Permalink
Browse files

clarifying some points in the previous code comment as well as some m…

…inor

style changes
  • Loading branch information...
1 parent 515d71d commit 5ba2c3733ac0def4ccf0c8e5e71702bb24154144 jinmei committed May 8, 2007
Showing with 10 additions and 7 deletions.
  1. +10 −7 kame/sys/netinet6/ip6_input.c
@@ -1,4 +1,4 @@
-/* $KAME: ip6_input.c,v 1.372 2007/05/06 14:35:08 itojun Exp $ */
+/* $KAME: ip6_input.c,v 1.373 2007/05/08 02:53:12 jinmei Exp $ */
/*
* Copyright (C) 1995, 1996, 1997, and 1998 WIDE Project.
@@ -1091,16 +1091,19 @@ ip6_input(m)
#endif
/*
- * here we try to reject packets with more than 1 routing
+ * Here we try to reject packets with more than 1 routing
* headers. we do this here (instead of tagging mbuf route6.c)
* for the sake of computational costs, such as malloc().
*
- * the code could be too restrictive - there could be
- * actual use of more than 1 routing headers on a packet,
+ * The code could be too restrictive - there could be
+ * actual use of more than 1 routing headers on a packet
* which cannot be used to do bad things unlike
- * IPV6_RTHDR_TYPE_0. we will revisit it later when update
- * to RFC2460 gets published.
- * Sun May 6 23:34:20 JST 2007
+ * IPV6_RTHDR_TYPE_0. This code could also prohibit a mixed use
+ * of TYPE_x and TYPE_y routing headers (x != y) even if it is
+ * safe.
+ *
+ * We may need to revisit this behavior if and when a new type
+ * of routing header is defined.
*/
if (nxt == IPPROTO_ROUTING) {
if (rh_present++) {

0 comments on commit 5ba2c37

Please sign in to comment.