Excerpts from Fabiano Rosas's message of February 13, 2021 6:33 am: > Nicholas Piggin <npiggin@xxxxxxxxx> writes: > >> Move the GUEST_MODE_SKIP logic into KVM code. This is quite a KVM >> internal detail that has no real need to be in common handlers. > > LGTM, > >> >> (XXX: Need to confirm CBE handlers etc) > > Do these interrupts exist only in Cell? I see that they set HSRRs and > MSR_HV, but CPU_FTRS_CELL does not contain CPU_HVMODE. So I don't get > why they use the KVM macros. Good question, I asked Michael Ellerman and he said there is a bare metal Cell which predates or otherwise does not define HVMODE. However it does not support KVM. So I think we can remove it. > And for the instruction_breakpoint (0x1300) I think it would help if we > could at least restrict when it is built. But I don't know what > ISA/processor version it is from. Yeah we should be documenting these non-architected handlers a little better IMO. I think we can get rid of the kvm skip from this one though. I've done that in a separate patch in the next series. Thanks, Nick
