is closer at the top.
50-130 cpu cycles faster depending on which kind of direct MBs are meassured.

Originally committed as revision 21788 to svn://svn.ffmpeg.org/ffmpeg/trunk

A little  faster and needed for future optimizations.
This sadly leads to some code duplication (which i hope i can factor out
again after the optimizations on the direcr mv code are done)

Originally committed as revision 21787 to svn://svn.ffmpeg.org/ffmpeg/trunk

Patch by Ramiro

Originally committed as revision 21786 to svn://svn.ffmpeg.org/ffmpeg/trunk

a little less than 200 cpu cycles faster with the cathedral sample.

Originally committed as revision 21785 to svn://svn.ffmpeg.org/ffmpeg/trunk

and is a bit faster (5-10 cpu cycles depending on what is meassured).

Originally committed as revision 21784 to svn://svn.ffmpeg.org/ffmpeg/trunk

Fixes issue1722

Originally committed as revision 21783 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21782 to svn://svn.ffmpeg.org/ffmpeg/trunk

3.6% faster on Elephants_Dream_HD-q7-aq7.ogg on my penryn

Originally committed as revision 21781 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21780 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21779 to svn://svn.ffmpeg.org/ffmpeg/trunk

3% faster on Elephants_Dream_HD-q7-aq7.ogg on my penryn

Originally committed as revision 21778 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21777 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21776 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21775 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21774 to svn://svn.ffmpeg.org/ffmpeg/trunk

Add this flag to img2 (fixes -vcodec copy to image2 in some cases)

Originally committed as revision 21773 to svn://svn.ffmpeg.org/ffmpeg/trunk

Fixes issue1585

Originally committed as revision 21772 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21771 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21770 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21769 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21768 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21767 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21766 to svn://svn.ffmpeg.org/ffmpeg/trunk

Spotted by Ramiro.

Originally committed as revision 21765 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21764 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21763 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21762 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21761 to svn://svn.ffmpeg.org/ffmpeg/trunk

Patch by Wolfram Gloger, wmglo A dent D med D uni-muenchen D de

Originally committed as revision 21760 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21756 to svn://svn.ffmpeg.org/ffmpeg/trunk

that output may be written as uint32_t since output buffer may not be aligned
(and it's silly to force alignment on it) and it does not work in that case
properly on some architectures.

Originally committed as revision 21754 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21753 to svn://svn.ffmpeg.org/ffmpeg/trunk

complain about missing function prototype.

Originally committed as revision 21752 to svn://svn.ffmpeg.org/ffmpeg/trunk

conversion and other information exposed by the pixdesc API.

Originally committed as revision 21751 to svn://svn.ffmpeg.org/ffmpeg/trunk

GCC is sometimes able to optimise constant calls to these functions,
incorrectly indicating that they exist.  Unoptimised calls will then
fail to link.

Originally committed as revision 21749 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21748 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21747 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21746 to svn://svn.ffmpeg.org/ffmpeg/trunk

floating point.
Should fix different results between PPC and x86 for the idroq-video-encode
FATE test.

Originally committed as revision 21745 to svn://svn.ffmpeg.org/ffmpeg/trunk

calculations in the later one are not correct with large time stamps.

Originally committed as revision 21744 to svn://svn.ffmpeg.org/ffmpeg/trunk

Patch by Timo Teräs (timo DOT teras AT iki DOT fi)

Originally committed as revision 21743 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21742 to svn://svn.ffmpeg.org/ffmpeg/trunk

not supported in FFmpeg. This will cause crashes later because the samplerate
is used to initialize the timebase.

Originally committed as revision 21741 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21740 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21739 to svn://svn.ffmpeg.org/ffmpeg/trunk

fixes issue1729

Originally committed as revision 21738 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 21737 to svn://svn.ffmpeg.org/ffmpeg/trunk

SVQ1 chroma has been special-cased aligned to 16-bytes since at least r15466
Other architectures also assume 16-byte alignment here too but set STRIDE_ALIGN
to 16.

Originally committed as revision 21736 to svn://svn.ffmpeg.org/ffmpeg/trunk

This fixes building on some broken systems.

Originally committed as revision 21735 to svn://svn.ffmpeg.org/ffmpeg/trunk

These are often, contrary to standards, implemented only as macros
or compiler-builtin functions without an actual symbol definition.

Originally committed as revision 21734 to svn://svn.ffmpeg.org/ffmpeg/trunk