🚀 go-pugleaf

RetroBBS NetNews Server

Inspired by RockSolid Light RIP Retro Guy

Article View: mod.computers.apollo
Article #395

Submission for mod-computers-apollo

#395
From: i91 Fons Rademakers
Date: Fri, 06 Mar 1987 16:23
92 lines
2968 bytes
Path: nikhefh!i91
From: i91@nikhefh.UUCP (Fons Rademakers)
Newsgroups: mod.computers.apollo
Subject: DOMAIN/IX Unix and inlib -> crash!!!
Message-ID: <257@nikhefh.UUCP>
Date: 6 Mar 87 16:23:41 GMT
Reply-To: i91@nikhefh.UUCP (Fons Rademakers)
Distribution: world
Organization: Nikhef-H, Amsterdam (the Netherlands).
Lines: 81

Could somebody (from Apollo) tell me why the following
crashes Apollo's Unix (csh and Bourne):

Compile these two little routines:
==========================================
      subroutine c1
      common // b(5)
               
      print *, 'c1: b(1)= ', b(1)
      print *, 'c1: b(5)= ', b(5)

      call c2

      return
      end

      subroutine c2
      common // c(5)

      print *, 'c2: c(1)= ', c(1)
      print *, 'c2: c(5)= ', c(5)

      c(1) = 111
      c(5) = 111

      return
      end
==========================================

Compile this little main program:
==========================================
*
*-- main program to intialize static data
*
      common // b(5)

      print *, 'blank common initialized'

      stop 
      end
==========================================

Make an installed library from these three
routines:
==========================================
#! /bin/sh

bind -b foo.inlib -<<\!
main.bin
-looks -all
-marks -all
-allmark
c1.bin
!
==========================================

Install the library (in DOMAIN/IX Unix):
==========================================
inlib foo.inlib
==========================================

And flash, super crash.

Note, in the Aegis (/com/sh) this works fine.
BTW, the contents of the  two little routines
does not matter. The only problem is the main program that 
initializes the static data. Remove the main program 
(and change the routine c1 so it uses a parameter to get
its value for b) and everything works well. 

I am running SR9.2.4 and inprocess is set in the csh.

Is this a BUG or a feature???

Cheers,
    Fons Rademakers
-- 
Org:    NIKHEF-H, National Institute for Nuclear and High-Energy Physics.
Mail:   Kruislaan 409, P.O. Box 41882, 1009 DB Amsterdam, the Netherlands
BITNET: U00029@HASARA5.BITNET                       Telex: 10262 (hef nl)
UUCP:   {decvax,cernvax,uck,unido,vmucnam,seismo}!mcvax!nikhefh!i91
Technical Details
Message-ID: <8703061623.AA12542@nikhefh.uucp>
Path: arc1-ca.pugleaf.net!utzoo!mnetor!seismo!lll-lcc!mordor!sri-spam!ames!ucbcad!ucbvax!i91@seismo.CSS.GOV@nikhefh.UUCP

Show Full Headers
Complete Article Headers
Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: arc1-ca.pugleaf.net!utzoo!mnetor!seismo!lll-lcc!mordor!sri-spam!ames!ucbcad!ucbvax!i91@seismo.CSS.GOV@nikhefh.UUCP
From: i91@seismo.CSS.GOV@nikhefh.UUCP (Fons Rademakers)
Newsgroups: mod.computers.apollo
Subject: Submission for mod-computers-apollo
Message-ID: <8703061623.AA12542@nikhefh.uucp>
Date: Fri, 6-Mar-87 11:23:42 EST
Article-I.D.: nikhefh.8703061623.AA12542
Posted: Fri Mar  6 11:23:42 1987
Date-Received: Sun, 8-Mar-87 10:46:59 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 92
Approved: apollo@yale-comix.arpa