From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Korsgaard Date: Wed, 15 Jun 2011 09:09:52 +0200 Subject: [Buildroot] [PATCH 7/8] package/autotools: add --{enable, disable}-{shared, static} automatically In-Reply-To: <20110615084756.5d787906@skate> (Thomas Petazzoni's message of "Wed, 15 Jun 2011 08:47:56 +0200") References: <87k4cq7rdq.fsf@macbook.be.48ers.dk> <20110613001630.1281e5c3@skate> <874o3t9zs2.fsf@macbook.be.48ers.dk> <20110614085430.3229006e@skate> <87wrgn8usg.fsf@macbook.be.48ers.dk> <20110615084756.5d787906@skate> Message-ID: <87r56v8ta7.fsf@macbook.be.48ers.dk> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: buildroot@busybox.net >>>>> "Thomas" == Thomas Petazzoni writes: Hi, Thomas> Ok. Thanks a lot for doing this testing! >> >> Another more serious issue is that some non-library packages like bmon >> interprete --enable-static as link the binary statically, leading to >> bloat (E.G. ~900KB instead of ~100KB). Thomas> Is --enable-static supposed to have this semantic ? If not, then we Thomas> should fix the package. If yes, then we can fix those packages manually Thomas> by adding --disable-static to _CONF_OPT, no ? I haven't been able to find an official specification of the --enable-static semantic, but so far I've only noticed bmon having this issue, so it might be local - It does kind of make sense though. Thomas> Or maybe this isn't worth the trouble after all, and we could just use Thomas> --disable-static by default. I'm not personally using static linking, Thomas> so adding the production of .a files was just done because it appeared Thomas> to be cost-free and could have been useful to some users. So I wouldn't Thomas> complain if we decide to revert back to --disable-static for target Thomas> packages. Ok. I'll test a bit more and see if I find other packages misbehaving as well. If I don't then I'l just add the workaround for bmon. -- Bye, Peter Korsgaard