On Tue, Feb 17, 1998 at 03:10:43AM -0800, Mike Smith wrote:
> > This might be a clue for the strange ED overwrites (or it might be
> > something completely unrelated :-)
>
> It looks kinda interesting, at any rate:
[code snippet deleted]
>
> Weird. The destination should be on the stack - are you comfortable
> that you haven't overrun the end of the stack?
I'm not certain about anything right now - I started reproducing this
about an hour ago. It might even be a hardware failure (but I've
tried with different cards of the same type, and all fail, while a
Kingston Ne2000-clone works flawlessly).
> > #4 0xf01b34ae in ed_pio_readmem (sc=0xf01ed400, src=25088,
> > dst=0xefbfffc0 "'", amount=4) at machine/cpufunc.h:185
>
> In my copy of cpufunc.h, line 185 is insb(). Is this an 8-bit card?
It's a "Thrust" NE2000-clone, based on the RTL8019AS.
And in the relevant header file here line 185 is insw().
Eivind, going back to debugging.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message