riscv: De-Arm a couple of names

[jrtc27]

Summary

These names were inherited from the arm64 port and should be changed to
the RISC-V terminology.

[markj]

sys/riscv/riscv/trap.c

354

data_abort() also seems like an ARMism.

[kp]

You missed the opportunity for a disarming pun.

[jrtc27]

In D26671#594266, @kp wrote:

You missed the opportunity for a disarming pun.

Dearm is also in the OED, even if it's marked obsolete :)

[jrtc27]

Also rename data_abort.

[jrtc27]

sys/riscv/riscv/trap.c

354

TIL :)

[jrtc27]

sys/riscv/riscv/trap.c

290

It is odd that we don't handle instruction page faults here... I assume because of the way we map the kernel we know we _shouldn't_ get them, but currently if we do we will end up taking the default case and printing "Unknown kernel exception", which is fine but not technically true...

[markj]

sys/riscv/riscv/trap.c

290

I was wondering about that too. I can't find a reason to not pass kernel-mode instruction fetch page faults to page_fault_handler(). vm_fault_trap() should trigger a panic in that case.

[jhb]

sys/riscv/riscv/trap.c

290

+1