Windows...

Windows...

[M Afifi]

I've read the FAQ and bits of information on the Xen website and in particular,

----------------------------------------------------------------------------------------------------------------------------

1.4 Does Xen support Microsoft Windows?

Unfortunately there are no plans to support any versions of Windows in
the near future. Furthermore, a port of Windows would be encumbered by
licensing issues. Longer term, virtualisation features in
next-generation CPUs should make it much easier to support unmodified
OSes: at that time we will reconsider Windows support.

----------------------------------------------------------------------------------------------------------------------------

Can I know a little bit more about,

How much changes are involved in the O.S.?
Was Microsoft even approached with this project to see what their take on it is?
Next generation CPU / Intel's VT, how soon can we expect a Xen release
to support this?

Given all of the above, what are the other challenges involved in
supporting Windows on Xen and what sort of timeline are we looking at?

Cheers,

M Afifi


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Re: Windows...

[Mark A. Williamson]

> How much changes are involved in the O.S.?

For a native Xen port:

* The OS memory management code must be modified to handle running in 
non-contigous physical memory (rather hard)
* The OS context switching code needs to be changed to call into Xen
* the platform initialisation code needs mostly removing and replacing with a 
small amount of Xen-specific stuff
* add various support infrastructure control interface, event channels, 
virtual interrupts
* Virtual device drivers need to be written (OK if someone takes the time to 
understand the Windows driver API)

Am I missing anything guys?

Anyhow, this requires source code access, which is the problem :-(  I heard 
talk of possibly getting Windows running under some extra virtualisation 
features without XP src code access but I think that's more of a theory at 
the moment, with no plans to work on it.

> Was Microsoft even approached with this project to see what their take on
> it is?

MS Research in Cambridge provided support for a port of Win XP to Xen 1.2 to 
be carried out locally some time ago.  I think they were really interested in 
it as a research platform rather than a product, which is fair enough - 
Microsoft have their own virtualisation platform, so I imagine they'd 
concentrate on that.

> Next generation CPU / Intel's VT, how soon can we expect a Xen 
> release to support this?

In-Xen support for VT is present, although I'm not sure the corresponding 
userspace support code is checked in yet.  I don't think Intel are disclosing 
when VT hardware will be released, however.

> Given all of the above, what are the other challenges involved in
> supporting Windows on Xen and what sort of timeline are we looking at?

Since there's VT code being integrated into Xen currently it seems reasonable 
to assume that when VT hardware comes out, Windows will Just Work (TM) but I 
don't know the concrete plans.  I may be wrong...

HTH,
Mark


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Re: Windows...

[Derrik Pates]

Mark A. Williamson wrote:
> Since there's VT code being integrated into Xen currently it seems reasonable 
> to assume that when VT hardware comes out, Windows will Just Work (TM) but I 
> don't know the concrete plans.  I may be wrong...

Well, I think that even if Windows itself can run unmodified, there'd 
still need to be a virtual BIOS, virtual VGA device, virtual IDE 
controller, virtual NIC device, etc. They'd need to be emulations of 
real devices, at least they'd have to be availabe until drivers can be 
loaded (and written, of course) for Windows that can take over and 
bypass the emulated devices. So the support for Intel VT/AMD Pacifica is 
just the first step of many, really.

-- 
Derrik Pates
demon@devrandom.net


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Re: Windows...

[Steven Hand]

>Mark A. Williamson wrote:
>> Since there's VT code being integrated into Xen currently it seems 
>>reasonable to assume that when VT hardware comes out, Windows will Just 
>>Work (TM) but I don't know the concrete plans.  I may be wrong...
>
>Well, I think that even if Windows itself can run unmodified, there'd 
>still need to be a virtual BIOS, virtual VGA device, virtual IDE 
>controller, virtual NIC device, etc. They'd need to be emulations of 
>real devices, at least they'd have to be availabe until drivers can be 
>loaded (and written, of course) for Windows that can take over and 
>bypass the emulated devices. So the support for Intel VT/AMD Pacifica is 
>just the first step of many, really.

Some support for fully virtual devices is already checked into the 
-unstable tree (see tools/ioemu). This is useful for start-of-day, 
but in practice for performance one would replace these drivers with
Xen-aware paravirtualized ones after install. 

cheers, 

S.


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Fwd: Windows...

[M Afifi]

Thanks for the information everyone.  It's a shame that Microsoft are
seeing it mainly as a research opportunity when clearly Xen has
already attracted some rather big players and customers.


On Tue, 18 Jan 2005 07:41:38 +0000, Steven Hand
<Steven.Hand@cl.cam.ac.uk> wrote:
>
> >Mark A. Williamson wrote:
> >> Since there's VT code being integrated into Xen currently it seems
> >>reasonable to assume that when VT hardware comes out, Windows will Just
> >>Work (TM) but I don't know the concrete plans.  I may be wrong...
> >
> >Well, I think that even if Windows itself can run unmodified, there'd
> >still need to be a virtual BIOS, virtual VGA device, virtual IDE
> >controller, virtual NIC device, etc. They'd need to be emulations of
> >real devices, at least they'd have to be availabe until drivers can be
> >loaded (and written, of course) for Windows that can take over and
> >bypass the emulated devices. So the support for Intel VT/AMD Pacifica is
> >just the first step of many, really.
>
> Some support for fully virtual devices is already checked into the
> -unstable tree (see tools/ioemu). This is useful for start-of-day,
> but in practice for performance one would replace these drivers with
> Xen-aware paravirtualized ones after install.
>
> cheers,
>
> S.
>


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt

Re: Windows...

[Jared Rhine]

> Some support for fully virtual devices is already checked into the 
> -unstable tree (see tools/ioemu).

As this apparently adds a g++ build requirement, would it be possible to
add a switch or build option or something to avoid building ioemu?  My
build process is currently breaking because of the (appreciated)
addition of ioemu/g++/X11 build requirements.

-- 
Jared Rhine <jared@wordzoo.com>



-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl

Re: Windows

[Viswesh S]

[-- Attachment #1: Type: text/plain, Size: 2913 bytes --]

I had tried the grub configuration which you have suggested before, from the wiki.
But still there wasnt any use :(
Viswesh



----- Original Message ----
From: Javier Martín <lordhabbit@gmail.com>
To: The development of GRUB 2 <grub-devel@gnu.org>
Sent: Wednesday, 30 July, 2008 11:59:34 PM
Subject: Re: Windows

2008/7/30 Viswesh S <viswesh_vichu@yahoo.com>:
> The problem is that, I am not able to boot windows by this configuration.
>
>
>
> I am getting the error
>
>
>
> "A disk read Error Occurred"
>
>
>
> If I make the windows disk as the first boot device,then windows is coming
> up properly.
Drivemap module to the rescue!! ^^
Hoping to resurrect the discussion about it, I developed the module
for just that purpose: I have a similar configuration, in which
Windows refuses to boot if its drive is not "the first BIOS drive".

In some cases and versions of Windows you can make it work without the
drivemap module with:
menuentry "Windows" {
    set root=(hd1,1)
    chainloader +1
}
But not all versions of NTLDR obey the "boot drive" parameter passed
in DL: in GRUB legacy, this was solved by the "map" command which made
Windows believed it was booting from the first BIOS drive by
intercepting all BIOS I/O calls and mapping the drive numbers - the
proposed "drivemap" module does the same in GRUB2, but it's not been
accepted yet (ahem xD).

>
>
>
> Hence I feel this is something to do with the grub2 configuration which is
> missing.
>
>
>
> If any needs any information(about my machine) to configure grub2 for
> windows,please let me know.
>
>
>
> Regards,
>
>
>
> Viswesh
>
> ----- Original Message ----
> From: Pavel Roskin <proski@gnu.org>
> To: The development of GRUB 2 <grub-devel@gnu.org>
> Sent: Wednesday, 30 July, 2008 11:39:56 PM
> Subject: Re: Windows
>
> On Wed, 2008-07-30 at 11:00 -0700, Viswesh S wrote:
>> menuentry "Windows" {
>>    chainloader (hd1,1)+1
>> }
>>
>> Is all this is sufficient or is there anything else I should add.
>
> I think you are better off describing the problem, not asking others to
> guess what could be wrong without knowing what the problem is.
>
> --
> Regards,
> Pavel Roskin
>
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/grub-devel
>
> ________________________________
> Add more friends to your messenger and enjoy! Invite them now.
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/grub-devel
>
>


_______________________________________________
Grub-devel mailing list
Grub-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/grub-devel



      Did you know? You can CHAT without downloading messenger. Go to http://in.webmessenger.yahoo.com/

[-- Attachment #2: Type: text/html, Size: 4566 bytes --]

Re: Windows

[Viswesh S]

[-- Attachment #1: Type: text/plain, Size: 1252 bytes --]

The problem is that, I am not able to boot windows by this configuration.
I am getting the error
"A disk read Error Occurred"
If I make the windows disk as the first boot device,then windows is coming up properly.
Hence I feel this is something to do with the grub2 configuration which is missing.
If any needs any information(about my machine) to configure grub2 for windows,please let me know.
Regards,
Viswesh



----- Original Message ----
From: Pavel Roskin <proski@gnu.org>
To: The development of GRUB 2 <grub-devel@gnu.org>
Sent: Wednesday, 30 July, 2008 11:39:56 PM
Subject: Re: Windows

On Wed, 2008-07-30 at 11:00 -0700, Viswesh S wrote:
> menuentry "Windows" {
>    chainloader (hd1,1)+1
> }
>  
> Is all this is sufficient or is there anything else I should add.

I think you are better off describing the problem, not asking others to
guess what could be wrong without knowing what the problem is.

-- 
Regards,
Pavel Roskin


_______________________________________________
Grub-devel mailing list
Grub-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/grub-devel



      Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/

[-- Attachment #2: Type: text/html, Size: 2068 bytes --]

Re: Windows

[Javier Martín]

2008/7/30 Viswesh S <viswesh_vichu@yahoo.com>:
> The problem is that, I am not able to boot windows by this configuration.
>
>
>
> I am getting the error
>
>
>
> "A disk read Error Occurred"
>
>
>
> If I make the windows disk as the first boot device,then windows is coming
> up properly.
Drivemap module to the rescue!! ^^
Hoping to resurrect the discussion about it, I developed the module
for just that purpose: I have a similar configuration, in which
Windows refuses to boot if its drive is not "the first BIOS drive".

In some cases and versions of Windows you can make it work without the
drivemap module with:
menuentry "Windows" {
    set root=(hd1,1)
    chainloader +1
}
But not all versions of NTLDR obey the "boot drive" parameter passed
in DL: in GRUB legacy, this was solved by the "map" command which made
Windows believed it was booting from the first BIOS drive by
intercepting all BIOS I/O calls and mapping the drive numbers - the
proposed "drivemap" module does the same in GRUB2, but it's not been
accepted yet (ahem xD).

>
>
>
> Hence I feel this is something to do with the grub2 configuration which is
> missing.
>
>
>
> If any needs any information(about my machine) to configure grub2 for
> windows,please let me know.
>
>
>
> Regards,
>
>
>
> Viswesh
>
> ----- Original Message ----
> From: Pavel Roskin <proski@gnu.org>
> To: The development of GRUB 2 <grub-devel@gnu.org>
> Sent: Wednesday, 30 July, 2008 11:39:56 PM
> Subject: Re: Windows
>
> On Wed, 2008-07-30 at 11:00 -0700, Viswesh S wrote:
>> menuentry "Windows" {
>>    chainloader (hd1,1)+1
>> }
>>
>> Is all this is sufficient or is there anything else I should add.
>
> I think you are better off describing the problem, not asking others to
> guess what could be wrong without knowing what the problem is.
>
> --
> Regards,
> Pavel Roskin
>
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/grub-devel
>
> ________________________________
> Add more friends to your messenger and enjoy! Invite them now.
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> http://lists.gnu.org/mailman/listinfo/grub-devel
>
>

Windows

[Viswesh S]

[-- Attachment #1: Type: text/plain, Size: 2021 bytes --]

Hi,
With grub2,what are the things I should take care, if I have another hard disk (ide , primary) with windows.
My initial boot device is a sata one with linux.
This is my device map 
(hd0)   /dev/sda
(hd1)   /dev/hda

My fdisk
Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
   Device Boot      Start         End      Blocks   Id  System
/dev/sda1   *           1        3647    29294496   83  Linux
/dev/sda2            3648        4133     3903795   82  Linux swap / Solaris
/dev/sda3            4134       10322    49713142+  83  Linux
/dev/sda4           10323       19457    73376887+   5  Extended
/dev/sda5           10323       19457    73376856   83  Linux
Disk /dev/hda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
   Device Boot      Start         End      Blocks   Id  System
/dev/hda1   *           1        9729    78148161    7  HPFS/NTFS
/dev/hda2            9730       19458    78140416    7  HPFS/NTFS

This is my grub.cfg for windows
menuentry "GNU/Linux, linux 2.6.15-51-amd64-k8" {
    linux   (hd0,1)/boot/vmlinuz-2.6.15-51-amd64-k8 root=/dev/sda1 ro
    initrd  (hd0,1)/boot/initrd.img-2.6.15-51-amd64-k8
}
menuentry "GNU/Linux, linux 2.6.15-51-amd64-k8 (single-user mode)" {
    linux   (hd0,1)/boot/vmlinuz-2.6.15-51-amd64-k8 root=/dev/sda1 ro single
    initrd  (hd0,1)/boot/initrd.img-2.6.15-51-amd64-k8
}
menuentry "Windows" {
    chainloader (hd1,1)+1
}
Is all this is sufficient or is there anything else I should add.
Viswesh


      Connect with friends all over the world. Get Yahoo! India Messenger at http://in.messenger.yahoo.com/?wm=n/

[-- Attachment #2: Type: text/html, Size: 3727 bytes --]

Re: Windows

[Pavel Roskin]

On Wed, 2008-07-30 at 11:00 -0700, Viswesh S wrote:
> menuentry "Windows" {
>     chainloader (hd1,1)+1
> }
>  
> Is all this is sufficient or is there anything else I should add.

I think you are better off describing the problem, not asking others to
guess what could be wrong without knowing what the problem is.

-- 
Regards,
Pavel Roskin

windows

[Jody Belka]

I was wondering what the current state is with windows on xen?

Secondly, i had a go at building the vmware vmmon kernel module
within dom0. It built ok, but trying to insert it into the running
kernel with insmod caused a segmentation fault from insmod, and
left the module in an initialising state, unusable, and also unable
to be removed without a reboot. I was wondering if anyone else has
had a go at this and had better luck?


-- 
Jody Belka
knew (at) pimb (dot) org


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click

Re: windows

[Ian Pratt]

> I was wondering what the current state is with windows on xen?

I don't think we're ever likely to see the current port available
on 'general release'.

It may (or may not) be possible to re-implement the XP port as
just a new HAL, exploiting Xen's shadow pagetable support. It's
something to look in to...

> Secondly, i had a go at building the vmware vmmon kernel module
> within dom0. It built ok, but trying to insert it into the running
> kernel with insmod caused a segmentation fault from insmod, and
> left the module in an initialising state, unusable, and also unable
> to be removed without a reboot. I was wondering if anyone else has
> had a go at this and had better luck?

vmmon assumes its running in ring 0 on the bare metal. There's no
chance of VMware running under Xen (as opposed to the other way
round). Our aim with XenLinux was to implement the full user
space binary API. vmmon would require a very complete x86
virtualization to work.  [Incidentally, does VMware even work
recursively?]


Ian


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click

Re: windows

[Tim Freeman]

On Wed, 14 Jul 2004 21:44:18 +0100
Ian Pratt <Ian.Pratt@cl.cam.ac.uk> wrote:

> [Incidentally, does VMware even work recursively?]

According to this the installer detects it is running inside another
VMWare:

http://www.eecs.harvard.edu/~konrad/References/notes/CS253-VMStacking-Presentation.pdf


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click

Re: windows

[Jody Belka]

On Wed, Jul 14, 2004 at 09:44:18PM +0100, Ian Pratt wrote:
> > I was wondering what the current state is with windows on xen?
> 
> I don't think we're ever likely to see the current port available
> on 'general release'.
> 
> It may (or may not) be possible to re-implement the XP port as
> just a new HAL, exploiting Xen's shadow pagetable support. It's
> something to look in to...

any idea then if/when this might happen if it does? and how
long it might possibly take if so? having tried out QEmu, to
see if it'd be feasible, it's horribly slow. without windows
support for xen i won't be able to run xen on my primary
server, which would be a shame.

> [Incidentally, does VMware even work recursively?]

Nope, appears not having googled a bit.


-- 
Jody Belka
knew (at) pimb (dot) org


-------------------------------------------------------
This SF.Net email is sponsored by BEA Weblogic Workshop
FREE Java Enterprise J2EE developer tools!
Get your free copy of BEA WebLogic Workshop 8.1 today.
http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click