linux-kernel - Re: [PATCH 1/3] [RESEND] ARM: pass -march=armv7-a when building NEON files with clang

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <alpine.LFD.2.20.1602191205000.13632@knanqh.ubzr>
Date:	Fri, 19 Feb 2016 12:08:23 -0500 (EST)
From:	Nicolas Pitre <nicolas.pitre@...aro.org>
To:	Arnd Bergmann <arnd@...db.de>
cc:	Russell King <linux@....linux.org.uk>,
	linux-arm-kernel@...ts.infradead.org,
	Ard Biesheuvel <ard.biesheuvel@...aro.org>,
	Jon Medhurst <tixy@...aro.org>,
	Marc Zyngier <marc.zyngier@....com>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/3] [RESEND] ARM: pass -march=armv7-a when building NEON
 files with clang

On Fri, 19 Feb 2016, Arnd Bergmann wrote:

> On Thursday 18 February 2016 12:31:35 Nicolas Pitre wrote:
> > On Thu, 18 Feb 2016, Arnd Bergmann wrote:
> > 
> > > clang ignores the -mfpu=neon flag when building with -march=armv6:
> > > 
> > > In file included from lib/raid6/neon1.c:27:
> > > clang/3.8.0/include/arm_neon.h:28:2: error: "NEON support not enabled"
> > > 
> > > There is no real need to build the file with -march=armv6 in a
> > > multi-CPU enabled kernel, as nothing in here will ever get called
> > > on an ARMv6 CPU. Adding -march=armv7 doesn't hurt and can only
> > > improve the code quality.
> > 
> > Is this enough to override a previous -mfpu for gcc?
> 
> I did not see any build failures on gcc with this, so I assume it
> has no effect. I could move the -march=armv7-a in front of
> -mfpu=neon if you think that would be safer though.

I don't know what's safer.  That's why I'm asking.  :-)


Nicolas