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


[Xen-ia64-devel] [patch 00/16] Kexec v20080125 - xen

To: xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-ia64-devel] [patch 00/16] Kexec v20080125 - xen
From: Simon Horman <horms@xxxxxxxxxxxx>
Date: Fri, 25 Jan 2008 17:19:58 +0900
Delivery-date: Fri, 25 Jan 2008 00:39:06 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
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
User-agent: quilt/0.46-1

Sorry that it has been so long since I have posted patches here.
And doubly sorry that this batch still has problems. But here goes anyway.

The major changes:
- Fix up the EFI RID code
  - I was mistakenly using rr[6] and rr[7] instead of rr[6<<61] and
    rr[7<<61]. Or in other words I was setting the RID of region 0
    (twice) instead of the RID of regions 6 and 7. Once I fixed
    this minor oversight I was left with very much not working code.
  - Implemented repining of TR entries when switching rr[7] to EFI_RID.
  - Turned off VHPT for EFI_RID.
  - This all took way to long!
- Unpinning
  - The unpinning of both the rebooting cpu and other cpus was wrong.
    I have fixed this (I hope).
- I also fixed some problems that Alex Williamson saw on an HP rx3600.
  I hope they haven't regressed in the mean time.


This patch is based on ext/ia64/xen-unstable.hg revision 16216:9adec82f7bfe.
It is an old revision becuase the following patch adds a TR entry for
the heap, which needs to be unpinned on kexec. I made a botched
attempt to do this, then got side-tracked by the RID offset issue above.
I believe that subsequently thee xen heap tr patch has been removed.
In any case, I indend to upport soon :-)


This code has been tested on an HP rx2620 and a Tiger 4
(n.b I usually use a Tiger 2).


Kexecing from Xen -> Xen has been the major focus and that seems to work.
In testing just before the release I noticed that both Kexec Xen -> Linux
and Kdump (Xen -> Linux) seem to get stuck in purgatory - I hate it when
that happens! I will investigate this problem.

For now, unless you're feeling particularly giving with your time,
please just test Xen -> Xen kexec.

Other known problems are listed in the PROBLEMS file.

Other patches:

There is a xen-linux patch to relocate_new_kernel. This is required.
For the companion patch in this series.


I will be attending Linux.Conf.Au in Melbourne next week.
If anyone who is attending wants to catch up please let me know.


Xen-ia64-devel mailing list