FS#39381 - [dhcpcd] segmentation fault with package version 6.3.1-1 -> Network Manager fails

Attached to Project: Arch Linux
Opened by Michael (Zeth) - Thursday, 13 March 2014, 11:06 GMT
Last edited by Anatol Pomozov (anatolik) - Friday, 14 March 2014, 14:39 GMT
Task Type Bug Report
Category Packages: Core
Status Closed
Assigned To Ronald van Haren (pressh)
Anatol Pomozov (anatolik)
Architecture x86_64
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:
Today I upgrade my system (last full update was 3 days ago) and after I woke up my laptop from suspension, network didn't come up again (also tried a full system restart).

The log said
Mar 13 19:14:50 XXXX NetworkManager[423]: <info> Activation (enp2s0) Beginning DHCPv4 transaction (timeout in 45 seconds)
Mar 13 19:14:50 XXXX dhcpcd[3522]: version 6.3.1 starting
Mar 13 19:14:50 XXXX NetworkManager[423]: <info> dhcpcd started with pid 3522
Mar 13 19:14:50 XXXX NetworkManager[423]: <info> Activation (enp2s0) Stage 3 of 5 (IP Configure Start) complete.
Mar 13 19:14:50 XXXX dhcpcd[3522]: DUID <removed>
Mar 13 19:14:50 XXXX dhcpcd[3522]: enp2s0: IAID 85:39:b9:63
Mar 13 19:14:50 XXXX dhcpcd[3522]: enp2s0: soliciting an IPv6 router
Mar 13 19:14:50 XXXX NetworkManager[423]: <info> (enp2s0): DHCPv4 state changed nbi -> preinit
Mar 13 19:14:50 XXXX dhcpcd[3522]: enp2s0: rebinding lease of 192.168.1.2
Mar 13 19:14:51 XXXX dhcpcd[3522]: enp2s0: Router Advertisement from <removed>
Mar 13 19:14:51 XXXX dhcpcd[3522]: enp2s0: requesting DHCPv6 information
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> (enp2s0): DHCPv4 state changed preinit -> (null)
Mar 13 19:14:51 XXXX kernel: dhcpcd[3522]: segfault at 10 ip 000000000042086c sp 00007fffcd28e910 error 4 in dhcpcd[400000+30000]
Mar 13 19:14:51 XXXX systemd-coredump[3525]: Process 3522 (dhcpcd) dumped core.
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> (enp2s0): DHCPv4 client pid 3522 exited with status -1
Mar 13 19:14:51 XXXX NetworkManager[423]: <warn> DHCP client died abnormally
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> Activation (enp2s0) Stage 4 of 5 (IPv4 Configure Timeout) scheduled...
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> Activation (enp2s0) Stage 4 of 5 (IPv4 Configure Timeout) started...
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> (enp2s0): device state change: ip-config -> failed (reason 'ip-config-unavailable') [70 120 5]
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> NetworkManager state is now DISCONNECTED
Mar 13 19:14:51 XXXX NetworkManager[423]: <info> Marking connection 'DHCP' invalid.
Mar 13 19:14:51 XXXX NetworkManager[423]: <warn> Activation (enp2s0) failed for connection 'DHCP'

Calling
# dhcpcd enp2s0
resulted in the same segfault message.

Fixed IP settings work.

I downgraded to 6.2.1-1 and it works again for now.

Additional info:
* package version:
dhcpcd 6.3.1-1
networkmanager 0.9.8.8-3
* latest Gnome 3 desktop environment
* x86_64
* ethernet hardware:
02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 03)
Subsystem: Hewlett-Packard Company Device 1486
Kernel driver in use: r8169
Kernel modules: r8169
03:00.0 Network controller: Broadcom Corporation BCM4313 802.11bgn Wireless Network Adapter (rev 01)
Subsystem: Hewlett-Packard Company Device 145c
Kernel driver in use: bcma-pci-bridge
Kernel modules: bcma

Steps to reproduce:
Upgrade to dhcpcd 6.3.1-1 and restart network connection.
Not sure if this happens with every hardware.

Please tell me if you need any other information.
This task depends upon

Closed by  Anatol Pomozov (anatolik)
Friday, 14 March 2014, 14:39 GMT
Reason for closing:  Fixed
Additional comments about closing:  dhcpcd-6.3.1-2
Comment by Doug Newgard (Scimmia) - Thursday, 13 March 2014, 15:28 GMT
Other people reporting the same thing: https://bbs.archlinux.org/viewtopic.php?id=178381
Comment by Anatol Pomozov (anatolik) - Thursday, 13 March 2014, 15:34 GMT
Haven't seen this issue before.

Could you please post the stack trace? "systemd-coredumpctl gdb" and then type "bt full".

Does it look like this issue? http://roy.marples.name/archives/dhcpcd-discuss/2014/0580.html
Comment by Anatol Pomozov (anatolik) - Thursday, 13 March 2014, 15:35 GMT Comment by Anatol Pomozov (anatolik) - Friday, 14 March 2014, 02:07 GMT
I am pushing to [testing] version 6.3.1-2. It contains 2 crash fixes from upstream and potentially might fix this problem. Please install this package and let me know if the crash still exists.
Comment by Michael (Zeth) - Friday, 14 March 2014, 04:05 GMT
Yup, 6.3.1-2 from testing does not crash anymore.

The trace looks a bit like the latter linked crash report 'cause of the dhcp6_handledata but not sure as there are no debugging symbols in the binary.

Anyway, fixed with the new version.
Thanks a lot!

TIME PID UID GID SIG EXE
Fr 2014-03-14 12:49:54 JST 16510 0 0 11 /usr/bin/dhcpcd
GNU gdb (GDB) 7.7
Copyright (C) 2014 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-unknown-linux-gnu".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
<http://www.gnu.org/software/gdb/documentation/>.
For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from /usr/bin/dhcpcd...(no debugging symbols found)...done.
[New LWP 16510]

warning: Could not load shared library symbols for linux-vdso.so.1.
Do you need "set solib-search-path" or "set sysroot"?
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Core was generated by `dhcpcd enp2s0'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x000000000042086c in dhcp6_handledata ()
(gdb) bt full
#0 0x000000000042086c in dhcp6_handledata ()
No symbol table info available.
#1 0x0000000000406e05 in eloop_start ()
No symbol table info available.
#2 0x0000000000402ef9 in main ()
No symbol table info available.
(gdb)
Comment by Ronald van Haren (pressh) - Friday, 14 March 2014, 06:58 GMT
Thanks for taking care of the fix for me Anatol.
Comment by Roy Marples (rsmarples) - Friday, 14 March 2014, 11:05 GMT
dhcpcd-6.3.2 has been released as well which should fix this and quite a few more issues.
Comment by Anatol Pomozov (anatolik) - Friday, 14 March 2014, 14:39 GMT
6.3.1-2 is moved to stable.

6.3.2-1 just pushed to [testing].

Loading...