Date: 26 Feb 2002 12:46:36 -0600 From: Bob Van Valzah <Bob@Talarian.Com> To: smp@FreeBSD.Org Subject: Performance vs. Stable Message-ID: <1014749196.231.48.camel@NewStorm.WhiteBarn.Com>
next in thread | raw e-mail | index | archive | help
I have run lmbench on 4.5-STABLE and 5.0-CURRENT (as of late last week).
I believe this might give some insight into the relative performance of
these kernels on uniprocessor micro benchmarks. (I know there a lot of
other things I could benchmark and I don't mean to suggest that this is
a particularly *important* area. It was just an easy area to do first.)
I am being cautious in interpreting the results since I don't have much
experience running lmbench and I haven't yet done anything to confirm my
results. I have some thoughts about where I should go next in
benchmarking, but I welcome your input too.
(Best numbers are starred, i.e., *123)
Processor, Processes - factor slower than the best
--------------------------------------------------
Host OS Mhz Null Null Simple /bin/sh Mmap 2-proc
8-proc
Syscall Process Process Process lat
ctxsw ctxsw
--------- ------------- ---- ------- ------- ------- ------- ---- ------
------
TwinHead FreeBSD 4.5-S 228 *61 *3.7K *13.6K *24.3K *121
*27 *30
TwinHead. FreeBSD 5.0-2 232 1.8 1.4 1.5 1.5 1.8
3.7 4.9
If I interpret the above output correctly, the largest slowdown comes in
context switching. I haven't yet looked with kgmon/gprof to figure out
where the difference comes from, but that might be a reasonable next
step.
Bob
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1014749196.231.48.camel>