Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Robin van Leeuwen]

When i boot my domU on amd64 (debian unstable)
i get during startup in xend.log:

[2005-08-27 20:21:06 xend] DEBUG (image:153) initDomain: cpu=-1 
mem_kb=131072 ssidref=-1 dom=3
[2005-08-27 20:21:06 xend] DEBUG (XendDomainInfo:780) init_domain> 
Created domain=3 name=samba-ha memory=128
[2005-08-27 20:21:06 xend] INFO (XendRoot:141) EVENT> xend.domain.create 
['samba-ha', 3]
[2005-08-27 20:21:06 xend] INFO (image:208) buildDomain os=linux dom=3 
vcpus=1
[2005-08-27 20:21:06 xend] INFO (xsobj:326) DBMap>introduceDomain> 3 
5882 <EventChannel dom1:0:18 dom2:3:2> /domain/838ae36b-66f8$[2005-08-27 
20:21:06 xend] DEBUG (XendDomainInfo:640) writing key 
availability=online to path /cpu/0 in store
[2005-08-27 20:21:06 xend] WARNING (XendDomainInfo:1032) Unknown config 
field boot
[2005-08-27 20:21:06 xend] WARNING (XendDomainInfo:1032) Unknown config 
field display
[2005-08-27 20:21:06 xend] ERROR (XendDomainInfo:425) ['vif', ['mac', 
'aa:00:00:7e:cb:8c'], ['bridge', 'xen-br0']]
[2005-08-27 20:21:06 xend] ERROR (XendDomainInfo:445) aa:00:00:7e:cb:8c
[2005-08-27 20:21:06 xend] INFO (XendRoot:141) EVENT> 
xend.domain.unpause ['samba-ha', 3]

But my domain works fine. After a while (+/- 5 or 8 minutes the 
following message appears:
(i'm not sure if the n/ testing main was generated by the bug, i was 
editting /etc/apt/sources.list
when it happend)

----------- [cut here ] --------- [please bite here ] ---------
Kernel BUG at "drivers/xen/blkfront/blkfront.c":95n/ testing main
invalid operand: 0000 [1] SMP
CPU 0 Modules linked in: md5 ipv6 ext3 jbd
Pid: 30, comm: pdflush

The domain stops responding on console and network. xm list shows:
samba-ha           3      128    0      1   -----   2496.0

xm top shows:
3 ------       2739   98.4     131072    6.3     131072       6.3     
1    1      110     3422    0

so it's pretty musy doing something...

any more info needed?

kind regards,

Robin van Leeuwen

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Robin van Leeuwen]

Some gaps in my previous mail:
in my xen0 and xenU i disabled

    XEN_NETDEV_GRANT_TX=n
    XEN_NETDEV_GRANT_RX=n

To resolve bugzilla bug id 176
(http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=176)

Last pull of hg was patch of :  Fri Aug 26 15:06:49 2005


Robin van Leeuwen wrote:

> When i boot my domU on amd64 (debian unstable)
> i get during startup in xend.log:
>
> [2005-08-27 20:21:06 xend] DEBUG (image:153) initDomain: cpu=-1 
> mem_kb=131072 ssidref=-1 dom=3
> [2005-08-27 20:21:06 xend] DEBUG (XendDomainInfo:780) init_domain> 
> Created domain=3 name=samba-ha memory=128
> [2005-08-27 20:21:06 xend] INFO (XendRoot:141) EVENT> 
> xend.domain.create ['samba-ha', 3]
> [2005-08-27 20:21:06 xend] INFO (image:208) buildDomain os=linux dom=3 
> vcpus=1
> [2005-08-27 20:21:06 xend] INFO (xsobj:326) DBMap>introduceDomain> 3 
> 5882 <EventChannel dom1:0:18 dom2:3:2> 
> /domain/838ae36b-66f8$[2005-08-27 20:21:06 xend] DEBUG 
> (XendDomainInfo:640) writing key availability=online to path /cpu/0 in 
> store
> [2005-08-27 20:21:06 xend] WARNING (XendDomainInfo:1032) Unknown 
> config field boot
> [2005-08-27 20:21:06 xend] WARNING (XendDomainInfo:1032) Unknown 
> config field display
> [2005-08-27 20:21:06 xend] ERROR (XendDomainInfo:425) ['vif', ['mac', 
> 'aa:00:00:7e:cb:8c'], ['bridge', 'xen-br0']]
> [2005-08-27 20:21:06 xend] ERROR (XendDomainInfo:445) aa:00:00:7e:cb:8c
> [2005-08-27 20:21:06 xend] INFO (XendRoot:141) EVENT> 
> xend.domain.unpause ['samba-ha', 3]
>
> But my domain works fine. After a while (+/- 5 or 8 minutes the 
> following message appears:
> (i'm not sure if the n/ testing main was generated by the bug, i was 
> editting /etc/apt/sources.list
> when it happend)
>
> ----------- [cut here ] --------- [please bite here ] ---------
> Kernel BUG at "drivers/xen/blkfront/blkfront.c":95n/ testing main
> invalid operand: 0000 [1] SMP
> CPU 0 Modules linked in: md5 ipv6 ext3 jbd
> Pid: 30, comm: pdflush
>
> The domain stops responding on console and network. xm list shows:
> samba-ha           3      128    0      1   -----   2496.0
>
> xm top shows:
> 3 ------       2739   98.4     131072    6.3     131072       6.3     
> 1    1      110     3422    0
>
> so it's pretty musy doing something...
>
> any more info needed?
>
> kind regards,
>
> Robin van Leeuwen
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Keir Fraser]

On 27 Aug 2005, at 20:16, Robin van Leeuwen wrote:

> When i boot my domU on amd64 (debian unstable)
> i get during startup in xend.log:

How old is this version of Xen? I thought I fixed this bug a few days 
ago.

  -- Keir

RE: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Sylvain COUTANT]

> > When i boot my domU on amd64 (debian unstable)
> > i get during startup in xend.log:
> 
> How old is this version of Xen? I thought I fixed this bug a few days
> ago.

I have almost the same problem with unstable download from 26/08.


Regards,

--
Sylvain.

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Keir Fraser]

> > > When i boot my domU on amd64 (debian unstable)
> > > i get during startup in xend.log:
> > 
> > How old is this version of Xen? I thought I fixed this bug a few days
> > ago.
> 
> I have almost the same problem with unstable download from 26/08.

Can you try with the most recent repository version? I think I've now
fixed this.

 -- Keir

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[David F Barrera]

On Mon, 2005-08-29 at 15:15 +0100, Keir Fraser wrote:
> > > > When i boot my domU on amd64 (debian unstable)
> > > > i get during startup in xend.log:
> > > 
> > > How old is this version of Xen? I thought I fixed this bug a few days
> > > ago.
> > 
> > I have almost the same problem with unstable download from 26/08.
> 
> Can you try with the most recent repository version? I think I've now
> fixed this.
I tried it on a RHEL 4 box this morning as of the following changesets,
and DomU still crashed:

changeset:   6458:98de1d5fe5fbeedec536f3ed1618c6234e5186d1
tag:         tip
user:        shand@ubuntu.eng.hq.xensource.com
date:        Mon Aug 29 00:50:04 2005
summary:     Minor (partial) build fix for 2.4.

changeset:   6457:b26d8e1b443690049282c909a1ed41c3ba33bf5c
user:        shand@ubuntu.eng.hq.xensource.com
date:        Mon Aug 29 00:40:36 2005
summary:     Fix nasty tools race between barking xu_autoreap() and
xpopen3.wait() -

I don't know if this is the same bug. What I am seeing on the DomU
console is: Kernel panic - not syncing: BUG!
On the serial console, I see the message: smp_send_stop
disable_local_APIC. I thought this problem and the one reported on this
thread were the same, and I've filed them both under bugzilla #190.

> 
>  -- Keir
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
> 
-- 
Regards,

David F Barrera
Linux Technology Center
Systems and Technology Group, IBM

"The wisest men follow their own direction. "
                                                        Euripides

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Keir Fraser]

On 29 Aug 2005, at 16:42, David F Barrera wrote:

> I tried it on a RHEL 4 box this morning as of the following changesets,
> and DomU still crashed:

That's one changeset too old! The very next changeset contains the 
(likely) fix.

It has a comment starting "Allocate a request-id space per vbd in 
blkfront.".

  -- Keir

Re: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[David F Barrera]

On Mon, 2005-08-29 at 17:11 +0100, Keir Fraser wrote:
> On 29 Aug 2005, at 16:42, David F Barrera wrote:
> 
> > I tried it on a RHEL 4 box this morning as of the following changesets,
> > and DomU still crashed:
> 
> That's one changeset too old! The very next changese
> t contains the 
> (likely) fix.
That's what I though might be the case. I retested using a later pull,
and the problem has been resolved. Thx.

changeset:   6461:936be0ae823fd3a94ea06d92fdacad65b85d6014
tag:         tip
user:        kaf24@firebug.cl.cam.ac.uk
date:        Mon Aug 29 09:53:38 2005
summary:     Fix page_is_ram() start-of-day checks, and the watch
routine


> 
> It has a comment starting "Allocate a request-id space per vbd in 
> blkfront.".
> 
>   -- Keir
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
> 
-- 
Regards,

David F Barrera
Linux Technology Center
Systems and Technology Group, IBM

"The wisest men follow their own direction. "
                                                        Euripides

RE: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[Ian Pratt]

> But my domain works fine. After a while (+/- 5 or 8 minutes the 
> following message appears:
> (i'm not sure if the n/ testing main was generated by the bug, i was 
> editting /etc/apt/sources.list
> when it happend)

Please can you file this on bugzilla.
 
> ----------- [cut here ] --------- [please bite here ] ---------
> Kernel BUG at "drivers/xen/blkfront/blkfront.c":95n/ testing main
> invalid operand: 0000 [1] SMP
> CPU 0 Modules linked in: md5 ipv6 ext3 jbd
> Pid: 30, comm: pdflush

This is incomplete -- we probably forget to flush console before calling
crash().

Ian

RE: Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

[David F Barrera]

On Sat, 2005-08-27 at 22:29 +0100, Ian Pratt wrote:
> > But my domain works fine. After a while (+/- 5 or 8 minutes the 
> > following message appears:
> > (i'm not sure if the n/ testing main was generated by the bug, i was 
> > editting /etc/apt/sources.list
> > when it happend)
> 
> Please can you file this on bugzilla.

I've already filed in on Bugzilla, on 8/25: 
Bugzilla Bug 190    Kernel BUG at "drivers/xen/blkfront/blkfront.c":95

David
>  
> > ----------- [cut here ] --------- [please bite here ] ---------
> > Kernel BUG at "drivers/xen/blkfront/blkfront.c":95n/ testing main
> > invalid operand: 0000 [1] SMP
> > CPU 0 Modules linked in: md5 ipv6 ext3 jbd
> > Pid: 30, comm: pdflush
> 
> This is incomplete -- we probably forget to flush console before calling
> crash().
> 
> Ian
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
> 
-- 
Regards,

David F Barrera
Linux Technology Center
Systems and Technology Group, IBM

"The wisest men follow their own direction. "
                                                        Euripides