Home Original page

[PATCH] Gold: Treat symbols with version index 0 as unversioned

Michael Matz matz@suse.de
Tue Nov 18 13:46:26 GMT 2025 
Hello,

On Tue, 18 Nov 2025, H.J. Lu wrote:

> > they then can used for resolving.  If they could, then there'd be an
> > internal inconsistency in the above two items: symbols defined in an
> > object (and that don't have a version otherwise), are to use index 1.  So,
> > for index 0 to be used the symbol in question must be (a) unversioned and
> > (b) not be defined (in the sense of being available to resolve against)
> > by the object (lest it had index 1).
> >
> > So, no, I don't think you should resolve against defined symbols with
> > NDX_LOCAL.  They have, as stated everywhere, local scope.
> >
> 
> The key point is that VER_NDX_LOCAL doesn't change symbol
> binding which is determined by the st_other field

Repeating this point doesn't make it more true.  The specs from the 
beginning of their existence always said something else, so this:

> as implemented
> by ld and ld.so in glibc.

... is a bug, strictly speak.  Now, I don't say we should change glibc.  
Nor do I think we should change ld.bfd in the way it deals with 
NDX_LOCAL undefined symbols.  But I do think that ld.bfd should _not_ 
produce NDX_LOCAL defined symbols, for the time being.

What you are proposing is a change in the spec (the linux one, not the Sun 
one), one I can go along with.  But as any such change we can't just do 
that when we feel like it.


Ciao,
Michael.
More information about the Binutils mailing list