Greeting, We noticed a -32% regression of aim7.jobs-per-min due to commit: commit: ae75f0ca76d8ad07f9e902f0f9e46ec042acf47c ("f2fs: introduce free nid bitmap") 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: RAID0 fs: f2fs test: disk_cp 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 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-RAID0-f2fs-disk_cp-3000-performance/lkp-ivb-ep01 8559922435601e8b ae75f0ca76d8ad07f9e902f0f9 ---------------- -------------------------- %stddev change %stddev \ | \ 117068 -32% 79734 aim7.jobs-per-min 154 47% 226 aim7.time.elapsed_time 154 47% 226 aim7.time.elapsed_time.max 22.64 38% 31.13 ± 8% aim7.time.user_time 5387 14% 6147 aim7.time.system_time 399647 11% 442359 interrupts.CAL:Function_call_interrupts 1080 ± 10% -17% 898 ± 5% iostat.md0.w/s 21463 -31% 14725 iostat.md0.wkB/s 7688 -32% 5263 vmstat.io.bo 47120 -4% 45046 vmstat.system.in 18949 ± 5% -25% 14119 vmstat.system.cs 38.21 -8% 35.11 turbostat.RAMWatt 87.77 -21% 69.39 turbostat.%Busy 158 -32% 107 turbostat.PkgWatt 130.42 -38% 80.32 ± 3% turbostat.CorWatt 2642 -41% 1570 ± 3% turbostat.Avg_MHz 0.19 28% 0.25 ± 3% perf-stat.branch-miss-rate% 1.899e+09 22% 2.317e+09 perf-stat.branch-misses 752803 19% 897581 perf-stat.page-faults 752798 19% 897578 perf-stat.minor-faults 2951975 ± 4% 9% 3215525 ± 4% perf-stat.context-switches 0.27 8% 0.29 perf-stat.ipc 3.115e+09 -4% 2.99e+09 perf-stat.node-store-misses 4.396e+12 -4% 4.199e+12 perf-stat.instructions 31.26 -5% 29.61 perf-stat.cache-miss-rate% 9.801e+11 -5% 9.323e+11 perf-stat.branch-instructions 4.499e+09 -6% 4.215e+09 perf-stat.node-stores 1.191e+12 -7% 1.107e+12 perf-stat.dTLB-loads 822029 ± 5% -12% 725505 ± 4% perf-stat.cpu-migrations 1.634e+13 -12% 1.439e+13 perf-stat.cpu-cycles 495167 ±121% -4e+05 124905 ±150% latency_stats.avg.do_write_page.[f2fs].write_data_page.[f2fs].f2fs_convert_inline_page.[f2fs].f2fs_convert_inline_inode.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.do_syscall_64.return_from_SYSCALL_64 1119 ±164% 5e+03 5832 ±113% latency_stats.avg.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath 28663 ± 81% -1e+04 14258 ± 27% latency_stats.avg.call_rwsem_down_read_failed.build_free_nids.[f2fs].alloc_nid.[f2fs].f2fs_new_inode.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open.SyS_creat.entry_SYSCALL_64_fastpath 21190 ±116% -2e+04 3463 ± 8% latency_stats.avg.call_rwsem_down_read_failed.build_free_nids.[f2fs].f2fs_balance_fs_bg.[f2fs].f2fs_balance_fs.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open.SyS_creat.entry_SYSCALL_64_fastpath 37304 ±109% 4e+05 418035 ± 23% latency_stats.max.call_rwsem_down_read_failed.build_free_nids.[f2fs].f2fs_balance_fs_bg.[f2fs].f2fs_balance_fs.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open.SyS_creat.entry_SYSCALL_64_fastpath 0 2e+05 186976 ± 18% latency_stats.max.call_rwsem_down_read_failed.build_free_nids.[f2fs].f2fs_balance_fs_bg.[f2fs].f2fs_balance_fs.[f2fs].f2fs_unlink.[f2fs].vfs_unlink.do_unlinkat.SyS_unlink.entry_SYSCALL_64_fastpath 495167 ±121% -4e+05 124905 ±150% latency_stats.max.do_write_page.[f2fs].write_data_page.[f2fs].f2fs_convert_inline_page.[f2fs].f2fs_convert_inline_inode.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.do_syscall_64.return_from_SYSCALL_64 3255 ±167% 2e+04 19600 ±137% latency_stats.max.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath 7.859e+09 ± 12% 2e+10 3.16e+10 latency_stats.sum.call_rwsem_down_write_failed.do_unlinkat.SyS_unlink.entry_SYSCALL_64_fastpath 102988 ± 98% 7e+06 7189879 ± 7% latency_stats.sum.call_rwsem_down_read_failed.build_free_nids.[f2fs].f2fs_balance_fs_bg.[f2fs].f2fs_balance_fs.[f2fs].f2fs_create.[f2fs].path_openat.do_filp_open.do_sys_open.SyS_creat.entry_SYSCALL_64_fastpath 0 2e+06 2172771 ± 13% latency_stats.sum.call_rwsem_down_read_failed.build_free_nids.[f2fs].f2fs_balance_fs_bg.[f2fs].f2fs_balance_fs.[f2fs].f2fs_unlink.[f2fs].vfs_unlink.do_unlinkat.SyS_unlink.entry_SYSCALL_64_fastpath 495167 ±121% -4e+05 124905 ±150% latency_stats.sum.do_write_page.[f2fs].write_data_page.[f2fs].f2fs_convert_inline_page.[f2fs].f2fs_convert_inline_inode.[f2fs].f2fs_preallocate_blocks.[f2fs].f2fs_file_write_iter.[f2fs].__vfs_write.vfs_write.SyS_write.do_syscall_64.return_from_SYSCALL_64 3312 ±167% 2e+04 19776 ±135% latency_stats.sum.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath 41381473 ± 71% -2e+07 19676236 ± 29% latency_stats.sum.invalidate_blocks.[f2fs].truncate_data_blocks_range.[f2fs].truncate_blocks.[f2fs].f2fs_truncate.[f2fs].f2fs_evict_inode.[f2fs].evict.iput.dentry_unlink_inode.__dentry_kill.dput.__fput.____fput aim7.jobs-per-min 140000 ++-----------------------------------------------------------------+ |.* *. *. .* .**. .* .* | 120000 *+ *.**.* * **.**.**.** *.**.* **.** * *.* **.**.**.**.**.* | : : | 100000 ++ : : | | : : | 80000 O+O O O OO OO OO OO OO OO OO : : | | : : | 60000 ++ :: | | :: | 40000 ++ :: | | :: | 20000 ++ : | | : | 0 ++-O--O-O-----------------------------------------*----------------+ turbostat.Avg_MHz 3000 ++-------------------------------------------------------------------+ | .**. *. *. *. .** .* | 2500 *+** **.*.* **.* *.**.**.**.*.**.**.**.* * : *.**.*.** *.**.* | : : | | : : | 2000 ++ : : | O O O O O O : : | 1500 ++ O O OO O OO O O OO O : : | | : : | 1000 ++ :: | | :: | | : | 500 ++ : | | : | 0 ++-O--O-O-------------------------------------------*----------------+ turbostat._Busy 90 *+**-**-*-**-*-**-**-*-**-**-*-**-*-**-**-*-**-**-*-*--**-*-**-*-**-**-* | : : | 80 ++ : : | 70 O+O O OO O OO OO O OO OO O OO : : | | : : | 60 ++ : : | 50 ++ : : | | :: | 40 ++ :: | 30 ++ :: | | :: | 20 ++ : | 10 ++ : | | : | 0 ++-O--O-O--------------------------------------------*-----------------+ turbostat.PkgWatt 180 ++--------------------------------------------------------------------+ | .* .* | 160 *+**.**.*.**.**.*.**.**.**.*.**.**.*.**.**.*.**.** : *.**.**.* *.**.* 140 ++ : : | | : : | 120 O+O O O O O O O : : | 100 ++ O OO O OO O O OO : : | | : : | 80 ++ : : | 60 ++ :: | | :: | 40 ++ : | 20 ++ : | | : | 0 ++-O--O-O--------------------------------------------*----------------+ turbostat.CorWatt 140 ++--------------------------------------------------------------------+ *.**.**.*.**.**.*.**.**.**.*.**.**.*.**.**.*.**.**.* *.**.**.*.**.**.* 120 ++ : : | | : : | 100 ++ : : | | O : : | 80 O+O OO OO O OO OO OO O OO O : : | | : : | 60 ++ :: | | :: | 40 ++ :: | | : | 20 ++ : | | : | 0 ++-O--O-O--------------------------------------------*----------------+ [*] 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