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 ACA873858D39; Wed, 28 Feb 2024 16:51:53 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org ACA873858D39
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1709139113;
	bh=BhuXZ82lJO9M/jC2YKBhbM8wIVIJ4rIDsbqV3O9oOEg=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=oexJOXatpdUlR5Pw6pkrbFok/4vrL9A7RDFO53Z49eO5bmN47dESzPV4lMuWjVtUE
	 ssslzsNoVo/woMN6HrzpBe6fJPpvS4czmO3/WMz8MV6PXJRn3iguILWW5aw2hiIdXc
	 cS4FLNVASvnzz3oKkTc4bI6hvSuSp9Ggq4IA2Ydw=
From: "tnfchris at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/114151] [14 Regression] weird and inefficient
 codegen and addressing modes since g:a0b1798042d033fd2cc2c806afbb77875dd2909b
Date: Wed, 28 Feb 2024 16:51:53 +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: 14.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: tnfchris 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: 14.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-114151-4-cyh91423Io@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-114151-4@http.gcc.gnu.org/bugzilla/>
References: <bug-114151-4@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=3D114151

--- Comment #3 from Tamar Christina <tnfchris at gcc dot gnu.org> ---
>=20
> This was a correctness fix btw, so I'm not sure we can easily recover - we
> could try using niter information for CHREC_VARIABLE but then there's
> variable niter here so I don't see a chance.
>=20

It looks like it's mostly SVE suffering here. Adv. SIMD an scalar codegen s=
eems
to have vastly improved with it. we see ~10% improvements due to better
addressing for scalar.

It also only happens at -O3 but -O2 is fine, which is weird as I was expect=
ed
IVopts to be enabled at -O2 too.

>=20
> OTOH the +1 could make it overflow for large size.
>=20
> Can you test the above?  It should be an incremental improvement.
>=20

Applying the changes does not seem to make a difference for the final codeg=
en
:(=