On 26 July 2015 at 13:46, Leif Lindholm <[email protected]> wrote: > On Sun, Jul 26, 2015 at 01:08:20PM +0200, Ard Biesheuvel wrote: >> > So ... had a look through, looks sane, have a couple of minor comments >> > (will add inline). >> > But what I don't have at home is a useful hardware platform to test >> > this on. Has anyone else tested on 32-bit/64-bit hardware? >> >> Only under KVM, which is not very useful to you. > > Not sufficiently. >
I tried it under FVP (after spotting and fixing a kernel bug, see separate email) and it works fine. >> I was kind of hoping you could push it through the CI Olivier always >> talked about. Or was that his personal setup? > > For all intents and purposes. > >> Do note that, even if they are non-trivial patches, the only actual >> change to non-virt platforms is that, for non-SEC modules, the result >> of the GIC detection is cached in a global. > > Sure. I'm just being paranoid. > > Well, I'll have a look tomorrow, while looking around for any hw > people might not miss... > If you do mind the ArmGicArchSecLib clone and subsequent change of ArmGicArchLib, we could drop that, since the other patches still allow the virt platforms to use the GIC revision specified in the device tree (which is the primary motivation for this series) -- Ard. _______________________________________________ edk2-devel mailing list [email protected] https://lists.01.org/mailman/listinfo/edk2-devel
