Jean-Pierre Dussault wrote:
>
> Good evening!
>
> A week ago, I reported a memory bug in Mesa-3.1. Since I use Mesa from
> within the Sather programming language, I found it difficult to provide
> you with a "small" program exhibiting the bug. Moreover, I feared that
> the Sather programming language might do incompatible things with
> respect to Mesa.
>
> I found a (not so small) "approved" program exhibiting the bug in one of
> the demos in the glut distribution: the glutmech demo, in
> "glut-3.7/progs/demos/glutmech". Here follows its resident size, as
> displayed by "top" on Sun/Solaris comparing the sises using Mesa-3.1 and
> Mesa-3.0.
>
> 4209 root -25 0 46M 19M run 0:51 39.19% 48.47%
> glutmech3.1
> 4208 root -15 0 7072K 6152K run 0:46 33.98% 45.21%
> glutmech3.0
>
> I searched for a demo using display lists, since I suspect that it is in
> the implementation of display lists that the bug origins.
>
> I hope this observation may help you!
Keith recently checked in some changes to the display list code
to reduce the amount of memory used. This happened since the
Mesa 3.1 beta 3 release.
Could you try the latest Mesa CVS sources with your app?
-Brian
_______________________________________________
Mesa-bug maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-bug
_______________________________________________
Mesa-dev maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-dev
