This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


RE: [Xen-ia64-devel] elilo allocations

To: "Luck, Tony" <tony.luck@xxxxxxxxx>, <xen-ia64-devel@xxxxxxxxxxxxxxxxxxx>
Subject: RE: [Xen-ia64-devel] elilo allocations
From: "Tian, Kevin" <kevin.tian@xxxxxxxxx>
Date: Fri, 9 Sep 2005 08:32:12 +0800
Delivery-date: Fri, 09 Sep 2005 00:29:57 +0000
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-ia64-devel-request@lists.xensource.com?subject=help>
List-id: Discussion of the ia64 port of Xen <xen-ia64-devel.lists.xensource.com>
List-post: <mailto:xen-ia64-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-ia64-devel>, <mailto:xen-ia64-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx
Thread-index: AcW0Id4vtWV493v5R++EH8sijc3mtwAeGklAAA64+pA=
Thread-topic: [Xen-ia64-devel] elilo allocations
>From: Luck, Tony
>>I think normally elilo will place parameters, initrd, etc. in higher
>>of physical memory, and it's unlikely to map all memories by one TR.
>>Actually the alt dtlb miss fault is wanted for Xen to access all the
>>physical memories. BTW, can you check whether 0x617ef90010 is legal
>>address in a memory bank on your box?
>Elilo just uses EFI allocator ... current implementation of EFI is
>allocating from higher addresses (on Tiger it appears to choose high
>addresses, but doesn't go over 2G/4G ... difficult to see whether it
>is a 31 or 32 bit limit as there is no memory in the 2-4G range).

Is EFI allocator platform independent? If true, then reasonable to see
limitation under 2G/4G for accommodating 32bit platform. ;-) Also is it
possible that specific box uses specific version of Elilo to behave
differently out of 4G?

>I think this may just be an implementation quirk.  The allocation
>could be at any address, so Linux must be prepared to cope with
>whatever it is passed.  It is definitely true that the existing
>implementation ends up with these out of range of DTR[0] that is
>mapping the kernel itself, so the Alt_DTLB handler must provide the
>mapping when these objects are accessed.



Xen-ia64-devel mailing list

<Prev in Thread] Current Thread [Next in Thread>