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 74D233851ABB; Tue,  7 Mar 2023 20:49:36 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 74D233851ABB
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1678222176;
	bh=TbmocKYXbPzt+mTYyhpIu5du9y9P4tOTZf2yvnjPt3U=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=NCjTADAj67d5dPmLRgAISqYld6+Ay3FrCGv0YTpVq5lDYbc5O9CGZsiOw8i7deGXu
	 qckLTzIggBvN9BrHLZuIEnV+LyAX8++pGa/ZRLG7h8zUFKmaYtYR8dNR0sZr/uRwMq
	 uHPwhXI6+p0ZmGpTifgjnj9E9G0qn8eDVX3yAXWY=
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/109008] [13 Regression] Wrong code in scipy
 package since r13-3926-gd4c2f1d376da6f
Date: Tue, 07 Mar 2023 20:49:36 +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: 13.0
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: 13.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-109008-4-KxJiLNX1nT@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-109008-4@http.gcc.gnu.org/bugzilla/>
References: <bug-109008-4@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=3D109008

--- Comment #25 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I guess more debugging tomorrow.  Because with the frange_nextafter it come=
s up
for the f1 to range
p[frange] double
[-2.22044604925031283432823045461545143383482334912930322712e-16
(-0x0.fffffffffffff8p-52),
4.44089209850062566865646090923090286766964669825860645425e-16
(0x0.fffffffffffff8p-51)]
But as can be seen on
double f1 (double eps) { double d =3D 1. + eps; if (d =3D=3D 1.) return eps=
; return
0.0; }
int
main ()
{
  __builtin_printf ("%.32a\n", f1 (-0x0.fffffffffffff8p-52));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter
(-0x0.fffffffffffff8p-52, -42.0)));
  __builtin_printf ("%.32a\n", f1 (0x0.fffffffffffff8p-51));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter (0x0.fffffffffffff8p=
-51,
42.0)));
  __builtin_printf ("%.32a\n", f1 (-0x1.0p-54));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter (-0x1.0p-54, -42.0))=
);
  __builtin_printf ("%.32a\n", f1 (0x1.0p-53));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter (0x1.0p-53, 42.0)));
  __builtin_printf ("%.32a\n", f1 (-0x0.8p-53));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter (-0x0.8p-53, -42.0))=
);
  __builtin_printf ("%.32a\n", f1 (0x0.8p-52));
  __builtin_printf ("%.32a\n", f1 (__builtin_nextafter (0x0.8p-52, 42.0)));
}
the correct exact range is what is written in the comment, i.e. [-0x1.0p-54,
0x1.0p-53]
aka [-0x0.8p-53, 0x0.8p-52].=