mupen64plus-oldsvn/z64/rgl_geometry.cpp

582 lines
17 KiB
C++

/*
* z64
*
* This program is free software; you can redistribute it and/
* or modify it under the terms of the GNU General Public Li-
* cence as published by the Free Software Foundation; either
* version 2 of the Licence, or any later version.
*
* This program is distributed in the hope that it will be use-
* ful, but WITHOUT ANY WARRANTY; without even the implied war-
* ranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
* See the GNU General Public Licence for more details.
*
* You should have received a copy of the GNU General Public
* Licence along with this program; if not, write to the Free
* Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139,
* USA.
*
**/
#include "Gfx #1.3.h"
#include "rdp.h"
#include "rgl.h"
#include <SDL/SDL.h>
inline float _zscale(uint16_t z)
{
uint32_t res;
int e = z>>16-3;
int m = (z>>2)&((1<<11)-1);
static struct {
int shift;
long add;
} z_format[8] = {
6, 0x00000,
5, 0x20000,
4, 0x30000,
3, 0x38000,
2, 0x3c000,
1, 0x3e000,
0, 0x3f000,
0, 0x3f800,
};
res = (m << z_format[e].shift) +
z_format[e].add;
return float(res)/0x3ffff;
}
inline float zscale(uint16_t z)
{
return float(z)/0xffff;
}
//#define zscale _zscale
float rglZscale(uint16_t z)
{
return _zscale(z);
}
void rglTextureRectangle(rdpTexRect_t * rect, int flip)
{
int tilenum = rect->tilenum;
int x1,x2,y1,y2,z;
int s, t;
int dx, dy;
// if (tilenum == 7) {
// LOG("Fixing tilenum from 7 to 0\n");
// tilenum = 0;
// }
x1 = (rect->xh);
x2 = (rect->xl);
y1 = (rect->yh);
y2 = (rect->yl);
s = int(rect->s)<<5;
t = int(rect->t)<<5;
DUMP("texrect %d x %d --> %d x %d s %d t %d flip %d\n",
x1, y1, x2, y2, s, t, flip);
if (RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) == RDP_CYCLE_TYPE_FILL ||
RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) == RDP_CYCLE_TYPE_COPY)
{
rect->dsdx /= 4;
//s /= 4;
x2 += 4;
y2 += 4;
} else {
x2 += 1;
y2 += 1;
}
x1 /= 4;
x2 /= 4;
y1 /= 4;
y2 /= 4;
if (x2 < x1) x2 = x1+1; // black gauge in SCARS (E)
int t1 = rglT1Usage(rdpState)? RGL_STRIP_TEX1:0;
int t2 = (rect->tilenum < 7 && rglT2Usage(rdpState))? RGL_STRIP_TEX2:0;
if (t1)
rglPrepareRendering(1, (tilenum==7 && RDP_GETOM_CYCLE_TYPE(rdpState.otherModes)==1)? 0:tilenum, y2-y1, 1);
if (t2)
rglPrepareRendering(1, tilenum+1, y2-y1, 1);
else if (!t1)
rglPrepareRendering(0, 0, 0, 1);
// TO BE REMOVED when we implement depth texture writing
curRBuffer->flags |= RGL_RB_HASTRIANGLES;
// TO CHECK should this before or after the rescaling above ?
// s -= (rdpTiles[tilenum].sl << 8);
// t -= (rdpTiles[tilenum].tl << 8);
// if (/*!tile.ms && */tile.mask_s)
// s &= (1<<tile.mask_s+10) - 1;
// if (/*!tile.mt && */tile.mask_t)
// t &= (1<<tile.mask_t+10) - 1;
#define XSCALE(x) (float(x))
#define YSCALE(y) (float(y))
#define ZSCALE(z) (zscale(z))
#define SSCALE(s) (float(s)/(1 << 10))
#define TSCALE(s) (float(s)/(1 << 10))
// #define glTexCoord2f(s, t) printf("tex %g %g\n", s, t), glTexCoord2f(s, t)
// #define glVertex3f(s, t, z) printf("vert %g %g %g\n", s, t, z), glVertex3f(s, t, z)
dx = x2 - x1;
dy = y2 - y1;
if (RDP_GETOM_Z_SOURCE_SEL(rdpState.otherModes))
z = rdpState.primitiveZ;
else
z = 0xffff;
// if (dump)
// fprintf(stderr, "fillrect cycle %d\n", other_modes.cycle_type);
rglStrip_t * strip = strips + nbStrips++;
rglAssert(nbStrips < MAX_STRIPS);
curChunk->nbStrips++;
rglVertex_t * vtx = vtxs + nbVtxs;
strip->flags = t1 | t2 | RGL_STRIP_ZBUFFER;
strip->vtxs = vtx;
strip->tilenum = tilenum;
float s2, tr;
s2 = s+int(rect->dsdx)*dx;
tr = t+int(rect->dtdy)*dy;
//LOG("%d %d\n", rect->dsdx, rect->dtdy);
if (0 && RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) < 2)
{
//if (rect->dsdx == (1<<10))
{
s += 1<<9;
s2 -= 1<<9;
}
//if (rect->dtdy == (1<<10))
{
t += 1<<9;
tr -= 1<<9;
}
}
if (flip) { vtx->t = SSCALE(s2); vtx->s = TSCALE(t);
} else { vtx->s = SSCALE(s2); vtx->t = TSCALE(t); }
vtx->x = XSCALE(x2); vtx->y = YSCALE(y1); vtx->z = ZSCALE(z); vtx++->w = 1;
if (flip) { vtx->t = SSCALE(s); vtx->s = TSCALE(t);
} else { vtx->s = SSCALE(s); vtx->t = TSCALE(t); }
vtx->x = XSCALE(x1); vtx->y = YSCALE(y1); vtx->z = ZSCALE(z); vtx++->w = 1;
if (flip) { vtx->t = SSCALE(s2); vtx->s = TSCALE(tr);
} else { vtx->s = SSCALE(s2); vtx->t = TSCALE(tr); }
vtx->x = XSCALE(x2); vtx->y = YSCALE(y2); vtx->z = ZSCALE(z); vtx++->w = 1;
if (flip) { vtx->t = SSCALE(s); vtx->s = TSCALE(tr);
} else { vtx->s = SSCALE(s); vtx->t = TSCALE(tr); }
vtx->x = XSCALE(x1); vtx->y = YSCALE(y2); vtx->z = ZSCALE(z); vtx++->w = 1;
strip->nbVtxs = vtx - strip->vtxs;
nbVtxs = vtx - vtxs;
}
void rglFillRectangle(rdpRect_t * rect)
{
int x1,x2,y1,y2,z;
int s, t;
int dx, dy;
rglPrepareRendering(0, 0, 0, 1);
DUMP("fillrect curRBuffer->flags %x %x %x\n", curRBuffer->flags, curRBuffer->addressStart, rdpZbAddress);
// if (/*(curRBuffer->flags & RGL_RB_DEPTH) &&*/
// RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) == RDP_CYCLE_TYPE_FILL &&
// rect->xh-4 <= rdpState.clip.xh && rect->xl+8 >= rdpState.clip.xl &&
// rect->yh-4 <= rdpState.clip.yh && rect->yl+8 >= rdpState.clip.yl
// ) {
// curChunk->flags |= RGL_CHUNK_CLEAR;
// return;
// }
x1 = (rect->xh / 4);
x2 = (rect->xl / 4);
y1 = (rect->yh / 4);
y2 = (rect->yl / 4);
if (RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) == RDP_CYCLE_TYPE_FILL ||
RDP_GETOM_CYCLE_TYPE(rdpState.otherModes) == RDP_CYCLE_TYPE_COPY)
{
x2 += 1;
y2 += 1;
} else {
//rglAssert(!(curRBuffer->flags & RGL_RB_DEPTH));
// x2 -= 1;
// y2 -= 1;
}
if (x2 < x1) x2 = x1+1; // black gauge in SCARS (E)
#define XSCALE(x) (float(x))
#define YSCALE(y) (float(y))
#define ZSCALE(z) (zscale(z))
if (RDP_GETOM_Z_SOURCE_SEL(rdpState.otherModes))
z = rdpState.primitiveZ;
else
z = 0xffff;
// if (dump)
// fprintf(stderr, "fillrect cycle %d\n", other_modes.cycle_type);
rglStrip_t * strip = strips + nbStrips++;
rglAssert(nbStrips < MAX_STRIPS);
curChunk->nbStrips++;
rglVertex_t * vtx = vtxs + nbVtxs;
strip->flags = RGL_STRIP_ZBUFFER;
strip->vtxs = vtx;
vtx->x = XSCALE(x2); vtx->y = YSCALE(y1); vtx->z = ZSCALE(z); vtx++->w = 1;
vtx->x = XSCALE(x1); vtx->y = YSCALE(y1); vtx->z = ZSCALE(z); vtx++->w = 1;
vtx->x = XSCALE(x2); vtx->y = YSCALE(y2); vtx->z = ZSCALE(z); vtx++->w = 1;
vtx->x = XSCALE(x1); vtx->y = YSCALE(y2); vtx->z = ZSCALE(z); vtx++->w = 1;
strip->nbVtxs = vtx - strip->vtxs;
nbVtxs = vtx - vtxs;
}
void rglTriangle(uint32_t w1, uint32_t w2, int shade, int texture, int zbuffer,
uint32_t * rdp_cmd)
{
int tilenum = (w1 >> 16) & 0x7;
// if (tilenum == 7) {
// LOG("Fixing tilenum from 7 to 0\n");
// tilenum = 0;
// }
int j;
int xleft, xright, xleft_inc, xright_inc;
int xstart, xend;
int r, g, b, a, z, s, t, w;
int dr, dg, db, da;
int drdx = 0, dgdx = 0, dbdx = 0, dadx = 0, dzdx = 0, dsdx = 0, dtdx = 0, dwdx = 0;
int drdy = 0, dgdy = 0, dbdy = 0, dady = 0, dzdy = 0, dsdy = 0, dtdy = 0, dwdy = 0;
int drde = 0, dgde = 0, dbde = 0, dade = 0, dzde = 0, dsde = 0, dtde = 0, dwde = 0;
int flip = (w1 & 0x800000) ? 1 : 0;
int32_t yl, ym, yh;
int32_t xl, xm, xh;
int64_t dxldy, dxhdy, dxmdy;
uint32_t w3, w4, w5, w6, w7, w8;
uint32_t * shade_base = rdp_cmd + 8;
uint32_t * texture_base = rdp_cmd + 8;
uint32_t * zbuffer_base = rdp_cmd + 8;
int t1 = (texture && rglT1Usage(rdpState))? RGL_STRIP_TEX1:0;
int t2 = (texture && tilenum < 7 && rglT2Usage(rdpState))? RGL_STRIP_TEX2:0;
if (t1)
rglPrepareRendering(1, (tilenum==7 && RDP_GETOM_CYCLE_TYPE(rdpState.otherModes)==1)? 0:tilenum, 0, zbuffer);
if (t2)
rglPrepareRendering(1, tilenum+1, 0, zbuffer);
else if (!t1)
rglPrepareRendering(0, 0, 0, zbuffer);
curRBuffer->flags |= RGL_RB_HASTRIANGLES;
if (shade)
{
texture_base += 16;
zbuffer_base += 16;
}
if (texture)
{
zbuffer_base += 16;
}
w3 = rdp_cmd[2];
w4 = rdp_cmd[3];
w5 = rdp_cmd[4];
w6 = rdp_cmd[5];
w7 = rdp_cmd[6];
w8 = rdp_cmd[7];
yl = (w1 & 0x3fff);
ym = ((w2 >> 16) & 0x3fff);
yh = ((w2 >> 0) & 0x3fff);
xl = (int32_t)(w3);
xh = (int32_t)(w5);
xm = (int32_t)(w7);
dxldy = (int32_t)(w4);
dxhdy = (int32_t)(w6);
dxmdy = (int32_t)(w8);
if (yl & (0x800<<2)) yl |= 0xfffff000<<2;
if (ym & (0x800<<2)) ym |= 0xfffff000<<2;
if (yh & (0x800<<2)) yh |= 0xfffff000<<2;
yh &= ~3;
r = 0xff; g = 0xff; b = 0xff; a = 0xff; z = 0xffff0000; s = 0; t = 0; w = 0x30000;
dr = 0; dg = 0; db = 0; da = 0;
if (shade)
{
r = (shade_base[0] & 0xffff0000) | ((shade_base[+4 ] >> 16) & 0x0000ffff);
g = ((shade_base[0 ] << 16) & 0xffff0000) | (shade_base[4 ] & 0x0000ffff);
b = (shade_base[1 ] & 0xffff0000) | ((shade_base[5 ] >> 16) & 0x0000ffff);
a = ((shade_base[1 ] << 16) & 0xffff0000) | (shade_base[5 ] & 0x0000ffff);
drdx = (shade_base[2 ] & 0xffff0000) | ((shade_base[6 ] >> 16) & 0x0000ffff);
dgdx = ((shade_base[2 ] << 16) & 0xffff0000) | (shade_base[6 ] & 0x0000ffff);
dbdx = (shade_base[3 ] & 0xffff0000) | ((shade_base[7 ] >> 16) & 0x0000ffff);
dadx = ((shade_base[3 ] << 16) & 0xffff0000) | (shade_base[7 ] & 0x0000ffff);
drde = (shade_base[8 ] & 0xffff0000) | ((shade_base[12] >> 16) & 0x0000ffff);
dgde = ((shade_base[8 ] << 16) & 0xffff0000) | (shade_base[12] & 0x0000ffff);
dbde = (shade_base[9 ] & 0xffff0000) | ((shade_base[13] >> 16) & 0x0000ffff);
dade = ((shade_base[9 ] << 16) & 0xffff0000) | (shade_base[13] & 0x0000ffff);
drdy = (shade_base[10] & 0xffff0000) | ((shade_base[14] >> 16) & 0x0000ffff);
dgdy = ((shade_base[10] << 16) & 0xffff0000) | (shade_base[14] & 0x0000ffff);
dbdy = (shade_base[11] & 0xffff0000) | ((shade_base[15] >> 16) & 0x0000ffff);
dady = ((shade_base[11] << 16) & 0xffff0000) | (shade_base[15] & 0x0000ffff);
}
if (texture)
{
s = (texture_base[0 ] & 0xffff0000) | ((texture_base[4 ] >> 16) & 0x0000ffff);
t = ((texture_base[0 ] << 16) & 0xffff0000) | (texture_base[4 ] & 0x0000ffff);
w = (texture_base[1 ] & 0xffff0000) | ((texture_base[5 ] >> 16) & 0x0000ffff);
dsdx = (texture_base[2 ] & 0xffff0000) | ((texture_base[6 ] >> 16) & 0x0000ffff);
dtdx = ((texture_base[2 ] << 16) & 0xffff0000) | (texture_base[6 ] & 0x0000ffff);
dwdx = (texture_base[3 ] & 0xffff0000) | ((texture_base[7 ] >> 16) & 0x0000ffff);
dsde = (texture_base[8 ] & 0xffff0000) | ((texture_base[12] >> 16) & 0x0000ffff);
dtde = ((texture_base[8 ] << 16) & 0xffff0000) | (texture_base[12] & 0x0000ffff);
dwde = (texture_base[9 ] & 0xffff0000) | ((texture_base[13] >> 16) & 0x0000ffff);
dsdy = (texture_base[10] & 0xffff0000) | ((texture_base[14] >> 16) & 0x0000ffff);
dtdy = ((texture_base[10] << 16) & 0xffff0000) | (texture_base[14] & 0x0000ffff);
dwdy = (texture_base[11] & 0xffff0000) | ((texture_base[15] >> 16) & 0x0000ffff);
}
if (zbuffer)
{
//rglAssert(!(curRBuffer->flags & RGL_RB_DEPTH));
z = zbuffer_base[0];
dzdx = zbuffer_base[1];
dzde = zbuffer_base[2];
dzdy = zbuffer_base[3];
}
xh <<= 2; xm <<= 2; xl <<= 2;
r <<= 2; g <<= 2; b <<= 2; a <<= 2;
dsde >>= 2; dtde >>= 2; dsdx >>= 2; dtdx >>= 2;
dzdx >>= 2; dzde >>= 2; dzdy >>= 2;
dwdx >>= 2; dwde >>= 2; dwdy >>= 2;
// #define tile rdpTiles[tilenum]
// s -= (rdpTiles[tilenum].sl << 8);
// t -= (rdpTiles[tilenum].tl << 8);
// if (/*!tile.ms && */tile.mask_s)
// s &= (1<<tile.mask_s+10) - 1;
// if (/*!tile.mt && */tile.mask_t)
// t &= (1<<tile.mask_t+10) - 1;
// #undef tile
xleft = xm;
xright = xh;
xleft_inc = dxmdy;
xright_inc = dxhdy;
while (yh<ym &&
!((!flip && xleft < xright+0x10000) ||
(flip && xleft > xright-0x10000))) {
xleft += xleft_inc; xright += xright_inc;
s += dsde; t += dtde; w += dwde;
r += drde; g += dgde; b += dbde; a += dade;
z += dzde;
yh++;
}
j = ym-yh;
//rglAssert(j >= 0);
#define XSCALE(x) (float(x)/(1<<18))
#define YSCALE(y) (float(y)/(1<<2))
#define ZSCALE(z) (RDP_GETOM_Z_SOURCE_SEL(rdpState.otherModes)? zscale(rdpState.primitiveZ) : zscale(z>>16))
#define WSCALE(z) 1.0f/(RDP_GETOM_PERSP_TEX_EN(rdpState.otherModes)? (float(uint32_t(z) + 0x10000)/0xffff0000) : 1.0f)
//#define WSCALE(w) (RDP_GETOM_PERSP_TEX_EN(rdpState.otherModes)? 65536.0f*65536.0f/float((w+ 0x10000)) : 1.0f)
#define CSCALE(c) (((c)>0x3ff0000? 0x3ff0000:((c)<0? 0 : (c)))>>18)
#define _PERSP(w) ( w )
#define PERSP(s, w) ( ((int64_t)(s) << 20) / (_PERSP(w)? _PERSP(w):1) )
#define SSCALE(s, _w) (RDP_GETOM_PERSP_TEX_EN(rdpState.otherModes)? float(PERSP(s, _w))/(1 << 10) : float(s)/(1<<21))
#define TSCALE(s, w) (RDP_GETOM_PERSP_TEX_EN(rdpState.otherModes)? float(PERSP(s, w))/(1 << 10) : float(s)/(1<<21))
rglStrip_t * strip = strips + nbStrips++;
rglAssert(nbStrips < MAX_STRIPS);
curChunk->nbStrips++;
rglVertex_t * vtx = vtxs + nbVtxs;
strip->flags = (shade? RGL_STRIP_SHADE : 0) | t1 | t2
| RGL_STRIP_ZBUFFER;
//| (zbuffer? RGL_STRIP_ZBUFFER : 0);
strip->vtxs = vtx;
strip->tilenum = tilenum;
int sw;
if (j > 0)
{
int dx = (xleft-xright>>16);
if ((!flip && xleft < xright) ||
(flip/* && xleft > xright*/))
{
if (shade) {
vtx->r = CSCALE(r+drdx*dx);
vtx->g = CSCALE(g+dgdx*dx);
vtx->b = CSCALE(b+dbdx*dx);
vtx->a = CSCALE(a+dadx*dx);
}
if (texture) {
vtx->s = SSCALE(s+dsdx*dx, w+dwdx*dx);
vtx->t = TSCALE(t+dtdx*dx, w+dwdx*dx);
}
vtx->x = XSCALE(xleft);
vtx->y = YSCALE(yh);
vtx->z = ZSCALE(z+dzdx*dx);
vtx->w = WSCALE(w+dwdx*dx);
vtx++;
}
if ((!flip/* && xleft < xright*/) ||
(flip && xleft > xright))
{
if (shade) {
vtx->r = CSCALE(r);
vtx->g = CSCALE(g);
vtx->b = CSCALE(b);
vtx->a = CSCALE(a);
}
if (texture) {
vtx->s = SSCALE(s, w);
vtx->t = TSCALE(t, w);
}
vtx->x = XSCALE(xright);
vtx->y = YSCALE(yh);
vtx->z = ZSCALE(z);
vtx->w = WSCALE(w);
vtx++;
}
}
xleft += xleft_inc*j; xright += xright_inc*j;
s += dsde*j; t += dtde*j; w += dwde*j;
r += drde*j; g += dgde*j; b += dbde*j; a += dade*j;
z += dzde*j;
// render ...
xleft = xl;
//if (yl-ym > 0)
{
int dx = (xleft-xright>>16);
if ((!flip && xleft <= xright) ||
(flip/* && xleft >= xright*/))
{
if (shade) {
vtx->r = CSCALE(r+drdx*dx);
vtx->g = CSCALE(g+dgdx*dx);
vtx->b = CSCALE(b+dbdx*dx);
vtx->a = CSCALE(a+dadx*dx);
}
if (texture) {
vtx->s = SSCALE(s+dsdx*dx, w+dwdx*dx);
vtx->t = TSCALE(t+dtdx*dx, w+dwdx*dx);
}
vtx->x = XSCALE(xleft);
vtx->y = YSCALE(ym);
vtx->z = ZSCALE(z+dzdx*dx);
vtx->w = WSCALE(w+dwdx*dx);
vtx++;
}
if ((!flip/* && xleft <= xright*/) ||
(flip && xleft >= xright))
{
if (shade) {
vtx->r = CSCALE(r);
vtx->g = CSCALE(g);
vtx->b = CSCALE(b);
vtx->a = CSCALE(a);
}
if (texture) {
vtx->s = SSCALE(s, w);
vtx->t = TSCALE(t, w);
}
vtx->x = XSCALE(xright);
vtx->y = YSCALE(ym);
vtx->z = ZSCALE(z);
vtx->w = WSCALE(w);
vtx++;
}
}
xleft_inc = dxldy;
xright_inc = dxhdy;
j = yl-ym;
//rglAssert(j >= 0);
//j--; // ?
xleft += xleft_inc*j; xright += xright_inc*j;
s += dsde*j; t += dtde*j; w += dwde*j;
r += drde*j; g += dgde*j; b += dbde*j; a += dade*j;
z += dzde*j;
while (yl>ym &&
!((!flip && xleft < xright+0x10000) ||
(flip && xleft > xright-0x10000))) {
xleft -= xleft_inc; xright -= xright_inc;
s -= dsde; t -= dtde; w -= dwde;
r -= drde; g -= dgde; b -= dbde; a -= dade;
z -= dzde;
j--;
yl--;
}
// render ...
if (j >= 0) {
int dx = (xleft-xright>>16);
if ((!flip && xleft <= xright) ||
(flip/* && xleft >= xright*/))
{
if (shade) {
vtx->r = CSCALE(r+drdx*dx);
vtx->g = CSCALE(g+dgdx*dx);
vtx->b = CSCALE(b+dbdx*dx);
vtx->a = CSCALE(a+dadx*dx);
}
if (texture) {
vtx->s = SSCALE(s+dsdx*dx, w+dwdx*dx);
vtx->t = TSCALE(t+dtdx*dx, w+dwdx*dx);
}
vtx->x = XSCALE(xleft);
vtx->y = YSCALE(yl);
vtx->z = ZSCALE(z+dzdx*dx);
vtx->w = WSCALE(w+dwdx*dx);
vtx++;
}
if ((!flip/* && xleft <= xright*/) ||
(flip && xleft >= xright))
{
if (shade) {
vtx->r = CSCALE(r);
vtx->g = CSCALE(g);
vtx->b = CSCALE(b);
vtx->a = CSCALE(a);
}
if (texture) {
vtx->s = SSCALE(s, w);
vtx->t = TSCALE(t, w);
}
vtx->x = XSCALE(xright);
vtx->y = YSCALE(yl);
vtx->z = ZSCALE(z);
vtx->w = WSCALE(w);
vtx++;
}
}
strip->nbVtxs = vtx - strip->vtxs;
nbVtxs = vtx - vtxs;
}