From: "Li, Aubrey" <aubrey.li@linux.intel.com>
To: "Song Bao Hua (Barry Song)" <song.bao.hua@hisilicon.com>,
Vincent Guittot <vincent.guittot@linaro.org>
Cc: Ingo Molnar <mingo@redhat.com>,
Peter Zijlstra <peterz@infradead.org>,
Juri Lelli <juri.lelli@redhat.com>,
Mel Gorman <mgorman@techsingularity.net>,
Valentin Schneider <valentin.schneider@arm.com>,
Qais Yousef <qais.yousef@arm.com>,
Dietmar Eggemann <dietmar.eggemann@arm.com>,
Steven Rostedt <rostedt@goodmis.org>,
Ben Segall <bsegall@google.com>,
Tim Chen <tim.c.chen@linux.intel.com>,
linux-kernel <linux-kernel@vger.kernel.org>,
Mel Gorman <mgorman@suse.de>, Jiang Biao <benbjiang@gmail.com>
Subject: Re: [RFC PATCH v8] sched/fair: select idle cpu from idle cpumask for task wakeup
Date: Tue, 15 Dec 2020 20:41:29 +0800 [thread overview]
Message-ID: <d3184ead-b82f-f784-4c07-9e36d4cc630c@linux.intel.com> (raw)
In-Reply-To: <121565627e944f8e9dde4080d19d5b02@hisilicon.com>
Hi Bao Hua,
Sorry I almost missed this message, :(
On 2020/12/14 7:29, Song Bao Hua (Barry Song) wrote:
>
> Hi Aubrey,
>
> The patch looks great. But I didn't find any hackbench improvement
> on kunpeng 920 which has 24 cores for each llc span. Llc span is also
> one numa node. The topology is like:
> # numactl --hardware
> available: 4 nodes (0-3)
> node 0 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
> node 0 size: 128669 MB
> node 0 free: 126995 MB
> node 1 cpus: 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
> 43 44 45 46 47
> node 1 size: 128997 MB
> node 1 free: 127539 MB
> node 2 cpus: 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
> 67 68 69 70 71
> node 2 size: 129021 MB
> node 2 free: 127106 MB
> node 3 cpus: 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90
> 91 92 93 94 95
> node 3 size: 127993 MB
> node 3 free: 126739 MB
> node distances:
> node 0 1 2 3
> 0: 10 12 20 22
> 1: 12 10 22 24
> 2: 20 22 10 12
> 3: 22 24 12 10
>
> Benchmark command:
> numactl -N 0-1 hackbench -p -T -l 20000 -g $1
>
> for each g, I ran 10 times to get the average time. And I tested
> g from 1 to 10.
>
> g 1 2 3 4 5 6 7 8 9 10
> w/o 1.4733 1.5992 1.9353 2.1563 2.8448 3.3305 3.9616 4.4870 5.0786 5.6983
> w/ 1.4709 1.6152 1.9474 2.1512 2.8298 3.2998 3.9472 4.4803 5.0462 5.6505
>
> Is it because the core number is small in llc span in my test?
I guess it is with SIS_PROP, when the system is very busy that idle cpu scan
loop is throttled by nr(4). The patch actually reduces 4 times scan so the
data change looks marginal. Vincent mentioned a notable change at here:
https://lkml.org/lkml/2020/12/14/109
Maybe you can increase the group number to see if it can be reproduced on
your side.
Thanks,
-Aubrey
next prev parent reply other threads:[~2020-12-15 12:43 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-12-10 1:43 [RFC PATCH v8] sched/fair: select idle cpu from idle cpumask for task wakeup Aubrey Li
2020-12-11 15:07 ` Vincent Guittot
2020-12-11 15:18 ` Li, Aubrey
2020-12-11 15:22 ` Vincent Guittot
2020-12-11 15:24 ` Li, Aubrey
2020-12-13 23:29 ` Song Bao Hua (Barry Song)
2020-12-15 12:41 ` Li, Aubrey [this message]
2021-03-04 13:51 ` Li, Aubrey
2021-03-08 11:30 ` Vincent Guittot
2021-03-08 13:50 ` Li, Aubrey
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=d3184ead-b82f-f784-4c07-9e36d4cc630c@linux.intel.com \
--to=aubrey.li@linux.intel.com \
--cc=benbjiang@gmail.com \
--cc=bsegall@google.com \
--cc=dietmar.eggemann@arm.com \
--cc=juri.lelli@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mgorman@suse.de \
--cc=mgorman@techsingularity.net \
--cc=mingo@redhat.com \
--cc=peterz@infradead.org \
--cc=qais.yousef@arm.com \
--cc=rostedt@goodmis.org \
--cc=song.bao.hua@hisilicon.com \
--cc=tim.c.chen@linux.intel.com \
--cc=valentin.schneider@arm.com \
--cc=vincent.guittot@linaro.org \
/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).