[linus:master] [ext4] 7f373083b4: fio.write_iops 54.8% improvement

From: kernel test robot

Date: Sun Nov 08 2026 - 03:09:31 EST




Hello,

kernel test robot noticed a 54.8% improvement of fio.write_iops on:


commit: 7f373083b45757b19f36a939f255239d40ccfb24 ("ext4: avoid unnecessary transaction in ext4_map_blocks() for unwritten extents") https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master

testcase: fio-basic
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:

runtime: 300s
disk: 1HDD
fs: ext4
nr_task: 1
test_size: 128G
rw: write
bs: 4k
ioengine: falloc
cpufreq_governor: performance


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


The kernel config and materials to reproduce are available at:
https://download.01.org/0day-ci/archive/20261108/202611080817.b5c132ae-lkp@xxxxxxxxx

=========================================================================================
bs/compiler/cpufreq_governor/disk/fs/ioengine/kconfig/nr_task/rootfs/runtime/rw/tbox_group/test_size/testcase:
4k/gcc-14/performance/1HDD/ext4/falloc/x86_64-rhel-9.4/1/debian-13-x86_64-20250902.cgz/300s/write/lkp-icl-2sp9/128G/fio-basic

commit:
6585cea3cf ("ext4: allow ext4_map_blocks() to start its own transaction handle")
7f373083b4 ("ext4: avoid unnecessary transaction in ext4_map_blocks() for unwritten extents")

6585cea3cf58b8e8 7f373083b45757b19f36a939f25
---------------- ---------------------------
%stddev %change %stddev
\ | \
0.08 ± 5% -0.1 0.02 ± 4% fio.latency_2us%
0.05 ± 2% -0.0 0.02 ± 5% fio.latency_4us%
27.91 -34.9% 18.16 fio.time.elapsed_time
27.91 -34.9% 18.16 fio.time.elapsed_time.max
17.90 -46.7% 9.54 fio.time.system_time
3013 -34.2% 1984 fio.time.voluntary_context_switches
4758 +54.8% 7366 fio.write_bw_MBps
648.00 -44.7% 358.67 fio.write_clat_90%_ns
652.00 -44.4% 362.67 fio.write_clat_95%_ns
664.00 -43.9% 372.67 fio.write_clat_99%_ns
632.44 -43.9% 354.58 fio.write_clat_mean_ns
1218093 +54.8% 1885792 fio.write_iops


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