[linus:master] [xfs] cbc06310c3: filebench.sum_operations/s -40.3% regression

[linus:master] [xfs] cbc06310c3: filebench.sum_operations/s -40.3% regression

[kernel test robot]

Hello,

kernel test robot noticed a -40.3% regression of filebench.sum_operations/s on:


commit: cbc06310c36f73a5f3b0c6f0d974d60cf66d816b ("xfs: reinstate the old i_version counter as STATX_CHANGE_COOKIE")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

testcase: filebench
test machine: 96 threads 2 sockets Intel(R) Xeon(R) Platinum 8260L CPU @ 2.40GHz (Cascade Lake) with 128G memory
parameters:

	disk: 1HDD
	fs: xfs
	fs2: nfsv4
	test: filemicro_rread.f
	cpufreq_governor: performance




If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <oliver.sang@intel.com>
| Closes: https://lore.kernel.org/oe-lkp/202310201455.d9efae-oliver.sang@intel.com


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


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20231020/202310201455.d9efae-oliver.sang@intel.com

=========================================================================================
compiler/cpufreq_governor/disk/fs2/fs/kconfig/rootfs/tbox_group/test/testcase:
  gcc-12/performance/1HDD/nfsv4/xfs/x86_64-rhel-8.3/debian-11.1-x86_64-20220510.cgz/lkp-csl-2sp3/filemicro_rread.f/filebench

commit: 
  f93b930030 ("xfs: Remove duplicate include")
  cbc06310c3 ("xfs: reinstate the old i_version counter as STATX_CHANGE_COOKIE")

f93b9300301d30f2 cbc06310c36f73a5f3b0c6f0d97 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
      0.05 ±  3%      +0.0        0.05 ±  6%  mpstat.cpu.all.sys%
      1801 ±  3%    +192.8%       5272 ±  5%  vmstat.system.cs
     34136 ±  3%     +47.7%      50422 ±  5%  meminfo.Active
    264.26 ± 11%   +6281.8%      16864 ± 20%  meminfo.Active(file)
     22783 ±  7%   +1234.4%     304018 ±  7%  turbostat.C1
      0.03 ± 17%      +0.0        0.06 ± 11%  turbostat.C1%
     60181 ± 10%     +93.2%     116249 ±  7%  turbostat.POLL
     68.91            +1.0%      69.58        turbostat.PkgWatt
     64.00           -40.2%      38.24 ± 13%  filebench.sum_bytes_mb/s
     32765           -40.3%      19568 ± 13%  filebench.sum_operations/s
     32765           -40.3%      19568 ± 13%  filebench.sum_reads/s
      0.00 ± 33%   +1911.1%       0.03 ± 19%  filebench.sum_time_ms/op
     20275 ± 31%    +269.7%      74961 ±  8%  filebench.time.voluntary_context_switches
     65.23 ± 10%   +6225.9%       4126 ± 16%  proc-vmstat.nr_active_file
    517254            -1.5%     509753        proc-vmstat.nr_inactive_file
     65.23 ± 10%   +6225.9%       4126 ± 16%  proc-vmstat.nr_zone_active_file
    517254            -1.5%     509753        proc-vmstat.nr_zone_inactive_file
   1199999           +24.8%    1497636        proc-vmstat.numa_hit
     13948         +1878.7%     276000        proc-vmstat.pgactivate
   1767189           +17.2%    2070794        proc-vmstat.pgalloc_normal
   1733917           +17.5%    2037105        proc-vmstat.pgfree
 1.462e+08            +4.8%  1.532e+08        perf-stat.i.branch-instructions
      1683 ±  3%    +210.4%       5224 ±  5%  perf-stat.i.context-switches
 1.943e+08            +4.9%  2.038e+08        perf-stat.i.dTLB-loads
  98635751            +5.4%  1.039e+08        perf-stat.i.dTLB-stores
    177093           +18.5%     209883 ±  3%  perf-stat.i.iTLB-loads
 7.228e+08            +4.5%  7.552e+08        perf-stat.i.instructions
      3.91 ±  3%      +6.2%       4.15 ±  2%  perf-stat.i.metric.M/sec
     10.86            -0.5       10.34        perf-stat.overall.branch-miss-rate%
      6.91 ±  2%      -3.6%       6.66        perf-stat.overall.cpi
     80.11            -2.6       77.55        perf-stat.overall.iTLB-load-miss-rate%
      1012            +2.9%       1042 ±  2%  perf-stat.overall.instructions-per-iTLB-miss
 1.455e+08            +4.8%  1.524e+08        perf-stat.ps.branch-instructions
      1674 ±  3%    +209.9%       5188 ±  5%  perf-stat.ps.context-switches
 1.933e+08            +4.9%  2.028e+08        perf-stat.ps.dTLB-loads
  98135969            +5.3%  1.034e+08        perf-stat.ps.dTLB-stores
    176255           +18.5%     208779 ±  3%  perf-stat.ps.iTLB-loads
  7.19e+08            +4.5%  7.512e+08        perf-stat.ps.instructions




Disclaimer:
Results have been estimated based on internal Intel analysis and are provided
for informational purposes only. Any difference in system hardware or software
design or configuration may affect actual performance.


-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki