lavf/mpegts: rename ff_mpegts_parse_*() to avpriv_mpegts_parse_*()

The were wrongly being exported and used by libavdevice

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'b50173a4dd47b9c3c89845b781fa958ccf860929'

* commit 'b50173a4dd47b9c3c89845b781fa958ccf860929':
  movenc: fix QT chapter track character encoding

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

Merge commit 'da9cc22d5bd5f59756c2037b02966376da2cf323'

* commit 'da9cc22d5bd5f59756c2037b02966376da2cf323':
  movenc: add track title to tracks

Conflicts:
libavformat/movenc.c

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

Merge commit 'd9432789bd119f0e37bcf65cebda05d36aafd4ed'

* commit 'd9432789bd119f0e37bcf65cebda05d36aafd4ed':
  movenc: remove pointless loop around BITEXACT test

Conflicts:
libavformat/movenc.c

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

avcodec/svq1dec: Fix multiple bugs from "svq1: do not modify the input packet"

Add padding, clear size, use the correct pointer.

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

Merge commit '7b588bb691644e1b3c168b99accf74248a24e3cf'

* commit '7b588bb691644e1b3c168b99accf74248a24e3cf':
  svq1: do not modify the input packet

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

Merge commit 'c7d9b473e28238d4a4ef1b7e8b42c1cca256da36'

* commit 'c7d9b473e28238d4a4ef1b7e8b42c1cca256da36':
  cdgraphics: do not return 0 from the decode function

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

Merge commit 'a1599f3f7ea8478d1f6a95e59e3bc6bc86d5f812'

* commit 'a1599f3f7ea8478d1f6a95e59e3bc6bc86d5f812':
  cdgraphics: switch to bytestream2

Conflicts:
libavcodec/cdgraphics.c

See: ad002e1a13a8df934bd6cb2c84175a4780ab8942
Merged-by: Michael Niedermayer <michaelni@gmx.at>

libavcodec/rawdec: remove old ff_find_pix_fmt() function

It's been dead code for more than a year

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avformat/dtsdec: check more of the dca headers in dts_probe()

Fixes misdetection of 664461_flac_16_44100_2.wav
Fixes Ticket3830

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

avcodec/dca: Make ff_dca_convert_bitstream() available to libavformat, needed for dts_probe()

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

movenc: fix QT chapter track character encoding

An encoding ("encd") box is required to tell QT that the string is UTF8

movenc: add track title to tracks

movenc: remove pointless loop around BITEXACT test

svq1: do not modify the input packet

The input data must remain constant, make a copy instead. This is in
theory a performance hit, but since I failed to find any samples
using this feature, this should not matter in practice.

Also, check the size of the header, avoiding invalid reads on truncated
data.

CC:libav-stable@libav.org

Merge commit 'ed6d9ce914d552eeda16af857da97c4b1aea1e3f'

* commit 'ed6d9ce914d552eeda16af857da97c4b1aea1e3f':
  configure: Include the armcc build number in the compiler identification

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

cdgraphics: do not return 0 from the decode function

0 means no data consumed, so it can trigger an infinite loop in the
caller.

CC:libav-stable@libav.org

cdgraphics: switch to bytestream2

Fixes possible invalid memory accesses on corrupted data.

CC:libav-stable@libav.org
Bug-ID: CVE-2013-3674

Merge commit 'cbc808d726afdf53d866264722785c1304c17390'

* commit 'cbc808d726afdf53d866264722785c1304c17390':
  jpeg2000: enable 4 component pixel formats

See: f0358dc1d30cd4f4862489ab95c4d408b00a8b0d
Merged-by: Michael Niedermayer <michaelni@gmx.at>

x86: sbrdsp/fft: reuse ps_neg constant

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

x86: diracdsp: reuse constants

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

x86: dwt: better share constants

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

x86: better share ff_pw_2

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

x86: vpx/h264/hevc/mpeg2: share constants

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

configure: Include the armcc build number in the compiler identification

This tries to find the most expressive part of the output of
armcc --vsn to include, giving a compiler identification of
"ARM Compiler 5.04 update 2 (build 82)" instead of just
"ARM Compiler 5.04" for armcc 5.0.

4.x versions of armcc output the following, for "armcc --vsn":
ARM C/C++ Compiler, RVCT4.0 [Build 925]
For evaluation purposes only
Software supplied by: ARM Limited

ARM C/C++ Compiler, 4.1 [Build 894]
For evaluation purposes only
Software supplied by: ARM Limited

5.0 versions output this:
Product: ARM Compiler 5.04
Component: ARM Compiler 5.04 update 2 (build 82)
Tool: armcc [5040081]
For evaluation purposes only
Software supplied by: ARM Limited

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

avcodec: Change get_pixels() to ptrdiff_t linesize

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

avutil/cpu: check av_parse_cpu_caps() table during cpu-test

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

avutil/cpu: Make cpu flag names match between cpu-test and av_parse_cpu_caps() tables

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

avutil/cpu: add aarch64 entries to 2nd table

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

jpeg2000: enable 4 component pixel formats

Bug-Id: 721
CC: libav-stable@libav.org
Sample-Id: 31230.mov

hevc_deblock: change tc type

The x86 asm expects int32_t so use that type.

Reviewed-by: Mickaël Raulet <mraulet@insa-rennes.fr>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

transcode_aac: fix const return value

Fixes Trac #3613.

Signed-off-by: Timothy Gu <timothygu99@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

doc/muxers: document which applications are known to need disable_chpl.

Thanks to: John Stebbins <stebbins@jetheaddev.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avutil/cpu: output cpu data to stdout

Errors go to stderr, but the cpu stats are non error output for cputest

This fixes echoing the cpu test results

Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

tests/fate/libavutil: switch cpu-test back to runecho so its results are vissible

(needs next commt too)

Reviewed-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avutil/pixelutils: check for malloc failure

Found-by: Daemon404
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avutil/pixelutils: avoid on stack arrays

The arrays are fairly large and could cause problems on some embedded systems
also they are not endian safe as they mix 32 and 8bit

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

avutil/cpu: fix cpu-test to work with ffmpegs cpuflags syntax

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

Merge commit 'a7153444df9040bf6ae103e0bbf6104b66f974cb'

* commit 'a7153444df9040bf6ae103e0bbf6104b66f974cb':
  huffyuvdec: check width size for yuv422p

Conflicts:
libavcodec/huffyuvdec.c

See: 6abb9a901fca27da14d4fffbb01948288b5da3ba
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '70cd3b8e659c3522eea5c16a65d14b8658894a94'

* commit '70cd3b8e659c3522eea5c16a65d14b8658894a94':
  mmvideo: check horizontal coordinate too

See: 8d3c99e825317b7efda5fd12e69896b47c700303
Merged-by: Michael Niedermayer <michaelni@gmx.at>

huffyuvdec: check width size for yuv422p

Avoid out of array accesses.

CC: libav-stable@libav.org
Bug-Id: CVE-2013-0848
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

mmvideo: check horizontal coordinate too

Fixes out of array accesses.

Bug-Id: CVE-2013-3672
Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>

avfilter/select: fix "t" eval variable comment

avfilter/select: remove optional avcodec dependency

avfilter/mp_decimate: remove avcodec dependency

avfilter/deshake: remove avcodec dependency

build: remove wrong avcodec dep in decimate

This was probably a mistake with mpdecimate. The decimate filter doesn't
depend on avcodec.

avutil: add pixelutils API

Merge commit '849b9d34c7ef70b370c53e7af3940f51cbc07d0f'

* commit '849b9d34c7ef70b370c53e7af3940f51cbc07d0f':
  wmalosslessdec: fix mclms_coeffs* array size

See: ec9578d54d09b64bf112c2bf7a34b1ef3b93dbd3
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '15201e256035a3e8f9d3d7b96fc327467e1a8ead'

* commit '15201e256035a3e8f9d3d7b96fc327467e1a8ead':
  configure: check $as first before using $gas as GNU as

Conflicts:
configure

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

Merge commit 'd395895cdb2ac8c95bd488549e7f893bd4dcc248'

* commit 'd395895cdb2ac8c95bd488549e7f893bd4dcc248':
  fate: generate tests/pixfmts.mak for all targets requiring it

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

avformat/iff: Fix "source comment"

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

avfilter/vf_noise: make constant things const

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

wmalosslessdec: fix mclms_coeffs* array size

Fixes corruption of context

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
CC:libav-stable@libav.org
Bug-Id: CVE-2014-2098
Signed-off-by: Anton Khirnov <anton@khirnov.net>

avfilter/avf_showwaves: New p2p mode for showwaves filter

Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '0897d2fdc7755849c3ae58d8b543ef734c200d3c'

* commit '0897d2fdc7755849c3ae58d8b543ef734c200d3c':
  movenc: Add option to disable nero chapters

Conflicts:
doc/muxers.texi

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

Merge commit '2601a9447efb778930b01ac58d2cc1d02498db4e'

* commit '2601a9447efb778930b01ac58d2cc1d02498db4e':
  mpegts: Add HEVC definitions

Conflicts:
libavformat/mpegts.h

See: b2880a3de9e3e160c47c069abcc8cb60015407b6
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '89616408e38ac7257e36976723df0e23d6ee1157'

* commit '89616408e38ac7257e36976723df0e23d6ee1157':
  mpegts: Define the section length with a constant

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

Merge commit '701e8b42e12ad625c64ceae2252acb1de390278c'

* commit '701e8b42e12ad625c64ceae2252acb1de390278c':
  vc-1: Optimise parser (with special attention to ARM)

Conflicts:
libavcodec/vc1_parser.c

See: a0d7f9ec9a10d935b83a47dbcb2cd85dcc007738
Merged-by: Michael Niedermayer <michaelni@gmx.at>

configure: check $as first before using $gas as GNU as

llvm's integrated assembler supports the AArch64 asm on darwin since
August 2014. So check $as first before using gas-preprocessor.pl via
$gas. Makes the checks specific for that the architecture specific asm
needs. PPC Altivec and AArch64 needs on ':vararg' for macro arguments.
Arm needs in addition the '.altmacro' directive.

Merge commit 'adf8227cf4e7b4fccb2ad88e1e09b6dc00dd00ed'

* commit 'adf8227cf4e7b4fccb2ad88e1e09b6dc00dd00ed':
  vc-1: Add platform-specific start code search routine to VC1DSPContext.

Conflicts:
configure
libavcodec/arm/vc1dsp_init_arm.c
libavcodec/vc1dsp.c
libavcodec/vc1dsp.h

See: 9d8ecdd8ca6d248e7439e8fdf255e39eda14e0f2
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'db7f1c7c5a1d37e7f4da64a79a97bea1c4b6e9f8'

* commit 'db7f1c7c5a1d37e7f4da64a79a97bea1c4b6e9f8':
  h264: Move start code search functions into separate source files.

Conflicts:
libavcodec/arm/Makefile
libavcodec/arm/h264dsp_init_arm.c
libavcodec/h264_parser.c
libavcodec/h264dsp.c
libavcodec/startcode.c
libavcodec/startcode.h

See: 270cede3f3772117454a14b620803d731036942d
Merged-by: Michael Niedermayer <michaelni@gmx.at>

fate: generate tests/pixfmts.mak for all targets requiring it

All subtargets which should run the fate-filter-pixdesc% need to
generate and include tests/pixfmts.mak. Most noteable missing target was
fate itself.

movenc: Add option to disable nero chapters

And add flag to muxer documentation.
Nero chapters break some taggers (mp3tag and iTunes).

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

mpegts: Add HEVC definitions

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

x86/vp9lpf: use fewer instructions in SPLATB_MIX

Signed-off-by: James Almer <jamrial@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fix compile error on bfin.

After the removal of all Blackfin architecture optimizations in
http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b55d3bbeed375f7b74442c4dd274d116a3e3d2e1
some includes were left behind leading to a compile error:

CC      libavformat/adtsenc.o
In file included from ./libavcodec/get_bits.h:35,
                 from ./libavcodec/ac3_parser.h:27,
                 from libavformat/ac3dec.c:23:
./libavcodec/mathops.h:43:29: error: bfin/mathops.h: No such file or directory

This compile error was found by buildroot autobuild system:
http://autobuild.buildroot.net/results/ae0/ae056f267e907091d09d2a1546d6f1ae02fa23b9/

Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

swscale/swscale_internal: fix ya16 -> rgba

Fixes Ticket644

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

lavd/avfoundation: Add device category.

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

Merge commit 'eee813eec7d3c0b0689f80665d3f796401742935'

* commit 'eee813eec7d3c0b0689f80665d3f796401742935':
  fate: Only generate tests/pixfmts.mak if some pixfmts fate test is run

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

Merge commit '990e2f3555bb0adae122d3599938ddcfee46e2d1'

* commit '990e2f3555bb0adae122d3599938ddcfee46e2d1':
  avcodec: Suppress deprecation warnings from DTG code scheduled for removal

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

Merge commit '086a541857812b2e0435db38fbabfddc7d19e53c'

* commit '086a541857812b2e0435db38fbabfddc7d19e53c':
  fate: png-suite

Conflicts:
tests/fate/image.mak

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

swscale/swscale_unscaled: fix ya16 input

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

Merge commit '60cbd6ad84de0931314030b81df87b69d6196587'

* commit '60cbd6ad84de0931314030b81df87b69d6196587':
  tiff: support reading gray+alpha at 8 bits

Conflicts:
libavcodec/tiff.c

See: f746f3790168bf898b629942868bda703e7f1045
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'bcc5f69b33e27b5e11aaea8304ee02f8d895cdab'

* commit 'bcc5f69b33e27b5e11aaea8304ee02f8d895cdab':
  tiff: support reading gray+alpha at 16 bits

Conflicts:
libavcodec/tiff.c

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

Merge commit 'e64f0bf2d2b1347ec9461f0e82852a62e8c6ffbe'

* commit 'e64f0bf2d2b1347ec9461f0e82852a62e8c6ffbe':
  png: support reading gray+alpha at 16 bits

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

Merge commit '2257165bff243534982f4ddabae4e65e2a35f2ab'

* commit '2257165bff243534982f4ddabae4e65e2a35f2ab':
  png: disable broken MMX/SIMD code for bpp <= 2

See: d1c48025a71e9e190d79115759c10f5807f4d0a7
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit 'f84a1b597c29dc035b8d5529ef88c2d7ff057820'

* commit 'f84a1b597c29dc035b8d5529ef88c2d7ff057820':
  swscale: support AV_PIX_FMT_YA16 as input

Conflicts:
libswscale/swscale_unscaled.c

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

mpegts: Define the section length with a constant

The specification says the value is expressed in 10 bits including
the 4-byte CRC.

vc-1: Optimise parser (with special attention to ARM)

The previous implementation of the parser made four passes over each input
buffer (reduced to two if the container format already guaranteed the input
buffer corresponded to frames, such as with MKV). But these buffers are
often 200K in size, certainly enough to flush the data out of L1 cache, and
for many CPUs, all the way out to main memory. The passes were:

1) locate frame boundaries (not needed for MKV etc)
2) copy the data into a contiguous block (not needed for MKV etc)
3) locate the start codes within each frame
4) unescape the data between start codes

After this, the unescaped data was parsed to extract certain header fields,
but because the unescape operation was so large, this was usually also
effectively operating on uncached memory. Most of the unescaped data was
simply thrown away and never processed further. Only step 2 - because it
used memcpy - was using prefetch, making things even worse.

This patch reorganises these steps so that, aside from the copying, the
operations are performed in parallel, maximising cache utilisation. No more
than the worst-case number of bytes needed for header parsing is unescaped.
Most of the data is, in practice, only read in order to search for a start
code, for which optimised implementations already existed in the H264 codec
(notably the ARM version uses prefetch, so we end up doing both remaining
passes at maximum speed). For MKV files, we know when we've found the last
start code of interest in a given frame, so we are able to avoid doing even
that one remaining pass for most of the buffer.

In some use-cases (such as the Raspberry Pi) video decode is handled by the
GPU, but the entire elementary stream is still fed through the parser to
pick out certain elements of the header which are necessary to manage the
decode process. As you might expect, in these cases, the performance of the
parser is significant.

To measure parser performance, I used the same VC-1 elementary stream in
either an MPEG-2 transport stream or a MKV file, and fed it through avconv
with -c:v copy -c:a copy -f null. These are the gperftools counts for
those streams, both filtered to only include vc1_parse() and its callees,
and unfiltered (to include the whole binary). Lower numbers are better:

                Before          After
File  Filtered  Mean   StdDev   Mean   StdDev  Confidence  Change
M2TS  No        861.7  8.2      650.5  8.1     100.0%      +32.5%
MKV   No        868.9  7.4      731.7  9.0     100.0%      +18.8%
M2TS  Yes       250.0  11.2     27.2   3.4     100.0%      +817.9%
MKV   Yes       149.0  12.8     1.7    0.8     100.0%      +8526.3%

Yes, that last case shows vc1_parse() running 86 times faster! The M2TS
case does show a larger absolute improvement though, since it was worse
to begin with.

This patch has been tested with the FATE suite (albeit on x86 for speed).

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

vc-1: Add platform-specific start code search routine to VC1DSPContext.

Initialise VC1DSPContext for parser as well as for decoder.
Note, the VC-1 code doesn't actually use the function pointer yet.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

h264: Move start code search functions into separate source files.

This permits re-use with parsers for codecs which use similar start codes.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Merge commit 'e9abafca278f87e1ecc6d50091d99b73ff63c9a6'

* commit 'e9abafca278f87e1ecc6d50091d99b73ff63c9a6':
  avutil: add AV_PIX_FMT_YA16 pixel format

Conflicts:
libavutil/pixdesc.c
libavutil/pixfmt.h
libavutil/version.h

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

Merge commit 'e96c3b81cadd0ba84d43b1f3a54980df3785d9a5'

* commit 'e96c3b81cadd0ba84d43b1f3a54980df3785d9a5':
  avutil: rename AV_PIX_FMT_Y400A to AV_PIX_FMT_YA8

Conflicts:
libavcodec/libopenjpegdec.c
libavcodec/libopenjpegenc.c
libavcodec/raw.c
libavutil/pixdesc.c
libavutil/pixfmt.h
libavutil/version.h
libswscale/swscale_internal.h
libswscale/swscale_unscaled.c

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

Merge commit 'e0d73fda361729d207e51fa2afd8ba6102e58d98'

* commit 'e0d73fda361729d207e51fa2afd8ba6102e58d98':
  avutil: add alias names for gray 8/16 colour spaces

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

Merge commit 'd2962e9f89cca6ff40f0c9d5ffc9c4397b8b1b26'

* commit 'd2962e9f89cca6ff40f0c9d5ffc9c4397b8b1b26':
  pixdesc: Support pixelformat aliases

Conflicts:
doc/APIchanges
libavutil/pixdesc.c
libavutil/version.h

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

Merge commit '5420099cab1e915b191cceccec4364f54cec6e52'

* commit '5420099cab1e915b191cceccec4364f54cec6e52':
  swscale: correctly pad destination buffer in rgb conversion

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

fate: Only generate tests/pixfmts.mak if some pixfmts fate test is run

avcodec: Suppress deprecation warnings from DTG code scheduled for removal

avformat/util: change av_find_default_stream_index() to use a score based system

Disfavor video streams with unknown resolution and no packets
Fixes seeking in audio-only-speex.flv

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

x86: hevc_deblock: remove unnecessary masking

The unpacks/shuffles later on makes it unnecessary.

Before:
1508 decicycles in h, 2096759 runs, 393 skips
2512 decicycles in v, 2095422 runs, 1730 skips

After:
1477 decicycles in h, 2096745 runs, 407 skips
2484 decicycles in v, 2095297 runs, 1855 skips

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

x86/hevc_mc: use fewer instructions in hevc_put_hevc_{uni, bi}_w[24]_{8, 10, 12}

Signed-off-by: James Almer <jamrial@gmail.com>
Reviewed-by: Christophe Gisquet <christophe.gisquet@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

x86/hevc_mc: remove an unnecessary pxor

Signed-off-by: James Almer <jamrial@gmail.com>
Reviewed-by: Mickaël Raulet <mraulet@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '23dfa00b88fc927d4c1854ab4fc60f5c6398f3ac'

* commit '23dfa00b88fc927d4c1854ab4fc60f5c6398f3ac':
  fate: explicitly set the default THREADS value

Conflicts:
tests/Makefile

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

Merge commit 'eb9244f20210fd420fb9b3c98126f9cae525d1cc'

* commit 'eb9244f20210fd420fb9b3c98126f9cae525d1cc':
  Add Icecast protocol

Conflicts:
Changelog
configure
doc/protocols.texi
libavformat/icecast.c
libavformat/version.h

See: e3dc2c86fc4178b100484c54f12c88705cdf6724
Merged-by: Michael Niedermayer <michaelni@gmx.at>

fate: png-suite

Every supported format is converted to RGB.

tiff: support reading gray+alpha at 8 bits

Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

tiff: support reading gray+alpha at 16 bits

png: support reading gray+alpha at 16 bits

png: disable broken MMX/SIMD code for bpp <= 2

The decoder was producing different results when ASM was disabled.
Based on a long debug session with Kostya.