Greeting, We noticed a -10% regression of aim7.jobs-per-min due to commit: commit: bb95d9ab2a9d4afd03b59a603cccb2c601f68b78 ("f2fs: drop exist_data for inline_data when truncated to 0") https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master 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: 4BRD_12G md: RAID1 fs: f2fs test: creat-clo load: 1500 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 git://git.kernel.org/pub/scm/linux/kernel/git/wfg/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: aim7/4BRD_12G-RAID1-f2fs-creat-clo-1500-performance/lkp-ivb-ep01 363fa4e078cbdc97 bb95d9ab2a9d4afd03b59a603c ---------------- -------------------------- %stddev change %stddev \ | \ 69986 ± 3% -10% 63066 aim7.jobs-per-min 226578 ± 8% 87% 423593 ± 4% aim7.time.involuntary_context_switches 4502 ± 3% 15% 5157 aim7.time.system_time 129 ± 3% 11% 142 aim7.time.elapsed_time 129 ± 3% 11% 142 aim7.time.elapsed_time.max 644595 ± 17% -28% 462051 ± 3% aim7.time.voluntary_context_switches 12748 ± 8% -11% 11310 ± 3% vmstat.system.cs 4192 ± 3% -10% 3772 iostat.md0.wkB/s 2622 5% 2747 turbostat.Avg_MHz 87.84 3% 90.68 turbostat.%Busy 128 3% 132 turbostat.CorWatt 155 159 turbostat.PkgWatt 316869 ± 24% 7e+05 972706 ± 23% latency_stats.avg.call_rwsem_down_write_failed.path_openat.do_filp_open.do_sys_open.SyS_creat.do_syscall_64.return_from_SYSCALL_64 3618 ± 83% 1e+04 18282 ± 43% latency_stats.avg.perf_event_alloc.SYSC_perf_event_open.SyS_perf_event_open.entry_SYSCALL_64_fastpath 3618 ± 83% 1e+04 18282 ± 43% latency_stats.max.perf_event_alloc.SYSC_perf_event_open.SyS_perf_event_open.entry_SYSCALL_64_fastpath 3618 ± 83% 1e+04 18282 ± 43% latency_stats.sum.perf_event_alloc.SYSC_perf_event_open.SyS_perf_event_open.entry_SYSCALL_64_fastpath 0.32 ± 23% 40% 0.45 ± 9% perf-stat.dTLB-load-miss-rate% 0.16 38% 0.22 perf-stat.branch-miss-rate% 1.854e+09 ± 3% 23% 2.282e+09 perf-stat.branch-misses 1.362e+13 ± 4% 16% 1.574e+13 perf-stat.cpu-cycles 1.176e+09 ± 3% 14% 1.343e+09 perf-stat.node-store-misses 1.447e+10 ± 7% 14% 1.645e+10 ± 4% perf-stat.cache-references 1.579e+09 ± 4% 9% 1.72e+09 perf-stat.node-stores 3.156e+11 3% 3.261e+11 perf-stat.dTLB-stores 42.68 43.85 perf-stat.node-store-miss-rate% 45.56 45.00 perf-stat.node-load-miss-rate% 25.79 -3% 24.97 perf-stat.cache-miss-rate% 1.189e+12 ± 4% -11% 1.058e+12 perf-stat.branch-instructions 5.785e+12 ± 4% -16% 4.837e+12 perf-stat.instructions 1.552e+12 ± 4% -17% 1.286e+12 perf-stat.dTLB-loads 0.42 -28% 0.31 perf-stat.ipc perf-stat.node-store-miss-rate_ 44.6 ++-------------------------------------------------------------------+ 44.4 ++ O O O | | O O O O O O O | 44.2 O+ O O O O O O | 44 ++ O O O O O O O O | 43.8 ++ O O O O 43.6 ++ | | | 43.4 ++ | 43.2 ++ | 43 ++ | 42.8 ++ | *. .*..*. .*..*. .*..*.*..*.*.*..*.*.. .*.. .*..*.* | 42.6 ++*..* *..* *..* * * | 42.4 ++-------------------------------------------------------------------+ perf-stat.ipc 0.44 ++-------------*----*-*----------------------------------------------+ |.*..*.*.. .*. *. *. .*. .*.*.*..*.*..*.*..*.*.. .* | 0.42 *+ * *. *. * | 0.4 ++ | | | 0.38 ++ | 0.36 ++ | | | 0.34 ++ | 0.32 ++ | | O O O O 0.3 O+ O O O O O O O O O | 0.28 ++O O O O O O O O O O O O | | O O O | 0.26 ++-------------------------------------------------------------------+ turbostat.Avg_MHz 2800 ++-------------------------------------------------------------------+ | O O O O O O | | O O O O O O O O O O 2750 ++O O O O O O O O O O | O O O | | | 2700 ++ | | | 2650 ++ | * * * | |+ .. : *.* .*.. .*.. .*..*.*.. + | 2600 ++* : .. + .*..* * *.*.*.. .*..* * | | * + .* * | | *.*. | 2550 ++-------------------------------------------------------------------+ turbostat._Busy 91.5 ++-------------------------------------------------------------------+ 91 ++ O O O | | O O O O O O O O O O O O O O O O O O O 90.5 O+ O O O O O O | 90 ++ | | | 89.5 ++ | 89 ++ | 88.5 ++ | | | 88 *+*..*. *.*.. .*.. .*.*.. .* | 87.5 ++ *..*.* .*. .. * *.*.*.. .*.. .*. * | | + *. * * * | 87 ++ + + | 86.5 ++-------------*-----------------------------------------------------+ aim7.time.involuntary_context_switches 600000 ++-----------------------------------------------------------------+ | O O | 550000 ++ O O O O | 500000 ++ O O | | O O O O O O O | 450000 O+O O O O O O O O O O 400000 ++ O O | | O | 350000 ++ | 300000 ++ | | | 250000 *+*..*.*..*. .*.. *.*..*.*..*. *.* .*. | 200000 ++ *. .* + *. .. + .*.*. *.* | | *. * * *. | 150000 ++-----------------------------------------------------------------+ [*] 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