Re: [LKP] Re: 2463a604a8: netperf.Throughput_tps 12.8% improvement

[Xing Zhengjun]

On 5/26/2021 12:44 AM, Bart Van Assche wrote:
> On 5/24/21 11:44 PM, kernel test robot wrote:
> > FYI, we noticed a 12.8% improvement of netperf.Throughput_tps due to commit:
> >
> > commit: 2463a604a86728777ce4284214a52de46a808c9e ("[PATCH v3 2/3] Introduce enums for the SAM, message, host and driver status codes")
> > url: https://github.com/0day-ci/linux/commits/Bart-Van-Assche/Introduce-enums-for-SCSI-status-codes/20210524-105751
> > base: https://git.kernel.org/cgit/linux/kernel/git/mkp/scsi.git for-next
> >
> > in testcase: netperf
> > on test machine: 192 threads 4 sockets Intel(R) Xeon(R) Platinum 9242 CPU @ 2.30GHz with 192G memory
> > with following parameters:
> >
> > 	ip: ipv4
> > 	runtime: 300s
> > 	nr_threads: 16
> > 	cluster: cs-localhost
> > 	test: TCP_CRR
> > 	cpufreq_governor: performance
> > 	ucode: 0x5003006
> >
> > test-description: Netperf is a benchmark that can be use to measure various aspect of networking performance.
> > test-url: http://www.netperf.org/netperf/
> The above email reports a performance improvement for the networking
> subsystem while my patch only affects the SCSI subsystem and should not
> have any performance impact. I'm confused by the above feedback ...

I suspect it related with cache alignment, 2463a604a8 changes "u8" 
(size:1) to "enum xxx_status" (size: 4),  the cache alignment is better 
than before , so cause the improvement.
> Bart.
> _______________________________________________
> LKP mailing list -- lkp@xxxxxxxxxxxx
> To unsubscribe send an email to lkp-leave@xxxxxxxxxxxx

--
Zhengjun Xing