Re: wait_on_page_bit_common(TASK_KILLABLE, EXCLUSIVE) can miss wakeup?

[Oleg Nesterov]

On 06/30, Linus Torvalds wrote:
>
> On Tue, Jun 30, 2020 at 4:51 AM Oleg Nesterov <oleg@xxxxxxxxxx> wrote:
> >
> > On 06/30, Oleg Nesterov wrote:
> > >
> > > may be someting like this
> >
> > or this ...
>
> Guys, I'd suggest we either
>
>  (a) do the minimal ugly one-liner workaround
>
> or
>
>  (b) do something *much* more radical.

I agree, and let me repeat I think your one-liner is fine.

I just tried to discuss the possible alternatives to that on-liner for the
start.

>    the wake_page_function() would actually do

I swear, I too thought about changing wake_page_function() but got lost ;)

The very fact it returns -1 if test_bit() is true suggests it can do more.

Oleg.