4.10-rc1 on Nokia N900: regression, WARN_ON() omap_l3_smx.c

4.10-rc1 on Nokia N900: regression, WARN_ON() omap_l3_smx.c

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 666 bytes --]

Hi!

I forgot I had v4.10-rc1 running, and now I got warning on all the
consoles (hand-copied).


Unhandled fault: external abort on non-linefetch (0x1028) at
0xfa0ab060
...
Comm: kworker/0:0 Not tainted.
Workqueue: events musb_irq_work
...
PC is at musb_default_readb().
...
WARNING: CPU: 0 ... at drivers/bus/omap_l3_smx.c:166
omap3_l3_app_irq+0xcc/...
Tainted: GDW.

I do have patches to allow nfsroot over usb. But they worked ok in
v4.9... Does anyone see it, too?

Best regards,

									Pavel

-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: 4.10-rc1 on Nokia N900: regression, WARN_ON() omap_l3_smx.c

[Tony Lindgren]

* Pavel Machek <pavel@ucw.cz> [161229 15:14]:
> Hi!
> 
> I forgot I had v4.10-rc1 running, and now I got warning on all the
> consoles (hand-copied).
> 
> 
> Unhandled fault: external abort on non-linefetch (0x1028) at
> 0xfa0ab060
> ...
> Comm: kworker/0:0 Not tainted.
> Workqueue: events musb_irq_work
> ...
> PC is at musb_default_readb().
> ...

This means the clocks are not enabled at that point.

> WARNING: CPU: 0 ... at drivers/bus/omap_l3_smx.c:166
> omap3_l3_app_irq+0xcc/...
> Tainted: GDW.

If you comment out postcore_initcall_sync(omap3_l3_init);
in drivers/bus/omap_l3_smx.c you'll see the proper stack
trace instead of the l3 interrupt trace. The system will
hang at that point most likely.

> I do have patches to allow nfsroot over usb. But they worked ok in
> v4.9... Does anyone see it, too?

Hmm not much has changed since v4.9. Are you sure you
had v4.9 or some earlier v4.9-rc version?

Regards,

Tony

Re: 4.10-rc1 on Nokia N900: regression, WARN_ON() omap_l3_smx.c

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 1359 bytes --]

Hi!

> > I forgot I had v4.10-rc1 running, and now I got warning on all the
> > consoles (hand-copied).
> > 
> > 
> > Unhandled fault: external abort on non-linefetch (0x1028) at
> > 0xfa0ab060
> > ...
> > Comm: kworker/0:0 Not tainted.
> > Workqueue: events musb_irq_work
> > ...
> > PC is at musb_default_readb().
> > ...
> 
> This means the clocks are not enabled at that point.
> 
> > WARNING: CPU: 0 ... at drivers/bus/omap_l3_smx.c:166
> > omap3_l3_app_irq+0xcc/...
> > Tainted: GDW.
> 
> If you comment out postcore_initcall_sync(omap3_l3_init);
> in drivers/bus/omap_l3_smx.c you'll see the proper stack
> trace instead of the l3 interrupt trace. The system will
> hang at that point most likely.
> 
> > I do have patches to allow nfsroot over usb. But they worked ok in
> > v4.9... Does anyone see it, too?
> 
> Hmm not much has changed since v4.9. Are you sure you
> had v4.9 or some earlier v4.9-rc version?

I believe v4.9 works.

But... this may be tricky to reproduce. It happened once so
far... after I reconnected N900 to computer when it was running for a
while. I do have USB cable with power meter on it, thus "flakey". Lets
see if it reappears...
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: 4.10-rc1 on Nokia N900: regression, WARN_ON() omap_l3_smx.c

[Pavel Machek]

[-- Attachment #1: Type: text/plain, Size: 1382 bytes --]

Hi!

> * Pavel Machek <pavel@ucw.cz> [161229 15:14]:
> > Hi!
> > 
> > I forgot I had v4.10-rc1 running, and now I got warning on all the
> > consoles (hand-copied).
> > 
> > 
> > Unhandled fault: external abort on non-linefetch (0x1028) at
> > 0xfa0ab060
> > ...
> > Comm: kworker/0:0 Not tainted.
> > Workqueue: events musb_irq_work
> > ...
> > PC is at musb_default_readb().
> > ...
> 
> This means the clocks are not enabled at that point.
> 
> > WARNING: CPU: 0 ... at drivers/bus/omap_l3_smx.c:166
> > omap3_l3_app_irq+0xcc/...
> > Tainted: GDW.
> 
> If you comment out postcore_initcall_sync(omap3_l3_init);
> in drivers/bus/omap_l3_smx.c you'll see the proper stack
> trace instead of the l3 interrupt trace. The system will
> hang at that point most likely.
> 
> > I do have patches to allow nfsroot over usb. But they worked ok in
> > v4.9... Does anyone see it, too?
> 
> Hmm not much has changed since v4.9. Are you sure you
> had v4.9 or some earlier v4.9-rc version?

It happened again in -rc5. I have full dmesg this time.

I started new thread with "

Subject: v4.9 to v4.10 regression: oops when USB cable is plugged in.

". It does no longer look completely scary.
 
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]