From: kernel test robot <oliver.sang@intel.com>
To: Dave Chinner <dchinner@redhat.com>
Cc: <oe-lkp@lists.linux.dev>, <lkp@intel.com>,
<linux-kernel@vger.kernel.org>, Carlos Maiolino <cem@kernel.org>,
Jan Kara <jack@suse.cz>, Christoph Hellwig <hch@lst.de>,
<linux-xfs@vger.kernel.org>, <oliver.sang@intel.com>
Subject: [linus:master] [xfs] c91d38b57f: stress-ng.chown.ops_per_sec 70.2% improvement
Date: Thu, 2 Oct 2025 16:11:29 +0800 [thread overview]
Message-ID: <202510020917.2ead7cfe-lkp@intel.com> (raw)
Hello,
kernel test robot noticed a 70.2% improvement of stress-ng.chown.ops_per_sec on:
commit: c91d38b57f2c4784d885c874b2a1234a01361afd ("xfs: rework datasync tracking and execution")
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) Platinum 8468V CPU @ 2.4GHz (Sapphire Rapids) with 384G memory
parameters:
nr_threads: 100%
disk: 1SSD
testtime: 60s
fs: xfs
test: chown
cpufreq_governor: performance
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20251002/202510020917.2ead7cfe-lkp@intel.com
=========================================================================================
compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
gcc-14/performance/1SSD/xfs/x86_64-rhel-9.4/100%/debian-13-x86_64-20250902.cgz/igk-spr-2sp1/chown/stress-ng/60s
commit:
bc7d684fea ("xfs: rearrange code in xfs_inode_item_precommit")
c91d38b57f ("xfs: rework datasync tracking and execution")
bc7d684fea18cc48 c91d38b57f2c4784d885c874b2a
---------------- ---------------------------
%stddev %change %stddev
\ | \
21446 -15.8% 18048 uptime.idle
1.152e+10 -29.2% 8.152e+09 ± 2% cpuidle..time
8263083 ± 5% +11.3% 9194591 ± 2% cpuidle..usage
98.28 -26.1% 72.66 ± 2% iostat.cpu.idle
1.35 ± 9% +1890.7% 26.82 ± 5% iostat.cpu.system
19060 ± 85% +613.5% 135997 ± 28% numa-meminfo.node0.Shmem
297720 ± 7% +43.6% 427508 ± 8% numa-meminfo.node1.Shmem
4765 ± 85% +613.7% 34012 ± 28% numa-vmstat.node0.nr_shmem
74477 ± 7% +43.6% 106962 ± 8% numa-vmstat.node1.nr_shmem
1004889 +25.4% 1260516 meminfo.Active
1004873 +25.4% 1260499 meminfo.Active(anon)
119627 ± 2% +81.8% 217452 ± 2% meminfo.Mapped
316668 ± 4% +77.8% 563129 meminfo.Shmem
87.17 ± 15% +151.4% 219.17 ± 7% perf-c2c.DRAM.local
1506 ± 9% +105.9% 3102 ± 2% perf-c2c.DRAM.remote
1846 ± 8% +74.6% 3225 ± 3% perf-c2c.HITM.local
967.83 ± 10% +111.5% 2046 ± 3% perf-c2c.HITM.remote
2814 ± 3% +87.3% 5271 ± 2% perf-c2c.HITM.total
6.91 ± 6% -40.1% 4.14 perf-sched.total_wait_and_delay.average.ms
428372 ± 5% +71.9% 736472 perf-sched.total_wait_and_delay.count.ms
6.90 ± 6% -40.2% 4.13 perf-sched.total_wait_time.average.ms
6.91 ± 6% -40.1% 4.14 perf-sched.wait_and_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
428372 ± 5% +71.9% 736472 perf-sched.wait_and_delay.count.[unknown].[unknown].[unknown].[unknown].[unknown]
6.90 ± 6% -40.2% 4.13 perf-sched.wait_time.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
702197 ± 6% +70.2% 1195115 stress-ng.chown.ops
11705 ± 6% +70.2% 19923 stress-ng.chown.ops_per_sec
181.50 ± 22% +1974.8% 3765 ± 4% stress-ng.time.involuntary_context_switches
279.00 ± 4% +1929.8% 5663 ± 5% stress-ng.time.percent_of_cpu_this_job_got
166.63 ± 4% +1941.2% 3401 ± 5% stress-ng.time.system_time
5209159 ± 7% +80.3% 9389854 stress-ng.time.voluntary_context_switches
98.26 -26.4 71.83 ± 2% mpstat.cpu.all.idle%
0.13 ± 5% +0.3 0.43 ± 5% mpstat.cpu.all.irq%
0.07 ± 11% +0.0 0.11 mpstat.cpu.all.soft%
1.15 ± 11% +26.0 27.11 ± 6% mpstat.cpu.all.sys%
0.38 ± 5% +0.1 0.53 ± 2% mpstat.cpu.all.usr%
1.00 +1533.3% 16.33 ± 86% mpstat.max_utilization.seconds
3.40 ± 6% +950.6% 35.70 ± 6% mpstat.max_utilization_pct
77.00 +1065.8% 897.67 ± 4% turbostat.Avg_MHz
2.56 ± 3% +28.4 30.96 ± 4% turbostat.Busy%
3008 -3.6% 2900 turbostat.Bzy_MHz
0.15 ± 25% +2.2 2.32 ± 2% turbostat.C1%
11.82 ± 8% -4.1 7.67 turbostat.C1E%
85.55 -26.4 59.13 ± 2% turbostat.C6%
77.00 ± 6% -40.5% 45.80 ± 4% turbostat.CPU%c1
17.96 ± 28% -88.6% 2.05 ± 8% turbostat.CPU%c6
0.53 ± 2% -43.3% 0.30 turbostat.IPC
4796749 ± 5% +186.9% 13763385 ± 2% turbostat.IRQ
268338 ± 2% +636.9% 1977304 ± 5% turbostat.NMI
0.27 ± 13% -30.6% 0.18 ± 18% turbostat.Pkg%pc6
50.33 +8.6% 54.67 turbostat.PkgTmp
399.48 +17.5% 469.39 turbostat.PkgWatt
9.47 +3.3% 9.79 turbostat.RAMWatt
251227 +25.5% 315208 proc-vmstat.nr_active_anon
172983 +1.4% 175383 proc-vmstat.nr_anon_pages
1003874 +6.1% 1065584 proc-vmstat.nr_file_pages
29851 +81.4% 54161 ± 3% proc-vmstat.nr_mapped
79196 ± 4% +77.9% 140909 proc-vmstat.nr_shmem
119037 +1.8% 121217 proc-vmstat.nr_slab_unreclaimable
251227 +25.5% 315208 proc-vmstat.nr_zone_active_anon
878478 +13.7% 998673 proc-vmstat.numa_hit
679877 +17.7% 800107 proc-vmstat.numa_local
10806 ±116% +199.2% 32333 ± 31% proc-vmstat.numa_pages_migrated
932088 +13.4% 1057269 proc-vmstat.pgalloc_normal
498583 ± 2% +6.7% 532123 ± 2% proc-vmstat.pgfault
797828 +3.6% 826753 proc-vmstat.pgfree
10806 ±116% +199.2% 32333 ± 31% proc-vmstat.pgmigrate_success
2667340 ± 6% -13.3% 2311308 ± 2% proc-vmstat.pgpgout
27316 ± 5% +19.9% 32752 ± 11% proc-vmstat.pgreuse
3.06 ± 13% -71.8% 0.86 ± 4% perf-stat.i.MPKI
1.695e+09 +535.8% 1.078e+10 ± 3% perf-stat.i.branch-instructions
2.07 ± 3% -1.5 0.58 ± 3% perf-stat.i.branch-miss-rate%
45649257 ± 2% +32.1% 60312544 perf-stat.i.branch-misses
20782898 ± 13% +118.8% 45464342 perf-stat.i.cache-misses
74939994 ± 3% +196.9% 2.225e+08 perf-stat.i.cache-references
211956 ± 7% +63.6% 346733 perf-stat.i.context-switches
2.09 ± 2% +55.1% 3.24 perf-stat.i.cpi
1.487e+10 ± 2% +1070.8% 1.741e+11 ± 4% perf-stat.i.cpu-cycles
633.37 ± 12% +397.0% 3147 ± 8% perf-stat.i.cpu-migrations
756.33 ± 16% +409.7% 3855 ± 5% perf-stat.i.cycles-between-cache-misses
8.097e+09 +561.3% 5.354e+10 ± 3% perf-stat.i.instructions
0.53 ± 3% -40.3% 0.31 perf-stat.i.ipc
0.99 ± 12% +84.4% 1.83 perf-stat.i.metric.K/sec
6687 ± 3% +9.2% 7305 ± 3% perf-stat.i.minor-faults
6687 ± 3% +9.2% 7305 ± 3% perf-stat.i.page-faults
2.57 ± 14% -66.9% 0.85 ± 4% perf-stat.overall.MPKI
2.68 ± 2% -2.1 0.56 ± 3% perf-stat.overall.branch-miss-rate%
1.84 ± 3% +76.9% 3.25 perf-stat.overall.cpi
730.19 ± 14% +425.1% 3833 ± 5% perf-stat.overall.cycles-between-cache-misses
0.54 ± 3% -43.5% 0.31 perf-stat.overall.ipc
1.668e+09 +535.6% 1.06e+10 ± 3% perf-stat.ps.branch-instructions
44770981 ± 2% +31.7% 58956996 perf-stat.ps.branch-misses
20427788 ± 13% +118.8% 44692670 perf-stat.ps.cache-misses
73770990 ± 3% +196.6% 2.188e+08 perf-stat.ps.cache-references
208371 ± 7% +63.5% 340716 perf-stat.ps.context-switches
1.463e+10 ± 2% +1070.4% 1.713e+11 ± 4% perf-stat.ps.cpu-cycles
623.35 ± 12% +396.3% 3093 ± 8% perf-stat.ps.cpu-migrations
7.964e+09 +561.1% 5.265e+10 ± 3% perf-stat.ps.instructions
6538 ± 3% +8.3% 7081 ± 3% perf-stat.ps.minor-faults
6538 ± 3% +8.3% 7081 ± 3% perf-stat.ps.page-faults
4.839e+11 +561.9% 3.203e+12 ± 4% perf-stat.total.instructions
5485 ± 39% +9425.4% 522524 ± 9% sched_debug.cfs_rq:/.avg_vruntime.avg
70276 ± 23% +1083.8% 831952 ± 3% sched_debug.cfs_rq:/.avg_vruntime.max
613.05 ± 40% +15770.4% 97293 ± 20% sched_debug.cfs_rq:/.avg_vruntime.min
7853 ± 18% +1329.3% 112241 ± 10% sched_debug.cfs_rq:/.avg_vruntime.stddev
0.05 ± 21% +283.3% 0.21 ± 12% sched_debug.cfs_rq:/.h_nr_queued.avg
1.00 +66.7% 1.67 ± 14% sched_debug.cfs_rq:/.h_nr_queued.max
0.22 ± 10% +80.1% 0.40 ± 5% sched_debug.cfs_rq:/.h_nr_queued.stddev
0.05 ± 21% +250.0% 0.19 ± 12% sched_debug.cfs_rq:/.h_nr_runnable.avg
0.22 ± 10% +60.3% 0.36 ± 5% sched_debug.cfs_rq:/.h_nr_runnable.stddev
0.00 +1.9e+12% 19199 ± 38% sched_debug.cfs_rq:/.left_deadline.avg
0.00 +6e+13% 598730 ± 12% sched_debug.cfs_rq:/.left_deadline.max
0.00 +1.9e+12% 19199 ± 38% sched_debug.cfs_rq:/.left_vruntime.avg
0.00 +6e+13% 598716 ± 12% sched_debug.cfs_rq:/.left_vruntime.max
41172 ±215% +4030.9% 1700764 ± 36% sched_debug.cfs_rq:/.load.avg
7612936 ±222% +499.5% 45642610 sched_debug.cfs_rq:/.load.max
551304 ±220% +1439.7% 8488365 ± 16% sched_debug.cfs_rq:/.load.stddev
40.23 ± 28% +3177.8% 1318 ± 7% sched_debug.cfs_rq:/.load_avg.avg
847.08 ± 30% +519.5% 5248 ± 9% sched_debug.cfs_rq:/.load_avg.max
140.26 ± 32% +812.7% 1280 ± 4% sched_debug.cfs_rq:/.load_avg.stddev
5485 ± 39% +9425.4% 522524 ± 9% sched_debug.cfs_rq:/.min_vruntime.avg
70276 ± 23% +1083.8% 831952 ± 3% sched_debug.cfs_rq:/.min_vruntime.max
613.05 ± 40% +15770.4% 97293 ± 20% sched_debug.cfs_rq:/.min_vruntime.min
7853 ± 18% +1329.3% 112241 ± 10% sched_debug.cfs_rq:/.min_vruntime.stddev
0.05 ± 21% +284.1% 0.21 ± 12% sched_debug.cfs_rq:/.nr_queued.avg
1.00 +66.7% 1.67 ± 14% sched_debug.cfs_rq:/.nr_queued.max
0.22 ± 10% +80.5% 0.40 ± 4% sched_debug.cfs_rq:/.nr_queued.stddev
0.00 +1.9e+12% 19199 ± 38% sched_debug.cfs_rq:/.right_vruntime.avg
0.00 +6e+13% 598716 ± 12% sched_debug.cfs_rq:/.right_vruntime.max
142.69 ± 23% +60.8% 229.41 ± 5% sched_debug.cfs_rq:/.runnable_avg.avg
142.61 ± 23% +60.4% 228.74 ± 5% sched_debug.cfs_rq:/.util_avg.avg
12.83 ± 25% +324.8% 54.52 ± 19% sched_debug.cfs_rq:/.util_est.avg
83.56 ± 17% +46.0% 122.02 ± 8% sched_debug.cfs_rq:/.util_est.stddev
1202994 ± 10% -16.1% 1009100 ± 3% sched_debug.cpu.avg_idle.avg
1139 ± 12% +20.7% 1375 sched_debug.cpu.clock_task.stddev
342.10 ± 29% +413.8% 1757 ± 13% sched_debug.cpu.curr->pid.avg
8154 ± 11% +16.4% 9490 sched_debug.cpu.curr->pid.max
1530 ± 14% +104.1% 3124 ± 5% sched_debug.cpu.curr->pid.stddev
0.05 ± 29% +306.0% 0.20 ± 15% sched_debug.cpu.nr_running.avg
1.00 +66.7% 1.67 ± 14% sched_debug.cpu.nr_running.max
0.21 ± 14% +86.5% 0.40 ± 6% sched_debug.cpu.nr_running.stddev
11939 ±125% +364.2% 55424 sched_debug.cpu.nr_switches.avg
296.42 ± 64% +11974.7% 35791 ± 12% sched_debug.cpu.nr_switches.min
29.50 ± 12% +66.9% 49.25 ± 11% sched_debug.cpu.nr_uninterruptible.max
-20.75 +385.1% -100.67 sched_debug.cpu.nr_uninterruptible.min
5.33 ± 8% +102.5% 10.80 ± 9% 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
next reply other threads:[~2025-10-02 8:11 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-10-02 8:11 kernel test robot [this message]
2025-10-03 7:56 ` [linus:master] [xfs] c91d38b57f: stress-ng.chown.ops_per_sec 70.2% improvement Christoph Hellwig
2025-10-03 13:38 ` Mateusz Guzik
2025-10-05 11:54 ` Dave Chinner
2025-10-07 6:44 ` Christoph Hellwig
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=202510020917.2ead7cfe-lkp@intel.com \
--to=oliver.sang@intel.com \
--cc=cem@kernel.org \
--cc=dchinner@redhat.com \
--cc=hch@lst.de \
--cc=jack@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=lkp@intel.com \
--cc=oe-lkp@lists.linux.dev \
/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.