On Sat, 24 Apr 1999, Alan Cox wrote:
> > Then it must be both drivers ... I just tried ne2k-pci and couldn't
> > reproduce the mysterious crashing, so this would hint towards 3c509 and
> > eexpress being broken. BTW, 3c509 does not use 82586, but you probably
> > already knew that.
>
> We are chasing a weird problem with the 3c509 driver and SMP lockups. That
> isnt just you alone. The 82586 problem report is new. Probably nobody else
> has used that configuration
M.H.VanLeeuwen (sp?) has gotten more info (in another thread on l-k). One
processor goes el3_start_xmit() -> diasble_irq() -> synchronize_irq()
where it possibly stops in wait_on_irq(). That would make me think the
other processor is running eternally in interrupt context and the other
decides to wait for it. My head is aching of reading the 3c509.c (let
alone the interrupt code for x86), so I might be well wrong.
The details on eexpress lockup are still a bit hazy. It seems to fail in
different ways (if I have time I'll try to whip up a report).
Happy hacking,
Taneli <taneli@firmament.fi>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/