Greeting, FYI, we noticed a -17.0% regression of vm-scalability.median due to commit: commit: 9266baf8eea588a78e5a0aa4ccd6cae36ff7349d ("mm: Optimized hugepage zeroing & copying from user") url: https://github.com/0day-ci/linux/commits/UPDATE-20200501-184514/Prathu-Baronia/mm-Optimized-hugepage-zeroing-copying-from-user/20200415-041712 in testcase: vm-scalability on test machine: 16 threads Intel(R) Xeon(R) E-2278G CPU @ 3.40GHz with 32G memory with following parameters: runtime: 300s size: 8T test: anon-w-seq-hugetlb cpufreq_governor: performance ucode: 0xca 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/ In addition to that, the commit also has significant impact on the following tests: +------------------+----------------------------------------------------------------------+ | testcase: change | will-it-scale: will-it-scale.per_process_ops -4.9% regression | | test machine | 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory | | test parameters | cpufreq_governor=performance | | | mode=process | | | nr_task=16 | | | test=page_fault1 | | | ucode=0x21 | +------------------+----------------------------------------------------------------------+ | testcase: change | vm-scalability: vm-scalability.median 8.4% improvement | | test machine | 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory | | test parameters | cpufreq_governor=performance | | | runtime=300s | | | size=8T | | | test=anon-w-seq-mt | | | ucode=0x21 | +------------------+----------------------------------------------------------------------+ | testcase: change | vm-scalability: vm-scalability.median -2.6% regression | | test machine | 96 threads Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz with 256G memory | | test parameters | cpufreq_governor=performance | | | runtime=300s | | | size=8T | | | test=anon-cow-seq-hugetlb | | | ucode=0x500002c | +------------------+----------------------------------------------------------------------+ | testcase: change | will-it-scale: will-it-scale.per_thread_ops -3.5% regression | | test machine | 8 threads Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz with 16G memory | | test parameters | cpufreq_governor=performance | | | mode=thread | | | nr_task=100% | | | test=page_fault1 | | | ucode=0x21 | +------------------+----------------------------------------------------------------------+ | testcase: change | will-it-scale: will-it-scale.per_process_ops -5.3% regression | | test machine | 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory | | test parameters | cpufreq_governor=performance | | | test=page_fault1 | | | ucode=0x21 | +------------------+----------------------------------------------------------------------+ If you fix the issue, kindly add following tag Reported-by: kernel test robot 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-7.6/debian-x86_64-20191114.cgz/300s/8T/lkp-cfl-e1/anon-w-seq-hugetlb/vm-scalability/0xca commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 dmesg.WARNING:at#for_ip_interrupt_entry/0x %stddev %change %stddev \ | \ 0.00 +35.0% 0.00 ± 41% vm-scalability.free_time 885731 -17.0% 735379 vm-scalability.median 14052443 -16.6% 11724633 vm-scalability.throughput 130581 ± 2% -9.0% 118824 ± 3% vm-scalability.time.involuntary_context_switches 2059540 -15.5% 1739286 vm-scalability.time.minor_page_faults 1466 +1.4% 1487 vm-scalability.time.percent_of_cpu_this_job_got 1430 +17.8% 1685 vm-scalability.time.system_time 2993 -6.5% 2797 vm-scalability.time.user_time 16555 -12.1% 14554 vm-scalability.time.voluntary_context_switches 3.937e+09 -15.6% 3.323e+09 vm-scalability.workload 322.62 -0.9% 319.64 boot-time.idle 5.46 ± 2% -1.0 4.50 ± 2% mpstat.cpu.all.idle% 31.64 +5.1 36.71 mpstat.cpu.all.sys% 62.00 -6.5% 58.00 vmstat.cpu.us 2220 ± 2% -10.8% 1980 vmstat.system.cs 1140 ± 5% -12.6% 997.00 ± 7% slabinfo.kmalloc-rcl-96.active_objs 1140 ± 5% -12.6% 997.00 ± 7% slabinfo.kmalloc-rcl-96.num_objs 6311 -9.9% 5685 slabinfo.vmap_area.active_objs 6311 -9.9% 5685 slabinfo.vmap_area.num_objs 132283 ± 9% -12.8% 115384 softirqs.CPU0.TIMER 137945 ± 7% -12.9% 120092 softirqs.CPU4.TIMER 26018 ± 4% -11.9% 22933 ± 9% softirqs.CPU5.RCU 134170 ± 8% -11.2% 119154 softirqs.CPU8.TIMER 354273 ± 7% -14.5% 303080 ± 7% cpuidle.C1.time 1473619 ± 37% -29.0% 1045628 ± 2% cpuidle.C1E.time 4610710 ± 51% -48.9% 2357842 ± 23% cpuidle.C3.time 23383 ± 41% -39.8% 14070 ± 11% cpuidle.C3.usage 2.605e+08 -16.8% 2.166e+08 cpuidle.C6.time 552177 -16.6% 460777 cpuidle.C6.usage 16488 ± 9% -30.4% 11472 ± 14% cpuidle.POLL.usage 8308 ± 5% -11.4% 7364 ± 6% interrupts.CAL:Function_call_interrupts 479.00 ± 2% -12.6% 418.50 ± 3% interrupts.CPU0.CAL:Function_call_interrupts 514.50 ± 4% -13.0% 447.50 ± 9% interrupts.CPU1.CAL:Function_call_interrupts 529.50 ± 4% -15.3% 448.75 ± 6% interrupts.CPU10.CAL:Function_call_interrupts 528.50 ± 4% -13.0% 459.75 ± 6% interrupts.CPU14.CAL:Function_call_interrupts 526.25 ± 6% -15.7% 443.50 ± 9% interrupts.CPU2.CAL:Function_call_interrupts 523.75 ± 3% -9.6% 473.25 ± 6% interrupts.CPU5.CAL:Function_call_interrupts 518.75 ± 6% -10.9% 462.25 ± 6% interrupts.CPU6.CAL:Function_call_interrupts 3608 ± 30% +101.4% 7268 ± 20% interrupts.CPU6.RES:Rescheduling_interrupts 81528 ± 4% -12.2% 71606 ± 5% interrupts.RES:Rescheduling_interrupts 373.33 ± 28% +47.7% 551.38 ± 5% sched_debug.cfs_rq:/.load_avg.max 90.00 ± 34% +50.7% 135.66 sched_debug.cfs_rq:/.load_avg.stddev 13.32 ± 13% +23.0% 16.39 ± 6% sched_debug.cfs_rq:/.nr_spread_over.avg 4816 ± 17% -31.8% 3286 ± 22% sched_debug.cpu.curr->pid.min 39104 ± 3% -10.7% 34937 sched_debug.cpu.nr_switches.max 7502 ± 7% -12.5% 6566 ± 3% sched_debug.cpu.nr_switches.stddev 34374 ± 5% -14.3% 29462 ± 5% sched_debug.cpu.sched_count.max 6717 ± 12% -16.4% 5617 ± 4% sched_debug.cpu.sched_count.stddev 2224 ± 5% -13.4% 1925 ± 4% sched_debug.cpu.sched_goidle.avg 1468 ± 5% -19.5% 1181 ± 3% sched_debug.cpu.sched_goidle.min 17663 ± 7% -14.0% 15196 ± 4% sched_debug.cpu.ttwu_count.max 3887 ± 16% -16.2% 3258 ± 4% sched_debug.cpu.ttwu_count.stddev 11677 ± 6% -18.3% 9544 ± 9% sched_debug.cpu.ttwu_local.max 2596 ± 12% -17.1% 2151 ± 6% sched_debug.cpu.ttwu_local.stddev 1709412 -15.6% 1442880 proc-vmstat.htlb_buddy_alloc_success 167.50 -11.0% 149.00 proc-vmstat.nr_dirtied 387471 +2.2% 395950 proc-vmstat.nr_dirty_background_threshold 775891 +2.2% 792870 proc-vmstat.nr_dirty_threshold 3943955 +2.2% 4030727 proc-vmstat.nr_free_pages 8037 -2.4% 7847 proc-vmstat.nr_inactive_anon 7418 -1.2% 7328 proc-vmstat.nr_mapped 11491 -1.8% 11288 proc-vmstat.nr_shmem 155.50 -11.3% 138.00 proc-vmstat.nr_written 8037 -2.4% 7847 proc-vmstat.nr_zone_inactive_anon 2351084 -13.1% 2042972 proc-vmstat.numa_hit 2351084 -13.1% 2042972 proc-vmstat.numa_local 8.759e+08 -15.6% 7.394e+08 proc-vmstat.pgalloc_normal 2451008 -13.1% 2129664 proc-vmstat.pgfault 8.749e+08 -15.7% 7.373e+08 proc-vmstat.pgfree 25.20 ± 2% -25.2 0.00 perf-profile.calltrace.cycles-pp.clear_subpage.clear_huge_page.hugetlb_no_page.hugetlb_fault.handle_mm_fault 24.59 ± 2% -24.6 0.00 perf-profile.calltrace.cycles-pp.clear_page_erms.clear_subpage.clear_huge_page.hugetlb_no_page.hugetlb_fault 26.17 -4.7 21.52 ± 2% perf-profile.calltrace.cycles-pp.do_rw_once 3.15 ± 5% -0.5 2.61 ± 5% perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.alloc_fresh_huge_page.alloc_surplus_huge_page.hugetlb_acct_memory.hugetlb_reserve_pages 3.15 ± 5% -0.5 2.60 ± 5% perf-profile.calltrace.cycles-pp.get_page_from_freelist.__alloc_pages_nodemask.alloc_fresh_huge_page.alloc_surplus_huge_page.hugetlb_acct_memory 3.17 ± 5% -0.5 2.63 ± 5% perf-profile.calltrace.cycles-pp.alloc_fresh_huge_page.alloc_surplus_huge_page.hugetlb_acct_memory.hugetlb_reserve_pages.hugetlbfs_file_mmap 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.mmap64 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe.mmap64 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.calltrace.cycles-pp.mmap64 3.18 ± 5% -0.5 2.64 ± 5% perf-profile.calltrace.cycles-pp.alloc_surplus_huge_page.hugetlb_acct_memory.hugetlb_reserve_pages.hugetlbfs_file_mmap.mmap_region 3.20 ± 4% -0.5 2.66 ± 5% perf-profile.calltrace.cycles-pp.hugetlb_acct_memory.hugetlb_reserve_pages.hugetlbfs_file_mmap.mmap_region.do_mmap 3.20 ± 4% -0.5 2.66 ± 5% perf-profile.calltrace.cycles-pp.hugetlbfs_file_mmap.mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff 3.20 ± 4% -0.5 2.66 ± 5% perf-profile.calltrace.cycles-pp.hugetlb_reserve_pages.hugetlbfs_file_mmap.mmap_region.do_mmap.vm_mmap_pgoff 2.79 ± 5% -0.5 2.33 ± 5% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.get_page_from_freelist.__alloc_pages_nodemask.alloc_fresh_huge_page.alloc_surplus_huge_page 2.78 ± 5% -0.5 2.33 ± 5% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.get_page_from_freelist.__alloc_pages_nodemask.alloc_fresh_huge_page 1.51 ± 8% -0.3 1.18 ± 3% perf-profile.calltrace.cycles-pp.ret_from_fork 1.51 ± 8% -0.3 1.18 ± 3% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork 1.49 ± 8% -0.3 1.17 ± 3% perf-profile.calltrace.cycles-pp.drm_fb_helper_dirty_work.process_one_work.worker_thread.kthread.ret_from_fork 1.50 ± 8% -0.3 1.18 ± 3% perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork 1.50 ± 8% -0.3 1.18 ± 3% perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork 1.45 ± 8% -0.3 1.14 ± 3% perf-profile.calltrace.cycles-pp.memcpy_erms.drm_fb_helper_dirty_work.process_one_work.worker_thread.kthread 1.05 ± 5% -0.3 0.79 ± 5% perf-profile.calltrace.cycles-pp.console_unlock.vprintk_emit.devkmsg_emit.devkmsg_write.new_sync_write 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.vprintk_emit.devkmsg_emit.devkmsg_write.new_sync_write.vfs_write 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.new_sync_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.write._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.devkmsg_write.new_sync_write.vfs_write.ksys_write.do_syscall_64 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.calltrace.cycles-pp.devkmsg_emit.devkmsg_write.new_sync_write.vfs_write.ksys_write 0.89 ± 6% -0.2 0.66 ± 5% perf-profile.calltrace.cycles-pp.serial8250_console_write.console_unlock.vprintk_emit.devkmsg_emit.devkmsg_write 0.83 ± 7% -0.2 0.60 ± 5% perf-profile.calltrace.cycles-pp.uart_console_write.serial8250_console_write.console_unlock.vprintk_emit.devkmsg_emit 1.42 ± 3% -0.2 1.24 ± 6% perf-profile.calltrace.cycles-pp.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.42 ± 3% -0.2 1.24 ± 6% perf-profile.calltrace.cycles-pp.do_group_exit.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.42 ± 3% -0.2 1.24 ± 6% perf-profile.calltrace.cycles-pp.do_exit.do_group_exit.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.41 ± 2% -0.2 1.23 ± 6% perf-profile.calltrace.cycles-pp.mmput.do_exit.do_group_exit.__x64_sys_exit_group.do_syscall_64 1.41 ± 2% -0.2 1.23 ± 6% perf-profile.calltrace.cycles-pp.exit_mmap.mmput.do_exit.do_group_exit.__x64_sys_exit_group 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.calltrace.cycles-pp.tlb_finish_mmu.exit_mmap.mmput.do_exit.do_group_exit 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.calltrace.cycles-pp.tlb_flush_mmu.tlb_finish_mmu.exit_mmap.mmput.do_exit 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.calltrace.cycles-pp.release_pages.tlb_flush_mmu.tlb_finish_mmu.exit_mmap.mmput 1.38 ± 2% -0.2 1.20 ± 6% perf-profile.calltrace.cycles-pp.__free_huge_page.release_pages.tlb_flush_mmu.tlb_finish_mmu.exit_mmap 1.44 ± 2% -0.2 1.27 ± 5% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 1.44 ± 2% -0.2 1.27 ± 5% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.05 ± 3% -0.2 0.88 ± 8% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__free_huge_page.release_pages.tlb_flush_mmu 1.05 ± 3% -0.2 0.88 ± 8% perf-profile.calltrace.cycles-pp._raw_spin_lock.__free_huge_page.release_pages.tlb_flush_mmu.tlb_finish_mmu 0.78 ± 7% -0.1 0.66 ± 4% perf-profile.calltrace.cycles-pp.secondary_startup_64 0.74 ± 8% -0.1 0.62 ± 6% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary 0.75 ± 8% -0.1 0.63 ± 6% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64 0.75 ± 8% -0.1 0.63 ± 6% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 0.75 ± 8% -0.1 0.63 ± 6% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64 0.74 ± 8% -0.1 0.62 ± 6% perf-profile.calltrace.cycles-pp.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 0.00 +0.9 0.95 perf-profile.calltrace.cycles-pp.mutex_spin_on_owner.__mutex_lock.hugetlb_fault.handle_mm_fault.do_page_fault 0.00 +1.0 0.97 perf-profile.calltrace.cycles-pp.__mutex_lock.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault 27.03 ± 2% +5.1 32.11 ± 4% perf-profile.calltrace.cycles-pp.hugetlb_no_page.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault 26.48 ± 2% +5.3 31.80 ± 4% perf-profile.calltrace.cycles-pp.clear_huge_page.hugetlb_no_page.hugetlb_fault.handle_mm_fault.do_page_fault 25.54 +5.8 31.33 ± 2% perf-profile.calltrace.cycles-pp.page_fault.do_access 25.48 +5.8 31.28 ± 2% perf-profile.calltrace.cycles-pp.do_page_fault.page_fault.do_access 25.41 +5.8 31.22 ± 2% perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault.do_access 25.38 +5.8 31.19 ± 2% perf-profile.calltrace.cycles-pp.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault.do_access 89.02 +9.2 98.24 ± 2% perf-profile.calltrace.cycles-pp.do_access 25.46 ± 2% -25.5 0.00 perf-profile.children.cycles-pp.clear_subpage 24.72 ± 2% -24.7 0.00 perf-profile.children.cycles-pp.clear_page_erms 5.90 ± 2% -1.0 4.92 ± 2% perf-profile.children.cycles-pp.do_syscall_64 5.90 ± 2% -1.0 4.93 ± 3% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 0.73 ± 3% -0.7 0.06 ± 7% perf-profile.children.cycles-pp.___might_sleep 3.91 ± 3% -0.6 3.29 ± 4% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.children.cycles-pp.mmap64 3.17 ± 5% -0.5 2.64 ± 5% perf-profile.children.cycles-pp.alloc_fresh_huge_page 3.16 ± 5% -0.5 2.62 ± 5% perf-profile.children.cycles-pp.get_page_from_freelist 3.21 ± 4% -0.5 2.68 ± 5% perf-profile.children.cycles-pp.vm_mmap_pgoff 3.18 ± 5% -0.5 2.65 ± 5% perf-profile.children.cycles-pp.alloc_surplus_huge_page 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.children.cycles-pp.hugetlbfs_file_mmap 3.20 ± 4% -0.5 2.67 ± 5% perf-profile.children.cycles-pp.hugetlb_reserve_pages 3.20 ± 4% -0.5 2.68 ± 5% perf-profile.children.cycles-pp.ksys_mmap_pgoff 3.20 ± 4% -0.5 2.68 ± 5% perf-profile.children.cycles-pp.do_mmap 3.20 ± 4% -0.5 2.68 ± 5% perf-profile.children.cycles-pp.mmap_region 3.16 ± 5% -0.5 2.63 ± 5% perf-profile.children.cycles-pp.__alloc_pages_nodemask 3.20 ± 4% -0.5 2.68 ± 5% perf-profile.children.cycles-pp.hugetlb_acct_memory 2.84 ± 4% -0.5 2.35 ± 5% perf-profile.children.cycles-pp._raw_spin_lock_irqsave 1.51 ± 8% -0.3 1.18 ± 3% perf-profile.children.cycles-pp.kthread 1.49 ± 8% -0.3 1.17 ± 3% perf-profile.children.cycles-pp.memcpy_erms 1.49 ± 8% -0.3 1.17 ± 3% perf-profile.children.cycles-pp.drm_fb_helper_dirty_work 1.51 ± 8% -0.3 1.19 ± 3% perf-profile.children.cycles-pp.ret_from_fork 1.50 ± 8% -0.3 1.18 ± 3% perf-profile.children.cycles-pp.process_one_work 1.50 ± 8% -0.3 1.18 ± 3% perf-profile.children.cycles-pp.worker_thread 1.18 ± 4% -0.3 0.89 ± 2% perf-profile.children.cycles-pp.console_unlock 1.01 ± 5% -0.3 0.75 perf-profile.children.cycles-pp.serial8250_console_write 1.19 ± 4% -0.3 0.92 ± 3% perf-profile.children.cycles-pp.vprintk_emit 1.08 ± 4% -0.3 0.83 ± 6% perf-profile.children.cycles-pp.write 1.09 ± 5% -0.3 0.84 ± 7% perf-profile.children.cycles-pp.new_sync_write 1.10 ± 5% -0.3 0.85 ± 7% perf-profile.children.cycles-pp.ksys_write 1.10 ± 5% -0.3 0.85 ± 7% perf-profile.children.cycles-pp.vfs_write 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.children.cycles-pp.devkmsg_write 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.children.cycles-pp._fini 1.06 ± 5% -0.3 0.80 ± 6% perf-profile.children.cycles-pp.devkmsg_emit 0.94 ± 6% -0.2 0.69 perf-profile.children.cycles-pp.uart_console_write 0.81 ± 5% -0.2 0.60 ± 2% perf-profile.children.cycles-pp.wait_for_xmitr 0.79 ± 5% -0.2 0.58 ± 2% perf-profile.children.cycles-pp.io_serial_in 0.75 ± 5% -0.2 0.55 perf-profile.children.cycles-pp.serial8250_console_putchar 1.43 ± 2% -0.2 1.25 ± 6% perf-profile.children.cycles-pp.__x64_sys_exit_group 1.43 ± 2% -0.2 1.25 ± 6% perf-profile.children.cycles-pp.do_group_exit 1.43 ± 2% -0.2 1.25 ± 6% perf-profile.children.cycles-pp.do_exit 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.children.cycles-pp.tlb_finish_mmu 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.children.cycles-pp.tlb_flush_mmu 1.42 ± 2% -0.2 1.25 ± 6% perf-profile.children.cycles-pp.mmput 1.42 ± 2% -0.2 1.25 ± 6% perf-profile.children.cycles-pp.exit_mmap 1.39 ± 2% -0.2 1.21 ± 6% perf-profile.children.cycles-pp.release_pages 1.38 ± 2% -0.2 1.20 ± 6% perf-profile.children.cycles-pp.__free_huge_page 1.19 ± 3% -0.2 1.02 ± 7% perf-profile.children.cycles-pp._raw_spin_lock 0.78 ± 7% -0.1 0.65 ± 4% perf-profile.children.cycles-pp.cpuidle_enter 0.78 ± 7% -0.1 0.65 ± 4% perf-profile.children.cycles-pp.cpuidle_enter_state 0.78 ± 7% -0.1 0.66 ± 4% perf-profile.children.cycles-pp.secondary_startup_64 0.78 ± 7% -0.1 0.66 ± 4% perf-profile.children.cycles-pp.cpu_startup_entry 0.78 ± 7% -0.1 0.66 ± 4% perf-profile.children.cycles-pp.do_idle 0.75 ± 8% -0.1 0.63 ± 6% perf-profile.children.cycles-pp.start_secondary 0.74 ± 7% -0.1 0.63 ± 4% perf-profile.children.cycles-pp.intel_idle 0.20 ± 8% -0.1 0.15 ± 7% perf-profile.children.cycles-pp.io_serial_out 0.48 ± 5% -0.1 0.43 ± 5% perf-profile.children.cycles-pp.tick_sched_handle 0.46 ± 6% -0.0 0.42 ± 6% perf-profile.children.cycles-pp.update_process_times 0.15 ± 7% -0.0 0.12 ± 5% perf-profile.children.cycles-pp.alloc_huge_page 0.08 ± 5% -0.0 0.07 perf-profile.children.cycles-pp.update_curr 0.43 ± 4% +0.6 1.01 ± 8% perf-profile.children.cycles-pp.mutex_spin_on_owner 0.44 ± 5% +0.6 1.04 ± 7% perf-profile.children.cycles-pp.__mutex_lock 27.03 ± 2% +5.1 32.17 ± 4% perf-profile.children.cycles-pp.hugetlb_no_page 26.78 ± 2% +5.2 31.95 ± 4% perf-profile.children.cycles-pp.clear_huge_page 27.75 ± 2% +5.7 33.48 ± 4% perf-profile.children.cycles-pp.page_fault 27.68 ± 2% +5.7 33.42 ± 4% perf-profile.children.cycles-pp.do_page_fault 27.61 ± 2% +5.7 33.35 ± 4% perf-profile.children.cycles-pp.handle_mm_fault 27.54 ± 2% +5.7 33.28 ± 4% perf-profile.children.cycles-pp.hugetlb_fault 64.78 +5.9 70.72 ± 2% perf-profile.children.cycles-pp.do_access 24.19 ± 2% -24.2 0.00 perf-profile.self.cycles-pp.clear_page_erms 13.76 -3.1 10.71 ± 2% perf-profile.self.cycles-pp.do_access 3.91 ± 3% -0.6 3.28 ± 4% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath 0.46 ± 4% -0.4 0.04 ± 58% perf-profile.self.cycles-pp.___might_sleep 1.47 ± 8% -0.3 1.16 ± 3% perf-profile.self.cycles-pp.memcpy_erms 0.78 ± 4% -0.2 0.57 ± 2% perf-profile.self.cycles-pp.io_serial_in 0.74 ± 7% -0.1 0.63 ± 4% perf-profile.self.cycles-pp.intel_idle 0.30 ± 6% -0.1 0.24 ± 3% perf-profile.self.cycles-pp.get_page_from_freelist 0.20 ± 8% -0.1 0.15 ± 7% perf-profile.self.cycles-pp.io_serial_out 0.43 ± 5% +0.6 1.00 ± 8% perf-profile.self.cycles-pp.mutex_spin_on_owner 0.32 ± 2% +31.2 31.57 ± 4% perf-profile.self.cycles-pp.clear_huge_page vm-scalability.time.user_time 3050 +--------------------------------------------------------------------+ | | 3000 |-+ .+.. .+.. .+.. .+. .+.. .+..+.. .+.. .+..| |..+.+..+..+ +..+ +..+ +. +..+ +..+ + +..+ | | | 2950 |-+ | | | 2900 |-+ | | | 2850 |-+ | | | | O O O O | 2800 |-+O O O O O O O O O O O O O O O O O O O O | | | 2750 +--------------------------------------------------------------------+ vm-scalability.time.system_time 1750 +--------------------------------------------------------------------+ | | 1700 |-+O O O O O O O O O O O O O O O | | O O O O O O O O O | 1650 |-+ | | | 1600 |-+ | | | 1550 |-+ | | | 1500 |-+ | | | 1450 |..+.+..+..+. .+..+. .+.+..+.+..+..+. .+. .+..| | +. +..+..+.+..+. +..+. +..+..+ | 1400 +--------------------------------------------------------------------+ vm-scalability.time.percent_of_cpu_this_job_got 1490 +--------------------------------------------------------------------+ | O O O O O O O | | O O O O O O O O O O O O O | 1485 |-+ O O | | | | | 1480 |-+ | | | 1475 |-+ | | | | | 1470 |-+ .+ +. .+.. .+. + | |.. .+..+. + .+ .+.. .. +..+ +..+. .+. +.. : + | | + +..+. + .+ + +. +.. : +| 1465 +--------------------------------------------------------------------+ vm-scalability.time.minor_page_faults 2.1e+06 +----------------------------------------------------------------+ |..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+.. .+..+.| 2.05e+06 |-+ + | | | 2e+06 |-+ | 1.95e+06 |-+ | | | 1.9e+06 |-+ | | | 1.85e+06 |-+ | 1.8e+06 |-+ | | | 1.75e+06 |-+ O | | O O O O O O O O O O O O O O O O O O O O O O O | 1.7e+06 +----------------------------------------------------------------+ vm-scalability.time.voluntary_context_switches 17000 +-------------------------------------------------------------------+ | +.. .+..+.. .+. .+..+.+..+.+..+..+. .+.. | 16500 |..+.+..+. .. + +.+. +. +. +.+..+.+..| | + | | | 16000 |-+ | | | 15500 |-+ | | | 15000 |-+ O | | O O O | | O O O O O O O O | 14500 |-+O O O O O O O O O O O O | | | 14000 +-------------------------------------------------------------------+ vm-scalability.throughput 1.45e+07 +----------------------------------------------------------------+ | | 1.4e+07 |..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.+..+.| | | | | 1.35e+07 |-+ | | | 1.3e+07 |-+ | | | 1.25e+07 |-+ | | | | | 1.2e+07 |-+ | | O O O O O O O O O O O O O O O O O O O O O O O O | 1.15e+07 +----------------------------------------------------------------+ vm-scalability.median 900000 +------------------------------------------------------------------+ |..+.+..+.+..+.+..+..+.+..+.+..+..+.+..+.+..+.+..+..+.+..+.+..+.+..| 880000 |-+ | 860000 |-+ | | | 840000 |-+ | 820000 |-+ | | | 800000 |-+ | 780000 |-+ | | | 760000 |-+ | 740000 |-+ | | O O O O O O O O O O O O O O O O O O O O O O O O | 720000 +------------------------------------------------------------------+ vm-scalability.workload 4e+09 +-----------------------------------------------------------------+ |..+.+..+.+..+.+..+.+..+.+..+.+..+..+.+..+.+..+.+..+.+..+.+..+.+..| 3.9e+09 |-+ | | | 3.8e+09 |-+ | | | 3.7e+09 |-+ | | | 3.6e+09 |-+ | | | 3.5e+09 |-+ | | | 3.4e+09 |-+ | | O | 3.3e+09 +-----------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample *************************************************************************************************** lkp-ivb-d02: 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/process/16/debian-x86_64-20191114.cgz/lkp-ivb-d02/page_fault1/will-it-scale/0x21 commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 1:4 0% 1:4 perf-profile.children.cycles-pp.error_entry %stddev %change %stddev \ | \ 238214 -4.9% 226603 will-it-scale.per_process_ops 3811432 -4.9% 3625669 will-it-scale.workload 25486 ± 8% +16.8% 29758 ± 7% softirqs.CPU3.RCU 20.96 ± 2% +5.5% 22.11 ± 2% boot-time.boot 55.42 ± 2% +6.0% 58.77 ± 2% boot-time.idle 4186 ± 16% +29.2% 5409 ± 12% interrupts.CPU2.RES:Rescheduling_interrupts 5630 ± 7% -14.9% 4788 ± 12% interrupts.CPU3.RES:Rescheduling_interrupts 98.92 ± 25% -27.7% 71.56 ± 18% sched_debug.cfs_rq:/.load_avg.stddev 515190 ± 15% +21.8% 627617 ± 8% sched_debug.cpu.avg_idle.avg 11670 ± 6% +17.0% 13659 sched_debug.cpu.ttwu_count.min 9696 ± 7% +13.9% 11046 ± 4% sched_debug.cpu.ttwu_local.min 313013 +1.0% 316254 proc-vmstat.nr_active_anon 309576 +1.1% 313054 proc-vmstat.nr_anon_pages 313008 +1.0% 316246 proc-vmstat.nr_zone_active_anon 22747924 -4.8% 21653355 proc-vmstat.numa_hit 22747924 -4.8% 21653355 proc-vmstat.numa_local 1.15e+09 -4.9% 1.093e+09 proc-vmstat.pgalloc_normal 20511827 -4.9% 19515779 proc-vmstat.pgfault 1.15e+09 -5.0% 1.093e+09 proc-vmstat.pgfree 2206073 -4.9% 2097199 proc-vmstat.thp_fault_alloc 82.50 -82.5 0.00 perf-profile.calltrace.cycles-pp.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 80.56 -80.6 0.00 perf-profile.calltrace.cycles-pp.clear_page_erms.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault 3.16 -0.3 2.87 ± 3% perf-profile.calltrace.cycles-pp.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 3.10 -0.3 2.81 ± 3% perf-profile.calltrace.cycles-pp.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault 3.12 -0.3 2.84 ± 3% perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 93.63 +0.3 93.98 perf-profile.calltrace.cycles-pp.page_fault 93.45 +0.4 93.81 perf-profile.calltrace.cycles-pp.do_page_fault.page_fault 92.88 +0.4 93.31 perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault 92.65 +0.5 93.12 perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault 85.28 +0.6 85.83 perf-profile.calltrace.cycles-pp.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 89.69 +0.7 90.36 perf-profile.calltrace.cycles-pp.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault 82.68 -82.7 0.00 perf-profile.children.cycles-pp.clear_subpage 82.08 -80.9 1.19 perf-profile.children.cycles-pp.clear_page_erms 0.81 -0.6 0.20 ± 4% perf-profile.children.cycles-pp.rcu_all_qs 5.07 -0.4 4.64 perf-profile.children.cycles-pp.get_page_from_freelist 5.20 -0.4 4.79 perf-profile.children.cycles-pp.__alloc_pages_nodemask 4.92 -0.4 4.54 ± 2% perf-profile.children.cycles-pp.alloc_pages_vma 1.08 -0.1 0.96 ± 4% perf-profile.children.cycles-pp._cond_resched 0.40 ± 11% -0.1 0.31 ± 4% perf-profile.children.cycles-pp.mem_cgroup_try_charge_delay 1.90 -0.1 1.81 perf-profile.children.cycles-pp.prep_new_page 0.47 ± 4% -0.1 0.41 ± 6% perf-profile.children.cycles-pp.unmap_vmas 0.46 ± 4% -0.1 0.41 ± 6% perf-profile.children.cycles-pp.unmap_page_range 0.38 ± 8% -0.0 0.33 ± 6% perf-profile.children.cycles-pp.pte_alloc_one 0.10 ± 7% -0.0 0.07 ± 7% perf-profile.children.cycles-pp.mem_cgroup_throttle_swaprate 0.16 ± 5% -0.0 0.13 ± 19% perf-profile.children.cycles-pp._raw_spin_lock 0.13 ± 6% -0.0 0.10 ± 11% perf-profile.children.cycles-pp.zap_huge_pmd 93.72 +0.3 94.05 perf-profile.children.cycles-pp.page_fault 93.53 +0.3 93.88 perf-profile.children.cycles-pp.do_page_fault 92.94 +0.4 93.36 perf-profile.children.cycles-pp.handle_mm_fault 92.72 +0.5 93.18 perf-profile.children.cycles-pp.__handle_mm_fault 89.69 +0.7 90.36 perf-profile.children.cycles-pp.do_huge_pmd_anonymous_page 85.41 +1.1 86.54 perf-profile.children.cycles-pp.clear_huge_page 81.23 -80.1 1.17 ± 2% perf-profile.self.cycles-pp.clear_page_erms 0.60 ± 3% -0.4 0.15 ± 2% perf-profile.self.cycles-pp.rcu_all_qs 1.56 ± 5% -0.4 1.12 ± 4% perf-profile.self.cycles-pp.___might_sleep 2.44 -0.3 2.18 ± 4% perf-profile.self.cycles-pp.get_page_from_freelist 2.07 ± 4% -0.1 1.96 perf-profile.self.cycles-pp.__free_pages_ok 0.08 ± 6% -0.0 0.05 perf-profile.self.cycles-pp.mem_cgroup_throttle_swaprate 0.31 ± 8% +0.1 0.46 ± 2% perf-profile.self.cycles-pp._cond_resched 0.21 ± 8% +83.9 84.10 perf-profile.self.cycles-pp.clear_huge_page *************************************************************************************************** lkp-ivb-d02: 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory ========================================================================================= compiler/cpufreq_governor/kconfig/rootfs/runtime/size/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/debian-x86_64-20191114.cgz/300s/8T/lkp-ivb-d02/anon-w-seq-mt/vm-scalability/0x21 commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- %stddev %change %stddev \ | \ 0.00 +1.8% 0.00 vm-scalability.free_time 1148096 +8.4% 1244718 vm-scalability.median 4588586 +7.8% 4948645 vm-scalability.throughput 1475456 +7.5% 1586296 vm-scalability.time.minor_page_faults 270.13 -15.6% 227.88 ± 2% vm-scalability.time.system_time 868.34 +4.7% 908.82 vm-scalability.time.user_time 1.326e+09 +7.8% 1.43e+09 vm-scalability.workload 0.01 ± 37% +0.0 0.02 ± 76% mpstat.cpu.all.soft% 23.34 -3.5 19.88 ± 2% mpstat.cpu.all.sys% 73.00 +5.1% 76.75 vmstat.cpu.us 6969 +5.2% 7329 ± 2% vmstat.system.cs 627528 ± 2% -16.5% 524174 ± 4% sched_debug.cpu.avg_idle.avg -12.75 +25.5% -16.00 sched_debug.cpu.nr_uninterruptible.min 10.19 ± 6% +24.2% 12.65 ± 18% sched_debug.cpu.nr_uninterruptible.stddev 19223068 ± 15% -21.8% 15037169 ± 10% cpuidle.C6.time 33333 ± 3% -16.3% 27901 ± 6% cpuidle.C6.usage 8465 ± 39% -58.8% 3487 ± 15% cpuidle.POLL.time 2649 ± 23% -54.6% 1203 ± 21% cpuidle.POLL.usage 26729 ± 7% +78.6% 47740 ± 39% softirqs.CPU0.RCU 4574 ± 15% +79.7% 8219 ± 55% softirqs.CPU1.NET_RX 28386 ± 10% +63.3% 46347 ± 40% softirqs.CPU1.RCU 27498 ± 6% +75.2% 48178 ± 40% softirqs.CPU2.RCU 27869 ± 5% +73.8% 48433 ± 39% softirqs.CPU3.RCU 4577 ± 15% +79.6% 8222 ± 55% softirqs.NET_RX 110484 ± 4% +72.6% 190699 ± 39% softirqs.RCU 3455928 ± 2% -36.0% 2212394 meminfo.Active 3452870 ± 2% -36.0% 2209318 meminfo.Active(anon) 3281881 ± 2% -37.5% 2049676 meminfo.AnonHugePages 3425083 ± 2% -35.9% 2195653 meminfo.AnonPages 52607 ± 16% +165.2% 139515 ± 2% meminfo.CmaFree 3259754 ± 2% +38.2% 4504331 meminfo.MemAvailable 3368788 ± 2% +36.9% 4613377 meminfo.MemFree 4662470 ± 2% -26.7% 3417882 meminfo.Memused 10268 -24.1% 7794 meminfo.PageTables 8448 ± 16% +82.5% 15416 ± 55% interrupts.27:PCI-MSI.1572864-edge.eth0 4211 -13.1% 3661 interrupts.CAL:Function_call_interrupts 1058 -14.5% 904.50 ± 2% interrupts.CPU0.CAL:Function_call_interrupts 8448 ± 16% +82.5% 15416 ± 55% interrupts.CPU1.27:PCI-MSI.1572864-edge.eth0 1042 ± 4% -14.6% 889.75 ± 3% interrupts.CPU1.CAL:Function_call_interrupts 6167 +52.0% 9373 ± 29% interrupts.CPU1.NMI:Non-maskable_interrupts 6167 +52.0% 9373 ± 29% interrupts.CPU1.PMI:Performance_monitoring_interrupts 1076 -10.1% 968.25 ± 2% interrupts.CPU2.CAL:Function_call_interrupts 1034 ± 2% -13.1% 898.75 ± 4% interrupts.CPU3.CAL:Function_call_interrupts 6172 +53.3% 9464 ± 32% interrupts.CPU3.NMI:Non-maskable_interrupts 6172 +53.3% 9464 ± 32% interrupts.CPU3.PMI:Performance_monitoring_interrupts 825701 ± 3% -33.6% 548074 proc-vmstat.nr_active_anon 810670 ± 3% -32.9% 543797 proc-vmstat.nr_anon_pages 1516 ± 3% -34.5% 992.25 proc-vmstat.nr_anon_transparent_hugepages 85789 ± 3% +30.6% 112073 proc-vmstat.nr_dirty_background_threshold 171789 ± 3% +30.6% 224421 proc-vmstat.nr_dirty_threshold 15623 ± 19% +127.4% 35525 proc-vmstat.nr_free_cma 879784 ± 3% +31.6% 1157608 proc-vmstat.nr_free_pages 3558 -0.7% 3535 proc-vmstat.nr_kernel_stack 2481 ± 2% -21.7% 1944 proc-vmstat.nr_page_table_pages 825696 ± 3% -33.6% 548054 proc-vmstat.nr_zone_active_anon 2326268 +6.5% 2476331 proc-vmstat.numa_hit 2326268 +6.5% 2476331 proc-vmstat.numa_local 72116983 +5.5% 76117229 ± 3% proc-vmstat.pgalloc_dma32 2.237e+08 +8.6% 2.429e+08 proc-vmstat.pgalloc_normal 1859214 +5.9% 1969537 proc-vmstat.pgfault 2.948e+08 +8.1% 3.186e+08 proc-vmstat.pgfree 574297 +7.8% 619284 proc-vmstat.thp_fault_alloc 21.56 ± 7% -21.6 0.00 perf-profile.calltrace.cycles-pp.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 20.80 ± 7% -20.8 0.00 perf-profile.calltrace.cycles-pp.clear_page_erms.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault 23.47 ± 7% -5.2 18.26 ± 2% perf-profile.calltrace.cycles-pp.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault 22.35 ± 7% -5.2 17.18 ± 2% perf-profile.calltrace.cycles-pp.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 22.65 ± 3% -4.5 18.18 perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault.do_access 22.74 ± 3% -4.5 18.27 perf-profile.calltrace.cycles-pp.do_page_fault.page_fault.do_access 22.61 ± 3% -4.5 18.14 perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault.do_access 22.77 ± 3% -4.5 18.31 perf-profile.calltrace.cycles-pp.page_fault.do_access 0.73 ± 8% -0.1 0.63 ± 5% perf-profile.calltrace.cycles-pp.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 0.72 ± 8% -0.1 0.62 ± 5% perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 0.70 ± 9% -0.1 0.61 ± 5% perf-profile.calltrace.cycles-pp.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault 62.06 ± 2% +2.1 64.16 perf-profile.calltrace.cycles-pp.do_rw_once 21.78 ± 7% -21.8 0.00 perf-profile.children.cycles-pp.clear_subpage 21.35 ± 7% -21.3 0.06 ± 16% perf-profile.children.cycles-pp.clear_page_erms 23.62 ± 7% -5.2 18.45 ± 2% perf-profile.children.cycles-pp.do_huge_pmd_anonymous_page 23.93 ± 6% -5.1 18.82 ± 2% perf-profile.children.cycles-pp.do_page_fault 23.82 ± 6% -5.1 18.73 ± 2% perf-profile.children.cycles-pp.handle_mm_fault 23.98 ± 6% -5.1 18.89 ± 2% perf-profile.children.cycles-pp.page_fault 23.77 ± 7% -5.1 18.69 ± 2% perf-profile.children.cycles-pp.__handle_mm_fault 22.54 ± 7% -5.0 17.52 perf-profile.children.cycles-pp.clear_huge_page 48.63 -1.0 47.68 perf-profile.children.cycles-pp.do_access 1.40 ± 8% -0.4 1.03 ± 19% perf-profile.children.cycles-pp.apic_timer_interrupt 1.30 ± 7% -0.4 0.95 ± 21% perf-profile.children.cycles-pp.smp_apic_timer_interrupt 1.09 ± 9% -0.3 0.75 ± 22% perf-profile.children.cycles-pp.hrtimer_interrupt 0.91 ± 11% -0.3 0.64 ± 22% perf-profile.children.cycles-pp.__hrtimer_run_queues 0.67 ± 9% -0.2 0.46 ± 20% perf-profile.children.cycles-pp.tick_sched_timer 0.58 ± 10% -0.2 0.39 ± 18% perf-profile.children.cycles-pp.tick_sched_handle 0.56 ± 10% -0.2 0.37 ± 20% perf-profile.children.cycles-pp.update_process_times 0.21 ± 6% -0.1 0.07 ± 17% perf-profile.children.cycles-pp.rcu_all_qs 0.34 ± 6% -0.1 0.23 ± 12% perf-profile.children.cycles-pp.scheduler_tick 0.46 ± 13% -0.1 0.35 ± 8% perf-profile.children.cycles-pp.___might_sleep 0.80 ± 7% -0.1 0.71 ± 6% perf-profile.children.cycles-pp.alloc_pages_vma 0.25 ± 8% -0.1 0.17 ± 14% perf-profile.children.cycles-pp.task_tick_fair 0.28 ± 9% -0.1 0.21 ± 5% perf-profile.children.cycles-pp._cond_resched 0.15 ± 11% -0.0 0.11 ± 22% perf-profile.children.cycles-pp.swapgs_restore_regs_and_return_to_usermode 0.08 ± 8% +0.0 0.12 ± 10% perf-profile.children.cycles-pp.__do_execve_file 0.08 ± 8% +0.0 0.12 ± 12% perf-profile.children.cycles-pp.__x64_sys_execve 0.08 ± 10% +0.0 0.12 ± 8% perf-profile.children.cycles-pp.execve 0.03 ±102% +0.0 0.08 ± 19% perf-profile.children.cycles-pp.search_binary_handler 0.01 ±173% +0.1 0.06 ± 6% perf-profile.children.cycles-pp.drm_fb_helper_dirty 0.11 ± 33% +0.1 0.18 ± 16% perf-profile.children.cycles-pp.handle_pte_fault 1.27 ± 7% +0.2 1.47 ± 4% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 1.27 ± 7% +0.2 1.47 ± 4% perf-profile.children.cycles-pp.do_syscall_64 63.21 +3.4 66.66 perf-profile.children.cycles-pp.do_rw_once 20.97 ± 6% -20.9 0.06 ± 16% perf-profile.self.cycles-pp.clear_page_erms 0.44 ± 16% -0.2 0.28 ± 11% perf-profile.self.cycles-pp.___might_sleep 0.16 ± 13% -0.1 0.05 ± 58% perf-profile.self.cycles-pp.rcu_all_qs 0.08 ± 23% -0.1 0.03 ±100% perf-profile.self.cycles-pp.task_tick_fair 0.06 ± 14% +0.0 0.08 ± 10% perf-profile.self.cycles-pp._raw_spin_lock_irqsave 17.35 ± 5% +1.8 19.11 perf-profile.self.cycles-pp.do_access 55.41 +3.1 58.55 perf-profile.self.cycles-pp.do_rw_once 0.09 ± 27% +16.8 16.94 ± 2% perf-profile.self.cycles-pp.clear_huge_page *************************************************************************************************** lkp-csl-2sp6: 96 threads Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz with 256G memory ========================================================================================= compiler/cpufreq_governor/kconfig/rootfs/runtime/size/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/debian-x86_64-20191114.cgz/300s/8T/lkp-csl-2sp6/anon-cow-seq-hugetlb/vm-scalability/0x500002c commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :4 75% 3:4 dmesg.WARNING:at#for_ip_swapgs_restore_regs_and_return_to_usermode/0x :4 75% 3:4 dmesg.WARNING:stack_recursion %stddev %change %stddev \ | \ 361704 -2.6% 352245 vm-scalability.median 36376079 -4.6% 34694751 vm-scalability.throughput 4265227 -5.4% 4036714 vm-scalability.time.minor_page_faults 13472 -2.7% 13115 vm-scalability.time.system_time 8845 +2.1% 9035 vm-scalability.time.user_time 354016 -5.3% 335133 vm-scalability.time.voluntary_context_switches 9.105e+09 -5.4% 8.616e+09 vm-scalability.workload 4848 -2.0% 4751 vmstat.system.cs 2477460 ± 3% -8.2% 2274167 ± 4% numa-numastat.node0.local_node 2508460 ± 3% -8.3% 2301032 ± 4% numa-numastat.node0.numa_hit 1386 ± 4% -7.1% 1288 ± 5% slabinfo.Acpi-Parse.active_objs 1386 ± 4% -7.1% 1288 ± 5% slabinfo.Acpi-Parse.num_objs 3371 ± 11% +26.7% 4271 ± 10% slabinfo.eventpoll_pwq.active_objs 3371 ± 11% +26.7% 4271 ± 10% slabinfo.eventpoll_pwq.num_objs 812.33 ± 9% -15.8% 684.33 ± 4% slabinfo.kmem_cache_node.active_objs 852.33 ± 9% -14.9% 725.00 ± 4% slabinfo.kmem_cache_node.num_objs 3993284 -5.4% 3778602 proc-vmstat.htlb_buddy_alloc_success 179.00 -3.5% 172.67 proc-vmstat.nr_dirtied 45924 -1.1% 45436 proc-vmstat.nr_shmem 166.00 -3.6% 160.00 proc-vmstat.nr_written 5075790 -4.8% 4832494 proc-vmstat.numa_hit 5044600 -4.8% 4801279 proc-vmstat.numa_local 2.046e+09 -5.4% 1.936e+09 proc-vmstat.pgalloc_normal 5095148 -4.6% 4860017 proc-vmstat.pgfault 2.045e+09 -5.3% 1.936e+09 proc-vmstat.pgfree 1299 ± 71% +185.8% 3715 ± 35% sched_debug.cfs_rq:/.load.min 0.12 ± 71% +181.8% 0.34 ± 35% sched_debug.cfs_rq:/.nr_running.min 0.29 ± 5% -33.3% 0.19 ± 11% sched_debug.cfs_rq:/.nr_running.stddev 1164 ± 6% +20.0% 1397 ± 4% sched_debug.cfs_rq:/.util_avg.max 244.31 ± 4% -12.5% 213.67 ± 2% sched_debug.cfs_rq:/.util_avg.stddev 0.12 ± 71% +181.8% 0.34 ± 35% sched_debug.cfs_rq:/.util_est_enqueued.min 8.71 ± 35% +62.1% 14.13 ± 20% sched_debug.cpu.clock.stddev 8.71 ± 35% +62.1% 14.13 ± 20% sched_debug.cpu.clock_task.stddev 2263 ± 13% -33.6% 1503 ± 12% sched_debug.cpu.curr->pid.stddev 1.18 +30.2% 1.53 ± 9% sched_debug.cpu.nr_running.max 0.12 ± 71% +236.4% 0.41 ± 16% sched_debug.cpu.nr_running.min 0.30 ± 2% -24.7% 0.23 ± 12% sched_debug.cpu.nr_running.stddev 0.06 ± 18% -75.4% 0.02 ± 40% sched_debug.cpu.nr_uninterruptible.avg -50.38 -19.7% -40.47 sched_debug.cpu.nr_uninterruptible.min 9247 ± 13% +20.9% 11175 ± 7% sched_debug.cpu.ttwu_count.max 13116 ± 9% +12.9% 14811 ± 6% softirqs.CPU34.SCHED 13634 ± 9% -14.8% 11609 ± 10% softirqs.CPU4.SCHED 123480 ± 6% -10.8% 110147 softirqs.CPU4.TIMER 123091 ± 6% -10.4% 110308 softirqs.CPU52.TIMER 12976 ± 13% -13.9% 11174 ± 14% softirqs.CPU53.SCHED 12600 ± 10% -12.2% 11058 ± 15% softirqs.CPU56.SCHED 13147 ± 13% -14.2% 11275 ± 13% softirqs.CPU57.SCHED 13284 ± 14% -18.0% 10898 ± 15% softirqs.CPU58.SCHED 121769 ± 6% -10.3% 109258 softirqs.CPU58.TIMER 13140 ± 12% -19.6% 10561 ± 15% softirqs.CPU60.SCHED 13017 ± 13% -13.4% 11274 ± 12% softirqs.CPU64.SCHED 12951 ± 12% -17.2% 10724 ± 14% softirqs.CPU68.SCHED 13327 ± 14% -15.8% 11216 ± 15% softirqs.CPU69.SCHED 12891 ± 14% +14.7% 14789 ± 7% softirqs.CPU84.SCHED 13383 ± 9% -10.6% 11968 ± 9% softirqs.CPU9.SCHED 39.20 -39.2 0.00 perf-profile.calltrace.cycles-pp.copy_subpage.copy_user_huge_page.hugetlb_cow.hugetlb_fault.handle_mm_fault 38.70 -38.7 0.00 perf-profile.calltrace.cycles-pp.copy_page.copy_subpage.copy_user_huge_page.hugetlb_cow.hugetlb_fault 83.46 -17.1 66.37 ± 9% perf-profile.calltrace.cycles-pp.do_access 52.26 -12.6 39.62 ± 9% perf-profile.calltrace.cycles-pp.page_fault.do_access 52.22 -12.6 39.59 ± 9% perf-profile.calltrace.cycles-pp.do_page_fault.page_fault.do_access 52.16 -12.6 39.54 ± 9% perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault.do_access 52.13 -12.6 39.52 ± 9% perf-profile.calltrace.cycles-pp.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault.do_access 14.42 ± 2% -3.6 10.85 ± 7% perf-profile.calltrace.cycles-pp.do_rw_once 0.69 ± 7% -0.3 0.37 ± 70% perf-profile.calltrace.cycles-pp.alloc_huge_page.hugetlb_cow.hugetlb_fault.handle_mm_fault.do_page_fault 0.00 +0.8 0.80 ± 24% perf-profile.calltrace.cycles-pp.apic_timer_interrupt.copy_page.copy_user_huge_page.hugetlb_cow.hugetlb_fault 39.37 +6.0 45.35 perf-profile.calltrace.cycles-pp.copy_user_huge_page.hugetlb_cow.hugetlb_fault.handle_mm_fault.do_page_fault 40.32 +6.1 46.44 perf-profile.calltrace.cycles-pp.hugetlb_cow.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault 6.40 ± 36% +18.0 24.41 ± 23% perf-profile.calltrace.cycles-pp.hugetlb_fault.handle_mm_fault.do_page_fault.page_fault 6.41 ± 36% +18.0 24.44 ± 23% perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault 6.42 ± 36% +18.0 24.47 ± 23% perf-profile.calltrace.cycles-pp.do_page_fault.page_fault 6.42 ± 36% +18.1 24.48 ± 23% perf-profile.calltrace.cycles-pp.page_fault 0.00 +44.6 44.64 perf-profile.calltrace.cycles-pp.copy_page.copy_user_huge_page.hugetlb_cow.hugetlb_fault.handle_mm_fault 39.21 -39.2 0.00 perf-profile.children.cycles-pp.copy_subpage 72.47 -16.0 56.47 ± 9% perf-profile.children.cycles-pp.do_access 25.92 ± 2% -4.8 21.11 ± 8% perf-profile.children.cycles-pp.do_rw_once 0.11 +0.0 0.13 ± 6% perf-profile.children.cycles-pp.native_irq_return_iret 0.07 ± 11% +0.0 0.10 ± 4% perf-profile.children.cycles-pp.native_write_msr 0.11 ± 4% +0.0 0.14 ± 3% perf-profile.children.cycles-pp.update_and_free_page 0.28 ± 5% +0.0 0.32 perf-profile.children.cycles-pp.process_one_work 0.27 ± 4% +0.0 0.31 perf-profile.children.cycles-pp.drm_fb_helper_dirty_work 0.37 ± 3% +0.0 0.41 ± 6% perf-profile.children.cycles-pp.task_tick_fair 0.29 ± 4% +0.0 0.33 perf-profile.children.cycles-pp.ret_from_fork 0.29 ± 4% +0.0 0.33 perf-profile.children.cycles-pp.kthread 0.28 ± 5% +0.0 0.32 perf-profile.children.cycles-pp.worker_thread 0.27 ± 4% +0.0 0.32 perf-profile.children.cycles-pp.memcpy_erms 0.09 ± 24% +0.0 0.13 ± 14% perf-profile.children.cycles-pp.perf_mux_hrtimer_handler 0.02 ±141% +0.1 0.07 ± 7% perf-profile.children.cycles-pp.__intel_pmu_enable_all 0.00 +0.1 0.05 perf-profile.children.cycles-pp.run_local_timers 0.02 ±141% +0.1 0.07 ± 23% perf-profile.children.cycles-pp.menu_select 0.58 ± 2% +0.1 0.66 ± 2% perf-profile.children.cycles-pp.__alloc_pages_nodemask 0.56 ± 2% +0.1 0.64 perf-profile.children.cycles-pp.get_page_from_freelist 0.62 ± 3% +0.1 0.72 perf-profile.children.cycles-pp.alloc_fresh_huge_page 0.05 ± 74% +0.1 0.14 ± 32% perf-profile.children.cycles-pp.start_kernel 1.02 ± 2% +0.1 1.13 ± 6% perf-profile.children.cycles-pp.__hrtimer_run_queues 0.71 +0.1 0.82 ± 3% perf-profile.children.cycles-pp.alloc_surplus_huge_page 0.84 +0.1 0.96 ± 3% perf-profile.children.cycles-pp.alloc_huge_page 1.26 ± 2% +0.1 1.40 ± 2% perf-profile.children.cycles-pp.hrtimer_interrupt 1.47 ± 2% +0.1 1.62 perf-profile.children.cycles-pp.smp_apic_timer_interrupt 1.91 +0.2 2.14 perf-profile.children.cycles-pp.apic_timer_interrupt 58.58 ± 3% +5.4 63.99 ± 4% perf-profile.children.cycles-pp.handle_mm_fault 58.60 ± 3% +5.4 64.02 ± 4% perf-profile.children.cycles-pp.hugetlb_fault 58.70 ± 3% +5.4 64.12 ± 4% perf-profile.children.cycles-pp.page_fault 58.65 ± 3% +5.4 64.08 ± 4% perf-profile.children.cycles-pp.do_page_fault 39.38 +6.0 45.36 perf-profile.children.cycles-pp.copy_user_huge_page 39.05 +6.0 45.07 perf-profile.children.cycles-pp.copy_page 40.33 +6.1 46.44 perf-profile.children.cycles-pp.hugetlb_cow 25.22 ± 2% -4.6 20.61 ± 8% perf-profile.self.cycles-pp.do_rw_once 8.36 -2.1 6.29 ± 8% perf-profile.self.cycles-pp.do_access 0.11 +0.0 0.13 ± 6% perf-profile.self.cycles-pp.native_irq_return_iret 0.07 ± 11% +0.0 0.10 ± 4% perf-profile.self.cycles-pp.native_write_msr 0.11 ± 4% +0.0 0.14 ± 3% perf-profile.self.cycles-pp.update_and_free_page 0.27 ± 5% +0.0 0.31 perf-profile.self.cycles-pp.memcpy_erms 0.00 +0.1 0.05 perf-profile.self.cycles-pp.run_local_timers 0.33 +0.1 0.40 perf-profile.self.cycles-pp.get_page_from_freelist 0.00 +0.1 0.08 ± 10% perf-profile.self.cycles-pp.clear_huge_page 0.00 +0.1 0.12 perf-profile.self.cycles-pp.copy_user_huge_page 38.43 +5.9 44.38 perf-profile.self.cycles-pp.copy_page 291.00 ± 9% -11.9% 256.33 ± 3% interrupts.35:PCI-MSI.31981568-edge.i40e-0000:3d:00.0:misc 297.00 ±116% -99.0% 3.00 ±141% interrupts.46:PCI-MSI.31981579-edge.i40e-eth0-TxRx-10 296.67 ±116% -99.1% 2.67 ±141% interrupts.CPU10.46:PCI-MSI.31981579-edge.i40e-eth0-TxRx-10 501.00 ± 3% +22.0% 611.00 ± 8% interrupts.CPU12.RES:Rescheduling_interrupts 6041 ± 7% -37.7% 3766 ± 20% interrupts.CPU2.NMI:Non-maskable_interrupts 6041 ± 7% -37.7% 3766 ± 20% interrupts.CPU2.PMI:Performance_monitoring_interrupts 5468 ± 8% +10.8% 6057 ± 10% interrupts.CPU21.NMI:Non-maskable_interrupts 5468 ± 8% +10.8% 6057 ± 10% interrupts.CPU21.PMI:Performance_monitoring_interrupts 1547 ± 63% +186.7% 4435 ± 17% interrupts.CPU24.TLB:TLB_shootdowns 585.67 -23.1% 450.33 ± 20% interrupts.CPU25.CAL:Function_call_interrupts 6241 ± 4% -23.6% 4767 ± 26% interrupts.CPU25.NMI:Non-maskable_interrupts 6241 ± 4% -23.6% 4767 ± 26% interrupts.CPU25.PMI:Performance_monitoring_interrupts 1089 ± 68% -56.9% 469.67 ± 2% interrupts.CPU26.RES:Rescheduling_interrupts 6424 -11.8% 5667 ± 4% interrupts.CPU28.NMI:Non-maskable_interrupts 6424 -11.8% 5667 ± 4% interrupts.CPU28.PMI:Performance_monitoring_interrupts 6076 ± 8% -22.9% 4684 ± 20% interrupts.CPU30.NMI:Non-maskable_interrupts 6076 ± 8% -22.9% 4684 ± 20% interrupts.CPU30.PMI:Performance_monitoring_interrupts 895.33 ± 73% +105.6% 1840 ± 3% interrupts.CPU30.TLB:TLB_shootdowns 6224 ± 9% -23.3% 4777 ± 26% interrupts.CPU31.NMI:Non-maskable_interrupts 6224 ± 9% -23.3% 4777 ± 26% interrupts.CPU31.PMI:Performance_monitoring_interrupts 529.00 ± 6% -27.2% 385.00 ± 26% interrupts.CPU38.RES:Rescheduling_interrupts 970.00 ± 69% +103.5% 1974 ± 9% interrupts.CPU38.TLB:TLB_shootdowns 4676 ± 26% +33.3% 6233 ± 10% interrupts.CPU4.NMI:Non-maskable_interrupts 4676 ± 26% +33.3% 6233 ± 10% interrupts.CPU4.PMI:Performance_monitoring_interrupts 518.67 ± 22% -25.4% 387.00 ± 25% interrupts.CPU40.RES:Rescheduling_interrupts 450.67 ± 20% -23.0% 347.00 ± 21% interrupts.CPU41.RES:Rescheduling_interrupts 6379 ± 7% -8.6% 5832 ± 7% interrupts.CPU42.NMI:Non-maskable_interrupts 6379 ± 7% -8.6% 5832 ± 7% interrupts.CPU42.PMI:Performance_monitoring_interrupts 509.67 ± 13% -34.7% 332.67 ± 22% interrupts.CPU42.RES:Rescheduling_interrupts 527.33 ± 9% -44.6% 292.00 ± 19% interrupts.CPU43.RES:Rescheduling_interrupts 545.33 ± 27% -41.1% 321.33 ± 10% interrupts.CPU45.RES:Rescheduling_interrupts 507.67 ± 25% -34.6% 332.00 ± 22% interrupts.CPU46.RES:Rescheduling_interrupts 435.00 ± 22% -21.5% 341.67 ± 6% interrupts.CPU48.RES:Rescheduling_interrupts 1142 ± 65% -72.2% 317.33 ± 26% interrupts.CPU56.RES:Rescheduling_interrupts 602.00 ± 64% -56.0% 265.00 ± 5% interrupts.CPU60.RES:Rescheduling_interrupts 282.00 ± 6% -16.2% 236.33 ± 2% interrupts.CPU61.RES:Rescheduling_interrupts 64.33 ± 67% +293.8% 253.33 ± 43% interrupts.CPU61.TLB:TLB_shootdowns 771.00 ± 62% -66.1% 261.33 ± 19% interrupts.CPU66.RES:Rescheduling_interrupts 81.00 ± 83% +135.0% 190.33 ± 27% interrupts.CPU66.TLB:TLB_shootdowns 4613 ± 21% +35.3% 6242 ± 10% interrupts.CPU68.NMI:Non-maskable_interrupts 4613 ± 21% +35.3% 6242 ± 10% interrupts.CPU68.PMI:Performance_monitoring_interrupts 4684 ± 35% +35.8% 6359 ± 8% interrupts.CPU69.NMI:Non-maskable_interrupts 4684 ± 35% +35.8% 6359 ± 8% interrupts.CPU69.PMI:Performance_monitoring_interrupts 291.00 ± 9% -11.9% 256.33 ± 3% interrupts.CPU7.35:PCI-MSI.31981568-edge.i40e-0000:3d:00.0:misc 102.00 ± 71% +105.2% 209.33 ± 26% interrupts.CPU70.TLB:TLB_shootdowns 245.00 ± 61% +156.5% 628.33 ± 5% interrupts.CPU72.TLB:TLB_shootdowns 140.67 ± 53% +183.9% 399.33 ± 47% interrupts.CPU74.TLB:TLB_shootdowns 6394 ± 2% -8.2% 5870 ± 6% interrupts.CPU76.NMI:Non-maskable_interrupts 6394 ± 2% -8.2% 5870 ± 6% interrupts.CPU76.PMI:Performance_monitoring_interrupts 173.33 ± 54% +251.9% 610.00 ± 32% interrupts.CPU76.TLB:TLB_shootdowns 74.00 ± 36% +544.6% 477.00 ± 14% interrupts.CPU77.TLB:TLB_shootdowns 317.00 ± 24% -26.4% 233.33 ± 20% interrupts.CPU79.RES:Rescheduling_interrupts 114.67 ±103% +126.7% 260.00 ± 23% interrupts.CPU79.TLB:TLB_shootdowns 152.00 ± 85% +110.3% 319.67 ± 11% interrupts.CPU82.TLB:TLB_shootdowns 284.67 -31.6% 194.67 ± 32% interrupts.CPU83.RES:Rescheduling_interrupts 267.67 ± 8% -19.7% 215.00 ± 17% interrupts.CPU84.RES:Rescheduling_interrupts 99.67 ±115% +134.1% 233.33 ± 40% interrupts.CPU84.TLB:TLB_shootdowns 76.00 ± 87% +200.4% 228.33 ± 32% interrupts.CPU86.TLB:TLB_shootdowns 6324 ± 6% -21.9% 4936 ± 29% interrupts.CPU87.NMI:Non-maskable_interrupts 6324 ± 6% -21.9% 4936 ± 29% interrupts.CPU87.PMI:Performance_monitoring_interrupts 38.67 ±112% +550.0% 251.33 ± 58% interrupts.CPU87.TLB:TLB_shootdowns 720.33 ± 66% -69.8% 217.67 ± 19% interrupts.CPU89.RES:Rescheduling_interrupts 309.67 ± 14% -35.5% 199.67 ± 19% interrupts.CPU90.RES:Rescheduling_interrupts 31.67 ± 70% +533.7% 200.67 ± 15% interrupts.CPU92.TLB:TLB_shootdowns 6189 ± 9% -24.8% 4656 ± 29% interrupts.CPU94.NMI:Non-maskable_interrupts 6189 ± 9% -24.8% 4656 ± 29% interrupts.CPU94.PMI:Performance_monitoring_interrupts 73.00 ± 65% +444.3% 397.33 ± 24% interrupts.CPU94.TLB:TLB_shootdowns 6313 ± 5% -9.7% 5703 ± 4% interrupts.CPU95.NMI:Non-maskable_interrupts 6313 ± 5% -9.7% 5703 ± 4% interrupts.CPU95.PMI:Performance_monitoring_interrupts 49044 ± 10% -15.0% 41667 ± 12% interrupts.RES:Rescheduling_interrupts *************************************************************************************************** lkp-ivb-d01: 8 threads Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz with 16G memory ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/thread/100%/debian-x86_64-20191114.cgz/lkp-ivb-d01/page_fault1/will-it-scale/0x21 commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 dmesg.RIP:cpuidle_enter_state :4 25% 1:4 dmesg.RIP:simple_getattr :4 25% 1:4 kmsg.b135d8f>]usb_hcd_irq 1:4 -25% :4 kmsg.c48>]usb_hcd_irq 1:4 -25% :4 kmsg.ca#>]usb_hcd_irq 1:4 -25% :4 kmsg.cb32>]usb_hcd_irq :4 25% 1:4 kmsg.d428ae1>]usb_hcd_irq 1:4 -25% :4 kmsg.ddef>]usb_hcd_irq %stddev %change %stddev \ | \ 534714 -3.5% 515893 will-it-scale.per_thread_ops 2517464 -3.5% 2428934 will-it-scale.time.minor_page_faults 518727 -4.8% 493620 will-it-scale.time.voluntary_context_switches 4277722 -3.5% 4127153 will-it-scale.workload 9481 -1.3% 9358 vmstat.system.cs 2764362 +196.4% 8193435 ± 28% cpuidle.C1.time 63195 ± 3% +126.3% 143040 ± 41% cpuidle.C1.usage 18199 ± 7% +12.8% 20532 ± 9% softirqs.CPU3.RCU 17571 ± 10% +21.0% 21255 ± 7% softirqs.CPU5.RCU 248287 -6.0% 233281 interrupts.CAL:Function_call_interrupts 2273 ± 17% -18.5% 1852 interrupts.CPU1.RES:Rescheduling_interrupts 1965 ± 7% -7.7% 1815 ± 3% interrupts.CPU2.RES:Rescheduling_interrupts 4.43 ± 7% -19.6% 3.56 ± 15% sched_debug.cfs_rq:/.nr_spread_over.avg 32.42 ± 12% -41.5% 18.96 ± 32% sched_debug.cfs_rq:/.nr_spread_over.max 10.63 ± 12% -39.0% 6.48 ± 25% sched_debug.cfs_rq:/.nr_spread_over.stddev 5397479 -3.2% 5222071 proc-vmstat.numa_hit 5397479 -3.2% 5222071 proc-vmstat.numa_local 1.289e+09 -3.5% 1.243e+09 proc-vmstat.pgalloc_normal 2902685 -2.9% 2817148 proc-vmstat.pgfault 1.289e+09 -3.5% 1.243e+09 proc-vmstat.pgfree 2511246 -3.5% 2422977 proc-vmstat.thp_fault_alloc 88.56 -88.6 0.00 perf-profile.calltrace.cycles-pp.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 86.61 -86.6 0.00 perf-profile.calltrace.cycles-pp.clear_page_erms.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault 90.95 -0.4 90.57 perf-profile.calltrace.cycles-pp.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 2.48 ± 3% -0.2 2.28 ± 2% perf-profile.calltrace.cycles-pp.__do_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.37 ± 3% -0.2 2.17 ± 2% perf-profile.calltrace.cycles-pp.release_pages.tlb_flush_mmu.tlb_finish_mmu.unmap_region.__do_munmap 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.munmap 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.calltrace.cycles-pp.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.calltrace.cycles-pp.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.munmap 2.46 ± 3% -0.2 2.25 ± 2% perf-profile.calltrace.cycles-pp.unmap_region.__do_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64 2.52 ± 3% -0.2 2.32 ± 2% perf-profile.calltrace.cycles-pp.munmap 2.42 ± 3% -0.2 2.22 ± 2% perf-profile.calltrace.cycles-pp.tlb_finish_mmu.unmap_region.__do_munmap.__vm_munmap.__x64_sys_munmap 2.42 ± 3% -0.2 2.22 ± 2% perf-profile.calltrace.cycles-pp.tlb_flush_mmu.tlb_finish_mmu.unmap_region.__do_munmap.__vm_munmap 2.25 ± 3% -0.2 2.04 ± 3% perf-profile.calltrace.cycles-pp.__free_pages_ok.release_pages.tlb_flush_mmu.tlb_finish_mmu.unmap_region 3.34 -0.1 3.22 perf-profile.calltrace.cycles-pp.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault 3.36 -0.1 3.25 perf-profile.calltrace.cycles-pp.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault 3.35 -0.1 3.24 perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.alloc_pages_vma.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 0.61 ± 3% +0.2 0.78 ± 2% perf-profile.calltrace.cycles-pp._cond_resched.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 95.16 +0.3 95.45 perf-profile.calltrace.cycles-pp.page_fault 94.98 +0.3 95.28 perf-profile.calltrace.cycles-pp.handle_mm_fault.do_page_fault.page_fault 95.12 +0.3 95.42 perf-profile.calltrace.cycles-pp.do_page_fault.page_fault 94.95 +0.3 95.26 perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault 94.89 +0.3 95.22 perf-profile.calltrace.cycles-pp.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault.do_page_fault.page_fault 88.69 -88.7 0.00 perf-profile.children.cycles-pp.clear_subpage 87.05 -86.9 0.12 ± 9% perf-profile.children.cycles-pp.clear_page_erms 0.88 ± 3% -0.8 0.10 ± 15% perf-profile.children.cycles-pp.rcu_all_qs 2.85 ± 3% -0.2 2.63 ± 3% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 2.85 ± 3% -0.2 2.63 ± 3% perf-profile.children.cycles-pp.do_syscall_64 2.49 ± 3% -0.2 2.28 ± 2% perf-profile.children.cycles-pp.__do_munmap 2.43 ± 3% -0.2 2.22 ± 2% perf-profile.children.cycles-pp.tlb_finish_mmu 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.children.cycles-pp.__x64_sys_munmap 2.46 ± 3% -0.2 2.26 ± 2% perf-profile.children.cycles-pp.unmap_region 2.43 ± 3% -0.2 2.22 ± 2% perf-profile.children.cycles-pp.tlb_flush_mmu 2.38 ± 3% -0.2 2.18 ± 2% perf-profile.children.cycles-pp.release_pages 2.52 ± 3% -0.2 2.32 ± 2% perf-profile.children.cycles-pp.munmap 2.52 ± 3% -0.2 2.31 ± 2% perf-profile.children.cycles-pp.__vm_munmap 2.25 ± 3% -0.2 2.05 ± 3% perf-profile.children.cycles-pp.__free_pages_ok 3.58 -0.1 3.43 perf-profile.children.cycles-pp.__alloc_pages_nodemask 3.55 -0.1 3.41 perf-profile.children.cycles-pp.get_page_from_freelist 0.95 ± 4% -0.1 0.84 ± 2% perf-profile.children.cycles-pp._cond_resched 3.37 -0.1 3.27 perf-profile.children.cycles-pp.alloc_pages_vma 0.68 -0.1 0.59 ± 2% perf-profile.children.cycles-pp.apic_timer_interrupt 0.33 ± 3% -0.1 0.27 ± 8% perf-profile.children.cycles-pp.hrtimer_interrupt 0.44 ± 2% -0.1 0.37 ± 3% perf-profile.children.cycles-pp.smp_apic_timer_interrupt 0.26 ± 3% -0.1 0.20 ± 9% perf-profile.children.cycles-pp.__hrtimer_run_queues 0.19 ± 3% -0.0 0.14 ± 3% perf-profile.children.cycles-pp.tick_sched_timer 0.16 ± 5% -0.0 0.12 ± 3% perf-profile.children.cycles-pp.tick_sched_handle 0.22 ± 5% -0.0 0.19 ± 6% perf-profile.children.cycles-pp.pte_alloc_one 0.15 ± 5% -0.0 0.12 ± 3% perf-profile.children.cycles-pp.update_process_times 95.14 +0.3 95.44 perf-profile.children.cycles-pp.do_page_fault 95.01 +0.3 95.31 perf-profile.children.cycles-pp.handle_mm_fault 95.19 +0.3 95.49 perf-profile.children.cycles-pp.page_fault 94.98 +0.3 95.29 perf-profile.children.cycles-pp.__handle_mm_fault 94.89 +0.3 95.22 perf-profile.children.cycles-pp.do_huge_pmd_anonymous_page 91.02 +0.5 91.52 perf-profile.children.cycles-pp.clear_huge_page 86.50 -86.4 0.12 ± 9% perf-profile.self.cycles-pp.clear_page_erms 0.57 ± 4% -0.5 0.08 ± 13% perf-profile.self.cycles-pp.rcu_all_qs 1.26 ± 4% -0.5 0.78 ± 3% perf-profile.self.cycles-pp.___might_sleep 2.16 ± 3% -0.2 1.96 ± 3% perf-profile.self.cycles-pp.__free_pages_ok 2.09 -0.1 2.00 perf-profile.self.cycles-pp.get_page_from_freelist 0.36 ± 4% +0.1 0.50 ± 4% perf-profile.self.cycles-pp._cond_resched 0.11 ± 11% +89.6 89.74 perf-profile.self.cycles-pp.clear_huge_page *************************************************************************************************** lkp-ivb-d02: 4 threads Intel(R) Core(TM) i3-3220 CPU @ 3.30GHz with 8G memory ========================================================================================= compiler/cpufreq_governor/kconfig/rootfs/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/debian-x86_64-20191114.cgz/lkp-ivb-d02/page_fault1/will-it-scale/0x21 commit: 8632e9b564 (" hyperv-fixes for 5.7-rc1") 9266baf8ee ("mm: Optimized hugepage zeroing & copying from user") 8632e9b5645bbc23 9266baf8eea588a78e5a0aa4ccd ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 0:4 2% 0:4 perf-profile.self.cycles-pp.error_entry %stddev %change %stddev \ | \ 2061889 -5.3% 1951865 will-it-scale.per_process_ops 2171644 -5.7% 2048044 will-it-scale.per_thread_ops 0.23 +2.1% 0.23 will-it-scale.scalability 1232399 -5.2% 1168571 will-it-scale.time.minor_page_faults 16.41 -6.3% 15.38 will-it-scale.time.user_time 57433 -6.5% 53713 will-it-scale.time.voluntary_context_switches 32371162 -5.1% 30719792 will-it-scale.workload 1138201 ± 16% -10.8% 1015835 ± 16% cpuidle.C1E.usage 532.00 ± 11% -22.6% 412.00 ± 20% slabinfo.skbuff_head_cache.num_objs 24570 +2.8% 25263 interrupts.CAL:Function_call_interrupts 1823 ± 20% +26.7% 2309 ± 11% interrupts.CPU0.RES:Rescheduling_interrupts 28931 ± 11% -15.5% 24434 ± 14% softirqs.CPU1.SCHED 14949 ± 11% +17.2% 17527 ± 6% softirqs.CPU2.RCU 14575349 -4.8% 13875471 proc-vmstat.numa_hit 14575349 -4.8% 13875471 proc-vmstat.numa_local 1.228e+09 -5.0% 1.167e+09 proc-vmstat.pgalloc_normal 12186704 -4.7% 11609076 proc-vmstat.pgfault 1.228e+09 -5.0% 1.167e+09 proc-vmstat.pgfree 2374618 -5.0% 2256894 proc-vmstat.thp_fault_alloc 0.00 +2e+11% 2041 ± 82% sched_debug.cfs_rq:/.MIN_vruntime.avg 0.00 +7.5e+11% 7472 ± 94% sched_debug.cfs_rq:/.MIN_vruntime.max 266202 ± 2% +16.4% 309784 ± 5% sched_debug.cfs_rq:/.load.avg 0.00 +2e+11% 2041 ± 82% sched_debug.cfs_rq:/.max_vruntime.avg 0.00 +7.5e+11% 7472 ± 94% sched_debug.cfs_rq:/.max_vruntime.max 155407 ± 7% -19.4% 125275 ± 15% sched_debug.cfs_rq:/.min_vruntime.min 58058 ± 16% +28.5% 74586 ± 10% sched_debug.cfs_rq:/.min_vruntime.stddev 4.17 ± 36% +46.0% 6.08 ± 17% sched_debug.cfs_rq:/.nr_spread_over.max 22843 ±100% +333.2% 98964 ± 51% sched_debug.cfs_rq:/.spread0.max 58060 ± 16% +28.5% 74587 ± 10% sched_debug.cfs_rq:/.spread0.stddev 58.60 ± 8% -58.6 0.00 perf-profile.calltrace.cycles-pp.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault.handle_mm_fault 57.50 ± 8% -57.5 0.00 perf-profile.calltrace.cycles-pp.clear_page_erms.clear_subpage.clear_huge_page.do_huge_pmd_anonymous_page.__handle_mm_fault 0.67 ± 4% -0.1 0.59 ± 6% perf-profile.calltrace.cycles-pp.prep_new_page.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.handle_pte_fault 0.65 ± 3% -0.1 0.57 ± 7% perf-profile.calltrace.cycles-pp.clear_page_erms.prep_new_page.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma 58.68 ± 8% -58.7 0.00 perf-profile.children.cycles-pp.clear_subpage 58.41 ± 8% -57.7 0.69 ± 7% perf-profile.children.cycles-pp.clear_page_erms 0.08 ± 8% -0.0 0.06 ± 6% perf-profile.children.cycles-pp.free_pages_and_swap_cache 0.10 ± 11% +0.0 0.13 ± 3% perf-profile.children.cycles-pp._raw_spin_lock 0.00 +0.1 0.09 ± 29% perf-profile.children.cycles-pp.forkshell 0.23 ± 19% +0.3 0.50 ± 5% perf-profile.children.cycles-pp.rcu_all_qs 57.92 ± 8% -57.2 0.69 ± 6% perf-profile.self.cycles-pp.clear_page_erms 0.08 ± 5% -0.0 0.06 ± 6% perf-profile.self.cycles-pp.free_pages_and_swap_cache 0.18 ± 13% +0.1 0.26 ± 7% perf-profile.self.cycles-pp.rcu_all_qs 0.72 ± 13% +60.5 61.22 ± 6% perf-profile.self.cycles-pp.clear_huge_page 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