root@candle.uucp (Bruce Momjian) (05/24/91)
I am using AT&T 386 Unix 3.1. When cu and uucp use the modem, they respawn uugetty. When Pcomm 1.2 uses the modem, it does not respawn uugetty, and the next time someone tries to log in, they get a connect but no login prompt. To fix this, I currently run pcomm from a script, which also calls another script with set-uid to root and which executes: UUPID=`/bin/ps -ef |\ /bin/grep " *root.*uugetty[^x]*$" | \ /bin/sed 's/^.*root *\([0-9]*\).*$/\1/'` if [ "$UUPID" -gt "1" ] then /bin/kill $UUPID fi ... which kills the uugetty ( I only have one uugetty running) and respawns. Obviously, this is a kludge. Does anyone know how cu and uucp respawn the uugetty properly, so I may add this to the pcomm source. (BTW, is there a more recent pcomm than 1.2).
jpr@jpradley.jpr.com (Jean-Pierre Radley) (06/06/91)
In article <1991May24.155552.19129@candle.uucp> root@candle.uucp (Bruce Momjian) writes: > >I am using AT&T 386 Unix 3.1. When cu and uucp use the modem, they respawn >uugetty. When Pcomm 1.2 uses the modem, it does not respawn uugetty, and >the next time someone tries to log in, they get a connect but no login prompt. > >To fix this, I currently run pcomm from a script, which also calls >another script with set-uid to root and which executes: > >UUPID=`/bin/ps -ef |\ > /bin/grep " *root.*uugetty[^x]*$" | \ > /bin/sed 's/^.*root *\([0-9]*\).*$/\1/'` >if [ "$UUPID" -gt "1" ] > then /bin/kill $UUPID >fi > >... which kills the uugetty ( I only have one uugetty running) and respawns. >Obviously, this is a kludge. > >Does anyone know how cu and uucp respawn the uugetty properly, so I may add >this to the pcomm source. (BTW, is there a more recent pcomm than 1.2). I don't know about AT&T Unix 3.1 I do know about SCO, and I just posted this message in CompuServe's UnixForum: In SCO Xenix 2.2, there is an ungetty program which can be called to suspend a running getty; ungetty -r will then restart getty. In SCO Xenix 2.3, the getty is like the uugetty in SCO Unix. For either of those, sending the getty/uugetty SIGUSR1 will suspend it, and then sending it a SIGUSR2 will zap it and cause a new getty/uugetty to start up. This is the scheme implemented in xcmalt's xcport.c Jean-Pierre Radley Unix in NYC jpr@jpr.com jpradley!jpr CIS: 72160,1341