public inbox for gcc-prs@sourceware.org
help / color / mirror / Atom feed
* c++/8797: g++ accepts in-class initialization of non-integral static member
@ 2002-12-03  7:46 bangerth
  0 siblings, 0 replies; 2+ messages in thread
From: bangerth @ 2002-12-03  7:46 UTC (permalink / raw)
  To: gcc-gnats


>Number:         8797
>Category:       c++
>Synopsis:       g++ accepts in-class initialization of non-integral static member
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          accepts-illegal
>Submitter-Id:   net
>Arrival-Date:   Tue Dec 03 07:46:02 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Wolfgang Bangerth
>Release:        unknown-1.0
>Organization:
>Environment:
2.95 through 3.3CVS, Linux (but front end matter anyway)
>Description:
Based on the discussion on PR 8778 (to be found in the audit
trail of that report):

gcc accepts this although the standard says that only 
integral _scalars_ can be initialized in-class:
-----------------------------
struct A {
    static const int i[] = { 1,2 };
};
-----------------------------
I _do_ get a warning if I use -pedantic. However, I maintain
that I should get an unconditional _error_ instead.

On the other hand, if this is a gcc extension, it should
- be documented in the manual
- generate an _error_ on -pedantic -ansi


To make things a little worse: templatizing the class like
---------------------------------------
template <int> struct A {
    static const int i[] = { 1,2 };
};

template struct A<1>;
---------------------------------------
yields the following, slightly convoluted error message
stating the same content twice in different words:

tmp/g> /home/bangerth/bin/gcc-3.3-pre/bin/c++ -pedantic -ansi -c x.cc
x.cc:2: warning: ISO C++ forbids initialization of member constant `i' of
   non-integral type `const int[]'
x.cc: In instantiation of `A<1>':
x.cc:5:   instantiated from here
x.cc:2: warning: ISO C++ forbids initialization of member constant `A<1>::i' of
   non-integral type `const int[2]'
>How-To-Repeat:

>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: c++/8797: g++ accepts in-class initialization of non-integral static member
@ 2003-01-13 23:26 reichelt
  0 siblings, 0 replies; 2+ messages in thread
From: reichelt @ 2003-01-13 23:26 UTC (permalink / raw)
  To: bangerth, gcc-bugs, gcc-prs, nobody

Synopsis: g++ accepts in-class initialization of non-integral static member

State-Changed-From-To: open->closed
State-Changed-By: reichelt
State-Changed-When: Mon Jan 13 15:26:30 2003
State-Changed-Why:
    Fixed with the new parser.
    We now get the following error message (even without -pedantic):
    
    bug.cc:2: error: a brace-enclosed initializer is not allowed here
    bug.cc:2: error: invalid in-class initialization of static data member of 
       non-integral type `const int[]'

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=8797


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2003-01-13 23:26 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2002-12-03  7:46 c++/8797: g++ accepts in-class initialization of non-integral static member bangerth
2003-01-13 23:26 reichelt

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).