On Mon, Jun 12, 2023 at 9:42 AM Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>
> Well, it makes sense for the bug, but that whole
>
> preempt_disable();
> current->flags &= ~PF_IO_WORKER;
> preempt_enable();
>
> thin in io_worker_exit() does *not* make sense to me.
Oh, it looks like that goes back to the original io_worker
implementation, when io_worker_start() would set the PF_IO_WORKER bit,
and io_worker_exit() would clear it, and they were otherwise created
like normal kthreads.
It *looks* all entirely historical to me, and removing those three
lines as left-over baggage smells like the right thing.
Linus
