Greeting,
FYI, we noticed a 8.7% improvement of stress-ng.memfd.ops_per_sec due to commit:
commit: 802f1d522d5fdaefc2b935141bc8fe03d43a99ab ("mm: page_counter: re-layout structure to reduce false sharing")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
in testcase: stress-ng
on test machine: 96 threads Intel(R) Xeon(R) Gold 6252 CPU @ 2.10GHz with 192G memory
with following parameters:
nr_threads: 10%
disk: 1HDD
testtime: 60s
fs: ext4
class: os
test: memfd
cpufreq_governor: performance
ucode: 0x5003006
Details are as below:
-------------------------------------------------------------------------------------------------->
To reproduce:
git clone https://github.com/intel/lkp-tests.git
cd lkp-tests
bin/lkp install job.yaml # job file is attached in this email
bin/lkp split-job --compatible job.yaml
bin/lkp run compatible-job.yaml
=========================================================================================
class/compiler/cpufreq_governor/disk/fs/kconfig/nr_threads/rootfs/tbox_group/test/testcase/testtime/ucode:
os/gcc-9/performance/1HDD/ext4/x86_64-rhel-8.3/10%/debian-10.4-x86_64-20200603.cgz/lkp-csl-2sp5/memfd/stress-ng/60s/0x5003006
commit:
c1a660dea3 ("mm: kmem: make __memcg_kmem_(un)charge static")
802f1d522d ("mm: page_counter: re-layout structure to reduce false sharing")
c1a660dea3fa6164 802f1d522d5fdaefc2b935141bc
---------------- ---------------------------
%stddev %change %stddev
\ | \
330.68 +8.7% 359.30 stress-ng.memfd.ops_per_sec
4417 ? 9% -40.7% 2618 ? 11% stress-ng.time.involuntary_context_switches
1.473e+08 +8.7% 1.6e+08 stress-ng.time.minor_page_faults
0.12 ? 5% +0.0 0.14 ? 3% mpstat.cpu.all.soft%
0.03 ? 18% +208.6% 0.10 ? 93% perf-sched.wait_and_delay.avg.ms.pipe_read.new_sync_read.vfs_read.ksys_read
18.75 ? 25% +77.9% 33.37 ? 29% sched_debug.cfs_rq:/.removed.load_avg.avg
265085 ? 4% -24.6% 199773 ? 3% softirqs.RCU
1938 -3.5% 1870 vmstat.system.cs
55992 -1.3% 55237 proc-vmstat.nr_slab_unreclaimable
1.532e+08 +8.7% 1.664e+08 proc-vmstat.numa_hit
1.531e+08 +8.7% 1.663e+08 proc-vmstat.numa_local
1.35e+08 +8.7% 1.467e+08 proc-vmstat.pgactivate
1.544e+08 +8.6% 1.678e+08 proc-vmstat.pgalloc_normal
1.475e+08 +8.6% 1.603e+08 proc-vmstat.pgfault
1.543e+08 +8.7% 1.677e+08 proc-vmstat.pgfree
31687 ? 2% -13.2% 27517 ? 5% slabinfo.lsm_file_cache.active_objs
31687 ? 2% -13.2% 27517 ? 5% slabinfo.lsm_file_cache.num_objs
58130 ? 4% -7.7% 53639 ? 2% slabinfo.radix_tree_node.active_objs
1045 ? 4% -7.6% 965.50 ? 2% slabinfo.radix_tree_node.active_slabs
58556 ? 4% -7.6% 54089 ? 2% slabinfo.radix_tree_node.num_objs
1045 ? 4% -7.6% 965.50 ? 2% slabinfo.radix_tree_node.num_slabs
45005 ? 3% -9.7% 40651 ? 3% slabinfo.shmem_inode_cache.active_objs
45296 ? 3% -9.6% 40945 ? 3% slabinfo.shmem_inode_cache.num_objs
2342 ? 80% -81.4% 435.17 ?137% interrupts.CPU13.NMI:Non-maskable_interrupts
2342 ? 80% -81.4% 435.17 ?137% interrupts.CPU13.PMI:Performance_monitoring_interrupts
354.67 ? 79% -74.2% 91.33 ? 68% interrupts.CPU13.TLB:TLB_shootdowns
483.33 ? 85% -87.7% 59.33 ? 70% interrupts.CPU17.TLB:TLB_shootdowns
550.83 ?145% -88.3% 64.50 ?103% interrupts.CPU26.TLB:TLB_shootdowns
286.33 ? 57% -75.9% 69.00 ?101% interrupts.CPU43.TLB:TLB_shootdowns
357.00 ? 94% -73.9% 93.17 ?105% interrupts.CPU60.TLB:TLB_shootdowns
312.83 ? 95% -76.7% 73.00 ? 65% interrupts.CPU61.TLB:TLB_shootdowns
447.83 ?103% -92.1% 35.33 ?129% interrupts.CPU65.TLB:TLB_shootdowns
528.17 ?132% -93.9% 32.33 ? 91% interrupts.CPU71.TLB:TLB_shootdowns
189.00 ? 60% -82.8% 32.50 ?150% interrupts.CPU75.TLB:TLB_shootdowns
377.33 ?103% -83.4% 62.67 ? 86% interrupts.CPU86.TLB:TLB_shootdowns
213.83 ?107% -86.3% 29.33 ?144% interrupts.CPU87.TLB:TLB_shootdowns
213.00 ? 53% -69.9% 64.17 ?130% interrupts.CPU88.TLB:TLB_shootdowns
30649 ? 59% -69.9% 9231 ? 58% interrupts.TLB:TLB_shootdowns
4.696e+09 +7.8% 5.065e+09 perf-stat.i.branch-instructions
22361825 +5.2% 23516325 ? 2% perf-stat.i.branch-misses
33234349 ? 3% -6.9% 30950008 ? 2% perf-stat.i.cache-misses
1794 -4.0% 1722 perf-stat.i.context-switches
1.16 -6.9% 1.08 ? 3% perf-stat.i.cpi
118.87 -3.8% 114.34 perf-stat.i.cpu-migrations
6.19e+09 +7.5% 6.656e+09 perf-stat.i.dTLB-loads
50621 +12.1% 56758 ? 10% perf-stat.i.dTLB-store-misses
3.265e+09 +8.1% 3.529e+09 perf-stat.i.dTLB-stores
2.453e+10 +7.9% 2.648e+10 perf-stat.i.instructions
3008 +7.4% 3231 ? 4% perf-stat.i.instructions-per-iTLB-miss
0.87 +8.6% 0.94 ? 3% perf-stat.i.ipc
148.89 +7.6% 160.22 perf-stat.i.metric.M/sec
57.89 ? 2% -7.0 50.87 ? 2% perf-stat.i.node-load-miss-rate%
4033634 ? 4% -26.2% 2977641 ? 2% perf-stat.i.node-load-misses
6050869 +6.9% 6466248 perf-stat.i.node-stores
4.86 ? 6% -15.0% 4.13 ? 4% perf-stat.overall.MPKI
0.48 -0.0 0.46 ? 2% perf-stat.overall.branch-miss-rate%
1.14 -8.3% 1.05 ? 3% perf-stat.overall.cpi
841.90 ? 3% +6.2% 894.00 perf-stat.overall.cycles-between-cache-misses
3050 +7.3% 3273 ? 4% perf-stat.overall.instructions-per-iTLB-miss
0.88 +9.1% 0.96 ? 3% perf-stat.overall.ipc
57.39 ? 2% -7.3 50.05 perf-stat.overall.node-load-miss-rate%
36.32 ? 4% -3.4 32.97 perf-stat.overall.node-store-miss-rate%
4.621e+09 +7.8% 4.984e+09 perf-stat.ps.branch-instructions
22001899 +5.2% 23138207 ? 2% perf-stat.ps.branch-misses
32702285 ? 3% -6.9% 30456230 ? 2% perf-stat.ps.cache-misses
1766 -4.1% 1694 perf-stat.ps.context-switches
116.93 -3.8% 112.49 perf-stat.ps.cpu-migrations
6.091e+09 +7.5% 6.55e+09 perf-stat.ps.dTLB-loads
49806 +12.1% 55833 ? 10% perf-stat.ps.dTLB-store-misses
3.213e+09 +8.1% 3.472e+09 perf-stat.ps.dTLB-stores
2.414e+10 +8.0% 2.606e+10 perf-stat.ps.instructions
3969042 ? 4% -26.2% 2930192 ? 2% perf-stat.ps.node-load-misses
5954304 +6.9% 6363077 perf-stat.ps.node-stores
1.522e+12 +7.4% 1.635e+12 perf-stat.total.instructions
38.48 ? 16% -5.4 33.08 ? 5% perf-profile.calltrace.cycles-pp.secondary_startup_64_no_verify
2.08 ? 13% -0.8 1.30 ? 3% perf-profile.calltrace.cycles-pp.workingset_age_nonresident.workingset_activation.mark_page_accessed.zap_pte_range.unmap_page_range
1.26 ? 43% -0.8 0.50 ? 47% perf-profile.calltrace.cycles-pp.lock_page_lruvec_irqsave.release_pages.__pagevec_release.shmem_undo_range.shmem_truncate_range
1.25 ? 42% -0.8 0.49 ? 47% perf-profile.calltrace.cycles-pp._raw_spin_lock_irqsave.lock_page_lruvec_irqsave.release_pages.__pagevec_release.shmem_undo_range
2.31 ? 13% -0.7 1.57 ? 2% perf-profile.calltrace.cycles-pp.workingset_activation.mark_page_accessed.zap_pte_range.unmap_page_range.unmap_vmas
1.40 ? 11% -0.3 1.06 ? 3% perf-profile.calltrace.cycles-pp.release_pages.__pagevec_release.shmem_undo_range.shmem_truncate_range.shmem_fallocate
1.40 ? 11% -0.3 1.07 ? 3% perf-profile.calltrace.cycles-pp.__pagevec_release.shmem_undo_range.shmem_truncate_range.shmem_fallocate.vfs_fallocate
0.94 ? 13% -0.3 0.68 ? 7% perf-profile.calltrace.cycles-pp.try_charge.mem_cgroup_charge.shmem_add_to_page_cache.shmem_getpage_gfp.shmem_fault
0.67 ? 11% +0.1 0.81 ? 4% perf-profile.calltrace.cycles-pp.shmem_seek_hole_data.shmem_file_llseek.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.76 ? 11% +0.2 0.91 ? 2% perf-profile.calltrace.cycles-pp.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.73 ? 9% +0.2 0.89 ? 6% perf-profile.calltrace.cycles-pp.unlock_page.filemap_map_pages.do_fault.__handle_mm_fault.handle_mm_fault
0.93 ? 8% +0.2 1.10 ? 3% perf-profile.calltrace.cycles-pp.rmqueue_bulk.rmqueue.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma
0.87 ? 9% +0.2 1.04 ? 4% perf-profile.calltrace.cycles-pp.shmem_file_llseek.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.94 ? 11% +0.2 1.12 ? 4% perf-profile.calltrace.cycles-pp.follow_page_pte.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff
0.51 ? 45% +0.2 0.70 ? 3% perf-profile.calltrace.cycles-pp.mmap_region.do_mmap.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64
0.99 ? 9% +0.2 1.18 ? 4% perf-profile.calltrace.cycles-pp.ksys_lseek.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.49 ? 45% +0.2 0.71 ? 4% perf-profile.calltrace.cycles-pp.__list_del_entry_valid.rmqueue_bulk.rmqueue.get_page_from_freelist.__alloc_pages_nodemask
1.31 ? 10% +0.2 1.55 ? 4% perf-profile.calltrace.cycles-pp.native_flush_tlb_one_user.flush_tlb_func_common.flush_tlb_mm_range.tlb_flush_mmu.tlb_finish_mmu
1.36 ? 10% +0.2 1.60 ? 4% perf-profile.calltrace.cycles-pp.flush_tlb_func_common.flush_tlb_mm_range.tlb_flush_mmu.tlb_finish_mmu.unmap_region
1.35 ? 9% +0.3 1.61 ? 4% perf-profile.calltrace.cycles-pp.rmqueue.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.shmem_alloc_page
1.41 ? 10% +0.3 1.67 ? 4% perf-profile.calltrace.cycles-pp.flush_tlb_mm_range.tlb_flush_mmu.tlb_finish_mmu.unmap_region.__do_munmap
1.59 ? 9% +0.3 1.91 ? 5% perf-profile.calltrace.cycles-pp.get_page_from_freelist.__alloc_pages_nodemask.alloc_pages_vma.shmem_alloc_page.shmem_alloc_and_acct_page
1.96 ? 9% +0.4 2.35 ? 4% perf-profile.calltrace.cycles-pp.__alloc_pages_nodemask.alloc_pages_vma.shmem_alloc_page.shmem_alloc_and_acct_page.shmem_getpage_gfp
2.37 ? 10% +0.4 2.79 ? 3% perf-profile.calltrace.cycles-pp.next_uptodate_page.filemap_map_pages.do_fault.__handle_mm_fault.handle_mm_fault
2.20 ? 9% +0.4 2.63 ? 4% perf-profile.calltrace.cycles-pp.alloc_pages_vma.shmem_alloc_page.shmem_alloc_and_acct_page.shmem_getpage_gfp.shmem_fault
2.53 ? 8% +0.5 3.02 ? 4% perf-profile.calltrace.cycles-pp.shmem_alloc_page.shmem_alloc_and_acct_page.shmem_getpage_gfp.shmem_fault.__do_fault
2.35 ? 9% +0.5 2.87 ? 4% perf-profile.calltrace.cycles-pp.clear_page_erms.shmem_getpage_gfp.shmem_fault.__do_fault.do_fault
2.99 ? 9% +0.6 3.55 ? 4% perf-profile.calltrace.cycles-pp.shmem_alloc_and_acct_page.shmem_getpage_gfp.shmem_fault.__do_fault.do_fault
5.24 ? 10% +1.0 6.22 ? 3% perf-profile.calltrace.cycles-pp.filemap_map_pages.do_fault.__handle_mm_fault.handle_mm_fault.__get_user_pages
22.72 ? 10% +3.6 26.28 ? 3% perf-profile.calltrace.cycles-pp.shmem_getpage_gfp.shmem_fault.__do_fault.do_fault.__handle_mm_fault
22.90 ? 10% +3.6 26.50 ? 3% perf-profile.calltrace.cycles-pp.shmem_fault.__do_fault.do_fault.__handle_mm_fault.handle_mm_fault
22.97 ? 10% +3.6 26.59 ? 3% perf-profile.calltrace.cycles-pp.__do_fault.do_fault.__handle_mm_fault.handle_mm_fault.__get_user_pages
29.85 ? 10% +4.8 34.61 ? 3% perf-profile.calltrace.cycles-pp.do_fault.__handle_mm_fault.handle_mm_fault.__get_user_pages.populate_vma_page_range
30.40 ? 10% +4.9 35.28 ? 3% perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.__get_user_pages.populate_vma_page_range.__mm_populate
30.84 ? 10% +4.9 35.76 ? 3% perf-profile.calltrace.cycles-pp.handle_mm_fault.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff
32.67 ? 10% +5.3 37.94 ? 3% perf-profile.calltrace.cycles-pp.__get_user_pages.populate_vma_page_range.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff
32.71 ? 10% +5.3 37.98 ? 3% perf-profile.calltrace.cycles-pp.populate_vma_page_range.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64
32.76 ? 10% +5.3 38.04 ? 3% perf-profile.calltrace.cycles-pp.__mm_populate.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
33.62 ? 10% +5.4 39.06 ? 3% perf-profile.calltrace.cycles-pp.vm_mmap_pgoff.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
33.65 ? 10% +5.5 39.10 ? 3% perf-profile.calltrace.cycles-pp.ksys_mmap_pgoff.do_syscall_64.entry_SYSCALL_64_after_hwframe
38.48 ? 16% -5.4 33.08 ? 5% perf-profile.children.cycles-pp.secondary_startup_64_no_verify
38.48 ? 16% -5.4 33.08 ? 5% perf-profile.children.cycles-pp.cpu_startup_entry
38.48 ? 16% -5.4 33.08 ? 5% perf-profile.children.cycles-pp.do_idle
4.18 ? 22% -1.6 2.58 ? 13% perf-profile.children.cycles-pp.native_queued_spin_lock_slowpath
4.16 ? 21% -1.6 2.59 ? 13% perf-profile.children.cycles-pp.lock_page_lruvec_irqsave
4.33 ? 21% -1.6 2.78 ? 12% perf-profile.children.cycles-pp._raw_spin_lock_irqsave
2.23 ? 13% -0.8 1.42 ? 3% perf-profile.children.cycles-pp.workingset_age_nonresident
2.48 ? 13% -0.8 1.71 ? 2% perf-profile.children.cycles-pp.workingset_activation
0.96 ? 13% -0.3 0.69 ? 6% perf-profile.children.cycles-pp.try_charge
0.80 ? 14% -0.2 0.60 ? 7% perf-profile.children.cycles-pp.page_counter_try_charge
0.33 ? 12% -0.2 0.17 ? 23% perf-profile.children.cycles-pp.smpboot_thread_fn
0.32 ? 13% -0.2 0.16 ? 26% perf-profile.children.cycles-pp.run_ksoftirqd
0.37 ? 12% -0.2 0.21 ? 19% perf-profile.children.cycles-pp.kthread
0.37 ? 12% -0.2 0.21 ? 18% perf-profile.children.cycles-pp.ret_from_fork
0.09 ? 11% +0.0 0.11 ? 6% perf-profile.children.cycles-pp.__entry_text_start
0.11 ? 11% +0.0 0.13 ? 6% perf-profile.children.cycles-pp.shmem_mmap
0.11 ? 10% +0.0 0.13 ? 5% perf-profile.children.cycles-pp.vm_unmapped_area
0.08 ? 16% +0.0 0.10 ? 9% perf-profile.children.cycles-pp.fsnotify_destroy_marks
0.11 ? 8% +0.0 0.14 ? 8% perf-profile.children.cycles-pp.xas_find_conflict
0.10 ? 11% +0.0 0.13 ? 6% perf-profile.children.cycles-pp.touch_atime
0.13 ? 11% +0.0 0.16 ? 6% perf-profile.children.cycles-pp.perf_event_mmap
0.08 ? 14% +0.0 0.10 ? 9% perf-profile.children.cycles-pp.fsnotify_grab_connector
0.13 ? 9% +0.0 0.16 ? 4% perf-profile.children.cycles-pp.arch_get_unmapped_area_topdown
0.05 ? 46% +0.0 0.08 ? 11% perf-profile.children.cycles-pp.__srcu_read_lock
0.13 ? 6% +0.0 0.17 ? 4% perf-profile.children.cycles-pp.shmem_get_unmapped_area
0.09 ? 14% +0.0 0.13 ? 5% perf-profile.children.cycles-pp.__destroy_inode
0.04 ? 71% +0.0 0.07 ? 8% perf-profile.children.cycles-pp.__mnt_want_write
0.10 ? 13% +0.0 0.14 ? 9% perf-profile.children.cycles-pp.apparmor_file_free_security
0.09 ? 15% +0.0 0.13 ? 6% perf-profile.children.cycles-pp.destroy_inode
0.10 ? 13% +0.0 0.14 ? 9% perf-profile.children.cycles-pp.security_file_free
0.15 ? 13% +0.0 0.19 ? 7% perf-profile.children.cycles-pp.free_unref_page_commit
0.16 ? 12% +0.0 0.20 ? 6% perf-profile.children.cycles-pp.rcu_read_unlock_strict
0.11 ? 18% +0.0 0.15 ? 12% perf-profile.children.cycles-pp.get_obj_cgroup_from_current
0.22 ? 7% +0.0 0.26 ? 5% perf-profile.children.cycles-pp.xas_start
0.25 ? 9% +0.0 0.30 ? 5% perf-profile.children.cycles-pp.__mod_zone_page_state
0.22 ? 12% +0.1 0.27 ? 6% perf-profile.children.cycles-pp._raw_spin_unlock_irqrestore
0.15 ? 19% +0.1 0.20 ? 4% perf-profile.children.cycles-pp.vm_area_alloc
0.26 ? 9% +0.1 0.31 ? 5% perf-profile.children.cycles-pp.follow_pmd_mask
0.26 ? 11% +0.1 0.32 ? 4% perf-profile.children.cycles-pp.___might_sleep
0.35 ? 10% +0.1 0.41 ? 5% perf-profile.children.cycles-pp.do_sys_ftruncate
0.28 ? 10% +0.1 0.34 ? 6% perf-profile.children.cycles-pp.follow_page_mask
0.29 ? 8% +0.1 0.35 ? 5% perf-profile.children.cycles-pp.scheduler_tick
0.38 ? 7% +0.1 0.44 ? 5% perf-profile.children.cycles-pp.xas_load
0.28 ? 10% +0.1 0.35 ? 8% perf-profile.children.cycles-pp.xas_clear_mark
0.37 ? 12% +0.1 0.44 ? 2% perf-profile.children.cycles-pp._raw_spin_lock_irq
0.34 ? 12% +0.1 0.41 ? 5% perf-profile.children.cycles-pp.page_mapping
0.33 ? 14% +0.1 0.41 ? 5% perf-profile.children.cycles-pp.propagate_protected_usage
0.36 ? 12% +0.1 0.44 ? 8% perf-profile.children.cycles-pp.xas_find
0.49 ? 6% +0.1 0.58 ? 7% perf-profile.children.cycles-pp.update_process_times
0.62 ? 7% +0.1 0.71 ? 5% perf-profile.children.cycles-pp.tick_sched_timer
0.50 ? 7% +0.1 0.59 ? 7% perf-profile.children.cycles-pp.tick_sched_handle
0.40 ? 10% +0.1 0.50 ? 7% perf-profile.children.cycles-pp.xas_init_marks
0.59 ? 12% +0.1 0.71 ? 3% perf-profile.children.cycles-pp.mmap_region
0.84 ? 7% +0.1 0.96 ? 3% perf-profile.children.cycles-pp.__hrtimer_run_queues
0.68 ? 11% +0.1 0.81 ? 4% perf-profile.children.cycles-pp.shmem_seek_hole_data
0.77 ? 11% +0.2 0.92 ? 2% perf-profile.children.cycles-pp.do_mmap
0.84 ? 11% +0.2 1.00 ? 4% perf-profile.children.cycles-pp.find_get_entries
0.75 ? 11% +0.2 0.92 ? 3% perf-profile.children.cycles-pp.__mod_node_page_state
0.87 ? 9% +0.2 1.05 ? 4% perf-profile.children.cycles-pp.shmem_file_llseek
0.72 ? 14% +0.2 0.90 ? 8% perf-profile.children.cycles-pp.rcu_core
0.72 ? 14% +0.2 0.90 ? 8% perf-profile.children.cycles-pp.rcu_do_batch
0.98 ? 11% +0.2 1.16 ? 4% perf-profile.children.cycles-pp.follow_page_pte
1.55 ? 9% +0.2 1.75 ? 2% perf-profile.children.cycles-pp.rmqueue_bulk
0.99 ? 9% +0.2 1.18 ? 4% perf-profile.children.cycles-pp.ksys_lseek
1.29 ? 10% +0.2 1.48 perf-profile.children.cycles-pp.__list_del_entry_valid
1.03 ? 9% +0.2 1.24 ? 5% perf-profile.children.cycles-pp.unlock_page
0.99 ? 11% +0.2 1.20 ? 6% perf-profile.children.cycles-pp.__softirqentry_text_start
0.79 ? 11% +0.2 1.00 ? 8% perf-profile.children.cycles-pp.kmem_cache_free
1.22 ? 11% +0.2 1.45 ? 3% perf-profile.children.cycles-pp.free_unref_page_list
1.04 ? 11% +0.2 1.27 ? 4% perf-profile.children.cycles-pp.__mod_lruvec_state
1.37 ? 10% +0.2 1.61 ? 4% perf-profile.children.cycles-pp.native_flush_tlb_one_user
1.43 ? 10% +0.2 1.68 ? 4% perf-profile.children.cycles-pp.flush_tlb_func_common
2.00 ? 9% +0.3 2.28 ? 3% perf-profile.children.cycles-pp.rmqueue
1.50 ? 10% +0.3 1.78 ? 3% perf-profile.children.cycles-pp.flush_tlb_mm_range
2.28 ? 9% +0.4 2.63 ? 3% perf-profile.children.cycles-pp.get_page_from_freelist
0.67 ? 13% +0.4 1.03 ? 9% perf-profile.children.cycles-pp.do_softirq_own_stack
0.76 ? 11% +0.4 1.14 ? 8% perf-profile.children.cycles-pp.irq_exit_rcu
2.38 ? 9% +0.4 2.80 ? 3% perf-profile.children.cycles-pp.next_uptodate_page
2.71 ? 9% +0.4 3.14 ? 3% perf-profile.children.cycles-pp.__alloc_pages_nodemask
2.23 ? 9% +0.4 2.66 ? 4% perf-profile.children.cycles-pp.alloc_pages_vma
2.54 ? 9% +0.5 3.03 ? 4% perf-profile.children.cycles-pp.shmem_alloc_page
2.40 ? 9% +0.5 2.92 ? 4% perf-profile.children.cycles-pp.clear_page_erms
3.00 ? 8% +0.6 3.57 ? 4% perf-profile.children.cycles-pp.shmem_alloc_and_acct_page
5.31 ? 10% +1.0 6.30 ? 3% perf-profile.children.cycles-pp.filemap_map_pages
22.75 ? 10% +3.6 26.31 ? 3% perf-profile.children.cycles-pp.shmem_getpage_gfp
22.91 ? 10% +3.6 26.51 ? 3% perf-profile.children.cycles-pp.shmem_fault
22.97 ? 10% +3.6 26.60 ? 3% perf-profile.children.cycles-pp.__do_fault
29.89 ? 10% +4.8 34.66 ? 3% perf-profile.children.cycles-pp.do_fault
30.43 ? 10% +4.9 35.32 ? 3% perf-profile.children.cycles-pp.__handle_mm_fault
30.89 ? 10% +4.9 35.81 ? 3% perf-profile.children.cycles-pp.handle_mm_fault
32.71 ? 10% +5.3 37.97 ? 3% perf-profile.children.cycles-pp.__get_user_pages
32.71 ? 10% +5.3 37.98 ? 3% perf-profile.children.cycles-pp.populate_vma_page_range
32.76 ? 10% +5.3 38.04 ? 3% perf-profile.children.cycles-pp.__mm_populate
33.62 ? 10% +5.5 39.07 ? 3% perf-profile.children.cycles-pp.vm_mmap_pgoff
33.65 ? 10% +5.5 39.10 ? 3% perf-profile.children.cycles-pp.ksys_mmap_pgoff
4.17 ? 22% -1.6 2.57 ? 14% perf-profile.self.cycles-pp.native_queued_spin_lock_slowpath
2.21 ? 13% -0.8 1.41 ? 3% perf-profile.self.cycles-pp.workingset_age_nonresident
0.73 ? 14% -0.2 0.51 ? 8% perf-profile.self.cycles-pp.page_counter_try_charge
0.08 ? 8% +0.0 0.10 ? 7% perf-profile.self.cycles-pp.__might_sleep
0.08 ? 9% +0.0 0.09 ? 5% perf-profile.self.cycles-pp.xas_find_conflict
0.09 ? 10% +0.0 0.10 ? 11% perf-profile.self.cycles-pp.percpu_counter_add_batch
0.07 ? 14% +0.0 0.09 ? 5% perf-profile.self.cycles-pp.__do_fault
0.09 ? 11% +0.0 0.11 ? 6% perf-profile.self.cycles-pp.__entry_text_start
0.08 ? 12% +0.0 0.10 ? 10% perf-profile.self.cycles-pp.xas_create_range
0.05 ? 45% +0.0 0.07 ? 8% perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
0.12 ? 10% +0.0 0.15 ? 6% perf-profile.self.cycles-pp.PageHuge
0.18 ? 11% +0.0 0.20 ? 4% perf-profile.self.cycles-pp.lru_cache_add
0.14 ? 10% +0.0 0.17 ? 4% perf-profile.self.cycles-pp.page_cache_free_page
0.07 ? 17% +0.0 0.10 ? 6% perf-profile.self.cycles-pp.__destroy_inode
0.12 ? 10% +0.0 0.15 ? 7% perf-profile.self.cycles-pp.__delete_from_page_cache
0.05 ? 46% +0.0 0.08 ? 13% perf-profile.self.cycles-pp.__srcu_read_lock
0.12 ? 8% +0.0 0.15 ? 7% perf-profile.self.cycles-pp.shmem_file_llseek
0.04 ? 71% +0.0 0.07 ? 5% perf-profile.self.cycles-pp.__mnt_want_write
0.19 ? 8% +0.0 0.23 ? 5% perf-profile.self.cycles-pp.xas_start
0.13 ? 15% +0.0 0.17 ? 7% perf-profile.self.cycles-pp.free_unref_page_list
0.11 ? 18% +0.0 0.15 ? 13% perf-profile.self.cycles-pp.get_obj_cgroup_from_current
0.10 ? 15% +0.0 0.14 ? 9% perf-profile.self.cycles-pp.apparmor_file_free_security
0.15 ? 12% +0.0 0.20 ? 5% perf-profile.self.cycles-pp.shmem_fault
0.15 ? 13% +0.0 0.20 ? 11% perf-profile.self.cycles-pp.kmem_cache_alloc
0.23 ? 11% +0.0 0.28 ? 5% perf-profile.self.cycles-pp.__mod_zone_page_state
0.25 ? 9% +0.0 0.30 ? 6% perf-profile.self.cycles-pp.shmem_pseudo_vma_init
0.22 ? 11% +0.0 0.27 ? 7% perf-profile.self.cycles-pp.get_page_from_freelist
0.25 ? 10% +0.0 0.30 ? 3% perf-profile.self.cycles-pp.___might_sleep
0.25 ? 10% +0.1 0.30 ? 7% perf-profile.self.cycles-pp.follow_page_mask
0.23 ? 11% +0.1 0.29 ? 5% perf-profile.self.cycles-pp.follow_pmd_mask
0.27 ? 10% +0.1 0.34 ? 7% perf-profile.self.cycles-pp.xas_clear_mark
0.36 ? 13% +0.1 0.43 ? 2% perf-profile.self.cycles-pp._raw_spin_lock_irq
0.33 ? 10% +0.1 0.40 ? 4% perf-profile.self.cycles-pp.page_mapping
0.33 ? 14% +0.1 0.40 ? 5% perf-profile.self.cycles-pp.propagate_protected_usage
0.46 ? 11% +0.1 0.54 ? 2% perf-profile.self.cycles-pp.shmem_add_to_page_cache
0.35 ? 11% +0.1 0.43 perf-profile.self.cycles-pp.follow_page_pte
0.33 ? 9% +0.1 0.45 ? 11% perf-profile.self.cycles-pp.kmem_cache_free
0.74 ? 11% +0.1 0.88 ? 3% perf-profile.self.cycles-pp.find_get_entries
0.84 ? 11% +0.1 0.97 ? 3% perf-profile.self.cycles-pp._raw_spin_lock
0.85 ? 9% +0.2 1.00 ? 4% perf-profile.self.cycles-pp.mark_page_accessed
0.72 ? 11% +0.2 0.89 ? 3% perf-profile.self.cycles-pp.__mod_node_page_state
1.27 ? 10% +0.2 1.45 ? 2% perf-profile.self.cycles-pp.__list_del_entry_valid
0.97 ? 9% +0.2 1.16 ? 4% perf-profile.self.cycles-pp.unlock_page
1.36 ? 10% +0.2 1.60 ? 4% perf-profile.self.cycles-pp.native_flush_tlb_one_user
1.71 ? 12% +0.3 2.01 ? 3% perf-profile.self.cycles-pp.filemap_map_pages
2.35 ? 9% +0.4 2.75 ? 3% perf-profile.self.cycles-pp.next_uptodate_page
2.37 ? 9% +0.5 2.88 ? 4% perf-profile.self.cycles-pp.clear_page_erms
10.31 ? 10% +2.4 12.66 ? 3% perf-profile.self.cycles-pp.shmem_getpage_gfp
stress-ng.memfd.ops_per_sec
365 +---------------------------------------------------------------------+
360 |-+ O O O OO O O O |
| OO O OO O O O O OO OO O O O O O OO |
355 |-+ O O O O O |
350 |-+ O |
| |
345 |-+ |
340 |-+ |
335 |-+ + + +.+ |
| ++. +.+ +. :+ :: + : ++.+ +. +.|
330 |-+ + + + + + +. :: : : : + : + |
325 |.++.++.+ + + ++.+ .+.++.+ .+.++.: :: :: + |
| + + + + + + :: |
320 |-+ + |
315 +---------------------------------------------------------------------+
stress-ng.time.minor_page_faults
1.65e+08 +----------------------------------------------------------------+
| |
| O O OO O O O |
1.6e+08 |O+O O OO O O OO O O OO O O O O O OO |
| O O O O O O |
| O |
1.55e+08 |-+ |
| |
1.5e+08 |-+ .+ |
| .+ .+ .+ + ++ : + .+ + |
| .++.++ : + + + :: : : :+ :+ +.+|
1.45e+08 |-.++.++ +. : + .+++. .++.+ : :: :: + |
|+ + ++.++ ++ + + :: |
| + |
1.4e+08 +----------------------------------------------------------------+
[*] bisect-good sample
[O] bisect-bad sample
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.
---
0DAY/LKP+ Test Infrastructure Open Source Technology Center
https://lists.01.org/hyperkitty/list/[email protected] Intel Corporation
Thanks,
Oliver Sang