- Unrolled to 16 pixels
- Take constants via structure, allowing different colorspace and channel order
- Use ADDHN to add 16.5 and take upper 8 bits of 16 bit values, narrowing to 8 bits
- clang-format applied, affecting mips code
On Cortex A510
Was RAWToJ400_Opt (1623 ms)
Now RAWToJ400_Opt (862 ms)
C RAWToJ400_Opt (1627 ms)
Bug: b/220171611
Change-Id: I06a9baf9650ebe2802fb6ff6dfbd524e2c06ada0
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3534023
Reviewed-by: Wan-Teh Chang <wtc@google.com>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
When used in C enum keyword can't be eliminated.
Bug: libyuv:872
Change-Id: Iacff5a8bd84ec7caa1f90889e48f81ffc10071ae
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3513317
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
This patch adds some deleted control macros so that these MSA
optimization functions can be called normally on mips platform.
There are also some modifications to adapt to the clang compiler.
Bug: libyuv:918
Change-Id: I6ffadc6582682b5eaeae2e0f4033d66d370b48b9
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3494667
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
This patch fixes compilation errors caused by the removal of kUVBias
and two failed test cases of LibYUVConvertTest.RGB565ToI420_Opt and
LibYUVConvertTest.ARGB1555ToI420_Opt.
Bug: libyuv:918
Change-Id: I1a66bcd7ef616aacbeca5b4015013015ccdf0f18
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3477416
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Add support for MM21 to NV12 and I420 conversion, and add SIMD
optimizations for arm, aarch64, SSE2, and SSSE3 machines.
Bug: libyuv:915, b/215425056
Change-Id: Iecb0c33287f35766a6169d4adf3b7397f1ba8b5d
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3433269
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Commit-Queue: Justin Green <greenjustin@google.com>
Optimize 20 functions in source/scale_lsx.cc file.
All test cases passed on loongarch platform.
Bug: libyuv:913
Change-Id: I85bcb3b0bfd9461bb6f93202546507352cbd624a
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3351469
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Optimize two functions in source/rotate_lsx.cc file.
All test cases passed on loongarch platform.
Bug: libyuv:913
Change-Id: Idf670a1bc078f6284a499a292e0cb795f5b603b4
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3351468
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Optimize 44 functions in source/row_lsx.cc file.
All test cases passed on loongarch platform.
Bug: libyuv:913
Change-Id: Ic80a5751314adc2e9bd435f2bbd928ab017a90f9
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3351467
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
1. Add supports for LSX and LASX.
2. Three optimization functions are added in loongarch/row_lasx.cc file:
I422ToARGBRow_LASX,I422ToRGBARow_LASX,I422AlphaToARGBRow_LASX.
Bug: libyuv:912, Bug: libyuv:913
Change-Id: I043c2704f99a5215724b5c0b7f97e6bf5f7a199b
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3329189
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
- adapted from Android420ToI420, adding a rotation parameter
- SplitRotateUV added to rotate and split the UV channel of NV12 or NV21
- rename RotateUV functions to SplitRotateUV
Bug: b/203549508
Change-Id: I6774da5fb5908fdf1fc12393f0001f41bbda9851
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3251282
Reviewed-by: richard winterton <rrwinterton@gmail.com>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
- ubsan complains on unaligned tests when an int16 or int32 is stored unaligned in C.
Although current Intel, ARM, Mips and PPC can do unaligned load/store, its not guaranteed
and could crash a CPU that doesnt support it.
- unaligned tests use offset of 2 or 4, which ubsan accepts.
- unittest fills in random buffer with 2 bytes at a time instead of a short.
- row common functions for int16 types use 2 shorts instead of 1 int.
Bug: libyuv:908, b/203243873
Change-Id: Idf13fa901647d7b0975f1947291caa781999a9bc
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3229782
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: Mirko Bonadei <mbonadei@chromium.org>
- reenable Intel SIMD unaffected by BIT_EXACT
- add bit exact version of ARGBAttenuate, which uses ARM version of formula.
- add bit exact version of ARGBUnatenuate, which mimics the AVX code.
Apply clang format to cleanup code.
Bug: libyuv:908, b/202888439
Change-Id: Ie842b1b3956b48f4190858e61c02998caedc2897
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3224702
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: richard winterton <rrwinterton@gmail.com>
- C code use ARM path, so NEON and C match
- C used on Intel platforms, disabling AVX.
Bug: libyuv:908, b/202888439
Change-Id: Ie035a150a60d3cf4ee7c849a96819d43640cf020
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3223507
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: richard winterton <rrwinterton@gmail.com>
With "m" GCC generates a memory address with offset which is
not allowed with ld1 on aarch64. Change constraint to "Q" to
force address without offset.
Bug: chromium:819294, libyuv:903
Change-Id: Iaae24bc6882cdef823259040a37fdbfc31f91185
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2922146
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
- swap U and V when crop x is odd
- document YUY2 and UYVY formats
- apply clang-format
Bug: libyuv:902
Change-Id: I045e44c907f4a9eb625d7c024b669bb308055f32
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/3039549
Reviewed-by: Mirko Bonadei <mbonadei@chromium.org>
Was
[ RUN ] LibYUVConvertTest.ARGB1555ToI420_Any
third_party/libyuv/files/unit_test/convert_test.cc:1139: Failure
Expected equality of these values:
dst_uv_c[i * kStrideUV + j]
Which is: '\x8B' (139)
dst_uv_opt[i * kStrideUV + j]
Which is: '\x92' (146)
third_party/libyuv/files/unit_test/convert_test.cc:1139: Failure
[ FAILED ] LibYUVConvertTest.ARGB1555ToI420_Any
Now
[ RUN ] LibYUVConvertTest.ARGB1555ToI420_Any
[ OK ] LibYUVConvertTest.ARGB1555ToI420_Any (0 ms)
Bug: libyuv:894, b/155722711
Change-Id: I12dcacd0ecfff4ede5693a2554e9bb10dc8586c1
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2870484
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: richard winterton <rrwinterton@gmail.com>
Add tests of all macros used by libyuv public headers
When a 1 step conversion is added, a 2 step test can compare
the old 2 step method to the 1 step. A 1 step unittest is
also added which compares C to SIMD. Making the 2 step
conversions measure performance of the 2 steps allows the
old 2 step performance to be compared to 1 step.
All macros used in public headers are added to an ifdef test.
Showing them in a unittest allows some diagnostics when
a test is failing.
Bug: libyuv:901
Change-Id: I7ffa6ed0cb3b506fa1b7fd4b7b1b729658c3c266
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2857916
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: Mirko Bonadei <mbonadei@chromium.org>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Use unsigned coefficient and signed UV value in YUVTORGB.
R=fbarchard@chromium.org
Bug: libyuv:862, libyuv:863
Change-Id: I32e58b2cee383fb98104c055beb0867a7ad05bfe
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2850016
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: Wan-Teh Chang <wtc@google.com>
Commit-Queue: Frank Barchard <fbarchard@chromium.org>
Use unsigned coefficients on Intel.
Make C, NEON and AVX2 match under LIBYUV_UNLIMITED_DATA.
Bug: libyuv:862, libyuv:863
Change-Id: I6c02147ea3c1875c4fc23863435aea86dcf5880a
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2830180
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
Reviewed-by: richard winterton <rrwinterton@gmail.com>
Refactor NEON YUVToRGB Assembly to support HBD data as input and output.
Work on YUV444 internally, remove subsampling in I444ToARGB.
libyuv_unittest --gtest_filter=*.NV??ToARGB_Opt:*UYVYToARGB_Opt:*YUY2ToARGB_Opt:*I4*ToARGB_Opt
Bug: libyuv:895, libyuv:862, libyuv:863
Change-Id: I05b56ea8ea56d9e523720b842fa6e4b122ed4115
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/2810060
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
