This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: Fix Relocs for x86-64
- To: aj at suse dot de
- Subject: Re: Fix Relocs for x86-64
- From: Hans-Peter Nilsson <hans-peter dot nilsson at axis dot com>
- Date: Sun, 11 Mar 2001 16:24:45 +0100
- CC: hans-peter dot nilsson at axis dot com, binutils at sources dot redhat dot com
> From: Andreas Jaeger <aj@suse.de>
> Date: 10 Mar 2001 10:02:48 +0100
> Hans-Peter Nilsson <hans-peter.nilsson@axis.com> writes:
> > [C comment for relocate_section elided]
>
> That description looks fine. What do others think? I'm in favour of
> including it for x86-64.
I actually hoped to lure you into adding more comments like
that, for the nontrivial condition parts above that one, on the
assumption that you had it in fresh memory. :-)
I believe those conditions can be replaced with
(sec->output_section == NULL && (sec->owner->flags & DYNAMIC) != 0)
for x64-64 as well as all targets with a lookalike
relocate_section.
I might forget some tricky cases; I would be happy to know.
brgds, H-P