The container has no timestamps and the framerate isn't stored in the
data either.
The decoder sets codec timebase to experimentally found value 1/15. Do
the same for the demuxer too, it should at least be better than the
default 1/90000.
The fields "Number of Bytes" and "Number of Frames" are mixed up. "Bytes"
come first, "Frames" behind.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Alex Converse <alex.converse@gmail.com>
This makes the packetization spec compliant for cases where one single
GOB doesn't fit into an RTP packet.
Signed-off-by: Martin Storsjö <martin@martin.st>
This fixes cases where the user had specified one desired MTU
via an option, and the protocol indicates another one.
Signed-off-by: Martin Storsjö <martin@martin.st>
Frame sizes in ID3v2.3 are not synchsafe, they are simply 32be numbers.
In practice this bug is not noticeable unless the frame size takes more
than 7 bits (which is almost never for text frames).
Seeking back on EOF will reset the EOF flag, causing us to re-enter
the loop to find the next marker in the ASF file, thus potentially
causing an infinite loop.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
Provide a way to wrap around the segment index so pseudostreaming
live through a web server and html5 browser is simpler.
Also ensure that 0 (disable) is a valid value across the options
providing wrap around.
By validating the index entries while reading, we don't need to
seek at startup to validate the entries. If the error in the
index entries is not pointing to (our definition of) the start
of packets, and there is an index entry pointing at some of the
first packets after the metadata, the invalid index can be discarded
almost immediately.
Signed-off-by: Martin Storsjö <martin@martin.st>
This returns 200 OK for OPTIONS requests and 501 Not Implemented
for all other requests.
Even though this doesn't do much actual handling of the requests,
it makes the code properly identify server requests as such, instead
of interpreting it as a reply to the client's request as it did
before.
Signed-off-by: Martin Storsjö <martin@martin.st>
For encoding, AVCodecContext.frame_size is the number of input samples to
send to the encoder and does not necessarily correspond directly to the
timestamps of the output packets.
This prevents certain tags with a default value assigned to them (as per
the EBML syntax elements) from ever being assigned a NULL value. Other
parts of the code rely on these being non-NULL (i.e. they don't check for
NULL before e.g. using the string in strcmp() or similar), and thus in
effect this prevents crashes when reading of such specific tags fails,
either because of low memory or because of targeted file corruption.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
compute_pkt_fields() is for unreliable estimates or guessing. The
keyframe information from the parser is (at least in theory) reliable,
so it should be used even when the other guessing is disabled with the
AVFMT_FLAG_NOFILLIN flag.
Therefore, move setting the packet keyframe flag based on parser
information from compute_pkt_fields() to read_frame_internal().
fixes a memleak for Vorbis and Theora, where the comment header from
avpriv_split_xiph_headers() is replaced by a buffer that must be freed
separately.
This allows opting for a lower MTU than what the AVIOContext
indicated, and allows writing into outputs that don't indicate
an MTU at all (such as plain files, which is useful for testing).
This also allows querying for the MTU via the avoption.
Signed-off-by: Martin Storsjö <martin@martin.st>
According to newer RFCs, this packetization scheme should only
be used for interfacing with legacy systems.
Implementing this packetization mode properly requires parsing
the full H263 bitstream to find macroblock boundaries (and knowing
their macroblock and gob numbers and motion vector predictors).
This implementation tries to look for GOB headers (which
can be inserted by using -ps <small number>), but if the GOBs
aren't small enough to fit into the MTU, the packetizer blindly
splits packets at any offset and claims it to be a GOB boundary
(by using Mode A from the RFC). While not correct, this seems
to work with some receivers.
Signed-off-by: Martin Storsjö <martin@martin.st>
Specifically, prevent jumping back in the file for the next index, since
this can lead to infinite loops where we jump between indexes referring
to each other, and don't read indexes that don't fit in the file.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
It is not supposed to be done outside lavc.
This is basically a revert of 818062f2f3.
It is unclear what issue this was supposed to fix, if it reappears again
it will have to be fixed in a more proper place.
The wtv-demux test change is because the sample starts with a B-frame.
We read sub_packet_h / 2 packets per line of data (during deinterleaving),
which equals zero if sub_packet_h <= 1, thus causing us to not read any
data, leading to an infinite loop.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
This allows writing QuickTime-compatible fragmented mp4 (with
a non-empty moov atom) to a non-seekable output.
This buffers the mdat for the initial fragment just as it does
for all normal fragments, too. Previously, the resulting
atom structure was mdat,moov, moof,mdat ..., while it now
is moov,mdat, moof,mdat.
Signed-off-by: Martin Storsjö <martin@martin.st>
In nonseekable files, we already stop parsing the toplevel atoms
after finding moov and one mdat. In large seekable files (or files
that are seekable, but slowly, e.g. http), reading all the fragments
at the start can take a considerable amount of time. This allows
opting out from this behaviour.
Signed-off-by: Martin Storsjö <martin@martin.st>
If parsing moov+mdat in a non-seekable file, we currently
abort parsing directly after parsing the header of the mdat
atom. If we want to continue parsing later (if looking to
parse later fragments), we need to skip past the content of the
mdat atom, otherwise we end up parsing the content of the mdat
atom as root level atoms.
Signed-off-by: Martin Storsjö <martin@martin.st>
For video, mark the first sample in a trun which doesn't have the
sample-is-non-sync-sample flag set as a keyframe.
In particular, the "sample does not depend on other samples" flag
isn't enough to make it a keyframe, since later frames still can
reference frames prior to that one (the flag only says that that
particular frame doesn't depend on other frames).
This fixes bug 215.
Signed-off-by: Martin Storsjö <martin@martin.st>
This avoids reading any old data in the AVIOContext buffer after
the seek, and indicates to the mpegts demuxer that we've seeked,
avoiding continuity check errors.
Signed-off-by: Martin Storsjö <martin@martin.st>
Enhance seeking by demuxing until the requested timestamp is
reached within the segment selected by the seek code using the
playlist info.
Some mpegts streams don't have dts set for all packets though,
this seeking method doesn't work well for that case.
Signed-off-by: Martin Storsjö <martin@martin.st>
This prevents failed assertions further down in the packet processing
where we require non-negative values for packet_size_left.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org
H263 in RTP can be packetized in two formats (RFC 2190, RFC
2429/4629). The former normally uses the static payload type 34,
while the latter normally uses dynamic payload types with the
SDP format names H263-1998 or H263-2000.
Look for packets that don't look like proper RFC 2190 packets and
switch to depacketizing them according to the new format if they
match some heuristic criteria.
Signed-off-by: Martin Storsjö <martin@martin.st>
According to 14496-12, the duration should be all 1s if
the duration is unknown. This is the case if writing a moov
atom without any samples described in it (e.g. as in ismv files).
Signed-off-by: Martin Storsjö <martin@martin.st>
According to unofficial documentation, the video rate is locked to the audio
sample rate. This results in proper synchronization of audio and video
timestamps from the demuxer. This only works if the first audio packet occurs
before the first video packet or the audio sample rate is the default rate of
11111 Hz, both of which are true for all samples in our archive.
This avoids initializing a stream with dummy values or when the file does not
contain audio.
Also set duration for audio packets, using the sample rate as the time base.
Update FATE reference to account for now non-existent palette packet.
This also fixes the FATE test if frame data is not initialized in
get_buffer(), so update comment in avconv accordingly.
The rtp demuxer which listens for RTP packets and detects the
RTP payload type will currently get confused if the first packet
received is an RTCP packet. Thus ignore such packets.
Signed-off-by: Martin Storsjö <martin@martin.st>
Prefix the functions/tables brktimegm, pcm_read_seek,
dv_offset_reset, voc_get_packet, codec_movaudio_tags,
codec_movvideo_tags.
After this, lavf has no global symbols without the proper prefix.
Signed-off-by: Martin Storsjö <martin@martin.st>
Keep the old protocol name around for backwards compatibility
until the next bump.
Deprecate the method of implicitly assuming the nested protocol.
For applehttp://server/path, it might have felt logical, but
supporting hls://server/path isn't quite as intuitive. Therefore
only support hls+http://server/path from now on.
Using this protocol at all is discouraged, since the hls demuxer
is more complete and fits into the architecture better. There
have been cases where the protocol implementation worked better
than the demuxer, but this should no longer be the case.
Signed-off-by: Martin Storsjö <martin@martin.st>
When this demuxer was created, there didn't seem to be any
consensus of a common short name for this protocol. Now
the consensus seems to be to call it hls.
Signed-off-by: Martin Storsjö <martin@martin.st>
This is different from the "modern" RTP payload formats for H263
as defined by RFC 4629, 2429 and 3555. According to the newer RFCs,
this old one is to be considered deprecated and only be used for
interoperating with legacy systems.
Signed-off-by: Martin Storsjö <martin@martin.st>
This allows easily differentiating between both implementations within the build
system and combining the native implementation for plain RTMP with librtmp for
the RTMPE, RTMPS, RTMPT, RTMPTE protocol variants.
If muxing into mpegts, 4 byte startcodes for the first NAL
of an access unit is required. Thus it is simplest for the
RTP depacketizer to just use 4 byte startcodes everywhere.
Signed-off-by: Martin Storsjö <martin@martin.st>
When segmenting the output from the mpegts muxer, one can
now set this option when cutting to a new segment, to make sure
the next segment starts with PAT/PMT/SDT.
Signed-off-by: Martin Storsjö <martin@martin.st>
The last entry is the total size of the essence container.
Previously a TemporalOffset error would be logged, even though
segments like these are expected.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
These are common with audio atoms. Without this the demuxer would read two
bytes at a time for a mono 16-bit file.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Specially crafted files can lead the parsing code to take too long.
We fix a lot of these problems by not allowing local tags to extend
past the end of the set and not allowing other KLVs to be read past
the end of themselves.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
This can happen if an index table segment has a very large IndexStartPosition.
zzuf3.mxf is an example of such a file.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Only the OPAtom demuxing logic is guaranteed to have index tables,
meaning OP1a files that lack an index would cause SIGSEGV.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Calculates based on total file size and wavetaillength from the header.
Falls back to multiplying finalframeblocks by 8 instead of 4 so that it will
at least be overestimating for 24-bit. Currently it can underestimate the
final packet size, leading to decoding errors.
Right now those muxers use the default timebase in all cases(1/90000).
This patch avoid unnecessary rescaling and makes the printed timestamps
more readable.
Also, extend the printed information to include the timebases and packet
pts/duration and align the columns.
Obviously changes the results of all fate tests which use those two
muxers.
Normally, the actual payload data contains sequence headers, too,
and the parser can extract this and set it as extradata. However,
the data in the dvc1 atom is the "official" extradata for the file.
This is required for proper stream copy of vc1 from ismv to ismv.
Signed-off-by: Martin Storsjö <martin@martin.st>
Do not use AVStream's duration for dts generation since it contains in
some cases the duration of the whole file instead of duration of the
samples in the moov. This happens if the mdhd holds the duration of the
whole file but has no entries or a zero duration in its stts.
dv: Fix null pointer dereference due to ach=0
Fixes part2 of CVE-2011-3929
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Reviewed-by: Roman Shaposhnik <roman@shaposhnik.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Alex Converse <alex.converse@gmail.com>
dv: check stype
Fixes part1 of CVE-2011-3929
Possibly fixes part of CVE-2011-3936
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Reviewed-by: Roman Shaposhnik <roman@shaposhnik.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Alex Converse <alex.converse@gmail.com>
This way, if the AVCodecContext is allocated for a specific codec, the
caller doesn't need to store this codec separately and then pass it
again to avcodec_open2().
It also allows to set codec private options using av_opt_set_* before
opening the codec.
It allows to check whether an AVCodecContext is open in a documented
way. Right now the undocumented way this check is done in lavf/lavc is
by checking whether AVCodecContext.codec is NULL. However it's desirable
to be able to set AVCodecContext.codec before avcodec_open2().
This makes the first packet of a track fragment run to get
the keyframe flag set properly if sample_degradation_priority
is nonzero.
This makes the keyframes flag be set properly for ismv files
created by Microsoft.
Signed-off-by: Martin Storsjö <martin@martin.st>
Previously, we've only passed the key string on to the recursive
amf_parse_object for the mixedarray type, not for 'object'. By
passing the key string on, the recursive amf_parse_object can
store the amf objects as metadata.
This kind of data was seen in data from XSplit Broadcaster, received
over RTMP via Wowza. This patch allows reading this metadata.
Signed-off-by: Martin Storsjö <martin@martin.st>
Check results for av_malloc() and fix an overflow in one call.
Related to CVE-2011-3940.
Based in part on work from Michael Niedermayer.
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Fixes CVE-2011-3940 (Out of bounds read resulting in out of bounds write)
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
(cherry picked from commit 5c011706bc)
Signed-off-by: Alex Converse <alex.converse@gmail.com>
-vbsf doesn't exist anymore. It got renamed to -bsf somewhere along the
line. Update print statement accordingly.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Current demuxer recognizes several colorspace formats that begin with 'C420'
but does not yet recognize plain 'C420'. GStreamer's y4menc component
generates .y4m files with a 'C420' colorspace. This new comparison is
placed after the other 'C420' checks so that it doesn't interfere with
them.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
In this mode, no seeks will be done except for within moov/moof
fragments, which should fit within the AVIOContext buffer.
This allows pushing live smooth streaming format data to
a live publishing point on IIS over http.
Signed-off-by: Martin Storsjö <martin@martin.st>
If no data was seen for a stream decoder are returning 0 when fed with
empty packets for flushing. We can stop flushing when the decoder does
not return delayed delayed frames anymore. Changes try_decode_frame()
return value to got_picture or negative error.
CC: libav-stable@libav.org
The field frame_size isn't written to the output anywhere except
than in mov.
This facilitates stream copy from formats that don't set frame_size.
Signed-off-by: Martin Storsjö <martin@martin.st>
Also add some space around operators and wrap a comment
that extends past the 80 char "limit"/guideline.
Signed-off-by: Martin Storsjö <martin@martin.st>
It can also optionally split the file into individual fragments,
which allows it to be served from any web server without any
server side support.
Signed-off-by: Martin Storsjö <martin@martin.st>
This fixes calculation of trackDuration if the MOVIentry array
is cleared. This is required by the fragmentation support in the
next patch.
Signed-off-by: Martin Storsjö <martin@martin.st>
