freebsd-skq/share/man/man4/netgraph.4

.\" Copyright (c) 1996-1999 Whistle Communications, Inc.
.\" All rights reserved.
.\" 
.\" Subject to the following obligations and disclaimer of warranty, use and
.\" redistribution of this software, in source or object code forms, with or
.\" without modifications are expressly permitted by Whistle Communications;
.\" provided, however, that:
.\" 1. Any and all reproductions of the source or object code must include the
.\"    copyright notice above and the following disclaimer of warranties; and
.\" 2. No rights are granted, in any manner or form, to use Whistle
.\"    Communications, Inc. trademarks, including the mark "WHISTLE
.\"    COMMUNICATIONS" on advertising, endorsements, or otherwise except as
.\"    such appears in the above copyright notice or in the software.
.\" 
.\" THIS SOFTWARE IS BEING PROVIDED BY WHISTLE COMMUNICATIONS "AS IS", AND
.\" TO THE MAXIMUM EXTENT PERMITTED BY LAW, WHISTLE COMMUNICATIONS MAKES NO
.\" REPRESENTATIONS OR WARRANTIES, EXPRESS OR IMPLIED, REGARDING THIS SOFTWARE,
.\" INCLUDING WITHOUT LIMITATION, ANY AND ALL IMPLIED WARRANTIES OF
.\" MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT.
.\" WHISTLE COMMUNICATIONS DOES NOT WARRANT, GUARANTEE, OR MAKE ANY
.\" REPRESENTATIONS REGARDING THE USE OF, OR THE RESULTS OF THE USE OF THIS
.\" SOFTWARE IN TERMS OF ITS CORRECTNESS, ACCURACY, RELIABILITY OR OTHERWISE.
.\" IN NO EVENT SHALL WHISTLE COMMUNICATIONS BE LIABLE FOR ANY DAMAGES
.\" RESULTING FROM OR ARISING OUT OF ANY USE OF THIS SOFTWARE, INCLUDING
.\" WITHOUT LIMITATION, ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY,
.\" PUNITIVE, OR CONSEQUENTIAL DAMAGES, PROCUREMENT OF SUBSTITUTE GOODS OR
.\" SERVICES, LOSS OF USE, DATA OR PROFITS, HOWEVER CAUSED AND UNDER ANY
.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
.\" THIS SOFTWARE, EVEN IF WHISTLE COMMUNICATIONS IS ADVISED OF THE POSSIBILITY
.\" OF SUCH DAMAGE.
.\" 
.\" Authors: Julian Elischer <julian@whistle.com>
.\"          Archie Cobbs <archie@whistle.com>
.\"
.\" $FreeBSD$
.\" $Whistle: netgraph.4,v 1.7 1999/01/28 23:54:52 julian Exp $
.\"
.Dd January 19, 1999
.Dt NETGRAPH 4
.Os FreeBSD
.Sh NAME
.Nm netgraph
.Nd graph based kernel networking subsystem
.Sh DESCRIPTION
The
.Nm 
system provides a uniform and modular system for the implementation
of kernel objects which perform various networking functions. The objects,
known as 
.Em nodes ,
can be arranged into arbitrarily complicated graphs. Nodes have
.Em hooks
which are used to connect two nodes together, forming the edges in the graph.
Nodes communicate along the edges to process data, implement protocols, etc.
.Pp
The aim of
.Nm 
is to supplement rather than replace the existing kernel networking
infrastructure. It provides:
.Pp
.Bl -bullet -compact -offset 2n
.It
A flexible way of combining protocol and link level drivers
.It
A modular way to implement new protocols
.It
A common framework for kernel entities to inter-communicate
.It
A reasonably fast, kernel-based implementation
.El
.Sh Nodes and Types
The most fundamental concept in
.Nm
is that of a
.Em node .
All nodes implement a number of predefined methods which allow them
to interact with other nodes in a well defined manner.
.Pp
Each node has a
.Em type ,
which is a static property of the node determined at node creation time.
A node's type is described by a unique ASCII type name.
The type implies what the node does and how it may be connected
to other nodes.
.Pp
In object-oriented language, types are classes and nodes are instances
of their respective class. All node types are subclasses of the generic node
type, and hence inherit certain common functionality and capabilities
(e.g., the ability to have an ASCII name).
.Pp
Nodes may be assigned a globally unique ASCII name which can be
used to refer to the node.
The name must not contain the characters ``.'' or  ``:'' and is limited to
.Dv "NG_NODELEN + 1"
characters (including NUL byte).
.Pp
Each node instance has a unique
.Em ID number
which is expressed as a 32-bit hex value. This value may be used to
refer to a node when there is no ASCII name assigned to it.
.Sh Hooks
Nodes are connected to other nodes by connecting a pair of
.Em hooks ,
one from each node. Data flows bidirectionally between nodes along
connected pairs of hooks.  A node may have as many hooks as it
needs, and may assign whatever meaning it wants to a hook.
.Pp
Hooks have these properties:
.Pp
.Bl -bullet -compact -offset 2n
.It
A hook has an ASCII name which is unique among all hooks
on that node (other hooks on other nodes may have the same name).
The name must not contain a ``.'' or a ``:''  and is
limited to
.Dv "NG_HOOKLEN + 1"
characters (including NUL byte).
.It
A hook is always connected to another hook. That is, hooks are
created at the time they are connected, and breaking an edge by
removing either hook destroys both hooks.
.El
.Pp
A node may decide to assign special meaning to some hooks. 
For example, connecting to the hook named ``debug'' might trigger
the node to start sending debugging information to that hook.
.Sh Data Flow
Two types of information flow between nodes: data messages and
control messages. Data messages are passed in mbuf chains along the edges
in the graph, one edge at a time. The first mbuf in a chain must have the
.Dv M_PKTHDR
flag set. Each node decides how to handle data coming in on its hooks.
.Pp
Control messages are type-specific structures sent from one node directly
to an arbitrary other node. There are two ways to address such a message. If
there is a sequence of edges connecting the two nodes, the message
may be ``source routed'' by specifying the corresponding sequence
of hooks as the destination address for the message (relative
addressing).  Otherwise, the recipient node global ASCII name
(or equivalent ID based name) is used as the destination address
for the message (absolute addressing).  The two types of addressing
may be combined, by specifying an absolute start node and a sequence
of hooks.
.Pp
Messages often represent commands that are followed by a reply message
in the reverse direction. To facilitate this, the recipient of a
control message is supplied with a ``return address'' that is suitable
for addressing a reply.
.Pp
Each control message contains a 32 bit value called a
.Em typecookie
indicating the type of the message, i.e., how to interpret it.
Typically each type defines a unique typecookie for the messages
that it understands.  However, a node may choose to recognize and
implement more than one type of message.
.Sh Netgraph is Functional
In order to minimize latency, most
.Nm netgraph
operations are functional.
That is, data and control messages are delivered by making function
calls rather than by using queues and mailboxes.  For example, if node
A wishes to send a data mbuf to neighboring node B, it calls the
generic
.Nm
data delivery function. This function in turn locates
node B and calls B's ``receive data'' method. While this mode of operation
results in good performance, it has a few implications for node
developers:
.Pp
.Bl -bullet -compact -offset 2n
.It
Whenever a node delivers a data or control message, the node
may need to allow for the possibility of receiving a returning message
before the original delivery function call returns.
.It
Netgraph nodes and support routines generally run at
.Dv "splnet()" .
However, some nodes may want to send data and control messages
from a different priority level. Netgraph supplies queueing routines which
utilize the NETISR system to move message delivery to 
.Dv "splnet()" .
Note that messages are always received at
.Dv "splnet()" .
.It
It's possible for an infinite loop to occur if the graph contains cycles.
.El
.Pp
So far, these issues have not proven problematical in practice.
.Sh Interaction With Other Parts of the Kernel
A node may have a hidden interaction with other components of the
kernel outside of the
.Nm
subsystem, such as device hardware,
kernel protocol stacks, etc.  In fact, one of the benefits of
.Nm
is the ability to join disparate kernel networking entities together in a
consistent communication framework.
.Pp
An example is the node type
.Em socket 
which is both a netgraph node and a
.Xr socket 2
BSD socket in the protocol family
.Dv PF_NETGRAPH .
Socket nodes allow user processes to participate in
.Nm netgraph .
Other nodes communicate with socket nodes using the usual methods, and the
node hides the fact that it is also passing information to and from a
cooperating user process.
.Pp
Another example is a device driver that presents
a node interface to the hardware.
.Sh Node Methods
Nodes are notified of the following actions via function calls
to the following node methods (all at
.Dv "splnet()" )
and may accept or reject that action (by returning the appropriate
error code):
.Bl -tag -width xxx
.It Creation of a new node
The constructor for the type is called. If creation of a new node is 
allowed, the constructor must call the generic node creation
function (in object-oriented terms, the superclass constructor)
and then allocate any special resources it needs. For nodes that
correspond to hardware, this is typically done during the device
attach routine. Often a global ASCII name corresponding to the
device name is assigned here as well.
.It Creation of a new hook
The hook is created and tentatively
linked to the node, and the node is told about the name that will be 
used to describe this hook. The node sets up any special data structures
it needs, or may reject the connection, based on the name of the hook.
.It Successful connection of two hooks
After both ends have accepted their
hooks, and the links have been made, the nodes get a chance to
find out who their peer is across the link and can then decide to reject
the connection. Tear-down is automatic.
.It Destruction of a hook
The node is notified of a broken connection. The node may consider some hooks
to be critical to operation and others to be expendable: the disconnection
of one hook may be an acceptable event while for another it
may effect a total shutdown for the node.
.It Shutdown of a node
This method allows a node to clean up
and to ensure that any actions that need to be performed
at this time are taken. The method must call the generic (i.e., superclass)
node destructor to get rid of the generic components of the node.
Some nodes (usually associated with a piece of hardware) may be
.Em persistent
in that a shutdown breaks all edges and resets the node,
but doesn't remove it, in which case the generic destructor is not called.
.El
.Sh Sending and Receiving Data
Three other methods are also supported by all nodes:
.Bl -tag -width xxx
.It Receive data message
An mbuf chain is passed to the node.
The node is notified on which hook the data arrived,
and can use this information in its processing decision.
The node must must always 
.Dv m_freem()
the mbuf chain on completion or error, or pass it on to another node
(or kernel module) which will then be responsible for freeing it.
.Pp
In addition to the mbuf chain itself there is also a pointer to a 
structure describing meta-data about the message
(e.g. priority information). This pointer may be
.Dv NULL
if there is no additional information. The format for this information is
described in 
.Dv netgraph.h .
The memory for meta-data must allocated via
.Dv malloc()
with type
.Dv M_NETGRAPH .
As with the data itself, it is the receiver's responsibility to
.Dv free()
the meta-data. If the mbuf chain is freed the meta-data must
be freed at the same time. If the meta-data is freed but the
real data on is passed on, then a
.Dv NULL
pointer must be substituted.
.Pp
The receiving node may decide to defer the data by queueing it in the
.Nm
NETISR system (see below).
.Pp
The structure and use of meta-data is still experimental, but is presently used in
frame-relay to indicate that management packets should be queued for transmission
at a higher priority than data packets. This is required for
conformance with Frame Relay standards.
.Pp
.It Receive queued data message
Usually this will be the same function as 
.Em Receive data message.
This is the entry point called when a data message is being handed to 
the node after having been queued in the NETISR system.
This allows a node to decide in the 
.Em Receive data message
method that a message should be deferred and queued,
and be sure that when it is processed from the queue,
it will not be queued again.
.It Receive control message
This method is called when a control message is addressed to the node.
A return address is always supplied, giving the address of the node
that originated the message so a reply message can be sent anytime later.
.Pp
It is possible for a synchronous reply to be made, and in fact this
is more common in practice.
This is done by setting a pointer (supplied as an extra function parameter)
to point to the reply.
Then when the control message delivery function returns,
the caller can check if this pointer has been made non-NULL,
and if so then it points to the reply message allocated via
.Dv malloc()
and containing the synchronous response. In both directions, 
(request and response) it is up to the 
receiver of that message to 
.Dv free()
the control message buffer. All control messages and replies are
allocated with
.Dv malloc()
type
.Dv M_NETGRAPH .
.El
.Pp
Much use has been made of reference counts, so that nodes being
free'd of all references are automatically freed, and this behaviour
has been tested and debugged to present a consistent and trustworthy
framework for the ``type module'' writer to use.
.Sh Addressing
The 
.Nm
framework provides an unambiguous and simple to use method of specifically
addressing any single node in the graph. The naming of a node is 
independent of its type, in that another node, or external component
need not know anything about the node's type in order to address it so as 
to send it a generic message type. Node and hook names should be
chosen so as to make addresses meaningful.
.Pp
Addresses are either absolute or relative. An absolute address begins
with a node name, (or ID), followed by a colon, followed by a sequence of hook
names separated by periods. This addresses the node reached by starting
at the named node and following the specified sequence of hooks.
A relative address includes only the sequence of hook names, implicitly
starting hook traversal at the local node.
.Pp
There are a couple of special possibilities for the node name.
The name ``.'' (referred to as ``.:'') always refers to the local node.
Also, nodes that have no global name may be addressed by their ID numbers,
by enclosing the hex representation of the ID number within square brackets.
Here are some examples of valid netgraph addresses:
.Bd -literal -offset 4n -compact

  .:
  foo:
  .:hook1
  foo:hook1.hook2
  [f057cd80]:hook1
.Ed
.Pp
Consider the following set of nodes might be created for a site with
a single physical frame relay line having two active logical DLCI channels,
with RFC-1490 frames on DLCI 16 and PPP frames over DLCI 20:
.Pp
.Bd -literal
[type SYNC ]                  [type FRAME]                 [type RFC1490]
[ "Frame1" ](uplink)<-->(data)[<un-named>](dlci16)<-->(mux)[<un-named>  ]
[    A     ]                  [    B     ](dlci20)<---+    [     C      ]
                                                      |
                                                      |      [ type PPP ]
                                                      +>(mux)[<un-named>]
                                                             [    D     ]
.Ed
.Pp
One could always send a control message to node C from anywhere
by using the name
.Em "Frame1:uplink.dlci16" .
Similarly, 
.Em "Frame1:uplink.dlci20"
could reliably be used to reach node D, and node A could refer
to node B as
.Em ".:uplink" ,
or simply
.Em "uplink" .
Conversely, B can refer to A as
.Em "data" .
The address
.Em "mux.data"
could be used by both nodes C and D to address a message to node A.
.Pp
Note that this is only for
.Em control messages .
Data messages are routed one hop at a time, by specifying the departing
hook, with each node making the next routing decision. So when B
receives a frame on hook
.Em data
it decodes the frame relay header to determine the DLCI,
and then forwards the unwrapped frame to either C or D.
.Pp
A similar graph might be used to represent multi-link PPP running
over an ISDN line:
.Pp
.Bd -literal
[ type BRI ](B1)<--->(link1)[ type MPP  ]
[  "ISDN1" ](B2)<--->(link2)[ (no name) ]
[          ](D) <-+
                  |
 +----------------+
 |
 +->(switch)[ type Q.921 ](term1)<---->(datalink)[ type Q.931 ]
            [ (no name)  ]                       [ (no name)  ]
.Ed
.Sh Netgraph Structures
Interesting members of the node and hook structures are shown below:
.Bd -literal
struct  ng_node {
  char    *name;                /* Optional globally unique name */
  void    *private;             /* Node implementation private info */
  struct  ng_type *type;        /* The type of this node */
  int     refs;                 /* Number of references to this struct */
  int     numhooks;             /* Number of connected hooks */
  hook_p  hooks;                /* Linked list of (connected) hooks */
};
typedef struct ng_node *node_p;

struct  ng_hook {
  char           *name;         /* This node's name for this hook */
  void           *private;      /* Node implementation private info */
  int            refs;          /* Number of references to this struct */
  struct ng_node *node;         /* The node this hook is attached to */
  struct ng_hook *peer;         /* The other hook in this connected pair */
  struct ng_hook *next;         /* Next in list of hooks for this node */
};
typedef struct ng_hook *hook_p;
.Ed
.Pp
The maintenance of the name pointers, reference counts, and linked list
of hooks for each node is handled automatically by the
.Nm
subsystem.
Typically a node's private info contains a back-pointer to the node or hook
structure, which counts as a new reference that must be registered by
incrementing
.Dv "node->refs" .
.Pp
From a hook you can obtain the corresponding node, and from
a node the list of all active hooks.
.Pp
Node types are described by this structure:
.Bd -literal
struct ng_type {
  u_int32_t version;                    /* Must equal NG_VERSION */
  const  char *name;                    /* Unique type name */

  /* Module event handler */
  modeventhand_t  mod_event;            /* Handle load/unload (optional) */

  /* Constructor */
  int    (*constructor)(node_p *node);  /* Create a new node */

  /** Methods using the node **/
  int    (*rcvmsg)(node_p node,         /* Receive control message */
            struct ng_mesg *msg,                /* The message */
            const char *retaddr,                /* Return address */
            struct ng_mesg **resp);             /* Synchronous response */
  int    (*shutdown)(node_p node);      /* Shutdown this node */
  int    (*newhook)(node_p node,        /* create a new hook */
            hook_p hook,                        /* Pre-allocated struct */
            const char *name);                  /* Name for new hook */

  /** Methods using the hook **/
  int    (*connect)(hook_p hook);       /* Confirm new hook attachment */
  int    (*rcvdata)(hook_p hook,        /* Receive data on a hook */
            struct mbuf *m,                     /* The data in an mbuf */
            meta_p meta);                       /* Meta-data, if any */
  int    (*disconnect)(hook_p hook);    /* Notify disconnection of hook */
};
.Ed
.Pp
Control messages have the following structure:
.Bd -literal
#define NG_CMDSTRLEN    15      /* Max command string (16 with null) */

struct ng_mesg {
  struct ng_msghdr {
    u_char      version;        /* Must equal NG_VERSION */
    u_char      spare;          /* Pad to 2 bytes */
    u_short     arglen;         /* Length of cmd/resp data */
    u_long      flags;          /* Message status flags */
    u_long      token;          /* Reply should have the same token */
    u_long      typecookie;     /* Node type understanding this message */
    u_long      cmd;            /* Command identifier */
    u_char      cmdstr[NG_CMDSTRLEN+1]; /* Cmd string (for debug) */
  } header;
  char  data[0];                /* Start of cmd/resp data */
};

#define NG_VERSION      1               /* Netgraph version */
#define NGF_ORIG        0x0000          /* Command */
#define NGF_RESP        0x0001          /* Response */
.Ed
.Pp
Control messages have the fixed header shown above, followed by a 
variable length data section which depends on the type cookie
and the command. Each field is explained below:
.Bl -tag -width xxx
.It Dv version
Indicates the version of netgraph itself. The current version is
.Dv NG_VERSION .
.It Dv arglen
This is the length of any extra arguments, which begin at
.Dv data .
.It Dv flags
Indicates whether this is a command or a response control message.
.It Dv token
The
.Dv token
is a means by which a sender can match a reply message to the
corresponding command message; the reply always has the same token.
.Pp
.It Dv typecookie
The corresponding node type's unique 32-bit value.
If a node doesn't recognize the type cookie it must reject the message
by returning
.Er EINVAL .
.Pp
Each type should have an include file that defines the commands,
argument format, and cookie for its own messages.
The typecookie
insures that the same header file was included by both sender and
receiver; when an incompatible change in the header file is made,
the typecookie
.Em must
be changed.
The de facto method for generating unique type cookies is to take the
seconds from the epoch at the time the header file is written
(i.e., the output of
.Dv "date -u +'%s'" ")."
.Pp
There is a predefined typecookie
.Dv NGM_GENERIC_COOKIE
for the ``generic'' node type, and
a corresponding set of generic messages which all nodes understand.
The handling of these messages is automatic.
.It Dv command
The identifier for the message command. This is type specific,
and is defined in the same header file as the typecookie.
.It Dv cmdstr
Room for a short human readable version of ``command'' (for debugging
purposes only).
.El
.Pp
Some modules may choose to implement messages from more than one 
of the header files and thus recognize more than one type cookie. 
.Sh Generic Control Messages
There are a number of standard predefined messages that will work
for any node, as they are supported directly by the framework itself.
These are defined in
.Dv ng_message.h
along with the basic layout of messages and other similar information.
.Bl -tag -width xxx
.It Dv NGM_CONNECT
Connect to another node, using the supplied hook names on either end.
.It Dv NGM_MKPEER
Construct a node of the given type and then connect to it using the
supplied hook names.
.It Dv NGM_SHUTDOWN
The target node should disconnect from all its neighbours and shut down.
Persistent nodes such as those representing physical hardware
might not disappear from the node namespace, but only reset themselves.
The node must disconnect all of its hooks.
This may result in neighbors shutting themselves down, and possibly a
cascading shutdown of the entire connected graph.
.It Dv NGM_NAME
Assign a name to a node. Nodes can exist without having a name, and this
is the default for nodes created using the
.Dv NGM_MKPEER
method. Such nodes can only be addressed relatively or by their ID number.
.It Dv NGM_RMHOOK
Ask the node to break a hook connection to one of its neighbours.
Both nodes will have their ``disconnect'' method invoked.
Either node may elect to totally shut down as a result.
.It Dv NGM_NODEINFO
Asks the target node to describe itself. The four returned fields
are the node name (if named), the node type, the node ID and the
number of hooks attached. The ID is an internal number unique to that node.
.It Dv NGM_LISTHOOKS
This returns the information given by
.Dv NGM_NODEINFO ,
but in addition 
includes an array of fields describing each link, and the description for
the node at the far end of that link.
.It Dv NGM_LISTNAMES
This returns an array of node descriptions (as for
.Dv NGM_NODEINFO ")"
where each entry of the array describes a named node.
All named nodes will be described.
.It Dv NGM_LISTNODES
This is the same as
.Dv NGM_LISTNAMES
except that all nodes are listed regardless of whether they have a name or not.
.It Dv NGM_LISTTYPES
This returns a list of all currently installed netgraph types.
.It Dv NGM_TEXT_STATUS
The node may return a text formatted status message.
The status information is determined entirely by the node type.
It is the only "generic" message
that requires any support within the node itself and as such the node may
elect to not support this message. The text response must be less than
.Dv NG_TEXTRESPONSE
bytes in length (presently 1024). This can be used to return general
status information in human readable form.
.El
.Sh Metadata
Data moving through the
.Nm
system can be accompanied by meta-data that describes some
aspect of that data. The form of the meta-data is a fixed header,
which contains enough information for most uses, and can optionally 
be supplemented by trailing
.Em option
structures, which contain a 
.Em cookie
(see the section on control messages), an identifier, a length and optional
data. If a node does not recognize the cookie associated with an option,
it should ignore that option.
.Pp
Meta data might include such things as priority, discard eligibility,
or special processing requirements. It might also mark a packet for
debug status, etc. The use of meta-data is still experimental.
.Sh INITIALIZATION
The base
.Nm
code may either be statically compiled
into the kernel or else loaded dynamically as a KLD via
.Xr kldload 8 .
In the former case, include
.Bd -literal -offset 4n -compact

   options NETGRAPH

.Ed
in your kernel configuration file. You may also include selected
node types in the kernel compilation, for example:
.Bd -literal -offset 4n -compact

   options NETGRAPH
   options NETGRAPH_SOCKET
   options NETGRAPH_ECHO

.Ed
.Pp
Once the
.Nm
subsystem is loaded, individual node types may be loaded at any time
as KLD modules via
.Xr kldload 8 .
Moreover,
.Nm
knows how to automatically do this; when a request to create a new
node of unknown type
.Em type
is made,
.Nm
will attempt to load the KLD module
.Dv ng_type.ko .
.Pp
Types can also be installed at boot time, as certain device drivers
may want to export each instance of the device as a netgraph node.
.Pp
In general, new types can be installed at any time from within the
kernel by calling
.Dv ng_newtype() ,
supplying a pointer to the type's
.Dv struct ng_type
structure.
.Pp
The
.Dv "NETGRAPH_INIT()"
macro automates this process by using a linker set.
.Sh EXISTING NODE TYPES
Several node types currently exist. Each is fully documented
in its own man page:
.Bl -tag -width xxx
.It SOCKET
The socket type implements two new sockets in the new protocol domain
.Dv PF_NETGRAPH .
The new sockets protocols are
.Dv NG_DATA
and
.Dv NG_CONTROL ,
both of type
.Dv SOCK_DGRAM .
Typically one of each is associated with a socket node.
When both sockets have closed, the node will shut down. The
.Dv NG_DATA
socket is used for sending and receiving data, while the
.Dv NG_CONTROL
socket is used for sending and receiving control messages.
Data and control messages are passed using the
.Xr sendto 2
and
.Xr recvfrom 2
calls, using a
.Dv struct sockaddr_ng
socket address.
.Pp
.It HOLE
Responds only to generic messages and is a ``black hole'' for data,
Useful for testing. Always accepts new hooks.
.Pp
.It ECHO
Responds only to generic messages and always echoes data back through the
hook from which it arrived. Returns any non generic messages as their
own response. Useful for testing.  Always accepts new hooks.
.Pp
.It TEE
This node is useful for ``snooping.'' It has 4 hooks:
.Dv left ,
.Dv right ,
.Dv left2right ,
and
.Dv right2left .
Data entering from the right is passed to the left and duplicated on
.Dv right2left,
and data entering from the left is passed to the right and
duplicated on
.Dv left2right .
Data entering from
.Dv left2right
is sent to the right and data from
.Dv right2left
to left. 
.Pp
.It RFC1490 MUX
Encapsulates/de-encapsulates frames encoded according to RFC 1490.
Has a hook for the encapsulated packets (``downstream'') and one hook
for each protocol (i.e., IP, PPP, etc.).
.Pp
.It FRAME RELAY MUX
Encapsulates/de-encapsulates Frame Relay frames.
Has a hook for the encapsulated packets (``downstream'') and one hook
for each DLCI.
.Pp
.It FRAME RELAY LMI
Automatically handles frame relay
``LMI'' (link management interface) operations and packets.
Automatically probes and detects which of several LMI standards
is in use at the exchange.
.Pp
.It TTY
This node is also a line discipline. It simply converts between mbuf
frames and sequential serial data, allowing a tty to appear as a netgraph
node. It has a programmable ``hotkey'' character.
.Pp
.It ASYNC
This node encapsulates and de-encapsulates asynchronous frames
according to RFC 1662. This is used in conjunction with the TTY node
type for supporting PPP links over asynchronous serial lines.
.Pp
.It INTERFACE
This node is also a system networking interface. It has hooks representing
each protocol family (IP, AppleTalk, IPX, etc.) and appears in the output of
.Xr ifconfig 8 .
The interfaces are named
.Em ng0 ,
.Em ng1 ,
etc.
.El
.Sh NOTES
Whether a named node exists can be checked by trying to send a control message
to it (e.g.,
.Dv NGM_NODEINFO
).
If it does not exist,
.Er ENOENT
will be returned.
.Pp
All data messages are mbuf chains with the M_PKTHDR flag set.
.Pp
Nodes are responsible for freeing what they allocate.
There are three exceptions:
.Bl -tag -width xxxx
.It 1
Mbufs sent across a data link are never to be freed by the sender. 
.It 2
Any meta-data information traveling with the data has the same restriction.
It might be freed by any node the data passes through, and a
.Dv NULL
passed onwards, but the caller will never free it.
Two macros
.Dv "NG_FREE_META(meta)"
and
.Dv "NG_FREE_DATA(m, meta)"
should be used if possible to free data and meta data (see
.Dv netgraph.h ")."
.It 3
Messages sent using
.Dv ng_send_message()
are freed by the callee. As in the case above, the addresses
associated with the message are freed by whatever allocated them so the 
recipient should copy them if it wants to keep that information.
.El
.Sh FILES
.Bl -tag -width xxxxx -compact
.It Pa /sys/netgraph/netgraph.h
Definitions for use solely within the kernel by
.Nm
nodes.
.It Pa /sys/netgraph/ng_message.h
Definitions needed by any file that needs to deal with 
.Nm 
messages.
.It Pa /sys/netgraph/ng_socket.h
Definitions needed to use 
.Nm
socket type nodes.
.It Pa /sys/netgraph/ng_{type}.h
Definitions needed to use 
.Nm
{type}
nodes, including the type cookie definition.
.It Pa /modules/netgraph.ko
Netgraph subsystem loadable KLD module.
.It Pa /modules/ng_{type}.ko
Loadable KLD module for node type {type}.
.El
.Sh USER MODE SUPPORT
There is a library for supporting user-mode programs that wish
to interact with the netgraph system. See
.Xr netgraph 3
for details.
.Pp
Two user-mode support programs,
.Xr ngctl 8
and
.Xr nghook 8 ,
are available to assist manual configuration and debugging.
.Pp
There are a few useful techniques for debugging new node types.
First, implementing new node types in user-mode first
makes debugging easier.
The
.Em tee
node type is also useful for debugging, especially in conjunction with
.Xr ngctl 8
and
.Xr nghook 8 .
.Sh SEE ALSO
.Xr socket 2 ,
.Xr netgraph 3 ,
.Xr ngctl 8 ,
.Xr nghook 8 ,
.Xr ng_async 8 ,
.Xr ng_cisco 8 ,
.Xr ng_echo 8 ,
.Xr ng_frame_relay 8 ,
.Xr ng_hole 8 ,
.Xr ng_iface 8 ,
.Xr ng_ksocket 8 ,
.Xr ng_lmi 8 ,
.Xr ng_ppp 8 ,
.Xr ng_pppoe 8 ,
.Xr ng_rfc1490 8 ,
.Xr ng_socket 8 ,
.Xr ng_tee 8 ,
.Xr ng_tty 8 ,
.Xr ng_UI 8 ,
.Xr ng_vjc 8 ,
.Xr ng_{type} 8 .
.Sh HISTORY
The
.Nm
system was designed and first implemented at Whistle Communications, Inc.
in a version FreeBSD 2.2 customized for the Whistle InterJet.
.Sh AUTHORS
Julian Elischer <julian@whistle.com>, with contributions by
Archie Cobbs <archie@whistle.com>.