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 D7651385829E; Mon, 27 Nov 2023 22:44:15 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org D7651385829E
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1701125055;
	bh=moyVKsWC3rJbueL1xxrBddji3zqfIKqRTf9ZHpgjrHg=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=XGE1Nsfc7ZrQy3jsFez0a2XWL/ImOteIU5pAgWlB4ExK9vhLPl4KRxvNdPLlEguR3
	 BvmEnkowrAjOENH2kZifW0941knNizM/e1+2eWCWTW8xDGfcBx+fxv8o3oe9n98g/z
	 sB705vAq5zwiU6+AnQA0LQzd0uncsjOyVtoriWr0=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/112736] [14 Regression] vectorizer is
 introducing out of bounds memory access
Date: Mon, 27 Nov 2023 22:44:15 +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: 14.0
X-Bugzilla-Keywords: needs-bisection, wrong-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: pinskia at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 14.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: short_desc cf_known_to_work cf_known_to_fail
 cf_reconfirmed_on keywords bug_status everconfirmed target_milestone
Message-ID: <bug-112736-4-WUmR6Mjfb3@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-112736-4@http.gcc.gnu.org/bugzilla/>
References: <bug-112736-4@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
List-Id: <gcc-bugs.sourceware.org>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D112736

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|vectorizer is introducing   |[14 Regression] vectorizer
                   |out of bounds memory access |is introducing out of
                   |                            |bounds memory access
      Known to work|                            |13.1.0
      Known to fail|                            |14.0
   Last reconfirmed|                            |2023-11-27
           Keywords|                            |needs-bisection, wrong-code
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
   Target Milestone|---                         |14.0

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
  vect__14.12_2 =3D MEM <vector(4) int> [(int *)&b + -4B];
  vect__14.14_16 =3D VEC_PERM_EXPR <vect__14.12_2, vect__14.12_2, { 3, 3, 3=
, 3
}>;


This might be ok, unless before b is unaligned and what is before is unmapp=
ed.

  # vectp_b.10_23 =3D PHI <vectp_b.10_13(5), &MEM <int[3]> [(void *)&b + -4=
B](2)>
  vect__14.12_1 =3D MEM <vector(4) int> [(int *)vectp_b.10_23];
  vect__14.13_10 =3D VEC_PERM_EXPR <vect__14.12_1, vect__14.12_1, { 3, 2, 1=
, 0
}>;
  vectp_b.10_11 =3D vectp_b.10_23 + 12;
  vect__14.14_12 =3D VEC_PERM_EXPR <vect__14.13_10, vect__14.13_10, { 0, 0,=
 0, 0
}>;


Note GCC 13 was ok:
  _1 =3D b[2];
  _2 =3D {_1, _1, _1, _1};
  MEM <vector(4) int> [(int *)&c + 40B] =3D _2;=