Intel ( i845G ) profiling

Simon Thum simon.thum at gmx.de
Thu Mar 13 14:50:16 PDT 2008


Matthias Hopf wrote:
> On Mar 08, 08 01:03:32 +0100, Simon Thum wrote:
>> I don't think a correct filter is what would make a phd these days, but 
>> anyway - no-one says it is xorg's responsibility to provide one. Though 
>> it would be nice. TBH I guessed that scaling is in SW anyway - I'm 
>> amazed this is done in HW.
> 
> ?
> All texturing is about is actually filtering and scaling...
> And bilinear filters were there in graphics hardware almost from day
> one as soon as textures were involved.
> 
> With anisotropic filtering even the low-level TEX commands of the
> fragment programs do higher order filtering in hardware. With fragment
> programs arbitrary filters are pretty easy to achieve anyway,
> Hardwiger's paper was more about multipass algorithms, you can also find
> some papers about arbitrary wavelet filters on GPUs written by me, so
> this is pretty old knowledge by now.

Rasterman was referring to XRender scaling:

> ten i can get onto my next favorite game and bitching about the quality of the
> rendering ( e.g. filtered downscaling in xrender :):) )

I hadn't assumed this runs through the 3D pipeline at all. Also, XRender 
accel is a domain of the driver, so things may vary. Does it?




More information about the xorg mailing list