The Linux Kernel Mailing List
 help / color / mirror / Atom feed
From: kernel test robot <oliver.sang@intel.com>
To: Zecheng Li <zecheng@google.com>
Cc: <oe-lkp@lists.linux.dev>, <lkp@intel.com>,
	<linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Zecheng Li <zli94@ncsu.edu>,
	"K Prateek Nayak" <kprateek.nayak@amd.com>,
	Josh Don <joshdon@google.com>, <aubrey.li@linux.intel.com>,
	<yu.c.chen@intel.com>, <oliver.sang@intel.com>
Subject: [linus:master] [sched/fair]  b8fea7af0e: stress-ng.session.ops_per_sec 91.5% improvement
Date: Fri, 26 Jun 2026 17:11:58 +0800	[thread overview]
Message-ID: <202606261633.32eca12c-lkp@intel.com> (raw)



Hello,

kernel test robot noticed a 91.5% improvement of stress-ng.session.ops_per_sec on:


commit: b8fea7af0e40feb6d9cbbd60b66ff0ec265e868f ("sched/fair: Allocate cfs_tg_state with percpu allocator")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master


testcase: stress-ng
config: x86_64-rhel-9.4
compiler: gcc-14
test machine: 192 threads 2 sockets Intel(R) Xeon(R) 6740E  CPU @ 2.4GHz (Sierra Forest) with 256G memory
parameters:

	nr_threads: 100%
	testtime: 60s
	test: session
	cpufreq_governor: performance


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


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20260626/202606261633.32eca12c-lkp@intel.com

=========================================================================================
compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
  gcc-14/performance/x86_64-rhel-9.4/100%/debian-13-x86_64-20250902.cgz/lkp-srf-2sp3/session/stress-ng/60s

commit: 
  89e1f67186 ("sched/fair: Remove task_group->se pointer array")
  b8fea7af0e ("sched/fair: Allocate cfs_tg_state with percpu allocator")

89e1f67186baca35 b8fea7af0e40feb6d9cbbd60b66 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
    610356           +91.5%    1168547 ±  2%  stress-ng.session.ops
     10187           +91.5%      19511 ±  2%  stress-ng.session.ops_per_sec
    682208           +23.9%     845283 ±  2%  stress-ng.time.involuntary_context_switches
      9315 ± 36%    +196.0%      27570 ± 46%  stress-ng.time.major_page_faults
  74894407 ±  3%     +91.4%  1.433e+08 ±  4%  stress-ng.time.minor_page_faults
     13490            -2.5%      13152        stress-ng.time.percent_of_cpu_this_job_got
      8076            -3.1%       7829        stress-ng.time.system_time
     33.20 ±  2%    +130.4%      76.48 ±  3%  stress-ng.time.user_time
   3019455          +147.5%    7471743 ±  2%  stress-ng.time.voluntary_context_switches
 2.619e+09 ±  6%     +41.1%  3.697e+09 ±  3%  cpuidle..time
   4789881          +106.5%    9891495 ±  2%  cpuidle..usage
     22.08 ±  4%     +44.9%      31.98 ±  3%  vmstat.cpu.id
   4955056            +9.8%    5439814        vmstat.memory.cache
    169.44 ±  2%     -13.4%     146.67 ±  2%  vmstat.procs.r
    124532          +120.7%     274835 ±  2%  vmstat.system.cs
    455602           +21.4%     553218        vmstat.system.in
     19.73 ±  5%     +10.0       29.75 ±  4%  mpstat.cpu.all.idle%
      0.03 ±  8%      -0.0        0.00 ± 81%  mpstat.cpu.all.iowait%
      0.37            +0.2        0.56 ±  3%  mpstat.cpu.all.irq%
     11.52            -9.3        2.22 ±  2%  mpstat.cpu.all.soft%
      0.89            +0.3        1.16        mpstat.cpu.all.usr%
     87.80            -6.3%      82.23        mpstat.max_utilization_pct
  23158413 ±  2%     +91.2%   44277233 ±  3%  numa-numastat.node0.local_node
  23346017 ±  2%     +95.5%   45639044 ±  3%  numa-numastat.node0.numa_hit
    187646 ± 17%    +625.7%    1361768 ±  6%  numa-numastat.node0.other_node
  21382055 ±  2%     +87.6%   40123301 ±  6%  numa-numastat.node1.local_node
  21639867 ±  2%     +92.2%   41600375 ±  6%  numa-numastat.node1.numa_hit
    257307 ± 11%    +474.1%    1477073 ±  6%  numa-numastat.node1.other_node
      2540           -12.4%       2225        turbostat.Avg_MHz
     79.44            -9.9       69.53        turbostat.Busy%
     22.02 ±  5%      +9.2       31.19 ±  3%  turbostat.C1%
     20.36 ±  5%     +48.8%      30.30 ±  3%  turbostat.CPU%c1
      0.15           +60.0%       0.24        turbostat.IPC
  30697679           +24.4%   38174002        turbostat.IRQ
    313.32            +2.6%     321.34        turbostat.PkgWatt
     24.18           +18.1%      28.56        turbostat.RAMWatt
   1537143           +32.6%    2038945        meminfo.Active
   1537110           +32.6%    2038912        meminfo.Active(anon)
   4808401            +9.8%    5281769        meminfo.Cached
   5073192            +8.4%    5498934        meminfo.Committed_AS
    391992           -15.3%     332050        meminfo.Mapped
     93966 ±  4%     +21.0%     113702 ±  5%  meminfo.PageTables
    172086 ±  7%    +325.6%     732317        meminfo.Percpu
   1095287           -56.6%     475325        meminfo.SUnreclaim
    716648           +66.1%    1190009        meminfo.Shmem
   1237725           -50.0%     618879        meminfo.Slab
      0.23           -66.5%       0.08        perf-sched.sch_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
     56.29 ± 20%     -37.3%      35.30 ± 13%  perf-sched.sch_delay.max.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
      0.23           -66.5%       0.08        perf-sched.total_sch_delay.average.ms
     56.29 ± 20%     -37.3%      35.30 ± 13%  perf-sched.total_sch_delay.max.ms
      9.88           -51.7%       4.78        perf-sched.total_wait_and_delay.average.ms
    399971          +106.1%     824497        perf-sched.total_wait_and_delay.count.ms
      9.65           -51.3%       4.70        perf-sched.total_wait_time.average.ms
      9.88           -51.7%       4.78        perf-sched.wait_and_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
    399971          +106.1%     824497        perf-sched.wait_and_delay.count.[unknown].[unknown].[unknown].[unknown].[unknown]
      9.65           -51.3%       4.70        perf-sched.wait_time.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
    133243          +116.4%     288278 ±  3%  perf-stat.i.context-switches
    194336            -1.1%     192149        perf-stat.i.cpu-clock
     27884           +91.6%      53428 ±  3%  perf-stat.i.cpu-migrations
    158.09 ± 37%    +185.5%     451.31 ± 46%  perf-stat.i.major-faults
   1256057 ±  3%     +88.1%    2362437 ±  4%  perf-stat.i.minor-faults
   1256215 ±  3%     +88.1%    2362888 ±  4%  perf-stat.i.page-faults
    194336            -1.1%     192149        perf-stat.i.task-clock
    128658          +118.2%     280754 ±  3%  perf-stat.ps.context-switches
     26831           +93.3%      51879 ±  3%  perf-stat.ps.cpu-migrations
    150.27 ± 37%    +193.4%     440.89 ± 46%  perf-stat.ps.major-faults
   1212558 ±  3%     +89.3%    2295418 ±  3%  perf-stat.ps.minor-faults
   1212708 ±  3%     +89.3%    2295859 ±  3%  perf-stat.ps.page-faults
    167364 ± 66%    +787.8%    1485853 ± 18%  numa-meminfo.node0.Active
    167345 ± 66%    +787.9%    1485838 ± 18%  numa-meminfo.node0.Active(anon)
     49987 ± 92%    +324.2%     212032 ± 19%  numa-meminfo.node0.Mapped
    619973 ±  6%     -58.1%     259551 ± 26%  numa-meminfo.node0.SUnreclaim
     29716 ± 78%   +3008.8%     923832 ±  4%  numa-meminfo.node0.Shmem
    705775 ±  7%     -54.1%     323802 ± 30%  numa-meminfo.node0.Slab
   1370362 ±  8%     -59.5%     555194 ± 51%  numa-meminfo.node1.Active
   1370348 ±  8%     -59.5%     555175 ± 51%  numa-meminfo.node1.Active(anon)
    362187 ± 10%     -74.2%      93542 ±159%  numa-meminfo.node1.AnonHugePages
    341339 ± 12%     -64.6%     120912 ± 35%  numa-meminfo.node1.Mapped
     45901 ± 16%     +23.8%      56827 ±  4%  numa-meminfo.node1.PageTables
    476900 ± 10%     -54.7%     215830 ± 31%  numa-meminfo.node1.SUnreclaim
    686948 ±  3%     -61.1%     267149 ± 15%  numa-meminfo.node1.Shmem
    533533 ± 10%     -44.7%     295138 ± 33%  numa-meminfo.node1.Slab
    383948           +32.8%     509815        proc-vmstat.nr_active_anon
    206513            +3.4%     213464        proc-vmstat.nr_anon_pages
   1201861            +9.9%    1320471        proc-vmstat.nr_file_pages
     97614 ±  2%     -14.9%      83031        proc-vmstat.nr_mapped
     23581 ±  3%     +20.7%      28463 ±  4%  proc-vmstat.nr_page_table_pages
    178922           +66.3%     297531        proc-vmstat.nr_shmem
    274192           -56.7%     118831        proc-vmstat.nr_slab_unreclaimable
     80759            +3.2%      83318        proc-vmstat.nr_vmalloc
    383948           +32.8%     509815        proc-vmstat.nr_zone_active_anon
  44989050 ±  2%     +93.9%   87242239 ±  2%  proc-vmstat.numa_hit
  44543635 ±  2%     +89.5%   84403353 ±  2%  proc-vmstat.numa_local
    444953 ±  3%    +538.0%    2838842        proc-vmstat.numa_other
  48904767 ±  2%     +85.6%   90766237 ±  2%  proc-vmstat.pgalloc_normal
  75424210 ±  3%     +90.8%  1.439e+08 ±  4%  proc-vmstat.pgfault
  48116497 ±  2%     +86.5%   89739194 ±  2%  proc-vmstat.pgfree
   4261223 ± 10%     +96.1%    8358094 ±  6%  proc-vmstat.pgreuse
     41819 ± 66%    +788.2%     371430 ± 18%  numa-vmstat.node0.nr_active_anon
     12503 ± 92%    +324.2%      53034 ± 19%  numa-vmstat.node0.nr_mapped
      7429 ± 78%   +3008.7%     230968 ±  4%  numa-vmstat.node0.nr_shmem
    155024 ±  6%     -58.1%      64888 ± 26%  numa-vmstat.node0.nr_slab_unreclaimable
     41818 ± 66%    +788.2%     371428 ± 18%  numa-vmstat.node0.nr_zone_active_anon
  23345970 ±  2%     +95.5%   45639379 ±  3%  numa-vmstat.node0.numa_hit
  23158366 ±  2%     +91.2%   44277567 ±  3%  numa-vmstat.node0.numa_local
    187646 ± 17%    +625.7%    1361768 ±  6%  numa-vmstat.node0.numa_other
    342699 ±  8%     -59.5%     138864 ± 51%  numa-vmstat.node1.nr_active_anon
    176.85 ± 10%     -74.2%      45.68 ±159%  numa-vmstat.node1.nr_anon_transparent_hugepages
     85411 ± 12%     -64.6%      30246 ± 35%  numa-vmstat.node1.nr_mapped
     11500 ± 16%     +23.8%      14240 ±  4%  numa-vmstat.node1.nr_page_table_pages
    171755 ±  3%     -61.1%      66798 ± 15%  numa-vmstat.node1.nr_shmem
    119206 ± 10%     -54.7%      53962 ± 31%  numa-vmstat.node1.nr_slab_unreclaimable
    342698 ±  8%     -59.5%     138863 ± 51%  numa-vmstat.node1.nr_zone_active_anon
  21639864 ±  2%     +92.2%   41600300 ±  6%  numa-vmstat.node1.numa_hit
  21382052 ±  2%     +87.6%   40123226 ±  6%  numa-vmstat.node1.numa_local
    257307 ± 11%    +474.1%    1477073 ±  6%  numa-vmstat.node1.numa_other
      0.50 ± 10%     -16.4%       0.42 ±  8%  sched_debug.cfs_rq:/.h_nr_queued.avg
      0.49 ± 11%     -15.5%       0.41 ±  8%  sched_debug.cfs_rq:/.h_nr_runnable.avg
     32477 ± 11%     -20.0%      25967 ± 15%  sched_debug.cfs_rq:/.left_deadline.max
     32462 ± 11%     -20.1%      25950 ± 15%  sched_debug.cfs_rq:/.left_vruntime.max
    495516 ± 11%     -18.3%     404915 ±  9%  sched_debug.cfs_rq:/.load.avg
      1675 ± 12%     -22.9%       1291 ± 24%  sched_debug.cfs_rq:/.load_avg.avg
     17.83 ± 37%     +73.3%      30.90 ± 29%  sched_debug.cfs_rq:/.load_avg.min
      0.50 ± 10%     -16.5%       0.42 ±  8%  sched_debug.cfs_rq:/.nr_queued.avg
     32462 ± 11%     -20.1%      25950 ± 15%  sched_debug.cfs_rq:/.right_vruntime.max
 8.253e+11 ± 46%     -68.5%    2.6e+11 ± 44%  sched_debug.cfs_rq:/.sum_w_vruntime.max
-1.142e+12           -79.2% -2.372e+11        sched_debug.cfs_rq:/.sum_w_vruntime.min
 1.641e+11 ± 14%     -80.5%  3.196e+10 ± 30%  sched_debug.cfs_rq:/.sum_w_vruntime.stddev
    562.81 ±  4%      -6.3%     527.61 ±  6%  sched_debug.cpu.clock_task.stddev
    408623 ± 13%     +44.3%     589658 ±  7%  sched_debug.cpu.curr->pid.avg
    609584           +90.0%    1157918 ±  2%  sched_debug.cpu.curr->pid.max
    278101 ±  6%    +106.4%     574007 ±  2%  sched_debug.cpu.curr->pid.stddev
      0.50 ± 10%     -16.5%       0.42 ±  8%  sched_debug.cpu.nr_running.avg
     22206          +111.3%      46925 ±  3%  sched_debug.cpu.nr_switches.avg
     59463 ± 30%     +66.6%      99041 ± 11%  sched_debug.cpu.nr_switches.max
     16171 ± 11%     +62.8%      26329 ± 21%  sched_debug.cpu.nr_switches.min
   -115.25          +180.9%    -323.70        sched_debug.cpu.nr_uninterruptible.min
     47.53 ± 13%    +134.1%     111.26 ± 21%  sched_debug.cpu.nr_uninterruptible.stddev




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.


-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki


                 reply	other threads:[~2026-06-26  9:12 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=202606261633.32eca12c-lkp@intel.com \
    --to=oliver.sang@intel.com \
    --cc=aubrey.li@linux.intel.com \
    --cc=joshdon@google.com \
    --cc=kprateek.nayak@amd.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lkp@intel.com \
    --cc=oe-lkp@lists.linux.dev \
    --cc=peterz@infradead.org \
    --cc=yu.c.chen@intel.com \
    --cc=zecheng@google.com \
    --cc=zli94@ncsu.edu \
    /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