[pvrusb2] Kernel oops?

Mike Isely isely at isely.net
Tue Sep 20 22:27:07 CDT 2005


Another thing:

There is more to the pvrusb2 log that's missing.  I need to match up all
the buffer addresses being reported when the buffers are created.  There
are normally 32 of them but what you posted only starts with the 16th
buffer - I can't see anything for the first 16 buffers.  This is immensely
important because it will tell us if that bad buffer is from corrupted
memory, deleted memory, or if that is not really a buffer but a wild
pointer.

Oh heck.  You say there's 60MB of log data?  I've got a 1.5Mb DSL downlink
here.  Can you just put it up and I'll suck it down?  Then I'll be able to
match up the USB addresses as well.

  -Mike


On Tue, 20 Sep 2005, Mike Isely wrote:

>
> This line is at the top of your kernel log just before the oops:
>
>   Sep 21 02:39:38 t42p ehci_hcd 0000:00:1d.7: devpath 4.1 ep4in 3strikes
>
> What the $##@! is that?
>
> Can you check for me that the USB address listed there matches the pvrusb2
> device's location?  Check the connection messages logged when you first
> plug it.  I suspect it is the same.  If so, I'm going to go digging
> through the kernel sources to see what is going on here...
>
> Also, I'd like you send me your pvrusb2.ko binary module.  Just send it to
> me directly.  It may be useful to cross-reference back some EIP values.
>
> Still looking...
>
>   -Mike
>
>
>
> On Tue, 20 Sep 2005, Mike Isely wrote:
>
> >
> > Thanks for catching this!  I've grabbed the logs, and will investigate.
> >
> >   -Mike
> >
> >
> > On Wed, 21 Sep 2005, Arnaud Boulan wrote:
> >
> > > Le Dimanche 18 Septembre 2005 20:08, Mike Isely a écrit :
> > >
> > > > Cute.
> > > >
> > > > Haven't seen any reports yet of this problem happening again,
> > > > unfortunately....
> > > >
> > > >   -Mike
> > >
> > > Yippeeeeeeeee! It happened again!! :o)
> > >
> > > Sep 21 02:39:38 t42p pvrusb2 Buffer cd717f70 is bad
> > > at /root/admin/drivers/pvr/pvrusb2-mci-20050911/driver/pvrusb2-io.c:211
> > > Sep 21 02:39:38 t42p pvrusb2 buffer BadSig cd717f70 state=unknown
> > > id=-706763648 status=1 stream=00004000 purb=0000001d sig=0xd5dfa200
> > > Sep 21 02:39:38 t42p kernel BUG
> > > at /root/admin/drivers/pvr/pvrusb2-mci-20050911/driver/pvrusb2-io.c:211!
> > >
> > > Here are my logs: (not attached because it's too big for the list)
> > >
> > > http://www.boulan.net/pvr/oops1/pvrusb2.log-443-09-21.gz
> > > http://www.boulan.net/pvr/oops1/kern.log-09-21
> > >
> > > I hope it will help...
> > > Although, i can see no logs about buffer "cd717f70" since last time my pvrusb2
> > > was plugged. I have some logs about "stream=cd717f80", maybe it's related?
> > > I keep the whole logs (~60MB compressed :) in case you need it.
> > >
> > > Arnaud
> > >
> > > _______________________________________________
> > > pvrusb2 mailing list
> > > pvrusb2 at isely.net
> > > http://www.isely.net/cgi-bin/mailman/listinfo/pvrusb2
> > >
> >
> >
>
>

-- 
                        |         Mike Isely          |     PGP fingerprint
     Spammers Die!!     |                             | 03 54 43 4D 75 E5 CC 92
                        |   isely @ pobox (dot) com   | 71 16 01 E2 B5 F5 C1 E8
                        |                             |


More information about the pvrusb2 mailing list