coffee/libyuv
1
0
Fork 0
mirror of https://chromium.googlesource.com/libyuv/libyuv synced 2025-12-06 16:56:55 +08:00
Code Issues Packages Projects Releases Wiki Activity
1,597 Commits 3 Branches 0 Tags BSD-3-Clause 60 MiB
Commit Graph

335 Commits

Author SHA1 Message Date
Frank Barchard
23c6a83561 Fix ifdef mismatch for mirroruv 
Macro define and macro ifdef didnt match, leading to C code
being used.  Make macro match function name.

TBR=harryjin@google.com
BUG=libyuv:543

Review URL: https://codereview.chromium.org/1579023002 .
 2016-01-11 16:33:36 -08:00
Frank Barchard
3f4d86053e avx2 interpolate use 8 bit 
BUG=libyuv:535
R=dhrosa@google.com

Review URL: https://codereview.chromium.org/1535833003 .
 2015-12-21 10:57:32 -08:00
Frank Barchard
f4447745ae Add rounding to InterpolateRow for improved quality and consistency. 
Remove inaccurate specializations for 1/4 and 3/4, since they round
incorrectly.  Specialize for 100% and 50% are kept due to performance.
Make C and ARM code match SSSE3.
Make unittests expect zero difference.

BUG=libyuv:535
R=harryjin@google.com

Review URL: https://codereview.chromium.org/1533643005 .
 2015-12-17 15:24:06 -08:00
Frank Barchard
b3bbcc1f4e add ifdef for AVX2 so vs2010 can still compile 
R=harryjin@google.com
BUG=libyuv:531

Review URL: https://codereview.chromium.org/1515503005 .
 2015-12-09 15:23:51 -08:00
Frank Barchard
dee77a4ebe Optimize yuv alpha blend AVX2 code to do 32 pixels at time. 
out/Release/libyuv_unittest --libyuv_width=1280 --libyuv_height=720 --libyuv_repeat=9999 --libyuv_flags=-1 --gtest_filter=*I420Blend_Opt

Was LibYUVPlanarTest.I420Blend_Opt (2335 ms)
Now LibYUVPlanarTest.I420Blend_Opt (1937 ms)

vs SSSE3
LibYUVPlanarTest.I420Blend_Opt (2599 ms)

BUG=libyuv:527
R=dhrosa@google.com

Review URL: https://codereview.chromium.org/1505673003 .
 2015-12-08 18:20:30 -08:00
Frank Barchard
bea690b3e0 AVX2 YUV alpha blender and improved unittests 
AVX2 version can process 16 pixels at a time for improved memory bandwidth and fewer instructions.

unittests improved to test unaligned memory, and test exactness when alpha is 0 or 255.

R=dhrosa@google.com, harryjin@google.com
BUG=libyuv:527

Review URL: https://codereview.chromium.org/1505433002 .
 2015-12-05 22:23:29 -08:00
Frank Barchard
8af0ebf816 planar blend use signed images 
R=dhrosa@google.com, harryjin@google.com, jzern@chromium.org
BUG=libyuv:527

Review URL: https://codereview.chromium.org/1491533002 .
 2015-12-02 14:20:17 -08:00
Frank Barchard
5eefbe2330 Fix for drmemory failure on I411ToARGB 
Before
I420ToARGB_Opt (594 ms)
I422ToARGB_Opt (483 ms)
I411ToARGB_Opt (748 ms) ***
I444ToARGB_Opt (452 ms)
I400ToARGB_Opt (218 ms)

After
I420ToARGB_Opt (591 ms)
I422ToARGB_Opt (454 ms)
I411ToARGB_Opt (502 ms)  ***
I444ToARGB_Opt (441 ms)
I400ToARGB_Opt (216 ms)

TBR=harryjin@google.com
BUG=libyuv:525

Review URL: https://codereview.chromium.org/1459513002 .
 2015-11-17 18:00:52 -08:00
Frank Barchard
0815568a50 test for unaligned vs aligned for CopyRow_SSE2 
improves performance on older CPUs where movdqa is faster.
TBR=harryjin@google.com
BUG=libyuv:492

Review URL: https://codereview.chromium.org/1455463002 .
 2015-11-17 00:04:03 -08:00
Frank Barchard
ce4c2fad1d Raw 24 bit RGB to RGB24 (bgr) 
Add unittests that do 1 step conversion vs 2 step conversion.

Tests end swapping versions match direct conversions.

R=harryjin@google.com
BUG=libyuv:518

Review URL: https://codereview.chromium.org/1419103007 .
 2015-11-03 10:30:30 -08:00
Frank Barchard
87926cec8b remove store bgra, abgr, raw unused macros 
TBR=harryjin@google.com
BUG=libyuv:518

Review URL: https://codereview.chromium.org/1420033004 .
 2015-11-02 10:40:03 -08:00
Frank Barchard
2c7aa0070a remove I422ToBGRA and use I422ToRGBA internally 
Removes low levels for I420ToBGRA and I420ToRAW and reimplements them as I420ToRGBA and I420ToRGB24 with transposed color matrix.

Adds unittests that do 1 step conversion vs 2 steps to test end swapping versions match direct conversions.

R=harryjin@google.com
BUG=libyuv:518

Review URL: https://codereview.chromium.org/1427993004 .
 2015-11-02 10:24:12 -08:00
Frank Barchard
5d97b93369 refactor I420ToABGR to use I420ToARGBRow 
Using a transposed conversion matrix, I420ToARGB can output ABGR.

R=harryjin@google.com, xhwang@chromium.org
BUG=libyuv:473

Review URL: https://codereview.chromium.org/1413573010 .
 2015-10-30 11:56:57 -07:00
Frank Barchard
b86dbf24d3 refactor I420AlphaToABGR to use I420AlphaToARGB internally 
swap U and V and transpose conversion matrix, so I420AlphaToARGB and
I420AlphaToABGR share low level code.

Having less code with same performance allows more focused
optimization for future ARM versions.

R=harryjin@google.com
TBR=harryjin@chromium.org
BUG=libyuv:473,libyuv:516

Review URL: https://codereview.chromium.org/1422263002 .
 2015-10-27 14:17:21 -07:00
Frank Barchard
cf160cdbaa implement I444ToABGR by swapping uv and transpose matrix 
U contributes to B and G.  V contributes to R and G.
By swapping U and V, they contribute to the opposite channels.  Adjust the matrix so the U contribution is in the matrix location such that it till contribute to the
new B channel and vice versa.
This allows ABGR versions of YUV conversion to use the same low level code as ARGB, just using a different matrix and swapping U and V pointers.

As a result the existing I444ToABGRRow functions are no longer needed and are removed.

Previously this function was only Intel AVX2 optimized for Windwos.  Now it is also optimized for Arm and GCC.

ARMv7 Neon
Was LibYUVConvertTest.I444ToABGR_Opt (75971 ms)
Now LibYUVConvertTest.I444ToABGR_Opt (3672 ms)
20.6 times faster.

R=xhwang@chromium.org
BUG=libyuv:515

Review URL: https://codereview.chromium.org/1414133006 .
 2015-10-27 10:21:21 -07:00
Frank Barchard
2e4466e282 change all pix parameters to width for consistency 
TBR=harryjin@google.com
BUG=none

Review URL: https://codereview.chromium.org/1398633002 .
 2015-10-07 22:30:36 -07:00
Frank Barchard
76a599ec3b fix jpeg and bt.709 yuvconstants for neon64. 
yuv constants for bt.601 were previously ported to neon64, as well
as the code to respect other color spaces.  But the jpeg and bt.709
colour conversion constants were still in armv7 form.  This changes
the constants for aarch64 builds to be compatible with the code.

yuv constants are now passed as const *

Remove Yvu constants which were used for older version on nv21 but not new code.

TBR=harryjin@google.com
BUG=none

Review URL: https://codereview.chromium.org/1398623002 .
 2015-10-07 19:46:56 -07:00
Frank Barchard
cc89e3a77b port ARGB to 565 dithering SSE2 code to GCC. 
Previously the assembly code was only available to Windows.
This CL ports the SSE2 code to GCC syntax.

When running a profiler on all the unittests, this function
was the slowest of all functions that still ran in C code.
   3.71%  libyuv_unittest  libyuv_unittest      [.] ARGBToRGB565DitherRow_C

Was
ARGBToRGB565Dither_Opt (2894 ms)
Now
ARGBToRGB565Dither_Opt (432 ms)

TBR=harryjin@google.com
BUG=libyuv:492

Review URL: https://codereview.chromium.org/1397673002 .
 2015-10-07 18:24:50 -07:00
Frank Barchard
914a9856c7 Reimplement NV21ToARGB to allow different color matrix. 
Low level for NV21ToARGB written to accept yuv matrix used by
other YUV to ARGB functions.
Previously NV21 was implemented for Windows using NV12 with a different
matrix that swapped U and V.  But the Arm version of the low level does
not allow the matrix U and V contributions to be swapped.
Using a new low level function that reads NV21 and uses the same
yuvconstants as other YUV conversion functions allows an Arm port of
this function.

TBR=harryjin@google.com
BUG=libyuv:500

Review URL: https://codereview.chromium.org/1388273002 .
 2015-10-06 20:34:44 -07:00
Frank Barchard
2cc1a2b233 Remove sse2 functions that also have ssse3 
ARGBBlendRow_SSE2, ARGBAttenuateRow_SSE2, and MirrorRow_SSE2
Since vast majority of CPUs have SSSE3 now, removing the SSE2
improves the performance of CPU dispatching.

R=harryjin@google.com
BUG=none

Review URL: https://codereview.chromium.org/1377053003 .
 2015-09-30 14:24:44 -07:00
Frank Barchard
febc26a2c9 win64 version of I422AlphaToARGB. 
Was
I420AlphaToARGB_Premult (8861 ms)
I420AlphaToARGB_Opt (7119 ms)
Now
I420AlphaToABGR_Premult (2840 ms)
I420AlphaToARGB_Opt (484 ms)

C function switched to 1 step.
Was
I420AlphaToARGB_Premult (8862 ms)
I420AlphaToABGR_Opt (6718 ms)

Now
I420AlphaToARGB_Premult (8706 ms)
I420AlphaToARGB_Opt (6541 ms)

R=harryjin@google.com
BUG=libyuv:496, libyuv:473

Review URL: https://codereview.chromium.org/1359183003 .
 2015-09-25 15:06:41 -07:00
Frank Barchard
9a0e12f5f1 AVX2 1 step I422AlphaToARGB for gcc and win. 
C     I420AlphaToARGB_Opt (5169 ms)
SSSE3 I420AlphaToARGB_Opt (432 ms)
AVX2  I420AlphaToARGB_Opt (358 ms)

and with premultiplication as 2 step process:
I420AlphaToARGB_Premult (7029 ms)
I420AlphaToARGB_Premult (757 ms)
I420AlphaToARGB_Premult (508 ms)

R=harryjin@google.com
BUG=libyuv:496,libyuv:473

Review URL: https://codereview.chromium.org/1372653003 .
 2015-09-25 13:37:42 -07:00
Frank Barchard
e365cdde3b I420Alpha row function in 1 pass. 
API change - I420AlphaToARGB takes flag indicating if RGB should be
premultiplied by alpha.

This version implements an efficient SSSE3 version for Windows.
C version done in 2 steps.

Was
libyuvTest.I420AlphaToARGB_Any (1136 ms)
libyuvTest.I420AlphaToARGB_Unaligned (1210 ms)
libyuvTest.I420AlphaToARGB_Invert (966 ms)
libyuvTest.I420AlphaToARGB_Opt (1031 ms)
libyuvTest.I420AlphaToABGR_Any (1020 ms)
libyuvTest.I420AlphaToABGR_Unaligned (1359 ms)
libyuvTest.I420AlphaToABGR_Invert (1082 ms)
libyuvTest.I420AlphaToABGR_Opt (986 ms)

R=harryjin@google.com
BUG=libyuv:496

Review URL: https://codereview.chromium.org/1367093002 .
 2015-09-25 10:29:20 -07:00
Frank Barchard
d4594beefc switch from ebp to ebx. 
ebx encodes more efficiently (1 byte less) for most address modes, than ebp.
previously it was used for 411 format, but the reader uses pinsrw now avoiding
gpr register.

BUG=libyuv:488
R=harryjin@google.com

Review URL: https://codereview.chromium.org/1365003003 .
 2015-09-24 17:25:11 -07:00
Frank Barchard
000cf89ca8 YUY2ToARGB avx2 in 1 step conversion. 
Includes UYVYToARGB ssse3 fix.

Was
YUY2ToARGB_Opt (433 ms)
69.79%  libyuv_unittest  libyuv_unittest      [.] I422ToARGBRow_AVX2
20.73%  libyuv_unittest  libyuv_unittest      [.] YUY2ToUV422Row_AVX2
 6.04%  libyuv_unittest  libyuv_unittest      [.] YUY2ToYRow_AVX2
 0.77%  libyuv_unittest  libyuv_unittest      [.] YUY2ToARGBRow_AVX2

Now
YUY2ToARGB_Opt (280 ms)
95.66%  libyuv_unittest  libyuv_unittest      [.] YUY2ToARGBRow_AVX2

BUG=libyuv:494
R=harryjin@google.com

Review URL: https://codereview.chromium.org/1364813002 .
 2015-09-23 11:15:18 -07:00
Frank Barchard
2b92ec8d0f Fix git markers introduced on landing previous CL 
BUG=none

Review URL: https://codereview.chromium.org/1359023003 .
 2015-09-22 15:00:57 -07:00
Frank Barchard
5f3d4270d1 yuy2 to rgb gcc versions 
read in read function for yuv conversion

R=harryjin@google.com
BUG=libyuv:488

Review URL: https://codereview.chromium.org/1355393002 .
 2015-09-22 14:27:33 -07:00
Frank Barchard
03cd8584e7 Read Y channel in read function for yuv conversion. 
Allows reader to support YUY2 format.
Also contains fix for win64 build for yuv conversion.

TBR=harryjin@google.com
BUG=libyuv:488

Review URL: https://codereview.chromium.org/1355333002 .
 2015-09-22 12:05:16 -07:00
Frank Barchard
f96890a0be yuvconstants for all YUV to RGB conversion functions. 
R=harryjin@google.com
BUG=libyuv:488

Review URL: https://codereview.chromium.org/1363503002 .
 2015-09-22 10:26:03 -07:00
Frank Barchard
62c49dc811 move constants into common 
R=harryjin@google.com
BUG=libyuv:488

Review URL: https://codereview.chromium.org/1359443005 .
 2015-09-18 16:28:44 -07:00
Frank Barchard
28427a53e2 I444ToABGR for android 
Reimplements I444ToARGB as a matrix function.
new I444ToABGR as matrix functions with wrappers and any functions.
Allows for future J444 and H444 versions.
I444ToABGR user level function added.

BUG=libyuv:490, libyuv:449
R=harryjin@google.com

Review URL: https://codereview.chromium.org/1355733002 .
 2015-09-18 11:20:58 -07:00
Frank Barchard
316e1ab996 avx2 width parameter bug fix 
R=harryjin@google.com
BUG=libyuv:489

Review URL: https://codereview.chromium.org/1321773004 .
 2015-09-09 11:56:35 -07:00
Frank Barchard
ed55d24d9f H420 functionality 
R=harryjin@google.com
BUG=libyuv:488

Review URL: https://webrtc-codereview.appspot.com/54869004 .
 2015-09-06 11:01:40 -07:00
Frank Barchard
7060e0d826 I420ToABGRMatrix functions with J420ToABGR wrapper. 
Allows direct conversion from JPeg to ABGR for android.

BUG=libyuv:488
R=harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/55719004 .
 2015-09-03 10:42:36 -07:00
Frank Barchard
925c3d9e26 I420ToARGB conversion with matrix. 
Take color conversion constants as a parameter to row function for I420ToARGBMatrixRow_SSSE3.
Allows future variations of color space using a single low level.

R=harryjin@google.com
BUG=libyuv:488

Review URL: https://webrtc-codereview.appspot.com/56669004 .
 2015-09-02 10:45:42 -07:00
Frank Barchard
0bc626a5d7 nolint removed 
R=harryjin@google.com
BUG=none

Review URL: https://webrtc-codereview.appspot.com/59389004.
 2015-08-31 10:52:13 -07:00
Frank Barchard
0735245c52 pinsrw instruction allows reading 2 bytes directly into an xmm register. 
Saving a gpr register allows the register to not be pushed for now, and in future it can be used to point to color conversion matrix or alpha channel.

R=harryjin@google.com
BUG=libyuv:488

Review URL: https://webrtc-codereview.appspot.com/52789004.
 2015-08-28 17:03:54 -07:00
Frank Barchard
be11f500f0 Use ebp to point to conversion table. 
Proof of concept that conversions can table color matrix as a parameter.

R=harryjin@google.com

BUG=libyuv:472, libyuv:488

Review URL: https://webrtc-codereview.appspot.com/58489004.
 2015-08-28 12:00:49 -07:00
Frank Barchard
3c4f5735ce use pointer to inverse table for clangcl 
R=harryjin@google.com
TBR=harryjin@google.com
BUG=none

Review URL: https://webrtc-codereview.appspot.com/54859004.
 2015-08-26 12:53:03 -07:00
Frank Barchard
5452cce452 port row to clangcl 
BUG=libyuv:487
R=harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/53799005.
 2015-08-25 16:15:42 -07:00
Frank Barchard
fa7ce4af3f fixed table for clangcl 
R=harryjin@google.com
BUG=libyuv:487

Review URL: https://webrtc-codereview.appspot.com/53799004.
 2015-08-25 10:47:30 -07:00
Frank Barchard
ee9aaea02f i422torgb565 is asm for clangcl as well 
Merge branch 'master' of https://chromium.googlesource.com/libyuv/libyuv into convertcl

allow lto for llvm but not gcc

R=harryjin@google.com
BUG=libyuv:469

Review URL: https://webrtc-codereview.appspot.com/52769004.
 2015-08-19 10:46:30 -07:00
Frank Barchard
715a29195b vpermq for avx2 ARGB4444ToARGB, ARGB1555ToARGB and RGB565ToARGB 
R=harryjin@google.com
BUG=libyuv:462

Review URL: https://webrtc-codereview.appspot.com/52759004.
 2015-07-07 17:06:04 -07:00
Frank Barchard
97b35daf75 disable faulty avx2 in argb conversions and box filter. and extend temporary buffer to 128 for an avx2 any function. 
R=harryjin@google.com
BUG=libyuv:462
TESTED=libyuv_unittest run on haswell laptop

Review URL: https://webrtc-codereview.appspot.com/53759004.
 2015-07-07 15:40:24 -07:00
Frank Barchard
0686f26938 blend remove alignment 1 pixel loop for less overhead. 
R=tpsiaki@google.com
BUG=none
TESTED=libyuvTest.ARGBBlend_Opt

Review URL: https://webrtc-codereview.appspot.com/50289005.
 2015-06-24 11:34:12 -07:00
fbarchard@google.com
2e9f3e5cf5 rename source files from row_posix.cc etc to row_gcc.cc to avoid gyp build filtering out source files from build when on windows with clang. The source code contained in row_gcc.cc is gcc syntax inline assembly available for any platform that supports gcc or clang for intel cpus. 
BUG=440
TESTED=try bots
R=harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/56579004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1430 16f28f9a-4ce2-e073-06de-1de4eb20be90
 2015-06-09 17:27:52 +00:00
fbarchard@google.com
b07de879b6 enable intrinsics for clangcl if -mssse3 is enabled. 
BUG=451
TESTED=untested
R=harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/52699004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1427 16f28f9a-4ce2-e073-06de-1de4eb20be90
 2015-06-08 22:48:18 +00:00
fbarchard@google.com
cfce47efc8 Change Sobel to use JPeg Luma calculation instead of extracting G channel. Using luma produces a better sobel that respects all 3 channels of RGB. Historically the G channel was used to improve performance, and because the luma of I420 is a constrained range, hurting quality. Using the JPeg variation of YUV, the luma is more accurate, including cross platform, better optimized for AVX2 and odd widths, and full range. 
BUG=444
TESTED=ARGBSobelXY_Opt
R=harryjin@google.com

Review URL: https://webrtc-codereview.appspot.com/57479004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1414 16f28f9a-4ce2-e073-06de-1de4eb20be90
 2015-05-27 22:32:26 +00:00
fbarchard@google.com
632c50f29c include posix source for 64 bit clang builds. 
BUG=440
TESTED=ninja -C out\Release_x64
R=bcornell@google.com

Review URL: https://webrtc-codereview.appspot.com/46259004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1407 16f28f9a-4ce2-e073-06de-1de4eb20be90
 2015-05-14 21:40:05 +00:00
fbarchard@google.com
2c44965e8d make row_win windows code built with clangcl include the _posix source code. 
depot_tools excludes these source files now, so they need to be manually
included.
BUG=435
TESTED=clangcl local build on windows
R=tpsiaki@google.com

Review URL: https://webrtc-codereview.appspot.com/49879004

git-svn-id: http://libyuv.googlecode.com/svn/trunk@1395 16f28f9a-4ce2-e073-06de-1de4eb20be90
 2015-05-02 00:39:29 +00:00