From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751682AbbJLQxu (ORCPT ); Mon, 12 Oct 2015 12:53:50 -0400 Received: from e39.co.us.ibm.com ([32.97.110.160]:45135 "EHLO e39.co.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750995AbbJLQxt (ORCPT ); Mon, 12 Oct 2015 12:53:49 -0400 X-IBM-Helo: d03dlp02.boulder.ibm.com X-IBM-MailFrom: paulmck@linux.vnet.ibm.com X-IBM-RcptTo: linux-kernel@vger.kernel.org Date: Mon, 12 Oct 2015 09:53:53 -0700 From: "Paul E. McKenney" To: Frederic Weisbecker Cc: Ingo Molnar , LKML , Chris Metcalf , Peter Zijlstra , Mike Galbraith , Dave Jones , Thomas Gleixner , Oleg Nesterov , Christoph Lameter , Alexey Dobriyan , Rik van Riel , Andrew Morton Subject: Re: [PATCH] nohz: Revert "nohz: Set isolcpus when nohz_full is set" Message-ID: <20151012165353.GF3910@linux.vnet.ibm.com> Reply-To: paulmck@linux.vnet.ibm.com References: <1444663283-30068-1-git-send-email-fweisbec@gmail.com> <20151012153202.GB3910@linux.vnet.ibm.com> <20151012162001.GA32228@lerouge> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20151012162001.GA32228@lerouge> User-Agent: Mutt/1.5.21 (2010-09-15) X-TM-AS-MML: disable X-Content-Scanned: Fidelis XPS MAILER x-cbid: 15101216-0033-0000-0000-000006578A31 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Oct 12, 2015 at 06:20:03PM +0200, Frederic Weisbecker wrote: > On Mon, Oct 12, 2015 at 08:32:02AM -0700, Paul E. McKenney wrote: > > On Mon, Oct 12, 2015 at 05:21:23PM +0200, Frederic Weisbecker wrote: > > > This reverts commit 8cb9764fc88b41db11f251e8b2a0d006578b7eb4. > > > > > > We assumed that nohz full users always want scheduler isolation on full > > > dynticks CPUs, therefore we included nohz full CPUs on cpu_isolated_map. > > > This means that tasks run by default on CPUs outside the nohz_full range > > > unless their affinity is explicity overwritten. > > > > > > This suits pure isolation workloads but when the machine is needed to > > > run common workloads, the available sets of CPUs to run common tasks > > > becomes reduced. > > > > > > We reach an extreme case when CONFIG_NO_HZ_FULL_ALL is enabled as it > > > leaves only CPU 0 for non-isolation tasks, which makes people think that > > > their supercomputer regressed to 90's UP. > > > > > > Some nohz full users appear to be interested in running normal workloads > > > either before or after an isolation workload. Nohz full isn't optimized > > > toward normal workloads but it's still better than UP performance. > > > > > > We are reaching a limitation in kernel presets here. Lets revert this > > > cpu_isolated_map inclusion and let userspace do its own scheduler > > > isolation using cpusets or explicit affinity settings. > > > > > > Reported-by: Ingo Molnar > > > Reported-by: Mike Galbraith > > > Cc: Chris Metcalf > > > Cc: Rik van Riel > > > Cc: Christoph Lameter > > > Cc: Mike Galbraith > > > Cc: Peter Zijlstra > > > Cc: Dave Jones > > > Cc: Oleg Nesterov > > > Cc: Paul E. McKenney > > > Cc: Thomas Gleixner > > > Cc: Ingo Molnar > > > Cc: Alexey Dobriyan > > > Cc: Andrew Morton > > > Signed-off-by: Frederic Weisbecker > > > --- > > > kernel/sched/core.c | 3 --- > > > 1 file changed, 3 deletions(-) > > > > > > diff --git a/kernel/sched/core.c b/kernel/sched/core.c > > > index 6159531..3c35b5f 100644 > > > --- a/kernel/sched/core.c > > > +++ b/kernel/sched/core.c > > > @@ -7238,9 +7238,6 @@ void __init sched_init_smp(void) > > > alloc_cpumask_var(&non_isolated_cpus, GFP_KERNEL); > > > alloc_cpumask_var(&fallback_doms, GFP_KERNEL); > > > > > > - /* nohz_full won't take effect without isolating the cpus. */ > > > - tick_nohz_full_add_cpus_to(cpu_isolated_map); > > > - > > > > Why not make this controlled by a boot parameter? That preserves > > the ease of use for those needing it, but avoids problems from people > > doing "make randconfig". > > Well it is already. As you pass nohz_full=1-32, you can pass as well isolcpus=1-32 True enough. Not sure that having to repeat the CPU list twice qualifies as "easy to use", though. Why not a nohz_full_iso or some such that isolates whatever CPUs you specified? But we really need people using this to weigh in. Thanx, Paul