glxgears is a benchmark, just doesn't benchmark what people often expect. I
often use it as a present/swapbuffers benchmark. So the whole blurb about not
using glxgears as a benchmark is false.
Again, detecting software rendering by the string is unreliable.
Maybe it help if you could provide some context for your motivation for doing
this. Is the purpose here educate users, or avoid invalid bug reports, or what?
The warning about direct rendering is ok, but should be printed to stderr, and
no need for capital "WARNING" as glxgears doesn't produces almost no output so
any warning should standout easily.
Jose
----- Original Message -----
> From: Matthias Hopf <mh...@suse.de>
>
> ---
> src/xdemos/glxgears.c | 8 ++++++++
> 1 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/src/xdemos/glxgears.c b/src/xdemos/glxgears.c
> index cff92b0..c64c2c9 100644
> --- a/src/xdemos/glxgears.c
> +++ b/src/xdemos/glxgears.c
> @@ -766,6 +766,14 @@ main(int argc, char *argv[])
> printf("GL_EXTENSIONS = %s\n", (char *)
> glGetString(GL_EXTENSIONS));
> }
>
> + if (! glXIsDirect(dpy, ctx))
> + printf ("\n***\n*** WARNING: Direct Rendering is NOT
> enabled\n***\n");
> + else if (strcasestr ((char *) glGetString(GL_RENDERER),
> "software"))
> + printf ("\n***\n*** WARNING: Hardware acceleration is NOT
> active\n***\n");
> + printf ("\n*** NOTE: Don't use glxgears as a benchmark.\n"
> + " OpenGL implementations are not optimized for frame rates >>
> 60fps,\n"
> + " thus these numbers are meaningless when compared between
> vendors.\n\n");
> +
> init();
>
> /* Set initial projection/viewing transformation.
> --
> 1.7.7
>
> _______________________________________________
> mesa-dev mailing list
> mesa-dev@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
>
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/mesa-dev
