System calls with cuda

I’m trying to integrate CPU code into GPU code. In particular, I want to do something like system calls embedded in GPU code.

I have already read in a paper that I can only run the Linux system calls on the CPU. Unfortunately, there is no mention of how this is implemented in CUDA code. Can anyone help me?

Hi,

It’s recommended to divide the CUDA implementation into two parts and use the CPU to launch the system call.

For example:

launch cuda code 1
wait
system call
launch cuda code 2
wait

Thanks.

okay

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.