[Bug middle-end/17042] New: [regression 3.5] Duplicated symbol in assembly output

[Bug middle-end/17042] New: [regression 3.5] Duplicated symbol in assembly output

[carlo at gcc dot gnu dot org]

I cannot manage to write a small test case that still shows this bug.
However - it must be a bug because this code compiles with all other
compiler versions and did compile fine with CVS of May 2004.

To reproduce compile the attached libcw_app.ii:

[...]
/usr/src/GNU/install/bin/../libexec/gcc/i686-redhat-linux/3.5.0/cc1plus
-fpreprocessed libcw_app.ii -quiet -dumpbase libcw_app.cc -mtune=pentiumpro
-auxbase-strip .libs/libcw_app.o -g -Wall -Woverloaded-virtual -Wundef
-Wpointer-arith -Wwrite-strings -Werror -Winline -version -fPIC -o libcw_app.s
GNU C++ version 3.5.0 20040813 (experimental) (i686-redhat-linux)
        compiled by GNU C version 3.5.0 20040813 (experimental).
GGC heuristics: --param ggc-min-expand=90 --param ggc-min-heapsize=113141
 as -V -Qy -o .libs/libcw_app.o libcw_app.s
GNU assembler version 2.14.90.0.8 (i386-redhat-linux) using BFD version
2.14.90.0.8 20040114
libcw_app.s: Assembler messages:
libcw_app.s:191: Error: symbol `__dso_handle' is already defined


The problem is that `__dso_handle' is emitted twice in libcw_app.s,
while it should only be emitted once.

-- 
           Summary: [regression 3.5] Duplicated symbol in assembly output
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Keywords: wrong-code
          Severity: normal
          Priority: P2
         Component: middle-end
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: carlo at gcc dot gnu dot org
                CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: i686-redhat-gnu-linux


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug middle-end/17042] [regression 3.5] Duplicated symbol in assembly output

[carlo at gcc dot gnu dot org]

------- Additional Comments From carlo at gcc dot gnu dot org  2004-08-15 23:33 -------
Created an attachment (id=6934)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=6934&action=view)
bzip2 compressed preprocessed source code

The .ii file is compressed because it is larger than 1 Mbyte.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [3.5 regression] Duplicated symbol in assembly output

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|middle-end                  |c++
            Summary|[regression 3.5] Duplicated |[3.5 regression] Duplicated
                   |symbol in assembly output   |symbol in assembly output
   Target Milestone|---                         |3.5.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [3.5 regression] Duplicated symbol in assembly output

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-08-15 23:44 -------
Reduced to:
struct A
{  A();  ~A(); };
A a;
extern "C" { void* __dso_handle __attribute__ ((__weak__)); }
void f()
{  __dso_handle = 0; }

Note the code it self is not right for what you want to do, instead it should be:
extern void* __dso_handle __attribute__ ((__weak__));

But the C++ front-end should be picking up the user variable and not making a new one.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2004-08-15 23:44:09
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [3.5 regression] Duplicated symbol in assembly output

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-08-16 17:34 -------
Using Phil's regression hunter (the hard way)
between 2004-07-29 and 2004-07-30 is when this broke.
Caused most likely by:
2004-07-29  Mark Mitchell  <mark@codesourcery.com>
        
        * cp-tree.h (IDENTIFIER_REPO_CHOSEN): Define.
        (lang_decl_flags): Narrow the width of "languages".  Add
        repo_available_p.
        (DECL_NEEDED_P): Remove. 
        (FOR_EACH_CLONE): New macro.
        (DECL_REPO_AVAILABLE_P): Likewise.
        (DECL_TINFO_P): Likewise.
        (set_linkage_according_to_type): Declare.
        (import_export_vtable): Remove.
        (import_export_tinfo): Likewise.
        ....

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mark at codesourcery dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [3.5 regression] Duplicated symbol in assembly output

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-08-17 06:25 -------
First, I am not sure that I agree that this is a bug:  __dso_handle is a magic
variable outside the user's namespace.

Second, I cannot reproduce this bug with today's compiler on i686-pc-linux-gnu
using the reduced test case.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [3.5 regression] Duplicated symbol in assembly output

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-08-17 07:19 -------
I still can reproduce it on the mainline (note that you have to have --enable-__cxa_atexit enabled):
        .type   __dso_handle, @object
        .size   __dso_handle, 4
__dso_handle:
        .zero   4
        .weak   __dso_handle
        .align 4
        .type   __dso_handle, @object
        .size   __dso_handle, 4
__dso_handle:
        .zero   4

Still this is a QOI issue, if we do anything we should warn or error out on this situation.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [4.0 regression] Duplicated symbol in assembly output

[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=17042

[Bug c++/17042] [4.0 regression] Duplicated symbol in assembly output

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-10-15 21:11 -------
Subject: Bug 17042

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2004-10-15 21:11:50

Modified files:
	gcc/cp         : ChangeLog decl.c parser.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/g++.dg/init: dso_handle1.C 
	gcc/testsuite/g++.dg/parse: pragma1.C typedef6.C 

Log message:
	PR c++/17042
	* decl.c (declare_global_var): Use the return value from pushdecl.
	
	PR c++/14667
	* parser.c (cp_parser_simple_declaration): Do not diagnose invalid
	type names if we have already found a valid type.
	(cp_parser_member_declaration): Likewise.
	
	PR c++/17852
	* parser.c (cp_parser_member_specification_opt): Handle
	CPP_PRAGMA.
	
	PR c++/17042
	* g++.dg/init/dso_handle1.C: New test.
	
	PR c++/17852
	* g++.dg/parse/pragma1.C: New test.
	
	PR c++/14667
	* g++.dg/parse/typedef6.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4437&r2=1.4438
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/decl.c.diff?cvsroot=gcc&r1=1.1315&r2=1.1316
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/parser.c.diff?cvsroot=gcc&r1=1.263&r2=1.264
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4459&r2=1.4460
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/init/dso_handle1.C.diff?cvsroot=gcc&r1=NONE&r2=1.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/parse/pragma1.C.diff?cvsroot=gcc&r1=NONE&r2=1.1
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/parse/typedef6.C.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [4.0 regression] Duplicated symbol in assembly output

[mmitchel at gcc dot gnu dot org]

------- Additional Comments From mmitchel at gcc dot gnu dot org  2004-10-15 21:15 -------
Fixed in GCC 4.0.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042

[Bug c++/17042] [4.0 regression] Duplicated symbol in assembly output

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-10-15 21:18 -------
Subject: Bug 17042

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	mmitchel@gcc.gnu.org	2004-10-15 21:18:08

Modified files:
	gcc/cp         : ChangeLog 
	gcc/testsuite  : ChangeLog 
	gcc/testsuite/g++.dg/parse: pragma1.C 

Log message:
	PR c++/17042
	* decl.c (declare_global_var): Use the return value from pushdecl.
	
	PR c++/14667
	* parser.c (cp_parser_simple_declaration): Do not diagnose invalid
	type names if we have already found a valid type.
	(cp_parser_member_declaration): Likewise.
	
	PR c++/17852
	* parser.c (cp_parser_member_specification_opt): Handle
	CPP_PRAGMA.
	
	PR c++/17042
	* g++.dg/init/dso_handle1.C: New test.
	
	PR c++/17852
	* g++.dg/parse/pragma1.C: New test.
	
	PR c++/14667
	* g++.dg/parse/typedef6.C: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/cp/ChangeLog.diff?cvsroot=gcc&r1=1.4438&r2=1.4439
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4460&r2=1.4461
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/g++.dg/parse/pragma1.C.diff?cvsroot=gcc&r1=1.1&r2=1.2



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17042