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-users] Problem with interrupts/APIC in xen-unstable

To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-users] Problem with interrupts/APIC in xen-unstable
From: Tomas Kopal <Tomas.Kopal@xxxxxxxx>
Date: Wed, 16 Nov 2005 13:11:35 +0100
Delivery-date: Wed, 16 Nov 2005 16:33:31 +0000
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-users-request@lists.xensource.com?subject=help>
List-id: Xen user discussion <xen-users.lists.xensource.com>
List-post: <mailto:xen-users@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mozilla Thunderbird 1.0.7 (Windows/20050923)
I have got a problem with network in clean dom0, just after boot. The
card does not receive even DHCP response. After looking into archives, I
started playing with ACPI/APIC and found that using "noapic" flag for
xen hypervisor in grub helps - the network then starts and works.

My setup:
xen-unstable, snapshot 14/11/05
xen hypervisor built from source
linux kernel 2.6.12 built from source with xen patches
all running in VMware workstation 5.0.0 build 13124
Ethernet card pcnet (PCnet/PCI II 79C970A) - simulated by VMware
xen flags (net not working): dom0_mem=98304
xen flags (net working): dom0_mem=98304 noapic
kernel flags: root=/dev/md1

This setup works correctly even with APIC enabled on xen-2.0.7.

The ethtool trick to turn off tx/rx checksums, also suggested earlier,
does not work as the driver reports "Operation not supported".

I can't rule out that the problem is caused by running within VMware,
but as similar reports appeared earlier without using VMware, I would
guess it is not.

This may be connected to bug #76
(http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=76), but the
xen-unstable sources I have have the patch mentioned already included.
Also, I have the problem immediately after boot, not after some time, as
reported in the bug report.

The network card is on IRQ 9, shared with ACPI - again corresponds to
the bug report and may be reason for the problem to be immediate. The
interrupt table looks like this:
  1:       606      Phys-irq    i8042
  9:         0      Phys-irq    acpi, eth0
 14:       959      Phys-irq    ide0
 15:      1867      Phys-irq    ide1
256:      2099   Dynamic-irq    timer0
257:     38696   Dynamic-irq    xenbus
258:         0   Dynamic-irq    console
259:         0   Dynamic-irq    net-be-dbg
NMI:         0
ERR:         0

There are message during boot related to this:
ACPI: PCI Interrupt Link [LNKB] enabled at IRQ 9
PCI: setting IRQ 9 as level-triggered
ACPI: PCI Interrupt 0000:00:10.0[A] -> Link [LNKB] -> GSI 9 (level, low)
-> IRQ 9
(so the IRQ seems to be level triggered after all)

This is my testing setup, so if you want me to investigate further to
find the culprit, let me know what do you want me to do.



Xen-users mailing list

<Prev in Thread] Current Thread [Next in Thread>
  • [Xen-users] Problem with interrupts/APIC in xen-unstable, Tomas Kopal <=