[Bug c++/58590] New: [C++11] Hidden typename not ill-formed under SFINAE conditions

[Bug c++/58590] New: [C++11] Hidden typename not ill-formed under SFINAE conditions

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

            Bug ID: 58590
           Summary: [C++11] Hidden typename not ill-formed under SFINAE
                    conditions
           Product: gcc
           Version: 4.9.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: daniel.kruegler at googlemail dot com

gcc 4.9.0 20130922 (experimental) compiled with the flags

-std=c++11 -Wall -pedantic-errors

rejects the following code:

//---------------------
struct S { int member; struct member{}; };

template<class T, class = typename T::member>
auto f(int) -> char(&)[1];

template<class T>
auto f(...) -> char(&)[2];

static_assert(sizeof(f<S>(0)) == 2, "");

int main() {
}
//---------------------

"main.cpp|9|error: static assertion failed: |"

After discussion with Richard Smith:

The presence of 'typename' before 'T::member' does not influence name lookup,
and 'S::member' as a template-argument names the non-static data member, not
the type, by [basic.scope.hiding]p2. The relevant rule in [temp.res]p4.

It seems that GCC uses the elaborated-type-specifier lookup instead here.

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #1 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Not having investigated this issue at all, I doubt that it should be considered
a SFINAE proper issue, that is I think it should be possible to demonstrate it
in an even smaller testcase, not using the usual SFINAE-exploiting structure (+
very minor nit: the definition of main is totally redundant and certainly will
not be in an eventual testcase)

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #2 from Daniel Krügler <daniel.kruegler at googlemail dot com> ---
(In reply to Paolo Carlini from comment #1)
> Not having investigated this issue at all, I doubt that it should be
> considered a SFINAE proper issue,

Well, the actual programmer problem occurred in this context of sfinae
situation.

> that is I think it should be possible to
> demonstrate it in an even smaller testcase, not using the usual
> SFINAE-exploiting structure (+ very minor nit: the definition of main is
> totally redundant and certainly will not be in an eventual testcase)

If you replace the current main definition, gcc accepts:

int main() {
   f<S, typename S::member>(0); // OK
}

which it clearly shouldn't. Especially if you extend to

int main() {
   f<S, typename S::member>(0); // OK
   f<S, S::member>(0); // Error
}

where the second line is rejected (as expected).
>From gcc-bugs-return-430925-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Tue Oct 01 22:47:56 2013
Return-Path: <gcc-bugs-return-430925-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 23499 invoked by alias); 1 Oct 2013 22:47:56 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 23468 invoked by uid 48); 1 Oct 2013 22:47:54 -0000
From: "ppluzhnikov at google dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/58591] GCC emits bad mangling for template function with a parameter pack
Date: Tue, 01 Oct 2013 22:47:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c++
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ppluzhnikov at google dot com
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-58591-4-ZDz33pxGA7@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-58591-4@http.gcc.gnu.org/bugzilla/>
References: <bug-58591-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-10/txt/msg00069.txt.bz2
Content-length: 147

http://gcc.gnu.org/bugzilla/show_bug.cgi?idX591

--- Comment #1 from Paul Pluzhnikov <ppluzhnikov at google dot com> ---
Google ref: b/10860844

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #3 from Paolo Carlini <paolo.carlini at oracle dot com> ---
I still believe this isn't a SFINAE proper issue. Also note that in the
original testcase main was *empty*.

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #4 from Daniel Krügler <daniel.kruegler at googlemail dot com> ---
(In reply to Paolo Carlini from comment #3)
> I still believe this isn't a SFINAE proper issue.

Could you please elaborate why?

> Also note that in the original testcase main was *empty*.

Sure, because I didn't want to see any error (including linker errors).
>From gcc-bugs-return-430929-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Tue Oct 01 23:08:26 2013
Return-Path: <gcc-bugs-return-430929-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 7081 invoked by alias); 1 Oct 2013 23:08:26 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 7038 invoked by uid 48); 1 Oct 2013 23:08:22 -0000
From: "paolo.carlini at oracle dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions
Date: Tue, 01 Oct 2013 23:08:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c++
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords: rejects-valid
X-Bugzilla-Severity: normal
X-Bugzilla-Who: paolo.carlini at oracle dot com
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-58590-4-Cr4MhD9iZL@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-58590-4@http.gcc.gnu.org/bugzilla/>
References: <bug-58590-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-10/txt/msg00073.txt.bz2
Content-length: 623

http://gcc.gnu.org/bugzilla/show_bug.cgi?idX590

--- Comment #5 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Because a SFINAE proper error is when you have an hard error, essentially by
definition from the implementation point of view, not when a static_assert
triggers. And this is not the case here. Unless you have reason to believe that
the front-end is doing something special for this kind of code, eg in overload
resolution, etc, which it wouldn't do outside sfinae.

If you didn't want to see errors you wanted -c, nothing special, what will
happen anyway in the testsuite for this kind of testcase.

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

Paolo Carlini <paolo.carlini at oracle dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2013-10-01
     Ever confirmed|0                           |1

--- Comment #6 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Consider the below. The fact that we do *not* reject it is a much smaller,
definitely non-SFINAE, but equivalent issue, IMO:

//---------------------------
struct S { int member; struct member{}; };

template<class T, class = typename T::member>
auto f(int) -> char(&)[1];

auto v = f<S>(0);
//---------------------------

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #7 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Or, I don't know, simple C++98 which we should also reject:

//------------------
struct S { int member; struct member{}; };

template<class T, class = typename T::member>
struct C { };

template class C<S>;
//------------------

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #8 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Or we don't reject:

struct S { int member; struct member{}; };

template<class T>
struct C { typedef typename T::member U; };

template class C<S>;

but you see my point. I'm surprised that we don't already have something in
bugzilla about this, though.

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[daniel.kruegler at googlemail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #9 from Daniel Krügler <daniel.kruegler at googlemail dot com> ---
(In reply to Paolo Carlini from comment #5)
> Because a SFINAE proper error is when you have an hard error, essentially by
> definition from the implementation point of view, not when a static_assert
> triggers. And this is not the case here. Unless you have reason to believe
> that the front-end is doing something special for this kind of code, eg in
> overload resolution, etc, which it wouldn't do outside sfinae.

This is a different as the one that I got from Jason a long time in the past a
ago: The errors tagged with SFINAE don't necessarily mean that the program is
ill-formed or not, but that any program that is under sfinae control (such as
the originally suggested example in this issue) compiles or compiles not
according to expectations based on the language rules. Basically all my issues
are based on such compile-time expressions combined with static_assert, if
possible, starting from the time where Jason implemented constexpr.
>From gcc-bugs-return-430935-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Wed Oct 02 05:50:20 2013
Return-Path: <gcc-bugs-return-430935-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 7536 invoked by alias); 2 Oct 2013 05:50:18 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 7320 invoked by uid 48); 2 Oct 2013 05:50:14 -0000
From: "Joost.VandeVondele at mat dot ethz.ch" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug fortran/58593] New: [4.9 Regression] internal compiler error: in gfc_conv_string_tmp, at fortran/trans-expr.c:2360
Date: Wed, 02 Oct 2013 05:50:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: fortran
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: Joost.VandeVondele at mat dot ethz.ch
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status bug_severity priority component assigned_to reporter
Message-ID: <bug-58593-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-10/txt/msg00079.txt.bz2
Content-length: 2096

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58593

            Bug ID: 58593
           Summary: [4.9 Regression] internal compiler error: in
                    gfc_conv_string_tmp, at fortran/trans-expr.c:2360
           Product: gcc
           Version: 4.9.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: fortran
          Assignee: unassigned at gcc dot gnu.org
          Reporter: Joost.VandeVondele at mat dot ethz.ch

Most likely caused by the fix for PR fortran/58579

http://gcc.gnu.org/viewcvs?rev=203088&root=gcc&view=rev


> cat bug.f90

    CHARACTER(len=20)                        :: tmpStr
    INTEGER :: output_unit=6
       WRITE (UNIT=output_unit,FMT="(T2,A,T61,A20)")&
         "DFT| Self-interaction correction (SIC)",ADJUSTR(TRIM(tmpstr))
    END


bug.f90: In function ‘MAIN__’:
bug.f90:4:0: internal compiler error: in gfc_conv_string_tmp, at
fortran/trans-expr.c:2360
          "DFT| Self-interaction correction (SIC)",ADJUSTR(TRIM(tmpstr))
 ^
0x6224dd gfc_conv_string_tmp(gfc_se*, tree_node*, tree_node*)
        ../../gcc/gcc/fortran/trans-expr.c:2360
0x634984 gfc_conv_intrinsic_adjust
        ../../gcc/gcc/fortran/trans-intrinsic.c:5328
0x64397a gfc_conv_intrinsic_function(gfc_se*, gfc_expr*)
        ../../gcc/gcc/fortran/trans-intrinsic.c:6644
0x631642 gfc_conv_function_expr
        ../../gcc/gcc/fortran/trans-expr.c:5547
0x626e0f gfc_conv_expr_reference(gfc_se*, gfc_expr*)
        ../../gcc/gcc/fortran/trans-expr.c:6359
0x64a871 gfc_trans_transfer(gfc_code*)
        ../../gcc/gcc/fortran/trans-io.c:2302
0x5f4787 trans_code
        ../../gcc/gcc/fortran/trans.c:1825
0x648278 build_dt
        ../../gcc/gcc/fortran/trans-io.c:1835
0x5f47a7 trans_code
        ../../gcc/gcc/fortran/trans.c:1797
0x61e115 gfc_generate_function_code(gfc_namespace*)
        ../../gcc/gcc/fortran/trans-decl.c:5528
0x5b3256 translate_all_program_units
        ../../gcc/gcc/fortran/parse.c:4509
0x5b3256 gfc_parse_file()
        ../../gcc/gcc/fortran/parse.c:4706
0x5f0445 gfc_be_parse_file
>From gcc-bugs-return-430936-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Wed Oct 02 05:51:16 2013
Return-Path: <gcc-bugs-return-430936-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 8518 invoked by alias); 2 Oct 2013 05:51:16 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 8463 invoked by uid 48); 2 Oct 2013 05:51:13 -0000
From: "Joost.VandeVondele at mat dot ethz.ch" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug fortran/58593] [4.9 Regression] internal compiler error: in gfc_conv_string_tmp, at fortran/trans-expr.c:2360
Date: Wed, 02 Oct 2013 05:51:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: fortran
X-Bugzilla-Version: 4.9.0
X-Bugzilla-Keywords: ice-on-valid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: Joost.VandeVondele at mat dot ethz.ch
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 4.9.0
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: cf_known_to_work target_milestone cf_known_to_fail keywords cf_reconfirmed_on cc
Message-ID: <bug-58593-4-4H2TiXxTje@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-58593-4@http.gcc.gnu.org/bugzilla/>
References: <bug-58593-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2013-10/txt/msg00080.txt.bz2
Content-length: 758

http://gcc.gnu.org/bugzilla/show_bug.cgi?idX593

Joost VandeVondele <Joost.VandeVondele at mat dot ethz.ch> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |4.8.1
   Target Milestone|---                         |4.9.0
      Known to fail|                            |4.9.0
           Keywords|                            |ice-on-valid-code
   Last reconfirmed|                            |2013-10-02
                 CC|                            |burnus at gcc dot gnu.org,
                   |                            |Joost.VandeVondele at mat dot ethz
                   |                            |.ch

[Bug c++/58590] [C++11] Hidden typename not ill-formed under SFINAE conditions

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

--- Comment #10 from Paolo Carlini <paolo.carlini at oracle dot com> ---
Personally, I never discussed with anybody what exactly using the word SFINAE
in a subject means. I only fix bugs ;) Anyway, in practice, as far as this
issue is concerned, it should be clear by now that, as I suspected, it doesn't
have much to do with SFINAE, can be easily reproduced in all sorts of contexts,
with very simple testcases.

[Bug c++/58590] Hidden typename not ill-formed

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

Marek Polacek <mpolacek at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mpolacek at gcc dot gnu.org

--- Comment #11 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
The original test is still rejected:

58590.C:9:31: error: static assertion failed
    9 | static_assert(sizeof(f<S>(0)) == 2, "");
      |               ~~~~~~~~~~~~~~~~^~~~

[Bug c++/58590] Hidden typename not ill-formed

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58590

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |DUPLICATE

--- Comment #12 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Dup of bug 32066.

*** This bug has been marked as a duplicate of bug 32066 ***