Comment 4 for bug 16137

We had a setup with kernel-image-2.6.9-2-k7 2.6.9-5 and iproute
20010824-8woody1 and the same problem as described. Then we upgraded to
the latest unstable versions:

kernel-image-2.6.10-1-k7 2.6.10-4
iproute 20041019-2

but still encounter the same problem: "ip -4 neigh flush dev eth0" hangs
up.

We noticed that it only happens if there are entries in the ARP-Table.
If the ARP cache is empty, it exits normally saying "Nothing to flush".
Our ARP-Cache is also being populated very quickly, maybe there is a
race-condition here?

If we leave the hung ip neigh flush dev eth0 running and in another
console issue something like:

# for a in `arp -n | awk '/^[0-9]/ { print $1; }'`; do arp -d $a; done

some times in a row (it cleans up all entries in the arp-cache), the
hung ip command awakes and finishes.