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 3492F3858CDA; Wed, 26 Apr 2023 10:07:28 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 3492F3858CDA
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1682503648;
	bh=f5JEnT+LYZs+TmdpdI7Qf1w/ffOsj65I0OvVS1D0uq8=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=iUu3owIj2glXnY12Rtnm76qmhZF3ddzvZ10M1TwlJl32ZPQMPuU/ChdzkaDYDzrKH
	 QB3fIQMplnv6zzpO68HebdENoKIsM8FBOY2T4BMl2/Kx4IxiwtYzj7h8rcSJams9Mf
	 qUSZa8E0CWlBijc8FCAcBhfjyT65kIJbsXUx8geM=
From: "jakub at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/109154] [13/14 regression] jump threading
 de-optimizes nested floating point comparisons
Date: Wed, 26 Apr 2023 10:07:25 +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: 13.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: jakub at gcc dot gnu.org
X-Bugzilla-Status: NEW
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: tnfchris at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 13.2
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-109154-4-FLFZjRQUD1@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-109154-4@http.gcc.gnu.org/bugzilla/>
References: <bug-109154-4@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=3D109154

--- Comment #58 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
As a different testcase showing what still needs to be done is e.g.
void
foo (int *p, int *q, int *r, int *s, int *t, int *u)
{
  #pragma omp simd
  for (int i =3D 0; i < 1024; i++)
    {
      int vp =3D p[i];
      int vq =3D q[i];
      int vr =3D r[i];
      int vs =3D s[i];
      int vt =3D t[i];
      int vu =3D u[i];
      int vw;
      if (vp !=3D 0)
        {
          if (vp > 100)
            {
              if (vq < 200)
                vw =3D 1;
              else if (vr)
                vw =3D 2;
              else
                vw =3D 3;
            }
          else if (vs > 100)
            {
              if (vq < 180)
                vw =3D 4;
              else if (vr > 20)
                vw =3D 5;
              else
                vw =3D 6;
            }
          else
            {
              if (vq < -100)
                vw =3D 7;
              else if (vr < -20)
                vw =3D 8;
              else
                vw =3D 9;
            }
        }
      else if (vt > 10)
        {
          if (vu > 100)
            vw =3D 10;
          else if (vu < -100)
            vw =3D 11;
          else
            vw =3D 12;
        }
      else
        vw =3D 13;
      u[i] =3D vw;
    }
}
with -O2 -fopenmp-simd.
I think we still need 12 VEC_COND_EXPRs to merge it all together, but if we
follow what the source is doing (or rediscover it), we can certainly
avoid so many useless &s on the conditions by merging it together in the ri=
ght
order.=