Greeting, FYI, we noticed a +12% improvement of aim7.jobs-per-min due to commit: commit: 59c9081bc86ef0b273a41abf2c1f413301429a6d ("f2fs: allow write page cache when writting cp") https://git.kernel.org/cgit/linux/kernel/git/jaegeuk/f2fs.git dev-test in testcase: aim7 on test machine: 40 threads Intel(R) Xeon(R) CPU E5-2690 v2 @ 3.00GHz with 384G memory with following parameters: disk: 1BRD_48G fs: f2fs test: disk_rw load: 3000 cpufreq_governor: performance test-description: AIM7 is a traditional UNIX system level benchmark suite which is used to test and measure the performance of multiuser system. test-url: https://sourceforge.net/projects/aimbench/files/aim-suite7/ Details are as below: --------------------------------------------------------------------------------------------------> To reproduce: git clone https://github.com/01org/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: aim7/1BRD_48G-f2fs-disk_rw-3000-performance/lkp-ivb-ep01 22588f8773f68744 59c9081bc86ef0b273a41abf2c ---------------- -------------------------- 108504 12% 121167 aim7.jobs-per-min 827142 8% 889247 aim7.time.voluntary_context_switches 166 -10% 149 aim7.time.elapsed_time 166 -10% 149 aim7.time.elapsed_time.max 5892 -12% 5176 aim7.time.system_time 575818 ± 4% -14% 495473 ± 3% aim7.time.involuntary_context_switches 426869 7% 455589 interrupts.CAL:Function_call_interrupts 7118 12% 7943 vmstat.io.bo 46543 47718 vmstat.system.in 1054904 ± 89% -1e+06 53098 ±173% latency_stats.avg.call_rwsem_down_write_failed.do_unlinkat.SyS_unlink.do_syscall_64.return_from_SYSCALL_64 4691 ±115% -5e+03 78 ± 6% latency_stats.max.kthread_park.watchdog_park_threads.watchdog_enable_all_cpus.proc_watchdog_update.proc_watchdog_common.proc_nmi_watchdog.proc_sys_call_handler.proc_sys_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath 5978 ± 90% -5e+03 1239 ± 13% latency_stats.sum.kthread_park.watchdog_park_threads.watchdog_enable_all_cpus.proc_watchdog_update.proc_watchdog_common.proc_nmi_watchdog.proc_sys_call_handler.proc_sys_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath 89.24 87.60 turbostat.%Busy 159 156 turbostat.PkgWatt 131 128 turbostat.CorWatt 2690 2624 turbostat.Avg_MHz 92438778 ± 28% 69% 1.564e+08 ± 19% perf-stat.iTLB-loads 0.19 14% 0.21 ± 5% perf-stat.branch-miss-rate% 0.27 6% 0.29 perf-stat.ipc 6.082e+09 -3% 5.894e+09 perf-stat.node-stores 4.699e+09 -5% 4.485e+09 ± 3% perf-stat.node-store-misses 1.351e+12 -6% 1.266e+12 perf-stat.dTLB-loads 4.872e+12 -7% 4.516e+12 perf-stat.instructions 2687098 ± 3% -8% 2473467 ± 3% perf-stat.context-switches 1.068e+12 -8% 9.801e+11 perf-stat.branch-instructions 0.14 ± 6% -10% 0.13 ± 5% perf-stat.dTLB-store-miss-rate% 6.616e+08 ± 6% -10% 5.979e+08 ± 5% perf-stat.dTLB-store-misses 1.785e+13 -13% 1.56e+13 perf-stat.cpu-cycles 2.855e+09 ± 13% -19% 2.303e+09 ± 14% perf-stat.node-loads 2.122e+09 ± 16% -22% 1.656e+09 ± 17% perf-stat.node-load-misses 38.81 ± 22% -29% 27.70 ± 22% perf-stat.iTLB-load-miss-rate% aim7.jobs-per-min 140000 ++-----------------------------------------------------------------+ | | 120000 O+O O O O O O O O O O O O O O O O O O O O O O O O O O | | *.*. .*.*..*.*.*.*.*.*.* O *.*..*.*.*.*.*.* 100000 ++: * : : | | : : : | 80000 ++ : *.*.. .*.* *.*. .*.* | |: : : * : : * | 60000 ++ : : : : | |: : : : : | 40000 ++ : : : : | |: : : : : | 20000 ++ :: :: | | : : | 0 *+-------------------------*------------*--------------------------+ [*] 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