mirror of
https://github.com/DrBeef/JKXR.git
synced 2024-11-15 08:52:15 +00:00
1738 lines
42 KiB
C++
1738 lines
42 KiB
C++
/*
|
|
===========================================================================
|
|
Copyright (C) 1999 - 2005, Id Software, Inc.
|
|
Copyright (C) 2000 - 2013, Raven Software, Inc.
|
|
Copyright (C) 2001 - 2013, Activision, Inc.
|
|
Copyright (C) 2005 - 2015, ioquake3 contributors
|
|
Copyright (C) 2013 - 2015, OpenJK contributors
|
|
|
|
This file is part of the OpenJK source code.
|
|
|
|
OpenJK is free software; you can redistribute it and/or modify it
|
|
under the terms of the GNU General Public License version 2 as
|
|
published by the Free Software Foundation.
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, see <http://www.gnu.org/licenses/>.
|
|
===========================================================================
|
|
*/
|
|
|
|
// tr_image.c
|
|
#include "tr_local.h"
|
|
#include "../rd-common/tr_common.h"
|
|
#include <png.h>
|
|
#include <map>
|
|
using namespace std;
|
|
|
|
#if !defined(GL_RGB5)
|
|
#define GL_RGB5 0x8050 // same as GL_RGB5_EXT
|
|
#endif
|
|
#if !defined(GL_RGBA4)
|
|
#define GL_RGBA4 0x8056 // same as GL_RGBA4_EXT and GL_RGBA4_OES
|
|
#endif
|
|
|
|
static byte s_intensitytable[256];
|
|
static unsigned char s_gammatable[256];
|
|
|
|
int gl_filter_min = GL_LINEAR_MIPMAP_NEAREST;
|
|
int gl_filter_max = GL_LINEAR;
|
|
|
|
#define FILE_HASH_SIZE 1024 // actually, the shader code needs this (from another module, great).
|
|
//static image_t* hashTable[FILE_HASH_SIZE];
|
|
|
|
/*
|
|
** R_GammaCorrect
|
|
*/
|
|
void R_GammaCorrect( byte *buffer, int bufSize ) {
|
|
int i;
|
|
|
|
for ( i = 0; i < bufSize; i++ ) {
|
|
buffer[i] = s_gammatable[buffer[i]];
|
|
}
|
|
}
|
|
|
|
typedef struct textureMode_s {
|
|
const char *name;
|
|
int minimize, maximize;
|
|
} textureMode_t;
|
|
|
|
textureMode_t modes[] = {
|
|
{"GL_NEAREST", GL_NEAREST, GL_NEAREST},
|
|
{"GL_LINEAR", GL_LINEAR, GL_LINEAR},
|
|
{"GL_NEAREST_MIPMAP_NEAREST", GL_NEAREST_MIPMAP_NEAREST, GL_NEAREST},
|
|
{"GL_LINEAR_MIPMAP_NEAREST", GL_LINEAR_MIPMAP_NEAREST, GL_LINEAR},
|
|
{"GL_NEAREST_MIPMAP_LINEAR", GL_NEAREST_MIPMAP_LINEAR, GL_NEAREST},
|
|
{"GL_LINEAR_MIPMAP_LINEAR", GL_LINEAR_MIPMAP_LINEAR, GL_LINEAR}
|
|
};
|
|
|
|
static const size_t numTextureModes = ARRAY_LEN(modes);
|
|
|
|
/*
|
|
================
|
|
return a hash value for the filename
|
|
================
|
|
*/
|
|
long generateHashValue( const char *fname ) {
|
|
int i;
|
|
long hash;
|
|
char letter;
|
|
|
|
hash = 0;
|
|
i = 0;
|
|
while (fname[i] != '\0') {
|
|
letter = tolower(fname[i]);
|
|
if (letter =='.') break; // don't include extension
|
|
if (letter =='\\') letter = '/'; // damn path names
|
|
hash+=(long)(letter)*(i+119);
|
|
i++;
|
|
}
|
|
hash &= (FILE_HASH_SIZE-1);
|
|
return hash;
|
|
}
|
|
|
|
// makeup a nice clean, consistant name to query for and file under, for map<> usage...
|
|
//
|
|
char *GenerateImageMappingName( const char *name )
|
|
{
|
|
static char sName[MAX_QPATH];
|
|
int i=0;
|
|
char letter;
|
|
|
|
while (name[i] != '\0' && i<MAX_QPATH-1)
|
|
{
|
|
letter = tolower(name[i]);
|
|
if (letter =='.') break; // don't include extension
|
|
if (letter =='\\') letter = '/'; // damn path names
|
|
sName[i++] = letter;
|
|
}
|
|
sName[i]=0;
|
|
|
|
return &sName[0];
|
|
}
|
|
|
|
/*
|
|
===============
|
|
GL_TextureMode
|
|
===============
|
|
*/
|
|
void GL_TextureMode( const char *string ) {
|
|
size_t i;
|
|
image_t *glt;
|
|
|
|
for ( i = 0; i < numTextureModes ; i++ ) {
|
|
if ( !Q_stricmp( modes[i].name, string ) ) {
|
|
break;
|
|
}
|
|
}
|
|
|
|
if ( i == numTextureModes ) {
|
|
ri.Printf( PRINT_ALL, "bad filter name\n" );
|
|
for ( i = 0; i < numTextureModes ; i++ ) {
|
|
ri.Printf( PRINT_ALL, "%s\n", modes[i].name );
|
|
}
|
|
return;
|
|
}
|
|
|
|
gl_filter_min = modes[i].minimize;
|
|
gl_filter_max = modes[i].maximize;
|
|
|
|
// If the level they requested is less than possible, set the max possible...
|
|
if ( r_ext_texture_filter_anisotropic->value > glConfig.maxTextureFilterAnisotropy )
|
|
ri.Cvar_SetValue( "r_ext_texture_filter_anisotropic", glConfig.maxTextureFilterAnisotropy );
|
|
|
|
// change all the existing mipmap texture objects
|
|
R_Images_StartIteration();
|
|
while ( (glt = R_Images_GetNextIteration()) != NULL)
|
|
{
|
|
if ( glt->mipmap ) {
|
|
GL_Bind (glt);
|
|
qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, gl_filter_min);
|
|
qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, gl_filter_max);
|
|
|
|
if(glConfig.maxTextureFilterAnisotropy>0) {
|
|
if(r_ext_texture_filter_anisotropic->integer>1) {
|
|
qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAX_ANISOTROPY_EXT, r_ext_texture_filter_anisotropic->value);
|
|
} else {
|
|
qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAX_ANISOTROPY_EXT, 1.0f);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
static float R_BytesPerTex (int format)
|
|
{
|
|
switch ( format ) {
|
|
case 1:
|
|
//"I "
|
|
return 1;
|
|
break;
|
|
case 2:
|
|
//"IA "
|
|
return 2;
|
|
break;
|
|
case 3:
|
|
//"RGB "
|
|
#ifdef HAVE_GLES
|
|
return 3;
|
|
#else
|
|
return glConfig.colorBits/8.0f;
|
|
#endif
|
|
break;
|
|
case 4:
|
|
//"RGBA "
|
|
#ifdef HAVE_GLES
|
|
return 4;
|
|
#else
|
|
return glConfig.colorBits/8.0f;
|
|
break;
|
|
|
|
case GL_RGBA4:
|
|
//"RGBA4"
|
|
return 2;
|
|
break;
|
|
case GL_RGB5:
|
|
//"RGB5 "
|
|
return 2;
|
|
#endif
|
|
break;
|
|
|
|
#ifndef HAVE_GLES
|
|
case GL_RGBA8:
|
|
//"RGBA8"
|
|
return 4;
|
|
break;
|
|
case GL_RGB8:
|
|
//"RGB8"
|
|
return 4;
|
|
break;
|
|
|
|
case GL_RGB4_S3TC:
|
|
//"S3TC "
|
|
return 0.33333f;
|
|
break;
|
|
case GL_COMPRESSED_RGB_S3TC_DXT1_EXT:
|
|
//"DXT1 "
|
|
return 0.33333f;
|
|
break;
|
|
case GL_COMPRESSED_RGBA_S3TC_DXT5_EXT:
|
|
//"DXT5 "
|
|
return 1;
|
|
break;
|
|
|
|
#endif
|
|
default:
|
|
//"???? "
|
|
return 4;
|
|
}
|
|
}
|
|
|
|
/*
|
|
===============
|
|
R_SumOfUsedImages
|
|
===============
|
|
*/
|
|
float R_SumOfUsedImages( qboolean bUseFormat )
|
|
{
|
|
int total = 0;
|
|
image_t *pImage;
|
|
|
|
R_Images_StartIteration();
|
|
while ( (pImage = R_Images_GetNextIteration()) != NULL)
|
|
{
|
|
if ( pImage->frameUsed == tr.frameCount- 1 ) {//it has already been advanced for the next frame, so...
|
|
if (bUseFormat)
|
|
{
|
|
float bytePerTex = R_BytesPerTex (pImage->internalFormat);
|
|
total += bytePerTex * (pImage->width * pImage->height);
|
|
}
|
|
else
|
|
{
|
|
total += pImage->width * pImage->height;
|
|
}
|
|
}
|
|
}
|
|
|
|
return total;
|
|
}
|
|
|
|
/*
|
|
===============
|
|
R_ImageList_f
|
|
===============
|
|
*/
|
|
void R_ImageList_f( void ) {
|
|
int i=0;
|
|
image_t *image;
|
|
int texels=0;
|
|
float texBytes = 0.0f;
|
|
const char *yesno[] = {"no ", "yes"};
|
|
|
|
ri.Printf( PRINT_ALL, "\n -w-- -h-- -mm- -if-- wrap --name-------\n");
|
|
|
|
int iNumImages = R_Images_StartIteration();
|
|
while ( (image = R_Images_GetNextIteration()) != NULL)
|
|
{
|
|
texels += image->width*image->height;
|
|
texBytes += image->width*image->height * R_BytesPerTex (image->internalFormat);
|
|
ri.Printf( PRINT_ALL, "%4i: %4i %4i %s ",
|
|
i, image->width, image->height, yesno[image->mipmap] );
|
|
switch ( image->internalFormat ) {
|
|
case 1:
|
|
ri.Printf( PRINT_ALL, "I " );
|
|
break;
|
|
case 2:
|
|
ri.Printf( PRINT_ALL, "IA " );
|
|
break;
|
|
case 3:
|
|
ri.Printf( PRINT_ALL, "RGB " );
|
|
break;
|
|
case 4:
|
|
ri.Printf( PRINT_ALL, "RGBA " );
|
|
break;
|
|
#ifndef HAVE_GLES
|
|
case GL_RGBA8:
|
|
ri.Printf( PRINT_ALL, "RGBA8" );
|
|
break;
|
|
case GL_RGB8:
|
|
ri.Printf( PRINT_ALL, "RGB8" );
|
|
break;
|
|
case GL_RGB4_S3TC:
|
|
ri.Printf( PRINT_ALL, "S3TC " );
|
|
break;
|
|
case GL_COMPRESSED_RGB_S3TC_DXT1_EXT:
|
|
ri.Printf( PRINT_ALL, "DXT1 " );
|
|
break;
|
|
case GL_COMPRESSED_RGBA_S3TC_DXT5_EXT:
|
|
ri.Printf( PRINT_ALL, "DXT5 " );
|
|
break;
|
|
case GL_RGBA4:
|
|
ri.Printf( PRINT_ALL, "RGBA4" );
|
|
break;
|
|
case GL_RGB5:
|
|
ri.Printf( PRINT_ALL, "RGB5 " );
|
|
break;
|
|
|
|
#endif
|
|
default:
|
|
ri.Printf( PRINT_ALL, "???? " );
|
|
}
|
|
|
|
switch ( image->wrapClampMode ) {
|
|
case GL_REPEAT:
|
|
ri.Printf( PRINT_ALL, "rept " );
|
|
break;
|
|
#ifndef HAVE_GLES
|
|
case GL_CLAMP:
|
|
ri.Printf( PRINT_ALL, "clmp " );
|
|
break;
|
|
#endif
|
|
case GL_CLAMP_TO_EDGE:
|
|
ri.Printf( PRINT_ALL, "clpE " );
|
|
break;
|
|
default:
|
|
ri.Printf( PRINT_ALL, "%4i ", image->wrapClampMode );
|
|
break;
|
|
}
|
|
|
|
ri.Printf( PRINT_ALL, "%s\n", image->imgName );
|
|
i++;
|
|
}
|
|
ri.Printf( PRINT_ALL, " ---------\n");
|
|
ri.Printf( PRINT_ALL, " -w-- -h-- -mm- -if- wrap --name-------\n");
|
|
ri.Printf( PRINT_ALL, " %i total texels (not including mipmaps)\n", texels );
|
|
ri.Printf( PRINT_ALL, " %.2fMB total texture mem (not including mipmaps)\n", texBytes/1048576.0f );
|
|
ri.Printf( PRINT_ALL, " %i total images\n\n", iNumImages );
|
|
}
|
|
|
|
//=======================================================================
|
|
|
|
|
|
/*
|
|
================
|
|
R_LightScaleTexture
|
|
|
|
Scale up the pixel values in a texture to increase the
|
|
lighting range
|
|
================
|
|
*/
|
|
void R_LightScaleTexture (unsigned *in, int inwidth, int inheight, qboolean only_gamma )
|
|
{
|
|
if ( only_gamma )
|
|
{
|
|
if ( !glConfig.deviceSupportsGamma )
|
|
{
|
|
int i, c;
|
|
byte *p;
|
|
|
|
p = (byte *)in;
|
|
|
|
c = inwidth*inheight;
|
|
for (i=0 ; i<c ; i++, p+=4)
|
|
{
|
|
p[0] = s_gammatable[p[0]];
|
|
p[1] = s_gammatable[p[1]];
|
|
p[2] = s_gammatable[p[2]];
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
int i, c;
|
|
byte *p;
|
|
|
|
p = (byte *)in;
|
|
|
|
c = inwidth*inheight;
|
|
|
|
if ( glConfig.deviceSupportsGamma )
|
|
{
|
|
for (i=0 ; i<c ; i++, p+=4)
|
|
{
|
|
p[0] = s_intensitytable[p[0]];
|
|
p[1] = s_intensitytable[p[1]];
|
|
p[2] = s_intensitytable[p[2]];
|
|
}
|
|
}
|
|
else
|
|
{
|
|
for (i=0 ; i<c ; i++, p+=4)
|
|
{
|
|
p[0] = s_gammatable[s_intensitytable[p[0]]];
|
|
p[1] = s_gammatable[s_intensitytable[p[1]]];
|
|
p[2] = s_gammatable[s_intensitytable[p[2]]];
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
================
|
|
R_MipMap2
|
|
|
|
Operates in place, quartering the size of the texture
|
|
Proper linear filter
|
|
================
|
|
*/
|
|
static void R_MipMap2( unsigned *in, int inWidth, int inHeight ) {
|
|
int i, j, k;
|
|
byte *outpix;
|
|
int inWidthMask, inHeightMask;
|
|
int total;
|
|
int outWidth, outHeight;
|
|
unsigned *temp;
|
|
|
|
outWidth = inWidth >> 1;
|
|
outHeight = inHeight >> 1;
|
|
temp = (unsigned int *)R_Malloc( outWidth * outHeight * 4, TAG_TEMP_WORKSPACE, qfalse );
|
|
|
|
inWidthMask = inWidth - 1;
|
|
inHeightMask = inHeight - 1;
|
|
|
|
for ( i = 0 ; i < outHeight ; i++ ) {
|
|
for ( j = 0 ; j < outWidth ; j++ ) {
|
|
outpix = (byte *) ( temp + i * outWidth + j );
|
|
for ( k = 0 ; k < 4 ; k++ ) {
|
|
total =
|
|
1 * ((byte *)&in[ ((i*2-1)&inHeightMask)*inWidth + ((j*2-1)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2-1)&inHeightMask)*inWidth + ((j*2)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2-1)&inHeightMask)*inWidth + ((j*2+1)&inWidthMask) ])[k] +
|
|
1 * ((byte *)&in[ ((i*2-1)&inHeightMask)*inWidth + ((j*2+2)&inWidthMask) ])[k] +
|
|
|
|
2 * ((byte *)&in[ ((i*2)&inHeightMask)*inWidth + ((j*2-1)&inWidthMask) ])[k] +
|
|
4 * ((byte *)&in[ ((i*2)&inHeightMask)*inWidth + ((j*2)&inWidthMask) ])[k] +
|
|
4 * ((byte *)&in[ ((i*2)&inHeightMask)*inWidth + ((j*2+1)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2)&inHeightMask)*inWidth + ((j*2+2)&inWidthMask) ])[k] +
|
|
|
|
2 * ((byte *)&in[ ((i*2+1)&inHeightMask)*inWidth + ((j*2-1)&inWidthMask) ])[k] +
|
|
4 * ((byte *)&in[ ((i*2+1)&inHeightMask)*inWidth + ((j*2)&inWidthMask) ])[k] +
|
|
4 * ((byte *)&in[ ((i*2+1)&inHeightMask)*inWidth + ((j*2+1)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2+1)&inHeightMask)*inWidth + ((j*2+2)&inWidthMask) ])[k] +
|
|
|
|
1 * ((byte *)&in[ ((i*2+2)&inHeightMask)*inWidth + ((j*2-1)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2+2)&inHeightMask)*inWidth + ((j*2)&inWidthMask) ])[k] +
|
|
2 * ((byte *)&in[ ((i*2+2)&inHeightMask)*inWidth + ((j*2+1)&inWidthMask) ])[k] +
|
|
1 * ((byte *)&in[ ((i*2+2)&inHeightMask)*inWidth + ((j*2+2)&inWidthMask) ])[k];
|
|
outpix[k] = total / 36;
|
|
}
|
|
}
|
|
}
|
|
|
|
memcpy( in, temp, outWidth * outHeight * 4 );
|
|
R_Free( temp );
|
|
}
|
|
|
|
/*
|
|
================
|
|
R_MipMap
|
|
|
|
Operates in place, quartering the size of the texture
|
|
================
|
|
*/
|
|
static void R_MipMap (byte *in, int width, int height) {
|
|
int i, j;
|
|
byte *out;
|
|
int row;
|
|
|
|
if ( !r_simpleMipMaps->integer ) {
|
|
R_MipMap2( (unsigned *)in, width, height );
|
|
return;
|
|
}
|
|
|
|
if ( width == 1 && height == 1 ) {
|
|
return;
|
|
}
|
|
|
|
row = width * 4;
|
|
out = in;
|
|
width >>= 1;
|
|
height >>= 1;
|
|
|
|
if ( width == 0 || height == 0 ) {
|
|
width += height; // get largest
|
|
for (i=0 ; i<width ; i++, out+=4, in+=8 ) {
|
|
out[0] = ( in[0] + in[4] )>>1;
|
|
out[1] = ( in[1] + in[5] )>>1;
|
|
out[2] = ( in[2] + in[6] )>>1;
|
|
out[3] = ( in[3] + in[7] )>>1;
|
|
}
|
|
return;
|
|
}
|
|
|
|
for (i=0 ; i<height ; i++, in+=row) {
|
|
for (j=0 ; j<width ; j++, out+=4, in+=8) {
|
|
out[0] = (in[0] + in[4] + in[row+0] + in[row+4])>>2;
|
|
out[1] = (in[1] + in[5] + in[row+1] + in[row+5])>>2;
|
|
out[2] = (in[2] + in[6] + in[row+2] + in[row+6])>>2;
|
|
out[3] = (in[3] + in[7] + in[row+3] + in[row+7])>>2;
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
==================
|
|
R_BlendOverTexture
|
|
|
|
Apply a color blend over a set of pixels
|
|
==================
|
|
*/
|
|
static void R_BlendOverTexture( byte *data, int pixelCount, byte blend[4] ) {
|
|
int i;
|
|
int inverseAlpha;
|
|
int premult[3];
|
|
|
|
inverseAlpha = 255 - blend[3];
|
|
premult[0] = blend[0] * blend[3];
|
|
premult[1] = blend[1] * blend[3];
|
|
premult[2] = blend[2] * blend[3];
|
|
|
|
for ( i = 0 ; i < pixelCount ; i++, data+=4 ) {
|
|
data[0] = ( data[0] * inverseAlpha + premult[0] ) >> 9;
|
|
data[1] = ( data[1] * inverseAlpha + premult[1] ) >> 9;
|
|
data[2] = ( data[2] * inverseAlpha + premult[2] ) >> 9;
|
|
}
|
|
}
|
|
|
|
byte mipBlendColors[16][4] = {
|
|
{0,0,0,0},
|
|
{255,0,0,128},
|
|
{0,255,0,128},
|
|
{0,0,255,128},
|
|
{255,0,0,128},
|
|
{0,255,0,128},
|
|
{0,0,255,128},
|
|
{255,0,0,128},
|
|
{0,255,0,128},
|
|
{0,0,255,128},
|
|
{255,0,0,128},
|
|
{0,255,0,128},
|
|
{0,0,255,128},
|
|
{255,0,0,128},
|
|
{0,255,0,128},
|
|
{0,0,255,128},
|
|
};
|
|
|
|
#ifdef HAVE_GLES
|
|
// helper function for GLES format conversions
|
|
byte * gles_convertRGB(byte * data, int width, int height)
|
|
{
|
|
byte * temp = (byte *) R_Malloc (width*height*3, TAG_TEMP_WORKSPACE, qfalse);
|
|
byte *src = data;
|
|
byte *dst = temp;
|
|
|
|
for (int i=0; i<width*height; i++) {
|
|
for (int j=0; j<3; j++)
|
|
*(dst++) = *(src++);
|
|
src++;
|
|
}
|
|
|
|
return temp;
|
|
}
|
|
byte * gles_convertRGBA4(byte * data, int width, int height)
|
|
{
|
|
byte * temp = (byte *) R_Malloc (width*height*2, TAG_TEMP_WORKSPACE, qfalse);
|
|
|
|
unsigned int * input = ( unsigned int *)(data);
|
|
unsigned short* output = (unsigned short*)(temp);
|
|
for (int i = 0; i < width*height; i++) {
|
|
unsigned int pixel = input[i];
|
|
// Unpack the source data as 8 bit values
|
|
unsigned int r = pixel & 0xff;
|
|
unsigned int g = (pixel >> 8) & 0xff;
|
|
unsigned int b = (pixel >> 16) & 0xff;
|
|
unsigned int a = (pixel >> 24) & 0xff;
|
|
// Convert to 4 bit vales
|
|
r >>= 4; g >>= 4; b >>= 4; a >>= 4;
|
|
output[i] = r << 12 | g << 8 | b << 4 | a;
|
|
}
|
|
return temp;
|
|
}
|
|
byte * gles_convertRGB5(byte * data, int width, int height)
|
|
{
|
|
byte * temp = (byte *) R_Malloc (width*height*2, TAG_TEMP_WORKSPACE, qfalse);
|
|
byte *src = data;
|
|
byte *dst = temp;
|
|
byte r,g,b;
|
|
|
|
unsigned int * input = ( unsigned int *)(data);
|
|
unsigned short* output = (unsigned short*)(temp);
|
|
for (int i = 0; i < width*height; i++) {
|
|
unsigned int pixel = input[i];
|
|
// Unpack the source data as 8 bit values
|
|
unsigned int r = pixel & 0xff;
|
|
unsigned int g = (pixel >> 8) & 0xff;
|
|
unsigned int b = (pixel >> 16) & 0xff;
|
|
// Convert to 4 bit vales
|
|
r >>= 3; g >>= 2; b >>= 3;
|
|
output[i] = r << 11 | g << 5 | b;
|
|
}
|
|
return temp;
|
|
}
|
|
byte * gles_convertLuminance(byte * data, int width, int height)
|
|
{
|
|
byte * temp = (byte *) R_Malloc (width*height, TAG_TEMP_WORKSPACE, qfalse);
|
|
byte *src = data;
|
|
byte *dst = temp;
|
|
byte r,g,b;
|
|
int i;
|
|
|
|
unsigned int * input = ( unsigned int *)(data);
|
|
byte* output = (byte*)(temp);
|
|
for (i = 0; i < width*height; i++) {
|
|
unsigned int pixel = input[i];
|
|
// Unpack the source data as 8 bit values
|
|
unsigned int r = pixel & 0xff;
|
|
output[i] = r;
|
|
}
|
|
return temp;
|
|
}
|
|
byte * gles_convertLuminanceAlpha(byte * data, int width, int height)
|
|
{
|
|
byte * temp = (byte *) R_Malloc (width*height*2, TAG_TEMP_WORKSPACE, qfalse);
|
|
byte *src = data;
|
|
byte *dst = temp;
|
|
byte r,g,b;
|
|
int i;
|
|
|
|
unsigned int * input = ( unsigned int *)(data);
|
|
unsigned short* output = (unsigned short*)(temp);
|
|
for (i = 0; i < width*height; i++) {
|
|
unsigned int pixel = input[i];
|
|
// Unpack the source data as 8 bit values
|
|
unsigned int r = pixel & 0xff;
|
|
unsigned int a = (pixel >> 24) & 0xff;
|
|
output[i] = r | a<<8;
|
|
}
|
|
return temp;
|
|
}
|
|
#endif
|
|
|
|
|
|
|
|
class CStringComparator
|
|
{
|
|
public:
|
|
bool operator()(const char *s1, const char *s2) const { return(strcmp(s1, s2) < 0); }
|
|
};
|
|
|
|
typedef map <const char *, image_t *, CStringComparator> AllocatedImages_t;
|
|
AllocatedImages_t AllocatedImages;
|
|
AllocatedImages_t::iterator itAllocatedImages;
|
|
int giTextureBindNum = 1024; // will be set to this anyway at runtime, but wtf?
|
|
|
|
|
|
// return = number of images in the list, for those interested
|
|
//
|
|
int R_Images_StartIteration(void)
|
|
{
|
|
itAllocatedImages = AllocatedImages.begin();
|
|
return AllocatedImages.size();
|
|
}
|
|
|
|
image_t *R_Images_GetNextIteration(void)
|
|
{
|
|
if (itAllocatedImages == AllocatedImages.end())
|
|
return NULL;
|
|
|
|
image_t *pImage = (*itAllocatedImages).second;
|
|
++itAllocatedImages;
|
|
return pImage;
|
|
}
|
|
|
|
// clean up anything to do with an image_t struct, but caller will have to clear the internal to an image_t struct ready for either struct free() or overwrite...
|
|
//
|
|
// (avoid using ri.xxxx stuff here in case running on dedicated)
|
|
//
|
|
static void R_Images_DeleteImageContents( image_t *pImage )
|
|
{
|
|
assert(pImage); // should never be called with NULL
|
|
if (pImage)
|
|
{
|
|
qglDeleteTextures( 1, &pImage->texnum );
|
|
R_Free(pImage);
|
|
}
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
===============
|
|
Upload32
|
|
|
|
===============
|
|
*/
|
|
extern qboolean charSet;
|
|
static void Upload32( unsigned *data,
|
|
GLenum format,
|
|
qboolean mipmap,
|
|
qboolean picmip,
|
|
qboolean isLightmap,
|
|
qboolean allowTC,
|
|
int *pformat,
|
|
word *pUploadWidth, word *pUploadHeight, bool bRectangle = false )
|
|
{
|
|
GLuint uiTarget = GL_TEXTURE_2D;
|
|
#ifndef HAVE_GLES
|
|
if ( bRectangle )
|
|
{
|
|
uiTarget = GL_TEXTURE_RECTANGLE_EXT;
|
|
}
|
|
#else
|
|
//Humm..
|
|
#endif
|
|
if (format == GL_RGBA)
|
|
{
|
|
int samples;
|
|
int i, c;
|
|
byte *scan;
|
|
float rMax = 0, gMax = 0, bMax = 0;
|
|
int width = *pUploadWidth;
|
|
int height = *pUploadHeight;
|
|
|
|
//
|
|
// perform optional picmip operation
|
|
//
|
|
if ( picmip ) {
|
|
for(i = 0; i < r_picmip->integer; i++) {
|
|
|
|
if ((width>16) && (height>16)) {
|
|
R_MipMap( (byte *)data, width, height );
|
|
width >>= 1;
|
|
height >>= 1;
|
|
if (width < 1) {
|
|
width = 1;
|
|
}
|
|
if (height < 1) {
|
|
height = 1;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
//
|
|
// clamp to the current upper OpenGL limit
|
|
// scale both axis down equally so we don't have to
|
|
// deal with a half mip resampling
|
|
//
|
|
while ( width > glConfig.maxTextureSize || height > glConfig.maxTextureSize ) {
|
|
R_MipMap( (byte *)data, width, height );
|
|
width >>= 1;
|
|
height >>= 1;
|
|
}
|
|
|
|
//
|
|
// scan the texture for each channel's max values
|
|
// and verify if the alpha channel is being used or not
|
|
//
|
|
c = width*height;
|
|
scan = ((byte *)data);
|
|
samples = 3;
|
|
for ( i = 0; i < c; i++ )
|
|
{
|
|
if ( scan[i*4+0] > rMax )
|
|
{
|
|
rMax = scan[i*4+0];
|
|
}
|
|
if ( scan[i*4+1] > gMax )
|
|
{
|
|
gMax = scan[i*4+1];
|
|
}
|
|
if ( scan[i*4+2] > bMax )
|
|
{
|
|
bMax = scan[i*4+2];
|
|
}
|
|
if ( scan[i*4 + 3] != 255 )
|
|
{
|
|
samples = 4;
|
|
break;
|
|
}
|
|
}
|
|
|
|
// select proper internal format
|
|
if ( samples == 3 )
|
|
{
|
|
if ( glConfig.textureCompression == TC_S3TC && allowTC )
|
|
{
|
|
#ifdef HAVE_GLES
|
|
assert(0);
|
|
#else
|
|
*pformat = GL_RGB4_S3TC;
|
|
#endif
|
|
}
|
|
else if ( glConfig.textureCompression == TC_S3TC_DXT && allowTC )
|
|
{ // Compress purely color - no alpha
|
|
if ( r_texturebits->integer == 16 ) {
|
|
#ifdef HAVE_GLES
|
|
assert(0);
|
|
#else
|
|
*pformat = GL_COMPRESSED_RGB_S3TC_DXT1_EXT; //this format cuts to 16 bit
|
|
#endif
|
|
}
|
|
else {//if we aren't using 16 bit then, use 32 bit compression
|
|
#ifdef HAVE_GLES
|
|
assert(0);
|
|
#else
|
|
*pformat = GL_COMPRESSED_RGBA_S3TC_DXT5_EXT;
|
|
#endif
|
|
}
|
|
}
|
|
else if ( isLightmap && r_texturebitslm->integer > 0 )
|
|
{
|
|
int lmBits = r_texturebitslm->integer & 0x30; // 16 or 32
|
|
// Allow different bit depth when we are a lightmap
|
|
if ( lmBits == 16 )
|
|
*pformat = GL_RGB5;
|
|
else
|
|
#ifdef HAVE_GLES
|
|
*pformat = GL_RGB;
|
|
#else
|
|
*pformat = GL_RGB8;
|
|
#endif
|
|
}
|
|
else if ( r_texturebits->integer == 16 )
|
|
{
|
|
*pformat = GL_RGB5;
|
|
}
|
|
else if ( r_texturebits->integer == 32 )
|
|
{
|
|
#ifdef HAVE_GLES
|
|
*pformat = GL_RGB;
|
|
#else
|
|
*pformat = GL_RGB8;
|
|
#endif
|
|
}
|
|
else
|
|
{
|
|
#ifdef HAVE_GLES
|
|
*pformat = GL_RGB;
|
|
#else
|
|
*pformat = 3;
|
|
#endif
|
|
}
|
|
}
|
|
else if ( samples == 4 )
|
|
{
|
|
if ( glConfig.textureCompression == TC_S3TC_DXT && allowTC)
|
|
{ // Compress both alpha and color
|
|
#ifdef HAVE_GLES
|
|
assert(0);
|
|
#else
|
|
*pformat = GL_COMPRESSED_RGBA_S3TC_DXT5_EXT;
|
|
#endif
|
|
}
|
|
else if ( r_texturebits->integer == 16 )
|
|
{
|
|
*pformat = GL_RGBA4;
|
|
}
|
|
else if ( r_texturebits->integer == 32 )
|
|
{
|
|
#ifdef HAVE_GLES
|
|
*pformat = GL_RGBA;
|
|
#else
|
|
*pformat = GL_RGBA8;
|
|
#endif
|
|
}
|
|
else
|
|
{
|
|
#ifdef HAVE_GLES
|
|
*pformat = GL_RGBA;
|
|
#else
|
|
*pformat = 4;
|
|
#endif
|
|
}
|
|
}
|
|
|
|
*pUploadWidth = width;
|
|
*pUploadHeight = height;
|
|
|
|
// copy or resample data as appropriate for first MIP level
|
|
#ifdef HAVE_GLES
|
|
#ifdef PANDORA
|
|
// SGX Workaround#2 : if width or height<=16 => make it bigger
|
|
#endif
|
|
//*pformat = GL_RGBA;
|
|
R_LightScaleTexture (data, width, height, mipmap?qfalse:qtrue );
|
|
|
|
glTexParameteri( GL_TEXTURE_2D, GL_GENERATE_MIPMAP, (mipmap)?GL_TRUE:GL_FALSE );
|
|
|
|
// and now, convert if needed and upload
|
|
// GLES doesn't do convertion itself, so we have to handle that
|
|
byte *temp;
|
|
switch ( *pformat ) {
|
|
case GL_RGB5:
|
|
temp = gles_convertRGB5((byte*)data, width, height);
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_RGB, width, height, 0, GL_RGB, GL_UNSIGNED_SHORT_5_6_5, temp);
|
|
R_Free(temp);
|
|
break;
|
|
case GL_RGBA4:
|
|
temp = gles_convertRGBA4((byte*)data, width, height);
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_RGBA, width, height, 0, GL_RGBA, GL_UNSIGNED_SHORT_4_4_4_4, temp);
|
|
R_Free(temp);
|
|
break;
|
|
case 3:
|
|
case GL_RGB:
|
|
temp = gles_convertRGB((byte*)data, width, height);
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_RGB, width, height, 0, GL_RGB, GL_UNSIGNED_BYTE, temp);
|
|
R_Free(temp);
|
|
break;
|
|
case 1:
|
|
temp = gles_convertLuminance((byte*)data, width, height);
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_LUMINANCE, width, height, 0, GL_LUMINANCE, GL_UNSIGNED_BYTE, temp);
|
|
R_Free(temp);
|
|
break;
|
|
case 2:
|
|
temp = gles_convertLuminanceAlpha((byte*)data, width, height);
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_LUMINANCE_ALPHA, width, height, 0, GL_LUMINANCE_ALPHA, GL_UNSIGNED_BYTE, temp);
|
|
R_Free(temp);
|
|
break;
|
|
default:
|
|
*pformat = GL_RGBA;
|
|
qglTexImage2D (GL_TEXTURE_2D, 0, GL_RGBA, width, height, 0, GL_RGBA, GL_UNSIGNED_BYTE, data);
|
|
}
|
|
|
|
|
|
// qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, GL_LINEAR );
|
|
// qglTexParameterf(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, GL_LINEAR );
|
|
#else
|
|
if (!mipmap)
|
|
{
|
|
qglTexImage2D( uiTarget, 0, *pformat, width, height, 0, GL_RGBA, GL_UNSIGNED_BYTE, data );
|
|
goto done;
|
|
}
|
|
|
|
R_LightScaleTexture (data, width, height, (qboolean)!mipmap );
|
|
|
|
qglTexImage2D( uiTarget, 0, *pformat, width, height, 0, GL_RGBA, GL_UNSIGNED_BYTE, data );
|
|
|
|
if (mipmap)
|
|
{
|
|
int miplevel;
|
|
|
|
miplevel = 0;
|
|
while (width > 1 || height > 1)
|
|
{
|
|
R_MipMap( (byte *)data, width, height );
|
|
width >>= 1;
|
|
height >>= 1;
|
|
if (width < 1)
|
|
width = 1;
|
|
if (height < 1)
|
|
height = 1;
|
|
miplevel++;
|
|
|
|
if ( r_colorMipLevels->integer )
|
|
{
|
|
R_BlendOverTexture( (byte *)data, width * height, mipBlendColors[miplevel] );
|
|
}
|
|
|
|
qglTexImage2D( uiTarget, miplevel, *pformat, width, height, 0, GL_RGBA, GL_UNSIGNED_BYTE, data );
|
|
}
|
|
}
|
|
#endif
|
|
}
|
|
else
|
|
{
|
|
}
|
|
|
|
done:
|
|
|
|
if (mipmap)
|
|
{
|
|
qglTexParameterf(uiTarget, GL_TEXTURE_MIN_FILTER, gl_filter_min);
|
|
qglTexParameterf(uiTarget, GL_TEXTURE_MAG_FILTER, gl_filter_max);
|
|
if(r_ext_texture_filter_anisotropic->integer>1 && glConfig.maxTextureFilterAnisotropy>0)
|
|
{
|
|
qglTexParameterf( GL_TEXTURE_2D, GL_TEXTURE_MAX_ANISOTROPY_EXT, r_ext_texture_filter_anisotropic->value );
|
|
}
|
|
}
|
|
else
|
|
{
|
|
qglTexParameterf(uiTarget, GL_TEXTURE_MIN_FILTER, GL_LINEAR );
|
|
qglTexParameterf(uiTarget, GL_TEXTURE_MAG_FILTER, GL_LINEAR );
|
|
}
|
|
|
|
GL_CheckErrors();
|
|
}
|
|
|
|
static void GL_ResetBinds(void)
|
|
{
|
|
memset( glState.currenttextures, 0, sizeof( glState.currenttextures ) );
|
|
if ( qglActiveTextureARB ) {
|
|
GL_SelectTexture( 1 );
|
|
qglBindTexture( GL_TEXTURE_2D, 0 );
|
|
GL_SelectTexture( 0 );
|
|
qglBindTexture( GL_TEXTURE_2D, 0 );
|
|
} else {
|
|
qglBindTexture( GL_TEXTURE_2D, 0 );
|
|
}
|
|
}
|
|
|
|
|
|
// special function used in conjunction with "devmapbsp"...
|
|
//
|
|
// (avoid using ri.xxxx stuff here in case running on dedicated)
|
|
//
|
|
void R_Images_DeleteLightMaps(void)
|
|
{
|
|
for (AllocatedImages_t::iterator itImage = AllocatedImages.begin(); itImage != AllocatedImages.end(); /* empty */)
|
|
{
|
|
image_t *pImage = (*itImage).second;
|
|
|
|
if (pImage->imgName[0] == '*' && strstr(pImage->imgName,"lightmap")) // loose check, but should be ok
|
|
{
|
|
R_Images_DeleteImageContents(pImage);
|
|
|
|
AllocatedImages.erase(itImage++);
|
|
}
|
|
else
|
|
{
|
|
++itImage;
|
|
}
|
|
}
|
|
|
|
GL_ResetBinds();
|
|
}
|
|
|
|
// special function currently only called by Dissolve code...
|
|
//
|
|
void R_Images_DeleteImage(image_t *pImage)
|
|
{
|
|
// Even though we supply the image handle, we need to get the corresponding iterator entry...
|
|
//
|
|
AllocatedImages_t::iterator itImage = AllocatedImages.find(pImage->imgName);
|
|
if (itImage != AllocatedImages.end())
|
|
{
|
|
R_Images_DeleteImageContents(pImage);
|
|
AllocatedImages.erase(itImage);
|
|
}
|
|
else
|
|
{
|
|
assert(0);
|
|
}
|
|
}
|
|
|
|
// called only at app startup, vid_restart, app-exit
|
|
//
|
|
void R_Images_Clear(void)
|
|
{
|
|
image_t *pImage;
|
|
// int iNumImages =
|
|
R_Images_StartIteration();
|
|
while ( (pImage = R_Images_GetNextIteration()) != NULL)
|
|
{
|
|
R_Images_DeleteImageContents(pImage);
|
|
}
|
|
|
|
AllocatedImages.clear();
|
|
|
|
giTextureBindNum = 1024;
|
|
}
|
|
|
|
|
|
void RE_RegisterImages_Info_f( void )
|
|
{
|
|
image_t *pImage = NULL;
|
|
int iImage = 0;
|
|
int iTexels = 0;
|
|
|
|
int iNumImages = R_Images_StartIteration();
|
|
while ( (pImage = R_Images_GetNextIteration()) != NULL)
|
|
{
|
|
ri.Printf( PRINT_ALL, "%d: (%4dx%4dy) \"%s\"",iImage, pImage->width, pImage->height, pImage->imgName);
|
|
ri.Printf( PRINT_DEVELOPER, S_COLOR_RED ", levused %d",pImage->iLastLevelUsedOn);
|
|
ri.Printf( PRINT_ALL, "\n");
|
|
|
|
iTexels += pImage->width * pImage->height;
|
|
iImage++;
|
|
}
|
|
ri.Printf( PRINT_ALL, "%d Images. %d (%.2fMB) texels total, (not including mipmaps)\n",iNumImages, iTexels, (float)iTexels / 1024.0f / 1024.0f);
|
|
ri.Printf( PRINT_DEVELOPER, S_COLOR_RED "RE_RegisterMedia_GetLevel(): %d",RE_RegisterMedia_GetLevel());
|
|
}
|
|
|
|
// currently, this just goes through all the images and dumps any not referenced on this level...
|
|
//
|
|
qboolean RE_RegisterImages_LevelLoadEnd(void)
|
|
{
|
|
ri.Printf( PRINT_DEVELOPER, S_COLOR_RED "RE_RegisterImages_LevelLoadEnd():\n");
|
|
|
|
// int iNumImages = AllocatedImages.size(); // more for curiosity, really.
|
|
|
|
qboolean imageDeleted = qtrue;
|
|
for (AllocatedImages_t::iterator itImage = AllocatedImages.begin(); itImage != AllocatedImages.end(); /* blank */)
|
|
{
|
|
qboolean bEraseOccured = qfalse;
|
|
|
|
image_t *pImage = (*itImage).second;
|
|
|
|
// don't un-register system shaders (*fog, *dlight, *white, *default), but DO de-register lightmaps ("*<mapname>/lightmap%d")
|
|
if (pImage->imgName[0] != '*' || strchr(pImage->imgName,'/'))
|
|
{
|
|
// image used on this level?
|
|
//
|
|
if ( pImage->iLastLevelUsedOn != RE_RegisterMedia_GetLevel() )
|
|
{
|
|
// nope, so dump it...
|
|
//
|
|
ri.Printf( PRINT_DEVELOPER, S_COLOR_RED "Dumping image \"%s\"\n",pImage->imgName);
|
|
|
|
R_Images_DeleteImageContents(pImage);
|
|
|
|
AllocatedImages.erase(itImage++);
|
|
bEraseOccured = qtrue;
|
|
imageDeleted = qtrue;
|
|
}
|
|
}
|
|
|
|
if ( !bEraseOccured )
|
|
{
|
|
++itImage;
|
|
}
|
|
}
|
|
|
|
|
|
// this check can be deleted AFAIC, it seems to be just a quake thing...
|
|
//
|
|
// iNumImages = R_Images_StartIteration();
|
|
// if (iNumImages > MAX_DRAWIMAGES)
|
|
// {
|
|
// ri.Printf( PRINT_ALL, S_COLOR_YELLOW "Level uses %d images, old limit was MAX_DRAWIMAGES (%d)\n", iNumImages, MAX_DRAWIMAGES);
|
|
// }
|
|
|
|
ri.Printf( PRINT_DEVELOPER, S_COLOR_RED "RE_RegisterImages_LevelLoadEnd(): Ok\n");
|
|
|
|
GL_ResetBinds();
|
|
|
|
return imageDeleted;
|
|
}
|
|
|
|
|
|
|
|
// returns image_t struct if we already have this, else NULL. No disk-open performed
|
|
// (important for creating default images).
|
|
//
|
|
// This is called by both R_FindImageFile and anything that creates default images...
|
|
//
|
|
static image_t *R_FindImageFile_NoLoad(const char *name, qboolean mipmap, qboolean allowPicmip, qboolean allowTC, int glWrapClampMode )
|
|
{
|
|
if (!name) {
|
|
return NULL;
|
|
}
|
|
|
|
char *pName = GenerateImageMappingName(name);
|
|
|
|
//
|
|
// see if the image is already loaded
|
|
//
|
|
AllocatedImages_t::iterator itAllocatedImage = AllocatedImages.find(pName);
|
|
if (itAllocatedImage != AllocatedImages.end())
|
|
{
|
|
image_t *pImage = (*itAllocatedImage).second;
|
|
|
|
// the white image can be used with any set of parms, but other mismatches are errors...
|
|
//
|
|
if ( strcmp( pName, "*white" ) ) {
|
|
if ( pImage->mipmap != !!mipmap ) {
|
|
ri.Printf( PRINT_ALL, S_COLOR_YELLOW "WARNING: reused image %s with mixed mipmap parm\n", pName );
|
|
}
|
|
if ( pImage->allowPicmip != !!allowPicmip ) {
|
|
ri.Printf( PRINT_ALL, S_COLOR_YELLOW "WARNING: reused image %s with mixed allowPicmip parm\n", pName );
|
|
}
|
|
if ( pImage->wrapClampMode != glWrapClampMode ) {
|
|
ri.Printf( PRINT_ALL, S_COLOR_YELLOW "WARNING: reused image %s with mixed glWrapClampMode parm\n", pName );
|
|
}
|
|
}
|
|
|
|
pImage->iLastLevelUsedOn = RE_RegisterMedia_GetLevel();
|
|
|
|
return pImage;
|
|
}
|
|
|
|
return NULL;
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
================
|
|
R_CreateImage
|
|
|
|
This is the only way any image_t are created
|
|
================
|
|
*/
|
|
image_t *R_CreateImage( const char *name, const byte *pic, int width, int height,
|
|
GLenum format, qboolean mipmap, qboolean allowPicmip, qboolean allowTC, int glWrapClampMode, bool bRectangle )
|
|
{
|
|
image_t *image;
|
|
qboolean isLightmap = qfalse;
|
|
|
|
if (strlen(name) >= MAX_QPATH ) {
|
|
Com_Error (ERR_DROP, "R_CreateImage: \"%s\" is too long\n", name);
|
|
}
|
|
|
|
if(glConfig.clampToEdgeAvailable && glWrapClampMode == GL_CLAMP) {
|
|
glWrapClampMode = GL_CLAMP_TO_EDGE;
|
|
}
|
|
|
|
if (name[0] == '*')
|
|
{
|
|
const char *psLightMapNameSearchPos = strrchr(name,'/');
|
|
if ( psLightMapNameSearchPos && !strncmp( psLightMapNameSearchPos+1, "lightmap", 8 ) ) {
|
|
isLightmap = qtrue;
|
|
}
|
|
}
|
|
|
|
if ( (width&(width-1)) || (height&(height-1)) )
|
|
{
|
|
Com_Error( ERR_FATAL, "R_CreateImage: %s dimensions (%i x %i) not power of 2!\n",name,width,height);
|
|
}
|
|
|
|
image = R_FindImageFile_NoLoad(name, mipmap, allowPicmip, allowTC, glWrapClampMode );
|
|
if (image) {
|
|
return image;
|
|
}
|
|
|
|
image = (image_t*) R_Malloc( sizeof( image_t ), TAG_IMAGE_T, qtrue );
|
|
// memset(image,0,sizeof(*image)); // qtrue above does this
|
|
|
|
image->texnum = 1024 + giTextureBindNum++; // ++ is of course staggeringly important...
|
|
|
|
// record which map it was used on...
|
|
//
|
|
image->iLastLevelUsedOn = RE_RegisterMedia_GetLevel();
|
|
|
|
image->mipmap = !!mipmap;
|
|
image->allowPicmip = !!allowPicmip;
|
|
|
|
Q_strncpyz(image->imgName, name, sizeof(image->imgName));
|
|
|
|
image->width = width;
|
|
image->height = height;
|
|
image->wrapClampMode = glWrapClampMode;
|
|
|
|
if ( qglActiveTextureARB ) {
|
|
GL_SelectTexture( 0 );
|
|
}
|
|
|
|
GLuint uiTarget = GL_TEXTURE_2D;
|
|
#ifndef HAVE_GLES
|
|
if ( bRectangle )
|
|
{
|
|
qglDisable( uiTarget );
|
|
uiTarget = GL_TEXTURE_RECTANGLE_EXT;
|
|
qglEnable( uiTarget );
|
|
glWrapClampMode = GL_CLAMP_TO_EDGE; // default mode supported by rectangle.
|
|
qglBindTexture( uiTarget, image->texnum );
|
|
}
|
|
else
|
|
#endif
|
|
{
|
|
GL_Bind(image);
|
|
}
|
|
|
|
Upload32( (unsigned *)pic, format,
|
|
(qboolean)image->mipmap,
|
|
allowPicmip,
|
|
isLightmap,
|
|
allowTC,
|
|
&image->internalFormat,
|
|
&image->width,
|
|
&image->height, bRectangle );
|
|
|
|
qglTexParameterf( uiTarget, GL_TEXTURE_WRAP_S, glWrapClampMode );
|
|
qglTexParameterf( uiTarget, GL_TEXTURE_WRAP_T, glWrapClampMode );
|
|
|
|
qglBindTexture( uiTarget, 0 ); //jfm: i don't know why this is here, but it breaks lightmaps when there's only 1
|
|
glState.currenttextures[glState.currenttmu] = 0; //mark it not bound
|
|
|
|
const char *psNewName = GenerateImageMappingName(name);
|
|
Q_strncpyz(image->imgName, psNewName, sizeof(image->imgName));
|
|
AllocatedImages[ image->imgName ] = image;
|
|
|
|
if ( bRectangle )
|
|
{
|
|
qglDisable( uiTarget );
|
|
qglEnable( GL_TEXTURE_2D );
|
|
}
|
|
|
|
return image;
|
|
}
|
|
|
|
//rwwRMG - added
|
|
void R_CreateAutomapImage( const char *name, const byte *pic, int width, int height, qboolean mipmap, qboolean allowPicmip, qboolean allowTC, int glWrapClampMode )
|
|
{
|
|
R_CreateImage(name, pic, width, height, GL_RGBA, mipmap, allowPicmip, allowTC, glWrapClampMode);
|
|
}
|
|
|
|
/*
|
|
===============
|
|
R_FindImageFile
|
|
|
|
Finds or loads the given image.
|
|
Returns NULL if it fails, not a default image.
|
|
==============
|
|
*/
|
|
image_t *R_FindImageFile( const char *name, qboolean mipmap, qboolean allowPicmip, qboolean allowTC, int glWrapClampMode ) {
|
|
image_t *image;
|
|
int width, height;
|
|
byte *pic;
|
|
|
|
if (!name || ri.Cvar_VariableIntegerValue( "dedicated" ) ) // stop ghoul2 horribleness as regards image loading from server
|
|
{
|
|
return NULL;
|
|
}
|
|
|
|
// need to do this here as well as in R_CreateImage, or R_FindImageFile_NoLoad() may complain about
|
|
// different clamp parms used...
|
|
//
|
|
if(glConfig.clampToEdgeAvailable && glWrapClampMode == GL_CLAMP) {
|
|
glWrapClampMode = GL_CLAMP_TO_EDGE;
|
|
}
|
|
|
|
image = R_FindImageFile_NoLoad(name, mipmap, allowPicmip, allowTC, glWrapClampMode );
|
|
if (image) {
|
|
return image;
|
|
}
|
|
|
|
//
|
|
// load the pic from disk
|
|
//
|
|
R_LoadImage( name, &pic, &width, &height );
|
|
if ( pic == NULL ) { // if we dont get a successful load
|
|
return NULL; // bail
|
|
}
|
|
|
|
|
|
// refuse to find any files not power of 2 dims...
|
|
//
|
|
if ( (width&(width-1)) || (height&(height-1)) )
|
|
{
|
|
ri.Printf( PRINT_ALL, "Refusing to load non-power-2-dims(%d,%d) pic \"%s\"...\n", width,height,name );
|
|
return NULL;
|
|
}
|
|
|
|
image = R_CreateImage( ( char * ) name, pic, width, height, GL_RGBA, mipmap, allowPicmip, allowTC, glWrapClampMode );
|
|
R_Free( pic );
|
|
return image;
|
|
}
|
|
|
|
|
|
/*
|
|
================
|
|
R_CreateDlightImage
|
|
================
|
|
*/
|
|
#define DLIGHT_SIZE 16
|
|
static void R_CreateDlightImage( void )
|
|
{
|
|
int width, height;
|
|
byte *pic;
|
|
|
|
R_LoadImage("gfx/2d/dlight", &pic, &width, &height);
|
|
if (pic)
|
|
{
|
|
tr.dlightImage = R_CreateImage("*dlight", pic, width, height, GL_RGBA, qfalse, qfalse, qfalse, GL_CLAMP );
|
|
R_Free(pic);
|
|
}
|
|
else
|
|
{ // if we dont get a successful load
|
|
int x,y;
|
|
byte data[DLIGHT_SIZE][DLIGHT_SIZE][4];
|
|
int b;
|
|
|
|
// make a centered inverse-square falloff blob for dynamic lighting
|
|
for (x=0 ; x<DLIGHT_SIZE ; x++) {
|
|
for (y=0 ; y<DLIGHT_SIZE ; y++) {
|
|
float d;
|
|
|
|
d = ( DLIGHT_SIZE/2 - 0.5f - x ) * ( DLIGHT_SIZE/2 - 0.5f - x ) +
|
|
( DLIGHT_SIZE/2 - 0.5f - y ) * ( DLIGHT_SIZE/2 - 0.5f - y );
|
|
b = 4000 / d;
|
|
if (b > 255) {
|
|
b = 255;
|
|
} else if ( b < 75 ) {
|
|
b = 0;
|
|
}
|
|
data[y][x][0] =
|
|
data[y][x][1] =
|
|
data[y][x][2] = b;
|
|
data[y][x][3] = 255;
|
|
}
|
|
}
|
|
tr.dlightImage = R_CreateImage("*dlight", (byte *)data, DLIGHT_SIZE, DLIGHT_SIZE, GL_RGBA, qfalse, qfalse, qfalse, GL_CLAMP );
|
|
}
|
|
}
|
|
|
|
|
|
/*
|
|
=================
|
|
R_InitFogTable
|
|
=================
|
|
*/
|
|
void R_InitFogTable( void ) {
|
|
int i;
|
|
float d;
|
|
float exp;
|
|
|
|
exp = 0.5;
|
|
|
|
for ( i = 0 ; i < FOG_TABLE_SIZE ; i++ ) {
|
|
d = pow ( (float)i/(FOG_TABLE_SIZE-1), exp );
|
|
|
|
tr.fogTable[i] = d;
|
|
}
|
|
}
|
|
|
|
/*
|
|
================
|
|
R_FogFactor
|
|
|
|
Returns a 0.0 to 1.0 fog density value
|
|
This is called for each texel of the fog texture on startup
|
|
and for each vertex of transparent shaders in fog dynamically
|
|
================
|
|
*/
|
|
float R_FogFactor( float s, float t ) {
|
|
float d;
|
|
|
|
s -= 1.0/512;
|
|
if ( s < 0 ) {
|
|
return 0;
|
|
}
|
|
if ( t < 1.0/32 ) {
|
|
return 0;
|
|
}
|
|
if ( t < 31.0/32 ) {
|
|
s *= (t - 1.0f/32.0f) / (30.0f/32.0f);
|
|
}
|
|
|
|
// we need to leave a lot of clamp range
|
|
s *= 8;
|
|
|
|
if ( s > 1.0 ) {
|
|
s = 1.0;
|
|
}
|
|
|
|
d = tr.fogTable[ (int)(s * (FOG_TABLE_SIZE-1)) ];
|
|
|
|
return d;
|
|
}
|
|
|
|
/*
|
|
================
|
|
R_CreateFogImage
|
|
================
|
|
*/
|
|
#define FOG_S 256
|
|
#define FOG_T 32
|
|
static void R_CreateFogImage( void ) {
|
|
int x,y;
|
|
byte *data;
|
|
float d;
|
|
float borderColor[4];
|
|
|
|
data = (unsigned char *)R_Malloc( FOG_S * FOG_T * 4, TAG_TEMP_WORKSPACE, qfalse );
|
|
|
|
// S is distance, T is depth
|
|
for (x=0 ; x<FOG_S ; x++) {
|
|
for (y=0 ; y<FOG_T ; y++) {
|
|
d = R_FogFactor( ( x + 0.5f ) / FOG_S, ( y + 0.5f ) / FOG_T );
|
|
|
|
data[(y*FOG_S+x)*4+0] =
|
|
data[(y*FOG_S+x)*4+1] =
|
|
data[(y*FOG_S+x)*4+2] = 255;
|
|
data[(y*FOG_S+x)*4+3] = 255*d;
|
|
}
|
|
}
|
|
// standard openGL clamping doesn't really do what we want -- it includes
|
|
// the border color at the edges. OpenGL 1.2 has clamp-to-edge, which does
|
|
// what we want.
|
|
tr.fogImage = R_CreateImage("*fog", (byte *)data, FOG_S, FOG_T, GL_RGBA, qfalse, qfalse, qfalse, GL_CLAMP );
|
|
R_Free( data );
|
|
|
|
borderColor[0] = 1.0;
|
|
borderColor[1] = 1.0;
|
|
borderColor[2] = 1.0;
|
|
borderColor[3] = 1;
|
|
|
|
#ifndef HAVE_GLES
|
|
qglTexParameterfv( GL_TEXTURE_2D, GL_TEXTURE_BORDER_COLOR, borderColor );
|
|
#endif
|
|
}
|
|
|
|
/*
|
|
==================
|
|
R_CreateDefaultImage
|
|
==================
|
|
*/
|
|
#define DEFAULT_SIZE 16
|
|
static void R_CreateDefaultImage( void ) {
|
|
int x;
|
|
byte data[DEFAULT_SIZE][DEFAULT_SIZE][4];
|
|
|
|
// the default image will be a box, to allow you to see the mapping coordinates
|
|
memset( data, 32, sizeof( data ) );
|
|
for ( x = 0 ; x < DEFAULT_SIZE ; x++ ) {
|
|
data[0][x][0] =
|
|
data[0][x][1] =
|
|
data[0][x][2] =
|
|
data[0][x][3] = 255;
|
|
|
|
data[x][0][0] =
|
|
data[x][0][1] =
|
|
data[x][0][2] =
|
|
data[x][0][3] = 255;
|
|
|
|
data[DEFAULT_SIZE-1][x][0] =
|
|
data[DEFAULT_SIZE-1][x][1] =
|
|
data[DEFAULT_SIZE-1][x][2] =
|
|
data[DEFAULT_SIZE-1][x][3] = 255;
|
|
|
|
data[x][DEFAULT_SIZE-1][0] =
|
|
data[x][DEFAULT_SIZE-1][1] =
|
|
data[x][DEFAULT_SIZE-1][2] =
|
|
data[x][DEFAULT_SIZE-1][3] = 255;
|
|
}
|
|
tr.defaultImage = R_CreateImage("*default", (byte *)data, DEFAULT_SIZE, DEFAULT_SIZE, GL_RGBA, qtrue, qfalse, qfalse, GL_REPEAT );
|
|
}
|
|
|
|
/*
|
|
==================
|
|
R_CreateBuiltinImages
|
|
==================
|
|
*/
|
|
void R_CreateBuiltinImages( void ) {
|
|
int x,y;
|
|
byte data[DEFAULT_SIZE][DEFAULT_SIZE][4];
|
|
|
|
R_CreateDefaultImage();
|
|
|
|
// we use a solid white image instead of disabling texturing
|
|
memset( data, 255, sizeof( data ) );
|
|
tr.whiteImage = R_CreateImage("*white", (byte *)data, 8, 8, GL_RGBA, qfalse, qfalse, qfalse, GL_REPEAT);
|
|
|
|
tr.screenImage = R_CreateImage("*screen", (byte *)data, 8, 8, GL_RGBA, qfalse, qfalse, qfalse, GL_REPEAT );
|
|
|
|
|
|
#ifndef HAVE_GLES
|
|
// Create the scene glow image. - AReis
|
|
tr.screenGlow = 1024 + giTextureBindNum++;
|
|
qglDisable( GL_TEXTURE_2D );
|
|
qglEnable( GL_TEXTURE_RECTANGLE_EXT );
|
|
qglBindTexture( GL_TEXTURE_RECTANGLE_EXT, tr.screenGlow );
|
|
qglTexImage2D( GL_TEXTURE_RECTANGLE_EXT, 0, GL_RGBA16, glConfig.vidWidth, glConfig.vidHeight, 0, GL_RGB, GL_FLOAT, 0 );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MIN_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MAG_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_S, GL_CLAMP );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_T, GL_CLAMP );
|
|
|
|
// Create the scene image. - AReis
|
|
tr.sceneImage = 1024 + giTextureBindNum++;
|
|
qglBindTexture( GL_TEXTURE_RECTANGLE_EXT, tr.sceneImage );
|
|
qglTexImage2D( GL_TEXTURE_RECTANGLE_EXT, 0, GL_RGBA16, glConfig.vidWidth, glConfig.vidHeight, 0, GL_RGB, GL_FLOAT, 0 );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MIN_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MAG_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_S, GL_CLAMP );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_T, GL_CLAMP );
|
|
|
|
// Create the minimized scene blur image.
|
|
if ( r_DynamicGlowWidth->integer > glConfig.vidWidth )
|
|
{
|
|
r_DynamicGlowWidth->integer = glConfig.vidWidth;
|
|
}
|
|
if ( r_DynamicGlowHeight->integer > glConfig.vidHeight )
|
|
{
|
|
r_DynamicGlowHeight->integer = glConfig.vidHeight;
|
|
}
|
|
tr.blurImage = 1024 + giTextureBindNum++;
|
|
qglBindTexture( GL_TEXTURE_RECTANGLE_EXT, tr.blurImage );
|
|
qglTexImage2D( GL_TEXTURE_RECTANGLE_EXT, 0, GL_RGBA16, r_DynamicGlowWidth->integer, r_DynamicGlowHeight->integer, 0, GL_RGB, GL_FLOAT, 0 );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MIN_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_MAG_FILTER, GL_LINEAR );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_S, GL_CLAMP );
|
|
qglTexParameteri( GL_TEXTURE_RECTANGLE_EXT, GL_TEXTURE_WRAP_T, GL_CLAMP );
|
|
qglDisable( GL_TEXTURE_RECTANGLE_EXT );
|
|
qglEnable( GL_TEXTURE_2D );
|
|
|
|
|
|
// with overbright bits active, we need an image which is some fraction of full color,
|
|
// for default lightmaps, etc
|
|
for (x=0 ; x<DEFAULT_SIZE ; x++) {
|
|
for (y=0 ; y<DEFAULT_SIZE ; y++) {
|
|
data[y][x][0] =
|
|
data[y][x][1] =
|
|
data[y][x][2] = tr.identityLightByte;
|
|
data[y][x][3] = 255;
|
|
}
|
|
}
|
|
|
|
#endif // HAVE_GLES
|
|
|
|
tr.identityLightImage = R_CreateImage("*identityLight", (byte *)data, 8, 8, GL_RGBA, qfalse, qfalse, qfalse, GL_REPEAT);
|
|
|
|
for(x=0;x<NUM_SCRATCH_IMAGES;x++) {
|
|
// scratchimage is usually used for cinematic drawing
|
|
tr.scratchImage[x] = R_CreateImage(va("*scratch%d",x), (byte *)data, DEFAULT_SIZE, DEFAULT_SIZE, GL_RGBA, qfalse, qtrue, qfalse, GL_CLAMP);
|
|
}
|
|
|
|
R_CreateDlightImage();
|
|
|
|
R_CreateFogImage();
|
|
}
|
|
|
|
|
|
/*
|
|
===============
|
|
R_SetColorMappings
|
|
===============
|
|
*/
|
|
void R_SetColorMappings( void ) {
|
|
int i, j;
|
|
float g;
|
|
int inf;
|
|
int shift;
|
|
|
|
// setup the overbright lighting
|
|
tr.overbrightBits = r_overBrightBits->integer;
|
|
if ( !glConfig.deviceSupportsGamma ) {
|
|
tr.overbrightBits = 0; // need hardware gamma for overbright
|
|
}
|
|
|
|
// never overbright in windowed mode
|
|
if ( !glConfig.isFullscreen )
|
|
{
|
|
tr.overbrightBits = 0;
|
|
}
|
|
|
|
if ( tr.overbrightBits > 1 ) {
|
|
tr.overbrightBits = 1;
|
|
}
|
|
|
|
if ( tr.overbrightBits < 0 ) {
|
|
tr.overbrightBits = 0;
|
|
}
|
|
|
|
tr.identityLight = 1.0f / ( 1 << tr.overbrightBits );
|
|
tr.identityLightByte = 255 * tr.identityLight;
|
|
|
|
|
|
if ( r_intensity->value < 1.0f ) {
|
|
ri.Cvar_Set( "r_intensity", "1" );
|
|
}
|
|
|
|
if ( r_gamma->value < 0.5f ) {
|
|
ri.Cvar_Set( "r_gamma", "0.5" );
|
|
} else if ( r_gamma->value > 3.0f ) {
|
|
ri.Cvar_Set( "r_gamma", "3.0" );
|
|
}
|
|
|
|
g = r_gamma->value;
|
|
|
|
shift = tr.overbrightBits;
|
|
|
|
for ( i = 0; i < 256; i++ ) {
|
|
if ( g == 1 ) {
|
|
inf = i;
|
|
} else {
|
|
inf = 255 * pow ( i/255.0f, 1.0f / g ) + 0.5f;
|
|
}
|
|
inf <<= shift;
|
|
if (inf < 0) {
|
|
inf = 0;
|
|
}
|
|
if (inf > 255) {
|
|
inf = 255;
|
|
}
|
|
s_gammatable[i] = inf;
|
|
}
|
|
|
|
for (i=0 ; i<256 ; i++) {
|
|
j = i * r_intensity->value;
|
|
if (j > 255) {
|
|
j = 255;
|
|
}
|
|
s_intensitytable[i] = j;
|
|
}
|
|
|
|
if ( glConfig.deviceSupportsGamma )
|
|
{
|
|
ri.WIN_SetGamma( &glConfig, s_gammatable, s_gammatable, s_gammatable );
|
|
}
|
|
}
|
|
|
|
/*
|
|
===============
|
|
R_InitImages
|
|
===============
|
|
*/
|
|
void R_InitImages( void ) {
|
|
//memset(hashTable, 0, sizeof(hashTable)); // DO NOT DO THIS NOW (because of image cacheing) -ste.
|
|
// build brightness translation tables
|
|
R_SetColorMappings();
|
|
|
|
// create default texture and white texture
|
|
R_CreateBuiltinImages();
|
|
}
|
|
|
|
/*
|
|
===============
|
|
R_DeleteTextures
|
|
===============
|
|
*/
|
|
// (only gets called during vid_restart now (and app exit), not during map load)
|
|
//
|
|
void R_DeleteTextures( void ) {
|
|
|
|
R_Images_Clear();
|
|
GL_ResetBinds();
|
|
}
|
|
|