Aligned all vertex components to 4 byte boundaries - maybe ATI likes that? Renamed and reorganized misc things.

git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@981 8ced0084-cf51-0410-be5f-012b33b47a6e

Source/Core/VideoCommon/Src/NativeVertexFormat.h

@@ -75,7 +75,6 @@ public:
 
 	// TODO: move these in under private:
 	int m_VBVertexStride;  // PC-side vertex stride
-	int m_VBStridePad;
 	u32 m_components;  // VB_HAS_X. Bitmask telling what vertex components are present.
 };
 

Source/Core/VideoCommon/Src/TextureDecoder.cpp

@@ -284,15 +284,15 @@ inline void decodebytesARGB8_4(u32 *dst, const u16 *src, const u16 *src2)
 
 inline u32 makecol(int r, int g, int b, int a)
 {
-    return ((a&255)<<24)|((r&255)<<16)|((g&255)<<8)|((b&255));
+    return (a<<24)|(r<<16)|(g<<8)|b;
 }
 
 void decodeDXTBlock(u32 *dst, const DXTBlock *src, int pitch)
 {
     u16 c1 = Common::swap16(src->color1);
     u16 c2 = Common::swap16(src->color2);
-    int blue1 = lut5to8[c1&0x1F];
-    int blue2 = lut5to8[c2&0x1F];
+    int blue1 = lut5to8[c1 & 0x1F];
+    int blue2 = lut5to8[c2 & 0x1F];
     int green1 = lut6to8[(c1>>5) & 0x3F];
     int green2 = lut6to8[(c2>>5) & 0x3F];
     int red1 = lut5to8[(c1>>11) & 0x1F];
@@ -320,7 +320,7 @@ void decodeDXTBlock(u32 *dst, const DXTBlock *src, int pitch)
         int val = src->lines[y];
         for (int x = 0; x < 4; x++)
         {
-            dst[x] = colors[(val>>6) & 3];
+            dst[x] = colors[(val >> 6) & 3];
             val <<= 2;
         }
         dst += pitch;
@@ -515,11 +515,12 @@ PC_TexFormat TexDecoder_Decode(u8 *dst, const u8 *src, int width, int height, in
 			}
         }
         return PC_TEX_FMT_BGRA32;
-    case GX_TF_CMPR:
+    case GX_TF_CMPR:  // speed critical
         {
 			// TODO: Shuffle to PC S3TC (DXTC) format instead of converting
             // 11111111 22222222 55555555 66666666
             // 33333333 44444444 77777777 88888888
+			// The metroid games use this format almost exclusively.
             for (int y = 0; y < height; y += 8)
                 for (int x = 0; x < width; x += 8)
                 {

Source/Core/VideoCommon/Src/XFMemory.cpp

@@ -19,4 +19,4 @@
 
 // STATE_TO_SAVE
 XFRegisters xfregs;
-u32 xfmem[XFMEM_SIZE];
+u32 xfmem[XFMEM_SIZE];

Source/Core/VideoCommon/Src/XFMemory.h

@@ -173,6 +173,8 @@ struct XFRegisters
     ColorChannel colChans[2]; //C0A0 C1A1
     TexCoordInfo texcoords[8];
     bool bEnableDualTexTransform;
+	float rawViewport[6];
+	float rawProjection[7];
 };
 
 #define XFMEM_SIZE               0x8000

Source/Plugins/Plugin_VideoDX9/Plugin_VideoDX9.vcproj

@@ -1205,14 +1205,6 @@
 				RelativePath=".\Src\OpcodeDecoding.cpp"
 				>
 			</File>
-			<File
-				RelativePath=".\Src\VertexHandler.cpp"
-				>
-			</File>
-			<File
-				RelativePath=".\Src\VertexHandler.h"
-				>
-			</File>
 			<File
 				RelativePath=".\Src\VertexLoader.cpp"
 				>
@@ -1245,6 +1237,14 @@
 				RelativePath=".\Src\VertexLoader_TextCoord.h"
 				>
 			</File>
+			<File
+				RelativePath=".\Src\VertexManager.cpp"
+				>
+			</File>
+			<File
+				RelativePath=".\Src\VertexManager.h"
+				>
+			</File>
 			<File
 				RelativePath=".\Src\XFStructs.cpp"
 				>

Source/Plugins/Plugin_VideoDX9/Src/BPStructs.cpp

@@ -24,7 +24,7 @@
 #include "OpcodeDecoding.h"
 #include "TextureCache.h"
 #include "TextureDecoder.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "PixelShader.h"
 #include "Utils.h"
 
@@ -112,7 +112,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_GENMODE:
 		if (changes)
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			
 			// dev->SetRenderState(D3DRS_CULLMODE, d3dCullModes[bpmem.genMode.cullmode]);
@@ -146,7 +146,7 @@ void BPWritten(int addr, int changes, int newval)
     case BPMEM_IND_MTX+7:
     case BPMEM_IND_MTX+8:
         if (changes) {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
             // PixelShaderMngr::SetIndMatrixChanged((addr-BPMEM_IND_MTX)/3);
         }
@@ -154,7 +154,7 @@ void BPWritten(int addr, int changes, int newval)
     case BPMEM_RAS1_SS0:
     case BPMEM_RAS1_SS1:
         if (changes) {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
             // PixelShaderMngr::SetIndTexScaleChanged();
         }
@@ -163,7 +163,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_ZMODE:
 		if (changes)
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			if (bpmem.zmode.testenable)
 			{
@@ -190,7 +190,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_ALPHACOMPARE:
 		if (changes)
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			float f[4] =
 			{
@@ -217,7 +217,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_CONSTANTALPHA:
 		if (changes)
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			float f[4] = {
 				bpmem.dstalpha.alpha/255.0f,0,0,0
@@ -238,7 +238,7 @@ void BPWritten(int addr, int changes, int newval)
 	
 	case 0x43:
         if (changes) {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
         }
         break;
@@ -246,7 +246,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_BLENDMODE:	
 		if (changes & 0xFFFF)
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			if (changes & 1)
 			{
@@ -328,7 +328,7 @@ void BPWritten(int addr, int changes, int newval)
 		{
 			// u32 fogATemp = bpmem.fog.a<<12;
 			// float fogA = *(float*)(&fogATemp);
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
 		}
 		break;
@@ -336,7 +336,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_FOGBEXPONENT: 
 	case BPMEM_FOGBMAGNITUDE:
 		{
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
 		}
 		break;
@@ -345,7 +345,7 @@ void BPWritten(int addr, int changes, int newval)
 		//fog settings
 		if(changes) {
 			static bool bFog = false;
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			
 			if(!renderFog)
@@ -393,7 +393,7 @@ void BPWritten(int addr, int changes, int newval)
 
 	case BPMEM_FOGCOLOR:
 		 if(changes) {
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 
 			if(!renderFog)
@@ -412,7 +412,7 @@ void BPWritten(int addr, int changes, int newval)
 
 	case BPMEM_SCISSOROFFSET: //TODO: investigate
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 		}
 		break;
@@ -420,7 +420,7 @@ void BPWritten(int addr, int changes, int newval)
 	case BPMEM_SCISSORTL:
 	case BPMEM_SCISSORBR:
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[addr] = newval;
 			int xoff = bpmem.scissorOffset.x*2-342;
 			int yoff = bpmem.scissorOffset.y*2-342;
@@ -445,7 +445,7 @@ void BPWritten(int addr, int changes, int newval)
 		break;
 	case BPMEM_ZTEX1:
         if (changes) {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
             //PRIM_LOG("ztex bias=0x%x\n", bpmem.ztex1.bias);
             //PixelShaderMngr::SetZTextureBias(bpmem.ztex1.bias);
@@ -453,7 +453,7 @@ void BPWritten(int addr, int changes, int newval)
 		break;
 	case BPMEM_ZTEX2:
         if (changes) {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
 #ifdef _DEBUG
             const char* pzop[] = {"DISABLE", "ADD", "REPLACE", "?"};
@@ -473,7 +473,7 @@ void BPWritten(int addr, int changes, int newval)
     case 0xfd: // ksel7
         if (changes)
         {
-            CVertexHandler::Flush();
+            VertexManager::Flush();
             ((u32*)&bpmem)[addr] = newval;
             // PixelShaderMngr::SetTevKSelChanged(addr-0xf6);
         }
@@ -486,7 +486,7 @@ void BPWritten(int addr, int changes, int newval)
 		case 0xA0: 
 			if (changes)
 			{
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 				((u32*)&bpmem)[addr] = newval;
 				FourTexUnits &tex = bpmem.tex[(addr&0xE0)==0xA0];
 				int stage = (addr&3);//(addr>>4)&2;
@@ -534,7 +534,7 @@ void BPWritten(int addr, int changes, int newval)
 			if (changes)
 			{
 				textureChanged[((addr&0xE0)==0xA0)*4+(addr&3)] = true;
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 			}
 			break;
 		case 0x8C://TEX IMAGE 1
@@ -542,7 +542,7 @@ void BPWritten(int addr, int changes, int newval)
 			if (changes)
 			{
 				textureChanged[((addr&0xE0)==0xA0)*4+(addr&3)] = true;
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 			}
 			break;
 		case 0x90://TEX IMAGE 2
@@ -550,7 +550,7 @@ void BPWritten(int addr, int changes, int newval)
 			if (changes)
 			{
 				textureChanged[((addr&0xE0)==0xA0)*4+(addr&3)] = true;
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 			}
 			break;
 		case 0x94://TEX IMAGE 3
@@ -558,7 +558,7 @@ void BPWritten(int addr, int changes, int newval)
 			if (changes)
 			{
 				textureChanged[((addr&0xE0)==0xA0)*4+(addr&3)] = true;
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 			}
 			break;
 		case 0x98://TEX TLUT
@@ -566,7 +566,7 @@ void BPWritten(int addr, int changes, int newval)
 			if (changes)
 			{
 				textureChanged[((addr&0xE0)==0xA0)*4+(addr&3)] = true;
-				CVertexHandler::Flush();
+				VertexManager::Flush();
 			}
 			break;
 		case 0x9C://TEX UNKNOWN
@@ -588,7 +588,7 @@ void BPWritten(int addr, int changes, int newval)
             case 0xD0:
                 if (changes)
                 {
-                    CVertexHandler::Flush();
+                    VertexManager::Flush();
                     ((u32*)&bpmem)[addr] = newval;
                     // PixelShaderMngr::SetTevCombinerChanged((addr&0x1f)/2);
                 }
@@ -599,7 +599,7 @@ void BPWritten(int addr, int changes, int newval)
 				{
 					if (addr&1)
 					{
-						CVertexHandler::Flush();
+						VertexManager::Flush();
 						((u32*)&bpmem)[addr] = newval;
 						static int lastRGBA[2][4] = {
 							{0xEEEEEEEE, 0xEEEEEEEE, 0xEEEEEEEE, 0xEEEEEEEE},
@@ -616,7 +616,7 @@ void BPWritten(int addr, int changes, int newval)
 						int rgba = ((a<<24) | (r << 16) | (g << 8) | b) & 0xFCFCFCFC; //let's not detect minimal changes
 						if (rgba != lastRGBA[type][num])
 						{
-							CVertexHandler::Flush();
+							VertexManager::Flush();
 							lastRGBA[type][num] = rgba;
 							float temp[4] = {
 								r/255.0f, g/255.0f, b/255.0f, a/255.0f
@@ -634,7 +634,7 @@ void BPWritten(int addr, int changes, int newval)
 			default:
 				if (changes)
 				{
-					CVertexHandler::Flush();
+					VertexManager::Flush();
 					((u32*)&bpmem)[addr] = newval;
 				}
 				break;
@@ -665,7 +665,7 @@ void LoadBPReg(u32 value0)
 	switch (opcode)
 	{
 	case 0x45: //GXSetDrawDone
-		CVertexHandler::Flush();
+		VertexManager::Flush();
 		switch (value0 & 0xFF)
 		{
 		case 0x02:
@@ -694,7 +694,7 @@ void LoadBPReg(u32 value0)
 
 	case 0x52:
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 
 			((u32*)&bpmem)[opcode] = newval;
 			RECT rc = {
@@ -778,7 +778,7 @@ void LoadBPReg(u32 value0)
 
 	case 0x65: //GXLoadTlut
 		{
-			CVertexHandler::Flush();
+			VertexManager::Flush();
 			((u32*)&bpmem)[opcode] = newval;
 
 			u32 tlutTMemAddr = (value0&0x3FF)<<9;

Source/Plugins/Plugin_VideoDX9/Src/CPStructs.cpp

@@ -20,7 +20,7 @@
 #include "CPStructs.h"
 #include "XFStructs.h"
 #include "TransformEngine.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "VertexLoader.h"
 
 // PROBLEM - matrix switching within vbuffers may be stateful!
@@ -59,8 +59,12 @@ void LoadCPReg(u32 SubCmd, u32 Value)
 		CPUpdateMatricesB();
 		break;			
 
-	case 0x50: CVertexHandler::Flush(); VertexLoader::SetVtxDesc_Lo(Value); break;
-	case 0x60: CVertexHandler::Flush(); VertexLoader::SetVtxDesc_Hi(Value); break;
+	case 0x50:
+		VertexManager::Flush(); VertexLoader::SetVtxDesc_Lo(Value);
+		break;
+	case 0x60:
+		VertexManager::Flush(); VertexLoader::SetVtxDesc_Hi(Value);
+		break;
 
 	case 0x70: g_VertexLoaders[SubCmd & 7].SetVAT_group0(Value); _assert_((SubCmd & 0x0F) < 8); break;
 	case 0x80: g_VertexLoaders[SubCmd & 7].SetVAT_group1(Value); _assert_((SubCmd & 0x0F) < 8); break;

Source/Plugins/Plugin_VideoDX9/Src/CPStructs.h

@@ -1,3 +1,20 @@
+// Copyright (C) 2003-2008 Dolphin Project.
+
+// This program is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, version 2.0.
+
+// This program is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License 2.0 for more details.
+
+// A copy of the GPL 2.0 should have been included with the program.
+// If not, see http://www.gnu.org/licenses/
+
+// Official SVN repository and contact information can be found at
+// http://code.google.com/p/dolphin-emu/
+
 #ifndef _CPSTRUCTS_H
 #define _CPSTRUCTS_H
 

Source/Plugins/Plugin_VideoDX9/Src/OpcodeDecoding.cpp

@@ -29,7 +29,7 @@
 #include "Common.h"
 #include "Statistics.h"
 #include "Profiler.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "TransformEngine.h"
 #include "OpcodeDecoding.h"
 #include "TextureCache.h"
@@ -288,9 +288,9 @@ static void Decode(void)
 			int vsize = vtxLoader.GetVertexSize();
 			vtxLoader.RunVertices(numVertices);
 
-			// draw vertices
+			// add vertices
 			int primitive = (Cmd & GX_PRIMITIVE_MASK) >> GX_PRIMITIVE_SHIFT;
-			CVertexHandler::DrawVertices(primitive, numVertices, &tempvarray);
+			VertexManager::AddVertices(primitive, numVertices, &tempvarray);
 		}
 		else
 		{

Source/Plugins/Plugin_VideoDX9/Src/Render.cpp

@@ -22,7 +22,7 @@
 
 #include "Config.h"
 #include "main.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "Render.h"
 #include "OpcodeDecoding.h"
 #include "BPStructs.h"
@@ -105,7 +105,7 @@ void Renderer::Initialize(void)
 	Postprocess::Initialize();
 	Postprocess::BeginFrame();
 	D3D::BeginFrame(true, 0);
-	CVertexHandler::BeginFrame();
+	VertexManager::BeginFrame();
 }
 
 void Renderer::AddMessage(const std::string &message, u32 ms)
@@ -274,7 +274,7 @@ void Renderer::SwapBuffers(void)
 	// D3D::EnableAlphaToCoverage();
 
 	Postprocess::BeginFrame();
-	CVertexHandler::BeginFrame();
+	VertexManager::BeginFrame();
 
 	if (g_Config.bOldCard)
 		D3D::font.SetRenderStates(); //compatibility with low end cards

Source/Plugins/Plugin_VideoDX9/Src/TransformEngine.cpp

@@ -22,7 +22,7 @@
 // #include "Globals.h"
 #include "Vec3.h"
 #include "TransformEngine.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "VertexLoader.h"
 
 #include "BPStructs.h"

Source/Plugins/Plugin_VideoDX9/Src/TransformEngine.h

@@ -22,7 +22,7 @@
 //as much work as possible will be delegated to vertex shaders later
 //to take full advantage of current PC HW
 
-#include "VertexHandler.h"	
+#include "VertexManager.h"	
 #include "DecodedVArray.h"
 
 class CTransformEngine

Source/Plugins/Plugin_VideoDX9/Src/VertexLoader.cpp

@@ -22,7 +22,7 @@
 #include "Common.h"
 #include "LookUpTables.h"
 #include "Profiler.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "VertexLoader.h"
 #include "XFStructs.h"
 #include "BPStructs.h"
@@ -207,15 +207,6 @@ void VertexLoader::Setup()
 		              m_VtxAttr.texCoord[i].Format, 
 		              m_VtxAttr.texCoord[i].Elements,
 					  m_VtxAttr.texCoord[i].Frac);
-
-	Compile();
-	//RET();
-/*
-	char temp[256];
-	sprintf(temp,"Size: %08x Pos: %i Norm:%i Col0: %i Col1: %i",m_VertexSize,m_VtxDesc.Position,m_VtxDesc.Normal,m_VtxDesc.Color0,m_VtxDesc.Color1);
-	g_VideoInitialize.pLog(temp);
-	sprintf(temp,"Pos: %i Norm:%i Col0: %i Col1: %i",_VtxAttr.PosFormat,_VtxAttr.NormalFormat,_VtxAttr.color[0].Comp,_VtxAttr.color[1].Comp);
-	g_VideoInitialize.pLog(temp);*/
 }
 
 void VertexLoader::SetupColor(int num, int mode, int format, int elements)
@@ -280,14 +271,14 @@ void VertexLoader::SetupTexCoord(int num, int mode, int format, int elements, in
 			int sizePro=0;
 			switch (format)
 			{
-			case FORMAT_UBYTE:	sizePro=1; WriteCall(TexCoord_ReadDirect_UByte);  break;
-			case FORMAT_BYTE:	sizePro=1; WriteCall(TexCoord_ReadDirect_Byte);   break;
-			case FORMAT_USHORT:	sizePro=2; WriteCall(TexCoord_ReadDirect_UShort); break;
-			case FORMAT_SHORT:	sizePro=2; WriteCall(TexCoord_ReadDirect_Short);  break;
-			case FORMAT_FLOAT:	sizePro=4; WriteCall(TexCoord_ReadDirect_Float);  break;
+			case FORMAT_UBYTE:	sizePro = 1; WriteCall(TexCoord_ReadDirect_UByte);  break;
+			case FORMAT_BYTE:	sizePro = 1; WriteCall(TexCoord_ReadDirect_Byte);   break;
+			case FORMAT_USHORT:	sizePro = 2; WriteCall(TexCoord_ReadDirect_UShort); break;
+			case FORMAT_SHORT:	sizePro = 2; WriteCall(TexCoord_ReadDirect_Short);  break;
+			case FORMAT_FLOAT:	sizePro = 4; WriteCall(TexCoord_ReadDirect_Float);  break;
 			default: _assert_(0); break;
 			}
-			m_VertexSize += sizePro * (elements?2:1);
+			m_VertexSize += sizePro * (elements ? 2 : 1);
 		}
 		break;
 	case INDEX8:	
@@ -324,37 +315,6 @@ void VertexLoader::WriteCall(TPipelineFunction func)
 
 using namespace Gen;
 
-// See comment in RunVertices
-void VertexLoader::Compile()
-{
-	return;
-/*	Gen::SetCodePtr(m_compiledCode);
-	//INT3();
-	Gen::Util::EmitPrologue(0);
-	
-	const u8 *loopStart = Gen::GetCodePtr();
-	MOV(32, M(&tcIndex), Imm32(0));
-	MOV(32, M(&colIndex), Imm32(0));
-
-	for (int i = 0; i < m_numPipelineStates; i++)
-	{
-		ABI_AlignStack(1 * 4);
-		PUSH(32, Imm32((u32)&m_VtxAttr));
-		CALL(m_PipelineStates[i]);
-		ABI_RestoreStack(1 * 4);
-	}
-
-	ADD(32, M(&varray->count), Imm8(1));
-	SUB(32, M(&this->m_counter), Imm8(1));
-	J_CC(CC_NZ, loopStart, true);
-	// Epilogue
-	Gen::Util::EmitEpilogue(0);
-	if (Gen::GetCodePtr()-(u8*)m_compiledCode > sizeof(m_compiledCode))
-	{
-		Crash();
-	}*/
-}
-
 void VertexLoader::PrepareRun()
 {
 	posScale = shiftLookup[m_VtxAttr.PosFrac];

Source/Plugins/Plugin_VideoDX9/Src/VertexLoader.h

@@ -18,29 +18,6 @@
 #ifndef _VERTEXLOADER_H
 #define _VERTEXLOADER_H
 
-/*
-Ideas for new intermediate vertex format
-
-Produce a mask specifying with components are present.
-
-Decoder always produces a canonical format for all components:
-
-Ubyte4 for matrix indices (x,y,z,w) and colors
-
-FloatN for all others, texcoords are always XY at this stage
-
-The decoders will write a continuous stream of vertices, to take maximum
-advantage of write combining hardware
-
-a map<mask, vdecl> will keep track of vertex declarations
-
-this time, we are going to minimize transfers to the gfx card
-
-it turns out that with PS2.0 we can easily do all the pipeline in hardware.
-The decision will be how granular to be with the number of shaders and lighting settin
-
-*/
-
 int ComputeVertexSize(u32 components);
 
 #include "CPStructs.h"
@@ -97,7 +74,6 @@ private:
 	int m_numPipelineStates;	
 	int m_VertexSize;
 	int m_counter;
-	u8 m_compiledCode[1024];
 
 	u32 m_components;
 
@@ -106,7 +82,6 @@ private:
 	UVAT_group2 m_group2;
 
 	bool m_AttrDirty;
-
 	TVtxAttr m_VtxAttr; //Decoded into easy format
 
 	//common for all loaders
@@ -124,7 +99,6 @@ public:
 	// run the pipeline 
 	static void SetVArray(DecodedVArray *_varray);
 	void Setup();
-	void Compile();
 	void PrepareRun();
 	void RunVertices(int count);
 	void WriteCall(TPipelineFunction func);

Source/Plugins/Plugin_VideoDX9/Src/VertexManager.cpp

@@ -21,7 +21,7 @@
 
 #include "Statistics.h"
 #include "Profiler.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "OpcodeDecoding.h"
 #include "TransformEngine.h"
 #include "IndexGenerator.h"
@@ -32,13 +32,17 @@
 
 using namespace D3D;
 
-IndexGenerator indexGen;
-Collection CVertexHandler::collection;
 
-LPDIRECT3DVERTEXDECLARATION9 vDecl;
+namespace VertexManager
+{
 
-D3DVertex *fakeVBuffer;
-u16 *fakeIBuffer;
+static IndexGenerator indexGen;
+static Collection collection;
+
+static LPDIRECT3DVERTEXDECLARATION9 vDecl;
+
+static D3DVertex *fakeVBuffer;
+static u16 *fakeIBuffer;
 
 #define MAXVBUFFERSIZE 65536*3
 #define MAXIBUFFERSIZE 65536*3
@@ -60,23 +64,24 @@ const D3DVERTEXELEMENT9 decl[] =
     D3DDECL_END()
 };
 
-void CVertexHandler::Init()
+
+bool Init()
 {
 	collection = C_NOTHING;
 	fakeVBuffer = new D3DVertex[65536];
 	fakeIBuffer = new u16[65536];
 	CreateDeviceObjects();
+	return true;
 }
 
-
-void CVertexHandler::Shutdown()
+void Shutdown()
 {
 	DestroyDeviceObjects();
 	delete [] fakeVBuffer;
 	delete [] fakeIBuffer;
 }
 
-void CVertexHandler::CreateDeviceObjects()
+void CreateDeviceObjects()
 {
 	HRESULT hr;
 	if (FAILED(hr = D3D::dev->CreateVertexDeclaration(decl, &vDecl)))
@@ -86,13 +91,13 @@ void CVertexHandler::CreateDeviceObjects()
     }
 }
 
-void CVertexHandler::BeginFrame()
+void BeginFrame()
 {
 	D3D::dev->SetVertexDeclaration(vDecl);
 	//D3D::dev->SetStreamSource(0,vBuffer,0,sizeof(D3DVertex));
 }
 
-void CVertexHandler::DestroyDeviceObjects()
+void DestroyDeviceObjects()
 {
 	if (vDecl)
 		vDecl->Release();
@@ -100,7 +105,7 @@ void CVertexHandler::DestroyDeviceObjects()
 }
 
 
-void CVertexHandler::AddIndices(int _primitive, int _numVertices)
+void AddIndices(int _primitive, int _numVertices)
 {
 	switch(_primitive) {
 	case GX_DRAW_QUADS:          indexGen.AddQuads(_numVertices); return;    
@@ -127,7 +132,7 @@ const Collection collectionTypeLUT[8] =
 
 D3DVertex *vbufferwrite;
 
-void CVertexHandler::DrawVertices(int _primitive, int _numVertices, const DecodedVArray *varray)
+void AddVertices(int _primitive, int _numVertices, const DecodedVArray *varray)
 {
 	if (_numVertices <= 0) //This check is pretty stupid... 
 		return;
@@ -163,7 +168,7 @@ void CVertexHandler::DrawVertices(int _primitive, int _numVertices, const Decode
 		if (_numVertices >= MAXVBUFFERSIZE)
 			MessageBox(NULL, "To much vertices for the buffer", "Video.DLL", MB_OK);
 
-		CTransformEngine::TransformVertices(_numVertices,varray,vbufferwrite);
+		CTransformEngine::TransformVertices(_numVertices, varray, vbufferwrite);
 	}
 	else //We are collecting the right type, keep going
 	{
@@ -171,11 +176,11 @@ void CVertexHandler::DrawVertices(int _primitive, int _numVertices, const Decode
 		INCSTAT(stats.numJoins);
 		//Success, keep adding to unlocked buffer
 		int last=indexGen.GetNumVerts();
-		AddIndices(_primitive,_numVertices);
+		AddIndices(_primitive, _numVertices);
 
 		if (_numVertices >= MAXVBUFFERSIZE)
 			MessageBox(NULL, "Too many vertices for the buffer", "Video.DLL", MB_OK);
-		CTransformEngine::TransformVertices(_numVertices,varray,vbufferwrite + last);
+		CTransformEngine::TransformVertices(_numVertices, varray, vbufferwrite + last);
 	}
 }
 
@@ -186,7 +191,7 @@ const D3DPRIMITIVETYPE pts[3] =
 	D3DPT_LINELIST,
 };
 
-void CVertexHandler::Flush()
+void Flush()
 {
     DVSTARTPROFILE();
 
@@ -226,46 +231,4 @@ void CVertexHandler::Flush()
 	}
 }
 
-
-/*
-	char szTmp[256];
-	sprintf(szTmp, "Batch size : %i",_numVertices);
-	g_VideoInitialize.pLog(szTmp);
-
-
-	char szTemp[256];
-	sprintf(szTemp, "count: %i", count);
-	SendMessage(g_VideoInitialize.hStatusBar, SB_SETTEXT,0,(LPARAM)szTemp);	
-
-	static bool bWaitStep = false;
-	if (count == 24800)
-	{
-		bWaitStep = true;
-		Renderer::Flush();
-	}
-	if (GetAsyncKeyState('J'))
-		bWaitStep = true;
-	if (GetAsyncKeyState('U'))
-		bWaitStep = false;
-
-	if (bWaitStep)
-	{	
-		static bool bKey = false;
-		while (GetAsyncKeyState('H') == 0)
-		{			
-			if (GetAsyncKeyState(VK_SPACE))
-				_asm int 3;
-			if (GetAsyncKeyState('U'))
-			{
-				bWaitStep = false;
-				break;
-			}
-		}
-		Renderer::Flush();
-		// wait for key release
-		while (GetAsyncKeyState ('H')){
-		};
-	}
-}
-*/
-//	old = bpmem.combiners[0].colorC.hex;
+}  // namespace

Source/Plugins/Plugin_VideoDX9/Src/VertexManager.h

@@ -41,26 +41,21 @@ enum Collection
 	C_LINES=2,
 	C_POINTS=3
 };
+
+namespace VertexManager
+{
+
 extern const Collection collectionTypeLUT[8];
 
-class CVertexHandler
-{
-private:
-	static Collection collection;
-	// Pipeline
+bool Init();
+void Shutdown();
 
-	static void PrepareRender();
-	static void AddIndices(int _primitive, int _numVertices);
+void BeginFrame();
 
-public:
-	static void Init();
-	static void Shutdown();
+void CreateDeviceObjects();
+void DestroyDeviceObjects();
 
-	static void BeginFrame();
-	
-	static void CreateDeviceObjects();
-	static void DestroyDeviceObjects();
+void AddVertices(int _primitive, int _numVertices, const DecodedVArray *varray);
+void Flush();
 
-	static void DrawVertices(int _primitive, int _numVertices, const DecodedVArray *varray);
-	static void Flush();
-};
+}  // namespace

Source/Plugins/Plugin_VideoDX9/Src/XFStructs.cpp

@@ -20,18 +20,15 @@
 #include "XFStructs.h"
 #include "Render.h"
 #include "main.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "Utils.h"
 
-float rawViewPort[6];
-float rawProjection[7];
-
 // LoadXFReg 0x10
 void LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
 {	
     DVSTARTPROFILE();
 	u32 address = baseAddress;
-	for (int i=0; i<(int)transferSize; i++)
+	for (int i = 0; i < (int)transferSize; i++)
 	{
 		address = baseAddress + i;
 
@@ -68,15 +65,15 @@ void LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
 				break;
 
 			case 0x101a: 
-				CVertexHandler::Flush();
-				memcpy(rawViewPort, &pData[i], sizeof(rawViewPort)); 
+				VertexManager::Flush();
+				memcpy(xfregs.rawViewport, &pData[i], sizeof(xfregs.rawViewport)); 
 				XFUpdateVP(); 
 				i += 6;
 				break;
 
 			case 0x1020: 
-				CVertexHandler::Flush();
-				memcpy(rawProjection, &pData[i], sizeof(rawProjection)); 
+				VertexManager::Flush();
+				memcpy(xfregs.rawProjection, &pData[i], sizeof(xfregs.rawProjection)); 
 				XFUpdatePJ(); 
 				i += 7;
 				return;
@@ -130,10 +127,10 @@ void LoadIndexedXF(u32 val, int array)
 
 void XFUpdateVP()
 {
-	Renderer::SetViewport(rawViewPort); 
+	Renderer::SetViewport(xfregs.rawViewport); 
 }
 
 void XFUpdatePJ()
 {
-	Renderer::SetProjection(rawProjection, 0); 
+	Renderer::SetProjection(xfregs.rawProjection, 0); 
 }

Source/Plugins/Plugin_VideoDX9/Src/XFStructs.h

@@ -22,15 +22,9 @@
 #include "Vec3.h"
 #include "XFMemory.h"
 
-extern float rawViewPort[6];
-extern float rawProjection[7];
-
 void XFUpdateVP();
 void XFUpdatePJ();
 void LoadXFReg(u32 transferSize, u32 address, u32 *pData);
 void LoadIndexedXF(u32 val, int array);
 
-#pragma pack()
-
-
 #endif

Source/Plugins/Plugin_VideoDX9/Src/main.cpp

@@ -29,7 +29,7 @@
 #include "OpcodeDecoding.h"
 #include "TextureCache.h"
 #include "BPStructs.h"
-#include "VertexHandler.h"
+#include "VertexManager.h"
 #include "TransformEngine.h"
 #include "DlgSettings.h"
 #include "D3DPostprocess.h"
@@ -213,7 +213,7 @@ void Video_Prepare(void)
 	TextureCache::Init();
 
 	BPInit();
-	CVertexHandler::Init();
+	VertexManager::Init();
 	Fifo_Init();
 	OpcodeDecoder_Init();
 }
@@ -221,7 +221,7 @@ void Video_Prepare(void)
 void Video_Shutdown(void) 
 {
 	Fifo_Shutdown();
-	CVertexHandler::Shutdown();
+	VertexManager::Shutdown();
 	TextureCache::Shutdown();
 	Renderer::Shutdown();
 	OpcodeDecoder_Shutdown();

Source/Plugins/Plugin_VideoOGL/Src/BPStructs.cpp

@@ -52,13 +52,17 @@ static const GLenum glSrcFactors[8] =
 
 static const GLenum glDestFactors[8] = {
     GL_ZERO, GL_ONE, GL_SRC_COLOR, GL_ONE_MINUS_SRC_COLOR,
-    GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA,  GL_DST_ALPHA, GL_ONE_MINUS_DST_ALPHA };
+    GL_SRC_ALPHA, GL_ONE_MINUS_SRC_ALPHA,  GL_DST_ALPHA, GL_ONE_MINUS_DST_ALPHA
+};
 
-static const GLenum glCmpFuncs[8] =  { GL_NEVER, GL_LESS, GL_EQUAL, GL_LEQUAL, GL_GREATER, GL_NOTEQUAL, GL_GEQUAL, GL_ALWAYS };
+static const GLenum glCmpFuncs[8] = {
+	GL_NEVER, GL_LESS, GL_EQUAL, GL_LEQUAL, GL_GREATER, GL_NOTEQUAL, GL_GEQUAL, GL_ALWAYS
+};
 
 static const GLenum glLogicOpCodes[16] = {
     GL_CLEAR, GL_SET, GL_COPY, GL_COPY_INVERTED, GL_NOOP, GL_INVERT, GL_AND, GL_NAND,
-    GL_OR, GL_NOR, GL_XOR, GL_EQUIV, GL_AND_REVERSE, GL_AND_INVERTED, GL_OR_REVERSE, GL_OR_INVERTED };
+    GL_OR, GL_NOR, GL_XOR, GL_EQUIV, GL_AND_REVERSE, GL_AND_INVERTED, GL_OR_REVERSE, GL_OR_INVERTED
+};
 
 void BPInit()
 {
@@ -88,7 +92,7 @@ void BPWritten(int addr, int changes, int newval)
             // none, ccw, cw, ccw
             if (bpmem.genMode.cullmode>0) {
                 glEnable(GL_CULL_FACE);
-                glFrontFace(bpmem.genMode.cullmode==2?GL_CCW:GL_CW);
+                glFrontFace(bpmem.genMode.cullmode == 2 ? GL_CCW : GL_CW);
             }
             else glDisable(GL_CULL_FACE);
 
@@ -256,7 +260,7 @@ void BPWritten(int addr, int changes, int newval)
                     glBlendFunc(glSrcFactors[bpmem.blendmode.srcfactor], glDestFactors[bpmem.blendmode.dstfactor]);
             }
             if (changes & 0x800) {
-                glBlendEquation(bpmem.blendmode.subtract?GL_FUNC_REVERSE_SUBTRACT:GL_FUNC_ADD);
+                glBlendEquation(bpmem.blendmode.subtract ? GL_FUNC_REVERSE_SUBTRACT : GL_FUNC_ADD);
                 if (bpmem.blendmode.subtract)
                     glBlendFunc(GL_ONE, GL_ONE);
                 else
@@ -269,7 +273,7 @@ void BPWritten(int addr, int changes, int newval)
         break;
 
 	case BPMEM_FOGRANGE:
-		if(changes) {
+		if (changes) {
 			// TODO(XK): Fog range format
 			//glFogi(GL_FOG_START, ...
 			//glFogi(GL_FOG_END, ...
@@ -474,25 +478,25 @@ void BPWritten(int addr, int changes, int newval)
 
                 VertexShaderMngr::SetViewportChanged();
 
-                // since clear operations use the source rectangle, have to do regular renders (glClear clears the entire buffer)
-                if (bpmem.blendmode.colorupdate || bpmem.blendmode.alphaupdate || bpmem.zmode.updateenable) {
-                    
+                // Since clear operations use the source rectangle, we have to do
+				// regular renders (glClear clears the entire buffer)
+                if (bpmem.blendmode.colorupdate || bpmem.blendmode.alphaupdate || bpmem.zmode.updateenable)
+				{                    
                     GLbitfield bits = 0;
                     if (bpmem.blendmode.colorupdate || bpmem.blendmode.alphaupdate) {
-                        u32 clearColor = (bpmem.clearcolorAR<<16)|bpmem.clearcolorGB;
-                        glClearColor(((clearColor>>16)&0xff)*(1/255.0f),((clearColor>>8)&0xff)*(1/255.0f),
-                            ((clearColor>>0)&0xff)*(1/255.0f),((clearColor>>24)&0xff)*(1/255.0f));
+                        u32 clearColor = (bpmem.clearcolorAR << 16) | bpmem.clearcolorGB;
+                        glClearColor(((clearColor>>16) & 0xff)*(1/255.0f),
+							         ((clearColor>>8 ) & 0xff)*(1/255.0f),
+                                     ((clearColor>>0 ) & 0xff)*(1/255.0f),
+									 ((clearColor>>24) & 0xff)*(1/255.0f));
                         bits |= GL_COLOR_BUFFER_BIT;
                     }
-
                     if (bpmem.zmode.updateenable) {
                         glClearDepth((float)(bpmem.clearZValue&0xFFFFFF) / float(0xFFFFFF));
                         bits |= GL_DEPTH_BUFFER_BIT;
                     }
-
                     if (nRestoreZBufferTarget )
                         glDrawBuffer(GL_COLOR_ATTACHMENT0_EXT); // don't clear ztarget here
-                    
                     glClear(bits);
                 }
 
@@ -500,12 +504,12 @@ void BPWritten(int addr, int changes, int newval)
 
                     glDrawBuffer(GL_COLOR_ATTACHMENT1_EXT);
                     GL_REPORT_ERRORD();
-
                     glColorMask(GL_TRUE,GL_TRUE,GL_TRUE,GL_TRUE);
                     
                     // red should probably be the LSB
-                    glClearColor(((bpmem.clearZValue>>0)&0xff)*(1/255.0f),((bpmem.clearZValue>>8)&0xff)*(1/255.0f),
-                        ((bpmem.clearZValue>>16)&0xff)*(1/255.0f), 0);
+                    glClearColor(((bpmem.clearZValue>>0)&0xff)*(1/255.0f),
+						         ((bpmem.clearZValue>>8)&0xff)*(1/255.0f),
+                                 ((bpmem.clearZValue>>16)&0xff)*(1/255.0f), 0);
                     glClear(GL_COLOR_BUFFER_BIT);
                     SetColorMask();
                     GL_REPORT_ERRORD();   
@@ -547,7 +551,7 @@ void BPWritten(int addr, int changes, int newval)
             {
                 VertexManager::Flush();
                 ((u32*)&bpmem)[addr] = newval;
-                PixelShaderMngr::SetTevOrderChanged(addr-0x28);
+                PixelShaderMngr::SetTevOrderChanged(addr - 0x28);
             }
             break;
 

Source/Plugins/Plugin_VideoOGL/Src/NativeVertexFormat.cpp

@@ -75,9 +75,8 @@ void NativeVertexFormat::Initialize(const TVtxDesc &vtx_desc, const TVtxAttr &vt
 	DVSTARTPROFILE();
 
 	if (m_VBVertexStride & 3) {
-		// make sure all strides are at least divisible by 4 (some gfx cards experience a 3x speed boost)
-		m_VBStridePad = 4 - (m_VBVertexStride & 3);
-		m_VBVertexStride += m_VBStridePad;
+		// We will not allow vertex components causing uneven strides.
+		PanicAlert("Uneven vertex stride: %i", m_VBVertexStride);
 	}
 
 	// compile the pointer set function - why?
@@ -86,9 +85,9 @@ void NativeVertexFormat::Initialize(const TVtxDesc &vtx_desc, const TVtxAttr &vt
 	Util::EmitPrologue(6);
 	int offset = 0;
 	
-	// Position
+	// Position, part 1
 	if (vtx_desc.Position != NOT_PRESENT) {  // TODO: Why the check? Always present, AFAIK!
-		CallCdeclFunction4_I(glVertexPointer, 3, GL_FLOAT, m_VBVertexStride, offset);
+		CallCdeclFunction4_I(glVertexPointer, 3, GL_FLOAT, m_VBVertexStride, 0);
 		offset += 12;
 	}
 
@@ -97,18 +96,18 @@ void NativeVertexFormat::Initialize(const TVtxDesc &vtx_desc, const TVtxAttr &vt
 		switch (vtx_attr.NormalFormat) {
 		case FORMAT_UBYTE:	
 		case FORMAT_BYTE:
-			CallCdeclFunction3_I(glNormalPointer, GL_BYTE, m_VBVertexStride, offset); offset += 3;
+			CallCdeclFunction3_I(glNormalPointer, GL_BYTE, m_VBVertexStride, offset); offset += 4;
 			if (vtx_attr.NormalElements) {
-				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM1_ATTRIB, 3, GL_BYTE, GL_TRUE, m_VBVertexStride, offset); offset += 3;
-				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM2_ATTRIB, 3, GL_BYTE, GL_TRUE, m_VBVertexStride, offset); offset += 3;
+				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM1_ATTRIB, 4, GL_BYTE, GL_TRUE, m_VBVertexStride, offset); offset += 4;
+				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM2_ATTRIB, 4, GL_BYTE, GL_TRUE, m_VBVertexStride, offset); offset += 4;
 			}
 			break;
 		case FORMAT_USHORT:
 		case FORMAT_SHORT:
-			CallCdeclFunction3_I(glNormalPointer, GL_SHORT, m_VBVertexStride, offset); offset += 6;
+			CallCdeclFunction3_I(glNormalPointer, GL_SHORT, m_VBVertexStride, offset); offset += 8;
 			if (vtx_attr.NormalElements) {
-				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM1_ATTRIB, 3, GL_SHORT, GL_TRUE, m_VBVertexStride, offset); offset += 6;
-				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM2_ATTRIB, 3, GL_SHORT, GL_TRUE, m_VBVertexStride, offset); offset += 6;
+				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM1_ATTRIB, 4, GL_SHORT, GL_TRUE, m_VBVertexStride, offset); offset += 8;
+				CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_NORM2_ATTRIB, 4, GL_SHORT, GL_TRUE, m_VBVertexStride, offset); offset += 8;
 			}
 			break;
 		case FORMAT_FLOAT:
@@ -164,19 +163,19 @@ void NativeVertexFormat::Initialize(const TVtxDesc &vtx_desc, const TVtxAttr &vt
 				}
 				else {
 					CallCdeclFunction4_I(glTexCoordPointer, 3, GL_SHORT, m_VBVertexStride, offset);
-					offset += 6;
+					offset += 8;
 				}
 			}
 			else {
 				CallCdeclFunction4_I(glTexCoordPointer, vtx_attr.texCoord[i].Elements ? 2 : 1, GL_FLOAT, m_VBVertexStride, offset);
-				offset += 4 * (vtx_attr.texCoord[i].Elements?2:1);
+				offset += 4 * (vtx_attr.texCoord[i].Elements ? 2 : 1);
 			}
 		}
 	}
 
 	if (vtx_desc.PosMatIdx) {
-		CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_POSMTX_ATTRIB, 1, GL_UNSIGNED_BYTE, GL_FALSE, m_VBVertexStride, offset);
-		offset += 1;
+		CallCdeclFunction6((void *)glVertexAttribPointer, SHADER_POSMTX_ATTRIB, 4, GL_UNSIGNED_BYTE, GL_FALSE, m_VBVertexStride, offset);
+		offset += 4;
 	}
 
 	_assert_(offset + m_VBStridePad == m_VBVertexStride);

Source/Plugins/Plugin_VideoOGL/Src/OpcodeDecoding.cpp

@@ -224,22 +224,22 @@ static void Decode()
             static u32 pData[16];
             for (int i = 0; i < dwTransferSize; i++)
                 pData[i] = DataReadU32();
-            VertexShaderMngr::LoadXFReg(dwTransferSize, dwAddress, pData);
+            LoadXFReg(dwTransferSize, dwAddress, pData);
 			INCSTAT(stats.thisFrame.numXFLoads);
         }
         break;
 
     case GX_LOAD_INDX_A: //used for position matrices
-        VertexShaderMngr::LoadIndexedXF(DataReadU32(), 0xC);
+        LoadIndexedXF(DataReadU32(), 0xC);
         break;
     case GX_LOAD_INDX_B: //used for normal matrices
-        VertexShaderMngr::LoadIndexedXF(DataReadU32(), 0xD);
+        LoadIndexedXF(DataReadU32(), 0xD);
         break;
     case GX_LOAD_INDX_C: //used for postmatrices
-        VertexShaderMngr::LoadIndexedXF(DataReadU32(), 0xE);
+        LoadIndexedXF(DataReadU32(), 0xE);
         break;
     case GX_LOAD_INDX_D: //used for lights
-        VertexShaderMngr::LoadIndexedXF(DataReadU32(), 0xF);
+        LoadIndexedXF(DataReadU32(), 0xF);
         break;
 
     case GX_CMD_CALL_DL:

Source/Plugins/Plugin_VideoOGL/Src/TextureMngr.cpp

@@ -67,28 +67,32 @@ const GLint c_MinLinearFilter[8] = {
 
 const GLint c_WrapSettings[4] = { GL_CLAMP_TO_EDGE, GL_REPEAT, GL_MIRRORED_REPEAT, GL_REPEAT };
 
-void TextureMngr::TCacheEntry::SetTextureParameters(TexMode0& newmode)
+void TextureMngr::TCacheEntry::SetTextureParameters(TexMode0 &newmode)
 {
     mode = newmode;
-    if( isNonPow2 ) {
+    if (isNonPow2) {
         // very limited!
-        glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MAG_FILTER, (newmode.mag_filter||g_Config.bForceFiltering)?GL_LINEAR:GL_NEAREST);
-        glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MIN_FILTER, (g_Config.bForceFiltering||newmode.min_filter>=4)?GL_LINEAR:GL_NEAREST);
-		if( newmode.wrap_s == 2 || newmode.wrap_t == 2 ) {
+        glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MAG_FILTER,
+			            (newmode.mag_filter || g_Config.bForceFiltering) ? GL_LINEAR : GL_NEAREST);
+        glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MIN_FILTER,
+			            (g_Config.bForceFiltering || newmode.min_filter >= 4) ? GL_LINEAR : GL_NEAREST);
+		if (newmode.wrap_s == 2 || newmode.wrap_t == 2) {
             DEBUG_LOG("cannot support mirrorred repeat mode\n");
 		}
     }
     else {
-        glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER, (newmode.mag_filter||g_Config.bForceFiltering)?GL_LINEAR:GL_NEAREST);
+        glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MAG_FILTER,
+			            (newmode.mag_filter || g_Config.bForceFiltering) ? GL_LINEAR : GL_NEAREST);
 
-        if( bHaveMipMaps ) {
+        if (bHaveMipMaps) {
             int filt = newmode.min_filter;
-            if( g_Config.bForceFiltering && newmode.min_filter < 4 )
+            if (g_Config.bForceFiltering && newmode.min_filter < 4)
                 newmode.min_filter += 4; // take equivalent forced linear
             glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, c_MinLinearFilter[filt]);
         }
         else
-            glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER, (g_Config.bForceFiltering||newmode.min_filter>=4)?GL_LINEAR:GL_NEAREST);
+            glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_MIN_FILTER,
+			                (g_Config.bForceFiltering || newmode.min_filter >= 4) ? GL_LINEAR : GL_NEAREST);
 
         glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_S, c_WrapSettings[newmode.wrap_s]);
         glTexParameteri(GL_TEXTURE_2D, GL_TEXTURE_WRAP_T, c_WrapSettings[newmode.wrap_t]);
@@ -195,7 +199,7 @@ TextureMngr::TCacheEntry* TextureMngr::Load(int texstage, u32 address, int width
         return NULL;
 
     TexCache::iterator iter = textures.find(address);
-    TexMode0 &tm0 = bpmem.tex[texstage>3].texMode0[texstage&3];
+    TexMode0 &tm0 = bpmem.tex[texstage > 3].texMode0[texstage & 3];
     u8 *ptr = g_VideoInitialize.pGetMemoryPointer(address);
 
     int palSize = TexDecoder_GetPaletteSize(format);
@@ -248,9 +252,9 @@ TextureMngr::TCacheEntry* TextureMngr::Load(int texstage, u32 address, int width
         }
     }
     
-    int bs = TexDecoder_GetBlockWidthInTexels(format)-1; //TexelSizeInNibbles(format)*width*height/16;
-    int expandedWidth = (width+bs) & (~bs);
-    PC_TexFormat dfmt = TexDecoder_Decode(temp,ptr,expandedWidth,height,format, tlutaddr, tlutfmt);
+    int bs = TexDecoder_GetBlockWidthInTexels(format) - 1; //TexelSizeInNibbles(format)*width*height/16;
+    int expandedWidth = (width + bs) & (~bs);
+    PC_TexFormat dfmt = TexDecoder_Decode(temp, ptr, expandedWidth, height, format, tlutaddr, tlutfmt);
 
     //Make an entry in the table
     TCacheEntry& entry = textures[address];
@@ -283,7 +287,7 @@ TextureMngr::TCacheEntry* TextureMngr::Load(int texstage, u32 address, int width
 		gl_type = GL_UNSIGNED_BYTE;
 		break;
 	}
-    if( !entry.isNonPow2 && ((tm0.min_filter&3)==1||(tm0.min_filter&3)==2) ) {
+    if (!entry.isNonPow2 && ((tm0.min_filter & 3) == 1 || (tm0.min_filter & 3) == 2)) {
         gluBuild2DMipmaps(GL_TEXTURE_2D, 4, width, height, gl_format, gl_type, temp);
         entry.bHaveMipMaps = true;
     }
@@ -294,9 +298,9 @@ TextureMngr::TCacheEntry* TextureMngr::Load(int texstage, u32 address, int width
         glPixelStorei(GL_UNPACK_ROW_LENGTH, 0);
 
     entry.frameCount = frameCount;
-    entry.w=width;
-    entry.h=height;
-    entry.fmt=format;
+    entry.w = width;
+    entry.h = height;
+    entry.fmt = format;
     entry.SetTextureParameters(tm0);
 
     if (g_Config.bDumpTextures) { // dump texture to file
@@ -346,7 +350,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
 
     GL_REPORT_ERRORD();
 
-    if( !bIsInit ) {
+    if (!bIsInit) {
         glGenTextures(1, (GLuint *)&entry.texture);
         glBindTexture(GL_TEXTURE_RECTANGLE_ARB, entry.texture);
         glTexImage2D(GL_TEXTURE_RECTANGLE_ARB, 0, 4, w, h, 0, GL_RGBA, GL_UNSIGNED_BYTE, NULL);
@@ -356,16 +360,16 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
         _assert_(entry.texture);
         bool bReInit = true;
 
-        if( entry.w == w && entry.h == h ) {
+        if (entry.w == w && entry.h == h) {
             glBindTexture(GL_TEXTURE_RECTANGLE_ARB, entry.texture);
             // for some reason mario sunshine errors here...
             GLenum err = GL_NO_ERROR;
             GL_REPORT_ERROR();
-            if( err == GL_NO_ERROR )
+            if (err == GL_NO_ERROR )
                 bReInit = false;
         }
 
-        if( bReInit ) {
+        if (bReInit) {
             // necessary, for some reason opengl gives errors when texture isn't deleted
             glDeleteTextures(1,(GLuint *)&entry.texture);
             glGenTextures(1, (GLuint *)&entry.texture);
@@ -375,13 +379,13 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
         }
     }
 
-    if( !bIsInit || !entry.isRenderTarget ) {
+    if (!bIsInit || !entry.isRenderTarget) {
         glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MAG_FILTER, GL_LINEAR);
         glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_MIN_FILTER, GL_LINEAR);
 
         glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);
         glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);
-        if( glGetError() != GL_NO_ERROR) {
+        if (glGetError() != GL_NO_ERROR) {
             glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_S, GL_CLAMP);
             glTexParameteri(GL_TEXTURE_RECTANGLE_ARB, GL_TEXTURE_WRAP_T, GL_CLAMP);
             GL_REPORT_ERRORD();
@@ -390,14 +394,14 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
     
     entry.w = w;
     entry.h = h;
-    entry.isRenderTarget=true;
+    entry.isRenderTarget = true;
     entry.fmt = copyfmt;
 
     float colmat[16];
     float fConstAdd[4] = {0};
     memset(colmat, 0, sizeof(colmat));
 
-    if( bFromZBuffer ) {
+    if (bFromZBuffer) {
         switch(copyfmt) {
             case 0: // Z4
             case 1: // Z8
@@ -428,18 +432,19 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
                 break;
         }
     }
-    else if( bIsIntensityFmt ) {
+    else if (bIsIntensityFmt) {
         fConstAdd[0] = fConstAdd[1] = fConstAdd[2] = 16.0f/255.0f;
         switch(copyfmt) {
             case 0: // I4
             case 1: // I8
             case 2: // IA4
             case 3: // IA8
+				// TODO - verify these coefficients
                 colmat[0] = 0.257f; colmat[1] = 0.504f; colmat[2] = 0.098f;
                 colmat[4] = 0.257f; colmat[5] = 0.504f; colmat[6] = 0.098f;
                 colmat[8] = 0.257f; colmat[9] = 0.504f; colmat[10] = 0.098f;
-                if( copyfmt < 2 ) {
-                    fConstAdd[3] = 16.0f/255.0f;
+                if (copyfmt < 2) {
+                    fConstAdd[3] = 16.0f / 255.0f;
                     colmat[12] = 0.257f; colmat[13] = 0.504f; colmat[14] = 0.098f;
                 }
                 else { // alpha
@@ -453,7 +458,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
         }
     }
     else {
-        switch(copyfmt) {
+        switch (copyfmt) {
             case 0: // R4
             case 8: // R8
                 colmat[0] = colmat[4] = colmat[8] = colmat[12] = 1;
@@ -495,7 +500,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
         }
     }
 
-//    if( bCopyToTarget ) {
+//    if (bCopyToTarget) {
 //        _assert_( glCheckFramebufferStatusEXT(GL_FRAMEBUFFER_EXT) == GL_FRAMEBUFFER_COMPLETE_EXT );
 //        glReadBuffer(GL_COLOR_ATTACHMENT0_EXT);
 //        GL_REPORT_ERRORD();
@@ -512,7 +517,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
 
     Renderer::ResetGLState(); // reset any game specific settings
 
-    if( s_TempFramebuffer == 0 )
+    if (s_TempFramebuffer == 0 )
         glGenFramebuffersEXT( 1, (GLuint *)&s_TempFramebuffer);
 
     Renderer::SetFramebuffer(s_TempFramebuffer);
@@ -525,7 +530,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
     if (itdepth == mapDepthTargets.end()) {
         DEPTHTARGET& depth = mapDepthTargets[(h << 16) | w];
         depth.framecount = frameCount;
-        glGenRenderbuffersEXT( 1, &depth.targ);
+        glGenRenderbuffersEXT(1, &depth.targ);
         glBindRenderbufferEXT(GL_RENDERBUFFER_EXT, depth.targ);
         glRenderbufferStorageEXT(GL_RENDERBUFFER_EXT, GL_DEPTH_COMPONENT/*GL_DEPTH24_STENCIL8_EXT*/, w, h);
         GL_REPORT_ERRORD();
@@ -565,7 +570,7 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
 
     TextureMngr::DisableStage(0);
 
-    if( bFromZBuffer )
+    if (bFromZBuffer )
         Renderer::SetZBufferRender(); // notify for future settings
 
     GL_REPORT_ERRORD();
@@ -575,11 +580,11 @@ void TextureMngr::CopyRenderTargetToTexture(u32 address, bool bFromZBuffer, bool
 
 void TextureMngr::EnableTex2D(int stage)
 {
-    if( !(nTex2DEnabled & (1<<stage)) ) {
+    if (!(nTex2DEnabled & (1<<stage))) {
         nTex2DEnabled |= (1<<stage);
         glEnable(GL_TEXTURE_2D);
     }
-    if( nTexRECTEnabled & (1<<stage) ) {
+    if (nTexRECTEnabled & (1<<stage)) {
         nTexRECTEnabled &= ~(1<<stage);
         glDisable(GL_TEXTURE_RECTANGLE_ARB);
     }
@@ -587,12 +592,12 @@ void TextureMngr::EnableTex2D(int stage)
 
 void TextureMngr::EnableTexRECT(int stage)
 {
-    if( (nTex2DEnabled & (1<<stage)) ) {
-        nTex2DEnabled &= ~(1<<stage);
+    if ((nTex2DEnabled & (1 << stage))) {
+        nTex2DEnabled &= ~(1 << stage);
         glDisable(GL_TEXTURE_2D);
     }
-    if( !(nTexRECTEnabled & (1<<stage)) ) {
-        nTexRECTEnabled |= (1<<stage);
+    if (!(nTexRECTEnabled & (1 << stage))) {
+        nTexRECTEnabled |= (1 << stage);
         glEnable(GL_TEXTURE_RECTANGLE_ARB);
     }
 }
@@ -600,15 +605,16 @@ void TextureMngr::EnableTexRECT(int stage)
 void TextureMngr::DisableStage(int stage)
 {
     bool bset = false;
-    if( nTex2DEnabled & (1<<stage) ) {
-        nTex2DEnabled &= ~(1<<stage);
-        glActiveTexture(GL_TEXTURE0+stage);
+    if (nTex2DEnabled & (1 << stage)) {
+        nTex2DEnabled &= ~(1 << stage);
+        glActiveTexture(GL_TEXTURE0 + stage);
         glDisable(GL_TEXTURE_2D);
         bset = true;
     }
-    if( nTexRECTEnabled & (1<<stage) ) {
-        nTexRECTEnabled &= ~(1<<stage);
-        if( !bset ) glActiveTexture(GL_TEXTURE0+stage);
+    if (nTexRECTEnabled & (1<<stage)) {
+        nTexRECTEnabled &= ~(1 << stage);
+        if (!bset)
+			glActiveTexture(GL_TEXTURE0 + stage);
         glDisable(GL_TEXTURE_RECTANGLE_ARB);
     }
 }

Source/Plugins/Plugin_VideoOGL/Src/VertexLoader.cpp

@@ -59,9 +59,9 @@ void LOADERDECL PosMtx_ReadDirect_UByte(const void *_p)
 void LOADERDECL PosMtx_Write(const void *_p)
 {
 	*VertexManager::s_pCurBufferPointer++ = s_curposmtx;
-	//*VertexManager::s_pCurBufferPointer++ = 0;
-	//*VertexManager::s_pCurBufferPointer++ = 0;
-	//*VertexManager::s_pCurBufferPointer++ = 0;
+	*VertexManager::s_pCurBufferPointer++ = 0;
+	*VertexManager::s_pCurBufferPointer++ = 0;
+	*VertexManager::s_pCurBufferPointer++ = 0;
 }
 
 void LOADERDECL TexMtx_ReadDirect_UByte(const void *_p)
@@ -89,7 +89,7 @@ void LOADERDECL TexMtx_Write_Short3(const void *_p)
 	((s16*)VertexManager::s_pCurBufferPointer)[0] = 0;
 	((s16*)VertexManager::s_pCurBufferPointer)[1] = 0;
 	((s16*)VertexManager::s_pCurBufferPointer)[2] = s_curtexmtx[s_texmtxwrite++];
-	VertexManager::s_pCurBufferPointer += 6;
+	VertexManager::s_pCurBufferPointer += 8;
 }
 
 #include "VertexLoader_Position.h"
@@ -101,6 +101,7 @@ VertexLoader::VertexLoader(const TVtxDesc &vtx_desc, const VAT &vtx_attr)
 {
 	m_VertexSize = 0;
 	m_numPipelineStages = 0;
+	m_NativeFmt = new NativeVertexFormat();
 	VertexLoader_Normal::Init();
 
 	m_VtxDesc = vtx_desc;
@@ -112,6 +113,7 @@ VertexLoader::VertexLoader(const TVtxDesc &vtx_desc, const VAT &vtx_attr)
 
 VertexLoader::~VertexLoader() 
 {
+	delete m_NativeFmt;
 }
 
 int VertexLoader::ComputeVertexSize()
@@ -227,30 +229,29 @@ void VertexLoader::CompileVertexTranslator()
 	m_numPipelineStages = 0;
 
 	// It's a bit ugly that we poke inside m_NativeFmt in this function. Planning to fix this.
-	m_NativeFmt.m_VBStridePad = 0;
-	m_NativeFmt.m_VBVertexStride = 0;
-	m_NativeFmt.m_components = 0;
+	m_NativeFmt->m_VBVertexStride = 0;
+	m_NativeFmt->m_components = 0;
 
 	// m_VBVertexStride for texmtx and posmtx is computed later when writing.
 	
 	// Position Matrix Index
 	if (m_VtxDesc.PosMatIdx) {
 		m_PipelineStages[m_numPipelineStages++] = PosMtx_ReadDirect_UByte;
-		m_NativeFmt.m_components |= VB_HAS_POSMTXIDX;
+		m_NativeFmt->m_components |= VB_HAS_POSMTXIDX;
 	}
 
-	if (m_VtxDesc.Tex0MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX0; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex1MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX1; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex2MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX2; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex3MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX3; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex4MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX4; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex5MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX5; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex6MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX6; WriteCall(TexMtx_ReadDirect_UByte); }
-	if (m_VtxDesc.Tex7MatIdx) {m_NativeFmt.m_components |= VB_HAS_TEXMTXIDX7; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex0MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX0; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex1MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX1; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex2MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX2; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex3MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX3; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex4MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX4; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex5MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX5; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex6MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX6; WriteCall(TexMtx_ReadDirect_UByte); }
+	if (m_VtxDesc.Tex7MatIdx) {m_NativeFmt->m_components |= VB_HAS_TEXMTXIDX7; WriteCall(TexMtx_ReadDirect_UByte); }
 
 	// Position
 	if (m_VtxDesc.Position != NOT_PRESENT)
-		m_NativeFmt.m_VBVertexStride += 12;
+		m_NativeFmt->m_VBVertexStride += 12;
 
 	switch (m_VtxDesc.Position) {
 	case NOT_PRESENT:	{_assert_msg_(0, "Vertex descriptor without position!", "WTF?");} break;
@@ -303,19 +304,19 @@ void VertexLoader::CompileVertexTranslator()
 		int sizePro = 0;
 		switch (m_VtxAttr.NormalFormat)
 		{
-		case FORMAT_UBYTE:	sizePro=1; break;
-		case FORMAT_BYTE:	sizePro=1; break;
-		case FORMAT_USHORT:	sizePro=2; break;
-		case FORMAT_SHORT:	sizePro=2; break;
-		case FORMAT_FLOAT:	sizePro=4; break;
+		case FORMAT_UBYTE:	sizePro = 1*4; break;
+		case FORMAT_BYTE:	sizePro = 1*4; break;
+		case FORMAT_USHORT:	sizePro = 2*4; break;
+		case FORMAT_SHORT:	sizePro = 2*4; break;
+		case FORMAT_FLOAT:	sizePro = 4*3; break;
 		default: _assert_(0); break;
 		}
-		m_NativeFmt.m_VBVertexStride += sizePro * 3 * (m_VtxAttr.NormalElements?3:1);
+		m_NativeFmt->m_VBVertexStride += sizePro * (m_VtxAttr.NormalElements?3:1);
 
 		int numNormals = (m_VtxAttr.NormalElements == 1) ? NRM_THREE : NRM_ONE;
-		m_NativeFmt.m_components |= VB_HAS_NRM0;
+		m_NativeFmt->m_components |= VB_HAS_NRM0;
 		if (numNormals == NRM_THREE)
-			m_NativeFmt.m_components |= VB_HAS_NRM1 | VB_HAS_NRM2;
+			m_NativeFmt->m_components |= VB_HAS_NRM1 | VB_HAS_NRM2;
 	}
 	
 	// Colors
@@ -324,7 +325,7 @@ void VertexLoader::CompileVertexTranslator()
 		SetupColor(i, col[i], m_VtxAttr.color[i].Comp, m_VtxAttr.color[i].Elements);
 
 		if (col[i] != NOT_PRESENT)
-			m_NativeFmt.m_VBVertexStride += 4;
+			m_NativeFmt->m_VBVertexStride += 4;
 	}
 
 	// TextureCoord
@@ -336,21 +337,21 @@ void VertexLoader::CompileVertexTranslator()
 	// Texture matrix indices (remove if corresponding texture coordinate isn't enabled)
 	for (int i = 0; i < 8; i++) {
 		SetupTexCoord(i, tc[i], m_VtxAttr.texCoord[i].Format, m_VtxAttr.texCoord[i].Elements, m_VtxAttr.texCoord[i].Frac);
-		if (m_NativeFmt.m_components & (VB_HAS_TEXMTXIDX0 << i)) {
+		if (m_NativeFmt->m_components & (VB_HAS_TEXMTXIDX0 << i)) {
 			if (tc[i] != NOT_PRESENT) {
 				// if texmtx is included, texcoord will always be 3 floats, z will be the texmtx index
 				WriteCall(m_VtxAttr.texCoord[i].Elements ? TexMtx_Write_Float : TexMtx_Write_Float2);
-				m_NativeFmt.m_VBVertexStride += 12;
+				m_NativeFmt->m_VBVertexStride += 12;
 			}
 			else {
 				WriteCall(TexMtx_Write_Short3);
-				m_NativeFmt.m_VBVertexStride += 6; // still include the texture coordinate, but this time as 6 bytes
-				m_NativeFmt.m_components |= VB_HAS_UV0 << i; // have to include since using now
+				m_NativeFmt->m_VBVertexStride += 8; // still include the texture coordinate, but this time as 6 bytes
+				m_NativeFmt->m_components |= VB_HAS_UV0 << i; // have to include since using now
 			}
 		}
 		else {
 			if (tc[i] != NOT_PRESENT)
-				m_NativeFmt.m_VBVertexStride += 4 * (m_VtxAttr.texCoord[i].Elements ? 2 : 1);
+				m_NativeFmt->m_VBVertexStride += 4 * (m_VtxAttr.texCoord[i].Elements ? 2 : 1);
 		}
 
 		if (tc[i] == NOT_PRESENT) {
@@ -362,30 +363,30 @@ void VertexLoader::CompileVertexTranslator()
 					break;
 				}
 			}
-			if (j == 8 && !((m_NativeFmt.m_components&VB_HAS_TEXMTXIDXALL) & (VB_HAS_TEXMTXIDXALL<<(i+1)))) // no more tex coords and tex matrices, so exit loop
+			if (j == 8 && !((m_NativeFmt->m_components&VB_HAS_TEXMTXIDXALL) & (VB_HAS_TEXMTXIDXALL<<(i+1)))) // no more tex coords and tex matrices, so exit loop
 				break;
 		}
 	}
 
 	if (m_VtxDesc.PosMatIdx) {
 		WriteCall(PosMtx_Write);
-		m_NativeFmt.m_VBVertexStride += 1;
+		m_NativeFmt->m_VBVertexStride += 4;
 	}
 
-	m_NativeFmt.Initialize(m_VtxDesc, m_VtxAttr);
+	m_NativeFmt->Initialize(m_VtxDesc, m_VtxAttr);
 }
 
 void VertexLoader::SetupColor(int num, int mode, int format, int elements)
 {
 	// if COL0 not present, then embed COL1 into COL0
-	if (num == 1 && !(m_NativeFmt.m_components & VB_HAS_COL0))
+	if (num == 1 && !(m_NativeFmt->m_components & VB_HAS_COL0))
 		num = 0;
 
-	m_NativeFmt.m_components |= VB_HAS_COL0 << num;
+	m_NativeFmt->m_components |= VB_HAS_COL0 << num;
 	switch (mode)
 	{
 	case NOT_PRESENT: 
-		m_NativeFmt.m_components &= ~(VB_HAS_COL0 << num);
+		m_NativeFmt->m_components &= ~(VB_HAS_COL0 << num);
 		break;
 	case DIRECT:
 		switch (format)
@@ -428,12 +429,12 @@ void VertexLoader::SetupColor(int num, int mode, int format, int elements)
 
 void VertexLoader::SetupTexCoord(int num, int mode, int format, int elements, int _iFrac)
 {
-	m_NativeFmt.m_components |= VB_HAS_UV0 << num;
+	m_NativeFmt->m_components |= VB_HAS_UV0 << num;
 	
 	switch (mode)
 	{
 	case NOT_PRESENT: 
-		m_NativeFmt.m_components &= ~(VB_HAS_UV0 << num);
+		m_NativeFmt->m_components &= ~(VB_HAS_UV0 << num);
 		break;
 	case DIRECT:
 		switch (format)
@@ -481,12 +482,12 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 	DVSTARTPROFILE();
 
 	// Flush if our vertex format is different from the currently set.
-	if (g_nativeVertexFmt != NULL && g_nativeVertexFmt != &m_NativeFmt)
+	if (g_nativeVertexFmt != NULL && g_nativeVertexFmt != m_NativeFmt)
 	{
 		VertexManager::Flush();
 		// Also move the Set() here?
 	}
-	g_nativeVertexFmt = &m_NativeFmt;
+	g_nativeVertexFmt = m_NativeFmt;
 
 	if (bpmem.genMode.cullmode == 3 && primitive < 5)
 	{
@@ -495,7 +496,7 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 		return;
 	}
 
-	VertexManager::EnableComponents(m_NativeFmt.m_components);
+	VertexManager::EnableComponents(m_NativeFmt->m_components);
 
 	// Load position and texcoord scale factors.
 	// TODO - figure out if we should leave these independent, or compile them into
@@ -511,7 +512,7 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 	m_VtxAttr.texCoord[7].Frac		= g_VtxAttr[vtx_attr_group].g2.Tex7Frac;
 
 	posScale = shiftLookup[m_VtxAttr.PosFrac];
-	if (m_NativeFmt.m_components & VB_HAS_UVALL) {
+	if (m_NativeFmt->m_components & VB_HAS_UVALL) {
 		for (int i = 0; i < 8; i++) {
 			tcScaleU[i] = shiftLookup[m_VtxAttr.texCoord[i].Frac];
 			tcScaleV[i] = shiftLookup[m_VtxAttr.texCoord[i].Frac];
@@ -525,11 +526,11 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 	switch (primitive) {
 		case 3: // strip
 		case 4: // fan
-			if (VertexManager::GetRemainingSize() < 3 * m_NativeFmt.m_VBVertexStride)
+			if (VertexManager::GetRemainingSize() < 3 * m_NativeFmt->m_VBVertexStride)
 				VertexManager::Flush();
 			break;
 		case 6: // line strip
-			if (VertexManager::GetRemainingSize() < 2 * m_NativeFmt.m_VBVertexStride)
+			if (VertexManager::GetRemainingSize() < 2 * m_NativeFmt->m_VBVertexStride)
 				VertexManager::Flush();
 			break;
 		case 0: // quads
@@ -544,11 +545,12 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 	}
 
 	int startv = 0, extraverts = 0;
+
 	for (int v = 0; v < count; v++)
 	{
 		if ((v % granularity) == 0)
 		{
-			if (VertexManager::GetRemainingSize() < granularity*m_NativeFmt.m_VBVertexStride) {
+			if (VertexManager::GetRemainingSize() < granularity*m_NativeFmt->m_VBVertexStride) {
 				// This buffer full - break current primitive and flush, to switch to the next buffer.
 				u8* plastptr = VertexManager::s_pCurBufferPointer;
 				if (v - startv > 0)
@@ -559,27 +561,27 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 					case 3: // triangle strip, copy last two vertices
 						// a little trick since we have to keep track of signs
 						if (v & 1) {
-							memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-2*m_NativeFmt.m_VBVertexStride, m_NativeFmt.m_VBVertexStride);
-							memcpy_gc(VertexManager::s_pCurBufferPointer+m_NativeFmt.m_VBVertexStride, plastptr-m_NativeFmt.m_VBVertexStride*2, 2*m_NativeFmt.m_VBVertexStride);
-							VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBVertexStride*3;
+							memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-2*m_NativeFmt->m_VBVertexStride, m_NativeFmt->m_VBVertexStride);
+							memcpy_gc(VertexManager::s_pCurBufferPointer+m_NativeFmt->m_VBVertexStride, plastptr-m_NativeFmt->m_VBVertexStride*2, 2*m_NativeFmt->m_VBVertexStride);
+							VertexManager::s_pCurBufferPointer += m_NativeFmt->m_VBVertexStride*3;
 							extraverts = 3;
 						}
 						else {
-							memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt.m_VBVertexStride*2, m_NativeFmt.m_VBVertexStride*2);
-							VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBVertexStride*2;
+							memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt->m_VBVertexStride*2, m_NativeFmt->m_VBVertexStride*2);
+							VertexManager::s_pCurBufferPointer += m_NativeFmt->m_VBVertexStride*2;
 							extraverts = 2;
 						}
 						break;
 					case 4: // tri fan, copy first and last vert
-						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt.m_VBVertexStride*(v-startv+extraverts), m_NativeFmt.m_VBVertexStride);
-						VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBVertexStride;
-						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt.m_VBVertexStride, m_NativeFmt.m_VBVertexStride);
-						VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBVertexStride;
+						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt->m_VBVertexStride*(v-startv+extraverts), m_NativeFmt->m_VBVertexStride);
+						VertexManager::s_pCurBufferPointer += m_NativeFmt->m_VBVertexStride;
+						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt->m_VBVertexStride, m_NativeFmt->m_VBVertexStride);
+						VertexManager::s_pCurBufferPointer += m_NativeFmt->m_VBVertexStride;
 						extraverts = 2;
 						break;
 					case 6: // line strip
-						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt.m_VBVertexStride, m_NativeFmt.m_VBVertexStride);
-						VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBVertexStride;
+						memcpy_gc(VertexManager::s_pCurBufferPointer, plastptr-m_NativeFmt->m_VBVertexStride, m_NativeFmt->m_VBVertexStride);
+						VertexManager::s_pCurBufferPointer += m_NativeFmt->m_VBVertexStride;
 						extraverts = 1;
 						break;
 					default:
@@ -596,7 +598,6 @@ void VertexLoader::RunVertices(int vtx_attr_group, int primitive, int count)
 		for (int i = 0; i < m_numPipelineStages; i++)
 			m_PipelineStages[i](&m_VtxAttr);
 
-		VertexManager::s_pCurBufferPointer += m_NativeFmt.m_VBStridePad;
 		PRIM_LOG("\n");
 	}
 

Source/Plugins/Plugin_VideoOGL/Src/VertexLoader.h

@@ -74,7 +74,7 @@ private:
 	TVtxDesc m_VtxDesc;  // Not really used currently - or well it is, but could be easily avoided.
 
 	// PC vertex format
-	NativeVertexFormat m_NativeFmt;
+	NativeVertexFormat *m_NativeFmt;
 
 	// Pipeline. To be JIT compiled in the future.
 	TPipelineFunction m_PipelineStages[32];  // TODO - figure out real max. it's lower.

Source/Plugins/Plugin_VideoOGL/Src/VertexLoaderManager.cpp

@@ -37,6 +37,8 @@ static VertexLoaderMap g_VertexLoaderMap;
 void Init()
 {
 	MarkAllDirty();
+	for (int i = 0; i < 8; i++)
+		g_VertexLoaders[i] = NULL;
 }
 
 void Shutdown()
@@ -53,9 +55,9 @@ void MarkAllDirty()
 	s_attr_dirty = 0xff;
 }
 
-void RefreshLoader(int vtx_attr_group)
+static void RefreshLoader(int vtx_attr_group)
 {
-	if (((s_attr_dirty >> vtx_attr_group) & 1) || !g_VertexLoaders[vtx_attr_group])
+	if ((s_attr_dirty >> vtx_attr_group) & 1)
 	{
 		VertexLoaderUID uid;
 		uid.InitFromCurrentState(vtx_attr_group);

Source/Plugins/Plugin_VideoOGL/Src/VertexLoader_Normal.cpp

@@ -123,7 +123,8 @@ void LOADERDECL VertexLoader_Normal::Normal_DirectByte(const void *_p)
     *VertexManager::s_pCurBufferPointer++ = DataReadU8();
     *VertexManager::s_pCurBufferPointer++ = DataReadU8();
     *VertexManager::s_pCurBufferPointer++ = DataReadU8();
-    LOG_NORM8();
+    VertexManager::s_pCurBufferPointer++;
+	LOG_NORM8();
 //    ((float*)VertexManager::s_pCurBufferPointer)[0] = ((float)(signed char)DataReadU8()+0.5f) / 127.5f;
 }
 
@@ -132,7 +133,7 @@ void LOADERDECL VertexLoader_Normal::Normal_DirectShort(const void *_p)
     ((u16*)VertexManager::s_pCurBufferPointer)[0] = DataReadU16();
     ((u16*)VertexManager::s_pCurBufferPointer)[1] = DataReadU16();
     ((u16*)VertexManager::s_pCurBufferPointer)[2] = DataReadU16();
-    VertexManager::s_pCurBufferPointer += 6;
+    VertexManager::s_pCurBufferPointer += 8;
     LOG_NORM16()
 //    ((float*)VertexManager::s_pCurBufferPointer)[0] = ((float)(signed short)DataReadU16()+0.5f) / 32767.5f;
 //    ((float*)VertexManager::s_pCurBufferPointer)[1] = ((float)(signed short)DataReadU16()+0.5f) / 32767.5f;
@@ -150,30 +151,31 @@ void LOADERDECL VertexLoader_Normal::Normal_DirectFloat(const void *_p)
 
 void LOADERDECL VertexLoader_Normal::Normal_DirectByte3(const void *_p)
 {
-    for (int i=0; i<3; i++)
+    for (int i = 0; i < 3; i++)
     {
         *VertexManager::s_pCurBufferPointer++ = DataReadU8();
         *VertexManager::s_pCurBufferPointer++ = DataReadU8();
         *VertexManager::s_pCurBufferPointer++ = DataReadU8();
+        VertexManager::s_pCurBufferPointer++;
         LOG_NORM8();
     }
 }
 
 void LOADERDECL VertexLoader_Normal::Normal_DirectShort3(const void *_p)
 {
-    for (int i=0; i<3; i++)
+    for (int i = 0; i < 3; i++)
     {
         ((u16*)VertexManager::s_pCurBufferPointer)[0] = DataReadU16();
         ((u16*)VertexManager::s_pCurBufferPointer)[1] = DataReadU16();
         ((u16*)VertexManager::s_pCurBufferPointer)[2] = DataReadU16();
-        VertexManager::s_pCurBufferPointer += 6;
+        VertexManager::s_pCurBufferPointer += 8;
         LOG_NORM16();
     }
 }
 
 void LOADERDECL VertexLoader_Normal::Normal_DirectFloat3(const void *_p)
 {
-    for (int i=0; i<3; i++)
+    for (int i = 0; i < 3; i++)
     {
         ((float*)VertexManager::s_pCurBufferPointer)[0] = DataReadF32();
         ((float*)VertexManager::s_pCurBufferPointer)[1] = DataReadF32();
@@ -193,6 +195,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Byte(const void *_p)
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+	VertexManager::s_pCurBufferPointer++;
 //    ((float*)VertexManager::s_pCurBufferPointer)[0] = ((float)(signed char)Memory_Read_U8(iAddress)+0.5f) / 127.5f;
 //    ((float*)VertexManager::s_pCurBufferPointer)[1] = ((float)(signed char)Memory_Read_U8(iAddress+1)+0.5f) / 127.5f;
 //    ((float*)VertexManager::s_pCurBufferPointer)[2] = ((float)(signed char)Memory_Read_U8(iAddress+2)+0.5f) / 127.5f;
@@ -207,7 +210,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Short(const void *_p)
     ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
     ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
     ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-    VertexManager::s_pCurBufferPointer += 6;
+    VertexManager::s_pCurBufferPointer += 8;
     LOG_NORM16();
 }
 
@@ -225,22 +228,24 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Float(const void *_p)
 void LOADERDECL VertexLoader_Normal::Normal_Index8_Byte3(const void *_p)
 {
     if (index3) {
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u8 Index = DataReadU8();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 1*3*i;
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+            *VertexManager::s_pCurBufferPointer++;
             LOG_NORM8();
         }
     }
     else {
         u8 Index = DataReadU8();
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 1*3*i;
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+			VertexManager::s_pCurBufferPointer++;
             LOG_NORM8();
         }
     }
@@ -249,24 +254,24 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Byte3(const void *_p)
 void LOADERDECL VertexLoader_Normal::Normal_Index8_Short3(const void *_p)
 {
     if (index3) {
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u8 Index = DataReadU8();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 2*3*i;
             ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
             ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
             ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-            VertexManager::s_pCurBufferPointer += 6;
+            VertexManager::s_pCurBufferPointer += 8;
             LOG_NORM16();
         }
     }
     else {
         u8 Index = DataReadU8();
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 2*3*i;
             ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
             ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
             ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-            VertexManager::s_pCurBufferPointer += 6;
+            VertexManager::s_pCurBufferPointer += 8;
             LOG_NORM16();
         }
     }
@@ -275,7 +280,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Short3(const void *_p)
 void LOADERDECL VertexLoader_Normal::Normal_Index8_Float3(const void *_p)
 {
     if (index3) {
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u8 Index = DataReadU8();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 4*3*i;
             ((float*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_Float(iAddress);
@@ -287,7 +292,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index8_Float3(const void *_p)
     }
     else {
         u8 Index = DataReadU8();
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 4*3*i;
             ((float*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_Float(iAddress);
             ((float*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_Float(iAddress+4);
@@ -309,6 +314,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Byte(const void *_p)
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
     *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+	VertexManager::s_pCurBufferPointer++;
     LOG_NORM8();
 }
 
@@ -319,7 +325,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Short(const void *_p)
     ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
     ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
     ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-    VertexManager::s_pCurBufferPointer += 6;
+    VertexManager::s_pCurBufferPointer += 8;
     LOG_NORM16();
 }
 
@@ -337,22 +343,24 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Float(const void *_p)
 void LOADERDECL VertexLoader_Normal::Normal_Index16_Byte3(const void *_p)
 {
     if (index3) {
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u16 Index = DataReadU16();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 1*3*i;
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+			VertexManager::s_pCurBufferPointer++;
             LOG_NORM8();
         }
     }
     else {
         u16 Index = DataReadU16();
-        for (int i=0; i<3; i++) {
+        for (int i = 0; i < 3; i++) {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 1*3*i;
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+1);
             *VertexManager::s_pCurBufferPointer++ = Memory_Read_U8(iAddress+2);
+			VertexManager::s_pCurBufferPointer++;
             LOG_NORM8();
         }
     }
@@ -362,27 +370,27 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Short3(const void *_p)
 {
     if (index3)
     {
-        for (int i=0; i<3; i++)
+        for (int i = 0; i < 3; i++)
         {
             u16 Index = DataReadU16();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 2*3*i;
             ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
             ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
             ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-            VertexManager::s_pCurBufferPointer += 6;
+            VertexManager::s_pCurBufferPointer += 8;
             LOG_NORM16();
         }
     }
     else
     {
         u16 Index = DataReadU16();
-        for (int i=0; i<3; i++)
+        for (int i = 0; i < 3; i++)
         {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 2*3*i;
             ((u16*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_U16(iAddress);
             ((u16*)VertexManager::s_pCurBufferPointer)[1] = Memory_Read_U16(iAddress+2);
             ((u16*)VertexManager::s_pCurBufferPointer)[2] = Memory_Read_U16(iAddress+4);
-            VertexManager::s_pCurBufferPointer += 6;
+            VertexManager::s_pCurBufferPointer += 8;
             LOG_NORM16();
         }
     }
@@ -392,7 +400,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Float3(const void *_p)
 {
     if (index3)
     {
-        for (int i=0; i<3; i++)
+        for (int i = 0; i < 3; i++)
         {
             u16 Index = DataReadU16();
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 4*3*i;
@@ -406,7 +414,7 @@ void LOADERDECL VertexLoader_Normal::Normal_Index16_Float3(const void *_p)
     else
     {
         u16 Index = DataReadU16();
-        for (int i=0; i<3; i++)
+        for (int i = 0; i < 3; i++)
         {
             u32 iAddress = arraybases[ARRAY_NORMAL] + (Index * arraystrides[ARRAY_NORMAL]) + 4*3*i;
             ((float*)VertexManager::s_pCurBufferPointer)[0] = Memory_Read_Float(iAddress);

Source/Plugins/Plugin_VideoOGL/Src/VertexManager.cpp

@@ -48,8 +48,6 @@ static const GLenum c_primitiveType[8] =
 
 bool Init()
 {
-	Destroy();
-
 	s_prevcomponents = 0;
 	s_pBaseBufferPointer = (u8*)AllocateMemoryPages(MAX_BUFFER_SIZE);
 	s_pCurBufferPointer = s_pBaseBufferPointer;
@@ -69,7 +67,7 @@ bool Init()
 	return true;
 }
 
-void Destroy()
+void Shutdown()
 {
 	FreeMemoryPages(s_pBaseBufferPointer, MAX_BUFFER_SIZE); s_pBaseBufferPointer = s_pCurBufferPointer = NULL;
 	glDeleteBuffers(ARRAYSIZE(s_vboBuffers), s_vboBuffers);

Source/Plugins/Plugin_VideoOGL/Src/VertexManager.h

@@ -25,21 +25,23 @@
 // Other functionality is moving out.
 namespace VertexManager
 {
-    bool Init();
-    void Destroy();
 
-    void ResetBuffer();
-    void ResetComponents();
+bool Init();
+void Shutdown();
 
-    void AddVertices(int primitive, int numvertices);
-    void Flush(); // flushes the current buffer
+void ResetBuffer();
+void ResetComponents();
 
-    int GetRemainingSize();  // remaining space in the current buffer.
+void AddVertices(int primitive, int numvertices);
+void Flush(); // flushes the current buffer
 
-	void EnableComponents(u32 components);  // very implementation specific - D3D9 won't need this one.
+int GetRemainingSize();  // remaining space in the current buffer.
+
+void EnableComponents(u32 components);  // very implementation specific - D3D9 won't need this one.
+
+// TODO: move, rename.
+extern u8* s_pCurBufferPointer;
 
-	// TODO: move, rename.
-    extern u8* s_pCurBufferPointer;
 };
 
 #endif  // _VERTEXMANAGER_H

Source/Plugins/Plugin_VideoOGL/Src/VertexShaderManager.cpp

@@ -49,8 +49,6 @@ extern int nBackbufferWidth, nBackbufferHeight;
 static int s_nMaxVertexInstructions;
 
 static float s_fMaterials[16];
-static float rawViewport[6] = {0};
-static float rawProjection[7] = {0};
 
 // track changes
 static bool bTexMatricesChanged[2], bPosNormalMatrixChanged, bProjectionChanged, bViewportChanged;
@@ -92,7 +90,7 @@ void VertexShaderMngr::Shutdown()
 }
 
 float VertexShaderMngr::GetPixelAspectRatio() {
-	return rawViewport[0] != 0 ? (float)Renderer::GetTargetWidth() / 640.0f : 1.0f;
+	return xfregs.rawViewport[0] != 0 ? (float)Renderer::GetTargetWidth() / 640.0f : 1.0f;
 }
 
 VERTEXSHADER* VertexShaderMngr::GetShader(u32 components)
@@ -378,8 +376,8 @@ void VertexShaderMngr::SetConstants()
 				wAdj = ratio;
 				hAdj = 1;
 
-				wid = ceil(fabs(2 * rawViewport[0]) / wAdj);
-				hei = ceil(fabs(2 * rawViewport[1]) / hAdj);
+				wid = ceil(fabs(2 * xfregs.rawViewport[0]) / wAdj);
+				hei = ceil(fabs(2 * xfregs.rawViewport[1]) / hAdj);
 
 				actualWid = ceil((float)winw / ratio);
 				actualRatiow = (float)actualWid / (float)wid; // the picture versus the screen
@@ -393,8 +391,8 @@ void VertexShaderMngr::SetConstants()
 				wAdj = 1;
 				hAdj = ratio;
 
-				wid = ceil(fabs(2 * rawViewport[0]) / wAdj);
-				hei = ceil(fabs(2 * rawViewport[1]) / hAdj);
+				wid = ceil(fabs(2 * xfregs.rawViewport[0]) / wAdj);
+				hei = ceil(fabs(2 * xfregs.rawViewport[1]) / hAdj);
 
 				actualHei = ceil((float)winh / ratio);
 				actualRatioh = (float)actualHei / (float)hei; // the picture versus the screen
@@ -404,52 +402,56 @@ void VertexShaderMngr::SetConstants()
 		}
 		else
 		{
-			wid = ceil(fabs(2 * rawViewport[0]));
-			hei = ceil(fabs(2 * rawViewport[1]));
+			wid = ceil(fabs(2 * xfregs.rawViewport[0]));
+			hei = ceil(fabs(2 * xfregs.rawViewport[1]));
 		}
 
 		if (g_Config.bStretchToFit)
 		{
 			glViewport(
-				(int)(rawViewport[3]-rawViewport[0]-342-scissorXOff) + xoffs,
-				Renderer::GetTargetHeight() - ((int)(rawViewport[4]-rawViewport[1]-342-scissorYOff)) + yoffs,
+				(int)(xfregs.rawViewport[3]-xfregs.rawViewport[0]-342-scissorXOff) + xoffs,
+				Renderer::GetTargetHeight() - ((int)(xfregs.rawViewport[4]-xfregs.rawViewport[1]-342-scissorYOff)) + yoffs,
 				wid, // width
 				hei // height
 				);
 		}
 		else
 		{
-			glViewport((int)(rawViewport[3]-rawViewport[0]-342-scissorXOff) * MValueX,
-				Renderer::GetTargetHeight()-((int)(rawViewport[4]-rawViewport[1]-342-scissorYOff))  * MValueY,
-				abs((int)(2 * rawViewport[0])) * MValueX, abs((int)(2 * rawViewport[1])) * MValueY);
+			glViewport((int)(xfregs.rawViewport[3]-xfregs.rawViewport[0]-342-scissorXOff) * MValueX,
+				Renderer::GetTargetHeight()-((int)(xfregs.rawViewport[4]-xfregs.rawViewport[1]-342-scissorYOff))  * MValueY,
+				abs((int)(2 * xfregs.rawViewport[0])) * MValueX, abs((int)(2 * xfregs.rawViewport[1])) * MValueY);
 		}
 
+		// Standard depth range
+		//glDepthRange(-(0.0f - (xfregs.rawViewport[5]-xfregs.rawViewport[2])/-16777215.0f), xfregs.rawViewport[5]/16777215.0f);
+
 		// Metroid Prime 1 & 2 likes this
-		glDepthRange((g_Config.bInvertDepth ? -1 : 1) * -(0.0f - (rawViewport[5]-rawViewport[2])/16777215.0f), rawViewport[5]/16777215.0f);
+		//glDepthRange(-(0.0f - (xfregs.rawViewport[5]-xfregs.rawViewport[2])/16777215.0f), xfregs.rawViewport[5]/16777215.0f);
+		glDepthRange((g_Config.bInvertDepth ? -1 : 1) * -(0.0f - (xfregs.rawViewport[5]-xfregs.rawViewport[2])/16777215.0f), xfregs.rawViewport[5]/16777215.0f);
 
 		// FZero stage likes this (a sonic hack)
-		// glDepthRange(-(0.0f - (rawViewport[5]-rawViewport[2])/-16777215.0f), rawViewport[5]/16777215.0f);
+		// glDepthRange(-(0.0f - (xfregs.rawViewport[5]-xfregs.rawViewport[2])/-16777215.0f), xfregs.rawViewport[5]/16777215.0f);
     }
 
     if (bProjectionChanged) {
         bProjectionChanged = false;
 
-        if (rawProjection[6] == 0) {
-            g_fProjectionMatrix[0] = rawProjection[0];
+        if (xfregs.rawProjection[6] == 0) {
+            g_fProjectionMatrix[0] = xfregs.rawProjection[0];
             g_fProjectionMatrix[1] = 0.0f;
-            g_fProjectionMatrix[2] = rawProjection[1];
+            g_fProjectionMatrix[2] = xfregs.rawProjection[1];
             g_fProjectionMatrix[3] = 0;
                         
             g_fProjectionMatrix[4] = 0.0f;
-            g_fProjectionMatrix[5] = rawProjection[2];
-            g_fProjectionMatrix[6] = rawProjection[3];
+            g_fProjectionMatrix[5] = xfregs.rawProjection[2];
+            g_fProjectionMatrix[6] = xfregs.rawProjection[3];
             g_fProjectionMatrix[7] = 0;
                         
             g_fProjectionMatrix[8] = 0.0f;
             g_fProjectionMatrix[9] = 0.0f;
-            g_fProjectionMatrix[10] = rawProjection[4];
+            g_fProjectionMatrix[10] = xfregs.rawProjection[4];
 			// Working bloom in ZTP
-            g_fProjectionMatrix[11] = -(0.0f - rawProjection[5]);  // Yes, it's important that it's done this way.
+            g_fProjectionMatrix[11] = -(0.0f - xfregs.rawProjection[5]);  // Yes, it's important that it's done this way.
 			// Working projection in PSO
 			// g_fProjectionMatrix[11] = -(1.0f - rawProjection[5]);
                         
@@ -459,23 +461,23 @@ void VertexShaderMngr::SetConstants()
             g_fProjectionMatrix[15] = 0.0f;
         }
         else {
-            g_fProjectionMatrix[0] = rawProjection[0];
+            g_fProjectionMatrix[0] = xfregs.rawProjection[0];
             g_fProjectionMatrix[1] = 0.0f;
             g_fProjectionMatrix[2] = 0.0f;
-            g_fProjectionMatrix[3] = rawProjection[1];
+            g_fProjectionMatrix[3] = xfregs.rawProjection[1];
 
             g_fProjectionMatrix[4] = 0.0f;
-            g_fProjectionMatrix[5] = rawProjection[2];
+            g_fProjectionMatrix[5] = xfregs.rawProjection[2];
             g_fProjectionMatrix[6] = 0.0f;
-            g_fProjectionMatrix[7] = rawProjection[3];
+            g_fProjectionMatrix[7] = xfregs.rawProjection[3];
 
             g_fProjectionMatrix[8] = 0.0f;
             g_fProjectionMatrix[9] = 0.0f;
-            g_fProjectionMatrix[10] = rawProjection[4];
+            g_fProjectionMatrix[10] = xfregs.rawProjection[4];
 			// Working bloom in ZTP
-            g_fProjectionMatrix[11] = -(-1.0f - rawProjection[5]);  // Yes, it's important that it's done this way.
-			// Working projection in PSO
-			// g_fProjectionMatrix[11] = -(0.0f - rawProjection[5]);
+            //g_fProjectionMatrix[11] = -(-1.0f - rawProjection[5]);  // Yes, it's important that it's done this way.
+			// Working projection in PSO, working Super Monkey Ball
+			g_fProjectionMatrix[11] = -(0.0f - xfregs.rawProjection[5]);
 
             g_fProjectionMatrix[12] = 0;
             g_fProjectionMatrix[13] = 0;
@@ -483,7 +485,7 @@ void VertexShaderMngr::SetConstants()
             g_fProjectionMatrix[15] = 1.0f;
         }
 
-        PRIM_LOG("Projection: %f %f %f %f %f %f\n", rawProjection[0], rawProjection[1], rawProjection[2], rawProjection[3], rawProjection[4], rawProjection[5]);
+        PRIM_LOG("Projection: %f %f %f %f %f %f\n", xfregs.rawProjection[0], xfregs.rawProjection[1], xfregs.rawProjection[2], xfregs.rawProjection[3], xfregs.rawProjection[4], xfregs.rawProjection[5]);
         SetVSConstant4fv(C_PROJECTION,   &g_fProjectionMatrix[0]);
         SetVSConstant4fv(C_PROJECTION+1, &g_fProjectionMatrix[4]);
         SetVSConstant4fv(C_PROJECTION+2, &g_fProjectionMatrix[8]);
@@ -591,11 +593,11 @@ void VertexShaderMngr::SetTexMatrixChangedB(u32 Value)
 void VertexShaderMngr::SetViewport(float* _Viewport)
 {
     // Workaround for paper mario, yep this is bizarre.
-    for (size_t i = 0; i < ARRAYSIZE(rawViewport); ++i) {
+    for (size_t i = 0; i < ARRAYSIZE(xfregs.rawViewport); ++i) {
         if (*(u32*)(_Viewport + i) == 0x7f800000)  // invalid fp number
             return;
     }
-    memcpy(rawViewport, _Viewport, sizeof(rawViewport));
+    memcpy(xfregs.rawViewport, _Viewport, sizeof(xfregs.rawViewport));
     bViewportChanged = true;
 }
 
@@ -606,12 +608,65 @@ void VertexShaderMngr::SetViewportChanged()
 
 void VertexShaderMngr::SetProjection(float* _pProjection, int constantIndex)
 {
-    memcpy(rawProjection, _pProjection, sizeof(rawProjection));
+    memcpy(xfregs.rawProjection, _pProjection, sizeof(xfregs.rawProjection));
     bProjectionChanged = true;
 }
 
+float* VertexShaderMngr::GetPosNormalMat()
+{
+    return (float*)xfmem + MatrixIndexA.PosNormalMtxIdx * 4;
+}
+
+// Mash together all the inputs that contribute to the code of a generated vertex shader into
+// a unique identifier, basically containing all the bits. Yup, it's a lot ....
+void VertexShaderMngr::GetVertexShaderId(VERTEXSHADERUID& vid, u32 components)
+{
+	u32 zbufrender = (bpmem.ztex2.op == ZTEXTURE_ADD) || Renderer::GetZBufferTarget() != 0;
+	vid.values[0] = components | 
+		(xfregs.numTexGens << 23) |
+		(xfregs.nNumChans << 27) |
+		((u32)xfregs.bEnableDualTexTransform << 29) |
+		(zbufrender << 30);
+
+	for (int i = 0; i < 2; ++i) {
+		vid.values[1+i] = xfregs.colChans[i].color.enablelighting ? 
+			(u32)xfregs.colChans[i].color.hex : 
+			(u32)xfregs.colChans[i].color.matsource;
+		vid.values[1+i] |= (xfregs.colChans[i].alpha.enablelighting ? 
+			(u32)xfregs.colChans[i].alpha.hex : 
+			(u32)xfregs.colChans[i].alpha.matsource) << 15;
+	}
+
+	// fog
+	vid.values[1] |= (((u32)bpmem.fog.c_proj_fsel.fsel & 3) << 30);
+	vid.values[2] |= (((u32)bpmem.fog.c_proj_fsel.fsel >> 2) << 30);
+
+	u32* pcurvalue = &vid.values[3];
+	for (int i = 0; i < xfregs.numTexGens; ++i) {
+		TexMtxInfo tinfo = xfregs.texcoords[i].texmtxinfo;
+		if (tinfo.texgentype != XF_TEXGEN_EMBOSS_MAP)
+			tinfo.hex &= 0x7ff;
+		if (tinfo.texgentype != XF_TEXGEN_REGULAR)
+			tinfo.projection = 0;
+
+		u32 val = ((tinfo.hex >> 1) & 0x1ffff);
+		if (xfregs.bEnableDualTexTransform && tinfo.texgentype == XF_TEXGEN_REGULAR) {
+			// rewrite normalization and post index
+			val |= ((u32)xfregs.texcoords[i].postmtxinfo.index << 17) | ((u32)xfregs.texcoords[i].postmtxinfo.normalize << 23);
+		}
+
+		switch (i & 3) {
+			case 0: pcurvalue[0] |= val; break;
+			case 1: pcurvalue[0] |= val << 24; pcurvalue[1] = val >> 8; ++pcurvalue; break;
+			case 2: pcurvalue[0] |= val << 16; pcurvalue[1] = val >> 16; ++pcurvalue; break;
+			case 3: pcurvalue[0] |= val << 8; ++pcurvalue; break;
+		}
+	}
+}
+
+
 // LoadXFReg 0x10
-void VertexShaderMngr::LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
+void LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
 {	
     u32 address = baseAddress;
     for (int i = 0; i < (int)transferSize; i++)
@@ -622,7 +677,7 @@ void VertexShaderMngr::LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
         if (address < 0x1000)
         {
             VertexManager::Flush();
-            InvalidateXFRange(address, address+transferSize);
+			VertexShaderMngr::InvalidateXFRange(address, address + transferSize);
             //PRIM_LOG("xfmem write: 0x%x-0x%x\n", address, address+transferSize);
 
             u32* p1 = &xfmem[address];
@@ -748,16 +803,16 @@ void VertexShaderMngr::LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
                 break;
             case 0x1018:
                 //_assert_msg_(GX_XF, 0, "XF matrixindex0");
-                SetTexMatrixChangedA(data); //?
+				VertexShaderMngr::SetTexMatrixChangedA(data); //?
                 break;
             case 0x1019:
                 //_assert_msg_(GX_XF, 0, "XF matrixindex1");
-                SetTexMatrixChangedB(data); //?
+                VertexShaderMngr::SetTexMatrixChangedB(data); //?
                 break;
 
             case 0x101a: 
                 VertexManager::Flush();
-                SetViewport((float*)&pData[i]);
+                VertexShaderMngr::SetViewport((float*)&pData[i]);
                 i += 6;
                 break;
 
@@ -812,7 +867,7 @@ void VertexShaderMngr::LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
                 break;
             }
         }
-        else if (address>=0x4000)
+        else if (address >= 0x4000)
         {
             // MessageBox(NULL, "1", "1", MB_OK);
             //4010 __GXSetGenMode
@@ -821,7 +876,7 @@ void VertexShaderMngr::LoadXFReg(u32 transferSize, u32 baseAddress, u32 *pData)
 }
 
 // TODO - verify that it is correct. Seems to work, though.
-void VertexShaderMngr::LoadIndexedXF(u32 val, int array)
+void LoadIndexedXF(u32 val, int array)
 {
     int index = val >> 16;
     int address = val & 0xFFF; //check mask
@@ -829,61 +884,9 @@ void VertexShaderMngr::LoadIndexedXF(u32 val, int array)
     //load stuff from array to address in xf mem
 
     VertexManager::Flush();
-    InvalidateXFRange(address, address+size);
+    VertexShaderMngr::InvalidateXFRange(address, address+size);
     //PRIM_LOG("xfmem iwrite: 0x%x-0x%x\n", address, address+size);
 
     for (int i = 0; i < size; i++)
         xfmem[address + i] = Memory_Read_U32(arraybases[array] + arraystrides[array]*index + i*4);
 }
-
-float* VertexShaderMngr::GetPosNormalMat()
-{
-    return (float*)xfmem + MatrixIndexA.PosNormalMtxIdx * 4;
-}
-
-// Mash together all the inputs that contribute to the code of a generated vertex shader into
-// a unique identifier, basically containing all the bits. Yup, it's a lot ....
-void VertexShaderMngr::GetVertexShaderId(VERTEXSHADERUID& vid, u32 components)
-{
-    u32 zbufrender = (bpmem.ztex2.op == ZTEXTURE_ADD) || Renderer::GetZBufferTarget() != 0;
-    vid.values[0] = components | 
-		(xfregs.numTexGens << 23) |
-		(xfregs.nNumChans << 27) |
-	    ((u32)xfregs.bEnableDualTexTransform << 29) |
-		(zbufrender << 30);
-
-    for (int i = 0; i < 2; ++i) {
-        vid.values[1+i] = xfregs.colChans[i].color.enablelighting ? 
-			(u32)xfregs.colChans[i].color.hex : 
-		    (u32)xfregs.colChans[i].color.matsource;
-        vid.values[1+i] |= (xfregs.colChans[i].alpha.enablelighting ? 
-			(u32)xfregs.colChans[i].alpha.hex : 
-		    (u32)xfregs.colChans[i].alpha.matsource) << 15;
-    }
-
-	// fog
-    vid.values[1] |= (((u32)bpmem.fog.c_proj_fsel.fsel & 3) << 30);
-    vid.values[2] |= (((u32)bpmem.fog.c_proj_fsel.fsel >> 2) << 30);
-
-    u32* pcurvalue = &vid.values[3];
-    for (int i = 0; i < xfregs.numTexGens; ++i) {
-        TexMtxInfo tinfo = xfregs.texcoords[i].texmtxinfo;
-        if (tinfo.texgentype != XF_TEXGEN_EMBOSS_MAP)
-            tinfo.hex &= 0x7ff;
-        if (tinfo.texgentype != XF_TEXGEN_REGULAR)
-            tinfo.projection = 0;
-
-        u32 val = ((tinfo.hex >> 1) & 0x1ffff);
-        if (xfregs.bEnableDualTexTransform && tinfo.texgentype == XF_TEXGEN_REGULAR) {
-            // rewrite normalization and post index
-            val |= ((u32)xfregs.texcoords[i].postmtxinfo.index << 17) | ((u32)xfregs.texcoords[i].postmtxinfo.normalize << 23);
-        }
-
-        switch (i & 3) {
-            case 0: pcurvalue[0] |= val; break;
-            case 1: pcurvalue[0] |= val << 24; pcurvalue[1] = val >> 8; ++pcurvalue; break;
-            case 2: pcurvalue[0] |= val << 16; pcurvalue[1] = val >> 16; ++pcurvalue; break;
-            case 3: pcurvalue[0] |= val << 8; ++pcurvalue; break;
-        }
-    }
-}

Source/Plugins/Plugin_VideoOGL/Src/VertexShaderManager.h

@@ -121,12 +121,11 @@ public:
     static void SetTexMatrixChangedA(u32 Value);
     static void SetTexMatrixChangedB(u32 Value);
 
-    static void LoadXFReg(u32 transferSize, u32 address, u32 *pData);
-    static void LoadIndexedXF(u32 val, int array);
-
     static float* GetPosNormalMat();
 	static float GetPixelAspectRatio();
 };
 
+void LoadXFReg(u32 transferSize, u32 address, u32 *pData);
+void LoadIndexedXF(u32 val, int array);
 
 #endif

Source/Plugins/Plugin_VideoOGL/Src/XFB.cpp

@@ -116,27 +116,12 @@ void XFB_Write(u8 *xfb_in_ram, const TRectangle& sourceRc, u32 dstWd, u32 dstHt,
 	Renderer::RestoreGLState();
     GL_REPORT_ERRORD();
 
-	// TODO - verify that XFB pixel order. probably reverse of GL pixel order
 	// TODO - use shader for conversion
 	ConvertToXFB((u32 *)xfb_in_ram, efb_buffer, dstWd, dstHt);
-
-	// old way
-	/*glActiveTexture(GL_TEXTURE0);
-	glBindTexture(GL_TEXTURE_RECTANGLE_ARB, Renderer::GetRenderTarget());
-    TextureMngr::EnableTexRECT(0);
-	
-	glGetTexImage(GL_TEXTURE_RECTANGLE_ARB, 0, GL_RGBA, GL_UNSIGNED_BYTE, efb_buffer);
-	
-	static u8 *efbEndAddress = efb_buffer + ((nBackbufferWidth * nBackbufferHeight) - 1) * 4;	
-	ConvertToXFBReverse((u32 *)xfb_in_ram, efbEndAddress, dstWd, dstHt);*/
-	
 }
 
 void XFB_Draw(u8 *xfb_in_ram, u32 width, u32 height, s32 yOffset)
 {
-	if(width == 0 || height == 0)
-		return;
-
 	OpenGL_Update(); // just updates the render window position and the backbuffer size
 
     Renderer::SetRenderMode(Renderer::RM_Normal);

Source/Plugins/Plugin_VideoOGL/Src/main.cpp

@@ -246,7 +246,7 @@ void Video_Shutdown(void)
     VertexShaderMngr::Shutdown();
     PixelShaderMngr::Shutdown();
     Fifo_Shutdown();
-    VertexManager::Destroy();
+    VertexManager::Shutdown();
     TextureMngr::Shutdown();
     OpcodeDecoder_Shutdown();
     Renderer::Shutdown();