George Steed 823d960afc [AArch64] Add SVE2 implementations of {P210,P410}ToAR30Row
Observed reductions in runtime compared to the existing Neon code:

            | P210ToAR30Row | P410ToAR30Row
Cortex-A510 |       -16.5%  |        -21.2%
Cortex-A520 | (!)    +2.7%  |         -8.7%
Cortex-A715 |        -6.1%  |         -6.1%
Cortex-A720 |        -6.2%  |         -5.9%
  Cortex-X2 |        -4.1%  |         -4.2%
  Cortex-X3 |        -4.2%  |         -4.2%
  Cortex-X4 |        -1.2%  |         -1.2%
Cortex-X925 |        -3.6%  |         -2.8%

Bug: b/42280942
Change-Id: I40723a370fad1ccb53f8ccd9d32cddb502500dd6
Reviewed-on: https://chromium-review.googlesource.com/c/libyuv/libyuv/+/6023036
Reviewed-by: Frank Barchard <fbarchard@chromium.org>
2024-11-14 16:52:21 +00:00
..
libyuv [AArch64] Add SVE2 implementations of {P210,P410}ToAR30Row 2024-11-14 16:52:21 +00:00
libyuv.h NV12 Copy, include scale_uv.h 2020-12-08 18:54:16 +00:00