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 66520385841A; Tue, 25 Jul 2023 06:34:01 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 66520385841A
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1690266841;
	bh=61ZNrisg6c/PWfuJqUWiinWqKBPz3gZixwDhRvZw3zI=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=guItjlapeqS31WRegl5aEJoELvdzztNZ6VaVd/uS5C1N9AoIobm3f0LWIQEiRFVjD
	 yI2wd0XPp+hKZEGWkJJDq7o+IKvXO77iYmgL5yLq3o7bP5D7P2RQ/sj7A+QyVdbBkT
	 h5yuEAGZYimTht3PtDt6hOUXOBFkuWUVN7A+gRJU=
From: "rguenther at suse dot de" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/110776] [14 Regression] powerpc-darwin bootstrap broken
 after r14-2490 with ICE rs6000.cc:5069 building libgfortran
Date: Tue, 25 Jul 2023 06:34:00 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 14.0
X-Bugzilla-Keywords: build, ice-on-valid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: rguenther at suse dot de
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: linkw at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 14.0
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-110776-4-jAV6dWLZ3V@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-110776-4@http.gcc.gnu.org/bugzilla/>
References: <bug-110776-4@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=3D110776

--- Comment #5 from rguenther at suse dot de <rguenther at suse dot de> ---
On Tue, 25 Jul 2023, linkw at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D110776
>=20
> Kewen Lin <linkw at gcc dot gnu.org> changed:
>=20
>            What    |Removed                     |Added
> -------------------------------------------------------------------------=
---
>                  CC|                            |rguenth at gcc dot gnu.o=
rg
>=20
> --- Comment #3 from Kewen Lin <linkw at gcc dot gnu.org> ---
> The failure exposed one special case: there is one stmt=20
>=20
> # VUSE <.MEM_404>
> _15 =3D *_14;
>=20
> its STMT_VINFO_STRIDED_P (stmt_info) is set, memory_access_type is
> VMAT_ELEMENTWISE and alignment_support_scheme is dr_unaligned_supported, =
its
> vector type is "vector(1) long int", so in the handling it's taken as:
>=20
>       /* Load vector(1) scalar_type if it's 1 element-wise vectype.  */
>       else if (nloads =3D=3D 1)
>         ltype =3D vectype;
>=20
> as its ltype is vector type, we cost it with=20
>=20
>                   if (VECTOR_TYPE_P (ltype))
>                     vect_get_load_cost (vinfo, stmt_info, 1,
>                                         alignment_support_scheme, misalig=
nment,
>                                         false, &inside_cost, nullptr, cos=
t_vec,
>                                         cost_vec, true);
>=20
> as it's dr_unaligned_supported, it's costed as unaligned_load then causes=
 the
> ICE. One idea is to teach rs6000_builtin_vectorization_cost about one sin=
gle
> lane vector unaligned load as scalar_load. But I think it's simple to tre=
at
> single lane vector load as scalar_load, as I expect veclower will lower it
> later.
>=20
> diff --git a/gcc/tree-vect-stmts.cc b/gcc/tree-vect-stmts.cc
> index 4622d6a04ef..bdf4c12cd03 100644
> --- a/gcc/tree-vect-stmts.cc
> +++ b/gcc/tree-vect-stmts.cc
> @@ -9985,7 +9985,9 @@ vectorizable_load (vec_info *vinfo,
>              {
>                if (costing_p)
>                  {
> -                  if (VECTOR_TYPE_P (ltype))
> +                  /* For a single lane vector type, we should cost it as
> +                     scalar_load to avoid ICE, see PR110776.  */
> +                  if (VECTOR_TYPE_P (ltype) && lnel > 1)
>                      vect_get_load_cost (vinfo, stmt_info, 1,
>                                          alignment_support_scheme,
> misalignment,
>                                          false, &inside_cost, nullptr,
> cost_vec,
>=20
> Hi Richi, what do you think of this?

I think apart from the consideration what a single element vector
is compared to a scalar, a more to-the-point fix is

  if (VECTOR_TYPE_P (ltype)
      && memory_access_type !=3D VMAT_ELEMENTWISE)=