| 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: Sun, 29 Nov 2020 23:07:22 +0800
From: kernel test robot <oliver.sang@...el.com>
To: Daniel Axtens <dja@...ens.net>
Cc: 0day robot <lkp@...el.com>, LKML <linux-kernel@...r.kernel.org>,
lkp@...ts.01.org, ying.huang@...el.com, feng.tang@...el.com,
zhengjun.xing@...el.com, viro@...iv.linux.org.uk,
linux-fsdevel@...r.kernel.org, David.Laight@...LAB.COM,
hch@...radead.org, Daniel Axtens <dja@...ens.net>
Subject: [fs/select.c] c191bb1b91: will-it-scale.per_thread_ops -2.3%
regression
Greeting,
FYI, we noticed a -2.3% regression of will-it-scale.per_thread_ops due to commit:
commit: c191bb1b91be4744f0cdf595dccbd4036cf24f5f ("[PATCH RESEND v2] fs/select.c: batch user writes in do_sys_poll")
url: https://github.com/0day-ci/linux/commits/Daniel-Axtens/fs-select-c-batch-user-writes-in-do_sys_poll/20201119-084654
base: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git c2e7554e1b85935d962127efa3c2a76483b0b3b6
in testcase: will-it-scale
on test machine: 144 threads Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz with 512G memory
with following parameters:
nr_task: 16
mode: thread
test: poll1
cpufreq_governor: performance
ucode: 0x16
test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two.
test-url: https://github.com/antonblanchard/will-it-scale
In addition to that, the commit also has significant impact on the following tests:
+------------------+------------------------------------------------------------------------+
| testcase: change | will-it-scale: will-it-scale.per_process_ops 6.7% improvement |
| test machine | 144 threads Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz with 512G memory |
| test parameters | cpufreq_governor=performance |
| | mode=process |
| | nr_task=100% |
| | test=poll2 |
| | ucode=0x16 |
+------------------+------------------------------------------------------------------------+
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 run job.yaml
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode:
gcc-9/performance/x86_64-rhel-8.3/thread/16/debian-10.4-x86_64-20200603.cgz/lkp-hsw-4ex1/poll1/will-it-scale/0x16
commit:
c2e7554e1b ("Merge tag 'gfs2-v5.10-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2")
c191bb1b91 ("fs/select.c: batch user writes in do_sys_poll")
c2e7554e1b85935d c191bb1b91be4744f0cdf595dcc
---------------- ---------------------------
%stddev %change %stddev
\ | \
1321041 -2.3% 1290627 will-it-scale.per_thread_ops
21136668 -2.3% 20650046 will-it-scale.workload
166093 ± 28% +89.3% 314338 ± 56% numa-numastat.node3.numa_hit
4844 -11.4% 4293 ± 11% proc-vmstat.pgactivate
461.00 +10.7% 510.50 syscalls.sys_poll.med
4.427e+09 +14.7% 5.077e+09 perf-stat.i.branch-instructions
53850560 ± 51% +35.5% 72945539 ± 4% perf-stat.i.cache-references
2.44 ± 3% -9.3% 2.21 perf-stat.i.cpi
6.426e+09 +6.8% 6.86e+09 perf-stat.i.dTLB-loads
0.39 -0.0 0.36 perf-stat.i.dTLB-store-miss-rate%
5.616e+09 ± 2% +5.4% 5.918e+09 perf-stat.i.dTLB-stores
92.10 -10.8 81.26 ± 2% perf-stat.i.iTLB-load-miss-rate%
28795351 ± 6% +16.8% 33645430 ± 6% perf-stat.i.iTLB-load-misses
2374389 ± 17% +230.8% 7853686 ± 12% perf-stat.i.iTLB-loads
2.18e+10 +11.7% 2.435e+10 perf-stat.i.instructions
0.42 ± 3% +9.3% 0.46 perf-stat.i.ipc
114.91 +8.5% 124.71 perf-stat.i.metric.M/sec
2.39 ± 2% -8.5% 2.18 perf-stat.overall.cpi
0.39 -0.0 0.36 perf-stat.overall.dTLB-store-miss-rate%
92.41 -11.3 81.08 ± 2% perf-stat.overall.iTLB-load-miss-rate%
0.42 ± 3% +9.2% 0.46 perf-stat.overall.ipc
312367 +14.1% 356394 perf-stat.overall.path-length
4.412e+09 +14.7% 5.061e+09 perf-stat.ps.branch-instructions
53671657 ± 51% +35.5% 72700879 ± 4% perf-stat.ps.cache-references
6.405e+09 +6.8% 6.837e+09 perf-stat.ps.dTLB-loads
5.598e+09 ± 2% +5.4% 5.898e+09 perf-stat.ps.dTLB-stores
28698099 ± 6% +16.8% 33530337 ± 6% perf-stat.ps.iTLB-load-misses
2366699 ± 17% +230.7% 7827623 ± 12% perf-stat.ps.iTLB-loads
2.173e+10 +11.7% 2.427e+10 perf-stat.ps.instructions
6.602e+12 +11.5% 7.36e+12 perf-stat.total.instructions
39.25 ±111% -88.5% 4.50 ± 45% interrupts.CPU110.RES:Rescheduling_interrupts
96.75 ± 4% +61.5% 156.25 ± 32% interrupts.CPU124.NMI:Non-maskable_interrupts
96.75 ± 4% +61.5% 156.25 ± 32% interrupts.CPU124.PMI:Performance_monitoring_interrupts
96.50 ± 35% +79.0% 172.75 ± 20% interrupts.CPU127.NMI:Non-maskable_interrupts
96.50 ± 35% +79.0% 172.75 ± 20% interrupts.CPU127.PMI:Performance_monitoring_interrupts
2.75 ± 69% +2118.2% 61.00 ±153% interrupts.CPU138.RES:Rescheduling_interrupts
99.50 ± 14% +52.8% 152.00 ± 35% interrupts.CPU139.NMI:Non-maskable_interrupts
99.50 ± 14% +52.8% 152.00 ± 35% interrupts.CPU139.PMI:Performance_monitoring_interrupts
121.25 ± 34% +41.4% 171.50 ± 25% interrupts.CPU142.NMI:Non-maskable_interrupts
121.25 ± 34% +41.4% 171.50 ± 25% interrupts.CPU142.PMI:Performance_monitoring_interrupts
3435 ± 7% +44.2% 4954 ± 20% interrupts.CPU15.NMI:Non-maskable_interrupts
3435 ± 7% +44.2% 4954 ± 20% interrupts.CPU15.PMI:Performance_monitoring_interrupts
7029 ± 8% -47.0% 3728 ± 13% interrupts.CPU5.NMI:Non-maskable_interrupts
7029 ± 8% -47.0% 3728 ± 13% interrupts.CPU5.PMI:Performance_monitoring_interrupts
112.50 ± 15% +50.7% 169.50 ± 21% interrupts.CPU71.NMI:Non-maskable_interrupts
112.50 ± 15% +50.7% 169.50 ± 21% interrupts.CPU71.PMI:Performance_monitoring_interrupts
5583 ± 27% -35.2% 3617 ± 34% interrupts.CPU75.NMI:Non-maskable_interrupts
5583 ± 27% -35.2% 3617 ± 34% interrupts.CPU75.PMI:Performance_monitoring_interrupts
233.50 ± 8% -37.6% 145.75 ± 39% interrupts.CPU81.TLB:TLB_shootdowns
4074 ± 22% +27.2% 5182 ± 21% interrupts.CPU9.NMI:Non-maskable_interrupts
4074 ± 22% +27.2% 5182 ± 21% interrupts.CPU9.PMI:Performance_monitoring_interrupts
291.00 ± 9% +25.4% 365.00 ± 14% interrupts.CPU9.TLB:TLB_shootdowns
4.25 ± 67% +764.7% 36.75 ±108% interrupts.CPU91.RES:Rescheduling_interrupts
87.75 ± 32% +125.4% 197.75 ± 14% interrupts.CPU94.NMI:Non-maskable_interrupts
87.75 ± 32% +125.4% 197.75 ± 14% interrupts.CPU94.PMI:Performance_monitoring_interrupts
9830 ± 12% -12.0% 8650 ± 3% interrupts.TLB:TLB_shootdowns
0.00 +0.7 0.68 ± 10% perf-profile.calltrace.cycles-pp._copy_to_user.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
5.02 ± 9% +1.0 5.97 ± 10% perf-profile.calltrace.cycles-pp.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
6.37 ± 9% +1.1 7.46 ± 10% perf-profile.calltrace.cycles-pp.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
13.52 ± 8% +1.5 15.05 ± 10% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
0.05 ± 8% +0.1 0.11 ± 14% perf-profile.children.cycles-pp.copy_user_enhanced_fast_string
0.15 ± 20% +0.1 0.24 ± 10% perf-profile.children.cycles-pp.check_stack_object
0.14 ± 9% +0.1 0.23 ± 11% perf-profile.children.cycles-pp.copy_user_generic_unrolled
0.33 ± 11% +0.1 0.47 ± 9% perf-profile.children.cycles-pp.___might_sleep
0.19 ± 11% +0.1 0.33 ± 14% perf-profile.children.cycles-pp.exit_to_user_mode_prepare
0.19 ± 13% +0.2 0.35 ± 9% perf-profile.children.cycles-pp.__might_sleep
0.32 ± 7% +0.2 0.56 ± 12% perf-profile.children.cycles-pp.__check_object_size
0.64 ± 12% +0.4 1.00 ± 9% perf-profile.children.cycles-pp.__might_fault
0.00 +0.7 0.70 ± 10% perf-profile.children.cycles-pp._copy_to_user
5.14 ± 9% +1.0 6.15 ± 10% perf-profile.children.cycles-pp.do_sys_poll
6.39 ± 9% +1.1 7.51 ± 10% perf-profile.children.cycles-pp.__x64_sys_poll
13.59 ± 8% +1.5 15.12 ± 10% perf-profile.children.cycles-pp.do_syscall_64
0.13 ± 11% +0.0 0.16 ± 11% perf-profile.self.cycles-pp.poll_freewait
0.15 ± 19% +0.1 0.21 ± 12% perf-profile.self.cycles-pp.check_stack_object
0.00 +0.1 0.06 ± 14% perf-profile.self.cycles-pp.copy_user_enhanced_fast_string
0.17 ± 13% +0.1 0.22 ± 12% perf-profile.self.cycles-pp.do_syscall_64
0.11 ± 11% +0.1 0.17 ± 9% perf-profile.self.cycles-pp.copy_user_generic_unrolled
0.13 ± 18% +0.1 0.21 ± 13% perf-profile.self.cycles-pp.__might_fault
0.16 ± 10% +0.1 0.30 ± 8% perf-profile.self.cycles-pp.__might_sleep
0.32 ± 11% +0.1 0.46 ± 11% perf-profile.self.cycles-pp.___might_sleep
0.60 ± 14% +0.1 0.74 ± 10% perf-profile.self.cycles-pp.ftrace_syscall_exit
0.16 ± 11% +0.1 0.30 ± 14% perf-profile.self.cycles-pp.exit_to_user_mode_prepare
0.17 ± 4% +0.2 0.33 ± 12% perf-profile.self.cycles-pp.__check_object_size
0.00 +0.2 0.21 ± 8% perf-profile.self.cycles-pp._copy_to_user
10910 ± 13% -17.9% 8955 ± 7% softirqs.CPU109.RCU
9932 ± 15% -17.0% 8240 ± 7% softirqs.CPU110.RCU
17714 ± 20% -53.4% 8261 ± 33% softirqs.CPU12.SCHED
11377 ± 27% -23.7% 8680 ± 5% softirqs.CPU128.RCU
18811 ± 3% -25.8% 13967 ± 13% softirqs.CPU13.SCHED
13684 ± 11% -21.6% 10728 ± 13% softirqs.CPU132.RCU
10948 ± 29% -24.8% 8238 ± 9% softirqs.CPU16.RCU
13961 ± 13% -19.8% 11198 ± 6% softirqs.CPU19.RCU
12708 ± 18% -18.8% 10318 ± 5% softirqs.CPU20.RCU
15297 ± 22% -30.5% 10629 ± 2% softirqs.CPU21.RCU
16571 ± 9% -20.1% 13236 ± 14% softirqs.CPU23.RCU
14472 ± 10% -16.3% 12114 ± 7% softirqs.CPU26.RCU
11203 ± 8% -22.4% 8688 ± 14% softirqs.CPU32.RCU
13804 ± 11% -19.1% 11171 ± 9% softirqs.CPU38.RCU
12160 ± 23% -20.9% 9618 ± 3% softirqs.CPU42.RCU
13930 ± 5% -15.6% 11758 ± 13% softirqs.CPU48.RCU
14228 ± 4% -23.0% 10954 ± 19% softirqs.CPU51.RCU
12704 ± 12% -21.0% 10038 ± 7% softirqs.CPU55.RCU
11699 ± 17% -18.1% 9583 ± 7% softirqs.CPU56.RCU
12408 ± 8% -22.6% 9598 ± 8% softirqs.CPU64.RCU
13203 ± 4% -17.2% 10926 ± 14% softirqs.CPU68.RCU
14688 ± 5% -9.7% 13269 ± 6% softirqs.CPU74.RCU
14748 ± 3% -13.0% 12829 ± 9% softirqs.CPU79.RCU
21729 ± 12% +30.0% 28254 ± 17% softirqs.CPU82.SCHED
20503 ± 10% +33.5% 27374 ± 23% softirqs.CPU83.SCHED
21808 ± 15% +59.9% 34871 ± 4% softirqs.CPU84.SCHED
20813 ± 2% +42.9% 29732 ± 15% softirqs.CPU85.SCHED
16729 ± 38% +49.6% 25033 ± 18% softirqs.CPU86.SCHED
10712 ± 23% -27.6% 7750 ± 3% softirqs.CPU89.RCU
9737 ± 16% -17.8% 8004 ± 7% softirqs.CPU92.RCU
13145 ± 12% -25.3% 9817 ± 12% softirqs.CPU96.RCU
14061 ± 7% -24.4% 10632 ± 15% softirqs.CPU97.RCU
will-it-scale.per_thread_ops
1.325e+06 +---------------------------------------------------------------+
| .+ +.+ +..+.+.+ + + .|
1.32e+06 |.+ .+.+. : : + + : .+ +..+ |
1.315e+06 |-+..+.+ : : +.+..+.+ : .+.+.+..+ |
| + : +. |
1.31e+06 |-+ + : |
1.305e+06 |-+ + |
| |
1.3e+06 |-+ |
1.295e+06 |-+ |
| O O O O O O O O |
1.29e+06 |-+ O O O O O O O O O |
1.285e+06 |-+ O O |
| O O O |
1.28e+06 +---------------------------------------------------------------+
syscalls.sys_poll.med
520 +---------------------------------------------------------------------+
| |
510 |-+O O O O O O O O O O O O O O O O O O O |
| O |
500 |-+ |
| O O |
490 |-+ |
| |
480 |-+ |
| |
470 |-+ |
| .+. .+.+ .+.. .+..+.+..+.+..+.+.. |
460 |-.+.+..+.+. +. + .+ + +.+..+.|
|. + .+.+..+.+. |
450 +---------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
***************************************************************************************************
lkp-hsw-4ex1: 144 threads Intel(R) Xeon(R) CPU E7-8890 v3 @ 2.50GHz with 512G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode:
gcc-9/performance/x86_64-rhel-8.3/process/100%/debian-10.4-x86_64-20200603.cgz/lkp-hsw-4ex1/poll2/will-it-scale/0x16
commit:
c2e7554e1b ("Merge tag 'gfs2-v5.10-rc4-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/gfs2/linux-gfs2")
c191bb1b91 ("fs/select.c: batch user writes in do_sys_poll")
c2e7554e1b85935d c191bb1b91be4744f0cdf595dcc
---------------- ---------------------------
%stddev %change %stddev
\ | \
274620 +6.7% 292996 will-it-scale.per_process_ops
39545400 +6.7% 42191543 will-it-scale.workload
5419 +8.5% 5882 ± 6% numa-meminfo.node3.KernelStack
239172 ± 3% -8.5% 218745 ± 7% numa-numastat.node0.numa_hit
135.64 ± 18% -33.2% 90.55 ± 3% sched_debug.cfs_rq:/.util_est_enqueued.stddev
18.00 +5.6% 19.00 vmstat.cpu.us
42108 ± 61% +57.4% 66280 ± 42% numa-vmstat.node0.numa_other
5420 +8.5% 5881 ± 6% numa-vmstat.node3.nr_kernel_stack
5691 ± 7% -17.4% 4702 ± 4% slabinfo.eventpoll_pwq.active_objs
5691 ± 7% -17.4% 4702 ± 4% slabinfo.eventpoll_pwq.num_objs
38.09 ± 9% -6.6% 35.57 boot-time.boot
26.35 -1.5% 25.96 boot-time.dhcp
4631 ± 10% -7.8% 4271 boot-time.idle
19539 ± 74% -86.7% 2595 ±103% syscalls.sys_close.max
592.00 ± 4% -53.4% 275.75 ±100% syscalls.sys_close.min
38616838 ± 24% -2.2e+07 16663575 ±102% syscalls.sys_close.noise.50%
1.829e+08 ± 12% +9.5e+07 2.781e+08 ± 17% syscalls.sys_mmap.noise.100%
2.962e+08 ± 11% +9.1e+07 3.87e+08 ± 10% syscalls.sys_mmap.noise.2%
2.757e+08 ± 13% +9.2e+07 3.678e+08 ± 11% syscalls.sys_mmap.noise.25%
2.952e+08 ± 12% +9.1e+07 3.858e+08 ± 10% syscalls.sys_mmap.noise.5%
2.397e+08 ± 17% +9.8e+07 3.377e+08 ± 11% syscalls.sys_mmap.noise.50%
82708 ± 19% +38.2% 114327 ± 14% syscalls.sys_openat.max
5465 ± 6% +7.6% 5881 ± 4% syscalls.sys_openat.min
1.735e+08 ± 11% +7.3e+07 2.463e+08 ± 15% syscalls.sys_openat.noise.100%
2.604e+08 ± 9% +8.8e+07 3.481e+08 ± 12% syscalls.sys_openat.noise.2%
2.389e+08 ± 8% +8.5e+07 3.244e+08 ± 12% syscalls.sys_openat.noise.25%
2.588e+08 ± 9% +8.9e+07 3.475e+08 ± 12% syscalls.sys_openat.noise.5%
2.139e+08 ± 14% +9.3e+07 3.07e+08 ± 13% syscalls.sys_openat.noise.50%
1.909e+08 ± 13% +8.3e+07 2.736e+08 ± 11% syscalls.sys_openat.noise.75%
5197 -10.2% 4667 syscalls.sys_poll.min
1.075e+11 -9.8% 9.693e+10 perf-stat.i.branch-instructions
0.30 -0.0 0.25 ± 2% perf-stat.i.branch-miss-rate%
3.112e+08 -24.3% 2.355e+08 perf-stat.i.branch-misses
0.81 +7.0% 0.87 perf-stat.i.cpi
1.081e+11 -3.3% 1.046e+11 perf-stat.i.dTLB-loads
0.07 +0.0 0.08 ± 2% perf-stat.i.dTLB-store-miss-rate%
39690700 +8.2% 42933145 ± 2% perf-stat.i.dTLB-store-misses
6.008e+10 -5.4% 5.683e+10 perf-stat.i.dTLB-stores
40551436 +27.0% 51498580 ± 2% perf-stat.i.iTLB-load-misses
4.92e+11 -6.5% 4.6e+11 perf-stat.i.instructions
12115 -26.2% 8944 ± 2% perf-stat.i.instructions-per-iTLB-miss
1.23 -6.7% 1.15 perf-stat.i.ipc
0.95 ± 2% -9.9% 0.85 perf-stat.i.metric.K/sec
1914 -6.3% 1794 perf-stat.i.metric.M/sec
861498 +5.0% 904380 perf-stat.i.node-stores
0.29 -0.0 0.24 perf-stat.overall.branch-miss-rate%
0.81 +7.3% 0.87 perf-stat.overall.cpi
0.07 +0.0 0.08 ± 2% perf-stat.overall.dTLB-store-miss-rate%
12138 -26.4% 8937 ± 2% perf-stat.overall.instructions-per-iTLB-miss
1.24 -6.8% 1.15 perf-stat.overall.ipc
3752635 -12.3% 3291386 perf-stat.overall.path-length
1.071e+11 -9.8% 9.659e+10 perf-stat.ps.branch-instructions
3.101e+08 -24.3% 2.347e+08 perf-stat.ps.branch-misses
1.078e+11 -3.3% 1.042e+11 perf-stat.ps.dTLB-loads
39554490 +8.2% 42784025 ± 2% perf-stat.ps.dTLB-store-misses
5.988e+10 -5.4% 5.664e+10 perf-stat.ps.dTLB-stores
40411523 +27.0% 51320791 ± 2% perf-stat.ps.iTLB-load-misses
4.904e+11 -6.5% 4.585e+11 perf-stat.ps.instructions
873555 +5.2% 918647 perf-stat.ps.node-stores
1.484e+14 -6.4% 1.389e+14 perf-stat.total.instructions
1034 ± 18% -25.0% 776.50 ± 21% interrupts.CPU0.CAL:Function_call_interrupts
3102 ± 29% -32.4% 2096 ± 31% interrupts.CPU109.CAL:Function_call_interrupts
309.00 +16.7% 360.75 ± 11% interrupts.CPU11.RES:Rescheduling_interrupts
3896 +51.8% 5915 ± 33% interrupts.CPU112.NMI:Non-maskable_interrupts
3896 +51.8% 5915 ± 33% interrupts.CPU112.PMI:Performance_monitoring_interrupts
3896 +76.3% 6870 ± 24% interrupts.CPU116.NMI:Non-maskable_interrupts
3896 +76.3% 6870 ± 24% interrupts.CPU116.PMI:Performance_monitoring_interrupts
3897 +77.1% 6900 ± 25% interrupts.CPU117.NMI:Non-maskable_interrupts
3897 +77.1% 6900 ± 25% interrupts.CPU117.PMI:Performance_monitoring_interrupts
403.00 ± 26% -25.6% 299.75 interrupts.CPU118.RES:Rescheduling_interrupts
4886 ± 34% +61.2% 7879 interrupts.CPU120.NMI:Non-maskable_interrupts
4886 ± 34% +61.2% 7879 interrupts.CPU120.PMI:Performance_monitoring_interrupts
4892 ± 34% +40.8% 6886 ± 24% interrupts.CPU122.NMI:Non-maskable_interrupts
4892 ± 34% +40.8% 6886 ± 24% interrupts.CPU122.PMI:Performance_monitoring_interrupts
303.00 ± 2% +122.4% 673.75 ± 66% interrupts.CPU130.RES:Rescheduling_interrupts
874.75 ± 22% -21.7% 685.00 interrupts.CPU133.CAL:Function_call_interrupts
6892 ± 24% -43.0% 3930 interrupts.CPU143.NMI:Non-maskable_interrupts
6892 ± 24% -43.0% 3930 interrupts.CPU143.PMI:Performance_monitoring_interrupts
4840 ± 34% +23.8% 5990 ± 29% interrupts.CPU15.NMI:Non-maskable_interrupts
4840 ± 34% +23.8% 5990 ± 29% interrupts.CPU15.PMI:Performance_monitoring_interrupts
4912 ± 31% +57.1% 7717 interrupts.CPU25.NMI:Non-maskable_interrupts
4912 ± 31% +57.1% 7717 interrupts.CPU25.PMI:Performance_monitoring_interrupts
852.25 ± 27% -19.7% 684.50 interrupts.CPU3.CAL:Function_call_interrupts
4913 ± 34% +30.2% 6396 ± 23% interrupts.CPU43.NMI:Non-maskable_interrupts
4913 ± 34% +30.2% 6396 ± 23% interrupts.CPU43.PMI:Performance_monitoring_interrupts
505.00 ± 27% -34.8% 329.25 ± 3% interrupts.CPU56.RES:Rescheduling_interrupts
4816 ± 34% +41.5% 6814 ± 24% interrupts.CPU74.NMI:Non-maskable_interrupts
4816 ± 34% +41.5% 6814 ± 24% interrupts.CPU74.PMI:Performance_monitoring_interrupts
4360 ± 19% +60.5% 6997 ± 19% interrupts.CPU78.NMI:Non-maskable_interrupts
4360 ± 19% +60.5% 6997 ± 19% interrupts.CPU78.PMI:Performance_monitoring_interrupts
570.25 ± 43% -44.1% 318.50 ± 7% interrupts.CPU79.RES:Rescheduling_interrupts
3887 +50.0% 5832 ± 32% interrupts.CPU8.NMI:Non-maskable_interrupts
3887 +50.0% 5832 ± 32% interrupts.CPU8.PMI:Performance_monitoring_interrupts
4831 ± 35% +61.6% 7807 interrupts.CPU81.NMI:Non-maskable_interrupts
4831 ± 35% +61.6% 7807 interrupts.CPU81.PMI:Performance_monitoring_interrupts
4828 ± 33% +55.5% 7508 ± 5% interrupts.CPU88.NMI:Non-maskable_interrupts
4828 ± 33% +55.5% 7508 ± 5% interrupts.CPU88.PMI:Performance_monitoring_interrupts
688.25 +37.5% 946.00 ± 38% interrupts.CPU94.CAL:Function_call_interrupts
6463 ± 24% -40.4% 3852 interrupts.CPU98.NMI:Non-maskable_interrupts
6463 ± 24% -40.4% 3852 interrupts.CPU98.PMI:Performance_monitoring_interrupts
24015 ± 3% -20.1% 19190 ± 11% softirqs.CPU0.RCU
23795 ± 13% -24.6% 17932 ± 11% softirqs.CPU1.RCU
20222 ± 4% -18.7% 16444 ± 11% softirqs.CPU10.RCU
19790 ± 5% -13.9% 17031 ± 12% softirqs.CPU100.RCU
20600 ± 7% -14.5% 17608 ± 13% softirqs.CPU103.RCU
19617 ± 5% -14.4% 16800 ± 12% softirqs.CPU104.RCU
20111 ± 4% -17.0% 16698 ± 11% softirqs.CPU105.RCU
20401 ± 4% -15.6% 17213 ± 12% softirqs.CPU106.RCU
20290 ± 5% -13.3% 17601 ± 9% softirqs.CPU107.RCU
15162 ± 8% -13.9% 13048 ± 7% softirqs.CPU111.RCU
19858 ± 7% -17.1% 16469 ± 13% softirqs.CPU112.RCU
21004 ± 4% -17.0% 17424 ± 16% softirqs.CPU113.RCU
20941 ± 6% -18.3% 17111 ± 13% softirqs.CPU114.RCU
20866 ± 5% -18.6% 16994 ± 14% softirqs.CPU115.RCU
20680 ± 8% -15.4% 17486 ± 16% softirqs.CPU116.RCU
20294 ± 6% -16.8% 16893 ± 14% softirqs.CPU118.RCU
18823 ± 7% -13.5% 16284 ± 13% softirqs.CPU119.RCU
18999 -19.2% 15347 ± 9% softirqs.CPU12.RCU
19549 ± 5% -12.9% 17033 ± 15% softirqs.CPU120.RCU
19347 ± 6% -14.9% 16459 ± 12% softirqs.CPU121.RCU
20077 -15.0% 17076 ± 11% softirqs.CPU122.RCU
18921 ± 6% -14.6% 16161 ± 13% softirqs.CPU123.RCU
19733 ± 7% -16.6% 16447 ± 14% softirqs.CPU124.RCU
18747 ± 7% -14.2% 16081 ± 12% softirqs.CPU125.RCU
19576 ± 2% -21.5% 15364 ± 15% softirqs.CPU13.RCU
19633 ± 6% -17.8% 16136 ± 11% softirqs.CPU14.RCU
19474 -17.0% 16157 ± 8% softirqs.CPU15.RCU
13032 ± 5% -13.3% 11296 ± 4% softirqs.CPU16.RCU
21118 ± 4% -24.6% 15931 ± 15% softirqs.CPU18.RCU
21481 ± 6% -21.6% 16848 ± 13% softirqs.CPU19.RCU
20905 ± 5% -18.9% 16950 ± 8% softirqs.CPU2.RCU
20451 ± 5% -19.4% 16484 ± 14% softirqs.CPU20.RCU
21874 ± 13% -22.6% 16931 ± 13% softirqs.CPU21.RCU
20134 ± 6% -19.9% 16132 ± 11% softirqs.CPU22.RCU
21091 ± 8% -19.1% 17072 ± 13% softirqs.CPU23.RCU
20274 ± 7% -16.0% 17021 ± 15% softirqs.CPU24.RCU
20052 ± 5% -19.0% 16238 ± 12% softirqs.CPU25.RCU
20312 ± 6% -17.6% 16742 ± 13% softirqs.CPU26.RCU
19908 ± 4% -17.9% 16344 ± 14% softirqs.CPU27.RCU
19007 ± 4% -14.6% 16225 ± 13% softirqs.CPU28.RCU
18961 ± 2% -14.8% 16153 ± 12% softirqs.CPU29.RCU
19179 ± 7% -14.1% 16480 ± 14% softirqs.CPU31.RCU
15175 ± 6% -14.3% 13008 ± 8% softirqs.CPU33.RCU
15497 ± 7% -15.5% 13091 ± 9% softirqs.CPU34.RCU
20075 ± 7% -18.6% 16337 ± 13% softirqs.CPU36.RCU
19733 ± 7% -16.6% 16461 ± 12% softirqs.CPU37.RCU
19773 ± 9% -18.5% 16109 ± 14% softirqs.CPU38.RCU
20016 ± 9% -18.7% 16264 ± 12% softirqs.CPU39.RCU
19876 ± 2% -15.4% 16818 ± 11% softirqs.CPU4.RCU
19783 ± 8% -17.9% 16241 ± 13% softirqs.CPU40.RCU
19899 ± 7% -18.7% 16184 ± 15% softirqs.CPU41.RCU
19607 ± 7% -19.6% 15756 ± 13% softirqs.CPU42.RCU
20270 ± 9% -20.1% 16197 ± 14% softirqs.CPU43.RCU
20019 ± 7% -17.8% 16447 ± 13% softirqs.CPU46.RCU
19490 ± 7% -13.7% 16826 ± 14% softirqs.CPU47.RCU
21764 ± 3% -21.0% 17185 ± 13% softirqs.CPU5.RCU
21386 ± 2% -20.0% 17107 ± 11% softirqs.CPU6.RCU
17298 ± 3% -12.3% 15177 ± 12% softirqs.CPU61.RCU
20348 ± 3% -17.5% 16796 ± 9% softirqs.CPU7.RCU
17202 ± 5% -12.7% 15017 ± 9% softirqs.CPU70.RCU
21532 ± 2% -12.9% 18755 ± 4% softirqs.CPU72.RCU
19809 ± 4% -14.8% 16880 ± 10% softirqs.CPU73.RCU
17594 ± 3% -14.4% 15055 ± 10% softirqs.CPU76.RCU
18105 ± 6% -16.7% 15084 ± 10% softirqs.CPU79.RCU
19810 ± 2% -17.1% 16424 ± 9% softirqs.CPU8.RCU
19135 ± 10% -21.5% 15018 ± 9% softirqs.CPU80.RCU
19213 ± 3% -18.4% 15684 ± 13% softirqs.CPU81.RCU
19420 ± 5% -18.6% 15806 ± 9% softirqs.CPU84.RCU
19353 ± 6% -17.9% 15894 ± 13% softirqs.CPU85.RCU
20126 ± 6% -21.3% 15845 ± 12% softirqs.CPU86.RCU
19408 ± 5% -17.9% 15934 ± 11% softirqs.CPU87.RCU
23453 ± 14% -32.5% 15824 ± 11% softirqs.CPU88.RCU
19281 ± 6% -19.8% 15460 ± 12% softirqs.CPU89.RCU
19960 ± 3% -19.2% 16124 ± 11% softirqs.CPU9.RCU
18900 ± 4% -17.5% 15590 ± 10% softirqs.CPU91.RCU
16433 ± 5% -16.2% 13779 ± 13% softirqs.CPU92.RCU
16210 ± 4% -15.2% 13750 ± 11% softirqs.CPU93.RCU
16136 ± 3% -15.3% 13664 ± 10% softirqs.CPU94.RCU
16855 ± 7% -15.3% 14283 ± 11% softirqs.CPU95.RCU
20517 ± 7% -16.2% 17190 ± 10% softirqs.CPU96.RCU
20516 ± 7% -19.7% 16468 ± 11% softirqs.CPU97.RCU
20661 ± 6% -18.1% 16923 ± 13% softirqs.CPU98.RCU
20716 ± 5% -18.2% 16952 ± 13% softirqs.CPU99.RCU
2696425 ± 3% -14.0% 2318720 ± 11% softirqs.RCU
6.06 -6.1 0.00 perf-profile.calltrace.cycles-pp.__put_user_nocheck_2.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
68.49 -1.9 66.62 perf-profile.calltrace.cycles-pp.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
74.30 -1.6 72.68 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
65.03 -1.1 63.96 perf-profile.calltrace.cycles-pp.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
86.64 -0.9 85.74 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__poll
92.74 -0.3 92.46 perf-profile.calltrace.cycles-pp.__poll
1.18 ± 2% -0.2 0.99 perf-profile.calltrace.cycles-pp.kfree.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.56 +0.0 0.59 perf-profile.calltrace.cycles-pp.trace_clock_x86_tsc.__rb_reserve_next.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_exit
0.71 +0.0 0.74 ± 2% perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe.__poll
0.63 +0.0 0.68 perf-profile.calltrace.cycles-pp.memcpy_erms.ftrace_syscall_enter.syscall_trace_enter.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.60 +0.0 0.65 ± 2% perf-profile.calltrace.cycles-pp.trace_clock_x86_tsc.__rb_reserve_next.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_enter
1.27 +0.1 1.34 ± 2% perf-profile.calltrace.cycles-pp.__rb_reserve_next.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_exit.syscall_exit_to_user_mode
1.76 +0.1 1.87 perf-profile.calltrace.cycles-pp.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_exit.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe
1.29 +0.1 1.41 ± 2% perf-profile.calltrace.cycles-pp.__rb_reserve_next.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_enter.syscall_trace_enter
2.15 +0.1 2.27 perf-profile.calltrace.cycles-pp.trace_buffer_lock_reserve.ftrace_syscall_exit.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe.__poll
2.43 +0.1 2.56 perf-profile.calltrace.cycles-pp.__entry_text_start.__poll
1.76 +0.1 1.90 ± 2% perf-profile.calltrace.cycles-pp.ring_buffer_lock_reserve.trace_buffer_lock_reserve.ftrace_syscall_enter.syscall_trace_enter.do_syscall_64
2.16 +0.2 2.32 ± 2% perf-profile.calltrace.cycles-pp.trace_buffer_lock_reserve.ftrace_syscall_enter.syscall_trace_enter.do_syscall_64.entry_SYSCALL_64_after_hwframe
2.70 +0.2 2.85 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.__poll
4.65 +0.2 4.84 perf-profile.calltrace.cycles-pp.__fdget.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
5.30 +0.2 5.51 perf-profile.calltrace.cycles-pp.syscall_trace_enter.do_syscall_64.entry_SYSCALL_64_after_hwframe.__poll
3.58 +0.3 3.83 ± 2% perf-profile.calltrace.cycles-pp.ftrace_syscall_exit.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe.__poll
6.68 ± 2% +0.3 6.94 perf-profile.calltrace.cycles-pp.testcase
2.39 ± 5% +0.3 2.66 perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64
3.48 ± 4% +0.3 3.79 perf-profile.calltrace.cycles-pp._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.52 ± 3% +0.4 0.92 ± 5% perf-profile.calltrace.cycles-pp.__virt_addr_valid.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64
0.00 +0.5 0.52 perf-profile.calltrace.cycles-pp.check_stack_object.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64
11.81 +0.7 12.52 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe.__poll
0.00 +0.7 0.72 perf-profile.calltrace.cycles-pp.__might_fault._copy_to_user.do_sys_poll.__x64_sys_poll.do_syscall_64
0.00 +1.1 1.07 perf-profile.calltrace.cycles-pp.__check_heap_object.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64
0.70 ± 2% +1.1 1.79 ± 6% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_safe_stack.__poll
24.95 +1.6 26.52 perf-profile.calltrace.cycles-pp.__fget_light.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.79 +2.1 3.93 perf-profile.calltrace.cycles-pp.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +2.2 2.25 perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string._copy_to_user.do_sys_poll.__x64_sys_poll.do_syscall_64
0.00 +3.3 3.29 perf-profile.calltrace.cycles-pp._copy_to_user.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe
5.23 -5.2 0.00 perf-profile.children.cycles-pp.__put_user_nocheck_2
67.28 -1.9 65.34 perf-profile.children.cycles-pp.do_sys_poll
68.53 -1.9 66.66 perf-profile.children.cycles-pp.__x64_sys_poll
74.38 -1.6 72.75 perf-profile.children.cycles-pp.do_syscall_64
86.71 -0.9 85.80 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
93.32 -0.3 93.05 perf-profile.children.cycles-pp.__poll
1.19 -0.2 0.99 perf-profile.children.cycles-pp.kfree
0.23 ± 3% -0.0 0.21 ± 2% perf-profile.children.cycles-pp.poll_freewait
0.22 +0.0 0.23 perf-profile.children.cycles-pp.poll@plt
0.22 +0.0 0.23 perf-profile.children.cycles-pp.__x86_retpoline_rax
0.52 ± 2% +0.0 0.54 ± 2% perf-profile.children.cycles-pp.tracing_generic_entry_update
0.27 +0.0 0.29 perf-profile.children.cycles-pp.syscall_enter_from_user_mode
0.72 +0.0 0.77 perf-profile.children.cycles-pp.exit_to_user_mode_prepare
1.18 +0.1 1.27 perf-profile.children.cycles-pp.trace_clock_x86_tsc
3.10 +0.2 3.27 perf-profile.children.cycles-pp.syscall_return_via_sysret
3.99 +0.2 4.17 perf-profile.children.cycles-pp.__fdget
2.59 +0.2 2.79 ± 2% perf-profile.children.cycles-pp.__rb_reserve_next
5.31 +0.2 5.53 perf-profile.children.cycles-pp.syscall_trace_enter
3.55 +0.2 3.80 ± 2% perf-profile.children.cycles-pp.ring_buffer_lock_reserve
3.63 +0.3 3.89 ± 2% perf-profile.children.cycles-pp.ftrace_syscall_exit
6.75 ± 2% +0.3 7.01 perf-profile.children.cycles-pp.testcase
0.34 +0.3 0.61 perf-profile.children.cycles-pp.__might_sleep
4.39 +0.3 4.67 ± 2% perf-profile.children.cycles-pp.trace_buffer_lock_reserve
0.71 ± 2% +0.3 1.04 ± 5% perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
3.58 ± 3% +0.3 3.91 perf-profile.children.cycles-pp._copy_from_user
0.22 ± 3% +0.3 0.56 perf-profile.children.cycles-pp.check_stack_object
0.49 +0.4 0.87 perf-profile.children.cycles-pp.___might_sleep
0.52 ± 3% +0.4 0.92 ± 5% perf-profile.children.cycles-pp.__virt_addr_valid
2.88 +0.4 3.32 perf-profile.children.cycles-pp.__entry_text_start
0.48 +0.6 1.10 perf-profile.children.cycles-pp.__check_heap_object
11.91 +0.7 12.62 perf-profile.children.cycles-pp.syscall_exit_to_user_mode
0.80 +0.8 1.60 perf-profile.children.cycles-pp.__might_fault
24.36 +1.7 26.05 perf-profile.children.cycles-pp.__fget_light
1.89 +2.3 4.14 perf-profile.children.cycles-pp.__check_object_size
2.43 ± 5% +2.5 4.97 perf-profile.children.cycles-pp.copy_user_enhanced_fast_string
0.00 +3.4 3.37 perf-profile.children.cycles-pp._copy_to_user
30.87 -5.8 25.09 perf-profile.self.cycles-pp.do_sys_poll
1.17 ± 2% -0.2 0.98 perf-profile.self.cycles-pp.kfree
0.21 ± 2% -0.0 0.19 perf-profile.self.cycles-pp.poll_freewait
0.48 +0.0 0.50 ± 2% perf-profile.self.cycles-pp.tracing_generic_entry_update
0.23 ± 3% +0.0 0.26 perf-profile.self.cycles-pp.syscall_enter_from_user_mode
0.40 ± 2% +0.0 0.43 perf-profile.self.cycles-pp._copy_from_user
0.60 +0.0 0.64 perf-profile.self.cycles-pp.exit_to_user_mode_prepare
1.08 +0.1 1.15 ± 2% perf-profile.self.cycles-pp.__x64_sys_poll
1.13 +0.1 1.20 ± 2% perf-profile.self.cycles-pp.trace_clock_x86_tsc
0.59 ± 3% +0.1 0.68 ± 3% perf-profile.self.cycles-pp.ftrace_syscall_exit
1.70 +0.1 1.80 perf-profile.self.cycles-pp.__fdget
1.29 +0.1 1.40 ± 2% perf-profile.self.cycles-pp.__rb_reserve_next
0.23 ± 2% +0.2 0.40 ± 2% perf-profile.self.cycles-pp.__might_fault
3.08 +0.2 3.26 perf-profile.self.cycles-pp.syscall_return_via_sysret
0.31 +0.3 0.56 perf-profile.self.cycles-pp.__might_sleep
6.56 ± 2% +0.3 6.83 perf-profile.self.cycles-pp.testcase
0.17 ± 2% +0.3 0.46 perf-profile.self.cycles-pp.check_stack_object
0.00 +0.4 0.35 perf-profile.self.cycles-pp._copy_to_user
0.48 +0.4 0.86 perf-profile.self.cycles-pp.___might_sleep
0.49 ± 3% +0.4 0.87 ± 6% perf-profile.self.cycles-pp.__virt_addr_valid
2.44 +0.4 2.86 perf-profile.self.cycles-pp.__entry_text_start
7.64 +0.4 8.07 perf-profile.self.cycles-pp.syscall_exit_to_user_mode
0.46 +0.6 1.07 perf-profile.self.cycles-pp.__check_heap_object
0.71 ± 4% +0.9 1.66 perf-profile.self.cycles-pp.__check_object_size
22.62 +1.6 24.22 perf-profile.self.cycles-pp.__fget_light
2.35 ± 6% +2.5 4.81 perf-profile.self.cycles-pp.copy_user_enhanced_fast_string
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.
Thanks,
Oliver Sang
View attachment "config-5.10.0-rc4-00108-gc191bb1b91be" of type "text/plain" (170113 bytes)
View attachment "job-script" of type "text/plain" (7981 bytes)
View attachment "job.yaml" of type "text/plain" (5377 bytes)
View attachment "reproduce" of type "text/plain" (336 bytes)
Powered by blists - more mailing lists