[eepro100] "card reports no resources" fixes?

[Khyron]

Okay, noted.

The driver in my pristine 2.2.18 sources is:

static const char *version =
"eepro100.c:v1.09j-t 9/29/99 Donald Becker
http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html\n"
"eepro100.c: $Revision: 1.20.2.10 $ 2000/05/31 Modified by Andrey V.
Savochkin <saw@saw.sw.com.sg> and others\n"
"eepro100.c: VA Linux custom, Dragan Stancevic <visitor@valinux.com>
2000/11/15\n";

Becker 1.09j-t/Savochkin 1.20.2.10

I always see these errors on highly loaded (network) systems.
Never when idle. But the intermittent nature is a biznatch.

For example, using 2 clusters of identical machines running
CA810E desktop boards from Intel (and onboard EEPro100s), we
see this incessantly on 1 cluster yet sparsely if ever on
the other. Both are running the *exact* same kernel, driver,
application software. Everything is identical except one has
the errors so bad no work can get done and the other isn't
fazed. In fact, one cluster ran for 2 weeks under regular
high network load (backups) and only failed after 2 weeks
of being pounded on.

The above driver (Becker 1.09j-t/Savochkin 1.20.2.10) has been
problematic using 2.2.15pre16, as has been the Intel e100
1.5.5 on 2.2.18.

If memory serves, you said...

> Khyron wrote:
> >
> > My question is, for everyone who has encountered this problem,
> > how did you resolve it? Ideally, if you could explain your
> > configuration as well (hardware, kernel version w/ patches, etc.)
> > that would be helpful.
>
> ive had that problem off and on for almost a year. i have never
> resolved it. luckily it has been verry intermittant. all of my systems
> are very low traffic and the problem only ever occured when booting
> the system. since i changed motherboards (Asus P3V4X P3800 to Tyan KT133A
> based board with Athlon) the problem (sofar) has been non existant.
> i also did not have the problem on my BP6(before i upgraded to
> the asus) last year with dual celerons(all using the same physical
> NIC). I have an Asus P3V4X at work as well(my main desktop), because
> of the eepro troubles i was sure to get a 3COM 3C905C instead, works
> much better for me.
>
> in short .......from what ive read on this list in the last ~6 months
> nobody has come up with a complete fix. i read every message that
> comes accross the list (i have some other issues with eepro cards
> too..). wish there was a way to better isolate the problem.
>
> a workaround, in my case anyways is to restart networking(on debian:
> /etc/init.d/networking restart) fixed it for me. that doesn't help
> when your trying to boot and mounting nfs filesystems automatically
> though it can take a good 20 minutes before i get a login screen
> while the eepro spits out errors.
>
> i've never had the error occur after the system was already up and
> running though, maybe others have.
>
> haven't seen the error in quite a while..probably since april at
> the most recent. between now and then ive been running linux 2.2.18
> and 2.2.19 with the driver thats integrated into the kernel haven't
> been patching the eepro on that system that had the problems.
>
> if anyone knows why having subinterfaces on (what seems to be) a
> dual port eepro100 NIC works on eth0, but crashes the system hard
> on eth1 lemme know..id like to be able to have subinterfaces on
> both! (i can reference my original posts if needed).
>
> nate
>
> --
> Nate Amsden
> System Administrator
> GraphOn
> http://www.graphon.com
>
> _______________________________________________
> eepro100 mailing list
> eepro100@scyld.com
> http://www.scyld.com/mailman/listinfo/eepro100
>

////////////////////////////////////////////////////////////////////
Khyron					    mailto:khyron@khyron.com
Key fingerprint = 53BB 08CA 6A4B 8AF8 DF9B  7E71 2D20 AD30 6684 E82D
			"Drama free in 2001!"
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\