Hi all,
Here’s an example of my code
global void eample(int *array)
{
…
int tx = blockIdx.x * blockDim.x + threadIdx.x; //break point is here to verify tx’s value
int ty = blockIdx.y * blockDim.y + threadIdx.y;
if(tx < 10)
{
int position = tx+1; //break point is here to verify position’s value
array[position] = position+3; //break point is here to catch incase top break point fails
}
…
}
After stepping into the global function from the kernel call using F11 on visual studios, all the break points are missed and the control returns back to my main().
However, verifying the array values back in main shows that the global function has indeed been executed correctly.
Can anyone help me out with this? Is there a fix? I need to debug within the global function because of the math calculations involved. The calculation example in my example above is highly simplified.
I’m very new to CUDA gpu programming, please go easy on the terms and steps in the explanation. :) Thank you!