From mboxrd@z Thu Jan  1 00:00:00 1970
From: Segher Boessenkool <segher@kernel.crashing.org>
Subject: Re: [cryptodev:master 130/134] aes_generic.c:undefined reference to `_restgpr_31_x'
Date: Sun, 14 Jan 2018 18:21:09 -0600
Message-ID: <20180115002109.GX21977@gate.crashing.org>
References: <201801122249.RHvvGQJ6%fengguang.wu@intel.com> <CAK8P3a0Rfbo=wJgEAm3vVFY1MpPWLE3HMGcUdoOXxqOVm8jRyw@mail.gmail.com> <20180112163946.GE21977@gate.crashing.org> <CAK8P3a39=PVpwiSk_3CjRnYBawhXqMatHKSsHCd2KHCgR+Zmyg@mail.gmail.com> <20180112204154.GM21977@gate.crashing.org> <CAK8P3a2YVVseBFNoDbwOtqR3-JHFuZ-WZD6-PzztNTO_X_soaw@mail.gmail.com> <20180112214111.GQ21977@gate.crashing.org> <CAK8P3a3XKwrVpXH9=8n1=qAKtqHE+1sh6OhaJ_39GZcWSP-j=Q@mail.gmail.com> <20180112221050.GR21977@gate.crashing.org> <CAK8P3a30k-JjMJ0M9KWXqUMT3eHmuuFvAe1KuKdqNoAiSZrgog@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: "open list:HARDWARE RANDOM NUMBER GENERATOR CORE"
        <linux-crypto@vger.kernel.org>,
        kbuild test robot <fengguang.wu@intel.com>,
        linuxppc-dev <linuxppc-dev@lists.ozlabs.org>,
        Herbert Xu <herbert@gondor.apana.org.au>
To: Arnd Bergmann <arnd@arndb.de>
Return-path: <linux-crypto-owner@vger.kernel.org>
Received: from gate.crashing.org ([63.228.1.57]:46406 "EHLO gate.crashing.org"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1752205AbeAOAVz (ORCPT <rfc822;linux-crypto@vger.kernel.org>);
        Sun, 14 Jan 2018 19:21:55 -0500
Content-Disposition: inline
In-Reply-To: <CAK8P3a30k-JjMJ0M9KWXqUMT3eHmuuFvAe1KuKdqNoAiSZrgog@mail.gmail.com>
Sender: linux-crypto-owner@vger.kernel.org
List-ID: <linux-crypto.vger.kernel.org>

On Sun, Jan 14, 2018 at 10:40:36PM +0100, Arnd Bergmann wrote:
> Right. I've done some more investigation anyway, starting over with the
> analysis of the gcc options that change it. I've found now that turning
> off '-fcode-hoisting' but leaving on the other options I had suspected
> earlier (-O2 instead of -Os, -ftree-sra, -ftree-pre) also fixes the
> stack problem, and appears to result in the best performance so
> far.

Oh nice!

> I need to rerun the whole test matrix, but that seems rather
> promising, and the result may also help debug what's really happening.

-fcode-hoisting moves all expression evaluation to as early as possible;
for this AES code that means it will increase register pressure a lot,
causing a lot of spilling (well, that is my guess).  If that is so, then
we need to dial down -fcode-hoisting a bit, maybe make it aware of
register pressure.

Glad you found a smoking gun,


Segher