Return-Path: Message-Id: <5.1.0.14.2.20030307094003.06de3de0@mail1.qualcomm.com> Date: Fri, 07 Mar 2003 09:44:57 -0800 To: david.libault@inventel.fr, Marcel Holtmann From: Max Krasnyansky Subject: Re: [Bluez-devel] BlueZ Qualification Cc: Daryl Van Vorst , bluez-devel@lists.sourceforge.net, Halam.Rose@7layers-UK.com In-Reply-To: <3E26CE2101C0104A@mel-rta6.wanadoo.fr> (added by postmaster@wanadoo.fr) References: <5.1.0.14.2.20030306133725.025f46d8@mail1.qualcomm.com> <1046706518.14907.133.camel@linux> <5.1.0.14.2.20030306133725.025f46d8@mail1.qualcomm.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" List-ID: At 12:47 AM 3/7/2003, David LIBAULT wrote: >Le Jeudi 6 Mars 2003 23:06, Max Krasnyansky a =E9crit : >> At 10:06 AM 3/3/2003, David LIBAULT wrote: >> >> 2. Have own patches against the stable kernel for qualification >> > >> >IMHO, it is possible to have L2CAP qualified without messing up that mu= ch >> > the code, and without changing the behavior/performance when "the othe= r >> > side" is working properly. >> >> May be. But we don't seem to know that for sure. What about that L2CAP M= TU >> negotiation thing for example ? > >I propose that we take this issue as the first step to qualification. Agreed. >> >Next step : L2CAP qualification. >> > >> >The main issue is to have the tester side... >> > >> > 1) write a small user-mode program that can open and close an A= CL >> > channel, and send any kind of packet over this link >> >> Don't you guys have that already ? If not I'll write this one. >> Do you have any specific features in mind ? > >I have sent to you a very small program (l2capsession.c). You can make it=20 >GPL. To make sure you have it, I have attached it to this message. This=20 >program uses as an input a very simple script file that tells what to send= ,=20 >to what address etc... > >BUT > >This is usable only for "over L2CAP" protocols like SDP, BNEP, RFCOMM etc.= .. >To do L2CAP tests, we need the same type of application but that sends dat= a=20 >directly over an ACL link. Is it a special type of socket that needs to be= =20 >opened to do so ? Yes raw HCI socket. I'll write that one. >This ACLsession program will help us do the L2CAP MTU negotiation test and= =20 >we'll all see if fixing it in the main stream code is that much of a big=20 >deal, and what the 7layers BQB can reasonably accept. Ok. >> > 2) write a script for each test case >> >> Anybody willing to contribute that ? >> > >Inventel can contribute with some test cases scripts using l2capsession.=20 >We'll do that if l2capsession "kind of way to do things" is retained.=20 >For the L2CAP tests (ACLsession) we have nothing yet. Sounds good. >> It sounds like we need to create a module in CVS and start putting thing= s >> that we need together. Any suggestion for a name ? "qube" ? Ok. I'll implement this ACL thing and add create cvs tree (including your l= 2session). Max =20 =20