platform/upstream/libav.git
10 years agoarm: Use the matching endfunc macro instead of the assembler directive directly
Martin Storsjö [Fri, 3 Jan 2014 09:39:10 +0000 (11:39 +0200)]
arm: Use the matching endfunc macro instead of the assembler directive directly

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoarm: Add a missing endfunc macro call
Martin Storsjö [Fri, 3 Jan 2014 09:36:08 +0000 (11:36 +0200)]
arm: Add a missing endfunc macro call

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agompegvideo: remove disabled bfin asm
Anton Khirnov [Tue, 3 Dec 2013 07:07:24 +0000 (08:07 +0100)]
mpegvideo: remove disabled bfin asm

It has been disabled as 'broken' over 3 years ago in b716a792

10 years agompegvideo: move dct_unquantize functions up to avoid forward declarations
Anton Khirnov [Tue, 3 Dec 2013 06:52:33 +0000 (07:52 +0100)]
mpegvideo: move dct_unquantize functions up to avoid forward declarations

10 years agompegvideo: split the encoding-only parts of frame_start() into a separate function
Anton Khirnov [Tue, 3 Dec 2013 06:48:09 +0000 (07:48 +0100)]
mpegvideo: split the encoding-only parts of frame_start() into a separate function

This introduces some code duplication. However, much of it should go
away once the decoders stop using MpegEncContext.

10 years agoFATE: add a test for the lavr mixing case fixed in fc6a3ef40d34ce8443ae57c2452f3f273d...
Anton Khirnov [Mon, 16 Dec 2013 20:30:39 +0000 (21:30 +0100)]
FATE: add a test for the lavr mixing case fixed in fc6a3ef40d34ce8443ae57c2452f3f273d7d4891

10 years agotests/Makefile: allow FILTER* to be called with lists of filter names
Anton Khirnov [Mon, 16 Dec 2013 21:21:01 +0000 (22:21 +0100)]
tests/Makefile: allow FILTER* to be called with lists of filter names

10 years agoaudio_mix: print (SKIP) instead of 0.0 for matrix columns removed along with output...
Anton Khirnov [Mon, 16 Dec 2013 18:05:45 +0000 (19:05 +0100)]
audio_mix: print (SKIP) instead of 0.0 for matrix columns removed along with output zeroing

10 years agoaudio_mix: fix zeroing output channels in certain cases
Anton Khirnov [Mon, 16 Dec 2013 17:44:15 +0000 (18:44 +0100)]
audio_mix: fix zeroing output channels in certain cases

Specifically, when the corresponding input channel exists and its matrix
column is all-zero (which is necessary for zeroing the output), the
matrix column must be removed from the matrix.

This is not done currently, so the mixing code would end up using
uninitialized pointers from stack.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
10 years agoaudio_mix: initialize the data pointers to NULL
Anton Khirnov [Mon, 16 Dec 2013 17:20:41 +0000 (18:20 +0100)]
audio_mix: initialize the data pointers to NULL

This should make it easier to catch problems where some of those
pointers are used uninitialized, since reading from NULL should always
crash, while random numbers from stack can turn out to be valid
pointers, so random memory may be silently overwritten.

10 years agoh264: reset data_partitioning if decoding the slice header for NAL_DPA fails
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
h264: reset data_partitioning if decoding the slice header for NAL_DPA fails

If it was set before then we can end up trying to decode a slice without
a valid slice header, which can lead to invalid memory access.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agolzw: switch to bytestream2
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
lzw: switch to bytestream2

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agopmpdec: check that there is at least one audio packet.
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
pmpdec: check that there is at least one audio packet.

The code cannot handle there being none, but that should not happen for
valid files.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agomathematics: remove asserts from av_rescale_rnd()
Anton Khirnov [Thu, 12 Dec 2013 06:34:13 +0000 (07:34 +0100)]
mathematics: remove asserts from av_rescale_rnd()

It is a public function, it must not assert on its parameters.

10 years agoeacmv: check the framerate before setting it.
Anton Khirnov [Thu, 12 Dec 2013 06:31:26 +0000 (07:31 +0100)]
eacmv: check the framerate before setting it.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agoadx: check that the offset is not negative
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
adx: check that the offset is not negative

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agorv30: fix extradata size check.
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
rv30: fix extradata size check.

It has been checking the number of bits in the offset instead of the
actual offset.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agompegvideo: set reference/pict_type on generated reference frames
Anton Khirnov [Thu, 28 Nov 2013 09:54:35 +0000 (10:54 +0100)]
mpegvideo: set reference/pict_type on generated reference frames

Otherwise the generic code will unref them, which can then result in
last_picture_ptr == current_picture_ptr, which causes deadlocks at least
in rv40.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org

10 years agoMakefile: add html template files to the doxygen deps
Anton Khirnov [Mon, 16 Dec 2013 08:30:17 +0000 (09:30 +0100)]
Makefile: add html template files to the doxygen deps

10 years agolavu: Move preprocessor macros in a separate file
Luca Barbato [Fri, 27 Dec 2013 01:57:48 +0000 (02:57 +0100)]
lavu: Move preprocessor macros in a separate file

And remove all the circular inclusions of avutil.h while at it.

10 years agoh264: namespace the decode function
Luca Barbato [Fri, 27 Dec 2013 06:48:13 +0000 (07:48 +0100)]
h264: namespace the decode function

Make much easier debugging.

10 years agodoxy: Update the css to have a flat style
Luca Barbato [Mon, 16 Dec 2013 00:18:56 +0000 (01:18 +0100)]
doxy: Update the css to have a flat style

Drop references to the doxy image gradients and style a the code blocks to
keep the whitespace and indent properly.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agodoxy: Fix link in badge color
Luca Barbato [Mon, 16 Dec 2013 00:24:15 +0000 (01:24 +0100)]
doxy: Fix link in badge color

Green on blue was not exactly optimal.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agoavutil: Move library version related macros to version.h
Diego Biurrun [Wed, 25 Dec 2013 16:16:05 +0000 (17:16 +0100)]
avutil: Move library version related macros to version.h

This is a more sensible place for these macros.

10 years agomms: Remove non-utf8 characters
Luca Barbato [Mon, 16 Dec 2013 05:40:24 +0000 (06:40 +0100)]
mms: Remove non-utf8 characters

Certain softwares get badly confused.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agoconfigure: Explicitly disable w32threads if the test for it fails
Diego Biurrun [Mon, 23 Dec 2013 03:24:52 +0000 (04:24 +0100)]
configure: Explicitly disable w32threads if the test for it fails

This avoids false positive enabling of w32threads if explicitly
requested on the command line, but dependencies are missing.

10 years agoerror.h: Do not circularly depend on avutil.h
Diego Biurrun [Fri, 13 Dec 2013 17:50:48 +0000 (18:50 +0100)]
error.h: Do not circularly depend on avutil.h

10 years agolavu: fix typo in documentation.
Tim Walker [Tue, 17 Dec 2013 15:47:31 +0000 (15:47 +0000)]
lavu: fix typo in documentation.

10 years agodoc: fix link to the ISC license text in the developer documentation.
Tim Walker [Sun, 22 Dec 2013 16:52:05 +0000 (16:52 +0000)]
doc: fix link to the ISC license text in the developer documentation.

10 years agodoc: fix a coding style error in the developer documentation.
Tim Walker [Sun, 22 Dec 2013 17:07:35 +0000 (17:07 +0000)]
doc: fix a coding style error in the developer documentation.

10 years agovc1: Fix intensity compensation performance regression
Mason Carter [Sun, 22 Dec 2013 01:27:18 +0000 (17:27 -0800)]
vc1: Fix intensity compensation performance regression

Introduced by 28243b0d35b47bbf9abbd454fc444a6e0a9e7b71

Intensity compensation is always used once it was encountered, because
v->next_use_ic is never set back to zero.

Reset v->next_use_ic, when resetting v->next_luty/uv.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agohevc: move DSP declarations from hevc.h into hevcdsp.h
Guillaume Martres [Fri, 20 Dec 2013 23:08:50 +0000 (00:08 +0100)]
hevc: move DSP declarations from hevc.h into hevcdsp.h

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agoAdd a libwebp encoder
Justin Ruggles [Fri, 22 Nov 2013 16:05:13 +0000 (11:05 -0500)]
Add a libwebp encoder

10 years agoarm: Don't clobber callee saved registers in scalarproduct
Martin Storsjö [Fri, 20 Dec 2013 13:02:35 +0000 (15:02 +0200)]
arm: Don't clobber callee saved registers in scalarproduct

q4-q7/d8-d15 are supposed to not be clobbered by the callee.

CC: libav-stable@libav.org
Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agohevc: support luma bit depth != chroma bit depth for PCM coding units
Mickaël Raulet [Sat, 14 Dec 2013 14:50:32 +0000 (15:50 +0100)]
hevc: support luma bit depth != chroma bit depth for PCM coding units

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: warn when an unknown profile is used
Guillaume Martres [Sat, 14 Dec 2013 14:50:31 +0000 (15:50 +0100)]
hevc: warn when an unknown profile is used

Bitstreams conforming to the spec should not use profiles not defined
in it.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc_ps: fix indentation
Gildas Cocherel [Fri, 20 Dec 2013 18:39:10 +0000 (19:39 +0100)]
hevc_ps: fix indentation

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: refactor Profile Tier Level
Gildas Cocherel [Wed, 18 Dec 2013 23:36:45 +0000 (00:36 +0100)]
hevc: refactor Profile Tier Level

Also store a few PTL flags which were skipped before

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: don't check for errors in PTL code
Guillaume Martres [Sat, 14 Dec 2013 14:50:29 +0000 (15:50 +0100)]
hevc: don't check for errors in PTL code

According to the spec, the value of XXX_reserved_zero_44bits should be
ignored, so don't report an error when it's not zero.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: store profile and level in AVCodecContext
Gildas Cocherel [Sat, 14 Dec 2013 14:50:28 +0000 (15:50 +0100)]
hevc: store profile and level in AVCodecContext

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agolavc: add HEVC profiles names
Gildas Cocherel [Wed, 18 Dec 2013 23:30:24 +0000 (00:30 +0100)]
lavc: add HEVC profiles names

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agoflv: Use the correct type to hold the file offset
Luca Barbato [Thu, 19 Dec 2013 06:38:15 +0000 (07:38 +0100)]
flv: Use the correct type to hold the file offset

10 years agojvdec: Do not feed the decoder with known wrong data
Luca Barbato [Fri, 13 Dec 2013 02:07:57 +0000 (03:07 +0100)]
jvdec: Do not feed the decoder with known wrong data

Still assume the size value is right in non-explode mode.

10 years agojvdec: Fix memory leak of jv->frames
Paul B Mahol [Tue, 3 Jul 2012 09:51:44 +0000 (09:51 +0000)]
jvdec: Fix memory leak of jv->frames

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
10 years agojvdec: Return EOF on end of file
Luca Barbato [Fri, 13 Dec 2013 02:10:16 +0000 (03:10 +0100)]
jvdec: Return EOF on end of file

10 years agox86: mpegvideo: move denoise_dct asm to mpegvideoenc
Anton Khirnov [Fri, 20 Dec 2013 12:00:22 +0000 (13:00 +0100)]
x86: mpegvideo: move denoise_dct asm to mpegvideoenc

This function is encoding-only.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
10 years agovc1: arm: Add NEON no_rnd chroma MC
Mason Carter [Sun, 15 Dec 2013 01:32:57 +0000 (17:32 -0800)]
vc1: arm: Add NEON no_rnd chroma MC

Apply David Conrad's old patch to the modern codebase.

http://ffmpeg.org/pipermail/ffmpeg-devel/2009-April/059877.html

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agovc1: arm: Add NEON assembly
Mason Carter [Sun, 15 Dec 2013 01:32:56 +0000 (17:32 -0800)]
vc1: arm: Add NEON assembly

For:

ff_vc1_inv_trans_{8,4}x{8,4}_{dc_,}neon
ff_put_pixels8x8_neon
ff_put_vc1_mspel_mc{0,1,2,3}{0,1,2,3}_neon (except for 00)

Based on ARM assembly code in libavcodec/arm by Rob Clark and Mans
Rullgard.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoatomics: cosmetics: Restructure ifdefs for greater clarity
Diego Biurrun [Wed, 18 Dec 2013 21:28:11 +0000 (22:28 +0100)]
atomics: cosmetics: Restructure ifdefs for greater clarity

Also fix the #endif comment in a few places and #include config.h to
avoid assuming it is #included implicitly.

10 years agoconfigure: Express atomics/thread deps through the dependency system
Diego Biurrun [Wed, 18 Dec 2013 20:52:28 +0000 (21:52 +0100)]
configure: Express atomics/thread deps through the dependency system

10 years agofate: Explicitly specify the pixel format for the pngparser test
Martin Storsjö [Wed, 18 Dec 2013 16:27:07 +0000 (18:27 +0200)]
fate: Explicitly specify the pixel format for the pngparser test

The RGB32 pixel format is RGBA/BGRA depending on target
endianness - make sure to convert it to one specific format for
the framecrc tests.

This fixes the pngparser fate test on big endian.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoPrepare release notes for Libav 10 "Eks" v10_alpha1
Anton Khirnov [Wed, 18 Dec 2013 02:52:13 +0000 (21:52 -0500)]
Prepare release notes for Libav 10 "Eks"

Additional editing by Diego Biurrun

10 years agopng_parser: Fix parsing on big endian
Martin Storsjö [Wed, 18 Dec 2013 08:39:47 +0000 (10:39 +0200)]
png_parser: Fix parsing on big endian

Since pc.state is populated by shifting in from the end of the
32 bit word, the content within pc.state is already in native endian
and should not be read with the AV_R{L,B} functions.

This was already done correctly for state64 above.

This fixes the fate-corepng test on big endian.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agovc1: Fix mb_height for field pictures
Michael Niedermayer [Sat, 14 Dec 2013 13:52:31 +0000 (14:52 +0100)]
vc1: Fix mb_height for field pictures

Tables are always allocated now with sufficient space for either progressive
or interlaced content. The alternative would be to detect a change
and reallocate.

This fixes decoding of a sample.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoapidoc: fix warning from stereo3d.h
Vittorio Giovara [Mon, 16 Dec 2013 13:40:06 +0000 (14:40 +0100)]
apidoc: fix warning from stereo3d.h

10 years agoh264: Refactor decode_frame_packing_arrangement
Vittorio Giovara [Fri, 13 Dec 2013 17:12:59 +0000 (18:12 +0100)]
h264: Refactor decode_frame_packing_arrangement

Directly set the fields when necessary.

10 years agohevc: Refactor decode_nal_sei_frame_packing_arrangement
Luca Barbato [Fri, 13 Dec 2013 15:39:45 +0000 (15:39 +0000)]
hevc: Refactor decode_nal_sei_frame_packing_arrangement

Directly set the fields when necessary.

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
10 years agofate: add a pngparser test
Vittorio Giovara [Wed, 11 Dec 2013 09:40:13 +0000 (10:40 +0100)]
fate: add a pngparser test

10 years agofate: add utility function to test parser, demuxer, and decoder
Vittorio Giovara [Thu, 12 Dec 2013 15:08:31 +0000 (16:08 +0100)]
fate: add utility function to test parser, demuxer, and decoder

10 years agopng: add a standalone parser
Peter Holik [Thu, 12 Dec 2013 15:11:35 +0000 (16:11 +0100)]
png: add a standalone parser

Useful for reading png images from a pipe.

10 years agoavutil: Remove deprecated intfloat_readwrite code
Diego Biurrun [Mon, 16 Dec 2013 14:25:05 +0000 (15:25 +0100)]
avutil: Remove deprecated intfloat_readwrite code

It was deprecated over two years ago.

10 years agobuild: Define __printf__ to __gnu_printf__ on MinGW*/gcc
Diego Biurrun [Fri, 13 Dec 2013 22:19:55 +0000 (23:19 +0100)]
build: Define __printf__ to __gnu_printf__ on MinGW*/gcc

This is necessary to avoid GCC assuming MS style printf arguments.

10 years agoBump major of libavfilter for reference counted buffer API changes.
Reinhard Tartler [Sat, 14 Dec 2013 21:31:03 +0000 (16:31 -0500)]
Bump major of libavfilter for reference counted buffer API changes.

Applications that have been linked against an older release of Libav and pick up
updated libraries experience segmentation faults because they pick up the new
libavfilter, which assumes AVFrames have been allocated by libavutil and thus
contain new reference-counting related fields. This will break for AVFrames that
have been allocated by old libavcodec.

All scheduled API changes are deferred to the next bump.

10 years agoBump libavutil major version to account for the LLS API/ABI changes.
Reinhard Tartler [Sat, 14 Dec 2013 21:30:25 +0000 (16:30 -0500)]
Bump libavutil major version to account for the LLS API/ABI changes.

Commit 41578f70cf8aec8e7565fba1ca7e07f3dc46c3d2 changed the LLS API, which was
called from libavcodec. Thus using an old libavcodec with a new libavutil will
break.

All scheduled API changes are deferred to the next bump.

10 years agortpdec_h264: Check the return value of functions doing allocations
Michael Niedermayer [Fri, 13 Dec 2013 00:27:26 +0000 (01:27 +0100)]
rtpdec_h264: Check the return value of functions doing allocations

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agosdp: Check theora colorspace before producing the configuration string
Martin Storsjö [Fri, 13 Dec 2013 22:27:26 +0000 (00:27 +0200)]
sdp: Check theora colorspace before producing the configuration string

This avoids a memory leak (or having to worry about freeing the
config string) if the colorspace isn't accepted.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agohdsenc: Check the init_file() return code
Michael Niedermayer [Thu, 12 Dec 2013 20:04:24 +0000 (21:04 +0100)]
hdsenc: Check the init_file() return code

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agohdsenc: Fix an off by one error in an array size check
Michael Niedermayer [Wed, 11 Dec 2013 19:17:23 +0000 (20:17 +0100)]
hdsenc: Fix an off by one error in an array size check

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agohdsenc: Avoid integer overflow
Michael Niedermayer [Wed, 11 Dec 2013 15:33:26 +0000 (16:33 +0100)]
hdsenc: Avoid integer overflow

Also remove a silly leftover pair of parentheses.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agoconfigure: Detect Solaris libc in an OpenIndiana/illumos compatible way
Diego Biurrun [Thu, 12 Dec 2013 11:14:30 +0000 (12:14 +0100)]
configure: Detect Solaris libc in an OpenIndiana/illumos compatible way

10 years agodoxy: Define a group for libswscale documentation
Luca Barbato [Thu, 12 Dec 2013 18:53:00 +0000 (19:53 +0100)]
doxy: Define a group for libswscale documentation

Uniform the style and eliminate a warning on newer doxygen version..

10 years agoconfigure: Reorder pthreads checks
Luca Barbato [Thu, 12 Dec 2013 16:05:34 +0000 (17:05 +0100)]
configure: Reorder pthreads checks

Some pthreads symbols might be present in libc (as shown on various *BSD)
but not all of them, leading to false positives.

Check for the most common compiler flags before the plain symbol check
to avoid known pitfalls.

10 years agoconfigure: Move the bz2 and zlib checks below phtreads
Luca Barbato [Wed, 11 Dec 2013 13:21:48 +0000 (14:21 +0100)]
configure: Move the bz2 and zlib checks below phtreads

There are alternate implementations of those libraries that use pthreads.

10 years agoavutil: Add deprecation ifdefs around obsolete intfloat code
Diego Biurrun [Fri, 13 Dec 2013 12:34:07 +0000 (13:34 +0100)]
avutil: Add deprecation ifdefs around obsolete intfloat code

This long-deprecated code should disappear at the next version bump.

10 years agobuild: Merge pthreads/w32threads OBJS declarations
Diego Biurrun [Thu, 12 Dec 2013 15:27:30 +0000 (16:27 +0100)]
build: Merge pthreads/w32threads OBJS declarations

10 years agoconfigure: Add -D__USE_MINGW_ANSI_STDIO=1 to CPPFLAGS on MinGW32
Diego Biurrun [Wed, 11 Dec 2013 23:39:43 +0000 (00:39 +0100)]
configure: Add -D__USE_MINGW_ANSI_STDIO=1 to CPPFLAGS on MinGW32

This enables a more C99-compliant implementation of printf et al.

10 years agoconfigure: Detect newer (>=4.1) versions of MinGW32
Diego Biurrun [Wed, 11 Dec 2013 23:34:43 +0000 (00:34 +0100)]
configure: Detect newer (>=4.1) versions of MinGW32

The MinGW32 project is deprecating its old version macros.

10 years agobuild: Do not add -g to HOSTCFLAGS
Diego Biurrun [Wed, 11 Dec 2013 23:04:27 +0000 (00:04 +0100)]
build: Do not add -g to HOSTCFLAGS

The flag is not unconditionally added to CFLAGS either and the
programs built on the host do not get habitually fed to a debugger.

10 years agompegts: add HEVC registration descriptor
Carl Eugen Hoyos [Wed, 16 Oct 2013 02:12:27 +0000 (04:12 +0200)]
mpegts: add HEVC registration descriptor

10 years agochangelog: drop redundant new attribute
Vittorio Giovara [Thu, 28 Nov 2013 14:01:50 +0000 (15:01 +0100)]
changelog: drop redundant new attribute

10 years agoimg2: add stereo 3d still picture file extensions
Martin Lambers [Wed, 16 Oct 2013 15:37:02 +0000 (17:37 +0200)]
img2: add stereo 3d still picture file extensions

10 years agortsp: suppress a incompatible pointer types warning
Vittorio Giovara [Tue, 10 Dec 2013 22:48:19 +0000 (22:48 +0000)]
rtsp: suppress a incompatible pointer types warning

10 years agoparseutils: fix discarding const attribute warning
Vittorio Giovara [Tue, 3 Dec 2013 11:57:51 +0000 (12:57 +0100)]
parseutils: fix discarding const attribute warning

10 years agomp3adu: Set the channel layout properly
Martin Storsjö [Wed, 11 Dec 2013 16:52:44 +0000 (18:52 +0200)]
mp3adu: Set the channel layout properly

This fixes decoding, broken since 7e35037.

This is similar to what was done for the normal mp3 decoder in
f4a86bc9.

Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agortspdec: Fix keep-alive request for ACTi cameras
Rumin Sam [Mon, 9 Dec 2013 20:43:51 +0000 (17:43 -0300)]
rtspdec: Fix keep-alive request for ACTi cameras

Some ACTi cameras fail if "*" is passed as the URI.

Signed-off-by: Ismael Luceno <ismael.luceno@corp.bluecherry.net>
Signed-off-by: Martin Storsjö <martin@martin.st>
10 years agolavc: deprecate avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 11:00:57 +0000 (12:00 +0100)]
lavc: deprecate avcodec_get_frame_defaults().

Also bump libavcodec micro and add an APIchanges entry saying that
av_frame_* should now be used instead of the lavc AVFrame functions.

10 years agoh264: call av_frame_unref() instead of avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:51:31 +0000 (11:51 +0100)]
h264: call av_frame_unref() instead of avcodec_get_frame_defaults().

This is a temporary workaround to allow deprecating
avcodec_get_frame_defaults(). The proper solution will be using a
properly allocated AVFrame in Picture.

10 years agoapi-example: remove an unneeded call to avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:55:57 +0000 (11:55 +0100)]
api-example: remove an unneeded call to avcodec_get_frame_defaults().

avcodec_decode_audio4() resets the frame itself.

10 years agolavf: remove an unneeded call to avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:55:57 +0000 (11:55 +0100)]
lavf: remove an unneeded call to avcodec_get_frame_defaults().

avcodec_decode_*() resets the frame itself.

10 years agompegvideo: call av_frame_unref() instead of avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:51:31 +0000 (11:51 +0100)]
mpegvideo: call av_frame_unref() instead of avcodec_get_frame_defaults().

This is a temporary workaround to allow deprecating
avcodec_get_frame_defaults(). The proper solution will be using a
properly allocated AVFrame in Picture.

10 years agompegvideo: remove an unneeded call to avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:54:46 +0000 (11:54 +0100)]
mpegvideo: remove an unneeded call to avcodec_get_frame_defaults().

ff_mpeg_unref_picture() already resets the frame.

10 years agoavplay: do not call avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:49:23 +0000 (11:49 +0100)]
avplay: do not call avcodec_get_frame_defaults().

avcodec_decode_audio4() will reset the frame itself.

10 years agolavc: call av_frame_unref() instead of avcodec_get_frame_defaults().
Anton Khirnov [Mon, 9 Dec 2013 10:48:25 +0000 (11:48 +0100)]
lavc: call av_frame_unref() instead of avcodec_get_frame_defaults().

avcodec_get_frame_defaults() will be deprecated.

10 years agoavconv: do not call avcodec_get_frame_defaults()
Anton Khirnov [Tue, 3 Dec 2013 11:00:32 +0000 (12:00 +0100)]
avconv: do not call avcodec_get_frame_defaults()

Strictly speaking it is not correct to call it on refcounted frames, in
any case it is unnecessary, since filtered_frame is always unreferenced
after poll_filter() returns.

10 years agolavc: deprecate avcodec_free_frame()
Anton Khirnov [Fri, 29 Nov 2013 20:43:56 +0000 (21:43 +0100)]
lavc: deprecate avcodec_free_frame()

av_frame_free() should be used instead.

10 years agoReplace all uses of avcodec_free_frame with av_frame_free().
Anton Khirnov [Fri, 29 Nov 2013 20:41:52 +0000 (21:41 +0100)]
Replace all uses of avcodec_free_frame with av_frame_free().

10 years agohevc: set time_base when possible
Guillaume Martres [Tue, 10 Dec 2013 14:42:29 +0000 (15:42 +0100)]
hevc: set time_base when possible

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: store the VPS list as an AVBufferRef, just like the others *PS
Gildas Cocherel [Tue, 10 Dec 2013 14:42:28 +0000 (15:42 +0100)]
hevc: store the VPS list as an AVBufferRef, just like the others *PS

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agohevc: do not dereference pointer before NULL check in verify_md5()
Michael Niedermayer [Tue, 10 Dec 2013 14:42:26 +0000 (15:42 +0100)]
hevc: do not dereference pointer before NULL check in verify_md5()

Signed-off-by: Anton Khirnov <anton@khirnov.net>
10 years agoconfigure: Add -D__USE_MINGW_ANSI_STDIO=1 to CPPFLAGS on MinGW64
Diego Biurrun [Sun, 8 Dec 2013 23:50:39 +0000 (00:50 +0100)]
configure: Add -D__USE_MINGW_ANSI_STDIO=1 to CPPFLAGS on MinGW64

This enables a more POSIX-compliant printf implementation, that
allows using length modifiers like 'z'.