After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 715181 - NMManager:startup is broken due to stalled device activation
NMManager:startup is broken due to stalled device activation
Status: RESOLVED FIXED
Product: NetworkManager
Classification: Platform
Component: general
git master
Other Linux
: Normal normal
: ---
Assigned To: NetworkManager maintainer(s)
NetworkManager maintainer(s)
Depends on:
Blocks:
 
 
Reported: 2013-11-25 16:50 UTC by Dan Winship
Modified: 2013-11-27 13:55 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
devices: if IP4 and IP6 config both fail right away, fail the device (1.17 KB, patch)
2013-11-26 17:19 UTC, Dan Winship
committed Details | Review

Description Dan Winship 2013-11-25 16:50:40 UTC
just noticed while debugging something else:

72795] [nm-manager.c:767] check_if_startup_complete(): check_if_startup_complete returns FALSE because of lo

danw@laptop:~> nmcli -f STARTUP -t gen status   
starting 

(This is well after startup.)
Comment 1 Dan Winship 2013-11-26 17:19:43 UTC
actually, lo is innocent (it unblocks startup shortly after). It was virbr0 that was causing the problem
Comment 2 Dan Winship 2013-11-26 17:19:58 UTC
Created attachment 262884 [details] [review]
devices: if IP4 and IP6 config both fail right away, fail the device

Previously, the device activation would stall in this case, because
the code wasn't expecting it to happen. In particular, this happens
when trying to assume a device that is up but has no IP config.
Comment 3 Thomas Haller 2013-11-26 18:49:08 UTC
looks good to me
Comment 4 Dan Williams 2013-11-26 23:19:19 UTC
The current patch looks correct.

However, at some point (soon?) we want to support Layer2-only configurations for at least ethernet devices, if only for bridge/bond and DCB.  Jiri Pirko posted some patches starting on that earlier this fall to the mailing list, which I haven't had time to look at yet.  This patch would break that, but there are also other assumptions in the code that need to be fixed up first (ie, these devices can never be 'default' since they don't have IP configuration, and the policy needs to handle that).

We should file another bug for that though, and attach jpirko's patches, and then assign it to somebody :)
Comment 5 Dan Winship 2013-11-27 13:55:05 UTC
Attachment 262884 [details] pushed as ff93ed6 - devices: if IP4 and IP6 config both fail right away, fail the device