I've been trying to figure this out for some time now... I have several systems that have 8- and 16-port RocketPort boards in them, connected to banks of modems. (ISA cards on all of them - I just got a PCI RP board, but haven't installed it anywhere yet.) These modems make a lot of outgoing calls (300+ each per day), and a small (2-3 each/day) amount of incoming calls. Every so often, I end up with a mgetty process that refuses to go away: 37021 ?? IE 0:00.04 /usr/local/sbin/mgetty cuaR1 38379 ?? I 0:00.04 /usr/local/sbin/mgetty cuaR3 38642 ?? I 0:00.04 /usr/local/sbin/mgetty cuaR0 (PID 37021 is the one that's stuck.) Note the "E" in the status field (ps ax). From the ps man page, "E" means "The process is trying to exit." "kill -9" doesn't make it die... Sometimes, cycling the modem power will fix the problem, but only about half of the time. The rest of the time, a reboot is required. I have a little prog that shows what the computer thinks the hardware handshaking lines say - on a "stuck" line, it says that DSR & CTS are low, where they're normally high on a working modem. Funny thing is, my little Radio Smack RS-232 tester says that the DSR & CTS lines are actually high, as do the modem LEDs. Any ideas? (My next step is to grab a voltmeter and measure the juice on those pins - I'm assuming it's close enough to the +-12V to give a valid signal, but who knows.... Even so, I'd think there should be SOME way to use a software fix to release the port. If "kill -9" doesn't kill a process, I'd almost consider that a bug.) mike To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
