How to switch driver from wireless extension to cfg/mac80211

How to switch driver from wireless extension to cfg/mac80211

[Philipp Hortmann]

Hi,

after cutting away a lot of dead code from the rtl8192e I would like to 
understand the way forward to migrate the driver from wireless extension 
to cfg/mac80211.

What is the best way forward to migrate the driver rtl8192e towards the
drivers/net/wireless/realtek/rtlwifi/rtl8192* ?

Does it have to be a hard cut?

What to start with?
  - start to change the used structs towards ieee80211_hw?

Thanks for your support.

Bye Philipp

_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: How to switch driver from wireless extension to cfg/mac80211

[jim.cromie]

On Sat, Oct 7, 2023 at 1:14 PM Philipp Hortmann
<philipp.g.hortmann@gmail.com> wrote:
>
> Hi,
>
> after cutting away a lot of dead code from the rtl8192e I would like to
> understand the way forward to migrate the driver from wireless extension
> to cfg/mac80211.
>

I think you need more email recipients for definitive answers.
In particular, is anyone else doing this ?
if they are, you should find them, make the job easier.

> What is the best way forward to migrate the driver rtl8192e towards the
> drivers/net/wireless/realtek/rtlwifi/rtl8192* ?
>

Id start with this search:
$ git log --grep=ieee80211_ -p -- drivers/net/wireless/realtek/ | less
+/ieee80211_

then narrow with a particular api-call that sounds like it would be central
to the switchover.

It might be useful to search for many api calls at once,
rank them by how many of the calls they add
(presuming more to be better, more "comprehensive")


> Does it have to be a hard cut?

not sure what you mean, but I'd guess no - smaller incremental changes
are preferred

>
> What to start with?
>   - start to change the used structs towards ieee80211_hw?

the old adage:  look at the data structs 1st.
not sure how well that applies in any given (or your) situation.

adapting old code to new structs does sound sane, and incrementally testable.

>
> Thanks for your support.
>
> Bye Philipp
>
> _______________________________________________
> Kernelnewbies mailing list
> Kernelnewbies@kernelnewbies.org
> https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies