This is the mail archive of the
mailing list for the binutils project.
Re: [RFC] alpha-elf vs copy_indirect_symbol
- From: Alan Modra <amodra at gmail dot com>
- To: Richard Henderson <rth at redhat dot com>
- Cc: binutils at sourceware dot org
- Date: Wed, 15 Jun 2011 12:50:21 +0930
- Subject: Re: [RFC] alpha-elf vs copy_indirect_symbol
- References: <4DF788CA.firstname.lastname@example.org>
On Tue, Jun 14, 2011 at 09:14:02AM -0700, Richard Henderson wrote:
> What concerns me is that copy_indirect_symbol is called for things
> that aren't indirect symbols. In particular, we call this hook for
> defweak and defined symbols.
The weak and strong symbols in question are both non-function symbols
defined in a shared lib, eg. timezone and _timezone.
> Which might be fine, I suppose, if we
> then *replaced* the defweak symbol in the sym_hashes table so that
> we never ever see it again. I have no idea what the current usage
> is supposed to accomplish.
It started life as a means to copy over symbol flags from the weak to
strong sym. Without doing some archaeology I'm not sure why we
also transfer over dyn_relocs on most targets.
Australia Development Lab, IBM