[Bug c++/18073] New: [4.0 regression] mmintrin.h rejected by C++ frontend

[Bug c++/18073] New: [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

Mark, since your patch
http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01166.html
the C++ frontend rejects the following code:

=====================================
#include<mmintrin.h>
=====================================

I get gazillions of messages like
/Work/reichelt/GCC/FARM/gcc-4.0-20041020/bin/../lib/gcc/i686-pc-linux-gnu/4.0.0/include/mmintrin.h:
In function `int __vector__ _mm_cvtsi32_si64(int)':
/Work/reichelt/GCC/FARM/gcc-4.0-20041020/bin/../lib/gcc/i686-pc-linux-gnu/4.0.0/include/mmintrin.h:62:
error: invalid cast from type 'long long int' to type 'int __vector__'

-- 
           Summary: [4.0 regression] mmintrin.h rejected by C++ frontend
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Keywords: rejects-valid
          Severity: normal
          Priority: P2
         Component: c++
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: reichelt at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org,mark at codesourcery dot
                    com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-10-20 12:00 -------
Confirmed.  This should be accepted.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |critical
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2004-10-20 12:00:47
               date|                            |
   Target Milestone|---                         |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-10-20 13:38 -------
Many C++ headers also fail when compiled with -pedantic:

==================
#include<iosfwd>
==================

/Work/reichelt/GCC/FARM/gcc-4.0-20041020/bin/../lib/gcc/i686-pc-linux-gnu/4.0.0/../../../../include/c++/4.0.0/i686-pc-linux-gnu/bits/gthr-default.h:
In function `int __gthread_active_p()':
/Work/reichelt/GCC/FARM/gcc-4.0-20041020/bin/../lib/gcc/i686-pc-linux-gnu/4.0.0/../../../../include/c++/4.0.0/i686-pc-linux-gnu/bits/gthr-default.h:108:
error: ISO C++ forbids casting between pointer-to-function and pointer-to-object


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[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=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-10-21 07:53 -------
Just for the record: The trouble with casting between pointer-to-function
and void* has been resolved by Nathan's patch

http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01199.html

The trouble with mmintrin.h remains, though.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-10-21 21:23 -------
Subject: Bug 18073

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2004-10-21 21:23:42

Modified files:
	gcc/testsuite  : ChangeLog 
	gcc/cp         : ChangeLog call.c class.c cp-tree.h cvt.c 
	                 typeck.c 
	gcc/testsuite/g++.dg/overload: pmf1.C 
Added files:
	gcc/testsuite/g++.dg/conversion: cast1.C 

Log message:
	PR c++/18073
	PR c++/10841
	* cp-tree.h (convert_to_base): Change prototype.
	(build_ptrmemfunc): Likewise.
	(convert_ptrmem): New function.
	* call.c (struct conversion): Adjust documentation for base_p.
	(standard_conversion): Set base_p for ck_pmem conversions as
	appropriate.
	(convert_like_real): Use convert_to_base for ck_pmem and ck_ptr
	conversions.
	* class.c (convert_to_base): Handle both pointers and objects.
	Add nonnull parameter.
	(build_vfield_ref): Adjust call to convert_to_base.
	* cvt.c (cp_convert_to_pointer): Adjust call to build_ptrmemfunc.
	(convert_force): Likewise.
	* typeck.c (build_unary_op): Likewise.
	(convert_ptrmem): New function.
	(build_static_cast_1): Use it.
	(build_reinterpret_cast): Allow conversions to vector types.
	(get_delta_difference): Add c_cast_p parameter.
	(build_ptrmemfunc): Likewise.  Adjust calls to
	get_delta_difference.
	
	PR c++/10841
	* g++.dg/conversion/cast1.C: New test.
	* g++.dg/overload/pmf1.C: Adjust error marker.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4483&r2=1.4484
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4450&r2=1.4451
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/call.c.diff?cvsroot=gcc&r1=1.515&r2=1.516
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/class.c.diff?cvsroot=gcc&r1=1.685&r2=1.686
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/cp-tree.h.diff?cvsroot=gcc&r1=1.1066&r2=1.1067
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/cvt.c.diff?cvsroot=gcc&r1=1.169&r2=1.170
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/typeck.c.diff?cvsroot=gcc&r1=1.587&r2=1.588
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/conversion/cast1.C.diff?cvsroot=gcc&r1=NONE&r2=1.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/overload/pmf1.C.diff?cvsroot=gcc&r1=1.2&r2=1.3



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-10-22 08:00 -------
Fixed on mainline.

Mark, your patch http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01201.html
adds __extension__ to __gthread_active_p to solve the issue in comment #2.

However, this was already fixed by Nathan's patch
http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01199.html
so the __extension__ shouldn't be necessary.

Are you going to remove it again, or would you rather keep it?


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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2004-10-22 16:26 -------
Subject: Re:  [4.0 regression] mmintrin.h rejected by C++ frontend

reichelt at gcc dot gnu dot org wrote:

>------- Additional Comments From reichelt at gcc dot gnu dot org  2004-10-22 08:00 -------
>Fixed on mainline.
>
>Mark, your patch http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01201.html
>adds __extension__ to __gthread_active_p to solve the issue in comment #2.
>
>However, this was already fixed by Nathan's patch
>http://gcc.gnu.org/ml/gcc-cvs/2004-10/msg01199.html
>so the __extension__ shouldn't be necessary.
>
>Are you going to remove it again, or would you rather keep it?
>
Yes, Nathan and I stepped on each other there.  I think we should keep 
the __extension__ just for clarity; it is an extension, and Jason, 
Nathan, and I have been contemplating issuing a diagnostic, but having 
it be a warning.  It probably wouldn't trigger since this is a system 
header, but I think it's best to leave the marker there.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[tbptbp at gmail dot com]

------- Additional Comments From tbptbp at gmail dot com  2004-12-01 13:41 -------
I don't think it's resolved at all, got bitten by exactly the same symptoms with
snapshots gcc-4.0-20041128 & gcc-4.0-20041121 on cygwin.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[nathan at gcc dot gnu dot org]

------- Additional Comments From nathan at gcc dot gnu dot org  2004-12-07 16:11 -------
not fixed. still rejecting
./include/mmintrin.h: In function 'int _mm_cvtsi64_si32(int __vector__)':
./include/mmintrin.h:91: error: invalid cast from type 'int __vector__' to type
'long long int'


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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-12-07 17:29 -------
Mark, it looks like your patch only fixed conversion from integer to
vector but not the other way round. The following patch fixes the problem
for me. Does this look right?


Index: typeck.c
===================================================================
RCS file: /home/reichelt/GCC/CVS/gcc-cvs/gcc/gcc/cp/typeck.c,v
retrieving revision 1.600
diff -u -p -r1.600 typeck.c
--- typeck.c	26 Nov 2004 15:45:47 -0000	1.600
+++ typeck.c	7 Dec 2004 17:24:27 -0000
@@ -4970,7 +4970,8 @@ build_reinterpret_cast_1 (tree type, tre
       expr = decl_constant_value (expr);
       return fold_if_not_in_template (build_nop (type, expr));
     }
-  else if (TREE_CODE (type) == VECTOR_TYPE)
+  else if (TREE_CODE (type) == VECTOR_TYPE
+           || TREE_CODE (intype) == VECTOR_TYPE)
     return fold_if_not_in_template (convert_to_vector (type, expr));
   else
     {
===================================================================


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |ASSIGNED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[mark at codesourcery dot com]

------- Additional Comments From mark at codesourcery dot com  2004-12-08 03:46 -------
Subject: Re:  [4.0 regression] mmintrin.h rejected by C++ frontend

reichelt at gcc dot gnu dot org wrote:
> ------- Additional Comments From reichelt at gcc dot gnu dot org  2004-12-07 17:29 -------
> Mark, it looks like your patch only fixed conversion from integer to
> vector but not the other way round. The following patch fixes the problem
> for me. Does this look right?

Yes, but convert_to_vector is badly misnamed (as this usage on your part 
is really convert_from_vector).  Also, I'm disturbed to see that 
convert_to_vector doesn't check that the "type" parameter is either a 
VECTOR_TYPE or an INTEGER_TYPE.

Your patch is OK if it passes testing, but you might also want to 
consider cleaning up the issues with convert_to_vector.  That's not your 
responsibility, though.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-12-08 16:54 -------
Patch here:
http://gcc.gnu.org/ml/gcc-patches/2004-12/msg00599.html

Mark, the posted patch is a little different from the one in comment #9.
(That's why I didn't apply it yet.)
It uses convert_to_integer for the conversion *from* the vector to
address your concern about the naming of convert_to_vector. Apart from
some additional checks at the beginning the effect should be the same.

I have to agree that convert_to_vector misses some sanity checks, but
this seems to hold for convert_to_vector as well, since I was able to
do the conversion *to* vector with convert_to_integer, too. :-(


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|mark at codesourcery dot com|reichelt at gcc dot gnu dot
                   |                            |org
           Keywords|                            |patch


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-12-09 10:34 -------
Subject: Bug 18073

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	reichelt@gcc.gnu.org	2004-12-09 10:34:24

Modified files:
	gcc/cp         : typeck.c ChangeLog 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/other: mmintrin.C 

Log message:
	PR c++/18073
	* typeck.c (build_reinterpret_cast_1): Allow cast from vector type.
	
	* g++.dg/other/mmintrin.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/typeck.c.diff?cvsroot=gcc&r1=1.602&r2=1.603
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4523&r2=1.4524
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/other/mmintrin.C.diff?cvsroot=gcc&r1=NONE&r2=1.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4728&r2=1.4729



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-12-09 10:39 -------
Fixed.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073

[Bug c++/18073] [4.0 regression] mmintrin.h rejected by C++ frontend

[reichelt at gcc dot gnu dot org]

------- Additional Comments From reichelt at gcc dot gnu dot org  2004-12-15 17:08 -------
There's still trouble on x86_64, but this problem is tracked in PR18967.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=18073