Re: [PATCH] locking/osq: Drop the overload of osq lock

[Boqun Feng <boqun.feng@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1419 bytes --]

On Sat, Jun 25, 2016 at 06:15:40PM +0200, Peter Zijlstra wrote:
> On Sat, Jun 25, 2016 at 11:21:30PM +0800, Boqun Feng wrote:
> > So on PPC, we have lppaca::yield_count to detect when an vcpu is
> > preempted, if the yield_count is even, the vcpu is running, otherwise it
> > is preempted(__spin_yield() is a user of this).
> > 
> > Therefore it makes more sense we
> > 
> > 		if (need_resched() || vcpu_is_preempted(old))
> > 
> > here, and implement vcpu_is_preempted() on PPC as
> > 
> > bool vcpu_is_preempted(int cpu)
> > {
> > 	return !!(be32_to_cpu(lppaca_of(cpu).yield_count) & 1)
> > }
> > 
> > Thoughts?
> 
> Would that not have issues where the owner cpu is kept running but the
> spinner (ie. _this_ vcpu) gets preempted? I would think that in that
> case we too want to stop spinning.
> 

I don't think we want(or need) to stop the spinning of _this_ vcpu in
that case? Because it has already been preempted, when it gets back to
run, the owner may still be running and haven't set ->locked to 1 yet,
which means spinning on this vcpu is still worthwhile.

I think the proper logic here is that in the optimistic spin queue, if
any one found its predecessor's vcpu was preempted, it should stop
spinning, because it's very likely that it would not see ->locked
becoming 1 in a short time.

> Although, if all vcpus are scheduled equal, it might not matter on
> average.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 473 bytes --]