firewire: net: style changes
authorStefan Richter <stefanr@s5r6.in-berlin.de>
Sun, 7 Jun 2009 20:57:53 +0000 (22:57 +0200)
committerStefan Richter <stefanr@s5r6.in-berlin.de>
Sun, 14 Jun 2009 12:26:29 +0000 (14:26 +0200)
Change names of types, variables, functions.
Omit debug code.
Use get_unaligned*, put_unaligned*.
Annotate big endian data.
Handle errors in __init.
Change whitespace.

Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>
drivers/firewire/core-card.c
drivers/firewire/net.c
include/linux/firewire.h

index cdab32b..8c45e43 100644 (file)
@@ -430,7 +430,7 @@ void fw_card_initialize(struct fw_card *card,
 
        INIT_DELAYED_WORK(&card->work, fw_card_bm_work);
        card->netdev = NULL;
-       INIT_LIST_HEAD(&card->ipv4_nodes);
+       INIT_LIST_HEAD(&card->peer_list);
 }
 EXPORT_SYMBOL(fw_card_initialize);
 
index 1535388..ba6f924 100644 (file)
@@ -6,6 +6,7 @@
  * based on eth1394 by Ben Collins et al
  */
 
+#include <linux/bug.h>
 #include <linux/device.h>
 #include <linux/ethtool.h>
 #include <linux/firewire.h>
@@ -13,6 +14,7 @@
 #include <linux/highmem.h>
 #include <linux/in.h>
 #include <linux/ip.h>
+#include <linux/jiffies.h>
 #include <linux/mod_devicetable.h>
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <asm/unaligned.h>
 #include <net/arp.h>
 
-/* Things to potentially make runtime cofigurable */
-/* must be at least as large as our maximum receive size */
-#define FIFO_SIZE 4096
-/* Network timeout in glibbles */
-#define IPV4_TIMEOUT       100000
+#define FWNET_MAX_FRAGMENTS    25      /* arbitrary limit */
+#define FWNET_ISO_PAGE_COUNT   (PAGE_SIZE < 16 * 1024 ? 4 : 2)
 
-/* Runitme configurable paramaters */
-static int ipv4_mpd = 25;
-static int ipv4_max_xmt = 0;
-/* 16k for receiving arp and broadcast packets.  Enough? */
-static int ipv4_iso_page_count = 4;
+#define IEEE1394_BROADCAST_CHANNEL     31
+#define IEEE1394_ALL_NODES             (0xffc0 | 0x003f)
+#define IEEE1394_MAX_PAYLOAD_S100      512
+#define FWNET_NO_FIFO_ADDR             (~0ULL)
 
-MODULE_AUTHOR("Jay Fenlason (fenlason@redhat.com)");
-MODULE_DESCRIPTION("Firewire IPv4 Driver (IPv4-over-IEEE1394 as per RFC 2734)");
-MODULE_LICENSE("GPL");
-MODULE_DEVICE_TABLE(ieee1394, ipv4_id_table);
-module_param_named(max_partial_datagrams, ipv4_mpd, int, S_IRUGO | S_IWUSR);
-MODULE_PARM_DESC(max_partial_datagrams, "Maximum number of received"
- " incomplete fragmented datagrams (default = 25).");
-
-/* Max xmt is useful for forcing fragmentation, which makes testing easier. */
-module_param_named(max_transmit, ipv4_max_xmt, int, S_IRUGO | S_IWUSR);
-MODULE_PARM_DESC(max_transmit, "Maximum datagram size to transmit"
- " (larger datagrams will be fragmented) (default = 0 (use hardware defaults).");
-
-/* iso page count controls how many pages will be used for receiving broadcast packets. */
-module_param_named(iso_pages, ipv4_iso_page_count, int, S_IRUGO | S_IWUSR);
-MODULE_PARM_DESC(iso_pages, "Number of pages to use for receiving broadcast packets"
- " (default = 4).");
-
-/* uncomment this line to do debugging */
-#define fw_debug(s, args...) printk(KERN_DEBUG KBUILD_MODNAME ": " s, ## args)
-
-/* comment out these lines to do debugging. */
-/* #undef fw_debug */
-/* #define fw_debug(s...) */
-/* #define print_hex_dump(l...) */
-
-/* Define a fake hardware header format for the networking core.  Note that
- * header size cannot exceed 16 bytes as that is the size of the header cache.
- * Also, we do not need the source address in the header so we omit it and
- * keep the header to under 16 bytes */
-#define IPV4_ALEN (8)
-/* This must equal sizeof(struct ipv4_ether_hdr) */
-#define IPV4_HLEN (10)
-
-/* FIXME: what's a good size for this? */
-#define INVALID_FIFO_ADDR (u64)~0ULL
-
-/* Things specified by standards */
-#define BROADCAST_CHANNEL 31
-
-#define S100_BUFFER_SIZE 512
-#define MAX_BUFFER_SIZE 4096
-
-#define IPV4_GASP_SPECIFIER_ID 0x00005EU
-#define IPV4_GASP_VERSION      0x00000001U
-
-#define IPV4_GASP_OVERHEAD (2 * sizeof(u32)) /* for GASP header */
-
-#define IPV4_UNFRAG_HDR_SIZE   sizeof(u32)
-#define IPV4_FRAG_HDR_SIZE     (2 * sizeof(u32))
-#define IPV4_FRAG_OVERHEAD     sizeof(u32)
-
-#define ALL_NODES (0xffc0 | 0x003f)
-
-#define IPV4_HDR_UNFRAG                0       /* unfragmented         */
-#define IPV4_HDR_FIRSTFRAG     1       /* first fragment       */
-#define IPV4_HDR_LASTFRAG      2       /* last fragment        */
-#define IPV4_HDR_INTFRAG       3       /* interior fragment    */
-
-/* Our arp packet (ARPHRD_IEEE1394) */
-/* FIXME: note that this is probably bogus on weird-endian machines */
-struct ipv4_arp {
-       u16 hw_type;            /* 0x0018       */
-       u16 proto_type;         /* 0x0806       */
-       u8 hw_addr_len;         /* 16           */
-       u8 ip_addr_len;         /* 4            */
-       u16 opcode;             /* ARP Opcode   */
-       /* Above is exactly the same format as struct arphdr */
-
-       u64 s_uniq_id;          /* Sender's 64bit EUI                   */
-       u8 max_rec;             /* Sender's max packet size             */
-       u8 sspd;                /* Sender's max speed                   */
-       u16 fifo_hi;            /* hi 16bits of sender's FIFO addr      */
-       u32 fifo_lo;            /* lo 32bits of sender's FIFO addr      */
-       u32 sip;                /* Sender's IP Address                  */
-       u32 tip;                /* IP Address of requested hw addr      */
-} __attribute__((packed));
+#define IANA_SPECIFIER_ID              0x00005eU
+#define RFC2734_SW_VERSION             0x000001U
 
-struct ipv4_ether_hdr {
-       unsigned char   h_dest[IPV4_ALEN];      /* destination address */
-       unsigned short  h_proto;                /* packet type ID field */
-}  __attribute__((packed));
+#define IEEE1394_GASP_HDR_SIZE 8
 
-static inline struct ipv4_ether_hdr *ipv4_ether_hdr(const struct sk_buff *skb)
-{
-       return (struct ipv4_ether_hdr *)skb_mac_header(skb);
-}
+#define RFC2374_UNFRAG_HDR_SIZE        4
+#define RFC2374_FRAG_HDR_SIZE  8
+#define RFC2374_FRAG_OVERHEAD  4
 
-enum ipv4_tx_type {
-       IPV4_UNKNOWN = 0,
-       IPV4_GASP = 1,
-       IPV4_WRREQ = 2,
-};
+#define RFC2374_HDR_UNFRAG     0       /* unfragmented         */
+#define RFC2374_HDR_FIRSTFRAG  1       /* first fragment       */
+#define RFC2374_HDR_LASTFRAG   2       /* last fragment        */
+#define RFC2374_HDR_INTFRAG    3       /* interior fragment    */
 
-enum ipv4_broadcast_state {
-       IPV4_BROADCAST_ERROR,
-       IPV4_BROADCAST_RUNNING,
-       IPV4_BROADCAST_STOPPED,
-};
+#define RFC2734_HW_ADDR_LEN    16
 
-#define ipv4_get_hdr_lf(h)             (((h)->w0&0xC0000000)>>30)
-#define ipv4_get_hdr_ether_type(h)     (((h)->w0&0x0000FFFF)    )
-#define ipv4_get_hdr_dg_size(h)                (((h)->w0&0x0FFF0000)>>16)
-#define ipv4_get_hdr_fg_off(h)         (((h)->w0&0x00000FFF)    )
-#define ipv4_get_hdr_dgl(h)            (((h)->w1&0xFFFF0000)>>16)
+struct rfc2734_arp {
+       __be16 hw_type;         /* 0x0018       */
+       __be16 proto_type;      /* 0x0806       */
+       u8 hw_addr_len;         /* 16           */
+       u8 ip_addr_len;         /* 4            */
+       __be16 opcode;          /* ARP Opcode   */
+       /* Above is exactly the same format as struct arphdr */
 
-#define ipv4_set_hdr_lf(lf)            (( lf)<<30)
-#define ipv4_set_hdr_ether_type(et)    (( et)    )
-#define ipv4_set_hdr_dg_size(dgs)      ((dgs)<<16)
-#define ipv4_set_hdr_fg_off(fgo)       ((fgo)    )
+       __be64 s_uniq_id;       /* Sender's 64bit EUI                   */
+       u8 max_rec;             /* Sender's max packet size             */
+       u8 sspd;                /* Sender's max speed                   */
+       __be16 fifo_hi;         /* hi 16bits of sender's FIFO addr      */
+       __be32 fifo_lo;         /* lo 32bits of sender's FIFO addr      */
+       __be32 sip;             /* Sender's IP Address                  */
+       __be32 tip;             /* IP Address of requested hw addr      */
+} __attribute__((packed));
 
-#define ipv4_set_hdr_dgl(dgl)          ((dgl)<<16)
+/* This header format is specific to this driver implementation. */
+#define FWNET_ALEN     8
+#define FWNET_HLEN     10
+struct fwnet_header {
+       u8 h_dest[FWNET_ALEN];  /* destination address */
+       __be16 h_proto;         /* packet type ID field */
+} __attribute__((packed));
 
-struct ipv4_hdr {
+/* IPv4 and IPv6 encapsulation header */
+struct rfc2734_header {
        u32 w0;
        u32 w1;
 };
 
-static inline void ipv4_make_uf_hdr( struct ipv4_hdr *hdr, unsigned ether_type) {
-       hdr->w0 = ipv4_set_hdr_lf(IPV4_HDR_UNFRAG)
-                  |ipv4_set_hdr_ether_type(ether_type);
-       fw_debug ( "Setting unfragmented header %p to %x\n", hdr, hdr->w0 );
-}
+#define fwnet_get_hdr_lf(h)            (((h)->w0 & 0xc0000000) >> 30)
+#define fwnet_get_hdr_ether_type(h)    (((h)->w0 & 0x0000ffff))
+#define fwnet_get_hdr_dg_size(h)       (((h)->w0 & 0x0fff0000) >> 16)
+#define fwnet_get_hdr_fg_off(h)                (((h)->w0 & 0x00000fff))
+#define fwnet_get_hdr_dgl(h)           (((h)->w1 & 0xffff0000) >> 16)
 
-static inline void ipv4_make_ff_hdr ( struct ipv4_hdr *hdr, unsigned ether_type, unsigned dg_size, unsigned dgl ) {
-       hdr->w0 = ipv4_set_hdr_lf(IPV4_HDR_FIRSTFRAG)
-                  |ipv4_set_hdr_dg_size(dg_size)
-                  |ipv4_set_hdr_ether_type(ether_type);
-       hdr->w1 = ipv4_set_hdr_dgl(dgl);
-       fw_debug ( "Setting fragmented header %p to first_frag %x,%x (et %x, dgs %x, dgl %x)\n", hdr, hdr->w0, hdr->w1,
- ether_type, dg_size, dgl );
-}
+#define fwnet_set_hdr_lf(lf)           ((lf)  << 30)
+#define fwnet_set_hdr_ether_type(et)   (et)
+#define fwnet_set_hdr_dg_size(dgs)     ((dgs) << 16)
+#define fwnet_set_hdr_fg_off(fgo)      (fgo)
 
-static inline void ipv4_make_sf_hdr ( struct ipv4_hdr *hdr, unsigned lf, unsigned dg_size, unsigned fg_off, unsigned dgl) {
-       hdr->w0 = ipv4_set_hdr_lf(lf)
-                |ipv4_set_hdr_dg_size(dg_size)
-                |ipv4_set_hdr_fg_off(fg_off);
-       hdr->w1 = ipv4_set_hdr_dgl(dgl);
-       fw_debug ( "Setting fragmented header %p to %x,%x (lf %x, dgs %x, fo %x dgl %x)\n",
- hdr, hdr->w0, hdr->w1,
- lf, dg_size, fg_off, dgl );
-}
+#define fwnet_set_hdr_dgl(dgl)         ((dgl) << 16)
 
-/* End of IP1394 headers */
+static inline void fwnet_make_uf_hdr(struct rfc2734_header *hdr,
+               unsigned ether_type)
+{
+       hdr->w0 = fwnet_set_hdr_lf(RFC2374_HDR_UNFRAG)
+                 | fwnet_set_hdr_ether_type(ether_type);
+}
 
-/* Fragment types */
-#define ETH1394_HDR_LF_UF      0       /* unfragmented         */
-#define ETH1394_HDR_LF_FF      1       /* first fragment       */
-#define ETH1394_HDR_LF_LF      2       /* last fragment        */
-#define ETH1394_HDR_LF_IF      3       /* interior fragment    */
+static inline void fwnet_make_ff_hdr(struct rfc2734_header *hdr,
+               unsigned ether_type, unsigned dg_size, unsigned dgl)
+{
+       hdr->w0 = fwnet_set_hdr_lf(RFC2374_HDR_FIRSTFRAG)
+                 | fwnet_set_hdr_dg_size(dg_size)
+                 | fwnet_set_hdr_ether_type(ether_type);
+       hdr->w1 = fwnet_set_hdr_dgl(dgl);
+}
 
-#define IP1394_HW_ADDR_LEN     16      /* As per RFC           */
+static inline void fwnet_make_sf_hdr(struct rfc2734_header *hdr,
+               unsigned lf, unsigned dg_size, unsigned fg_off, unsigned dgl)
+{
+       hdr->w0 = fwnet_set_hdr_lf(lf)
+                 | fwnet_set_hdr_dg_size(dg_size)
+                 | fwnet_set_hdr_fg_off(fg_off);
+       hdr->w1 = fwnet_set_hdr_dgl(dgl);
+}
 
 /* This list keeps track of what parts of the datagram have been filled in */
-struct ipv4_fragment_info {
-        struct list_head fragment_info;
+struct fwnet_fragment_info {
+       struct list_head fi_link;
        u16 offset;
        u16 len;
 };
 
-struct ipv4_partial_datagram {
-       struct list_head pdg_list;
-       struct list_head fragment_info;
+struct fwnet_partial_datagram {
+       struct list_head pd_link;
+       struct list_head fi_list;
        struct sk_buff *skb;
        /* FIXME Why not use skb->data? */
        char *pbuf;
@@ -208,40 +138,43 @@ struct ipv4_partial_datagram {
 /*
  * We keep one of these for each IPv4 capable device attached to a fw_card.
  * The list of them is stored in the fw_card structure rather than in the
- * ipv4_priv because the remote IPv4 nodes may be probed before the card is,
- * so we need a place to store them before the ipv4_priv structure is
+ * fwnet_device because the remote IPv4 nodes may be probed before the card is,
+ * so we need a place to store them before the fwnet_device structure is
  * allocated.
  */
-struct ipv4_node {
-       struct list_head ipv4_nodes;
-       /* guid of the remote node */
+struct fwnet_peer {
+       struct list_head peer_link;
+       /* guid of the remote peer */
        u64 guid;
-       /* FIFO address to transmit datagrams to, or INVALID_FIFO_ADDR */
+       /* FIFO address to transmit datagrams to, or FWNET_NO_FIFO_ADDR */
        u64 fifo;
 
        spinlock_t pdg_lock;    /* partial datagram lock                */
-       /* List of partial datagrams received from this node */
-       struct list_head pdg_list;
-       /* Number of entries in pdg_list at the moment */
+       /* List of partial datagrams received from this peer */
+       struct list_head pd_list;
+       /* Number of entries in pd_list at the moment */
        unsigned pdg_size;
 
-       /* max payload to transmit to this remote node */
-       /* This already includes the IPV4_FRAG_HDR_SIZE overhead */
+       /* max payload to transmit to this remote peer */
+       /* This already includes the RFC2374_FRAG_HDR_SIZE overhead */
        u16 max_payload;
        /* outgoing datagram label */
        u16 datagram_label;
-       /* Current node_id of the remote node */
-       u16 nodeid;
-       /* current generation of the remote node */
+       /* Current node_id of the remote peer */
+       u16 node_id;
+       /* current generation of the remote peer */
        u8 generation;
-       /* max speed that this node can receive at */
+       /* max speed that this peer can receive at */
        u8 xmt_speed;
 };
 
-struct ipv4_priv {
+struct fwnet_device {
        spinlock_t lock;
-
-       enum ipv4_broadcast_state broadcast_state;
+       enum {
+               FWNET_BROADCAST_ERROR,
+               FWNET_BROADCAST_RUNNING,
+               FWNET_BROADCAST_STOPPED,
+       } broadcast_state;
        struct fw_iso_context *broadcast_rcv_context;
        struct fw_iso_buffer broadcast_rcv_buffer;
        void **broadcast_rcv_buffer_ptrs;
@@ -257,14 +190,12 @@ struct ipv4_priv {
        u16 broadcast_xmt_datagramlabel;
 
        /*
-        * The csr address that remote nodes must send datagrams to for us to
+        * The CSR address that remote nodes must send datagrams to for us to
         * receive them.
         */
        struct fw_address_handler handler;
        u64 local_fifo;
 
-       /* Wake up to xmt        */
-        /* struct work_struct wake;*/
        /* List of packets to be sent */
        struct list_head packet_list;
        /*
@@ -279,17 +210,17 @@ struct ipv4_priv {
 };
 
 /* This is our task struct. It's used for the packet complete callback.  */
-struct ipv4_packet_task {
+struct fwnet_packet_task {
        /*
-        * ptask can actually be on priv->packet_list, priv->broadcasted_list,
-        * or priv->sent_list depending on its current state.
+        * ptask can actually be on dev->packet_list, dev->broadcasted_list,
+        * or dev->sent_list depending on its current state.
         */
-       struct list_head packet_list;
+       struct list_head pt_link;
        struct fw_transaction transaction;
-       struct ipv4_hdr hdr;
+       struct rfc2734_header hdr;
        struct sk_buff *skb;
-       struct ipv4_priv *priv;
-       enum ipv4_tx_type tx_type;
+       struct fwnet_device *dev;
+
        int outstanding_pkts;
        unsigned max_payload;
        u64 fifo_addr;
@@ -298,243 +229,192 @@ struct ipv4_packet_task {
        u8 speed;
 };
 
-static struct kmem_cache *ipv4_packet_task_cache;
-
-static const char ipv4_driver_name[] = "firewire-ipv4";
-
-static const struct ieee1394_device_id ipv4_id_table[] = {
-       {
-               .match_flags  = IEEE1394_MATCH_SPECIFIER_ID |
-                               IEEE1394_MATCH_VERSION,
-               .specifier_id = IPV4_GASP_SPECIFIER_ID,
-               .version      = IPV4_GASP_VERSION,
-       },
-       { }
-};
-
-static u32 ipv4_unit_directory_data[] = {
-       0x00040000,                                     /* unit directory */
-       0x12000000 | IPV4_GASP_SPECIFIER_ID,    /* specifier ID */
-       0x81000003,                                     /* text descriptor */
-       0x13000000 | IPV4_GASP_VERSION,         /* version */
-       0x81000005,                                     /* text descriptor */
-
-       0x00030000,                                     /* Three quadlets */
-       0x00000000,                                     /* Text */
-       0x00000000,                                     /* Language 0 */
-       0x49414e41,                                     /* I A N A */
-       0x00030000,                                     /* Three quadlets */
-       0x00000000,                                     /* Text */
-       0x00000000,                                     /* Language 0 */
-       0x49507634,                                     /* I P v 4 */
-};
-
-static struct fw_descriptor ipv4_unit_directory = {
-       .length = ARRAY_SIZE(ipv4_unit_directory_data),
-       .key = 0xd1000000,
-       .data = ipv4_unit_directory_data
-};
-
-static int ipv4_send_packet(struct ipv4_packet_task *ptask );
-
-/* ------------------------------------------------------------------ */
-/******************************************
- * HW Header net device functions
- ******************************************/
-  /* These functions have been adapted from net/ethernet/eth.c */
-
-/* Create a fake MAC header for an arbitrary protocol layer.
- * saddr=NULL means use device source address
- * daddr=NULL means leave destination address (eg unresolved arp). */
+/*
+ * saddr == NULL means use device source address.
+ * daddr == NULL means leave destination address (eg unresolved arp).
+ */
+static int fwnet_header_create(struct sk_buff *skb, struct net_device *net,
+                       unsigned short type, const void *daddr,
+                       const void *saddr, unsigned len)
+{
+       struct fwnet_header *h;
 
-static int ipv4_header ( struct sk_buff *skb, struct net_device *dev,
-                      unsigned short type, const void *daddr,
-                      const void *saddr, unsigned len) {
-       struct ipv4_ether_hdr *eth;
+       h = (struct fwnet_header *)skb_push(skb, sizeof(*h));
+       put_unaligned_be16(type, &h->h_proto);
 
-       eth = (struct ipv4_ether_hdr *)skb_push(skb, sizeof(*eth));
-       eth->h_proto = htons(type);
+       if (net->flags & (IFF_LOOPBACK | IFF_NOARP)) {
+               memset(h->h_dest, 0, net->addr_len);
 
-       if (dev->flags & (IFF_LOOPBACK | IFF_NOARP)) {
-               memset(eth->h_dest, 0, dev->addr_len);
-               return dev->hard_header_len;
+               return net->hard_header_len;
        }
 
        if (daddr) {
-               memcpy(eth->h_dest, daddr, dev->addr_len);
-               return dev->hard_header_len;
+               memcpy(h->h_dest, daddr, net->addr_len);
+
+               return net->hard_header_len;
        }
 
-       return -dev->hard_header_len;
+       return -net->hard_header_len;
 }
 
-/* Rebuild the faked MAC header. This is called after an ARP
- * (or in future other address resolution) has completed on this
- * sk_buff. We now let ARP fill in the other fields.
- *
- * This routine CANNOT use cached dst->neigh!
- * Really, it is used only when dst->neigh is wrong.
- */
-
-static int ipv4_rebuild_header(struct sk_buff *skb)
+static int fwnet_header_rebuild(struct sk_buff *skb)
 {
-       struct ipv4_ether_hdr *eth;
+       struct fwnet_header *h = (struct fwnet_header *)skb->data;
 
-       eth = (struct ipv4_ether_hdr *)skb->data;
-       if (eth->h_proto == htons(ETH_P_IP))
-               return arp_find((unsigned char *)&eth->h_dest, skb);
+       if (get_unaligned_be16(&h->h_proto) == ETH_P_IP)
+               return arp_find((unsigned char *)&h->h_dest, skb);
 
-       fw_notify ( "%s: unable to resolve type %04x addresses\n",
-                  skb->dev->name,ntohs(eth->h_proto) );
+       fw_notify("%s: unable to resolve type %04x addresses\n",
+                 skb->dev->name, be16_to_cpu(h->h_proto));
        return 0;
 }
 
-static int ipv4_header_cache(const struct neighbour *neigh, struct hh_cache *hh) {
-       unsigned short type = hh->hh_type;
-       struct net_device *dev;
-       struct ipv4_ether_hdr *eth;
+static int fwnet_header_cache(const struct neighbour *neigh,
+                             struct hh_cache *hh)
+{
+       struct net_device *net;
+       struct fwnet_header *h;
 
-       if (type == htons(ETH_P_802_3))
+       if (hh->hh_type == cpu_to_be16(ETH_P_802_3))
                return -1;
-       dev = neigh->dev;
-       eth = (struct ipv4_ether_hdr *)((u8 *)hh->hh_data + 16 - sizeof(*eth));
-       eth->h_proto = type;
-       memcpy(eth->h_dest, neigh->ha, dev->addr_len);
+       net = neigh->dev;
+       h = (struct fwnet_header *)((u8 *)hh->hh_data + 16 - sizeof(*h));
+       h->h_proto = hh->hh_type;
+       memcpy(h->h_dest, neigh->ha, net->addr_len);
+       hh->hh_len = FWNET_HLEN;
 
-       hh->hh_len = IPV4_HLEN;
        return 0;
 }
 
 /* Called by Address Resolution module to notify changes in address. */
-static void ipv4_header_cache_update(struct hh_cache *hh, const struct net_device *dev, const unsigned char * haddr ) {
-       memcpy((u8 *)hh->hh_data + 16 - IPV4_HLEN, haddr, dev->addr_len);
+static void fwnet_header_cache_update(struct hh_cache *hh,
+               const struct net_device *net, const unsigned char *haddr)
+{
+       memcpy((u8 *)hh->hh_data + 16 - FWNET_HLEN, haddr, net->addr_len);
 }
 
-static int ipv4_header_parse(const struct sk_buff *skb, unsigned char *haddr) {
-       memcpy(haddr, skb->dev->dev_addr, IPV4_ALEN);
-       return IPV4_ALEN;
+static int fwnet_header_parse(const struct sk_buff *skb, unsigned char *haddr)
+{
+       memcpy(haddr, skb->dev->dev_addr, FWNET_ALEN);
+
+       return FWNET_ALEN;
 }
 
-static const struct header_ops ipv4_header_ops = {
-       .create         = ipv4_header,
-       .rebuild        = ipv4_rebuild_header,
-       .cache          = ipv4_header_cache,
-       .cache_update   = ipv4_header_cache_update,
-       .parse          = ipv4_header_parse,
+static const struct header_ops fwnet_header_ops = {
+       .create         = fwnet_header_create,
+       .rebuild        = fwnet_header_rebuild,
+       .cache          = fwnet_header_cache,
+       .cache_update   = fwnet_header_cache_update,
+       .parse          = fwnet_header_parse,
 };
 
-/* ------------------------------------------------------------------ */
-
 /* FIXME: is this correct for all cases? */
-static bool ipv4_frag_overlap(struct ipv4_partial_datagram *pd, unsigned offset, unsigned len)
+static bool fwnet_frag_overlap(struct fwnet_partial_datagram *pd,
+                              unsigned offset, unsigned len)
 {
-        struct ipv4_fragment_info *fi;
+       struct fwnet_fragment_info *fi;
        unsigned end = offset + len;
 
-       list_for_each_entry(fi, &pd->fragment_info, fragment_info) {
-               if (offset < fi->offset + fi->len && end > fi->offset) {
-                       fw_debug ( "frag_overlap pd %p fi %p (%x@%x) with %x@%x\n", pd, fi, fi->len, fi->offset, len, offset );
+       list_for_each_entry(fi, &pd->fi_list, fi_link)
+               if (offset < fi->offset + fi->len && end > fi->offset)
                        return true;
-               }
-       }
-       fw_debug ( "frag_overlap %p does not overlap with %x@%x\n", pd, len, offset );
+
        return false;
 }
 
 /* Assumes that new fragment does not overlap any existing fragments */
-static struct ipv4_fragment_info *ipv4_frag_new ( struct ipv4_partial_datagram *pd, unsigned offset, unsigned len ) {
-       struct ipv4_fragment_info *fi, *fi2, *new;
+static struct fwnet_fragment_info *fwnet_frag_new(
+       struct fwnet_partial_datagram *pd, unsigned offset, unsigned len)
+{
+       struct fwnet_fragment_info *fi, *fi2, *new;
        struct list_head *list;
 
-       fw_debug ( "frag_new pd %p %x@%x\n", pd, len, offset );
-       list = &pd->fragment_info;
-       list_for_each_entry(fi, &pd->fragment_info, fragment_info) {
+       list = &pd->fi_list;
+       list_for_each_entry(fi, &pd->fi_list, fi_link) {
                if (fi->offset + fi->len == offset) {
                        /* The new fragment can be tacked on to the end */
                        /* Did the new fragment plug a hole? */
-                       fi2 = list_entry(fi->fragment_info.next, struct ipv4_fragment_info, fragment_info);
+                       fi2 = list_entry(fi->fi_link.next,
+                                        struct fwnet_fragment_info, fi_link);
                        if (fi->offset + fi->len == fi2->offset) {
-                               fw_debug ( "pd %p: hole filling %p (%x@%x) and %p(%x@%x): now %x@%x\n", pd, fi, fi->len, fi->offset,
-                               fi2, fi2->len, fi2->offset, fi->len + len + fi2->len, fi->offset );
                                /* glue fragments together */
                                fi->len += len + fi2->len;
-                               list_del(&fi2->fragment_info);
+                               list_del(&fi2->fi_link);
                                kfree(fi2);
                        } else {
-                               fw_debug ( "pd %p: extending %p from %x@%x to %x@%x\n", pd, fi, fi->len, fi->offset, fi->len+len, fi->offset );
                                fi->len += len;
                        }
+
                        return fi;
                }
                if (offset + len == fi->offset) {
                        /* The new fragment can be tacked on to the beginning */
                        /* Did the new fragment plug a hole? */
-                       fi2 = list_entry(fi->fragment_info.prev, struct ipv4_fragment_info, fragment_info);
+                       fi2 = list_entry(fi->fi_link.prev,
+                                        struct fwnet_fragment_info, fi_link);
                        if (fi2->offset + fi2->len == fi->offset) {
                                /* glue fragments together */
-                               fw_debug ( "pd %p: extending %p and merging with %p from %x@%x to %x@%x\n",
- pd, fi2, fi, fi2->len, fi2->offset, fi2->len + fi->len + len, fi2->offset );
                                fi2->len += fi->len + len;
-                               list_del(&fi->fragment_info);
+                               list_del(&fi->fi_link);
                                kfree(fi);
+
                                return fi2;
                        }
-                       fw_debug ( "pd %p: extending %p from %x@%x to %x@%x\n", pd, fi, fi->len, fi->offset, offset, fi->len + len );
                        fi->offset = offset;
                        fi->len += len;
+
                        return fi;
                }
                if (offset > fi->offset + fi->len) {
-                       list = &fi->fragment_info;
+                       list = &fi->fi_link;
                        break;
                }
                if (offset + len < fi->offset) {
-                       list = fi->fragment_info.prev;
+                       list = fi->fi_link.prev;
                        break;
                }
        }
 
        new = kmalloc(sizeof(*new), GFP_ATOMIC);
        if (!new) {
-               fw_error ( "out of memory in fragment handling!\n" );
+               fw_error("out of memory\n");
                return NULL;
        }
 
        new->offset = offset;
        new->len = len;
-       list_add(&new->fragment_info, list);
-       fw_debug ( "pd %p: new frag %p %x@%x\n", pd, new, new->len, new->offset );
-       list_for_each_entry( fi, &pd->fragment_info, fragment_info )
-               fw_debug ( "fi %p %x@%x\n", fi, fi->len, fi->offset );
+       list_add(&new->fi_link, list);
+
        return new;
 }
 
-/* ------------------------------------------------------------------ */
-
-static struct ipv4_partial_datagram *ipv4_pd_new(struct net_device *netdev,
- struct ipv4_node *node, u16 datagram_label, unsigned dg_size, u32 *frag_buf,
- unsigned frag_off, unsigned frag_len) {
-       struct ipv4_partial_datagram *new;
-       struct ipv4_fragment_info *fi;
+static struct fwnet_partial_datagram *fwnet_pd_new(struct net_device *net,
+               struct fwnet_peer *peer, u16 datagram_label, unsigned dg_size,
+               void *frag_buf, unsigned frag_off, unsigned frag_len)
+{
+       struct fwnet_partial_datagram *new;
+       struct fwnet_fragment_info *fi;
 
        new = kmalloc(sizeof(*new), GFP_ATOMIC);
        if (!new)
                goto fail;
-       INIT_LIST_HEAD(&new->fragment_info);
-       fi = ipv4_frag_new ( new, frag_off, frag_len);
-       if ( fi == NULL )
+
+       INIT_LIST_HEAD(&new->fi_list);
+       fi = fwnet_frag_new(new, frag_off, frag_len);
+       if (fi == NULL)
                goto fail_w_new;
+
        new->datagram_label = datagram_label;
        new->datagram_size = dg_size;
-       new->skb = dev_alloc_skb(dg_size + netdev->hard_header_len + 15);
-       if ( new->skb == NULL )
+       new->skb = dev_alloc_skb(dg_size + net->hard_header_len + 15);
+       if (new->skb == NULL)
                goto fail_w_fi;
-       skb_reserve(new->skb, (netdev->hard_header_len + 15) & ~15);
+
+       skb_reserve(new->skb, (net->hard_header_len + 15) & ~15);
        new->pbuf = skb_put(new->skb, dg_size);
        memcpy(new->pbuf + frag_off, frag_buf, frag_len);
-       list_add_tail(&new->pdg_list, &node->pdg_list);
-       fw_debug ( "pd_new: new pd %p { dgl %u, dg_size %u, skb %p, pbuf %p } on node %p\n",
- new, new->datagram_label, new->datagram_size, new->skb, new->pbuf, node );
+       list_add_tail(&new->pd_link, &peer->pd_list);
+
        return new;
 
 fail_w_fi:
@@ -542,174 +422,171 @@ fail_w_fi:
 fail_w_new:
        kfree(new);
 fail:
-       fw_error("ipv4_pd_new: no memory\n");
+       fw_error("out of memory\n");
+
        return NULL;
 }
 
-static struct ipv4_partial_datagram *ipv4_pd_find(struct ipv4_node *node, u16 datagram_label) {
-       struct ipv4_partial_datagram *pd;
+static struct fwnet_partial_datagram *fwnet_pd_find(struct fwnet_peer *peer,
+                                                   u16 datagram_label)
+{
+       struct fwnet_partial_datagram *pd;
 
-       list_for_each_entry(pd, &node->pdg_list, pdg_list) {
-               if ( pd->datagram_label == datagram_label ) {
-                       fw_debug ( "pd_find(node %p, label %u): pd %p\n", node, datagram_label, pd );
+       list_for_each_entry(pd, &peer->pd_list, pd_link)
+               if (pd->datagram_label == datagram_label)
                        return pd;
-               }
-       }
-       fw_debug ( "pd_find(node %p, label %u) no entry\n", node, datagram_label );
+
        return NULL;
 }
 
 
-static void ipv4_pd_delete ( struct ipv4_partial_datagram *old ) {
-       struct ipv4_fragment_info *fi, *n;
+static void fwnet_pd_delete(struct fwnet_partial_datagram *old)
+{
+       struct fwnet_fragment_info *fi, *n;
 
-       fw_debug ( "pd_delete %p\n", old );
-       list_for_each_entry_safe(fi, n, &old->fragment_info, fragment_info) {
-               fw_debug ( "Freeing fi %p\n", fi );
+       list_for_each_entry_safe(fi, n, &old->fi_list, fi_link)
                kfree(fi);
-       }
-       list_del(&old->pdg_list);
+
+       list_del(&old->pd_link);
        dev_kfree_skb_any(old->skb);
        kfree(old);
 }
 
-static bool ipv4_pd_update ( struct ipv4_node *node, struct ipv4_partial_datagram *pd,
- u32 *frag_buf, unsigned frag_off, unsigned frag_len) {
-       fw_debug ( "pd_update node %p, pd %p, frag_buf %p, %x@%x\n", node, pd, frag_buf, frag_len, frag_off );
-       if ( ipv4_frag_new ( pd, frag_off, frag_len ) == NULL)
+static bool fwnet_pd_update(struct fwnet_peer *peer,
+               struct fwnet_partial_datagram *pd, void *frag_buf,
+               unsigned frag_off, unsigned frag_len)
+{
+       if (fwnet_frag_new(pd, frag_off, frag_len) == NULL)
                return false;
+
        memcpy(pd->pbuf + frag_off, frag_buf, frag_len);
 
        /*
         * Move list entry to beginnig of list so that oldest partial
         * datagrams percolate to the end of the list
         */
-       list_move_tail(&pd->pdg_list, &node->pdg_list);
-       fw_debug ( "New pd list:\n" );
-       list_for_each_entry ( pd, &node->pdg_list, pdg_list ) {
-               fw_debug ( "pd %p\n", pd );
-       }
+       list_move_tail(&pd->pd_link, &peer->pd_list);
+
        return true;
 }
 
-static bool ipv4_pd_is_complete ( struct ipv4_partial_datagram *pd ) {
-       struct ipv4_fragment_info *fi;
-       bool ret;
+static bool fwnet_pd_is_complete(struct fwnet_partial_datagram *pd)
+{
+       struct fwnet_fragment_info *fi;
 
-       fi = list_entry(pd->fragment_info.next, struct ipv4_fragment_info, fragment_info);
+       fi = list_entry(pd->fi_list.next, struct fwnet_fragment_info, fi_link);
 
-       ret = (fi->len == pd->datagram_size);
-       fw_debug ( "pd_is_complete (pd %p, dgs %x): fi %p (%x@%x) %s\n", pd, pd->datagram_size, fi, fi->len, fi->offset, ret ? "yes" : "no" );
-       return ret;
+       return fi->len == pd->datagram_size;
 }
 
-/* ------------------------------------------------------------------ */
+static int fwnet_peer_new(struct fw_card *card, struct fw_device *device)
+{
+       struct fwnet_peer *peer;
 
-static int ipv4_node_new ( struct fw_card *card, struct fw_device *device ) {
-       struct ipv4_node *node;
+       peer = kmalloc(sizeof(*peer), GFP_KERNEL);
+       if (!peer) {
+               fw_error("out of memory\n");
 
-       node = kmalloc ( sizeof(*node), GFP_KERNEL );
-       if ( ! node ) {
-               fw_error ( "allocate new node failed\n" );
                return -ENOMEM;
        }
-       node->guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
-       node->fifo = INVALID_FIFO_ADDR;
-       INIT_LIST_HEAD(&node->pdg_list);
-       spin_lock_init(&node->pdg_lock);
-       node->pdg_size = 0;
-       node->generation = device->generation;
+       peer->guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
+       peer->fifo = FWNET_NO_FIFO_ADDR;
+       INIT_LIST_HEAD(&peer->pd_list);
+       spin_lock_init(&peer->pdg_lock);
+       peer->pdg_size = 0;
+       peer->generation = device->generation;
        rmb();
-       node->nodeid = device->node_id;
+       peer->node_id = device->node_id;
         /* FIXME what should it really be? */
-       node->max_payload = S100_BUFFER_SIZE - IPV4_UNFRAG_HDR_SIZE;
-       node->datagram_label = 0U;
-       node->xmt_speed = device->max_speed;
-       list_add_tail ( &node->ipv4_nodes, &card->ipv4_nodes );
-       fw_debug ( "node_new: %p { guid %016llx, generation %u, nodeid %x, max_payload %x, xmt_speed %x } added\n",
- node, (unsigned long long)node->guid, node->generation, node->nodeid, node->max_payload, node->xmt_speed );
+       peer->max_payload = IEEE1394_MAX_PAYLOAD_S100 - RFC2374_UNFRAG_HDR_SIZE;
+       peer->datagram_label = 0U;
+       peer->xmt_speed = device->max_speed;
+       list_add_tail(&peer->peer_link, &card->peer_list);
+
        return 0;
 }
 
-static struct ipv4_node *ipv4_node_find_by_guid(struct ipv4_priv *priv, u64 guid) {
-       struct ipv4_node *node;
+/* FIXME caller must take the lock, or peer needs to be reference-counted */
+static struct fwnet_peer *fwnet_peer_find_by_guid(struct fwnet_device *dev,
+                                                 u64 guid)
+{
+       struct fwnet_peer *p, *peer = NULL;
        unsigned long flags;
 
-       spin_lock_irqsave(&priv->lock, flags);
-       list_for_each_entry(node, &priv->card->ipv4_nodes, ipv4_nodes)
-               if (node->guid == guid) {
-                       /* FIXME: lock the node first? */
-                       spin_unlock_irqrestore ( &priv->lock, flags );
-                       fw_debug ( "node_find_by_guid (%016llx) found %p\n", (unsigned long long)guid, node );
-                       return node;
+       spin_lock_irqsave(&dev->lock, flags);
+       list_for_each_entry(p, &dev->card->peer_list, peer_link)
+               if (p->guid == guid) {
+                       peer = p;
+                       break;
                }
+       spin_unlock_irqrestore(&dev->lock, flags);
 
-       spin_unlock_irqrestore ( &priv->lock, flags );
-       fw_debug ( "node_find_by_guid (%016llx) not found\n", (unsigned long long)guid );
-       return NULL;
+       return peer;
 }
 
-static struct ipv4_node *ipv4_node_find_by_nodeid(struct ipv4_priv *priv, u16 nodeid) {
-       struct ipv4_node *node;
+/* FIXME caller must take the lock, or peer needs to be reference-counted */
+/* FIXME node_id doesn't mean anything without generation */
+static struct fwnet_peer *fwnet_peer_find_by_node_id(struct fwnet_device *dev,
+                                                    u16 node_id)
+{
+       struct fwnet_peer *p, *peer = NULL;
        unsigned long flags;
 
-       spin_lock_irqsave(&priv->lock, flags);
-       list_for_each_entry(node, &priv->card->ipv4_nodes, ipv4_nodes)
-               if (node->nodeid == nodeid) {
-                       /* FIXME: lock the node first? */
-                       spin_unlock_irqrestore ( &priv->lock, flags );
-                       fw_debug ( "node_find_by_nodeid (%x) found %p\n", nodeid, node );
-                       return node;
+       spin_lock_irqsave(&dev->lock, flags);
+       list_for_each_entry(p, &dev->card->peer_list, peer_link)
+               if (p->node_id == node_id) {
+                       peer = p;
+                       break;
                }
-       fw_debug ( "node_find_by_nodeid (%x) not found\n", nodeid );
-       spin_unlock_irqrestore ( &priv->lock, flags );
-       return NULL;
+       spin_unlock_irqrestore(&dev->lock, flags);
+
+       return peer;
 }
 
-/* This is only complicated because we can't assume priv exists */
-static void ipv4_node_delete ( struct fw_card *card, struct fw_device *device ) {
-       struct net_device *netdev;
-       struct ipv4_priv *priv;
-       struct ipv4_node *node;
+/* FIXME */
+static void fwnet_peer_delete(struct fw_card *card, struct fw_device *device)
+{
+       struct net_device *net;
+       struct fwnet_device *dev;
+       struct fwnet_peer *peer;
        u64 guid;
        unsigned long flags;
-       struct ipv4_partial_datagram *pd, *pd_next;
+       struct fwnet_partial_datagram *pd, *pd_next;
 
        guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
-       netdev = card->netdev;
-       if ( netdev )
-               priv = netdev_priv ( netdev );
+       net = card->netdev;
+       if (net)
+               dev = netdev_priv(net);
        else
-               priv = NULL;
-       if ( priv )
-               spin_lock_irqsave ( &priv->lock, flags );
-       list_for_each_entry( node, &card->ipv4_nodes, ipv4_nodes ) {
-               if ( node->guid == guid ) {
-                       list_del ( &node->ipv4_nodes );
-                       list_for_each_entry_safe( pd, pd_next, &node->pdg_list, pdg_list )
-                               ipv4_pd_delete ( pd );
+               dev = NULL;
+       if (dev)
+               spin_lock_irqsave(&dev->lock, flags);
+
+       list_for_each_entry(peer, &card->peer_list, peer_link) {
+               if (peer->guid == guid) {
+                       list_del(&peer->peer_link);
+                       list_for_each_entry_safe(pd, pd_next, &peer->pd_list,
+                                                pd_link)
+                               fwnet_pd_delete(pd);
                        break;
                }
        }
-       if ( priv )
-               spin_unlock_irqrestore ( &priv->lock, flags );
+       if (dev)
+               spin_unlock_irqrestore(&dev->lock, flags);
 }
 
-/* ------------------------------------------------------------------ */
-
-
-static int ipv4_finish_incoming_packet ( struct net_device *netdev,
- struct sk_buff *skb, u16 source_node_id, bool is_broadcast, u16 ether_type ) {
-       struct ipv4_priv *priv;
-       static u64 broadcast_hw = ~0ULL;
+static int fwnet_finish_incoming_packet(struct net_device *net,
+                                       struct sk_buff *skb, u16 source_node_id,
+                                       bool is_broadcast, u16 ether_type)
+{
+       struct fwnet_device *dev;
+       static const __be64 broadcast_hw = cpu_to_be64(~0ULL);
        int status;
-       u64 guid;
+       __be64 guid;
 
-       fw_debug ( "ipv4_finish_incoming_packet(%p, %p, %x, %s, %x\n",
- netdev, skb, source_node_id, is_broadcast ? "true" : "false", ether_type );
-       priv = netdev_priv(netdev);
+       dev = netdev_priv(net);
        /* Write metadata, and then pass to the receive level */
-       skb->dev = netdev;
+       skb->dev = net;
        skb->ip_summed = CHECKSUM_UNNECESSARY;  /* don't check it */
 
        /*
@@ -724,73 +601,75 @@ static int ipv4_finish_incoming_packet ( struct net_device *netdev,
         * about the sending machine.
         */
        if (ether_type == ETH_P_ARP) {
-               struct ipv4_arp *arp1394;
+               struct rfc2734_arp *arp1394;
                struct arphdr *arp;
                unsigned char *arp_ptr;
                u64 fifo_addr;
+               u64 peer_guid;
                u8 max_rec;
                u8 sspd;
                u16 max_payload;
-               struct ipv4_node *node;
-               static const u16 ipv4_speed_to_max_payload[] = {
+               struct fwnet_peer *peer;
+               static const u16 fwnet_speed_to_max_payload[] = {
                        /* S100, S200, S400, S800, S1600, S3200 */
                            512, 1024, 2048, 4096,  4096,  4096
                };
 
-               /* fw_debug ( "ARP packet\n" ); */
-               arp1394 = (struct ipv4_arp *)skb->data;
+               arp1394 = (struct rfc2734_arp *)skb->data;
                arp = (struct arphdr *)skb->data;
                arp_ptr = (unsigned char *)(arp + 1);
-               fifo_addr = (u64)ntohs(arp1394->fifo_hi) << 32 |
- ntohl(arp1394->fifo_lo);
-               max_rec = priv->card->max_receive;
-               if ( arp1394->max_rec < max_rec )
+               fifo_addr = (u64)ntohs(arp1394->fifo_hi) << 32
                              | ntohl(arp1394->fifo_lo);
+               max_rec = dev->card->max_receive;
+               if (arp1394->max_rec < max_rec)
                        max_rec = arp1394->max_rec;
                sspd = arp1394->sspd;
-               /*
-                * Sanity check. MacOSX seems to be sending us 131 in this
-                * field (atleast on my Panther G5). Not sure why.
-                */
-               if (sspd > 5 ) {
-                       fw_notify ( "sspd %x out of range\n", sspd );
+               /* Sanity check.  OS X 10.3 PPC reportedly sends 131. */
+               if (sspd > SCODE_3200) {
+                       fw_notify("sspd %x out of range\n", sspd);
                        sspd = 0;
                }
 
-               max_payload = min(ipv4_speed_to_max_payload[sspd],
(u16)(1 << (max_rec + 1))) - IPV4_UNFRAG_HDR_SIZE;
+               max_payload = min(fwnet_speed_to_max_payload[sspd],
                      (u16)(1 << (max_rec + 1))) - RFC2374_UNFRAG_HDR_SIZE;
 
-               guid = be64_to_cpu(get_unaligned(&arp1394->s_uniq_id));
-               node = ipv4_node_find_by_guid(priv, guid);
-               if (!node) {
-                       fw_notify ( "No node for ARP packet from %llx\n", guid );
+               peer_guid = get_unaligned_be64(&arp1394->s_uniq_id);
+               peer = fwnet_peer_find_by_guid(dev, peer_guid);
+               if (!peer) {
+                       fw_notify("No peer for ARP packet from %016llx\n",
+                                 (unsigned long long)peer_guid);
                        goto failed_proto;
                }
-               if ( node->nodeid != source_node_id || node->generation != priv->card->generation ) {
-                       fw_notify ( "Internal error: node->nodeid (%x) != soucre_node_id (%x) or node->generation (%x) != priv->card->generation(%x)\n",
- node->nodeid, source_node_id, node->generation, priv->card->generation );
-                       node->nodeid = source_node_id;
-                       node->generation = priv->card->generation;
+
+               /* FIXME don't use card->generation */
+               if (peer->node_id != source_node_id ||
+                   peer->generation != dev->card->generation) {
+                       fw_notify("Internal error: peer->node_id (%x) != "
+                                 "source_node_id (%x) or peer->generation (%x)"
+                                 " != dev->card->generation(%x)\n",
+                                 peer->node_id, source_node_id,
+                                 peer->generation, dev->card->generation);
+                       peer->node_id = source_node_id;
+                       peer->generation = dev->card->generation;
                }
 
                /* FIXME: for debugging */
-               if ( sspd > SCODE_400 )
+               if (sspd > SCODE_400)
                        sspd = SCODE_400;
                /* Update our speed/payload/fifo_offset table */
                /*
                 * FIXME: this does not handle cases where two high-speed endpoints must use a slower speed because of
                 * a lower speed hub between them.  We need to look at the actual topology map here.
                 */
-               fw_debug ( "Setting node %p fifo %llx (was %llx), max_payload %x (was %x), speed %x (was %x)\n",
- node, fifo_addr, node->fifo, max_payload, node->max_payload, sspd, node->xmt_speed );
-               node->fifo =    fifo_addr;
-               node->max_payload = max_payload;
+               peer->fifo = fifo_addr;
+               peer->max_payload = max_payload;
                /*
                 * Only allow speeds to go down from their initial value.
-                * Otherwise a local node that can only do S400 or slower may
-                * be told to transmit at S800 to a faster remote node.
+                * Otherwise a local peer that can only do S400 or slower may
+                * be told to transmit at S800 to a faster remote peer.
                 */
-               if ( node->xmt_speed > sspd )
-                       node->xmt_speed = sspd;
+               if (peer->xmt_speed > sspd)
+                       peer->xmt_speed = sspd;
 
                /*
                 * Now that we're done with the 1394 specific stuff, we'll
@@ -805,248 +684,257 @@ static int ipv4_finish_incoming_packet ( struct net_device *netdev,
                 */
 
                arp->ar_hln = 8;
-               arp_ptr += arp->ar_hln;         /* skip over sender unique id */
-               *(u32 *)arp_ptr = arp1394->sip; /* move sender IP addr */
-               arp_ptr += arp->ar_pln;         /* skip over sender IP addr */
+               /* skip over sender unique id */
+               arp_ptr += arp->ar_hln;
+               /* move sender IP addr */
+               put_unaligned(arp1394->sip, (u32 *)arp_ptr);
+               /* skip over sender IP addr */
+               arp_ptr += arp->ar_pln;
 
                if (arp->ar_op == htons(ARPOP_REQUEST))
                        memset(arp_ptr, 0, sizeof(u64));
                else
-                       memcpy(arp_ptr, netdev->dev_addr, sizeof(u64));
+                       memcpy(arp_ptr, net->dev_addr, sizeof(u64));
        }
 
        /* Now add the ethernet header. */
-       guid = cpu_to_be64(priv->card->guid);
-       if (dev_hard_header(skb, netdev, ether_type, is_broadcast ? &broadcast_hw : &guid, NULL,
- skb->len) >= 0) {
-               struct ipv4_ether_hdr *eth;
+       guid = cpu_to_be64(dev->card->guid);
+       if (dev_hard_header(skb, net, ether_type,
+                          is_broadcast ? &broadcast_hw : &guid,
+                          NULL, skb->len) >= 0) {
+               struct fwnet_header *eth;
                u16 *rawp;
                __be16 protocol;
 
                skb_reset_mac_header(skb);
                skb_pull(skb, sizeof(*eth));
-               eth = ipv4_ether_hdr(skb);
+               eth = (struct fwnet_header *)skb_mac_header(skb);
                if (*eth->h_dest & 1) {
-                       if (memcmp(eth->h_dest, netdev->broadcast, netdev->addr_len) == 0) {
-                               fw_debug ( "Broadcast\n" );
+                       if (memcmp(eth->h_dest, net->broadcast,
+                                  net->addr_len) == 0)
                                skb->pkt_type = PACKET_BROADCAST;
-                       }
 #if 0
                        else
                                skb->pkt_type = PACKET_MULTICAST;
 #endif
                } else {
-                       if (memcmp(eth->h_dest, netdev->dev_addr, netdev->addr_len)) {
+                       if (memcmp(eth->h_dest, net->dev_addr, net->addr_len)) {
                                u64 a1, a2;
 
-                               memcpy ( &a1, eth->h_dest, sizeof(u64));
-                               memcpy ( &a2, netdev->dev_addr, sizeof(u64));
-                               fw_debug ( "Otherhost %llx %llx %x\n", a1, a2, netdev->addr_len );
+                               memcpy(&a1, eth->h_dest, sizeof(u64));
+                               memcpy(&a2, net->dev_addr, sizeof(u64));
                                skb->pkt_type = PACKET_OTHERHOST;
                        }
                }
                if (ntohs(eth->h_proto) >= 1536) {
-                       fw_debug ( " proto %x %x\n", eth->h_proto, ntohs(eth->h_proto) );
                        protocol = eth->h_proto;
                } else {
                        rawp = (u16 *)skb->data;
-                       if (*rawp == 0xFFFF) {
-                               fw_debug ( "proto 802_3\n" );
+                       if (*rawp == 0xffff)
                                protocol = htons(ETH_P_802_3);
-                       } else {
-                               fw_debug ( "proto 802_2\n" );
+                       else
                                protocol = htons(ETH_P_802_2);
-                       }
                }
                skb->protocol = protocol;
        }
        status = netif_rx(skb);
-       if ( status == NET_RX_DROP) {
-               netdev->stats.rx_errors++;
-               netdev->stats.rx_dropped++;
+       if (status == NET_RX_DROP) {
+               net->stats.rx_errors++;
+               net->stats.rx_dropped++;
        } else {
-               netdev->stats.rx_packets++;
-               netdev->stats.rx_bytes += skb->len;
+               net->stats.rx_packets++;
+               net->stats.rx_bytes += skb->len;
        }
-       if (netif_queue_stopped(netdev))
-               netif_wake_queue(netdev);
+       if (netif_queue_stopped(net))
+               netif_wake_queue(net);
+
        return 0;
 
  failed_proto:
-       netdev->stats.rx_errors++;
-       netdev->stats.rx_dropped++;
+       net->stats.rx_errors++;
+       net->stats.rx_dropped++;
+
        dev_kfree_skb_any(skb);
-       if (netif_queue_stopped(netdev))
-               netif_wake_queue(netdev);
-       netdev->last_rx = jiffies;
+       if (netif_queue_stopped(net))
+               netif_wake_queue(net);
+
+       net->last_rx = jiffies;
+
        return 0;
 }
 
-/* ------------------------------------------------------------------ */
-
-static int ipv4_incoming_packet ( struct ipv4_priv *priv, u32 *buf, int len, u16 source_node_id, bool is_broadcast ) {
+static int fwnet_incoming_packet(struct fwnet_device *dev, __be32 *buf, int len,
+                                u16 source_node_id, bool is_broadcast)
+{
        struct sk_buff *skb;
-       struct net_device *netdev;
-       struct ipv4_hdr hdr;
+       struct net_device *net;
+       struct rfc2734_header hdr;
        unsigned lf;
        unsigned long flags;
-       struct ipv4_node *node;
-       struct ipv4_partial_datagram *pd;
+       struct fwnet_peer *peer;
+       struct fwnet_partial_datagram *pd;
        int fg_off;
        int dg_size;
        u16 datagram_label;
        int retval;
        u16 ether_type;
 
-       fw_debug ( "ipv4_incoming_packet(%p, %p, %d, %x, %s)\n", priv, buf, len, source_node_id, is_broadcast ? "true" : "false" );
-       netdev = priv->card->netdev;
+       net = dev->card->netdev;
 
-       hdr.w0 = ntohl(buf[0]);
-       lf = ipv4_get_hdr_lf(&hdr);
-       if ( lf == IPV4_HDR_UNFRAG ) {
+       hdr.w0 = be32_to_cpu(buf[0]);
+       lf = fwnet_get_hdr_lf(&hdr);
+       if (lf == RFC2374_HDR_UNFRAG) {
                /*
                 * An unfragmented datagram has been received by the ieee1394
                 * bus. Build an skbuff around it so we can pass it to the
                 * high level network layer.
                 */
-               ether_type = ipv4_get_hdr_ether_type(&hdr);
-               fw_debug ( "header w0 = %x, lf = %x, ether_type = %x\n", hdr.w0, lf, ether_type );
+               ether_type = fwnet_get_hdr_ether_type(&hdr);
                buf++;
-               len -= IPV4_UNFRAG_HDR_SIZE;
+               len -= RFC2374_UNFRAG_HDR_SIZE;
 
-               skb = dev_alloc_skb(len + netdev->hard_header_len + 15);
+               skb = dev_alloc_skb(len + net->hard_header_len + 15);
                if (unlikely(!skb)) {
-                       fw_error ( "Out of memory for incoming packet\n");
-                       netdev->stats.rx_dropped++;
+                       fw_error("out of memory\n");
+                       net->stats.rx_dropped++;
+
                        return -1;
                }
-               skb_reserve(skb, (netdev->hard_header_len + 15) & ~15);
-               memcpy(skb_put(skb, len), buf, len );
-               return ipv4_finish_incoming_packet(netdev, skb, source_node_id, is_broadcast, ether_type );
+               skb_reserve(skb, (net->hard_header_len + 15) & ~15);
+               memcpy(skb_put(skb, len), buf, len);
+
+               return fwnet_finish_incoming_packet(net, skb, source_node_id,
+                                                   is_broadcast, ether_type);
        }
        /* A datagram fragment has been received, now the fun begins. */
        hdr.w1 = ntohl(buf[1]);
-       buf +=2;
-       len -= IPV4_FRAG_HDR_SIZE;
-       if ( lf ==IPV4_HDR_FIRSTFRAG ) {
-               ether_type = ipv4_get_hdr_ether_type(&hdr);
+       buf += 2;
+       len -= RFC2374_FRAG_HDR_SIZE;
+       if (lf == RFC2374_HDR_FIRSTFRAG) {
+               ether_type = fwnet_get_hdr_ether_type(&hdr);
                fg_off = 0;
        } else {
-               fg_off = ipv4_get_hdr_fg_off(&hdr);
-               ether_type = 0; /* Shut up compiler! */
+               ether_type = 0;
+               fg_off = fwnet_get_hdr_fg_off(&hdr);
        }
-       datagram_label = ipv4_get_hdr_dgl(&hdr);
-       dg_size = ipv4_get_hdr_dg_size(&hdr); /* ??? + 1 */
-       fw_debug ( "fragmented: %x.%x = lf %x, ether_type %x, fg_off %x, dgl %x, dg_size %x\n", hdr.w0, hdr.w1, lf, ether_type, fg_off, datagram_label, dg_size );
-       node = ipv4_node_find_by_nodeid ( priv, source_node_id);
-       spin_lock_irqsave(&node->pdg_lock, flags);
-       pd = ipv4_pd_find( node, datagram_label );
+       datagram_label = fwnet_get_hdr_dgl(&hdr);
+       dg_size = fwnet_get_hdr_dg_size(&hdr); /* ??? + 1 */
+       peer = fwnet_peer_find_by_node_id(dev, source_node_id);
+
+       spin_lock_irqsave(&peer->pdg_lock, flags);
+
+       pd = fwnet_pd_find(peer, datagram_label);
        if (pd == NULL) {
-               while ( node->pdg_size >= ipv4_mpd ) {
+               while (peer->pdg_size >= FWNET_MAX_FRAGMENTS) {
                        /* remove the oldest */
-                       ipv4_pd_delete ( list_first_entry(&node->pdg_list, struct ipv4_partial_datagram, pdg_list) );
-                       node->pdg_size--;
+                       fwnet_pd_delete(list_first_entry(&peer->pd_list,
+                               struct fwnet_partial_datagram, pd_link));
+                       peer->pdg_size--;
                }
-               pd = ipv4_pd_new ( netdev, node, datagram_label, dg_size,
- buf, fg_off, len);
-               if ( pd == NULL) {
+               pd = fwnet_pd_new(net, peer, datagram_label,
                                dg_size, buf, fg_off, len);
+               if (pd == NULL) {
                        retval = -ENOMEM;
                        goto bad_proto;
                }
-               node->pdg_size++;
+               peer->pdg_size++;
        } else {
-               if (ipv4_frag_overlap(pd, fg_off, len) || pd->datagram_size != dg_size) {
+               if (fwnet_frag_overlap(pd, fg_off, len) ||
+                   pd->datagram_size != dg_size) {
                        /*
                         * Differing datagram sizes or overlapping fragments,
-                        * Either way the remote machine is playing silly buggers
-                        * with us: obliterate the old datagram and start a new one.
+                        * discard old datagram and start a new one.
                         */
-                       ipv4_pd_delete ( pd );
-                       pd = ipv4_pd_new ( netdev, node, datagram_label,
- dg_size, buf, fg_off, len);
-                       if ( pd == NULL ) {
+                       fwnet_pd_delete(pd);
+                       pd = fwnet_pd_new(net, peer, datagram_label,
                                        dg_size, buf, fg_off, len);
+                       if (pd == NULL) {
                                retval = -ENOMEM;
-                               node->pdg_size--;
+                               peer->pdg_size--;
                                goto bad_proto;
                        }
                } else {
-                       bool worked;
-
-                       worked = ipv4_pd_update ( node, pd,
- buf, fg_off, len );
-                       if ( ! worked ) {
+                       if (!fwnet_pd_update(peer, pd, buf, fg_off, len)) {
                                /*
                                 * Couldn't save off fragment anyway
                                 * so might as well obliterate the
                                 * datagram now.
                                 */
-                               ipv4_pd_delete ( pd );
-                               node->pdg_size--;
+                               fwnet_pd_delete(pd);
+                               peer->pdg_size--;
                                goto bad_proto;
                        }
                }
        } /* new datagram or add to existing one */
 
-       if ( lf == IPV4_HDR_FIRSTFRAG )
+       if (lf == RFC2374_HDR_FIRSTFRAG)
                pd->ether_type = ether_type;
-       if ( ipv4_pd_is_complete ( pd ) ) {
+
+       if (fwnet_pd_is_complete(pd)) {
                ether_type = pd->ether_type;
-               node->pdg_size--;
+               peer->pdg_size--;
                skb = skb_get(pd->skb);
-               ipv4_pd_delete ( pd );
-               spin_unlock_irqrestore(&node->pdg_lock, flags);
-               return ipv4_finish_incoming_packet ( netdev, skb, source_node_id, false, ether_type );
+               fwnet_pd_delete(pd);
+
+               spin_unlock_irqrestore(&peer->pdg_lock, flags);
+
+               return fwnet_finish_incoming_packet(net, skb, source_node_id,
+                                                   false, ether_type);
        }
        /*
         * Datagram is not complete, we're done for the
         * moment.
         */
-       spin_unlock_irqrestore(&node->pdg_lock, flags);
+       spin_unlock_irqrestore(&peer->pdg_lock, flags);
+
        return 0;
 
  bad_proto:
-       spin_unlock_irqrestore(&node->pdg_lock, flags);
-       if (netif_queue_stopped(netdev))
-               netif_wake_queue(netdev);
+       spin_unlock_irqrestore(&peer->pdg_lock, flags);
+
+       if (netif_queue_stopped(net))
+               netif_wake_queue(net);
+
        return 0;
 }
 
-static void ipv4_receive_packet ( struct fw_card *card, struct fw_request *r,
- int tcode, int destination, int source, int generation, int speed,
- unsigned long long offset, void *payload, size_t length, void *callback_data ) {
-       struct ipv4_priv *priv;
+static void fwnet_receive_packet(struct fw_card *card, struct fw_request *r,
+               int tcode, int destination, int source, int generation,
+               int speed, unsigned long long offset, void *payload,
+               size_t length, void *callback_data)
+{
+       struct fwnet_device *dev;
        int status;
 
-       fw_debug ( "ipv4_receive_packet(%p,%p,%x,%x,%x,%x,%x,%llx,%p,%lx,%p)\n",
- card, r, tcode, destination, source, generation, speed, offset, payload,
- (unsigned long)length, callback_data);
-       print_hex_dump ( KERN_DEBUG, "header: ", DUMP_PREFIX_OFFSET, 32, 1, payload, length, false );
-       priv = callback_data;
-       if (   tcode != TCODE_WRITE_BLOCK_REQUEST
-           || destination != card->node_id
-           || generation != card->generation
-           || offset != priv->handler.offset ) {
+       dev = callback_data;
+       if (tcode != TCODE_WRITE_BLOCK_REQUEST
+           || destination != card->node_id     /* <- FIXME */
+           || generation != card->generation   /* <- FIXME */
+           || offset != dev->handler.offset) {
                fw_send_response(card, r, RCODE_CONFLICT_ERROR);
-               fw_debug("Conflict error card node_id=%x, card generation=%x, local offset %llx\n",
- card->node_id, card->generation, (unsigned long long)priv->handler.offset );
+
                return;
        }
-       status = ipv4_incoming_packet ( priv, payload, length, source, false );
-       if ( status != 0 ) {
-               fw_error ( "Incoming packet failure\n" );
-               fw_send_response ( card, r, RCODE_CONFLICT_ERROR );
+
+       status = fwnet_incoming_packet(dev, payload, length, source, false);
+       if (status != 0) {
+               fw_error("Incoming packet failure\n");
+               fw_send_response(card, r, RCODE_CONFLICT_ERROR);
+
                return;
        }
-       fw_send_response ( card, r, RCODE_COMPLETE );
+
+       fw_send_response(card, r, RCODE_COMPLETE);
 }
 
-static void ipv4_receive_broadcast(struct fw_iso_context *context, u32 cycle,
- size_t header_length, void *header, void *data) {
-       struct ipv4_priv *priv;
+static void fwnet_receive_broadcast(struct fw_iso_context *context,
+               u32 cycle, size_t header_length, void *header, void *data)
+{
+       struct fwnet_device *dev;
        struct fw_iso_packet packet;
        struct fw_card *card;
-       u16 *hdr_ptr;
-       u32 *buf_ptr;
+       __be16 *hdr_ptr;
+       __be32 *buf_ptr;
        int retval;
        u32 length;
        u16 source_node_id;
@@ -1055,70 +943,68 @@ static void ipv4_receive_broadcast(struct fw_iso_context *context, u32 cycle,
        unsigned long offset;
        unsigned long flags;
 
-       fw_debug ( "ipv4_receive_broadcast ( context=%p, cycle=%x, header_length=%lx, header=%p, data=%p )\n", context, cycle, (unsigned long)header_length, header, data );
-       print_hex_dump ( KERN_DEBUG, "header: ", DUMP_PREFIX_OFFSET, 32, 1, header, header_length, false );
-       priv = data;
-       card = priv->card;
+       dev = data;
+       card = dev->card;
        hdr_ptr = header;
-       length = ntohs(hdr_ptr[0]);
-       spin_lock_irqsave(&priv->lock,flags);
-       offset = priv->rcv_buffer_size * priv->broadcast_rcv_next_ptr;
-       buf_ptr = priv->broadcast_rcv_buffer_ptrs[priv->broadcast_rcv_next_ptr++];
-       if ( priv->broadcast_rcv_next_ptr == priv->num_broadcast_rcv_ptrs )
-               priv->broadcast_rcv_next_ptr = 0;
-       spin_unlock_irqrestore(&priv->lock,flags);
-       fw_debug ( "length %u at %p\n", length, buf_ptr );
-       print_hex_dump ( KERN_DEBUG, "buffer: ", DUMP_PREFIX_OFFSET, 32, 1, buf_ptr, length, false );
+       length = be16_to_cpup(hdr_ptr);
+
+       spin_lock_irqsave(&dev->lock, flags);
+
+       offset = dev->rcv_buffer_size * dev->broadcast_rcv_next_ptr;
+       buf_ptr = dev->broadcast_rcv_buffer_ptrs[dev->broadcast_rcv_next_ptr++];
+       if (dev->broadcast_rcv_next_ptr == dev->num_broadcast_rcv_ptrs)
+               dev->broadcast_rcv_next_ptr = 0;
+
+       spin_unlock_irqrestore(&dev->lock, flags);
 
        specifier_id =    (be32_to_cpu(buf_ptr[0]) & 0xffff) << 8
                        | (be32_to_cpu(buf_ptr[1]) & 0xff000000) >> 24;
-       ver = be32_to_cpu(buf_ptr[1]) & 0xFFFFFF;
+       ver = be32_to_cpu(buf_ptr[1]) & 0xffffff;
        source_node_id = be32_to_cpu(buf_ptr[0]) >> 16;
-       /* fw_debug ( "source %x SpecID %x ver %x\n", source_node_id, specifier_id, ver ); */
-       if ( specifier_id == IPV4_GASP_SPECIFIER_ID && ver == IPV4_GASP_VERSION ) {
+
+       if (specifier_id == IANA_SPECIFIER_ID && ver == RFC2734_SW_VERSION) {
                buf_ptr += 2;
-               length -= IPV4_GASP_OVERHEAD;
-               ipv4_incoming_packet(priv, buf_ptr, length, source_node_id, true);
-       } else
-               fw_debug ( "Ignoring packet: not GASP\n" );
-       packet.payload_length = priv->rcv_buffer_size;
+               length -= IEEE1394_GASP_HDR_SIZE;
+               fwnet_incoming_packet(dev, buf_ptr, length,
+                                     source_node_id, true);
+       }
+
+       packet.payload_length = dev->rcv_buffer_size;
        packet.interrupt = 1;
        packet.skip = 0;
        packet.tag = 3;
        packet.sy = 0;
-       packet.header_length = IPV4_GASP_OVERHEAD;
-       spin_lock_irqsave(&priv->lock,flags);
-       retval = fw_iso_context_queue ( priv->broadcast_rcv_context, &packet,
- &priv->broadcast_rcv_buffer, offset );
-       spin_unlock_irqrestore(&priv->lock,flags);
-       if ( retval < 0 )
-               fw_error ( "requeue failed\n" );
-}
+       packet.header_length = IEEE1394_GASP_HDR_SIZE;
+
+       spin_lock_irqsave(&dev->lock, flags);
 
-static void debug_ptask ( struct ipv4_packet_task *ptask ) {
-       static const char *tx_types[] = { "Unknown", "GASP", "Write" };
-
-       fw_debug ( "packet %p { hdr { w0 %x w1 %x }, skb %p, priv %p,"
- " tx_type %s, outstanding_pkts %d, max_payload %x, fifo %llx,"
- " speed %x, dest_node %x, generation %x }\n",
- ptask, ptask->hdr.w0, ptask->hdr.w1, ptask->skb, ptask->priv,
- ptask->tx_type > IPV4_WRREQ ? "Invalid" : tx_types[ptask->tx_type],
- ptask->outstanding_pkts,  ptask->max_payload,
- ptask->fifo_addr, ptask->speed, ptask->dest_node, ptask->generation );
-       print_hex_dump ( KERN_DEBUG, "packet :", DUMP_PREFIX_OFFSET, 32, 1,
- ptask->skb->data, ptask->skb->len, false );
+       retval = fw_iso_context_queue(dev->broadcast_rcv_context, &packet,
+                                     &dev->broadcast_rcv_buffer, offset);
+
+       spin_unlock_irqrestore(&dev->lock, flags);
+
+       if (retval < 0)
+               fw_error("requeue failed\n");
 }
 
-static void ipv4_transmit_packet_done ( struct ipv4_packet_task *ptask ) {
-       struct ipv4_priv *priv;
+static struct kmem_cache *fwnet_packet_task_cache;
+
+static int fwnet_send_packet(struct fwnet_packet_task *ptask);
+
+static void fwnet_transmit_packet_done(struct fwnet_packet_task *ptask)
+{
+       struct fwnet_device *dev;
        unsigned long flags;
 
-       priv = ptask->priv;
-       spin_lock_irqsave ( &priv->lock, flags );
-       list_del ( &ptask->packet_list );
-       spin_unlock_irqrestore ( &priv->lock, flags );
-       ptask->outstanding_pkts--;
-       if ( ptask->outstanding_pkts > 0 ) {
+       dev = ptask->dev;
+
+       spin_lock_irqsave(&dev->lock, flags);
+       list_del(&ptask->pt_link);
+       spin_unlock_irqrestore(&dev->lock, flags);
+
+       ptask->outstanding_pkts--; /* FIXME access inside lock */
+
+       if (ptask->outstanding_pkts > 0) {
                u16 dg_size;
                u16 fg_off;
                u16 datagram_label;
@@ -1126,133 +1012,139 @@ static void ipv4_transmit_packet_done ( struct ipv4_packet_task *ptask ) {
                struct sk_buff *skb;
 
                /* Update the ptask to point to the next fragment and send it */
-               lf = ipv4_get_hdr_lf(&ptask->hdr);
+               lf = fwnet_get_hdr_lf(&ptask->hdr);
                switch (lf) {
-               case IPV4_HDR_LASTFRAG:
-               case IPV4_HDR_UNFRAG:
+               case RFC2374_HDR_LASTFRAG:
+               case RFC2374_HDR_UNFRAG:
                default:
-                       fw_error ( "Outstanding packet %x lf %x, header %x,%x\n", ptask->outstanding_pkts, lf, ptask->hdr.w0, ptask->hdr.w1 );
+                       fw_error("Outstanding packet %x lf %x, header %x,%x\n",
+                                ptask->outstanding_pkts, lf, ptask->hdr.w0,
+                                ptask->hdr.w1);
                        BUG();
 
-               case IPV4_HDR_FIRSTFRAG:
+               case RFC2374_HDR_FIRSTFRAG:
                        /* Set frag type here for future interior fragments */
-                       dg_size = ipv4_get_hdr_dg_size(&ptask->hdr);
-                       fg_off = ptask->max_payload - IPV4_FRAG_HDR_SIZE;
-                       datagram_label = ipv4_get_hdr_dgl(&ptask->hdr);
+                       dg_size = fwnet_get_hdr_dg_size(&ptask->hdr);
+                       fg_off = ptask->max_payload - RFC2374_FRAG_HDR_SIZE;
+                       datagram_label = fwnet_get_hdr_dgl(&ptask->hdr);
                        break;
 
-               case IPV4_HDR_INTFRAG:
-                       dg_size = ipv4_get_hdr_dg_size(&ptask->hdr);
-                       fg_off = ipv4_get_hdr_fg_off(&ptask->hdr) + ptask->max_payload - IPV4_FRAG_HDR_SIZE;
-                       datagram_label = ipv4_get_hdr_dgl(&ptask->hdr);
+               case RFC2374_HDR_INTFRAG:
+                       dg_size = fwnet_get_hdr_dg_size(&ptask->hdr);
+                       fg_off = fwnet_get_hdr_fg_off(&ptask->hdr)
+                                 + ptask->max_payload - RFC2374_FRAG_HDR_SIZE;
+                       datagram_label = fwnet_get_hdr_dgl(&ptask->hdr);
                        break;
                }
                skb = ptask->skb;
-               skb_pull ( skb, ptask->max_payload );
-               if ( ptask->outstanding_pkts > 1 ) {
-                       ipv4_make_sf_hdr ( &ptask->hdr,
-  IPV4_HDR_INTFRAG, dg_size, fg_off, datagram_label );
+               skb_pull(skb, ptask->max_payload);
+               if (ptask->outstanding_pkts > 1) {
+                       fwnet_make_sf_hdr(&ptask->hdr, RFC2374_HDR_INTFRAG,
+                                         dg_size, fg_off, datagram_label);
                } else {
-                       ipv4_make_sf_hdr ( &ptask->hdr,
-  IPV4_HDR_LASTFRAG, dg_size, fg_off, datagram_label );
-                       ptask->max_payload = skb->len + IPV4_FRAG_HDR_SIZE;
-
+                       fwnet_make_sf_hdr(&ptask->hdr, RFC2374_HDR_LASTFRAG,
+                                         dg_size, fg_off, datagram_label);
+                       ptask->max_payload = skb->len + RFC2374_FRAG_HDR_SIZE;
                }
-               ipv4_send_packet ( ptask );
+               fwnet_send_packet(ptask);
        } else {
-               dev_kfree_skb_any ( ptask->skb );
-               kmem_cache_free( ipv4_packet_task_cache, ptask );
+               dev_kfree_skb_any(ptask->skb);
+               kmem_cache_free(fwnet_packet_task_cache, ptask);
        }
 }
 
-static void ipv4_write_complete ( struct fw_card *card, int rcode,
- void *payload, size_t length, void *data ) {
-       struct ipv4_packet_task *ptask;
+static void fwnet_write_complete(struct fw_card *card, int rcode,
+                                void *payload, size_t length, void *data)
+{
+       struct fwnet_packet_task *ptask;
 
        ptask = data;
-       fw_debug ( "ipv4_write_complete ( %p, %x, %p, %lx, %p )\n",
- card, rcode, payload, (unsigned long)length, data );
-       debug_ptask ( ptask );
 
-       if ( rcode == RCODE_COMPLETE ) {
-               ipv4_transmit_packet_done ( ptask );
-       } else {
-               fw_error ( "ipv4_write_complete: failed: %x\n", rcode );
+       if (rcode == RCODE_COMPLETE)
+               fwnet_transmit_packet_done(ptask);
+       else
+               fw_error("fwnet_write_complete: failed: %x\n", rcode);
                /* ??? error recovery */
-       }
 }
 
-static int ipv4_send_packet ( struct ipv4_packet_task *ptask ) {
-       struct ipv4_priv *priv;
+static int fwnet_send_packet(struct fwnet_packet_task *ptask)
+{
+       struct fwnet_device *dev;
        unsigned tx_len;
-       struct ipv4_hdr *bufhdr;
+       struct rfc2734_header *bufhdr;
        unsigned long flags;
-       struct net_device *netdev;
-#if 0 /* stefanr */
-       int retval;
-#endif
+       struct net_device *net;
 
-       fw_debug ( "ipv4_send_packet\n" );
-       debug_ptask ( ptask );
-       priv = ptask->priv;
+       dev = ptask->dev;
        tx_len = ptask->max_payload;
-       switch (ipv4_get_hdr_lf(&ptask->hdr)) {
-       case IPV4_HDR_UNFRAG:
-               bufhdr = (struct ipv4_hdr *)skb_push(ptask->skb, IPV4_UNFRAG_HDR_SIZE);
-               bufhdr->w0 = htonl(ptask->hdr.w0);
+       switch (fwnet_get_hdr_lf(&ptask->hdr)) {
+       case RFC2374_HDR_UNFRAG:
+               bufhdr = (struct rfc2734_header *)
+                               skb_push(ptask->skb, RFC2374_UNFRAG_HDR_SIZE);
+               put_unaligned_be32(ptask->hdr.w0, &bufhdr->w0);
                break;
 
-       case IPV4_HDR_FIRSTFRAG:
-       case IPV4_HDR_INTFRAG:
-       case IPV4_HDR_LASTFRAG:
-               bufhdr = (struct ipv4_hdr *)skb_push(ptask->skb, IPV4_FRAG_HDR_SIZE);
-               bufhdr->w0 = htonl(ptask->hdr.w0);
-               bufhdr->w1 = htonl(ptask->hdr.w1);
+       case RFC2374_HDR_FIRSTFRAG:
+       case RFC2374_HDR_INTFRAG:
+       case RFC2374_HDR_LASTFRAG:
+               bufhdr = (struct rfc2734_header *)
+                               skb_push(ptask->skb, RFC2374_FRAG_HDR_SIZE);
+               put_unaligned_be32(ptask->hdr.w0, &bufhdr->w0);
+               put_unaligned_be32(ptask->hdr.w1, &bufhdr->w1);
                break;
 
        default:
                BUG();
        }
-       if ( ptask->tx_type == IPV4_GASP ) {
-               u32 *packets;
+       if (ptask->dest_node == IEEE1394_ALL_NODES) {
+               u8 *p;
                int generation;
-               int nodeid;
+               int node_id;
 
                /* ptask->generation may not have been set yet */
-               generation = priv->card->generation;
+               generation = dev->card->generation;
                smp_rmb();
-               nodeid = priv->card->node_id;
-               packets = (u32 *)skb_push(ptask->skb, sizeof(u32)*2);
-               packets[0] = htonl(nodeid << 16 | (IPV4_GASP_SPECIFIER_ID>>8));
-               packets[1] = htonl((IPV4_GASP_SPECIFIER_ID & 0xFF) << 24 | IPV4_GASP_VERSION);
-               fw_send_request ( priv->card, &ptask->transaction, TCODE_STREAM_DATA,
- fw_stream_packet_destination_id(3, BROADCAST_CHANNEL, 0),
- generation, SCODE_100, 0ULL, ptask->skb->data, tx_len + 8, ipv4_write_complete, ptask );
-               spin_lock_irqsave(&priv->lock,flags);
-               list_add_tail ( &ptask->packet_list, &priv->broadcasted_list );
-               spin_unlock_irqrestore(&priv->lock,flags);
-#if 0 /* stefanr */
-               return retval;
-#else
+               node_id = dev->card->node_id;
+
+               p = skb_push(ptask->skb, 8);
+               put_unaligned_be32(node_id << 16 | IANA_SPECIFIER_ID >> 8, p);
+               put_unaligned_be32((IANA_SPECIFIER_ID & 0xff) << 24
+                                               | RFC2734_SW_VERSION, &p[4]);
+
+               /* We should not transmit if broadcast_channel.valid == 0. */
+               fw_send_request(dev->card, &ptask->transaction,
+                               TCODE_STREAM_DATA,
+                               fw_stream_packet_destination_id(3,
+                                               IEEE1394_BROADCAST_CHANNEL, 0),
+                               generation, SCODE_100, 0ULL, ptask->skb->data,
+                               tx_len + 8, fwnet_write_complete, ptask);
+
+               /* FIXME race? */
+               spin_lock_irqsave(&dev->lock, flags);
+               list_add_tail(&ptask->pt_link, &dev->broadcasted_list);
+               spin_unlock_irqrestore(&dev->lock, flags);
+
                return 0;
-#endif
        }
-       fw_debug("send_request (%p, %p, WRITE_BLOCK, %x, %x, %x, %llx, %p, %d, %p, %p\n",
- priv->card, &ptask->transaction, ptask->dest_node, ptask->generation,
- ptask->speed, (unsigned long long)ptask->fifo_addr, ptask->skb->data, tx_len,
- ipv4_write_complete, ptask );
-       fw_send_request ( priv->card, &ptask->transaction,
- TCODE_WRITE_BLOCK_REQUEST, ptask->dest_node, ptask->generation, ptask->speed,
- ptask->fifo_addr, ptask->skb->data, tx_len, ipv4_write_complete, ptask );
-       spin_lock_irqsave(&priv->lock,flags);
-       list_add_tail ( &ptask->packet_list, &priv->sent_list );
-       spin_unlock_irqrestore(&priv->lock,flags);
-       netdev = priv->card->netdev;
-       netdev->trans_start = jiffies;
+
+       fw_send_request(dev->card, &ptask->transaction,
+                       TCODE_WRITE_BLOCK_REQUEST, ptask->dest_node,
+                       ptask->generation, ptask->speed, ptask->fifo_addr,
+                       ptask->skb->data, tx_len, fwnet_write_complete, ptask);
+
+       /* FIXME race? */
+       spin_lock_irqsave(&dev->lock, flags);
+       list_add_tail(&ptask->pt_link, &dev->sent_list);
+       spin_unlock_irqrestore(&dev->lock, flags);
+
+       net = dev->card->netdev;
+       net->trans_start = jiffies;
+
        return 0;
 }
 
-static int ipv4_broadcast_start ( struct ipv4_priv *priv ) {
+static int fwnet_broadcast_start(struct fwnet_device *dev)
+{
        struct fw_iso_context *context;
        int retval;
        unsigned num_packets;
@@ -1260,150 +1152,151 @@ static int ipv4_broadcast_start ( struct ipv4_priv *priv ) {
        struct fw_iso_packet packet;
        unsigned long offset;
        unsigned u;
-       /* unsigned transmit_speed; */
 
-#if 0 /* stefanr */
-       if ( priv->card->broadcast_channel != (BROADCAST_CHANNEL_VALID|BROADCAST_CHANNEL_INITIAL)) {
-               fw_notify ( "Invalid broadcast channel %x\n", priv->card->broadcast_channel );
-               /* FIXME: try again later? */
-               /* return -EINVAL; */
-       }
-#endif
-       if ( priv->local_fifo == INVALID_FIFO_ADDR ) {
-               struct fw_address_region region;
-
-               priv->handler.length = FIFO_SIZE;
-               priv->handler.address_callback = ipv4_receive_packet;
-               priv->handler.callback_data = priv;
-               /* FIXME: this is OHCI, but what about others? */
-               region.start = 0xffff00000000ULL;
-               region.end =   0xfffffffffffcULL;
-
-               retval = fw_core_add_address_handler ( &priv->handler, &region );
-               if ( retval < 0 )
+       if (dev->local_fifo == FWNET_NO_FIFO_ADDR) {
+               /* outside OHCI posted write area? */
+               static const struct fw_address_region region = {
+                       .start = 0xffff00000000ULL,
+                       .end   = CSR_REGISTER_BASE,
+               };
+
+               dev->handler.length = 4096;
+               dev->handler.address_callback = fwnet_receive_packet;
+               dev->handler.callback_data = dev;
+
+               retval = fw_core_add_address_handler(&dev->handler, &region);
+               if (retval < 0)
                        goto failed_initial;
-               priv->local_fifo = priv->handler.offset;
+
+               dev->local_fifo = dev->handler.offset;
        }
 
-       /*
-        * FIXME: rawiso limits us to PAGE_SIZE.  This only matters if we ever have
-        * a machine with PAGE_SIZE < 4096
-        */
-       max_receive = 1U << (priv->card->max_receive + 1);
-       num_packets = ( ipv4_iso_page_count * PAGE_SIZE ) / max_receive;
-       if ( ! priv->broadcast_rcv_context ) {
+       max_receive = 1U << (dev->card->max_receive + 1);
+       num_packets = (FWNET_ISO_PAGE_COUNT * PAGE_SIZE) / max_receive;
+
+       if (!dev->broadcast_rcv_context) {
                void **ptrptr;
 
-               context = fw_iso_context_create ( priv->card,
FW_ISO_CONTEXT_RECEIVE, BROADCAST_CHANNEL,
priv->card->link_speed, 8, ipv4_receive_broadcast, priv );
+               context = fw_iso_context_create(dev->card,
                  FW_ISO_CONTEXT_RECEIVE, IEEE1394_BROADCAST_CHANNEL,
                  dev->card->link_speed, 8, fwnet_receive_broadcast, dev);
                if (IS_ERR(context)) {
                        retval = PTR_ERR(context);
                        goto failed_context_create;
                }
-               retval = fw_iso_buffer_init ( &priv->broadcast_rcv_buffer,
- priv->card, ipv4_iso_page_count, DMA_FROM_DEVICE );
-               if ( retval < 0 )
+
+               retval = fw_iso_buffer_init(&dev->broadcast_rcv_buffer,
+                   dev->card, FWNET_ISO_PAGE_COUNT, DMA_FROM_DEVICE);
+               if (retval < 0)
                        goto failed_buffer_init;
-               ptrptr = kmalloc ( sizeof(void*)*num_packets, GFP_KERNEL );
-               if ( ! ptrptr ) {
+
+               ptrptr = kmalloc(sizeof(void *) * num_packets, GFP_KERNEL);
+               if (!ptrptr) {
                        retval = -ENOMEM;
                        goto failed_ptrs_alloc;
                }
-               priv->broadcast_rcv_buffer_ptrs = ptrptr;
-               for ( u = 0; u < ipv4_iso_page_count; u++ ) {
+
+               dev->broadcast_rcv_buffer_ptrs = ptrptr;
+               for (u = 0; u < FWNET_ISO_PAGE_COUNT; u++) {
                        void *ptr;
                        unsigned v;
 
-                       ptr = kmap ( priv->broadcast_rcv_buffer.pages[u] );
-                       for ( v = 0; v < num_packets / ipv4_iso_page_count; v++ )
-                               *ptrptr++ = (void *)((char *)ptr + v * max_receive);
+                       ptr = kmap(dev->broadcast_rcv_buffer.pages[u]);
+                       for (v = 0; v < num_packets / FWNET_ISO_PAGE_COUNT; v++)
+                               *ptrptr++ = (void *)
+                                               ((char *)ptr + v * max_receive);
                }
-               priv->broadcast_rcv_context = context;
-       } else
-               context = priv->broadcast_rcv_context;
+               dev->broadcast_rcv_context = context;
+       } else {
+               context = dev->broadcast_rcv_context;
+       }
 
        packet.payload_length = max_receive;
        packet.interrupt = 1;
        packet.skip = 0;
        packet.tag = 3;
        packet.sy = 0;
-       packet.header_length = IPV4_GASP_OVERHEAD;
+       packet.header_length = IEEE1394_GASP_HDR_SIZE;
        offset = 0;
-       for ( u = 0; u < num_packets; u++ ) {
-               retval = fw_iso_context_queue ( context, &packet,
- &priv->broadcast_rcv_buffer, offset );
-               if ( retval < 0 )
+
+       for (u = 0; u < num_packets; u++) {
+               retval = fw_iso_context_queue(context, &packet,
+                               &dev->broadcast_rcv_buffer, offset);
+               if (retval < 0)
                        goto failed_rcv_queue;
+
                offset += max_receive;
        }
-       priv->num_broadcast_rcv_ptrs = num_packets;
-       priv->rcv_buffer_size = max_receive;
-       priv->broadcast_rcv_next_ptr = 0U;
-       retval = fw_iso_context_start ( context, -1, 0, FW_ISO_CONTEXT_MATCH_ALL_TAGS ); /* ??? sync */
-       if ( retval < 0 )
+       dev->num_broadcast_rcv_ptrs = num_packets;
+       dev->rcv_buffer_size = max_receive;
+       dev->broadcast_rcv_next_ptr = 0U;
+       retval = fw_iso_context_start(context, -1, 0,
+                       FW_ISO_CONTEXT_MATCH_ALL_TAGS); /* ??? sync */
+       if (retval < 0)
                goto failed_rcv_queue;
-       /* FIXME: adjust this when we know the max receive speeds of all other IP nodes on the bus. */
-       /* since we only xmt at S100 ??? */
-       priv->broadcast_xmt_max_payload = S100_BUFFER_SIZE - IPV4_GASP_OVERHEAD - IPV4_UNFRAG_HDR_SIZE;
-       priv->broadcast_state = IPV4_BROADCAST_RUNNING;
+
+       /* FIXME: adjust it according to the min. speed of all known peers? */
+       dev->broadcast_xmt_max_payload = IEEE1394_MAX_PAYLOAD_S100
+                       - IEEE1394_GASP_HDR_SIZE - RFC2374_UNFRAG_HDR_SIZE;
+       dev->broadcast_state = FWNET_BROADCAST_RUNNING;
+
        return 0;
 
  failed_rcv_queue:
-       kfree ( priv->broadcast_rcv_buffer_ptrs );
-       priv->broadcast_rcv_buffer_ptrs = NULL;
+       kfree(dev->broadcast_rcv_buffer_ptrs);
+       dev->broadcast_rcv_buffer_ptrs = NULL;
  failed_ptrs_alloc:
-       fw_iso_buffer_destroy ( &priv->broadcast_rcv_buffer, priv->card );
+       fw_iso_buffer_destroy(&dev->broadcast_rcv_buffer, dev->card);
  failed_buffer_init:
-       fw_iso_context_destroy ( context );
-       priv->broadcast_rcv_context = NULL;
+       fw_iso_context_destroy(context);
+       dev->broadcast_rcv_context = NULL;
  failed_context_create:
-       fw_core_remove_address_handler ( &priv->handler );
+       fw_core_remove_address_handler(&dev->handler);
  failed_initial:
-       priv->local_fifo = INVALID_FIFO_ADDR;
+       dev->local_fifo = FWNET_NO_FIFO_ADDR;
+
        return retval;
 }
 
-/* This is called after an "ifup" */
-static int ipv4_open(struct net_device *dev) {
-       struct ipv4_priv *priv;
+/* ifup */
+static int fwnet_open(struct net_device *net)
+{
+       struct fwnet_device *dev = netdev_priv(net);
        int ret;
 
-       priv = netdev_priv(dev);
-       if (priv->broadcast_state == IPV4_BROADCAST_ERROR) {
-               ret = ipv4_broadcast_start ( priv );
+       if (dev->broadcast_state == FWNET_BROADCAST_ERROR) {
+               ret = fwnet_broadcast_start(dev);
                if (ret)
                        return ret;
        }
-       netif_start_queue(dev);
+       netif_start_queue(net);
+
        return 0;
 }
 
-/* This is called after an "ifdown" */
-static int ipv4_stop(struct net_device *netdev)
+/* ifdown */
+static int fwnet_stop(struct net_device *net)
 {
-       /* flush priv->wake */
-       /* flush_scheduled_work(); */
+       netif_stop_queue(net);
+
+       /* Deallocate iso context for use by other applications? */
 
-       netif_stop_queue(netdev);
        return 0;
 }
 
-/* Transmit a packet (called by kernel) */
-static int ipv4_tx(struct sk_buff *skb, struct net_device *netdev)
+static int fwnet_tx(struct sk_buff *skb, struct net_device *net)
 {
-       struct ipv4_ether_hdr hdr_buf;
-       struct ipv4_priv *priv = netdev_priv(netdev);
+       struct fwnet_header hdr_buf;
+       struct fwnet_device *dev = netdev_priv(net);
        __be16 proto;
        u16 dest_node;
-       enum ipv4_tx_type tx_type;
        unsigned max_payload;
        u16 dg_size;
        u16 *datagram_label_ptr;
-       struct ipv4_packet_task *ptask;
-       struct ipv4_node *node = NULL;
+       struct fwnet_packet_task *ptask;
+       struct fwnet_peer *peer = NULL;
 
-       ptask = kmem_cache_alloc(ipv4_packet_task_cache, GFP_ATOMIC);
+       ptask = kmem_cache_alloc(fwnet_packet_task_cache, GFP_ATOMIC);
        if (ptask == NULL)
                goto fail;
 
@@ -1412,7 +1305,7 @@ static int ipv4_tx(struct sk_buff *skb, struct net_device *netdev)
                goto fail;
 
        /*
-        * Get rid of the fake ipv4 header, but first make a copy.
+        * Make a copy of the driver-specific header.
         * We might need to rebuild the header on tx failure.
         */
        memcpy(&hdr_buf, skb->data, sizeof(hdr_buf));
@@ -1425,110 +1318,95 @@ static int ipv4_tx(struct sk_buff *skb, struct net_device *netdev)
         * Set the transmission type for the packet.  ARP packets and IP
         * broadcast packets are sent via GASP.
         */
-       if (   memcmp(hdr_buf.h_dest, netdev->broadcast, IPV4_ALEN) == 0
+       if (memcmp(hdr_buf.h_dest, net->broadcast, FWNET_ALEN) == 0
            || proto == htons(ETH_P_ARP)
-           || (   proto == htons(ETH_P_IP)
-               && IN_MULTICAST(ntohl(ip_hdr(skb)->daddr)) ) ) {
-               /* fw_debug ( "transmitting arp or multicast packet\n" );*/
-               tx_type = IPV4_GASP;
-               dest_node = ALL_NODES;
-               max_payload = priv->broadcast_xmt_max_payload;
-               /* BUG_ON(max_payload < S100_BUFFER_SIZE - IPV4_GASP_OVERHEAD); */
-               datagram_label_ptr = &priv->broadcast_xmt_datagramlabel;
-               ptask->fifo_addr = INVALID_FIFO_ADDR;
-               ptask->generation = 0U;
-               ptask->dest_node = 0U;
-               ptask->speed = 0;
+           || (proto == htons(ETH_P_IP)
+               && IN_MULTICAST(ntohl(ip_hdr(skb)->daddr)))) {
+               max_payload = dev->broadcast_xmt_max_payload;
+               datagram_label_ptr = &dev->broadcast_xmt_datagramlabel;
+
+               ptask->fifo_addr = FWNET_NO_FIFO_ADDR;
+               ptask->generation = 0;
+               ptask->dest_node = IEEE1394_ALL_NODES;
+               ptask->speed = SCODE_100;
        } else {
-               __be64 guid = get_unaligned((u64 *)hdr_buf.h_dest);
+               __be64 guid = get_unaligned((__be64 *)hdr_buf.h_dest);
                u8 generation;
 
-               node = ipv4_node_find_by_guid(priv, be64_to_cpu(guid));
-               if (!node) {
-                       fw_debug ( "Normal packet but no node\n" );
+               peer = fwnet_peer_find_by_guid(dev, be64_to_cpu(guid));
+               if (!peer)
                        goto fail;
-               }
 
-               if (node->fifo == INVALID_FIFO_ADDR) {
-                       fw_debug ( "Normal packet but no fifo addr\n" );
+               if (peer->fifo == FWNET_NO_FIFO_ADDR)
                        goto fail;
-               }
 
-               /* fw_debug ( "Transmitting normal packet to %x at %llxx\n", node->nodeid, node->fifo ); */
-               generation = node->generation;
-               dest_node = node->nodeid;
-               max_payload = node->max_payload;
-               /* BUG_ON(max_payload < S100_BUFFER_SIZE - IPV4_FRAG_HDR_SIZE); */
+               generation = peer->generation;
+               smp_rmb();
+               dest_node = peer->node_id;
+
+               max_payload = peer->max_payload;
+               datagram_label_ptr = &peer->datagram_label;
 
-               datagram_label_ptr = &node->datagram_label;
-               tx_type = IPV4_WRREQ;
-               ptask->fifo_addr = node->fifo;
+               ptask->fifo_addr = peer->fifo;
                ptask->generation = generation;
                ptask->dest_node = dest_node;
-               ptask->speed = node->xmt_speed;
+               ptask->speed = peer->xmt_speed;
        }
 
        /* If this is an ARP packet, convert it */
        if (proto == htons(ETH_P_ARP)) {
-               /* Convert a standard ARP packet to 1394 ARP. The first 8 bytes (the entire
-                * arphdr) is the same format as the ip1394 header, so they overlap.  The rest
-                * needs to be munged a bit.  The remainder of the arphdr is formatted based
-                * on hwaddr len and ipaddr len.  We know what they'll be, so it's easy to
-                * judge.
-                *
-                * Now that the EUI is used for the hardware address all we need to do to make
-                * this work for 1394 is to insert 2 quadlets that contain max_rec size,
-                * speed, and unicast FIFO address information between the sender_unique_id
-                * and the IP addresses.
-                */
                struct arphdr *arp = (struct arphdr *)skb->data;
                unsigned char *arp_ptr = (unsigned char *)(arp + 1);
-               struct ipv4_arp *arp1394 = (struct ipv4_arp *)skb->data;
-               u32 ipaddr;
-
-               ipaddr = *(u32*)(arp_ptr + IPV4_ALEN);
-               arp1394->hw_addr_len    = 16;
-               arp1394->max_rec        = priv->card->max_receive;
-               arp1394->sspd           = priv->card->link_speed;
-               arp1394->fifo_hi        = htons(priv->local_fifo >> 32);
-               arp1394->fifo_lo        = htonl(priv->local_fifo & 0xFFFFFFFF);
-               arp1394->sip            = ipaddr;
+               struct rfc2734_arp *arp1394 = (struct rfc2734_arp *)skb->data;
+               __be32 ipaddr;
+
+               ipaddr = get_unaligned((__be32 *)(arp_ptr + FWNET_ALEN));
+
+               arp1394->hw_addr_len    = RFC2734_HW_ADDR_LEN;
+               arp1394->max_rec        = dev->card->max_receive;
+               arp1394->sspd           = dev->card->link_speed;
+
+               put_unaligned_be16(dev->local_fifo >> 32,
+                                  &arp1394->fifo_hi);
+               put_unaligned_be32(dev->local_fifo & 0xffffffff,
+                                  &arp1394->fifo_lo);
+               put_unaligned(ipaddr, &arp1394->sip);
        }
-       if ( ipv4_max_xmt && max_payload > ipv4_max_xmt )
-               max_payload = ipv4_max_xmt;
 
        ptask->hdr.w0 = 0;
        ptask->hdr.w1 = 0;
        ptask->skb = skb;
-       ptask->priv = priv;
-        ptask->tx_type = tx_type;
+       ptask->dev = dev;
+
        /* Does it all fit in one packet? */
-       if ( dg_size <= max_payload ) {
-               ipv4_make_uf_hdr(&ptask->hdr, be16_to_cpu(proto));
+       if (dg_size <= max_payload) {
+               fwnet_make_uf_hdr(&ptask->hdr, ntohs(proto));
                ptask->outstanding_pkts = 1;
-               max_payload = dg_size + IPV4_UNFRAG_HDR_SIZE;
+               max_payload = dg_size + RFC2374_UNFRAG_HDR_SIZE;
        } else {
                u16 datagram_label;
 
-               max_payload -= IPV4_FRAG_OVERHEAD;
+               max_payload -= RFC2374_FRAG_OVERHEAD;
                datagram_label = (*datagram_label_ptr)++;
-               ipv4_make_ff_hdr(&ptask->hdr, be16_to_cpu(proto), dg_size, datagram_label );
+               fwnet_make_ff_hdr(&ptask->hdr, ntohs(proto), dg_size,
+                                 datagram_label);
                ptask->outstanding_pkts = DIV_ROUND_UP(dg_size, max_payload);
-               max_payload += IPV4_FRAG_HDR_SIZE;
+               max_payload += RFC2374_FRAG_HDR_SIZE;
        }
        ptask->max_payload = max_payload;
-       ipv4_send_packet ( ptask );
+       fwnet_send_packet(ptask);
+
        return NETDEV_TX_OK;
 
  fail:
        if (ptask)
-               kmem_cache_free(ipv4_packet_task_cache, ptask);
+               kmem_cache_free(fwnet_packet_task_cache, ptask);
 
        if (skb != NULL)
                dev_kfree_skb(skb);
 
-       netdev->stats.tx_dropped++;
-       netdev->stats.tx_errors++;
+       net->stats.tx_dropped++;
+       net->stats.tx_errors++;
 
        /*
         * FIXME: According to a patch from 2003-02-26, "returning non-zero
@@ -1540,280 +1418,291 @@ static int ipv4_tx(struct sk_buff *skb, struct net_device *netdev)
        return NETDEV_TX_OK;
 }
 
-/*
- * FIXME: What to do if we timeout? I think a host reset is probably in order,
- * so that's what we do. Should we increment the stat counters too?
- */
-static void ipv4_tx_timeout(struct net_device *dev) {
-       struct ipv4_priv *priv;
+static void fwnet_tx_timeout(struct net_device *net)
+{
+       fw_error("%s: timeout\n", net->name);
 
-       priv = netdev_priv(dev);
-       fw_error ( "%s: Timeout, resetting host\n", dev->name );
-#if 0 /* stefanr */
-       fw_core_initiate_bus_reset ( priv->card, 1 );
-#endif
+       /* FIXME: What to do if we timeout? */
 }
 
-static int ipv4_change_mtu ( struct net_device *dev, int new_mtu ) {
-#if 0
-       int max_mtu;
-       struct ipv4_priv *priv;
-#endif
-
+static int fwnet_change_mtu(struct net_device *net, int new_mtu)
+{
        if (new_mtu < 68)
                return -EINVAL;
 
-#if 0
-       priv = netdev_priv(dev);
-       /* This is not actually true because we can fragment packets at the firewire layer */
-       max_mtu = (1 << (priv->card->max_receive + 1))
-                               - sizeof(struct ipv4_hdr) - IPV4_GASP_OVERHEAD;
-       if (new_mtu > max_mtu) {
-               fw_notify ( "%s: Local node constrains MTU to %d\n", dev->name, max_mtu);
-               return -ERANGE;
-       }
-#endif
-       dev->mtu = new_mtu;
+       net->mtu = new_mtu;
        return 0;
 }
 
-static void ipv4_get_drvinfo(struct net_device *dev,
-struct ethtool_drvinfo *info) {
-       strcpy(info->driver, ipv4_driver_name);
-       strcpy(info->bus_info, "ieee1394"); /* FIXME provide more detail? */
+static void fwnet_get_drvinfo(struct net_device *net,
+                             struct ethtool_drvinfo *info)
+{
+       strcpy(info->driver, KBUILD_MODNAME);
+       strcpy(info->bus_info, "ieee1394");
 }
 
-static struct ethtool_ops ipv4_ethtool_ops = {
-       .get_drvinfo = ipv4_get_drvinfo,
+static struct ethtool_ops fwnet_ethtool_ops = {
+       .get_drvinfo = fwnet_get_drvinfo,
 };
 
-static const struct net_device_ops ipv4_netdev_ops = {
-       .ndo_open       = ipv4_open,
-       .ndo_stop       = ipv4_stop,
-       .ndo_start_xmit = ipv4_tx,
-       .ndo_tx_timeout = ipv4_tx_timeout,
-       .ndo_change_mtu = ipv4_change_mtu,
+static const struct net_device_ops fwnet_netdev_ops = {
+       .ndo_open       = fwnet_open,
+       .ndo_stop       = fwnet_stop,
+       .ndo_start_xmit = fwnet_tx,
+       .ndo_tx_timeout = fwnet_tx_timeout,
+       .ndo_change_mtu = fwnet_change_mtu,
 };
 
-static void ipv4_init_dev ( struct net_device *dev ) {
-       dev->header_ops         = &ipv4_header_ops;
-       dev->netdev_ops         = &ipv4_netdev_ops;
-       SET_ETHTOOL_OPS(dev, &ipv4_ethtool_ops);
-
-       dev->watchdog_timeo     = IPV4_TIMEOUT;
-       dev->flags              = IFF_BROADCAST | IFF_MULTICAST;
-       dev->features           = NETIF_F_HIGHDMA;
-       dev->addr_len           = IPV4_ALEN;
-       dev->hard_header_len    = IPV4_HLEN;
-       dev->type               = ARPHRD_IEEE1394;
-
-       /* FIXME: This value was copied from ether_setup(). Is it too much? */
-       dev->tx_queue_len       = 1000;
+static void fwnet_init_dev(struct net_device *net)
+{
+       net->header_ops         = &fwnet_header_ops;
+       net->netdev_ops         = &fwnet_netdev_ops;
+       net->watchdog_timeo     = 100000; /* ? FIXME */
+       net->flags              = IFF_BROADCAST | IFF_MULTICAST;
+       net->features           = NETIF_F_HIGHDMA;
+       net->addr_len           = FWNET_ALEN;
+       net->hard_header_len    = FWNET_HLEN;
+       net->type               = ARPHRD_IEEE1394;
+       net->tx_queue_len       = 1000; /* ? FIXME */
+       SET_ETHTOOL_OPS(net, &fwnet_ethtool_ops);
 }
 
-static int ipv4_probe ( struct device *dev ) {
-       struct fw_unit * unit;
-       struct fw_device *device;
-       struct fw_card *card;
-       struct net_device *netdev;
-       struct ipv4_priv *priv;
+/* FIXME create netdev upon first fw_unit of a card, not upon local fw_unit */
+static int fwnet_probe(struct device *_dev)
+{
+       struct fw_unit *unit = fw_unit(_dev);
+       struct fw_device *device = fw_parent_device(unit);
+       struct fw_card *card = device->card;
+       struct net_device *net;
+       struct fwnet_device *dev;
        unsigned max_mtu;
-       __be64 guid;
-
-       fw_debug("ipv4 Probing\n" );
-       unit = fw_unit ( dev );
-       device = fw_device ( unit->device.parent );
-       card = device->card;
 
-       if ( ! device->is_local ) {
+       if (!device->is_local) {
                int added;
 
-               fw_debug ( "Non-local, adding remote node entry\n" );
-               added = ipv4_node_new ( card, device );
+               added = fwnet_peer_new(card, device);
                return added;
        }
-       fw_debug("ipv4 Local: adding netdev\n" );
-       netdev = alloc_netdev ( sizeof(*priv), "firewire%d", ipv4_init_dev );
-       if ( netdev == NULL) {
-               fw_error( "Out of memory\n");
+       net = alloc_netdev(sizeof(*dev), "firewire%d", fwnet_init_dev);
+       if (net == NULL) {
+               fw_error("out of memory\n");
                goto out;
        }
 
-       SET_NETDEV_DEV(netdev, card->device);
-       priv = netdev_priv(netdev);
+       SET_NETDEV_DEV(net, card->device);
+       dev = netdev_priv(net);
 
-       spin_lock_init(&priv->lock);
-       priv->broadcast_state = IPV4_BROADCAST_ERROR;
-       priv->broadcast_rcv_context = NULL;
-       priv->broadcast_xmt_max_payload = 0;
-       priv->broadcast_xmt_datagramlabel = 0;
+       spin_lock_init(&dev->lock);
+       dev->broadcast_state = FWNET_BROADCAST_ERROR;
+       dev->broadcast_rcv_context = NULL;
+       dev->broadcast_xmt_max_payload = 0;
+       dev->broadcast_xmt_datagramlabel = 0;
 
-       priv->local_fifo = INVALID_FIFO_ADDR;
+       dev->local_fifo = FWNET_NO_FIFO_ADDR;
 
-       /* INIT_WORK(&priv->wake, ipv4_handle_queue);*/
-       INIT_LIST_HEAD(&priv->packet_list);
-       INIT_LIST_HEAD(&priv->broadcasted_list);
-       INIT_LIST_HEAD(&priv->sent_list );
+       /* INIT_WORK(&dev->wake, fwnet_handle_queue);*/
+       INIT_LIST_HEAD(&dev->packet_list);
+       INIT_LIST_HEAD(&dev->broadcasted_list);
+       INIT_LIST_HEAD(&dev->sent_list);
 
-       priv->card = card;
+       dev->card = card;
 
        /*
         * Use the RFC 2734 default 1500 octets or the maximum payload
         * as initial MTU
         */
        max_mtu = (1 << (card->max_receive + 1))
-                 - sizeof(struct ipv4_hdr) - IPV4_GASP_OVERHEAD;
-       netdev->mtu = min(1500U, max_mtu);
+                 - sizeof(struct rfc2734_header) - IEEE1394_GASP_HDR_SIZE;
+       net->mtu = min(1500U, max_mtu);
 
        /* Set our hardware address while we're at it */
-       guid = cpu_to_be64(card->guid);
-       memcpy(netdev->dev_addr, &guid, sizeof(u64));
-       memset(netdev->broadcast, 0xff, sizeof(u64));
-       if ( register_netdev ( netdev ) ) {
-               fw_error ( "Cannot register the driver\n");
+       put_unaligned_be64(card->guid, net->dev_addr);
+       put_unaligned_be64(~0ULL, net->broadcast);
+       if (register_netdev(net)) {
+               fw_error("Cannot register the driver\n");
                goto out;
        }
 
-       fw_notify ( "%s: IPv4 over Firewire on device %016llx\n",
netdev->name, card->guid );
-       card->netdev = netdev;
+       fw_notify("%s: IPv4 over FireWire on device %016llx\n",
                net->name, (unsigned long long)card->guid);
+       card->netdev = net;
 
-       return 0 /* ipv4_new_node ( ud ) */;
+       return 0;
  out:
-       if ( netdev )
-               free_netdev ( netdev );
+       if (net)
+               free_netdev(net);
+
        return -ENOENT;
 }
 
+static int fwnet_remove(struct device *_dev)
+{
+       struct fw_unit *unit = fw_unit(_dev);
+       struct fw_device *device = fw_parent_device(unit);
+       struct fw_card *card = device->card;
+       struct net_device *net;
+       struct fwnet_device *dev;
+       struct fwnet_peer *peer;
+       struct fwnet_partial_datagram *pd, *pd_next;
+       struct fwnet_packet_task *ptask, *pt_next;
+
+       if (!device->is_local) {
+               fwnet_peer_delete(card, device);
 
-static int ipv4_remove ( struct device *dev ) {
-       struct fw_unit * unit;
-       struct fw_device *device;
-       struct fw_card *card;
-       struct net_device *netdev;
-       struct ipv4_priv *priv;
-       struct ipv4_node *node;
-       struct ipv4_partial_datagram *pd, *pd_next;
-       struct ipv4_packet_task *ptask, *pt_next;
-
-       fw_debug("ipv4 Removing\n" );
-       unit = fw_unit ( dev );
-       device = fw_device ( unit->device.parent );
-       card = device->card;
-
-       if ( ! device->is_local ) {
-               fw_debug ( "Node %x is non-local, removing remote node entry\n", device->node_id );
-               ipv4_node_delete ( card, device );
                return 0;
        }
-       netdev = card->netdev;
-       if ( netdev ) {
-               fw_debug ( "Node %x is local: deleting netdev\n", device->node_id );
-               priv = netdev_priv ( netdev );
-               unregister_netdev ( netdev );
-               fw_debug ( "unregistered\n" );
-               if ( priv->local_fifo != INVALID_FIFO_ADDR )
-                       fw_core_remove_address_handler ( &priv->handler );
-               fw_debug ( "address handler gone\n" );
-               if ( priv->broadcast_rcv_context ) {
-                       fw_iso_context_stop ( priv->broadcast_rcv_context );
-                       fw_iso_buffer_destroy ( &priv->broadcast_rcv_buffer, priv->card );
-                       fw_iso_context_destroy ( priv->broadcast_rcv_context );
-                       fw_debug ( "rcv stopped\n" );
+
+       net = card->netdev;
+       if (net) {
+               dev = netdev_priv(net);
+               unregister_netdev(net);
+
+               if (dev->local_fifo != FWNET_NO_FIFO_ADDR)
+                       fw_core_remove_address_handler(&dev->handler);
+               if (dev->broadcast_rcv_context) {
+                       fw_iso_context_stop(dev->broadcast_rcv_context);
+                       fw_iso_buffer_destroy(&dev->broadcast_rcv_buffer,
+                                             dev->card);
+                       fw_iso_context_destroy(dev->broadcast_rcv_context);
                }
-               list_for_each_entry_safe( ptask, pt_next, &priv->packet_list, packet_list ) {
-                       dev_kfree_skb_any ( ptask->skb );
-                       kmem_cache_free( ipv4_packet_task_cache, ptask );
+               list_for_each_entry_safe(ptask, pt_next,
+                                        &dev->packet_list, pt_link) {
+                       dev_kfree_skb_any(ptask->skb);
+                       kmem_cache_free(fwnet_packet_task_cache, ptask);
                }
-               list_for_each_entry_safe( ptask, pt_next, &priv->broadcasted_list, packet_list ) {
-                       dev_kfree_skb_any ( ptask->skb );
-                       kmem_cache_free( ipv4_packet_task_cache, ptask );
+               list_for_each_entry_safe(ptask, pt_next,
+                                        &dev->broadcasted_list, pt_link) {
+                       dev_kfree_skb_any(ptask->skb);
+                       kmem_cache_free(fwnet_packet_task_cache, ptask);
                }
-               list_for_each_entry_safe( ptask, pt_next, &priv->sent_list, packet_list ) {
-                       dev_kfree_skb_any ( ptask->skb );
-                       kmem_cache_free( ipv4_packet_task_cache, ptask );
+               list_for_each_entry_safe(ptask, pt_next,
+                                        &dev->sent_list, pt_link) {
+                       dev_kfree_skb_any(ptask->skb);
+                       kmem_cache_free(fwnet_packet_task_cache, ptask);
                }
-               fw_debug ( "lists emptied\n" );
-               list_for_each_entry( node, &card->ipv4_nodes, ipv4_nodes ) {
-                       if ( node->pdg_size ) {
-                               list_for_each_entry_safe( pd, pd_next, &node->pdg_list, pdg_list )
-                                       ipv4_pd_delete ( pd );
-                               node->pdg_size = 0;
+               list_for_each_entry(peer, &card->peer_list, peer_link) {
+                       if (peer->pdg_size) {
+                               list_for_each_entry_safe(pd, pd_next,
+                                               &peer->pd_list, pd_link)
+                                       fwnet_pd_delete(pd);
+                               peer->pdg_size = 0;
                        }
-                       node->fifo = INVALID_FIFO_ADDR;
+                       peer->fifo = FWNET_NO_FIFO_ADDR;
                }
-               fw_debug ( "nodes cleaned up\n" );
-               free_netdev ( netdev );
+               free_netdev(net);
                card->netdev = NULL;
-               fw_debug ( "done\n" );
        }
+
        return 0;
 }
 
-static void ipv4_update ( struct fw_unit *unit ) {
-       struct fw_device *device;
-       struct fw_card *card;
+/*
+ * FIXME abort partially sent fragmented datagrams,
+ * discard partially received fragmented datagrams
+ */
+static void fwnet_update(struct fw_unit *unit)
+{
+       struct fw_device *device = fw_parent_device(unit);
+       struct net_device *net = device->card->netdev;
+       struct fwnet_device *dev;
+       struct fwnet_peer *peer;
+       u64 guid;
 
-       fw_debug ( "ipv4_update unit %p\n", unit );
-       device = fw_device ( unit->device.parent );
-       card = device->card;
-       if ( ! device->is_local ) {
-               struct ipv4_node *node;
-               u64 guid;
-               struct net_device *netdev;
-               struct ipv4_priv *priv;
-
-               netdev = card->netdev;
-               if ( netdev ) {
-                       priv = netdev_priv ( netdev );
-                       guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
-                       node = ipv4_node_find_by_guid ( priv, guid );
-                       if ( ! node ) {
-                               fw_error ( "ipv4_update: no node for device %llx\n", guid );
-                               return;
-                       }
-                       fw_debug ( "Non-local, updating remote node entry for guid %llx old generation %x, old nodeid %x\n", guid, node->generation, node->nodeid );
-                       node->generation = device->generation;
-                       rmb();
-                       node->nodeid = device->node_id;
-                       fw_debug ( "New generation %x, new nodeid %x\n", node->generation, node->nodeid );
-               } else
-                       fw_error ( "nonlocal, but no netdev?  How can that be?\n" );
-       } else {
-               /* FIXME: What do we need to do on bus reset? */
-               fw_debug ( "Local, doing nothing\n" );
+       if (net && !device->is_local) {
+               dev = netdev_priv(net);
+               guid = (u64)device->config_rom[3] << 32 | device->config_rom[4];
+               peer = fwnet_peer_find_by_guid(dev, guid);
+               if (!peer) {
+                       fw_error("fwnet_update: no peer for device %016llx\n",
+                                (unsigned long long)guid);
+                       return;
+               }
+               peer->generation = device->generation;
+               rmb();
+               peer->node_id = device->node_id;
        }
 }
 
-static struct fw_driver ipv4_driver = {
+static const struct ieee1394_device_id fwnet_id_table[] = {
+       {
+               .match_flags  = IEEE1394_MATCH_SPECIFIER_ID |
+                               IEEE1394_MATCH_VERSION,
+               .specifier_id = IANA_SPECIFIER_ID,
+               .version      = RFC2734_SW_VERSION,
+       },
+       { }
+};
+
+static struct fw_driver fwnet_driver = {
        .driver = {
-               .owner = THIS_MODULE,
-               .name = ipv4_driver_name,
-               .bus = &fw_bus_type,
-               .probe = ipv4_probe,
-               .remove = ipv4_remove,
+               .owner  = THIS_MODULE,
+               .name   = "net",
+               .bus    = &fw_bus_type,
+               .probe  = fwnet_probe,
+               .remove = fwnet_remove,
        },
-       .update = ipv4_update,
-       .id_table = ipv4_id_table,
+       .update   = fwnet_update,
+       .id_table = fwnet_id_table,
+};
+
+static const u32 rfc2374_unit_directory_data[] = {
+       0x00040000,     /* directory_length             */
+       0x1200005e,     /* unit_specifier_id: IANA      */
+       0x81000003,     /* textual descriptor offset    */
+       0x13000001,     /* unit_sw_version: RFC 2734    */
+       0x81000005,     /* textual descriptor offset    */
+       0x00030000,     /* descriptor_length            */
+       0x00000000,     /* text                         */
+       0x00000000,     /* minimal ASCII, en            */
+       0x49414e41,     /* I A N A                      */
+       0x00030000,     /* descriptor_length            */
+       0x00000000,     /* text                         */
+       0x00000000,     /* minimal ASCII, en            */
+       0x49507634,     /* I P v 4                      */
+};
+
+static struct fw_descriptor rfc2374_unit_directory = {
+       .length = ARRAY_SIZE(rfc2374_unit_directory_data),
+       .key    = (CSR_DIRECTORY | CSR_UNIT) << 24,
+       .data   = rfc2374_unit_directory_data
 };
 
-static int __init ipv4_init ( void ) {
-       int added;
+static int __init fwnet_init(void)
+{
+       int err;
+
+       err = fw_core_add_descriptor(&rfc2374_unit_directory);
+       if (err)
+               return err;
 
-       added = fw_core_add_descriptor ( &ipv4_unit_directory );
-       if ( added < 0 )
-               fw_error ( "Failed to add descriptor" );
-       ipv4_packet_task_cache = kmem_cache_create("packet_task",
- sizeof(struct ipv4_packet_task), 0, 0, NULL);
-       fw_debug("Adding ipv4 module\n" );
-       return driver_register ( &ipv4_driver.driver );
+       fwnet_packet_task_cache = kmem_cache_create("packet_task",
+                       sizeof(struct fwnet_packet_task), 0, 0, NULL);
+       if (!fwnet_packet_task_cache) {
+               err = -ENOMEM;
+               goto out;
+       }
+
+       err = driver_register(&fwnet_driver.driver);
+       if (!err)
+               return 0;
+
+       kmem_cache_destroy(fwnet_packet_task_cache);
+out:
+       fw_core_remove_descriptor(&rfc2374_unit_directory);
+
+       return err;
 }
+module_init(fwnet_init);
 
-static void __exit ipv4_cleanup ( void ) {
-       fw_core_remove_descriptor ( &ipv4_unit_directory );
-       fw_debug("Removing ipv4 module\n" );
-       driver_unregister ( &ipv4_driver.driver );
+static void __exit fwnet_cleanup(void)
+{
+       driver_unregister(&fwnet_driver.driver);
+       kmem_cache_destroy(fwnet_packet_task_cache);
+       fw_core_remove_descriptor(&rfc2374_unit_directory);
 }
+module_exit(fwnet_cleanup);
 
-module_init(ipv4_init);
-module_exit(ipv4_cleanup);
+MODULE_AUTHOR("Jay Fenlason <fenlason@redhat.com>");
+MODULE_DESCRIPTION("IPv4 over IEEE1394 as per RFC 2734");
+MODULE_LICENSE("GPL");
+MODULE_DEVICE_TABLE(ieee1394, fwnet_id_table);
index d44f47d..5cb0c15 100644 (file)
@@ -131,13 +131,10 @@ struct fw_card {
        bool broadcast_channel_allocated;
        u32 broadcast_channel;
        u32 topology_map[(CSR_TOPOLOGY_MAP_END - CSR_TOPOLOGY_MAP) / 4];
-       /* Only non-NULL if firewire-ipv4 is active on this card. */
+
+       /* firewire-net driver data */
        void *netdev;
-       /*
-        * The nodes get probed before the card, so we need a place to store
-        * them independent of card->netdev
-        */
-       struct list_head ipv4_nodes;
+       struct list_head peer_list;
 };
 
 static inline struct fw_card *fw_card_get(struct fw_card *card)