| 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: <c0be63f1-d3f5-46fb-9633-cd860939b34d@suse.cz>
Date: Thu, 20 Nov 2025 14:27:09 +0100
From: Vlastimil Babka <vbabka@...e.cz>
To: kernel test robot <oliver.sang@...el.com>
Cc: oe-lkp@...ts.linux.dev, lkp@...el.com, linux-kernel@...r.kernel.org,
Harry Yoo <harry.yoo@...cle.com>, Suren Baghdasaryan <surenb@...gle.com>,
linux-mm@...ck.org
Subject: Re: [linus:master] [slab] ec66e0d599: stress-ng.mmapfiles.ops_per_sec
51.6% regression
On 11/20/25 09:47, kernel test robot wrote:
>
>
> Hello,
>
> kernel test robot noticed a 51.6% regression of stress-ng.mmapfiles.ops_per_sec on:
>
>
> commit: ec66e0d599520ab414db745544e25d80d0ac5054 ("slab: add sheaf support for batching kfree_rcu() operations")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
This is weird because at that commit, sheaves are not yet enabled for any
cache. So it just adds some cpu overhead (without lock contention or
anything) to kfree_rcu() but that shouldn't regress anything that much and
also wouldn't eplain some of the perf details below (why is there more drm
code visible in profile? etc). Puzzling.
> [still regression on linus/master 23cb64fb76257309e396ea4cec8396d4a1dbae68]
> [still regression on linux-next/master fe4d0dea039f2befb93f27569593ec209843b0f5]
>
> testcase: stress-ng
> config: x86_64-rhel-9.4
> compiler: gcc-14
> test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
> parameters:
>
> nr_threads: 100%
> testtime: 60s
> test: mmapfiles
> 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@...el.com>
> | Closes: https://lore.kernel.org/oe-lkp/202511201550.3efbe5c2-lkp@intel.com
>
>
> Details are as below:
> -------------------------------------------------------------------------------------------------->
>
>
> The kernel config and materials to reproduce are available at:
> https://download.01.org/0day-ci/archive/20251120/202511201550.3efbe5c2-lkp@intel.com
>
> =========================================================================================
> compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
> gcc-14/performance/x86_64-rhel-9.4/100%/debian-13-x86_64-20250902.cgz/lkp-icl-2sp7/mmapfiles/stress-ng/60s
>
> commit:
> 2d517aa09b ("slab: add opt-in caching layer of percpu sheaves")
> ec66e0d599 ("slab: add sheaf support for batching kfree_rcu() operations")
>
> 2d517aa09bbc4203 ec66e0d599520ab414db745544e
> ---------------- ---------------------------
> %stddev %change %stddev
> \ | \
> 5101 ± 3% +31.9% 6730 uptime.idle
> 1.945e+09 ± 8% +84.4% 3.587e+09 cpuidle..time
> 1066989 ± 4% -52.2% 509858 ± 2% cpuidle..usage
> 111039 ± 8% -23.1% 85351 meminfo.Mapped
> 117474 ± 9% -64.3% 41938 ± 5% meminfo.Shmem
> 20973 ± 52% -66.9% 6935 ± 51% numa-meminfo.node0.Shmem
> 96825 ± 14% -64.2% 34686 ± 12% numa-meminfo.node1.Shmem
> 805213 ± 6% -37.8% 500443 ± 8% numa-numastat.node0.local_node
> 846564 ± 6% -36.2% 539718 ± 5% numa-numastat.node0.numa_hit
> 926087 ± 5% -43.7% 521121 ± 8% numa-numastat.node1.local_node
> 950918 ± 5% -42.4% 548005 ± 5% numa-numastat.node1.numa_hit
> 17592 ± 14% -86.9% 2310 ± 62% perf-c2c.DRAM.remote
> 26251 ± 15% -97.9% 560.50 ± 42% perf-c2c.HITM.local
> 12343 ± 21% -98.3% 212.83 ± 33% perf-c2c.HITM.remote
> 38594 ± 13% -98.0% 773.33 ± 28% perf-c2c.HITM.total
> 49.06 ± 8% +38.2 87.22 mpstat.cpu.all.idle%
> 0.26 ± 4% -0.2 0.07 ± 2% mpstat.cpu.all.irq%
> 0.34 ± 5% -0.2 0.14 ± 2% mpstat.cpu.all.soft%
> 46.59 ± 8% -37.0 9.54 ± 2% mpstat.cpu.all.sys%
> 3.75 ± 3% -0.7 3.03 ± 2% mpstat.cpu.all.usr%
> 5301 ± 52% -67.5% 1725 ± 50% numa-vmstat.node0.nr_shmem
> 846980 ± 6% -36.3% 539228 ± 5% numa-vmstat.node0.numa_hit
> 805629 ± 6% -37.9% 499953 ± 8% numa-vmstat.node0.numa_local
> 24340 ± 14% -64.7% 8598 ± 13% numa-vmstat.node1.nr_shmem
> 950685 ± 5% -42.5% 546781 ± 5% numa-vmstat.node1.numa_hit
> 925853 ± 5% -43.8% 519898 ± 8% numa-vmstat.node1.numa_local
> 42.78 ± 20% +128.7% 97.82 ± 6% perf-sched.total_wait_and_delay.average.ms
> 27411 ± 17% -73.7% 7217 ± 7% perf-sched.total_wait_and_delay.count.ms
> 42.68 ± 20% +129.2% 97.80 ± 6% perf-sched.total_wait_time.average.ms
> 42.78 ± 20% +128.7% 97.82 ± 6% perf-sched.wait_and_delay.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
> 27411 ± 17% -73.7% 7217 ± 7% perf-sched.wait_and_delay.count.[unknown].[unknown].[unknown].[unknown].[unknown]
> 42.68 ± 20% +129.2% 97.80 ± 6% perf-sched.wait_time.avg.ms.[unknown].[unknown].[unknown].[unknown].[unknown]
> 948660 -2.0% 929444 proc-vmstat.nr_file_pages
> 28101 ± 8% -23.3% 21556 proc-vmstat.nr_mapped
> 29493 ± 9% -65.0% 10329 ± 4% proc-vmstat.nr_shmem
> 37019 ± 36% -66.8% 12299 ± 70% proc-vmstat.numa_hint_faults_local
> 1799984 ± 2% -39.5% 1088731 proc-vmstat.numa_hit
> 1733801 ± 2% -41.0% 1022572 proc-vmstat.numa_local
> 4027921 ± 2% -43.2% 2287803 proc-vmstat.pgalloc_normal
> 515117 ± 6% -14.7% 439270 ± 3% proc-vmstat.pgfault
> 3875358 ± 2% -43.8% 2177651 proc-vmstat.pgfree
> 162347 ± 2% +6.6% 173058 stress-ng.mmapfiles.file_mmaps_per_sec
> 227714 ± 2% +18.3% 269275 stress-ng.mmapfiles.file_munmap_per_sec
> 6730769 ± 2% +6.5% 7165751 stress-ng.mmapfiles.file_pages_mmap'd_per_sec
> 9440223 ± 2% +18.1% 11149909 stress-ng.mmapfiles.file_pages_munmap'd_per_sec
> 14347199 ± 4% -50.6% 7081102 ± 2% stress-ng.mmapfiles.ops
> 238147 ± 4% -51.6% 115229 ± 2% stress-ng.mmapfiles.ops_per_sec
> 40725 ± 3% -61.7% 15591 ± 15% stress-ng.time.involuntary_context_switches
> 130760 ± 5% -44.6% 72390 ± 9% stress-ng.time.minor_page_faults
> 3062 ± 8% -80.1% 609.00 ± 2% stress-ng.time.percent_of_cpu_this_job_got
> 1822 ± 8% -79.6% 371.22 ± 2% stress-ng.time.system_time
> 27.22 ± 7% -62.5% 10.20 ± 3% stress-ng.time.user_time
> 128468 ± 14% -92.6% 9544 ± 13% stress-ng.time.voluntary_context_switches
> 1582 ± 8% -74.7% 400.00 turbostat.Avg_MHz
> 51.16 ± 8% -38.2 12.94 turbostat.Busy%
> 1.84 ± 7% -1.5 0.32 ± 8% turbostat.C1E%
> 47.08 ± 8% +39.7 86.77 turbostat.C6%
> 17.56 ± 14% +35.8% 23.85 turbostat.CPU%c1
> 10.91 ± 22% +419.9% 56.73 turbostat.CPU%c6
> 66.50 -9.5% 60.17 ± 2% turbostat.CoreTmp
> 0.25 ± 8% +98.0% 0.50 turbostat.IPC
> 6165713 ± 6% -76.8% 1429713 ± 3% turbostat.IRQ
> 1607348 ± 8% -85.6% 231500 ± 8% turbostat.NMI
> 0.21 ± 64% +151.2% 0.52 ± 57% turbostat.Pkg%pc2
> 67.33 -8.9% 61.33 ± 2% turbostat.PkgTmp
> 193.47 -26.9% 141.49 turbostat.PkgWatt
> 14.64 -2.8% 14.23 turbostat.RAMWatt
> 692012 ± 11% -95.6% 30567 ±176% sched_debug.cfs_rq:/.avg_vruntime.avg
> 843998 ± 11% -87.6% 105070 ± 59% sched_debug.cfs_rq:/.avg_vruntime.max
> 586404 ± 12% -96.3% 21456 ±217% sched_debug.cfs_rq:/.avg_vruntime.min
> 50250 ± 9% -73.3% 13437 ± 22% sched_debug.cfs_rq:/.avg_vruntime.stddev
> 0.43 ± 26% -49.1% 0.22 ± 22% sched_debug.cfs_rq:/.h_nr_queued.avg
> 0.43 ± 26% -48.9% 0.22 ± 22% sched_debug.cfs_rq:/.h_nr_runnable.avg
> 191.51 ± 25% +49.4% 286.08 ± 21% sched_debug.cfs_rq:/.load_avg.stddev
> 692012 ± 11% -95.6% 30567 ±176% sched_debug.cfs_rq:/.min_vruntime.avg
> 843998 ± 11% -87.6% 105070 ± 59% sched_debug.cfs_rq:/.min_vruntime.max
> 586404 ± 12% -96.3% 21456 ±217% sched_debug.cfs_rq:/.min_vruntime.min
> 50250 ± 9% -73.3% 13437 ± 22% sched_debug.cfs_rq:/.min_vruntime.stddev
> 0.43 ± 25% -48.9% 0.22 ± 21% sched_debug.cfs_rq:/.nr_queued.avg
> 125.12 ± 14% +100.7% 251.12 ± 26% sched_debug.cfs_rq:/.removed.load_avg.stddev
> 140.75 ± 38% -63.2% 51.76 ± 18% sched_debug.cfs_rq:/.util_est.avg
> 1589174 ± 17% -31.7% 1085437 ± 24% sched_debug.cpu.avg_idle.avg
> 84100 -30.2% 58725 ± 19% sched_debug.cpu.clock.avg
> 84105 -30.2% 58729 ± 19% sched_debug.cpu.clock.max
> 84094 -30.2% 58716 ± 19% sched_debug.cpu.clock.min
> 83671 -30.1% 58446 ± 19% sched_debug.cpu.clock_task.avg
> 83965 -30.1% 58706 ± 19% sched_debug.cpu.clock_task.max
> 75446 -33.2% 50369 ± 22% sched_debug.cpu.clock_task.min
> 2514 ± 27% -57.3% 1074 ± 18% sched_debug.cpu.curr->pid.avg
> 1144485 -19.6% 919879 ± 8% sched_debug.cpu.max_idle_balance_cost.avg
> 915997 ± 8% -45.4% 500000 sched_debug.cpu.max_idle_balance_cost.min
> 180422 ± 9% +95.0% 351823 ± 5% sched_debug.cpu.max_idle_balance_cost.stddev
> 0.42 ± 25% -48.3% 0.22 ± 19% sched_debug.cpu.nr_running.avg
> 7294 ± 4% -63.9% 2636 ± 24% sched_debug.cpu.nr_switches.avg
> 16562 ± 7% -38.9% 10112 ± 13% sched_debug.cpu.nr_switches.max
> 3591 ± 13% -89.3% 383.00 ± 35% sched_debug.cpu.nr_switches.min
> 84094 -30.2% 58721 ± 19% sched_debug.cpu_clk
> 83382 -30.4% 58008 ± 19% sched_debug.ktime
> 85326 -29.9% 59776 ± 18% sched_debug.sched_clk
> 1.30 ± 3% +12.6% 1.46 perf-stat.i.MPKI
> 5.487e+09 -49.8% 2.754e+09 ± 2% perf-stat.i.branch-instructions
> 1.62 ± 4% +1.6 3.26 perf-stat.i.branch-miss-rate%
> 82380569 ± 3% -9.7% 74393638 ± 2% perf-stat.i.branch-misses
> 38.92 ± 2% +21.7 60.64 perf-stat.i.cache-miss-rate%
> 33723559 ± 2% -42.5% 19401136 ± 2% perf-stat.i.cache-misses
> 91887377 ± 7% -64.2% 32858366 ± 2% perf-stat.i.cache-references
> 10263 ± 6% -68.9% 3193 ± 5% perf-stat.i.context-switches
> 4.28 ± 9% -60.7% 1.68 ± 2% perf-stat.i.cpi
> 1.026e+11 ± 8% -74.6% 2.603e+10 perf-stat.i.cpu-cycles
> 271.97 ± 6% -55.8% 120.25 ± 3% perf-stat.i.cpu-migrations
> 3718 ± 11% -66.4% 1248 ± 4% perf-stat.i.cycles-between-cache-misses
> 2.512e+10 -48.8% 1.286e+10 ± 2% perf-stat.i.instructions
> 0.28 ± 7% +158.0% 0.72 perf-stat.i.ipc
> 6900 ± 9% -18.4% 5628 ± 3% perf-stat.i.minor-faults
> 6900 ± 9% -18.4% 5628 ± 3% perf-stat.i.page-faults
> 1.34 ± 3% +12.4% 1.51 perf-stat.overall.MPKI
> 1.50 ± 4% +1.2 2.70 perf-stat.overall.branch-miss-rate%
> 36.77 ± 5% +22.2 59.01 perf-stat.overall.cache-miss-rate%
> 4.09 ± 7% -50.5% 2.03 perf-stat.overall.cpi
> 3050 ± 8% -56.0% 1342 perf-stat.overall.cycles-between-cache-misses
> 0.25 ± 7% +100.8% 0.49 perf-stat.overall.ipc
> 5.39e+09 -49.7% 2.709e+09 ± 2% perf-stat.ps.branch-instructions
> 80929629 ± 3% -9.6% 73166120 ± 2% perf-stat.ps.branch-misses
> 33114569 ± 2% -42.4% 19077924 ± 2% perf-stat.ps.cache-misses
> 90432277 ± 7% -64.2% 32336218 ± 2% perf-stat.ps.cache-references
> 10086 ± 6% -68.8% 3143 ± 5% perf-stat.ps.context-switches
> 1.009e+11 ± 8% -74.6% 2.561e+10 perf-stat.ps.cpu-cycles
> 267.93 ± 6% -55.8% 118.42 ± 3% perf-stat.ps.cpu-migrations
> 2.468e+10 -48.7% 1.265e+10 ± 2% perf-stat.ps.instructions
> 6756 ± 9% -18.1% 5531 ± 3% perf-stat.ps.minor-faults
> 6756 ± 9% -18.1% 5531 ± 3% perf-stat.ps.page-faults
> 1.515e+12 -46.8% 8.056e+11 ± 2% perf-stat.total.instructions
> 58.02 ± 29% -49.4 8.66 ± 48% perf-profile.calltrace.cycles-pp.link_path_walk.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
> 70.23 ± 20% -48.0 22.23 ± 38% perf-profile.calltrace.cycles-pp.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64
> 70.28 ± 20% -47.9 22.34 ± 38% perf-profile.calltrace.cycles-pp.do_filp_open.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 61.36 ± 19% -40.7 20.67 ± 37% perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 61.38 ± 19% -40.7 20.71 ± 37% perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 64.54 ± 16% -30.3 34.20 ± 20% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 64.58 ± 16% -30.3 34.29 ± 20% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe
> 20.45 ± 64% -16.8 3.69 ± 55% perf-profile.calltrace.cycles-pp.inode_permission.link_path_walk.path_openat.do_filp_open.do_sys_openat2
> 18.05 ± 69% -16.1 1.92 ± 54% perf-profile.calltrace.cycles-pp.kernfs_iop_permission.inode_permission.link_path_walk.path_openat.do_filp_open
> 17.76 ± 67% -15.6 2.13 ± 42% perf-profile.calltrace.cycles-pp.lookup_fast.walk_component.link_path_walk.path_openat.do_filp_open
> 17.87 ± 66% -15.6 2.29 ± 42% perf-profile.calltrace.cycles-pp.walk_component.link_path_walk.path_openat.do_filp_open.do_sys_openat2
> 15.70 ± 80% -14.7 1.02 ± 58% perf-profile.calltrace.cycles-pp.step_into.link_path_walk.path_openat.do_filp_open.do_sys_openat2
> 14.93 ± 86% -14.5 0.38 ±103% perf-profile.calltrace.cycles-pp.__d_lookup.lookup_fast.walk_component.link_path_walk.path_openat
> 14.74 ± 87% -14.5 0.21 ±144% perf-profile.calltrace.cycles-pp._raw_spin_lock.__d_lookup.lookup_fast.walk_component.link_path_walk
> 14.77 ± 85% -14.3 0.42 ± 74% perf-profile.calltrace.cycles-pp.dput.step_into.link_path_walk.path_openat.do_filp_open
> 13.41 ± 96% -13.4 0.00 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.__d_lookup.lookup_fast.walk_component
> 13.02 ± 99% -13.0 0.00 perf-profile.calltrace.cycles-pp._raw_spin_lock.dput.step_into.link_path_walk.path_openat
> 12.96 ± 99% -13.0 0.00 perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock.dput.step_into.link_path_walk
> 9.36 ± 74% -8.8 0.55 ± 85% perf-profile.calltrace.cycles-pp.down_read.kernfs_iop_permission.inode_permission.link_path_walk.path_openat
> 7.71 ± 68% -7.2 0.53 ±109% perf-profile.calltrace.cycles-pp.up_read.kernfs_iop_permission.inode_permission.link_path_walk.path_openat
> 9.55 ± 23% -6.6 2.95 ± 41% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
> 9.55 ± 23% -6.6 2.95 ± 41% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__open64_nocancel
> 9.50 ± 23% -6.6 2.92 ± 42% perf-profile.calltrace.cycles-pp.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
> 9.49 ± 23% -6.6 2.91 ± 41% perf-profile.calltrace.cycles-pp.do_sys_openat2.__x64_sys_openat.do_syscall_64.entry_SYSCALL_64_after_hwframe.__open64_nocancel
> 9.60 ± 22% -6.5 3.05 ± 42% perf-profile.calltrace.cycles-pp.__open64_nocancel
> 1.91 ± 78% -1.5 0.39 ±101% perf-profile.calltrace.cycles-pp.down_read.kernfs_dop_revalidate.lookup_fast.walk_component.link_path_walk
> 0.38 ±108% +0.9 1.28 ± 29% perf-profile.calltrace.cycles-pp.kernfs_fop_open.do_dentry_open.vfs_open.do_open.path_openat
> 0.00 +1.0 0.99 ± 46% perf-profile.calltrace.cycles-pp.generic_fillattr.kernfs_iop_getattr.vfs_getattr_nosec.vfs_fstat.__do_sys_newfstat
> 0.24 ±150% +1.1 1.36 ± 33% perf-profile.calltrace.cycles-pp.intel_idle.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle
> 0.12 ±223% +1.2 1.37 ± 52% perf-profile.calltrace.cycles-pp.copy_folio_from_iter_atomic.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.cpu_startup_entry.rest_init.start_kernel.x86_64_start_reservations.x86_64_start_kernel
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.rest_init.start_kernel.x86_64_start_reservations
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.rest_init.start_kernel.x86_64_start_reservations.x86_64_start_kernel.common_startup_64
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.start_kernel.x86_64_start_reservations.x86_64_start_kernel.common_startup_64
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.x86_64_start_kernel.common_startup_64
> 0.00 +1.3 1.34 ± 52% perf-profile.calltrace.cycles-pp.x86_64_start_reservations.x86_64_start_kernel.common_startup_64
> 0.00 +1.4 1.38 ± 29% perf-profile.calltrace.cycles-pp.errseq_sample.do_dentry_open.vfs_open.do_open.path_openat
> 0.65 ±100% +1.5 2.19 ± 40% perf-profile.calltrace.cycles-pp.record__pushfn.perf_mmap__push.record__mmap_read_evlist.cmd_record.perf_c2c__record
> 0.61 ±100% +1.6 2.18 ± 40% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.record__pushfn.perf_mmap__push.record__mmap_read_evlist.cmd_record
> 0.60 ±100% +1.6 2.18 ± 40% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.record__pushfn.perf_mmap__push.record__mmap_read_evlist
> 0.58 ±101% +1.6 2.18 ± 40% perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.record__pushfn.perf_mmap__push
> 0.58 ± 83% +1.6 2.20 ± 44% perf-profile.calltrace.cycles-pp.kernfs_fop_readdir.iterate_dir.__x64_sys_getdents64.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 0.55 ±101% +1.6 2.17 ± 40% perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.record__pushfn
> 0.53 ±101% +1.6 2.16 ± 40% perf-profile.calltrace.cycles-pp.shmem_file_write_iter.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
> 0.00 +1.7 1.66 ± 47% perf-profile.calltrace.cycles-pp.console_flush_all.console_unlock.vprintk_emit._printk.irq_work_run_list
> 0.00 +1.7 1.66 ± 47% perf-profile.calltrace.cycles-pp.console_unlock.vprintk_emit._printk.irq_work_run_list.irq_work_run
> 0.50 ±102% +1.7 2.16 ± 40% perf-profile.calltrace.cycles-pp.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write.do_syscall_64
> 0.00 +1.7 1.69 ± 44% perf-profile.calltrace.cycles-pp._printk.irq_work_run_list.irq_work_run.__sysvec_irq_work.sysvec_irq_work
> 0.00 +1.7 1.69 ± 44% perf-profile.calltrace.cycles-pp.irq_work_run_list.irq_work_run.__sysvec_irq_work.sysvec_irq_work.asm_sysvec_irq_work
> 0.00 +1.7 1.69 ± 44% perf-profile.calltrace.cycles-pp.vprintk_emit._printk.irq_work_run_list.irq_work_run.__sysvec_irq_work
> 0.70 ± 86% +1.8 2.53 ± 34% perf-profile.calltrace.cycles-pp.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary
> 0.74 ± 85% +2.0 2.70 ± 34% perf-profile.calltrace.cycles-pp.cpuidle_idle_call.do_idle.cpu_startup_entry.start_secondary.common_startup_64
> 0.85 ± 83% +2.1 2.99 ± 32% perf-profile.calltrace.cycles-pp.do_idle.cpu_startup_entry.start_secondary.common_startup_64
> 0.85 ± 83% +2.1 3.00 ± 32% perf-profile.calltrace.cycles-pp.cpu_startup_entry.start_secondary.common_startup_64
> 0.85 ± 83% +2.1 3.00 ± 32% perf-profile.calltrace.cycles-pp.start_secondary.common_startup_64
> 0.69 ± 85% +2.2 2.91 ± 39% perf-profile.calltrace.cycles-pp.cpuidle_enter_state.cpuidle_enter.cpuidle_idle_call.do_idle.cpu_startup_entry
> 0.97 ± 83% +2.2 3.20 ± 42% perf-profile.calltrace.cycles-pp.handle_internal_command.main
> 0.97 ± 83% +2.2 3.20 ± 42% perf-profile.calltrace.cycles-pp.main
> 0.97 ± 83% +2.2 3.20 ± 42% perf-profile.calltrace.cycles-pp.run_builtin.handle_internal_command.main
> 0.32 ±103% +2.4 2.77 ± 41% perf-profile.calltrace.cycles-pp.__d_lookup.lookup_fast.open_last_lookups.path_openat.do_filp_open
> 1.10 ± 51% +3.3 4.36 ± 37% perf-profile.calltrace.cycles-pp.lookup_fast.open_last_lookups.path_openat.do_filp_open.do_sys_openat2
> 1.30 ± 36% +3.3 4.58 ± 38% perf-profile.calltrace.cycles-pp.open_last_lookups.path_openat.do_filp_open.do_sys_openat2.__x64_sys_openat
> 1.04 ± 68% +3.3 4.33 ± 37% perf-profile.calltrace.cycles-pp.common_startup_64
> 0.26 ±223% +11.6 11.89 ±118% perf-profile.calltrace.cycles-pp.ast_mode_config_helper_atomic_commit_tail.commit_tail.drm_atomic_helper_commit.drm_atomic_commit.drm_atomic_helper_dirtyfb
> 0.26 ±223% +11.6 11.89 ±118% perf-profile.calltrace.cycles-pp.drm_atomic_helper_commit_tail.ast_mode_config_helper_atomic_commit_tail.commit_tail.drm_atomic_helper_commit.drm_atomic_commit
> 0.26 ±223% +11.6 11.89 ±118% perf-profile.calltrace.cycles-pp.commit_tail.drm_atomic_helper_commit.drm_atomic_commit.drm_atomic_helper_dirtyfb.drm_fbdev_shmem_helper_fb_dirty
> 0.26 ±223% +11.6 11.89 ±118% perf-profile.calltrace.cycles-pp.drm_atomic_helper_commit.drm_atomic_commit.drm_atomic_helper_dirtyfb.drm_fbdev_shmem_helper_fb_dirty.drm_fb_helper_damage_work
> 0.26 ±223% +11.6 11.90 ±118% perf-profile.calltrace.cycles-pp.drm_atomic_commit.drm_atomic_helper_dirtyfb.drm_fbdev_shmem_helper_fb_dirty.drm_fb_helper_damage_work.process_one_work
> 0.26 ±223% +11.6 11.91 ±118% perf-profile.calltrace.cycles-pp.drm_atomic_helper_dirtyfb.drm_fbdev_shmem_helper_fb_dirty.drm_fb_helper_damage_work.process_one_work.worker_thread
> 0.26 ±223% +11.6 11.91 ±118% perf-profile.calltrace.cycles-pp.drm_fbdev_shmem_helper_fb_dirty.drm_fb_helper_damage_work.process_one_work.worker_thread.kthread
> 0.26 ±223% +11.7 11.91 ±118% perf-profile.calltrace.cycles-pp.drm_fb_helper_damage_work.process_one_work.worker_thread.kthread.ret_from_fork
> 0.47 ±128% +11.7 12.21 ±115% perf-profile.calltrace.cycles-pp.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
> 0.67 ± 93% +11.8 12.44 ±112% perf-profile.calltrace.cycles-pp.kthread.ret_from_fork.ret_from_fork_asm
> 0.67 ± 93% +11.8 12.44 ±112% perf-profile.calltrace.cycles-pp.ret_from_fork.ret_from_fork_asm
> 0.67 ± 93% +11.8 12.44 ±112% perf-profile.calltrace.cycles-pp.ret_from_fork_asm
> 0.27 ±223% +11.9 12.17 ±115% perf-profile.calltrace.cycles-pp.process_one_work.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
> 58.10 ± 29% -49.2 8.92 ± 45% perf-profile.children.cycles-pp.link_path_walk
> 70.25 ± 20% -48.0 22.27 ± 38% perf-profile.children.cycles-pp.path_openat
> 70.29 ± 20% -47.9 22.36 ± 38% perf-profile.children.cycles-pp.do_filp_open
> 70.86 ± 20% -47.3 23.61 ± 38% perf-profile.children.cycles-pp.do_sys_openat2
> 70.88 ± 20% -47.2 23.65 ± 38% perf-profile.children.cycles-pp.__x64_sys_openat
> 35.18 ± 65% -35.1 0.07 ±141% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
> 36.83 ± 63% -34.8 1.99 ± 37% perf-profile.children.cycles-pp._raw_spin_lock
> 95.60 -19.4 76.22 ± 15% perf-profile.children.cycles-pp.do_syscall_64
> 95.71 -19.2 76.47 ± 15% perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
> 20.38 ± 69% -17.8 2.57 ± 49% perf-profile.children.cycles-pp.kernfs_iop_permission
> 18.81 ± 62% -17.1 1.71 ± 44% perf-profile.children.cycles-pp.dput
> 17.89 ± 66% -15.3 2.55 ± 39% perf-profile.children.cycles-pp.walk_component
> 15.83 ± 79% -14.4 1.47 ± 45% perf-profile.children.cycles-pp.step_into
> 15.19 ± 70% -13.3 1.84 ± 55% perf-profile.children.cycles-pp.down_read
> 19.12 ± 60% -12.2 6.88 ± 35% perf-profile.children.cycles-pp.lookup_fast
> 15.62 ± 81% -11.8 3.87 ± 35% perf-profile.children.cycles-pp.__d_lookup
> 9.24 ± 66% -7.6 1.61 ± 49% perf-profile.children.cycles-pp.up_read
> 9.61 ± 22% -6.5 3.12 ± 42% perf-profile.children.cycles-pp.__open64_nocancel
> 3.56 ±131% -3.1 0.43 ± 67% perf-profile.children.cycles-pp.lockref_get_not_dead
> 3.12 ±132% -3.1 0.05 ± 73% perf-profile.children.cycles-pp.__mutex_lock
> 2.68 ± 33% -1.5 1.15 ± 58% perf-profile.children.cycles-pp.lockref_put_return
> 0.47 ± 51% -0.3 0.20 ± 73% perf-profile.children.cycles-pp.set_nlink
> 0.17 ± 44% -0.1 0.04 ± 72% perf-profile.children.cycles-pp.mutex_spin_on_owner
> 0.02 ±144% +0.1 0.09 ± 26% perf-profile.children.cycles-pp.schedule_idle
> 0.02 ±145% +0.1 0.11 ± 40% perf-profile.children.cycles-pp.__pte_offset_map_lock
> 0.05 ±117% +0.1 0.14 ± 34% perf-profile.children.cycles-pp.kfree_rcu_work
> 0.00 +0.1 0.10 ± 49% perf-profile.children.cycles-pp.__evlist__enable
> 0.00 +0.1 0.10 ± 30% perf-profile.children.cycles-pp.load_elf_binary
> 0.00 +0.1 0.10 ± 29% perf-profile.children.cycles-pp.exec_binprm
> 0.00 +0.1 0.11 ± 21% perf-profile.children.cycles-pp.shmem_add_to_page_cache
> 0.08 ± 23% +0.1 0.19 ± 26% perf-profile.children.cycles-pp.native_sched_clock
> 0.10 ± 30% +0.1 0.21 ± 29% perf-profile.children.cycles-pp.seq_open
> 0.01 ±223% +0.1 0.12 ± 36% perf-profile.children.cycles-pp.__kernfs_iattrs
> 0.11 ± 53% +0.1 0.23 ± 29% perf-profile.children.cycles-pp.build_detached_freelist
> 0.00 +0.1 0.12 ± 34% perf-profile.children.cycles-pp.bprm_execve
> 0.05 ±107% +0.1 0.17 ± 24% perf-profile.children.cycles-pp.sched_balance_newidle
> 0.01 ±223% +0.1 0.14 ± 31% perf-profile.children.cycles-pp.update_rq_clock_task
> 0.06 ± 55% +0.1 0.19 ± 48% perf-profile.children.cycles-pp.strcmp
> 0.00 +0.1 0.13 ± 54% perf-profile.children.cycles-pp.__x64_sys_exit_group
> 0.00 +0.1 0.13 ± 54% perf-profile.children.cycles-pp.do_exit
> 0.00 +0.1 0.13 ± 54% perf-profile.children.cycles-pp.do_group_exit
> 0.08 ± 68% +0.1 0.22 ± 20% perf-profile.children.cycles-pp.__pick_next_task
> 0.04 ± 75% +0.1 0.18 ± 39% perf-profile.children.cycles-pp.raw_spin_rq_lock_nested
> 0.03 ±103% +0.1 0.17 ± 32% perf-profile.children.cycles-pp.__rmqueue_pcplist
> 0.00 +0.1 0.14 ± 40% perf-profile.children.cycles-pp.read_counters
> 0.00 +0.1 0.14 ± 59% perf-profile.children.cycles-pp.evlist_cpu_iterator__next
> 0.11 ± 68% +0.1 0.25 ± 24% perf-profile.children.cycles-pp.schedule
> 0.02 ±223% +0.2 0.17 ± 46% perf-profile.children.cycles-pp.sysvec_call_function_single
> 0.05 ±110% +0.2 0.20 ± 29% perf-profile.children.cycles-pp.zap_pte_range
> 0.00 +0.2 0.16 ± 43% perf-profile.children.cycles-pp.process_interval
> 0.06 ± 55% +0.2 0.22 ± 55% perf-profile.children.cycles-pp.get_state_synchronize_rcu_full
> 0.00 +0.2 0.16 ± 41% perf-profile.children.cycles-pp.cmd_stat
> 0.00 +0.2 0.16 ± 41% perf-profile.children.cycles-pp.dispatch_events
> 0.08 ± 64% +0.2 0.24 ± 40% perf-profile.children.cycles-pp.sched_balance_update_blocked_averages
> 0.00 +0.2 0.16 ± 52% perf-profile.children.cycles-pp.idle_cpu
> 0.07 ± 72% +0.2 0.23 ± 37% perf-profile.children.cycles-pp.kernfs_refresh_inode
> 0.11 ± 38% +0.2 0.27 ± 35% perf-profile.children.cycles-pp.obj_cgroup_charge_account
> 0.02 ±223% +0.2 0.18 ± 46% perf-profile.children.cycles-pp.asm_sysvec_call_function_single
> 0.07 ± 85% +0.2 0.25 ± 30% perf-profile.children.cycles-pp.zap_pmd_range
> 0.00 +0.2 0.18 ± 39% perf-profile.children.cycles-pp.__x64_sys_execve
> 0.00 +0.2 0.18 ± 39% perf-profile.children.cycles-pp.do_execveat_common
> 0.00 +0.2 0.18 ± 39% perf-profile.children.cycles-pp.execve
> 0.06 ± 74% +0.2 0.26 ± 31% perf-profile.children.cycles-pp.strlen
> 0.09 ± 52% +0.2 0.30 ± 50% perf-profile.children.cycles-pp.x64_sys_call
> 0.15 ± 57% +0.2 0.37 ± 27% perf-profile.children.cycles-pp.__schedule
> 0.00 +0.2 0.22 ± 65% perf-profile.children.cycles-pp.sched_setaffinity
> 0.12 ± 67% +0.2 0.35 ± 27% perf-profile.children.cycles-pp.unmap_page_range
> 0.13 ± 41% +0.2 0.36 ± 48% perf-profile.children.cycles-pp.kvfree_call_rcu
> 0.17 ± 33% +0.2 0.41 ± 42% perf-profile.children.cycles-pp.cp_new_stat
> 0.03 ±103% +0.3 0.29 ± 43% perf-profile.children.cycles-pp.__handle_mm_fault
> 0.02 ±141% +0.3 0.28 ± 44% perf-profile.children.cycles-pp.vfs_read
> 0.25 ± 25% +0.3 0.52 ± 31% perf-profile.children.cycles-pp.native_flush_tlb_one_user
> 0.02 ±141% +0.3 0.30 ± 47% perf-profile.children.cycles-pp.ksys_read
> 0.10 ± 37% +0.3 0.38 ± 27% perf-profile.children.cycles-pp.shmem_alloc_and_add_folio
> 0.03 ±103% +0.3 0.31 ± 42% perf-profile.children.cycles-pp.handle_mm_fault
> 0.20 ± 48% +0.3 0.48 ± 23% perf-profile.children.cycles-pp.unmap_region
> 0.03 ±106% +0.3 0.33 ± 43% perf-profile.children.cycles-pp.do_user_addr_fault
> 0.03 ±106% +0.3 0.33 ± 43% perf-profile.children.cycles-pp.exc_page_fault
> 0.50 ± 26% +0.3 0.81 ± 22% perf-profile.children.cycles-pp.tick_nohz_handler
> 0.06 ± 74% +0.3 0.38 ± 46% perf-profile.children.cycles-pp.kernfs_dir_pos
> 0.04 ±109% +0.3 0.35 ± 41% perf-profile.children.cycles-pp.asm_exc_page_fault
> 0.19 ± 42% +0.3 0.52 ± 38% perf-profile.children.cycles-pp.filldir64
> 0.16 ± 35% +0.3 0.50 ± 31% perf-profile.children.cycles-pp.__kmalloc_cache_noprof
> 0.52 ± 26% +0.3 0.86 ± 21% perf-profile.children.cycles-pp.__hrtimer_run_queues
> 0.30 ± 36% +0.4 0.65 ± 36% perf-profile.children.cycles-pp.__memcg_slab_post_alloc_hook
> 0.07 ± 67% +0.4 0.43 ± 31% perf-profile.children.cycles-pp.update_sg_lb_stats
> 0.64 ± 25% +0.4 1.05 ± 24% perf-profile.children.cycles-pp.hrtimer_interrupt
> 0.64 ± 25% +0.4 1.06 ± 24% perf-profile.children.cycles-pp.__sysvec_apic_timer_interrupt
> 0.34 ± 66% +0.5 0.82 ± 29% perf-profile.children.cycles-pp.unmap_vmas
> 0.05 ± 76% +0.5 0.55 ± 88% perf-profile.children.cycles-pp.do_poll
> 0.07 ± 56% +0.6 0.62 ± 82% perf-profile.children.cycles-pp.__x64_sys_poll
> 0.07 ± 56% +0.6 0.62 ± 82% perf-profile.children.cycles-pp.do_sys_poll
> 0.08 ± 64% +0.6 0.64 ± 59% perf-profile.children.cycles-pp.update_sd_lb_stats
> 0.08 ± 67% +0.6 0.65 ± 59% perf-profile.children.cycles-pp.sched_balance_find_src_group
> 0.06 ± 90% +0.6 0.66 ± 64% perf-profile.children.cycles-pp.sched_balance_domains
> 0.02 ±223% +0.6 0.63 ±115% perf-profile.children.cycles-pp.bit_putcs
> 0.02 ±223% +0.6 0.64 ±114% perf-profile.children.cycles-pp.fbcon_putcs
> 0.11 ± 69% +0.6 0.76 ± 54% perf-profile.children.cycles-pp.sched_balance_rq
> 0.17 ± 46% +0.7 0.82 ± 42% perf-profile.children.cycles-pp.rb_next
> 0.08 ± 59% +0.7 0.74 ± 71% perf-profile.children.cycles-pp.do_softirq
> 0.02 ±223% +0.7 0.69 ±114% perf-profile.children.cycles-pp.fbcon_redraw
> 0.02 ±223% +0.7 0.72 ±114% perf-profile.children.cycles-pp.con_scroll
> 0.02 ±223% +0.7 0.72 ±114% perf-profile.children.cycles-pp.fbcon_scroll
> 0.02 ±223% +0.7 0.72 ±114% perf-profile.children.cycles-pp.lf
> 0.11 ± 59% +0.7 0.82 ± 66% perf-profile.children.cycles-pp.flush_smp_call_function_queue
> 0.02 ±223% +0.7 0.73 ±113% perf-profile.children.cycles-pp.vt_console_print
> 0.57 ± 48% +0.7 1.30 ± 28% perf-profile.children.cycles-pp.kernfs_fop_open
> 0.55 ± 28% +0.9 1.40 ± 37% perf-profile.children.cycles-pp.its_return_thunk
> 0.38 ± 57% +1.0 1.37 ± 52% perf-profile.children.cycles-pp.copy_folio_from_iter_atomic
> 0.12 ± 61% +1.0 1.12 ± 50% perf-profile.children.cycles-pp._nohz_idle_balance
> 0.43 ± 66% +1.1 1.49 ± 33% perf-profile.children.cycles-pp.intel_idle
> 0.11 ± 86% +1.2 1.34 ± 52% perf-profile.children.cycles-pp.rest_init
> 0.11 ± 86% +1.2 1.34 ± 52% perf-profile.children.cycles-pp.start_kernel
> 0.11 ± 86% +1.2 1.34 ± 52% perf-profile.children.cycles-pp.x86_64_start_kernel
> 0.11 ± 86% +1.2 1.34 ± 52% perf-profile.children.cycles-pp.x86_64_start_reservations
> 0.87 ± 50% +1.3 2.17 ± 33% perf-profile.children.cycles-pp.mas_empty_area_rev
> 0.76 ± 57% +1.4 2.17 ± 40% perf-profile.children.cycles-pp.shmem_file_write_iter
> 0.19 ± 63% +1.4 1.62 ± 30% perf-profile.children.cycles-pp.errseq_sample
> 0.72 ± 56% +1.4 2.16 ± 40% perf-profile.children.cycles-pp.generic_perform_write
> 0.68 ± 56% +1.5 2.21 ± 44% perf-profile.children.cycles-pp.kernfs_fop_readdir
> 0.37 ± 61% +1.6 2.00 ± 38% perf-profile.children.cycles-pp.generic_fillattr
> 0.18 ± 35% +1.6 1.82 ± 34% perf-profile.children.cycles-pp._printk
> 0.19 ± 33% +1.6 1.84 ± 33% perf-profile.children.cycles-pp.__sysvec_irq_work
> 0.19 ± 33% +1.6 1.84 ± 33% perf-profile.children.cycles-pp.irq_work_run
> 0.19 ± 33% +1.7 1.84 ± 33% perf-profile.children.cycles-pp.irq_work_run_list
> 0.19 ± 32% +1.7 1.84 ± 33% perf-profile.children.cycles-pp.asm_sysvec_irq_work
> 0.19 ± 33% +1.7 1.84 ± 33% perf-profile.children.cycles-pp.sysvec_irq_work
> 1.50 ± 36% +1.8 3.30 ± 26% perf-profile.children.cycles-pp.asm_sysvec_apic_timer_interrupt
> 0.96 ± 61% +2.0 3.00 ± 32% perf-profile.children.cycles-pp.start_secondary
> 1.10 ± 60% +2.1 3.20 ± 42% perf-profile.children.cycles-pp.handle_internal_command
> 1.10 ± 60% +2.1 3.20 ± 42% perf-profile.children.cycles-pp.main
> 1.10 ± 60% +2.1 3.20 ± 42% perf-profile.children.cycles-pp.run_builtin
> 0.86 ± 64% +2.2 3.09 ± 33% perf-profile.children.cycles-pp.cpuidle_enter_state
> 0.86 ± 64% +2.2 3.10 ± 33% perf-profile.children.cycles-pp.cpuidle_enter
> 0.91 ± 62% +2.4 3.32 ± 33% perf-profile.children.cycles-pp.cpuidle_idle_call
> 1.49 ± 36% +3.2 4.70 ± 34% perf-profile.children.cycles-pp.open_last_lookups
> 1.08 ± 61% +3.3 4.33 ± 37% perf-profile.children.cycles-pp.do_idle
> 1.08 ± 61% +3.3 4.33 ± 37% perf-profile.children.cycles-pp.common_startup_64
> 1.08 ± 61% +3.3 4.33 ± 37% perf-profile.children.cycles-pp.cpu_startup_entry
> 0.12 ±137% +3.7 3.78 ±128% perf-profile.children.cycles-pp.delay_tsc
> 0.29 ± 87% +6.2 6.45 ±104% perf-profile.children.cycles-pp.io_serial_in
> 0.42 ± 96% +9.9 10.30 ±112% perf-profile.children.cycles-pp.wait_for_lsr
> 0.46 ± 99% +11.1 11.60 ±114% perf-profile.children.cycles-pp.serial8250_console_write
> 0.42 ±118% +11.4 11.81 ±118% perf-profile.children.cycles-pp.memcpy_toio
> 0.42 ±119% +11.4 11.86 ±118% perf-profile.children.cycles-pp.ast_primary_plane_helper_atomic_update
> 0.42 ±119% +11.4 11.86 ±118% perf-profile.children.cycles-pp.drm_fb_memcpy
> 0.42 ±119% +11.5 11.89 ±118% perf-profile.children.cycles-pp.ast_mode_config_helper_atomic_commit_tail
> 0.42 ±119% +11.5 11.89 ±118% perf-profile.children.cycles-pp.drm_atomic_helper_commit_tail
> 0.42 ±118% +11.5 11.89 ±118% perf-profile.children.cycles-pp.drm_atomic_helper_commit
> 0.42 ±119% +11.5 11.89 ±118% perf-profile.children.cycles-pp.commit_tail
> 0.42 ±118% +11.5 11.90 ±118% perf-profile.children.cycles-pp.drm_atomic_commit
> 0.42 ±119% +11.5 11.91 ±118% perf-profile.children.cycles-pp.drm_atomic_helper_dirtyfb
> 0.42 ±119% +11.5 11.91 ±118% perf-profile.children.cycles-pp.drm_fbdev_shmem_helper_fb_dirty
> 0.42 ±119% +11.5 11.91 ±118% perf-profile.children.cycles-pp.drm_fb_helper_damage_work
> 0.59 ± 86% +11.6 12.21 ±115% perf-profile.children.cycles-pp.worker_thread
> 0.54 ± 92% +11.6 12.17 ±115% perf-profile.children.cycles-pp.process_one_work
> 0.79 ± 64% +11.6 12.44 ±112% perf-profile.children.cycles-pp.kthread
> 0.79 ± 64% +11.6 12.44 ±112% perf-profile.children.cycles-pp.ret_from_fork
> 0.79 ± 64% +11.6 12.44 ±112% perf-profile.children.cycles-pp.ret_from_fork_asm
> 1.14 ± 71% +11.7 12.83 ±109% perf-profile.children.cycles-pp.ksys_write
> 1.10 ± 71% +11.7 12.83 ±109% perf-profile.children.cycles-pp.vfs_write
> 0.49 ±100% +11.9 12.35 ±114% perf-profile.children.cycles-pp.console_flush_all
> 0.49 ±100% +11.9 12.35 ±114% perf-profile.children.cycles-pp.console_unlock
> 0.50 ± 98% +11.9 12.45 ±114% perf-profile.children.cycles-pp.vprintk_emit
> 34.91 ± 66% -34.8 0.06 ±141% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
> 14.53 ± 70% -13.0 1.57 ± 59% perf-profile.self.cycles-pp.down_read
> 9.15 ± 66% -7.6 1.54 ± 51% perf-profile.self.cycles-pp.up_read
> 2.59 ± 33% -1.5 1.11 ± 60% perf-profile.self.cycles-pp.lockref_put_return
> 1.04 ± 27% -0.6 0.41 ± 70% perf-profile.self.cycles-pp.lockref_get_not_dead
> 0.16 ± 44% -0.1 0.04 ± 72% perf-profile.self.cycles-pp.mutex_spin_on_owner
> 0.02 ±142% +0.1 0.11 ± 27% perf-profile.self.cycles-pp.obj_cgroup_charge_account
> 0.01 ±223% +0.1 0.10 ± 38% perf-profile.self.cycles-pp.__kernfs_iattrs
> 0.00 +0.1 0.10 ± 43% perf-profile.self.cycles-pp._nohz_idle_balance
> 0.07 ± 21% +0.1 0.17 ± 10% perf-profile.self.cycles-pp.dput
> 0.05 ± 85% +0.1 0.15 ± 21% perf-profile.self.cycles-pp.kernfs_iop_permission
> 0.08 ± 23% +0.1 0.19 ± 25% perf-profile.self.cycles-pp.native_sched_clock
> 0.06 ± 63% +0.1 0.17 ± 48% perf-profile.self.cycles-pp.strcmp
> 0.10 ± 51% +0.1 0.22 ± 30% perf-profile.self.cycles-pp.build_detached_freelist
> 0.05 ± 74% +0.1 0.17 ± 40% perf-profile.self.cycles-pp.__kmalloc_cache_noprof
> 0.06 ± 74% +0.1 0.20 ± 50% perf-profile.self.cycles-pp._copy_to_user
> 0.00 +0.1 0.14 ± 49% perf-profile.self.cycles-pp.idle_cpu
> 0.01 ±223% +0.1 0.16 ± 50% perf-profile.self.cycles-pp.d_same_name
> 0.05 ± 76% +0.2 0.21 ± 53% perf-profile.self.cycles-pp.get_state_synchronize_rcu_full
> 0.10 ± 23% +0.2 0.27 ± 31% perf-profile.self.cycles-pp.rcu_all_qs
> 0.06 ± 75% +0.2 0.25 ± 35% perf-profile.self.cycles-pp.strlen
> 0.04 ±104% +0.2 0.24 ± 42% perf-profile.self.cycles-pp.update_sg_lb_stats
> 0.08 ± 53% +0.2 0.31 ± 30% perf-profile.self.cycles-pp.mas_empty_area_rev
> 0.05 ±103% +0.2 0.30 ± 38% perf-profile.self.cycles-pp.shmem_write_end
> 0.14 ± 26% +0.3 0.40 ± 42% perf-profile.self.cycles-pp.step_into
> 0.14 ± 38% +0.3 0.40 ± 32% perf-profile.self.cycles-pp.filldir64
> 0.09 ± 47% +0.3 0.36 ± 30% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
> 0.25 ± 25% +0.3 0.52 ± 31% perf-profile.self.cycles-pp.native_flush_tlb_one_user
> 0.06 ± 74% +0.3 0.37 ± 46% perf-profile.self.cycles-pp.kernfs_dir_pos
> 0.16 ± 57% +0.4 0.56 ± 50% perf-profile.self.cycles-pp.kernfs_fop_readdir
> 0.14 ± 65% +0.5 0.62 ± 28% perf-profile.self.cycles-pp.kernfs_dop_revalidate
> 0.16 ± 44% +0.6 0.81 ± 43% perf-profile.self.cycles-pp.rb_next
> 0.36 ± 57% +0.7 1.02 ± 22% perf-profile.self.cycles-pp.copy_folio_from_iter_atomic
> 0.15 ± 38% +0.7 0.82 ± 43% perf-profile.self.cycles-pp.may_open
> 0.33 ± 37% +0.9 1.24 ± 36% perf-profile.self.cycles-pp.do_dentry_open
> 0.43 ± 66% +1.1 1.49 ± 33% perf-profile.self.cycles-pp.intel_idle
> 0.19 ± 62% +1.4 1.58 ± 30% perf-profile.self.cycles-pp.errseq_sample
> 0.32 ± 65% +1.6 1.92 ± 37% perf-profile.self.cycles-pp.generic_fillattr
> 0.40 ± 31% +2.4 2.82 ± 34% perf-profile.self.cycles-pp.__d_lookup
> 0.12 ±137% +3.7 3.78 ±128% perf-profile.self.cycles-pp.delay_tsc
> 0.29 ± 87% +6.2 6.45 ±104% perf-profile.self.cycles-pp.io_serial_in
> 0.39 ±118% +10.8 11.20 ±120% perf-profile.self.cycles-pp.memcpy_toio
>
>
>
>
> 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.
>
>
Powered by blists - more mailing lists