From: Tejun Heo <tj@kernel.org>
To: Michal Hocko <mhocko@suse.cz>
Cc: lizefan@huawei.com, containers@lists.linux-foundation.org,
cgroups@vger.kernel.org, koverstreet@google.com,
linux-kernel@vger.kernel.org, cl@linux-foundation.org,
Mike Snitzer <snitzer@redhat.com>,
Vivek Goyal <vgoyal@redhat.com>,
"Alasdair G. Kergon" <agk@redhat.com>,
Jens Axboe <axboe@kernel.dk>,
Mikulas Patocka <mpatocka@redhat.com>,
Glauber Costa <glommer@gmail.com>
Subject: Re: [PATCH 11/11] cgroup: use percpu refcnt for cgroup_subsys_states
Date: Fri, 14 Jun 2013 23:31:55 -0700 [thread overview]
Message-ID: <20130615063155.GC7017@htj.dyndns.org> (raw)
In-Reply-To: <20130615053522.GA7017@htj.dyndns.org>
On Fri, Jun 14, 2013 at 10:35:22PM -0700, Tejun Heo wrote:
> On Fri, Jun 14, 2013 at 03:31:25PM -0700, Tejun Heo wrote:
> > I'll play with it a bit more on an actual machine and post more
> > results. Test program attached.
>
> So, here are the results from the same test on a dual-socket 2-way
> NUMA opteron 8 core machine.
>
> Running on one CPU.
>
> copy size atomic percpu diff in pct
> 0 535964443 616756827 +15.07%
> 32 399988186 378678713 -5.33%
> 64 389067476 355073979 -8.74%
> 128 342192631 315615300 -7.77%
> 256 281208005 260598931 -7.33%
> 512 188070912 193225269 +2.74%
>
> Running on all eight cores.
>
> copy size atomic percpu diff in pct
> 0 121324328 4889425511 +3,930.05%
> 32 96170193 2999613380 +3,019.07%
> 64 98139061 2813894184 +2,767.25%
> 128 112610025 2503229487 +2,122.92%
> 256 96828114 2069865752 +2,037.67%
> 512 95858297 1537726109 +1,504.17%
>
> Ration of all cores / single core.
>
> copy size atomic percpu
> 0 0.23 7.93
> 32 0.24 7.92
> 64 0.25 7.92
> 128 0.33 7.93
> 256 0.34 7.94
> 512 0.51 7.96
I was testing with CONFIG_PREEMPT, which makes rcu_read_[un]lock()s
quite a bit more expensive. The following is the same test results
with CONFIG_PREEMPT_VOLUNTARY which would the most preemptive server
distros would get anyway.
One CPU.
copy size atomic percpu diff in pct
0 534583387 1521561724 +184.63%
32 399098138 615962137 + 54.34%
64 388128431 555599274 + 43.15%
128 341336474 464502792 + 36.08%
256 280471681 354186740 + 26.28%
512 203784802 240067596 + 17.80%
All eight CPUs.
copy size atomic percpu diff in pct
0 117213982 12488998111 +10,554.87%
32 103545751 4940695158 + 4,671.51%
64 98135094 4456370409 + 4,441.06%
128 117729659 3725434154 + 3,064.40%
256 95916768 2840992396 + 2,861.94%
512 95795993 1926044518 + 1,910.57%
Ration of all cores / single core.
copy size atomic percpu
0 0.22 8.21
32 0.26 8.02
64 0.25 8.02
128 0.34 8.02
256 0.34 8.02
512 0.47 8.02
So, it's faster even with only one CPU.
Thanks.
--
tejun
next prev parent reply other threads:[~2013-06-15 6:32 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-13 4:04 [PATCHSET v2 cgroup/for-3.11] cgroup: convert cgroup_subsys_state refcnt to percpu_ref Tejun Heo
2013-06-13 4:04 ` [PATCH 01/11] cgroup: remove now unused css_depth() Tejun Heo
2013-06-13 4:04 ` [PATCH 02/11] cgroup: consistently use @cset for struct css_set variables Tejun Heo
2013-06-13 4:04 ` [PATCH 03/11] cgroup: bring some sanity to naming around cg_cgroup_link Tejun Heo
2013-06-13 4:04 ` [PATCH 04/11] cgroup: use kzalloc() instead of kmalloc() Tejun Heo
2013-06-13 4:04 ` [PATCH 05/11] cgroup: clean up css_[try]get() and css_put() Tejun Heo
2013-06-13 4:04 ` [PATCH 06/11] cgroup: rename CGRP_REMOVED to CGRP_DEAD Tejun Heo
2013-06-13 4:04 ` [PATCH 07/11] cgroup: drop unnecessary RCU dancing from __put_css_set() Tejun Heo
2013-06-13 4:04 ` [PATCH 08/11] cgroup: remove cgroup->count and use Tejun Heo
2013-06-13 4:04 ` [PATCH 09/11] cgroup: reorder the operations in cgroup_destroy_locked() Tejun Heo
2013-06-13 4:04 ` [PATCH 10/11] cgroup: split cgroup destruction into two steps Tejun Heo
2013-06-13 4:04 ` [PATCH 11/11] cgroup: use percpu refcnt for cgroup_subsys_states Tejun Heo
2013-06-13 23:16 ` Kent Overstreet
2013-06-14 12:55 ` Michal Hocko
2013-06-14 14:15 ` Glauber Costa
2013-06-14 14:22 ` Michal Hocko
2013-06-14 13:20 ` Michal Hocko
2013-06-14 22:31 ` Tejun Heo
2013-06-15 5:35 ` Tejun Heo
2013-06-15 5:39 ` Tejun Heo
2013-06-15 6:31 ` Tejun Heo [this message]
2013-06-17 13:27 ` Michal Hocko
2013-06-17 17:16 ` Tejun Heo
2013-06-13 6:04 ` [PATCHSET v2 cgroup/for-3.11] cgroup: convert cgroup_subsys_state refcnt to percpu_ref Li Zefan
2013-06-13 17:56 ` Tejun Heo
2013-06-14 2:41 ` Tejun Heo
-- strict thread matches above, loose matches on Subject: below --
2013-06-12 21:03 [PATCHSET " Tejun Heo
2013-06-12 21:03 ` [PATCH 11/11] cgroup: use percpu refcnt for cgroup_subsys_states Tejun Heo
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=20130615063155.GC7017@htj.dyndns.org \
--to=tj@kernel.org \
--cc=agk@redhat.com \
--cc=axboe@kernel.dk \
--cc=cgroups@vger.kernel.org \
--cc=cl@linux-foundation.org \
--cc=containers@lists.linux-foundation.org \
--cc=glommer@gmail.com \
--cc=koverstreet@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=lizefan@huawei.com \
--cc=mhocko@suse.cz \
--cc=mpatocka@redhat.com \
--cc=snitzer@redhat.com \
--cc=vgoyal@redhat.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