Greeting, FYI, we noticed a -9.9% regression of unixbench.score due to commit: commit: d519329f72a6f36bc4f2b85452640cfe583b4f81 ("sched/fair: Update util_est only on util_avg updates") https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master in testcase: unixbench on test machine: 8 threads Intel(R) Core(TM) i7 CPU 870 @ 2.93GHz with 6G memory with following parameters: runtime: 300s nr_task: 100% test: execl test-description: UnixBench is the original BYTE UNIX benchmark suite aims to test performance of Unix-like system. test-url: https://github.com/kdlucas/byte-unixbench 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/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase: gcc-7/x86_64-rhel-7.2/100%/debian-x86_64-2016-08-31.cgz/300s/nhm-white/execl/unixbench commit: a07630b8b2 ("sched/cpufreq/schedutil: Use util_est for OPP selection") d519329f72 ("sched/fair: Update util_est only on util_avg updates") a07630b8b2c16f82 d519329f72a6f36bc4f2b85452 ---------------- -------------------------- %stddev %change %stddev \ | \ 4626 -9.9% 4167 unixbench.score 3495362 ± 4% +70.4% 5957769 ± 2% unixbench.time.involuntary_context_switches 2.866e+08 -11.6% 2.534e+08 unixbench.time.minor_page_faults 666.75 -9.7% 602.25 unixbench.time.percent_of_cpu_this_job_got 1830 -9.7% 1653 unixbench.time.system_time 395.13 -5.2% 374.58 unixbench.time.user_time 8611715 -58.9% 3537314 ± 3% unixbench.time.voluntary_context_switches 6639375 -9.1% 6033775 unixbench.workload 26025 +3849.3% 1027825 interrupts.CAL:Function_call_interrupts 4856 ± 14% -27.4% 3523 ± 11% slabinfo.filp.active_objs 3534356 -8.8% 3223918 softirqs.RCU 77929 -11.2% 69172 vmstat.system.cs 19489 ± 2% +7.5% 20956 vmstat.system.in 9.05 ± 9% +11.0% 10.05 ± 8% boot-time.dhcp 131.63 ± 4% +8.6% 142.89 ± 7% boot-time.idle 9.07 ± 9% +11.0% 10.07 ± 8% boot-time.kernel_boot 76288 ± 3% -12.8% 66560 ± 3% meminfo.DirectMap4k 16606 -13.1% 14433 meminfo.Inactive 16515 -13.2% 14341 meminfo.Inactive(anon) 11.87 ± 5% +7.8 19.63 ± 4% mpstat.cpu.idle% 0.07 ± 35% -0.0 0.04 ± 17% mpstat.cpu.soft% 68.91 -6.1 62.82 mpstat.cpu.sys% 29291570 +325.4% 1.246e+08 cpuidle.C1.time 8629105 -36.1% 5513780 cpuidle.C1.usage 668733 ± 12% +11215.3% 75668902 ± 2% cpuidle.C1E.time 9763 ± 12% +16572.7% 1627882 ± 2% cpuidle.C1E.usage 1.834e+08 ± 9% +23.1% 2.258e+08 ± 11% cpuidle.C3.time 222674 ± 8% +133.4% 519690 ± 6% cpuidle.C3.usage 4129 -13.3% 3581 proc-vmstat.nr_inactive_anon 4129 -13.3% 3581 proc-vmstat.nr_zone_inactive_anon 2.333e+08 -12.2% 2.049e+08 proc-vmstat.numa_hit 2.333e+08 -12.2% 2.049e+08 proc-vmstat.numa_local 6625 -10.9% 5905 proc-vmstat.pgactivate 2.392e+08 -12.1% 2.102e+08 proc-vmstat.pgalloc_normal 2.936e+08 -12.6% 2.566e+08 proc-vmstat.pgfault 2.392e+08 -12.1% 2.102e+08 proc-vmstat.pgfree 2850 -15.3% 2413 turbostat.Avg_MHz 8629013 -36.1% 5513569 turbostat.C1 1.09 +3.5 4.61 turbostat.C1% 9751 ± 12% +16593.0% 1627864 ± 2% turbostat.C1E 0.03 ± 19% +2.8 2.80 turbostat.C1E% 222574 ± 8% +133.4% 519558 ± 6% turbostat.C3 6.84 ± 8% +1.5 8.34 ± 10% turbostat.C3% 2.82 ± 7% +250.3% 9.87 ± 2% turbostat.CPU%c1 6552773 ± 3% +23.8% 8111699 ± 2% turbostat.IRQ 2.02 ± 11% +28.3% 2.58 ± 9% turbostat.Pkg%pc3 7.635e+11 -12.5% 6.682e+11 perf-stat.branch-instructions 3.881e+10 -12.9% 3.381e+10 perf-stat.branch-misses 2.09 -0.3 1.77 ± 4% perf-stat.cache-miss-rate% 1.551e+09 -15.1% 1.316e+09 ± 4% perf-stat.cache-misses 26177920 -10.5% 23428188 perf-stat.context-switches 1.99 -2.8% 1.93 perf-stat.cpi 7.553e+12 -14.7% 6.446e+12 perf-stat.cpu-cycles 522523 ± 2% +628.3% 3805664 perf-stat.cpu-migrations 2.425e+10 ± 4% -14.3% 2.078e+10 perf-stat.dTLB-load-misses 1.487e+12 -11.3% 1.319e+12 perf-stat.dTLB-loads 1.156e+10 ± 3% -7.7% 1.066e+10 perf-stat.dTLB-store-misses 6.657e+11 -11.1% 5.915e+11 perf-stat.dTLB-stores 0.15 +0.0 0.15 perf-stat.iTLB-load-miss-rate% 5.807e+09 -11.0% 5.166e+09 perf-stat.iTLB-load-misses 3.799e+12 -12.1% 3.34e+12 perf-stat.iTLB-loads 3.803e+12 -12.2% 3.338e+12 perf-stat.instructions 654.99 -1.4% 646.07 perf-stat.instructions-per-iTLB-miss 0.50 +2.8% 0.52 perf-stat.ipc 2.754e+08 -11.6% 2.435e+08 perf-stat.minor-faults 1.198e+08 ± 7% +73.1% 2.074e+08 ± 4% perf-stat.node-stores 2.754e+08 -11.6% 2.435e+08 perf-stat.page-faults 572928 -3.4% 553258 perf-stat.path-length unixbench.score 4800 +-+------------------------------------------------------------------+ |+ + + | 4700 +-+ + + :+ +. :+ + + | | + + + +. : + + + + + + + .+++++ .+ +| 4600 +-+ +++ :+++ + ++: : :+ +++ ++.++++ + ++++ ++ | | + + + ++ ++ + | 4500 +-+ | | | 4400 +-+ | | | 4300 +-+ | O | 4200 +-O O O OOOO OO OOO OOOO OOOO O O | |O OO OOOOO O O OO O O O O O OO | 4100 +-+------------------------------------------------------------------+ unixbench.workload 9e+06 +-+---------------------------------------------------------------+ | : | 8.5e+06 +-+ : | | : | 8e+06 +-+ : | | :: | 7.5e+06 +-+ : : + | | +: : : + | 7e+06 +-+ + + :: : :: + + : + + + + + | |:+ + + : :: : : :: : :+ : : ::+ :+ .+ :+ ++ ++ + ++ ::++| 6.5e+06 +-O+ +++ ++++ +++ + ++ +.+ + ++ + + + + + + + +.+++ + | O O O O O O O | 6e+06 +O+OOO O OOOOOOOO OOOO OO OOOOOOOOO O O O OO | | O | 5.5e+06 +-+---------------------------------------------------------------+ [*] 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. Thanks, Xiaolong