From 5801a5508aecc91df7d669f086d6977d70059c65 Mon Sep 17 00:00:00 2001 From: George Wort Date: Thu, 13 Dec 2018 17:50:26 +0000 Subject: COMPMID-1767: NEON: Implement Where/Select Change-Id: If8a1ab6d6a029a5c547b726e0692eecef9a2e97d Reviewed-on: https://review.mlplatform.org/415 Tested-by: Arm Jenkins Reviewed-by: Georgios Pinitas --- .../core/NEON/wrapper/intrinsics/greaterthan.h | 64 ++++++++++++++++++++++ 1 file changed, 64 insertions(+) create mode 100644 arm_compute/core/NEON/wrapper/intrinsics/greaterthan.h (limited to 'arm_compute/core/NEON/wrapper/intrinsics/greaterthan.h') diff --git a/arm_compute/core/NEON/wrapper/intrinsics/greaterthan.h b/arm_compute/core/NEON/wrapper/intrinsics/greaterthan.h new file mode 100644 index 0000000000..5ee7516a4e --- /dev/null +++ b/arm_compute/core/NEON/wrapper/intrinsics/greaterthan.h @@ -0,0 +1,64 @@ +/* + * Copyright (c) 2018 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. + */ +#ifndef __ARM_COMPUTE_WRAPPER_CGT_H__ +#define __ARM_COMPUTE_WRAPPER_CGT_H__ + +#include + +namespace arm_compute +{ +namespace wrapper +{ +#define VCGT_IMPL(stype, vtype, rtype, prefix, postfix) \ + inline rtype vgreaterthan(const vtype &a, const vtype &b) \ + { \ + return prefix##_##postfix(a, b); \ + } + +VCGT_IMPL(uint8_t, uint8x8_t, uint8x8_t, vcgt, u8) +VCGT_IMPL(int8_t, int8x8_t, uint8x8_t, vcgt, s8) +VCGT_IMPL(uint16_t, uint16x4_t, uint16x4_t, vcgt, u16) +VCGT_IMPL(int16_t, int16x4_t, uint16x4_t, vcgt, s16) +VCGT_IMPL(uint32_t, uint32x2_t, uint32x2_t, vcgt, u32) +VCGT_IMPL(int32_t, int32x2_t, uint32x2_t, vcgt, s32) +VCGT_IMPL(float32x2_t, float32x2_t, uint32x2_t, vcgt, f32) +#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC +VCGT_IMPL(float16x4_t, float16x4_t, uint16x4_t, vcgt, f16) +#endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC + +VCGT_IMPL(uint8_t, uint8x16_t, uint8x16_t, vcgtq, u8) +VCGT_IMPL(int8_t, int8x16_t, uint8x16_t, vcgtq, s8) +VCGT_IMPL(uint16_t, uint16x8_t, uint16x8_t, vcgtq, u16) +VCGT_IMPL(int16_t, int16x8_t, uint16x8_t, vcgtq, s16) +VCGT_IMPL(uint32_t, uint32x4_t, uint32x4_t, vcgtq, u32) +VCGT_IMPL(int32_t, int32x4_t, uint32x4_t, vcgtq, s32) +VCGT_IMPL(float32x4_t, float32x4_t, uint32x4_t, vcgtq, f32) +#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC +VCGT_IMPL(float16x8_t, float16x8_t, uint16x8_t, vcgtq, f16) +#endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC + +#undef VCGT_IMPL +} // namespace wrapper +} // namespace arm_compute +#endif /* __ARM_COMPUTE_WRAPPER_CGT_H__ */ -- cgit v1.2.1