918 lines
30 KiB
C
918 lines
30 KiB
C
/* ipcom_netif.h - IPCOM network interface library */
|
|
|
|
/*
|
|
******************************************************************************
|
|
* INTERPEAK API HEADER FILE
|
|
*
|
|
* Document no: @(#) $Name: VXWORKS_ITER40_2019102207 $ $RCSfile: ipcom_netif.h,v $ $Revision: 1.102.10.1 $
|
|
* $Source: /home/interpeak/CVSRoot/ipcom/include/ipcom_netif.h,v $
|
|
* $Author: svc-cmnet $ $Date: 2013-03-13 07:45:14 $
|
|
* $State: Exp $ $Locker: $
|
|
*
|
|
* INTERPEAK_COPYRIGHT_STRING
|
|
* Design and implementation by Lennart Bang <lob@interpeak.se>
|
|
******************************************************************************
|
|
*/
|
|
#ifndef IPCOM_NETIF_H
|
|
#define IPCOM_NETIF_H
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 1 DESCRIPTION
|
|
****************************************************************************
|
|
*/
|
|
|
|
/*
|
|
DESCRIPTION
|
|
This library contains the API for the IPCOM network interface.
|
|
INCLUDE FILES: ipcom_type.h, ipcom_pkt.h, ipcom_sock.h, ipcom_auth.h
|
|
*/
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 2 CONFIGURATION
|
|
****************************************************************************
|
|
*/
|
|
|
|
#include "ipcom_config.h"
|
|
#include "ipcom_mib2.h"
|
|
|
|
/* Maximum size of link address (e.g. Ethernet MAC) in bytes. */
|
|
#define IP_LINKNAMSIZ 8
|
|
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 3 INCLUDE FILES
|
|
****************************************************************************
|
|
*/
|
|
|
|
#include <ipcom_auth.h>
|
|
#include <ipcom_pkt.h>
|
|
#include <ipcom_sock.h>
|
|
#include <ipcom_type.h>
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 4 DEFINES
|
|
****************************************************************************
|
|
*/
|
|
|
|
/*
|
|
*===========================================================================
|
|
* constants
|
|
*===========================================================================
|
|
*/
|
|
|
|
/* Interface types: */
|
|
#define IP_IFT_OTHER 0x1 /* None of the following: */
|
|
#define IP_IFT_ETHER 0x6 /* Ethernet CSMACD */
|
|
#define IP_IFT_PPP 0x17 /* RFC 1331 */
|
|
#define IP_IFT_LOOP 0x18 /* Loopback */
|
|
#define IP_IFT_IEEE80211 0x47 /* Radio spread spectrum */
|
|
#define IP_IFT_MPLS_TUNNEL 0x83 /* MPLS tunnel interface */
|
|
#define IP_IFT_L2VLAN 0x87 /* Layer 2 Virtual LAN using 802.1Q */
|
|
#define IP_IFT_PPPOE 0xf0 /* PPPoE */
|
|
#define IP_IFT_TUNNEL 0xf1 /* IP over IP tunnel */
|
|
#define IP_IFT_SOCKDEV 0xf2 /* Device that has a 1-to-1 mapping to a
|
|
socket in the IP_AF_SOCKDEV domain */
|
|
#define IP_IFT_SIT 0xf3
|
|
#define IP_IFT_IPGRE 0xf4
|
|
#define IP_IFT_TUNNEL6 0xf5
|
|
#define IP_IFT_LINK_STUB 0xf6 /* Custom Link layer */
|
|
#define IP_IFT_PPPL2TP 0xf7 /* L2TPv2 PPP interface */
|
|
#define IP_IFT_L2BRIDGE 0xf8 /* 802.1D Bridge Interfaces */
|
|
#define IP_IFT_L2BRIDGE_PV 0xf9 /* 802.1Q Port Based VLAN Bridge Interfaces (a PVID is specified)*/
|
|
#define IP_IFT_L2BRIDGE_TV 0xfa /* 802.1Q Tag Based VLAN Bridge Interfaces */
|
|
|
|
/* Interface default MTUs: */
|
|
#define IP_ETH_MTU 1500
|
|
#define IP_PPP_MTU 1500
|
|
|
|
/* Link header sizes: */
|
|
#define IP_ETH_HDR_SIZE 14
|
|
#define IP_PPP_HDR_SIZE (4 + 2) /* PPP header + 2 bytes FCS */
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Generic interface specific ioctls
|
|
*===========================================================================
|
|
* Generic non-public Interface IO control commands.
|
|
*/
|
|
|
|
/* Link layer and driver layer IO control commands */
|
|
#define IP_SIOCXOPEN IP_IO(NETIF, 100, R, struct Ip_ifreq) /* Open lower layer. argument not used. */
|
|
#define IP_SIOCXCLOSE IP_IO(NETIF, 101, R, struct Ip_ifreq) /* Close lower layer. argument not used. */
|
|
#define IP_SIOCXPROMISC IP_IOW(NETIF, 102, W, struct Ip_ifreq) /* Activate/deactivate promiscuous mode 'Ip_bool' */
|
|
#define IP_SIOCXDHCPUP IP_IOWR(NETIF, 103, W, struct Ip_ifreq) /* DHCP enabled on link. IPLITE ONLY! */
|
|
#define IP_SIOCXDHCPDOWN IP_IOWR(NETIF, 104, W, struct Ip_ifreq) /* DHCP disabled on link.IPLITE ONLY! */
|
|
#define IP_SIOCXCHGFLAGS IP_IOWR(NETIF, 105, W, struct Ip_ifreq) /* Notify the driver about network interface flag(s) change */
|
|
|
|
/* Link-layer join/leave multicast address (note: 'R' flagged cause ok to change while interface UP). */
|
|
#define IP_SIOCXADDMULTI_IN IP_IOW(INET , 100, R, struct Ip_in_addr) /* Join the corresponding link addr */
|
|
#define IP_SIOCXDELMULTI_IN IP_IOW(INET , 101, R, struct Ip_in_addr) /* Leave the corresponding link addr */
|
|
#define IP_SIOCXADDMULTI_IN6 IP_IOW(INET6, 100, R, struct Ip_in6_addr) /* Join the corresponding link addr */
|
|
#define IP_SIOCXDELMULTI_IN6 IP_IOW(INET6, 101, R, struct Ip_in6_addr) /* Leave the corresponding link addr */
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ethernet specific ioctls
|
|
*===========================================================================
|
|
* IO control commands specific to the Ethernet driver.
|
|
*/
|
|
|
|
/* Add and delete multicast group. argp is a pointer to Ethernet hardware addr. */
|
|
typedef struct Ip_hwaddr_struct { Ip_u8 hwaddr[6]; } Ip_hwaddr; /*!!*/
|
|
#define IP_SIOCXADDMULTI IP_IOW(ETH, 1, W, struct Ip_ethreq) /* Join a multicast group. */
|
|
#define IP_SIOCXDELMULTI IP_IOW(ETH, 2, W, struct Ip_ethreq) /* Leave a multicast group. */
|
|
|
|
/* Get/set IPv4 address/mask/gateway from/to driver. */
|
|
#define IP_SIOCXETHGINET IP_IOWR(ETH, 12, R, struct Ip_ethreq) /* Get IPv4 addr&mask&gateway. */
|
|
#define IP_SIOCXETHSINET IP_IOWR(ETH, 13, W, struct Ip_ethreq) /* Set IPv4 addr&mask&gateway. */
|
|
|
|
#define IP_SIOCXETHGINET6 IP_IOWR(ETH, 20, R, struct Ip_ethreq) /* Get IPv6 addr&prefixlen. */
|
|
|
|
/* Enable/disable RARP (IPNET only). */
|
|
#define IP_SIOCXETHSRARP IP_IOWR(ETH, 14, W, struct Ip_ethreq) /* Enable/disable RARP. */
|
|
|
|
#ifdef IPCOM_VXWORKS_USE_MUX_L2
|
|
/* proprietary ioctls: get vlan information from layer 2 */
|
|
#define IP_SIOCXETHGVLAN IP_IOWR(ETH, 50, R, struct Ip_ethreq)
|
|
|
|
/* proprietary ioctls: join a port to a vlan membership */
|
|
#define IP_SIOCXETHSVLAN IP_IOWR(ETH, 51, W, struct Ip_ethreq)
|
|
|
|
/* proprietary ioctls: enable a port for layer 2 vlan */
|
|
#define IP_SIOCXETHSPORT IP_IOWR(ETH, 52, W, struct Ip_ethreq)
|
|
#endif /* IPCOM_VXWORKS_USE_MUX_L2 */
|
|
|
|
/* proprietary ioctls: enable/disable vlan filtering */
|
|
#define IP_SIOCXETHSVLANFILTER IP_IOWR(ETH, 30, W, struct Ip_ethreq)
|
|
|
|
/* proprietary ioctls: set/clr VID filtering */
|
|
#define IP_SIOCXETHSVIDFILTER IP_IOWR(ETH, 31, W, struct Ip_ethreq)
|
|
#define IP_SIOCXETHCVIDFILTER IP_IOWR(ETH, 32, W, struct Ip_ethreq)
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* io events
|
|
*===========================================================================
|
|
* IO events are passed from driver -> link layer -> IPNET/IPLITE.
|
|
* data = IP_NULL on all current events.
|
|
*/
|
|
|
|
#define IP_EIOXUP 0x1001 /* Interface is up. */
|
|
#define IP_EIOXDOWN 0x1002 /* Interface is down. */
|
|
#define IP_EIOXRUNNING 0x1003 /* Driver is running. */
|
|
#define IP_EIOXSTOP 0x1004 /* Driver is stopped. */
|
|
#define IP_EIOXCHANGE 0x1005 /* Interface flags (other than up/running) changed. */
|
|
#define IP_EIOXNEWADDR 0x1006 /* Address being added to interface. */
|
|
#define IP_EIOXDELADDR 0x1007 /* Address being deleted to interface. */
|
|
#ifdef IPCOM_VXWORKS_USE_MUX_L2
|
|
#define IP_EIOXVLAN 0x1008 /* vlan egress frame type changes */
|
|
#endif /* IPCOM_VXWORKS_USE_MUX_L2 */
|
|
|
|
|
|
/* Driver can calculate and store the transport layer checksum (UDP and TCP)
|
|
on IPv4 packets while transmitting.
|
|
The the stack has calculated the pseudo-header checksum and stored it in
|
|
the transport layer checksum field. */
|
|
#define IPCOM_IF_DRV_CAP_IP4_TRANSPORT_CSUM_TX (1 << 0)
|
|
/* Driver can calculate and store the transport layer checksum (UDP and TCP)
|
|
on IPv6 packets while transmitting.
|
|
The the stack has calculated the pseudo-header checksum and stored it in
|
|
the transport layer checksum field. */
|
|
#define IPCOM_IF_DRV_CAP_IP6_TRANSPORT_CSUM_TX (1 << 1)
|
|
/* Driver can create IPv4 TCP segments that is small enough to fit the MTU
|
|
in hardware. TSO = Tcp Segmentation Offload */
|
|
#define IPCOM_IF_DRV_CAP_IP4_TSO (1 << 2)
|
|
/* Driver can create IPv6 TCP segments that is small enough to fit the MTU
|
|
in hardware. TSO = Tcp Segmentation Offload */
|
|
#define IPCOM_IF_DRV_CAP_IP6_TSO (1 << 3)
|
|
/* Driver can handle slightly larger than normal frames (e.g. frames with
|
|
vlan tag)
|
|
*/
|
|
#define IPCOM_IF_DRV_CAP_VLAN_MTU (1 << 4)
|
|
/* Driver can insert the vlan tag to a frame on trasnmit */
|
|
#define IPCOM_IF_DRV_CAP_VLAN_HWTAGGING_TX (1 << 5)
|
|
/* Driver can strip the vlan tag from the frame on receive */
|
|
#define IPCOM_IF_DRV_CAP_VLAN_HWTAGGING_RX (1 << 6)
|
|
|
|
/* Interface supports MACsec SecY */
|
|
#define IPCOM_IF_DRV_CAP_MACSEC_SECY (1 << 7)
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 5 PORT OVERRIDE INCLUDE
|
|
****************************************************************************
|
|
*/
|
|
|
|
#if defined(IPCOM_USE_NETIF) && IPCOM_USE_NETIF == IPCOM_NETIF_NATIVE
|
|
#include "ipcom_pnetif.h"
|
|
#endif
|
|
|
|
#include "ipcom_arphrd.h"
|
|
|
|
/*
|
|
****************************************************************************p
|
|
* 6 TYPES
|
|
****************************************************************************
|
|
*/
|
|
|
|
/* Forward declaration */
|
|
struct Ipcom_netif_struct;
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_eth_inet
|
|
*===========================================================================
|
|
* ioctl structure for IP_SIOCXETHGINET option.
|
|
*/
|
|
typedef struct Ipcom_eth_inet_struct
|
|
{
|
|
struct Ip_in_addr addr; /* IPv4 address in network endian. */
|
|
struct Ip_in_addr mask; /* IPv4 mask in network endian. */
|
|
struct Ip_in_addr gateway; /* IPv4 gateway in network endian (0 if none). */
|
|
}
|
|
Ipcom_eth_inet;
|
|
|
|
#ifdef IPCOM_USE_INET6
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_eth_inet6
|
|
*===========================================================================
|
|
* ioctl structure for IP_SIOCXETHGINET6 option.
|
|
*/
|
|
typedef struct Ipcom_eth_inet6_struct
|
|
{
|
|
struct Ip_in6_addr addr; /* IPv6 address in network endian. */
|
|
Ip_u8 prefixlen; /* IPv6 prefixlen */
|
|
}
|
|
Ipcom_eth_inet6;
|
|
#endif /* IPCOM_USE_INET6 */
|
|
|
|
#ifdef IPCOM_VXWORKS_USE_MUX_L2
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_eth_vlan
|
|
*===========================================================================
|
|
* ioctl structure for IP_SIOCXETHGVLAN, IP_SIOCXETHSVLAN and
|
|
* IP_SIOCXETHSPORT options used exclusively by vlan pseudo-interface
|
|
* to communicate with layer 2.
|
|
*/
|
|
typedef struct Ipcom_eth_vlan_struct
|
|
{
|
|
Ip_u16 vid; /* vlan id */
|
|
Ip_bool tagged; /* send vlan tagged frame */
|
|
Ip_bool is_member; /* port is a member of the vlan */
|
|
}
|
|
Ipcom_eth_vlan;
|
|
#endif /* IPCOM_VXWORKS_USE_MUX_L2 */
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ip_ethreq
|
|
*===========================================================================
|
|
* Interface request structure used for Ethernet ioctl's. All interface
|
|
* ioctl's must have parameter definitions which begin with ifr_name and
|
|
* link_index. The remainder may be option specific.
|
|
*/
|
|
struct Ip_ethreq
|
|
{
|
|
char ethr_name[IP_IFNAMSIZ]; /* if name, e.g. "eth0" */
|
|
|
|
union
|
|
{
|
|
int rarp; /* 0 = disable else enable. */
|
|
Ipcom_eth_inet inet; /* IPv4 address & mask. */
|
|
#ifdef IPCOM_USE_INET6
|
|
Ipcom_eth_inet6 inet6; /* IPv6 address & prefixlen. */
|
|
#endif
|
|
#ifdef IPCOM_VXWORKS_USE_MUX_L2
|
|
Ipcom_eth_vlan vlan;
|
|
#endif
|
|
}
|
|
ethru;
|
|
};
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_eth_vlanreq
|
|
*===========================================================================
|
|
* Layer 2 link layer vlan notification structure used for IP_EIOXVLAN io
|
|
* event
|
|
*/
|
|
#ifdef IP_EIOXVLAN
|
|
typedef struct Ipcom_eth_vlanreq
|
|
{
|
|
Ip_u16 vid; /* VLAN identifier */
|
|
Ip_u8 pri; /* 3-bit VLAN priority, for port-based only */
|
|
|
|
Ip_bool vid_is_pvid; /* this is a port-based vlan */
|
|
Ip_bool vlan_tag_required; /* link layer indicates vlan tag is required */
|
|
}
|
|
Ipcom_eth_vlanreq;
|
|
#endif /* IP_EIOXVLAN */
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_link_ioevent - communicate IO events from the driver to link layer
|
|
*
|
|
* Communicates IO events from the driver to the link layer. The driver should
|
|
* use this function hook to signal events to the link layer. The supported
|
|
* events are:
|
|
*
|
|
* \ts
|
|
* IP_EIOXUP |interface is up |
|
|
* IP_EIOXDOWN |interface is down |
|
|
* IP_EIOXRUNNING |driver is running |
|
|
* IP_EIOXSTOP |driver is stopped |
|
|
* IP_EIOXVLAN | vlan egress frame type changes |
|
|
* \te
|
|
*
|
|
* Note: data is 'IP_NULL' on all current events.
|
|
*
|
|
* RETURNS: Returns 0 on success, -1 otherwise.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
typedef int (
|
|
*Ipcom_link_ioevent
|
|
)
|
|
(
|
|
struct Ipcom_netif_struct *netif,
|
|
int command,
|
|
void *data,
|
|
int fcflags
|
|
);
|
|
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_link_tap - call a driver-specific 'rx/tx' tap handler
|
|
*
|
|
* This routine calls a driver-specific 'rx/tx' tap handler. The tap handler
|
|
* makes a copy of the packet that can be used for packet sniffing or other
|
|
* purposes.
|
|
*
|
|
* RETURNS: \&N/A
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
typedef void (
|
|
*Ipcom_link_tap
|
|
)
|
|
(
|
|
struct Ipcom_netif_struct *netif,
|
|
Ipcom_pkt *pkt,
|
|
int pcap_dir,
|
|
Ip_u16 frame_type,
|
|
int mac,
|
|
int payload
|
|
);
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_drv_output
|
|
*===========================================================================
|
|
* Description: Output the packet to the driver.
|
|
* Parameters: netif - A pointer to the network interface for this driver.
|
|
* pkt - The packet to send out on the link. pkt->start is the
|
|
* offset to the link layer header.
|
|
* Returns: 0 = success, < 0 = error code (-IP_ERRNO_xxx)
|
|
*
|
|
*/
|
|
typedef int (*Ipcom_drv_output)(struct Ipcom_netif_struct *netif, Ipcom_pkt *pkt);
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_drv_ioctl
|
|
*===========================================================================
|
|
* Description: IO control for the driver.
|
|
* Parameters: netif - A pointer to the network interface structure for this
|
|
* driver.
|
|
* commmand - The IP_SIOCXxxx to perform (see above).
|
|
* data - Data for to command.
|
|
* Returns: 0 = success, < 0 = error code (-IP_ERRNO_xxx)
|
|
*
|
|
*/
|
|
typedef int (*Ipcom_drv_ioctl)(struct Ipcom_netif_struct *netif, Ip_u32 command, void *data);
|
|
|
|
|
|
|
|
#ifdef IPCOM_USE_MUX_DRV
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_drv_ip_resolve
|
|
*===========================================================================
|
|
* Description: To resolve an ipv4 address to link address
|
|
* Parameters: mux_cookie: pointer to END object
|
|
* m_blk: IPv4 packet which needs to be sent
|
|
* ip4_addr: the destination ipv4 address which needs to be
|
|
* resolved (in a sockaddr_in struct).
|
|
* pkt_type: 0 = UNICAST,1 = MULTICAST, 2 = BROADCAST
|
|
* frame_type: network service type ( ipv4 or ipv6)
|
|
* neigh: pointer to IPNET route entry ( can be used
|
|
* to cache already resolved addresses to avoid duplicate lookup)
|
|
* Returns: 0 = immediate success with result in dst_m_blkid,
|
|
* < 0 = Delayed resolution with ownership of m_blk taken
|
|
* by the driver.
|
|
*
|
|
*/
|
|
#ifdef IPCOM_USE_INET
|
|
typedef int (*Ipcom_drv_ip_resolve)(void *mux_cookie,void *m_blk,
|
|
struct Ip_sockaddr_in *ip4_addr,void *dst_m_blkid,
|
|
int pkt_type, int frame_type, void *neigh);
|
|
#endif
|
|
|
|
#ifdef IPCOM_USE_INET6
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_drv_ip6_resolve
|
|
*===========================================================================
|
|
* Description: To resolve an ipv4 address to link address
|
|
* Parameters: mux_cookie: pointer to END object
|
|
* m_blk: IPv6 packet which needs to be sent
|
|
* ip6_addr: the destination ipv6 address which needs to be
|
|
* resolved (in a sockaddr_in6 struct).
|
|
* pkt_type: 0 = UNICAST,1 = MULTICAST, 2 = BROADCAST
|
|
* frame_type: network service type ( ipv4 or ipv6)
|
|
* neigh: pointer to IPNET route entry ( can be used
|
|
* to cache already resolved addresses to avoid duplicate lookup)
|
|
* Returns: 0 = immediate success with result in dst_m_blkid,
|
|
* < 0 = Delayed resolution with ownership of m_blk taken
|
|
* by the driver.
|
|
*
|
|
*/
|
|
typedef int (*Ipcom_drv_ip6_resolve)(void *mux_cookie,void *m_blk,
|
|
struct Ip_sockaddr_in6 *ip6_addr,void *dst_m_blkid,
|
|
int pkt_type, int frame_type, void *neigh);
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_drv_ip6_create_addre
|
|
*===========================================================================
|
|
* Description: To create an ipv6 address from the link addr and
|
|
* network prefix.
|
|
* Parameters: prefix: ipv6 network prefix in a struct Ip_in6_addr
|
|
* ip6_addr: returns the created ipv6 address in a struct
|
|
* Ip_in6_addr
|
|
* Returns: void
|
|
*/
|
|
typedef void (*Ipcom_drv_ip6_create_addr)(void *prefix, void *ip6_addr);
|
|
|
|
#endif /* IPCOM_USE_INET6 */
|
|
#endif /* IPCOM_USE_MUX_DRV */
|
|
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* Ipcom_netif
|
|
*===========================================================================
|
|
*/
|
|
typedef struct Ipcom_netif_struct
|
|
{
|
|
char name[IP_IFNAMSIZ]; /* interface name (e.g. eth0, ppp1) */
|
|
Ip_u32 flags; /* IP_IFF_xxx flags */
|
|
Ip_u32 ifindex; /* Interface index. */
|
|
Ip_u32 link_index; /* Link index for MPPP. Currently 0. */
|
|
Ip_u32 mtu; /* The interface maximum transfer unit */
|
|
Ip_u32 metric; /* The interface metric */
|
|
Ip_u16 type; /* Interface type, one of IP_IFT_XXX */
|
|
Ip_u16 arphrd_type;
|
|
Ipcom_atomic_t ref_count; /* Reference counter */
|
|
|
|
|
|
/****** LINK VARIABLES ******/
|
|
void *plink; /* Link port data, e.g. PPP specific info. */
|
|
Ip_u16 link_hdr_size; /* Size of the link header in bytes */
|
|
Ip_u16 link_addr_size; /* Size of the link address in bytes */
|
|
Ip_u8 link_addr[IP_LINKNAMSIZ];
|
|
|
|
/* Link function called by driver layer */
|
|
Ipcom_link_ioevent link_ioevent;
|
|
|
|
/* Link tap function called by link layer */
|
|
Ipcom_link_tap link_tap;
|
|
|
|
/****** DRIVER VARIABLES ******/
|
|
void *pdrv; /* Driver port data, IPCOM/IPNET/IPLITE does not use this info. */
|
|
|
|
/* Driver functions called by link layer */
|
|
Ipcom_drv_output drv_output;
|
|
Ipcom_drv_ioctl drv_ioctl;
|
|
Ip_u32 drv_capabilities; /* IPCOM_IF_DRV_CAP_xxx */
|
|
|
|
#ifdef IPCOM_USE_MUX_DRV
|
|
#ifdef IPCOM_USE_INET
|
|
Ipcom_drv_ip_resolve drv_ip_resolve;
|
|
#endif
|
|
#ifdef IPCOM_USE_INET6
|
|
Ipcom_drv_ip6_resolve drv_ip6_resolve;
|
|
Ipcom_drv_ip6_create_addr drv_ip6_create_addr;
|
|
#endif
|
|
#endif /* IPCOM_USE_MUX_DRV */
|
|
|
|
#ifdef IPCOM_USE_MIB2
|
|
Ipcom_mib2_stats_netif mib2;
|
|
#endif
|
|
|
|
#ifdef IPWLAN
|
|
int (*wlan_ioctl)(void *, Ip_u32, void *);
|
|
Ip_u8 macaddr[6];
|
|
#endif
|
|
|
|
#ifdef IPCOM_USE_FORWARDER
|
|
int hw_ix; /* forwarder port index to use for outgoing traffic */
|
|
void *hw_priv;
|
|
#endif
|
|
}
|
|
Ipcom_netif;
|
|
|
|
#define IPCOM_ARPHRD_TYPE(netif) ((netif)->arphrd_type? (netif)->arphrd_type : (netif)->type)
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 7 FUNCTIONS
|
|
****************************************************************************
|
|
*/
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_malloc - allocate a new 'ipcom_netif' structure
|
|
*
|
|
* This routine allocates a new 'ipcom_netif' structure with enough space to hold
|
|
* the link-layer address and data. All fields are initialized, except
|
|
* 'drv_ioctl', 'drv_output', 'pdrv' and 'link_addr', which must be initialized
|
|
* before calling ipcom_if_attach().
|
|
*
|
|
* If an ipnet_config.c ipnet_conf_link_layer[x].init_func(netif) initialization
|
|
* routine has been registered for this interface type, it is now called to
|
|
* initialize the interface.
|
|
*
|
|
* The ipcom_if_malloc() routine sets a default interface name. The name can be
|
|
* changed before attaching the interface.
|
|
*
|
|
* Parameter:
|
|
* \is
|
|
* \i <type>
|
|
* Interface type, see 'ipcom_netif.h' for 'IP_IFT_'<xxx> types. Usually, either
|
|
* 'IP_IFT_ETHER' or 'IP_IFT_PPP'.
|
|
* \ie
|
|
*
|
|
* RETURNS: Pointer to a new 'ipcom_netif' structure on success, or 'IP_NULL' if
|
|
* not enough memory could be allocated.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC Ipcom_netif *ipcom_if_malloc
|
|
(
|
|
Ip_u16 type
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_free - free all resources allocated by ipcom_if_malloc()
|
|
*
|
|
* This routine frees all resources allocated by ipcom_if_malloc() for the
|
|
* specified network interface.
|
|
*
|
|
* Parameter:
|
|
* \is
|
|
* \i <netif>
|
|
* The interface to free.
|
|
* \ie
|
|
*
|
|
* RETURNS: 0 on success, otherwise a negative ERRNO.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC int ipcom_if_free
|
|
(
|
|
Ipcom_netif *netif
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_attach - add a specified network interface
|
|
*
|
|
* This routine adds a specified network interface to the list of available
|
|
* interfaces.
|
|
*
|
|
* The TCP/IP stack configures the route entries and addresses for the added
|
|
* interface using the configuration information in the TCP/IP stack 'config'
|
|
* directory, if this function is called at boot. If the function is called at a
|
|
* later time, interface addresses and routes must be added using socket system
|
|
* calls, such as ioctl() calls, for example.
|
|
*
|
|
* Parameter:
|
|
* \is
|
|
* \i <netif>
|
|
* The network interface to add.
|
|
* \ie
|
|
*
|
|
* RETURNS: The interface index number on success, or 'IP_ERRNO_EINVAL' if some
|
|
* of the fields in the 'ipcom_netif' structure are invalid.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC int ipcom_if_attach
|
|
(
|
|
Ipcom_netif *netif
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_attach_and_lock - add a specified network interface
|
|
*
|
|
* This routine adds a specified network interface to the list of available
|
|
* interfaces. The interface is locked by ipnet when attached. The caller must
|
|
* unlock it (i.e decrease reference count) when finished with it
|
|
*
|
|
* The TCP/IP stack configures the route entries and addresses for the added
|
|
* interface using the configuration information in the TCP/IP stack 'config'
|
|
* directory, if this function is called at boot. If the function is called at a
|
|
* later time, interface addresses and routes must be added using socket system
|
|
* calls, such as ioctl() calls, for example.
|
|
*
|
|
* Parameter:
|
|
* \is
|
|
* \i <netif>
|
|
* The network interface to add.
|
|
* \ie
|
|
*
|
|
* RETURNS: The interface index number on success, or 'IP_ERRNO_EINVAL' if some
|
|
* of the fields in the 'ipcom_netif' structure are invalid.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC int ipcom_if_attach_and_lock
|
|
(
|
|
Ipcom_netif *netif
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_detach - detach a previously attached network interface
|
|
*
|
|
* This routine detaches a previously attached network interface. The interface
|
|
* must be completely stopped (neither 'IFF_UP' or 'IFF_RUNNING' can be set)
|
|
* before being detached. An interface is brought down using
|
|
* 'ioctl(\'SIOCSIFLAGS\')' with 'IFF_UP' not set.
|
|
*
|
|
* The routine cleans the interface output queue and removes all IPv4 and IPv6
|
|
* addresses and routes on the interface. Processes waiting on sockets with any
|
|
* of the removed addresses are woken up, and the system call returns an error
|
|
* code.
|
|
*
|
|
* Once an interface has been detached, the interface memory must be released
|
|
* using the 'ipcom_if_free(netif)' routine.
|
|
*
|
|
* Parameter:
|
|
* \is
|
|
* \i <netif>
|
|
* The network interface to detach.
|
|
* \ie
|
|
*
|
|
* RETURNS: The interface index number on success, otherwise one of the
|
|
* following negative ERRNO codes:
|
|
* \is
|
|
* \i -'IP_ERRNO_EBUSY'
|
|
* The specified interface is still 'UP' or 'RUNNING'.
|
|
* \i -'IP_ERRNO_EINVAL'
|
|
* The specified interface <netif> pointer is invalid.
|
|
* \ie
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC int ipcom_if_detach
|
|
(
|
|
Ipcom_netif *netif
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_changename - change the name of a network interface
|
|
*
|
|
* This routine changes the name of an attached interface.
|
|
*
|
|
*
|
|
* Parameters:
|
|
* \is
|
|
* \i <ipcomif>
|
|
* The network interface whose name is to change.
|
|
* \i <new_name>
|
|
* The new name for the interface.
|
|
* \ie
|
|
*
|
|
* RETURNS: \&N/A
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC void ipcom_if_changename
|
|
(
|
|
Ipcom_netif *ipcomif,
|
|
const char *new_name
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_changemtu - change the mtu of a network interface
|
|
*
|
|
* This routine changes the mtu of an attached interface.
|
|
*
|
|
*
|
|
* Parameters:
|
|
* \is
|
|
* \i <ipcomif>
|
|
* The network interface whose name is to change.
|
|
* \i <new_mtu>
|
|
* The new mtu for the interface.
|
|
* \ie
|
|
*
|
|
* RETURNS: \&N/A
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC void ipcom_if_changemtu
|
|
(
|
|
Ipcom_netif *ipcomif,
|
|
Ip_u32 new_mtu
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_if_changelinkaddr - change the link layer addr of a network interface
|
|
*
|
|
* This routine changes the link layer address of an attached interface.
|
|
*
|
|
*
|
|
* Parameters:
|
|
* \is
|
|
* \i <ipcomif>
|
|
* The network interface whose name is to change.
|
|
* \i <lladdr>
|
|
* The new link layer address for the interface (byte array).
|
|
* \i <size>
|
|
* The size of the new link layer address.
|
|
* \ie
|
|
*
|
|
* RETURNS: \&N/A
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC void ipcom_if_changelinkaddr
|
|
(
|
|
Ipcom_netif *ipcomif,
|
|
Ip_u8 *lladdr,
|
|
Ip_u16 size
|
|
);
|
|
|
|
|
|
/*******************************************************************************
|
|
*
|
|
* ipcom_flow_fingerprint - create a flow fingerprint
|
|
*
|
|
* This routine calculates a fingerprint of a packet. Two packets that belongs
|
|
* to the same flow will have the same fingerprint.
|
|
* The 'pkt-\>start must' be the offset into 'pkt-\>data' where the link layer
|
|
* starts.
|
|
*
|
|
* Parameters:
|
|
* \is
|
|
* \i <ipcomif>
|
|
* The network interface that received the packet.
|
|
* \i <pkt>
|
|
* The packet for which a fingerprint should be calculated.
|
|
* \ie
|
|
*
|
|
* RETURNS: A flow fingerprint for the packet.
|
|
*
|
|
* ERRNO:
|
|
*/
|
|
IP_PUBLIC unsigned ipcom_flow_fingerprint(
|
|
Ipcom_netif *ipcomif,
|
|
Ipcom_pkt *pkt);
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* ipcom_drv_xxx_create
|
|
*===========================================================================
|
|
* Description:
|
|
* Parameters:
|
|
* Returns:
|
|
*
|
|
*/
|
|
IP_PUBLIC Ip_err ipcom_drv_eth_create(void);
|
|
IP_PUBLIC Ip_err ipcom_drv_ppp_create(void);
|
|
IP_PUBLIC Ip_err ipcom_drv_wlan_create(void);
|
|
|
|
#ifdef IPCOM_USE_LBP
|
|
IP_PUBLIC Ip_err ipcom_drv_lbp_create(void);
|
|
#endif
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* ipcom_drv_ppp_if_init
|
|
*===========================================================================
|
|
* Description:
|
|
* Parameters:
|
|
* Returns:
|
|
*
|
|
*/
|
|
IP_PUBLIC int
|
|
ipcom_drv_ppp_if_init(const char *ifname, const char *devname, int unit);
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* ipcom_if_nametonetif
|
|
*===========================================================================
|
|
* Description: Returns a pointer to the interface descriptor identified by
|
|
* its name.
|
|
* This function ensures that the interface is locked (reference counter
|
|
* has been increased).
|
|
* The user MUST self unlock the interface when finished with it.
|
|
*
|
|
* Parameters: vr - Actual VR to search for interface within.
|
|
* ifname - A buffer to hold the name, must be at least
|
|
* IP_IFNAMSIZ characters long.
|
|
* Returns: A pointer to the interface descriptor if found
|
|
* IP_NULL otherwise.
|
|
*
|
|
*/
|
|
IP_PUBLIC Ipcom_netif *ipcom_if_nametonetif
|
|
(
|
|
Ip_u16 vr,
|
|
IP_CONST char *ifname
|
|
);
|
|
|
|
|
|
/*
|
|
*===========================================================================
|
|
* ipcom_if_xxx LKM
|
|
*===========================================================================
|
|
* Description:
|
|
* Parameters:
|
|
* Returns:
|
|
*
|
|
*/
|
|
#ifdef IP_PORT_LKM
|
|
IP_PUBLIC void ipcom_netif_assign_vr(Ipcom_netif* netif, Ip_u16 vr);
|
|
IP_GLOBAL void ipcom_if_lock(Ipcom_netif *netif);
|
|
IP_GLOBAL void ipcom_if_unlock(Ipcom_netif *netif);
|
|
#endif
|
|
|
|
|
|
/*
|
|
****************************************************************************
|
|
* 8 VARIABLES
|
|
****************************************************************************
|
|
*/
|
|
|
|
/*
|
|
*===========================================================================
|
|
* ipcom_conf_max_link_hdr_size
|
|
*===========================================================================
|
|
* The largest link header size of all network interfaces the system have.
|
|
* Ex: 14 for a system with only Ethernet network interfaces. The value is
|
|
* calculated automatically by IPNET/IPLITE depending on supported interfaces.
|
|
*
|
|
* NOTE: An additional 4 bytes must be allocated due to driver 32-bit cookie!
|
|
*/
|
|
IP_EXTERN int ipcom_conf_max_link_hdr_size;
|
|
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|
|
|
|
#endif
|
|
|
|
|
|
/*
|
|
****************************************************************************
|
|
* END OF FILE
|
|
****************************************************************************
|
|
*/
|