Re: DRM update for testing

From: Eric Anholt <eta_at_lclark.edu>
Date: Tue, 20 Dec 2005 13:25:01 -0800
On Tue, 2005-12-20 at 10:43 +0100, Oliver Fromme wrote:
> Eric Anholt <eta_at_lclark.edu> wrote:
>  > Oliver Fromme wrote:
>  > > Freddie Cash <fcash_at_ocis.net> wrote:
>  > > > Running glxgears at 1024x768x16 gives me a decent 35 fps, compared
>  > > > to the 1 fps I used to get.  :)
>  > >  
>  > > Uhm.  Are you sure that you're running hardware-accelerated
>  > > OpenGL glxgears?
>  > >  
>  > > I get 48 fps at 1400x1050x32 -- in software, without any
>  > > hardware 3D acceleration.  (It's a 1.6GHz Centrino notebook
>  > > with shared i915 graphics.  2D acceleration is enabled, of
>  > > course.)
>  > 
>  > glxgears is not a benchmark.
> 
> That's right.  I was not suggesting to use it as a real-
> world benchmark, but rather as a rough indication whether
> hardware acceleration is enabled at all or not.
> 
> If someone gets 1 fps (!) out of glxgears, something
> is _seriously_ wrong, even with software rendering.

No, that's what I'm trying to say here.  There's nothing seriously wrong
with 1 fps software glxgears at full screen, or anything wrong at all.
I get about 9 fps at 1600x1200x16 on my amd64.  1 fps is totally
believable on a different graphics card and a slower cpu.

-- 
Eric Anholt                                     eta_at_lclark.edu
http://people.freebsd.org/~anholt/              anholt_at_FreeBSD.org

Received on Tue Dec 20 2005 - 20:25:14 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:49 UTC