Adapt to the great-refactor libctru branch

Makefile

@@ -28,6 +28,7 @@ INCLUDES	:=	include
 ARCH	:=	-march=armv6k -mtune=mpcore -mfloat-abi=hard
 
 CFLAGS	:=	-g -Wall -O2 -mword-relocations \
+			-ffunction-sections \
 			-fomit-frame-pointer -ffast-math \
 			$(ARCH)
 

examples/gputest/Makefile

@@ -46,7 +46,7 @@ CFLAGS	+=	$(INCLUDE) -DARM11 -D_3DS
 CXXFLAGS	:= $(CFLAGS) -fno-rtti -fno-exceptions -std=gnu++11
 
 ASFLAGS	:=	-g $(ARCH)
-LDFLAGS	=	-specs=3dsx.specs -g $(ARCH) -Wl,-Map,$(notdir $*.map)
+LDFLAGS	=	-specs=3dsx.specs -g $(ARCH) -Wl,-Map,$(notdir $*.map) -Wl,--gc-sections
 
 LIBS	:= -lcitro3d -lctru -lm
 

source/base.c

@@ -2,25 +2,6 @@
 
 C3D_Context __C3D_Context;
 
-u32 C3Di_Float24(float f)
-{
-	if (!f) return 0;
-	union { float t; u32 v; } u;
-	u.t = f;
-	u32 s = u.v >> 31;
-	u32 exp = ((u.v >> 23) & 0xFF) - 0x40;
-	u32 man = (u.v >> 7) & 0xFFFF;
-
-	return (exp >= 0) ? (man | (exp << 16) | (s << 23)) : (s << 23);
-}
-
-u32 C3Di_FloatInv24(u32 val)
-{
-	// Too lazy to copy & paste & cleanup the libctru function
-	extern u32 computeInvValue(u32 val);
-	return computeInvValue(val);
-}
-
 static void C3Di_SetTex(GPU_TEXUNIT unit, C3D_Tex* tex)
 {
 	u32 reg[4];
@@ -86,7 +67,6 @@ bool C3D_Init(size_t cmdBufSize)
 	ctx->cmdBuf = linearAlloc(cmdBufSize);
 	if (!ctx->cmdBuf) return false;
 
-	GPU_Reset(NULL, ctx->cmdBuf, ctx->cmdBufSize);
 	GPUCMD_SetBuffer(ctx->cmdBuf, ctx->cmdBufSize, 0);
 
 	ctx->flags = C3DiF_Active | C3DiF_TexEnvAll | C3DiF_Effect | C3DiF_TexAll;
@@ -116,10 +96,10 @@ void C3D_SetViewport(u32 x, u32 y, u32 w, u32 h)
 {
 	C3D_Context* ctx = C3Di_GetContext();
 	ctx->flags |= C3DiF_Viewport | C3DiF_Scissor;
-	ctx->viewport[0] = C3Di_Float24((float)w/2);
+	ctx->viewport[0] = f32tof24(w / 2.0f);
-	ctx->viewport[1] = C3Di_FloatInv24(w);
+	ctx->viewport[1] = f32tof31(2.0f / w) << 1;
-	ctx->viewport[2] = C3Di_Float24((float)h/2);
+	ctx->viewport[2] = f32tof24(h / 2.0f);
-	ctx->viewport[3] = C3Di_FloatInv24(h);
+	ctx->viewport[3] = f32tof31(2.0f / h) << 1;
 	ctx->viewport[4] = (y << 16) | (x & 0xFFFF);
 	ctx->scissor[0] = GPU_SCISSOR_DISABLE;
 }
@@ -199,7 +179,8 @@ void C3Di_UpdateContext(void)
 		}
 
 		ctx->flags &= ~C3DiF_TexAll;
-		GPU_SetTextureEnable(units);
+		GPUCMD_AddMaskedWrite(GPUREG_006F, 0x2, units<<8);         // enables texcoord outputs
+		GPUCMD_AddWrite(GPUREG_TEXUNITS_CONFIG, 0x00011000|units); // enables texture units
 	}
 
 	if (ctx->flags & C3DiF_TexEnvAll)
@@ -226,11 +207,13 @@ void C3D_FlushAsync(void)
 	if (ctx->flags & C3DiF_NeedFinishDrawing)
 	{
 		ctx->flags &= ~C3DiF_NeedFinishDrawing;
-		GPU_FinishDrawing();
+		GPUCMD_AddWrite(GPUREG_0111, 0x00000001);
+		GPUCMD_AddWrite(GPUREG_0110, 0x00000001);
+		GPUCMD_AddWrite(GPUREG_0063, 0x00000001);
 	}
 
 	GPUCMD_Finalize();
-	GPUCMD_FlushAndRun(NULL);
+	GPUCMD_FlushAndRun();
 	GPUCMD_SetBuffer(ctx->cmdBuf, ctx->cmdBufSize, 0);
 }
 

source/context.h

@@ -65,8 +65,6 @@ static inline C3D_Context* C3Di_GetContext(void)
 	return &__C3D_Context;
 }
 
-u32 C3Di_Float24(float f);
-u32 C3Di_FloatInv24(u32 val);
 void C3Di_UpdateContext(void);
 void C3Di_IBOBind(C3D_IBO* ibo);
 void C3Di_AttrInfoBind(C3D_AttrInfo* info);

source/effect.c

@@ -10,8 +10,8 @@ static inline C3D_Effect* getEffect()
 void C3D_DepthMap(float zScale, float zOffset)
 {
 	C3D_Effect* e = getEffect();
-	e->zScale = C3Di_Float24(zScale);
+	e->zScale  = f32tof24(zScale);
-	e->zOffset = C3Di_Float24(zOffset);
+	e->zOffset = f32tof24(zOffset);
 }
 
 void C3D_CullFace(GPU_CULLMODE mode)

source/ibo.c

@@ -27,7 +27,7 @@ bool C3D_IBOAddData(C3D_IBO* ibo, const void* data, int indexCount)
 void C3D_IBOFlush(C3D_IBO* ibo)
 {
 	int stride = ibo->format+1;
-	GSPGPU_FlushDataCache(NULL, ibo->data, ibo->indexCount*stride);
+	GSPGPU_FlushDataCache(ibo->data, ibo->indexCount*stride);
 }
 
 void C3Di_IBOBind(C3D_IBO* ibo)

source/renderbuffer.c

@@ -41,15 +41,15 @@ void C3D_RenderBufClearAsync(C3D_RenderBuf* rb)
 {
 	u32 colorBufSize = calcColorBufSize(rb->width, rb->height, rb->colorFmt);
 	u32 depthBufSize = calcDepthBufSize(rb->width, rb->height, rb->depthFmt);
-	GX_SetMemoryFill(NULL,
+	GX_SetMemoryFill(
 		(u32*)rb->colorBuf, rb->clearColor, (u32*)((u8*)rb->colorBuf+colorBufSize), BIT(0) | ((u32)colorFmtSizes[rb->colorFmt] << 8),
-		(u32*)rb->depthBuf, rb->clearDepth, (u32*)((u8*)rb->depthBuf+depthBufSize), BIT(0) | ((u32)colorFmtSizes[rb->depthFmt] << 8));
+		(u32*)rb->depthBuf, rb->clearDepth, (u32*)((u8*)rb->depthBuf+depthBufSize), BIT(0) | ((u32)depthFmtSizes[rb->depthFmt] << 8));
 }
 
 void C3D_RenderBufTransferAsync(C3D_RenderBuf* rb, u32* frameBuf, u32 flags)
 {
 	u32 dim = GX_BUFFER_DIM((u32)rb->width, (u32)rb->height);
-	GX_SetDisplayTransfer(NULL, (u32*)rb->colorBuf, dim, frameBuf, dim, flags);
+	GX_SetDisplayTransfer((u32*)rb->colorBuf, dim, frameBuf, dim, flags);
 }
 
 void C3D_RenderBufBind(C3D_RenderBuf* rb)

source/texture.c

@@ -69,7 +69,7 @@ void C3D_TexBind(int unitId, C3D_Tex* tex)
 void C3D_TexFlush(C3D_Tex* tex)
 {
 	if (tex->data)
-		GSPGPU_FlushDataCache(NULL, tex->data, tex->size);
+		GSPGPU_FlushDataCache(tex->data, tex->size);
 }
 
 void C3D_TexDelete(C3D_Tex* tex)

source/uniforms.c

@@ -36,7 +36,8 @@ void C3D_UpdateUniforms(GPU_SHADER_TYPE type)
 		*/
 
 		// Upload the uniforms
-		GPU_SetFloatUniform(type, i, (u32*)&C3D_FVUnif[i], j-i);
+		GPUCMD_AddWrite(GPUREG_VSH_FLOATUNIFORM_CONFIG+offset, 0x80000000|i);
+		GPUCMD_AddWrites(GPUREG_VSH_FLOATUNIFORM_DATA+offset, (u32*)&C3D_FVUnif[i], (j-i)*4);
 
 		// Clear the dirty flag
 		int k;

source/vbo.c

@@ -27,7 +27,7 @@ bool C3D_VBOAddData(C3D_VBO* vbo, const void* data, size_t size, int vertexCount
 void C3D_VBOFlush(C3D_VBO* vbo)
 {
 	if (vbo->data)
-		GSPGPU_FlushDataCache(NULL, vbo->data, vbo->size);
+		GSPGPU_FlushDataCache(vbo->data, vbo->size);
 }
 
 void C3D_VBOBind(C3D_VBO* vbo)