* [PATCH] perf: Remove setting of page->index and ->mapping @ 2024-08-09 20:10 Matthew Wilcox (Oracle) 2024-08-10 22:15 ` Peter Zijlstra 2024-08-16 6:35 ` kernel test robot 0 siblings, 2 replies; 5+ messages in thread From: Matthew Wilcox (Oracle) @ 2024-08-09 20:10 UTC (permalink / raw) To: Peter Zijlstra Cc: Matthew Wilcox (Oracle), Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim, linux-perf-users, linux-kernel I have no idea why perf is setting these fields. There should be no need to do this. Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org> --- kernel/events/core.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index aa3450bdc227..f55ab04f3df0 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6221,8 +6221,6 @@ static vm_fault_t perf_mmap_fault(struct vm_fault *vmf) goto unlock; get_page(vmf->page); - vmf->page->mapping = vmf->vma->vm_file->f_mapping; - vmf->page->index = vmf->pgoff; ret = 0; unlock: -- 2.43.0 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH] perf: Remove setting of page->index and ->mapping 2024-08-09 20:10 [PATCH] perf: Remove setting of page->index and ->mapping Matthew Wilcox (Oracle) @ 2024-08-10 22:15 ` Peter Zijlstra 2024-08-16 6:35 ` kernel test robot 1 sibling, 0 replies; 5+ messages in thread From: Peter Zijlstra @ 2024-08-10 22:15 UTC (permalink / raw) To: Matthew Wilcox (Oracle) Cc: Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim, linux-perf-users, linux-kernel On Fri, Aug 09, 2024 at 09:10:36PM +0100, Matthew Wilcox (Oracle) wrote: > I have no idea why perf is setting these fields. There should be > no need to do this. Because these are not shared pages and nobody will ever look at these fields? Yeah, I can't for the life of me remember why I did that. > Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org> > --- > kernel/events/core.c | 2 -- > 1 file changed, 2 deletions(-) > > diff --git a/kernel/events/core.c b/kernel/events/core.c > index aa3450bdc227..f55ab04f3df0 100644 > --- a/kernel/events/core.c > +++ b/kernel/events/core.c > @@ -6221,8 +6221,6 @@ static vm_fault_t perf_mmap_fault(struct vm_fault *vmf) > goto unlock; > > get_page(vmf->page); > - vmf->page->mapping = vmf->vma->vm_file->f_mapping; > - vmf->page->index = vmf->pgoff; > > ret = 0; > unlock: > -- > 2.43.0 > ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] perf: Remove setting of page->index and ->mapping 2024-08-09 20:10 [PATCH] perf: Remove setting of page->index and ->mapping Matthew Wilcox (Oracle) 2024-08-10 22:15 ` Peter Zijlstra @ 2024-08-16 6:35 ` kernel test robot 2024-08-16 17:51 ` Matthew Wilcox 1 sibling, 1 reply; 5+ messages in thread From: kernel test robot @ 2024-08-16 6:35 UTC (permalink / raw) To: Matthew Wilcox (Oracle) Cc: oe-lkp, lkp, linux-perf-users, linux-kernel, ying.huang, feng.tang, fengwei.yin, Peter Zijlstra, Matthew Wilcox (Oracle), Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim, oliver.sang Hello, from commit message, it seems just some useless setting have been removed. but quite some performance regression bisect by our bot finally point to this change (already removed some not so significant changes to avoid this report mail is too long). we don't have enough knowledge to explain these, so just report FYI. if you have any suggestion for our tests or need us run more tests, please just let us know. thanks a lot! kernel test robot noticed a -28.3% regression of stress-ng.dirmany.files_removed_per_sec on: commit: a9ac22fb7ac3486a5bd89c25a2f784330ee92dd7 ("[PATCH] perf: Remove setting of page->index and ->mapping") url: https://github.com/intel-lab-lkp/linux/commits/Matthew-Wilcox-Oracle/perf-Remove-setting-of-page-index-and-mapping/20240810-111222 base: https://git.kernel.org/cgit/linux/kernel/git/perf/perf-tools-next.git perf-tools-next patch link: https://lore.kernel.org/all/20240809201038.2648058-1-willy@infradead.org/ patch subject: [PATCH] perf: Remove setting of page->index and ->mapping testcase: stress-ng test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory parameters: nr_threads: 100% disk: 1HDD testtime: 60s fs: btrfs test: dirmany cpufreq_governor: performance In addition to that, the commit also has significant impact on the following tests: +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | aim9: aim9.exec_test.ops_per_sec -12.3% regression | | test machine | 48 threads 2 sockets Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz (Ivy Bridge-EP) with 64G memory | | test parameters | cpufreq_governor=performance | | | test=exec_test | | | testtime=300s | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | stress-ng: stress-ng.sem.sem_wait_calls_per_sec -56.8% regression | | test machine | 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 256G memory | | test parameters | cpufreq_governor=performance | | | nr_threads=100% | | | test=sem | | | testtime=60s | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | tbench: tbench.throughput_MB/s -42.5% regression | | test machine | 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory | | test parameters | cluster=cs-localhost | | | cpufreq_governor=performance | | | iterations=1 | | | nr_threads=100% | | | runtime=600 | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | uperf: uperf.Runstat_master_Mbps -25.4% regression | | test machine | 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory | | test parameters | cluster=cs-localhost | | | cpufreq_governor=performance | | | datasize=64 | | | nr_threads=1 | | | protocol=UDP | | | runtime=30s | | | winsize=32k | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | netperf: netperf.Throughput_tps -31.4% regression | | test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory | | test parameters | cluster=cs-localhost | | | cpufreq_governor=performance | | | ip=ipv4 | | | nr_threads=50% | | | runtime=300s | | | test=SCTP_RR | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | will-it-scale: will-it-scale.per_process_ops -61.9% regression | | test machine | 104 threads 2 sockets (Skylake) with 192G memory | | test parameters | cpufreq_governor=performance | | | mode=process | | | nr_task=100% | | | test=context_switch1 | +------------------+--------------------------------------------------------------------------------------------------------+ | testcase: change | fsmark: fsmark.files_per_sec -29.8% regression | | test machine | 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory | | test parameters | cpufreq_governor=performance | | | disk=1BRD_32G | | | filesize=4K | | | fs2=nfsv4 | | | fs=ext4 | | | iterations=1x | | | nr_files_per_directory=1fpd | | | nr_threads=1t | | | sync_method=fsyncBeforeClose | | | test_size=4G | +------------------+--------------------------------------------------------------------------------------------------------+ If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags | Reported-by: kernel test robot <oliver.sang@intel.com> | Closes: https://lore.kernel.org/oe-lkp/202408161305.a3a5158a-oliver.sang@intel.com Details are as below: --------------------------------------------------------------------------------------------------> The kernel config and materials to reproduce are available at: https://download.01.org/0day-ci/archive/20240816/202408161305.a3a5158a-oliver.sang@intel.com ========================================================================================= compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/1HDD/btrfs/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp8/dirmany/stress-ng/60s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 3.832e+09 ± 2% +21.3% 4.647e+09 ± 12% cpuidle..time 0.71 ± 4% +0.2 0.88 ± 3% mpstat.cpu.all.usr% 37575 ± 10% -86.8% 4959 ± 2% meminfo.Active(anon) 121149 ± 3% -28.6% 86511 meminfo.Shmem 103.55 +13.9% 117.98 ± 8% uptime.boot 5993 +13.0% 6775 ± 8% uptime.idle 238142 ± 6% -23.9% 181123 ± 11% vmstat.system.cs 157186 ± 7% +59.3% 250390 ± 10% vmstat.system.in 16677 ± 10% -86.6% 2238 ± 49% numa-meminfo.node0.Active(anon) 20916 ± 15% -87.0% 2724 ± 39% numa-meminfo.node1.Active(anon) 1379973 ± 75% +117.7% 3004793 ± 40% numa-meminfo.node1.FilePages 666634 ±158% +224.1% 2160724 ± 53% numa-meminfo.node1.Unevictable 1586 ± 12% -100.0% 0.00 perf-c2c.DRAM.local 13101 ± 15% -100.0% 0.00 perf-c2c.DRAM.remote 12067 ± 15% -100.0% 0.00 perf-c2c.HITM.local 8050 ± 16% -100.0% 0.00 perf-c2c.HITM.remote 20118 ± 15% -100.0% 0.00 perf-c2c.HITM.total 9403 ± 10% -86.8% 1240 ± 2% proc-vmstat.nr_active_anon 27583 -2.0% 27039 proc-vmstat.nr_mapped 30307 ± 3% -28.6% 21640 proc-vmstat.nr_shmem 9403 ± 10% -86.8% 1240 ± 2% proc-vmstat.nr_zone_active_anon 66223 -18.6% 53878 proc-vmstat.numa_other 263934 ± 3% +1e+05% 2.768e+08 ± 16% proc-vmstat.pgfault 67.43 ± 3% -14.6% 57.57 ± 9% stress-ng.dirmany.%_of_time_creating_files 32.57 ± 6% +30.3% 42.43 ± 12% stress-ng.dirmany.%_of_time_removing_file 703.32 -28.3% 504.06 ± 3% stress-ng.dirmany.files_removed_per_sec 63.61 ± 2% +22.9% 78.17 ± 13% stress-ng.time.elapsed_time 63.61 ± 2% +22.9% 78.17 ± 13% stress-ng.time.elapsed_time.max 422.17 ± 4% -35.4% 272.83 ± 17% stress-ng.time.percent_of_cpu_this_job_got 267.72 ± 5% -21.4% 210.45 ± 17% stress-ng.time.system_time 4177 ± 10% -86.6% 559.65 ± 49% numa-vmstat.node0.nr_active_anon 4177 ± 10% -86.6% 559.65 ± 49% numa-vmstat.node0.nr_zone_active_anon 5236 ± 15% -87.0% 681.46 ± 39% numa-vmstat.node1.nr_active_anon 345110 ± 75% +117.7% 751196 ± 40% numa-vmstat.node1.nr_file_pages 164075 ± 3% +16.8% 191570 ± 14% numa-vmstat.node1.nr_slab_reclaimable 166658 ±158% +224.1% 540181 ± 53% numa-vmstat.node1.nr_unevictable 5236 ± 15% -87.0% 681.46 ± 39% numa-vmstat.node1.nr_zone_active_anon 166658 ±158% +224.1% 540181 ± 53% numa-vmstat.node1.nr_zone_unevictable 76822 ± 10% +82.2% 140007 ± 19% sched_debug.cfs_rq:/.avg_vruntime.max 8327 ± 16% +138.0% 19815 ± 8% sched_debug.cfs_rq:/.avg_vruntime.stddev 2.25 ± 27% -100.0% 0.00 sched_debug.cfs_rq:/.load_avg.min 76822 ± 10% +82.2% 140007 ± 19% sched_debug.cfs_rq:/.min_vruntime.max 8327 ± 16% +138.0% 19815 ± 8% sched_debug.cfs_rq:/.min_vruntime.stddev 19.67 ± 57% -98.7% 0.25 ±223% sched_debug.cfs_rq:/.runnable_avg.min 217.81 ± 3% +27.8% 278.40 ± 4% sched_debug.cfs_rq:/.runnable_avg.stddev 1006 ± 7% +16.2% 1169 ± 6% sched_debug.cfs_rq:/.util_avg.max 19.75 ± 57% -99.2% 0.17 ±223% sched_debug.cfs_rq:/.util_avg.min 216.49 ± 3% +27.8% 276.68 ± 4% sched_debug.cfs_rq:/.util_avg.stddev 28.42 ± 11% +58.1% 44.94 ± 14% sched_debug.cfs_rq:/.util_est.avg 447.25 ± 12% +71.0% 764.67 sched_debug.cfs_rq:/.util_est.max 90.88 ± 9% +85.0% 168.09 ± 5% sched_debug.cfs_rq:/.util_est.stddev 415102 ± 6% +27.5% 529068 sched_debug.cpu.avg_idle.avg 752805 ± 25% +49.8% 1127548 ± 6% sched_debug.cpu.avg_idle.max 154552 ± 6% +53.2% 236733 ± 6% sched_debug.cpu.avg_idle.stddev 101993 ± 13% -40.9% 60236 ± 32% sched_debug.cpu.nr_switches.avg 113676 ± 14% -33.7% 75369 ± 27% sched_debug.cpu.nr_switches.max 93486 ± 14% -87.3% 11896 ± 26% sched_debug.cpu.nr_switches.min 3929 ± 34% +221.3% 12624 ± 28% sched_debug.cpu.nr_switches.stddev 2.27 ± 2% -64.5% 0.81 ± 11% perf-stat.i.MPKI 1.604e+09 ± 2% +138.1% 3.819e+09 ± 6% perf-stat.i.branch-instructions 2.33 ± 2% -1.6 0.72 ± 6% perf-stat.i.branch-miss-rate% 36868744 ± 3% -29.2% 26089842 ± 7% perf-stat.i.branch-misses 244416 ± 6% -24.1% 185618 ± 12% perf-stat.i.context-switches 1.90 ± 3% -39.4% 1.16 ± 3% perf-stat.i.cpi 10537 ± 6% -78.5% 2262 ± 46% perf-stat.i.cpu-migrations 1308 ± 11% +411.3% 6692 ± 26% perf-stat.i.cycles-between-cache-misses 8.205e+09 ± 2% +125.6% 1.851e+10 ± 6% perf-stat.i.instructions 0.88 ± 3% +11.6% 0.98 ± 2% perf-stat.i.ipc 3.78 ± 6% +2824.0% 110.64 ± 3% perf-stat.i.metric.K/sec 3035 ± 2% +1.1e+05% 3448996 ± 3% perf-stat.i.minor-faults 3035 ± 2% +1.1e+05% 3448996 ± 3% perf-stat.i.page-faults 2.82 ± 2% -60.8% 1.10 ± 7% perf-stat.overall.MPKI 2.30 ± 2% -1.6 0.68 ± 7% perf-stat.overall.branch-miss-rate% 2.58 ± 2% -51.0% 1.26 ± 3% perf-stat.overall.cpi 915.45 ± 4% +25.5% 1148 ± 5% perf-stat.overall.cycles-between-cache-misses 0.39 ± 2% +104.3% 0.79 ± 3% perf-stat.overall.ipc 1.579e+09 ± 2% +138.7% 3.769e+09 ± 6% perf-stat.ps.branch-instructions 36312783 ± 3% -29.1% 25735794 ± 7% perf-stat.ps.branch-misses 240154 ± 6% -23.7% 183188 ± 12% perf-stat.ps.context-switches 10351 ± 6% -78.4% 2237 ± 46% perf-stat.ps.cpu-migrations 8.078e+09 ± 2% +126.1% 1.827e+10 ± 6% perf-stat.ps.instructions 2990 ± 2% +1.1e+05% 3403503 ± 3% perf-stat.ps.minor-faults 2990 ± 2% +1.1e+05% 3403503 ± 3% perf-stat.ps.page-faults 5.224e+11 ± 2% +178.3% 1.454e+12 ± 15% perf-stat.total.instructions 24.93 ± 18% -24.9 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 24.55 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction 24.55 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction 24.54 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents 24.51 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.fsync 24.50 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.50 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.fsync 24.49 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.49 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_sync_file.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.48 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe 24.43 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync.do_syscall_64 24.43 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync 24.41 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file 23.94 ± 56% -23.9 0.00 perf-profile.calltrace.cycles-pp.submit_eb_page.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range 23.50 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 23.49 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 23.48 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 22.12 ± 17% -22.1 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 21.95 ± 19% -22.0 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 21.61 ± 20% -21.6 0.00 perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 20.64 ± 17% -20.6 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 20.29 ± 57% -20.3 0.00 perf-profile.calltrace.cycles-pp.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc 20.29 ± 57% -20.3 0.00 perf-profile.calltrace.cycles-pp.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages.do_writepages 17.17 ± 57% -17.2 0.00 perf-profile.calltrace.cycles-pp.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork_asm 13.30 ± 27% -13.3 0.00 perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 12.54 ± 58% -12.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page 12.31 ± 58% -12.3 0.00 perf-profile.calltrace.cycles-pp.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio 11.98 ± 32% -12.0 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 10.42 ± 37% -10.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 8.48 ± 57% -8.5 0.00 perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 7.77 ± 62% -7.8 0.00 perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 7.46 ± 58% -7.5 0.00 perf-profile.calltrace.cycles-pp.check_leaf_item.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk 5.40 ± 69% -5.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_posted_msi_notification.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 4.81 ± 57% -4.8 0.00 perf-profile.calltrace.cycles-pp.check_dir_item.check_leaf_item.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio 4.62 ± 36% -4.6 0.00 perf-profile.calltrace.cycles-pp.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state 4.52 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page 4.51 ± 56% -4.5 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 4.49 ± 56% -4.5 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 4.46 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.crc32c_pcl_intel_update.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio 4.46 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.crc_pcl.crc32c_pcl_intel_update.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk 55.50 ± 15% -55.5 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 55.46 ± 15% -55.5 0.00 perf-profile.children.cycles-pp.do_syscall_64 28.46 ± 53% -28.5 0.00 perf-profile.children.cycles-pp.btrfs_commit_transaction 24.93 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.common_startup_64 24.93 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 24.92 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.do_idle 24.61 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btrfs_write_and_wait_transaction 24.61 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btrfs_write_marked_extents 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.__filemap_fdatawrite_range 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btree_write_cache_pages 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.do_writepages 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.filemap_fdatawrite_wbc 24.51 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.fsync 24.49 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.__x64_sys_fsync 24.49 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.btrfs_sync_file 24.00 ± 55% -24.0 0.00 perf-profile.children.cycles-pp.submit_eb_page 23.53 ± 19% -23.5 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 23.50 ± 16% -23.5 0.00 perf-profile.children.cycles-pp.start_secondary 22.05 ± 19% -22.1 0.00 perf-profile.children.cycles-pp.cpuidle_enter 22.00 ± 19% -22.0 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 21.66 ± 20% -21.7 0.00 perf-profile.children.cycles-pp.acpi_idle_enter 21.63 ± 20% -21.6 0.00 perf-profile.children.cycles-pp.acpi_safe_halt 20.41 ± 55% -20.4 0.00 perf-profile.children.cycles-pp.btrfs_submit_bio 20.41 ± 55% -20.4 0.00 perf-profile.children.cycles-pp.btrfs_submit_chunk 17.35 ± 55% -17.4 0.00 perf-profile.children.cycles-pp.btree_csum_one_bio 14.74 ± 31% -14.7 0.00 perf-profile.children.cycles-pp.ret_from_fork_asm 14.71 ± 31% -14.7 0.00 perf-profile.children.cycles-pp.ret_from_fork 14.62 ± 31% -14.6 0.00 perf-profile.children.cycles-pp.kthread 12.66 ± 56% -12.7 0.00 perf-profile.children.cycles-pp.__btrfs_check_leaf 12.66 ± 56% -12.7 0.00 perf-profile.children.cycles-pp.btrfs_check_leaf 8.52 ± 29% -8.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_call_function_single 8.48 ± 57% -8.5 0.00 perf-profile.children.cycles-pp.worker_thread 7.78 ± 62% -7.8 0.00 perf-profile.children.cycles-pp.process_one_work 7.62 ± 56% -7.6 0.00 perf-profile.children.cycles-pp.check_leaf_item 7.50 ± 34% -7.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt 5.58 ± 53% -5.6 0.00 perf-profile.children.cycles-pp.crc32c_pcl_intel_update 5.20 ± 46% -5.2 0.00 perf-profile.children.cycles-pp.__irq_exit_rcu 5.11 ± 47% -5.1 0.00 perf-profile.children.cycles-pp.handle_softirqs 5.09 ± 54% -5.1 0.00 perf-profile.children.cycles-pp.check_dir_item 4.74 ± 34% -4.7 0.00 perf-profile.children.cycles-pp.sysvec_call_function_single 4.57 ± 53% -4.6 0.00 perf-profile.children.cycles-pp.csum_tree_block 4.54 ± 70% -4.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_posted_msi_notification 4.49 ± 53% -4.5 0.00 perf-profile.children.cycles-pp.crc_pcl 4.45 ± 56% -4.5 0.00 perf-profile.children.cycles-pp.btrfs_get_32 4.06 ± 34% -4.1 0.00 perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt 8.89 ± 14% -8.9 0.00 perf-profile.self.cycles-pp.acpi_safe_halt 4.47 ± 53% -4.5 0.00 perf-profile.self.cycles-pp.crc_pcl *************************************************************************************************** lkp-ivb-2ep2: 48 threads 2 sockets Intel(R) Xeon(R) CPU E5-2697 v2 @ 2.70GHz (Ivy Bridge-EP) with 64G memory ========================================================================================= compiler/cpufreq_governor/kconfig/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/x86_64-rhel-8.3/debian-12-x86_64-20240206.cgz/lkp-ivb-2ep2/exec_test/aim9/300s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 4849014 +148.9% 12069624 cpuidle..usage 29121 ± 41% -62.2% 11007 ± 94% numa-numastat.node1.other_node 97.58 -6.9% 90.83 vmstat.cpu.id 2.30 +115.5% 4.95 vmstat.procs.r 8730 -13.2% 7575 vmstat.system.cs 14729 +456.7% 82001 vmstat.system.in 124849 ± 2% -91.8% 10221 meminfo.Active 124801 ± 2% -91.8% 10173 meminfo.Active(anon) 876134 -9.4% 793689 meminfo.Committed_AS 7544 +20.4% 9085 meminfo.PageTables 128860 ± 2% -91.1% 11481 ± 2% meminfo.Shmem 15.00 ± 19% -100.0% 0.00 perf-c2c.DRAM.local 36.00 ± 15% -100.0% 0.00 perf-c2c.DRAM.remote 2.67 ± 51% -100.0% 0.00 perf-c2c.HIT.remote 64.83 ± 3% -100.0% 0.00 perf-c2c.HITM.local 30.00 ± 15% -100.0% 0.00 perf-c2c.HITM.remote 1224 -12.3% 1074 aim9.exec_test.ops_per_sec 31706751 -12.9% 27603136 aim9.time.minor_page_faults 97.00 -1.0% 96.00 aim9.time.percent_of_cpu_this_job_got 176.72 +14.8% 202.90 aim9.time.system_time 116.68 -25.8% 86.55 ± 4% aim9.time.user_time 733520 -12.2% 643730 aim9.time.voluntary_context_switches 0.25 +1.2 1.42 mpstat.cpu.all.irq% 0.07 ± 2% +0.1 0.12 ± 2% mpstat.cpu.all.soft% 1.66 +3.8 5.46 mpstat.cpu.all.sys% 0.68 +1.7 2.36 mpstat.cpu.all.usr% 1.00 +14516.7% 146.17 ± 51% mpstat.max_utilization.seconds 8.64 +20.2% 10.38 mpstat.max_utilization_pct 9398 ± 74% -82.2% 1675 ± 57% numa-meminfo.node0.Active 9366 ± 74% -82.3% 1659 ± 58% numa-meminfo.node0.Active(anon) 9860 ± 70% -78.5% 2121 ± 48% numa-meminfo.node0.Shmem 115389 ± 6% -92.6% 8534 ± 11% numa-meminfo.node1.Active 115373 ± 6% -92.6% 8502 ± 11% numa-meminfo.node1.Active(anon) 522369 ± 30% -26.9% 381731 ± 30% numa-meminfo.node1.AnonPages.max 118964 ± 5% -92.1% 9348 ± 11% numa-meminfo.node1.Shmem 2345 ± 74% -82.3% 414.77 ± 58% numa-vmstat.node0.nr_active_anon 2466 ± 70% -78.5% 530.32 ± 48% numa-vmstat.node0.nr_shmem 2345 ± 74% -82.3% 414.77 ± 58% numa-vmstat.node0.nr_zone_active_anon 28843 ± 6% -92.6% 2125 ± 11% numa-vmstat.node1.nr_active_anon 29739 ± 5% -92.1% 2337 ± 11% numa-vmstat.node1.nr_shmem 28843 ± 6% -92.6% 2125 ± 11% numa-vmstat.node1.nr_zone_active_anon 29121 ± 41% -62.2% 11007 ± 94% numa-vmstat.node1.numa_other 519.91 ± 29% +39.3% 724.40 ± 3% sched_debug.IIkworker/R-0000 11208 ± 3% +249.5% 39175 ± 19% sched_debug.cfs_rq:/.avg_vruntime.avg 52206 ± 19% +419.6% 271275 ± 14% sched_debug.cfs_rq:/.avg_vruntime.max 9447 ± 10% +573.4% 63622 ± 17% sched_debug.cfs_rq:/.avg_vruntime.stddev 221530 ± 22% -59.4% 89942 ± 48% sched_debug.cfs_rq:/.load.max 39358 ± 19% -45.8% 21349 ± 35% sched_debug.cfs_rq:/.load.stddev 11208 ± 3% +249.5% 39175 ± 19% sched_debug.cfs_rq:/.min_vruntime.avg 52206 ± 19% +419.6% 271275 ± 14% sched_debug.cfs_rq:/.min_vruntime.max 9447 ± 10% +573.4% 63622 ± 17% sched_debug.cfs_rq:/.min_vruntime.stddev 7.04 ± 16% +352.0% 31.80 ± 20% sched_debug.cfs_rq:/.util_est.avg 121.39 ± 30% +212.8% 379.76 ± 23% sched_debug.cfs_rq:/.util_est.max 23.68 ± 20% +318.5% 99.12 ± 21% sched_debug.cfs_rq:/.util_est.stddev 66860 ± 13% -37.7% 41635 ± 27% sched_debug.cpu.avg_idle.stddev 2.40 ± 21% -70.1% 0.72 ± 21% sched_debug.cpu.clock.stddev 451.30 ± 3% -26.7% 330.98 ± 28% sched_debug.cpu.clock_task.stddev 90057 ± 3% -24.8% 67696 ± 23% sched_debug.cpu.curr->pid.max 16561 ± 9% -23.8% 12621 ± 22% sched_debug.cpu.curr->pid.stddev 13987 ± 3% -25.6% 10412 ± 23% sched_debug.cpu.nr_switches.avg 38759 ± 11% -30.0% 27133 ± 25% sched_debug.cpu.nr_switches.max 2680 ± 14% -72.4% 738.87 ± 24% sched_debug.cpu.nr_switches.min 31186 ± 2% -91.9% 2540 proc-vmstat.nr_active_anon 160386 +2.0% 163528 proc-vmstat.nr_anon_pages 797165 -3.7% 767817 proc-vmstat.nr_file_pages 161314 +1.5% 163804 proc-vmstat.nr_inactive_anon 11375 +3.6% 11782 proc-vmstat.nr_kernel_stack 8362 -2.8% 8125 proc-vmstat.nr_mapped 1829 +25.4% 2294 proc-vmstat.nr_page_table_pages 32208 ± 2% -91.1% 2867 ± 2% proc-vmstat.nr_shmem 20413 -3.4% 19710 proc-vmstat.nr_slab_reclaimable 28051 +0.8% 28289 proc-vmstat.nr_slab_unreclaimable 31186 ± 2% -91.9% 2540 proc-vmstat.nr_zone_active_anon 161314 +1.5% 163804 proc-vmstat.nr_zone_inactive_anon 20425 ± 33% -83.5% 3376 ±182% proc-vmstat.numa_hint_faults 26596010 -13.6% 22977970 proc-vmstat.numa_hit 26542993 -13.6% 22936126 proc-vmstat.numa_local 49790 -18.7% 40495 proc-vmstat.numa_other 11951 ± 68% -99.0% 123.17 ±160% proc-vmstat.numa_pages_migrated 91437 ± 43% -79.1% 19121 ±115% proc-vmstat.numa_pte_updates 35248 ± 4% -99.0% 364.17 ± 4% proc-vmstat.pgactivate 27497487 -13.5% 23788404 proc-vmstat.pgalloc_normal 32590981 +2227.8% 7.587e+08 proc-vmstat.pgfault 27449503 -13.7% 23694730 proc-vmstat.pgfree 11951 ± 68% -99.0% 123.17 ±160% proc-vmstat.pgmigrate_success 1549666 ± 2% -12.4% 1356755 ± 2% proc-vmstat.pgreuse 2.17 -61.6% 0.83 ± 2% perf-stat.i.MPKI 5.641e+08 +225.5% 1.836e+09 perf-stat.i.branch-instructions 5.50 -3.5 1.98 perf-stat.i.branch-miss-rate% 31671027 +15.1% 36440585 perf-stat.i.branch-misses 8.81 +1.4 10.23 ± 2% perf-stat.i.cache-miss-rate% 5351586 +35.8% 7269406 ± 2% perf-stat.i.cache-misses 60638638 +16.8% 70801331 perf-stat.i.cache-references 8748 -13.3% 7589 perf-stat.i.context-switches 1.85 -13.4% 1.60 perf-stat.i.cpi 4.765e+09 +193.1% 1.397e+10 perf-stat.i.cpu-cycles 890.01 +117.1% 1931 perf-stat.i.cycles-between-cache-misses 2.791e+09 +212.5% 8.722e+09 perf-stat.i.instructions 0.56 +12.6% 0.63 perf-stat.i.ipc 0.02 ± 27% +90.8% 0.03 ± 46% perf-stat.i.major-faults 4.40 +2276.8% 104.67 perf-stat.i.metric.K/sec 105697 +2276.8% 2512206 perf-stat.i.minor-faults 105697 +2276.8% 2512206 perf-stat.i.page-faults 1.92 -56.5% 0.83 ± 2% perf-stat.overall.MPKI 5.61 -3.6 1.98 perf-stat.overall.branch-miss-rate% 8.82 +1.4 10.27 ± 2% perf-stat.overall.cache-miss-rate% 1.71 -6.2% 1.60 perf-stat.overall.cpi 890.60 +115.8% 1922 perf-stat.overall.cycles-between-cache-misses 0.59 +6.6% 0.62 perf-stat.overall.ipc 5.623e+08 +225.4% 1.83e+09 perf-stat.ps.branch-instructions 31568362 +15.0% 36300756 perf-stat.ps.branch-misses 5333517 +35.9% 7245937 ± 2% perf-stat.ps.cache-misses 60439198 +16.8% 70570262 perf-stat.ps.cache-references 8719 -13.3% 7563 perf-stat.ps.context-switches 4.749e+09 +193.2% 1.392e+10 perf-stat.ps.cpu-cycles 2.782e+09 +212.5% 8.692e+09 perf-stat.ps.instructions 0.02 ± 27% +90.9% 0.03 ± 46% perf-stat.ps.major-faults 105344 +2277.4% 2504483 perf-stat.ps.minor-faults 105344 +2277.4% 2504483 perf-stat.ps.page-faults 8.38e+11 +212.6% 2.62e+12 perf-stat.total.instructions 66.96 -67.0 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 65.60 -65.6 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 65.60 -65.6 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 65.54 -65.5 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 64.56 -64.6 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 62.17 -62.2 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 61.49 -61.5 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 49.76 -49.8 0.00 perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 11.21 ± 2% -11.2 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 10.60 ± 2% -10.6 0.00 perf-profile.calltrace.cycles-pp.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 8.26 ± 2% -8.3 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 8.24 ± 2% -8.2 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 7.96 ± 2% -8.0 0.00 perf-profile.calltrace.cycles-pp.do_execveat_common.__x64_sys_execve.do_syscall_64.entry_SYSCALL_64_after_hwframe.execve 7.18 ± 2% -7.2 0.00 perf-profile.calltrace.cycles-pp.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter 7.01 ± 2% -7.0 0.00 perf-profile.calltrace.cycles-pp.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state 6.33 -6.3 0.00 perf-profile.calltrace.cycles-pp.bprm_execve.do_execveat_common.__x64_sys_execve.do_syscall_64.entry_SYSCALL_64_after_hwframe 6.02 -6.0 0.00 perf-profile.calltrace.cycles-pp.exec_binprm.bprm_execve.do_execveat_common.__x64_sys_execve.do_syscall_64 6.00 -6.0 0.00 perf-profile.calltrace.cycles-pp.search_binary_handler.exec_binprm.bprm_execve.do_execveat_common.__x64_sys_execve 5.75 -5.7 0.00 perf-profile.calltrace.cycles-pp.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt 5.35 -5.3 0.00 perf-profile.calltrace.cycles-pp.execve 5.34 -5.3 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.execve 5.34 -5.3 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_execve.do_syscall_64.entry_SYSCALL_64_after_hwframe.execve 5.34 -5.3 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.execve 5.28 -5.3 0.00 perf-profile.calltrace.cycles-pp.load_elf_binary.search_binary_handler.exec_binprm.bprm_execve.do_execveat_common 66.96 -67.0 0.00 perf-profile.children.cycles-pp.common_startup_64 66.96 -67.0 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 66.96 -67.0 0.00 perf-profile.children.cycles-pp.do_idle 65.92 -65.9 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 65.60 -65.6 0.00 perf-profile.children.cycles-pp.start_secondary 62.74 -62.7 0.00 perf-profile.children.cycles-pp.cpuidle_enter 62.59 -62.6 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 49.76 -49.8 0.00 perf-profile.children.cycles-pp.intel_idle 21.70 -21.7 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 21.66 -21.7 0.00 perf-profile.children.cycles-pp.do_syscall_64 11.84 ± 2% -11.8 0.00 perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt 11.19 ± 2% -11.2 0.00 perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt 8.12 ± 2% -8.1 0.00 perf-profile.children.cycles-pp.__x64_sys_execve 8.07 ± 2% -8.1 0.00 perf-profile.children.cycles-pp.execve 8.02 ± 2% -8.0 0.00 perf-profile.children.cycles-pp.do_execveat_common 7.57 ± 2% -7.6 0.00 perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt 7.40 ± 2% -7.4 0.00 perf-profile.children.cycles-pp.hrtimer_interrupt 6.83 -6.8 0.00 perf-profile.children.cycles-pp.asm_exc_page_fault 6.43 -6.4 0.00 perf-profile.children.cycles-pp.bprm_execve 6.39 -6.4 0.00 perf-profile.children.cycles-pp.exc_page_fault 6.27 -6.3 0.00 perf-profile.children.cycles-pp.do_user_addr_fault 6.12 -6.1 0.00 perf-profile.children.cycles-pp.__hrtimer_run_queues 6.04 -6.0 0.00 perf-profile.children.cycles-pp.exec_binprm 6.02 -6.0 0.00 perf-profile.children.cycles-pp.search_binary_handler 5.64 -5.6 0.00 perf-profile.children.cycles-pp.handle_mm_fault 5.57 -5.6 0.00 perf-profile.children.cycles-pp.load_elf_binary 5.25 -5.2 0.00 perf-profile.children.cycles-pp.__handle_mm_fault 49.75 -49.8 0.00 perf-profile.self.cycles-pp.intel_idle *************************************************************************************************** lkp-spr-r02: 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 256G memory ========================================================================================= compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/lkp-spr-r02/sem/stress-ng/60s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 257775 ± 22% +124.5% 578661 ± 21% cpuidle..usage 2202062 ± 22% -90.8% 202614 ± 38% numa-numastat.node1.local_node 2296320 ± 19% -85.0% 345152 ± 19% numa-numastat.node1.numa_hit 19.92 -52.0% 9.55 ± 2% vmstat.cpu.us 1.856e+08 ± 3% -58.4% 77159822 ± 4% vmstat.system.cs 556587 +42.9% 795511 vmstat.system.in 0.58 ± 2% +0.2 0.73 mpstat.cpu.all.irq% 0.02 ± 3% -0.0 0.01 ± 6% mpstat.cpu.all.soft% 75.65 +10.9 86.52 mpstat.cpu.all.sys% 20.45 -10.8 9.70 ± 3% mpstat.cpu.all.usr% 1531 ± 54% -100.0% 0.00 perf-c2c.DRAM.local 2246 ± 32% -100.0% 0.00 perf-c2c.DRAM.remote 1173 ± 69% -100.0% 0.00 perf-c2c.HITM.local 549.17 ± 31% -100.0% 0.00 perf-c2c.HITM.remote 25808257 ± 2% -58.1% 10805602 ± 4% sched_debug.cpu.nr_switches.avg 30207217 -57.0% 12980041 sched_debug.cpu.nr_switches.max 3103000 ± 72% -88.6% 354272 ±153% sched_debug.cpu.nr_switches.min 4087097 ± 21% -40.3% 2439826 ± 18% sched_debug.cpu.nr_switches.stddev 1.429e+10 ± 3% -53.6% 6.627e+09 ± 5% stress-ng.sem.ops 2.382e+08 ± 3% -53.6% 1.104e+08 ± 5% stress-ng.sem.ops_per_sec 327589 -56.8% 141462 stress-ng.sem.sem_timedwait_calls_per_sec 327403 -56.8% 141294 stress-ng.sem.sem_trywait_calls_per_sec 327558 -56.8% 141438 stress-ng.sem.sem_wait_calls_per_sec 1.178e+10 ± 2% -58.5% 4.886e+09 ± 4% stress-ng.time.involuntary_context_switches 10499 +13.5% 11913 stress-ng.time.system_time 2767 -53.0% 1301 ± 2% stress-ng.time.user_time 473599 ± 19% -63.1% 174673 ± 55% numa-meminfo.node0.Mapped 422997 ± 14% -99.3% 2940 ± 68% numa-meminfo.node1.Active 422988 ± 14% -99.3% 2932 ± 69% numa-meminfo.node1.Active(anon) 4625028 ± 25% -96.2% 176220 ± 42% numa-meminfo.node1.FilePages 4814124 ± 26% -84.9% 728255 ± 20% numa-meminfo.node1.Inactive 4814082 ± 26% -84.9% 728212 ± 20% numa-meminfo.node1.Inactive(anon) 1000373 ± 11% -92.9% 71080 ±136% numa-meminfo.node1.Mapped 6804306 ± 19% -67.0% 2242499 ± 5% numa-meminfo.node1.MemUsed 21291 ± 41% -41.3% 12504 ± 70% numa-meminfo.node1.PageTables 4556992 ± 26% -98.4% 74595 ±132% numa-meminfo.node1.Shmem 118547 ± 19% -63.1% 43793 ± 54% numa-vmstat.node0.nr_mapped 105773 ± 13% -99.3% 733.51 ± 69% numa-vmstat.node1.nr_active_anon 1156277 ± 25% -96.2% 44056 ± 42% numa-vmstat.node1.nr_file_pages 1203588 ± 26% -84.9% 182078 ± 20% numa-vmstat.node1.nr_inactive_anon 249177 ± 11% -92.8% 17833 ±135% numa-vmstat.node1.nr_mapped 5333 ± 41% -41.6% 3116 ± 70% numa-vmstat.node1.nr_page_table_pages 1139269 ± 26% -98.4% 18649 ±132% numa-vmstat.node1.nr_shmem 105772 ± 13% -99.3% 733.51 ± 69% numa-vmstat.node1.nr_zone_active_anon 1203585 ± 26% -84.9% 182077 ± 20% numa-vmstat.node1.nr_zone_inactive_anon 2295346 ± 19% -85.0% 344087 ± 19% numa-vmstat.node1.numa_hit 2201088 ± 22% -90.8% 201553 ± 38% numa-vmstat.node1.numa_local 426533 ± 13% -98.5% 6257 ± 20% meminfo.Active 426484 ± 13% -98.5% 6209 ± 21% meminfo.Active(anon) 7723073 ± 16% -57.1% 3309547 meminfo.Cached 10911723 ± 11% -41.3% 6400110 meminfo.Committed_AS 5441802 ± 22% -75.4% 1340267 meminfo.Inactive 5441575 ± 22% -75.4% 1340041 meminfo.Inactive(anon) 152735 -10.7% 136341 meminfo.KReclaimable 1472940 ± 7% -83.4% 245093 meminfo.Mapped 12189055 ± 10% -37.8% 7581637 meminfo.Memused 152735 -10.7% 136341 meminfo.SReclaimable 4639245 ± 27% -95.1% 225811 meminfo.Shmem 12381239 ± 9% -32.7% 8329522 meminfo.max_used_kB 0.05 ± 14% -52.4% 0.02 ± 18% perf-stat.i.MPKI 1.34e+11 +20.6% 1.616e+11 perf-stat.i.branch-instructions 1.17 ± 2% -0.7 0.49 ± 4% perf-stat.i.branch-miss-rate% 1.535e+09 ± 3% -53.0% 7.219e+08 ± 2% perf-stat.i.branch-misses 37.03 ± 7% -14.9 22.14 ± 15% perf-stat.i.cache-miss-rate% 21811787 ± 11% -63.8% 7902212 ± 21% perf-stat.i.cache-misses 57650952 ± 7% -39.5% 34876714 ± 11% perf-stat.i.cache-references 1.944e+08 ± 3% -58.3% 81049117 ± 4% perf-stat.i.context-switches 1.04 -17.7% 0.85 perf-stat.i.cpi 366.44 ± 3% -14.1% 314.89 ± 3% perf-stat.i.cpu-migrations 39918 ± 13% +167.9% 106957 ± 21% perf-stat.i.cycles-between-cache-misses 6.228e+11 +22.2% 7.609e+11 perf-stat.i.instructions 0.97 +22.5% 1.19 perf-stat.i.ipc 848.31 ± 2% -56.7% 367.24 ± 5% perf-stat.i.metric.K/sec 17694 ± 4% +7343.5% 1317088 ± 25% perf-stat.i.minor-faults 17694 ± 4% +7343.5% 1317089 ± 25% perf-stat.i.page-faults 24418 ± 46% +244.4% 84106 ± 20% perf-stat.overall.cycles-between-cache-misses 0.81 ± 44% +47.4% 1.19 perf-stat.overall.ipc 1.091e+11 ± 44% +44.9% 1.581e+11 perf-stat.ps.branch-instructions 5.071e+11 ± 44% +46.8% 7.443e+11 perf-stat.ps.instructions 14428 ± 44% +8787.7% 1282380 ± 24% perf-stat.ps.minor-faults 14428 ± 44% +8787.7% 1282380 ± 24% perf-stat.ps.page-faults 3.113e+13 ± 44% +46.7% 4.566e+13 perf-stat.total.instructions 107212 ± 13% -98.6% 1552 ± 21% proc-vmstat.nr_active_anon 309617 ± 2% -8.7% 282833 proc-vmstat.nr_anon_pages 6248796 +1.8% 6363802 proc-vmstat.nr_dirty_background_threshold 12512871 +1.8% 12743164 proc-vmstat.nr_dirty_threshold 1931563 ± 16% -57.2% 827520 proc-vmstat.nr_file_pages 62864244 +1.8% 64015988 proc-vmstat.nr_free_pages 1360577 ± 22% -75.3% 335437 proc-vmstat.nr_inactive_anon 367857 ± 7% -83.3% 61606 proc-vmstat.nr_mapped 8911 -10.3% 7992 ± 5% proc-vmstat.nr_page_table_pages 1160604 ± 27% -95.1% 56584 proc-vmstat.nr_shmem 38183 -10.7% 34085 proc-vmstat.nr_slab_reclaimable 107212 ± 13% -98.6% 1552 ± 21% proc-vmstat.nr_zone_active_anon 1360577 ± 22% -75.3% 335437 proc-vmstat.nr_zone_inactive_anon 250896 ± 6% -91.1% 22308 ± 29% proc-vmstat.numa_hint_faults 123468 ± 9% -93.7% 7808 ± 3% proc-vmstat.numa_hint_faults_local 2834379 ± 16% -72.1% 789773 proc-vmstat.numa_hit 2602586 ± 18% -76.9% 601311 proc-vmstat.numa_local 231785 -18.7% 188526 proc-vmstat.numa_other 74727 ± 8% -33.0% 50062 ± 21% proc-vmstat.numa_pages_migrated 923809 ± 3% -53.9% 425642 ± 10% proc-vmstat.numa_pte_updates 104416 ± 18% -48.0% 54266 proc-vmstat.pgactivate 2989481 ± 14% -66.7% 996881 proc-vmstat.pgalloc_normal 1207566 ± 2% +6626.3% 81224725 ± 24% proc-vmstat.pgfault 909425 ± 2% -27.7% 657444 ± 5% proc-vmstat.pgfree 74727 ± 8% -33.0% 50062 ± 21% proc-vmstat.pgmigrate_success 38727 ± 7% -34.0% 25562 proc-vmstat.pgreuse 90.17 -90.2 0.00 perf-profile.calltrace.cycles-pp.__sched_yield 69.08 -69.1 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__sched_yield 68.94 -68.9 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 46.48 -46.5 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_sched_yield.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 36.80 -36.8 0.00 perf-profile.calltrace.cycles-pp.schedule.__x64_sys_sched_yield.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 34.49 -34.5 0.00 perf-profile.calltrace.cycles-pp.__schedule.schedule.__x64_sys_sched_yield.do_syscall_64.entry_SYSCALL_64_after_hwframe 18.83 ± 2% -18.8 0.00 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 17.91 -17.9 0.00 perf-profile.calltrace.cycles-pp.pick_next_task_fair.__schedule.schedule.__x64_sys_sched_yield.do_syscall_64 8.68 -8.7 0.00 perf-profile.calltrace.cycles-pp.do_sched_yield.__x64_sys_sched_yield.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 8.13 ± 2% -8.1 0.00 perf-profile.calltrace.cycles-pp.__rseq_handle_notify_resume.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 7.60 ± 2% -7.6 0.00 perf-profile.calltrace.cycles-pp.switch_fpu_return.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.__sched_yield 7.04 ± 3% -7.0 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.__sched_yield 6.38 ± 2% -6.4 0.00 perf-profile.calltrace.cycles-pp.yield_task_fair.do_sched_yield.__x64_sys_sched_yield.do_syscall_64.entry_SYSCALL_64_after_hwframe 6.24 ± 2% -6.2 0.00 perf-profile.calltrace.cycles-pp.put_prev_entity.pick_next_task_fair.__schedule.schedule.__x64_sys_sched_yield 6.09 ± 2% -6.1 0.00 perf-profile.calltrace.cycles-pp.restore_fpregs_from_fpstate.switch_fpu_return.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe 5.01 ± 2% -5.0 0.00 perf-profile.calltrace.cycles-pp.rseq_ip_fixup.__rseq_handle_notify_resume.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe 91.66 -91.7 0.00 perf-profile.children.cycles-pp.__sched_yield 70.10 -70.1 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 68.79 -68.8 0.00 perf-profile.children.cycles-pp.do_syscall_64 47.16 -47.2 0.00 perf-profile.children.cycles-pp.__x64_sys_sched_yield 36.94 -36.9 0.00 perf-profile.children.cycles-pp.schedule 35.62 -35.6 0.00 perf-profile.children.cycles-pp.__schedule 19.16 ± 2% -19.2 0.00 perf-profile.children.cycles-pp.syscall_exit_to_user_mode 18.43 -18.4 0.00 perf-profile.children.cycles-pp.pick_next_task_fair 8.91 -8.9 0.00 perf-profile.children.cycles-pp.do_sched_yield 8.28 ± 2% -8.3 0.00 perf-profile.children.cycles-pp.__rseq_handle_notify_resume 7.72 ± 2% -7.7 0.00 perf-profile.children.cycles-pp.switch_fpu_return 7.15 ± 2% -7.1 0.00 perf-profile.children.cycles-pp.update_curr 6.61 ± 2% -6.6 0.00 perf-profile.children.cycles-pp.yield_task_fair 6.57 ± 2% -6.6 0.00 perf-profile.children.cycles-pp.put_prev_entity 6.32 ± 2% -6.3 0.00 perf-profile.children.cycles-pp.update_load_avg 6.14 ± 2% -6.1 0.00 perf-profile.children.cycles-pp.restore_fpregs_from_fpstate 5.17 ± 2% -5.2 0.00 perf-profile.children.cycles-pp.rseq_ip_fixup 5.07 ± 2% -5.1 0.00 perf-profile.children.cycles-pp.prepare_task_switch 6.07 ± 2% -6.1 0.00 perf-profile.self.cycles-pp.restore_fpregs_from_fpstate 5.19 ± 2% -5.2 0.00 perf-profile.self.cycles-pp.__schedule *************************************************************************************************** lkp-icl-2sp9: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory ========================================================================================= cluster/compiler/cpufreq_governor/iterations/kconfig/nr_threads/rootfs/runtime/tbox_group/testcase: cs-localhost/gcc-12/performance/1/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/600/lkp-icl-2sp9/tbench commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 7848524 -35.6% 5055935 vmstat.system.cs 99625 ± 16% +320.7% 419100 ± 2% vmstat.system.in 0.35 ± 9% +0.2 0.51 ± 7% mpstat.cpu.all.irq% 18.73 +13.5 32.27 mpstat.cpu.all.soft% 68.31 -8.5 59.81 mpstat.cpu.all.sys% 12.13 -5.0 7.17 mpstat.cpu.all.usr% 1.335e+08 -35.0% 86802870 numa-numastat.node0.local_node 1.335e+08 -35.0% 86830895 numa-numastat.node0.numa_hit 1.35e+08 -36.3% 85986323 numa-numastat.node1.local_node 1.35e+08 -36.3% 86011207 numa-numastat.node1.numa_hit 240.00 ± 11% -100.0% 0.00 perf-c2c.DRAM.local 7529 ± 6% -100.0% 0.00 perf-c2c.DRAM.remote 7785 ± 6% -100.0% 0.00 perf-c2c.HITM.local 4494 ± 4% -100.0% 0.00 perf-c2c.HITM.remote 12280 ± 3% -100.0% 0.00 perf-c2c.HITM.total 20996 -42.5% 12072 tbench.throughput_MB/s 20996 -42.5% 12072 tbench.throughput_MB/s_harmonic_mean 1.499e+09 -36.6% 9.505e+08 tbench.time.involuntary_context_switches 3512 -19.7% 2820 tbench.time.percent_of_cpu_this_job_got 20247 -14.4% 17328 tbench.time.system_time 5054 -40.8% 2991 tbench.time.user_time 1.328e+09 -34.3% 8.725e+08 tbench.time.voluntary_context_switches 80062 ± 51% -96.2% 3047 ± 70% numa-meminfo.node0.Active 80041 ± 51% -96.3% 2993 ± 71% numa-meminfo.node0.Active(anon) 94532 ± 44% -84.6% 14602 ± 24% numa-meminfo.node0.Shmem 3416559 ± 24% -99.9% 3490 ± 56% numa-meminfo.node1.Active 3416370 ± 24% -99.9% 3343 ± 55% numa-meminfo.node1.Active(anon) 179580 ± 58% +185.9% 513453 ± 32% numa-meminfo.node1.AnonPages.max 5327384 ± 30% -76.3% 1264854 ±100% numa-meminfo.node1.FilePages 106230 ± 45% -86.1% 14817 ±110% numa-meminfo.node1.Mapped 5947008 ± 28% -60.9% 2327541 ± 50% numa-meminfo.node1.MemUsed 5973 ± 84% +73.8% 10382 ± 47% numa-meminfo.node1.PageTables 3586379 ± 25% -99.8% 7080 ± 50% numa-meminfo.node1.Shmem 3495980 ± 23% -99.8% 6538 ± 18% meminfo.Active 3495769 ± 23% -99.8% 6337 ± 18% meminfo.Active(anon) 6793049 ± 13% -53.9% 3133909 meminfo.Cached 5711472 ± 15% -63.6% 2077036 meminfo.Committed_AS 860774 ± 10% -19.1% 696070 meminfo.Inactive 860058 ± 10% -19.2% 695354 meminfo.Inactive(anon) 100084 ± 2% -11.6% 88500 meminfo.KReclaimable 140027 ± 29% -69.4% 42890 meminfo.Mapped 8473548 ± 10% -40.4% 5048986 meminfo.Memused 100084 ± 2% -11.6% 88500 meminfo.SReclaimable 3680778 ± 24% -99.4% 21686 meminfo.Shmem 8658135 ± 10% -39.6% 5229555 meminfo.max_used_kB 20021 ± 51% -96.3% 748.82 ± 71% numa-vmstat.node0.nr_active_anon 23635 ± 44% -84.6% 3651 ± 24% numa-vmstat.node0.nr_shmem 20021 ± 51% -96.3% 748.82 ± 71% numa-vmstat.node0.nr_zone_active_anon 1.335e+08 -35.0% 86830064 numa-vmstat.node0.numa_hit 1.335e+08 -35.0% 86802040 numa-vmstat.node0.numa_local 854348 ± 24% -99.9% 835.66 ± 55% numa-vmstat.node1.nr_active_anon 1331772 ± 30% -76.3% 316213 ±100% numa-vmstat.node1.nr_file_pages 26385 ± 45% -85.6% 3786 ±110% numa-vmstat.node1.nr_mapped 1493 ± 84% +73.7% 2595 ± 47% numa-vmstat.node1.nr_page_table_pages 896521 ± 25% -99.8% 1770 ± 50% numa-vmstat.node1.nr_shmem 854348 ± 24% -99.9% 835.66 ± 55% numa-vmstat.node1.nr_zone_active_anon 1.35e+08 -36.3% 86010078 numa-vmstat.node1.numa_hit 1.35e+08 -36.3% 85985195 numa-vmstat.node1.numa_local 874132 ± 23% -99.8% 1584 ± 18% proc-vmstat.nr_active_anon 6344454 +1.3% 6429945 proc-vmstat.nr_dirty_background_threshold 12704422 +1.3% 12875612 proc-vmstat.nr_dirty_threshold 1698254 ± 13% -53.9% 783477 proc-vmstat.nr_file_pages 63822128 +1.3% 64678294 proc-vmstat.nr_free_pages 214803 ± 10% -19.1% 173832 proc-vmstat.nr_inactive_anon 35091 ± 29% -68.7% 10970 proc-vmstat.nr_mapped 920185 ± 24% -99.4% 5420 proc-vmstat.nr_shmem 25021 ± 2% -11.6% 22125 proc-vmstat.nr_slab_reclaimable 874132 ± 23% -99.8% 1584 ± 18% proc-vmstat.nr_zone_active_anon 214803 ± 10% -19.1% 173832 proc-vmstat.nr_zone_inactive_anon 119453 ± 26% -95.7% 5091 ± 59% proc-vmstat.numa_hint_faults 65444 ± 36% -98.4% 1036 ± 25% proc-vmstat.numa_hint_faults_local 2.685e+08 -35.6% 1.728e+08 proc-vmstat.numa_hit 2.685e+08 -35.6% 1.728e+08 proc-vmstat.numa_local 66380 -18.6% 54024 proc-vmstat.numa_other 252531 ± 14% -60.8% 99021 ± 17% proc-vmstat.numa_pte_updates 1163823 ± 24% -99.9% 776.50 ± 23% proc-vmstat.pgactivate 2.129e+09 -35.5% 1.373e+09 proc-vmstat.pgalloc_normal 2063471 ± 2% +45810.5% 9.474e+08 ± 9% proc-vmstat.pgfault 2.128e+09 -35.5% 1.373e+09 proc-vmstat.pgfree 85685 ± 9% -11.0% 76256 proc-vmstat.pgreuse 17143370 -10.1% 15417411 sched_debug.cfs_rq:/.avg_vruntime.avg 16788080 -10.5% 15023795 sched_debug.cfs_rq:/.avg_vruntime.min 147023 ± 26% +96.3% 288623 ± 26% sched_debug.cfs_rq:/.avg_vruntime.stddev 8563 ± 18% -47.1% 4529 ± 42% sched_debug.cfs_rq:/.load.min 17143370 -10.1% 15417411 sched_debug.cfs_rq:/.min_vruntime.avg 16788080 -10.5% 15023795 sched_debug.cfs_rq:/.min_vruntime.min 147023 ± 26% +96.3% 288623 ± 26% sched_debug.cfs_rq:/.min_vruntime.stddev 1805 ± 3% +15.6% 2086 sched_debug.cfs_rq:/.runnable_avg.max 1079 ± 7% -15.8% 908.85 ± 6% sched_debug.cfs_rq:/.runnable_avg.min 98.24 ± 12% +92.3% 188.89 ± 3% sched_debug.cfs_rq:/.runnable_avg.stddev 51.17 ± 16% +35.0% 69.10 ± 9% sched_debug.cfs_rq:/.util_avg.stddev 1176 ± 8% +41.6% 1665 ± 4% sched_debug.cfs_rq:/.util_est.max 241.66 ± 3% +16.9% 282.59 ± 2% sched_debug.cfs_rq:/.util_est.stddev 620439 ± 19% +40.4% 871331 ± 4% sched_debug.cpu.avg_idle.max 6766 ± 3% +94.0% 13129 ± 4% sched_debug.cpu.avg_idle.min 8.69 ± 4% +31.4% 11.41 ± 6% sched_debug.cpu.clock.stddev 327685 -12.4% 287198 sched_debug.cpu.clock_task.avg 331298 -10.1% 297931 sched_debug.cpu.clock_task.max 319890 -12.6% 279574 sched_debug.cpu.clock_task.min 1712 ± 10% +72.8% 2960 ± 11% sched_debug.cpu.clock_task.stddev 2339 ± 10% -48.6% 1202 ± 74% sched_debug.cpu.max_idle_balance_cost.stddev 0.00 ± 6% +27.9% 0.00 ± 20% sched_debug.cpu.next_balance.stddev 44202960 -28.6% 31561513 sched_debug.cpu.nr_switches.avg 45127928 -27.9% 32529217 sched_debug.cpu.nr_switches.max 42497993 -35.6% 27381916 ± 4% sched_debug.cpu.nr_switches.min 489510 ± 7% +87.5% 918047 ± 17% sched_debug.cpu.nr_switches.stddev 0.20 ± 12% -42.7% 0.12 ± 7% perf-stat.i.MPKI 3.212e+10 +38.9% 4.463e+10 perf-stat.i.branch-instructions 1.01 -0.4 0.64 perf-stat.i.branch-miss-rate% 3.214e+08 -15.3% 2.721e+08 perf-stat.i.branch-misses 32.93 ± 4% -2.2 30.71 ± 2% perf-stat.i.cache-miss-rate% 32055029 ± 12% -26.8% 23468576 ± 7% perf-stat.i.cache-misses 1.042e+08 ± 13% -24.7% 78443526 ± 5% perf-stat.i.cache-references 7869989 -35.6% 5069963 perf-stat.i.context-switches 1.40 -24.6% 1.06 perf-stat.i.cpi 7138 ± 12% +40.2% 10010 ± 7% perf-stat.i.cycles-between-cache-misses 1.601e+11 +34.5% 2.153e+11 perf-stat.i.instructions 0.71 +34.8% 0.96 perf-stat.i.ipc 2740 ± 2% +47765.7% 1311791 ± 9% perf-stat.i.minor-faults 2740 ± 2% +47765.5% 1311791 ± 9% perf-stat.i.page-faults 0.20 ± 12% -45.6% 0.11 ± 7% perf-stat.overall.MPKI 1.00 -0.4 0.61 perf-stat.overall.branch-miss-rate% 1.40 -25.9% 1.04 perf-stat.overall.cpi 7104 ± 12% +35.0% 9592 ± 8% perf-stat.overall.cycles-between-cache-misses 0.71 +34.9% 0.96 perf-stat.overall.ipc 3.207e+10 +38.9% 4.455e+10 perf-stat.ps.branch-instructions 3.209e+08 -15.3% 2.717e+08 perf-stat.ps.branch-misses 32011532 ± 12% -26.8% 23433423 ± 7% perf-stat.ps.cache-misses 1.042e+08 ± 13% -24.8% 78378749 ± 5% perf-stat.ps.cache-references 7857872 -35.6% 5062068 perf-stat.ps.context-switches 1.598e+11 +34.5% 2.149e+11 perf-stat.ps.instructions 2738 ± 2% +47722.7% 1309682 ± 9% perf-stat.ps.minor-faults 2738 ± 2% +47722.5% 1309682 ± 9% perf-stat.ps.page-faults 1.153e+14 +34.5% 1.551e+14 perf-stat.total.instructions 96.07 -96.1 0.00 perf-profile.calltrace.cycles-pp.main 60.65 -60.7 0.00 perf-profile.calltrace.cycles-pp.create_procs.main 42.12 -42.1 0.00 perf-profile.calltrace.cycles-pp.child_run.create_procs.main 40.64 -40.6 0.00 perf-profile.calltrace.cycles-pp.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send 39.19 -39.2 0.00 perf-profile.calltrace.cycles-pp.tcp_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe 38.18 -38.2 0.00 perf-profile.calltrace.cycles-pp.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64 37.37 -37.4 0.00 perf-profile.calltrace.cycles-pp.child_op.child_run.create_procs.main 35.42 -35.4 0.00 perf-profile.calltrace.cycles-pp.listener.main 35.36 -35.4 0.00 perf-profile.calltrace.cycles-pp.server.listener.main 30.48 -30.5 0.00 perf-profile.calltrace.cycles-pp.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto.__x64_sys_sendto 30.20 -30.2 0.00 perf-profile.calltrace.cycles-pp.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg.__sys_sendto 27.92 -27.9 0.00 perf-profile.calltrace.cycles-pp.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg 24.93 -24.9 0.00 perf-profile.calltrace.cycles-pp.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked 23.36 -23.4 0.00 perf-profile.calltrace.cycles-pp.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames 22.77 -22.8 0.00 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb.tcp_write_xmit 21.45 -21.4 0.00 perf-profile.calltrace.cycles-pp.__send.server.listener.main 21.42 -21.4 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send.do_packets 20.65 -20.7 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__send.server.listener.main 20.61 -20.6 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send.server.listener 19.40 -19.4 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send.server 18.50 -18.5 0.00 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.__tcp_transmit_skb 18.31 ± 2% -18.3 0.00 perf-profile.calltrace.cycles-pp.kill.create_procs.main 18.30 -18.3 0.00 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit 18.07 -18.1 0.00 perf-profile.calltrace.cycles-pp.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2 17.70 ± 2% -17.7 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.kill.create_procs.main 17.67 ± 2% -17.7 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.kill.create_procs.main 17.14 ± 2% -17.1 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_kill.do_syscall_64.entry_SYSCALL_64_after_hwframe.kill.create_procs 17.05 -17.0 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit 16.77 ± 3% -16.8 0.00 perf-profile.calltrace.cycles-pp.kill_something_info.__x64_sys_kill.do_syscall_64.entry_SYSCALL_64_after_hwframe.kill 16.50 ± 3% -16.5 0.00 perf-profile.calltrace.cycles-pp.kill_pid_info_type.kill_something_info.__x64_sys_kill.do_syscall_64.entry_SYSCALL_64_after_hwframe 16.42 -16.4 0.00 perf-profile.calltrace.cycles-pp.__napi_poll.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip 16.31 -16.3 0.00 perf-profile.calltrace.cycles-pp.process_backlog.__napi_poll.net_rx_action.handle_softirqs.do_softirq 15.90 -15.9 0.00 perf-profile.calltrace.cycles-pp.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recv 15.76 -15.8 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action.handle_softirqs 13.94 -13.9 0.00 perf-profile.calltrace.cycles-pp.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe 13.43 -13.4 0.00 perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action 13.38 -13.4 0.00 perf-profile.calltrace.cycles-pp.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll 13.28 -13.3 0.00 perf-profile.calltrace.cycles-pp.inet_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64 13.15 -13.2 0.00 perf-profile.calltrace.cycles-pp.recv.server.listener.main 13.15 -13.1 0.00 perf-profile.calltrace.cycles-pp.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog 12.97 -13.0 0.00 perf-profile.calltrace.cycles-pp.tcp_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom 12.60 ± 3% -12.6 0.00 perf-profile.calltrace.cycles-pp.security_task_kill.kill_pid_info_type.kill_something_info.__x64_sys_kill.do_syscall_64 12.57 ± 3% -12.6 0.00 perf-profile.calltrace.cycles-pp.apparmor_task_kill.security_task_kill.kill_pid_info_type.kill_something_info.__x64_sys_kill 11.84 -11.8 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.recv.server.listener.main 11.76 -11.8 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.recv.server.listener 11.29 -11.3 0.00 perf-profile.calltrace.cycles-pp.tcp_v4_do_rcv.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core 11.21 -11.2 0.00 perf-profile.calltrace.cycles-pp.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg.sock_recvmsg.__sys_recvfrom 10.98 -11.0 0.00 perf-profile.calltrace.cycles-pp.tcp_rcv_established.tcp_v4_do_rcv.tcp_v4_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish 10.56 -10.6 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recv.server 9.59 -9.6 0.00 perf-profile.calltrace.cycles-pp.sio_readx.child_op.child_run.create_procs.main 8.73 -8.7 0.00 perf-profile.calltrace.cycles-pp.do_packets.sio_readx.child_op.child_run.create_procs 6.57 -6.6 0.00 perf-profile.calltrace.cycles-pp.__send.do_packets.sio_readx.child_op.child_run 6.36 -6.4 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__send.do_packets.sio_readx.child_op 6.35 -6.4 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__send.do_packets.sio_readx 6.06 -6.1 0.00 perf-profile.calltrace.cycles-pp.sio_createx.child_op.child_run.create_procs.main 5.90 -5.9 0.00 perf-profile.calltrace.cycles-pp.sk_wait_data.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg.sock_recvmsg 5.57 -5.6 0.00 perf-profile.calltrace.cycles-pp.do_packets.sio_createx.child_op.child_run.create_procs 5.49 -5.5 0.00 perf-profile.calltrace.cycles-pp.sio_qpathinfo.child_op.child_run.create_procs.main 5.09 -5.1 0.00 perf-profile.calltrace.cycles-pp.wait_woken.sk_wait_data.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg 5.04 -5.0 0.00 perf-profile.calltrace.cycles-pp.do_packets.sio_qpathinfo.child_op.child_run.create_procs 96.13 -96.1 0.00 perf-profile.children.cycles-pp.main 85.56 -85.6 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 84.98 -85.0 0.00 perf-profile.children.cycles-pp.do_syscall_64 60.65 -60.7 0.00 perf-profile.children.cycles-pp.create_procs 48.42 -48.4 0.00 perf-profile.children.cycles-pp.__send 42.26 -42.3 0.00 perf-profile.children.cycles-pp.child_run 41.59 -41.6 0.00 perf-profile.children.cycles-pp.__x64_sys_sendto 41.40 -41.4 0.00 perf-profile.children.cycles-pp.__sys_sendto 39.97 -40.0 0.00 perf-profile.children.cycles-pp.tcp_sendmsg 38.93 -38.9 0.00 perf-profile.children.cycles-pp.tcp_sendmsg_locked 37.51 -37.5 0.00 perf-profile.children.cycles-pp.child_op 35.42 -35.4 0.00 perf-profile.children.cycles-pp.listener 35.38 -35.4 0.00 perf-profile.children.cycles-pp.server 33.33 -33.3 0.00 perf-profile.children.cycles-pp.do_packets 32.27 -32.3 0.00 perf-profile.children.cycles-pp.tcp_write_xmit 32.00 -32.0 0.00 perf-profile.children.cycles-pp.__tcp_push_pending_frames 30.08 -30.1 0.00 perf-profile.children.cycles-pp.__tcp_transmit_skb 27.32 -27.3 0.00 perf-profile.children.cycles-pp.__ip_queue_xmit 25.59 -25.6 0.00 perf-profile.children.cycles-pp.ip_finish_output2 25.22 -25.2 0.00 perf-profile.children.cycles-pp.recv 24.96 -25.0 0.00 perf-profile.children.cycles-pp.__dev_queue_xmit 20.53 -20.5 0.00 perf-profile.children.cycles-pp.__local_bh_enable_ip 20.06 -20.1 0.00 perf-profile.children.cycles-pp.do_softirq 19.82 -19.8 0.00 perf-profile.children.cycles-pp.handle_softirqs 19.19 -19.2 0.00 perf-profile.children.cycles-pp.__x64_sys_recvfrom 18.92 -18.9 0.00 perf-profile.children.cycles-pp.__sys_recvfrom 18.67 -18.7 0.00 perf-profile.children.cycles-pp.net_rx_action 18.44 ± 2% -18.4 0.00 perf-profile.children.cycles-pp.kill 17.99 -18.0 0.00 perf-profile.children.cycles-pp.__napi_poll 17.88 -17.9 0.00 perf-profile.children.cycles-pp.process_backlog 17.27 -17.3 0.00 perf-profile.children.cycles-pp.__netif_receive_skb_one_core 17.16 ± 2% -17.2 0.00 perf-profile.children.cycles-pp.__x64_sys_kill 16.99 -17.0 0.00 perf-profile.children.cycles-pp.sock_recvmsg 16.78 ± 3% -16.8 0.00 perf-profile.children.cycles-pp.kill_something_info 16.51 ± 3% -16.5 0.00 perf-profile.children.cycles-pp.kill_pid_info_type 16.03 -16.0 0.00 perf-profile.children.cycles-pp.inet_recvmsg 15.70 -15.7 0.00 perf-profile.children.cycles-pp.tcp_recvmsg 15.25 -15.2 0.00 perf-profile.children.cycles-pp.ip_local_deliver_finish 15.20 -15.2 0.00 perf-profile.children.cycles-pp.ip_protocol_deliver_rcu 14.97 -15.0 0.00 perf-profile.children.cycles-pp.tcp_v4_rcv 13.87 -13.9 0.00 perf-profile.children.cycles-pp.tcp_recvmsg_locked 12.95 -12.9 0.00 perf-profile.children.cycles-pp.tcp_v4_do_rcv 12.66 -12.7 0.00 perf-profile.children.cycles-pp.tcp_rcv_established 12.61 ± 3% -12.6 0.00 perf-profile.children.cycles-pp.security_task_kill 12.58 ± 3% -12.6 0.00 perf-profile.children.cycles-pp.apparmor_task_kill 9.59 -9.6 0.00 perf-profile.children.cycles-pp.sio_readx 8.65 -8.6 0.00 perf-profile.children.cycles-pp.schedule 8.55 -8.6 0.00 perf-profile.children.cycles-pp.__schedule 7.75 -7.8 0.00 perf-profile.children.cycles-pp.sk_wait_data 6.60 -6.6 0.00 perf-profile.children.cycles-pp.wait_woken 6.23 -6.2 0.00 perf-profile.children.cycles-pp.schedule_timeout 6.07 -6.1 0.00 perf-profile.children.cycles-pp.sio_createx 5.70 -5.7 0.00 perf-profile.children.cycles-pp.syscall_exit_to_user_mode 5.49 -5.5 0.00 perf-profile.children.cycles-pp.sio_qpathinfo 5.08 -5.1 0.00 perf-profile.children.cycles-pp.sock_def_readable 10.63 ± 3% -10.6 0.00 perf-profile.self.cycles-pp.apparmor_task_kill *************************************************************************************************** lkp-icl-2sp9: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory ========================================================================================= cluster/compiler/cpufreq_governor/datasize/kconfig/nr_threads/protocol/rootfs/runtime/tbox_group/testcase/winsize: cs-localhost/gcc-12/performance/64/x86_64-rhel-8.3/1/UDP/debian-12-x86_64-20240206.cgz/30s/lkp-icl-2sp9/uperf/32k commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 6808243 -25.8% 5054865 cpuidle..usage 34.51 -1.7% 33.94 boot-time.boot 2003 -1.8% 1966 boot-time.idle 313841 ± 3% -72.4% 86605 ± 25% numa-numastat.node1.local_node 334568 -67.5% 108757 ± 15% numa-numastat.node1.numa_hit 3.53 ± 3% +55.6% 5.49 vmstat.procs.r 367355 ± 3% -24.3% 277978 vmstat.system.cs 197397 ± 3% -12.2% 173405 vmstat.system.in 0.17 ± 2% -0.0 0.15 mpstat.cpu.all.irq% 0.27 ± 3% +0.2 0.47 mpstat.cpu.all.soft% 1.07 ± 5% +2.9 3.97 mpstat.cpu.all.sys% 5.63 ± 6% +38.5% 7.80 ± 5% mpstat.max_utilization_pct 11.33 ± 41% -100.0% 0.00 perf-c2c.DRAM.local 65.17 ± 64% -100.0% 0.00 perf-c2c.DRAM.remote 2717 ± 8% -100.0% 0.00 perf-c2c.HITM.local 32.17 ± 90% -100.0% 0.00 perf-c2c.HITM.remote 31117 ± 24% -54.1% 14280 ± 87% numa-meminfo.node0.Mapped 505070 -99.7% 1522 ± 24% numa-meminfo.node1.Active 504998 -99.7% 1490 ± 24% numa-meminfo.node1.Active(anon) 44594 ± 30% -60.3% 17722 ± 69% numa-meminfo.node1.Mapped 595059 -99.6% 2232 ± 16% numa-meminfo.node1.Shmem 151.45 -25.4% 112.99 uperf.Netstat_lo_in_Mbps 151.45 -25.4% 112.99 uperf.Netstat_lo_out_Mbps 105.36 -25.4% 78.60 uperf.Runstat_127.0.0.1_Mbps 105.36 -25.4% 78.60 uperf.Runstat_master_Mbps 25.17 -12.6% 22.00 uperf.time.percent_of_cpu_this_job_got 3325771 -25.8% 2467986 uperf.time.voluntary_context_switches 7923 ± 25% -53.9% 3654 ± 86% numa-vmstat.node0.nr_mapped 126255 -99.7% 373.42 ± 24% numa-vmstat.node1.nr_active_anon 11390 ± 30% -59.6% 4604 ± 68% numa-vmstat.node1.nr_mapped 148846 -99.6% 558.20 ± 16% numa-vmstat.node1.nr_shmem 126255 -99.7% 373.42 ± 24% numa-vmstat.node1.nr_zone_active_anon 334029 -67.7% 107918 ± 15% numa-vmstat.node1.numa_hit 313302 ± 3% -72.6% 85796 ± 25% numa-vmstat.node1.numa_local 506511 -99.5% 2716 meminfo.Active 506389 -99.5% 2652 meminfo.Active(anon) 3666247 -16.2% 3072894 meminfo.Cached 1380898 -41.7% 804785 meminfo.Committed_AS 730235 -11.6% 645257 meminfo.Inactive 729851 -11.6% 645012 meminfo.Inactive(anon) 75195 ± 17% -57.7% 31826 meminfo.Mapped 5569857 -9.0% 5068038 meminfo.Memused 606477 -97.8% 13355 meminfo.Shmem 5802899 -11.8% 5115559 meminfo.max_used_kB 126602 -99.5% 664.74 proc-vmstat.nr_active_anon 916622 -16.2% 768228 proc-vmstat.nr_file_pages 182518 -11.7% 161242 proc-vmstat.nr_inactive_anon 19151 ± 17% -57.1% 8207 proc-vmstat.nr_mapped 1519 +3.7% 1575 proc-vmstat.nr_page_table_pages 151678 -97.8% 3342 proc-vmstat.nr_shmem 22637 -3.8% 21768 proc-vmstat.nr_slab_reclaimable 126602 -99.5% 664.74 proc-vmstat.nr_zone_active_anon 182518 -11.7% 161242 proc-vmstat.nr_zone_inactive_anon 465325 -52.6% 220544 ± 2% proc-vmstat.numa_hit 399117 -58.2% 166673 ± 3% proc-vmstat.numa_local 66208 -18.7% 53838 proc-vmstat.numa_other 206736 -100.0% 33.17 ± 20% proc-vmstat.pgactivate 492731 -49.5% 248776 ± 2% proc-vmstat.pgalloc_normal 208304 ± 5% +54244.8% 1.132e+08 proc-vmstat.pgfault 275792 ± 2% -51.0% 135053 ± 4% proc-vmstat.pgfree 0.23 ± 3% -59.3% 0.09 ± 5% perf-stat.i.MPKI 1.982e+09 +106.5% 4.094e+09 perf-stat.i.branch-instructions 3.04 ± 2% -1.6 1.48 perf-stat.i.branch-miss-rate% 68611062 ± 2% -18.7% 55765171 perf-stat.i.branch-misses 7.33 ± 4% -1.3 6.00 ± 4% perf-stat.i.cache-miss-rate% 2783561 ± 3% -39.1% 1695747 ± 5% perf-stat.i.cache-misses 35458378 -19.3% 28615257 ± 2% perf-stat.i.cache-references 399282 ± 3% -24.1% 303251 perf-stat.i.context-switches 1.05 -20.3% 0.84 perf-stat.i.cpi 9.455e+09 ± 2% +74.1% 1.646e+10 perf-stat.i.cpu-cycles 137.55 -4.4% 131.46 perf-stat.i.cpu-migrations 11616 ± 17% +77.4% 20608 ± 4% perf-stat.i.cycles-between-cache-misses 9.825e+09 +101.0% 1.975e+10 perf-stat.i.instructions 1.01 +19.0% 1.20 perf-stat.i.ipc 6.21 ± 3% +1629.6% 107.36 perf-stat.i.metric.K/sec 4804 ± 6% +68281.0% 3285377 perf-stat.i.minor-faults 4804 ± 6% +68280.4% 3285377 perf-stat.i.page-faults 0.28 ± 2% -69.8% 0.09 ± 5% perf-stat.overall.MPKI 3.46 -2.1 1.36 perf-stat.overall.branch-miss-rate% 7.86 ± 2% -1.9 5.92 ± 6% perf-stat.overall.cache-miss-rate% 0.96 -13.4% 0.83 perf-stat.overall.cpi 3395 ± 2% +187.6% 9765 ± 5% perf-stat.overall.cycles-between-cache-misses 1.04 +15.5% 1.20 perf-stat.overall.ipc 1.925e+09 +106.2% 3.971e+09 perf-stat.ps.branch-instructions 66548387 -19.0% 53931766 perf-stat.ps.branch-misses 2708553 ± 3% -39.4% 1641008 ± 5% perf-stat.ps.cache-misses 34460219 -19.4% 27762399 ± 2% perf-stat.ps.cache-references 387430 ± 3% -24.1% 294073 perf-stat.ps.context-switches 9.192e+09 ± 2% +73.8% 1.598e+10 perf-stat.ps.cpu-cycles 133.55 -4.5% 127.51 perf-stat.ps.cpu-migrations 9.543e+09 +100.7% 1.916e+10 perf-stat.ps.instructions 4668 ± 6% +68177.7% 3187819 perf-stat.ps.minor-faults 4668 ± 6% +68177.1% 3187819 perf-stat.ps.page-faults 3.268e+11 +94.9% 6.368e+11 perf-stat.total.instructions 63.98 ± 3% -64.0 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 47.37 ± 2% -47.4 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 45.61 ± 6% -45.6 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 45.57 ± 6% -45.6 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 45.47 ± 6% -45.5 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 41.88 ± 6% -41.9 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 41.79 ± 2% -41.8 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 41.07 ± 2% -41.1 0.00 perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 40.56 ± 6% -40.6 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 22.56 ± 2% -22.6 0.00 perf-profile.calltrace.cycles-pp.sendto 21.42 ± 2% -21.4 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.sendto 21.36 ± 2% -21.4 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto 21.01 ± 2% -21.0 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto 20.87 ± 2% -20.9 0.00 perf-profile.calltrace.cycles-pp.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendto 19.40 ± 2% -19.4 0.00 perf-profile.calltrace.cycles-pp.udpv6_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64.entry_SYSCALL_64_after_hwframe 19.07 ± 3% -19.1 0.00 perf-profile.calltrace.cycles-pp.recvfrom 18.86 ± 2% -18.9 0.00 perf-profile.calltrace.cycles-pp.udp_sendmsg.udpv6_sendmsg.__sys_sendto.__x64_sys_sendto.do_syscall_64 16.78 ± 4% -16.8 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.recvfrom 16.72 ± 4% -16.7 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom 15.37 ± 4% -15.4 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom 15.24 ± 4% -15.2 0.00 perf-profile.calltrace.cycles-pp.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvfrom 14.58 ± 5% -14.6 0.00 perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 13.60 ± 3% -13.6 0.00 perf-profile.calltrace.cycles-pp.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64.entry_SYSCALL_64_after_hwframe 13.43 ± 3% -13.4 0.00 perf-profile.calltrace.cycles-pp.inet6_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom.do_syscall_64 13.32 ± 3% -13.3 0.00 perf-profile.calltrace.cycles-pp.udpv6_recvmsg.inet6_recvmsg.sock_recvmsg.__sys_recvfrom.__x64_sys_recvfrom 12.52 ± 4% -12.5 0.00 perf-profile.calltrace.cycles-pp.udp_send_skb.udp_sendmsg.udpv6_sendmsg.__sys_sendto.__x64_sys_sendto 12.18 ± 4% -12.2 0.00 perf-profile.calltrace.cycles-pp.ip_send_skb.udp_send_skb.udp_sendmsg.udpv6_sendmsg.__sys_sendto 11.64 ± 4% -11.6 0.00 perf-profile.calltrace.cycles-pp.ip_finish_output2.ip_send_skb.udp_send_skb.udp_sendmsg.udpv6_sendmsg 11.35 ± 4% -11.3 0.00 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.ip_send_skb.udp_send_skb.udp_sendmsg 9.29 ± 4% -9.3 0.00 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.ip_send_skb.udp_send_skb 9.14 ± 4% -9.1 0.00 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.ip_send_skb 9.13 ± 9% -9.1 0.00 perf-profile.calltrace.cycles-pp.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state 8.98 ± 4% -9.0 0.00 perf-profile.calltrace.cycles-pp.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2 8.83 -8.8 0.00 perf-profile.calltrace.cycles-pp.__skb_recv_udp.udpv6_recvmsg.inet6_recvmsg.sock_recvmsg.__sys_recvfrom 8.27 ± 4% -8.3 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit 7.80 ± 4% -7.8 0.00 perf-profile.calltrace.cycles-pp.__napi_poll.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip 7.68 ± 4% -7.7 0.00 perf-profile.calltrace.cycles-pp.process_backlog.__napi_poll.net_rx_action.handle_softirqs.do_softirq 7.37 ± 4% -7.4 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action.handle_softirqs 6.77 ± 4% -6.8 0.00 perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action 6.73 ± 4% -6.7 0.00 perf-profile.calltrace.cycles-pp.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll 6.54 ± 4% -6.5 0.00 perf-profile.calltrace.cycles-pp.__udp4_lib_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog 6.40 ± 2% -6.4 0.00 perf-profile.calltrace.cycles-pp.__skb_wait_for_more_packets.__skb_recv_udp.udpv6_recvmsg.inet6_recvmsg.sock_recvmsg 6.23 ± 9% -6.2 0.00 perf-profile.calltrace.cycles-pp.__sysvec_call_function_single.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter 5.95 ± 4% -5.9 0.00 perf-profile.calltrace.cycles-pp.udp_unicast_rcv_skb.__udp4_lib_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core 5.89 ± 4% -5.9 0.00 perf-profile.calltrace.cycles-pp.udp_queue_rcv_one_skb.udp_unicast_rcv_skb.__udp4_lib_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish 5.72 ± 9% -5.7 0.00 perf-profile.calltrace.cycles-pp.__flush_smp_call_function_queue.__sysvec_call_function_single.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt 5.13 ± 5% -5.1 0.00 perf-profile.calltrace.cycles-pp.schedule_timeout.__skb_wait_for_more_packets.__skb_recv_udp.udpv6_recvmsg.inet6_recvmsg 4.97 ± 3% -5.0 0.00 perf-profile.calltrace.cycles-pp.__udp_enqueue_schedule_skb.udp_queue_rcv_one_skb.udp_unicast_rcv_skb.__udp4_lib_rcv.ip_protocol_deliver_rcu 47.37 ± 2% -47.4 0.00 perf-profile.children.cycles-pp.common_startup_64 47.37 ± 2% -47.4 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 47.27 ± 2% -47.3 0.00 perf-profile.children.cycles-pp.do_idle 45.61 ± 6% -45.6 0.00 perf-profile.children.cycles-pp.start_secondary 43.59 ± 2% -43.6 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 42.19 ± 2% -42.2 0.00 perf-profile.children.cycles-pp.cpuidle_enter 41.99 ± 2% -42.0 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 41.83 -41.8 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 41.55 -41.5 0.00 perf-profile.children.cycles-pp.do_syscall_64 41.23 ± 2% -41.2 0.00 perf-profile.children.cycles-pp.acpi_idle_enter 41.18 ± 2% -41.2 0.00 perf-profile.children.cycles-pp.acpi_safe_halt 37.04 ± 2% -37.0 0.00 perf-profile.children.cycles-pp.asm_sysvec_call_function_single 22.75 ± 2% -22.7 0.00 perf-profile.children.cycles-pp.sendto 21.04 ± 2% -21.0 0.00 perf-profile.children.cycles-pp.__x64_sys_sendto 20.92 ± 2% -20.9 0.00 perf-profile.children.cycles-pp.__sys_sendto 19.44 ± 2% -19.4 0.00 perf-profile.children.cycles-pp.udpv6_sendmsg 19.25 ± 3% -19.3 0.00 perf-profile.children.cycles-pp.recvfrom 18.95 ± 2% -19.0 0.00 perf-profile.children.cycles-pp.udp_sendmsg 15.40 ± 4% -15.4 0.00 perf-profile.children.cycles-pp.__x64_sys_recvfrom 15.27 ± 4% -15.3 0.00 perf-profile.children.cycles-pp.__sys_recvfrom 13.62 ± 3% -13.6 0.00 perf-profile.children.cycles-pp.sock_recvmsg 13.43 ± 3% -13.4 0.00 perf-profile.children.cycles-pp.inet6_recvmsg 13.34 ± 3% -13.3 0.00 perf-profile.children.cycles-pp.udpv6_recvmsg 12.54 ± 4% -12.5 0.00 perf-profile.children.cycles-pp.udp_send_skb 12.20 ± 4% -12.2 0.00 perf-profile.children.cycles-pp.ip_send_skb 11.66 ± 4% -11.7 0.00 perf-profile.children.cycles-pp.ip_finish_output2 11.39 ± 4% -11.4 0.00 perf-profile.children.cycles-pp.__dev_queue_xmit 9.53 ± 4% -9.5 0.00 perf-profile.children.cycles-pp.handle_softirqs 9.37 ± 8% -9.4 0.00 perf-profile.children.cycles-pp.sysvec_call_function_single 9.35 ± 4% -9.3 0.00 perf-profile.children.cycles-pp.__local_bh_enable_ip 9.15 ± 4% -9.2 0.00 perf-profile.children.cycles-pp.do_softirq 8.88 -8.9 0.00 perf-profile.children.cycles-pp.__skb_recv_udp 8.28 ± 4% -8.3 0.00 perf-profile.children.cycles-pp.net_rx_action 7.82 ± 4% -7.8 0.00 perf-profile.children.cycles-pp.__napi_poll 7.70 ± 4% -7.7 0.00 perf-profile.children.cycles-pp.process_backlog 7.38 ± 4% -7.4 0.00 perf-profile.children.cycles-pp.__netif_receive_skb_one_core 7.05 -7.0 0.00 perf-profile.children.cycles-pp.__schedule 6.78 ± 4% -6.8 0.00 perf-profile.children.cycles-pp.ip_local_deliver_finish 6.75 ± 4% -6.7 0.00 perf-profile.children.cycles-pp.ip_protocol_deliver_rcu 6.56 ± 4% -6.6 0.00 perf-profile.children.cycles-pp.__udp4_lib_rcv 6.42 ± 2% -6.4 0.00 perf-profile.children.cycles-pp.__skb_wait_for_more_packets 6.31 ± 8% -6.3 0.00 perf-profile.children.cycles-pp.__sysvec_call_function_single 5.96 ± 4% -6.0 0.00 perf-profile.children.cycles-pp.udp_unicast_rcv_skb 5.92 ± 4% -5.9 0.00 perf-profile.children.cycles-pp.udp_queue_rcv_one_skb 5.80 ± 8% -5.8 0.00 perf-profile.children.cycles-pp.__flush_smp_call_function_queue 5.16 ± 4% -5.2 0.00 perf-profile.children.cycles-pp.schedule_timeout 4.98 ± 3% -5.0 0.00 perf-profile.children.cycles-pp.__udp_enqueue_schedule_skb 29.86 ± 4% -29.9 0.00 perf-profile.self.cycles-pp.acpi_safe_halt *************************************************************************************************** lkp-icl-2sp2: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory ========================================================================================= cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase: cs-localhost/gcc-12/performance/ipv4/x86_64-rhel-8.3/50%/debian-12-x86_64-20240206.cgz/300s/lkp-icl-2sp2/SCTP_RR/netperf commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 20146 -14.1% 17300 ± 2% uptime.idle 1.568e+10 -20.5% 1.246e+10 cpuidle..time 1.702e+09 -32.2% 1.154e+09 cpuidle..usage 3115754 -88.3% 365170 ± 16% numa-numastat.node1.local_node 3215262 -87.0% 419236 ± 8% numa-numastat.node1.numa_hit 37.40 -16.6% 31.18 vmstat.cpu.id 11031758 -31.8% 7527987 vmstat.system.cs 5666321 -23.4% 4342549 vmstat.system.in 37.03 -6.3 30.75 mpstat.cpu.all.idle% 3.69 -0.7 3.02 mpstat.cpu.all.irq% 18.00 +7.2 25.18 mpstat.cpu.all.soft% 2.07 -0.7 1.38 mpstat.cpu.all.usr% 81.67 ± 19% -100.0% 0.00 perf-c2c.DRAM.local 3324 ± 14% -100.0% 0.00 perf-c2c.DRAM.remote 246856 -100.0% 0.00 perf-c2c.HITM.local 1771 ± 11% -100.0% 0.00 perf-c2c.HITM.remote 248627 -100.0% 0.00 perf-c2c.HITM.total 2836525 -31.4% 1944538 netperf.ThroughputBoth_total_tps 44320 -31.4% 30383 netperf.ThroughputBoth_tps 2836525 -31.4% 1944538 netperf.Throughput_total_tps 44320 -31.4% 30383 netperf.Throughput_tps 15538 ± 2% +1516.1% 251120 ± 5% netperf.time.involuntary_context_switches 2180 -18.6% 1774 netperf.time.percent_of_cpu_this_job_got 6401 -17.7% 5264 netperf.time.system_time 253.51 -40.6% 150.54 netperf.time.user_time 8.508e+08 -31.5% 5.829e+08 netperf.time.voluntary_context_switches 8.51e+08 -31.4% 5.834e+08 netperf.workload 3568833 ± 6% -99.9% 5313 ± 2% meminfo.Active 3568785 ± 6% -99.9% 5273 ± 2% meminfo.Active(anon) 10615057 -71.1% 3072650 meminfo.Cached 8430375 -89.9% 851992 meminfo.Committed_AS 4707200 ± 4% -85.5% 681046 meminfo.Inactive 4706988 ± 4% -85.5% 680826 meminfo.Inactive(anon) 130914 -16.5% 109250 meminfo.KReclaimable 2445242 ± 6% -98.6% 33828 meminfo.Mapped 13265505 -55.6% 5893945 meminfo.Memused 16829 ± 2% -26.3% 12400 meminfo.PageTables 130914 -16.5% 109250 meminfo.SReclaimable 7553109 -99.9% 11212 meminfo.Shmem 13506404 -54.0% 6207695 meminfo.max_used_kB 225772 ± 64% -38.0% 139999 ±100% numa-meminfo.node0.AnonPages 227536 ± 64% -37.7% 141653 ± 99% numa-meminfo.node0.Inactive 227365 ± 64% -37.8% 141435 ± 99% numa-meminfo.node0.Inactive(anon) 122078 ± 8% -85.9% 17153 ± 74% numa-meminfo.node0.Mapped 9967 ± 5% -38.4% 6143 ± 8% numa-meminfo.node0.PageTables 3547939 ± 6% -99.9% 2434 ± 19% numa-meminfo.node1.Active 3547931 ± 6% -99.9% 2434 ± 19% numa-meminfo.node1.Active(anon) 10022700 ± 6% -84.9% 1510608 ± 82% numa-meminfo.node1.FilePages 4491145 ± 7% -88.0% 539362 ± 26% numa-meminfo.node1.Inactive 4491105 ± 7% -88.0% 539359 ± 26% numa-meminfo.node1.Inactive(anon) 2331352 ± 6% -99.3% 16689 ± 75% numa-meminfo.node1.Mapped 11420187 ± 6% -73.5% 3022233 ± 39% numa-meminfo.node1.MemUsed 7538938 -100.0% 3512 ± 13% numa-meminfo.node1.Shmem 56484 ± 64% -38.0% 34994 ±100% numa-vmstat.node0.nr_anon_pages 56882 ± 64% -37.8% 35353 ± 99% numa-vmstat.node0.nr_inactive_anon 30662 ± 9% -85.5% 4434 ± 73% numa-vmstat.node0.nr_mapped 2501 ± 5% -38.5% 1537 ± 8% numa-vmstat.node0.nr_page_table_pages 56882 ± 64% -37.8% 35353 ± 99% numa-vmstat.node0.nr_zone_inactive_anon 889164 ± 6% -99.9% 608.62 ± 19% numa-vmstat.node1.nr_active_anon 2505432 ± 6% -84.9% 377651 ± 82% numa-vmstat.node1.nr_file_pages 1120376 ± 7% -88.0% 134862 ± 26% numa-vmstat.node1.nr_inactive_anon 581125 ± 6% -99.3% 4276 ± 75% numa-vmstat.node1.nr_mapped 1884492 -100.0% 877.88 ± 13% numa-vmstat.node1.nr_shmem 889164 ± 6% -99.9% 608.62 ± 19% numa-vmstat.node1.nr_zone_active_anon 1120376 ± 7% -88.0% 134862 ± 26% numa-vmstat.node1.nr_zone_inactive_anon 3212683 -87.0% 417119 ± 8% numa-vmstat.node1.numa_hit 3113203 -88.3% 363053 ± 16% numa-vmstat.node1.numa_local 889503 ± 6% -99.9% 1318 ± 2% proc-vmstat.nr_active_anon 181558 -6.6% 169658 proc-vmstat.nr_anon_pages 6224451 +3.0% 6408360 proc-vmstat.nr_dirty_background_threshold 12464123 +3.0% 12832389 proc-vmstat.nr_dirty_threshold 2652787 -71.0% 768163 proc-vmstat.nr_file_pages 62620501 +2.9% 64462280 proc-vmstat.nr_free_pages 1178408 ± 4% -85.6% 170197 proc-vmstat.nr_inactive_anon 612759 ± 6% -98.6% 8708 proc-vmstat.nr_mapped 4212 ± 2% -26.5% 3098 proc-vmstat.nr_page_table_pages 1887299 -99.9% 2803 proc-vmstat.nr_shmem 32725 -16.5% 27312 proc-vmstat.nr_slab_reclaimable 889503 ± 6% -99.9% 1318 ± 2% proc-vmstat.nr_zone_active_anon 1178408 ± 4% -85.6% 170197 proc-vmstat.nr_zone_inactive_anon 290672 ± 45% -96.8% 9303 ± 41% proc-vmstat.numa_hint_faults 175901 ± 76% -97.5% 4457 ± 89% proc-vmstat.numa_hint_faults_local 3663440 -77.3% 830893 proc-vmstat.numa_hit 3530845 -79.5% 723050 proc-vmstat.numa_local 132548 -18.6% 107828 proc-vmstat.numa_other 467705 ± 25% -87.5% 58338 ± 63% proc-vmstat.numa_pte_updates 2487950 -100.0% 114.50 ± 16% proc-vmstat.pgactivate 3816007 -74.2% 983809 proc-vmstat.pgalloc_normal 1529029 ± 8% +34293.1% 5.259e+08 proc-vmstat.pgfault 1080189 -31.0% 745470 proc-vmstat.pgfree 50047 ± 16% -19.9% 40097 proc-vmstat.pgreuse 0.13 ± 12% -55.6% 0.06 ± 9% perf-stat.i.MPKI 3.129e+10 +64.8% 5.157e+10 perf-stat.i.branch-instructions 1.12 -0.5 0.63 perf-stat.i.branch-miss-rate% 3.419e+08 -14.1% 2.935e+08 perf-stat.i.branch-misses 19925383 ± 13% -32.6% 13438806 ± 9% perf-stat.i.cache-misses 1.336e+09 -24.7% 1.006e+09 perf-stat.i.cache-references 11135389 -31.8% 7599205 perf-stat.i.context-switches 1.55 -28.8% 1.10 perf-stat.i.cpi 2.551e+11 +3.4% 2.637e+11 perf-stat.i.cpu-cycles 86514 ± 4% +26.7% 109601 perf-stat.i.cpu-migrations 14145 ± 17% +43.8% 20343 ± 8% perf-stat.i.cycles-between-cache-misses 1.648e+11 +51.5% 2.497e+11 perf-stat.i.instructions 0.65 +45.3% 0.95 perf-stat.i.ipc 87.60 -1.3% 86.48 perf-stat.i.metric.K/sec 4794 ± 9% +35590.7% 1711366 perf-stat.i.minor-faults 4795 ± 9% +35590.6% 1711366 perf-stat.i.page-faults 0.12 ± 13% -55.5% 0.05 ± 9% perf-stat.overall.MPKI 1.09 -0.5 0.57 perf-stat.overall.branch-miss-rate% 1.55 -31.8% 1.06 perf-stat.overall.cpi 13042 ± 14% +51.7% 19779 ± 8% perf-stat.overall.cycles-between-cache-misses 0.65 +46.6% 0.95 perf-stat.overall.ipc 59142 +121.0% 130686 perf-stat.overall.path-length 3.118e+10 +64.8% 5.139e+10 perf-stat.ps.branch-instructions 3.407e+08 -14.1% 2.925e+08 perf-stat.ps.branch-misses 19880579 ± 13% -32.6% 13396005 ± 9% perf-stat.ps.cache-misses 1.332e+09 -24.7% 1.003e+09 perf-stat.ps.cache-references 11093234 -31.7% 7573495 perf-stat.ps.context-switches 2.542e+11 +3.4% 2.628e+11 perf-stat.ps.cpu-cycles 87621 ± 3% +24.9% 109407 perf-stat.ps.cpu-migrations 1.642e+11 +51.6% 2.489e+11 perf-stat.ps.instructions 4770 ± 9% +35666.9% 1706137 perf-stat.ps.minor-faults 4770 ± 9% +35666.8% 1706137 perf-stat.ps.page-faults 5.033e+13 +51.5% 7.624e+13 perf-stat.total.instructions 3403878 -9.6% 3076953 sched_debug.cfs_rq:/.avg_vruntime.avg 5906828 ± 7% +51.1% 8927261 ± 9% sched_debug.cfs_rq:/.avg_vruntime.max 3312785 -12.6% 2896386 sched_debug.cfs_rq:/.avg_vruntime.min 243929 ± 9% +164.4% 644860 ± 7% sched_debug.cfs_rq:/.avg_vruntime.stddev 1.31 ± 11% +31.9% 1.72 ± 17% sched_debug.cfs_rq:/.h_nr_running.max 267222 ± 12% -37.6% 166754 ± 11% sched_debug.cfs_rq:/.left_deadline.avg 3510445 -9.8% 3168156 ± 4% sched_debug.cfs_rq:/.left_deadline.max 902490 ± 6% -25.0% 677076 ± 4% sched_debug.cfs_rq:/.left_deadline.stddev 267208 ± 12% -37.6% 166745 ± 11% sched_debug.cfs_rq:/.left_vruntime.avg 3510259 -9.8% 3167981 ± 4% sched_debug.cfs_rq:/.left_vruntime.max 902441 ± 6% -25.0% 677037 ± 4% sched_debug.cfs_rq:/.left_vruntime.stddev 7978 ± 2% +21.3% 9675 ± 11% sched_debug.cfs_rq:/.load.avg 22814 ± 14% +289.5% 88865 ±147% sched_debug.cfs_rq:/.load.max 3403878 -9.6% 3076953 sched_debug.cfs_rq:/.min_vruntime.avg 5906828 ± 7% +51.1% 8927261 ± 9% sched_debug.cfs_rq:/.min_vruntime.max 3312785 -12.6% 2896386 sched_debug.cfs_rq:/.min_vruntime.min 243929 ± 9% +164.4% 644860 ± 7% sched_debug.cfs_rq:/.min_vruntime.stddev 267208 ± 12% -37.6% 166745 ± 11% sched_debug.cfs_rq:/.right_vruntime.avg 3510259 -9.8% 3167983 ± 4% sched_debug.cfs_rq:/.right_vruntime.max 902441 ± 6% -25.0% 677037 ± 4% sched_debug.cfs_rq:/.right_vruntime.stddev 124.18 ± 5% +19.4% 148.27 ± 2% sched_debug.cfs_rq:/.runnable_avg.stddev 410.44 -8.7% 374.59 sched_debug.cfs_rq:/.util_avg.avg 123.79 ± 5% +19.4% 147.85 ± 2% sched_debug.cfs_rq:/.util_avg.stddev 688.89 ± 8% +38.7% 955.28 ± 3% sched_debug.cfs_rq:/.util_est.max 451880 ± 32% +68.2% 759893 ± 15% sched_debug.cpu.avg_idle.max 6812 ± 2% +66.2% 11326 sched_debug.cpu.avg_idle.min 57452 ± 22% +59.0% 91375 ± 12% sched_debug.cpu.avg_idle.stddev 1511 ± 10% +166.4% 4026 ± 6% sched_debug.cpu.clock_task.stddev 1.31 ± 11% +34.0% 1.75 ± 9% sched_debug.cpu.nr_running.max 13041493 -24.7% 9825289 sched_debug.cpu.nr_switches.avg 13244004 -23.9% 10073573 sched_debug.cpu.nr_switches.max 8013647 ± 10% -84.4% 1247149 ±125% sched_debug.cpu.nr_switches.min 499070 ± 10% +82.7% 911672 ± 9% sched_debug.cpu.nr_switches.stddev 9.34 ± 3% -11.6% 8.26 ± 5% sched_debug.cpu.nr_uninterruptible.stddev 46.47 -46.5 0.00 perf-profile.calltrace.cycles-pp.___sys_sendmsg.__sys_sendmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendmsg 45.66 -45.7 0.00 perf-profile.calltrace.cycles-pp.____sys_sendmsg.___sys_sendmsg.__sys_sendmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe 45.53 ± 18% -45.5 0.00 perf-profile.calltrace.cycles-pp.main 43.65 -43.6 0.00 perf-profile.calltrace.cycles-pp.sctp_sendmsg.____sys_sendmsg.___sys_sendmsg.__sys_sendmsg.do_syscall_64 42.48 -42.5 0.00 perf-profile.calltrace.cycles-pp.sctp_sendmsg_to_asoc.sctp_sendmsg.____sys_sendmsg.___sys_sendmsg.__sys_sendmsg 37.52 -37.5 0.00 perf-profile.calltrace.cycles-pp.sctp_primitive_SEND.sctp_sendmsg_to_asoc.sctp_sendmsg.____sys_sendmsg.___sys_sendmsg 35.70 -35.7 0.00 perf-profile.calltrace.cycles-pp.sctp_outq_flush.sctp_cmd_interpreter.sctp_do_sm.sctp_primitive_SEND.sctp_sendmsg_to_asoc 29.47 -29.5 0.00 perf-profile.calltrace.cycles-pp.sctp_packet_transmit.sctp_outq_flush.sctp_cmd_interpreter.sctp_do_sm.sctp_primitive_SEND 29.39 -29.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 26.68 -26.7 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 26.58 -26.6 0.00 perf-profile.calltrace.cycles-pp.__ip_queue_xmit.sctp_packet_transmit.sctp_outq_flush.sctp_cmd_interpreter.sctp_do_sm 26.51 -26.5 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 26.48 -26.5 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 26.42 -26.4 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 25.57 -25.6 0.00 perf-profile.calltrace.cycles-pp.ip_finish_output2.__ip_queue_xmit.sctp_packet_transmit.sctp_outq_flush.sctp_cmd_interpreter 25.24 -25.2 0.00 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.sctp_packet_transmit.sctp_outq_flush 24.20 -24.2 0.00 perf-profile.calltrace.cycles-pp.sendmsg.process_requests.spawn_child.accept_connection.accept_connections 24.11 -24.1 0.00 perf-profile.calltrace.cycles-pp.sendmsg.main 23.86 -23.9 0.00 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit.sctp_packet_transmit 23.73 -23.7 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.sendmsg.process_requests.spawn_child.accept_connection 23.72 -23.7 0.00 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2.__ip_queue_xmit 23.71 -23.7 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendmsg.process_requests.spawn_child 23.62 -23.6 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.sendmsg.main 23.60 -23.6 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendmsg.main 23.59 -23.6 0.00 perf-profile.calltrace.cycles-pp.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip_finish_output2 23.53 -23.5 0.00 perf-profile.calltrace.cycles-pp.__sys_sendmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendmsg.process_requests 23.43 -23.4 0.00 perf-profile.calltrace.cycles-pp.__sys_sendmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe.sendmsg.main 22.88 -22.9 0.00 perf-profile.calltrace.cycles-pp.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit 22.79 -22.8 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 22.46 -22.5 0.00 perf-profile.calltrace.cycles-pp.__napi_poll.net_rx_action.handle_softirqs.do_softirq.__local_bh_enable_ip 22.35 -22.3 0.00 perf-profile.calltrace.cycles-pp.process_backlog.__napi_poll.net_rx_action.handle_softirqs.do_softirq 22.02 -22.0 0.00 perf-profile.calltrace.cycles-pp.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action.handle_softirqs 21.80 -21.8 0.00 perf-profile.calltrace.cycles-pp.recvmsg 21.50 -21.5 0.00 perf-profile.calltrace.cycles-pp.sctp_cmd_interpreter.sctp_do_sm.sctp_primitive_SEND.sctp_sendmsg_to_asoc.sctp_sendmsg 21.31 -21.3 0.00 perf-profile.calltrace.cycles-pp.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll.net_rx_action 21.26 -21.3 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 21.24 -21.2 0.00 perf-profile.calltrace.cycles-pp.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog.__napi_poll 21.08 -21.1 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 20.96 -21.0 0.00 perf-profile.calltrace.cycles-pp.sctp_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core.process_backlog 20.92 ± 41% -20.9 0.00 perf-profile.calltrace.cycles-pp.accept_connection.accept_connections.main 20.92 ± 41% -20.9 0.00 perf-profile.calltrace.cycles-pp.accept_connections.main 20.92 ± 41% -20.9 0.00 perf-profile.calltrace.cycles-pp.process_requests.spawn_child.accept_connection.accept_connections.main 20.92 ± 41% -20.9 0.00 perf-profile.calltrace.cycles-pp.spawn_child.accept_connection.accept_connections.main 20.40 -20.4 0.00 perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 20.34 -20.3 0.00 perf-profile.calltrace.cycles-pp.sctp_do_sm.sctp_primitive_SEND.sctp_sendmsg_to_asoc.sctp_sendmsg.____sys_sendmsg 20.05 -20.1 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.recvmsg 20.02 -20.0 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvmsg 18.81 -18.8 0.00 perf-profile.calltrace.cycles-pp.__sys_recvmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvmsg 18.35 -18.4 0.00 perf-profile.calltrace.cycles-pp.___sys_recvmsg.__sys_recvmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe.recvmsg 17.76 -17.8 0.00 perf-profile.calltrace.cycles-pp.sctp_assoc_bh_rcv.sctp_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish.__netif_receive_skb_one_core 17.53 -17.5 0.00 perf-profile.calltrace.cycles-pp.____sys_recvmsg.___sys_recvmsg.__sys_recvmsg.do_syscall_64.entry_SYSCALL_64_after_hwframe 17.04 -17.0 0.00 perf-profile.calltrace.cycles-pp.sock_recvmsg.____sys_recvmsg.___sys_recvmsg.__sys_recvmsg.do_syscall_64 16.61 -16.6 0.00 perf-profile.calltrace.cycles-pp.inet_recvmsg.sock_recvmsg.____sys_recvmsg.___sys_recvmsg.__sys_recvmsg 16.40 -16.4 0.00 perf-profile.calltrace.cycles-pp.sctp_recvmsg.inet_recvmsg.sock_recvmsg.____sys_recvmsg.___sys_recvmsg 14.32 ± 2% -14.3 0.00 perf-profile.calltrace.cycles-pp.sctp_do_sm.sctp_assoc_bh_rcv.sctp_rcv.ip_protocol_deliver_rcu.ip_local_deliver_finish 13.68 ± 2% -13.7 0.00 perf-profile.calltrace.cycles-pp.sctp_cmd_interpreter.sctp_do_sm.sctp_assoc_bh_rcv.sctp_rcv.ip_protocol_deliver_rcu 9.63 -9.6 0.00 perf-profile.calltrace.cycles-pp.sctp_skb_recv_datagram.sctp_recvmsg.inet_recvmsg.sock_recvmsg.____sys_recvmsg 9.28 ± 2% -9.3 0.00 perf-profile.calltrace.cycles-pp.sctp_outq_sack.sctp_cmd_interpreter.sctp_do_sm.sctp_assoc_bh_rcv.sctp_rcv 9.08 -9.1 0.00 perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 7.30 -7.3 0.00 perf-profile.calltrace.cycles-pp.schedule_timeout.sctp_skb_recv_datagram.sctp_recvmsg.inet_recvmsg.sock_recvmsg 7.05 -7.1 0.00 perf-profile.calltrace.cycles-pp.schedule.schedule_timeout.sctp_skb_recv_datagram.sctp_recvmsg.inet_recvmsg 6.92 -6.9 0.00 perf-profile.calltrace.cycles-pp.__schedule.schedule.schedule_timeout.sctp_skb_recv_datagram.sctp_recvmsg 6.02 -6.0 0.00 perf-profile.calltrace.cycles-pp.sctp_outq_flush_data.sctp_outq_flush.sctp_cmd_interpreter.sctp_do_sm.sctp_primitive_SEND 5.78 ± 3% -5.8 0.00 perf-profile.calltrace.cycles-pp.sctp_chunk_put.sctp_outq_sack.sctp_cmd_interpreter.sctp_do_sm.sctp_assoc_bh_rcv 5.48 -5.5 0.00 perf-profile.calltrace.cycles-pp.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state 5.33 ± 4% -5.3 0.00 perf-profile.calltrace.cycles-pp.consume_skb.sctp_chunk_put.sctp_outq_sack.sctp_cmd_interpreter.sctp_do_sm 67.76 -67.8 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 67.49 -67.5 0.00 perf-profile.children.cycles-pp.do_syscall_64 48.53 -48.5 0.00 perf-profile.children.cycles-pp.sendmsg 47.03 -47.0 0.00 perf-profile.children.cycles-pp.__sys_sendmsg 46.55 -46.6 0.00 perf-profile.children.cycles-pp.___sys_sendmsg 45.75 -45.7 0.00 perf-profile.children.cycles-pp.____sys_sendmsg 45.59 ± 18% -45.6 0.00 perf-profile.children.cycles-pp.main 44.18 -44.2 0.00 perf-profile.children.cycles-pp.sctp_sendmsg 42.58 -42.6 0.00 perf-profile.children.cycles-pp.sctp_sendmsg_to_asoc 37.58 -37.6 0.00 perf-profile.children.cycles-pp.sctp_primitive_SEND 37.38 -37.4 0.00 perf-profile.children.cycles-pp.sctp_do_sm 36.84 -36.8 0.00 perf-profile.children.cycles-pp.sctp_cmd_interpreter 35.94 -35.9 0.00 perf-profile.children.cycles-pp.sctp_outq_flush 29.54 -29.5 0.00 perf-profile.children.cycles-pp.sctp_packet_transmit 26.68 -26.7 0.00 perf-profile.children.cycles-pp.common_startup_64 26.68 -26.7 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 26.63 -26.6 0.00 perf-profile.children.cycles-pp.__ip_queue_xmit 26.63 -26.6 0.00 perf-profile.children.cycles-pp.do_idle 26.51 -26.5 0.00 perf-profile.children.cycles-pp.start_secondary 25.62 -25.6 0.00 perf-profile.children.cycles-pp.ip_finish_output2 25.30 -25.3 0.00 perf-profile.children.cycles-pp.__dev_queue_xmit 24.78 -24.8 0.00 perf-profile.children.cycles-pp.accept_connection 24.78 -24.8 0.00 perf-profile.children.cycles-pp.accept_connections 24.78 -24.8 0.00 perf-profile.children.cycles-pp.process_requests 24.78 -24.8 0.00 perf-profile.children.cycles-pp.spawn_child 24.02 -24.0 0.00 perf-profile.children.cycles-pp.__local_bh_enable_ip 23.80 -23.8 0.00 perf-profile.children.cycles-pp.handle_softirqs 23.78 -23.8 0.00 perf-profile.children.cycles-pp.do_softirq 22.97 -23.0 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 22.92 -22.9 0.00 perf-profile.children.cycles-pp.net_rx_action 22.50 -22.5 0.00 perf-profile.children.cycles-pp.__napi_poll 22.38 -22.4 0.00 perf-profile.children.cycles-pp.process_backlog 22.06 -22.1 0.00 perf-profile.children.cycles-pp.__netif_receive_skb_one_core 21.93 -21.9 0.00 perf-profile.children.cycles-pp.recvmsg 21.41 -21.4 0.00 perf-profile.children.cycles-pp.cpuidle_enter 21.34 -21.3 0.00 perf-profile.children.cycles-pp.ip_local_deliver_finish 21.28 -21.3 0.00 perf-profile.children.cycles-pp.ip_protocol_deliver_rcu 21.26 -21.3 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 21.03 -21.0 0.00 perf-profile.children.cycles-pp.sctp_rcv 20.54 -20.5 0.00 perf-profile.children.cycles-pp.acpi_idle_enter 20.42 -20.4 0.00 perf-profile.children.cycles-pp.acpi_safe_halt 18.83 -18.8 0.00 perf-profile.children.cycles-pp.__sys_recvmsg 18.36 -18.4 0.00 perf-profile.children.cycles-pp.___sys_recvmsg 17.81 -17.8 0.00 perf-profile.children.cycles-pp.sctp_assoc_bh_rcv 17.75 -17.8 0.00 perf-profile.children.cycles-pp.asm_sysvec_call_function_single 17.55 -17.6 0.00 perf-profile.children.cycles-pp.____sys_recvmsg 17.06 -17.1 0.00 perf-profile.children.cycles-pp.sock_recvmsg 16.84 -16.8 0.00 perf-profile.children.cycles-pp.sctp_recvmsg 16.62 -16.6 0.00 perf-profile.children.cycles-pp.inet_recvmsg 9.68 -9.7 0.00 perf-profile.children.cycles-pp.sctp_skb_recv_datagram 9.54 -9.5 0.00 perf-profile.children.cycles-pp.__schedule 9.36 ± 2% -9.4 0.00 perf-profile.children.cycles-pp.sctp_outq_sack 8.24 ± 2% -8.2 0.00 perf-profile.children.cycles-pp.sctp_chunk_put 7.62 -7.6 0.00 perf-profile.children.cycles-pp.schedule_timeout 7.06 -7.1 0.00 perf-profile.children.cycles-pp.schedule 6.38 ± 3% -6.4 0.00 perf-profile.children.cycles-pp.consume_skb 6.14 ± 3% -6.1 0.00 perf-profile.children.cycles-pp.skb_release_head_state 6.14 -6.1 0.00 perf-profile.children.cycles-pp.sctp_outq_flush_data 5.54 -5.5 0.00 perf-profile.children.cycles-pp.sysvec_call_function_single 13.56 -13.6 0.00 perf-profile.self.cycles-pp.acpi_safe_halt *************************************************************************************************** lkp-skl-fpga01: 104 threads 2 sockets (Skylake) with 192G memory ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-12/performance/x86_64-rhel-8.3/process/100%/debian-12-x86_64-20240206.cgz/lkp-skl-fpga01/context_switch1/will-it-scale commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 64399 ± 2% +268.0% 236970 ± 13% cpuidle..usage 1317337 ± 5% -75.7% 320761 ± 39% numa-numastat.node1.local_node 1363498 ± 5% -72.5% 374923 ± 32% numa-numastat.node1.numa_hit 27.91 -60.0% 11.18 ± 3% vmstat.cpu.us 18301349 -60.6% 7203175 ± 4% vmstat.system.cs 129807 +465.4% 733957 ± 5% vmstat.system.in 18547591 -61.9% 7060612 ± 4% will-it-scale.104.processes 178341 -61.9% 67890 ± 4% will-it-scale.per_process_ops 18547591 -61.9% 7060612 ± 4% will-it-scale.workload 0.41 ± 4% +0.9 1.33 mpstat.cpu.all.irq% 0.01 ± 3% -0.0 0.01 ± 2% mpstat.cpu.all.soft% 70.70 +16.1 86.78 mpstat.cpu.all.sys% 28.09 -17.0 11.11 ± 4% mpstat.cpu.all.usr% 197.00 ± 65% -100.0% 0.00 perf-c2c.DRAM.local 345.00 ± 13% -100.0% 0.00 perf-c2c.DRAM.remote 6145 ± 30% -100.0% 0.00 perf-c2c.HITM.local 207.50 ± 9% -100.0% 0.00 perf-c2c.HITM.remote 6353 ± 29% -100.0% 0.00 perf-c2c.HITM.total 2719746 ± 3% -99.9% 3450 meminfo.Active 2719698 ± 3% -99.9% 3402 meminfo.Active(anon) 6617563 ± 2% -53.4% 3084370 meminfo.Cached 6890243 ± 2% -51.3% 3357208 meminfo.Committed_AS 1534356 ± 7% -54.6% 696751 meminfo.Inactive 1534138 ± 7% -54.6% 696537 meminfo.Inactive(anon) 107936 -12.2% 94813 meminfo.KReclaimable 528224 ± 12% -93.0% 36788 meminfo.Mapped 8481255 ± 2% -36.9% 5355841 meminfo.Memused 107936 -12.2% 94813 meminfo.SReclaimable 3544842 ± 4% -99.7% 11703 ± 3% meminfo.Shmem 9043642 ± 2% -38.6% 5554552 meminfo.max_used_kB 322297 ± 32% -50.7% 159027 ± 76% numa-meminfo.node0.AnonPages 422986 ± 24% -59.1% 173110 ± 74% numa-meminfo.node0.AnonPages.max 323931 ± 32% -50.6% 160140 ± 76% numa-meminfo.node0.Inactive 323754 ± 32% -50.6% 159998 ± 76% numa-meminfo.node0.Inactive(anon) 75677 ± 11% -66.8% 25143 ± 48% numa-meminfo.node0.Mapped 2716604 ± 3% -99.9% 2155 ± 21% numa-meminfo.node1.Active 2716596 ± 3% -99.9% 2139 ± 21% numa-meminfo.node1.Active(anon) 4505110 ± 29% -71.2% 1298660 ± 98% numa-meminfo.node1.FilePages 1211376 ± 16% -55.7% 536704 ± 22% numa-meminfo.node1.Inactive 1211335 ± 16% -55.7% 536631 ± 22% numa-meminfo.node1.Inactive(anon) 451272 ± 15% -97.4% 11673 ±103% numa-meminfo.node1.Mapped 5431186 ± 24% -52.3% 2592950 ± 46% numa-meminfo.node1.MemUsed 3537633 ± 4% -99.8% 6102 ± 19% numa-meminfo.node1.Shmem 80524 ± 32% -50.6% 39762 ± 76% numa-vmstat.node0.nr_anon_pages 80887 ± 32% -50.5% 40003 ± 76% numa-vmstat.node0.nr_inactive_anon 19126 ± 11% -66.2% 6470 ± 47% numa-vmstat.node0.nr_mapped 80887 ± 32% -50.5% 40003 ± 76% numa-vmstat.node0.nr_zone_inactive_anon 679371 ± 3% -99.9% 535.24 ± 21% numa-vmstat.node1.nr_active_anon 1126311 ± 29% -71.2% 324665 ± 98% numa-vmstat.node1.nr_file_pages 302619 ± 16% -55.7% 134149 ± 22% numa-vmstat.node1.nr_inactive_anon 113050 ± 15% -97.4% 2961 ±103% numa-vmstat.node1.nr_mapped 884442 ± 4% -99.8% 1526 ± 19% numa-vmstat.node1.nr_shmem 679371 ± 3% -99.9% 535.23 ± 21% numa-vmstat.node1.nr_zone_active_anon 302620 ± 16% -55.7% 134149 ± 22% numa-vmstat.node1.nr_zone_inactive_anon 1362527 ± 5% -72.5% 374747 ± 32% numa-vmstat.node1.numa_hit 1316366 ± 5% -75.6% 320586 ± 39% numa-vmstat.node1.numa_local 120076 ± 10% +44.5% 173481 ± 10% sched_debug.cfs_rq:/.avg_vruntime.stddev 1.29 -14.3% 1.11 ± 2% sched_debug.cfs_rq:/.h_nr_running.avg 1093 ±141% +350.1% 4923 ± 22% sched_debug.cfs_rq:/.load.min 120076 ± 10% +44.5% 173481 ± 10% sched_debug.cfs_rq:/.min_vruntime.stddev 1815 ± 4% +13.8% 2065 sched_debug.cfs_rq:/.runnable_avg.max 1209 -12.4% 1059 ± 2% sched_debug.cfs_rq:/.runnable_avg.min 97.73 ± 12% +66.6% 162.81 ± 5% sched_debug.cfs_rq:/.runnable_avg.stddev 804.36 ± 3% -10.7% 718.36 ± 3% sched_debug.cfs_rq:/.util_avg.min 618.13 +16.3% 719.15 ± 2% sched_debug.cfs_rq:/.util_est.avg 1100 ± 5% +36.4% 1501 ± 7% sched_debug.cfs_rq:/.util_est.max 793984 ± 2% -34.7% 518681 ± 3% sched_debug.cpu.avg_idle.avg 1164653 ± 8% -19.1% 942294 ± 8% sched_debug.cpu.avg_idle.max 187692 ± 10% -18.2% 153546 ± 7% sched_debug.cpu.avg_idle.stddev 2817 ± 7% -26.8% 2061 ± 8% sched_debug.cpu.curr->pid.min 26222425 -55.7% 11622454 ± 3% sched_debug.cpu.nr_switches.avg 27158282 -55.2% 12180443 ± 4% sched_debug.cpu.nr_switches.max 24464466 ± 2% -57.5% 10402856 ± 4% sched_debug.cpu.nr_switches.min 605107 ± 12% -44.4% 336330 ± 6% sched_debug.cpu.nr_switches.stddev 679317 ± 3% -99.9% 850.20 proc-vmstat.nr_active_anon 178194 -2.9% 172994 proc-vmstat.nr_anon_pages 4677129 +1.7% 4755122 proc-vmstat.nr_dirty_background_threshold 9365694 +1.7% 9521870 proc-vmstat.nr_dirty_threshold 1654099 ± 2% -53.4% 771092 proc-vmstat.nr_file_pages 47057782 +1.7% 47838859 proc-vmstat.nr_free_pages 383856 ± 7% -54.6% 174141 proc-vmstat.nr_inactive_anon 132404 ± 12% -92.9% 9427 proc-vmstat.nr_mapped 4649 -1.8% 4567 proc-vmstat.nr_page_table_pages 885917 ± 4% -99.7% 2924 ± 3% proc-vmstat.nr_shmem 26984 -12.2% 23703 proc-vmstat.nr_slab_reclaimable 679317 ± 3% -99.9% 850.20 proc-vmstat.nr_zone_active_anon 383856 ± 7% -54.6% 174141 proc-vmstat.nr_zone_inactive_anon 201323 ± 11% -96.8% 6359 ± 91% proc-vmstat.numa_hint_faults 98999 ± 21% -99.5% 520.33 ± 5% proc-vmstat.numa_hint_faults_local 1907797 ± 2% -58.8% 785657 proc-vmstat.numa_hit 1805585 ± 2% -61.4% 697853 proc-vmstat.numa_local 107648 -18.6% 87641 proc-vmstat.numa_other 399248 ± 16% -88.1% 47657 ± 70% proc-vmstat.numa_pte_updates 958792 ± 5% -100.0% 204.50 ± 8% proc-vmstat.pgactivate 2040912 ± 2% -54.9% 920241 proc-vmstat.pgalloc_normal 1329369 ± 2% +15694.5% 2.1e+08 proc-vmstat.pgfault 1004379 -27.6% 727403 proc-vmstat.pgfree 49086 ± 20% -18.3% 40086 ± 4% proc-vmstat.pgreuse 0.03 -33.3% 0.02 ± 17% perf-stat.i.MPKI 2.522e+10 +69.5% 4.275e+10 ± 3% perf-stat.i.branch-instructions 1.55 -0.6 0.94 ± 18% perf-stat.i.branch-miss-rate% 18477740 -60.7% 7266076 ± 4% perf-stat.i.context-switches 2.20 -37.2% 1.38 ± 3% perf-stat.i.cpi 2.835e+11 -2.3% 2.771e+11 perf-stat.i.cpu-cycles 189.99 +3.7% 197.08 perf-stat.i.cpu-migrations 154844 ± 2% -35.7% 99550 ± 14% perf-stat.i.cycles-between-cache-misses 1.292e+11 +59.0% 2.054e+11 ± 3% perf-stat.i.instructions 0.46 +62.4% 0.74 ± 3% perf-stat.i.ipc 177.68 -53.2% 83.14 ± 3% perf-stat.i.metric.K/sec 4186 ± 2% +16391.7% 690372 perf-stat.i.minor-faults 4186 ± 2% +16391.2% 690372 perf-stat.i.page-faults 0.03 ± 2% -42.7% 0.01 ± 19% perf-stat.overall.MPKI 1.54 -0.6 0.91 ± 19% perf-stat.overall.branch-miss-rate% 2.19 -38.5% 1.35 ± 3% perf-stat.overall.cpi 0.46 +62.7% 0.74 ± 3% perf-stat.overall.ipc 2105030 +318.8% 8815433 perf-stat.overall.path-length 2.513e+10 +69.6% 4.262e+10 ± 3% perf-stat.ps.branch-instructions 18416035 -60.7% 7236360 ± 4% perf-stat.ps.context-switches 2.825e+11 -2.3% 2.761e+11 perf-stat.ps.cpu-cycles 189.25 +3.8% 196.41 perf-stat.ps.cpu-migrations 1.288e+11 +59.0% 2.048e+11 ± 3% perf-stat.ps.instructions 4169 ± 2% +16406.3% 688277 perf-stat.ps.minor-faults 4169 ± 2% +16405.8% 688277 perf-stat.ps.page-faults 3.904e+13 +59.3% 6.221e+13 ± 3% perf-stat.total.instructions 53.50 -53.5 0.00 perf-profile.calltrace.cycles-pp.read 45.69 -45.7 0.00 perf-profile.calltrace.cycles-pp.write 35.13 -35.1 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read 31.89 -31.9 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 29.98 -30.0 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write 26.86 -26.9 0.00 perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 26.74 -26.7 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 25.88 -25.9 0.00 perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 23.71 -23.7 0.00 perf-profile.calltrace.cycles-pp.pipe_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe 21.52 -21.5 0.00 perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 20.61 -20.6 0.00 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 18.62 ± 2% -18.6 0.00 perf-profile.calltrace.cycles-pp.pipe_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe 16.92 -16.9 0.00 perf-profile.calltrace.cycles-pp.schedule.pipe_read.vfs_read.ksys_read.do_syscall_64 16.55 -16.6 0.00 perf-profile.calltrace.cycles-pp.__schedule.schedule.pipe_read.vfs_read.ksys_read 14.74 ± 2% -14.7 0.00 perf-profile.calltrace.cycles-pp.__wake_up_sync_key.pipe_write.vfs_write.ksys_write.do_syscall_64 14.31 ± 2% -14.3 0.00 perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_sync_key.pipe_write.vfs_write.ksys_write 14.10 ± 2% -14.1 0.00 perf-profile.calltrace.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write.vfs_write 13.74 ± 2% -13.7 0.00 perf-profile.calltrace.cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write 9.02 -9.0 0.00 perf-profile.calltrace.cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key 7.16 -7.2 0.00 perf-profile.calltrace.cycles-pp.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common 6.93 -6.9 0.00 perf-profile.calltrace.cycles-pp.dequeue_task_fair.__schedule.schedule.pipe_read.vfs_read 6.62 -6.6 0.00 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.read 6.14 ± 2% -6.1 0.00 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.write 65.61 -65.6 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 58.74 -58.7 0.00 perf-profile.children.cycles-pp.do_syscall_64 53.92 -53.9 0.00 perf-profile.children.cycles-pp.read 46.12 -46.1 0.00 perf-profile.children.cycles-pp.write 26.87 -26.9 0.00 perf-profile.children.cycles-pp.ksys_read 25.90 -25.9 0.00 perf-profile.children.cycles-pp.vfs_read 23.87 -23.9 0.00 perf-profile.children.cycles-pp.pipe_read 21.55 -21.5 0.00 perf-profile.children.cycles-pp.ksys_write 20.65 -20.7 0.00 perf-profile.children.cycles-pp.vfs_write 19.74 -19.7 0.00 perf-profile.children.cycles-pp.schedule 19.36 -19.4 0.00 perf-profile.children.cycles-pp.__schedule 18.67 ± 2% -18.7 0.00 perf-profile.children.cycles-pp.pipe_write 14.75 ± 2% -14.8 0.00 perf-profile.children.cycles-pp.__wake_up_sync_key 14.32 ± 2% -14.3 0.00 perf-profile.children.cycles-pp.__wake_up_common 14.10 ± 2% -14.1 0.00 perf-profile.children.cycles-pp.autoremove_wake_function 13.79 ± 2% -13.8 0.00 perf-profile.children.cycles-pp.try_to_wake_up 12.86 -12.9 0.00 perf-profile.children.cycles-pp.syscall_return_via_sysret 10.00 -10.0 0.00 perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack 9.04 -9.0 0.00 perf-profile.children.cycles-pp.ttwu_do_activate 8.44 -8.4 0.00 perf-profile.children.cycles-pp.syscall_exit_to_user_mode 7.17 -7.2 0.00 perf-profile.children.cycles-pp.enqueue_task_fair 6.95 -6.9 0.00 perf-profile.children.cycles-pp.dequeue_task_fair 6.09 -6.1 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64 12.54 -12.5 0.00 perf-profile.self.cycles-pp.syscall_return_via_sysret 9.91 -9.9 0.00 perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack 6.97 -7.0 0.00 perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe 5.47 -5.5 0.00 perf-profile.self.cycles-pp.entry_SYSCALL_64 *************************************************************************************************** lkp-icl-2sp9: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory ========================================================================================= compiler/cpufreq_governor/disk/filesize/fs2/fs/iterations/kconfig/nr_files_per_directory/nr_threads/rootfs/sync_method/tbox_group/test_size/testcase: gcc-12/performance/1BRD_32G/4K/nfsv4/ext4/1x/x86_64-rhel-8.3/1fpd/1t/debian-12-x86_64-20240206.cgz/fsyncBeforeClose/lkp-icl-2sp9/4G/fsmark commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 1.519e+10 +34.7% 2.046e+10 ± 3% cpuidle..time 97.75 -4.6% 93.24 iostat.cpu.idle 1.84 +220.2% 5.88 iostat.cpu.system 301.00 +33.4% 401.55 ± 3% uptime.boot 18660 +28.1% 23909 ± 3% uptime.idle 929516 ±111% +336.5% 4057213 ± 47% numa-numastat.node0.local_node 974371 ±104% +320.7% 4098811 ± 47% numa-numastat.node0.numa_hit 17513401 ± 5% -17.6% 14426818 ± 13% numa-numastat.node1.local_node 17535676 ± 5% -17.6% 14440697 ± 13% numa-numastat.node1.numa_hit 403587 -29.1% 286045 ± 3% vmstat.io.bo 2.25 +128.2% 5.14 vmstat.procs.r 324930 -27.4% 236047 ± 2% vmstat.system.cs 176422 +45.4% 256569 vmstat.system.in 227.83 ± 11% -100.0% 0.00 perf-c2c.DRAM.local 427.33 ± 83% -100.0% 0.00 perf-c2c.DRAM.remote 5358 ± 4% -100.0% 0.00 perf-c2c.HITM.local 236.50 ± 98% -100.0% 0.00 perf-c2c.HITM.remote 5595 ± 2% -100.0% 0.00 perf-c2c.HITM.total 76998903 +33.3% 1.027e+08 ± 3% fsmark.app_overhead 4243 -29.8% 2978 ± 3% fsmark.files_per_sec 239.76 +42.0% 340.44 ± 3% fsmark.time.elapsed_time 239.76 +42.0% 340.44 ± 3% fsmark.time.elapsed_time.max 31.30 +51.6% 47.44 ± 3% fsmark.time.system_time 5769288 -1.7% 5670289 fsmark.time.voluntary_context_switches 0.15 +0.0 0.20 ± 5% mpstat.cpu.all.irq% 0.19 +0.1 0.26 mpstat.cpu.all.soft% 1.50 +3.9 5.43 mpstat.cpu.all.sys% 0.09 ± 2% +0.5 0.58 mpstat.cpu.all.usr% 1.17 ± 31% +22328.6% 261.67 ± 2% mpstat.max_utilization.seconds 5.18 ± 3% +66.7% 8.63 mpstat.max_utilization_pct 379352 ± 59% +87.1% 709913 ± 32% numa-meminfo.node0.AnonPages.max 3104 ± 9% +28.1% 3975 ± 12% numa-meminfo.node0.PageTables 111563 ± 19% +62.9% 181723 ± 31% numa-meminfo.node0.SUnreclaim 517329 ± 2% -99.6% 2323 ± 44% numa-meminfo.node1.Active(anon) 277501 ± 6% -41.6% 162034 ± 18% numa-meminfo.node1.Dirty 526882 ± 2% -99.5% 2824 ± 42% numa-meminfo.node1.Shmem 643.83 ± 6% +39.5% 898.01 ± 20% numa-meminfo.node1.Writeback 3076105 -13.2% 2669099 meminfo.Active 524607 -98.8% 6139 ± 2% meminfo.Active(anon) 1246297 +11.7% 1391614 meminfo.AnonHugePages 3303829 -14.1% 2836865 meminfo.Committed_AS 284869 -28.9% 202566 ± 3% meminfo.Dirty 7768 +10.6% 8592 meminfo.PageTables 543894 -96.9% 16697 meminfo.Shmem 645.23 ± 5% +80.9% 1167 ± 18% meminfo.Writeback 87885 ±195% +615.6% 628897 ± 53% numa-vmstat.node0.nr_dirtied 1993 ±222% +420.2% 10370 ± 55% numa-vmstat.node0.nr_dirty 781.52 ± 8% +27.2% 994.01 ± 12% numa-vmstat.node0.nr_page_table_pages 27893 ± 19% +62.9% 45435 ± 31% numa-vmstat.node0.nr_slab_unreclaimable 87721 ±195% +600.0% 614086 ± 53% numa-vmstat.node0.nr_written 1998 ±222% +422.0% 10431 ± 55% numa-vmstat.node0.nr_zone_write_pending 973993 ±104% +320.8% 4098274 ± 47% numa-vmstat.node0.numa_hit 929139 ±111% +336.6% 4056676 ± 47% numa-vmstat.node0.numa_local 129283 ± 2% -99.6% 581.30 ± 44% numa-vmstat.node1.nr_active_anon 69155 ± 6% -41.7% 40315 ± 18% numa-vmstat.node1.nr_dirty 131705 ± 2% -99.5% 705.99 ± 42% numa-vmstat.node1.nr_shmem 158.63 ± 4% +35.6% 215.16 ± 15% numa-vmstat.node1.nr_writeback 129283 ± 2% -99.6% 581.30 ± 44% numa-vmstat.node1.nr_zone_active_anon 69315 ± 6% -41.5% 40531 ± 18% numa-vmstat.node1.nr_zone_write_pending 17534283 ± 5% -17.6% 14439527 ± 13% numa-vmstat.node1.numa_hit 17512007 ± 5% -17.6% 14425648 ± 13% numa-vmstat.node1.numa_local 1.74 ± 9% -65.6% 0.60 ± 15% perf-stat.i.MPKI 9.615e+08 +285.6% 3.707e+09 perf-stat.i.branch-instructions 0.69 -0.5 0.19 perf-stat.i.branch-miss-rate% 7798406 -11.5% 6900769 perf-stat.i.branch-misses 329459 -27.7% 238084 ± 2% perf-stat.i.context-switches 1.35 -30.1% 0.94 perf-stat.i.cpi 6.387e+09 +159.7% 1.659e+10 perf-stat.i.cpu-cycles 853.55 ± 3% +132.1% 1980 ± 17% perf-stat.i.cycles-between-cache-misses 4.847e+09 +266.3% 1.776e+10 perf-stat.i.instructions 0.75 +42.0% 1.07 perf-stat.i.ipc 5.14 +2508.3% 134.15 perf-stat.i.metric.K/sec 2541 +1.6e+05% 4173925 perf-stat.i.minor-faults 2541 +1.6e+05% 4173925 perf-stat.i.page-faults 673.07 ± 45% +148.1% 1669 ± 16% perf-stat.overall.cycles-between-cache-misses 0.63 ± 44% +69.6% 1.07 perf-stat.overall.ipc 7.964e+08 ± 44% +364.2% 3.697e+09 perf-stat.ps.branch-instructions 5.299e+09 ± 44% +212.2% 1.654e+10 perf-stat.ps.cpu-cycles 4.015e+09 ± 44% +341.1% 1.771e+10 perf-stat.ps.instructions 2107 ± 44% +2e+05% 4162878 perf-stat.ps.minor-faults 2107 ± 44% +2e+05% 4162878 perf-stat.ps.page-faults 9.689e+11 ± 44% +524.4% 6.05e+12 ± 3% perf-stat.total.instructions 131138 -98.8% 1536 ± 2% proc-vmstat.nr_active_anon 637951 +4.3% 665408 proc-vmstat.nr_active_file 412901 +5.0% 433490 proc-vmstat.nr_anon_pages 607.19 +11.8% 679.13 proc-vmstat.nr_anon_transparent_hugepages 71193 -28.7% 50743 ± 3% proc-vmstat.nr_dirty 415380 +4.4% 433762 proc-vmstat.nr_inactive_anon 1003928 +5.2% 1056379 proc-vmstat.nr_inactive_file 9194 -5.4% 8699 proc-vmstat.nr_mapped 1941 +10.8% 2151 proc-vmstat.nr_page_table_pages 135981 -96.9% 4176 proc-vmstat.nr_shmem 761716 +5.0% 799556 proc-vmstat.nr_slab_reclaimable 135618 +5.3% 142780 proc-vmstat.nr_slab_unreclaimable 165.40 ± 2% +72.0% 284.42 ± 19% proc-vmstat.nr_writeback 3050431 +2.5% 3128215 proc-vmstat.nr_written 131138 -98.8% 1536 ± 2% proc-vmstat.nr_zone_active_anon 637951 +4.3% 665408 proc-vmstat.nr_zone_active_file 415380 +4.4% 433762 proc-vmstat.nr_zone_inactive_anon 1003928 +5.2% 1056379 proc-vmstat.nr_zone_inactive_file 71359 -28.5% 51028 ± 3% proc-vmstat.nr_zone_write_pending 1065 ± 10% -19.5% 858.50 ± 5% proc-vmstat.numa_hint_faults 945.50 ± 10% -50.9% 464.50 ± 6% proc-vmstat.numa_hint_faults_local 668.33 ± 25% +47.3% 984.67 ± 2% proc-vmstat.numa_huge_pte_updates 66342 -18.5% 54068 proc-vmstat.numa_other 367076 ± 23% +38.1% 507046 ± 2% proc-vmstat.numa_pte_updates 150353 ± 2% -98.4% 2420 proc-vmstat.pgactivate 671082 +2.1e+05% 1.427e+09 ± 4% proc-vmstat.pgfault 29181 +34.0% 39102 ± 3% proc-vmstat.pgreuse 7675 ± 9% +401.5% 38488 ± 8% sched_debug.cfs_rq:/.avg_vruntime.avg 40528 ± 16% +958.7% 429065 ± 19% sched_debug.cfs_rq:/.avg_vruntime.max 705.29 ± 28% +155.6% 1802 ± 21% sched_debug.cfs_rq:/.avg_vruntime.min 7669 ± 14% +883.8% 75453 ± 12% sched_debug.cfs_rq:/.avg_vruntime.stddev 0.30 ± 5% +13.7% 0.34 ± 5% sched_debug.cfs_rq:/.h_nr_running.stddev 951.02 ± 14% -38.0% 590.09 ± 34% sched_debug.cfs_rq:/.load_avg.max 172.57 ± 9% -28.4% 123.61 ± 18% sched_debug.cfs_rq:/.load_avg.stddev 7675 ± 9% +401.5% 38488 ± 8% sched_debug.cfs_rq:/.min_vruntime.avg 40528 ± 16% +958.7% 429065 ± 19% sched_debug.cfs_rq:/.min_vruntime.max 705.29 ± 28% +155.6% 1802 ± 21% sched_debug.cfs_rq:/.min_vruntime.min 7669 ± 14% +883.8% 75453 ± 12% sched_debug.cfs_rq:/.min_vruntime.stddev 0.30 ± 5% +13.9% 0.34 ± 5% sched_debug.cfs_rq:/.nr_running.stddev 716.12 ± 10% +51.6% 1085 ± 3% sched_debug.cfs_rq:/.runnable_avg.max 147.84 ± 6% +64.0% 242.44 sched_debug.cfs_rq:/.runnable_avg.stddev 712.46 ± 10% +52.1% 1083 ± 3% sched_debug.cfs_rq:/.util_avg.max 147.41 ± 6% +64.2% 242.12 sched_debug.cfs_rq:/.util_avg.stddev 16.26 ± 17% +216.4% 51.45 ± 7% sched_debug.cfs_rq:/.util_est.avg 383.67 ± 19% +162.8% 1008 ± 3% sched_debug.cfs_rq:/.util_est.max 63.16 ± 17% +222.5% 203.68 ± 2% sched_debug.cfs_rq:/.util_est.stddev 165310 ± 8% +24.3% 205470 ± 5% sched_debug.cpu.clock.avg 165314 ± 8% +24.3% 205474 ± 5% sched_debug.cpu.clock.max 165279 ± 8% +24.3% 205454 ± 5% sched_debug.cpu.clock.min 5.18 ± 28% -36.4% 3.30 ± 18% sched_debug.cpu.clock.stddev 164698 ± 8% +24.2% 204576 ± 5% sched_debug.cpu.clock_task.avg 165287 ± 8% +24.3% 205370 ± 5% sched_debug.cpu.clock_task.max 155952 ± 9% +25.1% 195174 ± 5% sched_debug.cpu.clock_task.min 1185 ± 2% +8.9% 1290 ± 5% sched_debug.cpu.clock_task.stddev 274.22 ± 13% +32.4% 362.97 ± 7% sched_debug.cpu.curr->pid.avg 5453 ± 6% +18.4% 6455 ± 4% sched_debug.cpu.curr->pid.max 940.43 ± 9% +18.8% 1117 ± 2% sched_debug.cpu.curr->pid.stddev 0.11 ± 10% +25.0% 0.13 ± 9% sched_debug.cpu.nr_running.avg 0.30 ± 6% +13.5% 0.34 ± 4% sched_debug.cpu.nr_running.stddev 24.81 ± 14% +28.4% 31.85 ± 8% sched_debug.cpu.nr_uninterruptible.stddev 165310 ± 8% +24.3% 205469 ± 5% sched_debug.cpu_clk 164147 ± 8% +24.5% 204306 ± 5% sched_debug.ktime 165888 ± 8% +24.2% 206051 ± 5% sched_debug.sched_clk 50.27 -50.3 0.00 perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm 50.27 -50.3 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm 50.27 -50.3 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork_asm 43.03 ± 6% -43.0 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 37.89 ± 2% -37.9 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 36.23 ± 2% -36.2 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 36.20 ± 2% -36.2 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 36.14 ± 2% -36.1 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 33.61 ± 2% -33.6 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 33.08 ± 2% -33.1 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 31.89 ± 2% -31.9 0.00 perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 31.81 ± 2% -31.8 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 27.02 -27.0 0.00 perf-profile.calltrace.cycles-pp.nfsd.kthread.ret_from_fork.ret_from_fork_asm 26.83 -26.8 0.00 perf-profile.calltrace.cycles-pp.svc_recv.nfsd.kthread.ret_from_fork.ret_from_fork_asm 24.50 -24.5 0.00 perf-profile.calltrace.cycles-pp.svc_handle_xprt.svc_recv.nfsd.kthread.ret_from_fork 16.59 ± 3% -16.6 0.00 perf-profile.calltrace.cycles-pp.svc_process.svc_handle_xprt.svc_recv.nfsd.kthread 16.57 ± 3% -16.6 0.00 perf-profile.calltrace.cycles-pp.svc_process_common.svc_process.svc_handle_xprt.svc_recv.nfsd 16.02 ± 2% -16.0 0.00 perf-profile.calltrace.cycles-pp.nfsd_dispatch.svc_process_common.svc_process.svc_handle_xprt.svc_recv 15.33 ± 2% -15.3 0.00 perf-profile.calltrace.cycles-pp.nfsd4_proc_compound.nfsd_dispatch.svc_process_common.svc_process.svc_handle_xprt 13.29 ± 5% -13.3 0.00 perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 12.88 ± 4% -12.9 0.00 perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 10.99 ± 8% -11.0 0.00 perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 9.54 ± 5% -9.5 0.00 perf-profile.calltrace.cycles-pp.kjournald2.kthread.ret_from_fork.ret_from_fork_asm 8.75 ± 5% -8.7 0.00 perf-profile.calltrace.cycles-pp.jbd2_journal_commit_transaction.kjournald2.kthread.ret_from_fork.ret_from_fork_asm 6.78 ± 4% -6.8 0.00 perf-profile.calltrace.cycles-pp.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state 6.10 ± 3% -6.1 0.00 perf-profile.calltrace.cycles-pp.ip6_xmit.inet6_csk_xmit.__tcp_transmit_skb.tcp_write_xmit.__tcp_push_pending_frames 5.97 ± 4% -6.0 0.00 perf-profile.calltrace.cycles-pp.nfsd4_create.nfsd4_proc_compound.nfsd_dispatch.svc_process_common.svc_process 5.93 ± 4% -5.9 0.00 perf-profile.calltrace.cycles-pp.ip6_finish_output.ip6_xmit.inet6_csk_xmit.__tcp_transmit_skb.tcp_write_xmit 5.88 ± 7% -5.9 0.00 perf-profile.calltrace.cycles-pp.rpc_async_schedule.process_one_work.worker_thread.kthread.ret_from_fork 5.87 ± 5% -5.9 0.00 perf-profile.calltrace.cycles-pp.nfsd_create.nfsd4_create.nfsd4_proc_compound.nfsd_dispatch.svc_process_common 5.86 ± 4% -5.9 0.00 perf-profile.calltrace.cycles-pp.ip6_finish_output2.ip6_finish_output.ip6_xmit.inet6_csk_xmit.__tcp_transmit_skb 5.85 ± 7% -5.9 0.00 perf-profile.calltrace.cycles-pp.__rpc_execute.rpc_async_schedule.process_one_work.worker_thread.kthread 5.73 ± 4% -5.7 0.00 perf-profile.calltrace.cycles-pp.__dev_queue_xmit.ip6_finish_output2.ip6_finish_output.ip6_xmit.inet6_csk_xmit 5.40 ± 15% -5.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 5.10 ± 8% -5.1 0.00 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.__dev_queue_xmit.ip6_finish_output2.ip6_finish_output.ip6_xmit 5.08 ± 8% -5.1 0.00 perf-profile.calltrace.cycles-pp.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip6_finish_output2.ip6_finish_output 5.04 ± 9% -5.0 0.00 perf-profile.calltrace.cycles-pp.handle_softirqs.do_softirq.__local_bh_enable_ip.__dev_queue_xmit.ip6_finish_output2 4.92 ± 6% -4.9 0.00 perf-profile.calltrace.cycles-pp.xs_tcp_send_request.xprt_request_transmit.xprt_transmit.call_transmit.__rpc_execute 50.28 -50.3 0.00 perf-profile.children.cycles-pp.ret_from_fork_asm 50.27 -50.3 0.00 perf-profile.children.cycles-pp.ret_from_fork 50.27 -50.3 0.00 perf-profile.children.cycles-pp.kthread 37.89 ± 2% -37.9 0.00 perf-profile.children.cycles-pp.common_startup_64 37.89 ± 2% -37.9 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 37.83 ± 2% -37.8 0.00 perf-profile.children.cycles-pp.do_idle 36.23 ± 2% -36.2 0.00 perf-profile.children.cycles-pp.start_secondary 35.18 ± 2% -35.2 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 33.25 ± 2% -33.2 0.00 perf-profile.children.cycles-pp.cpuidle_enter 33.11 ± 2% -33.1 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 31.90 ± 2% -31.9 0.00 perf-profile.children.cycles-pp.acpi_idle_enter 31.85 ± 2% -31.8 0.00 perf-profile.children.cycles-pp.acpi_safe_halt 27.02 -27.0 0.00 perf-profile.children.cycles-pp.nfsd 26.84 -26.8 0.00 perf-profile.children.cycles-pp.svc_recv 25.33 ± 4% -25.3 0.00 perf-profile.children.cycles-pp.asm_sysvec_call_function_single 24.51 -24.5 0.00 perf-profile.children.cycles-pp.svc_handle_xprt 16.59 ± 3% -16.6 0.00 perf-profile.children.cycles-pp.svc_process 16.57 ± 3% -16.6 0.00 perf-profile.children.cycles-pp.svc_process_common 16.03 ± 2% -16.0 0.00 perf-profile.children.cycles-pp.nfsd_dispatch 15.35 ± 2% -15.4 0.00 perf-profile.children.cycles-pp.nfsd4_proc_compound 13.31 ± 5% -13.3 0.00 perf-profile.children.cycles-pp.worker_thread 11.01 ± 8% -11.0 0.00 perf-profile.children.cycles-pp.process_one_work 9.63 ± 6% -9.6 0.00 perf-profile.children.cycles-pp.__rpc_execute 9.55 ± 3% -9.6 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 9.54 ± 5% -9.5 0.00 perf-profile.children.cycles-pp.kjournald2 9.54 ± 3% -9.5 0.00 perf-profile.children.cycles-pp.do_syscall_64 8.78 ± 5% -8.8 0.00 perf-profile.children.cycles-pp.jbd2_journal_commit_transaction 7.36 ± 3% -7.4 0.00 perf-profile.children.cycles-pp.handle_softirqs 7.32 ± 3% -7.3 0.00 perf-profile.children.cycles-pp.__tcp_push_pending_frames 7.29 ± 3% -7.3 0.00 perf-profile.children.cycles-pp.tcp_write_xmit 6.86 ± 2% -6.9 0.00 perf-profile.children.cycles-pp.vfs_mkdir 6.84 ± 3% -6.8 0.00 perf-profile.children.cycles-pp.__tcp_transmit_skb 6.82 ± 4% -6.8 0.00 perf-profile.children.cycles-pp.sysvec_call_function_single 6.30 ± 3% -6.3 0.00 perf-profile.children.cycles-pp.inet6_csk_xmit 6.10 ± 3% -6.1 0.00 perf-profile.children.cycles-pp.ip6_xmit 5.97 ± 4% -6.0 0.00 perf-profile.children.cycles-pp.nfsd4_create 5.93 ± 4% -5.9 0.00 perf-profile.children.cycles-pp.ip6_finish_output 5.88 ± 7% -5.9 0.00 perf-profile.children.cycles-pp.rpc_async_schedule 5.88 ± 5% -5.9 0.00 perf-profile.children.cycles-pp.nfsd_create 5.86 ± 4% -5.9 0.00 perf-profile.children.cycles-pp.ip6_finish_output2 5.74 ± 4% -5.7 0.00 perf-profile.children.cycles-pp.__dev_queue_xmit 5.53 ± 4% -5.5 0.00 perf-profile.children.cycles-pp.__local_bh_enable_ip 5.51 ± 3% -5.5 0.00 perf-profile.children.cycles-pp.sock_sendmsg 5.45 ± 4% -5.4 0.00 perf-profile.children.cycles-pp.do_softirq 5.38 ± 3% -5.4 0.00 perf-profile.children.cycles-pp.tcp_sendmsg 5.31 ± 4% -5.3 0.00 perf-profile.children.cycles-pp.__schedule 5.18 ± 6% -5.2 0.00 perf-profile.children.cycles-pp.call_transmit 5.14 ± 6% -5.1 0.00 perf-profile.children.cycles-pp.xprt_transmit 5.10 ± 6% -5.1 0.00 perf-profile.children.cycles-pp.xprt_request_transmit 5.08 ± 2% -5.1 0.00 perf-profile.children.cycles-pp.tcp_sendmsg_locked 5.02 ± 5% -5.0 0.00 perf-profile.children.cycles-pp.net_rx_action 4.93 ± 5% -4.9 0.00 perf-profile.children.cycles-pp.xs_tcp_send_request 21.35 ± 6% -21.4 0.00 perf-profile.self.cycles-pp.acpi_safe_halt 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. -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH] perf: Remove setting of page->index and ->mapping 2024-08-16 6:35 ` kernel test robot @ 2024-08-16 17:51 ` Matthew Wilcox 2024-08-20 5:01 ` Oliver Sang 0 siblings, 1 reply; 5+ messages in thread From: Matthew Wilcox @ 2024-08-16 17:51 UTC (permalink / raw) To: kernel test robot Cc: oe-lkp, lkp, linux-perf-users, linux-kernel, ying.huang, feng.tang, fengwei.yin, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim On Fri, Aug 16, 2024 at 02:35:39PM +0800, kernel test robot wrote: > from commit message, it seems just some useless setting have been removed. > but quite some performance regression bisect by our bot finally point to this > change (already removed some not so significant changes to avoid this report > mail is too long). > > we don't have enough knowledge to explain these, so just report FYI. > if you have any suggestion for our tests or need us run more tests, please > just let us know. thanks a lot! Can you try this replacement patch? diff --git a/kernel/events/core.c b/kernel/events/core.c index c973e3c11e03..bc12f4fb7fa9 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6221,10 +6221,9 @@ static vm_fault_t perf_mmap_fault(struct vm_fault *vmf) goto unlock; get_page(vmf->page); - vmf->page->mapping = vmf->vma->vm_file->f_mapping; - vmf->page->index = vmf->pgoff; + lock_page(vmf->page); - ret = 0; + ret = VM_FAULT_LOCKED; unlock: rcu_read_unlock(); ^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH] perf: Remove setting of page->index and ->mapping 2024-08-16 17:51 ` Matthew Wilcox @ 2024-08-20 5:01 ` Oliver Sang 0 siblings, 0 replies; 5+ messages in thread From: Oliver Sang @ 2024-08-20 5:01 UTC (permalink / raw) To: Matthew Wilcox Cc: oe-lkp, lkp, linux-perf-users, linux-kernel, ying.huang, feng.tang, fengwei.yin, Peter Zijlstra, Ingo Molnar, Arnaldo Carvalho de Melo, Namhyung Kim, oliver.sang hi, Matthew Wilcox, On Fri, Aug 16, 2024 at 06:51:23PM +0100, Matthew Wilcox wrote: > On Fri, Aug 16, 2024 at 02:35:39PM +0800, kernel test robot wrote: > > from commit message, it seems just some useless setting have been removed. > > but quite some performance regression bisect by our bot finally point to this > > change (already removed some not so significant changes to avoid this report > > mail is too long). > > > > we don't have enough knowledge to explain these, so just report FYI. > > if you have any suggestion for our tests or need us run more tests, please > > just let us know. thanks a lot! > > Can you try this replacement patch? > > diff --git a/kernel/events/core.c b/kernel/events/core.c > index c973e3c11e03..bc12f4fb7fa9 100644 > --- a/kernel/events/core.c > +++ b/kernel/events/core.c > @@ -6221,10 +6221,9 @@ static vm_fault_t perf_mmap_fault(struct vm_fault *vmf) > goto unlock; > > get_page(vmf->page); > - vmf->page->mapping = vmf->vma->vm_file->f_mapping; > - vmf->page->index = vmf->pgoff; > + lock_page(vmf->page); > > - ret = 0; > + ret = VM_FAULT_LOCKED; > unlock: > rcu_read_unlock(); > > still similar regression. I applied your replacement patch as 0c365693d5482c855ba9acd45a9cffa4e12fc4c1 [1] still upon cb1898f58e ("perf annotate-data: Support --skip-empty option") as [2] I picked 2 regression to check. for first one, before replacement: ========================================================================================= compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/1HDD/btrfs/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp8/dirmany/stress-ng/60s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 703.32 -28.3% 504.06 ± 3% stress-ng.dirmany.files_removed_per_sec after replacement (full comparison is as [3]): ========================================================================================= compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/1HDD/btrfs/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp8/dirmany/stress-ng/60s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") 0c365693d5 <-- replacement patch cb1898f58e0f175d 0c365693d5482c855ba9acd45a9 ---------------- --------------------------- %stddev %change %stddev \ | \ 703.32 -28.9% 499.74 stress-ng.dirmany.files_removed_per_sec for second one, before replacement: ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-12/performance/x86_64-rhel-8.3/process/100%/debian-12-x86_64-20240206.cgz/lkp-skl-fpga01/context_switch1/will-it-scale commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") a9ac22fb7a ("perf: Remove setting of page->index and ->mapping") cb1898f58e0f175d a9ac22fb7ac3486a5bd89c25a2f ---------------- --------------------------- %stddev %change %stddev \ | \ 178341 -61.9% 67890 ± 4% will-it-scale.per_process_ops after replacement (full comparison is as [4]): ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-12/performance/x86_64-rhel-8.3/process/100%/debian-12-x86_64-20240206.cgz/lkp-skl-fpga01/context_switch1/will-it-scale commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") 0c365693d5 <-- replacement patch cb1898f58e0f175d 0c365693d5482c855ba9acd45a9 ---------------- --------------------------- %stddev %change %stddev \ | \ 178341 -61.2% 69248 ± 5% will-it-scale.per_process_ops [1] commit 0c365693d5482c855ba9acd45a9cffa4e12fc4c1 Author: 0day robot <lkp@intel.com> Date: Sun Aug 18 16:18:03 2024 +0800 replacement for a9ac22fb7a from Matthew Wilcox <willy@infradead.org> diff --git a/kernel/events/core.c b/kernel/events/core.c index aa3450bdc2276..b102c404526e9 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -6221,10 +6221,9 @@ static vm_fault_t perf_mmap_fault(struct vm_fault *vmf) goto unlock; get_page(vmf->page); - vmf->page->mapping = vmf->vma->vm_file->f_mapping; - vmf->page->index = vmf->pgoff; + lock_page(vmf->page); - ret = 0; + ret = VM_FAULT_LOCKED; unlock: rcu_read_unlock(); [2] $ git log --oneline --graph 0c365693d5482 * 0c365693d5482 replacement for a9ac22fb7a from Matthew Wilcox <willy@infradead.org> * cb1898f58e0f1 perf annotate-data: Support --skip-empty option ... [3] ========================================================================================= compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: gcc-12/performance/1HDD/btrfs/x86_64-rhel-8.3/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp8/dirmany/stress-ng/60s commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") 0c365693d5 <-- replacement patch cb1898f58e0f175d 0c365693d5482c855ba9acd45a9 ---------------- --------------------------- %stddev %change %stddev \ | \ 8914265 ± 4% -14.0% 7667270 ± 3% cpuidle..usage 1.03 ± 3% -11.3% 0.91 ± 4% iostat.cpu.iowait 530992 ± 2% -10.8% 473754 ± 5% numa-numastat.node0.numa_hit 37575 ± 10% -87.6% 4642 ± 11% meminfo.Active(anon) 121149 ± 3% -28.8% 86296 meminfo.Shmem 1.04 ± 3% -0.1 0.92 ± 4% mpstat.cpu.all.iowait% 0.71 ± 4% +0.2 0.90 ± 6% mpstat.cpu.all.usr% 422.17 ± 4% -37.7% 263.12 ± 7% time.percent_of_cpu_this_job_got 267.72 ± 5% -31.5% 183.32 ± 4% time.system_time 6952736 ± 4% -17.3% 5749237 ± 4% time.voluntary_context_switches 0.68 ± 3% -9.7% 0.62 ± 4% vmstat.procs.b 238142 ± 6% -24.9% 178933 ± 3% vmstat.system.cs 157186 ± 7% +51.5% 238200 ± 3% vmstat.system.in 16677 ± 10% -88.3% 1957 ± 43% numa-meminfo.node0.Active(anon) 5262 ± 33% +59.2% 8375 ± 7% numa-meminfo.node0.PageTables 20916 ± 15% -87.2% 2687 ± 39% numa-meminfo.node1.Active(anon) 6573 ± 26% -43.9% 3690 ± 15% numa-meminfo.node1.PageTables 69830 ± 47% -95.7% 3023 ± 41% numa-meminfo.node1.Shmem 1586 ± 12% -100.0% 0.00 perf-c2c.DRAM.local 13101 ± 15% -100.0% 0.00 perf-c2c.DRAM.remote 12067 ± 15% -100.0% 0.00 perf-c2c.HITM.local 8050 ± 16% -100.0% 0.00 perf-c2c.HITM.remote 20118 ± 15% -100.0% 0.00 perf-c2c.HITM.total 67.43 ± 3% -11.9% 59.40 ± 2% stress-ng.dirmany.%_of_time_creating_files 32.57 ± 6% +24.7% 40.60 ± 4% stress-ng.dirmany.%_of_time_removing_file 703.32 -28.9% 499.74 stress-ng.dirmany.files_removed_per_sec 14642 ± 5% -11.3% 12984 ± 5% stress-ng.dirmany.ops_per_sec 422.17 ± 4% -37.7% 263.12 ± 7% stress-ng.time.percent_of_cpu_this_job_got 267.72 ± 5% -31.5% 183.32 ± 4% stress-ng.time.system_time 6952736 ± 4% -17.3% 5749237 ± 4% stress-ng.time.voluntary_context_switches 4177 ± 10% -88.3% 489.29 ± 43% numa-vmstat.node0.nr_active_anon 1313 ± 33% +59.6% 2096 ± 7% numa-vmstat.node0.nr_page_table_pages 4177 ± 10% -88.3% 489.29 ± 43% numa-vmstat.node0.nr_zone_active_anon 530566 ± 2% -10.8% 473184 ± 5% numa-vmstat.node0.numa_hit 5236 ± 15% -87.2% 672.21 ± 39% numa-vmstat.node1.nr_active_anon 1640 ± 26% -43.6% 924.79 ± 15% numa-vmstat.node1.nr_page_table_pages 17423 ± 47% -95.7% 755.96 ± 41% numa-vmstat.node1.nr_shmem 5236 ± 15% -87.2% 672.21 ± 39% numa-vmstat.node1.nr_zone_active_anon 9403 ± 10% -87.7% 1161 ± 10% proc-vmstat.nr_active_anon 27583 -2.1% 27001 proc-vmstat.nr_mapped 30307 ± 3% -28.7% 21607 proc-vmstat.nr_shmem 340687 -2.8% 331094 proc-vmstat.nr_slab_reclaimable 9403 ± 10% -87.7% 1161 ± 10% proc-vmstat.nr_zone_active_anon 278.50 ± 16% +269.3% 1028 ±128% proc-vmstat.numa_hint_faults_local 1037475 -5.2% 983767 ± 2% proc-vmstat.numa_hit 971211 -4.3% 929901 ± 2% proc-vmstat.numa_local 66223 -18.7% 53867 proc-vmstat.numa_other 1392679 -4.5% 1330429 proc-vmstat.pgalloc_normal 263934 ± 3% +91013.5% 2.405e+08 ± 6% proc-vmstat.pgfault 840859 ± 2% -21.1% 663807 ± 9% proc-vmstat.pgfree 76822 ± 10% +69.3% 130087 ± 16% sched_debug.cfs_rq:/.avg_vruntime.max 8327 ± 16% +127.9% 18977 ± 12% sched_debug.cfs_rq:/.avg_vruntime.stddev 2.25 ± 27% -100.0% 0.00 sched_debug.cfs_rq:/.load_avg.min 76822 ± 10% +69.3% 130087 ± 16% sched_debug.cfs_rq:/.min_vruntime.max 8327 ± 16% +127.9% 18978 ± 12% sched_debug.cfs_rq:/.min_vruntime.stddev 217.83 ± 45% +67.1% 364.00 ± 31% sched_debug.cfs_rq:/.removed.runnable_avg.max 217.83 ± 45% +67.1% 364.00 ± 31% sched_debug.cfs_rq:/.removed.util_avg.max 1018 ± 9% +24.0% 1262 ± 9% sched_debug.cfs_rq:/.runnable_avg.max 19.67 ± 57% -100.0% 0.00 sched_debug.cfs_rq:/.runnable_avg.min 217.81 ± 3% +31.7% 286.94 ± 7% sched_debug.cfs_rq:/.runnable_avg.stddev 1006 ± 7% +23.3% 1240 ± 8% sched_debug.cfs_rq:/.util_avg.max 19.75 ± 57% -100.0% 0.00 sched_debug.cfs_rq:/.util_avg.min 216.49 ± 3% +31.9% 285.45 ± 7% sched_debug.cfs_rq:/.util_avg.stddev 28.42 ± 11% +71.2% 48.67 ± 10% sched_debug.cfs_rq:/.util_est.avg 447.25 ± 12% +72.5% 771.44 ± 2% sched_debug.cfs_rq:/.util_est.max 90.88 ± 9% +90.2% 172.82 ± 4% sched_debug.cfs_rq:/.util_est.stddev 415102 ± 6% +23.8% 513771 ± 2% sched_debug.cpu.avg_idle.avg 38570 ± 20% +63.8% 63180 ± 11% sched_debug.cpu.avg_idle.min 154552 ± 6% +44.4% 223103 ± 3% sched_debug.cpu.avg_idle.stddev 101993 ± 13% -30.4% 70952 ± 11% sched_debug.cpu.nr_switches.avg 113676 ± 14% -24.1% 86266 ± 10% sched_debug.cpu.nr_switches.max 93486 ± 14% -83.6% 15356 ± 41% sched_debug.cpu.nr_switches.min 3929 ± 34% +261.7% 14212 ± 19% sched_debug.cpu.nr_switches.stddev 2.27 ± 2% -64.8% 0.80 ± 4% perf-stat.i.MPKI 1.604e+09 ± 2% +131.1% 3.705e+09 ± 2% perf-stat.i.branch-instructions 2.33 ± 2% -1.6 0.74 ± 10% perf-stat.i.branch-miss-rate% 36868744 ± 3% -27.5% 26723115 ± 9% perf-stat.i.branch-misses 23122994 ± 4% -14.7% 19730202 ± 4% perf-stat.i.cache-misses 64086754 ± 3% -16.2% 53705975 ± 4% perf-stat.i.cache-references 244416 ± 6% -24.8% 183752 ± 3% perf-stat.i.context-switches 1.90 ± 3% -39.1% 1.16 perf-stat.i.cpi 2.112e+10 ± 2% +7.6% 2.272e+10 ± 3% perf-stat.i.cpu-cycles 10537 ± 6% -72.8% 2868 ± 48% perf-stat.i.cpu-migrations 1308 ± 11% +395.6% 6487 ± 11% perf-stat.i.cycles-between-cache-misses 8.205e+09 ± 2% +118.7% 1.795e+10 ± 2% perf-stat.i.instructions 0.88 ± 3% +12.0% 0.99 perf-stat.i.ipc 3.78 ± 6% +2738.0% 107.39 perf-stat.i.metric.K/sec 3035 ± 2% +1.1e+05% 3345880 perf-stat.i.minor-faults 3035 ± 2% +1.1e+05% 3345880 perf-stat.i.page-faults 2.82 ± 2% -60.9% 1.10 ± 3% perf-stat.overall.MPKI 2.30 ± 2% -1.6 0.72 ± 8% perf-stat.overall.branch-miss-rate% 2.58 ± 2% -50.8% 1.27 perf-stat.overall.cpi 915.45 ± 4% +25.9% 1152 ± 3% perf-stat.overall.cycles-between-cache-misses 0.39 ± 2% +103.2% 0.79 perf-stat.overall.ipc 1.579e+09 ± 2% +131.3% 3.652e+09 ± 2% perf-stat.ps.branch-instructions 36312783 ± 3% -27.5% 26326319 ± 9% perf-stat.ps.branch-misses 22753189 ± 4% -14.4% 19465477 ± 4% perf-stat.ps.cache-misses 63076730 ± 3% -16.0% 53010328 ± 4% perf-stat.ps.cache-references 240154 ± 6% -24.5% 181197 ± 3% perf-stat.ps.context-switches 2.08e+10 ± 2% +7.8% 2.242e+10 ± 3% perf-stat.ps.cpu-cycles 10351 ± 6% -72.6% 2838 ± 48% perf-stat.ps.cpu-migrations 8.078e+09 ± 2% +119.0% 1.769e+10 ± 2% perf-stat.ps.instructions 2990 ± 2% +1.1e+05% 3296800 perf-stat.ps.minor-faults 2990 ± 2% +1.1e+05% 3296800 perf-stat.ps.page-faults 5.224e+11 ± 2% +141.6% 1.262e+12 ± 3% perf-stat.total.instructions 24.93 ± 18% -24.9 0.00 perf-profile.calltrace.cycles-pp.common_startup_64 24.55 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction 24.55 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction 24.54 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range.btrfs_write_marked_extents 24.51 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.fsync 24.50 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.50 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.fsync 24.49 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.49 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_sync_file.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe.fsync 24.48 ± 56% -24.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync.do_syscall_64.entry_SYSCALL_64_after_hwframe 24.43 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync.do_syscall_64 24.43 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file.__x64_sys_fsync 24.41 ± 57% -24.4 0.00 perf-profile.calltrace.cycles-pp.__filemap_fdatawrite_range.btrfs_write_marked_extents.btrfs_write_and_wait_transaction.btrfs_commit_transaction.btrfs_sync_file 23.94 ± 56% -23.9 0.00 perf-profile.calltrace.cycles-pp.submit_eb_page.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range 23.50 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64 23.49 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64 23.48 ± 16% -23.5 0.00 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64 22.12 ± 17% -22.1 0.00 perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64 21.95 ± 19% -22.0 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry 21.61 ± 20% -21.6 0.00 perf-profile.calltrace.cycles-pp.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle 20.64 ± 17% -20.6 0.00 perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary 20.29 ± 57% -20.3 0.00 perf-profile.calltrace.cycles-pp.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc 20.29 ± 57% -20.3 0.00 perf-profile.calltrace.cycles-pp.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages.do_writepages 17.17 ± 57% -17.2 0.00 perf-profile.calltrace.cycles-pp.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page.btree_write_cache_pages 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm 14.62 ± 31% -14.6 0.00 perf-profile.calltrace.cycles-pp.ret_from_fork_asm 13.30 ± 27% -13.3 0.00 perf-profile.calltrace.cycles-pp.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call 12.54 ± 58% -12.5 0.00 perf-profile.calltrace.cycles-pp.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page 12.31 ± 58% -12.3 0.00 perf-profile.calltrace.cycles-pp.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio 11.98 ± 32% -12.0 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 10.42 ± 37% -10.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 8.48 ± 57% -8.5 0.00 perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 7.77 ± 62% -7.8 0.00 perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork.ret_from_fork_asm 7.46 ± 58% -7.5 0.00 perf-profile.calltrace.cycles-pp.check_leaf_item.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio.btrfs_submit_chunk 5.40 ± 69% -5.4 0.00 perf-profile.calltrace.cycles-pp.asm_sysvec_posted_msi_notification.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state.cpuidle_enter 4.81 ± 57% -4.8 0.00 perf-profile.calltrace.cycles-pp.check_dir_item.check_leaf_item.__btrfs_check_leaf.btrfs_check_leaf.btree_csum_one_bio 4.62 ± 36% -4.6 0.00 perf-profile.calltrace.cycles-pp.sysvec_call_function_single.asm_sysvec_call_function_single.acpi_safe_halt.acpi_idle_enter.cpuidle_enter_state 4.52 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio.submit_eb_page 4.51 ± 56% -4.5 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe 4.49 ± 56% -4.5 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe 4.46 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.crc32c_pcl_intel_update.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk.btrfs_submit_bio 4.46 ± 55% -4.5 0.00 perf-profile.calltrace.cycles-pp.crc_pcl.crc32c_pcl_intel_update.csum_tree_block.btree_csum_one_bio.btrfs_submit_chunk 55.50 ± 15% -55.5 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 55.46 ± 15% -55.5 0.00 perf-profile.children.cycles-pp.do_syscall_64 28.46 ± 53% -28.5 0.00 perf-profile.children.cycles-pp.btrfs_commit_transaction 24.93 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.common_startup_64 24.93 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.cpu_startup_entry 24.92 ± 18% -24.9 0.00 perf-profile.children.cycles-pp.do_idle 24.61 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btrfs_write_and_wait_transaction 24.61 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btrfs_write_marked_extents 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.__filemap_fdatawrite_range 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.btree_write_cache_pages 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.do_writepages 24.60 ± 55% -24.6 0.00 perf-profile.children.cycles-pp.filemap_fdatawrite_wbc 24.51 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.fsync 24.49 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.__x64_sys_fsync 24.49 ± 56% -24.5 0.00 perf-profile.children.cycles-pp.btrfs_sync_file 24.00 ± 55% -24.0 0.00 perf-profile.children.cycles-pp.submit_eb_page 23.53 ± 19% -23.5 0.00 perf-profile.children.cycles-pp.cpuidle_idle_call 23.50 ± 16% -23.5 0.00 perf-profile.children.cycles-pp.start_secondary 22.05 ± 19% -22.1 0.00 perf-profile.children.cycles-pp.cpuidle_enter 22.00 ± 19% -22.0 0.00 perf-profile.children.cycles-pp.cpuidle_enter_state 21.66 ± 20% -21.7 0.00 perf-profile.children.cycles-pp.acpi_idle_enter 21.63 ± 20% -21.6 0.00 perf-profile.children.cycles-pp.acpi_safe_halt 20.41 ± 55% -20.4 0.00 perf-profile.children.cycles-pp.btrfs_submit_bio 20.41 ± 55% -20.4 0.00 perf-profile.children.cycles-pp.btrfs_submit_chunk 17.35 ± 55% -17.4 0.00 perf-profile.children.cycles-pp.btree_csum_one_bio 14.74 ± 31% -14.7 0.00 perf-profile.children.cycles-pp.ret_from_fork_asm 14.71 ± 31% -14.7 0.00 perf-profile.children.cycles-pp.ret_from_fork 14.62 ± 31% -14.6 0.00 perf-profile.children.cycles-pp.kthread 12.66 ± 56% -12.7 0.00 perf-profile.children.cycles-pp.__btrfs_check_leaf 12.66 ± 56% -12.7 0.00 perf-profile.children.cycles-pp.btrfs_check_leaf 8.52 ± 29% -8.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_call_function_single 8.48 ± 57% -8.5 0.00 perf-profile.children.cycles-pp.worker_thread 7.78 ± 62% -7.8 0.00 perf-profile.children.cycles-pp.process_one_work 7.62 ± 56% -7.6 0.00 perf-profile.children.cycles-pp.check_leaf_item 7.50 ± 34% -7.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt 5.58 ± 53% -5.6 0.00 perf-profile.children.cycles-pp.crc32c_pcl_intel_update 5.20 ± 46% -5.2 0.00 perf-profile.children.cycles-pp.__irq_exit_rcu 5.11 ± 47% -5.1 0.00 perf-profile.children.cycles-pp.handle_softirqs 5.09 ± 54% -5.1 0.00 perf-profile.children.cycles-pp.check_dir_item 4.74 ± 34% -4.7 0.00 perf-profile.children.cycles-pp.sysvec_call_function_single 4.57 ± 53% -4.6 0.00 perf-profile.children.cycles-pp.csum_tree_block 4.54 ± 70% -4.5 0.00 perf-profile.children.cycles-pp.asm_sysvec_posted_msi_notification 4.49 ± 53% -4.5 0.00 perf-profile.children.cycles-pp.crc_pcl 4.45 ± 56% -4.5 0.00 perf-profile.children.cycles-pp.btrfs_get_32 4.06 ± 34% -4.1 0.00 perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt 8.89 ± 14% -8.9 0.00 perf-profile.self.cycles-pp.acpi_safe_halt 4.47 ± 53% -4.5 0.00 perf-profile.self.cycles-pp.crc_pcl [4] ========================================================================================= compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase: gcc-12/performance/x86_64-rhel-8.3/process/100%/debian-12-x86_64-20240206.cgz/lkp-skl-fpga01/context_switch1/will-it-scale commit: cb1898f58e ("perf annotate-data: Support --skip-empty option") 0c365693d5 <-- replacement patch cb1898f58e0f175d 0c365693d5482c855ba9acd45a9 ---------------- --------------------------- %stddev %change %stddev \ | \ 64399 ± 2% +258.9% 231157 ± 11% cpuidle..usage 1317337 ± 5% -77.3% 299607 ± 52% numa-numastat.node1.local_node 1363498 ± 5% -76.6% 319125 ± 46% numa-numastat.node1.numa_hit 27.91 -59.1% 11.42 ± 4% vmstat.cpu.us 18301349 -59.9% 7340407 ± 4% vmstat.system.cs 129807 +448.8% 712411 vmstat.system.in 18547591 -61.2% 7201865 ± 5% will-it-scale.104.processes 178341 -61.2% 69248 ± 5% will-it-scale.per_process_ops 18547591 -61.2% 7201865 ± 5% will-it-scale.workload 0.41 ± 4% +0.9 1.33 mpstat.cpu.all.irq% 0.01 ± 3% -0.0 0.01 ± 4% mpstat.cpu.all.soft% 70.70 +15.8 86.54 mpstat.cpu.all.sys% 28.09 -16.7 11.37 ± 4% mpstat.cpu.all.usr% 197.00 ± 65% -100.0% 0.00 perf-c2c.DRAM.local 345.00 ± 13% -100.0% 0.00 perf-c2c.DRAM.remote 6145 ± 30% -100.0% 0.00 perf-c2c.HITM.local 207.50 ± 9% -100.0% 0.00 perf-c2c.HITM.remote 6353 ± 29% -100.0% 0.00 perf-c2c.HITM.total 75677 ± 11% -75.1% 18806 ± 77% numa-meminfo.node0.Mapped 2716604 ± 3% -99.9% 1820 ± 47% numa-meminfo.node1.Active 2716596 ± 3% -99.9% 1812 ± 47% numa-meminfo.node1.Active(anon) 4505110 ± 29% -65.6% 1549919 ± 89% numa-meminfo.node1.FilePages 1211376 ± 16% -60.6% 477658 ± 43% numa-meminfo.node1.Inactive 1211335 ± 16% -60.6% 477589 ± 43% numa-meminfo.node1.Inactive(anon) 451272 ± 15% -96.0% 18078 ± 79% numa-meminfo.node1.Mapped 5431186 ± 24% -48.9% 2777179 ± 46% numa-meminfo.node1.MemUsed 3537633 ± 4% -99.9% 4987 ± 37% numa-meminfo.node1.Shmem 19126 ± 11% -74.9% 4809 ± 77% numa-vmstat.node0.nr_mapped 679371 ± 3% -99.9% 453.29 ± 47% numa-vmstat.node1.nr_active_anon 1126311 ± 29% -65.6% 387479 ± 89% numa-vmstat.node1.nr_file_pages 302619 ± 16% -60.5% 119398 ± 43% numa-vmstat.node1.nr_inactive_anon 113050 ± 15% -95.9% 4645 ± 79% numa-vmstat.node1.nr_mapped 884442 ± 4% -99.9% 1247 ± 37% numa-vmstat.node1.nr_shmem 679371 ± 3% -99.9% 453.29 ± 47% numa-vmstat.node1.nr_zone_active_anon 302620 ± 16% -60.5% 119398 ± 43% numa-vmstat.node1.nr_zone_inactive_anon 1362527 ± 5% -76.6% 318783 ± 46% numa-vmstat.node1.numa_hit 1316366 ± 5% -77.3% 299265 ± 52% numa-vmstat.node1.numa_local 2719746 ± 3% -99.9% 3460 ± 2% meminfo.Active 2719698 ± 3% -99.9% 3428 ± 2% meminfo.Active(anon) 6617563 ± 2% -53.4% 3084531 meminfo.Cached 6890243 ± 2% -51.3% 3357677 meminfo.Committed_AS 1534356 ± 7% -54.6% 697259 meminfo.Inactive 1534138 ± 7% -54.6% 697031 meminfo.Inactive(anon) 107936 -12.1% 94833 meminfo.KReclaimable 528224 ± 12% -93.0% 36860 meminfo.Mapped 8481255 ± 2% -36.8% 5356536 meminfo.Memused 107936 -12.1% 94833 meminfo.SReclaimable 3544842 ± 4% -99.7% 11867 meminfo.Shmem 9043642 ± 2% -38.3% 5581530 meminfo.max_used_kB 120076 ± 10% +43.6% 172392 ± 11% sched_debug.cfs_rq:/.avg_vruntime.stddev 1.29 -14.6% 1.10 ± 3% sched_debug.cfs_rq:/.h_nr_running.avg 1093 ±141% +306.5% 4446 ± 50% sched_debug.cfs_rq:/.load.min 120076 ± 10% +43.6% 172392 ± 11% sched_debug.cfs_rq:/.min_vruntime.stddev 1815 ± 4% +13.8% 2066 ± 4% sched_debug.cfs_rq:/.runnable_avg.max 1209 -14.1% 1038 ± 2% sched_debug.cfs_rq:/.runnable_avg.min 97.73 ± 12% +69.5% 165.67 ± 7% sched_debug.cfs_rq:/.runnable_avg.stddev 804.36 ± 3% -14.8% 685.33 ± 4% sched_debug.cfs_rq:/.util_avg.min 63.65 ± 6% +21.3% 77.20 ± 6% sched_debug.cfs_rq:/.util_avg.stddev 618.13 +18.0% 729.46 ± 2% sched_debug.cfs_rq:/.util_est.avg 1100 ± 5% +41.5% 1556 ± 7% sched_debug.cfs_rq:/.util_est.max 793984 ± 2% -32.0% 539895 ± 3% sched_debug.cpu.avg_idle.avg 2817 ± 7% -21.9% 2200 ± 5% sched_debug.cpu.curr->pid.min 2.39 ± 3% +12.5% 2.69 ± 3% sched_debug.cpu.nr_running.max 26222425 -55.0% 11804141 ± 4% sched_debug.cpu.nr_switches.avg 27158282 -54.5% 12358425 ± 4% sched_debug.cpu.nr_switches.max 24464466 ± 2% -56.0% 10752464 ± 5% sched_debug.cpu.nr_switches.min 605107 ± 12% -47.2% 319551 ± 15% sched_debug.cpu.nr_switches.stddev 679317 ± 3% -99.9% 856.63 ± 2% proc-vmstat.nr_active_anon 178194 -2.9% 173071 proc-vmstat.nr_anon_pages 4677129 +1.7% 4755105 proc-vmstat.nr_dirty_background_threshold 9365694 +1.7% 9521838 proc-vmstat.nr_dirty_threshold 1654099 ± 2% -53.4% 771132 proc-vmstat.nr_file_pages 47057782 +1.7% 47838697 proc-vmstat.nr_free_pages 383856 ± 7% -54.6% 174258 proc-vmstat.nr_inactive_anon 132404 ± 12% -92.9% 9449 proc-vmstat.nr_mapped 4649 -1.8% 4564 proc-vmstat.nr_page_table_pages 885917 ± 4% -99.7% 2965 proc-vmstat.nr_shmem 26984 -12.1% 23708 proc-vmstat.nr_slab_reclaimable 679317 ± 3% -99.9% 856.63 ± 2% proc-vmstat.nr_zone_active_anon 383856 ± 7% -54.6% 174258 proc-vmstat.nr_zone_inactive_anon 201323 ± 11% -98.7% 2700 ±111% proc-vmstat.numa_hint_faults 98999 ± 21% -98.5% 1473 ±161% proc-vmstat.numa_hint_faults_local 1907797 ± 2% -59.0% 781318 proc-vmstat.numa_hit 1805585 ± 2% -61.6% 693715 proc-vmstat.numa_local 107648 -18.6% 87591 proc-vmstat.numa_other 23053 ± 70% -82.8% 3957 ±249% proc-vmstat.numa_pages_migrated 399248 ± 16% -86.2% 55155 ±122% proc-vmstat.numa_pte_updates 958792 ± 5% -100.0% 225.75 ± 6% proc-vmstat.pgactivate 2040912 ± 2% -55.0% 919056 proc-vmstat.pgalloc_normal 1329369 ± 2% +15435.5% 2.065e+08 ± 2% proc-vmstat.pgfault 1004379 -27.6% 727209 ± 2% proc-vmstat.pgfree 23053 ± 70% -82.8% 3957 ±249% proc-vmstat.pgmigrate_success 49086 ± 20% -20.5% 39045 proc-vmstat.pgreuse 0.03 -38.8% 0.02 ± 6% perf-stat.i.MPKI 2.522e+10 +72.3% 4.344e+10 ± 3% perf-stat.i.branch-instructions 1.55 -0.7 0.88 ± 20% perf-stat.i.branch-miss-rate% 3289507 ± 2% -15.8% 2769952 ± 5% perf-stat.i.cache-misses 18477740 -59.9% 7405320 ± 4% perf-stat.i.context-switches 2.20 -38.0% 1.37 ± 3% perf-stat.i.cpi 2.835e+11 -2.0% 2.777e+11 perf-stat.i.cpu-cycles 189.99 +3.0% 195.65 perf-stat.i.cpu-migrations 154844 ± 2% -32.0% 105248 ± 6% perf-stat.i.cycles-between-cache-misses 1.292e+11 +61.4% 2.085e+11 ± 3% perf-stat.i.instructions 0.46 +64.4% 0.75 ± 3% perf-stat.i.ipc 177.68 -52.6% 84.26 ± 3% perf-stat.i.metric.K/sec 4186 ± 2% +16117.8% 678906 ± 2% perf-stat.i.minor-faults 4186 ± 2% +16117.3% 678906 ± 2% perf-stat.i.page-faults 0.03 ± 2% -47.8% 0.01 ± 7% perf-stat.overall.MPKI 1.54 -0.7 0.84 ± 22% perf-stat.overall.branch-miss-rate% 2.19 -39.2% 1.33 ± 3% perf-stat.overall.cpi 86165 ± 2% +16.8% 100667 ± 6% perf-stat.overall.cycles-between-cache-misses 0.46 +64.8% 0.75 ± 3% perf-stat.overall.ipc 2105030 +316.7% 8772170 perf-stat.overall.path-length 2.513e+10 +72.3% 4.331e+10 ± 3% perf-stat.ps.branch-instructions 3280483 ± 2% -15.9% 2760114 ± 5% perf-stat.ps.cache-misses 18416035 -60.0% 7375563 ± 4% perf-stat.ps.context-switches 2.825e+11 -2.0% 2.768e+11 perf-stat.ps.cpu-cycles 189.25 +3.0% 194.99 perf-stat.ps.cpu-migrations 1.288e+11 +61.4% 2.079e+11 ± 3% perf-stat.ps.instructions 4169 ± 2% +16132.3% 676851 ± 2% perf-stat.ps.minor-faults 4169 ± 2% +16131.8% 676851 ± 2% perf-stat.ps.page-faults 3.904e+13 +61.7% 6.313e+13 ± 3% perf-stat.total.instructions 53.50 -53.5 0.00 perf-profile.calltrace.cycles-pp.read 45.69 -45.7 0.00 perf-profile.calltrace.cycles-pp.write 35.13 -35.1 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read 31.89 -31.9 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 29.98 -30.0 0.00 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write 26.86 -26.9 0.00 perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 26.74 -26.7 0.00 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 25.88 -25.9 0.00 perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read 23.71 -23.7 0.00 perf-profile.calltrace.cycles-pp.pipe_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe 21.52 -21.5 0.00 perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 20.61 -20.6 0.00 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write 18.62 ± 2% -18.6 0.00 perf-profile.calltrace.cycles-pp.pipe_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe 16.92 -16.9 0.00 perf-profile.calltrace.cycles-pp.schedule.pipe_read.vfs_read.ksys_read.do_syscall_64 16.55 -16.6 0.00 perf-profile.calltrace.cycles-pp.__schedule.schedule.pipe_read.vfs_read.ksys_read 14.74 ± 2% -14.7 0.00 perf-profile.calltrace.cycles-pp.__wake_up_sync_key.pipe_write.vfs_write.ksys_write.do_syscall_64 14.31 ± 2% -14.3 0.00 perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_sync_key.pipe_write.vfs_write.ksys_write 14.10 ± 2% -14.1 0.00 perf-profile.calltrace.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write.vfs_write 13.74 ± 2% -13.7 0.00 perf-profile.calltrace.cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.pipe_write 9.02 -9.0 0.00 perf-profile.calltrace.cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key 7.16 -7.2 0.00 perf-profile.calltrace.cycles-pp.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common 6.93 -6.9 0.00 perf-profile.calltrace.cycles-pp.dequeue_task_fair.__schedule.schedule.pipe_read.vfs_read 6.62 -6.6 0.00 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.read 6.14 ± 2% -6.1 0.00 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.write 65.61 -65.6 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe 58.74 -58.7 0.00 perf-profile.children.cycles-pp.do_syscall_64 53.92 -53.9 0.00 perf-profile.children.cycles-pp.read 46.12 -46.1 0.00 perf-profile.children.cycles-pp.write 26.87 -26.9 0.00 perf-profile.children.cycles-pp.ksys_read 25.90 -25.9 0.00 perf-profile.children.cycles-pp.vfs_read 23.87 -23.9 0.00 perf-profile.children.cycles-pp.pipe_read 21.55 -21.5 0.00 perf-profile.children.cycles-pp.ksys_write 20.65 -20.7 0.00 perf-profile.children.cycles-pp.vfs_write 19.74 -19.7 0.00 perf-profile.children.cycles-pp.schedule 19.36 -19.4 0.00 perf-profile.children.cycles-pp.__schedule 18.67 ± 2% -18.7 0.00 perf-profile.children.cycles-pp.pipe_write 14.75 ± 2% -14.8 0.00 perf-profile.children.cycles-pp.__wake_up_sync_key 14.32 ± 2% -14.3 0.00 perf-profile.children.cycles-pp.__wake_up_common 14.10 ± 2% -14.1 0.00 perf-profile.children.cycles-pp.autoremove_wake_function 13.79 ± 2% -13.8 0.00 perf-profile.children.cycles-pp.try_to_wake_up 12.86 -12.9 0.00 perf-profile.children.cycles-pp.syscall_return_via_sysret 10.00 -10.0 0.00 perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack 9.04 -9.0 0.00 perf-profile.children.cycles-pp.ttwu_do_activate 8.44 -8.4 0.00 perf-profile.children.cycles-pp.syscall_exit_to_user_mode 7.17 -7.2 0.00 perf-profile.children.cycles-pp.enqueue_task_fair 6.95 -6.9 0.00 perf-profile.children.cycles-pp.dequeue_task_fair 6.09 -6.1 0.00 perf-profile.children.cycles-pp.entry_SYSCALL_64 12.54 -12.5 0.00 perf-profile.self.cycles-pp.syscall_return_via_sysret 9.91 -9.9 0.00 perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack 6.97 -7.0 0.00 perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe 5.47 -5.5 0.00 perf-profile.self.cycles-pp.entry_SYSCALL_64 ^ permalink raw reply related [flat|nested] 5+ messages in thread
end of thread, other threads:[~2024-08-20 5:01 UTC | newest] Thread overview: 5+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2024-08-09 20:10 [PATCH] perf: Remove setting of page->index and ->mapping Matthew Wilcox (Oracle) 2024-08-10 22:15 ` Peter Zijlstra 2024-08-16 6:35 ` kernel test robot 2024-08-16 17:51 ` Matthew Wilcox 2024-08-20 5:01 ` Oliver Sang
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox