Re: regulatory problems with ath5k

[Arnd Hannemann <hannemann@nets.rwth-aachen.de>]

Am 26.05.2010 19:11, schrieb Luis R. Rodriguez:
> On Wed, May 26, 2010 at 4:23 AM, Arnd Hannemann
> <hannemann@nets.rwth-aachen.de> wrote:
>   
>> Hi,
>>
>> I'm trying to use 802.11a channels in AP mode with my ath5k nics with kernel 2.6.34.
>> And after setting up crda, udev and wireless regdb,
>>     
> Which release of wireless-regdb?
>
>   
This morning's git tree. I just copied the regulatory.bin from there.
Last commit was:

commit f3ba942f073358d1bca82a389da9f7da6e8fe81f
Author: John W. Linville <linville@tuxdriver.com>
Date:   Fri Feb 5 13:47:41 2010 -0500

    wireless-regdb: update sha1sum.txt in Makefile and fix REGDB_CHANGED
   
    Signed-off-by: John W. Linville <linville@tuxdriver.com>




>> "iw list" still shows flags
>> "passive scanning, no IBSS" for most of the 802.11a channels (see below.)
>>
>> I wonder what I'm doing wrong here?
>>     
> Lets take a look.
>
>   
>>        Band 2:
>>                Frequencies:
>>                        * 5180 MHz [36] (20.0 dBm) (radar detection)
>>                        * 5200 MHz [40] (20.0 dBm) (radar detection)
>>                        * 5220 MHz [44] (20.0 dBm) (radar detection)
>>                        * 5240 MHz [48] (20.0 dBm) (radar detection)
>>                        * 5260 MHz [52] (20.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5280 MHz [56] (20.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5300 MHz [60] (20.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5320 MHz [64] (20.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5500 MHz [100] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5520 MHz [104] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5540 MHz [108] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5560 MHz [112] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5580 MHz [116] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5600 MHz [120] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5620 MHz [124] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5640 MHz [128] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5660 MHz [132] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5680 MHz [136] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>                        * 5700 MHz [140] (26.0 dBm) (passive scanning, no IBSS, radar detection)
>>     
>
> Please do: 'iw reg get' and paste the output here, the above channels
> should not be passive scan/no-ibss for 'DE'.
>   

hannemann@mrouter46:~ $ sudo iw reg get
country DE:
       (2400 - 2483 @ 40), (N/A, 20)
       (5150 - 5350 @ 40), (N/A, 20), NO-OUTDOOR, DFS
       (5470 - 5725 @ 40), (N/A, 26), NO-OUTDOOR, DFS


>> hannemann@mrouter46:~ $ sudo iw reg get
>> country DE:
>>        (2400 - 2483 @ 40), (N/A, 20)
>>        (5150 - 5350 @ 40), (N/A, 20), NO-OUTDOOR, DFS
>>        (5470 - 5725 @ 40), (N/A, 26), NO-OUTDOOR, DFS
>>
>> dmesg snippet:
>>
>> May 26 12:10:36 mrouter46 kernel: [   30.023089] console [netcon0] enabled
>> May 26 12:10:36 mrouter46 kernel: [   30.023103] netconsole: network logging started
>> May 26 12:10:36 mrouter46 kernel: [   30.314702] cfg80211: Calling CRDA to update world regulatory domain
>> May 26 12:10:37 mrouter46 kernel: [   30.555745] ath5k 0000:00:0c.0: registered as \'phy0\'
>> May 26 12:10:37 mrouter46 kernel: [   31.039151] ath: EEPROM regdomain: 0x8114
>>     
> WTF your EEPROM has 0x8114 set on the regulatory domain. David, have
> you ever seen such a thing? This kernel (2.6.34) should already have
> an ath5k EEPROM checksum check so the EEPROM should be valid. Lets
> figure this out first before digging more into this.
>
> Some more notes below though.
>   

You got me, I hacked this up:

--- linux-2.6.34/drivers/net/wireless/ath/regd.c    2010-05-16
23:17:36.000000000 +0200
+++ linux-2.6.34.x-um/drivers/net/wireless/ath/regd.c    2010-05-26
19:57:44.000000000 +0200
@@ -501,7 +501,7 @@ ath_regd_init(struct ath_regulatory *reg
         return -EINVAL;
 
     ath_regd_sanitize(reg);
-
+    reg->current_rd = CTRY_GERMANY | COUNTRY_ERD_FLAG;
     printk(KERN_DEBUG "ath: EEPROM regdomain: 0x%0x\n", reg->current_rd);
 
     if (!ath_regd_is_eeprom_valid(reg)) {


Do you think this is the cause?

>   
>> May 26 12:10:37 mrouter46 kernel: [   31.039168] ath: EEPROM indicates we should expect a country code
>> May 26 12:10:37 mrouter46 kernel: [   31.039188] ath: doing EEPROM country->regdmn map search
>> May 26 12:10:37 mrouter46 kernel: [   31.039205] ath: country maps to regdmn code: 0x37
>> May 26 12:10:37 mrouter46 kernel: [   31.039222] ath: Country alpha2 being used: DE
>> May 26 12:10:37 mrouter46 kernel: [   31.039237] ath: Regpair used: 0x37
>> May 26 12:10:37 mrouter46 kernel: [   31.054155] cfg80211: World regulatory domain updated:
>>     
> So cfg80211 *first* starts out with the world regulatory domain by
> default, and then it will call CRDA for DE since your regpair mapped
> to 'DE'.
>
>   
>> May 26 12:10:37 mrouter46 kernel: [   31.054178]     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>> May 26 12:10:37 mrouter46 kernel: [   31.054207]     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>> May 26 12:10:37 mrouter46 kernel: [   31.054232]     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
>> May 26 12:10:37 mrouter46 kernel: [   31.054258]     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
>> May 26 12:10:37 mrouter46 kernel: [   31.054284]     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>> May 26 12:10:37 mrouter46 kernel: [   31.054310]     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
>> May 26 12:10:38 mrouter46 kernel: [   31.212057] phy0: Selected rate control algorithm \'minstrel\'
>> May 26 12:10:38 mrouter46 kernel: [   31.212732] ath5k phy0: Atheros AR5213A chip found (MAC: 0x59, PHY: 0x43)
>> May 26 12:10:38 mrouter46 kernel: [   31.212732] ath5k phy0: RF5112B multiband radio found (0x36)
>> May 26 12:10:38 mrouter46 kernel: [   31.212732] ath5k 0000:00:0e.0: registered as \'phy1\'
>> May 26 12:10:38 mrouter46 kernel: [   31.622592] cfg80211: Calling CRDA for country: DE
>>     
> Here is the call for 'DE'
>
>   
>> May 26 12:10:38 mrouter46 kernel: [   31.688006] cfg80211: Regulatory domain changed to country: DE
>> May 26 12:10:38 mrouter46 kernel: [   31.688030]     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
>> May 26 12:10:38 mrouter46 kernel: [   31.688057]     (2400000 KHz - 2483500 KHz @ 40000 KHz), (N/A, 2000 mBm)
>> May 26 12:10:38 mrouter46 kernel: [   31.688080]     (5150000 KHz - 5350000 KHz @ 40000 KHz), (N/A, 2000 mBm)
>> May 26 12:10:38 mrouter46 kernel: [   31.688103]     (5470000 KHz - 5725000 KHz @ 40000 KHz), (N/A, 2698 mBm)
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: EEPROM regdomain: 0x8114
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: EEPROM indicates we should expect a country code
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: doing EEPROM country->regdmn map search
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: country maps to regdmn code: 0x37
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: Country alpha2 being used: DE
>> May 26 12:10:38 mrouter46 kernel: [   31.745694] ath: Regpair used: 0x37
>> May 26 12:10:38 mrouter46 kernel: [   31.758662] phy1: Selected rate control algorithm \'minstrel\'
>> May 26 12:10:38 mrouter46 kernel: [   31.759383] ath5k phy1: Atheros AR5213A chip found (MAC: 0x59, PHY: 0x43)
>> May 26 12:10:38 mrouter46 kernel: [   31.759383] ath5k phy1: RF5112B multiband radio found (0x36)
>>     
> Do you have two cards?
>   

Yes.

Best regards,
Arnd