[Bug c++/21619] New: [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[Bug c++/21619] New: [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[rguenth at gcc dot gnu dot org]

The testcase

int f[__builtin_constant_p(&"Hello"[0])?1:-1];

did work on gcc 3.4, but doesn't any longer on 4.0 and 4.1.

The parser gets an integer constant node from fold_builtin, but still
thinks it is not a compile time constant for the array bound.

-- 
           Summary: [4.0/4.1 regression]
                    __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time
                    constant
           Product: gcc
           Version: 4.1.0
            Status: UNCONFIRMED
          Keywords: rejects-valid, missed-optimization
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: rguenth at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org
 BugsThisDependsOn: 21595


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-05-17 11:43 -------
Confirmed.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-05-17 11:43:17
               date|                            |
   Target Milestone|---                         |4.0.1


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[rguenth at gcc dot gnu dot org]

-- 
Bug 21619 depends on bug 21595, which changed state.

Bug 21595 Summary: [4.0 Regression] __builtin_constant_p(&"Hello"[0]) is true for C but not for C++
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21595

           What    |Old Value                   |New Value
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-06-01 00:31 -------
: Search converges between 2004-08-30-trunk (#529) and 2004-08-31-trunk (#530).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[rguenth at tat dot physik dot uni-tuebingen dot de]

------- Additional Comments From rguenth at tat dot physik dot uni-tuebingen dot de  2005-06-01 08:16 -------
Subject: Re:  [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1
 not compile-time constant

On 1 Jun 2005, pinskia at gcc dot gnu dot org wrote:

>
> ------- Additional Comments From pinskia at gcc dot gnu dot org  2005-06-01 00:31 -------
> : Search converges between 2004-08-30-trunk (#529) and 2004-08-31-trunk (#530).

Top of cp/ChangeLog for these?  I point my finger at

2004-08-31  Richard Henderson  <rth@redhat.com>

        PR c++/17221
        * pt.c (tsubst_expr): Move OFFSETOF_EXPR handling ...
        (tsubst_copy_and_build): ... here.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mmitchel at gcc dot gnu dot org]

-- 
           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=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-06-05 16:33 -------
Subject: Bug 21619

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2005-06-05 16:33:45

Modified files:
	gcc/cp         : ChangeLog cp-tree.h parser.c tree.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/ext: builtin9.C 

Log message:
	PR c++/21619
	* cp-tree.h (DECL_IS_BUILTIN_CONSTANT_P): New macro.
	* parser.c (cp_parser_postfix_expression): Allow non-constant
	expressions as arguments to __builtin_constant_p.
	* tree.c (builtin_valid_in_constant_expr_p): Use
	DECL_IS_BUILTIN_CONSTANT_P.
	
	PR c++/21619
	* g++.dg/ext/builtin9.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4774&r2=1.4775
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/cp-tree.h.diff?cvsroot=gcc&r1=1.1140&r2=1.1141
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/parser.c.diff?cvsroot=gcc&r1=1.335&r2=1.336
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/tree.c.diff?cvsroot=gcc&r1=1.435&r2=1.436
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5597&r2=1.5598
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/ext/builtin9.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0/4.1 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-06-05 16:47 -------
Subject: Bug 21619

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-4_0-branch
Changes by:	mmitchel@gcc.gnu.org	2005-06-05 16:46:54

Modified files:
	gcc/cp         : ChangeLog cp-tree.h parser.c tree.c 

Log message:
	PR c++/21619
	* cp-tree.h (DECL_IS_BUILTIN_CONSTANT_P): New macro.
	* parser.c (cp_parser_postfix_expression): Allow non-constant
	expressions as arguments to __builtin_constant_p.
	* tree.c (builtin_valid_in_constant_expr_p): Use
	DECL_IS_BUILTIN_CONSTANT_P.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.4648.2.57&r2=1.4648.2.58
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/cp-tree.h.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.1106.2.7&r2=1.1106.2.8
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/parser.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.319.2.8&r2=1.319.2.9
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/tree.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.427.2.3&r2=1.427.2.4



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-06-05 16:48 -------
4.0.1 has a patch to tolerate array accesses in __builtin_constant_p arguments,
but __builtin_constant_p never returns 1 for arguments of pointer type on the
4.0 branch, so the original test case still will not compile on 4.0.  In 4.1,
__builtin_constant_p is more clever, so the test case does work there.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[4.0/4.1 regression]        |[4.0 regression]
                   |__builtin_constant_p(&"Hello|__builtin_constant_p(&"Hello
                   |"[0])?1:-1 not compile-time |"[0])?1:-1 not compile-time
                   |constant                    |constant


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-06-05 16:49 -------
Unassigning from myself; this is now a middle-end bug.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[rguenth at gcc dot gnu dot org]

------- Additional Comments From rguenth at gcc dot gnu dot org  2005-06-05 17:00 -------
The middle-ends fold_builtin_constant_p routines are the same for 4.0 branch and
HEAD - can you point me to the difference you are seeing?

Thanks,
Richard.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-06-05 17:30 -------
Mark this is fixed on the 4.0 branch, the middle-end part was PR 21595 and was commited to 4.0 
branch:
2005-05-27  Richard Guenther  <rguenth@gcc.gnu.org>

        PR middle-end/21595
        * builtins.c (fold_builtin_constant_p): Handle
        &"string cst"[0] as constant.

And the testcase here works now.

Were you using an old branch to test on?

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2005-06-05 17:35 -------
Subject: Re:  [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1
 not compile-time constant

pinskia at gcc dot gnu dot org wrote:
> ------- Additional Comments From pinskia at gcc dot gnu dot org  2005-06-05 17:30 -------
> Mark this is fixed on the 4.0 branch, the middle-end part was PR 21595 and was commited to 4.0 
> branch:
> 2005-05-27  Richard Guenther  <rguenth@gcc.gnu.org>
> 
>         PR middle-end/21595
>         * builtins.c (fold_builtin_constant_p): Handle
>         &"string cst"[0] as constant.
> 
> And the testcase here works now.
> 
> Were you using an old branch to test on?

Good thought; let me just confirm.  If so, I'll add the testcase, and 
close the PR.

Thanks!



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-06-05 17:56 -------
Subject: Bug 21619

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-4_0-branch
Changes by:	mmitchel@gcc.gnu.org	2005-06-05 17:56:48

Modified files:
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/ext: builtin9.C 

Log message:
	PR c++/21619
	* g++.dg/ext/builtin9.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.5084.2.230&r2=1.5084.2.231
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/ext/builtin9.C.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2005-06-05 17:57 -------
Fixed in 4.0.1.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619

[Bug c++/21619] [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1 not compile-time constant

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2005-06-05 17:57 -------
Subject: Re:  [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1
 not compile-time constant

mark at codesourcery dot com wrote:
> ------- Additional Comments From mark at codesourcery dot com  2005-06-05 17:35 -------
> Subject: Re:  [4.0 regression] __builtin_constant_p(&"Hello"[0])?1:-1
>  not compile-time constant
> 
> pinskia at gcc dot gnu dot org wrote:
> 
>>------- Additional Comments From pinskia at gcc dot gnu dot org  2005-06-05 17:30 -------
>>Mark this is fixed on the 4.0 branch, the middle-end part was PR 21595 and was commited to 4.0 
>>branch:
>>2005-05-27  Richard Guenther  <rguenth@gcc.gnu.org>
>>
>>        PR middle-end/21595
>>        * builtins.c (fold_builtin_constant_p): Handle
>>        &"string cst"[0] as constant.
>>
>>And the testcase here works now.
>>
>>Were you using an old branch to test on?

That was exactly the problem; I've now checked in the testcase and will 
close the PR.

Thanks!



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21619