Committed: don't run ld-scripts/pr14962 for mmix-knuth-mmixware

[Hans-Peter Nilsson]

This test fails for mmix-knuth-mmixware.  The default-output "mmo"
object format has no notion of symbol types; the types attached by BFD
are derived from the section associations, themselves an artificial
construct and not part of mmo proper.  So, the expected symbol in
.data can't be absolute.  From it's value, it's deduced to be a data
symbol, as espected:

0000000000002000 D x

(a "D" instead of the expected "A" in the test output).  So, best to
exclude the test from being run for this target.  I also took the
liberty of adding punctuation to the comment that follows the markup,
as per GNU coding standards.  Text like that, without punctuation,
looks just too much like there's a typo or patch application error.

Committed.

ld/testsuite:
	* ld-scripts/pr14962.d: Don't run for mmix-knuth-mmixware.  Add
	comment punctuation.

Index: ld-scripts/pr14962.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/pr14962.d,v
retrieving revision 1.1
diff -p -u -r1.1 pr14962.d
--- ld-scripts/pr14962.d	18 Dec 2012 13:38:10 -0000	1.1
+++ ld-scripts/pr14962.d	30 Dec 2012 01:58:09 -0000
@@ -2,9 +2,9 @@
 #source: pr14962a.s
 #source: pr14962b.s
 #nm: -n
-#notarget: rx-*-*
-# the reference to x would normally generate a cross-reference error
-# but the linker script converts x to absolute to avoid the error
+#notarget: rx-*-* mmix-knuth-mmixware
+# The reference to x would normally generate a cross-reference error
+# but the linker script converts x to absolute to avoid the error.

 #...
 0+2000 A x