failed to stop md devices on jaunty restart or shutdown

[pierreblueher]

I am quite new this, please excuse therefore my possibily bad description. I've installed the beta of Jaunty. Everything works fine so far, except the shutdown or restart. After initiating one of these the system hangs with a black screen. This occurs for now on every shutdown respectively restart after the shutdown splash progress bar had been shown. The systems does not really freeze, instead it seem to wait for something. When I press alt-ctrl-del the system responds with a message: failed to shutdown md devices. After that it restarts. Unplugging connected usb devices such as hard disk or web cam are recognize, but do not seem to be the problem issue, since their absence do not resolve the problem. The message is given with a number, but that seems to change ramdomly, for example i had 114.xxx as well as 334.xxx. If u wish to have an exact number please dont hesitate and let me know.

with regards

reported on Ubuntu 9.04 beta

[Alican Kerman (akerman)]

I have the exact same problem. I can also help if more info is needed.

[WiFi Ed (moran-ed)]

Same problem here with Ubuntu 9.04 x64 beta. When I press alt-ctrl-del the system responds with a message: failed to shutdown md devices. After that it restarts. Sometimes I see this:

Killing all remining processes... [fail]
Unmounting local filesystems...
mount: / is busy.

Sony VAIO SR290 notebook

[pierreblueher (pierre-blueher)]

I have some updates concerning this issue. I think what "we" experience is an side effect. On my board I have also the Intel Wifi Link 5300 installed, which got some issues with the Kernel, meaning quite often Kerneloops detected a crash linked to that card. I already filed my reports to the Kerneloops team. I figured out via internet, that people reporting the same problems link to Intel Wifi cards (infact I found someone reporting this problem with the 4965 series), had installed the kernel backports. I did this one as well, since then the problem to the faulty shutdown was no longer present. I am running that system since 5 days this way. The problem concerning the Kernel crash on the other hand was not solved.

I hope this one can be confirmed by anybody. Maybe some config details would be helpful:

Ubuntu 9.04 beta 64 bit,
on a Dell 1525 with a Intel Wifi Link 5300 AGN

[Moritz Naumann (mnaumann)]

Same issue here. When asked to power down or reboot via gnome menu, this process is initiated displaying the ubuntu shutdown screen, but then switches to text mode and only shows a blinking cursur and hangs there until ctrl-alt-break is hit, which will cause it to continue the shutdown/reboot process by means of disabling md devices followed by the actual reboot / power down.

Ubuntu Desktop 9.04 RC1 AMD64
linux-image-2.6.28-11-generic 2.6.28-11.42

This is on a HP Compaq 6730s wih BIOS version 0A...
http://kmuto.jp/debian/hcl/Compaq/6730s
http://www.linlap.com/wiki/HP-Compaq+6730S
http://www.lapspecs.com/wiki/hp-compaq+6730s

...which includes Intel PRO/Wireless 5100 AGN [Shiloh] onboard.
03:00.0 Network controller: Intel Corporation PRO/Wireless 5100 AGN [Shiloh] Network Connection
03:00.0 0280: 8086:4237

I, too, suspect the iwlagn module which drives the wireless NIC may be the may be the root cause of this issue, mostly since it's new and I think I've seen some error message (but can't find thatin the logs now) related to it.

Could be one of these:
iwlagn: Error sending REPLY_ADD_STA: time out after 500ms.
iwlagn: Error sending REPLY_ADD_STA: enqueue_hcmd failed: -5

But it could as well be one of the other modules:
Modules linked in: aes_x86_64 aes_generic binfmt_misc ppdev bridge stp bnep input_poll
dev lp parport snd_hda_intel snd_pcm_oss arc4 snd_mixer_oss ecb snd_pcm snd_seq_dummy snd_seq_oss snd_seq_midi iwlagn iwlcore snd_rawmidi snd_seq_midi_event
uvcvideo snd_seq snd_timer snd_seq_device leds_hp_disk compat_ioctl32 led_class mac80211 videodev joydev snd soundcore psmouse video iTCO_wdt v4l1_compat lis
3lv02d snd_page_alloc usbhid btusb cfg80211 serio_raw intel_agp pcspkr output iTCO_vendor_support fglrx(P) sky2 fbcon tileblit font bitblit softcursor

It's probably not fglrx since the same issue already existed before I activated fglrx.

[Moritz Naumann (mnaumann)]

Apparently it's really iwlagn. After unloading the module (sudo rmmod iwlagn) rebooting works fine. A single line message is printed on screen before the reboot actually takes place, but I am not able to read it.

I will now try the latest wireless module version using http://linuxwireless.org/en/users/Download#Gettingcompat-wirelessonUbuntu hoping this issue is a module issue (not one with Ubuntu kernel patches) and may have been fixed already.

[pierreblueher (pierre-blueher)]

Moritz Naumann schrieb:
> Apparently it's really iwlagn. After unloading the module (sudo rmmod
> iwlagn) rebooting works fine. A single line message is printed on screen
> before the reboot actually takes place, but I am not able to read it.
>
> I will now try the latest wireless module version using
> http://linuxwireless.org/en/users/Download#Gettingcompat-
> wirelessonUbuntu hoping this issue is a module issue (not one with
> Ubuntu kernel patches) and may have been fixed already.
>
>
before for u start to recompile the kernel, did u try the backports? for
me it did solve the problem.

[Moritz Naumann (mnaumann)]

For me, too, installing linux-backports-modules-jaunty 2.6.28.11.15 is a successful workaround for this issue. I have rebooted a couple times, but not had this issue again.

So there is a workaround, what's missing is a fix.

[perfectska04 (perfectska04-deactivatedaccount)]

I have this same issue, and installing the backports modules does fix it - however, my wireless (Intel 3945ABG) stops working whenever the backports modules are installed. It can see networks, but it can never connect to one.

So far, no other workaround has fixed the shutdown/reboot problem, and the one that does - completely breaks wireless.

[pierreblueher (pierre-blueher)]

As far as I figured a certain lbm firmware package has to be activated
to get the wifi card correctly working. mine are found under
/lib/firmware/2.6.28-11-generic, where i have three files ending with
ucode. for me it is lbm-iwlwifi-5000-1.ucode to get my card properly
working. According to intel the "-1" signals that this firmware is valid
for the installed wifi card. "-2" invalidates the firmware for the
current wifi card. maybe u can have a look if this one activates ur wifi
card by activating ur firmware setting the right number ("-1"), if not
already done by default. may it helps.

perfectska04 schrieb:
> I have this same issue, and installing the backports modules does fix it
> - however, my wireless (Intel 3945ABG) stops working whenever the
> backports modules are installed. It can see networks, but it can never
> connect to one.
>
> So far, no other workaround has fixed the shutdown/reboot problem, and
> the one that does - completely breaks wireless.
>
>

[perfectska04 (perfectska04-deactivatedaccount)]

@pierreblueher
Thanks for the info. I tried renaming the files to the correct number and restarted, but wireless was still broken as long as the backports are installed.

I guess I'll have to roll back to intrepid for now, and hope a future kernel upgrade fixes this.

[Tomoharu Fujita (tomoharu-fujita)]

that workaround looks working for my dell m1330 laptop (iwl3945) and wicd.
reboot/shutdown now works properly.

perfectska04 schrieb:
> I have this same issue, and installing the backports modules does fix it
> - however, my wireless (Intel 3945ABG) stops working whenever the
> backports modules are installed. It can see networks, but it can never
> connect to one.
>
> So far, no other workaround has fixed the shutdown/reboot problem, and
> the one that does - completely breaks wireless.
>

[Raúl Núñez de Arenas Coronado (dervishd)]

Same problem here, with Jaunty final and Intel WifiLink 5100 wireless card.

I still haven't tried the workaround, but the symptoms are the same, so...

[rwwoodjr (rwwoodjr04)]

Used the single item work around "sudo rmmod iwlagn" works fine on my M1330 with 9.04. It did disconnect the wireless connection upon install - but after a reboot everything works just fine (wireless included).

OAG

[utnr6 (neileritchey)]

read this from the toshiba linux support portal, havent tried it yet but it worked for me on 8.10....

i just don't know what it all means

 sudo modprobe -r iwl4965

cd /etc/modprobe.d/

touch iwl4965

gedit iwl4965

In the file enter the following

alias wlan0 iwl4965

options iwl4965 disable_hw_scan=1

modprobe iwl4965

ifconfig wlan0 up

http://linux.toshiba-dme.co.jp/linux/eng/pc/sat_PSU40_report.htm
reported by TOSHIBA of Canada Limited

[jonbonjovi (jonbonjovi-84)]

I can confirm that typing "sudo rmmod iwlagn" in the terminal before shutting down does work around the problem. I'm running Ubuntu 9.04 on an HP dv5.

[Marc G. (marc-gu)]

After installing Jaunty, I had the same problem as described there :
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/71004
(a white screen after shutdown which start to darken in some area...)

I removed the 'splash' option in menu.lst but kept the 'quiet' option. Instead of a white screen, I got the problem described in this bug by pierreblueher and WiFi Ed. After removing the 'quiet' option, the bug reappeared only in some rare occasion. When the computer shutdown succesfully, I can read in the console : "iwl3945: Error : Response NULL in 'REPLY_ADD_STA'".

I have Ubuntu Jaunty with kernel 2.6.28-13 install on a Dell Inspiron 1520. As you can see in the error message, I have an intel 3945 wireless card. I hope that can help you solving this bug.

[ju (clemenzi-santo-gmail)]

Identical problem for me: ubuntu jaunty 64bit.

If I stop the wireless card (intel 3945) before shutdown the system it works.

[Saivann Carignan (oxmosys)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. This particular bug has already been reported and is a duplicate of bug 365733, so it is being marked as such. Please look at the other bug report to see if there is any missing information that you can provide, or to see if there is a workaround for the bug. Additionally, any further discussion regarding the bug should occur in the other report. Feel free to continue to report any other bugs you may find.