Re: [Dri-users] Degugging DRI

[Devon M. <dev...@gm...>]

On 1/3/07, Aprak Ofatys <apr...@gm...> wrote:
> OK, I think I get the point. One can't make any assumptions and needs to
> keep an open mind towards any possibility.
>
> Does anybody know of any GL debugging methodology? Since I have the sources
> of the applications that causes the problem I was thinking of linking with
> the ld '--wrap' option to each gl function. The wrapped functions should log
> on file that they are accessed and the result of each call to the original
> function. Does this sound as a reasonable way to begin to approach this
> problem?

Since you didn't mention Ogre issuing any warnings, I would assume
none of the calls are failing. Probably one or more are claiming the
succeeded when they have actually failed..

Take a look at some of the example code from here http://opengl.org/code/
Your goal should be to find something really simple that works, but
the equivalent code under Ogre does not. Think something with the
relative complexity of Hello World.

Once you have something from opengl.org that works, but fails when
done in Ogre, wrap and log the GL calls. See if you can see what
call/parameter Ogre is making/passing that is different from the
sample code.

Most of all, be patient! Debugging hardware/firmware/drivers is rarely
as simple or straightforward as one would like.

dcm