(intrepid) When coming out of hibernate/suspend (resume) in a different physical location, wireless networks are not updated.

Bug #264683 reported by Daryl Lublink
96
This bug affects 14 people
Affects Status Importance Assigned to Milestone
network-manager (Baltix)
Invalid
Undecided
Unassigned
network-manager (Ubuntu)
Invalid
Medium
Unassigned
Nominated for Intrepid by Daryl Lublink
Nominated for Jaunty by Jonathan Rascher
Nominated for Lucid by Nilbus

Bug Description

To reproduce :

1. Connect to a wireless network on Ubuntu Intrepid
2. Hibernate the computer
3. Move computer to a new location where the original network is not available and a new network is available
4. Resume the computer

What happens :

 Network Manager Applet shows only the original wireless network and tries to reconnect to it.

What should happen :
 Network Manager Applet should show the networks in the new location and not try and connect to a wireless network that is not available.

What I think is happening :
I think the Network Manager is not scanning for new wireless networks when it resumes.

Revision history for this message
Daryl Lublink (dlublink) wrote :
Download full text (4.0 KiB)

david@david-laptop-intrepid:~$ lsmod
Module Size Used by
ipv6 264228 8
af_packet 25728 4
radeon 147616 2
drm 86056 3 radeon
rfcomm 44432 2
l2cap 30336 9 rfcomm
bluetooth 61540 4 rfcomm,l2cap
ppdev 15620 0
acpi_cpufreq 15500 0
cpufreq_powersave 9856 0
cpufreq_stats 13188 0
cpufreq_userspace 11396 0
cpufreq_ondemand 14988 1
cpufreq_conservative 14600 0
freq_table 12672 3 acpi_cpufreq,cpufreq_stats,cpufreq_ondemand
sbs 19464 0
pci_slot 12552 0
sbshc 13440 1 sbs
iptable_filter 10752 0
ip_tables 19600 1 iptable_filter
x_tables 22916 1 ip_tables
lp 17156 0
ath5k 107904 0
mac80211 217076 1 ath5k
led_class 12164 1 ath5k
cfg80211 32392 2 ath5k,mac80211
pcmcia 43052 0
wlan_scan_sta 20608 1
ath_rate_sample 19968 1
joydev 18368 0
snd_intel8x0 37532 4
snd_ac97_codec 111780 1 snd_intel8x0
ac97_bus 9856 1 snd_ac97_codec
snd_pcm_oss 46976 0
snd_mixer_oss 22784 2 snd_pcm_oss
snd_pcm 83204 3 snd_intel8x0,snd_ac97_codec,snd_pcm_oss
psmouse 45200 0
serio_raw 13444 0
parport_pc 39204 1
parport 42604 3 ppdev,lp,parport_pc
ath_pci 99096 0
yenta_socket 31756 3
rsrc_nonstatic 19072 1 yenta_socket
snd_seq_dummy 10884 0
pcmcia_core 43540 3 pcmcia,yenta_socket,rsrc_nonstatic
wlan 211824 4 wlan_scan_sta,ath_rate_sample,ath_pci
ath_hal 198864 3 ath_rate_sample,ath_pci
container 11520 0
video 25104 0
output 11008 1 video
snd_seq_oss 38528 0
snd_seq_midi 14336 0
snd_rawmidi 29952 1 snd_seq_midi
snd_seq_midi_event 15232 2 snd_seq_oss,snd_seq_midi
snd_seq 57776 6 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_seq_midi_event
wmi 14504 0
battery 18436 0
ac 12292 0
button 14224 0
snd_timer 29960 2 snd_pcm,snd_seq
snd_seq_device 15116 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi,snd_rawmidi,snd_seq
snd 63140 16 snd_intel8x0,snd_ac97_codec,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_seq_oss,snd_rawmidi,snd_seq,snd_timer,snd_seq_device
soundcore 15328 2 snd
snd_page_alloc 16136 2 snd_intel8x0,snd_pcm
intel_agp 33724 1
shpchp 37908 0
pci_hotplug 35236 1 shpchp
iTCO_wdt 18596 0
iTCO_vendor_support 11652 1 iTCO_wdt
agpgart 42184 2 drm,intel_agp
evdev 17696 11
reiserfs 238848 1
sg 39732 0
sr_mod 22212 0
cdrom ...

Read more...

Revision history for this message
Daryl Lublink (dlublink) wrote :

david@david-laptop-intrepid:~$ lspci
00:00.0 Host bridge: Intel Corporation 82855PM Processor to I/O Controller (rev 03)
00:01.0 PCI bridge: Intel Corporation 82855PM Processor to AGP Controller (rev 03)
00:1d.0 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #1 (rev 03)
00:1d.1 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #2 (rev 03)
00:1d.2 USB Controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) USB UHCI Controller #3 (rev 03)
00:1d.7 USB Controller: Intel Corporation 82801DB/DBM (ICH4/ICH4-M) USB2 EHCI Controller (rev 03)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev 83)
00:1f.0 ISA bridge: Intel Corporation 82801DBM (ICH4-M) LPC Interface Bridge (rev 03)
00:1f.1 IDE interface: Intel Corporation 82801DBM (ICH4-M) IDE Controller (rev 03)
00:1f.3 SMBus: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) SMBus Controller (rev 03)
00:1f.5 Multimedia audio controller: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller (rev 03)
00:1f.6 Modem: Intel Corporation 82801DB/DBL/DBM (ICH4/ICH4-L/ICH4-M) AC'97 Modem Controller (rev 03)
01:00.0 VGA compatible controller: ATI Technologies Inc RV350 [Mobility Radeon 9600 M10]
02:04.0 Ethernet controller: Atheros Communications Inc. AR5212/AR5213 Multiprotocol MAC/baseband processor (rev 01)
02:06.0 CardBus bridge: O2 Micro, Inc. OZ711M3/MC3 4-in-1 MemoryCardBus Controller
02:06.1 CardBus bridge: O2 Micro, Inc. OZ711M3/MC3 4-in-1 MemoryCardBus Controller
02:06.2 System peripheral: O2 Micro, Inc. OZ711Mx 4-in-1 MemoryCardBus Accelerator
02:06.3 CardBus bridge: O2 Micro, Inc. OZ711M3/MC3 4-in-1 MemoryCardBus Controller
02:0e.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5705M_2 Gigabit Ethernet (rev 03)

Revision history for this message
Daryl Lublink (dlublink) wrote :

The interface is called ath0, I believe it uses the atheros drivers.

Revision history for this message
Daryl Lublink (dlublink) wrote :

Given that Intrepid is aimed to better mobile computing and this severely impairs that ability, I would say this bug should have an important of 'Major'.

Revision history for this message
Alexander Sack (asac) wrote :

please attach the complete /var/log/syslog taken directly after reproducing this bug.

Changed in network-manager-applet:
status: New → Incomplete
Revision history for this message
Alexander Sack (asac) wrote :

also state which exact version of network-manager you are using

Revision history for this message
Daryl Lublink (dlublink) wrote :

root@david-laptop-interpid:/home/david# apt-cache show network-manager
Package: network-manager
Priority: optional
Section: net
Installed-Size: 1924
Maintainer: Ubuntu Core Dev Team <>
Original-Maintainer: Riccardo Setti <>
Architecture: i386
Version: 0.7~~svn20080908t183521+eni0-0ubuntu2
Replaces: network-manager-pptp (<< 0.7~~)
Depends: libc6 (>= 2.4), libdbus-1-3 (>= 1.0.2), libdbus-glib-1-2 (>= 0.74), libglib2.0-0 (>= 2.16.0), libhal1 (>= 0.5.8.1), libnl1, libnm-glib0 (>= 0.7~~svn20080908), libnm-util0 (>= 0.7~~svn20080908), libnspr4-0d (>= 1.8.0.10), libnss3-1d (>= 3.12.0~1.9b1), libpolkit-dbus2 (>= 0.7), libpolkit2 (>= 0.7), libuuid1, iproute, iputils-arping, lsb-base (>= 2.0-6), wpasupplicant (>= 0.6.1~), dbus (>= 0.60), hal (>= 0.5.7.1), update-notifier-common
Recommends: network-manager-gnome | network-manager-kde
Conflicts: network-manager-pptp (<< 0.7~~)
Filename: pool/main/n/network-manager/network-manager_0.7~~svn20080908t183521+eni0-0ubuntu2_i386.deb
Size: 247934
MD5sum: 750a69fd8b0da0ab0e97c4333c9352ed
SHA1: 1dfd67c14620d8a41c8f36ed906b4365d49bc264
SHA256: f3e3dbcbf7a9578f3dacc7141e7fdbce5e0ff17c6c4d609314b39dda444a3561
Description: network management framework daemon
 NetworkManager attempts to keep an active network connection available at all
 times. It is intended only for the desktop use-case, and is not intended for
 usage on servers. The point of NetworkManager is to make networking
 configuration and setup as painless and automatic as possible. If using DHCP,
 NetworkManager is _intended_ to replace default routes, obtain IP addresses
 from a DHCP server, and change nameservers whenever it sees fit.
 .
 This package provides the userspace daemons.
 .
  Homepage: http://www.gnome.org/projects/NetworkManager/
Bugs:
Origin: Ubuntu
Task: ubuntu-desktop, kubuntu-desktop, edubuntu-desktop, edubuntu-desktop-kde, xubuntu-desktop

Revision history for this message
Daryl Lublink (dlublink) wrote :

I included the complete syslog so you have complete information, but you can see using tail -n 20 syslog where it fails to connect.

In this case, it can not even connect to the same wireless access point as before the hibernate. I will attach syslog from a situation where I change physical locations this evening.

Changed in network-manager:
status: Incomplete → New
Revision history for this message
Daryl Lublink (dlublink) wrote :

I also notice that for the past few weeks it has prompted me to connect to a wireless network even though it already has the information. Is this a new feature or a bug?

Revision history for this message
Ruben Verweij (ruben-verweij) wrote :

I am also using the atheros drivers on a hp compaq nc6000 with Intrepid and nm-applet 0.7.0 and I can confirm that after hibernating, it can not connect to the same wireless network as before the hibernate. It only works after rebooting again.

Alexander Sack (asac)
Changed in network-manager:
importance: Undecided → Medium
status: New → Confirmed
Revision history for this message
Alexander Sack (asac) wrote :

bug 289796 is somewhat related ... but i think its not a dupe.

Revision history for this message
Red Five (nelson-butterworth) wrote :

I can confirm this issue even without moving to another SSID. I hibernate at home all the time, and NM 0.7.0 won't find and connect to my home AP after resuming from either suspend or hibernate. I have an Atheros WLAN card, and this is a completely fresh install of Intrepid final.

I figured out a workaround for the time being. I wrote a script to do the following, with both my WLAN and LAN adaptors:

1. Stop the NetworkManager service
2. Ifconfig down both eth0 and ath0 (I sometimes dock my Latitude, so I do occasionally use a wire)
3. Unload ath_pci and tg3 (the wired LAN driver)
4. Reload ath_pci and tg3
5. Ifconfig up eth0 and ath0
6. Start NetworkManager

I run the script after resuming, and in about 20 seconds, I'm back online. I tried making adjustments to the scripts in /etc/acpi, but nothing seemed to work, so I run this script manually every time I resume.

UPDATE: I just saw some info for the NM PPA, so I think I'll try using it to update my install.

Revision history for this message
Red Five (nelson-butterworth) wrote :

I just loaded the version available as of this writing from the NM PPA, and it doesn't seem to fix this issue. I rebooted after the update, logged back in and let everything settle down. I then suspended the machine, and woke it back up after about 1 minute, and it was apparent that it wasn't going to reconnect. So I ran my script, and it hooked right up, easy-peasy.

Revision history for this message
Sasquatch (sasquatch) wrote :

I don't use Network Manager, but instead just the interfaces file. My wireless is managed by wpa_supplicant and that too does not see new wireless networks. It 'connects' to the last wireless it was connected if it was connected before the hibernate.

It seems that the interfaces file isn't reloaded, but instead are brought down and up again using 'ifconfig'. This is very annoying, as I use static at home and DHCP at school and wrote a script specifically for that. Now it doesn't work as it should.

So it isn't an NM specific issue, but a global resume problem.

Revision history for this message
mark (notmarkus) wrote :

I'm running Kubuntu Intrepid on an EeePC 900 and if I choose Suspend to RAM, when the computer comes back from suspend, I have this same problem. However, if I hit the Fn+F1 "Sleep" hotkey, I don't have this problem. Does anyone know the scripts involved in these two processes? Comparing them may shed some light on the issue.

Revision history for this message
azdine (azdine-t) wrote :

Same problem for me, running Ubuntu Intrepid on an HP NC6000 using the atheros drivers: when I come out from hibernate: 1) wifi networks listed by network manager are wrong, 2) I can't reconnect to any wifi network.

The simplest workaround I found is to run "sudo rmmod ath_pci, followed by sudo modprobe ath_pci

Revision history for this message
Red Five (nelson-butterworth) wrote :

Has anyone seen this yet on anything other than Atheros mini-PCI cards? I've got a buddy running Intrepid on a ThinkPad tablet, using a Broadcom wifi adapter, and all he has to do is disable then enable networking from nm-applet. I also installed xubuntu 8.10 on a much older Sony VAIO that has an Atheros-based PC Card wlan adaptor (D-Link DWL-630G), and it doesn't seem to have any problems whatsoever with suspend or hibernate.

Revision history for this message
Sasquatch (sasquatch) wrote :

I have it with my Intel Pro Wireless 3945 ABG card. All I need to do is restart my networking (either init.d/networking restart or bring the interface down and back up again). Again, I don't use Network Manager.

Revision history for this message
Peterson Silva (petersonsilva) wrote :

Also happens to me - Ubuntu 8.10 in an Acer Aspire 3620 series - Atheros card. I can't reconnect to wireless networks after hibernating.

Don't know if it happens with suspend also though, or whether I'm able reconnect on same connection...

Revision history for this message
Jurgis Pralgauskis (jurgis-pralgauskis) wrote :

same on Asuss Z99H after suspend to RAM
doesn't connect to Wireless any more :/

what helped:
sudo rmmod ath_pci
sudo modprobe ath_pci

network-manager: 0.7~~svn20081018t105859-0ubuntu1

Revision history for this message
Peterson Silva (petersonsilva) wrote :

Doesn't work on suspend either. The same network, same physical place, but I can't reconnect after suspend.

This should be fixed really fast. As someone said before, if 8.10 is about mobility, then we can't let this go further...

Revision history for this message
Daryl Lublink (dlublink) wrote :

I also notice that on a normal boot, it no longer automatically connects. I have to hit "connect" each time. Is this happening to anyone else?

Revision history for this message
Red Five (nelson-butterworth) wrote :

Mine connects automatically, once I unload and reload the ath_pci driver. I have to do that after suspend/resume or hibernate. I was able to determine that I could just reload the driver, rather than having to down the interface, reload the driver, then restart NM. I looked at all the ACPI scripts, and they indicate that they should unload all the network drivers during suspend/hibernate and reload upon resume, but that doesn't seem to be happening with the Atheros driver. Maybe that means it's either an ACPI problem or an Atheros problem. I remember reading last week that there's a new Atheros driver available; I may compile it this evening and do some testing.

Revision history for this message
Red Five (nelson-butterworth) wrote :

I just found an interesting note at the following link: http://www.ubuntu.com/getubuntu/releasenotes/810#Wireless%20doesn%27t%20work%20after%20suspend%20with%20ath_pci%20driver

I'm going to try this tonight, before I waste any time with the ath5k driver.

Revision history for this message
Red Five (nelson-butterworth) wrote :

The link I posted earlier today works. For those interested and using Atheros cards, just create this file: /etc/pm/config.d/madwifi with contents "SUSPEND_MODULES=ath_pci". Then suspend or hibernate as you desire, and it should bring the network back up right away. In my case, it was back even before I unlocked the screen from hibernation.

Revision history for this message
Sasquatch (sasquatch) wrote :

Red Five:
That might be with your Atheros card, but do you also have a solution for an Intel Wireless card (could be as simple as doing your atheros fix for iwlwifi)?

Did any of you also check your wired connection with it's config? I doubt that if you put a static address on it, then run dhclient on it, go to hibernate and when you resume, that you have your static address like you entered in the /etc/network/interfaces file. That is what I have too, my wired config isn't reset either. It comes back with whatever IP setting was on it before going to hibernate (in this case, a DHCP address).
This is without the use of Network Manager, because NM applet would request a DHCP address or set the static address once it detects a link on it, but that isn't done if you don't run NM. This was done properly on Hardy. In fact, I think that even if you set your wired to DHCP, it will not request a new IP, which is troublesome if you move to a different subnet.

Revision history for this message
Peterson Silva (petersonsilva) wrote :

David Lublink:
No, boot is ok for me.

Red Five:
Will test it soon...

Revision history for this message
Jurgis Pralgauskis (jurgis-pralgauskis) wrote :
Revision history for this message
Peterson Silva (petersonsilva) wrote :

"Suspended modules" worked for me perfectly, at least in suspend. Didn't test yet with hibernate, but, can I assume...?

Revision history for this message
robert1968 (robert1968) wrote :

Many thanks to Red Five!!

On Fujitsu C500 I also had the "no wireless after suspend" issue, but
with the solution proposed by Red Five :
          create a file /etc/pm/config.d/madwifi with contents "SUSPEND_MODULES=ath_pci"

Everything is perfect.
:)))

Thanks again.
Robert

Revision history for this message
Chris Bainbridge (chris-bainbridge) wrote :

bug #275692 is the one that Red Five's fix relates to.

Revision history for this message
vetdoc (m-rihm) wrote :

Same problem here. ISL3886 (Prism Javelin/Prism Xbow) on a Fujitsu Siemens Amilo L1300. Network Manager applet 0.7.0. Static IP. After suspend NM is working and working and working and finally shows the little red cross, indicating no network connection. After clicking on applet and trying to connect manually (had to type in password again?!) several times, system freezes completely. CPU at 100 %, no reaction at all, I had to hard reboot. Not tested with hibernation.

Revision history for this message
vetdoc (m-rihm) wrote :

Sorry, I forgot: Xubuntu 8.10 fresh install (yesterday). And no sound either ... (OT)

Revision history for this message
Ruben Verweij (ruben-verweij) wrote :

Red Five's fix also works on a HP compaq nc6000, after changing the file all the problems cease to exist.

Revision history for this message
Sasquatch (sasquatch) wrote :

There was a link to a different bug report about network manager (I believe that Red Five posted that). It mentioned the file /usr/lib/pm.config/10networkmanager to change something. As I don't use network manager, I picked the 50modules file, and in the resume|thaw line, I added the command to restart the network. That worked for me.

Revision history for this message
Dereck Wonnacott (dereck) wrote :

Red's fix worked perfect for me as well, I wonder if we can't get this in for Jaunty?

Attached: lspci - vv

Changed in network-manager:
status: Confirmed → Triaged
Revision history for this message
Nilbus (nilbus) wrote :

Worked for me with SUSPEND_MODULES=ndiswrapper

Revision history for this message
Tony Espy (awe) wrote :

I'm going to change the status of this bug to Incomplete, as the last comment on this bug was made in March.

If testing proves that this is a generic problem ( ie. it happens across multiple drivers ) in Ubuntu Karmic / 9.10, then this should be re-opened.

Please note that there is a driver-specific bug ( bug #452571 ) that's very similar to this that describes the same issue with systems running Karmic and the Broadcom 'wl' / STA driver.

Changed in network-manager (Ubuntu):
status: Triaged → Incomplete
Revision history for this message
Daryl Lublink (dlublink) wrote :

Hey,

I no longer have that computer, I broke it a few weeks ago. I also went back to Hardy Haron, so I do not know if the bug still happens.

David

Revision history for this message
papukaija (papukaija) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. You reported this bug a while ago and there hasn't been any activity in it recently. We were wondering if this is still an issue for you. Can you try with the latest Ubuntu release? Thanks in advance.

Revision history for this message
Jon-o Addleman (jaddle) wrote :

This seems to be fixed in Karmic on my thinkpad X60s (iwl3945 driver). Thanks!

Revision history for this message
Daryl Lublink (dlublink) wrote :

I don't have this computer any more, so I don't know if it is still happening.

Revision history for this message
papukaija (papukaija) wrote :

This bug report is being closed due to your last comment regarding this
being fixed with an update. For future reference you can manage the
status of your own bugs by clicking on the current status in the yellow
line and then choosing a new status in the revealed drop down box. You
can learn more about bug statuses at https://wiki.ubuntu.com/Bugs/Status
. Thank you again for taking the time to report this bug and helping to
make Ubuntu better. Please submit any future bugs you may find.

Changed in network-manager (Baltix):
status: New → Invalid
Changed in network-manager (Ubuntu):
status: Incomplete → Invalid
Revision history for this message
Jamie Lokier (jamie-shareable) wrote :

I am under the impression this was a genuine bug, and it was fixed last year.
It was fixed in the kernel, in generic wireless code.
Upgrading the kernel is enough to fix it.

See "Making NetworkManager work with suspend/resume", http://lwn.net/Articles/321102/

This is consistent with my experience: The bug was fixed for me when I upgraded Ubuntu 9.04 to 9.10.
I didn't comment because there was nothing more to add.

Therefore this bug should be marked as Fixed, rather than Invalid.

Revision history for this message
Jamie Lokier (jamie-shareable) wrote :

Btw, Fixed rather than Invalid, because it's still an issue when using Ubuntu 9.04 and earlier.

Revision history for this message
Lanchon (lanchon) wrote :

i'm using ubuntu 9.10 on a dell mini 9 and have this problem. according to what i've read in a hurry, i think the problem might be the way the wifi driver timestamps the networks it detected and not network manager itself. btw, i'm using the non-free driver that ubuntu offered me.

anyway, i just wanted to post a dirty workaround i hacked that fixes the problem for me. just drop the attached file in the /usr/lib/pm-utils/sleep.d directory, it stops and restarts the net manager during suspend and resume.

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

Duplicates of this bug

Other bug subscribers

Remote bug watches

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