ffmpeg.git
8 years agoMerge remote-tracking branch 'qatar/master'
Michael Niedermayer [Tue, 4 Oct 2011 01:21:14 +0000 (03:21 +0200)]
Merge remote-tracking branch 'qatar/master'

* qatar/master:
  id3v2: remove pointless casts
  id3v2: read TXXX frames with two calls to decode_str() instead of one.
  id3v2: don't discard the whole tag when encountering empty frames.
  libvpx: fix build with older libvpx versions.
  ARM: check for inline asm 'y' operand modifier support

Conflicts:
libavcodec/libvpxenc.c
libavformat/id3v2.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
8 years agolibvpxenc: support vpxencs command line options.
Luca Barbato [Tue, 4 Oct 2011 01:01:06 +0000 (03:01 +0200)]
libvpxenc: support vpxencs command line options.
This simplifies useage for users who are used to vpxenc.
We continue to support the previous options with the exception of the flags
for which it would be slightly messy.

8 years agodeshake: nicer looking defaults
Michael Niedermayer [Tue, 4 Oct 2011 00:15:18 +0000 (02:15 +0200)]
deshake: nicer looking defaults

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agolibavfilter: add deshake filter
danielgtaylor [Mon, 3 Oct 2011 23:31:18 +0000 (01:31 +0200)]
libavfilter: add deshake filter

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agolibavfilter: image transform code
danielgtaylor [Tue, 4 Oct 2011 00:10:18 +0000 (02:10 +0200)]
libavfilter: image transform code

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoH264: hotfix for speedloss on frame threading and h264 files with slices.
Michael Niedermayer [Mon, 3 Oct 2011 22:14:48 +0000 (00:14 +0200)]
H264: hotfix for speedloss on frame threading and h264 files with slices.

This fix is not ideal as it still limits the multithreading on field pictures
to the 2nd field only.
Ill try to fix it properly to allow both fields to decode concurrently but this
needs more work.

This bug exists since and was caused by:
commit ea6331f8bbaf3343faec6ffe77a6218a016a3af5
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Mon Jun 20 10:24:33 2011 -0400

    h264-mt: fix deadlock in packets with multiple slices (e.g. MP4).

8 years agothreads: warn in case of multiple finished setup calls.
Michael Niedermayer [Mon, 3 Oct 2011 19:25:58 +0000 (21:25 +0200)]
threads: warn in case of multiple finished setup calls.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264_parser: print AU size on error.
Michael Niedermayer [Mon, 3 Oct 2011 18:40:17 +0000 (20:40 +0200)]
h264_parser: print AU size on error.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264_parser: AVC support.
Michael Niedermayer [Mon, 3 Oct 2011 18:25:43 +0000 (20:25 +0200)]
h264_parser: AVC support.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agofate: fix due to new yuv444 formats
Michael Niedermayer [Mon, 3 Oct 2011 21:30:02 +0000 (23:30 +0200)]
fate: fix due to new yuv444 formats

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoRELEASE: 0.8.5 has been released
Dominique Leuenberger [Mon, 3 Oct 2011 07:09:57 +0000 (09:09 +0200)]
RELEASE: 0.8.5 has been released

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoEnable conversions to YUV444P9/YUV444P10, they seem to work fine.
Reimar Döffinger [Mon, 3 Oct 2011 19:30:58 +0000 (21:30 +0200)]
Enable conversions to YUV444P9/YUV444P10, they seem to work fine.

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
8 years agolavf/utils: fix overestimation of the rational number density.
Michael Niedermayer [Mon, 3 Oct 2011 14:32:17 +0000 (16:32 +0200)]
lavf/utils: fix overestimation of the rational number density.
Fixes Ticket498

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agompegtsenc: Lift limit on PMT PID
Loren Osborn [Mon, 3 Oct 2011 14:03:29 +0000 (16:03 +0200)]
mpegtsenc: Lift limit on PMT PID
Fixes Ticket518

8 years agolibcdio: Fix missing avclass in context
chinshou [Mon, 3 Oct 2011 13:40:40 +0000 (15:40 +0200)]
libcdio: Fix missing avclass in context
Fixes Ticket526

8 years agoid3v2: remove pointless casts
Anton Khirnov [Sun, 2 Oct 2011 07:14:30 +0000 (09:14 +0200)]
id3v2: remove pointless casts

8 years agoid3v2: read TXXX frames with two calls to decode_str() instead of one.
Anton Khirnov [Sun, 2 Oct 2011 07:06:34 +0000 (09:06 +0200)]
id3v2: read TXXX frames with two calls to decode_str() instead of one.

Read the key in the first, value in the second.

This allows to avoid pointless strdups and simplify decode_str() by
dropping two of its parameters.

8 years agoid3v2: don't discard the whole tag when encountering empty frames.
Anton Khirnov [Sun, 2 Oct 2011 05:16:09 +0000 (07:16 +0200)]
id3v2: don't discard the whole tag when encountering empty frames.

While they're technically invalid, it's better to skip them and try to
read the rest of the tag.

8 years agolibvpx: fix build with older libvpx versions.
Anton Khirnov [Sun, 2 Oct 2011 14:10:53 +0000 (16:10 +0200)]
libvpx: fix build with older libvpx versions.

VPX_ERROR_RESILIENT_DEFAULT and VPX_ERROR_RESILIENT_PARTITIONS weren't
defined before 4cb0ebe5b27d35ccc2a78c1d16f2622ddef21f74 (CommitDate: Tue
Jun 28 11:10:17 2011)

8 years agoARM: check for inline asm 'y' operand modifier support
Mans Rullgard [Sun, 2 Oct 2011 22:38:32 +0000 (23:38 +0100)]
ARM: check for inline asm 'y' operand modifier support

The inline asm added in bf5d46d uses the 'y' modifier which
is only supported from gcc 4.5.  This check allows building
with older compilers.

Signed-off-by: Mans Rullgard <mans@mansr.com>
8 years agoVP8: armv6 optimizations.
Ronald S. Bultje [Wed, 24 Aug 2011 20:58:37 +0000 (13:58 -0700)]
VP8: armv6 optimizations.

From 52.503s (~40fps) to 27.973sec (~80fps) decoding of 480p sintel
trailer, i.e. a ~2x speedup overall, on a Nexus S.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoavutil: bump for av_get_default_channel_layout()
Michael Niedermayer [Sun, 2 Oct 2011 23:17:42 +0000 (01:17 +0200)]
avutil: bump for av_get_default_channel_layout()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoMerge remote-tracking branch 'qatar/master'
Michael Niedermayer [Sun, 2 Oct 2011 23:02:45 +0000 (01:02 +0200)]
Merge remote-tracking branch 'qatar/master'

* qatar/master:
  mpc8: check output buffer size before decoding
  mpc7: return error if packet is too small.
  mpc7: check output buffer size before decoding
  nellymoserdec: allocate float_buf only when decoding to int16
  nellymoserdec: use NELLY_BUF_LEN instead of 128
  nellymoserdec: use NELLY_BLOCK_LEN instead of 64 when appropriate.
  nellymoserdec: allow user to request SAMPLE_FMT_FLT for output samples.
  nellymoser: check output buffer size before decoding
  win32: improve threading algorithm warning

Conflicts:
libavcodec/nellymoserdec.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoMAINTAINERS: new ffplay maintainer
Michael Niedermayer [Sun, 2 Oct 2011 22:38:50 +0000 (00:38 +0200)]
MAINTAINERS: new ffplay maintainer

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoMerge remote-tracking branch 'cus/stable'
Michael Niedermayer [Sun, 2 Oct 2011 21:53:21 +0000 (23:53 +0200)]
Merge remote-tracking branch 'cus/stable'

* cus/stable:
  ffplay: use libswresample instead of av_audio_convert
  audioconvert: add av_get_default_channel_layout public function
  ffplay: use avctx->channels and avctx->freq before avcodec_open2 consistently
  ffplay: remove now unnecessary request_channels, we set it now with options
  ffplay: set request_channels to 2

Merged-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoriff: map 0x0038 to amrnb, works on http://video.mopoto.com/4/40/407/40709.avi
Compn [Sun, 2 Oct 2011 21:39:03 +0000 (17:39 -0400)]
riff: map 0x0038 to amrnb, works on video.mopoto.com/4/40/407/40709.avi

8 years agoffplay: use libswresample instead of av_audio_convert
Marton Balint [Tue, 16 Aug 2011 23:28:23 +0000 (01:28 +0200)]
ffplay: use libswresample instead of av_audio_convert

Previously ffplay expected SDL_AudioOpen to provide the requested sample rate
and channel number. This is no longer a requirement because this patch replaces
the audio convert function with libswresample's swr_convert which is capable of
handling different sample formats, sample rates and different number of
channels and different channel layouts.

The patch also removes the hardcoded 16bit samples assumption and uses
av_get_bytes_per_sample almost everywhere. The only exceptions are
the update_sample_display and video_audio_display functions, it
seemed too much of a headache to make them generic.

We also fix a tiny bug in sdl_audio_callback, we ensure that the number of
bytes when we put silence in the buffer is a multiple of the frame size.

8 years agoaudioconvert: add av_get_default_channel_layout public function
Marton Balint [Sun, 2 Oct 2011 14:28:36 +0000 (16:28 +0200)]
audioconvert: add av_get_default_channel_layout public function

8 years agoffplay: use avctx->channels and avctx->freq before avcodec_open2 consistently
Marton Balint [Sun, 2 Oct 2011 09:00:40 +0000 (11:00 +0200)]
ffplay: use avctx->channels and avctx->freq before avcodec_open2 consistently

8 years agoffplay: remove now unnecessary request_channels, we set it now with options
Marton Balint [Sun, 2 Oct 2011 10:05:20 +0000 (12:05 +0200)]
ffplay: remove now unnecessary request_channels, we set it now with options

8 years agoffplay: set request_channels to 2
Michael Niedermayer [Sat, 1 Oct 2011 23:25:14 +0000 (01:25 +0200)]
ffplay: set request_channels to 2

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: fix intra 16x16 mode check when using mbaff and constrained_intra_pred.
Laurent Aimar [Sun, 2 Oct 2011 19:07:41 +0000 (21:07 +0200)]
h264: fix intra 16x16 mode check when using mbaff and constrained_intra_pred.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agolibvpxenc: cosmetics to make code more similar to lucas.
Michael Niedermayer [Sun, 2 Oct 2011 18:38:38 +0000 (20:38 +0200)]
libvpxenc: cosmetics to make code more similar to lucas.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: check for invalid bit depth value.
Laurent Aimar [Sun, 2 Oct 2011 16:29:31 +0000 (18:29 +0200)]
h264: check for invalid bit depth value.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: add entries for 11 and 12 bits in ff_h264_chroma_qp[][]
Laurent Aimar [Sun, 2 Oct 2011 16:29:30 +0000 (18:29 +0200)]
h264: add entries for 11 and 12 bits in ff_h264_chroma_qp[][]

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: fix the check for invalid SPS:num_ref_frames.
Laurent Aimar [Sun, 2 Oct 2011 14:06:38 +0000 (16:06 +0200)]
h264: fix the check for invalid SPS:num_ref_frames.

This patch set the limit to 16.

For information, thoses previous commits:
41f7e2d11d2dca23842ee89d530ca9fa15cec9d8
5cbb0e70a0a2ee99eb3cb09e837b9a1f7355b9bc
assumed it was either 30 or 32.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: do not let invalid values in h->ref_count on ff_h264_decode_ref_pic_list_reorde...
Laurent Aimar [Sun, 2 Oct 2011 14:06:37 +0000 (16:06 +0200)]
h264: do not let invalid values in h->ref_count on ff_h264_decode_ref_pic_list_reordering() errors.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoReject video with non multiple of 16 width/height in the 4xm decoder.
Laurent Aimar [Sat, 1 Oct 2011 00:04:45 +0000 (02:04 +0200)]
Reject video with non multiple of 16 width/height in the 4xm decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agompc8: check output buffer size before decoding
Justin Ruggles [Wed, 14 Sep 2011 15:39:21 +0000 (11:39 -0400)]
mpc8: check output buffer size before decoding

8 years agompc7: return error if packet is too small.
Justin Ruggles [Wed, 14 Sep 2011 15:16:42 +0000 (11:16 -0400)]
mpc7: return error if packet is too small.

8 years agompc7: check output buffer size before decoding
Justin Ruggles [Tue, 13 Sep 2011 22:53:18 +0000 (18:53 -0400)]
mpc7: check output buffer size before decoding

8 years agonellymoserdec: allocate float_buf only when decoding to int16
Justin Ruggles [Wed, 14 Sep 2011 16:37:01 +0000 (12:37 -0400)]
nellymoserdec: allocate float_buf only when decoding to int16

8 years agonellymoserdec: use NELLY_BUF_LEN instead of 128
Justin Ruggles [Wed, 14 Sep 2011 16:28:41 +0000 (12:28 -0400)]
nellymoserdec: use NELLY_BUF_LEN instead of 128

8 years agonellymoserdec: use NELLY_BLOCK_LEN instead of 64 when appropriate.
Justin Ruggles [Wed, 14 Sep 2011 16:27:12 +0000 (12:27 -0400)]
nellymoserdec: use NELLY_BLOCK_LEN instead of 64 when appropriate.

8 years agonellymoserdec: allow user to request SAMPLE_FMT_FLT for output samples.
Justin Ruggles [Wed, 14 Sep 2011 16:25:00 +0000 (12:25 -0400)]
nellymoserdec: allow user to request SAMPLE_FMT_FLT for output samples.

8 years agonellymoser: check output buffer size before decoding
Justin Ruggles [Wed, 14 Sep 2011 16:16:05 +0000 (12:16 -0400)]
nellymoser: check output buffer size before decoding

8 years agovf_lut.c: fix wrong syntax in example
Stefano Sabatini [Sun, 2 Oct 2011 13:21:12 +0000 (15:21 +0200)]
vf_lut.c: fix wrong syntax in example

8 years agolavfi: remove usage of deprecated av_opt_set_defaults2() function
Stefano Sabatini [Sun, 2 Oct 2011 13:16:36 +0000 (15:16 +0200)]
lavfi: remove usage of deprecated av_opt_set_defaults2() function

Prefer av_opt_set_defaults() instead.

8 years agowin32: improve threading algorithm warning
Luca Barbato [Sun, 25 Sep 2011 22:51:14 +0000 (00:51 +0200)]
win32: improve threading algorithm warning

If no threading is requested do not issue warning about unsupported
threading algorithm.

8 years agoMerge remote-tracking branch 'qatar/master'
Michael Niedermayer [Sun, 2 Oct 2011 02:41:22 +0000 (04:41 +0200)]
Merge remote-tracking branch 'qatar/master'

* qatar/master:
  lavc: add video/audio/encoding flags to global_quality option
  libvpxenc: use libvpx's own defaults for some parameters
  vpxenc: add private options

Conflicts:
libavcodec/libvpxenc.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
8 years ago4xm decoder: fix data size for i2 frames.
Michael Niedermayer [Sun, 2 Oct 2011 01:40:26 +0000 (03:40 +0200)]
4xm decoder: fix data size for i2 frames.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years ago4xm decoder: print some error messages in case of errors.
Michael Niedermayer [Sun, 2 Oct 2011 01:39:59 +0000 (03:39 +0200)]
4xm decoder: print some error messages in case of errors.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound accesses in the 4xm decoder.
Laurent Aimar [Sat, 1 Oct 2011 22:38:27 +0000 (00:38 +0200)]
Check for out of bound accesses in the 4xm decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoPrevent block size from inreasing in the shorten decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:43:05 +0000 (00:43 +0200)]
Prevent block size from inreasing in the shorten decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound reads in PTX decoder.
Laurent Aimar [Sat, 1 Oct 2011 22:38:32 +0000 (00:38 +0200)]
Check for out of bound reads in PTX decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix the check for missing references in ff_er_frame_end() for H264.
Laurent Aimar [Sat, 1 Oct 2011 22:38:31 +0000 (00:38 +0200)]
Fix the check for missing references in ff_er_frame_end() for H264.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoClear FF_INPUT_BUFFER_PADDING_SIZE bytes at the end of the temporary buffers used...
Laurent Aimar [Sat, 1 Oct 2011 22:38:30 +0000 (00:38 +0200)]
Clear FF_INPUT_BUFFER_PADDING_SIZE bytes at the end of the temporary buffers used in 4xm decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoPrevent NULL dereference when the huffman table is invalid in the 4xm decoder.
Laurent Aimar [Sat, 1 Oct 2011 22:38:28 +0000 (00:38 +0200)]
Prevent NULL dereference when the huffman table is invalid in the 4xm decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix use of uninitialized memory in 4X Technologies demuxer.
Laurent Aimar [Sat, 1 Oct 2011 22:38:26 +0000 (00:38 +0200)]
Fix use of uninitialized memory in 4X Technologies demuxer.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: increase ref_poc size to 32 as it can be per field.
Michael Niedermayer [Sat, 1 Oct 2011 22:47:54 +0000 (00:47 +0200)]
h264: increase ref_poc size to 32 as it can be per field.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: set unused ref_counts to 0 as a precautionary meassure.
Michael Niedermayer [Sat, 1 Oct 2011 22:45:41 +0000 (00:45 +0200)]
h264: set unused ref_counts to 0 as a precautionary meassure.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoff_dot_product(): change return value to int64_t
Michael Niedermayer [Sat, 1 Oct 2011 21:09:48 +0000 (23:09 +0200)]
ff_dot_product(): change return value to int64_t

Idea-by: Vitor
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agog729dec: initialize bit reader with the correct buffer size
Vitor Sessak [Sat, 1 Oct 2011 08:47:11 +0000 (10:47 +0200)]
g729dec: initialize bit reader with the correct buffer size

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agog723.1_data: remove unneeded L
Michael Niedermayer [Sat, 1 Oct 2011 16:13:30 +0000 (18:13 +0200)]
g723.1_data: remove unneeded L

Found-by: Vitor
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoh264: try reading SPS with wrong escaping, it sometimes works better
Michael Niedermayer [Sat, 1 Oct 2011 15:41:28 +0000 (17:41 +0200)]
h264: try reading SPS with wrong escaping, it sometimes works better

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agolavc: add video/audio/encoding flags to global_quality option
Anton Khirnov [Fri, 30 Sep 2011 15:12:13 +0000 (17:12 +0200)]
lavc: add video/audio/encoding flags to global_quality option

8 years agolibvpxenc: use libvpx's own defaults for some parameters
Luca Barbato [Fri, 30 Sep 2011 10:15:59 +0000 (12:15 +0200)]
libvpxenc: use libvpx's own defaults for some parameters

Specifically, qmin/qmax, gop_size and keyint_min.

Fixes bug 47.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
8 years agovpxenc: add private options
Luca Barbato [Mon, 26 Sep 2011 08:22:38 +0000 (10:22 +0200)]
vpxenc: add private options

Make libvpx support close to the libx264 one. Thanks to
Jan Gerber <j@v2v.cc> for the support.

Signed-off-by: Anton Khirnov <anton@khirnov.net>
8 years agog729dec: prevent out of array access
Michael Niedermayer [Sat, 1 Oct 2011 04:07:05 +0000 (06:07 +0200)]
g729dec: prevent out of array access

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoavienc: Add a limit on the number of skiped frames muxed in a row.
Michael Niedermayer [Sat, 1 Oct 2011 01:48:33 +0000 (03:48 +0200)]
avienc: Add a limit on the number of skiped frames muxed in a row.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agovf_scale.c: propagate error code
Michael Niedermayer [Sat, 1 Oct 2011 01:16:32 +0000 (03:16 +0200)]
vf_scale.c: propagate error code

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoMerge remote-tracking branch 'qatar/master'
Michael Niedermayer [Sat, 1 Oct 2011 00:54:46 +0000 (02:54 +0200)]
Merge remote-tracking branch 'qatar/master'

* qatar/master: (34 commits)
  dpcm: return error if packet is too small
  dpcm: use smaller data types for static tables
  dpcm: use sol_table_16 directly instead of through the DPCMContext.
  dpcm: replace short with int16_t
  dpcm: check to make sure channels is 1 or 2.
  dpcm: misc pretty-printing
  dpcm: remove unnecessary variable by using bytestream functions.
  dpcm: move codec-specific variable declarations to their corresponding decoding blocks.
  dpcm: consistently use the variable name 'n' for the next input byte.
  dpcm: output AV_SAMPLE_FMT_U8 for Sol DPCM subcodecs 1 and 2.
  dpcm: calculate and check actual output data size prior to decoding.
  dpcm: factor out the stereo flag calculation
  dpcm: cosmetics: rename channel_number to ch
  avserver: Fix a bug where the socket is IPv4, but IPv6 is autoselected for the loopback address.
  lavf: Avoid using av_malloc(0) in av_dump_format
  dxva2_h264: pass the correct 8x8 scaling lists
  dca: NEON optimised high freq VQ decoding
  avcodec: reject audio packets with NULL data and non-zero size
  dxva: Add ability to enable workaround for older ATI cards
  latmenc: Set latmBufferFullness to largest value to indicate it is not used
  ...

Conflicts:
libavcodec/dxva2_h264.c

Merged-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix out of bound reads/writes in the TIFF decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:45:03 +0000 (00:45 +0200)]
Fix out of bound reads/writes in the TIFF decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound writes in the QDM2 decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:45:05 +0000 (00:45 +0200)]
Check for out of bound writes in the QDM2 decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix out of bound reads in the QDM2 decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:45:04 +0000 (00:45 +0200)]
Fix out of bound reads in the QDM2 decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix out of bound reads due to integer overflow in the ADPCM IMA Electronic Arts EACS...
Laurent Aimar [Fri, 30 Sep 2011 22:45:02 +0000 (00:45 +0200)]
Fix out of bound reads due to integer overflow in the ADPCM IMA Electronic Arts EACS decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound reads in the Electronic Arts CMV decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:45:01 +0000 (00:45 +0200)]
Check for out of bound reads in the Electronic Arts CMV decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoPrevent NULL dereferences when missing the reference frame in the Electronic Arts...
Laurent Aimar [Fri, 30 Sep 2011 22:45:00 +0000 (00:45 +0200)]
Prevent NULL dereferences when missing the reference frame in the Electronic Arts CMV decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix potential pointer arithmetic overflows in the Electronic Arts CMV decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:44:59 +0000 (00:44 +0200)]
Fix potential pointer arithmetic overflows in the Electronic Arts CMV decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoPrevent infinite loop in the ANM decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:44:58 +0000 (00:44 +0200)]
Prevent infinite loop in the ANM decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoFix double free on error in Deluxe Paint Animation demuxer.
Laurent Aimar [Fri, 30 Sep 2011 22:44:57 +0000 (00:44 +0200)]
Fix double free on error in Deluxe Paint Animation demuxer.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound reads in AVS decoder.
Laurent Aimar [Fri, 30 Sep 2011 22:44:56 +0000 (00:44 +0200)]
Check for out of bound reads in AVS decoder.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for out of bound writes in the avs demuxer.
Laurent Aimar [Fri, 30 Sep 2011 22:44:55 +0000 (00:44 +0200)]
Check for out of bound writes in the avs demuxer.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoCheck for corrupted data in avs demuxer.
Laurent Aimar [Fri, 30 Sep 2011 22:44:54 +0000 (00:44 +0200)]
Check for corrupted data in avs demuxer.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoswresample: check for invalid sample formats.
Michael Niedermayer [Fri, 30 Sep 2011 23:39:17 +0000 (01:39 +0200)]
swresample: check for invalid sample formats.

Bug-Found-by: Justin Ruggles
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoavformat_free_context: favor av_freep()
Michael Niedermayer [Fri, 30 Sep 2011 23:29:30 +0000 (01:29 +0200)]
avformat_free_context: favor av_freep()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agooggenc: favor av_freep()
Michael Niedermayer [Fri, 30 Sep 2011 23:26:55 +0000 (01:26 +0200)]
oggenc: favor av_freep()

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoswr_test: test planar audio
Michael Niedermayer [Thu, 29 Sep 2011 02:55:53 +0000 (04:55 +0200)]
swr_test: test planar audio

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agoswr: minor fixes to get planar audio working
Michael Niedermayer [Thu, 29 Sep 2011 02:53:50 +0000 (04:53 +0200)]
swr: minor fixes to get planar audio working

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agompegvideo: increase emu edge buffer size
Michael Niedermayer [Fri, 30 Sep 2011 22:58:01 +0000 (00:58 +0200)]
mpegvideo: increase emu edge buffer size
This fixes a crash with 422 H.264

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
8 years agodpcm: return error if packet is too small
Justin Ruggles [Sun, 11 Sep 2011 17:18:51 +0000 (13:18 -0400)]
dpcm: return error if packet is too small

8 years agodpcm: use smaller data types for static tables
Justin Ruggles [Sun, 11 Sep 2011 16:46:06 +0000 (12:46 -0400)]
dpcm: use smaller data types for static tables

8 years agodpcm: use sol_table_16 directly instead of through the DPCMContext.
Justin Ruggles [Sun, 11 Sep 2011 16:43:03 +0000 (12:43 -0400)]
dpcm: use sol_table_16 directly instead of through the DPCMContext.

8 years agodpcm: replace short with int16_t
Justin Ruggles [Sun, 11 Sep 2011 16:41:48 +0000 (12:41 -0400)]
dpcm: replace short with int16_t

8 years agodpcm: check to make sure channels is 1 or 2.
Justin Ruggles [Sun, 11 Sep 2011 16:40:40 +0000 (12:40 -0400)]
dpcm: check to make sure channels is 1 or 2.

8 years agodpcm: misc pretty-printing
Justin Ruggles [Sun, 11 Sep 2011 16:37:17 +0000 (12:37 -0400)]
dpcm: misc pretty-printing

8 years agodpcm: remove unnecessary variable by using bytestream functions.
Justin Ruggles [Sun, 11 Sep 2011 16:24:03 +0000 (12:24 -0400)]
dpcm: remove unnecessary variable by using bytestream functions.

Uses 'buf' directly instead of a separate iterator variable 'in'.

8 years agodpcm: move codec-specific variable declarations to their corresponding decoding blocks.
Justin Ruggles [Sun, 11 Sep 2011 16:13:13 +0000 (12:13 -0400)]
dpcm: move codec-specific variable declarations to their corresponding decoding blocks.

8 years agodpcm: consistently use the variable name 'n' for the next input byte.
Justin Ruggles [Sun, 11 Sep 2011 16:08:38 +0000 (12:08 -0400)]
dpcm: consistently use the variable name 'n' for the next input byte.

8 years agodpcm: output AV_SAMPLE_FMT_U8 for Sol DPCM subcodecs 1 and 2.
Justin Ruggles [Sun, 11 Sep 2011 16:04:46 +0000 (12:04 -0400)]
dpcm: output AV_SAMPLE_FMT_U8 for Sol DPCM subcodecs 1 and 2.

Uses the native sample format for the codec instead of left-shifting all
samples by 8.