In RISCV architecture, there are pmp registers that define and control the memory regions. It is stated in the spec "Although the PMP mechanism supports regions as small as four bytes, platforms may specify coarser PMP region". I understand that I can define the minimum region as 16-byte instead of 4-byte in my core (Linux bootable). In other words, I can specify that memory regions should be aligned 16-byte at least. I wonder about the advantages and disadvantages of making a coarser minimum PMP region.
On my side making minimum regions as large as the cache line decreases the hardware cost and complexity. However, I worry about memory usage which is increased by these coarser regions (I am not sure). I also wonder if any other drawbacks or problems making coarser regions.