wired network disabled after resume (AR9485)

[Mark Thornton]

I have both a wired network and wireless. On first boot both are functional, but on resuming from sleep the wired controller is disabled (wireless still works). This behaviour started with the latest kernel installed yesterday.

From syslog on resume

Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> Unmanaged Device found; state CONNECTED forced. (see http://bugs.launchpad.net/bugs/191889)
Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> Unmanaged Device found; state CONNECTED forced. (see http://bugs.launchpad.net/bugs/191889)
Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> (eth4): now managed
Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> (eth4): device state change: unmanaged -> unavailable (reason 'managed') [10 20 2]
Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> (eth4): bringing up device.
Feb 20 08:48:31 mark-PBL21 NetworkManager[1109]: <info> (eth4): deactivating device (reason 'managed') [2]

 uname -a
Linux mark-PBL21 3.5.0-24-generic #37-Ubuntu SMP Thu Feb 7 01:50:30 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

$ sudo lshw -C network
[sudo] password for mthornton:
  *-network
       description: Wireless interface
       product: AR9485 Wireless Network Adapter
       vendor: Atheros Communications Inc.
       physical id: 0
       bus info: pci@0000:03:00.0
       logical name: wlan3
       version: 01
       serial: dc:85:de:03:c8:ce
       width: 64 bits
       clock: 33MHz
       capabilities: pm msi pciexpress bus_master cap_list rom ethernet physical wireless
       configuration: broadcast=yes driver=ath9k driverversion=3.5.0-24-generic firmware=N/A ip=192.168.41.214 latency=0 link=yes multicast=yes wireless=IEEE 802.11bgn
       resources: irq:17 memory:f7900000-f797ffff memory:f7980000-f798ffff
  *-network DISABLED
       description: Ethernet interface
       product: AR8161 Gigabit Ethernet
       vendor: Atheros Communications Inc.
       physical id: 0
       bus info: pci@0000:04:00.0
       logical name: eth4
       version: 10
       serial: 10:bf:48:9f:e9:c6
       width: 64 bits
       clock: 33MHz
       capabilities: pm pciexpress msi msix bus_master cap_list ethernet physical
       configuration: broadcast=yes driver=alx latency=0 multicast=yes
       resources: irq:19 memory:f7800000-f783ffff ioport:d000(size=128)

[Mark Thornton (mthornton-2)]

Still present with next kernel

$ uname -a
Linux mark-PBL21 3.5.0-25-generic #38-Ubuntu SMP Mon Feb 18 23:27:42 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

[Mark Thornton (mthornton-2)]

The problem seems to be the alx driver. The workaround here:

http://www.webupd8.org/2013/01/fix-wireless-or-wired-network-not.html

works.

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Gabriel (cartajenaa)]

I have this bug too. I had this issue on 14.04, 14. 10, 15.04 and 15.10. Now I will download 12.04.2 LTS* maybe is more stable to our AR9485 Wireless Network Adapter. My PC is HP 1000 Notebook PC and googled somethings and can be usable:

1. *Certified systems: http://www.ubuntu.com/certification/catalog/component/pci/168c%3A0032/
2. Linux kernel backports: http://drvbp1.linux-foundation.org/~mcgrof/rel-html/backports/
Instructions: http://www.vivaolinux.com.br/dica/Configurando-Wireless-no-GNULinux
3. Stabilizing an atheros ar9485: http://pkill-9.com/stabilizing-atheros-ar9485-ath9k-connection-linux-wheezy/

But nothing worked as well.

[Gabriel (cartajenaa)]

There my SystemLog when the problem came: http://pastebin.com/uZ43a28T
I needed restart some times. This sucks :(

[Thorsten Greeb (thorsten-greeb)]

I also have this problem on several machines running Ubuntu 16.04. After resume from suspend no IPv4 address is assigned to the network interfaces (both wired and wireless). Seems to be not related to hardware or driver but to NetworkManager.

Maybe this (known) problem is corrected in version 1.2.6 (see change log).

[Thorsten Greeb (thorsten-greeb)]

Interesting: Today network-manager 1.2.4 was published in "Xenial (16.04) proposed" which contains the cherry-picked fix from 1.2.6...