add 3d layer h-scrolling

desmume/ChangeLog

@@ -53,6 +53,7 @@
  - Add MMU->GPU signal for when vram mappings change, which allows it to assume textures are unchanged unless vram has changed [zeromus]
  - Move lighting model to software instead of using opengl for more precision [zeromus]
  - Fix a bug in texture transformation mode 1 [zeromus]
+ - Add 3d layer h-scrolling [zeromus]
  Mac OS X port:
  - Fixed: Filenames and paths with unicode characters now work. [Jeff]
  - Fixed: Load state from file button works again. [Jeff]

desmume/src/GPU.cpp

@@ -746,7 +746,7 @@ INLINE void renderline_textBG(const GPU * gpu, u8 num, u8 * dst, u32 Y, u16 XBG,
 	u16 yoff;
 	u16 x      = 0;
 	u16 xfin;
-	u16 palette_size;
+	u16 palette_size_shift;
 	u16 mosaic = T1ReadWord((u8 *)&gpu->dispx_st->dispx_MISC.MOSAIC, 0);
 
 	s8 line_dir = 1;
@@ -897,10 +897,10 @@ INLINE void renderline_textBG(const GPU * gpu, u8 num, u8 * dst, u32 Y, u16 XBG,
 		return;
 	}
 
-	palette_size=0; // color: no extended palette
+	palette_size_shift=0; // color: no extended palette
 	if(dispCnt->ExBGxPalette_Enable)  // color: extended palette
 	{
-		palette_size=0x100;
+		palette_size_shift=8;
 		pal = ARM9Mem.ExtPal[gpu->core][gpu->BGExtPalSlot[num]];
 		if(!pal) return;
 	}
@@ -926,9 +926,10 @@ INLINE void renderline_textBG(const GPU * gpu, u8 num, u8 * dst, u32 Y, u16 XBG,
 		}
 		for(; x < xfin; )
 		{
-			color = T1ReadWord(pal, ((*line) + (tileentry.bits.Palette*palette_size)) << 1);
-			if (*line) 
+			color = T1ReadWord(pal, ((*line) + (tileentry.bits.Palette<<palette_size_shift)) << 1);
+			if (*line)
 				gpu->setFinalColorBck(gpu,0,num,dst,color,x,Y);
+			
 			dst += 2; x++; xoff++;
 
 			line += line_dir;
@@ -2166,7 +2167,15 @@ static INLINE void GPU_ligne_layer(NDS_Screen * screen, u16 l)
 					{
 						if (i16 == 0 && dispCnt->BG0_3D)
 						{
-							gpu3D->NDS_3D_GetLine (l, (u16*)dst);
+							//determine the 3d range to grab
+							BGxOFS * bgofs = &gpu->dispx_st->dispx_BGxOFS[i16];
+							s16 hofs = (s16)T1ReadWord((u8 *)&bgofs->BGxHOFS, 0);
+							int start, end, ofs;
+							if(hofs==0) { start = 0; end = 255; ofs = 0; }
+							else if(hofs<0) { start = -hofs; end=255; ofs=0; }
+							else { start = 0; end=255-hofs; ofs=hofs; }
+							
+							gpu3D->NDS_3D_GetLine (l, start, end, (u16*)dst+ofs);
 							continue;
 						}
 					}

desmume/src/OGLRender.cpp

@@ -1230,9 +1230,7 @@ static void GetLineCaptured(int line, u16* dst)
 }
 
 
-//NHerve mod3 - Fixed blending with 2D backgrounds (New Super Mario Bros looks better)
-//zeromus post-mod3: fix even better
-static void GetLine (int line, u16* dst)
+static void GetLine(int line, int start, int end_inclusive, u16* dst)
 {
 	assert(line<192 && line>=0);
 
@@ -1254,14 +1252,14 @@ static void GetLine (int line, u16* dst)
 	//this alpha compositing blending logic isnt thought through very much
 	//someone needs to think about what bitdepth it should take place at and how to do it efficiently
 
-	for(int i = 0; i < 256; i++)
+	for(int i = start, j=0; i <= end_inclusive; ++i, ++j)
 	{
 		u32 stencil = screenStencil[i];
 
 		//you would use this if you wanted to use the stencil buffer to make decisions here
 		if(!stencil) continue;
 
-		u16 oldcolor = dst[i];
+		u16 oldcolor = dst[j];
 		
 		int t=i<<2;
 		u32 dstpixel;
@@ -1296,7 +1294,7 @@ static void GetLine (int line, u16* dst)
 		newpix = mix[newpix][oldpix];
 		dstpixel |= (newpix<<10);
 
-		dst[i] = dstpixel;
+		dst[j] = dstpixel;
 	}
 }
 

desmume/src/render3D.cpp

@@ -23,6 +23,7 @@
 static void NDS_nullFunc1		(void){}
 static char NDS_nullFunc2		(void){ return 1; }
 static void NDS_nullFunc3		(int,unsigned short*) {}
+static void NDS_nullFunc4		(int,int,int,unsigned short*) {}
 
 GPU3DInterface gpu3DNull = { 
 	NDS_nullFunc2, //NDS_3D_Init
@@ -30,7 +31,7 @@ GPU3DInterface gpu3DNull = {
 	NDS_nullFunc1, //NDS_3D_Close
 	NDS_nullFunc1, //NDS_3D_Render
 	NDS_nullFunc1, //NDS_3D_VramReconfigureSignal
-	NDS_nullFunc3, //NDS_3D_GetLine
+	NDS_nullFunc4, //NDS_3D_GetLine
 	NDS_nullFunc3 //NDS_3D_GetLineCaptured
 };
 

desmume/src/render3D.h

@@ -42,7 +42,7 @@ typedef struct Render3DInterface
 	void (CALL_CONVENTION*  NDS_3D_VramReconfigureSignal)	();
 
 	//Retrieves a line of color buffer data
-	void (CALL_CONVENTION*	NDS_3D_GetLine)					(int line, unsigned short* dst);
+	void (CALL_CONVENTION*	NDS_3D_GetLine)					(int line, int start, int end_inclusive, unsigned short* dst);
 	
 	//Retrieves a line of color buffer data for capture
 	void (CALL_CONVENTION*	NDS_3D_GetLineCaptured)			(int line, unsigned short* dst);