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 191FF3858414; Mon, 28 Nov 2022 17:26:56 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 191FF3858414
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1669656416;
	bh=o0E07BD8q8bYRz+TePb7ZFDgDOd8UPc8sp4i6qqf58c=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=A1zSGYdCSuj+aCl/NItJDTlBlkEXGWaAk0qrP+Lfj88SEIntkGzcCipipYbY9Ef2M
	 hbd+UzyIJUPhfRc3UAHHcg98tzIE/+AhuGYc/lkbG/8p6dZVKFn9Fq8s8veD+DJB9X
	 DAtvRyPBig6cNcM8IyuDLmPsOZ4pSeaRjPNfumco=
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/107888] [12/13 Regression] Missed min/max
 transformation in phiopt due to VRP
Date: Mon, 28 Nov 2022 17:26:55 +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: 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: 12.3
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-107888-4-EADSk1m9eC@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-107888-4@http.gcc.gnu.org/bugzilla/>
References: <bug-107888-4@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=3D107888

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #1)
> which means we fail to optimize a > b ? 1 : b as well, no?

Yes that is correct.

Even for max, "a >=3D b ? a : 6;" would need to be "reverted" 6 back to b.

  <bb 5> [local count: 1073741824]:
  if (ab_2(D) >=3D bb_3(D))
    goto <bb 6>; [65.00%]
  else
    goto <bb 7>; [35.00%]

  <bb 6> [local count: 697932184]:

  <bb 7> [local count: 1073741824]:
  # c_1 =3D PHI <ab_2(D)(6), 6(5)>


The min/max patterns inside match needs to handle CST if the ranges of the =
two
operands overlap with one/two values.

Even though this is a regression, I don't know if this shows up in real code
and is a small optimization really so I would suspect a P4 for this really =
as
it requires a bigger change that most likely won't be backported. I filed i=
t as
it was showing up while I was working on the patch for PR 101805 (which won=
't
be submitted until GCC 14 anyways).=