On Sun, May 24, 2026 at 07:44:57AM -0700, Breno Leitao wrote:
> Throughput improves +6% to +28% and average write latency drops 5%
> to 22% across every configuration.
FWIW: Kernel Test Robot detected similar and slightly better numbers
with this patch, and reported 29.1% improvement on lmbench3
%stddev %change %stddev
\ | \
37316 +29.1% 48191 ± 2% lmbench3.PIPE.bandwidth.MB/sec
5690 ± 9% +153.5% 14426 ± 18%
lmbench3.time.involuntary_context_switches
https://lore.kernel.org/all/[email protected]/
