GNOME Bugzilla – Bug 676382
can't join networks in places with MANY same-named ESSID
Last modified: 2012-07-30 20:46:19 UTC
In the library at University of Kansas, there are about 20 wireless devices all named "JAYHAWK" that my Intel Centrino Ultimate 6300 can detect. I'm never able to join any of them using Debian Linux kernel 3.2 or newer (and the iwlwifi module & firmware delivered with that) and network-manager. I have had plenty of connection trouble and folks have told me the problem is in iwlwifi. I've found that kernel module options are helping the iwlwifi to get connected in most places I go. These options are working most places these days: options iwlwifi power_save=0 swcrypto=1 bt_coex_active=0 11n_disable=1 That works most places, except in the library, where the problems are different. In /var/log/kern.log, there is evidence that the network device is constantly disconnecting itself and trying to join up to a different router. Isn't this what the errors mean? May 18 14:18:51 pjlap-124 kernel: [27142.329603] iwlwifi 0000:03:00.0: PCI INT A disabled May 18 14:18:52 pjlap-124 kernel: [27142.848662] NET: Unregistered protocol family 31 May 18 14:19:23 pjlap-124 kernel: [27173.848045] Compat-wireless backport release: compat-wireless-2012-05-04 May 18 14:19:23 pjlap-124 kernel: [27173.848055] Backport based on linux-next.git next-20120510 May 18 14:19:23 pjlap-124 kernel: [27173.860216] cfg80211: Calling CRDA to update world regulatory domain May 18 14:19:23 pjlap-124 kernel: [27173.875120] cfg80211: World regulatory domain updated: May 18 14:19:23 pjlap-124 kernel: [27173.875133] cfg80211: (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp) May 18 14:19:23 pjlap-124 kernel: [27173.875144] cfg80211: (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) May 18 14:19:23 pjlap-124 kernel: [27173.875153] cfg80211: (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) May 18 14:19:23 pjlap-124 kernel: [27173.875162] cfg80211: (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm) May 18 14:19:23 pjlap-124 kernel: [27173.875170] cfg80211: (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) May 18 14:19:23 pjlap-124 kernel: [27173.875178] cfg80211: (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm) May 18 14:19:23 pjlap-124 kernel: [27173.893731] iwlwifi: Intel(R) Wireless WiFi Link AGN driver for Linux, in-tree: May 18 14:19:23 pjlap-124 kernel: [27173.893735] iwlwifi: Copyright(c) 2003-2012 Intel Corporation May 18 14:19:23 pjlap-124 kernel: [27173.893831] iwlwifi 0000:03:00.0: PCI INT A -> Link[LNKB] -> GSI 11 (level, low) -> IRQ 11 May 18 14:19:23 pjlap-124 kernel: [27173.893896] iwlwifi 0000:03:00.0: setting latency timer to 64 May 18 14:19:23 pjlap-124 kernel: [27173.893936] iwlwifi 0000:03:00.0: pci_resource_len = 0x00002000 May 18 14:19:23 pjlap-124 kernel: [27173.893937] iwlwifi 0000:03:00.0: pci_resource_base = ffffc90000668000 May 18 14:19:23 pjlap-124 kernel: [27173.893939] iwlwifi 0000:03:00.0: HW Revision ID = 0x35 May 18 14:19:23 pjlap-124 kernel: [27173.894213] iwlwifi 0000:03:00.0: irq 48 for MSI/MSI-X May 18 14:19:23 pjlap-124 kernel: [27173.906465] iwlwifi 0000:03:00.0: loaded firmware version 9.221.4.1 build 25532 May 18 14:19:23 pjlap-124 kernel: [27173.906783] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUG disabled May 18 14:19:23 pjlap-124 kernel: [27173.906787] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEBUGFS disabled May 18 14:19:23 pjlap-124 kernel: [27173.906789] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TRACING disabled May 18 14:19:23 pjlap-124 kernel: [27173.906792] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_DEVICE_TESTMODE disabled May 18 14:19:23 pjlap-124 kernel: [27173.906794] iwlwifi 0000:03:00.0: CONFIG_IWLWIFI_P2P enabled May 18 14:19:23 pjlap-124 kernel: [27173.906797] iwlwifi 0000:03:00.0: Detected Intel(R) Centrino(R) Ultimate-N 6300 AGN, REV=0x74 May 18 14:19:23 pjlap-124 kernel: [27173.906956] iwlwifi 0000:03:00.0: L1 Enabled; Disabling L0S May 18 14:19:23 pjlap-124 kernel: [27173.917621] iwlwifi 0000:03:00.0: device EEPROM VER=0x43a, CALIB=0x6 May 18 14:19:23 pjlap-124 kernel: [27173.917624] iwlwifi 0000:03:00.0: Device SKU: 0x1F0 May 18 14:19:23 pjlap-124 kernel: [27173.917626] iwlwifi 0000:03:00.0: Valid Tx ant: 0x7, Valid Rx ant: 0x7 May 18 14:19:23 pjlap-124 kernel: [27173.917640] iwlwifi 0000:03:00.0: Tunable channels: 13 802.11bg, 24 802.11a channels May 18 14:19:23 pjlap-124 kernel: [27173.917741] Registered led device: phy0-led May 18 14:19:23 pjlap-124 kernel: [27173.917814] ieee80211 phy0: Selected rate control algorithm 'iwl-agn-rs' May 18 14:19:23 pjlap-124 kernel: [27173.926204] iwlwifi 0000:03:00.0: L1 Enabled; Disabling L0S May 18 14:19:23 pjlap-124 kernel: [27173.926496] iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1 May 18 14:19:23 pjlap-124 kernel: [27174.170481] iwlwifi 0000:03:00.0: L1 Enabled; Disabling L0S May 18 14:19:23 pjlap-124 kernel: [27174.170698] iwlwifi 0000:03:00.0: Radio type=0x0-0x3-0x1 May 18 14:19:23 pjlap-124 kernel: [27174.291693] ADDRCONF(NETDEV_UP): wlan0: link is not ready May 18 14:19:31 pjlap-124 kernel: [27181.747268] wlan0: authenticate with 00:24:97:b6:de:30 May 18 14:19:31 pjlap-124 kernel: [27181.759078] wlan0: direct probe to 00:24:97:b6:de:30 (try 1/3) May 18 14:19:31 pjlap-124 kernel: [27181.959738] wlan0: direct probe to 00:24:97:b6:de:30 (try 2/3) May 18 14:19:31 pjlap-124 kernel: [27182.163677] wlan0: direct probe to 00:24:97:b6:de:30 (try 3/3) May 18 14:19:31 pjlap-124 kernel: [27182.367589] wlan0: authentication with 00:24:97:b6:de:30 timed out ... May 18 14:21:35 pjlap-124 kernel: [27306.003696] wlan0: direct probe to 00:24:97:b6:de:30 (try 1/3) May 18 14:21:35 pjlap-124 kernel: [27306.206614] wlan0: direct probe to 00:24:97:b6:de:30 (try 2/3) May 18 14:21:35 pjlap-124 kernel: [27306.410542] wlan0: direct probe to 00:24:97:b6:de:30 (try 3/3) May 18 14:21:36 pjlap-124 kernel: [27306.614475] wlan0: authentication with 00:24:97:b6:de:30 timed out May 18 14:21:36 pjlap-124 kernel: [27306.623103] iwlwifi 0000:03:00.0: ACTIVATE a non DRIVER active station id 0 addr 00:24:97:b6:de:30 May 18 14:21:43 pjlap-124 kernel: [27313.964754] wlan0: authenticate with 00:24:97:b7:01:d0 May 18 14:21:43 pjlap-124 kernel: [27314.014214] wlan0: direct probe to 00:24:97:b7:01:d0 (try 1/3) May 18 14:21:43 pjlap-124 kernel: [27314.215955] wlan0: direct probe to 00:24:97:b7:01:d0 (try 2/3) May 18 14:21:43 pjlap-124 kernel: [27314.419898] wlan0: direct probe to 00:24:97:b7:01:d0 (try 3/3) May 18 14:21:44 pjlap-124 kernel: [27314.623864] wlan0: authentication with 00:24:97:b7:01:d0 timed out May 18 14:21:44 pjlap-124 kernel: [27314.632528] iwlwifi 0000:03:00.0: ACTIVATE a non DRIVER active station id 0 addr 00:24:97:b7:01:d0 May 18 14:21:51 pjlap-124 kernel: [27322.004962] wlan0: authenticate with 00:24:97:b6:de:30 May 18 14:21:51 pjlap-124 kernel: [27322.053952] wlan0: direct probe to 00:24:97:b6:de:30 (try 1/3) May 18 14:21:51 pjlap-124 kernel: [27322.257341] wlan0: direct probe to 00:24:97:b6:de:30 (try 2/3) May 18 14:21:52 pjlap-124 kernel: [27322.461273] wlan0: direct probe to 00:24:97:b6:de:30 (try 3/3) May 18 14:21:52 pjlap-124 kernel: [27322.665151] wlan0: authentication with 00:24:97:b6:de:30 timed out May 18 14:21:52 pjlap-124 kernel: [27322.665795] iwlwifi 0000:03:00.0: ACTIVATE a non DRIVER active station id 0 addr 00:24:97:b6:de:30 May 18 14:21:59 pjlap-124 kernel: [27330.044421] wlan0: authenticate with 00:24:97:b6:de:30 May 18 14:21:59 pjlap-124 kernel: [27330.048851] wlan0: direct probe to 00:24:97:b6:de:30 (try 1/3) May 18 14:21:59 pjlap-124 kernel: [27330.250697] wlan0: direct probe to 00:24:97:b6:de:30 (try 2/3) May 18 14:22:00 pjlap-124 kernel: [27330.454628] wlan0: direct probe to 00:24:97:b6:de:30 (try 3/3) May 18 14:22:00 pjlap-124 kernel: [27330.658556] wlan0: authentication with 00:24:97:b6:de:30 timed out Yesterday, I was told to remove network-manager and try wicd. Guess what? It allows me to pick one of the many JAYHAWK ESSID's and join it. And when I go to other places, I no longer have to repeatedly load and unload the iwlwifi module to get the thing to connect. wicd lets me pick and, at least after trying 4 different spots where I used to have trouble with network-manager, I don't see the same trouble any more. This makes me conclude that at least a fraction of the connection trouble I've had is in network-manager, not just in iwlwifi or its firmware. Am I off base? PJ $ uname -a Linux pjlap-124 3.2.0-2-amd64 #1 SMP Mon Apr 30 05:20:23 UTC 2012 x86_64 GNU/Linux
What NetworkManager and wpa_supplicant version do you use? $ lspci -nn | grep Network It would be helpful to see debug traces from wpa_supplicant: https://live.gnome.org/NetworkManager/Debugging#wpa_supplicant07 and also logs from NM that should be in /var/log/daemon.log (or something)
Enterprise wifi networks are designed to let you roam between SSIDs of the same name. If you pick one AP, then when you move the laptop, you'll loose connectivity. NetworkManager simply asks wpa_supplicant to find the best AP and connect to it, and if/when you move around it's supposed to roam between networks. What we need to do is figure out why the initial connection is failing, and to do that we need to get the wpa_supplicant debug logs. You can do this by: 1) mv /usr/sbin/wpa_supplicant / 2) killall -TERM wpa_supplicant 3) /wpa_supplicant -dddtu and then attempt a reconnect, and you'll get a lot of spew frmo the terminal you're running wpa_supplicant in. That's good; that's what we want, and what you should attach the to the bug. If it's an authenticated network though, these logs may disclose your username, so feel free to XXXX that out or something if you see it. When you're done testing, you can: mv /wpa_supplicant /usr/sbin/ or back to wherever your distro had the supplicant.
Created attachment 219869 [details] shows failure to join wireless router called Caliope
Sorry, sent attachment before including explanation. I've been running wpa_supplicant in an Emacs eshell so it's pretty easy to package up sections to show you. If that's too much, tell me, I'll collect less. I had an interesting example today. Attachment is one where I was at one place, suspended the computer, and went to another where the wireless is called Caliope. You see in the attached that wpa_supplicant was unable to connect with Caliope, after many tries. Then I killed the wireless with the hardware switch, and re-started it. In the attached log, I typed the word "RESTARTED" so you can see where that happened, and then the connection worked almost right away. Tomorrow I'll be going to a couple of places that have the multiple wireless routers, and I'll collect logs.
Created attachment 219930 [details] fails to join wireless network at restaurant I'm frustrated because I can't reproduce errors all the time, and sometimes things work when I expect they would fail. The JAYHAWK network that I was complaining about when I started this report allowed me to connect today for the first time in a long time. But other routers that I ordinarily join won't let me in. Attachment wpa-ysfailed-20120730.txt.gz is the wpa_supplicant debug output for a restaurant where I can usually join a network. While that is going on, the dmesg output looks like this, over and over [269847.555742] wlan0: authenticate with 0a:90:7f:90:00:8a (try 1) [269847.563669] wlan0: authenticated [269847.564015] wlan0: failed to insert Dummy STA entry for the AP (error -17) [269850.620611] wlan0: deauthenticating from 0a:90:7f:90:00:8a by local choice (reason=2) [269850.670008] wlan0: authenticate with 0a:90:7f:90:00:8a (try 1) [269850.680160] wlan0: authenticated [269850.680322] wlan0: failed to insert Dummy STA entry for the AP (error -17) [269853.847519] wlan0: deauthenticating from 0a:90:7f:90:00:8a by local choice (reason=2) [269853.896844] wlan0: authenticate with 0a:90:7f:90:00:8a (try 1) [269853.903709] wlan0: authenticated [269853.904098] wlan0: failed to insert Dummy STA entry for the AP (error -17) [269862.597552] wlan0: deauthenticating from 0a:90:7f:90:00:8a by local choice (reason=2) [269862.646443] wlan0: authenticate with 0a:90:7f:90:00:8a (try 1) [269862.653555] wlan0: authenticated [269862.653859] wlan0: failed to insert Dummy STA entry for the AP (error -17) [269865.837947] wlan0: deauthenticating from 0a:90:7f:90:00:8a by local choice (reason=2) [269865.886898] wlan0: authenticate with 0a:90:7f:90:00:8a (try 1) [269865.897502] wlan0: authenticated [269865.898131] wlan0: failed to insert Dummy STA entry for the AP (error -17) [269869.031950] wlan0: deauthenticating from 0a:90:7f:90:00:8a by local choice (reason=2) System information. This is a Debian Squeeze system on which I've installed many newer packages from the testing (Wheezy) distribution. I'm guessing now that's changed something... $ uname -a Linux pjlap-124 3.2.0-3-amd64 #1 SMP Thu Jun 28 09:07:26 UTC 2012 x86_64 GNU/Linux Packages: wpasupplicant 1.0-2 network-manager 0.9.4.0-5 linux-image-3.2.0-3-amd64 3.2.21-3
Yeah, so the: [269847.564015] wlan0: failed to insert Dummy STA entry for the AP (error -17) indicates a problem with the driver in that kernel version. That's almost certainly what's causing the repeated failures to connect to the AP in the supplicant: 1343668826.845673: wlan0: Event AUTH (11) received 1343668826.845679: wlan0: SME: Authentication response: peer=0a:90:7f:90:00:8a auth_type=0 status_code=1 1343668826.845681: SME: Authentication response IEs - hexdump(len=0): [NULL] 1343668826.845685: wlan0: SME: Authentication failed (status code 1) 1343668826.845687: Added BSSID 0a:90:7f:90:00:8a into blacklist So about the best we can do is try an upgraded kernel version, or perhaps your distro will backport any fixes they can find. The "dummy STA" stuff got removed from the drivers for the 3.4 kernel, so I'd expect 3.4 to work here. Basically, we can call this a kernel driver/stack bug that neither NM or the supplicant can do much about. NM doesn't try to work around kernel bugs, since then we'd have a codebase littered with workarounds since NM is expected to work on a wide variety of kernel versions. So I'll close for now, if we still get this bug with kernel 3.4 then we can reopen and investigate what's going on.