From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
 id AB03B38930E9; Fri, 24 Apr 2020 08:34:01 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org AB03B38930E9
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
 s=default; t=1587717241;
 bh=c2XzAbFlgNBS7LUXbOgxC/rgkE66XytdElCiYHOGB/o=;
 h=From:To:Subject:Date:In-Reply-To:References:From;
 b=fpblvNO/NaoN9IFD2sNTuvv05YgeE6xqAMKkrdxvSO3GMszxTrKTrS5crxkrquE7U
 GE8Hr17n6rkZRkisEabkv/L3lXL+b9t6ZaUM7YXSfM14rxERV1jocgTzptCUJVdLJ2
 Bz7OH6szgQNMGK6gko4W4Q6RbBQHUjF9ZwMYPQ/I=
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/94734] [10 Regression] Program crashes when
 compiled with -O2 since r10-1892-gb9ef6a2e04bfd013
Date: Fri, 24 Apr 2020 08:34:01 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: 10.0
X-Bugzilla-Keywords: wrong-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P1
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 10.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-94734-4-kCusNZvjEg@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-94734-4@http.gcc.gnu.org/bugzilla/>
References: <bug-94734-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-BeenThere: gcc-bugs@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-bugs mailing list <gcc-bugs.gcc.gnu.org>
List-Unsubscribe: <http://gcc.gnu.org/mailman/options/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-request@gcc.gnu.org?subject=help>
List-Subscribe: <http://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Fri, 24 Apr 2020 08:34:01 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D94734

--- Comment #10 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Yeah.
add_or_mark_expr could be extended to handle more complex addresses (perhaps
get_inner_reference and hash on the decl + offset expression and taking into
account the bitpos/bitsize then?

Further testcase.  Both foo and bar are miscompiled, baz is fine because ar=
r[7]
is known to be always within the object bounds.

__attribute__((noipa)) int
foo (int n)
{
  int arr[16], s =3D 0;
  for (int i =3D 0; i < n; i++)
    {
      if (i < 16)
        arr[i] =3D i;
    }
  for (int i =3D 0; i < 16; i++)
    s +=3D arr[i];
  return s;
}

__attribute__((noipa)) int
bar (int n, int x, unsigned long y, unsigned long z)
{
  int arr[16], s =3D 0;
  arr[4] =3D 42;
  for (int i =3D 0; i < n; i++)
    {
      if (x =3D=3D (i & 0x25))
        arr[y] =3D i;
    }
  return arr[z];
}

__attribute__((noipa)) int
baz (int n, int x, unsigned long z)
{
  int arr[16], s =3D 0;
  arr[12] =3D 42;
  for (int i =3D 0; i < n; i++)
    {
      if (x =3D=3D (i & 0x25))
        arr[7] =3D i;
    }
  return arr[z];
}

int
main ()
{
  if (foo (10374) !=3D 15 * 16 / 2)
    __builtin_abort ();
  if (bar (25, 0x25, (unsigned long) 0xdeadbeefbeefdeadULL, 4) !=3D 42)
    __builtin_abort ();
  if (bar (25, 4, 17, 17) !=3D 22)
    __builtin_abort ();
  if (baz (25, 0x25, 12) !=3D 42)
    __builtin_abort ();
  if (baz (25, 4, 7) !=3D 22)
    __builtin_abort ();
  if (baz (25, 4, 12) !=3D 42)
    __builtin_abort ();
  return 0;
}=