Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 25 Aug 2019 04:56:34 +0000 (UTC)
From:      Cy Schubert <cy@FreeBSD.org>
To:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-11@freebsd.org
Subject:   svn commit: r351470 - in stable: 11/sys/contrib/ipfilter/netinet 12/sys/contrib/ipfilter/netinet
Message-ID:  <201908250456.x7P4uY6s006477@repo.freebsd.org>

next in thread | raw e-mail | index | archive | help
Author: cy
Date: Sun Aug 25 04:56:33 2019
New Revision: 351470
URL: https://svnweb.freebsd.org/changeset/base/351470

Log:
  MFC r350880:
  
  r272552 applied the patch from ipfilter upstream fil.c r1.129 to fix
  broken ipfilter rule matches (upstream bug #554). The upstream patch
  was incomplete, it resolved all but one rule compare issue. The issue
  fixed here is when "{to, reply-to, dup-to} interface" are used in
  conjuncion with "on interface". The match was only made if the on keyword
  was specified in the same order in each case referencing the same rule.
  This commit fixes this.
  
  The reason for this is that interface name strings and comment keyword
  comments are stored in a a variable length field starting at fr_names
  in the frentry struct. These strings are placed into this variable length
  in the order they are encountered by ipf_y.y and indexed through index
  pointers in fr_ifnames, fr_comment or one of the frdest struct fd_name
  fields. (Three frdest structs are within frentry.) Order matters and
  this patch takes this into account.
  
  While in here it was discovered that though ipfilter is designed to
  pport multiple interface specifiations per rule (up to four), this
  undocumented (the man page makes no mention of it) feature does not work.
  A todo is to fix the multiple interfaces feature at a later date. To
  understand the design decision as to why only four were intended, it is
  suspected that the decision was made because Sun workstations and PCs
  rarely if ever exceeded four NICs at the time, this is not true in 2019.
  
  PR:		238796
  Reported by:	WHR <msl0000023508@gmail.com>

Modified:
  stable/11/sys/contrib/ipfilter/netinet/fil.c
  stable/11/sys/contrib/ipfilter/netinet/ip_fil.h
Directory Properties:
  stable/11/   (props changed)

Changes in other areas also in this revision:
Modified:
  stable/12/sys/contrib/ipfilter/netinet/fil.c
  stable/12/sys/contrib/ipfilter/netinet/ip_fil.h
Directory Properties:
  stable/12/   (props changed)

Modified: stable/11/sys/contrib/ipfilter/netinet/fil.c
==============================================================================
--- stable/11/sys/contrib/ipfilter/netinet/fil.c	Sun Aug 25 02:38:45 2019	(r351469)
+++ stable/11/sys/contrib/ipfilter/netinet/fil.c	Sun Aug 25 04:56:33 2019	(r351470)
@@ -4418,6 +4418,28 @@ ipf_matchicmpqueryreply(v, ic, icmp, rev)
 }
 
 
+/*
+ * IFNAMES are located in the variable length field starting at
+ * frentry.fr_names. As pointers within the struct cannot be passed
+ * to the kernel from ipf(8), an offset is used. An offset of -1 means it
+ * is unused (invalid). If it is used (valid) it is an offset to the
+ * character string of an interface name or a comment. The following
+ * macros will assist those who follow to understand the code.
+ */
+#define IPF_IFNAME_VALID(_a)	(_a != -1)
+#define IPF_IFNAME_INVALID(_a)	(_a == -1)
+#define IPF_IFNAMES_DIFFERENT(_a)	\
+	!((IPF_IFNAME_INVALID(fr1->_a) &&	\
+	IPF_IFNAME_INVALID(fr2->_a)) ||	\
+	(IPF_IFNAME_VALID(fr1->_a) &&	\
+	IPF_IFNAME_VALID(fr2->_a) &&	\
+	!strcmp(FR_NAME(fr1, _a), FR_NAME(fr2, _a))))
+#define IPF_FRDEST_DIFFERENT(_a)	\
+	(memcmp(&fr1->_a.fd_addr, &fr2->_a.fd_addr,	\
+	offsetof(frdest_t, fd_name) - offsetof(frdest_t, fd_addr)) ||	\
+	IPF_IFNAMES_DIFFERENT(_a.fd_name))
+
+
 /* ------------------------------------------------------------------------ */
 /* Function:    ipf_rule_compare                                            */
 /* Parameters:  fr1(I) - first rule structure to compare                    */
@@ -4430,22 +4452,50 @@ ipf_matchicmpqueryreply(v, ic, icmp, rev)
 static int
 ipf_rule_compare(frentry_t *fr1, frentry_t *fr2)
 {
+	int i;
+
 	if (fr1->fr_cksum != fr2->fr_cksum)
 		return (1);
 	if (fr1->fr_size != fr2->fr_size)
 		return (2);
 	if (fr1->fr_dsize != fr2->fr_dsize)
 		return (3);
-	if (bcmp((char *)&fr1->fr_func, (char *)&fr2->fr_func, FR_CMPSIZ(fr1))
+	if (bcmp((char *)&fr1->fr_func, (char *)&fr2->fr_func, FR_CMPSIZ)
 	    != 0)
 		return (4);
+	/*
+	 * XXX:	There is still a bug here as different rules with the
+	 *	the same interfaces but in a different order will compare
+	 *	differently. But since multiple interfaces in a rule doesn't
+	 *	work anyway a simple straightforward compare is performed
+	 *	here. Ultimately frentry_t creation will need to be
+	 *	revisited in ipf_y.y. While the other issue, recognition
+	 *	of only the first interface in a list of interfaces will
+	 *	need to be separately addressed along with why only four.
+	 */
+	for (i = 0; i < FR_NUM(fr1->fr_ifnames); i++) {
+		/*
+		 * XXX:	It's either the same index or uninitialized.
+		 * 	We assume this because multiple interfaces
+		 *	referenced by the same rule doesn't work anyway.
+		 */
+		if (IPF_IFNAMES_DIFFERENT(fr_ifnames[i]))
+			return(5);
+	}
+
+	if (IPF_FRDEST_DIFFERENT(fr_tif))
+		return (6);
+	if (IPF_FRDEST_DIFFERENT(fr_rif))
+		return (7);
+	if (IPF_FRDEST_DIFFERENT(fr_dif))
+		return (8);
 	if (!fr1->fr_data && !fr2->fr_data)
 		return (0);	/* move along, nothing to see here */
 	if (fr1->fr_data && fr2->fr_data) {
 		if (bcmp(fr1->fr_caddr, fr2->fr_caddr, fr1->fr_dsize) == 0)
 			return (0);	/* same */
 	}
-	return (5);
+	return (9);
 }
 
 

Modified: stable/11/sys/contrib/ipfilter/netinet/ip_fil.h
==============================================================================
--- stable/11/sys/contrib/ipfilter/netinet/ip_fil.h	Sun Aug 25 02:38:45 2019	(r351469)
+++ stable/11/sys/contrib/ipfilter/netinet/ip_fil.h	Sun Aug 25 04:56:33 2019	(r351470)
@@ -735,12 +735,9 @@ typedef	struct	frentry {
 	u_char	fr_icode;	/* return ICMP code */
 	int	fr_group;	/* group to which this rule belongs */
 	int	fr_grhead;	/* group # which this rule starts */
-	int	fr_ifnames[4];
 	int	fr_isctag;
 	int	fr_rpc;		/* XID Filtering */ 
 	ipftag_t fr_nattag;
-	frdest_t fr_tifs[2];	/* "to"/"reply-to" interface */
-	frdest_t fr_dif;	/* duplicate packet interface */
 	/*
 	 * These are all options related to stateful filtering
 	 */
@@ -750,6 +747,12 @@ typedef	struct	frentry {
 	int	fr_icmphead;	/* ICMP group  for state options */
 	u_int	fr_age[2];	/* non-TCP state timeouts */
 	/*
+	 * These are compared separately.
+	 */
+	int	fr_ifnames[4];
+	frdest_t fr_tifs[2];	/* "to"/"reply-to" interface */
+	frdest_t fr_dif;	/* duplicate packet interface */
+	/*
 	 * How big is the name buffer at the end?
 	 */
 	int	fr_namelen;
@@ -827,9 +830,10 @@ typedef	struct	frentry {
 
 #define	FR_NOLOGTAG	0
 
-#define	FR_CMPSIZ(_f)	((_f)->fr_size - \
-			 offsetof(struct frentry, fr_func))
+#define	FR_CMPSIZ	(offsetof(struct frentry, fr_ifnames) - \
+			offsetof(struct frentry, fr_func))
 #define	FR_NAME(_f, _n)	(_f)->fr_names + (_f)->_n
+#define FR_NUM(_a)	(sizeof(_a) / sizeof(*_a))
 
 
 /*



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201908250456.x7P4uY6s006477>