Rendering to texture problem, part 2

Hi all,

I am trying to replicate (in TSE) the Minimap rendering from the RTS kit. After a little hunting around I was able to figure out how to render to a texture and got the terrain to show up. Unfortunately, I was only able to get part of it to draw.

The way the RTS minimap rendering code works is that it takes the mission area and divides it up into "chunks" based on the size of the texture. In my case I'm using a 200x200 texture, so the minimap rendering chops the terrain up into 4 equal rectangles, positions the camera in the center of those 4 chunks, and then renders a chunk to a 100x100 pixel section of the backbuffer. Then the code copies that 100x100 pixel section into the appropriate part of the rendered texture, and then does the next chunk. At least that's my understanding.

What I'm seeing is that the last chunk to get rendered gets stretched over the entire bitmap. Actually it seems like it's only a portion of the last chunk but it's kind of hard to tell. In any case, it's wiping out the rest of the bitmap (assuming they were ever copied in properly). I was just hoping somebody could look at the original RTS code and my TSE-converted version and tell me if I made an obvious mistake. I'm really clueless about the GFX library, so in some cases I was just taking my best guess. In particular I'm wondering if the GFX->drawBitmapSR function is the right one to use (and that I called it correctly).

So here's the rendering portion of the original RTS kit code:

Con::printf("Map generation beginning...");
   for (S32 v = 0; v < subDivs; v++)
   {
      sCurrRenderPos.y = start.y + v * renderSquareSize;
      for (S32 u = 0; u < subDivs; u++)
      {
         sCurrRenderPos.x = start.x + u * renderSquareSize;

         glClear(GL_DEPTH_BUFFER_BIT);
         Parent::onRender(mBounds.point, mBounds);

         glFinish();
         glBindTexture(GL_TEXTURE_2D, mRenderTexture.getGLName());
         glCopyTexSubImage2D(GL_TEXTURE_2D, 0,
            texelSquareSize*u, texelSquareSize*v, 0, 0, texelSquareSize, texelSquareSize);
      }
   }

And here's my version in TSE:

Con::printf("Map generation beginning...");
	
	GFX->startActiveRenderSurface(mRenderTexture);
	GFX->setZEnable( true );
	GFX->clear( GFXClearZBuffer | GFXClearStencil | GFXClearTarget, ColorI( 255, 0, 0 ), 1.0f, 0 );

	for (S32 v = 0; v < subDivs; v++)
	{
		sCurrRenderPos.y = start.y + v * renderSquareSize;
		for (S32 u = 0; u < subDivs; u++)
		{
			sCurrRenderPos.x = start.x + u * renderSquareSize;

			GFX->clear(GFXClearZBuffer | GFXClearStencil, ColorI( 255, 0, 0 ), 1.0f, 0);
			
			Parent::onRender(mBounds.point, mBounds);
			
			GFX->drawBitmapSR(mRenderTexture, Point2I(texelSquareSize*u, texelSquareSize*v),
				RectI(0, 0, texelSquareSize, texelSquareSize));
				 
		}
	}

	GFX->endActiveRenderSurface();

Thanks!