diff options
Diffstat (limited to 'src/core/NEON/kernels/arm_gemm/merges')
6 files changed, 37 insertions, 13 deletions
diff --git a/src/core/NEON/kernels/arm_gemm/merges/list-sve.hpp b/src/core/NEON/kernels/arm_gemm/merges/list-sve.hpp new file mode 100644 index 0000000000..aded4b3b8c --- /dev/null +++ b/src/core/NEON/kernels/arm_gemm/merges/list-sve.hpp @@ -0,0 +1,28 @@ +/* + * Copyright (c) 2021 Arm Limited. + * + * SPDX-License-Identifier: MIT + * + * Permission is hereby granted, free of charge, to any person obtaining a copy + * of this software and associated documentation files (the "Software"), to + * deal in the Software without restriction, including without limitation the + * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or + * sell copies of the Software, and to permit persons to whom the Software is + * furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in all + * copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include "sve_merge_fp16_3VLx8.hpp" +#include "sve_merge_fp32_3VLx8.hpp" +#include "sve_merge_s32_3VLx8.hpp" +#include "sve_merge_u32_3VLx8.hpp"
\ No newline at end of file diff --git a/src/core/NEON/kernels/arm_gemm/merges/list.hpp b/src/core/NEON/kernels/arm_gemm/merges/list.hpp index 825c2fd020..dae874ef94 100644 --- a/src/core/NEON/kernels/arm_gemm/merges/list.hpp +++ b/src/core/NEON/kernels/arm_gemm/merges/list.hpp @@ -27,8 +27,4 @@ #include "a64_merge_s32_12x8.hpp" #include "a64_merge_s32_4x4.hpp" #include "a64_merge_u32_12x8.hpp" -#include "a64_merge_u32_4x4.hpp" -#include "sve_merge_fp16_3VLx8.hpp" -#include "sve_merge_fp32_3VLx8.hpp" -#include "sve_merge_s32_3VLx8.hpp" -#include "sve_merge_u32_3VLx8.hpp" +#include "a64_merge_u32_4x4.hpp"
\ No newline at end of file diff --git a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp16_3VLx8.hpp b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp16_3VLx8.hpp index cf1d10329b..4da32b459c 100644 --- a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp16_3VLx8.hpp +++ b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp16_3VLx8.hpp @@ -23,7 +23,7 @@ */ #pragma once -#ifdef __ARM_FEATURE_SVE +#ifdef ARM_COMPUTE_ENABLE_SVE template<> void MergeResults<3, 8, true>(__fp16 *out, const __fp16 *in, const int ldout, const int y0, const int ymax, const int x0, const int xmax, const __fp16 *bias, Activation act, bool append) @@ -1872,4 +1872,4 @@ void MergeResults<3, 8, true>(__fp16 *out, const __fp16 *in, const int ldout, co } } -#endif // __ARM_FEATURE_SVE +#endif // ARM_COMPUTE_ENABLE_SVE diff --git a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp index b0d10c085d..5505f1efe4 100644 --- a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp +++ b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp @@ -23,7 +23,7 @@ */ #pragma once -#ifdef __ARM_FEATURE_SVE +#ifdef ARM_COMPUTE_ENABLE_SVE template<> void MergeResults<3, 8, true>(float *out, const float *in, const int ldout, const int y0, const int ymax, const int x0, const int xmax, const float *bias, Activation act, bool append) @@ -1872,4 +1872,4 @@ void MergeResults<3, 8, true>(float *out, const float *in, const int ldout, cons } } -#endif // __ARM_FEATURE_SVE +#endif // ARM_COMPUTE_ENABLE_SVE diff --git a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_s32_3VLx8.hpp b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_s32_3VLx8.hpp index 34b6fe3ef5..c009881254 100644 --- a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_s32_3VLx8.hpp +++ b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_s32_3VLx8.hpp @@ -23,7 +23,7 @@ */ #pragma once -#ifdef __ARM_FEATURE_SVE +#ifdef ARM_COMPUTE_ENABLE_SVE template<> void MergeResults<3, 8, true>(int32_t *out, const int32_t *in, const int ldout, const int y0, const int ymax, const int x0, const int xmax, const int32_t *bias, Activation , bool append) @@ -1394,4 +1394,4 @@ void MergeResults<3, 8, true>(int32_t *out, const int32_t *in, const int ldout, } } -#endif // __ARM_FEATURE_SVE +#endif // ARM_COMPUTE_ENABLE_SVE diff --git a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_u32_3VLx8.hpp b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_u32_3VLx8.hpp index c4b2bb56d6..e992f6722c 100644 --- a/src/core/NEON/kernels/arm_gemm/merges/sve_merge_u32_3VLx8.hpp +++ b/src/core/NEON/kernels/arm_gemm/merges/sve_merge_u32_3VLx8.hpp @@ -23,7 +23,7 @@ */ #pragma once -#ifdef __ARM_FEATURE_SVE +#ifdef ARM_COMPUTE_ENABLE_SVE template<> void MergeResults<3, 8, true>(uint32_t *out, const uint32_t *in, const int ldout, const int y0, const int ymax, const int x0, const int xmax, const uint32_t *bias, Activation , bool append) @@ -1394,4 +1394,4 @@ void MergeResults<3, 8, true>(uint32_t *out, const uint32_t *in, const int ldout } } -#endif // __ARM_FEATURE_SVE +#endif // ARM_COMPUTE_ENABLE_SVE |