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-bugs] [Bug 1687] New: Boot hang because of SIS ACPI bug workaround

To: xen-bugs@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-bugs] [Bug 1687] New: Boot hang because of SIS ACPI bug workaround
From: bugzilla-daemon@xxxxxxxxxxxxxxxxxxx
Date: Fri, 12 Nov 2010 04:31:35 -0800
Delivery-date: Fri, 12 Nov 2010 04:31:41 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-bugs-request@lists.xensource.com?subject=help>
List-id: Xen Bugzilla <xen-bugs.lists.xensource.com>
List-post: <mailto:xen-bugs@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-bugs>, <mailto:xen-bugs-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-bugs>, <mailto:xen-bugs-request@lists.xensource.com?subject=unsubscribe>
Reply-to: bugs@xxxxxxxxxxxxxxxxxx
Sender: xen-bugs-bounces@xxxxxxxxxxxxxxxxxxx

           Summary: Boot hang because of SIS ACPI bug workaround
           Product: Xen
           Version: unstable
          Platform: x86-64
        OS/Version: Linux-2.6
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Linux-pv_ops dom0
        AssignedTo: xen-bugs@xxxxxxxxxxxxxxxxxxx
        ReportedBy: agrenott@xxxxxxxxx

Trying to switch from xen 3.4.2/2.6.18 to 4.0.1/2.6.32, boot hangs here:

calling  sctp_init+0x0/0x7a8 @ 1
SCTP: Hash tables configured (established 4096 bind 4096)
initcall sctp_init+0x0/0x7a8 returned 0 after 342 usecs
calling  update_mp_table+0x0/0x6be @ 1
initcall update_mp_table+0x0/0x6be returned 0 after 1 usecs
calling  io_apic_bug_finalize+0x0/0x66 @ 1

It's a server in a datacenter where I don't have serial console access, so I
can't give more output (I got these via netconsole module).
However, I've been able to track the issue down to function
io_apic_bug_finalize in arch/x86/kernel/apic/io_apic-xen.c : crash happens when
calling VOID(HYPERVISOR_platform_op(&op));
In my case, it tries to set QUIRK_IOAPIC_BAD_REGSEL.

However, looking at xen 4.0.1 hypervisor source
(xen/arch/x86/platform_hypercall.c), I have:
#ifndef sis_apic_bug
            sis_apic_bug = (quirk_id == QUIRK_IOAPIC_BAD_REGSEL);
            dprintk(XENLOG_INFO, "Domain 0 says that IO-APIC REGSEL is %s\n",
                    sis_apic_bug ? "bad" : "good");
            BUG_ON(sis_apic_bug != (quirk_id == QUIRK_IOAPIC_BAD_REGSEL));

In my case, sis_apic_bug is defined (meaning work-around is not implemented I
guess), and thus I suppose system hangs because of the BUG_ON.

So it seems to me there's some incompatibility between the dom0 trying to
activate the workaround and the hypervisor refusing it. And indeed, removing
the HYPERVISOR_platform_op call from dom0 allows me to boot.

Please tell me if I need to provide more data.

Configure bugmail: 
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

Xen-bugs mailing list

<Prev in Thread] Current Thread [Next in Thread>
  • [Xen-bugs] [Bug 1687] New: Boot hang because of SIS ACPI bug workaround, bugzilla-daemon <=