[GRASS5] Re: NVIZ crashes

Markus Neteler neteler at itc.it
Tue Aug 14 08:50:26 EDT 2001

On Tue, Aug 14, 2001 at 08:50:49AM -0300, Bob Covill wrote:
> Markus Neteler wrote:
> > >
> > > [snip]
> > >
> > > To both Markus and Helena:
> > >
> > > What happens if you just disable loading "dri", but still load "glx"?
> > >
> > > If there are problems when using the XFree86 GLX extension, that
> > > should probably be investigated further. OTOH, if the problems are
> > > limited to DRI, that's to be expected. Although, you might want to add
> > > a "don't use DRI" notice, just in case anyone still hasn't figured
> > > that one out for themselves yet ;)
> > 
> > Thanks for the hint, I'll try. However, it must be video card dependent
> > as I had both "dri" and "glx" active in Hannover.
> > 
> > Later,
> > 
> >  Markus
> Markus and Helena;
> I finally got RedHat Linux 7.1 re-installed. To date I have had no
> problems with NVIZ. I have tried repeatedly opening and closing it with
> no problems. 
> The motherboard is an older single processor (300 MHz Pentium II). The
> video card is a Matrox Milleneum G200.
> Helena -- do you know the make and model of the video card in your
> system?
> Hopefully we can pull all these bits of information together for a good
> solution. You seem to be getting close Markus.

Bob and others,

we have tested today for a while. The problem is DRI (direct rendering,
needed for openGL hardware acceleration). If DRI is disabled, NVIZ
works on Xfree4.0.3 (redhat 7.1) even when glx is active.

If DRI is enabled, the mesa-glxdemo programs run, but nviz crashes.
It is talking about a bad integer initialization (integer out of range).
No idea, how to debug this.

On thursday I'll switch to Xfree4.1.0, which provides an improved DRI.
Eventually the problems will be solved then... if not, it is a NVIZ
problem (probably the TOGL).

Bob, if you are willing to upgrade the TOGL1.0 to TOGL1.5, eventually
this will help as well? Of course I would test it here! Do you expect
side effects from such upgrade?


More information about the grass-dev mailing list