Added profile blocks for GL.

Engine/source/gfx/gl/gfxGLCircularVolatileBuffer.h

@@ -21,6 +21,7 @@ public:
 
    void init(U32 start, U32 end)
    {  
+      PROFILE_SCOPE(GFXGLQueryFence_issue);
       mStart = start;
       mEnd = end;
       mSync = glFenceSync(GL_SYNC_GPU_COMMANDS_COMPLETE, 0);
@@ -36,6 +37,7 @@ public:
 
    void wait()
    {   
+      PROFILE_SCOPE(GFXGLQueryFence_block);
       GLbitfield waitFlags = 0;
       GLuint64 waitDuration = 0;
       while( 1 ) 

Engine/source/gfx/gl/gfxGLDevice.cpp

@@ -325,6 +325,7 @@ void GFXGLDevice::resurrect()
 
 GFXVertexBuffer* GFXGLDevice::findVolatileVBO(U32 numVerts, const GFXVertexFormat *vertexFormat, U32 vertSize)
 {
+   PROFILE_SCOPE(GFXGLDevice_findVBPool);
    for(U32 i = 0; i < mVolatileVBs.size(); i++)
       if (  mVolatileVBs[i]->mNumVerts >= numVerts &&
             mVolatileVBs[i]->mVertexFormat.isEqual( *vertexFormat ) &&
@@ -333,6 +334,7 @@ GFXVertexBuffer* GFXGLDevice::findVolatileVBO(U32 numVerts, const GFXVertexForma
          return mVolatileVBs[i];
 
    // No existing VB, so create one
+   PROFILE_SCOPE(GFXGLDevice_createVBPool);
    StrongRefPtr<GFXGLVertexBuffer> buf(new GFXGLVertexBuffer(GFX, numVerts, vertexFormat, vertSize, GFXBufferTypeVolatile));
    buf->registerResourceWithDevice(this);
    mVolatileVBs.push_back(buf);
@@ -358,6 +360,7 @@ GFXVertexBuffer *GFXGLDevice::allocVertexBuffer(   U32 numVerts,
                                                    GFXBufferType bufferType,
                                                    void* data )  
 {
+   PROFILE_SCOPE(GFXGLDevice_allocVertexBuffer);
    if(bufferType == GFXBufferTypeVolatile)
       return findVolatileVBO(numVerts, vertexFormat, vertSize);
          
@@ -523,6 +526,7 @@ inline GLsizei GFXGLDevice::primCountToIndexCount(GFXPrimitiveType primType, U32
 
 GFXVertexDecl* GFXGLDevice::allocVertexDecl( const GFXVertexFormat *vertexFormat ) 
 {
+   PROFILE_SCOPE(GFXGLDevice_allocVertexDecl);
    typedef Map<void*, GFXGLVertexDecl> GFXGLVertexDeclMap;
    static GFXGLVertexDeclMap declMap;   
    GFXGLVertexDeclMap::Iterator itr = declMap.find( (void*)vertexFormat->getDescription().c_str() ); // description string are interned, safe to use c_str()
@@ -855,6 +859,7 @@ void GFXGLDevice::setShader(GFXShader *shader, bool force)
 
 void GFXGLDevice::setShaderConstBufferInternal(GFXShaderConstBuffer* buffer)
 {
+   PROFILE_SCOPE(GFXGLDevice_setShaderConstBufferInternal);
    static_cast<GFXGLShaderConstBuffer*>(buffer)->activate();
 }
 

Engine/source/gfx/gl/gfxGLShader.cpp

@@ -344,6 +344,7 @@ void GFXGLShaderConstBuffer::set(GFXShaderConstHandle* handle, const MatrixF* ma
 
 void GFXGLShaderConstBuffer::activate()
 {
+   PROFILE_SCOPE(GFXGLShaderConstBuffer_activate);
    mShader->setConstantsFromBuffer(this);
    mWasLost = false;
 }
@@ -394,6 +395,7 @@ void GFXGLShader::clearShaders()
 
 bool GFXGLShader::_init()
 {
+   PROFILE_SCOPE(GFXGLShader_Init);
    // Don't initialize empty shaders.
    if ( mVertexFile.isEmpty() && mPixelFile.isEmpty() )
       return false;
@@ -1013,6 +1015,7 @@ bool GFXGLShader::initShader( const Torque::Path &file,
                               bool isVertex, 
                               const Vector<GFXShaderMacro> &macros )
 {
+   PROFILE_SCOPE(GFXGLShader_CompileShader);
    GLuint activeShader = glCreateShader(isVertex ? GL_VERTEX_SHADER : GL_FRAGMENT_SHADER);
    if(isVertex)
       mVertexShader = activeShader;
@@ -1072,6 +1075,7 @@ bool GFXGLShader::initShader( const Torque::Path &file,
 /// Returns our list of shader constants, the material can get this and just set the constants it knows about
 const Vector<GFXShaderConstDesc>& GFXGLShader::getShaderConstDesc() const
 {
+   PROFILE_SCOPE(GFXGLShader_GetShaderConstants);
    return mConstants;
 }
 

Engine/source/gfx/gl/gfxGLStateBlock.cpp

@@ -88,6 +88,7 @@ const GFXStateBlockDesc& GFXGLStateBlock::getDesc() const
 /// @param oldState  The current state, used to make sure we don't set redundant states on the device.  Pass NULL to reset all states.
 void GFXGLStateBlock::activate(const GFXGLStateBlock* oldState)
 {
+   PROFILE_SCOPE(GFXGLStateBlock_Activate);
    // Big scary warning copied from Apple docs 
    // http://developer.apple.com/documentation/GraphicsImaging/Conceptual/OpenGL-MacProgGuide/opengl_performance/chapter_13_section_2.html#//apple_ref/doc/uid/TP40001987-CH213-SW12
    // Don't set a state that's already set. Once a feature is enabled, it does not need to be enabled again.

Engine/source/gfx/gl/gfxGLTextureManager.cpp

@@ -234,6 +234,7 @@ static void _fastTextureLoad(GFXGLTextureObject* texture, GBitmap* pDL)
    
    if(pDL->getFormat() == GFXFormatR8G8B8A8 || pDL->getFormat() == GFXFormatR8G8B8X8)
    {
+      PROFILE_SCOPE(Swizzle32_Upload);
       FrameAllocatorMarker mem;
       U8* pboMemory = (U8*)mem.alloc(bufSize);
       GFX->getDeviceSwizzle32()->ToBuffer(pboMemory, pDL->getBits(0), bufSize);
@@ -241,6 +242,7 @@ static void _fastTextureLoad(GFXGLTextureObject* texture, GBitmap* pDL)
    }
    else
    {
+      PROFILE_SCOPE(SwizzleNull_Upload);
       glBufferSubData(GL_PIXEL_UNPACK_BUFFER_ARB, 0, bufSize, pDL->getBits(0) );
    }
    
@@ -262,6 +264,7 @@ static void _slowTextureLoad(GFXGLTextureObject* texture, GBitmap* pDL)
 
 bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, GBitmap *pDL)
 {
+   PROFILE_SCOPE(GFXGLTextureManager_loadTexture);
    GFXGLTextureObject *texture = static_cast<GFXGLTextureObject*>(aTexture);
    
    AssertFatal(texture->getBinding() == GL_TEXTURE_1D || texture->getBinding() == GL_TEXTURE_2D, 
@@ -291,6 +294,8 @@ bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, GBitmap *pDL)
 
 bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, DDSFile *dds)
 {
+   PROFILE_SCOPE(GFXGLTextureManager_loadTextureDDS);
+
    AssertFatal(!(dds->mFormat == GFXFormatDXT2 || dds->mFormat == GFXFormatDXT4), "GFXGLTextureManager::_loadTexture - OpenGL does not support DXT2 or DXT4 compressed textures");
    GFXGLTextureObject* texture = static_cast<GFXGLTextureObject*>(aTexture);
    
@@ -308,6 +313,8 @@ bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, DDSFile *dds)
       numMips = 1;
    for(U32 i = 0; i < numMips; i++)
    {
+      PROFILE_SCOPE(GFXGLTexMan_loadSurface);
+
       if(isCompressedFormat(dds->mFormat))
       {
          if((!isPow2(dds->getWidth()) || !isPow2(dds->getHeight())) && GFX->getCardProfiler()->queryProfile("GL::Workaround::noCompressedNPoTTextures"))
@@ -344,6 +351,7 @@ bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, DDSFile *dds)
 
 bool GFXGLTextureManager::_loadTexture(GFXTextureObject *aTexture, void *raw)
 {
+   PROFILE_SCOPE(GFXGLTextureManager_loadTextureRaw);
    if(aTexture->getDepth() < 1)
       return false;
    

Engine/source/gfx/gl/gfxGLTextureObject.cpp

@@ -96,6 +96,9 @@ void GFXGLTextureObject::unlock(U32 mipLevel)
    if(!mLockedRect.bits)
       return;
 
+   // I know this is in unlock, but in GL we actually do our submission in unlock.
+   PROFILE_SCOPE(GFXGLTextureObject_lockRT);
+
    PRESERVE_TEXTURE(mBinding);
    glBindTexture(mBinding, mHandle);
    glBindBuffer(GL_PIXEL_UNPACK_BUFFER_ARB, mBuffer);
@@ -137,6 +140,8 @@ void GFXGLTextureObject::reInit()
 
 bool GFXGLTextureObject::copyToBmp(GBitmap * bmp)
 {
+   PROFILE_SCOPE(GFXGLTextureObject_copyToBmp);
+
    if (!bmp)
       return false;
 
@@ -175,6 +180,7 @@ bool GFXGLTextureObject::copyToBmp(GBitmap * bmp)
 
    glGetTexImage(mBinding, 0, GFXGLTextureFormat[mFormat], GFXGLTextureType[mFormat], orig);
    
+   PROFILE_START(GFXGLTextureObject_copyToBmp_pixCopy);
    for(int i = 0; i < srcPixelCount; ++i)
    {
       dest[0] = orig[0];
@@ -186,6 +192,7 @@ bool GFXGLTextureObject::copyToBmp(GBitmap * bmp)
       orig += srcBytesPerPixel;
       dest += dstBytesPerPixel;
    }
+   PROFILE_END();
 
    return true;
 }