Bug 1988 - SYSASSERT (#05) 4001178014 0x41408035 0x00008904 1353
: SYSASSERT (#05) 4001178014 0x41408035 0x00008904 1353
Status: VERIFIED DUPLICATE of bug 1853
: iwlwifi
firmware error
: official kernel (2.6.*)
: 5100/5300 (Intel(R) WiFi Link 5100/5300) Ubuntu
: P2 blocker
Assigned To:
:
:
:
:
:
  Show dependency treegraph
 
Reported: 2009-05-04 14:56 by
Modified: 2009-08-02 09:59 (History)


Attachments
dmesg (123.95 KB, text/plain)
2009-05-04 14:56, Jarkko Lietolahti
Details
dmesg of debug50=0x1 & sudo ip link set wlan2 up (23.52 KB, text/plain)
2009-05-05 15:37, Jarkko Lietolahti
Details
dmesg with AP set to channel 13. (33.58 KB, text/plain)
2009-05-05 15:47, Jarkko Lietolahti
Details
dmesg of sudo modprobe iwlagn debug50=0x1 11n_disable50=1 (29.19 KB, text/plain)
2009-05-05 15:50, Jarkko Lietolahti
Details
dmesg AP ch13 & 20Mhz 11n_disable=0 and working internet connection (34.34 KB, text/plain)
2009-05-11 11:54, Jarkko Lietolahti
Details


Note

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


Description From 2009-05-04 14:56:09
Created an attachment (id=1979) [details]
dmesg

Finally I had all the required pieces and repeatable microcode sw error case so
that I managed to capture the event log for further debugging.

Today iwlagn didn't want to work without loading the module with option
11n_disable50=1, instead it crashed with microcode sw everytime. So I
recompiled the 2.6.30 ubuntu kernel with CONFIG_IWLWIFI_DEBUG=y and loaded the
module with debug50=0x43fff. 

[ 7209.285089] iwlagn 0000:06:00.0: Microcode SW error detected.  Restarting
0x2000000.
[ 7209.285108] iwlagn 0000:06:00.0: Start IWL Error Log Dump:
[ 7209.285114] iwlagn 0000:06:00.0: Status: 0x000212E4, count: 5
[ 7209.285154] iwlagn 0000:06:00.0: Desc                               Time    
  data1      data2      line
[ 7209.285163] iwlagn 0000:06:00.0: SYSASSERT                    (#05)
4001178014 0x41408035 0x00008904 1353
[ 7209.285169] iwlagn 0000:06:00.0: blink1  blink2  ilink1  ilink2
[ 7209.285175] iwlagn 0000:06:00.0: 0x03768 0x03768 0x0090E 0x00000
[ 7209.285205] iwlagn 0000:06:00.0: Start IWL Event Log Dump: display count
512, wraps 11
[ 7209.285224] iwlagn 0000:06:00.0: EVT_LOGT:0013642890:0x0000010d:0106
[ 7209.285241] iwlagn 0000:06:00.0: EVT_LOGT:0013643274:0x00000000:0356
[ 7209.285259] iwlagn 0000:06:00.0: EVT_LOGT:0013645479:0x0000010d:0106
[ 7209.285276] iwlagn 0000:06:00.0: EVT_LOGT:0013645863:0x00000000:0356
[ 7209.285293] iwlagn 0000:06:00.0: EVT_LOGT:0013647134:0x0000010d:0106


Full dmesg included as attachment
------- Comment #1 From 2009-05-04 15:24:13 -------
So this seems to be related to using Channel 13 which somehow doesn't work all
the way. 
I found this https://bugzilla.redhat.com/show_bug.cgi?id=494814 and after
reading it and changing the AP to use channel 10 (it was using autochannel
which had selected ch 13) I could load the iwlagn module without
11n_disable50=1  and the connection is working better.
------- Comment #2 From 2009-05-04 16:58:50 -------
Sorry, assign to Reinette.
------- Comment #3 From 2009-05-05 14:43:37 -------
Could you please load module with debug50=0x1 and send output of dmesg after
you have run "ip link set wlan0 up" ?
------- Comment #4 From 2009-05-05 15:37:42 -------
Created an attachment (id=1982) [details]
dmesg of debug50=0x1 &  sudo ip link set wlan2 up

Note: AP is configured NOT to use channel 13
------- Comment #5 From 2009-05-05 15:47:52 -------
Created an attachment (id=1983) [details]
dmesg with AP set to channel 13.

dmesg with AP set to channel 13. iwlagn is associated but the connection
doesn't work e.g. nothing gets through.
------- Comment #6 From 2009-05-05 15:50:20 -------
Created an attachment (id=1984) [details]
dmesg of sudo modprobe iwlagn debug50=0x1 11n_disable50=1

AP set to channel 13, 11n_disable50=1, iwlagn is associated and the connection
works normally. With 11n_disable50=0 it would not work.
------- Comment #7 From 2009-05-10 21:57:52 -------
(In reply to comment #6)
> Created an attachment (id=1984) [details] [details]
> dmesg of sudo modprobe iwlagn debug50=0x1 11n_disable50=1
> 
> AP set to channel 13, 11n_disable50=1, iwlagn is associated and the connection
> works normally. With 11n_disable50=0 it would not work.
> 

Could you please try this experiment:

- set AP to use channel 13
- force your AP to use 20Mhz
- load driver with 11n_disable=0

Thank you
------- Comment #8 From 2009-05-11 04:11:28 -------
In all the previous tests the AP was set to Auto 20/40Mhz. 
------- Comment #9 From 2009-05-11 08:12:18 -------
(In reply to comment #8)
> In all the previous tests the AP was set to Auto 20/40Mhz. 
> 

Do things work with 11n if you force the AP to use 20Mhz only?
------- Comment #10 From 2009-05-11 11:54:23 -------
Created an attachment (id=1989) [details]
dmest AP ch13 & 20Mhz 11n_disable=0 and working internet connection

Yes, the connection seems to work when AP set to channel 13 & 20Mhz, iwlagn
loaded with sudo modprobe  iwlagn 11n_disable=0 debug50=0x1

Connection works OK.
------- Comment #11 From 2009-05-11 20:59:28 -------
Wey-Yi,

Could you please take a look at this problem?

When user tries to use 40Mhz on channel 13 then the firmware SYSASSERTS. This
channel is marked as WIDE and works fine with 20Mhz. Is it perhaps trying to
transmit in channel 14 which is not allowed?

Thank you
------- Comment #12 From 2009-05-12 07:03:29 -------
(In reply to comment #11)
> Wey-Yi,
> Could you please take a look at this problem?
> When user tries to use 40Mhz on channel 13 then the firmware SYSASSERTS. This
> channel is marked as WIDE and works fine with 20Mhz. Is it perhaps trying to
> transmit in channel 14 which is not allowed?
> Thank you

Yes, I will try to re-pro this problem here
------- Comment #13 From 2009-05-21 07:17:46 -------
(In reply to comment #11)
> Wey-Yi,
> Could you please take a look at this problem?
> When user tries to use 40Mhz on channel 13 then the firmware SYSASSERTS. This
> channel is marked as WIDE and works fine with 20Mhz. Is it perhaps trying to
> transmit in channel 14 which is not allowed?
> Thank you

Looking at the dmesg, appear to me this is "pure 40 MHz" problem which already
has the patch submitted for the fix.

Question on why using channel 13? there are only 3 non-overlapped channel in
"g" mode which are 1, 6, and 11. Using channel 13 is not recommended 

------- Comment #14 From 2009-05-21 13:30:33 -------
> Question on why using channel 13? there are only 3 non-overlapped channel in
> "g" mode which are 1, 6, and 11. Using channel 13 is not recommended 

AP had auto channel selection and for some reason it decided to use the channel
13.
------- Comment #15 From 2009-06-08 09:20:54 -------
The patch mentioned in comment #13 is now upstream. Could you please test with
latest code?
------- Comment #16 From 2009-06-08 09:21:21 -------

*** This bug has been marked as a duplicate of bug 1853 ***
------- Comment #17 From 2009-08-02 09:59:59 -------
The new firmware 8.24.2.12 fixes my problem. Thank you.

grep iwlagn
[   16.579725] iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux,
1.3.27k
[   16.579727] iwlagn: Copyright(c) 2003-2009 Intel Corporation
[   16.579877] iwlagn 0000:06:00.0: PCI INT A -> GSI 19 (level, low) -> IRQ 19
[   16.579906] iwlagn 0000:06:00.0: setting latency timer to 64
[   16.579957] iwlagn 0000:06:00.0: Detected Intel Wireless WiFi Link 5100AGN
REV=0x54
[   16.616763] iwlagn 0000:06:00.0: Tunable channels: 13 802.11bg, 24 802.11a
channels
[   16.616833] iwlagn 0000:06:00.0: irq 31 for MSI/MSI-X
[   59.965227] iwlagn 0000:06:00.0: firmware: requesting iwlwifi-5000-2.ucode
[   60.021998] iwlagn 0000:06:00.0: loaded firmware version 8.24.2.12
[  215.931914] iwlagn 0000:06:00.0: iwl_tx_agg_start on ra = 00:21:91:ea:90:1f
tid = 0
[ 2475.968286] iwlagn 0000:06:00.0: iwl_tx_agg_start on ra = 00:21:91:ea:90:1f
tid = 6