Add the dropped packet count to the received packet count, so that the

counts work similarly to the way they work in BPF and 2.4 and later
Linux (and so that we don't run the risk of the dropped packet count
being bigger than the received packet count, which can result in dropped
packet percentages > 100).

Author: yuguy

pcap-dlpi.c

@@ -62,7 +62,7 @@
 
 #ifndef lint
 static const char rcsid[] _U_ =
-    "@(#) $Header: /tcpdump/master/libpcap/pcap-dlpi.c,v 1.102 2004-04-23 05:19:04 guy Exp $ (LBL)";
+    "@(#) $Header: /tcpdump/master/libpcap/pcap-dlpi.c,v 1.103 2004-05-21 09:22:16 guy Exp $ (LBL)";
 #endif
 
 #ifdef HAVE_CONFIG_H
@@ -186,15 +186,19 @@ pcap_stats_dlpi(pcap_t *p, struct pcap_stat *ps)
 	/*
 	 * "ps_recv" counts packets handed to the filter, not packets
 	 * that passed the filter.  As filtering is done in userland,
-	 * this does not include packets dropped because we ran out
-	 * of buffer space.
+	 * this would not include packets dropped because we ran out
+	 * of buffer space; in order to make this more like other
+	 * platforms (and to keep from confusing applications that,
+	 * for example, compute packet drop percentages), we also make it
+	 * count packets dropped by "bufmod" (otherwise we might run
+	 * the risk of the packet drop count being bigger than the
+	 * received-packet count).
 	 *
-	 * "ps_drop" counts packets dropped inside the DLPI service
-	 * provider device device because of flow control requirements
-	 * or resource exhaustion; it doesn't count packets dropped by
-	 * the interface driver, or packets dropped upstream.  As
-	 * filtering is done in userland, it counts packets regardless
-	 * of whether they would've passed the filter.
+	 * "ps_drop" counts packets dropped by "bufmod" because of
+	 * flow control requirements or resource exhaustion; it doesn't
+	 * count packets dropped by the interface driver, or packets
+	 * dropped upstream.  As filtering is done in userland, it counts
+	 * packets regardless of whether they would've passed the filter.
 	 *
 	 * These statistics don't include packets not yet read from
 	 * the kernel by libpcap, but they may include packets not
@@ -307,6 +311,15 @@ pcap_read_dlpi(pcap_t *p, int cnt, pcap_handler callback, u_char *user)
 #endif
 			sbp = (struct sb_hdr *)bp;
 		p->md.stat.ps_drop = sbp->sbh_drops;
+		/*
+		 * On at least some other platforms (Linux 2.4 and later,
+		 * BSDs with BPF), "ps_recv" counts packets that were
+		 * received but dropped due to running out of buffer
+		 * space; we do that here, so that applications that
+		 * compute packet drop percentages don't see more packets
+		 * dropped than received.
+		 */
+		p->md.stat.ps_recv += sbp->sbh_drops;
 		pk = bp + sizeof(*sbp);
 		bp += sbp->sbh_totlen;
 		origlen = sbp->sbh_origlen;