[0.7] [3G] "PIN secret invalid" even though PIN is correct

Bug #286859 reported by Dennis Heinson
16
This bug affects 2 people
Affects Status Importance Assigned to Milestone
network-manager (Ubuntu)
Invalid
Undecided
Unassigned
network-manager (openSUSE)
Won't Fix
Unknown

Bug Description

Binary package hint: network-manager
Version: 0.7~~svn20081015t224738-0ubuntu1

I am using a Sierra Aircard MC8781. When I try to connect to the network, I get asked for my PIN (which I enter correctly) but it is not accepted. I know this from the logs - I am attaching relevant log and console output to this bug report.
As a feature request, it would be nice if the user were informed why he is getting asked for the PIN once more.

Note: It will also ask me for the PIN when I enter it statically into the "saved networks" dialogue.

dennis@Computer:~$ hal-find-by-capability --capability modem
/org/freedesktop/Hal/devices/usb_device_1199_6856_noserial_if0_serial_usb_0

The second attachment (lshal.txt) is the output from:
dennis@Computer:~$ lshal -u `hal-find-by-capability --capability modem`

Revision history for this message
In , Stefan Seyfried (seife) wrote :

I think that Tanja has had a similar problem, but on a one-port Novatel XU870 Card

Revision history for this message
In , Tambet (tambet) wrote :

So these cards don't implement GSM-07.05 or GSM-07.07 standards correctly. Currently, there are no workarounds for any quirks in NetworkManager.

It's not really decided yet how to solve these issues. Not sure yet if HAL should mark non standard devices with some special attributes or to handle that in NM. Also, having workarounds for all the different cards in the world makes the code totally unmaintainable, so we're looking into using some out of band implementation or making NM pluggable so that every non standard device can have it's own plugin. Something that happens after 0.7 (and opensuse 11.0).

Revision history for this message
In , Stefan Seyfried (seife) wrote :

After looking at the sierra website today where they say that for some cards one should use ttyUSB2, for others ttyUSB0, i just tried this, the quick'n dirty way, by linking /dev/ttyUSB2 to ttyUSB0:

root@stoetzler:~# ls -l /dev/ttyUSB0
lrwxrwxrwx 1 root root 12 23. Mai 00:19 /dev/ttyUSB0 -> /dev/ttyUSB2

And this works fine!
But there is no way that you can get signal strength while connected with this setup (the website says that all AT commands should go via ttyUSB2 and pppd connections should go via ttyUSB0. But beware - if you run pppd on ttyUSB0, the "atdt*99#" also must go on ttyUSB0. We fell into that trap with umtsmon ;)

So how to proceed? Do we need to fix this in HAL? In NM? In both?

Revision history for this message
In , Tambet (tambet) wrote :

Can you get me the output of 'lshal -l' with these cards?

Revision history for this message
In , Stefan Seyfried (seife) wrote :

Unfortunately it is even harder and not easy to fix for those cards.

Danny and me have the same card (hp hs2300, a relabeled sierra wireless) in very similar machines (mine is a hp 2510p, danny's is a 2710p).

And now for the good part:
- my machine has ttyUSB2 as "command" and ttyUSB0 as "ppp" port
- dannys machine has it the other way round.

See also http://sourceforge.net/mailarchive/forum.php?thread_name=200807181937.19707.dkukawka%40suse.de&forum_name=umtsmon-develop

In umtsmon we solved it by querying the ports with "ATI" and parsing the answer (it is different on both ports).
However, this looks out of scope for NetworkManager.

Danny, could we add a small helper to HAL that probes the ports of such cards and then assigns the ports correctly (instead of statically adding them via a FDI file)?

Then NetworkManager can just use the corresponding tags on the ports to select which port to use for PIN entering and for ppp.

Revision history for this message
Dennis Heinson (dheinson) wrote :

Binary package hint: network-manager

I am using a Sierra Aircard MC8781. When I try to connect to the network, I get asked for my PIN (which I enter correctly) but it is not accepted. I know this from the logs - I am attaching relative log and console output to this bug report.
As a feature request, it would be nice if the user were informed why he is getting asked for the PIN once more.

Note: It will also ask me for the PIN when I enter it statically into the "saved networks" dialogue.

dennis@Computer:~$ hal-find-by-capability --capability modem
/org/freedesktop/Hal/devices/usb_device_1199_6856_noserial_if0_serial_usb_0

Revision history for this message
Dennis Heinson (dheinson) wrote :
Revision history for this message
Dennis Heinson (dheinson) wrote :
description: updated
description: updated
description: updated
Revision history for this message
Alexander Sack (asac) wrote :

could you please stop nm like: sudo /etc/init.d/NetworkManager stop

then tail -f -n0 /var/log/syslog > /tmp/syslog.txt

and start NetworkManager like:

export NM_SERIAL_DEBUG=1
sudo NetworkManager --no-daemon 2>&1 | tee /tmp/nm.txt

then reproduce your problems and when done please attach your syslog.txt and your nm.txt.

NOTE: remember to remove any sensible information like PIN from both logs before submitting!

Changed in network-manager:
status: New → Incomplete
Revision history for this message
Dennis Heinson (dheinson) wrote :

Thanks for the quick response. I am attaching the output as requested. Hope this can get sorted out before release.

Revision history for this message
Dennis Heinson (dheinson) wrote :
Revision history for this message
Alexander Sack (asac) wrote : Re: [Bug 286859] Re: [0.7] [3G] "PIN secret invalid" even though PIN is correct

On Tue, Oct 21, 2008 at 04:02:06PM -0000, Dennis Heinson wrote:
>
> ** Attachment added: "nm.txt"
> http://launchpadlibrarian.net/18771466/nm.txt
>

hm ... unfortunately the NM_SERIAL_DEBUG=1 environment was not enough
to get your serial debug. A bit strange I would say. Are you sure you
are running latest network manager package?

 - Alexander

Revision history for this message
Dennis Heinson (dheinson) wrote :

I think so. But I must add that before my recent distro-upgrade I was runnning a network manager from the PPA that included UMTS support. But that has been removed in the upgrading process, correct? Or do I need to delete some old configuration data? Is there any other way of obtaining the serial data?

dennis@Computer:~$ apt-cache policy network-manager
network-manager:
  Installed: 0.7~~svn20081018t105859-0ubuntu1
  Candidate: 0.7~~svn20081018t105859-0ubuntu1
  Version table:
 *** 0.7~~svn20081018t105859-0ubuntu1 0
        500 http://de.archive.ubuntu.com intrepid/main Packages
        100 /var/lib/dpkg/status

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

could you test the patch posted by Jerone in bug 278992 ?

Revision history for this message
In , Aj-novell (aj-novell) wrote :

Danny, Seife, what shall we do here?

Revision history for this message
In , Dkukawka (dkukawka) wrote :

Is this only needed for Sierra Wireless?

Revision history for this message
In , Dkukawka (dkukawka) wrote :

Btw. Is the probing something we can do in NM?

Revision history for this message
In , Tambet (tambet) wrote :

In ModemManager, yes. I just need to know the usb ids and serial device names.

Revision history for this message
In , Dkukawka (dkukawka) wrote :

You cant differ them on the USB IDs or the device name. You need a prober for all Sierra Wireless devices.

Revision history for this message
In , Stefan Seyfried (seife) wrote :

My device is:
  Bus 004 Device 002: ID 03f0:1e1d Hewlett-Packard

It has ttyUSB0=ppp, ttyUSB2=control

Danny's card is exactly the same, but has ttyUSB0=control, ttyUSB2=ppp. It apparently depends on the mainboard you plug the card in or even something different. I believe I have seen today a 2710p (same as Danny's) with ttyUSB0=ppp, ttyUSB2=control :-(

Holgi, could you please post the output of "lsusb" and "grep -A3 '\[device\]' /usr/lib/qt3/etc/settings/umtsmonrc" on your 2710p?

Revision history for this message
In , Holgi-novell (holgi-novell) wrote :

Created an attachment (id=251522)
lsusb.txt

Revision history for this message
In , Holgi-novell (holgi-novell) wrote :

Created an attachment (id=251524)
device.txt

Revision history for this message
In , Tambet (tambet) wrote :

Can't this be fixed in kernel somehow? So that serial device names stay the same? Like in case of regular network devices, the device name always stays the same.

Revision history for this message
In , Stefan Seyfried (seife) wrote :

I don't think so, as I cannot see any differences in lsusb output between the two devices, but I am no USB expert either...

Revision history for this message
Dennis Heinson (dheinson) wrote :

I discovered something new:
Via #290177 there was a hint to manually unlock the 3G modem. I tried this to no avail with ttyUSB0. However, using ttyUSB2 did the job. Then, the connection fails with a message: Could not set APN.

My conclusion is that NM is talking to the Sierra Modem via the wrong ttyUSB port. Could someone with the necessary knowledge check whether NM is initializing on the wrong USB port?

I am using sierra driver v.1.3.2, kernel 2.6.27-7-generic

Thanks!

Revision history for this message
Dennis Heinson (dheinson) wrote :
Revision history for this message
Dennis Heinson (dheinson) wrote :

After manually setting the PIN and APN, and removing the APN entry in the network config, everything works and NM connects instantly.

So the problem is, I suppose, that NM talks to the Aircard on the wrong USB port (as can be seen in the logs).

Thanks!

Revision history for this message
In , Gregkh (gregkh) wrote :

We have discussed a way for the kernel to export to userspace exactly which port does what (settings, data connection, etc.) but no one has implemented it yet, sorry.

It might be something that is done for SLE12, depending on if it becomes something that is a large problem. For now, HAL has been handling these kinds of things (which port is which), and Device Kit in the future might be the better place for it.

As for persistant names, yes, that can be done with a simple udev rule, but again, you don't know which port to name what ahead of time, so it is a case-by-case basis right now.

Sorry I can't be of more help, and glad it's now working for you.

Revision history for this message
In , Stefan Seyfried (seife) wrote :

Danny, Holgi's card has the same port ordering as mine. How about your machine?

Revision history for this message
Alexander Sack (asac) wrote : Re: [Bug 286859] Re: [0.7] [3G] "PIN secret invalid" even though PIN is correct

On Tue, Nov 18, 2008 at 10:02:05AM -0000, Dennis Heinson wrote:
> After manually setting the PIN and APN, and removing the APN entry in
> the network config, everything works and NM connects instantly.
>
> So the problem is, I suppose, that NM talks to the Aircard on the wrong
> USB port (as can be seen in the logs).
>

please open a bug against hal-info, adding the output of lsusb -v
.... and stating that interface.number 2 works better.

you could also try to adjust the 10-modem.fdi to validate that moving
your products mapping to interfaces.number 2 helps.

 - Alexander

Revision history for this message
In , Stefan Seyfried (seife) wrote :

Greg, I was unable to find out how to identify the ports with a HAL fdi rule.
Can you please tell us what to look for?

Revision history for this message
In , Gregkh (gregkh) wrote :

Look at /usr/share/hal/fdi/information/10freedesktop/10-modem.fdi for how to define your specific device and the port of which it is attached to.

Revision history for this message
In , Dkukawka (dkukawka) wrote :

(In reply to comment #15 from Stefan Seyfried)
> I don't think so, as I cannot see any differences in lsusb output between the
> two devices, but I am no USB expert either...

The card in my 2710p is the same as in the machine of Holger (info from openSUSE 11.0):
-------------------------
[device]
ATPortName=/dev/ttyUSB0
DeviceType=0x51322402
PPPPortName=/dev/ttyUSB2

Bus 004 Device 002: ID 03f0:1e1d Hewlett-Packard
-------------------------

And IIRC this is the same card like the card in seifes hp2510.

Revision history for this message
In , Stefan Seyfried (seife) wrote :

(In reply to comment #25 from Greg Kroah-Hartman)
> Look at /usr/share/hal/fdi/information/10freedesktop/10-modem.fdi for how to
> define your specific device and the port of which it is attached to.

Greg, this is not helpful. We do not know what to match on to distinguish the Data from the Control port. Hint: the serial.port property is not useful.

We have the same card which is once ttyUSB0=data ttyUSB2=control and the other one is ttyUSB0=control and ttyUSB2=data.

What do we need to match on? Is there a usb descriptor or anyhting?

Danny: the lsusb would also be helpful, so that we can diff against Holgi's.

Revision history for this message
In , Gregkh (gregkh) wrote :

Can you provide me the output of 'lsusb -v' on both of the machines to see if these two devices really are identical or not?

Revision history for this message
In , Dkukawka (dkukawka) wrote :

@Holger: can you please provide the lsusb -v output?

Revision history for this message
In , Holgi-novell (holgi-novell) wrote :

Created an attachment (id=258479)
lsusb-v.txt

Sure. Here is the lsusb -v output of my "untwisted" 2710p

Revision history for this message
In , Stefan Seyfried (seife) wrote :

JFTR: the hs2300 in my 2510p also has exactly the same lsusb output.

Revision history for this message
In , Stefan Seyfried (seife) wrote :

danny: one idea... ;-)
Can you please provide the output of "find /sys -name port_number|xargs grep ." with your sierra card activated?

Revision history for this message
In , Dkukawka (dkukawka) wrote :

find /sys -name port_number|xargs grep .

/sys/devices/pci0000:00/0000:00:1a.1/usb3/3-1/3-1:1.0/ttyUSB0/port_number:0
/sys/devices/pci0000:00/0000:00:1a.1/usb3/3-1/3-1:1.0/ttyUSB1/port_number:1
/sys/devices/pci0000:00/0000:00:1a.1/usb3/3-1/3-1:1.0/ttyUSB2/port_number:2

Revision history for this message
In , Stefan Seyfried (seife) wrote :

bad. So this also does not help us :-(
(it is exactly the same on my machine, which has the ports unreversed)

Revision history for this message
In , Dkukawka (dkukawka) wrote :

I guess there is no other way than probing (as umtsmon do).

Changed in network-manager:
status: Unknown → Confirmed
Revision history for this message
AM (macchi) wrote :

I confirm this on Ubuntu 8.10. N-M keeps asking for PIN and PUK codes although the SIM is unlocked.

Please give me a hint on how to tweak the ppp and chat scripts used with N-M?

Thanks,
/AMIII

PS: I am having problems with Network Manager and a Huawei E600 (PCMCIA) for mobile internet connections through 3G. It is recognized as E620 and the connection process is stuck at a PIN question for an unlocked SIM card.

Some reference information N-M 0.7~~svn20081018... & N-M-Gnome 0.7~~20081020... on the very same system work fine with a Huawei E169. Furthermore, the E600 works fine on GNU/Linux directly with ppp scripts or under win.

Revision history for this message
Alexander Sack (asac) wrote : Re: [Bug 286859] Re: [0.7] [3G] "PIN secret invalid" even though PIN is correct

On Wed, Jan 14, 2009 at 10:42:58AM -0000, Antonio Macchi III wrote:
> I confirm this on Ubuntu 8.10. N-M keeps asking for PIN and PUK codes
> although the SIM is unlocked.
>
> Please give me a hint on how to tweak the ppp and chat scripts used with
> N-M?
>
>
> Thanks,
> /AMIII
>
>
> PS: I am having problems with Network Manager and a Huawei E600 (PCMCIA) for mobile internet connections through 3G. It is recognized as E620 and the connection process is stuck at a PIN question for an unlocked SIM card.
>
> Some reference information N-M 0.7~~svn20081018... & N-M-Gnome
> 0.7~~20081020... on the very same system work fine with a Huawei E169.
> Furthermore, the E600 works fine on GNU/Linux directly with ppp scripts
> or under win.
>

sorry for the absense. I think i made a mistake on my intial
instructions you have to become root first i think.

sudo su
export NM_SERIAL_DEBUG=1
NetworkManager --no-daemon 2>&1 | tee /tmp/nm.serial.txt

and attach the serial txt file after reproducing this twice please.

Thanks ... sorry for that.

 - Alexander

Revision history for this message
Olivier Lemaire (olivier-lemaire) wrote :

I could find a way to make NM work on Intrepid with my f$*g Sierra card:
 * boot windows and use the Sierra tools to "deactivate PIN security" (set the card not to ask the PIN code at each usage)
 * re-boot intrepid and set nothing in the PIN field of my broadband card setting
 * launch NM broadband connection - OK

Sure it's not perfect, but helped me use this 3G card...

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

On Tue, Feb 03, 2009 at 08:51:12PM -0000, Olivier Lemaire wrote:
> I could find a way to make NM work on Intrepid with my f$*g Sierra card:
> * boot windows and use the Sierra tools to "deactivate PIN security" (set the card not to ask the PIN code at each usage)
> * re-boot intrepid and set nothing in the PIN field of my broadband card setting
> * launch NM broadband connection - OK
>
> Sure it's not perfect, but helped me use this 3G card...
>

Getting the SERIAL log as posted before would still be
appreciated. Thanks!

 - Alexander

Revision history for this message
Olivier Lemaire (olivier-lemaire) wrote :

Here's the result of a lshal -u `hal-find-by-capability --capability modem` :
udi = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0_serial_usb_0'
  info.capabilities = {'serial', 'modem'} (string list)
  info.category = 'serial' (string)
  info.parent = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0' (string)
  info.product = 'MC8780 Device' (string)
  info.subsystem = 'tty' (string)
  info.udi = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0_serial_usb_0' (string)
  linux.device_file = '/dev/ttyUSB0' (string)
  linux.hotplug_type = 2 (0x2) (int)
  linux.subsystem = 'tty' (string)
  linux.sysfs_path = '/sys/devices/pci0000:00/0000:00:1d.2/usb3/3-2/3-2:1.0/ttyUSB0/tty/ttyUSB0' (string)
  modem.command_sets = {'GSM-07.07', 'GSM-07.05'} (string list)
  serial.device = '/dev/ttyUSB0' (string)
  serial.originating_device = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0' (string)
  serial.port = 0 (0x0) (int)
  serial.type = 'usb' (string)

Hope this'll help.

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

On Mon, Feb 23, 2009 at 11:40:20PM -0000, Olivier Lemaire wrote:
> Here's the result of a lshal -u `hal-find-by-capability --capability modem` :
> udi = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0_serial_usb_0'
> info.capabilities = {'serial', 'modem'} (string list)
> info.category = 'serial' (string)
> info.parent = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0' (string)
> info.product = 'MC8780 Device' (string)
> info.subsystem = 'tty' (string)
> info.udi = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0_serial_usb_0' (string)
> linux.device_file = '/dev/ttyUSB0' (string)
> linux.hotplug_type = 2 (0x2) (int)
> linux.subsystem = 'tty' (string)
> linux.sysfs_path = '/sys/devices/pci0000:00/0000:00:1d.2/usb3/3-2/3-2:1.0/ttyUSB0/tty/ttyUSB0' (string)
> modem.command_sets = {'GSM-07.07', 'GSM-07.05'} (string list)
> serial.device = '/dev/ttyUSB0' (string)
> serial.originating_device = '/org/freedesktop/Hal/devices/usb_device_1199_6832_noserial_if0' (string)
> serial.port = 0 (0x0) (int)
> serial.type = 'usb' (string)
>
> Hope this'll help.
>

I asked for the serial log as described in the previous posts, but
thatnks anyway.

 - Alexander

Revision history for this message
In , Bili (bili) wrote :

It's long time, now Sierra Wireless card works fine in 11.2 or 11.3.
->Won't fix. If any issue free to contact me.

Changed in network-manager (openSUSE):
status: Confirmed → Won't Fix
Revision history for this message
Bryan Quigley (bryanquigley) wrote :

Thank you for taking the time to report this bug and helping to make Ubuntu better. We are sorry that we do not always have the capacity to look at all reported bugs in a timely manner. There have been many changes in Ubuntu since that time you reported the bug and your problem may have been fixed with some of the updates. It would help us a lot if you could test it on a currently supported Ubuntu version.

This bug was marked Incomplete some time ago and was supposed to autoclose after 90 days. Another bug prevented that so I'm closing it now.

Changed in network-manager (Ubuntu):
status: Incomplete → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

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