Greeting,
FYI, we noticed a 7.7% improvement of stress-ng.netdev.ops_per_sec due to commit:
commit: b0e99d03778b2418aec20db99d97d19d25d198b6 ("net: socket: remove register_gifconf")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
in testcase: stress-ng
on test machine: 96 threads 2 sockets Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz with 192G memory
with following parameters:
nr_threads: 100%
testtime: 60s
class: network
test: netdev
cpufreq_governor: performance
ucode: 0x5003006
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
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
bin/lkp run generated-yaml-file
=========================================================================================
class/compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime/ucode:
network/gcc-9/performance/x86_64-rhel-8.3/100%/debian-10.4-x86_64-20200603.cgz/lkp-csl-2sp5/netdev/stress-ng/60s/0x5003006
commit:
709566d792 ("net: socket: rework SIOC?IFMAP ioctls")
b0e99d0377 ("net: socket: remove register_gifconf")
709566d79209af92 b0e99d03778b2418aec20db99d9
---------------- ---------------------------
%stddev %change %stddev
\ | \
209784 +7.7% 225992 stress-ng.netdev.ops_per_sec
1366 ? 30% -49.9% 684.83 ? 8% interrupts.CPU1.CAL:Function_call_interrupts
544.67 ? 2% +18.5% 645.33 ? 23% interrupts.CPU17.CAL:Function_call_interrupts
1.24 ? 2% -9.6% 1.12 ? 4% perf-stat.i.MPKI
38247636 -12.5% 33449096 ? 5% perf-stat.i.cache-misses
80968985 -11.3% 71785319 ? 4% perf-stat.i.cache-references
6534 +14.9% 7508 ? 6% perf-stat.i.cycles-between-cache-misses
7.916e+08 +6.4% 8.421e+08 perf-stat.i.dTLB-stores
953.20 -9.7% 861.00 ? 3% perf-stat.i.metric.K/sec
3871790 +3.8% 4019825 perf-stat.i.node-store-misses
1.17 -10.8% 1.04 ? 2% perf-stat.overall.MPKI
6666 +14.7% 7648 ? 5% perf-stat.overall.cycles-between-cache-misses
37655911 -12.5% 32934829 ? 5% perf-stat.ps.cache-misses
79759886 -11.3% 70715432 ? 4% perf-stat.ps.cache-references
7.79e+08 +6.4% 8.286e+08 perf-stat.ps.dTLB-stores
3812882 +3.8% 3958393 perf-stat.ps.node-store-misses
32.12 -0.2 31.89 perf-profile.calltrace.cycles-pp.osq_lock.__mutex_lock.sock_do_ioctl.sock_ioctl.__x64_sys_ioctl
32.45 -0.2 32.23 perf-profile.calltrace.cycles-pp.__mutex_lock.sock_do_ioctl.sock_ioctl.__x64_sys_ioctl.do_syscall_64
97.94 -0.3 97.60 perf-profile.children.cycles-pp.__mutex_lock
96.98 -0.3 96.64 perf-profile.children.cycles-pp.osq_lock
0.15 ? 3% -0.1 0.09 ? 5% perf-profile.children.cycles-pp.dev_ifconf
0.36 ? 2% -0.0 0.32 ? 2% perf-profile.children.cycles-pp.mutex_spin_on_owner
0.09 ? 5% +0.0 0.11 ? 4% perf-profile.children.cycles-pp.__might_sleep
0.15 ? 4% +0.0 0.16 ? 3% perf-profile.children.cycles-pp.__might_fault
0.14 ? 4% +0.0 0.16 ? 2% perf-profile.children.cycles-pp._copy_from_user
0.20 ? 3% +0.0 0.22 ? 2% perf-profile.children.cycles-pp.mutex_unlock
0.27 ? 4% +0.0 0.32 ? 4% perf-profile.children.cycles-pp.__rtnl_unlock
0.43 ? 3% +0.1 0.48 ? 2% perf-profile.children.cycles-pp.netdev_run_todo
0.11 ? 7% +0.1 0.24 ? 89% perf-profile.children.cycles-pp.clockevents_program_event
0.43 ? 4% +0.4 0.84 ? 93% perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt
0.43 ? 4% +0.4 0.84 ? 93% perf-profile.children.cycles-pp.hrtimer_interrupt
0.45 ? 4% +0.4 0.86 ? 91% perf-profile.children.cycles-pp.sysvec_apic_timer_interrupt
0.49 ? 4% +0.4 0.91 ? 87% perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt
96.51 -0.7 95.79 perf-profile.self.cycles-pp.osq_lock
0.36 ? 2% -0.0 0.32 ? 2% perf-profile.self.cycles-pp.mutex_spin_on_owner
0.20 ? 3% +0.0 0.22 perf-profile.self.cycles-pp.mutex_unlock
0.00 +0.1 0.05 ? 7% perf-profile.self.cycles-pp.inet_gifconf
stress-ng.netdev.ops_per_sec
235000 +------------------------------------------------------------------+
| |
230000 |-O O O O O O O O O O O O |
225000 |-+ O O O O O O O O O O O O O O O |
| O |
220000 |-+ |
| |
215000 |-+ |
| +. .+.+.+.+.+. |
210000 |-+ : +. +..+.+.+.|
205000 |.+.+.+..+ +. .+. : |
| + + .+. .. + + |
200000 |-+ +.+.+.+.+..+.+.+.+ + |
| |
195000 +------------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
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