| 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
| ||
|
Message-ID: <202509281326.f605b4eb-lkp@intel.com>
Date: Sun, 28 Sep 2025 13:21:42 +0800
From: kernel test robot <oliver.sang@...el.com>
To: Eric Dumazet <edumazet@...gle.com>
CC: <oe-lkp@...ts.linux.dev>, <lkp@...el.com>, Paolo Abeni
<pabeni@...hat.com>, Willem de Bruijn <willemb@...gle.com>, David Ahern
<dsahern@...nel.org>, Kuniyuki Iwashima <kuniyu@...gle.com>, Jakub Kicinski
<kuba@...nel.org>, <netdev@...r.kernel.org>, <oliver.sang@...el.com>
Subject: [linux-next:master] [net] 4effb335b5: stress-ng.rawsock.ops_per_sec
53.2% regression
Hello,
kernel test robot noticed a 53.2% regression of stress-ng.rawsock.ops_per_sec on:
commit: 4effb335b5dab08cb6e2c38d038910f8b527cfc9 ("net: group sk_backlog and sk_receive_queue")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
[still regression on linux-next/master 262858079afde6d367ce3db183c74d8a43a0e83f]
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: rawsock
cpufreq_governor: performance
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <oliver.sang@...el.com>
| Closes: https://lore.kernel.org/oe-lkp/202509281326.f605b4eb-lkp@intel.com
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250928/202509281326.f605b4eb-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/rawsock/stress-ng/60s
commit:
faf7b4aefd ("udp: update sk_rmem_alloc before busylock acquisition")
4effb335b5 ("net: group sk_backlog and sk_receive_queue")
faf7b4aefd5be1d1 4effb335b5dab08cb6e2c38d038
---------------- ---------------------------
%stddev %change %stddev
\ | \
4.724e+08 +39.7% 6.599e+08 ± 3% cpuidle..time
370887 ± 4% +197.2% 1102311 ± 5% cpuidle..usage
0.15 -38.9% 0.09 ± 44% turbostat.IPC
365.00 -20.0% 292.04 ± 44% turbostat.PkgWatt
20.18 -24.4% 15.26 ± 44% turbostat.RAMWatt
2.45 ± 19% +2.5 5.00 ± 8% mpstat.cpu.all.idle%
0.77 ± 2% -0.3 0.51 ± 5% mpstat.cpu.all.iowait%
72.04 +12.2 84.22 mpstat.cpu.all.soft%
23.72 ± 2% -14.2 9.52 ± 2% mpstat.cpu.all.sys%
0.78 ± 2% -0.3 0.50 ± 4% mpstat.cpu.all.usr%
2.83 ± 2% -52.1% 1.36 ± 2% stress-ng.rawsock.MB_recv'd_per_sec
1.265e+09 ± 2% -53.2% 5.922e+08 ± 2% stress-ng.rawsock.ops
21105419 ± 2% -53.2% 9879110 ± 2% stress-ng.rawsock.ops_per_sec
1706591 ± 2% -42.6% 979134 ± 3% stress-ng.time.involuntary_context_switches
167007 -16.4% 139564 stress-ng.time.minor_page_faults
4696 ± 2% -61.0% 1833 ± 3% stress-ng.time.percent_of_cpu_this_job_got
2751 ± 2% -61.1% 1069 ± 3% stress-ng.time.system_time
71.44 -54.7% 32.39 ± 2% stress-ng.time.user_time
107720 ± 2% +825.7% 997196 ± 7% stress-ng.time.voluntary_context_switches
4.98 ± 5% -21.5% 3.91 ± 9% perf-sched.sch_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
4.98 ± 5% -21.5% 3.91 ± 9% perf-sched.total_sch_delay.average.ms
24.66 ± 4% -20.3% 19.64 ± 6% perf-sched.total_wait_and_delay.average.ms
183199 ± 4% +37.2% 251374 ± 7% perf-sched.total_wait_and_delay.count.ms
3016 ± 8% -20.2% 2408 ± 15% perf-sched.total_wait_and_delay.max.ms
19.68 ± 4% -20.1% 15.73 ± 6% perf-sched.total_wait_time.average.ms
24.66 ± 4% -20.3% 19.64 ± 6% perf-sched.wait_and_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
183199 ± 4% +37.2% 251374 ± 7% perf-sched.wait_and_delay.count.[unknown].[unknown].[unknown].[unknown].[unknown]
3016 ± 8% -20.2% 2408 ± 15% perf-sched.wait_and_delay.max.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
19.68 ± 4% -20.1% 15.73 ± 6% perf-sched.wait_time.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
286434 +3.6% 296783 ± 2% proc-vmstat.nr_active_anon
105242 ± 2% +9.9% 115646 ± 5% proc-vmstat.nr_shmem
286434 +3.6% 296783 ± 2% proc-vmstat.nr_zone_active_anon
2409442 ± 2% -5.9% 2266894 proc-vmstat.numa_hit
999.33 ± 9% -39.4% 606.00 proc-vmstat.numa_huge_pte_updates
2211541 ± 2% -6.4% 2068937 ± 2% proc-vmstat.numa_local
109879 ± 3% -34.2% 72337 ± 6% proc-vmstat.numa_pages_migrated
549890 ± 9% -37.0% 346160 ± 2% proc-vmstat.numa_pte_updates
4067603 ± 2% -8.0% 3744126 ± 2% proc-vmstat.pgalloc_normal
529475 -6.2% 496583 ± 4% proc-vmstat.pgfault
3743518 ± 2% -9.7% 3380255 ± 3% proc-vmstat.pgfree
109879 ± 3% -34.2% 72337 ± 6% proc-vmstat.pgmigrate_success
1929439 ± 3% -62.7% 720154 ± 2% sched_debug.cfs_rq:/.avg_vruntime.avg
3052555 ± 4% -46.0% 1647653 ± 19% sched_debug.cfs_rq:/.avg_vruntime.max
1136911 ± 8% -53.8% 524938 ± 3% sched_debug.cfs_rq:/.avg_vruntime.min
342016 ± 5% -60.9% 133755 ± 13% sched_debug.cfs_rq:/.avg_vruntime.stddev
10.92 ± 28% -55.0% 4.92 ± 10% sched_debug.cfs_rq:/.h_nr_queued.max
10.92 ± 28% -55.0% 4.92 ± 10% sched_debug.cfs_rq:/.h_nr_runnable.max
1929439 ± 3% -62.7% 720155 ± 2% sched_debug.cfs_rq:/.min_vruntime.avg
3052555 ± 4% -46.0% 1647653 ± 19% sched_debug.cfs_rq:/.min_vruntime.max
1136911 ± 8% -53.8% 524938 ± 3% sched_debug.cfs_rq:/.min_vruntime.min
342016 ± 5% -60.9% 133755 ± 13% sched_debug.cfs_rq:/.min_vruntime.stddev
10081 ± 27% -54.9% 4549 ± 15% sched_debug.cfs_rq:/.runnable_avg.max
9761 ± 28% -57.5% 4144 ± 15% sched_debug.cfs_rq:/.util_est.max
1460892 ± 7% -16.4% 1221901 ± 3% sched_debug.cpu.avg_idle.avg
477384 ± 4% +32.5% 632636 ± 4% sched_debug.cpu.avg_idle.stddev
1622 ± 2% -50.8% 797.70 ± 7% sched_debug.cpu.clock_task.stddev
11.00 ± 28% -55.3% 4.92 ± 10% sched_debug.cpu.nr_running.max
7423 ± 3% +32.7% 9850 ± 4% sched_debug.cpu.nr_switches.avg
3621 ± 8% +76.2% 6381 ± 5% sched_debug.cpu.nr_switches.min
30.25 ± 2% -17.5% 24.95 ± 7% sched_debug.cpu.nr_uninterruptible.stddev
2.21 -32.3% 1.50 ± 2% perf-stat.i.MPKI
1.943e+10 -19.6% 1.561e+10 perf-stat.i.branch-instructions
0.62 -0.2 0.42 ± 3% perf-stat.i.branch-miss-rate%
1.209e+08 ± 2% -45.6% 65759957 ± 2% perf-stat.i.branch-misses
11.52 ± 2% -2.6 8.92 ± 3% perf-stat.i.cache-miss-rate%
1.955e+08 ± 2% -48.8% 1.001e+08 ± 2% perf-stat.i.cache-misses
1.713e+09 -26.2% 1.264e+09 ± 2% perf-stat.i.cache-references
36879 ± 2% +48.4% 54743 ± 5% perf-stat.i.context-switches
6.80 +33.2% 9.05 perf-stat.i.cpi
1.939e+11 +2.0% 1.978e+11 perf-stat.i.cpu-clock
2055 ± 2% +122.7% 4577 ± 5% perf-stat.i.cpu-migrations
3093 ± 2% +107.7% 6424 ± 3% perf-stat.i.cycles-between-cache-misses
8.83e+10 -24.9% 6.634e+10 perf-stat.i.instructions
0.15 -24.9% 0.11 perf-stat.i.ipc
1.939e+11 +2.0% 1.978e+11 perf-stat.i.task-clock
2.21 -32.9% 1.49 ± 2% perf-stat.overall.MPKI
0.62 -0.2 0.42 ± 3% perf-stat.overall.branch-miss-rate%
11.40 ± 2% -3.8 7.64 ± 4% perf-stat.overall.cache-miss-rate%
6.80 +33.2% 9.06 perf-stat.overall.cpi
3074 ± 2% +98.3% 6098 perf-stat.overall.cycles-between-cache-misses
0.15 -24.9% 0.11 perf-stat.overall.ipc
1.896e+10 -21.0% 1.498e+10 perf-stat.ps.branch-instructions
1.18e+08 -47.2% 62243287 ± 2% perf-stat.ps.branch-misses
1.906e+08 ± 2% -50.6% 94253030 perf-stat.ps.cache-misses
1.672e+09 -26.1% 1.235e+09 ± 2% perf-stat.ps.cache-references
35748 ± 2% +41.7% 50645 ± 4% perf-stat.ps.context-switches
5.859e+11 -1.9% 5.746e+11 perf-stat.ps.cpu-cycles
1945 ± 2% +112.2% 4127 ± 6% perf-stat.ps.cpu-migrations
8.617e+10 -26.3% 6.347e+10 perf-stat.ps.instructions
5.239e+12 -26.8% 3.836e+12 perf-stat.total.instructions
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
Powered by blists - more mailing lists