Hello all,

I’m almost sure but want to check if my understanding is correct.

To use a memory mapped IO, we use functions like request_mem_region and ioremap, iounmap functions.

I guess when we call ioremap function, the kernel sets up page table for the region so that the region becomes non-cacheable, and I don’t also have  to worry about reordering problem.

Is my understanding correct?

Thanks!

Chan Kim

_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies