another 1.10 job control bug?

["Nicholas S Castellano" <entropy@terminator.rs.itd.umich.edu>]

>
> After having been on holiday last week I've digged myself through the pile
>of patches that arose in my mailbox yesterday, namely those of SRU, NOX and
>AS. I sort of remember there were also some fixes dealing with job control.
>Something seems to have gone wrong with them, cause now I'm only able to log
>in once at the console, every following getty gets stopped after some:
>
>pid  91 (getty): Fcntl(0, cmd=0x5401)
>pid  91 (getty): Fcntl mode 5401: calling ioctl
>pid  91 (getty): job control: tty pgrp is 16 proc pgrp is 91
>pid  91 (getty): killgroup 91 22
>
> This did not happen in the "clean" version of 1.10, and it does not happen
>on any vconsole (2 Juergen: looks like they're really good 4 something... :-)

In one of the patches sent out by nox, there was a 'fix' I sent out a
few weeks ago for a problem someone described in tty.c.  I think the
'fix' is actually worse than the problem, and that code should *not*
be disabled.

The code in question is to release the controlling tty (by setting
it's pgrp to 0) when the process group leader attached to that process
exits.

Try undoing that patch and see if it helps.  I generally run with the
TOSTOP bit turned off except when testing job control, so I hadn't
encountered exactly the problem you mentioned here.

--
entropy -- it's not just a good idea, it's the second law.
Personal mail:      entropy@gnu.ai.mit.edu
MiNT library mail:  entropy@terminator.rs.itd.umich.edu
"what do you have against octal?" -jrb