* [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
@ 2025-03-10 5:50 kernel test robot
2025-03-10 9:55 ` Vlastimil Babka
0 siblings, 1 reply; 6+ messages in thread
From: kernel test robot @ 2025-03-10 5:50 UTC (permalink / raw)
To: Alexei Starovoitov
Cc: oe-lkp, lkp, Michal Hocko, Vlastimil Babka, Shakeel Butt,
cgroups, linux-mm, oliver.sang
Hello,
kernel test robot noticed a 37.9% regression of netperf.Throughput_Mbps on:
commit: 01d37228d331047a0bbbd1026cec2ccabef6d88d ("memcg: Use trylock to access memcg stock_lock.")
https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
[test failed on linux-next/master 7ec162622e66a4ff886f8f28712ea1b13069e1aa]
testcase: netperf
config: x86_64-rhel-9.4
compiler: gcc-12
test machine: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
parameters:
ip: ipv4
runtime: 300s
nr_threads: 50%
cluster: cs-localhost
test: TCP_MAERTS
cpufreq_governor: performance
In addition to that, the commit also has significant impact on the following tests:
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | stress-ng: stress-ng.mmapfork.ops_per_sec 63.5% regression |
| test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
| test parameters | cpufreq_governor=performance |
| | nr_threads=100% |
| | test=mmapfork |
| | testtime=60s |
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | hackbench: hackbench.throughput 26.6% regression |
| test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
| test parameters | cpufreq_governor=performance |
| | ipc=socket |
| | iterations=4 |
| | mode=threads |
| | nr_threads=100% |
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | lmbench3: lmbench3.TCP.socket.bandwidth.64B.MB/sec 33.0% regression |
| test machine | 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 512G memory |
| test parameters | cpufreq_governor=performance |
| | mode=development |
| | nr_threads=100% |
| | test=TCP |
| | test_memory_size=50% |
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | vm-scalability: vm-scalability.throughput 86.8% regression |
| test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
| test parameters | cpufreq_governor=performance |
| | runtime=300s |
| | size=1T |
| | test=lru-shm |
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | netperf: netperf.Throughput_Mbps 39.9% improvement |
| test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
| test parameters | cluster=cs-localhost |
| | cpufreq_governor=performance |
| | ip=ipv4 |
| | nr_threads=200% |
| | runtime=300s |
| | test=TCP_MAERTS |
+------------------+----------------------------------------------------------------------------------------------------+
| testcase: change | will-it-scale: will-it-scale.per_thread_ops 68.8% regression |
| test machine | 104 threads 2 sockets (Skylake) with 192G memory |
| test parameters | cpufreq_governor=performance |
| | mode=thread |
| | nr_task=100% |
| | test=fallocate1 |
+------------------+----------------------------------------------------------------------------------------------------+
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/202503101254.cfd454df-lkp@intel.com
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250310/202503101254.cfd454df-lkp@intel.com
=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase:
cs-localhost/gcc-12/performance/ipv4/x86_64-rhel-9.4/50%/debian-12-x86_64-20240206.cgz/300s/lkp-icl-2sp2/TCP_MAERTS/netperf
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
88798 ± 2% +11.3% 98788 perf-c2c.HITM.total
11324 ± 9% +29.0% 14612 uptime.idle
5.698e+09 +62.0% 9.228e+09 cpuidle..time
6.409e+08 ± 2% -13.9% 5.517e+08 ± 4% cpuidle..usage
12.79 ± 2% +10.0 22.80 mpstat.cpu.all.idle%
2.92 ± 2% -0.4 2.55 mpstat.cpu.all.irq%
68.81 -8.5 60.34 mpstat.cpu.all.sys%
2.75 -1.1 1.61 ± 2% mpstat.cpu.all.usr%
8.542e+08 -36.5% 5.424e+08 numa-numastat.node0.local_node
8.541e+08 -36.5% 5.425e+08 numa-numastat.node0.numa_hit
8.262e+08 -39.5% 4.995e+08 ± 3% numa-numastat.node1.local_node
8.262e+08 -39.5% 4.996e+08 ± 3% numa-numastat.node1.numa_hit
13.41 ± 2% +73.9% 23.32 vmstat.cpu.id
110.55 -13.7% 95.45 vmstat.procs.r
4461013 ± 2% -12.9% 3883497 ± 4% vmstat.system.cs
2363200 ± 2% -12.3% 2073470 ± 4% vmstat.system.in
6829101 ± 4% -59.5% 2765741 ± 7% numa-meminfo.node1.Active
6829101 ± 4% -59.5% 2765741 ± 7% numa-meminfo.node1.Active(anon)
7426985 ± 22% -32.0% 5051150 ± 27% numa-meminfo.node1.FilePages
2764830 ± 11% -92.4% 209706 ± 21% numa-meminfo.node1.Mapped
8991931 ± 18% -29.4% 6351136 ± 20% numa-meminfo.node1.MemUsed
14170 ± 8% -41.4% 8302 ± 5% numa-meminfo.node1.PageTables
6214806 ± 3% -63.5% 2266447 ± 7% numa-meminfo.node1.Shmem
7077695 ± 2% -57.6% 2999270 ± 5% meminfo.Active
7077695 ± 2% -57.6% 2999270 ± 5% meminfo.Active(anon)
9791069 -41.0% 5777962 ± 2% meminfo.Cached
7238271 ± 2% -56.5% 3151650 ± 5% meminfo.Committed_AS
2812962 ± 11% -92.1% 223137 ± 10% meminfo.Mapped
12548272 -35.9% 8045050 ± 2% meminfo.Memused
22784 ± 3% -27.4% 16539 meminfo.PageTables
6286416 ± 2% -63.8% 2273611 ± 7% meminfo.Shmem
12766197 -32.4% 8626151 ± 2% meminfo.max_used_kB
8.541e+08 -36.5% 5.427e+08 numa-vmstat.node0.numa_hit
8.542e+08 -36.5% 5.425e+08 numa-vmstat.node0.numa_local
1707145 ± 4% -59.5% 691093 ± 7% numa-vmstat.node1.nr_active_anon
1856614 ± 22% -32.0% 1262417 ± 27% numa-vmstat.node1.nr_file_pages
691174 ± 11% -92.3% 52918 ± 21% numa-vmstat.node1.nr_mapped
3544 ± 8% -41.3% 2080 ± 5% numa-vmstat.node1.nr_page_table_pages
1553569 ± 3% -63.6% 566242 ± 7% numa-vmstat.node1.nr_shmem
1707145 ± 4% -59.5% 691093 ± 7% numa-vmstat.node1.nr_zone_active_anon
8.262e+08 -39.5% 4.997e+08 ± 3% numa-vmstat.node1.numa_hit
8.262e+08 -39.5% 4.997e+08 ± 3% numa-vmstat.node1.numa_local
22880 -37.9% 14205 ± 2% netperf.ThroughputBoth_Mbps
1464367 -37.9% 909168 ± 2% netperf.ThroughputBoth_total_Mbps
22880 -37.9% 14205 ± 2% netperf.Throughput_Mbps
1464367 -37.9% 909168 ± 2% netperf.Throughput_total_Mbps
94030 ± 15% +799.5% 845847 ± 16% netperf.time.involuntary_context_switches
35098 +11.3% 39072 ± 3% netperf.time.minor_page_faults
3619 -30.6% 2511 netperf.time.percent_of_cpu_this_job_got
10591 -31.1% 7296 netperf.time.system_time
307.43 -12.8% 268.12 netperf.time.user_time
6.797e+08 ± 2% -12.9% 5.922e+08 ± 4% netperf.time.voluntary_context_switches
3.352e+09 -37.9% 2.081e+09 ± 2% netperf.workload
1768827 ± 2% -57.6% 749641 ± 5% proc-vmstat.nr_active_anon
198757 -8.2% 182368 proc-vmstat.nr_anon_pages
6242276 +1.8% 6354594 proc-vmstat.nr_dirty_background_threshold
12499816 +1.8% 12724725 proc-vmstat.nr_dirty_threshold
2447152 -41.0% 1444280 ± 2% proc-vmstat.nr_file_pages
62798979 +1.8% 63923764 proc-vmstat.nr_free_pages
703005 ± 11% -92.0% 56220 ± 12% proc-vmstat.nr_mapped
5711 ± 3% -27.3% 4153 proc-vmstat.nr_page_table_pages
1570988 ± 2% -63.8% 568192 ± 7% proc-vmstat.nr_shmem
33010 -7.1% 30660 proc-vmstat.nr_slab_reclaimable
70932 -3.7% 68338 proc-vmstat.nr_slab_unreclaimable
1768827 ± 2% -57.6% 749641 ± 5% proc-vmstat.nr_zone_active_anon
351363 ± 32% -79.4% 72278 ± 16% proc-vmstat.numa_hint_faults
337005 ± 34% -82.9% 57525 ± 24% proc-vmstat.numa_hint_faults_local
1.679e+09 -37.9% 1.042e+09 ± 2% proc-vmstat.numa_hit
1.679e+09 -37.9% 1.042e+09 ± 2% proc-vmstat.numa_local
411756 ± 21% -63.5% 150280 ± 34% proc-vmstat.numa_pte_updates
1.34e+10 -37.9% 8.324e+09 ± 2% proc-vmstat.pgalloc_normal
1393623 ± 8% -23.4% 1067508 proc-vmstat.pgfault
1.34e+10 -37.9% 8.323e+09 ± 2% proc-vmstat.pgfree
11265047 -21.1% 8884763 sched_debug.cfs_rq:/.avg_vruntime.avg
13067285 ± 2% -26.3% 9630862 ± 2% sched_debug.cfs_rq:/.avg_vruntime.max
10675424 -23.9% 8119367 ± 3% sched_debug.cfs_rq:/.avg_vruntime.min
0.78 -13.5% 0.67 ± 2% sched_debug.cfs_rq:/.h_nr_queued.avg
0.37 +13.0% 0.42 ± 2% sched_debug.cfs_rq:/.h_nr_queued.stddev
0.77 -13.5% 0.67 ± 2% sched_debug.cfs_rq:/.h_nr_runnable.avg
0.37 ± 2% +12.8% 0.42 ± 2% sched_debug.cfs_rq:/.h_nr_runnable.stddev
8980 ± 10% -14.6% 7667 ± 7% sched_debug.cfs_rq:/.load.avg
11265047 -21.1% 8884763 sched_debug.cfs_rq:/.min_vruntime.avg
13067285 ± 2% -26.3% 9630862 ± 2% sched_debug.cfs_rq:/.min_vruntime.max
10675424 -23.9% 8119367 ± 3% sched_debug.cfs_rq:/.min_vruntime.min
0.75 -11.7% 0.66 sched_debug.cfs_rq:/.nr_queued.avg
0.33 ± 3% +23.5% 0.40 ± 2% sched_debug.cfs_rq:/.nr_queued.stddev
265.16 ± 2% +18.4% 313.92 sched_debug.cfs_rq:/.util_avg.stddev
628.03 -17.3% 519.63 ± 2% sched_debug.cfs_rq:/.util_est.avg
313.28 ± 2% +20.2% 376.59 sched_debug.cfs_rq:/.util_est.stddev
5339 ± 5% +79.9% 9606 ± 3% sched_debug.cpu.avg_idle.min
1441 ± 7% -17.8% 1185 ± 17% sched_debug.cpu.clock_task.stddev
2929 -10.2% 2632 sched_debug.cpu.curr->pid.avg
1380 ± 2% +17.9% 1628 sched_debug.cpu.curr->pid.stddev
0.76 -14.0% 0.66 ± 2% sched_debug.cpu.nr_running.avg
0.39 ± 2% +11.2% 0.43 sched_debug.cpu.nr_running.stddev
5279363 -14.3% 4526246 ± 3% sched_debug.cpu.nr_switches.avg
2.297e+10 -28.9% 1.634e+10 ± 2% perf-stat.i.branch-instructions
0.81 +0.1 0.93 perf-stat.i.branch-miss-rate%
1.832e+08 -18.4% 1.495e+08 ± 2% perf-stat.i.branch-misses
1.64 ± 7% +0.6 2.27 ± 13% perf-stat.i.cache-miss-rate%
6.943e+09 -34.2% 4.57e+09 ± 2% perf-stat.i.cache-references
4494744 ± 2% -13.0% 3911893 ± 4% perf-stat.i.context-switches
2.51 +36.1% 3.42 ± 2% perf-stat.i.cpi
2.932e+11 -4.2% 2.81e+11 perf-stat.i.cpu-cycles
2907 ± 16% +1723.9% 53022 ± 13% perf-stat.i.cpu-migrations
1.167e+11 -29.3% 8.249e+10 ± 2% perf-stat.i.instructions
0.40 -25.8% 0.30 ± 2% perf-stat.i.ipc
0.04 ± 37% -81.0% 0.01 ± 47% perf-stat.i.major-faults
35.11 ± 2% -12.9% 30.57 ± 4% perf-stat.i.metric.K/sec
4270 ± 8% -25.2% 3195 perf-stat.i.minor-faults
4270 ± 8% -25.2% 3195 perf-stat.i.page-faults
0.80 +0.1 0.92 perf-stat.overall.branch-miss-rate%
1.58 ± 7% +0.6 2.20 ± 13% perf-stat.overall.cache-miss-rate%
2.51 +35.7% 3.41 ± 2% perf-stat.overall.cpi
0.40 -26.3% 0.29 ± 2% perf-stat.overall.ipc
10488 +13.8% 11937 perf-stat.overall.path-length
2.289e+10 -28.9% 1.628e+10 ± 2% perf-stat.ps.branch-instructions
1.826e+08 -18.4% 1.49e+08 ± 2% perf-stat.ps.branch-misses
6.92e+09 -34.2% 4.555e+09 ± 2% perf-stat.ps.cache-references
4479829 ± 2% -13.0% 3898858 ± 4% perf-stat.ps.context-switches
2.923e+11 -4.2% 2.8e+11 perf-stat.ps.cpu-cycles
2901 ± 16% +1721.9% 52859 ± 13% perf-stat.ps.cpu-migrations
1.164e+11 -29.3% 8.222e+10 ± 2% perf-stat.ps.instructions
0.04 ± 36% -81.1% 0.01 ± 47% perf-stat.ps.major-faults
4246 ± 8% -25.2% 3175 perf-stat.ps.minor-faults
4246 ± 8% -25.2% 3175 perf-stat.ps.page-faults
3.515e+13 -29.3% 2.484e+13 ± 2% perf-stat.total.instructions
0.01 ± 36% +4025.9% 0.37 ±133% perf-sched.sch_delay.avg.ms.__cond_resched.__release_sock.release_sock.sk_wait_data.tcp_recvmsg_locked
0.10 ± 30% +275.0% 0.38 ± 23% perf-sched.sch_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
0.21 ±134% +26130.2% 54.60 ± 11% perf-sched.sch_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
0.13 ± 11% +4401.2% 6.02 ± 34% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.19 ±150% +29283.8% 54.51 ± 8% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
0.02 ± 42% +1730.4% 0.35 ± 60% perf-sched.sch_delay.avg.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
0.09 ± 61% +83629.3% 76.33 ± 49% perf-sched.sch_delay.avg.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
2.00 ± 67% +2740.7% 56.88 ± 93% perf-sched.sch_delay.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
0.06 ± 57% +1.6e+05% 97.93 ± 48% perf-sched.sch_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
0.37 ± 89% +4177.4% 15.90 ± 19% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
0.61 ± 87% +12221.9% 75.25 ± 56% perf-sched.sch_delay.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
0.01 ± 21% +289.8% 0.03 ± 5% perf-sched.sch_delay.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
3.12 ±102% +5221.9% 166.03 ± 7% perf-sched.sch_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.15 ±121% +4038.3% 6.25 ± 10% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
62.06 ± 52% +247.8% 215.86 ± 46% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.__sk_flush_backlog.tcp_recvmsg_locked.tcp_recvmsg
0.03 ± 64% +22004.8% 6.93 ±136% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.release_sock.sk_wait_data.tcp_recvmsg_locked
36.09 ± 57% +524.7% 225.47 ± 52% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
146.95 ± 33% +643.8% 1092 ± 27% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
3.47 ± 33% +29240.5% 1019 perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.56 ± 65% -88.1% 0.07 ±136% perf-sched.sch_delay.max.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
173.81 ± 42% +538.7% 1110 ± 46% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
0.89 ± 92% +3331.6% 30.64 ± 72% perf-sched.sch_delay.max.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
3.17 ±107% +20724.1% 660.12 ± 82% perf-sched.sch_delay.max.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
0.04 ± 17% -62.0% 0.01 ± 48% perf-sched.sch_delay.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
2.01 ± 52% +74985.5% 1509 ± 33% perf-sched.sch_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
63.02 ±125% +610.9% 448.01 ± 7% perf-sched.sch_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
138.54 ± 15% +265.7% 506.70 ± 15% perf-sched.sch_delay.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
783.25 ± 33% +317.5% 3269 ± 31% perf-sched.sch_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
437.71 ± 91% +197.5% 1302 ± 34% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.01 ± 44% +1817.5% 0.26 ± 6% perf-sched.total_sch_delay.average.ms
1168 ± 33% +265.4% 4269 ± 14% perf-sched.total_sch_delay.max.ms
0.27 ± 4% +239.8% 0.93 ± 8% perf-sched.total_wait_and_delay.average.ms
6426495 ± 3% -66.1% 2178159 ± 8% perf-sched.total_wait_and_delay.count.ms
4156 ± 6% +97.9% 8227 ± 21% perf-sched.total_wait_and_delay.max.ms
0.26 ± 4% +159.1% 0.68 ± 9% perf-sched.total_wait_time.average.ms
4156 ± 6% +29.0% 5361 ± 6% perf-sched.total_wait_time.max.ms
0.34 ± 18% +351.4% 1.53 ± 18% perf-sched.wait_and_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
0.37 ±168% +29845.0% 109.35 ± 12% perf-sched.wait_and_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
5.01 ± 24% +307.2% 20.41 ± 18% perf-sched.wait_and_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.33 ±181% +32588.3% 109.02 ± 8% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
0.25 ± 77% +60202.8% 152.67 ± 49% perf-sched.wait_and_delay.avg.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
37.01 ± 5% +185.4% 105.64 ± 37% perf-sched.wait_and_delay.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
22.34 ± 59% +1096.8% 267.33 ± 51% perf-sched.wait_and_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
28.51 ± 11% +657.4% 215.92 ± 65% perf-sched.wait_and_delay.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
5.47 ±111% +5490.8% 305.70 ± 37% perf-sched.wait_and_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
491.64 ± 7% +244.5% 1693 ± 60% perf-sched.wait_and_delay.avg.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
5.44 ± 13% +581.2% 37.03 ± 16% perf-sched.wait_and_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
1.37 ± 79% +10884.8% 150.49 ± 56% perf-sched.wait_and_delay.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
0.02 ± 14% +561.9% 0.16 ± 5% perf-sched.wait_and_delay.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
182.37 ± 9% +210.0% 565.41 ± 6% perf-sched.wait_and_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.44 ±113% +2996.3% 13.52 ± 10% perf-sched.wait_and_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
696.96 ± 6% +47.1% 1025 ± 24% perf-sched.wait_and_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
1078 ± 73% +218.9% 3438 ± 14% perf-sched.wait_and_delay.count.__cond_resched.__release_sock.__sk_flush_backlog.tcp_recvmsg_locked.tcp_recvmsg
5253 ± 23% -64.0% 1891 ± 13% perf-sched.wait_and_delay.count.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
663.00 ± 8% -54.6% 301.17 ± 19% perf-sched.wait_and_delay.count.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
12.50 ± 22% -82.7% 2.17 ± 90% perf-sched.wait_and_delay.count.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
4.83 ± 7% -79.3% 1.00 ±100% perf-sched.wait_and_delay.count.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
5.67 ± 8% -79.4% 1.17 ±104% perf-sched.wait_and_delay.count.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
115.50 ± 5% -78.9% 24.33 ± 13% perf-sched.wait_and_delay.count.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
117.50 ± 4% -73.8% 30.83 ± 8% perf-sched.wait_and_delay.count.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
11.33 ± 26% -88.2% 1.33 ±141% perf-sched.wait_and_delay.count.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
123.83 ± 49% +136.5% 292.83 ± 22% perf-sched.wait_and_delay.count.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
2233 ± 9% -85.5% 323.17 ± 94% perf-sched.wait_and_delay.count.pipe_read.vfs_read.ksys_read.do_syscall_64
22.67 ± 4% -63.2% 8.33 ± 49% perf-sched.wait_and_delay.count.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
12.50 ± 6% -61.3% 4.83 ± 36% perf-sched.wait_and_delay.count.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
82.50 ± 4% -71.5% 23.50 ± 17% perf-sched.wait_and_delay.count.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
19.17 ± 3% -77.4% 4.33 ± 17% perf-sched.wait_and_delay.count.schedule_timeout.kcompactd.kthread.ret_from_fork
1001 ± 6% -73.0% 270.67 ± 15% perf-sched.wait_and_delay.count.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
1790 ± 32% -99.6% 8.00 ± 19% perf-sched.wait_and_delay.count.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
6330659 ± 3% -65.9% 2157272 ± 8% perf-sched.wait_and_delay.count.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
5075 ± 13% -64.9% 1781 ± 9% perf-sched.wait_and_delay.count.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
38453 ± 49% -73.5% 10206 ± 16% perf-sched.wait_and_delay.count.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
766.67 ± 6% -70.7% 224.33 ± 17% perf-sched.wait_and_delay.count.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
108.98 ± 77% +421.4% 568.27 ± 31% perf-sched.wait_and_delay.max.ms.__cond_resched.__release_sock.__sk_flush_backlog.tcp_recvmsg_locked.tcp_recvmsg
73.56 ± 54% +619.8% 529.51 ± 35% perf-sched.wait_and_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
163.67 ±108% +1235.6% 2185 ± 27% perf-sched.wait_and_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
1001 +103.6% 2038 perf-sched.wait_and_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
174.95 ±126% +1169.0% 2220 ± 46% perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
6.13 ±115% +21438.1% 1320 ± 82% perf-sched.wait_and_delay.max.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
2788 ± 25% -62.0% 1058 ±102% perf-sched.wait_and_delay.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
741.28 ± 14% +125.8% 1673 ± 14% perf-sched.wait_and_delay.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
1003 +251.0% 3523 ± 61% perf-sched.wait_and_delay.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
342.95 ±136% +927.4% 3523 ± 39% perf-sched.wait_and_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
1008 ± 40% +374.8% 4787 ± 65% perf-sched.wait_and_delay.max.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
207.61 ± 65% +333.7% 900.49 ± 7% perf-sched.wait_and_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
277.11 ± 15% +268.2% 1020 ± 15% perf-sched.wait_and_delay.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
4119 ± 8% +73.3% 7139 ± 26% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
3146 ± 20% +89.9% 5976 ± 18% perf-sched.wait_and_delay.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
0.37 ±104% +369.8% 1.76 ± 15% perf-sched.wait_time.avg.ms.__cond_resched.__release_sock.__sk_flush_backlog.tcp_recvmsg_locked.tcp_recvmsg
0.02 ± 35% +9586.0% 1.61 ±149% perf-sched.wait_time.avg.ms.__cond_resched.__release_sock.release_sock.sk_wait_data.tcp_recvmsg_locked
0.24 ± 13% +383.1% 1.15 ± 17% perf-sched.wait_time.avg.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
0.23 ±128% +23772.2% 54.75 ± 12% perf-sched.wait_time.avg.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
4.88 ± 24% +195.1% 14.39 ± 14% perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.20 ±144% +26532.7% 54.51 ± 8% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
0.07 ± 35% +2174.0% 1.58 ± 52% perf-sched.wait_time.avg.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
0.23 ± 29% +33379.5% 76.33 ± 49% perf-sched.wait_time.avg.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
37.01 ± 5% +185.4% 105.64 ± 37% perf-sched.wait_time.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
22.28 ± 59% +765.9% 192.89 ± 46% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.39 ±101% -100.0% 0.00 perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
26.50 ± 12% +500.0% 159.04 ± 67% perf-sched.wait_time.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
498.82 ± 9% -70.4% 147.68 ±114% perf-sched.wait_time.avg.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
5.41 ±112% +3742.6% 207.76 ± 43% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
473.69 ± 3% +177.8% 1315 ± 45% perf-sched.wait_time.avg.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
5.06 ± 7% +317.1% 21.12 ± 13% perf-sched.wait_time.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
0.76 ± 72% +9807.3% 75.25 ± 56% perf-sched.wait_time.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
0.02 ± 10% +690.9% 0.13 ± 6% perf-sched.wait_time.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
179.25 ± 8% +122.8% 399.39 ± 5% perf-sched.wait_time.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.29 ±108% +2447.1% 7.27 ± 10% perf-sched.wait_time.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
63.71 ± 51% +604.4% 448.79 ± 18% perf-sched.wait_time.max.ms.__cond_resched.__release_sock.__sk_flush_backlog.tcp_recvmsg_locked.tcp_recvmsg
0.09 ±130% +36184.9% 33.26 ±175% perf-sched.wait_time.max.ms.__cond_resched.__release_sock.release_sock.sk_wait_data.tcp_recvmsg_locked
38.34 ± 49% +782.5% 338.36 ± 32% perf-sched.wait_time.max.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
146.97 ± 33% +643.7% 1092 ± 27% perf-sched.wait_time.max.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
0.62 ± 57% -89.2% 0.07 ±136% perf-sched.wait_time.max.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
173.81 ± 42% +538.7% 1110 ± 46% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
1.41 ± 42% +10955.3% 155.36 ± 76% perf-sched.wait_time.max.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
4.70 ± 52% +13934.8% 660.12 ± 82% perf-sched.wait_time.max.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
2788 ± 25% -62.0% 1058 ±102% perf-sched.wait_time.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
741.28 ± 14% +125.8% 1673 ± 14% perf-sched.wait_time.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
1002 +151.0% 2515 ± 44% perf-sched.wait_time.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.55 ± 78% -100.0% 0.00 perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
1688 ± 27% +98.5% 3350 ± 22% perf-sched.wait_time.max.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
341.76 ±136% +587.1% 2348 ± 31% perf-sched.wait_time.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
840.47 ± 28% +309.7% 3443 ± 44% perf-sched.wait_time.max.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
167.71 ± 40% +174.9% 460.98 ± 6% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
138.56 ± 15% +284.4% 532.69 ± 12% perf-sched.wait_time.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
4119 ± 8% +20.5% 4965 ± 7% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
2922 ± 15% +75.6% 5133 ± 10% perf-sched.wait_time.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
***************************************************************************************************
lkp-emr-2sp1: 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime:
gcc-12/performance/x86_64-rhel-9.4/100%/debian-12-x86_64-20240206.cgz/lkp-emr-2sp1/mmapfork/stress-ng/60s
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
61300134 ± 4% -11.2% 54464501 ± 3% meminfo.max_used_kB
100.42 +12.5% 112.99 ± 4% uptime.boot
1.93 ± 13% -1.8 0.16 ± 14% mpstat.cpu.all.soft%
17.02 ± 2% -5.5 11.49 ± 2% mpstat.cpu.all.usr%
5510 ± 48% +195.8% 16296 ± 23% perf-c2c.DRAM.remote
1484 ± 49% +138.9% 3545 ± 24% perf-c2c.HITM.remote
16.52 ± 2% -32.4% 11.16 ± 2% vmstat.cpu.us
79660 -17.3% 65897 ± 4% vmstat.system.cs
3.332e+08 ± 2% -54.9% 1.502e+08 ± 22% numa-numastat.node0.local_node
3.337e+08 -54.9% 1.504e+08 ± 22% numa-numastat.node0.numa_hit
3.329e+08 -72.4% 92005595 ± 52% numa-numastat.node1.local_node
3.335e+08 -72.4% 92128084 ± 52% numa-numastat.node1.numa_hit
3.205e+08 -51.8% 1.546e+08 ± 16% numa-numastat.node2.local_node
3.208e+08 -51.8% 1.547e+08 ± 16% numa-numastat.node2.numa_hit
3.173e+08 ± 2% -73.1% 85455746 ± 45% numa-numastat.node3.local_node
3.176e+08 ± 2% -73.1% 85574085 ± 45% numa-numastat.node3.numa_hit
12219 -63.1% 4511 ± 6% stress-ng.mmapfork.ops
202.59 -63.5% 73.93 ± 6% stress-ng.mmapfork.ops_per_sec
64.01 +19.7% 76.59 ± 7% stress-ng.time.elapsed_time
64.01 +19.7% 76.59 ± 7% stress-ng.time.elapsed_time.max
4100955 +10.0% 4509134 ± 4% stress-ng.time.involuntary_context_switches
1.3e+09 -63.1% 4.801e+08 ± 7% stress-ng.time.minor_page_faults
24509 +1.4% 24848 stress-ng.time.percent_of_cpu_this_job_got
12906 +30.3% 16810 ± 6% stress-ng.time.system_time
2783 ± 2% -20.4% 2216 ± 7% stress-ng.time.user_time
464362 -60.5% 183317 ± 5% stress-ng.time.voluntary_context_switches
5361967 +1.5% 5440921 proc-vmstat.nr_dirty_background_threshold
10737045 +1.5% 10895145 proc-vmstat.nr_dirty_threshold
7335955 -7.4% 6793970 ± 6% proc-vmstat.nr_file_pages
53947071 +1.4% 54723190 proc-vmstat.nr_free_pages
6453955 ± 2% -8.4% 5911969 ± 7% proc-vmstat.nr_shmem
60752 -5.1% 57679 proc-vmstat.nr_slab_reclaimable
210937 -7.3% 195559 ± 3% proc-vmstat.nr_slab_unreclaimable
1.306e+09 -63.0% 4.83e+08 ± 7% proc-vmstat.numa_hit
1.305e+09 -63.0% 4.824e+08 ± 7% proc-vmstat.numa_local
1.505e+09 -63.0% 5.563e+08 ± 7% proc-vmstat.pgalloc_normal
1.301e+09 -63.0% 4.807e+08 ± 7% proc-vmstat.pgfault
1.504e+09 -63.1% 5.55e+08 ± 7% proc-vmstat.pgfree
616613 ± 23% -54.7% 279591 ± 21% proc-vmstat.pgreuse
389483 -63.3% 142929 ± 6% proc-vmstat.thp_fault_alloc
5489228 ± 29% +62.7% 8928447 ± 38% numa-meminfo.node0.FilePages
10213877 ± 31% -50.5% 5060225 ± 37% numa-meminfo.node3.Active
10213877 ± 31% -50.5% 5060225 ± 37% numa-meminfo.node3.Active(anon)
1849516 ± 31% -58.7% 764271 ± 49% numa-meminfo.node3.AnonHugePages
2103284 ± 26% -54.3% 961145 ± 33% numa-meminfo.node3.AnonPages
4500176 ± 30% -62.0% 1708987 ± 31% numa-meminfo.node3.AnonPages.max
8719420 ± 29% -52.1% 4173994 ± 41% numa-meminfo.node3.FilePages
28893 ± 23% -38.5% 17773 ± 19% numa-meminfo.node3.KernelStack
6264266 ± 35% -49.9% 3139363 ± 52% numa-meminfo.node3.Mapped
52318919 ± 6% +11.7% 58419737 ± 3% numa-meminfo.node3.MemFree
13602300 ± 23% -44.9% 7501482 ± 28% numa-meminfo.node3.MemUsed
146208 ± 31% -56.0% 64374 ± 39% numa-meminfo.node3.PageTables
217141 ± 16% -33.3% 144858 ± 12% numa-meminfo.node3.SUnreclaim
8085495 ± 33% -49.3% 4097432 ± 39% numa-meminfo.node3.Shmem
279175 ± 13% -32.4% 188683 ± 16% numa-meminfo.node3.Slab
3.341e+08 -55.0% 1.504e+08 ± 22% numa-vmstat.node0.numa_hit
3.335e+08 ± 2% -55.0% 1.502e+08 ± 22% numa-vmstat.node0.numa_local
3.338e+08 -72.4% 92136512 ± 52% numa-vmstat.node1.numa_hit
3.332e+08 -72.4% 92014024 ± 52% numa-vmstat.node1.numa_local
3.211e+08 -51.8% 1.547e+08 ± 16% numa-vmstat.node2.numa_hit
3.208e+08 -51.8% 1.546e+08 ± 16% numa-vmstat.node2.numa_local
2548625 ± 32% -50.5% 1261365 ± 41% numa-vmstat.node3.nr_active_anon
526621 ± 27% -54.4% 239946 ± 36% numa-vmstat.node3.nr_anon_pages
904.50 ± 32% -59.0% 370.93 ± 53% numa-vmstat.node3.nr_anon_transparent_hugepages
2173840 ± 30% -52.2% 1040058 ± 45% numa-vmstat.node3.nr_file_pages
13073386 ± 6% +11.8% 14619853 ± 4% numa-vmstat.node3.nr_free_pages
28795 ± 23% -38.2% 17795 ± 20% numa-vmstat.node3.nr_kernel_stack
35596 ± 33% -54.7% 16125 ± 43% numa-vmstat.node3.nr_page_table_pages
2015361 ± 33% -49.3% 1020916 ± 44% numa-vmstat.node3.nr_shmem
53949 ± 16% -32.8% 36254 ± 14% numa-vmstat.node3.nr_slab_unreclaimable
2548649 ± 32% -50.5% 1261287 ± 41% numa-vmstat.node3.nr_zone_active_anon
3.179e+08 ± 2% -73.1% 85558747 ± 45% numa-vmstat.node3.numa_hit
3.176e+08 ± 2% -73.1% 85440409 ± 45% numa-vmstat.node3.numa_local
12034713 ± 2% +27.0% 15288847 ± 8% sched_debug.cfs_rq:/.avg_vruntime.max
1438718 ± 32% +154.6% 3662687 ± 29% sched_debug.cfs_rq:/.avg_vruntime.stddev
70491 ±190% -95.0% 3512 ± 44% sched_debug.cfs_rq:/.load.avg
743319 ±202% -97.4% 19566 ±124% sched_debug.cfs_rq:/.load.stddev
12034713 ± 2% +27.0% 15288847 ± 8% sched_debug.cfs_rq:/.min_vruntime.max
1438712 ± 32% +154.6% 3662687 ± 29% sched_debug.cfs_rq:/.min_vruntime.stddev
222.33 ± 6% -11.2% 197.52 ± 8% sched_debug.cfs_rq:/.util_avg.stddev
60.33 ± 23% -50.9% 29.61 ± 22% sched_debug.cfs_rq:/.util_est.avg
139547 ± 15% +69.7% 236820 ± 23% sched_debug.cpu.avg_idle.stddev
179893 ± 2% -65.7% 61761 ± 3% sched_debug.cpu.curr->pid.avg
192433 -64.8% 67681 ± 4% sched_debug.cpu.curr->pid.max
38238 ± 26% -72.7% 10420 ± 8% sched_debug.cpu.curr->pid.stddev
868454 ± 22% +35.9% 1180661 ± 19% sched_debug.cpu.max_idle_balance_cost.max
29004 ± 42% +201.6% 87490 ± 36% sched_debug.cpu.max_idle_balance_cost.stddev
10283 -12.3% 9017 ± 3% sched_debug.cpu.nr_switches.avg
8342 ± 4% -24.6% 6290 ± 4% sched_debug.cpu.nr_switches.min
1412 ± 47% +60.5% 2267 ± 35% sched_debug.cpu.nr_switches.stddev
57.00 ± 39% -50.0% 28.50 ± 49% sched_debug.cpu.nr_uninterruptible.max
7.07 ± 7% -34.3% 4.65 ± 18% sched_debug.cpu.nr_uninterruptible.stddev
4.289e+10 -70.6% 1.263e+10 ± 2% perf-stat.i.branch-instructions
0.27 +0.1 0.40 ± 4% perf-stat.i.branch-miss-rate%
86292456 ± 2% -49.7% 43436183 ± 6% perf-stat.i.branch-misses
52.01 -8.7 43.30 ± 5% perf-stat.i.cache-miss-rate%
1.09e+09 -72.1% 3.043e+08 ± 15% perf-stat.i.cache-misses
2.089e+09 -66.0% 7.105e+08 ± 20% perf-stat.i.cache-references
81002 -21.4% 63704 ± 5% perf-stat.i.context-switches
3.27 +240.9% 11.16 ± 2% perf-stat.i.cpi
7.08e+11 +1.7% 7.2e+11 perf-stat.i.cpu-cycles
7221 ± 5% -73.5% 1917 ± 5% perf-stat.i.cpu-migrations
642.37 +274.4% 2404 ± 18% perf-stat.i.cycles-between-cache-misses
2.136e+11 -70.2% 6.365e+10 ± 2% perf-stat.i.instructions
0.32 -63.8% 0.12 ± 3% perf-stat.i.ipc
1.64 ± 59% -94.6% 0.09 ± 88% perf-stat.i.major-faults
2.34 ± 4% -100.0% 0.00 perf-stat.i.metric.K/sec
314368 ± 2% -68.1% 100262 ± 2% perf-stat.i.minor-faults
314370 ± 2% -68.1% 100263 ± 2% perf-stat.i.page-faults
0.20 ± 2% +0.1 0.32 ± 9% perf-stat.overall.branch-miss-rate%
51.98 -8.5 43.45 ± 5% perf-stat.overall.cache-miss-rate%
3.33 +241.0% 11.35 ± 2% perf-stat.overall.cpi
651.94 +273.2% 2432 ± 18% perf-stat.overall.cycles-between-cache-misses
0.30 -70.7% 0.09 ± 2% perf-stat.overall.ipc
4.2e+10 -70.2% 1.251e+10 ± 2% perf-stat.ps.branch-instructions
83007046 ± 2% -52.3% 39570947 ± 8% perf-stat.ps.branch-misses
1.068e+09 -71.6% 3.03e+08 ± 16% perf-stat.ps.cache-misses
2.054e+09 -65.7% 7.048e+08 ± 20% perf-stat.ps.cache-references
79568 -19.1% 64350 ± 3% perf-stat.ps.context-switches
6.961e+11 +2.7% 7.152e+11 perf-stat.ps.cpu-cycles
7079 ± 5% -73.2% 1894 ± 5% perf-stat.ps.cpu-migrations
2.091e+11 -69.9% 6.303e+10 ± 2% perf-stat.ps.instructions
1.56 ± 59% -97.0% 0.05 ± 74% perf-stat.ps.major-faults
305825 ± 3% -68.3% 97054 perf-stat.ps.minor-faults
305826 ± 3% -68.3% 97054 perf-stat.ps.page-faults
1.349e+13 -64.1% 4.846e+12 ± 8% perf-stat.total.instructions
28.19 ± 10% +366.8% 131.59 ± 88% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pmd_alloc
34.75 ± 15% +77.7% 61.75 ± 36% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pud_alloc
34.03 ± 29% +54.9% 52.72 ± 20% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pgd_alloc
24.31 ± 9% +198.8% 72.64 ± 33% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pte_alloc_one
25.40 ±110% +244.6% 87.53 ± 42% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.allocate_slab.___slab_alloc
41.64 ± 10% +41.3% 58.86 ± 9% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
33.33 ± 8% +213.7% 104.54 ± 68% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.vma_alloc_folio_noprof
35.00 ± 28% +67.7% 58.69 ± 26% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_cache_node_noprof.__get_vm_area_node.__vmalloc_node_range_noprof.alloc_thread_stack_node
6.70 ± 98% +367.0% 31.28 ± 71% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
31.30 ± 30% +619.1% 225.07 ±179% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.__vmalloc_area_node.__vmalloc_node_range_noprof.alloc_thread_stack_node
34.75 ± 19% +198.6% 103.76 ± 94% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
38.41 ± 10% +65.4% 63.55 ± 24% perf-sched.sch_delay.avg.ms.__cond_resched.__put_anon_vma.unlink_anon_vmas.free_pgtables.exit_mmap
44.17 ± 6% +57.3% 69.45 ± 25% perf-sched.sch_delay.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
30.30 ± 17% +41.9% 43.00 ± 14% perf-sched.sch_delay.avg.ms.__cond_resched.__vmalloc_area_node.__vmalloc_node_range_noprof.alloc_thread_stack_node.dup_task_struct
1.48 ± 33% +818.4% 13.56 ±109% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
32.02 ± 28% +98.3% 63.48 ± 23% perf-sched.sch_delay.avg.ms.__cond_resched.cgroup_css_set_fork.cgroup_can_fork.copy_process.kernel_clone
33.07 ± 11% +184.3% 94.01 ± 52% perf-sched.sch_delay.avg.ms.__cond_resched.copy_page_range.dup_mmap.dup_mm.constprop
30.00 ± 9% +239.8% 101.93 ± 52% perf-sched.sch_delay.avg.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
6.51 ± 31% +348.8% 29.20 ± 40% perf-sched.sch_delay.avg.ms.__cond_resched.down_read.__mm_populate.vm_mmap_pgoff.do_syscall_64
1.37 ±123% +609.8% 9.72 ± 70% perf-sched.sch_delay.avg.ms.__cond_resched.down_read.walk_component.link_path_walk.part
31.18 ± 6% +197.2% 92.68 ± 52% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.anon_vma_clone.anon_vma_fork.dup_mmap
30.95 ± 10% +282.4% 118.34 ± 67% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.anon_vma_fork.dup_mmap.dup_mm
33.18 ± 9% +166.0% 88.25 ± 41% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.dup_mmap.dup_mm.constprop
39.63 ± 6% +67.8% 66.50 ± 21% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.free_pgtables.exit_mmap.__mmput
38.72 ± 9% +62.3% 62.86 ± 14% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
37.73 ± 12% +62.0% 61.14 ± 11% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.unlink_file_vma_batch_add.free_pgtables.exit_mmap
8.30 ± 50% +275.7% 31.20 ± 43% perf-sched.sch_delay.avg.ms.__cond_resched.down_write.vma_link_file.__mmap_new_vma.__mmap_region
0.67 ± 50% +290.9% 2.61 ± 46% perf-sched.sch_delay.avg.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
4.39 ±101% +698.1% 35.06 ± 49% perf-sched.sch_delay.avg.ms.__cond_resched.down_write_killable.vm_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
38.18 ± 25% +60.7% 61.36 ± 15% perf-sched.sch_delay.avg.ms.__cond_resched.dput.path_put.exit_fs.do_exit
0.48 ±135% +2338.4% 11.59 ± 64% perf-sched.sch_delay.avg.ms.__cond_resched.dput.step_into.link_path_walk.part
0.60 ±152% +1831.8% 11.56 ± 89% perf-sched.sch_delay.avg.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
42.64 ± 7% +52.1% 64.85 ± 20% perf-sched.sch_delay.avg.ms.__cond_resched.exit_mmap.__mmput.exit_mm.do_exit
21.22 ± 35% -47.9% 11.05 ± 56% perf-sched.sch_delay.avg.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
35.21 ± 16% +339.5% 154.78 ±119% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_bulk_noprof.mas_dup_alloc.isra.0
8.71 ± 54% +270.1% 32.24 ± 40% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_lru_noprof.__d_alloc.d_alloc_pseudo.alloc_file_pseudo
7.82 ± 51% +285.6% 30.14 ± 40% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.alloc_file_pseudo.__shmem_file_setup
1.73 ±141% +785.8% 15.36 ± 63% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.path_openat.do_filp_open
35.74 ± 31% +484.4% 208.89 ±103% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_pid.copy_process.kernel_clone
31.90 ± 10% +214.3% 100.25 ± 54% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
25.61 ± 29% +644.5% 190.67 ±164% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.copy_signal.copy_process.kernel_clone
37.35 ± 13% +61.0% 60.16 ± 19% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.dup_mm.constprop.0
0.75 ±223% +1417.4% 11.45 ±120% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.getname_flags.part.0
9.62 ± 46% +243.7% 33.06 ± 36% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.mas_alloc_nodes.mas_preallocate.__mmap_new_vma
27.94 ± 13% +207.2% 85.84 ± 27% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.prepare_creds.copy_creds.copy_process
7.69 ± 48% +288.3% 29.87 ± 34% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.security_inode_alloc.inode_init_always_gfp.alloc_inode
8.08 ± 23% +264.7% 29.45 ± 42% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
1.21 ±172% +1936.5% 24.64 ± 77% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.__split_vma.vms_gather_munmap_vmas
32.66 ± 9% +184.8% 93.02 ± 44% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.dup_mmap.dup_mm
38.89 ± 11% +85.5% 72.16 ± 34% perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock.futex_exit_release.exit_mm_release.exit_mm
4.15 ±223% +1063.1% 48.31 ± 33% perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock.perf_event_exit_task.do_exit.do_group_exit
32.51 ± 17% +291.0% 127.10 ± 66% perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
44.38 ± 16% +262.6% 160.91 ±101% perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.mm_init.dup_mm
35.49 ± 15% +67.2% 59.34 ± 26% perf-sched.sch_delay.avg.ms.__cond_resched.remove_vma.exit_mmap.__mmput.exit_mm
48.78 ± 10% +24.2% 60.57 ± 13% perf-sched.sch_delay.avg.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
45.39 ± 8% +53.8% 69.81 ± 12% perf-sched.sch_delay.avg.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
42.58 ± 30% +283.5% 163.28 ±140% perf-sched.sch_delay.avg.ms.__cond_resched.switch_task_namespaces.do_exit.do_group_exit.__x64_sys_exit_group
14.92 ±101% +260.8% 53.85 ± 29% perf-sched.sch_delay.avg.ms.__cond_resched.task_work_run.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
35.61 ± 16% +140.7% 85.71 ± 76% perf-sched.sch_delay.avg.ms.__cond_resched.uprobe_start_dup_mmap.dup_mm.constprop.0
29.84 ± 17% +83.0% 54.61 ± 9% perf-sched.sch_delay.avg.ms.__cond_resched.wp_page_copy.__handle_mm_fault.handle_mm_fault.do_user_addr_fault
46.38 ± 7% +35.6% 62.87 ± 8% perf-sched.sch_delay.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
11.27 ± 11% +933.5% 116.43 ±119% perf-sched.sch_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
11.15 ± 11% +459.2% 62.34 ± 58% perf-sched.sch_delay.avg.ms.io_schedule.folio_wait_bit_common.filemap_fault.__do_fault
34.39 ± 9% +59.8% 54.95 ± 9% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
30.52 ± 18% +41.8% 43.27 ± 10% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
2.28 ± 49% +1270.2% 31.31 ±114% perf-sched.sch_delay.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
7.76 ± 35% +144.3% 18.95 ± 31% perf-sched.sch_delay.avg.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
38.01 ± 23% +151.3% 95.53 ± 46% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.__put_anon_vma
29.46 ± 14% +201.3% 88.77 ± 63% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.anon_vma_clone
29.91 ± 10% +134.2% 70.04 ± 23% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.anon_vma_fork
32.15 ± 24% +102.8% 65.18 ± 23% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.dup_mmap
36.21 ± 20% +56.2% 56.56 ± 27% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.unlink_anon_vmas
14.35 ± 96% +344.7% 63.80 ± 23% perf-sched.sch_delay.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.unlink_file_vma_batch_final
0.08 ± 57% +5932.7% 4.71 ±121% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
37.96 ± 13% +84.3% 69.97 ± 37% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
4.77 ± 39% -70.7% 1.40 ± 99% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.ret_from_fork_asm.[unknown]
111.67 ± 20% +60.7% 179.44 ± 19% perf-sched.sch_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pgd_alloc
445.63 ± 20% +632.3% 3263 ± 57% perf-sched.sch_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
99.05 ± 17% +72.4% 170.81 ± 24% perf-sched.sch_delay.max.ms.__cond_resched.__dentry_kill.dput.__fput.task_work_run
483.91 ± 16% +592.2% 3349 ± 56% perf-sched.sch_delay.max.ms.__cond_resched.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff
148.35 ± 21% +1202.9% 1932 ±113% perf-sched.sch_delay.max.ms.__cond_resched.__kmalloc_cache_node_noprof.__get_vm_area_node.__vmalloc_node_range_noprof.alloc_thread_stack_node
362.40 ± 14% +921.6% 3702 ± 62% perf-sched.sch_delay.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range.zap_pmd_range
257.63 ±145% +1627.8% 4451 ±147% perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
260.06 ± 33% +4874.0% 12935 ±112% perf-sched.sch_delay.max.ms.__cond_resched.copy_page_range.dup_mmap.dup_mm.constprop
271.64 ± 21% +5159.3% 14286 ±101% perf-sched.sch_delay.max.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
33.52 ± 39% +228.4% 110.08 ± 37% perf-sched.sch_delay.max.ms.__cond_resched.down_read.__mm_populate.vm_mmap_pgoff.do_syscall_64
104.84 ± 33% +577.7% 710.49 ±173% perf-sched.sch_delay.max.ms.__cond_resched.down_read.acct_collect.do_exit.do_group_exit
1.80 ±104% +2305.8% 43.38 ± 67% perf-sched.sch_delay.max.ms.__cond_resched.down_read.walk_component.link_path_walk.part
19.72 ± 46% +391.9% 97.03 ± 42% perf-sched.sch_delay.max.ms.__cond_resched.down_write.__mmap_new_vma.__mmap_region.do_mmap
222.08 ± 11% +5794.4% 13090 ±108% perf-sched.sch_delay.max.ms.__cond_resched.down_write.anon_vma_clone.anon_vma_fork.dup_mmap
274.25 ± 5% +5170.7% 14454 ±111% perf-sched.sch_delay.max.ms.__cond_resched.down_write.dup_mmap.dup_mm.constprop
73.44 ± 48% +113.6% 156.88 ± 21% perf-sched.sch_delay.max.ms.__cond_resched.down_write.unlink_file_vma_batch_final.free_pgtables.exit_mmap
23.85 ± 50% +249.4% 83.33 ± 34% perf-sched.sch_delay.max.ms.__cond_resched.down_write.vma_link_file.__mmap_new_vma.__mmap_region
92.73 ± 29% +71.5% 159.00 ± 24% perf-sched.sch_delay.max.ms.__cond_resched.down_write.vms_gather_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap
4.07 ± 36% +430.7% 21.62 ± 42% perf-sched.sch_delay.max.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
5.62 ±121% +1304.8% 78.91 ± 39% perf-sched.sch_delay.max.ms.__cond_resched.down_write_killable.vm_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.01 ±140% +5492.8% 56.59 ± 42% perf-sched.sch_delay.max.ms.__cond_resched.dput.step_into.link_path_walk.part
0.75 ±163% +5258.4% 40.24 ± 77% perf-sched.sch_delay.max.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
34.61 ± 78% +215.9% 109.35 ± 30% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_lru_noprof.__d_alloc.d_alloc_pseudo.alloc_file_pseudo
44.21 ± 76% +252.4% 155.83 ± 36% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.dup_task_struct.copy_process.kernel_clone
1.78 ± 95% +591.1% 12.30 ± 60% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.__khugepaged_enter.do_huge_pmd_anonymous_page.__handle_mm_fault
35.91 ± 77% +159.0% 93.00 ± 31% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.alloc_file_pseudo.__shmem_file_setup
3.16 ±154% +1390.5% 47.14 ± 53% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.path_openat.do_filp_open
330.41 ± 24% +4614.7% 15577 ±102% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
0.75 ±223% +4317.2% 33.33 ±110% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.getname_flags.part.0
37.23 ± 75% +142.7% 90.36 ± 33% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.mas_alloc_nodes.mas_preallocate.__mmap_new_vma
145.93 ± 24% +9241.3% 13631 ±108% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.prepare_creds.copy_creds.copy_process
43.53 ± 79% +166.1% 115.84 ± 31% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
3.38 ±203% +1420.1% 51.46 ± 60% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.__split_vma.vms_gather_munmap_vmas
224.58 ± 18% +681.9% 1755 ±105% perf-sched.sch_delay.max.ms.__cond_resched.mutex_lock.futex_exit_release.exit_mm_release.exit_mm
4.15 ±223% +2488.3% 107.50 ± 30% perf-sched.sch_delay.max.ms.__cond_resched.mutex_lock.perf_event_exit_task.do_exit.do_group_exit
411.45 ±112% +3129.7% 13288 ±108% perf-sched.sch_delay.max.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
428.93 ± 16% +707.9% 3465 ± 50% perf-sched.sch_delay.max.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
40.64 ± 80% +240.6% 138.43 ± 54% perf-sched.sch_delay.max.ms.__cond_resched.shmem_inode_unacct_blocks.shmem_undo_range.shmem_evict_inode.evict
459.95 ± 14% +804.7% 4161 ± 57% perf-sched.sch_delay.max.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
91.90 ± 34% +659.5% 697.98 ±177% perf-sched.sch_delay.max.ms.__cond_resched.switch_task_namespaces.do_exit.do_group_exit.__x64_sys_exit_group
45.16 ± 95% +176.9% 125.05 ± 37% perf-sched.sch_delay.max.ms.__cond_resched.task_work_run.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
56.92 ± 47% +144.0% 138.85 ± 30% perf-sched.sch_delay.max.ms.__cond_resched.unmap_page_range.unmap_vmas.exit_mmap.__mmput
319.32 ± 23% +445.2% 1740 ± 63% perf-sched.sch_delay.max.ms.__cond_resched.zap_pmd_range.isra.0.unmap_page_range
348.62 ± 21% +721.6% 2864 ± 53% perf-sched.sch_delay.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
27.43 ± 75% +1687.1% 490.13 ±145% perf-sched.sch_delay.max.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
1363 ± 60% +1094.4% 16281 ±102% perf-sched.sch_delay.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
317.08 ± 29% +578.7% 2151 ± 82% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
451.54 ± 16% +831.0% 4203 ± 54% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
142.47 ± 23% +50.7% 214.69 ± 26% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
298.70 ± 20% +754.7% 2553 ± 53% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
717.11 ± 59% +1529.8% 11687 ±127% perf-sched.sch_delay.max.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
39.28 ± 45% +302.2% 158.00 ±122% perf-sched.sch_delay.max.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
17.73 ±112% +561.1% 117.22 ± 58% perf-sched.sch_delay.max.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.unlink_file_vma_batch_final
58.28 ± 53% +2240.3% 1363 ± 71% perf-sched.sch_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
47.17 ± 95% +15545.3% 7379 ±104% perf-sched.sch_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
314.10 ± 16% +4088.0% 13154 ±100% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
36.28 ± 78% +123.3% 81.01 ± 32% perf-sched.sch_delay.max.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
35.75 ± 9% +58.8% 56.76 ± 4% perf-sched.total_sch_delay.average.ms
1769 ± 70% +860.7% 17004 ± 96% perf-sched.total_sch_delay.max.ms
81.84 ± 8% +61.4% 132.10 ± 9% perf-sched.total_wait_and_delay.average.ms
446898 ± 7% +98.4% 886795 ± 17% perf-sched.total_wait_and_delay.count.ms
3840 ± 62% +804.8% 34745 ± 92% perf-sched.total_wait_and_delay.max.ms
46.09 ± 6% +63.5% 75.34 ± 13% perf-sched.total_wait_time.average.ms
3062 ± 25% +607.8% 21677 ± 69% perf-sched.total_wait_time.max.ms
32.02 ±141% +823.3% 295.64 ± 79% perf-sched.wait_and_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pmd_alloc
81.86 ± 10% +41.9% 116.16 ± 9% perf-sched.wait_and_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
28.49 ±141% +696.0% 226.79 ± 53% perf-sched.wait_and_delay.avg.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
9.67 ± 17% +170.4% 26.14 ± 56% perf-sched.wait_and_delay.avg.ms.__cond_resched.folio_zero_user.vma_alloc_anon_folio_pmd.__do_huge_pmd_anonymous_page.__handle_mm_fault
80.77 ± 79% +507.2% 490.43 ±109% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_pid.copy_process.kernel_clone
14.65 ±223% +1421.7% 222.99 ± 55% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
42.27 ±100% +374.5% 200.59 ± 42% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.dup_mmap.dup_mm
31.22 ±141% +755.9% 267.22 ± 63% perf-sched.wait_and_delay.avg.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
384.40 ± 40% -75.2% 95.38 ±145% perf-sched.wait_and_delay.avg.ms.__cond_resched.process_one_work.worker_thread.kthread.ret_from_fork
96.32 ± 10% +24.3% 119.73 ± 13% perf-sched.wait_and_delay.avg.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
90.77 ± 8% +50.2% 136.29 ± 11% perf-sched.wait_and_delay.avg.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
37.70 ± 7% +544.1% 242.86 ± 52% perf-sched.wait_and_delay.avg.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.15 ± 4% +572.3% 0.98 ± 56% perf-sched.wait_and_delay.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
133.70 ± 9% +250.2% 468.27 ± 64% perf-sched.wait_and_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
571.39 ± 7% +162.7% 1501 ± 69% perf-sched.wait_and_delay.avg.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
457.45 +55.6% 711.90 ± 45% perf-sched.wait_and_delay.avg.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
114.14 ± 4% +453.2% 631.37 ± 45% perf-sched.wait_and_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
83.38 ± 10% +86.2% 155.25 ± 39% perf-sched.wait_and_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
70.48 ± 2% +337.6% 308.40 ± 49% perf-sched.wait_and_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
29.17 ±142% +1403.4% 438.50 ± 28% perf-sched.wait_and_delay.count.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pmd_alloc
5278 ± 11% -88.4% 610.67 ±141% perf-sched.wait_and_delay.count.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
304.50 ±142% +923.2% 3115 ± 18% perf-sched.wait_and_delay.count.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
3084 ± 17% -83.9% 495.67 ±142% perf-sched.wait_and_delay.count.__cond_resched.exit_mmap.__mmput.exit_mm.do_exit
5591 ± 21% +132.7% 13007 ± 16% perf-sched.wait_and_delay.count.__cond_resched.folio_zero_user.vma_alloc_anon_folio_pmd.__do_huge_pmd_anonymous_page.__handle_mm_fault
286.67 ±223% +1574.3% 4799 ± 17% perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
192.50 ±100% +938.4% 1998 ± 18% perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.dup_mmap.dup_mm
97.50 ±142% +386.3% 474.17 ± 15% perf-sched.wait_and_delay.count.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
483.83 ± 26% -55.7% 214.50 ± 46% perf-sched.wait_and_delay.count.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.mm_init.dup_mm
31620 ± 11% +1464.8% 494787 ± 16% perf-sched.wait_and_delay.count.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
74608 ± 7% -60.6% 29368 ± 25% perf-sched.wait_and_delay.count.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
690.50 ± 17% -59.6% 279.17 ±102% perf-sched.wait_and_delay.count.__cond_resched.zap_pmd_range.isra.0.unmap_page_range
1929 ± 20% -59.0% 790.50 ±104% perf-sched.wait_and_delay.count.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
34736 ± 7% -27.1% 25322 ± 17% perf-sched.wait_and_delay.count.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
11.00 ± 9% +87.9% 20.67 ± 17% perf-sched.wait_and_delay.count.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
24.17 ± 9% +98.6% 48.00 ± 15% perf-sched.wait_and_delay.count.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
13.50 ± 9% +300.0% 54.00 ±102% perf-sched.wait_and_delay.count.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
43.50 ± 8% +139.8% 104.33 ± 16% perf-sched.wait_and_delay.count.schedule_timeout.kcompactd.kthread.ret_from_fork
20294 ± 11% -31.3% 13937 ± 19% perf-sched.wait_and_delay.count.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
762.80 ±145% +2865.9% 22623 ±136% perf-sched.wait_and_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pmd_alloc
891.25 ± 20% +511.4% 5448 ± 47% perf-sched.wait_and_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
967.82 ± 16% +556.8% 6356 ± 57% perf-sched.wait_and_delay.max.ms.__cond_resched.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff
724.81 ± 14% +848.3% 6873 ± 59% perf-sched.wait_and_delay.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range.zap_pmd_range
772.30 ±145% +3714.2% 29456 ± 96% perf-sched.wait_and_delay.max.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
762.95 ±223% +4012.3% 31374 ±101% perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
1069 ±132% +2709.8% 30059 ± 96% perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.dup_mmap.dup_mm
512.94 ±188% +5166.8% 27015 ±107% perf-sched.wait_and_delay.max.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
857.86 ± 16% +679.2% 6684 ± 55% perf-sched.wait_and_delay.max.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
919.89 ± 14% +764.9% 7956 ± 59% perf-sched.wait_and_delay.max.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
914.96 ± 12% +2199.9% 21043 ± 88% perf-sched.wait_and_delay.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
321.23 ± 19% +594.3% 2230 ± 74% perf-sched.wait_and_delay.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
2838 ± 53% +1052.5% 32709 ±102% perf-sched.wait_and_delay.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
903.08 ± 16% +1170.0% 11469 ± 67% perf-sched.wait_and_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
1727 ± 17% +1533.4% 28221 ± 97% perf-sched.wait_and_delay.max.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
1050 +1308.0% 14792 ±118% perf-sched.wait_and_delay.max.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
1567 ± 8% +1368.0% 23016 ± 87% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
1730 ± 53% +1443.6% 26709 ± 97% perf-sched.wait_and_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
2200 ± 22% +1164.5% 27827 ±105% perf-sched.wait_and_delay.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
53.28 ± 19% +207.9% 164.05 ± 72% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.__pmd_alloc
35.30 ± 12% +208.9% 109.06 ± 48% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pte_alloc_one
23.94 ±107% +265.6% 87.53 ± 42% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.allocate_slab.___slab_alloc
40.22 ± 10% +42.5% 57.29 ± 9% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
1.32 ±206% +2168.2% 29.87 ± 74% perf-sched.wait_time.avg.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
38.38 ± 10% +59.5% 61.24 ± 23% perf-sched.wait_time.avg.ms.__cond_resched.__put_anon_vma.unlink_anon_vmas.free_pgtables.exit_mmap
45.35 ± 9% +45.5% 65.96 ± 24% perf-sched.wait_time.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
4.21 ± 15% +280.3% 16.00 ± 91% perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
49.28 ± 9% +132.5% 114.60 ± 61% perf-sched.wait_time.avg.ms.__cond_resched.copy_page_range.dup_mmap.dup_mm.constprop
49.10 ± 7% +154.3% 124.87 ± 53% perf-sched.wait_time.avg.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
0.47 ±107% +5995.1% 28.58 ± 39% perf-sched.wait_time.avg.ms.__cond_resched.down_read.__mm_populate.vm_mmap_pgoff.do_syscall_64
0.60 ±177% +5535.0% 33.81 ± 48% perf-sched.wait_time.avg.ms.__cond_resched.down_write.__mmap_new_vma.__mmap_region.do_mmap
56.26 ± 18% +95.3% 109.85 ± 54% perf-sched.wait_time.avg.ms.__cond_resched.down_write.anon_vma_clone.anon_vma_fork.dup_mmap
45.90 ± 12% +198.5% 137.01 ± 62% perf-sched.wait_time.avg.ms.__cond_resched.down_write.anon_vma_fork.dup_mmap.dup_mm
50.84 ± 9% +114.6% 109.07 ± 45% perf-sched.wait_time.avg.ms.__cond_resched.down_write.dup_mmap.dup_mm.constprop
40.23 ± 7% +60.7% 64.65 ± 22% perf-sched.wait_time.avg.ms.__cond_resched.down_write.free_pgtables.exit_mmap.__mmput
39.00 ± 8% +57.8% 61.55 ± 14% perf-sched.wait_time.avg.ms.__cond_resched.down_write.unlink_anon_vmas.free_pgtables.exit_mmap
37.73 ± 12% +56.6% 59.08 ± 12% perf-sched.wait_time.avg.ms.__cond_resched.down_write.unlink_file_vma_batch_add.free_pgtables.exit_mmap
1.15 ±196% +2575.9% 30.79 ± 44% perf-sched.wait_time.avg.ms.__cond_resched.down_write.vma_link_file.__mmap_new_vma.__mmap_region
0.19 ± 64% +303.8% 0.76 ± 28% perf-sched.wait_time.avg.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
38.18 ± 25% +60.7% 61.36 ± 15% perf-sched.wait_time.avg.ms.__cond_resched.dput.path_put.exit_fs.do_exit
0.46 ±140% +1957.9% 9.55 ± 63% perf-sched.wait_time.avg.ms.__cond_resched.dput.step_into.link_path_walk.part
0.14 ±202% +7445.0% 10.61 ± 94% perf-sched.wait_time.avg.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
43.32 ± 7% +44.6% 62.65 ± 18% perf-sched.wait_time.avg.ms.__cond_resched.exit_mmap.__mmput.exit_mm.do_exit
0.75 ± 30% +1444.8% 11.62 ± 63% perf-sched.wait_time.avg.ms.__cond_resched.folio_zero_user.vma_alloc_anon_folio_pmd.__do_huge_pmd_anonymous_page.__handle_mm_fault
49.62 ± 32% -77.7% 11.05 ± 56% perf-sched.wait_time.avg.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
0.48 ± 95% +6560.8% 32.08 ± 40% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_lru_noprof.__d_alloc.d_alloc_pseudo.alloc_file_pseudo
4.89 ±174% +564.9% 32.54 ± 43% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_lru_noprof.shmem_alloc_inode.alloc_inode.new_inode
2.54 ±147% +1068.9% 29.69 ± 41% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.alloc_file_pseudo.__shmem_file_setup
0.59 ±138% +2094.7% 12.94 ± 81% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.path_openat.do_filp_open
50.54 ± 9% +142.8% 122.73 ± 56% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.anon_vma_fork.dup_mmap.dup_mm
28.71 ± 37% +576.9% 194.34 ±165% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.copy_signal.copy_process.kernel_clone
0.52 ±130% +6185.7% 32.63 ± 36% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.mas_alloc_nodes.mas_preallocate.__mmap_new_vma
41.66 ± 25% +150.9% 104.53 ± 22% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.prepare_creds.copy_creds.copy_process
1.48 ±147% +1847.1% 28.80 ± 35% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.security_inode_alloc.inode_init_always_gfp.alloc_inode
0.81 ± 48% +3501.5% 29.20 ± 42% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
0.31 ±172% +7598.9% 24.05 ± 81% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.__split_vma.vms_gather_munmap_vmas
50.76 ± 8% +111.9% 107.57 ± 40% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.dup_mmap.dup_mm
38.89 ± 11% +79.6% 69.87 ± 36% perf-sched.wait_time.avg.ms.__cond_resched.mutex_lock.futex_exit_release.exit_mm_release.exit_mm
4.15 ±223% +1063.1% 48.31 ± 33% perf-sched.wait_time.avg.ms.__cond_resched.mutex_lock.perf_event_exit_task.do_exit.do_group_exit
51.54 ± 18% +171.9% 140.12 ± 61% perf-sched.wait_time.avg.ms.__cond_resched.mutex_lock_killable.pcpu_alloc_noprof.__percpu_counter_init_many.mm_init
384.38 ± 40% -67.3% 125.66 ± 96% perf-sched.wait_time.avg.ms.__cond_resched.process_one_work.worker_thread.kthread.ret_from_fork
35.52 ± 15% +67.0% 59.33 ± 26% perf-sched.wait_time.avg.ms.__cond_resched.remove_vma.exit_mmap.__mmput.exit_mm
47.53 ± 10% +24.5% 59.17 ± 13% perf-sched.wait_time.avg.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
45.37 ± 8% +46.5% 66.48 ± 11% perf-sched.wait_time.avg.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
37.68 ± 7% +541.9% 241.86 ± 52% perf-sched.wait_time.avg.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
42.58 ± 30% +283.5% 163.28 ±140% perf-sched.wait_time.avg.ms.__cond_resched.switch_task_namespaces.do_exit.do_group_exit.__x64_sys_exit_group
5.03 ±115% +947.4% 52.66 ± 29% perf-sched.wait_time.avg.ms.__cond_resched.task_work_run.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
31.31 ± 18% +82.2% 57.03 ± 9% perf-sched.wait_time.avg.ms.__cond_resched.wp_page_copy.__handle_mm_fault.handle_mm_fault.do_user_addr_fault
46.45 ± 6% +29.1% 59.95 ± 7% perf-sched.wait_time.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
0.15 ± 4% +572.3% 0.98 ± 56% perf-sched.wait_time.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
122.43 ± 9% +187.4% 351.84 ± 46% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
1.72 ± 52% +3352.8% 59.24 ± 60% perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.filemap_fault.__do_fault
32.83 ± 9% +63.7% 53.76 ± 10% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
12.22 ± 20% +183.3% 34.61 ± 33% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
30.44 ± 18% +35.2% 41.15 ± 13% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
563.63 ± 7% +163.0% 1482 ± 70% perf-sched.wait_time.avg.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
38.02 ± 23% +151.3% 95.54 ± 46% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.__put_anon_vma
45.48 ± 12% +83.7% 83.53 ± 26% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.anon_vma_fork
46.77 ± 14% +134.7% 109.78 ± 67% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.dup_mmap
14.35 ± 96% +344.7% 63.80 ± 23% perf-sched.wait_time.avg.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.unlink_file_vma_batch_final
28.53 ± 31% +752.9% 243.34 ±109% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
457.42 +55.6% 711.84 ± 45% perf-sched.wait_time.avg.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
5.01 ± 3% +105.7% 10.31 ± 60% perf-sched.wait_time.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
114.07 ± 4% +452.5% 630.20 ± 45% perf-sched.wait_time.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
45.42 ± 8% +87.8% 85.28 ± 40% perf-sched.wait_time.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
2.09 ± 77% +416.8% 10.82 ± 56% perf-sched.wait_time.avg.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
70.30 ± 2% +335.3% 306.05 ± 49% perf-sched.wait_time.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
974.31 ± 50% +1374.0% 14360 ±104% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.alloc_pages_noprof.pte_alloc_one
445.63 ± 20% +511.4% 2724 ± 47% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
99.05 ± 17% +72.4% 170.81 ± 24% perf-sched.wait_time.max.ms.__cond_resched.__dentry_kill.dput.__fput.task_work_run
483.91 ± 16% +556.8% 3178 ± 57% perf-sched.wait_time.max.ms.__cond_resched.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff
5.26 ±206% +1212.1% 69.05 ± 45% perf-sched.wait_time.max.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
35.48 ±172% +15060.6% 5378 ±205% perf-sched.wait_time.max.ms.__cond_resched.__kmalloc_cache_noprof.vmstat_start.seq_read_iter.proc_reg_read_iter
362.40 ± 14% +848.3% 3436 ± 59% perf-sched.wait_time.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range.zap_pmd_range
1433 ± 86% +864.9% 13835 ±102% perf-sched.wait_time.max.ms.__cond_resched.copy_page_range.dup_mmap.dup_mm.constprop
1520 ± 41% +969.1% 16259 ± 90% perf-sched.wait_time.max.ms.__cond_resched.copy_pte_range.copy_p4d_range.copy_page_range.dup_mmap
6.37 ± 99% +1535.9% 104.18 ± 31% perf-sched.wait_time.max.ms.__cond_resched.down_read.__mm_populate.vm_mmap_pgoff.do_syscall_64
104.84 ± 33% +577.7% 710.49 ±173% perf-sched.wait_time.max.ms.__cond_resched.down_read.acct_collect.do_exit.do_group_exit
3.60 ±177% +2595.0% 97.03 ± 42% perf-sched.wait_time.max.ms.__cond_resched.down_write.__mmap_new_vma.__mmap_region.do_mmap
1546 ± 50% +953.2% 16284 ±100% perf-sched.wait_time.max.ms.__cond_resched.down_write.dup_mmap.dup_mm.constprop
73.44 ± 48% +113.6% 156.88 ± 21% perf-sched.wait_time.max.ms.__cond_resched.down_write.unlink_file_vma_batch_final.free_pgtables.exit_mmap
5.75 ±183% +1348.7% 83.33 ± 34% perf-sched.wait_time.max.ms.__cond_resched.down_write.vma_link_file.__mmap_new_vma.__mmap_region
92.73 ± 29% +71.5% 159.00 ± 24% perf-sched.wait_time.max.ms.__cond_resched.down_write.vms_gather_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap
3.00 ± 54% +545.4% 19.38 ± 52% perf-sched.wait_time.max.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
1.00 ±143% +5554.3% 56.59 ± 42% perf-sched.wait_time.max.ms.__cond_resched.dput.step_into.link_path_walk.part
0.27 ±210% +13877.5% 38.00 ± 86% perf-sched.wait_time.max.ms.__cond_resched.dput.terminate_walk.path_openat.do_filp_open
1935 ± 67% -87.7% 238.19 ±173% perf-sched.wait_time.max.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
6.59 ±123% +1558.4% 109.35 ± 30% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_lru_noprof.__d_alloc.d_alloc_pseudo.alloc_file_pseudo
44.21 ± 76% +269.4% 163.33 ± 33% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.dup_task_struct.copy_process.kernel_clone
26.78 ±122% +247.2% 93.00 ± 31% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.alloc_file_pseudo.__shmem_file_setup
0.89 ±150% +4891.0% 44.44 ± 64% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.alloc_empty_file.path_openat.do_filp_open
0.75 ±223% +2718.3% 21.27 ±180% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.getname_flags.part.0
4.17 ±127% +2067.1% 90.36 ± 33% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.mas_alloc_nodes.mas_preallocate.__mmap_new_vma
228.57 ± 45% +5935.1% 13794 ±106% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.prepare_creds.copy_creds.copy_process
13.81 ± 36% +738.7% 115.84 ± 31% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
0.49 ±137% +10424.4% 51.41 ± 60% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_dup.__split_vma.vms_gather_munmap_vmas
224.58 ± 18% +604.8% 1582 ±108% perf-sched.wait_time.max.ms.__cond_resched.mutex_lock.futex_exit_release.exit_mm_release.exit_mm
4.15 ±223% +2488.3% 107.50 ± 30% perf-sched.wait_time.max.ms.__cond_resched.mutex_lock.perf_event_exit_task.do_exit.do_group_exit
428.93 ± 16% +679.2% 3342 ± 55% perf-sched.wait_time.max.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
40.64 ± 80% +240.6% 138.43 ± 54% perf-sched.wait_time.max.ms.__cond_resched.shmem_inode_unacct_blocks.shmem_undo_range.shmem_evict_inode.evict
459.95 ± 14% +764.9% 3978 ± 59% perf-sched.wait_time.max.ms.__cond_resched.shmem_undo_range.shmem_evict_inode.evict.__dentry_kill
914.95 ± 12% +1849.7% 17838 ± 86% perf-sched.wait_time.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
91.90 ± 34% +659.5% 697.98 ±177% perf-sched.wait_time.max.ms.__cond_resched.switch_task_namespaces.do_exit.do_group_exit.__x64_sys_exit_group
24.81 ±116% +384.4% 120.21 ± 37% perf-sched.wait_time.max.ms.__cond_resched.task_work_run.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
270.14 ± 19% +2674.2% 7494 ±140% perf-sched.wait_time.max.ms.__cond_resched.task_work_run.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe
56.92 ± 47% +144.0% 138.85 ± 30% perf-sched.wait_time.max.ms.__cond_resched.unmap_page_range.unmap_vmas.exit_mmap.__mmput
549.17 ± 71% +379.4% 2632 ± 89% perf-sched.wait_time.max.ms.__cond_resched.wp_page_copy.__handle_mm_fault.handle_mm_fault.do_user_addr_fault
319.32 ± 23% +361.3% 1472 ± 62% perf-sched.wait_time.max.ms.__cond_resched.zap_pmd_range.isra.0.unmap_page_range
481.60 ± 52% +448.7% 2642 ± 55% perf-sched.wait_time.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
321.23 ± 19% +594.3% 2230 ± 74% perf-sched.wait_time.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
2179 ± 38% +675.4% 16899 ± 97% perf-sched.wait_time.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
317.08 ± 29% +599.1% 2216 ± 85% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
451.54 ± 16% +1653.3% 7916 ±102% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
277.33 ± 11% +804.4% 2508 ± 56% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
1523 ± 12% +1227.5% 20227 ± 69% perf-sched.wait_time.max.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
1025 +1338.8% 14757 ±118% perf-sched.wait_time.max.ms.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
645.52 ± 69% +473.8% 3703 ±157% perf-sched.wait_time.max.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.dup_mmap
17.73 ±112% +561.1% 117.22 ± 58% perf-sched.wait_time.max.ms.schedule_preempt_disabled.rwsem_down_write_slowpath.down_write.unlink_file_vma_batch_final
242.29 ± 40% +4359.2% 10804 ±131% perf-sched.wait_time.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
54.15 ± 66% +14372.4% 7836 ±106% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
1567 ± 8% +1116.1% 19067 ± 84% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
1711 ± 54% +841.3% 16107 ± 85% perf-sched.wait_time.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
32.68 ± 96% +147.9% 81.01 ± 32% perf-sched.wait_time.max.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
2200 ± 22% +728.3% 18226 ± 83% perf-sched.wait_time.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
***************************************************************************************************
lkp-icl-2sp2: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
=========================================================================================
compiler/cpufreq_governor/ipc/iterations/kconfig/mode/nr_threads/rootfs/tbox_group/testcase:
gcc-12/performance/socket/4/x86_64-rhel-9.4/threads/100%/debian-12-x86_64-20240206.cgz/lkp-icl-2sp2/hackbench
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
2931589 ± 14% -29.8% 2059056 ± 12% cpuidle..usage
274766 ± 4% -17.8% 225885 ± 13% numa-meminfo.node0.SUnreclaim
200.50 +26.2% 253.04 ± 2% uptime.boot
1992153 +114.3% 4268815 ± 4% vmstat.system.cs
1.88 ± 16% -0.5 1.36 ± 8% mpstat.cpu.all.idle%
0.02 ± 3% +0.0 0.02 ± 4% mpstat.cpu.all.soft%
128556 +31.8% 169412 meminfo.AnonHugePages
485113 -15.1% 411686 meminfo.SUnreclaim
614406 -12.0% 540564 meminfo.Slab
538737 ± 13% +114.5% 1155509 ± 13% numa-numastat.node0.local_node
615940 ± 15% +95.7% 1205124 ± 11% numa-numastat.node0.numa_hit
1008187 ± 19% +45.5% 1466818 ± 5% numa-numastat.node1.local_node
1063633 ± 18% +45.7% 1549639 ± 4% numa-numastat.node1.numa_hit
68502 ± 4% -17.6% 56417 ± 14% numa-vmstat.node0.nr_slab_unreclaimable
615163 ± 15% +95.9% 1204843 ± 11% numa-vmstat.node0.numa_hit
537960 ± 13% +114.7% 1155228 ± 13% numa-vmstat.node0.numa_local
1062574 ± 18% +45.8% 1548926 ± 4% numa-vmstat.node1.numa_hit
1007129 ± 19% +45.6% 1466105 ± 5% numa-vmstat.node1.numa_local
40153 ± 27% -83.8% 6498 ± 24% perf-c2c.DRAM.local
5474 ± 28% -45.5% 2981 ± 20% perf-c2c.DRAM.remote
73336 ± 26% -74.2% 18885 ± 20% perf-c2c.HITM.local
1539 ± 39% -42.5% 884.83 ± 26% perf-c2c.HITM.remote
74875 ± 26% -73.6% 19769 ± 21% perf-c2c.HITM.total
121167 -15.0% 102975 proc-vmstat.nr_slab_unreclaimable
1682677 ± 9% +63.9% 2757090 ± 5% proc-vmstat.numa_hit
1550029 ± 10% +69.3% 2624651 ± 5% proc-vmstat.numa_local
3252894 ± 7% +209.5% 10068916 ± 7% proc-vmstat.pgalloc_normal
2648888 ± 9% +262.9% 9612330 ± 8% proc-vmstat.pgfree
415614 -26.6% 304910 hackbench.throughput
405261 -25.9% 300194 hackbench.throughput_avg
415614 -26.6% 304910 hackbench.throughput_best
386952 -24.9% 290681 hackbench.throughput_worst
149.14 +34.8% 201.02 hackbench.time.elapsed_time
149.14 +34.8% 201.02 hackbench.time.elapsed_time.max
58196111 ± 6% +286.5% 2.249e+08 ± 7% hackbench.time.involuntary_context_switches
134003 ± 5% -16.3% 112130 hackbench.time.minor_page_faults
17596 +35.8% 23894 hackbench.time.system_time
1136 +28.7% 1463 hackbench.time.user_time
2.372e+08 +167.5% 6.346e+08 ± 3% hackbench.time.voluntary_context_switches
1.42 -34.8% 0.92 ± 13% perf-stat.i.MPKI
4.477e+10 -17.2% 3.707e+10 perf-stat.i.branch-instructions
0.41 +0.1 0.50 perf-stat.i.branch-miss-rate%
1.744e+08 +3.0% 1.796e+08 perf-stat.i.branch-misses
23.68 -8.5 15.20 ± 7% perf-stat.i.cache-miss-rate%
3.098e+08 -46.4% 1.661e+08 ± 11% perf-stat.i.cache-misses
1.318e+09 -16.2% 1.105e+09 ± 4% perf-stat.i.cache-references
1972433 +116.9% 4278376 ± 4% perf-stat.i.context-switches
1.48 +22.1% 1.81 perf-stat.i.cpi
3.239e+11 +1.3% 3.283e+11 perf-stat.i.cpu-cycles
47350 ± 13% +65.3% 78248 ± 19% perf-stat.i.cpu-migrations
1064 +91.4% 2037 ± 12% perf-stat.i.cycles-between-cache-misses
2.186e+11 -16.9% 1.816e+11 perf-stat.i.instructions
0.68 -18.0% 0.56 perf-stat.i.ipc
15.69 +115.3% 33.79 ± 3% perf-stat.i.metric.K/sec
4400 ± 6% -18.0% 3607 perf-stat.i.minor-faults
4400 ± 6% -18.0% 3607 perf-stat.i.page-faults
1.42 -35.3% 0.92 ± 13% perf-stat.overall.MPKI
0.39 +0.1 0.48 perf-stat.overall.branch-miss-rate%
23.53 -8.5 15.00 ± 7% perf-stat.overall.cache-miss-rate%
1.48 +21.9% 1.81 perf-stat.overall.cpi
1046 +91.4% 2004 ± 11% perf-stat.overall.cycles-between-cache-misses
0.67 -17.9% 0.55 perf-stat.overall.ipc
4.448e+10 -17.0% 3.691e+10 perf-stat.ps.branch-instructions
1.73e+08 +3.2% 1.786e+08 perf-stat.ps.branch-misses
3.075e+08 -46.3% 1.653e+08 ± 11% perf-stat.ps.cache-misses
1.307e+09 -15.9% 1.099e+09 ± 4% perf-stat.ps.cache-references
1958509 +116.8% 4245688 ± 4% perf-stat.ps.context-switches
3.219e+11 +1.5% 3.266e+11 perf-stat.ps.cpu-cycles
46201 ± 13% +68.6% 77874 ± 19% perf-stat.ps.cpu-migrations
2.172e+11 -16.7% 1.808e+11 perf-stat.ps.instructions
4287 ± 6% -17.1% 3552 perf-stat.ps.minor-faults
4287 ± 6% -17.1% 3552 perf-stat.ps.page-faults
3.263e+13 +12.0% 3.653e+13 perf-stat.total.instructions
7909525 ± 2% +52.6% 12071802 sched_debug.cfs_rq:/.avg_vruntime.avg
10481951 ± 9% +61.8% 16959043 ± 11% sched_debug.cfs_rq:/.avg_vruntime.max
7141207 +52.0% 10853360 ± 2% sched_debug.cfs_rq:/.avg_vruntime.min
22.46 +13.0% 25.38 sched_debug.cfs_rq:/.h_nr_queued.avg
5.93 ± 5% +19.3% 7.08 ± 3% sched_debug.cfs_rq:/.h_nr_queued.stddev
22.20 +14.1% 25.33 sched_debug.cfs_rq:/.h_nr_runnable.avg
6.03 ± 5% +17.9% 7.11 ± 3% sched_debug.cfs_rq:/.h_nr_runnable.stddev
355.83 -22.8% 274.75 sched_debug.cfs_rq:/.load_avg.max
7909525 ± 2% +52.6% 12071802 sched_debug.cfs_rq:/.min_vruntime.avg
10481951 ± 9% +61.8% 16959043 ± 11% sched_debug.cfs_rq:/.min_vruntime.max
7141207 +52.0% 10853360 ± 2% sched_debug.cfs_rq:/.min_vruntime.min
0.69 +11.8% 0.78 sched_debug.cfs_rq:/.nr_queued.avg
0.44 ± 35% +59.4% 0.71 ± 13% sched_debug.cfs_rq:/.nr_queued.min
0.12 ± 17% -28.0% 0.08 ± 14% sched_debug.cfs_rq:/.nr_queued.stddev
341.39 -25.0% 256.00 sched_debug.cfs_rq:/.removed.load_avg.max
174.00 -24.7% 131.00 ± 2% sched_debug.cfs_rq:/.removed.runnable_avg.max
174.00 -24.7% 131.00 ± 2% sched_debug.cfs_rq:/.removed.util_avg.max
198071 ±125% +348.4% 888081 ± 62% sched_debug.cfs_rq:/.runnable_avg.avg
1977052 ±141% +219.2% 6310648 ± 51% sched_debug.cfs_rq:/.runnable_avg.stddev
1871 +20.4% 2253 sched_debug.cfs_rq:/.util_est.avg
577483 ± 2% -12.5% 505023 ± 8% sched_debug.cpu.avg_idle.avg
113708 +27.4% 144846 ± 2% sched_debug.cpu.clock.avg
115037 +26.7% 145757 ± 2% sched_debug.cpu.clock.max
112235 +28.2% 143890 ± 2% sched_debug.cpu.clock.min
113310 +27.3% 144279 ± 2% sched_debug.cpu.clock_task.avg
114742 +26.6% 145320 ± 2% sched_debug.cpu.clock_task.max
103500 +30.2% 134733 ± 2% sched_debug.cpu.clock_task.min
12338 +15.5% 14254 sched_debug.cpu.curr->pid.avg
15248 +12.6% 17174 sched_debug.cpu.curr->pid.max
22.46 +13.0% 25.37 sched_debug.cpu.nr_running.avg
5.94 ± 5% +19.1% 7.07 ± 3% sched_debug.cpu.nr_running.stddev
944460 ± 2% +212.8% 2954732 ± 3% sched_debug.cpu.nr_switches.avg
1384690 ± 9% +173.0% 3780143 ± 3% sched_debug.cpu.nr_switches.max
779137 ± 2% +198.7% 2327029 ± 9% sched_debug.cpu.nr_switches.min
94939 ± 18% +274.7% 355703 ± 41% sched_debug.cpu.nr_switches.stddev
112191 +28.2% 143862 ± 2% sched_debug.cpu_clk
111036 +28.5% 142706 ± 2% sched_debug.ktime
113021 +28.0% 144688 ± 2% sched_debug.sched_clk
55.02 ± 69% +1310.2% 775.94 ± 89% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof
38.12 ± 77% +943.0% 397.61 ± 59% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
65.04 ± 17% +653.0% 489.75 ± 28% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
20.05 ± 24% +288.4% 77.88 ± 34% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
10.24 ± 95% -99.5% 0.05 ±191% perf-sched.sch_delay.avg.ms.__cond_resched.dput.__fput.__x64_sys_close.do_syscall_64
47.60 ± 34% +735.2% 397.53 ± 42% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
19.81 ±109% -75.7% 4.81 ± 60% perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
55.61 ± 32% +1002.8% 613.30 ± 20% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
0.37 ±217% +2.4e+05% 895.65 ±222% perf-sched.sch_delay.avg.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
27.14 ± 39% +347.6% 121.45 ± 42% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
66.58 ± 18% +1020.4% 745.92 ± 18% perf-sched.sch_delay.avg.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
5.48 ± 17% +213.6% 17.19 ± 16% perf-sched.sch_delay.avg.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
8.61 ± 22% +385.3% 41.78 ± 10% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
330.55 ±116% +2270.5% 7835 ± 55% perf-sched.sch_delay.max.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
5574 ± 16% +211.6% 17370 ± 8% perf-sched.sch_delay.max.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
17.26 ± 88% -99.7% 0.05 ±191% perf-sched.sch_delay.max.ms.__cond_resched.dput.__fput.__x64_sys_close.do_syscall_64
5162 ± 15% +236.2% 17354 ± 12% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
4727 ± 24% +239.1% 16028 ± 13% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
209.42 ±202% +3332.2% 7187 ±117% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
1343 ± 39% +152.0% 3384 ± 42% perf-sched.sch_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
5830 ± 16% +212.2% 18205 ± 8% perf-sched.sch_delay.max.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
5836 ± 15% +202.9% 17679 ± 8% perf-sched.sch_delay.max.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
5816 ± 15% +212.2% 18162 ± 8% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
10.88 ± 19% +232.3% 36.16 ± 11% perf-sched.total_sch_delay.average.ms
5942 ± 13% +206.8% 18228 ± 8% perf-sched.total_sch_delay.max.ms
27.55 ± 19% +232.7% 91.65 ± 12% perf-sched.total_wait_and_delay.average.ms
11902 ± 13% +206.3% 36457 ± 8% perf-sched.total_wait_and_delay.max.ms
16.67 ± 19% +232.9% 55.49 ± 12% perf-sched.total_wait_time.average.ms
6456 ± 4% +182.5% 18242 ± 8% perf-sched.total_wait_time.max.ms
113.95 ± 69% +1316.6% 1614 ± 82% perf-sched.wait_and_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof
79.96 ± 73% +1082.3% 945.34 ± 57% perf-sched.wait_and_delay.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
143.75 ± 16% +611.5% 1022 ± 27% perf-sched.wait_and_delay.avg.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
42.41 ± 20% +275.0% 159.04 ± 32% perf-sched.wait_and_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
101.84 ± 33% +721.9% 837.03 ± 41% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
124.93 ± 34% +929.0% 1285 ± 20% perf-sched.wait_and_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
74.56 ± 60% +705.1% 600.29 ± 41% perf-sched.wait_and_delay.avg.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
76.07 ± 36% +330.8% 327.70 ± 26% perf-sched.wait_and_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
146.24 ± 18% +989.6% 1593 ± 18% perf-sched.wait_and_delay.avg.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
16.40 ± 18% +229.2% 54.00 ± 16% perf-sched.wait_and_delay.avg.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
739.69 ± 13% +254.3% 2621 ± 44% perf-sched.wait_and_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
19.70 ± 23% +353.8% 89.38 ± 9% perf-sched.wait_and_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
837.21 ± 24% +166.9% 2234 ± 13% perf-sched.wait_and_delay.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
36.50 ± 86% -96.3% 1.33 ±223% perf-sched.wait_and_delay.count.__cond_resched.__dentry_kill.shrink_dentry_list.shrink_dcache_parent.d_invalidate
132.83 ± 58% -60.0% 53.17 ± 27% perf-sched.wait_and_delay.count.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
101624 ± 38% -81.9% 18404 ± 17% perf-sched.wait_and_delay.count.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
733.10 ± 64% -84.4% 114.45 ±223% perf-sched.wait_and_delay.max.ms.__cond_resched.__dentry_kill.shrink_dentry_list.shrink_dcache_parent.d_invalidate
661.77 ±116% +2433.4% 16765 ± 49% perf-sched.wait_and_delay.max.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
11149 ± 16% +211.6% 34740 ± 8% perf-sched.wait_and_delay.max.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
10324 ± 15% +236.2% 34709 ± 12% perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
9455 ± 24% +239.0% 32056 ± 13% perf-sched.wait_and_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
646.51 ± 59% +503.5% 3902 ± 68% perf-sched.wait_and_delay.max.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
2745 ± 39% +175.6% 7565 ± 33% perf-sched.wait_and_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
11661 ± 16% +212.2% 36412 ± 8% perf-sched.wait_and_delay.max.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
11674 ± 15% +203.0% 35373 ± 8% perf-sched.wait_and_delay.max.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
6384 ± 4% +184.5% 18161 ± 7% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
11633 ± 15% +212.2% 36324 ± 8% perf-sched.wait_and_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
6098 ± 9% +192.4% 17831 ± 7% perf-sched.wait_and_delay.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
58.92 ± 68% +1322.7% 838.26 ± 76% perf-sched.wait_time.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof
41.83 ± 69% +1209.3% 547.73 ± 62% perf-sched.wait_time.avg.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
78.71 ± 15% +577.2% 533.02 ± 26% perf-sched.wait_time.avg.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
22.36 ± 17% +263.0% 81.16 ± 29% perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
8.84 ±109% -99.5% 0.05 ±191% perf-sched.wait_time.avg.ms.__cond_resched.dput.__fput.__x64_sys_close.do_syscall_64
54.24 ± 33% +710.2% 439.49 ± 40% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
69.31 ± 35% +869.9% 672.25 ± 21% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
51.67 ± 62% +784.1% 456.81 ± 40% perf-sched.wait_time.avg.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
48.93 ± 35% +321.5% 206.25 ± 19% perf-sched.wait_time.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
79.66 ± 18% +963.8% 847.45 ± 18% perf-sched.wait_time.avg.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
10.92 ± 19% +237.0% 36.81 ± 16% perf-sched.wait_time.avg.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
739.68 ± 13% +254.3% 2621 ± 44% perf-sched.wait_time.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
11.09 ± 23% +329.3% 47.60 ± 8% perf-sched.wait_time.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
833.25 ± 24% +136.5% 1970 ± 7% perf-sched.wait_time.avg.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
331.23 ±116% +2838.5% 9733 ± 52% perf-sched.wait_time.max.ms.__cond_resched.__kmalloc_node_noprof.alloc_slab_obj_exts.allocate_slab.___slab_alloc
5575 ± 16% +212.8% 17440 ± 8% perf-sched.wait_time.max.ms.__cond_resched.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
11.46 ± 79% -99.6% 0.05 ±191% perf-sched.wait_time.max.ms.__cond_resched.dput.__fput.__x64_sys_close.do_syscall_64
5162 ± 15% +236.2% 17354 ± 12% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
4850 ± 25% +230.4% 16028 ± 13% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown].[unknown]
216.39 ±194% +3253.6% 7256 ±115% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown].[unknown]
386.99 ± 62% +582.5% 2641 ± 54% perf-sched.wait_time.max.ms.rcu_gp_kthread.kthread.ret_from_fork.ret_from_fork_asm
1402 ± 39% +198.2% 4181 ± 27% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
5830 ± 16% +212.3% 18206 ± 8% perf-sched.wait_time.max.ms.schedule_timeout.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
5840 ± 15% +204.1% 17759 ± 8% perf-sched.wait_time.max.ms.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
6384 ± 4% +184.5% 18161 ± 7% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
5818 ± 15% +212.2% 18162 ± 8% perf-sched.wait_time.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
6098 ± 9% +192.4% 17831 ± 7% perf-sched.wait_time.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
55.29 ± 2% -16.0 39.24 perf-profile.calltrace.cycles-pp.read
50.80 ± 2% -14.9 35.91 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.read
47.07 ± 3% -14.9 32.19 perf-profile.calltrace.cycles-pp.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
50.53 ± 2% -14.8 35.71 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
48.63 ± 3% -14.8 33.86 perf-profile.calltrace.cycles-pp.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
45.80 ± 3% -14.6 31.18 perf-profile.calltrace.cycles-pp.sock_read_iter.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
44.98 ± 3% -14.5 30.52 perf-profile.calltrace.cycles-pp.sock_recvmsg.sock_read_iter.vfs_read.ksys_read.do_syscall_64
44.54 ± 3% -14.4 30.17 perf-profile.calltrace.cycles-pp.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.vfs_read.ksys_read
44.16 ± 3% -14.3 29.87 perf-profile.calltrace.cycles-pp.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter.vfs_read
13.28 ± 10% -6.2 7.10 ± 2% perf-profile.calltrace.cycles-pp.kmem_cache_free.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
16.20 ± 8% -5.9 10.30 ± 2% perf-profile.calltrace.cycles-pp.consume_skb.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
12.86 ± 11% -5.2 7.64 ± 3% perf-profile.calltrace.cycles-pp.skb_release_data.consume_skb.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
11.84 ± 11% -4.7 7.19 ± 3% perf-profile.calltrace.cycles-pp.kfree.skb_release_data.consume_skb.unix_stream_read_generic.unix_stream_recvmsg
7.28 ± 23% -4.6 2.64 ± 5% perf-profile.calltrace.cycles-pp.__put_partials.kmem_cache_free.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
7.06 ± 23% -4.5 2.53 ± 6% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__put_partials.kmem_cache_free.unix_stream_read_generic.unix_stream_recvmsg
6.92 ± 23% -4.5 2.45 ± 6% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__put_partials.kmem_cache_free.unix_stream_read_generic
7.38 ± 21% -4.1 3.27 ± 8% perf-profile.calltrace.cycles-pp.__put_partials.kfree.skb_release_data.consume_skb.unix_stream_read_generic
7.17 ± 21% -4.0 3.16 ± 8% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.__put_partials.kfree.skb_release_data.consume_skb
7.03 ± 22% -3.9 3.08 ± 8% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.__put_partials.kfree.skb_release_data
5.59 ± 20% -3.4 2.22 ± 4% perf-profile.calltrace.cycles-pp.___slab_alloc.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
7.82 ± 2% -3.4 4.45 ± 2% perf-profile.calltrace.cycles-pp.unix_stream_read_actor.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
7.75 ± 2% -3.4 4.40 ± 2% perf-profile.calltrace.cycles-pp.skb_copy_datagram_iter.unix_stream_read_actor.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
7.60 ± 2% -3.3 4.28 ± 2% perf-profile.calltrace.cycles-pp.__skb_datagram_iter.skb_copy_datagram_iter.unix_stream_read_actor.unix_stream_read_generic.unix_stream_recvmsg
4.94 ± 22% -3.2 1.70 ± 5% perf-profile.calltrace.cycles-pp.get_partial_node.___slab_alloc.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags
4.53 ± 23% -3.0 1.51 ± 5% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.get_partial_node.___slab_alloc.kmem_cache_alloc_node_noprof.__alloc_skb
4.50 ± 23% -3.0 1.50 ± 5% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.get_partial_node.___slab_alloc.kmem_cache_alloc_node_noprof
5.74 ± 19% -2.9 2.80 ± 7% perf-profile.calltrace.cycles-pp.___slab_alloc.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
5.05 ± 21% -2.9 2.19 ± 7% perf-profile.calltrace.cycles-pp.get_partial_node.___slab_alloc.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb
4.64 ± 22% -2.7 1.99 ± 8% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.get_partial_node.___slab_alloc.__kmalloc_node_track_caller_noprof.kmalloc_reserve
4.61 ± 22% -2.6 1.96 ± 8% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.get_partial_node.___slab_alloc.__kmalloc_node_track_caller_noprof
3.91 ± 3% -1.8 2.12 ± 2% perf-profile.calltrace.cycles-pp._copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.unix_stream_read_actor.unix_stream_read_generic
3.03 ± 3% -1.2 1.79 perf-profile.calltrace.cycles-pp.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.unix_stream_read_actor.unix_stream_read_generic
2.85 ± 3% -1.2 1.64 perf-profile.calltrace.cycles-pp.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.unix_stream_read_actor
2.15 ± 4% -1.0 1.11 ± 2% perf-profile.calltrace.cycles-pp.check_heap_object.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter
2.25 ± 14% -0.7 1.57 ± 3% perf-profile.calltrace.cycles-pp._raw_spin_lock.unix_stream_sendmsg.sock_write_iter.vfs_write.ksys_write
3.24 ± 4% -0.6 2.59 perf-profile.calltrace.cycles-pp.skb_release_head_state.consume_skb.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
2.67 ± 3% -0.6 2.03 ± 2% perf-profile.calltrace.cycles-pp.skb_copy_datagram_from_iter.unix_stream_sendmsg.sock_write_iter.vfs_write.ksys_write
3.10 ± 4% -0.6 2.48 perf-profile.calltrace.cycles-pp.unix_destruct_scm.skb_release_head_state.consume_skb.unix_stream_read_generic.unix_stream_recvmsg
1.53 ± 2% -0.6 0.92 ± 28% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.write
1.52 ± 3% -0.6 0.92 ± 28% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.read
2.90 ± 4% -0.6 2.32 perf-profile.calltrace.cycles-pp.sock_wfree.unix_destruct_scm.skb_release_head_state.consume_skb.unix_stream_read_generic
2.98 ± 4% -0.6 2.43 ± 11% perf-profile.calltrace.cycles-pp.__memcg_slab_free_hook.kmem_cache_free.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
1.97 ± 3% -0.5 1.48 ± 2% perf-profile.calltrace.cycles-pp.clear_bhb_loop.write
1.96 ± 2% -0.4 1.53 ± 2% perf-profile.calltrace.cycles-pp.clear_bhb_loop.read
1.32 ± 5% -0.4 0.95 ± 2% perf-profile.calltrace.cycles-pp.__slab_free.kfree.skb_release_data.consume_skb.unix_stream_read_generic
1.23 ± 3% -0.3 0.91 ± 2% perf-profile.calltrace.cycles-pp._copy_from_iter.skb_copy_datagram_from_iter.unix_stream_sendmsg.sock_write_iter.vfs_write
1.11 ± 3% -0.3 0.84 ± 2% perf-profile.calltrace.cycles-pp.__check_object_size.skb_copy_datagram_from_iter.unix_stream_sendmsg.sock_write_iter.vfs_write
0.60 ± 3% +0.1 0.69 perf-profile.calltrace.cycles-pp.skb_unlink.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
0.61 ± 5% +0.1 0.71 ± 2% perf-profile.calltrace.cycles-pp.mutex_lock.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
1.09 ± 5% +0.4 1.47 ± 8% perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
0.43 ±100% +0.5 0.91 ± 5% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.skb_queue_tail.unix_stream_sendmsg.sock_write_iter.vfs_write
0.29 ±100% +0.5 0.78 ± 5% perf-profile.calltrace.cycles-pp.pick_next_task_fair.__pick_next_task.__schedule.schedule.schedule_timeout
0.30 ±100% +0.5 0.80 ± 4% perf-profile.calltrace.cycles-pp.__pick_next_task.__schedule.schedule.schedule_timeout.unix_stream_data_wait
0.00 +0.5 0.51 perf-profile.calltrace.cycles-pp._raw_spin_lock.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
0.08 ±223% +0.5 0.60 perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.skb_unlink.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
0.00 +0.6 0.56 ± 6% perf-profile.calltrace.cycles-pp.dequeue_entity.dequeue_entities.dequeue_task_fair.try_to_block_task.__schedule
0.00 +0.6 0.63 ± 12% perf-profile.calltrace.cycles-pp.__schedule.schedule.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.00 +0.7 0.65 ± 12% perf-profile.calltrace.cycles-pp.schedule.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
0.17 ±141% +0.7 0.83 ± 4% perf-profile.calltrace.cycles-pp.enqueue_entity.enqueue_task_fair.enqueue_task.ttwu_do_activate.try_to_wake_up
0.00 +0.7 0.67 ± 2% perf-profile.calltrace.cycles-pp.mutex_unlock.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
0.78 ± 27% +0.7 1.45 ± 5% perf-profile.calltrace.cycles-pp.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key
0.35 ±100% +0.7 1.04 ± 4% perf-profile.calltrace.cycles-pp.dequeue_entities.dequeue_task_fair.try_to_block_task.__schedule.schedule
0.36 ±100% +0.7 1.08 ± 3% perf-profile.calltrace.cycles-pp.dequeue_task_fair.try_to_block_task.__schedule.schedule.schedule_timeout
0.38 ±100% +0.7 1.12 ± 3% perf-profile.calltrace.cycles-pp.try_to_block_task.__schedule.schedule.schedule_timeout.unix_stream_data_wait
0.39 ±100% +0.8 1.15 ± 4% perf-profile.calltrace.cycles-pp.enqueue_task_fair.enqueue_task.ttwu_do_activate.try_to_wake_up.autoremove_wake_function
0.40 ±100% +0.8 1.18 ± 4% perf-profile.calltrace.cycles-pp.enqueue_task.ttwu_do_activate.try_to_wake_up.autoremove_wake_function.__wake_up_common
1.68 ± 46% +1.1 2.81 ± 8% perf-profile.calltrace.cycles-pp.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg.sock_write_iter.vfs_write
1.44 ± 42% +1.2 2.63 ± 7% perf-profile.calltrace.cycles-pp.try_to_wake_up.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.sock_def_readable
1.51 ± 42% +1.2 2.70 ± 7% perf-profile.calltrace.cycles-pp.__wake_up_common.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg.sock_write_iter
1.47 ± 42% +1.2 2.67 ± 7% perf-profile.calltrace.cycles-pp.autoremove_wake_function.__wake_up_common.__wake_up_sync_key.sock_def_readable.unix_stream_sendmsg
1.59 ± 29% +1.3 2.89 ± 3% perf-profile.calltrace.cycles-pp.__schedule.schedule.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic
2.29 ± 45% +1.3 3.61 ± 7% perf-profile.calltrace.cycles-pp.sock_def_readable.unix_stream_sendmsg.sock_write_iter.vfs_write.ksys_write
1.63 ± 29% +1.3 2.96 ± 3% perf-profile.calltrace.cycles-pp.schedule.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg
1.68 ± 28% +1.4 3.02 ± 3% perf-profile.calltrace.cycles-pp.schedule_timeout.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg
1.96 ± 29% +1.7 3.65 ± 3% perf-profile.calltrace.cycles-pp.unix_stream_data_wait.unix_stream_read_generic.unix_stream_recvmsg.sock_recvmsg.sock_read_iter
0.00 +4.8 4.85 ± 2% perf-profile.calltrace.cycles-pp.page_counter_try_charge.try_charge_memcg.obj_cgroup_charge.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof
8.90 ± 12% +5.0 13.87 ± 2% perf-profile.calltrace.cycles-pp.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg
2.16 ± 3% +8.6 10.80 ± 2% perf-profile.calltrace.cycles-pp.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
0.00 +8.7 8.70 ± 2% perf-profile.calltrace.cycles-pp.try_charge_memcg.obj_cgroup_charge.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof.__alloc_skb
0.44 ± 44% +8.8 9.20 ± 2% perf-profile.calltrace.cycles-pp.obj_cgroup_charge.__memcg_slab_post_alloc_hook.kmem_cache_alloc_node_noprof.__alloc_skb.alloc_skb_with_frags
0.00 +9.5 9.52 ± 2% perf-profile.calltrace.cycles-pp.page_counter_try_charge.try_charge_memcg.obj_cgroup_charge.__memcg_slab_post_alloc_hook.__kmalloc_node_track_caller_noprof
9.94 ± 10% +13.6 23.50 perf-profile.calltrace.cycles-pp.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg
9.38 ± 11% +13.8 23.19 perf-profile.calltrace.cycles-pp.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb
43.95 +15.4 59.35 perf-profile.calltrace.cycles-pp.write
37.71 +16.6 54.29 perf-profile.calltrace.cycles-pp.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
35.73 +16.6 52.34 perf-profile.calltrace.cycles-pp.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
39.75 +16.7 56.48 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.write
39.48 +16.8 56.28 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
2.49 ± 3% +16.9 19.37 ± 2% perf-profile.calltrace.cycles-pp.__memcg_slab_post_alloc_hook.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb.alloc_skb_with_frags
34.47 +16.9 51.41 perf-profile.calltrace.cycles-pp.sock_write_iter.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.70 ± 2% +17.1 17.82 ± 2% perf-profile.calltrace.cycles-pp.obj_cgroup_charge.__memcg_slab_post_alloc_hook.__kmalloc_node_track_caller_noprof.kmalloc_reserve.__alloc_skb
0.00 +17.1 17.13 ± 2% perf-profile.calltrace.cycles-pp.try_charge_memcg.obj_cgroup_charge.__memcg_slab_post_alloc_hook.__kmalloc_node_track_caller_noprof.kmalloc_reserve
33.13 +17.3 50.43 perf-profile.calltrace.cycles-pp.unix_stream_sendmsg.sock_write_iter.vfs_write.ksys_write.do_syscall_64
23.11 ± 7% +17.7 40.82 perf-profile.calltrace.cycles-pp.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter.vfs_write.ksys_write
20.87 ± 9% +17.9 38.78 perf-profile.calltrace.cycles-pp.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter.vfs_write
20.58 ± 9% +18.0 38.56 perf-profile.calltrace.cycles-pp.__alloc_skb.alloc_skb_with_frags.sock_alloc_send_pskb.unix_stream_sendmsg.sock_write_iter
55.43 ± 2% -15.6 39.81 perf-profile.children.cycles-pp.read
47.13 ± 3% -14.9 32.23 perf-profile.children.cycles-pp.vfs_read
48.70 ± 3% -14.8 33.92 perf-profile.children.cycles-pp.ksys_read
45.85 ± 3% -14.6 31.21 perf-profile.children.cycles-pp.sock_read_iter
45.04 ± 3% -14.5 30.57 perf-profile.children.cycles-pp.sock_recvmsg
44.58 ± 3% -14.4 30.20 perf-profile.children.cycles-pp.unix_stream_recvmsg
44.36 ± 3% -14.3 30.02 perf-profile.children.cycles-pp.unix_stream_read_generic
23.88 ± 19% -14.1 9.79 ± 5% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
25.27 ± 18% -13.7 11.62 ± 4% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
14.68 ± 21% -8.8 5.92 ± 6% perf-profile.children.cycles-pp.__put_partials
11.33 ± 19% -6.3 5.03 ± 5% perf-profile.children.cycles-pp.___slab_alloc
13.34 ± 10% -6.2 7.14 ± 2% perf-profile.children.cycles-pp.kmem_cache_free
16.27 ± 8% -5.9 10.35 ± 2% perf-profile.children.cycles-pp.consume_skb
10.21 ± 21% -5.9 4.35 ± 6% perf-profile.children.cycles-pp.get_partial_node
12.90 ± 11% -5.2 7.66 ± 3% perf-profile.children.cycles-pp.skb_release_data
11.90 ± 11% -4.7 7.23 ± 3% perf-profile.children.cycles-pp.kfree
7.86 ± 2% -3.4 4.48 ± 2% perf-profile.children.cycles-pp.unix_stream_read_actor
7.78 ± 2% -3.4 4.42 ± 2% perf-profile.children.cycles-pp.skb_copy_datagram_iter
7.65 ± 2% -3.3 4.32 ± 2% perf-profile.children.cycles-pp.__skb_datagram_iter
3.93 ± 3% -1.8 2.14 ± 2% perf-profile.children.cycles-pp._copy_to_iter
4.23 ± 3% -1.5 2.70 perf-profile.children.cycles-pp.__check_object_size
3.07 ± 3% -1.3 1.82 perf-profile.children.cycles-pp.simple_copy_to_iter
2.82 ± 4% -1.2 1.62 perf-profile.children.cycles-pp.check_heap_object
3.96 ± 2% -0.9 3.04 ± 2% perf-profile.children.cycles-pp.clear_bhb_loop
2.75 ± 3% -0.7 2.08 ± 2% perf-profile.children.cycles-pp.skb_copy_datagram_from_iter
3.28 ± 4% -0.7 2.62 perf-profile.children.cycles-pp.skb_release_head_state
3.16 ± 4% -0.6 2.53 perf-profile.children.cycles-pp.unix_destruct_scm
2.93 ± 4% -0.6 2.34 perf-profile.children.cycles-pp.sock_wfree
2.40 ± 5% -0.5 1.94 perf-profile.children.cycles-pp.__slab_free
1.74 ± 2% -0.4 1.35 ± 2% perf-profile.children.cycles-pp.entry_SYSCALL_64
1.34 ± 2% -0.3 1.01 ± 2% perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
1.25 ± 3% -0.3 0.93 ± 2% perf-profile.children.cycles-pp._copy_from_iter
1.08 ± 3% -0.3 0.79 ± 3% perf-profile.children.cycles-pp.rw_verify_area
0.62 ± 3% -0.3 0.36 perf-profile.children.cycles-pp.__build_skb_around
0.73 ± 3% -0.2 0.55 ± 2% perf-profile.children.cycles-pp.__check_heap_object
0.63 ± 3% -0.1 0.48 perf-profile.children.cycles-pp.__cond_resched
0.40 ± 5% -0.1 0.28 ± 5% perf-profile.children.cycles-pp.fsnotify_pre_content
0.20 ± 6% -0.1 0.10 ± 4% perf-profile.children.cycles-pp.put_cpu_partial
0.41 ± 3% -0.1 0.31 perf-profile.children.cycles-pp.x64_sys_call
0.53 ± 3% -0.1 0.43 ± 4% perf-profile.children.cycles-pp.__virt_addr_valid
0.61 ± 4% -0.1 0.52 perf-profile.children.cycles-pp.syscall_return_via_sysret
0.27 ± 2% -0.1 0.20 ± 2% perf-profile.children.cycles-pp.rcu_all_qs
0.28 ± 3% -0.1 0.22 ± 2% perf-profile.children.cycles-pp.__scm_recv_common
0.23 ± 3% -0.1 0.17 ± 2% perf-profile.children.cycles-pp.security_file_permission
0.30 ± 3% -0.0 0.25 ± 3% perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
0.21 ± 5% -0.0 0.16 ± 6% perf-profile.children.cycles-pp.kmalloc_size_roundup
0.16 ± 5% -0.0 0.12 ± 4% perf-profile.children.cycles-pp.maybe_add_creds
0.16 ± 4% -0.0 0.12 ± 4% perf-profile.children.cycles-pp.is_vmalloc_addr
0.17 ± 4% -0.0 0.13 perf-profile.children.cycles-pp.put_pid
0.15 ± 2% -0.0 0.11 perf-profile.children.cycles-pp.security_socket_recvmsg
0.14 ± 3% -0.0 0.10 perf-profile.children.cycles-pp.security_socket_getpeersec_dgram
0.21 ± 5% -0.0 0.17 ± 4% perf-profile.children.cycles-pp.syscall_exit_to_user_mode_prepare
0.13 ± 3% -0.0 0.10 ± 3% perf-profile.children.cycles-pp.security_socket_sendmsg
0.18 ± 2% -0.0 0.15 ± 3% perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
0.14 ± 3% -0.0 0.10 ± 4% perf-profile.children.cycles-pp.manage_oob
0.17 ± 4% -0.0 0.14 ± 2% perf-profile.children.cycles-pp.check_stack_object
0.11 ± 3% -0.0 0.08 perf-profile.children.cycles-pp.wait_for_unix_gc
0.17 ± 3% -0.0 0.14 ± 2% perf-profile.children.cycles-pp.unix_scm_to_skb
0.13 ± 5% -0.0 0.10 ± 7% perf-profile.children.cycles-pp.cgroup_rstat_updated
0.09 ± 5% -0.0 0.07 perf-profile.children.cycles-pp.skb_put
0.07 -0.0 0.05 perf-profile.children.cycles-pp.__x64_sys_read
0.07 -0.0 0.05 perf-profile.children.cycles-pp.__x64_sys_write
0.09 ± 5% -0.0 0.08 ± 6% perf-profile.children.cycles-pp.skb_free_head
0.07 ± 6% -0.0 0.06 ± 6% perf-profile.children.cycles-pp.kfree_skbmem
0.08 ± 21% +0.0 0.10 ± 7% perf-profile.children.cycles-pp.__get_user_8
0.07 ± 27% +0.0 0.10 ± 4% perf-profile.children.cycles-pp.avg_vruntime
0.09 ± 23% +0.0 0.13 ± 7% perf-profile.children.cycles-pp.rseq_get_rseq_cs
0.08 ± 16% +0.0 0.12 ± 6% perf-profile.children.cycles-pp.os_xsave
0.07 ± 25% +0.0 0.11 ± 6% perf-profile.children.cycles-pp.sched_clock
0.08 ± 35% +0.0 0.12 ± 3% perf-profile.children.cycles-pp.place_entity
0.02 ± 99% +0.0 0.07 ± 8% perf-profile.children.cycles-pp.__put_user_8
0.02 ± 99% +0.0 0.07 ± 8% perf-profile.children.cycles-pp.__wrgsbase_inactive
0.06 ± 19% +0.0 0.11 ± 10% perf-profile.children.cycles-pp.___perf_sw_event
0.08 ± 29% +0.0 0.12 ± 7% perf-profile.children.cycles-pp.sched_clock_cpu
0.25 ± 9% +0.0 0.30 ± 5% perf-profile.children.cycles-pp.restore_fpregs_from_fpstate
0.01 ±223% +0.0 0.06 ± 11% perf-profile.children.cycles-pp.ktime_get
0.04 ±100% +0.1 0.09 ± 5% perf-profile.children.cycles-pp.update_entity_lag
0.01 ±223% +0.1 0.06 ± 7% perf-profile.children.cycles-pp.update_curr_dl_se
0.01 ±223% +0.1 0.06 ± 11% perf-profile.children.cycles-pp.clockevents_program_event
0.04 ±100% +0.1 0.10 ± 5% perf-profile.children.cycles-pp.native_sched_clock
0.09 ± 31% +0.1 0.14 ± 8% perf-profile.children.cycles-pp.update_rq_clock
0.00 +0.1 0.06 ± 9% perf-profile.children.cycles-pp.__rb_insert_augmented
0.06 ± 14% +0.1 0.13 ± 9% perf-profile.children.cycles-pp.vruntime_eligible
0.08 ± 26% +0.1 0.14 ± 10% perf-profile.children.cycles-pp.put_prev_entity
0.04 ±101% +0.1 0.11 ± 6% perf-profile.children.cycles-pp.finish_wait
0.00 +0.1 0.07 ± 12% perf-profile.children.cycles-pp.charge_memcg
0.10 ± 18% +0.1 0.17 ± 9% perf-profile.children.cycles-pp.check_preempt_wakeup_fair
0.04 ± 72% +0.1 0.12 ± 4% perf-profile.children.cycles-pp.rseq_update_cpu_node_id
0.00 +0.1 0.08 ± 14% perf-profile.children.cycles-pp.__mem_cgroup_charge
0.64 ± 3% +0.1 0.72 perf-profile.children.cycles-pp.skb_unlink
0.00 +0.1 0.08 ± 10% perf-profile.children.cycles-pp.set_next_buddy
0.00 +0.1 0.08 ± 13% perf-profile.children.cycles-pp.shmem_alloc_and_add_folio
0.00 +0.1 0.09 ± 15% perf-profile.children.cycles-pp.shmem_get_folio_gfp
0.00 +0.1 0.09 ± 15% perf-profile.children.cycles-pp.shmem_write_begin
0.65 ± 5% +0.1 0.74 perf-profile.children.cycles-pp.mutex_lock
0.14 ± 31% +0.1 0.22 ± 7% perf-profile.children.cycles-pp.wakeup_preempt
0.05 +0.1 0.14 ± 10% perf-profile.children.cycles-pp.cmd_record
0.17 ± 22% +0.1 0.26 ± 5% perf-profile.children.cycles-pp.__update_load_avg_se
0.05 ± 7% +0.1 0.15 ± 9% perf-profile.children.cycles-pp.handle_internal_command
0.05 ± 7% +0.1 0.15 ± 9% perf-profile.children.cycles-pp.main
0.14 ± 22% +0.1 0.24 ± 6% perf-profile.children.cycles-pp.rseq_ip_fixup
0.05 ± 7% +0.1 0.15 ± 9% perf-profile.children.cycles-pp.run_builtin
0.13 ± 25% +0.1 0.23 ± 9% perf-profile.children.cycles-pp.__update_load_avg_cfs_rq
0.31 ± 13% +0.1 0.42 ± 5% perf-profile.children.cycles-pp.switch_fpu_return
0.00 +0.1 0.11 ± 11% perf-profile.children.cycles-pp.generic_perform_write
0.19 ± 11% +0.1 0.30 ± 7% perf-profile.children.cycles-pp.__dequeue_entity
0.00 +0.1 0.11 ± 13% perf-profile.children.cycles-pp.shmem_file_write_iter
0.05 ±100% +0.1 0.16 ± 5% perf-profile.children.cycles-pp.update_rq_clock_task
0.00 +0.1 0.12 ± 10% perf-profile.children.cycles-pp.record__pushfn
0.00 +0.1 0.12 ± 12% perf-profile.children.cycles-pp.writen
0.00 +0.1 0.14 ± 8% perf-profile.children.cycles-pp.perf_mmap__push
0.18 ± 15% +0.1 0.32 ± 6% perf-profile.children.cycles-pp.__enqueue_entity
0.16 ± 22% +0.1 0.30 ± 4% perf-profile.children.cycles-pp.prepare_task_switch
0.00 +0.1 0.14 ± 9% perf-profile.children.cycles-pp.record__mmap_read_evlist
0.14 ± 21% +0.1 0.27 ± 5% perf-profile.children.cycles-pp.__switch_to
0.21 ± 23% +0.2 0.37 ± 6% perf-profile.children.cycles-pp.__rseq_handle_notify_resume
0.35 ± 19% +0.2 0.52 ± 9% perf-profile.children.cycles-pp.__switch_to_asm
0.29 ± 15% +0.2 0.47 ± 6% perf-profile.children.cycles-pp.set_next_entity
0.00 +0.2 0.18 ± 4% perf-profile.children.cycles-pp.page_counter_cancel
0.00 +0.2 0.19 ± 4% perf-profile.children.cycles-pp.page_counter_uncharge
0.00 +0.2 0.19 ± 3% perf-profile.children.cycles-pp.drain_stock
0.15 ± 16% +0.2 0.36 ± 8% perf-profile.children.cycles-pp.pick_eevdf
0.22 ± 25% +0.2 0.45 ± 7% perf-profile.children.cycles-pp.pick_task_fair
0.34 ± 31% +0.2 0.59 ± 6% perf-profile.children.cycles-pp.dequeue_entity
0.21 ± 44% +0.3 0.47 ± 11% perf-profile.children.cycles-pp.get_any_partial
0.47 ± 28% +0.3 0.74 ± 7% perf-profile.children.cycles-pp.update_load_avg
0.00 +0.3 0.29 ± 3% perf-profile.children.cycles-pp.refill_stock
0.39 ± 6% +0.3 0.68 perf-profile.children.cycles-pp.mutex_unlock
0.37 ± 34% +0.3 0.68 ± 8% perf-profile.children.cycles-pp.update_curr
0.69 ± 6% +0.3 1.03 perf-profile.children.cycles-pp.fput
0.48 ± 25% +0.4 0.86 ± 4% perf-profile.children.cycles-pp.enqueue_entity
0.39 ± 4% +0.4 0.82 ± 28% perf-profile.children.cycles-pp.obj_cgroup_uncharge_pages
2.39 ± 9% +0.4 2.82 ± 5% perf-profile.children.cycles-pp.syscall_exit_to_user_mode
0.61 ± 32% +0.5 1.07 ± 4% perf-profile.children.cycles-pp.dequeue_entities
0.72 ± 31% +0.5 1.18 ± 4% perf-profile.children.cycles-pp.enqueue_task_fair
0.76 ± 31% +0.5 1.22 ± 5% perf-profile.children.cycles-pp.enqueue_task
0.72 ± 25% +0.5 1.19 ± 6% perf-profile.children.cycles-pp.pick_next_task_fair
0.74 ± 25% +0.5 1.21 ± 6% perf-profile.children.cycles-pp.__pick_next_task
0.62 ± 31% +0.5 1.10 ± 4% perf-profile.children.cycles-pp.dequeue_task_fair
0.64 ± 29% +0.5 1.14 ± 4% perf-profile.children.cycles-pp.try_to_block_task
0.32 ± 8% +0.5 0.84 ± 53% perf-profile.children.cycles-pp.__mod_memcg_lruvec_state
0.29 ± 6% +0.5 0.82 ± 62% perf-profile.children.cycles-pp.__mod_memcg_state
0.90 ± 30% +0.6 1.48 ± 5% perf-profile.children.cycles-pp.ttwu_do_activate
0.00 +0.7 0.67 ± 11% perf-profile.children.cycles-pp.propagate_protected_usage
1.58 ± 41% +1.1 2.68 ± 7% perf-profile.children.cycles-pp.try_to_wake_up
1.65 ± 41% +1.1 2.74 ± 8% perf-profile.children.cycles-pp.__wake_up_common
1.60 ± 41% +1.1 2.71 ± 7% perf-profile.children.cycles-pp.autoremove_wake_function
1.87 ± 28% +1.2 3.08 ± 3% perf-profile.children.cycles-pp.schedule_timeout
2.31 ± 44% +1.3 3.63 ± 7% perf-profile.children.cycles-pp.sock_def_readable
2.11 ± 29% +1.5 3.65 ± 4% perf-profile.children.cycles-pp.__schedule
2.10 ± 28% +1.6 3.69 ± 4% perf-profile.children.cycles-pp.schedule
1.96 ± 29% +1.7 3.67 ± 3% perf-profile.children.cycles-pp.unix_stream_data_wait
90.76 +1.9 92.65 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
90.24 +2.0 92.23 perf-profile.children.cycles-pp.do_syscall_64
8.98 ± 12% +5.0 13.93 ± 2% perf-profile.children.cycles-pp.kmem_cache_alloc_node_noprof
10.02 ± 10% +13.5 23.56 perf-profile.children.cycles-pp.kmalloc_reserve
9.60 ± 10% +13.6 23.25 perf-profile.children.cycles-pp.__kmalloc_node_track_caller_noprof
0.00 +14.4 14.44 ± 2% perf-profile.children.cycles-pp.page_counter_try_charge
44.12 +15.9 60.04 perf-profile.children.cycles-pp.write
37.81 +16.6 54.46 perf-profile.children.cycles-pp.ksys_write
35.82 +16.7 52.50 perf-profile.children.cycles-pp.vfs_write
34.55 +16.9 51.46 perf-profile.children.cycles-pp.sock_write_iter
33.40 +17.2 50.62 perf-profile.children.cycles-pp.unix_stream_sendmsg
23.17 ± 7% +17.7 40.86 perf-profile.children.cycles-pp.sock_alloc_send_pskb
20.93 ± 9% +17.9 38.82 perf-profile.children.cycles-pp.alloc_skb_with_frags
20.68 ± 9% +17.9 38.63 perf-profile.children.cycles-pp.__alloc_skb
4.77 ± 3% +25.5 30.26 ± 2% perf-profile.children.cycles-pp.__memcg_slab_post_alloc_hook
1.29 ± 3% +25.8 27.07 ± 2% perf-profile.children.cycles-pp.obj_cgroup_charge
0.13 ± 3% +25.8 25.92 ± 2% perf-profile.children.cycles-pp.try_charge_memcg
23.87 ± 19% -14.1 9.78 ± 5% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
3.89 ± 3% -1.8 2.12 ± 2% perf-profile.self.cycles-pp._copy_to_iter
3.22 ± 6% -1.3 1.92 ± 2% perf-profile.self.cycles-pp.__memcg_slab_free_hook
2.44 ± 3% -1.1 1.39 ± 2% perf-profile.self.cycles-pp.unix_stream_read_generic
2.13 ± 4% -1.1 1.07 perf-profile.self.cycles-pp.check_heap_object
3.92 ± 2% -0.9 3.00 ± 2% perf-profile.self.cycles-pp.clear_bhb_loop
1.89 ± 4% -0.9 1.02 perf-profile.self.cycles-pp.kmem_cache_free
2.29 ± 3% -0.6 1.66 perf-profile.self.cycles-pp.__memcg_slab_post_alloc_hook
2.34 ± 3% -0.6 1.78 perf-profile.self.cycles-pp.sock_wfree
0.92 ± 6% -0.5 0.38 ± 3% perf-profile.self.cycles-pp.skb_release_data
1.46 ± 16% -0.5 0.97 ± 5% perf-profile.self.cycles-pp.unix_stream_sendmsg
1.11 ± 4% -0.5 0.63 perf-profile.self.cycles-pp.___slab_alloc
2.35 ± 5% -0.4 1.90 perf-profile.self.cycles-pp.__slab_free
0.76 ± 10% -0.4 0.38 ± 5% perf-profile.self.cycles-pp.get_partial_node
1.28 ± 3% -0.4 0.92 ± 2% perf-profile.self.cycles-pp.__kmalloc_node_track_caller_noprof
1.30 ± 2% -0.3 0.98 ± 2% perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
1.21 ± 2% -0.3 0.90 ± 2% perf-profile.self.cycles-pp._copy_from_iter
1.10 ± 3% -0.3 0.81 ± 2% perf-profile.self.cycles-pp.__alloc_skb
0.68 ± 3% -0.3 0.39 ± 2% perf-profile.self.cycles-pp.__skb_datagram_iter
1.05 ± 3% -0.3 0.77 ± 2% perf-profile.self.cycles-pp.sock_write_iter
0.99 -0.3 0.72 ± 2% perf-profile.self.cycles-pp.kmem_cache_alloc_node_noprof
0.58 ± 3% -0.3 0.33 perf-profile.self.cycles-pp.__build_skb_around
0.92 ± 2% -0.2 0.70 ± 2% perf-profile.self.cycles-pp.read
0.92 ± 2% -0.2 0.70 ± 2% perf-profile.self.cycles-pp.write
1.34 ± 3% -0.2 1.14 perf-profile.self.cycles-pp.syscall_exit_to_user_mode
0.39 ± 8% -0.2 0.21 ± 3% perf-profile.self.cycles-pp.__put_partials
0.88 ± 2% -0.2 0.71 ± 4% perf-profile.self.cycles-pp.obj_cgroup_charge
0.69 ± 3% -0.2 0.52 ± 2% perf-profile.self.cycles-pp.__check_heap_object
0.70 ± 6% -0.2 0.54 ± 2% perf-profile.self.cycles-pp.vfs_write
0.80 ± 2% -0.2 0.64 ± 2% perf-profile.self.cycles-pp.sock_read_iter
0.79 ± 3% -0.1 0.65 ± 2% perf-profile.self.cycles-pp.do_syscall_64
0.53 ± 3% -0.1 0.39 ± 2% perf-profile.self.cycles-pp.rw_verify_area
0.57 ± 3% -0.1 0.44 ± 2% perf-profile.self.cycles-pp.__check_object_size
0.77 ± 3% -0.1 0.64 ± 2% perf-profile.self.cycles-pp.vfs_read
0.47 ± 5% -0.1 0.36 ± 2% perf-profile.self.cycles-pp.kfree
0.58 ± 3% -0.1 0.47 perf-profile.self.cycles-pp.syscall_return_via_sysret
0.34 ± 5% -0.1 0.24 perf-profile.self.cycles-pp.sock_alloc_send_pskb
0.51 ± 3% -0.1 0.41 ± 2% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
0.33 ± 6% -0.1 0.23 ± 6% perf-profile.self.cycles-pp.fsnotify_pre_content
0.20 ± 6% -0.1 0.10 ± 3% perf-profile.self.cycles-pp.put_cpu_partial
0.48 ± 3% -0.1 0.39 ± 4% perf-profile.self.cycles-pp.__virt_addr_valid
0.20 ± 3% -0.1 0.11 ± 18% perf-profile.self.cycles-pp.obj_cgroup_uncharge_pages
0.36 ± 3% -0.1 0.28 ± 2% perf-profile.self.cycles-pp.x64_sys_call
0.34 ± 2% -0.1 0.26 ± 2% perf-profile.self.cycles-pp.__cond_resched
0.28 ± 3% -0.1 0.21 ± 2% perf-profile.self.cycles-pp.ksys_write
0.43 ± 3% -0.1 0.35 ± 2% perf-profile.self.cycles-pp.entry_SYSCALL_64
0.24 ± 3% -0.1 0.18 ± 5% perf-profile.self.cycles-pp.kmalloc_reserve
0.27 ± 3% -0.1 0.20 ± 2% perf-profile.self.cycles-pp.alloc_skb_with_frags
0.30 ± 3% -0.1 0.24 ± 3% perf-profile.self.cycles-pp.skb_copy_datagram_from_iter
0.33 ± 3% -0.1 0.27 ± 3% perf-profile.self.cycles-pp.sock_recvmsg
0.29 ± 3% -0.1 0.23 ± 2% perf-profile.self.cycles-pp.ksys_read
0.21 ± 3% -0.1 0.16 ± 3% perf-profile.self.cycles-pp.rcu_all_qs
0.22 ± 4% -0.0 0.18 ± 2% perf-profile.self.cycles-pp.unix_stream_recvmsg
0.19 ± 3% -0.0 0.14 perf-profile.self.cycles-pp.security_file_permission
0.25 ± 3% -0.0 0.21 ± 3% perf-profile.self.cycles-pp._raw_spin_unlock_irqrestore
0.21 ± 3% -0.0 0.16 ± 3% perf-profile.self.cycles-pp.__scm_recv_common
0.17 ± 6% -0.0 0.13 ± 7% perf-profile.self.cycles-pp.kmalloc_size_roundup
0.14 ± 3% -0.0 0.10 perf-profile.self.cycles-pp.skb_unlink
0.13 ± 3% -0.0 0.10 ± 4% perf-profile.self.cycles-pp.skb_queue_tail
0.12 ± 7% -0.0 0.09 perf-profile.self.cycles-pp.maybe_add_creds
0.18 ± 4% -0.0 0.15 ± 3% perf-profile.self.cycles-pp.entry_SYSCALL_64_safe_stack
0.13 ± 3% -0.0 0.10 ± 3% perf-profile.self.cycles-pp.skb_copy_datagram_iter
0.13 ± 2% -0.0 0.10 ± 4% perf-profile.self.cycles-pp.consume_skb
0.18 ± 3% -0.0 0.15 ± 3% perf-profile.self.cycles-pp.unix_destruct_scm
0.12 ± 5% -0.0 0.09 perf-profile.self.cycles-pp.is_vmalloc_addr
0.11 ± 4% -0.0 0.08 ± 5% perf-profile.self.cycles-pp.security_socket_getpeersec_dgram
0.14 ± 3% -0.0 0.11 ± 3% perf-profile.self.cycles-pp.check_stack_object
0.10 ± 4% -0.0 0.07 perf-profile.self.cycles-pp.security_socket_sendmsg
0.17 ± 5% -0.0 0.14 ± 3% perf-profile.self.cycles-pp.syscall_exit_to_user_mode_prepare
0.06 -0.0 0.03 ± 70% perf-profile.self.cycles-pp.kfree_skbmem
0.09 ± 4% -0.0 0.06 ± 6% perf-profile.self.cycles-pp.wait_for_unix_gc
0.12 ± 4% -0.0 0.09 ± 5% perf-profile.self.cycles-pp.security_socket_recvmsg
0.15 ± 3% -0.0 0.12 ± 3% perf-profile.self.cycles-pp.unix_scm_to_skb
0.12 ± 4% -0.0 0.09 perf-profile.self.cycles-pp.manage_oob
0.10 ± 4% -0.0 0.08 perf-profile.self.cycles-pp.skb_release_head_state
0.12 ± 4% -0.0 0.09 perf-profile.self.cycles-pp.put_pid
0.08 ± 6% -0.0 0.06 ± 8% perf-profile.self.cycles-pp.skb_put
0.08 ± 6% -0.0 0.06 perf-profile.self.cycles-pp.skb_free_head
0.09 -0.0 0.08 ± 6% perf-profile.self.cycles-pp.simple_copy_to_iter
0.08 ± 6% -0.0 0.06 ± 6% perf-profile.self.cycles-pp.unix_stream_read_actor
0.07 ± 16% +0.0 0.10 ± 7% perf-profile.self.cycles-pp.__get_user_8
0.08 ± 19% +0.0 0.11 ± 4% perf-profile.self.cycles-pp.pick_next_task_fair
0.05 ± 47% +0.0 0.08 ± 4% perf-profile.self.cycles-pp.unix_stream_data_wait
0.02 ± 99% +0.0 0.07 ± 7% perf-profile.self.cycles-pp.__wrgsbase_inactive
0.08 ± 14% +0.0 0.12 ± 5% perf-profile.self.cycles-pp.os_xsave
0.02 ±141% +0.0 0.06 ± 6% perf-profile.self.cycles-pp.place_entity
0.07 ± 32% +0.0 0.12 ± 12% perf-profile.self.cycles-pp.enqueue_task_fair
0.03 ±100% +0.0 0.07 ± 10% perf-profile.self.cycles-pp.___perf_sw_event
0.05 ± 74% +0.0 0.10 ± 5% perf-profile.self.cycles-pp.avg_vruntime
0.06 ± 13% +0.1 0.11 ± 11% perf-profile.self.cycles-pp.vruntime_eligible
0.25 ± 8% +0.1 0.30 ± 5% perf-profile.self.cycles-pp.restore_fpregs_from_fpstate
0.00 +0.1 0.05 ± 7% perf-profile.self.cycles-pp.check_preempt_wakeup_fair
0.02 ±141% +0.1 0.07 ± 5% perf-profile.self.cycles-pp.select_task_rq_fair
0.01 ±223% +0.1 0.06 ± 6% perf-profile.self.cycles-pp.schedule
0.02 ±141% +0.1 0.07 ± 8% perf-profile.self.cycles-pp.__put_user_8
0.00 +0.1 0.06 ± 9% perf-profile.self.cycles-pp.__rb_insert_augmented
0.04 ±100% +0.1 0.09 ± 7% perf-profile.self.cycles-pp.native_sched_clock
0.04 ±100% +0.1 0.10 ± 4% perf-profile.self.cycles-pp.dequeue_entity
0.09 ± 21% +0.1 0.15 ± 7% perf-profile.self.cycles-pp.dequeue_entities
0.04 ± 72% +0.1 0.11 ± 8% perf-profile.self.cycles-pp.rseq_update_cpu_node_id
0.00 +0.1 0.07 ± 6% perf-profile.self.cycles-pp.refill_stock
0.43 ± 4% +0.1 0.50 ± 2% perf-profile.self.cycles-pp.unix_write_space
0.00 +0.1 0.08 ± 9% perf-profile.self.cycles-pp.set_next_buddy
0.04 ±100% +0.1 0.12 ± 5% perf-profile.self.cycles-pp.switch_fpu_return
0.15 ± 22% +0.1 0.24 ± 6% perf-profile.self.cycles-pp.__update_load_avg_se
0.10 ± 24% +0.1 0.19 ± 4% perf-profile.self.cycles-pp.prepare_task_switch
0.15 ± 11% +0.1 0.24 ± 7% perf-profile.self.cycles-pp.__dequeue_entity
0.13 ± 34% +0.1 0.22 ± 9% perf-profile.self.cycles-pp.update_curr
0.12 ± 23% +0.1 0.21 ± 9% perf-profile.self.cycles-pp.__update_load_avg_cfs_rq
0.04 ±100% +0.1 0.14 ± 4% perf-profile.self.cycles-pp.update_rq_clock_task
0.04 ±100% +0.1 0.14 ± 5% perf-profile.self.cycles-pp.prepare_to_wait
0.28 ± 24% +0.1 0.40 ± 5% perf-profile.self.cycles-pp.__schedule
0.18 ± 14% +0.1 0.32 ± 6% perf-profile.self.cycles-pp.__enqueue_entity
0.13 ± 19% +0.1 0.27 ± 5% perf-profile.self.cycles-pp.__switch_to
0.40 ± 6% +0.1 0.54 ± 2% perf-profile.self.cycles-pp.mutex_lock
0.11 ± 20% +0.2 0.28 ± 7% perf-profile.self.cycles-pp.pick_eevdf
0.35 ± 18% +0.2 0.52 ± 9% perf-profile.self.cycles-pp.__switch_to_asm
0.00 +0.2 0.18 ± 4% perf-profile.self.cycles-pp.page_counter_cancel
0.37 ± 6% +0.3 0.67 perf-profile.self.cycles-pp.mutex_unlock
0.65 ± 6% +0.3 0.99 perf-profile.self.cycles-pp.fput
1.68 ± 14% +0.4 2.08 ± 2% perf-profile.self.cycles-pp._raw_spin_lock_irqsave
0.27 ± 9% +0.5 0.80 ± 56% perf-profile.self.cycles-pp.__mod_memcg_lruvec_state
0.22 ± 7% +0.5 0.77 ± 66% perf-profile.self.cycles-pp.__mod_memcg_state
0.00 +0.7 0.65 ± 11% perf-profile.self.cycles-pp.propagate_protected_usage
0.10 ± 4% +11.3 11.43 ± 3% perf-profile.self.cycles-pp.try_charge_memcg
0.00 +13.7 13.72 ± 2% perf-profile.self.cycles-pp.page_counter_try_charge
***************************************************************************************************
lkp-spr-2sp4: 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 512G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_threads/rootfs/tbox_group/test/test_memory_size/testcase:
gcc-12/performance/x86_64-rhel-9.4/development/100%/debian-12-x86_64-20240206.cgz/lkp-spr-2sp4/TCP/50%/lmbench3
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
1594 ± 3% +19.3% 1901 meminfo.Mlocked
149.67 ± 4% +8.9% 163.00 ± 5% perf-sched.wait_and_delay.count.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
969.96 ± 6% +48.3% 1438 ± 2% uptime.boot
6.042e+08 -9.6% 5.46e+08 ± 4% numa-numastat.node0.local_node
6.043e+08 -9.6% 5.462e+08 ± 4% numa-numastat.node0.numa_hit
6.043e+08 -9.6% 5.462e+08 ± 4% numa-vmstat.node0.numa_hit
6.042e+08 -9.6% 5.46e+08 ± 4% numa-vmstat.node0.numa_local
4526689 ± 6% +310.8% 18596060 ± 2% vmstat.system.cs
303047 -8.1% 278503 vmstat.system.in
12.93 ± 5% -4.2 8.72 ± 3% mpstat.cpu.all.idle%
4.39 ± 4% +9.4 13.80 ± 2% mpstat.cpu.all.soft%
5.33 ± 3% -1.1 4.22 mpstat.cpu.all.usr%
184478 -34.4% 121014 ± 2% lmbench3.TCP.socket.bandwidth.10MB.MB/sec
12082 ± 4% -33.0% 8091 lmbench3.TCP.socket.bandwidth.64B.MB/sec
915.46 ± 7% +50.8% 1380 ± 2% lmbench3.time.elapsed_time
915.46 ± 7% +50.8% 1380 ± 2% lmbench3.time.elapsed_time.max
44831013 ± 7% -44.1% 25067866 ± 3% lmbench3.time.involuntary_context_switches
11254 ± 4% -27.5% 8155 ± 3% lmbench3.time.percent_of_cpu_this_job_got
6453 ± 6% +9.1% 7040 ± 2% lmbench3.time.user_time
1.802e+09 ± 5% +597.6% 1.257e+10 ± 2% lmbench3.time.voluntary_context_switches
397.85 ± 3% +19.4% 474.88 proc-vmstat.nr_mlock
8096 +2.9% 8335 proc-vmstat.nr_page_table_pages
1.205e+09 -8.9% 1.098e+09 proc-vmstat.numa_hit
1.205e+09 -8.9% 1.097e+09 proc-vmstat.numa_local
9.615e+09 -8.9% 8.756e+09 proc-vmstat.pgalloc_normal
3414882 ± 3% +32.2% 4513062 ± 2% proc-vmstat.pgfault
9.614e+09 -8.9% 8.755e+09 proc-vmstat.pgfree
146555 ± 3% +30.1% 190716 ± 4% proc-vmstat.pgreuse
0.65 ± 77% +0.6 1.21 ± 21% perf-profile.calltrace.cycles-pp.__tcp_cleanup_rbuf.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg.sock_recvmsg
0.58 ± 80% +1.7 2.23 ± 8% perf-profile.calltrace.cycles-pp.release_sock.tcp_recvmsg.inet_recvmsg.sock_recvmsg.sock_read_iter
0.23 ± 51% +0.1 0.36 ± 2% perf-profile.children.cycles-pp.record__pushfn
0.23 ± 51% +0.1 0.36 ± 2% perf-profile.children.cycles-pp.writen
0.21 ± 53% +0.1 0.35 ± 5% perf-profile.children.cycles-pp.shmem_file_write_iter
0.21 ± 52% +0.1 0.35 ± 6% perf-profile.children.cycles-pp.generic_perform_write
0.78 ± 49% +0.5 1.25 ± 21% perf-profile.children.cycles-pp.__tcp_cleanup_rbuf
0.17 ± 78% +0.3 0.45 ± 24% perf-profile.self.cycles-pp.__tcp_cleanup_rbuf
0.33 ± 99% +1.7 2.02 ± 10% perf-profile.self.cycles-pp.release_sock
83660333 ± 7% +37.1% 1.147e+08 ± 2% sched_debug.cfs_rq:/.avg_vruntime.avg
97337639 ± 9% +37.5% 1.338e+08 ± 2% sched_debug.cfs_rq:/.avg_vruntime.max
71425435 ± 6% +42.3% 1.016e+08 ± 5% sched_debug.cfs_rq:/.avg_vruntime.min
0.42 ± 4% +16.7% 0.50 ± 3% sched_debug.cfs_rq:/.h_nr_queued.stddev
83660333 ± 7% +37.1% 1.147e+08 ± 2% sched_debug.cfs_rq:/.min_vruntime.avg
97337639 ± 9% +37.5% 1.338e+08 ± 2% sched_debug.cfs_rq:/.min_vruntime.max
71425435 ± 6% +42.3% 1.016e+08 ± 5% sched_debug.cfs_rq:/.min_vruntime.min
104.29 ± 40% -41.8% 60.70 ± 40% sched_debug.cfs_rq:/.removed.load_avg.max
271.46 ± 7% +17.4% 318.60 ± 4% sched_debug.cfs_rq:/.util_est.stddev
839790 ± 5% -13.7% 724828 ± 3% sched_debug.cpu.avg_idle.avg
94203 ± 36% -46.8% 50142 ± 55% sched_debug.cpu.avg_idle.min
312614 ± 8% +28.3% 401108 ± 3% sched_debug.cpu.avg_idle.stddev
500309 ± 5% +45.6% 728380 sched_debug.cpu.clock.avg
500675 ± 5% +45.5% 728646 sched_debug.cpu.clock.max
499891 ± 5% +45.7% 728091 sched_debug.cpu.clock.min
472701 ± 5% +33.6% 631669 ± 2% sched_debug.cpu.clock_task.avg
480686 ± 5% +34.3% 645473 ± 2% sched_debug.cpu.clock_task.max
459106 ± 5% +33.2% 611651 ± 2% sched_debug.cpu.clock_task.min
2864 ± 8% +95.5% 5599 ± 13% sched_debug.cpu.clock_task.stddev
17343 ± 6% +30.8% 22685 ± 3% sched_debug.cpu.curr->pid.avg
20079 ± 3% +28.2% 25732 sched_debug.cpu.curr->pid.max
66128 ± 16% -32.6% 44587 ± 7% sched_debug.cpu.max_idle_balance_cost.stddev
0.42 ± 4% +17.4% 0.50 ± 3% sched_debug.cpu.nr_running.stddev
12746739 ± 6% +370.9% 60025838 sched_debug.cpu.nr_switches.avg
16962002 ± 6% +296.5% 67251221 ± 2% sched_debug.cpu.nr_switches.max
7092180 ± 12% +582.2% 48379329 ± 6% sched_debug.cpu.nr_switches.min
1956675 ± 10% +72.0% 3365465 ± 26% sched_debug.cpu.nr_switches.stddev
499883 ± 5% +45.7% 728085 sched_debug.cpu_clk
498838 ± 5% +45.7% 727039 sched_debug.ktime
500718 ± 5% +45.6% 728919 sched_debug.sched_clk
7.122e+10 ± 4% +9.8% 7.823e+10 ± 2% perf-stat.i.branch-instructions
0.33 ± 3% +0.0 0.37 perf-stat.i.branch-miss-rate%
1.092e+08 ± 3% +107.4% 2.265e+08 ± 2% perf-stat.i.branch-misses
13.88 ± 3% +7.1 20.93 ± 2% perf-stat.i.cache-miss-rate%
2.94e+08 ± 8% -33.3% 1.962e+08 ± 3% perf-stat.i.cache-misses
1.297e+09 ± 4% -56.4% 5.652e+08 ± 3% perf-stat.i.cache-references
4741398 ± 5% +304.2% 19164589 ± 2% perf-stat.i.context-switches
5.582e+11 +5.5% 5.891e+11 perf-stat.i.cpu-cycles
443841 ± 4% -36.7% 281171 ± 6% perf-stat.i.cpu-migrations
35296 ± 8% -26.9% 25797 ± 4% perf-stat.i.cycles-between-cache-misses
3.65e+11 ± 4% +10.9% 4.047e+11 ± 2% perf-stat.i.instructions
0.65 ± 8% -42.8% 0.37 ± 24% perf-stat.i.major-faults
23.04 ± 5% +276.7% 86.77 ± 2% perf-stat.i.metric.K/sec
3611 ± 4% -12.8% 3149 perf-stat.i.minor-faults
3612 ± 4% -12.8% 3149 perf-stat.i.page-faults
0.79 ± 5% -38.0% 0.49 ± 4% perf-stat.overall.MPKI
0.15 ± 2% +0.1 0.29 perf-stat.overall.branch-miss-rate%
22.67 ± 4% +12.4 35.07 ± 2% perf-stat.overall.cache-miss-rate%
2022 ± 10% +50.6% 3046 ± 4% perf-stat.overall.cycles-between-cache-misses
6.892e+10 ± 4% +10.9% 7.64e+10 ± 2% perf-stat.ps.branch-instructions
1.043e+08 ± 4% +110.3% 2.194e+08 ± 2% perf-stat.ps.branch-misses
2.803e+08 ± 9% -30.8% 1.94e+08 ± 4% perf-stat.ps.cache-misses
1.234e+09 ± 5% -55.2% 5.533e+08 ± 3% perf-stat.ps.cache-references
4509001 ± 6% +311.1% 18537174 ± 2% perf-stat.ps.context-switches
5.613e+11 +5.1% 5.9e+11 perf-stat.ps.cpu-cycles
412889 ± 4% -35.3% 267257 ± 6% perf-stat.ps.cpu-migrations
3.533e+11 ± 4% +11.9% 3.952e+11 ± 2% perf-stat.ps.instructions
0.69 ± 7% -41.4% 0.40 ± 23% perf-stat.ps.major-faults
3548 ± 4% -11.9% 3125 perf-stat.ps.minor-faults
3549 ± 4% -11.9% 3126 perf-stat.ps.page-faults
3.226e+14 ± 2% +69.2% 5.459e+14 perf-stat.total.instructions
***************************************************************************************************
lkp-emr-2sp1: 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/rootfs/runtime/size/tbox_group/test/testcase:
gcc-12/performance/x86_64-rhel-9.4/debian-12-x86_64-20240206.cgz/300s/1T/lkp-emr-2sp1/lru-shm/vm-scalability
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
4.666e+10 +12.1% 5.231e+10 ± 2% cpuidle..time
4452013 +145.5% 10928419 ± 4% cpuidle..usage
233.91 +46.5% 342.71 uptime.boot
54574 +10.4% 60225 uptime.idle
65451 ± 28% +86.7% 122214 ± 35% meminfo.AnonHugePages
4653322 ± 4% +338.3% 20396688 ± 6% meminfo.Mapped
22987 ± 2% +150.5% 57574 ± 5% meminfo.PageTables
1.00 ±100% +66550.0% 666.50 ±138% perf-c2c.DRAM.local
58.17 ± 21% +15319.8% 8969 ±127% perf-c2c.DRAM.remote
11.83 ± 80% +14446.5% 1721 ±138% perf-c2c.HITM.local
19.17 ± 32% +11917.4% 2303 ±125% perf-c2c.HITM.remote
91.76 -27.2% 66.77 ± 2% vmstat.cpu.id
22.81 ± 2% +280.7% 86.82 ± 3% vmstat.procs.r
6701 -9.9% 6040 ± 3% vmstat.system.cs
47395 ± 5% +192.6% 138675 ± 5% vmstat.system.in
91.77 -25.2 66.53 ± 2% mpstat.cpu.all.idle%
0.07 ± 10% +0.2 0.24 ± 42% mpstat.cpu.all.irq%
0.03 +0.0 0.06 ± 25% mpstat.cpu.all.soft%
5.84 +25.4 31.28 ± 4% mpstat.cpu.all.sys%
2.29 -0.4 1.89 ± 3% mpstat.cpu.all.usr%
45.47 +100.8% 91.30 ± 7% mpstat.max_utilization_pct
1186980 +334.3% 5154770 ± 10% numa-meminfo.node0.Mapped
6229 ± 11% +131.7% 14432 ± 3% numa-meminfo.node0.PageTables
1183346 ± 4% +317.9% 4945218 ± 6% numa-meminfo.node1.Mapped
5409 ± 14% +154.5% 13769 ± 4% numa-meminfo.node1.PageTables
1213253 ± 5% +323.8% 5142208 ± 7% numa-meminfo.node2.Mapped
6317 ± 17% +129.9% 14522 ± 6% numa-meminfo.node2.PageTables
1219782 ± 4% +311.7% 5022404 ± 10% numa-meminfo.node3.Mapped
5915 ± 10% +143.6% 14409 ± 9% numa-meminfo.node3.PageTables
293807 ± 5% +333.7% 1274238 ± 10% numa-vmstat.node0.nr_mapped
1529 ± 11% +134.5% 3587 ± 4% numa-vmstat.node0.nr_page_table_pages
293261 ± 2% +316.8% 1222310 ± 6% numa-vmstat.node1.nr_mapped
1356 ± 12% +152.5% 3424 ± 5% numa-vmstat.node1.nr_page_table_pages
300064 ± 5% +324.0% 1272313 ± 6% numa-vmstat.node2.nr_mapped
1566 ± 15% +130.9% 3617 ± 7% numa-vmstat.node2.nr_page_table_pages
299769 ± 5% +317.7% 1252275 ± 10% numa-vmstat.node3.nr_mapped
1466 ± 14% +145.5% 3600 ± 9% numa-vmstat.node3.nr_page_table_pages
43026 +4.4% 44910 proc-vmstat.nr_kernel_stack
1205005 ± 2% +321.4% 5077813 ± 6% proc-vmstat.nr_mapped
5917 +142.4% 14340 ± 5% proc-vmstat.nr_page_table_pages
79171 -3.6% 76354 proc-vmstat.nr_slab_reclaimable
21575 ± 22% +109.8% 45270 ± 8% proc-vmstat.numa_hint_faults
1.071e+09 -12.1% 9.412e+08 ± 4% proc-vmstat.numa_hit
1.063e+09 -11.6% 9.402e+08 ± 4% proc-vmstat.numa_local
198294 ± 27% +104.9% 406310 ± 37% proc-vmstat.numa_pte_updates
1.063e+09 -11.4% 9.423e+08 ± 4% proc-vmstat.pgalloc_normal
1.06e+09 -11.4% 9.389e+08 ± 4% proc-vmstat.pgfault
1.063e+09 -11.4% 9.421e+08 ± 4% proc-vmstat.pgfree
396442 -5.6% 374046 ± 3% proc-vmstat.pgreuse
5826 ± 3% +42.1% 8276 ± 9% proc-vmstat.unevictable_pgs_culled
0.01 +65.2% 0.01 ± 4% vm-scalability.free_time
1315486 -86.6% 175811 ± 4% vm-scalability.median
0.85 ± 17% +5.3 6.19 ± 57% vm-scalability.median_stddev%
1.24 ± 22% +4.7 5.92 ± 58% vm-scalability.stddev%
3.401e+08 -86.8% 45006220 ± 5% vm-scalability.throughput
197.44 +55.0% 305.97 vm-scalability.time.elapsed_time
197.44 +55.0% 305.97 vm-scalability.time.elapsed_time.max
30562 +186.6% 87583 ± 7% vm-scalability.time.involuntary_context_switches
1.059e+09 -11.5% 9.375e+08 ± 4% vm-scalability.time.minor_page_faults
1964 +326.5% 8377 ± 4% vm-scalability.time.percent_of_cpu_this_job_got
2761 +777.9% 24242 ± 4% vm-scalability.time.system_time
1117 +24.6% 1393 ± 4% vm-scalability.time.user_time
107496 -12.3% 94315 ± 6% vm-scalability.time.voluntary_context_switches
4.742e+09 -11.5% 4.199e+09 ± 4% vm-scalability.workload
3.489e+10 -43.5% 1.972e+10 ± 5% perf-stat.i.branch-instructions
0.27 -0.1 0.19 ± 3% perf-stat.i.branch-miss-rate%
26446125 -31.2% 18182985 ± 2% perf-stat.i.branch-misses
1.417e+08 ± 2% -39.7% 85414690 ± 11% perf-stat.i.cache-misses
3.788e+08 -45.6% 2.06e+08 ± 17% perf-stat.i.cache-references
6654 -10.2% 5975 ± 3% perf-stat.i.context-switches
0.69 ± 7% +239.3% 2.34 ± 5% perf-stat.i.cpi
6.799e+10 +259.9% 2.447e+11 ± 5% perf-stat.i.cpu-cycles
588.49 ± 3% -16.5% 491.20 ± 4% perf-stat.i.cpu-migrations
524.75 ± 9% +285.2% 2021 ± 13% perf-stat.i.cycles-between-cache-misses
1.258e+11 -43.1% 7.162e+10 ± 5% perf-stat.i.instructions
1.56 ± 3% -50.9% 0.77 ± 4% perf-stat.i.ipc
1.02 ± 6% -53.4% 0.47 ± 10% perf-stat.i.major-faults
41.68 -44.5% 23.14 ± 6% perf-stat.i.metric.K/sec
5383984 -44.8% 2971563 ± 6% perf-stat.i.minor-faults
5383985 -44.8% 2971564 ± 6% perf-stat.i.page-faults
0.08 +0.0 0.09 ± 2% perf-stat.overall.branch-miss-rate%
37.40 +4.6 41.96 ± 8% perf-stat.overall.cache-miss-rate%
0.54 +542.1% 3.47 ± 3% perf-stat.overall.cpi
480.08 +516.6% 2960 ± 13% perf-stat.overall.cycles-between-cache-misses
1.85 -84.4% 0.29 ± 3% perf-stat.overall.ipc
5222 +2.4% 5350 perf-stat.overall.path-length
3.465e+10 -41.7% 2.019e+10 ± 4% perf-stat.ps.branch-instructions
26086104 -30.8% 18041050 ± 2% perf-stat.ps.branch-misses
1.408e+08 ± 2% -38.1% 87092040 ± 12% perf-stat.ps.cache-misses
3.763e+08 -44.1% 2.104e+08 ± 17% perf-stat.ps.cache-references
6606 -9.6% 5969 ± 3% perf-stat.ps.context-switches
6.754e+10 +275.5% 2.536e+11 ± 3% perf-stat.ps.cpu-cycles
584.17 ± 3% -16.0% 490.74 ± 4% perf-stat.ps.cpu-migrations
1.25e+11 -41.5% 7.316e+10 ± 4% perf-stat.ps.instructions
1.01 ± 6% -53.0% 0.48 ± 9% perf-stat.ps.major-faults
5346919 -42.8% 3057237 ± 4% perf-stat.ps.minor-faults
5346920 -42.8% 3057238 ± 4% perf-stat.ps.page-faults
2.476e+13 -9.3% 2.246e+13 ± 4% perf-stat.total.instructions
931893 ± 12% +892.4% 9248299 ± 18% sched_debug.cfs_rq:/.avg_vruntime.avg
1446903 ± 12% +687.3% 11391069 ± 16% sched_debug.cfs_rq:/.avg_vruntime.max
481792 ± 17% +1257.4% 6539912 ± 25% sched_debug.cfs_rq:/.avg_vruntime.min
204874 ± 10% +338.4% 898146 ± 8% sched_debug.cfs_rq:/.avg_vruntime.stddev
0.09 ± 67% +287.2% 0.36 ± 16% sched_debug.cfs_rq:/.h_nr_queued.avg
0.09 ± 70% +296.6% 0.36 ± 16% sched_debug.cfs_rq:/.h_nr_runnable.avg
708874 ± 42% +532.5% 4483644 ± 67% sched_debug.cfs_rq:/.left_deadline.max
63364 ± 46% +468.8% 360399 ± 72% sched_debug.cfs_rq:/.left_deadline.stddev
708816 ± 42% +532.5% 4483542 ± 67% sched_debug.cfs_rq:/.left_vruntime.max
63359 ± 46% +468.8% 360393 ± 72% sched_debug.cfs_rq:/.left_vruntime.stddev
931893 ± 12% +892.4% 9248299 ± 18% sched_debug.cfs_rq:/.min_vruntime.avg
1446903 ± 12% +687.3% 11391069 ± 16% sched_debug.cfs_rq:/.min_vruntime.max
481792 ± 17% +1257.4% 6539912 ± 25% sched_debug.cfs_rq:/.min_vruntime.min
204874 ± 10% +338.4% 898146 ± 8% sched_debug.cfs_rq:/.min_vruntime.stddev
0.10 ± 64% +278.4% 0.36 ± 16% sched_debug.cfs_rq:/.nr_queued.avg
708816 ± 42% +532.5% 4483542 ± 67% sched_debug.cfs_rq:/.right_vruntime.max
63359 ± 46% +468.8% 360393 ± 72% sched_debug.cfs_rq:/.right_vruntime.stddev
103.73 ± 59% +264.8% 378.38 ± 16% sched_debug.cfs_rq:/.util_avg.avg
195.87 ± 22% +53.5% 300.66 ± 13% sched_debug.cfs_rq:/.util_avg.stddev
31.73 ± 86% +384.5% 153.73 ± 17% sched_debug.cfs_rq:/.util_est.avg
82.97 ± 51% +78.2% 147.83 ± 10% sched_debug.cfs_rq:/.util_est.stddev
183331 ± 20% +37.5% 252021 ± 19% sched_debug.cpu.avg_idle.min
120841 ± 9% +41.6% 171158 ± 8% sched_debug.cpu.clock.avg
120861 ± 9% +41.7% 171224 ± 8% sched_debug.cpu.clock.max
120779 ± 9% +41.6% 171052 ± 8% sched_debug.cpu.clock.min
13.07 ± 15% +239.4% 44.36 ± 49% sched_debug.cpu.clock.stddev
120651 ± 9% +41.5% 170689 ± 8% sched_debug.cpu.clock_task.avg
120788 ± 9% +41.6% 171006 ± 8% sched_debug.cpu.clock_task.max
105855 ± 10% +45.7% 154255 ± 9% sched_debug.cpu.clock_task.min
2562 ± 99% +325.1% 10895 ± 17% sched_debug.cpu.curr->pid.avg
4577 ± 51% +81.1% 8289 ± 16% sched_debug.cpu.curr->pid.stddev
19843 ± 57% +640.0% 146849 ±124% sched_debug.cpu.max_idle_balance_cost.stddev
0.00 ± 65% +146.2% 0.00 ± 53% sched_debug.cpu.next_balance.stddev
0.08 ± 72% +313.0% 0.35 ± 17% sched_debug.cpu.nr_running.avg
0.20 ± 21% +48.7% 0.30 ± 11% sched_debug.cpu.nr_running.stddev
2694 ± 10% +29.2% 3482 ± 10% sched_debug.cpu.nr_switches.avg
946.51 ± 11% +27.7% 1208 ± 13% sched_debug.cpu.nr_switches.min
2562 ± 8% +40.0% 3586 ± 22% sched_debug.cpu.nr_switches.stddev
120822 ± 9% +41.6% 171065 ± 8% sched_debug.cpu_clk
119783 ± 9% +41.9% 170023 ± 8% sched_debug.ktime
122204 ± 9% +41.1% 172446 ± 8% sched_debug.sched_clk
0.19 ± 12% +115.1% 0.40 ± 40% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.05 ± 41% +346.6% 0.21 ± 27% perf-sched.sch_delay.avg.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
0.01 ± 59% +577.5% 0.05 ± 30% perf-sched.sch_delay.avg.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.05 ± 92% +167.9% 0.12 ± 19% perf-sched.sch_delay.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
0.01 ± 14% +623.5% 0.10 ± 74% perf-sched.sch_delay.avg.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.01 ± 5% +798.7% 0.12 ± 21% perf-sched.sch_delay.avg.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
0.02 ± 38% +525.5% 0.16 ± 52% perf-sched.sch_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.01 ± 20% +223.8% 0.04 ± 47% perf-sched.sch_delay.avg.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
0.04 ± 38% +852.4% 0.36 ± 66% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
0.02 ± 18% +771.7% 0.14 ± 65% perf-sched.sch_delay.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
0.15 ± 11% -36.6% 0.10 ± 28% perf-sched.sch_delay.avg.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
0.03 ± 21% +865.9% 0.28 ± 82% perf-sched.sch_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
0.15 ± 17% -66.0% 0.05 ± 35% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
0.20 ± 50% +24616.3% 50.38 ± 80% perf-sched.sch_delay.max.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
0.02 ± 97% +1023.3% 0.17 ± 38% perf-sched.sch_delay.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
3.42 ± 12% +101.3% 6.88 ± 61% perf-sched.sch_delay.max.ms.__cond_resched.ww_mutex_lock.drm_gem_vunmap_unlocked.drm_gem_fb_vunmap.drm_atomic_helper_commit_planes
0.11 ± 74% +483.3% 0.62 ± 44% perf-sched.sch_delay.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
0.02 ± 15% +1525.2% 0.29 ± 68% perf-sched.sch_delay.max.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.02 ± 15% +2244.7% 0.40 ± 51% perf-sched.sch_delay.max.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
1.02 ±103% +1337.3% 14.66 ±118% perf-sched.sch_delay.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.02 ± 25% +418.0% 0.11 ± 54% perf-sched.sch_delay.max.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
0.68 ±101% +6041.8% 41.46 ±129% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
1.12 ± 60% +49035.6% 551.87 ± 84% perf-sched.sch_delay.max.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
0.42 ± 71% +12165.4% 50.94 ±116% perf-sched.sch_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
0.27 ± 32% +31119.3% 84.55 ±221% perf-sched.sch_delay.max.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
1.98 ± 14% +800.7% 17.87 ± 72% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
0.10 ± 18% +70.7% 0.18 ± 14% perf-sched.total_sch_delay.average.ms
120.83 ±144% +528.3% 759.23 ± 51% perf-sched.total_sch_delay.max.ms
0.12 ±109% +533.6% 0.74 ± 43% perf-sched.wait_and_delay.avg.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
0.51 ± 43% +568.9% 3.44 ± 49% perf-sched.wait_and_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.06 ±101% +1002.8% 0.66 ± 57% perf-sched.wait_and_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
1.39 ± 20% +425.5% 7.30 ± 12% perf-sched.wait_and_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
1.48 ±100% +754.1% 12.63 ± 13% perf-sched.wait_and_delay.avg.ms.sigsuspend.__x64_sys_rt_sigsuspend.do_syscall_64.entry_SYSCALL_64_after_hwframe
77.50 ±108% +341.7% 342.33 ± 10% perf-sched.wait_and_delay.count.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
70.33 ±100% +189.6% 203.67 ± 17% perf-sched.wait_and_delay.count.sigsuspend.__x64_sys_rt_sigsuspend.do_syscall_64.entry_SYSCALL_64_after_hwframe
1209 ± 34% +46.0% 1766 ± 4% perf-sched.wait_and_delay.count.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
0.84 ±101% +1408.4% 12.67 ± 56% perf-sched.wait_and_delay.max.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
1006 ± 36% -73.0% 272.14 ± 55% perf-sched.wait_and_delay.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
1.16 ±129% +5576.0% 66.07 ±143% perf-sched.wait_and_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
12.52 ±101% +426.3% 65.90 ± 75% perf-sched.wait_and_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
10.40 ±102% +926.8% 106.78 ± 31% perf-sched.wait_and_delay.max.ms.sigsuspend.__x64_sys_rt_sigsuspend.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.05 ± 36% +196.2% 0.16 ± 27% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
0.05 ± 31% +196.0% 0.16 ± 25% perf-sched.wait_time.avg.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
0.16 ± 57% +311.3% 0.65 ± 43% perf-sched.wait_time.avg.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
11.56 ±122% +807.8% 104.94 ± 93% perf-sched.wait_time.avg.ms.__cond_resched.ww_mutex_lock.drm_gem_vunmap_unlocked.drm_gem_fb_vunmap.drm_atomic_helper_commit_planes
0.49 ± 46% +570.7% 3.29 ± 50% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.01 ±129% +403.4% 0.05 ± 36% perf-sched.wait_time.avg.ms.io_schedule.folio_wait_bit_common.filemap_fault.__do_fault
0.05 ± 40% +494.9% 0.29 ± 50% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
1.36 ± 21% +416.0% 7.02 ± 9% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
1.45 ±100% +768.6% 12.57 ± 13% perf-sched.wait_time.avg.ms.sigsuspend.__x64_sys_rt_sigsuspend.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.37 ± 52% +362.9% 1.72 ± 37% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
0.23 ± 46% +4734.9% 11.31 ± 81% perf-sched.wait_time.max.ms.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fault
1.14 ± 46% +951.6% 12.02 ± 60% perf-sched.wait_time.max.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
92.03 ± 49% +336.8% 402.04 ± 64% perf-sched.wait_time.max.ms.__cond_resched.ww_mutex_lock.drm_gem_vunmap_unlocked.drm_gem_fb_vunmap.drm_atomic_helper_commit_planes
1006 ± 36% -73.0% 272.14 ± 55% perf-sched.wait_time.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
0.05 ±121% +406.2% 0.24 ± 54% perf-sched.wait_time.max.ms.io_schedule.folio_wait_bit_common.filemap_fault.__do_fault
0.76 ± 85% +3220.5% 25.09 ±190% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
10.39 ±102% +927.7% 106.76 ± 31% perf-sched.wait_time.max.ms.sigsuspend.__x64_sys_rt_sigsuspend.do_syscall_64.entry_SYSCALL_64_after_hwframe
***************************************************************************************************
lkp-icl-2sp2: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
=========================================================================================
cluster/compiler/cpufreq_governor/ip/kconfig/nr_threads/rootfs/runtime/tbox_group/test/testcase:
cs-localhost/gcc-12/performance/ipv4/x86_64-rhel-9.4/200%/debian-12-x86_64-20240206.cgz/300s/lkp-icl-2sp2/TCP_MAERTS/netperf
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
4.421e+08 ± 2% +22.9% 5.434e+08 ± 6% numa-numastat.node0.local_node
4.421e+08 ± 2% +22.9% 5.435e+08 ± 6% numa-numastat.node0.numa_hit
15990 ± 95% -97.3% 437.17 ± 20% perf-c2c.DRAM.local
9053 ± 9% -34.0% 5974 ± 20% perf-c2c.HITM.local
9888 ± 9% -31.2% 6799 ± 20% perf-c2c.HITM.total
462.20 -23.0% 355.75 vmstat.procs.r
70900 ± 2% +9238.1% 6620774 vmstat.system.cs
172829 +22.5% 211697 ± 2% vmstat.system.in
0.71 ± 3% -0.2 0.46 ± 7% mpstat.cpu.all.irq%
5.89 ± 4% +11.9 17.74 mpstat.cpu.all.soft%
90.68 -12.1 78.54 mpstat.cpu.all.sys%
1.76 ± 10% +0.6 2.37 ± 2% mpstat.cpu.all.usr%
1502311 ± 10% +50.7% 2263993 ± 13% meminfo.Active
1502311 ± 10% +50.7% 2263993 ± 13% meminfo.Active(anon)
4135865 ± 3% +18.9% 4915987 ± 5% meminfo.Cached
1777441 ± 9% +42.6% 2534680 ± 11% meminfo.Committed_AS
305395 ± 16% -31.8% 208208 ± 14% meminfo.Mapped
631369 ± 24% +123.6% 1411635 ± 20% meminfo.Shmem
113990 ± 48% -78.4% 24658 ±134% numa-meminfo.node0.Mapped
4394098 ± 35% -43.4% 2485612 ± 63% numa-meminfo.node0.MemUsed
178139 ± 70% -94.8% 9344 ± 30% numa-meminfo.node0.Shmem
1013397 ± 22% +92.7% 1952777 ± 19% numa-meminfo.node1.Active
1013397 ± 22% +92.7% 1952777 ± 19% numa-meminfo.node1.Active(anon)
1654163 ± 81% +121.6% 3665933 ± 45% numa-meminfo.node1.FilePages
452337 ± 52% +210.1% 1402518 ± 20% numa-meminfo.node1.Shmem
28672 ± 48% -78.2% 6247 ±135% numa-vmstat.node0.nr_mapped
44638 ± 70% -94.8% 2335 ± 30% numa-vmstat.node0.nr_shmem
4.422e+08 ± 2% +22.9% 5.434e+08 ± 6% numa-vmstat.node0.numa_hit
4.421e+08 ± 2% +22.9% 5.434e+08 ± 6% numa-vmstat.node0.numa_local
253743 ± 23% +92.3% 488048 ± 19% numa-vmstat.node1.nr_active_anon
413918 ± 81% +121.4% 916363 ± 45% numa-vmstat.node1.nr_file_pages
113461 ± 52% +208.9% 350509 ± 20% numa-vmstat.node1.nr_shmem
253743 ± 23% +92.3% 488048 ± 19% numa-vmstat.node1.nr_zone_active_anon
3009 ± 2% +39.9% 4209 netperf.ThroughputBoth_Mbps
770338 ± 2% +39.9% 1077702 netperf.ThroughputBoth_total_Mbps
3009 ± 2% +39.9% 4209 netperf.Throughput_Mbps
770338 ± 2% +39.9% 1077702 netperf.Throughput_total_Mbps
6400961 ± 3% -88.6% 729834 ± 8% netperf.time.involuntary_context_switches
110445 ± 3% -9.0% 100508 ± 2% netperf.time.minor_page_faults
6907 -62.3% 2601 netperf.time.percent_of_cpu_this_job_got
20894 -64.5% 7412 netperf.time.system_time
102.48 +346.5% 457.58 ± 3% netperf.time.user_time
4016592 ± 11% +25096.8% 1.012e+09 netperf.time.voluntary_context_switches
1.763e+09 ± 2% +39.9% 2.467e+09 netperf.workload
375497 ± 10% +50.7% 565977 ± 12% proc-vmstat.nr_active_anon
1033912 ± 3% +18.9% 1228968 ± 5% proc-vmstat.nr_file_pages
76489 ± 16% -31.5% 52364 ± 14% proc-vmstat.nr_mapped
157787 ± 24% +123.6% 352879 ± 20% proc-vmstat.nr_shmem
78136 -1.9% 76672 proc-vmstat.nr_slab_unreclaimable
375497 ± 10% +50.7% 565977 ± 12% proc-vmstat.nr_zone_active_anon
8.82e+08 ± 2% +11.7% 9.854e+08 proc-vmstat.numa_hit
8.819e+08 ± 2% +11.7% 9.853e+08 proc-vmstat.numa_local
7.042e+09 ± 2% +11.7% 7.868e+09 proc-vmstat.pgalloc_normal
1045876 +3.9% 1086806 proc-vmstat.pgfault
7.042e+09 ± 2% +11.7% 7.867e+09 proc-vmstat.pgfree
55939 ± 3% -7.1% 51982 ± 2% proc-vmstat.pgreuse
20348384 -15.8% 17136384 sched_debug.cfs_rq:/.avg_vruntime.avg
32341494 ± 2% -23.6% 24701136 ± 3% sched_debug.cfs_rq:/.avg_vruntime.max
16723296 ± 3% -25.4% 12475358 sched_debug.cfs_rq:/.avg_vruntime.min
2336769 ± 9% +58.2% 3697913 ± 9% sched_debug.cfs_rq:/.avg_vruntime.stddev
3.08 -22.1% 2.40 sched_debug.cfs_rq:/.h_nr_queued.avg
5.39 ± 4% -20.1% 4.31 ± 5% sched_debug.cfs_rq:/.h_nr_queued.max
1.01 ± 2% -24.6% 0.76 ± 3% sched_debug.cfs_rq:/.h_nr_queued.stddev
4.89 ± 3% -12.5% 4.28 ± 5% sched_debug.cfs_rq:/.h_nr_runnable.max
0.95 ± 2% -20.3% 0.76 ± 3% sched_debug.cfs_rq:/.h_nr_runnable.stddev
20348384 -15.8% 17136384 sched_debug.cfs_rq:/.min_vruntime.avg
32341494 ± 2% -23.6% 24701136 ± 3% sched_debug.cfs_rq:/.min_vruntime.max
16723296 ± 3% -25.4% 12475358 sched_debug.cfs_rq:/.min_vruntime.min
2336769 ± 9% +58.2% 3697913 ± 9% sched_debug.cfs_rq:/.min_vruntime.stddev
709.08 ± 33% +59.9% 1134 ± 12% sched_debug.cfs_rq:/.runnable_avg.min
128.48 ± 8% -13.5% 111.13 ± 3% sched_debug.cfs_rq:/.util_avg.stddev
1775 ± 2% -20.0% 1420 ± 2% sched_debug.cfs_rq:/.util_est.avg
3785 ± 3% -32.8% 2545 ± 3% sched_debug.cfs_rq:/.util_est.max
197.58 ± 45% +137.3% 468.89 ± 3% sched_debug.cfs_rq:/.util_est.min
705.82 ± 3% -41.9% 410.12 ± 3% sched_debug.cfs_rq:/.util_est.stddev
454866 ± 2% -40.2% 271876 ± 5% sched_debug.cpu.avg_idle.avg
36878 ± 7% -76.2% 8790 ± 7% sched_debug.cpu.avg_idle.min
266282 ± 3% -10.9% 237226 ± 5% sched_debug.cpu.avg_idle.stddev
65.17 ± 31% -50.7% 32.10 ± 13% sched_debug.cpu.clock.stddev
192576 -10.0% 173278 sched_debug.cpu.clock_task.avg
183688 -10.2% 164932 sched_debug.cpu.clock_task.min
860.66 ± 3% +39.5% 1200 ± 17% sched_debug.cpu.clock_task.stddev
0.00 ± 29% -44.2% 0.00 ± 12% sched_debug.cpu.next_balance.stddev
3.08 -22.8% 2.38 sched_debug.cpu.nr_running.avg
5.42 ± 3% -20.5% 4.31 ± 5% sched_debug.cpu.nr_running.max
1.00 ± 2% -23.6% 0.77 ± 4% sched_debug.cpu.nr_running.stddev
73412 ± 2% +10555.4% 7822363 sched_debug.cpu.nr_switches.avg
127466 ± 11% +8550.4% 11026390 sched_debug.cpu.nr_switches.max
61589 +7830.8% 4884533 ± 4% sched_debug.cpu.nr_switches.min
8655 ± 12% +23368.3% 2031246 ± 7% sched_debug.cpu.nr_switches.stddev
59.31 -81.0% 11.29 ± 2% perf-stat.i.MPKI
7.435e+09 ± 3% +194.0% 2.186e+10 perf-stat.i.branch-instructions
0.73 +0.3 1.00 perf-stat.i.branch-miss-rate%
54199426 ± 3% +291.1% 2.12e+08 perf-stat.i.branch-misses
60.28 -12.6 47.65 ± 2% perf-stat.i.cache-miss-rate%
2.227e+09 ± 2% -44.8% 1.228e+09 ± 2% perf-stat.i.cache-misses
3.682e+09 ± 2% -30.3% 2.566e+09 perf-stat.i.cache-references
68267 ± 5% +9653.1% 6658139 perf-stat.i.context-switches
8.72 ± 2% -65.8% 2.98 perf-stat.i.cpi
684.12 ± 9% +84.7% 1263 ± 7% perf-stat.i.cpu-migrations
151.44 +80.1% 272.70 ± 2% perf-stat.i.cycles-between-cache-misses
3.76e+10 ± 2% +195.0% 1.109e+11 perf-stat.i.instructions
0.12 ± 2% +184.1% 0.35 perf-stat.i.ipc
0.04 ± 36% -86.1% 0.01 ± 82% perf-stat.i.major-faults
0.07 ± 29% +73057.1% 51.98 perf-stat.i.metric.K/sec
3130 +3.6% 3244 perf-stat.i.minor-faults
3130 +3.6% 3244 perf-stat.i.page-faults
59.27 -81.3% 11.08 ± 2% perf-stat.overall.MPKI
0.73 +0.2 0.97 perf-stat.overall.branch-miss-rate%
60.49 -12.6 47.88 ± 2% perf-stat.overall.cache-miss-rate%
8.72 ± 2% -66.3% 2.94 perf-stat.overall.cpi
147.03 ± 2% +80.5% 265.35 ± 2% perf-stat.overall.cycles-between-cache-misses
0.11 ± 2% +196.4% 0.34 perf-stat.overall.ipc
6473 +110.0% 13594 perf-stat.overall.path-length
7.398e+09 ± 2% +194.3% 2.177e+10 perf-stat.ps.branch-instructions
53832186 ± 3% +292.2% 2.111e+08 perf-stat.ps.branch-misses
2.217e+09 ± 2% -44.8% 1.224e+09 ± 2% perf-stat.ps.cache-misses
3.665e+09 -30.3% 2.556e+09 perf-stat.ps.cache-references
67640 ± 4% +9705.3% 6632391 perf-stat.ps.context-switches
679.53 ± 10% +84.9% 1256 ± 7% perf-stat.ps.cpu-migrations
3.741e+10 ± 2% +195.4% 1.105e+11 perf-stat.ps.instructions
0.04 ± 36% -86.0% 0.01 ± 82% perf-stat.ps.major-faults
3080 +4.6% 3223 perf-stat.ps.minor-faults
3080 +4.6% 3223 perf-stat.ps.page-faults
1.141e+13 ± 2% +193.8% 3.353e+13 perf-stat.total.instructions
4.32 ± 95% -100.0% 0.00 ±223% perf-sched.sch_delay.avg.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
4.98 ± 6% -92.3% 0.38 ±116% perf-sched.sch_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
2.74 ± 28% -59.9% 1.10 ± 21% perf-sched.sch_delay.avg.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
0.59 ± 3% +5402.4% 32.50 ± 48% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.79 ± 60% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.wait_for_completion_state.kernel_clone.__x64_sys_vfork
1.36 ± 38% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.dput.step_into.link_path_walk.part
3.70 ± 6% -71.5% 1.05 ± 20% perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
1.57 ± 54% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
7.59 ± 5% -96.3% 0.28 ±129% perf-sched.sch_delay.avg.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
3.45 ± 55% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.mmput.m_stop.seq_read_iter.seq_read
7.85 ± 13% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.mutex_lock.perf_poll.do_poll.constprop
1.80 ± 93% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.remove_vma.vms_complete_munmap_vmas.__mmap_region.do_mmap
1.04 ± 16% -85.5% 0.15 ±216% perf-sched.sch_delay.avg.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
2.26 ± 53% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.unmap_vmas.vms_clear_ptes.part.0
3.58 ± 26% -100.0% 0.00 perf-sched.sch_delay.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
3.26 ± 37% -99.7% 0.01 ±223% perf-sched.sch_delay.avg.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
4.38 ± 10% -88.4% 0.51 ±223% perf-sched.sch_delay.avg.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
1.51 ± 11% -98.4% 0.02 ±145% perf-sched.sch_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.08 ± 23% -91.8% 0.01 ±143% perf-sched.sch_delay.avg.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
2.61 ± 64% -100.0% 0.00 perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
1.98 ± 31% -98.8% 0.02 ±223% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
5.98 ± 81% -100.0% 0.00 perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown]
3.48 ± 67% -97.9% 0.07 ±195% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
0.27 ± 10% +6194.5% 16.97 ±193% perf-sched.sch_delay.avg.ms.pipe_read.vfs_read.ksys_read.do_syscall_64
1.25 ± 14% -97.9% 0.03 ±138% perf-sched.sch_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
0.06 ± 12% -76.3% 0.01 ± 54% perf-sched.sch_delay.avg.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
0.11 ± 87% +98576.8% 104.10 ± 72% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
3.68 ± 2% +1523.9% 59.75 ± 36% perf-sched.sch_delay.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
7.44 ± 3% -92.5% 0.55 ± 21% perf-sched.sch_delay.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
5.16 ± 7% -76.7% 1.20 ± 23% perf-sched.sch_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
1.47 ± 18% -89.8% 0.15 ±196% perf-sched.sch_delay.avg.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
6.65 ± 57% -100.0% 0.00 ±223% perf-sched.sch_delay.max.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
48.97 ± 17% -87.5% 6.12 ±176% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
29.59 ± 49% +22847.2% 6789 ± 21% perf-sched.sch_delay.max.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
12.47 ± 31% +8381.5% 1057 perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
1.85 ± 63% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.__wait_for_common.wait_for_completion_state.kernel_clone.__x64_sys_vfork
1.89 ± 34% -86.5% 0.26 ±105% perf-sched.sch_delay.max.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
2.91 ± 45% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.dput.step_into.link_path_walk.part
44.99 ± 27% +14850.3% 6726 ± 27% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
14.01 ± 12% -29.2% 9.92 ± 46% perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.kmalloc_reserve.__alloc_skb.tcp_stream_alloc_skb
2.25 ± 44% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
20.58 ± 6% -93.4% 1.35 ±137% perf-sched.sch_delay.max.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
19.88 ± 29% +10263.8% 2060 ± 71% perf-sched.sch_delay.max.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
4.85 ± 54% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.mmput.m_stop.seq_read_iter.seq_read
11.41 ± 20% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.mutex_lock.perf_poll.do_poll.constprop
2.78 ± 71% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.remove_vma.vms_complete_munmap_vmas.__mmap_region.do_mmap
4.15 ± 16% -92.8% 0.30 ±218% perf-sched.sch_delay.max.ms.__cond_resched.stop_one_cpu.sched_exec.bprm_execve.part
4.57 ± 32% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.unmap_vmas.vms_clear_ptes.part.0
6.25 ± 35% -100.0% 0.00 perf-sched.sch_delay.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
6.35 ± 20% -99.9% 0.01 ±223% perf-sched.sch_delay.max.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
7.68 ± 18% -93.4% 0.51 ±223% perf-sched.sch_delay.max.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
7.37 ± 18% -98.1% 0.14 ±136% perf-sched.sch_delay.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.19 ± 37% -95.6% 0.01 ±141% perf-sched.sch_delay.max.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
4.31 ± 65% -100.0% 0.00 perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
6.15 ± 49% -99.6% 0.02 ±223% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
8.42 ± 62% -100.0% 0.00 perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown]
8.41 ± 66% -99.1% 0.07 ±192% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
10.83 ± 36% -87.0% 1.41 ±149% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown]
8.83 ± 15% -99.1% 0.08 ±128% perf-sched.sch_delay.max.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
0.16 ± 21% -87.1% 0.02 ± 61% perf-sched.sch_delay.max.ms.schedule_timeout.kcompactd.kthread.ret_from_fork
33.75 ±193% +4375.3% 1510 ± 74% perf-sched.sch_delay.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
129.45 ±106% +5168.2% 6819 ± 26% perf-sched.sch_delay.max.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
61.49 ± 28% +11976.0% 7425 ± 20% perf-sched.sch_delay.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
180.09 ±145% +4079.9% 7527 ± 26% perf-sched.sch_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
5.14 ± 21% -96.7% 0.17 ±169% perf-sched.sch_delay.max.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
4.61 ± 2% -79.7% 0.93 ± 19% perf-sched.total_sch_delay.average.ms
201.76 ±126% +3741.3% 7750 ± 22% perf-sched.total_sch_delay.max.ms
17.32 ± 2% -82.6% 3.01 ± 19% perf-sched.total_wait_and_delay.average.ms
298061 ± 3% +450.2% 1640034 ± 23% perf-sched.total_wait_and_delay.count.ms
2725 ± 27% +468.8% 15500 ± 22% perf-sched.total_wait_and_delay.max.ms
12.71 ± 3% -83.6% 2.08 ± 19% perf-sched.total_wait_time.average.ms
2725 ± 27% +198.3% 8130 ± 19% perf-sched.total_wait_time.max.ms
8.84 ± 7% -76.2% 2.11 ± 20% perf-sched.wait_and_delay.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
359.01 ±198% -99.1% 3.30 ±223% perf-sched.wait_and_delay.avg.ms.__cond_resched.shmem_get_folio_gfp.shmem_write_begin.generic_perform_write.shmem_file_write_iter
370.11 ± 37% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
229.06 ± 4% -100.0% 0.00 perf-sched.wait_and_delay.avg.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
422.68 ± 21% -92.2% 32.86 ±174% perf-sched.wait_and_delay.avg.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
11.18 ± 2% +969.4% 119.55 ± 36% perf-sched.wait_and_delay.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
15.98 ± 4% -86.3% 2.19 ± 19% perf-sched.wait_and_delay.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
10.75 ± 7% -77.4% 2.42 ± 23% perf-sched.wait_and_delay.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
9037 ± 14% -99.9% 4.67 ±223% perf-sched.wait_and_delay.count.__cond_resched.__release_sock.release_sock.tcp_recvmsg.inet_recvmsg
5102 ± 29% +3851.2% 201595 ± 23% perf-sched.wait_and_delay.count.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
11.17 ± 71% -98.5% 0.17 ±223% perf-sched.wait_and_delay.count.__cond_resched.shmem_get_folio_gfp.shmem_write_begin.generic_perform_write.shmem_file_write_iter
1.00 ±141% +816.7% 9.17 ± 26% perf-sched.wait_and_delay.count.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
5.00 -100.0% 0.00 perf-sched.wait_and_delay.count.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
18.33 ± 4% -100.0% 0.00 perf-sched.wait_and_delay.count.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
649.83 ± 20% -96.9% 19.83 ± 85% perf-sched.wait_and_delay.count.pipe_read.vfs_read.ksys_read.do_syscall_64
28.83 ± 11% -78.6% 6.17 ± 49% perf-sched.wait_and_delay.count.schedule_hrtimeout_range.do_poll.constprop.0.do_sys_poll
19.67 ± 2% -89.8% 2.00 ± 76% perf-sched.wait_and_delay.count.schedule_timeout.kcompactd.kthread.ret_from_fork
122545 -98.2% 2193 ± 30% perf-sched.wait_and_delay.count.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
37034 ± 21% +2124.3% 823774 ± 23% perf-sched.wait_and_delay.count.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
7013 ± 11% -41.7% 4087 ± 22% perf-sched.wait_and_delay.count.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
109813 ± 2% +271.1% 407480 ± 23% perf-sched.wait_and_delay.count.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
732.33 -89.7% 75.50 ± 30% perf-sched.wait_and_delay.count.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
62.30 ± 24% +21492.3% 13452 ± 27% perf-sched.wait_and_delay.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
521.50 ±141% -99.4% 3.30 ±223% perf-sched.wait_and_delay.max.ms.__cond_resched.shmem_get_folio_gfp.shmem_write_begin.generic_perform_write.shmem_file_write_iter
1007 -100.0% 0.00 perf-sched.wait_and_delay.max.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
915.75 ± 9% -100.0% 0.00 perf-sched.wait_and_delay.max.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
228.64 ±127% +5865.5% 13639 ± 26% perf-sched.wait_and_delay.max.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
97.53 ± 50% +15128.8% 14851 ± 20% perf-sched.wait_and_delay.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
1840 ± 10% +340.7% 8112 ± 19% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
1098 ± 17% +1271.0% 15055 ± 26% perf-sched.wait_and_delay.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
2293 ± 25% +218.7% 7309 ± 27% perf-sched.wait_and_delay.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
4.06 ±105% -100.0% 0.00 ±223% perf-sched.wait_time.avg.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
3.65 ± 28% -69.8% 1.10 ± 21% perf-sched.wait_time.avg.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
5.89 ± 6% +605.9% 41.54 ± 24% perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
7.37 ± 57% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.__wait_for_common.wait_for_completion_state.kernel_clone.__x64_sys_vfork
0.32 ± 71% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
0.99 ± 53% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.dput.step_into.link_path_walk.part
5.14 ± 8% -79.5% 1.06 ± 20% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
5.59 ± 9% -64.1% 2.01 ± 80% perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_node_noprof.kmalloc_reserve.__alloc_skb.tcp_stream_alloc_skb
1.48 ± 61% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
7.85 ± 5% -88.2% 0.93 ±141% perf-sched.wait_time.avg.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
3.45 ± 55% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.mmput.m_stop.seq_read_iter.seq_read
8.04 ± 13% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.mutex_lock.perf_poll.do_poll.constprop
1.69 ±103% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.remove_vma.vms_complete_munmap_vmas.__mmap_region.do_mmap
350.00 ±205% -99.5% 1.65 ±223% perf-sched.wait_time.avg.ms.__cond_resched.shmem_get_folio_gfp.shmem_write_begin.generic_perform_write.shmem_file_write_iter
1.93 ± 77% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.unmap_vmas.vms_clear_ptes.part.0
3.56 ± 26% -100.0% 0.00 perf-sched.wait_time.avg.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
366.85 ± 37% -100.0% 0.01 ±223% perf-sched.wait_time.avg.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
59.19 ±133% -99.1% 0.51 ±223% perf-sched.wait_time.avg.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
4.25 ± 9% -74.7% 1.08 ± 81% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
228.98 ± 4% -100.0% 0.07 ±141% perf-sched.wait_time.avg.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
1.78 ± 35% -98.6% 0.02 ±223% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
5.98 ± 81% -100.0% 0.00 perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown]
4.51 ± 66% -98.2% 0.08 ±175% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
422.15 ± 21% -95.0% 21.25 ±214% perf-sched.wait_time.avg.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
7.55 ± 7% +1856.9% 147.84 ± 40% perf-sched.wait_time.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
7.50 ± 3% +697.4% 59.80 ± 36% perf-sched.wait_time.avg.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
8.53 ± 5% -80.8% 1.64 ± 19% perf-sched.wait_time.avg.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
5.59 ± 6% -78.2% 1.22 ± 23% perf-sched.wait_time.avg.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
1.06 ± 28% -98.6% 0.02 ±138% perf-sched.wait_time.avg.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
130.58 ±116% -88.6% 14.92 ± 24% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
6.44 ± 63% -100.0% 0.00 ±223% perf-sched.wait_time.max.ms.__cond_resched.__kmalloc_cache_noprof.perf_event_mmap_event.perf_event_mmap.__mmap_region
26.97 ± 35% +25069.5% 6789 ± 21% perf-sched.wait_time.max.ms.__cond_resched.__release_sock.release_sock.tcp_sendmsg.__sys_sendto
10.50 ± 46% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.__wait_for_common.wait_for_completion_state.kernel_clone.__x64_sys_vfork
1.68 ± 49% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.down_write_killable.exec_mmap.begin_new_exec.load_elf_binary
2.87 ± 47% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.dput.step_into.link_path_walk.part
531.77 ± 89% -96.1% 20.68 ± 37% perf-sched.wait_time.max.ms.__cond_resched.generic_perform_write.shmem_file_write_iter.vfs_write.ksys_write
31.98 ± 26% +20930.2% 6726 ± 27% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.__alloc_skb.tcp_stream_alloc_skb.tcp_sendmsg_locked
16.76 ± 28% -40.7% 9.94 ± 46% perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_node_noprof.kmalloc_reserve.__alloc_skb.tcp_stream_alloc_skb
2.23 ± 45% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.kmem_cache_alloc_noprof.vm_area_alloc.__mmap_new_vma.__mmap_region
20.58 ± 6% -77.6% 4.62 ±138% perf-sched.wait_time.max.ms.__cond_resched.lock_sock_nested.tcp_recvmsg.inet_recvmsg.sock_recvmsg
20.91 ± 16% +9755.3% 2060 ± 71% perf-sched.wait_time.max.ms.__cond_resched.lock_sock_nested.tcp_sendmsg.__sys_sendto.__x64_sys_sendto
4.85 ± 54% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.mmput.m_stop.seq_read_iter.seq_read
12.63 ± 32% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.mutex_lock.perf_poll.do_poll.constprop
2.78 ± 71% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.remove_vma.vms_complete_munmap_vmas.__mmap_region.do_mmap
513.55 ±144% -99.7% 1.65 ±223% perf-sched.wait_time.max.ms.__cond_resched.shmem_get_folio_gfp.shmem_write_begin.generic_perform_write.shmem_file_write_iter
3.68 ± 64% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.unmap_vmas.vms_clear_ptes.part.0
6.25 ± 35% -100.0% 0.00 perf-sched.wait_time.max.ms.__cond_resched.zap_pte_range.zap_pmd_range.isra.0
1001 -100.0% 0.01 ±223% perf-sched.wait_time.max.ms.__x64_sys_pause.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
339.35 ±138% -99.9% 0.51 ±223% perf-sched.wait_time.max.ms.do_nanosleep.hrtimer_nanosleep.common_nsleep.__x64_sys_clock_nanosleep
21.71 ± 47% -87.3% 2.75 ± 88% perf-sched.wait_time.max.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
915.66 ± 9% -100.0% 0.14 ±141% perf-sched.wait_time.max.ms.irq_thread.kthread.ret_from_fork.ret_from_fork_asm
6.15 ± 49% -99.6% 0.02 ±223% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
8.42 ± 62% -100.0% 0.00 perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown]
9.31 ± 65% -99.1% 0.08 ±173% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_call_function_single.[unknown].[unknown]
11.32 ± 34% -87.5% 1.41 ±149% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown]
283.33 ± 21% +484.6% 1656 ± 66% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
121.31 ±117% +5522.0% 6819 ± 26% perf-sched.wait_time.max.ms.schedule_timeout.wait_woken.sk_stream_wait_memory.tcp_sendmsg_locked
54.73 ± 45% +14104.3% 7773 ± 22% perf-sched.wait_time.max.ms.schedule_timeout.wait_woken.sk_wait_data.tcp_recvmsg_locked
1840 ± 10% +337.4% 8051 ± 18% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
1008 +646.1% 7527 ± 26% perf-sched.wait_time.max.ms.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.[unknown]
5.09 ± 22% -99.3% 0.04 ±129% perf-sched.wait_time.max.ms.wait_for_partner.fifo_open.do_dentry_open.vfs_open
2293 ± 25% +218.7% 7309 ± 27% perf-sched.wait_time.max.ms.worker_thread.kthread.ret_from_fork.ret_from_fork_asm
***************************************************************************************************
lkp-skl-fpga01: 104 threads 2 sockets (Skylake) with 192G memory
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase:
gcc-12/performance/x86_64-rhel-9.4/thread/100%/debian-12-x86_64-20240206.cgz/lkp-skl-fpga01/fallocate1/will-it-scale
commit:
8c57b687e8 ("mm, bpf: Introduce free_pages_nolock()")
01d37228d3 ("memcg: Use trylock to access memcg stock_lock.")
8c57b687e8331eb8 01d37228d331047a0bbbd1026ce
---------------- ---------------------------
%stddev %change %stddev
\ | \
125406 -10.4% 112308 meminfo.KReclaimable
125406 -10.4% 112308 meminfo.SReclaimable
0.03 -0.0 0.02 ± 10% mpstat.cpu.all.soft%
2.85 ± 3% -1.6 1.27 ± 26% mpstat.cpu.all.usr%
3597 -42.0% 2086 ± 5% vmstat.system.cs
119319 -1.1% 117948 vmstat.system.in
4530881 ± 3% -68.8% 1411930 ± 12% will-it-scale.104.threads
43565 ± 3% -68.8% 13575 ± 12% will-it-scale.per_thread_ops
4530881 ± 3% -68.8% 1411930 ± 12% will-it-scale.workload
1.375e+09 ± 3% -64.9% 4.824e+08 ± 23% numa-numastat.node0.local_node
1.376e+09 ± 3% -64.9% 4.826e+08 ± 23% numa-numastat.node0.numa_hit
1.356e+09 ± 3% -72.8% 3.69e+08 ± 19% numa-numastat.node1.local_node
1.357e+09 ± 3% -72.8% 3.691e+08 ± 19% numa-numastat.node1.numa_hit
1.376e+09 ± 3% -64.9% 4.824e+08 ± 23% numa-vmstat.node0.numa_hit
1.375e+09 ± 3% -64.9% 4.823e+08 ± 23% numa-vmstat.node0.numa_local
1.357e+09 ± 3% -72.8% 3.69e+08 ± 19% numa-vmstat.node1.numa_hit
1.356e+09 ± 3% -72.8% 3.689e+08 ± 19% numa-vmstat.node1.numa_local
3542 ± 89% +1.7e+05% 6202792 ±222% sched_debug.cfs_rq:/.runnable_avg.max
6301 -34.5% 4126 ± 3% sched_debug.cpu.nr_switches.avg
4167 ± 2% -62.6% 1558 ± 14% sched_debug.cpu.nr_switches.min
2174 ± 7% +10.9% 2410 ± 6% sched_debug.cpu.nr_switches.stddev
136.67 ± 26% +803.4% 1234 ± 10% perf-c2c.DRAM.local
366.17 ± 21% +1413.7% 5542 ± 10% perf-c2c.DRAM.remote
6364 ± 6% +208.8% 19652 ± 2% perf-c2c.HITM.local
154.83 ± 6% +834.1% 1446 ± 5% perf-c2c.HITM.remote
6519 ± 6% +223.7% 21099 ± 2% perf-c2c.HITM.total
300086 -2.3% 293262 proc-vmstat.nr_active_anon
129160 -5.6% 121896 proc-vmstat.nr_shmem
31344 -10.4% 28076 proc-vmstat.nr_slab_reclaimable
300085 -2.3% 293262 proc-vmstat.nr_zone_active_anon
2.733e+09 ± 3% -68.8% 8.517e+08 ± 12% proc-vmstat.numa_hit
2.732e+09 ± 3% -68.8% 8.514e+08 ± 12% proc-vmstat.numa_local
2.73e+09 ± 3% -68.8% 8.514e+08 ± 12% proc-vmstat.pgalloc_normal
2.73e+09 ± 3% -68.8% 8.513e+08 ± 12% proc-vmstat.pgfree
0.18 ± 39% +3637.9% 6.73 ± 23% perf-stat.i.MPKI
1.074e+10 ± 3% -57.2% 4.591e+09 ± 5% perf-stat.i.branch-instructions
66071729 ± 3% -55.2% 29625865 ± 8% perf-stat.i.branch-misses
11.40 ± 27% +21.7 33.11 ± 7% perf-stat.i.cache-miss-rate%
9285062 ± 35% +1457.0% 1.446e+08 ± 27% perf-stat.i.cache-misses
79584428 ± 8% +454.4% 4.412e+08 ± 30% perf-stat.i.cache-references
3563 -42.6% 2043 ± 5% perf-stat.i.context-switches
5.51 ± 3% +146.8% 13.60 ± 6% perf-stat.i.cpi
148.01 -6.4% 138.61 perf-stat.i.cpu-migrations
35747 ± 29% -94.1% 2107 ± 18% perf-stat.i.cycles-between-cache-misses
5.253e+10 ± 3% -59.2% 2.144e+10 ± 6% perf-stat.i.instructions
0.18 ± 3% -58.7% 0.08 ± 6% perf-stat.i.ipc
0.18 ± 39% +3650.7% 6.72 ± 23% perf-stat.overall.MPKI
0.62 +0.0 0.64 ± 4% perf-stat.overall.branch-miss-rate%
11.42 ± 26% +21.7 33.11 ± 7% perf-stat.overall.cache-miss-rate%
5.52 ± 3% +145.9% 13.57 ± 5% perf-stat.overall.cpi
34676 ± 28% -93.9% 2109 ± 18% perf-stat.overall.cycles-between-cache-misses
0.18 ± 3% -59.2% 0.07 ± 6% perf-stat.overall.ipc
3496124 +31.6% 4602126 ± 6% perf-stat.overall.path-length
1.07e+10 ± 3% -57.3% 4.571e+09 ± 5% perf-stat.ps.branch-instructions
65846459 ± 3% -55.3% 29454957 ± 8% perf-stat.ps.branch-misses
9255594 ± 35% +1455.7% 1.44e+08 ± 27% perf-stat.ps.cache-misses
79350913 ± 8% +453.9% 4.395e+08 ± 30% perf-stat.ps.cache-references
3551 -42.6% 2038 ± 5% perf-stat.ps.context-switches
147.51 -6.6% 137.73 perf-stat.ps.cpu-migrations
5.236e+10 ± 3% -59.2% 2.135e+10 ± 6% perf-stat.ps.instructions
2724 -1.8% 2676 perf-stat.ps.minor-faults
2724 -1.8% 2676 perf-stat.ps.page-faults
1.584e+13 ± 3% -59.3% 6.45e+12 ± 6% perf-stat.total.instructions
0.07 ±164% +606.2% 0.52 ± 34% perf-sched.sch_delay.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.vma_alloc_folio_noprof
0.83 ± 11% +182.4% 2.35 ± 6% perf-sched.sch_delay.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
0.06 ±179% +2166.3% 1.29 ± 48% perf-sched.sch_delay.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.vms_clear_ptes.part
0.19 ± 20% +154.9% 0.49 ± 44% perf-sched.sch_delay.avg.ms.__cond_resched.__wait_for_common.affine_move_task.__set_cpus_allowed_ptr.__sched_setaffinity
0.02 ±223% +8165.9% 1.25 ±138% perf-sched.sch_delay.avg.ms.__cond_resched.down_read.walk_component.link_path_walk.part
0.47 ± 9% +111.3% 0.99 ± 10% perf-sched.sch_delay.avg.ms.__cond_resched.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64
0.50 ± 6% +92.8% 0.97 ± 8% perf-sched.sch_delay.avg.ms.__cond_resched.shmem_undo_range.shmem_setattr.notify_change.do_truncate
0.12 ± 13% +189.9% 0.33 ± 27% perf-sched.sch_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.19 ±104% +505.3% 1.13 ± 47% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
0.28 ± 68% +249.6% 0.97 ± 27% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
0.19 ± 52% +355.0% 0.86 ± 52% perf-sched.sch_delay.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown]
0.05 ± 57% +557.6% 0.30 ±107% perf-sched.sch_delay.avg.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
0.16 ± 14% +80.9% 0.30 ± 7% perf-sched.sch_delay.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
17.11 ±108% -99.9% 0.02 ±223% perf-sched.sch_delay.avg.ms.schedule_timeout.khugepaged_wait_work.khugepaged.kthread
0.02 ± 14% +109.3% 0.04 ± 15% perf-sched.sch_delay.avg.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
6.68 ± 62% -49.1% 3.40 ± 41% perf-sched.sch_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
0.12 ±179% +518.2% 0.76 ± 43% perf-sched.sch_delay.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.vma_alloc_folio_noprof
3.87 ± 7% +69.2% 6.54 ± 47% perf-sched.sch_delay.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
0.06 ±169% +3507.7% 2.19 ± 49% perf-sched.sch_delay.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.vms_clear_ptes.part
0.02 ±223% +9304.4% 1.43 ±119% perf-sched.sch_delay.max.ms.__cond_resched.down_read.walk_component.link_path_walk.part
3.19 ± 48% -78.0% 0.70 ±158% perf-sched.sch_delay.max.ms.__cond_resched.down_write.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate
0.26 ±132% +953.7% 2.71 ± 40% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
0.61 ± 64% +559.5% 4.03 ± 35% perf-sched.sch_delay.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
0.18 ± 48% +1582.7% 3.00 ±133% perf-sched.sch_delay.max.ms.schedule_hrtimeout_range.ep_poll.do_epoll_wait.__x64_sys_epoll_wait
17.11 ±108% -99.9% 0.02 ±223% perf-sched.sch_delay.max.ms.schedule_timeout.khugepaged_wait_work.khugepaged.kthread
0.35 ± 6% +87.0% 0.66 ± 7% perf-sched.total_sch_delay.average.ms
75.94 ± 3% +41.9% 107.78 ± 6% perf-sched.total_wait_and_delay.average.ms
20535 ± 4% -41.7% 11978 ± 9% perf-sched.total_wait_and_delay.count.ms
75.58 ± 3% +41.7% 107.13 ± 6% perf-sched.total_wait_time.average.ms
0.94 ± 9% +111.4% 1.98 ± 10% perf-sched.wait_and_delay.avg.ms.__cond_resched.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64
1.00 ± 6% +92.7% 1.93 ± 8% perf-sched.wait_and_delay.avg.ms.__cond_resched.shmem_undo_range.shmem_setattr.notify_change.do_truncate
87.71 ± 50% +365.7% 408.48 ± 14% perf-sched.wait_and_delay.avg.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.97 ± 30% -76.5% 0.23 ±141% perf-sched.wait_and_delay.avg.ms.__cond_resched.vfs_fallocate.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe
3.03 ±147% +782.6% 26.71 ± 21% perf-sched.wait_and_delay.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
2.05 ± 17% +94.1% 3.98 ± 15% perf-sched.wait_and_delay.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
175.25 ± 3% +199.2% 524.30 ± 11% perf-sched.wait_and_delay.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
3879 ± 5% -47.8% 2024 ± 7% perf-sched.wait_and_delay.count.__cond_resched.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64
328.67 ± 16% +181.5% 925.17 ± 29% perf-sched.wait_and_delay.count.__cond_resched.shmem_inode_acct_blocks.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate
4765 ± 3% -52.5% 2262 ± 3% perf-sched.wait_and_delay.count.__cond_resched.shmem_undo_range.shmem_setattr.notify_change.do_truncate
154.83 ± 13% -87.0% 20.17 ±141% perf-sched.wait_and_delay.count.__cond_resched.vfs_fallocate.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe
551.67 ± 10% -29.9% 386.83 ± 11% perf-sched.wait_and_delay.count.irqentry_exit_to_user_mode.asm_sysvec_apic_timer_interrupt.[unknown]
5573 ± 6% -75.9% 1341 ± 22% perf-sched.wait_and_delay.count.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
297.16 ± 49% +551.6% 1936 ± 28% perf-sched.wait_and_delay.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
34.48 ±114% -91.8% 2.84 ±141% perf-sched.wait_and_delay.max.ms.__cond_resched.vfs_fallocate.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe
370.78 ±148% +321.1% 1561 ± 48% perf-sched.wait_and_delay.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
1077 ± 3% +230.9% 3566 ± 10% perf-sched.wait_and_delay.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
0.07 ±162% +641.0% 0.50 ± 33% perf-sched.wait_time.avg.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.vma_alloc_folio_noprof
0.77 ± 11% +569.6% 5.18 ±125% perf-sched.wait_time.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
0.05 ±190% +2300.0% 1.29 ± 48% perf-sched.wait_time.avg.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.vms_clear_ptes.part
0.02 ±223% +7438.5% 1.14 ±156% perf-sched.wait_time.avg.ms.__cond_resched.down_read.walk_component.link_path_walk.part
0.47 ± 9% +111.3% 0.99 ± 10% perf-sched.wait_time.avg.ms.__cond_resched.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64
0.50 ± 6% +92.8% 0.97 ± 8% perf-sched.wait_time.avg.ms.__cond_resched.shmem_undo_range.shmem_setattr.notify_change.do_truncate
99.70 ± 24% +309.7% 408.42 ± 14% perf-sched.wait_time.avg.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
3.03 ±147% +782.6% 26.71 ± 21% perf-sched.wait_time.avg.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
1.93 ± 18% +88.5% 3.64 ± 14% perf-sched.wait_time.avg.ms.do_wait.kernel_wait4.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.13 ± 82% +634.0% 0.93 ± 68% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
0.16 ±129% +455.8% 0.88 ± 31% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
0.19 ± 52% +355.0% 0.86 ± 52% perf-sched.wait_time.avg.ms.irqentry_exit_to_user_mode.asm_sysvec_reschedule_ipi.[unknown]
4.61 ± 13% +90.0% 8.76 ± 23% perf-sched.wait_time.avg.ms.schedule_timeout.__wait_for_common.wait_for_completion_state.kernel_clone
17.15 ±108% -99.9% 0.02 ±223% perf-sched.wait_time.avg.ms.schedule_timeout.khugepaged_wait_work.khugepaged.kthread
175.21 ± 3% +199.2% 524.18 ± 11% perf-sched.wait_time.avg.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
6.68 ± 62% -49.1% 3.40 ± 41% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.shmem_alloc_folio
0.12 ±179% +518.2% 0.76 ± 43% perf-sched.wait_time.max.ms.__cond_resched.__alloc_frozen_pages_noprof.alloc_pages_mpol.folio_alloc_mpol_noprof.vma_alloc_folio_noprof
3.87 ± 7% +4323.6% 171.00 ±216% perf-sched.wait_time.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
0.06 ±169% +3507.7% 2.19 ± 49% perf-sched.wait_time.max.ms.__cond_resched.__tlb_batch_free_encoded_pages.tlb_finish_mmu.vms_clear_ptes.part
0.02 ±223% +8481.3% 1.30 ±134% perf-sched.wait_time.max.ms.__cond_resched.down_read.walk_component.link_path_walk.part
3.19 ± 48% -78.0% 0.70 ±158% perf-sched.wait_time.max.ms.__cond_resched.down_write.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate
329.79 ± 26% +487.1% 1936 ± 28% perf-sched.wait_time.max.ms.__cond_resched.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
370.78 ±148% +321.1% 1561 ± 48% perf-sched.wait_time.max.ms.do_task_dead.do_exit.do_group_exit.__x64_sys_exit_group.x64_sys_call
0.19 ±108% +1175.5% 2.48 ± 51% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown]
0.36 ±121% +1007.0% 4.03 ± 35% perf-sched.wait_time.max.ms.irqentry_exit_to_user_mode.asm_exc_page_fault.[unknown].[unknown]
17.15 ±108% -99.9% 0.02 ±223% perf-sched.wait_time.max.ms.schedule_timeout.khugepaged_wait_work.khugepaged.kthread
5.22 +11.9% 5.85 ± 4% perf-sched.wait_time.max.ms.schedule_timeout.rcu_gp_fqs_loop.rcu_gp_kthread.kthread
1077 ± 3% +230.9% 3566 ± 10% perf-sched.wait_time.max.ms.smpboot_thread_fn.kthread.ret_from_fork.ret_from_fork_asm
50.56 -6.0 44.54 ± 10% perf-profile.calltrace.cycles-pp.ftruncate64
50.54 -6.0 44.53 ± 10% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.ftruncate64
50.52 -6.0 44.53 ± 10% perf-profile.calltrace.cycles-pp.do_sys_ftruncate.do_syscall_64.entry_SYSCALL_64_after_hwframe.ftruncate64
50.52 -6.0 44.53 ± 10% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.ftruncate64
50.52 -6.0 44.52 ± 10% perf-profile.calltrace.cycles-pp.do_ftruncate.do_sys_ftruncate.do_syscall_64.entry_SYSCALL_64_after_hwframe.ftruncate64
50.52 -6.0 44.52 ± 10% perf-profile.calltrace.cycles-pp.do_truncate.do_ftruncate.do_sys_ftruncate.do_syscall_64.entry_SYSCALL_64_after_hwframe
50.52 -6.0 44.52 ± 10% perf-profile.calltrace.cycles-pp.notify_change.do_truncate.do_ftruncate.do_sys_ftruncate.do_syscall_64
50.51 -6.0 44.52 ± 10% perf-profile.calltrace.cycles-pp.shmem_setattr.notify_change.do_truncate.do_ftruncate.do_sys_ftruncate
50.48 -6.0 44.51 ± 10% perf-profile.calltrace.cycles-pp.shmem_undo_range.shmem_setattr.notify_change.do_truncate.do_ftruncate
41.58 -5.3 36.26 ± 11% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.folio_lruvec_lock_irqsave.__page_cache_release.folios_put_refs.shmem_undo_range
41.59 -5.3 36.26 ± 11% perf-profile.calltrace.cycles-pp.folio_lruvec_lock_irqsave.__page_cache_release.folios_put_refs.shmem_undo_range.shmem_setattr
41.56 -5.3 36.25 ± 12% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.folio_lruvec_lock_irqsave.__page_cache_release.folios_put_refs
43.12 -5.2 37.90 ± 11% perf-profile.calltrace.cycles-pp.folios_put_refs.shmem_undo_range.shmem_setattr.notify_change.do_truncate
42.26 -5.2 37.05 ± 11% perf-profile.calltrace.cycles-pp.__page_cache_release.folios_put_refs.shmem_undo_range.shmem_setattr.notify_change
4.64 ± 2% -0.5 4.13 ± 9% perf-profile.calltrace.cycles-pp.lru_add_drain_cpu.__folio_batch_release.shmem_undo_range.shmem_setattr.notify_change
4.65 ± 2% -0.5 4.13 ± 9% perf-profile.calltrace.cycles-pp.__folio_batch_release.shmem_undo_range.shmem_setattr.notify_change.do_truncate
4.64 ± 2% -0.5 4.13 ± 9% perf-profile.calltrace.cycles-pp.folio_batch_move_lru.lru_add_drain_cpu.__folio_batch_release.shmem_undo_range.shmem_setattr
4.58 ± 2% -0.5 4.08 ± 9% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.folio_lruvec_lock_irqsave.folio_batch_move_lru.lru_add_drain_cpu.__folio_batch_release
4.58 ± 2% -0.5 4.08 ± 9% perf-profile.calltrace.cycles-pp.folio_lruvec_lock_irqsave.folio_batch_move_lru.lru_add_drain_cpu.__folio_batch_release.shmem_undo_range
4.57 ± 2% -0.5 4.08 ± 9% perf-profile.calltrace.cycles-pp.native_queued_spin_lock_slowpath._raw_spin_lock_irqsave.folio_lruvec_lock_irqsave.folio_batch_move_lru.lru_add_drain_cpu
0.63 ± 3% +0.2 0.82 ± 7% perf-profile.calltrace.cycles-pp.lru_gen_add_folio.lru_add.folio_batch_move_lru.__folio_batch_add_and_move.shmem_alloc_and_add_folio
0.00 +0.6 0.63 ± 19% perf-profile.calltrace.cycles-pp.__lruvec_stat_mod_folio.shmem_add_to_page_cache.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate
1.47 ± 2% +0.8 2.23 ± 15% perf-profile.calltrace.cycles-pp.shmem_add_to_page_cache.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate.vfs_fallocate
0.00 +0.8 0.82 ± 32% perf-profile.calltrace.cycles-pp.propagate_protected_usage.page_counter_try_charge.try_charge_memcg.charge_memcg.__mem_cgroup_charge
0.00 +0.9 0.92 ± 9% perf-profile.calltrace.cycles-pp.__mod_memcg_lruvec_state.__lruvec_stat_mod_folio.shmem_update_stats.shmem_add_to_page_cache.shmem_alloc_and_add_folio
0.66 ± 10% +1.0 1.69 ± 14% perf-profile.calltrace.cycles-pp.filemap_unaccount_folio.__filemap_remove_folio.filemap_remove_folio.truncate_inode_folio.shmem_undo_range
0.59 ± 12% +1.1 1.68 ± 14% perf-profile.calltrace.cycles-pp.__lruvec_stat_mod_folio.filemap_unaccount_folio.__filemap_remove_folio.filemap_remove_folio.truncate_inode_folio
0.09 ±223% +1.2 1.28 ± 7% perf-profile.calltrace.cycles-pp.shmem_update_stats.shmem_add_to_page_cache.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate
0.08 ±223% +1.2 1.31 ± 18% perf-profile.calltrace.cycles-pp.__mod_memcg_lruvec_state.__lruvec_stat_mod_folio.filemap_unaccount_folio.__filemap_remove_folio.filemap_remove_folio
0.00 +1.3 1.27 ± 7% perf-profile.calltrace.cycles-pp.__lruvec_stat_mod_folio.shmem_update_stats.shmem_add_to_page_cache.shmem_alloc_and_add_folio.shmem_get_folio_gfp
49.10 +5.9 55.01 ± 8% perf-profile.calltrace.cycles-pp.fallocate64
46.15 +7.7 53.86 ± 7% perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.fallocate64
45.49 +8.1 53.55 ± 7% perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.fallocate64
0.00 +8.1 8.09 ± 56% perf-profile.calltrace.cycles-pp.page_counter_try_charge.try_charge_memcg.charge_memcg.__mem_cgroup_charge.shmem_alloc_and_add_folio
45.32 +8.2 53.49 ± 7% perf-profile.calltrace.cycles-pp.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe.fallocate64
45.09 +8.3 53.41 ± 7% perf-profile.calltrace.cycles-pp.vfs_fallocate.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe.fallocate64
44.92 +8.4 53.36 ± 7% perf-profile.calltrace.cycles-pp.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64.entry_SYSCALL_64_after_hwframe
43.94 +9.0 52.96 ± 7% perf-profile.calltrace.cycles-pp.shmem_get_folio_gfp.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate.do_syscall_64
43.39 +9.4 52.79 ± 7% perf-profile.calltrace.cycles-pp.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate.vfs_fallocate.__x64_sys_fallocate
0.00 +11.7 11.68 ± 57% perf-profile.calltrace.cycles-pp.try_charge_memcg.charge_memcg.__mem_cgroup_charge.shmem_alloc_and_add_folio.shmem_get_folio_gfp
0.00 +12.7 12.68 ± 50% perf-profile.calltrace.cycles-pp.charge_memcg.__mem_cgroup_charge.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate
0.34 ±103% +14.2 14.51 ± 39% perf-profile.calltrace.cycles-pp.__mem_cgroup_charge.shmem_alloc_and_add_folio.shmem_get_folio_gfp.shmem_fallocate.vfs_fallocate
83.55 -9.1 74.42 ± 9% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
83.55 -9.1 74.44 ± 9% perf-profile.children.cycles-pp.folio_lruvec_lock_irqsave
83.52 -9.0 74.51 ± 9% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
50.56 -6.0 44.54 ± 10% perf-profile.children.cycles-pp.ftruncate64
50.52 -6.0 44.52 ± 10% perf-profile.children.cycles-pp.do_ftruncate
50.52 -6.0 44.53 ± 10% perf-profile.children.cycles-pp.do_sys_ftruncate
50.52 -6.0 44.52 ± 10% perf-profile.children.cycles-pp.do_truncate
50.52 -6.0 44.52 ± 10% perf-profile.children.cycles-pp.notify_change
50.51 -6.0 44.52 ± 10% perf-profile.children.cycles-pp.shmem_setattr
50.50 -6.0 44.52 ± 10% perf-profile.children.cycles-pp.shmem_undo_range
43.26 -5.2 38.02 ± 11% perf-profile.children.cycles-pp.folios_put_refs
42.30 -5.2 37.12 ± 11% perf-profile.children.cycles-pp.__page_cache_release
1.22 ± 6% -0.8 0.39 ± 18% perf-profile.children.cycles-pp.shmem_inode_acct_blocks
1.16 ± 5% -0.8 0.38 ± 16% perf-profile.children.cycles-pp.shmem_alloc_folio
1.20 ± 6% -0.8 0.44 ± 23% perf-profile.children.cycles-pp.syscall_return_via_sysret
1.22 ± 5% -0.7 0.54 ± 34% perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
0.99 ± 6% -0.7 0.32 ± 15% perf-profile.children.cycles-pp.folio_alloc_mpol_noprof
0.96 ± 6% -0.6 0.32 ± 16% perf-profile.children.cycles-pp.alloc_pages_mpol
0.80 ± 4% -0.5 0.27 ± 19% perf-profile.children.cycles-pp.xas_store
0.78 ± 6% -0.5 0.26 ± 18% perf-profile.children.cycles-pp.__alloc_frozen_pages_noprof
4.65 ± 2% -0.5 4.13 ± 9% perf-profile.children.cycles-pp.__folio_batch_release
4.64 ± 2% -0.5 4.13 ± 9% perf-profile.children.cycles-pp.lru_add_drain_cpu
0.54 ± 6% -0.4 0.16 ± 10% perf-profile.children.cycles-pp.security_vm_enough_memory_mm
0.50 ± 6% -0.3 0.16 ± 13% perf-profile.children.cycles-pp.entry_SYSCALL_64
0.48 ± 5% -0.3 0.17 ± 18% perf-profile.children.cycles-pp.get_page_from_freelist
0.36 -0.3 0.11 ± 16% perf-profile.children.cycles-pp.free_unref_folios
0.32 ± 4% -0.2 0.10 ± 15% perf-profile.children.cycles-pp.xas_load
0.38 ± 2% -0.2 0.16 ± 28% perf-profile.children.cycles-pp.find_lock_entries
0.31 ± 6% -0.2 0.11 ± 20% perf-profile.children.cycles-pp.rmqueue
0.24 ± 3% -0.2 0.07 ± 16% perf-profile.children.cycles-pp.xas_clear_mark
0.23 ± 4% -0.2 0.07 ± 16% perf-profile.children.cycles-pp.filemap_get_entry
0.21 ± 4% -0.1 0.07 ± 14% perf-profile.children.cycles-pp.xas_init_marks
0.16 ± 4% -0.1 0.03 ±100% perf-profile.children.cycles-pp.__mod_node_page_state
0.19 ± 7% -0.1 0.06 ± 14% perf-profile.children.cycles-pp.__cond_resched
0.20 ± 4% -0.1 0.08 ± 41% perf-profile.children.cycles-pp.__dquot_alloc_space
0.18 ± 3% -0.1 0.06 ± 23% perf-profile.children.cycles-pp.__folio_cancel_dirty
0.15 ± 7% -0.1 0.04 ± 45% perf-profile.children.cycles-pp.fdget
0.16 ± 8% -0.1 0.06 ± 13% perf-profile.children.cycles-pp.file_modified
0.15 ± 6% -0.1 0.05 ±111% perf-profile.children.cycles-pp.noop_dirty_folio
0.17 ± 4% -0.1 0.08 ± 4% perf-profile.children.cycles-pp.cgroup_rstat_updated
0.04 ± 44% +0.1 0.14 ± 16% perf-profile.children.cycles-pp.handle_internal_command
0.04 ± 44% +0.1 0.14 ± 16% perf-profile.children.cycles-pp.main
0.04 ± 44% +0.1 0.14 ± 16% perf-profile.children.cycles-pp.run_builtin
0.08 ± 5% +0.1 0.20 ± 62% perf-profile.children.cycles-pp.page_counter_cancel
0.00 +0.1 0.12 ± 20% perf-profile.children.cycles-pp.shmem_write_begin
0.00 +0.1 0.13 ± 19% perf-profile.children.cycles-pp.generic_perform_write
0.00 +0.1 0.13 ± 18% perf-profile.children.cycles-pp.shmem_file_write_iter
0.00 +0.1 0.13 ± 18% perf-profile.children.cycles-pp.vfs_write
0.01 ±223% +0.1 0.14 ± 16% perf-profile.children.cycles-pp.__cmd_record
0.01 ±223% +0.1 0.14 ± 16% perf-profile.children.cycles-pp.cmd_record
0.00 +0.1 0.13 ± 17% perf-profile.children.cycles-pp.ksys_write
0.00 +0.1 0.14 ± 15% perf-profile.children.cycles-pp.record__pushfn
0.00 +0.1 0.14 ± 15% perf-profile.children.cycles-pp.writen
0.00 +0.1 0.14 ± 18% perf-profile.children.cycles-pp.perf_mmap__push
0.00 +0.1 0.14 ± 18% perf-profile.children.cycles-pp.record__mmap_read_evlist
0.00 +0.1 0.14 ± 18% perf-profile.children.cycles-pp.write
0.09 ± 8% +0.1 0.23 ± 55% perf-profile.children.cycles-pp.page_counter_uncharge
0.10 ± 9% +0.2 0.26 ± 46% perf-profile.children.cycles-pp.uncharge_batch
0.66 ± 3% +0.2 0.86 ± 7% perf-profile.children.cycles-pp.lru_gen_add_folio
0.23 ± 8% +0.4 0.58 ± 19% perf-profile.children.cycles-pp.__mem_cgroup_uncharge_folios
1.50 ± 2% +0.7 2.24 ± 15% perf-profile.children.cycles-pp.shmem_add_to_page_cache
0.46 ± 6% +0.8 1.28 ± 7% perf-profile.children.cycles-pp.shmem_update_stats
0.34 ± 36% +0.8 1.18 ± 39% perf-profile.children.cycles-pp.get_mem_cgroup_from_mm
0.00 +0.9 0.85 ± 32% perf-profile.children.cycles-pp.propagate_protected_usage
0.67 ± 10% +1.0 1.69 ± 14% perf-profile.children.cycles-pp.filemap_unaccount_folio
96.78 +1.8 98.62 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
96.10 +2.2 98.31 perf-profile.children.cycles-pp.do_syscall_64
1.31 ± 8% +2.3 3.59 ± 11% perf-profile.children.cycles-pp.__lruvec_stat_mod_folio
1.14 ± 15% +2.8 3.97 ± 14% perf-profile.children.cycles-pp.__mod_memcg_lruvec_state
49.17 +5.9 55.03 ± 8% perf-profile.children.cycles-pp.fallocate64
0.00 +8.1 8.14 ± 56% perf-profile.children.cycles-pp.page_counter_try_charge
45.32 +8.2 53.49 ± 7% perf-profile.children.cycles-pp.__x64_sys_fallocate
45.09 +8.3 53.41 ± 7% perf-profile.children.cycles-pp.vfs_fallocate
44.94 +8.4 53.36 ± 7% perf-profile.children.cycles-pp.shmem_fallocate
43.99 +9.1 53.08 ± 7% perf-profile.children.cycles-pp.shmem_get_folio_gfp
43.48 +9.5 52.93 ± 7% perf-profile.children.cycles-pp.shmem_alloc_and_add_folio
0.14 ± 6% +11.6 11.74 ± 57% perf-profile.children.cycles-pp.try_charge_memcg
0.20 ± 9% +12.5 12.72 ± 50% perf-profile.children.cycles-pp.charge_memcg
0.58 ± 24% +14.0 14.54 ± 39% perf-profile.children.cycles-pp.__mem_cgroup_charge
83.52 -9.0 74.51 ± 9% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
1.19 ± 6% -0.8 0.44 ± 23% perf-profile.self.cycles-pp.syscall_return_via_sysret
1.21 ± 5% -0.7 0.54 ± 35% perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
0.68 ± 6% -0.4 0.32 ± 38% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
0.46 ± 5% -0.3 0.13 ± 16% perf-profile.self.cycles-pp.xas_store
0.44 ± 6% -0.3 0.14 ± 13% perf-profile.self.cycles-pp.entry_SYSCALL_64
0.40 ± 8% -0.2 0.15 ± 47% perf-profile.self.cycles-pp.lru_gen_add_folio
0.35 ± 2% -0.2 0.14 ± 34% perf-profile.self.cycles-pp.lru_gen_del_folio
0.31 ± 6% -0.2 0.10 ± 22% perf-profile.self.cycles-pp.shmem_fallocate
0.28 ± 5% -0.2 0.09 ± 13% perf-profile.self.cycles-pp.security_vm_enough_memory_mm
0.31 ± 4% -0.2 0.14 ± 29% perf-profile.self.cycles-pp.find_lock_entries
0.30 ± 7% -0.2 0.12 ± 63% perf-profile.self.cycles-pp.shmem_add_to_page_cache
0.24 ± 7% -0.2 0.08 ± 12% perf-profile.self.cycles-pp.__alloc_frozen_pages_noprof
0.22 ± 5% -0.2 0.07 ± 14% perf-profile.self.cycles-pp.xas_load
0.22 ± 3% -0.2 0.07 ± 17% perf-profile.self.cycles-pp.xas_clear_mark
0.26 ± 9% -0.1 0.11 ± 49% perf-profile.self.cycles-pp.lru_add
0.26 -0.1 0.11 ± 30% perf-profile.self.cycles-pp.folios_put_refs
0.19 ± 7% -0.1 0.06 ± 13% perf-profile.self.cycles-pp.shmem_alloc_and_add_folio
0.17 ± 6% -0.1 0.04 ± 71% perf-profile.self.cycles-pp.shmem_get_folio_gfp
0.18 ± 2% -0.1 0.05 ± 47% perf-profile.self.cycles-pp.free_unref_folios
0.18 ± 7% -0.1 0.06 ± 16% perf-profile.self.cycles-pp.shmem_inode_acct_blocks
0.15 ± 7% -0.1 0.04 ± 71% perf-profile.self.cycles-pp.fdget
0.15 ± 5% -0.1 0.05 ± 47% perf-profile.self.cycles-pp.fallocate64
0.16 ± 3% -0.1 0.06 ± 6% perf-profile.self.cycles-pp.cgroup_rstat_updated
0.08 ± 5% +0.1 0.20 ± 61% perf-profile.self.cycles-pp.page_counter_cancel
0.40 ± 3% +0.5 0.87 ± 53% perf-profile.self.cycles-pp.__lruvec_stat_mod_folio
0.04 ± 44% +0.6 0.65 ± 60% perf-profile.self.cycles-pp.__mem_cgroup_charge
0.00 +0.8 0.84 ± 32% perf-profile.self.cycles-pp.propagate_protected_usage
0.33 ± 37% +0.8 1.18 ± 39% perf-profile.self.cycles-pp.get_mem_cgroup_from_mm
0.03 ±141% +1.0 1.01 ± 44% perf-profile.self.cycles-pp.charge_memcg
0.98 ± 17% +2.9 3.90 ± 15% perf-profile.self.cycles-pp.__mod_memcg_lruvec_state
0.10 ± 8% +3.5 3.60 ± 60% perf-profile.self.cycles-pp.try_charge_memcg
0.00 +7.3 7.30 ± 59% perf-profile.self.cycles-pp.page_counter_try_charge
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] 6+ messages in thread
* Re: [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
2025-03-10 5:50 [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression kernel test robot
@ 2025-03-10 9:55 ` Vlastimil Babka
2025-03-10 10:18 ` Alexei Starovoitov
0 siblings, 1 reply; 6+ messages in thread
From: Vlastimil Babka @ 2025-03-10 9:55 UTC (permalink / raw)
To: kernel test robot, Alexei Starovoitov
Cc: oe-lkp, lkp, Michal Hocko, Shakeel Butt, cgroups, linux-mm
On 3/10/25 06:50, kernel test robot wrote:
>
>
> Hello,
>
> kernel test robot noticed a 37.9% regression of netperf.Throughput_Mbps on:
I assume this is some network receive context where gfpflags do not allow
blocking.
> commit: 01d37228d331047a0bbbd1026cec2ccabef6d88d ("memcg: Use trylock to access memcg stock_lock.")
> https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
>
> [test failed on linux-next/master 7ec162622e66a4ff886f8f28712ea1b13069e1aa]
>
> testcase: netperf
> config: x86_64-rhel-9.4
> compiler: gcc-12
> test machine: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
> parameters:
>
> ip: ipv4
> runtime: 300s
> nr_threads: 50%
> cluster: cs-localhost
> test: TCP_MAERTS
> cpufreq_governor: performance
>
>
> In addition to that, the commit also has significant impact on the following tests:
>
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | stress-ng: stress-ng.mmapfork.ops_per_sec 63.5% regression |
Hm interesting, this one at least from the name would be a GFP_KERNEL context?
> | test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
> | test parameters | cpufreq_governor=performance |
> | | nr_threads=100% |
> | | test=mmapfork |
> | | testtime=60s |
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | hackbench: hackbench.throughput 26.6% regression |
> | test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
> | test parameters | cpufreq_governor=performance |
> | | ipc=socket |
> | | iterations=4 |
> | | mode=threads |
> | | nr_threads=100% |
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | lmbench3: lmbench3.TCP.socket.bandwidth.64B.MB/sec 33.0% regression |
> | test machine | 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 512G memory |
> | test parameters | cpufreq_governor=performance |
> | | mode=development |
> | | nr_threads=100% |
> | | test=TCP |
> | | test_memory_size=50% |
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | vm-scalability: vm-scalability.throughput 86.8% regression |
> | test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
> | test parameters | cpufreq_governor=performance |
> | | runtime=300s |
> | | size=1T |
> | | test=lru-shm |
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | netperf: netperf.Throughput_Mbps 39.9% improvement |
An improvement? Weird.
> | test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
> | test parameters | cluster=cs-localhost |
> | | cpufreq_governor=performance |
> | | ip=ipv4 |
> | | nr_threads=200% |
> | | runtime=300s |
> | | test=TCP_MAERTS |
> +------------------+----------------------------------------------------------------------------------------------------+
> | testcase: change | will-it-scale: will-it-scale.per_thread_ops 68.8% regression |
> | test machine | 104 threads 2 sockets (Skylake) with 192G memory |
> | test parameters | cpufreq_governor=performance |
> | | mode=thread |
> | | nr_task=100% |
> | | test=fallocate1 |
> +------------------+----------------------------------------------------------------------------------------------------+
Some of those as well.
Anyway we should not be expecting the localtry_trylock_irqsave() itself be
failing and resulting in a slow path, as that woulre require an allocation
attempt from a nmi. So what else the commit does?
> 0.10 ± 4% +11.3 11.43 ± 3% perf-profile.self.cycles-pp.try_charge_memcg
> 0.00 +13.7 13.72 ± 2% perf-profile.self.cycles-pp.page_counter_try_charge
This does suggest more time spent in try_charge_memcg() because consume_stock() has failed.
And I suspect this:
+ if (!gfpflags_allow_spinning(gfp_mask))
+ /* Avoid the refill and flush of the older stock */
+ batch = nr_pages;
because this will affect the refill even if consume_stock() fails not due to
a trylock failure (which should not be happening), but also just because the
stock was of a wrong memcg or depleted. So in the nowait context we deny the
refill even if we have the memory. Attached patch could be used to see if it
if fixes things. I'm not sure about the testcases where it doesn't look like
nowait context would be used though, let's see.
I've also found this:
https://lore.kernel.org/all/7s6fbpwsynadnzybhdqg3jwhls4pq2sptyxuyghxpaufhissj5@iadb6ibzscjj/
>
> BTW after the done_restock tag in try_charge_memcg(), we will another
> gfpflags_allow_spinning() check to avoid schedule_work() and
> mem_cgroup_handle_over_high(). Maybe simply return early for
> gfpflags_allow_spinning() without checking high marks.
looks like a small possible optimization that was forgotten?
----8<----
From 29e7d18645577ce13d8a0140c0df050ce1ce0f95 Mon Sep 17 00:00:00 2001
From: Vlastimil Babka <vbabka@suse.cz>
Date: Mon, 10 Mar 2025 10:32:14 +0100
Subject: [PATCH] memcg: Avoid stock refill only if stock_lock can't be
acquired
Since commit 01d37228d331 ("memcg: Use trylock to access memcg
stock_lock.") consume_stock() can fail if it can't obtain
memcg_stock.stock_lock. In that case try_charge_memcg() also avoids
refilling or flushing the stock when gfp flags indicate we are in the
context where obtaining the lock could fail.
However consume_stock() can also fail because the stock was depleted, or
belonged to a different memcg. Avoiding the stock refill then reduces
the caching efficiency, as the refill could still succeed with memory
available. This has caused various regressions to be reported by the
kernel test robot.
To fix this, make the decision to avoid stock refill more precise by
making consume_stock() return -EBUSY when it fails to obtain stock_lock,
and using that for the no-refill decision.
Fixes: 01d37228d331 ("memcg: Use trylock to access memcg stock_lock.")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202503101254.cfd454df-lkp@intel.com
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
---
mm/memcontrol.c | 22 ++++++++++++++--------
1 file changed, 14 insertions(+), 8 deletions(-)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 092cab99dec7..a8371a22c7f4 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -1772,22 +1772,23 @@ static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
* stock, and at least @nr_pages are available in that stock. Failure to
* service an allocation will refill the stock.
*
- * returns true if successful, false otherwise.
+ * returns 0 if successful, -EBUSY if lock cannot be acquired, or -ENOMEM
+ * if the memcg does not match or there are not enough pages
*/
-static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
+static int consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
gfp_t gfp_mask)
{
struct memcg_stock_pcp *stock;
unsigned int stock_pages;
unsigned long flags;
- bool ret = false;
+ bool ret = -ENOMEM;
if (nr_pages > MEMCG_CHARGE_BATCH)
return ret;
if (!localtry_trylock_irqsave(&memcg_stock.stock_lock, flags)) {
if (!gfpflags_allow_spinning(gfp_mask))
- return ret;
+ return -EBUSY;
localtry_lock_irqsave(&memcg_stock.stock_lock, flags);
}
@@ -1795,7 +1796,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
stock_pages = READ_ONCE(stock->nr_pages);
if (memcg == READ_ONCE(stock->cached) && stock_pages >= nr_pages) {
WRITE_ONCE(stock->nr_pages, stock_pages - nr_pages);
- ret = true;
+ ret = 0;
}
localtry_unlock_irqrestore(&memcg_stock.stock_lock, flags);
@@ -2228,13 +2229,18 @@ int try_charge_memcg(struct mem_cgroup *memcg, gfp_t gfp_mask,
bool drained = false;
bool raised_max_event = false;
unsigned long pflags;
+ int consume_ret;
retry:
- if (consume_stock(memcg, nr_pages, gfp_mask))
+ consume_ret = consume_stock(memcg, nr_pages, gfp_mask);
+ if (!consume_ret)
return 0;
- if (!gfpflags_allow_spinning(gfp_mask))
- /* Avoid the refill and flush of the older stock */
+ /*
+ * Avoid the refill and flush of the older stock if we failed to acquire
+ * the stock_lock
+ */
+ if (consume_ret == -EBUSY)
batch = nr_pages;
if (!do_memsw_account() ||
--
2.48.1
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
2025-03-10 9:55 ` Vlastimil Babka
@ 2025-03-10 10:18 ` Alexei Starovoitov
2025-03-10 10:34 ` Vlastimil Babka
0 siblings, 1 reply; 6+ messages in thread
From: Alexei Starovoitov @ 2025-03-10 10:18 UTC (permalink / raw)
To: Vlastimil Babka
Cc: kernel test robot, Alexei Starovoitov, oe-lkp, kbuild test robot,
Michal Hocko, Shakeel Butt, open list:CONTROL GROUP (CGROUP),
linux-mm
On Mon, Mar 10, 2025 at 10:55 AM Vlastimil Babka <vbabka@suse.cz> wrote:
>
> On 3/10/25 06:50, kernel test robot wrote:
> >
> >
> > Hello,
> >
> > kernel test robot noticed a 37.9% regression of netperf.Throughput_Mbps on:
>
> I assume this is some network receive context where gfpflags do not allow
> blocking.
gfpflags_allow_spinning() should be true for all current callers
including networking.
> > commit: 01d37228d331047a0bbbd1026cec2ccabef6d88d ("memcg: Use trylock to access memcg stock_lock.")
> > https://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git master
> >
> > [test failed on linux-next/master 7ec162622e66a4ff886f8f28712ea1b13069e1aa]
> >
> > testcase: netperf
> > config: x86_64-rhel-9.4
> > compiler: gcc-12
> > test machine: 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory
> > parameters:
> >
> > ip: ipv4
> > runtime: 300s
> > nr_threads: 50%
> > cluster: cs-localhost
> > test: TCP_MAERTS
> > cpufreq_governor: performance
> >
> >
> > In addition to that, the commit also has significant impact on the following tests:
> >
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | stress-ng: stress-ng.mmapfork.ops_per_sec 63.5% regression |
>
> Hm interesting, this one at least from the name would be a GFP_KERNEL context?
weird indeed.
> > | test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
> > | test parameters | cpufreq_governor=performance |
> > | | nr_threads=100% |
> > | | test=mmapfork |
> > | | testtime=60s |
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | hackbench: hackbench.throughput 26.6% regression |
> > | test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
> > | test parameters | cpufreq_governor=performance |
> > | | ipc=socket |
> > | | iterations=4 |
> > | | mode=threads |
> > | | nr_threads=100% |
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | lmbench3: lmbench3.TCP.socket.bandwidth.64B.MB/sec 33.0% regression |
> > | test machine | 224 threads 2 sockets Intel(R) Xeon(R) Platinum 8480CTDX (Sapphire Rapids) with 512G memory |
> > | test parameters | cpufreq_governor=performance |
> > | | mode=development |
> > | | nr_threads=100% |
> > | | test=TCP |
> > | | test_memory_size=50% |
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | vm-scalability: vm-scalability.throughput 86.8% regression |
> > | test machine | 256 threads 4 sockets INTEL(R) XEON(R) PLATINUM 8592+ (Emerald Rapids) with 256G memory |
> > | test parameters | cpufreq_governor=performance |
> > | | runtime=300s |
> > | | size=1T |
> > | | test=lru-shm |
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | netperf: netperf.Throughput_Mbps 39.9% improvement |
>
> An improvement? Weird.
Even more weird and makes no sense to me so far.
>
> > | test machine | 128 threads 2 sockets Intel(R) Xeon(R) Gold 6338 CPU @ 2.00GHz (Ice Lake) with 256G memory |
> > | test parameters | cluster=cs-localhost |
> > | | cpufreq_governor=performance |
> > | | ip=ipv4 |
> > | | nr_threads=200% |
> > | | runtime=300s |
> > | | test=TCP_MAERTS |
> > +------------------+----------------------------------------------------------------------------------------------------+
> > | testcase: change | will-it-scale: will-it-scale.per_thread_ops 68.8% regression |
> > | test machine | 104 threads 2 sockets (Skylake) with 192G memory |
> > | test parameters | cpufreq_governor=performance |
> > | | mode=thread |
> > | | nr_task=100% |
> > | | test=fallocate1 |
> > +------------------+----------------------------------------------------------------------------------------------------+
>
> Some of those as well.
>
> Anyway we should not be expecting the localtry_trylock_irqsave() itself be
> failing and resulting in a slow path, as that woulre require an allocation
> attempt from a nmi. So what else the commit does?
>
> > 0.10 ± 4% +11.3 11.43 ± 3% perf-profile.self.cycles-pp.try_charge_memcg
> > 0.00 +13.7 13.72 ± 2% perf-profile.self.cycles-pp.page_counter_try_charge
>
> This does suggest more time spent in try_charge_memcg() because consume_stock() has failed.
>
> And I suspect this:
>
> + if (!gfpflags_allow_spinning(gfp_mask))
> + /* Avoid the refill and flush of the older stock */
> + batch = nr_pages;
>
> because this will affect the refill even if consume_stock() fails not due to
> a trylock failure (which should not be happening), but also just because the
> stock was of a wrong memcg or depleted. So in the nowait context we deny the
> refill even if we have the memory. Attached patch could be used to see if it
> if fixes things. I'm not sure about the testcases where it doesn't look like
> nowait context would be used though, let's see.
Not quite.
GFP_NOWAIT includes __GFP_KSWAPD_RECLAIM,
so gfpflags_allow_spinning() will return true.
So 'batch' won't change.
> I've also found this:
> https://lore.kernel.org/all/7s6fbpwsynadnzybhdqg3jwhls4pq2sptyxuyghxpaufhissj5@iadb6ibzscjj/
Right. And notice Shakeel's suggestion doesn't include '!' in
the condition.
I assumed it's a typo.
Hence added it as "if (!gfpflags_allow_spinning(gfp_mask))"
> >
> > BTW after the done_restock tag in try_charge_memcg(), we will another
> > gfpflags_allow_spinning() check to avoid schedule_work() and
> > mem_cgroup_handle_over_high(). Maybe simply return early for
> > gfpflags_allow_spinning() without checking high marks.
>
> looks like a small possible optimization that was forgotten?
> ----8<----
> From 29e7d18645577ce13d8a0140c0df050ce1ce0f95 Mon Sep 17 00:00:00 2001
> From: Vlastimil Babka <vbabka@suse.cz>
> Date: Mon, 10 Mar 2025 10:32:14 +0100
> Subject: [PATCH] memcg: Avoid stock refill only if stock_lock can't be
> acquired
>
> Since commit 01d37228d331 ("memcg: Use trylock to access memcg
> stock_lock.") consume_stock() can fail if it can't obtain
> memcg_stock.stock_lock. In that case try_charge_memcg() also avoids
> refilling or flushing the stock when gfp flags indicate we are in the
> context where obtaining the lock could fail.
>
> However consume_stock() can also fail because the stock was depleted, or
> belonged to a different memcg. Avoiding the stock refill then reduces
> the caching efficiency, as the refill could still succeed with memory
> available. This has caused various regressions to be reported by the
> kernel test robot.
>
> To fix this, make the decision to avoid stock refill more precise by
> making consume_stock() return -EBUSY when it fails to obtain stock_lock,
> and using that for the no-refill decision.
>
> Fixes: 01d37228d331 ("memcg: Use trylock to access memcg stock_lock.")
> Reported-by: kernel test robot <oliver.sang@intel.com>
> Closes: https://lore.kernel.org/oe-lkp/202503101254.cfd454df-lkp@intel.com
>
> Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
> ---
> mm/memcontrol.c | 22 ++++++++++++++--------
> 1 file changed, 14 insertions(+), 8 deletions(-)
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index 092cab99dec7..a8371a22c7f4 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -1772,22 +1772,23 @@ static bool obj_stock_flush_required(struct memcg_stock_pcp *stock,
> * stock, and at least @nr_pages are available in that stock. Failure to
> * service an allocation will refill the stock.
> *
> - * returns true if successful, false otherwise.
> + * returns 0 if successful, -EBUSY if lock cannot be acquired, or -ENOMEM
> + * if the memcg does not match or there are not enough pages
> */
> -static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
> +static int consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
> gfp_t gfp_mask)
> {
> struct memcg_stock_pcp *stock;
> unsigned int stock_pages;
> unsigned long flags;
> - bool ret = false;
> + bool ret = -ENOMEM;
>
> if (nr_pages > MEMCG_CHARGE_BATCH)
> return ret;
>
> if (!localtry_trylock_irqsave(&memcg_stock.stock_lock, flags)) {
> if (!gfpflags_allow_spinning(gfp_mask))
> - return ret;
> + return -EBUSY;
> localtry_lock_irqsave(&memcg_stock.stock_lock, flags);
> }
>
> @@ -1795,7 +1796,7 @@ static bool consume_stock(struct mem_cgroup *memcg, unsigned int nr_pages,
> stock_pages = READ_ONCE(stock->nr_pages);
> if (memcg == READ_ONCE(stock->cached) && stock_pages >= nr_pages) {
> WRITE_ONCE(stock->nr_pages, stock_pages - nr_pages);
> - ret = true;
> + ret = 0;
> }
>
> localtry_unlock_irqrestore(&memcg_stock.stock_lock, flags);
> @@ -2228,13 +2229,18 @@ int try_charge_memcg(struct mem_cgroup *memcg, gfp_t gfp_mask,
> bool drained = false;
> bool raised_max_event = false;
> unsigned long pflags;
> + int consume_ret;
>
> retry:
> - if (consume_stock(memcg, nr_pages, gfp_mask))
> + consume_ret = consume_stock(memcg, nr_pages, gfp_mask);
> + if (!consume_ret)
> return 0;
>
> - if (!gfpflags_allow_spinning(gfp_mask))
> - /* Avoid the refill and flush of the older stock */
> + /*
> + * Avoid the refill and flush of the older stock if we failed to acquire
> + * the stock_lock
> + */
> + if (consume_ret == -EBUSY)
> batch = nr_pages;
Sure.
I think it's a good optimization, but I don't think it will make
any difference here.
Fixes tag is not appropriate and the commit log is off too.
I have strong suspicion that this bot report is bogus.
I'll try to repro anyway.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
2025-03-10 10:18 ` Alexei Starovoitov
@ 2025-03-10 10:34 ` Vlastimil Babka
2025-03-10 10:56 ` Alexei Starovoitov
0 siblings, 1 reply; 6+ messages in thread
From: Vlastimil Babka @ 2025-03-10 10:34 UTC (permalink / raw)
To: Alexei Starovoitov
Cc: kernel test robot, Alexei Starovoitov, oe-lkp, kbuild test robot,
Michal Hocko, Shakeel Butt, open list:CONTROL GROUP (CGROUP),
linux-mm
On 3/10/25 11:18, Alexei Starovoitov wrote:
>> because this will affect the refill even if consume_stock() fails not due to
>> a trylock failure (which should not be happening), but also just because the
>> stock was of a wrong memcg or depleted. So in the nowait context we deny the
>> refill even if we have the memory. Attached patch could be used to see if it
>> if fixes things. I'm not sure about the testcases where it doesn't look like
>> nowait context would be used though, let's see.
>
> Not quite.
> GFP_NOWAIT includes __GFP_KSWAPD_RECLAIM,
> so gfpflags_allow_spinning() will return true.
Uh right, it's the new gfpflags_allow_spinning(), not the
gfpflags_allow_blocking() I'm used to and implicitly assumed, sorry.
But then it's very simple because it has a bug:
gfpflags_allow_spinning() does
return !(gfp_flags & __GFP_RECLAIM);
should be !!
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
2025-03-10 10:34 ` Vlastimil Babka
@ 2025-03-10 10:56 ` Alexei Starovoitov
2025-03-10 11:03 ` Vlastimil Babka
0 siblings, 1 reply; 6+ messages in thread
From: Alexei Starovoitov @ 2025-03-10 10:56 UTC (permalink / raw)
To: Vlastimil Babka
Cc: kernel test robot, Alexei Starovoitov, oe-lkp, kbuild test robot,
Michal Hocko, Shakeel Butt, open list:CONTROL GROUP (CGROUP),
linux-mm
On Mon, Mar 10, 2025 at 11:34 AM Vlastimil Babka <vbabka@suse.cz> wrote:
>
> On 3/10/25 11:18, Alexei Starovoitov wrote:
> >> because this will affect the refill even if consume_stock() fails not due to
> >> a trylock failure (which should not be happening), but also just because the
> >> stock was of a wrong memcg or depleted. So in the nowait context we deny the
> >> refill even if we have the memory. Attached patch could be used to see if it
> >> if fixes things. I'm not sure about the testcases where it doesn't look like
> >> nowait context would be used though, let's see.
> >
> > Not quite.
> > GFP_NOWAIT includes __GFP_KSWAPD_RECLAIM,
> > so gfpflags_allow_spinning() will return true.
>
> Uh right, it's the new gfpflags_allow_spinning(), not the
> gfpflags_allow_blocking() I'm used to and implicitly assumed, sorry.
>
> But then it's very simple because it has a bug:
> gfpflags_allow_spinning() does
>
> return !(gfp_flags & __GFP_RECLAIM);
>
> should be !!
Ouch.
So I accidentally exposed the whole linux-next to this stress testing
of new trylock facilities :(
But the silver lining is that this is the only thing that blew up :)
Could you send a patch or I will do it later today.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression
2025-03-10 10:56 ` Alexei Starovoitov
@ 2025-03-10 11:03 ` Vlastimil Babka
0 siblings, 0 replies; 6+ messages in thread
From: Vlastimil Babka @ 2025-03-10 11:03 UTC (permalink / raw)
To: Alexei Starovoitov
Cc: kernel test robot, Alexei Starovoitov, oe-lkp, kbuild test robot,
Michal Hocko, Shakeel Butt, open list:CONTROL GROUP (CGROUP),
linux-mm
On 3/10/25 11:56, Alexei Starovoitov wrote:
> On Mon, Mar 10, 2025 at 11:34 AM Vlastimil Babka <vbabka@suse.cz> wrote:
>>
>> On 3/10/25 11:18, Alexei Starovoitov wrote:
>> >> because this will affect the refill even if consume_stock() fails not due to
>> >> a trylock failure (which should not be happening), but also just because the
>> >> stock was of a wrong memcg or depleted. So in the nowait context we deny the
>> >> refill even if we have the memory. Attached patch could be used to see if it
>> >> if fixes things. I'm not sure about the testcases where it doesn't look like
>> >> nowait context would be used though, let's see.
>> >
>> > Not quite.
>> > GFP_NOWAIT includes __GFP_KSWAPD_RECLAIM,
>> > so gfpflags_allow_spinning() will return true.
>>
>> Uh right, it's the new gfpflags_allow_spinning(), not the
>> gfpflags_allow_blocking() I'm used to and implicitly assumed, sorry.
>>
>> But then it's very simple because it has a bug:
>> gfpflags_allow_spinning() does
>>
>> return !(gfp_flags & __GFP_RECLAIM);
>>
>> should be !!
>
> Ouch.
> So I accidentally exposed the whole linux-next to this stress testing
> of new trylock facilities :(
> But the silver lining is that this is the only thing that blew up :)
> Could you send a patch or I will do it later today.
OK
----8<----
From 69b3d1631645c82d9d88f17fb01184d24034df2b Mon Sep 17 00:00:00 2001
From: Vlastimil Babka <vbabka@suse.cz>
Date: Mon, 10 Mar 2025 11:57:52 +0100
Subject: [PATCH] mm: Fix the flipped condition in gfpflags_allow_spinning()
The function gfpflags_allow_spinning() has a bug that makes it return
the opposite result than intended. This could contribute to deadlocks as
usage profilerates, for now it was noticed as a performance regression
due to try_charge_memcg() not refilling memcg stock when it could. Fix
the flipped condition.
Fixes: 97769a53f117 ("mm, bpf: Introduce try_alloc_pages() for opportunistic page allocation")
Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202503101254.cfd454df-lkp@intel.com
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
---
include/linux/gfp.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/gfp.h b/include/linux/gfp.h
index ceb226c2e25c..c9fa6309c903 100644
--- a/include/linux/gfp.h
+++ b/include/linux/gfp.h
@@ -55,7 +55,7 @@ static inline bool gfpflags_allow_spinning(const gfp_t gfp_flags)
* regular page allocator doesn't fully support this
* allocation mode.
*/
- return !(gfp_flags & __GFP_RECLAIM);
+ return !!(gfp_flags & __GFP_RECLAIM);
}
#ifdef CONFIG_HIGHMEM
--
2.48.1
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2025-03-10 11:03 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2025-03-10 5:50 [linux-next:master] [memcg] 01d37228d3: netperf.Throughput_Mbps 37.9% regression kernel test robot
2025-03-10 9:55 ` Vlastimil Babka
2025-03-10 10:18 ` Alexei Starovoitov
2025-03-10 10:34 ` Vlastimil Babka
2025-03-10 10:56 ` Alexei Starovoitov
2025-03-10 11:03 ` Vlastimil Babka
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox