mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* [linus:master] [btrfs]  32c523c578:  reaim.jobs_per_min 16.1% regression
@ 2025-06-12  6:26 kernel test robot
  2025-06-12 13:17 ` Filipe Manana
  0 siblings, 1 reply; 2+ messages in thread
From: kernel test robot @ 2025-06-12  6:26 UTC (permalink / raw)
  To: Filipe Manana
  Cc: oe-lkp, lkp, linux-kernel, David Sterba, linux-btrfs, oliver.sang



Hello,

kernel test robot noticed a 16.1% regression of reaim.jobs_per_min on:


commit: 32c523c578e8489f55663ce8a8860079c8deb414 ("btrfs: allow folios to be released while ordered extent is finishing")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

[still regression on linus/master      5abc7438f1e9d62e91ad775cc83c9594c48d2282]
[still regression on linux-next/master 911483b25612c8bc32a706ba940738cc43299496]

testcase: reaim
config: x86_64-rhel-9.4
compiler: gcc-12
test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
parameters:

	runtime: 300s
	nr_task: 100%
	disk: 1HDD
	fs: btrfs
	test: disk
	cpufreq_governor: performance


In addition to that, the commit also has significant impact on the following tests:

+------------------+-------------------------------------------------------------------------------------------+
| testcase: change | reaim: reaim.jobs_per_min  5.3% regression                                                |
| test machine     | 8 threads 1 sockets Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (Ivy Bridge) with 16G memory |
| test parameters  | cpufreq_governor=performance                                                              |
|                  | disk=1HDD                                                                                 |
|                  | fs=btrfs                                                                                  |
|                  | nr_task=100%                                                                              |
|                  | runtime=300s                                                                              |
|                  | test=disk                                                                                 |
+------------------+-------------------------------------------------------------------------------------------+


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/202506121329.9d1f4d50-lkp@intel.com


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


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250612/202506121329.9d1f4d50-lkp@intel.com

=========================================================================================
compiler/cpufreq_governor/disk/fs/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase:
  gcc-12/performance/1HDD/btrfs/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/300s/lkp-icl-2sp7/disk/reaim

commit: 
  cbfb4cbf45 ("btrfs: update comment for try_release_extent_state()")
  32c523c578 ("btrfs: allow folios to be released while ordered extent is finishing")

cbfb4cbf459d9be4 32c523c578e8489f55663ce8a88 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
     71009           -14.7%      60593        meminfo.Shmem
 2.344e+10           -10.8%   2.09e+10        cpuidle..time
   6491212           -20.8%    5141230        cpuidle..usage
     93.14            -1.5%      91.75        iostat.cpu.idle
      5.81           +24.8%       7.24        iostat.cpu.iowait
    419.83            -9.5%     380.03        uptime.boot
     24988           -10.7%      22325        uptime.idle
    584.67 ±  3%     -23.6%     446.83 ±  5%  perf-c2c.DRAM.remote
    551.00 ± 10%     -19.1%     445.50 ±  9%  perf-c2c.HITM.local
    407.83 ±  5%     -25.4%     304.17 ±  6%  perf-c2c.HITM.remote
      5.84            +1.5        7.29        mpstat.cpu.all.iowait%
      0.09            -0.0        0.07        mpstat.cpu.all.irq%
      0.53            -0.1        0.46        mpstat.cpu.all.sys%
      0.36 ±  2%      +0.0        0.40 ±  4%  mpstat.cpu.all.usr%
   2110659 ±  3%     -22.4%    1638521 ±  3%  numa-numastat.node0.local_node
   2132066 ±  3%     -21.5%    1673684 ±  3%  numa-numastat.node0.numa_hit
   2054843 ±  4%     -16.8%    1709588 ±  2%  numa-numastat.node1.local_node
   2099801 ±  3%     -17.1%    1740752 ±  2%  numa-numastat.node1.numa_hit
     61.66          +186.4%     176.60        vmstat.io.bi
     17553           -16.2%      14712        vmstat.io.bo
      3.91 ±  3%     +25.7%       4.92 ±  4%  vmstat.procs.b
      9668           -12.3%       8476        vmstat.system.cs
     13868           -10.9%      12357        vmstat.system.in
    521965 ±  4%     -27.8%     376885 ±  2%  numa-vmstat.node0.nr_dirtied
    500007 ±  4%     -27.9%     360465 ±  2%  numa-vmstat.node0.nr_written
   2131472 ±  3%     -21.5%    1673240 ±  3%  numa-vmstat.node0.numa_hit
   2110065 ±  3%     -22.4%    1638077 ±  3%  numa-vmstat.node0.numa_local
    423228 ±  4%     -21.9%     330636 ±  4%  numa-vmstat.node1.nr_dirtied
    405391 ±  4%     -21.9%     316667 ±  4%  numa-vmstat.node1.nr_written
   2099495 ±  4%     -17.1%    1739892 ±  2%  numa-vmstat.node1.numa_hit
   2054537 ±  4%     -16.8%    1708728 ±  2%  numa-vmstat.node1.numa_local
      4262           -16.1%       3575        reaim.jobs_per_min
     66.61           -16.1%      55.87        reaim.jobs_per_min_child
      4305           -15.6%       3635        reaim.max_jobs_per_min
     90.09           +19.2%     107.42        reaim.parent_time
      1.07 ±  4%     -15.8%       0.90 ±  3%  reaim.std_dev_percent
    369.06           -10.9%     328.84        reaim.time.elapsed_time
    369.06           -10.9%     328.84        reaim.time.elapsed_time.max
     45840          +155.6%     117178        reaim.time.file_system_inputs
   5943434           -25.9%    4401353        reaim.time.file_system_outputs
     18556           -29.3%      13119        reaim.time.involuntary_context_switches
   3738859           -25.0%    2804459        reaim.time.minor_page_faults
     28.83           -16.8%      24.00        reaim.time.percent_of_cpu_this_job_got
     93.94           -24.8%      70.62        reaim.time.system_time
   1061461           -24.9%     797506        reaim.time.voluntary_context_switches
     25600           -25.0%      19200        reaim.workload
    187046            -1.3%     184574        proc-vmstat.nr_active_anon
    945195           -25.1%     707517        proc-vmstat.nr_dirtied
   1133127            -1.6%    1115274        proc-vmstat.nr_file_pages
    214899            -7.1%     199637        proc-vmstat.nr_inactive_file
     17745           -14.7%      15145        proc-vmstat.nr_shmem
    905398           -25.2%     677133        proc-vmstat.nr_written
    187046            -1.3%     184574        proc-vmstat.nr_zone_active_anon
    214899            -7.1%     199637        proc-vmstat.nr_zone_inactive_file
   4233781           -19.3%    3415500        proc-vmstat.numa_hit
   4167415           -19.6%    3349180        proc-vmstat.numa_local
   4401990           -19.1%    3559895        proc-vmstat.pgalloc_normal
   4837041           -21.2%    3813117        proc-vmstat.pgfault
   4058646           -20.7%    3220351        proc-vmstat.pgfree
     22920          +155.6%      58589        proc-vmstat.pgpgin
   6526981           -25.2%    4881946        proc-vmstat.pgpgout
     48196            -8.6%      44072 ±  3%  proc-vmstat.pgreuse
      2.95            +0.1        3.06        perf-stat.i.branch-miss-rate%
   9010149            +3.5%    9323057        perf-stat.i.branch-misses
  15272312            -6.6%   14260618        perf-stat.i.cache-references
      9701           -12.3%       8509        perf-stat.i.context-switches
      1.97            -1.7%       1.94        perf-stat.i.cpi
 2.028e+09            -3.4%  1.958e+09        perf-stat.i.cpu-cycles
    345.48           -12.7%     301.68        perf-stat.i.cpu-migrations
      1061            +2.2%       1083        perf-stat.i.cycles-between-cache-misses
      0.53            +2.0%       0.54        perf-stat.i.ipc
      3.86 ±  2%     -20.1%       3.09 ±  3%  perf-stat.i.major-faults
     12520           -11.9%      11027        perf-stat.i.minor-faults
     12524           -11.9%      11030        perf-stat.i.page-faults
      3.66            +0.1        3.81        perf-stat.overall.branch-miss-rate%
     16.24 ±  5%      +1.2       17.46 ±  6%  perf-stat.overall.cache-miss-rate%
      1.64            -3.2%       1.59        perf-stat.overall.cpi
      0.61            +3.3%       0.63        perf-stat.overall.ipc
  17792434           +18.5%   21081619        perf-stat.overall.path-length
   8979944            +3.5%    9292382        perf-stat.ps.branch-misses
  15229067            -6.6%   14216663        perf-stat.ps.cache-references
      9674           -12.3%       8483        perf-stat.ps.context-switches
 2.022e+09            -3.5%  1.953e+09        perf-stat.ps.cpu-cycles
    344.62           -12.7%     300.82        perf-stat.ps.cpu-migrations
      3.85 ±  2%     -20.2%       3.08 ±  4%  perf-stat.ps.major-faults
     12486           -12.0%      10993        perf-stat.ps.minor-faults
     12490           -12.0%      10996        perf-stat.ps.page-faults
 4.555e+11           -11.1%  4.048e+11        perf-stat.total.instructions
     24864 ±  2%     -18.6%      20241 ±  6%  sched_debug.cfs_rq:/.avg_vruntime.avg
      9646 ±  3%     -32.1%       6545 ± 11%  sched_debug.cfs_rq:/.avg_vruntime.min
     88.62 ±  7%     +17.8%     104.42 ±  6%  sched_debug.cfs_rq:/.load_avg.avg
    165.70 ±  8%     +21.4%     201.23 ±  6%  sched_debug.cfs_rq:/.load_avg.stddev
     24864 ±  2%     -18.6%      20241 ±  6%  sched_debug.cfs_rq:/.min_vruntime.avg
      9646 ±  3%     -32.1%       6545 ± 11%  sched_debug.cfs_rq:/.min_vruntime.min
      7.59 ± 18%     +40.4%      10.66 ±  7%  sched_debug.cfs_rq:/.util_est.avg
    197.81 ± 13%     +32.3%     261.71 ± 17%  sched_debug.cfs_rq:/.util_est.max
     31.28 ± 12%     +30.3%      40.77 ± 10%  sched_debug.cfs_rq:/.util_est.stddev
    229863           -17.2%     190282 ±  7%  sched_debug.cpu.clock.avg
    229868           -17.2%     190287 ±  7%  sched_debug.cpu.clock.max
    229857           -17.2%     190276 ±  7%  sched_debug.cpu.clock.min
    229303           -17.2%     189808 ±  7%  sched_debug.cpu.clock_task.avg
    229667           -17.2%     190143 ±  7%  sched_debug.cpu.clock_task.max
    219476           -17.8%     180349 ±  7%  sched_debug.cpu.clock_task.min
    749.88 ± 29%     -28.8%     533.95 ± 14%  sched_debug.cpu.curr->pid.avg
     32645           -29.7%      22956 ±  8%  sched_debug.cpu.curr->pid.max
      4481 ± 10%     -29.9%       3143 ± 10%  sched_debug.cpu.curr->pid.stddev
     29063           -30.0%      20341 ±  9%  sched_debug.cpu.nr_switches.avg
     83586 ±  2%     -29.8%      58716 ±  7%  sched_debug.cpu.nr_switches.max
     22149 ±  2%     -33.0%      14831 ± 11%  sched_debug.cpu.nr_switches.min
      7977 ±  2%     -25.6%       5937 ±  5%  sched_debug.cpu.nr_switches.stddev
     92.43 ± 28%     -47.8%      48.21 ± 10%  sched_debug.cpu.nr_uninterruptible.max
     25.24 ±  8%     -26.2%      18.62 ±  9%  sched_debug.cpu.nr_uninterruptible.stddev
    229859           -17.2%     190278 ±  7%  sched_debug.cpu_clk
    229148           -17.3%     189567 ±  7%  sched_debug.ktime
    230484           -17.2%     190908 ±  7%  sched_debug.sched_clk
      0.00 ±136%    +392.6%       0.02 ± 80%  perf-sched.sch_delay.avg.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
      0.03 ±114%     -56.9%       0.01 ±  5%  perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_read_slowpath.down_read.btrfs_tree_read_lock_nested
      0.25 ±209%     -95.1%       0.01 ± 13%  perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.open_last_lookups
      0.21 ± 40%     -38.8%       0.13 ± 11%  perf-sched.sch_delay.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
      0.07 ± 42%    +126.0%       0.16 ± 72%  perf-sched.sch_delay.max.ms.wait_extent_bit.__lock_extent.lock_extents_for_read.constprop.0
     26708 ±  2%      -9.5%      24158 ±  3%  perf-sched.total_wait_and_delay.count.ms
      3.20 ±  9%     +25.5%       4.02 ± 11%  perf-sched.wait_and_delay.avg.ms.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
      7.26 ± 15%     +50.7%      10.94 ± 19%  perf-sched.wait_and_delay.avg.ms.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
     19.16 ± 29%     -65.5%       6.61 ±142%  perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.btrfs_wait_ordered_roots
      0.02 ± 59%    +188.1%       0.05 ± 63%  perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.sync_inodes_sb
     15.73 ± 19%     +50.8%      23.72 ± 10%  perf-sched.wait_and_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
    256.17 ± 15%     +34.5%     344.47 ± 13%  perf-sched.wait_and_delay.avg.ms.wait_current_trans.start_transaction.btrfs_create_common.lookup_open.isra
     78.54 ±  8%     +18.1%      92.78 ± 11%  perf-sched.wait_and_delay.avg.ms.wait_for_commit.btrfs_commit_transaction.iterate_supers.ksys_sync
     67.31 ±  3%     +12.5%      75.72 ±  5%  perf-sched.wait_and_delay.avg.ms.wait_for_commit.btrfs_wait_for_commit.btrfs_attach_transaction_barrier.btrfs_sync_fs
    805.00 ± 11%     -21.9%     629.00 ± 11%  perf-sched.wait_and_delay.count.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
    833.33 ±  5%     -29.6%     586.50 ± 13%  perf-sched.wait_and_delay.count.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
      2990 ±  5%     -16.5%       2498 ±  3%  perf-sched.wait_and_delay.count.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
    151.17 ±  6%     -70.6%      44.50 ±141%  perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.btrfs_wait_ordered_roots
    553.50 ±  5%     -26.3%     407.67 ±  9%  perf-sched.wait_and_delay.count.schedule_timeout.__wait_for_common.btrfs_wait_ordered_extents.btrfs_wait_ordered_roots
    618.50 ±  9%     -25.4%     461.17 ± 12%  perf-sched.wait_and_delay.count.wait_current_trans.start_transaction.btrfs_sync_file.btrfs_do_write_iter
    955.00 ±  8%     -27.5%     692.17 ±  8%  perf-sched.wait_and_delay.count.wait_log_commit.btrfs_sync_log.btrfs_sync_file.btrfs_do_write_iter
    551.40 ± 10%     +33.8%     737.66 ±  8%  perf-sched.wait_and_delay.max.ms.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
    251.59 ± 18%     +74.5%     438.99 ± 14%  perf-sched.wait_and_delay.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
     11.70 ± 99%    -100.0%       0.00 ±223%  perf-sched.wait_time.avg.ms.__cond_resched.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range
     44.61 ±189%    +762.7%     384.84 ± 81%  perf-sched.wait_time.avg.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
      3.19 ±  9%     +25.6%       4.00 ± 11%  perf-sched.wait_time.avg.ms.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
      7.24 ± 15%     +50.9%      10.93 ± 19%  perf-sched.wait_time.avg.ms.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
      1.69 ± 79%    +178.9%       4.71 ± 62%  perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.extent_write_cache_pages.btrfs_writepages
     15.72 ± 19%     +50.9%      23.71 ± 10%  perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
     20.99 ±  2%     +38.2%      29.02 ± 15%  perf-sched.wait_time.avg.ms.schedule_timeout.btrfs_sync_log.btrfs_sync_file.btrfs_do_write_iter
     12.97 ± 13%     +49.0%      19.32 ± 16%  perf-sched.wait_time.avg.ms.schedule_timeout.io_schedule_timeout.__wait_for_common.barrier_all_devices
    256.15 ± 15%     +34.5%     344.45 ± 13%  perf-sched.wait_time.avg.ms.wait_current_trans.start_transaction.btrfs_create_common.lookup_open.isra
     10.94 ± 25%     +84.6%      20.19 ± 24%  perf-sched.wait_time.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
     78.51 ±  8%     +18.1%      92.75 ± 11%  perf-sched.wait_time.avg.ms.wait_for_commit.btrfs_commit_transaction.iterate_supers.ksys_sync
     67.26 ±  3%     +12.5%      75.66 ±  5%  perf-sched.wait_time.avg.ms.wait_for_commit.btrfs_wait_for_commit.btrfs_attach_transaction_barrier.btrfs_sync_fs
     19.50 ±128%    -100.0%       0.00 ±223%  perf-sched.wait_time.max.ms.__cond_resched.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range
     83.75 ±187%    +516.6%     516.38 ± 68%  perf-sched.wait_time.max.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
    551.39 ± 10%     +33.8%     737.64 ±  8%  perf-sched.wait_time.max.ms.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
    251.57 ± 18%     +74.5%     438.95 ± 14%  perf-sched.wait_time.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
     86.74 ± 14%     -27.4%      62.97 ± 11%  perf-sched.wait_time.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range


***************************************************************************************************
lkp-ivb-d01: 8 threads 1 sockets Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (Ivy Bridge) with 16G memory
=========================================================================================
compiler/cpufreq_governor/disk/fs/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase:
  gcc-12/performance/1HDD/btrfs/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/300s/lkp-ivb-d01/disk/reaim

commit: 
  cbfb4cbf45 ("btrfs: update comment for try_release_extent_state()")
  32c523c578 ("btrfs: allow folios to be released while ordered extent is finishing")

cbfb4cbf459d9be4 32c523c578e8489f55663ce8a88 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
     22777          +101.6%      45913 ± 13%  proc-vmstat.pgpgin
     12134 ±  4%     +18.9%      14426 ±  7%  sched_debug.cpu.nr_switches.stddev
     80.30            -1.8%      78.89        iostat.cpu.idle
     17.77            +8.1%      19.21        iostat.cpu.iowait
     54.02          +101.9%     109.06        vmstat.io.bi
      4061            -5.0%       3856        vmstat.io.bo
      2.60            +2.2%       2.65        reaim.child_systime
    348.91            -5.3%     330.27        reaim.jobs_per_min
     43.61            -5.3%      41.28        reaim.jobs_per_min_child
    356.27            -5.4%     336.88        reaim.max_jobs_per_min
    137.64            +5.6%     145.41        reaim.parent_time
     45554          +101.6%      91826 ± 13%  reaim.time.file_system_inputs
      0.17 ± 25%      -0.1        0.09 ± 31%  perf-profile.children.cycles-pp.select_task_rq_fair
      0.16 ± 29%      -0.1        0.10 ± 33%  perf-profile.children.cycles-pp.vfs_fstatat
      0.21 ± 16%      -0.1        0.15 ± 12%  perf-profile.children.cycles-pp.___perf_sw_event
      0.06 ± 79%      +0.1        0.16 ± 33%  perf-profile.children.cycles-pp.vms_complete_munmap_vmas
      0.07 ± 72%      +0.1        0.18 ± 16%  perf-profile.children.cycles-pp.vms_clear_ptes
      0.19 ± 18%      -0.1        0.12 ± 19%  perf-profile.self.cycles-pp.___perf_sw_event
      0.15 ± 30%      +0.1        0.23 ± 33%  perf-profile.self.cycles-pp.x86_pmu_disable
      0.04 ±  9%     +21.8%       0.05 ± 11%  perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.barrier_all_devices.write_all_supers.btrfs_sync_log
      0.05 ±  4%     +18.7%       0.06 ±  4%  perf-sched.sch_delay.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
      0.04 ± 69%    -100.0%       0.00        perf-sched.sch_delay.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
      0.05 ± 69%    -100.0%       0.00        perf-sched.sch_delay.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
     19.30 ±  9%     +17.0%      22.59 ±  6%  perf-sched.wait_and_delay.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
     19.25 ±  9%     +17.0%      22.53 ±  6%  perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
     15.74 ±107%    -100.0%       0.00        perf-sched.wait_time.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
     22.16 ± 79%    -100.0%       0.00        perf-sched.wait_time.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
      4.44            +1.9%       4.53        perf-stat.i.MPKI
      8.16            +0.2        8.41        perf-stat.i.branch-miss-rate%
      2313            +1.6%       2350        perf-stat.i.context-switches
      2.25            +1.9%       2.29        perf-stat.i.cpi
     73.48 ±  3%      -6.6%      68.61        perf-stat.i.cpu-migrations
      1972            -2.1%       1931        perf-stat.i.minor-faults
      1972            -2.1%       1931        perf-stat.i.page-faults
      2308            +1.6%       2344        perf-stat.ps.context-switches
     73.31 ±  3%      -6.6%      68.45        perf-stat.ps.cpu-migrations
      1967            -2.1%       1926        perf-stat.ps.minor-faults
      1968            -2.1%       1926        perf-stat.ps.page-faults





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


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [linus:master] [btrfs] 32c523c578: reaim.jobs_per_min 16.1% regression
  2025-06-12  6:26 [linus:master] [btrfs] 32c523c578: reaim.jobs_per_min 16.1% regression kernel test robot
@ 2025-06-12 13:17 ` Filipe Manana
  0 siblings, 0 replies; 2+ messages in thread
From: Filipe Manana @ 2025-06-12 13:17 UTC (permalink / raw)
  To: kernel test robot
  Cc: Filipe Manana, oe-lkp, lkp, linux-kernel, David Sterba, linux-btrfs

On Thu, Jun 12, 2025 at 7:26 AM kernel test robot <oliver.sang@intel.com> wrote:
>
>
>
> Hello,
>
> kernel test robot noticed a 16.1% regression of reaim.jobs_per_min on:
>
>
> commit: 32c523c578e8489f55663ce8a8860079c8deb414 ("btrfs: allow folios to be released while ordered extent is finishing")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
>
> [still regression on linus/master      5abc7438f1e9d62e91ad775cc83c9594c48d2282]
> [still regression on linux-next/master 911483b25612c8bc32a706ba940738cc43299496]
>
> testcase: reaim
> config: x86_64-rhel-9.4
> compiler: gcc-12
> test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
> parameters:
>
>         runtime: 300s
>         nr_task: 100%
>         disk: 1HDD
>         fs: btrfs
>         test: disk
>         cpufreq_governor: performance
>
>
> In addition to that, the commit also has significant impact on the following tests:
>
> +------------------+-------------------------------------------------------------------------------------------+
> | testcase: change | reaim: reaim.jobs_per_min  5.3% regression                                                |

Here it says 5.3% regression instead of 16.1%.

The phrasing right above suggests it's a different test, but it's
confusing since there's nothing here in the table description (test
parameters) that differs from the test details above.
It's the same test but on a different machine IIUC.

> | test machine     | 8 threads 1 sockets Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (Ivy Bridge) with 16G memory |
> | test parameters  | cpufreq_governor=performance                                                              |
> |                  | disk=1HDD                                                                                 |
> |                  | fs=btrfs                                                                                  |
> |                  | nr_task=100%                                                                              |
> |                  | runtime=300s                                                                              |
> |                  | test=disk                                                                                 |
> +------------------+-------------------------------------------------------------------------------------------+
>
>
> 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/202506121329.9d1f4d50-lkp@intel.com
>
>
> Details are as below:
> -------------------------------------------------------------------------------------------------->
>
>
> The kernel config and materials to reproduce are available at:
> https://download.01.org/0day-ci/archive/20250612/202506121329.9d1f4d50-lkp@intel.com

Following the instructions from the 'reproduce' file, it fails to
install the job:

fdmanana 13:28:49 ~/git/hub/lkp-tests (master)> sudo ./bin/lkp install job.yaml
distro=debian
version=13
Use: /home/fdmanana/git/hub/lkp-tests/distro/installer/debian install
bc debianutils gawk gzip kmod numactl rsync time automake bison
build-essential bzip2 ca-certificates cpio fakeroot flex git
libarchive-tools libc6-dev:i386 libc6-dev-x32 libipc-run-perl
libklibc-dev libssl-dev libtool linux-cpupower linux-libc-dev:i386
linux-perf openssl patch rsync ruby ruby-dev wget
Hit:1 http://deb.debian.org/debian testing InRelease
Reading package lists... Done
Reading package lists...
Building dependency tree...
Reading state information...
(...)
Reading state information...
E: Unable to locate package libpython2
E: Unable to locate package libpython3
Cannot install some packages of perf-c2c depends
fdmanana 13:29:05 ~/git/hub/lkp-tests (master)>

Alternatively, tried to run reaim directly following the script from
https://download.01.org/0day-ci/archive/20250612/202506121329.9d1f4d50-lkp@intel.com/repro-script

The problem is I can't find debian packages for reaim, and going to
sourceforge to get the source, and the latest is
osdl-aim-7.0.1.13.tar.gz (from 2004!), it doesn't compile, there are
tons on warnings and errors.
Applying the patch from lkp at
lkp-tests/programs/reaim/pkg/reaim.patch doesn't help either.




>
> =========================================================================================
> compiler/cpufreq_governor/disk/fs/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase:
>   gcc-12/performance/1HDD/btrfs/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/300s/lkp-icl-2sp7/disk/reaim
>
> commit:
>   cbfb4cbf45 ("btrfs: update comment for try_release_extent_state()")
>   32c523c578 ("btrfs: allow folios to be released while ordered extent is finishing")
>
> cbfb4cbf459d9be4 32c523c578e8489f55663ce8a88
> ---------------- ---------------------------
>          %stddev     %change         %stddev
>              \          |                \
>      71009           -14.7%      60593        meminfo.Shmem
>  2.344e+10           -10.8%   2.09e+10        cpuidle..time
>    6491212           -20.8%    5141230        cpuidle..usage
>      93.14            -1.5%      91.75        iostat.cpu.idle
>       5.81           +24.8%       7.24        iostat.cpu.iowait
>     419.83            -9.5%     380.03        uptime.boot
>      24988           -10.7%      22325        uptime.idle
>     584.67 ą  3%     -23.6%     446.83 ą  5%  perf-c2c.DRAM.remote
>     551.00 ą 10%     -19.1%     445.50 ą  9%  perf-c2c.HITM.local
>     407.83 ą  5%     -25.4%     304.17 ą  6%  perf-c2c.HITM.remote
>       5.84            +1.5        7.29        mpstat.cpu.all.iowait%
>       0.09            -0.0        0.07        mpstat.cpu.all.irq%
>       0.53            -0.1        0.46        mpstat.cpu.all.sys%
>       0.36 ą  2%      +0.0        0.40 ą  4%  mpstat.cpu.all.usr%
>    2110659 ą  3%     -22.4%    1638521 ą  3%  numa-numastat.node0.local_node
>    2132066 ą  3%     -21.5%    1673684 ą  3%  numa-numastat.node0.numa_hit
>    2054843 ą  4%     -16.8%    1709588 ą  2%  numa-numastat.node1.local_node
>    2099801 ą  3%     -17.1%    1740752 ą  2%  numa-numastat.node1.numa_hit
>      61.66          +186.4%     176.60        vmstat.io.bi
>      17553           -16.2%      14712        vmstat.io.bo
>       3.91 ą  3%     +25.7%       4.92 ą  4%  vmstat.procs.b
>       9668           -12.3%       8476        vmstat.system.cs
>      13868           -10.9%      12357        vmstat.system.in
>     521965 ą  4%     -27.8%     376885 ą  2%  numa-vmstat.node0.nr_dirtied
>     500007 ą  4%     -27.9%     360465 ą  2%  numa-vmstat.node0.nr_written
>    2131472 ą  3%     -21.5%    1673240 ą  3%  numa-vmstat.node0.numa_hit
>    2110065 ą  3%     -22.4%    1638077 ą  3%  numa-vmstat.node0.numa_local
>     423228 ą  4%     -21.9%     330636 ą  4%  numa-vmstat.node1.nr_dirtied
>     405391 ą  4%     -21.9%     316667 ą  4%  numa-vmstat.node1.nr_written
>    2099495 ą  4%     -17.1%    1739892 ą  2%  numa-vmstat.node1.numa_hit
>    2054537 ą  4%     -16.8%    1708728 ą  2%  numa-vmstat.node1.numa_local
>       4262           -16.1%       3575        reaim.jobs_per_min
>      66.61           -16.1%      55.87        reaim.jobs_per_min_child
>       4305           -15.6%       3635        reaim.max_jobs_per_min
>      90.09           +19.2%     107.42        reaim.parent_time
>       1.07 ą  4%     -15.8%       0.90 ą  3%  reaim.std_dev_percent
>     369.06           -10.9%     328.84        reaim.time.elapsed_time
>     369.06           -10.9%     328.84        reaim.time.elapsed_time.max
>      45840          +155.6%     117178        reaim.time.file_system_inputs
>    5943434           -25.9%    4401353        reaim.time.file_system_outputs
>      18556           -29.3%      13119        reaim.time.involuntary_context_switches
>    3738859           -25.0%    2804459        reaim.time.minor_page_faults
>      28.83           -16.8%      24.00        reaim.time.percent_of_cpu_this_job_got
>      93.94           -24.8%      70.62        reaim.time.system_time
>    1061461           -24.9%     797506        reaim.time.voluntary_context_switches
>      25600           -25.0%      19200        reaim.workload
>     187046            -1.3%     184574        proc-vmstat.nr_active_anon
>     945195           -25.1%     707517        proc-vmstat.nr_dirtied
>    1133127            -1.6%    1115274        proc-vmstat.nr_file_pages
>     214899            -7.1%     199637        proc-vmstat.nr_inactive_file
>      17745           -14.7%      15145        proc-vmstat.nr_shmem
>     905398           -25.2%     677133        proc-vmstat.nr_written
>     187046            -1.3%     184574        proc-vmstat.nr_zone_active_anon
>     214899            -7.1%     199637        proc-vmstat.nr_zone_inactive_file
>    4233781           -19.3%    3415500        proc-vmstat.numa_hit
>    4167415           -19.6%    3349180        proc-vmstat.numa_local
>    4401990           -19.1%    3559895        proc-vmstat.pgalloc_normal
>    4837041           -21.2%    3813117        proc-vmstat.pgfault
>    4058646           -20.7%    3220351        proc-vmstat.pgfree
>      22920          +155.6%      58589        proc-vmstat.pgpgin
>    6526981           -25.2%    4881946        proc-vmstat.pgpgout
>      48196            -8.6%      44072 ą  3%  proc-vmstat.pgreuse
>       2.95            +0.1        3.06        perf-stat.i.branch-miss-rate%
>    9010149            +3.5%    9323057        perf-stat.i.branch-misses
>   15272312            -6.6%   14260618        perf-stat.i.cache-references
>       9701           -12.3%       8509        perf-stat.i.context-switches
>       1.97            -1.7%       1.94        perf-stat.i.cpi
>  2.028e+09            -3.4%  1.958e+09        perf-stat.i.cpu-cycles
>     345.48           -12.7%     301.68        perf-stat.i.cpu-migrations
>       1061            +2.2%       1083        perf-stat.i.cycles-between-cache-misses
>       0.53            +2.0%       0.54        perf-stat.i.ipc
>       3.86 ą  2%     -20.1%       3.09 ą  3%  perf-stat.i.major-faults
>      12520           -11.9%      11027        perf-stat.i.minor-faults
>      12524           -11.9%      11030        perf-stat.i.page-faults
>       3.66            +0.1        3.81        perf-stat.overall.branch-miss-rate%
>      16.24 ą  5%      +1.2       17.46 ą  6%  perf-stat.overall.cache-miss-rate%
>       1.64            -3.2%       1.59        perf-stat.overall.cpi
>       0.61            +3.3%       0.63        perf-stat.overall.ipc
>   17792434           +18.5%   21081619        perf-stat.overall.path-length
>    8979944            +3.5%    9292382        perf-stat.ps.branch-misses
>   15229067            -6.6%   14216663        perf-stat.ps.cache-references
>       9674           -12.3%       8483        perf-stat.ps.context-switches
>  2.022e+09            -3.5%  1.953e+09        perf-stat.ps.cpu-cycles
>     344.62           -12.7%     300.82        perf-stat.ps.cpu-migrations
>       3.85 ą  2%     -20.2%       3.08 ą  4%  perf-stat.ps.major-faults
>      12486           -12.0%      10993        perf-stat.ps.minor-faults
>      12490           -12.0%      10996        perf-stat.ps.page-faults
>  4.555e+11           -11.1%  4.048e+11        perf-stat.total.instructions
>      24864 ą  2%     -18.6%      20241 ą  6%  sched_debug.cfs_rq:/.avg_vruntime.avg
>       9646 ą  3%     -32.1%       6545 ą 11%  sched_debug.cfs_rq:/.avg_vruntime.min
>      88.62 ą  7%     +17.8%     104.42 ą  6%  sched_debug.cfs_rq:/.load_avg.avg
>     165.70 ą  8%     +21.4%     201.23 ą  6%  sched_debug.cfs_rq:/.load_avg.stddev
>      24864 ą  2%     -18.6%      20241 ą  6%  sched_debug.cfs_rq:/.min_vruntime.avg
>       9646 ą  3%     -32.1%       6545 ą 11%  sched_debug.cfs_rq:/.min_vruntime.min
>       7.59 ą 18%     +40.4%      10.66 ą  7%  sched_debug.cfs_rq:/.util_est.avg
>     197.81 ą 13%     +32.3%     261.71 ą 17%  sched_debug.cfs_rq:/.util_est.max
>      31.28 ą 12%     +30.3%      40.77 ą 10%  sched_debug.cfs_rq:/.util_est.stddev
>     229863           -17.2%     190282 ą  7%  sched_debug.cpu.clock.avg
>     229868           -17.2%     190287 ą  7%  sched_debug.cpu.clock.max
>     229857           -17.2%     190276 ą  7%  sched_debug.cpu.clock.min
>     229303           -17.2%     189808 ą  7%  sched_debug.cpu.clock_task.avg
>     229667           -17.2%     190143 ą  7%  sched_debug.cpu.clock_task.max
>     219476           -17.8%     180349 ą  7%  sched_debug.cpu.clock_task.min
>     749.88 ą 29%     -28.8%     533.95 ą 14%  sched_debug.cpu.curr->pid.avg
>      32645           -29.7%      22956 ą  8%  sched_debug.cpu.curr->pid.max
>       4481 ą 10%     -29.9%       3143 ą 10%  sched_debug.cpu.curr->pid.stddev
>      29063           -30.0%      20341 ą  9%  sched_debug.cpu.nr_switches.avg
>      83586 ą  2%     -29.8%      58716 ą  7%  sched_debug.cpu.nr_switches.max
>      22149 ą  2%     -33.0%      14831 ą 11%  sched_debug.cpu.nr_switches.min
>       7977 ą  2%     -25.6%       5937 ą  5%  sched_debug.cpu.nr_switches.stddev
>      92.43 ą 28%     -47.8%      48.21 ą 10%  sched_debug.cpu.nr_uninterruptible.max
>      25.24 ą  8%     -26.2%      18.62 ą  9%  sched_debug.cpu.nr_uninterruptible.stddev
>     229859           -17.2%     190278 ą  7%  sched_debug.cpu_clk
>     229148           -17.3%     189567 ą  7%  sched_debug.ktime
>     230484           -17.2%     190908 ą  7%  sched_debug.sched_clk
>       0.00 ą136%    +392.6%       0.02 ą 80%  perf-sched.sch_delay.avg.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
>       0.03 ą114%     -56.9%       0.01 ą  5%  perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_read_slowpath.down_read.btrfs_tree_read_lock_nested
>       0.25 ą209%     -95.1%       0.01 ą 13%  perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.open_last_lookups
>       0.21 ą 40%     -38.8%       0.13 ą 11%  perf-sched.sch_delay.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
>       0.07 ą 42%    +126.0%       0.16 ą 72%  perf-sched.sch_delay.max.ms.wait_extent_bit.__lock_extent.lock_extents_for_read.constprop.0
>      26708 ą  2%      -9.5%      24158 ą  3%  perf-sched.total_wait_and_delay.count.ms
>       3.20 ą  9%     +25.5%       4.02 ą 11%  perf-sched.wait_and_delay.avg.ms.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
>       7.26 ą 15%     +50.7%      10.94 ą 19%  perf-sched.wait_and_delay.avg.ms.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
>      19.16 ą 29%     -65.5%       6.61 ą142%  perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.__mutex_lock.constprop.0.btrfs_wait_ordered_roots
>       0.02 ą 59%    +188.1%       0.05 ą 63%  perf-sched.wait_and_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.sync_inodes_sb
>      15.73 ą 19%     +50.8%      23.72 ą 10%  perf-sched.wait_and_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
>     256.17 ą 15%     +34.5%     344.47 ą 13%  perf-sched.wait_and_delay.avg.ms.wait_current_trans.start_transaction.btrfs_create_common.lookup_open.isra
>      78.54 ą  8%     +18.1%      92.78 ą 11%  perf-sched.wait_and_delay.avg.ms.wait_for_commit.btrfs_commit_transaction.iterate_supers.ksys_sync
>      67.31 ą  3%     +12.5%      75.72 ą  5%  perf-sched.wait_and_delay.avg.ms.wait_for_commit.btrfs_wait_for_commit.btrfs_attach_transaction_barrier.btrfs_sync_fs
>     805.00 ą 11%     -21.9%     629.00 ą 11%  perf-sched.wait_and_delay.count.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
>     833.33 ą  5%     -29.6%     586.50 ą 13%  perf-sched.wait_and_delay.count.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
>       2990 ą  5%     -16.5%       2498 ą  3%  perf-sched.wait_and_delay.count.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
>     151.17 ą  6%     -70.6%      44.50 ą141%  perf-sched.wait_and_delay.count.schedule_preempt_disabled.__mutex_lock.constprop.0.btrfs_wait_ordered_roots
>     553.50 ą  5%     -26.3%     407.67 ą  9%  perf-sched.wait_and_delay.count.schedule_timeout.__wait_for_common.btrfs_wait_ordered_extents.btrfs_wait_ordered_roots
>     618.50 ą  9%     -25.4%     461.17 ą 12%  perf-sched.wait_and_delay.count.wait_current_trans.start_transaction.btrfs_sync_file.btrfs_do_write_iter
>     955.00 ą  8%     -27.5%     692.17 ą  8%  perf-sched.wait_and_delay.count.wait_log_commit.btrfs_sync_log.btrfs_sync_file.btrfs_do_write_iter
>     551.40 ą 10%     +33.8%     737.66 ą  8%  perf-sched.wait_and_delay.max.ms.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
>     251.59 ą 18%     +74.5%     438.99 ą 14%  perf-sched.wait_and_delay.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
>      11.70 ą 99%    -100.0%       0.00 ą223%  perf-sched.wait_time.avg.ms.__cond_resched.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range
>      44.61 ą189%    +762.7%     384.84 ą 81%  perf-sched.wait_time.avg.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
>       3.19 ą  9%     +25.6%       4.00 ą 11%  perf-sched.wait_time.avg.ms.btrfs_commit_transaction.iterate_supers.ksys_sync.__x64_sys_sync
>       7.24 ą 15%     +50.9%      10.93 ą 19%  perf-sched.wait_time.avg.ms.btrfs_start_ordered_extent_nowriteback.btrfs_run_ordered_extent_work.btrfs_work_helper.process_one_work
>       1.69 ą 79%    +178.9%       4.71 ą 62%  perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.extent_write_cache_pages.btrfs_writepages
>      15.72 ą 19%     +50.9%      23.71 ą 10%  perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
>      20.99 ą  2%     +38.2%      29.02 ą 15%  perf-sched.wait_time.avg.ms.schedule_timeout.btrfs_sync_log.btrfs_sync_file.btrfs_do_write_iter
>      12.97 ą 13%     +49.0%      19.32 ą 16%  perf-sched.wait_time.avg.ms.schedule_timeout.io_schedule_timeout.__wait_for_common.barrier_all_devices
>     256.15 ą 15%     +34.5%     344.45 ą 13%  perf-sched.wait_time.avg.ms.wait_current_trans.start_transaction.btrfs_create_common.lookup_open.isra
>      10.94 ą 25%     +84.6%      20.19 ą 24%  perf-sched.wait_time.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>      78.51 ą  8%     +18.1%      92.75 ą 11%  perf-sched.wait_time.avg.ms.wait_for_commit.btrfs_commit_transaction.iterate_supers.ksys_sync
>      67.26 ą  3%     +12.5%      75.66 ą  5%  perf-sched.wait_time.avg.ms.wait_for_commit.btrfs_wait_for_commit.btrfs_attach_transaction_barrier.btrfs_sync_fs
>      19.50 ą128%    -100.0%       0.00 ą223%  perf-sched.wait_time.max.ms.__cond_resched.btree_write_cache_pages.do_writepages.filemap_fdatawrite_wbc.__filemap_fdatawrite_range
>      83.75 ą187%    +516.6%     516.38 ą 68%  perf-sched.wait_time.max.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
>     551.39 ą 10%     +33.8%     737.64 ą  8%  perf-sched.wait_time.max.ms.io_schedule.folio_wait_bit_common.folio_wait_writeback.__filemap_fdatawait_range
>     251.57 ą 18%     +74.5%     438.95 ą 14%  perf-sched.wait_time.max.ms.wait_current_trans.start_transaction.btrfs_attach_transaction_barrier.btrfs_sync_fs
>      86.74 ą 14%     -27.4%      62.97 ą 11%  perf-sched.wait_time.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>
>
> ***************************************************************************************************
> lkp-ivb-d01: 8 threads 1 sockets Intel(R) Core(TM) i7-3770K CPU @ 3.50GHz (Ivy Bridge) with 16G memory
> =========================================================================================
> compiler/cpufreq_governor/disk/fs/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase:
>   gcc-12/performance/1HDD/btrfs/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/300s/lkp-ivb-d01/disk/reaim
>
> commit:
>   cbfb4cbf45 ("btrfs: update comment for try_release_extent_state()")
>   32c523c578 ("btrfs: allow folios to be released while ordered extent is finishing")
>
> cbfb4cbf459d9be4 32c523c578e8489f55663ce8a88
> ---------------- ---------------------------
>          %stddev     %change         %stddev
>              \          |                \
>      22777          +101.6%      45913 ą 13%  proc-vmstat.pgpgin
>      12134 ą  4%     +18.9%      14426 ą  7%  sched_debug.cpu.nr_switches.stddev
>      80.30            -1.8%      78.89        iostat.cpu.idle
>      17.77            +8.1%      19.21        iostat.cpu.iowait
>      54.02          +101.9%     109.06        vmstat.io.bi
>       4061            -5.0%       3856        vmstat.io.bo
>       2.60            +2.2%       2.65        reaim.child_systime
>     348.91            -5.3%     330.27        reaim.jobs_per_min
>      43.61            -5.3%      41.28        reaim.jobs_per_min_child
>     356.27            -5.4%     336.88        reaim.max_jobs_per_min
>     137.64            +5.6%     145.41        reaim.parent_time
>      45554          +101.6%      91826 ą 13%  reaim.time.file_system_inputs
>       0.17 ą 25%      -0.1        0.09 ą 31%  perf-profile.children.cycles-pp.select_task_rq_fair
>       0.16 ą 29%      -0.1        0.10 ą 33%  perf-profile.children.cycles-pp.vfs_fstatat
>       0.21 ą 16%      -0.1        0.15 ą 12%  perf-profile.children.cycles-pp.___perf_sw_event
>       0.06 ą 79%      +0.1        0.16 ą 33%  perf-profile.children.cycles-pp.vms_complete_munmap_vmas
>       0.07 ą 72%      +0.1        0.18 ą 16%  perf-profile.children.cycles-pp.vms_clear_ptes
>       0.19 ą 18%      -0.1        0.12 ą 19%  perf-profile.self.cycles-pp.___perf_sw_event
>       0.15 ą 30%      +0.1        0.23 ą 33%  perf-profile.self.cycles-pp.x86_pmu_disable
>       0.04 ą  9%     +21.8%       0.05 ą 11%  perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.barrier_all_devices.write_all_supers.btrfs_sync_log
>       0.05 ą  4%     +18.7%       0.06 ą  4%  perf-sched.sch_delay.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
>       0.04 ą 69%    -100.0%       0.00        perf-sched.sch_delay.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>       0.05 ą 69%    -100.0%       0.00        perf-sched.sch_delay.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>      19.30 ą  9%     +17.0%      22.59 ą  6%  perf-sched.wait_and_delay.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
>      19.25 ą  9%     +17.0%      22.53 ą  6%  perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.write_all_supers.btrfs_sync_log
>      15.74 ą107%    -100.0%       0.00        perf-sched.wait_time.avg.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>      22.16 ą 79%    -100.0%       0.00        perf-sched.wait_time.max.ms.wait_extent_bit.__lock_extent.lock_and_cleanup_extent_if_need.copy_one_range
>       4.44            +1.9%       4.53        perf-stat.i.MPKI
>       8.16            +0.2        8.41        perf-stat.i.branch-miss-rate%
>       2313            +1.6%       2350        perf-stat.i.context-switches
>       2.25            +1.9%       2.29        perf-stat.i.cpi
>      73.48 ą  3%      -6.6%      68.61        perf-stat.i.cpu-migrations
>       1972            -2.1%       1931        perf-stat.i.minor-faults
>       1972            -2.1%       1931        perf-stat.i.page-faults
>       2308            +1.6%       2344        perf-stat.ps.context-switches
>      73.31 ą  3%      -6.6%      68.45        perf-stat.ps.cpu-migrations
>       1967            -2.1%       1926        perf-stat.ps.minor-faults
>       1968            -2.1%       1926        perf-stat.ps.page-faults
>
>
>
>
>
> 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
>
>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2025-06-12 13:17 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-06-12  6:26 [linus:master] [btrfs] 32c523c578: reaim.jobs_per_min 16.1% regression kernel test robot
2025-06-12 13:17 ` Filipe Manana

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®