This is the mail archive of the binutils@sources.redhat.com mailing list for the binutils project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: PATCH: BZ/175: Fix dynamic string offset


On Sat, May 22, 2004 at 11:27:22AM +0930, Alan Modra wrote:
> On Thu, May 20, 2004 at 08:30:24PM -0700, H. J. Lu wrote:
> > On Fri, May 21, 2004 at 10:55:03AM +0930, Alan Modra wrote:
> > > I think it quite possible that this patch will break other accesses
> > > into merged sections.  To properly fix the problem, you would need to
> > > distinguish this particular access from others.  Perhaps the reloc used
> > > might give you enough information.
> > 
> > Good idea. Here it is.
> 
> > 	* elfxx-ia64.c (elfNN_ia64_relax_section): Call
> > 	_bfd_merged_section_offset with irel->r_addend as dynamic
> > 	offset if it isn't a branch.
> 
> I doubt this patch is complete.  If you need to special case certain
> accesses in relax_section, how is it that relocate_section escapes a
> need for the same special case code?  This might prove more difficult,
> as you might need to pull apart _bfd_elf_rela_local_sym.  I'm also still
> concerned that other ia64 code sequences might use R_IA64_LTOFF22X in a
> way that requires you include the addend.  Please discuss this with Jim,
> I'm not familiar enough with ia64 to know.
> 

elfNN_ia64_relax_section shouldn't calls _bfd_merged_section_offset just
because tsec->sec_info_type == ELF_INFO_TYPE_MERGE. For local symbols,
it should also check

(tsec->flags & SEC_MERGE) && ELF_ST_TYPE (isym->st_info) == STT_SECTION



H.J.
----
2004-05-25  H.J. Lu  <hongjiu.lu@intel.com>

	* elfxx-ia64.c (elfNN_ia64_relax_section): Properly call
	_bfd_merged_section_offset for local symbols.

--- bfd/elfxx-ia64.c.merge	2004-05-19 08:30:21.000000000 -0700
+++ bfd/elfxx-ia64.c	2004-05-25 18:05:02.936045175 -0700
@@ -874,6 +874,16 @@ elfNN_ia64_relax_section (abfd, sec, lin
 
 	  toff = isym->st_value;
 	  dyn_i = get_dyn_sym_info (ia64_info, NULL, abfd, irel, FALSE);
+	  
+	  if ((tsec->flags & SEC_MERGE)
+	      && ELF_ST_TYPE (isym->st_info) == STT_SECTION
+	      && tsec->sec_info_type == ELF_INFO_TYPE_MERGE)
+	    toff = _bfd_merged_section_offset (abfd, &tsec,
+					       elf_section_data (tsec)->sec_info,
+					       toff + irel->r_addend,
+					       (bfd_vma) 0);
+	  else
+	    toff += irel->r_addend;
 	}
       else
 	{
@@ -918,16 +928,16 @@ elfNN_ia64_relax_section (abfd, sec, lin
 	      tsec = h->root.u.def.section;
 	      toff = h->root.u.def.value;
 	    }
+	  
+	  if (tsec->sec_info_type == ELF_INFO_TYPE_MERGE)
+	    toff = _bfd_merged_section_offset (abfd, &tsec,
+					       elf_section_data (tsec)->sec_info,
+					       toff + irel->r_addend,
+					       (bfd_vma) 0);
+	  else
+	    toff += irel->r_addend;
 	}
 
-      if (tsec->sec_info_type == ELF_INFO_TYPE_MERGE)
-	toff = _bfd_merged_section_offset (abfd, &tsec,
-					   elf_section_data (tsec)->sec_info,
-					   toff + irel->r_addend,
-					   (bfd_vma) 0);
-      else
-	toff += irel->r_addend;
-
       symaddr = tsec->output_section->vma + tsec->output_offset + toff;
 
       roff = irel->r_offset;


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]