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 897F4386188D; Mon, 21 Dec 2020 09:28:41 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 897F4386188D
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c/98404] ldist might punt on too large expressions for detected
 patterns with -Os
Date: Mon, 21 Dec 2020 09:28:41 +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: 10.2.0
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: REOPENED
X-Bugzilla-Resolution: 
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: short_desc everconfirmed cc cf_reconfirmed_on
 bug_status resolution
Message-ID: <bug-98404-4-pk0ciLPkT9@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-98404-4@http.gcc.gnu.org/bugzilla/>
References: <bug-98404-4@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: <https://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: <https://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Mon, 21 Dec 2020 09:28:41 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D98404

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Compiler emits unexpected   |ldist might punt on too
                   |function call that may      |large expressions for
                   |cause security problems     |detected patterns with -Os
     Ever confirmed|0                           |1
                 CC|                            |rguenth at gcc dot gnu.org
   Last reconfirmed|                            |2020-12-21
             Status|RESOLVED                    |REOPENED
         Resolution|INVALID                     |---

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Well, for -Os I guess we should reopen this for the case where the expressi=
ons
needed for the detected patterns look too big.
ldist in this case turns the loop into:
  _34 =3D _3 + -1;
  _33 =3D (unsigned int) _34;
  _32 =3D (sizetype) _33;
  _35 =3D _32 * 8;
  _36 =3D _3 > 0 ? _35 : 0;
  _37 =3D (long unsigned int) nm_25;
  _38 =3D _37 * 8;
  _39 =3D (sizetype) _33;
  _40 =3D 1 - _39;
  _41 =3D _40 * 8;
  _42 =3D _3 > 0 ? _41 : 8;
  _43 =3D _38 + _42;
  _44 =3D p_20 + _43;
  _45 =3D (long unsigned int) nm_25;
  _46 =3D _45 * 8;
  _47 =3D (unsigned int) _34;
  _48 =3D (sizetype) _47;
  _49 =3D 1 - _48;
  _50 =3D _49 * 8;
  _51 =3D _3 > 0 ? _50 : 8;
  _52 =3D _46 + _51;
  _53 =3D _52 + 18446744073709551608;
  _54 =3D p_20 + _53;
  __builtin_memmove (_44, _54, _36);
and while some of it is optimized later, we still have:
  _4 =3D (long unsigned int) nm_25;
  _5 =3D _4 * 8;
  _6 =3D p_20 + _5;
  _33 =3D (unsigned int) nm_25;
  _32 =3D (sizetype) _33;
  _35 =3D _32 * 8;
  _36 =3D _3 > 0 ? _35 : 0;
  _40 =3D 1 - _32;
  _41 =3D _40 * 8;
  _42 =3D _3 > 0 ? _41 : 8;
  _43 =3D _5 + _42;
  _44 =3D p_20 + _43;
  _9 =3D _5 + 18446744073709551608;
  _53 =3D _9 + _42;
  _54 =3D p_20 + _53;
in *.optimized.=