970901 - cross problem with libio/gen-params

Joel Sherrill joel@OARcorp.com
Wed Sep 3 10:30:00 GMT 1997


On Wed, 3 Sep 1997, Ian Lance Taylor wrote:

> I don't know if this is the problem, but here is a potential problem
> which might cause the symptoms you are reporting.

Sounds like this is the culprit.  I am trying your suggestion now.

> The current development sources of the binutils generate nm-new rather
> than nm.new.  This is for easier use on Windows systems, so that we
> can generate nm-new.exe without worrying about the multiple extensions
> in nm.new.exe (it's true that nm.new.exe will work on new MS file
> systems, but it's easier to sidestep the whole issue for the benefit
> of old 8.3 file systems).

Unfortunately binutils 2.8.1 actually builds nm.new ld.new and strip.new.
:(  

> The top level Makefile.in in the egcs release looks for nm-new.  If
> you are using this with, e.g., binutils 2.8.1, which produces nm.new,
> then the top level Makefile.in won't find the nm program in the build
> directory.  That will cause it to use nm from your path instead.

It looks like this is enought o get the build past this point.

> I don't know if this is really an egcs problem; the problem will only
> arise if you try to mix egcs with the binutils 2.8.1 release in the
> same directory tree.

This was the first egcs snapshot I could not build one-tree style with
egcs.

> When the egcs CVS server is up and running, I hope to start putting
> the development binutils releases on it as well.  At that time, it
> will be possible to get a single consistent tree.

Does this mean that we will eventually be looking at egcs being a full
toolset tree?

--joel




More information about the Gcc mailing list