net-device.h

Go to the documentation of this file.

/*
 * Copyright (c) 2005,2006 INRIA
 *
 * SPDX-License-Identifier: GPL-2.0-only
 *
 * Author: Mathieu Lacage <mathieu.lacage@sophia.inria.fr>
 * Modified by Emmanuelle Laprise to remove dependence on LLC headers
 */
#ifndef NET_DEVICE_H
#define NET_DEVICE_H
 
#include "address.h"
#include "packet.h"
 
#include "ns3/callback.h"
#include "ns3/ipv4-address.h"
#include "ns3/ipv6-address.h"
#include "ns3/object.h"
#include "ns3/ptr.h"
 
#include <stdint.h>
 
namespace ns3
{
 
class Node;
class Channel;
 
/**
 * \ingroup network
 * \defgroup netdevice Network Device
 */
 
/**
 * \ingroup netdevice
 *
 * \brief Network layer to device interface
 *
 * This interface defines the API which the IP and ARP
 * layers need to access to manage an instance of a network device
 * layer. It currently does not support MAC-level
 * multicast but this should not be too hard to add by adding
 * extra methods to register MAC multicast addresses to
 * filter out unwanted packets before handing them to the
 * higher layers.
 *
 * In Linux, this interface is analogous to the interface
 * just above dev_queue_xmit() (i.e., IP packet is fully
 * constructed with destination MAC address already selected).
 *
 * If you want to write a new MAC layer, you need to subclass
 * this base class and implement your own version of the
 * pure virtual methods in this class.
 *
 * This class was designed to hide as many MAC-level details as
 * possible from the perspective of layer 3 to allow a single layer 3
 * to work with any kind of MAC layer. Specifically, this class
 * encapsulates the specific format of MAC addresses used by a
 * device such that the layer 3 does not need any modification
 * to handle new address formats. This means obviously that the
 * NetDevice class must know about the address format of all potential
 * layer 3 protocols through its GetMulticast methods: the current
 * API has been optimized to make it easy to add new MAC protocols,
 * not to add new layer 3 protocols.
 *
 * Devices aiming to support flow control and dynamic queue limits must perform
 * the following operations:
 *   - in the NotifyNewAggregate method
 *     + cache the pointer to the netdevice queue interface aggregated to the
 *       device
 *     + set the select queue callback through the netdevice queue interface,
 *       if the device is multi-queue
 *   - anytime before initialization
 *     + set the number of device transmission queues (and optionally create them)
 *       through the netdevice queue interface, if the device is multi-queue
 *   - when the device queues have been created, invoke
 *     NetDeviceQueueInterface::ConnectQueueTraces, which
 *     + connects the Enqueue traced callback of the device queues to the
 *       PacketEnqueued static method of the NetDeviceQueue class
 *     + connects the Dequeue and DropAfterDequeue traced callback of the device
 *       queues to the PacketDequeued static method of the NetDeviceQueue
 *       class
 *     + connects the DropBeforeEnqueue traced callback of the device queues to
 *       the PacketDiscarded static method of the NetDeviceQueue class
 */
class NetDevice : public Object
{
  public:
    /**
     * \brief Get the type ID.
     * \return the object TypeId
     */
    static TypeId GetTypeId();
    ~NetDevice() override;
 
    /**
     * \param index ifIndex of the device
     */
    virtual void SetIfIndex(const uint32_t index) = 0;
    /**
     * \return index ifIndex of the device
     */
    virtual uint32_t GetIfIndex() const = 0;
 
    /**
     * \return the channel this NetDevice is connected to. The value
     *         returned can be zero if the NetDevice is not yet connected
     *         to any channel or if the underlying NetDevice has no
     *         concept of a channel. i.e., callers _must_ check for zero
     *         and be ready to handle it.
     */
    virtual Ptr<Channel> GetChannel() const = 0;
 
    /**
     * Set the address of this interface
     * \param address address to set
     */
    virtual void SetAddress(Address address) = 0;
 
    /**
     * \return the current Address of this interface.
     */
    virtual Address GetAddress() const = 0;
 
    /**
     * \param mtu MTU value, in bytes, to set for the device
     * \return whether the MTU value was within legal bounds
     *
     * Override for default MTU defined on a per-type basis.
     */
    virtual bool SetMtu(const uint16_t mtu) = 0;
    /**
     * \return the link-level MTU in bytes for this interface.
     *
     * This value is typically used by the IP layer to perform
     * IP fragmentation when needed.
     */
    virtual uint16_t GetMtu() const = 0;
    /**
     * \return true if link is up; false otherwise
     */
    virtual bool IsLinkUp() const = 0;
    /**
     * TracedCallback signature for link changed event.
     */
    typedef void (*LinkChangeTracedCallback)();
    /**
     * \param callback the callback to invoke
     *
     * Add a callback invoked whenever the link
     * status changes to UP. This callback is typically used
     * by the IP/ARP layer to flush the ARP cache and by IPv6 stack
     * to flush NDISC cache whenever the link goes up.
     */
    virtual void AddLinkChangeCallback(Callback<void> callback) = 0;
    /**
     * \return true if this interface supports a broadcast address,
     *         false otherwise.
     */
    virtual bool IsBroadcast() const = 0;
    /**
     * \return the broadcast address supported by
     *         this netdevice.
     *
     * Calling this method is invalid if IsBroadcast returns
     * not true.
     */
    virtual Address GetBroadcast() const = 0;
 
    /**
     * \return value of m_isMulticast flag
     */
    virtual bool IsMulticast() const = 0;
 
    /**
     * \brief Make and return a MAC multicast address using the provided
     *        multicast group
     *
     * \RFC{1112} says that an Ipv4 host group address is mapped to an Ethernet
     * multicast address by placing the low-order 23-bits of the IP address into
     * the low-order 23 bits of the Ethernet multicast address
     * 01-00-5E-00-00-00 (hex).  Similar RFCs exist for Ipv6 and Eui64 mappings.
     * This method performs the multicast address creation function appropriate
     * to the underlying MAC address of the device.  This MAC address is
     * encapsulated in an abstract Address to avoid dependencies on the exact
     * MAC address format.
     *
     * In the case of net devices that do not support
     * multicast, clients are expected to test NetDevice::IsMulticast and avoid
     * attempting to map multicast packets.  Subclasses of NetDevice that do
     * support multicasting are expected to override this method and provide an
     * implementation appropriate to the particular device.
     *
     * \param multicastGroup The IP address for the multicast group destination
     * of the packet.
     * \return The MAC multicast Address used to send packets to the provided
     * multicast group.
     *
     * \warning Calling this method is invalid if IsMulticast returns not true.
     * \see IsMulticast()
     */
    virtual Address GetMulticast(Ipv4Address multicastGroup) const = 0;
 
    /**
     * \brief Get the MAC multicast address corresponding
     * to the IPv6 address provided.
     * \param addr IPv6 address
     * \return the MAC multicast address
     * \warning Calling this method is invalid if IsMulticast returns not true.
     */
    virtual Address GetMulticast(Ipv6Address addr) const = 0;
 
    /**
     * \brief Return true if the net device is acting as a bridge.
     *
     * \return value of m_isBridge flag
     */
    virtual bool IsBridge() const = 0;
 
    /**
     * \brief Return true if the net device is on a point-to-point link.
     *
     * \return value of m_isPointToPoint flag
     */
    virtual bool IsPointToPoint() const = 0;
    /**
     * \param packet packet sent from above down to Network Device
     * \param dest mac address of the destination (already resolved)
     * \param protocolNumber identifies the type of payload contained in
     *        this packet. Used to call the right L3Protocol when the packet
     *        is received.
     *
     *  Called from higher layer to send packet into Network Device
     *  to the specified destination Address
     *
     * \return whether the Send operation succeeded
     */
    virtual bool Send(Ptr<Packet> packet, const Address& dest, uint16_t protocolNumber) = 0;
    /**
     * \param packet packet sent from above down to Network Device
     * \param source source mac address (so called "MAC spoofing")
     * \param dest mac address of the destination (already resolved)
     * \param protocolNumber identifies the type of payload contained in
     *        this packet. Used to call the right L3Protocol when the packet
     *        is received.
     *
     *  Called from higher layer to send packet into Network Device
     *  with the specified source and destination Addresses.
     *
     * \return whether the Send operation succeeded
     */
    virtual bool SendFrom(Ptr<Packet> packet,
                          const Address& source,
                          const Address& dest,
                          uint16_t protocolNumber) = 0;
    /**
     * \returns the node base class which contains this network
     *          interface.
     *
     * When a subclass needs to get access to the underlying node
     * base class to print the nodeid for example, it can invoke
     * this method.
     */
    virtual Ptr<Node> GetNode() const = 0;
 
    /**
     * \param node the node associated to this netdevice.
     *
     * This method is called from ns3::Node::AddDevice.
     */
    virtual void SetNode(Ptr<Node> node) = 0;
 
    /**
     * \returns true if ARP is needed, false otherwise.
     *
     * Called by higher-layers to check if this NetDevice requires
     * ARP to be used.
     */
    virtual bool NeedsArp() const = 0;
 
    /**
     * Packet types are used as they are in Linux.  GCC name resolution on
     * typedef enum {} PacketType is broken for the foreseeable future, so
     * if you need to use ns-3 PacketType in a driver that also uses the
     * Linux packet types you're hosed unless we define a shadow type,
     * which we do here.
     */
    enum PacketType
    {
        PACKET_HOST = 1, //!< Packet addressed to us
        NS3_PACKET_HOST = PACKET_HOST,
        PACKET_BROADCAST, //!< Packet addressed to all
        NS3_PACKET_BROADCAST = PACKET_BROADCAST,
        PACKET_MULTICAST, //!< Packet addressed to multicast group
        NS3_PACKET_MULTICAST = PACKET_MULTICAST,
        PACKET_OTHERHOST, //!< Packet addressed to someone else
        NS3_PACKET_OTHERHOST = PACKET_OTHERHOST,
    };
 
    /**
     * \param device a pointer to the net device which is calling this callback
     * \param packet the packet received
     * \param protocol the 16 bit protocol number associated with this packet.
     *        This protocol number is expected to be the same protocol number
     *        given to the Send method by the user on the sender side.
     * \param sender the address of the sender
     * \returns true if the callback could handle the packet successfully, false
     *          otherwise.
     */
    typedef Callback<bool, Ptr<NetDevice>, Ptr<const Packet>, uint16_t, const Address&>
        ReceiveCallback;
 
    /**
     * \param cb callback to invoke whenever a packet has been received and must
     *        be forwarded to the higher layers.
     *
     * Set the callback to be used to notify higher layers when a packet has been
     * received.
     */
    virtual void SetReceiveCallback(ReceiveCallback cb) = 0;
 
    /**
     * \param device a pointer to the net device which is calling this callback
     * \param packet the packet received
     * \param protocol the 16 bit protocol number associated with this packet.
     *        This protocol number is expected to be the same protocol number
     *        given to the Send method by the user on the sender side.
     * \param sender the address of the sender
     * \param receiver the address of the receiver
     * \param packetType type of packet received (broadcast/multicast/unicast/otherhost)
     * \returns true if the callback could handle the packet successfully, false
     *          otherwise.
     */
    typedef Callback<bool,
                     Ptr<NetDevice>,
                     Ptr<const Packet>,
                     uint16_t,
                     const Address&,
                     const Address&,
                     PacketType>
        PromiscReceiveCallback;
 
    /**
     * \param cb callback to invoke whenever a packet has been received in promiscuous mode and must
     *        be forwarded to the higher layers.
     *
     * Enables netdevice promiscuous mode and sets the callback that
     * will handle promiscuous mode packets.  Note, promiscuous mode
     * packets means _all_ packets, including those packets that can be
     * sensed by the netdevice but which are intended to be received by
     * other hosts.
     */
    virtual void SetPromiscReceiveCallback(PromiscReceiveCallback cb) = 0;
 
    /**
     * \return true if this interface supports a bridging mode, false otherwise.
     */
    virtual bool SupportsSendFrom() const = 0;
};
 
} // namespace ns3
 
#endif /* NET_DEVICE_H */