FS#60518 - coredump of NetworkManager

Attached to Project: Arch Linux
Opened by David (davlucasgmail.com) - Sunday, 21 October 2018, 14:01 GMT
Last edited by Doug Newgard (Scimmia) - Monday, 22 October 2018, 15:33 GMT
Task Type Bug Report
Category Packages: Testing
Status Closed
Assigned To No-one
Architecture x86_64
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 11
Private No

Details

Description: NetworkManager starts but automatically coredump after 2 seconds just after set the IP


Additional info:
networkmanager 1.14.2-1
*
* Evidence :

[root@pc david]# ifconfig -a
docker0: docker up
enp3s0: ethernet not connected
lo:
virbr0: qemu up
virbr1: qemu
virbr0-nic: qemu
virbr1-nic: Qemu
wlp0s20u4: Wifi UP

Steps to reproduce:
systemctl start NetworkManager.service

journalctl shows:

oct. 21 15:40:27 pc zeitgeist-daemon[1616]: [13:40:27.509601 CRITICAL] file ext-storage-monitor.c: line 1691: unexpected erro
r: GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: L’interface « org.freedesktop.NetworkManager » n’existe pas pour l’o
bjet à l’emplacement /org/freedesktop/NetworkManager (g-dbus-error-quark, 19)
ct. 21 15:40:27 pc NetworkManager[1682]: <info> [1540129227.5858] device (enp3s0): state change: unmanaged -> unavailable (
reason 'managed', sys-iface-state: 'external')
oct. 21 15:40:27 pc NetworkManager[1682]: <info> [1540129227.5899] manager: (virbr0): new Bridge device (/org/freedesktop/Ne
tworkManager/Devices/4)
oct. 21 15:40:27 pc audit[1682]: ANOM_ABEND auid=4294967295 uid=0 gid=0 ses=4294967295 pid=1682 comm="NetworkManager" exe="/u
sr/bin/NetworkManager" sig=11 res=1
oct. 21 15:40:27 pc kernel: NetworkManager[1682]: segfault at 8 ip 0000557838a2a8c5 sp 00007ffe204905a0 error 4 in NetworkMan
ager[5578389e8000+1aa000]
oct. 21 15:40:27 pc mission-control[1120]: Error calling GetManagedObjects() when name owner :1.149 for name org.freedesktop.
NetworkManager came back: GDBus.Error:org.freedesktop.DBus.Error.NoReply: Message recipient disconnected from message bus wit
hout replying
oct. 21 15:40:28 pc systemd-coredump[1711]: Process 1682 (NetworkManager) of user 0 dumped core.
#0 0x0000557838a2a8c5 n/a (NetworkManager)
#1 0x0000557838a0f555 n/a (NetworkManager)
#2 0x0000557838a10554 n/a (NetworkManager)
#3 0x0000557838a1233b n/a (NetworkManager)
#4 0x0000557838a15805 n/a (NetworkManager)
#5 0x00005578389e8c34 n/a (NetworkManager)
#7 0x00005578389e8f8e n/a (NetworkManager)



Functional issue :
kde does not work smoothly, it's ok with xfce or gnome

This task depends upon

Closed by  Doug Newgard (Scimmia)
Monday, 22 October 2018, 15:33 GMT
Reason for closing:  Fixed
Additional comments about closing:  networkmanager 1.14.3dev+15+g3fc592219-1
Comment by Chris (kitteboss) - Sunday, 21 October 2018, 14:23 GMT
Can confirm the segfault of NetworkManager from package version 1.14.2-1. KDE Plasma shows 'hickups' as consequence and will not start correctly, probably because of networkmanager-qt. For me the segfault occurres immediatley when '/usr/bin/NetworkManager --no-daemon' is executed, without delay
Comment by Kirill Zhirovsky (kirill9617) - Sunday, 21 October 2018, 15:55 GMT
I have this segfault too.
NetworkManager[24266]: segfault at 8 ip 000056244e3028c5 sp 00007ffe905f7c30 error 4 in NetworkManager[56244e2c0000+1aa000]
Code: 33 1c 25 28 00 00 00 48 89 e8 0f 85 9b 03 00 00 48 83 c4 68 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 44 00 00 48 8b 44 24 28 <48> 8b 58 08 6>
Comment by Chris (kitteboss) - Sunday, 21 October 2018, 17:16 GMT
Managed to get a backtrace with gdb including debug symbols:

Starting program: /usr/bin/NetworkManager --no-daemon
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
[New Thread 0x7ffff3b74700 (LWP 20695)]
[New Thread 0x7ffff3373700 (LWP 20696)]
[New Thread 0x7ffff2b72700 (LWP 20697)]
[Detaching after fork from child process 20699]

Thread 1 "NetworkManager" received signal SIGSEGV, Segmentation fault.
0x00005555555c5825 in nm_device_generate_connection (self=self@entry=0x5555559358b0, master=<optimized out>,
out_maybe_later=out_maybe_later@entry=0x7fffffffe740, error=error@entry=0x7fffffffe750) at ../NetworkManager/src/devices/nm-device.c:5490
5490 ../NetworkManager/src/devices/nm-device.c: No such file or directory.
(gdb) thread apply all bt

Thread 4 (Thread 0x7ffff2b72700 (LWP 20697)):
#0 0x00007ffff7445c21 in poll () at /usr/lib/libc.so.6
#1 0x00007ffff7ca7ee0 in () at /usr/lib/libglib-2.0.so.0
#2 0x00007ffff7ca8f62 in g_main_loop_run () at /usr/lib/libglib-2.0.so.0
#3 0x00007ffff7e0dc28 in () at /usr/lib/libgio-2.0.so.0
#4 0x00007ffff7c713eb in () at /usr/lib/libglib-2.0.so.0
#5 0x00007ffff7520a9d in start_thread () at /usr/lib/libpthread.so.0
#6 0x00007ffff7450b23 in clone () at /usr/lib/libc.so.6

Thread 3 (Thread 0x7ffff3373700 (LWP 20696)):
#0 0x00007ffff744b4ed in syscall () at /usr/lib/libc.so.6
#1 0x00007ffff7c56f61 in g_cond_wait_until () at /usr/lib/libglib-2.0.so.0
#2 0x00007ffff7cd4cb3 in () at /usr/lib/libglib-2.0.so.0
#3 0x00007ffff7c78d2b in () at /usr/lib/libglib-2.0.so.0
#4 0x00007ffff7c713eb in () at /usr/lib/libglib-2.0.so.0
#5 0x00007ffff7520a9d in start_thread () at /usr/lib/libpthread.so.0
#6 0x00007ffff7450b23 in clone () at /usr/lib/libc.so.6

Thread 2 (Thread 0x7ffff3b74700 (LWP 20695)):
#0 0x00007ffff7445c21 in poll () at /usr/lib/libc.so.6
#1 0x00007ffff7ca7ee0 in () at /usr/lib/libglib-2.0.so.0
#2 0x00007ffff7ca7fce in g_main_context_iteration () at /usr/lib/libglib-2.0.so.0
#3 0x00007ffff7ca8022 in () at /usr/lib/libglib-2.0.so.0
#4 0x00007ffff7c713eb in () at /usr/lib/libglib-2.0.so.0
#5 0x00007ffff7520a9d in start_thread () at /usr/lib/libpthread.so.0
#6 0x00007ffff7450b23 in clone () at /usr/lib/libc.so.6

Thread 1 (Thread 0x7ffff3b75ac0 (LWP 20691)):
#0 0x00005555555c5825 in nm_device_generate_connection
(self=self@entry=0x5555559358b0, master=<optimized out>, out_maybe_later=out_maybe_later@entry=0x7fffffffe740, error=error@entry=0x7fffffffe750)
at ../NetworkManager/src/devices/nm-device.c:5490
#1 0x00005555555a8dd5 in get_existing_connection (out_generated=<synthetic pointer>, device=0x5555559358b0, self=0x55555587b050)
at ../NetworkManager/src/nm-manager.c:2482
#2 0x00005555555a8dd5 in recheck_assume_connection (self=self@entry=0x55555587b050, device=device@entry=0x5555559358b0)
at ../NetworkManager/src/nm-manager.c:2645
#3 0x00005555555a9e34 in _device_realize_finish (self=0x55555587b050, device=0x5555559358b0, plink=0x555555866038)
at ../NetworkManager/src/nm-manager.c:2859
#4 0x00005555555abbcc in platform_link_added
(self=self@entry=0x55555587b050, ifindex=ifindex@entry=4, plink=plink@entry=0x555555866038, guess_assume=guess_assume@entry=0, dev_state=dev_state@e--Type <RET> for more, q to quit, c to continue without paging--
ntry=0x0) at ../NetworkManager/src/nm-manager.c:3185
#5 0x00005555555af7d5 in platform_link_added (dev_state=0x0, guess_assume=0, plink=0x555555866038, ifindex=4, self=0x55555587b050)
at ../NetworkManager/src/nm-manager.c:3071
#6 0x00005555555af7d5 in platform_query_devices (self=0x55555587b050) at ../NetworkManager/src/nm-manager.c:3297
#7 0x00005555555af7d5 in nm_manager_start (self=0x55555587b050, error=<optimized out>) at ../NetworkManager/src/nm-manager.c:6362
#8 0x000055555557fcb4 in main (argc=<optimized out>, argv=<optimized out>) at ../NetworkManager/src/main.c:417
Comment by David (davlucasgmail.com) - Sunday, 21 October 2018, 18:35 GMT
I found a workaround so KDE can work "smoothly" :
I disabled any network applets (by disabling first fully networkmanager)
And also, I disabled the service dbus-org.freedesktop.nm-dispatcher.service
#systemctl disable dbus-org.freedesktop.nm-dispatcher.service
#reboot

Just a workaround but useless for the ones who really need to work on KDE
Comment by Kevin Sopp (baraclese) - Monday, 22 October 2018, 10:34 GMT
As a workaround I boot the linux-lts kernel.
Comment by Chris (kitteboss) - Monday, 22 October 2018, 11:08 GMT
The best "workaround" should be to downgrade networkmanager and its dependencies to the version currently in stable. This should fix the problems.
The bug itself is a null pointer dereference of the variable local in ../NetworkManager/src/devices/nm-device.c:5490, which can be easily fixed by checking if the pointer is null.
I have a tentative patch which seems to work but as I am not familiar with the code I prefer not to publish it.
I assume one of the package maintainers will file a bug report upstream as soon as they are aware of the bug report.
Comment by Patrick Young (kmahyyg) - Monday, 22 October 2018, 11:31 GMT
This package is now merged into extra instead of leaving at testing, still cause segmentfault on KDE. You may need to recall 1.4.2-1 .
Comment by Mal Haak (insanemal) - Monday, 22 October 2018, 12:21 GMT
Hit this a few moments ago when I updated to the package in Extra.

Can confirm, its toast.

Can I get logs, will that help?

Downgrade fixed things.
Comment by David (davlucasgmail.com) - Monday, 22 October 2018, 13:34 GMT
Tanks, I never think on downgrade as an option, thanks to remind me

I've downgraded to ./networkmanager-1.12.2-1-x86_64.pkg.tar.xz , it works fine

Seems like kitteboss found the cause.






Comment by Peter Fern (pdf) - Monday, 22 October 2018, 13:35 GMT
Confirmed that downgrading to 1.14.0-1 does not produce coredumps under these conditions, please yank 1.14.2-1 ASAP.
Comment by Chris (kitteboss) - Monday, 22 October 2018, 14:45 GMT
1.14.3dev+15+g3fc592219-1 fixed the issue for me. Thx

Loading...