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 79D083857019; Thu,  8 Jun 2023 21:00:54 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 79D083857019
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1686258054;
	bh=dxtPPl5xG9ZzETOivvhQZ58/eXgiF/NKtb/NjKQBjD4=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=Nn3eslU00LcfR4FNq8JgLEgfEDJALxLiCkUI5A5vTJLXCfnxIZcRSdyz4HRdvQGH1
	 BqvoxpBnRCKWMT73RmfXMk6FJDeXDKm4H8GIi3So0jC3hnjwgh/Jwm5cS8T/TfBkC/
	 0y45q/rnY9esXWv8Oh8w3tf+iv1yGS8pcASJU/vc=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/110182] Vector(2) cast from double to float and
 subtraction seems to produce incorrect results at -O2 and above
Date: Thu, 08 Jun 2023 21:00:54 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 13.1.1
X-Bugzilla-Keywords: 
X-Bugzilla-Severity: normal
X-Bugzilla-Who: pinskia at gcc dot gnu.org
X-Bugzilla-Status: UNCONFIRMED
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: component short_desc
Message-ID: <bug-110182-4-V92vyKWsUn@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-110182-4@http.gcc.gnu.org/bugzilla/>
References: <bug-110182-4@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=3D110182

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|c++                         |target
            Summary|GCC generates incorrect     |Vector(2) cast from double
                   |results for simple Eigen    |to float and subtraction
                   |Casts / Subtractions at -O2 |seems to produce incorrect
                   |or above for a 3            |results at -O2 and above
                   |dimensional vector          |

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Looks like a target issue:
The gimple level looks like this in GCC 12 for low:
  vect__178.810_465 =3D (vector(2) float) _165;
  vect__202.814_467 =3D _165 - _165;
  vect__203.815_466 =3D (vector(2) float) vect__202.814_467;
  _172 =3D SR.808_30 * 6.371e+6;
  MEM[(struct plain_array *)&high] =3D{v} {CLOBBER};
  MEM <vector(2) float> [(float &)&high] =3D vect__178.810_465;
  _184 =3D (float) _172;
  MEM[(float &)&high + 8] =3D _184;
  MEM[(struct plain_array *)&low] =3D{v} {CLOBBER};
  MEM <vector(2) float> [(float &)&low] =3D vect__203.815_466;


While for GCC 11 we have the casting as scalar:
  _173 =3D (float) _298;
  MEM[(float &)&high] =3D _173;
  _296 =3D BIT_FIELD_REF <_160, 64, 64>;
  _177 =3D (float) _296;
  MEM[(float &)&high + 4] =3D _177;
  _179 =3D (float) _167;
  MEM[(float &)&high + 8] =3D _179;
  MEM[(struct plain_array *)&low] =3D{v} {CLOBBER};
  _192 =3D (double) _173;
  _197 =3D _298 - _192;
  _198 =3D (float) _197;
  MEM[(float &)&low] =3D _198;
  _224 =3D (double) _177;
  _229 =3D _296 - _224;
  _230 =3D (float) _229;
  MEM[(float &)&low + 4] =3D _230;
  _208 =3D (double) _179;
  _213 =3D _167 - _208;
  _214 =3D (float) _213;
  MEM[(float &)&low + 8] =3D _214;=