[MPlayer-dev-eng] Video output for RV280+Mesa3D

[malc]

On Sat, 6 Dec 2008, Reimar D?ffinger wrote:

> On Sat, Dec 06, 2008 at 01:19:16PM +0300, malc wrote:
>> On Sat, 6 Dec 2008, Reimar D?ffinger wrote:
>>> Actually it was much more stupid, I forgot to set
>>> GL_UNPACK_CLIENT_STORAGE_APPLE (fixed in SVN).  Still does not make
>>> a difference with my i945...
>>
>> You are mistaken. GL_UNPACK_CLIENT_STORAGE_APPLE has no effect
>> whatsoever unless the stride is sufficiently aligned (again at least
>> with r200).
>
> Yes, I don't doubt that and I made sure of that, but for my sample that
> was the case all along (and forgetting GL_UNPACK_CLIENT_STORAGE_APPLE
> definitely is the more stupid error).
>
>>  Anyhow the whole reliance on `-vf format=yuy2' makes sure
>> that vo_gl will never approach vo_rv280's speed since there's an extra
>> copying data around going on (decode -> convert -> copy to tex instead
>> of decode -> covnert to tex) In fact simply adding `-vf format=yuy2'
>> to `-vo rv280' increases the time taken to show that clip by 3
>> seconds.
>
> That is because your rv280 does not support direct-rendering (-dr
> option), vo_gl does, so with that option no extra copying is involved
> (except when the codec decodes directly to yuy2 and uses reference
> frames - but in that case vo_rv280 copies as well, actually even
> regardless of whether they are reference frames or not).

I see, well i converted the aforementioned clip to 768x576 and ran it with:
./mplayer /home/malc/x/vid/moo.avi -vf format=yuy2 -vo gl:ycbcr:mesa-buffer:rectangle=1:force-pbo:swapinterval=0 -dr -nosound -benchmark

yielding
BENCHMARKs: VC:   4.456s VO:  11.587s A:   0.000s Sys:   0.091s =   16.134s
BENCHMARK%: VC: 27.6162% VO: 71.8168% A:  0.0000% Sys:  0.5670% = 100.0000%

and ./mplayer /home/malc/x/vid/moo.avi -vo rv280 -nosound -benchmark
BENCHMARKs: VC:   3.606s VO:   1.799s A:   0.000s Sys:   0.088s =    5.492s
BENCHMARK%: VC: 65.6490% VO: 32.7538% A:  0.0000% Sys:  1.5971% = 100.0000%

I'd venture a guess that the difference in timing compared to the
previously posted results is due to the fact that i converted it to
avi containing mpeg4 decoding of which is much better, than theora,
optimized in mplayer. Relative difference is in line with previous
results.

-- 
mailto:av1474 at comtv.ru