Janne Grunau [Thu, 8 Dec 2011 19:28:49 +0000 (20:28 +0100)]
mxfdec: hybrid demuxing/seeking solution
This uses the old demuxing code for OP1a and separate demuxing code for OPAtom.
Timestamp output is added to the old demuxing code.
The seeking code is made to seek to the start of the desired EditUnit only,
from which the normal demuxing code takes over (if OP1a). This means we
do not use delta entries or slices, only StreamOffsets. OPAtom seeking
basically works like before.
This also makes D-10 seeking behave the same way as OP1a and OPAtom. In other
words, we allow seeking before the start or past the end for D-10 too.
Based on several patches by Tomas Härdin <tomas.hardin@codemill.se> and
Reimar Döffinger <Reimar.Doeffinger@gmx.de>.
Changed av_calloc to av_mallocz, added overflow checks.
Tomas Härdin [Mon, 14 Mar 2011 16:00:21 +0000 (17:00 +0100)]
mxfdec: Speed up metadata and index parsing.
Specifically, this means parsing as before until we run into essence.
At that point we seek to the footer and parse until EOF. After that we start
seeking backward to the previous partition and parse that until we run into
essence or the next partition. This procedure is repeated until we encounter
the last partition we parsed in the forward direction.
The end result of all this is that large essence containers are not needlessly
parsed. This speeds up parsing large files a lot.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
Philip de Nier [Mon, 12 Sep 2011 10:13:44 +0000 (11:13 +0100)]
mxfdec: skip to end of structural sets
This fixes reading of partition packs. The code stops reading after the
operational pattern and should skip the array of essence container
labels that follow.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
Anton Khirnov [Thu, 12 Jan 2012 08:51:23 +0000 (09:51 +0100)]
avserver: fix build after the next bump.
Now that 0.8 is out we can reapply this commit. It breaks shared
avserver builds due to avserver using internal libavformat symbols,
which are now hidden, so this commit also disables avserver with
--enable-shared.
Dmitry Volyntsev [Wed, 18 Jan 2012 09:46:37 +0000 (13:46 +0400)]
rtsp: Use a random offset for trying to open UDP ports for RTP
This avoids (for all practical cases) the issue of reusing
the same UDP port as for an earlier connection. If the remote
doesn't know the previous session was closed, he might keep
on sending packets to that port. If we always start off trying
to open the same UDP port, we might get those packets intermixed
with the new ones.
This is occasionally an issue when testing RTSP stuff with
DSS, perhaps also with other servers.
Martin Storsjö [Sat, 21 Jan 2012 15:20:45 +0000 (17:20 +0200)]
rtpdec: Use our own SSRC in the SDES field when sending RRs
The s->ssrc field is the sender's SSRC, we use ssrc + 1 to get
a collision free "unique" SSRC for ourselves in the RR part.
The SDES block in the RTCP packet should describe ourselves,
not the sender.
This was fixed for the RR part in 952139a3226b, but wasn't
fixed for the SDES part until now.
This could cause some Axis cameras to send RTCP BYE packets
to us due to the SSRC collision.
Janne Grunau [Wed, 18 Jan 2012 09:53:41 +0000 (10:53 +0100)]
threads: change the default for threads back to 1
Using threaded decoding by default breaks backward compatibility if
AVHWAccel is used or if an appliction sets threadunsafe callbacks.
Avconv and avplay still use -threads auto if not specified.
Anton Khirnov [Sat, 21 Jan 2012 07:40:47 +0000 (08:40 +0100)]
Revert "avserver: fix build after the next bump."
This temporarily (until 0.8 is released) reverts commit 8e1340abc316e038bb89e5a3b46e92ff58c98a88. That commit breaks shared
builds because of symbol hiding. Reverting it will enable shared builds
for 0.8
John Brooks [Tue, 13 Dec 2011 00:04:14 +0000 (17:04 -0700)]
mpegvideo: fix invalid memory access for small video dimensions
When either video dimension is only one macroblock, subtractions
based on v_edge_pos and the macroblock size may be negative. In
that situation, an unsigned comparison isn't sufficent to test for
MV overruns, because a limit of (unsigned)-1 will let any other
value pass.
Martin Storsjö [Fri, 20 Jan 2012 18:48:19 +0000 (20:48 +0200)]
movenc: Reorder entries in the MOVIentry struct, for tigheter packing
Originally, sizeof(struct MOVIentry) was 48, after the reordering,
it is 40 in my build configuration.
When writing really long mov/mp4 files, this can make a difference
- this saves a bit over 2 MB of memory per hour of video (down to
10.3 MB per hour from 12.3 MB per hour initially) for a video with
75 packets per second - 25 fps + 50 audio packets (which is the
case for AMR audio).
Janne Grunau [Wed, 18 Jan 2012 19:32:32 +0000 (20:32 +0100)]
lavf: force single-threaded decoding in avformat_find_stream_info
The H.264 decoder needs SPS and PPS for initialization during
multi-threaded decoding. When probed single-threaded SPS and PPS are
copied to extradata and are available for proper initialization of
the decoder before the first frame is decoded.
Extract processing of intra 16x16 blocks from intra macroblock
processing.
Also implement a function performing inverse transform and block
reconstruction for DC-only blocks in 1 pass instead of 2.
Split inter/intra macroblock handling code. This will allow further
optimizations such as performing inverse transform and block reconstruction
in a single pass as well as specialize code.
Signed-off-by: Janne Grunau <janne-libav@jannau.net>
Reinhard Tartler [Sun, 15 Jan 2012 19:32:11 +0000 (20:32 +0100)]
lavc: Relax API strictness in avcodec_decode_audio3 with a custom get_buffer()
Do not fail audio decoding with avcodec_decode_audio3 if user has set a
custom get_buffer. Strictly speaking, this was never allowed by the API,
but it seems that some software packages did so anyways. In order to
unbreak applications (cf. http://bugs.debian.org/655890), this change
clarifies the API and overrides the custom get_buffer() with the defaults.
Kostya Shishkov [Sun, 15 Jan 2012 13:23:28 +0000 (14:23 +0100)]
wavpack: fix clipping for 32-bit lossy mode
Reference decoder clips data before shifting it to final range and also
forces 32-bit lossy mode to be actually 24-bit lossy mode in order to be
able to perform proper clipping.
Martin Storsjö [Fri, 13 Jan 2012 21:41:16 +0000 (23:41 +0200)]
libx264: Don't leave max_b_frames as -1 if the user didn't set it
max_b_frames is initialized to -1 for libx264, to allow
distinguishing between an explicit user set 0 and a default not
touched 0 (see bb73cda2).
If max_b_frames is left as -1, this affects dts generation (where
expressions like max_b_frames != 0 are used), so make sure it is
left at the default 0 after the libx264 init function returns.
This avoids unnecessarily producing dts != pts when using
profile=baseline.
Mans Rullgard [Fri, 13 Jan 2012 18:56:36 +0000 (18:56 +0000)]
ARM: 4-byte align start of all asm functions
Due to apprent bugs in the GNU assembler and/or linker, relocations
can be incorrectly processed if the alignment of a Thumb instruction
is changed in the output file compared to the input object.
This fixes crashes in h264 decoding with Thumb enabled. No effect in
ARM mode since everything is 4-byte aligned there.