Re: [PATCH v2 2/7] xen: credit: (micro) optimize csched_runq_steal().

[Dario Faggioli <dario.faggioli@citrix.com>]

[-- Attachment #1.1: Type: text/plain, Size: 1500 bytes --]

On Fri, 2017-04-07 at 11:49 +0100, George Dunlap wrote:
> On 06/04/17 09:16, Dario Faggioli wrote:
> > --- a/xen/common/sched_credit.c
> > +++ b/xen/common/sched_credit.c
> > @@ -708,12 +708,10 @@ static inline int
> >  __csched_vcpu_is_migrateable(struct vcpu *vc, int dest_cpu,
> > cpumask_t *mask)
> >  {
> >      /*
> > -     * Don't pick up work that's in the peer's scheduling tail or
> > hot on
> > -     * peer PCPU. Only pick up work that prefers and/or is allowed
> > to run
> > -     * on our CPU.
> > +     * Don't pick up work that's hot on peer PCPU, or that can't
> > (or
> > +     * would prefer not to) run on cpu.
> >       */
> > -    return !vc->is_running &&
> > -           !__csched_vcpu_is_cache_hot(vc) &&
> > +    return !__csched_vcpu_is_cache_hot(vc) &&
> >             cpumask_test_cpu(dest_cpu, mask);
> 
> +1 to moving the check earlier; but there's a risk that this function
> will be used somewhere else and that the is_running check will be
> missed.
> 
> Should we maybe add an ASSERT() here (perhaps with a comment that the
> check can be added back in if necessary)?
> 
Ok, makes sense. Will do.

Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Senior Software Engineer, Citrix Systems R&D Ltd., Cambridge (UK)

[-- Attachment #1.2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

[-- Attachment #2: Type: text/plain, Size: 127 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
https://lists.xen.org/xen-devel