[2.1.1p4] utmp related patches plus unresolved bugs description

Ben Lindstrom mouring at pconline.com
Thu Jul 27 06:10:07 EST 2000

On Wed, 26 Jul 2000, Charles Levert wrote:

> Remaining:
> 	-- At logout, the utmp entry cannot be cleared since ttyslot
> 	cannot find the offset for the terminal since it is the parent
> 	process that calls this stuff and it does not execute in the
> 	context of the terminal.  It works at login since it is then
> 	done by the child process (the one that will eventually exec
> 	the user shell).  The solution involves having the child
> 	process inform the parent process of the ttyslot and of the
> 	fact that a utmp entry was written at login and needs to be
> 	removed.
This is pretty much the same problem I had after getting logins to work
under NeXT.  I could not see a clean way of doing it besides tracking the
orignal ttyslot() output until it was needed to clear the utmp file.  I
ended up just tracking what tty the user was on and wrote a primative
ttyslot(char *) that took in the tty and returned the slot to change.

That patch should help NeXT (I did something like that before, but it
was not as clean nor was finished) resolve it's utmp/wtmp issues.  


More information about the openssh-unix-dev mailing list