lkml.org 
[lkml]   [2020]   [Jul]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: [LKP] [xfs] a5949d3fae: aim7.jobs-per-min -33.6% regression
From
Date


On 7/7/2020 2:30 AM, Darrick J. Wong wrote:
> On Wed, Jul 01, 2020 at 03:49:52PM +0800, Xing Zhengjun wrote:
>>
>>
>> On 6/10/2020 11:07 AM, Xing Zhengjun wrote:
>>> Hi Darrick,
>>>
>>>    Do you have time to take a look at this? Thanks.
>>>
>>
>>
>> Ping...
>
> Yes, that decrease is the expected end result of making the write path
> take a longer route to avoid a file corruption vector.
>
> --D
>

Thanks for the explanation, We will stop tracking it.

>>
>>
>>> On 6/6/2020 11:48 PM, kernel test robot wrote:
>>>> Greeting,
>>>>
>>>> FYI, we noticed a -33.6% regression of aim7.jobs-per-min due to commit:
>>>>
>>>>
>>>> commit: a5949d3faedf492fa7863b914da408047ab46eb0 ("xfs: force writes
>>>> to delalloc regions to unwritten")
>>>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
>>>>
>>>> in testcase: aim7
>>>> on test machine: 48 threads Intel(R) Xeon(R) CPU E5-2697 v2 @
>>>> 2.70GHz with 64G memory
>>>> with following parameters:
>>>>
>>>>     disk: 1BRD_48G
>>>>     fs: xfs
>>>>     test: sync_disk_rw
>>>>     load: 600
>>>>     cpufreq_governor: performance
>>>>     ucode: 0x42e
>>>>
>>>> test-description: AIM7 is a traditional UNIX system level benchmark
>>>> suite which is used to test and measure the performance of multiuser
>>>> system.
>>>> test-url: https://sourceforge.net/projects/aimbench/files/aim-suite7/
>>>>
>>>>
>>>>
>>>> If you fix the issue, kindly add following tag
>>>> Reported-by: kernel test robot <rong.a.chen@intel.com>
>>>>
>>>>
>>>> Details are as below:
>>>> -------------------------------------------------------------------------------------------------->
>>>>
>>>>
>>>>
>>>> To reproduce:
>>>>
>>>>          git clone https://github.com/intel/lkp-tests.git
>>>>          cd lkp-tests
>>>>          bin/lkp install job.yaml  # job file is attached in this email
>>>>          bin/lkp run     job.yaml
>>>>
>>>> =========================================================================================
>>>>
>>>> compiler/cpufreq_governor/disk/fs/kconfig/load/rootfs/tbox_group/test/testcase/ucode:
>>>>
>>>> gcc-9/performance/1BRD_48G/xfs/x86_64-rhel-7.6/600/debian-x86_64-20191114.cgz/lkp-ivb-2ep1/sync_disk_rw/aim7/0x42e
>>>>
>>>>
>>>> commit:
>>>>    590b16516e ("xfs: refactor xfs_iomap_prealloc_size")
>>>>    a5949d3fae ("xfs: force writes to delalloc regions to unwritten")
>>>>
>>>> 590b16516ef38e2e a5949d3faedf492fa7863b914da
>>>> ---------------- ---------------------------
>>>>         fail:runs  %reproduction    fail:runs
>>>>             |             |             |
>>>>             :4           50%           2:4
>>>> dmesg.WARNING:at#for_ip_swapgs_restore_regs_and_return_to_usermode/0x
>>>>           %stddev     %change         %stddev
>>>>               \          |                \
>>>>       35272           -33.6%      23430        aim7.jobs-per-min
>>>>      102.13           +50.5%     153.75        aim7.time.elapsed_time
>>>>      102.13           +50.5%     153.75        aim7.time.elapsed_time.max
>>>>     1388038           +40.2%    1945838
>>>> aim7.time.involuntary_context_switches
>>>>       43420 ±  2%     +13.4%      49255 ±  2%
>>>> aim7.time.minor_page_faults
>>>>        3123           +44.2%       4504 ±  2%  aim7.time.system_time
>>>>       59.31            +6.5%      63.18        aim7.time.user_time
>>>>    48595108           +58.6%   77064959
>>>> aim7.time.voluntary_context_switches
>>>>        1.44           -28.8%       1.02        iostat.cpu.user
>>>>        0.07 ±  6%      +0.4        0.44 ±  7%  mpstat.cpu.all.iowait%
>>>>        1.44            -0.4        1.02        mpstat.cpu.all.usr%
>>>>        8632 ± 50%     +75.6%      15156 ± 34%
>>>> numa-meminfo.node0.KernelStack
>>>>        6583 ±136%    +106.0%      13562 ± 82%
>>>> numa-meminfo.node0.PageTables
>>>>       63325 ± 11%     +14.3%      72352 ± 12%
>>>> numa-meminfo.node0.SUnreclaim
>>>>        8647 ± 50%     +75.3%      15156 ± 34%
>>>> numa-vmstat.node0.nr_kernel_stack
>>>>        1656 ±136%    +104.6%       3389 ± 82%
>>>> numa-vmstat.node0.nr_page_table_pages
>>>>       15831 ± 11%     +14.3%      18087 ± 12%
>>>> numa-vmstat.node0.nr_slab_unreclaimable
>>>>       93640 ±  3%     +41.2%     132211 ±  2%  meminfo.AnonHugePages
>>>>       21641           +39.9%      30271 ±  4%  meminfo.KernelStack
>>>>      129269           +12.3%     145114        meminfo.SUnreclaim
>>>>       28000           -31.2%      19275        meminfo.max_used_kB
>>>>     1269307           -26.9%     927657        vmstat.io.bo
>>>>      149.75 ±  3%     -17.4%     123.75 ±  4%  vmstat.procs.r
>>>>      718992           +13.3%     814567        vmstat.system.cs
>>>>      231397            -9.3%     209881 ±  2%  vmstat.system.in
>>>>   6.774e+08           +70.0%  1.152e+09        cpuidle.C1.time
>>>>    18203372           +60.4%   29198744        cpuidle.C1.usage
>>>>   2.569e+08 ± 18%     +81.8%  4.672e+08 ±  5%  cpuidle.C1E.time
>>>>     2691402 ± 13%     +98.7%    5346901 ±  3%  cpuidle.C1E.usage
>>>>      990350           +95.0%    1931226 ±  2%  cpuidle.POLL.time
>>>>      520061           +97.7%    1028004 ±  2%  cpuidle.POLL.usage
>>>>       77231            +1.8%      78602        proc-vmstat.nr_active_anon
>>>>       19868            +3.8%      20615        proc-vmstat.nr_dirty
>>>>      381302            +1.0%     384969        proc-vmstat.nr_file_pages
>>>>        4388            -2.7%       4270
>>>> proc-vmstat.nr_inactive_anon
>>>>       69865            +4.7%      73155
>>>> proc-vmstat.nr_inactive_file
>>>>       21615           +40.0%      30251 ±  4%
>>>> proc-vmstat.nr_kernel_stack
>>>>        7363            -3.2%       7127        proc-vmstat.nr_mapped
>>>>       12595 ±  3%      +5.2%      13255 ±  4%  proc-vmstat.nr_shmem
>>>>       19619            +3.2%      20247
>>>> proc-vmstat.nr_slab_reclaimable
>>>>       32316           +12.3%      36280
>>>> proc-vmstat.nr_slab_unreclaimable
>>>>       77231            +1.8%      78602
>>>> proc-vmstat.nr_zone_active_anon
>>>>        4388            -2.7%       4270
>>>> proc-vmstat.nr_zone_inactive_anon
>>>>       69865            +4.7%      73155
>>>> proc-vmstat.nr_zone_inactive_file
>>>>       19165            +3.4%      19824
>>>> proc-vmstat.nr_zone_write_pending
>>>>        9845 ± 27%     +68.0%      16539 ± 37%
>>>> proc-vmstat.numa_hint_faults_local
>>>>     8195135            +2.5%    8402537        proc-vmstat.numa_hit
>>>>     8179414            +2.5%    8386209        proc-vmstat.numa_local
>>>>       15720            +3.9%      16328        proc-vmstat.numa_other
>>>>     8281098            +2.4%    8477157        proc-vmstat.pgalloc_normal
>>>>      342925 ±  2%     +39.8%     479279        proc-vmstat.pgfault
>>>>     8132789            +2.0%    8298189        proc-vmstat.pgfree
>>>>   1.327e+08            +9.7%  1.455e+08        proc-vmstat.pgpgout
>>>>      394.00          +151.7%     991.75 ±  5%
>>>> slabinfo.biovec-max.active_objs
>>>>      394.00          +152.1%     993.25 ±  5%
>>>> slabinfo.biovec-max.num_objs
>>>>       21028 ±  6%      -7.8%      19389 ±  5%  slabinfo.filp.active_objs
>>>>       21520 ±  5%      -7.6%      19886 ±  5%  slabinfo.filp.num_objs
>>>>        4955          +234.9%      16594 ±  2%
>>>> slabinfo.kmalloc-256.active_objs
>>>>       77.25          +238.5%     261.50 ±  2%
>>>> slabinfo.kmalloc-256.active_slabs
>>>>        4955          +238.4%      16768 ±  2%
>>>> slabinfo.kmalloc-256.num_objs
>>>>       77.25          +238.5%     261.50 ±  2%
>>>> slabinfo.kmalloc-256.num_slabs
>>>>        3368 ±  2%     +22.6%       4128 ±  2%
>>>> slabinfo.kmalloc-2k.active_objs
>>>>        3446 ±  2%     +21.9%       4200 ±  2%
>>>> slabinfo.kmalloc-2k.num_objs
>>>>       11815 ±  8%     +29.0%      15239 ±  3%
>>>> slabinfo.kmalloc-512.active_objs
>>>>       11824 ±  9%     +29.0%      15249 ±  3%
>>>> slabinfo.kmalloc-512.num_objs
>>>>        6849 ±  5%      +9.0%       7463 ±  2%
>>>> slabinfo.kmalloc-96.active_objs
>>>>        6906 ±  5%      +8.9%       7521 ±  2%
>>>> slabinfo.kmalloc-96.num_objs
>>>>      765.00 ±  4%     +25.0%     956.25 ±  5%
>>>> slabinfo.mnt_cache.active_objs
>>>>      765.00 ±  4%     +25.0%     956.25 ±  5%
>>>> slabinfo.mnt_cache.num_objs
>>>>       13866           +12.5%      15594
>>>> slabinfo.proc_inode_cache.active_objs
>>>>       13905           +12.2%      15598
>>>> slabinfo.proc_inode_cache.num_objs
>>>>        1954           +24.4%       2431 ±  3%
>>>> slabinfo.sighand_cache.active_objs
>>>>        1954           +24.4%       2431 ±  3%
>>>> slabinfo.sighand_cache.num_objs
>>>>        2596           +13.7%       2951 ±  2%
>>>> slabinfo.signal_cache.active_objs
>>>>        2596           +13.7%       2951 ±  2%
>>>> slabinfo.signal_cache.num_objs
>>>>        4368 ±  6%     -11.7%       3856 ±  2%
>>>> slabinfo.skbuff_head_cache.num_objs
>>>>      822.25 ± 11%     +21.4%     998.50 ±  3%
>>>> slabinfo.task_group.active_objs
>>>>      822.25 ± 11%     +21.4%     998.50 ±  3%
>>>> slabinfo.task_group.num_objs
>>>>        1612           +32.1%       2129 ±  3%
>>>> slabinfo.task_struct.active_objs
>>>>      540.50           +31.6%     711.25 ±  3%
>>>> slabinfo.task_struct.active_slabs
>>>>        1622           +31.6%       2134 ±  3%
>>>> slabinfo.task_struct.num_objs
>>>>      540.50           +31.6%     711.25 ±  3%
>>>> slabinfo.task_struct.num_slabs
>>>>        6905           +20.0%       8289
>>>> slabinfo.vmap_area.active_objs
>>>>        6933           +20.0%       8322
>>>> slabinfo.vmap_area.num_objs
>>>>      646.25 ±  7%    +195.2%       1907
>>>> slabinfo.xfs_buf.active_objs
>>>>      646.25 ±  7%    +195.2%       1907        slabinfo.xfs_buf.num_objs
>>>>        1.78 ±  2%      -0.3        1.51 ±  2%
>>>> perf-stat.i.branch-miss-rate%
>>>>    64079048            +3.9%   66551536        perf-stat.i.branch-misses
>>>>       17.99            -4.2       13.84
>>>> perf-stat.i.cache-miss-rate%
>>>>    66630506           -16.5%   55647472        perf-stat.i.cache-misses
>>>>   3.899e+08            +9.4%  4.265e+08
>>>> perf-stat.i.cache-references
>>>>      731029           +13.1%     826756
>>>> perf-stat.i.context-switches
>>>>        3.79            -1.7%       3.73        perf-stat.i.cpi
>>>>      113320            +8.3%     122693        perf-stat.i.cpu-migrations
>>>>        1498           +20.4%       1803
>>>> perf-stat.i.cycles-between-cache-misses
>>>>    50450547 ±  4%      -8.0%   46402743 ±  5%
>>>> perf-stat.i.dTLB-load-misses
>>>>        0.20 ±  3%      -0.0        0.16 ±  2%
>>>> perf-stat.i.dTLB-store-miss-rate%
>>>>     4587096 ±  3%     -22.9%    3538422 ±  2%
>>>> perf-stat.i.dTLB-store-misses
>>>>   2.297e+09            -5.6%  2.168e+09        perf-stat.i.dTLB-stores
>>>>       86.68            -2.4       84.25
>>>> perf-stat.i.iTLB-load-miss-rate%
>>>>    19484791            -6.1%   18297125
>>>> perf-stat.i.iTLB-load-misses
>>>>     3060023           +14.4%    3501587 ±  4%  perf-stat.i.iTLB-loads
>>>>        0.29            -1.5%       0.28        perf-stat.i.ipc
>>>>        0.21 ±  5%     -20.5%       0.17 ±  6%  perf-stat.i.metric.K/sec
>>>>        3120            -5.8%       2938        perf-stat.i.minor-faults
>>>>       42.61            +1.2       43.83
>>>> perf-stat.i.node-load-miss-rate%
>>>>    24887573           -10.1%   22384693
>>>> perf-stat.i.node-load-misses
>>>>    31814752           -12.5%   27830232        perf-stat.i.node-loads
>>>>       36.62            +0.8       37.45
>>>> perf-stat.i.node-store-miss-rate%
>>>>    22617213           -20.3%   18015607
>>>> perf-stat.i.node-store-misses
>>>>    35707079           -19.9%   28604167        perf-stat.i.node-stores
>>>>        3120            -5.8%       2938        perf-stat.i.page-faults
>>>>       13.96            +9.3%      15.25        perf-stat.overall.MPKI
>>>>        0.99            +0.0        1.03
>>>> perf-stat.overall.branch-miss-rate%
>>>>       17.09            -4.0       13.05
>>>> perf-stat.overall.cache-miss-rate%
>>>>        1555           +19.4%       1856
>>>> perf-stat.overall.cycles-between-cache-misses
>>>>        0.69 ±  4%      -0.1        0.64 ±  5%
>>>> perf-stat.overall.dTLB-load-miss-rate%
>>>>        0.20 ±  3%      -0.0        0.16 ±  2%
>>>> perf-stat.overall.dTLB-store-miss-rate%
>>>>       86.43            -2.5       83.94
>>>> perf-stat.overall.iTLB-load-miss-rate%
>>>>        1433            +6.7%       1528
>>>> perf-stat.overall.instructions-per-iTLB-miss
>>>>       43.89            +0.7       44.58
>>>> perf-stat.overall.node-load-miss-rate%
>>>>   6.389e+09            +0.8%  6.438e+09
>>>> perf-stat.ps.branch-instructions
>>>>    63452800            +4.2%   66110454        perf-stat.ps.branch-misses
>>>>    65977234           -16.2%   55277568        perf-stat.ps.cache-misses
>>>>   3.861e+08            +9.8%  4.237e+08
>>>> perf-stat.ps.cache-references
>>>>      723870           +13.5%     821265
>>>> perf-stat.ps.context-switches
>>>>      112209            +8.6%     121881
>>>> perf-stat.ps.cpu-migrations
>>>>    49955941 ±  4%      -7.7%   46094243 ±  5%
>>>> perf-stat.ps.dTLB-load-misses
>>>>     4542094 ±  3%     -22.6%    3514899 ±  2%
>>>> perf-stat.ps.dTLB-store-misses
>>>>   2.275e+09            -5.3%  2.154e+09        perf-stat.ps.dTLB-stores
>>>>    19294019            -5.8%   18175580
>>>> perf-stat.ps.iTLB-load-misses
>>>>     3030049           +14.8%    3478314 ±  4%  perf-stat.ps.iTLB-loads
>>>>        3089            -5.5%       2918        perf-stat.ps.minor-faults
>>>>    24643655            -9.8%   22235787
>>>> perf-stat.ps.node-load-misses
>>>>    31502901           -12.2%   27645143        perf-stat.ps.node-loads
>>>>    22395376           -20.1%   17895417
>>>> perf-stat.ps.node-store-misses
>>>>    35356897           -19.6%   28413366        perf-stat.ps.node-stores
>>>>        3089            -5.5%       2918        perf-stat.ps.page-faults
>>>>    2.85e+12           +50.6%  4.293e+12
>>>> perf-stat.total.instructions
>>>>       84193 ± 36%    +135.9%     198575 ± 25%
>>>> sched_debug.cfs_rq:/.MIN_vruntime.avg
>>>>      919053           +98.2%    1821460 ±  2%
>>>> sched_debug.cfs_rq:/.MIN_vruntime.max
>>>>      256686 ± 16%    +114.7%     551162 ± 13%
>>>> sched_debug.cfs_rq:/.MIN_vruntime.stddev
>>>>       19497          +106.7%      40307
>>>> sched_debug.cfs_rq:/.exec_clock.avg
>>>>       21240          +101.1%      42717
>>>> sched_debug.cfs_rq:/.exec_clock.max
>>>>       19079          +107.3%      39542
>>>> sched_debug.cfs_rq:/.exec_clock.min
>>>>      394.52 ±  2%     +54.3%     608.79 ± 13%
>>>> sched_debug.cfs_rq:/.exec_clock.stddev
>>>>      965.25 ± 15%     +59.7%       1541 ±  4%
>>>> sched_debug.cfs_rq:/.load_avg.avg
>>>>        2866 ± 21%     +57.1%       4503 ± 10%
>>>> sched_debug.cfs_rq:/.load_avg.max
>>>>      659.77 ± 15%     +68.9%       1114 ±  9%
>>>> sched_debug.cfs_rq:/.load_avg.stddev
>>>>       84193 ± 36%    +135.9%     198575 ± 25%
>>>> sched_debug.cfs_rq:/.max_vruntime.avg
>>>>      919053           +98.2%    1821462 ±  2%
>>>> sched_debug.cfs_rq:/.max_vruntime.max
>>>>      256686 ± 16%    +114.7%     551163 ± 13%
>>>> sched_debug.cfs_rq:/.max_vruntime.stddev
>>>>      901765           +97.1%    1777741 ±  2%
>>>> sched_debug.cfs_rq:/.min_vruntime.avg
>>>>      941794           +95.8%    1844445 ±  2%
>>>> sched_debug.cfs_rq:/.min_vruntime.max
>>>>      872925           +96.5%    1715658 ±  2%
>>>> sched_debug.cfs_rq:/.min_vruntime.min
>>>>       18358 ± 44%    +146.6%      45281 ± 34%
>>>> sched_debug.cfs_rq:/.min_vruntime.stddev
>>>>        1.22 ±  8%    +159.9%       3.18 ± 20%
>>>> sched_debug.cfs_rq:/.nr_spread_over.avg
>>>>        6.75 ± 23%     +65.4%      11.17 ± 17%
>>>> sched_debug.cfs_rq:/.nr_spread_over.max
>>>>        1.38 ± 11%     +61.4%       2.22 ± 12%
>>>> sched_debug.cfs_rq:/.nr_spread_over.stddev
>>>>       18357 ± 44%    +146.5%      45250 ± 34%
>>>> sched_debug.cfs_rq:/.spread0.stddev
>>>>      773.25 ±  3%     +13.2%     875.02
>>>> sched_debug.cfs_rq:/.util_avg.avg
>>>>      285624           -30.3%     199111 ±  4%
>>>> sched_debug.cpu.avg_idle.avg
>>>>      576507 ±  2%     -26.2%     425271 ±  7%
>>>> sched_debug.cpu.avg_idle.max
>>>>       51029 ± 68%     -83.4%       8469 ± 29%
>>>> sched_debug.cpu.avg_idle.min
>>>>      135149 ±  7%     -23.4%     103460 ± 10%
>>>> sched_debug.cpu.avg_idle.stddev
>>>>       57492           +51.0%      86835        sched_debug.cpu.clock.avg
>>>>       57495           +51.0%      86841        sched_debug.cpu.clock.max
>>>>       57487           +51.0%      86828        sched_debug.cpu.clock.min
>>>>        1.98 ±  4%     +74.8%       3.46 ± 10%
>>>> sched_debug.cpu.clock.stddev
>>>>       57492           +51.0%      86835
>>>> sched_debug.cpu.clock_task.avg
>>>>       57495           +51.0%      86841
>>>> sched_debug.cpu.clock_task.max
>>>>       57487           +51.0%      86828
>>>> sched_debug.cpu.clock_task.min
>>>>        1.98 ±  4%     +74.8%       3.46 ± 10%
>>>> sched_debug.cpu.clock_task.stddev
>>>>        2831           +43.0%       4049
>>>> sched_debug.cpu.curr->pid.max
>>>>      417519          +133.3%     974078
>>>> sched_debug.cpu.nr_switches.avg
>>>>      427854          +134.0%    1001016
>>>> sched_debug.cpu.nr_switches.max
>>>>      409942          +131.8%     950218
>>>> sched_debug.cpu.nr_switches.min
>>>>        4770 ± 42%    +238.7%      16156 ± 22%
>>>> sched_debug.cpu.nr_switches.stddev
>>>>        4.70 ± 23%     +37.1%       6.44 ±  4%
>>>> sched_debug.cpu.nr_uninterruptible.avg
>>>>      776.62 ± 19%     +86.1%       1445 ± 12%
>>>> sched_debug.cpu.nr_uninterruptible.max
>>>>     -472.62           +98.3%    -937.33
>>>> sched_debug.cpu.nr_uninterruptible.min
>>>>      289.80 ±  5%    +109.7%     607.68 ±  3%
>>>> sched_debug.cpu.nr_uninterruptible.stddev
>>>>      415893          +133.8%     972526
>>>> sched_debug.cpu.sched_count.avg
>>>>      423960          +135.5%     998457
>>>> sched_debug.cpu.sched_count.max
>>>>      409113          +132.0%     949068
>>>> sched_debug.cpu.sched_count.min
>>>>        4194 ± 50%    +274.3%      15698 ± 22%
>>>> sched_debug.cpu.sched_count.stddev
>>>>      110094          +125.8%     248571
>>>> sched_debug.cpu.sched_goidle.avg
>>>>      112236          +125.6%     253163
>>>> sched_debug.cpu.sched_goidle.max
>>>>      108281          +125.8%     244506
>>>> sched_debug.cpu.sched_goidle.min
>>>>      772.04 ± 16%    +175.8%       2129 ± 18%
>>>> sched_debug.cpu.sched_goidle.stddev
>>>>      297267          +137.2%     705173
>>>> sched_debug.cpu.ttwu_count.avg
>>>>      320112          +135.9%     755096
>>>> sched_debug.cpu.ttwu_count.max
>>>>      272487 ±  2%    +142.2%     660099
>>>> sched_debug.cpu.ttwu_count.min
>>>>       10662 ± 10%    +110.9%      22491 ± 18%
>>>> sched_debug.cpu.ttwu_count.stddev
>>>>       12295 ±  2%    +834.9%     114954 ±  2%
>>>> sched_debug.cpu.ttwu_local.avg
>>>>       13726 ±  2%    +769.7%     119376 ±  2%
>>>> sched_debug.cpu.ttwu_local.max
>>>>       10620 ±  4%    +942.4%     110710
>>>> sched_debug.cpu.ttwu_local.min
>>>>      693.29 ± 14%    +228.7%       2278 ± 18%
>>>> sched_debug.cpu.ttwu_local.stddev
>>>>       57488           +51.0%      86829        sched_debug.cpu_clk
>>>>       56998           +51.5%      86337        sched_debug.ktime
>>>>       57977           +50.6%      87317        sched_debug.sched_clk
>>>>       72.50 ± 37%     +55.5%     112.75 ± 26%
>>>> interrupts.36:PCI-MSI.2621442-edge.eth0-TxRx-1
>>>>       53.25 ±  2%     +84.5%      98.25 ± 22%
>>>> interrupts.37:PCI-MSI.2621443-edge.eth0-TxRx-2
>>>>       73.50 ± 26%    +116.3%     159.00 ± 60%
>>>> interrupts.38:PCI-MSI.2621444-edge.eth0-TxRx-3
>>>>      204541           +49.0%     304702 ±  2%
>>>> interrupts.CPU0.LOC:Local_timer_interrupts
>>>>      290268 ±  2%     +29.1%     374641 ±  4%
>>>> interrupts.CPU0.RES:Rescheduling_interrupts
>>>>      204580           +48.9%     304667 ±  2%
>>>> interrupts.CPU1.LOC:Local_timer_interrupts
>>>>      290342 ±  2%     +28.5%     373110 ±  4%
>>>> interrupts.CPU1.RES:Rescheduling_interrupts
>>>>      204537           +48.9%     304619 ±  2%
>>>> interrupts.CPU10.LOC:Local_timer_interrupts
>>>>      291605 ±  2%     +28.3%     374190 ±  4%
>>>> interrupts.CPU10.RES:Rescheduling_interrupts
>>>>      204140           +49.2%     304587 ±  2%
>>>> interrupts.CPU11.LOC:Local_timer_interrupts
>>>>      291988 ±  2%     +28.1%     374024 ±  4%
>>>> interrupts.CPU11.RES:Rescheduling_interrupts
>>>>      204367           +49.0%     304559 ±  2%
>>>> interrupts.CPU12.LOC:Local_timer_interrupts
>>>>      288632           +28.4%     370669 ±  3%
>>>> interrupts.CPU12.RES:Rescheduling_interrupts
>>>>      204555           +48.9%     304669 ±  2%
>>>> interrupts.CPU13.LOC:Local_timer_interrupts
>>>>      289354 ±  2%     +28.5%     371782 ±  3%
>>>> interrupts.CPU13.RES:Rescheduling_interrupts
>>>>      204536           +49.0%     304726 ±  2%
>>>> interrupts.CPU14.LOC:Local_timer_interrupts
>>>>        4937 ± 25%     +41.6%       6991 ±  2%
>>>> interrupts.CPU14.NMI:Non-maskable_interrupts
>>>>        4937 ± 25%     +41.6%       6991 ±  2%
>>>> interrupts.CPU14.PMI:Performance_monitoring_interrupts
>>>>      287759           +28.2%     368986 ±  3%
>>>> interrupts.CPU14.RES:Rescheduling_interrupts
>>>>      204527           +49.0%     304707 ±  2%
>>>> interrupts.CPU15.LOC:Local_timer_interrupts
>>>>      286950           +28.6%     369104 ±  3%
>>>> interrupts.CPU15.RES:Rescheduling_interrupts
>>>>      204556           +49.0%     304718 ±  2%
>>>> interrupts.CPU16.LOC:Local_timer_interrupts
>>>>        5444 ± 32%     -19.9%       4360 ± 32%
>>>> interrupts.CPU16.NMI:Non-maskable_interrupts
>>>>        5444 ± 32%     -19.9%       4360 ± 32%
>>>> interrupts.CPU16.PMI:Performance_monitoring_interrupts
>>>>      287623           +28.1%     368583 ±  3%
>>>> interrupts.CPU16.RES:Rescheduling_interrupts
>>>>      204528           +49.0%     304728 ±  2%
>>>> interrupts.CPU17.LOC:Local_timer_interrupts
>>>>      287734           +28.2%     368920 ±  3%
>>>> interrupts.CPU17.RES:Rescheduling_interrupts
>>>>      204465           +49.0%     304707 ±  2%
>>>> interrupts.CPU18.LOC:Local_timer_interrupts
>>>>        5382 ± 31%     -21.8%       4211 ± 27%
>>>> interrupts.CPU18.NMI:Non-maskable_interrupts
>>>>        5382 ± 31%     -21.8%       4211 ± 27%
>>>> interrupts.CPU18.PMI:Performance_monitoring_interrupts
>>>>      288778           +28.1%     369871 ±  3%
>>>> interrupts.CPU18.RES:Rescheduling_interrupts
>>>>      204302           +49.2%     304727 ±  2%
>>>> interrupts.CPU19.LOC:Local_timer_interrupts
>>>>      289794           +27.8%     370247 ±  3%
>>>> interrupts.CPU19.RES:Rescheduling_interrupts
>>>>      204552           +49.0%     304821 ±  2%
>>>> interrupts.CPU2.LOC:Local_timer_interrupts
>>>>      291056 ±  2%     +28.3%     373290 ±  4%
>>>> interrupts.CPU2.RES:Rescheduling_interrupts
>>>>      204468           +48.9%     304484 ±  2%
>>>> interrupts.CPU20.LOC:Local_timer_interrupts
>>>>      290778           +27.6%     371106 ±  3%
>>>> interrupts.CPU20.RES:Rescheduling_interrupts
>>>>      204546           +48.9%     304619 ±  2%
>>>> interrupts.CPU21.LOC:Local_timer_interrupts
>>>>      289974           +28.0%     371074 ±  3%
>>>> interrupts.CPU21.RES:Rescheduling_interrupts
>>>>      204694           +48.8%     304581 ±  2%
>>>> interrupts.CPU22.LOC:Local_timer_interrupts
>>>>      290312           +28.0%     371734 ±  3%
>>>> interrupts.CPU22.RES:Rescheduling_interrupts
>>>>      204374           +49.0%     304493 ±  2%
>>>> interrupts.CPU23.LOC:Local_timer_interrupts
>>>>      290266           +28.5%     372865 ±  3%
>>>> interrupts.CPU23.RES:Rescheduling_interrupts
>>>>      204518           +49.0%     304653 ±  2%
>>>> interrupts.CPU24.LOC:Local_timer_interrupts
>>>>      288815 ±  2%     +28.7%     371621 ±  4%
>>>> interrupts.CPU24.RES:Rescheduling_interrupts
>>>>      204414           +49.0%     304640 ±  2%
>>>> interrupts.CPU25.LOC:Local_timer_interrupts
>>>>      288779 ±  2%     +28.2%     370202 ±  4%
>>>> interrupts.CPU25.RES:Rescheduling_interrupts
>>>>      204536           +49.0%     304730 ±  2%
>>>> interrupts.CPU26.LOC:Local_timer_interrupts
>>>>      287615 ±  2%     +28.6%     369843 ±  4%
>>>> interrupts.CPU26.RES:Rescheduling_interrupts
>>>>       72.50 ± 37%     +55.5%     112.75 ± 26%
>>>> interrupts.CPU27.36:PCI-MSI.2621442-edge.eth0-TxRx-1
>>>>      204514           +49.0%     304668 ±  2%
>>>> interrupts.CPU27.LOC:Local_timer_interrupts
>>>>      287014 ±  2%     +28.5%     368825 ±  4%
>>>> interrupts.CPU27.RES:Rescheduling_interrupts
>>>>       53.25 ±  2%     +84.5%      98.25 ± 22%
>>>> interrupts.CPU28.37:PCI-MSI.2621443-edge.eth0-TxRx-2
>>>>      204331           +49.1%     304759 ±  2%
>>>> interrupts.CPU28.LOC:Local_timer_interrupts
>>>>      287278 ±  2%     +28.4%     368876 ±  4%
>>>> interrupts.CPU28.RES:Rescheduling_interrupts
>>>>       73.50 ± 26%    +116.3%     159.00 ± 60%
>>>> interrupts.CPU29.38:PCI-MSI.2621444-edge.eth0-TxRx-3
>>>>      204427           +49.0%     304651 ±  2%
>>>> interrupts.CPU29.LOC:Local_timer_interrupts
>>>>      287382 ±  2%     +28.9%     370434 ±  4%
>>>> interrupts.CPU29.RES:Rescheduling_interrupts
>>>>      204538           +49.0%     304765 ±  2%
>>>> interrupts.CPU3.LOC:Local_timer_interrupts
>>>>        5034 ± 27%     -29.4%       3552
>>>> interrupts.CPU3.NMI:Non-maskable_interrupts
>>>>        5034 ± 27%     -29.4%       3552
>>>> interrupts.CPU3.PMI:Performance_monitoring_interrupts
>>>>      288881 ±  2%     +28.3%     370740 ±  4%
>>>> interrupts.CPU3.RES:Rescheduling_interrupts
>>>>      204750           +48.8%     304674 ±  2%
>>>> interrupts.CPU30.LOC:Local_timer_interrupts
>>>>      287139 ±  2%     +28.8%     369825 ±  4%
>>>> interrupts.CPU30.RES:Rescheduling_interrupts
>>>>      204429           +48.9%     304416 ±  2%
>>>> interrupts.CPU31.LOC:Local_timer_interrupts
>>>>      288411 ±  2%     +28.4%     370199 ±  4%
>>>> interrupts.CPU31.RES:Rescheduling_interrupts
>>>>      204395           +49.3%     305106 ±  2%
>>>> interrupts.CPU32.LOC:Local_timer_interrupts
>>>>      289178 ±  2%     +28.6%     371830 ±  4%
>>>> interrupts.CPU32.RES:Rescheduling_interrupts
>>>>      204347           +49.2%     304799 ±  2%
>>>> interrupts.CPU33.LOC:Local_timer_interrupts
>>>>      289689 ±  2%     +28.2%     371280 ±  4%
>>>> interrupts.CPU33.RES:Rescheduling_interrupts
>>>>      204280           +49.1%     304566 ±  2%
>>>> interrupts.CPU34.LOC:Local_timer_interrupts
>>>>        6418 ± 24%     -44.5%       3564
>>>> interrupts.CPU34.NMI:Non-maskable_interrupts
>>>>        6418 ± 24%     -44.5%       3564
>>>> interrupts.CPU34.PMI:Performance_monitoring_interrupts
>>>>      290382 ±  2%     +28.3%     372606 ±  4%
>>>> interrupts.CPU34.RES:Rescheduling_interrupts
>>>>      204331           +49.0%     304517 ±  2%
>>>> interrupts.CPU35.LOC:Local_timer_interrupts
>>>>      289736 ±  2%     +28.3%     371869 ±  4%
>>>> interrupts.CPU35.RES:Rescheduling_interrupts
>>>>      204578           +48.8%     304423 ±  2%
>>>> interrupts.CPU36.LOC:Local_timer_interrupts
>>>>        7298           -39.0%       4450 ± 34%
>>>> interrupts.CPU36.NMI:Non-maskable_interrupts
>>>>        7298           -39.0%       4450 ± 34%
>>>> interrupts.CPU36.PMI:Performance_monitoring_interrupts
>>>>      287187           +28.3%     368516 ±  3%
>>>> interrupts.CPU36.RES:Rescheduling_interrupts
>>>>      204538           +49.0%     304729 ±  2%
>>>> interrupts.CPU37.LOC:Local_timer_interrupts
>>>>      288676           +28.1%     369792 ±  3%
>>>> interrupts.CPU37.RES:Rescheduling_interrupts
>>>>      204504           +49.0%     304687 ±  2%
>>>> interrupts.CPU38.LOC:Local_timer_interrupts
>>>>        6301 ± 24%     -29.8%       4424 ± 34%
>>>> interrupts.CPU38.NMI:Non-maskable_interrupts
>>>>        6301 ± 24%     -29.8%       4424 ± 34%
>>>> interrupts.CPU38.PMI:Performance_monitoring_interrupts
>>>>      287510           +28.1%     368349 ±  3%
>>>> interrupts.CPU38.RES:Rescheduling_interrupts
>>>>      204566           +48.9%     304668 ±  2%
>>>> interrupts.CPU39.LOC:Local_timer_interrupts
>>>>      286909           +28.1%     367622 ±  3%
>>>> interrupts.CPU39.RES:Rescheduling_interrupts
>>>>      204445           +49.0%     304686 ±  2%
>>>> interrupts.CPU4.LOC:Local_timer_interrupts
>>>>      288608 ±  2%     +28.7%     371493 ±  3%
>>>> interrupts.CPU4.RES:Rescheduling_interrupts
>>>>      204538           +48.9%     304582 ±  2%
>>>> interrupts.CPU40.LOC:Local_timer_interrupts
>>>>      286219           +27.9%     365983 ±  3%
>>>> interrupts.CPU40.RES:Rescheduling_interrupts
>>>>      204551           +49.0%     304739 ±  2%
>>>> interrupts.CPU41.LOC:Local_timer_interrupts
>>>>        5170 ± 29%     -18.5%       4215 ± 27%
>>>> interrupts.CPU41.NMI:Non-maskable_interrupts
>>>>        5170 ± 29%     -18.5%       4215 ± 27%
>>>> interrupts.CPU41.PMI:Performance_monitoring_interrupts
>>>>      286946           +27.8%     366717 ±  3%
>>>> interrupts.CPU41.RES:Rescheduling_interrupts
>>>>      204492           +48.9%     304442 ±  2%
>>>> interrupts.CPU42.LOC:Local_timer_interrupts
>>>>      287384           +28.0%     367805 ±  3%
>>>> interrupts.CPU42.RES:Rescheduling_interrupts
>>>>      204545           +49.0%     304679 ±  2%
>>>> interrupts.CPU43.LOC:Local_timer_interrupts
>>>>      289100           +27.3%     368059 ±  3%
>>>> interrupts.CPU43.RES:Rescheduling_interrupts
>>>>      204489           +49.0%     304635 ±  2%
>>>> interrupts.CPU44.LOC:Local_timer_interrupts
>>>>        6393 ± 24%     -32.4%       4319 ± 31%
>>>> interrupts.CPU44.NMI:Non-maskable_interrupts
>>>>        6393 ± 24%     -32.4%       4319 ± 31%
>>>> interrupts.CPU44.PMI:Performance_monitoring_interrupts
>>>>      288835           +27.8%     369182 ±  3%
>>>> interrupts.CPU44.RES:Rescheduling_interrupts
>>>>      204473           +49.2%     304983 ±  2%
>>>> interrupts.CPU45.LOC:Local_timer_interrupts
>>>>      289778           +27.4%     369059 ±  3%
>>>> interrupts.CPU45.RES:Rescheduling_interrupts
>>>>      204493           +49.0%     304598 ±  2%
>>>> interrupts.CPU46.LOC:Local_timer_interrupts
>>>>      289485           +27.9%     370172 ±  3%
>>>> interrupts.CPU46.RES:Rescheduling_interrupts
>>>>      204511           +49.2%     305090 ±  2%
>>>> interrupts.CPU47.LOC:Local_timer_interrupts
>>>>      286611           +28.8%     369021 ±  3%
>>>> interrupts.CPU47.RES:Rescheduling_interrupts
>>>>      204441           +49.1%     304794 ±  2%
>>>> interrupts.CPU5.LOC:Local_timer_interrupts
>>>>      289245 ±  2%     +28.4%     371335 ±  4%
>>>> interrupts.CPU5.RES:Rescheduling_interrupts
>>>>      204246           +49.0%     304414 ±  2%
>>>> interrupts.CPU6.LOC:Local_timer_interrupts
>>>>      290006 ±  2%     +28.3%     372180 ±  4%
>>>> interrupts.CPU6.RES:Rescheduling_interrupts
>>>>      204445           +49.0%     304719 ±  2%
>>>> interrupts.CPU7.LOC:Local_timer_interrupts
>>>>      290524 ±  2%     +28.2%     372414 ±  4%
>>>> interrupts.CPU7.RES:Rescheduling_interrupts
>>>>      204178           +49.2%     304663 ±  2%
>>>> interrupts.CPU8.LOC:Local_timer_interrupts
>>>>      291625 ±  2%     +27.8%     372747 ±  4%
>>>> interrupts.CPU8.RES:Rescheduling_interrupts
>>>>      204502           +48.9%     304407 ±  3%
>>>> interrupts.CPU9.LOC:Local_timer_interrupts
>>>>      291615 ±  2%     +27.9%     373024 ±  4%
>>>> interrupts.CPU9.RES:Rescheduling_interrupts
>>>>     9814627           +49.0%   14624016 ±  2%
>>>> interrupts.LOC:Local_timer_interrupts
>>>>    13869791           +28.2%   17785831 ±  3%
>>>> interrupts.RES:Rescheduling_interrupts
>>>>       25728           +47.6%      37969        softirqs.CPU0.RCU
>>>>       34944           +41.7%      49503 ±  3%  softirqs.CPU0.SCHED
>>>>       42375           +38.5%      58673 ±  2%  softirqs.CPU0.TIMER
>>>>       25348           +53.8%      38976 ±  5%  softirqs.CPU1.RCU
>>>>       30986           +46.2%      45315 ±  3%  softirqs.CPU1.SCHED
>>>>       41265           +44.3%      59538 ±  3%  softirqs.CPU1.TIMER
>>>>       25149           +48.3%      37288        softirqs.CPU10.RCU
>>>>       29752           +47.1%      43764 ±  2%  softirqs.CPU10.SCHED
>>>>       40609           +41.5%      57449        softirqs.CPU10.TIMER
>>>>       25218           +48.3%      37399        softirqs.CPU11.RCU
>>>>       30294           +44.6%      43800        softirqs.CPU11.SCHED
>>>>       41371           +39.2%      57578        softirqs.CPU11.TIMER
>>>>       24427           +48.9%      36361        softirqs.CPU12.RCU
>>>>       30438           +44.0%      43835        softirqs.CPU12.SCHED
>>>>       41045 ±  3%     +43.0%      58687        softirqs.CPU12.TIMER
>>>>       24247           +47.0%      35645        softirqs.CPU13.RCU
>>>>       30155           +44.5%      43577        softirqs.CPU13.SCHED
>>>>       41148 ±  2%     +40.3%      57720        softirqs.CPU13.TIMER
>>>>       24506           +46.3%      35856        softirqs.CPU14.RCU
>>>>       29991           +44.1%      43229        softirqs.CPU14.SCHED
>>>>       41190 ±  2%     +39.7%      57552        softirqs.CPU14.TIMER
>>>>       24412           +46.7%      35800        softirqs.CPU15.RCU
>>>>       30135           +44.1%      43430        softirqs.CPU15.SCHED
>>>>       45081 ± 15%     +28.0%      57684        softirqs.CPU15.TIMER
>>>>       25327           +47.0%      37225        softirqs.CPU16.RCU
>>>>       30230           +43.8%      43461        softirqs.CPU16.SCHED
>>>>       41157           +39.6%      57448        softirqs.CPU16.TIMER
>>>>       25136           +47.8%      37142        softirqs.CPU17.RCU
>>>>       29894           +44.5%      43212        softirqs.CPU17.SCHED
>>>>       41961 ±  5%     +36.5%      57281        softirqs.CPU17.TIMER
>>>>       25431           +46.1%      37160        softirqs.CPU18.RCU
>>>>       30062           +44.0%      43281        softirqs.CPU18.SCHED
>>>>       40956 ±  2%     +42.1%      58217        softirqs.CPU18.TIMER
>>>>       25496           +44.8%      36923 ±  2%  softirqs.CPU19.RCU
>>>>       29989           +46.5%      43946        softirqs.CPU19.SCHED
>>>>       41740 ±  2%     +67.8%      70057 ± 17%  softirqs.CPU19.TIMER
>>>>       25414           +48.3%      37678        softirqs.CPU2.RCU
>>>>       30905           +43.4%      44309 ±  2%  softirqs.CPU2.SCHED
>>>>       41083           +40.6%      57768 ±  2%  softirqs.CPU2.TIMER
>>>>       25124           +46.8%      36892        softirqs.CPU20.RCU
>>>>       30001           +44.7%      43416        softirqs.CPU20.SCHED
>>>>       41060 ±  7%     +39.2%      57151        softirqs.CPU20.TIMER
>>>>       24992           +47.6%      36896        softirqs.CPU21.RCU
>>>>       30109           +44.4%      43467        softirqs.CPU21.SCHED
>>>>       40055 ±  3%     +42.5%      57081 ±  2%  softirqs.CPU21.TIMER
>>>>       25072           +46.2%      36656        softirqs.CPU22.RCU
>>>>       30296           +43.3%      43425        softirqs.CPU22.SCHED
>>>>       24882           +48.8%      37031        softirqs.CPU23.RCU
>>>>       29900           +43.7%      42971        softirqs.CPU23.SCHED
>>>>       40388 ±  3%     +44.8%      58494 ±  3%  softirqs.CPU23.TIMER
>>>>       25007           +48.1%      37036        softirqs.CPU24.RCU
>>>>       29755           +45.9%      43399        softirqs.CPU24.SCHED
>>>>       40786           +40.6%      57353        softirqs.CPU24.TIMER
>>>>       25031           +47.3%      36875        softirqs.CPU25.RCU
>>>>       29841 ±  2%     +45.3%      43355        softirqs.CPU25.SCHED
>>>>       40971           +39.8%      57266        softirqs.CPU25.TIMER
>>>>       25145           +47.6%      37114        softirqs.CPU26.RCU
>>>>       29462 ±  2%     +48.5%      43739 ±  2%  softirqs.CPU26.SCHED
>>>>       40553           +41.8%      57521        softirqs.CPU26.TIMER
>>>>       25184           +47.6%      37169        softirqs.CPU27.RCU
>>>>       29867           +45.5%      43450 ±  3%  softirqs.CPU27.SCHED
>>>>       41118           +39.3%      57261        softirqs.CPU27.TIMER
>>>>       25452           +47.1%      37440        softirqs.CPU28.RCU
>>>>       29714           +46.2%      43440 ±  2%  softirqs.CPU28.SCHED
>>>>       41209 ±  2%     +37.8%      56783 ±  2%  softirqs.CPU28.TIMER
>>>>       25334           +48.8%      37696 ±  2%  softirqs.CPU29.RCU
>>>>       29894           +46.3%      43749        softirqs.CPU29.SCHED
>>>>       41413 ±  2%     +43.3%      59346 ±  5%  softirqs.CPU29.TIMER
>>>>       25514           +47.0%      37495        softirqs.CPU3.RCU
>>>>       30016           +46.0%      43831 ±  2%  softirqs.CPU3.SCHED
>>>>       41371 ±  2%     +39.3%      57611        softirqs.CPU3.TIMER
>>>>       25206           +47.4%      37149        softirqs.CPU30.RCU
>>>>       30167           +44.0%      43451 ±  2%  softirqs.CPU30.SCHED
>>>>       41166 ±  2%     +58.1%      65092 ± 19%  softirqs.CPU30.TIMER
>>>>       25211           +46.8%      37016        softirqs.CPU31.RCU
>>>>       30004 ±  2%     +45.2%      43569        softirqs.CPU31.SCHED
>>>>       41541 ±  2%     +36.9%      56874        softirqs.CPU31.TIMER
>>>>       24432           +46.9%      35885        softirqs.CPU32.RCU
>>>>       30206           +43.4%      43322 ±  2%  softirqs.CPU32.SCHED
>>>>       41069           +40.7%      57781 ±  2%  softirqs.CPU32.TIMER
>>>>       24303           +48.0%      35973        softirqs.CPU33.RCU
>>>>       30238           +43.8%      43489        softirqs.CPU33.SCHED
>>>>       41161           +38.2%      56872        softirqs.CPU33.TIMER
>>>>       24308           +47.6%      35876        softirqs.CPU34.RCU
>>>>       29803           +46.5%      43647 ±  2%  softirqs.CPU34.SCHED
>>>>       40809           +39.2%      56817        softirqs.CPU34.TIMER
>>>>       24291           +47.7%      35877        softirqs.CPU35.RCU
>>>>       30092           +44.2%      43386        softirqs.CPU35.SCHED
>>>>       40607           +41.5%      57480 ±  2%  softirqs.CPU35.TIMER
>>>>       25605           +49.5%      38286        softirqs.CPU36.RCU
>>>>       29900           +44.6%      43238        softirqs.CPU36.SCHED
>>>>       45207 ± 15%     +27.4%      57602        softirqs.CPU36.TIMER
>>>>       25555 ±  2%     +49.0%      38070        softirqs.CPU37.RCU
>>>>       29926           +45.2%      43460        softirqs.CPU37.SCHED
>>>>       42619 ±  8%     +36.7%      58242        softirqs.CPU37.TIMER
>>>>       27683 ±  9%     +50.2%      41577 ±  4%  softirqs.CPU38.RCU
>>>>       29700           +45.6%      43230        softirqs.CPU38.SCHED
>>>>       41779 ±  4%     +43.5%      59948        softirqs.CPU38.TIMER
>>>>       26699 ±  7%     +42.1%      37951        softirqs.CPU39.RCU
>>>>       30061           +43.4%      43102        softirqs.CPU39.SCHED
>>>>       42312 ±  5%     +36.1%      57581 ±  2%  softirqs.CPU39.TIMER
>>>>       25219           +47.7%      37247        softirqs.CPU4.RCU
>>>>       30363 ±  3%     +45.9%      44295        softirqs.CPU4.SCHED
>>>>       41108 ±  2%     +39.6%      57371 ±  2%  softirqs.CPU4.TIMER
>>>>       25413           +50.6%      38268        softirqs.CPU40.RCU
>>>>       29910           +46.0%      43665        softirqs.CPU40.SCHED
>>>>       39987 ±  2%     +45.9%      58322        softirqs.CPU40.TIMER
>>>>       25569           +49.5%      38238        softirqs.CPU41.RCU
>>>>       29867           +44.8%      43247        softirqs.CPU41.SCHED
>>>>       25492           +49.8%      38185        softirqs.CPU42.RCU
>>>>       29871           +45.4%      43440        softirqs.CPU42.SCHED
>>>>       40076           +44.8%      58020        softirqs.CPU42.TIMER
>>>>       25281           +48.9%      37640        softirqs.CPU43.RCU
>>>>       30025           +43.9%      43220        softirqs.CPU43.SCHED
>>>>       40222 ±  3%     +44.7%      58195        softirqs.CPU43.TIMER
>>>>       25183           +49.3%      37611        softirqs.CPU44.RCU
>>>>       30123           +43.6%      43249        softirqs.CPU44.SCHED
>>>>       41090 ±  6%     +42.5%      58537        softirqs.CPU44.TIMER
>>>>       25314           +49.6%      37863        softirqs.CPU45.RCU
>>>>       29998           +43.2%      42957        softirqs.CPU45.SCHED
>>>>       39705 ±  3%     +43.6%      57033        softirqs.CPU45.TIMER
>>>>       25500           +46.2%      37274        softirqs.CPU46.RCU
>>>>       29921           +44.7%      43286        softirqs.CPU46.SCHED
>>>>       41088 ±  5%     +39.9%      57463 ±  3%  softirqs.CPU46.TIMER
>>>>       25564           +48.4%      37931        softirqs.CPU47.RCU
>>>>       30178           +42.4%      42980        softirqs.CPU47.SCHED
>>>>       40738 ±  4%     +42.2%      57939 ±  3%  softirqs.CPU47.TIMER
>>>>       26467 ±  8%     +40.8%      37278        softirqs.CPU5.RCU
>>>>       30334 ±  2%     +44.1%      43714        softirqs.CPU5.SCHED
>>>>       42268 ±  4%     +35.8%      57394        softirqs.CPU5.TIMER
>>>>       25343           +47.0%      37245        softirqs.CPU6.RCU
>>>>       30100           +45.0%      43643 ±  2%  softirqs.CPU6.SCHED
>>>>       41270           +43.6%      59249 ±  5%  softirqs.CPU6.TIMER
>>>>       25331           +47.7%      37418        softirqs.CPU7.RCU
>>>>       30210           +44.7%      43727 ±  2%  softirqs.CPU7.SCHED
>>>>       41028           +41.9%      58206 ±  2%  softirqs.CPU7.TIMER
>>>>       25483           +47.2%      37514        softirqs.CPU8.RCU
>>>>       30324 ±  2%     +43.4%      43483 ±  2%  softirqs.CPU8.SCHED
>>>>       41027           +40.4%      57591 ±  2%  softirqs.CPU8.TIMER
>>>>       25546           +47.5%      37687        softirqs.CPU9.RCU
>>>>       30218           +44.8%      43771 ±  2%  softirqs.CPU9.SCHED
>>>>       41193           +40.1%      57716 ±  2%  softirqs.CPU9.TIMER
>>>>     1212606           +47.7%    1790808        softirqs.RCU
>>>>     1448198           +44.7%    2095500        softirqs.SCHED
>>>>     1990064           +40.7%    2800202        softirqs.TIMER
>>>>        8.56 ±  3%      -4.3        4.25 perf-profile.calltrace.cycles-pp.__filemap_fdatawrite_range.file_write_and_wait_range.xfs_file_fsync.xfs_file_buffered_aio_write.new_sync_write
>>>>
>>>>        8.50 ±  3%      -4.3        4.20 perf-profile.calltrace.cycles-pp.do_writepages.__filemap_fdatawrite_range.file_write_and_wait_range.xfs_file_fsync.xfs_file_buffered_aio_write
>>>>
>>>>        8.47 ±  3%      -4.3        4.17 perf-profile.calltrace.cycles-pp.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.file_write_and_wait_range.xfs_file_fsync
>>>>
>>>>       78.55            -4.1       74.44
>>>> perf-profile.calltrace.cycles-pp.write
>>>>        8.74 ±  3%      -4.0        4.70 perf-profile.calltrace.cycles-pp.file_write_and_wait_range.xfs_file_fsync.xfs_file_buffered_aio_write.new_sync_write.vfs_write
>>>>
>>>>       77.69            -3.9       73.75
>>>> perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write
>>>>       77.64            -3.9       73.71 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
>>>>
>>>>       77.13            -3.8       73.36 perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
>>>>
>>>>       77.08            -3.8       73.31 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
>>>>
>>>>       76.61            -3.7       72.96 perf-profile.calltrace.cycles-pp.new_sync_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
>>>>
>>>>       76.43            -3.6       72.80 perf-profile.calltrace.cycles-pp.xfs_file_buffered_aio_write.new_sync_write.vfs_write.ksys_write.do_syscall_64
>>>>
>>>>       69.22            -2.6       66.62 perf-profile.calltrace.cycles-pp.xfs_file_fsync.xfs_file_buffered_aio_write.new_sync_write.vfs_write.ksys_write
>>>>
>>>>        4.70 ±  2%      -2.2        2.50 perf-profile.calltrace.cycles-pp.iomap_submit_ioend.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.file_write_and_wait_range
>>>>
>>>>        4.63 ±  2%      -2.2        2.45 perf-profile.calltrace.cycles-pp.submit_bio.iomap_submit_ioend.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range
>>>>
>>>>        4.60 ±  2%      -2.2        2.43 perf-profile.calltrace.cycles-pp.generic_make_request.submit_bio.iomap_submit_ioend.xfs_vm_writepages.do_writepages
>>>>
>>>>        4.32 ±  2%      -2.1        2.22 perf-profile.calltrace.cycles-pp.brd_make_request.generic_make_request.submit_bio.iomap_submit_ioend.xfs_vm_writepages
>>>>
>>>>        3.72 ±  4%      -2.1        1.63 perf-profile.calltrace.cycles-pp.iomap_writepages.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.file_write_and_wait_range
>>>>
>>>>        3.71 ±  4%      -2.1        1.62 perf-profile.calltrace.cycles-pp.write_cache_pages.iomap_writepages.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range
>>>>
>>>>        2.63 ±  4%      -2.0        0.63 ±  3% perf-profile.calltrace.cycles-pp.iomap_finish_ioend.brd_make_request.generic_make_request.submit_bio.iomap_submit_ioend
>>>>
>>>>        7.72            -2.0        5.74 ±  2% perf-profile.calltrace.cycles-pp.__flush_work.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync.xfs_file_buffered_aio_write
>>>>
>>>>        3.08 ±  5%      -1.9        1.13 perf-profile.calltrace.cycles-pp.iomap_writepage_map.write_cache_pages.iomap_writepages.xfs_vm_writepages.do_writepages
>>>>
>>>>        2.49 ±  5%      -1.9        0.55 ±  3% perf-profile.calltrace.cycles-pp.end_page_writeback.iomap_finish_ioend.brd_make_request.generic_make_request.submit_bio
>>>>
>>>>        2.48 ±  5%      -1.9        0.55 ±  3% perf-profile.calltrace.cycles-pp.test_clear_page_writeback.end_page_writeback.iomap_finish_ioend.brd_make_request.generic_make_request
>>>>
>>>>        2.41 ±  6%      -1.8        0.61 ±  2% perf-profile.calltrace.cycles-pp.__test_set_page_writeback.iomap_writepage_map.write_cache_pages.iomap_writepages.xfs_vm_writepages
>>>>
>>>>        4.90            -1.5        3.40 ±  3% perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.__flush_work.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync
>>>>
>>>>        4.78            -1.5        3.30 ±  3% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irq.__flush_work.xlog_cil_force_lsn.xfs_log_force_lsn
>>>>
>>>>        2.35            -0.6        1.77 ±  2% perf-profile.calltrace.cycles-pp.iomap_apply.iomap_file_buffered_write.xfs_file_buffered_aio_write.new_sync_write.vfs_write
>>>>
>>>>        2.36            -0.6        1.79 perf-profile.calltrace.cycles-pp.iomap_file_buffered_write.xfs_file_buffered_aio_write.new_sync_write.vfs_write.ksys_write
>>>>
>>>>        3.40            -0.5        2.87 ±  6% perf-profile.calltrace.cycles-pp.__xfs_trans_commit.xfs_vn_update_time.file_update_time.xfs_file_aio_write_checks.xfs_file_buffered_aio_write
>>>>
>>>>        1.83            -0.5        1.33 ±  2% perf-profile.calltrace.cycles-pp.iomap_write_actor.iomap_apply.iomap_file_buffered_write.xfs_file_buffered_aio_write.new_sync_write
>>>>
>>>>        3.26            -0.5        2.75 ±  6% perf-profile.calltrace.cycles-pp.xfs_log_commit_cil.__xfs_trans_commit.xfs_vn_update_time.file_update_time.xfs_file_aio_write_checks
>>>>
>>>>        2.85            -0.4        2.40 perf-profile.calltrace.cycles-pp.complete.process_one_work.worker_thread.kthread.ret_from_fork
>>>>
>>>>        2.79            -0.4        2.34 perf-profile.calltrace.cycles-pp.try_to_wake_up.swake_up_locked.complete.process_one_work.worker_thread
>>>>
>>>>        2.83            -0.4        2.38 perf-profile.calltrace.cycles-pp.swake_up_locked.complete.process_one_work.worker_thread.kthread
>>>>
>>>>        1.48            -0.4        1.05 ±  2% perf-profile.calltrace.cycles-pp.brd_do_bvec.brd_make_request.generic_make_request.submit_bio.iomap_submit_ioend
>>>>
>>>>        0.61 ±  2%      -0.4        0.26 ±100% perf-profile.calltrace.cycles-pp.sched_ttwu_pending.scheduler_ipi.reschedule_interrupt.native_queued_spin_lock_slowpath._raw_spin_lock
>>>>
>>>>        1.89            -0.3        1.54 perf-profile.calltrace.cycles-pp.wait_for_completion.__flush_work.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync
>>>>
>>>>        1.71            -0.3        1.38 perf-profile.calltrace.cycles-pp.schedule_timeout.wait_for_completion.__flush_work.xlog_cil_force_lsn.xfs_log_force_lsn
>>>>
>>>>        1.65            -0.3        1.34 perf-profile.calltrace.cycles-pp.schedule.schedule_timeout.wait_for_completion.__flush_work.xlog_cil_force_lsn
>>>>
>>>>        0.69            -0.3        0.38 ± 57% perf-profile.calltrace.cycles-pp.iomap_write_begin.iomap_write_actor.iomap_apply.iomap_file_buffered_write.xfs_file_buffered_aio_write
>>>>
>>>>        1.59            -0.3        1.29 perf-profile.calltrace.cycles-pp.__sched_text_start.schedule.schedule_timeout.wait_for_completion.__flush_work
>>>>
>>>>        0.86            -0.3        0.58 ±  2% perf-profile.calltrace.cycles-pp.rwsem_down_read_slowpath.xfs_log_commit_cil.__xfs_trans_commit.xfs_vn_update_time.file_update_time
>>>>
>>>>        1.52            -0.3        1.26 ±  3% perf-profile.calltrace.cycles-pp.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit.xfs_vn_update_time.file_update_time
>>>>
>>>>        1.75            -0.2        1.50 ±  2% perf-profile.calltrace.cycles-pp.ttwu_do_activate.try_to_wake_up.swake_up_locked.complete.process_one_work
>>>>
>>>>        1.74            -0.2        1.49 ±  2% perf-profile.calltrace.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up.swake_up_locked
>>>>
>>>>        1.74            -0.2        1.50 ±  2% perf-profile.calltrace.cycles-pp.activate_task.ttwu_do_activate.try_to_wake_up.swake_up_locked.complete
>>>>
>>>>        2.15            -0.2        1.94 perf-profile.calltrace.cycles-pp.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate
>>>>
>>>>        0.79 ±  3%      -0.2        0.60 ±  4% perf-profile.calltrace.cycles-pp._raw_spin_lock.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit.xfs_vn_update_time
>>>>
>>>>        1.81            -0.2        1.62 perf-profile.calltrace.cycles-pp.arch_stack_walk.stack_trace_save_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair
>>>>
>>>>        1.87            -0.2        1.68 perf-profile.calltrace.cycles-pp.stack_trace_save_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.activate_task
>>>>
>>>>        0.70 ±  4%      -0.2        0.54 ±  4% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.xlog_cil_insert_items.xfs_log_commit_cil.__xfs_trans_commit
>>>>
>>>>        0.81            -0.1        0.68 perf-profile.calltrace.cycles-pp.unwind_next_frame.arch_stack_walk.stack_trace_save_tsk.__account_scheduler_latency.enqueue_entity
>>>>
>>>>        0.65 ±  2%      -0.1        0.55 ±  2% perf-profile.calltrace.cycles-pp.reschedule_interrupt.native_queued_spin_lock_slowpath._raw_spin_lock.xlog_cil_force_lsn.xfs_log_force_lsn
>>>>
>>>>        0.62 ±  2%      -0.1        0.52 ±  3% perf-profile.calltrace.cycles-pp.scheduler_ipi.reschedule_interrupt.native_queued_spin_lock_slowpath._raw_spin_lock.xlog_cil_force_lsn
>>>>
>>>>        1.18            -0.0        1.16 perf-profile.calltrace.cycles-pp.sched_ttwu_pending.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
>>>>
>>>>        1.01            -0.0        0.99 perf-profile.calltrace.cycles-pp.activate_task.ttwu_do_activate.sched_ttwu_pending.do_idle.cpu_startup_entry
>>>>
>>>>        1.00            -0.0        0.98 perf-profile.calltrace.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.sched_ttwu_pending.do_idle
>>>>
>>>>        0.96            -0.0        0.93 perf-profile.calltrace.cycles-pp.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate.sched_ttwu_pending
>>>>
>>>>        1.01            -0.0        0.99 perf-profile.calltrace.cycles-pp.ttwu_do_activate.sched_ttwu_pending.do_idle.cpu_startup_entry.start_secondary
>>>>
>>>>        0.59 ±  3%      +0.1        0.68 perf-profile.calltrace.cycles-pp.schedule_idle.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
>>>>
>>>>        0.58 ±  3%      +0.1        0.67 perf-profile.calltrace.cycles-pp.__sched_text_start.schedule_idle.do_idle.cpu_startup_entry.start_secondary
>>>>
>>>>        0.53            +0.1        0.63 ±  3% perf-profile.calltrace.cycles-pp.try_to_wake_up.__wake_up_common.__wake_up_common_lock.xlog_cil_push_work.process_one_work
>>>>
>>>>        0.91 ±  3%      +0.1        1.02 ±  2% perf-profile.calltrace.cycles-pp.xlog_write.xlog_cil_push_work.process_one_work.worker_thread.kthread
>>>>
>>>>        0.54            +0.1        0.65 ±  3% perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_common_lock.xlog_cil_push_work.process_one_work.worker_thread
>>>>
>>>>        0.57            +0.1        0.69 ±  3% perf-profile.calltrace.cycles-pp.__wake_up_common_lock.xlog_cil_push_work.process_one_work.worker_thread.kthread
>>>>
>>>>        1.83            +0.2        1.99 perf-profile.calltrace.cycles-pp.xlog_cil_push_work.process_one_work.worker_thread.kthread.ret_from_fork
>>>>
>>>>        0.00            +0.6        0.57 ±  2% perf-profile.calltrace.cycles-pp.activate_task.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common
>>>>
>>>>        0.00            +0.6        0.57 ±  2% perf-profile.calltrace.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up.autoremove_wake_function
>>>>
>>>>        0.00            +0.6        0.57 ±  2% perf-profile.calltrace.cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common.wake_up_page_bit
>>>>
>>>>        0.00            +0.7        0.70 ±  2% perf-profile.calltrace.cycles-pp.xfs_bmap_add_extent_unwritten_real.xfs_bmapi_convert_unwritten.xfs_bmapi_write.xfs_iomap_write_unwritten.xfs_end_ioend
>>>>
>>>>        0.00            +0.8        0.75 ±  4% perf-profile.calltrace.cycles-pp.xfs_bmapi_convert_unwritten.xfs_bmapi_write.xfs_iomap_write_unwritten.xfs_end_ioend.xfs_end_io
>>>>
>>>>        0.00            +0.8        0.80 perf-profile.calltrace.cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.wake_up_page_bit.iomap_finish_ioend
>>>>
>>>>        0.00            +0.8        0.80 perf-profile.calltrace.cycles-pp.autoremove_wake_function.__wake_up_common.wake_up_page_bit.iomap_finish_ioend.iomap_finish_ioends
>>>>
>>>>        0.00            +0.8        0.80 ± 30% perf-profile.calltrace.cycles-pp.xfs_trans_reserve.xfs_trans_alloc.xfs_vn_update_time.file_update_time.xfs_file_aio_write_checks
>>>>
>>>>        0.00            +0.8        0.82 perf-profile.calltrace.cycles-pp.__wake_up_common.wake_up_page_bit.iomap_finish_ioend.iomap_finish_ioends.xfs_end_ioend
>>>>
>>>>        0.00            +0.8        0.85 perf-profile.calltrace.cycles-pp.wake_up_page_bit.iomap_finish_ioend.iomap_finish_ioends.xfs_end_ioend.xfs_end_io
>>>>
>>>>        0.00            +0.9        0.89 ±  2% perf-profile.calltrace.cycles-pp.xfs_bmapi_write.xfs_iomap_write_unwritten.xfs_end_ioend.xfs_end_io.process_one_work
>>>>
>>>>        0.00            +0.9        0.92 ± 21% perf-profile.calltrace.cycles-pp.xfs_log_commit_cil.__xfs_trans_commit.xfs_iomap_write_unwritten.xfs_end_ioend.xfs_end_io
>>>>
>>>>        0.00            +1.0        0.97 ± 20% perf-profile.calltrace.cycles-pp.__xfs_trans_commit.xfs_iomap_write_unwritten.xfs_end_ioend.xfs_end_io.process_one_work
>>>>
>>>>       35.68            +1.0       36.68 perf-profile.calltrace.cycles-pp.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync.xfs_file_buffered_aio_write.new_sync_write
>>>>
>>>>        0.00            +1.0        1.04 perf-profile.calltrace.cycles-pp.iomap_finish_ioend.iomap_finish_ioends.xfs_end_ioend.xfs_end_io.process_one_work
>>>>
>>>>        0.00            +1.0        1.05 perf-profile.calltrace.cycles-pp.iomap_finish_ioends.xfs_end_ioend.xfs_end_io.process_one_work.worker_thread
>>>>
>>>>        9.35            +1.2       10.54 perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry
>>>>
>>>>        9.61            +1.2       10.85 perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary
>>>>
>>>>        9.63            +1.2       10.87 perf-profile.calltrace.cycles-pp.cpuidle_enter.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
>>>>
>>>>       12.24            +1.3       13.58 perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.secondary_startup_64
>>>>
>>>>       12.25            +1.3       13.60
>>>> perf-profile.calltrace.cycles-pp.start_secondary.secondary_startup_64
>>>>       12.25            +1.3       13.60 perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.secondary_startup_64
>>>>
>>>>       12.49            +1.4       13.89
>>>> perf-profile.calltrace.cycles-pp.secondary_startup_64
>>>>       14.38            +1.4       15.81 ±  2% perf-profile.calltrace.cycles-pp.xlog_wait_on_iclog.__xfs_log_force_lsn.xfs_log_force_lsn.xfs_file_fsync.xfs_file_buffered_aio_write
>>>>
>>>>       13.41            +1.4       14.86 ±  2% perf-profile.calltrace.cycles-pp.remove_wait_queue.xlog_wait_on_iclog.__xfs_log_force_lsn.xfs_log_force_lsn.xfs_file_fsync
>>>>
>>>>       13.37            +1.5       14.82 ±  2% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.remove_wait_queue.xlog_wait_on_iclog.__xfs_log_force_lsn.xfs_log_force_lsn
>>>>
>>>>       13.34            +1.5       14.80 ±  2% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.remove_wait_queue.xlog_wait_on_iclog.__xfs_log_force_lsn
>>>>
>>>>       58.91            +1.6       60.56 perf-profile.calltrace.cycles-pp.xfs_log_force_lsn.xfs_file_fsync.xfs_file_buffered_aio_write.new_sync_write.vfs_write
>>>>
>>>>        0.00            +2.2        2.24 ± 13% perf-profile.calltrace.cycles-pp.xfs_iomap_write_unwritten.xfs_end_ioend.xfs_end_io.process_one_work.worker_thread
>>>>
>>>>        5.63            +2.5        8.15 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.remove_wait_queue.xlog_cil_force_lsn.xfs_log_force_lsn
>>>>
>>>>        5.65            +2.5        8.19 perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.remove_wait_queue.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync
>>>>
>>>>        5.69            +2.5        8.23 perf-profile.calltrace.cycles-pp.remove_wait_queue.xlog_cil_force_lsn.xfs_log_force_lsn.xfs_file_fsync.xfs_file_buffered_aio_write
>>>>
>>>>        7.04            +2.8        9.80 ±  2% perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork
>>>>
>>>>        7.54            +2.9       10.45 ±  2%
>>>> perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork
>>>>        7.55            +2.9       10.46 ±  2%
>>>> perf-profile.calltrace.cycles-pp.ret_from_fork
>>>>        7.55            +2.9       10.46 ±  2%
>>>> perf-profile.calltrace.cycles-pp.kthread.ret_from_fork
>>>>        0.00            +3.3        3.30 ±  9% perf-profile.calltrace.cycles-pp.xfs_end_ioend.xfs_end_io.process_one_work.worker_thread.kthread
>>>>
>>>>        0.00            +3.3        3.32 ±  8% perf-profile.calltrace.cycles-pp.xfs_end_io.process_one_work.worker_thread.kthread.ret_from_fork
>>>>
>>>>        8.56 ±  3%      -4.3        4.25
>>>> perf-profile.children.cycles-pp.__filemap_fdatawrite_range
>>>>        8.47 ±  3%      -4.3        4.18
>>>> perf-profile.children.cycles-pp.xfs_vm_writepages
>>>>        8.50 ±  3%      -4.3        4.20
>>>> perf-profile.children.cycles-pp.do_writepages
>>>>       78.64            -4.1       74.51
>>>> perf-profile.children.cycles-pp.write
>>>>       78.55            -4.1       74.44
>>>> perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
>>>>       78.50            -4.1       74.41
>>>> perf-profile.children.cycles-pp.do_syscall_64
>>>>        8.75 ±  3%      -4.1        4.70
>>>> perf-profile.children.cycles-pp.file_write_and_wait_range
>>>>       77.14            -3.8       73.37
>>>> perf-profile.children.cycles-pp.ksys_write
>>>>       77.09            -3.8       73.33
>>>> perf-profile.children.cycles-pp.vfs_write
>>>>       76.62            -3.7       72.97
>>>> perf-profile.children.cycles-pp.new_sync_write
>>>>       76.44            -3.6       72.81
>>>> perf-profile.children.cycles-pp.xfs_file_buffered_aio_write
>>>>       69.23            -2.6       66.62
>>>> perf-profile.children.cycles-pp.xfs_file_fsync
>>>>        4.70 ±  2%      -2.2        2.50
>>>> perf-profile.children.cycles-pp.iomap_submit_ioend
>>>>        4.89 ±  2%      -2.1        2.75
>>>> perf-profile.children.cycles-pp.submit_bio
>>>>        4.85 ±  2%      -2.1        2.71
>>>> perf-profile.children.cycles-pp.generic_make_request
>>>>        3.73 ±  4%      -2.1        1.63
>>>> perf-profile.children.cycles-pp.iomap_writepages
>>>>        3.72 ±  4%      -2.1        1.63
>>>> perf-profile.children.cycles-pp.write_cache_pages
>>>>        4.43 ±  2%      -2.1        2.38
>>>> perf-profile.children.cycles-pp.brd_make_request
>>>>        7.73            -2.0        5.75 ±  2%
>>>> perf-profile.children.cycles-pp.__flush_work
>>>>        3.08 ±  4%      -1.9        1.14
>>>> perf-profile.children.cycles-pp.iomap_writepage_map
>>>>        6.00            -1.8        4.20 ±  2%
>>>> perf-profile.children.cycles-pp._raw_spin_lock_irq
>>>>        2.50 ±  4%      -1.8        0.70 ±  3%
>>>> perf-profile.children.cycles-pp.end_page_writeback
>>>>        2.50 ±  5%      -1.8        0.70 ±  4%
>>>> perf-profile.children.cycles-pp.test_clear_page_writeback
>>>>        2.42 ±  6%      -1.8        0.61 ±  2%
>>>> perf-profile.children.cycles-pp.__test_set_page_writeback
>>>>        2.08 ±  7%      -1.6        0.44 ±  4%
>>>> perf-profile.children.cycles-pp.sb_mark_inode_writeback
>>>>        2.08 ±  6%      -1.6        0.44 ±  5%
>>>> perf-profile.children.cycles-pp.sb_clear_inode_writeback
>>>>        2.70 ±  4%      -1.0        1.67 ±  2%
>>>> perf-profile.children.cycles-pp.iomap_finish_ioend
>>>>        2.36            -0.6        1.79
>>>> perf-profile.children.cycles-pp.iomap_file_buffered_write
>>>>        2.35            -0.6        1.78
>>>> perf-profile.children.cycles-pp.iomap_apply
>>>>        1.84            -0.5        1.33 ±  2%
>>>> perf-profile.children.cycles-pp.iomap_write_actor
>>>>        2.41 ±  4%      -0.5        1.94 ±  2%
>>>> perf-profile.children.cycles-pp.reschedule_interrupt
>>>>        3.44 ±  2%      -0.5        2.98
>>>> perf-profile.children.cycles-pp.sched_ttwu_pending
>>>>        2.87            -0.5        2.42
>>>> perf-profile.children.cycles-pp.complete
>>>>        2.83            -0.4        2.38
>>>> perf-profile.children.cycles-pp.swake_up_locked
>>>>        2.28 ±  4%      -0.4        1.84 ±  2%
>>>> perf-profile.children.cycles-pp.scheduler_ipi
>>>>        1.56            -0.4        1.17 ±  2%
>>>> perf-profile.children.cycles-pp.brd_do_bvec
>>>>        1.89            -0.4        1.54
>>>> perf-profile.children.cycles-pp.wait_for_completion
>>>>        1.71            -0.3        1.38
>>>> perf-profile.children.cycles-pp.schedule_timeout
>>>>        0.88            -0.2        0.68 ±  2%
>>>> perf-profile.children.cycles-pp.rwsem_down_read_slowpath
>>>>        0.69            -0.2        0.51 ±  2%
>>>> perf-profile.children.cycles-pp.iomap_write_begin
>>>>        0.67 ±  2%      -0.2        0.49
>>>> perf-profile.children.cycles-pp.lseek64
>>>>        1.33 ±  5%      -0.2        1.17 ±  3%
>>>> perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
>>>>        0.51            -0.2        0.36
>>>> perf-profile.children.cycles-pp.grab_cache_page_write_begin
>>>>        0.47            -0.1        0.33
>>>> perf-profile.children.cycles-pp.pagecache_get_page
>>>>        0.51 ±  3%      -0.1        0.38 ±  2%
>>>> perf-profile.children.cycles-pp.iomap_write_end
>>>>        0.57 ±  2%      -0.1        0.46
>>>> perf-profile.children.cycles-pp.blkdev_issue_flush
>>>>        0.38            -0.1        0.26 ±  5%
>>>> perf-profile.children.cycles-pp.iov_iter_copy_from_user_atomic
>>>>        0.34            -0.1        0.23 ±  6%
>>>> perf-profile.children.cycles-pp.copyin
>>>>        0.33            -0.1        0.22 ±  6%
>>>> perf-profile.children.cycles-pp.copy_user_enhanced_fast_string
>>>>        0.65 ±  2%      -0.1        0.56 ±  2%
>>>> perf-profile.children.cycles-pp.rwsem_optimistic_spin
>>>>        0.37            -0.1        0.28 ±  2%
>>>> perf-profile.children.cycles-pp.bio_alloc_bioset
>>>>        0.31 ±  4%      -0.1        0.22
>>>> perf-profile.children.cycles-pp.syscall_return_via_sysret
>>>>        0.56 ±  3%      -0.1        0.48 ±  3%
>>>> perf-profile.children.cycles-pp.osq_lock
>>>>        0.26 ±  3%      -0.1        0.18 ±  4%
>>>> perf-profile.children.cycles-pp.entry_SYSCALL_64
>>>>        0.28 ±  2%      -0.1        0.21 ±  5%
>>>> perf-profile.children.cycles-pp.mempool_alloc
>>>>        0.26 ±  6%      -0.1        0.18 ±  8%
>>>> perf-profile.children.cycles-pp.xlog_state_get_iclog_space
>>>>        0.29            -0.1        0.22 ±  3%
>>>> perf-profile.children.cycles-pp.find_get_entry
>>>>        0.34 ±  2%      -0.1        0.27 ±  4%
>>>> perf-profile.children.cycles-pp.generic_make_request_checks
>>>>        0.30 ±  5%      -0.1        0.23 ±  3%
>>>> perf-profile.children.cycles-pp.iomap_set_page_dirty
>>>>        0.32 ±  2%      -0.1        0.26 ±  5%
>>>> perf-profile.children.cycles-pp.brd_lookup_page
>>>>        0.18 ±  2%      -0.1        0.12 ±  6%
>>>> perf-profile.children.cycles-pp.iomap_set_range_uptodate
>>>>        0.15 ±  2%      -0.1        0.09 ±  4%
>>>> perf-profile.children.cycles-pp.__mod_lruvec_state
>>>>        0.15 ±  5%      -0.1        0.10 ±  9%
>>>> perf-profile.children.cycles-pp.wake_up_q
>>>>        0.21 ±  5%      -0.1        0.15 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_ilock
>>>>        0.13 ±  3%      -0.0        0.08 ± 10%
>>>> perf-profile.children.cycles-pp.rwsem_wake
>>>>        0.20 ±  2%      -0.0        0.15 ±  5%
>>>> perf-profile.children.cycles-pp.ksys_lseek
>>>>        0.20 ±  3%      -0.0        0.15 ±  3%
>>>> perf-profile.children.cycles-pp.xas_load
>>>>        0.20 ±  5%      -0.0        0.16 ±  2%
>>>> perf-profile.children.cycles-pp.__set_page_dirty
>>>>        0.19 ±  2%      -0.0        0.15 ±  2%
>>>> perf-profile.children.cycles-pp.xfs_iunlock
>>>>        0.40            -0.0        0.36 ±  4%
>>>> perf-profile.children.cycles-pp.xfs_buffered_write_iomap_begin
>>>>        0.17 ±  2%      -0.0        0.12 ±  4%
>>>> perf-profile.children.cycles-pp.clear_page_dirty_for_io
>>>>        0.20 ±  3%      -0.0        0.16 ±  4%
>>>> perf-profile.children.cycles-pp.submit_bio_wait
>>>>        0.18 ±  4%      -0.0        0.15 ±  7%
>>>> perf-profile.children.cycles-pp.security_file_permission
>>>>        0.10            -0.0        0.06 ±  6%
>>>> perf-profile.children.cycles-pp.evict
>>>>        0.10            -0.0        0.06 ±  6%
>>>> perf-profile.children.cycles-pp.truncate_inode_pages_range
>>>>        0.06 ±  6%      -0.0        0.03 ±100%
>>>> perf-profile.children.cycles-pp.vfs_unlink
>>>>        0.08 ±  6%      -0.0        0.04 ± 58%
>>>> perf-profile.children.cycles-pp.add_to_page_cache_lru
>>>>        0.26 ±  2%      -0.0        0.23 ±  3%
>>>> perf-profile.children.cycles-pp.reweight_entity
>>>>        0.12 ±  6%      -0.0        0.09 ±  4%
>>>> perf-profile.children.cycles-pp.bio_endio
>>>>        0.10 ±  7%      -0.0        0.07 ±  7%
>>>> perf-profile.children.cycles-pp.do_sys_open
>>>>        0.10 ±  7%      -0.0        0.07 ±  7%
>>>> perf-profile.children.cycles-pp.do_sys_openat2
>>>>        0.17 ±  2%      -0.0        0.14
>>>> perf-profile.children.cycles-pp.__pagevec_release
>>>>        0.13 ±  5%      -0.0        0.10 ±  4%
>>>> perf-profile.children.cycles-pp.down_write
>>>>        0.10 ±  5%      -0.0        0.06 ±  6%
>>>> perf-profile.children.cycles-pp.do_filp_open
>>>>        0.15 ±  4%      -0.0        0.12 ±  5%
>>>> perf-profile.children.cycles-pp.tag_pages_for_writeback
>>>>        0.11 ±  9%      -0.0        0.08 ±  5%
>>>> perf-profile.children.cycles-pp.bio_associate_blkg
>>>>        0.10 ±  5%      -0.0        0.07 ±  7%
>>>> perf-profile.children.cycles-pp.open64
>>>>        0.09 ±  4%      -0.0        0.06 ±  6%
>>>> perf-profile.children.cycles-pp.path_openat
>>>>        0.16 ±  8%      -0.0        0.13 ±  3%
>>>> perf-profile.children.cycles-pp.pagevec_lookup_range_tag
>>>>        0.12 ±  8%      -0.0        0.10 ±  5%
>>>> perf-profile.children.cycles-pp.account_page_dirtied
>>>>        0.10 ±  4%      -0.0        0.08 ±  6%
>>>> perf-profile.children.cycles-pp.__fdget_pos
>>>>        0.14 ±  3%      -0.0        0.11 ±  7%
>>>> perf-profile.children.cycles-pp.memset_erms
>>>>        0.11 ±  3%      -0.0        0.09 ±  4%
>>>> perf-profile.children.cycles-pp.switch_fpu_return
>>>>        0.23            -0.0        0.21 ±  3%
>>>> perf-profile.children.cycles-pp.__radix_tree_lookup
>>>>        0.18 ±  2%      -0.0        0.16 ±  9%
>>>> perf-profile.children.cycles-pp.insert_work
>>>>        0.12 ±  4%      -0.0        0.10 ±  7%
>>>> perf-profile.children.cycles-pp.common_file_perm
>>>>        0.09 ±  4%      -0.0        0.07 ±  6%
>>>> perf-profile.children.cycles-pp.__fget_light
>>>>        0.09 ±  4%      -0.0        0.07 ±  6%
>>>> perf-profile.children.cycles-pp.bvec_alloc
>>>>        0.15 ±  7%      -0.0        0.12 ±  3%
>>>> perf-profile.children.cycles-pp.find_get_pages_range_tag
>>>>        0.08 ±  8%      -0.0        0.06 ±  7%
>>>> perf-profile.children.cycles-pp.iov_iter_fault_in_readable
>>>>        0.09 ±  5%      -0.0        0.06 ± 13%
>>>> perf-profile.children.cycles-pp.blk_throtl_bio
>>>>        0.24 ±  4%      -0.0        0.22 ±  3%
>>>> perf-profile.children.cycles-pp._cond_resched
>>>>        0.13 ±  3%      -0.0        0.11 ±  4%
>>>> perf-profile.children.cycles-pp.release_pages
>>>>        0.10 ±  4%      -0.0        0.08 ±  6%
>>>> perf-profile.children.cycles-pp.xfs_bmbt_to_iomap
>>>>        0.08 ± 10%      -0.0        0.06
>>>> perf-profile.children.cycles-pp.unlink
>>>>        0.08 ± 10%      -0.0        0.06 ± 11%
>>>> perf-profile.children.cycles-pp.__sb_start_write
>>>>        0.08 ±  5%      -0.0        0.06 ±  7%
>>>> perf-profile.children.cycles-pp.generic_write_checks
>>>>        0.08 ±  8%      -0.0        0.06
>>>> perf-profile.children.cycles-pp.do_unlinkat
>>>>        0.07 ± 10%      -0.0        0.05
>>>> perf-profile.children.cycles-pp.xfs_break_layouts
>>>>        0.18 ±  2%      -0.0        0.16 ±  2%
>>>> perf-profile.children.cycles-pp.__list_add_valid
>>>>        0.12 ±  3%      -0.0        0.10 ±  8%
>>>> perf-profile.children.cycles-pp.__might_sleep
>>>>        0.14 ±  3%      -0.0        0.12 ±  7%
>>>> perf-profile.children.cycles-pp.xas_find_marked
>>>>        0.07 ± 10%      -0.0        0.05 ±  8%
>>>> perf-profile.children.cycles-pp.bio_associate_blkg_from_css
>>>>        0.08 ±  5%      -0.0        0.06 ± 11%
>>>> perf-profile.children.cycles-pp.bio_free
>>>>        0.08 ±  6%      -0.0        0.06
>>>> perf-profile.children.cycles-pp.xfs_file_llseek
>>>>        0.08 ±  5%      -0.0        0.06 ±  6%
>>>> perf-profile.children.cycles-pp.xfs_file_write_iter
>>>>        0.16 ±  2%      +0.0        0.17 ±  2%
>>>> perf-profile.children.cycles-pp.__switch_to_asm
>>>>        0.05            +0.0        0.07 ±  7%
>>>> perf-profile.children.cycles-pp.cpumask_next_wrap
>>>>        0.10 ±  4%      +0.0        0.12 ±  3%
>>>> perf-profile.children.cycles-pp._find_next_bit
>>>>        0.07 ±  7%      +0.0        0.09 ±  5%
>>>> perf-profile.children.cycles-pp.xlog_state_done_syncing
>>>>        0.16 ±  2%      +0.0        0.19 ±  2%
>>>> perf-profile.children.cycles-pp.check_preempt_curr
>>>>        0.19 ±  4%      +0.0        0.22
>>>> perf-profile.children.cycles-pp.ttwu_do_wakeup
>>>>        0.56 ±  3%      +0.0        0.60
>>>> perf-profile.children.cycles-pp.dequeue_entity
>>>>        0.20 ±  6%      +0.0        0.23 ±  4%
>>>> perf-profile.children.cycles-pp.xfs_trans_committed_bulk
>>>>        0.07 ±  7%      +0.0        0.10 ±  7%
>>>> perf-profile.children.cycles-pp.check_preempt_wakeup
>>>>        0.82 ±  2%      +0.0        0.86
>>>> perf-profile.children.cycles-pp.update_load_avg
>>>>        0.31 ±  3%      +0.0        0.35
>>>> perf-profile.children.cycles-pp.__unwind_start
>>>>        0.08 ±  8%      +0.0        0.12 ±  3%
>>>> perf-profile.children.cycles-pp.destroy_inode
>>>>        0.08 ± 10%      +0.0        0.12 ±  3%
>>>> perf-profile.children.cycles-pp.xfs_fs_destroy_inode
>>>>        0.08 ± 10%      +0.0        0.12 ±  3%
>>>> perf-profile.children.cycles-pp.xfs_inactive
>>>>        0.08 ± 15%      +0.0        0.12 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_bmap_validate_extent
>>>>        3.96            +0.0        4.01
>>>> perf-profile.children.cycles-pp.stack_trace_save_tsk
>>>>        0.28            +0.0        0.33
>>>> perf-profile.children.cycles-pp.set_next_entity
>>>>        0.00            +0.1        0.05
>>>> perf-profile.children.cycles-pp.xfs_trans_roll
>>>>        0.00            +0.1        0.05 ±  8%
>>>> perf-profile.children.cycles-pp.xfs_btree_delrec
>>>>        0.00            +0.1        0.05 ±  8%
>>>> perf-profile.children.cycles-pp.xfs_btree_insert
>>>>        0.16 ±  4%      +0.1        0.21 ±  5%
>>>> perf-profile.children.cycles-pp.memcpy_erms
>>>>        0.04 ± 57%      +0.1        0.10 ±  4%
>>>> perf-profile.children.cycles-pp.xfs_itruncate_extents_flags
>>>>        0.00            +0.1        0.06 ± 11%
>>>> perf-profile.children.cycles-pp.xfs_buf_item_format_segment
>>>>        0.00            +0.1        0.06 ± 11%
>>>> perf-profile.children.cycles-pp.xfs_btree_delete
>>>>        0.00            +0.1        0.06
>>>> perf-profile.children.cycles-pp.xfs_btree_update
>>>>        0.00            +0.1        0.07 ±  7%
>>>> perf-profile.children.cycles-pp.xfs_cil_prepare_item
>>>>        0.00            +0.1        0.07 ±  6%
>>>> perf-profile.children.cycles-pp.xfs_bmbt_update
>>>>        0.00            +0.1        0.07 ± 10%
>>>> perf-profile.children.cycles-pp.xfs_buf_item_format
>>>>        0.56            +0.1        0.64 ±  2%
>>>> perf-profile.children.cycles-pp.pick_next_task_fair
>>>>        0.03 ±100%      +0.1        0.10 ±  4%
>>>> perf-profile.children.cycles-pp.xfs_inactive_truncate
>>>>        0.08 ±  5%      +0.1        0.16 ± 11%
>>>> perf-profile.children.cycles-pp.xfs_iext_lookup_extent
>>>>        0.00            +0.1        0.08 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_defer_finish
>>>>        0.00            +0.1        0.08 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_bmap_check_leaf_extents
>>>>        0.00            +0.1        0.08 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_verify_agbno
>>>>        0.00            +0.1        0.09 ±  9%
>>>> perf-profile.children.cycles-pp.xfs_defer_finish_noroll
>>>>        0.60 ±  3%      +0.1        0.70
>>>> perf-profile.children.cycles-pp.schedule_idle
>>>>        0.39 ±  2%      +0.1        0.49
>>>> perf-profile.children.cycles-pp.xlog_state_release_iclog
>>>>        0.00            +0.1        0.10 ±  9%
>>>> perf-profile.children.cycles-pp.xfs_buf_find
>>>>        0.93 ±  3%      +0.1        1.03 ±  2%
>>>> perf-profile.children.cycles-pp.xlog_write
>>>>        2.58            +0.1        2.69
>>>> perf-profile.children.cycles-pp.schedule
>>>>        1.52            +0.1        1.63
>>>> perf-profile.children.cycles-pp.__wake_up_common_lock
>>>>        0.00            +0.1        0.11 ±  7%
>>>> perf-profile.children.cycles-pp.xfs_buf_get_map
>>>>        4.62            +0.1        4.73
>>>> perf-profile.children.cycles-pp.__account_scheduler_latency
>>>>        0.00            +0.1        0.12 ± 10%
>>>> perf-profile.children.cycles-pp.xfs_buf_read_map
>>>>        1.83            +0.2        1.99
>>>> perf-profile.children.cycles-pp.xlog_cil_push_work
>>>>        5.94            +0.2        6.12
>>>> perf-profile.children.cycles-pp.ttwu_do_activate
>>>>        0.00            +0.2        0.17 ±  6%
>>>> perf-profile.children.cycles-pp.xfs_btree_read_buf_block
>>>>        5.93            +0.2        6.11
>>>> perf-profile.children.cycles-pp.activate_task
>>>>        5.88            +0.2        6.06
>>>> perf-profile.children.cycles-pp.enqueue_task_fair
>>>>        0.00            +0.2        0.18 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_trans_read_buf_map
>>>>        5.41            +0.2        5.59
>>>> perf-profile.children.cycles-pp.enqueue_entity
>>>>        3.14            +0.2        3.33
>>>> perf-profile.children.cycles-pp.__sched_text_start
>>>>        0.00            +0.2        0.21 ±  6%
>>>> perf-profile.children.cycles-pp.xfs_btree_lookup_get_block
>>>>        0.00            +0.2        0.23 ± 57%
>>>> perf-profile.children.cycles-pp.xlog_grant_head_wait
>>>>        0.00            +0.2        0.24 ±  6%
>>>> perf-profile.children.cycles-pp.io_schedule
>>>>        0.11 ±  3%      +0.3        0.38 ±  4%
>>>> perf-profile.children.cycles-pp.__filemap_fdatawait_range
>>>>        0.00            +0.3        0.27 ±  8%
>>>> perf-profile.children.cycles-pp.wait_on_page_bit
>>>>        0.00            +0.3        0.34 ±  5%
>>>> perf-profile.children.cycles-pp.xfs_btree_lookup
>>>>        0.08 ± 14%      +0.4        0.43
>>>> perf-profile.children.cycles-pp.__queue_work
>>>>        0.08 ±  8%      +0.4        0.44
>>>> perf-profile.children.cycles-pp.queue_work_on
>>>>        0.00            +0.4        0.41
>>>> perf-profile.children.cycles-pp.xfs_end_bio
>>>>        3.51            +0.4        3.94 ±  9%
>>>> perf-profile.children.cycles-pp.__xfs_trans_commit
>>>>        0.20 ±  9%      +0.5        0.72 ± 54%
>>>> perf-profile.children.cycles-pp.xfs_log_ticket_ungrant
>>>>        0.06 ± 11%      +0.5        0.58 ± 67%
>>>> perf-profile.children.cycles-pp.xfs_log_space_wake
>>>>        0.45 ±  5%      +0.6        1.04 ± 36%
>>>> perf-profile.children.cycles-pp.xfs_log_reserve
>>>>        0.65 ±  3%      +0.6        1.25 ± 29%
>>>> perf-profile.children.cycles-pp.xfs_trans_alloc
>>>>        0.48 ±  5%      +0.6        1.09 ± 34%
>>>> perf-profile.children.cycles-pp.xfs_trans_reserve
>>>>        0.04 ± 57%      +0.6        0.65 ± 60%
>>>> perf-profile.children.cycles-pp.xlog_grant_head_check
>>>>        4.83            +0.7        5.52
>>>> perf-profile.children.cycles-pp.try_to_wake_up
>>>>        0.00            +0.7        0.70 ±  2%
>>>> perf-profile.children.cycles-pp.xfs_bmap_add_extent_unwritten_real
>>>>        0.00            +0.8        0.75 ±  4%
>>>> perf-profile.children.cycles-pp.xfs_bmapi_convert_unwritten
>>>>        0.04 ± 57%      +0.8        0.80
>>>> perf-profile.children.cycles-pp.autoremove_wake_function
>>>>        0.05 ±  8%      +0.8        0.85
>>>> perf-profile.children.cycles-pp.wake_up_page_bit
>>>>        1.50            +0.9        2.36
>>>> perf-profile.children.cycles-pp.__wake_up_common
>>>>        0.00            +0.9        0.89 ±  2%
>>>> perf-profile.children.cycles-pp.xfs_bmapi_write
>>>>       24.29            +0.9       25.18
>>>> perf-profile.children.cycles-pp._raw_spin_lock_irqsave
>>>>        0.07 ±  5%      +1.0        1.05
>>>> perf-profile.children.cycles-pp.iomap_finish_ioends
>>>>       35.68            +1.0       36.69
>>>> perf-profile.children.cycles-pp.xlog_cil_force_lsn
>>>>        9.53            +1.2       10.76
>>>> perf-profile.children.cycles-pp.intel_idle
>>>>        9.81            +1.3       11.10
>>>> perf-profile.children.cycles-pp.cpuidle_enter_state
>>>>        9.82            +1.3       11.11
>>>> perf-profile.children.cycles-pp.cpuidle_enter
>>>>       12.25            +1.3       13.60
>>>> perf-profile.children.cycles-pp.start_secondary
>>>>       12.49            +1.4       13.89
>>>> perf-profile.children.cycles-pp.secondary_startup_64
>>>>       12.49            +1.4       13.89
>>>> perf-profile.children.cycles-pp.cpu_startup_entry
>>>>       12.48            +1.4       13.88
>>>> perf-profile.children.cycles-pp.do_idle
>>>>       14.78            +1.4       16.20 ±  2%
>>>> perf-profile.children.cycles-pp.xlog_wait_on_iclog
>>>>       58.91            +1.7       60.57
>>>> perf-profile.children.cycles-pp.xfs_log_force_lsn
>>>>        0.00            +2.2        2.24 ± 13%
>>>> perf-profile.children.cycles-pp.xfs_iomap_write_unwritten
>>>>        7.05            +2.8        9.81 ±  2%
>>>> perf-profile.children.cycles-pp.process_one_work
>>>>        7.54            +2.9       10.45 ±  2%
>>>> perf-profile.children.cycles-pp.worker_thread
>>>>        7.55            +2.9       10.46 ±  2%
>>>> perf-profile.children.cycles-pp.ret_from_fork
>>>>        7.55            +2.9       10.46 ±  2%
>>>> perf-profile.children.cycles-pp.kthread
>>>>        0.10            +3.2        3.30 ±  9%
>>>> perf-profile.children.cycles-pp.xfs_end_ioend
>>>>        0.10 ±  4%      +3.2        3.32 ±  8%
>>>> perf-profile.children.cycles-pp.xfs_end_io
>>>>       19.57            +4.0       23.62
>>>> perf-profile.children.cycles-pp.remove_wait_queue
>>>>        1.20            -0.3        0.90 ±  2%
>>>> perf-profile.self.cycles-pp.brd_do_bvec
>>>>        0.55 ±  3%      -0.2        0.39 ±  4%
>>>> perf-profile.self.cycles-pp.do_syscall_64
>>>>        0.31            -0.1        0.22 ±  5%
>>>> perf-profile.self.cycles-pp.copy_user_enhanced_fast_string
>>>>        0.31 ±  3%      -0.1        0.22
>>>> perf-profile.self.cycles-pp.syscall_return_via_sysret
>>>>        0.54 ±  3%      -0.1        0.47 ±  3%
>>>> perf-profile.self.cycles-pp.osq_lock
>>>>        0.23 ±  3%      -0.1        0.16 ±  5%
>>>> perf-profile.self.cycles-pp.entry_SYSCALL_64
>>>>        0.10 ±  4%      -0.1        0.04 ± 57%
>>>> perf-profile.self.cycles-pp.__mod_lruvec_state
>>>>        0.18 ±  4%      -0.1        0.12 ±  5%
>>>> perf-profile.self.cycles-pp.iomap_set_range_uptodate
>>>>        0.60 ±  4%      -0.1        0.54 ±  2%
>>>> perf-profile.self.cycles-pp._raw_spin_lock
>>>>        0.77            -0.0        0.72
>>>> perf-profile.self.cycles-pp._raw_spin_lock_irqsave
>>>>        0.14            -0.0        0.10 ±  8%
>>>> perf-profile.self.cycles-pp.find_get_entry
>>>>        0.14 ±  3%      -0.0        0.11 ±  4%
>>>> perf-profile.self.cycles-pp.xas_load
>>>>        0.17 ±  4%      -0.0        0.14 ±  3%
>>>> perf-profile.self.cycles-pp.write
>>>>        0.30 ±  2%      -0.0        0.27 ±  6%
>>>> perf-profile.self.cycles-pp._raw_spin_lock_irq
>>>>        0.18 ±  5%      -0.0        0.15 ±  7%
>>>> perf-profile.self.cycles-pp.__flush_work
>>>>        0.08 ±  5%      -0.0        0.05
>>>> perf-profile.self.cycles-pp.xfs_inode_item_format_data_fork
>>>>        0.12 ±  3%      -0.0        0.09 ±  4%
>>>> perf-profile.self.cycles-pp.brd_lookup_page
>>>>        0.12 ±  7%      -0.0        0.09 ±  9%
>>>> perf-profile.self.cycles-pp.vfs_write
>>>>        0.18 ±  4%      -0.0        0.15 ±  7%
>>>> perf-profile.self.cycles-pp.insert_work
>>>>        0.13            -0.0        0.10 ±  8%
>>>> perf-profile.self.cycles-pp.memset_erms
>>>>        0.11 ±  4%      -0.0        0.08
>>>> perf-profile.self.cycles-pp.xfs_file_buffered_aio_write
>>>>        0.11 ±  6%      -0.0        0.08 ±  5%
>>>> perf-profile.self.cycles-pp.xfs_buffered_write_iomap_begin
>>>>        0.12 ±  3%      -0.0        0.09 ±  4%
>>>> perf-profile.self.cycles-pp.iomap_writepage_map
>>>>        0.08 ±  8%      -0.0        0.06 ± 15%
>>>> perf-profile.self.cycles-pp.bio_alloc_bioset
>>>>        0.06 ±  6%      -0.0        0.04 ± 57%
>>>> perf-profile.self.cycles-pp.xfs_bmbt_to_iomap
>>>>        0.09 ±  5%      -0.0        0.06 ±  6%
>>>> perf-profile.self.cycles-pp.__fget_light
>>>>        0.06            -0.0        0.04 ± 57%
>>>> perf-profile.self.cycles-pp.xfs_map_blocks
>>>>        0.18 ±  2%      -0.0        0.15 ±  3%
>>>> perf-profile.self.cycles-pp.__list_add_valid
>>>>        0.08 ±  6%      -0.0        0.06 ±  9%
>>>> perf-profile.self.cycles-pp.xfs_ilock
>>>>        0.08 ±  6%      -0.0        0.06 ±  9%
>>>> perf-profile.self.cycles-pp.iov_iter_fault_in_readable
>>>>        0.11 ±  4%      -0.0        0.09 ±  4%
>>>> perf-profile.self.cycles-pp.switch_fpu_return
>>>>        0.07            -0.0        0.05
>>>> perf-profile.self.cycles-pp.iomap_write_actor
>>>>        0.23 ±  2%      -0.0        0.21 ±  5%
>>>> perf-profile.self.cycles-pp.__radix_tree_lookup
>>>>        0.18 ±  4%      -0.0        0.16 ±  6%
>>>> perf-profile.self.cycles-pp.xlog_cil_force_lsn
>>>>        0.20 ±  2%      -0.0        0.18 ±  4%
>>>> perf-profile.self.cycles-pp.kmem_cache_alloc
>>>>        0.10 ±  5%      -0.0        0.08 ± 11%
>>>> perf-profile.self.cycles-pp.xfs_file_fsync
>>>>        0.07 ��  5%      -0.0        0.05 ±  9%
>>>> perf-profile.self.cycles-pp.clear_page_dirty_for_io
>>>>        0.07 ±  6%      -0.0        0.05
>>>> perf-profile.self.cycles-pp.__test_set_page_writeback
>>>>        0.11            -0.0        0.09 ±  8%
>>>> perf-profile.self.cycles-pp.common_file_perm
>>>>        0.23            -0.0        0.21 ±  4%
>>>> perf-profile.self.cycles-pp.update_rq_clock
>>>>        0.08 ±  6%      -0.0        0.06
>>>> perf-profile.self.cycles-pp.xfs_file_write_iter
>>>>        0.10 ±  5%      -0.0        0.08 ±  8%
>>>> perf-profile.self.cycles-pp.up_read
>>>>        0.08 ±  5%      -0.0        0.07
>>>> perf-profile.self.cycles-pp.write_cache_pages
>>>>        0.10            -0.0        0.09
>>>> perf-profile.self.cycles-pp.reweight_entity
>>>>        0.05 ±  8%      +0.0        0.07 ±  7%
>>>> perf-profile.self.cycles-pp.dequeue_entity
>>>>        0.16 ±  4%      +0.0        0.17 ±  2%
>>>> perf-profile.self.cycles-pp.__switch_to_asm
>>>>        0.11 ±  4%      +0.0        0.12 ±  3%
>>>> perf-profile.self.cycles-pp.set_next_entity
>>>>        0.15 ±  3%      +0.0        0.16 ±  2%
>>>> perf-profile.self.cycles-pp.orc_find
>>>>        0.09 ±  4%      +0.0        0.11 ±  6%
>>>> perf-profile.self.cycles-pp.enqueue_entity
>>>>        0.10 ±  4%      +0.0        0.12 ±  3%
>>>> perf-profile.self.cycles-pp._find_next_bit
>>>>        0.01 ±173%      +0.0        0.06 ±  7%
>>>> perf-profile.self.cycles-pp.check_preempt_wakeup
>>>>        0.01 ±173%      +0.0        0.06 ± 11%
>>>> perf-profile.self.cycles-pp.xfs_trans_committed_bulk
>>>>        0.45 ±  2%      +0.1        0.51 ±  2%
>>>> perf-profile.self.cycles-pp.__sched_text_start
>>>>        0.15 ±  6%      +0.1        0.21 ±  3%
>>>> perf-profile.self.cycles-pp.memcpy_erms
>>>>        0.00            +0.1        0.06 ±  6%
>>>> perf-profile.self.cycles-pp.xfs_verify_agbno
>>>>        0.08 ±  8%      +0.1        0.16 ±  9%
>>>> perf-profile.self.cycles-pp.xfs_iext_lookup_extent
>>>>        9.53            +1.2       10.76
>>>> perf-profile.self.cycles-pp.intel_idle
>>>>
>>>>
>>>>                                    aim7.jobs-per-min
>>>>    45000
>>>> +-------------------------------------------------------------------+
>>>>          |        +
>>>> +                                                       |
>>>>    40000 |.+.+    :   + .+.   +.
>>>> .+.     |
>>>>    35000 |-+ :    :    +   +.+
>>>> +.+.+.++.+.+.+.++.+.+.+.++.+.+.++.+.+   ++.+.|
>>>>          |   :
>>>> :                                                          |
>>>>    30000 |-+ :
>>>> :                                                           |
>>>>    25000 |-O :O O:O
>>>> OO                                                     |
>>>>          |   :   :  O    O O OO O O O
>>>> O                                      |
>>>>    20000 |-+  :
>>>> :                                                           |
>>>>    15000 |-+  :
>>>> :                                                           |
>>>>          |    :
>>>> :                                                           |
>>>>    10000 |-+  :
>>>> :                                                            |
>>>>     5000 |-+  :
>>>> :                                                            |
>>>>          |    :
>>>> :                                                            |
>>>>        0
>>>> +-------------------------------------------------------------------+
>>>>                                 aim7.time.system_time
>>>>    5000
>>>> +--------------------------------------------------------------------+
>>>>    4500 |-+        O   OO   O    O O
>>>> O                                       |
>>>>         | O OO O O   O    O   O
>>>> O                                            |
>>>>    4000
>>>> |-+
>>>> |
>>>>    3500
>>>> |-+
>>>> |
>>>>         |             .+ .+.+. .++.+.+.+.+.++.+.+.+.++.+.+.+.+
>>>> .+.+.+. .++.+.|
>>>>    3000 |.+.+       .+  +     +                               +
>>>> +      |
>>>>    2500 |-+ :
>>>> +.+                                                         |
>>>>    2000 |-+ :
>>>> :                                                           |
>>>>         |   :
>>>> :                                                            |
>>>>    1500 |-+  :
>>>> :                                                            |
>>>>    1000 |-+  :
>>>> :                                                            |
>>>>         |    :
>>>> :                                                            |
>>>>     500 |-+  :
>>>> :                                                             |
>>>>       0
>>>> +--------------------------------------------------------------------+
>>>>                                aim7.time.elapsed_time
>>>>    160
>>>> +---------------------------------------------------------------------+
>>>>        |   O OO O O O O OO O O O
>>>> O                                         |
>>>>    140
>>>> |-O
>>>> |
>>>>    120
>>>> |-+
>>>> |
>>>> |                                                                     |
>>>>    100 |-+           .+. +.+.
>>>> .+.+.++.+.+.+.+.+.++.+.+.+.+.++.+.+.+.+. +.+.+.|
>>>>        |.+.+    +. .+   +    +
>>>> +      |
>>>>     80 |-+ :    :
>>>> +                                                          |
>>>>        |    :
>>>> :                                                            |
>>>>     60 |-+  :
>>>> :                                                             |
>>>>     40 |-+  :
>>>> :                                                             |
>>>>        |    :
>>>> :                                                             |
>>>>     20 |-+   :
>>>> :                                                             |
>>>>        |
>>>> ::                                                              |
>>>>      0
>>>> +---------------------------------------------------------------------+
>>>>                              aim7.time.elapsed_time.max
>>>>    160
>>>> +---------------------------------------------------------------------+
>>>>        |   O OO O O O O OO O O O
>>>> O                                         |
>>>>    140
>>>> |-O
>>>> |
>>>>    120
>>>> |-+
>>>> |
>>>> |                                                                     |
>>>>    100 |-+           .+. +.+.
>>>> .+.+.++.+.+.+.+.+.++.+.+.+.+.++.+.+.+.+. +.+.+.|
>>>>        |.+.+    +. .+   +    +
>>>> +      |
>>>>     80 |-+ :    :
>>>> +                                                          |
>>>>        |    :
>>>> :                                                            |
>>>>     60 |-+  :
>>>> :                                                             |
>>>>     40 |-+  :
>>>> :                                                             |
>>>>        |    :
>>>> :                                                             |
>>>>     20 |-+   :
>>>> :                                                             |
>>>>        |
>>>> ::                                                              |
>>>>      0
>>>> +---------------------------------------------------------------------+
>>>>                          aim7.time.voluntary_context_switches
>>>>    8e+07
>>>> +-------------------------------------------------------------------+
>>>>          | O          O  O O OO O
>>>> O                                        |
>>>>    7e+07
>>>> |-+
>>>> |
>>>>    6e+07
>>>> |-+
>>>> |
>>>> |                                                                   |
>>>>    5e+07 |-+           +.+.+.++.+.+.+.++.+.+. .+   .+.+.    .+.
>>>> .+.   +.+.|
>>>>          |.+.+       .+                      +  +.+     ++.+   ++.+
>>>> +.+    |
>>>>    4e+07 |-+ :
>>>> +.+                                                        |
>>>>          |   :
>>>> :                                                          |
>>>>    3e+07 |-+ :
>>>> :                                                           |
>>>>    2e+07 |-+  :
>>>> :                                                           |
>>>>          |    :
>>>> :                                                           |
>>>>    1e+07 |-+  :
>>>> :                                                           |
>>>>          |    :
>>>> :                                                            |
>>>>        0
>>>> +-------------------------------------------------------------------+
>>>> [*] bisect-good sample
>>>> [O] bisect-bad  sample
>>>>
>>>>
>>>>
>>>> Disclaimer:
>>>> Results have been estimated based on internal Intel analysis and are
>>>> provided
>>>> for informational purposes only. Any difference in system hardware
>>>> or software
>>>> design or configuration may affect actual performance.
>>>>
>>>>
>>>> Thanks,
>>>> Rong Chen
>>>>
>>>>
>>>> _______________________________________________
>>>> LKP mailing list -- lkp@lists.01.org
>>>> To unsubscribe send an email to lkp-leave@lists.01.org
>>>>
>>>
>>
>> --
>> Zhengjun Xing

--
Zhengjun Xing

\
 
 \ /
  Last update: 2020-07-22 09:22    [W:0.227 / U:0.212 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site