On Thu Jun 15, 2023 at 7:19 PM AEST, BALATON Zoltan wrote: > On Thu, 15 Jun 2023, Nicholas Piggin wrote: > > On Thu Jun 15, 2023 at 7:34 AM AEST, BALATON Zoltan wrote: > >> Changing the parameter of cpu_interrupt_exittb() from CPUState to env > >> allows removing some more local CPUState variables in callers. > > > > I think it's more consistent to keep cs, which is same as > > cpu_interrupt(). > > But with this patch it's more consistent with the other functions devlared > in helper_regs.h and gets rid of the #ifdef in hreg_store_msr() so I'd > still like to keep this patch. Callers already have env so it should not > matter.
Being consistent with functions of the same file is not important or really makes sense. It's important to be consistent with functions of similar type. cpu_interrupt_exittb() is a helper to call cpu_interrupt() so makes sense to be similar. At best it seems like pointless churn. Thanks, Nick