Committed, fix failing ld-mmix/bpo-6 and ld-mmix/bpo-19

[Hans-Peter Nilsson]

Looks like these tests weren't adjusted with the latest change
causing local absolute symbols to (again?) appear in the symbol
table, in this case "i".  The only test now failing is bpo-10
but there's a PR for that.

	* ld-mmix/bpo-6.d, ld-mmix/bpo-19.d: Adjust for symbols between
	section symbols and the _start symbol.

Index: ld-mmix/bpo-19.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-mmix/bpo-19.d,v
retrieving revision 1.6
diff -p -u -r1.6 bpo-19.d
--- ld-mmix/bpo-19.d	25 May 2005 06:49:53 -0000	1.6
+++ ld-mmix/bpo-19.d	26 Oct 2005 01:28:33 -0000
@@ -15,6 +15,7 @@ SYMBOL TABLE:
 0+ l    d  \*ABS\*	0+ (|\.shstrtab)
 0+ l    d  \*ABS\*	0+ (|\.symtab)
 0+ l    d  \*ABS\*	0+ (|\.strtab)
+#...
 0+ g       \.text	0+ _start
 #...

Index: ld-mmix/bpo-6.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-mmix/bpo-6.d,v
retrieving revision 1.6
diff -p -u -r1.6 bpo-6.d
--- ld-mmix/bpo-6.d	25 May 2005 06:49:53 -0000	1.6
+++ ld-mmix/bpo-6.d	26 Oct 2005 01:28:33 -0000
@@ -14,6 +14,7 @@ SYMBOL TABLE:
 0+ l    d  \*ABS\*	0+ (|\.shstrtab)
 0+ l    d  \*ABS\*	0+ (|\.symtab)
 0+ l    d  \*ABS\*	0+ (|\.strtab)
+#...
 0+ g       \.text	0+ _start
 #...


brgds, H-P