wireless network not working now

When problems occur with wireless card, can sometime ifdown ra0 (even though it doesn't have IP address from dhcp) then ifup ra0, it sometimes gets IP address, but with following message :-run-parts: /etc/network/if-up.d/avahi-autoipd exited with return code 2

root@feisty-64b:/home/kit284# ifup ra0
There is already a pid file /var/run/dhclient.ra0.pid with pid 6847440
Internet Systems Consortium DHCP Client V3.0.4
Copyright 2004-2006 Internet Systems Consortium.
All rights reserved.
For info, please visit http://www.isc.org/sw/dhcp/

Listening on LPF/ra0/00:13:d3:6e:ae:37
Sending on LPF/ra0/00:13:d3:6e:ae:37
Sending on Socket/fallback
DHCPDISCOVER on ra0 to 255.255.255.255 port 67 interval 4
DHCPDISCOVER on ra0 to 255.255.255.255 port 67 interval 8
DHCPOFFER from 192.168.0.1
DHCPREQUEST on ra0 to 255.255.255.255 port 67
DHCPACK from 192.168.0.1
bound to 192.168.0.76 -- renewal in 244555 seconds.
RTNETLINK answers: File exists
run-parts: /etc/network/if-up.d/avahi-autoipd exited with return code 2

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

I have given up with the network drivers and now compile my own modules. The standard ra0 driver does not work with the "iwpriv" commands which is a real pain for me.

I now use the ra0 driver from the web site where the original (with iwpriv) commands work. The reason for this is that I now use the bonding module, and bond the ra0 and the eth0 into the bond0 interface. This allows me work on gigabit wired interface when available, but if I need to move it to another room, I disconnect the wired interface and it fails over to the wireless interface. WIth bonding, the MAC address of the active interface (only one running at a time) is the same, correspondingly the DHCP address is also the same, so that my MySQL server on the laptop with a public LAN address is still available on wither wired or wireless. The same MAC and IP is important whichever network is active.

Bonding does not work with wpa-supplicant which is the standard with Ubuntu driver for my wireless card running WPA or WPA2.

I would not be happy going to WEP or no security just to fit in with the standard distribution, especially after they have removed functionality that was (and still is) important to the way I work.

Summary:- the fault is still there, functionality has been removed for whatever reason, and I have to hack the system with my own compiled modules to get it to do what it did before in previous releases. If you have another work-around that meets the above specification, please feel free to contact me.

(uname -a ==> Linux pio 2.6.22-14-generic #1 SMP Thu Jan 31 23:33:13 UTC 2008 x86_64 GNU/Linux)

Hello,
Ralink drivers are in kernel tree since 2.4.24 (linux kernel version included in Hardy).Is posible that this bug is a duplicate of bug 134660

However this issue is one that should be reproducible with the live environment of the Desktop CD of the development release - Hardy Heron. It would help us greatly if you could test with it so we can work on getting it fixed in the actively developed release. You can find out more about the development release at http://www.ubuntu.com/testing/

Hi there - the problem is that up until herd 3/4 (see above) the module for the ralink wireless was rt2500 which had private ioctls that allows on chip WPA encoding/decoding of WPA. SInce then, the standard Ubuntu has gone with the rt2500pci/rt2x00lib/rt2x00pci modules that do not have private ioctls.

Even though the hardware still has the ability to encode/decode WPA traffic on-chip, the newer modules have the functionality removed.

This means that now (since gutsy), when the rt2500pci/rt2x00lib/rt2x00pci modules are loaded, you must also configure wpa_supplicant to decode/decode the WPA wireless traffic. When you have bonded interfaces, where one of the interfaces is a wireless network as mine is, wpa_supplicant cannot work, so my only option is to blacklist the current modules, and recompile the rt2500 modules from serialmonkey so the functionality is restored.

It is not possible that this is a duplicate of bug 134660 - they are describing something different, although some of the comments do refer to this problem.
----------------------------------------
TESTING WITH HARDY ALPHA 4
the command iwpriv gives the following:-
root@ubuntu:/home/ubuntu# iwpriv
lo no private ioctls.

eth0 no private ioctls.

irda0 no private ioctls.

wmaster0 no private ioctls.

wlan0 no private ioctls.
---------------------------------------
TESTING WITH GUTSY WITH STANDARD MODULES BLACKLISTED AND RT2500 COMPILED AND LOADED
root@pio:/tmp# iwpriv
lo no private ioctls.

eth0 no private ioctls.

wlan0 Available private ioctls :
          set (8BE2) : set 1024 char & get 0
          bbp (8BE3) : set 1024 char & get 1024 char
          mac (8BE5) : set 1024 char & get 1024 char
          e2p (8BE7) : set 1024 char & get 1024 char
          rfmontx (8BED) : set 2 int & get 1 char

irda0 no private ioctls.

bond0 no private ioctls.

--------------------------------------------------------------------

You can see the private ioctls for the wlan0 interface which are used by the rt2500 network module. THESE ARE MISSING FROM THE CURRENT GUTSY MODULES AND THE MODULES IN HARDY HERON ALPHA 4 RELEASE. (downloaded and run on Sun 10th Feb 2008)

To get this to work again, someone would need to recode the current rt2500pci, rt2x00pci and rt2x00lib modules in gutsy/hardy to put back the private ioctls.
(these modules are in /lib/modules/2.6.22-14-generic/ubuntu/wireless/rt2x00/ directory)

Unless this is done, this should be still considered as a bug, as the functionality of the private ioctls has been removed in favour of using wpa_supplicant to encode/decode WPA traffic.modules/2.6.22-14-generic/ubuntu/wireless/rt2x00/

Hope this makes it clearer, if not please contact me again. It would be good to get this going again, otherwise I am burdened with having to black list standard modules and recompiling the serialmonkey variety, which is a real pain everytime a new kernel version is released.

I have a similar problem with an iwl4965 wireless device. wpa_supplicant works fine with WEP and WPA when I use the device independently, however when it is bonded, only WEP works, and WPA does not. Here is the output from dmesg:

Mar 30 13:20:55 localhost kernel: wlan0: Initial auth_alg=0
Mar 30 13:20:55 localhost kernel: wlan0: authenticate with AP 00:1c:10:41:cd:9e
Mar 30 13:20:55 localhost kernel: wlan0: Initial auth_alg=0
Mar 30 13:20:55 localhost kernel: wlan0: authenticate with AP 00:1c:10:41:cd:9e
Mar 30 13:20:55 localhost kernel: wlan0: RX authentication from 00:1c:10:41:cd:9e (alg=0 transaction=2 status=0)
Mar 30 13:20:55 localhost kernel: wlan0: authenticated
Mar 30 13:20:55 localhost kernel: wlan0: associate with AP 00:1c:10:41:cd:9e
Mar 30 13:20:55 localhost kernel: wlan0: authentication frame received from 00:1c:10:41:cd:9e, but not in authenticate state - ignored
Mar 30 13:20:55 localhost kernel: wlan0: authentication frame received from 00:1c:10:41:cd:9e, but not in authenticate state - ignored
Mar 30 13:20:55 localhost kernel: wlan0: RX AssocResp from 00:1c:10:41:cd:9e (capab=0x431 status=0 aid=1)
Mar 30 13:20:55 localhost kernel: wlan0: associated
Mar 30 13:20:55 localhost kernel: wlan0: switched to long barker preamble (BSSID=00:1c:10:41:cd:9e)
Mar 30 13:20:55 localhost kernel: bonding: bond0: link status definitely up for interface wlan0.
Mar 30 13:20:59 localhost kernel: wlan0: RX deauthentication from 00:1c:10:41:cd:9e (reason=15)
Mar 30 13:20:59 localhost kernel: wlan0: deauthenticated
Mar 30 13:20:59 localhost kernel: bonding: bond0: link status definitely down for interface wlan0, disabling it
Mar 30 13:21:00 localhost kernel: wlan0: authenticate with AP 00:1c:10:41:cd:9e
Mar 30 13:21:00 localhost kernel: wlan0: RX authentication from 00:1c:10:41:cd:9e (alg=0 transaction=2 status=0)
Mar 30 13:21:00 localhost kernel: wlan0: authenticated
Mar 30 13:21:00 localhost kernel: wlan0: associate with AP 00:1c:10:41:cd:9e
Mar 30 13:21:00 localhost kernel: wlan0: RX ReassocResp from 00:1c:10:41:cd:9e (capab=0x431 status=0 aid=1)
Mar 30 13:21:00 localhost kernel: wlan0: associated
Mar 30 13:21:00 localhost kernel: wlan0: switched to long barker preamble (BSSID=00:1c:10:41:cd:9e)
Mar 30 13:21:00 localhost kernel: iwl4965: Error sending REPLY_RXON_ASSOC: Already sending a host command
Mar 30 13:21:00 localhost kernel: bonding: bond0: link status definitely up for interface wlan0.
Mar 30 13:21:04 localhost kernel: wlan0: RX deauthentication from 00:1c:10:41:cd:9e (reason=15)
Mar 30 13:21:04 localhost kernel: wlan0: deauthenticated
Mar 30 13:21:04 localhost kernel: bonding: bond0: link status definitely down for interface wlan0, disabling it

This repeats in a loop - successful association with the AP, then 4 seconds after the bonding device notes that the wireless link is definitely up, there is a deauthentication with reason=15.

As I said above, WEP works just fine, only WPA is failing. WPA works as long as bonding isn't involved.

I'm using kernel 2.6.24.1, and have tried with the 2.6.24.1 iwl4965 module, as well as two of the more recent development snapshots from the compat-wireless project. wpasupplicant is version 0.6.3-1

The Ubuntu Kernel Team is planning to move to the 2.6.27 kernel for the upcoming Intrepid Ibex 8.10 release. As a result, the kernel team would appreciate it if you could please test this newer 2.6.27 Ubuntu kernel. There are one of two ways you should be able to test:

1) If you are comfortable installing packages on your own, the linux-image-2.6.27-* package is currently available for you to install and test.

--or--

2) The upcoming Alpha5 for Intrepid Ibex 8.10 will contain this newer 2.6.27 Ubuntu kernel. Alpha5 is set to be released Thursday Sept 4. Please watch http://www.ubuntu.com/testing for Alpha5 to be announced. You should then be able to test via a LiveCD.

Please let us know immediately if this newer 2.6.27 kernel resolves the bug reported here or if the issue remains. More importantly, please open a new bug report for each new bug/regression introduced by the 2.6.27 kernel and tag the bug report with 'linux-2.6.27'. Also, please specifically note if the issue does or does not appear in the 2.6.26 kernel. Thanks again, we really appreicate your help and feedback.

i'm running ubuntu 8.10 and the problem is still there

@Cereal can you comment which driver you are using and also attach and updated dmesg output which captures the errors. Thanks.

Leann Ogasawara wrote:
> @Cereal can you comment which driver you are using and also attach and
> updated dmesg output which captures the errors. Thanks.
>
>
comments from bug report copied to here:-

==============start of copy==============

Hi there - the problem is that up until herd 3/4 (see above) the module
for the ralink wireless was rt2500 which had private ioctls that allows
on chip WPA encoding/decoding of WPA. SInce then, the standard Ubuntu
has gone with the rt2500pci/rt2x00lib/rt2x00pci modules that do not have
private ioctls.

Even though the hardware still has the ability to encode/decode WPA
traffic on-chip, the newer modules have the functionality removed.

This means that now (since gutsy), when the
rt2500pci/rt2x00lib/rt2x00pci modules are loaded, you must also
configure wpa_supplicant to decode/decode the WPA wireless traffic. When
you have bonded interfaces, where one of the interfaces is a wireless
network as mine is, wpa_supplicant cannot work, so my only option is to
blacklist the current modules, and recompile the rt2500 modules from
serialmonkey so the functionality is restored.

It is not possible that this is a duplicate of bug 134660
<https://bugs.launchpad.net/bugs/134660> - they are describing something
different, although some of the comments do refer to this problem.
===============end of copy============

The problem is still EXACTLY the same - NO PRIVATE IOCTL for the
wireless lan driver in the supplied modules.

@Scott Barker:

I have the _exact_ same problem on the same wireless chipset as well as on another machine with ipw2200. Both intel chipsets FWIW. Did you ever resolve this issue?

No - research shows that the chip inbuilt WPA and WEP crypt/decrypt are
not supported in the general driver scheme they have introduced for the
APIs. In any case, the inbuilt crypt/decrypt are limited to WEP and
WPA, whereas the standard API allows for wpa_supplicant (WEP,WPA and
WPA2, and other soon to be introduced), which the inbuilt drivers don't.

It is a pain in the ass, because I have used the bonding module to bond
the standard wired interface and the wireless interface, to give a
uniform IP address and MAC address no matter whether I am on the wired
or wireless connection. This is important for mysql servers, files
servers and mounting remote file systems from my laptop, as the network
would failover without hiccup from wired to wireless. (and back again)

Now, I have problems when I change from wired to wireless, I have to
stop my servers, unmount all remote file systems, shut down applications
for remote file systems such as mythtv, (not in that order), change to
wireless, then start them up again.

It appears that the people running the bug list don't understand what is
going on - it is clear from their roadmap that they are standardising
the driver APIs so that applications such as wpa_supplicant can be
used. This precludes the use of the inbuilt routines.

I have (in the past) blacklisted the normal module for the wireless,
recompiled the old module (from serialmonkey web site), and kept on
going, but it was becoming a pain with every module or kernel update.

If you need further instruction on the previous (outdated) version of
the driver, let me know, and I can point you in the right direction. I
am hoping that soon they will fix the problems with wpa_supplicant so
that I can go back to bonded interfaces and have it work
"out-of-the-box", and be immune from problems with kernel/module updates.

I have explained the situation via the bug reports and other emails
until I am blue in the face (and fingers), but the recipients don't seem
to understand the problem I face. (just like talking to a help desk
"please power off then power on to see if that fixes the problem...But
my computer is on fire!!! etc.etc.etc....."

I will probably let them close this bug due to apathy on my part and
reluctance to help on their part - c'est la vie.

Hope this helps.

Brian J. Murrell wrote:
> @Scott Barker:
>
> I have the _exact_ same problem on the same wireless chipset as well as
> on another machine with ipw2200. Both intel chipsets FWIW. Did you
> ever resolve this issue?
>
>

I have not yet been able to use WPA2 over my iwl4965 when bonded to my wired interface.

Ralink Tech adapters are now supported with standard mac80211 drivers.

I see here that a fixed was released.

It is now 2009 of August. I am on 9.04 jaunty

I am trying to use /etc/network/interfaces

this is what it looks like...

auto wlan0
iface wlan0 inet dhcp
        pre-up ifconfig wlan0 down
        pre-up ifconfig wlan0 up
        pre-up iwconfig wlan0 essid myap
        pre-up iwconfig wlan0 mode Managed
        pre-up iwconfig wlan0 channel 11
        pre-up iwpriv wlan0 set AuthMode=WPA2PSK
       pre-up iwpriv wlan0 set EncrypType=AES
       pre-up iwpriv wlan0 set WPAPSK=jdfkasjfkajsfpajfdakjf;lkaj;flja;lfjijpeijpfj

This produces the errors

wlan0 no private ioctls.

I'm going to file a seperate bug as well but just wanted to try here first.

Thanks.