hang on reboot with 3.6-rc1

hang on reboot with 3.6-rc1

[David Ahern]

[-- Attachment #1: Type: text/plain, Size: 130 bytes --]

Not sure if KVM is the culprit, but it is the last line shown on the 
console. I have to power cycle the server to reboot.

David

[-- Attachment #2: kvm-3.6-rc1.png --]
[-- Type: image/png, Size: 37339 bytes --]

Re: hang on reboot with 3.6-rc1

[Avi Kivity]

On 08/08/2012 07:27 PM, David Ahern wrote:
> Not sure if KVM is the culprit, but it is the last line shown on the
> console. I have to power cycle the server to reboot.

Have you tried rmmoding the kvm modules before reboot?

Were any guests running during this?


-- 
error compiling committee.c: too many arguments to function

Re: hang on reboot with 3.6-rc1

[David Ahern]

On 8/9/12 3:42 AM, Avi Kivity wrote:
> On 08/08/2012 07:27 PM, David Ahern wrote:
>> Not sure if KVM is the culprit, but it is the last line shown on the
>> console. I have to power cycle the server to reboot.
>
> Have you tried rmmoding the kvm modules before reboot?
>
> Were any guests running during this?

hmm.... seems to be a side effect of https://lkml.org/lkml/2012/8/3/370.

David

Re: hang on reboot with 3.6-rc1

[Alex Williamson]

On Thu, 2012-08-09 at 11:10 -0600, David Ahern wrote:
> On 8/9/12 3:42 AM, Avi Kivity wrote:
> > On 08/08/2012 07:27 PM, David Ahern wrote:
> >> Not sure if KVM is the culprit, but it is the last line shown on the
> >> console. I have to power cycle the server to reboot.
> >
> > Have you tried rmmoding the kvm modules before reboot?
> >
> > Were any guests running during this?
> 
> hmm.... seems to be a side effect of https://lkml.org/lkml/2012/8/3/370.

Is it fixed with the patches for that?  We're operating on a NULL
pointer with that bug on your system, so you probably don't want to be
assuming the system is entirely healthy after you hit it.  Thanks,

Alex

Re: hang on reboot with 3.6-rc1

[David Ahern]

On 8/9/12 11:26 AM, Alex Williamson wrote:
> On Thu, 2012-08-09 at 11:10 -0600, David Ahern wrote:
>> On 8/9/12 3:42 AM, Avi Kivity wrote:
>>> On 08/08/2012 07:27 PM, David Ahern wrote:
>>>> Not sure if KVM is the culprit, but it is the last line shown on the
>>>> console. I have to power cycle the server to reboot.
>>>
>>> Have you tried rmmoding the kvm modules before reboot?
>>>
>>> Were any guests running during this?
>>
>> hmm.... seems to be a side effect of https://lkml.org/lkml/2012/8/3/370.
>
> Is it fixed with the patches for that?  We're operating on a NULL
> pointer with that bug on your system, so you probably don't want to be
> assuming the system is entirely healthy after you hit it.  Thanks,


Agreed and yes.

Latest Linus tree has the iommu patch. Still need to get the other one 
in (https://lkml.org/lkml/2012/8/4/141) or whatever alternative is 
deemed better.

Re: hang on reboot with 3.6-rc1

[Alex Williamson]

On Thu, 2012-08-09 at 12:17 -0600, David Ahern wrote:
> On 8/9/12 11:26 AM, Alex Williamson wrote:
> > On Thu, 2012-08-09 at 11:10 -0600, David Ahern wrote:
> >> On 8/9/12 3:42 AM, Avi Kivity wrote:
> >>> On 08/08/2012 07:27 PM, David Ahern wrote:
> >>>> Not sure if KVM is the culprit, but it is the last line shown on the
> >>>> console. I have to power cycle the server to reboot.
> >>>
> >>> Have you tried rmmoding the kvm modules before reboot?
> >>>
> >>> Were any guests running during this?
> >>
> >> hmm.... seems to be a side effect of https://lkml.org/lkml/2012/8/3/370.
> >
> > Is it fixed with the patches for that?  We're operating on a NULL
> > pointer with that bug on your system, so you probably don't want to be
> > assuming the system is entirely healthy after you hit it.  Thanks,
> 
> 
> Agreed and yes.
> 
> Latest Linus tree has the iommu patch. Still need to get the other one 
> in (https://lkml.org/lkml/2012/8/4/141) or whatever alternative is 
> deemed better.

I think you should be ok with just what is in Linus' tree now.  The
change in pci_acs_path_enabled is the same kind of bug, but I don't
think you'll hit it with the fixed iommu code.  Please let me know if
that's not the case.  Thanks,

Alex

Re: hang on reboot with 3.6-rc1

[David Ahern]

On 8/9/12 1:03 PM, Alex Williamson wrote:

> I think you should be ok with just what is in Linus' tree now.  The
> change in pci_acs_path_enabled is the same kind of bug, but I don't
> think you'll hit it with the fixed iommu code.  Please let me know if
> that's not the case.  Thanks,

Indeed, latest tree is clean. no oops, VF shows fine and passthrough works.

David

Re: hang on reboot with 3.6-rc1

[Alex Williamson]

On Thu, 2012-08-09 at 13:35 -0600, David Ahern wrote:
> On 8/9/12 1:03 PM, Alex Williamson wrote:
> 
> > I think you should be ok with just what is in Linus' tree now.  The
> > change in pci_acs_path_enabled is the same kind of bug, but I don't
> > think you'll hit it with the fixed iommu code.  Please let me know if
> > that's not the case.  Thanks,
> 
> Indeed, latest tree is clean. no oops, VF shows fine and passthrough works.

Great, thanks!

Alex