Intel ( i845G ) profiling

Simon Thum simon.thum at
Wed Mar 12 04:52:08 PDT 2008

Stephane Marchesin wrote:
> On 3/7/08, Simon Thum <simon.thum at> wrote:
>> Stephane Marchesin wrote:
>>  > On 3/7/08, The Rasterman Carsten Haitzler <raster at> wrote:
>>  >
>>  >>  of course a pure box filter would be even better than this... and i suspect
>>  >>  some hardware may support that. beyond that you might not find much hardware
>>  >>  support and if its back to software in xrender with what i consider "phd
>>  >>  research level filter experiments" then i dont think this is something xorg
>>  >>  developers should be worrying about. if you want such a filter - it will end up
>>  >>  in software anyway and so you may as well implement it yourself and pre-scale in
>>  >>  software.
>>  >
>>  > Not really, there are a couple of papers on the topic of implementing
>>  > better filters on existing hardware which prove that this is possible
>>  > :
>>  >
>>  >
>>  >
>>  > So what you're discussing is definitely feasible today, it's just that
>>  > people are not ready to invest much time into the implementation.
>> I don't think that argument extends to the linear filters I described
>>  (the abstracts don't mention linearity aspects so they'll probably
>>  ignore it), leaving as an implementation option only software or some
>>  form of GPGPU programming. So we're again with the integrated 2D/3D
>>  pipeline everyone wants :)
> Well, I implemented this, so I know that it does arbitrary FIR
> filters. But maybe you know better from just reading the abstract...
> Stephane
Sorry I didn't mean to offend you. I made a guess which wasn't properly 
I meant 'linear' as in gamma-corrected or 'linear with luminance'. This 
essentially means you need at least 10 bits of precision or tricky 
lookups (or accept loss). I don't say thats infeasible or you/the 
authors didn't do so, but usually this is mentioned. Though 
'High-quality' could refer to that, I don't know.



More information about the xorg mailing list