Re: Partial symbol export vs --export-dynamic

[Hans-Peter Nilsson <hans-peter dot nilsson at axis dot com>]

> Date: Mon, 18 Jun 2001 08:21:21 -0700
> From: "H . J . Lu" <hjl@lucon.org>

> On Mon, Jun 18, 2001 at 09:13:22AM +0200, Hans-Peter Nilsson wrote:
> > > From: Ben Elliston <bje@redhat.com>
> > > Date: Mon, 18 Jun 2001 12:38:00 +1000 (EST)
> > 
> > > I see no reason why it shouldn't be possible for you to list C++
> > > symbols in an unmangled form and let the linker mangle them as the
> > > export file is processed.  Am I missing something, anyone?
> > 
> > Perhaps something like that there's only a *de*mangler built-in
> > through libiberty?  A mangler generally requires a C++-parser,
> > IIUC.  Or some other trick currently not implemented.
> 
> The version script has been supporting C++ since 1999:

I should have read the code before replying.  I see how it works
now: it iterates over the list of symbols found at linking and
demangles those marked C++ before matching them to the patterns
in the version-script.  So it can indeed match a "real" C++-name
and the version script does not need to know of any mangled
symbols.  Cool.  Shouldn't this be used on the GCC C++ DSO:s?

brgds, H-P