[RFC PATCH v0.1] net driver: mpc52xx fec

[Juergen Beisert]

On Thursday 27 September 2007 20:43, Scott Wood wrote:
> Jon Smirl wrote:
> > The call to msleep() is inside a block protected with
> >
> > :#define in_interrupt()          (irq_count())
> >
> > if (!in_interrupt)
> >
> > The stack trace looks like it is in a timer interrupt so shouldn't
> > irq_count be non-zero?
> > Could there be some lack of coordination on irq_count and the timer
> > tick with the preempt patch applied?  Or does irq_count() not count
> > soft irqs?
> >
> > (!in_interrupt) may be the wrong way to protect this code.
>
> I think in_atomic() is what you want.

I tried with in_atomic(). The BUG report is gone, but the problem still 
exists. 

While network stress testing: 

[...]
NETDEV WATCHDOG: eth0: transmit timed out
net eth0: transmit timed out
net eth0: queues didn't drain
net eth0:   tx: index: 35, outdex: 36
net eth0:   rx: index: 24, outdex: 25
PHY: f0003000:00 - Link is Down
PHY: f0003000:00 - Link is Up - 100/Full

The link is up again, but any connection is dead (no answers to ping etc.). 
But the serial console is still working. I'm not sure if the RT-Preempt patch 
*causes* this behavior or only *discover* it. Any idea?

Juergen
-- 
Dipl.-Ing. Juergen Beisert | http://www.pengutronix.de
 Pengutronix - Linux Solutions for Science and Industry
    Handelsregister: Amtsgericht Hildesheim, HRA 2686
         Vertretung Sued/Muenchen, Germany
   Phone: +49-8766-939 228 |  Fax: +49-5121-206917-9