Greeting,
FYI, we noticed a 2.4% improvement of unixbench.score due to commit:
commit: 85c62c8c3749eec02ba81217bdcac26867dc262e ("tracing: Have existing event_command.parse() implementations use helpers")
https://git.kernel.org/cgit/linux/kernel/git/rostedt/linux-trace.git ftrace/core
in testcase: unixbench
on test machine: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz with 256G memory
with following parameters:
runtime: 300s
nr_task: 30%
test: pipe
cpufreq_governor: performance
ucode: 0xd000280
test-description: UnixBench is the original BYTE UNIX benchmark suite aims to test performance of Unix-like system.
test-url: https://github.com/kdlucas/byte-unixbench
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
sudo bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml # generate the yaml file for lkp run
sudo bin/lkp run generated-yaml-file
# if come across any failure that blocks the test,
# please remove ~/.lkp and /lkp dir to run from a clean state.
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_task/rootfs/runtime/tbox_group/test/testcase/ucode:
gcc-9/performance/x86_64-rhel-8.3/30%/debian-10.4-x86_64-20200603.cgz/300s/lkp-icl-2sp2/pipe/unixbench/0xd000280
commit:
70b3c6b691 ("tracing: Add helper functions to simplify event_command.parse() callback handling")
85c62c8c37 ("tracing: Have existing event_command.parse() implementations use helpers")
70b3c6b69157039d 85c62c8c3749eec02ba81217bdc
---------------- ---------------------------
%stddev %change %stddev
\ | \
60620 +2.4% 62065 unixbench.score
2.947e+10 +2.3% 3.016e+10 unixbench.workload
36554 ? 7% -10.8% 32602 ? 4% softirqs.CPU123.SCHED
46257 ? 4% -10.3% 41506 ? 7% softirqs.CPU7.SCHED
32380 ? 12% +14.7% 37134 ? 6% softirqs.CPU81.SCHED
2.794e+10 +2.3% 2.858e+10 perf-stat.i.branch-instructions
13343338 -23.8% 10167419 ? 2% perf-stat.i.branch-misses
662823 -17.5% 546757 perf-stat.i.cache-misses
7192051 -20.0% 5751089 ? 2% perf-stat.i.cache-references
129484 +35.1% 174921 perf-stat.i.cycles-between-cache-misses
0.00 ? 2% -0.0 0.00 ? 2% perf-stat.i.dTLB-load-miss-rate%
665427 -46.7% 354693 perf-stat.i.dTLB-load-misses
4.435e+10 +2.3% 4.537e+10 perf-stat.i.dTLB-loads
2.665e+10 +2.3% 2.726e+10 perf-stat.i.dTLB-stores
1.438e+11 +2.3% 1.471e+11 perf-stat.i.instructions
1.58 +2.0% 1.62 perf-stat.i.ipc
59.02 -20.2% 47.12 ? 2% perf-stat.i.metric.K/sec
772.85 +2.3% 790.56 perf-stat.i.metric.M/sec
89.73 -2.2 87.53 perf-stat.i.node-load-miss-rate%
128386 -21.9% 100239 ? 2% perf-stat.i.node-load-misses
88458 ? 5% -34.6% 57889 ? 8% perf-stat.i.node-store-misses
0.05 -21.8% 0.04 ? 2% perf-stat.overall.MPKI
0.05 -0.0 0.04 ? 2% perf-stat.overall.branch-miss-rate%
0.54 -2.3% 0.53 perf-stat.overall.cpi
117918 +21.1% 142796 perf-stat.overall.cycles-between-cache-misses
0.00 -0.0 0.00 perf-stat.overall.dTLB-load-miss-rate%
1.84 +2.4% 1.89 perf-stat.overall.ipc
86.82 -3.5 83.34 perf-stat.overall.node-load-miss-rate%
48.41 ? 6% -9.0 39.39 ? 8% perf-stat.overall.node-store-miss-rate%
2.794e+10 +2.3% 2.857e+10 perf-stat.ps.branch-instructions
13329890 -23.8% 10154768 ? 2% perf-stat.ps.branch-misses
661626 -17.5% 545780 perf-stat.ps.cache-misses
7176900 -20.1% 5735850 ? 2% perf-stat.ps.cache-references
665319 -46.7% 354481 perf-stat.ps.dTLB-load-misses
4.435e+10 +2.3% 4.535e+10 perf-stat.ps.dTLB-loads
2.665e+10 +2.3% 2.725e+10 perf-stat.ps.dTLB-stores
1.438e+11 +2.3% 1.47e+11 perf-stat.ps.instructions
128207 -21.9% 100146 ? 2% perf-stat.ps.node-load-misses
88398 ? 5% -34.6% 57822 ? 8% perf-stat.ps.node-store-misses
5.652e+13 +2.3% 5.785e+13 perf-stat.total.instructions
3.19 -0.3 2.86 perf-profile.calltrace.cycles-pp.security_file_permission.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
2.91 -0.3 2.62 perf-profile.calltrace.cycles-pp.security_file_permission.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
6.37 -0.2 6.13 perf-profile.calltrace.cycles-pp.__entry_text_start.read
33.05 -0.2 32.82 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
1.86 ? 2% -0.2 1.66 perf-profile.calltrace.cycles-pp.common_file_perm.security_file_permission.vfs_read.ksys_read.do_syscall_64
1.84 -0.2 1.66 perf-profile.calltrace.cycles-pp.common_file_perm.security_file_permission.vfs_write.ksys_write.do_syscall_64
1.63 -0.1 1.53 ? 2% perf-profile.calltrace.cycles-pp._raw_spin_lock_irq.pipe_write.new_sync_write.vfs_write.ksys_write
1.99 -0.1 1.91 perf-profile.calltrace.cycles-pp.main
0.85 ? 2% -0.1 0.77 ? 2% perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write
1.26 -0.0 1.22 ? 2% perf-profile.calltrace.cycles-pp.current_time.file_update_time.pipe_write.new_sync_write.vfs_write
1.12 +0.0 1.16 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
1.73 +0.1 1.80 ? 2% perf-profile.calltrace.cycles-pp.__fget_light.__fdget_pos.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
2.91 +0.1 2.98 perf-profile.calltrace.cycles-pp.mutex_lock.pipe_read.new_sync_read.vfs_read.ksys_read
4.16 +0.1 4.23 perf-profile.calltrace.cycles-pp.atime_needs_update.touch_atime.pipe_read.new_sync_read.vfs_read
2.02 +0.1 2.10 perf-profile.calltrace.cycles-pp.__fdget_pos.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
4.89 +0.1 4.99 perf-profile.calltrace.cycles-pp.touch_atime.pipe_read.new_sync_read.vfs_read.ksys_read
2.42 +0.1 2.53 perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write
4.50 +0.1 4.63 perf-profile.calltrace.cycles-pp.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write.ksys_write
37.34 +0.3 37.63 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
38.45 +0.3 38.76 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read
24.57 +0.4 24.93 perf-profile.calltrace.cycles-pp.pipe_read.new_sync_read.vfs_read.ksys_read.do_syscall_64
26.85 +0.4 27.25 perf-profile.calltrace.cycles-pp.new_sync_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +0.5 0.52 perf-profile.calltrace.cycles-pp._raw_spin_unlock_irqrestore.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write
6.40 -0.6 5.79 perf-profile.children.cycles-pp.security_file_permission
4.12 -0.4 3.76 perf-profile.children.cycles-pp.common_file_perm
50.62 -0.3 50.34 perf-profile.children.cycles-pp.write
33.42 -0.2 33.22 perf-profile.children.cycles-pp.vfs_write
7.06 -0.2 6.85 perf-profile.children.cycles-pp.__entry_text_start
0.90 ? 2% -0.1 0.81 perf-profile.children.cycles-pp.__wake_up_common
2.14 -0.1 2.06 perf-profile.children.cycles-pp.main
1.52 -0.1 1.46 perf-profile.children.cycles-pp.__cond_resched
1.09 -0.0 1.05 perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
0.13 ? 5% -0.0 0.11 perf-profile.children.cycles-pp.write@plt
1.20 +0.0 1.23 perf-profile.children.cycles-pp.exit_to_user_mode_prepare
0.56 +0.0 0.61 perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
2.67 +0.1 2.73 perf-profile.children.cycles-pp.mutex_unlock
2.70 +0.1 2.77 perf-profile.children.cycles-pp.syscall_exit_to_user_mode
4.58 +0.1 4.68 perf-profile.children.cycles-pp.atime_needs_update
5.11 +0.1 5.22 perf-profile.children.cycles-pp.touch_atime
2.54 +0.1 2.66 perf-profile.children.cycles-pp._raw_spin_lock_irqsave
6.30 +0.1 6.41 perf-profile.children.cycles-pp.mutex_lock
4.87 +0.1 5.00 perf-profile.children.cycles-pp.__wake_up_common_lock
25.30 +0.4 25.65 perf-profile.children.cycles-pp.pipe_read
27.07 +0.4 27.48 perf-profile.children.cycles-pp.new_sync_read
2.44 ? 2% -0.4 2.08 perf-profile.self.cycles-pp.common_file_perm
1.98 ? 2% -0.3 1.72 perf-profile.self.cycles-pp.security_file_permission
3.06 -0.1 2.95 perf-profile.self.cycles-pp.write
1.62 -0.1 1.53 ? 2% perf-profile.self.cycles-pp.__entry_text_start
3.89 -0.1 3.79 perf-profile.self.cycles-pp.pipe_write
1.99 -0.1 1.91 perf-profile.self.cycles-pp.main
1.72 -0.1 1.65 perf-profile.self.cycles-pp.copy_page_from_iter
2.22 -0.1 2.17 perf-profile.self.cycles-pp.copy_page_to_iter
0.78 ? 2% -0.0 0.73 perf-profile.self.cycles-pp.__wake_up_common
0.32 ? 3% -0.0 0.29 ? 4% perf-profile.self.cycles-pp.__might_fault
0.22 ? 3% +0.0 0.24 perf-profile.self.cycles-pp.make_kgid
1.56 +0.0 1.60 perf-profile.self.cycles-pp.new_sync_write
1.92 +0.0 1.97 perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
1.54 +0.1 1.60 perf-profile.self.cycles-pp.new_sync_read
3.36 +0.1 3.46 perf-profile.self.cycles-pp.mutex_lock
2.46 +0.1 2.57 perf-profile.self.cycles-pp._raw_spin_lock_irqsave
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.
---
0DAY/LKP+ Test Infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/[email protected] Intel Corporation
Thanks,
Oliver Sang