diff options
| author | Michalis Spyrou <michalis.spyrou@arm.com> | 2020-05-06 17:03:59 +0100 |
|---|---|---|
| committer | Michalis Spyrou <michalis.spyrou@arm.com> | 2020-05-20 21:48:34 +0000 |
| commit | 18e20ff8bc31bd835e96b46c0beb5435c76572bd (patch) | |
| tree | fcd826b6f528da07d4b82d0300a19f78a6ac7333 /arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h | |
| parent | 11bfe0e99b3af0edbed1bfd2b89242b3b4e45d72 (diff) | |
| download | ComputeLibrary-18e20ff8bc31bd835e96b46c0beb5435c76572bd.tar.gz | |
COMPMID-3460: Refactor NEElementwiseUnaryKernel
Removed most of the templates and refactored the code.
Performance is the same but the libary size dropped by
52Kb.
Change-Id: I41ff0c0853c923d925cdaeb05f4a58c9086fff94
Signed-off-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/3190
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h')
| -rw-r--r-- | arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h | 10 |
1 files changed, 9 insertions, 1 deletions
diff --git a/arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h b/arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h index 2bf9f52dbe..77adcf7b8c 100644 --- a/arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h +++ b/arm_compute/core/NEON/wrapper/intrinsics/invsqrt.h @@ -1,5 +1,5 @@ /* - * Copyright (c) 2018-2019 ARM Limited. + * Copyright (c) 2018-2020 ARM Limited. * * SPDX-License-Identifier: MIT * @@ -37,10 +37,18 @@ namespace wrapper return prefix##_##postfix(a); \ } +#define VINVSQRT_IMPL_INT(stype, vtype, prefix, postfix) \ + inline vtype vinvsqrt(const vtype &a) \ + { \ + ARM_COMPUTE_UNUSED(a); \ + ARM_COMPUTE_ERROR("Not supported"); \ + } + VINVSQRT_IMPL(float, float32x2_t, vinvsqrt, f32) #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC VINVSQRT_IMPL(float16_t, float16x4_t, vinvsqrt, f16) #endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC +VINVSQRT_IMPL_INT(int, int32x4_t, vinvsqrt, s32) VINVSQRT_IMPL(float, float32x4_t, vinvsqrtq, f32) #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC |