Received: by 2002:a05:6358:3188:b0:123:57c1:9b43 with SMTP id q8csp5602485rwd; Wed, 24 May 2023 04:24:49 -0700 (PDT) X-Google-Smtp-Source: ACHHUZ6lGI5SJ8PnrEoDgH4nVD7C6sNgo1zetw0b70c9pNuLEftDKNtjTl7r9w/l6X1I674Rw38a X-Received: by 2002:a17:903:32c3:b0:1ab:8f4:af2b with SMTP id i3-20020a17090332c300b001ab08f4af2bmr19211514plr.38.1684927488641; Wed, 24 May 2023 04:24:48 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1684927488; cv=none; d=google.com; s=arc-20160816; b=f0F8Yd0Mi95bhRdNiVhjRMHOufh1lHd1sMEFJlkZqdcSHmR+cT2vkVsS1J2vXQqLml clc29HWiBfcyVCbKC6ehsEMj/zG1VOTAhoE6N/xejquPvi0b5nN1NDOhNI2BbMxYjoYV +grPUTy0wFIMQW0mHdfzmXG8csbE6azpC736Nb2uC2tojNCyXyn7pHQGT8nIIkRdw0oh CyTyZcTwvJfHkNGj/XtjjR/mlZfe9O5s/KFSDckjVzJZyzk7/ucrY1+FLRtq8iL817rE hbhwSLa3JFVolEo0NeK8ErVyLa34smiHfx3/89aCdYZrCDd2DFyzb5N6BBdrhWKaVDtH Pv3w== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:in-reply-to:references :cc:to:from:content-language:subject:mime-version:date :dkim-signature:message-id; bh=k4hIKekwZ63JzfKT+4+2e+oZbwx87RUDyZhuORZ7Thw=; b=eysGd9W3UVZUse1RzxFDY4ZQJ6eOJ4scO9pjAitYqbeUZVNItcrrJ3YsIiLMRfXNrJ YKUeH3efePubbvP95zY6JzSNJIChJ89gPoOfKCQAHJSSq3C78NkpZWGV9Vdcu/htOc2W oyMSipovFL8GZcx/ACaUdwv+5+FMl1Uz5hQB8p1//EcIpLNn02ig7NnjU+l1foyqRWRK lBwgHa2f6q98E+90w0EeRxQxjZZqZpeFtsT4xUxawW00SdrNmDQ0zoVd6320cnA2Uw/w li9EgrLASIhTLygodNaLjztAztPAT/mqGps6M1km1ViT0MPfWXCISuksqP5mM51UEl7n UFLw== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@linux.dev header.s=key1 header.b=vPrv2mS5; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=linux.dev Return-Path: Received: from out1.vger.email (out1.vger.email. [2620:137:e000::1:20]) by mx.google.com with ESMTP id e11-20020a170902b78b00b001ac896ff65dsi270196pls.480.2023.05.24.04.24.33; Wed, 24 May 2023 04:24:48 -0700 (PDT) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20; Authentication-Results: mx.google.com; dkim=pass header.i=@linux.dev header.s=key1 header.b=vPrv2mS5; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=linux.dev Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S233824AbjEXLI2 (ORCPT + 99 others); Wed, 24 May 2023 07:08:28 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:41668 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232155AbjEXLI1 (ORCPT ); Wed, 24 May 2023 07:08:27 -0400 X-Greylist: delayed 28423 seconds by postgrey-1.37 at lindbergh.monkeyblade.net; Wed, 24 May 2023 04:08:20 PDT Received: from out-61.mta0.migadu.com (out-61.mta0.migadu.com [91.218.175.61]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 31A4012F for ; Wed, 24 May 2023 04:08:19 -0700 (PDT) Message-ID: DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1684926497; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=k4hIKekwZ63JzfKT+4+2e+oZbwx87RUDyZhuORZ7Thw=; b=vPrv2mS5k6f3otZHhwWzOrkzuND7I9gJ09jdmyIkdUYPXnSLiTWX8BvyIygt8P6o5qsYCe TMXwxMLh4pYreD2MEFWLtoDLHzGdI3FQPWXSqq/3pS7sa2/DMiY85OVHt1i3IYmiQlmYL7 JvmVwgHtRyEKPAjrgwEHAK7KzHXfWQg= Date: Wed, 24 May 2023 19:08:07 +0800 MIME-Version: 1.0 Subject: Re: [linus:master] [mm] f95bdb700b: stress-ng.ramfs.ops_per_sec -88.8% regression Content-Language: en-US X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. From: Qi Zheng To: Yujie Liu , RCU Cc: oe-lkp@lists.linux.dev, lkp@intel.com, linux-kernel@vger.kernel.org, Andrew Morton , Vlastimil Babka , Kirill Tkhai , Roman Gushchin , =?UTF-8?Q?Christian_K=c3=b6nig?= , David Hildenbrand , Davidlohr Bueso , Johannes Weiner , Michal Hocko , Muchun Song , "Paul E. McKenney" , Shakeel Butt , Yang Shi , linux-mm@kvack.org, ying.huang@intel.com, feng.tang@intel.com, fengwei.yin@intel.com References: <202305230837.db2c233f-yujie.liu@intel.com> <896bbb09-d400-ec73-ba3a-b64c6e9bbe46@linux.dev> In-Reply-To: <896bbb09-d400-ec73-ba3a-b64c6e9bbe46@linux.dev> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit X-Migadu-Flow: FLOW_OUT X-Spam-Status: No, score=-2.1 required=5.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF,SPF_HELO_NONE,SPF_PASS, T_SCC_BODY_TEXT_LINE autolearn=ham autolearn_force=no version=3.4.6 X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on lindbergh.monkeyblade.net Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2023/5/24 16:36, Qi Zheng wrote: > Hi Yujie, > > On 2023/5/24 15:31, Yujie Liu wrote: >> Hi Qi, >> >> On Tue, May 23, 2023 at 11:05:51AM +0800, Qi Zheng wrote: >>> On 2023/5/23 09:35, kernel test robot wrote: >>>> Hello, >>>> >>>> kernel test robot noticed a -88.8% regression of >>>> stress-ng.ramfs.ops_per_sec on: >>>> >>>> commit: f95bdb700bc6bb74e1199b1f5f90c613e152cfa7 ("mm: vmscan: make >>>> global slab shrink lockless") >>>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master >>> >>> Hi, >>> >>> I see that the reproduction program you pasted is the following: >>> >>> ``` >>> dmsetup remove_all >>> wipefs -a --force /dev/sda1 >>> mkfs -t ext4 -q -F /dev/sda1 >>> mkdir -p /fs/sda1 >>> mount -t ext4 /dev/sda1 /fs/sda1 >>> >>> for cpu_dir in /sys/devices/system/cpu/cpu[0-9]* >>> do >>>     online_file="$cpu_dir"/online >>>     [ -f "$online_file" ] && [ "$(cat "$online_file")" -eq 0 ] && >>> continue >>> >>>     file="$cpu_dir"/cpufreq/scaling_governor >>>     [ -f "$file" ] && echo "performance" > "$file" >>> done >>> >>>   "mkdir" "-p" "/mnt/stress-ng" >>>   "mount" "/dev/sda1" "/mnt/stress-ng" >>>   "stress-ng" "--timeout" "60" "--times" "--verify" "--metrics-brief" >>> "--ramfs" "9" >>> ``` >>> >>> Looks like it will do mount and umount stress tests. >>> The commit 475733dda5ae ("mm: vmscan: add shrinker_srcu_generation") >>> has optimized the umount, can you test it again with this commit? >> >> We tested commit 475733dda5ae and the regression persists on it. Please >> check the details as below: > > Thanks for the testing. > > After looking at the test data below more carefully, I found that there > is no perf hotspot of the shrink_slab(). That is, is the memory > reclamation not triggered during the test? > > If this is the case, the only difference between 42c9db3970483 and > f95bdb700bc6b is that there is one more call to synchronize_srcu(), and > no one is in the critical section of shrinker_srcu at this time. It is > strange that this can cause such a large performance regression, I will > try to reproduce and continue to investigate the cause. Well, I just ran the following command and reproduced the result: stress-ng --timeout 60 --times --verify --metrics-brief --ramfs 9 & 1) with commit 42c9db3970483: stress-ng: info: [11023] setting to a 60 second run per stressor stress-ng: info: [11023] dispatching hogs: 9 ramfs stress-ng: info: [11023] stressor bogo ops real time usr time sys time bogo ops/s bogo ops/s stress-ng: info: [11023] (secs) (secs) (secs) (real time) (usr+sys time) stress-ng: info: [11023] ramfs 774966 60.00 10.18 169.45 12915.89 4314.26 stress-ng: info: [11023] for a 60.00s run time: stress-ng: info: [11023] 1920.11s available CPU time stress-ng: info: [11023] 10.18s user time ( 0.53%) stress-ng: info: [11023] 169.44s system time ( 8.82%) stress-ng: info: [11023] 179.62s total time ( 9.35%) stress-ng: info: [11023] load average: 8.99 2.69 0.93 stress-ng: info: [11023] successful run completed in 60.00s (1 min, 0.00 secs) 2) with commit f95bdb700bc6b: stress-ng: info: [37676] dispatching hogs: 9 ramfs stress-ng: info: [37676] stressor bogo ops real time usr time sys time bogo ops/s bogo ops/s stress-ng: info: [37676] (secs) (secs) (secs) (real time) (usr+sys time) stress-ng: info: [37676] ramfs 168673 60.00 1.61 39.66 2811.08 4087.47 stress-ng: info: [37676] for a 60.10s run time: stress-ng: info: [37676] 1923.36s available CPU time stress-ng: info: [37676] 1.60s user time ( 0.08%) stress-ng: info: [37676] 39.66s system time ( 2.06%) stress-ng: info: [37676] 41.26s total time ( 2.15%) stress-ng: info: [37676] load average: 7.69 3.63 2.36 stress-ng: info: [37676] successful run completed in 60.10s (1 min, 0.10 secs) The bogo ops/s (real time) did drop significantly. And the memory reclaimation was not triggered in the whole process. so theoretically no one is in the read critical section of shrinker_srcu. Then I found that some stress-ng-ramfs processes were in TASK_UNINTERRUPTIBLE state for a long time: root 42313 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42314 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42315 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42316 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42317 7.8 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42318 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42319 7.8 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42320 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42321 7.8 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42322 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42323 7.8 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42324 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42325 7.8 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42326 0.0 0.0 69592 2068 pts/0 S 19:00 0:00 stress-ng-ramfs [run] root 42327 7.9 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42328 7.9 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42329 7.9 0.0 69592 1812 pts/0 D 19:00 0:02 stress-ng-ramfs [run] root 42330 7.9 0.0 69592 1556 pts/0 D 19:00 0:02 stress-ng-ramfs [run] Their call stack is as follows: cat /proc/42330/stack [<0>] __synchronize_srcu.part.21+0x83/0xb0 [<0>] unregister_shrinker+0x85/0xb0 [<0>] deactivate_locked_super+0x27/0x70 [<0>] cleanup_mnt+0xb8/0x140 [<0>] task_work_run+0x65/0x90 [<0>] exit_to_user_mode_prepare+0x1ba/0x1c0 [<0>] syscall_exit_to_user_mode+0x1b/0x40 [<0>] do_syscall_64+0x44/0x80 [<0>] entry_SYSCALL_64_after_hwframe+0x63/0xcd + RCU folks, Is this result as expected? I would have thought that synchronize_srcu() should return quickly if no one is in the read critical section. :( Thanks, Qi > > Thanks, > Qi > >> >> ========================================================================================= >> class/compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime: >> >> os/gcc-11/performance/1HDD/ext4/x86_64-rhel-8.3/10%/debian-11.1-x86_64-20220510.cgz/lkp-csl-2sp7/ramfs/stress-ng/60s >> >> commit: >>    42c9db3970483 ("mm: vmscan: add a map_nr_max field to shrinker_info") >>    f95bdb700bc6b ("mm: vmscan: make global slab shrink lockless") >>    475733dda5aed ("mm: vmscan: add shrinker_srcu_generation") >> >> 42c9db39704839ee f95bdb700bc6bb74e1199b1f5f9 475733dda5aedba9e086379aafe >> ---------------- --------------------------- --------------------------- >>           %stddev     %change         %stddev     %change         %stddev >>               \          |                \          |                \ >>     1408764           -88.8%     157151 ±  8%     -89.9%     142123 ± >> 2%  stress-ng.ramfs.ops >>       23479           -88.8%       2619 ±  8%     -89.9%       2368 ± >> 2%  stress-ng.ramfs.ops_per_sec >>     1310766           -88.3%     153888 ±  9%     -89.6%     136815 ± >> 4%  stress-ng.time.involuntary_context_switches >>      425.00           -82.7%      73.67 ± 16%     -86.4%      57.83 ± >> 11%  stress-ng.time.percent_of_cpu_this_job_got >>      251.56           -82.7%      43.61 ± 15%     -86.2%      34.63 ± >> 10%  stress-ng.time.system_time >>       12.68 ± 20%     -80.1%       2.52 ± 34%     -88.3%       1.48 ± >> 33%  stress-ng.time.user_time >>     4306933           -88.1%     513985 ±  7%     -89.1%     469986 ± >> 3%  stress-ng.time.voluntary_context_switches >>        1.09 ±  9%      +3.7%       1.13 ± 10%      -3.5%       1.05 ± >> 8%  boot-time.smp_boot >>     9358706           -27.4%    6793950           -28.5% >> 6686823        cpuidle..usage >>       93.32            +4.7%      97.73            +4.7% >> 97.68        iostat.cpu.idle >>        5.78           -70.0%       1.74 ±  4%     -69.2%       1.78 ± >> 6%  iostat.cpu.system >>        4.33 ± 10%     -73.1%       1.17 ± 31%     -76.9% >> 1.00        vmstat.procs.r >>      250072           -84.7%      38245 ±  5%     -85.3% >> 36826        vmstat.system.cs >>      229190           -54.6%     104141           -54.8% >> 103578        vmstat.system.in >>        9218 ± 15%     -31.3%       6336 ±  5%     -34.9% >> 5999        meminfo.Active >>        5681 ±  2%     -39.8%       3421 ± 10%     -46.2% >> 3059        meminfo.Active(anon) >>       53760 ±  4%     -15.0%      45717           -15.6% >> 45385        meminfo.Percpu >>       12396           -19.4%       9993 ±  3%     -22.1% >> 9655        meminfo.Shmem >>        0.75            +0.5        1.27 ±  9%      +0.6        1.38 ± >> 9%  mpstat.cpu.all.irq% >>        0.13 ±  3%      -0.0        0.10 ±  4%      -0.0        0.11 ± >> 10%  mpstat.cpu.all.soft% >>        5.00            -4.7        0.34 ± 16%      -4.7        0.27 ± >> 15%  mpstat.cpu.all.sys% >>        0.88 ±  4%      -0.4        0.51 ±  4%      -0.4        0.51 ± >> 8%  mpstat.cpu.all.usr% >>      494448 ± 13%     -55.9%     217823 ±  9%     -58.0%     207899 ± >> 15%  numa-numastat.node0.local_node >>      553385 ± 12%     -52.2%     264460 ±  9%     -55.7%     245273 ± >> 9%  numa-numastat.node0.numa_hit >>      598515 ± 11%     -68.7%     187452 ± 12%     -69.6%     181919 ± >> 17%  numa-numastat.node1.local_node >>      632713 ± 11%     -63.9%     228349 ± 11%     -63.3%     231957 ± >> 10%  numa-numastat.node1.numa_hit >>     1310766           -88.3%     153888 ±  9%     -89.6%     136815 ± >> 4%  time.involuntary_context_switches >>       10797            -2.2%      10561            -2.3% >> 10551        time.minor_page_faults >>      425.00           -82.7%      73.67 ± 16%     -86.4%      57.83 ± >> 11%  time.percent_of_cpu_this_job_got >>      251.56           -82.7%      43.61 ± 15%     -86.2%      34.63 ± >> 10%  time.system_time >>       12.68 ± 20%     -80.1%       2.52 ± 34%     -88.3%       1.48 ± >> 33%  time.user_time >>     4306933           -88.1%     513985 ±  7%     -89.1%     469986 ± >> 3%  time.voluntary_context_switches >>        6462 ± 14%     -39.9%       3884 ± 15%     -48.0%       3362 ± >> 21%  numa-meminfo.node1.Active >>        4789 ±  5%     -52.6%       2272 ± 14%     -61.3%       1854 ± >> 11%  numa-meminfo.node1.Active(anon) >>     1451867 ± 84%     -91.1%     128582 ± 72%      -6.3%    1360404 ± >> 99%  numa-meminfo.node1.FilePages >>     2362122 ± 53%     -55.5%    1050305 ± 16%      -0.1%    2360235 ± >> 55%  numa-meminfo.node1.MemUsed >>        6655 ± 20%     -40.3%       3974 ± 17%     -45.8%       3605 ± >> 33%  numa-meminfo.node1.Shmem >>      159872 ± 13%     -23.2%     122702 ± 15%     -11.3%     141742 ± >> 17%  numa-meminfo.node1.Slab >>     1441661 ± 85%     -91.5%     122024 ± 77%      -6.0%    1354658 ± >> 99%  numa-meminfo.node1.Unevictable >>      553499 ± 12%     -52.2%     264583 ±  9%     -55.7%     245366 ± >> 9%  numa-vmstat.node0.numa_hit >>      494562 ± 13%     -55.9%     217947 ±  9%     -57.9%     207992 ± >> 15%  numa-vmstat.node0.numa_local >>        1196 ±  5%     -52.6%     567.83 ± 14%     -61.3%     463.17 ± >> 11%  numa-vmstat.node1.nr_active_anon >>      362966 ± 84%     -91.1%      32145 ± 72%      -6.3%     340100 ± >> 99%  numa-vmstat.node1.nr_file_pages >>        1663 ± 20%     -40.3%     993.33 ± 17%     -45.8%     901.00 ± >> 33%  numa-vmstat.node1.nr_shmem >>      360414 ± 85%     -91.5%      30505 ± 77%      -6.0%     338664 ± >> 99%  numa-vmstat.node1.nr_unevictable >>        1196 ±  5%     -52.6%     567.83 ± 14%     -61.3%     463.17 ± >> 11%  numa-vmstat.node1.nr_zone_active_anon >>      360414 ± 85%     -91.5%      30505 ± 77%      -6.0%     338664 ± >> 99%  numa-vmstat.node1.nr_zone_unevictable >>      632698 ± 11%     -63.9%     228516 ± 11%     -63.3%     231989 ± >> 10%  numa-vmstat.node1.numa_hit >>      598500 ± 11%     -68.7%     187619 ± 12%     -69.6%     181950 ± >> 17%  numa-vmstat.node1.numa_local >>        1420 ±  2%     -39.8%     855.00 ± 10%     -46.2% >> 764.33        proc-vmstat.nr_active_anon >>       20431            -1.0%      20219            -1.3% >> 20173        proc-vmstat.nr_kernel_stack >>        3098           -19.4%       2498 ±  3%     -22.1% >> 2413        proc-vmstat.nr_shmem >>       30843            -7.7%      28469            -8.0% >> 28365        proc-vmstat.nr_slab_reclaimable >>       54154            -7.7%      49970            -8.3% >> 49633        proc-vmstat.nr_slab_unreclaimable >>        1420 ±  2%     -39.8%     855.00 ± 10%     -46.2% >> 764.33        proc-vmstat.nr_zone_active_anon >>     1187707           -58.4%     494370 ±  2%     -59.7% >> 478988        proc-vmstat.numa_hit >>     1096738           -62.9%     406835 ±  3%     -64.3% >> 391574        proc-vmstat.numa_local >>       93101            -6.0%      87525            -6.1% >> 87417        proc-vmstat.numa_other >>        3089 ± 11%     -69.4%     945.17 ±  5%     -70.6%     906.83 ± >> 4%  proc-vmstat.pgactivate >>     1431073           -62.4%     538512 ±  3%     -63.2% >> 525958        proc-vmstat.pgalloc_normal >>     1272095           -69.3%     390780 ±  4%     -70.4%     376657 ± >> 2%  proc-vmstat.pgfree >>      221.33           -72.0%      62.00 ±  3%     -71.2%      63.83 ± >> 4%  turbostat.Avg_MHz >>        8.02            -4.2        3.80 ±  3%      -4.1        3.89 ± >> 5%  turbostat.Busy% >>        2761           -40.6%       1641           -40.3%       1648 ± >> 3%  turbostat.Bzy_MHz >>     2052119 ±  9%     -96.5%      72643 ± 20%     -95.9%      83725 ± >> 21%  turbostat.C1 >>        2.78 ±  4%      -2.5        0.30 ± 18%      -2.4        0.37 ± >> 16%  turbostat.C1% >>     1984776 ±  2%     -58.3%     827902 ±  7%     -61.2%     770934 ± >> 2%  turbostat.C1E >>     5247470           +11.4%    5843977           +10.3% >> 5785798        turbostat.C6 >>        1.89 ±  7%    +350.5%       8.50 ± 54%    +642.9%      14.02 ± >> 35%  turbostat.CPU%c6 >>       52.00 ±  2%     -11.2%      46.17 ±  3%      -5.4%      49.17 ± >> 2%  turbostat.CoreTmp >>        0.20 ±  2%     +21.2%       0.24 ±  3%     +21.2%       0.24 ± >> 2%  turbostat.IPC >>    14915294           -54.2%    6833094 ±  2%     -55.2% >> 6681768        turbostat.IRQ >>       52057           -47.7%      27209 ±  8%     -49.2% >> 26458        turbostat.POLL >>        0.01            -0.0        0.00            -0.0 >> 0.00        turbostat.POLL% >>        0.77 ±  8%    +516.9%       4.75 ± 94%   +1219.9%      10.16 ± >> 53%  turbostat.Pkg%pc2 >>       52.00 ±  2%     -11.2%      46.17 ±  3%      -5.4%      49.17 ± >> 2%  turbostat.PkgTmp >>      168.06           -21.8%     131.38           -22.0% >> 131.08        turbostat.PkgWatt >>       37.48            -3.0%      36.38            -2.8% >> 36.42        turbostat.RAMWatt >>   2.822e+09 ±  2%     -68.8%  8.811e+08 ±  5%     -68.0%  9.035e+08 ± >> 5%  perf-stat.i.branch-instructions >>        1.19            +0.8        1.96 ±  8%      +0.8        2.01 ± >> 7%  perf-stat.i.branch-miss-rate% >>    32471421 ±  2%     -48.5%   16730947 ±  7%     -44.0%   18182845 ± >> 8%  perf-stat.i.branch-misses >>       21.01 ±  5%     -14.8        6.21 ± 18%     -15.5        5.54 ± >> 12%  perf-stat.i.cache-miss-rate% >>    35115410           -91.4%    3022306 ± 16%     -91.3%    3067843 ± >> 13%  perf-stat.i.cache-misses >>   1.658e+08 ±  6%     -69.7%   50263818 ± 13%     -65.6%   57128404 ± >> 15%  perf-stat.i.cache-references >>      261530           -84.8%      39659 ±  6%     -85.5% >> 37841        perf-stat.i.context-switches >>        1.45 ±  2%     -10.4%       1.30 ±  7%     -13.9%       1.25 ± >> 5%  perf-stat.i.cpi >>    2.07e+10           -74.6%  5.248e+09 ±  2%     -74.2%  5.344e+09 ± >> 3%  perf-stat.i.cpu-cycles >>        1143 ±  6%     -55.0%     514.25 ± 18%     -59.2%     466.64 ± >> 16%  perf-stat.i.cpu-migrations >>      613.99 ±  2%    +206.8%       1883 ± 18%    +200.7%       1845 ± >> 13%  perf-stat.i.cycles-between-cache-misses >>        0.04 ±  8%      +0.0        0.09 ± 23%      +0.1        0.10 ± >> 28%  perf-stat.i.dTLB-load-miss-rate% >>     1337737 ±  6%     -38.9%     817401 ± 21%     -21.6%    1049177 ± >> 26%  perf-stat.i.dTLB-load-misses >>   3.312e+09 ±  3%     -69.4%  1.014e+09 ±  5%     -69.0%  1.027e+09 ± >> 5%  perf-stat.i.dTLB-loads >>      245827 ±  3%     -66.3%      82725 ± 16%     -58.6%     101780 ± >> 23%  perf-stat.i.dTLB-store-misses >>   1.604e+09 ±  2%     -68.3%  5.086e+08 ±  5%     -67.9%  5.151e+08 ± >> 5%  perf-stat.i.dTLB-stores >>       54.44            -9.1       45.36 ±  5%      -7.7       46.69 ± >> 2%  perf-stat.i.iTLB-load-miss-rate% >>     3408503           -64.7%    1204595 ±  6%     -65.7%    1169269 ± >> 2%  perf-stat.i.iTLB-load-misses >>     2867132           -48.3%    1481327 ±  7%     -52.8%    1352174 ± >> 6%  perf-stat.i.iTLB-loads >>   1.421e+10 ±  2%     -70.2%  4.231e+09 ±  5%     -69.7%  4.306e+09 ± >> 5%  perf-stat.i.instructions >>        4234 ±  3%     -16.3%       3542 ±  6%     -12.4%       3707 ± >> 3%  perf-stat.i.instructions-per-iTLB-miss >>        0.69           +16.4%       0.80 ±  4%     +16.4%       0.81 ± >> 4%  perf-stat.i.ipc >>        0.22           -74.7%       0.05 ±  2%     -74.2%       0.06 ± >> 4%  perf-stat.i.metric.GHz >>      372.98           +51.6%     565.34 ± 12%     +67.2%     623.73 ± >> 14%  perf-stat.i.metric.K/sec >>       82.31 ±  2%     -69.6%      25.02 ±  5%     -69.1%      25.47 ± >> 5%  perf-stat.i.metric.M/sec >>    10453834 ±  3%     -92.0%     832079 ± 20%     -93.6%     669247 ± >> 20%  perf-stat.i.node-load-misses >>      879828           -91.8%      71803 ± 17%     -93.3%      59285 ± >> 21%  perf-stat.i.node-loads >>       96.34           -10.8       85.52 ±  6%     -17.0       79.29 ± >> 7%  perf-stat.i.node-store-miss-rate% >>    19057141 ±  2%     -96.0%     771338 ± 33%     -96.8%     610125 ± >> 33%  perf-stat.i.node-store-misses >>      379581 ±  4%     -80.4%      74536 ± 16%     -75.8%      91901 ± >> 10%  perf-stat.i.node-stores >>        1.15            +0.8        1.90 ±  6%      +0.9        2.02 ± >> 7%  perf-stat.overall.branch-miss-rate% >>       21.25 ±  6%     -15.2        6.08 ± 17%     -15.8        5.43 ± >> 12%  perf-stat.overall.cache-miss-rate% >>        1.46 ±  2%     -14.7%       1.24 ±  4%     -14.7%       1.24 ± >> 5%  perf-stat.overall.cpi >>      589.71 ±  2%    +203.1%       1787 ± 17%    +200.3%       1771 ± >> 12%  perf-stat.overall.cycles-between-cache-misses >>        0.04 ±  9%      +0.0        0.08 ± 21%      +0.1        0.10 ± >> 28%  perf-stat.overall.dTLB-load-miss-rate% >>       54.31            -9.4       44.87 ±  5%      -7.9       46.43 ± >> 2%  perf-stat.overall.iTLB-load-miss-rate% >>        4169 ±  3%     -15.5%       3521 ±  6%     -11.7%       3680 ± >> 3%  perf-stat.overall.instructions-per-iTLB-miss >>        0.69 ±  2%     +17.5%       0.81 ±  4%     +17.5%       0.81 ± >> 4%  perf-stat.overall.ipc >>       98.05            -8.1       89.98 ±  4%     -12.6       85.46 ± >> 6%  perf-stat.overall.node-store-miss-rate% >>   2.777e+09 ±  2%     -68.8%   8.67e+08 ±  5%     -68.0%  8.889e+08 ± >> 5%  perf-stat.ps.branch-instructions >>    31948261 ±  2%     -48.5%   16462168 ±  7%     -44.0%   17887178 ± >> 8%  perf-stat.ps.branch-misses >>    34551107           -91.4%    2973956 ± 15%     -91.3%    3018186 ± >> 13%  perf-stat.ps.cache-misses >>   1.632e+08 ±  6%     -69.7%   49465945 ± 13%     -65.6%   56209703 ± >> 15%  perf-stat.ps.cache-references >>      257331           -84.8%      39027 ±  6%     -85.5% >> 37233        perf-stat.ps.context-switches >>   2.037e+10           -74.6%  5.165e+09 ±  2%     -74.2%  5.257e+09 ± >> 3%  perf-stat.ps.cpu-cycles >>        1124 ±  6%     -55.0%     506.01 ± 18%     -59.2%     459.11 ± >> 16%  perf-stat.ps.cpu-migrations >>     1316371 ±  6%     -38.9%     804455 ± 21%     -21.6%    1032308 ± >> 26%  perf-stat.ps.dTLB-load-misses >>   3.258e+09 ±  3%     -69.4%  9.975e+08 ±  5%     -69.0%   1.01e+09 ± >> 5%  perf-stat.ps.dTLB-loads >>      241892 ±  3%     -66.3%      81415 ± 16%     -58.6%     100141 ± >> 23%  perf-stat.ps.dTLB-store-misses >>   1.578e+09 ±  2%     -68.3%  5.004e+08 ±  5%     -67.9%  5.068e+08 ± >> 5%  perf-stat.ps.dTLB-stores >>     3353793           -64.7%    1185377 ±  6%     -65.7%    1150426 ± >> 2%  perf-stat.ps.iTLB-load-misses >>     2821090           -48.3%    1457697 ±  7%     -52.8%    1330388 ± >> 6%  perf-stat.ps.iTLB-loads >>   1.398e+10 ±  2%     -70.2%  4.163e+09 ±  5%     -69.7%  4.236e+09 ± >> 5%  perf-stat.ps.instructions >>    10285826 ±  3%     -92.0%     818754 ± 20%     -93.6%     658429 ± >> 20%  perf-stat.ps.node-load-misses >>      865684           -91.8%      70644 ± 17%     -93.3%      58310 ± >> 21%  perf-stat.ps.node-loads >>    18750865 ±  2%     -96.0%     758931 ± 33%     -96.8%     600247 ± >> 33%  perf-stat.ps.node-store-misses >>      373508 ±  4%     -80.4%      73345 ± 16%     -75.8%      90413 ± >> 10%  perf-stat.ps.node-stores >>    8.75e+11 ±  2%     -69.9%  2.637e+11 ±  6%     -69.8%   2.64e+11 ± >> 5%  perf-stat.total.instructions >>       18.61           -10.1        8.48 ± 19%     -10.9        7.73 ± >> 7%  perf-profile.calltrace.cycles-pp.mount >>       18.56           -10.1        8.46 ± 19%     -10.8        7.71 ± >> 7% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.mount >>       18.56           -10.1        8.46 ± 19%     -10.8        7.71 ± >> 7%  perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.mount >>       18.52           -10.1        8.44 ± 19%     -10.8        7.70 ± >> 7% >> perf-profile.calltrace.cycles-pp.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe.mount >>       17.64            -9.4        8.22 ± 19%     -10.2        7.48 ± >> 7% >> perf-profile.calltrace.cycles-pp.do_new_mount.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe.mount >>       15.74            -7.5        8.29 ± 22%      -8.6        7.16 ± >> 7%  perf-profile.calltrace.cycles-pp.umount2 >>       15.58            -7.4        8.14 ± 21%      -8.5        7.07 ± >> 7% >> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.umount2 >>       15.56            -7.4        8.13 ± 21%      -8.5        7.06 ± >> 7% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>       13.19 ±  2%      -6.9        6.33 ± 20%      -7.7        5.51 ± >> 7% >> perf-profile.calltrace.cycles-pp.__x64_sys_umount.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>       11.28 ±  3%      -5.6        5.64 ± 22%      -6.4        4.89 ± >> 7% >> perf-profile.calltrace.cycles-pp.path_umount.__x64_sys_umount.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>        9.21 ±  2%      -5.5        3.71 ± 12%      -5.8        3.45 ± >> 5%  perf-profile.calltrace.cycles-pp.kthread.ret_from_fork >>        9.21 ±  2%      -5.5        3.71 ± 12%      -5.8        3.45 ± >> 5%  perf-profile.calltrace.cycles-pp.ret_from_fork >>       10.68 ±  3%      -5.3        5.34 ± 22%      -6.1        4.62 ± >> 7% >> perf-profile.calltrace.cycles-pp.do_umount.path_umount.__x64_sys_umount.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        5.36            -3.7        1.61 ± 15%      -3.9        1.50 ± >> 7% >> perf-profile.calltrace.cycles-pp.attach_recursive_mnt.do_new_mount.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        4.80            -3.4        1.40 ± 14%      -3.5        1.31 ± >> 7% >> perf-profile.calltrace.cycles-pp.propagate_mnt.attach_recursive_mnt.do_new_mount.__x64_sys_mount.do_syscall_64 >>        4.68            -3.3        1.37 ± 14%      -3.4        1.27 ± >> 8% >> perf-profile.calltrace.cycles-pp.propagate_one.propagate_mnt.attach_recursive_mnt.do_new_mount.__x64_sys_mount >>        3.70 ±  4%      -3.2        0.53 ± 46%      -3.2        0.49 ± >> 45% >> perf-profile.calltrace.cycles-pp.rcu_do_batch.rcu_core.__do_softirq.run_ksoftirqd.smpboot_thread_fn >>        4.49            -3.2        1.33 ± 14%      -3.3        1.23 ± >> 10% >> perf-profile.calltrace.cycles-pp.copy_tree.propagate_one.propagate_mnt.attach_recursive_mnt.do_new_mount >>        4.48            -3.2        1.33 ± 14%      -3.3        1.22 ± >> 10% >> perf-profile.calltrace.cycles-pp.clone_mnt.copy_tree.propagate_one.propagate_mnt.attach_recursive_mnt >>        3.75 ±  4%      -3.1        0.64 ± 15%      -3.2        0.51 ± >> 45% >> perf-profile.calltrace.cycles-pp.__do_softirq.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork >>        3.75 ±  4%      -3.1        0.65 ± 15%      -3.2        0.51 ± >> 45% >> perf-profile.calltrace.cycles-pp.run_ksoftirqd.smpboot_thread_fn.kthread.ret_from_fork >>        3.72 ±  4%      -3.1        0.63 ± 14%      -3.2        0.50 ± >> 45% >> perf-profile.calltrace.cycles-pp.rcu_core.__do_softirq.run_ksoftirqd.smpboot_thread_fn.kthread >>        3.82 ±  4%      -3.0        0.87 ± 10%      -3.0        0.86 ± >> 10% >> perf-profile.calltrace.cycles-pp.smpboot_thread_fn.kthread.ret_from_fork >>        4.95 ±  2%      -2.8        2.11 ± 11%      -3.1        1.90 ± >> 5% >> perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork >>        4.05            -2.8        1.22 ± 14%      -2.9        1.12 ± >> 10% >> perf-profile.calltrace.cycles-pp.alloc_vfsmnt.clone_mnt.copy_tree.propagate_one.propagate_mnt >>        5.00            -2.8        2.24 ± 29%      -3.0        2.03 ± >> 9% >> perf-profile.calltrace.cycles-pp.vfs_get_super.vfs_get_tree.do_new_mount.__x64_sys_mount.do_syscall_64 >>        5.00            -2.8        2.25 ± 29%      -3.0        2.03 ± >> 9% >> perf-profile.calltrace.cycles-pp.vfs_get_tree.do_new_mount.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        5.34 ±  2%      -2.6        2.73 ± 13%      -2.9        2.48 ± >> 5%  perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork >>        3.49 ±  2%      -2.4        1.05 ± 13%      -2.5        0.96 ± >> 11% >> perf-profile.calltrace.cycles-pp.pcpu_alloc.alloc_vfsmnt.clone_mnt.copy_tree.propagate_one >>        4.63 ±  2%      -2.3        2.30 ± 24%      -2.6        2.01 ± >> 8% >> perf-profile.calltrace.cycles-pp.rwsem_down_write_slowpath.down_write.do_umount.path_umount.__x64_sys_umount >>        4.66 ±  3%      -2.3        2.36 ± 22%      -2.6        2.02 ± >> 8% >> perf-profile.calltrace.cycles-pp.down_write.do_umount.path_umount.__x64_sys_umount.do_syscall_64 >>        4.05            -2.3        1.78 ± 29%      -2.4        1.64 ± >> 9% >> perf-profile.calltrace.cycles-pp.sget_fc.vfs_get_super.vfs_get_tree.do_new_mount.__x64_sys_mount >>        3.92            -2.2        1.72 ± 30%      -2.3        1.58 ± >> 9% >> perf-profile.calltrace.cycles-pp.alloc_super.sget_fc.vfs_get_super.vfs_get_tree.do_new_mount >>        3.63 ±  2%      -2.1        1.58 ± 31%      -2.2        1.47 ± >> 11% >> perf-profile.calltrace.cycles-pp.__percpu_init_rwsem.alloc_super.sget_fc.vfs_get_super.vfs_get_tree >>        3.57 ±  2%      -2.0        1.56 ± 31%      -2.1        1.46 ± >> 11% >> perf-profile.calltrace.cycles-pp.pcpu_alloc.__percpu_init_rwsem.alloc_super.sget_fc.vfs_get_super >>        4.10 ±  5%      -1.9        2.20 ± 22%      -2.2        1.90 ± >> 8% >> perf-profile.calltrace.cycles-pp.namespace_unlock.do_umount.path_umount.__x64_sys_umount.do_syscall_64 >>        3.62            -1.6        2.00 ± 25%      -1.9        1.70 ± >> 11% >> perf-profile.calltrace.cycles-pp.rwsem_optimistic_spin.rwsem_down_write_slowpath.down_write.do_umount.path_umount >>        2.80            -1.5        1.26 ± 10%      -1.7        1.15 ± >> 13% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64 >>        2.81            -1.5        1.26 ± 10%      -1.7        1.15 ± >> 13% >> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.open64 >>        2.72            -1.5        1.18 ±  7%      -1.6        1.09 ± >> 11% >> perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64 >>        2.72            -1.5        1.17 ±  7%      -1.6        1.08 ± >> 11% >> perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.open64 >>        2.84            -1.5        1.30 ± 10%      -1.6        1.20 ± >> 13%  perf-profile.calltrace.cycles-pp.open64 >>        2.02            -1.5        0.52 ± 49%      -1.7        0.36 ± >> 71% >> perf-profile.calltrace.cycles-pp.memset_erms.pcpu_alloc.__percpu_init_rwsem.alloc_super.sget_fc >>        2.02 ±  2%      -1.5        0.56 ± 45%      -1.4        0.61 ± >> 11% >> perf-profile.calltrace.cycles-pp.memset_erms.pcpu_alloc.alloc_vfsmnt.clone_mnt.copy_tree >>        1.71 ±  2%      -1.4        0.32 ±101%      -1.2        0.47 ± >> 45% >> perf-profile.calltrace.cycles-pp.vfs_create_mount.do_new_mount.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        1.90 ±  3%      -1.3        0.61 ± 46%      -1.3        0.61 ± >> 13% >> perf-profile.calltrace.cycles-pp.user_path_at_empty.__x64_sys_umount.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>        4.78 ±  2%      -1.3        3.51 ± 17%      -1.6        3.16 ± >> 7% >> perf-profile.calltrace.cycles-pp.lock_mount.do_new_mount.__x64_sys_mount.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        2.62            -1.2        1.44 ± 17%      -1.5        1.15 ± >> 19% >> perf-profile.calltrace.cycles-pp.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64 >>        2.63            -1.2        1.45 ± 18%      -1.5        1.16 ± >> 19% >> perf-profile.calltrace.cycles-pp.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        4.57 ±  2%      -1.1        3.46 ± 17%      -1.4        3.12 ± >> 7% >> perf-profile.calltrace.cycles-pp.down_write.lock_mount.do_new_mount.__x64_sys_mount.do_syscall_64 >>        4.50 ±  2%      -1.1        3.44 ± 17%      -1.4        3.11 ± >> 7% >> perf-profile.calltrace.cycles-pp.rwsem_down_write_slowpath.down_write.lock_mount.do_new_mount.__x64_sys_mount >>        1.27 ±  5%      -1.1        0.21 ±142%      -1.0        0.27 >> ±100% >> perf-profile.calltrace.cycles-pp.alloc_vfsmnt.vfs_create_mount.do_new_mount.__x64_sys_mount.do_syscall_64 >>        1.37 ±  3%      -1.0        0.32 ±103%      -1.1        0.27 >> ±100% >> perf-profile.calltrace.cycles-pp.propagate_mount_busy.do_umount.path_umount.__x64_sys_umount.do_syscall_64 >>        2.28 ±  3%      -1.0        1.25 ± 19%      -1.2        1.10 ± >> 8% >> perf-profile.calltrace.cycles-pp.mntput_no_expire.namespace_unlock.do_umount.path_umount.__x64_sys_umount >>        1.26 ±  3%      -0.9        0.38 ±100%      -1.0        0.27 >> ±100% >> perf-profile.calltrace.cycles-pp.__sync_rcu_exp_select_node_cpus.process_one_work.worker_thread.kthread.ret_from_fork >>        1.13 ±  7%      -0.8        0.36 ±100%      -1.0        0.10 >> ±223%  perf-profile.calltrace.cycles-pp.unlink >>        1.10 ±  7%      -0.8        0.35 ±100%      -1.0        0.10 >> ±223% >> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.unlink >>        1.06 ±  4%      -0.8        0.31 ±100%      -1.0        0.10 >> ±223% >> perf-profile.calltrace.cycles-pp.__x64_sys_unlink.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink >>        1.10 ±  7%      -0.8        0.35 ±100%      -1.0        0.10 >> ±223% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink >>        1.01 ±  5%      -0.7        0.30 ±100%      -0.9        0.09 >> ±223% >> perf-profile.calltrace.cycles-pp.do_unlinkat.__x64_sys_unlink.do_syscall_64.entry_SYSCALL_64_after_hwframe.unlink >>        1.55 ±  9%      -0.7        0.85 ± 25%      -0.8        0.74 ± >> 9% >> perf-profile.calltrace.cycles-pp.synchronize_rcu_expedited.namespace_unlock.do_umount.path_umount.__x64_sys_umount >>        1.56 ±  3%      -0.7        0.86 ± 25%      -0.8        0.72 ± >> 8% >> perf-profile.calltrace.cycles-pp.rwsem_spin_on_owner.rwsem_optimistic_spin.rwsem_down_write_slowpath.down_write.lock_mount >>        0.97 ±  7%      -0.6        0.37 ± 70%      -0.8        0.18 >> ±141%  perf-profile.calltrace.cycles-pp.__lxstat64 >>        2.26 ±  2%      -0.6        1.68 ± 23%      -0.7        1.51 ± >> 8% >> 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 >>        0.94 ±  7%      -0.6        0.36 ± 70%      -0.8        0.17 >> ±141% >> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__lxstat64 >>        2.36 ±  6%      -0.6        1.79 ± 26%      -0.8        1.54 ± >> 7% >> perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>        0.93 ±  7%      -0.6        0.36 ± 70%      -0.8        0.09 >> ±223% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__lxstat64 >>        2.27 ±  2%      -0.6        1.70 ± 22%      -0.7        1.53 ± >> 8% >> perf-profile.calltrace.cycles-pp.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.umount2 >>        2.24 ±  2%      -0.6        1.67 ± 23%      -0.7        1.51 ± >> 7% >> perf-profile.calltrace.cycles-pp.task_work_run.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.do_syscall_64 >>        0.82 ±  3%      -0.5        0.29 ±100%      -0.8 >> 0.00 >> perf-profile.calltrace.cycles-pp.smp_call_function_single.__sync_rcu_exp_select_node_cpus.process_one_work.worker_thread.kthread >>        2.14            -0.5        1.63 ± 23%      -0.7        1.46 ± >> 8% >> perf-profile.calltrace.cycles-pp.cleanup_mnt.task_work_run.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode >>        1.18 ±  5%      -0.4        0.78 ± 19%      -0.5        0.65 ± >> 10% >> perf-profile.calltrace.cycles-pp._raw_spin_lock.mntput_no_expire.namespace_unlock.do_umount.path_umount >>        1.02 ±  6%      -0.3        0.72 ± 20%      -0.4        0.61 ± >> 9% >> perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.mntput_no_expire.namespace_unlock.do_umount >>        0.00            +0.4        0.43 ± 73%      +0.7        0.74 ± >> 11% >> perf-profile.calltrace.cycles-pp.tick_nohz_get_sleep_length.menu_select.cpuidle_idle_call.do_idle.cpu_startup_entry >>        0.00            +0.5        0.55 ±  7%      +0.4        0.42 ± >> 71% >> perf-profile.calltrace.cycles-pp.clockevents_program_event.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt >>        0.00            +0.6        0.65 ± 19%      +0.2        0.23 >> ±145%  perf-profile.calltrace.cycles-pp.__close >>        1.68 ±  6%      +0.7        2.37 ± 17%      +0.5        2.19 ± >> 9% >> perf-profile.calltrace.cycles-pp.osq_lock.rwsem_optimistic_spin.rwsem_down_write_slowpath.down_write.lock_mount >>        0.00            +0.7        0.69 ± 15%      +0.7        0.67 ± >> 9% >> perf-profile.calltrace.cycles-pp.rebalance_domains.__do_softirq.__irq_exit_rcu.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt >>        0.00            +0.7        0.70 ± 15%      +0.5        0.51 ± >> 45% >> perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.openat64 >>        0.00            +0.7        0.70 ± 14%      +0.5        0.52 ± >> 45% >> perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.openat64.device_set_syspath >>        0.00            +0.8        0.79 ± 15%      +0.8        0.79 ± >> 7% >> perf-profile.calltrace.cycles-pp.perf_mux_hrtimer_handler.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt >>        0.00            +0.8        0.80 ± 13%      +0.5        0.55 ± >> 46% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.openat64.device_set_syspath.sd_device_new_from_syspath >>        0.00            +0.8        0.81 ± 13%      +0.6        0.55 ± >> 46% >> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.openat64.device_set_syspath.sd_device_new_from_syspath >>        0.00            +0.9        0.86 ± 26%      +0.8        0.82 ± >> 21%  perf-profile.calltrace.cycles-pp.siphash24_compress >>        0.00            +0.9        0.87 ± 13%      +0.6        0.60 ± >> 46% >> perf-profile.calltrace.cycles-pp.openat64.device_set_syspath.sd_device_new_from_syspath >>        0.00            +1.2        1.15 ±  9%      +0.9        0.88 ± >> 23% >> perf-profile.calltrace.cycles-pp.show_mountinfo.seq_read_iter.vfs_read.ksys_read.do_syscall_64 >>        0.18 ±141%      +1.2        1.39 ± 11%      +0.8        1.03 ± >> 23%  perf-profile.calltrace.cycles-pp.sd_device_new_from_syspath >>        0.00            +1.2        1.24 ± 13%      +0.9        0.92 ± >> 23% >> perf-profile.calltrace.cycles-pp.device_set_syspath.sd_device_new_from_syspath >>        0.00            +1.4        1.43 ±  9%      +1.1        1.09 ± >> 23% >> perf-profile.calltrace.cycles-pp.seq_read_iter.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe >>        0.18 ±141%      +1.5        1.68 ±  7%      +1.3        1.46 ± >> 11% >> perf-profile.calltrace.cycles-pp.scheduler_tick.update_process_times.tick_sched_handle.tick_sched_timer.__hrtimer_run_queues >>        0.00            +1.6        1.55 ± 10%      +1.2        1.20 ± >> 21% >> perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read >>        0.00            +1.6        1.56 ± 10%      +1.2        1.21 ± >> 20% >> perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read >>        0.00            +1.6        1.61 ± 10%      +1.2        1.24 ± >> 22% >> perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read >>        0.00            +1.6        1.62 ± 10%      +1.2        1.24 ± >> 22%  perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read >>        0.00            +1.6        1.64 ± 10%      +1.3        1.26 ± >> 22%  perf-profile.calltrace.cycles-pp.read >>        0.00            +1.7        1.66 ± 12%      +1.8        1.81 ± >> 6% >> perf-profile.calltrace.cycles-pp.menu_select.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary >>        0.00            +1.8        1.83 ±  9%      +1.9        1.86 ± >> 9% >> perf-profile.calltrace.cycles-pp.__do_softirq.__irq_exit_rcu.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state >>        0.85 ±  4%      +2.1        2.92 ±  4%      +1.8        2.64 ± >> 8% >> perf-profile.calltrace.cycles-pp.update_process_times.tick_sched_handle.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt >>        0.00            +2.1        2.08 ±  8%      +2.1        2.11 ± >> 8% >> perf-profile.calltrace.cycles-pp.__irq_exit_rcu.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter >>        0.86 ±  4%      +2.2        3.05 ±  4%      +1.9        2.80 ± >> 9% >> perf-profile.calltrace.cycles-pp.tick_sched_handle.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt >>        0.99 ±  5%      +2.6        3.57 ±  3%      +2.3        3.31 ± >> 10% >> perf-profile.calltrace.cycles-pp.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt >>        1.22 ±  5%      +3.8        5.03 ±  4%      +3.7        4.92 ± >> 8% >> perf-profile.calltrace.cycles-pp.__hrtimer_run_queues.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt >>        1.41 ±  5%      +4.8        6.24 ±  3%      +4.8        6.19 ± >> 8% >> perf-profile.calltrace.cycles-pp.hrtimer_interrupt.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state >>        1.42 ±  5%      +4.9        6.35 ±  3%      +4.9        6.34 ± >> 8% >> perf-profile.calltrace.cycles-pp.__sysvec_apic_timer_interrupt.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter >>        2.01 ±  4%      +7.2        9.24 ±  5%      +7.4        9.42 ± >> 7% >> perf-profile.calltrace.cycles-pp.sysvec_apic_timer_interrupt.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call >>        2.14 ±  3%      +8.6       10.71 ±  5%      +8.9       11.06 ± >> 6% >> perf-profile.calltrace.cycles-pp.asm_sysvec_apic_timer_interrupt.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle >>       36.36            +9.8       46.14 ±  5%     +12.7       49.03 ± >> 3% >> perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle >>       40.38           +17.3       57.69 ±  6%     +20.8       61.13 ± >> 3% >> perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry >>       40.51           +17.8       58.31 ±  5%     +21.5       62.00 ± >> 3% >> perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary >>       42.41           +19.3       61.74 ±  5%     +23.2       65.66 ± >> 3% >> perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64_no_verify >>       42.42           +19.4       61.78 ±  5%     +23.3       65.71 ± >> 3% >> perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64_no_verify >>       42.42           +19.4       61.78 ±  5%     +23.3       65.71 ± >> 3% >> perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64_no_verify >>       41.01           +19.4       60.43 ±  6%     +23.3       64.35 ± >> 3% >> perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64_no_verify >>       42.74           +19.6       62.33 ±  5%     +23.4       66.10 ± >> 3%  perf-profile.calltrace.cycles-pp.secondary_startup_64_no_verify >>       43.68           -18.1       25.61 ± 15%     -21.4       22.32 ± >> 5%  perf-profile.children.cycles-pp.do_syscall_64 >>       43.73           -18.0       25.70 ± 15%     -21.3       22.39 ± >> 5%  perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe >>       18.62           -10.1        8.49 ± 19%     -10.9        7.74 ± >> 7%  perf-profile.children.cycles-pp.mount >>       18.52           -10.1        8.44 ± 19%     -10.8        7.70 ± >> 7%  perf-profile.children.cycles-pp.__x64_sys_mount >>       17.64            -9.4        8.22 ± 19%     -10.2        7.48 ± >> 7%  perf-profile.children.cycles-pp.do_new_mount >>       15.76            -7.5        8.30 ± 22%      -8.6        7.18 ± >> 7%  perf-profile.children.cycles-pp.umount2 >>       13.19 ±  2%      -6.9        6.34 ± 20%      -7.7        5.51 ± >> 7%  perf-profile.children.cycles-pp.__x64_sys_umount >>       11.28 ±  3%      -5.6        5.64 ± 22%      -6.4        4.89 ± >> 7%  perf-profile.children.cycles-pp.path_umount >>        7.26 ±  2%      -5.6        1.67 ± 19%      -5.7        1.56 ± >> 9%  perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath >>        9.21 ±  2%      -5.5        3.71 ± 12%      -5.8        3.45 ± >> 5%  perf-profile.children.cycles-pp.kthread >>        9.21 ±  2%      -5.5        3.72 ± 12%      -5.8        3.46 ± >> 5%  perf-profile.children.cycles-pp.ret_from_fork >>        8.72            -5.4        3.32 ± 23%      -5.7        3.06 ± >> 9%  perf-profile.children.cycles-pp.pcpu_alloc >>        6.56 ±  3%      -5.4        1.21 ± 16%      -5.4        1.20 ± >> 12%  perf-profile.children.cycles-pp._raw_spin_lock_irqsave >>       10.69 ±  3%      -5.3        5.34 ± 22%      -6.1        4.62 ± >> 7%  perf-profile.children.cycles-pp.do_umount >>        5.24 ±  5%      -4.6        0.67 ± 24%      -4.6        0.66 ± >> 10%  perf-profile.children.cycles-pp.free_percpu >>        5.36            -3.7        1.61 ± 15%      -3.9        1.50 ± >> 7%  perf-profile.children.cycles-pp.attach_recursive_mnt >>        5.32 ±  2%      -3.6        1.70 ± 16%      -3.7        1.58 ± >> 11%  perf-profile.children.cycles-pp.alloc_vfsmnt >>        5.12            -3.5        1.58 ± 19%      -3.6        1.48 ± >> 9%  perf-profile.children.cycles-pp.memset_erms >>        9.34 ±  2%      -3.5        5.87 ± 19%      -4.1        5.20 ± >> 7%  perf-profile.children.cycles-pp.down_write >>        4.80            -3.4        1.40 ± 14%      -3.5        1.31 ± >> 7%  perf-profile.children.cycles-pp.propagate_mnt >>        9.13 ±  2%      -3.4        5.74 ± 20%      -4.0        5.15 ± >> 7%  perf-profile.children.cycles-pp.rwsem_down_write_slowpath >>        4.69            -3.3        1.37 ± 14%      -3.4        1.27 ± >> 8%  perf-profile.children.cycles-pp.propagate_one >>        4.49            -3.2        1.33 ± 14%      -3.3        1.23 ± >> 10%  perf-profile.children.cycles-pp.copy_tree >>        4.48            -3.2        1.33 ± 14%      -3.3        1.22 ± >> 10%  perf-profile.children.cycles-pp.clone_mnt >>        3.75 ±  4%      -3.1        0.65 ± 15%      -3.2        0.60 ± >> 10%  perf-profile.children.cycles-pp.run_ksoftirqd >>        4.02 ±  4%      -3.1        0.94 ± 11%      -3.1        0.89 ± >> 8%  perf-profile.children.cycles-pp.rcu_do_batch >>        3.82 ±  4%      -3.0        0.87 ± 10%      -3.0        0.86 ± >> 10%  perf-profile.children.cycles-pp.smpboot_thread_fn >>        3.41 ±  4%      -2.9        0.47 ± 21%      -3.0        0.46 ± >> 9%  perf-profile.children.cycles-pp.delayed_free_vfsmnt >>        4.08 ±  4%      -2.9        1.18 ± 10%      -3.0        1.12 ± >> 7%  perf-profile.children.cycles-pp.rcu_core >>        3.96 ±  3%      -2.9        1.10 ± 15%      -3.0        0.98 ± >> 9%  perf-profile.children.cycles-pp.path_init >>        4.95 ±  2%      -2.8        2.11 ± 11%      -3.1        1.90 ± >> 5%  perf-profile.children.cycles-pp.process_one_work >>        5.00            -2.8        2.24 ± 29%      -3.0        2.03 ± >> 9%  perf-profile.children.cycles-pp.vfs_get_super >>        5.00            -2.8        2.25 ± 29%      -3.0        2.03 ± >> 9%  perf-profile.children.cycles-pp.vfs_get_tree >>        3.86            -2.7        1.21 ± 21%      -2.8        1.08 ± >> 9%  perf-profile.children.cycles-pp.rwsem_spin_on_owner >>        5.35 ±  2%      -2.6        2.73 ± 13%      -2.9        2.48 ± >> 5%  perf-profile.children.cycles-pp.worker_thread >>        4.05            -2.3        1.78 ± 29%      -2.4        1.64 ± >> 9%  perf-profile.children.cycles-pp.sget_fc >>        3.93            -2.2        1.72 ± 30%      -2.3        1.58 ± >> 9%  perf-profile.children.cycles-pp.alloc_super >>        4.35 ±  4%      -2.1        2.27 ± 22%      -2.4        1.94 ± >> 8%  perf-profile.children.cycles-pp.namespace_unlock >>        3.63 ±  2%      -2.1        1.58 ± 31%      -2.2        1.48 ± >> 11%  perf-profile.children.cycles-pp.__percpu_init_rwsem >>        5.42 ± 12%      -2.0        3.45 ± 13%      -2.2        3.18 ± >> 5%  perf-profile.children.cycles-pp.__schedule >>        4.44 ±  3%      -1.8        2.61 ±  7%      -1.9        2.57 ± >> 5%  perf-profile.children.cycles-pp.__do_softirq >>        4.15 ±  2%      -1.8        2.34 ± 12%      -2.1        2.08 ± >> 5%  perf-profile.children.cycles-pp._raw_spin_lock >>        3.17 ±  3%      -1.8        1.38 ± 13%      -2.0        1.18 ± >> 10%  perf-profile.children.cycles-pp.filename_lookup >>        2.46 ± 17%      -1.8        0.67 ± 19%      -1.8        0.62 ± >> 6%  perf-profile.children.cycles-pp.schedule_preempt_disabled >>        3.41 ±  3%      -1.7        1.68 ± 10%      -2.0        1.44 ± >> 11%  perf-profile.children.cycles-pp.path_lookupat >>        6.95 ±  2%      -1.6        5.31 ± 20%      -2.3        4.70 ± >> 8%  perf-profile.children.cycles-pp.rwsem_optimistic_spin >>        1.87 ±  6%      -1.6        0.26 ± 27%      -1.6        0.26 ± >> 17%  perf-profile.children.cycles-pp.percpu_free_rwsem >>        1.87 ±  6%      -1.6        0.26 ± 27%      -1.6        0.26 ± >> 18%  perf-profile.children.cycles-pp.destroy_super_work >>        2.68 ±  2%      -1.6        1.10 ± 15%      -1.7        0.93 ± >> 9%  perf-profile.children.cycles-pp.user_path_at_empty >>        2.84            -1.5        1.31 ± 10%      -1.6        1.22 ± >> 13%  perf-profile.children.cycles-pp.open64 >>        4.09 ± 13%      -1.5        2.63 ± 17%      -1.6        2.47 ± >> 7%  perf-profile.children.cycles-pp.schedule >>        2.97            -1.3        1.65 ±  6%      -1.5        1.49 ± >> 11%  perf-profile.children.cycles-pp.path_openat >>        2.98            -1.3        1.67 ±  6%      -1.5        1.50 ± >> 11%  perf-profile.children.cycles-pp.do_filp_open >>        2.19 ±  2%      -1.3        0.91 ± 24%      -1.3        0.89 ± >> 7%  perf-profile.children.cycles-pp.mnt_get_count >>        4.78 ±  2%      -1.3        3.51 ± 17%      -1.6        3.16 ± >> 7%  perf-profile.children.cycles-pp.lock_mount >>        2.72 ±  3%      -1.2        1.49 ± 19%      -1.4        1.33 ± >> 7%  perf-profile.children.cycles-pp.mntput_no_expire >>        1.43 ±  8%      -1.2        0.23 ± 41%      -1.1        0.31 ± >> 40%  perf-profile.children.cycles-pp.intel_idle_irq >>        1.71 ±  2%      -1.2        0.53 ± 23%      -1.2        0.54 ± >> 11%  perf-profile.children.cycles-pp.vfs_create_mount >>        3.11            -1.2        1.95 ±  5%      -1.4        1.74 ± >> 11%  perf-profile.children.cycles-pp.do_sys_openat2 >>        3.11            -1.1        1.97 ±  5%      -1.4        1.76 ± >> 11%  perf-profile.children.cycles-pp.__x64_sys_openat >>        3.52 ± 16%      -1.1        2.39 ± 18%      -1.3        2.25 ± >> 7%  perf-profile.children.cycles-pp.pick_next_task_fair >>        1.74 ±  2%      -1.1        0.65 ±  7%      -1.1        0.60 ± >> 7%  perf-profile.children.cycles-pp.step_into >>        2.43 ±  3%      -1.0        1.39 ±  8%      -1.3        1.16 ± >> 10%  perf-profile.children.cycles-pp.link_path_walk >>        1.68 ±  6%      -1.0        0.69 ± 18%      -1.1        0.59 ± >> 8%  perf-profile.children.cycles-pp.filename_parentat >>        3.21 ± 17%      -1.0        2.22 ± 18%      -1.1        2.07 ± >> 6%  perf-profile.children.cycles-pp.newidle_balance >>        1.66 ±  6%      -1.0        0.68 ± 18%      -1.1        0.58 ± >> 7%  perf-profile.children.cycles-pp.path_parentat >>        1.26 ±  4%      -0.9        0.38 ± 16%      -0.9        0.34 ± >> 7%  perf-profile.children.cycles-pp.__traverse_mounts >>        1.50 ±  4%      -0.9        0.63 ± 24%      -0.9        0.57 ± >> 6%  perf-profile.children.cycles-pp.__sync_rcu_exp_select_node_cpus >>        1.17 ±  3%      -0.9        0.30 ± 12%      -0.9        0.30 ± >> 10%  perf-profile.children.cycles-pp.asm_sysvec_call_function_single >>        1.37 ±  3%      -0.9        0.51 ± 31%      -0.9        0.50 ± >> 9%  perf-profile.children.cycles-pp.propagate_mount_busy >>        1.21 ±  3%      -0.8        0.37 ± 18%      -0.9        0.31 ± >> 5%  perf-profile.children.cycles-pp.lookup_mnt >>        1.55 ±  9%      -0.7        0.85 ± 25%      -0.8        0.74 ± >> 9%  perf-profile.children.cycles-pp.synchronize_rcu_expedited >>        1.05 ±  2%      -0.7        0.38 ± 23%      -0.7        0.36 ± >> 8%  perf-profile.children.cycles-pp.filename_create >>        2.74 ±  2%      -0.6        2.13 ± 17%      -0.8        1.92 ± >> 7%  perf-profile.children.cycles-pp.exit_to_user_mode_loop >>        2.78 ±  2%      -0.6        2.17 ± 16%      -0.8        1.96 ± >> 7%  perf-profile.children.cycles-pp.exit_to_user_mode_prepare >>        1.04 ±  4%      -0.6        0.47 ± 24%      -0.6        0.40 ± >> 7%  perf-profile.children.cycles-pp.symlink >>        1.14 ±  7%      -0.6        0.58 ± 24%      -0.7        0.48 ± >> 13%  perf-profile.children.cycles-pp.unlink >>        0.99 ±  3%      -0.6        0.43 ± 21%      -0.6        0.38 ± >> 5%  perf-profile.children.cycles-pp.__x64_sys_symlink >>        1.16            -0.6        0.61 ± 24%      -0.6        0.60 ± >> 5%  perf-profile.children.cycles-pp._find_next_bit >>        0.79 ±  8%      -0.5        0.25 ± 19%      -0.5        0.28 ± >> 14%  perf-profile.children.cycles-pp.wait_rcu_exp_gp >>        0.79 ±  8%      -0.5        0.25 ± 18%      -0.5        0.28 ± >> 14%  perf-profile.children.cycles-pp.sync_rcu_exp_select_cpus >>        0.94 ±  3%      -0.5        0.40 ± 22%      -0.6        0.35 ± >> 6%  perf-profile.children.cycles-pp.do_symlinkat >>        0.71 ±  5%      -0.5        0.18 ± 17%      -0.5        0.18 ± >> 13%  perf-profile.children.cycles-pp.sysvec_call_function_single >>        1.06 ±  4%      -0.5        0.54 ± 18%      -0.6        0.45 ± >> 14%  perf-profile.children.cycles-pp.__x64_sys_unlink >>        0.91 ±  3%      -0.5        0.40 ±  9%      -0.6        0.33 ± >> 12%  perf-profile.children.cycles-pp.__cond_resched >>        2.14            -0.5        1.63 ± 23%      -0.7        1.46 ± >> 8%  perf-profile.children.cycles-pp.cleanup_mnt >>        2.56 ±  2%      -0.5        2.05 ± 17%      -0.7        1.86 ± >> 7%  perf-profile.children.cycles-pp.task_work_run >>        1.01 ±  5%      -0.5        0.51 ± 19%      -0.6        0.42 ± >> 14%  perf-profile.children.cycles-pp.do_unlinkat >>        1.01 ±  2%      -0.5        0.52 ± 25%      -0.5        0.48 ± >> 10%  perf-profile.children.cycles-pp.stress_rndstr >>        0.98 ±  5%      -0.5        0.49 ± 25%      -0.6        0.42 ± >> 8%  perf-profile.children.cycles-pp.smp_call_function_single >>        1.01 ±  7%      -0.5        0.55 ± 11%      -0.5        0.50 ± >> 12%  perf-profile.children.cycles-pp.__lxstat64 >>        0.92 ±  4%      -0.4        0.47 ± 10%      -0.5        0.45 ± >> 10%  perf-profile.children.cycles-pp.__do_sys_newlstat >>        1.13 ±  5%      -0.4        0.68 ± 10%      -0.5        0.64 ± >> 12%  perf-profile.children.cycles-pp.vfs_statx >>        0.78 ±  2%      -0.4        0.34 ± 18%      -0.5        0.31 ± >> 7%  perf-profile.children.cycles-pp.generic_shutdown_super >>        0.78 ±  2%      -0.4        0.35 ± 18%      -0.5        0.32 ± >> 8%  perf-profile.children.cycles-pp.kill_litter_super >>        0.57 ±  6%      -0.4        0.14 ± 19%      -0.4        0.14 ± >> 18%  perf-profile.children.cycles-pp.__sysvec_call_function_single >>        0.55 ±  3%      -0.4        0.13 ± 36%      -0.4        0.13 ± >> 22%  perf-profile.children.cycles-pp.rcu_report_exp_cpu_mult >>        0.58 ±  4%      -0.4        0.16 ± 16%      -0.4        0.17 ± >> 18%  perf-profile.children.cycles-pp.__flush_smp_call_function_queue >>        1.20 ±  5%      -0.4        0.78 ± 11%      -0.5        0.73 ± >> 12%  perf-profile.children.cycles-pp.vfs_fstatat >>        0.78 ±  5%      -0.4        0.37 ± 36%      -0.5        0.31 ± >> 15%  perf-profile.children.cycles-pp.shmem_fill_super >>        0.61 ±  6%      -0.4        0.20 ± 25%      -0.4        0.18 ± >> 20%  perf-profile.children.cycles-pp.pcpu_alloc_area >>        0.88 ±  3%      -0.4        0.48 ±  4%      -0.4        0.48 ± >> 12%  perf-profile.children.cycles-pp.dput >>        0.58 ±  4%      -0.4        0.19 ± 35%      -0.4        0.13 ± >> 14%  perf-profile.children.cycles-pp.up_write >>        0.50 ±  3%      -0.4        0.12 ± 27%      -0.4        0.11 ± >> 27%  perf-profile.children.cycles-pp.rcu_note_context_switch >>        0.60 ±  5%      -0.4        0.23 ± 11%      -0.4        0.19 ± >> 7%  perf-profile.children.cycles-pp.__legitimize_mnt >>        0.71 ±  3%      -0.4        0.35 ± 27%      -0.4        0.32 ± >> 13%  perf-profile.children.cycles-pp.mnt_get_writers >>        0.56 ±  2%      -0.4        0.20 ± 21%      -0.4        0.18 ± >> 17%  perf-profile.children.cycles-pp.rcu_exp_wait_wake >>        0.66 ±  6%      -0.4        0.31 ± 29%      -0.4        0.28 ± >> 4%  perf-profile.children.cycles-pp.mkdir >>        0.62 ±  4%      -0.3        0.28 ± 32%      -0.4        0.26 ± >> 4%  perf-profile.children.cycles-pp.__x64_sys_mkdir >>        0.59 ±  4%      -0.3        0.26 ± 35%      -0.3        0.26 ± >> 8%  perf-profile.children.cycles-pp.do_mkdirat >>        0.89 ±  3%      -0.3        0.59 ± 14%      -0.4        0.46 ± >> 9%  perf-profile.children.cycles-pp.try_to_wake_up >>        0.52 ±  2%      -0.3        0.23 ± 23%      -0.2        0.27 ± >> 12%  perf-profile.children.cycles-pp.open_last_lookups >>        0.44 ±  5%      -0.3        0.15 ± 34%      -0.3        0.11 ± >> 10%  perf-profile.children.cycles-pp.rwsem_wake >>        0.45 ±  3%      -0.3        0.18 ± 23%      -0.3        0.17 ± >> 20%  perf-profile.children.cycles-pp.new_inode >>        0.39 ±  8%      -0.3        0.13 ± 55%      -0.3        0.12 ± >> 16%  perf-profile.children.cycles-pp.ida_alloc_range >>        0.52 ±  8%      -0.3        0.26 ± 25%      -0.3        0.26 ± >> 19%  perf-profile.children.cycles-pp.rmdir >>        0.54 ±  4%      -0.3        0.28 ± 37%      -0.3        0.23 ± >> 12%  perf-profile.children.cycles-pp.__percpu_counter_init >>        0.34 ±  3%      -0.3        0.08 ± 59%      -0.3        0.06 ± >> 13%  perf-profile.children.cycles-pp.unlock_mount >>        0.48 ±  4%      -0.2        0.24 ± 20%      -0.3        0.23 ± >> 18%  perf-profile.children.cycles-pp.__x64_sys_rmdir >>        0.40 ±  4%      -0.2        0.16 ± 15%      -0.2        0.17 ± >> 10%  perf-profile.children.cycles-pp.sync_filesystem >>        0.46 ±  4%      -0.2        0.22 ± 20%      -0.2        0.21 ± >> 18%  perf-profile.children.cycles-pp.do_rmdir >>        0.39 ±  5%      -0.2        0.15 ± 14%      -0.2        0.16 ± >> 12%  perf-profile.children.cycles-pp.writeback_inodes_sb >>        0.39 ±  4%      -0.2        0.15 ± 34%      -0.2        0.18 ± >> 12%  perf-profile.children.cycles-pp.lookup_open >>        0.33 ± 10%      -0.2        0.10 ± 35%      -0.2        0.10 ± >> 23%  perf-profile.children.cycles-pp.pcpu_free_area >>        0.38 ±  6%      -0.2        0.15 ± 11%      -0.2        0.16 ± >> 12%  perf-profile.children.cycles-pp.get_nr_dirty_inodes >>        0.45            -0.2        0.22 ± 26%      -0.3        0.15 ± >> 21%  perf-profile.children.cycles-pp.umount_tree >>        0.46 ±  3%      -0.2        0.24 ± 26%      -0.2        0.24 ± >> 20%  perf-profile.children.cycles-pp.kmem_cache_alloc_lru >>        0.34 ±  5%      -0.2        0.13 ± 32%      -0.2        0.10 ± >> 14%  perf-profile.children.cycles-pp.wake_up_q >>        0.36 ±  2%      -0.2        0.16 ± 21%      -0.2        0.16 ± >> 26%  perf-profile.children.cycles-pp.alloc_inode >>        0.25 ±  8%      -0.2        0.05 ± 53%      -0.2        0.05 ± >> 75%  perf-profile.children.cycles-pp.get_mountpoint >>        0.32 ±  6%      -0.2        0.13 ± 22%      -0.2        0.11 ± >> 22%  perf-profile.children.cycles-pp.commit_tree >>        0.28 ±  4%      -0.2        0.10 ± 38%      -0.2        0.09 ± >> 25%  perf-profile.children.cycles-pp.__list_del_entry_valid >>        0.80 ±  2%      -0.2        0.63 ±  9%      -0.2        0.59 ± >> 11%  perf-profile.children.cycles-pp.getname_flags >>        0.22 ±  6%      -0.2        0.04 ± 78%      -0.1        0.07 ± >> 20%  perf-profile.children.cycles-pp.mutex_lock_killable >>        0.28 ±  2%      -0.2        0.11 ± 27%      -0.2        0.11 ± >> 11%  perf-profile.children.cycles-pp.ida_free >>        0.27 ±  4%      -0.2        0.10 ± 25%      -0.2        0.09 ± >> 29%  perf-profile.children.cycles-pp.synchronize_rcu_expedited_wait >>        0.27 ±  9%      -0.2        0.10 ±  9%      -0.2        0.08 ± >> 17%  perf-profile.children.cycles-pp.__slab_free >>        0.19 ±  7%      -0.2        0.03 ±101%      -0.2        0.03 >> ±102%  perf-profile.children.cycles-pp.rcu_exp_handler >>        0.32 ±  4%      -0.2        0.16 ± 27%      -0.1        0.18 ± >> 24%  perf-profile.children.cycles-pp.slab_pre_alloc_hook >>        0.25 ±  3%      -0.2        0.10 ± 24%      -0.2        0.08 ± >> 27%  perf-profile.children.cycles-pp.synchronize_rcu_expedited_wait_once >>        0.29 ±  4%      -0.2        0.13 ± 26%      -0.2        0.11 ± >> 26%  perf-profile.children.cycles-pp.shmem_get_inode >>        0.34 ±  5%      -0.2        0.18 ± 16%      -0.1        0.19 ± >> 16%  perf-profile.children.cycles-pp.dentry_kill >>        0.23 ±  8%      -0.2        0.08 ± 25%      -0.2        0.04 ± >> 77%  perf-profile.children.cycles-pp.generic_exec_single >>        0.48 ±  5%      -0.2        0.33 ± 16%      -0.1        0.34 ± >> 30%  perf-profile.children.cycles-pp.flush_smp_call_function_queue >>        0.25 ±  8%      -0.1        0.10 ± 25%      -0.2        0.09 ± >> 26%  perf-profile.children.cycles-pp.llist_add_batch >>        0.50 ±  2%      -0.1        0.35 ± 14%      -0.2        0.28 ± >> 14%  perf-profile.children.cycles-pp.__wake_up_common_lock >>        0.31 ±  4%      -0.1        0.16 ± 19%      -0.1        0.18 ± >> 19%  perf-profile.children.cycles-pp.__dentry_kill >>        0.62 ±  3%      -0.1        0.48 ± 12%      -0.2        0.45 ± >> 11%  perf-profile.children.cycles-pp.strncpy_from_user >>        0.24 ±  7%      -0.1        0.10 ± 31%      -0.1        0.09 ± >> 30%  perf-profile.children.cycles-pp._find_next_zero_bit >>        0.31 ±  6%      -0.1        0.17 ± 11%      -0.1        0.16 ± >> 19%  perf-profile.children.cycles-pp.stress_mwc32 >>        0.61 ± 13%      -0.1        0.48 ± 25%      -0.2        0.38 ± >> 13%  perf-profile.children.cycles-pp.exp_funnel_lock >>        0.23 ±  6%      -0.1        0.09 ± 20%      -0.1        0.10 ± >> 21%  perf-profile.children.cycles-pp.ramfs_get_inode >>        0.29 ±  8%      -0.1        0.16 ± 18%      -0.1        0.16 ± >> 22%  perf-profile.children.cycles-pp.mutex_spin_on_owner >>        0.30 ±  5%      -0.1        0.17 ± 24%      -0.1        0.16 ± >> 19%  perf-profile.children.cycles-pp.evict >>        0.48 ± 19%      -0.1        0.35 ± 14%      -0.2        0.33 ± >> 8%  perf-profile.children.cycles-pp.idle_cpu >>        0.27 ±  6%      -0.1        0.14 ± 13%      -0.1        0.15 ± >> 19%  perf-profile.children.cycles-pp.__d_alloc >>        0.30 ±  6%      -0.1        0.18 ± 16%      -0.1        0.17 ± >> 18%  perf-profile.children.cycles-pp.queue_work_on >>        0.15 ±  6%      -0.1        0.03 ±100%      -0.1        0.05 ± >> 82%  perf-profile.children.cycles-pp.llist_reverse_order >>        0.19 ±  7%      -0.1        0.07 ± 16%      -0.1        0.06 ± >> 27%  perf-profile.children.cycles-pp.lockref_get >>        0.39 ±  8%      -0.1        0.28 ± 21%      -0.1        0.25 ± >> 17%  perf-profile.children.cycles-pp.sched_ttwu_pending >>        0.26 ±  7%      -0.1        0.15 ± 10%      -0.1        0.15 ± >> 30%  perf-profile.children.cycles-pp.__call_rcu_common >>        0.20 ±  7%      -0.1        0.08 ± 16%      -0.1        0.09 ± >> 20%  perf-profile.children.cycles-pp.get_nr_inodes >>        0.22 ± 11%      -0.1        0.10 ± 15%      -0.1        0.11 ± >> 31%  perf-profile.children.cycles-pp.d_alloc >>        0.27 ±  8%      -0.1        0.16 ± 14%      -0.1        0.12 ± >> 13%  perf-profile.children.cycles-pp.__legitimize_path >>        0.22 ±  6%      -0.1        0.11 ±  9%      -0.1        0.11 ± >> 32%  perf-profile.children.cycles-pp.__flush_work >>        0.34 ± 11%      -0.1        0.23 ± 23%      -0.1        0.22 ± >> 16%  perf-profile.children.cycles-pp.dequeue_task_fair >>        0.21 ±  3%      -0.1        0.10 ± 36%      -0.1        0.11 ± >> 27%  perf-profile.children.cycles-pp.memcg_list_lru_alloc >>        0.48 ±  8%      -0.1        0.37 ± 18%      -0.2        0.30 ± >> 4%  perf-profile.children.cycles-pp.ttwu_do_activate >>        0.16 ±  5%      -0.1        0.06 ± 47%      -0.1        0.06 ± >> 24%  perf-profile.children.cycles-pp.__list_add_valid >>        0.15 ±  6%      -0.1        0.04 ± 75%      -0.1        0.05 ± >> 47%  perf-profile.children.cycles-pp.pcpu_next_fit_region >>        0.14 ± 15%      -0.1        0.04 ± 73%      -0.1        0.02 >> ±142%  perf-profile.children.cycles-pp.swake_up_one >>        0.16 ±  4%      -0.1        0.06 ± 49%      -0.1        0.06 ± >> 47%  perf-profile.children.cycles-pp.pcpu_find_block_fit >>        0.17 ±  4%      -0.1        0.07 ± 57%      -0.1        0.08 ± >> 18%  perf-profile.children.cycles-pp.shmem_mknod >>        0.21 ± 11%      -0.1        0.11 ± 19%      -0.1        0.10 ± >> 19%  perf-profile.children.cycles-pp.ttwu_queue_wakelist >>        0.28 ±  8%      -0.1        0.18 ±  9%      -0.1        0.14 ± >> 17%  perf-profile.children.cycles-pp.try_to_unlazy >>        0.23 ± 14%      -0.1        0.13 ± 29%      -0.1        0.13 ± >> 19%  perf-profile.children.cycles-pp.schedule_hrtimeout_range_clock >>        0.14 ±  7%      -0.1        0.04 ± 75%      -0.1        0.04 ± >> 82%  perf-profile.children.cycles-pp.pcpu_block_update_hint_alloc >>        0.19 ±  8%      -0.1        0.10 ± 27%      -0.1        0.08 ± >> 25%  perf-profile.children.cycles-pp.complete_walk >>        0.12 ± 12%      -0.1        0.03 ±100%      -0.1        0.05 ± >> 46%  perf-profile.children.cycles-pp.__lookup_mnt >>        0.28 ±  3%      -0.1        0.19 ± 23%      -0.1        0.14 ± >> 15%  perf-profile.children.cycles-pp.vfs_symlink >>        0.14 ± 14%      -0.1        0.05 ± 74%      -0.1        0.01 >> ±223%  perf-profile.children.cycles-pp.irqentry_exit_to_user_mode >>        0.48 ±  4%      -0.1        0.38 ± 12%      -0.1        0.35 ± >> 17%  perf-profile.children.cycles-pp.kmem_cache_alloc >>        0.12 ±  6%      -0.1        0.03 ±108%      -0.1 >> 0.00        perf-profile.children.cycles-pp.__bitmap_clear >>        0.28 ± 11%      -0.1        0.19 ± 27%      -0.1        0.20 ± >> 15%  perf-profile.children.cycles-pp.dequeue_entity >>        0.42 ± 10%      -0.1        0.34 ± 17%      -0.1        0.27 ± >> 3%  perf-profile.children.cycles-pp.enqueue_task_fair >>        0.21 ± 18%      -0.1        0.12 ± 25%      -0.1        0.11 ± >> 28%  perf-profile.children.cycles-pp.__d_lookup >>        0.18 ±  5%      -0.1        0.10 ± 19%      -0.1        0.10 ± >> 18%  perf-profile.children.cycles-pp.__d_lookup_rcu >>        0.19 ±  8%      -0.1        0.11 ±  9%      -0.1        0.13 ± >> 24%  perf-profile.children.cycles-pp.update_curr >>        0.11 ± 12%      -0.1        0.03 ±105%      -0.1 >> 0.00        perf-profile.children.cycles-pp.shmem_put_super >>        0.14 ± 11%      -0.1        0.06 ± 55%      -0.1        0.04 ± >> 86%  perf-profile.children.cycles-pp.shmem_alloc_inode >>        0.20 ±  7%      -0.1        0.11 ± 34%      -0.1        0.09 ± >> 26%  perf-profile.children.cycles-pp.list_lru_destroy >>        0.17 ± 10%      -0.1        0.09 ± 24%      -0.1        0.09 ± >> 32%  perf-profile.children.cycles-pp.truncate_inode_pages_range >>        0.11 ± 10%      -0.1        0.03 ±103%      -0.1        0.02 >> ±141%  perf-profile.children.cycles-pp.apparmor_capable >>        0.11 ±  9%      -0.1        0.03 ±103%      -0.1        0.02 >> ±146%  perf-profile.children.cycles-pp.kstrdup >>        0.12 ±  9%      -0.1        0.04 ± 72%      -0.1        0.03 >> ±101%  perf-profile.children.cycles-pp.ns_capable >>        0.14 ± 10%      -0.1        0.06 ± 52%      -0.1        0.07 ± >> 8%  perf-profile.children.cycles-pp.d_make_root >>        0.16 ±  7%      -0.1        0.08 ± 51%      -0.1        0.09 ± >> 20%  perf-profile.children.cycles-pp.ramfs_fill_super >>        0.10 ± 13%      -0.1        0.03 ±102%      -0.1 >> 0.00        perf-profile.children.cycles-pp.percpu_counter_destroy >>        0.12 ±  9%      -0.1        0.04 ± 77%      -0.1        0.03 >> ±100%  perf-profile.children.cycles-pp.security_capable >>        0.13 ±  7%      -0.1        0.05 ± 71%      -0.1        0.06 ± >> 45%  perf-profile.children.cycles-pp.vfs_mkdir >>        0.19 ±  3%      -0.1        0.12 ± 22%      -0.1        0.11 ± >> 22%  perf-profile.children.cycles-pp.lockref_put_return >>        0.16 ±  7%      -0.1        0.09 ± 17%      -0.1        0.08 ± >> 49%  perf-profile.children.cycles-pp.mutex_unlock >>        0.10 ± 10%      -0.1        0.03 ±101%      -0.1        0.03 ± >> 70%  perf-profile.children.cycles-pp.propagate_umount >>        0.14 ± 10%      -0.1        0.06 ± 45%      -0.1        0.06 ± >> 49%  perf-profile.children.cycles-pp.nd_jump_root >>        0.31 ±  7%      -0.1        0.24 ± 16%      -0.1        0.20 ± >> 19%  perf-profile.children.cycles-pp.__kmem_cache_alloc_node >>        0.13 ±  5%      -0.1        0.06 ± 57%      -0.1        0.04 ± >> 76%  perf-profile.children.cycles-pp.perf_trace_sched_wakeup_template >>        0.08 ± 11%      -0.1        0.01 ±223%      -0.1        0.02 ± >> 99%  perf-profile.children.cycles-pp.strndup_user >>        0.14 ± 11%      -0.1        0.08 ± 33%      -0.1        0.06 ± >> 50%  perf-profile.children.cycles-pp.__lookup_hash >>        0.15 ±  8%      -0.1        0.08 ± 14%      -0.1        0.08 ± >> 8%  perf-profile.children.cycles-pp.task_work_add >>        0.11 ±  6%      -0.1        0.04 ±101%      -0.1        0.04 ± >> 75%  perf-profile.children.cycles-pp.autoremove_wake_function >>        0.12 ± 12%      -0.1        0.06 ± 78%      -0.1        0.06 ± >> 24%  perf-profile.children.cycles-pp.vfs_rmdir >>        0.16 ±  4%      -0.1        0.10 ± 21%      -0.1        0.06 ± >> 50%  perf-profile.children.cycles-pp.shrink_dcache_for_umount >>        0.14 ± 12%      -0.1        0.08 ± 24%      -0.1        0.09 ± >> 33%  perf-profile.children.cycles-pp.d_alloc_parallel >>        0.10 ± 17%      -0.1        0.04 ± 72%      -0.1        0.04 ± >> 73%  perf-profile.children.cycles-pp.vfs_unlink >>        0.20 ±  5%      -0.1        0.13 ± 22%      -0.1        0.10 ± >> 20%  perf-profile.children.cycles-pp.ramfs_symlink >>        0.08 ±  4%      -0.1        0.02 ±142%      -0.1        0.03 >> ±102%  perf-profile.children.cycles-pp.lockref_put_or_lock >>        0.26 ± 12%      -0.1        0.20 ± 28%      -0.1        0.18 ± >> 7%  perf-profile.children.cycles-pp.ep_poll >>        0.10 ± 11%      -0.1        0.04 ±101%      -0.1        0.04 ± >> 77%  perf-profile.children.cycles-pp.__pagevec_release >>        0.13 ± 10%      -0.1        0.07 ± 46%      -0.1        0.05 ± >> 73%  perf-profile.children.cycles-pp.xas_store >>        0.11 ±  8%      -0.1        0.05 ± 47%      -0.1        0.05 ± >> 74%  perf-profile.children.cycles-pp.set_root >>        0.16 ±  8%      -0.1        0.11 ± 27%      -0.1        0.09 ± >> 15%  perf-profile.children.cycles-pp.__kmalloc_node_track_caller >>        0.27 ± 12%      -0.1        0.21 ± 29%      -0.1        0.20 ± >> 8%  perf-profile.children.cycles-pp.do_epoll_wait >>        0.12 ±  8%      -0.1        0.06 ± 21%      -0.1        0.06 ± >> 19%  perf-profile.children.cycles-pp.__smp_call_single_queue >>        0.27 ± 12%      -0.0        0.22 ± 29%      -0.1        0.21 ± >> 6%  perf-profile.children.cycles-pp.__x64_sys_epoll_wait >>        0.16 ±  6%      -0.0        0.11 ± 25%      -0.1        0.09 ± >> 23%  perf-profile.children.cycles-pp.page_symlink >>        0.13 ±  5%      -0.0        0.08 ± 26%      -0.1        0.07 ± >> 50%  perf-profile.children.cycles-pp.prepare_task_switch >>        0.12 ± 10%      -0.0        0.07 ± 30%      -0.1        0.04 ± >> 79%  perf-profile.children.cycles-pp._raw_read_lock_irqsave >>        0.07 ±  6%      -0.0        0.03 ±102%      -0.0 >> 0.05        perf-profile.children.cycles-pp.fsnotify_destroy_marks >>        0.12 ±  9%      -0.0        0.07 ± 53%      -0.1        0.06 ± >> 45%  perf-profile.children.cycles-pp.vfs_parse_fs_string >>        0.31 ±  3%      -0.0        0.26 ± 16%      -0.1        0.19 ± >> 22%  perf-profile.children.cycles-pp.__wake_up_common >>        0.09 ±  7%      -0.0        0.05 ± 78%      -0.1        0.04 ± >> 75%  perf-profile.children.cycles-pp.mutex_lock >>        0.12 ± 11%      -0.0        0.08 ±118%      -0.1        0.04 ± >> 76%  perf-profile.children.cycles-pp.__list_lru_init >>        0.12 ±  9%      -0.0        0.08 ± 23%      -0.1        0.07 ± >> 30%  perf-profile.children.cycles-pp.kmalloc_trace >>        0.12 ±  5%      -0.0        0.08 ± 17%      -0.1        0.03 >> ±101%  perf-profile.children.cycles-pp.rcu_all_qs >>        0.10 ±  9%      -0.0        0.07 ± 50%      -0.0        0.06 ± >> 14%  perf-profile.children.cycles-pp.terminate_walk >>        0.22 ±  6%      -0.0        0.18 ± 20%      -0.1        0.16 ± >> 20%  perf-profile.children.cycles-pp.__queue_work >>        0.11 ±  8%      -0.0        0.08 ± 24%      -0.1        0.06 ± >> 48%  perf-profile.children.cycles-pp.simple_write_begin >>        0.10 ± 14%      -0.0        0.07 ± 32%      -0.1        0.03 >> ±102%  perf-profile.children.cycles-pp.memcg_destroy_list_lru >>        0.10 ±  8%      -0.0        0.06 ± 50%      -0.1        0.04 ± >> 72%  perf-profile.children.cycles-pp.pagecache_get_page >>        0.12 ± 10%      -0.0        0.08 ± 20%      -0.1        0.06 ± >> 65%  perf-profile.children.cycles-pp.__switch_to >>        0.30 ±  4%      -0.0        0.27 ± 18%      -0.1        0.24 ± >> 9%  perf-profile.children.cycles-pp.update_load_avg >>        0.11 ±  8%      -0.0        0.08 ± 11%      -0.0        0.06 ± >> 52%  perf-profile.children.cycles-pp.set_next_entity >>        0.30 ±  6%      -0.0        0.28 ± 13%      -0.1        0.22 ± >> 8%  perf-profile.children.cycles-pp.enqueue_entity >>        0.11 ±  6%      -0.0        0.10 ± 18%      -0.0        0.07 ± >> 21%  perf-profile.children.cycles-pp.__filemap_get_folio >>        0.10 ±  8%      -0.0        0.10 ± 21%      -0.0        0.07 ± >> 16%  perf-profile.children.cycles-pp.get_obj_cgroup_from_current >>        0.00            +0.0        0.01 ±223%      +0.1        0.06 ± >> 29%  perf-profile.children.cycles-pp.irq_work_tick >>        0.47 ±  5%      +0.0        0.49 ± 17%      -0.1        0.40 ± >> 15%  perf-profile.children.cycles-pp.walk_component >>        0.07 ± 51%      +0.0        0.11 ± 25%      +0.1        0.14 ± >> 22%  perf-profile.children.cycles-pp.__libc_recvmsg >>        0.05 ± 46%      +0.0        0.09 ± 26%      +0.1        0.12 ± >> 45%  perf-profile.children.cycles-pp.__pthread_mutex_lock >>        0.10 ± 13%      +0.0        0.14 ± 16%      +0.0        0.11 ± >> 18%  perf-profile.children.cycles-pp.do_open >>        0.04 ± 45%      +0.0        0.09 ± 22%      +0.0        0.08 ± >> 21%  perf-profile.children.cycles-pp.cp_new_stat >>        0.00            +0.0        0.04 ± 75%      +0.1        0.07 ± >> 29%  perf-profile.children.cycles-pp.vm_mmap_pgoff >>        0.08 ± 16%      +0.0        0.12 ± 17%      +0.0        0.11 ± >> 21%  perf-profile.children.cycles-pp.do_dentry_open >>        0.09 ± 19%      +0.0        0.13 ± 12%      +0.0        0.13 ± >> 16%  perf-profile.children.cycles-pp.copy_user_enhanced_fast_string >>        0.05 ± 46%      +0.0        0.09 ± 55%      +0.1        0.14 ± >> 30%  perf-profile.children.cycles-pp.__pthread_mutex_unlock_usercnt >>        0.00            +0.0        0.05 ± 79%      +0.1        0.11 ± >> 24%  perf-profile.children.cycles-pp.nr_iowait_cpu >>        0.02 ±142%      +0.0        0.07 ± 45%      +0.1        0.08 ± >> 37%  perf-profile.children.cycles-pp.consume_skb >>        0.00            +0.0        0.05 ±107%      +0.1        0.06 ± >> 28%  perf-profile.children.cycles-pp.native_apic_mem_write >>        0.00            +0.1        0.05 ± 49%      +0.1        0.06 ± >> 23%  perf-profile.children.cycles-pp.arch_cpu_idle_exit >>        0.00            +0.1        0.05 ± 83%      +0.1        0.08 ± >> 27%  perf-profile.children.cycles-pp.rb_next >>        0.01 ±223%      +0.1        0.06 ± 14%      +0.0        0.06 ± >> 51%  perf-profile.children.cycles-pp.syscall_return_via_sysret >>        0.00            +0.1        0.05 ± 46%      +0.1        0.06 ± >> 11%  perf-profile.children.cycles-pp.load_elf_binary >>        0.00            +0.1        0.05 ± 46%      +0.1        0.06 ± >> 11%  perf-profile.children.cycles-pp.search_binary_handler >>        0.00            +0.1        0.05 ± 50%      +0.1        0.09 ± >> 14%  perf-profile.children.cycles-pp.error_entry >>        0.00            +0.1        0.06 ± 45%      +0.1        0.07 ± >> 11%  perf-profile.children.cycles-pp.exec_binprm >>        0.00            +0.1        0.06 ± 11%      +0.1        0.06 ± >> 20%  perf-profile.children.cycles-pp.fput >>        0.11 ± 14%      +0.1        0.17 ± 26%      +0.1        0.21 ± >> 32%  perf-profile.children.cycles-pp.sock_recvmsg >>        0.00            +0.1        0.06 ± 94%      +0.1        0.11 ± >> 32% >> perf-profile.children.cycles-pp.tick_check_oneshot_broadcast_this_cpu >>        0.10 ± 21%      +0.1        0.16 ± 27%      +0.1        0.20 ± >> 32%  perf-profile.children.cycles-pp.unix_stream_read_generic >>        0.00            +0.1        0.06 ± 76%      +0.1        0.08 ± >> 25%  perf-profile.children.cycles-pp._dbus_first_type_in_signature >>        0.11 ±  8%      +0.1        0.17 ± 22%      +0.0        0.13 ± >> 20%  perf-profile.children.cycles-pp.down_read >>        0.10 ± 21%      +0.1        0.16 ± 27%      +0.1        0.20 ± >> 31%  perf-profile.children.cycles-pp.unix_stream_recvmsg >>        0.00            +0.1        0.06 ± 24%      +0.0        0.03 >> ±100%  perf-profile.children.cycles-pp._copy_to_user >>        0.00            +0.1        0.06 ± 32%      +0.1        0.06 ± >> 45%  perf-profile.children.cycles-pp.filp_close >>        0.00            +0.1        0.06 ± 45%      +0.1        0.08 ± >> 20%  perf-profile.children.cycles-pp.bprm_execve >>        0.04 ± 44%      +0.1        0.11 ± 16%      +0.1        0.13 ± >> 33%  perf-profile.children.cycles-pp.unit_file_find_dropin_paths >>        0.00            +0.1        0.07 ± 15%      +0.0        0.04 ± >> 71%  perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack >>        0.00            +0.1        0.07 ± 13%      +0.1        0.05 ± >> 49%  perf-profile.children.cycles-pp.hrtimer_get_next_event >>        0.00            +0.1        0.07 ± 32%      +0.1        0.06 ± >> 60%  perf-profile.children.cycles-pp.kernfs_iop_get_link >>        0.11 ± 17%      +0.1        0.18 ± 27%      +0.1        0.22 ± >> 32%  perf-profile.children.cycles-pp.____sys_recvmsg >>        0.00            +0.1        0.07 ± 47%      +0.1        0.08 ± >> 29%  perf-profile.children.cycles-pp.get_cpu_device >>        0.00            +0.1        0.07 ± 56%      +0.1        0.11 ± >> 13%  perf-profile.children.cycles-pp.tsc_verify_tsc_adjust >>        0.00            +0.1        0.08 ± 20%      +0.1 >> 0.08        perf-profile.children.cycles-pp.wait_for_xmitr >>        0.07 ± 23%      +0.1        0.14 ± 29%      +0.1        0.16 ± >> 44%  perf-profile.children.cycles-pp.recvmsg >>        0.00            +0.1        0.08 ± 23%      +0.1        0.08 ± >> 26%  perf-profile.children.cycles-pp.call_cpuidle >>        0.16 ±  6%      +0.1        0.24 ± 15%      +0.0        0.18 ± >> 12%  perf-profile.children.cycles-pp.ct_kernel_exit_state >>        0.00            +0.1        0.08 ± 22%      +0.1        0.10 ± >> 83%  perf-profile.children.cycles-pp.strlen@plt >>        0.00            +0.1        0.08 ± 35%      +0.1        0.11 ± >> 23%  perf-profile.children.cycles-pp.rb_insert_color >>        0.12 ± 16%      +0.1        0.20 ± 26%      +0.1        0.24 ± >> 30%  perf-profile.children.cycles-pp.___sys_recvmsg >>        0.00            +0.1        0.08 ± 24%      +0.1        0.06 ± >> 54%  perf-profile.children.cycles-pp.__close_nocancel >>        0.00            +0.1        0.08 ± 17%      +0.1        0.06 ± >> 54%  perf-profile.children.cycles-pp._hashmap_iterate >>        0.00            +0.1        0.08 ± 55%      +0.1        0.12 ± >> 14%  perf-profile.children.cycles-pp.arch_cpu_idle_enter >>        0.00            +0.1        0.08 ± 21%      +0.1        0.06 ± >> 47%  perf-profile.children.cycles-pp.seq_puts >>        0.11 ± 10%      +0.1        0.19 ± 21%      +0.1        0.18 ± >> 12%  perf-profile.children.cycles-pp.__entry_text_start >>        0.00            +0.1        0.08 ± 55%      +0.1        0.10 ± >> 23%  perf-profile.children.cycles-pp.irqentry_enter >>        0.00            +0.1        0.08 ± 83%      +0.1        0.11 ± >> 17%  perf-profile.children.cycles-pp.rb_erase >>        0.00            +0.1        0.08 ± 35%      +0.1        0.12 ± >> 33%  perf-profile.children.cycles-pp.perf_event_task_tick >>        0.00            +0.1        0.08 ± 26%      +0.1        0.07 ± >> 49%  perf-profile.children.cycles-pp.device_add_property_aux >>        0.02 ±141%      +0.1        0.10 ± 18%      +0.1        0.09 ± >> 35%  perf-profile.children.cycles-pp.sock_def_readable >>        0.00            +0.1        0.08 ± 30%      +0.1        0.09 ± >> 22%  perf-profile.children.cycles-pp.siphash24_init >>        0.00            +0.1        0.08 ± 32%      +0.1        0.09 ± >> 14%  perf-profile.children.cycles-pp.tick_nohz_tick_stopped >>        0.07 ± 48%      +0.1        0.16 ± 33%      +0.1        0.17 ± >> 30%  perf-profile.children.cycles-pp.__libc_sendmsg >>        0.12 ± 15%      +0.1        0.21 ± 26%      +0.1        0.25 ± >> 29%  perf-profile.children.cycles-pp.__sys_recvmsg >>        0.01 ±223%      +0.1        0.10 ± 27%      +0.1        0.10 ± >> 53%  perf-profile.children.cycles-pp.string_table_lookup >>        0.00            +0.1        0.09 ± 25%      +0.1        0.10 ± >> 17%  perf-profile.children.cycles-pp.tick_nohz_irq_exit >>        0.20 ±  9%      +0.1        0.29 ±  7%      +0.0        0.23 ± >> 13%  perf-profile.children.cycles-pp.inode_permission >>        0.00            +0.1        0.09 ± 27%      +0.1        0.07 ± >> 49%  perf-profile.children.cycles-pp.__handle_mm_fault >>        0.06 ± 14%      +0.1        0.16 ± 14%      +0.0        0.10 ± >> 19%  perf-profile.children.cycles-pp.vfs_fstat >>        0.00            +0.1        0.09 ± 64%      +0.1        0.12 ± >> 34%  perf-profile.children.cycles-pp.run_posix_cpu_timers >>        0.00            +0.1        0.09 ± 33%      +0.1        0.09 ± >> 35%  perf-profile.children.cycles-pp.vfs_readlink >>        0.00            +0.1        0.10 ± 19%      +0.1        0.06 ± >> 53%  perf-profile.children.cycles-pp.uevent_show >>        0.00            +0.1        0.10 ± 25%      +0.1        0.08 ± >> 47%  perf-profile.children.cycles-pp.handle_mm_fault >>        0.00            +0.1        0.10 ± 10%      +0.1        0.08 ± >> 19%  perf-profile.children.cycles-pp.up_read >>        0.00            +0.1        0.10 ± 30%      +0.1        0.11 ± >> 29%  perf-profile.children.cycles-pp.read_counters >>        0.00            +0.1        0.10 ± 20%      +0.1        0.08 ± >> 22%  perf-profile.children.cycles-pp.strchr >>        0.00            +0.1        0.10 ± 33%      +0.1        0.11 ± >> 29%  perf-profile.children.cycles-pp.path_hash_func >>        0.00            +0.1        0.10 ± 29%      +0.1        0.11 ± >> 29%  perf-profile.children.cycles-pp.cmd_stat >>        0.00            +0.1        0.10 ± 29%      +0.1        0.11 ± >> 29%  perf-profile.children.cycles-pp.dispatch_events >>        0.00            +0.1        0.10 ± 29%      +0.1        0.11 ± >> 29%  perf-profile.children.cycles-pp.process_interval >>        0.07 ± 11%      +0.1        0.18 ± 19%      +0.1        0.17 ± >> 48%  perf-profile.children.cycles-pp.memcpy_erms >>        0.00            +0.1        0.11 ± 19%      +0.1        0.08 ± >> 45%  perf-profile.children.cycles-pp.exc_page_fault >>        0.00            +0.1        0.11 ± 19%      +0.1        0.08 ± >> 45%  perf-profile.children.cycles-pp.do_user_addr_fault >>        0.00            +0.1        0.11 ±  6%      +0.1        0.08 ± >> 22%  perf-profile.children.cycles-pp.show_type >>        0.01 ±223%      +0.1        0.12 ± 16%      +0.1        0.12 ± >> 14%  perf-profile.children.cycles-pp.do_execveat_common >>        0.01 ±223%      +0.1        0.12 ± 16%      +0.1        0.12 ± >> 15%  perf-profile.children.cycles-pp.execve >>        0.01 ±223%      +0.1        0.12 ± 16%      +0.1        0.12 ± >> 15%  perf-profile.children.cycles-pp.__x64_sys_execve >>        0.03 ±100%      +0.1        0.14 ± 18%      +0.1        0.09 ± >> 46%  perf-profile.children.cycles-pp.do_faccessat >>        0.00            +0.1        0.11 ± 14%      +0.1        0.08 ± >> 37%  perf-profile.children.cycles-pp.dev_attr_show >>        0.00            +0.1        0.11 ± 28%      +0.1        0.12 ± >> 31%  perf-profile.children.cycles-pp.__libc_start_main >>        0.00            +0.1        0.11 ± 28%      +0.1        0.12 ± >> 31%  perf-profile.children.cycles-pp.main >>        0.00            +0.1        0.11 ± 28%      +0.1        0.12 ± >> 31%  perf-profile.children.cycles-pp.run_builtin >>        0.10 ± 28%      +0.1        0.22 ± 15%      +0.1        0.16 ± >> 59%  perf-profile.children.cycles-pp.sendmsg >>        0.00            +0.1        0.11 ± 13%      +0.1        0.08 ± >> 35%  perf-profile.children.cycles-pp.sysfs_kf_seq_show >>        0.11 ± 11%      +0.1        0.22 ±  8%      +0.1        0.22 ± >> 16%  perf-profile.children.cycles-pp._raw_spin_trylock >>        0.10 ±  3%      +0.1        0.22 ± 13%      +0.1        0.17 ± >> 20%  perf-profile.children.cycles-pp.__do_sys_newfstat >>        0.11 ± 15%      +0.1        0.23 ± 17%      +0.1        0.21 ± >> 23%  perf-profile.children.cycles-pp.__update_blocked_fair >>        0.00            +0.1        0.12 ± 27%      +0.1        0.09 ± >> 26%  perf-profile.children.cycles-pp.strlen >>        0.00            +0.1        0.12 ± 22%      +0.1        0.13 ± >> 24%  perf-profile.children.cycles-pp.rdrand >>        0.06 ± 18%      +0.1        0.18 ±  8%      +0.1        0.18 ± >> 23%  perf-profile.children.cycles-pp.cfree >>        0.00            +0.1        0.12 ± 13%      +0.1        0.07 ± >> 31%  perf-profile.children.cycles-pp.number >>        0.00            +0.1        0.12 ± 17%      +0.2        0.17 ± >> 23%  perf-profile.children.cycles-pp.hrtimer_update_next_event >>        0.03 ±100%      +0.1        0.15 ± 14%      +0.1        0.10 ± >> 26%  perf-profile.children.cycles-pp.access >>        0.01 ±223%      +0.1        0.13 ± 14%      +0.1        0.09 ± >> 51%  perf-profile.children.cycles-pp.kernfs_iop_permission >>        0.00            +0.1        0.13 ± 22%      +0.1        0.08 ± >> 30%  perf-profile.children.cycles-pp.string_escape_mem >>        0.00            +0.1        0.13 ± 30%      +0.1        0.11 ± >> 20%  perf-profile.children.cycles-pp.try_check_zero >>        0.00            +0.1        0.13 ± 12%      +0.1        0.11 ± >> 38%  perf-profile.children.cycles-pp.asm_exc_page_fault >>        0.00            +0.1        0.13 ± 30%      +0.1        0.12 ± >> 17%  perf-profile.children.cycles-pp.srcu_advance_state >>        0.01 ±223%      +0.1        0.14 ± 25%      +0.1        0.15 ± >> 24%  perf-profile.children.cycles-pp.local_clock >>        0.06 ± 15%      +0.1        0.19 ± 15%      +0.1        0.12 ± >> 44%  perf-profile.children.cycles-pp.prepend_path >>        0.00            +0.1        0.14 ± 26%      +0.1        0.14 ± >> 23%  perf-profile.children.cycles-pp.irqtime_account_process_tick >>        0.03 ± 70%      +0.1        0.17 ± 25%      +0.1        0.16 ± >> 27%  perf-profile.children.cycles-pp.note_gp_changes >>        0.03 ±102%      +0.1        0.17 ± 31%      +0.1        0.17 ± >> 33%  perf-profile.children.cycles-pp.strchr@plt >>        0.00            +0.1        0.14 ± 20%      +0.1        0.09 ± >> 33%  perf-profile.children.cycles-pp.seq_escape_mem >>        0.05 ±  7%      +0.1        0.19 ± 20%      +0.1        0.18 ± >> 14%  perf-profile.children.cycles-pp.siphash24_finalize >>        0.15 ± 10%      +0.1        0.30 ± 31%      +0.1        0.22 ± >> 21%  perf-profile.children.cycles-pp.trigger_load_balance >>        0.00            +0.1        0.15 ± 17%      +0.1        0.10 ± >> 23%  perf-profile.children.cycles-pp.format_decode >>        0.19 ±  5%      +0.2        0.34 ± 21%      +0.1        0.26 ± >> 7%  perf-profile.children.cycles-pp.ct_kernel_enter >>        0.13 ± 14%      +0.2        0.28 ± 19%      +0.1        0.26 ± >> 39%  perf-profile.children.cycles-pp.unix_stream_sendmsg >>        0.14 ± 18%      +0.2        0.29 ± 17%      +0.1        0.27 ± >> 37%  perf-profile.children.cycles-pp.sock_sendmsg >>        0.00            +0.2        0.15 ± 28%      +0.1        0.14 ± >> 19%  perf-profile.children.cycles-pp.process_srcu >>        0.14 ± 17%      +0.2        0.30 ± 19%      +0.1        0.27 ± >> 38%  perf-profile.children.cycles-pp.____sys_sendmsg >>        0.00            +0.2        0.16 ± 52%      +0.2        0.21 ± >> 19%  perf-profile.children.cycles-pp.timerqueue_del >>        0.00            +0.2        0.16 ± 21%      +0.2        0.18 ± >> 9%  perf-profile.children.cycles-pp.get_next_timer_interrupt >>        0.03 ±100%      +0.2        0.19 ± 17%      +0.1        0.12 ± >> 45%  perf-profile.children.cycles-pp.__d_path >>        0.01 ±223%      +0.2        0.18 ± 15%      +0.1        0.16 ± >> 12%  perf-profile.children.cycles-pp.delay_tsc >>        0.02 ±141%      +0.2        0.18 ± 28%      +0.2        0.19 ± >> 17%  perf-profile.children.cycles-pp.cpuidle_governor_latency_req >>        0.00            +0.2        0.17 ± 28%      +0.2        0.23 ± >> 13%  perf-profile.children.cycles-pp.hrtimer_next_event_without >>        0.00            +0.2        0.17 ± 47%      +0.1        0.10 ± >> 19%  perf-profile.children.cycles-pp.kernfs_dop_revalidate >>        0.16 ± 18%      +0.2        0.32 ± 21%      +0.1        0.29 ± >> 37%  perf-profile.children.cycles-pp.___sys_sendmsg >>        0.16 ± 15%      +0.2        0.33 ± 22%      +0.1        0.30 ± >> 37%  perf-profile.children.cycles-pp.__sys_sendmsg >>        0.03 ±100%      +0.2        0.20 ± 16%      +0.2        0.20 ± >> 14%  perf-profile.children.cycles-pp.check_cpu_stall >>        0.00            +0.2        0.18 ± 26%      +0.2        0.23 ± >> 23%  perf-profile.children.cycles-pp.__hrtimer_next_event_base >>        0.00            +0.2        0.18 ± 20%      +0.2        0.24 ± >> 23%  perf-profile.children.cycles-pp.timerqueue_add >>        0.09 ± 13%      +0.2        0.28 ± 37%      +0.1        0.19 ± >> 29%  perf-profile.children.cycles-pp.__x64_sys_readlinkat >>        0.07 ± 17%      +0.2        0.26 ± 13%      +0.1        0.18 ± >> 34%  perf-profile.children.cycles-pp.seq_path_root >>        0.30 ±  3%      +0.2        0.49 ± 12%      +0.2        0.48 ± >> 13%  perf-profile.children.cycles-pp.update_rq_clock >>        0.19 ±  6%      +0.2        0.38 ± 18%      +0.1        0.32 ± >> 8%  perf-profile.children.cycles-pp.ct_idle_exit >>        0.21 ± 11%      +0.2        0.40 ±  4%      +0.2        0.38 ± >> 17%  perf-profile.children.cycles-pp.sched_clock_cpu >>        0.07 ± 17%      +0.2        0.26 ± 22%      +0.2        0.24 ± >> 15%  perf-profile.children.cycles-pp.update_irq_load_avg >>        0.10 ± 14%      +0.2        0.29 ± 34%      +0.1        0.21 ± >> 27%  perf-profile.children.cycles-pp.do_readlinkat >>        0.11 ± 34%      +0.2        0.31 ± 29%      +0.2        0.30 ± >> 27%  perf-profile.children.cycles-pp.tick_sched_do_timer >>        0.10 ± 15%      +0.2        0.30 ± 31%      +0.1        0.20 ± >> 33%  perf-profile.children.cycles-pp.readlinkat >>        0.02 ±141%      +0.2        0.22 ± 19%      +0.1        0.16 ± >> 21%  perf-profile.children.cycles-pp.device_read_db_internal_filename >>        0.01 ±223%      +0.2        0.22 ± 18%      +0.3        0.29 ± >> 22%  perf-profile.children.cycles-pp.enqueue_hrtimer >>        0.22 ± 12%      +0.2        0.43 ±  5%      +0.2        0.46 ± >> 12%  perf-profile.children.cycles-pp.native_sched_clock >>        0.11 ± 14%      +0.2        0.33 ± 19%      +0.2        0.34 ± >> 27%  perf-profile.children.cycles-pp.path_compare >>        0.05 ± 46%      +0.2        0.29 ± 10%      +0.2        0.27 ± >> 15%  perf-profile.children.cycles-pp.__intel_pmu_enable_all >>        0.12 ± 11%      +0.2        0.36 ±  7%      +0.2        0.37 ± >> 19%  perf-profile.children.cycles-pp.irqtime_account_irq >>        0.04 ± 72%      +0.2        0.28 ± 31%      +0.2        0.23 ± >> 12%  perf-profile.children.cycles-pp.ktime_get_update_offsets_now >>        0.16 ±  6%      +0.2        0.40 ±  8%      +0.2        0.40 ± >> 6%  perf-profile.children.cycles-pp.rcu_pending >>        0.22 ±  4%      +0.2        0.46 ±  6%      +0.2        0.42 ± >> 8%  perf-profile.children.cycles-pp.native_irq_return_iret >>        0.18 ± 12%      +0.3        0.44 ± 24%      +0.2        0.37 ± >> 25%  perf-profile.children.cycles-pp.strcmp@plt >>        0.08 ±  8%      +0.3        0.36 ± 15%      +0.2        0.32 ± >> 29%  perf-profile.children.cycles-pp.seq_printf >>        0.39 ± 10%      +0.3        0.67 ± 19%      +0.2        0.54 ± >> 30%  perf-profile.children.cycles-pp.__close >>        0.09 ± 21%      +0.3        0.37 ± 24%      +0.4        0.47 ± >> 10%  perf-profile.children.cycles-pp.tick_nohz_next_event >>        0.18 ±  9%      +0.3        0.47 ± 34%      +0.3        0.50 ± >> 30%  perf-profile.children.cycles-pp.unit_name_is_valid >>        0.18 ± 22%      +0.3        0.47 ± 33%      +0.1        0.30 ± >> 45%  perf-profile.children.cycles-pp.__fxstat64 >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23%  perf-profile.children.cycles-pp.memcpy_toio >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23%  perf-profile.children.cycles-pp.drm_atomic_helper_commit_tail_rpm >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23%  perf-profile.children.cycles-pp.drm_atomic_helper_commit_planes >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23% >> perf-profile.children.cycles-pp.ast_primary_plane_helper_atomic_update >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23%  perf-profile.children.cycles-pp.drm_fb_memcpy >>        0.22 ±  7%      +0.3        0.52 ± 20%      +0.3        0.49 ± >> 20%  perf-profile.children.cycles-pp.update_blocked_averages >>        0.00            +0.3        0.31 ± 23%      +0.3        0.26 ± >> 23%  perf-profile.children.cycles-pp.commit_tail >>        0.00            +0.3        0.31 ± 23%      +0.3        0.26 ± >> 23% >> perf-profile.children.cycles-pp.ast_mode_config_helper_atomic_commit_tail >>        0.07 ±  6%      +0.3        0.39 ± 10%      +0.3        0.35 ± >> 13%  perf-profile.children.cycles-pp.read_tsc >>        0.32 ±  5%      +0.3        0.65 ± 15%      +0.3        0.58 ± >> 17%  perf-profile.children.cycles-pp.schedule_timeout >>        0.00            +0.3        0.33 ± 24%      +0.3        0.27 ± >> 25%  perf-profile.children.cycles-pp.drm_atomic_commit >>        0.00            +0.3        0.33 ± 24%      +0.3        0.27 ± >> 25%  perf-profile.children.cycles-pp.drm_atomic_helper_commit >>        0.05 ± 46%      +0.3        0.38 ± 32%      +0.4        0.48 ± >> 15%  perf-profile.children.cycles-pp.tick_irq_enter >>        0.00            +0.3        0.33 ± 24%      +0.3        0.27 ± >> 25%  perf-profile.children.cycles-pp.drm_atomic_helper_dirtyfb >>        0.06 ± 45%      +0.3        0.40 ± 27%      +0.4        0.50 ± >> 16%  perf-profile.children.cycles-pp.irq_enter_rcu >>        0.10 ±  8%      +0.3        0.45 ± 12%      +0.3        0.36 ± >> 23%  perf-profile.children.cycles-pp.vsnprintf >>        0.08 ± 11%      +0.4        0.44 ± 11%      +0.4        0.46 ± >> 15%  perf-profile.children.cycles-pp.lapic_next_deadline >>        0.09 ± 15%      +0.4        0.45 ± 25%      +0.3        0.44 ± >> 22%  perf-profile.children.cycles-pp.run_rebalance_domains >>        0.02 ±141%      +0.4        0.39 ± 27%      +0.3        0.32 ± >> 23%  perf-profile.children.cycles-pp.drm_fb_helper_damage_work >>        0.02 ±141%      +0.4        0.39 ± 27%      +0.3        0.32 ± >> 23%  perf-profile.children.cycles-pp.drm_fbdev_fb_dirty >>        0.18 ±  7%      +0.4        0.55 ±  6%      +0.3        0.52 ± >> 7%  perf-profile.children.cycles-pp.rcu_sched_clock_irq >>        0.21 ± 13%      +0.4        0.58 ± 11%      +0.3        0.47 ± >> 19%  perf-profile.children.cycles-pp.arch_scale_freq_tick >>        0.18 ±  8%      +0.4        0.60 ± 14%      +0.4        0.55 ± >> 17%  perf-profile.children.cycles-pp.__wait_for_common >>        0.10 ± 10%      +0.4        0.52 ± 17%      +0.5        0.56 ± >> 11%  perf-profile.children.cycles-pp.perf_rotate_context >>        0.14 ± 12%      +0.4        0.57 ± 21%      +0.4        0.50 ± >> 24%  perf-profile.children.cycles-pp.malloc >>        0.11 ± 14%      +0.5        0.58 ±  9%      +0.5        0.60 ± >> 13%  perf-profile.children.cycles-pp.clockevents_program_event >>        0.13 ± 20%      +0.5        0.61 ± 19%      +0.6        0.75 ± >> 10%  perf-profile.children.cycles-pp.tick_nohz_get_sleep_length >>        0.37 ±  9%      +0.5        0.89 ± 13%      +0.3        0.68 ± >> 22%  perf-profile.children.cycles-pp.openat64 >>        0.18 ±  8%      +0.5        0.72 ± 14%      +0.5        0.69 ± >> 7%  perf-profile.children.cycles-pp.rebalance_domains >>        0.07 ± 11%      +0.6        0.63 ± 21%      +0.5        0.55 ± >> 15%  perf-profile.children.cycles-pp.unregister_shrinker >>        0.16 ±  7%      +0.6        0.74 ±  8%      +0.5        0.64 ± >> 12%  perf-profile.children.cycles-pp.ktime_get >>        0.00            +0.6        0.59 ± 18%      +0.5        0.51 ± >> 16%  perf-profile.children.cycles-pp.__synchronize_srcu >>        0.09 ± 13%      +0.6        0.70 ± 18%      +0.6        0.73 ± >> 8%  perf-profile.children.cycles-pp.io_serial_in >>        0.25 ±  3%      +0.6        0.88 ± 26%      +0.6        0.84 ± >> 21%  perf-profile.children.cycles-pp.siphash24_compress >>        0.16 ± 10%      +0.7        0.82 ± 12%      +0.7        0.81 ± >> 7%  perf-profile.children.cycles-pp.perf_mux_hrtimer_handler >>        0.11 ±  6%      +0.8        0.88 ± 15%      +0.8        0.88 ± >> 8%  perf-profile.children.cycles-pp.wait_for_lsr >>        0.12 ±  6%      +0.8        0.90 ± 15%      +0.8        0.92 ± >> 8%  perf-profile.children.cycles-pp.serial8250_console_write >>        0.12 ±  7%      +0.8        0.92 ± 14%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.irq_work_run >>        0.12 ±  7%      +0.8        0.92 ± 14%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp._printk >>        0.12 ±  7%      +0.8        0.92 ± 14%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.vprintk_emit >>        0.12 ±  7%      +0.8        0.92 ± 14%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.console_unlock >>        0.12 ±  7%      +0.8        0.92 ± 14%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.console_flush_all >>        0.12 ±  7%      +0.8        0.92 ± 15%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.sysvec_irq_work >>        0.12 ±  7%      +0.8        0.92 ± 15%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.__sysvec_irq_work >>        0.12 ±  7%      +0.8        0.92 ± 15%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.irq_work_single >>        0.12 ±  7%      +0.8        0.92 ± 15%      +0.8        0.94 ± >> 8%  perf-profile.children.cycles-pp.asm_sysvec_irq_work >>        0.44 ±  9%      +0.8        1.24 ± 13%      +0.5        0.92 ± >> 23%  perf-profile.children.cycles-pp.device_set_syspath >>        0.12 ±  5%      +0.8        0.94 ± 15%      +0.8        0.97 ± >> 9%  perf-profile.children.cycles-pp.irq_work_run_list >>        0.29 ± 10%      +0.9        1.16 ± 10%      +0.6        0.88 ± >> 23%  perf-profile.children.cycles-pp.show_mountinfo >>        0.48 ±  9%      +0.9        1.39 ± 11%      +0.5        1.03 ± >> 23%  perf-profile.children.cycles-pp.sd_device_new_from_syspath >>        0.40 ±  9%      +1.0        1.45 ±  9%      +0.7        1.10 ± >> 23%  perf-profile.children.cycles-pp.seq_read_iter >>        0.61 ±  4%      +1.1        1.76 ±  7%      +1.0        1.56 ± >> 10%  perf-profile.children.cycles-pp.scheduler_tick >>        0.42 ±  9%      +1.2        1.58 ± 10%      +0.8        1.23 ± >> 20%  perf-profile.children.cycles-pp.vfs_read >>        0.42 ±  8%      +1.2        1.59 ± 10%      +0.8        1.24 ± >> 20%  perf-profile.children.cycles-pp.ksys_read >>        0.44 ± 10%      +1.2        1.64 ± 10%      +0.8        1.26 ± >> 22%  perf-profile.children.cycles-pp.read >>        0.40 ±  8%      +1.3        1.70 ± 12%      +1.5        1.87 ± >> 5%  perf-profile.children.cycles-pp.menu_select >>        3.06 ±  5%      +1.4        4.44 ± 22%      +0.9        3.92 ± >> 11%  perf-profile.children.cycles-pp.osq_lock >>        0.84 ±  4%      +1.4        2.24 ±  8%      +1.4        2.26 ± >> 7%  perf-profile.children.cycles-pp.__irq_exit_rcu >>        1.02 ±  3%      +2.0        3.06 ±  5%      +1.8        2.79 ± >> 8%  perf-profile.children.cycles-pp.update_process_times >>        1.04 ±  3%      +2.1        3.17 ±  4%      +1.9        2.93 ± >> 7%  perf-profile.children.cycles-pp.tick_sched_handle >>        1.18 ±  4%      +2.5        3.71 ±  4%      +2.3        3.45 ± >> 9%  perf-profile.children.cycles-pp.tick_sched_timer >>        1.46 ±  3%      +3.8        5.24 ±  4%      +3.7        5.13 ± >> 7%  perf-profile.children.cycles-pp.__hrtimer_run_queues >>        1.70 ±  4%      +4.8        6.46 ±  3%      +4.7        6.42 ± >> 7%  perf-profile.children.cycles-pp.hrtimer_interrupt >>        1.71 ±  4%      +4.9        6.57 ±  3%      +4.8        6.55 ± >> 7%  perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt >>        2.61 ±  3%      +7.0        9.63 ±  4%      +7.2        9.80 ± >> 7%  perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt >>        3.30 ±  2%      +7.3       10.55 ±  5%      +7.6       10.90 ± >> 6%  perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt >>       36.61            +9.7       46.29 ±  5%     +12.6       49.18 ± >> 3%  perf-profile.children.cycles-pp.intel_idle >>       40.81           +17.9       58.69 ±  5%     +21.4       62.22 ± >> 3%  perf-profile.children.cycles-pp.cpuidle_enter_state >>       40.82           +17.9       58.74 ±  5%     +21.5       62.29 ± >> 3%  perf-profile.children.cycles-pp.cpuidle_enter >>       42.42           +19.4       61.78 ±  5%     +23.3       65.71 ± >> 3%  perf-profile.children.cycles-pp.start_secondary >>       42.74           +19.6       62.33 ±  5%     +23.4       66.10 ± >> 3%  perf-profile.children.cycles-pp.do_idle >>       42.74           +19.6       62.33 ±  5%     +23.4       66.10 ± >> 3%  perf-profile.children.cycles-pp.secondary_startup_64_no_verify >>       42.74           +19.6       62.33 ±  5%     +23.4       66.10 ± >> 3%  perf-profile.children.cycles-pp.cpu_startup_entry >>       41.33           +19.6       60.93 ±  6%     +23.4       64.71 ± >> 3%  perf-profile.children.cycles-pp.cpuidle_idle_call >>        7.20 ±  2%      -5.5        1.65 ± 19%      -5.7        1.54 ± >> 9%  perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath >>        4.96            -3.5        1.45 ± 20%      -3.5        1.41 ± >> 9%  perf-profile.self.cycles-pp.memset_erms >>        3.74 ±  3%      -2.7        1.01 ± 17%      -2.9        0.89 ± >> 12%  perf-profile.self.cycles-pp.path_init >>        3.76            -2.6        1.12 ± 22%      -2.7        1.05 ± >> 9%  perf-profile.self.cycles-pp.rwsem_spin_on_owner >>        1.83 ±  3%      -1.1        0.72 ± 20%      -1.1        0.71 ± >> 6%  perf-profile.self.cycles-pp.mnt_get_count >>        1.82            -1.1        0.72 ± 13%      -1.1        0.73 ± >> 12%  perf-profile.self.cycles-pp._raw_spin_lock_irqsave >>        1.22 ±  9%      -1.1        0.16 ± 57%      -1.0        0.25 ± >> 56%  perf-profile.self.cycles-pp.intel_idle_irq >>        1.15 ±  2%      -0.7        0.48 ± 24%      -0.7        0.42 ± >> 16%  perf-profile.self.cycles-pp.pcpu_alloc >>        0.81 ±  3%      -0.6        0.24 ± 23%      -0.6        0.19 ± >> 13%  perf-profile.self.cycles-pp.lookup_mnt >>        1.02            -0.5        0.52 ± 23%      -0.5        0.54 ± >> 5%  perf-profile.self.cycles-pp._find_next_bit >>        0.94 ±  2%      -0.5        0.47 ± 13%      -0.5        0.46 ± >> 11%  perf-profile.self.cycles-pp.stress_rndstr >>        1.57 ±  2%      -0.4        1.14 ± 13%      -0.6        0.98 ± >> 5%  perf-profile.self.cycles-pp._raw_spin_lock >>        0.49 ±  4%      -0.4        0.14 ± 31%      -0.3        0.14 ± >> 26%  perf-profile.self.cycles-pp.__sync_rcu_exp_select_node_cpus >>        0.74 ±  5%      -0.3        0.41 ± 28%      -0.4        0.37 ± >> 10%  perf-profile.self.cycles-pp.smp_call_function_single >>        0.55 ±  5%      -0.3        0.27 ± 25%      -0.3        0.24 ± >> 14%  perf-profile.self.cycles-pp.mnt_get_writers >>        0.30 ±  8%      -0.2        0.06 ± 54%      -0.2        0.06 ± >> 19%  perf-profile.self.cycles-pp.free_percpu >>        0.25 ±  7%      -0.2        0.07 ± 48%      -0.2        0.06 ± >> 50%  perf-profile.self.cycles-pp.pcpu_alloc_area >>        0.27 ±  4%      -0.2        0.09 ± 38%      -0.2        0.09 ± >> 26%  perf-profile.self.cycles-pp.__list_del_entry_valid >>        0.23 ±  4%      -0.2        0.05 ± 76%      -0.2        0.05 ± >> 59%  perf-profile.self.cycles-pp.__flush_smp_call_function_queue >>        0.27 ±  9%      -0.2        0.10 ±  9%      -0.2        0.08 ± >> 17%  perf-profile.self.cycles-pp.__slab_free >>        0.19 ±  7%      -0.2        0.03 ±101%      -0.2        0.03 >> ±102%  perf-profile.self.cycles-pp.rcu_exp_handler >>        0.30 ± 11%      -0.2        0.14 ± 23%      -0.2        0.13 ± >> 16%  perf-profile.self.cycles-pp.__schedule >>        0.25 ±  9%      -0.1        0.10 ± 25%      -0.2        0.09 ± >> 28%  perf-profile.self.cycles-pp.llist_add_batch >>        0.23 ±  7%      -0.1        0.10 ± 31%      -0.1        0.09 ± >> 30%  perf-profile.self.cycles-pp._find_next_zero_bit >>        0.48 ± 19%      -0.1        0.35 ± 14%      -0.2        0.32 ± >> 9%  perf-profile.self.cycles-pp.idle_cpu >>        0.19 ± 10%      -0.1        0.06 ± 83%      -0.1        0.05 ± >> 77%  perf-profile.self.cycles-pp.rcu_report_exp_cpu_mult >>        0.28 ±  7%      -0.1        0.16 ± 18%      -0.1        0.16 ± >> 22%  perf-profile.self.cycles-pp.mutex_spin_on_owner >>        0.22 ±  6%      -0.1        0.09 ± 29%      -0.1        0.09 ± >> 23%  perf-profile.self.cycles-pp.mntput_no_expire >>        0.26 ±  7%      -0.1        0.13 ± 21%      -0.2        0.09 ± >> 23%  perf-profile.self.cycles-pp.__legitimize_mnt >>        0.44            -0.1        0.32 ± 11%      -0.2        0.28 ± >> 14%  perf-profile.self.cycles-pp.strncpy_from_user >>        0.15 ±  6%      -0.1        0.03 ±106%      -0.1        0.04 ± >> 73%  perf-profile.self.cycles-pp.mutex_lock_killable >>        0.21 ± 11%      -0.1        0.09 ± 15%      -0.1        0.09 ± >> 46%  perf-profile.self.cycles-pp.step_into >>        0.15 ±  7%      -0.1        0.03 ±100%      -0.1        0.05 ± >> 82%  perf-profile.self.cycles-pp.llist_reverse_order >>        0.18 ±  5%      -0.1        0.07 ± 16%      -0.1        0.06 ± >> 27%  perf-profile.self.cycles-pp.lockref_get >>        0.16 ±  5%      -0.1        0.04 ± 73%      -0.1        0.03 >> ±100%  perf-profile.self.cycles-pp.down_write >>        0.16 ±  7%      -0.1        0.04 ± 73%      -0.1        0.06 ± >> 24%  perf-profile.self.cycles-pp.__list_add_valid >>        0.15 ±  8%      -0.1        0.04 ± 73%      -0.1        0.05 ± >> 47%  perf-profile.self.cycles-pp.get_nr_dirty_inodes >>        0.15 ±  6%      -0.1        0.04 ± 75%      -0.1        0.05 ± >> 47%  perf-profile.self.cycles-pp.pcpu_next_fit_region >>        0.14 ±  6%      -0.1        0.05 ± 76%      -0.1        0.02 ± >> 99%  perf-profile.self.cycles-pp.up_write >>        0.23 ±  6%      -0.1        0.13 ± 10%      -0.1        0.13 ± >> 20%  perf-profile.self.cycles-pp.stress_mwc32 >>        0.16 ±  7%      -0.1        0.06 ± 11%      -0.1        0.07 ± >> 25%  perf-profile.self.cycles-pp.get_nr_inodes >>        0.12 ± 13%      -0.1        0.03 ±100%      -0.1        0.05 ± >> 46%  perf-profile.self.cycles-pp.__lookup_mnt >>        0.12 ±  6%      -0.1        0.03 ±108%      -0.1 >> 0.00        perf-profile.self.cycles-pp.__bitmap_clear >>        0.18 ±  5%      -0.1        0.10 ± 18%      -0.1        0.10 ± >> 18%  perf-profile.self.cycles-pp.__d_lookup_rcu >>        0.18 ±  4%      -0.1        0.11 ± 27%      -0.1        0.10 ± >> 25%  perf-profile.self.cycles-pp.lockref_put_return >>        0.11 ± 11%      -0.1        0.03 ±101%      -0.1        0.02 >> ±141%  perf-profile.self.cycles-pp.apparmor_capable >>        0.16 ±  6%      -0.1        0.08 ± 17%      -0.1        0.08 ± >> 49%  perf-profile.self.cycles-pp.mutex_unlock >>        0.15 ±  8%      -0.1        0.08 ± 17%      -0.1        0.08 ± >> 9%  perf-profile.self.cycles-pp.task_work_add >>        0.16 ±  8%      -0.1        0.09 ± 14%      -0.1        0.07 ± >> 56%  perf-profile.self.cycles-pp.__call_rcu_common >>        0.12 ± 23%      -0.1        0.06 ± 52%      -0.1        0.03 >> ±101%  perf-profile.self.cycles-pp.enqueue_task_fair >>        0.12 ± 10%      -0.0        0.07 ± 30%      -0.1        0.04 ± >> 79%  perf-profile.self.cycles-pp._raw_read_lock_irqsave >>        0.08 ±  6%      -0.0        0.04 ± 72%      -0.0        0.05 ± >> 46%  perf-profile.self.cycles-pp.generic_permission >>        0.11 ±  9%      -0.0        0.08 ± 21%      -0.1        0.06 ± >> 65%  perf-profile.self.cycles-pp.__switch_to >>        0.09 ±  9%      -0.0        0.06 ± 14%      -0.1        0.01 >> ±223%  perf-profile.self.cycles-pp.rcu_all_qs >>        0.08 ± 20%      +0.0        0.12 ± 15%      +0.0        0.12 ± >> 14%  perf-profile.self.cycles-pp.copy_user_enhanced_fast_string >>        0.00            +0.0        0.04 ±102%      +0.1        0.07 ± >> 26%  perf-profile.self.cycles-pp.get_next_timer_interrupt >>        0.04 ± 45%      +0.0        0.08 ± 25%      +0.0        0.08 ± >> 20%  perf-profile.self.cycles-pp.__entry_text_start >>        0.08 ± 10%      +0.0        0.12 ± 15%      +0.1        0.13 ± >> 34%  perf-profile.self.cycles-pp.load_balance >>        0.04 ± 45%      +0.0        0.08 ± 58%      +0.1        0.13 ± >> 30%  perf-profile.self.cycles-pp.__pthread_mutex_unlock_usercnt >>        0.00            +0.0        0.04 ± 73%      +0.1        0.07 ± >> 18%  perf-profile.self.cycles-pp.clockevents_program_event >>        0.16 ±  3%      +0.0        0.20 ± 26%      +0.0        0.19 ± >> 7%  perf-profile.self.cycles-pp.kmem_cache_free >>        0.00            +0.0        0.04 ± 77%      +0.1        0.07 ± >> 14%  perf-profile.self.cycles-pp.tick_sched_timer >>        0.00            +0.0        0.04 ±108%      +0.1        0.08 ± >> 24%  perf-profile.self.cycles-pp.rb_next >>        0.00            +0.0        0.04 ±103%      +0.1        0.08 ± >> 17%  perf-profile.self.cycles-pp.tsc_verify_tsc_adjust >>        0.00            +0.0        0.04 ± 77%      +0.1        0.11 ± >> 22%  perf-profile.self.cycles-pp.nr_iowait_cpu >>        0.00            +0.0        0.05 ±107%      +0.1        0.06 ± >> 28%  perf-profile.self.cycles-pp.native_apic_mem_write >>        0.00            +0.1        0.05 ± 51%      +0.1        0.08 ± >> 17%  perf-profile.self.cycles-pp.error_entry >>        0.00            +0.1        0.06 ±100%      +0.1        0.11 ± >> 32%  perf-profile.self.cycles-pp.tick_check_oneshot_broadcast_this_cpu >>        0.03 ±101%      +0.1        0.09 ± 26%      +0.1        0.12 ± >> 36%  perf-profile.self.cycles-pp.__pthread_mutex_lock >>        0.00            +0.1        0.06 ± 51%      +0.1        0.09 ± >> 18%  perf-profile.self.cycles-pp.asm_sysvec_apic_timer_interrupt >>        0.00            +0.1        0.06 ± 57%      +0.1        0.11 ± >> 22%  perf-profile.self.cycles-pp.rb_insert_color >>        0.00            +0.1        0.07 ± 48%      +0.1        0.08 ± >> 26%  perf-profile.self.cycles-pp.call_cpuidle >>        0.00            +0.1        0.07 ± 21%      +0.0        0.05 ± >> 46%  perf-profile.self.cycles-pp.show_mountinfo >>        0.00            +0.1        0.07 ± 18%      +0.1        0.06 ± >> 49%  perf-profile.self.cycles-pp._hashmap_iterate >>        0.00            +0.1        0.07 ± 27%      +0.1        0.06 ± >> 19%  perf-profile.self.cycles-pp.tick_nohz_tick_stopped >>        0.00            +0.1        0.07 ± 29%      +0.1        0.08 ± >> 24%  perf-profile.self.cycles-pp.siphash24_init >>        0.00            +0.1        0.07 ± 45%      +0.1        0.08 ± >> 29%  perf-profile.self.cycles-pp.get_cpu_device >>        0.00            +0.1        0.08 ± 21%      +0.0        0.04 >> ±105%  perf-profile.self.cycles-pp.string_escape_mem >>        0.00            +0.1        0.08 ± 16%      +0.1        0.06 ± >> 51%  perf-profile.self.cycles-pp.up_read >>        0.00            +0.1        0.08 ± 32%      +0.1        0.13 ± >> 14%  perf-profile.self.cycles-pp.sysvec_apic_timer_interrupt >>        0.11 ±  6%      +0.1        0.18 ± 14%      +0.1        0.19 ± >> 17%  perf-profile.self.cycles-pp.rcu_pending >>        0.16 ±  6%      +0.1        0.24 ± 16%      +0.0        0.18 ± >> 13%  perf-profile.self.cycles-pp.ct_kernel_exit_state >>        0.00            +0.1        0.08 ± 27%      +0.0        0.05 ± >> 73%  perf-profile.self.cycles-pp.down_read >>        0.00            +0.1        0.08 ± 81%      +0.1        0.11 ± >> 15%  perf-profile.self.cycles-pp.rb_erase >>        0.04 ± 71%      +0.1        0.12 ± 27%      +0.1        0.10 ± >> 29%  perf-profile.self.cycles-pp.__update_blocked_fair >>        0.04 ± 44%      +0.1        0.13 ± 17%      +0.1        0.16 ± >> 15%  perf-profile.self.cycles-pp.do_idle >>        0.00            +0.1        0.09 ± 32%      +0.1        0.06 ± >> 47%  perf-profile.self.cycles-pp.__irq_exit_rcu >>        0.00            +0.1        0.09 ± 52%      +0.1        0.10 ± >> 28%  perf-profile.self.cycles-pp.path_hash_func >>        0.00            +0.1        0.09 ± 24%      +0.1        0.08 ± >> 17%  perf-profile.self.cycles-pp.__do_softirq >>        0.00            +0.1        0.09 ± 58%      +0.1        0.09 ± >> 27%  perf-profile.self.cycles-pp.try_check_zero >>        0.00            +0.1        0.09 ± 44%      +0.1        0.11 ± >> 16%  perf-profile.self.cycles-pp.rebalance_domains >>        0.00            +0.1        0.09 ± 64%      +0.1        0.12 ± >> 34%  perf-profile.self.cycles-pp.run_posix_cpu_timers >>        0.00            +0.1        0.09 ± 13%      +0.1        0.07 ± >> 47%  perf-profile.self.cycles-pp.strchr >>        0.01 ±223%      +0.1        0.10 ± 22%      +0.0        0.06 ± >> 76%  perf-profile.self.cycles-pp.prepend_path >>        0.00            +0.1        0.10 ± 23%      +0.1        0.05 ± >> 49%  perf-profile.self.cycles-pp.number >>        0.06 ± 11%      +0.1        0.17 ± 17%      +0.1        0.14 ± >> 21%  perf-profile.self.cycles-pp.memcpy_erms >>        0.04 ± 71%      +0.1        0.14 ± 14%      +0.1        0.14 ± >> 24%  perf-profile.self.cycles-pp.cfree >>        0.00            +0.1        0.11 ± 12%      +0.1        0.09 ± >> 33%  perf-profile.self.cycles-pp.vsnprintf >>        0.00            +0.1        0.11 ± 26%      +0.1        0.09 ± >> 27%  perf-profile.self.cycles-pp.strlen >>        0.04 ± 71%      +0.1        0.15 ± 34%      +0.1        0.12 ± >> 20%  perf-profile.self.cycles-pp.ct_kernel_enter >>        0.00            +0.1        0.11 ± 22%      +0.1        0.14 ± >> 24%  perf-profile.self.cycles-pp.timerqueue_add >>        0.00            +0.1        0.11 ± 15%      +0.2        0.15 ± >> 8%  perf-profile.self.cycles-pp.hrtimer_interrupt >>        0.11 ± 12%      +0.1        0.22 ±  8%      +0.1        0.22 ± >> 16%  perf-profile.self.cycles-pp._raw_spin_trylock >>        0.00            +0.1        0.12 ± 29%      +0.2        0.17 ± >> 13%  perf-profile.self.cycles-pp.scheduler_tick >>        0.00            +0.1        0.12 ± 24%      +0.1        0.12 ± >> 26%  perf-profile.self.cycles-pp.rdrand >>        0.01 ±223%      +0.1        0.13 ± 23%      +0.2        0.17 ± >> 11%  perf-profile.self.cycles-pp.tick_nohz_next_event >>        0.04 ± 71%      +0.1        0.16 ± 14%      +0.1        0.18 ± >> 16%  perf-profile.self.cycles-pp.irqtime_account_irq >>        0.00            +0.1        0.13 ± 17%      +0.1        0.09 ± >> 23%  perf-profile.self.cycles-pp.rcu_sched_clock_irq >>        0.00            +0.1        0.14 ± 26%      +0.1        0.14 ± >> 23%  perf-profile.self.cycles-pp.irqtime_account_process_tick >>        0.00            +0.1        0.14 ± 17%      +0.1        0.10 ± >> 21%  perf-profile.self.cycles-pp.format_decode >>        0.15 ± 10%      +0.1        0.29 ± 33%      +0.1        0.20 ± >> 24%  perf-profile.self.cycles-pp.trigger_load_balance >>        0.03 ± 70%      +0.1        0.18 ± 21%      +0.1        0.17 ± >> 14%  perf-profile.self.cycles-pp.siphash24_finalize >>        0.00            +0.1        0.14 ± 34%      +0.2        0.19 ± >> 24%  perf-profile.self.cycles-pp.perf_rotate_context >>        0.01 ±223%      +0.1        0.16 ± 28%      +0.1        0.16 ± >> 29%  perf-profile.self.cycles-pp.strchr@plt >>        0.00            +0.2        0.15 ± 24%      +0.2        0.20 ± >> 28%  perf-profile.self.cycles-pp.__hrtimer_next_event_base >>        0.06 ± 13%      +0.2        0.21 ± 21%      +0.1        0.20 ± >> 29%  perf-profile.self.cycles-pp.path_compare >>        0.10 ± 37%      +0.2        0.26 ± 28%      +0.2        0.25 ± >> 28%  perf-profile.self.cycles-pp.tick_sched_do_timer >>        0.01 ±223%      +0.2        0.17 ± 24%      +0.2        0.18 ± >> 23%  perf-profile.self.cycles-pp.update_process_times >>        0.02 ± 99%      +0.2        0.18 ± 21%      +0.2        0.22 ± >> 9%  perf-profile.self.cycles-pp.cpuidle_idle_call >>        0.01 ±223%      +0.2        0.18 ± 15%      +0.1        0.16 ± >> 12%  perf-profile.self.cycles-pp.delay_tsc >>        0.00            +0.2        0.17 ± 21%      +0.1        0.12 ± >> 16%  perf-profile.self.cycles-pp.device_read_db_internal_filename >>        0.03 ±100%      +0.2        0.20 ± 16%      +0.2        0.20 ± >> 14%  perf-profile.self.cycles-pp.check_cpu_stall >>        0.07 ± 17%      +0.2        0.25 ± 24%      +0.2        0.24 ± >> 16%  perf-profile.self.cycles-pp.update_irq_load_avg >>        0.00            +0.2        0.19 ± 20%      +0.2        0.24 ± >> 21%  perf-profile.self.cycles-pp.__hrtimer_run_queues >>        0.02 ±141%      +0.2        0.22 ± 31%      +0.2        0.18 ± >> 28%  perf-profile.self.cycles-pp.ktime_get_update_offsets_now >>        0.21 ± 13%      +0.2        0.42 ±  5%      +0.2        0.45 ± >> 12%  perf-profile.self.cycles-pp.native_sched_clock >>        0.05 ± 46%      +0.2        0.29 ± 10%      +0.2        0.27 ± >> 15%  perf-profile.self.cycles-pp.__intel_pmu_enable_all >>        0.22 ±  4%      +0.2        0.46 ±  6%      +0.2        0.42 ± >> 8%  perf-profile.self.cycles-pp.native_irq_return_iret >>        0.17 ±  9%      +0.3        0.46 ± 35%      +0.3        0.48 ± >> 28%  perf-profile.self.cycles-pp.unit_name_is_valid >>        0.00            +0.3        0.30 ± 19%      +0.2        0.25 ± >> 23%  perf-profile.self.cycles-pp.memcpy_toio >>        0.10 ± 10%      +0.3        0.41 ± 17%      +0.2        0.35 ± >> 24%  perf-profile.self.cycles-pp.ktime_get >>        0.07 ±  9%      +0.3        0.38 ± 10%      +0.3        0.34 ± >> 12%  perf-profile.self.cycles-pp.read_tsc >>        0.08 ± 11%      +0.4        0.44 ± 11%      +0.4        0.45 ± >> 15%  perf-profile.self.cycles-pp.lapic_next_deadline >>        0.21 ± 13%      +0.4        0.58 ± 11%      +0.3        0.47 ± >> 19%  perf-profile.self.cycles-pp.arch_scale_freq_tick >>        0.12 ± 10%      +0.4        0.55 ± 21%      +0.4        0.48 ± >> 24%  perf-profile.self.cycles-pp.malloc >>        0.09 ± 13%      +0.6        0.70 ± 18%      +0.6        0.73 ± >> 8%  perf-profile.self.cycles-pp.io_serial_in >>        0.24 ±  6%      +0.6        0.84 ± 26%      +0.6        0.80 ± >> 22%  perf-profile.self.cycles-pp.siphash24_compress >>        0.21 ± 10%      +0.6        0.86 ±  6%      +0.6        0.86 ± >> 5%  perf-profile.self.cycles-pp.menu_select >>        0.20 ±  2%      +0.9        1.13 ± 12%      +1.0        1.23 ± >> 4%  perf-profile.self.cycles-pp.cpuidle_enter_state >>        3.02 ±  5%      +1.4        4.40 ± 22%      +0.9        3.89 ± >> 11%  perf-profile.self.cycles-pp.osq_lock >>       36.61            +9.7       46.29 ±  5%     +12.6       49.18 ± >> 3%  perf-profile.self.cycles-pp.intel_idle >> >> -- >> Best Regards, >> Yujie > -- Thanks, Qi