mirror/tww
1
0
Fork 0
mirror of https://github.com/zeldaret/tww.git synced 2026-09-09 19:51:20 -04:00
Code Issues Packages Projects Releases Wiki Activity

blockenc and encode_s3tc progress (#1199)

Browse Source 
* blockenc and encode_s3tc progress

* cleanup
This commit is contained in:
LagoLunatic
2026-09-09 14:56:37 -04:00
committed by GitHub
parent 99662e663f
commit e3c6c8143b
1 changed files with 113 additions and 180 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/m_Do/m_Do_graphic.cpp
+113 -180
View File
@@ -915,32 +915,15 @@ void blockenc(u8* block) {
/* Nonmatching */
u8 colors[16*3]; // sp18
u32 color_mags[4]; // sp08
u32 i;
u32 r9;
u32 pix_idx;
u32 color_num;
u32 i;
u32 color_0_idx;
u32 color_1_idx;
u32 r23;
u32 r18;
u32 r0;
u32 g0;
u32 b0;
u32 r1;
u32 g1;
u32 b1;
u32 r2;
u32 g2;
u32 b2;
u32 r3;
u32 g3;
u32 b3;
u32 r4;
u32 g4;
u32 b4;
u32 r21;
u32 r20;
u32 r22;
@@ -961,63 +944,52 @@ void blockenc(u8* block) {
i = 0; // r8
pix_idx = 0; // r10
for (; i < 0x30; i += 3, pix_idx += 4) {
u32 r9 = 0; // r9
u8* pix = &block[pix_idx];
r9 = 0; // r9
for (int j = 0; j < color_num; j++) {
u8* temp = &colors[r9*3];
if (temp[0] == pix[0] && temp[1] == pix[1] && temp[2] == pix[2]) {
if (colors[r9*3+0] == block[pix_idx+0] && colors[r9*3+1] == block[pix_idx+1] && colors[r9*3+2] == block[pix_idx+2]) {
break;
}
r9++;
}
if (r9 == color_num) {
colors[r9*3 + 0] = pix[0];
colors[r9*3 + 1] = pix[1];
colors[r9*3 + 2] = pix[2];
colors[r9*3 + 0] = block[pix_idx+0];
colors[r9*3 + 1] = block[pix_idx+1];
colors[r9*3 + 2] = block[pix_idx+2];
color_num += 1;
}
}
color_0_idx = 0; // r17
color_1_idx = 0; // r16
u32 color_0_idx = 0; // r17
u32 color_1_idx = 0; // r16
if (color_num > 1) {
r23 = 0;
r18 = INT32_MAX;
for (; r23 < color_num*3 - 3; r23 += 3) {
r0 = colors[r23+0] * 30;
g0 = colors[r23+1] * 59;
b0 = colors[r23+2] * 11;
for (r21 = r23+3; r21 < color_num*3 - 3; r21 += 3) {
r1 = colors[r21+0] * 30;
g1 = colors[r21+1] * 59;
b1 = colors[r21+2] * 11;
// r3 = (r0 * 5 + r1 * 3) / 8;
// g3 = (g0 * 5 + g1 * 3) / 8;
// b3 = (b0 * 5 + b1 * 3) / 8;
// r4 = (r1 * 5 + r0 * 3) / 8;
// g4 = (g1 * 5 + g0 * 3) / 8;
// b4 = (b1 * 5 + b0 * 3) / 8;
u32 r0 = colors[r23+0] * 30;
u32 g0 = colors[r23+1] * 59;
u32 b0 = colors[r23+2] * 11;
for (r21 = r23+3; r21 < color_num*3; r21 += 3) {
u32 r1 = colors[r21+0] * 30;
u32 g1 = colors[r21+1] * 59;
u32 b1 = colors[r21+2] * 11;
r20 = 0; // r20
pix_idx = r20; // r19
r22 = pix_idx; // r22
for (; r22 < 0x10; r22++, pix_idx += 4) {
// b2 = block[pix_idx + 2] * 11;
// r2 = block[pix_idx + 0] * 30;
// g2 = block[pix_idx + 1] * 59;
u32 r2 = (r0 * 5 + r1 * 3) / 8;
u32 g2 = (g0 * 5 + g1 * 3) / 8;
u32 b2 = (b0 * 5 + b1 * 3) / 8;
u32 r3 = (r1 * 5 + r0 * 3) / 8;
u32 g3 = (g1 * 5 + g0 * 3) / 8;
u32 b3 = (b1 * 5 + b0 * 3) / 8;
color_mags[0] = COLOR_SQUARE_MAG(r0, g0, b0, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[1] = COLOR_SQUARE_MAG(r1, g1, b1, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
r3 = (r0 * 5 + r1 * 3) / 8;
g3 = (g0 * 5 + g1 * 3) / 8;
b3 = (b0 * 5 + b1 * 3) / 8;
color_mags[2] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
r4 = (r1 * 5 + r0 * 3) / 8;
g4 = (g1 * 5 + g0 * 3) / 8;
b4 = (b1 * 5 + b0 * 3) / 8;
color_mags[3] = COLOR_SQUARE_MAG(r4, g4, b4, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[2] = COLOR_SQUARE_MAG(r2, g2, b2, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[3] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
u32 r5 = 0;
u32 r6 = INT32_MAX;
for (; r5 < ARRAY_SIZE(color_mags); r5++) {
for (; r5 < 4; r5++) {
if (color_mags[r5] < r6) {
r6 = color_mags[r5];
}
@@ -1053,35 +1025,31 @@ void blockenc(u8* block) {
block[0x42] = r4_ >> 8;
block[0x43] = r4_ & 0xFF;
r0 = ((r11 >> 8) & 0xF8) * 30;
g0 = ((r11 >> 3) & 0xFC) * 59;
b0 = ((r11 << 3) & 0xFC) * 11;
r1 = ((r4_ >> 8) & 0xF8) * 30;
g1 = ((r4_ >> 3) & 0xFC) * 59;
b1 = ((r4_ << 3) & 0xFC) * 11;
r3 = (r0 * 5 + r1 * 3) / 8;
g3 = (g0 * 5 + g1 * 3) / 8;
b3 = (b0 * 5 + b1 * 3) / 8;
r4 = (r1 * 5 + r0 * 3) / 8;
g4 = (g1 * 5 + g0 * 3) / 8;
b4 = (b1 * 5 + b0 * 3) / 8;
u32 r0 = ((r11 >> 8) & 0xF8) * 30;
u32 g0 = ((r11 >> 3) & 0xFC) * 59;
u32 b0 = ((r11 << 3) & 0xFC) * 11;
u32 r1 = ((r4_ >> 8) & 0xF8) * 30;
u32 g1 = ((r4_ >> 3) & 0xFC) * 59;
u32 b1 = ((r4_ << 3) & 0xFC) * 11;
u32 r2 = (r0 * 5 + r1 * 3) / 8;
u32 g2 = (g0 * 5 + g1 * 3) / 8;
u32 b2 = (b0 * 5 + b1 * 3) / 8;
u32 r3 = (r1 * 5 + r0 * 3) / 8;
u32 g3 = (g1 * 5 + g0 * 3) / 8;
u32 b3 = (b1 * 5 + b0 * 3) / 8;
r6 = 30; // r6 (bit offset within r8)
r8 = 0; // r8 (bitfield of color indexes)
pix_idx = r8;
for (; pix_idx < 0x40; pix_idx += 4, r6 -= 2) {
// b2 = block[pix_idx + 2] * 11;
// r2 = block[pix_idx + 0] * 30;
// g2 = block[pix_idx + 1] * 59;
color_mags[0] = COLOR_SQUARE_MAG(r0, g0, b0, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[1] = COLOR_SQUARE_MAG(r1, g1, b1, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[2] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[3] = COLOR_SQUARE_MAG(r4, g4, b4, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[2] = COLOR_SQUARE_MAG(r2, g2, b2, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[3] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
r24 = 0;
r5 = 0;
r5 = r24;
r25 = INT32_MAX;
u32 n = ARRAY_SIZE(color_mags) - r5;
for (; r5 < n; r5++) {
for (; r5 < 4; r5++) {
if (color_mags[r5] < r25) {
r25 = color_mags[r5];
r24 = r5;
@@ -1095,20 +1063,18 @@ void blockenc(u8* block) {
color_num = pix_idx; // r11
i = color_num; // r8
for (; i < 0x30; i += 3, pix_idx += 4) {
u8* pix = &block[pix_idx];
if (pix[3] == 0xFF) {
u32 r9 = 0; // r9
if (block[pix_idx+3] == 0xFF) {
r9 = 0; // r9
for (int j = 0; j < color_num; j++) {
u8* temp = &colors[r9*3];
if (temp[0] == pix[0] && temp[1] == pix[1] && temp[2] == pix[2]) {
if (colors[r9*3+0] == block[pix_idx+0] && colors[r9*3+1] == block[pix_idx+1] && colors[r9*3+2] == block[pix_idx+2]) {
break;
}
r9++;
}
if (r9 == color_num) {
colors[r9*3 + 0] = pix[0];
colors[r9*3 + 1] = pix[1];
colors[r9*3 + 2] = pix[2];
colors[r9*3 + 0] = block[pix_idx+0];
colors[r9*3 + 1] = block[pix_idx+1];
colors[r9*3 + 2] = block[pix_idx+2];
color_num += 1;
}
}
@@ -1119,43 +1085,34 @@ void blockenc(u8* block) {
(colors[2] & 0xF8) >> 3;
r11 = r4_;
color_0_idx = 0; // r17
color_1_idx = 0; // r16
// color_0_idx = 0; // r17
// color_1_idx = 0; // r16
if (color_num > 1) {
r23 = 0;
r18 = INT32_MAX;
for (; r23 < color_num*3 - 3; r23 += 3) {
u8* colors_r10 = &colors[r23];
r0 = colors_r10[0] * 30;
g0 = colors_r10[1] * 59;
b0 = colors_r10[2] * 11;
u32 r0 = colors[r23+0] * 30;
u32 g0 = colors[r23+1] * 59;
u32 b0 = colors[r23+2] * 11;
for (r21 = r23+3; r21 < color_num*3 - 3; r21 += 3) {
u8* colors_r11 = &colors[r21];
r1 = colors_r11[0] * 30;
g1 = colors_r11[1] * 59;
b1 = colors_r11[2] * 11;
r3 = (r0 + r1) / 2;
g3 = (g0 + g1) / 2;
b3 = (b0 + b1) / 2;
u32 r1 = colors[r21+0] * 30;
u32 g1 = colors[r21+1] * 59;
u32 b1 = colors[r21+2] * 11;
u32 r3 = (r0 + r1) / 2;
u32 g3 = (g0 + g1) / 2;
u32 b3 = (b0 + b1) / 2;
r20 = 0; // r20
pix_idx = r20; // r19
r22 = pix_idx; // r22
for (; r22 < 0x10; r22++, pix_idx += 4) {
if (block[pix_idx + 3] == 0xFF) {
// b2 = block[pix_idx + 2] * 11;
// r2 = block[pix_idx + 0] * 30;
// g2 = block[pix_idx + 1] * 59;
color_mags[0] = COLOR_SQUARE_MAG(r0, g0, b0, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[1] = COLOR_SQUARE_MAG(r1, g1, b1, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[2] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
// r4 = (r1 * 5 + r0 * 3) / 8;
// g4 = (g1 * 5 + g0 * 3) / 8;
// b4 = (b1 * 5 + b0 * 3) / 8;
// color_mags[3] = COLOR_SQUARE_MAG(r4, g4, b4, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
u32 r5 = 0;
u32 r6 = INT32_MAX;
for (; r5 < ARRAY_SIZE(color_mags); r5++) {
for (; r5 < 3; r5++) {
if (color_mags[r5] < r6) {
r6 = color_mags[r5];
}
@@ -1169,12 +1126,12 @@ void blockenc(u8* block) {
r18 = r20;
// color_0_idx = r23;
// color_1_idx = r21;
r11 = (colors_r10[0] & 0xF8) << 8 |
(colors_r10[1] & 0xFC) << 3 |
(colors_r10[2] & 0xF8) >> 3;
r4_ = (colors_r11[0] & 0xF8) << 8 |
(colors_r11[1] & 0xFC) << 3 |
(colors_r11[2] & 0xF8) >> 3;
r11 = (colors[r23+0] & 0xF8) << 8 |
(colors[r23+1] & 0xFC) << 3 |
(colors[r23+2] & 0xF8) >> 3;
r4_ = (colors[r21+0] & 0xF8) << 8 |
(colors[r21+1] & 0xFC) << 3 |
(colors[r21+2] & 0xF8) >> 3;
}
}
}
@@ -1185,50 +1142,42 @@ void blockenc(u8* block) {
r11 = r4_;
r4_ = temp;
} else if (r11 == r4_) {
r4_ = 0;
r11 = 0;
}
block[0x40] = r11 >> 8;
block[0x41] = r11 & 0xFF;
block[0x42] = r4_ >> 8;
block[0x43] = r4_ & 0xFF;
r0 = (r11 >> 8) & 0xF8;
g0 = (r11 >> 3) & 0xFC;
b0 = (r11 << 3) & 0xFC;
r1 = (r4_ >> 8) & 0xF8;
g1 = (r4_ >> 3) & 0xFC;
b1 = (r4_ << 3) & 0xFC;
u32 r0 = (r11 >> 8) & 0xF8;
u32 g0 = (r11 >> 3) & 0xFC;
u32 b0 = (r11 << 3) & 0xFC;
u32 r1 = (r4_ >> 8) & 0xF8;
u32 g1 = (r4_ >> 3) & 0xFC;
u32 b1 = (r4_ << 3) & 0xFC;
r0 *= 30;
g0 *= 59;
b0 *= 11;
r1 *= 30;
g1 *= 59;
b1 *= 11;
r3 = (r0 + r1) / 2;
g3 = (g0 + g1) / 2;
b3 = (b0 + b1) / 2;
u32 r3 = (r0 + r1) / 2;
u32 g3 = (g0 + g1) / 2;
u32 b3 = (b0 + b1) / 2;
// r6 = 30; // r6 (bit offset within r8)
r8 = 0; // r8 (bitfield of color indexes)
pix_idx = r8;
i = r8;
pix_idx = i;
for (; i < 0x10; i++, pix_idx += 4) {
if (block[pix_idx+3] == 0xFF) {
// b2 = block[pix_idx + 2] * 11;
// r2 = block[pix_idx + 0] * 30;
// g2 = block[pix_idx + 1] * 59;
color_mags[0] = COLOR_SQUARE_MAG(r0, g0, b0, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[1] = COLOR_SQUARE_MAG(r1, g1, b1, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
color_mags[2] = COLOR_SQUARE_MAG(r3, g3, b3, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
// r4 = (r1 * 5 + r0 * 3) / 8;
// g4 = (g1 * 5 + g0 * 3) / 8;
// b4 = (b1 * 5 + b0 * 3) / 8;
// color_mags[3] = COLOR_SQUARE_MAG(r4, g4, b4, block[pix_idx+0] * 30, block[pix_idx+1] * 59, block[pix_idx+2] * 11);
r5 = 0;
r24 = 0;
r24 = r5;
r25 = INT32_MAX;
u32 n = 3 - r24;
for (; r24 < n; r24++) {
for (; r24 < 3; r24++) {
if (color_mags[r24] < r25) {
r25 = color_mags[r24];
r5 = r24;
@@ -1254,35 +1203,31 @@ u32 encode_s3tc(u8* r25, u8* r26, int r27, int r28, GXTexFmt fmt) {
// params: capture_buffer, texture_buffer, width, height, fmt
u8 block[0x40 + 0x8]; // sp08;
u32 r31 = 0;
u32 r5_r8_r4;
int r3_r9_r5;
if (fmt == GX_TF_I8) {
u32 i8low = 255; // r30
u32 i8high = 0; // r29
for (int r18 = 0; r18 < r28; r18 += 8) {
for (int r19 = 0, r24 = 0; r19 < r27; r19 += 8, r24 += 0x20) {
for (int r19 = 0; r19 < r27; r19 += 8) {
for (int r20 = 0; r20 < 8; r20 += 4) {
int r21 = 0;
u32 r23 = r24 + r27 * (r18 + r20);
for (; r21 < 8; r21 += 4) {
u32 r5 = 0;
int r3 = 0;
for (; r3 < 4; r3++) {
u32 r10 = r21 + r23;
u32 r9 = 219;
u32 r8 = 255; // r8 (alpha)
for (int i = 0; i < 4; r5++, i++) {
u32 r4 = r3*8 + r10;
u32 r11 = r25[r4 + i];
for (int r21 = 0; r21 < 8; r21 += 4) {
r5_r8_r4 = 0;
for (r3_r9_r5 = 0; r3_r9_r5 < 4; r3_r9_r5++) {
u32 r4 = (r19 * 4) + r27 * (r18 + r20) + r21 + r3_r9_r5*8;
for (int i = 0; i < 4; r5_r8_r4 += 4, i++, r4++) {
u32 r11 = r25[r4];
if (i8low > r11) {
i8low = r11;
}
if (i8high < r11) {
i8high = r11;
}
u32 r0 = ((r11 - 16) * 255) / r9;
block[r5*4+0] = r0 & 0xF8;
block[r5*4+1] = r0 & 0xFC;
block[r5*4+2] = r0 & 0xF8;
block[r5*4+3] = r8;
u32 r0 = ((r11 - 16) * 255) / 219;
block[r5_r8_r4+0] = r0 & 0xF8;
block[r5_r8_r4+1] = r0 & 0xFC;
block[r5_r8_r4+2] = r0 & 0xF8;
block[r5_r8_r4+3] = 255; // r8 (alpha)
}
}
blockenc(block);
@@ -1297,22 +1242,16 @@ u32 encode_s3tc(u8* r25, u8* r26, int r27, int r28, GXTexFmt fmt) {
for (int r24 = 0; r24 < r28; r24 += 8) {
for (int r29 = 0; r29 < r27; r29 += 8) {
for (int r30 = 0; r30 < 8; r30 += 4) {
int r21 = 0;
u32 r22 = 2 * (r27 * (r24 + r30));
for (; r21 < 8; r21 += 4) {
u32 r8 = 0;
int r9 = 0;
// u32 r3 = 0;
u32 r6 = r22 + 8 * (r29 + r21);
u32 r5 = 255; // r5 (alpha)
for (; r9 < 4; r9++) {//, r3 += 8) {
u32 r10 = r9*8 + r6;
for (int i = 0; i < 4; i++, r8++, r10 += 2) {
for (int r21 = 0; r21 < 8; r21 += 4) {
r5_r8_r4 = 0;
for (r3_r9_r5 = 0; r3_r9_r5 < 4; r3_r9_r5++) {
u32 r10 = ((r27 * (r24 + r30))*2) + (r29 + r21)*8 + r3_r9_r5*8;
for (int i = 0; i < 4; i++, r5_r8_r4 += 4, r10 += 2) {
u32 r11 = *(u16*)(&r25[r10]);
block[r8*4+0] = (r11 >> 8) & 0xF8;
block[r8*4+1] = (r11 >> 3) & 0xFC;
block[r8*4+2] = (r11 << 3) & 0xF8;
block[r8*4+3] = r5;
block[r5_r8_r4+0] = (r11 >> 8) & 0xF8;
block[r5_r8_r4+1] = (r11 >> 3) & 0xFC;
block[r5_r8_r4+2] = (r11 << 3) & 0xF8;
block[r5_r8_r4+3] = 255; // r5 (alpha)
}
}
blockenc(block);
@@ -1326,23 +1265,17 @@ u32 encode_s3tc(u8* r25, u8* r26, int r27, int r28, GXTexFmt fmt) {
for (int r24 = 0; r24 < r28; r24 += 8) {
for (int r29 = 0; r29 < r27; r29 += 8) {
for (int r30 = 0; r30 < 8; r30 += 4) {
int r21 = 0;
u32 r22 = 4 * (r27 * (r24 + r30));
for (; r21 < 8; r21 += 4) {
u32 r4 = 0;
for (int r5 = 0; r5 < 4; r5++) {
u32 r10 = r22 + 16 * (r29 + r21);
u32 r6 = r5 * 8 + r10;
for (int r21 = 0; r21 < 8; r21 += 4) {
r5_r8_r4 = 0;
for (r3_r9_r5 = 0; r3_r9_r5 < 4; r3_r9_r5++) {
u32 r6 = 16 * (r29 + r21) + (4 * (r27 * (r24 + r30))) + r3_r9_r5 * 8;
u32 r7 = r6 + 0x20;
u32 r9 = 255; // r9 (alpha)
for (int i = 0; i < 4; i++, r4++, r6 += 2, r7 += 2) {
u8* r3 = &r25[r6];
block[r4*4+0] = r3[1] & 0xF8;
u8* r12 = &r25[r7];
block[r4*4+1] = r12[0] & 0xFC;
block[r4*4+2] = r12[1] & 0xF8;
block[r4*4+3] = r3[0];
block[r4*4+3] = r9;
for (int i = 0; i < 4; i++, r5_r8_r4 += 4, r6 += 2, r7 += 2) {
block[r5_r8_r4+0] = r25[r6+1] & 0xF8;
block[r5_r8_r4+1] = r25[r7+0] & 0xFC;
block[r5_r8_r4+2] = r25[r7+1] & 0xF8;
block[r5_r8_r4+3] = r25[r6+0];
block[r5_r8_r4+3] = 255; // r9 (alpha)
}
}
blockenc(block);
@@ -1477,7 +1410,7 @@ out:
}
/* 8000AAC4-8000AB1C .text mCaptureProc__FPv */
u32 mCaptureProc(void* dummy) {
u32 mCaptureProc(void*) {
u32 bytesCopied = encode_s3tc(mCaptureCaptureBuffer, mCaptureTextureBuffer, mCaptureSizeWidth, mCaptureSizeHeight, (GXTexFmt)mCaptureCaptureFormat);
#if VERSION == VERSION_JPN
DCFlushRange(mCaptureTextureBuffer, mCaptureTextureSize);