* commit '3ee462dc':
  examples/qsvdec: do not free the surfaces in the frame_free() callback
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit '2c32eace':
  qsvdec: close the MFX decoder on uninit
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'd0c8c380':
  qsv: document AVQSVContext members
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit '5d2daebf':
  qsvenc: mark the encoders as INIT_CLEANUP
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit '83847cc8':
  qsvenc: do not try to close the encoder if the session is NULL
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit '741b352b':
  qsvdec: list supported pixel formats
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'ae25413d':
  lavfi: do not exclude hwaccel formats from ff_all_formats()
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'cc285521':
  checkasm/x86: Correctly handle variadic functions
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit '5405584b':
  checkasm: Use a self-balancing tree
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'a41e5e19':
  vdpau: fix constrained baseline fallback
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'cc8db760':
  mpegts: use avcodec_get_type() to set codec_type
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'a5d58fea':
  lavc: reimplement avcodec_get_type() using codec descriptors
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

* commit 'b10b6ac7':
  vdpau: deprecate av_vdpau_get_profile()
Merged-by: Hendrik Leppkes <h.leppkes@gmail.com>

Fixes crash
Fixes: flicker-1.scout3d21443372922.28.m4a
Found-by: Dale Curtis <dalecurtis@google.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

pix_fmt was declared presumably to shorten the argument passed to the function.
However, it is currently not being used for such a purpose.
This patch simply removes it instead.
This fixes -Wunused-but-set-variable reported at e.g:
http://fate.ffmpeg.org/log.cgi?time=20150919194249&log=compile&slot=x86_64-darwin-gcc-4.9.
Signed-off-by: Ganesh Ajjanagadde <gajjanagadde@gmail.com>
Reviewed-by: wm4 <nfxjfg@googlemail.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

In case of resync, always free the packet, but retry only if the resync
did not get to the end of the file. Otherwise, there is a memory leak when the
last packet in the file is corrupted.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Fixes #4887.
Signed-off-by: Paul B Mahol <onemda@gmail.com>

The previous implementation was behaving incorrectly in some corner cases.

Even though libmfx might not need them anymore, avcodec might still
access the surfaces. So free them separately at the end.

Should fix some leaks on init failure.

This is useful for detecting QSV-enabled decoders.

It should be possible to pass hwaccel frames through lavfi.

The System V ABI on x86-64 specifies that the al register contains an upper
bound of the number of arguments passed in vector registers when calling
variadic functions, so we aren't allowed to clobber it.

checkasm_fail_func() is a variadic function so also zero al before calling it.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

The code assumed that the outermost interpolated pixels were always in
the fuzzy area defined by the band option. However if the band value
is small, there may be no fuzzy area on a given plane. In that case,
option show did not work, no rectangle was drawn (or only on the luma
plane, depending on the band value and chroma plane subsampling
factors.)

Fix the problem by not making any assumption on where the outermost
interpolated pixels will be.

The new code was verified to produce the same result as the original
code when the band value is not small.
Signed-off-by: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>

Tested functions are internally kept in a binary search tree for efficient
lookups. The downside of the current implementation is that the tree quickly
becomes unbalanced which causes an unneccessary amount of comparisons between
nodes. Improve this by changing the tree into a self-balancing left-leaning
red-black tree with a worst case lookup/insertion time complexity of O(log n).

Significantly reduces the recursion depth and makes the tests run around 10%
faster overall. The relative performance improvement compared to the existing
non-balanced tree will also most likely increase as more tests are added.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

It appears vdpau drivers can return constrained baseline as unsupported,
even if libvdpau knows about the symbol, and the main profile is
supported.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Note that this slightly changes behavior: it sets AVMEDIA_TYPE_UNKNOWN
if the codec type is unknown. This should be ok.
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

The appropriate flag for HEASLR (--high-entropy-va) was added in
binutils 2.25.

Also set the image base >4GB so higher entropy gets applied to image
base randomization when used with HEASLR (8 -> 17 bits of
randomization).  Windows does this for compatibility because of "latent
pointer truncation issues".
Signed-off-by: Alex Smith <alex.smith@warpsharp.info>

This function can intrinsically not deal with codec profile fallback
(for H.264 Constrained Baseline especially), and was made redundant
by av_vdpau_bind_context().
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Each line is padded by the format, and errors are now reported and
stop the decoding. Around 5% speedup.
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>