Earliest failing kernel version: 2.6.24.2 Distribution: Slack 12.0.0 Hardware Environment: i686 Intel(R) Pentium(R) 4 CPU 2.80GHz 00:0c.0 Network controller: Broadcom Corporation BCM4306 802.11b/g Wireless LAN Controller (rev 03) rfkill_input 4096 0 b43 97824 0 rfkill 6160 3 rfkill_input,b43 mac80211 101260 1 b43 cfg80211 11016 1 mac80211 led_class 4228 1 b43 input_polldev 4104 1 b43 bcm43xx 409248 0 firmware_class 7680 3 b43,bttv,bcm43xx ieee80211softmac 26496 1 bcm43xx ieee80211 27464 2 bcm43xx,ieee80211softmac ieee80211_crypt 5120 1 ieee80211 ssb 24324 2 b43,ohci_hcd Note the oddball use counts for rfkill_input, b43, rfkill, bcm43xx. These are the shown values with the link UP. Problem Description: After some time wireless dies off, with the following... Mar 12 11:04:15 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 12 11:23:52 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 12 14:01:39 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 12 22:31:53 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 13 14:55:23 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 13 20:51:53 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 14 00:11:01 shit kernel: wlan0: association frame received from 00:17:3f:9c:dd:1a, but not in associate state - ignored Mar 14 04:57:49 shit kernel: wlan0: No ProbeResp from current AP 00:17:3f:9c:dd:1a - assume out of range Mar 14 04:57:51 shit kernel: wlan0: RX deauthentication from 00:17:3f:9c:dd:1a (reason=3) Mar 14 04:57:51 shit kernel: wlan0: deauthenticated Mar 14 04:57:51 shit kernel: wlan0: RX deauthentication from 00:17:3f:9c:dd:1a (reason=3) Mar 14 04:57:51 shit last message repeated 6 times I then issue '/etc/rc.d/rc.inet1 wlan0_restart and it comes back... word has it this is a driver issue. Mar 14 09:22:29 shit kernel: wlan0: Initial auth_alg=0 Mar 14 09:22:29 shit kernel: wlan0: authenticate with AP 00:17:3f:9c:dd:1a Mar 14 09:22:29 shit kernel: wlan0: Initial auth_alg=0 Mar 14 09:22:29 shit kernel: wlan0: authenticate with AP 00:17:3f:9c:dd:1a Mar 14 09:22:29 shit kernel: wlan0: RX authentication from 00:17:3f:9c:dd:1a (alg=0 transaction=2 status=0) Mar 14 09:22:29 shit kernel: wlan0: authenticated Mar 14 09:22:29 shit kernel: wlan0: associate with AP 00:17:3f:9c:dd:1a Mar 14 09:22:29 shit kernel: wlan0: authentication frame received from 00:17:3f:9c:dd:1a, but not in authenticate state - ignored Mar 14 09:22:29 shit kernel: wlan0: RX AssocResp from 00:17:3f:9c:dd:1a (capab=0x401 status=0 aid=1) Mar 14 09:22:29 shit kernel: wlan0: associated Mar 14 09:22:29 shit kernel: wlan0: switched to short barker preamble (BSSID=00:17:3f:9c:dd:1a) Mar 14 09:22:40 shit kernel: wlan0: no IPv6 routers present Will see if the same problem occurs in 2.6.24.3, and will advise.
I'm not that familiar with Slackware, does the "rc.inet wlan0_restart" actually reload modules? Or does it just kill-off dhclient, ifconfig down, then start over?
No dhcp used, as it is using a static ip... :-) in short, to restart, all the script does is down then up the interface. Modules are NOT disposed of, or reloaded. I will provide for you the scripts in question as an attachment.
Created attachment 15263 [details] /etc/rc.d/rc.inet1
Created attachment 15264 [details] /etc/rc.d/rc.inet1.conf
Created attachment 15265 [details] /etc/rc.d/rc.wireless
Created attachment 15266 [details] /etc/rc.d/rc.wireless.conf
And yes, I am really actually running without any encryption, those are the scripts and actual configs. If you need any other information (such as version numbers on various tools/libc/etc) don't hesitate to ask. If you have any additional settings I can try to mitagate the problem, please advise, but also note the problem did not exist for a few revisions in the older bcm drivers. I also do not recall it happening in the kernel just previous to the current one either.
Well, sorry to have gone dark on you. Are you still seeing this problem with 2.6.25?
I have stopped using the device as it was unstable enough for me to use in production, however it will be placed in a crash box fairly soon with Slackware 12.1, and once it is in there, I'll be beating it to death and can relay anything I may notice.
Fedora 9 with 2.6.26.5-45.fc9.i686.PAE, system boots up and wlan0 is configure on startup: wlan0: authenticate with AP 00:15:6d:54:45:e6 wlan0: authenticated wlan0: associate with AP 00:15:6d:54:45:e6 wlan0: RX AssocResp from 00:15:6d:54:45:e6 (capab=0x31 status=0 aid=4) wlan0: associated wlan0: no IPv6 routers present After random time kernel is loosing wireless connection (here was after few days): wlan0: No ProbeResp from current AP 00:15:6d:54:45:e6 - assume out of range # ifdown wlan0 # ifup wlan0 wlan0: authenticate with AP 00:15:6d:54:45:e6 wlan0: authenticate with AP 00:15:6d:54:45:e6 wlan0: authenticate with AP 00:15:6d:54:45:e6 wlan0: authentication with AP 00:15:6d:54:45:e6 timed out phy0 -> rt2x00queue_write_tx_frame: Error - Arrived at non-free entry in the non-full queue 0. Please file bug report to http://rt2x00.serialmonkey.com. I'm unable to re-associate with AP and kernel is freezing in less that a minute.
Module used for my wireless card in my previous post is: # modinfo rt61pci filename: /lib/modules/2.6.26.5-45.fc9.i686.PAE/kernel/drivers/net/wireless/rt2x00/rt61pci.ko license: GPL firmware: rt2661.bin firmware: rt2561s.bin firmware: rt2561.bin description: Ralink RT61 PCI & PCMCIA Wireless LAN driver. version: 2.1.8 author: http://rt2x00.serialmonkey.com srcversion: D2569A91ED8BE041886E5EB alias: pci:v00001814d00000401sv*sd*bc*sc*i* alias: pci:v00001814d00000302sv*sd*bc*sc*i* alias: pci:v00001814d00000301sv*sd*bc*sc*i* depends: rt2x00lib,rt2x00pci,eeprom_93cx6,crc-itu-t vermagic: 2.6.26.5-45.fc9.i686.PAE SMP mod_unload 686 4KSTACKS
I finally have the crashbox set up, and I will begin beating the card to a pulp starting this Monday. Sorry for the huge time delay, life, etc...
My box just lost connection again, any hints?
Mikolaj, why aren't you reporting to buzilla.redhat.com? Are you willing to build and test an upstream kernel? Have you tried 2.6.27?
Thanks John, I think that's good idea. I will follow up with RedHat. https://bugzilla.redhat.com/show_bug.cgi?id=468377