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 3C150385842C; Wed,  4 Oct 2023 19:09:26 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 3C150385842C
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1696446566;
	bh=Uj/GtIc/k1+64Wtif5RKI7m0uuFjlruU8hNPo15NLLw=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=xW97QIpgd4k2x1RN0JP65BDH/HOCf8EoHiB3SLmkmiabfQu134I4No8ZwY8vsQ4Zf
	 SNjG+ntE4pV5Nom9rrlMkGekL5XRJ6u+0xW8tIt1cmnbmQGC4L5s/Zdm0xzn5AGLb7
	 FAY8r3DwxGUlYf0VRn3tVdBEn0CWoY7exmeDsvuc=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/111694] [13/14 Regression] Wrong behavior for
 signbit of negative zero when optimizing
Date: Wed, 04 Oct 2023 19:09:25 +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: 13.2.0
X-Bugzilla-Keywords: 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: 13.3
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: target_milestone everconfirmed cf_reconfirmed_on
 bug_status
Message-ID: <bug-111694-4-8CSeoEgzgo@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-111694-4@http.gcc.gnu.org/bugzilla/>
References: <bug-111694-4@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=3D111694

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |13.3
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2023-10-04
             Status|UNCONFIRMED                 |NEW

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.=20


EVRP removes the second __bultin_signbit incorrectly.
Most likely due to:
```

Folding statement: if (_3 !=3D 0)
 Registering value_relation (l_11(D) =3D=3D r_12(D)) on (2->3)
```

which then we get:
```
Predicate evaluates to: DON'T KNOW
Matching expression match.pd:2679, gimple-match-2.cc:35
Matching expression match.pd:2682, gimple-match-1.cc:66
Matching expression match.pd:2689, gimple-match-2.cc:96
Not folded
Folding statement: _20 =3D __builtin_signbit (r_12(D));
Queued stmt for removal.  Folds to: 0
Folding statement: if (_20 !=3D 0)
gimple_simplified to if (0 !=3D 0)
gimple_simplified to if (0 !=3D 0)
Folded into: if (0 !=3D 0)
```

For -O1 DOM does:
```
1>>> STMT 1 =3D _8 le_expr 0
1>>> STMT 1 =3D _8 ge_expr 0
1>>> STMT 1 =3D _8 eq_expr 0
1>>> STMT 0 =3D _8 ne_expr 0
0>>> COPY _8 =3D 0
Optimizing statement _10 =3D __builtin_signbit (r_6(D));
  Replaced 'r_6(D)' with constant '0.0'
gimple_simplified to _10 =3D 0;
  Folded to: _10 =3D 0;
LKUP STMT _10 =3D 0
=3D=3D=3D=3D ASGN _10 =3D 0
Optimizing statement if (_10 !=3D 0)
  Replaced '_10' with constant '0'
gimple_simplified to if (0 !=3D 0)
  Folded to: if (0 !=3D 0)
LKUP STMT 0 ne_expr 0
<<<< STMT 0 =3D _8 ne_expr 0
<<<< STMT 1 =3D _8 eq_expr 0
<<<< STMT 1 =3D _8 ge_expr 0
<<<< STMT 1 =3D _8 le_expr 0
<<<< COPY _8 =3D 0
```
Which is wrong too.=