On 02/21/19 13:38, Peter Maydell wrote: > On Thu, 21 Feb 2019 at 12:07, Laszlo Ersek <ler...@redhat.com> wrote: >> since we're talking "reset_flash", I'll note that there is no actual >> reset handler for cfi.pflash01. I found out recently, via: >> >> https://bugzilla.redhat.com/show_bug.cgi?id=1678713 > > Yes; this isn't uncommon for some of the really old > device models. It should definitely have one added. > > You are correct also that the timer in the pflash_cfi01 > model is dead code -- it has always been so, since the > device was added in 2007. The reason it is there is that > pflash_cfi01 was created as a copy-and-hack of the > cfi02 device. In cfi02 we do use the timer, as a way > of simulating "make full-chip and sector erases take a > guest-visible amount of time rather than completing > instantaneously". cfi01 doesn't do that (and I think > may not implement anything other than block erase), > but the timer initialization code was left in rather > than being deleted as part of the copy-and-hack.
Thank you, I've linked this back into the RHBZ. Laszlo
