From mboxrd@z Thu Jan 1 00:00:00 1970 From: Randy Dunlap Subject: Re: mmotm 2014-07-02-15-07 uploaded (stack protector) Date: Wed, 02 Jul 2014 18:42:53 -0700 Message-ID: <53B4B51D.2060508@infradead.org> References: <53b482d3.gR7nYB/K7hPREviI%akpm@linux-foundation.org> <53B493E2.2030905@infradead.org> <20140703005719.GJ5412@outflux.net> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Return-path: Received: from bombadil.infradead.org ([198.137.202.9]:52344 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753175AbaGCBmz (ORCPT ); Wed, 2 Jul 2014 21:42:55 -0400 In-Reply-To: <20140703005719.GJ5412@outflux.net> Sender: linux-next-owner@vger.kernel.org List-ID: To: Kees Cook Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-next@vger.kernel.org, sfr@canb.auug.org.au, mhocko@suse.cz, Andi Kleen On 07/02/14 17:57, Kees Cook wrote: > Hi Randy, > > And to explain more specifically what's happening: > > On Wed, Jul 02, 2014 at 04:21:06PM -0700, Randy Dunlap wrote: >> config-r8736:CONFIG_CC_STACKPROTECTOR_STRONG=y >> [...] >> build-r8736.out:/local/lnx/MM/mmotm-2014-0702-1507/Makefile:662: Cannot use CONFIG_CC_STACKPROTECTOR_STRONG: -fstack-protector-strong not supported by compiler >> build-r8736.out:/local/lnx/MM/mmotm-2014-0702-1507/Makefile:662: Cannot use CONFIG_CC_STACKPROTECTOR_STRONG: -fstack-protector-strong not supported by compiler >> gcc: error: unrecognized command line option '-fstack-protector-strong' > > That makes sense: a feature that requires a specific compiler version was > chosen, so you get a warning and a failed build. (The first warning was > during the .config parsing pass, the second warning was during the binary > build pass.) > >> config-r8737:# CONFIG_CC_STACKPROTECTOR_STRONG is not set >> [...] >> build-r8737.out:/local/lnx/MM/mmotm-2014-0702-1507/Makefile:662: Cannot use CONFIG_CC_STACKPROTECTOR_STRONG: -fstack-protector-strong not supported by compiler > > This is the same issue: the prior .config was still there with STRONG set, > but after finishing the new .config parsing pass, everything is happy for > the binary build pass. And, I assume, the r8737 didn't fail to build. True, r8737 built successfully. Maybe I should use 'make mrproper' instead of 'make clean' between each randconfig build. Thanks. > > The trouble seems to be how kbuild handle .config changes, but I was never > able to sort it out. -- ~Randy