This is the mail archive of the
mailing list for the libstdc++ project.
- To: Gabriel Dos Reis <gdr at codesourcery dot com>
- Subject: Re: PR3042
- From: David Edelsohn <dje at watson dot ibm dot com>
- Date: Sun, 10 Jun 2001 18:41:00 -0400
- cc: Mark Mitchell <mark at codesourcery dot com>, Gabriel dot Dos-Reis at cmla dot ens-cachan dot fr, bkoz at nabi dot net, jason at cygnus dot com, libstdc++ at gcc dot gnu dot org
>>>>> Gabriel Dos Reis writes:
Gabriel> I'm confused as to why you're bringing "xlC name mangling" into this
Because MARK(!) brought them into the discussion. This is really
David> Your proposal breaks G++ compatibility with IBM's
David> compilers in a gratuitous way.
Mark> This is a non-issue -- there is already no compatibility with IBM's
Mark> compilers. There's nothing to break here. IBM's compilers use
Mark> entirely different methods of managing template instantiation from
Mark> G++, and this makes everything different, including what storage
Mark> classes it makes sense to use for particular symbols.
While xlC uses different name mangling, Mark's statement about
template instantiation is false. xlC has various template instantiation
options. The default behaves similar to G++'s current behavior. G++
currently does not implement the more complicated solutions which would
allow all of the correct semantics on AIX.
Arguing that because G++ is not compatible with xlC in some
(important) ways, does not give G++ license to gratuitously add more
incompatibilities in behavior. There *IS* some compatibility between G++
and xlC. Just because one cannot link the object files together does not
mean that the rest of the argument can be ignored.