Hello,
kernel test robot noticed a 15.4% regression of filebench.sum_operations/s on:
commit: 7334c4df7a384b31f30a61adb60243a8614f8ff0 ("[PATCH 1/3] nfsd: add TIME_DELEG_ACCESS and TIME_DELEG_MODIFY to writeable attrs")
url: https://github.com/intel-lab-lkp/linux/commits/Jeff-Layton/nfsd-add-TIME_DELEG_ACCESS-and-TIME_DELEG_MODIFY-to-writeable-attrs/20241019-024741
patch link: https://lore.kernel.org/all/20241018-delstid-v1-1-c6021b75ff3e@xxxxxxxxxx/
patch subject: [PATCH 1/3] nfsd: add TIME_DELEG_ACCESS and TIME_DELEG_MODIFY to writeable attrs
testcase: filebench
config: x86_64-rhel-8.3
compiler: gcc-12
test machine: 128 threads 2 sockets Intel(R) Xeon(R) Platinum 8358 CPU @ 2.60GHz (Ice Lake) with 128G memory
parameters:
disk: 1HDD
fs: ext4
fs2: nfsv4
test: webproxy.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@xxxxxxxxx>
| Closes: https://lore.kernel.org/oe-lkp/202410281526.4971befc-oliver.sang@xxxxxxxxx
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20241028/202410281526.4971befc-oliver.sang@xxxxxxxxx
=========================================================================================
compiler/cpufreq_governor/disk/fs2/fs/kconfig/rootfs/tbox_group/test/testcase:
gcc-12/performance/1HDD/nfsv4/ext4/x86_64-rhel-8.3/debian-12-x86_64-20240206.cgz/lkp-icl-2sp6/webproxy.f/filebench
commit:
0f8b1a4184 ("lockd: Remove unneeded initialization of file_lock::c.flc_flags")
7334c4df7a ("nfsd: add TIME_DELEG_ACCESS and TIME_DELEG_MODIFY to writeable attrs")
0f8b1a41842544ec 7334c4df7a384b31f30a61adb60
---------------- ---------------------------
%stddev %change %stddev
\ | \
1.19 ±181% -83.9% 0.19 ± 6% perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
2514 ± 95% -88.8% 280.49 ±168% sched_debug.cpu.max_idle_balance_cost.stddev
2152 -3.4% 2080 vmstat.system.cs
2189 +7.3% 2349 vmstat.system.in
25682 ± 41% +59.9% 41078 ± 4% numa-meminfo.node0.Shmem
20709 ± 91% -82.5% 3615 ±111% numa-meminfo.node1.Mapped
46372 ± 22% -29.8% 32541 ± 5% numa-meminfo.node1.Shmem
6419 ± 41% +60.0% 10272 ± 4% numa-vmstat.node0.nr_shmem
5241 ± 90% -81.1% 988.55 ±104% numa-vmstat.node1.nr_mapped
11592 ± 22% -29.8% 8135 ± 5% numa-vmstat.node1.nr_shmem
0.70 -14.3% 0.60 filebench.sum_bytes_mb/s
9121 -15.4% 7720 filebench.sum_operations
152.01 -15.4% 128.66 filebench.sum_operations/s
39.67 -16.0% 33.33 filebench.sum_reads/s
620.37 +19.7% 742.65 filebench.sum_time_ms/op
8.00 -12.5% 7.00 filebench.sum_writes/s
1370 +6.3% 1456 filebench.time.elapsed_time
1370 +6.3% 1456 filebench.time.elapsed_time.max
42175 -2.1% 41280 filebench.time.voluntary_context_switches
31030536 -1.2% 30643382 perf-stat.i.branch-instructions
4.86 +0.1 4.96 perf-stat.i.branch-miss-rate%
7534035 +4.1% 7839877 perf-stat.i.cache-references
2139 -3.4% 2067 perf-stat.i.context-switches
1.509e+08 -1.2% 1.491e+08 perf-stat.i.instructions
1.33 +1.1% 1.35 perf-stat.overall.cpi
30980974 -1.2% 30597842 perf-stat.ps.branch-instructions
7527450 +4.1% 7833510 perf-stat.ps.cache-references
2137 -3.3% 2066 perf-stat.ps.context-switches
1.507e+08 -1.2% 1.489e+08 perf-stat.ps.instructions
2.069e+11 +4.9% 2.171e+11 perf-stat.total.instructions
27282 ± 2% +8.6% 29623 proc-vmstat.nr_active_file
79891 +2.6% 81962 proc-vmstat.nr_dirtied
18010 +2.2% 18405 proc-vmstat.nr_shmem
79856 +2.6% 81922 proc-vmstat.nr_written
27282 ± 2% +8.6% 29623 proc-vmstat.nr_zone_active_file
2853705 +4.5% 2982082 proc-vmstat.numa_hit
2721089 +4.7% 2849533 proc-vmstat.numa_local
3437788 +3.8% 3567439 proc-vmstat.pgalloc_normal
3344808 +5.8% 3540131 proc-vmstat.pgfault
3343237 +3.7% 3466101 proc-vmstat.pgfree
923426 +4.7% 966532 proc-vmstat.pgpgout
156450 +5.7% 165442 proc-vmstat.pgreuse
3.55 ± 7% -0.6 2.96 ± 8% perf-profile.children.cycles-pp.sched_balance_rq
3.10 ± 9% -0.5 2.64 ± 5% perf-profile.children.cycles-pp.sched_balance_find_src_group
0.67 ± 34% -0.4 0.26 ± 46% perf-profile.children.cycles-pp.__wake_up_common
1.20 ± 11% -0.4 0.82 ± 9% perf-profile.children.cycles-pp.copy_process
0.57 ± 22% -0.3 0.30 ± 21% perf-profile.children.cycles-pp.dup_mm
0.65 ± 28% -0.2 0.41 ± 38% perf-profile.children.cycles-pp.free_pgtables
0.02 ±141% +0.1 0.10 ± 30% perf-profile.children.cycles-pp.rpc_run_task
0.02 ±223% +0.1 0.12 ± 39% perf-profile.children.cycles-pp.security_cred_free
0.02 ±141% +0.1 0.16 ± 50% perf-profile.children.cycles-pp.tick_nohz_tick_stopped
0.07 ± 93% +0.2 0.23 ± 25% perf-profile.children.cycles-pp.__evlist__disable
0.34 ± 29% +0.2 0.51 ± 13% perf-profile.children.cycles-pp.ct_kernel_enter
5.00 ± 9% +0.7 5.73 ± 8% perf-profile.children.cycles-pp.__irq_exit_rcu
1.90 ± 8% -0.3 1.61 ± 10% perf-profile.self.cycles-pp.update_sg_lb_stats
0.02 ±141% +0.1 0.10 ± 46% perf-profile.self.cycles-pp.__pte_offset_map
0.22 ± 36% +0.1 0.36 ± 19% perf-profile.self.cycles-pp.ct_kernel_enter
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
