From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Kiper Date: Wed, 26 Mar 2014 13:48:45 +0000 Subject: Re: [PATCH v3 3/5] x86: Call efi_memblock_x86_reserve_range() on native EFI platform only Message-Id: <20140326134845.GN3454@olila.local.net-space.pl> List-Id: References: <1395781076-12000-1-git-send-email-daniel.kiper@oracle.com> <1395781076-12000-4-git-send-email-daniel.kiper@oracle.com> <20140326130009.GB24856@console-pimps.org> <5332E2B9020000780000255D@nat28.tlf.novell.com> <20140326133119.GD24856@console-pimps.org> <5332E6AE02000078000025F0@nat28.tlf.novell.com> In-Reply-To: <5332E6AE02000078000025F0@nat28.tlf.novell.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Jan Beulich Cc: Matt Fleming , david.vrabel@citrix.com, ian.campbell@citrix.com, stefano.stabellini@eu.citrix.com, jeremy@goop.org, fenghua.yu@intel.com, matt.fleming@intel.com, tony.luck@intel.com, x86@kernel.org, tglx@linutronix.de, xen-devel@lists.xenproject.org, boris.ostrovsky@oracle.com, konrad.wilk@oracle.com, eshelton@pobox.com, mingo@redhat.com, linux-efi@vger.kernel.org, linux-ia64@vger.kernel.org, linux-kernel@vger.kernel.org, hpa@zytor.com On Wed, Mar 26, 2014 at 01:39:42PM +0000, Jan Beulich wrote: > >>> On 26.03.14 at 14:31, wrote: > > On Wed, 26 Mar, at 01:22:49PM, Jan Beulich wrote: > >> >>> On 26.03.14 at 14:00, wrote: > >> > > >> > This could do with a little bit more explanation. Why is it not > >> > necessary to mark the EFI memory map that was passed to the kernel as > >> > reserved in memblock? > >> > >> Because that's in memory Dom0 doesn't even see: The EFI memory > >> map is visible to the hypervisor only. > > > > So where does boot_params.efi_info.efi_memmap point? > > > > If nowhere (i.e. it's NULL) that's no problem because memblock_reserve() > > handles zero size regions just fine. > > That's a question to Daniel - in our implementation (with a separate > Xen kernel that can't run on bare hardware) boot_params as a whole > simply doesn't exist. On my machine this function crashes on Xen so that is why I have changed condition. However, if you say that this issue could be solved in another way I will investigate it further. Daniel