WARNING - OLD ARCHIVES

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/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-devel

[Xen-devel] pci-passthrough problem using vt-d

To: xen-devel@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-devel] pci-passthrough problem using vt-d
From: "news.gmane.org" <chris.ace@xxxxxxx>
Date: Tue, 27 Jul 2010 20:32:36 +0200
Delivery-date: Tue, 27 Jul 2010 11:34:06 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
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/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; de; rv:1.9.2.7) Gecko/20100713 Thunderbird/3.1.1
I'm trying to pass-through a PCI ISDN card (HFC-PCI) to a HVM domU using VTd. It's reported by lspci, but the driver (hisax) does not load. It recognizes the card but then fails to initialize it. I've seen several posts where passing this card to a PV domain using pcifront failed due to timing issues, but I thought I'd be on the safe side with VTd.

The guest is a Linux 2.4 image (where I cannot adjust the kernel unfortunately!), so my only option is HVM. But I also tried with a 6.0.1 Knoppix (2.6.28 kernel) as HVM domU where I can reproduce the issue (see attached screenshot from Linux 2.6 HVM domU).

Xen is current 4.1-unstable, dom0 was Linux-2.6.34-xen from Gentoo (based on Novell's forward ported patches to Xenlinux), but using the latest pvops (xen/next) produces the same results.

No logs in dom0 dmesg, but xm dmesg shows (not sure if this is related, but 0xc110 is the actual IOMEM of the PCI card in the domU):

(XEN) vlapic.c:684:d1 Local APIC Write to read-only register 0x30
(XEN) vlapic.c:684:d1 Local APIC Write to read-only register 0x20
(XEN) vlapic.c:684:d1 Local APIC Write to read-only register 0x20
(XEN) domctl.c:1061:d0 ioport_map:remove f_gport=c110 f_mport=ec00 np=8
(XEN) domctl.c:1037:d0 ioport_map:add f_gport=c110 f_mport=ec00 np=8
(XEN) irq.c:243: Dom1 PCI link 0 changed 5 -> 0
(XEN) irq.c:243: Dom1 PCI link 1 changed 10 -> 0
(XEN) irq.c:243: Dom1 PCI link 2 changed 11 -> 0
(XEN) irq.c:243: Dom1 PCI link 3 changed 5 -> 0
(XEN) multi.c:3470:d1 write to pagetable during event injection: cr2=0xf6785f88, mfn=0x1b7d85 (XEN) multi.c:3470:d1 write to pagetable during event injection: cr2=0xf6787f88, mfn=0x1b7d87
(XEN) domctl.c:1061:d0 ioport_map:remove f_gport=c110 f_mport=ec00 np=8
(XEN) domctl.c:991:d0 memory_map:remove: gfn=f2001 mfn=febff nr_mfns=1

(the last two lines are being printed during the failed driver load)

I guess that my 2.4 image vs. Knoppix with 2.6.28 are different enough to exclude problems on the domU end and the same applies for Gentoo-2.6.34-xen vs. pvops as dom0, so I think this is Xen-specific...

Any ideas?



Best regards,
   Christian


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel

<Prev in Thread] Current Thread [Next in Thread>
  • [Xen-devel] pci-passthrough problem using vt-d, news.gmane.org <=