On 2013/5/9 04:11 AM, Bernhard Reutner-Fischer wrote: > On 8 May 2013 15:11:18 Jan Hubicka <hubi...@ucw.cz> wrote: >> > 2013-05-08 Nathan Sidwell <nat...@codesourcery.com> >> > gcc/ >> > * varasm.c (default_use_anchors_for_symbol_p): Reject WEAK. >> > gcc/testsuite/ >> > * gcc.dg/visibility-21.c: New. >> > Index: gcc/varasm.c >> > =================================================================== >> > --- gcc/varasm.c (revision 410150) >> > +++ gcc/varasm.c (working copy) >> > @@ -6871,6 +6871,11 @@ default_use_anchors_for_symbol_p (const_ >> > if (!targetm.binds_local_p (decl)) >> > return false; >> > + /* Weak decls might be overridden, but could still be local to >> > + the module. */ >> > + if (DECL_WEAK (decl)) >> > + return false; >> >> With LTO we play more of similar tricks, by making use of the >> resolution file. >> I.e. for COMMON and EXTERNAL. Does it matter here? >> >> I would preffer the renaming excercise, since the name confused me few >> times, >> too and the other predicate would be useful for IPA code :) >> >> Thanks a lot for looking into this! >> Honza > > Does this regress PR32219 ? > http://gcc.gnu.org/ml/gcc-patches/2010-03/msg00665.html > I don't remember if the patch and test case were applied yet and don't > have the sources at hand.. > > Thanks,
I don't think so (Nathan and I were also discussing that PR in the context of this fix). The testcase in that PR still segfaults, which I have another fix. Will post later. Chung-Lin
