Greeting,
FYI, we noticed a -2.9% regression of aim9.signal_test.ops_per_sec due to commit:
commit: 9e81889c7648d48dd5fe13f41cbc99f3c362484a ("sched: Fix affine_move_task() self-concurrency")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
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: signal_test
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 <[email protected]>
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/signal_test/aim9/300s/0xffff0190
commit:
3f1bc119cd ("sched: Optimize migration_cpu_stop()")
9e81889c76 ("sched: Fix affine_move_task() self-concurrency")
3f1bc119cd7fc987 9e81889c7648d48dd5fe13f41cb
---------------- ---------------------------
%stddev %change %stddev
\ | \
112561 -2.9% 109339 aim9.signal_test.ops_per_sec
215.24 +1.3% 218.12 aim9.time.system_time
66.75 -4.1% 64.00 aim9.time.user_time
0.07 ? 27% -20.8% 0.05 ? 6% perf-sched.sch_delay.avg.ms.devkmsg_read.vfs_read.ksys_read.do_syscall_64
52972282 +1.8% 53921623 perf-stat.i.iTLB-load-misses
52791114 +1.8% 53737428 perf-stat.ps.iTLB-load-misses
419.00 ? 4% -39.6% 253.00 ? 29% interrupts.CPU100.NMI:Non-maskable_interrupts
419.00 ? 4% -39.6% 253.00 ? 29% interrupts.CPU100.PMI:Performance_monitoring_interrupts
348.33 ? 29% -41.3% 204.50 ? 4% interrupts.CPU103.NMI:Non-maskable_interrupts
348.33 ? 29% -41.3% 204.50 ? 4% interrupts.CPU103.PMI:Performance_monitoring_interrupts
167.17 ? 5% +81.2% 302.83 ? 21% interrupts.CPU146.NMI:Non-maskable_interrupts
167.17 ? 5% +81.2% 302.83 ? 21% interrupts.CPU146.PMI:Performance_monitoring_interrupts
29.50 ?124% -76.8% 6.83 ? 43% interrupts.CPU148.RES:Rescheduling_interrupts
196.50 ? 30% +173.7% 537.83 ?113% interrupts.CPU153.NMI:Non-maskable_interrupts
196.50 ? 30% +173.7% 537.83 ?113% interrupts.CPU153.PMI:Performance_monitoring_interrupts
306.67 ? 32% -25.2% 229.50 ? 29% interrupts.CPU167.NMI:Non-maskable_interrupts
306.67 ? 32% -25.2% 229.50 ? 29% interrupts.CPU167.PMI:Performance_monitoring_interrupts
396.67 ? 3% -42.6% 227.67 ? 27% interrupts.CPU179.NMI:Non-maskable_interrupts
396.67 ? 3% -42.6% 227.67 ? 27% interrupts.CPU179.PMI:Performance_monitoring_interrupts
417.17 ? 3% -49.8% 209.50 ? 2% interrupts.CPU183.NMI:Non-maskable_interrupts
417.17 ? 3% -49.8% 209.50 ? 2% interrupts.CPU183.PMI:Performance_monitoring_interrupts
379.50 ? 20% -45.5% 206.83 interrupts.CPU188.NMI:Non-maskable_interrupts
379.50 ? 20% -45.5% 206.83 interrupts.CPU188.PMI:Performance_monitoring_interrupts
352.17 ? 28% -42.1% 204.00 ? 2% interrupts.CPU189.NMI:Non-maskable_interrupts
352.17 ? 28% -42.1% 204.00 ? 2% interrupts.CPU189.PMI:Performance_monitoring_interrupts
301.33 ? 18% -40.3% 179.83 ? 30% interrupts.CPU236.NMI:Non-maskable_interrupts
301.33 ? 18% -40.3% 179.83 ? 30% interrupts.CPU236.PMI:Performance_monitoring_interrupts
1598 ? 6% +11.3% 1779 ? 5% interrupts.CPU26.CAL:Function_call_interrupts
1609 ? 3% +18.3% 1903 ? 11% interrupts.CPU30.CAL:Function_call_interrupts
1372 ? 2% +31.9% 1809 ? 23% interrupts.CPU51.CAL:Function_call_interrupts
1477 ? 24% +64.5% 2430 ? 30% interrupts.CPU53.CAL:Function_call_interrupts
260.17 ? 27% -33.6% 172.83 ? 4% interrupts.CPU82.NMI:Non-maskable_interrupts
260.17 ? 27% -33.6% 172.83 ? 4% interrupts.CPU82.PMI:Performance_monitoring_interrupts
299.33 ? 31% -35.4% 193.33 ? 4% interrupts.CPU98.NMI:Non-maskable_interrupts
299.33 ? 31% -35.4% 193.33 ? 4% interrupts.CPU98.PMI:Performance_monitoring_interrupts
364.50 ? 19% -39.3% 221.17 ? 29% interrupts.CPU99.NMI:Non-maskable_interrupts
364.50 ? 19% -39.3% 221.17 ? 29% interrupts.CPU99.PMI:Performance_monitoring_interrupts
5828 ? 12% -13.9% 5016 ? 4% interrupts.RES:Rescheduling_interrupts
92.10 -0.6 91.48 perf-profile.calltrace.cycles-pp.secondary_startup_64_no_verify
0.81 ? 5% +0.1 0.90 ? 7% perf-profile.calltrace.cycles-pp.sighandler
0.79 ? 6% +0.1 0.90 ? 7% perf-profile.calltrace.cycles-pp.kill_pid_info.kill_something_info.__x64_sys_kill.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.89 ? 4% +0.1 1.02 ? 6% perf-profile.calltrace.cycles-pp.kill_something_info.__x64_sys_kill.do_syscall_64.entry_SYSCALL_64_after_hwframe.kill
0.46 ? 45% +0.2 0.67 ? 13% perf-profile.calltrace.cycles-pp.ret_from_fork
0.46 ? 45% +0.2 0.67 ? 13% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork
92.10 -0.6 91.48 perf-profile.children.cycles-pp.secondary_startup_64_no_verify
92.10 -0.6 91.48 perf-profile.children.cycles-pp.cpu_startup_entry
92.10 -0.6 91.48 perf-profile.children.cycles-pp.do_idle
4.69 ? 10% -0.5 4.22 ? 3% perf-profile.children.cycles-pp.ktime_get
0.50 ? 8% -0.1 0.39 ? 9% perf-profile.children.cycles-pp.tick_check_broadcast_expired
0.08 ? 12% -0.0 0.06 ? 9% perf-profile.children.cycles-pp.tick_nohz_idle_retain_tick
0.05 ? 8% +0.0 0.07 ? 9% perf-profile.children.cycles-pp.sched_clock_tick
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.drm_atomic_helper_commit
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.commit_tail
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.drm_atomic_helper_commit_tail
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.drm_atomic_helper_commit_planes
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.mgag200_simple_display_pipe_update
0.09 ? 36% +0.1 0.14 ? 31% perf-profile.children.cycles-pp.mgag200_handle_damage
0.09 ? 36% +0.1 0.15 ? 30% perf-profile.children.cycles-pp.drm_atomic_helper_dirtyfb
0.52 ? 4% +0.1 0.59 ? 7% perf-profile.children.cycles-pp.nr_iowait_cpu
0.68 ? 3% +0.1 0.76 ? 5% perf-profile.children.cycles-pp.update_ts_time_stats
0.33 ? 9% +0.1 0.43 ? 14% perf-profile.children.cycles-pp.run_posix_cpu_timers
0.80 ? 6% +0.1 0.91 ? 7% perf-profile.children.cycles-pp.kill_pid_info
0.89 ? 4% +0.1 1.03 ? 6% perf-profile.children.cycles-pp.kill_something_info
0.54 ? 9% +0.1 0.67 ? 13% perf-profile.children.cycles-pp.kthread
0.54 ? 10% +0.1 0.68 ? 13% perf-profile.children.cycles-pp.ret_from_fork
3.39 ? 14% -0.5 2.88 ? 5% perf-profile.self.cycles-pp.ktime_get
0.50 ? 7% -0.1 0.38 ? 9% perf-profile.self.cycles-pp.tick_check_broadcast_expired
0.07 ? 12% -0.0 0.04 ? 45% perf-profile.self.cycles-pp.exit_to_user_mode_prepare
0.14 ? 4% -0.0 0.12 ? 9% perf-profile.self.cycles-pp.arch_cpu_idle_enter
0.35 ? 4% +0.0 0.38 ? 5% perf-profile.self.cycles-pp.irq_enter_rcu
0.40 ? 2% +0.1 0.45 ? 3% perf-profile.self.cycles-pp.update_ts_time_stats
0.50 ? 4% +0.1 0.57 ? 7% perf-profile.self.cycles-pp.nr_iowait_cpu
0.33 ? 10% +0.1 0.43 ? 14% perf-profile.self.cycles-pp.run_posix_cpu_timers
aim9.signal_test.ops_per_sec
118000 +------------------------------------------------------------------+
| + |
116000 |-+ :: |
114000 |-+ +. + +. : : .+ |
|+.+ + ++.+ +.++.+ ++.+ :+ + .+ |
112000 |-+ : : + + O |
110000 |-+ : : O O O O O |
| +.++ O O O O OO O O OO O|
108000 |-+ O O OO OO OO O O O |
106000 |-+ OO OO O O O |
| O O |
104000 |-+ O O |
102000 |-+O O O |
|O O O |
100000 +------------------------------------------------------------------+
[*] 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/[email protected] Intel Corporation
Thanks,
Oliver Sang