Re: Q. about POD with const member (or g++ vs. Comeau)

[Andrew Haley <aph at redhat dot com>]

Stefan Naewe wrote:
> On 6/27/2008 12:34 PM, Andrew Haley wrote:
>> Stefan Naewe wrote:
>>> Given the following:
>>>
>>>
>>> // ++ CODE
>>> struct S
>>> {
>>>   const int ci;
>>> };
>>>
>>> int main()
>>> {
>>>   S  * s = new S;
>>> }
>>> // -- CODE
>>>
>>>
>>> Comeau Online gives me:
>>>
>>> // ++ output
>>> Comeau C/C++ 4.3.10.1 (May 29 2008 09:37:15) for ONLINE_EVALUATION_BETA1
>>> Copyright 1988-2008 Comeau Computing.  All rights reserved.
>>> MODE:strict errors C++ C++0x_extensions
>>>
>>> "ComeauTest.c", line 1: warning: class "S" defines no constructor to
>>> initialize the
>>>           following:
>>>             const member "S::ci"
>>>   struct S
>>>          ^
>>>
>>> "ComeauTest.c", line 8: error: object has an uninitialized const or
>>> reference member
>>>     S  * s = new S;
>>>                  ^
>>>
>>> 1 error detected in the compilation of "ComeauTest.c".
>>>
>>> // -- output
>>>
>>>
>>> Different g++ versions
>>> 4.2.3
>>> 4.1.2 20061115 (prerelease) (Debian 4.1.1-21)
>>> gcc version 4.3.1 20080404 (prerelease) [gcc-4_3-branch revision 133917]
>>>         (Ubuntu 20080404-0ubuntu1)
>>>
>>> give:
>>>
>>> // ++ output
>>> constmember.cc: In function 'int main()':
>>> constmember.cc:8: warning: unused variable 's'
>>> // -- output
>>>
>>>
>>> Reading 5.3.4.15 (and 8.5.5) of the holy standard
>>> I'd say Comeau is right, the program is ill-formed, isn't it?
>> Yes.
> 
> 'Yes'
>   as in
> 'Yes, the program is ill formed

> or
> 
> 'Yes'
>   as in
> 'Yes, the program is ill formed and there's a bug in GCC'
> 
> ??

It's ill-formed.  I think gcc should issue a diagnostic, so yes, GCC has
a bug, but I am uncertain.

Andrew.