ld's sole user of update_definedness() vs documentation

[Alan Modra]

On Mon, Feb 22, 2021 at 08:35:52AM +0100, Jan Beulich wrote:
> On 21.02.2021 05:05, Alan Modra wrote:
> >> Perhaps, if there are scripts in the wild which do this, this
> >> could at least be a warning for the time being?
> > 
> > Yes, I didn't mean to leave this go so long without doing something.
> > I just committed the following.
> 
> Thanks for the quick fix. Wouldn't it be better to also adjust the
> PROVIDE() doc then, replacing "error" by "warning"? I can certainly
> do this, if you agree.

Yes please.

-- 
Alan Modra
Australia Development Lab, IBM