platform/upstream/libav.git
13 years agoconfigure: use same CPPFLAGS in kFreeBSD as Linux
Reinhard Tartler [Sat, 14 May 2011 14:10:30 +0000 (16:10 +0200)]
configure: use same CPPFLAGS in kFreeBSD as Linux

046f081b46c8479820409cf8f530b988221bd15b reorganized the CPPFLAGS to no
longer add -D_POSIX_C_SOURCE unconditionally, but only on systems (e.g.,
glibc based ones) that require it.  As kFreeBSD uses glibc, it needs to
be treated similar.

Additionally, _BSD_SOURCE is turned on to enable some additional types
such as caddr_t, which are normally enabled on BSD but not with glibc.

13 years agoFix FSF address copy paste error in some license headers.
Diego Biurrun [Sat, 14 May 2011 19:32:31 +0000 (21:32 +0200)]
Fix FSF address copy paste error in some license headers.

13 years agoAdd an aac sample which uses LTP to fate-aac.
Justin Ruggles [Tue, 26 Apr 2011 22:19:52 +0000 (18:19 -0400)]
Add an aac sample which uses LTP to fate-aac.

13 years ago[PATCH] Update pixdesc_be fate refs after adding 9/10bit YUV420P formats.
Ronald S. Bultje [Fri, 13 May 2011 23:01:50 +0000 (19:01 -0400)]
[PATCH] Update pixdesc_be fate refs after adding 9/10bit YUV420P formats.

Also remove code that overwrites the C versions of functions in
sws_init_swScale_altivec(), so that it uses the C functions of files
if no altivec-optimized version exists.

13 years agoarm: properly mark external symbol call
Martin Aumüller [Sat, 14 May 2011 08:10:27 +0000 (10:10 +0200)]
arm: properly mark external symbol call

Surround memset and ff_vp8_dct_cat_prob by X() in order to fix iOS build

Includes patch by Luca Barbato <lu_zero@gentoo.org>.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
13 years agoswscale: properly inline bits/endianness in yuv2yuvX16inC().
Ronald S. Bultje [Fri, 13 May 2011 14:28:49 +0000 (10:28 -0400)]
swscale: properly inline bits/endianness in yuv2yuvX16inC().

13 years agoswscale: fix clipping of 9/10bit YUV420P.
Ronald S. Bultje [Fri, 13 May 2011 20:45:04 +0000 (16:45 -0400)]
swscale: fix clipping of 9/10bit YUV420P.

13 years agoAdd av_clip_uintp2() function
Mans Rullgard [Fri, 13 May 2011 15:39:17 +0000 (16:39 +0100)]
Add av_clip_uintp2() function

Signed-off-by: Mans Rullgard <mans@mansr.com>
13 years agodfa: fix buffer overflow checks to avoid integer overflows.
Reimar Döffinger [Tue, 29 Mar 2011 19:14:55 +0000 (21:14 +0200)]
dfa: fix buffer overflow checks to avoid integer overflows.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agomovenc: always write esds descriptor length using 4 bytes.
Baptiste Coudurier [Wed, 30 Mar 2011 21:08:16 +0000 (14:08 -0700)]
movenc: always write esds descriptor length using 4 bytes.

ipod shuffle doesn't support anything else.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: use parse_number_and_die() when it makes sense
Stefano Sabatini [Mon, 11 Apr 2011 11:16:07 +0000 (13:16 +0200)]
ffmpeg: use parse_number_and_die() when it makes sense

Prefer parse_number_or_die() over atoi()/atol() parsing for the options:
-pass, -top, -vc, and -qscale.

Improve input validation.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: get rid of the 'q' key schizofrenia
Anton Khirnov [Thu, 12 May 2011 13:15:07 +0000 (15:15 +0200)]
ffmpeg: get rid of the 'q' key schizofrenia

SIGINT for quitting should be enough for everybody.

13 years agortpdec_qdm2: Don't try to parse data packet if no configuration is received
Martin Storsjö [Thu, 12 May 2011 19:47:53 +0000 (22:47 +0300)]
rtpdec_qdm2: Don't try to parse data packet if no configuration is received

The later parsing of payload data depends on the configuration
being present. If it hasn't been configured properly yet,
parsing a data packet may lead to a crash.

Signed-off-by: Martin Storsjö <martin@martin.st>
13 years agoac3enc: put the counting of stereo rematrixing bits in the same place to
Justin Ruggles [Sat, 30 Apr 2011 19:55:18 +0000 (15:55 -0400)]
ac3enc: put the counting of stereo rematrixing bits in the same place to
make the code easier to understand.

13 years agoac3enc: clean up count_frame_bits() and count_frame_bits_fixed()
Justin Ruggles [Sat, 30 Apr 2011 19:52:55 +0000 (15:52 -0400)]
ac3enc: clean up count_frame_bits() and count_frame_bits_fixed()

13 years agompegvideo: make FF_DEBUG_DCT_COEFF output coeffs via av_log() instead of just via...
Michael Niedermayer [Wed, 23 Feb 2011 22:38:52 +0000 (23:38 +0100)]
mpegvideo: make FF_DEBUG_DCT_COEFF output coeffs via av_log() instead of just via AVFrame.

This allows the values to be used without changing C code and is closer to how
the other DEBUG flags work.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agosrtdec: make sure we don't write past the end of buffer
Aurelien Jacobs [Fri, 25 Mar 2011 00:02:53 +0000 (01:02 +0100)]
srtdec: make sure we don't write past the end of buffer

Signed-off-by: Aurelien Jacobs <aurel@gnuage.org>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agowmaenc: improve channel count and bitrate error handling in encode_init()
Tomas Härdin [Mon, 21 Mar 2011 09:52:36 +0000 (10:52 +0100)]
wmaenc: improve channel count and bitrate error handling in encode_init()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agomatroskaenc: make sure we don't produce invalid file with no codec ID
Aurelien Jacobs [Thu, 24 Mar 2011 00:56:26 +0000 (01:56 +0100)]
matroskaenc: make sure we don't produce invalid file with no codec ID

Signed-off-by: Aurelien Jacobs <aurel@gnuage.org>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agomatroskadec: check that pointers were initialized before accessing them
Aurelien Jacobs [Wed, 23 Mar 2011 23:28:19 +0000 (00:28 +0100)]
matroskadec: check that pointers were initialized before accessing them

Signed-off-by: Aurelien Jacobs <aurel@gnuage.org>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agolavf: fix function name in compute_pkt_fields2 av_dlog message
Stefano Sabatini [Thu, 17 Mar 2011 11:52:40 +0000 (12:52 +0100)]
lavf: fix function name in compute_pkt_fields2 av_dlog message

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agolavf: fix av_find_best_stream when providing a wanted stream.
Marton Balint [Sun, 20 Feb 2011 00:18:30 +0000 (01:18 +0100)]
lavf: fix av_find_best_stream when providing a wanted stream.

In the main loop, stream_number is incremented after checking the stream type,
so the search usually will not find the wanted stream.

This patch eliminates the useless stream_number variable and introduces a new
one, called real_stream_index to store the real stream index of the current
stream, no matter if we are looping through all the streams or only the streams
of a program.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agolavf: fix av_find_best_stream when decoder_ret is given and using a related stream.
Marton Balint [Sun, 20 Feb 2011 00:18:49 +0000 (01:18 +0100)]
lavf: fix av_find_best_stream when decoder_ret is given and using a related stream.

Yet another fix for the code originally designed for use without related_stream.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: factorize quality calculation
Michael Niedermayer [Wed, 6 Apr 2011 19:17:32 +0000 (21:17 +0200)]
ffmpeg: factorize quality calculation

Calculate quality value once per stream in print_report().
Also fix segfault, as coded_frame can be NULL.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agotiff: add support for SamplesPerPixel tag in tiff_decode_tag()
Stefano Sabatini [Mon, 9 May 2011 10:51:24 +0000 (12:51 +0200)]
tiff: add support for SamplesPerPixel tag in tiff_decode_tag()

Format detection and internal frame initialization is moved to a
separate init_image() function, which is called when all the tags have
been read, and so both BitsPerSample and SamplesPerPixel information
has been collected.

This fixes decoding of the file 11.tiff from roundup issue #1925.

Based on a patch by Kostya Shishkov <kostya.shishkov@gmail.com>.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agotiff: Prefer enum TiffCompr over int for TiffContext.compr.
Stefano Sabatini [Mon, 9 May 2011 19:05:42 +0000 (21:05 +0200)]
tiff: Prefer enum TiffCompr over int for TiffContext.compr.

This is safer and helps debugging.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agomov: Support edit list atom version 1.
Yusuke Nakamura [Wed, 11 May 2011 20:28:27 +0000 (05:28 +0900)]
mov: Support edit list atom version 1.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agoconfigure: Enable libpostproc automatically if GPL code is enabled. v0.7b2
Måns Rullgård [Thu, 12 May 2011 12:21:49 +0000 (13:21 +0100)]
configure: Enable libpostproc automatically if GPL code is enabled.

Enabling libpostproc automatically should give it more compile coverage
and save the FATE box maintainers some configuration hassles.

13 years agoCosmetics: fix prototypes in oggdec
Clément Bœsch [Wed, 11 May 2011 18:43:29 +0000 (20:43 +0200)]
Cosmetics: fix prototypes in oggdec

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agooggdec: fix memleak with continuous streams.
Clément Bœsch [Wed, 11 May 2011 18:43:27 +0000 (20:43 +0200)]
oggdec: fix memleak with continuous streams.

This avoids the creation of a new AVStream instead of replacing it when
a stream reset occurs (track change with some webradios for example).

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agomatroskaenc: add missing new line in av_log() call
Aurelien Jacobs [Thu, 24 Mar 2011 00:54:15 +0000 (01:54 +0100)]
matroskaenc: add missing new line in av_log() call

Signed-off-by: Aurelien Jacobs <aurel@gnuage.org>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agodnxhdenc: add AVClass in private context.
Baptiste Coudurier [Wed, 30 Mar 2011 15:47:30 +0000 (08:47 -0700)]
dnxhdenc: add AVClass in private context.

Fixes private options.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agomovenc: fix yuv range in avid atoms used by dnxhd.
Baptiste Coudurier [Wed, 23 Mar 2011 17:22:05 +0000 (10:22 -0700)]
movenc: fix yuv range in avid atoms used by dnxhd.

yuv range: full 1 / normal 2

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agomovenc: fix adpcm mono muxing.
Alex Converse [Mon, 28 Feb 2011 00:29:21 +0000 (16:29 -0800)]
movenc: fix adpcm mono muxing.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agompegts: do not output known sized packet if an unbounded packet is already queued.
Baptiste Coudurier [Mon, 28 Feb 2011 00:19:50 +0000 (16:19 -0800)]
mpegts: do not output known sized packet if an unbounded packet is already queued.

Fix issue #2624.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffprobe: tweak error message in open_input_file()
Stefano Sabatini [Sun, 13 Mar 2011 20:22:59 +0000 (21:22 +0100)]
ffprobe: tweak error message in open_input_file()

Replace "codec (id=%d)" with "codec with id %d", slightly more
readable.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agopatcheck: warn about assert usage
Peter Ross [Fri, 18 Mar 2011 11:28:52 +0000 (22:28 +1100)]
patcheck: warn about assert usage

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoconfigure: sort filter deps entries
Stefano Sabatini [Fri, 25 Mar 2011 17:49:24 +0000 (18:49 +0100)]
configure: sort filter deps entries

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoReplace int_fast integer types with their sized standard posix counterparts.
Diego Biurrun [Fri, 29 Apr 2011 14:08:20 +0000 (16:08 +0200)]
Replace int_fast integer types with their sized standard posix counterparts.

The _fast integer types provide no realworld benefits, but may introduce
portability issues and are just plain ugly.

13 years agojfdct: Replace sized int_fast integer types with plain int/unsigned.
Diego Biurrun [Fri, 29 Apr 2011 18:17:43 +0000 (20:17 +0200)]
jfdct: Replace sized int_fast integer types with plain int/unsigned.

int/unsigned is the natural memory access type for CPUs, using sized types
for temporary variables, counters and similar just increases code size and
can possibly cause a slowdown.

13 years agoconfigure: Do not unconditionally add -D_POSIX_C_SOURCE to CPPFLAGS.
Diego Biurrun [Wed, 11 May 2011 15:52:51 +0000 (17:52 +0200)]
configure: Do not unconditionally add -D_POSIX_C_SOURCE to CPPFLAGS.

Adding _POSIX_C_SOURCE to CPPFLAGS globally produces all sorts of problems
since it causes certain system functions to be hidden on some (BSD) systems.
The solution is to only add the flag on systems that really require it, i.e.
glibc-based ones.

This change makes BSD systems compile out-of-the-box without the need for
adding specific flags manually.  It also allows dropping a number of flags
set manually on a file-per-file basis, but were only present to work around
breakage introduced by the presence of _POSIX_C_SOURCE.

Also add _XOPEN_SOURCE to CPPFLAGS for glibc systems.  We use XSI extensions
in several places already, so it is preferable to define it globally instead
of littering source files with individual #defines only needed for glibc.

13 years agonsv: Fix misdetection of MP3 files as NSV.
Thierry Foucu [Thu, 12 May 2011 04:27:04 +0000 (21:27 -0700)]
nsv: Fix misdetection of MP3 files as NSV.

If an MP3 file contains the string NSVs, the NSV probe will confuse it for an
NSV file. Check for 0xBEEF after a Video/Audio chunk to achieve more accuracy.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agooggdec: use explicit AVInputFormat struct initializers
Clément Bœsch [Wed, 11 May 2011 18:43:31 +0000 (20:43 +0200)]
oggdec: use explicit AVInputFormat struct initializers

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agooggdec: simplify ogg_probe
Clément Bœsch [Wed, 11 May 2011 18:43:30 +0000 (20:43 +0200)]
oggdec: simplify ogg_probe

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agoswscale: fix YUV420P 9/10bit support.
Ronald S. Bultje [Wed, 11 May 2011 23:15:14 +0000 (19:15 -0400)]
swscale: fix YUV420P 9/10bit support.

Fix handling of input if not in native endianness, and add support for
9/10-bit output. This allows us to force endianness of YUV420P 9/10bit
in the H264/10bit fate tests, which should fix them on big-endian
systems.

13 years ago10-bit H.264 x86 chroma v loopfilter asm
Jason Garrett-Glaser [Wed, 11 May 2011 17:11:55 +0000 (10:11 -0700)]
10-bit H.264 x86 chroma v loopfilter asm

Also delete some unused deblock asm macros.

13 years agoPort SMPTE S302M audio decoder from FFmbc 0.3.
Baptiste Coudurier [Tue, 12 Apr 2011 22:29:09 +0000 (15:29 -0700)]
Port SMPTE S302M audio decoder from FFmbc 0.3.

13 years agoFix crash of interlaced MPEG2 decoding
Anatoly Nenashev [Sun, 27 Mar 2011 19:41:48 +0000 (21:41 +0200)]
Fix crash of interlaced MPEG2 decoding

Problem description, preliminary review discussion at
http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/127731

13 years agoh264pred: fix one more aliasing violation.
Ronald S. Bultje [Wed, 11 May 2011 12:30:02 +0000 (08:30 -0400)]
h264pred: fix one more aliasing violation.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agodoc/APIchanges: fill in missing hashes and dates.
Anton Khirnov [Wed, 11 May 2011 11:56:42 +0000 (13:56 +0200)]
doc/APIchanges: fill in missing hashes and dates.

13 years agoflacenc: use proper initializers for AVOption default values.
Anton Khirnov [Wed, 11 May 2011 11:45:12 +0000 (13:45 +0200)]
flacenc: use proper initializers for AVOption default values.

default_val was recently changes from double to a union, current code
wasn't updated for that.

13 years agolavc: deprecate named constants for deprecated antialias_algo.
Anton Khirnov [Wed, 11 May 2011 06:45:17 +0000 (08:45 +0200)]
lavc: deprecate named constants for deprecated antialias_algo.

13 years agoaac: workaround for compilation on cygwin
Reinhard Tartler [Wed, 11 May 2011 11:51:11 +0000 (13:51 +0200)]
aac: workaround for compilation on cygwin

On cygwin, math.h needs to be included before float.h because of a bug
in the system headers. Including libavutil/libm.h first works around
this issue.

Longer discussion of the topic:
http://thread.gmane.org/gmane.comp.video.ffmpeg.devel/128582

13 years agoswscale: extend YUV422p support to 10bits depth
Baptiste Coudurier [Wed, 11 May 2011 08:00:50 +0000 (10:00 +0200)]
swscale: extend YUV422p support to 10bits depth

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
13 years agotiff: add support for inverted FillOrder for uncompressed data
Stefano Sabatini [Mon, 9 May 2011 19:59:20 +0000 (21:59 +0200)]
tiff: add support for inverted FillOrder for uncompressed data

Fix decoding of file b.tif, trac issue #168.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
13 years agoRemove unused softfloat implementation.
Diego Biurrun [Wed, 11 May 2011 10:22:28 +0000 (12:22 +0200)]
Remove unused softfloat implementation.

The softfloat functionality is unused, not installed and incomplete.
On platforms without floating point units, the compiler provides a softfloat
implementation so there is no point in carrying this code around locally.

13 years agoh264pred: fix aliasing violations.
Ronald S. Bultje [Wed, 11 May 2011 00:23:08 +0000 (20:23 -0400)]
h264pred: fix aliasing violations.

Tested to fix Haiku H264/10bit fate failures, may also fix others.

13 years agorotozoom: Eliminate French variable name.
Diego Biurrun [Mon, 9 May 2011 11:00:04 +0000 (13:00 +0200)]
rotozoom: Eliminate French variable name.

13 years agorotozoom: Check return value of fread().
Diego Biurrun [Mon, 9 May 2011 10:42:15 +0000 (12:42 +0200)]
rotozoom: Check return value of fread().

This fixes the warnings:
tests/rotozoom.c:252: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result
tests/rotozoom.c:254: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result

13 years agorotozoom: Return an error value instead of calling exit().
Diego Biurrun [Mon, 9 May 2011 09:16:55 +0000 (11:16 +0200)]
rotozoom: Return an error value instead of calling exit().

13 years agorotozoom: Make init_demo() return int and check for errors on invocation.
Diego Biurrun [Mon, 9 May 2011 10:33:45 +0000 (12:33 +0200)]
rotozoom: Make init_demo() return int and check for errors on invocation.

13 years agorotozoom: Drop silly UINT8 typedef.
Diego Biurrun [Mon, 9 May 2011 10:26:00 +0000 (12:26 +0200)]
rotozoom: Drop silly UINT8 typedef.

13 years agorotozoom: Drop some unnecessary parentheses.
Diego Biurrun [Mon, 9 May 2011 10:23:55 +0000 (12:23 +0200)]
rotozoom: Drop some unnecessary parentheses.

13 years agorotozoom: K&R coding style cosmetics
Diego Biurrun [Mon, 9 May 2011 09:45:13 +0000 (11:45 +0200)]
rotozoom: K&R coding style cosmetics

13 years agortsp: Only do keepalive using GET_PARAMETER if the server supports it
Martin Storsjö [Mon, 9 May 2011 17:11:16 +0000 (20:11 +0300)]
rtsp: Only do keepalive using GET_PARAMETER if the server supports it

This is more like what VLC does. If the server doesn't mention
supporting GET_PARAMETER in response to an OPTIONS request,
VLC doesn't send any keepalive requests at all. After this patch,
libavformat will still send OPTIONS keepalives if GET_PARAMETER
isn't explicitly said to be supported.

Some RTSP cameras don't support GET_PARAMETER, and will
close the connection if this is sent as keepalive request
(but support OPTIONS just fine, but probably don't need any
keepalive at all). Some other cameras don't support using
OPTIONS as keepalive, but require GET_PARAMETER instead.

Signed-off-by: Martin Storsjö <martin@martin.st>
13 years agoffmpeg: call pre_process_video_frame() only if decoding is needed
Stefano Sabatini [Wed, 20 Apr 2011 11:13:09 +0000 (13:13 +0200)]
ffmpeg: call pre_process_video_frame() only if decoding is needed

In output_packet(), move the pre_process_video_frame() call inside the
if (ist->decoding_needed) { } block. This way
pre_process_video_frame() is not called when stream-copy has been
selected.

Also simplify.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: reformat resample condition code in transcode()
Stefano Sabatini [Sat, 16 Apr 2011 21:18:22 +0000 (23:18 +0200)]
ffmpeg: reformat resample condition code in transcode()

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: factorize resampling condition check in do_video_out()
Stefano Sabatini [Sat, 16 Apr 2011 21:11:01 +0000 (23:11 +0200)]
ffmpeg: factorize resampling condition check in do_video_out()

Simplify and improve readability.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: prefer "dec" over "ist->st->codec" in do_video_out() snippet
Stefano Sabatini [Sat, 16 Apr 2011 23:38:09 +0000 (01:38 +0200)]
ffmpeg: prefer "dec" over "ist->st->codec" in do_video_out() snippet

Simplify, ease readability.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: improve reporting if size/pixel format changes
Stefano Sabatini [Sat, 16 Apr 2011 20:58:13 +0000 (22:58 +0200)]
ffmpeg: improve reporting if size/pixel format changes

Use av_log() rather than fprintf(stderr, ...), and show information
related to the previous size/pixel format configuration.

Consistent with the corresponding message issued in case of audio
configuration change.

Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffmpeg: warns the user when the selected pixel format is ignored
Stefano Sabatini [Sat, 26 Mar 2011 14:26:45 +0000 (15:26 +0100)]
ffmpeg: warns the user when the selected pixel format is ignored

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agoffplay: add a dummy option -i so that it is easy to switch between ffmpeg -i "file...
Benjamin Larsson [Tue, 8 Mar 2011 14:29:46 +0000 (15:29 +0100)]
ffplay: add a dummy option -i so that it is easy to switch between ffmpeg -i "file" and ffplay -i "file".

Signed-off-by: Anton Khirnov <anton@khirnov.net>
13 years agocosmetics: Fix crazy formatting in resample.
Alex Converse [Tue, 10 May 2011 23:58:01 +0000 (16:58 -0700)]
cosmetics: Fix crazy formatting in resample.

13 years agoAllow resampling with no channel count change for up to 8 channels.
Alex Converse [Tue, 10 May 2011 21:24:05 +0000 (14:24 -0700)]
Allow resampling with no channel count change for up to 8 channels.

13 years agoDon't allow unsupported resampling configurations.
Alex Converse [Tue, 10 May 2011 22:10:31 +0000 (15:10 -0700)]
Don't allow unsupported resampling configurations.

13 years agoPort x86 10-bit H.264 deblock asm from x264
Jason Garrett-Glaser [Tue, 10 May 2011 15:55:12 +0000 (08:55 -0700)]
Port x86 10-bit H.264 deblock asm from x264

13 years agoUpdate x86 H.264 deblock asm
Jason Garrett-Glaser [Tue, 10 May 2011 14:08:24 +0000 (07:08 -0700)]
Update x86 H.264 deblock asm

Includes AVX versions from x264.

13 years agoAVOptions: make default_val a union, as proposed in AVOption2.
Anton Khirnov [Sat, 30 Apr 2011 05:43:26 +0000 (07:43 +0200)]
AVOptions: make default_val a union, as proposed in AVOption2.

This breaks API and ABI.

13 years agoarm/h264pred: add missing argument type.
Ronald S. Bultje [Tue, 10 May 2011 12:44:49 +0000 (08:44 -0400)]
arm/h264pred: add missing argument type.

13 years agoh264dsp_mmx: place bracket outside #if/#endif block.
Ronald S. Bultje [Tue, 10 May 2011 12:39:38 +0000 (08:39 -0400)]
h264dsp_mmx: place bracket outside #if/#endif block.

Should fix compile on systems missing yasm/nasm.

13 years agolavf/utils: fix ff_interleave_compare_dts corner case.
Vitor Sessak [Sat, 7 May 2011 20:48:29 +0000 (22:48 +0200)]
lavf/utils: fix ff_interleave_compare_dts corner case.

This should fix behavior introduced by commit
96573c0d7605672d69b42ae1dcf18764ce47c71a. Av_rescale_rnd() is not
lossless so if two timestamps are equal after being rescaled they are
not always actually identical. This patch use av_compare_ts() to get
always a correct result.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agofate: add 10-bit H264 tests.
Ronald S. Bultje [Sat, 7 May 2011 01:46:54 +0000 (21:46 -0400)]
fate: add 10-bit H264 tests.

13 years agoh264: do not print "too many references" warning for intra-only.
Ronald S. Bultje [Fri, 6 May 2011 20:31:26 +0000 (16:31 -0400)]
h264: do not print "too many references" warning for intra-only.

Fixes issue 2679.

13 years agoEnable decoding of high bit depth h264.
Oskar Arvidsson [Tue, 29 Mar 2011 15:49:00 +0000 (17:49 +0200)]
Enable decoding of high bit depth h264.

This patch completes the high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoAdds 8-, 9- and 10-bit versions of some of the functions used by the h264 decoder.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:59 +0000 (17:48 +0200)]
Adds 8-, 9- and 10-bit versions of some of the functions used by the h264 decoder.

This patch lets e.g. dsputil_init chose dsp functions with respect to
the bit depth to decode. The naming scheme of bit depth dependent
functions is <base name>_<bit depth>[_<prefix>] (i.e. the old
clear_blocks_c is now named clear_blocks_8_c).

Note: Some of the functions for high bit depth is not dependent on the
bit depth, but only on the pixel size. This leaves some room for
optimizing binary size.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoAdd support for higher QP values in h264.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:58 +0000 (17:48 +0200)]
Add support for higher QP values in h264.

In high bit depth, the QP values may now be up to (51 + 6*(bit_depth-8)).

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoAdd the notion of pixel size in h264 related functions.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:57 +0000 (17:48 +0200)]
Add the notion of pixel size in h264 related functions.

In high bit depth the pixels will not be stored in uint8_t like in the
normal case, but in uint16_t. The pixel size is thus 1 in normal bit
depth and 2 in high bit depth.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoMake the h264 loop filter bit depth aware.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:56 +0000 (17:48 +0200)]
Make the h264 loop filter bit depth aware.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoTemplate dsputil_template.c with respect to pixel size, etc.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:55 +0000 (17:48 +0200)]
Template dsputil_template.c with respect to pixel size, etc.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoTemplate h264idct_template.c with respect to pixel size, etc.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:54 +0000 (17:48 +0200)]
Template h264idct_template.c with respect to pixel size, etc.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoPreparatory patch for high bit depth h264 decoding support.
Oskar Arvidsson [Mon, 9 May 2011 15:18:37 +0000 (11:18 -0400)]
Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoMove some functions in dsputil.c into a new file dsputil_template.c.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:51 +0000 (17:48 +0200)]
Move some functions in dsputil.c into a new file dsputil_template.c.

The functions moved are used when decoding h264.
Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoMove the functions in h264idct into a new file h264idct_template.c.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:50 +0000 (17:48 +0200)]
Move the functions in h264idct into a new file h264idct_template.c.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoMove the functions in h264pred.c into a new file h264pred_template.c.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:49 +0000 (17:48 +0200)]
Move the functions in h264pred.c into a new file h264pred_template.c.

Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoPreparatory patch for high bit depth h264 decoding support.
Oskar Arvidsson [Mon, 9 May 2011 14:58:40 +0000 (10:58 -0400)]
Preparatory patch for high bit depth h264 decoding support.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoAdd pixel formats for 9- and 10-bit yuv420p.
Oskar Arvidsson [Tue, 29 Mar 2011 15:48:47 +0000 (17:48 +0200)]
Add pixel formats for 9- and 10-bit yuv420p.

Also add support for these formats in libswscale.

Needed for high bit depth h264 decoding.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoChoose h264 chroma dc dequant function dynamically.
Oskar Arvidsson [Sun, 10 Apr 2011 20:42:36 +0000 (22:42 +0200)]
Choose h264 chroma dc dequant function dynamically.

Needed for high bit depth h264 decoding.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agoh264: DSP'ize MBAFF loopfilter.
Ronald S. Bultje [Mon, 9 May 2011 13:48:21 +0000 (09:48 -0400)]
h264: DSP'ize MBAFF loopfilter.

13 years agoffmpeg: Simplify decode loop condition to not use next_pts
Alexander Strange [Tue, 10 May 2011 05:53:46 +0000 (01:53 -0400)]
ffmpeg: Simplify decode loop condition to not use next_pts

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
13 years agolavc: move some flac-specific options to its private context.
Anton Khirnov [Sat, 2 Apr 2011 10:28:01 +0000 (12:28 +0200)]
lavc: move some flac-specific options to its private context.