[3438] in Athena Bugs

home help back first fref pref prev next nref lref last post

Re: strange "crash"

daemon@ATHENA.MIT.EDU (probe@ATHENA.MIT.EDU)
Wed Oct 18 00:27:41 1989

From: probe@ATHENA.MIT.EDU
Date: Wed, 18 Oct 89 00:27:20 -0400
To: saeed@ATHENA.MIT.EDU
Cc: bugs@ATHENA.MIT.EDU, raeburn@ATHENA.MIT.EDU, beth@ATHENA.MIT.EDU,
In-Reply-To: 's message of Tue, 17 Oct 89 15:38:25 -0400,
Reply-To: Richard Basch   <probe@ATHENA.MIT.EDU>

  From: saeed@ATHENA.MIT.EDU
  Date: Tue, 17 Oct 89 15:38:25 -0400


  For the past two weeks I have been having the exact same problem that
  the user lishu has been having.  I am pretty positive it is not a
  panic or some kind of automatic reboot.  My station does dump core
  either.

  Basically what happens is that when I login under my customized files
  (I have a .path, .environment, .cshrc.mine, .startup.X), everything
  goes fine until the machine (X server) tries to put up my windows.  It
  usually gets about halfway through my .startup.X file and then
  suddenly, my cursor freezes for about 3 seconds, then the screen goes
  absolutely blank.  About 5 seconds later the toehold prompt comes up:
  "Hit any key to start."

  I thought this happened because of certain workstations, but not at
  all.  I have not been able to log onto ANY workstation with my files.
  For the past two weeks I have been logging under System.  Even PVAXen
  do the same thing with me.

  Yes, yes, I have reported this to bugs and hotline.  I have received 3
  responses.  Two helpful responses and one telling me to mind my own
  business, use my brain to fix the problem myself, and not to tell bugs
  about my problem.  I have tried to follow up on the two helpful ones,
  but have been pretty slow at it myself.  I apologize for this -- I
  just do not have the time.  Ken Raeburn has offered me help, and when
  I get the chance I will go over to his office and display my problem.
  I know this is a serious bug we have here (now that another user has
  confirmed it, I know it might not be just my dotfile configuration),
  but I have been very wrapped up in academics recently.

  Now the logistics of the problem exactly:

  <i am logging into the PVAX COPILOT.MIT.EDU to try this>

  i login.
  sipb mounted
  athenadoc mounted
  consult mounted
  bjaspan mounted

  "Starting Mwm . . "
  "Running Standard Startup Activities . . "

  [here the cursor changes]

  .Xresources sourced (i know this cause i have an somthing that tells
                  me this)

  [Mwm starts -- border change noticeable on console window]

  "Running standard startup activities . . "

  "Running custom startup activites listed in ~/.startup.X"

  [1]
  [2]
  [3]
  [4]
  [5]
  [6]
  [7]
  [8]
  [9]
  [10]

  (notice I still have a few more background processes to go in my
  .startup.X file)

  [windows start popping up: xterms (3), emacs (1), etc.]
  [moving cursor to make sure it works -- suddenly it freezes completely
  -- really i mean this]
  [screen blanks out]
  [toehold prompt comes up]

  Notice -- no error messages were sent to the console window -- it just
  ended me.  But that's the scary part -- it really didn't.

  I relogin under control-p.  I "ps aux" and get that my session_gate
  and Xsession are still active (if that's the terminaology).  It did
  not end my session.

  I'm cc'ing this to bugs, raeburn, beth, carla, 

  Saeed


  Another thing:  Copilot has been noted to be crashing -- but this DOES
  happen on EVERY machine (even on RT's) for me -- not just copilot.

  If anybody wishes me to demonstrate, I'll be glad to show them.  


  Saeed Jaffer
  Project Athena Consultant
  saeed@athena.mit.edu
  Baker House 105
  225-7105 (x5-7105)


First guess... look at /core... I bet it is a new file and that it is a
core dump of the X server.  There are several known problems with R3,
and it would be interesting to see if this is one of them.
Unfortunately, I do not have solutions to any of these problems, and I
do know that most of them will be fixed when R4 comes out (but we can't
take the code from the R4 server, because many of the internal
structures and argument passing changed).

-Richard

home help back first fref pref prev next nref lref last post