Date: Thu, 27 Oct 2005 10:53:27 +0200 From: "Poul-Henning Kamp" <phk@phk.freebsd.dk> To: "Yuriy N. Shkandybin" <jura@networks.ru> Cc: David Xu <davidxu@freebsd.org>, current@freebsd.org Subject: Re: MySQL Performance 6.0rc1 Message-ID: <21264.1130403207@critter.freebsd.dk> In-Reply-To: Your message of "Thu, 27 Oct 2005 12:24:30 %2B0400." <00a801c5dacf$db3b7700$6504010a@Jura>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <00a801c5dacf$db3b7700$6504010a@Jura>, "Yuriy N. Shkandybin" writes:
>>>Check gettimeofday syscall, it follows every I/O syscall, I think
>>>our gettimeofday is tooooooo expensive, if we can directly get time from
>>>memory, the performance will be improved further.
>It's true:
>run next on same PC от freebsd and linux and compare
I can produce unprecise timestamps just as fast as Linux.
Heck, I can do it even faster:
static __inline int
gettimeofday(struct timeval *tp, void *tz __unused)
{
tp.tv_sec = 0;
tp.tv_usec = 0;
return (0);
}
A very important part of learning how to do benchmarks is learning
how to measure the target and not the benchmarking code.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?21264.1130403207>