[Bug tree-optimization/60115] wrong code at -O3 on x86_64-linux-gnu

[Bug tree-optimization/60115] wrong code at -O3 on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #1 from Zhendong Su <su at cs dot ucdavis.edu> ---
The reported testcase seems to reveal an independent issue: GCC does warn the
out-of-bound array access at -Os and above, but not at -O0 and -O1, while clang
warns at all optimization levels. 

$ gcc-trunk -Warray-bounds -O0 small.c
$ gcc-trunk -Warray-bounds -O1 small.c 
$     
$ gcc-trunk -Warray-bounds -Os small.c
small.c: In function ‘main’:
small.c:8:10: warning: array subscript is above array bounds [-Warray-bounds]
     if (b[613])
          ^
$
$ clang-trunk -Warray-bounds -O0 small.c
small.c:8:9: warning: array index 613 is past the end of the array (which
contains 1 element) [-Warray-bounds]
    if (b[613])
        ^ ~~~
small.c:1:1: note: array 'b' declared here
int a, b[1];
^
1 warning generated.
$
>From gcc-bugs-return-443020-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org Sat Feb 08 07:23:16 2014
Return-Path: <gcc-bugs-return-443020-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Delivered-To: listarch-gcc-bugs@gcc.gnu.org
Received: (qmail 32539 invoked by alias); 8 Feb 2014 07:23:15 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Delivered-To: mailing list gcc-bugs@gcc.gnu.org
Received: (qmail 32513 invoked by uid 48); 8 Feb 2014 07:23:09 -0000
From: "glisse at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c++/60113] Internal compiler error
Date: Sat, 08 Feb 2014 07:23:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: c++
X-Bugzilla-Version: 4.6.3
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: glisse at gcc dot gnu.org
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields:
Message-ID: <bug-60113-4-56CshDmUDh@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-60113-4@http.gcc.gnu.org/bugzilla/>
References: <bug-60113-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2014-02/txt/msg00777.txt.bz2
Content-length: 446

http://gcc.gnu.org/bugzilla/show_bug.cgi?id`113

--- Comment #1 from Marc Glisse <glisse at gcc dot gnu.org> ---
(In reply to lmat from comment #0)
> Attached is the output from command line along with the
> preprocessed code that I was trying to compile.

Looks like you forgot to attach it?

> Please let me know if I can provide any more information!

http://gcc.gnu.org/bugs/
We also want to know the command line, the target system, etc.

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-02-08
                 CC|                            |jakub at gcc dot gnu.org
   Target Milestone|---                         |4.9.0
            Summary|wrong code at -O3 on        |[4.9 Regression] wrong code
                   |x86_64-linux-gnu            |at -O3 on x86_64-linux-gnu
     Ever confirmed|0                           |1

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I have to use something like:
int a, b[1];

int
main ()
{
lbl:
  for (; a; a--)
    if (b[10000000])
      goto lbl;

  return 0;
}

to actually successfully reproduce it.  Anyway, started with r204458.

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #4 from Zhendong Su <su at cs dot ucdavis.edu> ---
(In reply to Jakub Jelinek from comment #2)
> I have to use something like:
> int a, b[1];
> 
> int
> main ()
> {
> lbl:
>   for (; a; a--)
>     if (b[10000000])
>       goto lbl;
> 
>   return 0;
> }
> 
> to actually successfully reproduce it.  Anyway, started with r204458.

Should I report the out-out-bound array access warning issue as a separate
report?  Thanks.

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #5 from Zhendong Su <su at cs dot ucdavis.edu> ---
(In reply to Jakub Jelinek from comment #2)
> I have to use something like:
> int a, b[1];
> 
> int
> main ()
> {
> lbl:
>   for (; a; a--)
>     if (b[10000000])
>       goto lbl;
> 
>   return 0;
> }
> 
> to actually successfully reproduce it.  Anyway, started with r204458.

It also affects the current 4.8 branch that I just built (but not the 4.8
releases).  

$ gcc -v
Using built-in specs.
COLLECT_GCC=gcc
COLLECT_LTO_WRAPPER=/home/su/software/local/gcc-4.8/bin/../libexec/gcc/x86_64-unknown-linux-gnu/4.8.3/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-4.8/configure --enable-languages=c,c++
Thread model: posix
gcc version 4.8.3 20140208 (prerelease) [gcc-4_8-branch revision 207632] (GCC) 
$
$ gcc -O3 small.c; a.out
Segmentation fault (core dumped)
$
$ gcc-4.8.2 -O3 small.c; a.out
$ gcc-4.8.1 -O3 small.c; a.out
$ gcc-4.8.0 -O3 small.c; a.out

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #6 from Zhendong Su <su at cs dot ucdavis.edu> ---
So, it's also a 4.8 regression.

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[su at cs dot ucdavis.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #7 from Zhendong Su <su at cs dot ucdavis.edu> ---
> Should I report the out-out-bound array access warning issue as a separate
> report?  Thanks.

Reported it as 

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60121

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #8 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
LIM is pulling the out of bounds load out of the loop.

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |fmartinez at gmv dot com

--- Comment #9 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
*** Bug 60122 has been marked as a duplicate of this bug. ***

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #10 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
*** Bug 60121 has been marked as a duplicate of this bug. ***

[Bug tree-optimization/60115] [4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org

--- Comment #11 from Richard Biener <rguenth at gcc dot gnu.org> ---
Mine then.

[Bug tree-optimization/60115] [4.8/4.9 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #12 from Richard Biener <rguenth at gcc dot gnu.org> ---
The issue is that tree_could_trap_p, while handling

    case ARRAY_REF:
      base = TREE_OPERAND (expr, 0);
      if (tree_could_trap_p (base))
        return true;
      if (TREE_THIS_NOTRAP (expr))
        return false;
      return !in_array_bounds_p (expr);

doesn't handle out-of-bound accesses via MEM_REF at all.

    case MEM_REF:
      if (TREE_CODE (TREE_OPERAND (expr, 0)) == ADDR_EXPR)
        return false;

I have a fix.

[Bug tree-optimization/60115] [4.8 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |4.9.0
            Summary|[4.8/4.9 Regression] wrong  |[4.8 Regression] wrong code
                   |code at -O3 on              |at -O3 on x86_64-linux-gnu
                   |x86_64-linux-gnu            |
      Known to fail|4.9.0                       |

--- Comment #13 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed on trunk sofar.

[Bug tree-optimization/60115] [4.8 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #14 from Richard Biener <rguenth at gcc dot gnu.org> ---
Author: rguenth
Date: Mon Feb 10 13:51:23 2014
New Revision: 207656

URL: http://gcc.gnu.org/viewcvs?rev=207656&root=gcc&view=rev
Log:
2014-02-10  Richard Biener  <rguenther@suse.de>

    PR tree-optimization/60115
    * tree-eh.c (tree_could_trap_p): Unify TARGET_MEM_REF and
    MEM_REF handling.  Properly verify that the accesses are not
    out of the objects bound.

    * gcc.dg/torture/pr60115.c: New testcase.

Added:
    trunk/gcc/testsuite/gcc.dg/torture/pr60115.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-eh.c

[Bug tree-optimization/60115] [4.8 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

--- Comment #15 from Richard Biener <rguenth at gcc dot gnu.org> ---
Author: rguenth
Date: Mon Feb 10 15:20:30 2014
New Revision: 207659

URL: http://gcc.gnu.org/viewcvs?rev=207659&root=gcc&view=rev
Log:
2014-02-10  Richard Biener  <rguenther@suse.de>

    Backport from mainline
    2014-01-30  Richard Biener  <rguenther@suse.de>

    PR tree-optimization/59903
    * tree-vect-loop.c (vect_transform_loop): Guard multiple-types
    check properly.

    * gcc.dg/torture/pr59903.c: New testcase.

    2014-02-10  Richard Biener  <rguenther@suse.de>

    PR tree-optimization/60115
    * tree-eh.c (tree_could_trap_p): Unify TARGET_MEM_REF and
    MEM_REF handling.  Properly verify that the accesses are not
    out of the objects bound.

    * gcc.dg/torture/pr60115.c: New testcase.

Added:
    branches/gcc-4_8-branch/gcc/testsuite/gcc.dg/torture/pr59903.c
    branches/gcc-4_8-branch/gcc/testsuite/gcc.dg/torture/pr60115.c
Modified:
    branches/gcc-4_8-branch/gcc/ChangeLog
    branches/gcc-4_8-branch/gcc/testsuite/ChangeLog
    branches/gcc-4_8-branch/gcc/tree-eh.c
    branches/gcc-4_8-branch/gcc/tree-vect-loop.c

[Bug tree-optimization/60115] [4.8 Regression] wrong code at -O3 on x86_64-linux-gnu

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60115

Richard Biener <rguenth at gcc dot gnu.org> changed:

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

--- Comment #16 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed.