From: "Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com>
To: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: mingo@kernel.org, pjt@google.com, paul@paulmenage.org,
akpm@linux-foundation.org, rjw@sisk.pl, nacc@us.ibm.com,
paulmck@linux.vnet.ibm.com, tglx@linutronix.de,
seto.hidetoshi@jp.fujitsu.com, rob@landley.net, tj@kernel.org,
mschmidt@redhat.com, berrange@redhat.com,
nikunj@linux.vnet.ibm.com, vatsa@linux.vnet.ibm.com,
linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
linux-pm@vger.kernel.org
Subject: Re: [PATCH v2 0/7] CPU hotplug, cpusets: Fix issues with cpusets handling upon CPU hotplug
Date: Sat, 05 May 2012 01:28:33 +0530 [thread overview]
Message-ID: <4FA434E9.6000305@linux.vnet.ibm.com> (raw)
In-Reply-To: <1336159496.6509.51.camel@twins>
On 05/05/2012 12:54 AM, Peter Zijlstra wrote:
>
>> Documentation/cgroups/cpusets.txt | 43 +++--
>> include/linux/cpuset.h | 4
>> kernel/cpuset.c | 317 ++++++++++++++++++++++++++++---------
>> kernel/sched/core.c | 4
>> 4 files changed, 274 insertions(+), 94 deletions(-)
>
> Bah, I really hate this complexity you've created for a problem that
> really doesn't exist.
>
Doesn't exist? Well, I believe we do have a problem and a serious one
at that too!
The heart of the problem can be summarized in 2 sentences:
o During a CPU hotplug, tasks can move between cpusets, and never
come back to their original cpuset.
o Tasks might get pinned to lesser number of cpus, unreasonably.
Both these are undesirable from a system-admin point of view.
Moreover, having workarounds for this from userspace is way too messy and
ugly, if not impossible.
> So why not fix the active mask crap?
Because I doubt if that is the right way to approach this problem.
An updated cpu_active_mask not being the necessary and sufficient condition
for all scheduler related activities, is a different problem altogether, IMHO.
(Btw, Ingo had also suggested reworking this whole cpuset thing, while
reviewing the previous version of this fix.
http://thread.gmane.org/gmane.linux.kernel/1250097/focus=1252133)
Also, we need to fix this problem at the CPU Hotplug level itself, and
not just for the suspend/resume case. Because, we have had numerous bug
reports and people complaining about this issue, in various scenarios,
including those that didn't involve suspend/resume.
I am sure some of the people in Cc will have more to add to this, but in
general, when the CPU hotplug (maybe even cpu offline + online) and the
cpuset administration are done asynchronously, it leads to nasty surprises.
In fact, there have been reports where people spent inordinate amounts of
time before they figured out that a long-forgotten cpu hotplug operation
which was performed, was the root-cause of a low-performing workload!.
All these only suggest that it is time that we cleaned this up thoroughly,
and at the root cause level itself.
Btw, though there are 7 patches in this series, I don't think this patchset
increases the complexity of the code.. In fact, it makes many things simpler
and saner/cleaner, IMHO.
Regards,
Srivatsa S. Bhat
next prev parent reply other threads:[~2012-05-04 19:59 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-05-04 19:17 [PATCH v2 0/7] CPU hotplug, cpusets: Fix issues with cpusets handling upon CPU hotplug Srivatsa S. Bhat
2012-05-04 19:17 ` [PATCH v2 1/7] cpusets, hotplug: Implement cpuset tree traversal in a helper function Srivatsa S. Bhat
2012-05-04 19:18 ` [PATCH v2 2/7] cpusets, hotplug: Restructure functions that are invoked during hotplug Srivatsa S. Bhat
2012-05-04 19:18 ` [PATCH v2 3/7] cpusets: Introduce 'user_cpus_allowed' and rework the semantics of 'cpus_allowed' Srivatsa S. Bhat
2012-05-04 19:19 ` [PATCH v2 4/7] CPU hotplug, cpusets: Workout hotplug handling for cpusets Srivatsa S. Bhat
2012-05-04 19:19 ` [PATCH v2 5/7] Docs, cpusets: Update the cpuset documentation Srivatsa S. Bhat
2012-05-04 22:28 ` Rob Landley
2012-05-04 19:20 ` [PATCH v2 6/7] cpusets: Optimize the implementation of guarantee_online_cpus() Srivatsa S. Bhat
2012-05-04 19:20 ` [PATCH v2 7/7] cpusets: Remove out-dated comment about cpuset_track_online_cpus Srivatsa S. Bhat
2012-05-04 19:24 ` [PATCH v2 0/7] CPU hotplug, cpusets: Fix issues with cpusets handling upon CPU hotplug Peter Zijlstra
2012-05-04 19:58 ` Srivatsa S. Bhat [this message]
2012-05-04 20:14 ` Peter Zijlstra
2012-05-04 20:28 ` Peter Zijlstra
2012-05-04 20:49 ` Nishanth Aravamudan
2012-05-04 21:01 ` Peter Zijlstra
2012-05-04 21:27 ` Nishanth Aravamudan
2012-05-04 21:32 ` Peter Zijlstra
2012-05-04 21:34 ` Peter Zijlstra
2012-05-04 21:57 ` Nishanth Aravamudan
2012-05-04 21:38 ` Peter Zijlstra
2012-05-04 20:46 ` Nishanth Aravamudan
2012-05-04 20:56 ` Peter Zijlstra
2012-05-04 21:30 ` Nishanth Aravamudan
2012-05-04 21:44 ` Peter Zijlstra
2012-05-05 15:24 ` Alan Stern
2012-05-05 17:44 ` Paul E. McKenney
2012-05-05 18:56 ` Rafael J. Wysocki
2012-05-08 13:07 ` Daniel P. Berrange
2012-05-05 4:39 ` Mike Galbraith
2012-05-05 17:15 ` Srivatsa S. Bhat
2012-05-07 15:26 ` Jiang Liu
2012-05-09 9:12 ` Srivatsa S. Bhat
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4FA434E9.6000305@linux.vnet.ibm.com \
--to=srivatsa.bhat@linux.vnet.ibm.com \
--cc=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=berrange@redhat.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pm@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=mschmidt@redhat.com \
--cc=nacc@us.ibm.com \
--cc=nikunj@linux.vnet.ibm.com \
--cc=paul@paulmenage.org \
--cc=paulmck@linux.vnet.ibm.com \
--cc=pjt@google.com \
--cc=rjw@sisk.pl \
--cc=rob@landley.net \
--cc=seto.hidetoshi@jp.fujitsu.com \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
--cc=vatsa@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).