Date: Mon, 25 Jun 2001 19:31:22 +0300 (IDT) From: Eli Zaretskii X-Sender: eliz AT is To: Tim Van Holder cc: djgpp-workers AT delorie DOT com, Andris Pavenis , lauras AT softhome DOT net Subject: Re: gcc 3.0 released In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Reply-To: djgpp-workers AT delorie DOT com Errors-To: nobody AT delorie DOT com X-Mailing-List: djgpp-workers AT delorie DOT com X-Unsubscribes-To: listserv AT delorie DOT com Precedence: bulk On Mon, 25 Jun 2001, Tim Van Holder wrote: > > We've been through that in the past: the problem with the linker script > > is that, unlike specs, it is releated to both the compiler and to > > Binutils. > > I don't agree - linker scripts are only for the linker (ie binutils). Then please explain how come we now need to change it because of a new version of GCC. > Even if a new compiler needs updated linker scripts, that support must be > added to binutils, and it is the updated binutils that will bring the new > scripts with it. Yes, if GCC, Binutils, and djdev were all released together, like it was in the old days when DJ was doing it all single-handedly, then everybody would be happy, and, in fact, it would have been immaterial which one of these packages comes with the linker script. But that's not how these packages are released. And I don't think we should assume that every new release of the compiler will be accompanied by new Binutils. I don't think we can rely on that, as long as different people maintain these ports. > For example, if gcc 3.1 for whatever reason needs linker script capabilities > only provided by binutils 2.15, then our gcc 3.1 package, just like any > other, would require binutils 2.15. Such calamities is something we should IMHO try to prevent. It is unrealistic to expect that these two packages are always released in perfect sync.