[Ada 3.3] Backport compilation fixes

Gabriel Dos Reis gdr@integrable-solutions.net
Wed Dec 17 19:41:00 GMT 2003


kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:

|     If Ada folks think that ability to build previous versions of GNAT
|     with recent bootstrapping compilers is of no value for their
|     community, I see no reason to endorse that initiative -- even though
|     I can fully see the benefits for GCC developers.
| 
| I don't want to speak for Arno, but it didn't sound like that's what he
| was saying.  My interpretation was that he was expressing concern that,
| in general, compiling GNAT with a newer version of the compiler was something
| that's dubious.  But if it can be made to work in a particular case,
| there's no reason not to do it.

OK.  Thanks for the clarification.  
The issue here is whether it is OK by Ada maintainers to have a patch
that makes GNAT/GCC-3.3.x buildable by a bootstrapping from GCC-3.4.
If that is not OK, is it OK to have the configury machinery disable
Ada by default?  I can make decisions either way, but I would prefer
to have a clear statement from Ada maintainers -- because that may
settle some bugs as regression in future versions of GCC.

-- Gaby


   



More information about the Gcc-patches mailing list