All of lore.kernel.org
 help / color / mirror / Atom feed
From: Fengguang Wu <fengguang.wu@intel.com>
To: Alex Shi <alex.shi@linaro.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	lkp@linux.intel.com
Subject: Re: [sched] 73628fba4: +69% context switches
Date: Thu, 16 Jan 2014 09:23:01 +0800	[thread overview]
Message-ID: <20140116012301.GA2535@localhost> (raw)
In-Reply-To: <52D3804A.8060700@linaro.org>

On Mon, Jan 13, 2014 at 01:57:30PM +0800, Alex Shi wrote:
> On 01/11/2014 06:19 PM, Fengguang Wu wrote:
> > Alex,
> > 
> > FYI, we find much increased interrupts and context switches introduced
> > by commit 73628fba4 ("sched: unify imbalance bias for target group")
> > in your noload branch:
> 
> Many thanks for the generous and quick testing! :)
> 
> few questions for the results and give a testing patch for try. it also push on github. 
> 
> What's about the aim7 shell_rtns_1 and shared throughput?

The throughputs remain the same.  We only report changed numbers.

> > 7bea8c18805a5f1  73628fba451ae72221b155696  
> > ---------------  -------------------------  
> >      14979 ~ 4%   +1304.8%     210434 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
> >       2748 ~ 5%    +977.4%      29607 ~ 0%  nhm-white/micro/aim7/shared
> >      17727        +1254.1%     240041       TOTAL interrupts.RES
> 
> RES interrupt increased about 200,000, but vmstat's interrupt
> increased a little. guess the vmstat data is per seconds, right? If

Yes.

> so, it is better give how long time the vmstat running.

vmstat runs throughout the aim7 test run.

> The same problem on the time.involuntary_context_switches and vmstat cs.

ditto.

> According to involuntary CS definition in time, RES interrupt will cause involuntary CS. but here 29607 RES of aim7/shared cause 233218 time inv CS, does sth I missed or the data is incorrect?
> 
> > 
> > 7bea8c18805a5f1  73628fba451ae72221b155696  
> > ---------------  -------------------------  
> >       3617 ~ 0%     +69.2%       6118 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
> >       3617          +69.2%       6118       TOTAL vmstat.system.in
> > 
> > 7bea8c18805a5f1  73628fba451ae72221b155696  
> > ---------------  -------------------------  
> >     132746 ~ 0%     +69.0%     224346 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
> >     220038 ~ 0%      +6.0%     233218 ~ 0%  nhm-white/micro/aim7/shared
> >     352785          +29.7%     457564       TOTAL time.involuntary_context_switches
> > 
> > 7bea8c18805a5f1  73628fba451ae72221b155696  
> > ---------------  -------------------------  
> >    1424581 ~ 0%      +8.6%    1546786 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
> >    1424581           +8.6%    1546786       TOTAL time.voluntary_context_switches
> > 
> > 7bea8c18805a5f1  73628fba451ae72221b155696  
> > ---------------  -------------------------  
> >      20982 ~ 0%     +12.5%      23599 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
> >       6005 ~ 0%      +4.2%       6256 ~ 0%  nhm-white/micro/aim7/shared
> >      26988          +10.6%      29856       TOTAL vmstat.system.cs
> > 
> >                                   vmstat.system.cs
> > 
> 
> commit c5a8778a132cfa882609fbccb4ee6542eac9866d
> Author: Alex Shi <alex.shi@linaro.org>
> Date:   Mon Jan 13 13:54:30 2014 +0800
> 
>     more bias towards local cpu group


In your rebased tree, there are still much increased interrupts and
context switches:

Comparing 91dd08bd65052b483f96f75ab ("sched: more bias towards local
cpu group") with upstream:

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
        48 ~13%    +485.4%        281 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
        48         +485.4%        281       TOTAL numa-vmstat.node1.nr_shmem

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     10142 ~ 4%   +2737.8%     287830 ~ 2%  lkp-ne04/micro/aim7/shell_rtns_1
      1349 ~16%   +1479.3%      21319 ~ 4%  nhm-white/micro/aim7/dbase
     11492        +2590.0%     309149       TOTAL interrupts.RES

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
       194 ~12%    +472.9%       1115 ~ 2%  lkp-ne04/micro/aim7/shell_rtns_1
       194         +472.9%       1115       TOTAL numa-meminfo.node1.Shmem

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
    118634 ~ 2%     -73.0%      32081 ~ 0%  nhm-white/micro/aim7/dbase
    118634          -73.0%      32081       TOTAL interrupts.IWI

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
       327 ~ 1%     -71.0%         95 ~ 4%  lkp-ne04/micro/aim7/shell_rtns_1
       327          -71.0%         95       TOTAL numa-vmstat.node0.nr_shmem

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
      1312 ~ 1%     -70.8%        383 ~ 4%  lkp-ne04/micro/aim7/shell_rtns_1
      1312          -70.8%        383       TOTAL numa-meminfo.node0.Shmem

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     11786 ~ 9%    +141.4%      28451 ~ 3%  lkp-ne04/micro/aim7/shell_rtns_1
     11786         +141.4%      28451       TOTAL sched_debug.cfs_rq[10]:/.min_vruntime

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     30715 ~ 0%     -50.4%      15249 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     30715          -50.4%      15249       TOTAL numa-meminfo.node0.KernelStack

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
      3823 ~ 0%     -50.1%       1908 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
      3823          -50.1%       1908       TOTAL numa-vmstat.node0.nr_kernel_stack

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     22270 ~ 3%     +91.0%      42527 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     22270          +91.0%      42527       TOTAL sched_debug.cfs_rq[10]:/.blocked_load_avg

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
      1967 ~ 0%     +90.5%       3749 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
      1967          +90.5%       3749       TOTAL numa-vmstat.node1.nr_kernel_stack

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     15819 ~ 1%     +89.1%      29910 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     15819          +89.1%      29910       TOTAL numa-meminfo.node1.KernelStack

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     41191 ~ 4%     +75.7%      72359 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     41191          +75.7%      72359       TOTAL numa-meminfo.node1.SUnreclaim

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     10264 ~ 3%     +76.5%      18121 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     10264          +76.5%      18121       TOTAL numa-vmstat.node1.nr_slab_unreclaimable

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     17782 ~ 2%     -40.0%      10670 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
     17782          -40.0%      10670       TOTAL numa-vmstat.node0.nr_slab_unreclaimable

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     71301 ~ 1%     -40.0%      42766 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
     71301          -40.0%      42766       TOTAL numa-meminfo.node0.SUnreclaim

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     59103 ~ 3%     +49.8%      88528 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     59103          +49.8%      88528       TOTAL numa-meminfo.node1.Slab

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     89141 ~ 1%     -30.2%      62248 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     89141          -30.2%      62248       TOTAL numa-meminfo.node0.Slab

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     38773 ~ 2%     +29.7%      50304 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
     38773          +29.7%      50304       TOTAL sched_debug.cpu#10.nr_switches

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     35293 ~ 1%     +24.8%      44029 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     35293          +24.8%      44029       TOTAL sched_debug.cpu#15.nr_switches

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     35772 ~ 1%     +23.9%      44317 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
     35772          +23.9%      44317       TOTAL sched_debug.cpu#14.nr_switches

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
    946595 ~ 0%     -14.2%     812568 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
    946595          -14.2%     812568       TOTAL numa-meminfo.node0.MemUsed

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
      5966 ~ 0%     +11.1%       6627 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
      5966          +11.1%       6627       TOTAL slabinfo.mm_struct.active_objs

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
    131624 ~ 0%     +99.0%     261914 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
    131624          +99.0%     261914       TOTAL time.involuntary_context_switches

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
      3542 ~ 0%    +104.2%       7233 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
      2407 ~ 0%     -10.4%       2156 ~ 0%  nhm-white/micro/aim7/dbase
      5950          +57.8%       9389       TOTAL vmstat.system.in

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
     20993 ~ 0%     +17.1%      24588 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
      2166 ~ 0%      +7.8%       2335 ~ 0%  nhm-white/micro/aim7/dbase
     23159          +16.3%      26923       TOTAL vmstat.system.cs

      v3.13-rc8  91dd08bd65052b483f96f75ab
---------------  -------------------------
   1423954 ~ 0%     +12.2%    1598169 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
   1423954          +12.2%    1598169       TOTAL time.voluntary_context_switches

Comparing e69bd3470ef8ab10bcb9d6ad4 ("sched: unify imbalance bias for
target group") with its parent:

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
     15604 ~ 3%   +1268.6%     213573 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
      1532 ~16%    +647.7%      11455 ~ 6%  nhm-white/micro/aim7/dbase
     17136        +1213.1%     225029       TOTAL interrupts.RES

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
    107861 ~ 1%     -63.0%      39865 ~ 2%  nhm-white/micro/aim7/dbase
    107861          -63.0%      39865       TOTAL interrupts.IWI

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
    133536 ~ 0%     +69.4%     226257 ~ 1%  lkp-ne04/micro/aim7/shell_rtns_1
    133536          +69.4%     226257       TOTAL time.involuntary_context_switches

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
      2359 ~ 0%      -9.5%       2134 ~ 0%  nhm-white/micro/aim7/dbase
      2359           -9.5%       2134       TOTAL vmstat.system.in

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
   1431362 ~ 0%      +8.5%    1552345 ~ 0%  lkp-ne04/micro/aim7/shell_rtns_1
   1431362           +8.5%    1552345       TOTAL time.voluntary_context_switches

64651525f152c05  e69bd3470ef8ab10bcb9d6ad4  
---------------  -------------------------  
      2162 ~ 0%      +5.2%       2274 ~ 0%  nhm-white/micro/aim7/dbase
      2162           +5.2%       2274       TOTAL vmstat.system.cs


  reply	other threads:[~2014-01-16  1:23 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-01-11 10:19 [sched] 73628fba4: +69% context switches Fengguang Wu
2014-01-13  5:57 ` Alex Shi
2014-01-16  1:23   ` Fengguang Wu [this message]
2014-01-16  2:02     ` Alex Shi
2014-01-16  2:18       ` Fengguang Wu
2014-01-16  2:46         ` Alex Shi
2014-01-16  3:01           ` Fengguang Wu
2014-01-16  3:07             ` Alex Shi

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=20140116012301.GA2535@localhost \
    --to=fengguang.wu@intel.com \
    --cc=alex.shi@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lkp@linux.intel.com \
    --cc=peterz@infradead.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.