Date: Wed, 23 May 2001 00:05:46 -0700 From: "Ted Mittelstaedt" <tedm@toybox.placo.com> To: "David Banning" <david@banning.com> Cc: <questions@FreeBSD.ORG> Subject: RE: digiboard serial port behaviour problem Message-ID: <001a01c0e356$ca516b40$1401a8c0@tedm.placo.com> In-Reply-To: <20010522111648.A6103@yahoo.com>
next in thread | previous in thread | raw e-mail | index | archive | help
>-----Original Message----- >From: David Banning [mailto:sky_tracker@yahoo.com] >Sent: Tuesday, May 22, 2001 4:17 AM >To: Ted Mittelstaedt >Cc: David Banning; questions@FreeBSD.ORG >Subject: Re: digiboard serial port behaviour problem > > >> But, this database program hangs when it comes time >> for the opening screen. > >Yes. > >> >> My guess is that the database program is doing something during >> the opening screen that has nothing whatsoever to do with the >> terminal session, but is making the process crash. Since this >> is an SCO program that is most likely a function call into the >> SCO libraries, and thence to the ibcs2 emulator. > >If that is the case, wouldn't the program also crash on the console? yes, it would. >It runs fine on the system console. It also runs fine over a network >connection running on a X-Win32 xterm session. > Well, so much for that guess. > >Could it be that the database program >is attempting to work with FreeBSD's serial-port terminal >operation and thinks it's SCO? >The database does have it's own termcap file in >/usr/accell/lib > If it has it's own termcap (not uncommon with SCO programs) then it's probably opening the serial port in raw mode, because it expects to be doing all of the serial line handling. It sounds like a buffer overflow somewhere, possibly in between the program and the ibcs2 emulation layer, or between the emulation layer and the FreeBSD sio drivers. You did say, after all, that a terminal connected to com2 also locked up. I think, though, that I can recommend a solution for you, although I don't know how much your gonna like it. What you need to do is jettison the Digiboard and all other serial connections to the FreeBSD system. Then, you get on Ebay and obtain a Cisco 2511 terminal server plus Octal cable, which you can get for about $1000. You might also be able to dump the Digiboard there too, although you would be incredibly lucky to get $50 for it. You then plug the Terminals into the Cisco, and load a simple Cisco configuration into the 2511. Now, when people sit down and switch on their terminals, they get a FreeBSD Login: prompt. (you set the 2511 to Telnet the port into the FreeBSD system when a terminal is switched on) There are other terminal servers than Cisco 2511's, of course, even Digoboard makes some. Also, if you just can't bear to part with it, you could attempt to stick the Digiboard into a SECOND BSD system that is basically made up as a terminal server. But I think this is silly, your much better off getting those terminals onto a hardware terminal server, and the 2511 is designed for this. Ted > >> >> >> Ted Mittelstaedt tedm@toybox.placo.com >> Author of: The FreeBSD Corporate Networker's Guide >> Book website: http://www.freebsd-corp-net-guide.com >> >> >> >-----Original Message----- >> >From: David Banning [mailto:sky_tracker@yahoo.com] >> >Sent: Monday, May 21, 2001 7:12 PM >> >To: Ted Mittelstaedt >> >Cc: David Banning; questions@FreeBSD.ORG >> >Subject: Re: digiboard serial port behaviour problem >> > >> > >> >> >I am converting an old SCO system to FreeBSD and the database they >> >> >are running is an old SCO 3.2 binary. >> >> > >> >> >The digiboard PC/xe board seems to work with vi and mutt fine but >> >> >when there t comes time for the opening screen on this >database program >> >> > (I guess there is a burst of a lot of characters at one time) >> >> >the terminal hangs. >> >> > >> >> >> >> Are you ABSOLUTELY sure that the TERMINAL hangs and not the session? >> > >> >After some testing, it appears to be the session that is hanging. >> >Turning the wyse50 terminal off and on again does not produce >> >a working line again. >> >I even try to kill it with a "kill -9" and it takes quite a while >> >to die. >> > >> >> This would seem to be a session thing, not a terminal problem. But, >> >> if it really is a terminal problem, then you might be able to >> >> specify padding in the termcap entry for the terminal that would >> >> fix the problem. Padding was often used for older 3-wire >> >installations that >> >> used xon/xoff and the older terminals that >> >> only supported xon/xoff. >> > >> >Well I wonder about this now. I have tried a test that man >> >termcap suggests (in vi delete 16 lines, then hit u several >times quickly) >> >and it appears to work just fine. >> >It also appears that I have eliminated the digiboard as the culprit. >> >I attached a dumb terminal to /dev/cuaa0 and I get the same problem. >> >I also tried connecting a dumb terminal directly to a full (7 wire?) >> >serial line, and I have the problem there too. >> > >> >So now I am stumped even as to where to look for the problem. >> >I did a "stty -f /dev/cuaD00" on one of the ports and got; >> > >> >speed 38400 baud; >> >lflags: -icanon -isig -iexten -echo >> >iflags: -icrnl -ixon -ixany -imaxbel -brkint >> >oflags: -opost onocr onlret >> >cflags: cs8 -parenb >> > >> >Any ideas where to go from here? >> > >> >_________________________________________________________ >> >Do You Yahoo!? >> >Get your free @yahoo.com address at http://mail.yahoo.com >> > >> > >> > >-- >Croll's Query: > If tin whistles are made of tin, what are foghorns made of? > >_________________________________________________________ >Do You Yahoo!? >Get your free @yahoo.com address at http://mail.yahoo.com > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?001a01c0e356$ca516b40$1401a8c0>