| 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: <20160628014647.GE19084@yexl-desktop>
Date: Tue, 28 Jun 2016 09:46:47 +0800
From: kernel test robot <xiaolong.ye@...el.com>
To: "Kirill A. Shutemov" <kirill.shutemov@...ux.intel.com>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
"Huang, Ying" <ying.huang@...el.com>,
Rik van Riel <riel@...hat.com>, Mel Gorman <mgorman@...e.de>,
Michal Hocko <mhocko@...e.com>,
Minchan Kim <minchan@...nel.org>,
Vinayak Menon <vinmenon@...eaurora.org>,
Dave Hansen <dave.hansen@...ux.intel.com>,
Andrew Morton <akpm@...ux-foundation.org>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: [lkp] [mm] 315d09bf30: vm-scalability.throughput 64.6% improvement
FYI, we noticed a 64.6% improvement of vm-scalability.throughput due to commit:
commit 315d09bf30c2b436a1fdac86d31c24380cd56c4f ("Revert "mm: make faultaround produce old ptes"")
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
in testcase: vm-scalability
on test machine: 56 threads Grantley Haswell-EP with 64G memory
with following parameters: cpufreq_governor=performance/runtime=300s/size=2T/test=shm-pread-seq-mt
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.
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/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/rootfs/runtime/size/tbox_group/test/testcase:
gcc-4.9/performance/x86_64-rhel/debian-x86_64-2015-02-07.cgz/300s/2T/lkp-hsw01/shm-pread-seq-mt/vm-scalability
commit:
1f08fe2665 ("mailmap: add Boris Brezillon's email")
315d09bf30 ("Revert "mm: make faultaround produce old ptes"")
1f08fe266560fc2d 315d09bf30c2b436a1fdac86d3
---------------- --------------------------
%stddev %change %stddev
\ | \
27531872 ± 1% +64.6% 45326104 ± 0% vm-scalability.throughput
191.47 ± 1% -29.9% 134.16 ± 0% vm-scalability.time.elapsed_time
191.47 ± 1% -29.9% 134.16 ± 0% vm-scalability.time.elapsed_time.max
1.128e+08 ± 8% +68.0% 1.896e+08 ± 0% vm-scalability.time.minor_page_faults
4222 ± 0% -55.7% 1872 ± 0% vm-scalability.time.percent_of_cpu_this_job_got
961.95 ± 8% +19.8% 1152 ± 0% vm-scalability.time.system_time
7125 ± 1% -80.9% 1360 ± 0% vm-scalability.time.user_time
2736269 ± 25% +515.5% 16841224 ± 0% vm-scalability.time.voluntary_context_switches
98568 ± 21% -75.1% 24512 ± 13% latency_stats.hits.pipe_wait.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
4730 ± 30% +330.8% 20381 ± 18% latency_stats.sum.wait_on_page_bit.__migration_entry_wait.migration_entry_wait.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
5191 ± 2% -11.0% 4619 ± 1% slabinfo.cred_jar.active_objs
5191 ± 2% -11.0% 4619 ± 1% slabinfo.cred_jar.num_objs
7431191 ± 43% +60.2% 11904412 ± 4% numa-numastat.node1.local_node
7431191 ± 43% +60.2% 11904412 ± 4% numa-numastat.node1.numa_hit
280000 ± 90% -97.7% 6324 ± 47% numa-numastat.node1.numa_miss
17912 ± 86% -55.0% 8061 ± 2% softirqs.NET_RX
530133 ± 18% -75.4% 130251 ± 3% softirqs.RCU
453163 ± 26% +62.7% 737395 ± 1% softirqs.SCHED
4242892 ± 3% -66.3% 1430811 ± 2% softirqs.TIMER
0.00 ± 0% +Inf% 16.50 ± 3% vmstat.procs.b
43.00 ± 0% -16.9% 35.75 ± 1% vmstat.procs.r
29992 ± 22% +660.8% 228180 ± 0% vmstat.system.cs
70572 ± 0% -9.9% 63600 ± 0% vmstat.system.in
6259519 ± 1% +50.9% 9445759 ± 0% meminfo.Active
6243829 ± 1% +50.8% 9418336 ± 0% meminfo.Active(anon)
15689 ± 0% +74.8% 27422 ± 0% meminfo.Active(file)
13808412 ± 0% -12.8% 12044218 ± 0% meminfo.Inactive
13335827 ± 0% -13.1% 11583348 ± 0% meminfo.Inactive(anon)
45367988 ± 24% +432.4% 2.415e+08 ± 0% cpuidle.C1-HSW.time
2582764 ± 26% +426.1% 13588841 ± 0% cpuidle.C1-HSW.usage
2201153 ± 6% +1200.6% 28628523 ± 2% cpuidle.C1E-HSW.time
12004 ± 4% +1174.9% 153044 ± 3% cpuidle.C1E-HSW.usage
8211145 ± 13% +30.1% 10684935 ± 2% cpuidle.C3-HSW.time
18031 ± 7% +59.5% 28755 ± 2% cpuidle.C3-HSW.usage
2.493e+09 ± 1% +80.2% 4.492e+09 ± 0% cpuidle.C6-HSW.time
2637843 ± 1% +81.4% 4785113 ± 0% cpuidle.C6-HSW.usage
7111668 ± 29% +724.6% 58641885 ± 7% cpuidle.POLL.time
53935 ± 27% +444.0% 293412 ± 0% cpuidle.POLL.usage
75.85 ± 0% -53.1% 35.60 ± 0% turbostat.%Busy
2341 ± 0% -53.5% 1087 ± 0% turbostat.Avg_MHz
8.31 ± 4% +134.7% 19.49 ± 0% turbostat.CPU%c1
0.02 ± 24% +671.4% 0.14 ± 3% turbostat.CPU%c3
15.83 ± 1% +182.8% 44.77 ± 0% turbostat.CPU%c6
59.25 ± 3% -8.4% 54.25 ± 2% turbostat.CoreTmp
7.91 ± 3% +188.9% 22.86 ± 4% turbostat.Pkg%pc2
0.12 ± 61% +102.0% 0.25 ± 3% turbostat.Pkg%pc6
64.75 ± 2% -8.9% 59.00 ± 1% turbostat.PkgTmp
202.30 ± 0% -16.7% 168.50 ± 0% turbostat.PkgWatt
33.96 ± 1% -18.1% 27.82 ± 0% turbostat.RAMWatt
1967 ± 2% +76.6% 3475 ± 0% numa-vmstat.node0.nr_active_file
2158153 ± 22% -39.4% 1308350 ± 5% numa-vmstat.node0.nr_inactive_anon
613817 ± 47% +111.0% 1294953 ± 4% numa-vmstat.node1.nr_active_anon
1954 ± 2% +72.9% 3379 ± 0% numa-vmstat.node1.nr_active_file
2159 ± 11% +31.8% 2845 ± 19% numa-vmstat.node1.nr_anon_pages
1857510 ± 42% +58.3% 2941100 ± 4% numa-vmstat.node1.nr_file_pages
6082460 ± 13% -18.8% 4940032 ± 2% numa-vmstat.node1.nr_free_pages
1748193 ± 43% +59.7% 2791232 ± 4% numa-vmstat.node1.nr_mapped
256705 ± 1% +20.4% 309136 ± 3% numa-vmstat.node1.nr_page_table_pages
1796503 ± 43% +60.3% 2880117 ± 4% numa-vmstat.node1.nr_shmem
10775 ± 15% +23.9% 13351 ± 4% numa-vmstat.node1.nr_slab_reclaimable
167634 ± 99% -86.9% 21950 ± 13% numa-vmstat.node1.numa_miss
1560884 ± 1% +50.8% 2354512 ± 0% proc-vmstat.nr_active_anon
3922 ± 0% +74.8% 6855 ± 0% proc-vmstat.nr_active_file
3333754 ± 0% -13.1% 2895782 ± 0% proc-vmstat.nr_inactive_anon
370024 ± 97% -97.1% 10844 ± 0% proc-vmstat.numa_foreign
40219 ± 2% -43.1% 22897 ± 0% proc-vmstat.numa_hint_faults
36222 ± 4% -66.0% 12298 ± 1% proc-vmstat.numa_hint_faults_local
18994704 ± 1% +13.9% 21637801 ± 0% proc-vmstat.numa_hit
18994703 ± 1% +13.9% 21637800 ± 0% proc-vmstat.numa_local
370024 ± 97% -97.1% 10844 ± 0% proc-vmstat.numa_miss
508.00 ± 69% +2896.3% 15221 ± 50% proc-vmstat.numa_pages_migrated
18756368 ± 1% +12.6% 21122610 ± 0% proc-vmstat.pgalloc_normal
1.133e+08 ± 8% +67.6% 1.899e+08 ± 0% proc-vmstat.pgfault
508.00 ± 69% +2896.3% 15221 ± 50% proc-vmstat.pgmigrate_success
7872 ± 2% +76.6% 13904 ± 0% numa-meminfo.node0.Active(file)
8851043 ± 22% -38.3% 5464236 ± 5% numa-meminfo.node0.Inactive
8614666 ± 22% -39.2% 5233778 ± 5% numa-meminfo.node0.Inactive(anon)
2449740 ± 47% +112.0% 5192532 ± 4% numa-meminfo.node1.Active
2441920 ± 47% +112.1% 5179011 ± 4% numa-meminfo.node1.Active(anon)
7819 ± 2% +72.9% 13520 ± 0% numa-meminfo.node1.Active(file)
155.25 ±173% +1290.3% 2158 ± 70% numa-meminfo.node1.AnonHugePages
8631 ± 11% +31.8% 11378 ± 19% numa-meminfo.node1.AnonPages
7411490 ± 42% +58.7% 11763830 ± 4% numa-meminfo.node1.FilePages
6975983 ± 43% +60.0% 11164872 ± 4% numa-meminfo.node1.Mapped
24347447 ± 13% -18.8% 19760755 ± 2% numa-meminfo.node1.MemFree
8652684 ± 36% +53.0% 13239372 ± 3% numa-meminfo.node1.MemUsed
1027932 ± 1% +20.3% 1236525 ± 3% numa-meminfo.node1.PageTables
43069 ± 15% +24.0% 53409 ± 4% numa-meminfo.node1.SReclaimable
7167463 ± 43% +60.7% 11519899 ± 4% numa-meminfo.node1.Shmem
7.053e+10 ± 1% +21.6% 8.58e+10 ± 1% perf-stat.L1-dcache-load-misses
2.156e+12 ± 1% +8.2% 2.332e+12 ± 0% perf-stat.L1-dcache-loads
6.164e+11 ± 0% +12.8% 6.95e+11 ± 0% perf-stat.L1-dcache-stores
4.013e+09 ± 11% +122.1% 8.914e+09 ± 1% perf-stat.L1-icache-load-misses
3.219e+09 ± 8% -84.0% 5.152e+08 ± 3% perf-stat.LLC-load-misses
9.899e+08 ± 0% -85.8% 1.407e+08 ± 1% perf-stat.LLC-store-misses
2.061e+09 ± 8% +77.8% 3.664e+09 ± 0% perf-stat.LLC-stores
3.034e+12 ± 0% +5.7% 3.206e+12 ± 1% perf-stat.branch-instructions
1.105e+09 ± 10% +208.8% 3.414e+09 ± 0% perf-stat.branch-load-misses
3.045e+12 ± 0% +6.3% 3.238e+12 ± 1% perf-stat.branch-loads
1.132e+09 ± 14% +207.0% 3.476e+09 ± 1% perf-stat.branch-misses
8.041e+11 ± 2% -67.5% 2.615e+11 ± 1% perf-stat.bus-cycles
4.172e+09 ± 5% -83.2% 7.014e+08 ± 3% perf-stat.cache-misses
1.993e+10 ± 4% +21.4% 2.418e+10 ± 1% perf-stat.cache-references
5824116 ± 23% +434.7% 31142455 ± 0% perf-stat.context-switches
2.51e+13 ± 1% -67.9% 8.061e+12 ± 1% perf-stat.cpu-cycles
92604 ± 12% +2823.3% 2707058 ± 0% perf-stat.cpu-migrations
9.796e+09 ± 1% -72.9% 2.653e+09 ± 8% perf-stat.dTLB-load-misses
2.15e+12 ± 0% +8.9% 2.342e+12 ± 0% perf-stat.dTLB-loads
45788288 ± 21% +192.7% 1.34e+08 ± 18% perf-stat.dTLB-store-misses
6.166e+11 ± 0% +11.9% 6.896e+11 ± 1% perf-stat.dTLB-stores
51173097 ± 9% +98.2% 1.014e+08 ± 1% perf-stat.iTLB-load-misses
72582096 ± 19% +317.2% 3.028e+08 ± 1% perf-stat.iTLB-loads
9.447e+12 ± 0% +10.4% 1.043e+13 ± 1% perf-stat.instructions
1.133e+08 ± 8% +67.6% 1.899e+08 ± 0% perf-stat.minor-faults
2.655e+09 ± 8% -85.1% 3.957e+08 ± 3% perf-stat.node-load-misses
4.726e+08 ± 11% -72.8% 1.287e+08 ± 13% perf-stat.node-loads
7.493e+08 ± 2% -92.7% 54657056 ± 4% perf-stat.node-store-misses
2.531e+08 ± 1% -65.4% 87598099 ± 1% perf-stat.node-stores
1.133e+08 ± 8% +67.6% 1.899e+08 ± 0% perf-stat.page-faults
2.09e+13 ± 2% -67.4% 6.805e+12 ± 0% perf-stat.ref-cycles
0.46 ± 60% +801.6% 4.17 ± 1% perf-profile.cycles-pp.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate
5.78 ± 4% +64.0% 9.47 ± 0% perf-profile.cycles-pp.__do_fault.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault
82.11 ± 1% -10.9% 73.20 ± 0% perf-profile.cycles-pp.__do_page_fault.do_page_fault.page_fault
1.93 ± 8% -100.0% 0.00 ± -1% perf-profile.cycles-pp.__hrtimer_run_queues.hrtimer_interrupt.local_apic_timer_interrupt.smp_apic_timer_interrupt.apic_timer_interrupt
0.00 ± -1% +Inf% 2.18 ± 2% perf-profile.cycles-pp.__kernel_text_address.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency
1.28 ± 31% +338.1% 5.61 ± 1% perf-profile.cycles-pp.__lock_page.find_lock_entry.shmem_getpage_gfp.shmem_fault.__do_fault
0.15 ±173% +873.3% 1.46 ± 2% perf-profile.cycles-pp.__schedule.schedule.schedule_timeout.io_schedule_timeout.bit_wait_io
1.27 ± 31% +334.4% 5.50 ± 1% perf-profile.cycles-pp.__wait_on_bit_lock.__lock_page.find_lock_entry.shmem_getpage_gfp.shmem_fault
4.78 ± 38% +447.3% 26.17 ± 0% perf-profile.cycles-pp.__wake_up.__wake_up_bit.unlock_page.filemap_map_pages.handle_pte_fault
2.54 ± 28% +375.6% 12.06 ± 1% perf-profile.cycles-pp.__wake_up.__wake_up_bit.unlock_page.handle_pte_fault.handle_mm_fault
7.35 ± 22% +273.5% 27.46 ± 0% perf-profile.cycles-pp.__wake_up_bit.unlock_page.filemap_map_pages.handle_pte_fault.handle_mm_fault
2.60 ± 27% +369.5% 12.20 ± 1% perf-profile.cycles-pp.__wake_up_bit.unlock_page.handle_pte_fault.handle_mm_fault.__do_page_fault
0.35 ±100% +661.4% 2.67 ± 1% perf-profile.cycles-pp.__wake_up_common.__wake_up.__wake_up_bit.unlock_page.filemap_map_pages
1.05 ± 20% +287.7% 4.09 ± 0% perf-profile.cycles-pp.__wake_up_common.__wake_up.__wake_up_bit.unlock_page.handle_pte_fault
2.77 ± 30% -100.0% 0.00 ± -1% perf-profile.cycles-pp._raw_spin_lock.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
4.19 ± 40% +454.5% 23.23 ± 0% perf-profile.cycles-pp._raw_spin_lock_irqsave.__wake_up.__wake_up_bit.unlock_page.filemap_map_pages
1.45 ± 34% +434.1% 7.76 ± 2% perf-profile.cycles-pp._raw_spin_lock_irqsave.__wake_up.__wake_up_bit.unlock_page.handle_pte_fault
0.19 ±173% +1307.9% 2.67 ± 2% perf-profile.cycles-pp._raw_spin_lock_irqsave.prepare_to_wait_exclusive.__wait_on_bit_lock.__lock_page.find_lock_entry
0.55 ± 60% +786.4% 4.88 ± 0% perf-profile.cycles-pp.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function
2.92 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.apic_timer_interrupt
1.33 ± 39% +383.1% 6.44 ± 0% perf-profile.cycles-pp.autoremove_wake_function.wake_bit_function.__wake_up_common.__wake_up.__wake_up_bit
0.47 ± 60% +275.1% 1.77 ± 2% perf-profile.cycles-pp.bit_wait_io.__wait_on_bit_lock.__lock_page.find_lock_entry.shmem_getpage_gfp
9.62 ± 12% +113.9% 20.59 ± 2% perf-profile.cycles-pp.call_cpuidle.cpu_startup_entry.start_secondary
1.25 ± 12% -45.7% 0.68 ± 2% perf-profile.cycles-pp.clear_page_c_e.shmem_fault.__do_fault.handle_pte_fault.handle_mm_fault
10.55 ± 10% +118.2% 23.03 ± 1% perf-profile.cycles-pp.cpu_startup_entry.start_secondary
9.61 ± 12% +114.2% 20.57 ± 2% perf-profile.cycles-pp.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
9.59 ± 11% +109.5% 20.09 ± 2% perf-profile.cycles-pp.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry.start_secondary
1.32 ± 39% +384.8% 6.39 ± 0% perf-profile.cycles-pp.default_wake_function.autoremove_wake_function.wake_bit_function.__wake_up_common.__wake_up
82.31 ± 1% -10.8% 73.41 ± 0% perf-profile.cycles-pp.do_page_fault.page_fault
4.28 ± 8% -49.7% 2.15 ± 1% perf-profile.cycles-pp.do_set_pte.filemap_map_pages.handle_pte_fault.handle_mm_fault.__do_page_fault
3.66 ± 8% -59.7% 1.48 ± 4% perf-profile.cycles-pp.down_read_trylock.__do_page_fault.do_page_fault.page_fault
0.30 ±101% +1184.7% 3.79 ± 1% perf-profile.cycles-pp.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair
0.54 ± 60% +771.2% 4.68 ± 1% perf-profile.cycles-pp.enqueue_entity.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up
0.54 ± 60% +782.0% 4.79 ± 0% perf-profile.cycles-pp.enqueue_task_fair.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function
33.12 ± 6% +32.2% 43.77 ± 0% perf-profile.cycles-pp.filemap_map_pages.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault
1.54 ± 29% +341.1% 6.79 ± 1% perf-profile.cycles-pp.find_lock_entry.shmem_getpage_gfp.shmem_fault.__do_fault.handle_pte_fault
43.86 ± 5% +53.2% 67.21 ± 0% perf-profile.cycles-pp.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
2.25 ± 10% -100.0% 0.00 ± -1% perf-profile.cycles-pp.hrtimer_interrupt.local_apic_timer_interrupt.smp_apic_timer_interrupt.apic_timer_interrupt
9.15 ± 11% +76.0% 16.10 ± 2% perf-profile.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry
0.47 ± 60% +275.0% 1.76 ± 2% perf-profile.cycles-pp.io_schedule_timeout.bit_wait_io.__wait_on_bit_lock.__lock_page.find_lock_entry
2.41 ± 10% -100.0% 0.00 ± -1% perf-profile.cycles-pp.local_apic_timer_interrupt.smp_apic_timer_interrupt.apic_timer_interrupt
5.57 ± 38% +450.7% 30.67 ± 0% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__wake_up.__wake_up_bit.unlock_page
0.17 ±173% +1345.7% 2.53 ± 3% perf-profile.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.prepare_to_wait_exclusive.__wait_on_bit_lock.__lock_page
2.89 ± 7% -56.4% 1.26 ± 2% perf-profile.cycles-pp.page_add_file_rmap.do_set_pte.filemap_map_pages.handle_pte_fault.handle_mm_fault
82.61 ± 1% -11.0% 73.50 ± 0% perf-profile.cycles-pp.page_fault
0.17 ±173% +2070.0% 3.80 ± 2% perf-profile.cycles-pp.poll_idle.cpuidle_enter_state.cpuidle_enter.call_cpuidle.cpu_startup_entry
0.48 ± 64% +499.5% 2.89 ± 3% perf-profile.cycles-pp.prepare_to_wait_exclusive.__wait_on_bit_lock.__lock_page.find_lock_entry.shmem_getpage_gfp
0.14 ±173% +2282.8% 3.46 ± 1% perf-profile.cycles-pp.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity
5.64 ± 27% -42.5% 3.25 ± 2% perf-profile.cycles-pp.radix_tree_next_chunk.filemap_map_pages.handle_pte_fault.handle_mm_fault.__do_page_fault
0.30 ±101% +1185.6% 3.79 ± 1% perf-profile.cycles-pp.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.activate_task
0.29 ±101% +425.0% 1.52 ± 1% perf-profile.cycles-pp.schedule.schedule_timeout.io_schedule_timeout.bit_wait_io.__wait_on_bit_lock
0.30 ±101% +421.2% 1.54 ± 2% perf-profile.cycles-pp.schedule_timeout.io_schedule_timeout.bit_wait_io.__wait_on_bit_lock.__lock_page
1.02 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.scheduler_tick.update_process_times.tick_sched_handle.isra.17.tick_sched_timer.__hrtimer_run_queues
5.74 ± 4% +64.2% 9.43 ± 0% perf-profile.cycles-pp.shmem_fault.__do_fault.handle_pte_fault.handle_mm_fault.__do_page_fault
4.29 ± 8% +100.6% 8.61 ± 1% perf-profile.cycles-pp.shmem_getpage_gfp.shmem_fault.__do_fault.handle_pte_fault.handle_mm_fault
2.86 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.smp_apic_timer_interrupt.apic_timer_interrupt
10.58 ± 10% +118.7% 23.12 ± 1% perf-profile.cycles-pp.start_secondary
1.37 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.tick_sched_handle.isra.17.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.local_apic_timer_interrupt
1.43 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt.local_apic_timer_interrupt.smp_apic_timer_interrupt
1.29 ± 39% +391.1% 6.32 ± 0% perf-profile.cycles-pp.try_to_wake_up.default_wake_function.autoremove_wake_function.wake_bit_function.__wake_up_common
0.71 ± 71% +616.6% 5.07 ± 0% perf-profile.cycles-pp.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function.wake_bit_function
10.42 ± 13% +180.7% 29.26 ± 0% perf-profile.cycles-pp.unlock_page.filemap_map_pages.handle_pte_fault.handle_mm_fault.__do_page_fault
2.90 ± 24% +333.1% 12.57 ± 1% perf-profile.cycles-pp.unlock_page.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault
1.58 ± 6% -41.2% 0.93 ± 1% perf-profile.cycles-pp.up_read.__do_page_fault.do_page_fault.page_fault
1.35 ± 9% -100.0% 0.00 ± -1% perf-profile.cycles-pp.update_process_times.tick_sched_handle.isra.17.tick_sched_timer.__hrtimer_run_queues.hrtimer_interrupt
1.04 ± 31% -100.0% 0.00 ± -1% perf-profile.cycles-pp.vmacache_find.__do_page_fault.do_page_fault.page_fault
1.33 ± 39% +382.8% 6.44 ± 0% perf-profile.cycles-pp.wake_bit_function.__wake_up_common.__wake_up.__wake_up_bit.unlock_page
73746 ± 1% -69.7% 22350 ± 0% sched_debug.cfs_rq:/.exec_clock.avg
83538 ± 1% -57.8% 35283 ± 6% sched_debug.cfs_rq:/.exec_clock.max
71361 ± 1% -74.1% 18505 ± 3% sched_debug.cfs_rq:/.exec_clock.min
1820 ± 6% +88.8% 3436 ± 18% sched_debug.cfs_rq:/.exec_clock.stddev
556906 ± 1% -78.4% 120294 ± 14% sched_debug.cfs_rq:/.load.avg
250112 ± 6% +22.5% 306369 ± 5% sched_debug.cfs_rq:/.load.stddev
459.38 ± 4% -89.0% 50.62 ± 17% sched_debug.cfs_rq:/.load_avg.avg
168.50 ± 20% -99.5% 0.92 ± 78% sched_debug.cfs_rq:/.load_avg.min
78306 ± 1% -69.0% 24277 ± 0% sched_debug.cfs_rq:/.min_vruntime.avg
92855 ± 1% -55.6% 41189 ± 2% sched_debug.cfs_rq:/.min_vruntime.max
75262 ± 1% -73.8% 19753 ± 3% sched_debug.cfs_rq:/.min_vruntime.min
2945 ± 5% +48.6% 4377 ± 14% sched_debug.cfs_rq:/.min_vruntime.stddev
0.56 ± 1% -78.9% 0.12 ± 14% sched_debug.cfs_rq:/.nr_running.avg
0.25 ± 6% +20.5% 0.30 ± 6% sched_debug.cfs_rq:/.nr_running.stddev
428.46 ± 2% -95.6% 18.75 ± 4% sched_debug.cfs_rq:/.runnable_load_avg.avg
143.31 ± 3% -29.4% 101.24 ± 1% sched_debug.cfs_rq:/.runnable_load_avg.stddev
4249 ± 26% +113.1% 9054 ± 16% sched_debug.cfs_rq:/.spread0.max
2946 ± 5% +48.6% 4377 ± 14% sched_debug.cfs_rq:/.spread0.stddev
519.54 ± 1% -91.1% 46.27 ± 12% sched_debug.cfs_rq:/.util_avg.avg
205.31 ± 24% -99.1% 1.75 ± 70% sched_debug.cfs_rq:/.util_avg.min
108.98 ± 11% +18.5% 129.19 ± 4% sched_debug.cfs_rq:/.util_avg.stddev
671352 ± 6% +27.6% 856754 ± 1% sched_debug.cpu.avg_idle.avg
318488 ± 5% -50.7% 157168 ± 12% sched_debug.cpu.avg_idle.stddev
129100 ± 3% -21.2% 101762 ± 0% sched_debug.cpu.clock.avg
129118 ± 3% -21.2% 101781 ± 0% sched_debug.cpu.clock.max
129012 ± 3% -21.3% 101477 ± 0% sched_debug.cpu.clock.min
129100 ± 3% -21.2% 101762 ± 0% sched_debug.cpu.clock_task.avg
129118 ± 3% -21.2% 101781 ± 0% sched_debug.cpu.clock_task.max
129012 ± 3% -21.3% 101477 ± 0% sched_debug.cpu.clock_task.min
395.11 ± 4% -95.4% 18.19 ± 5% sched_debug.cpu.cpu_load[0].avg
194.11 ± 2% -48.0% 101.01 ± 1% sched_debug.cpu.cpu_load[0].stddev
438.43 ± 2% -94.8% 22.93 ± 10% sched_debug.cpu.cpu_load[1].avg
106.31 ± 32% -100.0% 0.00 ± -1% sched_debug.cpu.cpu_load[1].min
123.48 ± 5% -14.2% 105.97 ± 7% sched_debug.cpu.cpu_load[1].stddev
436.49 ± 2% -95.4% 20.15 ± 7% sched_debug.cpu.cpu_load[2].avg
89.88 ± 31% -100.0% 0.00 ± -1% sched_debug.cpu.cpu_load[2].min
120.46 ± 5% -17.4% 99.52 ± 3% sched_debug.cpu.cpu_load[2].stddev
434.60 ± 2% -96.0% 17.60 ± 6% sched_debug.cpu.cpu_load[3].avg
87.44 ± 19% -100.0% 0.00 ± -1% sched_debug.cpu.cpu_load[3].min
118.05 ± 5% -18.9% 95.69 ± 1% sched_debug.cpu.cpu_load[3].stddev
433.47 ± 2% -96.4% 15.61 ± 5% sched_debug.cpu.cpu_load[4].avg
96.94 ± 21% -100.0% 0.00 ± -1% sched_debug.cpu.cpu_load[4].min
114.65 ± 6% -18.9% 93.01 ± 1% sched_debug.cpu.cpu_load[4].stddev
1524 ± 1% -87.4% 192.58 ± 7% sched_debug.cpu.curr->pid.avg
3752 ± 3% -17.7% 3087 ± 0% sched_debug.cpu.curr->pid.max
554671 ± 1% -79.4% 114266 ± 16% sched_debug.cpu.load.avg
256059 ± 5% +18.1% 302374 ± 6% sched_debug.cpu.load.stddev
78790 ± 2% -61.2% 30537 ± 2% sched_debug.cpu.nr_load_updates.avg
90782 ± 2% -50.6% 44880 ± 2% sched_debug.cpu.nr_load_updates.max
74885 ± 2% -64.5% 26594 ± 3% sched_debug.cpu.nr_load_updates.min
2752 ± 5% +21.0% 3331 ± 7% sched_debug.cpu.nr_load_updates.stddev
0.58 ± 0% -79.6% 0.12 ± 14% sched_debug.cpu.nr_running.avg
1.75 ± 10% -33.3% 1.17 ± 14% sched_debug.cpu.nr_running.max
53270 ± 24% +424.6% 279439 ± 0% sched_debug.cpu.nr_switches.avg
103276 ± 22% +226.9% 337585 ± 3% sched_debug.cpu.nr_switches.max
15721 ± 70% +1299.9% 220080 ± 3% sched_debug.cpu.nr_switches.min
48.94 ± 13% +559.2% 322.58 ± 9% sched_debug.cpu.nr_uninterruptible.max
-67.00 ±-11% +378.4% -320.50 ± -4% sched_debug.cpu.nr_uninterruptible.min
25.35 ± 11% +439.0% 136.64 ± 5% sched_debug.cpu.nr_uninterruptible.stddev
53707 ± 24% +421.5% 280087 ± 0% sched_debug.cpu.sched_count.avg
123598 ± 22% +196.8% 366809 ± 4% sched_debug.cpu.sched_count.max
15428 ± 70% +1326.2% 220035 ± 3% sched_debug.cpu.sched_count.min
24829 ± 26% +407.6% 126032 ± 0% sched_debug.cpu.sched_goidle.avg
46268 ± 24% +228.0% 151781 ± 3% sched_debug.cpu.sched_goidle.max
7174 ± 74% +1278.3% 98887 ± 3% sched_debug.cpu.sched_goidle.min
27051 ± 24% +472.6% 154899 ± 0% sched_debug.cpu.ttwu_count.avg
52173 ± 22% +254.8% 185098 ± 2% sched_debug.cpu.ttwu_count.max
7978 ± 77% +1449.6% 123640 ± 3% sched_debug.cpu.ttwu_count.min
1414 ± 12% +833.5% 13208 ± 1% sched_debug.cpu.ttwu_local.avg
9625 ± 27% +76.1% 16948 ± 4% sched_debug.cpu.ttwu_local.max
402.12 ± 30% +2413.0% 10105 ± 2% sched_debug.cpu.ttwu_local.min
129106 ± 3% -21.2% 101776 ± 0% sched_debug.cpu_clk
126482 ± 2% -21.6% 99142 ± 0% sched_debug.ktime
0.00 ±173% +566.7% 0.01 ± 34% sched_debug.rt_rq:/.rt_nr_running.avg
0.06 ±173% +433.3% 0.33 ± 0% sched_debug.rt_rq:/.rt_nr_running.max
0.01 ±173% +486.8% 0.05 ± 15% sched_debug.rt_rq:/.rt_nr_running.stddev
0.03 ± 26% +58.7% 0.04 ± 10% sched_debug.rt_rq:/.rt_time.avg
0.79 ± 41% +103.5% 1.62 ± 13% sched_debug.rt_rq:/.rt_time.max
0.13 ± 30% +72.3% 0.23 ± 12% sched_debug.rt_rq:/.rt_time.stddev
129106 ± 3% -21.2% 101776 ± 0% sched_debug.sched_clk
vm-scalability.throughput
5e+07 ++----------------------------------------------------------------+
4.5e+07 OO O OO OO OO OO OO OOO OO |
| |
4e+07 ++ |
3.5e+07 ++ |
| |
3e+07 **.**.**.**. *.**.**.* *.**.**.**.* .**.**.***.**. *.**.**.**.**. *
2.5e+07 ++ * * * * *|
2e+07 ++ |
| |
1.5e+07 ++ |
1e+07 ++ |
| |
5e+06 ++ |
0 ++--O-------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
Thanks,
Xiaolong
View attachment "config-4.7.0-rc4-00102-g315d09b" of type "text/plain" (150949 bytes)
View attachment "job.yaml" of type "text/plain" (3471 bytes)
View attachment "reproduce" of type "text/plain" (6325 bytes)
Powered by blists - more mailing lists