swscale/utils: remove unused define

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

avformat/dvenc: Implement 32khz & 44.1khz for 25/50fps

30000/1001 fps isnt implemented as i dont know which are the correct interleaving parameters,
pointers to the appropriate spec and page are welcome

Fixes Ticket2597

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

avformat/seek: use av_malloc_array()

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

avfilter/avfilter: use av_malloc(z)_array()

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

avcodec/hevc_ps: Check abs_delta_rps

Fixes integer overflow

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

avcodec/hevc: check nb_cpb

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

hevc: use intreadwrite

When dealing with MVs, both components may be processed at a time.

On Win64, 560 to 539 cycles for derive_spatial_merge_candidates.

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

libavutil: document side effects of macros

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

hevc: report more precise progress

There's a lag of one CTB line for SAO behind deblocking filter, except for
last line. However, once SAO has been completed on a line, all its pixels,
i.e. up to y+ctb_size are filtered and ready to be used as reference.

Without SAO, when deblocking filter finishes a CTB line, only the bottom
bottom 4 pixels may be filtered when next CTB is process by the deblocing.

The await_progess for hevc then checks whether the bottom pixels of a PU
requires access beyond that point, so the reporting should effectively
report up to the the above limits.

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

Merge commit '1a583c0c60240adb8fa6620c6df33f1a0a0fe5d9'

* commit '1a583c0c60240adb8fa6620c6df33f1a0a0fe5d9':
  fdct: Move ppc-specific declarations to a header in the ppc directory

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

Merge commit '5dcc201505f71b1e73e9eef12ce89d4eed252ad0'

* commit '5dcc201505f71b1e73e9eef12ce89d4eed252ad0':
  simple_idct: Move x86-specific declarations to a header in the x86 directory

Conflicts:
libavcodec/x86/simple_idct.c

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

Merge commit '85cabb8d002f2cd100ced5cc17d87bfc9460d314'

* commit '85cabb8d002f2cd100ced5cc17d87bfc9460d314':
  fdct: Move x86-specific declarations to a header in the x86 directory

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

Merge commit 'd6e1d37100af568211f28ec0bcf7958a3a2a299e'

* commit 'd6e1d37100af568211f28ec0bcf7958a3a2a299e':
  oss_audio: Split muxer and demuxer

Conflicts:
libavdevice/Makefile
libavdevice/oss_audio.c

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

swscale/x86/hscale_fast_bilinear_simd: add inline asm guards

Should fix MSVC build

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

fdct: Move ppc-specific declarations to a header in the ppc directory

simple_idct: Move x86-specific declarations to a header in the x86 directory

fdct: Move x86-specific declarations to a header in the x86 directory

sws: move inline asm hscale to MMX_OBJS

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

sws/x86: split mmxext fast bilinear scaler out

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

sws: Move fast bilinear C code into seperate file

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

swscale/swscale_internal: add needed version.h

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

Merge commit 'e0a2e60c0a6cbcceef08e94af5081d2aa8e6a52f'

* commit 'e0a2e60c0a6cbcceef08e94af5081d2aa8e6a52f':
  dct-test: Reuse enum idct_permutation_type instead of duplicating it

Conflicts:
libavcodec/dct-test.c

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

oss_audio: use a macro to simplify ioctl() error checking

Also add a note about SNDCTL_DSP_GETFMTS which may fail even if OSS is
available.

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

hevc: fix transquant_bypass

cherry picked from commit 4a16cb2c70728a55d2fd723aff01b13ea259c4df
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

oss_audio: Split muxer and demuxer

Signed-off-by: Diego Biurrun <diego@biurrun.de>

hevc/rext: put a warning log message instead of an error log message

cherry picked from commit 243cb99cff727d6a14c32cdff2748f6c255dbaf4
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Parse dri when receiving jpg via rtp.

Based on a patch by contact at iridiummobile ru.
Fixes ticket #3780.

Write valid free atom when -moov_size is specified.

Fixes ticket #3769.

avcodec/hevc: fix pict_type

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

hevc/sao: optimze sao implementation

- adding one extra pixel all around the frame
- do not copy when SAO is not applied

5% improvement

cherry picked from commit 10fc29fc19a12c4d8168fbe1a954b76386db12d0
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '913fa85a2c1fcd71802f7decdef5f355e0d9c757'

* commit '913fa85a2c1fcd71802f7decdef5f355e0d9c757':
  dct-test: Skip indirection for MMX IDCT permutation

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

Merge commit '9e0b29911f1f167381a7dbdfca68bf417b8c767b'

* commit '9e0b29911f1f167381a7dbdfca68bf417b8c767b':
  x86: dnxhdenc: Eliminate some unnecessary ifdefs

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

Merge commit 'b4edbe9a0eb48423fbf249b0ca7fb62838bdb7f5'

* commit 'b4edbe9a0eb48423fbf249b0ca7fb62838bdb7f5':
  libgsm: Split decoder and encoder

Conflicts:
libavcodec/libgsmenc.c

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

Merge commit '8b0dd4942aac320d1ca3c40fa7ea1be342c71273'

* commit '8b0dd4942aac320d1ca3c40fa7ea1be342c71273':
  idctdsp: prettyprinting cosmetics

Conflicts:
libavcodec/idctdsp.c
libavcodec/ppc/idctdsp.c
libavcodec/x86/idctdsp_init.c

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

dct-test: Reuse enum idct_permutation_type instead of duplicating it

Merge commit 'b4987f72197e0c62cf2633bf835a9c32d2a445ae'

* commit 'b4987f72197e0c62cf2633bf835a9c32d2a445ae':
  idct: Convert IDCT permutation #defines to an enum

Conflicts:
libavcodec/idctdsp.c
libavcodec/x86/cavsdsp.c

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

Merge commit 'c0de9159a7ba5707aa0a5c2bc73ae78b7b87ec46'

* commit 'c0de9159a7ba5707aa0a5c2bc73ae78b7b87ec46':
  avdevice: Give names to anonymously typedeffed structs

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

Merge commit '24af1aa0f70362a66cda04c9d7cd012e019f5572'

* commit '24af1aa0f70362a66cda04c9d7cd012e019f5572':
  fft: Convert FFT/MDCT permutation type #defines to enums

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

Merge commit '746ad4e0df7faf93329804e412ec53c1d929a75b'

* commit '746ad4e0df7faf93329804e412ec53c1d929a75b':
  dct-test: Improve CPU flags struct member name

Conflicts:
libavcodec/dct-test.c

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

Merge commit 'cb44b21da1f59923be577f08c267ec270529be97'

* commit 'cb44b21da1f59923be577f08c267ec270529be97':
  dct-test: Move cpu_flags variable out of global scope

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

fate: Add test for wav Peak Envelope Chunk encoder (levl chunk).

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

Fix make checkheaders for libavcodec/proresdsp.h.

dct-test: Skip indirection for MMX IDCT permutation

x86: dnxhdenc: Eliminate some unnecessary ifdefs

libgsm: Split decoder and encoder

Signed-off-by: Diego Biurrun <diego@biurrun.de>

idctdsp: prettyprinting cosmetics

idct: Convert IDCT permutation #defines to an enum

Also rename the enum values to be consistent with other DCT permutations.

avdevice: Give names to anonymously typedeffed structs

Anonymous structs can cause trouble in header files, so try to
avoid them altogether as a matter of good style.

fft: Convert FFT/MDCT permutation type #defines to enums

dct-test: Improve CPU flags struct member name

dct-test: Move cpu_flags variable out of global scope

lavfi/drawtext: add alias "expr_int_format" to expansion function "eif"

Reviewed-by: Stefano Sabatini <stefasab@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

smacker: remove dead code

Signed-off-by: Timothy Gu <timothygu99@gmail.com>
Reviewed-by: Paul B Mahol <onemda@gmail.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '7e18a727d2c2a19f22fcf68875d1b05fd2eafcef'

* commit '7e18a727d2c2a19f22fcf68875d1b05fd2eafcef':
  arm: cosmetics: Consistently use lowercase for shift operators

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

Merge commit 'fe67f3fbb5f9f6a6b60f837f6bc5e087ac11f3bf'

* commit 'fe67f3fbb5f9f6a6b60f837f6bc5e087ac11f3bf':
  arm: cosmetics: Fix a misaligned asm operand

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

arm: cosmetics: Consistently use lowercase for shift operators

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

arm: cosmetics: Fix a misaligned asm operand

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

alpha/me_cmp: move code out of dsputil

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

alpha/idctdsp: move disabled cruft out of dsputil

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

avcodec/a64multienc use av_freep()

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

a64multienc: remove dead assignment

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

kerndeint: remove dead assignments

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

Merge commit '5a272190a04666f0fe41be767396b30712638c21'

* commit '5a272190a04666f0fe41be767396b30712638c21':
  armv6: Accelerate butterflies_float

See: 57641410d1a386937bec3fddd6c75119550916ec
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '5edad2c4a1f46bcc56be755af86ab355c2f1b37f'

* commit '5edad2c4a1f46bcc56be755af86ab355c2f1b37f':
  armv6: Accelerate vector_fmul_window

See: 649c666137f43542b45941f42034ab3f44a31d38
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '87552d54d3337c3241e8a9e1a05df16eaa821496'

* commit '87552d54d3337c3241e8a9e1a05df16eaa821496':
  armv6: Accelerate ff_fft_calc for general case (nbits != 4)

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

Merge commit '5c22e8e4ad0852d61d5c4ba8d67d33fd72339497'

* commit '5c22e8e4ad0852d61d5c4ba8d67d33fd72339497':
  armv6: Accelerate ff_imdct_half for general case (mdct_bits != 6)

See: 42c1cc35b7623ce76c7b55c6bc100f135e17cd4f
Merged-by: Michael Niedermayer <michaelni@gmx.at>

avcodec/dvenc: clear all dsp contexts

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

armv6: Accelerate butterflies_float

I benchmarked the result by measuring the number of gperftools samples that
hit anywhere in the AAC decoder (starting from aac_decode_frame()) or
specifically in butterflies_float_c() / ff_butterflies_float_vfp() for the
same sample AAC stream:

                   Before          After
                   Mean   StdDev   Mean   StdDev  Confidence  Change
Audio decode       1542.8 43.7     1470.5 41.5    100.0%      +4.9%
butterflies_float  130.0  11.9     70.2   12.1    100.0%      +85.2%

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

armv6: Accelerate vector_fmul_window

I benchmarked the result by measuring the number of gperftools samples that
hit anywhere in the AAC decoder (starting from aac_decode_frame()) or
specifically in vector_fmul_window_c() / ff_vector_fmul_window_vfp() for the
same sample AAC stream:

                    Before          After
                    Mean   StdDev   Mean   StdDev  Confidence  Change
Audio decode        1598.2 47.4     1529.2 25.4    100.0%      +4.5%
vector_fmul_window  244.0  22.1     188.9  22.3    100.0%      +29.2%

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

armv6: Accelerate ff_fft_calc for general case (nbits != 4)

The previous implementation targeted DTS Coherent Acoustics, which only
requires nbits == 4 (fft16()). This case was (and still is) linked directly
rather than being indirected through ff_fft_calc_vfp(), but now the full
range from radix-4 up to radix-65536 is available. This benefits other codecs
such as AAC and AC3.

The implementaion is based upon the C version, with each routine larger than
radix-16 calling a hierarchy of smaller FFT functions, then performing a
post-processing pass. This pass benefits a lot from loop unrolling to
counter the long pipelines in the VFP. A relaxed calling standard also
reduces the overhead of the call hierarchy, and avoiding the excessive
inlining performed by GCC probably helps with I-cache utilisation too.

I benchmarked the result by measuring the number of gperftools samples that
hit anywhere in the AAC decoder (starting from aac_decode_frame()) or
specifically in the FFT routines (fft4() to fft512() and pass()) for the
same sample AAC stream:

              Before          After
              Mean   StdDev   Mean   StdDev  Confidence  Change
Audio decode  2245.5 53.1     1599.6 43.8    100.0%      +40.4%
FFT routines  940.6  22.0     348.1  20.8    100.0%      +170.2%

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

armv6: Accelerate ff_imdct_half for general case (mdct_bits != 6)

The previous implementation targeted DTS Coherent Acoustics, which only
requires mdct_bits == 6. This relatively small size lent itself to
unrolling the loops a small number of times, and encoding offsets
calculated at assembly time within the load/store instructions of each
iteration.

In the more general case (codecs such as AAC and AC3) much larger arrays
are used - mdct_bits == [8, 9, 11]. The old method does not scale for
these cases, so more integer registers are used with non-unrolled versions
of the loops (and with some stack spillage). The postrotation filter loop
is still unrolled by a factor of 2 to permit the double-buffering of some
VFP registers to facilitate overlap of neighbouring iterations.

I benchmarked the result by measuring the number of gperftools samples
that hit anywhere in the AAC decoder (starting from aac_decode_frame())
or specifically in ff_imdct_half_c / ff_imdct_half_vfp, for the same
example AAC stream:

                  Before          After
                  Mean   StdDev   Mean   StdDev  Confidence  Change
aac_decode_frame  2368.1 35.8     2117.2 35.3    100.0%      +11.8%
ff_imdct_half_*   457.5  22.4     251.2  16.2    100.0%      +82.1%

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

avcodec/me_cmp: restore author attribution and copyrights

These where removed by libav in

See: git show -C 2d60444331fca1910510038dd3817bea885c2367
diff --git a/libavcodec/dsputil.c b/libavcodec/me_cmp.c
similarity index 98%
rename from libavcodec/dsputil.c
rename to libavcodec/me_cmp.c
index ba71a99..9fcc937 100644
--- a/libavcodec/dsputil.c
+++ b/libavcodec/me_cmp.c
@@ -1,8 +1,4 @@
 /*
- * DSP utils
- * Copyright (c) 2000, 2001 Fabrice Bellard
- * Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
- *
  * This file is part of Libav.
  *
  * Libav is free software; you can redistribute it and/or

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

Merge commit '2d60444331fca1910510038dd3817bea885c2367'

* commit '2d60444331fca1910510038dd3817bea885c2367':
  dsputil: Split motion estimation compare bits off into their own context

Conflicts:
configure
libavcodec/Makefile
libavcodec/arm/Makefile
libavcodec/dvenc.c
libavcodec/error_resilience.c
libavcodec/h264.h
libavcodec/h264_slice.c
libavcodec/me_cmp.c
libavcodec/me_cmp.h
libavcodec/motion_est.c
libavcodec/motion_est_template.c
libavcodec/mpeg4videoenc.c
libavcodec/mpegvideo.c
libavcodec/mpegvideo_enc.c
libavcodec/x86/Makefile
libavcodec/x86/me_cmp_init.c

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

Merge commit 'a578b0407dc983aecd72028e1127062689b67089'

* commit 'a578b0407dc983aecd72028e1127062689b67089':
  configure: Assume runtime cpu detection on arm on --target-os=android as well

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

Merge commit 'c23ce454b3e33634a188d6facfd2b7182af5af93'

* commit 'c23ce454b3e33634a188d6facfd2b7182af5af93':
  x86: dsputil: Coalesce all init files

Conflicts:
libavcodec/x86/dsputil_init.c
libavcodec/x86/dsputil_x86.h
libavcodec/x86/motion_est.c

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

lavd/x11grab: reindent after last commit.

lavfi: check refcount before merging.

When merging the formats around the automatically inserted
convert filters, the refcount of the format lists can not be 0.
Coverity does not detect it, and suspects a memory leak,
because if refcount is 0 the newly allocated lists are not
stored anywhere. That gives CIDs 1224282, 1224283 and 1224284.
Lists with refcount 0 are used in can_merge_formats(), so the
asserts can not be moved inside the merge functions.

lavd/x11grab: add an option to disable MIT-SHM.

With remote displays supporting the MIT-SHM extension,
the extension is detected and used, but attaching fails
asynchronously.

lavd/x11grab: check 32-bits color masks.

The X11 servers by VNC, at 32-bits depths, has the following masks:
R:0x000007ff G:0x003ff800 B:0xffc00000
This is not compatible with AV_PIX_FMT_0RGB32, and the result
is success with completely wrong colors.

lavd/x11grab: change error code for unsupported visuals.

lavd/x11grab: disable drawing mouse without XFixes.

Fix a segfault if the XFixes extension is not available on
the X11 server.
Can be reproduced using the VNC server.

dsputil: Split motion estimation compare bits off into their own context

avcodec/resample: Use av_m/realloc_array()

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

avcodec/resample2: Use av_malloc(z)_array()

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

ffserver: use av_mallocz_array()

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

lavf/segment: set segment end time when the first packet arrives

Avoid negative durations in case there is a single packet in the current
segment, since in that case the end time is still set to the previous
segment end time.

lavf/segment: fix weird indent

lavf/segment: show incoming packet duration in debug message

lavf/segment: only use reference frames for computing the segment end time

This avoids a systematic overestimate of the segments duration when there
are several streams.

Fix trac ticket #3724.

configure: Assume runtime cpu detection on arm on --target-os=android as well

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

avformat/dv: implement fallback in dv_extract_pack()

Fixes Ticket2340
Fixes Ticket2341

Based-on mail from Dave Rice <dave@dericed.com>
Tested-by: Dave Rice <dave@dericed.com>
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '324ff59444ff5470bb325ff1e2be7c4b054fc944'

* commit '324ff59444ff5470bb325ff1e2be7c4b054fc944':
  avpacket: Check for and return errors in ff_interleave_add_packet()

Conflicts:
libavformat/audiointerleave.c
libavformat/internal.h
libavformat/mux.c

See: 4d7c71c36467331f1e0c0f17af9f371d33308a9c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

Merge commit '2db953f84671997e936f91140ffb5143c1537844'

* commit '2db953f84671997e936f91140ffb5143c1537844':
  h264: K&R formatting cosmetics

Conflicts:
libavcodec/h264.c
libavcodec/h264_direct.c
libavcodec/h264_slice.c

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

Merge commit 'a11ef610eec5241e226efbd1c5bb5eb45c88d2c2'

* commit 'a11ef610eec5241e226efbd1c5bb5eb45c88d2c2':
  h264: Remove some commented-out, broken cruft

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

doc/filters/drawtext: simplify example, use clip() eval function

lavfi/drawtext: add fontcolor_expr option

Allow to dynamically evaluate the font color.

Signed-off-by: Stefano Sabatini <stefasab@gmail.com>

x86: dsputil: Coalesce all init files

This makes the init files match the structure of the dsputil split.

avpacket: Check for and return errors in ff_interleave_add_packet()

Signed-off-by: Diego Biurrun <diego@biurrun.de>

h264: K&R formatting cosmetics

Signed-off-by: Diego Biurrun <diego@biurrun.de>
Signed-off-by: Vittorio Giovara <vittorio.giovara@gmail.com>

lavfi/drawtext: add expansion function eif()

It evaluates expression and outputs it as integer value, using specified
format.

Address trac ticket #3699.

Signed-off-by: Stefano Sabatini <stefasab@gmail.com>