Re: c++/4884: g++ 3.0.2 problem with -fvolatile

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[lerdsuwa]

Synopsis: g++ 3.0.2 problem with -fvolatile

State-Changed-From-To: open->analyzed
State-Changed-By: lerdsuwa
State-Changed-When: Sat Nov 17 00:14:24 2001
State-Changed-Why:
    Confirmed.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&pr=4884&database=gcc

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[Mark Mitchell]

The following reply was made to PR c++/4884; it has been noted by GNATS.

From: Mark Mitchell <mark@codesourcery.com>
To: "Ashif S. Harji" <asharji@plg2.math.uwaterloo.ca>,
   "mmitchel@gcc.gnu.org" <mmitchel@gcc.gnu.org>,
   "asharji@uwaterloo.ca" <asharji@uwaterloo.ca>,
   "gcc-bugs@gcc.gnu.org" <gcc-bugs@gcc.gnu.org>,
   "gcc-prs@gcc.gnu.org" <gcc-prs@gcc.gnu.org>,
   "pabuhr@uwaterloo.ca" <pabuhr@uwaterloo.ca>,
   "gcc-gnats@gcc.gnu.org" <gcc-gnats@gcc.gnu.org>
Cc: gcc-patches@gcc.gnu.org
Subject: Re: c++/4884: g++ 3.0.2 problem with -fvolatile
Date: Wed, 27 Mar 2002 11:14:35 -0800

 --On Tuesday, March 26, 2002 06:21:10 PM -0500 "Ashif S. Harji" 
 <asharji@plg2.math.uwaterloo.ca> wrote:
 
 >
 > Thanks, for the prompt response.
 >
 > I downloaded the latest version from CVS.  Unfortunately, there is still a
 > problem in the build_op_delete_call function.
 >
 > The compilation of the following program fails with a segault when
 > -fvolatile is used but not without.  Note the addition of the constructor
 > to class bar.
 >
 > 2>@awk[114]% more test3.cc
 >
 > class bar {
 >   public :
 >     bar() { }
 >     void * operator new ( unsigned int , void * storage ) { return (void
 > *)1;} };
 >
 > class foo {
 >   public:
 >     void mem ( ) {
 >         new ( 0 ) bar;
 >     }
 > };
 >
 
 Well, that one wasn't in your original bug report. :-)
 
 Fixed with the attached patch.  Tested on i686-pc-linux-gnu, applied
 on the mainline and on the branch.
 
 --
 Mark Mitchell                   mark@codesourcery.com
 CodeSourcery, LLC               http://www.codesourcery.com
 
 2002-03-27  Mark Mitchell  <mark@codesourcery.com>
 
 	PR c++/4884
 	* call.c (build_op_delete_call): Allow for the fact the placement
 	may be a COMPOUND_EXPR.
 	
 Index: cp/call.c
 ===================================================================
 RCS file: /cvs/gcc/gcc/gcc/cp/call.c,v
 retrieving revision 1.307.2.4
 diff -c -p -r1.307.2.4 call.c
 *** call.c	2002/03/18 16:44:07	1.307.2.4
 --- call.c	2002/03/27 18:38:06
 *************** build_op_delete_call (code, addr, size,
 *** 3623,3637 ****
 
     if (placement)
       {
 !       /* placement is a CALL_EXPR around an ADDR_EXPR around a function. 
 */
 
         /* Extract the function.  */
 !       argtypes = TREE_OPERAND (TREE_OPERAND (placement, 0), 0);
         /* Then the second parm type.  */
 !       argtypes = TREE_CHAIN (TYPE_ARG_TYPES (TREE_TYPE (argtypes)));
 !
         /* Also the second argument.  */
 !       args = TREE_CHAIN (TREE_OPERAND (placement, 1));
       }
     else
       {
 --- 3623,3644 ----
 
     if (placement)
       {
 !       tree alloc_fn;
 !       tree call_expr;
 
 +       /* Find the allocation function that is being called. */
 +       call_expr = placement;
 +       /* Sometimes we have a COMPOUND_EXPR, rather than a simple
 + 	 CALL_EXPR. */
 +       while (TREE_CODE (call_expr) == COMPOUND_EXPR)
 + 	call_expr = TREE_OPERAND (call_expr, 1);
         /* Extract the function.  */
 !       alloc_fn = get_callee_fndecl (call_expr);
 !       my_friendly_assert (alloc_fn != NULL_TREE, 20020327);
         /* Then the second parm type.  */
 !       argtypes = TREE_CHAIN (TYPE_ARG_TYPES (TREE_TYPE (alloc_fn)));
         /* Also the second argument.  */
 !       args = TREE_CHAIN (TREE_OPERAND (call_expr, 1));
       }
     else
       {
 Index: testsuite/g++.dg/init/new2.C
 ===================================================================
 RCS file: new2.C
 diff -N new2.C
 *** /dev/null	Tue May  5 13:32:27 1998
 --- new2.C	Wed Mar 27 10:38:08 2002
 ***************
 *** 0 ****
 --- 1,18 ----
 + // Origin: asharji@uwaterloo.ca
 +
 + // { dg-do compile }
 + // { dg-options "-fvolatile" }
 +
 + class bar {
 +   public :
 +     bar() { }
 +     void * operator new ( __SIZE_TYPE__ , void * storage )
 +      { return (void *)1;}
 + };
 +
 + class foo {
 +   public:
 +     void mem ( ) {
 +         new ( 0 ) bar;
 +     }
 + };
 

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[Ashif S. Harji]

The following reply was made to PR c++/4884; it has been noted by GNATS.

From: "Ashif S. Harji" <asharji@plg2.math.uwaterloo.ca>
To: mmitchel@gcc.gnu.org, <asharji@uwaterloo.ca>, <gcc-bugs@gcc.gnu.org>,
        <gcc-prs@gcc.gnu.org>, <mmitchel@gcc.gnu.org>, <pabuhr@uwaterloo.ca>,
        <gcc-gnats@gcc.gnu.org>
Cc:  
Subject: Re: c++/4884: g++ 3.0.2 problem with -fvolatile
Date: Tue, 26 Mar 2002 18:21:10 -0500 (EST)

 Thanks, for the prompt response.
 
 I downloaded the latest version from CVS.  Unfortunately, there is still a
 problem in the build_op_delete_call function.
 
 The compilation of the following program fails with a segault when
 -fvolatile is used but not without.  Note the addition of the constructor
 to class bar.
 
 2>@awk[114]% more test3.cc
 
 class bar {
   public :
     bar() { }
     void * operator new ( unsigned int , void * storage ) { return (void *)1;}
 };
 
 class foo {
   public:
     void mem ( ) {
         new ( 0 ) bar;
     }
 };
 
 2>@awk[115]% g++ test3.cc -c
 
 2>@awk[116]% g++ test3.cc -fvolatile -c
 test3.cc: In member function `void foo::mem()':
 test3.cc:11: internal error: Segmentation fault
 Please submit a full bug report,
 with preprocessed source if appropriate.
 See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.
 
 2>@awk[117]% g++ -v
 Reading specs from ~/gccinstall/lib/gcc-lib/i686-pc-linux-gnu/3.2/specs
 Configured with: ../gcc/configure --prefix=~/gccinstall --enable-languages=c,c++ :
 Thread model: single
 gcc version 3.2 20020326 (experimental)
 
 thanks,
 ashif.
 
 On 26 Mar 2002 mmitchel@gcc.gnu.org wrote:
 
 > Synopsis: g++ 3.0.2 problem with -fvolatile
 >
 > State-Changed-From-To: analyzed->closed
 > State-Changed-By: mmitchel
 > State-Changed-When: Tue Mar 26 10:38:32 2002
 > State-Changed-Why:
 >     Fixed.
 >
 > http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=4884
 >
 

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[mmitchel]

Synopsis: g++ 3.0.2 problem with -fvolatile

State-Changed-From-To: analyzed->closed
State-Changed-By: mmitchel
State-Changed-When: Tue Mar 26 10:38:32 2002
State-Changed-Why:
    Fixed.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=4884

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[mmitchel]

Synopsis: g++ 3.0.2 problem with -fvolatile

Responsible-Changed-From-To: unassigned->mmitchel
Responsible-Changed-By: mmitchel
Responsible-Changed-When: Mon Mar 25 10:53:09 2002
Responsible-Changed-Why:
    Working on a fix.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=4884

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[lerdsuwa]

The following reply was made to PR c++/4884; it has been noted by GNATS.

From: lerdsuwa@gcc.gnu.org
To: asharji@uwaterloo.ca, gcc-bugs@gcc.gnu.org, gcc-gnats@gcc.gnu.org,
  gcc-prs@gcc.gnu.org, nobody@gcc.gnu.org, pabuhr@uwaterloo.ca
Cc:  
Subject: Re: c++/4884: g++ 3.0.2 problem with -fvolatile
Date: 17 Nov 2001 08:14:25 -0000

 Synopsis: g++ 3.0.2 problem with -fvolatile
 
 State-Changed-From-To: open->analyzed
 State-Changed-By: lerdsuwa
 State-Changed-When: Sat Nov 17 00:14:24 2001
 State-Changed-Why:
     Confirmed.
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&pr=4884&database=gcc

Re: c++/4884: g++ 3.0.2 problem with -fvolatile

[lerdsuwa]

The following reply was made to PR c++/4884; it has been noted by GNATS.

From: lerdsuwa@gcc.gnu.org
To: asharji@uwaterloo.ca, gcc-bugs@gcc.gnu.org, gcc-gnats@gcc.gnu.org,
  gcc-prs@gcc.gnu.org, nobody@gcc.gnu.org, pabuhr@uwaterloo.ca
Cc:  
Subject: Re: c++/4884: g++ 3.0.2 problem with -fvolatile
Date: 17 Nov 2001 08:14:25 -0000

 Synopsis: g++ 3.0.2 problem with -fvolatile
 
 State-Changed-From-To: open->analyzed
 State-Changed-By: lerdsuwa
 State-Changed-When: Sat Nov 17 00:14:24 2001
 State-Changed-Why:
     Confirmed.
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&pr=4884&database=gcc

c++/4884: g++ 3.0.2 problem with -fvolatile

[asharji]

>Number:         4884
>Category:       c++
>Synopsis:       g++ 3.0.2 problem with -fvolatile
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          ice-on-legal-code
>Submitter-Id:   net
>Arrival-Date:   Thu Nov 15 12:26:01 PST 2001
>Closed-Date:
>Last-Modified:
>Originator:     Ashif Harji
>Release:        gcc version 3.0.2
>Organization:
>Environment:
operating system version: Redhat 7.1
hardware architecture: athalon 1.4 ghz
>Description:
I am having a problem with g++ 3.0.2 using -fvolatile.

1>@awk[114]% g++ -v
Reading specs from /u/asharji/gcc/gcc302/bin/../lib/gcc-lib/i686-pc-linux-gnu/3.0.2/specs
Configured with: /u/asharji/gcc/gcc-3.0.2/configure --prefix=/u/asharji/gcc/gcc302/
Thread model: single
gcc version 3.0.2

When I compile without -fvolatile, it works:

1>@awk[117]% g++ test1.cc -c

But not when I compile with -fvolatile:

1>@awk[118]% g++ test1.cc -c -fvolatile
test1.cc: In member function `void foo::mem()':
test1.cc:14: Internal error #20000521.
test1.cc:14: Internal compiler error in build_new_1, at cp/init.c:2426
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://www.gnu.org/software/gcc/bugs.html> for instructions.


This error does not occur with gcc-2.95.3:

@tcl[104]% g++ -v
Reading specs from /u/asharji/gcc/lib/gcc-lib/i686-pc-linux-gnu/2.95.3/specs
gcc version 2.95.3 20010315 (release)

@tcl[110]% g++ test1.cc -c

@tcl[111]% g++ test1.cc -c -fvolatile

@tcl[112]% 
>How-To-Repeat:
1>@awk[113]% cat test1.cc 

typedef unsigned int size_t;

class bar {
    int i;
  public :
    void * operator new ( size_t , void * storage );
};

class foo {
    int storage[ 5 ];
  public:
    void mem ( ) {
        bar *s = new ( ( void * ) & storage ) bar;
    }
};

1>@awk[118]% g++ test1.cc -c -fvolatile
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted: