Re: Is iwn drivers working on Current ?

From: Brandon Gooch <jamesbrandongooch_at_gmail.com>
Date: Fri, 13 Mar 2009 07:00:43 -1000
On Mar 13, 2009, at 5:52 AM, David Horn <dhorn2000_at_gmail.com> wrote:

> On Fri, Mar 13, 2009 at 2:23 AM, Olivier Cochard-Labbé
> <olivier_at_freenas.org> wrote:
>> Hi,
>>
>> I try to use the iwn drivers on my Dell Latitude D630 laptop on a
>> up-to-date current, but I can't.
>>
>> I've an Intel 4965AGN card, and here is the result of pciconf -lv :
>>
>> iwn0_at_pci0:12:0:0:       class=0x028000 card=0x11218086  
>> chip=0x42298086
>> rev=0x61 hdr=0x00
>>    vendor     = 'Intel Corporation'
>>    device     = 'Intel 4965AGN Intel Wireless WiFi Link
>> 4965AGN(supporting 802.11a/b/g/Draft-N)'
>>    class      = network
>>
>> I've put on /boot/loader.conf the magics lines:
>> legal.intel_iwn.license_ack=1
>
> The license ack for iwn is not needed anymore btw.
>
>> if_iwn_load="YES"
>>
>> my /etc/rc.conf is configured like that:
>> wlans_iwn0="wlan0"
>> ifconfig_wlan0="country FR WPA DHCP"
>>
>> and the /etc/wpa_supplicant.conf was tested using 2 others differents
>> wirelles card (USB rum and PCMCIA ath) without problem.
>> But I have this error message on dmesg:
>>
>> iwn0: need multicast update callback
>> iwn0: error, INTR=2000000<SW_ERROR> STATUS=0x0
>> iwn0: error, INTR=2000000<SW_ERROR> STATUS=0x0
>> iwn0: iwn_config: configure command failed, error 35
>> iwn0: iwn_init_locked: could not configure device, error 35
>> iwn0: iwn_config: configure command failed, error 35
>> iwn0: iwn_init_locked: could not configure device, error 35
>> iwn0: iwn_transfer_firmware: timeout waiting for first alive  
>> notice, error 35
>> iwn0: iwn_init_locked: could not load firmware, error 35
>
> I have seen these exact same error 35 issues.
>
>>
>> (The wireless switch is on the ON position, and this wireless works
>> when booting with Linux Ubuntu).
>>
>> And not association possible...
>>
>> Here are some other information:
>>
>> [root_at_d630]~#ifconfig wlan0
>> wlan0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0  
>> mtu 1500
>>        ether 00:1d:e0:72:10:01
>>        inet6 fe80::21d:e0ff:fe72:1001%wlan0 prefixlen 64 scopeid 0x7
>>        media: IEEE 802.11 Wireless Ethernet autoselect (autoselect)
>>        status: no carrier
>>        ssid "" channel 13 (2472 Mhz 11g)
>>        regdomain ETSI country F2 authmode WPA1+WPA2/802.11i privacy  
>> ON
>>        deftxkey UNDEF txpower 30 bmiss 10 scanvalid 60 protmode CTS  
>> wme
>>        roaming MANUAL
>> [root_at_d630]~#ifconfig wlan0 list scan
>> [root_at_d630]~#ifconfig wlan0 list caps
>> drivercaps= 
>> 581e001<STA,TXPMGT,SHSLOT,SHPREAMBLE,MONITOR,WPA1,WPA2,WME>
>> [root_at_d630]~#uname -a
>> FreeBSD d630.freenas.org 8.0-CURRENT FreeBSD 8.0-CURRENT #14: Thu Mar
>> 12 14:17:22 CET 2009
> <snip>
>> Does someone know where the problem is ?
>
> Yes, I have seen this problem with iwn on current, but unfortunately
> do not know a real "fix" at the moment.
>
> For me, it seems to be timing related at boot time, caused by IPv6
> rtsol starting on the interface before iwn had settled.  iwn0 should
> never be initialized by rtsol, as it is the hardware 802.11 interface.
> wlan0 should be initialized instead. (-current vap code)
>
> Questions to help see if you are seeing the same problem:
>
> 1)  Do you have ipv6_enabled="yes" in rc.conf ?
> 2)  Report on the results of "ifconfig -l"
>
> In my case, iwn0 was the first interface in the list, resulting in
> rtsol starting on iwn0 (and not some other interface).  There are
> several unique issues here, but I believe that a work-around patch
> that I wrote to /etc/network.subr that might help is attached to PR
> conf/132179:
>
> http://www.freebsd.org/cgi/query-pr.cgi?pr=conf/132179
>
> The real issue is likely a bug in the iwn driver.  Someone with real
> knowledge of the iwn code would probably give better insight than me.
> Sam ?  Brandon ?
>
> Good Luck.
>
> --_Dave
>
>>
>> Thanks for your help,
>>
>> Olivier
> _______________________________________________
> freebsd-current_at_freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-current
> To unsubscribe, send any mail to "freebsd-current-unsubscribe_at_freebsd.org 
> "

I made an attempt to update the driver by grabbing a recent snapshot  
of the OpenBSD driver from CVS -- I don't remember the exact date, but  
it was after Damien Bergamini's "complete rewrite", and it includes  
support for newer cards as well.

After reading and working with the source in an attempt to at least  
get it to build, I stalled back in February and only yesterday did I  
get to look at the code again.

I was initially working with 7.1-STABLE, but recently moved to using  
current. Incidentally, many of the issues I was having with the de- 
vapified, backported version of the driver disappears after moving to  
current, allowing me to use the wireless card for daily work. It still  
requires a kld-unload,load occasionally...

So, essentially I have a heavily commented set of source files with no  
vap (and no locking) that won't build.

It has been an intense learning experience, though. Robert Watson's  
fxr.watson.org is invaluable!

Oh, and I did grab the newest firmware from Damien's site and got that  
to "build" ok -- of course it doesn't load with the current driver.

-brandon
Received on Fri Mar 13 2009 - 16:01:00 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:43 UTC