X-Recipient: archive-cygwin AT delorie DOT com X-SWARE-Spam-Status: No, hits=-2.5 required=5.0 tests=AWL,BAYES_00,SARE_SUB_ENC_UTF8,SPF_HELO_PASS,SPF_PASS X-Spam-Check-By: sourceware.org To: cygwin AT cygwin DOT com From: Eric Blake Subject: Re: Updated: =?utf-8?b?bGlic2lnc2VndnsyfTIuOC0x?= Date: Tue, 15 Dec 2009 20:54:27 +0000 (UTC) Lines: 43 Message-ID: References: <4B266F13 DOT 6070305 AT x-ray DOT at> <4B2788CB DOT 3050306 AT byu DOT net> <4B27DDAB DOT 3080503 AT x-ray DOT at> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit User-Agent: Loom/3.14 (http://gmane.org/) X-IsSubscribed: yes Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Id: List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com Reini Urban x-ray.at> writes: > >> The DLL revision is bumped from 1 to 2. > > > > I'm curious why you bumped the dll revision when building 2.8. I don't > > see any backwards-incompatible API changes compared to 2.6 that would have > > required the bump; am I overlooking something? But the barn door is open, > > now that you have released clisp depending on the new dll revision. So > > I'll now have to rebuild m4 to pick up the new dll version. > > Me neither. It was not me, it was Bruno Haible, upstream. > But there must be some changes. > LIBSIGSEGV_VERSION_INFO = 2:1:0 > 2.7 had 2:0:0 > And 2.6++ started with 1:0:0 According to the libtool versioning scheme, http://www.gnu.org/software/libtool/manual/libtool.html#Updating-version-info, the change from 2.6 (1:0:0) to 2.7 (2:0:0) requires a dll version bump on the surface (the 'current' field changed from 1 to 2 to mark an interface change, and the 'age' field reset to 0 meaning that the change is not backwards- compatible). But reading the NEWS file, the only documented change was Linux- specific ("the type 'stackoverflow_context_t' is now typedefed to 'ucontext_t *' rather than 'struct sigcontext *'"), so we probably could have survived without a dll rev bump. Meanwhile, the change from 2.7 (2:0:0) to 2.8 (2:1:0) implies complete backwards compatibility (the 'revision' field changed from 0 to 1 to reflect that code changed to fix bugs, but neither 'current' nor 'age' changed so nothing was backwards-incompatible). And the NEWS appears to back this up, with no mention of any interface changes. But certainly an argument for leaving the dll version unchanged. Just so I'm clear, is the dll version something you manually pick based on the libtool version changing, or is it something generated automatically by the default build process (either by libtool or by cygport)? I guess that if the version suffix is automatic, based on the fact that the 'current' field changed from 1 to 2 between 2.6 and 2.8, then you did the right thing, even if the interface change that Bruno based his version bump has no effect on cygwin. -- Eric Blake -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple