Re: Section symbols not getting created. Bug? Is attached patch correct fix?

[John Marino <binutils at marino dot st>]

Alan Modra wrote:
> On Wed, Apr 06, 2011 at 08:16:44AM +0200, John Marino wrote:
> > At the very least, shouldn't there be a switch that could be used to
> > override the current behavior and go back to the original behavior?
>
> I'd rather not add yet another command line switch.  I suppose you
> could add an internal flag that selects the old behaviour, and set the
> flag just for your target in a .em file.
>
> > What was the rationale for making this breaking change in the first place?
>
> Avoiding useless clutter in the symbol table.

Well, I think this is a case where the baby was thrown out with the bath 
water.  I would have thought that backwards capability would have been 
provided with this change.

Regarding the internal flag / .em suggestion, isn't that more or less 
equivalent to the patch on ldlang.c I provided earlier?  Basically it 
would revert the behavior to work as binutils 2.17?

If we're going to patch the binutils codebase locally anyway, maybe it's 
better to switch the behavior of lang_insert_orphan based on the value 
of link_info.shared.  If -shared or -Bshareable passed via command-line, 
use exp_assign function (e.g. 2.17 behavior) and if not, use the 
exp_provide function (e.g. trunk behavior).

Do you see any fallacy with that logic?

Regards,
John