Anybody else seeing 'dc0: watchdog timeout' since SMPng integration?
Just after the removal of SPLs on alpha, my UP1000 started
occasionally spewing 'dc0: watchdog timeout' and effectively dropping
off the network when linking 23+ MB alpha kernel.debug's over NFSv3 to
a reasonably snappy NFS server. This happens on maybe 1 in 5 kernel
links. I don't have tulip cards in any PCs, so it is hard for me to
tell if this is an alpha issue or an if_dc driver issue..
When it happens, I see the following:
Oct 23 17:37:58 thunder /kernel.test: dc0: TX underrun -- increasing TX threshold
Oct 23 17:54:44 thunder /kernel.test: dc0: watchdog timeout
Oct 23 17:54:59 thunder /kernel.test: nfs server xxx:/export/xxx: not responding
Oct 23 17:55:02 thunder /kernel.test: nfs server xxx:/export/xxx: not responding
<...>
The card in question is a gen-u-ine DEC DE500 of some vintage:
# pciconf -l | grep ^dc
dc0@pci0:9:0: class=0x020000 card=0x500b1011 chip=0x00191011 rev=0x30 hdr=0x00
It identifies itself as:
dc0: <Intel 21143 10/100BaseTX> port 0x10100-0x1017f mem 0x41353100-0x4135317f irq 10
at device 9.0 on pci0
dc0: interrupting at ISA irq 10
dc0: Ethernet address: 00:00:f8:07:b6:45
miibus0: <MII bus> on dc0
dcphy0: <Intel 21143 NWAY media interface> on miibus0
dcphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
<...>
I'm running in hardcoded 100baseTX full-duplex mode:
dc0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
inet 152.3.x.y netmask 0xffffff00 broadcast 152.3.x.255
ether 00:00:f8:07:b6:45
media: 100baseTX <full-duplex> status: active
supported media: autoselect 100baseTX <full-duplex> 100baseTX 10baseT/UTP
<full-duplex> 10baseT/UTP 100baseTX <hw-loopback> none
Thanks,
Drew
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
