<div data-ntes="ntes_mail_body_root" style="line-height:1.7;color:#000000;font-size:14px;font-family:Arial"><div id="spnEditorContent"><p style="margin: 0;">Thank for the patch, I have a comment,</p><p style="margin: 0;"><br></p><pre style="width: 1238.04px; word-break: break-word !important;">--- a/source/common/aarch64/blockcopy8-sve.S
+++ b/source/common/aarch64/blockcopy8-sve.S
@@ -965,7 +965,7 @@ function PFX(cpy2Dto1D_shl_64x64_sve)
whilelt p0.h, x9, x8
b.first .L_cpy2Dto1D_shl_64x64
add x1, x1, x2, lsl #1
- addvl x0, x0, #1
+ add x0, x0, x8, lsl #1
cbnz w12, .L_init_cpy2Dto1D_shl_64x64
ret</pre><pre style="width: 1238.04px; word-break: break-word !important;">In here, the x8 is constant, why not add constant directly?</pre><pre style="width: 1238.04px; word-break: break-word !important;"><br></pre><pre style="width: 1238.04px; word-break: break-word !important;">At 2025-02-07 21:47:07, "Gerda Zsejke More" <gerdazsejke.more@arm.com> wrote:</pre></div><pre>>Enable tests for block size 64x64 for various pixel functions.
>
>Add missing implementation for ssd_s for block size 64x64.
>
>Arpad Panyik (1):
> pixelharness.cpp: Enable tests for 64x64 TU primitives
>
>Gerda Zsejke More (1):
> AArch64: Add Neon asm impl. for SBD ssd_s block size 64x64
>
> source/common/aarch64/asm-primitives.cpp | 2 +
> source/common/aarch64/blockcopy8-sve.S | 4 +-
> source/common/aarch64/ssd-a.S | 35 ++++++
> source/test/pixelharness.cpp | 135 +++++++++++------------
> 4 files changed, 105 insertions(+), 71 deletions(-)
>
>--
>2.39.5 (Apple Git-154)
>
>_______________________________________________
>x265-devel mailing list
>x265-devel@videolan.org
>https://mailman.videolan.org/listinfo/x265-devel
</pre></div>