Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 16 May 2001 17:49:37 +1000 (EST)
From:      Bruce Evans <bde@zeta.org.au>
To:        Dave Leimbach <dleimbac@MPI-Softtech.Com>
Cc:        dleimbac@earthlink.net, wes@softweyr.com, freebsd-questions@FreeBSD.ORG, arch@FreeBSD.ORG
Subject:   Re: Gettimeofday Again...
Message-ID:  <Pine.BSF.4.21.0105161741590.8023-100000@besplex.bde.org>
In-Reply-To: <200105151440.JAA01279@MPI-Softtech.Com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 15 May 2001, Dave Leimbach wrote:

> I think SMP is a way more important topic so long as gettimeofday doesn't get 
> called in thread context switches :)..  This was mentioned before and the 

gettimeofday() (actually the internal kernel part of it) certainly does get
called in thread context switches.  This has a fairly small impact on
context switching time.  According to lmbench2 for 2 tiny  processes (see
my previous mail for some details on the machine):

Linux 2.2.9               1 usec
Linux 2.4.0.something     1 usec
FreeBSD-4.0.something     1 usec
FreeBSD-current           2 usec

microtime() in the context switch takes about 0.4 usec.  This accounts
for about half of the pessimizations in SMPng according to the above
measurements.  But this may misleading since most of the times in the above
are rounded to the nearest usec.

Bruce


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0105161741590.8023-100000>