Greeting, FYI, we noticed a 12.5% improvement of vm-scalability.median due to commit: commit: af53d3e9e04024885de5b4fda51e5fa362ae2bd8 ("mm: swapoff: shmem_unuse() stop eviction without igrab()") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master in testcase: vm-scalability on test machine: 88 threads Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz with 128G memory with following parameters: runtime: 300s test: small-allocs 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/tbox_group/test/testcase: gcc-7/performance/x86_64-rhel-7.6/debian-x86_64-2018-04-03.cgz/300s/lkp-bdw-ep2/small-allocs/vm-scalability commit: 64165b1aff ("mm: swapoff: take notice of completion sooner") af53d3e9e0 ("mm: swapoff: shmem_unuse() stop eviction without igrab()") 64165b1affc5bc16 af53d3e9e04024885de5b4fda51 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 1:4 -25% :4 dmesg.WARNING:at_ip___perf_sw_event/0x 1:4 6% 1:4 perf-profile.children.cycles-pp.error_entry 1:4 5% 1:4 perf-profile.self.cycles-pp.error_entry %stddev %change %stddev \ | \ 192772 +12.5% 216822 vm-scalability.median 16968696 +12.5% 19084289 vm-scalability.throughput 343.22 -8.5% 314.03 vm-scalability.time.elapsed_time 343.22 -8.5% 314.03 vm-scalability.time.elapsed_time.max 47020 ± 2% -5.5% 44434 vm-scalability.time.involuntary_context_switches 26179 -8.7% 23908 ± 2% vm-scalability.time.system_time 2.83 ± 3% -18.0% 2.32 ± 13% turbostat.CPU%c1 34401 ± 4% +16.4% 40057 ± 7% vmstat.system.cs 74998 -1.3% 74052 proc-vmstat.nr_active_anon 313.67 ±116% -82.6% 54.67 proc-vmstat.nr_dirtied 332.33 ± 4% -5.2% 315.00 proc-vmstat.nr_inactive_file 32972 ± 2% -3.6% 31772 proc-vmstat.nr_shmem 74998 -1.3% 74052 proc-vmstat.nr_zone_active_anon 332.33 ± 4% -5.2% 315.00 proc-vmstat.nr_zone_inactive_file 5135 ± 13% -20.4% 4087 ± 5% proc-vmstat.numa_hint_faults 22938 ± 32% -55.2% 10283 ± 87% proc-vmstat.numa_pages_migrated 100642 ± 5% -12.4% 88212 ± 7% proc-vmstat.numa_pte_updates 22938 ± 32% -55.2% 10283 ± 87% proc-vmstat.pgmigrate_success 2211 ± 6% -11.3% 1962 ± 3% slabinfo.avc_xperms_data.active_objs 2211 ± 6% -11.3% 1962 ± 3% slabinfo.avc_xperms_data.num_objs 2982 ± 4% +22.6% 3657 ± 8% slabinfo.eventpoll_pwq.active_objs 2982 ± 4% +22.6% 3657 ± 8% slabinfo.eventpoll_pwq.num_objs 832.33 ± 5% -18.9% 675.00 ± 6% slabinfo.file_lock_cache.active_objs 832.33 ± 5% -18.9% 675.00 ± 6% slabinfo.file_lock_cache.num_objs 357.67 ± 13% +21.9% 436.00 ± 11% slabinfo.kernfs_iattrs_cache.active_objs 357.67 ± 13% +21.9% 436.00 ± 11% slabinfo.kernfs_iattrs_cache.num_objs 657.00 ± 10% +84.7% 1213 ± 25% slabinfo.mnt_cache.active_objs 657.00 ± 10% +84.7% 1213 ± 25% slabinfo.mnt_cache.num_objs 1083 ± 9% +50.2% 1628 ± 17% slabinfo.pool_workqueue.active_objs 1084 ± 9% +50.2% 1628 ± 17% slabinfo.pool_workqueue.num_objs 82049 ± 97% -73.0% 22126 ± 3% sched_debug.cfs_rq:/.load.max 20.68 ± 14% -15.9% 17.39 ± 10% sched_debug.cfs_rq:/.load_avg.avg 2.39 ± 51% -55.7% 1.06 ± 57% sched_debug.cfs_rq:/.removed.util_avg.avg 38.78 ± 45% -40.6% 23.03 ± 2% sched_debug.cfs_rq:/.runnable_load_avg.max 1407 ± 2% -9.9% 1267 ± 5% sched_debug.cfs_rq:/.util_avg.max 1302 ± 4% -13.0% 1133 ± 4% sched_debug.cfs_rq:/.util_est_enqueued.max 4087 ± 49% -53.2% 1913 ± 4% sched_debug.cpu.avg_idle.min 40.17 ± 41% -36.5% 25.50 ± 5% sched_debug.cpu.cpu_load[0].max 4.28 ± 12% -25.2% 3.20 ± 25% sched_debug.cpu.cpu_load[3].min 4.56 ± 8% -27.1% 3.32 ± 21% sched_debug.cpu.cpu_load[4].min 82050 ± 97% -73.0% 22126 ± 3% sched_debug.cpu.load.max 0.24 ± 4% +10.5% 0.27 ± 5% sched_debug.cpu.nr_running.stddev 361090 ± 15% +27.1% 458815 ± 16% sched_debug.cpu.nr_switches.max 53498 ± 16% +20.8% 64606 ± 11% sched_debug.cpu.nr_switches.stddev -15.94 -15.9% -13.41 sched_debug.cpu.nr_uninterruptible.min 3.687e+10 +3.0% 3.8e+10 perf-stat.i.branch-instructions 18198224 +6.1% 19313004 perf-stat.i.branch-misses 45.75 -2.2 43.54 perf-stat.i.cache-miss-rate% 1.335e+08 +7.4% 1.434e+08 perf-stat.i.cache-references 34686 ± 4% +16.5% 40418 ± 7% perf-stat.i.context-switches 19532295 +8.0% 21094254 ± 2% perf-stat.i.dTLB-load-misses 0.03 ± 6% +0.0 0.04 ± 4% perf-stat.i.dTLB-store-miss-rate% 710649 ± 4% +38.9% 987128 ± 4% perf-stat.i.dTLB-store-misses 2.761e+09 +8.9% 3.007e+09 perf-stat.i.dTLB-stores 7145505 ± 2% +9.0% 7788838 ± 2% perf-stat.i.iTLB-load-misses 3102317 +8.7% 3371015 perf-stat.i.minor-faults 74.43 -2.5 71.93 perf-stat.i.node-load-miss-rate% 3972987 ± 3% +13.5% 4508220 ± 4% perf-stat.i.node-loads 54.67 -1.8 52.85 ± 2% perf-stat.i.node-store-miss-rate% 3102316 +8.7% 3371023 perf-stat.i.page-faults 0.83 ± 2% +5.4% 0.87 perf-stat.overall.MPKI 0.05 +0.0 0.05 perf-stat.overall.branch-miss-rate% 46.43 -2.2 44.21 perf-stat.overall.cache-miss-rate% 1.46 -1.6% 1.43 perf-stat.overall.cpi 0.04 +0.0 0.04 perf-stat.overall.dTLB-load-miss-rate% 0.03 ± 5% +0.0 0.03 ± 5% perf-stat.overall.dTLB-store-miss-rate% 22573 -6.5% 21098 ± 2% perf-stat.overall.instructions-per-iTLB-miss 0.69 +1.6% 0.70 perf-stat.overall.ipc 71.97 -2.3 69.66 perf-stat.overall.node-load-miss-rate% 11560 -6.3% 10837 perf-stat.overall.path-length 3.675e+10 +3.0% 3.787e+10 perf-stat.ps.branch-instructions 18140894 +6.1% 19251328 perf-stat.ps.branch-misses 1.333e+08 +7.3% 1.43e+08 perf-stat.ps.cache-references 34570 ± 4% +16.5% 40280 ± 7% perf-stat.ps.context-switches 19467989 +8.0% 21024486 ± 2% perf-stat.ps.dTLB-load-misses 708456 ± 4% +38.9% 983923 ± 4% perf-stat.ps.dTLB-store-misses 2.756e+09 +8.8% 2.998e+09 perf-stat.ps.dTLB-stores 7122526 ± 2% +9.0% 7762640 ± 2% perf-stat.ps.iTLB-load-misses 3092107 +8.7% 3359671 perf-stat.ps.minor-faults 3971985 ± 3% +13.2% 4498255 ± 4% perf-stat.ps.node-loads 3092107 +8.7% 3359671 perf-stat.ps.page-faults 5.586e+13 -6.3% 5.236e+13 perf-stat.total.instructions 775.33 ± 12% -18.8% 629.67 ± 13% interrupts.34:IR-PCI-MSI.1572864-edge.eth0-TxRx-0 7072 ±132% -97.6% 169.00 ± 3% interrupts.35:IR-PCI-MSI.1572865-edge.eth0-TxRx-1 343.67 ± 51% -51.3% 167.33 ± 4% interrupts.44:IR-PCI-MSI.1572874-edge.eth0-TxRx-10 240.33 ± 17% -31.3% 165.00 ± 4% interrupts.48:IR-PCI-MSI.1572878-edge.eth0-TxRx-14 197.00 ± 8% -14.0% 169.33 ± 3% interrupts.49:IR-PCI-MSI.1572879-edge.eth0-TxRx-15 173.00 -9.6% 156.33 interrupts.76:IR-PCI-MSI.1572904-edge.eth0-TxRx-40 197.00 ± 13% -20.6% 156.33 interrupts.82:IR-PCI-MSI.1572910-edge.eth0-TxRx-46 174.00 ± 2% -10.2% 156.33 interrupts.85:IR-PCI-MSI.1572913-edge.eth0-TxRx-49 308670 -5.9% 290349 interrupts.CAL:Function_call_interrupts 775.33 ± 12% -18.8% 629.67 ± 13% interrupts.CPU0.34:IR-PCI-MSI.1572864-edge.eth0-TxRx-0 7072 ±132% -97.6% 169.00 ± 3% interrupts.CPU1.35:IR-PCI-MSI.1572865-edge.eth0-TxRx-1 343.67 ± 51% -51.3% 167.33 ± 4% interrupts.CPU10.44:IR-PCI-MSI.1572874-edge.eth0-TxRx-10 3958 ± 15% +35.0% 5344 ± 18% interrupts.CPU11.RES:Rescheduling_interrupts 240.33 ± 17% -31.3% 165.00 ± 4% interrupts.CPU14.48:IR-PCI-MSI.1572878-edge.eth0-TxRx-14 197.00 ± 8% -14.0% 169.33 ± 3% interrupts.CPU15.49:IR-PCI-MSI.1572879-edge.eth0-TxRx-15 5240 ± 34% +49.8% 7850 interrupts.CPU15.NMI:Non-maskable_interrupts 5240 ± 34% +49.8% 7850 interrupts.CPU15.PMI:Performance_monitoring_interrupts 5270 ± 34% +49.2% 7863 interrupts.CPU16.NMI:Non-maskable_interrupts 5270 ± 34% +49.2% 7863 interrupts.CPU16.PMI:Performance_monitoring_interrupts 5234 ± 35% +49.9% 7844 interrupts.CPU19.NMI:Non-maskable_interrupts 5234 ± 35% +49.9% 7844 interrupts.CPU19.PMI:Performance_monitoring_interrupts 3987 ± 14% +64.1% 6544 ± 26% interrupts.CPU2.RES:Rescheduling_interrupts 5227 ± 34% +50.5% 7866 interrupts.CPU21.NMI:Non-maskable_interrupts 5227 ± 34% +50.5% 7866 interrupts.CPU21.PMI:Performance_monitoring_interrupts 5251 ± 34% +50.1% 7882 interrupts.CPU22.NMI:Non-maskable_interrupts 5251 ± 34% +50.1% 7882 interrupts.CPU22.PMI:Performance_monitoring_interrupts 3591 -14.8% 3060 ± 6% interrupts.CPU24.CAL:Function_call_interrupts 3752 ± 30% -33.3% 2503 ± 20% interrupts.CPU35.RES:Rescheduling_interrupts 2400 ± 18% +82.7% 4385 ± 38% interrupts.CPU37.RES:Rescheduling_interrupts 173.00 -9.6% 156.33 interrupts.CPU40.76:IR-PCI-MSI.1572904-edge.eth0-TxRx-40 197.00 ± 13% -20.6% 156.33 interrupts.CPU46.82:IR-PCI-MSI.1572910-edge.eth0-TxRx-46 6496 ± 43% -44.3% 3618 interrupts.CPU48.RES:Rescheduling_interrupts 174.00 ± 2% -10.2% 156.33 interrupts.CPU49.85:IR-PCI-MSI.1572913-edge.eth0-TxRx-49 2967 ± 12% +62.3% 4817 ± 29% interrupts.CPU51.RES:Rescheduling_interrupts 3538 ± 8% +35.3% 4788 ± 12% interrupts.CPU54.RES:Rescheduling_interrupts 2709 ± 10% +216.7% 8579 ± 49% interrupts.CPU55.RES:Rescheduling_interrupts 3410 ± 10% +77.2% 6043 ± 38% interrupts.CPU59.RES:Rescheduling_interrupts 3125 ± 14% +30.8% 4086 ± 5% interrupts.CPU61.RES:Rescheduling_interrupts 3489 ± 4% -8.1% 3206 interrupts.CPU66.CAL:Function_call_interrupts 4396 ± 30% -48.4% 2268 ± 19% interrupts.CPU67.RES:Rescheduling_interrupts 3547 ± 29% -49.5% 1790 ± 15% interrupts.CPU69.RES:Rescheduling_interrupts 6372 ± 71% -64.1% 2289 ± 31% interrupts.CPU71.RES:Rescheduling_interrupts 5363 ± 32% +42.8% 7658 ± 13% interrupts.CPU8.RES:Rescheduling_interrupts 1847 ± 38% +120.9% 4079 ± 22% interrupts.CPU81.RES:Rescheduling_interrupts 2386 ± 27% +169.3% 6425 ± 21% interrupts.CPU83.RES:Rescheduling_interrupts 3577 -10.8% 3190 ± 4% interrupts.CPU85.CAL:Function_call_interrupts 3007 ± 28% -36.4% 1912 ± 27% interrupts.CPU85.RES:Rescheduling_interrupts 127386 ± 4% -15.8% 107206 ± 9% softirqs.CPU0.TIMER 4781 ±131% -96.2% 182.67 ± 7% softirqs.CPU1.NET_RX 119534 ± 7% -11.6% 105649 ± 9% softirqs.CPU1.TIMER 116298 -12.5% 101784 ± 8% softirqs.CPU12.TIMER 121370 ± 4% -13.8% 104620 ± 7% softirqs.CPU16.TIMER 115479 ± 2% -7.6% 106670 ± 3% softirqs.CPU2.TIMER 117370 ± 4% -12.6% 102596 ± 8% softirqs.CPU20.TIMER 141857 ± 23% -27.4% 102983 ± 5% softirqs.CPU23.TIMER 121250 ± 4% -15.9% 101922 ± 3% softirqs.CPU24.TIMER 113802 ± 4% -12.3% 99762 ± 3% softirqs.CPU25.TIMER 115962 ± 7% -11.8% 102314 ± 5% softirqs.CPU27.TIMER 118600 ± 2% -15.9% 99746 ± 3% softirqs.CPU28.TIMER 111281 ± 3% -9.7% 100442 ± 2% softirqs.CPU29.TIMER 114727 ± 7% -11.7% 101284 ± 6% softirqs.CPU31.TIMER 117706 ± 2% -16.3% 98561 ± 2% softirqs.CPU32.TIMER 111704 ± 2% -11.1% 99308 ± 2% softirqs.CPU33.TIMER 113769 ± 5% -9.7% 102742 ± 3% softirqs.CPU34.TIMER 118381 ± 6% -14.3% 101509 ± 5% softirqs.CPU35.TIMER 124979 ± 5% -19.5% 100653 ± 3% softirqs.CPU36.TIMER 118029 ± 4% -14.1% 101349 ± 2% softirqs.CPU37.TIMER 116608 ± 5% -10.3% 104644 softirqs.CPU38.TIMER 120916 ± 7% -15.8% 101849 ± 5% softirqs.CPU39.TIMER 121219 ± 3% -14.4% 103712 ± 7% softirqs.CPU4.TIMER 124203 ± 5% -19.6% 99810 ± 5% softirqs.CPU40.TIMER 117777 ± 5% -14.0% 101305 ± 3% softirqs.CPU41.TIMER 15501 ± 6% +10.1% 17060 ± 7% softirqs.CPU42.RCU 117204 ± 3% -11.1% 104163 ± 3% softirqs.CPU42.TIMER 123928 ± 6% -16.4% 103591 ± 5% softirqs.CPU43.TIMER 121932 ± 4% -13.1% 105966 ± 9% softirqs.CPU44.TIMER 119465 ± 2% -14.0% 102690 ± 9% softirqs.CPU48.TIMER 15988 ± 9% -13.4% 13839 ± 3% softirqs.CPU49.RCU 118678 -14.5% 101519 ± 8% softirqs.CPU52.TIMER 115458 -12.3% 101281 ± 8% softirqs.CPU56.TIMER 119449 ± 3% -13.2% 103647 ± 7% softirqs.CPU60.TIMER 124537 ± 10% -17.8% 102344 ± 4% softirqs.CPU67.TIMER 121448 ± 5% -17.7% 99965 ± 3% softirqs.CPU68.TIMER 112604 ± 4% -12.8% 98184 ± 3% softirqs.CPU69.TIMER 116708 ± 7% -13.4% 101059 ± 5% softirqs.CPU71.TIMER 118026 ± 2% -16.5% 98503 ± 3% softirqs.CPU72.TIMER 110472 ± 3% -9.1% 100419 ± 3% softirqs.CPU73.TIMER 114615 ± 8% -12.3% 100546 ± 6% softirqs.CPU75.TIMER 117420 ± 3% -15.5% 99212 ± 3% softirqs.CPU76.TIMER 111314 -10.1% 100038 ± 3% softirqs.CPU77.TIMER 114170 ± 5% -11.5% 100999 ± 2% softirqs.CPU78.TIMER 118013 ± 7% -13.5% 102039 ± 6% softirqs.CPU79.TIMER 119115 -14.2% 102215 ± 9% softirqs.CPU8.TIMER 124833 ± 5% -18.8% 101420 ± 3% softirqs.CPU80.TIMER 116041 ± 5% -12.9% 101060 ± 3% softirqs.CPU81.TIMER 115255 ± 5% -9.2% 104659 ± 2% softirqs.CPU82.TIMER 119809 ± 7% -15.8% 100850 ± 5% softirqs.CPU83.TIMER 122483 ± 5% -19.4% 98724 ± 4% softirqs.CPU84.TIMER 116504 ± 4% -13.7% 100577 ± 3% softirqs.CPU85.TIMER 116582 ± 4% -12.0% 102642 ± 3% softirqs.CPU86.TIMER 123119 ± 8% -17.1% 102019 ± 6% softirqs.CPU87.TIMER 10286028 -11.2% 9128903 ± 5% softirqs.TIMER 67.05 ± 5% -7.1 59.96 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 67.05 ± 5% -7.1 59.97 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.mmap64 67.26 ± 5% -7.1 60.18 perf-profile.calltrace.cycles-pp.mmap64 66.19 ± 5% -6.9 59.30 perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 66.26 ± 5% -6.9 59.37 perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 79.90 -2.2 77.73 perf-profile.calltrace.cycles-pp.osq_lock.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write.vma_link 81.13 -2.2 78.97 perf-profile.calltrace.cycles-pp.down_write.vma_link.mmap_region.do_mmap.vm_mmap_pgoff 81.05 -2.2 78.89 perf-profile.calltrace.cycles-pp.call_rwsem_down_write_failed.down_write.vma_link.mmap_region.do_mmap 81.04 -2.2 78.89 perf-profile.calltrace.cycles-pp.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write.vma_link.mmap_region 82.13 -2.1 80.05 perf-profile.calltrace.cycles-pp.vma_link.mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff 82.52 -2.0 80.51 perf-profile.calltrace.cycles-pp.mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64 82.90 -2.0 80.94 perf-profile.calltrace.cycles-pp.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.92 -0.2 0.70 ± 4% perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault 0.78 -0.2 0.57 ± 5% perf-profile.calltrace.cycles-pp.exit_to_usermode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 0.78 -0.2 0.57 ± 5% perf-profile.calltrace.cycles-pp.task_work_run.exit_to_usermode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 0.77 ± 2% -0.2 0.57 ± 5% perf-profile.calltrace.cycles-pp.task_numa_work.task_work_run.exit_to_usermode_loop.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.11 -0.2 0.92 ± 2% perf-profile.calltrace.cycles-pp.handle_mm_fault.__do_page_fault.do_page_fault.page_fault.do_access 1.85 -0.1 1.77 perf-profile.calltrace.cycles-pp.__do_page_fault.do_page_fault.page_fault.do_access 3.83 +0.2 4.06 perf-profile.calltrace.cycles-pp.page_fault.do_access 0.17 ±141% +0.4 0.54 ± 2% perf-profile.calltrace.cycles-pp.swapgs_restore_regs_and_return_to_usermode.do_access 8.94 +1.1 9.99 perf-profile.calltrace.cycles-pp.do_access 10.60 +1.5 12.09 perf-profile.calltrace.cycles-pp.do_rw_once 16.81 ± 20% +5.0 21.76 ± 4% perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 17.05 ± 20% +5.0 22.01 ± 4% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 17.05 ± 20% +5.0 22.01 ± 4% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 16.82 ± 20% +5.0 21.79 ± 4% perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 67.27 ± 5% -7.1 60.19 perf-profile.children.cycles-pp.mmap64 79.93 -2.2 77.75 perf-profile.children.cycles-pp.osq_lock 81.13 -2.2 78.97 perf-profile.children.cycles-pp.down_write 81.05 -2.2 78.89 perf-profile.children.cycles-pp.call_rwsem_down_write_failed 81.05 -2.2 78.89 perf-profile.children.cycles-pp.rwsem_down_write_failed 84.15 -2.1 82.01 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 84.14 -2.1 82.01 perf-profile.children.cycles-pp.do_syscall_64 82.13 -2.1 80.05 perf-profile.children.cycles-pp.vma_link 82.53 -2.0 80.51 perf-profile.children.cycles-pp.mmap_region 82.90 -2.0 80.94 perf-profile.children.cycles-pp.do_mmap 82.99 -1.9 81.06 perf-profile.children.cycles-pp.vm_mmap_pgoff 83.09 -1.9 81.17 perf-profile.children.cycles-pp.ksys_mmap_pgoff 1.18 ± 3% -0.2 0.94 ± 5% perf-profile.children.cycles-pp.exit_to_usermode_loop 1.18 ± 3% -0.2 0.94 ± 5% perf-profile.children.cycles-pp.task_work_run 1.18 ± 3% -0.2 0.94 ± 5% perf-profile.children.cycles-pp.task_numa_work 0.94 -0.2 0.72 ± 3% perf-profile.children.cycles-pp.__handle_mm_fault 1.12 -0.2 0.94 ± 2% perf-profile.children.cycles-pp.handle_mm_fault 1.87 -0.1 1.79 perf-profile.children.cycles-pp.__do_page_fault 0.11 ± 4% -0.1 0.06 ± 8% perf-profile.children.cycles-pp.osq_unlock 0.40 -0.0 0.38 ± 2% perf-profile.children.cycles-pp.smp_apic_timer_interrupt 0.23 ± 2% -0.0 0.22 perf-profile.children.cycles-pp.__hrtimer_run_queues 0.05 +0.0 0.06 perf-profile.children.cycles-pp.selinux_mmap_file 0.05 +0.0 0.06 perf-profile.children.cycles-pp.__fget 0.05 +0.0 0.06 perf-profile.children.cycles-pp.__slab_alloc 0.05 +0.0 0.06 perf-profile.children.cycles-pp.___slab_alloc 0.05 +0.0 0.06 perf-profile.children.cycles-pp.new_slab 0.07 +0.0 0.08 perf-profile.children.cycles-pp.security_mmap_file 0.10 ± 4% +0.0 0.11 perf-profile.children.cycles-pp.vm_area_alloc 0.07 ± 6% +0.0 0.09 ± 5% perf-profile.children.cycles-pp.syscall_return_via_sysret 0.09 +0.0 0.10 ± 4% perf-profile.children.cycles-pp.vmacache_find 0.07 ± 7% +0.0 0.08 perf-profile.children.cycles-pp.kmem_cache_alloc 0.06 +0.0 0.07 ± 6% perf-profile.children.cycles-pp._raw_spin_lock 0.09 +0.0 0.11 ± 4% perf-profile.children.cycles-pp.sync_regs 0.07 ± 7% +0.0 0.09 ± 5% perf-profile.children.cycles-pp.__rb_insert_augmented 0.14 ± 3% +0.0 0.17 ± 5% perf-profile.children.cycles-pp.vma_policy_mof 0.22 ± 2% +0.0 0.25 ± 3% perf-profile.children.cycles-pp.find_vma 0.18 ± 2% +0.0 0.22 ± 2% perf-profile.children.cycles-pp.perf_event_mmap 0.22 ± 3% +0.0 0.26 perf-profile.children.cycles-pp.___perf_sw_event 0.28 +0.0 0.32 perf-profile.children.cycles-pp.unmapped_area_topdown 0.58 ± 2% +0.0 0.62 ± 4% perf-profile.children.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.31 +0.0 0.35 ± 2% perf-profile.children.cycles-pp.arch_get_unmapped_area_topdown 0.18 ± 2% +0.0 0.23 ± 4% perf-profile.children.cycles-pp.do_anonymous_page 0.33 +0.0 0.38 perf-profile.children.cycles-pp.get_unmapped_area 0.00 +0.1 0.05 perf-profile.children.cycles-pp.vma_interval_tree_augment_rotate 0.00 +0.1 0.05 perf-profile.children.cycles-pp.down_read_trylock 0.00 +0.1 0.05 perf-profile.children.cycles-pp.perf_iterate_sb 0.00 +0.1 0.05 perf-profile.children.cycles-pp.___might_sleep 0.29 +0.1 0.34 ± 2% perf-profile.children.cycles-pp.__perf_sw_event 0.44 +0.1 0.50 ± 2% perf-profile.children.cycles-pp.vma_interval_tree_insert 0.00 +0.1 0.06 perf-profile.children.cycles-pp.__vma_link_list 0.55 +0.1 0.64 perf-profile.children.cycles-pp.native_irq_return_iret 3.09 +0.1 3.20 perf-profile.children.cycles-pp.page_fault 0.78 +0.1 0.91 perf-profile.children.cycles-pp.rwsem_spin_on_owner 10.80 +1.3 12.07 perf-profile.children.cycles-pp.do_access 9.23 +1.3 10.55 perf-profile.children.cycles-pp.do_rw_once 79.58 -2.2 77.43 perf-profile.self.cycles-pp.osq_lock 0.71 -0.3 0.44 ± 6% perf-profile.self.cycles-pp.__handle_mm_fault 1.04 ± 3% -0.3 0.77 ± 5% perf-profile.self.cycles-pp.task_numa_work 0.21 -0.1 0.14 ± 5% perf-profile.self.cycles-pp.rwsem_down_write_failed 0.11 ± 4% -0.1 0.06 ± 8% perf-profile.self.cycles-pp.osq_unlock 0.18 ± 5% -0.0 0.14 ± 3% perf-profile.self.cycles-pp.up_write 0.05 +0.0 0.06 perf-profile.self.cycles-pp.__fget 0.07 ± 6% +0.0 0.09 ± 5% perf-profile.self.cycles-pp.syscall_return_via_sysret 0.09 ± 5% +0.0 0.10 perf-profile.self.cycles-pp.vmacache_find 0.07 +0.0 0.09 ± 5% perf-profile.self.cycles-pp.__perf_sw_event 0.06 ± 7% +0.0 0.08 perf-profile.self.cycles-pp.__rb_insert_augmented 0.08 +0.0 0.10 ± 4% perf-profile.self.cycles-pp.sync_regs 0.10 ± 4% +0.0 0.12 ± 4% perf-profile.self.cycles-pp.mmap_region 0.09 ± 5% +0.0 0.11 ± 8% perf-profile.self.cycles-pp.vma_policy_mof 0.08 ± 5% +0.0 0.10 ± 4% perf-profile.self.cycles-pp.do_anonymous_page 0.12 ± 4% +0.0 0.14 ± 6% perf-profile.self.cycles-pp.find_vma 0.17 ± 2% +0.0 0.19 ± 2% perf-profile.self.cycles-pp.handle_mm_fault 0.19 ± 4% +0.0 0.22 ± 2% perf-profile.self.cycles-pp.___perf_sw_event 0.25 +0.0 0.28 ± 3% perf-profile.self.cycles-pp.__do_page_fault 0.28 +0.0 0.32 perf-profile.self.cycles-pp.unmapped_area_topdown 0.00 +0.1 0.05 perf-profile.self.cycles-pp.page_fault 0.00 +0.1 0.05 perf-profile.self.cycles-pp.vma_interval_tree_augment_rotate 0.00 +0.1 0.05 perf-profile.self.cycles-pp.___might_sleep 0.44 +0.1 0.49 perf-profile.self.cycles-pp.vma_interval_tree_insert 0.00 +0.1 0.06 perf-profile.self.cycles-pp.__vma_link_list 0.37 ± 2% +0.1 0.44 perf-profile.self.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.55 +0.1 0.64 perf-profile.self.cycles-pp.native_irq_return_iret 0.78 +0.1 0.91 perf-profile.self.cycles-pp.rwsem_spin_on_owner 5.50 +0.8 6.31 perf-profile.self.cycles-pp.do_access 6.00 +0.9 6.92 perf-profile.self.cycles-pp.do_rw_once vm-scalability.throughput 2e+07 +-+-------OO-------O---O------------------------------------------+ 1.8e+07 O-OO OO O OO O O O O OO OO O | |.++.++.+.++.++.+.++.+.++.++.+.++.++.+.+ ++.+.++.+.+ ++.+.++.++.| 1.6e+07 +-+ : : : : | 1.4e+07 +-+ : : : : | | : : : : | 1.2e+07 +-+ : : : : | 1e+07 +-+ : : : : | 8e+06 +-+ :: :: | | :: :: | 6e+06 +-+ :: :: | 4e+06 +-+ :: :: | | : : | 2e+06 +-+ : : | 0 +-+--------------------------O----O-------------------------------+ vm-scalability.median 250000 +-+----------------------------------------------------------------+ | | O OO O OO OO O OO O OO O OO O O O O | 200000 +-++.+.++.++.+.++.+.++.+.++.++.+.++.+.++ +.+.++.+.++ ++.++.+.++.| | : : : : | | : : : : | 150000 +-+ : : : : | | : : : : | 100000 +-+ : : : : | | :: : : | | :: : : | 50000 +-+ :: : : | | : : | | : : | 0 +-+--------------------------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, Rong Chen