Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

p cannot be calculated before av_dup_packet since that one
might change avpkt->data, causing invalid reads and a
non-working range check.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* qatar/master:
  http: Remove the custom function for disabling chunked posts
  rtsp: Disable chunked http post through AVOptions
  movdec: Set frame_size for AMR
  h264_weight: remove duplication functions.
  swscale: align vertical filtersize by 2 on x86.
  libavfilter: reindent.
  matroskadec: empty blocks are in fact valid.
  avfilter: don't abort() on zero-size allocations.
  h264: improve calculation of codec delay.
  movenc: Set a correct packet size for AMR-NB mode 15, "no data"
  avformat: Add functions for doing global network initialization
  avformat: Add the https protocol
  avformat: Add the tls protocol, using OpenSSL or gnutls
  avformat: Initialize gnutls in ff_tls_init()
  w32threads: Wrap the mutex functions in inline functions returning int
  configure: Allow linking to the gnutls library
  avformat: Add ff_tls_init()/deinit() that initialize OpenSSL
  configure: Allow linking to openssl
  avcodec: Allow locking and unlocking an avformat specific mutex
  avformat: Split out functions from network.h to a new file, network.c

Conflicts:
	Changelog
	configure
	doc/APIchanges
	libavcodec/internal.h
	libavcodec/version.h
	libavfilter/formats.c
	libavformat/matroskadec.c
	libavformat/mov.c
	libavformat/version.h
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avctx->width/height remain right/bottom cropped as previous behaviour.

Hardware decoders need to know the uncropped data to allocate surfaces
of correct height. Some hardware is picky and fails to decoder properly
if a surface larger than needed is used during decode, so just aligning
up is not enough.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

When constructing a channel layout string from individual channel names,
+ is a more practical separator than |, because it is usually not a shell
special character.
Signed-off-by: Nicolas George <nicolas.george@normalesup.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Nicolas George <nicolas.george@normalesup.org>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fixed Ticket432
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

GCC 4.6.2 at least still seems to fail to put them in .rodata though,
see also http://gcc.gnu.org/bugzilla/show_bug.cgi?id=37303Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Otherwise it might return a write-only frame which would break
decoding completely.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

This fixes visible corruption in the incomplete last frame
of the FATE sample.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

There were multiple issues, for example might we have to re-run
the decompression when the size of the buffer increased,
we should always use a decompression buffer large enough for
the header (so we do not get stuck when the size is too small).
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Since it contains dimensions parsing an invalid one has rather
annoying effects.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

This avoids crashes due to reading out-of-bounds.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Treating them like 0 is safest, current code would invoke
undefined pointer arithmetic behaviour in this case.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

I am not sure these new values are correct, not am I sure
the semantics are a good idea since we do not seem to make any
use of them but they caused a lot of confusion, but this
seems to make things closer to matching the documentation.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

* cus/stable:
  ffplay: Avoid useage of deprecated url_fileno().
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Martin Storsjö <martin@martin.st>

This avoids having to use a private function.
Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>