Greeting, FYI, we noticed a -2.9% regression of will-it-scale.per_process_ops due to commit: commit: a9d57ef15cbe327fe54416dd194ee0ea66ae53a4 ("x86/retpolines: Disable switch jump tables when retpolines are enabled") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master in testcase: will-it-scale on test machine: 88 threads Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz with 64G memory with following parameters: nr_task: 50% mode: process test: poll2 cpufreq_governor: performance test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two. test-url: https://github.com/antonblanchard/will-it-scale 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 1.9% improvement | | test machine | 104 threads Skylake with 192G memory | | test parameters | cpufreq_governor=performance | | | mode=process | | | nr_task=100% | | | test=poll2 | +------------------+---------------------------------------------------------------+ Details are as below: --------------------------------------------------------------------------------------------------> To reproduce: git clone https://github.com/intel/lkp-tests.git cd lkp-tests bin/lkp install job.yaml # job file is attached in this email bin/lkp run job.yaml ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-7/performance/x86_64-rhel-7.6/process/50%/debian-x86_64-2018-04-03-no-ucode.cgz/lkp-bdw-ep3d/poll2/will-it-scale commit: b929a500d6 ("x86/realmode: Don't leak the trampoline kernel address") a9d57ef15c ("x86/retpolines: Disable switch jump tables when retpolines are enabled") b929a500d6847916 a9d57ef15cbe327fe54416dd194 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 1:4 -25% :4 dmesg.WARNING:at#for_ip_interrupt_entry/0x %stddev %change %stddev \ | \ 330139 -2.9% 320625 will-it-scale.per_process_ops 14526142 -2.9% 14107522 will-it-scale.workload 54022 ± 5% +8.4% 58586 ± 5% meminfo.Shmem 246.31 -2.5% 240.07 turbostat.PkgWatt 264657 ± 32% -69.6% 80393 ±136% sched_debug.cpu.avg_idle.min 125898 ± 7% +27.2% 160086 ± 13% sched_debug.cpu.avg_idle.stddev 1373 ± 4% -9.3% 1245 ± 8% slabinfo.task_group.active_objs 1373 ± 4% -9.3% 1245 ± 8% slabinfo.task_group.num_objs 4457 +1.2% 4509 proc-vmstat.nr_inactive_anon 13502 ± 5% +8.4% 14640 ± 5% proc-vmstat.nr_shmem 4457 +1.2% 4509 proc-vmstat.nr_zone_inactive_anon 12930 ± 6% +11.9% 14468 ± 7% proc-vmstat.pgactivate 28524 ± 48% -46.5% 15253 ± 99% softirqs.CPU21.SCHED 16821 ± 13% +43.7% 24179 ± 20% softirqs.CPU29.RCU 10383 ± 5% -5.2% 9847 ± 4% softirqs.CPU51.RCU 4607 +237.0% 15526 ± 96% softirqs.CPU73.SCHED 9012 ± 4% +36.3% 12280 ± 24% softirqs.CPU79.RCU 9673 ± 5% -15.6% 8165 ± 11% softirqs.CPU86.RCU 18590 ± 4% -11.5% 16454 ± 5% softirqs.CPU9.RCU 68.51 ± 7% -10.9 57.58 ± 2% perf-profile.calltrace.cycles-pp.poll 64.19 ± 7% -10.1 54.05 ± 2% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.poll 64.02 ± 7% -10.1 53.91 ± 2% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 63.87 ± 7% -10.1 53.78 ± 2% perf-profile.calltrace.cycles-pp.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 62.25 ± 7% -9.7 52.51 ± 2% perf-profile.calltrace.cycles-pp.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 8.65 ± 7% -1.6 7.02 perf-profile.calltrace.cycles-pp.__fget_light.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.14 ± 7% -0.6 2.57 ± 2% perf-profile.calltrace.cycles-pp.testcase 2.36 ± 8% -0.5 1.90 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.poll 2.18 ± 7% -0.4 1.77 perf-profile.calltrace.cycles-pp.__fdget.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.12 ± 7% -0.4 1.72 ± 2% perf-profile.calltrace.cycles-pp._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 1.78 ± 8% -0.3 1.46 ± 3% perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.poll 1.64 ± 7% -0.3 1.35 ± 2% perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64 0.63 ± 8% -0.3 0.34 ± 70% perf-profile.calltrace.cycles-pp.__kmalloc.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.53 ± 63% +0.4 0.96 ± 2% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel 0.53 ± 63% +0.4 0.96 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64 0.53 ± 63% +0.4 0.96 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_kernel.secondary_startup_64 0.53 ± 63% +0.4 0.96 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_kernel.secondary_startup_64 0.53 ± 63% +0.4 0.96 perf-profile.calltrace.cycles-pp.start_kernel.secondary_startup_64 27.23 ± 18% +11.2 38.42 ± 3% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary 27.36 ± 19% +11.2 38.56 ± 3% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64 27.36 ± 19% +11.2 38.56 ± 3% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 27.36 ± 19% +11.2 38.56 ± 3% perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64 27.30 ± 19% +11.2 38.50 ± 3% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64 27.97 ± 19% +11.6 39.53 ± 3% perf-profile.calltrace.cycles-pp.secondary_startup_64 0.08 ± 2% +7.5% 0.09 ± 2% perf-stat.i.MPKI 3.464e+10 -2.9% 3.365e+10 perf-stat.i.branch-instructions 0.27 +0.0 0.27 perf-stat.i.branch-miss-rate% 91626903 -1.3% 90398973 perf-stat.i.branch-misses 201334 +7.1% 215711 ± 3% perf-stat.i.cache-misses 12934851 ± 2% +4.6% 13531824 ± 2% perf-stat.i.cache-references 0.75 +2.9% 0.78 perf-stat.i.cpi 692709 ± 2% -7.2% 642795 ± 4% perf-stat.i.cycles-between-cache-misses 13369807 -4.0% 12834557 perf-stat.i.dTLB-load-misses 3.619e+10 -2.9% 3.515e+10 perf-stat.i.dTLB-loads 0.00 ± 7% +0.0 0.00 perf-stat.i.dTLB-store-miss-rate% 9252 ± 3% +13.0% 10458 ± 2% perf-stat.i.dTLB-store-misses 1.856e+10 -3.0% 1.801e+10 perf-stat.i.dTLB-stores 89.90 -0.7 89.21 perf-stat.i.iTLB-load-miss-rate% 1440658 ± 2% +8.6% 1564319 ± 2% perf-stat.i.iTLB-loads 1.636e+11 -2.9% 1.589e+11 perf-stat.i.instructions 1.33 -2.9% 1.29 perf-stat.i.ipc 0.08 ± 2% +7.7% 0.09 ± 2% perf-stat.overall.MPKI 0.26 +0.0 0.27 perf-stat.overall.branch-miss-rate% 0.75 +2.9% 0.78 perf-stat.overall.cpi 612206 -6.6% 571711 ± 3% perf-stat.overall.cycles-between-cache-misses 0.00 ± 3% +0.0 0.00 ± 2% perf-stat.overall.dTLB-store-miss-rate% 89.91 -0.7 89.22 perf-stat.overall.iTLB-load-miss-rate% 1.33 -2.9% 1.29 perf-stat.overall.ipc 3.453e+10 -2.9% 3.354e+10 perf-stat.ps.branch-instructions 91321651 -1.3% 90098532 perf-stat.ps.branch-misses 200758 +7.1% 215087 ± 3% perf-stat.ps.cache-misses 12895211 ± 2% +4.6% 13490789 ± 2% perf-stat.ps.cache-references 13324749 -4.0% 12791285 perf-stat.ps.dTLB-load-misses 3.607e+10 -2.9% 3.504e+10 perf-stat.ps.dTLB-loads 9258 ± 3% +13.1% 10473 ± 2% perf-stat.ps.dTLB-store-misses 1.85e+10 -3.0% 1.795e+10 perf-stat.ps.dTLB-stores 1435810 ± 2% +8.6% 1559044 ± 2% perf-stat.ps.iTLB-loads 1.631e+11 -2.9% 1.584e+11 perf-stat.ps.instructions 4.916e+13 -2.9% 4.772e+13 perf-stat.total.instructions 3164 ± 42% +69.6% 5368 ± 9% interrupts.CPU0.NMI:Non-maskable_interrupts 3164 ± 42% +69.6% 5368 ± 9% interrupts.CPU0.PMI:Performance_monitoring_interrupts 3089 ± 42% +72.8% 5338 ± 10% interrupts.CPU1.NMI:Non-maskable_interrupts 3089 ± 42% +72.8% 5338 ± 10% interrupts.CPU1.PMI:Performance_monitoring_interrupts 2442 ± 24% +118.0% 5325 ± 10% interrupts.CPU10.NMI:Non-maskable_interrupts 2442 ± 24% +118.0% 5325 ± 10% interrupts.CPU10.PMI:Performance_monitoring_interrupts 1193 ± 67% -67.0% 394.00 ±130% interrupts.CPU12.RES:Rescheduling_interrupts 145.75 ± 67% -89.5% 15.33 ± 76% interrupts.CPU13.RES:Rescheduling_interrupts 28.50 ± 98% +1725.7% 520.33 ± 65% interrupts.CPU16.RES:Rescheduling_interrupts 2432 ± 24% +120.4% 5360 ± 9% interrupts.CPU17.NMI:Non-maskable_interrupts 2432 ± 24% +120.4% 5360 ± 9% interrupts.CPU17.PMI:Performance_monitoring_interrupts 2450 ± 24% +117.9% 5341 ± 10% interrupts.CPU18.NMI:Non-maskable_interrupts 2450 ± 24% +117.9% 5341 ± 10% interrupts.CPU18.PMI:Performance_monitoring_interrupts 3077 ± 42% +73.3% 5333 ± 10% interrupts.CPU19.NMI:Non-maskable_interrupts 3077 ± 42% +73.3% 5333 ± 10% interrupts.CPU19.PMI:Performance_monitoring_interrupts 3073 ± 42% +72.3% 5295 ± 11% interrupts.CPU2.NMI:Non-maskable_interrupts 3073 ± 42% +72.3% 5295 ± 11% interrupts.CPU2.PMI:Performance_monitoring_interrupts 3422 ± 27% +56.8% 5365 ± 9% interrupts.CPU20.NMI:Non-maskable_interrupts 3422 ± 27% +56.8% 5365 ± 9% interrupts.CPU20.PMI:Performance_monitoring_interrupts 238.00 ±111% +227.2% 778.67 ± 35% interrupts.CPU20.RES:Rescheduling_interrupts 3433 ± 28% +56.1% 5360 ± 9% interrupts.CPU24.NMI:Non-maskable_interrupts 3433 ± 28% +56.1% 5360 ± 9% interrupts.CPU24.PMI:Performance_monitoring_interrupts 3436 ± 28% +55.3% 5336 ± 10% interrupts.CPU25.NMI:Non-maskable_interrupts 3436 ± 28% +55.3% 5336 ± 10% interrupts.CPU25.PMI:Performance_monitoring_interrupts 3428 ± 27% +55.0% 5314 ± 11% interrupts.CPU26.NMI:Non-maskable_interrupts 3428 ± 27% +55.0% 5314 ± 11% interrupts.CPU26.PMI:Performance_monitoring_interrupts 3442 ± 28% +55.2% 5344 ± 10% interrupts.CPU28.NMI:Non-maskable_interrupts 3442 ± 28% +55.2% 5344 ± 10% interrupts.CPU28.PMI:Performance_monitoring_interrupts 3428 ± 27% +88.4% 6461 ± 16% interrupts.CPU29.NMI:Non-maskable_interrupts 3428 ± 27% +88.4% 6461 ± 16% interrupts.CPU29.PMI:Performance_monitoring_interrupts 3103 ± 42% +71.6% 5325 ± 10% interrupts.CPU3.NMI:Non-maskable_interrupts 3103 ± 42% +71.6% 5325 ± 10% interrupts.CPU3.PMI:Performance_monitoring_interrupts 3445 ± 27% +54.6% 5325 ± 10% interrupts.CPU38.NMI:Non-maskable_interrupts 3445 ± 27% +54.6% 5325 ± 10% interrupts.CPU38.PMI:Performance_monitoring_interrupts 3086 ± 42% +74.5% 5387 ± 8% interrupts.CPU4.NMI:Non-maskable_interrupts 3086 ± 42% +74.5% 5387 ± 8% interrupts.CPU4.PMI:Performance_monitoring_interrupts 3078 ± 42% +109.6% 6452 ± 15% interrupts.CPU5.NMI:Non-maskable_interrupts 3078 ± 42% +109.6% 6452 ± 15% interrupts.CPU5.PMI:Performance_monitoring_interrupts 3085 ± 42% +73.7% 5361 ± 9% interrupts.CPU6.NMI:Non-maskable_interrupts 3085 ± 42% +73.7% 5361 ± 9% interrupts.CPU6.PMI:Performance_monitoring_interrupts 3064 ± 41% +75.6% 5380 ± 9% interrupts.CPU7.NMI:Non-maskable_interrupts 3064 ± 41% +75.6% 5380 ± 9% interrupts.CPU7.PMI:Performance_monitoring_interrupts 618.50 ± 95% -97.5% 15.33 ±100% interrupts.CPU7.RES:Rescheduling_interrupts 3092 ± 43% +71.3% 5297 ± 11% interrupts.CPU8.NMI:Non-maskable_interrupts 3092 ± 43% +71.3% 5297 ± 11% interrupts.CPU8.PMI:Performance_monitoring_interrupts 3087 ± 42% +71.3% 5287 ± 11% interrupts.CPU9.NMI:Non-maskable_interrupts 3087 ± 42% +71.3% 5287 ± 11% interrupts.CPU9.PMI:Performance_monitoring_interrupts 485459 ± 7% +13.3% 549830 ± 8% interrupts.NMI:Non-maskable_interrupts 485459 ± 7% +13.3% 549830 ± 8% interrupts.PMI:Performance_monitoring_interrupts will-it-scale.per_process_ops 350000 +-+----------------------------------------------------------------+ |..O.O..O.O..+.O O + O +.O..+..O.O..O.+..O.+..+..+.+..+.+..+.+..| 300000 +-+ : : : | | : : : | 250000 +-+ : :: : | | : : : : | 200000 +-+ : : : : | | : : : : | 150000 +-+ : : : : | | : : : : | 100000 +-+ : : : : | | : : :: | 50000 +-+ : : | | : : | 0 O-+----------O-------O----O----O---------O-------------------------+ will-it-scale.workload 1.6e+07 +-+---------------------------------------------------------------+ |..+.+..+.+..+.+ + +..+.+..+..+.+..+.+..+.+..+.+..+.+..+.+..| 1.4e+07 +-+O O O O O O : O : O O O O O | 1.2e+07 +-+ : : : | | : :: : | 1e+07 +-+ : : : : | | : : : : | 8e+06 +-+ : : : : | | : : : : | 6e+06 +-+ : : : : | 4e+06 +-+ : : : : | | :: :: | 2e+06 +-+ : : | | : : | 0 O-+----------O------O----O----O----------O------------------------+ [*] bisect-good sample [O] bisect-bad sample *************************************************************************************************** lkp-skl-fpga01: 104 threads Skylake with 192G memory ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-7/performance/x86_64-rhel-7.6/process/100%/debian-x86_64-2018-04-03.cgz/lkp-skl-fpga01/poll2/will-it-scale commit: b929a500d6 ("x86/realmode: Don't leak the trampoline kernel address") a9d57ef15c ("x86/retpolines: Disable switch jump tables when retpolines are enabled") b929a500d6847916 a9d57ef15cbe327fe54416dd194 ---------------- --------------------------- fail:runs %reproduction fail:runs | | | 2:4 -50% :4 dmesg.WARNING:at_ip_do_sys_poll/0x %stddev %change %stddev \ | \ 232834 +1.9% 237246 will-it-scale.per_process_ops 24214809 +1.9% 24673704 will-it-scale.workload 6197 ± 63% -87.7% 762.25 ± 57% proc-vmstat.numa_hint_faults 22.54 ± 8% -16.8% 18.75 ± 3% sched_debug.cpu.cpu_load[2].max 29.57 +1.2% 29.91 boot-time.boot 2498 +1.9% 2547 boot-time.idle 4158 ± 4% -10.5% 3723 ± 5% slabinfo.sock_inode_cache.active_objs 4158 ± 4% -10.5% 3723 ± 5% slabinfo.sock_inode_cache.num_objs 154503 ± 5% +397.0% 767833 ±133% cpuidle.C1.time 36996508 ± 8% -56.2% 16209414 ± 80% cpuidle.C1E.time 86639 ± 6% -42.1% 50184 ± 51% cpuidle.C1E.usage 97889223 +18.1% 1.156e+08 ± 10% cpuidle.C6.time 2726 +1.5% 2767 turbostat.Avg_MHz 84477 ± 8% -43.7% 47581 ± 55% turbostat.C1E 0.12 ± 11% -0.1 0.05 ± 81% turbostat.C1E% 0.29 ± 2% +0.1 0.35 ± 10% turbostat.C6% 0.75 -43.2% 0.43 ± 9% turbostat.CPU%c1 0.16 ± 4% +34.4% 0.22 ± 12% turbostat.CPU%c6 197.15 -81.0 116.19 ± 7% turbostat.PKG_% 576.75 ± 43% -70.4% 171.00 ± 39% interrupts.CPU17.RES:Rescheduling_interrupts 758.00 ± 72% -65.4% 262.50 ±108% interrupts.CPU50.RES:Rescheduling_interrupts 5757 ± 2% -14.2% 4941 ± 4% interrupts.CPU51.RES:Rescheduling_interrupts 174.50 ±140% -87.7% 21.50 ± 33% interrupts.CPU64.RES:Rescheduling_interrupts 45.50 ± 95% +185.2% 129.75 ± 70% interrupts.CPU65.RES:Rescheduling_interrupts 97.25 ±115% +460.9% 545.50 ± 65% interrupts.CPU84.RES:Rescheduling_interrupts 281.50 ± 68% -86.5% 38.00 ± 44% interrupts.CPU87.RES:Rescheduling_interrupts 40.00 ± 28% +138.8% 95.50 ± 41% interrupts.CPU95.RES:Rescheduling_interrupts 800.00 ±102% -96.4% 29.00 ± 33% interrupts.CPU96.RES:Rescheduling_interrupts 171.25 ± 3% -47.9% 89.25 ± 40% interrupts.TLB:TLB_shootdowns 18.91 -0.7 18.20 perf-profile.calltrace.cycles-pp.__fget_light.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 90.57 -0.7 89.90 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 90.36 -0.7 89.69 perf-profile.calltrace.cycles-pp.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 90.74 -0.7 90.08 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.poll 95.88 -0.6 95.27 perf-profile.calltrace.cycles-pp.poll 88.47 -0.5 88.00 perf-profile.calltrace.cycles-pp.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe.poll 2.74 -0.4 2.34 perf-profile.calltrace.cycles-pp.__fdget.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.82 ± 2% -0.1 0.77 ± 4% perf-profile.calltrace.cycles-pp.kfree.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.52 +0.0 0.55 perf-profile.calltrace.cycles-pp.__virt_addr_valid.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64 1.21 +0.1 1.27 perf-profile.calltrace.cycles-pp.__check_object_size.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 2.64 +0.1 2.71 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.poll 1.76 +0.1 1.88 perf-profile.calltrace.cycles-pp.copy_user_enhanced_fast_string._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64 2.52 +0.2 2.68 perf-profile.calltrace.cycles-pp._copy_from_user.do_sys_poll.__x64_sys_poll.do_syscall_64.entry_SYSCALL_64_after_hwframe 3.65 +0.6 4.24 perf-profile.calltrace.cycles-pp.testcase 5.765e+10 +2.0% 5.878e+10 perf-stat.i.branch-instructions 0.27 +0.0 0.28 perf-stat.i.branch-miss-rate% 1.537e+08 +6.2% 1.633e+08 perf-stat.i.branch-misses 2.83e+11 +1.5% 2.874e+11 perf-stat.i.cpu-cycles 5.912e+10 +1.9% 6.026e+10 perf-stat.i.dTLB-loads 2.858e+10 +1.9% 2.912e+10 perf-stat.i.dTLB-stores 24052836 +1.8% 24490119 perf-stat.i.iTLB-load-misses 2.723e+11 +2.0% 2.776e+11 perf-stat.i.instructions 0.27 +0.0 0.28 perf-stat.overall.branch-miss-rate% 5.746e+10 +2.0% 5.859e+10 perf-stat.ps.branch-instructions 1.532e+08 +6.2% 1.627e+08 perf-stat.ps.branch-misses 2.821e+11 +1.5% 2.864e+11 perf-stat.ps.cpu-cycles 5.892e+10 +1.9% 6.006e+10 perf-stat.ps.dTLB-loads 2.848e+10 +1.9% 2.902e+10 perf-stat.ps.dTLB-stores 23972265 +1.8% 24408198 perf-stat.ps.iTLB-load-misses 2.713e+11 +2.0% 2.767e+11 perf-stat.ps.instructions 8.175e+13 +1.9% 8.333e+13 perf-stat.total.instructions 6416 ± 3% +50.9% 9679 ± 24% softirqs.CPU103.RCU 7745 ± 5% +19.5% 9256 ± 19% softirqs.CPU12.RCU 8620 ± 4% +13.0% 9740 ± 5% softirqs.CPU25.RCU 8519 ± 2% +52.0% 12944 ± 32% softirqs.CPU33.RCU 8146 ± 2% +13.8% 9267 ± 9% softirqs.CPU35.RCU 7760 ± 2% +22.8% 9531 ± 20% softirqs.CPU39.RCU 8021 ± 4% +26.9% 10183 ± 18% softirqs.CPU40.RCU 7931 ± 3% +15.6% 9167 ± 10% softirqs.CPU42.RCU 7759 ± 2% +36.7% 10607 ± 23% softirqs.CPU44.RCU 7746 ± 3% +23.5% 9570 ± 23% softirqs.CPU45.RCU 7697 ± 2% +19.7% 9215 ± 6% softirqs.CPU47.RCU 8387 ± 2% +12.0% 9398 ± 6% softirqs.CPU58.RCU 7738 ± 3% +19.8% 9267 ± 5% softirqs.CPU60.RCU 7418 ± 3% +34.0% 9937 ± 22% softirqs.CPU72.RCU 7224 ± 2% +28.3% 9269 ± 18% softirqs.CPU82.RCU 7274 ± 2% +24.5% 9055 ± 20% softirqs.CPU88.RCU 7527 ± 3% +36.6% 10284 ± 16% softirqs.CPU89.RCU 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