Path: utzoo!attcan!uunet!tut.cis.ohio-state.edu!ucsd!usc!cs.utexas.edu!rutgers!orstcs!jacobs.cs.orst.edu!keffert
From: keffert@jacobs.cs.orst.edu (Thomas Keffer)
Newsgroups: comp.lang.c++
Subject: Does "const" mean "const"??
Message-ID: <19078@orstcs.CS.ORST.EDU>
Date: 26 Jun 90 23:33:22 GMT
Sender: usenet@orstcs.CS.ORST.EDU
Reply-To: keffert@jacobs.cs.orst.edu (Thomas Keffer)
Organization: Oregon State University - CS - Corvallis Oregon
Lines: 27

Consider the following:

void foo(const int* j)
{
  *(int*)j = 5;	// Ignore the "constness" of j
}

main()
{
  i = 0;
  foo(&i);
  printf("i = %d\n");	// What's the value of i?
}

The question is: Can an aggressive optimizing compiler assume that
a variable passed to a function as a const pointer will remain 
unchanged after invokation of the function?  Or, if you prefer
a more philosophical pitch: is "const" a documentation aid, or
a hint to the compiler?

-tk
___
Thomas Keffer
Rogue Wave Assoc.
P.O. Box 2328
Corvallis, OR 97339
(503) 745-5908