Commit Graph

169 Commits

Author SHA1 Message Date
Michael Niedermayer
33a67bd6c7 Fix overflow.
Originally committed as revision 27573 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 20:47:16 +00:00
Michael Niedermayer
ff9a056d1a Fix build failure due to %%eip on x86_64.
Originally committed as revision 27569 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 17:50:49 +00:00
Michael Niedermayer
6b79dbce0c Change RGB2YUV_SHIFT from 16 to 15 to make it able to work
with 16bit signed constants (like SIMD might use).

Originally committed as revision 27568 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 16:39:56 +00:00
Michael Niedermayer
ba83d86299 Add bgr/rgb15/16/32->UV-half to the macro so theres less code duplication
at the source level.

Originally committed as revision 27567 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 13:55:51 +00:00
Michael Niedermayer
a0baa07aef Factorize RGB/BGR15/16/32->UV by using the preprocessor.
Originally committed as revision 27566 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 03:56:06 +00:00
Michael Niedermayer
214892ee42 Factorize rgb/bgr15/16/32->Y by using the preprocessor.
Originally committed as revision 27565 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 03:42:45 +00:00
Michael Niedermayer
2f60f6297c Make SWS_FULL_CHR_H_INP work.
Originally committed as revision 27564 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-10 02:28:38 +00:00
Michael Niedermayer
dfb09bd1d4 Rewrite bgr24->yuv mmx code, the new code is cleaner, more accurate,
and does not throw half the chroma away.

Originally committed as revision 27561 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-09 23:30:06 +00:00
Michael Niedermayer
35ab2b6445 More correct rounding for the rgb/bgr->yuv converters.
Originally committed as revision 27557 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-09 16:55:58 +00:00
Michael Niedermayer
aa13b0fc55 Prevent overflows during mpeg->jpeg yuv.
Originally committed as revision 27552 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-09 03:35:07 +00:00
Michael Niedermayer
6bc0c7928a Fix jpeg yuv.
Fixes issue504.

Originally committed as revision 27547 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-08 12:36:38 +00:00
Michael Niedermayer
1625216eaa Fix accurate rounding mode on x86_64.
Fixes issue222.

Originally committed as revision 27545 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-07 21:06:21 +00:00
Luca Barbato
97b93389af Make ToY and ToUV family of function consistent part I
Convert width argument from int to long (note: srcW is still an int).

Originally committed as revision 27544 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-07 18:22:29 +00:00
Michael Niedermayer
b0880d5da4 Make 16bit grayscale output work.
Originally committed as revision 27524 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-04 22:58:16 +00:00
Michael Niedermayer
9990e4269c Support PIX_FMT_RGB32_1 and PIX_FMT_BGR32_1.
Fixes issue248.

Originally committed as revision 27522 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-04 21:59:15 +00:00
Michael Niedermayer
27a90b0450 Fix rgb15/16 vs. bgr part2.
Originally committed as revision 27517 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-09-04 18:24:25 +00:00
Michael Niedermayer
a1f3ffa3ce Make C code in yuv2yuv1() do accurate rounding, this could be split
depending on SWS_ACCURATE as well if someone wants.

Originally committed as revision 27323 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-18 00:09:09 +00:00
Michael Niedermayer
d78c1ea16f indent
Originally committed as revision 27322 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-17 23:43:48 +00:00
Michael Niedermayer
bf2bdde67a Forgotten accurate rounding function YSCALEYUV2YV121_ACCURATE.
Originally committed as revision 27321 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-17 23:42:06 +00:00
Michael Niedermayer
7bae01c62c simplify yuv2yuv1()
Originally committed as revision 27320 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-17 23:35:36 +00:00
Luca Barbato
fcc402b1c9 Remove ASSERT() macro. SoC Patch from Keiji Costantini
Originally committed as revision 27214 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-06 03:32:47 +00:00
Diego Biurrun
8a3227968c spelling/grammar/wording overhaul
Originally committed as revision 27190 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-07-04 13:49:45 +00:00
Baptiste Coudurier
87cf861c7d cast to correct type, suppress warnings
Originally committed as revision 26946 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-06-01 03:28:15 +00:00
Diego Biurrun
4cf16bbed5 cosmetics: comment typo fixes
Originally committed as revision 26269 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-03-22 17:35:32 +00:00
Baptiste Coudurier
4b0c30b79c remove redundant SwScaler text since av_log uses AVClass context
Originally committed as revision 26182 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-03-06 09:51:25 +00:00
Michael Niedermayer
8b2fce0d3f Turn ancient V offset numerical constants into named ones.
Add a check that checks that the width is within the choosen constant.
This might have been exploitable.

Originally committed as revision 26165 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-03-05 03:00:25 +00:00
Reimar Döffinger
f018bc1000 __asm __volatile -> asm volatile part 3
Originally committed as revision 26091 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-02-24 14:59:16 +00:00
Carl Eugen Hoyos
0cb2559445 Discard two symbols from libswscale.
Originally committed as revision 26060 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-02-22 08:08:19 +00:00
Reimar Döffinger
5802683a97 Make some assembler constants global instead of declaring them multiple times.
Originally committed as revision 25910 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-01-29 14:58:10 +00:00
Benoit Fouet
1b0a4572f4 Change (a == NULL) condition to (!a) and (a != NULL) condition to (a).
Originally committed as revision 25780 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-01-17 15:49:31 +00:00
Benoit Fouet
e5091488a5 Remove non cosmetic spaces inside parentheses.
Originally committed as revision 25777 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2008-01-17 08:57:15 +00:00
Diego Biurrun
bc279024c2 Comment some #endif directives.
Originally committed as revision 25211 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-11-30 13:21:20 +00:00
Diego Biurrun
a680708d2d The function names of [rgb|bgr]1[56]to[UV|Y] had rgb<->bgr flipped.
Rename them to match the actual implementation. Fixes issue 162.

Originally committed as revision 24965 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-11-04 12:34:46 +00:00
Diego Biurrun
b19bcbaa78 license header consistency cosmetics
Originally committed as revision 23722 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-07-05 10:18:58 +00:00
Diego Biurrun
ddda6fcddc comment grammar fixes
Originally committed as revision 23685 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-06-28 11:18:16 +00:00
Ivo van Poorten
fa88429489 remove the need of one extra shift in rgb16ToUV
Originally committed as revision 23281 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-05-09 21:34:05 +00:00
Ivo van Poorten
f96829d2d0 Fix rgb15ToUV. Correct order of components and use one shift less.
Originally committed as revision 23279 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-05-09 17:29:20 +00:00
Diego Biurrun
2da0d70d5e cosmetics attack, part IV: Remove all tabs and prettyprint/reindent the code.
Originally committed as revision 23179 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-04-30 08:05:29 +00:00
Diego Biurrun
6a4970abd1 cosmetics: Remove trailing whitespace.
Originally committed as revision 23147 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-04-26 23:07:11 +00:00
Michael Niedermayer
fa65e2f63a improve horizontal chroma resolution with pal
Originally committed as revision 22994 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-04-15 01:27:34 +00:00
Michael Niedermayer
21c08a3f1c convert palette to yuv instead of converting each pixel after pal->rgb
Originally committed as revision 22993 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-04-15 00:53:32 +00:00
Diego Biurrun
911406f2c4 typo
Originally committed as revision 22932 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-04-07 14:07:51 +00:00
Reimar Döffinger
9340a99588 Fix a possible crash on 64 bit systems when the lumSrcPtr or chrSrcPtr
does not fit in 32 bits. The SWS_ACCURATE_RND is still broken though.

Originally committed as revision 22401 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-03-01 19:48:40 +00:00
Reimar Döffinger
adcec46a83 Add av_ prefix to clip functions
Originally committed as revision 22339 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-02-25 10:27:43 +00:00
Michael Niedermayer
18064f5cf3 BGR/RGB4 byte formats as input
fixing isRGB/BGR() for the byte formats

Originally committed as revision 22244 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-02-17 11:36:02 +00:00
Michael Niedermayer
e28630fc70 pal8 input
this is not optimized or anything (that is easy to add just use the existing unscaled converters if possible ...)

Originally committed as revision 22236 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-02-16 21:17:15 +00:00
Luca Abeni
2d529db50f Pass a context to av_log(), when possible
Originally committed as revision 21999 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2007-01-23 13:41:45 +00:00
Luca Abeni
b2afecbe02 change all the occurrences of "FFMIN(FFMAX())" to clip_uint8() or clip()
Originally committed as revision 21776 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-12-26 18:08:40 +00:00
Luca Abeni
7d19628b14 Add some explicit casts to avoid some warnings
Originally committed as revision 21772 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-12-26 16:21:13 +00:00
Luca Abeni
0683a5c558 Fix compilation of non-MMX code with gcc 2.95
Originally committed as revision 21702 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-12-20 15:58:15 +00:00
Reimar Döffinger
e96da13bc7 Change "p" asm constraints to "g", since "p" was a no longer necessary hack to
make AMD64 compilation work and ICC can not handle "p".

Originally committed as revision 21341 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-27 21:59:13 +00:00
Luca Abeni
807e0c6645 Clarify that some of the non-SIMD code is now LGPLed.
All the authours of such code agreed.

Originally committed as revision 21045 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-19 10:38:12 +00:00
Michael Niedermayer
6264a515d0 forgotten 2 converters (yeah svn up, svn di svn ci isnt enough i should actually look at the code after svn up not just the diff ...)
Originally committed as revision 20962 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-16 14:49:04 +00:00
Michael Niedermayer
c2271987f8 fixing the lamest bug in swscale, all the rgb/bgr->* code did 2x2 downsampling for chroma, it should just be 2x1 (the rest of the code also belived its 2x1 ...)
Originally committed as revision 20961 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-16 14:44:00 +00:00
Nigel Pearson
d904b5fc66 Fix MacIntel build: "/nop" is illegal on Apple's older version of GAS
Patch by Nigel Pearson % nigel A ind P tansu P com P au %
Original thread:
Date: Nov 6, 2006 10:43 PM
Subject: [Ffmpeg-devel] [PATCH] swscale uses /nop, illegal for old versions of GAS

Originally committed as revision 20740 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-07 07:07:05 +00:00
Luca Abeni
a43fb6b37e Add support for conversions from the rgb565 and rgb555 formats
Originally committed as revision 20605 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-02 09:01:01 +00:00
Diego Biurrun
3d6a30d948 Use common define for x86_32 and x86_64.
Originally committed as revision 20592 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-11-01 18:36:20 +00:00
Kostya Shishkov
4884b9e50d 16-bit grayscale support
Originally committed as revision 20427 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-10-24 05:02:33 +00:00
Diego Biurrun
d026b45e73 Change license headers to say 'FFmpeg' instead of 'this program'.
Originally committed as revision 20100 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-10-07 15:33:14 +00:00
Diego Biurrun
f8d6112850 Fix unused variable warning when compiling with MMX disabled.
Originally committed as revision 20021 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-10-03 17:11:43 +00:00
Luca Abeni
e9e12f0e11 Remove the dependency of libswscale on img_format.h
Originally committed as revision 19878 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-09-17 15:15:13 +00:00
Michael Niedermayer
d1a7ae1a37 make the C code of the swscaler which i wrote LGPL
Originally committed as revision 19600 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-08-30 10:50:20 +00:00
Uoti Urpala
91d0bda26d Fix compilation with -no-PIC and without -fomit-frame-pointer (used by
--enable-debug).
Fix from Jason Tackaberry for his previous -fPIC patch.

Originally committed as revision 19406 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-08-15 14:50:24 +00:00
Jason Tackaberry
83c89c7885 -fPIC support for libswscale
patch by Jason Tackaberry, tack /at/ urandom /dot/ ca

Originally committed as revision 19402 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-08-15 08:47:03 +00:00
Diego Biurrun
4bff9ef9d0 Replace asmalign.h hack by ASMALIGN cpp macros from config.h.
Originally committed as revision 19378 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-08-13 00:21:14 +00:00
Luca Abeni
6f209cb549 Replace MIN() and MAX() with FFMIN() and FFMAX()
Originally committed as revision 19186 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-07-26 13:19:02 +00:00
Michael Niedermayer
8422aa888b a tiny bit of cleanup
Originally committed as revision 19178 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-07-24 10:48:54 +00:00
Michael Niedermayer
bca11e75fb vertical scaler with accurate rounding, some people on doom9 can see +-1 errors
the +-1 issue is limited to >2tap vertical filters, so bilinear upscale was unaffected
the new code is sometime faster sometimes slower but the difference is significant (~20%) so its optional and enabled with arnd=1

Originally committed as revision 19177 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-07-24 10:36:06 +00:00
Luca Abeni
73435f006b Move postproc ---> libswscale
Originally committed as revision 18866 to svn://svn.mplayerhq.hu/mplayer/trunk/libswscale
2006-06-30 12:00:31 +00:00