Commit Graph

35129 Commits

Author SHA1 Message Date
Michael Niedermayer
c20a696306 cavsdec: check for changing w/h.
Our decoder does not support changing w/h.

Fixes CVE-2012-2777 and CVE-2012-2784.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Anton Khirnov <anton@khirnov.net>
2012-09-29 09:26:04 +02:00
Anton Khirnov
12e1e83461 lavc: set channel count from channel layout in avcodec_open2().
Some decoders (e.g. nellymoser) only set channel_layout and do not set
channel count.
2012-09-29 09:25:40 +02:00
Derek Buitenhuis
7d1d446990 doc/platform: Rework the Visual Studio linking section
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:52:21 -04:00
Derek Buitenhuis
ed8a2ddeea doc/faq: Change the Visual Studio entry to reflect current status
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:50:45 -04:00
Derek Buitenhuis
f45b54437a doc/platform: Replace Visual Studio section with build instructions
Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:50:32 -04:00
Derek Buitenhuis
3e071551af doc/platform: Nuke section on linking static MinGW-built libs with MSVC
This practice is not supported by the MinGW developers, and even requires
patching the MinGW runtimes in newer versions. Furthermore, we now support
build with MSVC, so this section is rendered useless.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:27:59 -04:00
Derek Buitenhuis
55254a3f7f doc/platform: Remove false claim about MinGW installer
This works just fine, and has for quite a while, using the mingw-get
installer.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:27:25 -04:00
Derek Buitenhuis
6a3078bb89 doc/platform: Mention MinGW-w64
This is the toolchain most, including Libav's nightlies, use to build.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-28 15:27:13 -04:00
Michael Niedermayer
791b5954bc dsputil_mmx: fix reading prior of the src array in sub_hfyu_median_prediction()
This should fix the utvideoenc valgrind failure

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-09-28 12:25:07 -04:00
Luca Barbato
97cfa55eea mpegaudiodec: fix short_start calculation
The value should be always 3, as it follows from the specification.

Fix a stack buffer overflow in exponents_from_scale_factors as reported
by asan. Thanks to Dale Curtis for the sample vector.
2012-09-28 16:02:39 +02:00
Nathan Caldwell
d16860a237 libopus: Remap channels using libopus' internal remapping.
This way we can directly remap channels from Opus' channel order to
libav's internal channel order, instead of mapping channels from
Opus' order to Vorbis' order then to libav's order.

Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-28 11:03:32 +02:00
Nicolas George
44617d6ec9 Opus decoder using libopus
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-28 11:03:20 +02:00
Nathan Caldwell
e4aa3831b7 avcodec: document the use of AVCodecContext.delay for audio decoding
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-28 10:50:07 +02:00
Kostya Shishkov
4dc8c8386e vc1dec: add flush function for WMV9 and VC-1 decoders
CC: libav-stable@libav.org
2012-09-28 07:08:54 +02:00
Duncan Salerno
8bd324e9e4 http: Increase buffer sizes to cope with longer URIs
Use the MAX_URL_SIZE define where applicable. Increase buffer
sizes for all buffers that need to fit a long pathname - buffers
that need to fit only the hostname (and other short strings, but
not the pathname - such as "headers" in http_connect) are kept
at 1024 bytes for now.

Also increase the max line length in http_read_header, since it
might need to contain a full url for Location: redirects.

Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-28 00:09:38 +03:00
Diego Biurrun
bc66827fb1 nutenc: const correctness for ff_put_v_trace/put_s_trace function arguments 2012-09-27 19:32:47 +02:00
Diego Biurrun
6c5b0517e0 h264_refs: Fix debug tprintf argument types 2012-09-27 19:10:10 +02:00
Diego Biurrun
80412997c8 golomb: const correctness for get_ue()/get_se() function arguments 2012-09-27 19:10:10 +02:00
Diego Biurrun
e00215040a get_bits: const correctness for get_bits_trace()/get_xbits_trace() arguments 2012-09-27 19:10:10 +02:00
Diego Biurrun
4e5b7f31f7 dwt: Drop unused functions spatial_compose{53|97}i() 2012-09-27 12:08:09 +02:00
Diego Biurrun
83655442fa nutdec: Remove unused and broken debug function stub 2012-09-27 12:02:33 +02:00
Diego Biurrun
0ae3ba8daa avcodec: Drop long-deprecated imgconvert.h header 2012-09-27 12:02:33 +02:00
Nathan Caldwell
bcc1f7caeb Add Opus support to the Ogg muxer.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-27 10:48:35 +02:00
Andrew D'Addesio
2a3d82ab46 Add Opus codec id and codec description.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-27 10:48:35 +02:00
Dale Curtis
0575772fa0 avformat: Identify anonymous AVIO typedef structs.
Anonymous typedef structs prevent forward declaration, this
change gives the AVIOContext and AVIOInterruptCB structures
a name.  These structures are now in line with other common
structures such as AVFormatContext and AVCodecContext.

Signed-off-by: Dale Curtis <dalecurtis@chromium.org>
Signed-off-by: Diego Biurrun <diego@biurrun.de>
2012-09-27 09:14:22 +02:00
Diego Biurrun
93e81ee81c nutdec: const correctness for get_v_trace/get_s_trace function arguments 2012-09-26 19:10:13 +02:00
Diego Biurrun
2dc1310484 truemotion2: Request samples for old TM2 headers 2012-09-26 19:10:13 +02:00
Martin Storsjö
69673138c5 rtpdec: Remove a useless ff_ prefix from a static symbol
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-26 19:05:18 +03:00
Dmitry Samonenko
b6bf1490da rtpdec: Support depacketizing speex
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-26 19:05:10 +03:00
Dmitry Samonenko
490ae95aa8 rtpenc: Add support for packetizing speex
This packetization scheme simply places the full packets into the
RTP packet without any extra header bytes.

Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-26 19:04:57 +03:00
Alexandre Colucci
92d2b909a0 xsub: feed init_get_bits the whole buffer
Do not use rlelen field for buffer size in init_get_bits, it is
only the size of the data for the first field.
Since it is not reliable, just use the size of the whole buffer.

Additional comments add removal of unused rlelen variable by
Reimar Döffinger.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
2012-09-25 22:28:53 +02:00
Martin Storsjö
375bff7084 libfdk-aac: Allow setting VBR modes via a private option
This avoids using the global_quality field and QSCALE flag for
passing the VBR modes, since the value range of the global_quality
field doesn't really map cleanly to this codec's VBR modes.

Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-25 20:51:47 +03:00
Martin Storsjö
a10190dcd8 libfdk-aac: Warn the user that the VBR modes are unsupported
These modes were not originally exposed by the library at all.
In practice, only a few of them work for each sample rate/profile
combination, and they don't work at all for the more uncommon
sample rates.

Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-25 20:51:46 +03:00
Derek Buitenhuis
b62d4c16c9 Revert "cbrt_tablegen: Include libm.h"
This code runs on the host and should not use libm.h, which is meant
for the target.

This reverts commit 2768b71798.
2012-09-25 12:54:26 -04:00
Anton Khirnov
7a1a9dd56c pthread: make sure AVFrame.extended_data is set properly.
Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
2012-09-25 12:03:23 +02:00
Nathan Caldwell
c8ba8be27e libfdk-aac: reindent after last commit
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-25 08:08:05 +03:00
Nathan Caldwell
94a69dee64 libfdk-aac: Limit to supported sample rates.
Signed-off-by: Martin Storsjö <martin@martin.st>
2012-09-25 08:08:05 +03:00
Derek Buitenhuis
2768b71798 cbrt_tablegen: Include libm.h
Needed for cbrtf fallback on systems which lack it.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
2012-09-24 22:47:47 -04:00
Luca Barbato
0336dea2ef oggparsetheora: make it more robust 2012-09-24 22:35:29 +02:00
Michael Niedermayer
bb146bb57b ogg: prevent NULL pointer deference in theora gptopts
Additional safety in case a special ogg stream is crafted
with the proper number of

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
2012-09-24 22:35:29 +02:00
Luca Barbato
d1f05dd183 ogg: calculate the start position once all the headers are parsed
The fisbone packets can be muxed in any order as long the last one
comes before the first data packet.
2012-09-24 22:35:29 +02:00
Luca Barbato
7751e4693d ogg: check that the expected number of headers had been parsed
Not having the header for a codec is a tell-tale of a broken file.
2012-09-24 22:35:29 +02:00
John Van Sickle
a716006a7d libx264: change default to closed gop to match x264cli
open-gop can be enabled with "-flags -cgop"

Signed-off-by: Anton Khirnov <anton@khirnov.net>
2012-09-24 19:23:03 +02:00
Anton Khirnov
11d1ca4b2c Use avcodec_free_frame() to free AVFrames. 2012-09-24 12:31:25 +02:00
Anton Khirnov
9eb296572e lavf: use a malloced AVFrame in try_decode_frame().
This allows using avcodec_free_frame() to free it properly.
2012-09-24 12:31:24 +02:00
Anton Khirnov
a42aadabc6 lavc: add avcodec_free_frame().
Since an AVFrame now has malloced members (extended_data), it must have
a destructor.
2012-09-24 12:31:24 +02:00
Anton Khirnov
b437cec143 lavc: ensure extended_data is set properly on decoding 2012-09-24 12:31:24 +02:00
Anton Khirnov
2bc0de3858 lavc: initialize AVFrame.extended_data in avcodec_get_frame_defaults() 2012-09-24 12:31:24 +02:00
Anton Khirnov
c084a975aa lavc: use av_mallocz to allocate AVFrames.
Otherwise the frame is uninitialized, so avcodec_get_frame_defaults()
cannot determine whether to free extended_data.
2012-09-24 12:31:24 +02:00
Anton Khirnov
a83499b13b lavc: rename the argument of avcodec_alloc_frame/get_frame_defaults
AVFrame is used for both audio and video, so calling the argument 'pic'
is misleading.
2012-09-24 12:31:24 +02:00