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 4C2BD385841E; Thu, 29 Jun 2023 15:13:57 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 4C2BD385841E
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org;
	s=default; t=1688051637;
	bh=z0il/R2H64IigSrmoJdGQUdYGzuLwc7R50WyieLo25g=;
	h=From:To:Subject:Date:In-Reply-To:References:From;
	b=IL/rSKQn+5KKKVr8gxdheAVWKFO7OSroFRKHmnsoHSw67Kok+vMJJW0Gsl+zwiz4i
	 u2nd/rrEu/lSD4tRYHAhzulh5StraJiUtDX7TE4r5F7J1pD2QACSVhHNLsEKswn2vI
	 BvUSBisvn0Mj5mDXBlG3kkrcL6mr1mTHiLjdJD4E=
From: "ryanpholt at me dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug middle-end/110472] 60% slowdown with fwrapv when using openmp
Date: Thu, 29 Jun 2023 15:13:53 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: middle-end
X-Bugzilla-Version: 10.2.1
X-Bugzilla-Keywords: missed-optimization, openmp
X-Bugzilla-Severity: normal
X-Bugzilla-Who: ryanpholt at me 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-110472-4-PfJhkzLImm@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-110472-4@http.gcc.gnu.org/bugzilla/>
References: <bug-110472-4@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=3D110472

--- Comment #2 from Ryan Holt <ryanpholt at me dot com> ---
(In reply to Andrew Pinski from comment #1)
> I think it is just wrong iv-opt choices.
>=20
> Works just fine on aarch64-linux-gnu too:
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ~/upstream-gcc/bin/gcc t4.c -O2
> -fopenmp
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ./a.out ;echo
> time: 15.191220
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ~/upstream-gcc/bin/gcc t4.c -O2
> -fopenmp -fwrapv
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ./a.out ;echo
> time: 18.854280
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ./a.out ;echo
> time: 16.705876
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ~/upstream-gcc/bin/gcc t4.c -O2
> -fopenmp
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ./a.out ;echo
> time: 17.491387
> ubuntu@ubuntu:~/src/upstream-gcc-aarch64\# ./a.out ;echo
> time: 17.519264

I forgot to explicitly call out that I can only reproduce the big speedup
without fwrapv when compiling with -O3. I noticed you were using O2 on
aarch64-linux-gnu.=