[m-dev.] Re: Things seg faulting everywhere

[Simon Taylor]

> Andrew Bromage <bromage at cs.mu.OZ.AU> writes:
> 
> >However, diff still doesn't run when compiled on hydra, mercury and
> >kryten.  I've checked it in, since it works on the alphas.
> 
> The problem seems to be in getopt. It is seg faulting in the call to
> getopt__process_options.

No, the problem is the call to `copy/2' in globals.m - Tyson reported
this bug a couple of weeks ago.

Simon.


Estimated hours taken: 0.25

samples/diff/globals.m:
	Convert a call to copy to unsafe_promise_unique, since copy
	is buggy at the moment.


Index: globals.m
===================================================================
RCS file: /home/staff/zs/imp/mercury/samples/diff/globals.m,v
retrieving revision 1.1
diff -u -t -u -r1.1 globals.m
--- 1.1	1998/01/13 00:52:03
+++ globals.m	1998/09/21 06:38:42
@@ -168,7 +168,7 @@
         }.
 
 globals__io_set_globals(Globals) -->
-        { copy(Globals, UniqGlobals) },
+        { unsafe_promise_unique(Globals, UniqGlobals) },
         { type_to_univ(UniqGlobals, UnivGlobals) },
         io__set_globals(UnivGlobals).