From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcelo Tosatti Subject: Re: [PATCH v8 0/6] cgroup/cpuset: Add new cpuset partition type & empty effecitve cpus Date: Wed, 10 Nov 2021 13:14:29 -0300 Message-ID: <20211110161429.GA22855@fuller.cnet> References: <20211018143619.205065-1-longman@redhat.com> <20211110111357.17617-1-felix.moessbauer@siemens.com> <20211110135653.GD20566@blackbody.suse.cz> <20211110161020.GA20101@fuller.cnet> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Return-path: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1636560916; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=pujqTrjEVvmX1Jx376eksYnuy7UjFSMiFW1jzdmHKRE=; b=HKyWsFTy5AbnOGPcwQnm/29BaJ3R93f7cbU9kIxrHEirA+OFSH4zgICWg6CCJNUc76Gfer qz0O6jCs8Zr3Mp88KjTmwTg4fu4JJth7lcyv1JWioggh9w7wAfdid1u2cFP6hIJfekyLv0 /jtjTkmZN7WLCTENVv2eBIGJVQkWPTI= Content-Disposition: inline In-Reply-To: <20211110161020.GA20101@fuller.cnet> List-ID: Content-Type: text/plain; charset="iso-8859-1" To: "Moessbauer, Felix" Cc: Michal =?iso-8859-1?Q?Koutn=FD?= , "longman@redhat.com" , "akpm@linux-foundation.org" , "cgroups@vger.kernel.org" , "corbet@lwn.net" , "frederic@kernel.org" , "guro@fb.com" , "hannes@cmpxchg.org" , "juri.lelli@redhat.com" , "linux-doc@vger.kernel.org" , "linux-kernel@vger.kernel.org" , "linux-kselftest@vger.kernel.org" , "lizefan.x@bytedance.com" , "pauld@redhat.com" , "peterz@infradead.org" , "shuah@kernel.org" On Wed, Nov 10, 2021 at 01:10:20PM -0300, Marcelo Tosatti wrote: > On Wed, Nov 10, 2021 at 03:21:54PM +0000, Moessbauer, Felix wrote: > >=20 > >=20 > > > -----Original Message----- > > > From: Michal Koutn=FD > > > Sent: Wednesday, November 10, 2021 2:57 PM > > > To: Moessbauer, Felix (T RDA IOT SES-DE) > > > Cc: longman@redhat.com; akpm@linux-foundation.org; > > > cgroups@vger.kernel.org; corbet@lwn.net; frederic@kernel.org; guro@fb= .com; > > > hannes@cmpxchg.org; juri.lelli@redhat.com; linux-doc@vger.kernel.org;= linux- > > > kernel@vger.kernel.org; linux-kselftest@vger.kernel.org; > > > lizefan.x@bytedance.com; mtosatti@redhat.com; pauld@redhat.com; > > > peterz@infradead.org; shuah@kernel.org; tj@kernel.org; Kiszka, Jan (T= RDA > > > IOT) ; Schild, Henning (T RDA IOT SES-DE) > > > > > > Subject: Re: [PATCH v8 0/6] cgroup/cpuset: Add new cpuset partition t= ype & > > > empty effecitve cpus > > >=20 > > > Hello. > > >=20 > > > On Wed, Nov 10, 2021 at 12:13:57PM +0100, Felix Moessbauer > > > wrote: > > > > However, I was not able to see any latency improvements when using > > > > cpuset.cpus.partition=3Disolated. > > >=20 > > > Interesting. What was the baseline against which you compared it (iso= lcpus, no > > > cpusets,...)? > >=20 > > For this test, I just compared both settings cpuset.cpus.partition=3Dis= olated|root. > > There, I did not see a significant difference (but I know, RT tuning de= pends on a ton of things). > >=20 > > >=20 > > > > The test was performed with jitterdebugger on CPUs 1-3 and the foll= owing > > > cmdline: > > > > rcu_nocbs=3D1-4 nohz_full=3D1-4 irqaffinity=3D0,5-6,11 intel_pstate= =3Ddisable > > > > On the other cpus, stress-ng was executed to generate load. > > > > [...] > > >=20 > > > > This requires cgroup.type=3Dthreaded on both cgroups and changes to= the > > > > application (threads have to be born in non-rt group and moved to r= t-group). > > >=20 > > > But even with isolcpus the application would need to set affinity of = threads to > > > the selected CPUs (cf cgroup migrating). Do I miss anything? > >=20 > > Yes, that's true. But there are two differences (given that you use iso= lcpus): > > 1. the application only has to set the affinity for rt threads. > > Threads that do not explicitly set the affinity are automatically excl= uded from the isolated cores. > > Even common rt test applications like jitterdebugger do not pin their = non-rt threads. > > 2. Threads can be started on non-rt CPUs and then bound to a specific r= t CPU. > > This binding can be specified before thread creation via pthread_create. > > By that, you can make sure that at no point in time a thread has a "for= bidden" CPU in its affinities. > >=20 > > With cgroup2, you cannot guarantee the second aspect, as thread creatio= n and moving to a cgroup is not an atomic operation. > > Also - please correct me if I'm wrong - you first have to create a thre= ad before moving it into a group. > > At creation time, you cannot set the final affinity mask (as you create= it in the non-rt group and there the CPU is not in the cpuset.cpus). > > Once you move the thread to the rt cgroup, it has a default mask and by= that can be executed on other rt cores. >=20 > man clone3: >=20 > CLONE_NEWCGROUP (since Linux 4.6) > Create the process in a new cgroup namespace. If this= flag is not set, then (as with fork(2)) the > process is created in the same cgroup namespaces as the cal= ling process. >=20 > For further information on cgroup namespaces, see cgroup_na= mespaces(7). >=20 > Only a privileged process (CAP_SYS_ADMIN) can employ CLONE_= NEWCGROUP. >=20 Err, CLONE_INTO_CGROUP.