[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu dot org]

------- Comment #2 from redi at gcc dot gnu dot org  2005-10-02 16:25 -------
P.S. in case it isn't obvious, the last paragraph should say "... F::f ..."
rather than "... F::F ..."


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[mmitchel at gcc dot gnu dot org]

-- 

mmitchel at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |mark at codesourcery dot com
                   |dot org                     |
             Status|NEW                         |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[mmitchel at gcc dot gnu dot org]

------- Comment #3 from mmitchel at gcc dot gnu dot org  2005-10-03 00:16 -------
Until DR228 the name following ".template" had to be a member template, so the
code would have been invalid on those grounds.

However, even after DR228, I don't think this is valid code.

The name lookup done after ".template" must include templates in the global
scope; DR 228 makes that clear.

EDG 3.6 rejects the program in its strict mode, as well, giving two messages:

"pr24161.cpp", line 7: error: name following "template" must be a member
          template
     t.template f< int >();
                ^

"pr24161.cpp", line 7: error: too few arguments for class template "f"
     t.template f< int >();
                       ^


-- 

mmitchel at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu dot org]

------- Comment #4 from redi at gcc dot gnu dot org  2005-10-03 16:54 -------
> Until DR228 the name following ".template" had to be a member template, so the
> code would have been invalid on those grounds.

Erm, the name following .template *is* a member template, that's my whole
point!  :-)

I don't understand why DR228 applies, that is about whether non-member
templates can follow the .template syntax, not about whether 3.4.5 para 2
applies to templates.  Additionally, I believe the apparently correct behaviour
that resulted in my bug report makes the use of "template" even harder to teach
- something that was apparently a motivation for DR228.  It also means that the
template function "f" is dependent on include ordering and whether a global "f"
has been declared.

However, I'll take the question to a more appropriate forum and try to get an
explanation there, rather than banging on about it here.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu dot org]

------- Comment #5 from redi at gcc dot gnu dot org  2005-10-03 16:56 -------
Oops, sorry, I meant:

template function "call_f" is dependent on include ordering and whether a
global "f" has been declared.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[pinskia at gcc dot gnu dot org]

------- Comment #6 from pinskia at gcc dot gnu dot org  2005-10-03 17:06 -------
For what is worth Comeau rejects the code:


"ComeauTest.c", line 8: error: name following "template" must be a member
template
     t.template f< int >();
                ^

"ComeauTest.c", line 8: error: too few arguments for class template "f"
     t.template f< int >();
                       ^

"ComeauTest.c", line 8: warning: ambiguous class member reference -- function
          template "F::f" (declared at line 14) used in preference to class
          template "f" (declared at line 2)
     t.template f< int >();
                ^
          detected during instantiation of "void call_f(T &) [with T=F]" 

Removing the template class above does "fix" it too there.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[mark at codesourcery dot com]

------- Comment #7 from mark at codesourcery dot com  2005-10-03 17:08 -------
Subject: Re:  [3.4/4.0/4.1 Regression] Lookup of template member
 function finds global type.

redi at gcc dot gnu dot org wrote:
> ------- Comment #4 from redi at gcc dot gnu dot org  2005-10-03 16:54 -------
> 
>>Until DR228 the name following ".template" had to be a member template, so the
>>code would have been invalid on those grounds.
> 
> 
> Erm, the name following .template *is* a member template, that's my whole
> point!  :-)

:-)

The problem is that we can't do the lookup, because the expression
before the "." is dependent.  But, we still have to do global-scope
lookup, because "t.template S<T>::f()" would be valid, if "S" were a
global template, and redundant "template" keywords are specifically allowed.

I do see your point, though.  It's possible that the lookup after
"t.template" should be restricted so that if "t" is dependent, then no
lookup is performed.

> However, I'll take the question to a more appropriate forum and try to get an
> explanation there, rather than banging on about it here.

If you do get a DR opened about this, please let me know!


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu dot org]

------- Comment #8 from redi at gcc dot gnu dot org  2005-10-03 17:54 -------
Aah, gotcha.  I now see why the global scope has to be considered, and why
DR228 is relevant.  Thanks!  I still think it's of dubious value having the
standard defined like that, so might take it to the UK panel and see if anyone
there cares about it.

Andrew, another interesting variation is to add two template arguments to the
function call.  Then you don't get the "wrong number of template arguments" you
get one saying "invalid use of struct f", which is because it's not valid to
refer to a typename there - but then everyone except the compiler already knew
that!  ;-)


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[anthony dot ajw at gmail dot com]

------- Comment #9 from anthony dot ajw at gmail dot com  2005-10-04 21:43 -------
The name following "t." must be the name of a member of either T, or one of its
base classes.

t.template f<int> can only refer to a global template f if it is then followed
by a nested name specifier (e.g. t.template f<int>::X) so that it is a
qualified name for member of T.

This is just the same as t.f always refers to a member, even if there is a
non-member f in scope, unless one writes t.f::X (for some X).

t.template f<int>() can thus only refer to a member template, and the global
template should not be considered.


-- 

anthony dot ajw at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |anthony dot ajw at gmail dot
                   |                            |com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu dot org]

------- Comment #10 from redi at gcc dot gnu dot org  2005-10-05 21:16 -------
Yes, I agree with Anthony, DR228 is about nested-name-specifiers, this bug is
about an unqualified-id, so 3.4.5p2 should still apply.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

--- Comment #13 from Jonathan Wakely <redi at gcc dot gnu.org> ---
EDG gives the same error as GCC

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[cameron314 at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

Cameron <cameron314 at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |cameron314 at gmail dot com

--- Comment #12 from Cameron <cameron314 at gmail dot com> ---
VS2013 accepts this too.

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|                            |

--- Comment #11 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-03-01 01:32:26 UTC ---
FWIW clang++ accepts this

[Bug c++/24161] [3.4/4.0/4.1 Regression] Lookup of template member function finds global type.

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-10-01 15:49 -------
Confirmed, 3.3 accepted the code.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
      Known to fail|                            |4.0.0 4.1.0 3.4.0
      Known to work|                            |3.3.3
   Last reconfirmed|0000-00-00 00:00:00         |2005-10-01 15:49:01
               date|                            |
            Summary|Lookup of template member   |[3.4/4.0/4.1 Regression]
                   |function finds global type. |Lookup of template member
                   |                            |function finds global type.
   Target Milestone|---                         |4.0.3


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24161