diff options
| author | Michael Tyler <michael.tyler@arm.com> | 2023-04-12 17:43:17 +0100 |
|---|---|---|
| committer | michael.tyler <michael.tyler@arm.com> | 2023-06-05 15:57:58 +0000 |
| commit | 74921eee924625426429044decefe3673561b174 (patch) | |
| tree | 654da1a95e3d42d6af8ad1ff27bb40d77b1fd8c5 /src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp | |
| parent | df5d9878008be9b60586df97ebfff197abb5195e (diff) | |
| download | ComputeLibrary-74921eee924625426429044decefe3673561b174.tar.gz | |
Update CPU kernel implementations and guard directives
Resolves COMPMID-6023
Change-Id: I868975d14c4f98af6716726feda22405a6a4c891
Signed-off-by: Michael Tyler <michael.tyler@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9686
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp')
| -rw-r--r-- | src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp | 7 |
1 files changed, 3 insertions, 4 deletions
diff --git a/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp b/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp index dafa53eec3..ac67467240 100644 --- a/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp +++ b/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_16_2x4.hpp @@ -24,7 +24,7 @@ #pragma once -#ifdef __aarch64__ +#if defined(__aarch64__) namespace { @@ -320,7 +320,6 @@ void a64_transpose_interleave_16_2x4(uint16_t *out, const uint16_t *in, size_t w "bge 1b\n" "cbz %x[height], 20f\n" "10:" // Main loop skip - "11:" // Tail row loop: Head "mov x9, %x[in]\n" "add x28, x9, %x[in_stride]\n" @@ -486,7 +485,6 @@ void a64_transpose_interleave_16_2x4(uint16_t *out, const uint16_t *in, size_t w "add %x[out], %x[out], #0x80\n" "bge 11b\n" "20:" // Done - : [height] "+&r" (height), [in] "+&r" (in), [out] "+&r" (out) : [in_stride] "r" (in_stride), [out_stride] "r" (out_stride), [pad_row] "r" (pad_row), [width] "r" (width) : "cc", "memory", "v0", "v1", "v2", "v3", "v4", "v5", "v6", "v7", "v8", "v9", "v10", "v11", "v12", "v13", "v14", "v15", "v16", "v17", "v18", "v19", "v20", "v21", "v22", "v23", "v24", "v25", "v26", "v27", "v28", "v29", "v30", "v31", "x9", "x20", "x21", "x22", "x23", "x24", "x25", "x26", "x27", "x28" @@ -508,4 +506,5 @@ void Transform<16, 4, true, VLType::None>( ); } -#endif + +#endif // defined(__aarch64__) |