On Mon 10-12-18 15:47:11, Peter Zijlstra wrote:
> On Mon, Dec 10, 2018 at 03:13:37PM +0100, Michal Hocko wrote:
> > I do not see any scheduler guys Cced and it would be really great to get
> > their opinion here.
> >
> > On Mon 10-12-18 11:36:39, Daniel Vetter wrote:
> > > In some special cases we must not block, but there's not a
> > > spinlock, preempt-off, irqs-off or similar critical section already
> > > that arms the might_sleep() debug checks. Add a non_block_start/end()
> > > pair to annotate these.
> > >
> > > This will be used in the oom paths of mmu-notifiers, where blocking is
> > > not allowed to make sure there's forward progress.
> >
> > Considering the only alternative would be to abuse
> > preempt_{disable,enable}, and that really has a different semantic, I
> > think this makes some sense. The cotext is preemptible but we do not
> > want notifier to sleep on any locks, WQ etc.
>
> I'm confused... what is this supposed to do?
>
> And what does 'block' mean here? Without preempt_disable/IRQ-off we're
> subject to regular preemption and execution can stall for arbitrary
> amounts of time.
The notifier is called from quite a restricted context - oom_reaper -
which shouldn't depend on any locks or sleepable conditionals. The code
should be swift as well but we mostly do care about it to make a forward
progress. Checking for sleepable context is the best thing we could come
up with that would describe these demands at least partially.
--
Michal Hocko
SUSE Labs
