lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Date:	Mon, 20 Jun 2016 15:16:27 +0800
From:	kernel test robot <xiaolong.ye@...el.com>
To:	Trond Myklebust <trond.myklebust@...marydata.com>
Cc:	0day robot <fengguang.wu@...el.com>,
	LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp] [SUNRPC]  818e8e4109: fsmark.files_per_sec -7.7% regression


FYI, we noticed a -7.7% regression of fsmark.files_per_sec due to commit:

commit 818e8e4109d58cc55c884223a1d93426024d23d8 ("SUNRPC: RPC transport queue must be low latency")
https://github.com/0day-ci/linux Trond-Myklebust/SUNRPC-Small-optimisation-of-client-receive/20160615-035043

in testcase: fsmark
on test machine: ivb44: 48 threads Ivytown Ivy Bridge-EP with 64G memory
with following parameters: cpufreq_governor=performance/disk=1BRD_48G/filesize=4M/fs2=nfsv4/fs=xfs/iterations=1x/nr_threads=64t/sync_method=NoSync/test_size=40G



Details are as below:
-------------------------------------------------------------------------------------------------->


To reproduce:

        git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/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/filesize/fs2/fs/iterations/kconfig/nr_threads/rootfs/sync_method/tbox_group/test_size/testcase:
  gcc-4.9/performance/1BRD_48G/4M/nfsv4/xfs/1x/x86_64-rhel/64t/debian-x86_64-2015-02-07.cgz/NoSync/ivb44/40G/fsmark

commit: 
  6b25579f3a0c53fc63884d1487e3e8434670def4
  818e8e4109d58cc55c884223a1d93426024d23d8

6b25579f3a0c53fc 818e8e4109d58cc55c884223a1 
---------------- -------------------------- 
         %stddev     %change         %stddev
             \          |                \  
  10245512 ±  2%     -54.4%    4675370 ±  2%  fsmark.app_overhead
     83.20 ±  0%      -7.7%      76.80 ±  0%  fsmark.files_per_sec
    123.40 ±  0%      +6.2%     131.06 ±  0%  fsmark.time.elapsed_time
    123.40 ±  0%      +6.2%     131.06 ±  0%  fsmark.time.elapsed_time.max
     45128 ±  0%     -68.3%      14323 ±  1%  fsmark.time.involuntary_context_switches
     37.50 ±  1%     -17.3%      31.00 ±  0%  fsmark.time.percent_of_cpu_this_job_got
     46.66 ±  0%     -11.5%      41.30 ±  0%  fsmark.time.system_time
   1488713 ±  0%      -6.6%    1390819 ±  0%  fsmark.time.voluntary_context_switches
    280541 ±  2%     +62.5%     455972 ±  0%  softirqs.NET_RX
    334642 ±  0%      -5.7%     315581 ±  0%  vmstat.io.bo
   5821075 ±  0%     +21.6%    7079435 ±  0%  vmstat.memory.cache
     46892 ±  0%      -5.7%      44240 ±  0%  vmstat.system.cs
     21396 ± 28%     -67.4%       6979 ± 11%  latency_stats.max.call_rwsem_down_write_failed.filename_create.SyS_mkdir.entry_SYSCALL_64_fastpath
     10905 ± 16%    +474.2%      62621 ±165%  latency_stats.max.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_lookup.[nfsv4].nfs4_proc_lookup_common.[nfsv4].nfs4_proc_lookup.[nfsv4].nfs_lookup_revalidate.nfs4_lookup_revalidate.lookup_dcache.__lookup_hash
   3782146 ±  6%     -70.4%    1120937 ±  2%  latency_stats.sum.call_rwsem_down_write_failed.filename_create.SyS_mkdir.entry_SYSCALL_64_fastpath
    208165 ± 10%     -75.5%      51062 ± 13%  latency_stats.sum.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_access.[nfsv4].nfs4_proc_access.[nfsv4].nfs_do_access.nfs_permission.__inode_permission.inode_permission.link_path_walk
      7.46 ±  0%      -6.5%       6.98 ±  0%  turbostat.%Busy
    158.75 ±  0%      -5.8%     149.50 ±  0%  turbostat.Avg_MHz
      0.31 ±  6%     -35.0%       0.20 ±  3%  turbostat.CPU%c3
      4.73 ±  0%      -1.5%       4.65 ±  0%  turbostat.RAMWatt
  37004192 ±  2%     -73.8%    9682574 ±  1%  cpuidle.C1E-IVT.time
    120363 ±  1%     -52.9%      56683 ±  1%  cpuidle.C1E-IVT.usage
  23302776 ±  3%     -24.2%   17659372 ±  2%  cpuidle.C3-IVT.time
     82331 ±  1%     -21.0%      65040 ±  0%  cpuidle.C3-IVT.usage
     36963 ±  1%     -25.0%      27736 ±  2%  cpuidle.POLL.usage
   5420047 ±  1%     +23.2%    6677171 ±  0%  meminfo.Cached
      3081 ±  6%     -46.5%       1648 ± 12%  meminfo.Dirty
   5358458 ±  1%     +23.3%    6605440 ±  0%  meminfo.Inactive
   5349323 ±  1%     +23.3%    6596293 ±  0%  meminfo.Inactive(file)
      3093 ± 13%     -56.6%       1344 ±  1%  meminfo.NFS_Unstable
    145243 ±  1%     +15.9%     168335 ±  0%  meminfo.SReclaimable
     44381 ±  3%     +22.4%      54339 ±  2%  meminfo.Unevictable
      2696 ±  6%     -49.5%       1362 ±  7%  meminfo.Writeback
    615701 ±  1%     +25.7%     773764 ±  0%  slabinfo.buffer_head.active_objs
     15787 ±  1%     +25.7%      19840 ±  0%  slabinfo.buffer_head.active_slabs
    615714 ±  1%     +25.7%     773776 ±  0%  slabinfo.buffer_head.num_objs
     15787 ±  1%     +25.7%      19840 ±  0%  slabinfo.buffer_head.num_slabs
     57899 ±  2%     +21.8%      70515 ±  1%  slabinfo.radix_tree_node.active_objs
      1037 ±  2%     +21.3%       1258 ±  1%  slabinfo.radix_tree_node.active_slabs
     58118 ±  2%     +21.3%      70515 ±  1%  slabinfo.radix_tree_node.num_objs
      1037 ±  2%     +21.3%       1258 ±  1%  slabinfo.radix_tree_node.num_slabs
      1266 ±  2%     -10.9%       1128 ±  3%  slabinfo.scsi_data_buffer.active_objs
      1266 ±  2%     -10.9%       1128 ±  3%  slabinfo.scsi_data_buffer.num_objs
    760.25 ±  2%     +22.2%     929.00 ±  1%  proc-vmstat.kswapd_low_wmark_hit_quickly
    760.00 ±  4%     -44.9%     418.50 ±  8%  proc-vmstat.nr_dirty
   1355408 ±  1%     +23.2%    1669485 ±  0%  proc-vmstat.nr_file_pages
   1337536 ±  1%     +23.3%    1649068 ±  0%  proc-vmstat.nr_inactive_file
     36313 ±  1%     +15.9%      42082 ±  0%  proc-vmstat.nr_slab_reclaimable
     11094 ±  3%     +22.4%      13584 ±  2%  proc-vmstat.nr_unevictable
    756.00 ± 13%     -55.8%     334.25 ±  2%  proc-vmstat.nr_unstable
    681.75 ±  9%     -48.2%     353.25 ±  5%  proc-vmstat.nr_writeback
      1488 ± 34%     -90.8%     136.25 ±173%  proc-vmstat.pgscan_direct_dma32
     34551 ± 22%     -91.0%       3113 ±173%  proc-vmstat.pgscan_direct_normal
      1004 ± 49%     -86.4%     136.25 ±173%  proc-vmstat.pgsteal_direct_dma32
      7921 ± 64%     -83.3%       1322 ±173%  proc-vmstat.pgsteal_direct_normal
  9.06e+08 ±  2%     -10.8%  8.078e+08 ±  4%  perf-stat.L1-dcache-prefetch-misses
 4.514e+09 ±  0%      +5.6%  4.766e+09 ±  1%  perf-stat.L1-icache-load-misses
  1.53e+09 ±  5%     -20.4%  1.218e+09 ±  0%  perf-stat.LLC-load-misses
 2.351e+09 ±  3%     +10.2%  2.592e+09 ±  2%  perf-stat.LLC-prefetches
  2.48e+09 ±  3%     +10.7%  2.746e+09 ±  2%  perf-stat.LLC-store-misses
 3.651e+09 ±  1%      +5.6%  3.855e+09 ±  2%  perf-stat.LLC-stores
 6.288e+10 ±  4%     +11.5%   7.01e+10 ±  0%  perf-stat.branch-instructions
 1.176e+09 ±  2%     +11.5%  1.311e+09 ±  4%  perf-stat.branch-load-misses
 5.597e+10 ± 10%     +13.1%  6.332e+10 ±  6%  perf-stat.branch-loads
 1.241e+09 ±  2%     +11.6%  1.386e+09 ±  4%  perf-stat.branch-misses
 1.038e+11 ±  2%      +9.3%  1.134e+11 ±  2%  perf-stat.dTLB-loads
  92930638 ±  1%      +4.9%   97490872 ±  2%  perf-stat.iTLB-load-misses
    277756 ±  0%      +5.3%     292450 ±  0%  perf-stat.minor-faults
 1.138e+09 ± 10%     -26.1%  8.414e+08 ±  5%  perf-stat.node-loads
 6.039e+08 ±  9%     -18.9%  4.896e+08 ±  4%  perf-stat.node-prefetches
 5.084e+08 ±  2%     +25.3%  6.372e+08 ±  5%  perf-stat.node-store-misses
    277802 ±  0%      +5.3%     292431 ±  0%  perf-stat.page-faults
      1398 ± 12%     -49.8%     701.75 ± 12%  numa-meminfo.node0.Dirty
   2673371 ±  2%     +21.0%    3233672 ±  6%  numa-meminfo.node0.FilePages
   2642305 ±  2%     +21.0%    3197794 ±  6%  numa-meminfo.node0.Inactive
   2639728 ±  1%     +21.1%    3197516 ±  6%  numa-meminfo.node0.Inactive(file)
   4995759 ±  2%     +21.7%    6082060 ±  2%  numa-meminfo.node0.MemUsed
      1521 ±  5%     -59.6%     614.25 ± 11%  numa-meminfo.node0.NFS_Unstable
     21760 ±  3%     +23.2%      26805 ±  0%  numa-meminfo.node0.Unevictable
      1438 ± 17%     -51.5%     698.00 ± 13%  numa-meminfo.node0.Writeback
      3517 ± 61%     +56.7%       5511 ± 15%  numa-meminfo.node1.AnonHugePages
      1634 ±  9%     -46.2%     879.00 ±  9%  numa-meminfo.node1.Dirty
   2726731 ±  1%     +25.8%    3429717 ±  5%  numa-meminfo.node1.FilePages
   2696295 ±  1%     +25.9%    3393847 ±  5%  numa-meminfo.node1.Inactive
   2689711 ±  1%     +25.8%    3384959 ±  5%  numa-meminfo.node1.Inactive(file)
   4917935 ±  2%     +23.5%    6072040 ±  2%  numa-meminfo.node1.MemUsed
      1384 ±  2%     -48.5%     713.75 ± 21%  numa-meminfo.node1.NFS_Unstable
     69455 ±  1%     +22.6%      85179 ±  7%  numa-meminfo.node1.SReclaimable
    120279 ±  3%     +14.6%     137894 ±  8%  numa-meminfo.node1.Slab
     21762 ±  3%     +23.3%      26831 ±  0%  numa-meminfo.node1.Unevictable
      1702 ±  9%     -58.3%     709.50 ± 11%  numa-meminfo.node1.Writeback
    975493 ±  2%     +21.0%    1180348 ±  5%  numa-vmstat.node0.nr_dirtied
    380.50 ±  9%     -55.2%     170.50 ± 13%  numa-vmstat.node0.nr_dirty
    668385 ±  2%     +20.9%     808309 ±  6%  numa-vmstat.node0.nr_file_pages
    659973 ±  1%     +21.1%     799269 ±  6%  numa-vmstat.node0.nr_inactive_file
      5439 ±  3%     +23.2%       6701 ±  0%  numa-vmstat.node0.nr_unevictable
    369.25 ±  7%     -59.4%     150.00 ±  9%  numa-vmstat.node0.nr_unstable
    379.00 ± 15%     -56.6%     164.50 ± 11%  numa-vmstat.node0.nr_writeback
    974763 ±  2%     +21.1%    1180042 ±  5%  numa-vmstat.node0.nr_written
   1655473 ±  2%     +21.5%    2011897 ±  6%  numa-vmstat.node0.numa_hit
   1655473 ±  2%     +21.5%    2011897 ±  6%  numa-vmstat.node0.numa_local
    967615 ±  3%     +28.0%    1238876 ±  5%  numa-vmstat.node1.nr_dirtied
    367.00 ±  0%     -40.3%     219.00 ±  9%  numa-vmstat.node1.nr_dirty
    681556 ±  1%     +25.8%     857306 ±  5%  numa-vmstat.node1.nr_file_pages
    672299 ±  1%     +25.9%     846118 ±  5%  numa-vmstat.node1.nr_inactive_file
     17358 ±  1%     +22.7%      21291 ±  7%  numa-vmstat.node1.nr_slab_reclaimable
      5440 ±  3%     +23.3%       6708 ±  0%  numa-vmstat.node1.nr_unevictable
    348.75 ± 10%     -45.3%     190.75 ± 17%  numa-vmstat.node1.nr_unstable
    389.75 ± 13%     -48.5%     200.75 ±  4%  numa-vmstat.node1.nr_writeback
    966871 ±  3%     +28.1%    1238490 ±  5%  numa-vmstat.node1.nr_written
   1594506 ±  4%     +27.2%    2028330 ±  5%  numa-vmstat.node1.numa_hit
   1594505 ±  4%     +27.2%    2028329 ±  5%  numa-vmstat.node1.numa_local
    976.82 ±  1%     -26.3%     719.87 ±  1%  sched_debug.cfs_rq:/.exec_clock.avg
      2411 ±  3%     -12.9%       2101 ±  3%  sched_debug.cfs_rq:/.exec_clock.max
     58.45 ± 11%     -60.9%      22.88 ±  8%  sched_debug.cfs_rq:/.exec_clock.min
    887.20 ±  1%     -20.3%     706.71 ±  0%  sched_debug.cfs_rq:/.exec_clock.stddev
    330.47 ±  5%     -60.8%     129.43 ±  6%  sched_debug.cfs_rq:/.load_avg.avg
      1765 ±  7%     -56.7%     763.92 ±  9%  sched_debug.cfs_rq:/.load_avg.max
    456.70 ±  7%     -56.4%     199.20 ±  7%  sched_debug.cfs_rq:/.load_avg.stddev
     49.53 ± 14%     -33.3%      33.04 ± 10%  sched_debug.cfs_rq:/.runnable_load_avg.avg
    140.97 ± 23%     -26.0%     104.31 ±  4%  sched_debug.cfs_rq:/.runnable_load_avg.stddev
    746.08 ±  5%     -12.9%     649.58 ±  2%  sched_debug.cfs_rq:/.util_avg.max
    153.35 ±  6%     -16.3%     128.31 ±  3%  sched_debug.cfs_rq:/.util_avg.stddev
     36.90 ± 11%     -21.6%      28.93 ± 13%  sched_debug.cpu.cpu_load[0].avg
    108.16 ±  5%     -10.7%      96.61 ±  6%  sched_debug.cpu.cpu_load[0].stddev
     67.83 ±  6%     -22.6%      52.51 ±  6%  sched_debug.cpu.cpu_load[1].avg
    133.54 ±  7%     -17.7%     109.90 ± 12%  sched_debug.cpu.cpu_load[1].stddev
     62.46 ±  7%     -23.9%      47.54 ±  3%  sched_debug.cpu.cpu_load[2].avg
    632.17 ± 11%     -14.3%     541.67 ±  9%  sched_debug.cpu.cpu_load[2].max
    119.72 ±  7%     -18.3%      97.77 ±  7%  sched_debug.cpu.cpu_load[2].stddev
     55.54 ±  7%     -24.7%      41.80 ±  2%  sched_debug.cpu.cpu_load[3].avg
    589.17 ±  6%     -17.2%     488.00 ±  4%  sched_debug.cpu.cpu_load[3].max
    107.89 ±  5%     -19.1%      87.30 ±  4%  sched_debug.cpu.cpu_load[3].stddev
     50.93 ±  8%     -26.0%      37.71 ±  2%  sched_debug.cpu.cpu_load[4].avg
    543.75 ±  4%     -16.7%     453.08 ±  4%  sched_debug.cpu.cpu_load[4].max
     99.09 ±  3%     -18.6%      80.63 ±  3%  sched_debug.cpu.cpu_load[4].stddev
      0.00 ± 44%     -43.6%       0.00 ± 13%  sched_debug.cpu.next_balance.stddev
      2415 ±  6%     -11.7%       2133 ±  4%  sched_debug.cpu.nr_load_updates.stddev
     37473 ±  1%     -27.5%      27179 ±  0%  sched_debug.cpu.nr_switches.avg
     87510 ±  7%     -23.2%      67204 ±  6%  sched_debug.cpu.nr_switches.max
      1512 ± 15%     -58.6%     625.33 ±  8%  sched_debug.cpu.nr_switches.min
     35187 ±  1%     -24.7%      26504 ±  1%  sched_debug.cpu.nr_switches.stddev
     12.59 ± 10%     -27.2%       9.16 ±  8%  sched_debug.cpu.nr_uninterruptible.stddev
     36793 ±  1%     -28.0%      26494 ±  0%  sched_debug.cpu.sched_count.avg
     90630 ±  4%     -22.0%      70672 ±  9%  sched_debug.cpu.sched_count.max
      1189 ± 16%     -71.9%     333.92 ± 13%  sched_debug.cpu.sched_count.min
     35241 ±  1%     -24.8%      26503 ±  1%  sched_debug.cpu.sched_count.stddev
     16975 ±  1%     -28.4%      12159 ±  1%  sched_debug.cpu.sched_goidle.avg
     39073 ±  6%     -22.9%      30133 ±  6%  sched_debug.cpu.sched_goidle.max
    517.50 ± 17%     -75.5%     126.83 ± 13%  sched_debug.cpu.sched_goidle.min
     16091 ±  1%     -25.3%      12019 ±  1%  sched_debug.cpu.sched_goidle.stddev
     18581 ±  1%     -27.9%      13391 ±  1%  sched_debug.cpu.ttwu_count.avg
     43020 ±  5%     -25.7%      31949 ±  7%  sched_debug.cpu.ttwu_count.max
    919.08 ± 15%     -77.3%     208.75 ±  2%  sched_debug.cpu.ttwu_count.min
     16996 ±  1%     -23.8%      12956 ±  1%  sched_debug.cpu.ttwu_count.stddev
      1647 ±  0%     -18.3%       1347 ±  0%  sched_debug.cpu.ttwu_local.avg
      3739 ±  3%     -14.2%       3207 ±  3%  sched_debug.cpu.ttwu_local.max
    120.75 ±  4%     -37.2%      75.83 ±  3%  sched_debug.cpu.ttwu_local.min
      1476 ±  1%     -13.7%       1273 ±  1%  sched_debug.cpu.ttwu_local.stddev
      1.05 ±  4%     -12.8%       0.92 ±  2%  perf-profile.cycles-pp.__filemap_fdatawait_range.filemap_fdatawait_range.filemap_write_and_wait_range.xfs_file_fsync.vfs_fsync_range
     12.69 ±  1%     -25.3%       9.47 ±  1%  perf-profile.cycles-pp.__filemap_fdatawrite_range.filemap_write_and_wait_range.xfs_file_fsync.vfs_fsync_range.nfsd_commit
     10.17 ±  5%     -17.3%       8.41 ±  5%  perf-profile.cycles-pp.__generic_file_write_iter.generic_file_write_iter.nfs_file_write.__vfs_write.vfs_write
      1.70 ±  2%     +17.8%       2.00 ±  5%  perf-profile.cycles-pp.__rpc_execute.rpc_async_schedule.process_one_work.worker_thread.kthread
     10.41 ±  5%     -18.1%       8.52 ±  4%  perf-profile.cycles-pp.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath
      1.25 ±  6%     +52.7%       1.90 ±  5%  perf-profile.cycles-pp.apic_timer_interrupt.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
      7.69 ±  2%     -30.9%       5.31 ±  3%  perf-profile.cycles-pp.brd_do_bvec.brd_make_request.generic_make_request.submit_bio.xfs_add_to_ioend
      2.61 ±  2%     -28.3%       1.87 ±  5%  perf-profile.cycles-pp.brd_do_bvec.brd_make_request.generic_make_request.submit_bio.xfs_submit_ioend.isra.17
      4.68 ±  1%     -19.5%       3.77 ±  2%  perf-profile.cycles-pp.brd_insert_page.part.4.brd_do_bvec.brd_make_request.generic_make_request.submit_bio
      7.80 ±  2%     -30.5%       5.42 ±  3%  perf-profile.cycles-pp.brd_make_request.generic_make_request.submit_bio.xfs_add_to_ioend.xfs_do_writepage
      2.64 ±  2%     -26.3%       1.94 ±  2%  perf-profile.cycles-pp.brd_make_request.generic_make_request.submit_bio.xfs_submit_ioend.isra.17.xfs_vm_writepages
      3.15 ±  7%    +121.1%       6.96 ±  2%  perf-profile.cycles-pp.call_rwsem_down_write_failed.down_write.xfs_file_buffered_aio_write.xfs_file_write_iter.do_iter_readv_writev
      1.31 ±  4%     +26.0%       1.66 ±  5%  perf-profile.cycles-pp.call_transmit.__rpc_execute.rpc_async_schedule.process_one_work.worker_thread
      3.58 ±  7%     -16.2%       3.00 ±  3%  perf-profile.cycles-pp.copy_page_to_iter.skb_copy_datagram_iter.tcp_recvmsg.inet_recvmsg.sock_recvmsg
      3.89 ±  5%     -13.6%       3.36 ±  5%  perf-profile.cycles-pp.copy_user_enhanced_fast_string.generic_perform_write.__generic_file_write_iter.generic_file_write_iter.nfs_file_write
     13.17 ±  6%     +26.5%      16.66 ±  1%  perf-profile.cycles-pp.do_iter_readv_writev.do_readv_writev.vfs_writev.nfsd_vfs_write.nfsd4_write
     13.21 ±  6%     +26.3%      16.69 ±  1%  perf-profile.cycles-pp.do_readv_writev.vfs_writev.nfsd_vfs_write.nfsd4_write.nfsd4_proc_compound
     12.68 ±  1%     -25.3%       9.47 ±  1%  perf-profile.cycles-pp.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.xfs_file_fsync.vfs_fsync_range
      3.15 ±  7%    +120.8%       6.96 ±  2%  perf-profile.cycles-pp.down_write.xfs_file_buffered_aio_write.xfs_file_write_iter.do_iter_readv_writev.do_readv_writev
     16.75 ±  3%     -13.9%      14.42 ±  4%  perf-profile.cycles-pp.entry_SYSCALL_64_fastpath
      1.07 ±  4%     -12.9%       0.93 ±  3%  perf-profile.cycles-pp.filemap_fdatawait_range.filemap_write_and_wait_range.xfs_file_fsync.vfs_fsync_range.nfsd_commit
      3.51 ±  4%      -7.4%       3.25 ±  5%  perf-profile.cycles-pp.filemap_write_and_wait_range.nfs_file_fsync.vfs_fsync_range.vfs_fsync.nfs4_file_flush
     13.75 ±  1%     -24.3%      10.41 ±  1%  perf-profile.cycles-pp.filemap_write_and_wait_range.xfs_file_fsync.vfs_fsync_range.nfsd_commit.nfsd4_commit
     10.23 ±  5%     -17.4%       8.46 ±  5%  perf-profile.cycles-pp.generic_file_write_iter.nfs_file_write.__vfs_write.vfs_write.sys_write
      7.84 ±  1%     -30.3%       5.47 ±  3%  perf-profile.cycles-pp.generic_make_request.submit_bio.xfs_add_to_ioend.xfs_do_writepage.write_cache_pages
      2.64 ±  2%     -26.0%       1.95 ±  2%  perf-profile.cycles-pp.generic_make_request.submit_bio.xfs_submit_ioend.isra.17.xfs_vm_writepages.do_writepages
     10.05 ±  5%     -17.8%       8.26 ±  5%  perf-profile.cycles-pp.generic_perform_write.__generic_file_write_iter.generic_file_write_iter.nfs_file_write.__vfs_write
      4.65 ±  7%     -13.6%       4.02 ±  2%  perf-profile.cycles-pp.inet_recvmsg.sock_recvmsg.kernel_recvmsg.svc_recvfrom.isra.9.svc_tcp_recvfrom
      1.00 ±  4%     +27.6%       1.28 ±  8%  perf-profile.cycles-pp.inet_sendpage.xs_sendpages.xs_tcp_send_request.xprt_transmit.call_transmit
      4.66 ±  7%     -13.3%       4.04 ±  2%  perf-profile.cycles-pp.kernel_recvmsg.svc_recvfrom.isra.9.svc_tcp_recvfrom.svc_recv.nfsd
      3.45 ±  6%     -16.4%       2.89 ±  4%  perf-profile.cycles-pp.memcpy_erms.copy_page_to_iter.skb_copy_datagram_iter.tcp_recvmsg.inet_recvmsg
     10.39 ±  5%     -18.0%       8.52 ±  5%  perf-profile.cycles-pp.nfs_file_write.__vfs_write.vfs_write.sys_write.entry_SYSCALL_64_fastpath
      1.85 ±  3%     -15.4%       1.57 ±  6%  perf-profile.cycles-pp.nfs_updatepage.nfs_write_end.generic_perform_write.__generic_file_write_iter.generic_file_write_iter
      3.80 ±  4%     -24.3%       2.88 ±  7%  perf-profile.cycles-pp.nfs_write_end.generic_perform_write.__generic_file_write_iter.generic_file_write_iter.nfs_file_write
     13.81 ±  1%     -24.0%      10.50 ±  1%  perf-profile.cycles-pp.nfsd4_commit.nfsd4_proc_compound.nfsd_dispatch.svc_process_common.svc_process
     13.25 ±  6%     +26.2%      16.73 ±  1%  perf-profile.cycles-pp.nfsd4_write.nfsd4_proc_compound.nfsd_dispatch.svc_process_common.svc_process
     13.81 ±  1%     -24.0%      10.50 ±  1%  perf-profile.cycles-pp.nfsd_commit.nfsd4_commit.nfsd4_proc_compound.nfsd_dispatch.svc_process_common
     13.22 ±  6%     +26.3%      16.70 ±  1%  perf-profile.cycles-pp.nfsd_vfs_write.nfsd4_write.nfsd4_proc_compound.nfsd_dispatch.svc_process_common
      0.00 ± -1%      +Inf%       1.56 ±  3%  perf-profile.cycles-pp.osq_lock.rwsem_optimistic_spin.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write
      1.71 ±  3%     +17.8%       2.02 ±  4%  perf-profile.cycles-pp.rpc_async_schedule.process_one_work.worker_thread.kthread.ret_from_fork
      3.15 ±  7%    +121.0%       6.96 ±  2%  perf-profile.cycles-pp.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write.xfs_file_buffered_aio_write.xfs_file_write_iter
      3.15 ±  7%    +121.0%       6.96 ±  2%  perf-profile.cycles-pp.rwsem_optimistic_spin.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write.xfs_file_buffered_aio_write
      2.73 ±  6%     +93.7%       5.29 ±  2%  perf-profile.cycles-pp.rwsem_spin_on_owner.rwsem_optimistic_spin.rwsem_down_write_failed.call_rwsem_down_write_failed.down_write
      3.81 ±  6%     -15.5%       3.22 ±  3%  perf-profile.cycles-pp.skb_copy_datagram_iter.tcp_recvmsg.inet_recvmsg.sock_recvmsg.kernel_recvmsg
      1.19 ±  7%     +54.9%       1.85 ±  5%  perf-profile.cycles-pp.smp_apic_timer_interrupt.apic_timer_interrupt.cpuidle_enter.call_cpuidle.cpu_startup_entry
      4.66 ±  7%     -13.3%       4.04 ±  2%  perf-profile.cycles-pp.sock_recvmsg.kernel_recvmsg.svc_recvfrom.isra.9.svc_tcp_recvfrom.svc_recv
      7.84 ±  1%     -30.2%       5.47 ±  3%  perf-profile.cycles-pp.submit_bio.xfs_add_to_ioend.xfs_do_writepage.write_cache_pages.xfs_vm_writepages
      2.64 ±  2%     -26.0%       1.95 ±  2%  perf-profile.cycles-pp.submit_bio.xfs_submit_ioend.isra.17.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range
      4.67 ±  7%     -13.3%       4.05 ±  2%  perf-profile.cycles-pp.svc_recvfrom.isra.9.svc_tcp_recvfrom.svc_recv.nfsd.kthread
     10.51 ±  5%     -17.7%       8.65 ±  5%  perf-profile.cycles-pp.sys_write.entry_SYSCALL_64_fastpath
      4.64 ±  7%     -13.8%       4.00 ±  2%  perf-profile.cycles-pp.tcp_recvmsg.inet_recvmsg.sock_recvmsg.kernel_recvmsg.svc_recvfrom.isra.9
      0.96 ±  4%     +29.1%       1.24 ±  7%  perf-profile.cycles-pp.tcp_sendpage.inet_sendpage.xs_sendpages.xs_tcp_send_request.xprt_transmit
     13.81 ±  1%     -24.1%      10.48 ±  1%  perf-profile.cycles-pp.vfs_fsync_range.nfsd_commit.nfsd4_commit.nfsd4_proc_compound.nfsd_dispatch
     10.49 ±  5%     -17.9%       8.62 ±  5%  perf-profile.cycles-pp.vfs_write.sys_write.entry_SYSCALL_64_fastpath
     13.21 ±  6%     +26.3%      16.69 ±  1%  perf-profile.cycles-pp.vfs_writev.nfsd_vfs_write.nfsd4_write.nfsd4_proc_compound.nfsd_dispatch
     10.03 ±  1%     -25.1%       7.51 ±  1%  perf-profile.cycles-pp.write_cache_pages.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range
      8.03 ±  2%     -30.2%       5.61 ±  3%  perf-profile.cycles-pp.xfs_add_to_ioend.xfs_do_writepage.write_cache_pages.xfs_vm_writepages.do_writepages
      9.43 ±  1%     -26.4%       6.93 ±  2%  perf-profile.cycles-pp.xfs_do_writepage.write_cache_pages.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range
     13.17 ±  6%     +26.5%      16.65 ±  1%  perf-profile.cycles-pp.xfs_file_buffered_aio_write.xfs_file_write_iter.do_iter_readv_writev.do_readv_writev.vfs_writev
     13.80 ±  0%     -24.1%      10.48 ±  1%  perf-profile.cycles-pp.xfs_file_fsync.vfs_fsync_range.nfsd_commit.nfsd4_commit.nfsd4_proc_compound
     13.17 ±  6%     +26.5%      16.66 ±  1%  perf-profile.cycles-pp.xfs_file_write_iter.do_iter_readv_writev.do_readv_writev.vfs_writev.nfsd_vfs_write
      2.65 ±  2%     -26.0%       1.96 ±  2%  perf-profile.cycles-pp.xfs_submit_ioend.isra.17.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range
     12.68 ±  1%     -25.3%       9.47 ±  1%  perf-profile.cycles-pp.xfs_vm_writepages.do_writepages.__filemap_fdatawrite_range.filemap_write_and_wait_range.xfs_file_fsync
      1.26 ±  4%     +27.8%       1.61 ±  5%  perf-profile.cycles-pp.xprt_transmit.call_transmit.__rpc_execute.rpc_async_schedule.process_one_work
      1.24 ±  3%     +28.0%       1.59 ±  5%  perf-profile.cycles-pp.xs_sendpages.xs_tcp_send_request.xprt_transmit.call_transmit.__rpc_execute
      1.24 ±  3%     +28.4%       1.59 ±  5%  perf-profile.cycles-pp.xs_tcp_send_request.xprt_transmit.call_transmit.__rpc_execute.rpc_async_schedule






                             fsmark.time.system_time

  48 ++---------------------------------------------------------------------+
     |      *.            *                      *.           *..           |
  47 ++    +  *         .. +        .*         ..  *         +   *    .*.. .*
  46 ++  .*    +      .*    *. .*..*  +      .*     +      .*     + .*    * |
     *.*.       *..*.*        *        *..*.*        *.*..*        *        |
  45 ++                                                                     |
     |                                                                      |
  44 ++                                                                     |
     |                                                                      |
  43 ++                                                                     |
  42 ++                                                                     |
     |                                                        O             |
  41 ++         O  O O O  O O   O    O O  O O O  O O O O  O O    O          |
     O O  O   O               O    O                                        |
  40 ++-----O---------------------------------------------------------------+


                     fsmark.time.percent_of_cpu_this_job_got

  38 ++-----*-*-----------*----------*-----------*-*----------*--*-----*----*
     |     +   +        .. +        + +        ..   +        +    +   +    +|
  37 *+*..*     *..*.*.*    *.*.*..*   *..*.*.*      *.*..*.*      *.*    * |
  36 ++                                                                     |
     |                                                                      |
  35 ++                                                                     |
     |                                                                      |
  34 ++                                                                     |
     |                                                                      |
  33 ++                                                                     |
  32 ++                                                                     |
     |                                                                      |
  31 ++         O  O O O  O O O O  O O O  O O O  O O O O  O O O  O          |
     |                                                                      |
  30 O+O--O-O-O-------------------------------------------------------------+


                             fsmark.time.elapsed_time

  134 ++---------O----------------------------------------------------------+
      O O  O O O    O                                                       |
  132 ++                                                                    |
      |                                                     O O             |
      |               O O    O O O O  O O O O  O O O O  O O      O          |
  130 ++                  O                                                 |
      |                                                                     |
  128 ++                                                                    |
      |                                                                     |
  126 ++                                                                    |
      |                                                                     |
      | *..                                 *..           *.                |
  124 ++   *.*.          .*..     .*..*.   +   *.*.      +  *.*..   .*.*.. .*
      *        *.*..*.*.*    *.*.*      *.*        *.*..*        *.*      * |
  122 ++--------------------------------------------------------------------+


                           fsmark.time.elapsed_time.max

  134 ++---------O----------------------------------------------------------+
      O O  O O O    O                                                       |
  132 ++                                                                    |
      |                                                     O O             |
      |               O O    O O O O  O O O O  O O O O  O O      O          |
  130 ++                  O                                                 |
      |                                                                     |
  128 ++                                                                    |
      |                                                                     |
  126 ++                                                                    |
      |                                                                     |
      | *..                                 *..           *.                |
  124 ++   *.*.          .*..     .*..*.   +   *.*.      +  *.*..   .*.*.. .*
      *        *.*..*.*.*    *.*.*      *.*        *.*..*        *.*      * |
  122 ++--------------------------------------------------------------------+


                      fsmark.time.involuntary_context_switches

  50000 ++------------------------------------------------------------------+
        |                                                                   |
  45000 *+*.*..*.*.*.*.*..*.*.*.*.*..*.*.*.*.*.*..*.*.*.*.*..*.*.*.*.*..*.*.*
  40000 ++                                                                  |
        |                                                                   |
  35000 ++                                                                  |
        |                                                                   |
  30000 ++                                                                  |
        |                                                                   |
  25000 ++                                                                  |
  20000 ++                                                                  |
        |                                                                   |
  15000 ++  O  O   O O O  O O O O O  O O O O   O  O O   O O    O O          |
        O O      O                           O        O      O              |
  10000 ++------------------------------------------------------------------+



                              fsmark.files_per_sec

  84 ++---------------------------------------------------------------------+
     *.*..*.*.*.*..*.*.*..*.*.*.*..*.*.*..*.*.*..*.*.*.*..*.*.*..*.*.*.*..*.*
  83 ++                                                                     |
  82 ++                                                                     |
     |                                                                      |
  81 ++                                                                     |
     |                                                                      |
  80 ++                                                                     |
     |                                                                      |
  79 ++                                                                     |
  78 ++                                                                     |
     |                                                                      |
  77 ++                                                                     |
     O O  O O O O  O O O  O O O O  O O O  O O O  O O O O  O O O  O          |
  76 ++---------------------------------------------------------------------+


                                  fsmark.app_overhead

  1.6e+07 ++----------------------------------------------------------------+
          |            *                                                    |
  1.4e+07 ++           :                                                    |
          |           : :                                                   |
          |           : :                                                   |
  1.2e+07 ++         :  :                                                   |
          |.*.*.    .*   :  .*.          .*. .*.*.    .*. .*.*.   .*..   .*.|
    1e+07 *+    *..*     *.*   *.*..*.*.*   *     *..*   *     *.*    *.*   *
          |                                                                 |
    8e+06 ++                                                                |
          |                                                                 |
          |                                                                 |
    6e+06 ++                                                                |
          O O O O  O O O O O O O O  O O O O O O O O  O O O O O O O          |
    4e+06 ++----------------------------------------------------------------+


	[*] 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,
Xiaolong

View attachment "config-4.7.0-rc3-00003-g818e8e4" of type "text/plain" (150917 bytes)

View attachment "job.yaml" of type "text/plain" (3895 bytes)

View attachment "reproduce" of type "text/plain" (5914 bytes)

Powered by blists - more mailing lists