From: Huang Ying <ying.huang@intel.com>
To: lkp@lists.01.org
Subject: [sched] d2345b70b2f: -1.1% hackbench.throughput, +58.1% hackbench.time.involuntary_context_switches
Date: Thu, 04 Jun 2015 08:52:37 +0800 [thread overview]
Message-ID: <1433379157.7032.40.camel@intel.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 41433 bytes --]
FYI, we noticed the below changes on
git://bee.sh.intel.com/git/ydu19/tip for-lkp
commit d2345b70b2f9dc2963f417921d8ad53bc2ce0799 ("sched: Rewrite runnable load and utilization average tracking")
testcase/path_params/tbox_group: hackbench/performance-1600%-process-socket/lkp-ne04
53a5dba0cf8766d0 d2345b70b2f9dc2963f417921d
---------------- --------------------------
%stddev %change %stddev
\ | \
71948 ± 0% -1.1% 71183 ± 0% hackbench.throughput
14259643 ± 9% +58.1% 22546553 ± 4% hackbench.time.involuntary_context_switches
395 ± 0% -3.4% 382 ± 2% hackbench.time.user_time
1.106e+08 ± 1% +8.4% 1.199e+08 ± 1% hackbench.time.voluntary_context_switches
1715335 ± 3% +54.1% 2642893 ± 6% proc-vmstat.pgalloc_dma32
14259643 ± 9% +58.1% 22546553 ± 4% time.involuntary_context_switches
665068 ± 2% +14.7% 762941 ± 2% softirqs.RCU
1.64 ± 1% +38.5% 2.27 ± 0% turbostat.CPU%c1
0.15 ± 7% -44.8% 0.08 ± 8% turbostat.CPU%c3
0.70 ± 3% -63.9% 0.25 ± 9% turbostat.CPU%c6
0 ± 0% +Inf% 5 ± 28% vmstat.procs.b
203727 ± 2% +15.0% 234375 ± 1% vmstat.system.cs
20533 ± 1% +5.2% 21598 ± 1% vmstat.system.in
3031995 ± 4% +58.4% 4802543 ± 7% numa-numastat.node0.numa_hit
3032004 ± 4% +58.4% 4802526 ± 7% numa-numastat.node0.local_node
7303437 ± 1% -26.0% 5402219 ± 9% numa-numastat.node1.local_node
7303467 ± 1% -26.0% 5402244 ± 9% numa-numastat.node1.numa_hit
14612 ± 2% +8.4% 15841 ± 2% slabinfo.kmalloc-128.active_objs
6889 ± 2% -10.8% 6142 ± 0% slabinfo.kmalloc-256.active_slabs
6889 ± 2% -10.8% 6142 ± 0% slabinfo.kmalloc-256.num_slabs
220462 ± 2% -10.8% 196565 ± 0% slabinfo.kmalloc-256.num_objs
6927 ± 2% -10.7% 6188 ± 0% slabinfo.kmalloc-512.num_slabs
221687 ± 2% -10.7% 198053 ± 0% slabinfo.kmalloc-512.num_objs
6927 ± 2% -10.7% 6188 ± 0% slabinfo.kmalloc-512.active_slabs
698677 ± 2% +68.2% 1175487 ± 3% cpuidle.C1-NHM.usage
1.112e+08 ± 1% -27.2% 80921308 ± 3% cpuidle.C1-NHM.time
81180 ± 4% +116.7% 175922 ± 5% cpuidle.C1E-NHM.usage
14459073 ± 8% +630.9% 1.057e+08 ± 4% cpuidle.C1E-NHM.time
145745 ± 1% -86.2% 20169 ± 20% cpuidle.C3-NHM.usage
20382464 ± 2% -32.9% 13670003 ± 8% cpuidle.C3-NHM.time
99811142 ± 1% -44.5% 55397968 ± 6% cpuidle.C6-NHM.time
34051 ± 3% -27.3% 24753 ± 6% cpuidle.C6-NHM.usage
197 ± 4% +157.6% 507 ± 7% cpuidle.POLL.usage
1260 ± 0% +59.7% 2012 ± 20% numa-vmstat.node0.nr_mapped
1707663 ± 6% +42.8% 2438291 ± 8% numa-vmstat.node0.numa_hit
501 ± 41% +760.9% 4319 ± 12% numa-vmstat.node0.nr_kernel_stack
26088 ± 4% +78.6% 46595 ± 7% numa-vmstat.node0.nr_slab_unreclaimable
28873 ± 4% +32.6% 38294 ± 5% numa-vmstat.node0.nr_anon_pages
16980 ± 5% +79.5% 30480 ± 8% numa-vmstat.node0.nr_page_table_pages
30952 ± 4% +27.2% 39380 ± 6% numa-vmstat.node0.nr_active_anon
1646861 ± 6% +45.2% 2391626 ± 9% numa-vmstat.node0.numa_local
3867845 ± 2% -22.1% 3012126 ± 5% numa-vmstat.node1.numa_hit
53281 ± 2% -18.1% 43650 ± 7% numa-vmstat.node1.nr_active_anon
8376 ± 3% -44.0% 4691 ± 12% numa-vmstat.node1.nr_kernel_stack
51202 ± 2% -20.7% 40600 ± 8% numa-vmstat.node1.nr_anon_pages
3862823 ± 2% -22.5% 2992841 ± 5% numa-vmstat.node1.numa_local
48137 ± 2% -31.4% 32998 ± 10% numa-vmstat.node1.nr_page_table_pages
71449 ± 2% -32.8% 48007 ± 7% numa-vmstat.node1.nr_slab_unreclaimable
6534 ± 6% -13.3% 5663 ± 3% numa-vmstat.node1.nr_slab_reclaimable
123850 ± 3% +27.9% 158364 ± 7% numa-meminfo.node0.Active(anon)
8359 ± 42% +732.7% 69614 ± 15% numa-meminfo.node0.KernelStack
5060 ± 0% +59.0% 8047 ± 20% numa-meminfo.node0.Mapped
103613 ± 5% +80.7% 187188 ± 8% numa-meminfo.node0.SUnreclaim
115570 ± 5% +33.3% 154023 ± 7% numa-meminfo.node0.AnonPages
170936 ± 3% +21.5% 207698 ± 5% numa-meminfo.node0.Active
68046 ± 6% +80.0% 122507 ± 10% numa-meminfo.node0.PageTables
597746 ± 2% +53.8% 919321 ± 6% numa-meminfo.node0.MemUsed
121251 ± 4% +71.7% 208224 ± 7% numa-meminfo.node0.Slab
26084 ± 6% -13.1% 22655 ± 4% numa-meminfo.node1.SReclaimable
133675 ± 2% -43.4% 75633 ± 12% numa-meminfo.node1.KernelStack
205113 ± 1% -20.6% 162848 ± 7% numa-meminfo.node1.AnonPages
192341 ± 1% -31.1% 132448 ± 11% numa-meminfo.node1.PageTables
310539 ± 1% -30.8% 215008 ± 7% numa-meminfo.node1.Slab
262277 ± 3% -15.3% 222255 ± 5% numa-meminfo.node1.Active
284455 ± 1% -32.4% 192352 ± 7% numa-meminfo.node1.SUnreclaim
1286365 ± 0% -26.1% 950143 ± 6% numa-meminfo.node1.MemUsed
213339 ± 3% -18.0% 175024 ± 7% numa-meminfo.node1.Active(anon)
32 ± 34% +123.1% 72 ± 17% latency_stats.avg.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
560 ± 2% -11.2% 497 ± 1% latency_stats.avg.do_wait.SyS_wait4.system_call_fastpath
2347 ± 0% +2.7% 2411 ± 0% latency_stats.avg.poll_schedule_timeout.do_sys_poll.SyS_poll.system_call_fastpath
227 ± 2% -13.0% 197 ± 5% latency_stats.avg.pipe_wait.pipe_read.__vfs_read.vfs_read.SyS_read.system_call_fastpath
232 ± 14% -34.5% 152 ± 3% latency_stats.avg.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_access.[nfsv4].nfs4_proc_access.[nfsv4].nfs_do_access.nfs_permission.__inode_permission.inode_permission.may_open
2022 ± 3% -24.2% 1534 ± 3% latency_stats.avg.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.do_last.path_openat.do_filp_open.do_sys_open.SyS_open.system_call_fastpath
277 ± 30% +290.6% 1084 ± 14% latency_stats.avg.ep_poll.SyS_epoll_wait.system_call_fastpath
44572 ± 35% +604.9% 314192 ± 12% latency_stats.avg.wait_on_page_bit.__migration_entry_wait.migration_entry_wait.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
503 ± 0% -7.5% 465 ± 1% latency_stats.avg.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.__vfs_read.vfs_read.SyS_read.system_call_fastpath
1837 ± 0% +3.4% 1899 ± 0% latency_stats.avg.sock_alloc_send_pskb.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.__vfs_write.vfs_write.SyS_write.system_call_fastpath
2418 ± 47% +442.6% 13123 ± 15% latency_stats.hits.wait_on_page_bit.__migration_entry_wait.migration_entry_wait.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
12 ± 18% +75.5% 21 ± 42% latency_stats.hits.call_rwsem_down_read_failed.page_lock_anon_vma_read.rmap_walk.try_to_unmap.migrate_pages.migrate_misplaced_page.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
16 ± 17% +104.5% 33 ± 13% latency_stats.hits.call_rwsem_down_read_failed.rmap_walk.remove_migration_ptes.migrate_pages.migrate_misplaced_page.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
852 ± 5% -27.3% 619 ± 22% latency_stats.hits.wait_on_page_bit_killable.__lock_page_or_retry.filemap_fault.__do_fault.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
67862917 ± 2% +17.9% 80007991 ± 2% latency_stats.hits.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.__vfs_read.vfs_read.SyS_read.system_call_fastpath
1 ± 33% -100.0% 0 ± 0% latency_stats.hits.stop_one_cpu.sched_exec.do_execveat_common.SyS_execve.return_from_execve
225174 ± 45% +567.0% 1501845 ± 10% latency_stats.max.wait_on_page_bit.__migration_entry_wait.migration_entry_wait.handle_pte_fault.handle_mm_fault.__do_page_fault.do_page_fault.page_fault
2561 ± 4% -17.7% 2108 ± 6% latency_stats.max.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.do_last.path_openat.do_filp_open.do_sys_open.SyS_open.system_call_fastpath
3368 ± 30% +36.2% 4587 ± 3% latency_stats.max.wait_woken.inotify_read.__vfs_read.vfs_read.SyS_read.system_call_fastpath
1100786 ± 3% -15.7% 927862 ± 5% latency_stats.sum.pipe_wait.pipe_read.__vfs_read.vfs_read.SyS_read.system_call_fastpath
20232 ± 3% -24.2% 15345 ± 3% latency_stats.sum.pipe_wait.wait_for_partner.fifo_open.do_dentry_open.vfs_open.do_last.path_openat.do_filp_open.do_sys_open.SyS_open.system_call_fastpath
3.416e+10 ± 1% +9.1% 3.727e+10 ± 1% latency_stats.sum.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.__vfs_read.vfs_read.SyS_read.system_call_fastpath
7891 ± 30% +256.1% 28099 ± 33% latency_stats.sum.ep_poll.SyS_epoll_wait.system_call_fastpath
6761 ± 14% -32.8% 4544 ± 4% latency_stats.sum.rpc_wait_bit_killable.__rpc_execute.rpc_execute.rpc_run_task.nfs4_call_sync_sequence.[nfsv4]._nfs4_proc_access.[nfsv4].nfs4_proc_access.[nfsv4].nfs_do_access.nfs_permission.__inode_permission.inode_permission.may_open
1726737 ± 3% -15.8% 1454425 ± 6% latency_stats.sum.do_wait.SyS_wait4.system_call_fastpath
3.183e+10 ± 0% -3.9% 3.06e+10 ± 2% latency_stats.sum.sock_alloc_send_pskb.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.__vfs_write.vfs_write.SyS_write.system_call_fastpath
2.00 ± 18% +69.8% 3.40 ± 13% perf-profile.cpu-cycles.enqueue_entity.enqueue_task_fair.enqueue_task.activate_task.ttwu_do_activate
8.90 ± 11% +20.8% 10.75 ± 1% perf-profile.cpu-cycles.sock_alloc_send_pskb.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.__vfs_write
0.84 ± 15% -60.6% 0.33 ± 49% perf-profile.cpu-cycles.__kmalloc_node_track_caller.__kmalloc_reserve.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
10.41 ± 14% -50.8% 5.12 ± 45% perf-profile.cpu-cycles.__vfs_write.vfs_write.sys_write.system_call_fastpath.__write_nocancel
0.97 ± 14% -59.4% 0.39 ± 49% perf-profile.cpu-cycles.__kmalloc_reserve.isra.27.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg
30.18 ± 4% +15.3% 34.81 ± 3% perf-profile.cpu-cycles.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.__vfs_write.vfs_write
33.61 ± 4% +11.9% 37.61 ± 4% perf-profile.cpu-cycles.sock_sendmsg.sock_write_iter.__vfs_write.vfs_write.sys_write
2.10 ± 17% +74.2% 3.67 ± 13% perf-profile.cpu-cycles.ttwu_do_activate.constprop.85.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common
1.37 ± 14% +76.1% 2.42 ± 13% perf-profile.cpu-cycles.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity
3.30 ± 11% +49.2% 4.93 ± 11% perf-profile.cpu-cycles.sock_def_readable.unix_stream_sendmsg.sock_sendmsg.sock_write_iter.__vfs_write
2.58 ± 8% -37.9% 1.60 ± 27% perf-profile.cpu-cycles.skb_copy_datagram_iter.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.__vfs_read
2.01 ± 17% +72.6% 3.47 ± 14% perf-profile.cpu-cycles.enqueue_task_fair.enqueue_task.activate_task.ttwu_do_activate.try_to_wake_up
2.05 ± 17% +73.0% 3.55 ± 14% perf-profile.cpu-cycles.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function.autoremove_wake_function
2.54 ± 12% +69.8% 4.30 ± 11% perf-profile.cpu-cycles.__wake_up_common.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg.sock_sendmsg
0.69 ± 12% +77.1% 1.22 ± 15% perf-profile.cpu-cycles.__kernel_text_address.print_context_stack.dump_trace.save_stack_trace_tsk.__account_scheduler_latency
2.47 ± 12% +71.7% 4.25 ± 11% perf-profile.cpu-cycles.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.sock_def_readable
5.02 ± 3% -16.2% 4.21 ± 9% perf-profile.cpu-cycles.skb_release_all.consume_skb.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
1.57 ± 17% +76.6% 2.78 ± 14% perf-profile.cpu-cycles.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.enqueue_task
2.48 ± 12% +72.0% 4.26 ± 11% perf-profile.cpu-cycles.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg
1.57 ± 17% +76.5% 2.76 ± 14% perf-profile.cpu-cycles.dump_trace.save_stack_trace_tsk.__account_scheduler_latency.enqueue_entity.enqueue_task_fair
2.72 ± 13% +67.1% 4.55 ± 11% perf-profile.cpu-cycles.try_to_wake_up.default_wake_function.autoremove_wake_function.__wake_up_common.__wake_up_sync_key
35.38 ± 4% +10.0% 38.92 ± 3% perf-profile.cpu-cycles.sock_write_iter.__vfs_write.vfs_write.sys_write.system_call_fastpath
2.06 ± 17% +71.5% 3.53 ± 14% perf-profile.cpu-cycles.enqueue_task.activate_task.ttwu_do_activate.try_to_wake_up.default_wake_function
1.74 ± 19% +74.1% 3.03 ± 14% perf-profile.cpu-cycles.__account_scheduler_latency.enqueue_entity.enqueue_task_fair.enqueue_task.activate_task
2.83 ± 11% +60.8% 4.55 ± 10% perf-profile.cpu-cycles.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg.sock_sendmsg.sock_write_iter
56 ± 2% -100.0% 0 ± 0% sched_debug.cfs_rq[0]:/.runnable_load_avg
7860 ± 20% -87.7% 965 ± 2% sched_debug.cfs_rq[0]:/.tg_load_avg
62 ± 10% -13.9% 54 ± 1% sched_debug.cfs_rq[0]:/.load
549 ± 19% -100.0% 0 ± 0% sched_debug.cfs_rq[0]:/.utilization_load_avg
225 ± 13% -100.0% 0 ± 0% sched_debug.cfs_rq[10]:/.utilization_load_avg
41338 ± 12% -97.5% 1022 ± 3% sched_debug.cfs_rq[10]:/.tg_load_avg
54 ± 19% -100.0% 0 ± 0% sched_debug.cfs_rq[10]:/.runnable_load_avg
3762 ± 27% -100.0% 0 ± 0% sched_debug.cfs_rq[10]:/.tg_load_contrib
3703 ± 28% -100.0% 0 ± 0% sched_debug.cfs_rq[10]:/.blocked_load_avg
42 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[11]:/.runnable_load_avg
39600 ± 6% -97.4% 1019 ± 3% sched_debug.cfs_rq[11]:/.tg_load_avg
2197 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[11]:/.tg_load_contrib
2151 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[11]:/.blocked_load_avg
322 ± 26% -100.0% 0 ± 0% sched_debug.cfs_rq[11]:/.utilization_load_avg
68 ± 16% -100.0% 0 ± 0% sched_debug.cfs_rq[12]:/.runnable_load_avg
39471 ± 5% -97.4% 1018 ± 4% sched_debug.cfs_rq[12]:/.tg_load_avg
3 ± 24% +107.1% 7 ± 32% sched_debug.cfs_rq[12]:/.nr_spread_over
3093 ± 22% -100.0% 0 ± 0% sched_debug.cfs_rq[12]:/.tg_load_contrib
382 ± 21% -100.0% 0 ± 0% sched_debug.cfs_rq[12]:/.utilization_load_avg
3024 ± 22% -100.0% 0 ± 0% sched_debug.cfs_rq[12]:/.blocked_load_avg
48 ± 27% -100.0% 0 ± 0% sched_debug.cfs_rq[13]:/.runnable_load_avg
2946 ± 30% -100.0% 0 ± 0% sched_debug.cfs_rq[13]:/.tg_load_contrib
374 ± 16% -100.0% 0 ± 0% sched_debug.cfs_rq[13]:/.utilization_load_avg
38273 ± 10% -97.3% 1023 ± 4% sched_debug.cfs_rq[13]:/.tg_load_avg
2897 ± 31% -100.0% 0 ± 0% sched_debug.cfs_rq[13]:/.blocked_load_avg
38999 ± 8% -97.4% 1033 ± 3% sched_debug.cfs_rq[14]:/.tg_load_avg
1763 ± 39% -100.0% 0 ± 0% sched_debug.cfs_rq[14]:/.blocked_load_avg
71 ± 7% -100.0% 0 ± 0% sched_debug.cfs_rq[14]:/.runnable_load_avg
355 ± 27% -100.0% 0 ± 0% sched_debug.cfs_rq[14]:/.utilization_load_avg
1832 ± 38% -100.0% 0 ± 0% sched_debug.cfs_rq[14]:/.tg_load_contrib
2372 ± 46% -100.0% 0 ± 0% sched_debug.cfs_rq[15]:/.tg_load_contrib
46 ± 14% -100.0% 0 ± 0% sched_debug.cfs_rq[15]:/.runnable_load_avg
2315 ± 48% -100.0% 0 ± 0% sched_debug.cfs_rq[15]:/.blocked_load_avg
37699 ± 9% -97.3% 1035 ± 4% sched_debug.cfs_rq[15]:/.tg_load_avg
344 ± 21% -100.0% 0 ± 0% sched_debug.cfs_rq[15]:/.utilization_load_avg
10747 ± 18% -91.0% 966 ± 2% sched_debug.cfs_rq[1]:/.tg_load_avg
52 ± 2% -100.0% 0 ± 0% sched_debug.cfs_rq[1]:/.runnable_load_avg
704 ± 40% -100.0% 0 ± 0% sched_debug.cfs_rq[1]:/.utilization_load_avg
5093773 ± 4% -9.6% 4606679 ± 4% sched_debug.cfs_rq[1]:/.min_vruntime
1401 ± 47% -100.0% 0 ± 0% sched_debug.cfs_rq[2]:/.blocked_load_avg
1460 ± 45% -100.0% 0 ± 0% sched_debug.cfs_rq[2]:/.tg_load_contrib
22584 ± 33% -95.7% 974 ± 3% sched_debug.cfs_rq[2]:/.tg_load_avg
59 ± 5% -100.0% 0 ± 0% sched_debug.cfs_rq[2]:/.runnable_load_avg
1484 ± 41% -100.0% 0 ± 0% sched_debug.cfs_rq[3]:/.blocked_load_avg
1536 ± 40% -100.0% 0 ± 0% sched_debug.cfs_rq[3]:/.tg_load_contrib
644 ± 35% -100.0% 0 ± 0% sched_debug.cfs_rq[3]:/.utilization_load_avg
5279976 ± 3% -10.8% 4708651 ± 2% sched_debug.cfs_rq[3]:/.min_vruntime
52 ± 10% -100.0% 0 ± 0% sched_debug.cfs_rq[3]:/.runnable_load_avg
30501 ± 8% -96.8% 981 ± 3% sched_debug.cfs_rq[3]:/.tg_load_avg
1967 ± 42% -100.0% 0 ± 0% sched_debug.cfs_rq[4]:/.tg_load_contrib
295 ± 15% -100.0% 0 ± 0% sched_debug.cfs_rq[4]:/.utilization_load_avg
1871 ± 42% -100.0% 0 ± 0% sched_debug.cfs_rq[4]:/.blocked_load_avg
57 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[4]:/.runnable_load_avg
31557 ± 5% -96.9% 984 ± 3% sched_debug.cfs_rq[4]:/.tg_load_avg
54 ± 21% -100.0% 0 ± 0% sched_debug.cfs_rq[5]:/.runnable_load_avg
14 ± 27% +64.9% 23 ± 32% sched_debug.cfs_rq[5]:/.nr_spread_over
456 ± 16% -100.0% 0 ± 0% sched_debug.cfs_rq[5]:/.utilization_load_avg
35890 ± 9% -97.2% 1018 ± 2% sched_debug.cfs_rq[5]:/.tg_load_avg
3902 ± 30% -100.0% 0 ± 0% sched_debug.cfs_rq[5]:/.tg_load_contrib
3815 ± 30% -100.0% 0 ± 0% sched_debug.cfs_rq[5]:/.blocked_load_avg
360 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[6]:/.utilization_load_avg
60 ± 17% -100.0% 0 ± 0% sched_debug.cfs_rq[6]:/.runnable_load_avg
1601 ± 41% -100.0% 0 ± 0% sched_debug.cfs_rq[6]:/.tg_load_contrib
36619 ± 13% -97.2% 1010 ± 3% sched_debug.cfs_rq[6]:/.tg_load_avg
1541 ± 43% -100.0% 0 ± 0% sched_debug.cfs_rq[6]:/.blocked_load_avg
47 ± 9% -100.0% 0 ± 0% sched_debug.cfs_rq[7]:/.runnable_load_avg
16 ± 30% +116.7% 35 ± 33% sched_debug.cfs_rq[7]:/.nr_spread_over
38064 ± 13% -97.3% 1022 ± 4% sched_debug.cfs_rq[7]:/.tg_load_avg
1765 ± 39% -100.0% 0 ± 0% sched_debug.cfs_rq[7]:/.blocked_load_avg
1813 ± 38% -100.0% 0 ± 0% sched_debug.cfs_rq[7]:/.tg_load_contrib
5414683 ± 2% -8.5% 4953680 ± 1% sched_debug.cfs_rq[7]:/.min_vruntime
326 ± 39% -100.0% 0 ± 0% sched_debug.cfs_rq[7]:/.utilization_load_avg
361 ± 38% -100.0% 0 ± 0% sched_debug.cfs_rq[8]:/.utilization_load_avg
2114 ± 42% -100.0% 0 ± 0% sched_debug.cfs_rq[8]:/.tg_load_contrib
61 ± 12% -100.0% 0 ± 0% sched_debug.cfs_rq[8]:/.runnable_load_avg
38503 ± 12% -97.3% 1022 ± 4% sched_debug.cfs_rq[8]:/.tg_load_avg
2052 ± 43% -100.0% 0 ± 0% sched_debug.cfs_rq[8]:/.blocked_load_avg
5301883 ± 6% -12.3% 4647905 ± 2% sched_debug.cfs_rq[9]:/.min_vruntime
2170 ± 22% -100.0% 0 ± 0% sched_debug.cfs_rq[9]:/.tg_load_contrib
54 ± 9% -100.0% 0 ± 0% sched_debug.cfs_rq[9]:/.runnable_load_avg
2115 ± 23% -100.0% 0 ± 0% sched_debug.cfs_rq[9]:/.blocked_load_avg
370 ± 41% -100.0% 0 ± 0% sched_debug.cfs_rq[9]:/.utilization_load_avg
40533 ± 12% -97.5% 1013 ± 3% sched_debug.cfs_rq[9]:/.tg_load_avg
3429160 ± 4% +12.4% 3852887 ± 4% sched_debug.cpu#0.nr_switches
3460339 ± 3% +11.5% 3859077 ± 4% sched_debug.cpu#0.sched_count
3255972 ± 1% +19.9% 3904059 ± 3% sched_debug.cpu#1.nr_switches
2972896 ± 1% +12.1% 3331611 ± 1% sched_debug.cpu#1.ttwu_count
955576 ± 8% -24.1% 724961 ± 17% sched_debug.cpu#1.avg_idle
1123082 ± 12% -55.5% 500000 ± 0% sched_debug.cpu#1.max_idle_balance_cost
18121 ± 12% +95.1% 35352 ± 7% sched_debug.cpu#1.sched_goidle
3299562 ± 1% +19.8% 3953430 ± 2% sched_debug.cpu#1.sched_count
32042 ± 18% +29.4% 41467 ± 10% sched_debug.cpu#10.sched_goidle
3626516 ± 3% +22.6% 4446037 ± 2% sched_debug.cpu#11.sched_count
18988 ± 12% +95.7% 37163 ± 20% sched_debug.cpu#11.sched_goidle
891501 ± 14% -43.9% 500000 ± 0% sched_debug.cpu#11.max_idle_balance_cost
3590230 ± 3% +23.6% 4436784 ± 2% sched_debug.cpu#11.nr_switches
3067383 ± 2% +17.8% 3614738 ± 2% sched_debug.cpu#11.ttwu_count
2718845 ± 2% +13.2% 3076775 ± 1% sched_debug.cpu#11.ttwu_local
3058328 ± 2% +17.8% 3602585 ± 2% sched_debug.cpu#13.ttwu_count
3565846 ± 4% +25.6% 4478570 ± 3% sched_debug.cpu#13.sched_count
3541852 ± 4% +24.3% 4403611 ± 4% sched_debug.cpu#13.nr_switches
916506 ± 21% -45.4% 500000 ± 0% sched_debug.cpu#13.max_idle_balance_cost
20844 ± 13% +106.7% 43081 ± 21% sched_debug.cpu#13.sched_goidle
3533050 ± 2% +26.3% 4461776 ± 1% sched_debug.cpu#15.sched_count
23322 ± 20% +80.3% 42054 ± 17% sched_debug.cpu#15.sched_goidle
2673770 ± 2% +12.2% 3000835 ± 1% sched_debug.cpu#15.ttwu_local
3042506 ± 2% +16.0% 3530112 ± 1% sched_debug.cpu#15.ttwu_count
997258 ± 20% -49.9% 500000 ± 0% sched_debug.cpu#15.max_idle_balance_cost
3529886 ± 2% +23.5% 4360892 ± 1% sched_debug.cpu#15.nr_switches
14 ± 14% -24.1% 11 ± 11% sched_debug.cpu#15.nr_running
3586920 ± 3% +9.1% 3914639 ± 4% sched_debug.cpu#2.nr_switches
1111366 ± 23% -32.2% 753103 ± 14% sched_debug.cpu#3.avg_idle
2650771 ± 1% +9.8% 2909625 ± 2% sched_debug.cpu#3.ttwu_local
3498894 ± 1% +20.0% 4197900 ± 2% sched_debug.cpu#3.sched_count
3459439 ± 1% +20.8% 4177304 ± 3% sched_debug.cpu#3.nr_switches
929880 ± 23% -46.2% 500000 ± 0% sched_debug.cpu#3.max_idle_balance_cost
3043439 ± 1% +14.8% 3492383 ± 3% sched_debug.cpu#3.ttwu_count
21837 ± 15% +80.9% 39499 ± 16% sched_debug.cpu#3.sched_goidle
3174785 ± 2% +8.2% 3434347 ± 2% sched_debug.cpu#4.ttwu_count
3453169 ± 3% +23.9% 4277320 ± 3% sched_debug.cpu#5.nr_switches
981248 ± 26% -49.0% 500000 ± 0% sched_debug.cpu#5.max_idle_balance_cost
47 ± 2% +14.7% 54 ± 7% sched_debug.cpu#5.cpu_load[0]
3472968 ± 3% +23.6% 4291182 ± 3% sched_debug.cpu#5.sched_count
3065614 ± 2% +12.3% 3441364 ± 2% sched_debug.cpu#5.ttwu_count
2667436 ± 1% +9.8% 2930164 ± 3% sched_debug.cpu#5.ttwu_local
1242788 ± 18% -47.8% 648516 ± 19% sched_debug.cpu#5.avg_idle
20042 ± 13% +134.9% 47081 ± 14% sched_debug.cpu#5.sched_goidle
3654050 ± 3% +11.7% 4081164 ± 5% sched_debug.cpu#6.sched_count
3635987 ± 3% +12.1% 4076422 ± 5% sched_debug.cpu#6.nr_switches
21598 ± 6% +77.9% 38426 ± 19% sched_debug.cpu#7.sched_goidle
3477232 ± 4% +21.8% 4234219 ± 2% sched_debug.cpu#7.nr_switches
48 ± 10% +22.3% 59 ± 5% sched_debug.cpu#7.cpu_load[2]
48 ± 10% +22.3% 59 ± 8% sched_debug.cpu#7.cpu_load[1]
3048037 ± 2% +17.9% 3594543 ± 2% sched_debug.cpu#7.ttwu_count
2644522 ± 1% +11.4% 2946167 ± 1% sched_debug.cpu#7.ttwu_local
47 ± 9% +22.1% 58 ± 11% sched_debug.cpu#7.cpu_load[0]
3487732 ± 4% +25.4% 4373083 ± 3% sched_debug.cpu#7.sched_count
47 ± 7% +22.0% 58 ± 3% sched_debug.cpu#7.cpu_load[4]
925614 ± 7% -46.0% 500000 ± 0% sched_debug.cpu#7.max_idle_balance_cost
47 ± 9% +22.6% 58 ± 4% sched_debug.cpu#7.cpu_load[3]
3063652 ± 0% +12.6% 3450027 ± 3% sched_debug.cpu#9.ttwu_count
3476055 ± 4% +21.0% 4207345 ± 3% sched_debug.cpu#9.sched_count
21060 ± 2% -13.0% 18322 ± 2% sched_debug.cpu#9.curr->pid
16048 ± 12% +145.6% 39415 ± 13% sched_debug.cpu#9.sched_goidle
827315 ± 15% -39.6% 500000 ± 0% sched_debug.cpu#9.max_idle_balance_cost
2672155 ± 0% +12.1% 2994475 ± 2% sched_debug.cpu#9.ttwu_local
1206528 ± 23% -44.8% 666329 ± 11% sched_debug.cpu#9.avg_idle
3423880 ± 4% +22.8% 4206112 ± 3% sched_debug.cpu#9.nr_switches
lkp-ne04: Nehalem-EP
Memory: 12G
hackbench.time.involuntary_context_switches
2.4e+07 O+--O-----------O----------O--------------------------------------+
| O O O |
2.2e+07 ++ O |
| O |
2e+07 ++ |
| |
1.8e+07 ++ |
| *.. |
1.6e+07 ++ + . .*. |
| .*...*... + *.. .. .. .*.. |
1.4e+07 ++ *... .. *... + * .. .|
*. : * * * *
1.2e+07 ++.. : |
| ..*...*... : |
1e+07 ++--*-----------*-------------------------------------------------+
4000 ++-------------------------------------------------------------------+
O O O O O O O O O |
3500 ++ |
3000 ++ |
| |
2500 ++ |
| |
2000 ++ |
| |
1500 ++ |
1000 ++ *. |
| + .. |
500 ++ + |
| + *...|
0 *+--*---*---*---*---*---*---*---*----*---*---*---*---*---*-----------*
cpuidle.POLL.usage
600 ++--------------------------------------------------------------------+
| O |
550 ++ O |
500 ++ O |
| O |
450 ++ O |
400 O+ O O O |
| |
350 ++ |
300 ++ |
| |
250 ++ ..*.. |
200 *+. ..*...*.. .*. . .*... ..*...*...|
| . ..*...*.. . .. *...*... .. *. *
150 ++--*---*--------------------*-------------------*--------------------+
cpuidle.C1-NHM.usage
1.3e+06 ++----------------------------------------------------------------+
| O |
1.2e+06 ++ O O |
O O O O O |
1.1e+06 ++ O |
1e+06 ++ |
| |
900000 ++ |
| |
800000 ++ |
700000 *+..*...*... .*.. .*... .*...*...*... |
| *...*.. .. . .. *...*...*. *...*
600000 ++ * *...* |
| |
500000 ++----------------------------------------------------------------+
cpuidle.C1E-NHM.time
1.2e+08 ++----------------------------------------------------------------+
1.1e+08 ++ O |
O O O O O O O |
1e+08 ++ O |
9e+07 ++ |
8e+07 ++ |
7e+07 ++ |
| |
6e+07 ++ |
5e+07 ++ |
4e+07 ++ |
3e+07 ++ |
| |
2e+07 *+..*...*...*...*..*... ..*... ..*...*.. ..*...*...*...|
1e+07 ++---------------------*---*---*-------*----------*---------------*
cpuidle.C1E-NHM.usage
200000 ++------O----------------------------------------------------------+
| O O O |
180000 O+ O |
| O O |
160000 ++ O |
| |
140000 ++ |
| |
120000 ++ |
| |
100000 ++ ..*.. |
| ..*. . |
80000 ++..*...*. *...*... .*... ..*...*...*...*...*...*...*
*. *...*. *. |
60000 ++-----------------------------------------------------------------+
cpuidle.C3-NHM.usage
160000 ++-----------------------------------------------------------------+
*...*...*...*...*... ..*... ..*...*...*...*...|
140000 ++ *...*...*...*..*...*. *. *
120000 ++ |
| |
100000 ++ |
| |
80000 ++ |
| |
60000 ++ |
40000 ++ |
| |
20000 O+ O O O O O O O |
| O |
0 ++-----------------------------------------------------------------+
cpuidle.C6-NHM.time
1.5e+08 ++----------------------------------------------------------------+
1.4e+08 ++ * |
| : : |
1.3e+08 ++ : : |
1.2e+08 ++ : : |
| : : |
1.1e+08 ++ : : ..*.. .*.. |
1e+08 ++..* *. ..*...*... ..*... .. *...*...*...*...*
9e+07 *+ *...*. *. * |
| |
8e+07 ++ |
7e+07 ++ |
| |
6e+07 O+ O O O O O O O |
5e+07 ++---------------------O------------------------------------------+
cpuidle.C6-NHM.usage
36000 ++------------------------------------------------------------------+
| ..*...*... *... ..*.. .*
34000 *+..*...*. *.. .. *...*...*... ..*.. .*. . .. |
32000 ++ . . *. . .. * |
| * * |
30000 ++ |
| |
28000 ++ O |
O O O O O |
26000 ++ O |
24000 ++ O |
| |
22000 ++ O |
| |
20000 ++------------------------------------------------------------------+
turbostat.CPU%c1
2.4 ++--------------------------------------------------------------------+
| O |
2.3 O+ O O O O O O O |
2.2 ++ |
| |
2.1 ++ |
2 ++ |
| |
1.9 ++ |
1.8 ++ |
| |
1.7 ++..*...*...*... .*.. .*... ..*...|
1.6 *+ *....*... ..*... .. . .. *....*...*...*. *
| *. * * |
1.5 ++--------------------------------------------------------------------+
turbostat.CPU%c6
1.2 ++--------------------------------------------------------------------+
1.1 ++ * |
| : : |
1 ++ : : |
0.9 ++ : : |
| : : |
0.8 ++ : : .*.... ..*.. .*.... |
0.7 ++..* : .. *...*...*. . ..*... .. *...*...*...*...*
0.6 *+ * *. * |
| |
0.5 ++ |
0.4 ++ |
| O |
0.3 O+ O O O O O O |
0.2 ++-----------------------O--------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
To reproduce:
apt-get install ruby
git clone git://git.kernel.org/pub/scm/linux/kernel/git/wfg/lkp-tests.git
cd lkp-tests
bin/setup-local job.yaml # the job file attached in this email
bin/run-local job.yaml
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,
Ying Huang
-------------------------------------
lkp(a)eclists.intel.com
https://eclists.intel.com/sympa/info/lkp
Unsubscribe by sending email to sympa(a)eclists.intel.com with subject "Unsubscribe lkp"
[-- Attachment #2: job.yaml --]
[-- Type: text/plain, Size: 3100 bytes --]
---
LKP_SERVER: inn
testcase: hackbench
default-monitors:
wait: pre-test
uptime:
iostat:
vmstat:
numa-numastat:
numa-vmstat:
numa-meminfo:
proc-vmstat:
proc-stat:
interval: 10
meminfo:
slabinfo:
interrupts:
lock_stat:
latency_stats:
softirqs:
bdi_dev_mapping:
diskstats:
nfsstat:
cpuidle:
cpufreq-stats:
turbostat:
pmeter:
sched_debug:
interval: 60
default-watchdogs:
watch-oom:
watchdog:
cpufreq_governor: performance
model: Nehalem-EP
memory: 12G
hdd_partitions: "/dev/disk/by-id/ata-ST3500514NS_9WJ03EBA-part3"
swap_partitions: "/dev/disk/by-id/ata-ST3120026AS_5MS07HA2-part2"
rootfs_partition: "/dev/disk/by-id/ata-ST3500514NS_9WJ03EBA-part1"
nr_threads: 1600%
perf-profile:
freq: 800
hackbench:
mode: process
ipc: socket
branch: yuyang-tip/for-lkp
commit: 4c59e142fee20a57617cff2250bd1b1125f0efdf
repeat_to: 5
testbox: lkp-ne04
tbox_group: lkp-ne04
kconfig: x86_64-rhel
enqueue_time: 2015-06-01 18:48:12.328289425 +08:00
user: lkp
queue: ydu19
compiler: gcc-4.9
kernel: "/pkg/linux/x86_64-rhel/gcc-4.9/4c59e142fee20a57617cff2250bd1b1125f0efdf/vmlinuz-4.1.0-rc4-01964-g4c59e14"
rootfs: debian-x86_64-2015-02-07.cgz
result_root: "/result/hackbench/performance-1600%-process-socket/lkp-ne04/debian-x86_64-2015-02-07.cgz/x86_64-rhel/gcc-4.9/4c59e142fee20a57617cff2250bd1b1125f0efdf/0"
job_file: "/lkp/scheduled/lkp-ne04/ydu19_hackbench-performance-1600%-process-socket-x86_64-rhel-4c59e142fee20a57617cff2250bd1b1125f0efdf-0-20150601-47374-972w61.yaml"
dequeue_time: 2015-06-01 22:46:50.623747550 +08:00
nr_cpu: "$(nproc)"
max_uptime: 2400
initrd: "/osimage/debian/debian-x86_64-2015-02-07.cgz"
bootloader_append:
- root=/dev/ram0
- user=lkp
- job=/lkp/scheduled/lkp-ne04/ydu19_hackbench-performance-1600%-process-socket-x86_64-rhel-4c59e142fee20a57617cff2250bd1b1125f0efdf-0-20150601-47374-972w61.yaml
- ARCH=x86_64
- kconfig=x86_64-rhel
- branch=yuyang-tip/for-lkp
- commit=4c59e142fee20a57617cff2250bd1b1125f0efdf
- BOOT_IMAGE=/pkg/linux/x86_64-rhel/gcc-4.9/4c59e142fee20a57617cff2250bd1b1125f0efdf/vmlinuz-4.1.0-rc4-01964-g4c59e14
- max_uptime=2400
- RESULT_ROOT=/result/hackbench/performance-1600%-process-socket/lkp-ne04/debian-x86_64-2015-02-07.cgz/x86_64-rhel/gcc-4.9/4c59e142fee20a57617cff2250bd1b1125f0efdf/0
- LKP_SERVER=inn
- |2-
earlyprintk=ttyS0,115200 systemd.log_level=err
debug apic=debug sysrq_always_enabled rcupdate.rcu_cpu_stall_timeout=100
panic=-1 softlockup_panic=1 nmi_watchdog=panic oops=panic load_ramdisk=2 prompt_ramdisk=0
console=ttyS0,115200 console=tty0 vga=normal
rw
lkp_initrd: "/lkp/lkp/lkp-x86_64.cgz"
modules_initrd: "/pkg/linux/x86_64-rhel/gcc-4.9/4c59e142fee20a57617cff2250bd1b1125f0efdf/modules.cgz"
bm_initrd: "/osimage/deps/debian-x86_64-2015-02-07.cgz/lkp.cgz,/osimage/deps/debian-x86_64-2015-02-07.cgz/run-ipconfig.cgz,/osimage/deps/debian-x86_64-2015-02-07.cgz/turbostat.cgz,/lkp/benchmarks/turbostat.cgz"
job_state: finished
loadavg: 601.69 815.25 477.15 1/224 2727
start_time: '1433170051'
end_time: '1433170659'
version: "/lkp/lkp/.src-20150601-214511"
[-- Attachment #3: reproduce.ksh --]
[-- Type: text/plain, Size: 2142 bytes --]
echo performance > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu1/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu10/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu11/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu12/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu13/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu14/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu15/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu2/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu3/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu4/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu5/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu6/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu7/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu8/cpufreq/scaling_governor
echo performance > /sys/devices/system/cpu/cpu9/cpufreq/scaling_governor
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
/usr/bin/hackbench -g 256 --process -l 1875
reply other threads:[~2015-06-04 0:52 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=1433379157.7032.40.camel@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.