linux-kernel - Re: [PATCH v3 0/3] sched/topology: optimize topology_span

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <ZyzbRKEtOj54mdJQ@yury-ThinkPad>
Date: Thu, 7 Nov 2024 07:22:44 -0800
From: Yury Norov <yury.norov@...il.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: linux-kernel@...r.kernel.org,
	Christophe JAILLET <christophe.jaillet@...adoo.fr>,
	Chen Yu <yu.c.chen@...el.com>, Leonardo Bras <leobras@...hat.com>,
	Ingo Molnar <mingo@...hat.com>, Juri Lelli <juri.lelli@...hat.com>,
	Vincent Guittot <vincent.guittot@...aro.org>,
	Dietmar Eggemann <dietmar.eggemann@....com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Ben Segall <bsegall@...gle.com>, Mel Gorman <mgorman@...e.de>,
	Valentin Schneider <vschneid@...hat.com>
Subject: Re: [PATCH v3 0/3] sched/topology: optimize topology_span_sane()

On Wed, Nov 06, 2024 at 07:06:13PM +0100, Peter Zijlstra wrote:
> On Mon, Sep 02, 2024 at 11:36:04AM -0700, Yury Norov wrote:
> > The function may call cpumask_equal with tl->mask(cpu) == tl->mask(i),
> > even when cpu != i. In such case, cpumask_equal() would always return
> > true, and we can proceed to the next iteration immediately.
> > 
> > Valentin Schneider shares on it:
> > 
> >   PKG can potentially hit that condition, and so can any
> >   sched_domain_mask_f that relies on the node masks...
> > 
> >   I'm thinking ideally we should have checks in place to
> >   ensure all node_to_cpumask_map[] masks are disjoint,
> >   then we could entirely skip the levels that use these
> >   masks in topology_span_sane(), but there's unfortunately
> >   no nice way to flag them... Also there would be cases
> >   where there's no real difference between PKG and NODE
> >   other than NODE is still based on a per-cpu cpumask and
> >   PKG isn't, so I don't see a nicer way to go about this.
> > 
> > v1: https://lore.kernel.org/lkml/ZrJk00cmVaUIAr4G@yury-ThinkPad/T/
> > v2: https://lkml.org/lkml/2024/8/7/1299
> > v3:
> >  - add topology_cpumask_equal() helper in #3;
> >  - re-use 'cpu' as an iterator int the for_each_cpu() loop;
> >  - add proper versioning for all patches.
> > 
> > Yury Norov (3):
> >   sched/topology: pre-compute topology_span_sane() loop params
> >   sched/topology: optimize topology_span_sane()
> >   sched/topology: reorganize topology_span_sane() checking order
> 
> Why are we doing this? Subject says optimize, but I see no performance
> numbers?

Well, if we have a chance that cpumask_equal() is passed with
mask1 == mask2, the new vs old approach is O(1) vs O(N), and this
is the case of PKG. This should have performance impact for many-CPUs
setups (much more than BITS_PER_LONG).

I have no such machine, and can't give you perf numbers. I can drop
'optimization' wording, if you find it confusing without numbers.

Thanks,
Yury