Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT sourceware DOT cygnus DOT com Delivered-To: mailing list cygwin-developers AT sourceware DOT cygnus DOT com Message-Id: <200002081706.LAA06484@hp2.xraylith.wisc.edu> To: Chris Faylor cc: cygwin-developers AT sourceware DOT cygnus DOT com Subject: Re: Revisiting libcygwin.a/libc.a/libm.a In-reply-to: Your message of "Tue, 08 Feb 2000 11:36:55 EST." <20000208113655 DOT A4801 AT cygnus DOT com> Date: Tue, 08 Feb 2000 11:06:23 -0600 From: Mumit Khan Chris Faylor writes: > Would it be a good idea to eliminate the separate lib[cm].a for the next > net release? I've already got a modified Makefile for newlib and cygwin > which links libcygwin.a to libc.a and libm.a. > > The last I checked, I think that there were inexplicable problems > with constructors when you linked libcygwin.a to libc.a. > > Mumit, do you remember this? Do you have any further insight into the > problem? Do you still think that it's a good idea? Remember it way too well. I did do some tests since and it seemed to work with v1.0 CD version. I will check again this evening. I for one consider this to be the right thing, even with the various problems/misfeatures others have pointed out in the past (eg., if it's symlink, it's not usable by native tools, but I believe the advantage outweighs the problem here). One problem we have now is that most build tools add -lm if math library is needed, and we end up using statically linked versions of the math routines. More this evening. Regards, Mumit