From mboxrd@z Thu Jan 1 00:00:00 1970 From: Borislav Petkov Subject: Re: [PATCH 0/4] EFI 1:1 mapping Date: Tue, 4 Jun 2013 10:15:44 +0200 Message-ID: <20130604081544.GA23640@pd.tnic> References: <1370177770-26661-1-git-send-email-bp@alien8.de> <20130602225620.GA5496@srcf.ucam.org> <20130603081148.GB13607@nazgul.tnic> <20130603143252.GB20252@srcf.ucam.org> <20130603145412.GJ2004@console-pimps.org> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Return-path: Content-Disposition: inline In-Reply-To: <20130603145412.GJ2004-HNK1S37rvNbeXh+fF434Mdi2O/JbrIOy@public.gmane.org> Sender: linux-efi-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Matt Fleming Cc: Matthew Garrett , Linux EFI , Jiri Kosina , X86-ML , LKML , Borislav Petkov , James Bottomley List-Id: linux-efi@vger.kernel.org On Mon, Jun 03, 2013 at 03:54:12PM +0100, Matt Fleming wrote: > On Mon, 03 Jun, at 03:32:52PM, Matthew Garrett wrote: > > We can only pass one set of addresses to SetVirtualAddressMap(), but it > > doesn't seem like there's any intrinsic reason we can't the runtime > > regions mapped to multiple virtual addresses. > > Indeed. That's the approach I took with my 1:1 series from last year. If > Windows is mapping things at higher addresses like you said, then > they're probably doing the same. Right, the way I've done it now is to do the virtual mapping unconditionally, then do the 1:1 mapping and call SetVirtualAddressMap (f*cking camelcase is so nasty to type - that's why they need whole IDEs to program :-)) with that map, if "efi=1:1_map" has been passed on the cmd line. After the call, we use *only* the 1:1 map but the virtual mapping is still there. The initial approach to addressing the b0rked Macs would then be to never use the 1:1 map on them. -- Regards/Gruss, Boris. Sent from a fat crate under my desk. Formatting is fine. --