1/* 2 * Mesa 3-D graphics library 3 * 4 * Copyright (C) 1999-2007 Brian Paul All Rights Reserved. 5 * Copyright (C) 2018-2019 Intel Corporation 6 * 7 * Permission is hereby granted, free of charge, to any person obtaining a 8 * copy of this software and associated documentation files (the "Software"), 9 * to deal in the Software without restriction, including without limitation 10 * the rights to use, copy, modify, merge, publish, distribute, sublicense, 11 * and/or sell copies of the Software, and to permit persons to whom the 12 * Software is furnished to do so, subject to the following conditions: 13 * 14 * The above copyright notice and this permission notice shall be included 15 * in all copies or substantial portions of the Software. 16 * 17 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS 18 * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 19 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 20 * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR 21 * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, 22 * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR 23 * OTHER DEALINGS IN THE SOFTWARE. 24 */ 25 26#ifndef _HALF_FLOAT_H_ 27#define _HALF_FLOAT_H_ 28 29#include <stdbool.h> 30#include <stdint.h> 31#include <string.h> 32#include "util/u_cpu_detect.h" 33 34#if defined(USE_X86_64_ASM) 35#include <immintrin.h> 36#endif 37 38#ifdef __cplusplus 39extern "C" { 40#endif 41 42#define FP16_ONE ((uint16_t) 0x3c00) 43#define FP16_ZERO ((uint16_t) 0) 44 45uint16_t _mesa_float_to_half_slow(float val); 46float _mesa_half_to_float_slow(uint16_t val); 47uint8_t _mesa_half_to_unorm8(uint16_t v); 48uint16_t _mesa_uint16_div_64k_to_half(uint16_t v); 49 50/* 51 * _mesa_float_to_float16_rtz_slow is no more than a wrapper to the counterpart 52 * softfloat.h call. Still, softfloat.h conversion API is meant to be kept 53 * private. In other words, only use the API published here, instead of 54 * calling directly the softfloat.h one. 55 */ 56uint16_t _mesa_float_to_float16_rtz_slow(float val); 57 58static inline uint16_t 59_mesa_float_to_half(float val) 60{ 61#if defined(USE_X86_64_ASM) 62 if (util_get_cpu_caps()->has_f16c) { 63 __m128 in = {val}; 64 __m128i out; 65 66 /* $0 = round to nearest */ 67 __asm volatile("vcvtps2ph $0, %1, %0" : "=v"(out) : "v"(in)); 68 return out[0]; 69 } 70#endif 71 return _mesa_float_to_half_slow(val); 72} 73 74static inline float 75_mesa_half_to_float(uint16_t val) 76{ 77#if defined(USE_X86_64_ASM) 78 if (util_get_cpu_caps()->has_f16c) { 79 __m128i in = {val}; 80 __m128 out; 81 82 __asm volatile("vcvtph2ps %1, %0" : "=v"(out) : "v"(in)); 83 return out[0]; 84 } 85#endif 86 return _mesa_half_to_float_slow(val); 87} 88 89static inline uint16_t 90_mesa_float_to_float16_rtz(float val) 91{ 92#if defined(USE_X86_64_ASM) 93 if (util_get_cpu_caps()->has_f16c) { 94 __m128 in = {val}; 95 __m128i out; 96 97 /* $3 = round towards zero (truncate) */ 98 __asm volatile("vcvtps2ph $3, %1, %0" : "=v"(out) : "v"(in)); 99 return out[0]; 100 } 101#endif 102 return _mesa_float_to_float16_rtz_slow(val); 103} 104 105static inline uint16_t 106_mesa_float_to_float16_rtne(float val) 107{ 108 return _mesa_float_to_half(val); 109} 110 111static inline bool 112_mesa_half_is_negative(uint16_t h) 113{ 114 return !!(h & 0x8000); 115} 116 117 118#ifdef __cplusplus 119 120/* Helper class for disambiguating fp16 from uint16_t in C++ overloads */ 121 122struct float16_t { 123 uint16_t bits; 124 float16_t(float f) : bits(_mesa_float_to_half(f)) {} 125 float16_t(double d) : bits(_mesa_float_to_half(d)) {} 126 float16_t(uint16_t raw_bits) : bits(raw_bits) {} 127 static float16_t one() { return float16_t(FP16_ONE); } 128 static float16_t zero() { return float16_t(FP16_ZERO); } 129}; 130 131#endif 132 133 134#ifdef __cplusplus 135} /* extern C */ 136#endif 137 138#endif /* _HALF_FLOAT_H_ */ 139