Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uunet!snorkelwacker!bloom-beacon!LARRY.MCRCIM.MCGILL.EDU!mouse
From: mouse@LARRY.MCRCIM.MCGILL.EDU (der Mouse)
Newsgroups: comp.windows.x
Subject: server coredumps: followup
Message-ID: <9002120744.AA13566@Larry.McRCIM.McGill.EDU>
Date: 12 Feb 90 07:44:58 GMT
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 30

A while (a week or two) ago I sent a note to xpert to the effect that
the MIT R4 server was dumping core under ill-defined circumstances,
usually when changing keyboard focus from one window to another.
Someone (rws, I think) said I'd have to track it down a bit more.

Well, I have.  And I don't understand it.  From dix/window.c:

static void
MarkWindow(pWin)
    register WindowPtr pWin;
{
....
    val->before.oldAbsCorner.x = pWin->drawable.x;
    val->before.oldAbsCorner.y = pWin->drawable.y;  /* <<<<<< */
    val->before.borderVisible = NullRegion;
....
}

The indicated assignment is causing the corruption.  (I placed calls to
my sanity-checking routine every line throughout MarkWindow, and the
one before that line was happy, but the one after wasn't.)

I've taken three trial crash coredumps since adding my sanity checking,
and they've all been at that same line.  Ideas, anyone?  I can provide
more information if anyone wants it.

					der Mouse

			old: mcgill-vision!mouse
			new: mouse@larry.mcrcim.mcgill.edu