NetBSD-5.0.2/dist/libpcap/README.linux

In order for libpcap to be able to capture packets on a Linux system,
the "packet" protocol must be supported by your kernel.  If it is not,
you may get error messages such as

	modprobe: can't locate module net-pf-17

in "/var/adm/messages", or may get messages such as

	socket: Address family not supported by protocol

from applications using libpcap.

You must configure the kernel with the CONFIG_PACKET option for this
protocol; the following note is from the Linux "Configure.help" file for
the 2.0[.x] kernel:

	Packet socket
	CONFIG_PACKET
	  The Packet protocol is used by applications which communicate
	  directly with network devices without an intermediate network
	  protocol implemented in the kernel, e.g. tcpdump. If you want them
	  to work, choose Y. 

	  This driver is also available as a module called af_packet.o ( =
	  code which can be inserted in and removed from the running kernel
	  whenever you want). If you want to compile it as a module, say M
	  here and read Documentation/modules.txt; if you use modprobe or
	  kmod, you may also want to add "alias net-pf-17 af_packet" to 
	  /etc/modules.conf.

and the note for the 2.2[.x] kernel says:

	Packet socket
	CONFIG_PACKET
	  The Packet protocol is used by applications which communicate
	  directly with network devices without an intermediate network
	  protocol implemented in the kernel, e.g. tcpdump. If you want them
	  to work, choose Y. This driver is also available as a module called
	  af_packet.o ( = code which can be inserted in and removed from the
	  running kernel whenever you want). If you want to compile it as a
	  module, say M here and read Documentation/modules.txt.  You will
	  need to add 'alias net-pf-17 af_packet' to your /etc/conf.modules
	  file for the module version to function automatically.  If unsure,
	  say Y.

In addition, there is an option that, in 2.2 and later kernels, will
allow packet capture filters specified to programs such as tcpdump to be
executed in the kernel, so that packets that don't pass the filter won't
be copied from the kernel to the program, rather than having all packets
copied to the program and libpcap doing the filtering in user mode. 

Copying packets from the kernel to the program consumes a significant
amount of CPU, so filtering in the kernel can reduce the overhead of
capturing packets if a filter has been specified that discards a
significant number of packets.  (If no filter is specified, it makes no
difference whether the filtering isn't performed in the kernel or isn't
performed in user mode. :-))

The option for this is the CONFIG_FILTER option; the "Configure.help"
file says:

	Socket filtering
	CONFIG_FILTER
	  The Linux Socket Filter is derived from the Berkeley Packet Filter.
	  If you say Y here, user-space programs can attach a filter to any
	  socket and thereby tell the kernel that it should allow or disallow
	  certain types of data to get through the socket. Linux Socket
	  Filtering works on all socket types except TCP for now. See the text
	  file linux/Documentation/networking/filter.txt for more information.
	  If unsure, say N.


Statistics:
Statistics reported by pcap are platform specific.  The statistics
reported by pcap_stats on Linux are as follows:

2.2.x
=====
ps_recv   Number of packets that were accepted by the pcap filter
ps_drops  Always 0, this statistic is not gatherd on this platform

2.4.x
=====
ps_rec    Number of packets that were accepted by the pcap filter
ps_drops  Number of packets that had passed filtering but were not
          passed on to pcap due to things like buffer shortage, etc.
			 This is useful because these are packets you are interested in
			 but won't be reported by, for example, tcpdump output.