kernel panic while using compat-wireless-2.6.38.2-2

Hello linux-wireless list!

I am using a Sparklan WPEA-166G/W with Chipset AR5BXB63 Rev. 11 (it's a MiniPCI-card) in an embedded X86-System (Pipal 2140U, Intel ATOM processor). It's running voyage-0.6.5, which essentially is a Debian Lenny System.

I am running kernel 2.6.33.9-rt31:

demo00000001:~# uname -a
Linux SSB_wlan-demo00000001 2.6.33.9-rt31-atom-ssb #1 PREEMPT RT Mon May 9 18:25:28 CEST 2011 i686 GNU/Linux

I installed the compat-wireless drivers version 2.6.38.2-2 ( http://www.orbit-lab.org/kernel/compat-wireless-2.6-stable/v2.6.38/compat-wireless-2.6.38.2-2.tar.bz2 ) and installed everything I need according to this webpage: http://acx100.erley.org/stable.html

Once issueing modprobe ath5k, putting the interface up and waiting for ~30sec, I get a kernel panic. Could this be a bug or am I doing something wrong?

The error only occurrs after putting the interface up. If I don't do so, the error won't be there.

Here is some output from the machine:

demo00000001:~# modprobe ath5k && ifconfig wlan0 up
[   48.363894] Compat-wireless backport release: compat-wireless-2.6.38.2-2
[   48.372351] Backport based on linux-2.6-allstable.git v2.6.38.2
[   48.411748] cfg80211: Calling CRDA to update world regulatory domain
[   48.517969] ath5k 0000:09:00.0: using bridge 0000:00:1c.2 INT A to get IRQ 17
[   48.526919] ath5k 0000:09:00.0: PCI->APIC IRQ transform: INT A -> IRQ 17
[   48.534315] ath5k 0000:09:00.0: registered as 'phy0'
[   49.162919] ath5k phy0: Atheros AR2425 chip found (MAC: 0xe2, PHY: 0x70)
demo00000001:~# [   77.953752] ------------[ cut here ]------------
[   77.958711] WARNING: at /home/dborgmann/downloads/compat-wireless-2.6.38.2-2/net/mac80211/rx.c:2827 ieee80211_rx+0x38/0x6fe [mac80211]()
[   77.971856] Hardware name: Luna Pier CRB
[   77.976049] Modules linked in: [last unloaded: scsi_wait_scan]
[   77.982354] Pid: 9, comm: sirq-tasklet/0 Not tainted 2.6.33.9-rt31-atom-ssb #1
[   77.990090] Call Trace:
[   77.992732]  [<c011852f>] warn_slowpath_common+0x65/0x7c
[   77.998419]  [<e0024ad1>] ? ieee80211_rx+0x38/0x6fe [mac80211]
[   78.004672]  [<c0118553>] warn_slowpath_null+0xd/0x10
[   78.010119]  [<e0024ad1>] ieee80211_rx+0x38/0x6fe [mac80211]
[   78.016157]  [<c022d6e8>] ? __alloc_skb+0x4c/0x110
[   78.021296]  [<c0106297>] ? nommu_map_page+0x3c/0x51
[   78.026648]  [<e00992c9>] ? ath5k_rx_skb_alloc+0x99/0x10a [ath5k]
[   78.033190]  [<e009b989>] ath5k_tasklet_rx+0x59f/0x60b [ath5k]
[   78.039454]  [<c011c486>] __tasklet_action.clone.8+0x76/0xd1
[   78.045502]  [<c011c529>] tasklet_action+0x23/0x25
[   78.050640]  [<c011c285>] run_ksoftirqd+0xf0/0x19b
[   78.055788]  [<c011c195>] ? run_ksoftirqd+0x0/0x19b
[   78.061032]  [<c0127829>] kthread+0x5e/0x63
[   78.065538]  [<c01277cb>] ? kthread+0x0/0x63
[   78.070107]  [<c0102bb6>] kernel_thread_helper+0x6/0x10
[   78.075695] ---[ end trace 6b864a3c0aa1a8eb ]---

Best regards,
Dennis

Hello linux-wireless list!

Is there really no one with an idea? Is there no one among the developers, who is interested in this bug, if it is a bug?

I'm starting to lose my hope...

Dennis

Am 11.05.2011 12:38, schrieb Dennis Borgmann:

Can you reproduce this in the wireless-testing tree?

Ben

Ben Greear <[email protected]>
Candela Technologies Inc  http://www.candelatech.com

Problem has been solved by using latest compat-wireless-2.6.39-1 from last night.

Finally.

Thank you, linux-wireless folks!

Best regards,
Dennis

Am 23.05.2011 18:35, schrieb Ben Greear:

Dear linux-wireless list,

sorry, I did answer too early, problem still exists. I am now using compat-wireless-2.6.39-1. I hope, this is new enough (it seems to be from last night anyway...)

I now even exchanged the wireless card, now I am using a Gigabyte GN-WI01HT (AR5414 chip), the problem seems to be the same. I attached some dmesg info. Could this be a hardware error of the mainboard? Actually, I only have one hardware to test on and I somehow don't really trust it. Just a thought of mine...

What are your suggestions, or could this really be a bug?

Best regards,
Dennis

[    7.525432] Compat-wireless backport release: compat-wireless-v2.6.39-1
[    7.532737] Backport based on linux-2.6-allstable.git v2.6.39
[    7.576502] cfg80211: Calling CRDA to update world regulatory domain
[    7.640147] cfg80211: World regulatory domain updated:
[    7.645908] cfg80211:     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
[    7.656157] cfg80211:     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    7.664483] cfg80211:     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
[    7.672754] cfg80211:     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
[    7.674546] ath5k 0000:11:07.0: PCI->APIC IRQ transform: INT A -> IRQ 19
[    7.688710] cfg80211:     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    7.697046] cfg80211:     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    7.705453] ath5k 0000:11:07.0: registered as 'phy0'
[    8.351679] ath: EEPROM regdomain: 0x0
[    8.351685] ath: EEPROM indicates default country code should be used
[    8.351690] ath: doing EEPROM country->regdmn map search
[    8.351695] ath: country maps to regdmn code: 0x3a
[    8.351699] ath: Country alpha2 being used: US
[    8.351703] ath: Regpair used: 0x3a
[    8.360800] ieee80211 phy0: Selected rate control algorithm 'minstrel_ht'
[    8.362487] cfg80211: Calling CRDA for country: US
[    8.362536] ath5k phy0: Atheros AR5414 chip found (MAC: 0xa5, PHY: 0x61)
[    8.363657] modprobe used greatest stack depth: 5860 bytes left
[    8.388493] cfg80211: Regulatory domain changed to country: US
[    8.394937] cfg80211:     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
[    8.403724] cfg80211:     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2700 mBm)
[    8.412246] cfg80211:     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 1700 mBm)
[    8.421561] cfg80211:     (5250000 KHz - 5330000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    8.430451] cfg80211:     (5490000 KHz - 5600000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    8.438853] cfg80211:     (5650000 KHz - 5710000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[    8.447623] cfg80211:     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 3000 mBm)
[    8.496322] scsi 4:0:0:0: Direct-Access     Generic  Storage Device   0.00 PQ: 0 ANSI: 2
[    8.505354] sd 4:0:0:0: Attached scsi generic sg1 type 0
[    8.511989] sd 4:0:0:0: [sdb] Attached SCSI removable disk
[    8.796704] chown used greatest stack depth: 5824 bytes left
[   27.909475] cfg80211: Calling CRDA for country: DE
[   27.920079] cfg80211: Regulatory domain changed to country: DE
[   27.927544] cfg80211:     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
[   27.937610] cfg80211:     (2400000 KHz - 2483500 KHz @ 40000 KHz), (N/A, 2000 mBm)
[   27.945562] cfg80211:     (5150000 KHz - 5250000 KHz @ 40000 KHz), (N/A, 2000 mBm)
[   27.953475] cfg80211:     (5250000 KHz - 5350000 KHz @ 40000 KHz), (N/A, 2000 mBm)
[   27.961475] cfg80211:     (5470000 KHz - 5725000 KHz @ 40000 KHz), (N/A, 2698 mBm)
[   28.014622] hostapd used greatest stack depth: 5748 bytes left
[   33.195098] ------------[ cut here ]------------
[   33.201175] WARNING: at /mnt/entw/compat-wireless-2.6.39-1/net/mac80211/rx.c:2882 ieee80211_rx+0x39/0x660 [mac80211]()
[   33.213802] Hardware name: Luna Pier CRB
[   33.219331] Modules linked in: ath5k ath mac80211 cfg80211 compat [last unloaded: scsi_wait_scan]
[   33.226899] Pid: 24, comm: sirq-tasklet/1 Not tainted 2.6.33.7.2-rt30-atom-ssb #1
[   33.236137] Call Trace:
[   33.240118]  [<c1030666>] warn_slowpath_common+0x60/0x90
[   33.247071]  [<c10306a3>] warn_slowpath_null+0xd/0x10
[   33.253762]  [<dff60edd>] ieee80211_rx+0x39/0x660 [mac80211]
[   33.261111]  [<e031064c>] ath5k_tasklet_rx+0x540/0x5aa [ath5k]
[   33.268566]  [<c102b281>] ? set_next_entity+0x1e/0x89
[   33.273935]  [<c135941d>] ? __schedule+0x76f/0x7a4
[   33.280319]  [<c10356b4>] __tasklet_action+0x77/0xd7
[   33.285564]  [<c10357d2>] tasklet_action+0x5c/0x62
[   33.291956]  [<c1035c19>] run_ksoftirqd+0x131/0x247
[   33.297091]  [<c1035ae8>] ? run_ksoftirqd+0x0/0x247
[   33.303589]  [<c1046ab6>] kthread+0x61/0x66
[   33.308029]  [<c1046a55>] ? kthread+0x0/0x66
[   33.313884]  [<c1002db6>] kernel_thread_helper+0x6/0x10
[   33.319430] ---[ end trace 95aeede189daca6f ]---

Am 23.05.2011 18:35, schrieb Ben Greear:

I use the wireless-testing tree, not any compat stuff. We are testing ath5k and ath9k currently, and no troubles.

If you can reproduce it using wireless-testing, I'll look at it closer. Otherwise, it might be some problem with the backport.

Thanks,
Ben

Ben Greear <[email protected]>
Candela Technologies Inc  http://www.candelatech.com