Vec4.inl

Go to the documentation of this file.

// Jolt Physics Library (https://github.com/jrouwe/JoltPhysics)
// SPDX-FileCopyrightText: 2021 Jorrit Rouwe
// SPDX-License-Identifier: MIT
 
#include <Jolt/Math/Trigonometry.h>
#include <Jolt/Math/Vec3.h>
#include <Jolt/Math/UVec4.h>
 
JPH_NAMESPACE_BEGIN
 
// Constructor
Vec4::Vec4(Vec3Arg inRHS) :
    mValue(inRHS.mValue)
{
}
 
Vec4::Vec4(Vec3Arg inRHS, float inW)
{
#if defined(JPH_USE_SSE4_1)
    mValue = _mm_blend_ps(inRHS.mValue, _mm_set1_ps(inW), 8);
#elif defined(JPH_USE_NEON)
    mValue = vsetq_lane_f32(inW, inRHS.mValue, 3);
#else
    for (int i = 0; i < 3; i++)
        mF32[i] = inRHS.mF32[i];
    mF32[3] = inW;
#endif
}
 
Vec4::Vec4(float inX, float inY, float inZ, float inW)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_set_ps(inW, inZ, inY, inX);
#elif defined(JPH_USE_NEON)
    uint32x2_t xy = vcreate_u32(static_cast<uint64>(BitCast<uint32>(inX)) | (static_cast<uint64>(BitCast<uint32>(inY)) << 32));
    uint32x2_t zw = vcreate_u32(static_cast<uint64>(BitCast<uint32>(inZ)) | (static_cast<uint64>(BitCast<uint32>(inW)) << 32));
    mValue = vreinterpretq_f32_u32(vcombine_u32(xy, zw));
#else
    mF32[0] = inX;
    mF32[1] = inY;
    mF32[2] = inZ;
    mF32[3] = inW;
#endif
}
 
template<uint32 SwizzleX, uint32 SwizzleY, uint32 SwizzleZ, uint32 SwizzleW>
Vec4 Vec4::Swizzle() const
{
    static_assert(SwizzleX <= 3, "SwizzleX template parameter out of range");
    static_assert(SwizzleY <= 3, "SwizzleY template parameter out of range");
    static_assert(SwizzleZ <= 3, "SwizzleZ template parameter out of range");
    static_assert(SwizzleW <= 3, "SwizzleW template parameter out of range");
 
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(SwizzleW, SwizzleZ, SwizzleY, SwizzleX));
#elif defined(JPH_USE_NEON)
    return JPH_NEON_SHUFFLE_F32x4(mValue, mValue, SwizzleX, SwizzleY, SwizzleZ, SwizzleW);
#else
    return Vec4(mF32[SwizzleX], mF32[SwizzleY], mF32[SwizzleZ], mF32[SwizzleW]);
#endif
}
 
Vec4 Vec4::sZero()
{
#if defined(JPH_USE_SSE)
    return _mm_setzero_ps();
#elif defined(JPH_USE_NEON)
    return vdupq_n_f32(0);
#else
    return Vec4(0, 0, 0, 0);
#endif
}
 
Vec4 Vec4::sReplicate(float inV)
{
#if defined(JPH_USE_SSE)
    return _mm_set1_ps(inV);
#elif defined(JPH_USE_NEON)
    return vdupq_n_f32(inV);
#else
    return Vec4(inV, inV, inV, inV);
#endif
}
 
Vec4 Vec4::sOne()
{
    return sReplicate(1.0f);
}
 
Vec4 Vec4::sNaN()
{
    return sReplicate(numeric_limits<float>::quiet_NaN());
}
 
Vec4 Vec4::sLoadFloat4(const Float4 *inV)
{
#if defined(JPH_USE_SSE)
    return _mm_loadu_ps(&inV->x);
#elif defined(JPH_USE_NEON)
    return vld1q_f32(&inV->x);
#else
    return Vec4(inV->x, inV->y, inV->z, inV->w);
#endif
}
 
Vec4 Vec4::sLoadFloat4Aligned(const Float4 *inV)
{
#if defined(JPH_USE_SSE)
    return _mm_load_ps(&inV->x);
#elif defined(JPH_USE_NEON)
    return vld1q_f32(&inV->x);
#else
    return Vec4(inV->x, inV->y, inV->z, inV->w);
#endif
}
 
template <const int Scale>
Vec4 Vec4::sGatherFloat4(const float *inBase, UVec4Arg inOffsets)
{
#if defined(JPH_USE_SSE)
    #ifdef JPH_USE_AVX2
        return _mm_i32gather_ps(inBase, inOffsets.mValue, Scale);
    #else
        const uint8 *base = reinterpret_cast<const uint8 *>(inBase);
        Type x = _mm_load_ss(reinterpret_cast<const float *>(base + inOffsets.GetX() * Scale));
        Type y = _mm_load_ss(reinterpret_cast<const float *>(base + inOffsets.GetY() * Scale));
        Type xy = _mm_unpacklo_ps(x, y);
        Type z = _mm_load_ss(reinterpret_cast<const float *>(base + inOffsets.GetZ() * Scale));
        Type w = _mm_load_ss(reinterpret_cast<const float *>(base + inOffsets.GetW() * Scale));
        Type zw = _mm_unpacklo_ps(z, w);
        return _mm_movelh_ps(xy, zw);
    #endif
#else
    const uint8 *base = reinterpret_cast<const uint8 *>(inBase);
    float x = *reinterpret_cast<const float *>(base + inOffsets.GetX() * Scale);
    float y = *reinterpret_cast<const float *>(base + inOffsets.GetY() * Scale);
    float z = *reinterpret_cast<const float *>(base + inOffsets.GetZ() * Scale);
    float w = *reinterpret_cast<const float *>(base + inOffsets.GetW() * Scale);
    return Vec4(x, y, z, w);
#endif
}
 
Vec4 Vec4::sMin(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_min_ps(inV1.mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vminq_f32(inV1.mValue, inV2.mValue);
#else
    return Vec4(min(inV1.mF32[0], inV2.mF32[0]),
                min(inV1.mF32[1], inV2.mF32[1]),
                min(inV1.mF32[2], inV2.mF32[2]),
                min(inV1.mF32[3], inV2.mF32[3]));
#endif
}
 
Vec4 Vec4::sMax(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_max_ps(inV1.mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vmaxq_f32(inV1.mValue, inV2.mValue);
#else
    return Vec4(max(inV1.mF32[0], inV2.mF32[0]),
                max(inV1.mF32[1], inV2.mF32[1]),
                max(inV1.mF32[2], inV2.mF32[2]),
                max(inV1.mF32[3], inV2.mF32[3]));
#endif
}
 
Vec4 Vec4::sClamp(Vec4Arg inV, Vec4Arg inMin, Vec4Arg inMax)
{
    return sMax(sMin(inV, inMax), inMin);
}
 
UVec4 Vec4::sEquals(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_castps_si128(_mm_cmpeq_ps(inV1.mValue, inV2.mValue));
#elif defined(JPH_USE_NEON)
    return vceqq_f32(inV1.mValue, inV2.mValue);
#else
    return UVec4(inV1.mF32[0] == inV2.mF32[0]? 0xffffffffu : 0,
                 inV1.mF32[1] == inV2.mF32[1]? 0xffffffffu : 0,
                 inV1.mF32[2] == inV2.mF32[2]? 0xffffffffu : 0,
                 inV1.mF32[3] == inV2.mF32[3]? 0xffffffffu : 0);
#endif
}
 
UVec4 Vec4::sLess(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_castps_si128(_mm_cmplt_ps(inV1.mValue, inV2.mValue));
#elif defined(JPH_USE_NEON)
    return vcltq_f32(inV1.mValue, inV2.mValue);
#else
    return UVec4(inV1.mF32[0] < inV2.mF32[0]? 0xffffffffu : 0,
                 inV1.mF32[1] < inV2.mF32[1]? 0xffffffffu : 0,
                 inV1.mF32[2] < inV2.mF32[2]? 0xffffffffu : 0,
                 inV1.mF32[3] < inV2.mF32[3]? 0xffffffffu : 0);
#endif
}
 
UVec4 Vec4::sLessOrEqual(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_castps_si128(_mm_cmple_ps(inV1.mValue, inV2.mValue));
#elif defined(JPH_USE_NEON)
    return vcleq_f32(inV1.mValue, inV2.mValue);
#else
    return UVec4(inV1.mF32[0] <= inV2.mF32[0]? 0xffffffffu : 0,
                 inV1.mF32[1] <= inV2.mF32[1]? 0xffffffffu : 0,
                 inV1.mF32[2] <= inV2.mF32[2]? 0xffffffffu : 0,
                 inV1.mF32[3] <= inV2.mF32[3]? 0xffffffffu : 0);
#endif
}
 
UVec4 Vec4::sGreater(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_castps_si128(_mm_cmpgt_ps(inV1.mValue, inV2.mValue));
#elif defined(JPH_USE_NEON)
    return vcgtq_f32(inV1.mValue, inV2.mValue);
#else
    return UVec4(inV1.mF32[0] > inV2.mF32[0]? 0xffffffffu : 0,
                 inV1.mF32[1] > inV2.mF32[1]? 0xffffffffu : 0,
                 inV1.mF32[2] > inV2.mF32[2]? 0xffffffffu : 0,
                 inV1.mF32[3] > inV2.mF32[3]? 0xffffffffu : 0);
#endif
}
 
UVec4 Vec4::sGreaterOrEqual(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_castps_si128(_mm_cmpge_ps(inV1.mValue, inV2.mValue));
#elif defined(JPH_USE_NEON)
    return vcgeq_f32(inV1.mValue, inV2.mValue);
#else
    return UVec4(inV1.mF32[0] >= inV2.mF32[0]? 0xffffffffu : 0,
                 inV1.mF32[1] >= inV2.mF32[1]? 0xffffffffu : 0,
                 inV1.mF32[2] >= inV2.mF32[2]? 0xffffffffu : 0,
                 inV1.mF32[3] >= inV2.mF32[3]? 0xffffffffu : 0);
#endif
}
 
Vec4 Vec4::sFusedMultiplyAdd(Vec4Arg inMul1, Vec4Arg inMul2, Vec4Arg inAdd)
{
#if defined(JPH_USE_SSE)
    #ifdef JPH_USE_FMADD
        return _mm_fmadd_ps(inMul1.mValue, inMul2.mValue, inAdd.mValue);
    #else
        return _mm_add_ps(_mm_mul_ps(inMul1.mValue, inMul2.mValue), inAdd.mValue);
    #endif
#elif defined(JPH_USE_NEON)
    return vmlaq_f32(inAdd.mValue, inMul1.mValue, inMul2.mValue);
#else
    return Vec4(inMul1.mF32[0] * inMul2.mF32[0] + inAdd.mF32[0],
                inMul1.mF32[1] * inMul2.mF32[1] + inAdd.mF32[1],
                inMul1.mF32[2] * inMul2.mF32[2] + inAdd.mF32[2],
                inMul1.mF32[3] * inMul2.mF32[3] + inAdd.mF32[3]);
#endif
}
 
Vec4 Vec4::sSelect(Vec4Arg inNotSet, Vec4Arg inSet, UVec4Arg inControl)
{
#if defined(JPH_USE_SSE4_1) && !defined(JPH_PLATFORM_WASM) // _mm_blendv_ps has problems on FireFox
    return _mm_blendv_ps(inNotSet.mValue, inSet.mValue, _mm_castsi128_ps(inControl.mValue));
#elif defined(JPH_USE_SSE)
    __m128 is_set = _mm_castsi128_ps(_mm_srai_epi32(inControl.mValue, 31));
    return _mm_or_ps(_mm_and_ps(is_set, inSet.mValue), _mm_andnot_ps(is_set, inNotSet.mValue));
#elif defined(JPH_USE_NEON)
    return vbslq_f32(vreinterpretq_u32_s32(vshrq_n_s32(vreinterpretq_s32_u32(inControl.mValue), 31)), inSet.mValue, inNotSet.mValue);
#else
    Vec4 result;
    for (int i = 0; i < 4; i++)
        result.mF32[i] = (inControl.mU32[i] & 0x80000000u) ? inSet.mF32[i] : inNotSet.mF32[i];
    return result;
#endif
}
 
Vec4 Vec4::sOr(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_or_ps(inV1.mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vreinterpretq_f32_u32(vorrq_u32(vreinterpretq_u32_f32(inV1.mValue), vreinterpretq_u32_f32(inV2.mValue)));
#else
    return UVec4::sOr(inV1.ReinterpretAsInt(), inV2.ReinterpretAsInt()).ReinterpretAsFloat();
#endif
}
 
Vec4 Vec4::sXor(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_xor_ps(inV1.mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vreinterpretq_f32_u32(veorq_u32(vreinterpretq_u32_f32(inV1.mValue), vreinterpretq_u32_f32(inV2.mValue)));
#else
    return UVec4::sXor(inV1.ReinterpretAsInt(), inV2.ReinterpretAsInt()).ReinterpretAsFloat();
#endif
}
 
Vec4 Vec4::sAnd(Vec4Arg inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_and_ps(inV1.mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vreinterpretq_f32_u32(vandq_u32(vreinterpretq_u32_f32(inV1.mValue), vreinterpretq_u32_f32(inV2.mValue)));
#else
    return UVec4::sAnd(inV1.ReinterpretAsInt(), inV2.ReinterpretAsInt()).ReinterpretAsFloat();
#endif
}
 
void Vec4::sSort4(Vec4 &ioValue, UVec4 &ioIndex)
{
    // Pass 1, test 1st vs 3rd, 2nd vs 4th
    Vec4 v1 = ioValue.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>();
    UVec4 i1 = ioIndex.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>();
    UVec4 c1 = sLess(ioValue, v1).Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_Z, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v1, c1);
    ioIndex = UVec4::sSelect(ioIndex, i1, c1);
 
    // Pass 2, test 1st vs 2nd, 3rd vs 4th
    Vec4 v2 = ioValue.Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>();
    UVec4 i2 = ioIndex.Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>();
    UVec4 c2 = sLess(ioValue, v2).Swizzle<SWIZZLE_Y, SWIZZLE_Y, SWIZZLE_W, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v2, c2);
    ioIndex = UVec4::sSelect(ioIndex, i2, c2);
 
    // Pass 3, test 2nd vs 3rd component
    Vec4 v3 = ioValue.Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Y, SWIZZLE_W>();
    UVec4 i3 = ioIndex.Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Y, SWIZZLE_W>();
    UVec4 c3 = sLess(ioValue, v3).Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Z, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v3, c3);
    ioIndex = UVec4::sSelect(ioIndex, i3, c3);
}
 
void Vec4::sSort4Reverse(Vec4 &ioValue, UVec4 &ioIndex)
{
    // Pass 1, test 1st vs 3rd, 2nd vs 4th
    Vec4 v1 = ioValue.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>();
    UVec4 i1 = ioIndex.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>();
    UVec4 c1 = sGreater(ioValue, v1).Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_Z, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v1, c1);
    ioIndex = UVec4::sSelect(ioIndex, i1, c1);
 
    // Pass 2, test 1st vs 2nd, 3rd vs 4th
    Vec4 v2 = ioValue.Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>();
    UVec4 i2 = ioIndex.Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>();
    UVec4 c2 = sGreater(ioValue, v2).Swizzle<SWIZZLE_Y, SWIZZLE_Y, SWIZZLE_W, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v2, c2);
    ioIndex = UVec4::sSelect(ioIndex, i2, c2);
 
    // Pass 3, test 2nd vs 3rd component
    Vec4 v3 = ioValue.Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Y, SWIZZLE_W>();
    UVec4 i3 = ioIndex.Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Y, SWIZZLE_W>();
    UVec4 c3 = sGreater(ioValue, v3).Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_Z, SWIZZLE_W>();
    ioValue = sSelect(ioValue, v3, c3);
    ioIndex = UVec4::sSelect(ioIndex, i3, c3);
}
 
bool Vec4::operator == (Vec4Arg inV2) const
{
    return sEquals(*this, inV2).TestAllTrue();
}
 
bool Vec4::IsClose(Vec4Arg inV2, float inMaxDistSq) const
{
    return (inV2 - *this).LengthSq() <= inMaxDistSq;
}
 
bool Vec4::IsNearZero(float inMaxDistSq) const
{
    return LengthSq() <= inMaxDistSq;
}
 
bool Vec4::IsNormalized(float inTolerance) const
{
    return abs(LengthSq() - 1.0f) <= inTolerance;
}
 
bool Vec4::IsNaN() const
{
#if defined(JPH_USE_AVX512)
    return _mm_fpclass_ps_mask(mValue, 0b10000001) != 0;
#elif defined(JPH_USE_SSE)
    return _mm_movemask_ps(_mm_cmpunord_ps(mValue, mValue)) != 0;
#elif defined(JPH_USE_NEON)
    uint32x4_t is_equal = vceqq_f32(mValue, mValue); // If a number is not equal to itself it's a NaN
    return vaddvq_u32(vshrq_n_u32(is_equal, 31)) != 4;
#else
    return isnan(mF32[0]) || isnan(mF32[1]) || isnan(mF32[2]) || isnan(mF32[3]);
#endif
}
 
Vec4 Vec4::operator * (Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_mul_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vmulq_f32(mValue, inV2.mValue);
#else
    return Vec4(mF32[0] * inV2.mF32[0],
                mF32[1] * inV2.mF32[1],
                mF32[2] * inV2.mF32[2],
                mF32[3] * inV2.mF32[3]);
#endif
}
 
Vec4 Vec4::operator * (float inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_mul_ps(mValue, _mm_set1_ps(inV2));
#elif defined(JPH_USE_NEON)
    return vmulq_n_f32(mValue, inV2);
#else
    return Vec4(mF32[0] * inV2, mF32[1] * inV2, mF32[2] * inV2, mF32[3] * inV2);
#endif
}
 
Vec4 operator * (float inV1, Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    return _mm_mul_ps(_mm_set1_ps(inV1), inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vmulq_n_f32(inV2.mValue, inV1);
#else
    return Vec4(inV1 * inV2.mF32[0],
                inV1 * inV2.mF32[1],
                inV1 * inV2.mF32[2],
                inV1 * inV2.mF32[3]);
#endif
}
 
Vec4 Vec4::operator / (float inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_div_ps(mValue, _mm_set1_ps(inV2));
#elif defined(JPH_USE_NEON)
    return vdivq_f32(mValue, vdupq_n_f32(inV2));
#else
    return Vec4(mF32[0] / inV2, mF32[1] / inV2, mF32[2] / inV2, mF32[3] / inV2);
#endif
}
 
Vec4 &Vec4::operator *= (float inV2)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_mul_ps(mValue, _mm_set1_ps(inV2));
#elif defined(JPH_USE_NEON)
    mValue = vmulq_n_f32(mValue, inV2);
#else
    for (int i = 0; i < 4; ++i)
        mF32[i] *= inV2;
#endif
    return *this;
}
 
Vec4 &Vec4::operator *= (Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_mul_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    mValue = vmulq_f32(mValue, inV2.mValue);
#else
    for (int i = 0; i < 4; ++i)
        mF32[i] *= inV2.mF32[i];
#endif
    return *this;
}
 
Vec4 &Vec4::operator /= (float inV2)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_div_ps(mValue, _mm_set1_ps(inV2));
#elif defined(JPH_USE_NEON)
    mValue = vdivq_f32(mValue, vdupq_n_f32(inV2));
#else
    for (int i = 0; i < 4; ++i)
        mF32[i] /= inV2;
#endif
    return *this;
}
 
Vec4 Vec4::operator + (Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_add_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vaddq_f32(mValue, inV2.mValue);
#else
    return Vec4(mF32[0] + inV2.mF32[0],
                mF32[1] + inV2.mF32[1],
                mF32[2] + inV2.mF32[2],
                mF32[3] + inV2.mF32[3]);
#endif
}
 
Vec4 &Vec4::operator += (Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_add_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    mValue = vaddq_f32(mValue, inV2.mValue);
#else
    for (int i = 0; i < 4; ++i)
        mF32[i] += inV2.mF32[i];
#endif
    return *this;
}
 
Vec4 Vec4::operator - () const
{
#if defined(JPH_USE_SSE)
    return _mm_sub_ps(_mm_setzero_ps(), mValue);
#elif defined(JPH_USE_NEON)
    #ifdef JPH_CROSS_PLATFORM_DETERMINISTIC
        return vsubq_f32(vdupq_n_f32(0), mValue);
    #else
        return vnegq_f32(mValue);
    #endif
#else
    #ifdef JPH_CROSS_PLATFORM_DETERMINISTIC
        return Vec4(0.0f - mF32[0], 0.0f - mF32[1], 0.0f - mF32[2], 0.0f - mF32[3]);
    #else
        return Vec4(-mF32[0], -mF32[1], -mF32[2], -mF32[3]);
    #endif
#endif
}
 
Vec4 Vec4::operator - (Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_sub_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vsubq_f32(mValue, inV2.mValue);
#else
    return Vec4(mF32[0] - inV2.mF32[0],
                mF32[1] - inV2.mF32[1],
                mF32[2] - inV2.mF32[2],
                mF32[3] - inV2.mF32[3]);
#endif
}
 
Vec4 &Vec4::operator -= (Vec4Arg inV2)
{
#if defined(JPH_USE_SSE)
    mValue = _mm_sub_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    mValue = vsubq_f32(mValue, inV2.mValue);
#else
    for (int i = 0; i < 4; ++i)
        mF32[i] -= inV2.mF32[i];
#endif
    return *this;
}
 
Vec4 Vec4::operator / (Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE)
    return _mm_div_ps(mValue, inV2.mValue);
#elif defined(JPH_USE_NEON)
    return vdivq_f32(mValue, inV2.mValue);
#else
    return Vec4(mF32[0] / inV2.mF32[0],
                mF32[1] / inV2.mF32[1],
                mF32[2] / inV2.mF32[2],
                mF32[3] / inV2.mF32[3]);
#endif
}
 
Vec4 Vec4::SplatX() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(0, 0, 0, 0));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 0);
#else
    return Vec4(mF32[0], mF32[0], mF32[0], mF32[0]);
#endif
}
 
Vec4 Vec4::SplatY() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(1, 1, 1, 1));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 1);
#else
    return Vec4(mF32[1], mF32[1], mF32[1], mF32[1]);
#endif
}
 
Vec4 Vec4::SplatZ() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(2, 2, 2, 2));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 2);
#else
    return Vec4(mF32[2], mF32[2], mF32[2], mF32[2]);
#endif
}
 
Vec4 Vec4::SplatW() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(3, 3, 3, 3));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 3);
#else
    return Vec4(mF32[3], mF32[3], mF32[3], mF32[3]);
#endif
}
 
Vec3 Vec4::SplatX3() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(0, 0, 0, 0));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 0);
#else
    return Vec3(mF32[0], mF32[0], mF32[0]);
#endif
}
 
Vec3 Vec4::SplatY3() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(1, 1, 1, 1));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 1);
#else
    return Vec3(mF32[1], mF32[1], mF32[1]);
#endif
}
 
Vec3 Vec4::SplatZ3() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(2, 2, 2, 2));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 2);
#else
    return Vec3(mF32[2], mF32[2], mF32[2]);
#endif
}
 
Vec3 Vec4::SplatW3() const
{
#if defined(JPH_USE_SSE)
    return _mm_shuffle_ps(mValue, mValue, _MM_SHUFFLE(3, 3, 3, 3));
#elif defined(JPH_USE_NEON)
    return vdupq_laneq_f32(mValue, 3);
#else
    return Vec3(mF32[3], mF32[3], mF32[3]);
#endif
}
 
int Vec4::GetLowestComponentIndex() const
{
    // Get the minimum value in all 4 components
    Vec4 value = Vec4::sMin(*this, Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>());
    value = Vec4::sMin(value, value.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>());
 
    // Compare with the original vector to find which component is equal to the minimum value
    return CountTrailingZeros(Vec4::sEquals(*this, value).GetTrues());
}
 
int Vec4::GetHighestComponentIndex() const
{
    // Get the maximum value in all 4 components
    Vec4 value = Vec4::sMax(*this, Swizzle<SWIZZLE_Y, SWIZZLE_X, SWIZZLE_W, SWIZZLE_Z>());
    value = Vec4::sMax(value, value.Swizzle<SWIZZLE_Z, SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y>());
 
    // Compare with the original vector to find which component is equal to the maximum value
    return CountTrailingZeros(Vec4::sEquals(*this, value).GetTrues());
}
 
Vec4 Vec4::Abs() const
{
#if defined(JPH_USE_AVX512)
    return _mm_range_ps(mValue, mValue, 0b1000);
#elif defined(JPH_USE_SSE)
    return _mm_max_ps(_mm_sub_ps(_mm_setzero_ps(), mValue), mValue);
#elif defined(JPH_USE_NEON)
    return vabsq_f32(mValue);
#else
    return Vec4(abs(mF32[0]), abs(mF32[1]), abs(mF32[2]), abs(mF32[3]));
#endif
}
 
Vec4 Vec4::Reciprocal() const
{
    return sOne() / mValue;
}
 
Vec4 Vec4::DotV(Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE4_1)
    return _mm_dp_ps(mValue, inV2.mValue, 0xff);
#elif defined(JPH_USE_NEON)
    float32x4_t mul = vmulq_f32(mValue, inV2.mValue);
    return vdupq_n_f32(vaddvq_f32(mul));
#else
    // Brackets placed so that the order is consistent with the vectorized version
    return Vec4::sReplicate((mF32[0] * inV2.mF32[0] + mF32[1] * inV2.mF32[1]) + (mF32[2] * inV2.mF32[2] + mF32[3] * inV2.mF32[3]));
#endif
}
 
float Vec4::Dot(Vec4Arg inV2) const
{
#if defined(JPH_USE_SSE4_1)
    return _mm_cvtss_f32(_mm_dp_ps(mValue, inV2.mValue, 0xff));
#elif defined(JPH_USE_NEON)
    float32x4_t mul = vmulq_f32(mValue, inV2.mValue);
    return vaddvq_f32(mul);
#else
    // Brackets placed so that the order is consistent with the vectorized version
    return (mF32[0] * inV2.mF32[0] + mF32[1] * inV2.mF32[1]) + (mF32[2] * inV2.mF32[2] + mF32[3] * inV2.mF32[3]);
#endif
}
 
float Vec4::LengthSq() const
{
#if defined(JPH_USE_SSE4_1)
    return _mm_cvtss_f32(_mm_dp_ps(mValue, mValue, 0xff));
#elif defined(JPH_USE_NEON)
    float32x4_t mul = vmulq_f32(mValue, mValue);
    return vaddvq_f32(mul);
#else
    // Brackets placed so that the order is consistent with the vectorized version
    return (mF32[0] * mF32[0] + mF32[1] * mF32[1]) + (mF32[2] * mF32[2] + mF32[3] * mF32[3]);
#endif
}
 
float Vec4::Length() const
{
#if defined(JPH_USE_SSE4_1)
    return _mm_cvtss_f32(_mm_sqrt_ss(_mm_dp_ps(mValue, mValue, 0xff)));
#elif defined(JPH_USE_NEON)
    float32x4_t mul = vmulq_f32(mValue, mValue);
    float32x2_t sum = vdup_n_f32(vaddvq_f32(mul));
    return vget_lane_f32(vsqrt_f32(sum), 0);
#else
    // Brackets placed so that the order is consistent with the vectorized version
    return sqrt((mF32[0] * mF32[0] + mF32[1] * mF32[1]) + (mF32[2] * mF32[2] + mF32[3] * mF32[3]));
#endif
}
 
Vec4 Vec4::Sqrt() const
{
#if defined(JPH_USE_SSE)
    return _mm_sqrt_ps(mValue);
#elif defined(JPH_USE_NEON)
    return vsqrtq_f32(mValue);
#else
    return Vec4(sqrt(mF32[0]), sqrt(mF32[1]), sqrt(mF32[2]), sqrt(mF32[3]));
#endif
}
 
 
Vec4 Vec4::GetSign() const
{
#if defined(JPH_USE_AVX512)
    return _mm_fixupimm_ps(mValue, mValue, _mm_set1_epi32(0xA9A90A00), 0);
#elif defined(JPH_USE_SSE)
    Type minus_one = _mm_set1_ps(-1.0f);
    Type one = _mm_set1_ps(1.0f);
    return _mm_or_ps(_mm_and_ps(mValue, minus_one), one);
#elif defined(JPH_USE_NEON)
    Type minus_one = vdupq_n_f32(-1.0f);
    Type one = vdupq_n_f32(1.0f);
    return vreinterpretq_f32_u32(vorrq_u32(vandq_u32(vreinterpretq_u32_f32(mValue), vreinterpretq_u32_f32(minus_one)), vreinterpretq_u32_f32(one)));
#else
    return Vec4(std::signbit(mF32[0])? -1.0f : 1.0f,
                std::signbit(mF32[1])? -1.0f : 1.0f,
                std::signbit(mF32[2])? -1.0f : 1.0f,
                std::signbit(mF32[3])? -1.0f : 1.0f);
#endif
}
 
template <int X, int Y, int Z, int W>
JPH_INLINE Vec4 Vec4::FlipSign() const
{
    static_assert(X == 1 || X == -1, "X must be 1 or -1");
    static_assert(Y == 1 || Y == -1, "Y must be 1 or -1");
    static_assert(Z == 1 || Z == -1, "Z must be 1 or -1");
    static_assert(W == 1 || W == -1, "W must be 1 or -1");
    return Vec4::sXor(*this, Vec4(X > 0? 0.0f : -0.0f, Y > 0? 0.0f : -0.0f, Z > 0? 0.0f : -0.0f, W > 0? 0.0f : -0.0f));
}
 
Vec4 Vec4::Normalized() const
{
#if defined(JPH_USE_SSE4_1)
    return _mm_div_ps(mValue, _mm_sqrt_ps(_mm_dp_ps(mValue, mValue, 0xff)));
#elif defined(JPH_USE_NEON)
    float32x4_t mul = vmulq_f32(mValue, mValue);
    float32x4_t sum = vdupq_n_f32(vaddvq_f32(mul));
    return vdivq_f32(mValue, vsqrtq_f32(sum));
#else
    return *this / Length();
#endif
}
 
void Vec4::StoreFloat4(Float4 *outV) const
{
#if defined(JPH_USE_SSE)
    _mm_storeu_ps(&outV->x, mValue);
#elif defined(JPH_USE_NEON)
    vst1q_f32(&outV->x, mValue);
#else
    for (int i = 0; i < 4; ++i)
        (&outV->x)[i] = mF32[i];
#endif
}
 
UVec4 Vec4::ToInt() const
{
#if defined(JPH_USE_SSE)
    return _mm_cvttps_epi32(mValue);
#elif defined(JPH_USE_NEON)
    return vcvtq_u32_f32(mValue);
#else
    return UVec4(uint32(mF32[0]), uint32(mF32[1]), uint32(mF32[2]), uint32(mF32[3]));
#endif
}
 
UVec4 Vec4::ReinterpretAsInt() const
{
#if defined(JPH_USE_SSE)
    return UVec4(_mm_castps_si128(mValue));
#elif defined(JPH_USE_NEON)
    return vreinterpretq_u32_f32(mValue);
#else
    return *reinterpret_cast<const UVec4 *>(this);
#endif
}
 
int Vec4::GetSignBits() const
{
#if defined(JPH_USE_SSE)
    return _mm_movemask_ps(mValue);
#elif defined(JPH_USE_NEON)
    int32x4_t shift = JPH_NEON_INT32x4(0, 1, 2, 3);
    return vaddvq_u32(vshlq_u32(vshrq_n_u32(vreinterpretq_u32_f32(mValue), 31), shift));
#else
    return (std::signbit(mF32[0])? 1 : 0) | (std::signbit(mF32[1])? 2 : 0) | (std::signbit(mF32[2])? 4 : 0) | (std::signbit(mF32[3])? 8 : 0);
#endif
}
 
float Vec4::ReduceMin() const
{
    Vec4 v = sMin(mValue, Swizzle<SWIZZLE_Y, SWIZZLE_UNUSED, SWIZZLE_W, SWIZZLE_UNUSED>());
    v = sMin(v, v.Swizzle<SWIZZLE_Z, SWIZZLE_UNUSED, SWIZZLE_UNUSED, SWIZZLE_UNUSED>());
    return v.GetX();
}
 
float Vec4::ReduceMax() const
{
    Vec4 v = sMax(mValue, Swizzle<SWIZZLE_Y, SWIZZLE_UNUSED, SWIZZLE_W, SWIZZLE_UNUSED>());
    v = sMax(v, v.Swizzle<SWIZZLE_Z, SWIZZLE_UNUSED, SWIZZLE_UNUSED, SWIZZLE_UNUSED>());
    return v.GetX();
}
 
void Vec4::SinCos(Vec4 &outSin, Vec4 &outCos) const
{
    // Implementation based on sinf.c from the cephes library, combines sinf and cosf in a single function, changes octants to quadrants and vectorizes it
    // Original implementation by Stephen L. Moshier (See: http://www.moshier.net/)
 
    // Make argument positive and remember sign for sin only since cos is symmetric around x (highest bit of a float is the sign bit)
    UVec4 sin_sign = UVec4::sAnd(ReinterpretAsInt(), UVec4::sReplicate(0x80000000U));
    Vec4 x = Vec4::sXor(*this, sin_sign.ReinterpretAsFloat());
 
    // x / (PI / 2) rounded to nearest int gives us the quadrant closest to x
    UVec4 quadrant = (0.6366197723675814f * x + Vec4::sReplicate(0.5f)).ToInt();
 
    // Make x relative to the closest quadrant.
    // This does x = x - quadrant * PI / 2 using a two step Cody-Waite argument reduction.
    // This improves the accuracy of the result by avoiding loss of significant bits in the subtraction.
    // We start with x = x - quadrant * PI / 2, PI / 2 in hexadecimal notation is 0x3fc90fdb, we remove the lowest 16 bits to
    // get 0x3fc90000 (= 1.5703125) this means we can now multiply with a number of up to 2^16 without losing any bits.
    // This leaves us with: x = (x - quadrant * 1.5703125) - quadrant * (PI / 2 - 1.5703125).
    // PI / 2 - 1.5703125 in hexadecimal is 0x39fdaa22, stripping the lowest 12 bits we get 0x39fda000 (= 0.0004837512969970703125)
    // This leaves uw with: x = ((x - quadrant * 1.5703125) - quadrant * 0.0004837512969970703125) - quadrant * (PI / 2 - 1.5703125 - 0.0004837512969970703125)
    // See: https://stackoverflow.com/questions/42455143/sine-cosine-modular-extended-precision-arithmetic
    // After this we have x in the range [-PI / 4, PI / 4].
    Vec4 float_quadrant = quadrant.ToFloat();
    x = ((x - float_quadrant * 1.5703125f) - float_quadrant * 0.0004837512969970703125f) - float_quadrant * 7.549789948768648e-8f;
 
    // Calculate x2 = x^2
    Vec4 x2 = x * x;
 
    // Taylor expansion:
    // Cos(x) = 1 - x^2/2! + x^4/4! - x^6/6! + x^8/8! + ... = (((x2/8!- 1/6!) * x2 + 1/4!) * x2 - 1/2!) * x2 + 1
    Vec4 taylor_cos = ((2.443315711809948e-5f * x2 - Vec4::sReplicate(1.388731625493765e-3f)) * x2 + Vec4::sReplicate(4.166664568298827e-2f)) * x2 * x2 - 0.5f * x2 + Vec4::sOne();
    // Sin(x) = x - x^3/3! + x^5/5! - x^7/7! + ... = ((-x2/7! + 1/5!) * x2 - 1/3!) * x2 * x + x
    Vec4 taylor_sin = ((-1.9515295891e-4f * x2 + Vec4::sReplicate(8.3321608736e-3f)) * x2 - Vec4::sReplicate(1.6666654611e-1f)) * x2 * x + x;
 
    // The lowest 2 bits of quadrant indicate the quadrant that we are in.
    // Let x be the original input value and x' our value that has been mapped to the range [-PI / 4, PI / 4].
    // since cos(x) = sin(x - PI / 2) and since we want to use the Taylor expansion as close as possible to 0,
    // we can alternate between using the Taylor expansion for sin and cos according to the following table:
    //
    // quadrant  sin(x)      cos(x)
    // XXX00b    sin(x')     cos(x')
    // XXX01b    cos(x')    -sin(x')
    // XXX10b   -sin(x')    -cos(x')
    // XXX11b   -cos(x')     sin(x')
    //
    // So: sin_sign = bit2, cos_sign = bit1 ^ bit2, bit1 determines if we use sin or cos Taylor expansion
    UVec4 bit1 = quadrant.LogicalShiftLeft<31>();
    UVec4 bit2 = UVec4::sAnd(quadrant.LogicalShiftLeft<30>(), UVec4::sReplicate(0x80000000U));
 
    // Select which one of the results is sin and which one is cos
    Vec4 s = Vec4::sSelect(taylor_sin, taylor_cos, bit1);
    Vec4 c = Vec4::sSelect(taylor_cos, taylor_sin, bit1);
 
    // Update the signs
    sin_sign = UVec4::sXor(sin_sign, bit2);
    UVec4 cos_sign = UVec4::sXor(bit1, bit2);
 
    // Correct the signs
    outSin = Vec4::sXor(s, sin_sign.ReinterpretAsFloat());
    outCos = Vec4::sXor(c, cos_sign.ReinterpretAsFloat());
}
 
Vec4 Vec4::Tan() const
{
    // Implementation based on tanf.c from the cephes library, see Vec4::SinCos for further details
    // Original implementation by Stephen L. Moshier (See: http://www.moshier.net/)
 
    // Make argument positive
    UVec4 tan_sign = UVec4::sAnd(ReinterpretAsInt(), UVec4::sReplicate(0x80000000U));
    Vec4 x = Vec4::sXor(*this, tan_sign.ReinterpretAsFloat());
 
    // x / (PI / 2) rounded to nearest int gives us the quadrant closest to x
    UVec4 quadrant = (0.6366197723675814f * x + Vec4::sReplicate(0.5f)).ToInt();
 
    // Remap x to range [-PI / 4, PI / 4], see Vec4::SinCos
    Vec4 float_quadrant = quadrant.ToFloat();
    x = ((x - float_quadrant * 1.5703125f) - float_quadrant * 0.0004837512969970703125f) - float_quadrant * 7.549789948768648e-8f;
 
    // Calculate x2 = x^2
    Vec4 x2 = x * x;
 
    // Roughly equivalent to the Taylor expansion:
    // Tan(x) = x + x^3/3 + 2*x^5/15 + 17*x^7/315 + 62*x^9/2835 + ...
    Vec4 tan =
        (((((9.38540185543e-3f * x2 + Vec4::sReplicate(3.11992232697e-3f)) * x2 + Vec4::sReplicate(2.44301354525e-2f)) * x2
        + Vec4::sReplicate(5.34112807005e-2f)) * x2 + Vec4::sReplicate(1.33387994085e-1f)) * x2 + Vec4::sReplicate(3.33331568548e-1f)) * x2 * x + x;
 
    // For the 2nd and 4th quadrant we need to invert the value
    UVec4 bit1 = quadrant.LogicalShiftLeft<31>();
    tan = Vec4::sSelect(tan, Vec4::sReplicate(-1.0f) / (tan JPH_IF_FLOATING_POINT_EXCEPTIONS_ENABLED(+ Vec4::sReplicate(FLT_MIN))), bit1); // Add small epsilon to prevent div by zero, works because tan is always positive
 
    // Put the sign back
    return Vec4::sXor(tan, tan_sign.ReinterpretAsFloat());
}
 
Vec4 Vec4::ASin() const
{
    // Implementation based on asinf.c from the cephes library
    // Original implementation by Stephen L. Moshier (See: http://www.moshier.net/)
 
    // Make argument positive
    UVec4 asin_sign = UVec4::sAnd(ReinterpretAsInt(), UVec4::sReplicate(0x80000000U));
    Vec4 a = Vec4::sXor(*this, asin_sign.ReinterpretAsFloat());
 
    // ASin is not defined outside the range [-1, 1] but it often happens that a value is slightly above 1 so we just clamp here
    a = Vec4::sMin(a, Vec4::sOne());
 
    // When |x| <= 0.5 we use the asin approximation as is
    Vec4 z1 = a * a;
    Vec4 x1 = a;
 
    // When |x| > 0.5 we use the identity asin(x) = PI / 2 - 2 * asin(sqrt((1 - x) / 2))
    Vec4 z2 = 0.5f * (Vec4::sOne() - a);
    Vec4 x2 = z2.Sqrt();
 
    // Select which of the two situations we have
    UVec4 greater = Vec4::sGreater(a, Vec4::sReplicate(0.5f));
    Vec4 z = Vec4::sSelect(z1, z2, greater);
    Vec4 x = Vec4::sSelect(x1, x2, greater);
 
    // Polynomial approximation of asin
    z = ((((4.2163199048e-2f * z + Vec4::sReplicate(2.4181311049e-2f)) * z + Vec4::sReplicate(4.5470025998e-2f)) * z + Vec4::sReplicate(7.4953002686e-2f)) * z + Vec4::sReplicate(1.6666752422e-1f)) * z * x + x;
 
    // If |x| > 0.5 we need to apply the remainder of the identity above
    z = Vec4::sSelect(z, Vec4::sReplicate(0.5f * JPH_PI) - (z + z), greater);
 
    // Put the sign back
    return Vec4::sXor(z, asin_sign.ReinterpretAsFloat());
}
 
Vec4 Vec4::ACos() const
{
    // Not the most accurate, but simple
    return Vec4::sReplicate(0.5f * JPH_PI) - ASin();
}
 
Vec4 Vec4::ATan() const
{
    // Implementation based on atanf.c from the cephes library
    // Original implementation by Stephen L. Moshier (See: http://www.moshier.net/)
 
    // Make argument positive
    UVec4 atan_sign = UVec4::sAnd(ReinterpretAsInt(), UVec4::sReplicate(0x80000000U));
    Vec4 x = Vec4::sXor(*this, atan_sign.ReinterpretAsFloat());
    Vec4 y = Vec4::sZero();
 
    // If x > Tan(PI / 8)
    UVec4 greater1 = Vec4::sGreater(x, Vec4::sReplicate(0.4142135623730950f));
    Vec4 x1 = (x - Vec4::sOne()) / (x + Vec4::sOne());
 
    // If x > Tan(3 * PI / 8)
    UVec4 greater2 = Vec4::sGreater(x, Vec4::sReplicate(2.414213562373095f));
    Vec4 x2 = Vec4::sReplicate(-1.0f) / (x JPH_IF_FLOATING_POINT_EXCEPTIONS_ENABLED(+ Vec4::sReplicate(FLT_MIN))); // Add small epsilon to prevent div by zero, works because x is always positive
 
    // Apply first if
    x = Vec4::sSelect(x, x1, greater1);
    y = Vec4::sSelect(y, Vec4::sReplicate(0.25f * JPH_PI), greater1);
 
    // Apply second if
    x = Vec4::sSelect(x, x2, greater2);
    y = Vec4::sSelect(y, Vec4::sReplicate(0.5f * JPH_PI), greater2);
 
    // Polynomial approximation
    Vec4 z = x * x;
    y += (((8.05374449538e-2f * z - Vec4::sReplicate(1.38776856032e-1f)) * z + Vec4::sReplicate(1.99777106478e-1f)) * z - Vec4::sReplicate(3.33329491539e-1f)) * z * x + x;
 
    // Put the sign back
    return Vec4::sXor(y, atan_sign.ReinterpretAsFloat());
}
 
Vec4 Vec4::sATan2(Vec4Arg inY, Vec4Arg inX)
{
    UVec4 sign_mask = UVec4::sReplicate(0x80000000U);
 
    // Determine absolute value and sign of y
    UVec4 y_sign = UVec4::sAnd(inY.ReinterpretAsInt(), sign_mask);
    Vec4 y_abs = Vec4::sXor(inY, y_sign.ReinterpretAsFloat());
 
    // Determine absolute value and sign of x
    UVec4 x_sign = UVec4::sAnd(inX.ReinterpretAsInt(), sign_mask);
    Vec4 x_abs = Vec4::sXor(inX, x_sign.ReinterpretAsFloat());
 
    // Always divide smallest / largest to avoid dividing by zero
    UVec4 x_is_numerator = Vec4::sLess(x_abs, y_abs);
    Vec4 numerator = Vec4::sSelect(y_abs, x_abs, x_is_numerator);
    Vec4 denominator = Vec4::sSelect(x_abs, y_abs, x_is_numerator);
    Vec4 atan = (numerator / denominator).ATan();
 
    // If we calculated x / y instead of y / x the result is PI / 2 - result (note that this is true because we know the result is positive because the input was positive)
    atan = Vec4::sSelect(atan, Vec4::sReplicate(0.5f * JPH_PI) - atan, x_is_numerator);
 
    // Now we need to map to the correct quadrant
    // x_sign   y_sign  result
    // +1       +1      atan
    // -1       +1      -atan + PI
    // -1       -1      atan - PI
    // +1       -1      -atan
    // This can be written as: x_sign * y_sign * (atan - (x_sign < 0? PI : 0))
    atan -= Vec4::sAnd(x_sign.ArithmeticShiftRight<31>().ReinterpretAsFloat(), Vec4::sReplicate(JPH_PI));
    atan = Vec4::sXor(atan, UVec4::sXor(x_sign, y_sign).ReinterpretAsFloat());
    return atan;
}
 
uint32 Vec4::CompressUnitVector() const
{
    constexpr float cOneOverSqrt2 = 0.70710678f;
    constexpr uint cNumBits = 9;
    constexpr uint cMask = (1 << cNumBits) - 1;
 
    // Store sign bit
    Vec4 v = *this;
    uint32 max_element = v.Abs().GetHighestComponentIndex();
    uint32 value = 0;
    if (v[max_element] < 0.0f)
    {
        value = 0x80000000u;
        v = -v;
    }
 
    // Store highest component
    value |= max_element << 29;
 
    // Store the other three components in a compressed format
    UVec4 compressed = Vec4::sClamp((v + Vec4::sReplicate(cOneOverSqrt2)) * (float(cMask) / (2.0f * cOneOverSqrt2)) + Vec4::sReplicate(0.5f), Vec4::sZero(), Vec4::sReplicate(cMask)).ToInt();
    switch (max_element)
    {
    case 0:
        compressed = compressed.Swizzle<SWIZZLE_Y, SWIZZLE_Z, SWIZZLE_W, SWIZZLE_UNUSED>();
        break;
 
    case 1:
        compressed = compressed.Swizzle<SWIZZLE_X, SWIZZLE_Z, SWIZZLE_W, SWIZZLE_UNUSED>();
        break;
 
    case 2:
        compressed = compressed.Swizzle<SWIZZLE_X, SWIZZLE_Y, SWIZZLE_W, SWIZZLE_UNUSED>();
        break;
    }
 
    value |= compressed.GetX();
    value |= compressed.GetY() << cNumBits;
    value |= compressed.GetZ() << 2 * cNumBits;
    return value;
}
 
Vec4 Vec4::sDecompressUnitVector(uint32 inValue)
{
    constexpr float cOneOverSqrt2 = 0.70710678f;
    constexpr uint cNumBits = 9;
    constexpr uint cMask = (1u << cNumBits) - 1;
 
    // Restore three components
    Vec4 v = Vec4(UVec4(inValue & cMask, (inValue >> cNumBits) & cMask, (inValue >> (2 * cNumBits)) & cMask, 0).ToFloat()) * (2.0f * cOneOverSqrt2 / float(cMask)) - Vec4(cOneOverSqrt2, cOneOverSqrt2, cOneOverSqrt2, 0.0f);
    JPH_ASSERT(v.GetW() == 0.0f);
 
    // Restore the highest component
    v.SetW(sqrt(max(1.0f - v.LengthSq(), 0.0f)));
 
    // Extract sign
    if ((inValue & 0x80000000u) != 0)
        v = -v;
 
    // Swizzle the components in place
    switch ((inValue >> 29) & 3)
    {
    case 0:
        v = v.Swizzle<SWIZZLE_W, SWIZZLE_X, SWIZZLE_Y, SWIZZLE_Z>();
        break;
 
    case 1:
        v = v.Swizzle<SWIZZLE_X, SWIZZLE_W, SWIZZLE_Y, SWIZZLE_Z>();
        break;
 
    case 2:
        v = v.Swizzle<SWIZZLE_X, SWIZZLE_Y, SWIZZLE_W, SWIZZLE_Z>();
        break;
    }
 
    return v;
}
 
JPH_NAMESPACE_END