Skip to content
Permalink
master
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

newton/demo/source/poly_arm.cpp

Go to file
 
 
Cannot retrieve contributors at this time
878 lines (755 sloc) 19.4 KB
Raw Blame
Open in GitHub Desktop
#include <gba_video.h>
#include <gba_dma.h>
#include <mappy.h>
#include "poly.h"
#include "support.h"
#include "math.h"
#include "exstructs.h"
#include <gba_systemcalls.h>
#define SHIFT 16
#define LEFT 1
#define RIGHT 2
#define FILLER_SHIFT256_1X1
#define FILLER_TEX256_1X1
#define FILLER_TEX256_2X1
//#define FILLER_NMAP_1X1
//#define FILLER_NMAP_2X1
#define FILLER_EBUMP_1X1
//#define FILLER_EBUMP_2X1
namespace {
// stuff used by draw_part
const s16* global_light;
void hline_shift256_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_tex256_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_nmap_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_ebump_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_tex256_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_nmap_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_ebump_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) IWRAM_CODE;
void hline_shift256_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) {
#ifndef FILLER_SHIFT256_1X1
assert(0);
#else
const u8* tex = tex0;
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define ITERATION {\
u32 bg = *(u16*)dst; \
u32 index0 = (uv>>24)|(uv&0xFF00);\
u32 col0 = tex[index0];\
uv += duvdx;\
u32 index1 = (uv>>24)|(uv&0xFF00);\
u32 col1 = tex[index1];\
uv += duvdx;\
*(u16*)dst = ((bg & 0xFF) >> col0) | (((bg & 0xFF00) >> col1) & 0xFF00) ;\
dst += 2;\
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
u8 col = tex[index];
val = (val&0x00FF) | (((val & 0xFF00) >> col) & 0xFF00);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
register unsigned c = counter >> 4;
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
#undef ITERATION
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
u8 col = tex[index];
val = (val&0xFF00) | ((val & 0xFF) >> col);
x1++;
}
#endif
}
void hline_tex256_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex, const u8* tex1, int palOffs) {
#ifndef FILLER_TEX256_2X1
assert(0);
#else
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define ITERATION {\
u32 index = (uv>>24)|(uv&0xFF00);\
u8 col = tex[index];\
*dst = col;\
dst += 2;\
uv += (duvdx<<1); \
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
u8 col = tex[index];
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
register unsigned c = counter >> 4;
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
#undef ITERATION
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
u8 col = tex[index];
val = (val&0xFF00)|col;
x1++;
}
#endif
}
void hline_tex256_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) {
#ifndef FILLER_TEX256_1X1
assert(0);
#else
const u8* tex = tex0;
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define ITERATION {\
u32 index0 = (uv>>24)|(uv&0xFF00);\
u32 col0 = tex[index0];\
uv += duvdx;\
u32 index1 = (uv>>24)|(uv&0xFF00);\
u32 col1 = tex[index1];\
uv += duvdx;\
*(u16*)dst = (col1<<8) | col0;\
dst += 2;\
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
u8 col = tex[index];
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
register unsigned c = counter >> 3;
switch (counter&7) {
do {
ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
#undef ITERATION
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
u8 col = tex[index];
val = (val&0xFF00)|col;
x1++;
}
#endif
}
void hline_nmap_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) {
#ifndef FILLER_NMAP_2X1
assert(0);
#else
const s16* tex = (s16*)tex0;
s32 lightx = global_light[0];
s32 lighty = global_light[1];
s32 lightz = global_light[2];
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define STUFF(ut, inn) {\
s32 lol = 0; \
lol += lightx * (s32(inn<<16)>>27);\
lol += lighty * (s32(inn<<21)>>27);\
lol += lightz * (s32(inn<<26)>>27);\
lol >>= 8;\
ut = lol+palOffs;\
}
#define ITERATION {\
u32 index = (uv>>24)|(uv&0xFF00);\
s32 val = tex[index];\
STUFF(*dst, val);\
dst += 2;\
uv += (duvdx<<1); \
}
#define ITERATION_ASM \
asm ( \
"and r0, %0, #65280\n\t" \
"orr r0, r0, %0, lsr #24\n\t" \
"mov r0, r0, asl #1\n\t" \
"ldrsh r1, [r0, %2]\n\t" \
"mov r2, r1, asr #11\n\t" \
"mul r0, %3, r2\n\t" \
"mov r2, r1, asl #21\n\t" \
"mov r2, r2, asr #27\n\t" \
"mla r0, %4, r2, r0\n\t" \
"mov r2, r1, asl #26\n\t" \
"mov r2, r2, asr #27\n\t" \
"mla r0, %5, r2, r0\n\t" \
"add r0, %6, r0, asr #8\n\t" \
"strb r0, [%1], #2\n\t" \
"add %0, %0, %7" \
: "=r"(uv), "=r"(dst) \
: "r"(tex), "r"(lightx), "r"(lighty), "r"(lightz), "r"(palOffs), "r"(duvdx<<1), "0"(uv), "1"(dst) \
: "r0", "r1", "r2" \
)
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
s32 col = tex[index];
STUFF(col, col);
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
unsigned c = counter >> 4;
{
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION_ASM;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
}
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
s32 col = tex[index];
STUFF(col, col);
val = (val&0xFF00)|col;
x1++;
}
#undef STUFF
#undef ITERATION
#undef ITERATION_ASM
#endif
}
void hline_nmap_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0_8, const u8* tex1, int palOffs) {
#ifndef FILLER_NMAP_1X1
assert(0);
#else
const s16* tex = (s16*)tex0_8;
s32 lightx = global_light[0];
s32 lighty = global_light[1];
s32 lightz = global_light[2];
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define STUFF(ut, inn) {\
s32 lol = 0; \
lol += lightx * (s32(inn<<16)>>27);\
lol += lighty * (s32(inn<<21)>>27);\
lol += lightz * (s32(inn<<26)>>27);\
lol >>= 8;\
ut = lol+palOffs;\
}
#define ITERATION {\
u32 index0 = (uv>>24)|(uv&0xFF00);\
s32 tmp0 = tex[index0];\
u32 col0;\
STUFF(col0, tmp0);\
uv += duvdx;\
u32 index1 = (uv>>24)|(uv&0xFF00);\
s32 tmp1 = tex[index1];\
u32 col1;\
STUFF(col1, tmp1);\
uv += duvdx;\
*(u16*)dst = (col1<<8)|col0;\
dst += 2;\
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
s32 col = tex[index];
STUFF(col, col);
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
unsigned c = counter >> 4;
{
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
}
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
s32 col = tex[index];
STUFF(col, col);
val = (val&0xFF00)|col;
x1++;
}
#undef STUFF
#undef ITERATION
#endif
}
void hline_ebump_2x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0_8, const u8* tex1, int palOffs) {
#ifndef FILLER_EBUMP_2X1
assert(0);
#else
const s16* tex = (s16*)tex0_8;
s32 lightangle = -global_light[0];
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define STUFF(ut, inn) {\
int v = ((inn-lightangle)<<1)&0xFF00; \
int u = inn & 0xFF; \
ut = tex1[v|u]; \
}
//u32 index = (uv>>24)|(uv&0xFF00);
#define ITERATION {\
u32 index0 = (uv>>24)|(uv&0xFF00);\
u32 val0 = tex[index0];\
STUFF(*dst, val0);\
dst += 2;\
uv += (duvdx<<1); \
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
u16 col = tex[index];
STUFF(col, col);
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
register unsigned c = counter >> 4;
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
#undef ITERATION
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
u16 col = tex[index];
STUFF(col, col);
val = (val&0xFF00)|col;
x1++;
}
#endif
}
void hline_ebump_1x1(u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs) {
#ifndef FILLER_EBUMP_1X1
assert(0);
#else
const s16* tex = (s16*)tex0;
s32 lightangle = -global_light[0];
if (x2>GBA_WIDTH) x2=GBA_WIDTH;
if (x1<0) {
uv -= ((duvdx&0xFFFF0000)*x1)|u16(duvdx*x1);
x1=0;
}
if (x2 <= x1) return;
#define STUFF(ut, inn) {\
int v = ((inn-lightangle)<<1)&0xFF00; \
int u = inn & 0xFF; \
ut = tex1[v|u]; \
}
//u32 index = (uv>>24)|(uv&0xFF00);
#define ITERATION {\
u32 index0 = (uv>>24)|(uv&0xFF00);\
u32 tmp0 = tex[index0];\
u32 col0;\
STUFF(col0, tmp0);\
uv += duvdx;\
u32 index1 = (uv>>24)|(uv&0xFF00);\
u32 tmp1 = tex[index1];\
u32 col1;\
STUFF(col1, tmp1);\
uv += duvdx;\
*(u16*)dst = (col1<<8) | col0;\
dst += 2;\
}
if (x1&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = ((u16*)dst)[x1>>1];
u16 col = tex[index];
STUFF(col, col);
val = (val&0x00FF)|(u32(col)<<8);
x1++;
uv += duvdx;
}
dst += x1;
s32 counter = (x2-x1)>>1;
register unsigned c = counter >> 4;
switch (counter&15) {
do {
ITERATION;
case 15: ITERATION;
case 14: ITERATION;
case 13: ITERATION;
case 12: ITERATION;
case 11: ITERATION;
case 10: ITERATION;
case 9: ITERATION;
case 8: ITERATION;
case 7: ITERATION;
case 6: ITERATION;
case 5: ITERATION;
case 4: ITERATION;
case 3: ITERATION;
case 2: ITERATION;
case 1: ITERATION;
case 0:;
} while (c--);
}
#undef ITERATION
if (x2&1) {
u32 index = (uv>>24)|(uv&0xFF00);
u16& val = *((u16*)dst);
u16 col = tex[index];
STUFF(col, col);
val = (val&0xFF00)|col;
x1++;
}
#endif
}
struct XIter {
s32 x;
s32 delta_x;
INLINE void increment() {
x+=delta_x;
}
INLINE void increment(int count) {
x+=delta_x*count;
}
};
#define PACKED_INTERPOLATION
struct IterL {
private:
#ifdef PACKED_INTERPOLATION
u32 uv;
u32 delta_uv;
#else
u32 u, v;
u32 delta_u;
u32 delta_v;
#endif
public:
INLINE void increment()
{
#ifdef PACKED_INTERPOLATION
uv +=delta_uv;
#else
u += delta_u;
v += delta_v;
#endif
}
INLINE void increment(int count)
{
#ifdef PACKED_INTERPOLATION
uv += ((delta_uv&0xFFFF0000)*count)|u16(delta_uv*count);
#else
u += delta_u * count;
v += delta_v * count;
#endif
}
INLINE s32 setUv(s32 u, s32 v)
{
#ifdef PACKED_INTERPOLATION
this->uv = (u32(u) & 0xFFFF0000) | ((u32(v) >> 16) & 0xFFFF);
#else
this->u = u;
this->v = v;
#endif
}
INLINE s32 setUvDelta(s32 u_delta, s32 v_delta)
{
#ifdef PACKED_INTERPOLATION
this->delta_uv = (u32(u_delta) & 0xFFFF0000) | ((u32(v_delta) >> 16) & 0xFFFF);
#else
this->delta_u = u_delta;
this->delta_v = v_delta;
#endif
}
INLINE s32 getUv()
{
#ifdef PACKED_INTERPOLATION
return uv;
#else
return (u & 0xFFFF0000) | ((v >> 16) & 0xFFFF);
#endif
}
};
typedef void (*HLinePtr) (u8* dst, int x1, int x2, u32 uv, s32 duvdx, const u8* tex0, const u8* tex1, int palOffs);
INLINE void draw_part(u8* dst, IterL& iterL, XIter& xiterL, XIter& xiterR, int count, s32 duvdx, HLinePtr hLinePtr, const Material& mtl) {
for (; count; count--) {
int xl = xiterL.x>>16;
int xr = xiterR.x>>16;
hLinePtr(dst, xl, xr, iterL.getUv(), duvdx, mtl.tex[0], mtl.tex[1], mtl.palOffs);
// hLinePtr(dst, xl, xr, iterL.uv, duvdx, mtl.tex[0], mtl.tex[1], mtl.palOffs);
iterL.increment();
xiterL.increment();
xiterR.increment();
dst += GBA_WIDTH;
}
}
const unsigned char conf_tab[8*8] = {
0, 255, 255, 255, 255, 255, 255, 255,
RIGHT, 2, 1, 0, 1, 0, 255, 255,
RIGHT, 1, 0, 2, 0, 2, 255, 255,
LEFT, 1, 2, 0, 0, 2, 255, 255,
RIGHT, 0, 2, 1, 2, 1, 255, 255,
LEFT, 2, 0, 1, 1, 0, 255, 255,
LEFT, 0, 1, 2, 2, 1, 255, 255,
0, 255, 255, 255, 255, 255, 255, 255,
};
}; // namespace
void polyfiller(u8* target, PolyVertex v[3], const Material& mtl, const s16 light[3], int clip_top, int clip_bottom) {
global_light = light;
unsigned int index = 0;
if ((v[0].y >> 8) < (v[1].y >> 8)) index |= (1 << 2) * 8;
if ((v[1].y >> 8) < (v[2].y >> 8)) index |= (1 << 1) * 8;
if ((v[2].y >> 8) < (v[0].y >> 8)) index |= (1 << 0) * 8;
int type = conf_tab[index++];
const PolyVertex& top_vert = v[conf_tab[index++]];
const PolyVertex& center_vert = v[conf_tab[index++]];
const PolyVertex& bottom_vert = v[conf_tab[index++]];
const PolyVertex& right_vert = v[conf_tab[index++]];
const PolyVertex& left_vert = v[conf_tab[index++]];
int top = top_vert.y >> 8;
int center = center_vert.y >> 8;
int bottom = bottom_vert.y >> 8;
if (top>=clip_bottom || bottom<=clip_top) return;
s32 duvdx;
// s32 duvdx_half;
{
s32 dX10 = (v[1].x >> 8) -(v[0].x >> 8);
s32 dX21 = (v[2].x >> 8) -(v[1].x >> 8);
s32 dY10 = (v[1].y >> 8) -(v[0].y >> 8);
s32 dY21 = (v[2].y >> 8) -(v[1].y >> 8);
s32 area = dX21*dY10 - dX10*dY21;
if (area<1) return;
// s32 area_inv =s32((1 << 16) / (float(area) / 256));
s32 area_inv = rcp(area<<8);
s32 du10 = v[1].u - v[0].u;
s32 du21 = v[2].u - v[1].u;
s32 dv10 = v[1].v - v[0].v;
s32 dv21 = v[2].v - v[1].v;
#if 1
s32 dudx = (s64(du21*dY10 - du10*dY21) * area_inv) >> 8;
s32 dvdx = (s64(dv21*dY10 - dv10*dY21) * area_inv) >> 8;
#else
s32 dudx = int((float(du21*dY10 - du10*dY21) / area) * 65536);
s32 dvdx = int((float(dv21*dY10 - dv10*dY21) / area) * 65536);
#endif
duvdx = (dudx & 0xFFFF0000) | (u32(dvdx)>>16);
// duvdx_half = ((dudx>>1)&0xFFFF0000) | (u32(dvdx>>1)>>16);
}
s32 count;
u8* dst;
s32 long_rcp = ircp((bottom_vert.y >> 8) - (top_vert.y >> 8));
s32 left_rcp = ircp((left_vert.y >> 8) - (top_vert.y >> 8));
XIter xiter_long;
xiter_long.x = (top_vert.x>>8)<<SHIFT;
xiter_long.delta_x = ((bottom_vert.x >> 8) - (top_vert.x >> 8)) * long_rcp;
IterL iterL;
// finn uv'er
{
iterL.setUv(top_vert.u << 16, top_vert.v << 16);
s32 delta_u = (s64(left_vert.u - top_vert.u) * left_rcp);
s32 delta_v = (s64(left_vert.v - top_vert.v) * left_rcp);
iterL.setUvDelta(delta_u, delta_v);
}
// -y clipping
if ((top_vert.y >> 8)<clip_top) {
xiter_long.increment(clip_top-(top_vert.y>>8));
iterL.increment(clip_top-(top_vert.y>>8));
}
// select filler
HLinePtr hLinePtr=NULL;
switch(mtl.filler | (mtl.oneone<<7)) {
default:
case FILLER_TEX256:
hLinePtr = (HLinePtr)hline_tex256_2x1;
break;
case FILLER_TEX256 | (1<<7):
hLinePtr = (HLinePtr)hline_tex256_1x1;
break;
case FILLER_NMAP:
hLinePtr = (HLinePtr)hline_nmap_2x1;
break;
case FILLER_NMAP | (1<<7):
hLinePtr = (HLinePtr)hline_nmap_1x1;
break;
case FILLER_EBUMP:
hLinePtr = (HLinePtr)hline_ebump_2x1;
break;
case FILLER_EBUMP | (1<<7):
hLinePtr = (HLinePtr)hline_ebump_1x1;
break;
case FILLER_SHIFT | (1 << 7):
hLinePtr = (HLinePtr)hline_shift256_1x1;
break;
}
// UPPER
if (center>clip_top) { // upper is visible
if (center>clip_bottom) center = clip_bottom; // +y clipping
s32 short_rcp = ircp((center_vert.y >> 8) - (top_vert.y >> 8));
XIter xiter_short;
xiter_short.x = (top_vert.x>>8)<<SHIFT;
xiter_short.delta_x = ((center_vert.x >> 8) - (top_vert.x >> 8)) * short_rcp;
if ((top_vert.y>>8)<clip_top) { // -y clipping
top=clip_top; // -y stuff
xiter_short.increment(clip_top-(top_vert.y >> 8));
}
dst = &target[top*GBA_WIDTH];
count = center-top;
XIter& xiterL = (type==RIGHT) ? xiter_long : xiter_short;
XIter& xiterR = (type==RIGHT) ? xiter_short : xiter_long;
draw_part(dst, iterL, xiterL, xiterR, count, duvdx, hLinePtr, mtl);
}
// LOWER
if (center<clip_bottom) { // lower is visible
if (bottom>clip_bottom) bottom = clip_bottom; // +y clipping
s32 short_rcp = ircp((bottom_vert.y >> 8) - (center_vert.y >> 8));
XIter xiter_short;
xiter_short.x = (center_vert.x>>8)<<SHIFT;
xiter_short.delta_x = ((bottom_vert.x >> 8) - (center_vert.x >> 8)) * short_rcp;
// oppdater uv for lower
if (type==LEFT) {
iterL.setUv(center_vert.u << 16, center_vert.v << 16);
s32 delta_u = (s64(bottom_vert.u - center_vert.u) * short_rcp);
s32 delta_v = (s64(bottom_vert.v - center_vert.v) * short_rcp);
iterL.setUvDelta(delta_u, delta_v);
if ((center_vert.y >> 8)<clip_top) iterL.increment(clip_top-(center_vert.y >> 8));
}
if ((center_vert.y>>8)<clip_top) { // -y clipping
xiter_short.increment(clip_top-(center_vert.y>>8));
dst = &target[clip_top*GBA_WIDTH];
count = bottom-clip_top;
} else { // no -y clipping
dst = &target[center*GBA_WIDTH];
count = bottom-center;
}
XIter& xiterL = (type==RIGHT) ? xiter_long : xiter_short;
XIter& xiterR = (type==RIGHT) ? xiter_short : xiter_long;
draw_part(dst, iterL, xiterL, xiterR, count, duvdx, hLinePtr, mtl);
}
}