Xen users - Feb 2009 - Trouble enabling VT-D: "I/O virtualisation disabled"

I have a chipset which supports VT-d (Intel X58) and, according to my
motherboard manufacturers, VT-d is enabled in the BIOS and listed as "Intel
VT." The problem is that xm dmesg shows "I/O virtualisation
disabled" and I
get this error message when I try to create a domU with a pci device
assigned: ''Error: failed to assign device: maybe the platform
doesn''t
support VT-d, or VT-d isn''t enabled properly?".

I''ve made sure the PCI device backend driver is compiled into the
kernel,
and dmesg shows pciback seizing my device:
pciback 0000:03:00.0: seizing device
pciback 0000:03:00.0: enabling device (0000 -> 0003)
 
I also tried it with several other PCI devices. All were seized, but I got
the same results when trying to launch a domU with any of them. I''ve
tried
this with both Xen 3.3 and Xen unstable, with the 2.6.26 Debian Xen kernel.
I can''t use an earlier kernel as they don''t support my sata
chipset.

Does anyone have any idea what could be causing these errors?
-- 
View this message in context:
http://www.nabble.com/Trouble-enabling-VT-D%3A-%22I-O-virtualisation-disabled%22-tp22173121p22173121.html
Sent from the Xen - User mailing list archive at Nabble.com.


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

I too have the same problems.   Using ASUS P6T and Debian Lenny.  I
played around with it for weeks and just could not get it to work,
here is my original post on the issue.

http://www.nabble.com/ASUS-p6t,-i7-920,--vt-d-broken.....-td21624614.html

I''ve given up for the moment......


On Tue, Feb 24, 2009 at 11:06 AM, andrewyates <andrewyates@gmail.com>
wrote:
>
> I have a chipset which supports VT-d (Intel X58) and, according to my
> motherboard manufacturers, VT-d is enabled in the BIOS and listed as
"Intel
> VT." The problem is that xm dmesg shows "I/O virtualisation
disabled" and I
> get this error message when I try to create a domU with a pci device
> assigned: ''Error: failed to assign device: maybe the platform
doesn''t
> support VT-d, or VT-d isn''t enabled properly?".
>
> I''ve made sure the PCI device backend driver is compiled into the
kernel,
> and dmesg shows pciback seizing my device:
> pciback 0000:03:00.0: seizing device
> pciback 0000:03:00.0: enabling device (0000 -> 0003)
>
> I also tried it with several other PCI devices. All were seized, but I got
> the same results when trying to launch a domU with any of them.
I''ve tried
> this with both Xen 3.3 and Xen unstable, with the 2.6.26 Debian Xen kernel.
> I can''t use an earlier kernel as they don''t support my
sata chipset.
>
> Does anyone have any idea what could be causing these errors?
> --
> View this message in context:
http://www.nabble.com/Trouble-enabling-VT-D%3A-%22I-O-virtualisation-disabled%22-tp22173121p22173121.html
> Sent from the Xen - User mailing list archive at Nabble.com.
>
>
> _______________________________________________
> Xen-users mailing list
> Xen-users@lists.xensource.com
> http://lists.xensource.com/xen-users
>


-- 
Steve Reaver
http://www.stevereaver.com
Electronic Music Production

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

Did you compile Xen with pci utils?

-Alex

On Mon, Feb 23, 2009 at 7:06 PM, andrewyates <andrewyates@gmail.com>
wrote:
>
> I have a chipset which supports VT-d (Intel X58) and, according to my
> motherboard manufacturers, VT-d is enabled in the BIOS and listed as
"Intel
> VT." The problem is that xm dmesg shows "I/O virtualisation
disabled" and I
> get this error message when I try to create a domU with a pci device
> assigned: ''Error: failed to assign device: maybe the platform
doesn''t
> support VT-d, or VT-d isn''t enabled properly?".
>
> I''ve made sure the PCI device backend driver is compiled into the
kernel,
> and dmesg shows pciback seizing my device:
> pciback 0000:03:00.0: seizing device
> pciback 0000:03:00.0: enabling device (0000 -> 0003)
>
> I also tried it with several other PCI devices. All were seized, but I got
> the same results when trying to launch a domU with any of them.
I''ve tried
> this with both Xen 3.3 and Xen unstable, with the 2.6.26 Debian Xen kernel.
> I can''t use an earlier kernel as they don''t support my
sata chipset.
>
> Does anyone have any idea what could be causing these errors?
> --
> View this message in context:
http://www.nabble.com/Trouble-enabling-VT-D%3A-%22I-O-virtualisation-disabled%22-tp22173121p22173121.html
> Sent from the Xen - User mailing list archive at Nabble.com.
>
>
> _______________________________________________
> Xen-users mailing list
> Xen-users@lists.xensource.com
> http://lists.xensource.com/xen-users
>
_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

Well given that I do not know exactly what you mean, I most probably
did not. PCI Passthrough seems to be enabled however because I can see
it seizing the device, what more should I have done ?

Andrew - Sorry for hijacking your thread but it seems that we may have
the same problem.


Steve

On Tue, Feb 24, 2009 at 12:34 PM, Alexia Benington
<alexbenington@gmail.com> wrote:
> Did you compile Xen with pci utils?
>
> -Alex
>
_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

On Fedora, make sure the package pciutils-devel is already installed
before compiling Xen. I think the package is pciutils-dev for Debian.

-Alex

On Mon, Feb 23, 2009 at 9:30 PM, Steve Reaver <stevereaver@gmail.com>
wrote:
> Well given that I do not know exactly what you mean, I most probably
> did not. PCI Passthrough seems to be enabled however because I can see
> it seizing the device, what more should I have done ?
>
> Andrew - Sorry for hijacking your thread but it seems that we may have
> the same problem.
>
>
> Steve
>
> On Tue, Feb 24, 2009 at 12:34 PM, Alexia Benington
> <alexbenington@gmail.com> wrote:
>> Did you compile Xen with pci utils?
>>
>> -Alex
>>
>
_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users

Alexia Benington wrote:
> 
> Did you compile Xen with pci utils?
> 
Yes. I don''t see a way to verify that it was found, but
/usr/include/pci/
was present when I compiled it.

-- 
View this message in context:
http://www.nabble.com/Trouble-enabling-VT-D%3A-%22I-O-virtualisation-disabled%22-tp22173121p22175708.html
Sent from the Xen - User mailing list archive at Nabble.com.


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

andrewyates schrieb:
> I have a chipset which supports VT-d (Intel X58) and, according to my
> motherboard manufacturers, VT-d is enabled in the BIOS and listed as
"Intel
> VT." 
VT is not the same as VT-d, so if you enable VT you will essentially 
have the "vmx" feature of the CPU enabled which will allow hvm, but
that
has nothing to do with VT-d.
What board do you own?


Best regards,
    Christian


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

> I too have the same problems.   Using ASUS P6T and Debian Lenny.  I
> played around with it for weeks and just could not get it to work,
> here is my original post on the issue.
> 
> http://www.nabble.com/ASUS-p6t,-i7-920,--vt-d-broken.....-td21624614.html
> 
> I''ve given up for the moment......
Hi Steve,

I''m also having a problem with the Asus P6T, but since I''m
using
Xen-unstable I''m not sure if this is the same problem than you have.
Have a look at the "VT-D RMRR is incorrect" thread initially started
by
Yoshiharo Mori:
It looks like the ACPI-DMAR tables are broken on some (Supermicro, Asus) 
X58 boards.
This is a problem for Xen-unstable, disabling VT-d.

I already opened a case with Asus but they are telling me that Linux is 
not officially supported. I''ll continue to chase them as VT-d is (at 
least in theory) platform independent but it might help if others also 
open cases about this!


Unfortunately I can''t go back to Xen-stable for now to test if this is 
the same issue as I''m running on a paravirt_ops kernel that requires 
Xen-unstable (or at least heavy patching for 3.3).


Btw, just looked at your other post again, the iommu=1 needs to go on 
the xen line and the pciback stuff on the line together with the linux 
kernel... and again, VMX (aka VT) has nothing to do with VT-d...


Best regards,
    Christian


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

Christian Tramnitz wrote:
> 
> VT is not the same as VT-d, so if you enable VT you will essentially 
> have the "vmx" feature of the CPU enabled which will allow hvm,
but that
> has nothing to do with VT-d.
> What board do you own?
> 
It''s a MSI X58 Platinum. The only virtualization option shown in the
BIOS is
"Intel VT", but MSI support says that includes VT-d. I''m not
sure of a way
to tell if it''s actually enabled there or not.

-- 
View this message in context:
http://www.nabble.com/Trouble-enabling-VT-D%3A-%22I-O-virtualisation-disabled%22-tp22173121p22185384.html
Sent from the Xen - User mailing list archive at Nabble.com.


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