This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: gcj, shared libraries, and -Bsymbolic
Tom Tromey writes:
> rth> What you actually want is (at least for -findirect-dispatch) is to
> rth> not export any symbols whatsoever. And to do so as simply as not
> rth> marking the symbols TREE_PUBLIC in the front end.
>
> Andrew> I know. However, at present this breaks CNI.
>
> I was thinking about this yesterday and it occurred to me that CNI is
> already broken with BC. So I think we could make this change now
> without making things worse... assuming we have the time to do it.
I suppose we could try making all symbols LOCAL only when
-findirect-dispatch is used. It would probably work, but I'm not
certain.
> Andrew> It happens. :-) We know that the current scheme is rather inefficient
> Andrew> in that it uses slow linear searches, etc, but it has mostly the
> Andrew> correct semantics.
>
> Yeah. So far there is just PR 19285, and an as-yet undiagnosed bug in
> eclipse 3 that can be fixed by deleting resources.jar.so (meaning
> eclipse probably has an obscure dependency on the precise timing of
> linking). Profiling all this stuff would be wonderful, if only there
> were more hours in the day.
Indeed.
Andrew.