Doesn't restore interface state correctly on resume

Bug #37515 reported by Matt Zimmerman
12
Affects Status Importance Assigned to Milestone
acpi-support (Ubuntu)
Fix Released
Medium
Scott James Remnant (Canonical)

Bug Description

As discussed via email, acpi-support currently brings up all auto interfaces (ifup -a) rather than the ones that it actually brought down before suspending.

Matt Zimmerman (mdz)
Changed in acpi-support:
assignee: nobody → keybuk
Revision history for this message
Paul Sladen (sladen) wrote :

Currently it puts anything that looks like a network device to sleep and then brings them *all* up again, regardless of whether they were asleep before.

See also bug #37010.

Revision history for this message
Scott James Remnant (Canonical) (canonical-scott) wrote :

Easy fix

Changed in acpi-support:
status: Unconfirmed → Fix Released
Revision history for this message
Nicolò Chieffo (yelo3) wrote :

can you add a fix on reboot too?
because if I shut down an interface, I don't want it up after a reboot! thanks!

Revision history for this message
Scott James Remnant (Canonical) (canonical-scott) wrote :

If you don't want an interface to come up on boot, remove the "auto" line for it from /etc/network/interfaces

Revision history for this message
Nicolò Chieffo (yelo3) wrote : Re: [Bug 37515] Re: Doesn't restore interface state correctly on resume

I'm sorry I was talking about the card state, not the interface! this
might be the wrong bug...

On 4/6/06, Scott James Remnant <email address hidden> wrote:
> If you don't want an interface to come up on boot, remove the "auto" line for it from /etc/network/interfaces
> --
> Doesn't restore interface state correctly on resume
> https://launchpad.net/malone/bugs/37515
>

Revision history for this message
scoopex (ms-ubuntu) wrote :

Same problem here - Kubuntu 6.10 restores all interfaces which are defined in /etc/network/interfaces.

It seems that Ubuntu restores all network-interfaces after a resume or a suspend2disk - it doesn't matter
if you commented out the "auto eth0" line.

In this case i have two default routes after a resume and a not working internet connection.

My /etc/network/interfaces
---------------------------------------------------------------------------------------------------------------------
auto lo
iface lo inet loopback

auto eth1
iface eth1 inet static
 address 192.168.0.57
 netmask 255.255.255.0
 gateway 192.168.0.254
 wpa-driver wext
 wpa-conf /etc/wpa_supplicant/wpa_supplicant.conf

#auto eth0
iface eth0 inet static
  address 192.168.0.56
  netmask 255.255.255.0
  gateway 192.168.0.254

#auto eth2
iface eth2 inet dhcp

#auto ath0
iface ath0 inet dhcp

#auto wlan0
iface wlan0 inet dhcp

Revision history for this message
Paul Sladen (sladen) wrote :

There's still something slightly amiss with the logic---somebody reported this to me in-person a week ago on their laptop. Because of 'ifdown' and 'ifup' we loose the state of the interface if it was configured by a method that was not 'ifupdown' (network-manager or manual 'ifconfig').

Revision history for this message
Brad Langhorst (brad-langhorst) wrote :

it would be nice to try to do an "ifconfig vmnet1 up" if the ifup fails.
When devices are not listed in /etc/network/interfaces this is a problem.

to work around it, i've added a line
iface vmnet1 inet manual
 to /etc/network/interfaces

brad

Revision history for this message
Yann (lostec) wrote :

A good fix for this is, for wired interface, is to check if a cable is connected/link is up at pre-up stage (mii ethernet transceiver knows). See this fix I commited as it can also cause a long resume from STR problem:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/236066

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.