Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 7 Jan 1999 16:46:22 +0000 (GMT)
From:      Doug Rabson <dfr@nlsystems.com>
To:        Stuart Anderson <anderson@metrolink.com>
Cc:        freebsd-ports@FreeBSD.ORG, freebsd-alpha@FreeBSD.ORG
Subject:   Re: Mesa3 on alpha
Message-ID:  <Pine.BSF.4.01.9901071641320.391-100000@herring.nlsystems.com>
In-Reply-To: <Pine.BSF.3.96.990107105430.10087A-100000@trantor.sc.metrolink.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 7 Jan 1999, Stuart Anderson wrote:

> On Thu, 7 Jan 1999, Doug Rabson wrote:
> 
> > I think that the right way to go here is to change system compilers to
> > something more modern but that would be fairly hard :-)  How hard would it
> > be to make Mesa3 depend on the egcs port?
> > 
> > For what its worth, RedHat 5.0 (which used gcc 2.7.2.3) built Mesa without
> > optimisations.  
> 
> Does it run OK? When porting OpenGL, I found that egcs was the only compiler
> that would generate correct code for the Alpha. The others would always
> end up SIGFPEs.

It seems to run most of the tests ok (built non-optimised) but I can't
claim to have actually tested it much.  It probably needs to be built with
the software completion flags (-mtrap-precision=i -mfp-trap-mode=su) since
I can imagine generating a few denormalised numbers when rasterising a few
pathological triangles.

In alpha code, strange SIGFPEs often mean that your program needs software
completion since the chip won't perform calculations which either use or
generate non-finite or denormalised numbers.

--
Doug Rabson				Mail:  dfr@nlsystems.com
Nonlinear Systems Ltd.			Phone: +44 181 442 9037



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-ports" 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.4.01.9901071641320.391-100000>