platform/upstream/libav.git
11 years agofate: atrac: Place atrac1 and atrac3 tests in different groups
Diego Biurrun [Fri, 19 Oct 2012 23:09:33 +0000 (01:09 +0200)]
fate: atrac: Place atrac1 and atrac3 tests in different groups

11 years agofate: aac: cosmetics: Group AAC LATM tests together
Diego Biurrun [Sat, 20 Oct 2012 10:19:54 +0000 (12:19 +0200)]
fate: aac: cosmetics: Group AAC LATM tests together

11 years agox86: avresample: Add missing colons to assembly labels
Diego Biurrun [Tue, 6 Nov 2012 11:07:35 +0000 (12:07 +0100)]
x86: avresample: Add missing colons to assembly labels

YASM accepts labels without colons, but NASM issues warnings.

11 years agoFATE: add a 24-bit FLAC encoding test
Justin Ruggles [Mon, 5 Nov 2012 17:49:36 +0000 (12:49 -0500)]
FATE: add a 24-bit FLAC encoding test

11 years agoFATE: rename FLAC tests from flac-* to flac-16-*
Justin Ruggles [Sat, 27 Oct 2012 05:22:07 +0000 (01:22 -0400)]
FATE: rename FLAC tests from flac-* to flac-16-*

This will distinguish FLAC 16-bit tests from FLAC tests with other bit
depths.

11 years agoflacenc: use RICE2 entropy coding mode for 24-bit
Justin Ruggles [Sat, 27 Oct 2012 04:46:43 +0000 (00:46 -0400)]
flacenc: use RICE2 entropy coding mode for 24-bit

11 years agoflacenc: add 24-bit encoding
Justin Ruggles [Sat, 27 Oct 2012 04:26:02 +0000 (00:26 -0400)]
flacenc: add 24-bit encoding

11 years agoflacdsp: move lpc encoding from FLAC encoder to FLACDSPContext
Justin Ruggles [Sat, 27 Oct 2012 04:25:04 +0000 (00:25 -0400)]
flacdsp: move lpc encoding from FLAC encoder to FLACDSPContext

Also, templatize the functions for 16-bit and 32-bit sample range. This will
be used for 24-bit FLAC encoding.

11 years agoflacenc: use uint64_t for bit counts
Justin Ruggles [Sat, 27 Oct 2012 03:22:38 +0000 (23:22 -0400)]
flacenc: use uint64_t for bit counts

Needed to avoid integer overflows for 24-bit encoding.

11 years agoflacenc: remove wasted trailing 0 bits
Justin Ruggles [Thu, 25 Oct 2012 19:07:59 +0000 (15:07 -0400)]
flacenc: remove wasted trailing 0 bits

11 years agolavu: add av_ctz() for trailing zero bit count
Justin Ruggles [Fri, 26 Oct 2012 18:48:40 +0000 (14:48 -0400)]
lavu: add av_ctz() for trailing zero bit count

11 years agoflacenc: use a separate buffer for byte-swapping for MD5 checksum on big-endian
Justin Ruggles [Thu, 25 Oct 2012 19:00:10 +0000 (15:00 -0400)]
flacenc: use a separate buffer for byte-swapping for MD5 checksum on big-endian

This is much faster than calculating the MD5 one sample at a time.

11 years agofate: aac: Place LATM tests and general AAC tests in different groups
Diego Biurrun [Sat, 20 Oct 2012 10:21:41 +0000 (12:21 +0200)]
fate: aac: Place LATM tests and general AAC tests in different groups

11 years agobuild: The A64 muxer depends on rawenc.o for ff_raw_write_packet()
Diego Biurrun [Fri, 2 Nov 2012 21:49:04 +0000 (22:49 +0100)]
build: The A64 muxer depends on rawenc.o for ff_raw_write_packet()

11 years agox86: h264qpel: Only define mmxext QPEL functions if H264QPEL is enabled
Diego Biurrun [Fri, 2 Nov 2012 17:25:08 +0000 (18:25 +0100)]
x86: h264qpel: Only define mmxext QPEL functions if H264QPEL is enabled

This fixes compilation with --disable-everything and components enabled.

11 years agox86: PABSW: port to cpuflags
Diego Biurrun [Sun, 15 Jul 2012 15:59:26 +0000 (17:59 +0200)]
x86: PABSW: port to cpuflags

11 years agox86: vc1dsp: port to cpuflags
Diego Biurrun [Sun, 15 Jul 2012 16:49:44 +0000 (18:49 +0200)]
x86: vc1dsp: port to cpuflags

11 years agortmp: Use av_strlcat instead of strncat
Martin Storsjö [Mon, 5 Nov 2012 09:01:24 +0000 (11:01 +0200)]
rtmp: Use av_strlcat instead of strncat

Signed-off-by: Martin Storsjö <martin@martin.st>
11 years agodoc: add apidoc target for doxygen API documentation
Janne Grunau [Mon, 29 Oct 2012 15:26:59 +0000 (16:26 +0100)]
doc: add apidoc target for doxygen API documentation

Documentation includes only the externally visible API of the installed
headers.

Based on a patch by Anton Khirnov <anton@khirnov.net>.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
11 years agomatroskadec: do not use avpacket internals
Anton Khirnov [Thu, 1 Nov 2012 05:34:41 +0000 (06:34 +0100)]
matroskadec: do not use avpacket internals

11 years agolavf: fix av_interleaved_write_frame() doxy.
Anton Khirnov [Wed, 31 Oct 2012 15:54:16 +0000 (16:54 +0100)]
lavf: fix av_interleaved_write_frame() doxy.

A sentence was mistakenly inserted in the middle of another sentence.

11 years agolavf: clarify the lifetime of demuxed packets.
Anton Khirnov [Wed, 31 Oct 2012 18:59:53 +0000 (19:59 +0100)]
lavf: clarify the lifetime of demuxed packets.

11 years agoavconv: do not free muxed packet on streamcopy.
Anton Khirnov [Wed, 31 Oct 2012 20:15:41 +0000 (21:15 +0100)]
avconv: do not free muxed packet on streamcopy.

The packet belongs to lavf, the caller must not free it or access it in
any other way.
No change in practice, since destruct is set to NULL anyway.

11 years agocrc: move doxy to the header
Anton Khirnov [Fri, 2 Nov 2012 09:58:36 +0000 (10:58 +0100)]
crc: move doxy to the header

11 years agovf_drawtext: do not use deprecated av_tree_node_size
Anton Khirnov [Fri, 2 Nov 2012 09:56:30 +0000 (10:56 +0100)]
vf_drawtext: do not use deprecated av_tree_node_size

11 years agox86: Refactor PSWAPD fallback implementations and port to cpuflags
Diego Biurrun [Wed, 1 Aug 2012 22:55:34 +0000 (00:55 +0200)]
x86: Refactor PSWAPD fallback implementations and port to cpuflags

11 years agoparser: Move Doxygen documentation to the header files
Diego Biurrun [Fri, 2 Nov 2012 11:04:16 +0000 (12:04 +0100)]
parser: Move Doxygen documentation to the header files

11 years agoPGS subtitles: Expose forced flag
John Stebbins [Sat, 20 Oct 2012 16:56:11 +0000 (09:56 -0700)]
PGS subtitles: Expose forced flag

Useful for detection of subtitles displayed during foreign language
scenes.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
11 years agox86: PMINUB: port to cpuflags
Diego Biurrun [Sun, 15 Jul 2012 15:43:50 +0000 (17:43 +0200)]
x86: PMINUB: port to cpuflags

11 years agox86: ac3dsp: port to cpuflags
Diego Biurrun [Sun, 15 Jul 2012 13:41:30 +0000 (15:41 +0200)]
x86: ac3dsp: port to cpuflags

11 years agox86util: Add cpuflags_mmxext alias for cpuflags_mmx2
Diego Biurrun [Tue, 30 Oct 2012 00:11:14 +0000 (01:11 +0100)]
x86util: Add cpuflags_mmxext alias for cpuflags_mmx2

"mmxext" is a more sensible name and more common in outside projects.

11 years agox86inc: Only define program_name if the macro is unset
Diego Biurrun [Tue, 30 Oct 2012 01:54:25 +0000 (02:54 +0100)]
x86inc: Only define program_name if the macro is unset

This allows overriding the value from outside of the file.

11 years agodv1394: Swap the min and max values of the 'standard' option
Michael Niedermayer [Fri, 2 Nov 2012 01:07:15 +0000 (02:07 +0100)]
dv1394: Swap the min and max values of the 'standard' option

DV1394_NTSC has a lower value than DV1394_PAL.

Signed-off-by: Martin Storsjö <martin@martin.st>
11 years agortpdec_vp8: Don't parse fields that aren't used
Martin Storsjö [Mon, 8 Oct 2012 22:17:45 +0000 (01:17 +0300)]
rtpdec_vp8: Don't parse fields that aren't used

This avoids warnings about unused variables.

Signed-off-by: Martin Storsjö <martin@martin.st>
11 years agolavc: add some AVPacket doxy.
Anton Khirnov [Wed, 31 Oct 2012 16:27:24 +0000 (17:27 +0100)]
lavc: add some AVPacket doxy.

11 years agoaudiointerleave: deobfuscate a function call.
Anton Khirnov [Wed, 31 Oct 2012 21:10:41 +0000 (22:10 +0100)]
audiointerleave: deobfuscate a function call.

right above there is if (pkt) {<do stuff>; pkt = NULL}, so pkt is just a
fancy name for NULL at this point.

11 years agortpdec: factorize identical code used in several handlers
Anton Khirnov [Thu, 1 Nov 2012 13:03:04 +0000 (14:03 +0100)]
rtpdec: factorize identical code used in several handlers

11 years agoa64: remove interleaved mode.
Anton Khirnov [Wed, 31 Oct 2012 21:32:36 +0000 (22:32 +0100)]
a64: remove interleaved mode.

It has been disabled since it was added two years ago.

11 years agodoc: Point to the new location of the c99-to-c89 tool
Martin Storsjö [Thu, 1 Nov 2012 13:27:18 +0000 (15:27 +0200)]
doc: Point to the new location of the c99-to-c89 tool

This location contains prebuilt binaries as well.

Signed-off-by: Martin Storsjö <martin@martin.st>
11 years agodecode_audio3: initialize AVFrame
Ilkka Ollakka [Wed, 31 Oct 2012 09:24:36 +0000 (11:24 +0200)]
decode_audio3: initialize AVFrame

Same fix and issue as in a25d912dca9cd553440167e0476c47581359c0fc

Signed-off-by: Anton Khirnov <anton@khirnov.net>
11 years agows-snd1: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 21:22:53 +0000 (17:22 -0400)]
ws-snd1: set channel layout

11 years agowmavoice: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 21:19:25 +0000 (17:19 -0400)]
wmavoice: set channel layout

11 years agowmapro: use AVCodecContext.channels instead of keeping a private copy
Justin Ruggles [Tue, 23 Oct 2012 21:10:37 +0000 (17:10 -0400)]
wmapro: use AVCodecContext.channels instead of keeping a private copy

11 years agowma: do not keep private copies of some AVCodecContext fields
Justin Ruggles [Tue, 23 Oct 2012 20:30:59 +0000 (16:30 -0400)]
wma: do not keep private copies of some AVCodecContext fields

channels, sample_rate, bit_rate, and block_align can be used directly from
the AVCodecContext

11 years agovmdaudio: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 19:27:44 +0000 (15:27 -0400)]
vmdaudio: set channel layout

11 years agotwinvq: validate sample rate code
Justin Ruggles [Tue, 23 Oct 2012 17:17:50 +0000 (13:17 -0400)]
twinvq: validate sample rate code

A large invalid value could cause undefined behavior when left-shifted
by 8 later in the function.

11 years agotwinvq: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 17:17:17 +0000 (13:17 -0400)]
twinvq: set channel layout

11 years agotwinvq: validate that channels is not <= 0
Justin Ruggles [Tue, 23 Oct 2012 17:15:24 +0000 (13:15 -0400)]
twinvq: validate that channels is not <= 0

This could occur due to integer overflow when reading the channel count from
the extradata.

11 years agotruespeech: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 04:53:16 +0000 (00:53 -0400)]
truespeech: set channel layout

11 years agosipr: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 04:51:26 +0000 (00:51 -0400)]
sipr: set channel layout

11 years agoshorten: validate that the channel count in the header is not <= 0
Justin Ruggles [Tue, 23 Oct 2012 04:40:51 +0000 (00:40 -0400)]
shorten: validate that the channel count in the header is not <= 0

11 years agora288dec: set channel layout
Justin Ruggles [Tue, 23 Oct 2012 04:35:50 +0000 (00:35 -0400)]
ra288dec: set channel layout

11 years agora144dec: set channel layout
Justin Ruggles [Mon, 22 Oct 2012 22:57:39 +0000 (18:57 -0400)]
ra144dec: set channel layout

11 years agoqdm2: remove unneeded checks for channel count
Justin Ruggles [Mon, 22 Oct 2012 22:54:29 +0000 (18:54 -0400)]
qdm2: remove unneeded checks for channel count

11 years agoqdm2: make sure channels is not <= 0 and set channel layout
Justin Ruggles [Mon, 22 Oct 2012 22:53:19 +0000 (18:53 -0400)]
qdm2: make sure channels is not <= 0 and set channel layout

11 years agoqcelpdec: set channel layout
Justin Ruggles [Mon, 22 Oct 2012 22:25:46 +0000 (18:25 -0400)]
qcelpdec: set channel layout

11 years agonellymoserdec: set channels to 1
Justin Ruggles [Mon, 22 Oct 2012 21:58:24 +0000 (17:58 -0400)]
nellymoserdec: set channels to 1

11 years agolibopencore-amr: set channel layout for amr-nb or if not set by the user
Justin Ruggles [Mon, 22 Oct 2012 21:49:59 +0000 (17:49 -0400)]
libopencore-amr: set channel layout for amr-nb or if not set by the user

11 years agolibilbc: set channel layout
Justin Ruggles [Mon, 22 Oct 2012 21:41:47 +0000 (17:41 -0400)]
libilbc: set channel layout

11 years agodpcm: use AVCodecContext.channels instead of keeping a private copy
Justin Ruggles [Mon, 22 Oct 2012 20:03:20 +0000 (16:03 -0400)]
dpcm: use AVCodecContext.channels instead of keeping a private copy

11 years agoimc: set channels to 1 instead of validating it
Justin Ruggles [Mon, 22 Oct 2012 19:54:29 +0000 (15:54 -0400)]
imc: set channels to 1 instead of validating it

11 years agogsmdec: always set channel layout and sample rate at initialization
Justin Ruggles [Mon, 22 Oct 2012 19:45:38 +0000 (15:45 -0400)]
gsmdec: always set channel layout and sample rate at initialization

Only mono 8kHz is supported.

11 years agolibgsmdec: always set channel layout and sample rate at initialization
Justin Ruggles [Mon, 22 Oct 2012 19:43:59 +0000 (15:43 -0400)]
libgsmdec: always set channel layout and sample rate at initialization

Only mono 8kHz is supported.

11 years agog726dec: do not validate sample rate
Justin Ruggles [Mon, 22 Oct 2012 17:26:46 +0000 (13:26 -0400)]
g726dec: do not validate sample rate

For decoding it does not really matter what the sample rate is.

11 years agog726dec: set channel layout at initialization instead of validating it
Justin Ruggles [Mon, 22 Oct 2012 17:26:14 +0000 (13:26 -0400)]
g726dec: set channel layout at initialization instead of validating it

11 years agoflacdec: do not warn on sample rate change
Justin Ruggles [Sun, 21 Oct 2012 21:04:58 +0000 (17:04 -0400)]
flacdec: do not warn on sample rate change

11 years agoflacdec: allow mid-stream channel layout change
Justin Ruggles [Sun, 21 Oct 2012 21:02:28 +0000 (17:02 -0400)]
flacdec: allow mid-stream channel layout change

Although the libFLAC decoder cannot handle such a change, it is allowed by the
spec and could potentially occur with live streams.

11 years agoflacdec: use av_samples_* functions for sample buffer allocation
Justin Ruggles [Sun, 21 Oct 2012 20:15:34 +0000 (16:15 -0400)]
flacdec: use av_samples_* functions for sample buffer allocation

Also, return an error on allocation failure.

11 years agodcadec: allow the decoder to change the channel layout mid-stream
Justin Ruggles [Sun, 21 Oct 2012 19:12:36 +0000 (15:12 -0400)]
dcadec: allow the decoder to change the channel layout mid-stream

11 years agocook: use av_dlog() for debug logging instead of av_log() with AV_LOG_ERROR
Justin Ruggles [Sun, 21 Oct 2012 18:47:10 +0000 (14:47 -0400)]
cook: use av_dlog() for debug logging instead of av_log() with AV_LOG_ERROR

11 years agocook: move samples_per_frame from COOKSubpacket to where it is used
Justin Ruggles [Sun, 21 Oct 2012 18:43:02 +0000 (14:43 -0400)]
cook: move samples_per_frame from COOKSubpacket to where it is used

11 years agocook: use av_get_channel_layout_nb_channels() instead of cook_count_channels()
Justin Ruggles [Sun, 21 Oct 2012 18:36:38 +0000 (14:36 -0400)]
cook: use av_get_channel_layout_nb_channels() instead of cook_count_channels()

11 years agocook: reverse a condition so that the code makes more sense
Justin Ruggles [Sun, 21 Oct 2012 18:33:11 +0000 (14:33 -0400)]
cook: reverse a condition so that the code makes more sense

11 years agocook: remove unneeded COOKContext variable, sample_rate
Justin Ruggles [Sun, 21 Oct 2012 18:26:29 +0000 (14:26 -0400)]
cook: remove unneeded COOKContext variable, sample_rate

11 years agocook: remove unneeded COOKContext variable, bit_rate
Justin Ruggles [Sun, 21 Oct 2012 18:25:18 +0000 (14:25 -0400)]
cook: remove unneeded COOKContext variable, bit_rate

11 years agocook: use AVCodecContext.channels instead of keeping a private copy
Justin Ruggles [Sun, 21 Oct 2012 18:24:04 +0000 (14:24 -0400)]
cook: use AVCodecContext.channels instead of keeping a private copy

11 years agobmvaudio: set channel layout at init() rather than validating it
Justin Ruggles [Sun, 21 Oct 2012 17:48:50 +0000 (13:48 -0400)]
bmvaudio: set channel layout at init() rather than validating it

11 years agoatrac1: do not keep a copy of channel count in the private context
Justin Ruggles [Wed, 17 Oct 2012 15:45:49 +0000 (11:45 -0400)]
atrac1: do not keep a copy of channel count in the private context

11 years agodsicinaudio: set channels and channel layout
Justin Ruggles [Wed, 17 Oct 2012 15:29:04 +0000 (11:29 -0400)]
dsicinaudio: set channels and channel layout

11 years agog722dec: set channel layout at initialization instead of validating it
Justin Ruggles [Mon, 15 Oct 2012 20:40:44 +0000 (16:40 -0400)]
g722dec: set channel layout at initialization instead of validating it

11 years agoamrwbdec: set channels, channel_layout, and sample_rate
Justin Ruggles [Mon, 8 Oct 2012 01:19:28 +0000 (21:19 -0400)]
amrwbdec: set channels, channel_layout, and sample_rate

Only mono 16kHz is supported.

11 years agoamrnbdec: set channels, channel_layout, and sample_rate
Justin Ruggles [Mon, 8 Oct 2012 01:17:45 +0000 (21:17 -0400)]
amrnbdec: set channels, channel_layout, and sample_rate

Only mono 8kHz is supported.

11 years agodca_parser: allow the parser to change the sample rate
Justin Ruggles [Mon, 8 Oct 2012 00:52:35 +0000 (20:52 -0400)]
dca_parser: allow the parser to change the sample rate

11 years agolavc: check channel count after decoder init
Justin Ruggles [Tue, 23 Oct 2012 04:56:00 +0000 (00:56 -0400)]
lavc: check channel count after decoder init

Ensures the decoder did not set channel count to an insanely high value
during initialization, which could cause large memory usage when it tries to
get a buffer during decoding.

11 years agolavc: move SANE_NB_CHANNELS to internal.h and use it in the PCM decoders
Justin Ruggles [Mon, 22 Oct 2012 17:40:10 +0000 (13:40 -0400)]
lavc: move SANE_NB_CHANNELS to internal.h and use it in the PCM decoders

11 years agofate: add ac3/eac3 tests to FATE_SAMPLES_AVCONV
Janne Grunau [Wed, 31 Oct 2012 18:56:02 +0000 (19:56 +0100)]
fate: add ac3/eac3 tests to FATE_SAMPLES_AVCONV

Fixes typo in 2cbdd7c92958cb8226491d8eb23ed2d57d4b841e.

11 years agoavconv_opt, cmdutils: Add missing function parameter Doxygen
Diego Biurrun [Wed, 24 Oct 2012 17:20:13 +0000 (19:20 +0200)]
avconv_opt, cmdutils: Add missing function parameter Doxygen

11 years agox86: Move optimization suffix to end of function names
Diego Biurrun [Fri, 27 Jul 2012 12:09:52 +0000 (14:09 +0200)]
x86: Move optimization suffix to end of function names

This simplifies cpuflags porting.

11 years agox86: h264_chromamc_10bit: drop pointless PAVG %define
Diego Biurrun [Fri, 27 Jul 2012 11:43:33 +0000 (13:43 +0200)]
x86: h264_chromamc_10bit: drop pointless PAVG %define

It is only used in one place so there is no need for the abstraction.

11 years agox86: mmx2 ---> mmxext in function names
Diego Biurrun [Sun, 8 Jul 2012 17:56:10 +0000 (19:56 +0200)]
x86: mmx2 ---> mmxext in function names

11 years agoswscale: do not forget to swap data in formats with different endianness
Kostya Shishkov [Wed, 31 Oct 2012 10:26:32 +0000 (11:26 +0100)]
swscale: do not forget to swap data in formats with different endianness

Otherwise during scaling it will try to interpret input in the wrong way and
that leads to the test results disagreeing on different platforms and with
different optimizations.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
11 years agoaacdec: Drop some unused function arguments
Diego Biurrun [Wed, 24 Oct 2012 17:15:54 +0000 (19:15 +0200)]
aacdec: Drop some unused function arguments

11 years agox86: Fix assembly with NASM
Dave Yeo [Wed, 31 Oct 2012 06:48:26 +0000 (23:48 -0700)]
x86: Fix assembly with NASM

Unlike YASM, NASM only looks for include files in the current
directory, not in the directory that included files reside in.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
11 years agox86: fmtconvert: Refactor cvtps2pi emulation through cpuflags
Diego Biurrun [Sun, 15 Jul 2012 17:55:21 +0000 (19:55 +0200)]
x86: fmtconvert: Refactor cvtps2pi emulation through cpuflags

11 years agox86: fmtconvert: port to cpuflags
Diego Biurrun [Sun, 15 Jul 2012 13:42:17 +0000 (15:42 +0200)]
x86: fmtconvert: port to cpuflags

11 years agox86: MMX2 ---> MMXEXT in macro names
Diego Biurrun [Sun, 8 Jul 2012 17:28:57 +0000 (19:28 +0200)]
x86: MMX2 ---> MMXEXT in macro names

11 years agox86: mmx2 ---> mmxext in variable names
Diego Biurrun [Sun, 8 Jul 2012 17:55:31 +0000 (19:55 +0200)]
x86: mmx2 ---> mmxext in variable names

11 years agox86: mmx2 ---> mmxext in comments and messages
Diego Biurrun [Sun, 8 Jul 2012 17:16:20 +0000 (19:16 +0200)]
x86: mmx2 ---> mmxext in comments and messages

11 years agox86: yasm: Use complete source path for macro helper %includes
Diego Biurrun [Sun, 15 Jul 2012 09:48:21 +0000 (11:48 +0200)]
x86: yasm: Use complete source path for macro helper %includes

This is more consistent with the way we handle C #includes and
it simplifies the build system.

11 years agox86: include x86inc.asm in x86util.asm
Diego Biurrun [Sat, 7 Jul 2012 23:20:48 +0000 (01:20 +0200)]
x86: include x86inc.asm in x86util.asm

This is necessary to allow refactoring some x86util macros with cpuflags.