| 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
| ||
|
Date: Tue, 27 Apr 2021 15:34:48 +0800
From: kernel test robot <oliver.sang@...el.com>
To: Dennis Zhou <dennis@...nel.org>
Cc: Roman Gushchin <guro@...com>,
Pratik Sampat <psampat@...ux.ibm.com>,
LKML <linux-kernel@...r.kernel.org>, lkp@...ts.01.org,
lkp@...el.com, ying.huang@...el.com, feng.tang@...el.com,
zhengjun.xing@...el.com
Subject: [percpu] ace7e70901: aim9.sync_disk_rw.ops_per_sec -2.3% regression
Greeting,
FYI, we noticed a -2.3% regression of aim9.sync_disk_rw.ops_per_sec due to commit:
commit: ace7e7090137ee996757eb5eebc94439b0e2803a ("percpu: use reclaim threshold instead of running for every page")
https://git.kernel.org/cgit/linux/kernel/git/dennis/percpu.git for-5.14
in testcase: aim9
on test machine: 256 threads Intel(R) Genuine Intel(R) CPU 0000 @ 1.30GHz with 112G memory
with following parameters:
testtime: 300s
test: sync_disk_rw
cpufreq_governor: performance
ucode: 0xffff0190
test-description: Suite IX is the "AIM Independent Resource Benchmark:" the famous synthetic benchmark.
test-url: https://sourceforge.net/projects/aimbench/files/aim-suite9/
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <oliver.sang@...el.com>
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml
bin/lkp run compatible-job.yaml
=========================================================================================
compiler/cpufreq_governor/kconfig/rootfs/tbox_group/test/testcase/testtime/ucode:
gcc-9/performance/x86_64-rhel-8.3/debian-10.4-x86_64-20200603.cgz/lkp-knl-f1/sync_disk_rw/aim9/300s/0xffff0190
commit:
f183324133 ("percpu: implement partial chunk depopulation")
ace7e70901 ("percpu: use reclaim threshold instead of running for every page")
f183324133ea535d ace7e7090137ee996757eb5eebc
---------------- ---------------------------
%stddev %change %stddev
\ | \
103531 -2.3% 101102 aim9.sync_disk_rw.ops_per_sec
55.21 +3.0% 56.87 aim9.time.user_time
421.91 ± 9% +25.1% 527.64 ± 13% sched_debug.cfs_rq:/.load_avg.max
39017 ± 4% -5.8% 36738 ± 4% softirqs.CPU4.RCU
39128 ± 4% -7.3% 36268 ± 4% softirqs.CPU7.RCU
0.05 ± 14% +39.5% 0.07 ± 20% perf-sched.sch_delay.avg.ms.devkmsg_read.vfs_read.ksys_read.do_syscall_64
3970 ±105% +117.6% 8637 ± 36% perf-sched.wait_and_delay.max.ms.preempt_schedule_common.__cond_resched.wait_for_completion.affine_move_task.__set_cpus_allowed_ptr
0.05 ± 77% +349.6% 0.21 ± 75% perf-sched.wait_time.avg.ms.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
0.11 ± 84% +494.5% 0.68 ±130% perf-sched.wait_time.max.ms.exit_to_user_mode_prepare.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
3970 ±105% +117.6% 8637 ± 36% perf-sched.wait_time.max.ms.preempt_schedule_common.__cond_resched.wait_for_completion.affine_move_task.__set_cpus_allowed_ptr
8878045 -2.0% 8702700 proc-vmstat.numa_hit
8878044 -2.0% 8702696 proc-vmstat.numa_local
515593 -2.3% 503677 proc-vmstat.pgactivate
8967593 -2.0% 8788379 proc-vmstat.pgalloc_normal
8941947 -2.0% 8762451 proc-vmstat.pgfree
396.17 ± 23% -49.2% 201.14 ± 4% interrupts.CPU226.NMI:Non-maskable_interrupts
396.17 ± 23% -49.2% 201.14 ± 4% interrupts.CPU226.PMI:Performance_monitoring_interrupts
266.50 ± 27% -36.4% 169.43 ± 2% interrupts.CPU242.NMI:Non-maskable_interrupts
266.50 ± 27% -36.4% 169.43 ± 2% interrupts.CPU242.PMI:Performance_monitoring_interrupts
92.33 ± 85% -75.2% 22.86 ± 34% interrupts.CPU33.RES:Rescheduling_interrupts
26.83 ± 30% +130.5% 61.86 ± 61% interrupts.CPU45.RES:Rescheduling_interrupts
7.63 ± 2% +0.4 7.99 perf-stat.i.branch-miss-rate%
16.71 -0.5 16.18 perf-stat.i.cache-miss-rate%
8.36 -3.0% 8.11 perf-stat.i.cpi
7.61 ± 2% +0.4 7.97 perf-stat.overall.branch-miss-rate%
16.70 -0.5 16.17 perf-stat.overall.cache-miss-rate%
8.18 -2.8% 7.95 perf-stat.overall.cpi
0.09 ±223% +0.5 0.59 ± 5% perf-profile.calltrace.cycles-pp.tick_check_broadcast_expired.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64_no_verify
0.07 ± 5% -0.0 0.05 ± 41% perf-profile.children.cycles-pp.timerqueue_iterate_next
0.15 ± 4% +0.0 0.17 ± 7% perf-profile.children.cycles-pp.tick_nohz_idle_got_tick
0.05 ± 9% +0.0 0.10 ± 17% perf-profile.children.cycles-pp.cpuidle_get_cpu_driver
0.05 ± 7% +0.0 0.10 ± 14% perf-profile.children.cycles-pp.rcu_irq_exit
0.23 ± 11% +0.1 0.30 ± 8% perf-profile.children.cycles-pp.cpumask_next_and
0.34 ± 5% +0.1 0.41 ± 7% perf-profile.children.cycles-pp.rb_insert_color
0.40 ± 7% +0.1 0.49 ± 13% perf-profile.children.cycles-pp.hrtimer_forward
0.35 ± 8% +0.1 0.45 ± 6% perf-profile.children.cycles-pp.get_cpu_device
0.47 ± 6% +0.1 0.59 ± 4% perf-profile.children.cycles-pp.tick_check_broadcast_expired
0.60 ± 6% -0.0 0.55 ± 7% perf-profile.self.cycles-pp.perf_event_task_tick
0.26 ± 8% -0.0 0.22 ± 9% perf-profile.self.cycles-pp.tick_nohz_irq_exit
0.19 ± 3% +0.0 0.23 ± 7% perf-profile.self.cycles-pp.irqentry_enter
0.03 ± 70% +0.1 0.09 ± 16% perf-profile.self.cycles-pp.cpuidle_get_cpu_driver
0.32 ± 5% +0.1 0.39 ± 7% perf-profile.self.cycles-pp.rb_insert_color
0.06 ± 47% +0.1 0.13 ± 9% perf-profile.self.cycles-pp.cpumask_next_and
0.00 +0.1 0.08 ± 18% perf-profile.self.cycles-pp.rcu_irq_exit
0.34 ± 8% +0.1 0.44 ± 5% perf-profile.self.cycles-pp.get_cpu_device
0.47 ± 5% +0.1 0.59 ± 5% perf-profile.self.cycles-pp.tick_check_broadcast_expired
0.14 ± 9% +0.1 0.26 ± 10% perf-profile.self.cycles-pp.rcu_core
0.77 ± 5% +0.1 0.92 ± 3% perf-profile.self.cycles-pp.tick_nohz_next_event
aim9.sync_disk_rw.ops_per_sec
105000 +------------------------------------------------------------------+
| + |
104500 |-+ : : +.... |
104000 |-+ : : : + |
| : : + : + +|
103500 |-+ : : : : : + + |
103000 |-+ +. : : : : : + + |
| . .. : : : : : + |
102500 |-+ .. ..+.. .+ : : : : |
102000 |...+... . +. . .. +... : + |
| + O + + |
101500 |-+ O |
101000 |-+ O O |
| O O |
100500 +------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
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.
---
0DAY/LKP+ Test Infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/lkp@lists.01.org Intel Corporation
Thanks,
Oliver Sang
View attachment "config-5.12.0-rc7-00006-gace7e7090137" of type "text/plain" (172883 bytes)
View attachment "job-script" of type "text/plain" (7975 bytes)
View attachment "job.yaml" of type "text/plain" (5184 bytes)
View attachment "reproduce" of type "text/plain" (254 bytes)
Powered by blists - more mailing lists