[linus:master] [fs] 177fdbae39: fxmark.ssd_ext4_no_jnl_MRPL_4_bufferedio.works/sec 6.8% improvement

Hello,

kernel test robot noticed a 6.8% improvement of fxmark.ssd_ext4_no_jnl_MRPL_4_bufferedio.works/sec on:

commit: 177fdbae39ecccb441d45e5e5ab146ea35b03d49 ("fs: inline step_into() and walk_component()")
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

testcase: fxmark
config: x86_64-rhel-9.4
compiler: gcc-14
test machine: 64 threads 2 sockets Intel(R) Xeon(R) Gold 6346 CPU @ 3.10GHz (Ice Lake) with 256G memory
parameters:

	disk: 1SSD
	media: ssd
	test: MRPL
	fstype: ext4_no_jnl
	directio: bufferedio
	thread_nr: 4
	cpufreq_governor: performance


Details are as below:
-------------------------------------------------------------------------------------------------->


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20260121/[email protected]

=========================================================================================
compiler/cpufreq_governor/directio/disk/fstype/kconfig/media/rootfs/tbox_group/test/testcase/thread_nr:
  gcc-14/performance/bufferedio/1SSD/ext4_no_jnl/x86_64-rhel-9.4/ssd/debian-11.1-x86_64-20220510.cgz/lkp-icl-2sp8/MRPL/fxmark/4

commit: 
  9d2a6211a7 ("fs: tidy up step_into() & friends before inlining")
  177fdbae39 ("fs: inline step_into() and walk_component()")

9d2a6211a7b97256 177fdbae39ecccb441d45e5e5ab 
---------------- --------------------------- 
         %stddev     %change         %stddev
             \          |                \  
   7403946            +6.8%    7907512        fxmark.ssd_ext4_no_jnl_MRPL_4_bufferedio.works/sec
      8028 ±  3%     -13.1%       6980 ±  5%  numa-meminfo.node0.KernelStack
      5780 ±  5%     +18.5%       6851 ±  5%  numa-meminfo.node1.KernelStack
      8029 ±  3%     -13.0%       6982 ±  6%  numa-vmstat.node0.nr_kernel_stack
      5781 ±  5%     +18.5%       6852 ±  5%  numa-vmstat.node1.nr_kernel_stack
      3107 ± 13%     -14.2%       2665        perf-sched.total_wait_and_delay.max.ms
      3107 ± 13%     -14.2%       2665        perf-sched.total_wait_time.max.ms
   1610350            +2.0%    1642966        perf-stat.i.cache-references
   1587301            +2.0%    1619507        perf-stat.ps.cache-references
      6.94 ± 83%      -6.1        0.88 ±223%  perf-profile.calltrace.cycles-pp.iput.__dentry_kill.dput.__fput.task_work_run
      8.67 ± 84%      -4.8        3.89 ±143%  perf-profile.calltrace.cycles-pp.mutex_unlock.sw_perf_event_destroy.__free_event.perf_event_release_kernel.perf_release
      7.84 ± 83%      -3.3        4.52 ±163%  perf-profile.calltrace.cycles-pp.tlb_finish_mmu.exit_mmap.__mmput.exit_mm.do_exit
      5.17 ±117%      -2.3        2.86 ±144%  perf-profile.calltrace.cycles-pp.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput.exit_mm
      5.17 ±117%      -2.3        2.86 ±144%  perf-profile.calltrace.cycles-pp.folios_put_refs.free_pages_and_swap_cache.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap
      5.17 ±117%      -2.3        2.86 ±144%  perf-profile.calltrace.cycles-pp.free_pages_and_swap_cache.__tlb_batch_free_encoded_pages.tlb_finish_mmu.exit_mmap.__mmput
      6.94 ± 83%      -6.1        0.88 ±223%  perf-profile.children.cycles-pp.iput
     10.34 ± 99%      -5.3        5.00 ±152%  perf-profile.children.cycles-pp.mutex_unlock
      7.84 ± 83%      -3.3        4.52 ±163%  perf-profile.children.cycles-pp.tlb_finish_mmu
      5.17 ±117%      -2.3        2.86 ±144%  perf-profile.children.cycles-pp.__tlb_batch_free_encoded_pages
      5.17 ±117%      -2.3        2.86 ±144%  perf-profile.children.cycles-pp.free_pages_and_swap_cache
      6.94 ± 83%      -6.1        0.88 ±223%  perf-profile.self.cycles-pp.iput
     10.34 ± 99%      -5.3        5.00 ±152%  perf-profile.self.cycles-pp.mutex_unlock
      3.78 ±100%      -1.4        2.38 ±223%  perf-profile.self.cycles-pp.zap_present_ptes



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