From mboxrd@z Thu Jan 1 00:00:00 1970 From: arnd@arndb.de (Arnd Bergmann) Date: Fri, 3 Jan 2014 15:49:02 +0100 Subject: [BUG] shmobile build failure In-Reply-To: <20140102232325.GF7383@n2100.arm.linux.org.uk> References: <20140102172413.GB7383@n2100.arm.linux.org.uk> <20140102232325.GF7383@n2100.arm.linux.org.uk> Message-ID: <201401031549.02866.arnd@arndb.de> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Friday 03 January 2014, Russell King - ARM Linux wrote: > On Thu, Jan 02, 2014 at 03:20:36PM -0800, Olof Johansson wrote: > > On Thu, Jan 2, 2014 at 1:47 PM, Russell King - ARM Linux > > > Have you run any allmod or allyesconfig builds recently? The logs from > > > those builds are excessively big and unreadable due to this: > > > > > > warning: __mcount_loc already exists: object-file.o > > > > > > being spat out for every object file being built, and I can't figure out > > > what's causing that. > > > > > > Hm, no I have not. Do you have a feel for when it began? I can do a bisect. > > I don't know - I've only recently re-enabled those builds. There's > suggestions from other architectures that it may be toolchain related - > some people have blamed binutils for it, but I'm running binutils 2.22. > > Hence why I'm interested if others see this too. > I just tried building allmodconfig using linux-next-20131224 and arm-linux-gnueabihf-gcc-4.8.1-7ubuntu1. I had to disable a bunch of drivers to get it to build, but didn't see the __mcount_loc problem. What kernel tree and compiler are you using? Arnd