Red Hat Bugzilla – Bug 863936
dhcrelay segfaults when packet is received over interface without any IPv4 address assigned
Last modified: 2013-11-21 02:43:16 EST
Description of problem: dhcrelay segfaults when sending a DHCPDISCOVER from a directly connected host. System setup dell-per610-01 (em2) <---> ibm-dx360m4-01 (eth1) <---> 10.16.64.14 client dhcrelay DHCP root@ibm-dx360m4-01.rhts.eng.bos.redhat.com:~> gdb dhcrelay -ex 'set args -d -i eth0 -i eth1 10.16.64.14' GNU gdb (GDB) Red Hat Enterprise Linux (7.2-56.el6) Copyright (C) 2010 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-redhat-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>... Reading symbols from /usr/sbin/dhcrelay...Reading symbols from /usr/lib/debug/usr/sbin/dhcrelay.debug...done. done. (gdb) r Starting program: /usr/sbin/dhcrelay -d -i eth0 -i eth1 10.16.64.14 Internet Systems Consortium DHCP Relay Agent 4.1.1-P1 Copyright 2004-2010 Internet Systems Consortium. All rights reserved. For info, please visit https://www.isc.org/software/dhcp/ Listening on LPF/eth1/34:40:b5:b9:5f:9c Sending on LPF/eth1/34:40:b5:b9:5f:9c Listening on LPF/eth0/34:40:b5:b9:5f:9b Sending on LPF/eth0/34:40:b5:b9:5f:9b Sending on Socket/fallback Forwarded BOOTREQUEST for 00:14:5e:f4:78:38 to 10.16.64.14 Forwarded BOOTREPLY for 00:14:5e:f4:78:38 to 10.16.68.21 Forwarded BOOTREQUEST for 00:14:5e:f4:78:34 to 10.16.64.14 Forwarded BOOTREPLY for 00:14:5e:f4:78:34 to 255.255.255.255 Forwarded BOOTREQUEST for 00:14:5e:f4:78:34 to 10.16.64.14 Forwarded BOOTREPLY for 00:14:5e:f4:78:34 to 255.255.255.255 Program received signal SIGSEGV, Segmentation fault. 0x00007ffff7f9f467 in do_relay4 (ip=<value optimized out>, packet=0x7fffffffccc0, length=<value optimized out>, from_port=<value optimized out>, from=..., hfrom=<value optimized out>) at dhcrelay.c:660 660 packet->giaddr = ip->addresses[0]; (gdb) bt #0 0x00007ffff7f9f467 in do_relay4 (ip=<value optimized out>, packet=0x7fffffffccc0, length=<value optimized out>, from_port=<value optimized out>, from=..., hfrom=<value optimized out>) at dhcrelay.c:660 #1 0x00007ffff7fa6e1f in got_one (h=0x7ffff820da70) at discover.c:1400 #2 0x00007ffff7fd3353 in omapi_one_dispatch (wo=<value optimized out>, t=0x0) at dispatch.c:520 #3 0x00007ffff7fa8e5a in dispatch () at dispatch.c:92 #4 0x00007ffff7f9dbc1 in main (argc=<value optimized out>, argv=<value optimized out>) at dhcrelay.c:549 (gdb) Version-Release number of selected component (if applicable): dhcp-4.1.1-31.P1.el6.x86_64 How reproducible: always Steps to Reproduce: 1. on ibm-dx360m4-01 run: dhcrelay -i eth0 -i eth1 10.16.64.14 2. on dell-per610-01.lab.bos.redhat.com run: dhclient -v em2 3. Actual results: dhcrelay segfault This also happens when using /etc/sysconfig/dhcrelay cat /etc/sysconfig/dhcrelay # Command line options here DHCRELAYARGS="" # DHCPv4 only INTERFACES="eth0 eth1" # DHCPv4 only DHCPSERVERS="10.16.64.14"
I've compiled dhcp-4.2.4-P2 and tried again. Here's the output Starting program: /usr/sbin/dhcrelay -d -i eth0 -i eth1 10.16.64.14 Internet Systems Consortium DHCP Relay Agent 4.2.4-P2 Copyright 2004-2012 Internet Systems Consortium. All rights reserved. For info, please visit https://www.isc.org/software/dhcp/ Listening on LPF/eth1/34:40:b5:b9:5f:9c Sending on LPF/eth1/34:40:b5:b9:5f:9c Listening on LPF/eth0/34:40:b5:b9:5f:9b Sending on LPF/eth0/34:40:b5:b9:5f:9b Sending on Socket/fallback Forwarded BOOTREQUEST for 00:17:ef:dd:21:00 to 10.16.64.14 Forwarded BOOTREPLY for 00:17:ef:dd:21:00 to 10.16.65.209 Discarding packet received on eth1 interface that has no IPv4 address assigned. Discarding packet received on eth1 interface that has no IPv4 address assigned. Discarding packet received on eth1 interface that has no IPv4 address assigned. So the segfault probably results in em2 not having an IPv4 address assigned. But actually that's what I'm trying to do. Would be nice if you could contact me on irc as I'm working on reproducing bug #845515 and could use a little help with that. TIA Jiri!
I think I see my error now. I had ibm-dx360m4-01.rhts.eng.bos.redhat.com eth0 has in IP from DHCP eth1 (direct connection to dell-per610-01) just had link up, but no IP assigned. This causes the segfault and still should be fixed. Now that I've manually assigned an IP to eth1 dhcrelay works.
Created attachment 623346 [details] patch This has been fixed upstream in ISC dhcp-4.2.2b1 with the following note in changelog: - Relay no longer crashes, when DHCP packet is received over interface without any IPv4 address assigned. [ISC-Bugs #22409]
*** Bug 920559 has been marked as a duplicate of this bug. ***
Hi Jiri, please do you know when this patch will included? thanks
Michal, you can check the Flags yourself: rhel‑6.5.0? means that it's been proposed for 6.5 pm_ack? means that we are waiting for product management (pm) once rhel-6.x.0 is + instead of ? it means it'll be fixed in rhel-6.x.0
(In reply to comment #9) > Michal, you can check the Flags yourself: I've just realized that you probably can't see them, sorry. So ... yes, it's been proposed for 6.5, but it all depends on product management.
(In reply to comment #10) > (In reply to comment #9) > > Michal, you can check the Flags yourself: > > I've just realized that you probably can't see them, sorry. > So ... yes, it's been proposed for 6.5, but it all depends on product > management. Hi Jiri, yes, you're right. Flags field is empty for me even I've logged here. Thanks for info.
Verified with dhclient-4.1.1-38.P1.el6.x86_64. dhcrelay no longer segfaults when IP address is not present on network interfaces.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2013-1572.html