From: kernel test robot <oliver.sang@intel.com>
To: Mateusz Guzik <mjguzik@gmail.com>
Cc: <oe-lkp@lists.linux.dev>, <lkp@intel.com>, <oliver.sang@intel.com>
Subject: [mjguzik:lkp-bench] [x86] 76a45d86f7: will-it-scale.per_process_ops 8.3% improvement
Date: Fri, 18 Apr 2025 10:46:37 +0800 [thread overview]
Message-ID: <202504181042.54ea2b8a-lkp@intel.com> (raw)
Hello,
kernel test robot noticed a 8.3% improvement of will-it-scale.per_process_ops on:
commit: 76a45d86f78bc61a27e382c287a3ed4f046755bb ("x86: prevent gcc from emitting rep movsq/stosq for inlined ops")
https://github.com/mjguzik/linux lkp-bench
testcase: will-it-scale
config: x86_64-rhel-9.4
compiler: gcc-12
test machine: 224 threads 4 sockets Intel(R) Xeon(R) Platinum 8380H CPU @ 2.90GHz (Cooper Lake) with 192G memory
parameters:
nr_task: 100%
mode: process
test: page_fault3
cpufreq_governor: performance
In addition to that, the commit also has significant impact on the following tests:
+------------------+---------------------------------------------------------------------------------------------+
| testcase: change | lmbench3: lmbench3.TCP.socket.bandwidth.64B.MB/sec 3.7% improvement |
| 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=50% |
| | test=TCP |
| | test_memory_size=50% |
+------------------+---------------------------------------------------------------------------------------------+
| testcase: change | lmbench3: lmbench3.syscall.open/close.latency.us 1.5% 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=1 |
| | test=SYSCALL |
| | test_memory_size=50% |
+------------------+---------------------------------------------------------------------------------------------+
| testcase: change | stress-ng: stress-ng.icache.ops_per_sec 3.4% improvement |
| test machine | 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory |
| test parameters | cpufreq_governor=performance |
| | nr_threads=100% |
| | test=icache |
| | testtime=60s |
+------------------+---------------------------------------------------------------------------------------------+
Details are as below:
-------------------------------------------------------------------------------------------------->
The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20250418/202504181042.54ea2b8a-lkp@intel.com
=========================================================================================
compiler/cpufreq_governor/kconfig/mode/nr_task/rootfs/tbox_group/test/testcase:
gcc-12/performance/x86_64-rhel-9.4/process/100%/debian-12-x86_64-20240206.cgz/lkp-cpl-4sp2/page_fault3/will-it-scale
commit:
3b07108ada ("Merge tag 'linux_kselftest-fixes-6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest")
76a45d86f7 ("x86: prevent gcc from emitting rep movsq/stosq for inlined ops")
3b07108ada81a8eb 76a45d86f78bc61a27e382c287a
---------------- ---------------------------
%stddev %change %stddev
\ | \
1.831e+08 +8.3% 1.982e+08 will-it-scale.224.processes
817312 +8.3% 884836 will-it-scale.per_process_ops
1.831e+08 +8.3% 1.982e+08 will-it-scale.workload
1.199e+08 +7.6% 1.291e+08 proc-vmstat.numa_hit
1.196e+08 +7.7% 1.287e+08 proc-vmstat.numa_local
1.201e+08 +7.6% 1.293e+08 proc-vmstat.pgalloc_normal
5.507e+10 +8.3% 5.962e+10 proc-vmstat.pgfault
1.184e+08 ± 2% +8.8% 1.288e+08 proc-vmstat.pgfree
0.99 -3.0% 0.96 perf-stat.i.MPKI
1.093e+11 +9.9% 1.2e+11 perf-stat.i.branch-instructions
0.11 +0.0 0.11 perf-stat.i.branch-miss-rate%
1.011e+08 +17.4% 1.187e+08 perf-stat.i.branch-misses
5.488e+08 +8.0% 5.929e+08 perf-stat.i.cache-misses
6.914e+08 +8.0% 7.466e+08 perf-stat.i.cache-references
1.38 -10.5% 1.23 perf-stat.i.cpi
1388 -7.6% 1282 perf-stat.i.cycles-between-cache-misses
5.54e+11 +11.5% 6.176e+11 perf-stat.i.instructions
0.73 +11.7% 0.81 perf-stat.i.ipc
1626 +8.3% 1760 perf-stat.i.metric.K/sec
1.822e+08 +8.3% 1.972e+08 perf-stat.i.minor-faults
1.822e+08 +8.3% 1.972e+08 perf-stat.i.page-faults
0.99 -3.1% 0.96 perf-stat.overall.MPKI
0.09 +0.0 0.10 perf-stat.overall.branch-miss-rate%
1.38 -10.5% 1.23 perf-stat.overall.cpi
1390 -7.6% 1284 perf-stat.overall.cycles-between-cache-misses
0.73 +11.7% 0.81 perf-stat.overall.ipc
913121 +3.0% 940467 perf-stat.overall.path-length
1.089e+11 +9.9% 1.196e+11 perf-stat.ps.branch-instructions
1.008e+08 +17.4% 1.183e+08 perf-stat.ps.branch-misses
5.47e+08 +8.0% 5.909e+08 perf-stat.ps.cache-misses
6.891e+08 +8.0% 7.441e+08 perf-stat.ps.cache-references
5.521e+11 +11.5% 6.155e+11 perf-stat.ps.instructions
1.816e+08 +8.3% 1.966e+08 perf-stat.ps.minor-faults
1.816e+08 +8.3% 1.966e+08 perf-stat.ps.page-faults
1.672e+14 +11.5% 1.864e+14 perf-stat.total.instructions
6.56 -5.4 1.16 perf-profile.calltrace.cycles-pp.sync_regs.asm_exc_page_fault.testcase
75.84 -2.3 73.50 perf-profile.calltrace.cycles-pp.testcase
5.01 -0.5 4.56 perf-profile.calltrace.cycles-pp.lock_vma_under_rcu.do_user_addr_fault.exc_page_fault.asm_exc_page_fault.testcase
0.57 +0.0 0.59 perf-profile.calltrace.cycles-pp.___pte_offset_map.pte_offset_map_rw_nolock.handle_pte_fault.__handle_mm_fault.handle_mm_fault
0.58 +0.0 0.60 perf-profile.calltrace.cycles-pp.access_error.do_user_addr_fault.exc_page_fault.asm_exc_page_fault.testcase
0.92 +0.0 0.96 perf-profile.calltrace.cycles-pp.pte_offset_map_rw_nolock.handle_pte_fault.__handle_mm_fault.handle_mm_fault.do_user_addr_fault
2.21 +0.0 2.25 perf-profile.calltrace.cycles-pp.__perf_sw_event.do_user_addr_fault.exc_page_fault.asm_exc_page_fault.testcase
0.66 +0.1 0.71 perf-profile.calltrace.cycles-pp.folio_unlock.fault_dirty_shared_page.do_shared_fault.do_pte_missing.__handle_mm_fault
1.46 +0.1 1.52 perf-profile.calltrace.cycles-pp.___perf_sw_event.__perf_sw_event.do_user_addr_fault.exc_page_fault.asm_exc_page_fault
0.71 +0.1 0.77 perf-profile.calltrace.cycles-pp._raw_spin_lock.__pte_offset_map_lock.finish_fault.do_shared_fault.do_pte_missing
1.60 +0.1 1.66 perf-profile.calltrace.cycles-pp.handle_pte_fault.__handle_mm_fault.handle_mm_fault.do_user_addr_fault.exc_page_fault
0.63 +0.1 0.70 perf-profile.calltrace.cycles-pp.folios_put_refs.free_pages_and_swap_cache.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range
0.84 +0.1 0.92 perf-profile.calltrace.cycles-pp.irqentry_exit_to_user_mode.asm_exc_page_fault.testcase
1.46 +0.1 1.56 perf-profile.calltrace.cycles-pp.inode_needs_update_time.file_update_time.fault_dirty_shared_page.do_shared_fault.do_pte_missing
1.71 +0.1 1.81 perf-profile.calltrace.cycles-pp.__pte_offset_map_lock.finish_fault.do_shared_fault.do_pte_missing.__handle_mm_fault
0.93 +0.1 1.03 perf-profile.calltrace.cycles-pp.__mod_memcg_lruvec_state.__lruvec_stat_mod_folio.folio_remove_rmap_ptes.tlb_flush_rmap_batch.tlb_flush_rmaps
0.95 +0.1 1.05 perf-profile.calltrace.cycles-pp.error_entry.testcase
1.66 +0.1 1.77 perf-profile.calltrace.cycles-pp.file_update_time.fault_dirty_shared_page.do_shared_fault.do_pte_missing.__handle_mm_fault
0.97 +0.1 1.07 perf-profile.calltrace.cycles-pp.free_pages_and_swap_cache.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range.zap_pmd_range
0.99 +0.1 1.10 perf-profile.calltrace.cycles-pp.__tlb_batch_free_encoded_pages.tlb_flush_mmu.zap_pte_range.zap_pmd_range.unmap_page_range
0.99 +0.1 1.10 perf-profile.calltrace.cycles-pp.tlb_flush_mmu.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas
1.42 +0.1 1.54 perf-profile.calltrace.cycles-pp.___perf_sw_event.__perf_sw_event.handle_mm_fault.do_user_addr_fault.exc_page_fault
1.51 +0.2 1.67 perf-profile.calltrace.cycles-pp.__lruvec_stat_mod_folio.folio_remove_rmap_ptes.tlb_flush_rmap_batch.tlb_flush_rmaps.zap_pte_range
2.08 +0.2 2.27 perf-profile.calltrace.cycles-pp.__perf_sw_event.handle_mm_fault.do_user_addr_fault.exc_page_fault.asm_exc_page_fault
2.71 +0.2 2.90 perf-profile.calltrace.cycles-pp.mas_walk.lock_vma_under_rcu.do_user_addr_fault.exc_page_fault.asm_exc_page_fault
2.46 +0.2 2.66 perf-profile.calltrace.cycles-pp.tlb_flush_rmap_batch.tlb_flush_rmaps.zap_pte_range.zap_pmd_range.unmap_page_range
2.28 +0.2 2.49 perf-profile.calltrace.cycles-pp.folio_remove_rmap_ptes.tlb_flush_rmap_batch.tlb_flush_rmaps.zap_pte_range.zap_pmd_range
2.48 +0.2 2.68 perf-profile.calltrace.cycles-pp.tlb_flush_rmaps.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas
2.73 +0.2 2.94 perf-profile.calltrace.cycles-pp.xas_load.filemap_get_entry.shmem_get_folio_gfp.shmem_fault.__do_fault
3.40 +0.3 3.65 perf-profile.calltrace.cycles-pp.fault_dirty_shared_page.do_shared_fault.do_pte_missing.__handle_mm_fault.handle_mm_fault
4.28 ± 3% +0.3 4.56 ± 3% perf-profile.calltrace.cycles-pp.set_pte_range.finish_fault.do_shared_fault.do_pte_missing.__handle_mm_fault
2.99 +0.3 3.34 perf-profile.calltrace.cycles-pp.zap_present_ptes.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas
8.47 +0.6 9.04 perf-profile.calltrace.cycles-pp.finish_fault.do_shared_fault.do_pte_missing.__handle_mm_fault.handle_mm_fault
6.49 +0.6 7.11 perf-profile.calltrace.cycles-pp.filemap_get_entry.shmem_get_folio_gfp.shmem_fault.__do_fault.do_shared_fault
106.00 +0.6 106.64 perf-profile.calltrace.cycles-pp.asm_exc_page_fault.testcase
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_after_hwframe.__munmap
7.05 +0.7 7.76 perf-profile.calltrace.cycles-pp.zap_pte_range.zap_pmd_range.unmap_page_range.unmap_vmas.vms_clear_ptes
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.__munmap
7.10 +0.7 7.82 perf-profile.calltrace.cycles-pp.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap.__x64_sys_munmap
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.do_vmi_munmap.__vm_munmap.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe
7.09 +0.7 7.81 perf-profile.calltrace.cycles-pp.unmap_page_range.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap
7.09 +0.7 7.81 perf-profile.calltrace.cycles-pp.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap
7.09 +0.7 7.81 perf-profile.calltrace.cycles-pp.zap_pmd_range.unmap_page_range.unmap_vmas.vms_clear_ptes.vms_complete_munmap_vmas
7.11 +0.7 7.83 perf-profile.calltrace.cycles-pp.__x64_sys_munmap.do_syscall_64.entry_SYSCALL_64_after_hwframe.__munmap
7.10 +0.7 7.82 perf-profile.calltrace.cycles-pp.vms_clear_ptes.vms_complete_munmap_vmas.do_vmi_align_munmap.do_vmi_munmap.__vm_munmap
8.32 +0.8 9.08 perf-profile.calltrace.cycles-pp.shmem_get_folio_gfp.shmem_fault.__do_fault.do_shared_fault.do_pte_missing
9.07 +0.8 9.89 perf-profile.calltrace.cycles-pp.shmem_fault.__do_fault.do_shared_fault.do_pte_missing.__handle_mm_fault
9.57 +0.9 10.42 perf-profile.calltrace.cycles-pp.__do_fault.do_shared_fault.do_pte_missing.__handle_mm_fault.handle_mm_fault
44.51 +1.3 45.83 perf-profile.calltrace.cycles-pp.exc_page_fault.asm_exc_page_fault.testcase
27.32 +1.4 28.76 perf-profile.calltrace.cycles-pp.__handle_mm_fault.handle_mm_fault.do_user_addr_fault.exc_page_fault.asm_exc_page_fault
43.61 +1.5 45.07 perf-profile.calltrace.cycles-pp.do_user_addr_fault.exc_page_fault.asm_exc_page_fault.testcase
22.07 +1.7 23.78 perf-profile.calltrace.cycles-pp.do_shared_fault.do_pte_missing.__handle_mm_fault.handle_mm_fault.do_user_addr_fault
22.67 +1.8 24.42 perf-profile.calltrace.cycles-pp.do_pte_missing.__handle_mm_fault.handle_mm_fault.do_user_addr_fault.exc_page_fault
33.46 +1.9 35.35 perf-profile.calltrace.cycles-pp.handle_mm_fault.do_user_addr_fault.exc_page_fault.asm_exc_page_fault.testcase
6.75 -5.6 1.16 perf-profile.children.cycles-pp.sync_regs
79.90 -1.8 78.06 perf-profile.children.cycles-pp.asm_exc_page_fault
92.80 -0.7 92.08 perf-profile.children.cycles-pp.testcase
5.06 -0.4 4.61 perf-profile.children.cycles-pp.lock_vma_under_rcu
0.18 ± 2% -0.0 0.17 perf-profile.children.cycles-pp.irqentry_enter
0.05 +0.0 0.06 perf-profile.children.cycles-pp.restore_regs_and_return_to_kernel
0.22 +0.0 0.24 perf-profile.children.cycles-pp.__cond_resched
0.21 ± 2% +0.0 0.23 perf-profile.children.cycles-pp.alloc_pages_mpol
0.11 ± 4% +0.0 0.13 perf-profile.children.cycles-pp.free_swap_cache
0.10 ± 5% +0.0 0.11 ± 3% perf-profile.children.cycles-pp.queued_spin_unlock
0.22 ± 2% +0.0 0.24 perf-profile.children.cycles-pp.pte_alloc_one
0.21 ± 2% +0.0 0.23 perf-profile.children.cycles-pp.__alloc_frozen_pages_noprof
0.21 ± 2% +0.0 0.23 perf-profile.children.cycles-pp.alloc_pages_noprof
0.31 ± 2% +0.0 0.33 ± 2% perf-profile.children.cycles-pp.perf_exclude_event
0.18 +0.0 0.20 perf-profile.children.cycles-pp.__tlb_remove_folio_pages_size
0.43 +0.0 0.46 perf-profile.children.cycles-pp.ktime_get_coarse_real_ts64_mg
0.26 ± 3% +0.0 0.29 perf-profile.children.cycles-pp.folio_mapping
0.20 +0.0 0.23 ± 2% perf-profile.children.cycles-pp.vm_normal_page
0.20 +0.0 0.23 perf-profile.children.cycles-pp.folio_mark_accessed
1.03 +0.0 1.06 perf-profile.children.cycles-pp.___pte_offset_map
0.64 +0.0 0.67 perf-profile.children.cycles-pp.access_error
0.32 +0.0 0.35 perf-profile.children.cycles-pp.__irqentry_text_end
0.98 +0.0 1.02 perf-profile.children.cycles-pp.pte_offset_map_rw_nolock
0.66 +0.1 0.71 perf-profile.children.cycles-pp.folio_unlock
0.29 ± 2% +0.1 0.34 perf-profile.children.cycles-pp.noop_dirty_folio
0.76 +0.1 0.83 perf-profile.children.cycles-pp._raw_spin_lock
1.62 +0.1 1.68 perf-profile.children.cycles-pp.handle_pte_fault
0.63 +0.1 0.70 perf-profile.children.cycles-pp.folios_put_refs
0.69 ± 2% +0.1 0.77 perf-profile.children.cycles-pp.folio_mark_dirty
0.70 +0.1 0.78 perf-profile.children.cycles-pp.__mod_node_page_state
0.85 +0.1 0.93 perf-profile.children.cycles-pp.irqentry_exit_to_user_mode
1.52 +0.1 1.61 perf-profile.children.cycles-pp.inode_needs_update_time
1.78 +0.1 1.88 perf-profile.children.cycles-pp.__pte_offset_map_lock
1.70 +0.1 1.81 perf-profile.children.cycles-pp.file_update_time
0.99 +0.1 1.10 perf-profile.children.cycles-pp.free_pages_and_swap_cache
0.99 +0.1 1.10 perf-profile.children.cycles-pp.__tlb_batch_free_encoded_pages
0.99 +0.1 1.10 perf-profile.children.cycles-pp.tlb_flush_mmu
1.12 +0.1 1.26 perf-profile.children.cycles-pp.error_entry
2.74 +0.2 2.94 perf-profile.children.cycles-pp.mas_walk
3.03 +0.2 3.23 perf-profile.children.cycles-pp.___perf_sw_event
2.48 +0.2 2.68 perf-profile.children.cycles-pp.tlb_flush_rmaps
2.32 +0.2 2.52 perf-profile.children.cycles-pp.folio_remove_rmap_ptes
2.48 +0.2 2.68 perf-profile.children.cycles-pp.tlb_flush_rmap_batch
2.76 +0.2 2.98 perf-profile.children.cycles-pp.xas_load
4.33 +0.2 4.56 perf-profile.children.cycles-pp.__perf_sw_event
3.47 +0.3 3.74 perf-profile.children.cycles-pp.fault_dirty_shared_page
4.35 ± 3% +0.3 4.63 ± 2% perf-profile.children.cycles-pp.set_pte_range
3.07 +0.4 3.42 perf-profile.children.cycles-pp.zap_present_ptes
8.59 +0.6 9.16 perf-profile.children.cycles-pp.finish_fault
6.52 +0.6 7.15 perf-profile.children.cycles-pp.filemap_get_entry
7.12 +0.7 7.83 perf-profile.children.cycles-pp.__munmap
7.09 +0.7 7.81 perf-profile.children.cycles-pp.unmap_vmas
7.10 +0.7 7.82 perf-profile.children.cycles-pp.vms_complete_munmap_vmas
7.09 +0.7 7.81 perf-profile.children.cycles-pp.zap_pte_range
7.11 +0.7 7.83 perf-profile.children.cycles-pp.do_vmi_align_munmap
7.11 +0.7 7.83 perf-profile.children.cycles-pp.do_vmi_munmap
7.09 +0.7 7.81 perf-profile.children.cycles-pp.unmap_page_range
7.09 +0.7 7.81 perf-profile.children.cycles-pp.zap_pmd_range
7.17 +0.7 7.89 perf-profile.children.cycles-pp.do_syscall_64
7.11 +0.7 7.83 perf-profile.children.cycles-pp.__vm_munmap
7.11 +0.7 7.83 perf-profile.children.cycles-pp.__x64_sys_munmap
7.10 +0.7 7.82 perf-profile.children.cycles-pp.vms_clear_ptes
7.17 +0.7 7.89 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
8.35 +0.8 9.11 perf-profile.children.cycles-pp.shmem_get_folio_gfp
9.37 +0.8 10.18 perf-profile.children.cycles-pp.native_irq_return_iret
9.10 +0.8 9.92 perf-profile.children.cycles-pp.shmem_fault
9.60 +0.9 10.45 perf-profile.children.cycles-pp.__do_fault
44.64 +1.3 45.90 perf-profile.children.cycles-pp.exc_page_fault
27.47 +1.4 28.91 perf-profile.children.cycles-pp.__handle_mm_fault
43.78 +1.4 45.23 perf-profile.children.cycles-pp.do_user_addr_fault
22.14 +1.7 23.85 perf-profile.children.cycles-pp.do_shared_fault
22.69 +1.8 24.44 perf-profile.children.cycles-pp.do_pte_missing
33.56 +1.9 35.46 perf-profile.children.cycles-pp.handle_mm_fault
6.72 -5.6 1.11 perf-profile.self.cycles-pp.sync_regs
2.27 -0.6 1.64 perf-profile.self.cycles-pp.lock_vma_under_rcu
3.17 -0.4 2.78 perf-profile.self.cycles-pp.__handle_mm_fault
0.72 -0.2 0.54 perf-profile.self.cycles-pp.exc_page_fault
1.32 -0.1 1.23 perf-profile.self.cycles-pp.asm_exc_page_fault
0.09 +0.0 0.10 perf-profile.self.cycles-pp.clear_page_erms
0.06 +0.0 0.07 perf-profile.self.cycles-pp.error_return
0.07 ± 6% +0.0 0.08 ± 5% perf-profile.self.cycles-pp.queued_spin_unlock
0.10 ± 4% +0.0 0.11 perf-profile.self.cycles-pp.free_swap_cache
0.23 ± 3% +0.0 0.25 perf-profile.self.cycles-pp.folio_mapping
0.16 ± 2% +0.0 0.18 perf-profile.self.cycles-pp.__tlb_remove_folio_pages_size
0.44 +0.0 0.46 perf-profile.self.cycles-pp.pte_offset_map_rw_nolock
0.18 +0.0 0.20 perf-profile.self.cycles-pp.folio_mark_accessed
0.18 ± 2% +0.0 0.20 ± 2% perf-profile.self.cycles-pp.vm_normal_page
0.40 +0.0 0.42 perf-profile.self.cycles-pp.ktime_get_coarse_real_ts64_mg
0.26 +0.0 0.28 perf-profile.self.cycles-pp.free_pages_and_swap_cache
0.96 +0.0 0.98 perf-profile.self.cycles-pp.___pte_offset_map
0.56 +0.0 0.59 perf-profile.self.cycles-pp.handle_pte_fault
0.62 +0.0 0.65 perf-profile.self.cycles-pp.access_error
0.30 +0.0 0.34 perf-profile.self.cycles-pp.__irqentry_text_end
0.59 +0.0 0.63 perf-profile.self.cycles-pp.__pte_offset_map_lock
1.25 +0.0 1.29 perf-profile.self.cycles-pp.__perf_sw_event
0.76 +0.0 0.81 perf-profile.self.cycles-pp.folio_remove_rmap_ptes
0.54 ± 2% +0.0 0.59 ± 2% perf-profile.self.cycles-pp.do_pte_missing
0.43 ± 2% +0.0 0.48 perf-profile.self.cycles-pp.folio_mark_dirty
0.62 +0.1 0.67 perf-profile.self.cycles-pp.folio_unlock
0.25 ± 2% +0.1 0.30 perf-profile.self.cycles-pp.noop_dirty_folio
0.53 +0.1 0.58 perf-profile.self.cycles-pp.zap_pte_range
0.73 ± 2% +0.1 0.79 perf-profile.self.cycles-pp.shmem_fault
0.72 +0.1 0.79 perf-profile.self.cycles-pp._raw_spin_lock
0.83 ± 2% +0.1 0.89 perf-profile.self.cycles-pp.folio_add_file_rmap_ptes
0.62 +0.1 0.69 perf-profile.self.cycles-pp.folios_put_refs
0.67 +0.1 0.74 perf-profile.self.cycles-pp.__mod_node_page_state
0.79 +0.1 0.87 perf-profile.self.cycles-pp.irqentry_exit_to_user_mode
1.62 +0.1 1.73 perf-profile.self.cycles-pp.shmem_get_folio_gfp
1.76 +0.1 1.89 perf-profile.self.cycles-pp.finish_fault
1.04 +0.1 1.17 perf-profile.self.cycles-pp.error_entry
2.74 +0.2 2.91 perf-profile.self.cycles-pp.handle_mm_fault
2.27 +0.2 2.45 perf-profile.self.cycles-pp.___perf_sw_event
2.71 +0.2 2.89 perf-profile.self.cycles-pp.mas_walk
2.26 +0.2 2.45 perf-profile.self.cycles-pp.xas_load
2.15 +0.2 2.38 perf-profile.self.cycles-pp.zap_present_ptes
3.74 +0.4 4.16 perf-profile.self.cycles-pp.filemap_get_entry
9.36 +0.8 10.16 perf-profile.self.cycles-pp.native_irq_return_iret
28.42 +2.6 31.02 perf-profile.self.cycles-pp.testcase
***************************************************************************************************
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/50%/debian-12-x86_64-20240206.cgz/lkp-spr-2sp4/TCP/50%/lmbench3
commit:
3b07108ada ("Merge tag 'linux_kselftest-fixes-6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest")
76a45d86f7 ("x86: prevent gcc from emitting rep movsq/stosq for inlined ops")
3b07108ada81a8eb 76a45d86f78bc61a27e382c287a
---------------- ---------------------------
%stddev %change %stddev
\ | \
16039 +3.7% 16639 lmbench3.TCP.socket.bandwidth.64B.MB/sec
4.85 ± 2% -1.7 3.18 ± 2% perf-profile.calltrace.cycles-pp._raw_spin_lock_bh.release_sock.tcp_sendmsg.sock_write_iter.vfs_write
2.65 -0.5 2.16 perf-profile.calltrace.cycles-pp.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.tcp_recvmsg_locked.tcp_recvmsg
2.22 -0.5 1.76 perf-profile.calltrace.cycles-pp.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter.tcp_recvmsg_locked
1.54 -0.4 1.16 perf-profile.calltrace.cycles-pp.check_heap_object.__check_object_size.simple_copy_to_iter.__skb_datagram_iter.skb_copy_datagram_iter
2.76 -0.1 2.62 ± 2% perf-profile.calltrace.cycles-pp.tcp_send_mss.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter.vfs_write
2.04 ± 2% -0.1 1.92 ± 2% perf-profile.calltrace.cycles-pp.tcp_current_mss.tcp_send_mss.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter
0.61 -0.0 0.58 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.tcp_recvmsg.inet_recvmsg.sock_recvmsg.sock_read_iter
0.66 -0.0 0.64 perf-profile.calltrace.cycles-pp.__local_bh_enable_ip.tcp_sendmsg.sock_write_iter.vfs_write.ksys_write
0.73 +0.0 0.74 perf-profile.calltrace.cycles-pp.rw_verify_area.vfs_write.ksys_write.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.71 +0.0 0.74 perf-profile.calltrace.cycles-pp.rw_verify_area.vfs_read.ksys_read.do_syscall_64.entry_SYSCALL_64_after_hwframe
0.59 +0.0 0.61 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_safe_stack.write
0.56 +0.0 0.59 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64_safe_stack.read
1.83 +0.0 1.87 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.read
0.84 +0.0 0.88 perf-profile.calltrace.cycles-pp._raw_spin_lock_bh.release_sock.tcp_recvmsg.inet_recvmsg.sock_recvmsg
2.02 +0.0 2.07 perf-profile.calltrace.cycles-pp.ktime_get.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked.tcp_sendmsg
1.52 +0.0 1.57 perf-profile.calltrace.cycles-pp.read_tsc.ktime_get.tcp_write_xmit.__tcp_push_pending_frames.tcp_sendmsg_locked
1.91 +0.0 1.96 perf-profile.calltrace.cycles-pp.syscall_exit_to_user_mode.do_syscall_64.entry_SYSCALL_64_after_hwframe.write
1.50 +0.1 1.58 perf-profile.calltrace.cycles-pp.read_tsc.ktime_get.tcp_mstamp_refresh.tcp_rcv_space_adjust.tcp_recvmsg_locked
0.67 +0.1 0.74 ± 4% perf-profile.calltrace.cycles-pp.__virt_addr_valid.check_heap_object.__check_object_size.skb_do_copy_data_nocache.tcp_sendmsg_locked
1.89 +0.1 1.97 perf-profile.calltrace.cycles-pp.ktime_get.tcp_mstamp_refresh.tcp_rcv_space_adjust.tcp_recvmsg_locked.tcp_recvmsg
2.21 +0.1 2.31 perf-profile.calltrace.cycles-pp.tcp_mstamp_refresh.tcp_rcv_space_adjust.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg
2.63 +0.1 2.74 perf-profile.calltrace.cycles-pp.tcp_rcv_space_adjust.tcp_recvmsg_locked.tcp_recvmsg.inet_recvmsg.sock_recvmsg
1.64 +0.1 1.78 perf-profile.calltrace.cycles-pp.check_heap_object.__check_object_size.skb_do_copy_data_nocache.tcp_sendmsg_locked.tcp_sendmsg
2.38 +0.1 2.52 perf-profile.calltrace.cycles-pp.__check_object_size.skb_do_copy_data_nocache.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter
6.00 +0.2 6.18 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.write
3.06 +0.2 3.27 perf-profile.calltrace.cycles-pp._copy_from_iter.skb_do_copy_data_nocache.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter
5.78 +0.3 6.06 perf-profile.calltrace.cycles-pp.entry_SYSCALL_64.read
6.06 +0.4 6.42 perf-profile.calltrace.cycles-pp.skb_do_copy_data_nocache.tcp_sendmsg_locked.tcp_sendmsg.sock_write_iter.vfs_write
32.23 +0.5 32.69 perf-profile.calltrace.cycles-pp.tcp_sendmsg.sock_write_iter.vfs_write.ksys_write.do_syscall_64
6.06 ± 2% +1.4 7.46 perf-profile.calltrace.cycles-pp.release_sock.tcp_sendmsg.sock_write_iter.vfs_write.ksys_write
7.64 -0.7 6.91 perf-profile.children.cycles-pp._raw_spin_lock_bh
2.80 -0.5 2.31 perf-profile.children.cycles-pp.simple_copy_to_iter
83.66 -0.4 83.25 perf-profile.children.cycles-pp.do_syscall_64
84.74 -0.4 84.35 perf-profile.children.cycles-pp.entry_SYSCALL_64_after_hwframe
5.31 -0.3 4.97 perf-profile.children.cycles-pp.__check_object_size
3.46 -0.3 3.19 perf-profile.children.cycles-pp.check_heap_object
2.85 -0.1 2.71 ± 2% perf-profile.children.cycles-pp.tcp_send_mss
2.26 ± 2% -0.1 2.14 ± 2% perf-profile.children.cycles-pp.tcp_current_mss
0.91 -0.1 0.80 perf-profile.children.cycles-pp.__cond_resched
1.32 -0.1 1.21 ± 2% perf-profile.children.cycles-pp.__virt_addr_valid
0.27 ± 6% -0.1 0.18 ± 10% perf-profile.children.cycles-pp.__alloc_skb
0.49 -0.1 0.43 perf-profile.children.cycles-pp.rcu_all_qs
0.16 ± 8% -0.1 0.11 ± 10% perf-profile.children.cycles-pp.tcp_stream_alloc_skb
0.12 ± 6% -0.0 0.09 ± 8% perf-profile.children.cycles-pp.__tcp_send_ack
0.36 -0.0 0.34 perf-profile.children.cycles-pp.is_vmalloc_addr
0.50 -0.0 0.48 perf-profile.children.cycles-pp.tcp_push
0.19 ± 2% +0.0 0.21 ± 2% perf-profile.children.cycles-pp.read@plt
0.04 ± 33% +0.0 0.06 perf-profile.children.cycles-pp.kfree_skbmem
0.18 ± 2% +0.0 0.20 perf-profile.children.cycles-pp.inet_send_prepare
0.67 +0.0 0.68 perf-profile.children.cycles-pp.security_file_permission
1.04 +0.0 1.06 perf-profile.children.cycles-pp.x64_sys_call
0.40 +0.0 0.43 ± 2% perf-profile.children.cycles-pp.source
0.47 ± 2% +0.0 0.50 perf-profile.children.cycles-pp.inet_sendmsg
0.47 ± 2% +0.0 0.50 ± 4% perf-profile.children.cycles-pp.tick_nohz_handler
0.48 ± 2% +0.0 0.51 ± 4% perf-profile.children.cycles-pp.__hrtimer_run_queues
0.56 +0.0 0.60 perf-profile.children.cycles-pp.loop_transfer
1.66 +0.1 1.71 perf-profile.children.cycles-pp.rw_verify_area
0.35 +0.1 0.40 perf-profile.children.cycles-pp.security_socket_sendmsg
1.17 +0.1 1.22 perf-profile.children.cycles-pp.entry_SYSCALL_64_safe_stack
0.37 +0.1 0.43 perf-profile.children.cycles-pp.security_socket_recvmsg
3.95 +0.1 4.04 perf-profile.children.cycles-pp.syscall_exit_to_user_mode
2.93 +0.1 3.02 perf-profile.children.cycles-pp.entry_SYSRETQ_unsafe_stack
2.36 +0.1 2.46 perf-profile.children.cycles-pp.tcp_mstamp_refresh
2.72 +0.1 2.83 perf-profile.children.cycles-pp.tcp_rcv_space_adjust
3.21 +0.1 3.33 perf-profile.children.cycles-pp.read_tsc
4.41 +0.1 4.55 perf-profile.children.cycles-pp.ktime_get
3.11 +0.2 3.32 perf-profile.children.cycles-pp._copy_from_iter
6.96 +0.2 7.21 perf-profile.children.cycles-pp.entry_SYSCALL_64
6.25 +0.4 6.61 perf-profile.children.cycles-pp.skb_do_copy_data_nocache
32.48 +0.5 32.94 perf-profile.children.cycles-pp.tcp_sendmsg
9.63 +0.6 10.22 perf-profile.children.cycles-pp.release_sock
1.74 -0.6 1.15 perf-profile.self.cycles-pp.sock_read_iter
2.22 -0.5 1.72 perf-profile.self.cycles-pp.sock_write_iter
1.76 -0.1 1.63 perf-profile.self.cycles-pp.check_heap_object
1.21 ± 2% -0.1 1.11 ± 2% perf-profile.self.cycles-pp.__virt_addr_valid
1.19 -0.1 1.11 perf-profile.self.cycles-pp.__skb_datagram_iter
3.85 -0.1 3.78 perf-profile.self.cycles-pp.tcp_sendmsg_locked
1.46 ± 2% -0.1 1.39 perf-profile.self.cycles-pp.__check_object_size
1.32 -0.1 1.26 perf-profile.self.cycles-pp.__local_bh_enable_ip
0.55 -0.1 0.49 perf-profile.self.cycles-pp.lock_sock_nested
0.50 -0.1 0.44 perf-profile.self.cycles-pp.__cond_resched
0.38 ± 9% -0.0 0.33 ± 4% perf-profile.self.cycles-pp.tcp_rate_check_app_limited
0.83 -0.0 0.79 perf-profile.self.cycles-pp.tcp_current_mss
0.30 -0.0 0.26 ± 2% perf-profile.self.cycles-pp.rcu_all_qs
0.66 -0.0 0.63 perf-profile.self.cycles-pp.tcp_tso_segs
0.12 ± 4% -0.0 0.10 ± 6% perf-profile.self.cycles-pp.tcp_ack
0.27 -0.0 0.25 perf-profile.self.cycles-pp.is_vmalloc_addr
0.35 -0.0 0.34 perf-profile.self.cycles-pp._raw_spin_unlock_bh
0.40 -0.0 0.39 perf-profile.self.cycles-pp.tcp_push
0.11 +0.0 0.12 perf-profile.self.cycles-pp.__x64_sys_read
0.55 +0.0 0.57 perf-profile.self.cycles-pp.security_file_permission
0.36 +0.0 0.37 perf-profile.self.cycles-pp.tcp_release_cb
0.38 +0.0 0.39 perf-profile.self.cycles-pp.tcp_rcv_space_adjust
0.28 +0.0 0.30 perf-profile.self.cycles-pp.tcp_mstamp_refresh
0.93 +0.0 0.95 perf-profile.self.cycles-pp.x64_sys_call
0.77 +0.0 0.79 perf-profile.self.cycles-pp.ksys_write
0.04 ± 50% +0.0 0.06 perf-profile.self.cycles-pp.kfree_skbmem
1.18 +0.0 1.20 perf-profile.self.cycles-pp.entry_SYSCALL_64_after_hwframe
0.30 +0.0 0.33 ± 2% perf-profile.self.cycles-pp.source
0.37 +0.0 0.40 perf-profile.self.cycles-pp.inet_recvmsg
0.58 +0.0 0.61 perf-profile.self.cycles-pp.sock_recvmsg
0.76 +0.0 0.79 perf-profile.self.cycles-pp.entry_SYSCALL_64_safe_stack
0.25 +0.0 0.28 perf-profile.self.cycles-pp.security_socket_sendmsg
0.98 +0.0 1.01 perf-profile.self.cycles-pp.rw_verify_area
1.82 +0.0 1.86 perf-profile.self.cycles-pp.do_syscall_64
1.69 +0.0 1.73 perf-profile.self.cycles-pp.entry_SYSCALL_64
0.45 +0.0 0.49 perf-profile.self.cycles-pp.loop_transfer
0.26 +0.1 0.32 perf-profile.self.cycles-pp.security_socket_recvmsg
3.33 +0.1 3.40 perf-profile.self.cycles-pp.syscall_exit_to_user_mode
3.03 +0.1 3.10 perf-profile.self.cycles-pp.write
2.85 +0.1 2.94 perf-profile.self.cycles-pp.entry_SYSRETQ_unsafe_stack
3.00 +0.1 3.11 perf-profile.self.cycles-pp.read_tsc
2.98 +0.1 3.09 perf-profile.self.cycles-pp.read
4.48 +0.1 4.60 perf-profile.self.cycles-pp._raw_spin_lock_bh
2.99 +0.2 3.20 perf-profile.self.cycles-pp._copy_from_iter
4.48 ± 2% +0.4 4.92 ± 3% perf-profile.self.cycles-pp.release_sock
***************************************************************************************************
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/1/debian-12-x86_64-20240206.cgz/lkp-spr-2sp4/SYSCALL/50%/lmbench3
commit:
3b07108ada ("Merge tag 'linux_kselftest-fixes-6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest")
76a45d86f7 ("x86: prevent gcc from emitting rep movsq/stosq for inlined ops")
3b07108ada81a8eb 76a45d86f78bc61a27e382c287a
---------------- ---------------------------
%stddev %change %stddev
\ | \
44464148 ± 5% -4.0% 42700356 perf-stat.ps.cache-misses
0.13 -4.3% 0.12 lmbench3.syscall.fstat.latency.us
0.47 +1.5% 0.48 lmbench3.syscall.open/close.latency.us
2.08 ± 57% +1.8 3.88 ± 32% perf-profile.calltrace.cycles-pp.open64
3.54 ± 56% +2.6 6.17 ± 17% perf-profile.calltrace.cycles-pp.get_cpu_sleep_time_us.get_idle_time.uptime_proc_show.seq_read_iter.vfs_read
4.15 ± 57% +2.7 6.90 ± 18% perf-profile.calltrace.cycles-pp.get_idle_time.uptime_proc_show.seq_read_iter.vfs_read.ksys_read
2.15 ± 58% +1.8 3.96 ± 32% perf-profile.children.cycles-pp.open64
3.54 ± 56% +2.7 6.25 ± 16% perf-profile.children.cycles-pp.get_cpu_sleep_time_us
4.15 ± 57% +2.7 6.90 ± 18% perf-profile.children.cycles-pp.get_idle_time
2.58 ± 60% +2.1 4.64 ± 17% perf-profile.self.cycles-pp.get_cpu_sleep_time_us
***************************************************************************************************
lkp-icl-2sp7: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) 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-icl-2sp7/icache/stress-ng/60s
commit:
3b07108ada ("Merge tag 'linux_kselftest-fixes-6.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest")
76a45d86f7 ("x86: prevent gcc from emitting rep movsq/stosq for inlined ops")
3b07108ada81a8eb 76a45d86f78bc61a27e382c287a
---------------- ---------------------------
%stddev %change %stddev
\ | \
144451 ± 20% +60.0% 231102 ± 28% proc-vmstat.numa_pte_updates
602712 ± 6% +13.7% 685178 ± 8% proc-vmstat.pgfault
1140689 +3.4% 1179295 stress-ng.icache.ops
19011 +3.4% 19654 stress-ng.icache.ops_per_sec
3.932e+10 +4.0% 4.09e+10 perf-stat.i.branch-instructions
1.02 -4.2% 0.98 perf-stat.i.cpi
1.91e+11 +4.5% 1.995e+11 perf-stat.i.instructions
0.98 +4.4% 1.02 perf-stat.i.ipc
7998 ± 8% +16.3% 9301 ± 9% perf-stat.i.minor-faults
7998 ± 8% +16.3% 9301 ± 9% perf-stat.i.page-faults
1.02 -4.1% 0.98 perf-stat.overall.cpi
0.98 +4.3% 1.02 perf-stat.overall.ipc
3.862e+10 +4.0% 4.018e+10 perf-stat.ps.branch-instructions
1.876e+11 +4.5% 1.96e+11 perf-stat.ps.instructions
7789 ± 7% +17.5% 9148 ± 9% perf-stat.ps.minor-faults
7789 ± 7% +17.5% 9149 ± 9% perf-stat.ps.page-faults
1.157e+13 +3.9% 1.202e+13 perf-stat.total.instructions
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
reply other threads:[~2025-04-18 2:46 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=202504181042.54ea2b8a-lkp@intel.com \
--to=oliver.sang@intel.com \
--cc=lkp@intel.com \
--cc=mjguzik@gmail.com \
--cc=oe-lkp@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.