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 864FF3858406; Thu, 23 Sep 2021 07:16:28 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 864FF3858406
From: "rguenther at suse dot de" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/102452] Structs with flexible array members
 are not optimized on stack
Date: Thu, 23 Sep 2021 07:16:28 +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: 9.3.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: rguenther at suse dot de
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-102452-4-RIzGyO1QV3@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-102452-4@http.gcc.gnu.org/bugzilla/>
References: <bug-102452-4@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
X-BeenThere: gcc-bugs@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-bugs mailing list <gcc-bugs.gcc.gnu.org>
List-Unsubscribe: <https://gcc.gnu.org/mailman/options/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-request@gcc.gnu.org?subject=help>
List-Subscribe: <https://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Sep 2021 07:16:28 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D102452

--- Comment #5 from rguenther at suse dot de <rguenther at suse dot de> ---
On Thu, 23 Sep 2021, pinskia at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D102452
>=20
> --- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
> get_three after einline w/o USE_FLEX_ARR defined (this is changing the ty=
pe to
> int):
>=20
>   <bb 2> :
>   _5 =3D test_2(D)->is_a;
>   if (_5 !=3D 0)
>     goto <bb 3>; [50.00%]
>   else
>     goto <bb 4>; [50.00%]
>=20
>   <bb 3> :
>   D.3292 =3D *test_2(D);
>   goto <bb 5>; [100.00%]
>=20
>   <bb 4> :
>   _6 =3D test_2(D)->as.b.one;
>   _7 =3D (long int) _6;
>   _8 =3D test_2(D)->as.b.two;
>   _9 =3D (long int) _8;
>   _10 =3D test_2(D)->as.b.three;
>   _11 =3D (long int) _10;
>   D.3292.as.a.one =3D _7;
>   D.3292.as.a.two =3D _9;
>   D.3292.as.a.three =3D _11;
>=20
>   <bb 5> :
>   _4 =3D D.3292.as.a.three;
>   return _4;
>=20
> With USE_FLEX_ARR defined to 1:
>   <bb 2> :
>   _5 =3D test_2(D)->is_a;
>   if (_5 !=3D 0)
>     goto <bb 3>; [50.00%]
>   else
>     goto <bb 4>; [50.00%]
>=20
>   <bb 3> :
>   D.3293 =3D *test_2(D);
>   goto <bb 5>; [100.00%]
>=20
>   <bb 4> :
>   _6 =3D test_2(D)->as.b.one;
>   _7 =3D (long int) _6;
>   ret.as.a.one =3D _7;
>   _8 =3D test_2(D)->as.b.two;
>   _9 =3D (long int) _8;
>   ret.as.a.two =3D _9;
>   _10 =3D test_2(D)->as.b.three;
>   _11 =3D (long int) _10;
>   ret.as.a.three =3D _11;
>   D.3293 =3D ret;
>   ret =3D{v} {CLOBBER};
>=20
>   <bb 5> :
>   _4 =3D D.3293.as.a.three;
>   return _4;
>=20
>=20
> as far as I can tell the IR is the same before that,
> even RSO:
>   D.3293 =3D make_test_a (test_2(D)); [return slot optimization]
>=20
>=20
> But it looks in the case of the flex array case, RSO does not actually ha=
ppen.

Are you sure it's not SRA behaving differently?  The IL inlined early
is only showing in later dumps...=