irrlicht/source/Irrlicht/S4DVertex.h
cutealien 3280b3319e Add floor32, ceil32, round32 compatibility functions for burnings renderer.
To avoid changing burnings now those functions have no IRRLICHT_FAST_MATH anymore, 
there's a new header irrMathFastCompat.h which has ..._fast functions doing the old behavior.
With the troubles they have documented.
I changed burnings to use those functions throughout.
Or as much as possible... Burnings probably also uses classes like SColor which also have functions 
using those, but I don't plan  to adapt them.
Maybe IRRLICHT_FAST_MATH should be a flag exlusive to burnings in the future, I don't think it makes 
much sense otherwise anymore (it often expects 32-bit asm).

git-svn-id: svn://svn.code.sf.net/p/irrlicht/code/trunk@6012 dfc29bdd-3216-0410-991c-e03cc46cb475
2019-12-16 16:15:04 +00:00

694 lines
13 KiB
C++

// Copyright (C) 2002-2012 Nikolaus Gebhardt / Thomas Alten
// This file is part of the "Irrlicht Engine".
// For conditions of distribution and use, see copyright notice in irrlicht.h
#ifndef __S_4D_VERTEX_H_INCLUDED__
#define __S_4D_VERTEX_H_INCLUDED__
#include "SoftwareDriver2_compile_config.h"
#include "SoftwareDriver2_helper.h"
#include "irrAllocator.h"
namespace irr
{
namespace video
{
struct sVec2
{
f32 x;
f32 y;
sVec2 () {}
sVec2 ( f32 s) : x ( s ), y ( s ) {}
sVec2 ( f32 _x, f32 _y )
: x ( _x ), y ( _y ) {}
void set ( f32 _x, f32 _y )
{
x = _x;
y = _y;
}
// f = a * t + b * ( 1 - t )
void interpolate(const sVec2& a, const sVec2& b, const f32 t)
{
x = b.x + ( ( a.x - b.x ) * t );
y = b.y + ( ( a.y - b.y ) * t );
}
sVec2 operator-(const sVec2& other) const
{
return sVec2(x - other.x, y - other.y);
}
sVec2 operator+(const sVec2& other) const
{
return sVec2(x + other.x, y + other.y);
}
void operator+=(const sVec2& other)
{
x += other.x;
y += other.y;
}
sVec2 operator*(const f32 s) const
{
return sVec2(x * s , y * s);
}
void operator*=( const f32 s)
{
x *= s;
y *= s;
}
void operator=(const sVec2& other)
{
x = other.x;
y = other.y;
}
};
// A8R8G8B8
struct sVec4;
struct sCompressedVec4
{
u32 argb;
void setA8R8G8B8 ( u32 value )
{
argb = value;
}
void setColorf ( const video::SColorf & color )
{
argb = core::floor32_fast( color.a * 255.f ) << 24 |
core::floor32_fast( color.r * 255.f ) << 16 |
core::floor32_fast( color.g * 255.f ) << 8 |
core::floor32_fast( color.b * 255.f );
}
void setVec4 ( const sVec4 & v );
// f = a * t + b * ( 1 - t )
void interpolate(const sCompressedVec4& a, const sCompressedVec4& b, const f32 t)
{
argb = PixelBlend32 ( b.argb, a.argb, core::floor32_fast( t * 256.f ) );
}
};
struct sVec4
{
union
{
struct { f32 x, y, z, w; };
struct { f32 a, r, g, b; };
// struct { sVec2 xy, zw; }; // sorry, this does not compile with gcc
};
sVec4 () {}
sVec4 ( f32 s) : x ( s ), y ( s ), z ( s ), w ( s ) {}
sVec4 ( f32 _x, f32 _y, f32 _z, f32 _w )
: x ( _x ), y ( _y ), z( _z ), w ( _w ){}
void set ( f32 _x, f32 _y, f32 _z, f32 _w )
{
x = _x;
y = _y;
z = _z;
w = _w;
}
void setA8R8G8B8 ( u32 argb )
{
x = ( ( argb & 0xFF000000 ) >> 24 ) * ( 1.f / 255.f );
y = ( ( argb & 0x00FF0000 ) >> 16 ) * ( 1.f / 255.f );
z = ( ( argb & 0x0000FF00 ) >> 8 ) * ( 1.f / 255.f );
w = ( ( argb & 0x000000FF ) ) * ( 1.f / 255.f );
}
void setColorf ( const video::SColorf & color )
{
x = color.a;
y = color.r;
z = color.g;
w = color.b;
}
// f = a * t + b * ( 1 - t )
void interpolate(const sVec4& a, const sVec4& b, const f32 t)
{
x = b.x + ( ( a.x - b.x ) * t );
y = b.y + ( ( a.y - b.y ) * t );
z = b.z + ( ( a.z - b.z ) * t );
w = b.w + ( ( a.w - b.w ) * t );
}
f32 dotProduct(const sVec4& other) const
{
return x*other.x + y*other.y + z*other.z + w*other.w;
}
f32 dot_xyz( const sVec4& other) const
{
return x*other.x + y*other.y + z*other.z;
}
f32 get_length_xyz_square () const
{
return x * x + y * y + z * z;
}
f32 get_length_xyz () const
{
return core::squareroot ( x * x + y * y + z * z );
}
void normalize_xyz ()
{
const f32 l = core::reciprocal_squareroot ( x * x + y * y + z * z );
x *= l;
y *= l;
z *= l;
}
void project_xyz ()
{
w = core::reciprocal ( w );
x *= w;
y *= w;
z *= w;
}
sVec4 operator-(const sVec4& other) const
{
return sVec4(x - other.x, y - other.y, z - other.z,w - other.w);
}
sVec4 operator+(const sVec4& other) const
{
return sVec4(x + other.x, y + other.y, z + other.z,w + other.w);
}
void operator+=(const sVec4& other)
{
x += other.x;
y += other.y;
z += other.z;
w += other.w;
}
sVec4 operator*(const f32 s) const
{
return sVec4(x * s , y * s, z * s,w * s);
}
sVec4 operator*(const sVec4 &other) const
{
return sVec4(x * other.x , y * other.y, z * other.z,w * other.w);
}
void mulReciprocal ( f32 s )
{
const f32 i = core::reciprocal ( s );
x = (f32) ( x * i );
y = (f32) ( y * i );
z = (f32) ( z * i );
w = (f32) ( w * i );
}
void mul ( const f32 s )
{
x *= s;
y *= s;
z *= s;
w *= s;
}
/*
void operator*=(f32 s)
{
x *= s;
y *= s;
z *= s;
w *= s;
}
*/
void operator*=(const sVec4 &other)
{
x *= other.x;
y *= other.y;
z *= other.z;
w *= other.w;
}
void operator=(const sVec4& other)
{
x = other.x;
y = other.y;
z = other.z;
w = other.w;
}
};
struct sVec3
{
union
{
struct { f32 r, g, b; };
struct { f32 x, y, z; };
};
sVec3 () {}
sVec3 ( f32 _x, f32 _y, f32 _z )
: r ( _x ), g ( _y ), b( _z ) {}
sVec3 ( const sVec4 &v )
: r ( v.x ), g ( v.y ), b( v.z ) {}
void set ( f32 _r, f32 _g, f32 _b )
{
r = _r;
g = _g;
b = _b;
}
void setR8G8B8 ( u32 argb )
{
r = ( ( argb & 0x00FF0000 ) >> 16 ) * ( 1.f / 255.f );
g = ( ( argb & 0x0000FF00 ) >> 8 ) * ( 1.f / 255.f );
b = ( ( argb & 0x000000FF ) ) * ( 1.f / 255.f );
}
void setColorf ( const video::SColorf & color )
{
r = color.r;
g = color.g;
b = color.b;
}
void add (const sVec3& other)
{
r += other.r;
g += other.g;
b += other.b;
}
void mulAdd(const sVec3& other, const f32 v)
{
r += other.r * v;
g += other.g * v;
b += other.b * v;
}
void mulAdd(const sVec3& v0, const sVec3& v1)
{
r += v0.r * v1.r;
g += v0.g * v1.g;
b += v0.b * v1.b;
}
void saturate ( sVec4 &dest, u32 argb )
{
dest.x = ( ( argb & 0xFF000000 ) >> 24 ) * ( 1.f / 255.f );
dest.y = core::min_ ( r, 1.f );
dest.z = core::min_ ( g, 1.f );
dest.w = core::min_ ( b, 1.f );
}
// f = a * t + b * ( 1 - t )
void interpolate(const sVec3& v0, const sVec3& v1, const f32 t)
{
r = v1.r + ( ( v0.r - v1.r ) * t );
g = v1.g + ( ( v0.g - v1.g ) * t );
b = v1.b + ( ( v0.b - v1.b ) * t );
}
sVec3 operator-(const sVec3& other) const
{
return sVec3(r - other.r, b - other.b, g - other.g);
}
sVec3 operator+(const sVec3& other) const
{
return sVec3(r + other.r, g + other.g, b + other.b);
}
sVec3 operator*(const f32 s) const
{
return sVec3(r * s , g * s, b * s);
}
sVec3 operator/(const f32 s) const
{
f32 inv = 1.f / s;
return sVec3(r * inv , g * inv, b * inv);
}
sVec3 operator*(const sVec3 &other) const
{
return sVec3(r * other.r , b * other.b, g * other.g);
}
void operator+=(const sVec3& other)
{
r += other.r;
g += other.g;
b += other.b;
}
void setLength ( f32 len )
{
const f32 l = len * core::reciprocal_squareroot ( r * r + g * g + b * b );
r *= l;
g *= l;
b *= l;
}
};
inline void sCompressedVec4::setVec4 ( const sVec4 & v )
{
argb = core::floor32_fast( v.x * 255.f ) << 24 |
core::floor32_fast( v.y * 255.f ) << 16 |
core::floor32_fast( v.z * 255.f ) << 8 |
core::floor32_fast( v.w * 255.f );
}
enum e4DVertexFlag
{
VERTEX4D_INSIDE = 0x0000003F,
VERTEX4D_CLIPMASK = 0x0000003F,
VERTEX4D_PROJECTED = 0x00000100,
VERTEX4D_FORMAT_MASK = 0xFFFF0000,
VERTEX4D_FORMAT_MASK_TEXTURE = 0x000F0000,
VERTEX4D_FORMAT_TEXTURE_1 = 0x00010000,
VERTEX4D_FORMAT_TEXTURE_2 = 0x00020000,
VERTEX4D_FORMAT_TEXTURE_3 = 0x00030000,
VERTEX4D_FORMAT_TEXTURE_4 = 0x00040000,
VERTEX4D_FORMAT_MASK_COLOR = 0x00F00000,
VERTEX4D_FORMAT_COLOR_1 = 0x00100000,
VERTEX4D_FORMAT_COLOR_2 = 0x00200000,
VERTEX4D_FORMAT_MASK_BUMP = 0x0F000000,
VERTEX4D_FORMAT_BUMP_DOT3 = 0x01000000,
};
const u32 MATERIAL_MAX_COLORS = 1;
const u32 BURNING_MATERIAL_MAX_TEXTURES = 2;
const u32 BURNING_MATERIAL_MAX_TANGENT = 1;
// dummy Vertex. used for calculation vertex memory size
struct s4DVertex_proxy
{
u32 flag;
sVec4 Pos;
sVec2 Tex[BURNING_MATERIAL_MAX_TEXTURES];
#ifdef SOFTWARE_DRIVER_2_USE_VERTEX_COLOR
sVec4 Color[MATERIAL_MAX_COLORS];
#endif
sVec3 LightTangent[BURNING_MATERIAL_MAX_TANGENT];
};
#define SIZEOF_SVERTEX 64
#define SIZEOF_SVERTEX_LOG2 6
/*!
Internal BurningVideo Vertex
*/
struct s4DVertex
{
u32 flag;
sVec4 Pos;
sVec2 Tex[ BURNING_MATERIAL_MAX_TEXTURES ];
#ifdef SOFTWARE_DRIVER_2_USE_VERTEX_COLOR
sVec4 Color[ MATERIAL_MAX_COLORS ];
#endif
sVec3 LightTangent[BURNING_MATERIAL_MAX_TANGENT];
//u8 fill [ SIZEOF_SVERTEX - sizeof (s4DVertex_proxy) ];
// f = a * t + b * ( 1 - t )
void interpolate(const s4DVertex& b, const s4DVertex& a, const f32 t)
{
u32 i;
u32 size;
Pos.interpolate ( a.Pos, b.Pos, t );
#ifdef SOFTWARE_DRIVER_2_USE_VERTEX_COLOR
size = (flag & VERTEX4D_FORMAT_MASK_COLOR) >> 20;
for ( i = 0; i!= size; ++i )
{
Color[i].interpolate ( a.Color[i], b.Color[i], t );
}
#endif
size = (flag & VERTEX4D_FORMAT_MASK_TEXTURE) >> 16;
for ( i = 0; i!= size; ++i )
{
Tex[i].interpolate ( a.Tex[i], b.Tex[i], t );
}
size = (flag & VERTEX4D_FORMAT_MASK_BUMP) >> 24;
for ( i = 0; i!= size; ++i )
{
LightTangent[i].interpolate ( a.LightTangent[i], b.LightTangent[i], t );
}
}
};
// ----------------- Vertex Cache ---------------------------
struct SAlignedVertex
{
SAlignedVertex ( u32 element, u32 aligned )
: ElementSize ( element )
{
u32 byteSize = (ElementSize << SIZEOF_SVERTEX_LOG2 ) + aligned;
mem = new u8 [ byteSize ];
data = (s4DVertex*) mem;
}
virtual ~SAlignedVertex ()
{
delete [] mem;
}
s4DVertex *data;
u8 *mem;
u32 ElementSize;
};
// hold info for different Vertex Types
struct SVSize
{
u32 Format;
u32 Pitch;
u32 TexSize;
};
// a cache info
struct SCacheInfo
{
u32 index;
u32 hit;
};
#define VERTEXCACHE_ELEMENT 16
#define VERTEXCACHE_MISS 0xFFFFFFFF
struct SVertexCache
{
SVertexCache (): mem ( VERTEXCACHE_ELEMENT * 2, 128 ) {}
SCacheInfo info[VERTEXCACHE_ELEMENT];
// Transformed and lite, clipping state
// + Clipped, Projected
SAlignedVertex mem;
// source
const void* vertices;
u32 vertexCount;
const void* indices;
u32 indexCount;
u32 indicesIndex;
u32 indicesRun;
// primitives consist of x vertices
u32 primitivePitch;
u32 vType; //E_VERTEX_TYPE
u32 pType; //scene::E_PRIMITIVE_TYPE
u32 iType; //E_INDEX_TYPE iType
};
// swap 2 pointer
REALINLINE void swapVertexPointer(const s4DVertex** v1, const s4DVertex** v2)
{
const s4DVertex* b = *v1;
*v1 = *v2;
*v2 = b;
}
// ------------------------ Internal Scanline Rasterizer -----------------------------
// internal scan convert
struct sScanConvertData
{
u8 left; // major edge left/right
u8 right; // !left
f32 invDeltaY[3]; // inverse edge delta y
f32 x[2]; // x coordinate
f32 slopeX[2]; // x slope along edges
#if defined ( SOFTWARE_DRIVER_2_USE_WBUFFER ) || defined ( SOFTWARE_DRIVER_2_PERSPECTIVE_CORRECT )
f32 w[2]; // w coordinate
fp24 slopeW[2]; // w slope along edges
#else
f32 z[2]; // z coordinate
f32 slopeZ[2]; // z slope along edges
#endif
sVec4 c[MATERIAL_MAX_COLORS][2]; // color
sVec4 slopeC[MATERIAL_MAX_COLORS][2]; // color slope along edges
sVec2 t[BURNING_MATERIAL_MAX_TEXTURES][2]; // texture
sVec2 slopeT[BURNING_MATERIAL_MAX_TEXTURES][2]; // texture slope along edges
sVec3 l[BURNING_MATERIAL_MAX_TANGENT][2]; // Light Tangent
sVec3 slopeL[BURNING_MATERIAL_MAX_TEXTURES][2]; // tanget slope along edges
};
// passed to scan Line
struct sScanLineData
{
s32 y; // y position of scanline
f32 x[2]; // x start, x end of scanline
#if defined ( SOFTWARE_DRIVER_2_USE_WBUFFER ) || defined ( SOFTWARE_DRIVER_2_PERSPECTIVE_CORRECT )
f32 w[2]; // w start, w end of scanline
#else
f32 z[2]; // z start, z end of scanline
#endif
#ifdef SOFTWARE_DRIVER_2_USE_VERTEX_COLOR
sVec4 c[MATERIAL_MAX_COLORS][2]; // color start, color end of scanline
#endif
sVec2 t[BURNING_MATERIAL_MAX_TEXTURES][2]; // texture start, texture end of scanline
sVec3 l[BURNING_MATERIAL_MAX_TANGENT][2]; // Light Tangent start, end
};
// passed to pixel Shader
struct sPixelShaderData
{
tVideoSample *dst;
fp24 *z;
s32 xStart;
s32 xEnd;
s32 dx;
s32 i;
};
/*
load a color value
*/
inline void getTexel_plain2 ( tFixPoint &r, tFixPoint &g, tFixPoint &b,
const sVec4 &v
)
{
r = tofix(v.y, FIX_POINT_F32_MUL);
g = tofix(v.z, FIX_POINT_F32_MUL);
b = tofix(v.w, FIX_POINT_F32_MUL);
}
/*
load a color value
*/
inline void getSample_color ( tFixPoint &a, tFixPoint &r, tFixPoint &g, tFixPoint &b,
const sVec4 &v
)
{
a = tofix(v.x, FIX_POINT_F32_MUL);
r = tofix ( v.y, COLOR_MAX * FIX_POINT_F32_MUL);
g = tofix ( v.z, COLOR_MAX * FIX_POINT_F32_MUL);
b = tofix ( v.w, COLOR_MAX * FIX_POINT_F32_MUL);
}
/*
load a color value
*/
inline void getSample_color ( tFixPoint &r, tFixPoint &g, tFixPoint &b,const sVec4 &v )
{
r = tofix ( v.y, COLOR_MAX * FIX_POINT_F32_MUL);
g = tofix ( v.z, COLOR_MAX * FIX_POINT_F32_MUL);
b = tofix ( v.w, COLOR_MAX * FIX_POINT_F32_MUL);
}
/*
load a color value
*/
inline void getSample_color ( tFixPoint &r, tFixPoint &g, tFixPoint &b,
const sVec4 &v, const f32 mulby )
{
r = tofix ( v.y, mulby);
g = tofix ( v.z, mulby);
b = tofix ( v.w, mulby);
}
}
}
#endif