[Devel] [PATCH VZ10 v2] fs/fuse kio: fix hung connection on vstorage-mount exit

[Alexey Kuznetsov]

Hello!

On Fri, May 15, 2026 at 8:41 PM Konstantin Khorenko
<khorenko at virtuozzo.com> wrote:

> Or may be to use wait_event (non-interruptible), since kthreads typically don't need signal handling.

Agreed. I did not notice this. interruptible sleep will result in
deadlock if a signal sent by accident.


>      if (!test_bit(KTHREAD_SHOULD_STOP, ...))  // true, so SKIP threadfn
>          ret = threadfn(data);                 // krpc_sender_fn is NEVER called
>      do_exit(-EINTR)

This hypothetical dereference was my finding, when I tracked what happens
if threads were never scheduled before death. But I did not know that actually
threadfn is not even called. Frankly, redundant check would not hurt, this code
is executed once. But yes, it looks like it is really redundant.