Bug 303871 - iwl3945 stopped working. firmware problem?
Summary: iwl3945 stopped working. firmware problem?
Keywords:
Status: CLOSED RAWHIDE
Alias: None
Product: Fedora
Classification: Fedora
Component: kernel
Version: rawhide
Hardware: All
OS: Linux
low
high
Target Milestone: ---
Assignee: John W. Linville
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard:
Depends On:
Blocks: F8Blocker
TreeView+ depends on / blocked
 
Reported: 2007-09-24 19:32 UTC by Zack Cerza
Modified: 2007-11-30 22:12 UTC (History)
3 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2007-10-16 18:28:17 UTC
Type: ---
Embargoed:

Attachments (Terms of Use)

Description Zack Cerza 2007-09-24 19:32:05 UTC 
Description of problem:
I updated my rawhide installation on my Thinkpad Z61t today, and iwl3945 seems
to have broken. I'm including portions of /var/log/messages from booting, and
three 'iwlist scan's following that.

# Boot, no scans performed
Sep 24 15:08:50 megadoomer kernel: iwl3945: Intel(R) PRO/Wireless 3945ABG/BG
Network Connection driver for Linux, 0.1.15kds
Sep 24 15:08:50 megadoomer kernel: iwl3945: Copyright(c) 2003-2007 Intel Corporation
Sep 24 15:08:50 megadoomer kernel: iwl3945: Detected Intel PRO/Wireless 3945ABG
Network Connection
Sep 24 15:08:50 megadoomer kernel: iwl3945: Tunable channels: 11 802.11bg, 13
802.11a channels
Sep 24 15:09:05 megadoomer NetworkManager: <info>  wlan0: Device is
fully-supported using driver 'iwl3945'.

# First scan - no results
Sep 24 15:13:24 megadoomer kernel: iwl3945: Microcode SW error detected. 
Restarting 0x82000008.
Sep 24 15:13:24 megadoomer kernel: iwl3945: Error Reply type 0x00000000 cmd
REPLY_SCAN_CMD (0x80) seq 0x4438 ser 0x00000000
Sep 24 15:13:25 megadoomer kernel: iwl3945: Can't stop Rx DMA.
Sep 24 15:13:25 megadoomer kernel: iwl3945: No space for Tx
Sep 24 15:13:25 megadoomer kernel: iwl3945: Error sending REPLY_ADD_STA:
iwl_enqueue_hcmd failed: -28
Sep 24 15:13:25 megadoomer kernel: iwl3945: No space for Tx
Sep 24 15:13:25 megadoomer kernel: iwl3945: Error sending
REPLY_TX_PWR_TABLE_CMD: iwl_enqueue_hcmd failed: -28

# Second scan - 4 results (there are 17 APs within range)
Sep 24 15:14:41 megadoomer kernel: iwl3945: REPLY_ADD_STA failed

# Third scan - Failed to read scan data : Resource temporarily unavailable
Sep 24 15:15:18 megadoomer kernel: iwl3945: Microcode SW error detected. 
Restarting 0x82000008.
Sep 24 15:15:18 megadoomer kernel: iwl3945: Error Reply type 0x00000000 cmd
REPLY_SCAN_CMD (0x80) seq 0x4424 ser 0x00000000
Sep 24 15:15:19 megadoomer kernel: iwl3945: Can't stop Rx DMA.


rmmod iwl3945; modprobe iwl3945 does not seem to help.

Version-Release number of selected component (if applicable):
iwl3945-firmware-2.14.1.5-2
kernel-2.6.23-0.189.rc6.git8.fc8
Comment 1 drago01 2007-09-26 16:44:19 UTC 
this is not a firmware bug but a bug in the iwlwifi driver.
there is a upstream fix :
http://intellinuxwireless.org/repos/?p=iwlwifi.git;a=commit;h=7cf42e587d3a381d2655e1eb1cbee9603f337173
John can you add this patch to the kernel rpm? it should hit upstream
wireless-dev soon.
Comment 2 John W. Linville 2007-09-26 16:55:54 UTC 
Best would be if you could presuade Intel to post it upstream...
Comment 3 drago01 2007-09-26 17:03:22 UTC 
sent a mail to ipw3945-devel about it...
Comment 4 John W. Linville 2007-09-28 14:01:02 UTC 
kernel-2.6.23-0.212.rc8.git2.fc8 has a sizeable iwlwifi update.  Can you 
replicate the issue with that kernel?
Comment 5 Zack Cerza 2007-09-28 19:39:08 UTC 
Looks to be fixed in kernel-2.6.23-0.212.rc8.git2.fc8. Thanks!
Comment 6 Zack Cerza 2007-10-01 18:59:31 UTC 
OK, still on the same kernel and seeing this with iwlist wlan0 scan:

"wlan0    Failed to read scan data : Resource temporarily unavailable"

I haven't seen anything interesting in /var/log/messages this time, though.
Comment 7 John W. Linville 2007-10-01 19:45:08 UTC 
Is that a temporary condition?  Or is the device forever useless afterwards?
Comment 8 Zack Cerza 2007-10-01 20:02:15 UTC 
Rebooting seemed to give me some luck.
Comment 9 John W. Linville 2007-10-16 17:51:38 UTC 
Is this still a problem with current rawhide of fc8 kernels?
Comment 10 Zack Cerza 2007-10-16 18:28:17 UTC 
I've done a fair amount of testing, and haven't had any problems. Closing.
Note You need to log in before you can comment on or make changes to this bug.