Address MISSING_BREAK static analysis reports some more.

This addresses core findings 4 and 6.

Moises Silva helped me by stating that a break could be
safely added to the case where it is added in chan_dahdi.c

In say.c, I have added a comment indicating that static analysis
complains but that it is currently unknown if this is correct.

This fixes all core findings of this type.

(closes issue ASTERISK-19662)
reported by Matthew Jordan
........

Merged revisions 367027 from http://svn.asterisk.org/svn/asterisk/branches/1.8
........

Merged revisions 367028 from http://svn.asterisk.org/svn/asterisk/branches/10


git-svn-id: https://origsvn.digium.com/svn/asterisk/trunk@367029 65c4cc65-6c06-0410-ace0-fbb531ad65f3

channels/chan_dahdi.c

@@ -8051,6 +8051,7 @@ static struct ast_frame *dahdi_handle_event(struct ast_channel *ast)
 #else
 		ast_log(LOG_WARNING, "Received bits changed on %s signalling?\n", sig2str(p->sig));
 #endif
+		break;
 	case DAHDI_EVENT_PULSE_START:
 		/* Stop tone if there's a pulse start and the PBX isn't started */
 		if (!ast_channel_pbx(ast))

main/say.c

@@ -6127,6 +6127,9 @@ int ast_say_date_with_format_zh(struct ast_channel *chan, time_t t, const char *
 				if (tm.tm_hour < 10) {
 					res = wait_file(chan, ints, "digits/0", lang);
 				}
+				/* XXX Static analysis warns of no break here. No idea if this is
+				 * correct or not
+				 */
 			case 'k':
 				/* 24-Hour */
 				if (!(tm.tm_hour % 10) || tm.tm_hour < 10) {