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 D1ED33858D38; Wed, 11 Jan 2023 13:03:16 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org D1ED33858D38
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1673442196;
	bh=l8qVhgwb+/304j4vb8BfVYv5zQUHV1SN025VJ6SCWRI=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=yGm+3LmieowsoH9P9btr1Tq41yR1MFWxGLOsdrqMzU76x0X4hcYFrk6YJnOcv9vC7
	 pYrFzVSufHq26WHLg1OjYjYXi/RQdsCTCZlYfEOiEydYhHC5h1DrrHlNjdsdzwCxBA
	 G3orlzpaRJwEivls4Mrf46WIluc6LP6v73Ye2GDA=
From: "rguenth at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/105554] [10/11/12/13 Regression] ICE: in
 emit_block_move_hints, at expr.cc:1829 since r9-5509-g5928bc2ec06dd4e7
Date: Wed, 11 Jan 2023 13:03:16 +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: 13.0
X-Bugzilla-Keywords: ice-on-valid-code
X-Bugzilla-Severity: normal
X-Bugzilla-Who: rguenth at gcc dot gnu.org
X-Bugzilla-Status: ASSIGNED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P2
X-Bugzilla-Assigned-To: marxin at gcc dot gnu.org
X-Bugzilla-Target-Milestone: 10.5
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-105554-4-BwCIvVhY1x@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-105554-4@http.gcc.gnu.org/bugzilla/>
References: <bug-105554-4@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=3D105554

--- Comment #8 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #7)
> (In reply to Martin Li=C5=A1ka from comment #6)
> > We fail in the param assignment:
> >=20
> > (gdb) pp x
> > (reg:V4DI 82)
> > (gdb) pp y
> > (mem/c:BLK (reg/f:DI 76 virtual-incoming-args) [1 x+0 S32 A256])
> >=20
> > So we will likely need something similar to what we have in tree-inline=
.cc:
> >=20
> >   5928    /* For vector typed decls make sure to update DECL_MODE accor=
ding
> >   5929       to the new function context.  */
> >   5930    if (VECTOR_TYPE_P (TREE_TYPE (copy)))
> >   5931      SET_DECL_MODE (copy, TYPE_MODE (TREE_TYPE (copy)));
> >=20
> > @Richi: Do you have a clue where to adjust it?
>=20
> I think it goes wrong in use_register_for_decl (called from
> assign_parm_setup_block).
>=20
> diff --git a/gcc/function.cc b/gcc/function.cc
> index d975b001ec9..b54f9b33c6a 100644
> --- a/gcc/function.cc
> +++ b/gcc/function.cc
> @@ -2229,7 +2229,9 @@ use_register_for_decl (const_tree decl)
>      }
>=20=20
>    /* Only register-like things go in registers.  */
> -  if (DECL_MODE (decl) =3D=3D BLKmode)
> +  if (DECL_MODE (decl) =3D=3D BLKmode
> +      || (VECTOR_TYPE_P (TREE_TYPE (decl))
> +         && TYPE_MODE (TREE_TYPE (decl)) =3D=3D BLKmode))
>      return false;
>=20=20
>    /* If -ffloat-store specified, don't put explicit float variables
>=20
> fixes the ICE, not sure if we should adjust the PARM_DECLs mode somewhere
> in target cloning instead though?

Like in copy_arguments_nochange?=