Greeting, FYI, we noticed a 15.5% improvement of stress-ng.msg.ops_per_sec due to commit: commit: c5654df66d65f6b5f8967f15a0b61f89acb5941e ("[PATCH 09/13] nfs: define and use the NFS_INO_INVALID_XATTR flag") url: https://github.com/0day-ci/linux/commits/Frank-van-der-Linden/client-side-user-xattr-RFC8276-support/20200312-064740 base: git://git.linux-nfs.org/projects/trondmy/linux-nfs.git linux-next in testcase: stress-ng on test machine: 96 threads Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz with 192G memory with following parameters: nr_threads: 100% disk: 1HDD testtime: 1s class: scheduler cpufreq_governor: performance ucode: 0x500002c sc_pid_max: 4194304 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 ========================================================================================= class/compiler/cpufreq_governor/disk/kconfig/nr_threads/rootfs/sc_pid_max/tbox_group/testcase/testtime/ucode: scheduler/gcc-7/performance/1HDD/x86_64-rhel-7.6/100%/debian-x86_64-20191114.cgz/4194304/lkp-csl-2sp5/stress-ng/1s/0x500002c commit: 02936a79b1 ("nfs: modify update_changeattr to deal with regular files") c5654df66d ("nfs: define and use the NFS_INO_INVALID_XATTR flag") 02936a79b108f25d c5654df66d65f6b5f8967f15a0b ---------------- --------------------------- %stddev %change %stddev \ | \ 5946708 +15.6% 6876526 stress-ng.msg.ops 5872043 +15.5% 6782979 stress-ng.msg.ops_per_sec 6174 -2.0% 6050 stress-ng.time.percent_of_cpu_this_job_got 2707 -3.7% 2606 ± 2% stress-ng.time.system_time 14823 ± 3% -6.2% 13899 ± 4% proc-vmstat.nr_page_table_pages 1.702e+08 ± 12% +209.8% 5.272e+08 ± 62% cpuidle.C6.time 250855 ± 9% +146.5% 618483 ± 55% cpuidle.C6.usage 2424560 ± 2% -3.2% 2347822 perf-stat.i.context-switches 4214 ± 4% -8.1% 3873 ± 2% perf-stat.i.instructions-per-iTLB-miss 0.27 +4.0% 0.29 perf-stat.i.ipc 0.24 +3.1% 0.25 perf-stat.overall.ipc 13130807 +1.2% 13289826 perf-stat.ps.iTLB-loads 0.17 ± 9% -20.1% 0.14 ± 8% sched_debug.cfs_rq:/.nr_running.avg 0.40 ± 3% -10.0% 0.36 ± 5% sched_debug.cfs_rq:/.nr_running.stddev 246.08 ± 9% -20.2% 196.28 ± 6% sched_debug.cpu.curr->pid.avg 638.88 ± 4% -9.8% 576.10 ± 3% sched_debug.cpu.curr->pid.stddev 0.14 ± 8% -20.4% 0.11 ± 4% sched_debug.cpu.nr_running.avg 615.25 ± 12% -43.1% 350.25 ± 8% sched_debug.cpu.nr_switches.min 3332131 ± 3% -33.7% 2208077 ± 20% meminfo.Active 3329127 ± 3% -33.8% 2205066 ± 20% meminfo.Active(anon) 2232223 ± 4% -33.3% 1488680 ± 25% meminfo.AnonHugePages 2612610 ± 3% -37.3% 1639048 ± 28% meminfo.AnonPages 7991735 ± 5% -19.6% 6426850 ± 10% meminfo.Memused 62476 ± 3% -14.1% 53672 ± 7% meminfo.PageTables 1804845 -44.7% 998859 ± 45% meminfo.max_used_kB 25157 ± 2% -14.3% 21559 ± 8% slabinfo.Acpi-State.active_objs 25302 ± 2% -14.2% 21702 ± 8% slabinfo.Acpi-State.num_objs 50827 ± 6% -14.3% 43541 ± 3% slabinfo.filp.active_objs 1655 ± 5% -14.1% 1422 ± 3% slabinfo.filp.active_slabs 52972 ± 5% -14.1% 45525 ± 3% slabinfo.filp.num_objs 1655 ± 5% -14.1% 1422 ± 3% slabinfo.filp.num_slabs 650.00 +19.7% 778.00 ± 10% slabinfo.kmem_cache_node.active_objs 704.00 +18.2% 832.00 ± 9% slabinfo.kmem_cache_node.num_objs 5747 ± 8% +11.4% 6400 ± 7% slabinfo.task_struct.active_objs 5753 ± 8% +11.3% 6405 ± 7% slabinfo.task_struct.active_slabs 5753 ± 8% +11.3% 6405 ± 7% slabinfo.task_struct.num_objs 5753 ± 8% +11.3% 6405 ± 7% slabinfo.task_struct.num_slabs 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.write._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.devkmsg_write.new_sync_write.vfs_write.ksys_write.do_syscall_64 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.devkmsg_emit.devkmsg_write.new_sync_write.vfs_write.ksys_write 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.vprintk_emit.devkmsg_emit.devkmsg_write.new_sync_write.vfs_write 13.85 ±135% -3.7 10.19 ±173% perf-profile.calltrace.cycles-pp.console_unlock.vprintk_emit.devkmsg_emit.devkmsg_write.new_sync_write 11.26 ±142% -2.6 8.64 ±173% perf-profile.calltrace.cycles-pp.serial8250_console_write.console_unlock.vprintk_emit.devkmsg_emit.devkmsg_write 0.87 ±103% +1.8 2.70 ± 26% perf-profile.calltrace.cycles-pp.do_group_exit.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.87 ±103% +1.8 2.70 ± 26% perf-profile.calltrace.cycles-pp.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 0.00 +1.9 1.90 ± 22% perf-profile.calltrace.cycles-pp.mmput.do_exit.do_group_exit.__x64_sys_exit_group.do_syscall_64 0.00 +1.9 1.90 ± 22% perf-profile.calltrace.cycles-pp.exit_mmap.mmput.do_exit.do_group_exit.__x64_sys_exit_group 0.87 ±103% +2.4 3.30 ± 33% perf-profile.calltrace.cycles-pp.do_exit.do_group_exit.__x64_sys_exit_group.do_syscall_64.entry_SYSCALL_64_after_hwframe 13.85 ±135% -3.7 10.19 ±173% perf-profile.children.cycles-pp._fini 13.85 ±135% -3.7 10.19 ±173% perf-profile.children.cycles-pp.devkmsg_write 13.85 ±135% -3.7 10.19 ±173% perf-profile.children.cycles-pp.devkmsg_emit 13.85 ±135% -3.7 10.19 ±173% perf-profile.children.cycles-pp.vprintk_emit 13.85 ±135% -3.7 10.19 ±173% perf-profile.children.cycles-pp.console_unlock 11.26 ±142% -2.6 8.64 ±173% perf-profile.children.cycles-pp.serial8250_console_write 10.19 ±139% -2.2 8.03 ±173% perf-profile.children.cycles-pp.wait_for_xmitr 0.36 ±173% +1.9 2.21 ± 9% perf-profile.children.cycles-pp.__sched_text_start 1.02 ±100% +2.2 3.21 ± 24% perf-profile.children.cycles-pp.mmput 1.02 ±100% +2.2 3.21 ± 24% perf-profile.children.cycles-pp.exit_mmap 0.00 +2.2 2.21 ± 9% perf-profile.children.cycles-pp.pick_next_task_fair 0.87 ±103% +2.4 3.30 ± 33% perf-profile.children.cycles-pp.__x64_sys_exit_group 3047 ± 7% +30.9% 3988 ± 10% interrupts.CPU0.CAL:Function_call_interrupts 1608 ± 11% +71.7% 2761 ± 32% interrupts.CPU0.TLB:TLB_shootdowns 45602 ± 22% -56.6% 19783 ± 78% interrupts.CPU1.RES:Rescheduling_interrupts 31367 ± 58% -44.0% 17553 ± 86% interrupts.CPU12.RES:Rescheduling_interrupts 29637 ± 60% -66.0% 10090 ± 20% interrupts.CPU14.RES:Rescheduling_interrupts 1808 ± 16% +70.7% 3086 ± 26% interrupts.CPU14.TLB:TLB_shootdowns 10750 ± 11% +143.2% 26140 ± 57% interrupts.CPU19.RES:Rescheduling_interrupts 2808 ± 6% +15.0% 3229 ± 7% interrupts.CPU23.CAL:Function_call_interrupts 4405 ± 12% -25.0% 3302 ± 14% interrupts.CPU24.CAL:Function_call_interrupts 582.00 ± 23% -69.5% 177.75 ±110% interrupts.CPU24.IWI:IRQ_work_interrupts 4969 ± 16% -36.6% 3149 ± 12% interrupts.CPU25.CAL:Function_call_interrupts 492.75 ± 18% -71.8% 138.75 ±101% interrupts.CPU25.IWI:IRQ_work_interrupts 4057 ± 8% -29.0% 2879 ± 11% interrupts.CPU26.CAL:Function_call_interrupts 447.00 ± 76% -76.2% 106.50 ± 98% interrupts.CPU26.IWI:IRQ_work_interrupts 158.25 ± 21% -71.6% 45.00 ±111% interrupts.CPU27.IWI:IRQ_work_interrupts 13335 ± 16% -22.6% 10316 ± 17% interrupts.CPU30.RES:Rescheduling_interrupts 161649 ± 7% +172.6% 440679 ± 64% interrupts.CPU4.LOC:Local_timer_interrupts 2350 ± 21% -35.6% 1513 ± 26% interrupts.CPU44.TLB:TLB_shootdowns 2288 ± 24% -39.1% 1393 ± 29% interrupts.CPU45.TLB:TLB_shootdowns 3108 ± 8% -12.3% 2725 ± 6% interrupts.CPU46.CAL:Function_call_interrupts 12381 ± 9% +96.0% 24268 ± 68% interrupts.CPU46.RES:Rescheduling_interrupts 2537 ± 24% -39.2% 1541 ± 22% interrupts.CPU46.TLB:TLB_shootdowns 2987 ± 12% +32.3% 3952 ± 4% interrupts.CPU49.CAL:Function_call_interrupts 12127 ± 13% +118.7% 26525 ± 49% interrupts.CPU57.RES:Rescheduling_interrupts 2711 ± 13% +30.5% 3539 ± 17% interrupts.CPU58.CAL:Function_call_interrupts 41057 ± 40% -73.6% 10837 ± 14% interrupts.CPU58.RES:Rescheduling_interrupts 1784 ± 15% +40.3% 2503 ± 18% interrupts.CPU60.TLB:TLB_shootdowns 35545 ± 44% -45.1% 19500 ± 82% interrupts.CPU62.RES:Rescheduling_interrupts 3012 ± 3% +23.1% 3708 ± 9% interrupts.CPU64.CAL:Function_call_interrupts 1835 ± 13% +44.1% 2646 ± 21% interrupts.CPU66.TLB:TLB_shootdowns 12893 ± 7% -21.3% 10149 ± 10% interrupts.CPU67.RES:Rescheduling_interrupts 1617 ± 20% +41.5% 2289 ± 17% interrupts.CPU71.TLB:TLB_shootdowns 682.50 ±105% -85.8% 96.75 ±122% interrupts.CPU72.IWI:IRQ_work_interrupts 2163 ± 14% -25.8% 1604 ± 30% interrupts.CPU72.TLB:TLB_shootdowns 297.75 ± 6% -69.9% 89.75 ±116% interrupts.CPU73.IWI:IRQ_work_interrupts 612.00 ±117% -90.3% 59.25 ±101% interrupts.CPU74.IWI:IRQ_work_interrupts 85.00 ± 31% -72.4% 23.50 ±121% interrupts.CPU75.IWI:IRQ_work_interrupts 15016 ± 14% -30.8% 10398 ± 17% interrupts.CPU77.RES:Rescheduling_interrupts 3163 ± 6% -19.0% 2562 ± 11% interrupts.CPU78.CAL:Function_call_interrupts 3118 ± 7% -13.3% 2704 ± 5% interrupts.CPU79.CAL:Function_call_interrupts 28253 ± 70% -56.2% 12372 ± 16% interrupts.CPU81.RES:Rescheduling_interrupts 171332 ± 8% +119.5% 376071 ± 51% interrupts.CPU85.LOC:Local_timer_interrupts 2100 ± 17% -21.4% 1651 ± 9% interrupts.CPU85.TLB:TLB_shootdowns 39800 ± 40% -59.5% 16114 ± 65% interrupts.CPU9.RES:Rescheduling_interrupts 31.75 ±138% -100.0% 0.00 interrupts.CPU90.IWI:IRQ_work_interrupts 2535 ± 30% -36.2% 1616 ± 22% interrupts.CPU93.TLB:TLB_shootdowns 2243465 ± 4% -6.7% 2092631 ± 5% interrupts.RES:Rescheduling_interrupts stress-ng.msg.ops 8e+06 +-------------------------------------------------------------------+ | OO O O OO O O OO O O OO | 7e+06 |-+ O O OO O O OO O O OO | 6e+06 |-+ .+ +.| |.++.+.+.++.+.+.++.+.+.++.+.+.++ +.++.+.+.++.+ ++.+.+ +.+.+.+ | 5e+06 |-+ : : : : | | : : : : | 4e+06 |-+ : : : : | | : : : : | 3e+06 |-+ : : : : | 2e+06 |-+ : : : : | | : : : : | 1e+06 |-+ : : | | : : | 0 +-------------------------------------------------------------------+ stress-ng.msg.ops_per_sec 8e+06 +-------------------------------------------------------------------+ | OO O O OO O O OO O O OO | 7e+06 |-+ O O OO O O OO O O OO | 6e+06 |-+ | |.++.+.+.++.+.+.++.+.+.++.+.+.++ +.++.+.+.++.+ ++.+.+.++.+.+.++.| 5e+06 |-+ : : : : | | : : : : | 4e+06 |-+ : : : : | | : : : : | 3e+06 |-+ : : : : | 2e+06 |-+ : : : : | | : : : : | 1e+06 |-+ : : | | : : | 0 +-------------------------------------------------------------------+ [*] bisect-good sample [O] bisect-bad sample Disclaimer: Results have been estimated based on internal Intel analysis and are provided for informational purposes only. Any difference in system hardware or software design or configuration may affect actual performance. Thanks, Rong Chen