coffee/libyuv
1
0
Fork 0
mirror of https://chromium.googlesource.com/libyuv/libyuv synced 2026-01-01 03:12:16 +08:00
Code Issues Packages Projects Releases Wiki Activity
libyuv/source
History
George Steed 4ad050b5ec [AArch64] Unroll {I422,I422Alpha}ToARGBRow_SVE2 
Since the UV components are duplicated in I422 we end up wasting half of
the vector bandwidth processing the same elements twice. By unrolling
the kernel to process two vectors of Y per iteration we can fill a whole
vector of U/V components.

Rather than packing RGBA components into pairs during the narrowing we
now just narrow into individual component vectors and use ST4B instead.
This by itself is slower on some micro-architectures like Cortex-A510
but the benefit from unrolling significantly outweights this.

            | I422AlphaToARGBRow_SVE2 | I422ToARGBRow_SVE2
Cortex-A510 |                  -46.2% |             -48.8%
Cortex-A720 |                  -20.8% |             -21.0%
  Cortex-X2 |                  -11.3% |              -7.5%
  Cortex-X4 |                  -15.4% |             -15.5%

Bug: libyuv:973
Change-Id: I69389c4279861f7a460ae0c28186f023c728c4e8
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/5725173
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
2024-07-19 15:55:59 +00:00
..
compare_common.cc clang-tidy applied 2021-04-01 21:42:47 +00:00
compare_gcc.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
compare_msa.cc use unix line endings 2018-06-20 23:19:59 +00:00
compare_neon64.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
compare_neon.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
compare_win.cc Switch win32 to row_gcc for clangcl. 2021-04-22 19:32:32 +00:00
compare.cc [AArch64] Add Neon implementation of HashDjb2 2024-05-01 19:37:31 +00:00
convert_argb.cc [AArch64] Add SVE2 implementation of RGB24ToARGBRow 2024-07-08 20:12:05 +00:00
convert_from_argb.cc [AArch64] Add I8MM implementation of ARGBToUV444Row 2024-07-16 17:32:52 +00:00
convert_from.cc Change ScalePlane,ScalePlane_16,... to return int 2023-11-03 23:53:24 +00:00
convert_jpeg.cc PlaneScale, UVScale and ARGBScale test 3x and 4x down sample. 2020-10-28 20:41:59 +00:00
convert_to_argb.cc Remove M420 and refactor NV12ToI420 2020-05-26 18:48:00 +00:00
convert_to_i420.cc Fix ConvertToI420 when using YUY2 or UYVY with odd crop_x. 2021-07-19 22:22:22 +00:00
convert.cc [AArch64] Add SVE2 implementations for AYUVTo{UV,VU}Row 2024-06-04 18:18:07 +00:00
cpu_id.cc [AArch64] Enable SME feature detection on Apple Silicon 2024-07-08 16:19:27 +00:00
mjpeg_decoder.cc Add AMXINT8 cpu detect 2024-02-15 21:44:47 +00:00
mjpeg_validate.cc Update to r1732 for more robust jpeg 2019-07-01 22:32:36 +00:00
planar_functions.cc [AArch64] Add SVE2 implementation of RAWToRGB24Row 2024-07-08 15:55:14 +00:00
rotate_any.cc [AArch64] Fix rotate by odd sizes 2024-07-15 18:13:31 +00:00
rotate_argb.cc malloc return 1 for failures and assert for internal functions 2023-12-04 22:55:20 +00:00
rotate_common.cc [AArch64] Use full vectors in TransposeWx{8 => 16}_NEON 2024-05-21 07:46:42 +00:00
rotate_gcc.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
rotate_lsx.cc [AArch64] Use full vectors in TransposeWx{8 => 16}_NEON 2024-05-21 07:46:42 +00:00
rotate_msa.cc cpuid show vector length on ARM and RISCV 2024-07-02 18:10:56 +00:00
rotate_neon64.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
rotate_neon.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
rotate_sme.cc [AArch64] Add SME implementation of TransposeUVWxH 2024-07-19 12:15:40 +00:00
rotate_win.cc Switch win32 to row_gcc for clangcl. 2021-04-22 19:32:32 +00:00
rotate.cc [AArch64] Add SME implementation of TransposeUVWxH 2024-07-19 12:15:40 +00:00
row_any.cc [AArch64] Add I8MM implementation of ARGBToUV444Row 2024-07-16 17:32:52 +00:00
row_common.cc [RVV] Support AR64ToAB64 and RGBA-family color conversions 2023-09-05 22:44:48 +00:00
row_gcc.cc [AArch64] Fix SVE/SME vector length printing in cpuid 2024-07-02 19:44:41 +00:00
row_lasx.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
row_lsx.cc [AArch64] Fix SVE/SME vector length printing in cpuid 2024-07-02 19:44:41 +00:00
row_msa.cc Fix Bugs on mips platform V2. 2022-03-01 13:16:31 +00:00
row_neon64.cc [AArch64] Add I8MM implementation of ARGBToUV444Row 2024-07-16 17:32:52 +00:00
row_neon.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
row_rvv.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
row_sve.cc [AArch64] Unroll {I422,I422Alpha}ToARGBRow_SVE2 2024-07-19 15:55:59 +00:00
row_win.cc Fix tidy warning that uint32_t dither4 should not be const 2023-06-02 00:42:02 +00:00
scale_any.cc UVScale down by 2 fix for C and optimize for NEON 2023-04-12 22:49:20 +00:00
scale_argb.cc [AArch64] Add SVE implementation for I422ToARGBRow 2024-04-27 18:26:11 +00:00
scale_common.cc Fix warnings for missing prototypes 2023-06-30 17:46:56 +00:00
scale_gcc.cc cpuid show vector length on ARM and RISCV 2024-07-02 18:10:56 +00:00
scale_lsx.cc DetilePlane and unittest for NEON 2022-01-31 20:05:55 +00:00
scale_msa.cc Switch to C99 types 2018-01-23 19:16:05 +00:00
scale_neon64.cc [AArch64] Rework data loading in ScaleARGBFilterCols_NEON 2024-07-10 23:10:43 +00:00
scale_neon.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
scale_rgb.cc RGBScale function using 3 steps: RGB24ToARGB, ARGBScale, ARGBToRGB24 2022-03-19 01:44:06 +00:00
scale_rvv.cc Add volatile for gcc inline to avoid being removed 2024-07-02 01:25:24 +00:00
scale_uv.cc Disable RVV ScaleDownBy4 if compiler option is not enabled 2024-06-18 01:52:40 +00:00
scale_win.cc Switch win32 to row_gcc for clangcl. 2021-04-22 19:32:32 +00:00
scale.cc malloc return 1 for failures and assert for internal functions 2023-12-04 22:55:20 +00:00
test.sh Optimze ABGRToI420 for AVX2 2020-06-04 18:24:45 +00:00
video_common.cc Lint cleanup after C99 change CL 2018-01-24 19:16:03 +00:00