Program control doesn't enter into __global__ function in CUDA C

Accelerated Computing CUDA CUDA Programming and Performance
1

When the CUDA c program is executed on windows 7 , When the program control reaches to global function following error is displayed

Find Source tmpxft_00001588_00000000-2_kernel.cudafe1.stub.c

and in output variable value is displayed as 0 (Zero)

PRogram is correct n gets compiled also.

Kindly reply

2

“PRogram is correct n gets compiled also.”

compilation does not guarantee execution - the reason why debuggers are needed

although i can not confirm this: stub is generally associated with the kernel configuration - the pre-kernel launch evaluator

i) post your code

also:

a) use the debugger and step your program; note the first line of divergence

b) add error checking (cudaGetLastError()) before and after your kernel launch

c) add a breakpoint within your kernel, and see if you can reach that

d) interpret the results of a), b) and c)