From: Huang, Ying <ying.huang@intel.com>
To: lkp@lists.01.org
Subject: Re: [x86, paravirt] fd6f48529f: aim7.jobs-per-min -26.1% regression
Date: Thu, 08 Dec 2016 16:39:10 +0800 [thread overview]
Message-ID: <87bmwm6ci9.fsf@yhuang-dev.intel.com> (raw)
In-Reply-To: <20161207131740.GY3092@twins.programming.kicks-ass.net>
[-- Attachment #1: Type: text/plain, Size: 38262 bytes --]
Peter Zijlstra <peterz@infradead.org> writes:
> On Thu, Dec 01, 2016 at 02:07:29AM +0800, kernel test robot wrote:
>>
>> Greeting,
>>
>> FYI, we noticed a -26.1% regression of aim7.jobs-per-min due to commit:
>>
>>
>> commit: fd6f48529f4067c6bfbead88d3d7d117d51150f6 ("x86,paravirt: Optimize native pv_lock_ops.vcpu_is_preempted")
>> https://git.kernel.org/pub/scm/linux/kernel/git/peterz/queue.git locking/core
>>
>> in testcase: aim7
>> on test machine: qemu-system-x86_64 -enable-kvm -cpu host -smp 4 -m 5G
>> with following parameters:
>>
>> test: link_test
>> load: 2000
>>
>> test-description: AIM7 is a traditional UNIX system level benchmark
>> suite which is used to test and measure the performance of multiuser
>> system.
>> test-url: https://sourceforge.net/projects/aimbench/files/aim-suite7/
>>
>
> OK, so I'm having a very hard time reproducing this. How can I run this
> link_test stuff when I got aim7 compiled? This is all impenetrable gunk
> to me.
>
>
>
> Also, do you see the same problem with the upstream commit?
>
> 3cded4179481 ("x86/paravirt: Optimize native pv_lock_ops.vcpu_is_preempted()")
Here is the test result for the commit,
=========================================================================================
compiler/kconfig/load/rootfs/tbox_group/test/testcase:
gcc-6/x86_64-rhel-7.2/2000/debian-x86_64-2016-08-31.cgz/vm-lkp-a05/link_test/aim7
commit:
05ffc951392df57edecc2519327b169210c3df75
3cded41794818d788aa1dc028ede4a1c1222d937
05ffc951392df57e 3cded41794818d788aa1dc028e
---------------- --------------------------
%stddev %change %stddev
\ | \
9169 ± 2% -27.9% 6610 ± 1% aim7.jobs-per-min
1325 ± 2% +38.2% 1831 ± 1% aim7.time.elapsed_time
1325 ± 2% +38.2% 1831 ± 1% aim7.time.elapsed_time.max
1639 ± 2% +9.1% 1789 ± 2% aim7.time.system_time
2561 ± 2% +32.7% 3399 ± 2% aim7.time.user_time
528278 ± 2% +81.8% 960538 ± 0% aim7.time.voluntary_context_switches
813472 ± 1% +62.6% 1323099 ± 4% softirqs.RCU
34284 ± 1% +393.3% 169136 ± 9% softirqs.SCHED
1858312 ± 1% +30.8% 2430971 ± 1% softirqs.TIMER
65.00 ± 15% -47.7% 34.00 ± 20% vmstat.procs.r
3201 ± 0% -7.9% 2949 ± 0% vmstat.system.cs
5629 ± 1% -10.3% 5049 ± 1% vmstat.system.in
488672 ± 1% +82.2% 890241 ± 1% proc-vmstat.numa_hit
488610 ± 1% +82.2% 890164 ± 1% proc-vmstat.numa_local
531586 ± 1% +77.1% 941453 ± 1% proc-vmstat.pgalloc_normal
521563 ± 1% +83.6% 957670 ± 0% proc-vmstat.pgfault
524470 ± 1% +78.5% 936028 ± 1% proc-vmstat.pgfree
42797 ± 0% +10.2% 47167 ± 0% slabinfo.dentry.active_objs
3903 ± 0% +21.8% 4753 ± 0% slabinfo.kmalloc-96.active_objs
3903 ± 0% +22.0% 4761 ± 0% slabinfo.kmalloc-96.num_objs
2001 ± 1% +26.8% 2536 ± 0% slabinfo.shmem_inode_cache.active_objs
2291 ± 1% +18.1% 2706 ± 0% slabinfo.shmem_inode_cache.num_objs
1325 ± 2% +38.2% 1831 ± 1% time.elapsed_time
1325 ± 2% +38.2% 1831 ± 1% time.elapsed_time.max
316.33 ± 0% -10.6% 282.67 ± 1% time.percent_of_cpu_this_job_got
2561 ± 2% +32.7% 3399 ± 2% time.user_time
528278 ± 2% +81.8% 960538 ± 0% time.voluntary_context_switches
16345 ± 83% -92.9% 1153 ± 9% latency_stats.avg.call_rwsem_down_write_failed.unlink_file_vma.free_pgtables.exit_mmap.mmput.flush_old_exec.load_elf_binary.search_binary_handler.do_execveat_common.SyS_execve.do_syscall_64.return_from_SYSCALL_64
23013 ± 68% -93.0% 1600 ± 29% latency_stats.avg.call_rwsem_down_write_failed.vma_link.mmap_region.do_mmap.vm_mmap_pgoff.SyS_mmap_pgoff.SyS_mmap.entry_SYSCALL_64_fastpath
45454 ±111% -90.8% 4161 ±141% latency_stats.avg.nfs_wait_on_request.nfs_writepage_setup.nfs_updatepage.nfs_write_end.generic_perform_write.nfs_file_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
32652 ± 60% -99.0% 333.67 ± 16% latency_stats.avg.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
484699 ± 65% -92.3% 37212 ± 72% latency_stats.avg.wait_on_page_bit.__filemap_fdatawait_range.filemap_fdatawait_keep_errors.sync_inodes_sb.sync_inodes_one_sb.iterate_supers.sys_sync.entry_SYSCALL_64_fastpath
45.00 ± 9% +25319.3% 11438 ± 8% latency_stats.hits.pipe_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
5712 ± 82% +318.0% 23877 ± 66% latency_stats.max.call_rwsem_down_write_failed.__vma_adjust.__split_vma.split_vma.mprotect_fixup.do_mprotect_pkey.SyS_mprotect.entry_SYSCALL_64_fastpath
2760 ±141% +1049.6% 31733 ± 28% latency_stats.max.call_rwsem_down_write_failed.copy_process._do_fork.SyS_clone.do_syscall_64.return_from_SYSCALL_64
0.00 ± -1% +Inf% 10184 ± 64% latency_stats.max.call_rwsem_down_write_failed.unlink_file_vma.free_pgtables.unmap_region.do_munmap.mmap_region.do_mmap.vm_mmap_pgoff.SyS_mmap_pgoff.SyS_mmap.entry_SYSCALL_64_fastpath
39755 ± 68% -71.8% 11206 ± 27% latency_stats.max.call_rwsem_down_write_failed.vma_link.mmap_region.do_mmap.vm_mmap_pgoff.SyS_mmap_pgoff.SyS_mmap.entry_SYSCALL_64_fastpath
82135 ±124% -94.9% 4161 ±141% latency_stats.max.nfs_wait_on_request.nfs_writepage_setup.nfs_updatepage.nfs_write_end.generic_perform_write.nfs_file_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
592251 ±118% -83.3% 98662 ± 45% latency_stats.max.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
792307 ±103% -86.0% 110586 ± 22% latency_stats.max.pipe_wait.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
0.00 ± -1% +Inf% 12847 ± 70% latency_stats.max.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
484699 ± 65% -92.3% 37212 ± 72% latency_stats.max.wait_on_page_bit.__filemap_fdatawait_range.filemap_fdatawait_keep_errors.sync_inodes_sb.sync_inodes_one_sb.iterate_supers.sys_sync.entry_SYSCALL_64_fastpath
6136 ± 72% +760.8% 52822 ± 42% latency_stats.sum.call_rwsem_down_write_failed.__vma_adjust.__split_vma.split_vma.mprotect_fixup.do_mprotect_pkey.SyS_mprotect.entry_SYSCALL_64_fastpath
7297 ±141% +1151.9% 91355 ± 7% latency_stats.sum.call_rwsem_down_write_failed.copy_process._do_fork.SyS_clone.do_syscall_64.return_from_SYSCALL_64
0.00 ± -1% +Inf% 12156 ± 64% latency_stats.sum.call_rwsem_down_write_failed.unlink_file_vma.free_pgtables.unmap_region.do_munmap.mmap_region.do_mmap.vm_mmap_pgoff.SyS_mmap_pgoff.SyS_mmap.entry_SYSCALL_64_fastpath
122802 ±129% -96.6% 4161 ±141% latency_stats.sum.nfs_wait_on_request.nfs_writepage_setup.nfs_updatepage.nfs_write_end.generic_perform_write.nfs_file_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
0.00 ± -1% +Inf% 18300 ± 71% latency_stats.sum.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
292182 ± 32% +645.7% 2178683 ± 1% latency_stats.sum.pipe_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
484699 ± 65% -92.3% 37212 ± 72% latency_stats.sum.wait_on_page_bit.__filemap_fdatawait_range.filemap_fdatawait_keep_errors.sync_inodes_sb.sync_inodes_one_sb.iterate_supers.sys_sync.entry_SYSCALL_64_fastpath
694272 ± 2% +27.9% 887696 ± 2% sched_debug.cfs_rq:/.exec_clock.avg
697960 ± 2% +28.0% 893571 ± 2% sched_debug.cfs_rq:/.exec_clock.max
690173 ± 2% +28.1% 883901 ± 2% sched_debug.cfs_rq:/.exec_clock.min
53.01 ± 4% -17.6% 43.69 ± 3% sched_debug.cfs_rq:/.load_avg.stddev
3373615 ± 2% +20.5% 4065963 ± 2% sched_debug.cfs_rq:/.min_vruntime.avg
3393431 ± 2% +20.7% 4096926 ± 2% sched_debug.cfs_rq:/.min_vruntime.max
3348545 ± 1% +20.5% 4035538 ± 2% sched_debug.cfs_rq:/.min_vruntime.min
17027 ± 12% +41.4% 24071 ± 14% sched_debug.cfs_rq:/.min_vruntime.stddev
0.94 ± 5% -15.4% 0.80 ± 10% sched_debug.cfs_rq:/.nr_running.min
0.06 ± 20% +99.1% 0.12 ± 38% sched_debug.cfs_rq:/.nr_running.stddev
21.65 ± 32% +52.0% 32.92 ± 5% sched_debug.cfs_rq:/.nr_spread_over.avg
12.59 ± 7% +61.9% 20.38 ± 23% sched_debug.cfs_rq:/.nr_spread_over.min
165.33 ± 9% -9.1% 150.30 ± 7% sched_debug.cfs_rq:/.runnable_load_avg.min
-2516 ±-372% +1102.6% -30258 ± -5% sched_debug.cfs_rq:/.spread0.avg
13923 ± 69% -97.6% 332.39 ± 84% sched_debug.cfs_rq:/.spread0.max
-23392 ±-61% +156.6% -60032 ± -9% sched_debug.cfs_rq:/.spread0.min
14184 ± 15% +66.7% 23649 ± 10% sched_debug.cfs_rq:/.spread0.stddev
981.73 ± 1% -11.1% 872.34 ± 5% sched_debug.cfs_rq:/.util_avg.min
13.21 ± 50% +259.1% 47.44 ± 35% sched_debug.cfs_rq:/.util_avg.stddev
5791458 ± 3% -43.5% 3274044 ± 7% sched_debug.cpu.avg_idle.avg
9237536 ± 8% -45.0% 5079786 ± 6% sched_debug.cpu.avg_idle.max
2851452 ± 10% -32.2% 1933507 ± 8% sched_debug.cpu.avg_idle.min
2476106 ± 16% -50.6% 1223320 ± 8% sched_debug.cpu.avg_idle.stddev
799962 ± 2% +32.0% 1056226 ± 2% sched_debug.cpu.clock.avg
801460 ± 2% +31.9% 1056747 ± 2% sched_debug.cpu.clock.max
798223 ± 2% +32.2% 1055492 ± 2% sched_debug.cpu.clock.min
1204 ± 22% -60.4% 476.30 ± 54% sched_debug.cpu.clock.stddev
732345 ± 2% +33.3% 976382 ± 2% sched_debug.cpu.clock_task.avg
735188 ± 2% +33.3% 980230 ± 2% sched_debug.cpu.clock_task.max
729410 ± 2% +33.4% 972788 ± 2% sched_debug.cpu.clock_task.min
2143 ± 20% +33.0% 2849 ± 9% sched_debug.cpu.clock_task.stddev
51.61 ± 10% -17.4% 42.64 ± 11% sched_debug.cpu.cpu_load[1].stddev
2700 ± 8% +31.7% 3555 ± 9% sched_debug.cpu.curr->pid.avg
4641 ± 0% +45.0% 6732 ± 7% sched_debug.cpu.curr->pid.max
1557 ± 7% +64.5% 2562 ± 6% sched_debug.cpu.curr->pid.stddev
319378 ± 7% -14.0% 274755 ± 4% sched_debug.cpu.load.avg
617372 ± 17% -22.6% 478021 ± 10% sched_debug.cpu.load.max
198535 ± 26% -28.5% 142039 ± 12% sched_debug.cpu.load.stddev
6923798 ± 12% -21.5% 5435214 ± 2% sched_debug.cpu.max_idle_balance_cost.avg
12147808 ± 11% -28.2% 8724945 ± 8% sched_debug.cpu.max_idle_balance_cost.max
3456820 ± 13% -36.1% 2208165 ± 15% sched_debug.cpu.max_idle_balance_cost.stddev
0.00 ± 22% -59.3% 0.00 ± 52% sched_debug.cpu.next_balance.stddev
694723 ± 2% +29.2% 897791 ± 2% sched_debug.cpu.nr_load_updates.avg
701516 ± 2% +28.9% 904484 ± 1% sched_debug.cpu.nr_load_updates.max
688870 ± 2% +29.6% 892655 ± 2% sched_debug.cpu.nr_load_updates.min
27.46 ± 27% -55.3% 12.27 ± 45% sched_debug.cpu.nr_running.avg
38.22 ± 5% -49.0% 19.48 ± 8% sched_debug.cpu.nr_running.max
18.61 ± 62% -63.7% 6.75 ±115% sched_debug.cpu.nr_running.min
417974 ± 0% +42.1% 594084 ± 1% sched_debug.cpu.nr_switches.avg
434615 ± 0% +42.0% 617135 ± 2% sched_debug.cpu.nr_switches.max
402365 ± 1% +43.4% 576858 ± 2% sched_debug.cpu.nr_switches.min
480.59 ± 3% +17.0% 562.30 ± 7% sched_debug.cpu.nr_uninterruptible.max
41.82 ± 26% +97.5% 82.59 ± 26% sched_debug.cpu.nr_uninterruptible.stddev
424839 ± 0% +39.8% 593868 ± 1% sched_debug.cpu.sched_count.avg
440948 ± 0% +39.2% 613987 ± 2% sched_debug.cpu.sched_count.max
409377 ± 1% +41.2% 578031 ± 2% sched_debug.cpu.sched_count.min
3171 ± 4% +1083.4% 37528 ± 13% sched_debug.cpu.sched_goidle.avg
4199 ± 7% +833.0% 39181 ± 12% sched_debug.cpu.sched_goidle.max
2124 ± 19% +1579.8% 35690 ± 15% sched_debug.cpu.sched_goidle.min
781.90 ± 23% +73.8% 1358 ± 33% sched_debug.cpu.sched_goidle.stddev
160749 ± 2% +72.9% 277897 ± 2% sched_debug.cpu.ttwu_count.avg
168447 ± 1% +69.7% 285869 ± 2% sched_debug.cpu.ttwu_count.max
154130 ± 2% +76.3% 271775 ± 2% sched_debug.cpu.ttwu_count.min
57828 ± 2% +42.9% 82608 ± 0% sched_debug.cpu.ttwu_local.avg
64780 ± 1% +38.6% 89761 ± 1% sched_debug.cpu.ttwu_local.max
51318 ± 2% +49.9% 76907 ± 1% sched_debug.cpu.ttwu_local.min
796657 ± 2% +32.4% 1054652 ± 2% sched_debug.cpu_clk
793315 ± 2% +32.5% 1051328 ± 2% sched_debug.ktime
0.00 ± 81% +196.1% 0.01 ± 43% sched_debug.rt_rq:/.rt_time.min
796657 ± 2% +32.4% 1054652 ± 2% sched_debug.sched_clk
=========================================================================================
compiler/kconfig/rootfs/tbox_group/test/testcase:
gcc-6/x86_64-rhel-7.2/debian-x86_64-2016-08-31.cgz/vm-lkp-a03/writeseek3/will-it-scale
commit:
05ffc951392df57edecc2519327b169210c3df75
3cded41794818d788aa1dc028ede4a1c1222d937
05ffc951392df57e 3cded41794818d788aa1dc028e
---------------- --------------------------
%stddev %change %stddev
\ | \
133981 ± 7% -15.7% 112956 ± 7% will-it-scale.per_process_ops
135498 ± 0% -22.4% 105187 ± 1% will-it-scale.per_thread_ops
0.20 ± 5% -53.4% 0.09 ± 6% will-it-scale.scalability
8824 ± 5% -43.5% 4988 ± 9% will-it-scale.time.involuntary_context_switches
117.00 ± 0% -46.2% 63.00 ± 0% will-it-scale.time.percent_of_cpu_this_job_got
254.95 ± 0% -47.3% 134.44 ± 0% will-it-scale.time.system_time
259475 ± 4% +431.2% 1378354 ± 3% will-it-scale.time.voluntary_context_switches
1368 ± 24% +49.3% 2042 ± 7% interrupts.CAL:Function_call_interrupts
2426 ± 6% -15.4% 2052 ± 7% slabinfo.anon_vma.active_objs
5472 ± 4% +394.8% 27077 ± 2% vmstat.system.cs
3197 ± 1% +75.9% 5623 ± 4% vmstat.system.in
35005 ± 2% +48.1% 51829 ± 0% softirqs.RCU
21333 ± 2% +326.5% 90979 ± 0% softirqs.SCHED
308038 ± 0% -28.9% 219138 ± 2% softirqs.TIMER
8824 ± 5% -43.5% 4988 ± 9% time.involuntary_context_switches
117.00 ± 0% -46.2% 63.00 ± 0% time.percent_of_cpu_this_job_got
254.95 ± 0% -47.3% 134.44 ± 0% time.system_time
15.69 ± 0% -24.6% 11.84 ± 1% time.user_time
259475 ± 4% +431.2% 1378354 ± 3% time.voluntary_context_switches
5073 ± 86% +333.7% 21999 ±172% latency_stats.avg.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_getattr.[nfsv4].nfs4_proc_getattr.[nfsv4].__nfs_revalidate_inode.nfs_do_access.nfs_permission.__inode_permission.inode_permission
510237 ± 5% +496.3% 3042399 ± 3% latency_stats.hits.call_rwsem_down_write_failed.generic_file_write_iter.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
510237 ± 5% +496.3% 3042399 ± 3% latency_stats.hits.max
60936 ± 34% +303.6% 245959 ±128% latency_stats.max.max
0.00 ± -1% +Inf% 5416 ±100% latency_stats.max.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
30640 ±127% +545.3% 197733 ±173% latency_stats.max.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_getattr.[nfsv4].nfs4_proc_getattr.[nfsv4].__nfs_revalidate_inode.nfs_do_access.nfs_permission.__inode_permission.inode_permission
7424 ± 77% +321.6% 31304 ±169% latency_stats.max.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_mm_fault.__do_page_fault.trace_do_page_fault.do_async_page_fault.async_page_fault
31808074 ± 1% +808.0% 2.888e+08 ± 0% latency_stats.sum.call_rwsem_down_write_failed.generic_file_write_iter.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
31808074 ± 1% +808.0% 2.888e+08 ± 0% latency_stats.sum.max
0.00 ± -1% +Inf% 10281 ± 99% latency_stats.sum.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
973.00 ± 46% +29729.6% 290242 ± 13% latency_stats.sum.pipe_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
43549 ± 91% +354.6% 197969 ±172% latency_stats.sum.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_getattr.[nfsv4].nfs4_proc_getattr.[nfsv4].__nfs_revalidate_inode.nfs_do_access.nfs_permission.__inode_permission.inode_permission
11317 ± 87% +178.8% 31548 ±168% latency_stats.sum.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_mm_fault.__do_page_fault.trace_do_page_fault.do_async_page_fault.async_page_fault
46612 ± 2% -30.2% 32522 ± 3% sched_debug.cfs_rq:/.exec_clock.avg
87197 ± 0% -18.7% 70902 ± 0% sched_debug.cfs_rq:/.exec_clock.max
18125 ± 8% -33.2% 12115 ± 13% sched_debug.cfs_rq:/.exec_clock.min
25645 ± 2% -11.1% 22809 ± 2% sched_debug.cfs_rq:/.exec_clock.stddev
249.94 ± 4% -15.1% 212.27 ± 7% sched_debug.cfs_rq:/.load_avg.avg
167.19 ± 5% -24.6% 126.06 ± 8% sched_debug.cfs_rq:/.load_avg.min
166834 ± 2% -39.5% 100933 ± 2% sched_debug.cfs_rq:/.min_vruntime.avg
247458 ± 2% -29.3% 174873 ± 3% sched_debug.cfs_rq:/.min_vruntime.max
85072 ± 8% -36.3% 54167 ± 9% sched_debug.cfs_rq:/.min_vruntime.min
59318 ± 6% -23.3% 45506 ± 3% sched_debug.cfs_rq:/.min_vruntime.stddev
0.81 ± 39% -100.0% 0.00 ± -1% sched_debug.cfs_rq:/.nr_spread_over.min
-161804 ± -6% -25.4% -120730 ± -2% sched_debug.cfs_rq:/.spread0.min
59329 ± 6% -23.3% 45518 ± 3% sched_debug.cfs_rq:/.spread0.stddev
814.55 ± 2% -13.6% 703.81 ± 6% sched_debug.cfs_rq:/.util_avg.avg
992.19 ± 1% -11.6% 876.62 ± 2% sched_debug.cfs_rq:/.util_avg.max
651.12 ± 5% -16.8% 541.88 ± 12% sched_debug.cfs_rq:/.util_avg.min
266.14 ± 14% +57.2% 418.49 ± 7% sched_debug.cpu.clock_task.stddev
213.55 ± 9% -21.3% 168.08 ± 10% sched_debug.cpu.cpu_load[0].avg
230.91 ± 5% -21.8% 180.59 ± 10% sched_debug.cpu.cpu_load[1].avg
152.81 ± 3% -45.2% 83.69 ± 29% sched_debug.cpu.cpu_load[1].min
230.55 ± 4% -22.0% 179.84 ± 10% sched_debug.cpu.cpu_load[2].avg
150.19 ± 6% -41.3% 88.12 ± 23% sched_debug.cpu.cpu_load[2].min
228.58 ± 4% -21.8% 178.77 ± 9% sched_debug.cpu.cpu_load[3].avg
148.25 ± 9% -39.9% 89.12 ± 19% sched_debug.cpu.cpu_load[3].min
225.36 ± 4% -21.8% 176.28 ± 8% sched_debug.cpu.cpu_load[4].avg
142.75 ± 8% -37.3% 89.56 ± 16% sched_debug.cpu.cpu_load[4].min
30240 ± 7% -15.5% 25555 ± 6% sched_debug.cpu.nr_load_updates.min
56455 ± 1% +695.6% 449188 ± 3% sched_debug.cpu.nr_switches.avg
79408 ± 9% +667.6% 609573 ± 6% sched_debug.cpu.nr_switches.max
39865 ± 2% +127.2% 90564 ± 4% sched_debug.cpu.nr_switches.min
15379 ± 21% +1269.2% 210574 ± 4% sched_debug.cpu.nr_switches.stddev
2.12 ± 25% +50.0% 3.19 ± 11% sched_debug.cpu.nr_uninterruptible.max
1.54 ± 18% +54.9% 2.39 ± 19% sched_debug.cpu.nr_uninterruptible.stddev
50481 ± 1% +776.6% 442502 ± 3% sched_debug.cpu.sched_count.avg
74230 ± 9% +711.6% 602428 ± 6% sched_debug.cpu.sched_count.max
32590 ± 2% +162.7% 85601 ± 5% sched_debug.cpu.sched_count.min
16115 ± 20% +1198.7% 209295 ± 5% sched_debug.cpu.sched_count.stddev
14868 ± 6% +1322.9% 211568 ± 3% sched_debug.cpu.sched_goidle.avg
19320 ± 8% +1407.3% 291212 ± 7% sched_debug.cpu.sched_goidle.max
10486 ± 10% +211.8% 32696 ± 3% sched_debug.cpu.sched_goidle.min
3421 ± 14% +2960.2% 104701 ± 4% sched_debug.cpu.sched_goidle.stddev
26125 ± 2% +771.8% 227765 ± 2% sched_debug.cpu.ttwu_count.avg
38958 ± 10% +676.7% 302591 ± 3% sched_debug.cpu.ttwu_count.max
18584 ± 12% +148.7% 46227 ± 1% sched_debug.cpu.ttwu_count.min
7961 ± 30% +1225.3% 105512 ± 3% sched_debug.cpu.ttwu_count.stddev
9646 ± 10% -34.8% 6293 ± 7% sched_debug.cpu.ttwu_local.avg
25388 ± 16% -51.3% 12373 ± 8% sched_debug.cpu.ttwu_local.max
9497 ± 20% -56.8% 4102 ± 10% sched_debug.cpu.ttwu_local.stddev
=========================================================================================
compiler/kconfig/rootfs/tbox_group/test/testcase:
gcc-6/x86_64-rhel-7.2/debian-x86_64-2016-08-31.cgz/vm-lkp-a03/writeseek2/will-it-scale
commit:
05ffc951392df57edecc2519327b169210c3df75
3cded41794818d788aa1dc028ede4a1c1222d937
05ffc951392df57e 3cded41794818d788aa1dc028e
---------------- --------------------------
%stddev %change %stddev
\ | \
117707 ± 0% -10.4% 105488 ± 0% will-it-scale.per_thread_ops
0.14 ± 5% -30.0% 0.10 ± 5% will-it-scale.scalability
7386 ± 5% -27.7% 5340 ± 5% will-it-scale.time.involuntary_context_switches
114.50 ± 0% -45.4% 62.50 ± 0% will-it-scale.time.percent_of_cpu_this_job_got
252.41 ± 0% -47.3% 133.15 ± 0% will-it-scale.time.system_time
390729 ± 4% +233.0% 1301165 ± 5% will-it-scale.time.voluntary_context_switches
1385 ± 21% +49.6% 2072 ± 6% interrupts.CAL:Function_call_interrupts
2938 ± 42% +61.1% 4735 ± 29% slabinfo.kmalloc-32.active_objs
2938 ± 42% +61.1% 4735 ± 29% slabinfo.kmalloc-32.num_objs
7965 ± 4% +217.8% 25316 ± 4% vmstat.system.cs
3325 ± 0% +67.9% 5583 ± 5% vmstat.system.in
35250 ± 2% +49.4% 52681 ± 3% softirqs.RCU
24427 ± 2% +271.5% 90748 ± 2% softirqs.SCHED
305640 ± 0% -28.3% 219173 ± 0% softirqs.TIMER
7386 ± 5% -27.7% 5340 ± 5% time.involuntary_context_switches
114.50 ± 0% -45.4% 62.50 ± 0% time.percent_of_cpu_this_job_got
252.41 ± 0% -47.3% 133.15 ± 0% time.system_time
390729 ± 4% +233.0% 1301165 ± 5% time.voluntary_context_switches
0.00 ± -1% +Inf% 4556 ±105% latency_stats.avg.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
789155 ± 5% +259.8% 2839571 ± 4% latency_stats.hits.call_rwsem_down_write_failed.generic_file_write_iter.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
789155 ± 5% +259.8% 2839571 ± 4% latency_stats.hits.max
0.00 ± -1% +Inf% 7416 ±100% latency_stats.max.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
12955 ±106% -95.9% 536.50 ±129% latency_stats.max.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_mm_fault.__do_page_fault.trace_do_page_fault.do_async_page_fault.async_page_fault
45389051 ± 2% +541.5% 2.912e+08 ± 0% latency_stats.sum.call_rwsem_down_write_failed.generic_file_write_iter.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
45389051 ± 2% +541.5% 2.912e+08 ± 0% latency_stats.sum.max
24448 ± 22% +304.0% 98774 ± 10% latency_stats.sum.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
0.00 ± -1% +Inf% 14768 ±125% latency_stats.sum.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
5731 ± 51% +5184.2% 302877 ± 7% latency_stats.sum.pipe_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
17745 ±118% -96.0% 708.75 ±102% latency_stats.sum.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_mm_fault.__do_page_fault.trace_do_page_fault.do_async_page_fault.async_page_fault
46558 ± 2% -47.2% 24580 ± 57% sched_debug.cfs_rq:/.exec_clock.avg
87062 ± 0% -38.4% 53670 ± 57% sched_debug.cfs_rq:/.exec_clock.max
18511 ± 8% -52.9% 8711 ± 58% sched_debug.cfs_rq:/.exec_clock.min
25505 ± 2% -31.9% 17378 ± 57% sched_debug.cfs_rq:/.exec_clock.stddev
157973 ± 23% -73.0% 42690 ± 89% sched_debug.cfs_rq:/.load.min
248.97 ± 3% -34.5% 163.19 ± 58% sched_debug.cfs_rq:/.load_avg.avg
170.56 ± 4% -52.2% 81.50 ± 59% sched_debug.cfs_rq:/.load_avg.min
163483 ± 1% -55.0% 73569 ± 57% sched_debug.cfs_rq:/.min_vruntime.avg
244881 ± 1% -48.5% 126134 ± 57% sched_debug.cfs_rq:/.min_vruntime.max
82001 ± 7% -51.8% 39556 ± 58% sched_debug.cfs_rq:/.min_vruntime.min
59695 ± 5% -45.4% 32602 ± 57% sched_debug.cfs_rq:/.min_vruntime.stddev
0.94 ± 4% -33.3% 0.62 ± 57% sched_debug.cfs_rq:/.nr_running.avg
0.81 ± 13% -69.2% 0.25 ± 70% sched_debug.cfs_rq:/.nr_running.min
232.31 ± 2% -46.5% 124.30 ± 58% sched_debug.cfs_rq:/.runnable_load_avg.avg
370.56 ± 8% -39.2% 225.31 ± 58% sched_debug.cfs_rq:/.runnable_load_avg.max
151.12 ± 4% -76.1% 36.12 ± 58% sched_debug.cfs_rq:/.runnable_load_avg.min
-80526 ± -6% -34.9% -52383 ±-58% sched_debug.cfs_rq:/.spread0.avg
-162018 ± -5% -46.7% -86405 ±-57% sched_debug.cfs_rq:/.spread0.min
59706 ± 5% -45.4% 32609 ± 57% sched_debug.cfs_rq:/.spread0.stddev
815.88 ± 1% -39.2% 496.09 ± 58% sched_debug.cfs_rq:/.util_avg.avg
992.94 ± 1% -34.3% 652.12 ± 57% sched_debug.cfs_rq:/.util_avg.max
660.44 ± 2% -46.0% 356.75 ± 60% sched_debug.cfs_rq:/.util_avg.min
223.78 ± 4% -45.3% 122.33 ± 57% sched_debug.cpu.cpu_load[0].avg
365.50 ± 9% -38.2% 225.94 ± 58% sched_debug.cpu.cpu_load[0].max
126.19 ± 16% -68.8% 39.31 ± 84% sched_debug.cpu.cpu_load[0].min
223.53 ± 4% -42.2% 129.09 ± 57% sched_debug.cpu.cpu_load[1].avg
364.94 ± 9% -38.4% 224.69 ± 58% sched_debug.cpu.cpu_load[1].max
129.06 ± 14% -56.9% 55.56 ± 61% sched_debug.cpu.cpu_load[1].min
220.55 ± 4% -43.2% 125.22 ± 57% sched_debug.cpu.cpu_load[2].avg
363.94 ± 8% -38.8% 222.75 ± 59% sched_debug.cpu.cpu_load[2].max
128.06 ± 12% -57.5% 54.44 ± 61% sched_debug.cpu.cpu_load[2].min
218.84 ± 3% -44.4% 121.75 ± 57% sched_debug.cpu.cpu_load[3].avg
363.44 ± 6% -39.8% 218.94 ± 59% sched_debug.cpu.cpu_load[3].max
129.06 ± 9% -57.2% 55.25 ± 62% sched_debug.cpu.cpu_load[3].min
218.17 ± 2% -45.4% 119.05 ± 57% sched_debug.cpu.cpu_load[4].avg
362.81 ± 4% -41.1% 213.56 ± 59% sched_debug.cpu.cpu_load[4].max
130.88 ± 5% -56.2% 57.38 ± 62% sched_debug.cpu.cpu_load[4].min
1262 ± 6% -41.9% 733.83 ± 62% sched_debug.cpu.curr->pid.avg
562.94 ± 37% -84.7% 86.38 ±104% sched_debug.cpu.curr->pid.min
158697 ± 23% -70.4% 46952 ± 73% sched_debug.cpu.load.min
29618 ± 8% -34.9% 19275 ± 57% sched_debug.cpu.nr_load_updates.min
0.81 ± 13% -69.2% 0.25 ± 70% sched_debug.cpu.nr_running.min
10582 ± 9% -57.9% 4458 ± 58% sched_debug.cpu.ttwu_local.avg
29402 ± 13% -69.0% 9127 ± 58% sched_debug.cpu.ttwu_local.max
11214 ± 15% -72.0% 3144 ± 58% sched_debug.cpu.ttwu_local.stddev
=========================================================================================
compiler/kconfig/rootfs/tbox_group/test/testcase:
gcc-6/x86_64-rhel-7.2/debian-x86_64-2016-08-31.cgz/vm-lkp-a03/unlink1/will-it-scale
commit:
05ffc951392df57edecc2519327b169210c3df75
3cded41794818d788aa1dc028ede4a1c1222d937
05ffc951392df57e 3cded41794818d788aa1dc028e
---------------- --------------------------
%stddev %change %stddev
\ | \
16313 ± 6% -23.6% 12469 ± 0% will-it-scale.per_process_ops
16895 ± 0% -26.5% 12411 ± 0% will-it-scale.per_thread_ops
0.31 ± 1% -60.3% 0.12 ± 1% will-it-scale.scalability
16998 ± 4% -32.5% 11466 ± 8% will-it-scale.time.involuntary_context_switches
113.00 ± 0% -42.7% 64.75 ± 0% will-it-scale.time.percent_of_cpu_this_job_got
253.12 ± 0% -43.1% 144.03 ± 0% will-it-scale.time.system_time
78161 ± 0% +940.8% 813494 ± 1% will-it-scale.time.voluntary_context_switches
1368 ± 17% +36.4% 1866 ± 4% interrupts.CAL:Function_call_interrupts
2802 ± 1% +448.2% 15364 ± 0% vmstat.system.cs
3227 ± 1% +100.4% 6467 ± 0% vmstat.system.in
56760 ± 3% +79.7% 101973 ± 1% softirqs.RCU
36216 ± 3% +154.3% 92080 ± 0% softirqs.SCHED
315821 ± 0% -25.6% 235069 ± 0% softirqs.TIMER
1067890 ± 2% -43.1% 607496 ± 2% proc-vmstat.numa_hit
1067861 ± 2% -43.1% 607468 ± 2% proc-vmstat.numa_local
1762497 ± 2% -44.6% 976765 ± 1% proc-vmstat.pgalloc_normal
1759927 ± 2% -44.7% 974114 ± 1% proc-vmstat.pgfree
2922 ± 7% -38.1% 1808 ± 2% slabinfo.kmalloc-256.active_objs
2968 ± 7% -37.5% 1855 ± 2% slabinfo.kmalloc-256.num_objs
2743 ± 7% -38.8% 1680 ± 3% slabinfo.shmem_inode_cache.active_objs
2772 ± 7% -38.5% 1705 ± 3% slabinfo.shmem_inode_cache.num_objs
16998 ± 4% -32.5% 11466 ± 8% time.involuntary_context_switches
113.00 ± 0% -42.7% 64.75 ± 0% time.percent_of_cpu_this_job_got
253.12 ± 0% -43.1% 144.03 ± 0% time.system_time
8.11 ± 1% -27.4% 5.88 ± 3% time.user_time
78161 ± 0% +940.8% 813494 ± 1% time.voluntary_context_switches
18200 ± 82% -68.3% 5774 ± 62% latency_stats.avg.max
84687 ± 5% +775.3% 741232 ± 1% latency_stats.hits.call_rwsem_down_write_failed.do_unlinkat.SyS_unlink.entry_SYSCALL_64_fastpath
79821 ± 4% +998.5% 876839 ± 0% latency_stats.hits.call_rwsem_down_write_failed.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
84780 ± 5% +934.3% 876839 ± 0% latency_stats.hits.max
926.75 ±173% +490.0% 5467 ± 88% latency_stats.max.call_rwsem_down_write_failed.unlink_file_vma.free_pgtables.exit_mmap.mmput.flush_old_exec.load_elf_binary.search_binary_handler.do_execveat_common.SyS_execve.do_syscall_64.return_from_SYSCALL_64
90755 ± 73% -52.2% 43373 ± 9% latency_stats.max.max
8546 ± 38% -86.9% 1123 ± 36% latency_stats.max.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
7734 ± 85% -85.4% 1127 ± 47% latency_stats.max.stop_one_cpu.sched_exec.do_execveat_common.SyS_execve.do_syscall_64.return_from_SYSCALL_64
19824666 ± 3% +553.4% 1.295e+08 ± 0% latency_stats.sum.call_rwsem_down_write_failed.do_unlinkat.SyS_unlink.entry_SYSCALL_64_fastpath
18676250 ± 2% +666.0% 1.431e+08 ± 0% latency_stats.sum.call_rwsem_down_write_failed.path_openat.do_filp_open.do_sys_open.SyS_open.entry_SYSCALL_64_fastpath
926.75 ±173% +739.2% 7777 ± 99% latency_stats.sum.call_rwsem_down_write_failed.unlink_file_vma.free_pgtables.exit_mmap.mmput.flush_old_exec.load_elf_binary.search_binary_handler.do_execveat_common.SyS_execve.do_syscall_64.return_from_SYSCALL_64
19824666 ± 3% +621.6% 1.431e+08 ± 0% latency_stats.sum.max
21861 ± 35% +281.3% 83362 ± 7% latency_stats.sum.pipe_read.__vfs_read.vfs_read.SyS_read.entry_SYSCALL_64_fastpath
8887 ± 60% +3600.2% 328845 ± 3% latency_stats.sum.pipe_write.__vfs_write.vfs_write.SyS_write.entry_SYSCALL_64_fastpath
47251 ± 2% -26.7% 34657 ± 0% sched_debug.cfs_rq:/.exec_clock.avg
86652 ± 0% -16.5% 72331 ± 0% sched_debug.cfs_rq:/.exec_clock.max
22604 ± 8% -35.7% 14535 ± 9% sched_debug.cfs_rq:/.exec_clock.min
284.03 ± 6% -19.8% 227.78 ± 7% sched_debug.cfs_rq:/.load_avg.avg
187.00 ± 6% -38.2% 115.50 ± 8% sched_debug.cfs_rq:/.load_avg.min
156921 ± 2% -37.5% 98113 ± 0% sched_debug.cfs_rq:/.min_vruntime.avg
233602 ± 2% -29.5% 164691 ± 1% sched_debug.cfs_rq:/.min_vruntime.max
83618 ± 6% -35.4% 54026 ± 2% sched_debug.cfs_rq:/.min_vruntime.min
54783 ± 4% -24.5% 41351 ± 2% sched_debug.cfs_rq:/.min_vruntime.stddev
207.30 ± 3% -15.7% 174.77 ± 9% sched_debug.cfs_rq:/.runnable_load_avg.avg
97.31 ± 3% -43.8% 54.69 ± 17% sched_debug.cfs_rq:/.runnable_load_avg.min
-76348 ± -4% -13.0% -66417 ± -3% sched_debug.cfs_rq:/.spread0.avg
-149671 ± -4% -26.2% -110516 ± -2% sched_debug.cfs_rq:/.spread0.min
54797 ± 4% -24.5% 41361 ± 2% sched_debug.cfs_rq:/.spread0.stddev
838.38 ± 2% -20.6% 665.44 ± 3% sched_debug.cfs_rq:/.util_avg.avg
1009 ± 0% -13.5% 873.06 ± 2% sched_debug.cfs_rq:/.util_avg.max
698.25 ± 4% -27.7% 504.75 ± 8% sched_debug.cfs_rq:/.util_avg.min
267.78 ± 20% +144.8% 655.64 ± 16% sched_debug.cpu.clock_task.stddev
213.39 ± 6% -20.8% 169.06 ± 11% sched_debug.cpu.cpu_load[0].avg
116.38 ± 24% -68.8% 36.31 ± 54% sched_debug.cpu.cpu_load[0].min
234.58 ± 2% -18.5% 191.17 ± 8% sched_debug.cpu.cpu_load[1].avg
158.19 ± 3% -47.0% 83.81 ± 33% sched_debug.cpu.cpu_load[1].min
232.41 ± 3% -20.4% 185.09 ± 8% sched_debug.cpu.cpu_load[2].avg
154.25 ± 3% -48.6% 79.31 ± 25% sched_debug.cpu.cpu_load[2].min
229.50 ± 3% -21.3% 180.53 ± 7% sched_debug.cpu.cpu_load[3].avg
150.00 ± 3% -48.2% 77.62 ± 17% sched_debug.cpu.cpu_load[3].min
227.19 ± 3% -22.3% 176.44 ± 7% sched_debug.cpu.cpu_load[4].avg
147.00 ± 4% -46.1% 79.25 ± 14% sched_debug.cpu.cpu_load[4].min
418.31 ± 38% -80.6% 81.19 ± 2% sched_debug.cpu.curr->pid.min
528.82 ± 12% +31.8% 696.79 ± 1% sched_debug.cpu.curr->pid.stddev
38066 ± 2% -16.8% 31687 ± 10% sched_debug.cpu.nr_load_updates.min
24488 ± 3% +9.9% 26915 ± 3% sched_debug.cpu.nr_load_updates.stddev
1.30 ± 11% -18.1% 1.06 ± 10% sched_debug.cpu.nr_running.avg
49239 ± 1% +415.4% 253772 ± 1% sched_debug.cpu.nr_switches.avg
77869 ± 10% +320.9% 327779 ± 1% sched_debug.cpu.nr_switches.max
30323 ± 15% +158.7% 78446 ± 9% sched_debug.cpu.nr_switches.min
18253 ± 25% +464.5% 103041 ± 4% sched_debug.cpu.nr_switches.stddev
43018 ± 1% +474.6% 247180 ± 1% sched_debug.cpu.sched_count.avg
72309 ± 12% +342.4% 319861 ± 1% sched_debug.cpu.sched_count.max
22541 ± 19% +224.9% 73230 ± 10% sched_debug.cpu.sched_count.min
19100 ± 22% +432.6% 101731 ± 4% sched_debug.cpu.sched_count.stddev
11814 ± 7% +842.4% 111345 ± 1% sched_debug.cpu.sched_goidle.avg
19421 ± 17% +664.7% 148514 ± 2% sched_debug.cpu.sched_goidle.max
6494 ± 15% +278.2% 24559 ± 11% sched_debug.cpu.sched_goidle.min
5147 ± 19% +888.0% 50856 ± 3% sched_debug.cpu.sched_goidle.stddev
22211 ± 1% +475.1% 127736 ± 1% sched_debug.cpu.ttwu_count.avg
35498 ± 11% +365.5% 165259 ± 3% sched_debug.cpu.ttwu_count.max
14006 ± 17% +155.9% 35840 ± 9% sched_debug.cpu.ttwu_count.min
8309 ± 27% +544.9% 53589 ± 3% sched_debug.cpu.ttwu_count.stddev
Best Regards,
Huang, Ying
next prev parent reply other threads:[~2016-12-08 8:39 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-30 18:07 [x86, paravirt] fd6f48529f: aim7.jobs-per-min -26.1% regression kernel test robot
2016-12-01 7:04 ` Peter Zijlstra
2016-11-30 23:17 ` Ye Xiaolong
2016-12-01 7:21 ` Peter Zijlstra
2016-12-01 16:49 ` Ye Xiaolong
2016-12-07 13:17 ` Peter Zijlstra
2016-12-08 1:12 ` Huang, Ying
2016-12-08 8:39 ` Huang, Ying [this message]
2016-12-08 11:08 ` Peter Zijlstra
2016-12-09 2:36 ` Huang, Ying
2016-12-09 5:36 ` Peter Zijlstra
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87bmwm6ci9.fsf@yhuang-dev.intel.com \
--to=ying.huang@intel.com \
--cc=lkp@lists.01.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.