[ath9k-devel] [PATCH] Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

[Oleksij Rempel <linux@rempel-privat.de>]

Am 08.04.2014 09:51, schrieb Ashish Patro:
> Hi Oleksij,
> 
> I can take a pass at the code to reduce the duplicate code between ath9k and ath9k_htc. I?ll need to make changes to the current spectral.c code and add code into ath_common. Will it be ok for me to do this code refactoring? I?ve so far avoided doing significant refactoring of the existing code in the ath9k driver.

It will be great if you 'll do it. If patches are ok, they will be accepted.
Please separate all you stages to different patches. It will make
bisection easier. End result should looks some way like this:
PATCH 1 - move spectralscan related data from ath_softc to some common
struct and make all related changes in ath9k driver.
PATCH 2 - move all common spectral related functions to common-spectral.c
PATCH 3 - make use of common-spectral.c in ath9k_htc.

After each patch source should be compilable and driver should work.

> Thanks
> Ashish
> 
> On Apr 7, 2014, at 1:10 PM, Oleksij Rempel <linux@rempel-privat.de> wrote:
> 
>> Am 07.04.2014 19:21, schrieb Ashish Patro:
>>> Hi Oleksij,
>>>
>>> I didn't want to poke too much into the ath9k part of the driver code,
>>> so I wrote it in the current way. All the duplicated code is currently
>>> present in spectral_htc.c. The debugFS related code needs to be mostly
>>> duplicated to provide the same interface for ath9k and ath9k_htc drivers.
>>>
>>> The common code can be later moved into ath_common. The main issue is
>>> that the current ath9k code is using references to ath_softc, so the
>>> common code needs to be written to be independent of ath_softc and
>>> ath9k_htc_priv data structures.
>>
>> Correct. spectral.c has no too deep ath_softc dependencies, it can be
>> solved.
>> I don't like the idea of having patch which i already know, that i will
>> need rework it. Ath9k-htc has not enough developers, the goal is to
>> reduce work, not increase it.
>>
>>> On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel <linux@rempel-privat.de
>>> <mailto:linux@rempel-privat.de>> wrote:
>>>
>>>    Am 07.04.2014 01:23, schrieb Ashish Patro:
>>>> - Used the same mechanism to collect spectral scan data as used by
>>>    the existing ath9k code.
>>>> - Minor refactoring done to the code to separate the ath9k and
>>>    ath9k_htc related spectral scan functions into separate files.
>>>>
>>>> This work was done at the WiNGS Lab at the University of Wisconsin
>>>    Madison.
>>>
>>>    Current patch produce lots of duplicated code with ath9k. It will
>>>    increase work later. I guess you plan to fix? ;)
>>>
>>>    --
>>>    Regards,
>>>    Oleksij
>>>
>>>
>>>    _______________________________________________
>>>    ath9k-devel mailing list
>>>    ath9k-devel at lists.ath9k.org <mailto:ath9k-devel@lists.ath9k.org>
>>>    https://lists.ath9k.org/mailman/listinfo/ath9k-devel
>>>
>>>
>>
>>
>> -- 
>> Regards,
>> Oleksij
>>
> 


-- 
Regards,
Oleksij

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 278 bytes
Desc: OpenPGP digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140408/4c5fcdc4/attachment-0001.pgp