Greeting,
FYI, we noticed a -10.1% regression of netperf.Throughput_Mbps due to commit:
commit: 6c362d9657293d700a8299acbeb2f1e24378f488 ("sched/fair: Fix wake_affine() for !NUMA_BALANCING") git://anongit.freedesktop.org/drm-intel topic/core-for-CI
in testcase: netperf on test machine: 16 threads Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz with 8G memory with following parameters:
ip: ipv4 runtime: 300s nr_threads: 200% cluster: cs-localhost test: TCP_SENDFILE cpufreq_governor: performance
test-description: Netperf is a benchmark that can be use to measure various aspect of networking performance. test-url: http://www.netperf.org/netperf/
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
testcase/path_params/tbox_group/run: netperf/ipv4-300s-200%-cs-localhost-TCP_SENDFILE-performance/lkp-bdw-de1
0e8e8706dfe2a17a 6c362d9657293d700a8299acbe ---------------- -------------------------- 11039 -10% 9929 netperf.Throughput_Mbps 36660 ± 20% 1257% 497519 netperf.time.involuntary_context_switches 7069 6997 netperf.time.minor_page_faults 2899 -6% 2734 netperf.time.system_time 983 -6% 925 netperf.time.percent_of_cpu_this_job_got 62.39 -13% 54.04 netperf.time.user_time 3137334 -63% 1171532 netperf.time.voluntary_context_switches 16915 46% 24638 vmstat.system.in 31608 69% 53301 vmstat.system.cs 64003 ± 6% 443% 347394 perf-stat.cpu-migrations 35981 ± 6% 205% 109811 ± 4% perf-stat.instructions-per-iTLB-miss 0.16 63% 0.25 perf-stat.branch-miss-rate% 2.538e+09 57% 3.977e+09 perf-stat.branch-misses 0.81 -4% 0.77 perf-stat.ipc 1.24 5% 1.30 perf-stat.cpi 370027 364969 perf-stat.page-faults 370025 364952 perf-stat.minor-faults 2.094e+12 -6% 1.97e+12 perf-stat.dTLB-stores 3.212e+12 -5% 3.04e+12 perf-stat.dTLB-loads 1.632e+12 -4% 1.566e+12 perf-stat.branch-instructions 8.997e+12 -4% 8.606e+12 perf-stat.instructions 1.874e+11 11% 2.079e+11 perf-stat.cache-references 1.874e+11 11% 2.079e+11 perf-stat.cache-misses 8.82 ± 6% -82% 1.59 ± 5% perf-stat.iTLB-load-miss-rate% 0.00 157% 0.01 perf-stat.dTLB-load-miss-rate% 1.104e+08 143% 2.682e+08 perf-stat.dTLB-load-misses 2.511e+08 ± 6% -69% 78564886 ± 5% perf-stat.iTLB-load-misses 2.594e+09 88% 4.87e+09 perf-stat.iTLB-loads 9571397 69% 16145347 perf-stat.context-switches
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, Xiaolong