Hello,
kernel test robot noticed a 16.6% improvement of filebench.sum_operations/s on:
commit: 26a80762153ba0dc98258b5e6d2e9741178c5114 ("NFSD: Add a Kconfig setting to enable delegated timestamps")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
testcase: filebench
config: x86_64-rhel-9.4
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
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250529/202505291023.d4c802b1-lkp@xxxxxxxxx
=========================================================================================
compiler/cpufreq_governor/disk/fs2/fs/kconfig/rootfs/tbox_group/test/testcase:
gcc-12/performance/1HDD/nfsv4/ext4/x86_64-rhel-9.4/debian-12-x86_64-20240206.cgz/lkp-icl-2sp6/webproxy.f/filebench
commit:
87480a8ce5 ("sysctl: Fixes nsm_local_state bounds")
26a8076215 ("NFSD: Add a Kconfig setting to enable delegated timestamps")
87480a8ce567340a 26a80762153ba0dc98258b5e6d2
---------------- ---------------------------
%stddev %change %stddev
\ | \
5114423 -11.9% 4505604 cpuidle..usage
212.82 ± 7% -12.6% 186.03 ± 3% sched_debug.cpu.curr->pid.avg
2922 -7.5% 2703 vmstat.system.in
184.48 ± 9% -17.4% 152.45 ± 5% perf-sched.wait_and_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
184.41 ± 9% -17.4% 152.39 ± 5% perf-sched.wait_time.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
0.60 +16.7% 0.70 filebench.sum_bytes_mb/s
7676 +16.6% 8950 filebench.sum_operations
127.92 +16.6% 149.15 filebench.sum_operations/s
33.00 +18.7% 39.17 filebench.sum_reads/s
745.09 -15.3% 630.77 filebench.sum_time_ms/op
7.00 +14.3% 8.00 filebench.sum_writes/s
1465 -6.5% 1370 filebench.time.elapsed_time
1465 -6.5% 1370 filebench.time.elapsed_time.max
41322 +2.3% 42260 filebench.time.voluntary_context_switches
73915 -1.2% 73000 proc-vmstat.nr_dirtied
68893 -4.8% 65555 ± 2% proc-vmstat.nr_inactive_file
21211 -3.3% 20518 proc-vmstat.nr_shmem
73782 -1.5% 72646 proc-vmstat.nr_written
68893 -4.8% 65555 ± 2% proc-vmstat.nr_zone_inactive_file
3148302 -5.2% 2985206 proc-vmstat.numa_hit
3015702 -5.4% 2852716 proc-vmstat.numa_local
3697834 -4.5% 3531265 proc-vmstat.pgalloc_normal
3722673 -6.0% 3499153 proc-vmstat.pgfault
3588865 -4.4% 3429561 proc-vmstat.pgfree
940490 -4.3% 899874 proc-vmstat.pgpgout
162192 -5.9% 152665 proc-vmstat.pgreuse
5.37 -0.1 5.25 perf-stat.i.branch-miss-rate%
1681345 -1.1% 1662216 perf-stat.i.branch-misses
11085725 -5.0% 10535577 perf-stat.i.cache-references
2.25 -1.6% 2.21 perf-stat.i.cpi
2.976e+08 -1.8% 2.922e+08 perf-stat.i.cpu-cycles
0.46 +1.6% 0.47 perf-stat.i.ipc
5.18 -0.1 5.08 perf-stat.overall.branch-miss-rate%
1.87 -2.6% 1.82 perf-stat.overall.cpi
0.54 +2.6% 0.55 perf-stat.overall.ipc
1678730 -1.1% 1659448 perf-stat.ps.branch-misses
11076863 -5.0% 10526572 perf-stat.ps.cache-references
2.972e+08 -1.8% 2.919e+08 perf-stat.ps.cpu-cycles
2.334e+11 -5.8% 2.198e+11 perf-stat.total.instructions
6.43 ± 5% -0.4 6.01 ± 3% perf-profile.children.cycles-pp.__schedule
1.11 ± 11% -0.2 0.87 ± 13% perf-profile.children.cycles-pp.try_to_block_task
0.94 ± 7% -0.2 0.76 ± 9% perf-profile.children.cycles-pp.copy_mc_enhanced_fast_string
0.18 ± 14% +0.1 0.27 ± 26% perf-profile.children.cycles-pp.set_pte_range
0.04 ±100% +0.1 0.14 ± 12% perf-profile.children.cycles-pp.xprt_sock_sendmsg
0.13 ± 28% +0.1 0.25 ± 25% perf-profile.children.cycles-pp.devkmsg_read
0.38 ± 14% +0.1 0.50 ± 9% perf-profile.children.cycles-pp.arch_scale_freq_tick
0.03 ±145% +0.1 0.15 ± 16% perf-profile.children.cycles-pp.scnprintf
0.12 ± 27% +0.1 0.24 ± 26% perf-profile.children.cycles-pp.printk_get_next_message
1.45 ± 8% +0.2 1.63 ± 5% perf-profile.children.cycles-pp.copy_process
0.93 ± 6% -0.2 0.74 ± 10% perf-profile.self.cycles-pp.copy_mc_enhanced_fast_string
0.21 ± 27% -0.1 0.11 ± 32% perf-profile.self.cycles-pp.hrtimer_update_next_event
0.38 ± 14% +0.1 0.50 ± 9% perf-profile.self.cycles-pp.arch_scale_freq_tick
0.21 ± 20% +0.1 0.34 ± 19% perf-profile.self.cycles-pp.tsc_verify_tsc_adjust
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
