Re: Does the LD --wrap feature work for library internal references?

[Sebastian Huber <sebastian dot huber at embedded-brains dot de>]

On 29/01/2019 08:44, Sebastian Huber wrote:
> On 29/01/2019 01:31, Alan Modra wrote:
> > > Attached is a second version which does the wrapping in
> > > RELOC_FOR_GLOBAL_SYMBOL(). It removes 
> > > bfd_wrapped_link_hash_lookup(). I am
> > > not sure if this is a good idea. This function is used in various 
> > > places.
> > > Maybe the wrapping needs to be done on other spots and not only in
> > > RELOC_FOR_GLOBAL_SYMBOL().
> > Yes, wrapping should happen for more than just relocs.  For instance,
> > if a shared library references malloc and you're building an
> > executable wih --wrap=malloc and have a __wrap_malloc, then I think
> > the shared library reference ought to be satisfied by your
> > __wrap_malloc.  That has implications for ELF symbol merging. Note
> > that I'm not even sure whether this works currently, but it is
> > feasible that a shared library reference to malloc be satisfied by
> > your __wrap_malloc while your __real_malloc results in
> > malloc@@GLIBC_2.2.5  say.  If that does work currently we wouldn't want
> > to break it..
>
> Yes, there are LD testsuite failures with this patch exactly in this 
> area. If I keep the symbol wrapping as is and in addition add it to 
> RELOC_FOR_GLOBAL_SYMBOL(), then there is the problem that the 
> __real_SYM -> SYM information was lost and we wrap SYM always to 
> __wrap_SYM. This leads to an infinite recursion in:
>
> void __wrap_f(void)
> {
>     __real_f();
> }
>
> Looks like no matter how I try it, it almost works. There are probably 
> some reasons why it wasn't added for defined references in the last 20 
> years.

I think we have three options,

1. leave it as is,

2. remove the wrapping via the symbol hash table and add the wrapping to 
all symbol table consumers, e.g. global symbol relocations, and whatever, or

3. add a new option, e.g. --wrap-reloc which is orthogonal to the 
existing --wrap option and just deals with wrappings of global symbol 
relocations.

I don't know the linker code good enough to determine if 2. would work 
and how much effort it needs.

--
Sebastian Huber, embedded brains GmbH

Address : Dornierstr. 4, D-82178 Puchheim, Germany
Phone   : +49 89 189 47 41-16
Fax     : +49 89 189 47 41-09
E-Mail  : sebastian.huber@embedded-brains.de
PGP     : Public key available on request.

Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG.