dhclient segfaults after uplink restart

U

ultramage

I observed this new behavior yesterday for the first time after updating to freebsd14, after the ISP restarted the modem (and again after I restarted it myself):
Code: 
Jan 14 10:03:55 srv kernel: re1: link state changed to DOWN
Jan 14 10:04:41 srv kernel: re1: link state changed to UP
Jan 14 10:05:05 srv kernel: re1: link state changed to DOWN
Jan 14 10:05:07 srv kernel: re1: link state changed to UP
Jan 14 10:05:18 srv dhclient[79131]: New IP Address (re1): ***.***.202.139
Jan 14 10:05:18 srv dhclient[79135]: New Subnet Mask (re1): 255.255.240.0
Jan 14 10:05:18 srv dhclient[79139]: New Broadcast Address (re1): 255.255.255.255
Jan 14 10:05:18 srv dhclient[79143]: New Routers (re1): ***.***.192.1
Jan 14 10:05:19 srv dhclient[79148]: New Routers (re1): ***.***.192.1
Jan 14 10:05:31 srv dhclient[309]: connection closed
Jan 14 10:05:31 srv kernel: pid 373 (dhclient), jid 0, uid 65: exited on signal 11 (no core dump - other error)
Jan 14 10:05:31 srv dhclient[309]: exiting.
This results in re1 not being properly configured and the host becoming unreachable. I don't know what exactly is wrong as my configuration hasn't changed for a long time. The 'New Routers' line being repeated seems to be a freebsd14 quirk, probably not related. Based on the PIDs, I think the crash is the privileged dhclient instance and the disconnect is the syslog logger instance. I don't know why there's a 12-second gap there.
 
ultramage said:
I observed this new behavior yesterday for the first time after updating to freebsd14, after the ISP restarted the modem (and again after I restarted it myself):
Code: 
Jan 14 10:03:55 srv kernel: re1: link state changed to DOWN
Jan 14 10:04:41 srv kernel: re1: link state changed to UP
Jan 14 10:05:05 srv kernel: re1: link state changed to DOWN
Jan 14 10:05:07 srv kernel: re1: link state changed to UP
Jan 14 10:05:18 srv dhclient[79131]: New IP Address (re1): ***.***.202.139
Jan 14 10:05:18 srv dhclient[79135]: New Subnet Mask (re1): 255.255.240.0
Jan 14 10:05:18 srv dhclient[79139]: New Broadcast Address (re1): 255.255.255.255
Jan 14 10:05:18 srv dhclient[79143]: New Routers (re1): ***.***.192.1
Jan 14 10:05:19 srv dhclient[79148]: New Routers (re1): ***.***.192.1
Jan 14 10:05:31 srv dhclient[309]: connection closed
Jan 14 10:05:31 srv kernel: pid 373 (dhclient), jid 0, uid 65: exited on signal 11 (no core dump - other error)
Jan 14 10:05:31 srv dhclient[309]: exiting.
This results in re1 not being properly configured and the host becoming unreachable. I don't know what exactly is wrong as my configuration hasn't changed for a long time. The 'New Routers' line being repeated seems to be a freebsd14 quirk, probably not related. Based on the PIDs, I think the crash is the privileged dhclient instance and the disconnect is the syslog logger instance. I don't know why there's a 12-second gap there.
Click to expand...
I'm on `stable/14-n267010-acdbbde0274e` and just observed the same problem, without a real gap, however:

Code: 
…
Mar 25 11:35:23 creutzfeld dhclient[349]: DHCPOFFER from 192.168.0.1
Mar 25 11:35:25 creutzfeld dhclient[349]: connection closed
Mar 25 11:35:25 creutzfeld kernel: pid 428 (dhclient), jid 0, uid 65: exited on signal 11 (no core dump - bad address)
Mar 25 11:35:25 creutzfeld dhclient[349]: exiting.
 
Can you turn on core dumps and get a backtrace?

This seems important to fix.

Any way to reproduce it at home?
 
I will need to see if it works in vmware. The Hetzner server has an em(4) NIC (it's a physical server).
Strangely, IPv6 also stopped working, even though it does not use DHCP for assignment.

It's at 14.1-p3 amd64.

It doesn't get much traffic.
 
You must log in or register to reply here.
Back
Top