0

I have a question about how is possible to modify the Linux scheduler(Linux kernel in general) to specify resources (CPUs) to run an application, to be clear, I'm wondering is there any way (even not efficient and slow) to tell the kernel scheduler that executes the user part of the code on specific core and the kernel part on another one?

I know in terms of execution it may have no sense but in terms of debugging and studying specific subjects, it may help a lot.

I couldn't find any article, book, document about this, I appreciate it if refer me to some references too.

Elephant88
  • 117
  • 1
  • 12
  • I’m voting to close this question because it is too broad – 0___________ Sep 14 '21 at 07:41
  • Maybe one of these? https://stackoverflow.com/questions/33994983/assigning-a-cpu-core-to-a-process-linux https://www.xmodulo.com/run-program-process-specific-cpu-cores-linux.html https://stackpointer.io/unix/linux-assign-process-to-cpu-core/502/ – Aconcagua Sep 14 '21 at 08:18
  • @Aconcagua Thanks, but these are to execute the whole process within a set of cores, workers... I'm looking for executing the User part of the code and the Kernel part on two different resources if it would be possible. – Elephant88 Sep 14 '21 at 09:02
  • 2
    The syscall always runs on the same core as the user process because it needs access to user state. There is no separate scheduling done. – stark Sep 14 '21 at 11:01

0 Answers0