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-devel] hardware support

To: <xen-devel@xxxxxxxxxxxxxxxxxxx>
Subject: [Xen-devel] hardware support
From: "Ky Srinivasan" <ksrinivasan@xxxxxxxxxx>
Date: Mon, 08 May 2006 09:23:36 -0600
Delivery-date: Mon, 08 May 2006 08:24:10 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
We have had some problems bringing up SLES10/Xen on DL360 box (dual
processor). The Compaq boxes have this notion of "auto" mode for
MPS/APIC to deal with buggy OSes and under this setting (which is the
default setting), vanilla Linux (2.6.16) comes up on this box in UP mode
since it finds the local APIC disabled in bios. However, the xen based
kernel tries to come up MP and fails in a non-intuitive way by  hitting
a BUG in the hypervisor code. Following is the stack trace of the xen
based kernel:

[<ff10f8d6>] alloc_vcpu+0x36/0x120
  [<ff1518c0>] smp_prepare_cpus+0x460/0xd30
  [<ff132fcd>] intel_p6_mcheck_init+0xbd/0xe0
  [<ff1555b4>] get_mtrr_state+0xf4/0x110
  [<ff1508f1>] __start_xen+0x5d1/0x9e0
  [<ff123270>] new_tlbflush_clock_period+0x0/0x90
  [<ff10015a>] start_paging+0x52/0x54
  CPU0 FATAL TRAP 6 (invalid opcode), ERROR_CODE 0000, IN INTERRUPT
  System shutting down--need manual reset.

As opposed to the vanilla linux kernel (that fails to see the local
apic), xen sees the local apic (incorrectly) and fails to bring up the
secondary CPUs and finally takes the FATAL TRAP as described in the
trace above.

If however, the bios setting is changed to Full APIC mode, both the
vanilla linux kernel and the xen based kernel boot correctly. I wanted
to know if others have seen similar problems. Also, is somebody looking
at addressing  these kinds of issues?


K. Y

Xen-devel mailing list

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