Bug 1964 - Iwlwifi causes microcode error and OOPs when associating
: Iwlwifi causes microcode error and OOPs when associating
Status: VERIFIED FIXED
: iwlwifi
BSS
: official kernel (2.6.*)
: 4965 (Intel(R) WiFi Link 4965) Fedora 10
: P2 enhancement
Assigned To:
:
:
:
:
:
  Show dependency treegraph
 
Reported: 2009-04-11 06:38 by
Modified: 2009-05-30 19:27 (History)


Attachments
Better formatted dmesg output (5.66 KB, text/plain)
2009-04-11 06:39, Ben Gamari
Details
DMA fixes (17.96 KB, patch)
2009-04-13 10:25, reinette chatre
Details | Diff
small debug patch (2.21 KB, patch)
2009-04-17 15:31, reinette chatre
Details | Diff
DMA fixes (18.56 KB, patch)
2009-04-17 15:32, reinette chatre
Details | Diff


Note

You need to log in before you can comment on or make changes to this bug.


Description From 2009-04-11 06:38:10
Recently (since 2.9.29 or so) NetworkManager induces a Microcode Error when
attempting to associate. Shortly thereafter, it causes an oops which generally
causes the rest of the machine to lock up. 



Apr  8 14:34:34 mercury NetworkManager: <info>  Wireless now enabled by radio
killswitch
Apr  8 14:34:34 mercury NetworkManager: <info>  (wlan0): bringing up device.
Apr  8 14:34:34 mercury kernel: Bluetooth: RFCOMM socket layer initialized
Apr  8 14:34:34 mercury kernel: Bluetooth: RFCOMM TTY layer initialized
Apr  8 14:34:34 mercury kernel: Bluetooth: RFCOMM ver 1.11
Apr  8 14:34:34 mercury bluetoothd[3126]: Adapter /org/bluez/3126/hci0 has been
enabled
Apr  8 14:34:34 mercury kernel: Registered led device: iwl-phy0::radio
Apr  8 14:34:34 mercury kernel: Registered led device: iwl-phy0::assoc
Apr  8 14:34:34 mercury kernel: Registered led device: iwl-phy0::RX
Apr  8 14:34:34 mercury kernel: Registered led device: iwl-phy0::TX
Apr  8 14:34:34 mercury NetworkManager: <info>  (wlan0): device state change: 2
-> 3
Apr  8 14:34:34 mercury kernel: ADDRCONF(NETDEV_UP): wlan0: link is not ready
Apr  8 14:34:34 mercury NetworkManager: <info>  (wlan0): supplicant interface
state:  starting -> ready
Apr  8 14:34:35 mercury kernel: iwl3945 0000:0c:00.0: Microcode SW error
detected. Restarting 0x82000008.
Apr  8 14:34:35 mercury kernel: iwl3945 0000:0c:00.0: Error Reply type
0x00000005 cmd REPLY_PHY_CALIBRATION_CMD (0xB0) seq 0x013A ser 0x004E0000
Apr  8 14:34:35 mercury kernel: general protection fault: 0000 [#1] PREEMPT SMP 
Apr  8 14:34:35 mercury kernel: last sysfs file:
/sys/devices/pci0000:00/0000:00:1c.1/0000:0c:00.0/rfkill/rfkill0/state
Apr  8 14:34:35 mercury kernel: CPU 1 
Apr  8 14:34:35 mercury kernel: Modules linked in: rfcomm btusb bridge stp llc
bnep sco l2cap bluetooth coretemp hwmon sunrpc ip6table_filter ip6_tables ipv6
cpufreq_ondemand acpi_cpufreq freq_table dm_multipath uinput arc4 ecb iwl3945
iwlcore snd_hda_codec_idt mac80211 snd_hda_intel snd_hda_codec snd_hwdep
snd_pcm snd_timer snd lib80211 firewire_ohci yenta_socket iTCO_wdt cfg80211 tg3
battery soundcore dell_laptop iTCO_vendor_support video output wmi
rsrc_nonstatic i2c_i801 firewire_core crc_itu_t libphy pcspkr joydev rfkill
dcdbas ac snd_page_alloc i915 drm i2c_algo_bit cfbcopyarea i2c_core cfbimgblt
cfbfillrect [last unloaded: microcode]
Apr  8 14:34:35 mercury kernel: Pid: 2485, comm: iwl3945 Not tainted 2.6.29work
#29 Latitude D830                   
Apr  8 14:34:35 mercury kernel: RIP: 0010:[<ffffffff810a7b4c>] 
[<ffffffff810a7b4c>] put_page+0x16/0x99
Apr  8 14:34:35 mercury kernel: RSP: 0018:ffff880117d09d10  EFLAGS: 00010082
Apr  8 14:34:35 mercury kernel: RAX: ffff8801104c2bc0 RBX: ffff880107c02000
RCX: 0000000000000abe
Apr  8 14:34:35 mercury kernel: RDX: 0000000000000030 RSI: 0000000000000086
RDI: e8c589ef894cffff
Apr  8 14:34:35 mercury kernel: RBP: ffff880117d09d40 R08: ffff880117d09d80
R09: 0000000000000003
Apr  8 14:34:35 mercury kernel: R10: 0000000000000000 R11: 0000002d17da21c8
R12: 0000000000000001
Apr  8 14:34:35 mercury kernel: R13: ffff880117da5c38 R14: ffff880117da21e0
R15: 0000000000000000
Apr  8 14:34:35 mercury kernel: FS:  0000000000000000(0000)
GS:ffff88002803a000(0000) knlGS:0000000000000000
Apr  8 14:34:35 mercury kernel: CS:  0010 DS: 0018 ES: 0018 CR0:
000000008005003b
Apr  8 14:34:35 mercury kernel: CR2: 00007f550652fe60 CR3: 0000000000201000
CR4: 00000000000006e0
Apr  8 14:34:35 mercury kernel: DR0: 0000000000000000 DR1: 0000000000000000
DR2: 0000000000000000
Apr  8 14:34:35 mercury kernel: DR3: 0000000000000000 DR6: 00000000ffff0ff0
DR7: 0000000000000400
Apr  8 14:34:35 mercury kernel: Process iwl3945 (pid: 2485, threadinfo
ffff880117d08000, task ffff88011d5245c0)
Apr  8 14:34:35 mercury kernel: Stack:
Apr  8 14:34:35 mercury kernel: 0000000000000abe ffff880107c02000
0000000000000001 ffff880117da5c38
Apr  8 14:34:35 mercury kernel: ffff880117da21e0 0000000000000000
ffff880117d09d60 ffffffff8133da83
Apr  8 14:34:35 mercury kernel: ffff880107c02000 ffff880117da10e0
ffff880117d09d80 ffffffff8133d603
Apr  8 14:34:35 mercury kernel: Call Trace:
Apr  8 14:34:35 mercury kernel: [<ffffffff8133da83>] skb_release_data+0x77/0xce
Apr  8 14:34:35 mercury kernel: [<ffffffff8133d603>] __kfree_skb+0x1e/0x81
Apr  8 14:34:35 mercury kernel: [<ffffffff8133d693>] consume_skb+0x2d/0x2f
Apr  8 14:34:35 mercury kernel: [<ffffffffa01eca5b>]
iwl_rx_queue_reset+0x97/0x104 [iwlcore]
Apr  8 14:34:35 mercury kernel: [<ffffffffa021d6f3>]
iwl3945_hw_nic_init+0x16b/0x4bb [iwl3945]
Apr  8 14:34:35 mercury kernel: [<ffffffffa0216d30>] __iwl3945_up+0x157/0x2d0
[iwl3945]
Apr  8 14:34:35 mercury kernel: [<ffffffffa0216edc>] iwl3945_bg_up+0x33/0x48
[iwl3945]
Apr  8 14:34:35 mercury kernel: [<ffffffffa0216ea9>] ? iwl3945_bg_up+0x0/0x48
[iwl3945]
Apr  8 14:34:35 mercury kernel: [<ffffffff81059d45>] run_workqueue+0x94/0x131
Apr  8 14:34:35 mercury kernel: [<ffffffff81059ece>] worker_thread+0xec/0xfd
Apr  8 14:34:35 mercury kernel: [<ffffffff8105dac7>] ?
autoremove_wake_function+0x0/
Apr  8 14:34:35 mercury kernel: [<ffffffff81059de2>] ? worker_thread+0x0/0xfd
Apr  8 14:34:35 mercury kernel: [<ffffffff8105d6c6>] kthread+0x4d/0x78
Apr  8 14:34:35 mercury kernel: [<ffffffff81011e6a>] child_rip+0xa/0x20
Apr  8 14:34:35 mercury kernel: [<ffffffff81011801>] ? restore_args+0x0/0x30
Apr  8 14:34:35 mercury kernel: [<ffffffff8105d679>] ? kthread+0x0/0x78
Apr  8 14:34:35 mercury kernel: [<ffffffff81011e60>] ? child_rip+0x0/0x20
Apr  8 14:34:35 mercury kernel: Code: 89 e5 0f 1f 44 00 00 48 c7 c7 91 78 0a 81
e8 e8 2c fb ff c9 c3 55 48 89 e5 41 57 41 56 41 55 41 54 53 48 83 ec 08 0f 1f
44 00 00 <48> f7 07 00 60 00 00 48 89 fb 74 07 e8 0b f4 ff ff eb 64 f0 ff 
Apr  8 14:34:35 mercury kernel: RIP  [<ffffffff810a7b4c>] put_page+0x16/0x99
Apr  8 14:34:35 mercury kernel: RSP <ffff880117d09d10>
Apr  8 14:34:35 mercury kernel: ---[ end trace c75efee48ea390a2 ]---
Apr  8 14:34:35 mercury kernel: note: iwl3945[2485] exited with preempt_count 1
------- Comment #1 From 2009-04-11 06:39:04 -------
Created an attachment (id=1936) [details]
Better formatted dmesg output
------- Comment #2 From 2009-04-12 21:26:16 -------
Could you please recompile driver with CONFIG_IWLWIFI_DEBUG?

CONFIG_IWLWIFI_DEBUG=y

You then can reload driver with "modprobe iwlagn debug=0x43fff". Please attach
dmesg log after you see the connection stall.

Please see http://www.intellinuxwireless.org/?n=fw_error_report for detail
uCode error report.
------- Comment #3 From 2009-04-13 10:25:41 -------
Created an attachment (id=1940) [details]
DMA fixes

Could you please try the attached patch? 
------- Comment #4 From 2009-04-13 21:38:20 -------
(In reply to comment #3)
> Created an attachment (id=1940) [details] [details]
> DMA fixes
> 
> Could you please try the attached patch? 
> 

As it turns out, this seems to fix it. When will this be going into Linus's
master?
------- Comment #5 From 2009-04-14 08:50:22 -------
(In reply to comment #4)
> (In reply to comment #3)
> > Created an attachment (id=1940) [details] [details] [details]
> > DMA fixes
> > 
> > Could you please try the attached patch? 
> > 
> 
> As it turns out, this seems to fix it. When will this be going into Linus's
> master?
> 

That is the goal but unfortunately this patch breaks fragmentation on the 5100
device. We cannot send it upstream until this has been resolved. There thus
seems to still be problems with this patch. Could you please run with it for a
while to help us catch the issues?
------- Comment #6 From 2009-04-14 09:48:57 -------
(In reply to comment #5)
> That is the goal but unfortunately this patch breaks fragmentation on the 5100
> device. We cannot send it upstream until this has been resolved. There thus
> seems to still be problems with this patch. Could you please run with it for a
> while to help us catch the issues?
> 
Absolutely, I've been running it ever since my last message. No problems so
far. Let me know when there's another patch to be tested.
------- Comment #7 From 2009-04-17 15:31:10 -------
Created an attachment (id=1945) [details]
small debug patch

debug patch on which new DMA fix patch depends
------- Comment #8 From 2009-04-17 15:32:06 -------
Created an attachment (id=1946) [details]
DMA fixes

Could you please test this new version of the DMA fixes patch? It does depend
on a small debug patch that is also attached here.

Thanks
------- Comment #9 From 2009-05-28 15:23:56 -------
This patch is upstream now
------- Comment #10 From 2009-05-30 19:27:29 -------
Mark it as Verified.