Greeting, FYI, we noticed a 18.6% improvement of vm-scalability.median due to commit: commit: dd12385915f4f83b738467e599b053c33dffbd48 ("[PATCH] mm: fix COW faults after mlock()") url: https://github.com/0day-ci/linux/commits/Yury-Norov/mm-fix-COW-faults-after-mlock/20180925-174527 base: https://github.com/thesofproject/linux master in testcase: vm-scalability on test machine: 80 threads Skylake with 64G memory with following parameters: runtime: 300s size: 1T test: msync cpufreq_governor: performance test-description: The motivation behind this suite is to exercise functions and regions of the mm/ of the Linux kernel which are of interest to us. test-url: https://git.kernel.org/cgit/linux/kernel/git/wfg/vm-scalability.git/ 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/cpufreq_governor/kconfig/rootfs/runtime/size/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.2/debian-x86_64-2018-04-03.cgz/300s/1T/lkp-ivb-d02/msync/vm-scalability/0x20 commit: dd52cb8790 (" platform-drivers-x86 for v4.17-4") dd12385915 ("mm: fix COW faults after mlock()") dd52cb879063ca54 dd12385915f4f83b738467e599 ---------------- -------------------------- fail:runs %reproduction fail:runs | | | 1:4 -25% :4 dmesg.WARNING:at#for_ip_error_entry/0x 0:4 18% 1:4 perf-profile.calltrace.cycles-pp.error_entry 2:4 -18% 2:4 perf-profile.children.cycles-pp.error_entry 0:4 0% 0:4 perf-profile.children.cycles-pp.schedule_timeout 2:4 -17% 1:4 perf-profile.self.cycles-pp.error_entry %stddev %change %stddev \ | \ 277824 A+- 4% +18.6% 329586 A+- 6% vm-scalability.median 0.46 A+- 6% -46.9% 0.24 A+- 27% vm-scalability.median_stddev 2817 A+- 26% +5.7e+06% 1.608e+08 A+- 6% vm-scalability.time.file_system_inputs 6509 A+- 28% -73.7% 1714 A+- 3% vm-scalability.time.major_page_faults 142.50 A+- 5% -40.0% 85.50 A+- 4% vm-scalability.time.percent_of_cpu_this_job_got 281.87 A+- 6% -35.2% 182.76 A+- 5% vm-scalability.time.system_time 152.54 A+- 6% -46.3% 81.85 A+- 5% vm-scalability.time.user_time 448928 A+- 6% -31.1% 309105 A+- 8% vm-scalability.time.voluntary_context_switches 2.093e+08 A+- 6% -44.9% 1.154e+08 A+- 5% vm-scalability.workload 18807972 A+- 5% -27.3% 13664118 A+- 6% interrupts.CAL:Function_call_interrupts 88890 A+- 2% -10.4% 79618 A+- 2% softirqs.RCU 100860 A+- 2% +31.5% 132623 A+- 2% softirqs.SCHED 9.00 +2.8e+06% 256051 A+- 5% vmstat.io.bi 6.00 -50.0% 3.00 vmstat.memory.buff 67870 A+- 4% -25.0% 50893 A+- 5% vmstat.system.in 14.04 A+- 8% +2.4 16.47 A+- 7% mpstat.cpu.idle% 29.68 A+- 7% +13.5 43.20 A+- 2% mpstat.cpu.iowait% 42.68 A+- 5% -10.1 32.62 A+- 5% mpstat.cpu.sys% 13.58 A+- 4% -5.9 7.70 A+- 4% mpstat.cpu.usr% 294613 +88.5% 555281 slabinfo.buffer_head.active_objs 7559 +88.5% 14248 slabinfo.buffer_head.active_slabs 294837 +88.5% 555706 slabinfo.buffer_head.num_objs 7559 +88.5% 14248 slabinfo.buffer_head.num_slabs 817.75 A+- 6% -9.7% 738.25 A+- 7% slabinfo.proc_inode_cache.num_objs 454.00 A+- 8% +24.4% 565.00 A+- 18% slabinfo.skbuff_head_cache.active_objs 13651918 A+- 29% +56.9% 21419299 A+- 8% cpuidle.C1.time 1.931e+08 A+- 6% +51.1% 2.919e+08 A+- 2% cpuidle.C1E.time 508755 A+- 6% +33.4% 678785 A+- 2% cpuidle.C1E.usage 16723882 A+- 4% +73.1% 28943328 cpuidle.C3.time 21421 A+- 3% +78.8% 38298 cpuidle.C3.usage 3.002e+08 A+- 5% +29.2% 3.879e+08 A+- 4% cpuidle.C6.time 312454 A+- 4% +28.8% 402526 A+- 4% cpuidle.C6.usage 107432 A+- 20% +61.4% 173441 A+- 12% cpuidle.POLL.time 18727 A+- 8% -25.9% 13876 A+- 6% cpuidle.POLL.usage 1073224 A+- 2% +17.7% 1263701 meminfo.Active(file) 308786 A+- 3% +18.0% 364452 A+- 2% meminfo.Dirty 1342759 A+- 2% -20.9% 1061562 A+- 3% meminfo.Inactive 172701 A+- 16% -39.3% 104801 A+- 3% meminfo.Inactive(anon) 1170057 -18.2% 956761 A+- 3% meminfo.Inactive(file) 13762 -20.3% 10968 A+- 2% meminfo.PageTables 85607 +29.8% 111114 meminfo.SReclaimable 107046 +23.8% 132534 meminfo.Slab 37561 A+- 4% +23.5% 46379 A+- 4% meminfo.Writeback 84144 A+- 7% -26.0% 62266 A+- 5% sched_debug.cfs_rq:/.exec_clock.avg 85958 A+- 6% -24.6% 64822 A+- 5% sched_debug.cfs_rq:/.exec_clock.max 82818 A+- 7% -26.6% 60788 A+- 5% sched_debug.cfs_rq:/.exec_clock.min 209438 A+- 7% -41.9% 121723 A+- 4% sched_debug.cfs_rq:/.min_vruntime.avg 218172 A+- 8% -40.0% 130873 A+- 3% sched_debug.cfs_rq:/.min_vruntime.max 199461 A+- 6% -43.1% 113416 A+- 7% sched_debug.cfs_rq:/.min_vruntime.min 1399 A+- 2% -16.2% 1172 A+- 11% sched_debug.cpu.curr->pid.stddev 15.87 A+- 25% +104.2% 32.42 A+- 35% sched_debug.cpu.nr_uninterruptible.max -14.25 +91.2% -27.25 sched_debug.cpu.nr_uninterruptible.min 11.61 A+- 28% +98.0% 22.99 A+- 28% sched_debug.cpu.nr_uninterruptible.stddev 127503 A+- 14% +26.0% 160648 A+- 4% sched_debug.cpu.sched_goidle.min 0.00 A+- 94% -91.3% 0.00 A+-158% sched_debug.rt_rq:/.rt_time.stddev 1879 A+- 4% -27.7% 1358 A+- 5% turbostat.Avg_MHz 57.42 A+- 4% -15.7 41.70 A+- 4% turbostat.Busy% 1.11 A+- 30% +0.6 1.71 A+- 7% turbostat.C1% 508750 A+- 6% +33.4% 678785 A+- 2% turbostat.C1E 15.71 A+- 7% +7.6 23.32 A+- 2% turbostat.C1E% 21419 A+- 3% +78.8% 38298 turbostat.C3 1.36 A+- 5% +1.0 2.31 turbostat.C3% 312453 A+- 4% +28.8% 402527 A+- 4% turbostat.C6 24.42 A+- 6% +6.6 31.00 A+- 5% turbostat.C6% 30.15 A+- 8% +39.3% 42.01 A+- 3% turbostat.CPU%c1 1.09 A+- 5% +60.0% 1.75 A+- 6% turbostat.CPU%c3 11.33 A+- 9% +28.3% 14.54 A+- 6% turbostat.CPU%c6 12.77 A+- 3% -18.3% 10.43 A+- 3% turbostat.CorWatt 39798103 A+- 5% -25.5% 29652069 A+- 6% turbostat.IRQ 30.09 -8.0% 27.69 turbostat.PkgWatt 4.902e+11 A+- 5% -41.5% 2.869e+11 A+- 28% perf-stat.branch-instructions 3.636e+09 A+- 8% -32.3% 2.461e+09 A+- 28% perf-stat.branch-misses 66.28 -7.3 58.95 perf-stat.cache-miss-rate% 9.761e+09 A+- 7% -26.2% 7.2e+09 A+- 28% perf-stat.cache-misses 1.11 +5.9% 1.17 perf-stat.cpi 2.298e+12 A+- 6% -35.6% 1.48e+12 A+- 27% perf-stat.cpu-cycles 19198 A+- 5% +14.7% 22028 A+- 5% perf-stat.cpu-migrations 3.858e+09 A+- 15% -42.9% 2.205e+09 A+- 16% perf-stat.dTLB-load-misses 5.213e+11 A+- 5% -36.1% 3.332e+11 A+- 28% perf-stat.dTLB-loads 0.12 A+- 9% -0.0 0.09 A+- 18% perf-stat.dTLB-store-miss-rate% 3.446e+08 A+- 10% -52.5% 1.638e+08 A+- 17% perf-stat.dTLB-store-misses 2.839e+11 A+- 5% -35.2% 1.839e+11 A+- 28% perf-stat.dTLB-stores 74.39 +7.5 81.93 perf-stat.iTLB-load-miss-rate% 3.687e+08 A+- 8% -45.1% 2.024e+08 A+- 27% perf-stat.iTLB-load-misses 1.272e+08 A+- 10% -65.7% 43583231 A+- 21% perf-stat.iTLB-loads 2.077e+12 A+- 5% -39.0% 1.266e+12 A+- 28% perf-stat.instructions 5646 A+- 3% +10.3% 6228 A+- 4% perf-stat.instructions-per-iTLB-miss 0.90 -5.5% 0.85 perf-stat.ipc 47044927 A+- 6% -44.5% 26119814 A+- 5% perf-stat.minor-faults 47046188 A+- 6% -44.5% 26120941 A+- 5% perf-stat.page-faults 22566 A+- 39% -78.5% 4857 A+- 4% proc-vmstat.allocstall_movable 120471 A+- 14% -86.8% 15866 A+- 5% proc-vmstat.allocstall_normal 7661 A+- 28% -78.7% 1631 A+- 11% proc-vmstat.compact_fail 7682 A+- 28% -78.4% 1659 A+- 11% proc-vmstat.compact_stall 2752 A+- 5% -37.9% 1710 A+- 4% proc-vmstat.kswapd_low_wmark_hit_quickly 267103 A+- 2% +17.7% 314329 A+- 2% proc-vmstat.nr_active_file 78076 A+- 3% +17.3% 91559 proc-vmstat.nr_dirty 7811 A+- 6% -21.7% 6118 A+- 28% proc-vmstat.nr_free_cma 43289 A+- 16% -39.4% 26248 A+- 3% proc-vmstat.nr_inactive_anon 293783 -17.7% 241780 A+- 4% proc-vmstat.nr_inactive_file 3445 -20.5% 2740 proc-vmstat.nr_page_table_pages 21416 +29.8% 27796 proc-vmstat.nr_slab_reclaimable 9153 A+- 2% +26.1% 11542 A+- 2% proc-vmstat.nr_writeback 267072 A+- 2% +17.7% 314320 A+- 2% proc-vmstat.nr_zone_active_file 43289 A+- 16% -39.4% 26251 A+- 3% proc-vmstat.nr_zone_inactive_anon 293694 -17.7% 241712 A+- 4% proc-vmstat.nr_zone_inactive_file 87180 A+- 3% +18.2% 103074 proc-vmstat.nr_zone_write_pending 85522443 A+- 6% -43.5% 48347251 A+- 5% proc-vmstat.numa_hit 85522443 A+- 6% -43.5% 48347251 A+- 5% proc-vmstat.numa_local 2958 A+- 5% -35.3% 1913 A+- 3% proc-vmstat.pageoutrun 51287873 A+- 6% -43.2% 29135174 A+- 5% proc-vmstat.pgactivate 28504604 A+- 8% -18.2% 23322955 A+- 5% proc-vmstat.pgalloc_dma32 57168735 A+- 6% -56.0% 25151136 A+- 8% proc-vmstat.pgalloc_normal 41153300 A+- 7% -46.5% 22026100 A+- 6% proc-vmstat.pgdeactivate 85685675 A+- 6% -43.4% 48485032 A+- 5% proc-vmstat.pgfree 6509 A+- 28% -73.7% 1714 A+- 3% proc-vmstat.pgmajfault 2976 +2.7e+06% 80419921 A+- 6% proc-vmstat.pgpgin 41153300 A+- 7% -46.5% 22026099 A+- 6% proc-vmstat.pgrefill 477249 A+- 6% +72.3% 822527 A+- 11% proc-vmstat.pgrotated 43943485 A+- 5% -52.5% 20878181 A+- 5% proc-vmstat.pgscan_direct 84942346 A+- 6% -41.7% 49530834 A+- 5% proc-vmstat.pgscan_kswapd 8102736 A+- 7% -85.2% 1196177 A+- 5% proc-vmstat.pgsteal_direct 47558626 A+- 6% -39.2% 28921712 A+- 6% proc-vmstat.pgsteal_kswapd 14421 -13.1% 12538 A+- 3% proc-vmstat.slabs_scanned 2020457 A+- 15% -66.2% 682455 A+- 8% proc-vmstat.workingset_activate 17434531 A+- 8% -44.0% 9754788 A+- 5% proc-vmstat.workingset_refault 19.94 A+- 25% -17.5 2.42 A+-110% perf-profile.calltrace.cycles-pp.do_access 12.24 A+- 26% -10.7 1.50 A+-109% perf-profile.calltrace.cycles-pp.page_fault.do_access 12.22 A+- 26% -10.7 1.50 A+-109% perf-profile.calltrace.cycles-pp.do_page_fault.page_fault.do_access 12.21 A+- 26% -10.7 1.50 A+-109% perf-profile.calltrace.cycles-pp.__do_page_fault.do_page_fault.page_fault.do_access 11.34 A+- 26% -10.0 1.37 A+-109% perf-profile.calltrace.cycles-pp.handle_mm_fault.__do_page_fault.do_page_fault.page_fault.do_access 16.17 A+- 23% -8.6 7.58 A+- 25% perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 9.39 A+- 27% -8.1 1.29 A+-110% perf-profile.calltrace.cycles-pp.do_rw_once 7.59 A+- 23% -4.5 3.05 A+- 30% perf-profile.calltrace.cycles-pp.do_page_mkwrite.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 7.53 A+- 23% -4.5 3.00 A+- 30% perf-profile.calltrace.cycles-pp.__xfs_filemap_fault.do_page_mkwrite.__handle_mm_fault.handle_mm_fault.__do_page_fault 6.69 A+- 24% -3.3 3.44 A+- 23% perf-profile.calltrace.cycles-pp.__do_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 6.55 A+- 24% -3.2 3.38 A+- 23% perf-profile.calltrace.cycles-pp.__xfs_filemap_fault.__do_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault 14.30 A+- 5% -3.0 11.27 A+- 7% perf-profile.calltrace.cycles-pp.__do_page_cache_readahead.ondemand_readahead.filemap_fault.__xfs_filemap_fault.__do_fault 14.31 A+- 5% -3.0 11.29 A+- 8% perf-profile.calltrace.cycles-pp.ondemand_readahead.filemap_fault.__xfs_filemap_fault.__do_fault.__handle_mm_fault 15.40 A+- 3% -2.7 12.73 A+- 6% perf-profile.calltrace.cycles-pp.filemap_fault.__xfs_filemap_fault.__do_fault.__handle_mm_fault.handle_mm_fault 2.67 A+- 41% -2.1 0.56 A+-116% perf-profile.calltrace.cycles-pp.__alloc_pages_slowpath.__alloc_pages_nodemask.__do_page_cache_readahead.ondemand_readahead.filemap_fault 4.15 A+- 27% -2.1 2.10 A+- 12% perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.__do_page_cache_readahead.ondemand_readahead.filemap_fault.__xfs_filemap_fault 2.46 A+- 19% -1.6 0.89 A+- 68% perf-profile.calltrace.cycles-pp.shrink_page_list.shrink_inactive_list.shrink_node_memcg.shrink_node.do_try_to_free_pages 3.60 A+- 24% -1.5 2.06 A+- 40% perf-profile.calltrace.cycles-pp.shrink_inactive_list.shrink_node_memcg.shrink_node.do_try_to_free_pages.try_to_free_pages 9.38 A+- 3% -1.4 7.95 A+- 14% perf-profile.calltrace.cycles-pp.mpage_readpages.__do_page_cache_readahead.ondemand_readahead.filemap_fault.__xfs_filemap_fault 3.54 A+- 3% -0.6 2.95 A+- 15% perf-profile.calltrace.cycles-pp.add_to_page_cache_lru.mpage_readpages.__do_page_cache_readahead.ondemand_readahead.filemap_fault 0.31 A+-100% +0.5 0.81 A+- 17% perf-profile.calltrace.cycles-pp.xfs_bmap_add_extent_hole_delay.xfs_bmapi_reserve_delalloc.xfs_file_iomap_begin.iomap_apply.iomap_page_mkwrite 0.00 +0.6 0.63 A+- 22% perf-profile.calltrace.cycles-pp.follow_page_pte.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff 0.34 A+-102% +0.8 1.10 A+- 40% perf-profile.calltrace.cycles-pp.__set_page_dirty.mark_buffer_dirty.__block_commit_write.block_commit_write.iomap_page_mkwrite_actor 0.59 A+- 66% +1.0 1.59 A+- 26% perf-profile.calltrace.cycles-pp.mark_buffer_dirty.__block_commit_write.block_commit_write.iomap_page_mkwrite_actor.iomap_apply 0.00 +1.1 1.06 A+- 36% perf-profile.calltrace.cycles-pp.alloc_set_pte.finish_fault.__handle_mm_fault.handle_mm_fault.__get_user_pages 0.00 +1.1 1.11 A+- 34% perf-profile.calltrace.cycles-pp.finish_fault.__handle_mm_fault.handle_mm_fault.__get_user_pages.populate_vma_page_range 2.14 A+- 34% +1.2 3.35 A+- 19% perf-profile.calltrace.cycles-pp.xfs_file_iomap_begin.iomap_apply.iomap_page_mkwrite.__xfs_filemap_fault.do_page_mkwrite 1.14 A+- 34% +1.2 2.35 A+- 47% perf-profile.calltrace.cycles-pp.kmem_cache_alloc.alloc_buffer_head.alloc_page_buffers.create_empty_buffers.create_page_buffers 1.24 A+- 34% +1.2 2.46 A+- 45% perf-profile.calltrace.cycles-pp.alloc_page_buffers.create_empty_buffers.create_page_buffers.__block_write_begin_int.iomap_page_mkwrite_actor 1.19 A+- 33% +1.2 2.43 A+- 46% perf-profile.calltrace.cycles-pp.alloc_buffer_head.alloc_page_buffers.create_empty_buffers.create_page_buffers.__block_write_begin_int 1.57 A+- 33% +1.4 2.98 A+- 35% perf-profile.calltrace.cycles-pp.create_empty_buffers.create_page_buffers.__block_write_begin_int.iomap_page_mkwrite_actor.iomap_apply 1.60 A+- 34% +1.4 3.04 A+- 33% perf-profile.calltrace.cycles-pp.create_page_buffers.__block_write_begin_int.iomap_page_mkwrite_actor.iomap_apply.iomap_page_mkwrite 1.90 A+- 32% +2.0 3.86 A+- 38% perf-profile.calltrace.cycles-pp.__block_write_begin_int.iomap_page_mkwrite_actor.iomap_apply.iomap_page_mkwrite.__xfs_filemap_fault 0.00 +2.1 2.09 A+- 50% perf-profile.calltrace.cycles-pp.memcpy_erms.memcpy_to_page._copy_to_iter.copy_page_to_iter.shmem_file_read_iter 0.00 +2.1 2.11 A+- 50% perf-profile.calltrace.cycles-pp.memcpy_to_page._copy_to_iter.copy_page_to_iter.shmem_file_read_iter.do_iter_readv_writev 0.00 +2.2 2.19 A+- 50% perf-profile.calltrace.cycles-pp._copy_to_iter.copy_page_to_iter.shmem_file_read_iter.do_iter_readv_writev.do_iter_read 0.00 +2.2 2.22 A+- 50% perf-profile.calltrace.cycles-pp.copy_page_to_iter.shmem_file_read_iter.do_iter_readv_writev.do_iter_read.loop_queue_work 3.25 A+- 30% +2.9 6.13 A+- 26% perf-profile.calltrace.cycles-pp.iomap_page_mkwrite_actor.iomap_apply.iomap_page_mkwrite.__xfs_filemap_fault.do_page_mkwrite 0.00 +3.0 3.00 A+- 51% perf-profile.calltrace.cycles-pp.shmem_file_read_iter.do_iter_readv_writev.do_iter_read.loop_queue_work.kthread_worker_fn 0.00 +3.1 3.06 A+- 51% perf-profile.calltrace.cycles-pp.do_iter_readv_writev.do_iter_read.loop_queue_work.kthread_worker_fn.kthread 0.00 +3.3 3.26 A+- 52% perf-profile.calltrace.cycles-pp.do_iter_read.loop_queue_work.kthread_worker_fn.kthread.ret_from_fork 6.10 A+- 23% +4.1 10.16 A+- 21% perf-profile.calltrace.cycles-pp.iomap_apply.iomap_page_mkwrite.__xfs_filemap_fault.do_page_mkwrite.__handle_mm_fault 6.60 A+- 23% +4.4 11.01 A+- 19% perf-profile.calltrace.cycles-pp.iomap_page_mkwrite.__xfs_filemap_fault.do_page_mkwrite.__handle_mm_fault.handle_mm_fault 3.01 A+- 25% +5.9 8.94 A+- 38% perf-profile.calltrace.cycles-pp.secondary_startup_64 0.00 +9.5 9.53 A+- 27% perf-profile.calltrace.cycles-pp.__xfs_filemap_fault.do_page_mkwrite.__handle_mm_fault.handle_mm_fault.__get_user_pages 0.00 +9.6 9.59 A+- 27% perf-profile.calltrace.cycles-pp.do_page_mkwrite.__handle_mm_fault.handle_mm_fault.__get_user_pages.populate_vma_page_range 11.83 A+- 21% +10.1 21.96 A+- 18% perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__get_user_pages.populate_vma_page_range.__mm_populate 11.85 A+- 21% +10.4 22.23 A+- 18% perf-profile.calltrace.cycles-pp.handle_mm_fault.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff 12.52 A+- 21% +11.2 23.77 A+- 17% perf-profile.calltrace.cycles-pp.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff 12.52 A+- 21% +11.2 23.77 A+- 17% perf-profile.calltrace.cycles-pp.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 12.52 A+- 21% +11.2 23.77 A+- 17% perf-profile.calltrace.cycles-pp.populate_vma_page_range.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64 19.94 A+- 25% -17.5 2.42 A+-110% perf-profile.children.cycles-pp.do_access 17.94 A+- 23% -9.2 8.71 A+- 25% perf-profile.children.cycles-pp.__do_page_fault 17.89 A+- 23% -9.2 8.69 A+- 25% perf-profile.children.cycles-pp.do_page_fault 17.91 A+- 23% -9.2 8.73 A+- 25% perf-profile.children.cycles-pp.page_fault 9.39 A+- 27% -8.1 1.29 A+-110% perf-profile.children.cycles-pp.do_rw_once 14.54 A+- 6% -3.3 11.29 A+- 8% perf-profile.children.cycles-pp.__do_page_cache_readahead 14.52 A+- 5% -3.2 11.29 A+- 7% perf-profile.children.cycles-pp.ondemand_readahead 15.55 A+- 4% -2.8 12.78 A+- 7% perf-profile.children.cycles-pp.filemap_fault 16.04 A+- 4% -2.6 13.46 A+- 6% perf-profile.children.cycles-pp.__do_fault 7.32 A+- 17% -1.6 5.70 A+- 14% perf-profile.children.cycles-pp.__alloc_pages_nodemask 5.75 A+- 23% -1.6 4.13 A+- 17% perf-profile.children.cycles-pp.__alloc_pages_slowpath 5.38 A+- 22% -1.5 3.85 A+- 16% perf-profile.children.cycles-pp.try_to_free_pages 5.38 A+- 22% -1.5 3.85 A+- 16% perf-profile.children.cycles-pp.do_try_to_free_pages 4.18 A+- 7% -1.4 2.76 A+- 29% perf-profile.children.cycles-pp.page_vma_mapped_walk 9.57 -1.4 8.19 A+- 12% perf-profile.children.cycles-pp.mpage_readpages 1.76 A+- 30% -1.3 0.48 A+- 69% perf-profile.children.cycles-pp.pte_alloc_one 2.91 A+- 17% -1.1 1.77 A+- 35% perf-profile.children.cycles-pp.page_referenced_one 0.79 A+- 22% -0.7 0.07 A+-113% perf-profile.children.cycles-pp.filemap_map_pages 0.65 A+- 39% -0.6 0.08 A+- 70% perf-profile.children.cycles-pp.shrink_slab 1.12 A+- 32% -0.5 0.66 A+- 29% perf-profile.children.cycles-pp.free_pcppages_bulk 1.27 A+- 36% -0.5 0.82 A+- 38% perf-profile.children.cycles-pp.free_unref_page_list 3.68 A+- 2% -0.4 3.24 A+- 9% perf-profile.children.cycles-pp.add_to_page_cache_lru 0.75 A+- 22% -0.3 0.46 A+- 16% perf-profile.children.cycles-pp.__radix_tree_replace 0.30 A+- 24% -0.1 0.15 A+- 53% perf-profile.children.cycles-pp.find_vma 1.26 A+- 3% -0.1 1.12 A+- 6% perf-profile.children.cycles-pp.pagevec_lru_move_fn 0.29 A+- 9% -0.1 0.15 A+- 23% perf-profile.children.cycles-pp.replace_slot 1.31 A+- 3% -0.1 1.17 A+- 8% perf-profile.children.cycles-pp.__lru_cache_add 0.20 A+- 27% -0.1 0.06 A+- 65% perf-profile.children.cycles-pp.drain_local_pages_wq 0.20 A+- 27% -0.1 0.06 A+- 65% perf-profile.children.cycles-pp.drain_pages 0.20 A+- 27% -0.1 0.06 A+- 65% perf-profile.children.cycles-pp.drain_pages_zone 0.83 A+- 11% -0.1 0.69 A+- 16% perf-profile.children.cycles-pp.page_cache_tree_insert 0.59 A+- 7% -0.1 0.46 A+- 20% perf-profile.children.cycles-pp.unmap_page_range 0.58 A+- 5% -0.1 0.46 A+- 19% perf-profile.children.cycles-pp.unmap_vmas 0.21 A+- 18% -0.1 0.10 A+- 50% perf-profile.children.cycles-pp.page_mapped 0.13 A+- 38% -0.1 0.04 A+- 58% perf-profile.children.cycles-pp.super_cache_count 0.16 A+- 19% -0.1 0.09 A+- 27% perf-profile.children.cycles-pp.ptep_clear_flush_young 0.09 A+- 27% -0.1 0.03 A+-100% perf-profile.children.cycles-pp.down_read_trylock 0.00 +0.1 0.06 A+- 11% perf-profile.children.cycles-pp.find_next_and_bit 0.08 A+- 32% +0.1 0.14 A+- 34% perf-profile.children.cycles-pp.task_tick_fair 0.01 A+-173% +0.1 0.08 A+- 24% perf-profile.children.cycles-pp.cpumask_next_and 0.07 A+- 59% +0.1 0.13 A+- 30% perf-profile.children.cycles-pp.pmd_devmap_trans_unstable 0.02 A+-173% +0.1 0.09 A+- 21% perf-profile.children.cycles-pp.xfs_fsb_to_db 0.04 A+-101% +0.1 0.14 A+- 18% perf-profile.children.cycles-pp.kthread_queue_work 0.06 A+-101% +0.1 0.15 A+- 11% perf-profile.children.cycles-pp.loop_queue_rq 0.09 A+- 28% +0.1 0.19 A+- 27% perf-profile.children.cycles-pp.memset_erms 0.11 A+- 38% +0.1 0.22 A+- 15% perf-profile.children.cycles-pp.current_kernel_time64 0.07 A+- 70% +0.1 0.19 A+- 7% perf-profile.children.cycles-pp.blk_mq_dispatch_rq_list 0.00 +0.1 0.12 A+- 65% perf-profile.children.cycles-pp.touch_atime 0.14 A+- 18% +0.1 0.27 A+- 31% perf-profile.children.cycles-pp.xfs_file_iomap_end 0.11 A+- 70% +0.1 0.23 A+- 8% perf-profile.children.cycles-pp.blk_mq_run_hw_queue 0.10 A+- 69% +0.1 0.23 A+- 6% perf-profile.children.cycles-pp.__blk_mq_run_hw_queue 0.10 A+- 69% +0.1 0.23 A+- 6% perf-profile.children.cycles-pp.blk_mq_sched_dispatch_requests 0.10 A+- 65% +0.1 0.23 A+- 7% perf-profile.children.cycles-pp.blk_mq_do_dispatch_sched 0.19 A+- 15% +0.1 0.34 A+- 18% perf-profile.children.cycles-pp.current_time 0.35 A+- 25% +0.2 0.51 A+- 14% perf-profile.children.cycles-pp.try_to_wake_up 0.48 A+- 19% +0.2 0.63 A+- 5% perf-profile.children.cycles-pp.set_page_dirty 0.28 A+- 19% +0.2 0.44 A+- 24% perf-profile.children.cycles-pp.xfs_vn_update_time 0.07 A+-101% +0.2 0.23 A+- 7% perf-profile.children.cycles-pp.__blk_mq_delay_run_hw_queue 0.09 A+- 43% +0.2 0.25 A+- 46% perf-profile.children.cycles-pp.update_load_avg 0.80 A+- 9% +0.2 0.98 A+- 6% perf-profile.children.cycles-pp._raw_spin_lock_irqsave 0.55 A+- 17% +0.2 0.73 A+- 3% perf-profile.children.cycles-pp.radix_tree_tag_clear 0.46 A+- 20% +0.2 0.65 A+- 5% perf-profile.children.cycles-pp.xfs_iunlock 0.09 A+-102% +0.2 0.28 A+- 15% perf-profile.children.cycles-pp.blk_mq_flush_plug_list 0.10 A+- 79% +0.2 0.30 A+- 11% perf-profile.children.cycles-pp.blk_flush_plug_list 0.42 A+- 20% +0.2 0.61 A+- 19% perf-profile.children.cycles-pp.tick_sched_timer 0.49 A+- 13% +0.2 0.69 A+- 16% perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore 0.00 +0.2 0.20 A+- 19% perf-profile.children.cycles-pp.blk_finish_plug 0.00 +0.2 0.23 A+- 61% perf-profile.children.cycles-pp.mpage_end_io 1.46 A+- 13% +0.2 1.68 A+- 4% perf-profile.children.cycles-pp.__radix_tree_lookup 0.43 A+- 14% +0.3 0.70 A+- 18% perf-profile.children.cycles-pp.follow_page_pte 0.37 A+- 28% +0.3 0.64 A+- 18% perf-profile.children.cycles-pp.blk_mq_make_request 0.15 A+- 39% +0.3 0.42 A+- 83% perf-profile.children.cycles-pp.menu_select 0.65 A+- 8% +0.3 0.94 A+- 20% perf-profile.children.cycles-pp.file_update_time 0.45 A+- 12% +0.3 0.76 A+- 28% perf-profile.children.cycles-pp.radix_tree_tag_set 0.39 A+- 28% +0.3 0.70 A+- 17% perf-profile.children.cycles-pp.submit_bio 0.39 A+- 28% +0.3 0.70 A+- 17% perf-profile.children.cycles-pp.generic_make_request 0.73 A+- 10% +0.4 1.11 A+- 8% perf-profile.children.cycles-pp.radix_tree_lookup_slot 0.07 A+-119% +0.4 0.46 A+- 67% perf-profile.children.cycles-pp.__pte_alloc 0.65 A+- 14% +0.4 1.06 A+- 36% perf-profile.children.cycles-pp.__hrtimer_run_queues 0.22 A+- 37% +0.4 0.64 A+- 12% perf-profile.children.cycles-pp.follow_page_mask 0.92 A+- 6% +0.5 1.46 A+- 21% perf-profile.children.cycles-pp.alloc_set_pte 1.04 A+- 8% +0.5 1.59 A+- 19% perf-profile.children.cycles-pp.xfs_vm_set_page_dirty 1.18 A+- 8% +0.7 1.83 A+- 10% perf-profile.children.cycles-pp.find_get_entry 0.69 A+- 11% +0.7 1.42 A+- 19% perf-profile.children.cycles-pp.finish_fault 0.00 +0.8 0.77 A+- 70% perf-profile.children.cycles-pp.mempool_alloc 0.00 +0.8 0.77 A+- 71% perf-profile.children.cycles-pp.bio_alloc_bioset 1.22 A+- 16% +0.8 2.06 A+- 50% perf-profile.children.cycles-pp.smp_apic_timer_interrupt 1.00 A+- 28% +0.9 1.86 A+- 20% perf-profile.children.cycles-pp.mark_buffer_dirty 1.22 A+- 15% +0.9 2.09 A+- 50% perf-profile.children.cycles-pp.apic_timer_interrupt 1.22 A+- 25% +1.0 2.23 A+- 16% perf-profile.children.cycles-pp.__block_commit_write 1.18 A+- 26% +1.0 2.22 A+- 16% perf-profile.children.cycles-pp.block_commit_write 2.56 A+- 24% +1.1 3.69 A+- 16% perf-profile.children.cycles-pp.xfs_file_iomap_begin 1.32 A+- 27% +1.4 2.77 A+- 41% perf-profile.children.cycles-pp.alloc_buffer_head 1.41 A+- 26% +1.5 2.87 A+- 41% perf-profile.children.cycles-pp.alloc_page_buffers 1.75 A+- 26% +1.5 3.29 A+- 36% perf-profile.children.cycles-pp.create_empty_buffers 1.78 A+- 27% +1.6 3.34 A+- 34% perf-profile.children.cycles-pp.create_page_buffers 2.17 A+- 24% +1.8 4.00 A+- 35% perf-profile.children.cycles-pp.__block_write_begin_int 0.84 A+- 36% +1.9 2.71 A+- 31% perf-profile.children.cycles-pp.new_slab 0.88 A+- 35% +1.9 2.76 A+- 30% perf-profile.children.cycles-pp.___slab_alloc 0.88 A+- 35% +1.9 2.76 A+- 30% perf-profile.children.cycles-pp.__slab_alloc 0.00 +2.1 2.11 A+- 50% perf-profile.children.cycles-pp.memcpy_to_page 0.00 +2.2 2.20 A+- 50% perf-profile.children.cycles-pp._copy_to_iter 1.41 A+- 25% +2.2 3.61 A+- 26% perf-profile.children.cycles-pp.kmem_cache_alloc 0.00 +2.3 2.25 A+- 50% perf-profile.children.cycles-pp.copy_page_to_iter 3.42 A+- 23% +3.0 6.38 A+- 26% perf-profile.children.cycles-pp.iomap_page_mkwrite_actor 0.00 +3.0 3.01 A+- 51% perf-profile.children.cycles-pp.shmem_file_read_iter 0.00 +3.3 3.30 A+- 53% perf-profile.children.cycles-pp.do_iter_read 6.12 A+- 23% +4.2 10.27 A+- 21% perf-profile.children.cycles-pp.iomap_apply 2.14 A+- 22% +4.3 6.46 A+- 22% perf-profile.children.cycles-pp.intel_idle 6.65 A+- 24% +4.4 11.08 A+- 19% perf-profile.children.cycles-pp.iomap_page_mkwrite 7.61 A+- 24% +5.0 12.65 A+- 17% perf-profile.children.cycles-pp.do_page_mkwrite 2.75 A+- 26% +5.5 8.29 A+- 35% perf-profile.children.cycles-pp.cpuidle_enter_state 3.01 A+- 25% +5.9 8.94 A+- 38% perf-profile.children.cycles-pp.secondary_startup_64 3.01 A+- 25% +5.9 8.94 A+- 38% perf-profile.children.cycles-pp.cpu_startup_entry 3.01 A+- 25% +6.0 8.97 A+- 38% perf-profile.children.cycles-pp.do_idle 16.28 A+- 22% +10.9 27.21 A+- 15% perf-profile.children.cycles-pp.do_syscall_64 16.28 A+- 22% +11.0 27.24 A+- 15% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 12.56 A+- 21% +11.2 23.80 A+- 17% perf-profile.children.cycles-pp.vm_mmap_pgoff 12.52 A+- 21% +11.2 23.77 A+- 17% perf-profile.children.cycles-pp.__mm_populate 12.52 A+- 21% +11.2 23.77 A+- 17% perf-profile.children.cycles-pp.populate_vma_page_range 12.55 A+- 21% +11.3 23.80 A+- 17% perf-profile.children.cycles-pp.ksys_mmap_pgoff 12.54 A+- 21% +11.3 23.81 A+- 17% perf-profile.children.cycles-pp.__get_user_pages 9.01 A+- 26% -7.8 1.25 A+-110% perf-profile.self.cycles-pp.do_rw_once 3.84 A+- 24% -3.4 0.42 A+-108% perf-profile.self.cycles-pp.do_access 2.89 A+- 9% -1.0 1.88 A+- 30% perf-profile.self.cycles-pp.page_vma_mapped_walk 4.27 A+- 3% -0.9 3.35 A+- 18% perf-profile.self.cycles-pp.do_mpage_readpage 1.14 A+- 16% -0.4 0.70 A+- 41% perf-profile.self.cycles-pp.page_referenced_one 0.39 A+- 11% -0.1 0.24 A+- 28% perf-profile.self.cycles-pp.xfs_get_blocks 0.29 A+- 9% -0.1 0.15 A+- 23% perf-profile.self.cycles-pp.replace_slot 0.40 A+- 20% -0.1 0.26 A+- 33% perf-profile.self.cycles-pp.page_referenced 0.21 A+- 20% -0.1 0.10 A+- 50% perf-profile.self.cycles-pp.page_mapped 0.16 A+- 19% -0.1 0.09 A+- 27% perf-profile.self.cycles-pp.ptep_clear_flush_young 0.09 A+- 28% -0.1 0.03 A+-100% perf-profile.self.cycles-pp.down_read_trylock 0.17 A+- 3% -0.0 0.14 A+- 12% perf-profile.self.cycles-pp.mem_cgroup_commit_charge 0.07 A+- 14% +0.0 0.12 A+- 21% perf-profile.self.cycles-pp.kmem_cache_free 0.00 +0.1 0.06 A+- 11% perf-profile.self.cycles-pp.find_next_and_bit 0.09 A+- 27% +0.1 0.15 A+- 19% perf-profile.self.cycles-pp.iomap_page_mkwrite_actor 0.01 A+-173% +0.1 0.08 A+- 23% perf-profile.self.cycles-pp.radix_tree_lookup_slot 0.02 A+-173% +0.1 0.09 A+- 21% perf-profile.self.cycles-pp.xfs_fsb_to_db 0.09 A+- 36% +0.1 0.16 A+- 14% perf-profile.self.cycles-pp.__mark_inode_dirty 0.09 A+- 28% +0.1 0.16 A+- 31% perf-profile.self.cycles-pp.memset_erms 0.01 A+-173% +0.1 0.08 A+- 48% perf-profile.self.cycles-pp.memcpy_from_page 0.21 A+- 23% +0.1 0.28 A+- 12% perf-profile.self.cycles-pp.lock_page_memcg 0.13 A+- 32% +0.1 0.22 A+- 30% perf-profile.self.cycles-pp.mark_buffer_dirty 0.16 A+- 5% +0.1 0.25 A+- 12% perf-profile.self.cycles-pp.__block_commit_write 0.15 A+- 33% +0.1 0.25 A+- 32% perf-profile.self.cycles-pp.xfs_add_to_ioend 0.18 A+- 23% +0.1 0.28 A+- 15% perf-profile.self.cycles-pp.set_page_dirty 0.10 A+- 36% +0.1 0.22 A+- 15% perf-profile.self.cycles-pp.current_kernel_time64 0.07 A+- 69% +0.1 0.19 A+- 47% perf-profile.self.cycles-pp.security_file_permission 0.05 A+- 63% +0.1 0.18 A+- 71% perf-profile.self.cycles-pp.update_load_avg 0.13 A+- 27% +0.1 0.27 A+- 24% perf-profile.self.cycles-pp.__xfs_filemap_fault 0.23 A+- 27% +0.1 0.38 A+- 26% perf-profile.self.cycles-pp.xfs_ilock 0.44 A+- 20% +0.1 0.59 A+- 5% perf-profile.self.cycles-pp.radix_tree_tag_clear 0.17 A+- 29% +0.2 0.34 A+- 15% perf-profile.self.cycles-pp.follow_page_pte 0.00 +0.2 0.18 A+- 43% perf-profile.self.cycles-pp.shmem_file_read_iter 1.43 A+- 12% +0.2 1.62 A+- 3% perf-profile.self.cycles-pp.__radix_tree_lookup 0.16 A+- 42% +0.2 0.36 A+- 33% perf-profile.self.cycles-pp.xfs_bmapi_reserve_delalloc 0.33 A+- 19% +0.2 0.53 A+- 29% perf-profile.self.cycles-pp.kmem_cache_alloc 0.53 A+- 9% +0.2 0.76 A+- 10% perf-profile.self.cycles-pp._raw_spin_lock_irqsave 0.37 A+- 21% +0.2 0.61 A+- 23% perf-profile.self.cycles-pp.__block_write_begin_int 0.20 A+- 38% +0.3 0.48 A+- 13% perf-profile.self.cycles-pp.follow_page_mask 0.05 A+-112% +0.3 0.34 A+- 51% perf-profile.self.cycles-pp.__get_user_pages 0.45 A+- 12% +0.3 0.76 A+- 28% perf-profile.self.cycles-pp.radix_tree_tag_set 0.46 A+- 15% +0.3 0.79 A+- 11% perf-profile.self.cycles-pp.find_get_entry 0.46 A+- 33% +0.4 0.82 A+- 18% perf-profile.self.cycles-pp.xfs_file_iomap_begin 0.61 A+- 14% +0.4 0.98 A+- 16% perf-profile.self.cycles-pp.__handle_mm_fault 2.14 A+- 21% +4.3 6.46 A+- 22% perf-profile.self.cycles-pp.intel_idle vm-scalability.time.user_time 280 +-+-------------------------------------------------------------------+ | + + .+..+.+.. .+.. .+.. + +.. | 260 +-+ + + .+. +. +..+.+. +.. + +.. .+ | 240 +-+ + +. + +. | | + | 220 +-+ | | | 200 +-+ | | | 180 +-+ | 160 +-+ | | O O O O O O O | 140 O-+ O O O O O O O O O O O O O O O | O O O | 120 +-+-------------------------------------------------------------------+ vm-scalability.time.system_time 2600 +-+------------------------------------------------------------------+ | +.. .+. .+ .+.. | 2400 +-+ .. +. +..+..+. + .+. +.+..+..+ | 2200 +-+ +.+.. .+.+ +. | | .. +..+. | 2000 +-++ | | | 1800 +-+ | | | 1600 +-+ | 1400 +-+ | | O O | 1200 O-+O O O O O O O | | O O O O O O O O O O O O O O O O 1000 +-+------------------------------------------------------------------+ vm-scalability.time.percent_of_cpu_this_job_got 900 +-+-------------------------------------------------------------------+ 850 +-+ .+.. .+.. .+. .+.. .+.+.. | |.. .+.. .+ +. +. +. +..+. +..+..+ | 800 +-+ .+ +..+..+. | 750 +-++. | 700 +-+ | 650 +-+ | | | 600 +-+ | 550 +-+ | 500 +-+ | 450 +-+ O | O O O O O O O O O O O O O | 400 +-+ O O O O O O O O O O O O 350 +-+-------------------------------------------------------------------+ vm-scalability.time.major_page_faults 8000 +-+------------------------------------------------------------------+ | : | 7000 +-+ : | | : : | |: : | 6000 +-+ : +.. .+. .+..+. + +.. | | :.+.. + +. +..+. +..+.. .. + .. | 5000 +-+ + +..+.. + + + + | | +.+ | 4000 +-+ | | | | O O | 3000 +-+ O O O O O O O O O O O O O O O O O O O O O O O | 2000 +-+--------------------------------------O---------------------------+ vm-scalability.time.file_system_inputs 3e+08 +-+---------------------------------------------------------------+ | | 2.5e+08 +-+ O O | | | | O O O O 2e+08 +-+ O O | | | 1.5e+08 +-+ | | | 1e+08 +-+ | O O O O O O O O O O O O O O | | O O O O | 5e+07 +-+ | | | 0 +-+---------------------------------------------------------------+ vm-scalability.time.file_system_outputs 1.6e+09 +-+O-O--O--O------O-----O-O-----O----O-------O--------------------+ | | 1.5e+09 +-+ | O O O O O O O O | 1.4e+09 +-+ | | | 1.3e+09 +-+ O O O O O O O O | | 1.2e+09 +-+ | | | 1.1e+09 +-+ | |.. .+..+.. .+.+..+.. .+.. .+.. .+.. | 1e+09 +-++ +.+. + +..+.+. +.+. +.+..+.+ | | | 9e+08 +-+---------------------------------------------------------------+ vm-scalability.workload 6e+08 +-+---------------------------------------------------------------+ |.. +..+.. +.+..+.. +.. +.. +.. | | + .. + .. .. | 5.5e+08 +-++ +.+ + +..+.+ +.+ +.+..+.+ | | | | | 5e+08 +-+ | | | 4.5e+08 +-+ | | O O O O O O O O O O | | | 4e+08 O-+ O O O O O O O | | | | O O O O O O O O 3.5e+08 +-+---------------------------------------------------------------+ [*] 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, Rong Chen