RE: [EXT] Re: [PATCH v9 0/2] wifi: mwifiex: add code to support host mlme

[David Lin]

> From: David Lin
> Sent: Wednesday, March 20, 2024 9:00 AM
> To: Johannes Berg <johannes@xxxxxxxxxxxxxxxx>; Brian Norris
> <briannorris@xxxxxxxxxxxx>; Francesco Dolcini <francesco@xxxxxxxxxx>
> Cc: kvalo@xxxxxxxxxx; linux-wireless@xxxxxxxxxxxxxxx;
> linux-kernel@xxxxxxxxxxxxxxx; Pete Hsieh <tsung-hsien.hsieh@xxxxxxx>;
> rafael.beims <rafael.beims@xxxxxxxxxxx>; Francesco Dolcini
> <francesco.dolcini@xxxxxxxxxxx>
> Subject: RE: [EXT] Re: [PATCH v9 0/2] wifi: mwifiex: add code to support host
> mlme
> 
> > From: Johannes Berg <johannes@xxxxxxxxxxxxxxxx>
> > Sent: Tuesday, March 19, 2024 8:13 PM
> > To: Brian Norris <briannorris@xxxxxxxxxxxx>; Francesco Dolcini
> > <francesco@xxxxxxxxxx>
> > Cc: kvalo@xxxxxxxxxx; linux-wireless@xxxxxxxxxxxxxxx;
> > linux-kernel@xxxxxxxxxxxxxxx; David Lin <yu-hao.lin@xxxxxxx>; Pete
> > Hsieh <tsung-hsien.hsieh@xxxxxxx>; rafael.beims
> > <rafael.beims@xxxxxxxxxxx>; Francesco Dolcini
> > <francesco.dolcini@xxxxxxxxxxx>
> > Subject: [EXT] Re: [PATCH v9 0/2] wifi: mwifiex: add code to support
> > host mlme
> >
> > Caution: This is an external email. Please take care when clicking
> > links or opening attachments. When in doubt, report the message using
> > the 'Report this email' button
> >
> >
> > On Fri, 2024-03-15 at 17:49 -0700, Brian Norris wrote:
> > >
> > > Now that I've looked a bit closer today: I'm realizing this may(?)
> > > be one of the first "full MAC" drivers trying to implement its own
> > > MLME
> > > -- or at least, the auth/assoc bits.
> >
> > Hmm, yeah, why _is_ that? mwifiex was originally "sold" as a "full MAC"
> > driver, i.e. doing things in the firmware.
> >
> > We've said that "soft MAC" drivers should be using mac80211, but this
> > thing can't seem to decide?
> >
> > Also decl.h should probably _shrink_ rather than grow, a number of
> > things just replicate ieee80211.h (such as MWIFIEX_MGMT_HEADER_LEN
> > really is just
> > sizeof(ieee80211_mgmt) or so? Not quite correctly.)
> >
> 
> This can be done for feature patches.
> 
> > So yeah, agree with Brian, not only would this be the first, but it's
> > also something we don't really _want_. All other drivers that want
> > stuff like this are stuck in staging ...
> >
> > So why is this needed for a supposedly "firmware does it all" driver,
> > and why can it not be integrated with mac80211 if it's no longer "firmware
> does it all"?
> >
> > Johannes
> 
> Our proprietary driver is cfg80211 driver, it is very hard to create a brand new
> mac80211 driver and still can port all tested stuffs from our proprietary driver.
> 
> Thanks,
> David

BTW, vendor should have the choice to use cfg80211 or mac80211 for their chips, right?