Return-path: Received: from fk-out-0910.google.com ([209.85.128.187]:11100 "EHLO fk-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750908AbYISIDD (ORCPT ); Fri, 19 Sep 2008 04:03:03 -0400 Received: by fk-out-0910.google.com with SMTP id 18so298072fkq.5 for ; Fri, 19 Sep 2008 01:03:02 -0700 (PDT) Message-ID: <1ba2fa240809190103i69c66978he4762fd350ab2e5d@mail.gmail.com> (sfid-20080919_100312_773600_994670A4) Date: Fri, 19 Sep 2008 11:03:02 +0300 From: "Tomas Winkler" To: "Johannes Berg" Subject: Re: technology preview: scan with cfg80211 Cc: linux-wireless , "Dan Williams" In-Reply-To: <1221795715.10419.15.camel@johannes.berg> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 References: <1221794329.10419.9.camel@johannes.berg> <1221795715.10419.15.camel@johannes.berg> Sender: linux-wireless-owner@vger.kernel.org List-ID: On Fri, Sep 19, 2008 at 6:41 AM, Johannes Berg wrote: > On Fri, 2008-09-19 at 05:18 +0200, Johannes Berg wrote: >> This is totally incomplete: >> >> cfg80211 needs to >> * actually hash on the BSSID instead of putting things into a list >> * provide BSS list functions so mac80211 need not keep track >> * allow a "private" area in each scan result with hw-provided size >> * handle mesh properly >> * provide a way for the hw to limit the number of SSIDs in a scan Don't understand this requirement? * provide a way to limit the number of channels in a scan (for faster roaming candidate search) > > * provide a function to just pass in the whole beacon/probe response > frame and pick it apart itself > * handle aging > * provide a wext "set scan" hook that creates a request and passes it > off to ->scan(), provide wext "get scan" hook > * parse the supported rates into a bitmap like mac80211 does now, based > on the bitrate information the wiphy provides > >> in mac80211, we need to >> * pass the scan request to the hardware offload >> * handle multiple SSIDs in a scan >> * use BSS list stuff cfg80211 needs to provide >> * handle active/passive properly > > * use all those helpers I mentioned above > > Also, we need to think about > * how fullmac stuff can integrate, can it provide this IE-based > information, and if not, is it feasible to just synthesize it? > * whether this should really be per wiphy as it is now, or somehow more > per netdev, or not related to a netdev at all... > > johannes >