Greeting, FYI, we noticed a 28.7% improvement of will-it-scale.per_process_ops due to commit: commit: edf5d99c32537dd1257b0720546378b08bf55be7 (" iHUEABYIAB0WIQQeIjHXFZhlkJWE3YwfdExxlD3i+QUCXhOWDwAKCRAfdExxlD3i") https://github.com/sbates130272/linux-p2pmem.git remap_pages_cache_v2 in testcase: will-it-scale on test machine: 288 threads Intel(R) Xeon Phi(TM) CPU 7295 @ 1.50GHz with 80G memory with following parameters: nr_task: 100% mode: process test: lseek1 cpufreq_governor: performance ucode: 0x11 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 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 30.3% improvement | | test machine | 288 threads Intel(R) Xeon Phi(TM) CPU 7295 @ 1.50GHz with 80G memory | | test parameters | cpufreq_governor=performance | | | mode=process | | | nr_task=100% | | | test=lseek2 | | | ucode=0x11 | +------------------+----------------------------------------------------------------------+ 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/ucode: gcc-7/performance/x86_64-rhel-7.6/process/100%/debian-x86_64-2019-11-14.cgz/lkp-knm01/lseek1/will-it-scale/0x11 commit: 75fed70a5c (" iHUEABYIAB0WIQQeIjHXFZhlkJWE3YwfdExxlD3i+QUCXhOWDwAKCRAfdExxlD3i") edf5d99c32 (" iHUEABYIAB0WIQQeIjHXFZhlkJWE3YwfdExxlD3i+QUCXhOWDwAKCRAfdExxlD3i") 75fed70a5c4d2b25 edf5d99c32537dd1257b0720546 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 dmesg.WARNING:at_ip___perf_sw_event/0x %stddev %change %stddev \ | \ 328529 +28.7% 422923 will-it-scale.per_process_ops 94616746 +28.7% 1.218e+08 will-it-scale.workload 284.67 ± 23% +624.0% 2061 ±110% cpuidle.POLL.usage 144.78 +2.6% 148.56 turbostat.PkgWatt 51.67 -10.5% 46.25 ± 4% vmstat.cpu.sy 45.00 +11.1% 50.00 ± 4% vmstat.cpu.us 2378 ± 14% +20.2% 2858 sched_debug.cfs_rq:/.load.min 2378 ± 14% +12.7% 2681 ± 11% sched_debug.cfs_rq:/.runnable_weight.min -76613 -83.2% -12841 sched_debug.cfs_rq:/.spread0.avg 1320056 ± 4% -16.7% 1099298 ± 13% sched_debug.cpu.avg_idle.stddev 0.01 ± 99% +116.0% 0.01 ± 45% sched_debug.cpu.nr_uninterruptible.avg 124951 ± 11% +25.7% 157069 ± 8% softirqs.CPU118.TIMER 119157 ± 7% +4.9% 125043 ± 8% softirqs.CPU18.TIMER 105394 ± 3% +6.9% 112621 ± 3% softirqs.CPU19.TIMER 111056 ± 8% +6.1% 117797 ± 8% softirqs.CPU281.TIMER 114657 ± 10% +6.6% 122211 ± 12% softirqs.CPU283.TIMER 44.25 ± 2% -20.3% 35.25 perf-stat.i.MPKI 6.143e+09 +27.0% 7.804e+09 perf-stat.i.branch-instructions 8.11 +0.1 8.20 perf-stat.i.branch-miss-rate% 4.981e+08 +28.3% 6.391e+08 perf-stat.i.branch-misses 15.01 -20.8% 11.88 perf-stat.i.cpi 1.48 +0.0 1.52 perf-stat.i.iTLB-load-miss-rate% 4.361e+08 +29.6% 5.653e+08 perf-stat.i.iTLB-load-misses 2.896e+10 +26.8% 3.671e+10 perf-stat.i.iTLB-loads 2.9e+10 +26.8% 3.678e+10 perf-stat.i.instructions 66.66 -2.2% 65.21 perf-stat.i.instructions-per-iTLB-miss 0.07 +26.5% 0.08 perf-stat.i.ipc 44.38 ± 2% -20.6% 35.25 perf-stat.overall.MPKI 8.11 +0.1 8.19 perf-stat.overall.branch-miss-rate% 15.04 -21.3% 11.84 perf-stat.overall.cpi 1.48 +0.0 1.52 perf-stat.overall.iTLB-load-miss-rate% 66.49 -2.2% 65.05 perf-stat.overall.instructions-per-iTLB-miss 0.07 +27.0% 0.08 perf-stat.overall.ipc 95657 -1.2% 94500 perf-stat.overall.path-length 6.12e+09 +27.1% 7.777e+09 perf-stat.ps.branch-instructions 4.963e+08 +28.3% 6.37e+08 perf-stat.ps.branch-misses 4.345e+08 +29.7% 5.634e+08 perf-stat.ps.iTLB-load-misses 2.886e+10 +26.8% 3.659e+10 perf-stat.ps.iTLB-loads 2.889e+10 +26.9% 3.665e+10 perf-stat.ps.instructions 9.051e+12 +27.2% 1.151e+13 perf-stat.total.instructions 6.95 ± 77% -6.0 0.95 ±173% perf-profile.calltrace.cycles-pp.lseek64 7.34 ± 5% -4.9 2.45 ± 3% perf-profile.calltrace.cycles-pp.__fget_light.__fdget_pos.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 15.23 ± 9% -4.5 10.75 ± 4% perf-profile.calltrace.cycles-pp.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 7.46 ± 9% -4.3 3.19 ± 3% perf-profile.calltrace.cycles-pp.__fdget_pos.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 34.87 ± 7% -3.6 31.26 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 4.79 ± 9% -1.7 3.14 ± 4% perf-profile.calltrace.cycles-pp.shmem_file_llseek.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.36 ± 15% -1.3 2.08 ± 12% perf-profile.calltrace.cycles-pp.fpregs_assert_state_consistent.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.77 ± 10% +0.2 0.99 ± 4% perf-profile.calltrace.cycles-pp.__x64_sys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.39 ± 70% +0.4 0.82 ± 4% perf-profile.calltrace.cycles-pp.generic_file_llseek_size.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 29.77 ± 6% +2.5 32.28 ± 3% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64 18.23 ± 4% +4.6 22.88 ± 5% perf-profile.calltrace.cycles-pp.syscall_return_via_sysret 7.39 ± 76% -6.4 1.02 ±173% perf-profile.children.cycles-pp.lseek64 37.80 ± 2% -5.5 32.25 perf-profile.children.cycles-pp.do_syscall_64 16.42 ± 4% -5.2 11.18 ± 4% perf-profile.children.cycles-pp.ksys_lseek 7.41 ± 5% -4.9 2.56 ± 2% perf-profile.children.cycles-pp.__fget_light 8.10 ± 4% -4.6 3.48 ± 3% perf-profile.children.cycles-pp.__fdget_pos 41.21 ± 2% -4.3 36.91 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 5.10 ± 4% -1.9 3.17 ± 4% perf-profile.children.cycles-pp.shmem_file_llseek 3.66 ± 10% -1.4 2.22 ± 11% perf-profile.children.cycles-pp.fpregs_assert_state_consistent 0.13 ± 24% -0.1 0.07 ± 60% perf-profile.children.cycles-pp.ksoftirqd_running 0.12 ± 13% -0.0 0.07 ± 58% perf-profile.children.cycles-pp.run_posix_cpu_timers 0.57 ± 4% +0.2 0.77 ± 3% perf-profile.children.cycles-pp.__x86_indirect_thunk_rax 0.86 ± 5% +0.2 1.07 ± 4% perf-profile.children.cycles-pp.__x64_sys_lseek 0.65 ± 3% +0.3 0.93 ± 5% perf-profile.children.cycles-pp.generic_file_llseek_size 21.65 +4.1 25.77 ± 4% perf-profile.children.cycles-pp.syscall_return_via_sysret 6.95 ± 6% -4.7 2.29 ± 4% perf-profile.self.cycles-pp.__fget_light 4.77 ± 5% -1.9 2.92 ± 6% perf-profile.self.cycles-pp.shmem_file_llseek 3.53 ± 10% -1.5 2.04 ± 11% perf-profile.self.cycles-pp.fpregs_assert_state_consistent 0.13 ± 24% -0.1 0.07 ± 59% perf-profile.self.cycles-pp.ksoftirqd_running 0.12 ± 13% -0.0 0.07 ± 58% perf-profile.self.cycles-pp.run_posix_cpu_timers 0.47 ± 6% +0.2 0.62 ± 4% perf-profile.self.cycles-pp.__x86_indirect_thunk_rax 0.69 ± 6% +0.2 0.86 ± 6% perf-profile.self.cycles-pp.__x64_sys_lseek 0.64 ± 6% +0.2 0.89 ± 8% perf-profile.self.cycles-pp.__fdget_pos 0.62 ± 5% +0.3 0.88 ± 7% perf-profile.self.cycles-pp.generic_file_llseek_size 2.48 ± 6% +1.0 3.48 ± 7% perf-profile.self.cycles-pp.ksys_lseek 13.39 ± 2% +1.0 14.44 ± 4% perf-profile.self.cycles-pp.do_syscall_64 3.20 ± 6% +1.3 4.49 ± 7% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe 21.61 +4.1 25.70 ± 4% perf-profile.self.cycles-pp.syscall_return_via_sysret 15.33 ± 69% +229.3% 50.50 ± 81% interrupts.CPU10.RES:Rescheduling_interrupts 8274 ± 6% -50.8% 4068 ± 12% interrupts.CPU105.NMI:Non-maskable_interrupts 8274 ± 6% -50.8% 4068 ± 12% interrupts.CPU105.PMI:Performance_monitoring_interrupts 296.00 ± 83% -88.7% 33.50 ± 98% interrupts.CPU116.RES:Rescheduling_interrupts 6750 ± 23% -39.7% 4071 ± 12% interrupts.CPU121.NMI:Non-maskable_interrupts 6750 ± 23% -39.7% 4071 ± 12% interrupts.CPU121.PMI:Performance_monitoring_interrupts 6716 ± 24% -40.0% 4026 ± 12% interrupts.CPU126.NMI:Non-maskable_interrupts 6716 ± 24% -40.0% 4026 ± 12% interrupts.CPU126.PMI:Performance_monitoring_interrupts 207.33 ± 68% -98.4% 3.25 ± 70% interrupts.CPU126.RES:Rescheduling_interrupts 4.00 ± 54% +1631.2% 69.25 ± 89% interrupts.CPU129.RES:Rescheduling_interrupts 184.00 ±102% +121.2% 407.00 ± 41% interrupts.CPU134.RES:Rescheduling_interrupts 81.33 ±123% +213.8% 255.25 ± 71% interrupts.CPU135.RES:Rescheduling_interrupts 6545 ± 23% -37.7% 4077 ± 11% interrupts.CPU141.NMI:Non-maskable_interrupts 6545 ± 23% -37.7% 4077 ± 11% interrupts.CPU141.PMI:Performance_monitoring_interrupts 6897 ± 29% -40.7% 4087 ± 11% interrupts.CPU148.NMI:Non-maskable_interrupts 6897 ± 29% -40.7% 4087 ± 11% interrupts.CPU148.PMI:Performance_monitoring_interrupts 8224 ± 5% -50.8% 4050 ± 12% interrupts.CPU154.NMI:Non-maskable_interrupts 8224 ± 5% -50.8% 4050 ± 12% interrupts.CPU154.PMI:Performance_monitoring_interrupts 3.00 ± 47% +1600.0% 51.00 ± 73% interrupts.CPU165.RES:Rescheduling_interrupts 73.33 ± 48% -81.6% 13.50 ±143% interrupts.CPU169.RES:Rescheduling_interrupts 6832 ± 24% -40.4% 4072 ± 11% interrupts.CPU172.NMI:Non-maskable_interrupts 6832 ± 24% -40.4% 4072 ± 11% interrupts.CPU172.PMI:Performance_monitoring_interrupts 164.33 ± 57% -93.9% 10.00 ± 94% interrupts.CPU174.RES:Rescheduling_interrupts 6686 ± 23% -38.6% 4107 ± 10% interrupts.CPU175.NMI:Non-maskable_interrupts 6686 ± 23% -38.6% 4107 ± 10% interrupts.CPU175.PMI:Performance_monitoring_interrupts 401.00 ± 55% -89.2% 43.25 ±158% interrupts.CPU180.RES:Rescheduling_interrupts 8.00 ± 88% +1603.1% 136.25 ± 93% interrupts.CPU184.RES:Rescheduling_interrupts 119.67 ± 85% +304.0% 483.50 ± 15% interrupts.CPU185.RES:Rescheduling_interrupts 138.33 ± 52% -93.7% 8.75 ± 67% interrupts.CPU188.RES:Rescheduling_interrupts 85.33 ± 56% -91.5% 7.25 ± 24% interrupts.CPU190.RES:Rescheduling_interrupts 167.67 ±113% -94.0% 10.00 ±111% interrupts.CPU197.RES:Rescheduling_interrupts 4123 ± 6% +67.9% 6922 ± 16% interrupts.CPU198.NMI:Non-maskable_interrupts 4123 ± 6% +67.9% 6922 ± 16% interrupts.CPU198.PMI:Performance_monitoring_interrupts 329.00 ± 4% -64.5% 116.75 ±102% interrupts.CPU198.RES:Rescheduling_interrupts 6864 ± 24% -40.2% 4104 ± 11% interrupts.CPU20.NMI:Non-maskable_interrupts 6864 ± 24% -40.2% 4104 ± 11% interrupts.CPU20.PMI:Performance_monitoring_interrupts 56.00 ± 71% +1179.0% 716.25 ±102% interrupts.CPU206.RES:Rescheduling_interrupts 8212 ± 6% -50.4% 4071 ± 11% interrupts.CPU207.NMI:Non-maskable_interrupts 8212 ± 6% -50.4% 4071 ± 11% interrupts.CPU207.PMI:Performance_monitoring_interrupts 4204 ± 6% +95.8% 8233 ± 11% interrupts.CPU21.NMI:Non-maskable_interrupts 4204 ± 6% +95.8% 8233 ± 11% interrupts.CPU21.PMI:Performance_monitoring_interrupts 6844 ± 30% -40.3% 4082 ± 11% interrupts.CPU211.NMI:Non-maskable_interrupts 6844 ± 30% -40.3% 4082 ± 11% interrupts.CPU211.PMI:Performance_monitoring_interrupts 117.67 ± 81% +223.6% 380.75 ± 36% interrupts.CPU212.RES:Rescheduling_interrupts 6708 ± 24% -40.2% 4008 ± 12% interrupts.CPU218.NMI:Non-maskable_interrupts 6708 ± 24% -40.2% 4008 ± 12% interrupts.CPU218.PMI:Performance_monitoring_interrupts 287.33 ±126% -95.2% 13.75 ± 98% interrupts.CPU22.RES:Rescheduling_interrupts 8087 ± 7% -49.3% 4102 ± 11% interrupts.CPU221.NMI:Non-maskable_interrupts 8087 ± 7% -49.3% 4102 ± 11% interrupts.CPU221.PMI:Performance_monitoring_interrupts 5.00 ± 43% +1970.0% 103.50 ±130% interrupts.CPU223.RES:Rescheduling_interrupts 25.00 ±108% +1262.0% 340.50 ±138% interrupts.CPU236.RES:Rescheduling_interrupts 8324 ± 5% -27.9% 5998 ± 27% interrupts.CPU241.NMI:Non-maskable_interrupts 8324 ± 5% -27.9% 5998 ± 27% interrupts.CPU241.PMI:Performance_monitoring_interrupts 4180 ± 5% +64.0% 6855 ± 16% interrupts.CPU255.NMI:Non-maskable_interrupts 4180 ± 5% +64.0% 6855 ± 16% interrupts.CPU255.PMI:Performance_monitoring_interrupts 197.00 ± 58% -90.6% 18.50 ±148% interrupts.CPU256.RES:Rescheduling_interrupts 6885 ± 25% -40.2% 4115 ± 11% interrupts.CPU259.NMI:Non-maskable_interrupts 6885 ± 25% -40.2% 4115 ± 11% interrupts.CPU259.PMI:Performance_monitoring_interrupts 121.00 ±102% -90.7% 11.25 ±132% interrupts.CPU263.RES:Rescheduling_interrupts 76.00 ± 44% +188.5% 219.25 ± 28% interrupts.CPU266.RES:Rescheduling_interrupts 276.00 ± 47% -70.9% 80.25 ±123% interrupts.CPU27.RES:Rescheduling_interrupts 75.67 ± 67% -76.5% 17.75 ± 93% interrupts.CPU271.RES:Rescheduling_interrupts 7.33 ± 12% +2589.8% 197.25 ± 33% interrupts.CPU273.RES:Rescheduling_interrupts 4164 ± 6% +66.7% 6941 ± 16% interrupts.CPU287.NMI:Non-maskable_interrupts 4164 ± 6% +66.7% 6941 ± 16% interrupts.CPU287.PMI:Performance_monitoring_interrupts 333.00 ± 41% -78.0% 73.25 ±108% interrupts.CPU30.RES:Rescheduling_interrupts 8338 ± 7% -51.3% 4061 ± 12% interrupts.CPU37.NMI:Non-maskable_interrupts 8338 ± 7% -51.3% 4061 ± 12% interrupts.CPU37.PMI:Performance_monitoring_interrupts 8385 ± 6% -39.9% 5036 ± 32% interrupts.CPU49.NMI:Non-maskable_interrupts 8385 ± 6% -39.9% 5036 ± 32% interrupts.CPU49.PMI:Performance_monitoring_interrupts 367.33 ± 32% -74.5% 93.75 ± 97% interrupts.CPU5.RES:Rescheduling_interrupts 314.67 ± 27% -96.7% 10.25 ±107% interrupts.CPU53.RES:Rescheduling_interrupts 5.33 ± 31% +1250.0% 72.00 ±109% interrupts.CPU6.RES:Rescheduling_interrupts 5.00 +4265.0% 218.25 ±117% interrupts.CPU60.RES:Rescheduling_interrupts 8219 ± 5% -50.2% 4093 ± 11% interrupts.CPU64.NMI:Non-maskable_interrupts 8219 ± 5% -50.2% 4093 ± 11% interrupts.CPU64.PMI:Performance_monitoring_interrupts 4186 ± 5% +94.2% 8129 ± 12% interrupts.CPU89.NMI:Non-maskable_interrupts 4186 ± 5% +94.2% 8129 ± 12% interrupts.CPU89.PMI:Performance_monitoring_interrupts 6.00 ± 75% +579.2% 40.75 ±123% interrupts.CPU9.RES:Rescheduling_interrupts will-it-scale.per_process_ops 450000 +-+--O--O----------------------------------------------------------+ O O O O O O O O O O O O O O O O O O O O O O O | 400000 +-+ | 350000 +-+ +.+..+..+.+..+.+..+.. | |..+.+..+ : +.+..+.+ +..+..+.+ +.+..| 300000 +-+ : : : : : : | 250000 +-+ : : : : : : | | : : : : : : | 200000 +-+ : : : : : : | 150000 +-+ : : : : : : | | : : : : : : | 100000 +-+ : : : : : : | 50000 +-+ :: :: : : | | : : : : | 0 +-+----------------------------------------------------------------+ will-it-scale.workload 1.4e+08 +-+---------------------------------------------------------------+ | O O | 1.2e+08 O-+O O O O O O O O O O O O O O O O O O O O O O | | | 1e+08 +-+ .+.. +.+..+.+..+.+..+.+..+.. .+.. +. .+. .+..| | + + : + + : +. + + | 8e+07 +-+ : : : : : : | | : : : : : : | 6e+07 +-+ : : : : : : | | : : : : : : | 4e+07 +-+ : : : : : : | | : : : : : : | 2e+07 +-+ :: :: : : | | : : : : | 0 +-+---------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample *************************************************************************************************** lkp-knm01: 288 threads Intel(R) Xeon Phi(TM) CPU 7295 @ 1.50GHz with 80G memory ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase/ucode: gcc-7/performance/x86_64-rhel-7.6/process/100%/debian-x86_64-2019-11-14.cgz/lkp-knm01/lseek2/will-it-scale/0x11 commit: 75fed70a5c (" iHUEABYIAB0WIQQeIjHXFZhlkJWE3YwfdExxlD3i+QUCXhOWDwAKCRAfdExxlD3i") edf5d99c32 (" iHUEABYIAB0WIQQeIjHXFZhlkJWE3YwfdExxlD3i+QUCXhOWDwAKCRAfdExxlD3i") 75fed70a5c4d2b25 edf5d99c32537dd1257b0720546 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | :4 25% 1:4 dmesg.WARNING:at#for_ip_interrupt_entry/0x %stddev %change %stddev \ | \ 330632 +30.3% 430704 will-it-scale.per_process_ops 95222345 +30.3% 1.24e+08 will-it-scale.workload 4791 ± 2% -12.5% 4193 ± 6% numa-vmstat.node1.nr_slab_reclaimable 4.287e+08 ± 41% +44.9% 6.211e+08 ± 39% cpuidle.C6.time 17256 ± 30% +303.5% 69629 ±101% cpuidle.POLL.time 19164 ± 2% -12.5% 16774 ± 6% numa-meminfo.node1.KReclaimable 19164 ± 2% -12.5% 16774 ± 6% numa-meminfo.node1.SReclaimable 837582 +1.2% 847550 proc-vmstat.numa_hit 837582 +1.2% 847550 proc-vmstat.numa_local 106639 ± 4% +11.9% 119303 ± 8% softirqs.CPU19.TIMER 134988 ± 12% +17.8% 159032 ± 8% softirqs.CPU261.TIMER 52.00 -8.2% 47.75 vmstat.cpu.sy 46.00 +8.2% 49.75 vmstat.cpu.us 3.31 ± 4% +26.5% 4.18 ± 11% sched_debug.cfs_rq:/.load_avg.avg 75.90 ± 36% +280.5% 288.80 ± 45% sched_debug.cfs_rq:/.load_avg.max 5.72 ± 26% +217.7% 18.18 ± 42% sched_debug.cfs_rq:/.load_avg.stddev 0.34 ± 59% +0.2 0.55 ± 47% turbostat.C6% 0.18 ± 88% +100.0% 0.35 ± 60% turbostat.CPU%c6 145.07 +2.7% 148.94 turbostat.PkgWatt 7674 ± 4% -6.1% 7204 ± 5% slabinfo.khugepaged_mm_slot.active_objs 7674 ± 4% -6.1% 7204 ± 5% slabinfo.khugepaged_mm_slot.num_objs 1097 ± 4% +12.0% 1229 slabinfo.task_group.active_objs 1097 ± 4% +12.0% 1229 slabinfo.task_group.num_objs 42.39 ± 2% -16.7% 35.29 perf-stat.i.MPKI 6.24e+09 +28.5% 8.02e+09 perf-stat.i.branch-instructions 8.04 +0.1 8.14 perf-stat.i.branch-miss-rate% 5.019e+08 +30.0% 6.526e+08 perf-stat.i.branch-misses 4.02 ± 3% -0.2 3.85 ± 2% perf-stat.i.cache-miss-rate% 1.245e+09 +6.7% 1.328e+09 perf-stat.i.cache-references 14.87 -21.7% 11.64 perf-stat.i.cpi 1.47 +0.0 1.50 perf-stat.i.iTLB-load-miss-rate% 4.379e+08 +30.9% 5.732e+08 perf-stat.i.iTLB-load-misses 2.929e+10 +28.1% 3.751e+10 perf-stat.i.iTLB-loads 2.931e+10 +28.2% 3.757e+10 perf-stat.i.instructions 67.06 -2.0% 65.70 perf-stat.i.instructions-per-iTLB-miss 0.07 +27.4% 0.09 perf-stat.i.ipc 42.48 ± 2% -16.8% 35.34 perf-stat.overall.MPKI 8.04 +0.1 8.14 perf-stat.overall.branch-miss-rate% 3.91 ± 3% -0.2 3.71 ± 2% perf-stat.overall.cache-miss-rate% 14.90 -21.8% 11.66 perf-stat.overall.cpi 1.47 +0.0 1.51 perf-stat.overall.iTLB-load-miss-rate% 66.91 -2.0% 65.54 perf-stat.overall.instructions-per-iTLB-miss 0.07 +27.8% 0.09 perf-stat.overall.ipc 95601 -1.6% 94026 perf-stat.overall.path-length 6.216e+09 +28.5% 7.988e+09 perf-stat.ps.branch-instructions 5e+08 +30.0% 6.501e+08 perf-stat.ps.branch-misses 1.24e+09 +6.7% 1.322e+09 perf-stat.ps.cache-references 4.362e+08 +30.9% 5.709e+08 perf-stat.ps.iTLB-load-misses 2.917e+10 +28.1% 3.736e+10 perf-stat.ps.iTLB-loads 2.919e+10 +28.2% 3.742e+10 perf-stat.ps.instructions 9.103e+12 +28.1% 1.166e+13 perf-stat.total.instructions 6.84 ± 3% -4.4 2.48 ± 3% perf-profile.calltrace.cycles-pp.__fget_light.__fdget_pos.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 15.01 ± 2% -4.1 10.91 ± 3% perf-profile.calltrace.cycles-pp.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 7.33 ± 3% -4.1 3.24 ± 3% perf-profile.calltrace.cycles-pp.__fdget_pos.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 34.91 ± 2% -3.8 31.13 ± 2% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 38.16 ± 2% -2.3 35.82 ± 2% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 4.69 ± 2% -1.5 3.17 ± 3% perf-profile.calltrace.cycles-pp.shmem_file_llseek.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.10 ± 4% -1.1 2.04 ± 6% perf-profile.calltrace.cycles-pp.fpregs_assert_state_consistent.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.77 ± 2% +0.2 1.02 ± 2% perf-profile.calltrace.cycles-pp.__x64_sys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.56 ± 3% +0.3 0.84 ± 3% perf-profile.calltrace.cycles-pp.generic_file_llseek_size.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe 19.58 ± 3% +3.2 22.76 ± 2% perf-profile.calltrace.cycles-pp.syscall_return_via_sysret 36.88 -4.6 32.27 perf-profile.children.cycles-pp.do_syscall_64 7.15 -4.5 2.61 ± 3% perf-profile.children.cycles-pp.__fget_light 15.78 -4.3 11.46 ± 2% perf-profile.children.cycles-pp.ksys_lseek 7.77 -4.2 3.57 ± 4% perf-profile.children.cycles-pp.__fdget_pos 40.18 -3.1 37.05 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 4.86 -1.6 3.23 ± 2% perf-profile.children.cycles-pp.shmem_file_llseek 3.31 ± 3% -1.1 2.19 ± 5% perf-profile.children.cycles-pp.fpregs_assert_state_consistent 0.33 ± 10% -0.2 0.15 ± 66% perf-profile.children.cycles-pp.irq_enter 0.24 ± 10% -0.1 0.11 ± 65% perf-profile.children.cycles-pp.rcu_irq_enter 0.06 +0.0 0.08 ± 6% perf-profile.children.cycles-pp.vm_normal_page 0.61 ± 3% +0.1 0.67 ± 3% perf-profile.children.cycles-pp.change_protection 0.61 ± 3% +0.1 0.67 ± 3% perf-profile.children.cycles-pp.change_p4d_range 0.61 ± 3% +0.1 0.67 ± 3% perf-profile.children.cycles-pp.change_prot_numa 0.62 ± 3% +0.1 0.68 ± 3% perf-profile.children.cycles-pp.task_work_run 0.62 ± 3% +0.1 0.68 ± 2% perf-profile.children.cycles-pp.task_numa_work 0.63 ± 3% +0.1 0.70 ± 2% perf-profile.children.cycles-pp.exit_to_usermode_loop 0.36 ± 7% +0.1 0.43 ± 3% perf-profile.children.cycles-pp.prepare_exit_to_usermode 0.56 ± 2% +0.2 0.77 perf-profile.children.cycles-pp.__x86_indirect_thunk_rax 0.84 ± 2% +0.3 1.10 ± 2% perf-profile.children.cycles-pp.__x64_sys_lseek 0.64 ± 2% +0.3 0.96 ± 3% perf-profile.children.cycles-pp.generic_file_llseek_size 22.27 +3.6 25.83 ± 2% perf-profile.children.cycles-pp.syscall_return_via_sysret 6.64 -4.3 2.36 ± 6% perf-profile.self.cycles-pp.__fget_light 4.51 -1.5 2.99 ± 3% perf-profile.self.cycles-pp.shmem_file_llseek 3.19 ± 3% -1.2 2.02 ± 6% perf-profile.self.cycles-pp.fpregs_assert_state_consistent 0.73 ± 16% -0.3 0.42 ± 56% perf-profile.self.cycles-pp.__hrtimer_run_queues 0.49 ± 12% -0.3 0.24 ± 58% perf-profile.self.cycles-pp.smp_apic_timer_interrupt 0.24 ± 9% -0.1 0.11 ± 65% perf-profile.self.cycles-pp.rcu_irq_enter 0.20 ± 8% -0.1 0.13 ± 38% perf-profile.self.cycles-pp.perf_mux_hrtimer_handler 0.09 ± 7% -0.1 0.04 ±100% perf-profile.self.cycles-pp.irq_enter 0.46 ± 2% +0.2 0.64 ± 2% perf-profile.self.cycles-pp.__x86_indirect_thunk_rax 0.67 ± 2% +0.2 0.90 ± 4% perf-profile.self.cycles-pp.__x64_sys_lseek 0.60 ± 2% +0.3 0.92 ± 5% perf-profile.self.cycles-pp.generic_file_llseek_size 0.56 +0.4 0.92 ± 7% perf-profile.self.cycles-pp.__fdget_pos 2.42 +1.2 3.59 ± 4% perf-profile.self.cycles-pp.ksys_lseek 13.26 +1.3 14.55 ± 2% perf-profile.self.cycles-pp.do_syscall_64 3.10 ± 2% +1.5 4.62 ± 4% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe 22.23 +3.5 25.77 ± 2% perf-profile.self.cycles-pp.syscall_return_via_sysret 3780 +87.1% 7070 ± 22% interrupts.CPU101.NMI:Non-maskable_interrupts 3780 +87.1% 7070 ± 22% interrupts.CPU101.PMI:Performance_monitoring_interrupts 12.25 ± 67% +3938.8% 494.75 ±105% interrupts.CPU102.RES:Rescheduling_interrupts 4707 ± 32% +27.6% 6005 ± 25% interrupts.CPU116.NMI:Non-maskable_interrupts 4707 ± 32% +27.6% 6005 ± 25% interrupts.CPU116.PMI:Performance_monitoring_interrupts 5.25 ± 69% +8914.3% 473.25 ± 61% interrupts.CPU126.RES:Rescheduling_interrupts 3794 +55.7% 5907 ± 25% interrupts.CPU128.NMI:Non-maskable_interrupts 3794 +55.7% 5907 ± 25% interrupts.CPU128.PMI:Performance_monitoring_interrupts 3766 ± 2% +89.4% 7134 ± 23% interrupts.CPU129.NMI:Non-maskable_interrupts 3766 ± 2% +89.4% 7134 ± 23% interrupts.CPU129.PMI:Performance_monitoring_interrupts 3770 +84.9% 6970 ± 21% interrupts.CPU132.NMI:Non-maskable_interrupts 3770 +84.9% 6970 ± 21% interrupts.CPU132.PMI:Performance_monitoring_interrupts 4728 ± 34% +52.0% 7189 ± 24% interrupts.CPU137.NMI:Non-maskable_interrupts 4728 ± 34% +52.0% 7189 ± 24% interrupts.CPU137.PMI:Performance_monitoring_interrupts 3781 +33.8% 5060 ± 28% interrupts.CPU140.NMI:Non-maskable_interrupts 3781 +33.8% 5060 ± 28% interrupts.CPU140.PMI:Performance_monitoring_interrupts 14.50 ±133% +515.5% 89.25 ± 77% interrupts.CPU140.RES:Rescheduling_interrupts 3823 +58.0% 6039 ± 26% interrupts.CPU146.NMI:Non-maskable_interrupts 3823 +58.0% 6039 ± 26% interrupts.CPU146.PMI:Performance_monitoring_interrupts 4690 ± 31% +28.3% 6016 ± 25% interrupts.CPU149.NMI:Non-maskable_interrupts 4690 ± 31% +28.3% 6016 ± 25% interrupts.CPU149.PMI:Performance_monitoring_interrupts 107.75 ± 88% -78.7% 23.00 ± 61% interrupts.CPU149.RES:Rescheduling_interrupts 3825 +62.3% 6209 ± 32% interrupts.CPU150.NMI:Non-maskable_interrupts 3825 +62.3% 6209 ± 32% interrupts.CPU150.PMI:Performance_monitoring_interrupts 94.00 ±122% -93.9% 5.75 ± 46% interrupts.CPU159.RES:Rescheduling_interrupts 114.75 ±108% -78.9% 24.25 ±125% interrupts.CPU16.RES:Rescheduling_interrupts 23.25 ±112% +489.2% 137.00 ± 94% interrupts.CPU164.RES:Rescheduling_interrupts 7034 ± 5% -41.1% 4146 ± 8% interrupts.CPU171.NMI:Non-maskable_interrupts 7034 ± 5% -41.1% 4146 ± 8% interrupts.CPU171.PMI:Performance_monitoring_interrupts 12.00 ±101% +2462.5% 307.50 ±100% interrupts.CPU172.RES:Rescheduling_interrupts 481.75 ± 71% -94.8% 25.00 ± 48% interrupts.CPU177.RES:Rescheduling_interrupts 3773 +38.4% 5223 ± 39% interrupts.CPU187.NMI:Non-maskable_interrupts 3773 +38.4% 5223 ± 39% interrupts.CPU187.PMI:Performance_monitoring_interrupts 3702 ± 2% +36.7% 5061 ± 28% interrupts.CPU191.NMI:Non-maskable_interrupts 3702 ± 2% +36.7% 5061 ± 28% interrupts.CPU191.PMI:Performance_monitoring_interrupts 4617 ± 32% +38.7% 6402 ± 40% interrupts.CPU197.NMI:Non-maskable_interrupts 4617 ± 32% +38.7% 6402 ± 40% interrupts.CPU197.PMI:Performance_monitoring_interrupts 57.75 ±145% +201.7% 174.25 ± 68% interrupts.CPU198.RES:Rescheduling_interrupts 4776 ± 33% +49.7% 7152 ± 24% interrupts.CPU21.NMI:Non-maskable_interrupts 4776 ± 33% +49.7% 7152 ± 24% interrupts.CPU21.PMI:Performance_monitoring_interrupts 4725 ± 34% +31.0% 6190 ± 32% interrupts.CPU217.NMI:Non-maskable_interrupts 4725 ± 34% +31.0% 6190 ± 32% interrupts.CPU217.PMI:Performance_monitoring_interrupts 3830 +88.2% 7208 ± 24% interrupts.CPU226.NMI:Non-maskable_interrupts 3830 +88.2% 7208 ± 24% interrupts.CPU226.PMI:Performance_monitoring_interrupts 3776 +40.9% 5321 ± 41% interrupts.CPU227.NMI:Non-maskable_interrupts 3776 +40.9% 5321 ± 41% interrupts.CPU227.PMI:Performance_monitoring_interrupts 3863 +57.2% 6072 ± 26% interrupts.CPU23.NMI:Non-maskable_interrupts 3863 +57.2% 6072 ± 26% interrupts.CPU23.PMI:Performance_monitoring_interrupts 195.00 ±101% -97.1% 5.75 ± 33% interrupts.CPU23.RES:Rescheduling_interrupts 4677 ± 32% +76.8% 8268 ± 8% interrupts.CPU232.NMI:Non-maskable_interrupts 4677 ± 32% +76.8% 8268 ± 8% interrupts.CPU232.PMI:Performance_monitoring_interrupts 5325 ± 29% +54.5% 8225 ± 8% interrupts.CPU238.NMI:Non-maskable_interrupts 5325 ± 29% +54.5% 8225 ± 8% interrupts.CPU238.PMI:Performance_monitoring_interrupts 3765 ± 2% +68.2% 6331 ± 39% interrupts.CPU240.NMI:Non-maskable_interrupts 3765 ± 2% +68.2% 6331 ± 39% interrupts.CPU240.PMI:Performance_monitoring_interrupts 6.75 ± 42% +425.9% 35.50 ±121% interrupts.CPU240.RES:Rescheduling_interrupts 12.25 ± 89% +1575.5% 205.25 ± 84% interrupts.CPU243.RES:Rescheduling_interrupts 4.25 ± 30% +2811.8% 123.75 ± 86% interrupts.CPU244.RES:Rescheduling_interrupts 3817 +85.9% 7096 ± 21% interrupts.CPU256.NMI:Non-maskable_interrupts 3817 +85.9% 7096 ± 21% interrupts.CPU256.PMI:Performance_monitoring_interrupts 4691 ± 32% +28.0% 6005 ± 25% interrupts.CPU260.NMI:Non-maskable_interrupts 4691 ± 32% +28.0% 6005 ± 25% interrupts.CPU260.PMI:Performance_monitoring_interrupts 3799 +9.3% 4152 ± 7% interrupts.CPU265.NMI:Non-maskable_interrupts 3799 +9.3% 4152 ± 7% interrupts.CPU265.PMI:Performance_monitoring_interrupts 3804 +37.6% 5233 ± 38% interrupts.CPU267.NMI:Non-maskable_interrupts 3804 +37.6% 5233 ± 38% interrupts.CPU267.PMI:Performance_monitoring_interrupts 4496 ± 25% +59.7% 7180 ± 24% interrupts.CPU27.NMI:Non-maskable_interrupts 4496 ± 25% +59.7% 7180 ± 24% interrupts.CPU27.PMI:Performance_monitoring_interrupts 89.50 ± 97% +283.8% 343.50 ± 57% interrupts.CPU270.RES:Rescheduling_interrupts 4766 ± 32% +26.5% 6031 ± 25% interrupts.CPU28.NMI:Non-maskable_interrupts 4766 ± 32% +26.5% 6031 ± 25% interrupts.CPU28.PMI:Performance_monitoring_interrupts 3837 +57.5% 6045 ± 25% interrupts.CPU29.NMI:Non-maskable_interrupts 3837 +57.5% 6045 ± 25% interrupts.CPU29.PMI:Performance_monitoring_interrupts 7085 ± 5% -41.4% 4153 ± 9% interrupts.CPU45.NMI:Non-maskable_interrupts 7085 ± 5% -41.4% 4153 ± 9% interrupts.CPU45.PMI:Performance_monitoring_interrupts 406.75 ±104% -82.0% 73.25 ±118% interrupts.CPU53.RES:Rescheduling_interrupts 3828 +38.8% 5313 ± 41% interrupts.CPU55.NMI:Non-maskable_interrupts 3828 +38.8% 5313 ± 41% interrupts.CPU55.PMI:Performance_monitoring_interrupts 4507 ± 27% +39.0% 6264 ± 33% interrupts.CPU6.NMI:Non-maskable_interrupts 4507 ± 27% +39.0% 6264 ± 33% interrupts.CPU6.PMI:Performance_monitoring_interrupts 4663 ± 30% +30.3% 6078 ± 26% interrupts.CPU65.NMI:Non-maskable_interrupts 4663 ± 30% +30.3% 6078 ± 26% interrupts.CPU65.PMI:Performance_monitoring_interrupts 4693 ± 30% +54.1% 7234 ± 23% interrupts.CPU66.NMI:Non-maskable_interrupts 4693 ± 30% +54.1% 7234 ± 23% interrupts.CPU66.PMI:Performance_monitoring_interrupts 4551 ± 26% +58.3% 7206 ± 21% interrupts.CPU7.NMI:Non-maskable_interrupts 4551 ± 26% +58.3% 7206 ± 21% interrupts.CPU7.PMI:Performance_monitoring_interrupts 9.00 ±116% +2825.0% 263.25 ± 87% interrupts.CPU74.RES:Rescheduling_interrupts 638.50 ±144% -92.2% 49.50 ±145% interrupts.CPU9.RES:Rescheduling_interrupts 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, Oliver Sang