https://sourceware.org/bugzilla/show_bug.cgi?id=27311
Alan Modra <amodra at gmail dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |ASSIGNED
Assignee|unassigned at sourceware dot org |amodra at gmail dot com
--- Comment #6 from Alan Modra <amodra at gmail dot com> ---
What's happening here (using Michael's testcase, thanks!) is that
inlib1@@LIBFOO is seen as being a definition, triggering
if (!add_needed
&& matched
&& definition
&& ((dynsym
&& h->ref_regular_nonweak)
|| (old_bfd != NULL
&& (old_bfd->flags & BFD_PLUGIN) != 0
&& bind != STB_WEAK)
|| (h->ref_dynamic_nonweak
&& (elf_dyn_lib_class (abfd) & DYN_AS_NEEDED) != 0
&& !on_needed_list (elf_dt_name (abfd),
htab->needed, NULL))))
to pull in lib1.so. However, inlib1@@LIBFOO is actually overridden by inlib1
from the object file. Without symbol versioning we see inlib1 from lib1.so
setting override during the first _bfd_elf_merge_symbol call in
add_object_symbols and then this
if (override && matched)
definition = FALSE;
stops the non-versioned inlib1 from being considered a reason to pull in
lib1.so.
--
You are receiving this mail because:
You are on the CC list for the bug.
