[CUDA] Use appropriate return code for out of registers kernel launch

Change the returned error code for exiting the kernel launch entry point in CUDA
when exceeding the maximum available registers for execution on the SM.
Previously we were returning a misleading error code.

source/adapters/cuda/enqueue.cpp

@@ -256,7 +256,7 @@ setKernelParams(const ur_context_handle_t Context,
 
         if (hasExceededMaxRegistersPerBlock(Device, Kernel,
                                             KernelLocalWorkGroupSize)) {
-          return UR_RESULT_ERROR_INVALID_WORK_GROUP_SIZE;
+          return UR_RESULT_ERROR_OUT_OF_RESOURCES;
         }
       } else {
         guessLocalWorkSize(Device, ThreadsPerBlock, GlobalWorkSize, WorkDim,