Re: [syzbot] INFO: task hung in io_wqe_worker

[Pavel Begunkov]

On 10/22/21 14:57, Pavel Begunkov wrote:
> On 10/22/21 14:49, Pavel Begunkov wrote:
> > On 10/22/21 05:38, syzbot wrote:
> > > Hello,
> > >
> > > syzbot has tested the proposed patch but the reproducer is still triggering an issue:
> > > INFO: task hung in io_wqe_worker
> > >
> > > INFO: task iou-wrk-9392:9401 blocked for more than 143 seconds.
> > >        Not tainted 5.15.0-rc2-syzkaller #0
> > > "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> > > task:iou-wrk-9392    state:D stack:27952 pid: 9401 ppid:  7038 flags:0x00004004
> > > Call Trace:
> > >   context_switch kernel/sched/core.c:4940 [inline]
> > >   __schedule+0xb44/0x5960 kernel/sched/core.c:6287
> > >   schedule+0xd3/0x270 kernel/sched/core.c:6366
> > >   schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857
> > >   do_wait_for_common kernel/sched/completion.c:85 [inline]
> > >   __wait_for_common kernel/sched/completion.c:106 [inline]
> > >   wait_for_common kernel/sched/completion.c:117 [inline]
> > >   wait_for_completion+0x176/0x280 kernel/sched/completion.c:138
> > >   io_worker_exit fs/io-wq.c:183 [inline]
> > >   io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597
> > >   ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295

#syz test: https://github.com/isilence/linux.git syz_coredump



> > Easily reproducible, it's stuck in
> >
> > static void io_worker_exit(struct io_worker *worker)
> > {
> >      ...
> >      wait_for_completion(&worker->ref_done);
> >      ...
> > }
> >
> > The reference belongs to a create_worker_cb() task_work item. It's expected
> > to either be executed or cancelled by io_wq_exit_workers(), but the owner
> > task never goes __io_uring_cancel (called in do_exit()) and so never
> > reaches io_wq_exit_workers().
> >
> > Following the owner task, cat /proc/<pid>/stack:
> >
> > [<0>] do_coredump+0x1d0/0x10e0
> > [<0>] get_signal+0x4a3/0x960
> > [<0>] arch_do_signal_or_restart+0xc3/0x6d0
> > [<0>] exit_to_user_mode_prepare+0x10e/0x190
> > [<0>] irqentry_exit_to_user_mode+0x9/0x20
> > [<0>] irqentry_exit+0x36/0x40
> > [<0>] exc_page_fault+0x95/0x190
> > [<0>] asm_exc_page_fault+0x1e/0x30
> >
> > (gdb) l *(do_coredump+0x1d0-5)
> > 0xffffffff81343ccb is in do_coredump (fs/coredump.c:469).
> > 464
> > 465             if (core_waiters > 0) {
> > 466                     struct core_thread *ptr;
> > 467
> > 468                     freezer_do_not_count();
> > 469                     wait_for_completion(&core_state->startup);
> > 470                     freezer_count();
> >
> > Can't say anything more at the moment as not familiar with coredump
>
> A simple hack allowing task works to be executed from there
> workarounds the problem
>
>
> diff --git a/fs/coredump.c b/fs/coredump.c
> index 3224dee44d30..f6f9dfb02296 100644
> --- a/fs/coredump.c
> +++ b/fs/coredump.c
> @@ -466,7 +466,8 @@ static int coredump_wait(int exit_code, struct core_state *core_state)
>           struct core_thread *ptr;
>
>           freezer_do_not_count();
> -        wait_for_completion(&core_state->startup);
> +        while (wait_for_completion_interruptible(&core_state->startup))
> +            tracehook_notify_signal();
>           freezer_count();
>           /*
>            * Wait for all the threads to become inactive, so that

--
Pavel Begunkov