diff options
| author | Sheri Zhang <sheri.zhang@arm.com> | 2021-10-15 19:54:17 +0100 |
|---|---|---|
| committer | Sheri Zhang <sheri.zhang@arm.com> | 2021-10-18 17:36:47 +0000 |
| commit | 5dda2177800009b24e31550ed849b1ef3fca6167 (patch) | |
| tree | dfce69d52db6111d6751f4ee4add6ab172a3290d /src/core/NEON/NEMath.h | |
| parent | c9cecc0e565e7b4978cecc92e03e6c93bb8d0cb9 (diff) | |
| download | ComputeLibrary-5dda2177800009b24e31550ed849b1ef3fca6167.tar.gz | |
DirectConv3d support refine
- Decouple data support of CpuDirectConv3dKernel
- Update documentation for Conv3d
Signed-off-by: Sheri Zhang <sheri.zhang@arm.com>
Change-Id: I1d94aa28f821f45a1a3d39cc3335c8faeee89f0d
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6453
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/NEON/NEMath.h')
| -rw-r--r-- | src/core/NEON/NEMath.h | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/src/core/NEON/NEMath.h b/src/core/NEON/NEMath.h index 13484c9c15..8118c4701f 100644 --- a/src/core/NEON/NEMath.h +++ b/src/core/NEON/NEMath.h @@ -239,6 +239,14 @@ float32x4_t vsinq_f32(float32x4_t val); */ float32x2_t vsin_f32(float32x2_t val); +/** Reduce a vector to be a scalar by accumulating all lanes in the vector + * + * @param[in] v Vector to be reduced. + * + * @return the wrapped-around number. + */ +float vreduce(const float32x4_t &v); + #ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC /** Calculate hyperbolic tangent. * @@ -319,6 +327,13 @@ float16x8_t vpowq_f16(float16x8_t val, float16x8_t n); */ float16x8_t vsinq_f16(float16x8_t val); +/** Reduce a vector to be a scalar by accumulating all lanes in the vector + * + * @param[in] v Vector to be reduced. + * + * @return the wrapped-around number. + */ +float16_t vreduce(const float16x8_t &v); #endif /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */ } // namespace arm_compute #include "src/core/NEON/NEMath.inl" |