Re: [parisc-linux] 715/50 hang - more info

[Richard Hirst <rhirst@linuxcare.com>]

Hi,
  I have just committed changes to drivers/net/lasi_82596.c and
drivers/gsc/lan.c.  Fairly big changes to lasi_82596.c, partly
just tidyup, partly improvements to cache handling.  It is still
not perfect but is a _lot_ better than it was.  The change to
lan.c was to use a different probe function on older h/w (like
my 715/75), where we need to swap the MPU_PORT, as per the message
from prumpf.  If the driver seems completely dead for you, try
enabling/disabling that word swap, and please let me know that
you needed to.

All success or failure reports welcome :-)

Richard


On Thu, May 18, 2000 at 08:31:56AM -0400, Chris Snow wrote:
> > I had the same problem on my 715/33 with lasi_82596.c.
> > I emailed Helge Deller who made the last changes to
> > this driver and he told me that there seems to be the
> > problem that not all interrupts from the i82596 are
> > delivered to the interrupt routine. The result of this
> > is that all 16 xmit buffers will get filled and then
> > the kernel hangs.
> 
> That's what I came up with as well.
> 
> > He also told me that he will have no time until 23rd
> > of May (approx.) to find the reason for this and to
> > correct it. So right now there is no way to use
> > NFSROOT on a 715/old.
> 
> Well, I'm going to keep playing with it over the next few days, although
> I'm not sure if I'll have any success...
> 
> If I succeed in fixing anything, I'll be sure to let everyone know.
> 
> 
> Thanks,
> 
> 
>      Chris
> 
> ---------------------------------------------------------------------------
> To unsubscribe: send e-mail to parisc-linux-request@thepuffingroup.com with
> `unsubscribe' as the subject.
> 
>