shahriman AMS [Tue, 8 Nov 2011 08:47:32 +0000 (08:47 +0000)]
vc1dec: use correct hybrid prediction threshold.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
shahriman AMS [Tue, 8 Nov 2011 08:47:31 +0000 (08:47 +0000)]
vc1dec: Partial rewrite of vc1_pred_mv()
The current code is a generalization of the earlier progressive MV
prediction code. This was supposed to predict MVs for both interlaced
and progressive pictures. But the interlaced MV prediction is buggy
and works mostly by luck.
This partially fixes interlaced MV prediction.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
shahriman AMS [Tue, 8 Nov 2011 08:47:30 +0000 (08:47 +0000)]
vc1dec: take ME precision into account while scaling MV predictors.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Andrey Utkin [Tue, 8 Nov 2011 15:24:04 +0000 (17:24 +0200)]
lavf: don't leak corrupted packets
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Martin Storsjö [Mon, 7 Nov 2011 09:19:17 +0000 (11:19 +0200)]
tls: Use ERR_get_error() in do_tls_poll
The return value ret isn't an error code that can be passed
to ERR_error_string().
This makes the error messages printed actually contain useful
information.
Signed-off-by: Martin Storsjö <martin@martin.st>
Alex Converse [Tue, 8 Nov 2011 02:46:46 +0000 (18:46 -0800)]
indeo3: Fix a fencepost error.
Found with asan and the venerable 1-dog.avi sample.
Alex Converse [Mon, 7 Nov 2011 18:14:54 +0000 (10:14 -0800)]
mxfdec: Fix comparison of unsigned expression < 0.
'size' is populated by functions returning int64_t and int that return
negative error codes.
Anton Khirnov [Tue, 8 Nov 2011 13:13:30 +0000 (14:13 +0100)]
mpegts: set stream id on just created stream, not an unrelated variable
Bug introduced in
84ad31ff180fa089cd6bfd93c246336a16036455.
Thanks to Uoti Urpala for finding it.
Justin Ruggles [Fri, 4 Nov 2011 00:41:17 +0000 (20:41 -0400)]
ra288: return error if input buffer is too small
Justin Ruggles [Sat, 29 Oct 2011 04:42:48 +0000 (00:42 -0400)]
ra288: utilize DSPContext.vector_fmul()
Justin Ruggles [Sat, 29 Oct 2011 04:09:03 +0000 (00:09 -0400)]
ra288: use memcpy() to copy decoded samples to output
Justin Ruggles [Sat, 29 Oct 2011 01:27:36 +0000 (21:27 -0400)]
mace: only calculate output buffer size once
Diego Biurrun [Sun, 30 Oct 2011 21:08:47 +0000 (22:08 +0100)]
Remove redundant filename self-references inside files.
Filenames are brittle across renames and add no useful information.
Diego Biurrun [Tue, 8 Nov 2011 15:06:50 +0000 (16:06 +0100)]
indeo3data: add missing config.h #include for HAVE_BIGENDIAN
Diego Biurrun [Tue, 8 Nov 2011 12:46:18 +0000 (13:46 +0100)]
x86: drop pointless ARCH_X86 #ifdef from files in x86 subdirectory
Justin Ruggles [Tue, 8 Nov 2011 15:33:18 +0000 (10:33 -0500)]
avplay: reset rdft when closing stream.
this fixes a crash when cycling audio streams if the spectrograph is
displayed.
Anton Khirnov [Tue, 8 Nov 2011 07:07:51 +0000 (08:07 +0100)]
doc/git-howto: expand format-patch and send-email notes.
Anton Khirnov [Sun, 6 Nov 2011 14:16:18 +0000 (15:16 +0100)]
lavf: expand doxy for some AVFormatContext fields.
Martin Storsjö [Mon, 7 Nov 2011 21:07:39 +0000 (23:07 +0200)]
avformat: Avoid a warning about mixed declarations and code
Signed-off-by: Martin Storsjö <martin@martin.st>
Kostya Shishkov [Tue, 1 Nov 2011 10:58:44 +0000 (11:58 +0100)]
BMV demuxer and decoder
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Mon, 7 Nov 2011 19:06:01 +0000 (21:06 +0200)]
matroskaenc: Make sure the seekhead struct is freed even on seek failure
The caller expects the seekhead struct to be freed when calling
matroska_write_seekhead. Currently, the structure is leaked if the
seek fails.
Signed-off-by: Martin Storsjö <martin@martin.st>
Alex Converse [Thu, 3 Nov 2011 23:02:17 +0000 (16:02 -0700)]
mpeg12enc: Remove write-only variables.
Alex Converse [Thu, 3 Nov 2011 22:56:37 +0000 (15:56 -0700)]
mpeg12enc: Don't set up run-level info for level 0.
run: The number of zero coefficients preceding a non-zero coefficient,
in the scan order. The absolute value of the non-zero coefficient is
called "level".
The run-level code makes illegal reads when trying to set up tables for
nonsense level 0.
Alex Converse [Thu, 3 Nov 2011 22:57:23 +0000 (15:57 -0700)]
msmpeg4: Don't set up run-level info for level 0.
run: The number of zero coefficients preceding a non-zero coefficient,
in the scan order. The absolute value of the non-zero coefficient is
called "level".
The run-level code makes illegal reads when trying to set up tables for
nonsense level 0.
Martin Storsjö [Mon, 7 Nov 2011 11:57:33 +0000 (13:57 +0200)]
avformat: Warn about using network functions without calling avformat_network_init
This is to make developers aware of the fact that they will
start using the new init function at some point.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Mon, 7 Nov 2011 12:09:06 +0000 (14:09 +0200)]
avformat: Revise wording
It might make sense not to make the function completely mandatory
immediately at the next bump, which might be quite soon after
the function was introduced.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 6 Nov 2011 22:25:03 +0000 (00:25 +0200)]
rdt: Set AVFMT_NOFILE on ff_rdt_demuxer
This makes rdt work again, which has been broken since
603b8bc2a109978c8499b06d2556f1433306eca7. This commit made
opening a demuxer without a file (or in this case, with a filename
which can't be opened) fail, unless the demuxer actually declared
AVFMT_NOFILE.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 6 Nov 2011 21:51:19 +0000 (23:51 +0200)]
rdt: Check the return value of avformat_open
If it failed, return NULL. This avoids trying to use an
half-initialized RDTDemuxContext.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 6 Nov 2011 21:50:39 +0000 (23:50 +0200)]
rtsp: Discard the dynamic handler, if it has an alloc function which failed
Signed-off-by: Martin Storsjö <martin@martin.st>
Justin Ruggles [Sun, 6 Nov 2011 20:27:55 +0000 (15:27 -0500)]
dsputil: use cpuflags in x86 versions of vector_clip_int32()
Justin Ruggles [Sun, 6 Nov 2011 19:22:42 +0000 (14:22 -0500)]
x86inc: use sse versions of common macros instead of sse2 when applicable
Anton Khirnov [Sun, 6 Nov 2011 19:33:13 +0000 (20:33 +0100)]
doc/APIchanges: add missing dates and hashes
Anton Khirnov [Sun, 6 Nov 2011 18:13:28 +0000 (19:13 +0100)]
lavf: don't return from void av_update_cur_dts()
Anton Khirnov [Sun, 6 Nov 2011 13:22:22 +0000 (14:22 +0100)]
Changelog: add more entries.
Anton Khirnov [Sun, 6 Nov 2011 13:05:46 +0000 (14:05 +0100)]
Changelog: update ffmpeg/avconv incompatibility list.
Anton Khirnov [Sun, 6 Nov 2011 09:32:44 +0000 (10:32 +0100)]
avconv: remove some redundant temporary variables.
Anton Khirnov [Sun, 6 Nov 2011 09:27:43 +0000 (10:27 +0100)]
avconv: fix broken indentation
Anton Khirnov [Sun, 6 Nov 2011 09:26:56 +0000 (10:26 +0100)]
avconv: move copy_initial_nonkeyframes to the options context.
Also document it.
Anton Khirnov [Sun, 6 Nov 2011 12:00:13 +0000 (13:00 +0100)]
avconv: use file:stream instead of file.stream in log messages.
Anton Khirnov [Sat, 5 Nov 2011 15:23:23 +0000 (16:23 +0100)]
doc/avconv: elaborate on basic functionality.
Anton Khirnov [Sat, 5 Nov 2011 14:55:25 +0000 (15:55 +0100)]
doc/avconv: -sample_fmts, not -help sample_fmts prints the sample formats
Martin Storsjö [Sat, 5 Nov 2011 23:50:03 +0000 (01:50 +0200)]
openssl: Only use CRYPTO_set_id_callback on OpenSSL < 1.0.0
Since 1.0.0, this function is deprecated. A new function,
CRYPTO_THREADID_set_callback is available, but if not set at all,
it uses the address of errno as thread id, which should be
sufficient for most systems.
On windows, it never was necessary to use this function even
before 1.0.0, it used the right win32 API function for this
by default.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 6 Nov 2011 00:47:48 +0000 (02:47 +0200)]
Call avformat_network_init/deinit in the programs
Calling the init function will become mandatory at some later
point. By calling it, more heavy network init (such as SSL/TLS
library init) can be done once at startup, instead of implicitly
when used (which could lead to it being done a number of times).
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Fri, 4 Nov 2011 13:08:49 +0000 (15:08 +0200)]
Remove leftover includes of strings.h
Signed-off-by: Martin Storsjö <martin@martin.st>
Reimar Döffinger [Sat, 5 Nov 2011 15:21:04 +0000 (17:21 +0200)]
avutil: Don't allow using strcasecmp/strncasecmp
Signed-off-by: Martin Storsjö <martin@martin.st>
Reimar Döffinger [Wed, 2 Nov 2011 19:17:25 +0000 (20:17 +0100)]
Replace all usage of strcasecmp/strncasecmp
All current usages of it are incompatible with localization.
For example strcasecmp("i", "I") != 0 is possible, but would
break many of the places where it is used.
Instead use our own implementations that always treat the data
as ASCII.
Signed-off-by: Martin Storsjö <martin@martin.st>
Reimar Döffinger [Sat, 5 Nov 2011 15:20:41 +0000 (17:20 +0200)]
avstring: Add locale independent implementations of strcasecmp/strncasecmp
Signed-off-by: Martin Storsjö <martin@martin.st>
Luca Barbato [Sat, 5 Nov 2011 15:15:43 +0000 (17:15 +0200)]
avstring: Add locale independent implementations of toupper/tolower
Signed-off-by: Martin Storsjö <martin@martin.st>
Kostya Shishkov [Sat, 5 Nov 2011 13:34:23 +0000 (14:34 +0100)]
cosmetics: insert some spaces in explicit enum value assignments
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Kostya Shishkov [Sat, 5 Nov 2011 13:34:08 +0000 (14:34 +0100)]
move 8SVX audio codecs to the audio codec list part on the next bump
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Kostya Shishkov [Sat, 5 Nov 2011 13:33:45 +0000 (14:33 +0100)]
deprecate codec IDs that won't ever be used
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Ronald S. Bultje [Sun, 6 Nov 2011 00:31:40 +0000 (17:31 -0700)]
swscale: write yuv2plane1 MMX/SSE2/SSE4/AVX functions.
Loren Merritt [Wed, 2 Nov 2011 20:13:26 +0000 (20:13 +0000)]
Update x86inc.asm to latest x264 version, and add AVX symmetry.
We keep INIT_AVX (for backwards compatibility). 3arg AVX ops with
a memory arg can only have it in src2, whereas SSE emulation of
3arg prefers to have it in src1 (i.e. the mov). So, if the op is
symmetric and the wrong one is memory, swap them.
Martin Storsjö [Sun, 20 Jun 2010 20:59:15 +0000 (23:59 +0300)]
http: Remove the custom function for disabling chunked posts
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sun, 20 Jun 2010 20:58:22 +0000 (23:58 +0300)]
rtsp: Disable chunked http post through AVOptions
This avoids having to use a private function.
Signed-off-by: Martin Storsjö <martin@martin.st>
Carl Eugen Hoyos [Thu, 3 Nov 2011 19:20:58 +0000 (20:20 +0100)]
movdec: Set frame_size for AMR
Earlier, sc->samples_per_frame was used for setting the frame size,
but all files don't have that set properly. The frame size is a
known constant for these codecs.
If frame_size isn't set, the mov/3gp muxer refuses to mux it.
This fixes stream copy of audio from
https://roundup.libav.org/file1248/Video_With_AMR-NB_Audio.3gp
to another 3gp file (roundup issue 2468).
Signed-off-by: Martin Storsjö <martin@martin.st>
Ronald S. Bultje [Sun, 30 Oct 2011 17:25:25 +0000 (10:25 -0700)]
h264_weight: remove duplication functions.
Ronald S. Bultje [Fri, 4 Nov 2011 19:19:29 +0000 (12:19 -0700)]
swscale: align vertical filtersize by 2 on x86.
The vertical scaler handles 2 rows at a time and thus requires
alignment by 2, or else it'll read invalid memory and result in
corrupt output.
Ronald S. Bultje [Sat, 5 Nov 2011 14:01:39 +0000 (07:01 -0700)]
libavfilter: reindent.
Ronald S. Bultje [Sat, 29 Oct 2011 23:17:51 +0000 (16:17 -0700)]
matroskadec: empty blocks are in fact valid.
Ronald S. Bultje [Sat, 29 Oct 2011 23:17:27 +0000 (16:17 -0700)]
avfilter: don't abort() on zero-size allocations.
Ronald S. Bultje [Sat, 29 Oct 2011 20:44:50 +0000 (13:44 -0700)]
h264: improve calculation of codec delay.
Fixes the following conformance suite samples:
HCBP1_HHI_A.264, HCBP2_HHI_A.264, HCMP1_HHI_A.264 (main)
HCHP1_HHI_B.264, HCHP2_HHI_A.264, HCHP3_HHI_A.264 (frext)
Carl Eugen Hoyos [Wed, 19 Oct 2011 09:20:48 +0000 (11:20 +0200)]
movenc: Set a correct packet size for AMR-NB mode 15, "no data"
These packets are valid packets, and consist of 1 byte (which
contains the mode bits).
This had been analyzed and reported by Igor Levin, igor d levin comverse com.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Tue, 1 Nov 2011 11:40:04 +0000 (13:40 +0200)]
avformat: Add functions for doing global network initialization
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sat, 5 Feb 2011 22:20:26 +0000 (00:20 +0200)]
avformat: Add the https protocol
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sat, 5 Feb 2011 22:19:49 +0000 (00:19 +0200)]
avformat: Add the tls protocol, using OpenSSL or gnutls
Note, this protocol doesn't yet check verify the server
certificate against a local database of trusted CA root
certificates.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sat, 5 Feb 2011 23:06:49 +0000 (00:06 +0100)]
avformat: Initialize gnutls in ff_tls_init()
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Tue, 1 Nov 2011 13:23:03 +0000 (15:23 +0200)]
w32threads: Wrap the mutex functions in inline functions returning int
This allows using these wrappers in the gcrypt mutex callbacks.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sat, 5 Feb 2011 22:39:20 +0000 (23:39 +0100)]
configure: Allow linking to the gnutls library
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Fri, 4 Feb 2011 22:25:07 +0000 (00:25 +0200)]
avformat: Add ff_tls_init()/deinit() that initialize OpenSSL
If the application hasn't set up mutex callbacks, we set up
our own using pthreads (or w32pthreads).
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Wed, 2 Feb 2011 22:37:20 +0000 (00:37 +0200)]
configure: Allow linking to openssl
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Sat, 29 Oct 2011 22:20:15 +0000 (01:20 +0300)]
avcodec: Allow locking and unlocking an avformat specific mutex
This extends the lock manager in avcodec to manage two separate
mutexes via the user-specified lock functions.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Mon, 7 Feb 2011 10:59:50 +0000 (12:59 +0200)]
avformat: Split out functions from network.h to a new file, network.c
Signed-off-by: Martin Storsjö <martin@martin.st>
Justin Ruggles [Fri, 4 Nov 2011 02:51:07 +0000 (22:51 -0400)]
binkaudio: expand quant_table to accommodate all possible values
Martin Storsjö [Thu, 3 Nov 2011 12:14:09 +0000 (14:14 +0200)]
libx264: Set the default of the rc_lookahead option to -1
This allows it to use the defaults specified by preset/tune,
without overwriting it with the default value from the
AVCodecContext field.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Thu, 3 Nov 2011 12:12:14 +0000 (14:12 +0200)]
avcodec: Set flags2 default value depending on availability
This makes the code compile when FF_API_X264_GLOBAL_OPTS or
FF_API_LAME_GLOBAL_OPTS is 0.
Signed-off-by: Martin Storsjö <martin@martin.st>
Diego Biurrun [Thu, 3 Nov 2011 20:38:23 +0000 (21:38 +0100)]
configure: declare dependency of h264_vaapi_hwaccel on h264_decoder
Alex Converse [Thu, 3 Nov 2011 22:55:52 +0000 (15:55 -0700)]
vp6: Fix illegal read.
Alex Converse [Thu, 3 Nov 2011 22:54:04 +0000 (15:54 -0700)]
avfilter: Don't copy garbage from the stack when setting up video pictures.
Martin Storsjö [Thu, 3 Nov 2011 09:40:13 +0000 (11:40 +0200)]
avcodec: Make sure codec_type is set by avcodec_get_context_defaults2
This function used to set codec_type. With the current fallback
implementation based on avcodec_get_context_defaults3, codec_type
won't be set to the value passed in, but will be set to
AVMEDIA_TYPE_UNKNOWN. Legacy callers of this function might expect
this field to be set to the value passed in.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Wed, 2 Nov 2011 15:54:00 +0000 (17:54 +0200)]
avcodec: Remove a misplaced and useless attribute_deprecated
If attribute_deprecated is used in an enum declaration, it
should follow the 'enum' keyword, otherwise it's ignored
silently. This is the only case of attribute_deprecated for
enum declarations currently.
Currently, this attribute_deprecated doesn't have any effect.
If moved to the right place, it emits a warning every single
time avcodec.h is included, like this:
avcodec.h:2827: warning: ‘AVLPCType’ is deprecated (declared at avcodec.h:543)
There is already a working attribute_deprecated for the
corresponding field in AVCodecContext, so therefore this
one shouldn't be needed.
Signed-off-by: Martin Storsjö <martin@martin.st>
Anton Khirnov [Wed, 2 Nov 2011 20:50:54 +0000 (21:50 +0100)]
avconv: add -dump_attachment option.
Anton Khirnov [Sun, 30 Oct 2011 06:10:08 +0000 (07:10 +0100)]
avconv: add -attach option.
It allows attaching arbitrary files, e.g. fonts to Matroska files.
Anton Khirnov [Sat, 29 Oct 2011 05:58:23 +0000 (07:58 +0200)]
avconv: make negative mappings disable only streams from the specified file
Justin Ruggles [Tue, 1 Nov 2011 20:57:41 +0000 (16:57 -0400)]
fmtconvert: fix int32_to_float_fmul_scalar() for windows x86_64
The calling convention only allows 4 non-stack parameter, with each
float or int register being skipped if not used.
fixes Bug 64
Maxim Poliakovski [Wed, 2 Nov 2011 20:34:57 +0000 (21:34 +0100)]
replacement Indeo 3 decoder
The new decoder is much smaller and has better code quality.
Cleanup and fixes courtesy of Kostya Shishkov.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Justin Ruggles [Wed, 2 Nov 2011 22:30:49 +0000 (18:30 -0400)]
gsm demuxer: do not allocate packet twice.
fixes memleak with raw gsm demuxing.
Justin Ruggles [Wed, 26 Oct 2011 16:43:18 +0000 (12:43 -0400)]
flvenc: use first packet delay as global delay.
This keeps the streams sychronized. The packets must be interleaved per-DTS.
Justin Ruggles [Wed, 2 Nov 2011 14:38:10 +0000 (10:38 -0400)]
ac3enc: doxygen update.
Add some parameters to existing function documentation.
Remove some unneeded documentation.
Convert some static function documentation to non-doxygen style.
Justin Ruggles [Sat, 29 Oct 2011 15:16:01 +0000 (11:16 -0400)]
imc: return error codes instead of 0 for error conditions.
This fixes a bug where the whole buffer was returned as decoded audio due to
*data_size not being set to zero and the return value being >= 0.
Justin Ruggles [Fri, 28 Oct 2011 22:36:28 +0000 (18:36 -0400)]
imc: return meaningful error codes instead of -1
Justin Ruggles [Fri, 28 Oct 2011 22:32:40 +0000 (18:32 -0400)]
imc: do not set channel layout for stereo
we only support decoding of mono imc
Justin Ruggles [Fri, 28 Oct 2011 22:31:21 +0000 (18:31 -0400)]
imc: validate channel count
ask for a sample if not mono
Justin Ruggles [Fri, 28 Oct 2011 22:25:49 +0000 (18:25 -0400)]
imc: check for ff_fft_init() failure
Justin Ruggles [Fri, 28 Oct 2011 22:24:03 +0000 (18:24 -0400)]
imc: check output buffer size before decoding
Justin Ruggles [Fri, 28 Oct 2011 22:16:31 +0000 (18:16 -0400)]
imc: use DSPContext.bswap16_buf() to byte-swap packet data
John Brooks [Sat, 29 Oct 2011 21:41:24 +0000 (15:41 -0600)]
rtsp: add allowed_media_types option
Streams from RTSP or SDP that do not match an allowed type will
be skipped entirely, which allows video-only or audio-only
streaming from servers that provide both.
Signed-off-by: Martin Storsjö <martin@martin.st>
Justin Ruggles [Fri, 28 Oct 2011 14:35:15 +0000 (10:35 -0400)]
libgsm: add flush function to reset the decoder state when seeking
Justin Ruggles [Fri, 28 Oct 2011 14:28:41 +0000 (10:28 -0400)]
libgsm: simplify decoding by using a loop
Justin Ruggles [Fri, 28 Oct 2011 14:18:07 +0000 (10:18 -0400)]
gsm: log error message when packet is too small