Re: gold vs libc

[Alan Modra <amodra@gmail.com>]

On Wed, Sep 10, 2014 at 06:05:26PM -0400, Rafael Espíndola wrote:
> > If the .eh_frame data in crt1.o really does need to come before
> > __EH_FRAME_BEGIN__, another thing you could do is simply make it so
> > gold treats it as non-optimizable. Adding a null relocation to the
> > first word of the section should do it; inserting a zero-length entry
> > anywhere but the end would do it (if that doesn't have adverse affects
> > elsewhere).
> 
> Since the problem comes from an optimizations that knows what
> .eh_frame is, maybe it could learn that __EH_FRAME_BEGIN__ and
> __EH_FRAME_END__ are special symbols marking the start and end of the
> section?

I'm inclined to think this is the best solution (and made a comment to
that effect in pr17366 before reading this thread).  I don't like the
idea of making .eh_frame optimisation depend on --eh-frame-hdr.
Providing a sorted list of FDEs is really separate to optimising those
FDEs and CIEs.

-- 
Alan Modra
Australia Development Lab, IBM