| 1 | // |
| 2 | // Redistribution and use in source and binary forms, with or without |
| 3 | // modification, are permitted provided that the following conditions |
| 4 | // are met: |
| 5 | // * Redistributions of source code must retain the above copyright |
| 6 | // notice, this list of conditions and the following disclaimer. |
| 7 | // * Redistributions in binary form must reproduce the above copyright |
| 8 | // notice, this list of conditions and the following disclaimer in the |
| 9 | // documentation and/or other materials provided with the distribution. |
| 10 | // * Neither the name of NVIDIA CORPORATION nor the names of its |
| 11 | // contributors may be used to endorse or promote products derived |
| 12 | // from this software without specific prior written permission. |
| 13 | // |
| 14 | // THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ''AS IS'' AND ANY |
| 15 | // EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE |
| 16 | // IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR |
| 17 | // PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR |
| 18 | // CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, |
| 19 | // EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, |
| 20 | // PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR |
| 21 | // PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY |
| 22 | // OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT |
| 23 | // (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE |
| 24 | // OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. |
| 25 | // |
| 26 | // Copyright (c) 2008-2021 NVIDIA Corporation. All rights reserved. |
| 27 | // Copyright (c) 2004-2008 AGEIA Technologies, Inc. All rights reserved. |
| 28 | // Copyright (c) 2001-2004 NovodeX AG. All rights reserved. |
| 29 | |
| 30 | #ifndef PSFOUNDATION_PSUNIXINTRINSICS_H |
| 31 | #define PSFOUNDATION_PSUNIXINTRINSICS_H |
| 32 | |
| 33 | #include "Ps.h" |
| 34 | #include "foundation/PxAssert.h" |
| 35 | #include <math.h> |
| 36 | |
| 37 | #if PX_ANDROID |
| 38 | #include <signal.h> // for Ns::debugBreak() { raise(SIGTRAP); } |
| 39 | #endif |
| 40 | |
| 41 | #if 0 |
| 42 | #include <libkern/OSAtomic.h> |
| 43 | #endif |
| 44 | |
| 45 | // this file is for internal intrinsics - that is, intrinsics that are used in |
| 46 | // cross platform code but do not appear in the API |
| 47 | |
| 48 | #if !(PX_LINUX || PX_ANDROID || PX_PS4 || PX_APPLE_FAMILY) |
| 49 | #error "This file should only be included by unix builds!!" |
| 50 | #endif |
| 51 | |
| 52 | namespace physx |
| 53 | { |
| 54 | namespace shdfnd |
| 55 | { |
| 56 | |
| 57 | PX_FORCE_INLINE void memoryBarrier() |
| 58 | { |
| 59 | __sync_synchronize(); |
| 60 | } |
| 61 | |
| 62 | /*! |
| 63 | Return the index of the highest set bit. Undefined for zero arg. |
| 64 | */ |
| 65 | PX_INLINE uint32_t highestSetBitUnsafe(uint32_t v) |
| 66 | { |
| 67 | |
| 68 | return 31 - __builtin_clz(v); |
| 69 | } |
| 70 | |
| 71 | /*! |
| 72 | Return the index of the highest set bit. Undefined for zero arg. |
| 73 | */ |
| 74 | PX_INLINE int32_t lowestSetBitUnsafe(uint32_t v) |
| 75 | { |
| 76 | return __builtin_ctz(v); |
| 77 | } |
| 78 | |
| 79 | /*! |
| 80 | Returns the index of the highest set bit. Returns 32 for v=0. |
| 81 | */ |
| 82 | PX_INLINE uint32_t countLeadingZeros(uint32_t v) |
| 83 | { |
| 84 | if(v) |
| 85 | return __builtin_clz(v); |
| 86 | else |
| 87 | return 32; |
| 88 | } |
| 89 | |
| 90 | /*! |
| 91 | Prefetch aligned 64B x86, 32b ARM around \c ptr+offset. |
| 92 | */ |
| 93 | PX_FORCE_INLINE void prefetchLine(const void* ptr, uint32_t offset = 0) |
| 94 | { |
| 95 | __builtin_prefetch(reinterpret_cast<const char* PX_RESTRICT>(ptr) + offset, 0, 3); |
| 96 | } |
| 97 | |
| 98 | /*! |
| 99 | Prefetch \c count bytes starting at \c ptr. |
| 100 | */ |
| 101 | #if PX_ANDROID || PX_IOS |
| 102 | PX_FORCE_INLINE void prefetch(const void* ptr, uint32_t count = 1) |
| 103 | { |
| 104 | const char* cp = static_cast<const char*>(ptr); |
| 105 | size_t p = reinterpret_cast<size_t>(ptr); |
| 106 | uint32_t startLine = uint32_t(p >> 5), endLine = uint32_t((p + count - 1) >> 5); |
| 107 | uint32_t lines = endLine - startLine + 1; |
| 108 | do |
| 109 | { |
| 110 | prefetchLine(cp); |
| 111 | cp += 32; |
| 112 | } while(--lines); |
| 113 | } |
| 114 | #else |
| 115 | PX_FORCE_INLINE void prefetch(const void* ptr, uint32_t count = 1) |
| 116 | { |
| 117 | const char* cp = reinterpret_cast<const char*>(ptr); |
| 118 | uint64_t p = size_t(ptr); |
| 119 | uint64_t startLine = p >> 6, endLine = (p + count - 1) >> 6; |
| 120 | uint64_t lines = endLine - startLine + 1; |
| 121 | do |
| 122 | { |
| 123 | prefetchLine(ptr: cp); |
| 124 | cp += 64; |
| 125 | } while(--lines); |
| 126 | } |
| 127 | #endif |
| 128 | |
| 129 | //! \brief platform-specific reciprocal |
| 130 | PX_CUDA_CALLABLE PX_FORCE_INLINE float recipFast(float a) |
| 131 | { |
| 132 | return 1.0f / a; |
| 133 | } |
| 134 | |
| 135 | //! \brief platform-specific fast reciprocal square root |
| 136 | PX_CUDA_CALLABLE PX_FORCE_INLINE float recipSqrtFast(float a) |
| 137 | { |
| 138 | return 1.0f / ::sqrtf(x: a); |
| 139 | } |
| 140 | |
| 141 | //! \brief platform-specific floor |
| 142 | PX_CUDA_CALLABLE PX_FORCE_INLINE float floatFloor(float x) |
| 143 | { |
| 144 | return ::floorf(x: x); |
| 145 | } |
| 146 | |
| 147 | #define NS_EXPECT_TRUE(x) x |
| 148 | #define NS_EXPECT_FALSE(x) x |
| 149 | |
| 150 | } // namespace shdfnd |
| 151 | } // namespace physx |
| 152 | |
| 153 | #endif // #ifndef PSFOUNDATION_PSUNIXINTRINSICS_H |
| 154 | |