2006-04-27 Richard Braun <syn@hurdfr.org>

	    Manuel Menal  <mmenal@hurdfr.org>

* device/if_hdr.h (struct ifnet): Added new members `if_snd_port_list'
	and `if_snd_port_list_lock'.
	* device/net_io.c: Reworked to improve BPF support.  Filters can be
	applied to ingress packets, egress packets, or both.
	* device/subrs.c: Initialize the `if_snd_port_list' and
	`if_snd_port_list_lock'.
	* include/device/bpf.h [0]: Enable unconditionally.
	Include <sys/types.h>.
	(BPF_IN, BPF_OUT): New macros.
	* include/device/net_status.h (NETF_TYPE_MASK, NETF_IN, NETF_OUT): New
	macros.
	(struct net_rcv_msg): New member `sent'.
	* linux/dev/glue/net.c: Mark ingress packets as received and inject
	egress packets into the packet filters.

7 files changed, 313 insertions, 178 deletions

diff --git a/ChangeLog b/ChangeLog
index dcc33ad..d9e2617 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,21 @@
+2006-04-27  Richard Braun  <syn@hurdfr.org>
+	    Manuel Menal  <mmenal@hurdfr.org>
+
+	* device/if_hdr.h (struct ifnet): Added new members `if_snd_port_list'
+	and `if_snd_port_list_lock'.
+	* device/net_io.c: Reworked to improve BPF support.  Filters can be
+	applied to ingress packets, egress packets, or both.
+	* device/subrs.c: Initialize the `if_snd_port_list' and
+	`if_snd_port_list_lock'.
+	* include/device/bpf.h [0]: Enable unconditionally.
+	Include <sys/types.h>.
+	(BPF_IN, BPF_OUT): New macros.
+	* include/device/net_status.h (NETF_TYPE_MASK, NETF_IN, NETF_OUT): New
+	macros.
+	(struct net_rcv_msg): New member `sent'.
+	* linux/dev/glue/net.c: Mark ingress packets as received and inject
+	egress packets into the packet filters.
+
 2006-04-26  Thomas Schwinge  <tschwinge@gnu.org>
 
 	* Makefile.in: Replace `make' with `$(MAKE)'.
diff --git a/device/if_hdr.h b/device/if_hdr.h
index fa53fd3..6ed30a3 100644
--- a/device/if_hdr.h
+++ b/device/if_hdr.h
@@ -79,8 +79,11 @@ struct ifnet {
 	char	*if_address;		/* pointer to hardware address */
 	struct ifqueue if_snd;		/* output queue */
 	queue_head_t if_rcv_port_list;	/* input filter list */
+	queue_head_t if_snd_port_list;	/* output filter list */
 	decl_simple_lock_data(,
-		if_rcv_port_list_lock)	/* lock for filter list */
+		if_rcv_port_list_lock)	/* lock for input filter list */
+	decl_simple_lock_data(,
+		if_snd_port_list_lock)	/* lock for output filter list */
 /* statistics */
 	int	if_ipackets;		/* packets received */
 	int	if_ierrors;		/* input errors */
diff --git a/device/net_io.c b/device/net_io.c
index e17c679..2001d63 100644
--- a/device/net_io.c
+++ b/device/net_io.c
@@ -288,7 +288,8 @@ net_kmsg_more(void)
  * filter for a single session.
  */
 struct net_rcv_port {
-	queue_chain_t	chain;		/* list of open_descriptors */
+	queue_chain_t	input;		/* list of input open_descriptors */
+	queue_chain_t	output;		/* list of output open_descriptors */
 	ipc_port_t	rcv_port;	/* port to send packet to */
 	int		rcv_qlimit;	/* port's qlimit */
 	int		rcv_count;	/* number of packets received */
@@ -348,15 +349,15 @@ decl_simple_lock_data(,net_hash_header_lock)
 	} while ((elt) != (head));
 
 
-#define FILTER_ITERATE(ifp, fp, nextfp) \
-	for ((fp) = (net_rcv_port_t) queue_first(&(ifp)->if_rcv_port_list);\
-	     !queue_end(&(ifp)->if_rcv_port_list, (queue_entry_t)(fp));    \
-	     (fp) = (nextfp)) {						   \
-		(nextfp) = (net_rcv_port_t) queue_next(&(fp)->chain);
+#define FILTER_ITERATE(if_port_list, fp, nextfp, chain)	\
+	for ((fp) = (net_rcv_port_t) queue_first(if_port_list);	\
+	     !queue_end(if_port_list, (queue_entry_t)(fp));	\
+	     (fp) = (nextfp)) {					\
+		(nextfp) = (net_rcv_port_t) queue_next(chain);
 #define FILTER_ITERATE_END }
 
 /* entry_p must be net_rcv_port_t or net_hash_entry_t */
-#define ENQUEUE_DEAD(dead, entry_p) { \
+#define ENQUEUE_DEAD(dead, entry_p, chain) {			\
 	queue_next(&(entry_p)->chain) = (queue_entry_t) (dead);	\
 	(dead) = (queue_entry_t)(entry_p);			\
 }
@@ -711,23 +712,36 @@ net_filter(kmsg, send_list)
  	queue_entry_t		dead_entp = (queue_entry_t) 0;
  	unsigned int		ret_count;
 
+	queue_head_t *if_port_list;
+
 	int count = net_kmsg(kmsg)->net_rcv_msg_packet_count;
 	ifp = (struct ifnet *) kmsg->ikm_header.msgh_remote_port;
 	ipc_kmsg_queue_init(send_list);
 
+	if (net_kmsg(kmsg)->sent)
+	    if_port_list = &ifp->if_snd_port_list;
+	else
+	    if_port_list = &ifp->if_rcv_port_list;
+
 	/*
 	 * Unfortunately we can't allocate or deallocate memory
-	 * while holding this lock.  And we can't drop the lock
-	 * while examining the filter list.
+	 * while holding these locks. And we can't drop the locks
+	 * while examining the filter lists.
+	 * Both locks are hold in case a filter is removed from both
+	 * queues.
 	 */
 	simple_lock(&ifp->if_rcv_port_list_lock);
- 	FILTER_ITERATE(ifp, infp, nextfp)
- 	{
+	simple_lock(&ifp->if_snd_port_list_lock);
+	FILTER_ITERATE(if_port_list, infp, nextfp,
+		       net_kmsg(kmsg)->sent ? &infp->output : &infp->input)
+	{
  	    entp = (net_hash_entry_t) 0;
- 	    if (infp->filter[0] == NETF_BPF) {
- 		ret_count = bpf_do_filter(infp, net_kmsg(kmsg)->packet, count,
- 					  net_kmsg(kmsg)->header,
- 					  &hash_headp, &entp);
+ 	    if ((infp->filter[0] & NETF_TYPE_MASK) == NETF_BPF) {
+ 		ret_count = bpf_do_filter(infp, net_kmsg(kmsg)->packet
+					  + sizeof(struct packet_header),
+					  count, net_kmsg(kmsg)->header,
+					  ifp->if_header_size, &hash_headp,
+					  &entp);
 		if (entp == (net_hash_entry_t) 0)
 		  dest = infp->rcv_port;
 		else
@@ -754,9 +768,15 @@ net_filter(kmsg, send_list)
 		     */
 
  		    if (entp == (net_hash_entry_t) 0) {
- 			queue_remove(&ifp->if_rcv_port_list, infp,
- 				     net_rcv_port_t, chain);
- 			ENQUEUE_DEAD(dead_infp, infp);
+			if (infp->filter[0] & NETF_IN)
+			    queue_remove(&ifp->if_rcv_port_list, infp,
+					 net_rcv_port_t, input);
+			if (infp->filter[0] & NETF_OUT)
+			    queue_remove(&ifp->if_snd_port_list, infp,
+					 net_rcv_port_t, output);
+
+			/* Use input only for queues of dead filters. */
+ 			ENQUEUE_DEAD(dead_infp, infp, input);
  			continue;
  		    } else {
  			hash_ent_remove (ifp,
@@ -808,22 +828,27 @@ net_filter(kmsg, send_list)
 		 * See if ordering of filters is wrong
 		 */
 		if (infp->priority >= NET_HI_PRI) {
-		    prevfp = (net_rcv_port_t) queue_prev(&infp->chain);
-		    /*
-		     * If infp is not the first element on the queue,
-		     * and the previous element is at equal priority
-		     * but has a lower count, then promote infp to
-		     * be in front of prevfp.
-		     */
-		    if ((queue_t)prevfp != &ifp->if_rcv_port_list &&
-			infp->priority == prevfp->priority) {
-			/*
-			 * Threshold difference to prevent thrashing
-			 */
-			if (net_filter_queue_reorder
-			    && (100 + prevfp->rcv_count < rcount))
-				reorder_queue(&prevfp->chain, &infp->chain);
+#define REORDER_PRIO(chain)						\
+		    prevfp = (net_rcv_port_t) queue_prev(&infp->chain);	\
+		    /*							\
+		     * If infp is not the first element on the queue,	\
+		     * and the previous element is at equal priority	\
+		     * but has a lower count, then promote infp to	\
+		     * be in front of prevfp.				\
+		     */							\
+		    if ((queue_t)prevfp != if_port_list &&		\
+			infp->priority == prevfp->priority) {		\
+			/*						\
+			 * Threshold difference to prevent thrashing	\
+			 */						\
+			if (net_filter_queue_reorder			\
+			    && (100 + prevfp->rcv_count < rcount))	\
+			    reorder_queue(&prevfp->chain, &infp->chain);\
 		    }
+
+		    REORDER_PRIO(input);
+		    REORDER_PRIO(output);
+
 		    /*
 		     * High-priority filter -> no more deliveries
 		     */
@@ -833,7 +858,7 @@ net_filter(kmsg, send_list)
 	    }
 	}
 	FILTER_ITERATE_END
-
+	simple_unlock(&ifp->if_snd_port_list_lock);
 	simple_unlock(&ifp->if_rcv_port_list_lock);
 
 	/*
@@ -872,7 +897,7 @@ net_do_filter(infp, data, data_count, header)
 #define	header_word	((unsigned short *)header)
 
 	sp = &stack[NET_FILTER_STACK_DEPTH];
-	fp = &infp->filter[0];
+	fp = &infp->filter[1]; /* filter[0] used for flags */
 	fpe = infp->filter_end;
 
 	*sp = TRUE;
@@ -999,6 +1024,10 @@ parse_net_filter(filter, count)
 	register filter_t	*fpe = &filter[count];
 	register filter_t	op, arg;
 
+	/*
+	 * count is at least 1, and filter[0] is used for flags.
+	 */
+	filter++;
 	sp = NET_FILTER_STACK_DEPTH;
 
 	for (; filter < fpe; filter++) {
@@ -1099,6 +1128,7 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
     int				i;
     int				ret, is_new_infp;
     io_return_t			rval;
+    boolean_t			in, out;
 
     /*
      * Check the filter syntax.
@@ -1107,13 +1137,19 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
     filter_bytes = CSPF_BYTES(filter_count);
     match = (bpf_insn_t) 0;
 
-    if (filter_count > 0 && filter[0] == NETF_BPF) {
+    if (filter_count == 0) {
+	return (D_INVALID_OPERATION);
+    } else if (!((filter[0] & NETF_IN) || (filter[0] & NETF_OUT))) {
+	return (D_INVALID_OPERATION); /* NETF_IN or NETF_OUT required */
+    } else if ((filter[0] & NETF_TYPE_MASK) == NETF_BPF) {
 	ret = bpf_validate((bpf_insn_t)filter, filter_bytes, &match);
 	if (!ret)
 	    return (D_INVALID_OPERATION);
-    } else {
+    } else if ((filter[0] & NETF_TYPE_MASK) == 0) {
 	if (!parse_net_filter(filter, filter_count))
 	    return (D_INVALID_OPERATION);
+    } else {
+	return (D_INVALID_OPERATION);
     }
 
     rval = D_SUCCESS;			/* default return value */
@@ -1129,8 +1165,8 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
 	is_new_infp = TRUE;
     } else {
         /*
-	 * If there is a match instruction, we assume there will
-	 * multiple session with a common substructure and allocate
+	 * If there is a match instruction, we assume there will be
+	 * multiple sessions with a common substructure and allocate
 	 * a hash table to deal with them.
 	 */
 	my_infp = 0;
@@ -1143,70 +1179,87 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
      * Look for filters with dead ports (for GC).
      * Look for a filter with the same code except KEY insns.
      */
-    
-    simple_lock(&ifp->if_rcv_port_list_lock);
-    
-    FILTER_ITERATE(ifp, infp, nextfp)
+    void check_filter_list(queue_head_t *if_port_list)
     {
+	FILTER_ITERATE(if_port_list, infp, nextfp,
+                       (if_port_list == &ifp->if_rcv_port_list)
+                       ? &infp->input : &infp->output)
+	{
 	    if (infp->rcv_port == MACH_PORT_NULL) {
-		    if (match != 0
-			&& infp->priority == priority
-			&& my_infp == 0
-			&& (infp->filter_end - infp->filter) == filter_count
-			&& bpf_eq((bpf_insn_t)infp->filter,
-				  filter, filter_bytes))
-			    {
-				    my_infp = infp;
-			    }
-
-		    for (i = 0; i < NET_HASH_SIZE; i++) {
-			    head = &((net_hash_header_t) infp)->table[i];
-			    if (*head == 0)
-				    continue;
-
-			    /*
-			     * Check each hash entry to make sure the
-			     * destination port is still valid.  Remove
-			     * any invalid entries.
-			     */
-			    entp = *head;
-			    do {
-				    nextentp = (net_hash_entry_t) entp->he_next;
+		if (match != 0
+		    && infp->priority == priority
+		    && my_infp == 0
+		    && (infp->filter_end - infp->filter) == filter_count
+		    && bpf_eq((bpf_insn_t)infp->filter,
+			      filter, filter_bytes))
+		    my_infp = infp;
+
+		for (i = 0; i < NET_HASH_SIZE; i++) {
+		    head = &((net_hash_header_t) infp)->table[i];
+		    if (*head == 0)
+			continue;
+
+		    /*
+		     * Check each hash entry to make sure the
+		     * destination port is still valid.  Remove
+		     * any invalid entries.
+		     */
+		    entp = *head;
+		    do {
+			nextentp = (net_hash_entry_t) entp->he_next;
   
-				    /* checked without 
-				       ip_lock(entp->rcv_port) */
-				    if (entp->rcv_port == rcv_port
-					|| !IP_VALID(entp->rcv_port)
-					|| !ip_active(entp->rcv_port)) {
-				
-					    ret = hash_ent_remove (ifp,
-						(net_hash_header_t)infp,
-						(my_infp == infp),
-						head,
-						entp,
-						&dead_entp);
-					    if (ret)
-						    goto hash_loop_end;
-				    }
+			/* checked without 
+			   ip_lock(entp->rcv_port) */
+			if (entp->rcv_port == rcv_port
+			    || !IP_VALID(entp->rcv_port)
+			    || !ip_active(entp->rcv_port)) {
+			    ret = hash_ent_remove (ifp,
+				(net_hash_header_t)infp,
+				(my_infp == infp),
+				head,
+				entp,
+				&dead_entp);
+			    if (ret)
+				goto hash_loop_end;
+			}
 			
-				    entp = nextentp;
-			    /* While test checks head since hash_ent_remove
-			       might modify it.
-			       */
-			    } while (*head != 0 && entp != *head);
-		    }
+			entp = nextentp;
+		    /* While test checks head since hash_ent_remove
+		       might modify it.
+		     */
+		    } while (*head != 0 && entp != *head);
+		}
+
 		hash_loop_end:
 		    ;
-		    
 	    } else if (infp->rcv_port == rcv_port
 		       || !IP_VALID(infp->rcv_port)
 		       || !ip_active(infp->rcv_port)) {
-		    /* Remove the old filter from list */
-		    remqueue(&ifp->if_rcv_port_list, (queue_entry_t)infp);
-		    ENQUEUE_DEAD(dead_infp, infp);
+
+		    /* Remove the old filter from lists */
+		    if (infp->filter[0] & NETF_IN)
+			queue_remove(&ifp->if_rcv_port_list, infp,
+				     net_rcv_port_t, input);
+		    if (infp->filter[0] & NETF_OUT)
+			queue_remove(&ifp->if_snd_port_list, infp,
+				     net_rcv_port_t, output);
+
+		    ENQUEUE_DEAD(dead_infp, infp, input);
 	    }
+	}
+	FILTER_ITERATE_END
     }
-    FILTER_ITERATE_END
+
+    in = (filter[0] & NETF_IN) != 0;
+    out = (filter[0] & NETF_OUT) != 0;
+
+    simple_lock(&ifp->if_rcv_port_list_lock);
+    simple_lock(&ifp->if_snd_port_list_lock);
+
+    if (in)
+	check_filter_list(&ifp->if_rcv_port_list);
+    if (out)
+	check_filter_list(&ifp->if_snd_port_list);
 
     if (my_infp == 0) {
 	/* Allocate a dummy infp */
@@ -1217,6 +1270,7 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
 	}
 	if (i == N_NET_HASH) {
 	    simple_unlock(&net_hash_header_lock);
+	    simple_unlock(&ifp->if_snd_port_list_lock);
 	    simple_unlock(&ifp->if_rcv_port_list_lock);
 
             ipc_port_release_send(rcv_port);
@@ -1257,10 +1311,21 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
 	}
 
 	/* Insert my_infp according to priority */
-	queue_iterate(&ifp->if_rcv_port_list, infp, net_rcv_port_t, chain)
-	    if (priority > infp->priority)
-		break;
-	enqueue_tail((queue_t)&infp->chain, (queue_entry_t)my_infp);
+	if (in) {
+	    queue_iterate(&ifp->if_rcv_port_list, infp, net_rcv_port_t, input)
+		if (priority > infp->priority)
+		    break;
+
+	    queue_enter(&ifp->if_rcv_port_list, my_infp, net_rcv_port_t, input);
+	}
+
+	if (out) {
+	    queue_iterate(&ifp->if_snd_port_list, infp, net_rcv_port_t, output)
+		if (priority > infp->priority)
+		    break;
+
+	    queue_enter(&ifp->if_snd_port_list, my_infp, net_rcv_port_t, output);
+	}
     }
     
     if (match != 0)
@@ -1284,9 +1349,9 @@ net_set_filter(ifp, rcv_port, priority, filter, filter_count)
 
 	((net_hash_header_t)my_infp)->ref_count++;
 	hash_entp->rcv_qlimit = net_add_q_info(rcv_port);
-
     }
     
+    simple_unlock(&ifp->if_snd_port_list_lock);
     simple_unlock(&ifp->if_rcv_port_list_lock);
 
 clean_and_return:
@@ -1537,11 +1602,12 @@ net_io_init()
  */
 
 int
-bpf_do_filter(infp, p, wirelen, header, hash_headpp, entpp)
+bpf_do_filter(infp, p, wirelen, header, hlen, hash_headpp, entpp)
 	net_rcv_port_t	infp;
 	char *		p;		/* packet data */
 	unsigned int	wirelen;	/* data_count (in bytes) */
 	char *		header;
+	unsigned int    hlen;           /* header len (in bytes) */
 	net_hash_entry_t	**hash_headpp, *entpp;	/* out */
 {
 	register bpf_insn_t pc, pc_end;
@@ -1551,8 +1617,11 @@ bpf_do_filter(infp, p, wirelen, header, hash_headpp, entpp)
 	register int k;
 	long mem[BPF_MEMWORDS];
 
+	/* Generic pointer to either HEADER or P according to the specified offset. */
+	char *data = NULL;
+
 	pc = ((bpf_insn_t) infp->filter) + 1;
-					/* filter[0].code is BPF_BEGIN */
+				/* filter[0].code is (NETF_BPF | flags) */
 	pc_end = (bpf_insn_t)infp->filter_end;
 	buflen = NET_RCV_MAX;
 	*entpp = 0;			/* default */
@@ -1596,58 +1665,53 @@ bpf_do_filter(infp, p, wirelen, header, hash_headpp, entpp)
 
 		case BPF_LD|BPF_W|BPF_ABS:
 			k = pc->k;
-			if ((u_int)k + sizeof(long) <= buflen) {
-#ifdef BPF_ALIGN
-				if (((int)(p + k) & 3) != 0)
-					A = EXTRACT_LONG(&p[k]);
-				else
-#endif
-					A = ntohl(*(long *)(p + k));
-				continue;
-			}
 
-			k -= BPF_DLBASE;
-			if ((u_int)k + sizeof(long) <= NET_HDW_HDR_MAX) {
+		load_word:
+			if ((u_int)k + sizeof(long) <= hlen)
+			     data = header;
+			else if ((u_int)k + sizeof(long) <= buflen) {
+			     k -= hlen;
+			     data = p;
+			} else
+			     return 0;
+
 #ifdef BPF_ALIGN
-				if (((int)(header + k) & 3) != 0)
-					A = EXTRACT_LONG(&header[k]);
-				else
+			if (((int)(data + k) & 3) != 0)
+			     A = EXTRACT_LONG(&data[k]);
+			else
 #endif
-					A = ntohl(*(long *)(header + k));
-				continue;
-			} else {
-				return 0;
-			}
+			     A = ntohl(*(long *)(data + k));
+			continue;
 
 		case BPF_LD|BPF_H|BPF_ABS:
 			k = pc->k;
-			if ((u_int)k + sizeof(short) <= buflen) {
-				A = EXTRACT_SHORT(&p[k]);
-				continue;
-			}
 
-			k -= BPF_DLBASE;
-			if ((u_int)k + sizeof(short) <= NET_HDW_HDR_MAX) {
-				A = EXTRACT_SHORT(&header[k]);
-				continue;
-			} else {
-				return 0;
-			}
+		load_half:
+			if ((u_int)k + sizeof(short) <= hlen)
+			     data = header;
+			else if ((u_int)k + sizeof(short) <= buflen) {
+			     k -= hlen;
+			     data = p;
+			} else
+			     return 0;
+
+			A = EXTRACT_SHORT(&data[k]);
+			continue;
 
 		case BPF_LD|BPF_B|BPF_ABS:
-			k = pc->k;
-			if ((u_int)k < buflen) {
-				A = p[k];
-				continue;
-			}
-			
-			k -= BPF_DLBASE;
-			if ((u_int)k < NET_HDW_HDR_MAX) {
-				A = header[k];
-				continue;
-			} else {
-				return 0;
-			}
+		        k = pc->k;
+
+		load_byte:
+			if ((u_int)k < hlen)
+			     data = header;
+			else if ((u_int)k < buflen) {
+			     data = p;
+			     k -= hlen;
+			} else
+			     return 0;
+
+			A = data[k];
+			continue;
 
 		case BPF_LD|BPF_W|BPF_LEN:
 			A = wirelen;
@@ -1659,35 +1723,27 @@ bpf_do_filter(infp, p, wirelen, header, hash_headpp, entpp)
 
 		case BPF_LD|BPF_W|BPF_IND:
 			k = X + pc->k;
-			if (k + sizeof(long) > buflen)
-				return 0;
-#ifdef BPF_ALIGN
-			if (((int)(p + k) & 3) != 0)
-				A = EXTRACT_LONG(&p[k]);
-			else
-#endif
-				A = ntohl(*(long *)(p + k));
-			continue;
-
+			goto load_word;
+			
 		case BPF_LD|BPF_H|BPF_IND:
 			k = X + pc->k;
-			if (k + sizeof(short) > buflen)
-				return 0;
-			A = EXTRACT_SHORT(&p[k]);
-			continue;
+			goto load_half;
 
 		case BPF_LD|BPF_B|BPF_IND:
 			k = X + pc->k;
-			if (k >= buflen)
-				return 0;
-			A = p[k];
-			continue;
+			goto load_byte;
 
 		case BPF_LDX|BPF_MSH|BPF_B:
 			k = pc->k;
-			if (k >= buflen)
-				return 0;
-			X = (p[pc->k] & 0xf) << 2;
+			if (k < hlen)
+			     data = header;
+			else if (k < buflen) {
+			     data = p;
+			     k -= hlen;
+			} else
+			     return 0;
+
+			X = (data[k] & 0xf) << 2;
 			continue;
 
 		case BPF_LD|BPF_IMM:
@@ -1855,7 +1911,11 @@ bpf_validate(f, bytes, match)
 	register bpf_insn_t p;
 
 	len = BPF_BYTES2LEN(bytes);
-	/* f[0].code is already checked to be BPF_BEGIN. So skip f[0]. */
+
+	/*
+	 * f[0].code is already checked to be (NETF_BPF | flags).
+	 * So skip f[0].
+	 */
 
 	for (i = 1; i < len; ++i) {
 		/*
@@ -1984,7 +2044,7 @@ bpf_match (hash, n_keys, keys, hash_headpp, entpp)
 /*
  * Removes a hash entry (ENTP) from its queue (HEAD).
  * If the reference count of filter (HP) becomes zero and not USED,
- * HP is removed from ifp->if_rcv_port_list and is freed.
+ * HP is removed from the corresponding port lists and is freed.
  */
 
 int
@@ -1998,13 +2058,18 @@ hash_ent_remove (ifp, hp, used, head, entp, dead_p)
 	hp->ref_count--;
 
 	if (*head == entp) {
-
 		if (queue_empty((queue_t) entp)) {
 			*head = 0;
-			ENQUEUE_DEAD(*dead_p, entp);
+			ENQUEUE_DEAD(*dead_p, entp, chain);
 			if (hp->ref_count == 0 && !used) {
-				remqueue((queue_t) &ifp->if_rcv_port_list,
-					 (queue_entry_t)hp);
+				if (((net_rcv_port_t)hp)->filter[0] & NETF_IN)
+					queue_remove(&ifp->if_rcv_port_list,
+						     (net_rcv_port_t)hp,
+						     net_rcv_port_t, input);
+				if (((net_rcv_port_t)hp)->filter[0] & NETF_OUT)
+					queue_remove(&ifp->if_snd_port_list,
+						     (net_rcv_port_t)hp,
+						     net_rcv_port_t, output);
 				hp->n_keys = 0;
 				return TRUE;
 			}
@@ -2015,7 +2080,7 @@ hash_ent_remove (ifp, hp, used, head, entp, dead_p)
 	}
 
 	remqueue((queue_t)*head, (queue_entry_t)entp);
-	ENQUEUE_DEAD(*dead_p, entp);
+	ENQUEUE_DEAD(*dead_p, entp, chain);
 	return FALSE;
 }    
 
@@ -2069,7 +2134,7 @@ net_free_dead_infp (dead_infp)
 
 	for (infp = (net_rcv_port_t) dead_infp; infp != 0; infp = nextfp)
 	{
-		nextfp = (net_rcv_port_t) queue_next(&infp->chain);
+		nextfp = (net_rcv_port_t) queue_next(&infp->input);
 		ipc_port_release_send(infp->rcv_port);
 		net_del_q_info(infp->rcv_qlimit);
 		zfree(net_rcv_zone, (vm_offset_t) infp);
diff --git a/device/subrs.c b/device/subrs.c
index a65d228..e6e8b6b 100644
--- a/device/subrs.c
+++ b/device/subrs.c
@@ -82,7 +82,9 @@ void if_init_queues(ifp)
 {
 	IFQ_INIT(&ifp->if_snd);
 	queue_init(&ifp->if_rcv_port_list);
+	queue_init(&ifp->if_snd_port_list);
 	simple_lock_init(&ifp->if_rcv_port_list_lock);
+	simple_lock_init(&ifp->if_snd_port_list_lock);
 }
 
 
diff --git a/include/device/bpf.h b/include/device/bpf.h
index 05cbf9b..a4c4eef 100644
--- a/include/device/bpf.h
+++ b/include/device/bpf.h
@@ -72,7 +72,8 @@
 #ifndef _DEVICE_BPF_H_
 #define _DEVICE_BPF_H_
 
-#if 0	/* not used in MK now */
+#include <sys/types.h> /* u_short */
+
 /*
  * Alignment macros.  BPF_WORDALIGN rounds up to the next 
  * even multiple of BPF_ALIGNMENT. 
@@ -115,14 +116,14 @@ struct bpf_version {
 #define DLT_PPP		9	/* Point-to-point Protocol */
 #define DLT_FDDI	10	/* FDDI */
 
-#endif /* 0 */
-
 /*
  * The instruction encondings.
  */
 
-/* Magic number for the first instruction */
-#define BPF_BEGIN NETF_BPF
+/* Magic number and flags for the first instruction */
+#define BPF_BEGIN	NETF_BPF
+#define BPF_IN		NETF_IN
+#define BPF_OUT		NETF_OUT
 
 /* instruction classes */
 #define BPF_CLASS(code) ((code) & 0x07)
diff --git a/include/device/net_status.h b/include/device/net_status.h
index dbf5e39..5131ef5 100644
--- a/include/device/net_status.h
+++ b/include/device/net_status.h
@@ -98,6 +98,11 @@ struct net_status {
  *  If the final value of the filter operation is true, then the packet is
  *  accepted for the filter.
  *
+ *  The first filter_t object is a header which allows to set flags for the
+ *  filter code. Main flags concern the direction of packets. This header is
+ *  split in the same way NETF words are : the 6 MSB bits indicate the type
+ *  of filter while the 10 LSB bits are the flags. For native NETF filters,
+ *  clear the 6 MSB bits (which is why there is no dedicated macro).
  */
 
 typedef	unsigned short	filter_t;
@@ -112,6 +117,14 @@ typedef filter_t	*filter_array_t;
 #define	NETF_ARG(word)	((word) & 0x3ff)
 #define	NETF_OP(word)	(((word)>>NETF_NBPA)&0x3f)
 
+/*  filter types  */
+#define NETF_TYPE_MASK	(((1 << NETF_NBPO) - 1) << NETF_NBPA)
+#define NETF_BPF	(1 << NETF_NBPA)
+
+/*  flags  */
+#define NETF_IN		0x1
+#define NETF_OUT	0x2
+
 /*  binary operators  */
 #define NETF_NOP	(0<<NETF_NBPA)
 #define NETF_EQ		(1<<NETF_NBPA)
@@ -131,7 +144,6 @@ typedef filter_t	*filter_array_t;
 #define	NETF_RSH	(15<<NETF_NBPA)
 #define	NETF_ADD	(16<<NETF_NBPA)
 #define	NETF_SUB	(17<<NETF_NBPA)
-#define NETF_BPF	(((1 << NETF_NBPO) - 1) << NETF_NBPA)
 
 
 /*  stack arguments  */
@@ -178,6 +190,7 @@ struct net_rcv_msg {
 	char		header[NET_HDW_HDR_MAX];
 	mach_msg_type_t	packet_type;
 	char		packet[NET_RCV_MAX];
+	boolean_t	sent;
 };
 typedef struct net_rcv_msg 	*net_rcv_msg_t;
 #define	net_rcv_msg_packet_count packet_type.msgt_number
diff --git a/linux/dev/glue/net.c b/linux/dev/glue/net.c
index 93556db..4a989d3 100644
--- a/linux/dev/glue/net.c
+++ b/linux/dev/glue/net.c
@@ -290,6 +290,9 @@ netif_rx (struct sk_buff *skb)
   eh = (struct ether_header *) (net_kmsg (kmsg)->header);
   ph = (struct packet_header *) (net_kmsg (kmsg)->packet);
   memcpy (eh, skb->data, sizeof (struct ether_header));
+
+  /* packet is prefixed with a struct packet_header,
+     see include/device/net_status.h.  */
   memcpy (ph + 1, skb->data + sizeof (struct ether_header),
 	  skb->len - sizeof (struct ether_header));
   ph->type = eh->ether_type;
@@ -298,6 +301,8 @@ netif_rx (struct sk_buff *skb)
 
   dev_kfree_skb (skb, FREE_READ);
 
+  net_kmsg(kmsg)->sent = FALSE; /* Mark packet as received.  */
+
   /* Pass packet up to the microkernel.  */
   net_packet (&dev->net_data->ifnet, kmsg,
 	      ph->length, ethernet_priority (kmsg));
@@ -484,6 +489,34 @@ device_write (void *d, ipc_port_t reply_port,
     }
   splx (s);
 
+  /* Send packet to filters.  */
+  {
+    struct packet_header *packet;
+    struct ether_header *header;
+    ipc_kmsg_t kmsg;
+
+    kmsg = net_kmsg_get ();
+
+    if (kmsg != IKM_NULL)
+      {
+        /* Suitable for Ethernet only.  */
+        header = (struct ether_header *) (net_kmsg (kmsg)->header);
+        packet = (struct packet_header *) (net_kmsg (kmsg)->packet);
+        memcpy (header, skb->data, sizeof (struct ether_header));
+
+        /* packet is prefixed with a struct packet_header,
+           see include/device/net_status.h.  */
+        memcpy (packet + 1, skb->data + sizeof (struct ether_header),
+                skb->len - sizeof (struct ether_header));
+        packet->length = skb->len - sizeof (struct ether_header)
+                         + sizeof (struct packet_header);
+        packet->type = header->ether_type;
+        net_kmsg (kmsg)->sent = TRUE; /* Mark packet as sent.  */
+        net_packet (&dev->net_data->ifnet, kmsg, packet->length,
+                    ethernet_priority (kmsg));
+      }
+  }
+
   return MIG_NO_REPLY;
 }