public inbox for gcc-prs@sourceware.org
help / color / mirror / Atom feed
* c++/3136: Misleading error message / Bug?
@ 2001-06-12  6:16 philipp.bachmann
  0 siblings, 0 replies; 4+ messages in thread
From: philipp.bachmann @ 2001-06-12  6:16 UTC (permalink / raw)
  To: gcc-gnats

>Number:         3136
>Category:       c++
>Synopsis:       Misleading error message / Bug?
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Jun 12 06:16:01 PDT 2001
>Closed-Date:
>Last-Modified:
>Originator:     Philipp Bachmann
>Release:        g++ 2.95.3
>Organization:
>Environment:
Sun Solaris 7 SPARC
>Description:
When "CONST" in the sample code provided is defined
to be empty, then the following error message is
printed by g++:
copyConstr.c++: In function 'class derived_t whatHappensHere(base_t)':
copyConstr.c++:74: conversion from 'base_t' to non-scalar type 'derived_t' requested.

If either
(1) "CONST" is defined as "const" or
(2) line 74 is commented out and line 73 is commented in,
then the error message completely
vanishes.

I can understand the necessity for "const" (i.e. Fix (1))
- but that Fix (2) solves the problem I'm surprised about.

So my interpretation is twofold: The error message is misleading
and g++ has a small bug.
>How-To-Repeat:
via Email
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: text/plain; name="copyConstr.c++.txt"
Content-Disposition: inline; filename="copyConstr.c++.txt"

/* Very simple program to demonstrate misleading error message
   of GNU g++ 2.95.3. */

/* $Log: copyconstr.c++,v $
 * Revision 1.2  2001/06/12 13:58:07  bachlipp
 * Introduced "base_t::getMe() const".
 *
 * Revision 1.1  2001/06/12 13:32:39  bachlipp
 * Initial revision
 * */

static const char rcsid[]="@@(#)$Id: copyconstr.c++,v 1.2 2001/06/12 13:58:07 bachlipp Exp $";

#include<iostream>

#define CONST

// #define CONST const

class base_t {
public :
  base_t(void);
  base_t(const base_t &);
  virtual ~base_t(void);
  virtual base_t getMe(void) const;
};

class derived_t : public base_t {
public:
  derived_t(void);
  derived_t(const derived_t &);
  derived_t(CONST base_t &);
  ~derived_t(void);
};

derived_t whatHappensHere(base_t);

base_t::base_t(void) {
  std::cerr<<"\"base_t::Constructor\" called."<<std::endl;
}

base_t::base_t(const base_t &org) {
  std::cerr<<"\"base_t::Copyconstructor\" called."<<std::endl;
}

base_t::~base_t(void) {
  std::cerr<<"\"base_t::Destructor\" called."<<std::endl;
}

base_t base_t::getMe(void) const {
  std::cerr<<"\"base_t::getMe\" called."<<std::endl;
  return *this;
}

derived_t::derived_t(void) : base_t() {
  std::cerr<<"\"derived_t::Constructor\" called."<<std::endl;
}

derived_t::derived_t(const derived_t &org) : base_t(org) {
  std::cerr<<"\"derived_t::Copyconstructor\" called."<<std::endl;
}

derived_t::derived_t(CONST base_t &org) : base_t(org) {
  std::cerr<<"\"derived_t::PerverseCopyconstructor\" called."<<std::endl;
}

derived_t::~derived_t(void) {
  std::cerr<<"\"derived_t::Destructor\" called."<<std::endl;
}

derived_t whatHappensHere(base_t self) {
  std::cerr<<"\"whatHappensHere()\" started."<<std::endl;
  // return self;
  return self.getMe();
}

int main() {
  derived_t der;
  whatHappensHere(der);
  std::cerr<<"\"main()\": done."<<std::endl;
  return 0;
}


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: c++/3136: Misleading error message / Bug?
@ 2001-06-15 16:35 rodrigc
  0 siblings, 0 replies; 4+ messages in thread
From: rodrigc @ 2001-06-15 16:35 UTC (permalink / raw)
  To: gcc-bugs, gcc-prs, nobody, philipp.bachmann, rodrigc

Synopsis: Misleading error message / Bug?

State-Changed-From-To: open->closed
State-Changed-By: rodrigc
State-Changed-When: Fri Jun 15 16:35:57 2001
State-Changed-Why:
    gcc is not in error.  There is a bug in your code.
    The error message may be misleading though.
    On line 74, of your test case:
         return self.getMe();
    
    The compiler creates a temporary variable
    to store the result of getMe(), since self
    is not a const reference.
    What is happening is that on line 74:
    return self.getMe();
    
    When you return this temporary value, you are passing
    it to a copy constructor which takes derived_t&
    (without a const) as value.  This is not allowed in C++.
    
    Your options are to make self a const parameter to the 
    whatHappensHere() function.
    
    This is similar to a bug report described here:
    http://gcc.gnu.org/ml/gcc/2001-06/msg00765.html

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3136&database=gcc


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: c++/3136: Misleading error message / Bug?
@ 2001-06-15 16:20 rodrigc
  0 siblings, 0 replies; 4+ messages in thread
From: rodrigc @ 2001-06-15 16:20 UTC (permalink / raw)
  To: gcc-bugs, gcc-prs, nobody, philipp.bachmann, rodrigc

Synopsis: Misleading error message / Bug?

State-Changed-From-To: feedback->open
State-Changed-By: rodrigc
State-Changed-When: Fri Jun 15 16:20:20 2001
State-Changed-Why:
    Received test case.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3136&database=gcc


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: c++/3136: Misleading error message / Bug?
@ 2001-06-14 18:55 rodrigc
  0 siblings, 0 replies; 4+ messages in thread
From: rodrigc @ 2001-06-14 18:55 UTC (permalink / raw)
  To: gcc-bugs, gcc-prs, nobody, philipp.bachmann, rodrigc

Synopsis: Misleading error message / Bug?

State-Changed-From-To: open->feedback
State-Changed-By: rodrigc
State-Changed-When: Thu Jun 14 18:55:24 2001
State-Changed-Why:
    Your test case is empty and only contains one line
    with a comment.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view&pr=3136&database=gcc


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2001-06-15 16:35 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-06-12  6:16 c++/3136: Misleading error message / Bug? philipp.bachmann
2001-06-14 18:55 rodrigc
2001-06-15 16:20 rodrigc
2001-06-15 16:35 rodrigc

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).