diff options
author | Georgios Pinitas <georgios.pinitas@arm.com> | 2019-06-27 17:00:52 +0100 |
---|---|---|
committer | Georgios Pinitas <georgios.pinitas@arm.com> | 2019-07-26 11:55:15 +0000 |
commit | cfa2bba98169cb5ab1945462514be1b6badf7d98 (patch) | |
tree | 1635e6e9463e9798c7195f0aa71b5df3f2650df1 /src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp | |
parent | f59b16f42ef68bde877b70816ffb953d64c8baa3 (diff) | |
download | ComputeLibrary-cfa2bba98169cb5ab1945462514be1b6badf7d98.tar.gz |
COMPMID-2178: Update GEMM assembly code.
Perform offset reduction and requantization within the assembly wrapper.
Change-Id: I5d5b3e1f6f9ef4c71805362c57f88ff199c027a3
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1541
Comments-Addressed: Pablo Marquez <pablo.tello@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp')
-rw-r--r-- | src/core/NEON/kernels/arm_gemm/merges/sve_merge_fp32_3VLx8.hpp | 38 |
1 files changed, 19 insertions, 19 deletions
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 27084c3598..4b51066ebe 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 @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018 Arm Limited. + * Copyright (c) 2018-2019 Arm Limited. * * SPDX-License-Identifier: MIT * @@ -84,7 +84,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 2: { long w = xmax - i; @@ -135,7 +135,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 3: { long w = xmax - i; @@ -198,7 +198,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 4: { long w = xmax - i; @@ -272,7 +272,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 5: { long w = xmax - i; @@ -358,7 +358,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 6: { long w = xmax - i; @@ -456,7 +456,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 7: { long w = xmax - i; @@ -566,7 +566,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + default: case 8: { @@ -688,8 +688,8 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - - + + } } else @@ -739,7 +739,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 2: { long w = xmax - i; @@ -802,7 +802,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 3: { long w = xmax - i; @@ -883,7 +883,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 4: { long w = xmax - i; @@ -981,7 +981,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 5: { long w = xmax - i; @@ -1097,7 +1097,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 6: { long w = xmax - i; @@ -1231,7 +1231,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + case 7: { long w = xmax - i; @@ -1383,7 +1383,7 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - + default: case 8: { @@ -1553,8 +1553,8 @@ inline void MergeResults<3, 8, true>(float *out, const float *in, const int ldou ); } break; - - + + } } } |