Re: ifconfig on re0 panics Toshiba Satellite M115-S3094 with RealTek 8101E chipset

[Joe Talbott <josepht@xxxxxxxxxx>]

On Fri, Dec 15, 2006 at 08:14:54PM +0800, Sepherosa Ziehau wrote:
> On 12/15/06, Joe Talbott <josepht@cstone.net> wrote:
> >On Thu, Dec 14, 2006 at 08:22:06PM +0800, Sepherosa Ziehau wrote:
> >> On 12/14/06, Joe Talbott <josepht@cstone.net> wrote:
> >> >Okay.  I've learned a little more about this.  If I power down the
> >> >laptop and leave it off for a few minutes then power it up, login and
> >> >ifconfig re0 ... it panics.  I got a core dump with 'call dumpsys',
> >> >'panic' didn't work.  Next I entered 'reset' at the db> prompt.  I
> >> >logged in again and ifconfig re0 ... and it panics again.  Same db>
> >> >stuff as before and reset.  This time everything hangs after ifconfig
> >> >re0 ....  Power cycle and then I can ifconfig re0 ... and pass
> >> >traffic.  I did this all again (i.e. 2 panics, a hang, then open
> >> >sesame) to verify the process.  I have uploaded the kernel and vmcore
> >> >files for the panics onto leaf.dragonflybsd.org:/home/josepht/crash/.
> >>
> >> Two panics are in the same place (outl), so I suspect this may be the
> >> effect of writing to some reserved hardware registers.  Please test
> >> following patch:
> >> http://leaf.dragonflybsd.org/~sephe/re2.diff
> >>
> >
> >Same as before... Cold Boot -> Panic -> Panic -> freeze -> Network UP.
> >kernel.[78] and vmcore.[78] are here:
> >
> >leaf.dragonflybsd.org:/home/josepht/crash/
> 
> Kernel paniced when it wrote a valid hardware register with a valid
> value :-(, so I suspect something is wrong before this "valid"
> writing.  Please test this patch:
> http://leaf.dragonflybsd.org/~sephe/re3.diff
> 

Same again.  {vmcore,kernel}.10 are here:

leaf.dragonflybsd.org:/home/josepht/crash/

I've put some if_printf's in the re_init function to try to narrow it
down but haven't managed to figure anything out yet.

Joe