This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [committed] experimental top level bootstrap support


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Alexandre Oliva wrote:
| On Mar 15, 2004, neroden@twcny.rr.com (Nathanael Nerode) wrote:
|
|
|>2004-03-15  Nathanael Nerode  <neroden@gcc.gnu.org>
|
|
|>	* Makefile.tpl: Introduce experimental top level bootstrap support.
|>	* Makefile.in: Regenerate.
|>	* configure.in: Introduce support for top level bootstrap.
|>	* configure: Regenerate.
|
|
| Hmm...  Top-level bootstrap is sort of pointless unless you can

Well, it removes an awful lot of hair from the gcc Makefile, along with
messes of dangling links; the fact that it's reconfigured at each stage
removes a lot *more* mess, so I wouldn't call it 'pointless'.

| actually take advantage of it to add more tools (e.g., assembler,
| linker) and libraries (libiberty, libgcc after it's split out of gcc)
| in the bootstrap.
|
|  So I think it would be better to have
| stage[123]/gcc instead of stage[123]-gcc.  Am I making any sense?
Yes; unfortunately it's easier said than done.  I spent an inordinate
amount of time getting the details right on this version, and I did try
several other methods, including that.  If we can get all the bootstrap
support out of the gcc Makefile and into the top level, I believe that
that will allow for extensive cleanups which will make it much easier to
accomplish this goal.  So I'd rather concentrate on doing that first.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFAW7loRGZ0aC4lkIIRAjaoAJ9fHLF0ACC7AO5ZX7ZjbtLdy0FW5wCfWs5a
Nqd8xMRYj8L8vWOpj9V6iyg=
=np3V
-----END PGP SIGNATURE-----


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]