[PATCH] ld: Fix several 32-bit SPARC plugin tests

[Nick Clifton]

Hi Rainer,

My guess as to the reason for the abort() was that the writer wanted to
catch unhandled sparc machine values.  So please could you try changing
the patch to this:

diff --git a/bfd/elf32-sparc.c b/bfd/elf32-sparc.c
index 65c81ecccf..aba81fae1a 100644
--- a/bfd/elf32-sparc.c
+++ b/bfd/elf32-sparc.c
@@ -150,6 +150,8 @@ sparc_final_write_processing (bfd *abfd)
     case bfd_mach_sparc_sparclite_le :
       elf_elfheader (abfd)->e_flags |= EF_SPARC_LEDATA;
       break;
+    case 0: /* A non-sparc architecture - ignore.  */
+      break;
     default :
       abort ();
       break;

Or, if like me you do not like calls to abort() inside a library,
then how about this:

diff --git a/bfd/elf32-sparc.c b/bfd/elf32-sparc.c
index 65c81ecccf..e6c873ddf1 100644
--- a/bfd/elf32-sparc.c
+++ b/bfd/elf32-sparc.c
@@ -150,8 +150,11 @@ sparc_final_write_processing (bfd *abfd)
     case bfd_mach_sparc_sparclite_le :
       elf_elfheader (abfd)->e_flags |= EF_SPARC_LEDATA;
       break;
+    case 0: /* A non-sparc architecture - ignore.  */
+      break;
     default :
-      abort ();
+      _bfd_error_handler (_("%pB: unhandled spac machine value '%lu' detected during write processing"),
+                         abfd, (long) bfd_get_mach (abfd));
       break;
     }
 }

Either of these is pre-approved for committal if they work for you.

Cheers
  Nick