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 B4764384403C; Tue,  6 Apr 2021 16:33:27 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org B4764384403C
From: "schnetter at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/99912] Unnecessary / inefficient spilling of AVX2 ymm
 registers
Date: Tue, 06 Apr 2021 16:33:27 +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: 11.0
X-Bugzilla-Keywords: missed-optimization
X-Bugzilla-Severity: normal
X-Bugzilla-Who: schnetter at gmail dot com
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-99912-4-Pdp9IsgLZK@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-99912-4@http.gcc.gnu.org/bugzilla/>
References: <bug-99912-4@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: Tue, 06 Apr 2021 16:33:27 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D99912

--- Comment #5 from Erik Schnetter <schnetter at gmail dot com> ---
As you suggested, the problem is probably not caused by register spills, bu=
t by
stores into a struct that are not optimized away. In this case, the respect=
ive
struct elements are unused in the code.

I traced the results of the first __builtin_ia32_maskloadpd256:

  _63940 =3D __builtin_ia32_maskloadpd256 (_63955, prephitmp_86203);
  MEM <const vector(4) double> [(struct mat3 *)&vars + 992B] =3D _63940;
  _178613 =3D .FMA (_63940, _64752, _178609);
  MEM <const vector(4) double> [(struct mat3 *)&vars + 1312B] =3D _63940;

The respective struct locations (+ 992B, + 1312B) are indeed not used anywh=
ere
else.

The struct is of type z4c_vars. It (and its parent) are defined in lines 27=
9837
to 280818. It is large.

Is there e.g. a parameter I could set to make GCC try harder avoid unnecess=
ary
stores?=