Greeting,
FYI, we noticed a 32.6% improvement of stress-ng.icache.ops_per_sec due to commit:
commit: 64fe24a3e05e5f3ac56fcd45afd2fd1d9cc8fcb6 ("mm/mprotect: try avoiding write faults for exclusive anonymous pages when changing protection")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
in testcase: stress-ng
on test machine: 48 threads 2 sockets Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz with 112G memory
with following parameters:
nr_threads: 100%
testtime: 60s
class: cpu-cache
test: icache
cpufreq_governor: performance
ucode: 0x42e
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
sudo bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run
sudo bin/lkp run generated-yaml-file
# if come across any failure that blocks the test,
# please remove ~/.lkp and /lkp dir to run from a clean state.
=========================================================================================
class/compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime/ucode:
cpu-cache/gcc-11/performance/x86_64-rhel-8.3/100%/debian-11.1-x86_64-20220510.cgz/lkp-ivb-2ep1/icache/stress-ng/60s/0x42e
commit:
50b0f797ca ("userfaultfd: selftests: infinite loop in faulting_process")
64fe24a3e0 ("mm/mprotect: try avoiding write faults for exclusive anonymous pages when changing protection")
50b0f797cab6cb7d 64fe24a3e05e5f3ac56fcd45afd
---------------- ---------------------------
%stddev %change %stddev
\ | \
489641 +32.6% 649466 ? 2% stress-ng.icache.ops
8160 +32.6% 10824 ? 2% stress-ng.icache.ops_per_sec
5.009e+08 -100.0% 8334 stress-ng.time.minor_page_faults
2011 +4.0% 2090 stress-ng.time.system_time
798.42 -10.0% 718.89 stress-ng.time.user_time
27.00 -2.5 24.53 mpstat.cpu.all.usr%
0.13 +15.4% 0.15 ? 3% turbostat.IPC
7006434 +16.7% 8177937 vmstat.memory.cache
2458795 +28.4% 3156476 numa-numastat.node1.local_node
2489886 +27.7% 3179646 numa-numastat.node1.numa_hit
1981079 ? 3% -37.3% 1242491 ? 8% numa-meminfo.node1.Active
1981013 ? 3% -37.3% 1242454 ? 8% numa-meminfo.node1.Active(anon)
4292310 +24.8% 5358029 ? 3% numa-meminfo.node1.FilePages
2200308 ? 3% +85.0% 4071306 ? 4% numa-meminfo.node1.Inactive
2200185 ? 3% +85.0% 4071246 ? 4% numa-meminfo.node1.Inactive(anon)
253031 ? 4% -10.2% 227134 ? 3% numa-meminfo.node1.Mapped
4742774 +22.3% 5798233 ? 3% numa-meminfo.node1.MemUsed
4138936 +27.5% 5276782 numa-meminfo.node1.Shmem
493166 ? 2% -36.6% 312894 ? 8% numa-vmstat.node1.nr_active_anon
1068609 +26.2% 1348947 ? 2% numa-vmstat.node1.nr_file_pages
547588 ? 3% +87.2% 1024991 ? 4% numa-vmstat.node1.nr_inactive_anon
1030266 +29.0% 1328635 numa-vmstat.node1.nr_shmem
493166 ? 2% -36.6% 312893 ? 8% numa-vmstat.node1.nr_zone_active_anon
547588 ? 3% +87.2% 1024990 ? 4% numa-vmstat.node1.nr_zone_inactive_anon
2489709 +27.7% 3179664 numa-vmstat.node1.numa_hit
2458618 +28.4% 3156493 numa-vmstat.node1.numa_local
1975430 ? 3% -36.7% 1251331 ? 8% meminfo.Active
1975329 ? 3% -36.7% 1251235 ? 8% meminfo.Active(anon)
6806979 +17.4% 7989179 meminfo.Cached
5037390 +23.5% 6220076 meminfo.Committed_AS
2414454 ? 3% +78.9% 4319684 ? 4% meminfo.Inactive
2414271 ? 3% +78.9% 4319496 ? 4% meminfo.Inactive(anon)
7982613 +14.8% 9167457 meminfo.Memused
4124105 +28.7% 5306304 ? 2% meminfo.Shmem
7993824 +14.8% 9178322 meminfo.max_used_kB
492415 ? 3% -36.4% 313242 ? 7% proc-vmstat.nr_active_anon
2667152 -1.1% 2637032 proc-vmstat.nr_dirty_background_threshold
5340827 -1.1% 5280512 proc-vmstat.nr_dirty_threshold
1698751 +17.7% 1999689 proc-vmstat.nr_file_pages
26851564 -1.1% 26549917 proc-vmstat.nr_free_pages
601985 ? 3% +79.7% 1081844 ? 4% proc-vmstat.nr_inactive_anon
1028032 +29.3% 1328969 proc-vmstat.nr_shmem
24818 +2.7% 25489 proc-vmstat.nr_slab_reclaimable
492415 ? 3% -36.4% 313242 ? 7% proc-vmstat.nr_zone_active_anon
601985 ? 3% +79.7% 1081844 ? 4% proc-vmstat.nr_zone_inactive_anon
2637652 +25.9% 3321072 proc-vmstat.numa_hit
2594059 +26.3% 3275093 proc-vmstat.numa_local
76406 ? 2% -37.0% 48149 ? 7% proc-vmstat.pgactivate
2637756 +25.8% 3317990 proc-vmstat.pgalloc_normal
5.012e+08 -99.9% 275690 proc-vmstat.pgfault
5.009e+08 -100.0% 11431 proc-vmstat.pgreuse
0.53 ? 7% -20.0% 0.43 ? 4% perf-stat.i.MPKI
1.286e+10 +12.7% 1.449e+10 ? 2% perf-stat.i.branch-instructions
1.21 +0.1 1.28 perf-stat.i.branch-miss-rate%
1.336e+08 +20.0% 1.604e+08 perf-stat.i.branch-misses
27889096 ? 7% -16.8% 23200601 ? 3% perf-stat.i.cache-references
2.06 -11.4% 1.83 ? 2% perf-stat.i.cpi
86811 ? 25% +115.9% 187413 ? 34% perf-stat.i.cycles-between-cache-misses
1.879e+10 +12.7% 2.117e+10 ? 2% perf-stat.i.dTLB-loads
0.50 ? 2% -0.1 0.39 ? 3% perf-stat.i.dTLB-store-miss-rate%
66268800 ? 2% -15.5% 55967378 ? 4% perf-stat.i.dTLB-store-misses
1.282e+10 +8.3% 1.389e+10 ? 2% perf-stat.i.dTLB-stores
88.36 +3.6 91.94 perf-stat.i.iTLB-load-miss-rate%
3.064e+08 +29.0% 3.952e+08 ? 2% perf-stat.i.iTLB-load-misses
39715709 ? 2% -16.0% 33366018 ? 3% perf-stat.i.iTLB-loads
6.475e+10 +13.1% 7.326e+10 ? 2% perf-stat.i.instructions
0.49 +12.1% 0.55 perf-stat.i.ipc
1705 ? 2% -31.8% 1162 ? 3% perf-stat.i.metric.K/sec
927.63 +11.4% 1033 ? 2% perf-stat.i.metric.M/sec
7918528 -100.0% 3555 perf-stat.i.minor-faults
37.53 ? 4% -2.9 34.66 ? 4% perf-stat.i.node-load-miss-rate%
39.79 -3.1 36.71 ? 5% perf-stat.i.node-store-miss-rate%
3225256 ? 4% +12.9% 3642536 ? 5% perf-stat.i.node-stores
7918528 -100.0% 3555 perf-stat.i.page-faults
0.43 ? 8% -26.2% 0.32 ? 4% perf-stat.overall.MPKI
1.03 +0.1 1.10 perf-stat.overall.branch-miss-rate%
18.42 ? 3% +5.2 23.60 ? 4% perf-stat.overall.cache-miss-rate%
2.10 -11.5% 1.86 ? 2% perf-stat.overall.cpi
0.51 ? 2% -0.1 0.40 ? 3% perf-stat.overall.dTLB-store-miss-rate%
88.53 +3.7 92.22 perf-stat.overall.iTLB-load-miss-rate%
211.20 -12.3% 185.29 perf-stat.overall.instructions-per-iTLB-miss
0.48 +13.1% 0.54 ? 2% perf-stat.overall.ipc
25.53 ? 6% -6.2 19.30 ? 6% perf-stat.overall.node-store-miss-rate%
1.267e+10 +12.7% 1.428e+10 ? 2% perf-stat.ps.branch-instructions
1.309e+08 +20.1% 1.572e+08 perf-stat.ps.branch-misses
27625119 ? 7% -16.5% 23070398 ? 3% perf-stat.ps.cache-references
1.852e+10 +12.7% 2.087e+10 ? 2% perf-stat.ps.dTLB-loads
65319076 ? 2% -15.6% 55136962 ? 4% perf-stat.ps.dTLB-store-misses
1.264e+10 +8.3% 1.369e+10 ? 2% perf-stat.ps.dTLB-stores
3.021e+08 +29.0% 3.896e+08 ? 2% perf-stat.ps.iTLB-load-misses
39157738 ? 2% -16.0% 32873937 ? 3% perf-stat.ps.iTLB-loads
6.38e+10 +13.1% 7.219e+10 ? 2% perf-stat.ps.instructions
7806710 -100.0% 3451 perf-stat.ps.minor-faults
3241411 ? 5% +13.7% 3684615 ? 5% perf-stat.ps.node-stores
7806710 -100.0% 3451 perf-stat.ps.page-faults
4.088e+12 +12.8% 4.614e+12 ? 2% 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://01.org/lkp