Message-ID: <42e2c59c035142597f00038b8a67f13a9fb40292.camel@sipsolutions.net>
Subject: Re: [PATCH] cfg80211: Authentication offload to user space in AP
 mode
From:   Johannes Berg <johannes@sipsolutions.net>
To:     dasaris@codeaurora.org,
        "Otcheretianski, Andrei" <andrei.otcheretianski@intel.com>
Cc:     linux-wireless@vger.kernel.org
Date:   Tue, 15 Jan 2019 14:26:56 +0100
In-Reply-To: <7bfaa41538350a2b385b14b00bb0444e@codeaurora.org>
References: <1544779909-27758-1-git-send-email-dasaris@codeaurora.org>
         (sfid-20181214_103219_686121_BC072990) <dd3263b2b3f0850a26b316f9e8cb0cba81dc97c7.camel@sipsolutions.net>
         <B2CD83211F5FDD478D7EFFF030D107EDCD031EAB@HASMSX111.ger.corp.intel.com>
         <7bfaa41538350a2b385b14b00bb0444e@codeaurora.org>
Content-Type: text/plain; charset="UTF-8"
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-wireless-owner@vger.kernel.org
Precedence: bulk

On Thu, 2018-12-20 at 16:37 +0530, dasaris@codeaurora.org wrote:
> On 2018-12-18 19:10, Otcheretianski, Andrei wrote:
> > > So, interesting - Andrei also added the PMKID but also the PMK here, 
> > > in a
> > > separate patch:
> > > https://patchwork.kernel.org/patch/10732093/
> > 
> > Yeah, PMK is needed for the devices that offload 4way handshake.
> 
> Thanks. Fine with us too. Will have an updated patchset with the pmkid 
> part removed, as this is already added with the patch by Andrei.

Well, we're still discussing the whole thing :)

You can add the PMKID here as well, and then Andrei just has to rebase
accordingly?

> > > > [...]
> > > > -	if (!info->attrs[NL80211_ATTR_SSID])
> > > > +	if (!info->attrs[NL80211_ATTR_SSID] &&
> > > > +	    dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP &&
> > > > +	    dev->ieee80211_ptr->iftype != NL80211_IFTYPE_AP_VLAN &&
> > > > +	    dev->ieee80211_ptr->iftype != NL80211_IFTYPE_P2P_GO)
> > > >  		return -EINVAL;
> > > 
> > > Hm. do we really expect this on AP_VLAN?
> > 
> > Actually I'm not sure about GO neither. P2P doesn't use SAE etc..
> > 
> > Andrei
> 
> Since these vifs are similar to an AP, we have considered external 
> authentication to be applicable for them as well. Please let me know if 
> this is fine . I shall update the patch accordingly.

Sure, I think it's fine. Even if P2P right now doesn't actually
specify/use WPA3/SAE we can still support it, right?

My question was really more about AP_VLAN, because I was expecting the
handshake to still happen on the AP interface, not on the AP_VLAN?

johannes