| 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
| ||
|
Message-ID: <20190520063611.GC19312@shao2-debian>
Date: Mon, 20 May 2019 14:36:11 +0800
From: kernel test robot <rong.a.chen@...el.com>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: Borislav Petkov <bp@...e.de>, Jon Masters <jcm@...hat.com>,
LKML <linux-kernel@...r.kernel.org>,
Linus Torvalds <torvalds@...ux-foundation.org>, lkp@...org
Subject: [x86/speculation/mds] 22dd836508: will-it-scale.per_process_ops
-3.8% regression
Greeting,
FYI, we noticed a -3.8% regression of will-it-scale.per_process_ops due to commit:
commit: 22dd8365088b6403630b82423cf906491859b65e ("x86/speculation/mds: Add mitigation mode VMWERV")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
in testcase: will-it-scale
on test machine: 192 threads Skylake-SP with 256G memory
with following parameters:
nr_task: 100%
mode: process
test: getppid1
cpufreq_governor: performance
test-description: Will It Scale takes a testcase and runs it from 1 through to n parallel copies to see if the testcase will scale. It builds both a process and threads based test in order to see any differences between the two.
test-url: https://github.com/antonblanchard/will-it-scale
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
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase:
gcc-7/performance/x86_64-rhel-7.6/process/100%/debian-x86_64-2018-04-03.cgz/lkp-skl-4sp1/getppid1/will-it-scale
commit:
8a4b06d391 ("x86/speculation/mds: Add sysfs reporting for MDS")
22dd836508 ("x86/speculation/mds: Add mitigation mode VMWERV")
8a4b06d391b0a42a 22dd8365088b6403630b82423cf
---------------- ---------------------------
%stddev %change %stddev
\ | \
2728130 -3.8% 2623420 will-it-scale.per_process_ops
5.238e+08 -3.8% 5.037e+08 will-it-scale.workload
68815 ± 10% +473.5% 394687 ±136% cpuidle.C1.time
22197 ± 35% +88.8% 41910 ± 26% numa-numastat.node2.other_node
55.00 -3.2% 53.25 vmstat.cpu.sy
41.50 +4.2% 43.25 vmstat.cpu.us
2614 ± 55% +304.4% 10572 ± 68% numa-meminfo.node1.Inactive
1580 ± 63% +569.1% 10571 ± 68% numa-meminfo.node1.Inactive(anon)
18890 ±159% +154.2% 48017 ± 61% numa-meminfo.node1.Shmem
394.75 ± 63% +567.0% 2633 ± 69% numa-vmstat.node1.nr_inactive_anon
4722 ±159% +154.2% 12003 ± 61% numa-vmstat.node1.nr_shmem
394.75 ± 63% +567.0% 2633 ± 69% numa-vmstat.node1.nr_zone_inactive_anon
106360 ± 7% +18.7% 126214 ± 8% numa-vmstat.node2.numa_other
24497 ± 4% -9.0% 22294 ± 2% softirqs.CPU10.RCU
24475 ± 5% -9.2% 22230 ± 2% softirqs.CPU11.RCU
24738 ± 3% -8.2% 22701 ± 2% softirqs.CPU12.RCU
25279 ± 5% -9.3% 22924 ± 2% softirqs.CPU13.RCU
25111 ± 4% -9.3% 22766 softirqs.CPU15.RCU
26486 ± 3% -8.5% 24243 ± 8% softirqs.CPU168.RCU
24739 ± 6% -7.0% 23004 ± 7% softirqs.CPU171.RCU
24353 ± 4% -7.8% 22451 ± 3% softirqs.CPU3.RCU
24312 ± 4% -7.0% 22618 ± 3% softirqs.CPU4.RCU
24502 ± 4% -8.5% 22429 ± 3% softirqs.CPU5.RCU
24736 ± 3% -8.2% 22718 ± 2% softirqs.CPU7.RCU
24243 ± 4% -8.1% 22278 ± 2% softirqs.CPU9.RCU
46.04 -1.6 44.41 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.getppid
33.29 -1.2 32.10 perf-profile.calltrace.cycles-pp.syscall_return_via_sysret.getppid
3.08 -0.2 2.89 perf-profile.calltrace.cycles-pp.testcase
98.42 +0.1 98.56 perf-profile.calltrace.cycles-pp.getppid
14.84 ± 2% +2.1 16.97 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.getppid
17.02 ± 2% +2.9 19.92 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.getppid
41.12 -1.4 39.67 perf-profile.children.cycles-pp.entry_SYSCALL_64
38.91 -1.4 37.52 perf-profile.children.cycles-pp.syscall_return_via_sysret
2.16 -0.1 2.02 perf-profile.children.cycles-pp.testcase
0.53 ± 2% -0.0 0.50 ± 2% perf-profile.children.cycles-pp.__x86_indirect_thunk_rax
98.92 +0.1 99.01 perf-profile.children.cycles-pp.getppid
14.89 ± 2% +2.3 17.19 perf-profile.children.cycles-pp.do_syscall_64
17.08 ± 2% +3.0 20.06 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
38.91 -1.5 37.44 perf-profile.self.cycles-pp.syscall_return_via_sysret
36.15 -1.2 34.90 perf-profile.self.cycles-pp.entry_SYSCALL_64
6.12 -0.2 5.87 perf-profile.self.cycles-pp.getppid
1.01 -0.1 0.93 perf-profile.self.cycles-pp.testcase
0.52 ± 2% -0.0 0.49 ± 2% perf-profile.self.cycles-pp.__x86_indirect_thunk_rax
0.18 ± 2% +0.0 0.21 ± 3% perf-profile.self.cycles-pp.getppid@plt
2.16 +0.7 2.86 perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
2.65 +2.6 5.23 perf-profile.self.cycles-pp.do_syscall_64
1.903e+10 -6.5% 1.78e+10 perf-stat.i.branch-instructions
2.81 +0.1 2.91 ± 2% perf-stat.i.branch-miss-rate%
5.336e+08 -3.8% 5.135e+08 perf-stat.i.branch-misses
4.01 +3.8% 4.17 perf-stat.i.cpi
3.47e+10 -2.4% 3.386e+10 perf-stat.i.dTLB-loads
2.102e+10 -3.8% 2.022e+10 perf-stat.i.dTLB-stores
5.212e+08 -2.1% 5.105e+08 perf-stat.i.iTLB-load-misses
1.034e+11 -3.3% 9.994e+10 perf-stat.i.instructions
0.25 -3.4% 0.24 perf-stat.i.ipc
4198 -7.7% 3874 perf-stat.i.minor-faults
4203 -7.8% 3874 perf-stat.i.page-faults
2.80 +0.1 2.89 perf-stat.overall.branch-miss-rate%
4.01 +3.5% 4.15 perf-stat.overall.cpi
198.43 -1.3% 195.78 perf-stat.overall.instructions-per-iTLB-miss
0.25 -3.4% 0.24 perf-stat.overall.ipc
1.897e+10 -6.5% 1.774e+10 perf-stat.ps.branch-instructions
5.319e+08 -3.8% 5.118e+08 perf-stat.ps.branch-misses
3.458e+10 -2.4% 3.374e+10 perf-stat.ps.dTLB-loads
2.095e+10 -3.8% 2.014e+10 perf-stat.ps.dTLB-stores
5.193e+08 -2.0% 5.087e+08 perf-stat.ps.iTLB-load-misses
1.03e+11 -3.4% 9.959e+10 perf-stat.ps.instructions
4203 -8.1% 3864 perf-stat.ps.minor-faults
4208 -8.2% 3864 perf-stat.ps.page-faults
3.149e+13 -3.3% 3.044e+13 perf-stat.total.instructions
1159663 +7.1% 1241742 ± 6% interrupts.CAL:Function_call_interrupts
6042 +7.7% 6504 ± 6% interrupts.CPU0.CAL:Function_call_interrupts
6066 +7.4% 6514 ± 5% interrupts.CPU1.CAL:Function_call_interrupts
6059 +7.5% 6511 ± 5% interrupts.CPU10.CAL:Function_call_interrupts
6066 +7.1% 6496 ± 6% interrupts.CPU100.CAL:Function_call_interrupts
6067 +7.0% 6494 ± 6% interrupts.CPU101.CAL:Function_call_interrupts
6063 +7.1% 6493 ± 6% interrupts.CPU102.CAL:Function_call_interrupts
6065 +7.1% 6494 ± 6% interrupts.CPU103.CAL:Function_call_interrupts
6064 +7.1% 6495 ± 6% interrupts.CPU104.CAL:Function_call_interrupts
6032 +7.6% 6492 ± 6% interrupts.CPU105.CAL:Function_call_interrupts
6065 +7.0% 6487 ± 6% interrupts.CPU112.CAL:Function_call_interrupts
6068 +7.0% 6492 ± 6% interrupts.CPU114.CAL:Function_call_interrupts
6074 +7.0% 6500 ± 6% interrupts.CPU117.CAL:Function_call_interrupts
6065 +7.5% 6517 ± 5% interrupts.CPU119.CAL:Function_call_interrupts
6051 +7.7% 6519 ± 5% interrupts.CPU12.CAL:Function_call_interrupts
5863 ± 5% +10.6% 6487 ± 6% interrupts.CPU120.CAL:Function_call_interrupts
6059 +7.1% 6492 ± 6% interrupts.CPU123.CAL:Function_call_interrupts
6077 +7.0% 6502 ± 6% interrupts.CPU124.CAL:Function_call_interrupts
86.00 ± 45% -73.8% 22.50 ± 27% interrupts.CPU124.RES:Rescheduling_interrupts
6074 +7.0% 6501 ± 6% interrupts.CPU125.CAL:Function_call_interrupts
6070 +7.0% 6493 ± 6% interrupts.CPU126.CAL:Function_call_interrupts
6064 +7.8% 6536 ± 7% interrupts.CPU128.CAL:Function_call_interrupts
6051 +7.6% 6511 ± 5% interrupts.CPU13.CAL:Function_call_interrupts
206.00 ±111% -94.5% 11.25 ± 51% interrupts.CPU137.RES:Rescheduling_interrupts
6046 +7.6% 6508 ± 5% interrupts.CPU14.CAL:Function_call_interrupts
21.75 ± 40% +402.3% 109.25 ± 51% interrupts.CPU141.RES:Rescheduling_interrupts
6056 +7.5% 6510 ± 5% interrupts.CPU155.CAL:Function_call_interrupts
1498 ± 99% -99.3% 10.25 ± 66% interrupts.CPU159.RES:Rescheduling_interrupts
6041 +7.7% 6503 ± 5% interrupts.CPU16.CAL:Function_call_interrupts
6067 +7.1% 6500 ± 6% interrupts.CPU164.CAL:Function_call_interrupts
241.25 ±134% -90.6% 22.75 ± 95% interrupts.CPU165.RES:Rescheduling_interrupts
283.75 ± 93% -75.8% 68.75 ±101% interrupts.CPU168.RES:Rescheduling_interrupts
6047 +7.5% 6499 ± 5% interrupts.CPU17.CAL:Function_call_interrupts
47.50 ± 69% +788.4% 422.00 ± 76% interrupts.CPU170.RES:Rescheduling_interrupts
22.25 ± 18% +497.8% 133.00 ± 94% interrupts.CPU174.RES:Rescheduling_interrupts
6060 +7.3% 6501 ± 5% interrupts.CPU18.CAL:Function_call_interrupts
6036 +7.6% 6496 ± 6% interrupts.CPU181.CAL:Function_call_interrupts
6073 +6.9% 6491 ± 6% interrupts.CPU182.CAL:Function_call_interrupts
5516 ± 17% +17.8% 6497 ± 6% interrupts.CPU183.CAL:Function_call_interrupts
6062 +7.3% 6505 ± 6% interrupts.CPU189.CAL:Function_call_interrupts
6063 +7.4% 6510 ± 6% interrupts.CPU190.CAL:Function_call_interrupts
6054 +7.5% 6506 ± 6% interrupts.CPU191.CAL:Function_call_interrupts
6068 +7.3% 6510 ± 5% interrupts.CPU2.CAL:Function_call_interrupts
6065 +7.1% 6496 ± 5% interrupts.CPU21.CAL:Function_call_interrupts
6062 +7.9% 6541 ± 5% interrupts.CPU23.CAL:Function_call_interrupts
5669 ± 10% +15.6% 6554 ± 6% interrupts.CPU24.CAL:Function_call_interrupts
6062 +7.3% 6506 ± 5% interrupts.CPU3.CAL:Function_call_interrupts
347.50 ±115% -89.4% 36.75 ±112% interrupts.CPU35.RES:Rescheduling_interrupts
218.50 ± 72% -83.9% 35.25 ± 79% interrupts.CPU39.RES:Rescheduling_interrupts
6065 +7.3% 6509 ± 5% interrupts.CPU4.CAL:Function_call_interrupts
144.75 ± 59% -77.5% 32.50 ± 71% interrupts.CPU41.RES:Rescheduling_interrupts
6014 +8.2% 6507 ± 5% interrupts.CPU48.CAL:Function_call_interrupts
605.25 ± 61% -79.3% 125.50 ±146% interrupts.CPU48.RES:Rescheduling_interrupts
6063 +7.3% 6506 ± 5% interrupts.CPU5.CAL:Function_call_interrupts
6068 +7.4% 6517 ± 6% interrupts.CPU52.CAL:Function_call_interrupts
325.25 ± 83% -90.2% 32.00 ±100% interrupts.CPU52.RES:Rescheduling_interrupts
6060 +7.3% 6502 ± 6% interrupts.CPU54.CAL:Function_call_interrupts
6058 +7.3% 6497 ± 6% interrupts.CPU55.CAL:Function_call_interrupts
6057 +7.4% 6504 ± 6% interrupts.CPU57.CAL:Function_call_interrupts
564.25 ± 99% -90.4% 54.00 ±121% interrupts.CPU57.RES:Rescheduling_interrupts
6047 +7.7% 6512 ± 6% interrupts.CPU58.CAL:Function_call_interrupts
6042 +7.6% 6504 ± 6% interrupts.CPU59.CAL:Function_call_interrupts
6062 +7.5% 6515 ± 5% interrupts.CPU6.CAL:Function_call_interrupts
6042 +8.0% 6525 ± 6% interrupts.CPU60.CAL:Function_call_interrupts
6058 +7.5% 6515 ± 6% interrupts.CPU61.CAL:Function_call_interrupts
5839 ± 3% +11.5% 6513 ± 5% interrupts.CPU62.CAL:Function_call_interrupts
349.00 ± 69% -87.1% 45.00 ± 75% interrupts.CPU62.RES:Rescheduling_interrupts
4698 ± 28% +38.9% 6526 ± 6% interrupts.CPU63.CAL:Function_call_interrupts
977.75 ±112% -82.6% 170.50 ±158% interrupts.CPU63.RES:Rescheduling_interrupts
6050 +7.7% 6514 ± 6% interrupts.CPU64.CAL:Function_call_interrupts
5503 ± 17% +18.6% 6529 ± 6% interrupts.CPU65.CAL:Function_call_interrupts
611.75 ± 92% -91.5% 52.25 ±129% interrupts.CPU65.RES:Rescheduling_interrupts
6048 +7.8% 6522 ± 6% interrupts.CPU68.CAL:Function_call_interrupts
5958 ± 3% +9.3% 6512 ± 6% interrupts.CPU69.CAL:Function_call_interrupts
6061 +7.4% 6512 ± 5% interrupts.CPU7.CAL:Function_call_interrupts
6060 +7.5% 6517 ± 6% interrupts.CPU70.CAL:Function_call_interrupts
6052 +7.6% 6511 ± 6% interrupts.CPU71.CAL:Function_call_interrupts
6006 +8.4% 6510 ± 5% interrupts.CPU72.CAL:Function_call_interrupts
6066 +7.4% 6516 ± 6% interrupts.CPU75.CAL:Function_call_interrupts
5445 ± 24% -28.8% 3878 ± 34% interrupts.CPU75.NMI:Non-maskable_interrupts
5445 ± 24% -28.8% 3878 ± 34% interrupts.CPU75.PMI:Performance_monitoring_interrupts
6065 +7.4% 6514 ± 6% interrupts.CPU76.CAL:Function_call_interrupts
6066 +7.5% 6520 ± 6% interrupts.CPU77.CAL:Function_call_interrupts
5458 ± 24% -29.0% 3874 ± 34% interrupts.CPU77.NMI:Non-maskable_interrupts
5458 ± 24% -29.0% 3874 ± 34% interrupts.CPU77.PMI:Performance_monitoring_interrupts
5471 ± 24% -28.7% 3898 ± 34% interrupts.CPU78.NMI:Non-maskable_interrupts
5471 ± 24% -28.7% 3898 ± 34% interrupts.CPU78.PMI:Performance_monitoring_interrupts
6065 +7.5% 6521 ± 6% interrupts.CPU79.CAL:Function_call_interrupts
6066 +7.5% 6519 ± 6% interrupts.CPU80.CAL:Function_call_interrupts
6062 +7.6% 6521 ± 6% interrupts.CPU81.CAL:Function_call_interrupts
5459 ± 24% -28.7% 3894 ± 34% interrupts.CPU81.NMI:Non-maskable_interrupts
5459 ± 24% -28.7% 3894 ± 34% interrupts.CPU81.PMI:Performance_monitoring_interrupts
6070 +7.2% 6506 ± 6% interrupts.CPU82.CAL:Function_call_interrupts
6065 +7.7% 6529 ± 6% interrupts.CPU83.CAL:Function_call_interrupts
6070 +7.4% 6520 ± 6% interrupts.CPU84.CAL:Function_call_interrupts
6076 +7.2% 6514 ± 6% interrupts.CPU85.CAL:Function_call_interrupts
6068 +7.3% 6512 ± 6% interrupts.CPU86.CAL:Function_call_interrupts
6063 +7.4% 6511 ± 6% interrupts.CPU87.CAL:Function_call_interrupts
6074 +7.3% 6515 ± 6% interrupts.CPU88.CAL:Function_call_interrupts
6061 +7.4% 6511 ± 6% interrupts.CPU89.CAL:Function_call_interrupts
6074 +7.0% 6498 ± 5% interrupts.CPU9.CAL:Function_call_interrupts
5979 ± 2% +8.9% 6513 ± 6% interrupts.CPU90.CAL:Function_call_interrupts
5415 ± 21% +21.1% 6556 ± 5% interrupts.CPU91.CAL:Function_call_interrupts
6093 +6.8% 6509 ± 6% interrupts.CPU93.CAL:Function_call_interrupts
6001 ± 2% +8.5% 6514 ± 6% interrupts.CPU94.CAL:Function_call_interrupts
5484 ± 16% +18.8% 6516 ± 6% interrupts.CPU96.CAL:Function_call_interrupts
6068 +7.1% 6501 ± 6% interrupts.CPU97.CAL:Function_call_interrupts
6068 +7.1% 6501 ± 5% interrupts.CPU98.CAL:Function_call_interrupts
352.00 ± 61% -91.5% 29.75 ± 36% interrupts.CPU98.RES:Rescheduling_interrupts
6064 +7.1% 6497 ± 6% interrupts.CPU99.CAL:Function_call_interrupts
will-it-scale.per_process_ops
3e+06 +-+---------------------------------------------------------------+
|++.++++++.+++++.++++++.++++++.+++++.++ +++.+ +++.++ +++.++++++.++|
2.5e+06 OOO OOOOOO OOOOO OOOOOO OOOOOO OOO + : : : : |
| : : : : |
| : : : : |
2e+06 +-+ : : : : |
| : : : : |
1.5e+06 +-+ : : : : |
| : : |
1e+06 +-+ : : |
| : : |
| : : |
500000 +-+ : : |
| : : |
0 +-+---------------------------------------------------------------+
[*] 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-5.0.0-00012-g22dd8365" of type "text/plain" (187612 bytes)
View attachment "job-script" of type "text/plain" (7666 bytes)
View attachment "job.yaml" of type "text/plain" (5312 bytes)
View attachment "reproduce" of type "text/plain" (313 bytes)
Powered by blists - more mailing lists