https://bugzilla.novell.com/show_bug.cgi?id=441511 User gp@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c1 Gerald Pfeifer changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |deckel@novell.com, marc.ruehrschneck@novell.com --- Comment #1 from Gerald Pfeifer 2008-11-04 08:37:24 MST --- A snippet from /var/log/NetworkManager that seems to be relevant. Observe the roaming at 16:36:40 and back at 16:36:46. Nov 4 16:31:56 rana NetworkManager: <info> Activation (eth1) Stage 5 of 5 (IP Configure Commit) complete. Nov 4 16:36:37 rana NetworkManager: <info> (eth1): supplicant connection state change: 7 -> 3 Nov 4 16:36:38 rana NetworkManager: <info> (eth1): supplicant connection state change: 3 -> 0 Nov 4 16:36:40 rana NetworkManager: <debug> [1225813000.936278] periodic_update(): Roamed from BSSID 00:0B:86:C2:13:82 (Novell) to (none) ((none)) Nov 4 16:36:42 rana NetworkManager: <info> (eth1): supplicant connection state change: 0 -> 3 Nov 4 16:36:43 rana NetworkManager: <info> (eth1): supplicant connection state change: 3 -> 4 Nov 4 16:36:46 rana NetworkManager: <debug> [1225813006.940557] periodic_update(): Roamed from BSSID (none) ((none)) to 00:0B:86:C2:0B:E2 (Novell) Nov 4 16:36:53 rana NetworkManager: <info> (eth1): device state change: 8 -> 3 Nov 4 16:36:53 rana NetworkManager: <info> (eth1): deactivating device (reason: 11). Nov 4 16:36:53 rana NetworkManager: <info> eth1: canceled DHCP transaction, dhcp client pid 2430 Nov 4 16:36:53 rana NetworkManager: <WARN> check_one_route(): (eth1) error -34 returned from rtnl_route_del(): Sucess Nov 4 16:36:53 rana NetworkManager: <info> Activation (eth1) starting connection 'Auto Novell' Nov 4 16:36:53 rana NetworkManager: <info> (eth1): device state change: 3 -> 4 Nov 4 16:36:53 rana NetworkManager: <info> Activation (eth1) Stage 1 of 5 (Device Prepare) scheduled... Nov 4 16:36:53 rana NetworkManager: <info> (eth1): supplicant connection state change: 4 -> 0 Nov 4 16:36:53 rana NetworkManager: <info> Activation (eth1) Stage 1 of 5 (Device Prepare) started... Nov 4 16:36:53 rana NetworkManager: <info> Activation (eth1) Stage 2 of 5 (Device Configure) scheduled... Nov 4 16:36:53 rana NetworkManager: <info> Activation (eth1) Stage 1 of 5 (Device Prepare) complete. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User gp@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c3 --- Comment #3 from Gerald Pfeifer 2008-11-04 09:04:50 MST --- Good pointer, thank you Christian! If it is indeed this very issue, all the better (though it would be disappointing that we have not addressed this for FACTORY/11.1/SLED 11 in a proactive manner). Mobilistas, I'd be happy to give a test package a try. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 Helmut Schaa changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |gp@novell.com -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c6 Helmut Schaa changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|zoz@novell.com |hschaa@novell.com --- Comment #6 from Helmut Schaa 2008-11-05 23:36:39 MST --- (In reply to comment #5 from Gerald Pfeifer)

I'm afraid this didn't help:

:( Ok, I'll try to reproduce the issue here. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 Helmut Schaa changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P3 - Medium -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User gp@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c9 --- Comment #9 from Gerald Pfeifer 2008-11-27 19:57:59 MST --- Created an attachment (id=256360) --> (https://bugzilla.novell.com/attachment.cgi?id=256360) Also seen in a hotel (unencrypted connection, hopping from good AP to one with bad signal) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 Helmut Schaa changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #256360|application/octet-stream |text/plain mime type| | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User gp@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c11 Gerald Pfeifer changed: What |Removed |Added ---------------------------------------------------------------------------- Severity|Normal |Major Summary|Connection to "Novell" wireless network |ipw2100 constantly breaks connections via |consistenly breaks with ipw2100 |unnecessary roaming --- Comment #11 from Gerald Pfeifer 2008-12-20 20:57:22 MST --- Bumping severity (back) to major, since this stroke yet again, apparently even on a network with a single access point. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c13 --- Comment #13 from Helmut Schaa 2009-01-19 05:40:10 MST --- (In reply to comment #12)

That might qualify for P1, but certainly P2.

Ok, I'll try to allocate at least one day for this issue this week. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c15 --- Comment #15 from Helmut Schaa 2009-01-21 06:40:45 MST --- I was able to track the problem described in comment #14 down to a race in the ipw2100-driver (luckily not the firmware). The circumstances are as follows: 1) The firmware receives assoc response and moved internally to associated state. followed by an interrupt which informs the driver about that change. 2) The driver does _not_ update its internal state to ASSOCIATED in the interrupt handler. Instead it queues a work to be executed later. 3) The firmware receives the identity request from the AP and accepts the frame as it is in associated state. Followed by an interrupt to the driver. 4) Driver gets rx interrupt but drops the frame due to not being in associated state. 5) The queued work is executed and state changes to associated but the AP will not resend the EAP frame as the firmware already acked it. Now thinking about how to fix this issue. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c17 --- Comment #17 from Helmut Schaa 2009-01-23 05:16:19 MST --- No progress here since yesterday, see upstream discussion about details => Not fixable for RC3. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c19 Helmut Schaa changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |gp@novell.com --- Comment #19 from Helmut Schaa 2009-01-29 01:27:59 MST --- One thing we could try instead of fixing the driver is to disable PMK-caching in wpa_supplicant (of course I don't like to disable that for all cards but only for testing on your machine). Just add the following line to your /etc/wpa_supplicant/wpa_supplicant.conf: dot11RSNAConfigPMKLifetime=1 That might fix the problems you have with reconnecting. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User gp@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c21 --- Comment #21 from Gerald Pfeifer 2009-01-30 06:27:11 MST --- I will be able to test this Tuesday or Wednesday at the earliest since I do not have access to an environment where this triggers before then, and will be happy to do so, but if there is someone else who could test, that would speed up things... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c23 --- Comment #23 from Helmut Schaa 2009-02-04 07:20:59 MST --- Just discussed possible workarounds for this bug with Christian and Stefan. This issue could be fixed by disabling PMK-caching as a whole (either from NM or the supplicant). However, this might slow down connection attempts on _all_ drivers (of course only if the environment allows PMK-caching). Hence, we decided to not work around this issue but focus on fixing ipw2100 instead which will seriously take some more time. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=441511 User hschaa@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=441511#c25 --- Comment #25 from Helmut Schaa 2009-02-19 10:38:07 MST --- Just a quick status update: I have a quick'n'dirty hack for ipw2100, af_packet and wpa_supplicant ready now. WPA-EAP connections are much more stable with that, PMKSA caching works fine and as a bonus the initial authentication is speeded up by some seconds. However, the patch is not in an acceptable state for both upstream and SLED11 yet, it's just a quick hack and needs some more work. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.