Centrino Wireless-N 1000 support is also broken (Re: iwn(4) in -HEAD supporting Centrino Wireless-N 135)

[Adrian Chadd]

Ok, this is much, much clearer. I'll go look at r258035, which is the
new chip config stuff.


-a


On 29 April 2014 17:53, Kaho Toshikazu <kaho at elam.kais.kyoto-u.ac.jp> wrote:
>   Hello, thank you for your reply.
>
>> Would you mind clarifying what you wrote a little more?
>> What I'm reading this as:
>>
>> * If I grab an r258030 tree, it's broken on your NIC
>> * If I then do what, apply r258034, it works?
>> * If I apply 258035 -> 258085, it also works?
>>
>> I really would appreciate it if you would re-state what you've written
>> so I/we can better understand :-)
>>
>> Something like:
>>
>> * Revision X, doesn't work
>> * Revision X with revisions A, B and then C applied, does work
>> * Revision X with A, B and C applied, and a modified revision D
>> applied, does work (and attach the modified revision D)
>>
>> That way I/we can clearly reproduce what you've done. :-)
>>
>> Thanks!
>
>   I'm sorry for my ambiguous representation.
>
> Until Revision 257951, work.
>
> From r258030 to r258034, doesn't work. If antmask_2stream set to
> be IWN_ANT_AB, work.
>
> After r258035, doesn't work. antmask_2stream = IWN_ANT_AB doesn't help.
>
>> Ok. So r258030 changed the antenna configuration for transmitting to
>> use the tx chainmask from the EEPROM settings,r ather than hard-coding
>> it to IWN_ANT_AB. This is needed because a bunch of NICs actually
>> don't have two antennas and if you feed it IWN_ANT_AB it just plainly
>> doesn't work.
>>
>> So, if you try -HEAD, but you set antmask_2stream to be IWN_ANT_AB,
>> does it fix the problem?
>>
>> Ie, add this line
>>
>> linkq.antmsk_2stream = IWN_ANT_AB;
>>
>> .. just after this line:
>>
>> linkq.ampdu_limit = htole16(4000);
>
> Yes, it fix the problem until r258034, but doesn't fix after r258035.
> Rivision 258035 changed many lines and if_iwn_chip_cfg.h was created.
> I faced a trouble about where should I try to change.
>
> --
> Kaho Toshikazu