ld's sole user of update_definedness() vs documentation

[Jan Beulich]

After having read through PROVIDE()'s documentation I've come to
the conclusion that a linker script defining a symbol defined by
an object would be an error. After having made a change where I
would have expected such an error to surface (which I then had
planned to take care of), no such error actually appeared.

Considering commit 422f1c65c9e9 ("Report an error for script
multiply defined symbols"), the description of which amends the
title by "or maybe not just yet", wouldn't it be helpful for
documentation to match reality? It could still state that such
duplicates may become an error in the future, but I don't think
it should lead people to believe that such an error will in
fact get diagnosed as long as this doesn't actually happen.

Perhaps, if there are scripts in the wild which do this, this
could at least be a warning for the time being?

Jan