YUV2 -> RGB BGR for 565, 555 and 888 a.k.a. 24bit color.
Speed-up compared to C version compiled with -O3  187.28%
Patch by Marc Hoffman %mmh A pleasantst P com%
Original thread:
Date: May 9, 2007 2:46 AM
Subject: [FFmpeg-devel] PATCH BlackFin yuv2rgb color space conversion

Originally committed as revision 23307 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

Originally committed as revision 9018 to svn://svn.ffmpeg.org/ffmpeg/trunk

unrolls some loops, utilizing all 8 xmm registers. fft-test
shows ~10% speed up in (I)FFT and ~8% speed up in (I)MDCT on Dothan

Originally committed as revision 9017 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9016 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9015 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9014 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9013 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9012 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9011 to svn://svn.ffmpeg.org/ffmpeg/trunk

Original patch by Jason Millard   jsm174 _at_ gmail
Date: On Fri, 11 May 2007 11:14:01 -0400
Subject: [FFmpeg-devel] enable/disable bitstream filters?

Originally committed as revision 9010 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9009 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9008 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9007 to svn://svn.ffmpeg.org/ffmpeg/trunk

Based on patch by Jason Askew, jason dot askew at gmail dot com.

Originally committed as revision 9006 to svn://svn.ffmpeg.org/ffmpeg/trunk

present in the #included header file dsputil.h.

Originally committed as revision 9005 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9004 to svn://svn.ffmpeg.org/ffmpeg/trunk

core2: mmx2=154 sse2=73 ssse3=66 (cycles)
k8: mmx2=179 sse2=149
p4: mmx2=284 sse2=194

Originally committed as revision 9003 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 9002 to svn://svn.ffmpeg.org/ffmpeg/trunk

sum_abs_dctelem_* alone:
core2: c=186 mmx2=39 sse2=21 ssse3=13 (cycles)
k8: c=163 mmx2=33 sse2=31
p4: c=370 mmx2=60 sse2=60
 dct_sad including sum_abs_dctelem_*:
core2: c=405 mmx2=258 sse2=240 ssse3=232
k8: c=624 mmx2=394 sse2=392
p4: c=849 mmx2=556 sse2=556

Originally committed as revision 9001 to svn://svn.ffmpeg.org/ffmpeg/trunk

core2: before mmx2=193 cycles. after mmx2=174 sse2=122 ssse3=115 (cycles).
k8: before mmx2=205. after mmx2=184 sse2=180.
p4: before mmx2=342. after mmx2=314 sse2=309.

Originally committed as revision 9000 to svn://svn.ffmpeg.org/ffmpeg/trunk

EMU_EDGE broke memory alignment when the desired alignment is >32 bytes.

Originally committed as revision 8999 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8998 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 23296 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale

Originally committed as revision 8997 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8996 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8995 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8994 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8993 to svn://svn.ffmpeg.org/ffmpeg/trunk

16% faster on core2, 5% faster on p4. 10% slower (and thus disabled) on k8.

Originally committed as revision 8992 to svn://svn.ffmpeg.org/ffmpeg/trunk

40% faster on core2, 18% faster on k8, 5% faster on p4.

Originally committed as revision 8991 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8990 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8989 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8988 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8987 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8986 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8985 to svn://svn.ffmpeg.org/ffmpeg/trunk

6% faster on core2 and k8, no change on p4.

Originally committed as revision 8984 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8983 to svn://svn.ffmpeg.org/ffmpeg/trunk

Patch by Jindrich Makovicka, makovick at gmail dot com
Reference thread:
Subject: Re: [PATCH] make libavcodec use bytestream functions
Date: Thu, 10 May 2007 07:49:43 +0200

Originally committed as revision 8982 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8981 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8980 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8979 to svn://svn.ffmpeg.org/ffmpeg/trunk

ff_split_xiph_headers() are there.

Originally committed as revision 8978 to svn://svn.ffmpeg.org/ffmpeg/trunk

handle the AVERROR_NOENT error case in print_error

Originally committed as revision 8977 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8976 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8975 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8974 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8973 to svn://svn.ffmpeg.org/ffmpeg/trunk

Originally committed as revision 8972 to svn://svn.ffmpeg.org/ffmpeg/trunk

Patch by Limin Wang, edited by me

Originally committed as revision 8971 to svn://svn.ffmpeg.org/ffmpeg/trunk