[Openchrome-users] 3D acceleration

Jon Nettleton jon.nettleton
Thu Sep 14 14:29:37 PDT 2006


On Thu, 2006-09-14 at 16:07 -0500, Joel Means wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> I am wondering if I am getting 3D acceleration.  I am now using an
> MII10000 board (CLE266 chipset).  glxinfo indicates that direct
> rendering is enabled, but when I run glxgears, I am at ~100% cpu usage
> to get ~425fps.  It seems like I shouldn't be getting that kind of cpu
> usage if I am getting 3D acceleration. 


My guess is you are running at 24-bit color depth.  If that is the case
then those numbers are about right.  The 3d engine is not very powerful
on the via chipsets.  Running in 16-bit color depth you should get
between 700 and 800 fps.  

Don't mind glxgears pegging the cpu that is just what that program does.
It is a horrible benchmark for 3d engine performance, just a very simple
opengl program.


>  My xorg.conf has the following
> options set:
> 
> "EnableAGPDMA"
> "DisableIRQ"
> "RenderAccel" "true"
> 
> I have been through everything I can find on the net about setting up
> DRI stuff, but I just wonder why my cpu usage is so high.
> The main reason that I need to resolve this is that I am using MythTV
> with the OpenGL renderer and between mythfrontend and X, I am using all
> of the available cpu.  

Is this just the new OpenGL Mythtv gui in 0.20, or while watching video?

> This has the responsiveness of Myth being really
> slow.  Is this behavior unavoidable, or do I have something set up
> wrong?  I noticed on the website that it is recommended to use
> MesaLib-6.4, but I have 6.5 installed?  Is this the problem?  I would
> like some input before I try to roll back to 6.4.2 since I have Xorg-7.1
> installed and would have to roll that back to 7.0 meaning a reinstall of
> a whole lot of software.  Any advice is greatly appreciated.

I don't think you should need to roll everything back.  First off I
would try out 16-bit color depth...yes I know it isn't optimal but it
should help to see where the bottleneck is.  After that we can take some
other debugging steps.

Jon

> Thanks,
> Joel
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.3 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
> 
> iD8DBQFFCcSY0tbb2uIZyCURAjgHAJoDv1DH6h8jtnO6RkSR63dPGf7f4QCfd64q
> ty8VdlwZNT9wEHjiQpUt6+g=
> =8gve
> -----END PGP SIGNATURE-----
> 
> _______________________________________________
> openchrome-users mailing list
> openchrome-users at openchrome.org
> http://wiki.openchrome.org/mailman/listinfo/openchrome-users
> Main page:
> http://www.openchrome.org
> Wiki:
> http://wiki.openchrome.org
> User Forum:
> http://wiki.openchrome.org/tikiwiki/tiki-view_forum.php?forumId=1
> 
> 





More information about the Openchrome-users mailing list