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 6DA993858D35; Mon, 15 Apr 2024 08:18:06 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 6DA993858D35
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1713169086;
	bh=eKuErWQvq8IR5Too+SKgo7Rv40ELg/itA1HymG3zcuY=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=CD/Cx9I/8hv0FyKLuN4wt0sGiS1xVM8lBuRTlZRgNJ2IcpSS2Gf8ImbNPJpphobLg
	 OpsLWCmy5vXfpGUU9Bv9qJqe6QtuLHBQ2ohXEjb/nxholy4PIg4/kvaFpzmuIoX8KG
	 3F0ZVtEWS7KFqKC2h6FdaLrVGdskY9ZdVZgg2G08=
From: "rguenther at suse dot de" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/114635] OpenMP reductions fail dependency
 analysis
Date: Mon, 15 Apr 2024 08:18:02 +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, openmp
X-Bugzilla-Severity: normal
X-Bugzilla-Who: rguenther at suse dot de
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: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-114635-4-dmeHmMndrD@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-114635-4@http.gcc.gnu.org/bugzilla/>
References: <bug-114635-4@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=3D114635

--- Comment #17 from rguenther at suse dot de <rguenther at suse dot de> ---
On Mon, 15 Apr 2024, jakub at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D114635
>=20
> --- Comment #16 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
> (In reply to Richard Biener from comment #14)
> > I think
> >=20
> >   if (safelen)
> >     {
> >       poly_uint64 val;
> >       safelen =3D OMP_CLAUSE_SAFELEN_EXPR (safelen);
> >       if (!poly_int_tree_p (safelen, &val))
> >         safelen_int =3D 0;
> >       else
> >         safelen_int =3D MIN (constant_lower_bound (val), INT_MAX);
> >=20
> > should simply become
> >=20
> >         safelen_int =3D constant_upper_bound_with_limit (val, INT_MAX);
> >=20
> > ?  Usually targets do have a limit on the actual length but I see
> > constant_upper_bound_with_limit doesn't query such.  But it would
> > be a more appropriate way to say there might be an actual target limit =
here?
>=20
> OMP_CLAUSE_SAFELEN_EXPR is always an INTEGER_CST, the FEs verify that and=
 error
> if it is not.  So, I must say I don't really understand parts of the
> r8-5649-g9d2f08ab97be
> changes.  I can understand the intent to make max_vf a poly_int, but don't
> understand why safelen should be, what would it mean and when it would be=
 set
> that way?

It would be only to "better" encode "infinity".  But I see loop->safelen
is 'int' but only [0, MAX_INT] is specified so we'd conveniently have
say -1 to encode "infinity".  We'd have to special case that value
anyway?=