aboutsummaryrefslogtreecommitdiff
path: root/PxShared/include/foundation/nx/PxNXIntrinsics.h
blob: adf3bf5578161178eeb7821cad7f9bb847f69da6 (plain) (blame)
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
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
/*
 * Copyright (c) 2008-2015, NVIDIA CORPORATION.  All rights reserved.
 *
 * NVIDIA CORPORATION and its licensors retain all intellectual property
 * and proprietary rights in and to this software, related documentation
 * and any modifications thereto.  Any use, reproduction, disclosure or
 * distribution of this software and related documentation without an express
 * license agreement from NVIDIA CORPORATION is strictly prohibited.
 */
// Copyright (c) 2004-2008 AGEIA Technologies, Inc. All rights reserved.
// Copyright (c) 2001-2004 NovodeX AG. All rights reserved.  


#ifndef PX_FOUNDATION_PX_NX_INTRINSICS_H
#define PX_FOUNDATION_PX_NX_INTRINSICS_H

#include "foundation/Px.h"
#include "foundation/PxAssert.h"

#if !PX_NX
	#error "This file should only be included by NX builds!!"
#endif

#include <math.h>
#include <float.h>

#include "nn/cstd/cstd_CMath.h"

#if !PX_DOXYGEN
namespace physx
{
namespace intrinsics
{
#endif

	//! \brief platform-specific absolute value
	PX_CUDA_CALLABLE PX_FORCE_INLINE float abs(float a)						{	return ::fabsf(a);	}

	//! \brief platform-specific select float
	PX_CUDA_CALLABLE PX_FORCE_INLINE float fsel(float a, float b, float c)	{	return (a >= 0.0f) ? b : c;	}

	//! \brief platform-specific sign
	PX_CUDA_CALLABLE PX_FORCE_INLINE float sign(float a)					{	return (a >= 0.0f) ? 1.0f : -1.0f; }

	//! \brief platform-specific reciprocal
	PX_CUDA_CALLABLE PX_FORCE_INLINE float recip(float a)					{	return 1.0f/a;			}

	//! \brief platform-specific reciprocal estimate
	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipFast(float a)				{	return 1.0f/a;			}

	//! \brief platform-specific square root
	PX_CUDA_CALLABLE PX_FORCE_INLINE float sqrt(float a)					{	return ::sqrtf(a);	}

	//! \brief platform-specific reciprocal square root
	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipSqrt(float a)				{   return 1.0f/::sqrtf(a); }

	//! \brief platform-specific reciprocal square root estimate
	PX_CUDA_CALLABLE PX_FORCE_INLINE float recipSqrtFast(float a)			{	return 1.0f/::sqrtf(a); }

	//! \brief platform-specific sine
	PX_CUDA_CALLABLE PX_FORCE_INLINE float sin(float a)						{   return ::sinf(a); }

	//! \brief platform-specific cosine
	PX_CUDA_CALLABLE PX_FORCE_INLINE float cos(float a)						{   return ::cosf(a); }

	//! \brief platform-specific minimum
	PX_CUDA_CALLABLE PX_FORCE_INLINE float selectMin(float a, float b)		{	return a<b ? a : b;	}

	//! \brief platform-specific maximum
	PX_CUDA_CALLABLE PX_FORCE_INLINE float selectMax(float a, float b)		{	return a>b ? a : b; }

	//! \brief platform-specific finiteness check
	PX_CUDA_CALLABLE PX_FORCE_INLINE bool isFinite(float a)
	{
#ifdef __CUDACC__
		return isfinite(a) ? true : false;
#else
		return !nn::cstd::IsNan(a) && !nn::cstd::IsInf(a);
#endif
	}

	//! \brief platform-specific finiteness check
	PX_CUDA_CALLABLE PX_FORCE_INLINE bool isFinite(double a)
	{
#ifdef __CUDACC__
		return isfinite(a) ? true : false;
#else
		return !nn::cstd::IsNan(a) && !nn::cstd::IsInf(a);
#endif
	}

	/*!
	Sets \c count bytes starting at \c dst to zero.
	*/
	PX_FORCE_INLINE void* memZero(void* PX_RESTRICT dest, uint32_t count)
	{
		return memset(dest, 0, count);
	}

	/*!
	Sets \c count bytes starting at \c dst to \c c.
	*/
	PX_FORCE_INLINE void* memSet(void* PX_RESTRICT dest, int32_t c, uint32_t count)
	{
		return memset(dest, c, count);
	}

	/*!
	Copies \c count bytes from \c src to \c dst. User memMove if regions overlap.
	*/
	PX_FORCE_INLINE void* memCopy(void* PX_RESTRICT dest, const void* PX_RESTRICT src, uint32_t count)
	{
		return memcpy(dest, src, count);
	}

	/*!
	Copies \c count bytes from \c src to \c dst. Supports overlapping regions.
	*/
	PX_FORCE_INLINE void* memMove(void* PX_RESTRICT dest, const void* PX_RESTRICT src, uint32_t count)
	{
		return memmove(dest, src, count);
	}

	/*!
	Set 128B to zero starting at \c dst+offset. Must be aligned.
	*/
	PX_FORCE_INLINE void memZero128(void* PX_RESTRICT dest, uint32_t offset = 0)
	{
		PX_ASSERT(((size_t(dest)+offset) & 0x7f) == 0);
		memSet((char* PX_RESTRICT)dest+offset, 0, 128);
	}

#if !PX_DOXYGEN
} // namespace intrinsics
} // namespace physx
#endif

#endif