| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 2 Jan 2020 17:08:47 +0800
From: kernel test robot <rong.a.chen@...el.com>
To: Ttttabcd <ttttabcd@...tonmail.com>
Cc: Netdev <netdev@...r.kernel.org>,
"edumazet@...gle.com" <edumazet@...gle.com>,
David Miller <davem@...emloft.net>,
"kuznet@....inr.ac.ru" <kuznet@....inr.ac.ru>,
"yoshfuji@...ux-ipv6.org" <yoshfuji@...ux-ipv6.org>,
lkp@...ts.01.org
Subject: [tcp] abda73240d: apachebench.requests_per_second 14.3% improvement
Greeting,
FYI, we noticed a 14.3% improvement of apachebench.requests_per_second due to commit:
commit: abda73240d0cd1deb65b98a94b36e7c3e3a27d7e ("[PATCH] tcp: Fix tcp_max_syn_backlog limit on connection requests")
url: https://github.com/0day-ci/linux/commits/Ttttabcd/tcp-Fix-tcp_max_syn_backlog-limit-on-connection-requests/20191230-164004
in testcase: apachebench
on test machine: 16 threads Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz with 48G memory
with following parameters:
runtime: 300s
concurrency: 8000
cluster: cs-localhost
cpufreq_governor: performance
ucode: 0x7000019
test-description: apachebench is a tool for benchmarking your Apache Hypertext Transfer Protocol (HTTP) server.
test-url: https://httpd.apache.org/docs/2.4/programs/ab.html
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 run job.yaml
=========================================================================================
cluster/compiler/concurrency/cpufreq_governor/kconfig/rootfs/runtime/tbox_group/testcase/ucode:
cs-localhost/gcc-7/8000/performance/x86_64-rhel-7.6/debian-x86_64-2019-11-14.cgz/300s/lkp-bdw-de1/apachebench/0x7000019
commit:
bb3d0b8bf5 ("net_sched: sch_fq: properly set sk->sk_pacing_status")
abda73240d ("tcp: Fix tcp_max_syn_backlog limit on connection requests")
bb3d0b8bf5be61ab abda73240d0cd1deb65b98a94b3
---------------- ---------------------------
%stddev %change %stddev
\ | \
7093 -56.7% 3068 apachebench.connection_time.connect.max
36705 ± 4% -5.5% 34685 apachebench.connection_time.processing.max
40733 -12.3% 35710 apachebench.connection_time.total.max
36705 ± 4% -5.5% 34685 apachebench.connection_time.waiting.max
40733 -5022.8 35710 apachebench.max_latency.100%
24494 +14.3% 27995 apachebench.requests_per_second
41.83 -12.2% 36.71 apachebench.time.elapsed_time
41.83 -12.2% 36.71 apachebench.time.elapsed_time.max
50.75 +13.3% 57.50 apachebench.time.percent_of_cpu_this_job_got
23640 ± 3% +16.4% 27519 ± 6% apachebench.time.voluntary_context_switches
326.62 -12.5% 285.76 apachebench.time_per_request
263070 +14.3% 300683 apachebench.transfer_rate
20.21 -0.8% 20.06 boot-time.dhcp
94754 ± 6% +55.9% 147705 ± 39% cpuidle.POLL.time
45698 ± 2% +12.0% 51170 ± 2% meminfo.max_used_kB
8807 -1.6% 8663 proc-vmstat.nr_kernel_stack
26737 -1.2% 26427 proc-vmstat.nr_slab_unreclaimable
1.63 ± 5% +0.2 1.82 ± 2% mpstat.cpu.all.soft%
10.27 ± 2% +1.5 11.73 ± 2% mpstat.cpu.all.sys%
12.92 ± 4% +1.7 14.61 ± 3% mpstat.cpu.all.usr%
75.25 -3.7% 72.50 vmstat.cpu.id
12.00 ± 5% +14.6% 13.75 ± 3% vmstat.cpu.us
42421 ± 2% +11.0% 47073 ± 2% vmstat.system.cs
907.86 ± 45% -56.2% 397.73 ± 25% sched_debug.cfs_rq:/.MIN_vruntime.avg
3031 ± 57% -49.2% 1540 ± 25% sched_debug.cfs_rq:/.MIN_vruntime.stddev
907.86 ± 45% -55.9% 400.22 ± 24% sched_debug.cfs_rq:/.max_vruntime.avg
3031 ± 57% -48.9% 1550 ± 24% sched_debug.cfs_rq:/.max_vruntime.stddev
527.25 ± 5% -12.3% 462.50 ± 6% slabinfo.file_lock_cache.active_objs
527.25 ± 5% -12.3% 462.50 ± 6% slabinfo.file_lock_cache.num_objs
1091 -13.3% 946.25 slabinfo.request_sock_TCP.active_objs
1091 -13.3% 946.25 slabinfo.request_sock_TCP.num_objs
675.25 ± 3% +11.6% 753.50 ± 3% turbostat.Avg_MHz
28.83 ± 3% +2.9 31.75 ± 3% turbostat.Busy%
1604577 ± 2% -16.6% 1338213 ± 6% turbostat.IRQ
25.12 +5.0% 26.39 turbostat.PkgWatt
21584 ± 15% -19.7% 17333 ± 5% softirqs.CPU1.TIMER
9678 ± 3% -21.3% 7619 ± 17% softirqs.CPU13.SCHED
35783 ± 14% +203.8% 108714 ± 92% softirqs.CPU14.NET_RX
116833 ± 21% -44.7% 64579 ± 39% softirqs.CPU6.NET_RX
17951 ± 7% +10.3% 19804 ± 3% softirqs.CPU6.RCU
18282 ± 4% +9.1% 19938 ± 4% softirqs.CPU9.RCU
87603 ± 2% -19.4% 70574 ± 7% interrupts.CPU0.LOC:Local_timer_interrupts
456.00 ± 14% +21.4% 553.50 ± 12% interrupts.CPU0.RES:Rescheduling_interrupts
87468 ± 3% -18.8% 71005 ± 8% interrupts.CPU1.LOC:Local_timer_interrupts
87542 ± 2% -19.0% 70869 ± 8% interrupts.CPU10.LOC:Local_timer_interrupts
87603 ± 2% -19.1% 70888 ± 8% interrupts.CPU11.LOC:Local_timer_interrupts
87021 ± 2% -19.1% 70430 ± 8% interrupts.CPU12.LOC:Local_timer_interrupts
87655 ± 2% -19.9% 70200 ± 9% interrupts.CPU13.LOC:Local_timer_interrupts
708.50 ± 37% -32.6% 477.75 ± 6% interrupts.CPU13.RES:Rescheduling_interrupts
87592 ± 2% -18.8% 71095 ± 7% interrupts.CPU14.LOC:Local_timer_interrupts
87597 ± 2% -19.1% 70874 ± 8% interrupts.CPU15.LOC:Local_timer_interrupts
87372 ± 2% -18.9% 70868 ± 8% interrupts.CPU2.LOC:Local_timer_interrupts
87502 ± 3% -19.1% 70768 ± 7% interrupts.CPU3.LOC:Local_timer_interrupts
87572 ± 2% -19.1% 70837 ± 7% interrupts.CPU4.LOC:Local_timer_interrupts
87617 ± 2% -19.1% 70847 ± 8% interrupts.CPU5.LOC:Local_timer_interrupts
87466 ± 3% -19.1% 70733 ± 7% interrupts.CPU6.LOC:Local_timer_interrupts
87382 ± 2% -19.0% 70811 ± 8% interrupts.CPU7.LOC:Local_timer_interrupts
87519 ± 3% -19.2% 70710 ± 8% interrupts.CPU8.LOC:Local_timer_interrupts
87484 ± 2% -19.0% 70884 ± 7% interrupts.CPU9.LOC:Local_timer_interrupts
4.00 ± 58% +968.8% 42.75 ±118% interrupts.CPU9.TLB:TLB_shootdowns
1400004 ± 2% -19.1% 1132398 ± 8% interrupts.LOC:Local_timer_interrupts
0.39 ± 13% -0.1 0.34 ± 2% perf-profile.children.cycles-pp.__cant_sleep
0.52 ± 2% -0.0 0.48 ± 3% perf-profile.children.cycles-pp.menu_select
0.36 ± 3% -0.0 0.33 ± 3% perf-profile.children.cycles-pp.ap_update_vhost_from_headers
0.13 ± 3% -0.0 0.11 ± 6% perf-profile.children.cycles-pp.apr_brigade_create
0.11 -0.0 0.09 ± 8% perf-profile.children.cycles-pp.timerqueue_del
0.08 ± 5% +0.0 0.10 ± 9% perf-profile.children.cycles-pp.__switch_to_asm
0.12 ± 5% +0.0 0.14 ± 6% perf-profile.children.cycles-pp.__x64_sys_close
0.25 ± 2% +0.0 0.27 ± 6% perf-profile.children.cycles-pp.vma_link
0.09 ± 11% +0.0 0.13 ± 24% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
0.34 ± 6% +0.0 0.38 ± 4% perf-profile.children.cycles-pp.thread_group_cputime_adjusted
0.39 ± 3% +0.0 0.42 ± 2% perf-profile.children.cycles-pp.irq_exit
0.17 ± 7% +0.0 0.21 ± 3% perf-profile.children.cycles-pp.rcu_core
0.09 +0.0 0.13 ± 26% perf-profile.children.cycles-pp.finish_task_switch
0.01 ±173% +0.0 0.06 ± 9% perf-profile.children.cycles-pp.ap_discard_request_body
1.97 +0.1 2.05 ± 3% perf-profile.children.cycles-pp.ksys_mmap_pgoff
2.75 +0.1 2.89 ± 3% perf-profile.children.cycles-pp.__sched_text_start
0.01 ±173% +0.2 0.23 ±101% perf-profile.children.cycles-pp.set_memory_ro
0.04 ±102% +0.3 0.34 ±109% perf-profile.children.cycles-pp.bpf_prepare_filter
0.03 ±102% +0.3 0.34 ±109% perf-profile.children.cycles-pp.bpf_prog_select_runtime
0.03 ±102% +0.3 0.34 ±109% perf-profile.children.cycles-pp.bpf_int_jit_compile
0.04 ±103% +0.3 0.34 ±107% perf-profile.children.cycles-pp.do_seccomp
0.04 ±103% +0.3 0.34 ±106% perf-profile.children.cycles-pp.prctl
0.04 ±103% +0.3 0.34 ±106% perf-profile.children.cycles-pp.__x64_sys_prctl
0.04 ±102% +0.3 0.34 ±107% perf-profile.children.cycles-pp.bpf_prog_create_from_user
0.02 ±173% +0.3 0.33 ±113% perf-profile.children.cycles-pp.change_page_attr_set_clr
0.31 ± 5% -0.0 0.27 ± 2% perf-profile.self.cycles-pp.__cant_sleep
0.17 ± 8% -0.0 0.14 ± 5% perf-profile.self.cycles-pp.tcp_clean_rtx_queue
0.07 ± 10% +0.0 0.08 ± 5% perf-profile.self.cycles-pp.vfs_read
0.08 ± 5% +0.0 0.09 ± 4% perf-profile.self.cycles-pp.__switch_to_asm
0.05 ± 8% +0.0 0.07 ± 14% perf-profile.self.cycles-pp.finish_task_switch
0.11 ± 4% +0.0 0.14 ± 7% perf-profile.self.cycles-pp.syscall_trace_enter
0.09 ± 11% +0.0 0.13 ± 24% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
0.01 ±173% +0.0 0.06 ± 11% perf-profile.self.cycles-pp.apr_array_cat
9.752e+08 ± 2% +12.3% 1.096e+09 ± 2% perf-stat.i.branch-instructions
47821007 ± 4% +12.3% 53713491 ± 4% perf-stat.i.branch-misses
3.281e+08 ± 2% +12.0% 3.675e+08 ± 2% perf-stat.i.cache-misses
3.281e+08 ± 2% +12.0% 3.675e+08 ± 2% perf-stat.i.cache-references
45580 ± 2% +10.6% 50410 ± 2% perf-stat.i.context-switches
3.72 -3.9% 3.57 ± 3% perf-stat.i.cpi
1.077e+10 ± 2% +11.0% 1.195e+10 ± 2% perf-stat.i.cpu-cycles
54.33 ± 3% -9.4% 49.24 ± 3% perf-stat.i.cycles-between-cache-misses
7489450 ± 2% +8.3% 8110568 ± 2% perf-stat.i.dTLB-load-misses
1.331e+09 ± 3% +14.4% 1.523e+09 ± 3% perf-stat.i.dTLB-loads
1177728 ± 3% +12.6% 1325547 ± 7% perf-stat.i.dTLB-store-misses
7.693e+08 +10.8% 8.521e+08 perf-stat.i.dTLB-stores
7966949 ± 4% +15.2% 9177912 ± 4% perf-stat.i.iTLB-load-misses
6080309 ± 4% +9.7% 6672092 ± 3% perf-stat.i.iTLB-loads
4.704e+09 ± 2% +12.4% 5.289e+09 ± 2% perf-stat.i.instructions
1054 ± 10% +14.5% 1207 ± 4% perf-stat.i.instructions-per-iTLB-miss
0.35 +9.2% 0.38 perf-stat.i.ipc
30946 +11.1% 34367 ± 2% perf-stat.i.minor-faults
30946 +11.1% 34367 ± 2% perf-stat.i.page-faults
2.29 -1.3% 2.26 perf-stat.overall.cpi
0.56 -0.0 0.53 perf-stat.overall.dTLB-load-miss-rate%
9.532e+08 ± 2% +12.1% 1.068e+09 ± 2% perf-stat.ps.branch-instructions
46750366 ± 4% +12.0% 52371630 ± 4% perf-stat.ps.branch-misses
3.205e+08 ± 2% +11.8% 3.581e+08 ± 2% perf-stat.ps.cache-misses
3.205e+08 ± 2% +11.8% 3.581e+08 ± 2% perf-stat.ps.cache-references
44524 ± 2% +10.3% 49126 ± 2% perf-stat.ps.context-switches
1.052e+10 ± 2% +10.7% 1.165e+10 ± 2% perf-stat.ps.cpu-cycles
7313092 ± 2% +8.0% 7900741 ± 3% perf-stat.ps.dTLB-load-misses
1.301e+09 ± 3% +14.1% 1.484e+09 ± 3% perf-stat.ps.dTLB-loads
1150200 ± 3% +12.3% 1291517 ± 7% perf-stat.ps.dTLB-store-misses
7.517e+08 +10.5% 8.305e+08 perf-stat.ps.dTLB-stores
7781460 ± 4% +14.9% 8943399 ± 4% perf-stat.ps.iTLB-load-misses
5941419 ± 4% +9.5% 6504269 ± 3% perf-stat.ps.iTLB-loads
4.598e+09 ± 2% +12.1% 5.157e+09 ± 2% perf-stat.ps.instructions
30234 +10.8% 33497 ± 2% perf-stat.ps.minor-faults
30234 +10.8% 33497 ± 2% perf-stat.ps.page-faults
apachebench.time.percent_of_cpu_this_job_got
58 O-+----------------O---------O-O------O-O------------------------------+
| |
57 +-O O O O O O O O O O O O O |
56 +-+ O |
| |
55 +-+ |
54 +-+ |
| |
53 +-+ |
52 +-+ + |
| + + |
51 +-+.+.+.+.+..+.+.+. .+.+.+. .+ +. .+. .+.+.+. .|
50 +-+ +.+ +.+.+.+.+..+.+ + + +..+ + |
| + + |
49 +-+--------------------------------------------------------------------+
apachebench.time.elapsed_time
43 +-+--------------------------------------------------------------------+
| +. .+.+ .+. .+. .+. .+ |
42 +-+ +.+.+.+..+ + .+ + + + +.+.. .+.+ + .|
41 +-+ +.+ +.+. .+.+.. .+ + + + + |
| + + + + + |
40 +-+ + |
| |
39 +-+ |
| |
38 +-+ |
37 +-+ |
O O O O O O O O O O O O O O O |
36 +-+ O O O O |
| O |
35 +-+--------------------------------------------------------------------+
apachebench.time.elapsed_time.max
43 +-+--------------------------------------------------------------------+
| +. .+.+ .+. .+. .+. .+ |
42 +-+ +.+.+.+..+ + .+ + + + +.+.. .+.+ + .|
41 +-+ +.+ +.+. .+.+.. .+ + + + + |
| + + + + + |
40 +-+ + |
| |
39 +-+ |
| |
38 +-+ |
37 +-+ |
O O O O O O O O O O O O O O O |
36 +-+ O O O O |
| O |
35 +-+--------------------------------------------------------------------+
apachebench.requests_per_second
29000 +-+-----------------------------------------------------------------+
28500 +-+ O O O |
| O O O O O |
28000 O-+ O O O O O O O O O O O |
27500 +-+ |
| |
27000 +-+ |
26500 +-+ |
26000 +-+ |
| |
25500 +-+ .+ |
25000 +-+ +.+ + .+.+.+.+.+.+ +. .+ |
|. .+. .+. + + + + + +. .+ + .+.|
24500 +-+.+.+ + +.+ +.+.+ + + +.+ +.+.+ |
24000 +-+-----------------------------------------------------------------+
apachebench.time_per_request
340 +-+-------------------------------------------------------------------+
| |
330 +-+.+.+. .+.+.. .+. .+. .+.+. .+.+. |
| +.+.+ +. .+ +. .+ +. +. .+ +.|
320 +-+ +.+. + +.+.+.+.+.+ + + |
| + |
310 +-+ |
| |
300 +-+ |
| |
290 O-+ |
| O O O O O O O O O O O O O O O O |
280 +-+ O O O |
| |
270 +-+-------------------------------------------------------------------+
apachebench.transfer_rate
310000 +-+-------O--------------------------------------------------------+
| O O O O |
300000 +-O O O O O O O O O O O OO O |
O |
| |
290000 +-+ |
| |
280000 +-+ |
| |
270000 +-+ .+. |
|. +.+ +.+.+.+.+.++. .+. .+. .+.|
| +.+.+.+.+.+. + +.+.+ +. .+. .+.+ +.+.+ |
260000 +-+ +.+ + + |
| |
250000 +-+----------------------------------------------------------------+
apachebench.connection_time.total.max
42000 +-+-----------------------------------------------------------------+
| .+.+ .+. .+. |
41000 +-+.+ +.+.+ : .+.+.+ + + +. .+.+.+. .|
40000 +-+ : : : .+ +. .+. .+ + + +.+ + |
| : : + + + +.+ +.+ + |
39000 +-+ : : + |
| : : |
38000 +-+ : : |
| :: |
37000 +-+ : |
36000 +-+ + |
O O O O O O O O O O O O O O O |
35000 +-+ O O O O |
| O |
34000 +-+-----------------------------------------------------------------+
apachebench.max_latency.100_
42000 +-+-----------------------------------------------------------------+
| .+.+ .+. .+. |
41000 +-+.+ +.+.+ : .+.+.+ + + +. .+.+.+. .|
40000 +-+ : : : .+ +. .+. .+ + + +.+ + |
| : : + + + +.+ +.+ + |
39000 +-+ : : + |
| : : |
38000 +-+ : : |
| :: |
37000 +-+ : |
36000 +-+ + |
O O O O O O O O O O O O O O O |
35000 +-+ O O O O |
| O |
34000 +-+-----------------------------------------------------------------+
[*] 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.
Thanks,
Rong Chen
View attachment "config-5.5.0-rc2-00667-gabda73240d0cd" of type "text/plain" (202322 bytes)
View attachment "job-script" of type "text/plain" (8051 bytes)
View attachment "job.yaml" of type "text/plain" (5551 bytes)
View attachment "reproduce" of type "text/plain" (460 bytes)
Powered by blists - more mailing lists