fixed
	+ SSE / SSE 2 Detection routine added
	+ openGL Header files updated to current OpenGL release
	+ generated and updated glh_genext.h with extgen from (slightly modified to match libs in sl)
		- http://mongoose.dyndns.org/code/cs459/w-demo/inc/glh/extgen/?C=S;O=A
		- http://www.koders.com/info.aspx?c=ProjectInfo&pid=1REQGRL8XKLTF3VCTC8YR8YEMG&s=windows.h
	+ VRAM detection and dependencies of gGLManager.mVRAM fixed
		- now Texture RAM calculation and everything that depends on gGLManager.mVRAM should work
	


####################################################
## Changes
####################################################

linden\indra\newview\llviewerjointmesh.cpp
@840
---	sVectorizePerfTest = ( sVectorizePerfTest ? sVectorizePerfTest : gSavedSettings.getBOOL("VectorizePerfTest") );
	sVectorizeProcessor = gSavedSettings.getU32("VectorizeProcessor");
	BOOL vectorizeEnable = gSavedSettings.getBOOL("VectorizeEnable");
	BOOL vectorizeSkin = gSavedSettings.getBOOL("VectorizeSkin");
+++ sVectorizePerfTest = gSavedSettings.getBOOL("VectorizePerfTest");
	BOOL vectorizeEnable = gSavedSettings.getBOOL("VectorizeEnable");
	BOOL vectorizeSkin = gSavedSettings.getBOOL("VectorizeSkin");
	if (sVectorizePerfTest) {
		// get CPUInfos to ask hasSSE | hasSSE2 if cpu supports sse|sse2
		LLCPUInfo CPUInfo;
		llinfos << "CPU supports SSE " << ( CPUInfo.hasSSE() ? "TRUE" : "FALSE" ) << llendl;
		llinfos << "CPU supports SSE2 " << ( CPUInfo.hasSSE2() ? "TRUE" : "FALSE" ) << llendl;
		if (CPUInfo.hasSSE() && CPUInfo.hasSSE2()) { sVectorizeProcessor = 2; }
		if (CPUInfo.hasSSE() && !sVectorizeProcessor) { sVectorizeProcessor = 1; }
		if (sVectorizeProcessor != 0) { 
			vectorizeEnable = 1;
			vectorizeSkin = 1;
		}
	}

	
linden\indra\newview\llapviewerwin32.cpp
@465 ' gGLManager.mVRAM = gDXHardware.getVRAM();
- Hardware must be probed every startup to fill this var!! So i changed
@403
--- //if (FALSE == gSavedSettings.getBOOL("NoHardwareProbe"))
+++	if (FALSE == 0)

- Furthermore if gGLMananger.mVRAM is null
  LLViewerImageList::getMaxVideoRamSetting(bool get_recommended) will not test max Texture Mem!!
  llviwerimagelist.cpp
  @941, if (gGLManager.mVRAM != 0)
  
- Furthermore LLFloaterHardwareSettings::refreshEnabledState() in llfloaterhardwaresettings.cpp @92,93 will not work correctly 
  and GrapicsCardTextureMemory will not correctly updated during refresh


linden\indra\llrender\llgl.cpp
@112 --- // GL_ARB_point_parameters
@113 --- PFNGLPOINTPARAMETERFARBPROC glPointParameterfARB = NULL;
@114 --- PFNGLPOINTPARAMETERFVARBPROC glPointParameterfvARB = NULL;
@248 / 243
--- #if LL_WINDOWS
	PFNWGLSWAPINTERVALEXTPROC			wglSwapIntervalEXT = NULL;
	#endif
# Those changes where done cause glPointParameterfARB will be initiated during extensions call now (updated glh_genext.h defined there)

@591 / 597 # only for non ATI?? thats curios and one of some other things i dont understand
--- mHasPointParameters = !mIsATI && ExtensionExists("GL_ARB_point_parameters", gGLHExts.mSysExts);
+++ mHasPointParameters = ExtensionExists("GL_ARB_point_parameters", gGLHExts.mSysExts);
@707 / 720 // didn't noticed any performance issues so i disabled it (ATI has an mip-map opt feature @driver stage)
--- LL_INFOS("RenderInit") << "Disabling mip-map generation for ATI GPUs (performance opt)" << LL_ENDL;
    mHasMipMapGeneration = FALSE;
+++ LL_INFOS("RenderInit") << "__NOT__ Disabling mip-map generation for ATI GPUs (performance opt)" << LL_ENDL;
	//mHasMipMapGeneration = FALSE;

linden\indra\newview\llappviewer.cpp
@407
??? LLCubeMap::sUseCubeMaps = LLFeatureManager::getInstance()->isFeatureAvailable("RenderCubeMap");
# tests a feature that not in feature list by default. Would be if the featurestable.txt file would read before and
# this feature is added. Cause LLFeatureManager::loadFeatureTables() is called later it outputs a warning ... 
# The extension will be loaded regardless of that warning message so i dont care.



####################################################
## Notes
####################################################
linden\indra\llrender\llgl.cpp - @591 GL_ARB_point_parameters not for ATI???
featuretable.txt  - added "RenderCubeMap				1	1" right after "RenderAvatarVP				1	1" to all modes

## - Added those lines to gpu_tables.txt for recognizing  ATI Cards some still exists before
## - Copied from an existing jira note
ATI Mobility Radeon HD 2300		.*ATI.*Mobility.*HD.*23.*			1		1
ATI Mobility Radeon HD 2400		.*ATI.*Mobility.*HD.*24.*			1		1
ATI Mobility Radeon HD 2600		.*ATI.*Mobility.*HD.*26.*			3		1
ATI Mobility Radeon HD 3400		.*ATI.*Mobility.*HD.*34.*			1		1
ATI Mobility Radeon HD 3600		.*ATI.*Mobility.*HD.*36.*			3		1
ATI Mobility Radeon HD 3800		.*ATI.*Mobility.*HD.*38.*			3		1
ATI Radeon HD 2300				.*ATI.*Radeon HD 23.*				0		1
ATI Radeon HD 2400				.*ATI.*Radeon HD.*24.*				1		1
ATI Radeon HD 2600				.*ATI.*Radeon HD 26.*				2		1
ATI Radeon HD 2900				.*ATI.*Radeon HD 29.*				3		1
ATI Radeon HD 3200				.*ATI.*Radeon HD.*32.*				1		1
ATI Radeon HD 3300				.*ATI.*Radeon HD.*33.*				1		1
ATI Radeon HD 3400				.*ATI.*Radeon HD.*34.*				1		1
ATI Radeon HD 3600				.*ATI.*Radeon HD.*36.*				3		1
ATI Radeon HD 3800				.*ATI.*Radeon HD.*38.*				3		1
ATI Radeon HD 4300				.*ATI.*Radeon HD 43.*				1		1
ATI Radeon HD 4600				.*ATI.*Radeon HD 46.*				3		1
ATI Radeon HD 4800				.*ATI.*Radeon HD.*48.*				3		1


####################################################
## Current Stop Points - VS Stuff
####################################################
L939 llviewerimagelist.cpp
L321 llgl.cpp
L316 llgl.cpp
L311 llgl.cpp
L306 llgl.cpp
L075 glh_extensions.h
L114 llfeaturemanager.cpp
L212 llfeaturemanager.cpp
L290 llfeaturemanager.cpp
L407 llappviewer.cpp



####################################################
## Performance Tipps // just for testing purpose
####################################################
DisableVerticalSync: TRUE
RenderAppleUseMultiGL: FALSE
RenderAvatarCloth: TRUE
RenderAvatarVP: TRUE
RenderCubeMap: TRUE
RenderUseFBO: TRUE
RenderVBOEnable: TRUE
RunMultipleThreads: TRUE
VectorizeEnable: TRUE
VectorizeProcessor: TRUE
VectorizeSkin: TRUE