[Bug c++/23263] New: Fails to identify member definition

[Bug c++/23263] New: Fails to identify member definition

[igodard at pacbell dot net]

template<typename T> struct foo{ static T* p; };
template<typename T> T* foo<T>::p = 0;

template<typename T> struct A {};
template<typename T, template<typename> class C>
    struct bar{ static C<void>* p; };
template<typename T> A<void>* bar<T,A>::p = 0;

int main() {
    foo<int> f;
    bar<int, A> b;
    return 0;
    }

gets you:
~/ootbc/members/src$ g++ foo.cc
foo.cc:7: error: invalid use of undefined type `struct bar<T, A>'
foo.cc:6: error: declaration of `struct bar<T, A>'

As it works for the plain type case (foo), it looks like something's wrong in
the binding of template template parameters (bar).

Ivan

-- 
           Summary: Fails to identify member definition
           Product: gcc
           Version: 3.4.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: igodard at pacbell dot net
                CC: gcc-bugs at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Fails to identify member definition

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2005-08-06 20:29 -------
No, you try to partially specialize a member. That isn't allowed.  
You either have to completely specialize all template parameters, or  
none. Long standing annoyance in C++...  
  
W.  

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Fails to identify member definition

[igodard at pacbell dot net]

------- Additional Comments From igodard at pacbell dot net  2005-08-06 20:40 -------
Reopened as a complaint about the diagnostic. The naive (or even reasonably
sophisticated) user cannot figure out "illegal partial specialization" from that
message.

Comeau gives:

"ComeauTest.c", line 7: error: template argument list must match the parameter list
  template<typename T> A<void>* bar<T,A>::p = 0;
                                ^

which I admit is only a little better.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |UNCONFIRMED
         Resolution|INVALID                     |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Fails to identify member definition

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |enhancement
           Keywords|                            |diagnostic


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Incomprehensible message for invalid attempt to partially specialize a member

[bangerth at dealii dot org]

------- Additional Comments From bangerth at dealii dot org  2005-08-06 22:56 -------
If you know what's going on, the error message is actually clear:  
  foo.cc:7: error: invalid use of undefined type `struct bar<T, A>'  
This just means that the partially specialized class bar<T,A> has not  
yet been declared, and that you can't therefore define a member of it.  
  
But I grant that we could do better. For that purpose, this example 
may be better: 
--------------------- 
template <int A, int B> struct X { void f(); }; 
template <int A> void X<A,1>::f() {} 
--------------------- 
We presently get this: 
g/x> /home/bangerth/bin/gcc-4.1-pre/bin/c++ -c x.cc 
x.cc:2: error: invalid use of undefined type &#8216;struct X<A, 1>&#8217; 
x.cc:1: error: declaration of &#8216;struct X<A, 1>&#8217; 
x.cc:2: error: template definition of non-template &#8216;void X<A, 1>::f()&#8217; 
  
W.  

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-08-06 22:56:41
               date|                            |
            Summary|Fails to identify member    |Incomprehensible message for
                   |definition                  |invalid attempt to partially
                   |                            |specialize a member


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Incomprehensible message for invalid attempt to partially specialize a member

[igodard at pacbell dot net]

------- Additional Comments From igodard at pacbell dot net  2005-08-08 02:42 -------
See also #23281

Should this and other reports about diagnostic quality be treated as
"enhancement requests"? I suppose tht it depends on what you consider
"correctness". If that means "conforming to standard", then any diagnostic will
do, including "something wrong found somewhere in program". Personally I use a
different standard of "correct", which includes "usable", and so addresses
issues like compilation time measured in days as well as quality of diagnostic.

I understand that the gcc implementation community has a notion of "QOI"
(quality of implementation) which reflects much of what I call "usable". But I
would distinguish a "defect of QOI" from an "enhancement request"; the latter
being something nice to have, but the former impacting real usage.

I put this particular DR in the "defect of QOI" category. Even if you do not,
you probably can select a non-empty set of DRs that are truly "defects" and not
mere "enhancement requests. 

Consequently I'm filing this as a DR against the gcc DR reporting machinery
itself, rather than against the compiler in particular. There needs to be
categories for QOI defects of varying severity; either that, or complaints about
diagnostics and other QOI ussues should not by policy be filed as "enhancement
requests", and forgotten.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263

[Bug c++/23263] Incomprehensible message for invalid attempt to partially specialize a member

[manu at gcc dot gnu dot org]

------- Comment #5 from manu at gcc dot gnu dot org  2007-03-12 17:39 -------
(In reply to comment #4)
> See also PR 23281
[snip]
> Consequently I'm filing this as a DR against the gcc DR reporting machinery
> itself, rather than against the compiler in particular. There needs to be
> categories for QOI defects of varying severity; either that, or complaints about
> diagnostics and other QOI ussues should not by policy be filed as "enhancement
> requests", and forgotten.

It doesn't matter whether this is filled this way or that way. The problem is
that there are more tan 2 thousand bugs opened. Logically, most developers tend
to focus on fixing regressions, ICEs or optimisations rather on precise wording
of correct diagnostics.

We need more contributors in this area, you are welcome to participate!


-- 

manu at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |manu at gcc dot gnu dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23263