[Bug c++/54026] New: [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[Bug c++/54026] New: [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[ppluzhnikov at google dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

             Bug #: 54026
           Summary: [4.7/4.8 regression] template const struct with
                    mutable members erroneously emitted to .rodata
    Classification: Unclassified
           Product: gcc
           Version: unknown
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: ppluzhnikov@google.com


Reproduced with 4.7 and current trunk:
g++ (GCC) 4.8.0 20120719 (experimental)

// --- cut here ---
void non_const(int *);

template <typename T>
struct Foo {
  T x;
  mutable int y;
  void func() const { non_const(&y); }
};

struct Bar {
  int x;
  mutable int y;
  void func() const { non_const(&y); }
};

const Foo<int> foo = { 1, 2 };
const Bar bar = { 3, 4 };
// --- cut here ---

Results in:

    .data
    .align 4
    .type    _ZL3bar, @object
    .size    _ZL3bar, 8
_ZL3bar:
    .long    3
    .long    4

    .section    .rodata
    .align 4
    .type    _ZL3foo, @object
    .size    _ZL3foo, 8
_ZL3foo:
    .long    1
    .long    2


But it is a bug to place "foo" into .rodata, since it contains a mutable
member.

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

--- Comment #1 from Richard Guenther <rguenth at gcc dot gnu.org> 2012-07-19 10:08:08 UTC ---
I don't think 'mutable' overrides a const declaration specifier.

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[ppluzhnikov at google dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

--- Comment #2 from Paul Pluzhnikov <ppluzhnikov at google dot com> 2012-07-19 13:20:40 UTC ---
(In reply to comment #1)
> I don't think 'mutable' overrides a const declaration specifier.

I am not a language lawyer, but I believe the standard disagrees: from
INCITS+ISO+IEC+14882-2011-DRAFT.pdf (similar language in
INCITS+ISO+IEC+14882-1998.pdf):


3.10 Lvalues and rvalues
...
8 The referent of a const-qualified expression shall not be modified (through
that expression), except that if it is of class type and has a mutable
component, that component can be modified (7.1.6.1).


7.1.6.1 The cv-qualifiers
...
5 For another example

  struct X {
    mutable int i;
    int j;
  };

  struct Y {
    X x;
    Y();
  };

  const Y y;
  y.x.i++; // well-formed: mutable member can be modified

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jason at redhat dot com

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> 2012-07-19 14:46:02 UTC ---
It is caused by revision 176052:

http://gcc.gnu.org/ml/gcc-cvs/2011-07/msg00317.html

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-07-19
            Version|unknown                     |4.8.0
   Target Milestone|---                         |4.7.2
     Ever Confirmed|0                           |1

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-07-19 16:44:20 UTC ---
The problem is that for incomplete types TYPE_HAS_MUTABLE_P is never going to
be true, as finish_struct_1 hasn't been called yet, therefore not
check_bases_and_members and check_field_decls which sets CLASSTYPE_HAS_MUTABLE.

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |jason at gcc dot gnu.org
         Resolution|                            |FIXED

--- Comment #5 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:28:53 UTC ---
Checking in fix now.

--- Comment #6 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:29:19 UTC ---
Author: jason
Date: Fri Jul 20 06:29:13 2012
New Revision: 189701

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=189701
Log:
    PR c++/54026
    * typeck.c (cp_apply_type_quals_to_decl): Check COMPLETE_TYPE_P.

Added:
    trunk/gcc/testsuite/g++.dg/init/mutable1.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/typeck.c
    trunk/gcc/testsuite/ChangeLog

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |jason at gcc dot gnu.org
         Resolution|                            |FIXED

--- Comment #5 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:28:53 UTC ---
Checking in fix now.

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

Jason Merrill <jason at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |jason at gcc dot gnu.org
         Resolution|                            |FIXED

--- Comment #5 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:28:53 UTC ---
Checking in fix now.

--- Comment #6 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:29:19 UTC ---
Author: jason
Date: Fri Jul 20 06:29:13 2012
New Revision: 189701

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=189701
Log:
    PR c++/54026
    * typeck.c (cp_apply_type_quals_to_decl): Check COMPLETE_TYPE_P.

Added:
    trunk/gcc/testsuite/g++.dg/init/mutable1.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/typeck.c
    trunk/gcc/testsuite/ChangeLog

--- Comment #7 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:29:37 UTC ---
Author: jason
Date: Fri Jul 20 06:29:33 2012
New Revision: 189702

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=189702
Log:
    PR c++/54026
    * typeck.c (cp_apply_type_quals_to_decl): Check COMPLETE_TYPE_P.

Added:
    branches/gcc-4_7-branch/gcc/testsuite/g++.dg/init/mutable1.C
Modified:
    branches/gcc-4_7-branch/gcc/cp/ChangeLog
    branches/gcc-4_7-branch/gcc/cp/typeck.c
    branches/gcc-4_7-branch/gcc/testsuite/ChangeLog

[Bug c++/54026] [4.7/4.8 regression] template const struct with mutable members erroneously emitted to .rodata

[jason at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54026

--- Comment #7 from Jason Merrill <jason at gcc dot gnu.org> 2012-07-20 06:29:37 UTC ---
Author: jason
Date: Fri Jul 20 06:29:33 2012
New Revision: 189702

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=189702
Log:
    PR c++/54026
    * typeck.c (cp_apply_type_quals_to_decl): Check COMPLETE_TYPE_P.

Added:
    branches/gcc-4_7-branch/gcc/testsuite/g++.dg/init/mutable1.C
Modified:
    branches/gcc-4_7-branch/gcc/cp/ChangeLog
    branches/gcc-4_7-branch/gcc/cp/typeck.c
    branches/gcc-4_7-branch/gcc/testsuite/ChangeLog