From mboxrd@z Thu Jan 1 00:00:00 1970 From: Cathryn Mataga Subject: Re: kissattach/kissnetd versus ptyq ttyq and ptmx/pts Date: Thu, 06 Jul 2006 01:14:02 -0700 Message-ID: <44ACC64A.6000609@junglevision.com> References: <20060623130355.9211a397.washer@trlp.com> <44AB00AC.7000708@junglevision.com> <620c90570607041948l150297b6ue70236ee277d5ce8@mail.gmail.com> <44AB6C66.4050000@junglevision.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <44AB6C66.4050000@junglevision.com> Sender: linux-hams-owner@vger.kernel.org List-Id: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Cc: linux-hams@vger.kernel.org Cathryn Mataga wrote: > In my original Linux ham radio setup, I had the following... > > /sbin/kissnetd -f 1064 /dev/ptyq3 /dev/ptyq4& > sleep 4 > /sbin/kissattach -l /dev/ttyq3 lb1 192.168.0.2& > /sbin/kissattach -l /dev/ttyq4 lb2 192.168.0.3& > > for creating the loopback device. > > Seems that in Fedora 4 ptyq3 and ttyq3 don't exist anymore. I tried > using > MAKENODE ptyq3 and ttyq3 to create these device files, but I don't > think they're > working? I get this error. > > kissattach: open: No such device or address > > I've been googling the pseudo terminals for Linux and it seems that > the new way is > with ptmx/pts and ptyq3 and ttyq3 are the old style. But I'm not > sure, will kissattach > works with these? Or is there a way to get ptyq3 ttyq3 working in > Fedora 5? > > What's the right way to go about this, these days? > - Hmm, well I'm still kind of stuck on this a little bit. This is what I'm thinking is the situation, just puzzling at a few files here. 1. I don't see an obvious option to compile the kernel with ptyq/ttyq support. Maybe this is hidden somewhere? Or maybe this should be working, and I've missed it? This would be easiest. Maybe it's working on FC5 and I've confused myself into thinking it's not, but I don't know. 2. kissnetd and kissattach aren't that complicated. But it looks like they are designed to work with ptyq/ttyq style devices, and nothing related to ptmx exists. 3. I naively thought, well, I'll just put the servers on ptmx and fish around on /dev/pts/[0-9] to see if anything can connect, and this doesn't works. It seems to print some kind of messages but then it gets a read error. 4. It looks like from the man page ptmx needs a call to grantpt and to unlockpt. it looks like an open on /dev/ptmx creates a random slave name. So somehow this has to be printed out, or output to a file. and then use some shell magic to get those names to kissattach so it can connect to them? 5. This is speculation, but I think kissattach maybe doesn't have to change. But kissnetd does. There are some ioctl statements in kissattach, and I'm not sure of any of these.. 6. Is this the tip of the iceberg? I use virtual ttys with mkiss also. Maybe that's all I need. I'm thinking of going in and adding a few lines to kissnetd and see how that goes. Unless someone who 'knows what's going on' has a better idea.