If a spare is added to fix a degraded raid5 array and the resync does not run,
the kernel probably still has it as a spare.  The user should have the option to
remove this spare disk from the array.

Note: This patch should be applied on top of raid5_degrade_fix.patch:
http://evms.sourceforge.net/patches/2.5.5/engine/raid5_degrade_fix.patch

--- evms-2.5.5a/plugins/md/raid5_mgr.c	2006-05-10 11:33:10.173604904 -0500
+++ evms-2.5.5b/plugins/md/raid5_mgr.c	2006-05-10 11:27:27.348722208 -0500
@@ -3416,15 +3416,9 @@
 	/* Clear out the spare_disks list. */
 	EngFncs->delete_all_elements(spare_disks);
 
-	/*
-	 * If there is only one spare and the array is in degrade mode, then
-	 * the spare cannot be removed.
-	 */
-	if (vol->spare_disks <= 1) {
-		if (vol->flags & MD_DEGRADED) {
-			LOG_EXIT_INT(0);
-			return 0;
-		}
+	if ((vol->flags & MD_DEGRADED) && (vol->spare_disks > 0)) {
+			LOG_WARNING("%s is degraded, but there are %d spare disk(s).\n",
+			            vol->name, vol->spare_disks);
 	}
 
 	LIST_FOR_EACH(vol->members, iter, member) {