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 797425 - Wired connection stops working on resume from suspend (plasma-nm applet shows no connection)
Wired connection stops working on resume from suspend (plasma-nm applet shows...
Status: RESOLVED NOTGNOME
Product: NetworkManager
Classification: Platform
Component: general
unspecified
Other Linux
: Normal normal
: ---
Assigned To: NetworkManager maintainer(s)
NetworkManager maintainer(s)
Depends on:
Blocks:
 
 
Reported: 2019-02-14 14:44 UTC by Strangiato
Modified: 2019-04-24 18:26 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
screenshot (328.29 KB, image/png)
2019-02-14 14:44 UTC, Strangiato
Details
nm log (30.00 KB, text/plain)
2019-02-14 15:02 UTC, Strangiato
Details

Description Strangiato 2019-02-14 14:44:12 UTC
Created attachment 374195 [details]
screenshot

Bug was already reported in KDE tracker
https://bugs.kde.org/show_bug.cgi?id=404345

Jan Grulich said there is some bug with networkmanager:
"From the log it looks like a NM issue, your wired device goes into UNMANAGED state and later on into UNAVAILABLE. Please report this issue to NetworkManager."

STEPS TO REPRODUCE
1. start KDE Plasma
2. connect to a wired connection
3. suspend and resume your system

OBSERVED RESULT
plasma notification indicates connection failure, plasma-nm applet shows no connection despite the wired connection is listed in plasma-nm editor and network manager service is running. See the screenshot.
Wired connection only works again after reboot, restart plasma does not help.

EXPECTED RESULT
wired connection works

SOFTWARE/OS VERSIONS
Operating System: Arch Linux 
KDE Plasma Version: 5.15.0
KDE Frameworks Version: 5.55.0
Qt Version: 5.12.1
Kernel Version: 4.20.7-arch1-1-ARCH
Comment 1 Thomas Haller 2019-02-14 14:46:38 UTC
please attach a logfile with level=TRACE. see the hints at https://cgit.freedesktop.org/NetworkManager/NetworkManager/tree/contrib/fedora/rpm/NetworkManager.conf#n28

Thanks!
Comment 2 Strangiato 2019-02-14 15:02:42 UTC
Created attachment 374196 [details]
nm log
Comment 3 Strangiato 2019-03-16 19:03:46 UTC
bug persists on Arch Linux after update to nm 1.16
Comment 4 Strangiato 2019-04-23 18:51:45 UTC
Bug persists on Arch Linux after update to nm 1.18.
Comment 5 Thomas Haller 2019-04-24 11:48:10 UTC
in the attached log, you see that the device stays in state "unavailable". As you probably would also see in `nmcli device` output and `nmcli -f all device show enp4s0`.

The reason why the device is "unavailable" is because it has no carrier (cable not plugged in). You see that for example here:

  [1550155921.9009] platform: signal: link changed: 2: enp4s0 <UP;broadcast,multicast,up>

(it's "UP" but not "LOWER_UP"). You should see the same in `ip link` output.


The attached log contains only a short part, but presuming that the relevant portion is there:

  - ensure the cable is plugged in :)

  - looks like a driver/firmware issue.



that gives `ip -d link` when this occurs?
Comment 6 Strangiato 2019-04-24 12:30:36 UTC
The cable is always plugged in. Just reboot the computer is enough to make the connection work again.

Here is the output of 'ip -d link'

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 promiscuity 0 minmtu 0 maxmtu 0 addrgenmode eui64 numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535 
2: enp4s0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc fq_codel state DOWN mode DEFAULT group default qlen 1000
    link/ether fc:aa:14:fc:4b:1b brd ff:ff:ff:ff:ff:ff promiscuity 0 minmtu 60 maxmtu 9200 addrgenmode none numtxqueues 1 numrxqueues 1 gso_max_size 65536 gso_max_segs 65535
Comment 7 Thomas Haller 2019-04-24 14:19:56 UTC
> 2: enp4s0: <NO-CARRIER,

This is a driver/firmware/kernel issue.

There is the "ignore-carrier" setting in NetworkManager.conf. But that's not gonna help, because without carrier you cannot do DHCP.


I don't think NetworkManager can do anything about this.
Comment 8 Strangiato 2019-04-24 14:34:24 UTC
Humm... I tested both r8168 and r8169 (kernel) drivers with both 5.0.9 and 4.19.36 kernels and the problem persists.
Can it be some problem with my router?
Comment 9 Thomas Haller 2019-04-24 17:30:01 UTC
It could be. The switch may disable the port. For example, it may do so if it detects a routing loop or some forbidden traffic. But in this case, that seems *not* the likely issue. I don't know.
Comment 10 Strangiato 2019-04-24 18:00:47 UTC
I just solved the problem by resetting the BIOS settings on my motherboard.
I think that was some problem related to power management settings.
Thank you very much for your interest on this issue Thomas.
Have a nice day.
Comment 11 Thomas Haller 2019-04-24 18:26:23 UTC
Oh great. I am closing this bug. Please reopen if something's to add. Thank you!