[linux-next:master] [super] 1af3331764: stress-ng.unlink.unlink_calls_per_sec 23.3% regression
kernel test robot
oliver.sang at intel.com
Mon May 19 04:50:03 UTC 2025
Hello,
kernel test robot noticed a 23.3% regression of stress-ng.unlink.unlink_calls_per_sec on:
commit: 1af3331764b9356fadc4652af77bbbc97f3d7f78 ("super: add filesystem freezing helpers for suspend and hibernate")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
[test failed on linux-next/master 8566fc3b96539e3235909d6bdda198e1282beaed]
testcase: stress-ng
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:
nr_threads: 100%
testtime: 60s
test: unlink
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 at intel.com>
| Closes: https://lore.kernel.org/oe-lkp/202505191143.59950d28-lkp@intel.com
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250519/202505191143.59950d28-lkp@intel.com
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
gcc-12/performance/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp8/unlink/stress-ng/60s
commit:
62a2175ddf ("gfs2: pass through holder from the VFS for freeze/thaw")
1af3331764 ("super: add filesystem freezing helpers for suspend and hibernate")
62a2175ddf7e7294 1af3331764b9356fadc4652af77
---------------- ---------------------------
%stddev %change %stddev
\ | \
22349 ± 2% +6.5% 23809 vmstat.system.cs
159467 +2.1% 162851 vmstat.system.in
67494 +12.8% 76141 ± 20% proc-vmstat.nr_shmem
1367488 -2.8% 1329138 proc-vmstat.nr_slab_reclaimable
541527 -1.8% 531651 proc-vmstat.nr_slab_unreclaimable
316736 +17.4% 371828 stress-ng.time.voluntary_context_switches
47192 +1.4% 47854 stress-ng.unlink.ops
712.42 +1.8% 725.38 stress-ng.unlink.ops_per_sec
12343 -23.3% 9464 stress-ng.unlink.unlink_calls_per_sec
1.376e+10 -1.5% 1.355e+10 perf-stat.i.branch-instructions
50.73 -1.6 49.16 perf-stat.i.cache-miss-rate%
2.702e+08 +3.7% 2.802e+08 perf-stat.i.cache-references
23174 ± 2% +6.2% 24600 perf-stat.i.context-switches
1565 +16.8% 1828 perf-stat.i.cpu-migrations
6.418e+10 -1.5% 6.321e+10 perf-stat.i.instructions
0.29 -1.5% 0.29 perf-stat.i.ipc
2.11 +1.9% 2.15 perf-stat.overall.MPKI
50.08 -1.6 48.47 perf-stat.overall.cache-miss-rate%
3.52 +1.5% 3.58 perf-stat.overall.cpi
0.28 -1.5% 0.28 perf-stat.overall.ipc
1.356e+10 -1.5% 1.336e+10 perf-stat.ps.branch-instructions
2.664e+08 +3.7% 2.763e+08 perf-stat.ps.cache-references
22843 ± 2% +6.2% 24252 perf-stat.ps.context-switches
1544 +16.8% 1803 perf-stat.ps.cpu-migrations
6.328e+10 -1.5% 6.233e+10 perf-stat.ps.instructions
4.344e+12 -1.8% 4.268e+12 perf-stat.total.instructions
7.93 ± 3% -20.9% 6.27 ± 8% perf-sched.sch_delay.avg.ms.__cond_resched.__dentry_kill.dput.lookup_one_qstr_excl.do_unlinkat
5.39 ± 2% -24.0% 4.10 ± 3% perf-sched.sch_delay.avg.ms.__cond_resched.dput.do_unlinkat.__x64_sys_unlink.do_syscall_64
7.22 ± 5% -14.8% 6.15 ± 3% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_lru_noprof.__d_alloc.d_alloc.lookup_one_qstr_excl
7.73 ± 3% -8.9% 7.04 perf-sched.sch_delay.avg.ms.__cond_resched.mnt_want_write.do_unlinkat.__x64_sys_unlink.do_syscall_64
0.73 ± 5% -22.1% 0.57 ± 8% perf-sched.sch_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
6.43 ± 2% -14.6% 5.49 ± 3% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
15.68 ± 18% +25.0% 19.60 ± 9% perf-sched.sch_delay.max.ms.__cond_resched.down_read.unmap_mapping_range.simple_setattr.notify_change
24.13 ± 10% +44.0% 34.75 ± 23% perf-sched.sch_delay.max.ms.__cond_resched.down_write.do_truncate.do_open.path_openat
11.91 ± 2% -22.2% 9.26 ± 3% perf-sched.wait_and_delay.avg.ms.__cond_resched.dput.do_unlinkat.__x64_sys_unlink.do_syscall_64
71.98 ± 14% -30.4% 50.09 ± 15% perf-sched.wait_and_delay.avg.ms.anon_pipe_read.fifo_pipe_read.vfs_read.ksys_read
14.94 ± 5% -14.2% 12.82 ± 3% perf-sched.wait_and_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
27151 -18.5% 22138 perf-sched.wait_and_delay.count.__cond_resched.__dentry_kill.dput.__fput.__x64_sys_close
1001 ± 2% +77.4% 1776 ± 2% perf-sched.wait_and_delay.count.__cond_resched.down_write.vfs_unlink.do_unlinkat.__x64_sys_unlink
2584 ± 2% +19.3% 3082 ± 3% perf-sched.wait_and_delay.count.__cond_resched.dput.do_unlinkat.__x64_sys_unlink.do_syscall_64
1046 ± 2% -21.0% 826.67 ± 3% perf-sched.wait_and_delay.count.__cond_resched.dput.lookup_one_qstr_excl.do_unlinkat.__x64_sys_unlink
30647 -14.9% 26094 perf-sched.wait_and_delay.count.__cond_resched.dput.open_last_lookups.path_openat.do_filp_open
802.83 ± 3% +94.0% 1557 ± 3% perf-sched.wait_and_delay.count.__cond_resched.dput.simple_unlink.vfs_unlink.do_unlinkat
2993 +83.0% 5479 ± 2% perf-sched.wait_and_delay.count.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
1952 ± 2% +19.9% 2341 perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_lru_noprof.alloc_inode.new_inode.ramfs_get_inode
1026 ± 2% +59.1% 1632 ± 2% perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_noprof.security_inode_alloc.inode_init_always_gfp.alloc_inode
1831 ± 2% +70.4% 3120 perf-sched.wait_and_delay.count.__cond_resched.mnt_want_write.do_unlinkat.__x64_sys_unlink.do_syscall_64
1932 ± 2% +38.5% 2677 ± 2% perf-sched.wait_and_delay.count.__cond_resched.mnt_want_write.open_last_lookups.path_openat.do_filp_open
612.17 ± 15% +45.3% 889.33 ± 19% perf-sched.wait_and_delay.count.anon_pipe_read.fifo_pipe_read.vfs_read.ksys_read
7589 ± 2% +21.7% 9237 ± 2% perf-sched.wait_and_delay.count.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.do_unlinkat
3569 ± 2% +20.4% 4299 ± 2% perf-sched.wait_and_delay.count.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.filename_create
6.52 ± 2% -20.8% 5.16 ± 3% perf-sched.wait_time.avg.ms.__cond_resched.dput.do_unlinkat.__x64_sys_unlink.do_syscall_64
71.91 ± 14% -30.4% 50.04 ± 15% perf-sched.wait_time.avg.ms.anon_pipe_read.fifo_pipe_read.vfs_read.ksys_read
8.52 ± 7% -13.9% 7.33 ± 3% perf-sched.wait_time.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
23.62 ± 9% +26.8% 29.96 ± 8% perf-sched.wait_time.max.ms.__cond_resched.down_read.unmap_mapping_range.truncate_pagecache.simple_setattr
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
More information about the Linux-security-module-archive
mailing list