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 C40383858D35; Sat, 24 Jun 2023 22:12:43 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org C40383858D35
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1687644763;
	bh=/6ewfJr5bI+H4ZbljyV12/pioQFAIjW4Oa1TSw7lhdU=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=wpLDleoMctKvpHQ2u0WJUkThAlTTQQ0mkmLd6WwX9XQFcbqANQMEV2yTwVXAJHISg
	 pM7cMpW6b33DpKCcAf9fsWgxlhxb2LSMoHWZOxmg7RIG4OiGllvNXInNjOAPzbHGok
	 +xyExY4oUZNUo0Id9t2b033Z4d/9pELYxZAM7xrM=
From: "vanyacpp at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug middle-end/109986] missing fold (~a | b) ^ a => ~(a & b)
Date: Sat, 24 Jun 2023 22:12:43 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: middle-end
X-Bugzilla-Version: 14.0
X-Bugzilla-Keywords: easyhack, missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: vanyacpp at gmail dot com
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: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-109986-4-f9Ghe8NKQW@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-109986-4@http.gcc.gnu.org/bugzilla/>
References: <bug-109986-4@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=3D109986

--- Comment #3 from Ivan Sorokin <vanyacpp at gmail dot com> ---
I tried to investigate why GCC is able to simplify `(a | b) ^ a` and `(a | =
~b)
^ a` from comment 2, but not similarly looking `(~a | b) ^ a` from comment =
0.

`(a | b) ^ a` matches the following pattern from match.pd:

/* (X | Y) ^ X -> Y & ~ X*/
(simplify
 (bit_xor:c (convert1? (bit_ior:c @@0 @1)) (convert2? @0))
 (if (tree_nop_conversion_p (type, TREE_TYPE (@0)))
  (convert (bit_and @1 (bit_not @0)))))

`(a | ~b) ^ a` matches another pattern:

/* (~X | C) ^ D -> (X | C) ^ (~D ^ C) if (~D ^ C) can be simplified.  */
(simplify
 (bit_xor:c (bit_ior:cs (bit_not:s @0) @1) @2)
  (bit_xor (bit_ior @0 @1) (bit_xor! (bit_not! @2) @1)))

With substitution `X =3D b, C =3D a, D =3D a` it gives:

(b | a) ^ (~a ^ a)
(b | a) ^ -1
~(b | a)

`(~a | b) ^ a` is not simplifiable by this pattern because it requires that=
 `~D
^ C` is simplifiable further, but `~a ^ b` is not. In any case, even if it =
were
applicable it would produce `(a | b) ^ (~a ^ b)` which has more operations =
than
the original expression.=