Created attachment 25600 [details] dmesg of x61s thinkpad to regain connection need to reconnect with nm-applet: message on connection lost: [20396.771335] iwlagn 0000:03:00.0: iwl_tx_agg_start on ra = 00:25:84:02:e3:39 tid = 0 funny thing is that connections that where already alive still work for some seconds but then die off. no new ping nothing. need to check if reproducible with wpa_supplicant and dhclient. 2.6.32.10 is also affected. eventually both last stable release seem to have made the situation worth as now it happens much more often. lspci -vv -s 03:00.0 03:00.0 Network controller: Intel Corporation PRO/Wireless 4965 AG or AGN Network Connection (rev 61) Subsystem: Intel Corporation Device 1111 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx+ Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx- Latency: 0, Cache Line Size: 64 bytes Interrupt: pin A routed to IRQ 29 Region 0: Memory at f7f00000 (64-bit, non-prefetchable) [size=8K] Capabilities: <access denied> Kernel driver in use: iwlagn
Please try creating a file calle /etc/modprobe.d/iwlagn.conf with a line in it like this: options iwlagn 11n_disable=1 Does that help?
will test out, thanks.
I also encountered that bug on a 2.6.32.10, I would be disconnected after 2 or 3 minutes, but your workaround works. For the record: # lspci -vv -s 05:00.0 05:00.0 Network controller: Intel Corporation PRO/Wireless 4965 AG or AGN [Kedron] Network Connection (rev 61) Subsystem: Intel Corporation Device 1101 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx- Latency: 0, Cache Line Size: 64 bytes Interrupt: pin A routed to IRQ 29 Region 0: Memory at c4800000 (64-bit, non-prefetchable) [size=8K] Kernel driver in use: iwlagn Kernel modules: iwlagn
I see the same problem with dropping connections with an Intel 5300: 03:00.0 Network controller: Intel Corporation Wireless WiFi Link 5300 Subsystem: Intel Corporation Device 1011 Physical Slot: 1 Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx- Latency: 0, Cache Line Size: 64 bytes Interrupt: pin A routed to IRQ 33 Region 0: Memory at f2500000 (64-bit, non-prefetchable) [size=8K] Capabilities: <access denied> Kernel driver in use: iwlagn Kernel modules: iwlagn
Turning on 11n_disable does not help stability for me. I still get dropped after some period of time ranging between a few seconds and a few hours.
Created attachment 25717 [details] syslog messages pertaining to the failure
(In reply to comment #0) > Created an attachment (id=25600) [details] > dmesg of x61s thinkpad > (I believe the 4965 and 5300 issues are different, this update is for 5300). This log contains a significant of information related to wireless connections being created and stopped, often as requested by userspace. There even are some IBSS connections. Could you please trim the information so that it can be clear what and where the problem is? How easy is this to reproduce? Would it be possible to have debugging running while you reproduce this? Your driver is not compiled with debug support, so please do so first (CONFIG_IWLWIFI_DEBUG) and then load the module with "debug=0x4043fff".
(In reply to comment #5) > Turning on 11n_disable does not help stability for me. I still get dropped > after some period of time ranging between a few seconds and a few hours. (This update is for 5300 issue only) Could you please try out the patches in http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2037#c113 ?
setting "options iwlagn 11n_disable=1" fixed the bug sorry that i was slow in reporting bug, but sometimes it been hard to reproduce it. other days you would hit it continuously: ls -l /etc/modprobe.d/iwlagn.conf -rw-r--r-- 1 root root 29 Mar 25 17:10 /etc/modprobe.d/iwlagn.conf not reproduced since on the 4965, thanks. are the relevant patch for this in 2.6.32 stable queue?
(In reply to comment #9) > are the relevant patch for this in 2.6.32 stable queue? The option you enabled turned off all 11n support ... this is probably not a patch that you want in stable. Any chance in getting the information requested in comment #7?
it is not easy to reproduce no. some times it happens not at all, booted 2.6.32.10-rc1 with iwlwifi debug and will tell next days.
Created attachment 25779 [details] 2.6.32.10-rc1 dmesg of x61s with iwlwifi debug after boot
was unable to reproduce with debug on. maybe 2.6.32.11 had the fix, will retry next week without debug on that. thanks
fixed with 2.6.32.11 + "iwlwifi: counting number of tfds can be free for 4965" thanks
With my 5300 chip, Lenovo X200, 2.6.33.1 has a massive regression. Instead of the wifi link dropping, and forcing an rmmod/modprobe cycle, the machine now crashes hard and leaves the wifi status LED blinking rapidly. I have to hold down the power switch for 5 seconds to reboot.
Bryan, please open a new bug for the issue you describe in comment 15.
you are seeing a seperate bug, please report as such. 2.6.33.2 is out test against it before.
Created attachment 25886 [details] /var/log/messages for Linux 2.6.32.11 + 4965agn with debug=0x4043fff Sorry for replying late but I didn't have time to test before, the bug is *not* fixed for the 2.6.32.11 on my 4965agn, however for this kernel version the bug occurs with debug=0x4043fff (it would not occur on a 2.6.32.10 with that parameter). I didn't trim the log, but since the bug occurs within a few seconds after connection I think it's small enough to be usable. To post this message I'm using a 2.6.32.11 with 11n_disable=1, so far so good.
reread the closure message, you need 2.6.32.11 plus a patch that is queued for 2.6.32.12
Sorry, as I didn't see any patch attached to that bug report, I (mis)understood that it was already in 2.6.32.11