1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
|
/*========================== begin_copyright_notice ============================
Copyright (C) 2017-2021 Intel Corporation
SPDX-License-Identifier: MIT
============================= end_copyright_notice ===========================*/
#include "../include/BiF_Definitions.cl"
#include "../../Headers/spirv.h"
INLINE float3 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v3f32_v3f32, )(float3 p0, float3 p1 ){
float3 result;
result.x = SPIRV_OCL_BUILTIN(fma, _f32_f32_f32, )(p0.y, p1.z, -p0.z * p1.y );
result.y = SPIRV_OCL_BUILTIN(fma, _f32_f32_f32, )(p0.z, p1.x, -p0.x * p1.z );
result.z = SPIRV_OCL_BUILTIN(fma, _f32_f32_f32, )(p0.x, p1.y, -p0.y * p1.x );
return result;
}
INLINE float4 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v4f32_v4f32, )(float4 p0, float4 p1 ){
float4 result;
result.xyz = SPIRV_OCL_BUILTIN(cross, _v3f32_v3f32, )( p0.xyz, p1.xyz );
result.w = 0.0f;
return result;
}
#if defined(cl_khr_fp64)
INLINE double3 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v3f64_v3f64, )(double3 p0, double3 p1 ){
double3 result;
result.x = SPIRV_OCL_BUILTIN(fma, _f64_f64_f64, )(p0.y, p1.z, -p0.z * p1.y );
result.y = SPIRV_OCL_BUILTIN(fma, _f64_f64_f64, )(p0.z, p1.x, -p0.x * p1.z );
result.z = SPIRV_OCL_BUILTIN(fma, _f64_f64_f64, )(p0.x, p1.y, -p0.y * p1.x );
return result;
}
INLINE double4 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v4f64_v4f64, )(double4 p0, double4 p1 ){
double4 result;
result.xyz = SPIRV_OCL_BUILTIN(cross, _v3f64_v3f64, )( p0.xyz, p1.xyz );
result.w = (half) 0.0;
return result;
}
#endif // defined(cl_khr_fp64)
#if defined(cl_khr_fp16)
INLINE half3 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v3f16_v3f16, )(half3 p0, half3 p1 ){
float3 ret = SPIRV_OCL_BUILTIN(cross, _v3f32_v3f32, )(SPIRV_BUILTIN(FConvert, _v3f32_v3f16, _Rfloat3)(p0), SPIRV_BUILTIN(FConvert, _v3f32_v3f16, _Rfloat3)(p1));
return SPIRV_BUILTIN(FConvert, _v3f16_v3f32, _Rhalf3)(ret);
}
INLINE half4 SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(cross, _v4f16_v4f16, )(half4 p0, half4 p1 ){
half4 result;
result.xyz = SPIRV_OCL_BUILTIN(cross, _v3f16_v3f16, )( p0.xyz, p1.xyz );
result.w = (half)0.0f;
return result;
}
#endif // defined(cl_khr_fp16)
|
