From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stefan Schmidt Subject: Re: [PATCH] ieee802154: add rx LQI from userspace Date: Fri, 13 Jul 2018 11:43:49 -0400 Message-ID: <858d58d5-be9c-179e-c74c-3b47dee44993@datenfreihafen.org> References: <20180607140802.22666-1-peron.clem@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Cc: Romuald Cari , linux-wpan@vger.kernel.org, Alexander Aring , Stefan Schmidt , "David S . Miller" , netdev@vger.kernel.org, linux-kernel@vger.kernel.org, =?UTF-8?Q?Cl=c3=a9ment_Peron?= To: =?UTF-8?B?Q2zDqW1lbnQgUMOpcm9u?= Return-path: In-Reply-To: Content-Language: en-US Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Hello. On 10.07.2018 11:13, Clément Péron wrote: > Hi Alexander, Stefan, > > Thanks for your feedbacks, > > On Mon, 9 Jul 2018 at 10:49, Stefan Schmidt wrote: >> >> Hello Clement. >> >> Finally coming to review the patch. Sorry for the delay. >> >> On 07.06.2018 16:08, Clément Péron wrote: >>> From: Romuald CARI >>> >>> The Link Quality Indication data exposed by drivers could not be accessed from >>> userspace. Since this data is per-datagram received, it makes sense to make it >>> available to userspace application through the ancillary data mechanism in >>> recvmsg rather than through ioctls. This can be activated using the socket >>> option WPAN_WANTLQI under SOL_IEEE802154 protocol. >> >> I can see that it makes the application life a lot easier to have data >> send out and LQI value synced up by using the socket approach instead of >> dealing with socket and ioctl's. I am good with this patch in general. >> >> So you have some public code that uses this approach? I would be >> interesting in the userspace part of yours. A demo would be fine. If the >> network handling part of your application is public anyway that would be >> even better. :-) > > I'm sorry but the userspace code that use this isn't open. Fair enough. > I will check if I can share this part. > Just the idea is to compute an average LQI and when it reach a > threshold we allow the remote to control the device. Yes, I would be happy with a simple example just showing how to use the socket opt and a very basic algo on how this LQI would be used. regards Stefan Schmidt