The script updated OMKized Makefiles.

Result of run of ./ffmpeg-omkize.sh .
This can be compiled after
./configure4omk
by OMK build system.

Added copy of libswscale/Makefile for omkization.

Added copies of makefiles for OMK build.

Added script to configure FFMpeg for OMK build.

Added FFmpeg omkize script

Add Hurd to OS list and disable dv1394 in the Hurd case.
patch by Andres Mejia, mcitadel gmail com

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18938 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

honor restart interval in mjpeg, fix #861, SpectralFan.mov still decodes correctly

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18937 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mlp: Re-enable x86-optimized DSP functions.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18936 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mlp: Use LABEL_MANGLE() to export label symbols from inside asm block.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18935 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add LABEL_MANGLE() to export label symbols from inside asm blocks.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18934 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

CorePNG can be decoded since 18421

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18933 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

move DEBUG define before include to get dprintf and change printf to dprintf

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18932 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix unaligned dsputil call.

Should fix FATE corepng test on Solaris/Sparc.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18930 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fix get_str16_nolen with odd len, fix #1065

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18929 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

move DEBUG define before include to get dprintf and move guidcmp before print_guid, fix compilation with DEBUG

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18928 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

eatqi: move "block" variable into context to ensure sufficient alignment for
idct_put for compilers/architectures that can not align stack variables that much.
This is also consistent with similar code in eatgq.c

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18927 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Distinguish between non-coded blocks and decode errors in rtjpeg's get_block

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18926 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add a few size checks when decoding rtjpeg blocks.
Might avoid crashes in unlikely cases, but mostly avoids ugly artefacts
for partial frames.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18925 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Move "block" variable in rtjpeg decoder to context to ensure sufficient alignment
for idct_put on architectures/compilers that have problems aligning on the stack.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18924 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

H264: Don't deblock if alpha or beta is 0.
This ensures that the MMX loop filter is always bitexact with the C version.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18923 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Remove duplicated RGB_TO_[YUV]_CCIR macros, directly use those defined
in libavcodec/colorspace.h.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18922 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

indent

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18921 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix lowres with rv20.
Fixes issue1102.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18920 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix passing of unaligned addresses to dsputil diff_bytes() by
changeing calls to sub_left_prediction such that
the source frame data pointers are no longer offset to skip the initial
elements. This keeps them properly aligned for the subsequent call to diff_bytes.
Changes encoding offset to compensate.

Mailing list thread:
https://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2009-May/069952.html

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18919 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add offset argument to encode_422_bitstream() to control where
in the source data array to start encoding.
Change all calls to pass zero for this new argument for equivalent behavior.

This is a restructure to support a forthcoming bug fix.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18918 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

ARM: actually add VP3 loop filter

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18917 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

ARM: NEON VP3 Loop Filter

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18916 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix regression tests, diff was due to a patch to mpeg-2 parser timestamp fetching, pending,
current ref is technically more correct though.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18912 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fix hdlr descr in mp4 which use c string, patch by Larbi Joubala, larbi dot joubala at resonate-mp4 dot com

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18911 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

simplify pes timestamps fetching, based on a patch by corp186, chasedouglas at gmail dot com

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18910 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

compute total bitrate more accurately, patch by corp186, chasedouglas at gmail dot com

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18909 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fix qtrle encoding when previous frame linesize differs, fix #998

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18908 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

pad last audio frame when encoding

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18907 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

remove useless include

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18906 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Set progressive_sequence before MPV_common_init which cares about it when
setting mb_height for interlaced mpeg-2 encoding.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18905 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Revert part of r18903:
MLP DSP functions x86-optimized.
12.59% overall speedup in x86_32
 9.98% overall speedup in x86_64
compared to gcc 4.3.3

It breaks on x86_64 with PIC because the MANGLE macro appends (%%rip), which
is not wanted for labels.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18904 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

MLP DSP functions x86-optimized.
12.59% overall speedup in x86_32
 9.98% overall speedup in x86_64
compared to gcc 4.3.3

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18903 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fix mouse seeking when start time is not set, fix #1083

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18902 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Remove duplicated RGB_TO_[YUV] macros, use those included in
libavcodec/colorspace.h.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18901 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix grammar for avfilter_graph_parse() doxy.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18900 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix libtheora encoding for non-mod16 sizes

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18899 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Move ALIGN macro to libavutil/common.h and use it in various places

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18898 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

fail if stream fifo could not be allocated

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18897 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Move eatgq blocks array from the stack to the codec context and make sure
it is aligned as necessary for DSPContext's idct_put.
Just aligning it on the stack would have been possible but less reliable
and without any real benefit.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18896 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Set speex frame_size in ogg demuxer

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18895 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: Fix some typos.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18894 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Print out more verbose licensing information.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18893 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

check av_fifo_alloc return

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18892 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

correctly check for audio streams in mpeg ps muxer, fix multiple audio tracks

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18891 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

10l: typing mistake in r18888

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18890 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

eac3dec: fix bug in GAQ dequantization

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18889 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

eac3dec: make GAQ dequantization 24-bit

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18888 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

eac3dec: revert commit r18860.  keep the AHT IDCT 24-bit.  will make AHT GAQ
dequantization 24-bit in a separate commit.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18887 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mathops: sort arch-specific #include lines

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18886 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mathops: whitespace cosmetics

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18885 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Create shared libs correctly, fix mingw build

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18884 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add QCP demuxer.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18883 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Mention LGPL AC-3 decoder in the changelog.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18882 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix building from within subdirs

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18881 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Use IF1 frame sizes in AMR-WB.
Patch by Martin Storsjö: martin (remove that) at martin (oh and this too) st

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18880 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Use iMDCT output scaling to simplify ffvorbis and make it slightly faster.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18879 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix for a problem with inverted sign of output data from ffvorbis decoder.

Now the sign of audio samples in ffvorbis output is the same as in original
uncompressed audio file and this also allows the use of tiny_psnr to compare
ffvorbis with libvorbis/tremor.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18878 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Support for testing (i)MDCT output scale factor in fft-test tool.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18877 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

The AC-3 decoder is now LGPL. Exchange the license header and change references
in the build system and documentation that mark it as GPL.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18876 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

LGPL version of ac3_decode_transform_coeffs_ch, ~12.4% faster.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18875 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Define ARCH_MIPS64 when building for MIPS64

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18874 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Define HAVE_BIGENDIAN as 0 or 1 according to CPU type

This should replace the inconsistently named and defined
WORDS_BIGENDIAN.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18873 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

configure: add some MIPS CPUs

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18872 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add missing libavformat dependencies on mpeg4audio.o

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18871 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Move AMR-WB-specific code to the AMR-WB section of the file.
Fixes compilation when AMR-NB #includes are not available on the system.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18870 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: Move AMR-WB-specific code to the AMR-WB part of the file.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18869 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Log with level AV_LOG_ERROR if av_parse_color() cannot recognize a
color.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18868 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Remove useless casts of vec_ld() parameters.
Gets rid of 'cast discards qualifiers from pointer target type' warnings.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18867 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Indent

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18866 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Change from INIT_VLC_USE_STATIC to INIT_VLC_USE_NEW_STATIC in atrac3

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18865 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Indent

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18864 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Switch from INIT_VLC_USE_STATIC to INIT_VLC_USE_NEW_STATIC in qdm2.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18863 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Make TMV decoder check packet sizes before reading from it.

FATE test 295 might need updating.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18862 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add a forward declaration of mpeg1_decode_block_intra to fix compilation
broken by r18859 for e.g. all gcc 4 compilers.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18861 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

eac3dec: use 16-bit pre-mantissas instead of 24-bit in AHT decoding. it is
simpler and also fixes a bug in GAQ dequantization.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18860 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mpeg12: fix non-standard use of "extern inline"

Rename ff_mpeg1_decode_block_intra to mpeg1_decode_block_intra and
make it static.  Add ff_mpeg1_decode_block_intra as plain extern
function calling mpeg1_decode_block_intra.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18859 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix the M bit for multi-packet NALs.
Patch by Yao Peter (peter AT yuvad DOT com)

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18858 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

ac3dec: fix coupling range check.  the start subband must be less than
the end subband.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18857 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

oma: fix build if memcmp() is a macro

Any C library function may be a macro, so compound literals
passed to memcmp() must be surrounded by parens to avoid being
split on commas.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18856 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Support for getting (i)MDCT output multiplied by a constant scaling factor.

Scaling (i)MDCT output has no runtime overhead and can be used to improve
performance of audio codecs. All the changes are only needed in
'ff_mdct_init' function and slow down initialization a bit.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18855 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Move img_convert_ctx to VideoState and free it in stream_close().

Patch by Martin Storsjö, martin martin st

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18854 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add missing frees to stream_close() and do_exit().

Patch by Martin Storsjö, martin martin st

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18853 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Fix off-by-one error in MS RLE decoder which may result into writing past
picture buffer.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18852 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

move increment after debug print

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18851 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: K&R coding style, indent correctly, break overly long lines

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18850 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Replace 'x == 0' if conditions with '!x'.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18849 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: Remove pointless {}.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18848 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Surround AMR_WB encoding code by appropriate #ifdefs.
This will help support the opencore AMR library.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18847 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: Move some more code around to avoid #ifdeffery.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18846 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

cosmetics: Move functions around so that encoding and decoding functions are
grouped together. This will save some #ifdefs.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18845 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Remove redundant #if condition. Two CONFIG_LIBAMR_NB blocks were right after each other,
so that the second check was pointless.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18844 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Add av_cold attribute to amr_nb decoding/encoding initialization functions.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18843 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

Bugfix for bug introduced in the multichannel cook patchset.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18842 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mlp: Simplify adressing of state and coeffs arrays for both filters by making
the arrays sequential.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18841 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b

mlp: include mlp.h and use MAX_CHANNELS instead of hardcoding value 8.

git-svn-id: file:///var/local/repositories/ffmpeg/trunk@18840 9553f0bf-9b14-0410-a0b8-cfaf0461ba5b