@@ -12,6 +12,11 @@ THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES

 WITH REGARD TO THIS SOFTWARE.

 */

+/*

+      fgbg fgbg

+byte [0000 0000]

+*/

+

 static Uint8 blending[5][16] = {

 	{0, 0, 0, 0, 1, 0, 1, 1, 2, 2, 0, 2, 3, 3, 3, 0},

 	{0, 1, 2, 3, 0, 1, 2, 3, 0, 1, 2, 3, 0, 1, 2, 3},

@@ -31,8 +36,8 @@ ppu_clear(Ppu *p)

 	int x, y;

 	for(y = 0; y < p->height; ++y) {

 		for(x = 0; x < p->width; ++x) {

-			ppu_write(p, p->bg, x, y, 0);

-			ppu_write(p, p->fg, x, y, 0);

+			ppu_write(p, 0, x, y, 0);

+			ppu_write(p, 1, x, y, 0);

 		}

 	}

 }

@@ -40,34 +45,30 @@ ppu_clear(Ppu *p)

 Uint8

 ppu_read(Ppu *p, Uint16 x, Uint16 y)

 {

-	Uint16 ch1, ch2, row = ppu_row(p, x, y);

-	ch1 = (p->fg[row] >> (7 - x % 8)) & 1;

-	ch2 = (p->fg[row + 8] >> (7 - x % 8)) & 1;

-	if(!ch1 && !ch2) {

-		ch1 = (p->bg[row] >> (7 - x % 8)) & 1;

-		ch2 = (p->bg[row + 8] >> (7 - x % 8)) & 1;

-	}

-	return ch1 + (ch2 << 1);

+	int row = (x + y * p->width) / 0x2;

+	int shift = (1 - (x & 0x1)) * 0x4;

+

+	return p->pixels[row] & 0x3;

 }

 void

-ppu_write(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 color)

+ppu_write(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 color)

 {

-	Uint16 row = ppu_row(p, x, y), col = 7 - (x % 8);

-	if(x >= p->width || y >= p->height)

-		return;

-	if(color == 0 || color == 2)

-		layer[row] &= ~(1UL << col);

-	else

-		layer[row] |= 1UL << col;

-	if(color == 0 || color == 1)

-		layer[row + 8] &= ~(1UL << col);

-	else

-		layer[row + 8] |= 1UL << col;

+	int row = (x + y * p->width) / 0x2;

+	int original = p->pixels[row];

+	Uint8 next = 0x0;

+	if(x % 2) {

+		next |= original & 0xf0;

+		next |= color << (layer * 2);

+	} else {

+		next |= original & 0x0f;

+		next |= color << (4 + (layer * 2));

+	}

+	p->pixels[row] = next;

 }

 void

-ppu_1bpp(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy)

+ppu_1bpp(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy)

 {

 	Uint16 v, h;

 	for(v = 0; v < 8; v++)

@@ -83,7 +84,7 @@ ppu_1bpp(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, U

 }

 void

-ppu_2bpp(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy)

+ppu_2bpp(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy)

 {

 	Uint16 v, h;

 	for(v = 0; v < 8; v++)

@@ -108,9 +109,9 @@ ppu_set_size(Ppu *p, Uint16 width, Uint16 height)

 	ppu_clear(p);

 	p->width = width;

 	p->height = height;

-	p->pixels = realloc(p->bg, p->width / 4 * p->height * sizeof(Uint8) * 2);

+	p->pixels = realloc(p->bg, p->width * p->height * sizeof(Uint8) * 2);

 	p->bg = p->pixels;

-	p->fg = p->pixels + (p->width / 4 * p->height * sizeof(Uint8));

+	p->fg = p->pixels + (p->width * p->height * sizeof(Uint8));

 	ppu_clear(p);

 	return p->bg && p->fg;

 }

@@ -27,9 +27,9 @@ typedef struct Ppu {

 } Ppu;

 Uint8 ppu_read(Ppu *p, Uint16 x, Uint16 y);

-void ppu_write(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 color);

+void ppu_write(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 color);

 void ppu_frame(Ppu *p);

-void ppu_1bpp(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy);

-void ppu_2bpp(Ppu *p, Uint8 *layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy);

+void ppu_1bpp(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy);

+void ppu_2bpp(Ppu *p, Uint8 layer, Uint16 x, Uint16 y, Uint8 *sprite, Uint8 color, Uint8 flipx, Uint8 flipy);

 int ppu_set_size(Ppu *p, Uint16 width, Uint16 height);

@@ -195,25 +195,25 @@ draw_inspect(Ppu *p, Uint8 *stack, Uint8 wptr, Uint8 rptr, Uint8 *memory)

 	for(i = 0; i < 0x20; ++i) {

 		x = ((i % 8) * 3 + 1) * 8, y = (i / 8 + 1) * 8, b = stack[i];

 		/* working stack */

-		ppu_1bpp(p, ppu.fg, x, y, font[(b >> 4) & 0xf], 1 + (wptr == i) * 0x7, 0, 0);

-		ppu_1bpp(p, ppu.fg, x + 8, y, font[b & 0xf], 1 + (wptr == i) * 0x7, 0, 0);

+		ppu_1bpp(p, 1, x, y, font[(b >> 4) & 0xf], 1 + (wptr == i) * 0x7, 0, 0);

+		ppu_1bpp(p, 1, x + 8, y, font[b & 0xf], 1 + (wptr == i) * 0x7, 0, 0);

 		y = 0x28 + (i / 8 + 1) * 8;

 		b = memory[i];

 		/* return stack */

-		ppu_1bpp(p, ppu.fg, x, y, font[(b >> 4) & 0xf], 3, 0, 0);

-		ppu_1bpp(p, ppu.fg, x + 8, y, font[b & 0xf], 3, 0, 0);

+		ppu_1bpp(p, 1, x, y, font[(b >> 4) & 0xf], 3, 0, 0);

+		ppu_1bpp(p, 1, x + 8, y, font[b & 0xf], 3, 0, 0);

 	}

 	/* return pointer */

-	ppu_1bpp(p, ppu.fg, 0x8, y + 0x10, font[(rptr >> 4) & 0xf], 0x2, 0, 0);

-	ppu_1bpp(p, ppu.fg, 0x10, y + 0x10, font[rptr & 0xf], 0x2, 0, 0);

+	ppu_1bpp(p, 1, 0x8, y + 0x10, font[(rptr >> 4) & 0xf], 0x2, 0, 0);

+	ppu_1bpp(p, 1, 0x10, y + 0x10, font[rptr & 0xf], 0x2, 0, 0);

 	/* guides */

 	for(x = 0; x < 0x10; ++x) {

-		ppu_write(p, ppu.fg, x, p->height / 2, 2);

-		ppu_write(p, ppu.fg, p->width - x, p->height / 2, 2);

-		ppu_write(p, ppu.fg, p->width / 2, p->height - x, 2);

-		ppu_write(p, ppu.fg, p->width / 2, x, 2);

-		ppu_write(p, ppu.fg, p->width / 2 - 0x10 / 2 + x, p->height / 2, 2);

-		ppu_write(p, ppu.fg, p->width / 2, p->height / 2 - 0x10 / 2 + x, 2);

+		ppu_write(p, 1, x, p->height / 2, 2);

+		ppu_write(p, 1, p->width - x, p->height / 2, 2);

+		ppu_write(p, 1, p->width / 2, p->height - x, 2);

+		ppu_write(p, 1, p->width / 2, x, 2);

+		ppu_write(p, 1, p->width / 2 - 0x10 / 2 + x, p->height / 2, 2);

+		ppu_write(p, 1, p->width / 2, p->height / 2 - 0x10 / 2 + x, 2);

 	}

 }

@@ -391,7 +391,7 @@ screen_talk(Device *d, Uint8 b0, Uint8 w)

 			Uint16 x = peek16(d->dat, 0x8);

 			Uint16 y = peek16(d->dat, 0xa);

 			Uint8 layer = d->dat[0xe] & 0x40;

-			ppu_write(&ppu, layer ? ppu.fg : ppu.bg, x, y, d->dat[0xe] & 0x3);

+			ppu_write(&ppu, layer, x, y, d->dat[0xe] & 0x3);

 			if(d->dat[0x6] & 0x01) poke16(d->dat, 0x8, x + 1); /* auto x+1 */

 			if(d->dat[0x6] & 0x02) poke16(d->dat, 0xa, y + 1); /* auto y+1 */

 			break;

@@ -402,10 +402,10 @@ screen_talk(Device *d, Uint8 b0, Uint8 w)

 			Uint8 layer = d->dat[0xf] & 0x40;

 			Uint8 *addr = &d->mem[peek16(d->dat, 0xc)];

 			if(d->dat[0xf] & 0x80) {

-				ppu_2bpp(&ppu, layer ? ppu.fg : ppu.bg, x, y, addr, d->dat[0xf] & 0xf, d->dat[0xf] & 0x10, d->dat[0xf] & 0x20);

+				ppu_2bpp(&ppu, layer, x, y, addr, d->dat[0xf] & 0xf, d->dat[0xf] & 0x10, d->dat[0xf] & 0x20);

 				if(d->dat[0x6] & 0x04) poke16(d->dat, 0xc, peek16(d->dat, 0xc) + 16); /* auto addr+16 */

 			} else {

-				ppu_1bpp(&ppu, layer ? ppu.fg : ppu.bg, x, y, addr, d->dat[0xf] & 0xf, d->dat[0xf] & 0x10, d->dat[0xf] & 0x20);

+				ppu_1bpp(&ppu, layer, x, y, addr, d->dat[0xf] & 0xf, d->dat[0xf] & 0x10, d->dat[0xf] & 0x20);

 				if(d->dat[0x6] & 0x04) poke16(d->dat, 0xc, peek16(d->dat, 0xc) + 8); /* auto addr+8 */

 			}

 			if(d->dat[0x6] & 0x01) poke16(d->dat, 0x8, x + 8); /* auto x+8 */