lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Message-ID: <20160807141506.GA22624@yexl-desktop>
Date:	Sun, 7 Aug 2016 22:15:06 +0800
From:	kernel test robot <xiaolong.ye@...el.com>
To:	Mel Gorman <mgorman@...hsingularity.net>
Cc:	LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp] [mm, page_alloc]  194c9de225:  netperf.Throughput_Mbps 2.5%
 improvement



FYI, we noticed a 2.5% improvement of netperf.Throughput_Mbps due to commit:

commit 194c9de2256fa001eb69dfba48dbb6d1b2562bc7 ("mm, page_alloc: remove fair zone allocation policy")
https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master

in testcase: netperf
on test machine: 16 threads Broadwell-DE with 8G memory
with following parameters:

	ip: ipv4
	runtime: 900s
	nr_threads: 200%
	cluster: cs-localhost
	test: TCP_STREAM
	cpufreq_governor: performance


Disclaimer:
Results have been estimated based on internal Intel analysis and are provided
for informational purposes only. Any difference in system hardware or software
design or configuration may affect actual performance.

Details are as below:
-------------------------------------------------------------------------------------------------->


To reproduce:

        git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git
        cd lkp-tests
        bin/lkp install job.yaml  # job file is attached in this email
        bin/lkp run     job.yaml

=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase:
  cs-localhost/gcc-6/performance/ipv4/x86_64-rhel/200%/debian-x86_64-2015-02-07.cgz/900s/lkp-bdw-de1/TCP_STREAM/netperf

commit: 
  3f44f76ad2 ("mm, vmscan: add classzone information to tracepoints")
  194c9de225 ("mm, page_alloc: remove fair zone allocation policy")

3f44f76ad2128d47 194c9de2256fa001eb69dfba48 
---------------- -------------------------- 
       fail:runs  %reproduction    fail:runs
           |             |             |    
           :4           25%           1:4     kmsg.i2c_i2c-#:sendbytes:NAK_bailout
         %stddev     %change         %stddev
             \          |                \  
      2468 ±  0%      +2.5%       2529 ±  0%  netperf.Throughput_Mbps
    219017 ±  2%     -30.5%     152291 ±  5%  netperf.time.involuntary_context_switches
    578.25 ±  0%      -9.7%     522.00 ±  0%  netperf.time.percent_of_cpu_this_job_got
      5093 ±  0%      -9.8%       4593 ±  0%  netperf.time.system_time
    122.55 ±  0%      -8.1%     112.67 ±  0%  netperf.time.user_time
   2504387 ±  0%      +7.0%    2680007 ±  0%  netperf.time.voluntary_context_switches
    146422 ±  0%      -1.5%     144221 ±  0%  interrupts.CAL:Function_call_interrupts
    100998 ±  7%     -10.7%      90170 ±  0%  softirqs.SCHED
    647822 ± 82%     -68.2%     205753 ± 15%  cpuidle.C3-BDW.time
     15.50 ± 30%    +295.2%      61.25 ± 94%  cpuidle.POLL.usage
      3642 ±  7%     -11.5%       3222 ±  3%  slabinfo.cred_jar.active_objs
      3642 ±  7%     -11.5%       3222 ±  3%  slabinfo.cred_jar.num_objs
      0.08 ± 19%     +31.3%       0.11 ±  4%  turbostat.Pkg%pc6
     25.95 ±  0%      -1.0%      25.69 ±  0%  turbostat.RAMWatt
      2002 ±  2%    -100.0%       0.00 ± -1%  proc-vmstat.nr_alloc_batch
 5.247e+08 ±  0%    -100.0%       0.00 ±  0%  proc-vmstat.pgalloc_dma32
 1.601e+09 ±  0%     +36.0%  2.178e+09 ±  0%  proc-vmstat.pgalloc_normal
    509575 ±  0%     -83.9%      82003 ±  0%  meminfo.Active
      8151 ±  1%    +905.9%      81996 ±  0%  meminfo.Active(anon)
    501423 ±  0%    -100.0%       7.00 ±  0%  meminfo.Active(file)
     78991 ±  4%      -9.7%      71311 ±  5%  meminfo.DirectMap4k
     79524 ±  3%    +530.6%     501442 ±  0%  meminfo.Inactive(file)
      9.00 ±  7%     -18.6%       7.33 ±  2%  sched_debug.cpu.clock.stddev
      9.00 ±  7%     -18.6%       7.33 ±  2%  sched_debug.cpu.clock_task.stddev
      1408 ±  3%     -13.8%       1213 ± 11%  sched_debug.cpu.curr->pid.stddev
      3685 ± 35%     -66.1%       1248 ± 11%  sched_debug.cpu.nr_load_updates.stddev
      0.00 ±173%   +1424.9%       0.00 ± 52%  sched_debug.rt_rq:/.rt_time.min
      4.66 ±  1%      +9.0%       5.08 ±  1%  perf-profile.cycles-pp.__do_softirq.do_softirq_own_stack.do_softirq.part.13.__local_bh_enable_ip.ip_finish_output2
      4.72 ±  1%      +8.6%       5.12 ±  1%  perf-profile.cycles-pp.__local_bh_enable_ip.ip_finish_output2.ip_finish_output.ip_output.ip_local_out
     24.39 ±  1%     -10.0%      21.94 ±  1%  perf-profile.cycles-pp.copy_user_enhanced_fast_string.tcp_sendmsg.inet_sendmsg.sock_sendmsg.SYSC_sendto
      4.71 ±  1%      +8.9%       5.12 ±  1%  perf-profile.cycles-pp.do_softirq.part.13.__local_bh_enable_ip.ip_finish_output2.ip_finish_output.ip_output
      4.68 ±  1%      +9.1%       5.11 ±  1%  perf-profile.cycles-pp.do_softirq_own_stack.do_softirq.part.13.__local_bh_enable_ip.ip_finish_output2.ip_finish_output
      1.29 ±  4%     +19.1%       1.53 ±  2%  perf-profile.cycles-pp.free_one_page.__free_pages_ok.free_compound_page.__put_compound_page.__put_page
      4.56 ±  1%      +8.9%       4.96 ±  1%  perf-profile.cycles-pp.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq.part.13.__local_bh_enable_ip
      4.48 ±  1%      +8.8%       4.88 ±  1%  perf-profile.cycles-pp.process_backlog.net_rx_action.__do_softirq.do_softirq_own_stack.do_softirq.part.13
     34.16 ±  1%      -8.6%      31.23 ±  0%  perf-profile.cycles-pp.tcp_sendmsg.inet_sendmsg.sock_sendmsg.SYSC_sendto.sys_sendto
      1.03 ±  1%     -14.6%       0.88 ±  2%  perf-profile.func.cycles-pp.get_page_from_freelist
      1.18 ±  9%     +30.2%       1.53 ±  5%  perf-profile.func.cycles-pp.native_queued_spin_lock_slowpath
      2.57 ±  2%     -13.6%       2.22 ±  1%  perf-profile.func.cycles-pp.tcp_sendmsg
 5.257e+11 ±  0%      +2.5%   5.39e+11 ±  0%  perf-stat.branch-instructions
      0.55 ±  0%      +5.8%       0.58 ±  0%  perf-stat.branch-miss-rate
 2.869e+09 ±  0%      +8.5%  3.113e+09 ±  0%  perf-stat.branch-misses
      0.03 ±  4%      -6.1%       0.03 ±  4%  perf-stat.dTLB-load-miss-rate
 1.465e+12 ±  0%      +2.3%  1.499e+12 ±  0%  perf-stat.dTLB-loads
 1.121e+12 ±  0%      +2.3%  1.147e+12 ±  0%  perf-stat.dTLB-stores
     87.66 ±  8%     -18.9%      71.08 ±  7%  perf-stat.iTLB-load-miss-rate
 1.744e+08 ±  2%      +7.7%  1.878e+08 ±  2%  perf-stat.iTLB-load-misses
  26237761 ± 70%    +197.5%   78044825 ± 28%  perf-stat.iTLB-loads
 2.923e+12 ±  0%      +2.4%  2.993e+12 ±  0%  perf-stat.instructions
     16771 ±  2%      -4.9%      15949 ±  2%  perf-stat.instructions-per-iTLB-miss
      0.08 ±  0%      +2.4%       0.09 ±  0%  perf-stat.ipc



                              netperf.Throughput_Mbps

  3000 ++-------------------------------------------------------------------+
       |                                                                    |
  2500 OOO OOOOO OOOO.OOOOO.OOOOO.O*                             **.**** .***
       | *.*      ***              :  *.*      ***   **      *.**       *   |
       | : :      :                :  : :      : :   ::      :              |
  2000 ++: :      :                :  : :      : :   ::      :              |
       | : :      :                :  : :      :  :  ::      :              |
  1500 ++: :      :                :  : :      :  :  ::      :              |
       | : :      :                 : : :      :  :  ::      :              |
  1000 ++   :    :                  ::   :    :   : :  :    :               |
       |:   :    :                  ::   :    :   : :  :    :               |
       |:   :    :                  ::   :    :   : :  :    :               |
   500 ++   :    :                  ::   :    :    ::  :    :               |
       |:   :    :                  ::   :    :    ::  :    :               |
     0 **---****-*------------------**---***-**----**--*-****---------------+


	[*] bisect-good sample
	[O] bisect-bad  sample



Thanks,
Xiaolong

View attachment "config-4.7.0-rc7-00317-g194c9de" of type "text/plain" (151365 bytes)

View attachment "job.yaml" of type "text/plain" (4055 bytes)

View attachment "reproduce" of type "text/plain" (2472 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ