While this technically compiles in current ffmpeg, this is only
because ffmpeg is compiled in strict ISO C mode, which disables
the builtin 'vector' keyword for AltiVec/VSX. Instead this gets
replaced with a macro inside altivec.h, which defines vector to
be actually __vector, which accepts random types.
Normally, the vector keyword should be used only with plain
scalar non-typedef types, such as unsigned int. But we have the
vec_(s|u)(8|16|32) macros, which can be used in a portable manner,
in util_altivec.h in libavutil.
This is also consistent with other AltiVec/VSX code elsewhere in
the tree.
Fixes#7861.
Signed-off-by: Daniel Kolesa <daniel@octaforge.org>
Signed-off-by: Lauri Kasanen <cand@gmx.com>
The implementation is pretty straight-forward. Most of the existing
NV12 codepaths work regardless of subsampling and are re-used as is.
Where necessary I wrote the slightly different NV24 versions.
Finally, the one thing that confused me for a long time was the
asm specific x86 path that did an explicit exclusion check for NV12.
I replaced that with a semi-planar check and also updated the
equivalent PPC code, which Lauri kindly checked.
This function wouldn't benefit from VSX instructions, so I put it
under altivec.
./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt grayf32le \
-f null -vframes 100 -v error -nostats -
3743 UNITS in planar1, 65495 runs, 41 skips
-cpuflags 0
23511 UNITS in planar1, 65530 runs, 6 skips
grayf32be
4647 UNITS in planar1, 65449 runs, 87 skips
-cpuflags 0
28608 UNITS in planar1, 65530 runs, 6 skips
The native speedup is 6.28133, and the bswapping one 6.15623.
Fate passes, each format tested with an image to video conversion.
Signed-off-by: Lauri Kasanen <cand@gmx.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
Passes fate on LE (with "lavc/jrevdct: Avoid an aliasing violation" applied).
Signed-off-by: Lauri Kasanen <cand@gmx.com>
Tested-by: Michael Kostylev on BE
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
./ffmpeg_g -f rawvideo -pix_fmt rgb24 -s hd1080 -i /dev/zero -pix_fmt yuv420p \
-f null -vframes 100 -v error -nostats -
1158 UNITS in planar1, 65528 runs, 8 skips
-cpuflags 0
19082 UNITS in planar1, 65533 runs, 3 skips
16.48 speedup ratio. On x86, SSE2 is ~7. Curiously, the Power C version
takes as many cycles as the x86 SSE2 version, yikes it's fast.
Note that this function uses VSX instructions, but is not marked so.
This is because several existing functions also make that mistake.
I'll submit a patch moving them once this is reviewed.
Signed-off-by: Lauri Kasanen <cand@gmx.com>
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
GCC tool had a bug of PPC intrinsic interpret, which has been fixed in GCC 4.9.1. This bug lead to
errors in two of our previous patches. We found this when we update our GCC tools to 4.9.1 and by
reading the related info on GCC website. We fix our previous error in two separate commits
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
add marcos GET_LS() GET_VF() LOAD_FILTER() LOAD_L1() GET_VF4() FIRST_LOAD() UPDATE_PTR() LOAD_SRCV() LOAD_SRCV8() GET_VFD() for POWER LE
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
yuv4mpeg: return proper error codes.
Give all anonymously typedeffed structs in headers a name
fate: Add parseutils test
parseutils-test: Drop random colors from parsing test
vf_pad/scale: use double precision for aspect ratios.
build: error on variable-length arrays
ppc: swscale: rework yuv2planeX_altivec()
ppc: fmtconvert: kill VLA in float_to_int16_interleave_altivec()
x86: dsputil: kill VLA in gmc_mmx()
libspeexenc: Updated commentary to reflect recent changes
libspeexenc: Add an option for enabling DTX
doc/APIchanges: fill in missing dates and hashes.
lavr: bump major to 1 and declare it stable.
lavr: change the type of the data buffers to uint8_t**.
lavc: deprecate the audio resampling API.
Conflicts:
cmdutils.h
configure
doc/APIchanges
ffplay.c
libavcodec/dwt.h
libavcodec/libspeexenc.c
libavfilter/vf_pad.c
libavfilter/vf_scale.c
libavformat/asf.h
tests/fate/libavutil.mak
tests/ref/fate/parseutils
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This gets rid of the variable-length scratch buffer by filtering 16
pixels at a time and writing directly to the destination. The extra
loads this requires to load the source values are compensated by not
doing a round-trip to memory before shifting.
Signed-off-by: Mans Rullgard <mans@mansr.com>
* qatar/master:
v410dec: Implement explode mode support
zerocodec: fix direct rendering.
wav: init st to NULL to avoid a false-positive warning.
wavpack: set bits_per_raw_sample for S32 samples to properly identify 24-bit
h264: refactor NAL decode loop
RTMPTE protocol support
RTMPE protocol support
rtmp: Add ff_rtmp_calc_digest_pos()
rtmp: Rename rtmp_calc_digest to ff_rtmp_calc_digest and make it global
swscale: add missing HAVE_INLINE_ASM check.
lavfi: place x86 inline assembly under HAVE_INLINE_ASM.
vc1: Add a test for interlaced field pictures
swscale: Mark all init functions as av_cold
swscale: x86: Drop pointless _mmx suffix from filenames
lavf: use conditional notation for default codec in muxer declarations.
swscale: place inline assembly bilinear scaler under HAVE_INLINE_ASM.
dsputil: ppc: cosmetics: pretty-print
dsputil: x86: add SHUFFLE_MASK_W macro
configure: respect CC_O setting in check_cc
Conflicts:
Changelog
configure
libavcodec/v410dec.c
libavcodec/zerocodec.c
libavformat/asfenc.c
libavformat/version.h
libswscale/utils.c
libswscale/x86/swscale.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
mpegvideo_enc: only allocate output packet when we know there will be output
Add names for more channel layouts to the channel layout map.
sunrast: Add a sample request for RMP_RAW colormap.
avcodec: do not override pts or duration from the audio encoder
Add prores regression test.
Enable already existing rso regression test.
Add regression test for "sox" format muxer/demuxer.
Add dpx encoding regression test.
swscale: K&R formatting cosmetics for PowerPC code (part I/II)
img2: Use ff_guess_image2_codec(filename) shorthand where appropriate.
Clarify licensing information about files borrowed from libjpeg.
Mark mutable static data const where appropriate.
avplay: fix -threads option
dvbsubdec: avoid undefined signed left shift in RGBA macro
mlpdec: use av_log_ask_for_sample()
gif: K&R formatting cosmetics
png: make .long_name more descriptive
movdec: Adjust keyframe flagging in fragmented files
rv34: change most "int stride" into "ptrdiff_t stride".
Conflicts:
avprobe.c
ffplay.c
libavcodec/mlpdec.c
libavcodec/mpegvideo_enc.c
libavcodec/pngenc.c
libavcodec/x86/v210-init.c
libavfilter/vf_boxblur.c
libavfilter/vf_crop.c
libavfilter/vf_drawtext.c
libavfilter/vf_lut.c
libavfilter/vf_overlay.c
libavfilter/vf_pad.c
libavfilter/vf_scale.c
libavfilter/vf_select.c
libavfilter/vf_setpts.c
libavfilter/vf_settb.c
libavformat/img2.c
libavutil/audioconvert.c
tests/codec-regression.sh
tests/lavf-regression.sh
tests/ref/lavf/dpx
tests/ref/vsynth1/prores
tests/ref/vsynth2/prores
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master: (22 commits)
wma: Clip WMA1 and WMA2 frame length to 11 bits.
movenc: Don't require frame_size to be set for modes other than mov
doc: Update APIchanges with info on muxer flushing
movenc: Reindent a block
tools: Remove some unnecessary #undefs.
rv20: prevent calling ff_h263_decode_mba() with unset height/width
tools: K&R reformatting cosmetics
Ignore generated aviocat and ismindex tools.
build: Automatically include architecture-specific library Makefile snippets.
indeo5: prevent null pointer dereference on broken files
pktdumper: Use usleep instead of sleep
cosmetics: Remove some unnecessary block braces.
Drop unnecessary prefix from *sink* variable and struct names.
Add a tool for creating smooth streaming manifests
movdec: Calculate an average bit rate for fragmented streams, too
movenc: Write the sample rate instead of time scale in the stsd atom
movenc: Add a separate ismv/isma (smooth streaming) muxer
movenc: Allow the caller to decide on fragmentation
libavformat: Add a flag for muxers that support write_packet(NULL) for flushing
movenc: Add support for writing fragmented mov files
...
Conflicts:
Changelog
cmdutils.c
cmdutils.h
doc/APIchanges
ffmpeg.c
ffplay.c
libavfilter/Makefile
libavformat/Makefile
libavformat/avformat.h
libavformat/movenc.c
libavformat/movenc.h
libavformat/version.h
tools/graph2dot.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
id3v2: fix doxy comment - 'machine byte order' makes no sense on char arrays
VC1: restore mistakenly removed code
twinvq: check output buffer size before decoding
twinvq: return an error when the packet size is too small
lavf: export some forgotten symbols with non-av prefixes.
swscale: update altivec yuv2planeX asm to new per-plane API.
swscale: make yuv2yuvX_10_sse2/avx 8/9/16-bits aware.
yuv2planeX10 SIMD
swscale: decide whether to use yuv2plane1/X on a per-plane basis.
swscale: reintroduce full precision in 16-bit output.
Split up yuv2yuvX functions
Split out yuv2yuv1 luma and chroma in order to make them generic DSP functions
lavc: replace references to deprecated AVCodecContext.error_recognition to use AVCodecContext.err_recognition
lavc: translate non-flag-based er options into flag-based ef options at codec open
add -err_filter AVOptions to access flag-based error recognition
h264_weight: initialize "height" function argument properly.
presets: spelling error in libvpx 1080p50_60
avplay: fix fullscreen behaviour with SDL 1.2.14 on Mac OS X
Conflicts:
ffplay.c
libavformat/libavformat.v
libswscale/swscale.c
libswscale/x86/swscale_template.c
tests/ref/lavfi/pixfmts_scale
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
ppc: fix some pointer to integer casts
ppc: fix 32-bit PIC build
vmdaudio: fix decoding of 16-bit audio format.
lavf: do not set codec_tag for rawvideo
h264: check for out of bounds reads in ff_h264_decode_extradata().
flvdec: Check for overflow before allocating arrays
avconv: use correct output stream index when checking max_frames
avconv: remove fake coded_frame on streamcopy hack
Conflicts:
avconv.c
libavcodec/h264.c
libavcodec/ppc/asm.S
libavcodec/vmdav.c
libavformat/flvdec.c
libavformat/utils.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Use uintptr_t instead of plain int. Without this change, the
comparisons will come out wrong for pointers in certain ranges.
Fixes random failures on ppc64. Also fixes some compiler warnings.
Signed-off-by: Mans Rullgard <mans@mansr.com>
* qatar/master: (23 commits)
h264: hide reference frame errors unless requested
swscale: split hScale() function pointer into h[cy]Scale().
Move clipd macros to x86util.asm.
avconv: reindent.
avconv: rescue poor abused start_time global.
avconv: rescue poor abused recording_time global.
avconv: merge two loops in output_packet().
avconv: fix broken indentation.
avconv: get rid of the arbitrary MAX_FILES limit.
avconv: get rid of the output_streams_for_file vs. ost_table schizophrenia
avconv: add a wrapper for output AVFormatContexts and merge output_opts into it
avconv: make itsscale syntax consistent with other options.
avconv: factor out adding input streams.
avconv: Factorize combining auto vsync with format.
avconv: Factorize video resampling.
avconv: Don't unnecessarily convert ipts to a double.
ffmpeg: remove unsed variable nopts
RV3/4 parser: remove unused variable 'off'
add XMV demuxer
rmdec: parse FPS in RealMedia properly
...
Conflicts:
avconv.c
libavformat/version.h
libswscale/swscale.c
tests/ref/fate/lmlm4-demux
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This allows using more specific implementations for chroma/luma, e.g.
we can make assumptions on filterSize being constant, thus avoiding
that test at runtime.
* qatar/master:
swscale: add dithering to yuv2yuvX_altivec_real
rv34: free+allocate buffer instead of reallocating it to preserve alignment
h264: add missing brackets.
swscale: use 15-bit intermediates for 9/10-bit scaling.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
It just does that part in scalar form, I doubt using a vector store
over 2 array would speed it up particularly.
The function should be written to not use a scratch buffer.
* qatar/master: (21 commits)
swscale: Add Doxygen for hyscale_fast/hScale.
fate: enable lavfi-pixmt tests on big endian systems
PPC: swscale: disable altivec functions for unsupported formats
fate: merge identical pixdesc_be/le tests
swscale: Add Doxygen for yuv2planar*/yuv2packed* functions.
build: call texi2pod.pl with full path instead of symlink
build: include sub-makefiles using full path instead of symlinks
swscale: update big endian reference values after dff5a835.
wavpack: skip blocks with no samples
cosmetics: remove outdated comment that is no longer true
build: replace some addprefix/addsuffix with substitution refs
avutil: Remove unused arbitrary precision integer code.
configure: Drop check for availability of ten assembler operands.
aacenc: Save channel configuration for later use.
aacenc: Fix codebook trellising for zeroed bands.
swscale: change prototypes of scaled YUV output functions.
swscale: re-add support for non-native endianness.
swscale: disentangle yuv2rgbX_c_full() into small functions.
swscale: split yuv2packed[12X]_c() remainders into small functions.
swscale: split yuv2packedX_altivec in smaller functions.
...
Conflicts:
Makefile
configure
libavcodec/x86/dsputil_mmx.c
libavfilter/Makefile
libavformat/Makefile
libavutil/integer.c
libavutil/integer.h
libswscale/swscale.c
libswscale/swscale_internal.h
libswscale/x86/swscale_template.c
tests/ref/lavfi/pixdesc_le
tests/ref/lavfi/pixfmts_scale
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Remove unused variables "flags" and "dstFormat" in yuv2packed1,
merge source rows per plane for yuv2packed[12], and make every
source argument int16_t (some where invalidly set to uint16_t).
This prevents stack pollution and is part of the Great Evil Plan
to simplify swscale.
This will likely lead to a considerable performance boost,
since it removes a branch from the inner loop. Part of the
Great Evil Plan to simplify swscale.
* qatar/master:
build: improve rules for test programs
build: factor out the .c and .S compile commands as a macro
swscale: remove unused xInc/srcW arguments from hScale().
H.264: disable 2tap qpel with CODEC_FLAG2_FAST and >8-bit
H.264: make filter_mb_fast support 4:4:4
mpeg4videoenc: Remove disabled variant of mpeg4_encode_block().
configure: allow post-fixed cpu strings for athlon64, k8, and opteron when setting the -march flag.
Move some variable declarations below the proper #ifdefs.
Conflicts:
Makefile
ffplay.c
libswscale/swscale.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Daniel Kolesa
Philip Langdale
Lauri Kasanen
Sergey Lavrushkin
Michael Niedermayer
Pedro Arthur
Rong Yan
Christophe Gisquet
Michael Niedermayer
Diego Biurrun
Anton Khirnov
Mans Rullgard
Ronald S. Bultje
Kieran Kunhya
Luca Barbato