[x264-devel] [PATCH] AltiVec implementation of hadamard_ac routines

[Guillaume POIRIER]

Takashi-san,

On Mon, Feb 2, 2009 at 4:51 PM, Takashi Mochizuki
<mochi at da2.so-net.ne.jp> wrote:
> Guillaunme-san,
>
> Oops, an issue on my energy saver settings found; was set as "auto";
> Now adjust as "max".

That shouldn't matter, should it? Auto reduces frequency and voltage
of the processor only if it's not used. If you use it, it's set it to
max. :)


>> Please use the newer version attached to this mail, since it fixes
>> alignment issues that previous patch had. It may be faster than the
>> previous one you tried... In any case, it doesn't suffer from know
>> bugs so you should consider using it instead.
>
> I have tested your new patch, now with newer r1097.
> It seems to be quite cool, aggressive performance gain.
>
> Using same test as previous post,
> - new hadamard_ac patch shows about 5% gain overall.
> - around 2 hours, multiple tests, without any problem.

Not bad!


> Encode statistics show exactly same values.
> Also I just checked only each frame's byte lengthes, but they seem to
> be same.

Well, if you compute the md5sum of both encodes, they should be the
same if you used the same encoding settings and the same passlogfile.

Thanks for your tests.

Guillaume
-- 
Only a very small fraction of our DNA does anything; the rest is all
comments and ifdefs.

Gilda Radner  - "Adopted kids are such a pain - you have to teach them
how to look like you."