hey guys,

I am having a problem building my code. The kernel is simple, and it runs on my smaller machine with an integrated graphics card. My big machine has 2 NVIDIA GTX 485M cards in it, not running SLI. I get an error code of -11 when I build the program, but it runs just fine on my other system. Any ideas what would cause this?

relevant code
Code :
//setup and get the platforms
	vector<cl::Platform> platforms;
	err = cl::Platform::get(&platforms);
	//setup and get the context
	cl_context_properties properties[3] = { CL_CONTEXT_PLATFORM, (cl_context_properties)platforms[0](), 0 };
	cl::Context context = cl::Context(CL_DEVICE_TYPE_ALL, properties, 0, 0, &err);
	//get a handle to the GPU
	vector<cl::Device> deviceList = context.getInfo<CL_CONTEXT_DEVICES>(&err);
	//create the queue
	cl::CommandQueue queue = cl::CommandQueue(context, deviceList[0]);
	//get the kernel
	//setup the program
	cl::Program::Sources kernelSource(1, std::make_pair(kernelSourceCode, kernelSize));
	cl::Program program = cl::Program(context, kernelSource, &err);
	err = program.build(deviceList);
	std::string str = program.getBuildInfo<CL_PROGRAM_BUILD_LOG>(deviceList[0], &err);
the last line is where it fails, due to a null reference exception in getInfoFunctor1. its not due to an invalid kernel, because it runs on my other system, and the library is linked properly, with the most recent drivers installed. I dont know if it applies, but when I trace the program.build function, it returns 193 for the error, just before converting it to -11, and it doesnt have a null reference exception if I get rid of the buildInfo line, but it still returns err code -11 and the GPU code fails.