sk ethernet driver: watchdog timeout

[Palle Girgensohn]

--On onsdag, mars 17, 2004 00.21.44 +0100 "Arno J. Klaassen" 
<arno at heho.snv.jussieu.fr> wrote:

> Hello,
>
>> I have an ASUS motherboard A7V8X-E Deluxe with onboard 10/100/1000
>> Mbit/s NIC from Marvell Semiconductor.
>>
>> My problem is that it sometimes lock up with the error message
>>
>>  sk0: watchdog timeout
>
> I have a similar problem with 3Com cards on an ASUS A7N266;
> I just post in case this might be related (and in hope for
> a hint for a solution )

Hi again,

I've since this thread started tried this on more different systems, with 
exactly the same results. Anyone else experiencing this? Anything I can do 
to help fixing it?

Regards,
Palle





>> This is a fresh FreeBSD 4-STABLE system.
>
> xl0 = 3c905C-TX
> xl1 = 3c900-COMBO
> switch = Netgear FS105
>
> box netboots (using xl0) -Stable from a PPro server with same 3c905C-TX
> and works like a charm (network software development use, generating
> high network load)
>
> This weekend I pulled in an old scsi-card and disk and installed
> 5.2.1 and -current on it :
>
>  - networking through xl0 gives me about 6Kbps throughput, unworkable
>    but no freezes
>  - networking through xl1 works more or less OK as long as I stick
>    to NFS (udp-mounts), simple editing and serial makes;
>    the first "make -j X" and/or "ar srv mylib lots-of-ofiles" generates :
>
>     xl1: watchdog timeout
>
>    system does not come back from lost network, at least not for
>    the first couple of minutes
>
>    whenever I try a "cvsup -g -L 2 /etc/my-supfile" or "cvs -d faraway
> update"    I always get the same
>
>     xl1: watchdog timeout
>
>    after some dozens of seconds
>
> As said, same behaviour on 5.2.1 and -current from sunday.
> With or without acpi/apic/WITNESS does not make any difference.
> With or without CPUTYPE=athlon-xp either.
>
> Note, when I installed this system (at a co-worker) one year ago,
> I *had* to fix "AGP Aperture size" to 64M in the BIOS, otherwise
> the netboot failed.
> I just think about this, I did not test this weekend whether current
> -Stable (sic) works OK when changing this value.
>
> Thanx for your time, hope this helps to find a solution.
>
> I can bake a current kernel on another box and ask my coworker
> to test it.
>
> Regards,
>
>  Arno
> _______________________________________________
> freebsd-net at freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-net
> To unsubscribe, send any mail to "freebsd-net-unsubscribe at freebsd.org"