Several Huawei USB dongle don't work with kernel 2.6.31-12.40 (2.6.31.1 update related)

Affects me also with Huawei E169 and eeePC 1000H

Ubuntu 9.10 beta updated today
kernel: 2.6.31-12-generic #41
network-manager: 0.8~a~git.20091005t192303.1d28ad1-0ubuntu2
modemmanager: 0.2.git.20091003t063318.aa78b5f-0ubuntu1

Exactly the same issue on two machines after upgrading to 2.6.31-12.40/41 generic. Retaining kernel 2.6.31-11 as workaround and hoping later updates solve this problem.

Kernel is seeing dongle as a usb drive and not a modem. I'm sure its the kernel upgrade because NM and modem manager are the same versions.

BTW karmic alpha/beta is working very well for me, apart from this problem.

Problem exists on 2.6.31-13-generic #42

AND on Ubuntu Netbook Remix daily live (.iso) of 9-Oct-09 where Huawei E169 cannot be used at all as there is no previous kernel to boot!

Confirming problem remains with latest kernel 2.6.31-13.42.

Updating to 2.6.31-13#43 makes no difference

Below is part of dmesg (2.6.31-13#43) showing the 'loop':

g@KKsep22:~$ dmesg
[ 1261.396207] usb 2-2: new full speed USB device using uhci_hcd and address 4
[ 1261.616651] usb 2-2: configuration #1 chosen from 1 choice
[ 1261.634400] scsi86 : SCSI emulation for USB Mass Storage devices
[ 1261.637423] usb-storage: device found at 4
[ 1261.637438] usb-storage: waiting for device to settle before scanning
[ 1261.744264] usb 2-2: USB disconnect, address 4
[ 1262.480201] usb 2-2: new full speed USB device using uhci_hcd and address 5
[ 1262.645650] usb 2-2: configuration #1 chosen from 1 choice
[ 1262.653302] option 2-2:1.0: GSM modem (1-port) converter detected
[ 1262.653695] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB0
[ 1262.656890] option 2-2:1.1: GSM modem (1-port) converter detected
[ 1262.657342] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB1
[ 1262.666827] option 2-2:1.2: GSM modem (1-port) converter detected
[ 1262.667166] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB2
[ 1262.672207] scsi90 : SCSI emulation for USB Mass Storage devices
[ 1262.678576] usb-storage: device found at 5
[ 1262.678585] usb-storage: waiting for device to settle before scanning
[ 1267.678453] usb-storage: device scan complete
[ 1267.681422] scsi 90:0:0:0: CD-ROM HUAWEI Mass Storage 2.31 PQ: 0 ANSI: 2
[ 1267.684423] scsi 90:0:0:1: Direct-Access HUAWEI SD Storage 2.31 PQ: 0 ANSI: 2
[ 1267.715350] sr0: scsi-1 drive
[ 1267.715798] sr 90:0:0:0: Attached scsi CD-ROM sr0
[ 1267.716193] sr 90:0:0:0: Attached scsi generic sg2 type 5
[ 1267.716944] sd 90:0:0:1: Attached scsi generic sg3 type 0
[ 1267.749246] option: option_instat_callback: error -108
[ 1267.749530] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
[ 1267.749596] option 2-2:1.0: device disconnected
[ 1267.752224] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
[ 1267.752271] option 2-2:1.1: device disconnected
[ 1267.755231] option1 ttyUSB2: GSM modem (1-port) converter now disconnected from ttyUSB2
[ 1267.755279] option 2-2:1.2: device disconnected
[ 1267.864140] usb 2-2: reset full speed USB device using uhci_hcd and address 5
[ 1268.011659] option 2-2:1.2: GSM modem (1-port) converter detected
[ 1268.011950] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB0
[ 1268.012354] option 2-2:1.1: GSM modem (1-port) converter detected
[ 1268.012698] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB1
[ 1268.013127] option 2-2:1.0: GSM modem (1-port) converter detected
[ 1268.013513] usb 2-2: GSM modem (1-port) converter now attached to ttyUSB2
[ 1268.091233] option: option_instat_callback: error -108
[ 1268.099009] option1 ttyUSB2: GSM modem (1-port) converter now disconnected from ttyUSB2
[ 1268.099071] option 2-2:1.0: device disconnected
[ 1268.099265] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
[ 1268.099322] option 2-2:1.1: device disconnected
[ 1268.099495] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
[ 1268.099547] option 2-2:1.2: device disconnected
[ 1268.212136] usb 2-2: reset full speed USB dev...

Affects me to. Seems like the USB modem is mounted as an USB-memory stick. I can actually open the disk in Nautilus and browse the content. I tried to download http://www.greenhughes.com/files/e169g-switch_0.3_all.deb which resolved the issue for me in earlier versions of Ubuntu. The modem has been working flawless up until last week when I upgraded to 2.6.31-12 and today I upgraded to 2.6.31-13 and the modem is still being seen as usb-disk.

When inserting the usb-modem I get the following error message:

Unable to mount Tele2 3G-modem

Error mounting: mount exited with exit code 32: Mount: block
device /dev/sr1 is write protected, mounting read-only
mount: /dev/sr1 already mounted or /media/Tele2 3G-
modem busy

We identified the bug upstream, it's a 2.6.31.1 - >2.6.31.2 regression. Still looking for a proper fix (yeah, I have one of these critters too).

Discussion is mostly happening on lkml but we'll keep http://bugzilla.kernel.org/show_bug.cgi?id=14355 updated when a solution is found, hopefully in the next couple of days.

Thanks, Benjamin. Good luck with the fix. I have retained 2.6.31-11 and I am updating other kernels as they appear in repos.

2.6.31-13.44 doesn't fix the bug for me

We now know what the problem is, still debating how to properly fix it

The attached patch should make it work again, though it's a bit of a hack and probably not
what the final solution will be. In case you really don't want to stick to 2.6.31.1 :-)

Hi Benjamin, I do not want to 'gain' from your knowledge but if possible can you take a look to bug:
https://bugs.launchpad.net/bugs/408555
that has a similar 'looping' on creating/disconnecting ttyUSBx with ZTE MF636? The difference there is that ttyUSBx is advanced (+1) at a time reaching ttyUSB8-9-... and finally the system gets frozen.
Thanks in advance,
George

Ok I have taken the dirty fix benh referenced above and applied it to the karmic kernel and built some kernels for testing. Could those of you affected by this issue please try these kernels and report back here. The kernels can be found at the URL below:

     http://people.canonical.com/~apw/lp446146-karmic/

Andy's #44lp446146apw1 fixes the problem for me.

The problem is also solved for me using Andy's kernel.

Same good results for my E169 with 2.6.31-13-generic #44lp446146apw1

My problem is fixed with this Andy's #44lp446146apw1. Thank you!

Works for me too, except the E169 also loads as a USB drive.

Thanks

i got this to with my E220 usb modem using kernel 2.31.13 , hope this gets fixed in the future

Hi Russell Faull,
my opinion is that USB drive (virtual CDROM/ZeroCD) must exist but possibly not mounted automatically. It is not good to reject it as in the future a manufacturer may want to include any Linux (open source better) application program to control/update their modem (ex. like betavine s/w). The user can add a rule to eject/unmount it automatically. The same and more important is to get access on microSD card reader.
Regards,
George

Hi George

Good point about future Linux software, I hope more manufacturers adopt open source.

Perhaps not for this thread, but how do I control mount / unmount automatically, as you suggest?

Regards and thanks
Russell

Hi all,

I can confirm the issues with E220 and E270, although I haven't tested the fixes yet. Will try to update my box today/tomorrow ...

regards,
FraGe

Note that 2.6.31-13-generic #44lp446146apw1 is just '2.6.31-13' for grub.
A newer update will force 2.6.31-13-generic #45 to be used instead of the patched and the problem will come back!

@Russell Faull, post anything at http://ubuntuforums.org/showthread.php?t=1202430 to discuss.

I have an E220 and get the following whenever I attempt to use it;

Unable to mount Mobile_Connect
Error mounting: mount exited with exit code 32: mount: block device /dev/sr0 is write-protected, mounting read-only mount: /dev/sr0: can't read superblock

Bug remains with 2.6.31-14 #46 (Change log not available to the time of writing this post).

Thanks to GeorgeVita for warning about *31-13, which is now 'pinned'.

Here's the "final" patch that unless major problem or end of the world,
Greg will send to Linus and apply to -stable tomorrow:

http://marc.info/?l=linux-kernel&m=125540979118722&w=2

Just got latest kernal update and the problem still persists. Have to go back to -11 to use the USB dongle!

2.6.31-14.46 is still affected by the bug, and today it's kernel freeze time...
Does milestone -> karmic-updates mean that a fix for this bug will be released as a kernel update after the release of 9.10 ?
If the dongle is your only Internet connection you will have to save the kernel update packages on a flash disk and use them to upgrade after you installed 9.10.

This is indeed a concern. Those dongles are quite common, at least here in .au

My latest patch was accepted and is now in gregkh tree, it should hit the next stable release and Linus soon now, might be worth making sure it's in the released kernel

Loaded just new updates from main server.
still doesn't work.

It gives this error:
Error mounting: mount exited with exit code 32: mount: block device /dev/sr1 is write-protected, mounting read-only
mount: /dev/sr1 already mounted or /media/Mobile Partner busy

Without any updates only cd-install of karmic it works good.

It's a big problem, Those dongles are very common in Italy too

Patch now upstream ID b8430e1b82b7e514d76a88eb70a7d8831d50df1e and queued for next -stable

Ben, does this mean the patch is likely to be 2.6.31-14.48? And thanks for solving this annoying bug.

I have no idea in what ubuntu kernel it will show up :-) All I know is the upstream status :-)

This bug was also affecting me, thanks to Andy's .13 Kernel I am now online.
Look forward to a fix before the final release.

The patch is now pulled into our tree. It is most likely to release as an SRU immediatly after release.

Andy, do I understand you correctly that the fix for this problem will only be available in a kernel upgrade sometime after Karmic is available to the general public as a final release on 27 October?

If so, there will be many disappointed users of e169 modems. I expected the Ubuntu will also suffer pretty negative publicity. I hope I am misunderstanding your post..

BTW I am using the kernel with Ben's 'hack' patch you posted at #15. It works fine. Is there a chance of a new kernel complied with the final patch for testing through this forum?

Regards and thanks

SRU (https://wiki.ubuntu.com/StableReleaseUpdates) is unacceptable!

This must be fixed BEFORE final release. Imagine a new user (Ubuntu/linux 'newbie') without any other Internet access:

Get Ubuntu 9.10, install it, use windows to get a new .deb, copy it to your Ubuntu 9.10 Desktop and double click it to install. 'Ubuntu via windows'...

We 'throw away' dial up (no wvdial), we do not want ZTE MF636 (ok this is a minority) now we need to patch LiveCDs for a Huawei 3G modem (very popular worldwide)?

... and I think you understand that this is not OUR problem as we are involved in testing (development phase) and we gain a lot of knowledge (Thanks!). Now we can use over 5 methods to connect with our 3G modems.
What about a newcomer?

If the fix is in the Ubuntu 'tree', it seems a process failure not to be able to propagate it quickly through the repos. Today's kernal #48 added support for an iriver music player -- important, yes, but not as important as connecting to the net! IMHO.

hmm, I thought "released" means, that it is in the kernel now, but a todays daily-build doesn't work either - its not just the 169,
- I guess, nearly every huawei, if not all are affected - I use a 220 and a 270, both don't work right now.

greets,
FraGe

Sorry everyone. Accidentally clicked on the status change to Fixed Released. Reverted to original status of Fix Committed.

Please get this in karmic final itself and NOT as an SRU.

I am sure there are many people like me who depend on 3g modems for net connection and after i install Karmic and then it doesn't work. We'll be stuck. Especially for the new comers it will make a very abd impression.

It's incredibly fiddly, but I can eventually get my Huawei E220 modem going by using this method which I used years ago with some earlier version of Ubuntu...

sudo rmmod usb-storage
sudo modprobe usbserial vendor=0x12d1 product=0x1003

I completely depend upon my 3g connection. If I'd performed a fresh install over my Jaunty or upgraded to Karmic I'd have been left up the creek.

I don't think an average user would either want or be capable of resolving a connection problem like this, and there are MANY 3g users in Ireland. This would be a disaster if left unfixed in the final release.

Please include this fix on the CD, it is absolutely essential for the many people, especially students, who depend on their USB modems for internet connectivity!!

Just want to add yet another comment in favor of getting this in the release and seriously increase the importance. Many many people do use these USB dongles. Vodafone - the biggest pan-European mobile provider - among many others, uses these dongles for mobile Internet in most countries where they operate.
I can imagine lots of people having a crappy first experience with Ubuntu because of this. In fact, I am writing this on a brand new installation of Karmic where the person requested my help to solve this.

Many people in my part of the world (and I suspect elsewhere) are rapidly abandoning fixed line adsl(2) and traditional ISPs for 3G mobile broadband -- Virgin Mobile with Huawei E169 in .au in my case.

With one week to public release of Karmic, don't let it be Ubuntu's Vista. For most people, If it smells like shit, it probably is -- which would be pretty sad. They won't forgive easily and wait for an M$-type 'service pack' (read SRU).

As someone who's only option for affordable broadband access is via HSDPA I would vote for this being including in the official release. At the same time for every ubuntu release since hardy has included a last minute fix which introduced a regression which has bitten me (or a client). If proper QA can not occur between now and the release, I think it is worse to introduce another regression at the last minute than ship a product with a relnoted known issue.

Just a quick followup, I tested the AMD64 kernel post by Andy on 12Oct and it works as advertised. I would also be happy to test a current kernel with the patch too.

I know kernel and QA team are very busy close to release!
I also know that upgrading a kernel to archive so close to release is not a good idea.
But please consider how this will affect a big subset of the users, leaving them stranded without means of upgrade.
I would like to have the kernel and release team considering a critical measure and delay release until the kernel can be properly tested.

I also understand that if we would delay the release for each regression comes up, we would never do it, but this is an huge regression, and there's already a patch that is reported to be working.

Please take this into consideration.

Thanks.

I used to connect with my mobile and this patch doesn't solve my problem. I still can't connect with broadband mobile (no broad band mobile appear on network manager menu). The last time it works was when I have 2.6.28 installed from ubuntu 9.04.

For me it eventually works but only after plugging & unplugging device a few times

With my E169 the

sudo rmmod usb-storage
sudo modprobe usbserial vendor=0x12d1 product=0x1001

worked (other people will find their vendor and product IDs on https://wiki.ubuntu.com/NetworkManager/Hardware/3G)

however I had an additional isssue in that no DNS was picked up - /etc/resolv.conf was empty. This was solved by adding usepeerdns to /etc/ppp/options and then reconnecting.

+1 to those who say these issues must be sorted out prior to a release.

To add insult to injury inserting my E169 kills NetworkManager.
I have to remove my E169 stick before I can restart NetworkManager without having to reboot the system.
Using the commands above I can start NM but the applet does not come back properly and a reboot it needed to bring NM back fully.

i have the same problem with e169 - on amd64 - as every one probaby

Releasing Ubuntu with this kernel would really be a bad idea - if a newbie cannot connect to internet - he will surely erase his ubuntu install and work with windows7 (the 1st good windows for me since win2k)

On Karmic RC, rmmodding e modprobing usb-storage and serial won't work in order to get Huawei E169G Dongle to work. (see attachment for my details).

I won't add a word about how it's *vital* to have this as final behaviour for Karmic when insert the dongle.

Most on this forum agree that a fix is 'vital' for the final release of Karmic. I would be really interested in a view from someone on the release team / kernel team about the importance of mobile broadband access for new or less experienced users with this popular modem.

Come, guys, we have not heard from you since Andy's post #39 on 16 October.

Hi again, I would like to express some thoughts, just to be 'fair' and 'typical':

a. this is not a 'forum' but a 'bug report system'
>>> the 'bug' reported, a fix found and it is on the way to be established

b. most users above CAN FIX the problem as they involved in testing pre-released KK
>>> they/we complain about the BAD situation for a newcomer who has 3G (mobile) or land-wireless broadband as their main internet connection

c. #446146 is a real BUG as in many cases the creation/disconnection of the CDROM and/or ttyUSBx is looping (dmesg has new lines) and sometimes system freezes

d. this bug found 'in time' (8-Oct), discussed and confirmed at ubuntuforums.org and then reported at launchpad following the 'typical' procedure. Confirmed by developers and a patch released. This patch tested and found working (or at least 'working enough') 'in time' (12-Oct)

e. Kernel Freeze came at 15-Oct without this fix!

f. Possibly the purpose of this thread ('bug report') fulfilled after comment #16-20

Some reference links:
Karmic Koala Release Schedule http://ubuntuforums.org/showthread.php?t=1133054
canNOT use Huawei E169 with 2.6.31-12-generic #40 http://ubuntuforums.org/showthread.php?p=8068801

Regards,
George

P.S.: Release of 9.04 had wvdial removed from final .iso and a lot of dial-up users faced a major problem!

George, you are the voice of moderation.

This discussion should have moved to the forums after the fix was known. But the frustration expressed here is precisely because the fix was available before the kernel freeze on 15 October and is based on concern for new users and the reputation of Ubuntu.

Regards

Hi all,

after upgrading to the newest kernel ( http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/ ), downloaded from another machine I finally found a way to connect with my E220 using the mentioned

sudo rmmod usb-storage
sudo modprobe usbserial vendor=0x12d1 product=0x1003

this is ok for a Beta, but it's really inacceptable for the Final. Hope, this gets fixed and in the kernel before the release :(

regards,
FraGe

Looking at the changelog for 2.6.31.5 (http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.31.5) it appears Ben H's fix is included.

I have Linux BluBUG 2.6.31-02063105-generic #02063105 SMP Fri Oct 23 09:08:09 UTC 2009 x86_64 GNU/Linux
and still it doesnt work out of the box
attaching syslog

In my honest opinion, this is reason enough for a postponement of the
release.

What did I do wrong?

jyrki@jyrki-desktop:~$ sudo rmmod usb-storage
ERROR: Module usb_storage is in use
jyrki@jyrki-desktop:~$ sudo rmmod usb-storage
ERROR: Module usb_storage is in use
jyrki@jyrki-desktop:~$ sudo modprobe usbserial vendor=0x12d1 product=0x1001
jyrki@jyrki-desktop:~$

No it doesn't work.
I have E169

It's E169 hdspa

What goes wrong?

jyrki@jyrki-desktop:~$ rmmod usb-storage
ERROR: Module usb_storage is in use
jyrki@jyrki-desktop:~$ sudo umount /dev/sr1
[sudo] password for jyrki:
umount: /dev/sr1: not mounted
jyrki@jyrki-desktop:~$ sudo rmmod usb-storage
ERROR: Module usb_storage is in use
jyrki@jyrki-desktop:~$ sudo modprobe -r usb-storage
FATAL: Module usb_storage is in use.
jyrki@jyrki-desktop:~$ sudo umount /dev/sr1
umount: /dev/sr1: not mounted
jyrki@jyrki-desktop:~$ sudo umount /dev/sr1
umount: /dev/sr1: not mounted
jyrki@jyrki-desktop:~$ sudo modprobe -r usb-storage
FATAL: Module usb_storage is in use.
jyrki@jyrki-desktop:~$
jyrki@jyrki-desktop:~$
jyrki@jyrki-desktop:~$ sudo umount /dev/sr1
umount: /dev/sr1: not mounted
jyrki@jyrki-desktop:~$ sudo modprobe -r usb-storage
FATAL: Module usb_storage is in use.
jyrki@jyrki-desktop:~$ sudo rmmod usb-storage
ERROR: Module usb_storage is in use
jyrki@jyrki-desktop:~$

I have a Huawei E220 USB dongle and I am suffering the same post 2.6.31-11 problems as many others in this list.
The 2.6.31-02063105 from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/ did not solve the issue for me. I have fully functional system with the 2.6.31-11, but anything later will make system reset the dongle continuously. I have tested this with two different laptops.

lsusb identifies the dongle as:
Bus 003 Device 003: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem

Huawei E220 is very popular in Finland, so...

Kernel 2.6.31.5 mentioned by liro R-J works for me with Huawei E169. It should work ok because, according to the changelog, Ben H's patch has been applied.

Is a new bug report needed for the E220?

I haven't had a chance to test 2.6.31.5 on my E169, will do so later today, but it's possible that E220 suffers from a different bug (Huawei seems to have a pretty crappy firmware).

I would suggest you guys file a kernel bugzilla entry on kernel.org, CCing Alan Stern, and attach the usbmon output when plugging the device (and let it go through a couple of reset cycles). Alan's pretty responsive and found the problem with the E169 pretty quickly from my logs.

Feel free to also CC me.

Ok, so here's what I observed with 2.6.31.5 and my E169:

 - After boot, the first time I insert the E169, the SCSI part shows up but not the modem
 - Unplug it and re-plug it, and it works

I'm not 100% sure yet what we have wrong that causes that behaviour, but it seems to
be quite consistent. However, it's a lot less bad than what we had before.

The E220 may well be having an altogether different problem.

Ben, I have the same experience with 2.6.31.5. But at least it works.

Thanks and regards

I'm in .ie, I'm an end user, and mobile broadband is ESSENTIAL for my work. My vodafone Huawei E620 dongle has the same problem listed above, but the SUSE install I put on my backup laptop works flawlessly. If Karmic doesn't work on Thursday, I will be moving to another Distro. Fix this!

I reported the bug with the E220 a while back: https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/449394

It may be a duplicate but I'm not sure.

There will be a certain irony if Karmic, hyped as For The Cloud, is released without a way for many users to get online while on the move. Fingers crossed this is resolved. That is all.

Iød like to add that the E620 doesn't work either... This needs to work for
Lucid!!

Nicholas Ipsen

On Mon, Oct 26, 2009 at 10:02 AM, Gerard Cunningham <email address hidden>wrote:

> There will be a certain irony if Karmic, hyped as For The Cloud, is
> released without a way for many users to get online while on the move.
> Fingers crossed this is resolved. That is all.
>
> --
> Huawei E169 USB dongle not working with kernel 2.6.31-12.40
> https://bugs.launchpad.net/bugs/446146
> You received this bug notification because you are a direct subscriber
> of the bug.
>
> Status in The Linux Kernel: Confirmed
> Status in “linux” package in Ubuntu: Fix Committed
>
> Bug description:
> I'm using a Huawei E169 USB dongle to connect to a GSM mobile broadband
> network. It used to work out-of-the-box with Jaunty and previous versions of
> Karmic (kernel 2.6.31-11 and previous). However the device doesn't work any
> more with the latest kernel.
>
> Dmesg gives me repatedly this message:
>
> dmesg:
> option: option_instat_callback: error -108
>
> lsusb:
> Bus 005 Device 005: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB
> Modem
>
>
> My system: Dell XPS M1530, see attached result of "lshw"
>
> This bug apparently affects other people as well:
> http://ubuntuforums.org/showthread.php?p=8070873#post8070873
>

On Mon, 2009-10-26 at 09:28 +0000, Seph_VII wrote:
> Iød like to add that the E620 doesn't work either... This needs to work for
> Lucid!!

As I said earlier, if your HW doesn't work with what's in 2.6.31.5
(minus the glitch of having to unplug & replug which I'll try to
investigate when I have some time), then you should file an upstream bug
on kernel.org bugzilla, and make sure you CC Alan Stern and provide
usbmon logs. That's the only way it's going to get fixed.

Cheers,
Ben.

> Nicholas Ipsen
>
>
> On Mon, Oct 26, 2009 at 10:02 AM, Gerard Cunningham <email address hidden>wrote:
>
> > There will be a certain irony if Karmic, hyped as For The Cloud, is
> > released without a way for many users to get online while on the move.
> > Fingers crossed this is resolved. That is all.
> >
> > --
> > Huawei E169 USB dongle not working with kernel 2.6.31-12.40
> > https://bugs.launchpad.net/bugs/446146
> > You received this bug notification because you are a direct subscriber
> > of the bug.
> >
> > Status in The Linux Kernel: Confirmed
> > Status in “linux” package in Ubuntu: Fix Committed
> >
> > Bug description:
> > I'm using a Huawei E169 USB dongle to connect to a GSM mobile broadband
> > network. It used to work out-of-the-box with Jaunty and previous versions of
> > Karmic (kernel 2.6.31-11 and previous). However the device doesn't work any
> > more with the latest kernel.
> >
> > Dmesg gives me repatedly this message:
> >
> > dmesg:
> > option: option_instat_callback: error -108
> >
> > lsusb:
> > Bus 005 Device 005: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB
> > Modem
> >
> >
> > My system: Dell XPS M1530, see attached result of "lshw"
> >
> > This bug apparently affects other people as well:
> > http://ubuntuforums.org/showthread.php?p=8070873#post8070873
> >
>

A couple of clarifications for E220:
2.6.31-11 Ubuntu kernel does work directly without a glitch.
Ubuntu kernels -12, -13, -14 and 2.6.31.5 kernel does not work directly. It does not work either after several plug in / out cycles.

As Patrick Cost wrote, I can also make 2.6.31.5 work by:
sudo rmmod usb-storage
sudo modprobe usbserial vendor=0x12d1 product=0x1003

After that the device is working properly.

Any word on a release date for this patch, or is it roll your own kernel time?

In the mean time, why not create a PPA with the patched kernels?

Hi Daniel, you changed the status from "Fix Committed" to "Fix Released"... Can you please tell me which package version was released with the fix?

Thanks!

Hi.

I tried with my wifes E270 right now, and this works like a charm. Can anybody try this and give feedback?

E270 and E220 are normally known as the same modem/hardware, they look different, but seem to be the same - so it's a bit strange, that they behave different in this case.

After switching to my E220 without doing anything more I'll post again.

nope, it doesn't work then, had to rmmod usb-serial about 20 times and plug/unplug the device too :(

added a separate bug report for E220 in http://bugzilla.kernel.org/show_bug.cgi?id=14499

ah, perfect, we should talk about this one somewhere else, for example in
https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/449394 this report:

All these Huawei bugs are clearly not going to be patched before the final release.

Who is responsible for adding a 'Known Issues' comment on the Ubuntu homepage for the release?

I suspect there will be plenty of peeved users otherwise. I know of a number of user who would lose their only connectivity due to these bugs.

Great....
This is the worst one so far - in addition to suspend at times hanging my machine and wlan toggle not working properly on my Eee pcs. Guess what I use for internet connectivity - yup, a Huawei modem.

Regressions, regressions....It worked from Hardy onwards (with wvdial but still).

Well... there would be a way to fix all those regressions for release until a proper fix is found
upstream .... which would be to revert the patch that went in 2.6.31.2 that caused them all.

The result would be to go back to 2.6.31.1 state with the modem working but the storage
side of the device not working, which isn't a big deal, since it's bugs in the USB storage
implementation of the devices that seem to be causing this.

This is commit

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=d0defb855c8504c49b92bdc0203689ce9b4cf7ba

Cheers,
Ben.

Has anyone, apart from Ben H, liro R-J (who has a e220) and me, tried the kernel at http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/? Apart from the first time you insert the e169, both modem and usb storage seem to work flawlessly. (I'm using it now.)

I also have the current ubuntu kernel -- 2.6.31-14.48 -- installed and the meta packages so that I know when there is a new kernel in the repos,

I installed that kernel to fix my wlan problems, but didnt yet try my Huawei with it. I will report back when I have the time to test it - should be today.

On the Eee pc now - I have the Huawei E220 dongle, I installed the 2.6.31-5 kernel.
The modem worked on first connection, both the USB memory part and the modem showed up - the modem had been configured already previously on Jaunty so no configuration was needed now. It works normally as far as I can tell.

Cant say about E169, I dont have one of those.

I should add that before, on Intrepid and Jaunty, the modem worked perfectly, but the USB memory part never showed up, which it does now.

With kernel http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/, I can see CD and memory part (VFAT) and modem as well, but modem seems to be kind of blocked. I can't do any connection via e169. Anyone knows what I missed?

....
Oct 29 17:23:28 mini0001 modem-manager: (Huawei): GSM modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1 claimed port ttyUSB0
Oct 29 17:23:28 mini0001 modem-manager: Added modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1
Oct 29 17:23:28 mini0001 modem-manager: Exported modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1 as /org/freedesktop/ModemManager/Modems/2
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): new GSM device (driver: 'option1')
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): exported as /org/freedesktop/NetworkManager/Devices/6
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): now managed
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): device state change: 1 -> 2 (reason 2)
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): deactivating device (reason: 2).
Oct 29 17:23:28 mini0001 NetworkManager: flush_routes: assertion `iface_idx >= 0' failed
Oct 29 17:23:28 mini0001 NetworkManager: flush_addresses: assertion `iface_idx >= 0' failed
Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): device state change: 2 -> 3 (reason 0)
Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB1): re-checking support...
Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB1) opening serial device...
Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB2): re-checking support...
Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB2) opening serial device...
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1) closing serial device...
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1) opening serial device...
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1): probe requested by plugin 'Generic'
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2) closing serial device...
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2) opening serial device...
Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2): probe requested by plugin 'Generic'
Oct 29 17:23:35 mini0001 kernel: [ 1017.806404] sd 19:0:0:1: [sdf] 3842048 512-byte logical blocks: (1.96 GB/1.83 GiB)
Oct 29 17:23:35 mini0001 kernel: [ 1017.889379] sd 19:0:0:1: [sdf] Assuming drive cache: write through
Oct 29 17:23:35 mini0001 kernel: [ 1017.973318] sd 19:0:0:1: [sdf] Assuming drive cache: write through
Oct 29 17:23:35 mini0001 kernel: [ 1017.973330] sdf: sdf1
Oct 29 17:23:45 mini0001 modem-manager: (ttyUSB1) closing serial device...
Oct 29 17:23:46 mini0001 NetworkManager: <WARN> wait_for_connection_expired(): Connection (2) /org/freedesktop/NetworkManagerSettings/0 failed to activate (timeout): (0) Connection was not provided by any settings service
Oct 29 17:23:46 mini0001 modem-manager: (ttyUSB2) closing serial device...
Oct 29 17:23:46 mini0001 modem-manager: mm_modem_base_add_port: assertion `port == NULL' failed
Oct 29 17:23:46 mini0001 modem-manager: do_grab_port: plugin 'Generic' claimed to support tty/ttyUSB2 but couldn't: (-1) (unknown)
Oct 29 17:23:53 mini0001 NetworkManager: <WARN> wait_for_connection_expired(): Connectio...

On Thu, 2009-10-29 at 16:26 +0000, Marcin Giedz wrote:
> With kernel http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/, I
> can see CD and memory part (VFAT) and modem as well, but modem seems to
> be kind of blocked. I can't do any connection via e169. Anyone knows
> what I missed?

Weird. It looks like its working properly and something inside network
manager goes wrong. Tried popping it out and back in ? Trying multiple
times maybe ? Checking your various NM settings ?

Doesn't look related to the bug at hand.

Cheers,
Ben.

> ....
> Oct 29 17:23:28 mini0001 modem-manager: (Huawei): GSM modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1 claimed port ttyUSB0
> Oct 29 17:23:28 mini0001 modem-manager: Added modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1
> Oct 29 17:23:28 mini0001 modem-manager: Exported modem /sys/devices/pci0000:00/0000:00:1d.2/usb4/4-1 as /org/freedesktop/ModemManager/Modems/2
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): new GSM device (driver: 'option1')
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): exported as /org/freedesktop/NetworkManager/Devices/6
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): now managed
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): device state change: 1 -> 2 (reason 2)
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): deactivating device (reason: 2).
> Oct 29 17:23:28 mini0001 NetworkManager: flush_routes: assertion `iface_idx >= 0' failed
> Oct 29 17:23:28 mini0001 NetworkManager: flush_addresses: assertion `iface_idx >= 0' failed
> Oct 29 17:23:28 mini0001 NetworkManager: <info> (ttyUSB0): device state change: 2 -> 3 (reason 0)
> Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB1): re-checking support...
> Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB1) opening serial device...
> Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB2): re-checking support...
> Oct 29 17:23:29 mini0001 modem-manager: (ttyUSB2) opening serial device...
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1) closing serial device...
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1) opening serial device...
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB1): probe requested by plugin 'Generic'
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2) closing serial device...
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2) opening serial device...
> Oct 29 17:23:34 mini0001 modem-manager: (ttyUSB2): probe requested by plugin 'Generic'
> Oct 29 17:23:35 mini0001 kernel: [ 1017.806404] sd 19:0:0:1: [sdf] 3842048 512-byte logical blocks: (1.96 GB/1.83 GiB)
> Oct 29 17:23:35 mini0001 kernel: [ 1017.889379] sd 19:0:0:1: [sdf] Assuming drive cache: write through
> Oct 29 17:23:35 mini0001 kernel: [ 1017.973318] sd 19:0:0:1: [sdf] Assuming drive cache: write through
> Oct 29 17:23:35 mini0001 kernel: [ 1017.973330] sdf: sdf1
> Oct 29 17:23:45 mini0001 modem-manager: (ttyUSB1) closing serial device...
> Oct 29 17:23:46 mini0001 NetworkManager: <WARN> wait_for_connection_expired(): Connection (2) /org/freedesktop/NetworkManagerSettings/0 failed to activate (timeout): (0) Connection was not provided by any settings service
> Oct 29 17:23:46 mini0...

As expected, the final release kernel for Karmic is 2.6.31-14.48, so no Huawei modem connections for new or existing users (unless they find this bug report and can figure out what to do). A pretty poor outcome, given Ben's patch has been 'committed' for two weeks.

The E220 problem seems to be persisting with me. If I connect my E220 via my USB hub then I can eventually get online (I have set Network Manager to connect automatically). Usually the connection with switch on/off several times before finally settling. But it is far from stable, it will drop the connection without warning.

When I connect the E220 directly to my laptop I get the following message when I attempt to connect:

"Unable to mount Mobile_Connect

Error mounting: mount exited with exit code 32: mount: /dev/sr1 is not a valid block device"

So this is obviously a regression from 9.04, I will not be updating my other two machines until this is fixed as I rely completely on the E220 for Internet access.

..to Ben,

Yes tried many times, and I don;'t thing a have any strange settings in NM as wired/wifi connection works great. Should I file a new bug for this?

If you don't have access to other kernel version, is there a workaround? Saw something with rmmod and modprobe in the forums, but it did not work for me.

im using Huawei E220 modem to get get connected to internet. I installed Karmic yesterday(30 Oct 2009) the modem is not working i dunno wats wrong i think its something to do with the drivers cos in my laptop when i boot if i got o windows first and then restart immediately after the desktop appears in windows and then i go to ubuntu it works i dunno whats wrong with this. It was the same with my bluetooth connection too but luckily i was able to fix it by this http://rusiruboteju.wordpress.com/2009/08/08/phoenix-bios-and-bluetooth/
I want to fix this too can anyone help me cos this is the only way that im getting connected to internet and plus i hate to go to windows just to make my modem work the kernel version im using is 2.6.31-14-generic all the previous ubuntu versions like 9.04 and 8.10 worked well with this modem why is this not working?
and can anyone please tell me a workaround till this bug is fixed?
and also in the notification area i cant see the network connection notification if i remove it and add it again to the panel it works!
can anyone HELP ME!!

I've the same problem with my Huawei Technologies Co., Ltd. E620 USB Modem dongle. With my Nokia N93i I can go online (it use the module cdc_acm). But it's funny. Yesterday I'll get online with the Huawei dongle. Is it clear when a new kernel will come with the updates? My kernel 2.6.31-14-generic #48-Ubuntu SMP Fri Oct 16 14:05:01 UTC 2009 x86_64 GNU/Linux

"Fix Released" means that a package that fixes the bug has been released in karmic-updates, and your Update Manager will be able to download and install the update: the problem is solved and we can forget about the bug.
"Fix Committed" means that a patch that fixes the problem has been created and possibly some packages have been built, but these packages have not been approved and have not made their way to the official karmic-update repository yet.
The second case seems to me to be our case.
If we mark this bug as "Fix Released", distribution maintainers will think it's a problem completely solved, forget about it, and not take any further action: an updated package will not be released.
Please don't change the status to "Fix Released" unless you really know what you are doing, especially if you joined launchpad just a few days ago =).
You can keep a bookmark to https://edge.launchpad.net/ubuntu/karmic/+source/linux/ to check when a new kernel package has been released.

For those interested and with alternate access to the web, kernel in *.32 series at http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.32-rc5/ works with my e169 (at first glance, everything else seems ok too).

just one stupid question that is not directly related to the discussion: I had the same problem with the e620 modem and the alternative kernel I installed from this bugreport got it working. But can this be the cause of another usb device (namely a flash drive) not working anymore?

@Alexander Skiba - That's a different issue. Please open a new bug and attach the output of lsusb with and without the flash drive plugged in.

Same problem with the E620:

Details on dmesg-Output - Kernel tries permanently activating the hardware.

Nov 1 10:17:17 franz-laptop kernel: [ 7540.444181] usb 5-1: reset full speed USB device using uhci_hcd and address 10
Nov 1 10:17:17 franz-laptop kernel: [ 7540.589741] option 5-1:1.2: GSM modem (1-port) converter detected
Nov 1 10:17:17 franz-laptop kernel: [ 7540.589892] usb 5-1: GSM modem (1-port) converter now attached to ttyUSB0
Nov 1 10:17:17 franz-laptop kernel: [ 7540.590202] option 5-1:1.1: GSM modem (1-port) converter detected
Nov 1 10:17:17 franz-laptop kernel: [ 7540.590324] usb 5-1: GSM modem (1-port) converter now attached to ttyUSB1
Nov 1 10:17:17 franz-laptop kernel: [ 7540.590425] option 5-1:1.0: GSM modem (1-port) converter detected
Nov 1 10:17:17 franz-laptop kernel: [ 7540.590546] usb 5-1: GSM modem (1-port) converter now attached to ttyUSB2
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609588] option1 ttyUSB2: GSM modem (1-port) converter now disconnected from ttyUSB2
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609610] option 5-1:1.0: device disconnected
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609686] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609705] option 5-1:1.1: device disconnected
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609778] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
Nov 1 10:17:17 franz-laptop kernel: [ 7540.609795] option 5-1:1.2: device disconnected

Output of lsusb:
franz@franz-laptop:~$ lsusb
Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 006 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 001 Device 002: ID 064e:a101 Suyin Corp. Acer CrystalEye Webcam
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 005 Device 012: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

Error message:
Error mounting: mount exited with exit code 32: mount: block device /dev/sr1 is write-protected, mounting read-only
mount: /dev/sr1 already mounted or /media/Telekom Austria busy

Same issue with E220:

[ 4060.068051] usb 3-1: new full speed USB device using uhci_hcd and address 22
[ 4060.230258] usb 3-1: configuration #1 chosen from 1 choice
[ 4060.237777] scsi134 : SCSI emulation for USB Mass Storage devices
[ 4060.238017] usb-storage: device found at 22
[ 4060.238022] usb-storage: waiting for device to settle before scanning
[ 4060.424127] usb 3-1: USB disconnect, address 22
[ 4061.160061] usb 3-1: new full speed USB device using uhci_hcd and address 23
[ 4061.323258] usb 3-1: configuration #1 chosen from 1 choice
[ 4061.330765] option 3-1:1.0: GSM modem (1-port) converter detected
[ 4061.330850] usb 3-1: GSM modem (1-port) converter now attached to ttyUSB0
[ 4061.333206] option 3-1:1.1: GSM modem (1-port) converter detected
[ 4061.333274] usb 3-1: GSM modem (1-port) converter now attached to ttyUSB1
[ 4061.339421] scsi137 : SCSI emulation for USB Mass Storage devices
[ 4061.351221] usb-storage: device found at 23
[ 4061.351225] usb-storage: waiting for device to settle before scanning
[ 4066.350149] usb-storage: device scan complete
[ 4066.354135] scsi 137:0:0:0: CD-ROM HUAWEI Mass Storage 2.31 PQ: 0 ANSI: 2
[ 4066.383094] sr1: scsi-1 drive
[ 4066.383346] sr 137:0:0:0: Attached scsi CD-ROM sr1
[ 4066.383494] sr 137:0:0:0: Attached scsi generic sg2 type 5
[ 4077.753095] sr1: CDROM (ioctl) error, command: Get configuration 46 00 00 00 00 00 00 00 20 00
[ 4077.753122] sr: Sense Key : No Sense [current]
[ 4077.753131] sr: Add. Sense: No additional sense information
[ 4078.016109] sr1: CDROM (ioctl) error, command: Get configuration 46 00 00 00 00 00 00 00 20 00
[ 4078.016137] sr: Sense Key : No Sense [current]
[ 4078.016146] sr: Add. Sense: No additional sense information
[ 4078.155063] option: option_instat_callback: error -108
[ 4078.156301] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
[ 4078.156324] option 3-1:1.0: device disconnected
[ 4078.156388] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
[ 4078.156405] option 3-1:1.1: device disconnected
[ 4078.324045] usb 3-1: reset full speed USB device using uhci_hcd and address 23
[ 4078.471254] option 3-1:1.1: GSM modem (1-port) converter detected
[ 4078.471417] usb 3-1: GSM modem (1-port) converter now attached to ttyUSB0
[ 4078.479115] option 3-1:1.0: GSM modem (1-port) converter detected
[ 4078.479276] usb 3-1: GSM modem (1-port) converter now attached to ttyUSB1
[ 4078.876079] option: option_instat_callback: error -108
[ 4078.877639] option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
[ 4078.877683] option 3-1:1.0: device disconnected
[ 4078.877796] option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
[ 4078.877827] option 3-1:1.1: device disconnected
[ 4079.000069] usb 3-1: reset full speed USB device using uhci_hcd and address 23
[ 4079.124035] usb 3-1: device descriptor read/64, error -71
[ 4079.296744] usb 3-1: USB disconnect, address 23

cirne@cirne-laptop:/var/log$ lsusb
Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 026: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem
...

As I said multiple times already, the E220 bug (and possibly the E620 bug too) are a different problem (well, they seem to be) since you claim they aren't fixed by the existing patch in 2.6.31.5. Hence, you guys should file a bug report on bugzilla.kernel.org and attach usbmon logs (or attach such logs here and I can forward to BZ.k.o). Make sure to CC me and Alan Stern if you open the bug yourself.

This is happening to me too, I have the:
Bus 004 Device 005: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem

For now the workaround I using is:
http://ubuntuforums.org/showpost.php?p=8195214&postcount=2

My modem: ONDA MSA501HS
Kernel: 2.6.31-14-generic
Ubuntu 9.10 (my desktop OS version, installed as released in 29th Oct)

Same problem: after pluged in, shows it as a pendrive. Don't matter if you eject it, it's not reconized as a modem anymore. In Ubuntu 9.04 (my netbook OS version) still works.

Hi Benjamin, have a look at posts #86 and #87, are these what you're looking for ? I see http://bugzilla.kernel.org/show_bug.cgi?id=14499 claims to have fixed the issue in 2.6.31.6.

On Mon, 2009-11-02 at 09:14 +0000, Patrick Cost wrote:
> Hi Benjamin, have a look at posts #86 and #87, are these what you're
> looking for ? I see http://bugzilla.kernel.org/show_bug.cgi?id=14499
> claims to have fixed the issue in 2.6.31.6.

Ah, there may have been another fix I haven't seen. I'll have a look.
thanks.

Cheers,
Ben.

Hi, I had the same problem. My modem HUAWEI E169 is detected as a pendrive and not at all as a modem (sometimes it is not even recognised).
While waiting for the official patch, I installed kernel 2.6.32-rc5 and everything works well.

Andy, when will you release this update?

Not sure if that is related or somehow helps (as it seems that bug is already fixed): I have a Huawei E160G HSDPA Modem and I have to boot my laptop having it already attached (WLAN disabled) to get it to work correctly. Otherwise plugging it in when already logged in it mounts the drive on it and connecting does not really work (getting https://bugs.launchpad.net/bugs/452427 and https://bugs.launchpad.net/bugs/404351).

Hi!

I have huawei 169E and I have this problem that the machine goes coo-coo on that one message that it displays incessantly:

[23.318813] option: option_instat_callback: error -108

it also reported at one point:

no PSB or ACPI _PSS objects

I see your comments and thank yous but can't find the info to fix things.

So please help

Mikko

Hi All,

I saw some comments saying that the E220 problem also occurs with E620. Well, it seems that there are two different E620.

I have a E620 (a PCMCIA card) that it is seen by the kernel as a USB Device:
Bus 006 Device 002: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem

It worked out-of-the-box with no problem, also in live CD. ;-)

However, today I was showing Ubuntu 9.10 to a friend that has a E220 (vendor=0x12d1, product=0x1003) and I wanted to show that Ubuntu works with her USB Internet stick, even before installing it.

Well, bad luck and -1 eventual future ubuntu fan girl.

In fact, with this E220 device, the kernel see a memory device and not a modem.

I didn't try hard to circumvent the problem because Ubuntu was running in live CD and she didn't want to install it .... for now.

As Haldo comments above, using kernel 2.6.32-999-generic #200911011000 (2.6.32-rc5) my E169 can connect fine. Some other problems appear with this kernel so using it means: 'experimental'!

Bug affects me too.
My Device:
Bus 002 Device 008: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem

Can confirm that the workaround from post 110 is working on my system.

The workarounds with the usage of an Ubuntu Kernel from the 9.04 Version posted earlier
is no solution for me as with this the Internet connection is working but heavily graphics error and a not working sound card are the result.

Using Vodafone Ireland Huawei E620 dongle, Karmic Koala on a Thinkpad T60, the fix at #110 works for me. However, the same fix did not work pre-Karmic when I was on Ubuntu 9.04

I have the kernel 2.6.31-14-generic (which should be an up-to-date karmic) and my Huawei E160G needs to be plugged in already before startup of the system to make it work - at least for my last 3 attempts to use that mobile internet modem that was the only way to get it to work (did not try the manual loading of modules).

Please stop changing the status to "Fix Released" unless you really know what you are doing.

As Andrea already wrote on comment 103:
"Fix Released" means that a package that fixes the bug has been released in karmic-updates, and your Update Manager will be able to download and install the update: the problem is solved and we can forget about the bug.
"Fix Committed" means that a patch that fixes the problem has been created and possibly some packages have been built, but these packages have not been approved and have not made their way to the official karmic-update repository yet.
If we mark this bug as "Fix Released", distribution maintainers will think it's a problem completely solved, forget about it, and not take any further action: an updated package will not be released.

hey guys this might be a bit stupid but it worked for me
when i log in to ubuntu the modem doesn't work but it says or the notification area says that the network connection is established and available but cant browse or chat or do an installation using apt-get then wat i did was i logged out and logged in again vola it works fine :D i think its the easiest way to overcome this till this bug is fixed.

i still faces this problem on my dell XPS , can you help me ?

I also face the same issue. I am running 2.6.31-14-generic #48 in a Toshiba R600.
lsusb shows: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem
i also get "error mounting: mount exited with exit code 32: mount: block device /dev/sr1 is write-protected, mount read-only....blah"

dmesg:
option 3-1:1.2: GSM modem (1-port) converter detected
usb 3-1: GSM modem (1-port) converter now attached to ttyUSB0
option 3-1:1.1: GSM modem (1-port) converter detected
usb 3-1: GSM modem (1-port) converter now attached to ttyUSB1
option 3-1:1.0: GSM modem (1-port) converter detected
usb 3-1: GSM modem (1-port) converter now attached to ttyUSB2
option: option_instat_callback: error -108
option1 ttyUSB2: GSM modem (1-port) converter now disconnected from ttyUSB2
option 3-1:1.0: device disconnected
option1 ttyUSB1: GSM modem (1-port) converter now disconnected from ttyUSB1
option 3-1:1.1: device disconnected
option1 ttyUSB0: GSM modem (1-port) converter now disconnected from ttyUSB0
option 3-1:1.2: device disconnected
usb 3-1: reset full speed USB device using uhci_hcd and address 3

This repeats. I also get a "notification" that GSM is disconnected...never connected.

9.04 worked well,9.10 not so well.

usb_modeswitch is installed and working as best as I can tell....

I don't have the modeswitch installed - should that be installed in general for the mobile internet sticks to make them work correctly?

I think this bug should at least have a High importance and not medium. This is basically a show stopper for most people.

And, am I right in assuming that when this fix gets approved it will be released to us ASAP?

I'm basically holding back from installing karmic because of this issue.

Hi People!
I have Huawei UMTS e800 Express Card and it doesn't work since 2.6.31-12 Kernel.
Neither with the Fix from Andy Whitcroft.
I tried last kernels from kernel.org - 2.6.31.5, also 2.6.32-rc6.
I see Andy's fix in Source Files in that Kernels, but it doesn't help me.

I tried also
sudo rmmod usb-storage
sudo modprobe usbserial vendor=0x12d1 product=0x1001

but :
Error mounting: mount exited with exit code 32: mount: block device /dev/sr1 is write-protected, mounting read-only
mount: /dev/sr1: can't read superblock
Error mounting: mount exitedError mounting: mount exited with exit code 32: mount: block device /dev/sr1 is write-protected, mounting read-only
mount: /dev/sr1: can't read superblock with exit code 32: mount: block device /dev/sr1 is write-protected, mounting read-only
mount: /dev/sr1: can't read superblock
Anyone any Idea?
Thanks!

I found old kernel 2.6.31-11 amd64 to stay "on-line". But this is not pretty..
Thanks, Marcel

As craighewetson said, the importance should be bumped to High, there are gazillions of people who depend on a Huawei device to get connectivity. I've got a E220, E620, E870 and a K3520 around and none of them are working. Amazed that this hasn't been solved yet :S

It doesn't matter because it's going to be fixed, but It's true that "importance" must be set to "High".

Amazed that this hasn't been solved yet too. This bug is VERY SERIOUS!
These USB 3G dongles are used massively in Spain (and other European countries).
Many people like me are using 3G Internet at home because the ADSL coverage in our country is ridiculous.
Please stop changing to Fix Released, and give it High importance level!

Thank you very much!

is a fix is commited, where is it? so i can try it and give feedback.

i second the high importance of resolve this bug. If some help is needed, say it :)

Is there a plan on when this fix will be released in Karmic, are we talking
weeks, months?
I think if everyone has an idea of how long it will take it will give us a
warm feel :)

I have a contract E620 from vodafone, which is now effectively deadweight, and a pay-as-you-go E1550 from Three, which fortunately still works. Applying some of the fixes suggested in these threads fixes my E620 problem, but its an imperfect fix, I can't connect every time, and the fix breaks my 1550. I've had some issues with sound and boot freezes since applying the fix too, but i'm not sure if they're connected. Karmic Koala clean install on an IBM T60 Thinkpad.

After several attempts at workaround, I've decided to wait for Ubuntu to fix this problem, and use my pay-as-you-go back for on-road connectivity. This is a disaster. This year to date, I've helped three people migrate to Ubuntu. I won't be recommending it again for quite some time, and I'm looking at other distros myself.

The Huawei dongles are very common here in Sweden! Even bundled with minipc's and laptops.
This bug is fatal for all those upgrading to 9.10 and looses all internet conectivity.

This bug is also very serious when it comes to people finally giving Ubuntu their first try (lots of media
attention in Sweden regarding Win7 vs Ubuntu 9.10 releases). I imagine they won't be impressed when
their laptops are turned into typewriters.

Also, two of my recently Ubuntu'fied friends has just reinstalled XP because of this! Not good...
I don't even dare to think about how many outthere doing the same as we speak.

High importance is an understatement.

These usb dongles are very very common also here in Italy.
The importance of this bug should not be only "High" but maybe "Extreme"!

I hope the fix will be released ASAP.
Why not to insert the patched kernel in the -proposed repos, for the moment?

Previous comments told about other problems with the current kernel and/or the patch fixes E620 but breaks previously working E1550 - if I got the comment or Gerard Cunningham in #135 right.

@Martin Wildam, to clarify, the fix mentioned in Comment #110 fixed the E620 connection, but as a result my E1550 stopped working, I experienced other problems around the same time, whih may or may not have been related.

I'd rather not go futzing around with temporary fixes, and will now wait for a solution from Ubuntu. I just hope I don't have to wait too long.

I am an IT professional. I converted to Ubuntu 2 years ago because of the appalling state of Windows software. I persevered with Ubuntu 6 and 7 and was relieved with 9.04 when it just worked without any intervention. I now rely totally on my E220 dongle and upgraded to 9.10 a few days ago with only minor reservations as my faith in Ubuntu had been strengthened by recent releases.
Now I have no internet connection and have no way to deal with this.
GET A GRIP.

I think that, for now, we have to downgrade.

But that is almost unacceptable!
Jaunty was, by far, the best release, yet.
Shame on Canonical

On Thu, Nov 5, 2009 at 9:20 PM, Steve Lilleyman <
<email address hidden>> wrote:

> I am an IT professional. I converted to Ubuntu 2 years ago because of the
> appalling state of Windows software. I persevered with Ubuntu 6 and 7 and
> was relieved with 9.04 when it just worked without any intervention. I now
> rely totally on my E220 dongle and upgraded to 9.10 a few days ago with only
> minor reservations as my faith in Ubuntu had been strengthened by recent
> releases.
> Now I have no internet connection and have no way to deal with this.
> GET A GRIP.
>
> --
> Several Huawei USB dongle don't work with kernel 2.6.31-12.40 (2.6.31.1
> update related)
> https://bugs.launchpad.net/bugs/446146
> You received this bug notification because you are a direct subscriber
> of the bug.
>

--
João Marcelino
Investigador Principal com Habilitação
Vice-Presidente do Conselho Científico
Chefe do Núcleo de Barragens e Obras de Aterro - Departamento de Geotecnia
LNEC
Av. Brasil 101 - 1700-066- LISBOA

This problems was entirely preventable because the fix was known two weeks before karmic's release.

However, I'm wondering from comments above (post #129) that the kernel.org version 2.6.31.5 does not work with some modems, if the problem is wider than Ubuntu and affects other distros as well.

I also wonder if, as Ben Herrenschmidt commented earlier, we have multiple problems. My e169 works with Andy's patched kernel and kernel.org kernels after and including 2.6.31.5.

this is really an embarrassing bug, at least because working usb 3g -data modems are a basic requirement for the new distro for netbooks (ubuntu nbr) -- i switched to ubuntu about three years ago and keep campaigning for it, but with this issue it's really hard to argue.

...on the other hand by all critisicm: we should not forget that we get all this great ubuntu distro for free. for me, I'll turn a blind eye on that on hope for a really quick fix.

OK it is free, but one does not expect that a new release is worse then a
older one. I also try very hard to get my co-workers to work in Ubuntu (or
something like it) but this problem is very embarrassing.

On Thu, Nov 5, 2009 at 10:04 PM, hanzz <email address hidden> wrote:

> this is really an embarrassing bug, at least because working usb 3g
> -data modems are a basic requirement for the new distro for netbooks
> (ubuntu nbr) -- i switched to ubuntu about three years ago and keep
> campaigning for it, but with this issue it's really hard to argue.
>
> ...on the other hand by all critisicm: we should not forget that we get
> all this great ubuntu distro for free. for me, I'll turn a blind eye on
> that on hope for a really quick fix.
>
> --
> Several Huawei USB dongle don't work with kernel 2.6.31-12.40 (2.6.31.1
> update related)
> https://bugs.launchpad.net/bugs/446146
> You received this bug notification because you are a direct subscriber
> of the bug.
>

OK, I think they know now how important this issue is.

To get constructive again: What can we do to help clearing up the situation and to finally get a complete and comprehensive fix?

Whoever is working on this - what can we do for you to help - as it seems that the proposed fix is not fully working for everybody?

On Thu, 2009-11-05 at 22:46 +0000, Martin Wildam wrote:
> OK, I think they know now how important this issue is.
>
> To get constructive again: What can we do to help clearing up the
> situation and to finally get a complete and comprehensive fix?
>
> Whoever is working on this - what can we do for you to help - as it
> seems that the proposed fix is not fully working for everybody?

Well, what you can do is

 - Test if the proposed fix (or 2.6.31.5) works for you, if yes, then
just wait since I expect Canonical to release that update soon

 - If not, please send the usbmon logs I've requested about 3 times so
we can actually track down what the "other" problem might be :-)

As to how to do that usbmon logging, it's very easy:

 cat /sys/kernel/debug/usb/usbmon/0u >log_file.txt

And then plug the modem, wait about a minute for "things" to happen, and
ctrl-C the above command & send us the log file (along with the dmesg
output snapshot -after- the log has been taken).

Cheers,
Ben.

Ben,

As requested in #146, I have attached the files you requested. Please let me know if there is more I can do to help.

On Fri, 2009-11-06 at 00:08 +0000, Joel wrote:
> Ben,
>
> As requested in #146, I have attached the files you requested. Please
> let me know if there is more I can do to help.
>
> ** Attachment added: "usbmon & dmesg output"
> http://launchpadlibrarian.net/35202110/usb3gdongle.tar.gz
>

Yeah, please tell me precisely what modem model it is (and how it shows
up in lsusb). Thanks.

Also what kernel are you running (uname -a will do)

Cheers,
Ben.

@Ben,

Sorry, should have included all info in one post.....

The modem is a 3G vodafone jobby from Australia...
lsusb shows: Bus 003 Device 003: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem
uname -a shows: 2.6.31-14-generic #48-Ubuntu SMP Fri Oct 16 14:04:26 UTC 2009 i686 GNU/Linux

some more info is in post #126

Let me know if there is more....

@Ben,

Back of modem shows Model K3520

But do that kernel contain the proposed fix ? Or is this the current stock Karmic kernel ? IE. AFAIK, Ubuntu hasn't released a kernel with the patch so unless you are running Andy's build or 2.6.31.5 -or- manually applied my patch and rebuilt, you are not running with the proposed fix applied.

Have you tried with the 2.6.31.5 from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/ as well ?

We basically need to figure out whether a -different- problem is affecting E620 users or simply that people on this thread have been confused as to what kernel contains the fix and what kernels don't

Ben,

Thank you for your patience on this, it appears I am one of the confused :-( sorry. I am not running Andy's build nor have I applied your patch :-(

I am running the current stock kernel...
If you can point me in the direction of your patch (and how to apply) I will take this on.
I will then post back and let everyone know :-)

cheers again for your patience.....

Just try this one:

http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/

Thanks !

Ben.

Just for the record:

The Ubuntu kernel 2.6.31-14.48 is the Karmic Koala final iso version and does not work with E169.

The kernel complied by Andy is 2.6.31-13.44 (post #15) and it has Ben's patch applied -- it works for me.

Kernels 2.6.31-5 and later (2.6.32 from kernel.org, for example) have Ben's patch applied but have not been released by Ubuntu as yet. Again, they work for my E169 and Ben's modem, see earlier posts.

If in doubt about the patch, check for Ben's name in the release note with each kernel.

PS. Confusing, too, because E169 (as labelled) is identified as E620 in usb logs and using lsusb.

Hi.
I do this Outputs as a Help for fixing the Bug. (cat /sys/kernel/debug/usb/usbmon/0u and also output from /var/log/messages - i did following: plug in modem, wait, unplug, wait, plug in, wait unplug)

I have Huawei e800 Express Card UMTS Modem.
The outputs are from the newest kernel : 2.6.32-rc6 - i did build it for me from kernel.org

For other confused : "fix" it for the time of fixing the bug for you, as i did:
Search in goggle (in other PC, that is on line ) for some precompiled deb Package with Kernel for Example: "2.6.31-11 amd64 deb", transfer it to karmic system, install it with sudo dpkg -i <filename.deb> and after reboot use that downgraded kernel.
If you have not 64bit system then repalce in search the amd64 for i386

I hope it will help someone..
M.

I would also like to try the fix as it seems I am the only one here with a 160G. But using an unreleased kernel is something new for me. Do I have to install both the linux-image and linux-headers. And how can I then switch back to the released kernel? Using Grub menu?

You need only linux-image. Yes, grub menu is the place, where you choose, what kernel you wish to boot.

You can anytime deinstall this kernel with sudo dpkg -P linux-image-2.6.31-11-generic or something like this. BUT You have to boot in that case to other kernel, if you want to remove that 11!

If you write that deinstall statement , hit <TAB> key after ..2.6.31 System will fill in the rest of the line, if this is unique Package in database.

I plan to wait the fix - so I will install new kernels and fixes as they come. If they not works, i will boot again from GRUB menu into 11, deinstall the newest kernel (because i won't use it and also to gain place in /boot dir)
If it will be fixed in newest kernel/Updates, I deinstall that 11 kernel away..
M.

It is very likely that we get an official kernel update anyway: http://www.h-online.com/open/news/item/Hole-in-the-Linux-kernel-allows-root-access-850016.html

Hi,

had same Problem with Huawai E660 PCMCIA Datacard which my Laptop shows as USB-Device on WinXP. VirtualBox grabs the whole USB Port and forwards it to Ubsudo untu where it works until 9.04.

lsusb:
Bus 002 Device 002: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem

When booting with attached USB-Modem, the usb-storage mounts a new drive but UMTS doesn`t work.

--> Attached Modem after reboot --> Everything works! THANKS!

Allright, let's try to see more clearly before we got dig further.

First of all, the only valid tests are with either Andy test kernel or 2.6.31.5 since both have my fix.

Then, there's an additional problem that I've seen as well and not debugged yet, which sometimes for me makes the modem not work the first time it's plugged (or if plugged at boot). Yanking it out and back in makes it work fine. This is a -different- issue that the one we've been tracked and it has a simple workaround obviously.

So please, from now, report if and only if you have tested the bug fix kernel -and- tested removing and re-inserting the modem and it still doesn't work. At which point, I need logs. To MarcelID: Have you tried the removal/insertion trick ? Does it help for you with a patched kernel ? If you have and it still doesn't work, then we have at least one set of valid logs to study further and i'll send them to the USB maintainer.

Thanks !

Looks like I lost my ability to write proper english tonight... sorry :-)

To reiterate, the initial fix for the E169 issue is committed (only). It is not Fix Released, when it _is_ released it will be switched to that state automatically by the launchpad janitor as the packages are put into the archive. Please do not confuse people further by moving it Fix Released. Although there is much talk of other modems on this bug report, the fix tied to this bug number is currently only known for certain to fix E169 modems. When this bug goes Fix Released that will only confirm that _that_ one fix has released.

We need therefore to confirm specifically whether the fixes we are proposing to release fix all of the modem or if more action is required. There are unconfirmed reports that the fix for this bug also fixes the E220 modems and E629 modems. There is also a second fix coming in the 2.6.31.5 update which is reported to help more of these devices. To help confirm these reports and ensure will have all of these covered I have made TWO test kernels:

1) 2.6.31-16.50lp446146apw5 -- this contains both of the fixes above and represents what we are currrently planning on releasing as an update,

2) 2.6.31-16.50lp446146apw10 -- this contains an additional patch revert as suggested by BenH, should the previous kernel not work for you.

This bug has become very confused so we need testing from those who can test, plus very clear testing and reporting on those tests to ensure we are releasing the right fixes. So could those who are able to test these kernels, first test the *apw5* kernel at the URL below, please report your testing on that kernel stating clearly the Huawei model number you are testing with, and attaching the dmesg output to your report:

    http://people.canonical.com/~apw/lp446146-karmic/test1/

If and ONLY if that kernel does not work for you could you then test the second *apw10* kernel and report on that one too, again reporting the Huawei model number you are testing and attaching the dmesg output to your report:

    http://people.canonical.com/~apw/lp446146-karmic/test2/

Now that Karmic has opened and Lucid has released we have _two_ tasks on this bug one for each release. The linux (Ubuntu) task represents the updated Lucid kernel, the Karmic sub-task tracks Karmic release status.

To Benjamin:
No, the "reinsertion" of Express Card e800 did not help..
Neither with kernel 2.6.31.5 from kernel.org

I'm downloading already yours pw5 kernel. So let's see..
M.

Accepted linux into karmic-proposed, the package will build now and be available in a few hours. Please test and give feedback here. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you in advance!

Andy, I have test *apw5* with E169 and it works without any need for additional plug / unplug as described earlier by Ben.

Output of dmesg is attached.

Regards,
Russell

So, my test was unsuccessful with apw5 and HUAWEI e800 Express card.
/sys/kernel/debug/usb/usbmon/0u and /var/log/messages (in 1 File appended) attached.
I tried reinsertion - unsuccessful.

Will try apw10..
M.

I tried apw10, but I cannot install it - there is Error somewhere in deb..

boss@ntb:/data/kernel$ sudo dpkg -i linux-image-2.6.31-16-generic_2.6.31-16.50lp446146apw10_amd64.deb
Selecting previously deselected package linux-image-2.6.31-16-generic.
(Reading database ... 125152 files and directories currently installed.)
Unpacking linux-image-2.6.31-16-generic (from linux-image-2.6.31-16-generic_2.6.31-16.50lp446146apw10_amd64.deb) ...
Done.
dpkg-deb (subprocess): short read in buffer_copy (failed to write to pipe in copy)
dpkg-deb: subprocess paste returned error exit status 2
dpkg: error processing linux-image-2.6.31-16-generic_2.6.31-16.50lp446146apw10_amd64.deb (--install):
 short read in buffer_copy (backend dpkg-deb during `./lib/modules/2.6.31-16-generic/kernel/fs/xfs/xfs.ko')
Running postrm hook script /usr/sbin/update-grub.
Generating grub.cfg ...
Found linux image: /boot/vmlinuz-2.6.31-11-generic
Found initrd image: /boot/initrd.img-2.6.31-11-generic
Found memtest86+ image: /memtest86+.bin
Found Microsoft Windows XP Professional on /dev/sda1
done
Errors were encountered while processing:
 linux-image-2.6.31-16-generic_2.6.31-16.50lp446146apw10_amd64.deb
boss@ntb:/data/kernel$

M.

Andy, I tried several boot / reboot, insertion cycles. Confirming all works as expected with apw5 and E169. Is this the same kernel as Martin says will be in 'proposed' within hours?

I post Logs with old kernel 2.6.31-11. It's the last kernel, where my E800 works. It may be good for debugging..
M.

Hi,

Still not working with apw5 and huawei E220 .

Logs are attached.

J

Couldn't find new proposed kernel, but I can confirm that test kernel 2.6.31-16.50lp446146apw5_i386 do work with E169. However, the test kernel mount the USB disk 'Mobile Partner' containing windows setup files. That now desirable and differs from behavior from jaunty.

It's also seems that my Fujitsu-Siemens Amilo M1425 fails to wake up from hibernate with test kernel.
Attached log and dmesg files.

I can also confirm that the Huawei E620 modem works with kernel 2.6.31-16-generic #50lp446146apw5. lsusb shows this:

Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 001 Device 008: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem
Bus 001 Device 006: ID 064e:f111 Suyin Corp.
Bus 001 Device 002: ID 152d:2329 JMicron Technology Corp. / JMicron USA Technology Corp.
Bus 001 Device 005: ID 2040:7070 Hauppauge Nova-T Stick 3
Bus 001 Device 003: ID 0409:0059 NEC Corp. HighSpeed Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 006 Device 003: ID 08ff:1600 AuthenTec, Inc. AES1600
Bus 006 Device 002: ID 0b05:1751 ASUSTek Computer, Inc.
Bus 006 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 002: ID 1241:1122 Belkin Typhoon Stream Optical Mouse USB+PS/2
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

dmesg output is attached. I also see the additional USB disk appearing as Janne Huovari describes. The modem devices ttyUSB0, ttyUSB1 and ttyUSB2 do not disappear as they did with the original Karmic kernel.

Patrick

P.S.: This is on AMD64

As soon as I'll find the new proposed kernel, I'll test it.

@Ben, post #160:
As I said in my previous post, I am running kernel 2.6.32-rc5.
The additional problem you noted is present even with this kernel; i.e. sometimes the first time the modem is plugged in, it doesn't work. This happens just sometimes; unplugging and plugging in the modem again solves it.

Andy, I have tried both apw5 and apw10 and they both work with my e169 -- BUT they both break suspend on my Acer Aspire One. I have not had time to try them on my Dell 6400.

The best working kernel I have is the one you posted at #15. Other later kernels (2.6.32 etc) do not AppAmor correctly.

Same problems with the internet key Huawey 1692 (linux compatible!)

Andy,
It seems that E220 is still broken. Unfortunately I can not currently give you logs about your new kernel version, but I posted three usbmon logs to the kernel bugzilla one week ago as Ben proposed: http://bugzilla.kernel.org/show_bug.cgi?id=14499

1. Ubuntu 2.6.31-11 kernel. E220 inserted and Network manager connection
selected. Fully working internet connection.

2. 2.6.31.5 from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/ USB
inserted and after Network manager I waited couple of cycles of automatic
removal of the driver.

3. 2.6.31.5 from http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/ USB
inserted and commands:
    sudo rmmod usb-storage
    sudo modprobe usbserial vendor=0x12d1 product=0x1003
after selecting the right connection from Network Manager it starts to work.
This method does not work always.

Í hope these helps.

Andy, I can confirm 2.6.31-16.50lp446146apw5_i386 works (scsi and modem) with e169 on my Dell6400 and suspend also wakes up.

In my earlier post (#176) my AA1 fails to wake from suspend on this kernel.

Regards

E220 does not work with apw5
but works fine with apw10

Thanks, now waiting for update which it seem to be 2.6.31-15.49

https://launchpad.net/ubuntu/karmic/+source/linux/2.6.31-15.49

2.6.31-5.49 from link above installed and works with e169 (as described in post #172).

Thanks and regards, Russell

2.6.31-5.49 breaks resume from suspend on my Acer Aspire One.

I have Huawei E150E
lsusb shows
Bus 002 Device 005: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem

I installed this
http://kernel.ubuntu.com/~kernel-ppa/mainline/v2.6.31.5/

But I am not sure what the rest meant by worked. For my case, with the above kernel, the stick consistently became a usb-storage (while previously its always a half baked stage).

I tried rmmod usb-storage and modprobe usbserial, but still NetworkManager couldn't see anything.

Attaching my usbmon and dmesg

Tryed to install the *apw5* kernel, but could not start ubuntu after that. System stayed in text modus, screen flashed and I could not log on (lenovo T61, Nvidia Quadro NVS 140M).
lusb: Bus 005 Device 002: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem

Yiepi! With the awp5 Kernel my E620 works! I didn't tested the awp10. Here are some logs while plug in the device and an overview of the hardware

I can confirm that E620 now works correctly, i.e I can go online and browse etc, using Vodafone Mobile Connect in Kubuntu Karmic on AMD64 with the 2.6.31-15-generic kernel.

Plugged in and out, and went on and off line several times, it worked at every attempt. And, it reverted correctly to ethernet or wifi afterwards.

But I am using wicd, not network-manager!

There may be a slight anomaly, as it seems to occupy ttyUSB0, 1 and 2. Should it do that?

Acer Aspire 7520 (AMD Athlon X2, Nvidia Gforce 7000M, with 2xSATA drives and non-standard Broadcom wireless card.

uname -a (machine ID redacted):
Linux ********** 2.6.31-15-generic #49-Ubuntu SMP Fri Nov 6 09:55:13 UTC 2009 x86_64 GNU/Linux

lspci:
00:00.0 RAM memory: nVidia Corporation MCP67 Memory Controller (rev a2)
00:01.0 ISA bridge: nVidia Corporation MCP67 ISA Bridge (rev a2)
00:01.1 SMBus: nVidia Corporation MCP67 SMBus (rev a2)
00:01.2 RAM memory: nVidia Corporation MCP67 Memory Controller (rev a2)
00:01.3 Co-processor: nVidia Corporation MCP67 Co-processor (rev a2)
00:02.0 USB Controller: nVidia Corporation MCP67 OHCI USB 1.1 Controller (rev a2)
00:02.1 USB Controller: nVidia Corporation MCP67 EHCI USB 2.0 Controller (rev a2)
00:04.0 USB Controller: nVidia Corporation MCP67 OHCI USB 1.1 Controller (rev a2)
00:04.1 USB Controller: nVidia Corporation MCP67 EHCI USB 2.0 Controller (rev a2)
00:06.0 IDE interface: nVidia Corporation MCP67 IDE Controller (rev a1)
00:07.0 Audio device: nVidia Corporation MCP67 High Definition Audio (rev a1)
00:08.0 PCI bridge: nVidia Corporation MCP67 PCI Bridge (rev a2)
00:09.0 IDE interface: nVidia Corporation MCP67 AHCI Controller (rev a2)
00:0a.0 Ethernet controller: nVidia Corporation MCP67 Ethernet (rev a2)
00:0c.0 PCI bridge: nVidia Corporation MCP67 PCI Express Bridge (rev a2)
00:0d.0 PCI bridge: nVidia Corporation MCP67 PCI Express Bridge (rev a2)
00:12.0 VGA compatible controller: nVidia Corporation C67 [GeForce 7000M / nForce 610M] (rev a2)
00:18.0 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] HyperTransport Technology Configuration
00:18.1 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Address Map
00:18.2 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] DRAM Controller
00:18.3 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Miscellaneous Control
01:04.0 FireWire (IEEE 1394): Ricoh Co Ltd R5C832 IEEE 1394 Controller (rev 05)
01:04.1 SD Host controller: Ricoh Co Ltd R5C822 SD/SDIO/MMC/MS/MSPro Host Adapter (rev 22)
01:04.2 System peripheral: Ricoh Co Ltd R5C592 Memory Stick Bus Host Adapter (rev 12)
01:04.3 System peripheral: Ricoh Co Ltd xD-Picture Card Controller (rev 12)
05:00.0 Network controller: Broadcom Corporation BCM4311 802.11b/g WLAN (rev 01)

lsusb:
Bus 004 Device 005: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem

dmesg (tail):
  653.410135] usb 4-1: new full speed USB device using ohci_hcd and address 6
[ 653.644632] usb 4-1: configuration #1 chosen from 1 choice
[ 653.649214] scsi10 : SCSI emulation for USB Mass Storage devices
[ 653.649858] usb-storage: device found at 6
[ 653.649864] usb-storage: waiting for device to settle before scanning
[ 653.71...

In view of the many complaints and much distress above from people who depend on these modems, much more so than myself, can we now please have the updated kernel in the repositories as quickly as possible? And, also in an updated Karmic, quickly?

Most people will not be able to mess about with kernel installation themselves.

Thanks.

Huawei E220 modem.

lsusb:
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 004: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem
Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 002: ID 046d:c517 Logitech, Inc. LX710 Cordless Desktop Laser
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

I have tested both 2.6.31-16.50lp446146apw5 and 2.6.31-16.50lp446146apw10, and attached the logs (usbmon and dmesg).

Results:
apw5 - exactly the same response as release kernel.
apw10 - works, BUT modem must be physically unplugged and reconnected. Everything seems to work, but it will not connect to the network if it was plugged in when the system booted. After unplugging, and reconnecting, it works fine (adding this post connected over the modem, using apw10).

Thanks!
Justin

Can only seem to add one attachment per comment?

Here is the next log...

Next log...

Doh. resent the apw5 log. Try again...

Last one.

I have a huawei e220 modem i would like to use, but i have no idea how to do the fixes that are proposed, like tiger99 said in comment #187.

Anyone knows a usefull/good link to somewhere where i can learn how to install a kernel diff?

If the update is already in the repositories, what package should i search for?

You can download the deb files and double click on them, the package
installer will install them.
Just make sure you install i386 one if you are running 32bit.

Remember to restart :)

Just reverted from apw10 to 2.6.31-11.

E220 works well in apw10, but the computer (Intel Littlefalls 2 integrated mb/cpu) dies after around 30 minutes. The screen blanks, and the computer stops responding to keyboard, mouse and ACPI events, BUT it is still running (networking up, and logging to /var/log/messages).

I get the following errors in /var/log/messages:
Nov 7 20:54:09 mediapc kernel: [ 2280.864065] i915/3 D c08165c0 0 295 2 0x00000000
Nov 7 20:54:09 mediapc kernel: [ 2280.864076] f6559f04 00000046 00000000 c08165c0 f69cc168 c08165c0 1574e75d 000001e8
Nov 7 20:54:09 mediapc kernel: [ 2280.864091] c08165c0 c08165c0 f69cc168 c08165c0 1574d741 000001e8 c08165c0 f55601c0
Nov 7 20:54:09 mediapc kernel: [ 2280.864105] f69cbed0 f69a6c14 f69a6c18 ffffffff f6559f30 c056fe56 f708cb60 f69a6c1c
Nov 7 20:54:09 mediapc kernel: [ 2280.864119] Call Trace:
Nov 7 20:54:09 mediapc kernel: [ 2280.864137] [<c056fe56>] __mutex_lock_slowpath+0xc6/0x130
Nov 7 20:54:09 mediapc kernel: [ 2280.864145] [<c056fd70>] mutex_lock+0x20/0x40
Nov 7 20:54:09 mediapc kernel: [ 2280.864203] [<f824ec0a>] i915_gem_retire_work_handler+0x2a/0x70 [i915]
Nov 7 20:54:09 mediapc kernel: [ 2280.864214] [<c0157a3e>] run_workqueue+0x6e/0x140
Nov 7 20:54:09 mediapc kernel: [ 2280.864246] [<f824ebe0>] ? i915_gem_retire_work_handler+0x0/0x70 [i915]
Nov 7 20:54:09 mediapc kernel: [ 2280.864255] [<c0157b98>] worker_thread+0x88/0xe0
Nov 7 20:54:09 mediapc kernel: [ 2280.864264] [<c015c240>] ? autoremove_wake_function+0x0/0x40
Nov 7 20:54:09 mediapc kernel: [ 2280.864272] [<c0157b10>] ? worker_thread+0x0/0xe0
Nov 7 20:54:09 mediapc kernel: [ 2280.864279] [<c015bf4c>] kthread+0x7c/0x90
Nov 7 20:54:09 mediapc kernel: [ 2280.864286] [<c015bed0>] ? kthread+0x0/0x90
Nov 7 20:54:09 mediapc kernel: [ 2280.864295] [<c0104007>] kernel_thread_helper+0x7/0x10

Hi everyone.
I installed kernel 2.6.31-15.49 from the repositories. Now my E169 works!
Nevertheless, it is still present the little problem of posts #160 and #175, so sometimes it is necessary to unplug and plug again the modem before the system detects it.

If you need any sort of log, just ask.

2.6.31-15-generic #49-Ubuntu SMP Fri Nov 6 09:55:13 UTC 2009 x86_64 GNU/Linux from Karmic proposed repo works with E620 here

patrick@linuxps:~$ lsusb
Bus 007 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 006 Device 003: ID 08ff:1600 AuthenTec, Inc. AES1600
Bus 006 Device 002: ID 0b05:1751 ASUSTek Computer, Inc.
Bus 006 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 005 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 001 Device 008: ID 12d1:1001 Huawei Technologies Co., Ltd. E620 USB Modem
Bus 001 Device 006: ID 064e:f111 Suyin Corp.
Bus 001 Device 005: ID 2040:7070 Hauppauge Nova-T Stick 3
Bus 001 Device 003: ID 152d:2329 JMicron Technology Corp. / JMicron USA Technology Corp.
Bus 001 Device 002: ID 0409:0059 NEC Corp. HighSpeed Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 004 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
Bus 003 Device 002: ID 1241:1122 Belkin Typhoon Stream Optical Mouse USB+PS/2
Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

dmesg attached

Patrick

Hi,

I tried the "rmmod usb-storage" and "modprobe usbserial vendor=0x12d1 product=0x1003" fix before I tried to install a new kernel, and that did not work. After I could not find any linux2.6.31-15.49.deb (read comment #193) files anywhere (Although I think I just don't know where launchpad.net has hidden them, hehe, and I googled for it aswel), I thought of putting the "rmm...." fix into /etc/rc.local (so that it would do all that during startup) and that worked fine on kernel "2.6.31-14" which I think is under the normal repositories. This fix worked on two laptops (dell and a hp).

I can give a step by step how-to for any relative newby (like me) if someone has the same problem.

On Sat, 2009-11-07 at 19:39 +0000, Justin Schoeman wrote:
> Just reverted from apw10 to 2.6.31-11.
>
> E220 works well in apw10, but the computer (Intel Littlefalls 2
> integrated mb/cpu) dies after around 30 minutes. The screen blanks, and
> the computer stops responding to keyboard, mouse and ACPI events, BUT it
> is still running (networking up, and logging to /var/log/messages).

You should create a new bug report for that one. Also, I would suggest
you paste the -entire- log :-) (or at least the missing lines before
what you pasted) and add informations about the HW setup.

Cheers,
Ben.

> I get the following errors in /var/log/messages:
> Nov 7 20:54:09 mediapc kernel: [ 2280.864065] i915/3 D c08165c0 0 295 2 0x00000000
> Nov 7 20:54:09 mediapc kernel: [ 2280.864076] f6559f04 00000046 00000000 c08165c0 f69cc168 c08165c0 1574e75d 000001e8
> Nov 7 20:54:09 mediapc kernel: [ 2280.864091] c08165c0 c08165c0 f69cc168 c08165c0 1574d741 000001e8 c08165c0 f55601c0
> Nov 7 20:54:09 mediapc kernel: [ 2280.864105] f69cbed0 f69a6c14 f69a6c18 ffffffff f6559f30 c056fe56 f708cb60 f69a6c1c
> Nov 7 20:54:09 mediapc kernel: [ 2280.864119] Call Trace:
> Nov 7 20:54:09 mediapc kernel: [ 2280.864137] [<c056fe56>] __mutex_lock_slowpath+0xc6/0x130
> Nov 7 20:54:09 mediapc kernel: [ 2280.864145] [<c056fd70>] mutex_lock+0x20/0x40
> Nov 7 20:54:09 mediapc kernel: [ 2280.864203] [<f824ec0a>] i915_gem_retire_work_handler+0x2a/0x70 [i915]
> Nov 7 20:54:09 mediapc kernel: [ 2280.864214] [<c0157a3e>] run_workqueue+0x6e/0x140
> Nov 7 20:54:09 mediapc kernel: [ 2280.864246] [<f824ebe0>] ? i915_gem_retire_work_handler+0x0/0x70 [i915]
> Nov 7 20:54:09 mediapc kernel: [ 2280.864255] [<c0157b98>] worker_thread+0x88/0xe0
> Nov 7 20:54:09 mediapc kernel: [ 2280.864264] [<c015c240>] ? autoremove_wake_function+0x0/0x40
> Nov 7 20:54:09 mediapc kernel: [ 2280.864272] [<c0157b10>] ? worker_thread+0x0/0xe0
> Nov 7 20:54:09 mediapc kernel: [ 2280.864279] [<c015bf4c>] kthread+0x7c/0x90
> Nov 7 20:54:09 mediapc kernel: [ 2280.864286] [<c015bed0>] ? kthread+0x0/0x90
> Nov 7 20:54:09 mediapc kernel: [ 2280.864295] [<c0104007>] kernel_thread_helper+0x7/0x10
>

On Sat, 2009-11-07 at 15:53 +0000, Justin Schoeman wrote:
> Next log...
>
> ** Attachment added: "usbmon-apw5.bz2"
> http://launchpadlibrarian.net/35317459/usbmon-apw5.bz2

Thanks. I've referenced these from the kernel bugzilla entry

http://bugzilla.kernel.org/show_bug.cgi?id=14499

Where hopefully the kernel USB folks will follow up

Cheers,
Ben.

The kernel meta-packages (linux, linux-generic etc) do not seem to trigger an upgrade to *-15.49 although the new packages are shown in Synaptic after a reload in the what's new in repository section. If this is widespread, it does not seem normal behavior and will make it harder for novice users to upgrade. Running apt-get update and apt-get upgrade has similar results.

Are others experiencing this?

Russell Faull wrote:

> The kernel meta-packages (linux, linux-generic etc) do not seem to trigger an upgrade to *-15.49 although the
> new packages are shown in Synaptic after a reload in the what's new in repository section. If this is
> widespread, it does not seem normal behavior and will make it harder for novice users to upgrade. Running
> apt-get update and apt-get upgrade has similar results.

> Are others experiencing this?

Yes, I had the same issue. apt-get dist-upgrade did not pull these packages in. I only got aware that the packages are actually in the repo after I manually installed the kernel debs from https://launchpad.net/ubuntu/karmic/+source/linux/2.6.31-15.49 but forgot one of the kernel header files. apt-get dist-upgrade suggested that I execute "apt-get -f install". And voilà, when I did this, apt-get pulled the missing deb from the -proposed repo. To make sure everything fit together I removed the manually installed kernel and installed the packages with apt-get using their full names, which worked.

Patrick

@Patrick. Have you tried to remove kernel *-14? When I do with *-15 installed, Synaptic wants to remove the associated meta-package as well. (To remove the meta-package, would mean any future kernel upgrades would not be triggered automatically.)

With previous kernel upgrades, the earlier kernel can be removed and the meta-package is associated with the latest kernel. (apt-get -f install had no effect.)

Later, I uninstalled *-15 and it showed in Synaptic as an available package but, again, not as 'upgradeable'. Strange!

This really should have been fixed before release. Lots of people are having a difficult time because of this issue.
Ive just been trying to tell, over IRC, how to get the kernel 2.6.31-15 installed when you have no network connection as the Huawei modem was the sole way to connect.

Now she has it installed, Huawei e160 modem, but Network manager does not automatically see the modem, nor does it ask for the ISP info automatically. She tells me NM sees that there is some modem, but does not configure it correctly, nor does it bring up a connection. On Jaunty all this happened automatically.

I have a e220, this had problems also - NM did not automatically configure it, I needed to add the connection manually, and even then it did not tell me I have a Huawei connected until AFTER I had configured the connection (entered the ISP info and so on), and after that I had to tell NM to Auto-connect or it would not connect.

There is still something fishy going on. Im in a bit of a hurry now, Ill try to see about logs later. I cannot get to her logs, she lives across the country and has no net conn on the machine now (with the e160)

My lsusb:
Bus 002 Device 006: ID 12d1:1003 Huawei Technologies Co., Ltd. E220 HSDPA Modem / E270 HSDPA/HSUPA Modem

My connection now works, with the caveats above - ie. NM did not automatically detect it the first time, it did not show up in Mobile broadband devices list before creating a profile for it - with Generic device, then clicking Autoconnect on, at which point it did detect the Huawei and did connect.

My friend with the e160 is not reporting success yet, though I cannot know if she is trying to do the configuring correctly via NM.

Tested apw5 on a dell d830 using amd64 with e160, e169g and e220 all seem fine. I haven't noticed any other regressions caused by the fix in the last 24hrs.

I think the problem is simply that the meta-packages depend on the latest package from karmic-updates, but not on the latest released package from karmic-proposed. If they depended on the latest proposed package, this would cause a headache to people that don't have karmic-proposed enabled. So there is probably no bug whatsoever, this is just the way it should be. Can someone confirm this ?

New metapackages are also published in -proposed as part of a kernel SRU, but they generally lag behind the upload of the kernel itself. The updated metapackages should be available soon.

Thanks, Steve. I'll sleep on it and see what Monday brings as far as the meta-package issue goes.

I still believe the new kernel *-15 causes new regression with suspend/resume and possibly hibernation ----:) But, hey, the modem works, I can live with no suspend.

Activating 'karmic-proposed' and 'karmic-backports' I installed 2.26.31-15 kernel automacally. Now my E169 works perfectly. I can read in-built CD-data and use the SD-storage with a 16 Gb Micro-SD card ... and 3G modem works perfectly. But kernel.log (dmesg) shows an error:

modem-manager[1067]: segfault at 7ffffff7 ip b777d943 sp bffe4284 error 4 in libdbus-1.so.3.4.0[b7756000+37000]

In spite of this, everything seems to work perfectly.

Sorry for the late reply...I my modem and SD now work with new download :-) cheers to all...nice one Ben...

My Huawei E169 works now too! Activated proposed and installed only headers and image. I am however seeing errors in dmesg at boot that were not there before! Attaching dmesg output.?field.comment=My Huawei E169 works now too! Activated proposed and installed only headers and image. I am however seeing errors in dmesg at boot that were not there before! Attaching dmesg output.

My Huawei E169 works now too! Activated proposed and installed only headers and image. I am however seeing errors in dmesg at boot that were not there before! Attaching dmesg output.

Still no success with my Huawei E220 and the new 2.6.31-15 image.

finally the usb modem working fine , just i have upgrade to last kernel

thank you

I would suggest that the kernel is probably now fixed, but a number of people are still having trouble with Network Manager.

My recommendation is to install wicd, which will uninstall NM, to see if that fixes it. You can always reverse it later.

Also, use Vodafone Mobile Connect, https://forge.betavine.net/frs/?group_id=12. It does actually work with any ISP, not just Vodafone. It has the big advantage of providing a usage monitor, amongst other things. If it seems not to work, be aware that it sometimes pops up its initial configuration window behind the main one, and once you find it, there should not be any problem setting it up. You will need your PIN of course, typically 0000.

You would need to install ozerocdoff, usb-modeswitch and then vodafone-mobile-connect (try the latest svn version). The debs will install singly with Gdebi, but whatever package manager you use will offer to update usb-modeswitch to a later version, which seems to be ok to allow.

I am thinking that if sufficient people try this, it should help to locate the source of the remaining problems, and/or get those people working again, even if it is only a temporary fix. It works for me, or I would not suggest it.

Further to that, could Vodafone Mobile Connect be added to the repositories, and preferably the next CD image, as it is GPL? It would make it easier for inexperienced users, and more convenient for everyone.

Here are my test results with the Huawei E160G:

Test 1: Current officially released kernel
I have clicked 3 times manually on Network Manager->(Mobile Broadband) Drei and nothing helped.

The log1a is after about 5 minutes after suddenly (didn't expect it anymore) the connection has been established.

Test 2: apw5 Kernel
I have clicked 2 times manually on Network Manager->(Mobile Broadband) Drei and nothing helped.
In the logs log2 is with WLAN additionally enabled, log2a is without.

Test 3: apw10 Kernel
On the first attempt plugging the mobile internet stick nothing happened (log3).
On the second attempt (log3a) it worked.
I notice that the drive does not get mounted as in Test 1 (and unfortunately I don't remember how it was on test2).

Test 4: Most recent development kernel (2.6.32-020632rc6)
log4 is my first attempt. The drive from the stick gets mounted. After clicking manually on the Network Manager->(Mobile Broadband) Drei it works.
Second attempt fails - I can't get it to work any more even if clicking several times manually to connect.

Test 5: Most recent development kernel (2.6.32-020632rc6) - After another reboot
I plug in the stick and wait until the message is displayed that GSM is disconnected. I click on 'Drei'.
It does not work either (log5).
Second attempt without success (log5a). BTW: Drive from the stick gets mounted.
Third attempt without success (log5b). BTW: Drive from the stick gets mounted. I try unmount and then reconnect without success.
Fourth attempt (log5c) using other USB port - message that disconnected comes quite fast, then drive gets mounted - no success.
Fifth attempt (log5d)
Sixth attempt (log5e) logged in a separate session before login - anyway did not work with the modem plugged in before login.

Test 6: Back on apw10 Kernel
It works only the first time I connect the stick!
Each further attempt (I did a few times) it does not work any more (log6).

What I notice is that in general the animated icon of network manager hangs with the tooltip "Preparing mobile broadband connection 'Drei'" after I plugged in the.

==> However, the apw10 kernel seems the only one I can get a connection at least the first time - did not try more often. I remember that on Jaunty with some older kernel I sometimes had to plugin the stick also a few times until it worked.

==> Anyway, my opinion is that the last released kernel in Jaunty was the one where it worked best for me.

Ah, btw: The apw10 kernel - I did verify that - does not map the drive from the stick.
So if that gets not mapped it is a pretty good sign that it works (same behavior was also on Jaunty where it worked fine).

@Martin, have you tried 2.6.31-15 #49 from the 'proposed' repos?