Twice today during very heavy network I/O (dumping a large postgres
database over the ethernet to another machine on the same switch) I got
this error:
Mar 22 03:42:22 d01 kernel: bge0: watchdog timeout -- resetting
Mar 22 10:28:24 d01 kernel: bge0: watchdog timeout -- resetting
The most informative info I found searching google was this one:
http://unix.derkeiler.com/Mailing-Lists/FreeBSD/net/2003-08/0096.html
However, that seems to imply that a 5.4-PRERELEASE build March 15 would
not be caused by this particular bug. The dmesg says this of bge0:
bge0: <Broadcom BCM5704C Dual Gigabit Ethernet, ASIC rev. 0x2003> mem
0xfc9b0000-0xfc9bffff,0xfc9c0000-0xfc9cffff irq 24 at device 9.0 on
pci3
miibus0: <MII bus> on bge0
bge0: Ethernet address: 00:e0:81:2a:ac:5e
Any insights on where to look? This is a dual Opteron running
FreeBSD/amd64 with 4GB disk.
The only thing funny I did was build with CPUTYPE=opteron
optimizations. I'm rebuilding right now without that, but I've also
cvsup'd the sources... This is my SWAG as to the problem.
Anyone else still experiencing these timeouts?
Vivek Khera, Ph.D.
+1-301-869-4449 x806