[2133] in Release_7.7_team
acroread printing bug
daemon@ATHENA.MIT.EDU (Thomas Bushnell, BSG)
Mon Feb 14 14:34:02 2000
Date: Mon, 14 Feb 2000 14:33:50 -0500 (EST)
Message-Id: <200002141933.OAA337970@pusey.mit.edu>
From: tb@MIT.EDU (Thomas Bushnell, BSG)
To: release-team@mit.edu, source-developers@mit.edu
As I sent in a previous message, there is a bug in lprng's lpr which
gets triggered by acroread, causing acroread's printing feature to
spin forever on Linux. This has two effects right now:
1) Printing from acroread doesn't work, and once you try it, your
acroread loses and you will need to restart it.
2) The spinning lpr process hangs around even after you log out, and
because it's setuid root, cleanup doesn't clean it. If users su
and kill these processes, they will get their CPU back.
We have a patch for cleanup to fix problem (2) which will go out
probably in Linux 8.3.27.
I could hack lprng to avoid the proximate problem with the spin, but
from the bug, I expect that will not make printing work, it will just
fake it fail less spectacularly.
So the consulting people should be notified in some way about the
problem to avoid persistent user confusion.
It would be really nice if someone other than me who understands lprng
and/or acroread could work on debugging the underlying problem; I've
already got plenty to do.
Thomas
