Date: Tue, 4 Oct 2005 04:50:21 GMT From: "Maxwell N. Krohn" <krohn@MIT.EDU> To: freebsd-ports-bugs@FreeBSD.org Subject: Re: ports/86475: New Ports: devel/sfslite-noopt and devel/sfslite-dbg; sfslite with different build options Message-ID: <200510040450.j944oLoE013074@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
The following reply was made to PR ports/86475; it has been noted by GNATS. From: "Maxwell N. Krohn" <krohn@MIT.EDU> To: Renato Botelho <garga@FreeBSD.org> Cc: bug-followup@FreeBSD.org Subject: Re: ports/86475: New Ports: devel/sfslite-noopt and devel/sfslite-dbg; sfslite with different build options Date: Tue, 4 Oct 2005 00:41:04 -0400 (EDT) Hi Renato, Yes, there is a good reason. For those of us who use the SFS libraries to develop, we often want 3 different versions of the libraries around simultaneously. The default options (compile with -O2) make code that runs much faster, perhaps twice as fast as compiling without optimization. However, optimized code is very frustrating to debug. So we often have debug builds of the software around. Once a bug is fixed, it's back to the standard build until the next bug shows up. Along the same lines, running with DMALLOC enabled as in sfslite-dbg really slows things down, but is very useful for tracking down hard-to-find memory corruptions. It also has an extra dependency--- devel/dmalloc. The ports are set up now in such a way that all 3 build modes can coexist peacefully. Thus, the three ports have different package lists. I'm pretty sure that doing them as one package with different options would mean one set of libraries clobbering the other. Pav and I discussed some of the issues after I intially submitted devel/sfslite, and I think in the end he was convinced. Thanks for your help. Regards, Max On Thu, 29 Sep 2005, Renato Botelho wrote: > Is there a reason to don't include these new options > on sfslite port? > > Cheers > -- > Renato Botelho >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200510040450.j944oLoE013074>