block/blk-mq.c:177 [inline]
> ffff8880288261c0 (&q->q_usage_counter(queue)#67){++++}-{0:0}, at: blk_mq_freeze_queue+0x15/0x20 block/blk-mq.c:187
> 
> other info that might help us debug this:
>  Possible unsafe locking scenario:
> 
>        CPU0
>        ----
>   lock(&q->q_usage_counter(queue)#67);
>   lock(&q->q_usage_counter(queue)#67);
> 
>  *** DEADLOCK ***
> 
>  May be due to missing lock nesting notation
> 
> 3 locks held by udevd/6086:
>  #0: ffff888034a534c8 (&disk->open_mutex){+.+.}-{4:4}, at: bdev_open+0xf0/0xc50 block/bdev.c:904
>  #1: ffff888028826188 (&q->q_usage_counter(io)#81){+.+.}-{0:0}, at: blk_freeze_queue block/blk-mq.c:177 [inline]
>  #1: ffff888028826188 (&q->q_usage_counter(io)#81){+.+.}-{0:0}, at: blk_mq_freeze_queue+0x15/0x20 block/blk-mq.c:187
>  #2: ffff8880288261c0 (&q->q_usage_counter(queue)#67){++++}-{0:0}, at: blk_freeze_queue block/blk-mq.c:177 [inline]
>  #2: ffff8880288261c0 (&q->q_usage_counter(queue)#67){++++}-{0:0}, at: blk_mq_freeze_queue+0x15/0x20 block/blk-mq.c:187

Not get idea how blk_mq_freeze_queue is called in this context.

Is the blk_mq_unfreeze_queue() in sd_revalidate_disk() not released?

Anyway, please test the not-merged fixes.

#syz test: https://github.com/ming1/linux.git for-next


Thanks, 
Ming