No change in assembly code produced on x86_64.

Once fixed, the end_frame function does exactly what
avfilter_default_end_frame does; therefore, end_frame
can be removed to let avfilter_default_end_frame work.

Fixes ticket #1038.

* qatar/master: (29 commits)
  sbrdsp.asm: convert all instructions to float/SSE ones.
  dv: cosmetics.
  dv: check buffer size before reading profile.
  Revert "AAC SBR: group some writes."
  udp: Print an error message if bind fails
  cook: extend channel uncoupling tables so the full bit range is covered.
  roqvideo: cosmetics.
  roqvideo: convert to bytestream2 API.
  dca: don't use av_clip_uintp2().
  wmall: fix build with -DDEBUG enabled.
  smc: port to bytestream2 API.
  AAC SBR: group some writes.
  dsputil: remove shift parameter from scalarproduct_int16
  SBR DSP: unroll sum_square
  rv34: remove dead code in intra availability check
  rv34: clean a bit availability checks.
  v4l2: update documentation
  tgq: convert to bytestream2 API.
  parser: remove forward declaration of MpegEncContext
  dca: prevent accessing static arrays with invalid indexes.
  ...

Conflicts:
	doc/indevs.texi
	libavcodec/Makefile
	libavcodec/dca.c
	libavcodec/dvdata.c
	libavcodec/eatgq.c
	libavcodec/mmvideo.c
	libavcodec/roqvideodec.c
	libavcodec/smc.c
	libswscale/output.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

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

This allows future assumptions to be made without affecting non-intra files.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The "ECs != 1 -> OP1a" assumption was wrong. Luckily, the file that triggered
that behavior had two ECs, not zero. Hence distinguishing between them is
simple in this case.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This fixes rare cases where OPAtom may be treated as OP1a, causing all essence
to be read into RAM.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

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

Speed += 10% overall for pan=stereo:FL=FL+FR from wav to null.

Since the values are floats, using the float operations
makes sense, improves performance on some CPUs and
makes the code SSE compatible instead of needing SSE2.

Based on suggestion by Jason.
Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

This reverts commit ba36f14e. It
broke decoding on x86-32 on some systems.

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

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

The argument is not a literal, thus causing the ARM v6 or later
builds to break.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

There is only one caller, which does not need the shifting. Other use cases
are situations where different roundings would be needed.

The x86 and neon versions are modified accordingly.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

The length is even, so some unrolling can be performed. Timings are for x86:
- 32bits: 102c -> 82c
- 64bits:  82c -> 69c
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

This was an incorrect copy-and-paste to a code not needing the original code.
Spotted by Jason in a previous review but forgotten in the commit.
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Update the examples and remove stale information.
Video4linux support had been already removed.

This protects against input buffer overreads.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Signed-off-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

This way, it protects against overreads for 4bpp/2bpp content also.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Fixes bug 240, as well as several integer overflows (visible as glitches)
in other scaling output routines, e.g. YUV422.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC: libav-stable@libav.org

Fixes ticket #1040.

The +1 is there to skip the ','.
With \r\n, the +1 skips the \r but that is ok.
With only \n, the +1 skips it and all hell breaks loose.