From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============3759767735245733651==" MIME-Version: 1.0 From: Denis Kenzior Subject: Re: [PATCH 1/1] Huawei E176: Mark primary and secondary device at ofono.rules TODO: Fix sim_add detection, add E1552 idProduct to ofono.rules Date: Tue, 01 Jun 2010 14:09:10 -0500 Message-ID: <201006011409.11869.denkenz@gmail.com> In-Reply-To: <87d3wc2etx.fsf@potku.valot.fi> List-Id: To: ofono@ofono.org --===============3759767735245733651== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi Kalle, > Marcel Holtmann writes: > > Hi Kalle, > = > Hallo Marcel, > = > >> I personally would prefer a solution which would dynamically probe the > >> ports and choose them based on results. I believe modemmanager does > >> something like this, but I haven't looked in detail. > > > > we should do something like auto-detect at some point, but there are > > limits in it. > = > So what's the best option to go forward? > = > And related to this, can we rely on the port numbering order provided by > udev? For example, on my Huawei E1552 port 0 is the main chat port and > port 2 is the "event" port. Is it certain that port numbering will be > the same across all distributions and kernels? > = I'm not the kernel expert, but the answer is probably 'no'. However, doing = any sort of port auto-detection inside plugins/udev.c is a bit nasty. Best= we = can do is set all of the 2/3/4 ttys and let the plugin probe them. I = certainly have doubts we can do this reliably. If we can't, then we should = consider Florian's approach as a backup. Regards, -Denis --===============3759767735245733651==--