On 19/03/20 00:40, Krish Sadhukhan wrote: >> >> + /* Did the test attempt vmlaunch or vmresume? */ >> + bool vmlaunch; >> + /* Did the instruction VM-Fail? */ >> + bool vm_fail; > > > I still like the old name, "failure_early". To me, "vm_fail" and > "failed_vmentry" sound similar and confusing. > > The SDM calls this type of failure as "early failure" which is denoted > by an (instruction) error number, in order to distinguish it from the > failure that happens during guest state checking/loading. So, probably a > better naming is "vm_early_failure" or "vm_fail_early". Or may be, > "vm_instr_error" ? This is actually what the SDM calls VMfailValid (we should never get to VMfailInvalid in these tests), so vm_fail is appropriate. failed_vmentry is what the SDM calls "VM-entry failure". I agree that the names are similar and confusing, but there's some value in keeping them close to the SDM. Paolo
