On Fri, 23 Oct 2020 at 14:47, Leif Lindholm <leif@xxxxxxxxxxxx> wrote: > > On Fri, Oct 23, 2020 at 14:08:25 +0200, Ard Biesheuvel wrote: > > Now that we implemented supported for the LoadFile2 protocol for initrd > > loading, there is no longer a need to pass the initrd parameters via > > the device tree. This means there is no longer a reason to update the > > device tree in the first place, and so we can ignore it entirely. > > There is a change in behaviour here which I don't think matters, but > I'll call it out anyway: > If there was ever a kernel out there with an EFI stub that depended on > a chosen node existing in the DT, and the one provide by firmware did > not contain one, that setup would break from this *if* it didn't use > an initrd. > I checked the Linux source, and the original code contributed by Roy already contained the logic to create the /chosen node if it wants there already. So we should be fine here.
