Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 00218C61DA3 for ; Tue, 21 Feb 2023 16:57:49 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S234321AbjBUQ5s (ORCPT ); Tue, 21 Feb 2023 11:57:48 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:58978 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S233600AbjBUQ5q (ORCPT ); Tue, 21 Feb 2023 11:57:46 -0500 Received: from smtp-fw-9102.amazon.com (smtp-fw-9102.amazon.com [207.171.184.29]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 06EFE2CFD2 for ; Tue, 21 Feb 2023 08:57:37 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=amazon.de; i=@amazon.de; q=dns/txt; s=amazon201209; t=1676998657; x=1708534657; h=date:from:to:cc:subject:message-id:references: mime-version:in-reply-to:content-transfer-encoding; bh=fhEoY9IU3G7XCWBaSUwtYVs3CkU8ozJyP8tt7abOQGQ=; b=CaDT5EdsNxiEPFaXCPD56tDosbuVk6nfpA0OlxNYDHLYQVfLRSwaPx6S +zgUKPs3sh2FLpohzmUD2SslvWVuka+GqOteFT5mqXShPcDBCJNKNMHHG C5Q8jw0XyGuR3t5criQU5T6ygDCtzFaDVaW2lnHeNrI0whAUpyWiq5xu8 M=; X-IronPort-AV: E=Sophos;i="5.97,315,1669075200"; d="scan'208";a="313643089" Received: from pdx4-co-svc-p1-lb2-vlan2.amazon.com (HELO email-inbound-relay-iad-1e-m6i4x-0aba4706.us-east-1.amazon.com) ([10.25.36.210]) by smtp-border-fw-9102.sea19.amazon.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 21 Feb 2023 16:57:28 +0000 Received: from EX13D41EUC003.ant.amazon.com (iad12-ws-svc-p26-lb9-vlan2.iad.amazon.com [10.40.163.34]) by email-inbound-relay-iad-1e-m6i4x-0aba4706.us-east-1.amazon.com (Postfix) with ESMTPS id 3A1D2A2878; Tue, 21 Feb 2023 16:57:23 +0000 (UTC) Received: from EX19D033EUC004.ant.amazon.com (10.252.61.133) by EX13D41EUC003.ant.amazon.com (10.43.164.188) with Microsoft SMTP Server (TLS) id 15.0.1497.46; Tue, 21 Feb 2023 16:57:23 +0000 Received: from u40bc5e070a0153.ant.amazon.com (10.1.212.12) by EX19D033EUC004.ant.amazon.com (10.252.61.133) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.1118.24; Tue, 21 Feb 2023 16:57:18 +0000 Date: Tue, 21 Feb 2023 17:57:12 +0100 From: Roman Kagan To: Vincent Guittot , Peter Zijlstra CC: , Valentin Schneider , Zhang Qiao , Ben Segall , "Waiman Long" , Steven Rostedt , Mel Gorman , Dietmar Eggemann , "Daniel Bristot de Oliveira" , Ingo Molnar , Juri Lelli Subject: Re: [PATCH v3] sched/fair: sanitize vruntime of entity being placed Message-ID: Mail-Followup-To: Roman Kagan , Vincent Guittot , Peter Zijlstra , linux-kernel@vger.kernel.org, Valentin Schneider , Zhang Qiao , Ben Segall , Waiman Long , Steven Rostedt , Mel Gorman , Dietmar Eggemann , Daniel Bristot de Oliveira , Ingo Molnar , Juri Lelli References: <20230209193107.1432770-1-rkagan@amazon.de> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Disposition: inline In-Reply-To: X-Originating-IP: [10.1.212.12] X-ClientProxiedBy: EX19D041UWA003.ant.amazon.com (10.13.139.105) To EX19D033EUC004.ant.amazon.com (10.252.61.133) Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Feb 21, 2023 at 10:38:44AM +0100, Vincent Guittot wrote: > On Thu, 9 Feb 2023 at 20:31, Roman Kagan wrote: > > > > From: Zhang Qiao > > > > When a scheduling entity is placed onto cfs_rq, its vruntime is pulled > > to the base level (around cfs_rq->min_vruntime), so that the entity > > doesn't gain extra boost when placed backwards. > > > > However, if the entity being placed wasn't executed for a long time, its > > vruntime may get too far behind (e.g. while cfs_rq was executing a > > low-weight hog), which can inverse the vruntime comparison due to s64 > > overflow. This results in the entity being placed with its original > > vruntime way forwards, so that it will effectively never get to the cpu. > > > > To prevent that, ignore the vruntime of the entity being placed if it > > didn't execute for longer than the time that can lead to an overflow. > > > > Signed-off-by: Zhang Qiao > > [rkagan: formatted, adjusted commit log, comments, cutoff value] > > Co-developed-by: Roman Kagan > > Signed-off-by: Roman Kagan > = > Reviewed-by: Vincent Guittot > = > > --- > > v2 -> v3: > > - make cutoff less arbitrary and update comments [Vincent] > > > > v1 -> v2: > > - add Zhang Qiao's s-o-b > > - fix constant promotion on 32bit > > > > kernel/sched/fair.c | 21 +++++++++++++++++++-- > > 1 file changed, 19 insertions(+), 2 deletions(-) Turns out Peter took v2 through his tree, and it has already landed in Linus' master. What scares me, though, is that I've got a message from the test robot that this commit drammatically affected hackbench results, see the quote below. I expected the commit not to affect any benchmarks. Any idea what could have caused this change? Thanks, Roman. On Tue, Feb 21, 2023 at 03:34:16PM +0800, kernel test robot wrote: > FYI, we noticed a 125.5% improvement of hackbench.throughput due to commi= t: > = > commit: 829c1651e9c4a6f78398d3e67651cef9bb6b42cc ("sched/fair: sanitize v= runtime of entity being placed") > https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master > = > in testcase: hackbench > on test machine: 88 threads 2 sockets Intel(R) Xeon(R) Gold 6238M CPU @ 2= .10GHz (Cascade Lake) with 128G memory > with following parameters: > = > nr_threads: 50% > iterations: 8 > mode: process > ipc: pipe > cpufreq_governor: performance > = > test-description: Hackbench is both a benchmark and a stress test for the= Linux kernel scheduler. > test-url: https://github.com/linux-test-project/ltp/blob/master/testcases= /kernel/sched/cfs-scheduler/hackbench.c > = > In addition to that, the commit also has significant impact on the follow= ing tests: > = > +------------------+--------------------------------------------------+ > | testcase: change | hackbench: hackbench.throughput -8.1% regression | > | test machine | 104 threads 2 sockets (Skylake) with 192G memory | > | test parameters | cpufreq_governor=3Dperformance | > | | ipc=3Dsocket | > | | iterations=3D4 | > | | mode=3Dprocess | > | | nr_threads=3D100% | > +------------------+--------------------------------------------------+ > = > Details are as below: > = > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > compiler/cpufreq_governor/ipc/iterations/kconfig/mode/nr_threads/rootfs/t= box_group/testcase: > gcc-11/performance/pipe/8/x86_64-rhel-8.3/process/50%/debian-11.1-x86_6= 4-20220510.cgz/lkp-csl-2sp9/hackbench > = > commit: > a2e90611b9 ("sched/fair: Remove capacity inversion detection") > 829c1651e9 ("sched/fair: sanitize vruntime of entity being placed") > = > a2e90611b9f425ad 829c1651e9c4a6f78398d3e6765 > ---------------- --------------------------- > %stddev %change %stddev > \ | \ > 308887 =B1 5% +125.5% 696539 hackbench.throughput > 259291 =B1 2% +127.3% 589293 hackbench.throughput_avg > 308887 =B1 5% +125.5% 696539 hackbench.throughput_best > 198770 =B1 2% +105.5% 408552 =B1 4% hackbench.throughput_wo= rst > 319.60 =B1 2% -55.8% 141.24 hackbench.time.elapsed_ti= me > 319.60 =B1 2% -55.8% 141.24 hackbench.time.elapsed_ti= me.max > 1.298e+09 =B1 8% -87.6% 1.613e+08 =B1 7% hackbench.time.involunt= ary_context_switches > 477107 -12.5% 417660 hackbench.time.minor_page_f= aults > 24683 =B1 2% -57.2% 10562 hackbench.time.system_time > 2136 =B1 3% -45.0% 1174 hackbench.time.user_time > 3.21e+09 =B1 4% -83.0% 5.442e+08 =B1 3% hackbench.time.voluntar= y_context_switches > 5.28e+08 =B1 4% +8.4% 5.723e+08 =B1 3% cpuidle..time > 365.97 =B1 2% -48.9% 187.12 uptime.boot > 3322559 =B1 3% +34.3% 4463206 =B1 15% vmstat.memory.cache > 14194257 =B1 2% -62.8% 5279904 =B1 3% vmstat.system.cs > 2120781 =B1 3% -72.8% 576421 =B1 4% vmstat.system.in > 1.84 =B1 12% +2.6 4.48 =B1 5% mpstat.cpu.all.idle% > 2.49 =B1 3% -1.1 1.39 =B1 4% mpstat.cpu.all.irq% > 0.04 =B1 12% +0.0 0.05 mpstat.cpu.all.soft% > 7.36 +2.2 9.56 mpstat.cpu.all.usr% > 61555 =B1 6% -72.8% 16751 =B1 16% numa-meminfo.node1.Acti= ve > 61515 =B1 6% -72.8% 16717 =B1 16% numa-meminfo.node1.Acti= ve(anon) > 960182 =B1102% +225.6% 3125990 =B1 42% numa-meminfo.node1.File= Pages > 1754002 =B1 53% +137.9% 4173379 =B1 34% numa-meminfo.node1.MemU= sed > 35296824 =B1 6% +157.8% 91005048 numa-numastat.node0.local= _node > 35310119 =B1 6% +157.9% 91058472 numa-numastat.node0.numa_= hit > 35512423 =B1 5% +159.7% 92232951 numa-numastat.node1.local= _node > 35577275 =B1 4% +159.4% 92273266 numa-numastat.node1.numa_= hit > 35310253 =B1 6% +157.9% 91058211 numa-vmstat.node0.numa_hit > 35296958 =B1 6% +157.8% 91004787 numa-vmstat.node0.numa_lo= cal > 15337 =B1 6% -72.5% 4216 =B1 17% numa-vmstat.node1.nr_ac= tive_anon > 239988 =B1102% +225.7% 781607 =B1 42% numa-vmstat.node1.nr_fi= le_pages > 15337 =B1 6% -72.5% 4216 =B1 17% numa-vmstat.node1.nr_zo= ne_active_anon > 35577325 =B1 4% +159.4% 92273215 numa-vmstat.node1.numa_hit > 35512473 =B1 5% +159.7% 92232900 numa-vmstat.node1.numa_lo= cal > 64500 =B1 8% -61.8% 24643 =B1 32% meminfo.Active > 64422 =B1 8% -61.9% 24568 =B1 32% meminfo.Active(anon) > 140271 =B1 14% -38.0% 86979 =B1 24% meminfo.AnonHugePages > 372672 =B1 2% +13.3% 422069 meminfo.AnonPages > 3205235 =B1 3% +35.1% 4329061 =B1 15% meminfo.Cached > 1548601 =B1 7% +77.4% 2747319 =B1 24% meminfo.Committed_AS > 783193 =B1 14% +154.9% 1996137 =B1 33% meminfo.Inactive > 783010 =B1 14% +154.9% 1995951 =B1 33% meminfo.Inactive(anon) > 4986534 =B1 2% +28.2% 6394741 =B1 10% meminfo.Memused > 475092 =B1 22% +236.5% 1598918 =B1 41% meminfo.Shmem > 2777 -2.1% 2719 turbostat.Bzy_MHz > 11143123 =B1 6% +72.0% 19162667 turbostat.C1 > 0.24 =B1 7% +0.7 0.94 =B1 3% turbostat.C1% > 100440 =B1 18% +203.8% 305136 =B1 15% turbostat.C1E > 0.06 =B1 9% +0.1 0.18 =B1 11% turbostat.C1E% > 1.24 =B1 3% +1.6 2.81 =B1 4% turbostat.C6% > 1.38 =B1 3% +156.1% 3.55 =B1 3% turbostat.CPU%c1 > 0.33 =B1 5% +76.5% 0.58 =B1 7% turbostat.CPU%c6 > 0.16 +31.2% 0.21 turbostat.IPC > 6.866e+08 =B1 5% -87.8% 83575393 =B1 5% turbostat.IRQ > 0.33 =B1 27% +0.2 0.57 turbostat.POLL% > 0.12 =B1 10% +176.4% 0.33 =B1 12% turbostat.Pkg%pc2 > 0.09 =B1 7% -100.0% 0.00 turbostat.Pkg%pc6 > 61.33 +5.2% 64.50 =B1 2% turbostat.PkgTmp > 14.81 +2.0% 15.11 turbostat.RAMWatt > 16242 =B1 8% -62.0% 6179 =B1 32% proc-vmstat.nr_active_a= non > 93150 =B1 2% +13.2% 105429 proc-vmstat.nr_anon_pages > 801219 =B1 3% +35.1% 1082320 =B1 15% proc-vmstat.nr_file_pag= es > 195506 =B1 14% +155.2% 498919 =B1 33% proc-vmstat.nr_inactive= _anon > 118682 =B1 22% +236.9% 399783 =B1 41% proc-vmstat.nr_shmem > 16242 =B1 8% -62.0% 6179 =B1 32% proc-vmstat.nr_zone_act= ive_anon > 195506 =B1 14% +155.2% 498919 =B1 33% proc-vmstat.nr_zone_ina= ctive_anon > 70889233 =B1 5% +158.6% 1.833e+08 proc-vmstat.numa_hit > 70811086 =B1 5% +158.8% 1.832e+08 proc-vmstat.numa_local > 55885 =B1 22% -67.2% 18327 =B1 38% proc-vmstat.numa_pages_= migrated > 422312 =B1 10% -95.4% 19371 =B1 7% proc-vmstat.pgactivate > 71068460 =B1 5% +158.1% 1.834e+08 proc-vmstat.pgalloc_normal > 1554994 -19.6% 1250346 =B1 4% proc-vmstat.pgfault > 71011267 =B1 5% +155.9% 1.817e+08 proc-vmstat.pgfree > 55885 =B1 22% -67.2% 18327 =B1 38% proc-vmstat.pgmigrate_s= uccess > 111247 =B1 2% -35.0% 72355 =B1 2% proc-vmstat.pgreuse > 2506368 =B1 2% -53.1% 1176320 proc-vmstat.unevictable_p= gs_scanned > 20.06 =B1 10% -22.4% 15.56 =B1 8% sched_debug.cfs_rq:/.h_= nr_running.max > 0.81 =B1 32% -93.1% 0.06 =B1223% sched_debug.cfs_rq:/.h_= nr_running.min > 1917 =B1 34% -100.0% 0.00 sched_debug.cfs_rq:/.load= .min > 24.18 =B1 10% +39.0% 33.62 =B1 11% sched_debug.cfs_rq:/.lo= ad_avg.avg > 245.61 =B1 25% +66.3% 408.33 =B1 22% sched_debug.cfs_rq:/.lo= ad_avg.max > 47.52 =B1 13% +72.6% 82.03 =B1 8% sched_debug.cfs_rq:/.lo= ad_avg.stddev > 13431147 -64.9% 4717147 sched_debug.cfs_rq:/.min_vr= untime.avg > 18161799 =B1 7% -67.4% 5925316 =B1 6% sched_debug.cfs_rq:/.mi= n_vruntime.max > 12413026 -65.0% 4340952 sched_debug.cfs_rq:/.min_vr= untime.min > 739748 =B1 16% -66.6% 247410 =B1 17% sched_debug.cfs_rq:/.mi= n_vruntime.stddev > 0.85 -16.4% 0.71 sched_debug.cfs_rq:/.nr_run= ning.avg > 0.61 =B1 25% -90.9% 0.06 =B1223% sched_debug.cfs_rq:/.nr= _running.min > 0.10 =B1 25% +109.3% 0.22 =B1 7% sched_debug.cfs_rq:/.nr= _running.stddev > 169.22 +101.7% 341.33 sched_debug.cfs_rq:/.remove= d.load_avg.max > 32.41 =B1 24% +100.2% 64.90 =B1 16% sched_debug.cfs_rq:/.re= moved.load_avg.stddev > 82.92 =B1 10% +108.1% 172.56 sched_debug.cfs_rq:/.remo= ved.runnable_avg.max > 13.60 =B1 28% +114.0% 29.10 =B1 20% sched_debug.cfs_rq:/.re= moved.runnable_avg.stddev > 82.92 =B1 10% +108.1% 172.56 sched_debug.cfs_rq:/.remo= ved.util_avg.max > 13.60 =B1 28% +114.0% 29.10 =B1 20% sched_debug.cfs_rq:/.re= moved.util_avg.stddev > 2156 =B1 12% -36.6% 1368 =B1 27% sched_debug.cfs_rq:/.ru= nnable_avg.min > 2285 =B1 7% -19.8% 1833 =B1 6% sched_debug.cfs_rq:/.ru= nnable_avg.stddev > -2389921 -64.8% -840940 sched_debug.cfs_rq:/.spread= 0.min > 739781 =B1 16% -66.5% 247837 =B1 17% sched_debug.cfs_rq:/.sp= read0.stddev > 843.88 =B1 2% -20.5% 670.53 sched_debug.cfs_rq:/.util= _avg.avg > 433.64 =B1 7% -43.5% 244.83 =B1 17% sched_debug.cfs_rq:/.ut= il_avg.min > 187.00 =B1 6% +40.6% 263.02 =B1 4% sched_debug.cfs_rq:/.ut= il_avg.stddev > 394.15 =B1 14% -29.5% 278.06 =B1 3% sched_debug.cfs_rq:/.ut= il_est_enqueued.avg > 1128 =B1 12% -17.6% 930.39 =B1 5% sched_debug.cfs_rq:/.ut= il_est_enqueued.max > 38.36 =B1 29% -100.0% 0.00 sched_debug.cfs_rq:/.util= _est_enqueued.min > 3596 =B1 15% -39.5% 2175 =B1 7% sched_debug.cpu.avg_idl= e.min > 160647 =B1 9% -25.9% 118978 =B1 9% sched_debug.cpu.avg_idl= e.stddev > 197365 -46.2% 106170 sched_debug.cpu.clock.avg > 197450 -46.2% 106208 sched_debug.cpu.clock.max > 197281 -46.2% 106128 sched_debug.cpu.clock.min > 49.96 =B1 22% -53.1% 23.44 =B1 19% sched_debug.cpu.clock.s= tddev > 193146 -45.7% 104898 sched_debug.cpu.clock_task.= avg > 194592 -45.8% 105455 sched_debug.cpu.clock_task.= max > 177878 -49.3% 90211 sched_debug.cpu.clock_task.= min > 1794 =B1 5% -10.7% 1602 =B1 2% sched_debug.cpu.clock_t= ask.stddev > 13154 =B1 2% -20.3% 10479 sched_debug.cpu.curr->pid= .avg > 15059 -17.2% 12468 sched_debug.cpu.curr->pid.m= ax > 7263 =B1 33% -100.0% 0.00 sched_debug.cpu.curr->pid= .min > 9321 =B1 36% +98.2% 18478 =B1 44% sched_debug.cpu.max_idl= e_balance_cost.stddev > 0.00 =B1 17% -41.6% 0.00 =B1 13% sched_debug.cpu.next_ba= lance.stddev > 20.00 =B1 11% -21.4% 15.72 =B1 7% sched_debug.cpu.nr_runn= ing.max > 0.86 =B1 17% -87.1% 0.11 =B1141% sched_debug.cpu.nr_runn= ing.min > 25069883 -83.7% 4084117 =B1 4% sched_debug.cpu.nr_switch= es.avg > 26486718 -82.8% 4544009 =B1 4% sched_debug.cpu.nr_switch= es.max > 23680077 -84.5% 3663816 =B1 4% sched_debug.cpu.nr_switch= es.min > 589836 =B1 3% -68.7% 184621 =B1 16% sched_debug.cpu.nr_swit= ches.stddev > 197278 -46.2% 106128 sched_debug.cpu_clk > 194327 -46.9% 103176 sched_debug.ktime > 197967 -46.0% 106821 sched_debug.sched_clk > 14.91 -37.6% 9.31 perf-stat.i.MPKI > 2.657e+10 +25.0% 3.32e+10 perf-stat.i.branch-instruct= ions > 1.17 -0.4 0.78 perf-stat.i.branch-miss-rat= e% > 3.069e+08 -20.1% 2.454e+08 perf-stat.i.branch-misses > 6.43 =B1 8% +2.2 8.59 =B1 4% perf-stat.i.cache-miss-= rate% > 1.952e+09 -24.3% 1.478e+09 perf-stat.i.cache-references > 14344055 =B1 2% -58.6% 5932018 =B1 3% perf-stat.i.context-swi= tches > 1.83 -21.8% 1.43 perf-stat.i.cpi > 2.403e+11 -3.4% 2.322e+11 perf-stat.i.cpu-cycles > 1420139 =B1 2% -38.8% 869692 =B1 5% perf-stat.i.cpu-migrati= ons > 2619 =B1 7% -15.5% 2212 =B1 8% perf-stat.i.cycles-betw= een-cache-misses > 0.24 =B1 19% -0.1 0.10 =B1 17% perf-stat.i.dTLB-load-m= iss-rate% > 90403286 =B1 19% -55.8% 39926283 =B1 16% perf-stat.i.dTLB-load-m= isses > 3.823e+10 +28.6% 4.918e+10 perf-stat.i.dTLB-loads > 0.01 =B1 34% -0.0 0.01 =B1 33% perf-stat.i.dTLB-store-= miss-rate% > 2779663 =B1 34% -52.7% 1315899 =B1 31% perf-stat.i.dTLB-store-= misses > 2.19e+10 +24.2% 2.72e+10 perf-stat.i.dTLB-stores > 47.99 =B1 2% +28.0 75.94 perf-stat.i.iTLB-load-mis= s-rate% > 89417955 =B1 2% +38.7% 1.24e+08 =B1 4% perf-stat.i.iTLB-load-m= isses > 97721514 =B1 2% -58.2% 40865783 =B1 3% perf-stat.i.iTLB-loads > 1.329e+11 +26.3% 1.678e+11 perf-stat.i.instructions > 1503 -7.7% 1388 =B1 3% perf-stat.i.instructions-= per-iTLB-miss > 0.55 +30.2% 0.72 perf-stat.i.ipc > 1.64 =B1 18% +217.4% 5.20 =B1 11% perf-stat.i.major-faults > 2.73 -3.7% 2.63 perf-stat.i.metric.GHz > 1098 =B1 2% -7.1% 1020 =B1 3% perf-stat.i.metric.K/sec > 1008 +24.4% 1254 perf-stat.i.metric.M/sec > 4334 =B1 2% +90.5% 8257 =B1 7% perf-stat.i.minor-faults > 90.94 -14.9 75.99 perf-stat.i.node-load-miss-= rate% > 41932510 =B1 8% -43.0% 23899176 =B1 10% perf-stat.i.node-load-m= isses > 3366677 =B1 5% +86.2% 6267816 perf-stat.i.node-loads > 81.77 =B1 3% -36.3 45.52 =B1 3% perf-stat.i.node-store-= miss-rate% > 18498318 =B1 7% -31.8% 12613933 =B1 7% perf-stat.i.node-store-= misses > 3023556 =B1 10% +508.7% 18405880 =B1 2% perf-stat.i.node-stores > 4336 =B1 2% +90.5% 8262 =B1 7% perf-stat.i.page-faults > 14.70 -41.2% 8.65 perf-stat.overall.MPKI > 1.16 -0.4 0.72 perf-stat.overall.branch-mi= ss-rate% > 6.22 =B1 7% +2.4 8.59 =B1 4% perf-stat.overall.cache= -miss-rate% > 1.81 -24.3% 1.37 perf-stat.overall.cpi > 0.24 =B1 19% -0.2 0.07 =B1 15% perf-stat.overall.dTLB-= load-miss-rate% > 0.01 =B1 34% -0.0 0.00 =B1 29% perf-stat.overall.dTLB-= store-miss-rate% > 47.78 =B1 2% +29.3 77.12 perf-stat.overall.iTLB-lo= ad-miss-rate% > 1486 -9.1% 1351 =B1 4% perf-stat.overall.instruc= tions-per-iTLB-miss > 0.55 +32.0% 0.73 perf-stat.overall.ipc > 92.54 -15.4 77.16 =B1 2% perf-stat.overall.node-lo= ad-miss-rate% > 85.82 =B1 2% -48.1 37.76 =B1 5% perf-stat.overall.node-= store-miss-rate% > 2.648e+10 +25.2% 3.314e+10 perf-stat.ps.branch-instruc= tions > 3.06e+08 -22.1% 2.383e+08 perf-stat.ps.branch-misses > 1.947e+09 -25.5% 1.451e+09 perf-stat.ps.cache-referenc= es > 14298713 =B1 2% -62.5% 5359285 =B1 3% perf-stat.ps.context-sw= itches > 2.396e+11 -4.0% 2.299e+11 perf-stat.ps.cpu-cycles > 1415512 =B1 2% -42.2% 817981 =B1 4% perf-stat.ps.cpu-migrat= ions > 90073948 =B1 19% -60.4% 35711862 =B1 15% perf-stat.ps.dTLB-load-= misses > 3.811e+10 +29.7% 4.944e+10 perf-stat.ps.dTLB-loads > 2767291 =B1 34% -56.3% 1210210 =B1 29% perf-stat.ps.dTLB-store= -misses > 2.183e+10 +25.0% 2.729e+10 perf-stat.ps.dTLB-stores > 89118809 =B1 2% +39.6% 1.244e+08 =B1 4% perf-stat.ps.iTLB-load-= misses > 97404381 =B1 2% -62.2% 36860047 =B1 3% perf-stat.ps.iTLB-loads > 1.324e+11 +26.7% 1.678e+11 perf-stat.ps.instructions > 1.62 =B1 18% +164.7% 4.29 =B1 8% perf-stat.ps.major-faul= ts > 4310 =B1 2% +75.1% 7549 =B1 5% perf-stat.ps.minor-faul= ts > 41743097 =B1 8% -47.3% 21984450 =B1 9% perf-stat.ps.node-load-= misses > 3356259 =B1 5% +92.6% 6462631 perf-stat.ps.node-loads > 18414647 =B1 7% -35.7% 11833799 =B1 6% perf-stat.ps.node-store= -misses > 3019790 =B1 10% +545.0% 19478071 perf-stat.ps.node-stores > 4312 =B1 2% +75.2% 7553 =B1 5% perf-stat.ps.page-faults > 4.252e+13 -43.7% 2.395e+13 perf-stat.total.instructions > 29.92 =B1 4% -22.8 7.09 =B1 29% perf-profile.calltrace.= cycles-pp.__wake_up_common_lock.pipe_write.vfs_write.ksys_write.do_syscall_= 64 > 28.53 =B1 5% -21.6 6.92 =B1 29% perf-profile.calltrace.= cycles-pp.__wake_up_common.__wake_up_common_lock.pipe_write.vfs_write.ksys_= write > 27.86 =B1 5% -21.1 6.77 =B1 29% perf-profile.calltrace.= cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_common_lock.p= ipe_write.vfs_write > 27.55 =B1 5% -20.9 6.68 =B1 29% perf-profile.calltrace.= cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_u= p_common_lock.pipe_write > 22.28 =B1 4% -17.0 5.31 =B1 30% perf-profile.calltrace.= cycles-pp.schedule.pipe_read.vfs_read.ksys_read.do_syscall_64 > 21.98 =B1 4% -16.7 5.24 =B1 30% perf-profile.calltrace.= cycles-pp.__schedule.schedule.pipe_read.vfs_read.ksys_read > 12.62 =B1 4% -9.6 3.00 =B1 33% perf-profile.calltrace.= cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_u= p_common.__wake_up_common_lock > 34.09 -9.2 24.92 =B1 3% perf-profile.calltrace.cy= cles-pp.pipe_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_h= wframe > 11.48 =B1 5% -8.8 2.69 =B1 38% perf-profile.calltrace.= cycles-pp.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.autoremove_wake= _function.__wake_up_common > 9.60 =B1 7% -7.2 2.40 =B1 35% perf-profile.calltrace.= cycles-pp.dequeue_task_fair.__schedule.schedule.pipe_read.vfs_read > 36.39 -6.2 30.20 perf-profile.calltrace.cycl= es-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.__lib= c_read > 40.40 -6.1 34.28 perf-profile.calltrace.cycl= es-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read > 40.95 -5.7 35.26 perf-profile.calltrace.cycl= es-pp.entry_SYSCALL_64_after_hwframe.__libc_read > 37.43 -5.4 32.07 perf-profile.calltrace.cycl= es-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_read > 6.30 =B1 11% -5.2 1.09 =B1 36% perf-profile.calltrace.= cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64= .entry_SYSCALL_64_after_hwframe.__libc_write > 5.66 =B1 12% -5.1 0.58 =B1 75% perf-profile.calltrace.= cycles-pp.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_= user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe > 6.46 =B1 10% -5.1 1.40 =B1 28% perf-profile.calltrace.= cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hw= frame.__libc_write > 5.53 =B1 13% -5.0 0.56 =B1 75% perf-profile.calltrace.= cycles-pp.schedule.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall= _exit_to_user_mode.do_syscall_64 > 5.42 =B1 13% -4.9 0.56 =B1 75% perf-profile.calltrace.= cycles-pp.__schedule.schedule.exit_to_user_mode_loop.exit_to_user_mode_prep= are.syscall_exit_to_user_mode > 5.82 =B1 9% -4.7 1.10 =B1 37% perf-profile.calltrace.= cycles-pp._raw_spin_lock.try_to_wake_up.autoremove_wake_function.__wake_up_= common.__wake_up_common_lock > 5.86 =B1 16% -4.6 1.31 =B1 37% perf-profile.calltrace.= cycles-pp.enqueue_entity.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.= autoremove_wake_function > 5.26 =B1 9% -4.4 0.89 =B1 57% perf-profile.calltrace.= cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.try_to_wake_up.au= toremove_wake_function.__wake_up_common > 45.18 -3.5 41.68 perf-profile.calltrace.cycl= es-pp.__libc_read > 50.31 -3.2 47.12 perf-profile.calltrace.cycl= es-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write > 4.00 =B1 27% -2.9 1.09 =B1 40% perf-profile.calltrace.= cycles-pp.dequeue_entity.dequeue_task_fair.__schedule.schedule.pipe_read > 50.75 -2.7 48.06 perf-profile.calltrace.cycl= es-pp.entry_SYSCALL_64_after_hwframe.__libc_write > 40.80 -2.6 38.20 perf-profile.calltrace.cycl= es-pp.pipe_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_= hwframe > 3.10 =B1 15% -2.5 0.62 =B1103% perf-profile.calltrace.= cycles-pp.update_cfs_group.dequeue_task_fair.__schedule.schedule.pipe_read > 2.94 =B1 12% -2.3 0.62 =B1102% perf-profile.calltrace.= cycles-pp.update_cfs_group.enqueue_task_fair.ttwu_do_activate.try_to_wake_u= p.autoremove_wake_function > 2.38 =B1 9% -2.0 0.38 =B1102% perf-profile.calltrace.= cycles-pp._raw_spin_lock.__schedule.schedule.pipe_read.vfs_read > 2.24 =B1 7% -1.8 0.40 =B1 71% perf-profile.calltrace.= cycles-pp.prepare_to_wait_event.pipe_read.vfs_read.ksys_read.do_syscall_64 > 2.08 =B1 6% -1.8 0.29 =B1100% perf-profile.calltrace.= cycles-pp.pick_next_task_fair.__schedule.schedule.pipe_read.vfs_read > 2.10 =B1 10% -1.8 0.32 =B1104% perf-profile.calltrace.= cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__schedule.schedu= le.pipe_read > 2.76 =B1 7% -1.5 1.24 =B1 17% perf-profile.calltrace.= cycles-pp.select_task_rq.try_to_wake_up.autoremove_wake_function.__wake_up_= common.__wake_up_common_lock > 2.27 =B1 5% -1.4 0.88 =B1 11% perf-profile.calltrace.= cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64= .entry_SYSCALL_64_after_hwframe.__libc_read > 2.43 =B1 7% -1.3 1.16 =B1 17% perf-profile.calltrace.= cycles-pp.select_task_rq_fair.select_task_rq.try_to_wake_up.autoremove_wake= _function.__wake_up_common > 2.46 =B1 5% -1.3 1.20 =B1 7% perf-profile.calltrace.= cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hw= frame.__libc_read > 1.54 =B1 5% -1.2 0.32 =B1101% perf-profile.calltrace.= cycles-pp.update_load_avg.enqueue_entity.enqueue_task_fair.ttwu_do_activate= .try_to_wake_up > 0.97 =B1 9% -0.3 0.66 =B1 19% perf-profile.calltrace.= cycles-pp.select_idle_sibling.select_task_rq_fair.select_task_rq.try_to_wak= e_up.autoremove_wake_function > 0.86 =B1 6% +0.2 1.02 perf-profile.calltrace.cy= cles-pp.__might_fault._copy_from_iter.copy_page_from_iter.pipe_write.vfs_wr= ite > 0.64 =B1 9% +0.5 1.16 =B1 5% perf-profile.calltrace.= cycles-pp.mutex_unlock.pipe_read.vfs_read.ksys_read.do_syscall_64 > 0.47 =B1 45% +0.5 0.99 =B1 5% perf-profile.calltrace.= cycles-pp.__fget_light.__fdget_pos.ksys_read.do_syscall_64.entry_SYSCALL_64= _after_hwframe > 0.60 =B1 8% +0.5 1.13 =B1 5% perf-profile.calltrace.= cycles-pp.__fdget_pos.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwfram= e.__libc_read > 0.00 +0.5 0.54 =B1 5% perf-profile.calltrace.cy= cles-pp.current_time.file_update_time.pipe_write.vfs_write.ksys_write > 0.00 +0.6 0.56 =B1 4% perf-profile.calltrace.cy= cles-pp.__might_resched.__might_fault._copy_from_iter.copy_page_from_iter.p= ipe_write > 0.00 +0.6 0.56 =B1 7% perf-profile.calltrace.cy= cles-pp.__might_resched.__might_fault._copy_to_iter.copy_page_to_iter.pipe_= read > 0.00 +0.6 0.58 =B1 5% perf-profile.calltrace.cy= cles-pp.__might_resched.mutex_lock.pipe_write.vfs_write.ksys_write > 0.00 +0.6 0.62 =B1 3% perf-profile.calltrace.cy= cles-pp.__might_resched.mutex_lock.pipe_read.vfs_read.ksys_read > 0.00 +0.7 0.65 =B1 6% perf-profile.calltrace.cy= cles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.prepare_to_= wait_event.pipe_write.vfs_write > 0.00 +0.7 0.65 =B1 7% perf-profile.calltrace.cy= cles-pp.poll_idle.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_id= le > 0.57 =B1 5% +0.7 1.24 =B1 6% perf-profile.calltrace.= cycles-pp.__fget_light.__fdget_pos.ksys_write.do_syscall_64.entry_SYSCALL_6= 4_after_hwframe > 0.00 +0.7 0.72 =B1 6% perf-profile.calltrace.cy= cles-pp._raw_spin_lock_irqsave.prepare_to_wait_event.pipe_write.vfs_write.k= sys_write > 0.00 +0.8 0.75 =B1 6% perf-profile.calltrace.cy= cles-pp.mutex_spin_on_owner.__mutex_lock.pipe_write.vfs_write.ksys_write > 0.74 =B1 9% +0.8 1.48 =B1 5% perf-profile.calltrace.= cycles-pp.file_update_time.pipe_write.vfs_write.ksys_write.do_syscall_64 > 0.63 =B1 5% +0.8 1.40 =B1 5% perf-profile.calltrace.= cycles-pp.__fdget_pos.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwfra= me.__libc_write > 0.00 +0.8 0.78 =B1 19% perf-profile.calltrace.cy= cles-pp.reader__read_event.perf_session__process_events.record__finish_outp= ut.__cmd_record > 0.00 +0.8 0.78 =B1 19% perf-profile.calltrace.cy= cles-pp.record__finish_output.__cmd_record > 0.00 +0.8 0.78 =B1 19% perf-profile.calltrace.cy= cles-pp.perf_session__process_events.record__finish_output.__cmd_record > 0.00 +0.8 0.80 =B1 15% perf-profile.calltrace.cy= cles-pp.__cmd_record > 0.00 +0.8 0.82 =B1 11% perf-profile.calltrace.cy= cles-pp.intel_idle_irq.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.= do_idle > 0.00 +0.9 0.85 =B1 6% perf-profile.calltrace.cy= cles-pp.prepare_to_wait_event.pipe_write.vfs_write.ksys_write.do_syscall_64 > 0.00 +0.9 0.86 =B1 4% perf-profile.calltrace.cy= cles-pp.current_time.atime_needs_update.touch_atime.pipe_read.vfs_read > 0.00 +0.9 0.87 =B1 5% perf-profile.calltrace.cy= cles-pp.syscall_return_via_sysret.__libc_write > 0.00 +0.9 0.88 =B1 5% perf-profile.calltrace.cy= cles-pp.syscall_return_via_sysret.__libc_read > 0.26 =B1100% +1.0 1.22 =B1 10% perf-profile.calltrace.= cycles-pp.__schedule.schedule.pipe_write.vfs_write.ksys_write > 0.00 +1.0 0.96 =B1 6% perf-profile.calltrace.cy= cles-pp.__might_fault._copy_to_iter.copy_page_to_iter.pipe_read.vfs_read > 0.27 =B1100% +1.0 1.23 =B1 10% perf-profile.calltrace.= cycles-pp.schedule.pipe_write.vfs_write.ksys_write.do_syscall_64 > 0.00 +1.0 0.97 =B1 7% perf-profile.calltrace.cy= cles-pp.page_counter_uncharge.uncharge_batch.__mem_cgroup_uncharge.__folio_= put.pipe_read > 0.87 =B1 8% +1.1 1.98 =B1 5% perf-profile.calltrace.= cycles-pp.apparmor_file_permission.security_file_permission.vfs_read.ksys_r= ead.do_syscall_64 > 0.73 =B1 6% +1.1 1.85 =B1 5% perf-profile.calltrace.= cycles-pp.apparmor_file_permission.security_file_permission.vfs_write.ksys_= write.do_syscall_64 > 0.00 +1.2 1.15 =B1 7% perf-profile.calltrace.cy= cles-pp.uncharge_batch.__mem_cgroup_uncharge.__folio_put.pipe_read.vfs_read > 0.00 +1.2 1.23 =B1 6% perf-profile.calltrace.cy= cles-pp.__mem_cgroup_uncharge.__folio_put.pipe_read.vfs_read.ksys_read > 0.00 +1.2 1.24 =B1 7% perf-profile.calltrace.cy= cles-pp.__folio_put.pipe_read.vfs_read.ksys_read.do_syscall_64 > 0.48 =B1 45% +1.3 1.74 =B1 6% perf-profile.calltrace.= cycles-pp.atime_needs_update.touch_atime.pipe_read.vfs_read.ksys_read > 0.60 =B1 7% +1.3 1.87 =B1 8% perf-profile.calltrace.= cycles-pp.__wake_up_common_lock.pipe_read.vfs_read.ksys_read.do_syscall_64 > 1.23 =B1 7% +1.3 2.51 =B1 4% perf-profile.calltrace.= cycles-pp.mutex_lock.pipe_read.vfs_read.ksys_read.do_syscall_64 > 43.42 +1.3 44.75 perf-profile.calltrace.cycl= es-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write > 0.83 =B1 7% +1.3 2.17 =B1 5% perf-profile.calltrace.= cycles-pp.security_file_permission.vfs_write.ksys_write.do_syscall_64.entry= _SYSCALL_64_after_hwframe > 0.98 =B1 7% +1.4 2.36 =B1 6% perf-profile.calltrace.= cycles-pp.security_file_permission.vfs_read.ksys_read.do_syscall_64.entry_S= YSCALL_64_after_hwframe > 0.27 =B1100% +1.4 1.70 =B1 9% perf-profile.calltrace.= cycles-pp.__wake_up_common.__wake_up_common_lock.pipe_read.vfs_read.ksys_re= ad > 0.79 =B1 8% +1.4 2.23 =B1 6% perf-profile.calltrace.= cycles-pp.touch_atime.pipe_read.vfs_read.ksys_read.do_syscall_64 > 0.18 =B1141% +1.5 1.63 =B1 9% perf-profile.calltrace.= cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_u= p_common_lock.pipe_read > 0.18 =B1141% +1.5 1.67 =B1 9% perf-profile.calltrace.= cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_common_lock.p= ipe_read.vfs_read > 0.00 +1.6 1.57 =B1 10% perf-profile.calltrace.cy= cles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_sta= rtup_entry > 0.00 +1.6 1.57 =B1 10% perf-profile.calltrace.cy= cles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_sec= ondary > 1.05 =B1 8% +1.7 2.73 =B1 6% perf-profile.calltrace.= cycles-pp.copy_user_enhanced_fast_string.copyin._copy_from_iter.copy_page_f= rom_iter.pipe_write > 1.84 =B1 9% +1.7 3.56 =B1 5% perf-profile.calltrace.= cycles-pp.copy_user_enhanced_fast_string.copyout._copy_to_iter.copy_page_to= _iter.pipe_read > 1.41 =B1 9% +1.8 3.17 =B1 6% perf-profile.calltrace.= cycles-pp.copyin._copy_from_iter.copy_page_from_iter.pipe_write.vfs_write > 0.00 +1.8 1.79 =B1 9% perf-profile.calltrace.cy= cles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.seconda= ry_startup_64_no_verify > 1.99 =B1 9% +2.0 3.95 =B1 5% perf-profile.calltrace.= cycles-pp.copyout._copy_to_iter.copy_page_to_iter.pipe_read.vfs_read > 2.40 =B1 7% +2.4 4.82 =B1 5% perf-profile.calltrace.= cycles-pp._copy_from_iter.copy_page_from_iter.pipe_write.vfs_write.ksys_wri= te > 0.00 +2.5 2.50 =B1 7% perf-profile.calltrace.cy= cles-pp.__mutex_lock.pipe_write.vfs_write.ksys_write.do_syscall_64 > 2.89 =B1 8% +2.6 5.47 =B1 5% perf-profile.calltrace.= cycles-pp.copy_page_from_iter.pipe_write.vfs_write.ksys_write.do_syscall_64 > 1.04 =B1 30% +2.8 3.86 =B1 5% perf-profile.calltrace.= cycles-pp.__entry_text_start.__libc_write > 0.00 +2.9 2.90 =B1 11% perf-profile.calltrace.cy= cles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64_no_v= erify > 0.00 +2.9 2.91 =B1 11% perf-profile.calltrace.cy= cles-pp.cpu_startup_entry.start_secondary.secondary_startup_64_no_verify > 0.00 +2.9 2.91 =B1 11% perf-profile.calltrace.cy= cles-pp.start_secondary.secondary_startup_64_no_verify > 0.85 =B1 27% +2.9 3.80 =B1 5% perf-profile.calltrace.= cycles-pp.__entry_text_start.__libc_read > 0.00 +3.0 2.96 =B1 11% perf-profile.calltrace.cy= cles-pp.secondary_startup_64_no_verify > 2.60 =B1 9% +3.1 5.74 =B1 6% perf-profile.calltrace.= cycles-pp._copy_to_iter.copy_page_to_iter.pipe_read.vfs_read.ksys_read > 2.93 =B1 9% +3.7 6.66 =B1 5% perf-profile.calltrace.= cycles-pp.copy_page_to_iter.pipe_read.vfs_read.ksys_read.do_syscall_64 > 1.60 =B1 12% +4.6 6.18 =B1 7% perf-profile.calltrace.= cycles-pp.mutex_unlock.pipe_write.vfs_write.ksys_write.do_syscall_64 > 2.60 =B1 10% +4.6 7.24 =B1 5% perf-profile.calltrace.= cycles-pp.mutex_lock.pipe_write.vfs_write.ksys_write.do_syscall_64 > 28.75 =B1 5% -21.6 7.19 =B1 28% perf-profile.children.c= ycles-pp.schedule > 30.52 =B1 4% -21.6 8.97 =B1 22% perf-profile.children.c= ycles-pp.__wake_up_common_lock > 28.53 =B1 6% -21.0 7.56 =B1 26% perf-profile.children.c= ycles-pp.__schedule > 29.04 =B1 5% -20.4 8.63 =B1 23% perf-profile.children.c= ycles-pp.__wake_up_common > 28.37 =B1 5% -19.9 8.44 =B1 23% perf-profile.children.c= ycles-pp.autoremove_wake_function > 28.08 =B1 5% -19.7 8.33 =B1 23% perf-profile.children.c= ycles-pp.try_to_wake_up > 13.90 =B1 2% -10.2 3.75 =B1 28% perf-profile.children.c= ycles-pp.ttwu_do_activate > 12.66 =B1 3% -9.2 3.47 =B1 29% perf-profile.children.c= ycles-pp.enqueue_task_fair > 34.20 -9.2 25.05 =B1 3% perf-profile.children.cyc= les-pp.pipe_read > 90.86 -9.1 81.73 perf-profile.children.cycle= s-pp.do_syscall_64 > 91.80 -8.3 83.49 perf-profile.children.cycle= s-pp.entry_SYSCALL_64_after_hwframe > 10.28 =B1 7% -7.8 2.53 =B1 27% perf-profile.children.c= ycles-pp._raw_spin_lock > 9.85 =B1 7% -6.9 2.92 =B1 29% perf-profile.children.c= ycles-pp.dequeue_task_fair > 8.69 =B1 7% -6.6 2.05 =B1 24% perf-profile.children.c= ycles-pp.exit_to_user_mode_prepare > 8.99 =B1 6% -6.2 2.81 =B1 16% perf-profile.children.c= ycles-pp.syscall_exit_to_user_mode > 36.46 -6.1 30.34 perf-profile.children.cycle= s-pp.vfs_read > 8.38 =B1 8% -5.8 2.60 =B1 23% perf-profile.children.c= ycles-pp.native_queued_spin_lock_slowpath > 6.10 =B1 11% -5.4 0.66 =B1 61% perf-profile.children.c= ycles-pp.exit_to_user_mode_loop > 37.45 -5.3 32.13 perf-profile.children.cycle= s-pp.ksys_read > 6.50 =B1 35% -4.9 1.62 =B1 61% perf-profile.children.c= ycles-pp.update_curr > 6.56 =B1 15% -4.6 1.95 =B1 57% perf-profile.children.c= ycles-pp.update_cfs_group > 6.38 =B1 14% -4.5 1.91 =B1 28% perf-profile.children.c= ycles-pp.enqueue_entity > 5.74 =B1 5% -3.8 1.92 =B1 25% perf-profile.children.c= ycles-pp.update_load_avg > 45.56 -3.8 41.75 perf-profile.children.cycle= s-pp.__libc_read > 3.99 =B1 4% -3.1 0.92 =B1 24% perf-profile.children.c= ycles-pp.pick_next_task_fair > 4.12 =B1 27% -2.7 1.39 =B1 34% perf-profile.children.c= ycles-pp.dequeue_entity > 40.88 -2.5 38.37 perf-profile.children.cycle= s-pp.pipe_write > 3.11 =B1 4% -2.4 0.75 =B1 22% perf-profile.children.c= ycles-pp.switch_mm_irqs_off > 2.06 =B1 33% -1.8 0.27 =B1 27% perf-profile.children.c= ycles-pp.asm_sysvec_call_function_single > 2.38 =B1 41% -1.8 0.60 =B1 72% perf-profile.children.c= ycles-pp.perf_trace_sched_wakeup_template > 2.29 =B1 5% -1.7 0.60 =B1 25% perf-profile.children.c= ycles-pp.switch_fpu_return > 2.30 =B1 6% -1.6 0.68 =B1 18% perf-profile.children.c= ycles-pp.prepare_task_switch > 1.82 =B1 33% -1.6 0.22 =B1 31% perf-profile.children.c= ycles-pp.sysvec_call_function_single > 1.77 =B1 33% -1.6 0.20 =B1 32% perf-profile.children.c= ycles-pp.__sysvec_call_function_single > 1.96 =B1 5% -1.5 0.50 =B1 20% perf-profile.children.c= ycles-pp.reweight_entity > 2.80 =B1 7% -1.2 1.60 =B1 12% perf-profile.children.c= ycles-pp.select_task_rq > 1.61 =B1 6% -1.2 0.42 =B1 25% perf-profile.children.c= ycles-pp.restore_fpregs_from_fpstate > 1.34 =B1 9% -1.2 0.16 =B1 28% perf-profile.children.c= ycles-pp._raw_spin_unlock_irqrestore > 1.62 =B1 4% -1.2 0.45 =B1 22% perf-profile.children.c= ycles-pp.set_next_entity > 1.55 =B1 8% -1.1 0.43 =B1 12% perf-profile.children.c= ycles-pp.update_rq_clock > 1.49 =B1 8% -1.1 0.41 =B1 14% perf-profile.children.c= ycles-pp.__update_load_avg_cfs_rq > 1.30 =B1 20% -1.0 0.26 =B1 18% perf-profile.children.c= ycles-pp.finish_task_switch > 1.44 =B1 5% -1.0 0.42 =B1 19% perf-profile.children.c= ycles-pp.__switch_to_asm > 2.47 =B1 7% -1.0 1.50 =B1 12% perf-profile.children.c= ycles-pp.select_task_rq_fair > 2.33 =B1 7% -0.9 1.40 =B1 3% perf-profile.children.c= ycles-pp.prepare_to_wait_event > 1.24 =B1 7% -0.9 0.35 =B1 14% perf-profile.children.c= ycles-pp.__update_load_avg_se > 1.41 =B1 32% -0.9 0.56 =B1 24% perf-profile.children.c= ycles-pp.sched_ttwu_pending > 2.29 =B1 8% -0.8 1.45 =B1 3% perf-profile.children.c= ycles-pp._raw_spin_lock_irqsave > 1.04 =B1 7% -0.8 0.24 =B1 22% perf-profile.children.c= ycles-pp.check_preempt_curr > 1.01 =B1 3% -0.7 0.30 =B1 20% perf-profile.children.c= ycles-pp.__switch_to > 0.92 =B1 7% -0.7 0.26 =B1 12% perf-profile.children.c= ycles-pp.update_min_vruntime > 0.71 =B1 2% -0.6 0.08 =B1 75% perf-profile.children.c= ycles-pp.put_prev_entity > 0.76 =B1 6% -0.6 0.14 =B1 32% perf-profile.children.c= ycles-pp.check_preempt_wakeup > 0.81 =B1 66% -0.6 0.22 =B1 34% perf-profile.children.c= ycles-pp.set_task_cpu > 0.82 =B1 17% -0.6 0.23 =B1 10% perf-profile.children.c= ycles-pp.cpuacct_charge > 1.08 =B1 15% -0.6 0.51 =B1 10% perf-profile.children.c= ycles-pp.wake_affine > 0.56 =B1 15% -0.5 0.03 =B1100% perf-profile.children.c= ycles-pp.asm_sysvec_reschedule_ipi > 0.66 =B1 3% -0.5 0.15 =B1 28% perf-profile.children.c= ycles-pp.os_xsave > 0.52 =B1 44% -0.5 0.06 =B1151% perf-profile.children.c= ycles-pp.native_irq_return_iret > 0.55 =B1 5% -0.4 0.15 =B1 21% perf-profile.children.c= ycles-pp.__calc_delta > 0.56 =B1 10% -0.4 0.17 =B1 26% perf-profile.children.c= ycles-pp.___perf_sw_event > 0.70 =B1 15% -0.4 0.32 =B1 11% perf-profile.children.c= ycles-pp.task_h_load > 0.40 =B1 4% -0.3 0.06 =B1 49% perf-profile.children.c= ycles-pp.pick_next_entity > 0.57 =B1 6% -0.3 0.26 =B1 7% perf-profile.children.c= ycles-pp.__list_del_entry_valid > 0.39 =B1 8% -0.3 0.08 =B1 24% perf-profile.children.c= ycles-pp.set_next_buddy > 0.64 =B1 6% -0.3 0.36 =B1 6% perf-profile.children.c= ycles-pp._raw_spin_lock_irq > 0.53 =B1 20% -0.3 0.25 =B1 8% perf-profile.children.c= ycles-pp.ttwu_queue_wakelist > 0.36 =B1 8% -0.3 0.08 =B1 11% perf-profile.children.c= ycles-pp.rb_insert_color > 0.41 =B1 6% -0.3 0.14 =B1 17% perf-profile.children.c= ycles-pp.sched_clock_cpu > 0.36 =B1 33% -0.3 0.10 =B1 17% perf-profile.children.c= ycles-pp.__flush_smp_call_function_queue > 0.37 =B1 4% -0.2 0.13 =B1 16% perf-profile.children.c= ycles-pp.native_sched_clock > 0.28 =B1 5% -0.2 0.07 =B1 18% perf-profile.children.c= ycles-pp.rb_erase > 0.32 =B1 7% -0.2 0.12 =B1 10% perf-profile.children.c= ycles-pp.__list_add_valid > 0.23 =B1 6% -0.2 0.03 =B1103% perf-profile.children.c= ycles-pp.resched_curr > 0.27 =B1 5% -0.2 0.08 =B1 20% perf-profile.children.c= ycles-pp.__wrgsbase_inactive > 0.26 =B1 6% -0.2 0.08 =B1 17% perf-profile.children.c= ycles-pp.finish_wait > 0.26 =B1 4% -0.2 0.08 =B1 11% perf-profile.children.c= ycles-pp.rcu_note_context_switch > 0.33 =B1 21% -0.2 0.15 =B1 32% perf-profile.children.c= ycles-pp.migrate_task_rq_fair > 0.22 =B1 9% -0.2 0.07 =B1 22% perf-profile.children.c= ycles-pp.perf_trace_buf_update > 0.17 =B1 8% -0.1 0.03 =B1100% perf-profile.children.c= ycles-pp.rb_next > 0.15 =B1 32% -0.1 0.03 =B1100% perf-profile.children.c= ycles-pp.llist_reverse_order > 0.34 =B1 7% -0.1 0.26 =B1 3% perf-profile.children.c= ycles-pp.anon_pipe_buf_release > 0.14 =B1 6% -0.1 0.07 =B1 17% perf-profile.children.c= ycles-pp.read@plt > 0.10 =B1 17% -0.1 0.04 =B1 75% perf-profile.children.c= ycles-pp.remove_entity_load_avg > 0.07 =B1 10% -0.0 0.02 =B1 99% perf-profile.children.c= ycles-pp.generic_update_time > 0.11 =B1 6% -0.0 0.07 =B1 8% perf-profile.children.c= ycles-pp.__mark_inode_dirty > 0.00 +0.1 0.06 =B1 9% perf-profile.children.cyc= les-pp.load_balance > 0.00 +0.1 0.06 =B1 11% perf-profile.children.cyc= les-pp._raw_spin_trylock > 0.00 +0.1 0.06 =B1 7% perf-profile.children.cyc= les-pp.uncharge_folio > 0.00 +0.1 0.06 =B1 7% perf-profile.children.cyc= les-pp.__do_softirq > 0.00 +0.1 0.07 =B1 10% perf-profile.children.cyc= les-pp.tick_nohz_get_sleep_length > 0.00 +0.1 0.08 =B1 14% perf-profile.children.cyc= les-pp.__get_obj_cgroup_from_memcg > 0.15 =B1 23% +0.1 0.23 =B1 7% perf-profile.children.c= ycles-pp.task_tick_fair > 0.19 =B1 17% +0.1 0.28 =B1 7% perf-profile.children.c= ycles-pp.scheduler_tick > 0.00 +0.1 0.10 =B1 21% perf-profile.children.cyc= les-pp.select_idle_core > 0.00 +0.1 0.10 =B1 9% perf-profile.children.cyc= les-pp.osq_unlock > 0.23 =B1 12% +0.1 0.34 =B1 6% perf-profile.children.c= ycles-pp.update_process_times > 0.37 =B1 13% +0.1 0.48 =B1 5% perf-profile.children.c= ycles-pp.hrtimer_interrupt > 0.24 =B1 12% +0.1 0.35 =B1 6% perf-profile.children.c= ycles-pp.tick_sched_handle > 0.31 =B1 14% +0.1 0.43 =B1 4% perf-profile.children.c= ycles-pp.__hrtimer_run_queues > 0.37 =B1 12% +0.1 0.49 =B1 5% perf-profile.children.c= ycles-pp.__sysvec_apic_timer_interrupt > 0.00 +0.1 0.12 =B1 10% perf-profile.children.cyc= les-pp.__mod_memcg_state > 0.26 =B1 10% +0.1 0.38 =B1 6% perf-profile.children.c= ycles-pp.tick_sched_timer > 0.00 +0.1 0.13 =B1 7% perf-profile.children.cyc= les-pp.free_unref_page > 0.00 +0.1 0.14 =B1 8% perf-profile.children.cyc= les-pp.rmqueue > 0.15 =B1 8% +0.2 0.30 =B1 5% perf-profile.children.c= ycles-pp.rcu_all_qs > 0.16 =B1 6% +0.2 0.31 =B1 5% perf-profile.children.c= ycles-pp.__x64_sys_write > 0.00 +0.2 0.16 =B1 10% perf-profile.children.cyc= les-pp.propagate_protected_usage > 0.00 +0.2 0.16 =B1 10% perf-profile.children.cyc= les-pp.menu_select > 0.00 +0.2 0.16 =B1 9% perf-profile.children.cyc= les-pp.memcg_account_kmem > 0.42 =B1 12% +0.2 0.57 =B1 4% perf-profile.children.c= ycles-pp.sysvec_apic_timer_interrupt > 0.15 =B1 11% +0.2 0.31 =B1 8% perf-profile.children.c= ycles-pp.__x64_sys_read > 0.00 +0.2 0.17 =B1 8% perf-profile.children.cyc= les-pp.get_page_from_freelist > 0.44 =B1 11% +0.2 0.62 =B1 4% perf-profile.children.c= ycles-pp.asm_sysvec_apic_timer_interrupt > 0.10 =B1 31% +0.2 0.28 =B1 24% perf-profile.children.c= ycles-pp.mnt_user_ns > 0.16 =B1 4% +0.2 0.35 =B1 5% perf-profile.children.c= ycles-pp.kill_fasync > 0.20 =B1 10% +0.2 0.40 =B1 3% perf-profile.children.c= ycles-pp.syscall_exit_to_user_mode_prepare > 0.09 =B1 7% +0.2 0.29 =B1 4% perf-profile.children.c= ycles-pp.page_copy_sane > 0.08 =B1 8% +0.2 0.31 =B1 6% perf-profile.children.c= ycles-pp.rw_verify_area > 0.12 =B1 11% +0.2 0.36 =B1 8% perf-profile.children.c= ycles-pp.ktime_get_coarse_real_ts64 > 0.28 =B1 12% +0.2 0.52 =B1 5% perf-profile.children.c= ycles-pp.inode_needs_update_time > 0.00 +0.3 0.27 =B1 7% perf-profile.children.cyc= les-pp.__memcg_kmem_charge_page > 0.43 =B1 6% +0.3 0.73 =B1 5% perf-profile.children.c= ycles-pp.__cond_resched > 0.21 =B1 29% +0.3 0.54 =B1 15% perf-profile.children.c= ycles-pp.select_idle_cpu > 0.10 =B1 10% +0.3 0.43 =B1 17% perf-profile.children.c= ycles-pp.fsnotify_perm > 0.23 =B1 11% +0.3 0.56 =B1 6% perf-profile.children.c= ycles-pp.syscall_enter_from_user_mode > 0.06 =B1 75% +0.4 0.47 =B1 27% perf-profile.children.c= ycles-pp.queue_event > 0.21 =B1 9% +0.4 0.62 =B1 5% perf-profile.children.c= ycles-pp.entry_SYSCALL_64_safe_stack > 0.06 =B1 75% +0.4 0.48 =B1 26% perf-profile.children.c= ycles-pp.ordered_events__queue > 0.06 =B1 73% +0.4 0.50 =B1 24% perf-profile.children.c= ycles-pp.process_simple > 0.01 =B1223% +0.4 0.44 =B1 9% perf-profile.children.c= ycles-pp.schedule_idle > 0.05 =B1 8% +0.5 0.52 =B1 7% perf-profile.children.c= ycles-pp.__alloc_pages > 0.45 =B1 7% +0.5 0.94 =B1 5% perf-profile.children.c= ycles-pp.__get_task_ioprio > 0.89 =B1 8% +0.5 1.41 =B1 4% perf-profile.children.c= ycles-pp.__might_sleep > 0.01 =B1223% +0.5 0.54 =B1 21% perf-profile.children.c= ycles-pp.flush_smp_call_function_queue > 0.05 =B1 46% +0.5 0.60 =B1 7% perf-profile.children.c= ycles-pp.osq_lock > 0.34 =B1 8% +0.6 0.90 =B1 5% perf-profile.children.c= ycles-pp.aa_file_perm > 0.01 =B1223% +0.7 0.67 =B1 7% perf-profile.children.c= ycles-pp.poll_idle > 0.14 =B1 17% +0.7 0.82 =B1 6% perf-profile.children.c= ycles-pp.mutex_spin_on_owner > 0.12 =B1 12% +0.7 0.82 =B1 15% perf-profile.children.c= ycles-pp.__cmd_record > 0.07 =B1 72% +0.7 0.78 =B1 19% perf-profile.children.c= ycles-pp.reader__read_event > 0.07 =B1 72% +0.7 0.78 =B1 19% perf-profile.children.c= ycles-pp.record__finish_output > 0.07 =B1 72% +0.7 0.78 =B1 19% perf-profile.children.c= ycles-pp.perf_session__process_events > 0.76 =B1 8% +0.8 1.52 =B1 5% perf-profile.children.c= ycles-pp.file_update_time > 0.08 =B1 61% +0.8 0.85 =B1 11% perf-profile.children.c= ycles-pp.intel_idle_irq > 1.23 =B1 8% +0.9 2.11 =B1 4% perf-profile.children.c= ycles-pp.__might_fault > 0.02 =B1141% +1.0 0.97 =B1 7% perf-profile.children.c= ycles-pp.page_counter_uncharge > 0.51 =B1 9% +1.0 1.48 =B1 4% perf-profile.children.c= ycles-pp.current_time > 0.05 =B1 46% +1.1 1.15 =B1 7% perf-profile.children.c= ycles-pp.uncharge_batch > 1.12 =B1 6% +1.1 2.23 =B1 5% perf-profile.children.c= ycles-pp.__fget_light > 0.06 =B1 14% +1.2 1.23 =B1 6% perf-profile.children.c= ycles-pp.__mem_cgroup_uncharge > 0.06 =B1 14% +1.2 1.24 =B1 7% perf-profile.children.c= ycles-pp.__folio_put > 0.64 =B1 7% +1.2 1.83 =B1 5% perf-profile.children.c= ycles-pp.syscall_return_via_sysret > 1.19 =B1 8% +1.2 2.42 =B1 4% perf-profile.children.c= ycles-pp.__might_resched > 0.59 =B1 9% +1.3 1.84 =B1 6% perf-profile.children.c= ycles-pp.atime_needs_update > 43.47 +1.4 44.83 perf-profile.children.cycle= s-pp.ksys_write > 1.28 =B1 6% +1.4 2.68 =B1 5% perf-profile.children.c= ycles-pp.__fdget_pos > 0.80 =B1 8% +1.5 2.28 =B1 6% perf-profile.children.c= ycles-pp.touch_atime > 0.11 =B1 49% +1.5 1.59 =B1 9% perf-profile.children.c= ycles-pp.cpuidle_enter_state > 0.11 =B1 49% +1.5 1.60 =B1 9% perf-profile.children.c= ycles-pp.cpuidle_enter > 0.12 =B1 51% +1.7 1.81 =B1 9% perf-profile.children.c= ycles-pp.cpuidle_idle_call > 1.44 =B1 8% +1.8 3.22 =B1 6% perf-profile.children.c= ycles-pp.copyin > 2.00 =B1 9% +2.0 4.03 =B1 5% perf-profile.children.c= ycles-pp.copyout > 1.02 =B1 8% +2.0 3.07 =B1 5% perf-profile.children.c= ycles-pp.entry_SYSRETQ_unsafe_stack > 1.63 =B1 7% +2.3 3.90 =B1 5% perf-profile.children.c= ycles-pp.apparmor_file_permission > 2.64 =B1 8% +2.3 4.98 =B1 5% perf-profile.children.c= ycles-pp._copy_from_iter > 0.40 =B1 14% +2.5 2.92 =B1 7% perf-profile.children.c= ycles-pp.__mutex_lock > 2.91 =B1 8% +2.6 5.54 =B1 5% perf-profile.children.c= ycles-pp.copy_page_from_iter > 0.17 =B1 62% +2.7 2.91 =B1 11% perf-profile.children.c= ycles-pp.start_secondary > 1.83 =B1 7% +2.8 4.59 =B1 5% perf-profile.children.c= ycles-pp.security_file_permission > 0.17 =B1 60% +2.8 2.94 =B1 11% perf-profile.children.c= ycles-pp.do_idle > 0.17 =B1 60% +2.8 2.96 =B1 11% perf-profile.children.c= ycles-pp.secondary_startup_64_no_verify > 0.17 =B1 60% +2.8 2.96 =B1 11% perf-profile.children.c= ycles-pp.cpu_startup_entry > 2.62 =B1 9% +3.2 5.84 =B1 6% perf-profile.children.c= ycles-pp._copy_to_iter > 1.55 =B1 8% +3.2 4.79 =B1 5% perf-profile.children.c= ycles-pp.__entry_text_start > 3.09 =B1 8% +3.7 6.77 =B1 5% perf-profile.children.c= ycles-pp.copy_user_enhanced_fast_string > 2.95 =B1 9% +3.8 6.73 =B1 5% perf-profile.children.c= ycles-pp.copy_page_to_iter > 2.28 =B1 11% +5.1 7.40 =B1 6% perf-profile.children.c= ycles-pp.mutex_unlock > 3.92 =B1 9% +6.0 9.94 =B1 5% perf-profile.children.c= ycles-pp.mutex_lock > 8.37 =B1 9% -5.8 2.60 =B1 23% perf-profile.self.cycle= s-pp.native_queued_spin_lock_slowpath > 6.54 =B1 15% -4.6 1.95 =B1 57% perf-profile.self.cycle= s-pp.update_cfs_group > 3.08 =B1 4% -2.3 0.74 =B1 22% perf-profile.self.cycle= s-pp.switch_mm_irqs_off > 2.96 =B1 4% -1.8 1.13 =B1 33% perf-profile.self.cycle= s-pp.update_load_avg > 2.22 =B1 8% -1.5 0.74 =B1 12% perf-profile.self.cycle= s-pp._raw_spin_lock_irqsave > 1.96 =B1 9% -1.5 0.48 =B1 15% perf-profile.self.cycle= s-pp.update_curr > 1.94 =B1 5% -1.3 0.64 =B1 16% perf-profile.self.cycle= s-pp._raw_spin_lock > 1.78 =B1 5% -1.3 0.50 =B1 18% perf-profile.self.cycle= s-pp.__schedule > 1.59 =B1 7% -1.2 0.40 =B1 12% perf-profile.self.cycle= s-pp.enqueue_entity > 1.61 =B1 6% -1.2 0.42 =B1 25% perf-profile.self.cycle= s-pp.restore_fpregs_from_fpstate > 1.44 =B1 8% -1.0 0.39 =B1 14% perf-profile.self.cycle= s-pp.__update_load_avg_cfs_rq > 1.42 =B1 5% -1.0 0.41 =B1 19% perf-profile.self.cycle= s-pp.__switch_to_asm > 1.18 =B1 7% -0.9 0.33 =B1 14% perf-profile.self.cycle= s-pp.__update_load_avg_se > 1.14 =B1 10% -0.8 0.31 =B1 9% perf-profile.self.cycle= s-pp.update_rq_clock > 0.90 =B1 7% -0.7 0.19 =B1 21% perf-profile.self.cycle= s-pp.pick_next_task_fair > 1.04 =B1 7% -0.7 0.33 =B1 13% perf-profile.self.cycle= s-pp.prepare_task_switch > 0.98 =B1 4% -0.7 0.29 =B1 20% perf-profile.self.cycle= s-pp.__switch_to > 0.88 =B1 6% -0.7 0.20 =B1 17% perf-profile.self.cycle= s-pp.enqueue_task_fair > 1.01 =B1 6% -0.7 0.35 =B1 10% perf-profile.self.cycle= s-pp.prepare_to_wait_event > 0.90 =B1 8% -0.6 0.25 =B1 12% perf-profile.self.cycle= s-pp.update_min_vruntime > 0.79 =B1 17% -0.6 0.22 =B1 9% perf-profile.self.cycle= s-pp.cpuacct_charge > 1.10 =B1 5% -0.6 0.54 =B1 9% perf-profile.self.cycle= s-pp.try_to_wake_up > 0.66 =B1 3% -0.5 0.15 =B1 27% perf-profile.self.cycle= s-pp.os_xsave > 0.71 =B1 6% -0.5 0.22 =B1 18% perf-profile.self.cycle= s-pp.reweight_entity > 0.68 =B1 9% -0.5 0.19 =B1 10% perf-profile.self.cycle= s-pp.perf_trace_sched_wakeup_template > 0.67 =B1 9% -0.5 0.18 =B1 11% perf-profile.self.cycle= s-pp.__wake_up_common > 0.65 =B1 6% -0.5 0.17 =B1 23% perf-profile.self.cycle= s-pp.switch_fpu_return > 0.60 =B1 11% -0.5 0.14 =B1 28% perf-profile.self.cycle= s-pp.perf_tp_event > 0.52 =B1 44% -0.5 0.06 =B1151% perf-profile.self.cycle= s-pp.native_irq_return_iret > 0.52 =B1 7% -0.4 0.08 =B1 25% perf-profile.self.cycle= s-pp._raw_spin_unlock_irqrestore > 0.55 =B1 4% -0.4 0.15 =B1 22% perf-profile.self.cycle= s-pp.__calc_delta > 0.61 =B1 5% -0.4 0.21 =B1 12% perf-profile.self.cycle= s-pp.dequeue_task_fair > 0.69 =B1 14% -0.4 0.32 =B1 11% perf-profile.self.cycle= s-pp.task_h_load > 0.49 =B1 11% -0.3 0.15 =B1 29% perf-profile.self.cycle= s-pp.___perf_sw_event > 0.37 =B1 4% -0.3 0.05 =B1 73% perf-profile.self.cycle= s-pp.pick_next_entity > 0.50 =B1 3% -0.3 0.19 =B1 15% perf-profile.self.cycle= s-pp.select_idle_sibling > 0.38 =B1 9% -0.3 0.08 =B1 24% perf-profile.self.cycle= s-pp.set_next_buddy > 0.32 =B1 4% -0.3 0.03 =B1100% perf-profile.self.cycle= s-pp.put_prev_entity > 0.64 =B1 6% -0.3 0.35 =B1 7% perf-profile.self.cycle= s-pp._raw_spin_lock_irq > 0.52 =B1 5% -0.3 0.25 =B1 6% perf-profile.self.cycle= s-pp.__list_del_entry_valid > 0.34 =B1 5% -0.3 0.07 =B1 29% perf-profile.self.cycle= s-pp.schedule > 0.35 =B1 9% -0.3 0.08 =B1 10% perf-profile.self.cycle= s-pp.rb_insert_color > 0.40 =B1 5% -0.3 0.14 =B1 16% perf-profile.self.cycle= s-pp.select_task_rq_fair > 0.33 =B1 6% -0.3 0.08 =B1 16% perf-profile.self.cycle= s-pp.check_preempt_wakeup > 0.33 =B1 8% -0.2 0.10 =B1 16% perf-profile.self.cycle= s-pp.select_task_rq > 0.36 =B1 3% -0.2 0.13 =B1 16% perf-profile.self.cycle= s-pp.native_sched_clock > 0.32 =B1 7% -0.2 0.10 =B1 14% perf-profile.self.cycle= s-pp.finish_task_switch > 0.32 =B1 4% -0.2 0.11 =B1 13% perf-profile.self.cycle= s-pp.dequeue_entity > 0.32 =B1 8% -0.2 0.12 =B1 10% perf-profile.self.cycle= s-pp.__list_add_valid > 0.23 =B1 5% -0.2 0.03 =B1103% perf-profile.self.cycle= s-pp.resched_curr > 0.27 =B1 6% -0.2 0.07 =B1 21% perf-profile.self.cycle= s-pp.rb_erase > 0.27 =B1 5% -0.2 0.08 =B1 20% perf-profile.self.cycle= s-pp.__wrgsbase_inactive > 0.28 =B1 13% -0.2 0.09 =B1 12% perf-profile.self.cycle= s-pp.check_preempt_curr > 0.30 =B1 13% -0.2 0.12 =B1 7% perf-profile.self.cycle= s-pp.ttwu_queue_wakelist > 0.24 =B1 5% -0.2 0.06 =B1 19% perf-profile.self.cycle= s-pp.set_next_entity > 0.21 =B1 34% -0.2 0.04 =B1 71% perf-profile.self.cycle= s-pp.__flush_smp_call_function_queue > 0.25 =B1 5% -0.2 0.08 =B1 16% perf-profile.self.cycle= s-pp.rcu_note_context_switch > 0.19 =B1 26% -0.1 0.04 =B1 73% perf-profile.self.cycle= s-pp.perf_trace_sched_stat_runtime > 0.20 =B1 8% -0.1 0.06 =B1 13% perf-profile.self.cycle= s-pp.ttwu_do_activate > 0.17 =B1 8% -0.1 0.03 =B1100% perf-profile.self.cycle= s-pp.rb_next > 0.22 =B1 23% -0.1 0.09 =B1 31% perf-profile.self.cycle= s-pp.migrate_task_rq_fair > 0.15 =B1 32% -0.1 0.03 =B1100% perf-profile.self.cycle= s-pp.llist_reverse_order > 0.16 =B1 8% -0.1 0.06 =B1 14% perf-profile.self.cycle= s-pp.wake_affine > 0.10 =B1 31% -0.1 0.03 =B1100% perf-profile.self.cycle= s-pp.sched_ttwu_pending > 0.14 =B1 5% -0.1 0.07 =B1 20% perf-profile.self.cycle= s-pp.read@plt > 0.32 =B1 8% -0.1 0.26 =B1 3% perf-profile.self.cycle= s-pp.anon_pipe_buf_release > 0.10 =B1 6% -0.1 0.04 =B1 45% perf-profile.self.cycle= s-pp.__wake_up_common_lock > 0.10 =B1 9% -0.0 0.07 =B1 8% perf-profile.self.cycle= s-pp.__mark_inode_dirty > 0.00 +0.1 0.06 =B1 11% perf-profile.self.cycles-= pp.free_unref_page > 0.00 +0.1 0.06 =B1 6% perf-profile.self.cycles-= pp.__alloc_pages > 0.00 +0.1 0.06 =B1 11% perf-profile.self.cycles-= pp._raw_spin_trylock > 0.00 +0.1 0.06 =B1 7% perf-profile.self.cycles-= pp.uncharge_folio > 0.00 +0.1 0.06 =B1 11% perf-profile.self.cycles-= pp.uncharge_batch > 0.00 +0.1 0.07 =B1 10% perf-profile.self.cycles-= pp.menu_select > 0.00 +0.1 0.08 =B1 14% perf-profile.self.cycles-= pp.__get_obj_cgroup_from_memcg > 0.00 +0.1 0.08 =B1 7% perf-profile.self.cycles-= pp.__memcg_kmem_charge_page > 0.00 +0.1 0.10 =B1 10% perf-profile.self.cycles-= pp.osq_unlock > 0.07 =B1 5% +0.1 0.17 =B1 8% perf-profile.self.cycle= s-pp.copyin > 0.00 +0.1 0.11 =B1 11% perf-profile.self.cycles-= pp.__mod_memcg_state > 0.13 =B1 8% +0.1 0.24 =B1 6% perf-profile.self.cycle= s-pp.rcu_all_qs > 0.14 =B1 5% +0.1 0.28 =B1 5% perf-profile.self.cycle= s-pp.__x64_sys_write > 0.07 =B1 10% +0.1 0.21 =B1 5% perf-profile.self.cycle= s-pp.page_copy_sane > 0.13 =B1 12% +0.1 0.28 =B1 9% perf-profile.self.cycle= s-pp.__x64_sys_read > 0.00 +0.2 0.15 =B1 10% perf-profile.self.cycles-= pp.propagate_protected_usage > 0.18 =B1 9% +0.2 0.33 =B1 4% perf-profile.self.cycle= s-pp.syscall_exit_to_user_mode_prepare > 0.07 =B1 8% +0.2 0.23 =B1 5% perf-profile.self.cycle= s-pp.rw_verify_area > 0.08 =B1 34% +0.2 0.24 =B1 27% perf-profile.self.cycle= s-pp.mnt_user_ns > 0.13 =B1 5% +0.2 0.31 =B1 7% perf-profile.self.cycle= s-pp.kill_fasync > 0.21 =B1 8% +0.2 0.39 =B1 5% perf-profile.self.cycle= s-pp.__might_fault > 0.06 =B1 13% +0.2 0.26 =B1 9% perf-profile.self.cycle= s-pp.copyout > 0.10 =B1 11% +0.2 0.31 =B1 8% perf-profile.self.cycle= s-pp.ktime_get_coarse_real_ts64 > 0.26 =B1 13% +0.2 0.49 =B1 6% perf-profile.self.cycle= s-pp.inode_needs_update_time > 0.23 =B1 8% +0.2 0.47 =B1 5% perf-profile.self.cycle= s-pp.copy_page_from_iter > 0.14 =B1 7% +0.2 0.38 =B1 6% perf-profile.self.cycle= s-pp.file_update_time > 0.36 =B1 7% +0.3 0.62 =B1 4% perf-profile.self.cycle= s-pp.ksys_read > 0.54 =B1 13% +0.3 0.80 =B1 4% perf-profile.self.cycle= s-pp._copy_from_iter > 0.15 =B1 5% +0.3 0.41 =B1 8% perf-profile.self.cycle= s-pp.touch_atime > 0.14 =B1 5% +0.3 0.40 =B1 6% perf-profile.self.cycle= s-pp.__cond_resched > 0.18 =B1 5% +0.3 0.47 =B1 4% perf-profile.self.cycle= s-pp.syscall_exit_to_user_mode > 0.16 =B1 8% +0.3 0.46 =B1 6% perf-profile.self.cycle= s-pp.syscall_enter_from_user_mode > 0.16 =B1 9% +0.3 0.47 =B1 6% perf-profile.self.cycle= s-pp.__fdget_pos > 1.79 =B1 8% +0.3 2.12 =B1 3% perf-profile.self.cycle= s-pp.pipe_read > 0.10 =B1 8% +0.3 0.43 =B1 17% perf-profile.self.cycle= s-pp.fsnotify_perm > 0.20 =B1 4% +0.4 0.55 =B1 5% perf-profile.self.cycle= s-pp.ksys_write > 0.05 =B1 76% +0.4 0.46 =B1 27% perf-profile.self.cycle= s-pp.queue_event > 0.32 =B1 6% +0.4 0.73 =B1 6% perf-profile.self.cycle= s-pp.exit_to_user_mode_prepare > 0.21 =B1 9% +0.4 0.62 =B1 6% perf-profile.self.cycle= s-pp.entry_SYSCALL_64_safe_stack > 0.79 =B1 8% +0.4 1.22 =B1 4% perf-profile.self.cycle= s-pp.__might_sleep > 0.44 =B1 5% +0.4 0.88 =B1 7% perf-profile.self.cycle= s-pp.do_syscall_64 > 0.26 =B1 8% +0.4 0.70 =B1 4% perf-profile.self.cycle= s-pp.atime_needs_update > 0.42 =B1 7% +0.5 0.88 =B1 5% perf-profile.self.cycle= s-pp.__get_task_ioprio > 0.28 =B1 12% +0.5 0.75 =B1 5% perf-profile.self.cycle= s-pp.copy_page_to_iter > 0.19 =B1 6% +0.5 0.68 =B1 10% perf-profile.self.cycle= s-pp.security_file_permission > 0.31 =B1 8% +0.5 0.83 =B1 5% perf-profile.self.cycle= s-pp.aa_file_perm > 0.05 =B1 46% +0.5 0.59 =B1 8% perf-profile.self.cycle= s-pp.osq_lock > 0.30 =B1 7% +0.5 0.85 =B1 6% perf-profile.self.cycle= s-pp._copy_to_iter > 0.00 +0.6 0.59 =B1 6% perf-profile.self.cycles-= pp.poll_idle > 0.13 =B1 20% +0.7 0.81 =B1 6% perf-profile.self.cycle= s-pp.mutex_spin_on_owner > 0.38 =B1 9% +0.7 1.12 =B1 5% perf-profile.self.cycle= s-pp.current_time > 0.08 =B1 59% +0.8 0.82 =B1 11% perf-profile.self.cycle= s-pp.intel_idle_irq > 0.92 =B1 6% +0.8 1.72 =B1 4% perf-profile.self.cycle= s-pp.entry_SYSCALL_64_after_hwframe > 0.01 =B1223% +0.8 0.82 =B1 6% perf-profile.self.cycle= s-pp.page_counter_uncharge > 0.86 =B1 7% +1.1 1.91 =B1 4% perf-profile.self.cycle= s-pp.vfs_read > 1.07 =B1 6% +1.1 2.14 =B1 5% perf-profile.self.cycle= s-pp.__fget_light > 0.67 =B1 7% +1.1 1.74 =B1 6% perf-profile.self.cycle= s-pp.vfs_write > 0.15 =B1 12% +1.1 1.28 =B1 7% perf-profile.self.cycle= s-pp.__mutex_lock > 1.09 =B1 6% +1.1 2.22 =B1 5% perf-profile.self.cycle= s-pp.__libc_read > 0.62 =B1 6% +1.2 1.79 =B1 5% perf-profile.self.cycle= s-pp.syscall_return_via_sysret > 1.16 =B1 8% +1.2 2.38 =B1 4% perf-profile.self.cycle= s-pp.__might_resched > 0.91 =B1 7% +1.3 2.20 =B1 5% perf-profile.self.cycle= s-pp.__libc_write > 0.59 =B1 8% +1.3 1.93 =B1 6% perf-profile.self.cycle= s-pp.__entry_text_start > 1.27 =B1 7% +1.7 3.00 =B1 6% perf-profile.self.cycle= s-pp.apparmor_file_permission > 0.99 =B1 8% +2.0 2.98 =B1 5% perf-profile.self.cycle= s-pp.entry_SYSRETQ_unsafe_stack > 1.74 =B1 8% +3.4 5.15 =B1 6% perf-profile.self.cycle= s-pp.pipe_write > 2.98 =B1 8% +3.7 6.64 =B1 5% perf-profile.self.cycle= s-pp.copy_user_enhanced_fast_string > 2.62 =B1 10% +4.8 7.38 =B1 5% perf-profile.self.cycle= s-pp.mutex_lock > 2.20 =B1 10% +5.1 7.30 =B1 6% perf-profile.self.cycle= s-pp.mutex_unlock > = > = > *************************************************************************= ************************** > lkp-skl-fpga01: 104 threads 2 sockets (Skylake) with 192G memory > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > compiler/cpufreq_governor/ipc/iterations/kconfig/mode/nr_threads/rootfs/t= box_group/testcase: > gcc-11/performance/socket/4/x86_64-rhel-8.3/process/100%/debian-11.1-x8= 6_64-20220510.cgz/lkp-skl-fpga01/hackbench > = > commit: > a2e90611b9 ("sched/fair: Remove capacity inversion detection") > 829c1651e9 ("sched/fair: sanitize vruntime of entity being placed") > = > a2e90611b9f425ad 829c1651e9c4a6f78398d3e6765 > ---------------- --------------------------- > %stddev %change %stddev > \ | \ > 177139 -8.1% 162815 hackbench.throughput > 174484 -18.8% 141618 =B1 2% hackbench.throughput_avg > 177139 -8.1% 162815 hackbench.throughput_best > 168530 -37.3% 105615 =B1 3% hackbench.throughput_worst > 281.38 +23.1% 346.39 =B1 2% hackbench.time.elapsed_ti= me > 281.38 +23.1% 346.39 =B1 2% hackbench.time.elapsed_ti= me.max > 1.053e+08 =B1 2% +688.4% 8.302e+08 =B1 9% hackbench.time.involunt= ary_context_switches > 21992 +27.8% 28116 =B1 2% hackbench.time.system_time > 6652 +8.2% 7196 hackbench.time.user_time > 3.482e+08 +289.2% 1.355e+09 =B1 9% hackbench.time.voluntary_= context_switches > 2110813 =B1 5% +21.6% 2565791 =B1 3% cpuidle..usage > 333.95 +19.5% 399.05 uptime.boot > 0.03 -0.0 0.03 mpstat.cpu.all.soft% > 22.68 -2.9 19.77 mpstat.cpu.all.usr% > 561083 =B1 10% +45.5% 816171 =B1 12% numa-numastat.node0.loc= al_node > 614314 =B1 9% +36.9% 841173 =B1 12% numa-numastat.node0.num= a_hit > 1393279 =B1 7% -16.8% 1158997 =B1 2% numa-numastat.node1.loc= al_node > 1443679 =B1 5% -14.9% 1229074 =B1 3% numa-numastat.node1.num= a_hit > 4129900 =B1 8% -23.0% 3181115 vmstat.memory.cache > 1731 +30.8% 2265 vmstat.procs.r > 1598044 +290.3% 6237840 =B1 7% vmstat.system.cs > 320762 +60.5% 514672 =B1 8% vmstat.system.in > 962111 =B1 6% +46.0% 1404646 =B1 7% turbostat.C1 > 233987 =B1 5% +51.2% 353892 turbostat.C1E > 91515563 +97.3% 1.806e+08 =B1 10% turbostat.IRQ > 448466 =B1 14% -34.2% 294934 =B1 5% turbostat.POLL > 34.60 -7.3% 32.07 turbostat.RAMWatt > 514028 =B1 2% -14.0% 442125 =B1 2% meminfo.AnonPages > 4006312 =B1 8% -23.9% 3047078 meminfo.Cached > 3321064 =B1 10% -32.7% 2236362 =B1 2% meminfo.Committed_AS > 1714752 =B1 21% -60.3% 680479 =B1 8% meminfo.Inactive > 1714585 =B1 21% -60.3% 680305 =B1 8% meminfo.Inactive(anon) > 757124 =B1 18% -67.2% 248485 =B1 27% meminfo.Mapped > 6476123 =B1 6% -19.4% 5220738 meminfo.Memused > 1275724 =B1 26% -75.2% 316896 =B1 15% meminfo.Shmem > 6806047 =B1 3% -13.3% 5901974 meminfo.max_used_kB > 161311 =B1 23% +31.7% 212494 =B1 5% numa-meminfo.node0.Anon= Pages > 165693 =B1 22% +30.5% 216264 =B1 5% numa-meminfo.node0.Inac= tive > 165563 =B1 22% +30.6% 216232 =B1 5% numa-meminfo.node0.Inac= tive(anon) > 140638 =B1 19% -36.7% 89034 =B1 11% numa-meminfo.node0.Mapp= ed > 352173 =B1 14% -35.3% 227805 =B1 8% numa-meminfo.node1.Anon= Pages > 501396 =B1 11% -22.6% 388042 =B1 5% numa-meminfo.node1.Anon= Pages.max > 1702242 =B1 43% -77.8% 378325 =B1 22% numa-meminfo.node1.File= Pages > 1540803 =B1 25% -70.4% 455592 =B1 13% numa-meminfo.node1.Inac= tive > 1540767 =B1 25% -70.4% 455451 =B1 13% numa-meminfo.node1.Inac= tive(anon) > 612123 =B1 18% -74.9% 153752 =B1 37% numa-meminfo.node1.Mapp= ed > 3085231 =B1 24% -53.9% 1420940 =B1 14% numa-meminfo.node1.MemU= sed > 254052 =B1 4% -19.1% 205632 =B1 21% numa-meminfo.node1.SUnr= eclaim > 1259640 =B1 27% -75.9% 303123 =B1 15% numa-meminfo.node1.Shmem > 304597 =B1 7% -20.2% 242920 =B1 17% numa-meminfo.node1.Slab > 40345 =B1 23% +31.5% 53054 =B1 5% numa-vmstat.node0.nr_an= on_pages > 41412 =B1 22% +30.4% 53988 =B1 5% numa-vmstat.node0.nr_in= active_anon > 35261 =B1 19% -36.9% 22256 =B1 12% numa-vmstat.node0.nr_ma= pped > 41412 =B1 22% +30.4% 53988 =B1 5% numa-vmstat.node0.nr_zo= ne_inactive_anon > 614185 =B1 9% +36.9% 841065 =B1 12% numa-vmstat.node0.numa_= hit > 560955 =B1 11% +45.5% 816063 =B1 12% numa-vmstat.node0.numa_= local > 88129 =B1 14% -35.2% 57097 =B1 8% numa-vmstat.node1.nr_an= on_pages > 426425 =B1 43% -77.9% 94199 =B1 22% numa-vmstat.node1.nr_fi= le_pages > 386166 =B1 25% -70.5% 113880 =B1 13% numa-vmstat.node1.nr_in= active_anon > 153658 =B1 18% -75.3% 38021 =B1 37% numa-vmstat.node1.nr_ma= pped > 315775 =B1 27% -76.1% 75399 =B1 16% numa-vmstat.node1.nr_sh= mem > 63411 =B1 4% -18.6% 51593 =B1 21% numa-vmstat.node1.nr_sl= ab_unreclaimable > 386166 =B1 25% -70.5% 113880 =B1 13% numa-vmstat.node1.nr_zo= ne_inactive_anon > 1443470 =B1 5% -14.9% 1228740 =B1 3% numa-vmstat.node1.numa_= hit > 1393069 =B1 7% -16.8% 1158664 =B1 2% numa-vmstat.node1.numa_= local > 128457 =B1 2% -14.0% 110530 =B1 3% proc-vmstat.nr_anon_pag= es > 999461 =B1 8% -23.8% 761774 proc-vmstat.nr_file_pages > 426485 =B1 21% -60.1% 170237 =B1 9% proc-vmstat.nr_inactive= _anon > 82464 -2.6% 80281 proc-vmstat.nr_kernel_stack > 187777 =B1 18% -66.9% 62076 =B1 28% proc-vmstat.nr_mapped > 316813 =B1 27% -75.0% 79228 =B1 16% proc-vmstat.nr_shmem > 31469 -2.0% 30840 proc-vmstat.nr_slab_reclaim= able > 117889 -8.4% 108036 proc-vmstat.nr_slab_unrecla= imable > 426485 =B1 21% -60.1% 170237 =B1 9% proc-vmstat.nr_zone_ina= ctive_anon > 187187 =B1 12% -43.5% 105680 =B1 9% proc-vmstat.numa_hint_f= aults > 128363 =B1 15% -61.5% 49371 =B1 19% proc-vmstat.numa_hint_f= aults_local > 47314 =B1 22% +39.2% 65863 =B1 13% proc-vmstat.numa_pages_= migrated > 457026 =B1 9% -18.1% 374188 =B1 13% proc-vmstat.numa_pte_up= dates > 2586600 =B1 3% +27.7% 3302787 =B1 8% proc-vmstat.pgalloc_nor= mal > 1589970 -6.2% 1491838 proc-vmstat.pgfault > 2347186 =B1 10% +37.7% 3232369 =B1 8% proc-vmstat.pgfree > 47314 =B1 22% +39.2% 65863 =B1 13% proc-vmstat.pgmigrate_s= uccess > 112713 +7.0% 120630 =B1 3% proc-vmstat.pgreuse > 2189056 +22.2% 2674944 =B1 2% proc-vmstat.unevictable_p= gs_scanned > 14.08 =B1 2% +29.3% 18.20 =B1 5% sched_debug.cfs_rq:/.h_= nr_running.avg > 0.80 =B1 14% +179.2% 2.23 =B1 24% sched_debug.cfs_rq:/.h_= nr_running.min > 245.23 =B1 12% -19.7% 196.97 =B1 6% sched_debug.cfs_rq:/.lo= ad_avg.max > 2.27 =B1 16% +75.0% 3.97 =B1 4% sched_debug.cfs_rq:/.lo= ad_avg.min > 45.77 =B1 16% -17.8% 37.60 =B1 6% sched_debug.cfs_rq:/.lo= ad_avg.stddev > 11842707 +39.9% 16567992 sched_debug.cfs_rq:/.min_vr= untime.avg > 13773080 =B1 3% +113.9% 29460281 =B1 7% sched_debug.cfs_rq:/.mi= n_vruntime.max > 11423218 +30.3% 14885830 sched_debug.cfs_rq:/.min_vr= untime.min > 301190 =B1 12% +439.9% 1626088 =B1 10% sched_debug.cfs_rq:/.mi= n_vruntime.stddev > 203.83 -16.3% 170.67 sched_debug.cfs_rq:/.remove= d.load_avg.max > 14330 =B1 3% +30.9% 18756 =B1 5% sched_debug.cfs_rq:/.ru= nnable_avg.avg > 25115 =B1 4% +15.5% 28999 =B1 6% sched_debug.cfs_rq:/.ru= nnable_avg.max > 3811 =B1 11% +68.0% 6404 =B1 21% sched_debug.cfs_rq:/.ru= nnable_avg.min > 3818 =B1 6% +15.3% 4404 =B1 7% sched_debug.cfs_rq:/.ru= nnable_avg.stddev > -849635 +410.6% -4338612 sched_debug.cfs_rq:/.spread= 0.avg > 1092373 =B1 54% +691.1% 8641673 =B1 21% sched_debug.cfs_rq:/.sp= read0.max > -1263082 +378.1% -6038905 sched_debug.cfs_rq:/.spread= 0.min > 300764 =B1 12% +441.8% 1629507 =B1 9% sched_debug.cfs_rq:/.sp= read0.stddev > 1591 =B1 4% -11.1% 1413 =B1 3% sched_debug.cfs_rq:/.ut= il_avg.max > 288.90 =B1 11% +64.5% 475.23 =B1 13% sched_debug.cfs_rq:/.ut= il_avg.min > 240.33 =B1 2% -32.1% 163.09 =B1 3% sched_debug.cfs_rq:/.ut= il_avg.stddev > 494.27 =B1 3% +41.6% 699.85 =B1 3% sched_debug.cfs_rq:/.ut= il_est_enqueued.avg > 11.23 =B1 54% +634.1% 82.47 =B1 22% sched_debug.cfs_rq:/.ut= il_est_enqueued.min > 174576 +20.7% 210681 sched_debug.cpu.clock.avg > 174926 +21.2% 211944 sched_debug.cpu.clock.max > 174164 +20.3% 209436 sched_debug.cpu.clock.min > 230.84 =B1 33% +226.1% 752.67 =B1 20% sched_debug.cpu.clock.s= tddev > 172836 +20.6% 208504 sched_debug.cpu.clock_task.= avg > 173552 +21.0% 210079 sched_debug.cpu.clock_task.= max > 156807 +22.3% 191789 sched_debug.cpu.clock_task.= min > 1634 +17.1% 1914 =B1 5% sched_debug.cpu.clock_tas= k.stddev > 0.00 =B1 32% +220.1% 0.00 =B1 20% sched_debug.cpu.next_ba= lance.stddev > 14.12 =B1 2% +28.7% 18.18 =B1 5% sched_debug.cpu.nr_runn= ing.avg > 0.73 =B1 25% +213.6% 2.30 =B1 24% sched_debug.cpu.nr_runn= ing.min > 1810086 +461.3% 10159215 =B1 10% sched_debug.cpu.nr_switch= es.avg > 2315994 =B1 3% +515.6% 14258195 =B1 9% sched_debug.cpu.nr_swit= ches.max > 1529863 +380.3% 7348324 =B1 9% sched_debug.cpu.nr_switch= es.min > 167487 =B1 18% +770.8% 1458519 =B1 21% sched_debug.cpu.nr_swit= ches.stddev > 174149 +20.2% 209410 sched_debug.cpu_clk > 170980 +20.6% 206240 sched_debug.ktime > 174896 +20.2% 210153 sched_debug.sched_clk > 7.35 +24.9% 9.18 =B1 4% perf-stat.i.MPKI > 1.918e+10 +14.4% 2.194e+10 perf-stat.i.branch-instruct= ions > 2.16 -0.1 2.09 perf-stat.i.branch-miss-rat= e% > 4.133e+08 +6.6% 4.405e+08 perf-stat.i.branch-misses > 23.08 -9.2 13.86 =B1 7% perf-stat.i.cache-miss-ra= te% > 1.714e+08 -37.2% 1.076e+08 =B1 3% perf-stat.i.cache-misses > 7.497e+08 +33.7% 1.002e+09 =B1 5% perf-stat.i.cache-referen= ces > 1636365 +382.4% 7893858 =B1 5% perf-stat.i.context-switc= hes > 2.74 -6.8% 2.56 perf-stat.i.cpi > 131725 +288.0% 511159 =B1 10% perf-stat.i.cpu-migrations > 1672 +160.8% 4361 =B1 4% perf-stat.i.cycles-betwee= n-cache-misses > 0.49 +0.6 1.11 =B1 5% perf-stat.i.dTLB-load-mis= s-rate% > 1.417e+08 +158.7% 3.665e+08 =B1 5% perf-stat.i.dTLB-load-mis= ses > 2.908e+10 +9.1% 3.172e+10 perf-stat.i.dTLB-loads > 0.12 =B1 4% +0.1 0.20 =B1 4% perf-stat.i.dTLB-store-= miss-rate% > 20805655 =B1 4% +90.9% 39716345 =B1 4% perf-stat.i.dTLB-store-= misses > 1.755e+10 +8.6% 1.907e+10 perf-stat.i.dTLB-stores > 29.04 +3.6 32.62 =B1 2% perf-stat.i.iTLB-load-mis= s-rate% > 56676082 +60.4% 90917582 =B1 3% perf-stat.i.iTLB-load-mis= ses > 1.381e+08 +30.6% 1.804e+08 perf-stat.i.iTLB-loads > 1.03e+11 +10.5% 1.139e+11 perf-stat.i.instructions > 1840 -21.1% 1451 =B1 4% perf-stat.i.instructions-= per-iTLB-miss > 0.37 +10.9% 0.41 perf-stat.i.ipc > 1084 -4.5% 1035 =B1 2% perf-stat.i.metric.K/sec > 640.69 +10.3% 706.44 perf-stat.i.metric.M/sec > 5249 -9.3% 4762 =B1 3% perf-stat.i.minor-faults > 23.57 +18.7 42.30 =B1 8% perf-stat.i.node-load-mis= s-rate% > 40174555 -45.0% 22109431 =B1 10% perf-stat.i.node-loads > 8.84 =B1 2% +24.5 33.30 =B1 10% perf-stat.i.node-store-= miss-rate% > 2912322 +60.3% 4667137 =B1 16% perf-stat.i.node-store-mi= sses > 34046752 -50.6% 16826621 =B1 9% perf-stat.i.node-stores > 5278 -9.2% 4791 =B1 3% perf-stat.i.page-faults > 7.24 +12.1% 8.12 =B1 4% perf-stat.overall.MPKI > 2.15 -0.1 2.05 perf-stat.overall.branch-mi= ss-rate% > 22.92 -9.5 13.41 =B1 7% perf-stat.overall.cache-m= iss-rate% > 2.73 -6.3% 2.56 perf-stat.overall.cpi > 1644 +43.4% 2358 =B1 3% perf-stat.overall.cycles-= between-cache-misses > 0.48 +0.5 0.99 =B1 4% perf-stat.overall.dTLB-lo= ad-miss-rate% > 0.12 =B1 4% +0.1 0.19 =B1 4% perf-stat.overall.dTLB-= store-miss-rate% > 29.06 +2.9 32.01 =B1 2% perf-stat.overall.iTLB-lo= ad-miss-rate% > 1826 -26.6% 1340 =B1 4% perf-stat.overall.instruc= tions-per-iTLB-miss > 0.37 +6.8% 0.39 perf-stat.overall.ipc > 22.74 +6.8 29.53 =B1 13% perf-stat.overall.node-lo= ad-miss-rate% > 7.63 +8.4 16.02 =B1 20% perf-stat.overall.node-st= ore-miss-rate% > 1.915e+10 +9.0% 2.088e+10 perf-stat.ps.branch-instruc= tions > 4.119e+08 +3.9% 4.282e+08 perf-stat.ps.branch-misses > 1.707e+08 -30.5% 1.186e+08 =B1 3% perf-stat.ps.cache-misses > 7.446e+08 +19.2% 8.874e+08 =B1 4% perf-stat.ps.cache-refere= nces > 1611874 +289.1% 6271376 =B1 7% perf-stat.ps.context-swit= ches > 127362 +189.0% 368041 =B1 11% perf-stat.ps.cpu-migratio= ns > 1.407e+08 +116.2% 3.042e+08 =B1 5% perf-stat.ps.dTLB-load-mi= sses > 2.901e+10 +5.4% 3.057e+10 perf-stat.ps.dTLB-loads > 20667480 =B1 4% +66.8% 34473793 =B1 4% perf-stat.ps.dTLB-store= -misses > 1.751e+10 +5.1% 1.84e+10 perf-stat.ps.dTLB-stores > 56310692 +45.0% 81644183 =B1 4% perf-stat.ps.iTLB-load-mi= sses > 1.375e+08 +26.1% 1.733e+08 perf-stat.ps.iTLB-loads > 1.028e+11 +6.3% 1.093e+11 perf-stat.ps.instructions > 4929 -24.5% 3723 =B1 2% perf-stat.ps.minor-faults > 40134633 -32.9% 26946247 =B1 9% perf-stat.ps.node-loads > 2805073 +39.5% 3914304 =B1 16% perf-stat.ps.node-store-m= isses > 33938259 -38.9% 20726382 =B1 8% perf-stat.ps.node-stores > 4952 -24.5% 3741 =B1 2% perf-stat.ps.page-faults > 2.911e+13 +30.9% 3.809e+13 =B1 2% perf-stat.total.instructi= ons > 15.30 =B1 4% -8.6 6.66 =B1 5% perf-profile.calltrace.= cycles-pp.sock_alloc_send_pskb.unix_stream_sendmsg.sock_sendmsg.sock_write_= iter.vfs_write > 13.84 =B1 6% -7.9 5.98 =B1 6% perf-profile.calltrace.= cycles-pp.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg.soc= k_sendmsg.sock_write_iter > 13.61 =B1 6% -7.8 5.84 =B1 6% perf-profile.calltrace.= cycles-pp.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream= _sendmsg.sock_sendmsg > 9.00 =B1 2% -5.5 3.48 =B1 4% perf-profile.calltrace.= cycles-pp.consume_skb.unix_stream_read_generic.unix_stream_recvmsg.sock_rea= d_iter.vfs_read > 6.44 =B1 4% -4.3 2.14 =B1 6% perf-profile.calltrace.= cycles-pp.skb_release_data.consume_skb.unix_stream_read_generic.unix_stream= _recvmsg.sock_read_iter > 5.83 =B1 8% -3.4 2.44 =B1 5% perf-profile.calltrace.= cycles-pp.kmem_cache_alloc_node.__alloc_skb.alloc_skb_with_frags.sock_alloc= _send_pskb.unix_stream_sendmsg > 5.81 =B1 6% -3.3 2.48 =B1 6% perf-profile.calltrace.= cycles-pp.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_= pskb.unix_stream_sendmsg > 5.50 =B1 7% -3.2 2.32 =B1 6% perf-profile.calltrace.= cycles-pp.__kmalloc_node_track_caller.kmalloc_reserve.__alloc_skb.alloc_skb= _with_frags.sock_alloc_send_pskb > 5.07 =B1 8% -3.0 2.04 =B1 6% perf-profile.calltrace.= cycles-pp.__kmem_cache_alloc_node.__kmalloc_node_track_caller.kmalloc_reser= ve.__alloc_skb.alloc_skb_with_frags > 6.22 =B1 2% -2.9 3.33 =B1 3% perf-profile.calltrace.= cycles-pp.unix_stream_read_actor.unix_stream_read_generic.unix_stream_recvm= sg.sock_read_iter.vfs_read > 6.17 =B1 2% -2.9 3.30 =B1 3% perf-profile.calltrace.= cycles-pp.skb_copy_datagram_iter.unix_stream_read_actor.unix_stream_read_ge= neric.unix_stream_recvmsg.sock_read_iter > 6.11 =B1 2% -2.9 3.24 =B1 3% perf-profile.calltrace.= cycles-pp.__skb_datagram_iter.skb_copy_datagram_iter.unix_stream_read_actor= .unix_stream_read_generic.unix_stream_recvmsg > 50.99 -2.6 48.39 perf-profile.calltrace.cycl= es-pp.__libc_read > 5.66 =B1 3% -2.3 3.35 =B1 3% perf-profile.calltrace.= cycles-pp.syscall_return_via_sysret.__libc_read > 5.52 =B1 3% -2.3 3.27 =B1 3% perf-profile.calltrace.= cycles-pp.syscall_return_via_sysret.__libc_write > 3.14 =B1 2% -1.7 1.42 =B1 4% perf-profile.calltrace.= cycles-pp._copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.unix_str= eam_read_actor.unix_stream_read_generic > 2.73 =B1 2% -1.6 1.15 =B1 4% perf-profile.calltrace.= cycles-pp.copyout._copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.= unix_stream_read_actor > 2.59 =B1 2% -1.5 1.07 =B1 4% perf-profile.calltrace.= cycles-pp.copy_user_enhanced_fast_string.copyout._copy_to_iter.__skb_datagr= am_iter.skb_copy_datagram_iter > 2.72 =B1 3% -1.4 1.34 =B1 6% perf-profile.calltrace.= cycles-pp.kmem_cache_free.unix_stream_read_generic.unix_stream_recvmsg.sock= _read_iter.vfs_read > 41.50 -1.2 40.27 perf-profile.calltrace.cycl= es-pp.entry_SYSCALL_64_after_hwframe.__libc_read > 2.26 =B1 4% -1.1 1.12 perf-profile.calltrace.cy= cles-pp.skb_release_head_state.consume_skb.unix_stream_read_generic.unix_st= ream_recvmsg.sock_read_iter > 2.76 =B1 3% -1.1 1.63 =B1 3% perf-profile.calltrace.= cycles-pp.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_c= opy_datagram_iter.unix_stream_read_actor > 2.84 =B1 3% -1.1 1.71 =B1 2% perf-profile.calltrace.= cycles-pp.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.un= ix_stream_read_actor.unix_stream_read_generic > 2.20 =B1 4% -1.1 1.08 perf-profile.calltrace.cy= cles-pp.unix_destruct_scm.skb_release_head_state.consume_skb.unix_stream_re= ad_generic.unix_stream_recvmsg > 2.98 =B1 2% -1.1 1.90 =B1 6% perf-profile.calltrace.= cycles-pp.skb_copy_datagram_from_iter.unix_stream_sendmsg.sock_sendmsg.sock= _write_iter.vfs_write > 1.99 =B1 4% -1.1 0.92 =B1 2% perf-profile.calltrace.= cycles-pp.sock_wfree.unix_destruct_scm.skb_release_head_state.consume_skb.u= nix_stream_read_generic > 2.10 =B1 3% -1.0 1.08 =B1 4% perf-profile.calltrace.= cycles-pp.check_heap_object.__check_object_size.simple_copy_to_iter.__skb_d= atagram_iter.skb_copy_datagram_iter > 2.08 =B1 4% -0.8 1.24 =B1 3% perf-profile.calltrace.= cycles-pp.__entry_text_start.__libc_write > 2.16 =B1 3% -0.7 1.47 perf-profile.calltrace.cy= cles-pp.__entry_text_start.__libc_read > 2.20 =B1 2% -0.7 1.52 =B1 3% perf-profile.calltrace.= cycles-pp.__kmem_cache_free.skb_release_data.consume_skb.unix_stream_read_g= eneric.unix_stream_recvmsg > 1.46 =B1 3% -0.6 0.87 =B1 8% perf-profile.calltrace.= cycles-pp._copy_from_iter.skb_copy_datagram_from_iter.unix_stream_sendmsg.s= ock_sendmsg.sock_write_iter > 4.82 =B1 2% -0.6 4.24 perf-profile.calltrace.cy= cles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwfr= ame.__libc_read > 1.31 =B1 2% -0.4 0.90 =B1 4% perf-profile.calltrace.= cycles-pp.__check_object_size.skb_copy_datagram_from_iter.unix_stream_sendm= sg.sock_sendmsg.sock_write_iter > 0.96 =B1 3% -0.4 0.57 =B1 10% perf-profile.calltrace.= cycles-pp.copyin._copy_from_iter.skb_copy_datagram_from_iter.unix_stream_se= ndmsg.sock_sendmsg > 1.14 =B1 3% -0.4 0.76 =B1 5% perf-profile.calltrace.= cycles-pp.memcg_slab_post_alloc_hook.kmem_cache_alloc_node.__alloc_skb.allo= c_skb_with_frags.sock_alloc_send_pskb > 0.99 =B1 3% -0.3 0.65 =B1 8% perf-profile.calltrace.= cycles-pp.memcg_slab_post_alloc_hook.__kmem_cache_alloc_node.__kmalloc_node= _track_caller.kmalloc_reserve.__alloc_skb > 1.30 =B1 4% -0.3 0.99 =B1 3% perf-profile.calltrace.= cycles-pp.sock_recvmsg.sock_read_iter.vfs_read.ksys_read.do_syscall_64 > 0.98 =B1 2% -0.3 0.69 =B1 3% perf-profile.calltrace.= cycles-pp.security_file_permission.vfs_write.ksys_write.do_syscall_64.entry= _SYSCALL_64_after_hwframe > 0.67 -0.2 0.42 =B1 50% perf-profile.calltrace.cy= cles-pp.check_heap_object.__check_object_size.skb_copy_datagram_from_iter.u= nix_stream_sendmsg.sock_sendmsg > 0.56 =B1 4% -0.2 0.32 =B1 81% perf-profile.calltrace.= cycles-pp.__fdget_pos.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwfram= e.__libc_read > 0.86 =B1 2% -0.2 0.63 =B1 3% perf-profile.calltrace.= cycles-pp.apparmor_file_permission.security_file_permission.vfs_write.ksys_= write.do_syscall_64 > 1.15 =B1 4% -0.2 0.93 =B1 4% perf-profile.calltrace.= cycles-pp.security_socket_recvmsg.sock_recvmsg.sock_read_iter.vfs_read.ksys= _read > 0.90 -0.2 0.69 =B1 3% perf-profile.calltrace.cy= cles-pp.get_obj_cgroup_from_current.kmem_cache_alloc_node.__alloc_skb.alloc= _skb_with_frags.sock_alloc_send_pskb > 1.23 =B1 3% -0.2 1.07 =B1 3% perf-profile.calltrace.= cycles-pp.security_socket_sendmsg.sock_sendmsg.sock_write_iter.vfs_write.ks= ys_write > 1.05 =B1 2% -0.2 0.88 =B1 2% perf-profile.calltrace.= cycles-pp.security_file_permission.vfs_read.ksys_read.do_syscall_64.entry_S= YSCALL_64_after_hwframe > 0.84 =B1 4% -0.2 0.68 =B1 4% perf-profile.calltrace.= cycles-pp.aa_sk_perm.security_socket_recvmsg.sock_recvmsg.sock_read_iter.vf= s_read > 0.88 -0.1 0.78 =B1 5% perf-profile.calltrace.cy= cles-pp.apparmor_file_permission.security_file_permission.vfs_read.ksys_rea= d.do_syscall_64 > 0.94 =B1 3% -0.1 0.88 =B1 4% perf-profile.calltrace.= cycles-pp.aa_sk_perm.security_socket_sendmsg.sock_sendmsg.sock_write_iter.v= fs_write > 0.62 =B1 2% +0.3 0.90 =B1 2% perf-profile.calltrace.= cycles-pp.mutex_lock.unix_stream_read_generic.unix_stream_recvmsg.sock_read= _iter.vfs_read > 0.00 +0.6 0.58 =B1 3% perf-profile.calltrace.cy= cles-pp._raw_spin_lock_irqsave.prepare_to_wait.unix_stream_data_wait.unix_s= tream_read_generic.unix_stream_recvmsg > 0.00 +0.6 0.61 =B1 6% perf-profile.calltrace.cy= cles-pp.update_load_avg.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.a= utoremove_wake_function > 0.00 +0.6 0.62 =B1 4% perf-profile.calltrace.cy= cles-pp.set_next_entity.pick_next_task_fair.__schedule.schedule.exit_to_use= r_mode_loop > 0.00 +0.7 0.67 =B1 11% perf-profile.calltrace.cy= cles-pp.update_load_avg.dequeue_entity.dequeue_task_fair.__schedule.schedule > 0.00 +0.7 0.67 =B1 7% perf-profile.calltrace.cy= cles-pp.__switch_to_asm.__libc_write > 0.00 +0.8 0.76 =B1 4% perf-profile.calltrace.cy= cles-pp.reweight_entity.dequeue_task_fair.__schedule.schedule.schedule_time= out > 0.00 +0.8 0.77 =B1 4% perf-profile.calltrace.cy= cles-pp.___perf_sw_event.prepare_task_switch.__schedule.schedule.schedule_t= imeout > 0.00 +0.8 0.77 =B1 8% perf-profile.calltrace.cy= cles-pp.put_prev_entity.pick_next_task_fair.__schedule.schedule.exit_to_use= r_mode_loop > 0.00 +0.8 0.81 =B1 5% perf-profile.calltrace.cy= cles-pp.prepare_task_switch.__schedule.schedule.exit_to_user_mode_loop.exit= _to_user_mode_prepare > 0.00 +0.8 0.81 =B1 5% perf-profile.calltrace.cy= cles-pp.check_preempt_wakeup.check_preempt_curr.ttwu_do_activate.try_to_wak= e_up.autoremove_wake_function > 0.00 +0.8 0.82 =B1 2% perf-profile.calltrace.cy= cles-pp.__switch_to_asm.__libc_read > 0.00 +0.8 0.82 =B1 3% perf-profile.calltrace.cy= cles-pp.reweight_entity.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.a= utoremove_wake_function > 0.00 +0.9 0.86 =B1 5% perf-profile.calltrace.cy= cles-pp.perf_trace_sched_wakeup_template.try_to_wake_up.autoremove_wake_fun= ction.__wake_up_common.__wake_up_common_lock > 0.00 +0.9 0.87 =B1 8% perf-profile.calltrace.cy= cles-pp.update_load_avg.enqueue_entity.enqueue_task_fair.ttwu_do_activate.t= ry_to_wake_up > 29.66 +0.9 30.58 perf-profile.calltrace.cycl= es-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.__l= ibc_write > 0.00 +1.0 0.95 =B1 3% perf-profile.calltrace.cy= cles-pp.set_next_entity.pick_next_task_fair.__schedule.schedule.schedule_ti= meout > 0.00 +1.0 0.98 =B1 4% perf-profile.calltrace.cy= cles-pp.check_preempt_curr.ttwu_do_activate.try_to_wake_up.autoremove_wake_= function.__wake_up_common > 0.00 +1.0 0.99 =B1 3% perf-profile.calltrace.cy= cles-pp.update_curr.enqueue_entity.enqueue_task_fair.ttwu_do_activate.try_t= o_wake_up > 0.00 +1.0 1.05 =B1 4% perf-profile.calltrace.cy= cles-pp.prepare_to_wait.unix_stream_data_wait.unix_stream_read_generic.unix= _stream_recvmsg.sock_read_iter > 0.00 +1.1 1.07 =B1 12% perf-profile.calltrace.cy= cles-pp.select_idle_sibling.select_task_rq_fair.select_task_rq.try_to_wake_= up.autoremove_wake_function > 27.81 =B1 2% +1.2 28.98 perf-profile.calltrace.cy= cles-pp.unix_stream_recvmsg.sock_read_iter.vfs_read.ksys_read.do_syscall_64 > 27.36 =B1 2% +1.2 28.59 perf-profile.calltrace.cy= cles-pp.unix_stream_read_generic.unix_stream_recvmsg.sock_read_iter.vfs_rea= d.ksys_read > 0.00 +1.5 1.46 =B1 2% perf-profile.calltrace.cy= cles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.try_to_wake_up.auto= remove_wake_function.__wake_up_common > 0.00 +1.6 1.55 =B1 4% perf-profile.calltrace.cy= cles-pp.prepare_task_switch.__schedule.schedule.schedule_timeout.unix_strea= m_data_wait > 0.00 +1.6 1.60 =B1 4% perf-profile.calltrace.cy= cles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.e= ntry_SYSCALL_64_after_hwframe.__libc_read > 27.58 +1.6 29.19 perf-profile.calltrace.cycl= es-pp.sock_write_iter.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_a= fter_hwframe > 0.00 +1.6 1.63 =B1 5% perf-profile.calltrace.cy= cles-pp.update_curr.dequeue_entity.dequeue_task_fair.__schedule.schedule > 0.00 +1.6 1.65 =B1 5% perf-profile.calltrace.cy= cles-pp.restore_fpregs_from_fpstate.switch_fpu_return.exit_to_user_mode_pre= pare.syscall_exit_to_user_mode.do_syscall_64 > 0.00 +1.7 1.66 =B1 6% perf-profile.calltrace.cy= cles-pp.switch_mm_irqs_off.__schedule.schedule.exit_to_user_mode_loop.exit_= to_user_mode_prepare > 0.00 +1.8 1.80 perf-profile.calltrace.cycl= es-pp._raw_spin_lock.try_to_wake_up.autoremove_wake_function.__wake_up_comm= on.__wake_up_common_lock > 0.00 +1.8 1.84 =B1 2% perf-profile.calltrace.cy= cles-pp.pick_next_task_fair.__schedule.schedule.schedule_timeout.unix_strea= m_data_wait > 0.00 +2.0 1.97 =B1 2% perf-profile.calltrace.cy= cles-pp.switch_mm_irqs_off.__schedule.schedule.schedule_timeout.unix_stream= _data_wait > 26.63 =B1 2% +2.0 28.61 perf-profile.calltrace.cy= cles-pp.sock_sendmsg.sock_write_iter.vfs_write.ksys_write.do_syscall_64 > 0.00 +2.0 2.01 =B1 6% perf-profile.calltrace.cy= cles-pp.pick_next_task_fair.__schedule.schedule.exit_to_user_mode_loop.exit= _to_user_mode_prepare > 0.00 +2.1 2.09 =B1 6% perf-profile.calltrace.cy= cles-pp.select_task_rq_fair.select_task_rq.try_to_wake_up.autoremove_wake_f= unction.__wake_up_common > 0.00 +2.1 2.11 =B1 5% perf-profile.calltrace.cy= cles-pp.switch_fpu_return.exit_to_user_mode_prepare.syscall_exit_to_user_mo= de.do_syscall_64.entry_SYSCALL_64_after_hwframe > 25.21 =B1 2% +2.2 27.43 perf-profile.calltrace.cy= cles-pp.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.vfs_write.ksys_wri= te > 0.00 +2.4 2.43 =B1 5% perf-profile.calltrace.cy= cles-pp.select_task_rq.try_to_wake_up.autoremove_wake_function.__wake_up_co= mmon.__wake_up_common_lock > 48.00 +2.7 50.69 perf-profile.calltrace.cycl= es-pp.__libc_write > 0.00 +2.9 2.87 =B1 5% perf-profile.calltrace.cy= cles-pp.dequeue_entity.dequeue_task_fair.__schedule.schedule.schedule_timeo= ut > 0.09 =B1223% +3.4 3.47 =B1 3% perf-profile.calltrace.= cycles-pp.enqueue_entity.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.= autoremove_wake_function > 39.07 +4.8 43.84 perf-profile.calltrace.cycl= es-pp.entry_SYSCALL_64_after_hwframe.__libc_write > 0.66 =B1 18% +5.0 5.62 =B1 4% perf-profile.calltrace.= cycles-pp.dequeue_task_fair.__schedule.schedule.schedule_timeout.unix_strea= m_data_wait > 4.73 +5.1 9.88 =B1 3% perf-profile.calltrace.cy= cles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwfr= ame.__libc_write > 0.66 =B1 20% +5.3 5.98 =B1 3% perf-profile.calltrace.= cycles-pp.enqueue_task_fair.ttwu_do_activate.try_to_wake_up.autoremove_wake= _function.__wake_up_common > 35.96 +5.7 41.68 perf-profile.calltrace.cycl= es-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__libc_write > 0.00 +6.0 6.02 =B1 6% perf-profile.calltrace.cy= cles-pp.__schedule.schedule.exit_to_user_mode_loop.exit_to_user_mode_prepar= e.syscall_exit_to_user_mode > 0.00 +6.2 6.18 =B1 6% perf-profile.calltrace.cy= cles-pp.schedule.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_e= xit_to_user_mode.do_syscall_64 > 0.00 +6.4 6.36 =B1 6% perf-profile.calltrace.cy= cles-pp.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_us= er_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe > 0.78 =B1 19% +6.4 7.15 =B1 3% perf-profile.calltrace.= cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_u= p_common.__wake_up_common_lock > 0.18 =B1141% +7.0 7.18 =B1 6% perf-profile.calltrace.= cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64= .entry_SYSCALL_64_after_hwframe.__libc_write > 1.89 =B1 15% +12.1 13.96 =B1 3% perf-profile.calltrace.= cycles-pp.__schedule.schedule.schedule_timeout.unix_stream_data_wait.unix_s= tream_read_generic > 1.92 =B1 15% +12.3 14.23 =B1 3% perf-profile.calltrace.= cycles-pp.schedule.schedule_timeout.unix_stream_data_wait.unix_stream_read_= generic.unix_stream_recvmsg > 1.66 =B1 19% +12.4 14.06 =B1 2% perf-profile.calltrace.= cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_u= p_common_lock.sock_def_readable > 1.96 =B1 15% +12.5 14.48 =B1 3% perf-profile.calltrace.= cycles-pp.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.u= nix_stream_recvmsg.sock_read_iter > 1.69 =B1 19% +12.7 14.38 =B1 2% perf-profile.calltrace.= cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_common_lock.s= ock_def_readable.unix_stream_sendmsg > 1.75 =B1 19% +13.0 14.75 =B1 2% perf-profile.calltrace.= cycles-pp.__wake_up_common.__wake_up_common_lock.sock_def_readable.unix_str= eam_sendmsg.sock_sendmsg > 2.53 =B1 10% +13.4 15.90 =B1 2% perf-profile.calltrace.= cycles-pp.sock_def_readable.unix_stream_sendmsg.sock_sendmsg.sock_write_ite= r.vfs_write > 1.96 =B1 16% +13.5 15.42 =B1 2% perf-profile.calltrace.= cycles-pp.__wake_up_common_lock.sock_def_readable.unix_stream_sendmsg.sock_= sendmsg.sock_write_iter > 2.28 =B1 15% +14.6 16.86 =B1 3% perf-profile.calltrace.= cycles-pp.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvms= g.sock_read_iter.vfs_read > 15.31 =B1 4% -8.6 6.67 =B1 5% perf-profile.children.c= ycles-pp.sock_alloc_send_pskb > 13.85 =B1 6% -7.9 5.98 =B1 5% perf-profile.children.c= ycles-pp.alloc_skb_with_frags > 13.70 =B1 6% -7.8 5.89 =B1 6% perf-profile.children.c= ycles-pp.__alloc_skb > 9.01 =B1 2% -5.5 3.48 =B1 4% perf-profile.children.c= ycles-pp.consume_skb > 6.86 =B1 26% -4.7 2.15 =B1 2% perf-profile.children.c= ycles-pp._raw_spin_lock_irqsave > 11.27 =B1 3% -4.6 6.67 =B1 3% perf-profile.children.c= ycles-pp.syscall_return_via_sysret > 6.46 =B1 4% -4.3 2.15 =B1 6% perf-profile.children.c= ycles-pp.skb_release_data > 4.18 =B1 25% -4.0 0.15 =B1 69% perf-profile.children.c= ycles-pp.___slab_alloc > 5.76 =B1 32% -3.9 1.91 =B1 3% perf-profile.children.c= ycles-pp.native_queued_spin_lock_slowpath > 5.98 =B1 8% -3.5 2.52 =B1 5% perf-profile.children.c= ycles-pp.kmem_cache_alloc_node > 5.84 =B1 6% -3.3 2.50 =B1 6% perf-profile.children.c= ycles-pp.kmalloc_reserve > 3.33 =B1 30% -3.3 0.05 =B1 88% perf-profile.children.c= ycles-pp.get_partial_node > 5.63 =B1 7% -3.3 2.37 =B1 6% perf-profile.children.c= ycles-pp.__kmalloc_node_track_caller > 5.20 =B1 7% -3.1 2.12 =B1 6% perf-profile.children.c= ycles-pp.__kmem_cache_alloc_node > 6.23 =B1 2% -2.9 3.33 =B1 3% perf-profile.children.c= ycles-pp.unix_stream_read_actor > 6.18 =B1 2% -2.9 3.31 =B1 3% perf-profile.children.c= ycles-pp.skb_copy_datagram_iter > 6.11 =B1 2% -2.9 3.25 =B1 3% perf-profile.children.c= ycles-pp.__skb_datagram_iter > 51.39 -2.5 48.85 perf-profile.children.cycle= s-pp.__libc_read > 3.14 =B1 3% -2.5 0.61 =B1 13% perf-profile.children.c= ycles-pp.__slab_free > 5.34 =B1 3% -2.1 3.23 =B1 3% perf-profile.children.c= ycles-pp.__entry_text_start > 3.57 =B1 2% -1.9 1.66 =B1 6% perf-profile.children.c= ycles-pp.copy_user_enhanced_fast_string > 3.16 =B1 2% -1.7 1.43 =B1 4% perf-profile.children.c= ycles-pp._copy_to_iter > 2.74 =B1 2% -1.6 1.16 =B1 4% perf-profile.children.c= ycles-pp.copyout > 4.16 =B1 2% -1.5 2.62 =B1 3% perf-profile.children.c= ycles-pp.__check_object_size > 2.73 =B1 3% -1.4 1.35 =B1 6% perf-profile.children.c= ycles-pp.kmem_cache_free > 2.82 =B1 2% -1.2 1.63 =B1 3% perf-profile.children.c= ycles-pp.check_heap_object > 2.27 =B1 4% -1.1 1.13 =B1 2% perf-profile.children.c= ycles-pp.skb_release_head_state > 2.85 =B1 3% -1.1 1.72 =B1 2% perf-profile.children.c= ycles-pp.simple_copy_to_iter > 2.22 =B1 4% -1.1 1.10 perf-profile.children.cyc= les-pp.unix_destruct_scm > 3.00 =B1 2% -1.1 1.91 =B1 5% perf-profile.children.c= ycles-pp.skb_copy_datagram_from_iter > 2.00 =B1 4% -1.1 0.92 =B1 2% perf-profile.children.c= ycles-pp.sock_wfree > 2.16 =B1 3% -0.7 1.43 =B1 7% perf-profile.children.c= ycles-pp.memcg_slab_post_alloc_hook > 1.45 =B1 3% -0.7 0.73 =B1 7% perf-profile.children.c= ycles-pp.entry_SYSCALL_64_safe_stack > 2.21 =B1 2% -0.7 1.52 =B1 3% perf-profile.children.c= ycles-pp.__kmem_cache_free > 1.49 =B1 3% -0.6 0.89 =B1 8% perf-profile.children.c= ycles-pp._copy_from_iter > 1.40 =B1 3% -0.6 0.85 =B1 13% perf-profile.children.c= ycles-pp.mod_objcg_state > 0.74 -0.5 0.24 =B1 16% perf-profile.children.cyc= les-pp.__build_skb_around > 1.48 -0.5 1.01 =B1 2% perf-profile.children.cyc= les-pp.get_obj_cgroup_from_current > 2.05 =B1 2% -0.5 1.59 =B1 2% perf-profile.children.c= ycles-pp.security_file_permission > 0.98 =B1 2% -0.4 0.59 =B1 10% perf-profile.children.c= ycles-pp.copyin > 1.08 =B1 3% -0.4 0.72 =B1 3% perf-profile.children.c= ycles-pp.__might_resched > 1.75 -0.3 1.42 =B1 4% perf-profile.children.cyc= les-pp.apparmor_file_permission > 1.32 =B1 4% -0.3 1.00 =B1 3% perf-profile.children.c= ycles-pp.sock_recvmsg > 0.54 =B1 4% -0.3 0.25 =B1 6% perf-profile.children.c= ycles-pp.skb_unlink > 0.54 =B1 6% -0.3 0.26 =B1 3% perf-profile.children.c= ycles-pp.unix_write_space > 0.66 =B1 3% -0.3 0.39 =B1 4% perf-profile.children.c= ycles-pp.obj_cgroup_charge > 0.68 =B1 2% -0.3 0.41 =B1 4% perf-profile.children.c= ycles-pp.entry_SYSRETQ_unsafe_stack > 0.86 =B1 4% -0.3 0.59 =B1 3% perf-profile.children.c= ycles-pp.__check_heap_object > 0.75 =B1 9% -0.3 0.48 =B1 2% perf-profile.children.c= ycles-pp.skb_set_owner_w > 1.84 =B1 3% -0.3 1.58 =B1 4% perf-profile.children.c= ycles-pp.aa_sk_perm > 0.68 =B1 11% -0.2 0.44 =B1 3% perf-profile.children.c= ycles-pp.skb_queue_tail > 1.22 =B1 4% -0.2 0.99 =B1 5% perf-profile.children.c= ycles-pp.__fdget_pos > 0.70 =B1 2% -0.2 0.48 =B1 5% perf-profile.children.c= ycles-pp.__get_obj_cgroup_from_memcg > 1.16 =B1 4% -0.2 0.93 =B1 3% perf-profile.children.c= ycles-pp.security_socket_recvmsg > 0.48 =B1 3% -0.2 0.29 =B1 4% perf-profile.children.c= ycles-pp.__might_fault > 0.24 =B1 7% -0.2 0.05 =B1 56% perf-profile.children.c= ycles-pp.fsnotify_perm > 1.12 =B1 4% -0.2 0.93 =B1 6% perf-profile.children.c= ycles-pp.__fget_light > 1.24 =B1 3% -0.2 1.07 =B1 3% perf-profile.children.c= ycles-pp.security_socket_sendmsg > 0.61 =B1 3% -0.2 0.45 =B1 2% perf-profile.children.c= ycles-pp.__might_sleep > 0.33 =B1 5% -0.2 0.17 =B1 6% perf-profile.children.c= ycles-pp.refill_obj_stock > 0.40 =B1 2% -0.1 0.25 =B1 4% perf-profile.children.c= ycles-pp.kmalloc_slab > 0.57 =B1 2% -0.1 0.45 perf-profile.children.cyc= les-pp.asm_sysvec_apic_timer_interrupt > 0.54 =B1 3% -0.1 0.42 =B1 2% perf-profile.children.c= ycles-pp.wait_for_unix_gc > 0.42 =B1 2% -0.1 0.30 =B1 3% perf-profile.children.c= ycles-pp.is_vmalloc_addr > 1.00 =B1 2% -0.1 0.87 =B1 5% perf-profile.children.c= ycles-pp.__virt_addr_valid > 0.52 =B1 2% -0.1 0.41 perf-profile.children.cyc= les-pp.sysvec_apic_timer_interrupt > 0.33 =B1 3% -0.1 0.21 =B1 3% perf-profile.children.c= ycles-pp.tick_sched_handle > 0.36 =B1 2% -0.1 0.25 =B1 4% perf-profile.children.c= ycles-pp.tick_sched_timer > 0.47 =B1 2% -0.1 0.36 =B1 2% perf-profile.children.c= ycles-pp.hrtimer_interrupt > 0.48 =B1 2% -0.1 0.36 =B1 2% perf-profile.children.c= ycles-pp.__sysvec_apic_timer_interrupt > 0.32 =B1 3% -0.1 0.21 =B1 5% perf-profile.children.c= ycles-pp.update_process_times > 0.42 =B1 3% -0.1 0.31 =B1 2% perf-profile.children.c= ycles-pp.__hrtimer_run_queues > 0.26 =B1 6% -0.1 0.16 =B1 4% perf-profile.children.c= ycles-pp.kmalloc_size_roundup > 0.20 =B1 4% -0.1 0.10 =B1 9% perf-profile.children.c= ycles-pp.task_tick_fair > 0.24 =B1 3% -0.1 0.15 =B1 4% perf-profile.children.c= ycles-pp.scheduler_tick > 0.30 =B1 5% -0.1 0.21 =B1 8% perf-profile.children.c= ycles-pp.obj_cgroup_uncharge_pages > 0.20 =B1 2% -0.1 0.11 =B1 6% perf-profile.children.c= ycles-pp.should_failslab > 0.51 =B1 2% -0.1 0.43 =B1 6% perf-profile.children.c= ycles-pp.syscall_enter_from_user_mode > 0.15 =B1 8% -0.1 0.07 =B1 13% perf-profile.children.c= ycles-pp.syscall_exit_to_user_mode_prepare > 0.19 =B1 4% -0.1 0.12 =B1 5% perf-profile.children.c= ycles-pp.apparmor_socket_sendmsg > 0.20 =B1 4% -0.1 0.13 =B1 5% perf-profile.children.c= ycles-pp.aa_file_perm > 0.18 =B1 5% -0.1 0.12 =B1 5% perf-profile.children.c= ycles-pp.apparmor_socket_recvmsg > 0.14 =B1 13% -0.1 0.08 =B1 55% perf-profile.children.c= ycles-pp.__mod_memcg_lruvec_state > 0.24 =B1 4% -0.1 0.18 =B1 2% perf-profile.children.c= ycles-pp.rcu_all_qs > 0.18 =B1 10% -0.1 0.12 =B1 11% perf-profile.children.c= ycles-pp.memcg_account_kmem > 0.37 =B1 3% -0.1 0.31 =B1 3% perf-profile.children.c= ycles-pp.security_socket_getpeersec_dgram > 0.08 -0.0 0.06 =B1 8% perf-profile.children.cyc= les-pp.put_pid > 0.18 =B1 3% -0.0 0.16 =B1 4% perf-profile.children.c= ycles-pp.apparmor_socket_getpeersec_dgram > 0.21 =B1 3% +0.0 0.23 =B1 2% perf-profile.children.c= ycles-pp.__get_task_ioprio > 0.00 +0.1 0.05 perf-profile.children.cycle= s-pp.perf_exclude_event > 0.00 +0.1 0.06 =B1 7% perf-profile.children.cyc= les-pp.invalidate_user_asid > 0.00 +0.1 0.07 =B1 6% perf-profile.children.cyc= les-pp.__bitmap_and > 0.05 +0.1 0.13 =B1 8% perf-profile.children.cyc= les-pp.irqentry_exit_to_user_mode > 0.00 +0.1 0.08 =B1 7% perf-profile.children.cyc= les-pp.schedule_debug > 0.00 +0.1 0.08 =B1 13% perf-profile.children.cyc= les-pp.read@plt > 0.00 +0.1 0.08 =B1 5% perf-profile.children.cyc= les-pp.sysvec_reschedule_ipi > 0.00 +0.1 0.10 =B1 4% perf-profile.children.cyc= les-pp.tracing_gen_ctx_irq_test > 0.00 +0.1 0.10 =B1 4% perf-profile.children.cyc= les-pp.place_entity > 0.00 +0.1 0.12 =B1 10% perf-profile.children.cyc= les-pp.native_irq_return_iret > 0.07 =B1 14% +0.1 0.19 =B1 3% perf-profile.children.c= ycles-pp.__list_add_valid > 0.00 +0.1 0.13 =B1 6% perf-profile.children.cyc= les-pp.perf_trace_buf_alloc > 0.00 +0.1 0.13 =B1 34% perf-profile.children.cyc= les-pp._find_next_and_bit > 0.00 +0.1 0.14 =B1 5% perf-profile.children.cyc= les-pp.switch_ldt > 0.00 +0.1 0.15 =B1 5% perf-profile.children.cyc= les-pp.check_cfs_rq_runtime > 0.00 +0.1 0.15 =B1 30% perf-profile.children.cyc= les-pp.migrate_task_rq_fair > 0.00 +0.2 0.15 =B1 5% perf-profile.children.cyc= les-pp.__rdgsbase_inactive > 0.00 +0.2 0.16 =B1 3% perf-profile.children.cyc= les-pp.save_fpregs_to_fpstate > 0.00 +0.2 0.16 =B1 6% perf-profile.children.cyc= les-pp.ttwu_queue_wakelist > 0.00 +0.2 0.17 perf-profile.children.cycle= s-pp.perf_trace_buf_update > 0.00 +0.2 0.18 =B1 2% perf-profile.children.cyc= les-pp.rb_insert_color > 0.00 +0.2 0.18 =B1 4% perf-profile.children.cyc= les-pp.rb_next > 0.00 +0.2 0.18 =B1 21% perf-profile.children.cyc= les-pp.__cgroup_account_cputime > 0.01 =B1223% +0.2 0.21 =B1 28% perf-profile.children.c= ycles-pp.perf_trace_sched_switch > 0.00 +0.2 0.20 =B1 3% perf-profile.children.cyc= les-pp.select_idle_cpu > 0.00 +0.2 0.20 =B1 3% perf-profile.children.cyc= les-pp.rcu_note_context_switch > 0.00 +0.2 0.21 =B1 26% perf-profile.children.cyc= les-pp.set_task_cpu > 0.00 +0.2 0.22 =B1 8% perf-profile.children.cyc= les-pp.resched_curr > 0.08 =B1 5% +0.2 0.31 =B1 11% perf-profile.children.c= ycles-pp.task_h_load > 0.00 +0.2 0.24 =B1 3% perf-profile.children.cyc= les-pp.finish_wait > 0.04 =B1 44% +0.3 0.29 =B1 5% perf-profile.children.c= ycles-pp.rb_erase > 0.19 =B1 6% +0.3 0.46 perf-profile.children.cyc= les-pp._raw_spin_unlock_irqrestore > 0.20 =B1 6% +0.3 0.47 =B1 3% perf-profile.children.c= ycles-pp.__list_del_entry_valid > 0.00 +0.3 0.28 =B1 3% perf-profile.children.cyc= les-pp.__wrgsbase_inactive > 0.02 =B1141% +0.3 0.30 =B1 2% perf-profile.children.c= ycles-pp.native_sched_clock > 0.06 =B1 13% +0.3 0.34 =B1 2% perf-profile.children.c= ycles-pp.sched_clock_cpu > 0.64 =B1 2% +0.3 0.93 perf-profile.children.cyc= les-pp.mutex_lock > 0.00 +0.3 0.30 =B1 5% perf-profile.children.cyc= les-pp.cr4_update_irqsoff > 0.00 +0.3 0.30 =B1 4% perf-profile.children.cyc= les-pp.clear_buddies > 0.07 =B1 55% +0.3 0.37 =B1 5% perf-profile.children.c= ycles-pp.perf_trace_sched_stat_runtime > 0.10 =B1 66% +0.3 0.42 =B1 5% perf-profile.children.c= ycles-pp.perf_tp_event > 0.02 =B1142% +0.3 0.36 =B1 6% perf-profile.children.c= ycles-pp.cpuacct_charge > 0.12 =B1 9% +0.4 0.47 =B1 11% perf-profile.children.c= ycles-pp.wake_affine > 0.00 +0.4 0.36 =B1 13% perf-profile.children.cyc= les-pp.available_idle_cpu > 0.05 =B1 48% +0.4 0.42 =B1 6% perf-profile.children.c= ycles-pp.finish_task_switch > 0.12 =B1 4% +0.4 0.49 =B1 4% perf-profile.children.c= ycles-pp.asm_sysvec_reschedule_ipi > 0.07 =B1 17% +0.4 0.48 perf-profile.children.cyc= les-pp.__calc_delta > 0.03 =B1100% +0.5 0.49 =B1 4% perf-profile.children.c= ycles-pp.pick_next_entity > 0.00 +0.5 0.48 =B1 8% perf-profile.children.cyc= les-pp.set_next_buddy > 0.08 =B1 14% +0.6 0.66 =B1 4% perf-profile.children.c= ycles-pp.update_min_vruntime > 0.07 =B1 17% +0.6 0.68 =B1 2% perf-profile.children.c= ycles-pp.os_xsave > 0.29 =B1 7% +0.7 0.99 =B1 3% perf-profile.children.c= ycles-pp.update_cfs_group > 0.17 =B1 17% +0.7 0.87 =B1 4% perf-profile.children.c= ycles-pp.perf_trace_sched_wakeup_template > 0.14 =B1 7% +0.7 0.87 =B1 3% perf-profile.children.c= ycles-pp.__update_load_avg_se > 0.14 =B1 16% +0.8 0.90 =B1 2% perf-profile.children.c= ycles-pp.update_rq_clock > 0.08 =B1 17% +0.8 0.84 =B1 5% perf-profile.children.c= ycles-pp.check_preempt_wakeup > 0.12 =B1 14% +0.8 0.95 =B1 3% perf-profile.children.c= ycles-pp.__update_load_avg_cfs_rq > 0.22 =B1 5% +0.8 1.07 =B1 3% perf-profile.children.c= ycles-pp.prepare_to_wait > 0.10 =B1 18% +0.9 0.98 =B1 3% perf-profile.children.c= ycles-pp.check_preempt_curr > 29.72 +0.9 30.61 perf-profile.children.cycle= s-pp.vfs_write > 0.14 =B1 11% +0.9 1.03 =B1 4% perf-profile.children.c= ycles-pp.__switch_to > 0.07 =B1 20% +0.9 0.99 =B1 6% perf-profile.children.c= ycles-pp.put_prev_entity > 0.12 =B1 16% +1.0 1.13 =B1 5% perf-profile.children.c= ycles-pp.___perf_sw_event > 0.07 =B1 17% +1.0 1.10 =B1 13% perf-profile.children.c= ycles-pp.select_idle_sibling > 27.82 =B1 2% +1.2 28.99 perf-profile.children.cyc= les-pp.unix_stream_recvmsg > 27.41 =B1 2% +1.2 28.63 perf-profile.children.cyc= les-pp.unix_stream_read_generic > 0.20 =B1 15% +1.4 1.59 =B1 3% perf-profile.children.c= ycles-pp.reweight_entity > 0.21 =B1 13% +1.4 1.60 =B1 4% perf-profile.children.c= ycles-pp.__switch_to_asm > 0.23 =B1 10% +1.4 1.65 =B1 5% perf-profile.children.c= ycles-pp.restore_fpregs_from_fpstate > 0.20 =B1 13% +1.5 1.69 =B1 3% perf-profile.children.c= ycles-pp.set_next_entity > 27.59 +1.6 29.19 perf-profile.children.cycle= s-pp.sock_write_iter > 0.28 =B1 10% +1.8 2.12 =B1 5% perf-profile.children.c= ycles-pp.switch_fpu_return > 0.26 =B1 11% +1.8 2.10 =B1 6% perf-profile.children.c= ycles-pp.select_task_rq_fair > 26.66 =B1 2% +2.0 28.63 perf-profile.children.cyc= les-pp.sock_sendmsg > 0.31 =B1 12% +2.1 2.44 =B1 5% perf-profile.children.c= ycles-pp.select_task_rq > 0.30 =B1 14% +2.2 2.46 =B1 4% perf-profile.children.c= ycles-pp.prepare_task_switch > 25.27 =B1 2% +2.2 27.47 perf-profile.children.cyc= les-pp.unix_stream_sendmsg > 2.10 +2.3 4.38 =B1 2% perf-profile.children.cyc= les-pp._raw_spin_lock > 0.40 =B1 14% +2.5 2.92 =B1 5% perf-profile.children.c= ycles-pp.dequeue_entity > 48.40 +2.6 51.02 perf-profile.children.cycle= s-pp.__libc_write > 0.46 =B1 15% +3.1 3.51 =B1 3% perf-profile.children.c= ycles-pp.enqueue_entity > 0.49 =B1 10% +3.2 3.64 =B1 7% perf-profile.children.c= ycles-pp.update_load_avg > 0.53 =B1 20% +3.4 3.91 =B1 3% perf-profile.children.c= ycles-pp.update_curr > 80.81 +3.4 84.24 perf-profile.children.cycle= s-pp.entry_SYSCALL_64_after_hwframe > 0.50 =B1 12% +3.5 4.00 =B1 4% perf-profile.children.c= ycles-pp.switch_mm_irqs_off > 0.55 =B1 9% +3.8 4.38 =B1 4% perf-profile.children.c= ycles-pp.pick_next_task_fair > 9.60 +4.6 14.15 =B1 2% perf-profile.children.cyc= les-pp.syscall_exit_to_user_mode > 0.78 =B1 13% +4.9 5.65 =B1 4% perf-profile.children.c= ycles-pp.dequeue_task_fair > 0.78 =B1 15% +5.2 5.99 =B1 3% perf-profile.children.c= ycles-pp.enqueue_task_fair > 74.30 +5.6 79.86 perf-profile.children.cycle= s-pp.do_syscall_64 > 0.90 =B1 15% +6.3 7.16 =B1 3% perf-profile.children.c= ycles-pp.ttwu_do_activate > 0.33 =B1 31% +6.3 6.61 =B1 6% perf-profile.children.c= ycles-pp.exit_to_user_mode_loop > 0.82 =B1 15% +8.1 8.92 =B1 5% perf-profile.children.c= ycles-pp.exit_to_user_mode_prepare > 1.90 =B1 16% +12.2 14.10 =B1 2% perf-profile.children.c= ycles-pp.try_to_wake_up > 2.36 =B1 11% +12.2 14.60 =B1 3% perf-profile.children.c= ycles-pp.schedule_timeout > 1.95 =B1 15% +12.5 14.41 =B1 2% perf-profile.children.c= ycles-pp.autoremove_wake_function > 2.01 =B1 15% +12.8 14.76 =B1 2% perf-profile.children.c= ycles-pp.__wake_up_common > 2.23 =B1 13% +13.2 15.45 =B1 2% perf-profile.children.c= ycles-pp.__wake_up_common_lock > 2.53 =B1 10% +13.4 15.90 =B1 2% perf-profile.children.c= ycles-pp.sock_def_readable > 2.29 =B1 15% +14.6 16.93 =B1 3% perf-profile.children.c= ycles-pp.unix_stream_data_wait > 2.61 =B1 13% +18.0 20.65 =B1 4% perf-profile.children.c= ycles-pp.schedule > 2.66 =B1 13% +18.1 20.77 =B1 4% perf-profile.children.c= ycles-pp.__schedule > 11.25 =B1 3% -4.6 6.67 =B1 3% perf-profile.self.cycle= s-pp.syscall_return_via_sysret > 5.76 =B1 32% -3.9 1.90 =B1 3% perf-profile.self.cycle= s-pp.native_queued_spin_lock_slowpath > 8.69 =B1 3% -3.4 5.27 =B1 3% perf-profile.self.cycle= s-pp.syscall_exit_to_user_mode > 3.11 =B1 3% -2.5 0.60 =B1 13% perf-profile.self.cycle= s-pp.__slab_free > 6.65 =B1 2% -2.2 4.47 =B1 2% perf-profile.self.cycle= s-pp.entry_SYSCALL_64_after_hwframe > 4.78 =B1 3% -1.9 2.88 =B1 3% perf-profile.self.cycle= s-pp.__entry_text_start > 3.52 =B1 2% -1.9 1.64 =B1 6% perf-profile.self.cycle= s-pp.copy_user_enhanced_fast_string > 2.06 =B1 3% -1.1 0.96 =B1 5% perf-profile.self.cycle= s-pp.kmem_cache_free > 1.42 =B1 3% -1.0 0.46 =B1 10% perf-profile.self.cycle= s-pp.check_heap_object > 1.43 =B1 4% -0.8 0.64 perf-profile.self.cycles-= pp.sock_wfree > 0.99 =B1 3% -0.8 0.21 =B1 12% perf-profile.self.cycle= s-pp.skb_release_data > 0.84 =B1 8% -0.7 0.10 =B1 64% perf-profile.self.cycle= s-pp.___slab_alloc > 1.97 =B1 2% -0.6 1.32 perf-profile.self.cycles-= pp.unix_stream_read_generic > 1.60 =B1 3% -0.5 1.11 =B1 4% perf-profile.self.cycle= s-pp.memcg_slab_post_alloc_hook > 1.24 =B1 2% -0.5 0.75 =B1 11% perf-profile.self.cycle= s-pp.mod_objcg_state > 0.71 -0.5 0.23 =B1 15% perf-profile.self.cycles-= pp.__build_skb_around > 0.95 =B1 3% -0.5 0.50 =B1 6% perf-profile.self.cycle= s-pp.__alloc_skb > 0.97 =B1 4% -0.4 0.55 =B1 5% perf-profile.self.cycle= s-pp.kmem_cache_alloc_node > 0.99 =B1 3% -0.4 0.59 =B1 4% perf-profile.self.cycle= s-pp.vfs_write > 1.38 =B1 2% -0.4 0.99 perf-profile.self.cycles-= pp.__kmem_cache_free > 0.86 =B1 2% -0.4 0.50 =B1 3% perf-profile.self.cycle= s-pp.__kmem_cache_alloc_node > 0.92 =B1 4% -0.4 0.56 =B1 4% perf-profile.self.cycle= s-pp.sock_write_iter > 1.06 =B1 3% -0.4 0.70 =B1 3% perf-profile.self.cycle= s-pp.__might_resched > 0.73 =B1 4% -0.3 0.44 =B1 4% perf-profile.self.cycle= s-pp.__cond_resched > 0.85 =B1 3% -0.3 0.59 =B1 4% perf-profile.self.cycle= s-pp.__check_heap_object > 1.46 =B1 7% -0.3 1.20 =B1 2% perf-profile.self.cycle= s-pp.unix_stream_sendmsg > 0.73 =B1 9% -0.3 0.47 =B1 2% perf-profile.self.cycle= s-pp.skb_set_owner_w > 1.54 -0.3 1.28 =B1 4% perf-profile.self.cycles-= pp.apparmor_file_permission > 0.74 =B1 3% -0.2 0.50 =B1 2% perf-profile.self.cycle= s-pp.get_obj_cgroup_from_current > 1.15 =B1 3% -0.2 0.91 =B1 8% perf-profile.self.cycle= s-pp.aa_sk_perm > 0.60 -0.2 0.36 =B1 4% perf-profile.self.cycles-= pp.entry_SYSRETQ_unsafe_stack > 0.65 =B1 4% -0.2 0.45 =B1 6% perf-profile.self.cycle= s-pp.__get_obj_cgroup_from_memcg > 0.24 =B1 6% -0.2 0.05 =B1 56% perf-profile.self.cycle= s-pp.fsnotify_perm > 0.76 =B1 3% -0.2 0.58 =B1 2% perf-profile.self.cycle= s-pp.sock_read_iter > 1.10 =B1 4% -0.2 0.92 =B1 6% perf-profile.self.cycle= s-pp.__fget_light > 0.42 =B1 3% -0.2 0.25 =B1 4% perf-profile.self.cycle= s-pp.obj_cgroup_charge > 0.32 =B1 4% -0.2 0.17 =B1 6% perf-profile.self.cycle= s-pp.refill_obj_stock > 0.29 -0.2 0.14 =B1 8% perf-profile.self.cycles-= pp.__kmalloc_node_track_caller > 0.54 =B1 3% -0.1 0.40 =B1 2% perf-profile.self.cycle= s-pp.__might_sleep > 0.30 =B1 7% -0.1 0.16 =B1 22% perf-profile.self.cycle= s-pp.security_file_permission > 0.34 =B1 3% -0.1 0.21 =B1 6% perf-profile.self.cycle= s-pp.entry_SYSCALL_64_safe_stack > 0.41 =B1 3% -0.1 0.29 =B1 3% perf-profile.self.cycle= s-pp.is_vmalloc_addr > 0.27 =B1 3% -0.1 0.16 =B1 6% perf-profile.self.cycle= s-pp._copy_from_iter > 0.24 =B1 3% -0.1 0.12 =B1 3% perf-profile.self.cycle= s-pp.ksys_write > 0.95 =B1 2% -0.1 0.84 =B1 5% perf-profile.self.cycle= s-pp.__virt_addr_valid > 0.56 =B1 11% -0.1 0.46 =B1 4% perf-profile.self.cycle= s-pp.sock_def_readable > 0.16 =B1 7% -0.1 0.06 =B1 18% perf-profile.self.cycle= s-pp.sock_recvmsg > 0.22 =B1 5% -0.1 0.14 =B1 2% perf-profile.self.cycle= s-pp.ksys_read > 0.27 =B1 4% -0.1 0.19 =B1 5% perf-profile.self.cycle= s-pp.kmalloc_slab > 0.28 =B1 2% -0.1 0.20 =B1 2% perf-profile.self.cycle= s-pp.consume_skb > 0.35 =B1 2% -0.1 0.28 =B1 3% perf-profile.self.cycle= s-pp.__check_object_size > 0.13 =B1 8% -0.1 0.06 =B1 18% perf-profile.self.cycle= s-pp.syscall_exit_to_user_mode_prepare > 0.20 =B1 5% -0.1 0.12 =B1 6% perf-profile.self.cycle= s-pp.kmalloc_reserve > 0.26 =B1 5% -0.1 0.19 =B1 4% perf-profile.self.cycle= s-pp.sock_alloc_send_pskb > 0.42 =B1 2% -0.1 0.35 =B1 7% perf-profile.self.cycle= s-pp.syscall_enter_from_user_mode > 0.19 =B1 5% -0.1 0.12 =B1 6% perf-profile.self.cycle= s-pp.aa_file_perm > 0.16 =B1 4% -0.1 0.10 =B1 4% perf-profile.self.cycle= s-pp.skb_copy_datagram_from_iter > 0.18 =B1 4% -0.1 0.12 =B1 6% perf-profile.self.cycle= s-pp.apparmor_socket_sendmsg > 0.18 =B1 5% -0.1 0.12 =B1 4% perf-profile.self.cycle= s-pp.apparmor_socket_recvmsg > 0.15 =B1 5% -0.1 0.10 =B1 5% perf-profile.self.cycle= s-pp.alloc_skb_with_frags > 0.64 =B1 3% -0.1 0.59 perf-profile.self.cycles-= pp.__libc_write > 0.20 =B1 4% -0.1 0.15 =B1 3% perf-profile.self.cycle= s-pp._copy_to_iter > 0.15 =B1 5% -0.1 0.10 =B1 11% perf-profile.self.cycle= s-pp.sock_sendmsg > 0.08 =B1 4% -0.1 0.03 =B1 81% perf-profile.self.cycle= s-pp.copyout > 0.11 =B1 6% -0.0 0.06 =B1 7% perf-profile.self.cycle= s-pp.__fdget_pos > 0.12 =B1 5% -0.0 0.07 =B1 10% perf-profile.self.cycle= s-pp.kmalloc_size_roundup > 0.34 =B1 3% -0.0 0.29 perf-profile.self.cycles-= pp.do_syscall_64 > 0.20 =B1 4% -0.0 0.15 =B1 4% perf-profile.self.cycle= s-pp.rcu_all_qs > 0.41 =B1 3% -0.0 0.37 =B1 8% perf-profile.self.cycle= s-pp.unix_stream_recvmsg > 0.22 =B1 2% -0.0 0.17 =B1 4% perf-profile.self.cycle= s-pp.unix_destruct_scm > 0.09 =B1 4% -0.0 0.05 perf-profile.self.cycles-= pp.should_failslab > 0.10 =B1 15% -0.0 0.06 =B1 50% perf-profile.self.cycle= s-pp.__mod_memcg_lruvec_state > 0.11 =B1 4% -0.0 0.07 perf-profile.self.cycles-= pp.__might_fault > 0.16 =B1 2% -0.0 0.13 =B1 6% perf-profile.self.cycle= s-pp.obj_cgroup_uncharge_pages > 0.18 =B1 4% -0.0 0.16 =B1 3% perf-profile.self.cycle= s-pp.security_socket_getpeersec_dgram > 0.28 =B1 2% -0.0 0.25 =B1 2% perf-profile.self.cycle= s-pp.unix_write_space > 0.17 =B1 2% -0.0 0.15 =B1 5% perf-profile.self.cycle= s-pp.apparmor_socket_getpeersec_dgram > 0.08 =B1 6% -0.0 0.05 =B1 7% perf-profile.self.cycle= s-pp.security_socket_sendmsg > 0.12 =B1 4% -0.0 0.10 =B1 3% perf-profile.self.cycle= s-pp.__skb_datagram_iter > 0.24 =B1 2% -0.0 0.22 perf-profile.self.cycles-= pp.mutex_unlock > 0.08 =B1 5% +0.0 0.10 =B1 6% perf-profile.self.cycle= s-pp.scm_recv > 0.17 =B1 2% +0.0 0.19 =B1 3% perf-profile.self.cycle= s-pp.__x64_sys_read > 0.19 =B1 3% +0.0 0.22 =B1 2% perf-profile.self.cycle= s-pp.__get_task_ioprio > 0.00 +0.1 0.06 perf-profile.self.cycles-pp= .finish_wait > 0.00 +0.1 0.06 =B1 7% perf-profile.self.cycles-= pp.cr4_update_irqsoff > 0.00 +0.1 0.06 =B1 7% perf-profile.self.cycles-= pp.invalidate_user_asid > 0.00 +0.1 0.07 =B1 12% perf-profile.self.cycles-= pp.wake_affine > 0.00 +0.1 0.07 =B1 7% perf-profile.self.cycles-= pp.check_cfs_rq_runtime > 0.00 +0.1 0.07 =B1 5% perf-profile.self.cycles-= pp.perf_trace_buf_update > 0.00 +0.1 0.07 =B1 9% perf-profile.self.cycles-= pp.asm_sysvec_reschedule_ipi > 0.00 +0.1 0.07 =B1 10% perf-profile.self.cycles-= pp.__bitmap_and > 0.00 +0.1 0.08 =B1 10% perf-profile.self.cycles-= pp.schedule_debug > 0.00 +0.1 0.08 =B1 13% perf-profile.self.cycles-= pp.read@plt > 0.00 +0.1 0.08 =B1 12% perf-profile.self.cycles-= pp.perf_trace_buf_alloc > 0.00 +0.1 0.09 =B1 35% perf-profile.self.cycles-= pp.migrate_task_rq_fair > 0.00 +0.1 0.09 =B1 5% perf-profile.self.cycles-= pp.place_entity > 0.00 +0.1 0.10 =B1 4% perf-profile.self.cycles-= pp.tracing_gen_ctx_irq_test > 0.00 +0.1 0.10 perf-profile.self.cycles-pp= .__wake_up_common_lock > 0.07 =B1 17% +0.1 0.18 =B1 3% perf-profile.self.cycle= s-pp.__list_add_valid > 0.00 +0.1 0.11 =B1 8% perf-profile.self.cycles-= pp.native_irq_return_iret > 0.00 +0.1 0.12 =B1 6% perf-profile.self.cycles-= pp.select_idle_cpu > 0.00 +0.1 0.12 =B1 34% perf-profile.self.cycles-= pp._find_next_and_bit > 0.00 +0.1 0.13 =B1 25% perf-profile.self.cycles-= pp.__cgroup_account_cputime > 0.00 +0.1 0.13 =B1 7% perf-profile.self.cycles-= pp.switch_ldt > 0.00 +0.1 0.14 =B1 5% perf-profile.self.cycles-= pp.check_preempt_curr > 0.00 +0.1 0.15 =B1 2% perf-profile.self.cycles-= pp.save_fpregs_to_fpstate > 0.00 +0.1 0.15 =B1 5% perf-profile.self.cycles-= pp.__rdgsbase_inactive > 0.14 =B1 3% +0.2 0.29 perf-profile.self.cycles-= pp._raw_spin_unlock_irqrestore > 0.00 +0.2 0.15 =B1 7% perf-profile.self.cycles-= pp.ttwu_queue_wakelist > 0.00 +0.2 0.17 =B1 4% perf-profile.self.cycles-= pp.rb_insert_color > 0.00 +0.2 0.17 =B1 5% perf-profile.self.cycles-= pp.rb_next > 0.00 +0.2 0.18 =B1 2% perf-profile.self.cycles-= pp.autoremove_wake_function > 0.01 =B1223% +0.2 0.19 =B1 6% perf-profile.self.cycle= s-pp.ttwu_do_activate > 0.00 +0.2 0.20 =B1 2% perf-profile.self.cycles-= pp.rcu_note_context_switch > 0.00 +0.2 0.20 =B1 7% perf-profile.self.cycles-= pp.exit_to_user_mode_loop > 0.27 +0.2 0.47 =B1 3% perf-profile.self.cycles-= pp.mutex_lock > 0.00 +0.2 0.20 =B1 28% perf-profile.self.cycles-= pp.perf_trace_sched_switch > 0.00 +0.2 0.21 =B1 9% perf-profile.self.cycles-= pp.resched_curr > 0.04 =B1 45% +0.2 0.26 =B1 7% perf-profile.self.cycle= s-pp.perf_tp_event > 0.06 =B1 7% +0.2 0.28 =B1 8% perf-profile.self.cycle= s-pp.perf_trace_sched_wakeup_template > 0.19 =B1 7% +0.2 0.41 =B1 5% perf-profile.self.cycle= s-pp.__list_del_entry_valid > 0.08 =B1 5% +0.2 0.31 =B1 11% perf-profile.self.cycle= s-pp.task_h_load > 0.00 +0.2 0.23 =B1 5% perf-profile.self.cycles-= pp.finish_task_switch > 0.03 =B1 70% +0.2 0.27 =B1 5% perf-profile.self.cycle= s-pp.rb_erase > 0.02 =B1142% +0.3 0.29 =B1 2% perf-profile.self.cycle= s-pp.native_sched_clock > 0.00 +0.3 0.28 =B1 3% perf-profile.self.cycles-= pp.__wrgsbase_inactive > 0.00 +0.3 0.28 =B1 6% perf-profile.self.cycles-= pp.clear_buddies > 0.07 =B1 10% +0.3 0.35 =B1 3% perf-profile.self.cycle= s-pp.schedule_timeout > 0.03 =B1 70% +0.3 0.33 =B1 3% perf-profile.self.cycle= s-pp.select_task_rq > 0.06 =B1 13% +0.3 0.36 =B1 4% perf-profile.self.cycle= s-pp.__wake_up_common > 0.06 =B1 13% +0.3 0.36 =B1 3% perf-profile.self.cycle= s-pp.dequeue_entity > 0.06 =B1 18% +0.3 0.37 =B1 7% perf-profile.self.cycle= s-pp.perf_trace_sched_stat_runtime > 0.01 =B1223% +0.3 0.33 =B1 4% perf-profile.self.cycle= s-pp.schedule > 0.02 =B1142% +0.3 0.35 =B1 7% perf-profile.self.cycle= s-pp.cpuacct_charge > 0.01 =B1223% +0.3 0.35 perf-profile.self.cycles-= pp.set_next_entity > 0.00 +0.4 0.35 =B1 13% perf-profile.self.cycles-= pp.available_idle_cpu > 0.08 =B1 10% +0.4 0.44 =B1 5% perf-profile.self.cycle= s-pp.prepare_to_wait > 0.63 =B1 3% +0.4 1.00 =B1 4% perf-profile.self.cycle= s-pp.vfs_read > 0.02 =B1142% +0.4 0.40 =B1 4% perf-profile.self.cycle= s-pp.check_preempt_wakeup > 0.02 =B1141% +0.4 0.42 =B1 4% perf-profile.self.cycle= s-pp.pick_next_entity > 0.07 =B1 17% +0.4 0.48 perf-profile.self.cycles-= pp.__calc_delta > 0.06 =B1 14% +0.4 0.47 =B1 3% perf-profile.self.cycle= s-pp.unix_stream_data_wait > 0.04 =B1 45% +0.4 0.45 =B1 4% perf-profile.self.cycle= s-pp.switch_fpu_return > 0.00 +0.5 0.46 =B1 7% perf-profile.self.cycles-= pp.set_next_buddy > 0.07 =B1 17% +0.5 0.53 =B1 3% perf-profile.self.cycle= s-pp.select_task_rq_fair > 0.08 =B1 16% +0.5 0.55 =B1 4% perf-profile.self.cycle= s-pp.try_to_wake_up > 0.08 =B1 19% +0.5 0.56 =B1 3% perf-profile.self.cycle= s-pp.update_rq_clock > 0.02 =B1141% +0.5 0.50 =B1 10% perf-profile.self.cycle= s-pp.select_idle_sibling > 0.77 =B1 2% +0.5 1.25 =B1 2% perf-profile.self.cycle= s-pp.__libc_read > 0.09 =B1 19% +0.5 0.59 =B1 3% perf-profile.self.cycle= s-pp.reweight_entity > 0.08 =B1 14% +0.5 0.59 =B1 2% perf-profile.self.cycle= s-pp.dequeue_task_fair > 0.08 =B1 13% +0.6 0.64 =B1 5% perf-profile.self.cycle= s-pp.update_min_vruntime > 0.02 =B1141% +0.6 0.58 =B1 7% perf-profile.self.cycle= s-pp.put_prev_entity > 0.06 =B1 11% +0.6 0.64 =B1 4% perf-profile.self.cycle= s-pp.enqueue_task_fair > 0.07 =B1 18% +0.6 0.68 =B1 3% perf-profile.self.cycle= s-pp.os_xsave > 1.39 =B1 2% +0.7 2.06 =B1 3% perf-profile.self.cycle= s-pp._raw_spin_lock_irqsave > 0.28 =B1 8% +0.7 0.97 =B1 4% perf-profile.self.cycle= s-pp.update_cfs_group > 0.14 =B1 8% +0.7 0.83 =B1 3% perf-profile.self.cycle= s-pp.__update_load_avg_se > 1.76 =B1 3% +0.7 2.47 =B1 3% perf-profile.self.cycle= s-pp._raw_spin_lock > 0.12 =B1 12% +0.7 0.85 =B1 5% perf-profile.self.cycle= s-pp.prepare_task_switch > 0.12 =B1 12% +0.8 0.91 =B1 3% perf-profile.self.cycle= s-pp.__update_load_avg_cfs_rq > 0.13 =B1 12% +0.8 0.93 =B1 5% perf-profile.self.cycle= s-pp.pick_next_task_fair > 0.13 =B1 12% +0.9 0.98 =B1 4% perf-profile.self.cycle= s-pp.__switch_to > 0.11 =B1 18% +0.9 1.06 =B1 5% perf-profile.self.cycle= s-pp.___perf_sw_event > 0.16 =B1 11% +1.2 1.34 =B1 4% perf-profile.self.cycle= s-pp.enqueue_entity > 0.20 =B1 12% +1.4 1.58 =B1 4% perf-profile.self.cycle= s-pp.__switch_to_asm > 0.23 =B1 10% +1.4 1.65 =B1 5% perf-profile.self.cycle= s-pp.restore_fpregs_from_fpstate > 0.25 =B1 12% +1.5 1.77 =B1 4% perf-profile.self.cycle= s-pp.__schedule > 0.22 =B1 10% +1.6 1.78 =B1 10% perf-profile.self.cycle= s-pp.update_load_avg > 0.23 =B1 16% +1.7 1.91 =B1 7% perf-profile.self.cycle= s-pp.update_curr > 0.48 =B1 11% +3.4 3.86 =B1 4% perf-profile.self.cycle= s-pp.switch_mm_irqs_off > = > = > To reproduce: > = > git clone https://github.com/intel/lkp-tests.git > cd lkp-tests > sudo bin/lkp install job.yaml # job file is attached in= this email > bin/lkp split-job --compatible job.yaml # generate the yaml file = for lkp run > sudo bin/lkp run generated-yaml-file > = > # if come across any failure that blocks the test, > # please remove ~/.lkp and /lkp dir to run from a clean state. Amazon Development Center Germany GmbH Krausenstr. 38 10117 Berlin Geschaeftsfuehrung: Christian Schlaeger, Jonathan Weiss Eingetragen am Amtsgericht Charlottenburg unter HRB 149173 B Sitz: Berlin Ust-ID: DE 289 237 879