Rob Jenkins (robj++at++quid.csd.sgi.com)
Thu, 28 May 1998 07:51:42 -0700
There have been various texture manager problems in the past that can give
symptoms like this under certian conditions. What you describe sounds like a GL
level problem, not Performer, as you're seeing textures that aren't from the
program you're currently running. The texture manager for each platform is
different, the platform I've seen this kind of thing on most is Impact/OCTANE
but with recent patches almost all those went away so I would suggest that you
make sure you have the latest gfx patch for the platform with the problem:
Impact: patch 1935 ( there is a later patch 2677 but I don't think that's in a
patch set yet and doesn't seem to have texture manager fixes over 1935 ).
OCTANE: patch 2895
Onyx2 iR: patch 2922
Onyx iR: patch 2327
Onyx RE2: patch 2739
Also, for OCTANE/Impact at least I know there are extra fixes in the Irix 6.5
texture manager. If you have the latest patches for your machine and still see
a problem then please get the test case down to the smallest possible that
shows the problem ( although it's more important just to have a test case even
if it's big, reducing the size just makes life easier ) then log a support call
- are you a Beta test site for Irix 6.5 ?
Cheers
Rob
On May 27, 5:32pm, Scott Herod wrote:
> Subject: Elusive texture memory problem
> Are there any known problems with performer (or gl) paging
> textures back and forth between hardware texture memory?
> (Performer 2.0.4) In particular, we've got a performer scene
> with multiple small textures created from raster images. If
> we repeatedly play the scene, it runs fine for ~30 times (not
> always the same number) and then one texture replaces another.
>
> I've seen the problem occur between separate scenes, and
> it has been reported that once a Doom texture appeared where
> it couldn't possibly be created. (Fortunately not on my machine.)
>
> I have a lot of ideas as to what the problem is not...
> We're not confusing our pointers to the beginnings of the
> raster images. pfDelete is called religiously. We are not
> opening files of raster images.
>
> I have only seen textures replace those (supposedly) created
> later, never the other way around. I've only seen textures
> be replaced by similar sized textures.
>
> The pfuClearTextureMemory() routine which is part of
> pfuTextureMemory corrects the problem with some scenes if it
> is called after the scene is deleted but it does not effect
> other examples.
>
> Any suggestions of other places to look would be greatly
> appreciated.
>
> Scott Herod
> herod++at++evt.com
> =======================================================================
> List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com
>-- End of excerpt from Scott Herod
--
________________________________________________________________
Rob Jenkins mailto:robj++at++sgi.com
Silicon Graphics, Mtn View, California, USA
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:57:26 PDT