[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: tulip.c problems with 2.1.96+ versions
- From: "Wes Bauske" <wsb paralleldata com>
- To: axp-list redhat com
- Subject: Re: tulip.c problems with 2.1.96+ versions
- Date: Tue, 05 May 1998 00:06:43 -0500
Matti Aarnio wrote:
> Hello,
>
> At my PWS433a Alpha machine I have two Tulip driven network
> interfaces. One builtin 21142 "card", and one SMC 9332BDT(?)
> card with 21140 chip. ("?" because its been awhile since I
> looked at its labels...)
>
> Everything works fine with kernel 2.1.92, but not so since
> 2.1.96 (I didn't try any of ..93 to ..95, so I don't know
> when the change came).
>
> Well, TRANSMIT does not work at 2.1.96 and latter, but
> receive sure works. Running tcpdump does yield received
> frames from the network.
>
> If I install the module with option "debug=3", then I get
> following status reports to dmesg once per second:
>
> eth0: Transmit error, Tx status 7fffbc84.
> eth0: Transmit error, Tx status 7fffbc84.
> eth0: Transmit error, Tx status 7fffbc84.
> eth0: Transmit error, Tx status 7fffbc84.
>
> All of the base-address data et.al. appear to be same as
> with 2.1.92, which works.
>
Actually, it is working, at least under 2.1.9[78]. The statistics are just
screwy. I use it that way without problems and have a telent session
across it right now. I guess somehow the kernel statistics structure
and what tulip.c thinks it is are not matching?? Maybe someone
messed up the structure and declared something an int instead of
long?? Must be in the kernel, not tulip.c cause I tried my tulip.c
from an older kernel and it does the same thing. In fact I tried
three different tulip.c versions and they all show bad stats but
work OK.
Wes
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
[]