Leaked source code of windows server 2003
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

452 lines
13 KiB

//----------------------------------------------------------------------------
//
// d3dutil.cpp
//
// Miscellanous utility functions.
//
// Copyright (C) Microsoft Corporation, 1997.
//
//----------------------------------------------------------------------------
#include "pch.cpp"
#pragma hdrstop
#include <span.h>
DBG_DECLARE_FILE();
// Declare TextureDiff as an out-of-line function.
FLOAT FASTCALL
TextureDiff(FLOAT fTb, FLOAT fTa, INT iMode)
#include <texdiff.h>
//----------------------------------------------------------------------------
//
// DebugBreakFn
//
// Stub function that should never be called. Prints a warning and
// DebugBreaks. Can be inserted in any function table, although it
// will destroy the stack frame with callconv or argument mismatch.
// That's OK since if it's called something has gone wrong.
//
//----------------------------------------------------------------------------
void FASTCALL
DebugBreakFn(void)
{
GDPF(("!! DebugBreakFn called. Leaving this function may destroy\n"));
GDPF((" the stack frame. !!\n"));
#if DBG
DebugBreak();
#endif
}
//----------------------------------------------------------------------------
//
// OctagonNorm
//
// Returns a good approximation to sqrt(fX*fX + fY*fY)
//
//----------------------------------------------------------------------------
FLOAT FASTCALL
OctagonNorm(FLOAT fX, FLOAT fY)
{
fX = ABSF(fX);
fY = ABSF(fY);
return ((11.0f/32.0f)*(fX + fY) + (21.0f/32.0f)*max(fX, fY));
}
//----------------------------------------------------------------------------
//
// ComputeLOD
//
// Computes mipmap level for the given W by deriving U and V and
// then computing LOD from the dU and dV gradients.
//
//----------------------------------------------------------------------------
INT FASTCALL
ComputeLOD(PCD3DI_RASTCTX pCtx,
FLOAT fU, FLOAT fV, FLOAT fW,
FLOAT fDUoWDX, FLOAT fDVoWDX, FLOAT fDOoWDX,
FLOAT fDUoWDY, FLOAT fDVoWDY, FLOAT fDOoWDY)
{
// Compute coverage gradients.
FLOAT fDUDX = ABSF(fW * (fDUoWDX - fU * fDOoWDX));
FLOAT fDUDY = ABSF(fW * (fDUoWDY - fU * fDOoWDY));
FLOAT fDVDX = ABSF(fW * (fDVoWDX - fV * fDOoWDX));
FLOAT fDVDY = ABSF(fW * (fDVoWDY - fV * fDOoWDY));
// Scale gradients to texture LOD 0 size.
fDUDX *= (FLOAT)pCtx->pTexture[0]->iSizeU;
fDUDY *= (FLOAT)pCtx->pTexture[0]->iSizeU;
fDVDX *= (FLOAT)pCtx->pTexture[0]->iSizeV;
fDVDY *= (FLOAT)pCtx->pTexture[0]->iSizeV;
// Determine pixel coverage value to use.
FLOAT fCoverage;
// too fuzzy
#ifdef COVERAGE_MAXGRAD
fCoverage = max(fDUDX, fDUDY);
fCoverage = max(fCoverage, fDVDX);
fCoverage = max(fCoverage, fDVDY);
#endif
// too sharp, in particular, for aligned cases, fCoverage is always 0
// which leads to iLOD of LOD_MIN regardless of orientation
#ifdef COVERAGE_MINGRAD
fCoverage = min(fDUDX, fDUDY);
fCoverage = min(fCoverage, fDVDX);
fCoverage = min(fCoverage, fDVDY);
#endif
#define COVERAGE_AVERAGE 1
#ifdef COVERAGE_AVERAGE
// use OctagonNorm to approximate each length of parallelogram
// approximating texture coverage, and arithmetically average those to
// get the coverage.
fCoverage = (OctagonNorm(fDUDX, fDVDX) + OctagonNorm(fDUDY, fDVDY))/2.0f;
#endif
// Compute approximate log2 of coverage.
FLOAT fLOD = APPXLG2F(fCoverage);
// Apply LOD bias.
fLOD += pCtx->pTexture[0]->fLODBias;
INT iLOD = FTOI(fLOD * LOD_SCALE);
// Clamp to available levels. Not clamped to zero so that the span
// code can check for magnification cases with a sign check.
iLOD = min(iLOD, pCtx->pTexture[0]->iMaxScaledLOD);
return max(LOD_MIN, iLOD);
}
//----------------------------------------------------------------------------
//
// ComputeTableFog
//
// Computes table fog values based on render state and the given Z.
// ATTENTION - Brute force for non-linear modes. Should be optimized
// to use a table-based approximation.
//
//----------------------------------------------------------------------------
UINT FASTCALL
ComputeTableFog(PDWORD pdwRenderState,
FLOAT fZ)
{
double dPow;
switch(pdwRenderState[D3DRENDERSTATE_FOGTABLEMODE])
{
case D3DFOG_LINEAR:
{
FLOAT fFogStart = ASFLOAT(pdwRenderState[D3DRENDERSTATE_FOGTABLESTART]);
FLOAT fFogEnd = ASFLOAT(pdwRenderState[D3DRENDERSTATE_FOGTABLEEND]);
if (fZ >= fFogEnd)
{
return 0;
}
if (fZ <= fFogStart)
{
return FTOI(FOG_ONE_SCALE-1.0F);
}
return FTOI(((fFogEnd - fZ) / (fFogEnd - fFogStart)) * (FOG_ONE_SCALE-1.0F));
}
case D3DFOG_EXP:
dPow = (double)
(ASFLOAT(pdwRenderState[D3DRENDERSTATE_FOGTABLEDENSITY]) * fZ);
// note that exp(-x) returns a result in the range (0.0, 1.0]
// for x >= 0
dPow = exp(-dPow);
return FTOI((FLOAT)dPow * (FOG_ONE_SCALE-1.0F));
case D3DFOG_EXP2:
dPow = (double)
(ASFLOAT(pdwRenderState[D3DRENDERSTATE_FOGTABLEDENSITY]) * fZ);
dPow = exp(-dPow * dPow);
return FTOI((FLOAT)dPow * (FOG_ONE_SCALE-1.0F));
}
GASSERTMSG(FALSE, ("ComputeTableFog unreachable\n"));
return 0;
}
//----------------------------------------------------------------------------
//
// pVecNormalize2
//
// Normalizes the given D3DVECTOR. Supports in-place operation.
//
//----------------------------------------------------------------------------
void FASTCALL
pVecNormalize2(LPD3DVECTOR pVec, LPD3DVECTOR pRes)
{
FLOAT fLen;
fLen = pVecLenSq(pVec);
if (FLOAT_CMP_POS(fLen, <=, g_fNearZero))
{
pVecSet(pRes, 0.0f, 0.0f, 0.0f);
return;
}
fLen = ISQRTF(fLen);
pVecScale(pVec, fLen, pRes);
}
//-----------------------------------------------------------------------------
//
// IntLog2
//
// Do a quick, integer log2 for exact powers of 2.
//
//-----------------------------------------------------------------------------
UINT32 FASTCALL
IntLog2(UINT32 x)
{
UINT32 y = 0;
x >>= 1;
while(x != 0)
{
x >>= 1;
y++;
}
return y;
}
//---------------------------------------------------------------------
// Builds normalized plane equations going through 3 points
//
// Returns:
// 0 - if success
// -1 - if can not build plane
//
int MakePlane(D3DVECTOR *v1, D3DVECTOR *v2, D3DVECTOR *v3, D3DVECTORH *plane)
{
D3DVECTOR a;
D3DVECTOR b;
pVecSub(v2, v1, &a);
pVecSub(v3, v1, &b);
plane->x = a.y*b.z - a.z*b.y;
plane->y = a.z*b.x - a.x*b.z;
plane->z = a.x*b.y - a.y*b.x;
plane->w = - pVecDot(v1, plane);
double tmp = pVecDot(plane, plane);
if (tmp <= 0)
return -1;
tmp = 1.0/sqrt(tmp);
plane->x = (D3DVALUE)(plane->x * tmp);
plane->y = (D3DVALUE)(plane->y * tmp);
plane->z = (D3DVALUE)(plane->z * tmp);
plane->w = (D3DVALUE)(plane->w * tmp);
return 0;
}
//---------------------------------------------------------------------
// This function uses Cramer's Rule to calculate the matrix inverse.
// See nt\private\windows\opengl\serever\soft\so_math.c
//
// Returns:
// 0 - if success
// -1 - if input matrix is singular
//
int Inverse4x4(D3DMATRIX *src, D3DMATRIX *inverse)
{
double x00, x01, x02;
double x10, x11, x12;
double x20, x21, x22;
double rcp;
double x30, x31, x32;
double y01, y02, y03, y12, y13, y23;
double z02, z03, z12, z13, z22, z23, z32, z33;
#define x03 x01
#define x13 x11
#define x23 x21
#define x33 x31
#define z00 x02
#define z10 x12
#define z20 x22
#define z30 x32
#define z01 x03
#define z11 x13
#define z21 x23
#define z31 x33
/* read 1st two columns of matrix into registers */
x00 = src->_11;
x01 = src->_12;
x10 = src->_21;
x11 = src->_22;
x20 = src->_31;
x21 = src->_32;
x30 = src->_41;
x31 = src->_42;
/* compute all six 2x2 determinants of 1st two columns */
y01 = x00*x11 - x10*x01;
y02 = x00*x21 - x20*x01;
y03 = x00*x31 - x30*x01;
y12 = x10*x21 - x20*x11;
y13 = x10*x31 - x30*x11;
y23 = x20*x31 - x30*x21;
/* read 2nd two columns of matrix into registers */
x02 = src->_13;
x03 = src->_14;
x12 = src->_23;
x13 = src->_24;
x22 = src->_33;
x23 = src->_34;
x32 = src->_43;
x33 = src->_44;
/* compute all 3x3 cofactors for 2nd two columns */
z33 = x02*y12 - x12*y02 + x22*y01;
z23 = x12*y03 - x32*y01 - x02*y13;
z13 = x02*y23 - x22*y03 + x32*y02;
z03 = x22*y13 - x32*y12 - x12*y23;
z32 = x13*y02 - x23*y01 - x03*y12;
z22 = x03*y13 - x13*y03 + x33*y01;
z12 = x23*y03 - x33*y02 - x03*y23;
z02 = x13*y23 - x23*y13 + x33*y12;
/* compute all six 2x2 determinants of 2nd two columns */
y01 = x02*x13 - x12*x03;
y02 = x02*x23 - x22*x03;
y03 = x02*x33 - x32*x03;
y12 = x12*x23 - x22*x13;
y13 = x12*x33 - x32*x13;
y23 = x22*x33 - x32*x23;
/* read 1st two columns of matrix into registers */
x00 = src->_11;
x01 = src->_12;
x10 = src->_21;
x11 = src->_22;
x20 = src->_31;
x21 = src->_32;
x30 = src->_41;
x31 = src->_42;
/* compute all 3x3 cofactors for 1st column */
z30 = x11*y02 - x21*y01 - x01*y12;
z20 = x01*y13 - x11*y03 + x31*y01;
z10 = x21*y03 - x31*y02 - x01*y23;
z00 = x11*y23 - x21*y13 + x31*y12;
/* compute 4x4 determinant & its reciprocal */
rcp = x30*z30 + x20*z20 + x10*z10 + x00*z00;
if (rcp == (float)0)
return -1;
rcp = (float)1/rcp;
/* compute all 3x3 cofactors for 2nd column */
z31 = x00*y12 - x10*y02 + x20*y01;
z21 = x10*y03 - x30*y01 - x00*y13;
z11 = x00*y23 - x20*y03 + x30*y02;
z01 = x20*y13 - x30*y12 - x10*y23;
/* multiply all 3x3 cofactors by reciprocal */
inverse->_11 = (float)(z00*rcp);
inverse->_21 = (float)(z01*rcp);
inverse->_12 = (float)(z10*rcp);
inverse->_31 = (float)(z02*rcp);
inverse->_13 = (float)(z20*rcp);
inverse->_41 = (float)(z03*rcp);
inverse->_14 = (float)(z30*rcp);
inverse->_22 = (float)(z11*rcp);
inverse->_32 = (float)(z12*rcp);
inverse->_23 = (float)(z21*rcp);
inverse->_42 = (float)(z13*rcp);
inverse->_24 = (float)(z31*rcp);
inverse->_33 = (float)(z22*rcp);
inverse->_43 = (float)(z23*rcp);
inverse->_34 = (float)(z32*rcp);
inverse->_44 = (float)(z33*rcp);
return 0;
}
//---------------------------------------------------------------------
// Checks the FVF flags for errors and returns the stride in bytes between
// vertices.
//
// Returns:
// HRESULT and stride in bytes between vertices
//
//---------------------------------------------------------------------
HRESULT FASTCALL
FVFCheckAndStride(DWORD dwFVF, DWORD* pdwStride)
{
if (NULL == pdwStride)
{
return DDERR_INVALIDPARAMS;
}
if ( (dwFVF & (D3DFVF_RESERVED0 | D3DFVF_RESERVED1 | D3DFVF_RESERVED2 |
D3DFVF_NORMAL)) ||
((dwFVF & (D3DFVF_XYZ | D3DFVF_XYZRHW)) == 0) )
{
// can't set reserved bits, shouldn't have normals in
// output to rasterizers, and must have coordinates
return DDERR_INVALIDPARAMS;
}
DWORD dwStride;
if (dwFVF != D3DFVF_TLVERTEX)
{ // New (non TL)FVF vertex
// XYZ
dwStride = sizeof(D3DVALUE) * 3;
if (dwFVF & D3DFVF_XYZRHW)
{
dwStride += sizeof(D3DVALUE);
}
if (dwFVF & D3DFVF_DIFFUSE)
{
dwStride += sizeof(D3DCOLOR);
}
if (dwFVF & D3DFVF_SPECULAR)
{
dwStride += sizeof(D3DCOLOR);
}
INT iTexCount = (dwFVF & D3DFVF_TEXCOUNT_MASK) >> D3DFVF_TEXCOUNT_SHIFT;
dwStride += sizeof(D3DVALUE) * 2 * iTexCount;
}
else
{ // (Legacy) TL vertex
dwStride = sizeof(D3DTLVERTEX);
}
*pdwStride = dwStride;
return D3D_OK;
}
//---------------------------------------------------------------------
// Gets the value from DIRECT3D registry key
// Returns TRUE if success
// If fails value is not changed
//
BOOL GetD3DRegValue(DWORD type, char *valueName, LPVOID value, DWORD dwSize)
{
HKEY hKey = (HKEY) NULL;
if (ERROR_SUCCESS == RegOpenKey(HKEY_LOCAL_MACHINE, RESPATH_D3D, &hKey))
{
DWORD dwType;
LONG result;
result = RegQueryValueEx(hKey, valueName, NULL, &dwType,
(LPBYTE)value, &dwSize);
RegCloseKey(hKey);
return result == ERROR_SUCCESS && dwType == type;
}
else
return FALSE;
}