[linus:master] [xfs] 62334fab47: stress-ng.rename.ops_per_sec 4.5% improvement

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




Hello,

kernel test robot noticed a 4.5% improvement of stress-ng.rename.ops_per_sec on:


commit: 62334fab47621dd91ab30dd5bb6c43d78a8ec279 ("xfs: use per-mount cpumask to track nonempty percpu inodegc lists")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

testcase: stress-ng
test machine: 36 threads 1 sockets Intel(R) Core(TM) i9-9980XE CPU @ 3.00GHz (Skylake) with 32G memory
parameters:

	nr_threads: 10%
	disk: 1SSD
	testtime: 60s
	fs: xfs
	class: filesystem
	test: rename
	cpufreq_governor: performance


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

+------------------+------------------------------------------------------------------------------------------+
| testcase: change | stress-ng: stress-ng.symlink.ops_per_sec 2.1% improvement                                |
| test machine     | 36 threads 1 sockets Intel(R) Core(TM) i9-9980XE CPU @ 3.00GHz (Skylake) with 32G memory |
| test parameters  | class=filesystem                                                                         |
|                  | cpufreq_governor=performance                                                             |
|                  | disk=1SSD                                                                                |
|                  | fs=xfs                                                                                   |
|                  | nr_threads=10%                                                                           |
|                  | test=symlink                                                                             |
|                  | testtime=60s                                                                             |
+------------------+------------------------------------------------------------------------------------------+
| testcase: change | stress-ng: stress-ng.link.ops_per_sec 3.2% improvement                                   |
| test machine     | 36 threads 1 sockets Intel(R) Core(TM) i9-9980XE CPU @ 3.00GHz (Skylake) with 32G memory |
| test parameters  | class=filesystem                                                                         |
|                  | cpufreq_governor=performance                                                             |
|                  | disk=1SSD                                                                                |
|                  | fs=xfs                                                                                   |
|                  | nr_threads=10%                                                                           |
|                  | test=link                                                                                |
|                  | testtime=60s                                                                             |
+------------------+------------------------------------------------------------------------------------------+




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


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20231031/202310311328.77701e96-oliver.sang@xxxxxxxxx

=========================================================================================
class/compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
  filesystem/gcc-12/performance/1SSD/xfs/x86_64-rhel-8.3/10%/debian-11.1-x86_64-20220510.cgz/lkp-skl-d08/rename/stress-ng/60s

commit: 
  ecd49f7a36 ("xfs: fix per-cpu CIL structure aggregation racing with dying cpus")
  62334fab47 ("xfs: use per-mount cpumask to track nonempty percpu inodegc lists")

ecd49f7a36fbccc8 62334fab47621dd91ab30dd5bb6 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
      3065 ± 17%     +21.5%       3724 ±  8%  perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
      3065 ± 17%     +21.5%       3724 ±  8%  perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
  13731223            +4.5%   14352233        stress-ng.rename.ops
    228654            +4.5%     238992        stress-ng.rename.ops_per_sec
    208138            +4.1%     216733 ±  3%  stress-ng.time.voluntary_context_switches
      1.14            +0.0        1.17        perf-stat.i.branch-miss-rate%
      0.97            +1.5%       0.99        perf-stat.i.cpi
      0.03            +0.0        0.03        perf-stat.i.dTLB-load-miss-rate%
    798812            +3.5%     827138        perf-stat.i.dTLB-load-misses
      0.00 ± 11%      -0.0        0.00 ± 17%  perf-stat.i.dTLB-store-miss-rate%
 1.321e+09            +2.8%  1.359e+09        perf-stat.i.dTLB-stores
      1.03            -1.4%       1.02        perf-stat.i.ipc
    263342            +4.0%     273811 ±  2%  perf-stat.i.node-loads
      0.12 ±  2%      +3.6%       0.12        perf-stat.overall.MPKI
      0.97            +1.4%       0.99        perf-stat.overall.cpi
      0.03            +0.0        0.03        perf-stat.overall.dTLB-load-miss-rate%
      1.03            -1.4%       1.01        perf-stat.overall.ipc
    786211            +3.5%     814068        perf-stat.ps.dTLB-load-misses
   1.3e+09            +2.8%  1.337e+09        perf-stat.ps.dTLB-stores
    259165            +4.0%     269495 ±  2%  perf-stat.ps.node-loads
      2.99 ±  4%      -2.3        0.66 ± 10%  perf-profile.calltrace.cycles-pp.xfs_inodegc_queue_all.xfs_fs_statfs.statfs_by_dentry.user_statfs.__do_sys_statfs
      0.62 ±  6%      +0.1        0.71 ±  8%  perf-profile.calltrace.cycles-pp.lookup_dcache.lookup_one_qstr_excl.do_renameat2.__x64_sys_rename.do_syscall_64
      0.60 ±  5%      +0.1        0.68 ±  9%  perf-profile.calltrace.cycles-pp.d_lookup.lookup_dcache.lookup_one_qstr_excl.do_renameat2.__x64_sys_rename
      0.82 ±  6%      +0.1        0.94 ±  7%  perf-profile.calltrace.cycles-pp.d_move.vfs_rename.do_renameat2.__x64_sys_rename.do_syscall_64
      0.68 ±  6%      +0.1        0.81 ±  7%  perf-profile.calltrace.cycles-pp.kmem_cache_alloc_lru.__d_alloc.d_alloc.lookup_one_qstr_excl.do_renameat2
      1.58 ±  3%      +0.2        1.74 ±  6%  perf-profile.calltrace.cycles-pp.xfs_trans_reserve.xfs_trans_alloc.xfs_rename.xfs_vn_rename.vfs_rename
      0.98 ±  4%      +0.2        1.15 ±  7%  perf-profile.calltrace.cycles-pp.xfs_dir_removename.xfs_rename.xfs_vn_rename.vfs_rename.do_renameat2
      0.96 ± 11%      +0.2        1.20 ±  8%  perf-profile.calltrace.cycles-pp.mutex_spin_on_owner.__mutex_lock.do_renameat2.__x64_sys_rename.do_syscall_64
      1.53 ±  3%      +0.2        1.78 ±  7%  perf-profile.calltrace.cycles-pp.xfs_inode_item_format.xlog_cil_insert_format_items.xlog_cil_insert_items.xlog_cil_commit.__xfs_trans_commit
      1.94 ±  3%      +0.3        2.19 ±  6%  perf-profile.calltrace.cycles-pp.xfs_trans_alloc.xfs_rename.xfs_vn_rename.vfs_rename.do_renameat2
      1.19 ± 10%      +0.3        1.45 ±  9%  perf-profile.calltrace.cycles-pp.__mutex_lock.do_renameat2.__x64_sys_rename.do_syscall_64.entry_SYSCALL_64_after_hwframe
      1.78 ±  3%      +0.3        2.05 ±  7%  perf-profile.calltrace.cycles-pp.xlog_cil_insert_format_items.xlog_cil_insert_items.xlog_cil_commit.__xfs_trans_commit.xfs_rename
      2.24 ±  4%      +0.3        2.53 ±  8%  perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.__statfs
      2.61 ±  3%      +0.3        2.94 ±  6%  perf-profile.calltrace.cycles-pp.xlog_cil_insert_items.xlog_cil_commit.__xfs_trans_commit.xfs_rename.xfs_vn_rename
      2.41 ±  4%      +0.4        2.82 ± 10%  perf-profile.calltrace.cycles-pp.path_parentat.__filename_parentat.do_renameat2.__x64_sys_rename.do_syscall_64
      2.25 ±  4%      +0.4        2.69 ±  6%  perf-profile.calltrace.cycles-pp._find_next_or_bit.__percpu_counter_sum.xfs_fs_statfs.statfs_by_dentry.user_statfs
      2.74 ±  4%      +0.5        3.23 ±  9%  perf-profile.calltrace.cycles-pp.__filename_parentat.do_renameat2.__x64_sys_rename.do_syscall_64.entry_SYSCALL_64_after_hwframe
      0.08 ±223%      +0.5        0.58 ±  5%  perf-profile.calltrace.cycles-pp.path_init.path_parentat.__filename_parentat.do_renameat2.__x64_sys_rename
      4.69 ±  2%      +0.6        5.32 ±  8%  perf-profile.calltrace.cycles-pp.xlog_cil_commit.__xfs_trans_commit.xfs_rename.xfs_vn_rename.vfs_rename
      8.39 ±  4%      +1.2        9.54 ±  9%  perf-profile.calltrace.cycles-pp.__percpu_counter_sum.xfs_fs_statfs.statfs_by_dentry.user_statfs.__do_sys_statfs
     29.06 ±  2%      +3.9       32.91 ±  7%  perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.rename
     30.82 ±  2%      +4.1       34.94 ±  7%  perf-profile.calltrace.cycles-pp.rename
      3.00 ±  4%      -2.3        0.66 ± 10%  perf-profile.children.cycles-pp.xfs_inodegc_queue_all
      0.68 ±  7%      -0.5        0.13 ± 16%  perf-profile.children.cycles-pp._find_next_bit
      0.10 ± 12%      +0.0        0.13 ± 11%  perf-profile.children.cycles-pp.kfree
      0.08 ± 16%      +0.0        0.12 ± 12%  perf-profile.children.cycles-pp.__x64_sys_statfs
      0.31 ±  8%      +0.0        0.36 ±  8%  perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
      0.33 ±  7%      +0.1        0.40 ±  9%  perf-profile.children.cycles-pp.slab_pre_alloc_hook
      0.11 ± 11%      +0.1        0.18 ± 28%  perf-profile.children.cycles-pp.__d_lookup_rcu
      0.40 ±  7%      +0.1        0.47 ± 10%  perf-profile.children.cycles-pp.xfs_trans_del_item
      0.14 ± 12%      +0.1        0.22 ± 22%  perf-profile.children.cycles-pp.lookup_fast
      0.62 ±  5%      +0.1        0.71 ±  8%  perf-profile.children.cycles-pp.d_lookup
      0.48 ±  7%      +0.1        0.57 ±  7%  perf-profile.children.cycles-pp.__cond_resched
      0.64 ±  6%      +0.1        0.74 ±  7%  perf-profile.children.cycles-pp.lookup_dcache
      0.50 ±  3%      +0.1        0.60 ±  6%  perf-profile.children.cycles-pp.xfs_dir2_sf_removename
      0.48 ±  6%      +0.1        0.58 ± 11%  perf-profile.children.cycles-pp.xlog_cil_alloc_shadow_bufs
      0.49 ±  8%      +0.1        0.61 ±  9%  perf-profile.children.cycles-pp.__d_move
      0.82 ±  5%      +0.1        0.96 ±  7%  perf-profile.children.cycles-pp.d_move
      0.82 ±  4%      +0.1        0.96 ±  4%  perf-profile.children.cycles-pp.__d_alloc
      0.71 ±  5%      +0.1        0.86 ±  5%  perf-profile.children.cycles-pp.kmem_cache_alloc_lru
      0.48 ± 11%      +0.1        0.62 ± 14%  perf-profile.children.cycles-pp.up_write
      0.97 ±  5%      +0.2        1.13 ±  5%  perf-profile.children.cycles-pp.d_alloc
      1.88 ±  2%      +0.2        2.04 ±  9%  perf-profile.children.cycles-pp.strncpy_from_user
      1.00 ±  4%      +0.2        1.18 ±  7%  perf-profile.children.cycles-pp.xfs_dir_removename
      0.36 ± 24%      +0.2        0.58 ± 19%  perf-profile.children.cycles-pp.__legitimize_mnt
      0.96 ± 11%      +0.2        1.20 ±  8%  perf-profile.children.cycles-pp.mutex_spin_on_owner
      0.77 ± 11%      +0.2        1.02 ±  6%  perf-profile.children.cycles-pp.path_init
      1.58 ±  3%      +0.3        1.83 ±  6%  perf-profile.children.cycles-pp.xfs_inode_item_format
      2.02 ±  3%      +0.3        2.28 ±  6%  perf-profile.children.cycles-pp.xfs_trans_alloc
      2.23 ±  5%      +0.3        2.49 ±  4%  perf-profile.children.cycles-pp._raw_spin_lock
      1.19 ± 10%      +0.3        1.45 ±  9%  perf-profile.children.cycles-pp.__mutex_lock
      1.89 ±  2%      +0.3        2.17 ±  6%  perf-profile.children.cycles-pp.xlog_cil_insert_format_items
      2.73 ±  3%      +0.3        3.06 ±  6%  perf-profile.children.cycles-pp.xlog_cil_insert_items
      2.51 ±  4%      +0.4        2.92 ±  9%  perf-profile.children.cycles-pp.path_parentat
      2.27 ±  3%      +0.4        2.70 ±  6%  perf-profile.children.cycles-pp._find_next_or_bit
      3.61 ±  3%      +0.5        4.07 ±  6%  perf-profile.children.cycles-pp.syscall_return_via_sysret
      2.83 ±  4%      +0.5        3.31 ±  8%  perf-profile.children.cycles-pp.__filename_parentat
      4.97 ±  2%      +0.6        5.60 ±  7%  perf-profile.children.cycles-pp.xlog_cil_commit
      7.91 ±  2%      +0.9        8.81 ±  7%  perf-profile.children.cycles-pp.__xfs_trans_commit
      8.76 ±  3%      +1.2        9.99 ±  9%  perf-profile.children.cycles-pp.__percpu_counter_sum
     26.27 ±  2%      +3.5       29.73 ±  7%  perf-profile.children.cycles-pp.do_renameat2
     30.90 ±  2%      +4.1       35.03 ±  7%  perf-profile.children.cycles-pp.rename
      1.93 ±  4%      -1.8        0.11 ± 21%  perf-profile.self.cycles-pp.xfs_inodegc_queue_all
      0.66 ±  7%      -0.5        0.13 ± 20%  perf-profile.self.cycles-pp._find_next_bit
      0.08 ± 17%      +0.0        0.12 ± 11%  perf-profile.self.cycles-pp.kfree
      0.08 ± 18%      +0.0        0.12 ± 15%  perf-profile.self.cycles-pp.__x64_sys_statfs
      0.12 ± 18%      +0.0        0.17 ± 13%  perf-profile.self.cycles-pp.__percpu_counter_compare
      0.11 ± 11%      +0.1        0.17 ± 31%  perf-profile.self.cycles-pp.__d_lookup_rcu
      0.17 ± 13%      +0.1        0.24 ±  9%  perf-profile.self.cycles-pp.xfs_inode_item_format_data_fork
      0.17 ± 10%      +0.1        0.24 ± 16%  perf-profile.self.cycles-pp.xlog_cil_alloc_shadow_bufs
      0.28 ±  5%      +0.1        0.35 ±  9%  perf-profile.self.cycles-pp.__cond_resched
      0.32 ±  5%      +0.1        0.39 ±  5%  perf-profile.self.cycles-pp.__filename_parentat
      0.36 ± 23%      +0.2        0.57 ± 19%  perf-profile.self.cycles-pp.__legitimize_mnt
      1.52 ±  5%      +0.2        1.74 ±  8%  perf-profile.self.cycles-pp.__entry_text_start
      0.96 ± 11%      +0.2        1.20 ±  8%  perf-profile.self.cycles-pp.mutex_spin_on_owner
      2.09 ±  5%      +0.2        2.34 ±  4%  perf-profile.self.cycles-pp._raw_spin_lock
      0.75 ± 11%      +0.3        1.01 ±  6%  perf-profile.self.cycles-pp.path_init
      1.93 ±  5%      +0.4        2.30 ±  7%  perf-profile.self.cycles-pp._find_next_or_bit
      3.60 ±  3%      +0.5        4.06 ±  6%  perf-profile.self.cycles-pp.syscall_return_via_sysret
      5.79 ±  3%      +0.8        6.64 ± 10%  perf-profile.self.cycles-pp.__percpu_counter_sum


***************************************************************************************************
lkp-skl-d08: 36 threads 1 sockets Intel(R) Core(TM) i9-9980XE CPU @ 3.00GHz (Skylake) with 32G memory
=========================================================================================
class/compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
  filesystem/gcc-12/performance/1SSD/xfs/x86_64-rhel-8.3/10%/debian-11.1-x86_64-20220510.cgz/lkp-skl-d08/symlink/stress-ng/60s

commit: 
  ecd49f7a36 ("xfs: fix per-cpu CIL structure aggregation racing with dying cpus")
  62334fab47 ("xfs: use per-mount cpumask to track nonempty percpu inodegc lists")

ecd49f7a36fbccc8 62334fab47621dd91ab30dd5bb6 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
    117708            +1.9%     119936        vmstat.system.cs
   1092777            +1.7%    1110814        proc-vmstat.numa_hit
   1092614            +2.3%    1117304        proc-vmstat.numa_local
    474.50            +2.1%     484.50        stress-ng.symlink.ops
      7.90            +2.1%       8.06        stress-ng.symlink.ops_per_sec
      0.02 ±131%   +5345.1%       1.29 ±130%  perf-sched.sch_delay.max.ms.__cond_resched.down_read.xlog_cil_commit.__xfs_trans_commit.xfs_inactive_ifree
      0.01 ± 85%     -62.4%       0.01 ± 48%  perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc.__xfs_free_extent_later.xfs_bmap_del_extent_real.__xfs_bunmapi
      0.02 ± 51%    +422.5%       0.10 ±146%  perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc.xfs_trans_alloc.xfs_trans_alloc_dir.xfs_remove
      0.01 ± 24%    +392.7%       0.07 ±147%  perf-sched.wait_time.avg.ms.__cond_resched.down_read.xlog_cil_commit.__xfs_trans_commit.xfs_trans_roll
      0.77 ±145%    +129.3%       1.77 ± 90%  perf-sched.wait_time.max.ms.__cond_resched.mutex_lock.perf_event_ctx_lock_nested.constprop.0
     16.85            -0.5       16.33 ±  2%  perf-stat.i.cache-miss-rate%
    123504            +1.6%     125516        perf-stat.i.context-switches
      0.99            +1.0%       1.00        perf-stat.i.cpi
     16.87            -0.6       16.26 ±  2%  perf-stat.overall.cache-miss-rate%
      0.99            +1.0%       1.00        perf-stat.overall.cpi
    121555            +1.6%     123540        perf-stat.ps.context-switches
      3.17 ±  2%      -1.1        2.02 ±  2%  perf-profile.calltrace.cycles-pp.xfs_inodegc_queue_all.xfs_fs_statfs.statfs_by_dentry.user_statfs.__do_sys_statfs
     12.49            -1.0       11.44 ±  2%  perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__statfs
      8.02 ±  2%      -1.0        6.98 ±  2%  perf-profile.calltrace.cycles-pp.xfs_fs_statfs.statfs_by_dentry.user_statfs.__do_sys_statfs.do_syscall_64
     11.88            -1.0       10.84 ±  2%  perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
      8.14 ±  2%      -1.0        7.10 ±  2%  perf-profile.calltrace.cycles-pp.statfs_by_dentry.user_statfs.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe
     10.51            -1.0        9.48 ±  2%  perf-profile.calltrace.cycles-pp.user_statfs.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
     10.76            -1.0        9.74 ±  2%  perf-profile.calltrace.cycles-pp.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
     14.25            -1.0       13.23 ±  2%  perf-profile.calltrace.cycles-pp.__statfs
      3.19 ±  2%      -1.2        2.03 ±  2%  perf-profile.children.cycles-pp.xfs_inodegc_queue_all
      8.14 ±  2%      -1.0        7.10 ±  2%  perf-profile.children.cycles-pp.statfs_by_dentry
      8.05 ±  2%      -1.0        7.00 ±  2%  perf-profile.children.cycles-pp.xfs_fs_statfs
     14.34            -1.0       13.31 ±  2%  perf-profile.children.cycles-pp.__statfs
     10.51            -1.0        9.49 ±  2%  perf-profile.children.cycles-pp.user_statfs
     10.76            -1.0        9.74 ±  2%  perf-profile.children.cycles-pp.__do_sys_statfs
      0.33 ± 10%      -0.2        0.13 ±  8%  perf-profile.children.cycles-pp._find_next_bit
      0.08 ±  9%      +0.0        0.10 ±  6%  perf-profile.children.cycles-pp.xfs_btree_ptr_to_daddr
      0.34 ±  6%      +0.1        0.39 ±  7%  perf-profile.children.cycles-pp.xlog_ticket_alloc
      0.65 ±  4%      +0.1        0.73 ±  5%  perf-profile.children.cycles-pp.__filename_parentat
      1.12            -0.9        0.22 ±  7%  perf-profile.self.cycles-pp.xfs_inodegc_queue_all
      0.32 ± 11%      -0.2        0.13 ±  8%  perf-profile.self.cycles-pp._find_next_bit
      0.35 ±  5%      -0.0        0.30 ±  4%  perf-profile.self.cycles-pp.__list_del_entry_valid_or_report
      0.08 ±  9%      +0.0        0.10 ±  7%  perf-profile.self.cycles-pp.xfs_btree_ptr_to_daddr



***************************************************************************************************
lkp-skl-d08: 36 threads 1 sockets Intel(R) Core(TM) i9-9980XE CPU @ 3.00GHz (Skylake) with 32G memory
=========================================================================================
class/compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
  filesystem/gcc-12/performance/1SSD/xfs/x86_64-rhel-8.3/10%/debian-11.1-x86_64-20220510.cgz/lkp-skl-d08/link/stress-ng/60s

commit: 
  ecd49f7a36 ("xfs: fix per-cpu CIL structure aggregation racing with dying cpus")
  62334fab47 ("xfs: use per-mount cpumask to track nonempty percpu inodegc lists")

ecd49f7a36fbccc8 62334fab47621dd91ab30dd5bb6 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
      0.10 ± 27%     -37.0%       0.06 ± 20%  perf-sched.wait_time.avg.ms.__cond_resched.down.xfs_buf_lock.xfs_buf_find_lock.xfs_buf_lookup
    862.00            +3.2%     889.50        stress-ng.link.ops
     14.35            +3.2%      14.81        stress-ng.link.ops_per_sec
 2.423e+09            -1.3%   2.39e+09        perf-stat.i.branch-instructions
  49572825 ±  2%      +5.6%   52346561 ±  3%  perf-stat.i.cache-references
      0.03            +0.0        0.03        perf-stat.i.dTLB-load-miss-rate%
   1021147            +1.8%    1039653        perf-stat.i.dTLB-load-misses
 2.028e+09            +2.3%  2.074e+09        perf-stat.i.dTLB-stores
      0.03            +0.0        0.03        perf-stat.overall.dTLB-load-miss-rate%
 2.385e+09            -1.3%  2.353e+09        perf-stat.ps.branch-instructions
  48787960 ±  2%      +5.6%   51523373 ±  3%  perf-stat.ps.cache-references
   1004980            +1.8%    1023298        perf-stat.ps.dTLB-load-misses
 1.996e+09            +2.3%  2.041e+09        perf-stat.ps.dTLB-stores
     46299 ± 56%     +97.1%      91275 ±  5%  sched_debug.cfs_rq:/.avg_vruntime.max
      9861 ± 64%    +101.9%      19910 ±  3%  sched_debug.cfs_rq:/.avg_vruntime.stddev
     46299 ± 56%     +97.1%      91275 ±  5%  sched_debug.cfs_rq:/.min_vruntime.max
      9861 ± 64%    +101.9%      19910 ±  3%  sched_debug.cfs_rq:/.min_vruntime.stddev
     38310 ± 36%     +52.4%      58379        sched_debug.cpu.clock.avg
     38311 ± 36%     +52.4%      58381        sched_debug.cpu.clock.max
     38308 ± 36%     +52.4%      58378        sched_debug.cpu.clock.min
     38210 ± 36%     +52.1%      58133        sched_debug.cpu.clock_task.avg
     38238 ± 36%     +52.1%      58173        sched_debug.cpu.clock_task.max
     37820 ± 36%     +52.7%      57762        sched_debug.cpu.clock_task.min
      3535 ±  8%     +12.4%       3975        sched_debug.cpu.curr->pid.max
     14219 ± 93%    +132.7%      33087        sched_debug.cpu.nr_switches.avg
     31.08 ± 92%    +153.1%      78.67 ±  8%  sched_debug.cpu.nr_uninterruptible.max
    -23.83          +161.5%     -62.33        sched_debug.cpu.nr_uninterruptible.min
     10.35 ± 78%    +127.4%      23.54 ±  4%  sched_debug.cpu.nr_uninterruptible.stddev
     38309 ± 36%     +52.4%      58378        sched_debug.cpu_clk
     38230 ± 36%     +52.5%      58299        sched_debug.ktime
     38330 ± 36%     +52.4%      58401        sched_debug.sched_clk
     16.79 ±  2%      -2.1       14.74 ±  3%  perf-profile.calltrace.cycles-pp.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
     16.32 ±  2%      -2.0       14.28 ±  3%  perf-profile.calltrace.cycles-pp.user_statfs.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
     18.83 ±  2%      -2.0       16.84 ±  3%  perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__statfs
     19.91 ±  2%      -2.0       17.95 ±  3%  perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__statfs
     12.10 ±  2%      -1.9       10.18 ±  3%  perf-profile.calltrace.cycles-pp.xfs_fs_statfs.statfs_by_dentry.user_statfs.__do_sys_statfs.do_syscall_64
     12.29 ±  2%      -1.9       10.36 ±  3%  perf-profile.calltrace.cycles-pp.statfs_by_dentry.user_statfs.__do_sys_statfs.do_syscall_64.entry_SYSCALL_64_after_hwframe
     22.92            -1.7       21.20 ±  3%  perf-profile.calltrace.cycles-pp.__statfs
      0.74 ±  4%      +0.1        0.85        perf-profile.calltrace.cycles-pp.xfs_trans_read_buf_map.xfs_da_read_buf.xfs_da3_node_lookup_int.xfs_dir2_node_addname.xfs_dir_createname
      1.32 ±  5%      +0.1        1.44 ±  3%  perf-profile.calltrace.cycles-pp.xfs_da_read_buf.xfs_da3_node_lookup_int.xfs_dir2_node_addname.xfs_dir_createname.xfs_link
      1.88 ±  3%      +0.2        2.07 ±  2%  perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.__statfs
      2.34 ±  4%      -2.2        0.16 ± 16%  perf-profile.children.cycles-pp.xfs_inodegc_queue_all
     16.80 ±  2%      -2.1       14.74 ±  3%  perf-profile.children.cycles-pp.__do_sys_statfs
     16.33 ±  2%      -2.0       14.29 ±  3%  perf-profile.children.cycles-pp.user_statfs
     12.14 ±  2%      -1.9       10.21 ±  3%  perf-profile.children.cycles-pp.xfs_fs_statfs
     12.29 ±  2%      -1.9       10.36 ±  3%  perf-profile.children.cycles-pp.statfs_by_dentry
     23.06            -1.7       21.34 ±  3%  perf-profile.children.cycles-pp.__statfs
      0.58 ±  8%      -0.4        0.13 ± 14%  perf-profile.children.cycles-pp._find_next_bit
      0.08 ± 16%      +0.0        0.10 ± 13%  perf-profile.children.cycles-pp.xfs_qm_need_dqattach
      0.14 ± 18%      +0.0        0.18 ±  8%  perf-profile.children.cycles-pp.xfs_vn_getattr
      3.96 ±  3%      +0.2        4.19 ±  2%  perf-profile.children.cycles-pp.syscall_return_via_sysret
      0.57 ±  9%      -0.4        0.13 ± 15%  perf-profile.self.cycles-pp._find_next_bit
      0.06 ± 13%      +0.0        0.08 ± 18%  perf-profile.self.cycles-pp.xfs_dir2_node_find_freeblk
      0.07 ± 18%      +0.0        0.10 ±  9%  perf-profile.self.cycles-pp.xfs_mod_freecounter
      0.07 ± 16%      +0.0        0.10 ± 13%  perf-profile.self.cycles-pp.xfs_qm_need_dqattach
      0.17 ± 11%      +0.0        0.21 ±  7%  perf-profile.self.cycles-pp.xfs_link
      1.71 ±  5%      +0.1        1.84 ±  4%  perf-profile.self.cycles-pp._find_next_or_bit
      3.95 ±  3%      +0.2        4.19 ±  2%  perf-profile.self.cycles-pp.syscall_return_via_sysret



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




[Index of Archives]     [XFS Filesystem Development (older mail)]     [Linux Filesystem Development]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux RAID]     [Linux SCSI]


  Powered by Linux