gsdx sw: wrap GS memory

Cost ought to remain small. Worst case is 2 extra "and" operation by group of pixels in scanline renderer

I think PixelAddressN functions are mostly call in the init.

plugins/GSdx/GSDrawScanlineCodeGenerator.x86.cpp

@@ -611,6 +611,7 @@ void GSDrawScanlineCodeGenerator::TestZ(const Xmm& temp1, const Xmm& temp2)
 
 	mov(ebp, ptr[esi + 4]);
 	add(ebp, ptr[edi + 4]);
+	and(ebp, HALF_VM_SIZE - 1);
 
 	// GSVector4i zs = zi;
 
@@ -2360,6 +2361,7 @@ void GSDrawScanlineCodeGenerator::ReadFrame()
 
 	mov(ebx, ptr[esi]);
 	add(ebx, ptr[edi]);
+	and(ebx, HALF_VM_SIZE - 1);
 
 	if(!m_sel.rfb)
 	{