Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Boris Derzhavets]

[-- Attachment #1.1: Type: text/plain, Size: 4120 bytes --]

In response to Michael Young post fedora-xen list :-

> There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)

Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
Works as vanilla as well . Allows to run KVMs been built with standard fedora kernel.

Boris.

--- On Thu, 1/6/11, Pasi Kärkkäinen <pasik@iki.fi> wrote:

From: Pasi Kärkkäinen <pasik@iki.fi>
Subject: [Xen-devel] 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85
To: xen-devel@lists.xensource.com
Cc: "Ian Campbell" <Ian.Campbell@citrix.com>, "Konrad Rzeszutek Wilk" <konrad.wilk@oracle.com>
Date: Thursday, January 6, 2011, 12:36 PM

Hello,

First of all.. big thanks to everyone involved in upstreaming Xen dom0 bits! 

I just tried upstream pvops 2.6.37 dom0 kernel on one of my machines, and it worked
out-of-the-box ! Even nouveau kms/modesetting worked just fine.

There was non-fatal BUG/traceback though:

Full dmesg log available at: http://pasik.reaktio.net/xen/logs/dmesg-2.6.37-1.xendom0.fc15.x86_64-kms-modesetting.txt

[    9.930809] BUG: sleeping function called from invalid context at kernel/mutex.c:85
[    9.931616] in_atomic(): 0, irqs_disabled(): 1, pid: 572, name: kworker/0:2
[    9.932403] Pid: 572, comm: kworker/0:2 Not tainted 2.6.37-1.xendom0.fc15.x86_64 #1
[    9.933188] Call Trace:
[    9.933973]  [<ffffffff81042005>] __might_sleep+0xeb/0xf0
[    9.934758]  [<ffffffff81476e9e>] mutex_lock+0x24/0x50
[    9.935533]  [<ffffffff810abe14>] free_desc+0x2e/0x5f
[    9.936296]  [<ffffffff810abe80>] irq_free_descs+0x3b/0x71
[    9.937064]  [<ffffffff812ac0c8>] irq_free_desc+0x13/0x15
[    9.937812]  [<ffffffff812aca85>] xen_destroy_irq+0xe1/0xf5
[    9.938545]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
[    9.939270]  [<ffffffff813b70b9>] xen_teardown_msi_irq+0xe/0x10
[    9.939982]  [<ffffffff81255630>] default_teardown_msi_irqs+0x57/0x80
[    9.940674]  [<ffffffff812550f6>] free_msi_irqs+0x8b/0xe9
[    9.941352]  [<ffffffff81255ff4>] pci_disable_msix+0x35/0x39
[    9.942030]  [<ffffffffa005111a>] xhci_cleanup_msix+0x31/0x51 [xhci_hcd]
[    9.942705]  [<ffffffffa0051d35>] xhci_suspend+0x199/0x1ab [xhci_hcd]
[    9.943370]  [<ffffffffa00565ce>] xhci_pci_suspend+0x23/0x25 [xhci_hcd]
[    9.944029]  [<ffffffff81350355>] suspend_common+0x65/0xe2
[    9.944677]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
[    9.945318]  [<ffffffff813503eb>] hcd_pci_runtime_suspend+0x19/0x43
[    9.945955]  [<ffffffff8124549b>] pci_pm_runtime_suspend+0x58/0xea
[    9.946587]  [<ffffffff81245443>] ? pci_pm_runtime_suspend+0x0/0xea
[    9.947221]  [<ffffffff812ef7a0>] rpm_suspend+0x285/0x3c5
[    9.947836]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
[    9.948445]  [<ffffffff8107d31f>] ? arch_local_irq_restore+0xb/0xd
[    9.949051]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
[    9.949648]  [<ffffffff8107d32c>] ? arch_local_irq_disable+0xb/0xd
[    9.950230]  [<ffffffff812f0357>] __pm_runtime_suspend+0x51/0x6d
[    9.950801]  [<ffffffff8124515b>] ? pci_pm_runtime_idle+0x0/0x4d
[    9.951359]  [<ffffffff812451a2>] pci_pm_runtime_idle+0x47/0x4d
[    9.951906]  [<ffffffff812efaa5>] rpm_idle+0x14d/0x18f
[    9.952445]  [<ffffffff812f042e>] ? pm_runtime_work+0x0/0xa0
[    9.952970]  [<ffffffff812f0499>] pm_runtime_work+0x6b/0xa0
[    9.953490]  [<ffffffff810665ab>] process_one_work+0x18c/0x285
[    9.954010]  [<ffffffff810673b2>] worker_thread+0x104/0x1a4
[    9.954526]  [<ffffffff810672ae>] ? worker_thread+0x0/0x1a4
[    9.955038]  [<ffffffff8106a9e3>] kthread+0x82/0x8a
[    9.955545]  [<ffffffff8100bae4>] kernel_thread_helper+0x4/0x10
[    9.956060]  [<ffffffff8100aee3>] ? int_ret_from_sys_call+0x7/0x1b
[    9.956575]  [<ffffffff814782dd>] ? retint_restore_args+0x5/0x6
[    9.957088]  [<ffffffff8100bae0>] ? kernel_thread_helper+0x0/0x10


-- Pasi


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel



      

[-- Attachment #1.2: Type: text/html, Size: 5838 bytes --]

[-- Attachment #2: dmesg.37-1.xen.log.gz --]
[-- Type: application/x-gzip, Size: 16492 bytes --]

[-- Attachment #3: dmesg.37-1.vanilla.log.gz --]
[-- Type: application/x-gzip, Size: 16936 bytes --]

[-- Attachment #4: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Gerry Reno]

On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
> In response to Michael Young post fedora-xen list :-
>
> > There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
>
> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
> Works as vanilla as well . Allows to run KVMs been built with standard
> fedora kernel.
>
> Boris.
>

Haven't tried 2.6.37 yet but I have not been able to get X running with
any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
stack trace.

-Gerry

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Boris Derzhavets]

[-- Attachment #1.1: Type: text/plain, Size: 1339 bytes --]

The most  pvops kernel  2.6.32.27 works fine with X  under Xen 4.0.1 ( Video Radeon HD 4650)  with kms enabled on following systems  :-

1. Ubuntu 10.04  Server + Ubuntu Desktop
2. Ubuntu 10.10  Server + Ubuntu Desktop
3. Fedora 14
4. Scientific Linux 6 Server ALPHA 3 ( RHEL 6 Server)

What Video Controller you have been using ?

Boris.

--- On Thu, 1/6/11, Gerry Reno <greno@verizon.net> wrote:

From: Gerry Reno <greno@verizon.net>
Subject: Re: [Xen-devel] 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85
To: xen-devel@lists.xensource.com
Date: Thursday, January 6, 2011, 12:50 PM

On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
> In response to Michael Young post fedora-xen list :-
>
> > There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
>
> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
> Works as vanilla as well . Allows to run KVMs been built with standard
> fedora kernel.
>
> Boris.
>

Haven't tried 2.6.37 yet but I have not been able to get X running with
any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
stack trace.

-Gerry


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel



      

[-- Attachment #1.2: Type: text/html, Size: 1952 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Gerry Reno]

On 01/06/2011 12:58 PM, Boris Derzhavets wrote:
> The most  pvops kernel  2.6.32.27 works fine with X  under Xen 4.0.1 (
> Video Radeon HD 4650)  with kms enabled on following systems  :-
>
> 1. Ubuntu 10.04  Server + Ubuntu Desktop
> 2. Ubuntu 10.10  Server + Ubuntu Desktop
> 3. Fedora 14
> 4. Scientific Linux 6 Server ALPHA 3 ( RHEL 6 Server)
>
> What Video Controller you have been using ?
>
> Boris.
>

01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3300
Graphics
        Subsystem: Giga-byte Technology Device d000
        Flags: bus master, fast devsel, latency 0, IRQ 27
        Memory at d0000000 (32-bit, prefetchable) [size=256M]
        I/O ports at ee00 [size=256]
        Memory at fdfe0000 (32-bit, non-prefetchable) [size=64K]
        Memory at fde00000 (32-bit, non-prefetchable) [size=1M]
        Capabilities: [50] Power Management version 3
        Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit+
Queue=0/0 Enable+
        Kernel driver in use: fglrx_pci
        Kernel modules: fglrx, radeon


-Gerry

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Ian Campbell]

On Thu, 2011-01-06 at 18:02 +0000, Gerry Reno wrote: 
> On 01/06/2011 12:58 PM, Boris Derzhavets wrote:
> > The most  pvops kernel  2.6.32.27 works fine with X  under Xen 4.0.1 (
> > Video Radeon HD 4650)  with kms enabled on following systems  :-
> >
> > 1. Ubuntu 10.04  Server + Ubuntu Desktop
> > 2. Ubuntu 10.10  Server + Ubuntu Desktop
> > 3. Fedora 14
> > 4. Scientific Linux 6 Server ALPHA 3 ( RHEL 6 Server)
> >
> > What Video Controller you have been using ?
> >
> > Boris.
> >
> 
> 01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3300
> Graphics
>         Subsystem: Giga-byte Technology Device d000
>         Flags: bus master, fast devsel, latency 0, IRQ 27
>         Memory at d0000000 (32-bit, prefetchable) [size=256M]
>         I/O ports at ee00 [size=256]
>         Memory at fdfe0000 (32-bit, non-prefetchable) [size=64K]
>         Memory at fde00000 (32-bit, non-prefetchable) [size=1M]
>         Capabilities: [50] Power Management version 3
>         Capabilities: [a0] Message Signalled Interrupts: Mask- 64bit+
> Queue=0/0 Enable+
>         Kernel driver in use: fglrx_pci

fglrx is the ATI binary only/proprietary driver? I think that's your
problem. ATI are the only ones who can fix that to work with Xen.

The open source radeon driver does work just fine for me on 2.6.32 and
has done for some time. I don't believe the required fixes are in 2.6.37
though, although they might be queued for 2.6.38.

Ian.

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Pasi Kärkkäinen]

On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
> > In response to Michael Young post fedora-xen list :-
> >
> > > There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
> >
> > Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
> > Works as vanilla as well . Allows to run KVMs been built with standard
> > fedora kernel.
> >
> > Boris.
> >
> 
> Haven't tried 2.6.37 yet but I have not been able to get X running with
> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
> stack trace.
> 

Which kernels did you try?
Can you post the stack trace? 

-- Pasi

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Gerry Reno]

On 01/06/2011 01:00 PM, Pasi Kärkkäinen wrote:
> On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
>   
>> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
>>     
>>> In response to Michael Young post fedora-xen list :-
>>>
>>>       
>>>> There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
>>>>         
>>> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
>>> Works as vanilla as well . Allows to run KVMs been built with standard
>>> fedora kernel.
>>>
>>> Boris.
>>>
>>>       
>> Haven't tried 2.6.37 yet but I have not been able to get X running with
>> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
>> stack trace.
>>
>>     
> Which kernels did you try?
> Can you post the stack trace? 
>
> -- Pasi
>   

I'm using this machine currently so I can't reboot it to get a stack
trace.  Too much work open.  It's probably the same stack trace that
Boris got.
I tested using some 2.6.31 and 2.6.32 kernels.

-Gerry

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Pasi Kärkkäinen]

On Thu, Jan 06, 2011 at 01:12:26PM -0500, Gerry Reno wrote:
> On 01/06/2011 01:00 PM, Pasi Kärkkäinen wrote:
> > On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
> >   
> >> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
> >>     
> >>> In response to Michael Young post fedora-xen list :-
> >>>
> >>>       
> >>>> There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
> >>>>         
> >>> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
> >>> Works as vanilla as well . Allows to run KVMs been built with standard
> >>> fedora kernel.
> >>>
> >>> Boris.
> >>>
> >>>       
> >> Haven't tried 2.6.37 yet but I have not been able to get X running with
> >> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
> >> stack trace.
> >>
> >>     
> > Which kernels did you try?
> > Can you post the stack trace? 
> >
> > -- Pasi
> >   
> 
> I'm using this machine currently so I can't reboot it to get a stack
> trace.  Too much work open.  It's probably the same stack trace that
> Boris got.
> I tested using some 2.6.31 and 2.6.32 kernels.
> 

2.6.31 tree is outdated and it doesn't have the gfx related fixes in it.

Also only the pretty recent 2.6.32.x versions have the drm/kms/ttm 
gfx related fixes (by konrad) in it.

So if it was many months ago when you tested, you should re-test.

-- Pasi

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Boris Derzhavets]

[-- Attachment #1.1: Type: text/plain, Size: 1623 bytes --]

>I tested using some 2.6.31 and 2.6.32 kernels.
> -Gerry

Would suggest to try the most recent 2.6.32.27 pvops via Jeremy Git Repos

Boris.

--- On Thu, 1/6/11, Gerry Reno <greno@verizon.net> wrote:

From: Gerry Reno <greno@verizon.net>
Subject: Re: [Xen-devel] 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85
To: xen-devel@lists.xensource.com
Date: Thursday, January 6, 2011, 1:12 PM

On 01/06/2011 01:00 PM, Pasi Kärkkäinen wrote:
> On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
>   
>> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
>>     
>>> In response to Michael Young post fedora-xen list :-
>>>
>>>       
>>>> There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
>>>>         
>>> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
>>> Works as vanilla as well . Allows to run KVMs been built with standard
>>> fedora kernel.
>>>
>>> Boris.
>>>
>>>       
>> Haven't tried 2.6.37 yet but I have not been able to get X running with
>> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
>> stack trace.
>>
>>     
> Which kernels did you try?
> Can you post the stack trace? 
>
> -- Pasi
>   

I'm using this machine currently so I can't reboot it to get a stack
trace.  Too much work open.  It's probably the same stack trace that
Boris got.
I tested using some 2.6.31 and 2.6.32 kernels.

-Gerry


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel



      

[-- Attachment #1.2: Type: text/html, Size: 2537 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Konrad Rzeszutek Wilk]

On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
> > In response to Michael Young post fedora-xen list :-
> >
> > > There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
> >
> > Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
> > Works as vanilla as well . Allows to run KVMs been built with standard
> > fedora kernel.
> >
> > Boris.
> >
> 
> Haven't tried 2.6.37 yet but I have not been able to get X running with
> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
> stack trace.

Can you try:


git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git

devel/fix-amd-bootup

branch please?

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

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Gerry Reno]

On 01/06/2011 02:10 PM, Konrad Rzeszutek Wilk wrote:
> On Thu, Jan 06, 2011 at 12:50:16PM -0500, Gerry Reno wrote:
>   
>> On 01/06/2011 12:42 PM, Boris Derzhavets wrote:
>>     
>>> In response to Michael Young post fedora-xen list :-
>>>
>>>       
>>>> There is a 2.6.37 kernel to test (2.6.37-1.xendom0.fc15)
>>>>         
>>> Works for me under Xen 4.0.1 , but attempt to "startx" causes stack trace.
>>> Works as vanilla as well . Allows to run KVMs been built with standard
>>> fedora kernel.
>>>
>>> Boris.
>>>
>>>       
>> Haven't tried 2.6.37 yet but I have not been able to get X running with
>> any of the dom0 kernels while using Ubuntu Lucid.  I've always gotten a
>> stack trace.
>>     
> Can you try:
>
>
> git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git
>
> devel/fix-amd-bootup
>
> branch please?
>
>   
>
Konrad,
    Thanks.  Yes, I'll try it but it won't be for a couple days.

Gerry

2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Pasi Kärkkäinen]

Hello,

First of all.. big thanks to everyone involved in upstreaming Xen dom0 bits! 

I just tried upstream pvops 2.6.37 dom0 kernel on one of my machines, and it worked
out-of-the-box ! Even nouveau kms/modesetting worked just fine.

There was non-fatal BUG/traceback though:

Full dmesg log available at: http://pasik.reaktio.net/xen/logs/dmesg-2.6.37-1.xendom0.fc15.x86_64-kms-modesetting.txt

[    9.930809] BUG: sleeping function called from invalid context at kernel/mutex.c:85
[    9.931616] in_atomic(): 0, irqs_disabled(): 1, pid: 572, name: kworker/0:2
[    9.932403] Pid: 572, comm: kworker/0:2 Not tainted 2.6.37-1.xendom0.fc15.x86_64 #1
[    9.933188] Call Trace:
[    9.933973]  [<ffffffff81042005>] __might_sleep+0xeb/0xf0
[    9.934758]  [<ffffffff81476e9e>] mutex_lock+0x24/0x50
[    9.935533]  [<ffffffff810abe14>] free_desc+0x2e/0x5f
[    9.936296]  [<ffffffff810abe80>] irq_free_descs+0x3b/0x71
[    9.937064]  [<ffffffff812ac0c8>] irq_free_desc+0x13/0x15
[    9.937812]  [<ffffffff812aca85>] xen_destroy_irq+0xe1/0xf5
[    9.938545]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
[    9.939270]  [<ffffffff813b70b9>] xen_teardown_msi_irq+0xe/0x10
[    9.939982]  [<ffffffff81255630>] default_teardown_msi_irqs+0x57/0x80
[    9.940674]  [<ffffffff812550f6>] free_msi_irqs+0x8b/0xe9
[    9.941352]  [<ffffffff81255ff4>] pci_disable_msix+0x35/0x39
[    9.942030]  [<ffffffffa005111a>] xhci_cleanup_msix+0x31/0x51 [xhci_hcd]
[    9.942705]  [<ffffffffa0051d35>] xhci_suspend+0x199/0x1ab [xhci_hcd]
[    9.943370]  [<ffffffffa00565ce>] xhci_pci_suspend+0x23/0x25 [xhci_hcd]
[    9.944029]  [<ffffffff81350355>] suspend_common+0x65/0xe2
[    9.944677]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
[    9.945318]  [<ffffffff813503eb>] hcd_pci_runtime_suspend+0x19/0x43
[    9.945955]  [<ffffffff8124549b>] pci_pm_runtime_suspend+0x58/0xea
[    9.946587]  [<ffffffff81245443>] ? pci_pm_runtime_suspend+0x0/0xea
[    9.947221]  [<ffffffff812ef7a0>] rpm_suspend+0x285/0x3c5
[    9.947836]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
[    9.948445]  [<ffffffff8107d31f>] ? arch_local_irq_restore+0xb/0xd
[    9.949051]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
[    9.949648]  [<ffffffff8107d32c>] ? arch_local_irq_disable+0xb/0xd
[    9.950230]  [<ffffffff812f0357>] __pm_runtime_suspend+0x51/0x6d
[    9.950801]  [<ffffffff8124515b>] ? pci_pm_runtime_idle+0x0/0x4d
[    9.951359]  [<ffffffff812451a2>] pci_pm_runtime_idle+0x47/0x4d
[    9.951906]  [<ffffffff812efaa5>] rpm_idle+0x14d/0x18f
[    9.952445]  [<ffffffff812f042e>] ? pm_runtime_work+0x0/0xa0
[    9.952970]  [<ffffffff812f0499>] pm_runtime_work+0x6b/0xa0
[    9.953490]  [<ffffffff810665ab>] process_one_work+0x18c/0x285
[    9.954010]  [<ffffffff810673b2>] worker_thread+0x104/0x1a4
[    9.954526]  [<ffffffff810672ae>] ? worker_thread+0x0/0x1a4
[    9.955038]  [<ffffffff8106a9e3>] kthread+0x82/0x8a
[    9.955545]  [<ffffffff8100bae4>] kernel_thread_helper+0x4/0x10
[    9.956060]  [<ffffffff8100aee3>] ? int_ret_from_sys_call+0x7/0x1b
[    9.956575]  [<ffffffff814782dd>] ? retint_restore_args+0x5/0x6
[    9.957088]  [<ffffffff8100bae0>] ? kernel_thread_helper+0x0/0x10


-- Pasi

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Pasi Kärkkäinen]

On Thu, Jan 06, 2011 at 07:36:28PM +0200, Pasi Kärkkäinen wrote:
> Hello,
> 
> First of all.. big thanks to everyone involved in upstreaming Xen dom0 bits! 
> 
> I just tried upstream pvops 2.6.37 dom0 kernel on one of my machines, and it worked
> out-of-the-box ! Even nouveau kms/modesetting worked just fine.
> 
> There was non-fatal BUG/traceback though:
> 
> Full dmesg log available at: http://pasik.reaktio.net/xen/logs/dmesg-2.6.37-1.xendom0.fc15.x86_64-kms-modesetting.txt
> 

Additional data.. latest xen/stable-2.6.32.x dom0 kernel (2.6.32.27) doesn't have this BUG/traceback.

-- Pasi


> [    9.930809] BUG: sleeping function called from invalid context at kernel/mutex.c:85
> [    9.931616] in_atomic(): 0, irqs_disabled(): 1, pid: 572, name: kworker/0:2
> [    9.932403] Pid: 572, comm: kworker/0:2 Not tainted 2.6.37-1.xendom0.fc15.x86_64 #1
> [    9.933188] Call Trace:
> [    9.933973]  [<ffffffff81042005>] __might_sleep+0xeb/0xf0
> [    9.934758]  [<ffffffff81476e9e>] mutex_lock+0x24/0x50
> [    9.935533]  [<ffffffff810abe14>] free_desc+0x2e/0x5f
> [    9.936296]  [<ffffffff810abe80>] irq_free_descs+0x3b/0x71
> [    9.937064]  [<ffffffff812ac0c8>] irq_free_desc+0x13/0x15
> [    9.937812]  [<ffffffff812aca85>] xen_destroy_irq+0xe1/0xf5
> [    9.938545]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
> [    9.939270]  [<ffffffff813b70b9>] xen_teardown_msi_irq+0xe/0x10
> [    9.939982]  [<ffffffff81255630>] default_teardown_msi_irqs+0x57/0x80
> [    9.940674]  [<ffffffff812550f6>] free_msi_irqs+0x8b/0xe9
> [    9.941352]  [<ffffffff81255ff4>] pci_disable_msix+0x35/0x39
> [    9.942030]  [<ffffffffa005111a>] xhci_cleanup_msix+0x31/0x51 [xhci_hcd]
> [    9.942705]  [<ffffffffa0051d35>] xhci_suspend+0x199/0x1ab [xhci_hcd]
> [    9.943370]  [<ffffffffa00565ce>] xhci_pci_suspend+0x23/0x25 [xhci_hcd]
> [    9.944029]  [<ffffffff81350355>] suspend_common+0x65/0xe2
> [    9.944677]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
> [    9.945318]  [<ffffffff813503eb>] hcd_pci_runtime_suspend+0x19/0x43
> [    9.945955]  [<ffffffff8124549b>] pci_pm_runtime_suspend+0x58/0xea
> [    9.946587]  [<ffffffff81245443>] ? pci_pm_runtime_suspend+0x0/0xea
> [    9.947221]  [<ffffffff812ef7a0>] rpm_suspend+0x285/0x3c5
> [    9.947836]  [<ffffffff810073df>] ? xen_restore_fl_direct_end+0x0/0x1
> [    9.948445]  [<ffffffff8107d31f>] ? arch_local_irq_restore+0xb/0xd
> [    9.949051]  [<ffffffff81477d3f>] ? _raw_spin_unlock_irqrestore+0x17/0x19
> [    9.949648]  [<ffffffff8107d32c>] ? arch_local_irq_disable+0xb/0xd
> [    9.950230]  [<ffffffff812f0357>] __pm_runtime_suspend+0x51/0x6d
> [    9.950801]  [<ffffffff8124515b>] ? pci_pm_runtime_idle+0x0/0x4d
> [    9.951359]  [<ffffffff812451a2>] pci_pm_runtime_idle+0x47/0x4d
> [    9.951906]  [<ffffffff812efaa5>] rpm_idle+0x14d/0x18f
> [    9.952445]  [<ffffffff812f042e>] ? pm_runtime_work+0x0/0xa0
> [    9.952970]  [<ffffffff812f0499>] pm_runtime_work+0x6b/0xa0
> [    9.953490]  [<ffffffff810665ab>] process_one_work+0x18c/0x285
> [    9.954010]  [<ffffffff810673b2>] worker_thread+0x104/0x1a4
> [    9.954526]  [<ffffffff810672ae>] ? worker_thread+0x0/0x1a4
> [    9.955038]  [<ffffffff8106a9e3>] kthread+0x82/0x8a
> [    9.955545]  [<ffffffff8100bae4>] kernel_thread_helper+0x4/0x10
> [    9.956060]  [<ffffffff8100aee3>] ? int_ret_from_sys_call+0x7/0x1b
> [    9.956575]  [<ffffffff814782dd>] ? retint_restore_args+0x5/0x6
> [    9.957088]  [<ffffffff8100bae0>] ? kernel_thread_helper+0x0/0x10
> 
> 
> -- Pasi
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel

Re: 2.6.37 dom0 non-fatal BUG: sleeping function called from invalid context at kernel/mutex.c:85

[Pasi Kärkkäinen]

On Thu, Jan 06, 2011 at 08:23:14PM +0200, Pasi Kärkkäinen wrote:
> On Thu, Jan 06, 2011 at 07:36:28PM +0200, Pasi Kärkkäinen wrote:
> > Hello,
> > 
> > First of all.. big thanks to everyone involved in upstreaming Xen dom0 bits! 
> > 
> > I just tried upstream pvops 2.6.37 dom0 kernel on one of my machines, and it worked
> > out-of-the-box ! Even nouveau kms/modesetting worked just fine.
> > 
> > There was non-fatal BUG/traceback though:
> > 
> > Full dmesg log available at: http://pasik.reaktio.net/xen/logs/dmesg-2.6.37-1.xendom0.fc15.x86_64-kms-modesetting.txt
> > 
> 
> Additional data.. latest xen/stable-2.6.32.x dom0 kernel (2.6.32.27) doesn't have this BUG/traceback.
>

Sorry for the noise.. it seems this is a bug in the xhci driver in 2.6.37.
It's already fixed upstream but the fix is not in 2.6.37.

I noticed the bugreport from lkml, it seems to happen also on baremetal for other people.
(it didn't for me).

Well.. at least it was not xen dom0 specific bug :)

-- Pasi


From: Greg KH <gregkh@suse.de>
To: Robert Hancock <hancockrwd@gmail.com>
Cc: Sarah Sharp <sarah.a.sharp@linux.intel.com>,
	Linux-usb <linux-usb@vger.kernel.org>,
	linux-kernel <linux-kernel@vger.kernel.org>
Date: Wed, 5 Jan 2011 16:30:03 -0800
Subject: Re: Sleeping function called from invalid context in 2.6.37 xhci_hcd

On Wed, Jan 05, 2011 at 06:22:03PM -0600, Robert Hancock wrote:
> Just saw this on booting up on the latest Fedora Rawhide kernel
> based on 2.6.37. Looks like we were trying to do a runtime suspend
> on the xHCI USB controller and it disabled MSI-X while interrupts
> were disabled?


Known issue, patches came in too late to make it for .37, hopefully will
show up in .38-rc and then get backported to .37.  They are on the
linux-usb mailing list if you need them now.

Hope this helps,

greg k-h