pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Andreas Kinzler]

Hello Jeremy & Keir,

sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x 
still does not boot if x > 18 on my machines (and on many other 
machines, see this mailing list) is not solved yet.

Shouldn't this be a great focus of debugging and fixing?

What is the point of having releases for 4.0.2 or 4.1 if I don't have a 
dom0 kernel?

The last time I tried it was with snapshot 892d2f052e97.

Regards Andreas

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Pasi Kärkkäinen]

On Mon, Mar 14, 2011 at 11:43:41AM +0100, Andreas Kinzler wrote:
> Hello Jeremy & Keir,
>

Hello,

Added Konrad to CC list.

> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x  
> still does not boot if x > 18 on my machines (and on many other  
> machines, see this mailing list) is not solved yet.
>
> Shouldn't this be a great focus of debugging and fixing?
>
> What is the point of having releases for 4.0.2 or 4.1 if I don't have a  
> dom0 kernel?
>
> The last time I tried it was with snapshot 892d2f052e97.
>

So please remind about the issue again..?
Please post a Xen and dom0 kernel full serial logs..

Was this the issue you tried to bisect? Can you post the
latest changeset that works, and the first that fails? 

-- Pasi

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Andreas Kinzler]

On 14.03.2011 12:48, Pasi Kärkkäinen wrote:
>> Hello Jeremy&  Keir,
> Hello,
> Added Konrad to CC list.
>> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
>> still does not boot if x>  18 on my machines (and on many other
>> machines, see this mailing list) is not solved yet.
>> Shouldn't this be a great focus of debugging and fixing?
>> What is the point of having releases for 4.0.2 or 4.1 if I don't have a
>> dom0 kernel?
>> The last time I tried it was with snapshot 892d2f052e97.
> So please remind about the issue again..?
> Please post a Xen and dom0 kernel full serial logs..

Yes, this is just a reminder. The archives already have all the infos. See:
http://lists.xensource.com/archives/html/xen-devel/2010-09/msg00758.html
http://lists.xensource.com/archives/html/xen-devel/2010-10/msg00471.html

Regards Andreas

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Pasi Kärkkäinen]

On Tue, Mar 15, 2011 at 11:26:39AM +0100, Andreas Kinzler wrote:
> On 14.03.2011 12:48, Pasi Kärkkäinen wrote:
>>> Hello Jeremy&  Keir,
>> Hello,
>> Added Konrad to CC list.
>>> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
>>> still does not boot if x>  18 on my machines (and on many other
>>> machines, see this mailing list) is not solved yet.
>>> Shouldn't this be a great focus of debugging and fixing?
>>> What is the point of having releases for 4.0.2 or 4.1 if I don't have a
>>> dom0 kernel?
>>> The last time I tried it was with snapshot 892d2f052e97.
>> So please remind about the issue again..?
>> Please post a Xen and dom0 kernel full serial logs..
>
> Yes, this is just a reminder. The archives already have all the infos. See:
> http://lists.xensource.com/archives/html/xen-devel/2010-09/msg00758.html
> http://lists.xensource.com/archives/html/xen-devel/2010-10/msg00471.html
>

Thanks for the links.

I think Konrad was planning to get back to this issue later this week..
It has been reported by other users aswell.

-- Pasi

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Stefano Stabellini]

On Mon, 14 Mar 2011, Andreas Kinzler wrote:
> Hello Jeremy & Keir,
> 
> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x 
> still does not boot if x > 18 on my machines (and on many other 
> machines, see this mailing list) is not solved yet.
> 
> Shouldn't this be a great focus of debugging and fixing?
> 
> What is the point of having releases for 4.0.2 or 4.1 if I don't have a 
> dom0 kernel?
> 
> The last time I tried it was with snapshot 892d2f052e97.

can you try this simple patch please?

---



diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c
index 73ffd55..69f73c1 100644
--- a/arch/x86/mm/init.c
+++ b/arch/x86/mm/init.c
@@ -287,7 +287,7 @@ unsigned long __init_refok init_memory_mapping(unsigned long start,
 	load_cr3(swapper_pg_dir);
 #endif
 
-#ifdef CONFIG_X86_64
+#if 0
 	if (!after_bootmem && !start) {
 		pud_t *pud;
 		pmd_t *pmd;

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[sven]

On 03/14/2011 11:43 AM, Andreas Kinzler wrote:
> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
> still does not boot if x > 18 on my machines (and on many other
> machines, see this mailing list) is not solved yet.

Hello Andreas,

there seems to be good news. Today Konrad published a branch with 
possible fixes for this really-longtime-annoying showstopper.

For me, the 'get_phys_to_machine'-issue seems to be history.

You might want to try Konrad's "debug-code" branch, too:

http://git.kernel.org/?p=linux/kernel/git/konrad/xen.git


Regards,
Sven

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Andreas Kinzler]

On 16.03.2011 15:36, Stefano Stabellini wrote:
>> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
>> still does not boot if x>  18 on my machines (and on many other
>> machines, see this mailing list) is not solved yet.
>> Shouldn't this be a great focus of debugging and fixing?
>> What is the point of having releases for 4.0.2 or 4.1 if I don't have a
>> dom0 kernel?
>> The last time I tried it was with snapshot 892d2f052e97.
> can you try this simple patch please?

Yes, I can confirm that this patch seems to help. I tried snapshot 
ea954f6 (current commit as of now). With the patch it does boot and 
first runs of Xen VMs seem all OK. Without the patch it does not boot.

Is this the official fix? Any side-effects?

Regards Andreas

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Stefano Stabellini]

On Mon, 21 Mar 2011, Andreas Kinzler wrote:
> On 16.03.2011 15:36, Stefano Stabellini wrote:
> >> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
> >> still does not boot if x>  18 on my machines (and on many other
> >> machines, see this mailing list) is not solved yet.
> >> Shouldn't this be a great focus of debugging and fixing?
> >> What is the point of having releases for 4.0.2 or 4.1 if I don't have a
> >> dom0 kernel?
> >> The last time I tried it was with snapshot 892d2f052e97.
> > can you try this simple patch please?
> 
> Yes, I can confirm that this patch seems to help. I tried snapshot 
> ea954f6 (current commit as of now). With the patch it does boot and 
> first runs of Xen VMs seem all OK. Without the patch it does not boot.
> 
> Is this the official fix? Any side-effects?
 
No it is not, but it shouldn't have any bad side effects.

Re: pvops0 kernel 2.6.32.x still does not boot if x > 18, shouldn't that be fixed for release of Xen 4.0.2/4.1

[Pasi Kärkkäinen]

On Wed, Mar 16, 2011 at 11:23:22PM +0100, sven wrote:
> On 03/14/2011 11:43 AM, Andreas Kinzler wrote:
>> sorry to be persistent, but the problem that pvops0 kernel 2.6.32.x
>> still does not boot if x > 18 on my machines (and on many other
>> machines, see this mailing list) is not solved yet.
>
> Hello Andreas,
>
> there seems to be good news. Today Konrad published a branch with  
> possible fixes for this really-longtime-annoying showstopper.
>
> For me, the 'get_phys_to_machine'-issue seems to be history.
>
> You might want to try Konrad's "debug-code" branch, too:
>
> http://git.kernel.org/?p=linux/kernel/git/konrad/xen.git
>

Hello,

Do I remember correctly this bugfix was waiting for a suitable patch inclusion
to upstream Linux, and then a backport to pvops 2.6.32.x kernel? 

-- Pasi