This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: trying to glue Modula-3 onto egcs table based exception handl ing support
- To: rth at cygnus dot com
- Subject: Re: trying to glue Modula-3 onto egcs table based exception handl ing support
- From: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- Date: Thu, 1 Jun 00 06:01:31 EDT
- Cc: gcc at gcc dot gnu dot org
> Where precisely does m3.c tie into the rest of GCC?
Above the tree level.
> There may be other (political & legal) concerns with integrating
> this file depending on precisely where it hooks in and exactly
> how it works.
I wouldn't think it would be much different than Ada.
The m3 (real) front end writes out a simple text-based representation
of its (probably processed) AST, which is piped to m3.c, which converts
it into trees, and from there the normal things happen.
Note that GNAT has not done this for nearly six years. (There is code around
to read and write the tree, but this is used for ASIS, not the compilation
process. I don't even think it *could* be used for the compilation process
at this point.)
I imagine it would be possible, though perhaps painful, to link the
front end directly against the back end. I havn't looked at how GNAT
accomplishes this.
By just linking the whole thing together in one executable.