File: smoothstep.cl

package info (click to toggle)
intel-graphics-compiler 1.0.17791.18-1
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 102,312 kB
  • sloc: cpp: 935,343; lisp: 286,143; ansic: 16,196; python: 3,279; yacc: 2,487; lex: 1,642; pascal: 300; sh: 174; makefile: 27
file content (42 lines) | stat: -rw-r--r-- 1,665 bytes parent folder | download | duplicates (3)
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
/*========================== 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 float SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(smoothstep, _f32_f32_f32, )(float edge0, float edge1, float x ){
    float div = (x - edge0) / (edge1 - edge0);
    float temp = SPIRV_OCL_BUILTIN(fclamp, _f32_f32_f32, )( div, 0.0f, 1.0f );
    return temp * temp * SPIRV_OCL_BUILTIN(mad, _f32_f32_f32, )( -2.0f, temp, 3.0f );
}

GENERATE_SPIRV_OCL_VECTOR_FUNCTIONS_3ARGS( smoothstep, float, float, f32 )

#if defined(cl_khr_fp64)

INLINE double SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(smoothstep, _f64_f64_f64, )(double edge0, double edge1, double x ){
    double div = (x - edge0) / (edge1 - edge0);
    double temp = SPIRV_OCL_BUILTIN(fclamp, _f64_f64_f64, )( div, 0.0, 1.0 );
    return temp * temp * SPIRV_OCL_BUILTIN(mad, _f64_f64_f64, )( -2.0, temp, 3.0 );
}

GENERATE_SPIRV_OCL_VECTOR_FUNCTIONS_3ARGS( smoothstep, double, double, f64 )

#endif // defined(cl_khr_fp64)

#if defined(cl_khr_fp16)

INLINE half SPIRV_OVERLOADABLE SPIRV_OCL_BUILTIN(smoothstep, _f16_f16_f16, )(half edge0, half edge1, half x ){
    half div = (x - edge0) / (edge1 - edge0);
    half temp = SPIRV_OCL_BUILTIN(fclamp, _f16_f16_f16, )( div, (half)0.0f, (half)1.0f );
    return temp * temp * SPIRV_OCL_BUILTIN(mad, _f16_f16_f16, )( (half)-2.0f, temp, (half)3.0f );
}

GENERATE_SPIRV_OCL_VECTOR_FUNCTIONS_3ARGS( smoothstep, half, half, f16 )

#endif // defined(cl_khr_fp16)