Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!burl!codas!killer!billw
From: billw@killer.UUCP
Newsgroups: comp.sources.bugs
Subject: Berkeley fortune program on SysV
Message-ID: <2048@killer.UUCP>
Date: Mon, 9-Nov-87 17:38:39 EST
Article-I.D.: killer.2048
Posted: Mon Nov  9 17:38:39 1987
Date-Received: Thu, 12-Nov-87 03:44:30 EST
Reply-To: billw@killer.UUCP (Bill Wisner)
Distribution: world
Organization: HASA
Lines: 71

[ You can feed this entire article to patch ... ]

After a long, long delay I finally decided to get the Berkeley fortune program
from volume 1 of comp.sources.games running on killer. The program claimed
to run fine on SysV with appropriate defines, but it didn't quite work out
that way. The fixes are, of course, included in this article.

First of all, it seems that due to a doubled #include, <sys/types.h> is scanned
twice and all the typedefs repeated. Our compiler didn't like that. Quickie fix
is to remove the #include <sys/types.h> from strfile.h. Thanks to Steve Parker
for catching this one.

*** orig/strfile.h	Mon Nov  9 15:19:54 1987
--- strfile.h	Sun Nov  8 12:35:04 1987
***************
*** 3,10
  
  # define	__STRFILE__
  
- # include	<sys/types.h>
- 
  # define	MAXDELIMS	3
  
  /*

--- 3,8 -----
  
  # define	__STRFILE__
  
  # define	MAXDELIMS	3
  
  /*

So, I have the fortune program compiled and the fortunes database properly
assembled and installed, so I eagerly type fortune to see what gem of wisdom
I will get. Well, it wasn't quite what I expected:

Floating exception - core dumped

After a bit of mucking about with debuggers and the like, it became apparent
that due to inadequate variable checks, an operation identical to "0 % 0" was
being performed. That's a no-no.

Again, the fix is simple.

*** orig/fortune.c	Mon Nov  9 15:19:28 1987
--- fortune.c	Mon Nov  9 15:26:33 1987
***************
*** 312,318
  	 * Make sure all values are in range.
  	 */
  
! 	if (Tbl.str_delims[1] >= Tbl.str_delims[0])
  		Tbl.str_delims[1] %= Tbl.str_delims[0];
  	if (Tbl.str_delims[2] >= Tbl.str_numstr) {
  		Tbl.str_delims[2] -= Tbl.str_delims[0];

--- 312,319 -----
  	 * Make sure all values are in range.
  	 */
  
! 	if ((Tbl.str_delims[1] >= Tbl.str_delims[0]) &&
! 	((Tbl.str_delims[1] != 0) && (Tbl.str_delims[0] != 0)))
  		Tbl.str_delims[1] %= Tbl.str_delims[0];
  	if (Tbl.str_delims[2] >= Tbl.str_numstr) {
  		Tbl.str_delims[2] -= Tbl.str_delims[0];

Enjoy - I do.
-- 
Bill Wisner, HASA "A" Division		..{codas,ihnp4}!killer!billw
"I don't mind at all.." -- Bourgeois Tagg