From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Lameter Date: Tue, 03 Aug 2004 21:06:26 +0000 Subject: Re: Timer updates revision 7 (asm sets predicates/various fixes) Message-Id: List-Id: References: In-Reply-To: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-ia64@vger.kernel.org On Tue, 3 Aug 2004, David Mosberger wrote: > >>>>> On Mon, 2 Aug 2004 19:33:56 -0700 (PDT), christoph@lameter.com said: > > Christoph> On Mon, 2 Aug 2004, David Mosberger wrote: > >> >>>>> On Fri, 30 Jul 2004 08:30:19 -0700 (PDT), Christoph Lameter said: > >> > Christoph> On Thu, 29 Jul 2004, David Mosberger wrote: > >> >> It occurred to me now why your numbers are higher: I linked my > >> >> program statically, whereas yours is probably linked dynamically? > >> > Christoph> Correct. But why would this have an influence? > >> > >> I'm just trying to determine why our numbers were different. I dont like > >> unexplained differences. > > Christoph> The different numbers may be due to the different kinds of scaling > Christoph> applied to the clock frequency to produce the ITC frequency. > > No, I think they're precisely because you linked the program dynamically. > Try linking it statically. I'm fairly confident you'll get the > same/very similar numbers as I did. When I link it statically the fastcall handler is not used and the numbers are much higher. Maybe we are using different glibcs? This is with SUSE SLES9.