From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Petazzoni Date: Wed, 14 Nov 2012 09:21:03 +0100 Subject: [Buildroot] libgomp.so.1.0.0 not in target image In-Reply-To: <50A2D27A.5000706@mind.be> References: <50A284A8.9070304@mind.be> <20121113231021.6995bd28@skate> <50A2D27A.5000706@mind.be> Message-ID: <20121114092103.5f057441@skate> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net Arnout, On Wed, 14 Nov 2012 00:06:34 +0100, Arnout Vandecappelle wrote: > On 13/11/12 23:10, Thomas Petazzoni wrote: > > Arnout, > > > > On Tue, 13 Nov 2012 18:34:32 +0100, Arnout Vandecappelle wrote: > > > >> Thomas, I think this should be fixed by adding libgomp to > >> LIB_EXTERNAL_LIBS in ext-tool.mk. Is that right? Will that also > >> work if libgomp doesn't exist? > > > > Do we really need to be linked against libgomp? Under what > > conditions does that happen? Can't we avoid that even if the > > toolchain has OpenMP support? > > beecrypt links agains libgomp if libgomp is available; rpm uses > beecrypt. > > The point is not avoiding to link with libgomp (you have to do that > explicitly), the point is that if somebody links against libgomp, it > has to be in target. My point is precisely to avoid linking with libgomp. OpenMP in beecrypt is only used in one location, the RSA algorithm, and OpenMP support can be disabled by passing --disable-openmp. So instead of unconditionally copying a 232 KB library just to parallelize a single RSA algorithm in beecrypt, I'd rather advocate to disable OpenMP support in beecrypt. Best regards, Thomas -- Thomas Petazzoni, Free Electrons Kernel, drivers, real-time and embedded Linux development, consulting, training and support. http://free-electrons.com