Re: Is xen planning to support hvm guests booting under UEFI?

Re: Is xen planning to support hvm guests booting under UEFI?

[jim burns]

On Sat September 24 2011, 6:43:42 PM, Todd Deshane wrote:
> On Sat, Sep 24, 2011 at 5:34 PM, jim burns <jim_burn@bellsouth.net> wrote:
> > On Wed September 21 2011, 9:06:38 PM, jim burns wrote:
> >> Pls cc: me as I am not subscribed.
> > 
> > [...]
> > http://wiki.xensource.com/xenwiki/XenParavirtOps has been reworked
> > recently. Not sure if this was there before:
> > 
> > "Add EFI support in Linux pv-ops. Jan Beulich wrote patches to make Xen
> > hypervisor be able to compile as an EFI application. "Build xen.efi,
> > write up a config file for it to read (most importantly so it knows
> > what Dom0 kernel and initrd to use), and you should be good to go
> > (provided the EFI implementation isn't too flawed). This being an EFI
> > application you can simply run it from the shell prompt. Parameter for
> > Xen.efi is -cfg <file>, and <file> has: kernel=, ramdisk=, options=,
> > video=gfx-x. The Linux pv-ops needs at least to parse
> > XEN_VGATYPE_EFI_LFB data, E820 parsed (should be working) and give the
> > ACPI subsystem a pointer to the ACPI DSDT without consulting EBDA."
> > 
> > This doesn't solve the Windows 8 hvm problem, but at least EFI is being
> > considered.
> 
> There was a google summer of code project around UEFI:
> http://code.google.com/p/google-summer-of-code-2011-tianocore/downloads/deta
> il?name=Bei_Guan.tar.gz&can=2&q=
> 
> I don't know if it was submitted to xen-devel, but you can follow up
> with the xen-devel mailing list if you are interested.
> 
> Hope that helps.

The article I quoted was from http://www.itworld.com/it-
managementstrategy/205255/windows-8-oem-specs-
may-block-linux-booting :

"Red Hat's Matthew Garrett was one of the first to notice that according to 
the new logo rules, all Windows 8 machines will need to be have the Unified 
Extensible Firmware Interface (UEFI) instead of the venerable BIOS firmware 
layer. BIOS has been pretty much the sole firmware interface for PCs for a 
long time.

The EFI system has slowly been making headway in recent years, and right now 
EFI firmware is compatible with Windows supporting the GUID Partition Table 
(GPT), OS X/Intel, and Linux 2.6 and beyond machines. EFI is seen as a better 
hardware/software interface than BIOS, since it is platform-agnostic, runs in 
32- or 64-bit mode, and GPT machines can handle boot partitions of up to 9.4 
zettabytes. (That's 9.5 billion terabytes to you and me.)

EFI, and the later UEFI specification, is not the problem for Linux. The 
problem is Microsoft's other requirement for any Windows 8-certified client: 
the system must support secure booting. This hardened boot means that "all 
firmware and software in the boot process must be signed by a trusted 
Certificate Authority (CA)," according to slides from a recent presentation on 
the UEFI boot process made by Arie van der Hoeven, Microsoft Principal Lead 
Program Manager."

Putting aside the signing problem for now, the immediate problem is qemu-dm as 
the emulator for hvm guests, since it provides an emulated *bios*. Or 
hvmloader as the bootloader. The url Todd provides points to a project to 
modify 'ovmf', an EFI bootloader, to be xen hvm aware. Windows 8 will be a 
game changer, if the article is right. Hope plans are being made to 
incorporate some sort of EFI bootloader into xen.

RE: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Paul Durrant]

Since I was at the presentation in question, I asked about requirements for virtual platforms. I seem to recall the response being that UEFI was not a requirement for virtual platforms as yet and moreover I don't believe hyper-v VMs UEFI boot. 
That said I'm not sure about the requirements surrounding trusted boot; I guess they will become clearer as the HCK requirements are firmed up.

I'd certainly like to see HVM UEFI boot sooner rather than later as I think it will speed up Windows boot (and not just for Windows 8) quite a lot but I didn't see any need to panic just yet.

  Paul 

> -----Original Message-----
> From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-
> bounces@lists.xensource.com] On Behalf Of jim burns
> Sent: 25 September 2011 03:18
> To: xen-users@lists.xensource.com; xen-devel@lists.xensource.com
> Cc: Todd Deshane
> Subject: [Xen-devel] Re: [Xen-users] Is xen planning to support hvm
> guests booting under UEFI?
> 
> On Sat September 24 2011, 6:43:42 PM, Todd Deshane wrote:
> > On Sat, Sep 24, 2011 at 5:34 PM, jim burns
> <jim_burn@bellsouth.net> wrote:
> > > On Wed September 21 2011, 9:06:38 PM, jim burns wrote:
> > >> Pls cc: me as I am not subscribed.
> > >
> > > [...]
> > > http://wiki.xensource.com/xenwiki/XenParavirtOps has been
> reworked
> > > recently. Not sure if this was there before:
> > >
> > > "Add EFI support in Linux pv-ops. Jan Beulich wrote patches to
> make
> > > Xen hypervisor be able to compile as an EFI application. "Build
> > > xen.efi, write up a config file for it to read (most importantly
> so
> > > it knows what Dom0 kernel and initrd to use), and you should be
> good
> > > to go (provided the EFI implementation isn't too flawed). This
> being
> > > an EFI application you can simply run it from the shell prompt.
> > > Parameter for Xen.efi is -cfg <file>, and <file> has: kernel=,
> > > ramdisk=, options=, video=gfx-x. The Linux pv-ops needs at least
> to
> > > parse XEN_VGATYPE_EFI_LFB data, E820 parsed (should be working)
> and
> > > give the ACPI subsystem a pointer to the ACPI DSDT without
> consulting EBDA."
> > >
> > > This doesn't solve the Windows 8 hvm problem, but at least EFI
> is
> > > being considered.
> >
> > There was a google summer of code project around UEFI:
> > http://code.google.com/p/google-summer-of-code-2011-
> tianocore/download
> > s/deta
> > il?name=Bei_Guan.tar.gz&can=2&q=
> >
> > I don't know if it was submitted to xen-devel, but you can follow
> up
> > with the xen-devel mailing list if you are interested.
> >
> > Hope that helps.
> 
> The article I quoted was from http://www.itworld.com/it-
> managementstrategy/205255/windows-8-oem-specs-
> may-block-linux-booting :
> 
> "Red Hat's Matthew Garrett was one of the first to notice that
> according to the new logo rules, all Windows 8 machines will need to
> be have the Unified Extensible Firmware Interface (UEFI) instead of
> the venerable BIOS firmware layer. BIOS has been pretty much the
> sole firmware interface for PCs for a long time.
> 
> The EFI system has slowly been making headway in recent years, and
> right now EFI firmware is compatible with Windows supporting the
> GUID Partition Table (GPT), OS X/Intel, and Linux 2.6 and beyond
> machines. EFI is seen as a better hardware/software interface than
> BIOS, since it is platform-agnostic, runs in
> 32- or 64-bit mode, and GPT machines can handle boot partitions of
> up to 9.4 zettabytes. (That's 9.5 billion terabytes to you and me.)
> 
> EFI, and the later UEFI specification, is not the problem for Linux.
> The problem is Microsoft's other requirement for any Windows 8-
> certified client:
> the system must support secure booting. This hardened boot means
> that "all firmware and software in the boot process must be signed
> by a trusted Certificate Authority (CA)," according to slides from a
> recent presentation on the UEFI boot process made by Arie van der
> Hoeven, Microsoft Principal Lead Program Manager."
> 
> Putting aside the signing problem for now, the immediate problem is
> qemu-dm as the emulator for hvm guests, since it provides an
> emulated *bios*. Or hvmloader as the bootloader. The url Todd
> provides points to a project to modify 'ovmf', an EFI bootloader, to
> be xen hvm aware. Windows 8 will be a game changer, if the article
> is right. Hope plans are being made to incorporate some sort of EFI
> bootloader into xen.
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Keir Fraser]

The GSOC project student seemed to be making good progress, and there were
others on the OVMF side who were giving him support. I haven't heard
anything in about a month now, though.


On 25/09/2011 07:12, "Paul Durrant" <Paul.Durrant@citrix.com> wrote:

> Since I was at the presentation in question, I asked about requirements for
> virtual platforms. I seem to recall the response being that UEFI was not a
> requirement for virtual platforms as yet and moreover I don't believe hyper-v
> VMs UEFI boot. 
> That said I'm not sure about the requirements surrounding trusted boot; I
> guess they will become clearer as the HCK requirements are firmed up.
> 
> I'd certainly like to see HVM UEFI boot sooner rather than later as I think it
> will speed up Windows boot (and not just for Windows 8) quite a lot but I
> didn't see any need to panic just yet.
> 
>   Paul 
> 
>> -----Original Message-----
>> From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-
>> bounces@lists.xensource.com] On Behalf Of jim burns
>> Sent: 25 September 2011 03:18
>> To: xen-users@lists.xensource.com; xen-devel@lists.xensource.com
>> Cc: Todd Deshane
>> Subject: [Xen-devel] Re: [Xen-users] Is xen planning to support hvm
>> guests booting under UEFI?
>> 
>> On Sat September 24 2011, 6:43:42 PM, Todd Deshane wrote:
>>> On Sat, Sep 24, 2011 at 5:34 PM, jim burns
>> <jim_burn@bellsouth.net> wrote:
>>>> On Wed September 21 2011, 9:06:38 PM, jim burns wrote:
>>>>> Pls cc: me as I am not subscribed.
>>>> 
>>>> [...]
>>>> http://wiki.xensource.com/xenwiki/XenParavirtOps has been
>> reworked
>>>> recently. Not sure if this was there before:
>>>> 
>>>> "Add EFI support in Linux pv-ops. Jan Beulich wrote patches to
>> make
>>>> Xen hypervisor be able to compile as an EFI application. "Build
>>>> xen.efi, write up a config file for it to read (most importantly
>> so
>>>> it knows what Dom0 kernel and initrd to use), and you should be
>> good
>>>> to go (provided the EFI implementation isn't too flawed). This
>> being
>>>> an EFI application you can simply run it from the shell prompt.
>>>> Parameter for Xen.efi is -cfg <file>, and <file> has: kernel=,
>>>> ramdisk=, options=, video=gfx-x. The Linux pv-ops needs at least
>> to
>>>> parse XEN_VGATYPE_EFI_LFB data, E820 parsed (should be working)
>> and
>>>> give the ACPI subsystem a pointer to the ACPI DSDT without
>> consulting EBDA."
>>>> 
>>>> This doesn't solve the Windows 8 hvm problem, but at least EFI
>> is
>>>> being considered.
>>> 
>>> There was a google summer of code project around UEFI:
>>> http://code.google.com/p/google-summer-of-code-2011-
>> tianocore/download
>>> s/deta
>>> il?name=Bei_Guan.tar.gz&can=2&q=
>>> 
>>> I don't know if it was submitted to xen-devel, but you can follow
>> up
>>> with the xen-devel mailing list if you are interested.
>>> 
>>> Hope that helps.
>> 
>> The article I quoted was from http://www.itworld.com/it-
>> managementstrategy/205255/windows-8-oem-specs-
>> may-block-linux-booting :
>> 
>> "Red Hat's Matthew Garrett was one of the first to notice that
>> according to the new logo rules, all Windows 8 machines will need to
>> be have the Unified Extensible Firmware Interface (UEFI) instead of
>> the venerable BIOS firmware layer. BIOS has been pretty much the
>> sole firmware interface for PCs for a long time.
>> 
>> The EFI system has slowly been making headway in recent years, and
>> right now EFI firmware is compatible with Windows supporting the
>> GUID Partition Table (GPT), OS X/Intel, and Linux 2.6 and beyond
>> machines. EFI is seen as a better hardware/software interface than
>> BIOS, since it is platform-agnostic, runs in
>> 32- or 64-bit mode, and GPT machines can handle boot partitions of
>> up to 9.4 zettabytes. (That's 9.5 billion terabytes to you and me.)
>> 
>> EFI, and the later UEFI specification, is not the problem for Linux.
>> The problem is Microsoft's other requirement for any Windows 8-
>> certified client:
>> the system must support secure booting. This hardened boot means
>> that "all firmware and software in the boot process must be signed
>> by a trusted Certificate Authority (CA)," according to slides from a
>> recent presentation on the UEFI boot process made by Arie van der
>> Hoeven, Microsoft Principal Lead Program Manager."
>> 
>> Putting aside the signing problem for now, the immediate problem is
>> qemu-dm as the emulator for hvm guests, since it provides an
>> emulated *bios*. Or hvmloader as the bootloader. The url Todd
>> provides points to a project to modify 'ovmf', an EFI bootloader, to
>> be xen hvm aware. Windows 8 will be a game changer, if the article
>> is right. Hope plans are being made to incorporate some sort of EFI
>> bootloader into xen.
>> 
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xensource.com
>> http://lists.xensource.com/xen-devel
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Ian Campbell]

Adding Bei Guan and Andrey Warkentin who I think were student and mentor
respectively. Hi guys!

On Sun, 2011-09-25 at 15:50 +0100, Keir Fraser wrote:
> The GSOC project student seemed to be making good progress, and there were
> others on the OVMF side who were giving him support. I haven't heard
> anything in about a month now, though.

It would be very interesting to get an update (or a pointer to one) of
the status/result of this particular project now that the GSoC is over.

Thanks,
Ian.


> 
> 
> On 25/09/2011 07:12, "Paul Durrant" <Paul.Durrant@citrix.com> wrote:
> 
> > Since I was at the presentation in question, I asked about requirements for
> > virtual platforms. I seem to recall the response being that UEFI was not a
> > requirement for virtual platforms as yet and moreover I don't believe hyper-v
> > VMs UEFI boot. 
> > That said I'm not sure about the requirements surrounding trusted boot; I
> > guess they will become clearer as the HCK requirements are firmed up.
> > 
> > I'd certainly like to see HVM UEFI boot sooner rather than later as I think it
> > will speed up Windows boot (and not just for Windows 8) quite a lot but I
> > didn't see any need to panic just yet.
> > 
> >   Paul 
> > 
> >> -----Original Message-----
> >> From: xen-devel-bounces@lists.xensource.com [mailto:xen-devel-
> >> bounces@lists.xensource.com] On Behalf Of jim burns
> >> Sent: 25 September 2011 03:18
> >> To: xen-users@lists.xensource.com; xen-devel@lists.xensource.com
> >> Cc: Todd Deshane
> >> Subject: [Xen-devel] Re: [Xen-users] Is xen planning to support hvm
> >> guests booting under UEFI?
> >> 
> >> On Sat September 24 2011, 6:43:42 PM, Todd Deshane wrote:
> >>> On Sat, Sep 24, 2011 at 5:34 PM, jim burns
> >> <jim_burn@bellsouth.net> wrote:
> >>>> On Wed September 21 2011, 9:06:38 PM, jim burns wrote:
> >>>>> Pls cc: me as I am not subscribed.
> >>>> 
> >>>> [...]
> >>>> http://wiki.xensource.com/xenwiki/XenParavirtOps has been
> >> reworked
> >>>> recently. Not sure if this was there before:
> >>>> 
> >>>> "Add EFI support in Linux pv-ops. Jan Beulich wrote patches to
> >> make
> >>>> Xen hypervisor be able to compile as an EFI application. "Build
> >>>> xen.efi, write up a config file for it to read (most importantly
> >> so
> >>>> it knows what Dom0 kernel and initrd to use), and you should be
> >> good
> >>>> to go (provided the EFI implementation isn't too flawed). This
> >> being
> >>>> an EFI application you can simply run it from the shell prompt.
> >>>> Parameter for Xen.efi is -cfg <file>, and <file> has: kernel=,
> >>>> ramdisk=, options=, video=gfx-x. The Linux pv-ops needs at least
> >> to
> >>>> parse XEN_VGATYPE_EFI_LFB data, E820 parsed (should be working)
> >> and
> >>>> give the ACPI subsystem a pointer to the ACPI DSDT without
> >> consulting EBDA."
> >>>> 
> >>>> This doesn't solve the Windows 8 hvm problem, but at least EFI
> >> is
> >>>> being considered.
> >>> 
> >>> There was a google summer of code project around UEFI:
> >>> http://code.google.com/p/google-summer-of-code-2011-
> >> tianocore/download
> >>> s/deta
> >>> il?name=Bei_Guan.tar.gz&can=2&q=
> >>> 
> >>> I don't know if it was submitted to xen-devel, but you can follow
> >> up
> >>> with the xen-devel mailing list if you are interested.
> >>> 
> >>> Hope that helps.
> >> 
> >> The article I quoted was from http://www.itworld.com/it-
> >> managementstrategy/205255/windows-8-oem-specs-
> >> may-block-linux-booting :
> >> 
> >> "Red Hat's Matthew Garrett was one of the first to notice that
> >> according to the new logo rules, all Windows 8 machines will need to
> >> be have the Unified Extensible Firmware Interface (UEFI) instead of
> >> the venerable BIOS firmware layer. BIOS has been pretty much the
> >> sole firmware interface for PCs for a long time.
> >> 
> >> The EFI system has slowly been making headway in recent years, and
> >> right now EFI firmware is compatible with Windows supporting the
> >> GUID Partition Table (GPT), OS X/Intel, and Linux 2.6 and beyond
> >> machines. EFI is seen as a better hardware/software interface than
> >> BIOS, since it is platform-agnostic, runs in
> >> 32- or 64-bit mode, and GPT machines can handle boot partitions of
> >> up to 9.4 zettabytes. (That's 9.5 billion terabytes to you and me.)
> >> 
> >> EFI, and the later UEFI specification, is not the problem for Linux.
> >> The problem is Microsoft's other requirement for any Windows 8-
> >> certified client:
> >> the system must support secure booting. This hardened boot means
> >> that "all firmware and software in the boot process must be signed
> >> by a trusted Certificate Authority (CA)," according to slides from a
> >> recent presentation on the UEFI boot process made by Arie van der
> >> Hoeven, Microsoft Principal Lead Program Manager."
> >> 
> >> Putting aside the signing problem for now, the immediate problem is
> >> qemu-dm as the emulator for hvm guests, since it provides an
> >> emulated *bios*. Or hvmloader as the bootloader. The url Todd
> >> provides points to a project to modify 'ovmf', an EFI bootloader, to
> >> be xen hvm aware. Windows 8 will be a game changer, if the article
> >> is right. Hope plans are being made to incorporate some sort of EFI
> >> bootloader into xen.
> >> 
> >> _______________________________________________
> >> Xen-devel mailing list
> >> Xen-devel@lists.xensource.com
> >> http://lists.xensource.com/xen-devel
> > 
> > _______________________________________________
> > Xen-devel mailing list
> > Xen-devel@lists.xensource.com
> > http://lists.xensource.com/xen-devel
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: [Xen-devel] Re: Is xen planning to support hvm guests booting under UEFI?

[Andrei E. Warkentin]

Hi Ian, Keir, others -

I was the mentor for the project. We got to the point where OVMF on
Xen was indistinguishable from OVMF on QEMU as far
as emulated device support (Bei can confirm - is this true for
network/PXE boot as well?). ACPI and DMI info published from
the hvmloader is consumed by the firmware and published. I believe we
could boot RedHat. Last I checked Win7 was unbootable under QEMU with
OVMF (so we didn't check) and I don't know the status of that (if I
remember correctly, it BSODs in bootvid.sys (under QEMU) due to low
memory area not being reserved away, or because there is no fake IDT
set up - that seems to match up with the hoops I needed to do doing
similar work on Hyper-V).

OVMF is aware of Xen, and work was being done on a PV block device
(Bei got as far as hypercall support and xenstore). Bei, are you
working on the PV block driver? I want to say we're pretty close on
that one (PV block not very hard, and neither is the block device
abstraction for EFI).

I/Bei had patches against hvmloader (to support booting OVMF) and Bei
had patches against the management stack as well to expose the
functionality of picking OVMF as the firmware. I believe all of these
were sent to xen development list.

Jordan Justen (Intel) had been working on integrating the OVMF changes
last I checked, but I am not positive (I just changed jobs and
relocated, and still am not on the TianoCore mailing list). I am
adding him to the reply.

A


2011/9/26 Ian Campbell <Ian.Campbell@citrix.com>:
> Adding Bei Guan and Andrey Warkentin who I think were student and mentor
> respectively. Hi guys!
>
> On Sun, 2011-09-25 at 15:50 +0100, Keir Fraser wrote:
>> The GSOC project student seemed to be making good progress, and there were
>> others on the OVMF side who were giving him support. I haven't heard
>> anything in about a month now, though.
>
> It would be very interesting to get an update (or a pointer to one) of
> the status/result of this particular project now that the GSoC is over.
>
> Thanks,
> Ian.

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Ian Campbell]

On Mon, 2011-09-26 at 20:18 +0100, Andrei E. Warkentin wrote:
> Hi Ian, Keir, others -
> 
> I was the mentor for the project. We got to the point where OVMF on
> Xen was indistinguishable from OVMF on QEMU as far
> as emulated device support (Bei can confirm - is this true for
> network/PXE boot as well?). ACPI and DMI info published from
> the hvmloader is consumed by the firmware and published. I believe we
> could boot RedHat. Last I checked Win7 was unbootable under QEMU with
> OVMF (so we didn't check) and I don't know the status of that (if I
> remember correctly, it BSODs in bootvid.sys (under QEMU) due to low
> memory area not being reserved away, or because there is no fake IDT
> set up - that seems to match up with the hoops I needed to do doing
> similar work on Hyper-V).
> 
> OVMF is aware of Xen, and work was being done on a PV block device
> (Bei got as far as hypercall support and xenstore). Bei, are you
> working on the PV block driver? I want to say we're pretty close on
> that one (PV block not very hard, and neither is the block device
> abstraction for EFI).

Thanks for the update, it sounds like it's pretty far along.

The PV enhancements etc sound great but it would be nice if we could tie
up the baseline stuff so that folks can start using it.

Is Bei still working on the project since the end of GSoC? If not is
there an archive of what he accomplished so that we can find someone to
pick up on it if necessary?

> I/Bei had patches against hvmloader (to support booting OVMF) and Bei
> had patches against the management stack as well to expose the
> functionality of picking OVMF as the firmware. I believe all of these
> were sent to xen development list.

I remember these. A bunch got applied after some discussion but IIRC not
all of them were? Bei, are you able to resend the ones which did not?

Is there a wiki page or something somewhere detailing what a user needs
to do (i.e. build steps, guest cfg etc) in order to use the UEFI
support?

Cheers,
Ian.

> 
> Jordan Justen (Intel) had been working on integrating the OVMF changes
> last I checked, but I am not positive (I just changed jobs and
> relocated, and still am not on the TianoCore mailing list). I am
> adding him to the reply.
> 
> A
> 
> 
> 2011/9/26 Ian Campbell <Ian.Campbell@citrix.com>:
> > Adding Bei Guan and Andrey Warkentin who I think were student and mentor
> > respectively. Hi guys!
> >
> > On Sun, 2011-09-25 at 15:50 +0100, Keir Fraser wrote:
> >> The GSOC project student seemed to be making good progress, and there were
> >> others on the OVMF side who were giving him support. I haven't heard
> >> anything in about a month now, though.
> >
> > It would be very interesting to get an update (or a pointer to one) of
> > the status/result of this particular project now that the GSoC is over.
> >
> > Thanks,
> > Ian.

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Keir Fraser]

On 27/09/2011 07:41, "Ian Campbell" <Ian.Campbell@eu.citrix.com> wrote:

>> OVMF is aware of Xen, and work was being done on a PV block device
>> (Bei got as far as hypercall support and xenstore). Bei, are you
>> working on the PV block driver? I want to say we're pretty close on
>> that one (PV block not very hard, and neither is the block device
>> abstraction for EFI).
> 
> Thanks for the update, it sounds like it's pretty far along.
> 
> The PV enhancements etc sound great but it would be nice if we could tie
> up the baseline stuff so that folks can start using it.

Yes, we'd rather get something basically working into the tree, than wait
for everything to be absolutely perfect.

> Is Bei still working on the project since the end of GSoC? If not is
> there an archive of what he accomplished so that we can find someone to
> pick up on it if necessary?
> 
>> I/Bei had patches against hvmloader (to support booting OVMF) and Bei
>> had patches against the management stack as well to expose the
>> functionality of picking OVMF as the firmware. I believe all of these
>> were sent to xen development list.
> 
> I remember these. A bunch got applied after some discussion but IIRC not
> all of them were? Bei, are you able to resend the ones which did not?

Pretty much all the non-ovmf-specific changes went in (i.e., the most
contentious bits, to existing hvmloader code, should be pretty much checked
in already).

 -- Keir

> Is there a wiki page or something somewhere detailing what a user needs
> to do (i.e. build steps, guest cfg etc) in order to use the UEFI
> support?

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Andrei E. Warkentin]

Hi,

2011/9/27 Keir Fraser <keir.xen@gmail.com>:
> On 27/09/2011 07:41, "Ian Campbell" <Ian.Campbell@eu.citrix.com> wrote:
>
>>> OVMF is aware of Xen, and work was being done on a PV block device
>>> (Bei got as far as hypercall support and xenstore). Bei, are you
>>> working on the PV block driver? I want to say we're pretty close on
>>> that one (PV block not very hard, and neither is the block device
>>> abstraction for EFI).
>>
>> Thanks for the update, it sounds like it's pretty far along.
>>
>> The PV enhancements etc sound great but it would be nice if we could tie
>> up the baseline stuff so that folks can start using it.
>
> Yes, we'd rather get something basically working into the tree, than wait
> for everything to be absolutely perfect.
>
>> Is Bei still working on the project since the end of GSoC? If not is
>> there an archive of what he accomplished so that we can find someone to
>> pick up on it if necessary?
>>
>>> I/Bei had patches against hvmloader (to support booting OVMF) and Bei
>>> had patches against the management stack as well to expose the
>>> functionality of picking OVMF as the firmware. I believe all of these
>>> were sent to xen development list.
>>
>> I remember these. A bunch got applied after some discussion but IIRC not
>> all of them were? Bei, are you able to resend the ones which did not?
>
> Pretty much all the non-ovmf-specific changes went in (i.e., the most
> contentious bits, to existing hvmloader code, should be pretty much checked
> in already).
>

This is just a ping to let you know that I'm still on this. I'll
update you guys as soon as
I can get Bei's work on top of latest hg tree building and running.

Thanks,
A

Re: Re: [Xen-users] Is xen planning to support hvm guests booting under UEFI?

[Ian Campbell]

On Mon, 2011-10-10 at 19:52 +0100, Andrei E. Warkentin wrote:
> 
> This is just a ping to let you know that I'm still on this. I'll
> update you guys as soon as
> I can get Bei's work on top of latest hg tree building and running.

Thanks Andrei. If there's anything I can do to help please let me know.

Ian.