On Fri, Jun 12, 2009 at 03:03:58PM +0200, Harald Hoyer wrote:
> Nice find!
>
> Maybe we can run the "real world test suite" ("benchmark") before the
> next release and try to straighten out such odds.
>
> Most of the benchmark results they post are not showing scientific
> results, only when something is really odd, like here, they have their
> use.
>
It's almost certainly attributable to the default install using audit.
Roland and various others have done a lot of work improving things, but
there is always going to be a per-syscall overhead to this kind of
thing. A few extra usec a syscall adds up after a few hundred thousand
calls...
regards, Kyle
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
