From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <martin@MPA-Garching.MPG.DE>
Received: from srv-08.MPA-Garching.MPG.DE (srv-08.MPA-Garching.MPG.DE
 [130.183.82.27])
 by sourceware.org (Postfix) with ESMTPS id 89E97385701F
 for <gcc-help@gcc.gnu.org>; Tue, 23 Mar 2021 13:48:49 +0000 (GMT)
DMARC-Filter: OpenDMARC Filter v1.3.2 sourceware.org 89E97385701F
Authentication-Results: sourceware.org; dmarc=none (p=none dis=none)
 header.from=MPA-Garching.MPG.DE
Authentication-Results: sourceware.org;
 spf=pass smtp.mailfrom=martin@MPA-Garching.MPG.DE
Received: from [192.168.178.60] (aftr-62-216-202-20.dynamic.mnet-online.de
 [62.216.202.20]) (authenticated bits=0)
 by srv-08.MPA-Garching.MPG.DE (8.15.2/8.15.2) with ESMTPSA id 12NDml4c045367
 (version=TLSv1.3 cipher=TLS_AES_256_GCM_SHA384 bits=256 verify=NO);
 Tue, 23 Mar 2021 14:48:48 +0100
Subject: Re: g++: Suboptimal code generation for simple wrapper class around
 vector data type
To: Alexander Monakov <amonakov@ispras.ru>
Cc: gcc-help <gcc-help@gcc.gnu.org>
References: <af3d9ce6-b144-df55-883c-1958c0153358@mpa-garching.mpg.de>
 <b04719d9-594f-da2a-f498-859767767f76@mpa-garching.mpg.de>
 <alpine.LNX.2.20.13.2103231621320.25152@monopod.intra.ispras.ru>
From: Martin Reinecke <martin@MPA-Garching.MPG.DE>
Message-ID: <0673ddd7-4d96-393b-4370-66c43ca2872e@mpa-garching.mpg.de>
Date: Tue, 23 Mar 2021 14:48:47 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101
 Thunderbird/78.8.0
MIME-Version: 1.0
In-Reply-To: <alpine.LNX.2.20.13.2103231621320.25152@monopod.intra.ispras.ru>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.6.2
 (srv-08.MPA-Garching.MPG.DE [130.183.82.27]);
 Tue, 23 Mar 2021 14:48:48 +0100 (CET)
X-Spam-Status: No, score=1.6 required=5.0 tests=BAYES_00, KAM_DMARC_STATUS,
 NICE_REPLY_A, RCVD_IN_BARRACUDACENTRAL, SPF_HELO_NONE, SPF_PASS,
 TXREP autolearn=no autolearn_force=no version=3.4.2
X-Spam-Level: *
X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on
 server2.sourceware.org
X-BeenThere: gcc-help@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-help mailing list <gcc-help.gcc.gnu.org>
List-Unsubscribe: <https://gcc.gnu.org/mailman/options/gcc-help>,
 <mailto:gcc-help-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-help/>
List-Post: <mailto:gcc-help@gcc.gnu.org>
List-Help: <mailto:gcc-help-request@gcc.gnu.org?subject=help>
List-Subscribe: <https://gcc.gnu.org/mailman/listinfo/gcc-help>,
 <mailto:gcc-help-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Tue, 23 Mar 2021 13:48:50 -0000

Dear Alexander,

thanks a lot for the reply!

> Yes. Note that _m256d is declared with 'may_alias' attribute, so in general
> it should weaken optimizations by disabling type-based alias analysis,
> but in this case replacing _m256d with __v4df does not help. 
> 
> The following version should be easier to optimize, but still is not handled
> well (at -O2 -mfma). I'd suggest to present it in the PR in addition to your
> reduced version.

I have opened PR c++/99728 for this in the meantime and will add your
test case!

Thanks again,
  Martin