* [anolis-intel-cloud:devel-5.10] [sched/fair] eb34bf470a: hackbench.throughput 14.3% regression
@ 2025-03-31 14:32 kernel test robot
0 siblings, 0 replies; only message in thread
From: kernel test robot @ 2025-03-31 14:32 UTC (permalink / raw)
To: aubrey.li; +Cc: oe-lkp, lkp, oliver.sang
Hello,
kernel test robot noticed a 14.3% regression of hackbench.throughput on:
commit: eb34bf470aa84e3d7594928c150c56888b369e33 ("sched/fair: Bring back select_idle_smt(), but differently")
https://gitee.com/anolis/intel-cloud-kernel.git devel-5.10
testcase: hackbench
config: x86_64-anolis-intel-cloud-devel-5.10
compiler: gcc-12
test machine: 240 threads 1 sockets Genuine Intel(R) 0000 (Granite Rapids) with 192G memory
parameters:
nr_threads: 100%
iterations: 8
mode: process
ipc: pipe
cpufreq_governor: performance
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <oliver.sang@intel.com>
| Closes: https://lore.kernel.org/oe-lkp/202503312129.2a8ee063-lkp@intel.com
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250331/202503312129.2a8ee063-lkp@intel.com
=========================================================================================
compiler/cpufreq_governor/ipc/iterations/kconfig/mode/nr_threads/rootfs/tbox_group/testcase:
gcc-12/performance/pipe/8/x86_64-anolis-intel-cloud-devel-5.10/process/100%/debian-12-x86_64-20240206.cgz/lkp-gnr-1ap1/hackbench
commit:
80f5f5edac ("sched/fair: Optimize test_idle_cores() for !SMT")
eb34bf470a ("sched/fair: Bring back select_idle_smt(), but differently")
80f5f5edacb1cccc eb34bf470aa84e3d7594928c150
---------------- ---------------------------
%stddev %change %stddev
\ | \
4.55 ± 5% -0.6 3.92 ± 6% mpstat.cpu.all.idle%
65977799 +4.6% 69010684 vmstat.system.cs
1132850 ± 2% -18.2% 926883 vmstat.system.in
278.83 ± 6% -11.3% 247.38 ± 9% sched_debug.cpu.nr_uninterruptible.max
-295.17 -18.9% -239.29 sched_debug.cpu.nr_uninterruptible.min
101.50 ± 3% -19.2% 82.06 ± 4% sched_debug.cpu.nr_uninterruptible.stddev
1459169 +11.9% 1633494 ± 19% proc-vmstat.nr_file_pages
150994 +1.5% 153239 proc-vmstat.nr_kernel_stack
630767 ± 2% +27.6% 805092 ± 39% proc-vmstat.nr_shmem
1.099e+08 ± 2% -23.2% 84388885 ± 2% proc-vmstat.numa_hit
1.099e+08 ± 2% -23.2% 84388885 ± 2% proc-vmstat.numa_local
1.105e+08 ± 2% -23.1% 84933388 ± 2% proc-vmstat.pgalloc_normal
1.093e+08 ± 2% -23.6% 83588849 ± 2% proc-vmstat.pgfree
1339787 ± 3% -14.3% 1148604 hackbench.throughput
1178940 ± 2% -10.9% 1050422 ± 2% hackbench.throughput_avg
1339787 ± 3% -14.3% 1148604 hackbench.throughput_best
976391 ± 4% -12.1% 858353 ± 6% hackbench.throughput_worst
195.13 ± 2% +12.0% 218.48 hackbench.time.elapsed_time
195.13 ± 2% +12.0% 218.48 hackbench.time.elapsed_time.max
5.959e+09 ± 2% +19.3% 7.11e+09 ± 2% hackbench.time.involuntary_context_switches
40522 ± 2% +12.4% 45538 ± 2% hackbench.time.system_time
4462 +12.2% 5005 ± 2% hackbench.time.user_time
7.161e+09 +13.9% 8.155e+09 hackbench.time.voluntary_context_switches
39.26 ±217% +2252.7% 923.65 ± 97% perf-sched.sch_delay.avg.ms.rcu_gp_kthread.kthread.ret_from_fork
7752 ± 7% +13.0% 8762 ± 12% perf-sched.sch_delay.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe
194.31 ±220% +849.8% 1845 ± 84% perf-sched.sch_delay.max.ms.rcu_gp_kthread.kthread.ret_from_fork
7752 ± 7% +13.0% 8762 ± 12% perf-sched.total_sch_delay.max.ms
15485 ± 8% +12.8% 17468 ± 11% perf-sched.total_wait_and_delay.max.ms
7748 ± 8% +13.1% 8765 ± 11% perf-sched.total_wait_time.max.ms
15484 ± 8% +12.8% 17468 ± 11% perf-sched.wait_and_delay.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe
6659 ± 8% +18.8% 7908 ± 9% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork
0.15 ± 41% +2450.5% 3.94 ±134% perf-sched.wait_time.avg.ms.__cond_resched.mutex_lock.pipe_read.new_sync_read.vfs_read
5.02 ± 61% +5567.2% 284.51 ±102% perf-sched.wait_time.max.ms.__cond_resched.mutex_lock.pipe_read.new_sync_read.vfs_read
7742 ± 8% +13.1% 8752 ± 11% perf-sched.wait_time.max.ms.exit_to_user_mode_loop.exit_to_user_mode_prepare.syscall_exit_to_user_mode.entry_SYSCALL_64_after_hwframe
6659 ± 8% +18.8% 7908 ± 9% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork
0.03 ± 4% -14.6% 0.02 ± 4% perf-stat.i.MPKI
0.75 +0.0 0.76 perf-stat.i.branch-miss-rate%
7.963e+08 +2.5% 8.163e+08 perf-stat.i.branch-misses
0.38 ± 5% -0.0 0.34 ± 5% perf-stat.i.cache-miss-rate%
7820015 ± 3% -16.3% 6548408 ± 5% perf-stat.i.cache-misses
4.094e+09 +1.6% 4.158e+09 perf-stat.i.cache-references
66817468 +4.5% 69841496 perf-stat.i.context-switches
14015959 +18.0% 16545810 perf-stat.i.cpu-migrations
179705 ± 6% +34.8% 242161 ± 2% perf-stat.i.cycles-between-cache-misses
337.28 +6.9% 360.42 perf-stat.i.metric.K/sec
17632 ± 2% -11.7% 15578 ± 2% perf-stat.i.minor-faults
17691 ± 2% -11.6% 15634 ± 2% perf-stat.i.page-faults
0.58 ± 44% +0.1 0.72 perf-stat.overall.branch-miss-rate%
1.15 ± 44% +22.2% 1.40 perf-stat.overall.cpi
83187 ± 44% +43.7% 119539 ± 4% perf-stat.overall.cycles-between-cache-misses
6.626e+08 ± 44% +22.7% 8.132e+08 perf-stat.ps.branch-misses
3.392e+09 ± 44% +22.0% 4.138e+09 perf-stat.ps.cache-references
55705902 ± 44% +25.0% 69618145 perf-stat.ps.context-switches
11583377 ± 44% +41.7% 16416419 perf-stat.ps.cpu-migrations
9.264e+13 ± 44% +32.6% 1.229e+14 perf-stat.total.instructions
28.27 -1.0 27.28 perf-profile.calltrace.cycles-pp.new_sync_read.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
27.27 -0.9 26.41 perf-profile.calltrace.cycles-pp.pipe_read.new_sync_read.vfs_read.ksys_read.do_syscall_64
1.55 ± 4% -0.4 1.15 ± 17% perf-profile.calltrace.cycles-pp.copy_page_from_iter.pipe_write.new_sync_write.vfs_write.ksys_write
2.62 ± 3% -0.4 2.22 ± 10% perf-profile.calltrace.cycles-pp.copy_page_to_iter.pipe_read.new_sync_read.vfs_read.ksys_read
1.65 -0.1 1.51 ± 9% perf-profile.calltrace.cycles-pp.prepare_task_switch.__schedule.schedule.pipe_read.new_sync_read
0.53 ± 44% +0.5 1.03 ± 12% perf-profile.calltrace.cycles-pp.available_idle_cpu.id_idle_cpu.select_idle_sibling.select_task_rq_fair.try_to_wake_up
0.59 ± 44% +0.6 1.16 ± 12% perf-profile.calltrace.cycles-pp.id_idle_cpu.select_idle_sibling.select_task_rq_fair.try_to_wake_up.autoremove_wake_function
9.32 ± 3% +1.4 10.71 ± 6% perf-profile.calltrace.cycles-pp.select_idle_sibling.select_task_rq_fair.try_to_wake_up.autoremove_wake_function.__wake_up_common
11.63 ± 2% +1.7 13.31 ± 4% perf-profile.calltrace.cycles-pp.select_task_rq_fair.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_common_lock
35.66 +2.9 38.59 ± 3% perf-profile.calltrace.cycles-pp.new_sync_write.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
34.85 +3.1 37.97 ± 3% perf-profile.calltrace.cycles-pp.pipe_write.new_sync_write.vfs_write.ksys_write.do_syscall_64
24.62 ± 2% +3.5 28.11 ± 4% perf-profile.calltrace.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_common_lock.pipe_write.new_sync_write
24.48 ± 2% +3.5 27.97 ± 4% perf-profile.calltrace.cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_common_lock.pipe_write
25.53 ± 2% +3.6 29.14 ± 4% perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write
25.99 ± 2% +3.6 29.63 ± 4% perf-profile.calltrace.cycles-pp.__wake_up_common_lock.pipe_write.new_sync_write.vfs_write.ksys_write
32.49 -1.7 30.77 perf-profile.children.cycles-pp.ksys_read
31.32 -1.5 29.78 perf-profile.children.cycles-pp.vfs_read
28.35 -1.0 27.34 perf-profile.children.cycles-pp.new_sync_read
27.49 -0.9 26.60 perf-profile.children.cycles-pp.pipe_read
2.28 ± 9% -0.6 1.68 ± 16% perf-profile.children.cycles-pp.entry_SYSCALL_64
2.03 ± 6% -0.6 1.46 ± 13% perf-profile.children.cycles-pp.security_file_permission
1.43 ± 5% -0.4 1.02 ± 14% perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
1.64 ± 5% -0.4 1.26 ± 13% perf-profile.children.cycles-pp.copy_page_from_iter
2.72 ± 4% -0.4 2.36 ± 9% perf-profile.children.cycles-pp.copy_user_enhanced_fast_string
1.21 ± 6% -0.3 0.86 ± 14% perf-profile.children.cycles-pp.fsnotify
1.14 ± 6% -0.3 0.83 ± 12% perf-profile.children.cycles-pp.apparmor_file_permission
2.70 ± 3% -0.3 2.40 ± 5% perf-profile.children.cycles-pp.copy_page_to_iter
1.44 ± 2% -0.2 1.19 ± 7% perf-profile.children.cycles-pp.__fdget_pos
0.87 ± 5% -0.2 0.64 ± 13% perf-profile.children.cycles-pp.touch_atime
0.80 ± 6% -0.2 0.57 ± 13% perf-profile.children.cycles-pp.current_time
1.04 ± 4% -0.2 0.83 ± 14% perf-profile.children.cycles-pp.copyin
0.82 ± 4% -0.2 0.62 ± 12% perf-profile.children.cycles-pp.file_update_time
0.66 ± 6% -0.2 0.48 ± 13% perf-profile.children.cycles-pp.atime_needs_update
1.18 ± 2% -0.2 1.02 ± 6% perf-profile.children.cycles-pp.__fget_light
2.30 -0.2 2.13 ± 3% perf-profile.children.cycles-pp.__update_load_avg_cfs_rq
0.68 ± 4% -0.2 0.52 ± 10% perf-profile.children.cycles-pp.__cond_resched
0.51 ± 6% -0.1 0.37 ± 12% perf-profile.children.cycles-pp.aa_file_perm
0.50 ± 6% -0.1 0.40 ± 9% perf-profile.children.cycles-pp.syscall_enter_from_user_mode
0.51 ± 3% -0.1 0.42 ± 8% perf-profile.children.cycles-pp.syscall_return_via_sysret
0.34 ± 7% -0.1 0.25 ± 11% perf-profile.children.cycles-pp.ktime_get_coarse_real_ts64
0.24 ± 7% -0.1 0.16 ± 14% perf-profile.children.cycles-pp.rw_verify_area
0.23 ± 5% -0.1 0.16 ± 14% perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
0.27 ± 5% -0.1 0.21 ± 10% perf-profile.children.cycles-pp.rcu_all_qs
0.17 ± 6% -0.1 0.12 ± 16% perf-profile.children.cycles-pp.iov_iter_init
0.56 ± 2% -0.0 0.52 ± 3% perf-profile.children.cycles-pp.sched_clock_cpu
0.17 ± 6% -0.0 0.13 ± 13% perf-profile.children.cycles-pp.page_copy_sane
0.50 ± 2% -0.0 0.46 ± 3% perf-profile.children.cycles-pp.sched_clock
0.18 ± 5% -0.0 0.14 ± 11% perf-profile.children.cycles-pp.amd_clear_divider
0.22 ± 2% -0.0 0.18 ± 6% perf-profile.children.cycles-pp.rcu_read_unlock_strict
0.44 ± 3% -0.0 0.41 ± 3% perf-profile.children.cycles-pp.native_sched_clock
0.13 ± 4% -0.0 0.10 ± 10% perf-profile.children.cycles-pp.__x64_sys_read
0.15 ± 3% -0.0 0.12 ± 7% perf-profile.children.cycles-pp.__x64_sys_write
0.09 ± 5% -0.0 0.06 ± 7% perf-profile.children.cycles-pp.__bitmap_and
0.05 +0.0 0.07 ± 6% perf-profile.children.cycles-pp.group_balancer_enabled
0.16 ± 9% +0.0 0.20 ± 9% perf-profile.children.cycles-pp.attach_entity_load_avg
0.02 ± 99% +0.1 0.08 perf-profile.children.cycles-pp.cpus_share_cache
0.45 ± 6% +0.1 0.59 ± 12% perf-profile.children.cycles-pp.__cmd_record
0.45 ± 6% +0.1 0.59 ± 12% perf-profile.children.cycles-pp.cmd_record
0.29 ± 10% +0.1 0.44 ± 13% perf-profile.children.cycles-pp.remove_entity_load_avg
1.08 ± 4% +0.2 1.26 ± 6% perf-profile.children.cycles-pp.task_h_load
0.48 ± 6% +0.2 0.69 ± 6% perf-profile.children.cycles-pp._find_next_bit
1.78 ± 5% +0.2 2.02 ± 6% perf-profile.children.cycles-pp.wake_affine
0.94 ± 9% +0.3 1.20 ± 10% perf-profile.children.cycles-pp.migrate_task_rq_fair
0.00 +0.3 0.29 ± 6% perf-profile.children.cycles-pp.cpumask_next
1.67 ± 8% +0.4 2.12 ± 9% perf-profile.children.cycles-pp.set_task_cpu
4.40 ± 2% +0.6 5.00 ± 5% perf-profile.children.cycles-pp.available_idle_cpu
9.39 ± 3% +1.7 11.06 ± 6% perf-profile.children.cycles-pp.select_idle_sibling
11.70 ± 2% +1.9 13.63 ± 4% perf-profile.children.cycles-pp.select_task_rq_fair
38.66 +2.2 40.87 ± 2% perf-profile.children.cycles-pp.ksys_write
37.72 +2.4 40.12 ± 2% perf-profile.children.cycles-pp.vfs_write
26.41 ± 2% +2.7 29.16 ± 2% perf-profile.children.cycles-pp.autoremove_wake_function
26.39 ± 2% +2.8 29.16 ± 2% perf-profile.children.cycles-pp.try_to_wake_up
27.34 ± 2% +2.9 30.19 ± 2% perf-profile.children.cycles-pp.__wake_up_common
27.92 ± 2% +2.9 30.78 ± 2% perf-profile.children.cycles-pp.__wake_up_common_lock
35.89 +2.9 38.82 ± 3% perf-profile.children.cycles-pp.new_sync_write
35.00 +3.1 38.09 ± 3% perf-profile.children.cycles-pp.pipe_write
1.39 ± 5% -0.4 0.99 ± 14% perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
2.68 ± 4% -0.4 2.32 ± 9% perf-profile.self.cycles-pp.copy_user_enhanced_fast_string
1.14 ± 6% -0.3 0.81 ± 14% perf-profile.self.cycles-pp.fsnotify
0.95 ± 6% -0.3 0.68 ± 14% perf-profile.self.cycles-pp.entry_SYSCALL_64
1.15 ± 4% -0.2 0.95 ± 6% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
0.64 ± 5% -0.2 0.45 ± 14% perf-profile.self.cycles-pp.vfs_write
2.27 -0.2 2.10 ± 3% perf-profile.self.cycles-pp.__update_load_avg_cfs_rq
0.53 ± 6% -0.2 0.37 ± 15% perf-profile.self.cycles-pp.security_file_permission
0.59 ± 6% -0.2 0.44 ± 12% perf-profile.self.cycles-pp.apparmor_file_permission
0.63 ± 4% -0.1 0.48 ± 10% perf-profile.self.cycles-pp.new_sync_write
1.09 ± 2% -0.1 0.95 ± 5% perf-profile.self.cycles-pp.__fget_light
0.48 ± 6% -0.1 0.34 ± 14% perf-profile.self.cycles-pp.current_time
0.47 ± 6% -0.1 0.34 ± 13% perf-profile.self.cycles-pp.copy_page_from_iter
0.40 ± 6% -0.1 0.29 ± 13% perf-profile.self.cycles-pp.aa_file_perm
0.79 ± 2% -0.1 0.70 ± 4% perf-profile.self.cycles-pp.new_sync_read
0.51 ± 3% -0.1 0.42 ± 7% perf-profile.self.cycles-pp.syscall_return_via_sysret
0.44 ± 4% -0.1 0.35 ± 9% perf-profile.self.cycles-pp.do_syscall_64
0.34 ± 6% -0.1 0.25 ± 14% perf-profile.self.cycles-pp.file_update_time
0.35 ± 4% -0.1 0.26 ± 10% perf-profile.self.cycles-pp.__cond_resched
0.28 ± 6% -0.1 0.20 ± 13% perf-profile.self.cycles-pp.ksys_write
0.25 ± 6% -0.1 0.17 ± 15% perf-profile.self.cycles-pp.__fdget_pos
0.48 ± 3% -0.1 0.41 ± 6% perf-profile.self.cycles-pp.ksys_read
0.28 ± 5% -0.1 0.21 ± 11% perf-profile.self.cycles-pp.atime_needs_update
0.39 ± 6% -0.1 0.32 ± 8% perf-profile.self.cycles-pp.syscall_enter_from_user_mode
0.26 ± 7% -0.1 0.19 ± 11% perf-profile.self.cycles-pp.ktime_get_coarse_real_ts64
0.22 ± 5% -0.1 0.16 ± 12% perf-profile.self.cycles-pp.entry_SYSCALL_64_safe_stack
0.70 ± 2% -0.1 0.64 ± 4% perf-profile.self.cycles-pp.copy_page_to_iter
0.19 ± 7% -0.1 0.13 ± 16% perf-profile.self.cycles-pp.rw_verify_area
0.34 ± 3% -0.1 0.29 ± 7% perf-profile.self.cycles-pp.exit_to_user_mode_prepare
1.08 -0.1 1.03 ± 2% perf-profile.self.cycles-pp.dequeue_task_fair
0.58 ± 4% -0.0 0.53 ± 5% perf-profile.self.cycles-pp.syscall_exit_to_user_mode
0.19 ± 4% -0.0 0.15 ± 10% perf-profile.self.cycles-pp.rcu_all_qs
0.15 ± 5% -0.0 0.10 ± 13% perf-profile.self.cycles-pp.touch_atime
0.13 ± 8% -0.0 0.09 ± 14% perf-profile.self.cycles-pp.iov_iter_init
0.68 -0.0 0.64 ± 3% perf-profile.self.cycles-pp.dequeue_entity
0.10 ± 7% -0.0 0.08 ± 13% perf-profile.self.cycles-pp.page_copy_sane
0.08 ± 7% -0.0 0.06 ± 11% perf-profile.self.cycles-pp.copyin
0.08 ± 7% -0.0 0.06 ± 11% perf-profile.self.cycles-pp.copyout
0.12 ± 4% -0.0 0.10 ± 6% perf-profile.self.cycles-pp.rcu_read_unlock_strict
0.08 ± 4% -0.0 0.06 ± 6% perf-profile.self.cycles-pp.__bitmap_and
0.27 ± 6% +0.0 0.30 ± 7% perf-profile.self.cycles-pp.enqueue_task
0.16 ± 9% +0.0 0.20 ± 9% perf-profile.self.cycles-pp.attach_entity_load_avg
0.00 +0.1 0.05 perf-profile.self.cycles-pp.group_balancer_enabled
0.31 ± 4% +0.1 0.36 ± 5% perf-profile.self.cycles-pp.id_idle_cpu
0.00 +0.1 0.06 perf-profile.self.cycles-pp.cpus_share_cache
0.53 ± 5% +0.1 0.61 ± 6% perf-profile.self.cycles-pp.update_rq_clock
0.00 +0.1 0.08 ± 7% perf-profile.self.cycles-pp.cpumask_next
0.59 ± 8% +0.1 0.70 ± 8% perf-profile.self.cycles-pp.migrate_task_rq_fair
0.92 ± 4% +0.1 1.04 ± 5% perf-profile.self.cycles-pp.__wake_up_common
0.69 ± 8% +0.2 0.86 ± 8% perf-profile.self.cycles-pp.set_task_cpu
0.38 ± 7% +0.2 0.55 ± 6% perf-profile.self.cycles-pp._find_next_bit
1.07 ± 4% +0.2 1.24 ± 6% perf-profile.self.cycles-pp.task_h_load
2.22 ± 3% +0.2 2.44 ± 4% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
4.36 ± 2% +0.6 4.94 ± 5% perf-profile.self.cycles-pp.available_idle_cpu
0.86 ± 3% +0.9 1.74 ± 4% perf-profile.self.cycles-pp.select_idle_sibling
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.
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2025-03-31 14:32 UTC | newest]
Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-03-31 14:32 [anolis-intel-cloud:devel-5.10] [sched/fair] eb34bf470a: hackbench.throughput 14.3% regression kernel test robot
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.