avformat/utils: Check cur_dts in update_initial_timestamps() more
Fixes: runtime error: signed integer overflow: 18133149658382192 - -9223090561878065151 cannot be represented in type 'long long' Fixes: crbug 831552 Reported-by: Matt Wolenetz <wolenetz@google.com> Reviewed-by: Matt Wolenetz <wolenetz@google.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avcodec/utils: Enforce minimum width also for VP5/6
Fixes: out of array access Fixes: poc_0411 Found-by: GwanYeong Kim <gy741.kim@gmail.com> Tested-by: GwanYeong Kim <gy741.kim@gmail.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
avformat/utils: Fix integer overflow in end time calculation in update_stream_timings()
Fixes: crbug 829153 Reported-by: Matt Wolenetz <wolenetz@google.com> Reviewed-by: Matt Wolenetz <wolenetz@google.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Bela Bodecs [Fri, 13 Apr 2018 10:11:32 +0000 (12:11 +0200)]
avformat/utils: Stream specifier enhancement 2.
In some cases, mainly working with multiprogram mpeg-ts containers as
input, it would be handy to select sub stream of a specific program by
their metadata.
This patch makes it possible to narrow the stream selection among
streams of the specified program by stream metadata.
Examples:
p:601:m:language:hun will select all sub streams of program with id 601
where sub streams have metadata key named 'language' with value 'hun'.
p:602:m:guide will select all sub streams of program with id 602 where
sub streams have metadata key named 'guide'.
Signed-off-by: Bela Bodecs <bodecsb@vivanet.hu> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Steven Liu [Fri, 13 Apr 2018 07:39:49 +0000 (15:39 +0800)]
avformat/hlsenc: add option hls_delete_threshold
When using hls_list_size with hls_flags delete_segments, currently
hls_list_size * 2 +- segments remain on disk. With this new option,
the amount of disk space used can be controlled by the user.
fix ticket: #7131
Signed-off-by: Steven Liu <lq@chinaffmpeg.org> Signed-off-by: Aman Gupta <aman@tmm1.net>
avcodec/h264_slice: Fix integer overflow with last_poc
Fixes: signed integer overflow: 2147483646 - -2816 cannot be represented in type 'int' Fixes: crbug 823145 Reported-by: Matt Wolenetz <wolenetz@google.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Fixes: crbug 822705 Reported-by: Matt Wolenetz <wolenetz@google.com> Reviewed-by: Matt Wolenetz <wolenetz@google.com> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Bela Bodecs [Sun, 1 Apr 2018 21:29:14 +0000 (23:29 +0200)]
avformat/utils: Stream specifier enhancement
Currently when specifying the program id you can only decide to select
all stream of the specified program (e.g. p:103 will select all streams
of program 103) or narrow the selection to a specific stream sub index
(e.g. p:145:1 will select 2nd stream of program 145.) But you can not
specify like all audio streams of program 145 or 3rd video stream of
program 311.
In some case, mainly working with multiprogram mpeg-ts containers as
input, this feature would be handy.
This patch makes it possible to narrow the stream selection among
streams of the specified program by stream type and optionally its
index. Handled types: a, v, s, d.
Examples: p:601:a will select all audio streams of program 601,
p:603:a:1 will select 2nd audio streams of program 603,
p:604:v:0 will select first video stream of program 604.
Signed-off-by: Bela Bodecs <bodecsb@vivanet.hu> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Bela Bodecs [Wed, 11 Apr 2018 03:15:54 +0000 (11:15 +0800)]
avformat/hlsenc: fix handling of delete_segments when %v is present
When var_stream_map option is used, %v must appear either in segment
name template or in the directory path. This latter case currently is
not handled and delete_segments flag of hls_flags is broken now. This
patch fix this. The root cause of the bug was that HLSSegment struct
only stores the final filename part, but not the final directory path.
Most of the cases, final path info is unneded, It only necessary when
you want to delete old segments (e.g in case of live streaming).
Without variant streams it was unnecessary to store the final directory
path, because all segment were stored into the same directory. But
introducing %v in directory names either require to store the final
directory path into HLSSegment or associate segments with their variant
streams to be able deleting them later. I have choosen the second
solution and introduced a variant index data member into the segment
struct.
Signed-off-by: Bela Bodecs <bodecsb@vivanet.hu> Signed-off-by: Steven Liu <lq@onvideo.cn>
Timo Teräs [Tue, 3 Apr 2018 06:05:18 +0000 (09:05 +0300)]
avformat/movenc: use correct iTunes copyright atom
Support for writing copyright metadata was added in commit bed4fc54c9
for 3GP, MOV and iTunes metadata. 3GP and MOV cases are formally
specified. However, iTunes format does not have specification, and
it seems to have been assumed that it would use the same atom as
MOV (both being Apple formats).
However, Apple uses 'cprt' atom for iTunes metadata (do note that
the iTunes 'cprt' encoding is generic iTunes ItemList atom, not
the 3GP 'cprt' encoding. These are also inside different parent
atoms).
Most references trying to document iTunes atoms mention only
the 'cprt' tag. See:
- http://atomicparsley.sourceforge.net/mpeg-4files.html
- http://mutagen.readthedocs.io/en/latest/api/mp4.html
Same applies to other software supporting this tag. Most of them
encode and decode only the 'cprt' atom.
ffmpeg mov demuxer supports both atoms in this context. There are
few pieces of other software that support similarly both 'cprt' and
the incorrect '\251cpy' atom in this context. I believe they do it in
order to read the ffmpeg encoded incorrect copyright atom.
In light of the above this changes the copyright atom to 'cprt' as
it seems to be supported univerally and is the correct atom to use.
Signed-off-by: Timo Teräs <timo.teras@iki.fi> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Matt Wolenetz [Tue, 10 Apr 2018 20:59:25 +0000 (13:59 -0700)]
lavc/libopusdec: Allow avcodec_open2 to call .close
If there is a decoder initialization failure detected in avcodec_open2
after .init is called, allow graceful decoder .close to prevent leaking
libopus decoder allocations.
BUG=828526
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Limits based on 3GPP TS 26.245 V14.0.0 Fixes: Timeout Fixes: 6377/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_MOVTEXT_fuzzer-5175929115508736 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Reviewed-by: Philip Langdale <philipl@overt.org> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Zhong Li [Sat, 7 Apr 2018 17:38:55 +0000 (19:38 +0200)]
lavc/qsvdec: set complete_frame flags for progressive picture
Set the flag MFX_BITSTREAM_COMPLETE_FRAME when it is a progressive picture.
This can fix vc1 decoding segment fault issues because can't set the start
code correctly.
v2: fix some h264 interlaced clips regression
a. field_order of some h264 interlaced video (e.g: cama3_vtc_b.avc) is marked as AV_FIELD_UNKNOWN
in h264_parser.c. This is not a completed frames.
So only set the MFX_BITSTREAM_COMPLETE_FRAME when it is progressive.
b. some clips have both progressive and interlaced frames (e.g.CAPAMA3_Sand_F.264),
the parsed field_order maybe changed druing the decoding progress.
This patch has been verified for other codecs(mpeg2/hevc/vp8).
Signed-off-by: Zhong Li <zhong.li@intel.com> Signed-off-by: Maxym Dmytrychenko <maxim.d33@gmail.com>
Zhong Li [Wed, 4 Apr 2018 09:51:30 +0000 (17:51 +0800)]
lavf/qsvvpp: bypass vpp if not needed.
Currently vpp pipeline is always created, even for the unnecessary
cases such as setting the option "vpp_qsv=w=1280:h=720" for an input
with native resolution 1280x720. Thus introduces unnecessary performance
dropping, so bypass vpp if not needed.
Signed-off-by: Zhong Li <zhong.li@intel.com> Signed-off-by: Maxym Dmytrychenko <maxim.d33@gmail.com>
Zhong Li [Wed, 4 Apr 2018 09:51:29 +0000 (17:51 +0800)]
lavc/qsvdec: expose frame pic_type and key_frame
Currently pict_type and key_frame are unset.
Add an extra param to fetch the picture type from qsv decoder
The judgement “key frame is equal to IDR frame” only suitable for H264.
For HEVC, all IRAP frames are key frames, and other codecs have no IDR
frame.
Signed-off-by: ChaoX A Liu <chaox.a.liu@intel.com> Signed-off-by: Zhong Li <zhong.li@intel.com> Signed-off-by: Maxym Dmytrychenko <maxim.d33@gmail.com>
avcodec/aacsbr_fixed: Fix integer overflow in sbr_hf_assemble()
Fixes: runtime error: signed integer overflow: 2052929346 + 204817098 cannot be represented in type 'int'
This was missed in b1bef755f617af9685b592d866b3eb7f3c4b02b1 Fixes: 5275/clusterfuzz-testcase-minimized-5367635958038528 Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
James Almer [Wed, 4 Apr 2018 17:12:54 +0000 (14:12 -0300)]
avformat/matroskadec: use refcounted buffers in EbmlBin
Data in EbmlBin objects is never changed after being read from the
input file (save for two specific cases with encoded CodePrivate), so
using AVBufferRef we can prevent unnecessary copy of data by instead
creating new references to said constant data.
avcodec/decode: fix warning when decoding pseudo paletted formats
The pseudo palette allocation is optional now. But if it's still
allocated (like the internal get_buffer2 implementation does, for
compatibility), it shouldn't print a warning.
PSEUDOPAL pixel formats are not paletted, but carried a palette with the
intention of allowing code to treat unpaletted formats as paletted. The
palette simply mapped the byte values to the resulting RGB values,
making it some sort of LUT for RGB conversion.
It was used for 1 byte formats only: RGB4_BYTE, BGR4_BYTE, RGB8, BGR8,
GRAY8. The first 4 are awfully obscure, used only by some ancient bitmap
formats. The last one, GRAY8, is more common, but its treatment is
grossly incorrect. It considers full range GRAY8 only, so GRAY8 coming
from typical Y video planes was not mapped to the correct RGB values.
This cannot be fixed, because AVFrame.color_range can be freely changed
at runtime, and there is nothing to ensure the pseudo palette is
updated.
Also, nothing actually used the PSEUDOPAL palette data, except xwdenc
(trivially changed in the previous commit). All other code had to treat
it as a special case, just to ignore or to propagate palette data.
In conclusion, this was just a very strange old mechnaism that has no
real justification to exist anymore (although it may have been nice and
useful in the past). Now it's an artifact that makes the API harder to
use: API users who allocate their own pixel data have to be aware that
they need to allocate the palette, or FFmpeg will crash on them in
_some_ situations. On top of this, there was no API to allocate the
pseuo palette outside of av_frame_get_buffer().
This patch not only deprecates AV_PIX_FMT_FLAG_PSEUDOPAL, but also makes
the pseudo palette optional. Nothing accesses it anymore, though if it's
set, it's propagated. It's still allocated and initialized for
compatibility with API users that rely on this feature. But new API
users do not need to allocate it. This was an explicit goal of this
patch.
Most changes replace AV_PIX_FMT_FLAG_PSEUDOPAL with FF_PSEUDOPAL. I
first tried #ifdefing all code, but it was a mess. The FF_PSEUDOPAL
macro reduces the mess, and still allows defining FF_API_PSEUDOPAL to 0.
Passes FATE with FF_API_PSEUDOPAL enabled and disabled. In addition,
FATE passes with FF_API_PSEUDOPAL set to 1, but with allocation
functions manually changed to not allocating a palette.
projects / ffmpeg / log