Mailing List Archive

Support for vTPM in Xen-API
Hello!

For the Xen-API call tomorrow:

Here are modifications for better supporting the vTPM in the Xen API. I
took the classes for the networking (VIF) and block devices (VBD) as
examples and modeled the vTPM class similar to them. The result gives a
better approach to managing this type of device.

I thought about posting a patch for the latex source file for the
Xen-API document but had the impression that the contents of some
subsections were generated using a tool. So here are just modifications to
tables so far.

Comments are welcome.




2.6 Class : VM
2.6.1 Fields for class: VM

Name VM
Description a virtual machine (or 'guest')


Quals Field Type Description
[...]
RW VTPM (VTPM ref) Set reference to a set of
VTPMs
[...]

--> Remove TPM/instance and TPM/backend from current VM class.



NEW CLASS:

2.1x Class: VTPM
2.1x.1 Fields for class: VTPM

Name VTPM
Description: a virtual TPM device

Quals Field Type Description

RO_run uuid string unique identifier /
object reference
RW name string human-readable name
of the interface
RO_ins driver driver_type the style of driver
RO_ins VM VM ref the virtual machine
RO_ins backend VM ref the domain where the
backend is located
RO_ins instance int the instance number
of the virtual TPM the domain is associated with
RO_ins TPM TPM ref reference to a TPM
object representing the state of the device, if it can be
determined; may be
NIL

NEW CLASS:

2.1x+1 Class: TPM
2.1x+1.1 Fields for class: TPM

Name TPM
Description: a TPM device?s current state; the device may be a
hardware TPM or a virtual TPM whose state can be determined

Quals Field Type Description

RO_run uuid string unique
identifier / object reference
RO_run active boolean whether the
device is active
RO_run temp_deactivated boolean whether the
device has temporarily
been
deactivated
RO_run has_owner boolean whether the
device has an owner


Additional methods:

TakeOwnership(owner-pwd, srk-pwd) : Take ownership of the TPM device with
the given passwords for the owner and the storage root key.



Regards,
Stefan
Re: Support for vTPM in Xen-API [ In reply to ]
On Tue, Sep 12, 2006 at 08:53:12PM -0400, Stefan Berger wrote:

> Hello!
>
> For the Xen-API call tomorrow:
>
> Here are modifications for better supporting the vTPM in the Xen API. I
> took the classes for the networking (VIF) and block devices (VBD) as
> examples and modeled the vTPM class similar to them. The result gives a
> better approach to managing this type of device.
>
> I thought about posting a patch for the latex source file for the
> Xen-API document but had the impression that the contents of some
> subsections were generated using a tool. So here are just modifications to
> tables so far.

Stefan, thanks for all of that. I'll certainly include it in the next
rev. No, there's no tool involved (though if I'd realised how much work
it would have been when I started, I'd have probably written one!)
Anyway, if you've got a patch for the latex, that would be great -- the
source files are all in a tarball on the wiki.

Cheers,

Ewan.

_______________________________________________
xen-api mailing list
xen-api@lists.xensource.com
http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-api
Re: Support for vTPM in Xen-API [ In reply to ]
Ewan Mellor <ewan@xensource.com> wrote on 09/14/2006 01:47:27 AM:

> On Tue, Sep 12, 2006 at 08:53:12PM -0400, Stefan Berger wrote:
>
> > Hello!
> >
> > For the Xen-API call tomorrow:
> >
> > Here are modifications for better supporting the vTPM in the Xen
API. I
> > took the classes for the networking (VIF) and block devices (VBD) as
> > examples and modeled the vTPM class similar to them. The result gives
a
> > better approach to managing this type of device.
> >
> > I thought about posting a patch for the latex source file for the
> > Xen-API document but had the impression that the contents of some
> > subsections were generated using a tool. So here are just
modifications to
> > tables so far.
>
> Stefan, thanks for all of that. I'll certainly include it in the next
> rev. No, there's no tool involved (though if I'd realised how much work
> it would have been when I started, I'd have probably written one!)
> Anyway, if you've got a patch for the latex, that would be great -- the
> source files are all in a tarball on the wiki.

I will send you a diff against the latex then. It won't be the last
vTPM-related change.

Stefan
>
> Cheers,
>
> Ewan.