From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)
Description of problem:
I have used the acenic driver together with two alteon gigabit ethernet cards,
which one of them was configured by pump. after a short while of working properly, the card which
was configured by pump got stuck, sending ARP requests, but not responding to the answers (although
they were transmitted across the wires).
Steps to Reproduce:
1. configure a dhcp with pump as the client for an alteon card with acenic
2. mount a remote directory with NFS, and copy lots of files into it
3. now, delete all the files from the directory with rm -rf
Actual Results: the card always got stuck
Expected Results: normal work, no card hang
I have investigated a little and compared the way pump configures the device to the way ifconfig
does (if I configure the device with ifconfig, this doesn't happen), and i have found that the
cause for this is the way pump disable the device.
when setting up the device, pump first disables the device by removing "IFF_UP" and "IFF_RUNNING"
flags from the device.
ifconfig only disableds the "IFF_RUNNING" flag before disabling the device.
I have recompiled pump and removed the IFF_UP disable from the pumpDisableInterface function in
dhcp.c, and it seems that it fixed the problem.
I am not sure if this is a problem with the way pump disable the device, or a bug in the acenic
driver, but this is what causes it.
for additional information you can contact me-
thanks in advance- Ori
If pump works on other network cards with these settings, it really should work
on the alteon. Going to ask if this is actually a kernel driver problem...?
pump disables the device over-agressive indeed; this causes problems with
other networkcards as well. You could try using dhcpcd instead (should be on the
cd as well)