add more comments related to IDCODE/BYPASS behavior when detecting parts

git-svn-id: https://urjtag.svn.sourceforge.net/svnroot/urjtag/trunk@1897 b68d4a1b-bc3d-0410-92ed-d4ac073336b7

urjtag/ChangeLog

@@ -5,6 +5,10 @@
 
   * MAINTAINERS: Update paths to Blackfin related files.
 
+  * src/tap/detect.c (urj_tap_detect_parts): Clarify detect behavior with
+    parts that default to IDCODE or BYPASS since both are acceptable to the
+    JTAG IEEE standard.
+
 2011-02-24  Mike Frysinger  <vapier@gentoo.org>
 
   * src/tap/cable/ft2232.c: Rename user-facing cable string from

urjtag/src/tap/detect.c

@@ -284,7 +284,7 @@ urj_tap_detect_parts (urj_chain_t *chain, const char *db_path)
 
         if (urj_tap_register_compare (one, br) == 0)
         {
-            /* part with id */
+            /* Part that supports IDCODE */
             if (all_ids)
                 memcpy (id->data, &all_ids->data[i * 32 + 1], 31 * sizeof (id->data[0]));
             else
@@ -297,8 +297,14 @@ urj_tap_detect_parts (urj_chain_t *chain, const char *db_path)
             urj_log (URJ_LOG_LEVEL_NORMAL, _("Device Id: %s (0x%0*" PRIX64 ")\n"),
                      urj_tap_register_get_string (did), did->len / 4,
                      urj_tap_register_get_value (did));
-        } else
-            urj_log (URJ_LOG_LEVEL_NORMAL, _("Device Id: unknown as bit 0 was not a 1\n"));
+        }
+        else
+        {
+            /* If the device does not provide IDCODE, then it'll be in BYPASS
+             * mode after we reset the state machine.  So we'll get a 0 here.
+             * Not a bug, just a sad part :(.  */
+            urj_log (URJ_LOG_LEVEL_NORMAL, _("Device Id: not supported (bit 0 was not a 1)\n"));
+        }
 
         part = urj_part_alloc (did);
         if (part == NULL)