Greeting, FYI, we noticed a 43.3% improvement of will-it-scale.per_process_ops due to commit: commit: 815744d75152078cde5391fc1e3c2d4424323fb6 ("mm: memcontrol: don't batch updates of local VM stats and events") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master in testcase: will-it-scale on test machine: 192 threads Intel(R) Xeon(R) CPU @ 2.20GHz with 192G memory with following parameters: nr_task: 100% mode: process test: page_fault3 cpufreq_governor: performance test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two. test-url: https://github.com/antonblanchard/will-it-scale 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/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-7/performance/x86_64-rhel-7.6/process/100%/debian-x86_64-2019-05-14.cgz/lkp-csl-2ap1/page_fault3/will-it-scale commit: c11fb13a11 ("Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid";) 815744d751 ("mm: memcontrol: don't batch updates of local VM stats and events") c11fb13a117e5a67 815744d75152078cde5391fc1e3 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 16:4 183% 23:4 perf-profile.calltrace.cycles-pp.sync_regs.error_entry.testcase 18:4 204% 26:4 perf-profile.calltrace.cycles-pp.error_entry.testcase 0:4 3% 0:4 perf-profile.children.cycles-pp.error_exit 19:4 216% 28:4 perf-profile.children.cycles-pp.error_entry 0:4 2% 0:4 perf-profile.self.cycles-pp.error_exit 2:4 24% 3:4 perf-profile.self.cycles-pp.error_entry %stddev %change %stddev \ | \ 470022 +43.3% 673338 will-it-scale.per_process_ops 90244291 +43.3% 1.293e+08 will-it-scale.workload 33.13 -2.1% 32.44 boot-time.dhcp 13.47 ± 6% -1.0 12.47 mpstat.cpu.all.usr% 0.14 ± 10% +76.8% 0.25 ± 7% turbostat.CPU%c1 609.72 +9.7% 668.75 turbostat.PkgWatt 22201 ± 2% -11.1% 19733 ± 3% numa-meminfo.node2.Inactive 22201 ± 2% -12.0% 19545 ± 4% numa-meminfo.node2.Inactive(anon) 6060 ± 4% +15.5% 7000 ± 3% numa-meminfo.node2.KernelStack 30791 ± 32% -47.1% 16275 ± 22% numa-meminfo.node3.Inactive 30675 ± 33% -47.5% 16090 ± 22% numa-meminfo.node3.Inactive(anon) 15906 -1.4% 15683 proc-vmstat.nr_page_table_pages 60975063 +38.1% 84216668 proc-vmstat.numa_hit 60881605 +38.2% 84123351 proc-vmstat.numa_local 61181699 +38.1% 84469061 proc-vmstat.pgalloc_normal 2.713e+10 +43.3% 3.887e+10 proc-vmstat.pgfault 58818843 ± 5% +38.1% 81212599 ± 4% proc-vmstat.pgfree 14513924 ± 2% +41.8% 20585689 ± 2% numa-numastat.node0.local_node 14537287 ± 2% +41.7% 20601299 ± 2% numa-numastat.node0.numa_hit 15413984 ± 2% +37.3% 21168399 numa-numastat.node1.local_node 15437340 ± 2% +37.3% 21191634 numa-numastat.node1.numa_hit 15397026 +38.9% 21394060 numa-numastat.node2.local_node 15428108 +38.8% 21417417 numa-numastat.node2.numa_hit 15632711 +35.1% 21119563 numa-numastat.node3.local_node 15648357 +35.2% 21150655 numa-numastat.node3.numa_hit 444.19 ± 12% +41.5% 628.34 ± 3% sched_debug.cfs_rq:/.exec_clock.stddev 172846 ± 7% +19.1% 205841 sched_debug.cfs_rq:/.min_vruntime.stddev 0.05 ± 8% -22.1% 0.04 ± 5% sched_debug.cfs_rq:/.nr_running.stddev 171875 ± 7% +19.2% 204919 sched_debug.cfs_rq:/.spread0.stddev 280585 ± 89% -93.9% 17176 ± 64% sched_debug.cpu.avg_idle.min 24193 ± 31% +61.1% 38975 ± 24% sched_debug.cpu.nr_switches.max 1136 ± 13% +34.4% 1527 ± 16% sched_debug.cpu.ttwu_count.stddev 1018 ± 15% +44.2% 1467 ± 18% sched_debug.cpu.ttwu_local.stddev 8487551 ± 2% +35.9% 11538420 ± 2% numa-vmstat.node0.numa_hit 8464705 ± 2% +36.1% 11523132 ± 2% numa-vmstat.node0.numa_local 8798431 +32.5% 11661199 numa-vmstat.node1.numa_hit 8689509 ± 2% +32.9% 11552294 numa-vmstat.node1.numa_local 5568 ± 2% -12.9% 4847 ± 5% numa-vmstat.node2.nr_inactive_anon 6057 ± 4% +15.6% 7001 ± 3% numa-vmstat.node2.nr_kernel_stack 5571 ± 2% -12.9% 4851 ± 5% numa-vmstat.node2.nr_zone_inactive_anon 8648618 +35.6% 11724115 numa-vmstat.node2.numa_hit 8532135 +36.1% 11615170 numa-vmstat.node2.numa_local 7672 ± 33% -48.1% 3980 ± 17% numa-vmstat.node3.nr_inactive_anon 7673 ± 33% -48.1% 3983 ± 17% numa-vmstat.node3.nr_zone_inactive_anon 8891497 +30.8% 11626904 numa-vmstat.node3.numa_hit 8789979 +30.9% 11510256 numa-vmstat.node3.numa_local 130.25 ±165% -98.7% 1.75 ± 74% interrupts.CPU109.RES:Rescheduling_interrupts 131.00 ±137% -97.7% 3.00 ± 97% interrupts.CPU118.RES:Rescheduling_interrupts 598.75 ±121% -93.5% 38.75 ± 62% interrupts.CPU13.RES:Rescheduling_interrupts 10.25 ±156% +5961.0% 621.25 ±167% interrupts.CPU145.RES:Rescheduling_interrupts 688.50 ±129% -95.4% 31.50 ±107% interrupts.CPU16.RES:Rescheduling_interrupts 1.25 ± 34% +10960.0% 138.25 ±113% interrupts.CPU161.RES:Rescheduling_interrupts 779.50 ±149% -94.9% 39.75 ± 92% interrupts.CPU17.RES:Rescheduling_interrupts 104.25 ± 69% -93.8% 6.50 ± 35% interrupts.CPU177.RES:Rescheduling_interrupts 100.25 ± 90% -89.3% 10.75 ±118% interrupts.CPU182.RES:Rescheduling_interrupts 494.25 ± 60% -79.0% 103.75 ± 16% interrupts.CPU2.RES:Rescheduling_interrupts 4480 ± 16% -40.5% 2665 ± 55% interrupts.CPU24.CAL:Function_call_interrupts 19.50 ±152% +928.2% 200.50 ± 91% interrupts.CPU29.RES:Rescheduling_interrupts 5259 ± 34% +59.7% 8397 interrupts.CPU39.NMI:Non-maskable_interrupts 5259 ± 34% +59.7% 8397 interrupts.CPU39.PMI:Performance_monitoring_interrupts 5258 ± 34% +59.6% 8390 interrupts.CPU42.NMI:Non-maskable_interrupts 5258 ± 34% +59.6% 8390 interrupts.CPU42.PMI:Performance_monitoring_interrupts 5253 ± 34% +59.8% 8393 interrupts.CPU43.NMI:Non-maskable_interrupts 5253 ± 34% +59.8% 8393 interrupts.CPU43.PMI:Performance_monitoring_interrupts 5248 ± 34% +59.9% 8394 interrupts.CPU44.NMI:Non-maskable_interrupts 5248 ± 34% +59.9% 8394 interrupts.CPU44.PMI:Performance_monitoring_interrupts 5261 ± 34% +60.0% 8419 interrupts.CPU57.NMI:Non-maskable_interrupts 5261 ± 34% +60.0% 8419 interrupts.CPU57.PMI:Performance_monitoring_interrupts 7874 +20.1% 9459 ± 11% interrupts.CPU95.RES:Rescheduling_interrupts 217.00 ± 49% -80.2% 43.00 ± 93% interrupts.CPU96.RES:Rescheduling_interrupts 53003 ± 5% -13.3% 45949 ± 6% interrupts.RES:Rescheduling_interrupts 2.68 -31.4% 1.84 perf-stat.i.MPKI 4.694e+10 +43.1% 6.715e+10 perf-stat.i.branch-instructions 1.238e+08 +43.4% 1.775e+08 ± 2% perf-stat.i.branch-misses 59.57 +9.1 68.63 perf-stat.i.cache-miss-rate% 3.661e+08 +14.0% 4.173e+08 perf-stat.i.cache-misses 2.49 -31.0% 1.72 perf-stat.i.cpi 1557 -12.7% 1360 perf-stat.i.cycles-between-cache-misses 0.02 ±109% -0.0 0.00 ± 13% perf-stat.i.dTLB-load-miss-rate% 6.494e+10 +44.9% 9.411e+10 perf-stat.i.dTLB-loads 1.59e+09 +43.4% 2.28e+09 perf-stat.i.dTLB-store-misses 3.43e+10 +44.8% 4.965e+10 perf-stat.i.dTLB-stores 2.29e+11 +44.2% 3.301e+11 perf-stat.i.instructions 0.40 +44.8% 0.58 perf-stat.i.ipc 90220636 +43.3% 1.293e+08 perf-stat.i.minor-faults 6.99 ± 2% -6.0 0.96 ± 17% perf-stat.i.node-load-miss-rate% 2563994 ± 2% -67.3% 839546 ± 12% perf-stat.i.node-load-misses 34196685 ± 2% +154.9% 87152794 ± 5% perf-stat.i.node-loads 14.00 -7.5 6.48 perf-stat.i.node-store-miss-rate% 14863715 ± 2% -38.3% 9165413 ± 2% perf-stat.i.node-store-misses 91337940 +44.8% 1.323e+08 perf-stat.i.node-stores 90221535 +43.3% 1.293e+08 perf-stat.i.page-faults 2.68 -31.4% 1.84 perf-stat.overall.MPKI 59.57 +9.1 68.63 perf-stat.overall.cache-miss-rate% 2.49 -31.0% 1.72 perf-stat.overall.cpi 1557 -12.7% 1359 perf-stat.overall.cycles-between-cache-misses 0.02 ±109% -0.0 0.00 ± 13% perf-stat.overall.dTLB-load-miss-rate% 0.40 +44.8% 0.58 perf-stat.overall.ipc 6.98 ± 2% -6.0 0.96 ± 17% perf-stat.overall.node-load-miss-rate% 14.00 -7.5 6.48 perf-stat.overall.node-store-miss-rate% 4.677e+10 +43.1% 6.692e+10 perf-stat.ps.branch-instructions 1.234e+08 +43.4% 1.769e+08 ± 2% perf-stat.ps.branch-misses 3.647e+08 +14.0% 4.158e+08 perf-stat.ps.cache-misses 6.471e+10 +45.0% 9.379e+10 perf-stat.ps.dTLB-loads 1.584e+09 +43.5% 2.273e+09 perf-stat.ps.dTLB-store-misses 3.417e+10 +44.8% 4.949e+10 perf-stat.ps.dTLB-stores 2.281e+11 +44.2% 3.29e+11 perf-stat.ps.instructions 89897978 +43.3% 1.288e+08 perf-stat.ps.minor-faults 2554791 ± 2% -67.2% 836695 ± 12% perf-stat.ps.node-load-misses 34073965 ± 2% +154.9% 86855995 ± 5% perf-stat.ps.node-loads 14810461 ± 2% -38.3% 9134179 ± 2% perf-stat.ps.node-store-misses 91010555 +44.9% 1.318e+08 perf-stat.ps.node-stores 89898192 +43.3% 1.288e+08 perf-stat.ps.page-faults 6.812e+13 +44.2% 9.822e+13 perf-stat.total.instructions 11934 ± 6% +35.0% 16111 ± 12% softirqs.CPU120.RCU 11745 ± 4% +14.5% 13450 ± 5% softirqs.CPU122.RCU 11990 ± 7% +11.6% 13378 ± 4% softirqs.CPU124.RCU 11979 ± 5% +12.4% 13466 ± 4% softirqs.CPU126.RCU 11997 ± 6% +11.4% 13370 ± 5% softirqs.CPU127.RCU 12165 ± 3% +12.4% 13677 ± 4% softirqs.CPU128.RCU 12213 ± 4% +10.0% 13431 ± 4% softirqs.CPU129.RCU 11827 ± 4% +16.9% 13831 ± 5% softirqs.CPU130.RCU 11469 ± 9% +19.7% 13725 ± 4% softirqs.CPU131.RCU 11869 ± 5% +15.5% 13711 ± 3% softirqs.CPU132.RCU 11751 ± 4% +15.7% 13596 ± 4% softirqs.CPU134.RCU 11675 ± 5% +16.6% 13615 ± 6% softirqs.CPU135.RCU 11900 ± 5% +15.0% 13687 ± 3% softirqs.CPU136.RCU 11959 ± 5% +14.0% 13636 ± 4% softirqs.CPU137.RCU 11940 ± 5% +13.7% 13576 ± 3% softirqs.CPU138.RCU 11905 ± 6% +15.9% 13804 ± 5% softirqs.CPU139.RCU 12342 ± 5% +11.4% 13750 ± 6% softirqs.CPU140.RCU 11828 ± 4% +13.3% 13401 ± 3% softirqs.CPU141.RCU 11823 ± 4% +14.5% 13536 ± 3% softirqs.CPU142.RCU 11658 ± 7% +15.6% 13472 ± 3% softirqs.CPU143.RCU 133947 ± 19% -23.1% 102992 ± 3% softirqs.CPU143.TIMER 12294 ± 3% +6.5% 13090 ± 3% softirqs.CPU145.RCU 12118 ± 3% +7.3% 12999 softirqs.CPU146.RCU 12079 ± 3% +9.6% 13240 ± 2% softirqs.CPU149.RCU 11937 ± 3% +11.0% 13256 ± 2% softirqs.CPU155.RCU 12003 ± 3% +11.2% 13348 ± 4% softirqs.CPU156.RCU 11979 ± 6% +9.1% 13075 ± 4% softirqs.CPU158.RCU 11992 ± 3% +9.6% 13148 ± 4% softirqs.CPU159.RCU 12283 ± 5% +14.0% 13997 ± 9% softirqs.CPU167.RCU 11803 +12.4% 13267 ± 3% softirqs.CPU180.RCU 12018 ± 5% +6.8% 12838 ± 4% softirqs.CPU187.RCU 12493 ± 5% +13.6% 14192 ± 4% softirqs.CPU27.RCU 12587 ± 6% +13.8% 14328 ± 6% softirqs.CPU30.RCU 12864 ± 3% +9.6% 14103 ± 4% softirqs.CPU33.RCU 12555 ± 4% +12.9% 14181 ± 6% softirqs.CPU34.RCU 12422 ± 4% +17.1% 14545 ± 4% softirqs.CPU35.RCU 12235 +17.1% 14328 ± 2% softirqs.CPU36.RCU 12710 ± 5% +10.1% 13989 ± 3% softirqs.CPU37.RCU 12441 ± 4% +14.6% 14262 ± 3% softirqs.CPU38.RCU 12457 ± 4% +12.4% 14000 ± 3% softirqs.CPU39.RCU 12503 ± 5% +13.5% 14188 ± 2% softirqs.CPU40.RCU 12430 ± 4% +15.9% 14408 ± 3% softirqs.CPU41.RCU 12494 ± 5% +14.5% 14310 ± 3% softirqs.CPU42.RCU 12776 ± 4% +13.9% 14547 ± 5% softirqs.CPU43.RCU 12466 ± 4% +12.6% 14040 ± 2% softirqs.CPU45.RCU 12361 ± 5% +15.8% 14313 ± 3% softirqs.CPU46.RCU 11235 ± 8% +26.8% 14244 ± 2% softirqs.CPU47.RCU 118604 ± 8% -12.5% 103803 ± 2% softirqs.CPU47.TIMER 12832 ± 4% +8.0% 13857 ± 5% softirqs.CPU62.RCU 12303 ± 4% +7.4% 13208 ± 7% softirqs.CPU73.RCU 12603 ± 5% +19.3% 15040 ± 9% softirqs.CPU8.RCU 10873 ± 18% +24.4% 13522 ± 3% softirqs.CPU81.RCU 12259 +9.4% 13412 ± 3% softirqs.CPU83.RCU 12381 ± 2% +9.2% 13517 ± 4% softirqs.CPU87.RCU 12363 ± 2% +8.7% 13440 ± 3% softirqs.CPU88.RCU 12498 ± 2% +9.3% 13656 ± 3% softirqs.CPU89.RCU 12202 ± 2% +9.5% 13359 ± 5% softirqs.CPU92.RCU 12291 ± 3% +10.9% 13629 ± 3% softirqs.CPU94.RCU 15.61 -9.2 6.40 perf-profile.calltrace.cycles-pp.alloc_set_pte.finish_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault 15.89 -9.1 6.81 perf-profile.calltrace.cycles-pp.finish_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 40.86 -8.9 31.99 perf-profile.calltrace.cycles-pp.handle_mm_fault.__do_page_fault.do_page_fault.page_fault.testcase 12.32 -8.8 3.57 ± 3% perf-profile.calltrace.cycles-pp.page_add_file_rmap.alloc_set_pte.finish_fault.__handle_mm_fault.handle_mm_fault 47.45 -8.1 39.30 perf-profile.calltrace.cycles-pp.__do_page_fault.do_page_fault.page_fault.testcase 49.07 -7.6 41.47 perf-profile.calltrace.cycles-pp.do_page_fault.page_fault.testcase 78.70 -7.5 71.19 perf-profile.calltrace.cycles-pp.testcase 9.92 -6.2 3.72 ± 16% perf-profile.calltrace.cycles-pp.page_remove_rmap.unmap_page_range.unmap_vmas.unmap_region.__do_munmap 6.82 ± 5% -5.5 1.31 ± 8% perf-profile.calltrace.cycles-pp.__count_memcg_events.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 6.45 ± 3% -5.2 1.23 ± 25% perf-profile.calltrace.cycles-pp.__mod_memcg_state.__mod_lruvec_state.page_remove_rmap.unmap_page_range.unmap_vmas 6.23 ± 4% -5.2 1.06 ± 3% perf-profile.calltrace.cycles-pp.__mod_memcg_state.__mod_lruvec_state.page_add_file_rmap.alloc_set_pte.finish_fault 7.19 ± 2% -5.0 2.14 ± 18% perf-profile.calltrace.cycles-pp.__mod_lruvec_state.page_remove_rmap.unmap_page_range.unmap_vmas.unmap_region 7.00 ± 3% -5.0 2.04 ± 2% perf-profile.calltrace.cycles-pp.__mod_lruvec_state.page_add_file_rmap.alloc_set_pte.finish_fault.__handle_mm_fault 32.31 -4.0 28.28 perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 14.76 -3.8 10.98 ± 9% perf-profile.calltrace.cycles-pp.unmap_page_range.unmap_vmas.unmap_region.__do_munmap.__vm_munmap 14.80 -3.8 11.04 ± 9% perf-profile.calltrace.cycles-pp.unmap_vmas.unmap_region.__do_munmap.__vm_munmap.__x64_sys_munmap 14.83 -3.8 11.07 ± 9% perf-profile.calltrace.cycles-pp.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 14.83 -3.8 11.07 ± 9% perf-profile.calltrace.cycles-pp.__do_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe 14.83 -3.8 11.07 ± 9% perf-profile.calltrace.cycles-pp.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 14.83 -3.8 11.08 ± 9% perf-profile.calltrace.cycles-pp.munmap 14.82 -3.8 11.07 ± 9% perf-profile.calltrace.cycles-pp.unmap_region.__do_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64 14.83 -3.8 11.08 ± 9% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.munmap 14.83 -3.8 11.08 ± 9% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 3.96 ± 6% -3.7 0.27 ±100% perf-profile.calltrace.cycles-pp.lock_page_memcg.page_add_file_rmap.alloc_set_pte.finish_fault.__handle_mm_fault 1.72 -0.9 0.81 perf-profile.calltrace.cycles-pp._raw_spin_lock.alloc_set_pte.finish_fault.__handle_mm_fault.handle_mm_fault 1.30 ± 3% -0.6 0.69 perf-profile.calltrace.cycles-pp.down_read_trylock.__do_page_fault.do_page_fault.page_fault.testcase 1.31 -0.6 0.71 perf-profile.calltrace.cycles-pp.up_read.__do_page_fault.do_page_fault.page_fault.testcase 0.84 -0.2 0.64 perf-profile.calltrace.cycles-pp.unlock_page.fault_dirty_shared_page.__handle_mm_fault.handle_mm_fault.__do_page_fault 0.55 +0.1 0.67 ± 4% perf-profile.calltrace.cycles-pp.release_pages.tlb_flush_mmu.unmap_page_range.unmap_vmas.unmap_region 0.53 ± 3% +0.2 0.72 ± 2% perf-profile.calltrace.cycles-pp.current_time.file_update_time.__handle_mm_fault.handle_mm_fault.__do_page_fault 0.70 +0.2 0.90 ± 3% perf-profile.calltrace.cycles-pp.tlb_flush_mmu.unmap_page_range.unmap_vmas.unmap_region.__do_munmap 1.91 +0.3 2.18 perf-profile.calltrace.cycles-pp.fault_dirty_shared_page.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 0.53 ± 2% +0.3 0.82 ± 4% perf-profile.calltrace.cycles-pp.vmacache_find.find_vma.__do_page_fault.do_page_fault.page_fault 0.92 +0.3 1.24 ± 3% perf-profile.calltrace.cycles-pp.___perf_sw_event.__perf_sw_event.do_page_fault.page_fault.testcase 0.60 ± 2% +0.3 0.93 ± 4% perf-profile.calltrace.cycles-pp.find_vma.__do_page_fault.do_page_fault.page_fault.testcase 0.92 +0.3 1.25 perf-profile.calltrace.cycles-pp.file_update_time.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 0.41 ± 57% +0.4 0.77 ± 3% perf-profile.calltrace.cycles-pp.set_page_dirty.fault_dirty_shared_page.__handle_mm_fault.handle_mm_fault.__do_page_fault 0.96 +0.4 1.34 perf-profile.calltrace.cycles-pp.swapgs_restore_regs_and_return_to_usermode.testcase 1.37 +0.4 1.80 perf-profile.calltrace.cycles-pp.__perf_sw_event.do_page_fault.page_fault.testcase 0.84 ± 3% +0.5 1.30 ± 2% perf-profile.calltrace.cycles-pp.___perf_sw_event.__perf_sw_event.__do_page_fault.do_page_fault.page_fault 0.00 +0.6 0.55 ± 2% perf-profile.calltrace.cycles-pp.set_page_dirty.unmap_page_range.unmap_vmas.unmap_region.__do_munmap 0.00 +0.6 0.56 ± 4% perf-profile.calltrace.cycles-pp.__mod_node_page_state.__mod_lruvec_state.page_add_file_rmap.alloc_set_pte.finish_fault 0.00 +0.6 0.56 ± 2% perf-profile.calltrace.cycles-pp.page_mapping.set_page_dirty.fault_dirty_shared_page.__handle_mm_fault.handle_mm_fault 1.42 ± 2% +0.7 2.12 perf-profile.calltrace.cycles-pp.__perf_sw_event.__do_page_fault.do_page_fault.page_fault.testcase 1.74 +0.8 2.55 ± 2% perf-profile.calltrace.cycles-pp.xas_load.find_get_entry.find_lock_entry.shmem_getpage_gfp.shmem_fault 8.78 +2.7 11.48 ± 3% perf-profile.calltrace.cycles-pp.find_lock_entry.shmem_getpage_gfp.shmem_fault.__do_fault.__handle_mm_fault 9.54 +2.9 12.42 ± 3% perf-profile.calltrace.cycles-pp.shmem_getpage_gfp.shmem_fault.__do_fault.__handle_mm_fault.handle_mm_fault 10.25 +3.1 13.40 ± 3% perf-profile.calltrace.cycles-pp.shmem_fault.__do_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault 10.57 +3.2 13.80 ± 3% perf-profile.calltrace.cycles-pp.__do_fault.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault 4.94 ± 2% +4.4 9.29 ± 4% perf-profile.calltrace.cycles-pp.find_get_entry.find_lock_entry.shmem_getpage_gfp.shmem_fault.__do_fault 84.06 +12.0 96.02 perf-profile.calltrace.cycles-pp.page_fault.testcase 12.70 ± 3% -10.4 2.31 ± 13% perf-profile.children.cycles-pp.__mod_memcg_state 14.21 ± 2% -10.0 4.23 ± 10% perf-profile.children.cycles-pp.__mod_lruvec_state 15.70 -9.2 6.54 perf-profile.children.cycles-pp.alloc_set_pte 15.93 -9.1 6.86 perf-profile.children.cycles-pp.finish_fault 40.99 -8.8 32.19 perf-profile.children.cycles-pp.handle_mm_fault 12.36 -8.8 3.60 ± 3% perf-profile.children.cycles-pp.page_add_file_rmap 47.57 -8.1 39.47 perf-profile.children.cycles-pp.__do_page_fault 49.12 -7.6 41.53 perf-profile.children.cycles-pp.do_page_fault 9.96 -6.2 3.77 ± 16% perf-profile.children.cycles-pp.page_remove_rmap 6.82 ± 5% -5.5 1.32 ± 8% perf-profile.children.cycles-pp.__count_memcg_events 5.18 ± 6% -4.4 0.77 ± 14% perf-profile.children.cycles-pp.lock_page_memcg 32.44 -4.0 28.45 perf-profile.children.cycles-pp.__handle_mm_fault 14.80 -3.8 11.04 ± 9% perf-profile.children.cycles-pp.unmap_vmas 14.80 -3.8 11.04 ± 9% perf-profile.children.cycles-pp.unmap_page_range 14.88 -3.8 11.12 ± 9% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 14.88 -3.8 11.12 ± 9% perf-profile.children.cycles-pp.do_syscall_64 14.83 -3.8 11.07 ± 9% perf-profile.children.cycles-pp.__x64_sys_munmap 14.83 -3.8 11.07 ± 9% perf-profile.children.cycles-pp.__do_munmap 14.83 -3.8 11.07 ± 9% perf-profile.children.cycles-pp.__vm_munmap 14.82 -3.8 11.07 ± 9% perf-profile.children.cycles-pp.unmap_region 14.83 -3.8 11.08 ± 9% perf-profile.children.cycles-pp.munmap 10.73 ± 8% -1.6 9.09 perf-profile.children.cycles-pp.native_irq_return_iret 1.76 -0.9 0.86 perf-profile.children.cycles-pp._raw_spin_lock 1.31 -0.6 0.71 perf-profile.children.cycles-pp.up_read 1.31 ± 3% -0.6 0.72 perf-profile.children.cycles-pp.down_read_trylock 0.84 -0.2 0.64 perf-profile.children.cycles-pp.unlock_page 0.37 ± 5% -0.2 0.18 ± 4% perf-profile.children.cycles-pp.__unlock_page_memcg 0.05 ± 9% +0.0 0.07 perf-profile.children.cycles-pp.p4d_offset 0.06 ± 6% +0.0 0.09 ± 4% perf-profile.children.cycles-pp.get_page_from_freelist 0.07 +0.0 0.10 ± 4% perf-profile.children.cycles-pp.pte_alloc_one 0.07 ± 6% +0.0 0.10 ± 5% perf-profile.children.cycles-pp.__alloc_pages_nodemask 0.07 ± 5% +0.0 0.10 ± 4% perf-profile.children.cycles-pp.__x86_indirect_thunk_rax 0.04 ± 57% +0.0 0.07 ± 6% perf-profile.children.cycles-pp.prep_new_page 0.03 ±100% +0.0 0.07 ± 7% perf-profile.children.cycles-pp.clear_page_erms 0.10 ± 4% +0.0 0.14 ± 10% perf-profile.children.cycles-pp.task_tick_fair 0.01 ±173% +0.0 0.05 ± 9% perf-profile.children.cycles-pp.unlock_page_memcg 0.03 ±100% +0.0 0.07 ± 10% perf-profile.children.cycles-pp.native_set_pte_at 0.10 +0.0 0.15 ± 3% perf-profile.children.cycles-pp.PageHuge 0.12 ± 6% +0.0 0.17 ± 14% perf-profile.children.cycles-pp.scheduler_tick 0.15 ± 8% +0.0 0.20 ± 2% perf-profile.children.cycles-pp._vm_normal_page 0.13 ± 3% +0.1 0.18 perf-profile.children.cycles-pp.page_rmapping 0.14 ± 3% +0.1 0.19 ± 8% perf-profile.children.cycles-pp.ktime_get_update_offsets_now 0.16 ± 2% +0.1 0.21 ± 4% perf-profile.children.cycles-pp.fpregs_assert_state_consistent 0.12 ± 4% +0.1 0.18 ± 2% perf-profile.children.cycles-pp.rcu_all_qs 0.00 +0.1 0.06 ± 7% perf-profile.children.cycles-pp.hrtimer_active 0.17 ± 6% +0.1 0.22 ± 15% perf-profile.children.cycles-pp.tick_sched_handle 0.13 ± 3% +0.1 0.19 ± 2% perf-profile.children.cycles-pp.pmd_pfn 0.16 ± 6% +0.1 0.22 ± 14% perf-profile.children.cycles-pp.update_process_times 0.17 ± 2% +0.1 0.24 perf-profile.children.cycles-pp.pmd_page_vaddr 0.21 ± 8% +0.1 0.27 ± 2% perf-profile.children.cycles-pp.perf_exclude_event 0.11 ± 23% +0.1 0.17 ± 11% perf-profile.children.cycles-pp.timespec64_trunc 0.09 ± 4% +0.1 0.16 ± 32% perf-profile.children.cycles-pp.mem_cgroup_from_task 0.22 ± 7% +0.1 0.30 ± 15% perf-profile.children.cycles-pp.tick_sched_timer 0.15 ± 3% +0.1 0.24 perf-profile.children.cycles-pp.free_pages_and_swap_cache 0.20 ± 2% +0.1 0.29 ± 2% perf-profile.children.cycles-pp.mark_page_accessed 0.20 ± 14% +0.1 0.29 ± 15% perf-profile.children.cycles-pp.ktime_get_coarse_real_ts64 0.25 ± 4% +0.1 0.34 perf-profile.children.cycles-pp.__might_sleep 0.12 ± 4% +0.1 0.23 ± 2% perf-profile.children.cycles-pp.__tlb_remove_page_size 0.27 ± 4% +0.1 0.37 ± 12% perf-profile.children.cycles-pp.__hrtimer_run_queues 0.23 ± 2% +0.1 0.34 ± 2% perf-profile.children.cycles-pp._cond_resched 0.36 ± 2% +0.1 0.48 ± 2% perf-profile.children.cycles-pp.prepare_exit_to_usermode 0.31 ± 4% +0.1 0.43 ± 4% perf-profile.children.cycles-pp.xas_start 0.56 +0.1 0.69 ± 3% perf-profile.children.cycles-pp.release_pages 0.66 +0.1 0.79 perf-profile.children.cycles-pp.___might_sleep 0.38 +0.1 0.51 perf-profile.children.cycles-pp.__set_page_dirty_no_writeback 0.69 ± 4% +0.1 0.83 ± 11% perf-profile.children.cycles-pp.apic_timer_interrupt 0.58 ± 4% +0.2 0.75 ± 12% perf-profile.children.cycles-pp.hrtimer_interrupt 0.61 ± 4% +0.2 0.79 ± 11% perf-profile.children.cycles-pp.smp_apic_timer_interrupt 0.56 ± 2% +0.2 0.75 ± 2% perf-profile.children.cycles-pp.current_time 0.59 ± 4% +0.2 0.78 ± 5% perf-profile.children.cycles-pp.pmd_devmap_trans_unstable 0.71 +0.2 0.92 ± 3% perf-profile.children.cycles-pp.tlb_flush_mmu 1.97 +0.3 2.27 perf-profile.children.cycles-pp.fault_dirty_shared_page 0.55 ± 2% +0.3 0.85 ± 4% perf-profile.children.cycles-pp.vmacache_find 1.17 ± 2% +0.3 1.48 perf-profile.children.cycles-pp.page_mapping 0.94 +0.3 1.28 perf-profile.children.cycles-pp.file_update_time 0.72 ± 2% +0.3 1.06 ± 9% perf-profile.children.cycles-pp.__mod_node_page_state 0.63 ± 2% +0.3 0.97 ± 3% perf-profile.children.cycles-pp.find_vma 0.96 +0.4 1.34 perf-profile.children.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.93 ± 2% +0.4 1.38 ± 2% perf-profile.children.cycles-pp.set_page_dirty 1.89 ± 2% +0.8 2.68 perf-profile.children.cycles-pp.___perf_sw_event 1.79 +0.8 2.61 perf-profile.children.cycles-pp.xas_load 2.81 +1.1 3.94 perf-profile.children.cycles-pp.__perf_sw_event 4.28 +1.9 6.15 perf-profile.children.cycles-pp.sync_regs 67.10 +2.4 69.48 perf-profile.children.cycles-pp.page_fault 8.85 +2.7 11.58 ± 3% perf-profile.children.cycles-pp.find_lock_entry 9.56 +2.9 12.45 ± 3% perf-profile.children.cycles-pp.shmem_getpage_gfp 10.27 +3.1 13.42 ± 3% perf-profile.children.cycles-pp.shmem_fault 10.58 +3.2 13.81 ± 3% perf-profile.children.cycles-pp.__do_fault 85.11 +3.8 88.87 perf-profile.children.cycles-pp.testcase 4.97 ± 2% +4.4 9.34 ± 4% perf-profile.children.cycles-pp.find_get_entry 12.60 ± 3% -10.3 2.25 ± 14% perf-profile.self.cycles-pp.__mod_memcg_state 6.80 ± 5% -5.5 1.29 ± 8% perf-profile.self.cycles-pp.__count_memcg_events 5.11 ± 6% -4.4 0.71 ± 14% perf-profile.self.cycles-pp.lock_page_memcg 2.84 ± 2% -1.7 1.09 perf-profile.self.cycles-pp.find_lock_entry 10.73 ± 8% -1.6 9.08 perf-profile.self.cycles-pp.native_irq_return_iret 1.73 -0.9 0.83 perf-profile.self.cycles-pp._raw_spin_lock 1.29 ± 4% -0.6 0.68 perf-profile.self.cycles-pp.down_read_trylock 1.30 -0.6 0.70 perf-profile.self.cycles-pp.up_read 1.37 -0.3 1.05 ± 3% perf-profile.self.cycles-pp.page_add_file_rmap 0.82 -0.2 0.61 perf-profile.self.cycles-pp.unlock_page 0.35 ± 4% -0.2 0.17 ± 3% perf-profile.self.cycles-pp.__unlock_page_memcg 0.05 +0.0 0.07 ± 5% perf-profile.self.cycles-pp.__x86_indirect_thunk_rax 0.14 ± 6% +0.0 0.16 ± 10% perf-profile.self.cycles-pp.perf_swevent_event 0.08 +0.0 0.11 ± 3% perf-profile.self.cycles-pp.PageHuge 0.11 ± 4% +0.0 0.14 perf-profile.self.cycles-pp.page_rmapping 0.09 ± 4% +0.0 0.13 perf-profile.self.cycles-pp.find_vma 0.03 ±100% +0.0 0.07 ± 7% perf-profile.self.cycles-pp.clear_page_erms 0.22 ± 6% +0.0 0.26 perf-profile.self.cycles-pp.__do_fault 0.09 +0.0 0.13 ± 3% perf-profile.self.cycles-pp.rcu_all_qs 0.16 ± 2% +0.0 0.21 ± 3% perf-profile.self.cycles-pp.prepare_exit_to_usermode 0.01 ±173% +0.0 0.06 ± 14% perf-profile.self.cycles-pp.native_set_pte_at 0.13 ± 8% +0.0 0.18 ± 3% perf-profile.self.cycles-pp._vm_normal_page 0.13 ± 5% +0.0 0.18 ± 9% perf-profile.self.cycles-pp.ktime_get_update_offsets_now 0.17 ± 9% +0.0 0.22 perf-profile.self.cycles-pp.perf_exclude_event 0.11 +0.1 0.16 ± 4% perf-profile.self.cycles-pp._cond_resched 0.00 +0.1 0.05 perf-profile.self.cycles-pp.unlock_page_memcg 0.12 ± 3% +0.1 0.17 ± 2% perf-profile.self.cycles-pp.pmd_pfn 0.00 +0.1 0.05 ± 8% perf-profile.self.cycles-pp.pmd_devmap 0.15 ± 5% +0.1 0.21 ± 4% perf-profile.self.cycles-pp.fpregs_assert_state_consistent 0.16 ± 2% +0.1 0.22 perf-profile.self.cycles-pp.pmd_page_vaddr 0.00 +0.1 0.06 ± 7% perf-profile.self.cycles-pp.hrtimer_active 0.10 ± 19% +0.1 0.16 ± 9% perf-profile.self.cycles-pp.timespec64_trunc 0.08 ± 10% +0.1 0.15 ± 33% perf-profile.self.cycles-pp.mem_cgroup_from_task 0.80 +0.1 0.87 perf-profile.self.cycles-pp.__mod_lruvec_state 0.15 ± 2% +0.1 0.23 ± 3% perf-profile.self.cycles-pp.free_pages_and_swap_cache 0.19 ± 2% +0.1 0.27 ± 3% perf-profile.self.cycles-pp.mark_page_accessed 0.23 ± 5% +0.1 0.32 perf-profile.self.cycles-pp.__might_sleep 0.18 ± 15% +0.1 0.28 ± 16% perf-profile.self.cycles-pp.ktime_get_coarse_real_ts64 0.20 ± 3% +0.1 0.29 perf-profile.self.cycles-pp.do_page_fault 0.23 +0.1 0.33 perf-profile.self.cycles-pp.finish_fault 0.11 ± 4% +0.1 0.21 ± 4% perf-profile.self.cycles-pp.__tlb_remove_page_size 0.28 ± 4% +0.1 0.39 ± 3% perf-profile.self.cycles-pp.xas_start 0.21 ± 17% +0.1 0.32 ± 6% perf-profile.self.cycles-pp.fault_dirty_shared_page 0.54 +0.1 0.66 ± 4% perf-profile.self.cycles-pp.release_pages 0.34 +0.1 0.46 perf-profile.self.cycles-pp.__set_page_dirty_no_writeback 0.65 +0.1 0.78 perf-profile.self.cycles-pp.___might_sleep 0.39 +0.1 0.53 ± 2% perf-profile.self.cycles-pp.file_update_time 0.29 +0.2 0.44 perf-profile.self.cycles-pp.set_page_dirty 0.70 ± 6% +0.2 0.86 ± 3% perf-profile.self.cycles-pp.shmem_getpage_gfp 0.56 ± 4% +0.2 0.74 ± 5% perf-profile.self.cycles-pp.pmd_devmap_trans_unstable 0.71 ± 2% +0.3 0.97 perf-profile.self.cycles-pp.shmem_fault 0.59 +0.3 0.85 perf-profile.self.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.73 +0.3 1.00 perf-profile.self.cycles-pp.page_fault 1.12 ± 2% +0.3 1.41 perf-profile.self.cycles-pp.page_mapping 0.52 ± 2% +0.3 0.82 ± 4% perf-profile.self.cycles-pp.vmacache_find 0.90 +0.3 1.22 ± 2% perf-profile.self.cycles-pp.__perf_sw_event 0.71 ± 2% +0.3 1.05 ± 9% perf-profile.self.cycles-pp.__mod_node_page_state 0.91 +0.4 1.29 perf-profile.self.cycles-pp.alloc_set_pte 1.29 +0.6 1.88 perf-profile.self.cycles-pp.__do_page_fault 1.57 +0.7 2.23 ± 2% perf-profile.self.cycles-pp.handle_mm_fault 1.48 +0.7 2.17 ± 2% perf-profile.self.cycles-pp.xas_load 1.56 ± 3% +0.7 2.26 ± 2% perf-profile.self.cycles-pp.___perf_sw_event 2.55 ± 4% +1.1 3.64 perf-profile.self.cycles-pp.__handle_mm_fault 2.99 +1.8 4.83 ± 8% perf-profile.self.cycles-pp.unmap_page_range 4.27 +1.9 6.13 perf-profile.self.cycles-pp.sync_regs 3.14 ± 3% +3.5 6.62 ± 6% perf-profile.self.cycles-pp.find_get_entry 18.54 +10.1 28.67 perf-profile.self.cycles-pp.testcase will-it-scale.per_process_ops 700000 +-+---------------------O------------------------------------------+ | O O O O O O O O O O O O O | O O O O O O O | 650000 +-+ | | | | | 600000 +-+ | | | 550000 +-+ | | | | | 500000 +-+ | | .+.. .+.. +.. ..+..+..+..+.. .+.. .+.. | |..+..+. +. +.. .. +..+. +..+. +..+. .| 450000 +-+----------------------------------------------------------------+ will-it-scale.workload 1.35e+08 +-+--------------------------------------------------------------+ 1.3e+08 +-+ O O O O O O O O O O O O O | O O O O O O O O | 1.25e+08 +-+ | 1.2e+08 +-+ | | | 1.15e+08 +-+ | 1.1e+08 +-+ | 1.05e+08 +-+ | | | 1e+08 +-+ | 9.5e+07 +-+ | |..+.. .+.. .+.. .+.. .+.+..+..+.. .+.. .+.. .| 9e+07 +-+ +. +. +..+. +..+. +..+. +..+. +. | 8.5e+07 +-+--------------------------------------------------------------+ [*] 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