netfilter: nfnetlink: netns support
[safe/jmp/linux-2.6] / net / netfilter / nfnetlink_log.c
index 62c3f31..285e902 100644 (file)
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
- *
- * 2006-01-26 Harald Welte <laforge@netfilter.org>
- *     - Add optional local and global sequence number to detect lost
- *       events from userspace
- *
  */
 #include <linux/module.h>
 #include <linux/skbuff.h>
@@ -34,6 +29,8 @@
 #include <linux/jhash.h>
 #include <linux/random.h>
 #include <net/sock.h>
+#include <net/netfilter/nf_log.h>
+#include <net/netfilter/nfnetlink_log.h>
 
 #include <asm/atomic.h>
 
 #define NFULNL_NLBUFSIZ_DEFAULT        NLMSG_GOODSIZE
 #define NFULNL_TIMEOUT_DEFAULT         100     /* every second */
 #define NFULNL_QTHRESH_DEFAULT         100     /* 100 packets */
+#define NFULNL_COPY_RANGE_MAX  0xFFFF  /* max packet size is limited by 16-bit struct nfattr nfa_len field */
 
 #define PRINTR(x, args...)     do { if (net_ratelimit()) \
                                     printk(x, ## args); } while (0);
 
-#if 0
-#define UDEBUG(x, args ...)    printk(KERN_DEBUG "%s(%d):%s(): " x,       \
-                                       __FILE__, __LINE__, __FUNCTION__,  \
-                                       ## args)
-#else
-#define UDEBUG(x, ...)
-#endif
-
 struct nfulnl_instance {
        struct hlist_node hlist;        /* global list of instances */
        spinlock_t lock;
@@ -63,7 +53,6 @@ struct nfulnl_instance {
 
        unsigned int qlen;              /* number of nlmsgs in skb */
        struct sk_buff *skb;            /* pre-allocatd skb */
-       struct nlmsghdr *lastnlh;       /* netlink header of last msg in skb */
        struct timer_list timer;
        int peer_pid;                   /* PID of the peer process */
 
@@ -97,8 +86,6 @@ __instance_lookup(u_int16_t group_num)
        struct hlist_node *pos;
        struct nfulnl_instance *inst;
 
-       UDEBUG("entering (group_num=%u)\n", group_num);
-
        head = &instance_table[instance_hashfn(group_num)];
        hlist_for_each_entry(inst, pos, head, hlist) {
                if (inst->group_num == group_num)
@@ -131,8 +118,8 @@ static void
 instance_put(struct nfulnl_instance *inst)
 {
        if (inst && atomic_dec_and_test(&inst->use)) {
-               UDEBUG("kfree(inst=%p)\n", inst);
                kfree(inst);
+               module_put(THIS_MODULE);
        }
 }
 
@@ -142,30 +129,32 @@ static struct nfulnl_instance *
 instance_create(u_int16_t group_num, int pid)
 {
        struct nfulnl_instance *inst;
-
-       UDEBUG("entering (group_num=%u, pid=%d)\n", group_num,
-               pid);
+       int err;
 
        write_lock_bh(&instances_lock);
        if (__instance_lookup(group_num)) {
-               inst = NULL;
-               UDEBUG("aborting, instance already exists\n");
+               err = -EEXIST;
                goto out_unlock;
        }
 
        inst = kzalloc(sizeof(*inst), GFP_ATOMIC);
-       if (!inst)
+       if (!inst) {
+               err = -ENOMEM;
                goto out_unlock;
+       }
+
+       if (!try_module_get(THIS_MODULE)) {
+               kfree(inst);
+               err = -EAGAIN;
+               goto out_unlock;
+       }
 
        INIT_HLIST_NODE(&inst->hlist);
        spin_lock_init(&inst->lock);
        /* needs to be two, since we _put() after creation */
        atomic_set(&inst->use, 2);
 
-       init_timer(&inst->timer);
-       inst->timer.function = nfulnl_timer;
-       inst->timer.data = (unsigned long)inst;
-       /* don't start timer yet. (re)start it  with every packet */
+       setup_timer(&inst->timer, nfulnl_timer, (unsigned long)inst);
 
        inst->peer_pid = pid;
        inst->group_num = group_num;
@@ -174,74 +163,45 @@ instance_create(u_int16_t group_num, int pid)
        inst->flushtimeout      = NFULNL_TIMEOUT_DEFAULT;
        inst->nlbufsiz          = NFULNL_NLBUFSIZ_DEFAULT;
        inst->copy_mode         = NFULNL_COPY_PACKET;
-       inst->copy_range        = 0xffff;
-
-       if (!try_module_get(THIS_MODULE))
-               goto out_free;
+       inst->copy_range        = NFULNL_COPY_RANGE_MAX;
 
        hlist_add_head(&inst->hlist,
                       &instance_table[instance_hashfn(group_num)]);
 
-       UDEBUG("newly added node: %p, next=%p\n", &inst->hlist,
-               inst->hlist.next);
-
        write_unlock_bh(&instances_lock);
 
        return inst;
 
-out_free:
-       instance_put(inst);
 out_unlock:
        write_unlock_bh(&instances_lock);
-       return NULL;
+       return ERR_PTR(err);
 }
 
-static int __nfulnl_send(struct nfulnl_instance *inst);
+static void __nfulnl_flush(struct nfulnl_instance *inst);
 
 static void
-_instance_destroy2(struct nfulnl_instance *inst, int lock)
+__instance_destroy(struct nfulnl_instance *inst)
 {
        /* first pull it out of the global list */
-       if (lock)
-               write_lock_bh(&instances_lock);
-
-       UDEBUG("removing instance %p (queuenum=%u) from hash\n",
-               inst, inst->group_num);
-
        hlist_del(&inst->hlist);
 
-       if (lock)
-               write_unlock_bh(&instances_lock);
-
        /* then flush all pending packets from skb */
 
        spin_lock_bh(&inst->lock);
-       if (inst->skb) {
-               if (inst->qlen)
-                       __nfulnl_send(inst);
-               if (inst->skb) {
-                       kfree_skb(inst->skb);
-                       inst->skb = NULL;
-               }
-       }
+       if (inst->skb)
+               __nfulnl_flush(inst);
        spin_unlock_bh(&inst->lock);
 
        /* and finally put the refcount */
        instance_put(inst);
-
-       module_put(THIS_MODULE);
-}
-
-static inline void
-__instance_destroy(struct nfulnl_instance *inst)
-{
-       _instance_destroy2(inst, 0);
 }
 
 static inline void
 instance_destroy(struct nfulnl_instance *inst)
 {
-       _instance_destroy2(inst, 1);
+       write_lock_bh(&instances_lock);
+       __instance_destroy(inst);
+       write_unlock_bh(&instances_lock);
 }
 
 static int
@@ -261,11 +221,8 @@ nfulnl_set_mode(struct nfulnl_instance *inst, u_int8_t mode,
 
        case NFULNL_COPY_PACKET:
                inst->copy_mode = mode;
-               /* we're using struct nfattr which has 16bit nfa_len */
-               if (range > 0xffff)
-                       inst->copy_range = 0xffff;
-               else
-                       inst->copy_range = range;
+               inst->copy_range = min_t(unsigned int,
+                                        range, NFULNL_COPY_RANGE_MAX);
                break;
 
        default:
@@ -327,14 +284,12 @@ nfulnl_set_flags(struct nfulnl_instance *inst, u_int16_t flags)
        return 0;
 }
 
-static struct sk_buff *nfulnl_alloc_skb(unsigned int inst_size,
-                                       unsigned int pkt_size)
+static struct sk_buff *
+nfulnl_alloc_skb(unsigned int inst_size, unsigned int pkt_size)
 {
        struct sk_buff *skb;
        unsigned int n;
 
-       UDEBUG("entered (%u, %u)\n", inst_size, pkt_size);
-
        /* alloc skb which should be big enough for a whole multipart
         * message.  WARNING: has to be <= 128k due to slab restrictions */
 
@@ -361,40 +316,43 @@ static struct sk_buff *nfulnl_alloc_skb(unsigned int inst_size,
 static int
 __nfulnl_send(struct nfulnl_instance *inst)
 {
-       int status;
-
-       if (timer_pending(&inst->timer))
-               del_timer(&inst->timer);
-
-       if (!inst->skb)
-               return 0;
+       int status = -1;
 
        if (inst->qlen > 1)
-               inst->lastnlh->nlmsg_type = NLMSG_DONE;
+               NLMSG_PUT(inst->skb, 0, 0,
+                         NLMSG_DONE,
+                         sizeof(struct nfgenmsg));
 
-       status = nfnetlink_unicast(inst->skb, inst->peer_pid, MSG_DONTWAIT);
-       if (status < 0) {
-               UDEBUG("netlink_unicast() failed\n");
-               /* FIXME: statistics */
-       }
+       status = nfnetlink_unicast(inst->skb, &init_net, inst->peer_pid,
+                                  MSG_DONTWAIT);
 
        inst->qlen = 0;
        inst->skb = NULL;
-       inst->lastnlh = NULL;
 
+nlmsg_failure:
        return status;
 }
 
-static void nfulnl_timer(unsigned long data)
+static void
+__nfulnl_flush(struct nfulnl_instance *inst)
 {
-       struct nfulnl_instance *inst = (struct nfulnl_instance *)data;
+       /* timer holds a reference */
+       if (del_timer(&inst->timer))
+               instance_put(inst);
+       if (inst->skb)
+               __nfulnl_send(inst);
+}
 
-       UDEBUG("timer function called, flushing buffer\n");
+static void
+nfulnl_timer(unsigned long data)
+{
+       struct nfulnl_instance *inst = (struct nfulnl_instance *)data;
 
        spin_lock_bh(&inst->lock);
-       __nfulnl_send(inst);
-       instance_put(inst);
+       if (inst->skb)
+               __nfulnl_send(inst);
        spin_unlock_bh(&inst->lock);
+       instance_put(inst);
 }
 
 /* This is an inline function, we don't really care about a long
@@ -403,22 +361,19 @@ static inline int
 __build_packet_message(struct nfulnl_instance *inst,
                        const struct sk_buff *skb,
                        unsigned int data_len,
-                       unsigned int pf,
+                       u_int8_t pf,
                        unsigned int hooknum,
                        const struct net_device *indev,
                        const struct net_device *outdev,
                        const struct nf_loginfo *li,
                        const char *prefix, unsigned int plen)
 {
-       unsigned char *old_tail;
        struct nfulnl_msg_packet_hdr pmsg;
        struct nlmsghdr *nlh;
        struct nfgenmsg *nfmsg;
        __be32 tmp_uint;
+       sk_buff_data_t old_tail = inst->skb->tail;
 
-       UDEBUG("entered\n");
-
-       old_tail = inst->skb->tail;
        nlh = NLMSG_PUT(inst->skb, 0, 0,
                        NFNL_SUBSYS_ULOG << 8 | NFULNL_MSG_PACKET,
                        sizeof(struct nfgenmsg));
@@ -430,38 +385,33 @@ __build_packet_message(struct nfulnl_instance *inst,
        pmsg.hw_protocol        = skb->protocol;
        pmsg.hook               = hooknum;
 
-       NFA_PUT(inst->skb, NFULA_PACKET_HDR, sizeof(pmsg), &pmsg);
+       NLA_PUT(inst->skb, NFULA_PACKET_HDR, sizeof(pmsg), &pmsg);
 
        if (prefix)
-               NFA_PUT(inst->skb, NFULA_PREFIX, plen, prefix);
+               NLA_PUT(inst->skb, NFULA_PREFIX, plen, prefix);
 
        if (indev) {
-               tmp_uint = htonl(indev->ifindex);
 #ifndef CONFIG_BRIDGE_NETFILTER
-               NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV, sizeof(tmp_uint),
-                       &tmp_uint);
+               NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_INDEV,
+                            htonl(indev->ifindex));
 #else
                if (pf == PF_BRIDGE) {
                        /* Case 1: outdev is physical input device, we need to
                         * look for bridge group (when called from
                         * netfilter_bridge) */
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSINDEV,
-                               sizeof(tmp_uint), &tmp_uint);
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_PHYSINDEV,
+                                    htonl(indev->ifindex));
                        /* this is the bridge group "brX" */
-                       tmp_uint = htonl(indev->br_port->br->dev->ifindex);
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV,
-                               sizeof(tmp_uint), &tmp_uint);
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_INDEV,
+                                    htonl(indev->br_port->br->dev->ifindex));
                } else {
                        /* Case 2: indev is bridge group, we need to look for
                         * physical device (when called from ipv4) */
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_INDEV,
-                               sizeof(tmp_uint), &tmp_uint);
-                       if (skb->nf_bridge && skb->nf_bridge->physindev) {
-                               tmp_uint =
-                                   htonl(skb->nf_bridge->physindev->ifindex);
-                               NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSINDEV,
-                                       sizeof(tmp_uint), &tmp_uint);
-                       }
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_INDEV,
+                                    htonl(indev->ifindex));
+                       if (skb->nf_bridge && skb->nf_bridge->physindev)
+                               NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_PHYSINDEV,
+                                            htonl(skb->nf_bridge->physindev->ifindex));
                }
 #endif
        }
@@ -469,93 +419,97 @@ __build_packet_message(struct nfulnl_instance *inst,
        if (outdev) {
                tmp_uint = htonl(outdev->ifindex);
 #ifndef CONFIG_BRIDGE_NETFILTER
-               NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV, sizeof(tmp_uint),
-                       &tmp_uint);
+               NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_OUTDEV,
+                            htonl(outdev->ifindex));
 #else
                if (pf == PF_BRIDGE) {
                        /* Case 1: outdev is physical output device, we need to
                         * look for bridge group (when called from
                         * netfilter_bridge) */
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSOUTDEV,
-                               sizeof(tmp_uint), &tmp_uint);
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_PHYSOUTDEV,
+                                    htonl(outdev->ifindex));
                        /* this is the bridge group "brX" */
-                       tmp_uint = htonl(outdev->br_port->br->dev->ifindex);
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV,
-                               sizeof(tmp_uint), &tmp_uint);
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_OUTDEV,
+                                    htonl(outdev->br_port->br->dev->ifindex));
                } else {
                        /* Case 2: indev is a bridge group, we need to look
                         * for physical device (when called from ipv4) */
-                       NFA_PUT(inst->skb, NFULA_IFINDEX_OUTDEV,
-                               sizeof(tmp_uint), &tmp_uint);
-                       if (skb->nf_bridge) {
-                               tmp_uint =
-                                   htonl(skb->nf_bridge->physoutdev->ifindex);
-                               NFA_PUT(inst->skb, NFULA_IFINDEX_PHYSOUTDEV,
-                                       sizeof(tmp_uint), &tmp_uint);
-                       }
+                       NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_OUTDEV,
+                                    htonl(outdev->ifindex));
+                       if (skb->nf_bridge && skb->nf_bridge->physoutdev)
+                               NLA_PUT_BE32(inst->skb, NFULA_IFINDEX_PHYSOUTDEV,
+                                            htonl(skb->nf_bridge->physoutdev->ifindex));
                }
 #endif
        }
 
-       if (skb->mark) {
-               tmp_uint = htonl(skb->mark);
-               NFA_PUT(inst->skb, NFULA_MARK, sizeof(tmp_uint), &tmp_uint);
-       }
+       if (skb->mark)
+               NLA_PUT_BE32(inst->skb, NFULA_MARK, htonl(skb->mark));
 
-       if (indev && skb->dev && skb->dev->hard_header_parse) {
+       if (indev && skb->dev) {
                struct nfulnl_msg_packet_hw phw;
-               int len = skb->dev->hard_header_parse((struct sk_buff *)skb,
-                                                   phw.hw_addr);
-               phw.hw_addrlen = htons(len);
-               NFA_PUT(inst->skb, NFULA_HWADDR, sizeof(phw), &phw);
+               int len = dev_parse_header(skb, phw.hw_addr);
+               if (len > 0) {
+                       phw.hw_addrlen = htons(len);
+                       NLA_PUT(inst->skb, NFULA_HWADDR, sizeof(phw), &phw);
+               }
        }
 
-       if (skb->tstamp.off_sec) {
-               struct nfulnl_msg_packet_timestamp ts;
+       if (indev && skb_mac_header_was_set(skb)) {
+               NLA_PUT_BE16(inst->skb, NFULA_HWTYPE, htons(skb->dev->type));
+               NLA_PUT_BE16(inst->skb, NFULA_HWLEN,
+                            htons(skb->dev->hard_header_len));
+               NLA_PUT(inst->skb, NFULA_HWHEADER, skb->dev->hard_header_len,
+                       skb_mac_header(skb));
+       }
 
-               ts.sec = cpu_to_be64(skb->tstamp.off_sec);
-               ts.usec = cpu_to_be64(skb->tstamp.off_usec);
+       if (skb->tstamp.tv64) {
+               struct nfulnl_msg_packet_timestamp ts;
+               struct timeval tv = ktime_to_timeval(skb->tstamp);
+               ts.sec = cpu_to_be64(tv.tv_sec);
+               ts.usec = cpu_to_be64(tv.tv_usec);
 
-               NFA_PUT(inst->skb, NFULA_TIMESTAMP, sizeof(ts), &ts);
+               NLA_PUT(inst->skb, NFULA_TIMESTAMP, sizeof(ts), &ts);
        }
 
        /* UID */
        if (skb->sk) {
                read_lock_bh(&skb->sk->sk_callback_lock);
                if (skb->sk->sk_socket && skb->sk->sk_socket->file) {
-                       __be32 uid = htonl(skb->sk->sk_socket->file->f_uid);
-                       /* need to unlock here since NFA_PUT may goto */
+                       struct file *file = skb->sk->sk_socket->file;
+                       __be32 uid = htonl(file->f_cred->fsuid);
+                       __be32 gid = htonl(file->f_cred->fsgid);
+                       /* need to unlock here since NLA_PUT may goto */
                        read_unlock_bh(&skb->sk->sk_callback_lock);
-                       NFA_PUT(inst->skb, NFULA_UID, sizeof(uid), &uid);
+                       NLA_PUT_BE32(inst->skb, NFULA_UID, uid);
+                       NLA_PUT_BE32(inst->skb, NFULA_GID, gid);
                } else
                        read_unlock_bh(&skb->sk->sk_callback_lock);
        }
 
        /* local sequence number */
-       if (inst->flags & NFULNL_CFG_F_SEQ) {
-               tmp_uint = htonl(inst->seq++);
-               NFA_PUT(inst->skb, NFULA_SEQ, sizeof(tmp_uint), &tmp_uint);
-       }
+       if (inst->flags & NFULNL_CFG_F_SEQ)
+               NLA_PUT_BE32(inst->skb, NFULA_SEQ, htonl(inst->seq++));
+
        /* global sequence number */
-       if (inst->flags & NFULNL_CFG_F_SEQ_GLOBAL) {
-               tmp_uint = htonl(atomic_inc_return(&global_seq));
-               NFA_PUT(inst->skb, NFULA_SEQ_GLOBAL, sizeof(tmp_uint), &tmp_uint);
-       }
+       if (inst->flags & NFULNL_CFG_F_SEQ_GLOBAL)
+               NLA_PUT_BE32(inst->skb, NFULA_SEQ_GLOBAL,
+                            htonl(atomic_inc_return(&global_seq)));
 
        if (data_len) {
-               struct nfattr *nfa;
-               int size = NFA_LENGTH(data_len);
+               struct nlattr *nla;
+               int size = nla_attr_size(data_len);
 
-               if (skb_tailroom(inst->skb) < (int)NFA_SPACE(data_len)) {
+               if (skb_tailroom(inst->skb) < nla_total_size(data_len)) {
                        printk(KERN_WARNING "nfnetlink_log: no tailroom!\n");
                        goto nlmsg_failure;
                }
 
-               nfa = (struct nfattr *)skb_put(inst->skb, NFA_ALIGN(size));
-               nfa->nfa_type = NFULA_PAYLOAD;
-               nfa->nfa_len = size;
+               nla = (struct nlattr *)skb_put(inst->skb, nla_total_size(data_len));
+               nla->nla_type = NFULA_PAYLOAD;
+               nla->nla_len = size;
 
-               if (skb_copy_bits(skb, 0, NFA_DATA(nfa), data_len))
+               if (skb_copy_bits(skb, 0, nla_data(nla), data_len))
                        BUG();
        }
 
@@ -563,8 +517,7 @@ __build_packet_message(struct nfulnl_instance *inst,
        return 0;
 
 nlmsg_failure:
-       UDEBUG("nlmsg_failure\n");
-nfattr_failure:
+nla_put_failure:
        PRINTR(KERN_ERR "nfnetlink_log: error creating log nlmsg\n");
        return -1;
 }
@@ -583,8 +536,8 @@ static struct nf_loginfo default_loginfo = {
 };
 
 /* log handler for internal netfilter logging api */
-static void
-nfulnl_log_packet(unsigned int pf,
+void
+nfulnl_log_packet(u_int8_t pf,
                  unsigned int hooknum,
                  const struct sk_buff *skb,
                  const struct net_device *in,
@@ -596,7 +549,6 @@ nfulnl_log_packet(unsigned int pf,
        struct nfulnl_instance *inst;
        const struct nf_loginfo *li;
        unsigned int qthreshold;
-       unsigned int nlbufsiz;
        unsigned int plen;
 
        if (li_user && li_user->type == NF_LOG_TYPE_ULOG)
@@ -606,48 +558,49 @@ nfulnl_log_packet(unsigned int pf,
 
        inst = instance_lookup_get(li->u.ulog.group);
        if (!inst)
-               inst = instance_lookup_get(0);
-       if (!inst) {
-               PRINTR("nfnetlink_log: trying to log packet, "
-                       "but no instance for group %u\n", li->u.ulog.group);
                return;
-       }
 
        plen = 0;
        if (prefix)
-               plen = strlen(prefix);
+               plen = strlen(prefix) + 1;
 
-       /* all macros expand to constant values at compile time */
        /* FIXME: do we want to make the size calculation conditional based on
         * what is actually present?  way more branches and checks, but more
         * memory efficient... */
        size =    NLMSG_SPACE(sizeof(struct nfgenmsg))
-               + NFA_SPACE(sizeof(struct nfulnl_msg_packet_hdr))
-               + NFA_SPACE(sizeof(u_int32_t))  /* ifindex */
-               + NFA_SPACE(sizeof(u_int32_t))  /* ifindex */
+               + nla_total_size(sizeof(struct nfulnl_msg_packet_hdr))
+               + nla_total_size(sizeof(u_int32_t))     /* ifindex */
+               + nla_total_size(sizeof(u_int32_t))     /* ifindex */
 #ifdef CONFIG_BRIDGE_NETFILTER
-               + NFA_SPACE(sizeof(u_int32_t))  /* ifindex */
-               + NFA_SPACE(sizeof(u_int32_t))  /* ifindex */
+               + nla_total_size(sizeof(u_int32_t))     /* ifindex */
+               + nla_total_size(sizeof(u_int32_t))     /* ifindex */
 #endif
-               + NFA_SPACE(sizeof(u_int32_t))  /* mark */
-               + NFA_SPACE(sizeof(u_int32_t))  /* uid */
-               + NFA_SPACE(plen)               /* prefix */
-               + NFA_SPACE(sizeof(struct nfulnl_msg_packet_hw))
-               + NFA_SPACE(sizeof(struct nfulnl_msg_packet_timestamp));
+               + nla_total_size(sizeof(u_int32_t))     /* mark */
+               + nla_total_size(sizeof(u_int32_t))     /* uid */
+               + nla_total_size(sizeof(u_int32_t))     /* gid */
+               + nla_total_size(plen)                  /* prefix */
+               + nla_total_size(sizeof(struct nfulnl_msg_packet_hw))
+               + nla_total_size(sizeof(struct nfulnl_msg_packet_timestamp));
 
-       UDEBUG("initial size=%u\n", size);
+       if (in && skb_mac_header_was_set(skb)) {
+               size +=   nla_total_size(skb->dev->hard_header_len)
+                       + nla_total_size(sizeof(u_int16_t))     /* hwtype */
+                       + nla_total_size(sizeof(u_int16_t));    /* hwlen */
+       }
 
        spin_lock_bh(&inst->lock);
 
        if (inst->flags & NFULNL_CFG_F_SEQ)
-               size += NFA_SPACE(sizeof(u_int32_t));
+               size += nla_total_size(sizeof(u_int32_t));
        if (inst->flags & NFULNL_CFG_F_SEQ_GLOBAL)
-               size += NFA_SPACE(sizeof(u_int32_t));
+               size += nla_total_size(sizeof(u_int32_t));
 
        qthreshold = inst->qthreshold;
        /* per-rule qthreshold overrides per-instance */
-       if (qthreshold > li->u.ulog.qthreshold)
-               qthreshold = li->u.ulog.qthreshold;
+       if (li->u.ulog.qthreshold)
+               if (qthreshold > li->u.ulog.qthreshold)
+                       qthreshold = li->u.ulog.qthreshold;
+
 
        switch (inst->copy_mode) {
        case NFULNL_COPY_META:
@@ -662,51 +615,36 @@ nfulnl_log_packet(unsigned int pf,
                else
                        data_len = inst->copy_range;
 
-               size += NFA_SPACE(data_len);
-               UDEBUG("copy_packet, therefore size now %u\n", size);
+               size += nla_total_size(data_len);
                break;
 
        default:
-               spin_unlock_bh(&inst->lock);
-               instance_put(inst);
-               return;
+               goto unlock_and_release;
        }
 
-       if (size > inst->nlbufsiz)
-               nlbufsiz = size;
-       else
-               nlbufsiz = inst->nlbufsiz;
-
-       if (!inst->skb) {
-               if (!(inst->skb = nfulnl_alloc_skb(nlbufsiz, size))) {
-                       UDEBUG("error in nfulnl_alloc_skb(%u, %u)\n",
-                               inst->nlbufsiz, size);
-                       goto alloc_failure;
-               }
-       } else if (inst->qlen >= qthreshold ||
-                  size > skb_tailroom(inst->skb)) {
+       if (inst->skb &&
+           size > skb_tailroom(inst->skb) - sizeof(struct nfgenmsg)) {
                /* either the queue len is too high or we don't have
                 * enough room in the skb left. flush to userspace. */
-               UDEBUG("flushing old skb\n");
-
-               __nfulnl_send(inst);
+               __nfulnl_flush(inst);
+       }
 
-               if (!(inst->skb = nfulnl_alloc_skb(nlbufsiz, size))) {
-                       UDEBUG("error in nfulnl_alloc_skb(%u, %u)\n",
-                               inst->nlbufsiz, size);
+       if (!inst->skb) {
+               inst->skb = nfulnl_alloc_skb(inst->nlbufsiz, size);
+               if (!inst->skb)
                        goto alloc_failure;
-               }
        }
 
-       UDEBUG("qlen %d, qthreshold %d\n", inst->qlen, qthreshold);
        inst->qlen++;
 
        __build_packet_message(inst, skb, data_len, pf,
                                hooknum, in, out, li, prefix, plen);
 
+       if (inst->qlen >= qthreshold)
+               __nfulnl_flush(inst);
        /* timer_pending always called within inst->lock, so there
         * is no chance of a race here */
-       if (!timer_pending(&inst->timer)) {
+       else if (!timer_pending(&inst->timer)) {
                instance_get(inst);
                inst->timer.expires = jiffies + (inst->flushtimeout*HZ/100);
                add_timer(&inst->timer);
@@ -718,10 +656,10 @@ unlock_and_release:
        return;
 
 alloc_failure:
-       UDEBUG("error allocating skb\n");
        /* FIXME: statistics */
        goto unlock_and_release;
 }
+EXPORT_SYMBOL_GPL(nfulnl_log_packet);
 
 static int
 nfulnl_rcv_nl_event(struct notifier_block *this,
@@ -729,8 +667,7 @@ nfulnl_rcv_nl_event(struct notifier_block *this,
 {
        struct netlink_notify *n = ptr;
 
-       if (event == NETLINK_URELEASE &&
-           n->protocol == NETLINK_NETFILTER && n->pid) {
+       if (event == NETLINK_URELEASE && n->protocol == NETLINK_NETFILTER) {
                int i;
 
                /* destroy all instances for this pid */
@@ -741,8 +678,8 @@ nfulnl_rcv_nl_event(struct notifier_block *this,
                        struct hlist_head *head = &instance_table[i];
 
                        hlist_for_each_entry_safe(inst, tmp, t2, head, hlist) {
-                               UDEBUG("node = %p\n", inst);
-                               if (n->pid == inst->peer_pid)
+                               if ((net_eq(n->net, &init_net)) &&
+                                   (n->pid == inst->peer_pid))
                                        __instance_destroy(inst);
                        }
                }
@@ -757,65 +694,59 @@ static struct notifier_block nfulnl_rtnl_notifier = {
 
 static int
 nfulnl_recv_unsupp(struct sock *ctnl, struct sk_buff *skb,
-                 struct nlmsghdr *nlh, struct nfattr *nfqa[], int *errp)
+                  const struct nlmsghdr *nlh,
+                  const struct nlattr * const nfqa[])
 {
        return -ENOTSUPP;
 }
 
-static struct nf_logger nfulnl_logger = {
+static struct nf_logger nfulnl_logger __read_mostly = {
        .name   = "nfnetlink_log",
        .logfn  = &nfulnl_log_packet,
        .me     = THIS_MODULE,
 };
 
-static const int nfula_min[NFULA_MAX] = {
-       [NFULA_PACKET_HDR-1]    = sizeof(struct nfulnl_msg_packet_hdr),
-       [NFULA_MARK-1]          = sizeof(u_int32_t),
-       [NFULA_TIMESTAMP-1]     = sizeof(struct nfulnl_msg_packet_timestamp),
-       [NFULA_IFINDEX_INDEV-1] = sizeof(u_int32_t),
-       [NFULA_IFINDEX_OUTDEV-1]= sizeof(u_int32_t),
-       [NFULA_IFINDEX_PHYSINDEV-1]     = sizeof(u_int32_t),
-       [NFULA_IFINDEX_PHYSOUTDEV-1]    = sizeof(u_int32_t),
-       [NFULA_HWADDR-1]        = sizeof(struct nfulnl_msg_packet_hw),
-       [NFULA_PAYLOAD-1]       = 0,
-       [NFULA_PREFIX-1]        = 0,
-       [NFULA_UID-1]           = sizeof(u_int32_t),
-       [NFULA_SEQ-1]           = sizeof(u_int32_t),
-       [NFULA_SEQ_GLOBAL-1]    = sizeof(u_int32_t),
-};
-
-static const int nfula_cfg_min[NFULA_CFG_MAX] = {
-       [NFULA_CFG_CMD-1]       = sizeof(struct nfulnl_msg_config_cmd),
-       [NFULA_CFG_MODE-1]      = sizeof(struct nfulnl_msg_config_mode),
-       [NFULA_CFG_TIMEOUT-1]   = sizeof(u_int32_t),
-       [NFULA_CFG_QTHRESH-1]   = sizeof(u_int32_t),
-       [NFULA_CFG_NLBUFSIZ-1]  = sizeof(u_int32_t),
-       [NFULA_CFG_FLAGS-1]     = sizeof(u_int16_t),
+static const struct nla_policy nfula_cfg_policy[NFULA_CFG_MAX+1] = {
+       [NFULA_CFG_CMD]         = { .len = sizeof(struct nfulnl_msg_config_cmd) },
+       [NFULA_CFG_MODE]        = { .len = sizeof(struct nfulnl_msg_config_mode) },
+       [NFULA_CFG_TIMEOUT]     = { .type = NLA_U32 },
+       [NFULA_CFG_QTHRESH]     = { .type = NLA_U32 },
+       [NFULA_CFG_NLBUFSIZ]    = { .type = NLA_U32 },
+       [NFULA_CFG_FLAGS]       = { .type = NLA_U16 },
 };
 
 static int
 nfulnl_recv_config(struct sock *ctnl, struct sk_buff *skb,
-                  struct nlmsghdr *nlh, struct nfattr *nfula[], int *errp)
+                  const struct nlmsghdr *nlh,
+                  const struct nlattr * const nfula[])
 {
        struct nfgenmsg *nfmsg = NLMSG_DATA(nlh);
        u_int16_t group_num = ntohs(nfmsg->res_id);
        struct nfulnl_instance *inst;
+       struct nfulnl_msg_config_cmd *cmd = NULL;
        int ret = 0;
 
-       UDEBUG("entering for msg %u\n", NFNL_MSG_TYPE(nlh->nlmsg_type));
+       if (nfula[NFULA_CFG_CMD]) {
+               u_int8_t pf = nfmsg->nfgen_family;
+               cmd = nla_data(nfula[NFULA_CFG_CMD]);
 
-       if (nfattr_bad_size(nfula, NFULA_CFG_MAX, nfula_cfg_min)) {
-               UDEBUG("bad attribute size\n");
-               return -EINVAL;
+               /* Commands without queue context */
+               switch (cmd->command) {
+               case NFULNL_CFG_CMD_PF_BIND:
+                       return nf_log_bind_pf(pf, &nfulnl_logger);
+               case NFULNL_CFG_CMD_PF_UNBIND:
+                       nf_log_unbind_pf(pf);
+                       return 0;
+               }
        }
 
        inst = instance_lookup_get(group_num);
-       if (nfula[NFULA_CFG_CMD-1]) {
-               u_int8_t pf = nfmsg->nfgen_family;
-               struct nfulnl_msg_config_cmd *cmd;
-               cmd = NFA_DATA(nfula[NFULA_CFG_CMD-1]);
-               UDEBUG("found CFG_CMD for\n");
+       if (inst && inst->peer_pid != NETLINK_CB(skb).pid) {
+               ret = -EPERM;
+               goto out_put;
+       }
 
+       if (cmd != NULL) {
                switch (cmd->command) {
                case NFULNL_CFG_CMD_BIND:
                        if (inst) {
@@ -825,102 +756,92 @@ nfulnl_recv_config(struct sock *ctnl, struct sk_buff *skb,
 
                        inst = instance_create(group_num,
                                               NETLINK_CB(skb).pid);
-                       if (!inst) {
-                               ret = -EINVAL;
-                               goto out_put;
+                       if (IS_ERR(inst)) {
+                               ret = PTR_ERR(inst);
+                               goto out;
                        }
                        break;
                case NFULNL_CFG_CMD_UNBIND:
                        if (!inst) {
                                ret = -ENODEV;
-                               goto out_put;
-                       }
-
-                       if (inst->peer_pid != NETLINK_CB(skb).pid) {
-                               ret = -EPERM;
-                               goto out_put;
+                               goto out;
                        }
 
                        instance_destroy(inst);
-                       break;
-               case NFULNL_CFG_CMD_PF_BIND:
-                       UDEBUG("registering log handler for pf=%u\n", pf);
-                       ret = nf_log_register(pf, &nfulnl_logger);
-                       break;
-               case NFULNL_CFG_CMD_PF_UNBIND:
-                       UDEBUG("unregistering log handler for pf=%u\n", pf);
-                       /* This is a bug and a feature.  We cannot unregister
-                        * other handlers, like nfnetlink_inst can */
-                       nf_log_unregister_pf(pf);
-                       break;
+                       goto out;
                default:
-                       ret = -EINVAL;
+                       ret = -ENOTSUPP;
                        break;
                }
-       } else {
-               if (!inst) {
-                       UDEBUG("no config command, and no instance for "
-                               "group=%u pid=%u =>ENOENT\n",
-                               group_num, NETLINK_CB(skb).pid);
-                       ret = -ENOENT;
-                       goto out_put;
-               }
-
-               if (inst->peer_pid != NETLINK_CB(skb).pid) {
-                       UDEBUG("no config command, and wrong pid\n");
-                       ret = -EPERM;
-                       goto out_put;
-               }
        }
 
-       if (nfula[NFULA_CFG_MODE-1]) {
+       if (nfula[NFULA_CFG_MODE]) {
                struct nfulnl_msg_config_mode *params;
-               params = NFA_DATA(nfula[NFULA_CFG_MODE-1]);
+               params = nla_data(nfula[NFULA_CFG_MODE]);
 
+               if (!inst) {
+                       ret = -ENODEV;
+                       goto out;
+               }
                nfulnl_set_mode(inst, params->copy_mode,
                                ntohl(params->copy_range));
        }
 
-       if (nfula[NFULA_CFG_TIMEOUT-1]) {
-               __be32 timeout =
-                       *(__be32 *)NFA_DATA(nfula[NFULA_CFG_TIMEOUT-1]);
+       if (nfula[NFULA_CFG_TIMEOUT]) {
+               __be32 timeout = nla_get_be32(nfula[NFULA_CFG_TIMEOUT]);
 
+               if (!inst) {
+                       ret = -ENODEV;
+                       goto out;
+               }
                nfulnl_set_timeout(inst, ntohl(timeout));
        }
 
-       if (nfula[NFULA_CFG_NLBUFSIZ-1]) {
-               __be32 nlbufsiz =
-                       *(__be32 *)NFA_DATA(nfula[NFULA_CFG_NLBUFSIZ-1]);
+       if (nfula[NFULA_CFG_NLBUFSIZ]) {
+               __be32 nlbufsiz = nla_get_be32(nfula[NFULA_CFG_NLBUFSIZ]);
 
+               if (!inst) {
+                       ret = -ENODEV;
+                       goto out;
+               }
                nfulnl_set_nlbufsiz(inst, ntohl(nlbufsiz));
        }
 
-       if (nfula[NFULA_CFG_QTHRESH-1]) {
-               __be32 qthresh =
-                       *(__be32 *)NFA_DATA(nfula[NFULA_CFG_QTHRESH-1]);
+       if (nfula[NFULA_CFG_QTHRESH]) {
+               __be32 qthresh = nla_get_be32(nfula[NFULA_CFG_QTHRESH]);
 
+               if (!inst) {
+                       ret = -ENODEV;
+                       goto out;
+               }
                nfulnl_set_qthresh(inst, ntohl(qthresh));
        }
 
-       if (nfula[NFULA_CFG_FLAGS-1]) {
-               __be16 flags =
-                       *(__be16 *)NFA_DATA(nfula[NFULA_CFG_FLAGS-1]);
+       if (nfula[NFULA_CFG_FLAGS]) {
+               __be16 flags = nla_get_be16(nfula[NFULA_CFG_FLAGS]);
+
+               if (!inst) {
+                       ret = -ENODEV;
+                       goto out;
+               }
                nfulnl_set_flags(inst, ntohs(flags));
        }
 
 out_put:
        instance_put(inst);
+out:
        return ret;
 }
 
-static struct nfnl_callback nfulnl_cb[NFULNL_MSG_MAX] = {
+static const struct nfnl_callback nfulnl_cb[NFULNL_MSG_MAX] = {
        [NFULNL_MSG_PACKET]     = { .call = nfulnl_recv_unsupp,
                                    .attr_count = NFULA_MAX, },
        [NFULNL_MSG_CONFIG]     = { .call = nfulnl_recv_config,
-                                   .attr_count = NFULA_CFG_MAX, },
+                                   .attr_count = NFULA_CFG_MAX,
+                                   .policy = nfula_cfg_policy },
 };
 
-static struct nfnetlink_subsystem nfulnl_subsys = {
+static const struct nfnetlink_subsystem nfulnl_subsys = {
        .name           = "log",
        .subsys_id      = NFNL_SUBSYS_ULOG,
        .cb_count       = NFULNL_MSG_MAX,
@@ -932,10 +853,8 @@ struct iter_state {
        unsigned int bucket;
 };
 
-static struct hlist_node *get_first(struct seq_file *seq)
+static struct hlist_node *get_first(struct iter_state *st)
 {
-       struct iter_state *st = seq->private;
-
        if (!st)
                return NULL;
 
@@ -946,10 +865,8 @@ static struct hlist_node *get_first(struct seq_file *seq)
        return NULL;
 }
 
-static struct hlist_node *get_next(struct seq_file *seq, struct hlist_node *h)
+static struct hlist_node *get_next(struct iter_state *st, struct hlist_node *h)
 {
-       struct iter_state *st = seq->private;
-
        h = h->next;
        while (!h) {
                if (++st->bucket >= INSTANCE_BUCKETS)
@@ -960,30 +877,32 @@ static struct hlist_node *get_next(struct seq_file *seq, struct hlist_node *h)
        return h;
 }
 
-static struct hlist_node *get_idx(struct seq_file *seq, loff_t pos)
+static struct hlist_node *get_idx(struct iter_state *st, loff_t pos)
 {
        struct hlist_node *head;
-       head = get_first(seq);
+       head = get_first(st);
 
        if (head)
-               while (pos && (head = get_next(seq, head)))
+               while (pos && (head = get_next(st, head)))
                        pos--;
        return pos ? NULL : head;
 }
 
 static void *seq_start(struct seq_file *seq, loff_t *pos)
+       __acquires(instances_lock)
 {
        read_lock_bh(&instances_lock);
-       return get_idx(seq, *pos);
+       return get_idx(seq->private, *pos);
 }
 
 static void *seq_next(struct seq_file *s, void *v, loff_t *pos)
 {
        (*pos)++;
-       return get_next(s, v);
+       return get_next(s->private, v);
 }
 
 static void seq_stop(struct seq_file *s, void *v)
+       __releases(instances_lock)
 {
        read_unlock_bh(&instances_lock);
 }
@@ -999,7 +918,7 @@ static int seq_show(struct seq_file *s, void *v)
                          inst->flushtimeout, atomic_read(&inst->use));
 }
 
-static struct seq_operations nful_seq_ops = {
+static const struct seq_operations nful_seq_ops = {
        .start  = seq_start,
        .next   = seq_next,
        .stop   = seq_stop,
@@ -1008,22 +927,8 @@ static struct seq_operations nful_seq_ops = {
 
 static int nful_open(struct inode *inode, struct file *file)
 {
-       struct seq_file *seq;
-       struct iter_state *is;
-       int ret;
-
-       is = kzalloc(sizeof(*is), GFP_KERNEL);
-       if (!is)
-               return -ENOMEM;
-       ret = seq_open(file, &nful_seq_ops);
-       if (ret < 0)
-               goto out_free;
-       seq = file->private_data;
-       seq->private = is;
-       return ret;
-out_free:
-       kfree(is);
-       return ret;
+       return seq_open_private(file, &nful_seq_ops,
+                       sizeof(struct iter_state));
 }
 
 static const struct file_operations nful_file_ops = {
@@ -1039,9 +944,6 @@ static const struct file_operations nful_file_ops = {
 static int __init nfnetlink_log_init(void)
 {
        int i, status = -ENOMEM;
-#ifdef CONFIG_PROC_FS
-       struct proc_dir_entry *proc_nful;
-#endif
 
        for (i = 0; i < INSTANCE_BUCKETS; i++)
                INIT_HLIST_HEAD(&instance_table[i]);
@@ -1058,19 +960,25 @@ static int __init nfnetlink_log_init(void)
                goto cleanup_netlink_notifier;
        }
 
-#ifdef CONFIG_PROC_FS
-       proc_nful = create_proc_entry("nfnetlink_log", 0440,
-                                     proc_net_netfilter);
-       if (!proc_nful)
+       status = nf_log_register(NFPROTO_UNSPEC, &nfulnl_logger);
+       if (status < 0) {
+               printk(KERN_ERR "log: failed to register logger\n");
                goto cleanup_subsys;
-       proc_nful->proc_fops = &nful_file_ops;
+       }
+
+#ifdef CONFIG_PROC_FS
+       if (!proc_create("nfnetlink_log", 0440,
+                        proc_net_netfilter, &nful_file_ops))
+               goto cleanup_logger;
 #endif
        return status;
 
 #ifdef CONFIG_PROC_FS
+cleanup_logger:
+       nf_log_unregister(&nfulnl_logger);
+#endif
 cleanup_subsys:
        nfnetlink_subsys_unregister(&nfulnl_subsys);
-#endif
 cleanup_netlink_notifier:
        netlink_unregister_notifier(&nfulnl_rtnl_notifier);
        return status;