thoehne/PhanesEngine
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

SIMD improvements.

Browse Source
This commit is contained in:
THoehne 2024-08-28 14:26:40 +02:00
parent 245f1d87c5
commit c58e05373f
1 changed files with 150 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

150
Engine/Source/Runtime/Core/public/Math/SIMD/PhanesVectorMathSSE.hpp
View File

@ -102,6 +102,16 @@ namespace Phanes::Core::Math::SIMD
{
return _mm_cmpeq_pd(v1, v2);
}
/// <summary>
/// Sets the last component of the register to zero. <br>
/// The last component could hold unexpected values.
/// </summary>
/// <param name="v1"></param>
void vec3_fix(Phanes::Core::Types::Vec4f32Reg v1)
{
v1 = _mm_blend_ps(v1, _mm_setzero_ps(), 0x1);
}
}
@ -175,6 +185,11 @@ namespace Phanes::Core::Math::Detail
{
r.comp = _mm_sub_ps(v1.comp, _mm_set_ps1(s));
}
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& r, float s, const Phanes::Core::Math::TVector4<float, true>& v1)
{
r.comp = _mm_sub_ps(_mm_set_ps1(s), v1.comp);
}
};
template<>
@ -203,6 +218,11 @@ namespace Phanes::Core::Math::Detail
{
r.comp = _mm_div_ps(v1.comp, _mm_set_ps1(s));
}
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& r, float s, const Phanes::Core::Math::TVector4<float, true>& v1)
{
r.comp = _mm_div_ps(_mm_set_ps1(s), v1.comp);
}
};
template<>
@ -223,6 +243,62 @@ namespace Phanes::Core::Math::Detail
}
};
template<>
struct compute_vec4_mag<float, true>
{
static FORCEINLINE float map(const Phanes::Core::Math::TVector4<float, true>& v1)
{
__m128 tmp = _mm_mul_ps(v1.data, v1.data);
return sqrt(tmp.m128_f32[0] + tmp.m128_f32[1] + tmp.m128_f32[2] + tmp.m128_f32[3]);
}
};
template<>
struct compute_vec4_dotp<float, true>
{
static FORCEINLINE float map(const Phanes::Core::Math::TVector4<float, true>& v1, const Phanes::Core::Math::TVector4<float, true>& v2)
{
return SIMD::vec4_dot_cvtf32(v1.data, v2.data);
}
};
template<>
struct compute_vec4_set<float, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& v1, float x, float y, float z, float w)
{
v1.data = _mm_setr_ps(x, y, z, w);
}
};
template<>
struct compute_vec4_max<float, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& r, const Phanes::Core::Math::TVector4<float, true>& v1, const Phanes::Core::Math::TVector4<float, true>& v2)
{
r.data = _mm_max_ps(v1.data, v2.data);
}
};
template<>
struct compute_vec4_min<float, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& r, const Phanes::Core::Math::TVector4<float, true>& v1, const Phanes::Core::Math::TVector4<float, true>& v2)
{
r.data = _mm_min_ps(v1.data, v2.data);
}
};
template<>
struct compute_vec4_pdiv<float, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector4<float, true>& r, const Phanes::Core::Math::TVector4<float, true>& v1)
{
__m128 tmp = _mm_div_ps(v1.data, _mm_set_ps1(v1.w));
r.data = _mm_blend_ps(tmp, _mm_setzero_ps(), 0x1);
}
};
// ============ //
// TVector3 //
@ -260,6 +336,14 @@ namespace Phanes::Core::Math::Detail
}
};
template<>
struct compute_vec3_set<float, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector3<float, true>& v1, float x, float y, float z)
{
v1.data = _mm_setr_ps(x, y, z, 0.0f);
}
};
template<> struct compute_vec3_add<float, true> : public compute_vec4_add<float, true> {};
template<> struct compute_vec3_sub<float, true> : public compute_vec4_sub<float, true> {};
@ -267,6 +351,10 @@ namespace Phanes::Core::Math::Detail
template<> struct compute_vec3_div<float, true> : public compute_vec4_div<float, true> {};
template<> struct compute_vec3_inc<float, true> : public compute_vec4_inc<float, true> {};
template<> struct compute_vec3_dec<float, true> : public compute_vec4_dec<float, true> {};
template<> struct compute_vec3_mag<float, true> : public compute_vec4_mag<float, true> {};
template<> struct compute_vec3_dotp<float, true> : public compute_vec4_dotp<float, true> {};
template<> struct compute_vec3_max<float, true> : public compute_vec4_max<float, true> {};
template<> struct compute_vec3_min<float, true> : public compute_vec4_min<float, true> {};
template<>
struct compute_vec3_cross_p<float, true>
@ -336,6 +424,11 @@ namespace Phanes::Core::Math::Detail
{
r.comp = _mm_sub_pd(v1.comp, _mm_set1_pd(s));
}
static FORCEINLINE void map(Phanes::Core::Math::TVector2<double, true>& r, double s, const Phanes::Core::Math::TVector2<double, true>& v1)
{
r.comp = _mm_sub_pd(_mm_set1_pd(s), v1.comp);
}
};
template<>
@ -364,6 +457,11 @@ namespace Phanes::Core::Math::Detail
{
r.comp = _mm_div_pd(v1.comp, _mm_set1_pd(s));
}
static FORCEINLINE void map(Phanes::Core::Math::TVector2<double, true>& r, double s, const Phanes::Core::Math::TVector2<double, true>& v1)
{
r.comp = _mm_div_pd(_mm_set1_pd(s), v1.comp);
}
};
template<>
@ -384,6 +482,58 @@ namespace Phanes::Core::Math::Detail
}
};
// Magnitude
template<RealType T>
struct compute_vec2_mag<T, true>
{
static FORCEINLINE double map(const Phanes::Core::Math::TVector2<double, true>& v1)
{
__m128d tmp = _mm_mul_pd(v1.data, v1.data);
return sqrt(tmp.m128d_f64[0] + tmp.m128d_f64[1]);
}
};
// Dot product
template<>
struct compute_vec2_dotp<double, true>
{
static FORCEINLINE double map(const Phanes::Core::Math::TVector2<double, true>& v1)
{
__m128d tmp = _mm_mul_pd(v1.data, v1.data);
return tmp.m128d_f64[0] + tmp.m128d_f64[1];
}
};
// Max
template<>
struct compute_vec2_max<double, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector2<double, true>& r, const Phanes::Core::Math::TVector2<double, true>& v1, const Phanes::Core::Math::TVector2<double, true>& v2)
{
r.data = _mm_max_pd(v1.data, v2.data);
}
};
// Min
template<>
struct compute_vec2_min<double, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector2<double, true>& r, const Phanes::Core::Math::TVector2<double, true>& v1, const Phanes::Core::Math::TVector2<double, true>& v2)
{
r.data = _mm_min_pd(v1.data, v2.data);
}
};
// Set
template<>
struct compute_vec2_set<double, true>
{
static FORCEINLINE void map(Phanes::Core::Math::TVector2<double, true>& v1, double x, double y)
{
v1.data = _mm_setr_pd(x, y);
}
};
// =============== //
// TIntVector4 //