Re: [PATCH RFC v3] x86,mm,sched: make lazy TLB mode even lazier

[Rik van Riel <riel@redhat.com>]

[-- Attachment #1: Type: text/plain, Size: 1225 bytes --]

On Sat, 2016-08-27 at 10:03 +0200, Ingo Molnar wrote:
> * Rik van Riel <riel@redhat.com> wrote:
> 
> > On Thu, 25 Aug 2016 12:42:15 -0700
> > "H. Peter Anvin" <hpa@zytor.com> wrote:
> > 
> > > Why grabbing a lock instead of cmpxchg?
> > 
> > ... and some more cleanups later, this might actually be
> > good to merge, assuming it works for Benjamin :)
> > 
> > ---8<---
> 
> LGTM in principle (it's a pretty clever trick!), just some minor
> stylistic nits:

Thanks for the review. I have applied the stylistic nits, and
turned lazy_tlb_can_skip_flush into a big switch statement as
suggested by Linus.

> > +	 */
> > +	for_each_cpu(cpu, cpumask)
> > +		if (lazy_tlb_can_skip_flush(cpu))
> > +			cpumask_clear_cpu(cpu, (struct cpumask
> > *)cpumask);
> 
> Please remove the 'const' from the cpumask type definition instead of
> this ugly 
> cast!

I played around with this on Thursday already, and ran out of
time to clean that up before going to the next talk. This will
be fixed in the next version. 

> I'd also like to wait for the Tested-by from Benjamin as well before
> we can 
> proceeed.

Agreed.

Ben, a new version is coming up real soon.

-- 

All Rights Reversed.

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 473 bytes --]