[Bug c++/65327] New: GCC rejects "constexpr volatile int i = 5;"

[Bug c++/65327] New: GCC rejects "constexpr volatile int i = 5;"

[Casey at Carter dot net]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

            Bug ID: 65327
           Summary: GCC rejects "constexpr volatile int i = 5;"
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: Casey at Carter dot net

See
http://stackoverflow.com/questions/28883510/can-constexpr-be-combined-with-volatile
for discussion. GCC 5 (Identified as "5.0.0 HEAD 20150304" on Wandbox at
http://melpon.org/wandbox/permlink/u7BC3wFEfMv0zrVN) rejects this program:

int main() {
    constexpr volatile int i = 5;
}

with the diagnostic:

prog.cc: In function 'int main()':
prog.cc:3:28: error: both 'volatile' and 'constexpr' cannot be used here
    constexpr volatile int i = 5;

AFAICT, this program is well-formed: `constexpr` can be applied to literal
types, which includes scalar types, which includes cv-qualified arithmetic
types. Semantically, I *think* "constexpr volatile" is equivalent to "const
volatile" since "constexpr" objects are implicitly "const" but lvalue-to-rvalue
conversions applied to volatile glvalues cannot be part of a
core-constant-expression.

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[yaghmour.shafik at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

Shafik Yaghmour <yaghmour.shafik at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |yaghmour.shafik at gmail dot com

--- Comment #1 from Shafik Yaghmour <yaghmour.shafik at gmail dot com> ---
Defect report 1688:
http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2014/n4093.html#1688
confirms that this is valid, and it says:

"The combination is intentionally permitted and could be used in some
circumstances to force constant initialization."

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-03-16
                 CC|                            |mpolacek at gcc dot gnu.org
     Ever confirmed|0                           |1

--- Comment #2 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Confirmed then.  I wonder if

--- gcc/cp/decl.c
+++ gcc/cp/decl.c
@@ -10134,8 +10134,9 @@ grokdeclarator (const cp_declarator *declarator,
      the object as `const'.  */
   if (constexpr_p && innermost_code != cdk_function)
     {
-      if (type_quals & TYPE_QUAL_VOLATILE)
-        error ("both %<volatile%> and %<constexpr%> cannot be used here");
+      /* DR1688 says that a `constexpr' specifier in combination with
+     `volatile' is valid.  */
+
       if (TREE_CODE (type) != REFERENCE_TYPE)
     {
       type_quals |= TYPE_QUAL_CONST;

is enough...

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |mpolacek at gcc dot gnu.org
   Target Milestone|---                         |4.8.5

--- Comment #3 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Seems it is, taking.  4.8/4.9 have the same problem.

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

--- Comment #4 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Presumably started with r166013.

Note that in
constexpr volatile int a = 42;
constexpr int b = a;
the initialization of b should be rejected, but it is not.  This is a related
problem though, my patch doesn't change it.

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

--- Comment #5 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Author: mpolacek
Date: Mon Mar 16 18:30:49 2015
New Revision: 221463

URL: https://gcc.gnu.org/viewcvs?rev=221463&root=gcc&view=rev
Log:
    DR 1688
    PR c++/65327
    * decl.c (grokdeclarator): Allow volatile and constexpr together.

    * g++.dg/cpp0x/constexpr-object1.C: Change dg-error to dg-bogus.
    * g++.dg/cpp0x/pr65327.C: New test.

Added:
    trunk/gcc/testsuite/g++.dg/cpp0x/pr65327.C
Modified:
    trunk/gcc/cp/ChangeLog
    trunk/gcc/cp/decl.c
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/testsuite/g++.dg/cpp0x/constexpr-object1.C

[Bug c++/65327] GCC rejects "constexpr volatile int i = 5;"

[mpolacek at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65327

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

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

--- Comment #6 from Marek Polacek <mpolacek at gcc dot gnu.org> ---
Fixed for GCC 5.  Not going to backport it.