tge/engine/math/mMathFn.h
2025-02-17 23:17:30 -06:00

455 lines
11 KiB
C++
Executable File

//-----------------------------------------------------------------------------
// Torque Game Engine
// Copyright (C) GarageGames.com, Inc.
//-----------------------------------------------------------------------------
#ifndef _MMATHFN_H_
#define _MMATHFN_H_
#ifndef _PLATFORM_H_
#include "platform/platform.h"
#endif
#ifndef _MCONSTANTS_H_
#include "math/mConstants.h"
#endif
#include <math.h>
// Remove a couple of annoying macros, if they are present (In VC 6, they are.)
#ifdef min
#undef min
#endif
#ifdef max
#undef max
#endif
class MatrixF;
class PlaneF;
extern void MathConsoleInit();
//--------------------------------------
// Installable Library Prototypes
extern S32 (*m_mulDivS32)(S32 a, S32 b, S32 c);
extern U32 (*m_mulDivU32)(S32 a, S32 b, U32 c);
extern F32 (*m_catmullrom)(F32 t, F32 p0, F32 p1, F32 p2, F32 p3);
extern void (*m_point2F_normalize)(F32 *p);
extern void (*m_point2F_normalize_f)(F32 *p, F32 len);
extern void (*m_point2D_normalize)(F64 *p);
extern void (*m_point2D_normalize_f)(F64 *p, F64 len);
extern void (*m_point3F_normalize)(F32 *p);
extern void (*m_point3F_normalize_f)(F32 *p, F32 len);
extern void (*m_point3F_interpolate)(const F32 *from, const F32 *to, F32 factor, F32 *result);
extern void (*m_point3D_normalize)(F64 *p);
extern void (*m_point3D_normalize_f)(F64 *p, F64 len);
extern void (*m_point3D_interpolate)(const F64 *from, const F64 *to, F64 factor, F64 *result);
extern void (*m_point3F_bulk_dot)(const F32* refVector,
const F32* dotPoints,
const U32 numPoints,
const U32 pointStride,
F32* output);
extern void (*m_point3F_bulk_dot_indexed)(const F32* refVector,
const F32* dotPoints,
const U32 numPoints,
const U32 pointStride,
const U32* pointIndices,
F32* output);
extern void (*m_quatF_set_matF)( F32 x, F32 y, F32 z, F32 w, F32* m );
extern void (*m_matF_set_euler)(const F32 *e, F32 *result);
extern void (*m_matF_set_euler_point)(const F32 *e, const F32 *p, F32 *result);
extern void (*m_matF_identity)(F32 *m);
extern void (*m_matF_inverse)(F32 *m);
extern void (*m_matF_affineInverse)(F32 *m);
extern void (*m_matF_transpose)(F32 *m);
extern void (*m_matF_scale)(F32 *m,const F32* p);
extern void (*m_matF_normalize)(F32 *m);
extern F32 (*m_matF_determinant)(const F32 *m);
extern void (*m_matF_x_matF)(const F32 *a, const F32 *b, F32 *mresult);
// extern void (*m_matF_x_point3F)(const F32 *m, const F32 *p, F32 *presult);
// extern void (*m_matF_x_vectorF)(const F32 *m, const F32 *v, F32 *vresult);
extern void (*m_matF_x_point4F)(const F32 *m, const F32 *p, F32 *presult);
extern void (*m_matF_x_scale_x_planeF)(const F32 *m, const F32* s, const F32 *p, F32 *presult);
extern void (*m_matF_x_box3F)(const F32 *m, F32 *min, F32 *max);
// Note that x must point to at least 4 values for quartics, and 3 for cubics
extern U32 (*mSolveQuadratic)(F32 a, F32 b, F32 c, F32* x);
extern U32 (*mSolveCubic)(F32 a, F32 b, F32 c, F32 d, F32* x);
extern U32 (*mSolveQuartic)(F32 a, F32 b, F32 c, F32 d, F32 e, F32* x);
extern S32 mRandI(S32 i1, S32 i2); // random # from i1 to i2 inclusive
extern F32 mRandF(F32 f1, F32 f2); // random # from f1 to f2 inclusive
inline void m_matF_x_point3F(const F32 *m, const F32 *p, F32 *presult)
{
AssertFatal(p != presult, "Error, aliasing matrix mul pointers not allowed here!");
#if defined(TORQUE_SUPPORTS_GCC_INLINE_X86_ASM)
// inline assembly version because gcc's math optimization isn't as good
// JMQ: the profiler shows that with g++ 2.96, the difference
// between the asm and optimized c versions is minimal.
int u0, u1, u2;
__asm__ __volatile__ (
"flds 0x8(%%eax)\n"
"fmuls 0x8(%%ecx)\n"
"flds 0x4(%%eax)\n"
"fmuls 0x4(%%ecx)\n"
"faddp %%st,%%st(1)\n"
"flds (%%eax)\n"
"fmuls (%%ecx)\n"
"faddp %%st,%%st(1)\n"
"fadds 0xc(%%eax)\n"
"fstps (%%edx)\n"
"flds 0x18(%%eax)\n"
"fmuls 0x8(%%ecx)\n"
"flds 0x10(%%eax)\n"
"fmuls (%%ecx)\n"
"faddp %%st,%%st(1)\n"
"flds 0x14(%%eax)\n"
"fmuls 0x4(%%ecx)\n"
"faddp %%st,%%st(1)\n"
"fadds 0x1c(%%eax)\n"
"fstps 0x4(%%edx)\n"
"flds 0x28(%%eax)\n"
"fmuls 0x8(%%ecx)\n"
"flds 0x20(%%eax)\n"
"fmuls (%%ecx)\n"
"faddp %%st,%%st(1)\n"
"flds 0x24(%%eax)\n"
"fmuls 0x4(%%ecx)\n"
"faddp %%st,%%st(1)\n"
"fadds 0x2c(%%eax)\n"
"fstps 0x8(%%edx)\n"
: "=&a" (u0), "=&c" (u1), "=&d" (u2)
: "0" (m), "1" (p), "2" (presult)
: "memory" );
#else
presult[0] = m[0]*p[0] + m[1]*p[1] + m[2]*p[2] + m[3];
presult[1] = m[4]*p[0] + m[5]*p[1] + m[6]*p[2] + m[7];
presult[2] = m[8]*p[0] + m[9]*p[1] + m[10]*p[2] + m[11];
#endif
}
//--------------------------------------
inline void m_matF_x_vectorF(const F32 *m, const F32 *v, F32 *vresult)
{
AssertFatal(v != vresult, "Error, aliasing matrix mul pointers not allowed here!");
#if defined(TORQUE_SUPPORTS_GCC_INLINE_X86_ASM)
// inline assembly version because gcc's math optimization isn't as good
// JMQ: the profiler shows that with g++ 2.96, the difference
// between the asm and optimized c versions is minimal.
int u0, u1, u2;
__asm__ __volatile__ (
"flds 0x8(%%ecx)\n"
"fmuls 0x8(%%eax)\n"
"flds 0x4(%%ecx)\n"
"fmuls 0x4(%%eax)\n"
"faddp %%st,%%st(1)\n"
"flds (%%ecx)\n"
"fmuls (%%eax)\n"
"faddp %%st,%%st(1)\n"
"fstps (%%edx)\n"
"flds 0x18(%%ecx)\n"
"fmuls 0x8(%%eax)\n"
"flds 0x10(%%ecx)\n"
"fmuls (%%eax)\n"
"faddp %%st,%%st(1)\n"
"flds 0x14(%%ecx)\n"
"fmuls 0x4(%%eax)\n"
"faddp %%st,%%st(1)\n"
"fstps 0x4(%%edx)\n"
"flds 0x28(%%ecx)\n"
"fmuls 0x8(%%eax)\n"
"flds 0x20(%%ecx)\n"
"fmuls (%%eax)\n"
"faddp %%st,%%st(1)\n"
"flds 0x24(%%ecx)\n"
"fmuls 0x4(%%eax)\n"
"faddp %%st,%%st(1)\n"
"fstps 0x8(%%edx)\n"
: "=&c" (u0), "=&a" (u1), "=&d" (u2)
: "0" (m), "1" (v), "2" (vresult)
: "memory" );
#else
vresult[0] = m[0]*v[0] + m[1]*v[1] + m[2]*v[2];
vresult[1] = m[4]*v[0] + m[5]*v[1] + m[6]*v[2];
vresult[2] = m[8]*v[0] + m[9]*v[1] + m[10]*v[2];
#endif
}
//--------------------------------------
// Inlines
inline F32 mFloor(const F32 val)
{
return (F32) floor(val);
}
inline F32 mCeil(const F32 val)
{
return (F32) ceil(val);
}
inline F32 mFabs(const F32 val)
{
return (F32) fabs(val);
}
inline F32 mFmod(const F32 val, const F32 mod)
{
return (F32) fmod(val, mod);
}
inline S32 mAbs(const S32 val)
{
// Kinda lame, and disallows intrinsic inlining by the compiler. Maybe fix?
// DMM
if (val < 0)
return -val;
return val;
}
inline S32 mClamp(S32 val, S32 low, S32 high)
{
return getMax(getMin(val, high), low);
}
inline F32 mClampF(F32 val, F32 low, F32 high)
{
return (F32) getMax(getMin(val, high), low);
}
inline S32 mMulDiv(S32 a, S32 b, S32 c)
{
return m_mulDivS32(a, b, c);
}
inline U32 mMulDiv(S32 a, S32 b, U32 c)
{
return m_mulDivU32(a, b, c);
}
inline F32 mSin(const F32 angle)
{
return (F32) sin(angle);
}
inline F32 mCos(const F32 angle)
{
return (F32) cos(angle);
}
inline F32 mTan(const F32 angle)
{
return (F32) tan(angle);
}
inline F32 mAsin(const F32 val)
{
return (F32) asin(val);
}
inline F32 mAcos(const F32 val)
{
return (F32) acos(val);
}
inline F32 mAtan(const F32 x, const F32 y)
{
return (F32) atan2(x, y);
}
inline void mSinCos(const F32 angle, F32 &s, F32 &c)
{
s = mSin(angle);
c = mCos(angle);
}
inline F32 mTanh(const F32 angle)
{
return (F32) tanh(angle);
}
inline F32 mSqrt(const F32 val)
{
return (F32) sqrt(val);
}
inline F32 mPow(const F32 x, const F32 y)
{
return (F32) pow(x, y);
}
inline F32 mLog(const F32 val)
{
return (F32) log(val);
}
inline F64 mSin(const F64 angle)
{
return (F64) sin(angle);
}
inline F64 mCos(const F64 angle)
{
return (F64) cos(angle);
}
inline F64 mTan(const F64 angle)
{
return (F64) tan(angle);
}
inline F64 mAsin(const F64 val)
{
return (F64) asin(val);
}
inline F64 mAcos(const F64 val)
{
return (F64) acos(val);
}
inline F64 mAtan(const F64 x, const F64 y)
{
return (F64) atan2(x, y);
}
inline void mSinCos(const F64 angle, F64 &sin, F64 &cos)
{
sin = mSin(angle);
cos = mCos(angle);
}
inline F64 mTanh(const F64 angle)
{
return (F64) tanh(angle);
}
inline F64 mPow(const F64 x, const F64 y)
{
return (F64) pow(x, y);
}
inline F64 mLog(const F64 val)
{
return (F64) log(val);
}
inline F32 mCatmullrom(F32 t, F32 p0, F32 p1, F32 p2, F32 p3)
{
return m_catmullrom(t, p0, p1, p2, p3);
}
inline F64 mFabsD(const F64 val)
{
return (F64) fabs(val);
}
inline F64 mFmodD(const F64 val, const F64 mod)
{
return (F64) fmod(val, mod);
}
inline F64 mSqrtD(const F64 val)
{
return (F64) sqrt(val);
}
inline F64 mFloorD(const F64 val)
{
return (F64) floor(val);
}
inline F64 mCeilD(const F64 val)
{
return (F64) ceil(val);
}
//--------------------------------------
#ifndef _MPOINT_H_
#include "math/mPoint.h"
#endif
inline F32 mDot(const Point3F &p1, const Point3F &p2)
{
return (p1.x*p2.x + p1.y*p2.y + p1.z*p2.z);
}
inline void mCross(const Point3F &a, const Point3F &b, Point3F *res)
{
res->x = (a.y * b.z) - (a.z * b.y);
res->y = (a.z * b.x) - (a.x * b.z);
res->z = (a.x * b.y) - (a.y * b.x);
}
inline F64 mDot(const Point3D &p1, const Point3D &p2)
{
return (p1.x*p2.x + p1.y*p2.y + p1.z*p2.z);
}
inline void mCross(const Point3D &a, const Point3D &b, Point3D *res)
{
res->x = (a.y * b.z) - (a.z * b.y);
res->y = (a.z * b.x) - (a.x * b.z);
res->z = (a.x * b.y) - (a.y * b.x);
}
inline Point3F mCross(const Point3F &a, const Point3F &b)
{
Point3F ret;
mCross(a,b,&ret);
return ret;
}
inline void mCross(const F32* a, const F32* b, F32 *res)
{
res[0] = (a[1] * b[2]) - (a[2] * b[1]);
res[1] = (a[2] * b[0]) - (a[0] * b[2]);
res[2] = (a[0] * b[1]) - (a[1] * b[0]);
}
inline void mCross(const F64* a, const F64* b, F64* res)
{
res[0] = (a[1] * b[2]) - (a[2] * b[1]);
res[1] = (a[2] * b[0]) - (a[0] * b[2]);
res[2] = (a[0] * b[1]) - (a[1] * b[0]);
}
void mTransformPlane(const MatrixF& mat, const Point3F& scale, const PlaneF& plane, PlaneF* result);
//--------------------------------------
inline F32 mDegToRad(F32 d)
{
return F32((d * M_PI) / F32(180));
}
inline F32 mRadToDeg(F32 r)
{
return F32((r * 180.0) / M_PI);
}
inline F64 mDegToRad(F64 d)
{
return (d * M_PI) / F64(180);
}
inline F64 mRadToDeg(F64 r)
{
return (r * 180.0) / M_PI;
}
#endif //_MMATHFN_H_