[NETFILTER]: Core changes required by upcoming nfnetlink_queue code

- split netfiler verdict in 16bit verdict and 16bit queue number
- add 'queuenum' argument to nf_queue_outfn_t and its users ip[6]_queue
- move NFNL_SUBSYS_ definitions from enum to #define
- introduce autoloading for nfnetlink subsystem modules
- add MODULE_ALIAS_NFNL_SUBSYS macro
- add nf_unregister_queue_handlers() to register all handlers for a given
  nf_queue_outfn_t
- add more verbose DEBUGP macro definition to nfnetlink.c
- make nfnetlink_subsys_register fail if subsys already exists
- add some more comments and debug statements to nfnetlink.c

Signed-off-by: Harald Welte <laforge@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

6 files changed, 83 insertions, 25 deletions

diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h
index d163e20ca8d9..711e05f33d68 100644
--- a/include/linux/netfilter.h
+++ b/include/linux/netfilter.h
@@ -21,6 +21,16 @@
 #define NF_STOP 5
 #define NF_MAX_VERDICT NF_STOP
 
+/* we overload the higher bits for encoding auxiliary data such as the queue
+ * number. Not nice, but better than additional function arguments. */
+#define NF_VERDICT_MASK 0x0000ffff
+#define NF_VERDICT_BITS 16
+
+#define NF_VERDICT_QMASK 0xffff0000
+#define NF_VERDICT_QBITS 16
+
+#define NF_QUEUE_NR(x) ((x << NF_VERDICT_QBITS) & NF_VERDICT_QMASK || NF_QUEUE)
+
 /* only for userspace compatibility */
 #ifndef __KERNEL__
 /* Generic cache responses from hook functions.
@@ -179,10 +189,12 @@ int nf_getsockopt(struct sock *sk, int pf, int optval, char __user *opt,
 
 /* Packet queuing */
 typedef int (*nf_queue_outfn_t)(struct sk_buff *skb, 
-                                struct nf_info *info, void *data);
+                                struct nf_info *info,
+                                unsigned int queuenum, void *data);
 extern int nf_register_queue_handler(int pf, 
                                      nf_queue_outfn_t outfn, void *data);
 extern int nf_unregister_queue_handler(int pf);
+extern void nf_unregister_queue_handlers(nf_queue_outfn_t outfn);
 extern void nf_reinject(struct sk_buff *skb,
                         struct nf_info *info,
                         unsigned int verdict);
diff --git a/include/linux/netfilter/nfnetlink.h b/include/linux/netfilter/nfnetlink.h
index ace7a7be0742..561f9df28808 100644
--- a/include/linux/netfilter/nfnetlink.h
+++ b/include/linux/netfilter/nfnetlink.h
@@ -69,15 +69,14 @@ struct nfgenmsg {
 #define NFNL_SUBSYS_ID(x)       ((x & 0xff00) >> 8)
 #define NFNL_MSG_TYPE(x)        (x & 0x00ff)
 
-enum nfnl_subsys_id {
-        NFNL_SUBSYS_NONE = 0,
-        NFNL_SUBSYS_CTNETLINK,
-        NFNL_SUBSYS_CTNETLINK_EXP,
-        NFNL_SUBSYS_IPTNETLINK,
-        NFNL_SUBSYS_QUEUE,
-        NFNL_SUBSYS_ULOG,
-        NFNL_SUBSYS_COUNT,
-};
+/* No enum here, otherwise __stringify() trick of MODULE_ALIAS_NFNL_SUBSYS()
+ * won't work anymore */
+#define NFNL_SUBSYS_NONE                0
+#define NFNL_SUBSYS_CTNETLINK           1
+#define NFNL_SUBSYS_CTNETLINK_EXP       2
+#define NFNL_SUBSYS_QUEUE               3
+#define NFNL_SUBSYS_ULOG                4
+#define NFNL_SUBSYS_COUNT               5
 
 #ifdef __KERNEL__
 
@@ -142,5 +141,8 @@ extern int nfnetlink_send(struct sk_buff *skb, u32 pid, unsigned group,
                           int echo);
 extern int nfnetlink_unicast(struct sk_buff *skb, u_int32_t pid, int flags);
 
+#define MODULE_ALIAS_NFNL_SUBSYS(subsys) \
+        MODULE_ALIAS("nfnetlink-subsys-" __stringify(subsys))
+
 #endif  /* __KERNEL__ */
 #endif  /* _NFNETLINK_H */
diff --git a/net/core/netfilter.c b/net/core/netfilter.c
index 1ed4f3110421..3e38084ac2bd 100644
--- a/net/core/netfilter.c
+++ b/net/core/netfilter.c
@@ -221,7 +221,8 @@ static unsigned int nf_iterate(struct list_head *head,
                 verdict = elem->hook(hook, skb, indev, outdev, okfn);
                 if (verdict != NF_ACCEPT) {
 #ifdef CONFIG_NETFILTER_DEBUG
-                        if (unlikely(verdict > NF_MAX_VERDICT)) {
+                        if (unlikely((verdict & NF_VERDICT_MASK)
+                                                        > NF_MAX_VERDICT)) {
                                 NFDEBUG("Evil return from %p(%u).\n",
                                         elem->hook, hook);
                                 continue;
@@ -239,6 +240,9 @@ int nf_register_queue_handler(int pf, nf_queue_outfn_t outfn, void *data)
 {      
         int ret;
 
+        if (pf >= NPROTO)
+                return -EINVAL;
+
         write_lock_bh(&queue_handler_lock);
         if (queue_handler[pf].outfn)
                 ret = -EBUSY;
@@ -255,6 +259,9 @@ int nf_register_queue_handler(int pf, nf_queue_outfn_t outfn, void *data)
 /* The caller must flush their queue before this */
 int nf_unregister_queue_handler(int pf)
 {
+        if (pf >= NPROTO)
+                return -EINVAL;
+
         write_lock_bh(&queue_handler_lock);
         queue_handler[pf].outfn = NULL;
         queue_handler[pf].data = NULL;
@@ -286,6 +293,20 @@ int nf_unregister_queue_rerouter(int pf)
         return 0;
 }
 
+void nf_unregister_queue_handlers(nf_queue_outfn_t outfn)
+{
+        int pf;
+
+        write_lock_bh(&queue_handler_lock);
+        for (pf = 0; pf < NPROTO; pf++)  {
+                if (queue_handler[pf].outfn == outfn) {
+                        queue_handler[pf].outfn = NULL;
+                        queue_handler[pf].data = NULL;
+                }
+        }
+        write_unlock_bh(&queue_handler_lock);
+}
+
 /* 
  * Any packet that leaves via this function must come back 
  * through nf_reinject().
@@ -295,7 +316,8 @@ static int nf_queue(struct sk_buff **skb,
                     int pf, unsigned int hook,
                     struct net_device *indev,
                     struct net_device *outdev,
-                    int (*okfn)(struct sk_buff *))
+                    int (*okfn)(struct sk_buff *),
+                    unsigned int queuenum)
 {
         int status;
         struct nf_info *info;
@@ -347,7 +369,8 @@ static int nf_queue(struct sk_buff **skb,
         if (queue_rerouter[pf].save)
                 queue_rerouter[pf].save(*skb, info);
 
-        status = queue_handler[pf].outfn(*skb, info, queue_handler[pf].data);
+        status = queue_handler[pf].outfn(*skb, info, queuenum,
+                                         queue_handler[pf].data);
 
         if (status >= 0 && queue_rerouter[pf].reroute)
                 status = queue_rerouter[pf].reroute(skb, info);
@@ -397,9 +420,10 @@ next_hook:
         } else if (verdict == NF_DROP) {
                 kfree_skb(*pskb);
                 ret = -EPERM;
-        } else if (verdict == NF_QUEUE) {
+        } else if ((verdict & NF_VERDICT_MASK)  == NF_QUEUE) {
                 NFDEBUG("nf_hook: Verdict = QUEUE.\n");
-                if (!nf_queue(pskb, elem, pf, hook, indev, outdev, okfn))
+                if (!nf_queue(pskb, elem, pf, hook, indev, outdev, okfn,
+                              verdict >> NF_VERDICT_BITS))
                         goto next_hook;
         }
 unlock:
@@ -456,14 +480,15 @@ void nf_reinject(struct sk_buff *skb, struct nf_info *info,
                                      info->okfn, INT_MIN);
         }
 
-        switch (verdict) {
+        switch (verdict & NF_VERDICT_MASK) {
         case NF_ACCEPT:
                 info->okfn(skb);
                 break;
 
         case NF_QUEUE:
                 if (!nf_queue(&skb, elem, info->pf, info->hook, 
-                              info->indev, info->outdev, info->okfn))
+                              info->indev, info->outdev, info->okfn,
+                              verdict >> NF_VERDICT_BITS))
                         goto next_hook;
                 break;
         }
@@ -613,6 +638,7 @@ EXPORT_SYMBOL(nf_reinject);
 EXPORT_SYMBOL(nf_setsockopt);
 EXPORT_SYMBOL(nf_unregister_hook);
 EXPORT_SYMBOL(nf_unregister_queue_handler);
+EXPORT_SYMBOL_GPL(nf_unregister_queue_handlers);
 EXPORT_SYMBOL_GPL(nf_register_queue_rerouter);
 EXPORT_SYMBOL_GPL(nf_unregister_queue_rerouter);
 EXPORT_SYMBOL(nf_unregister_sockopt);
diff --git a/net/ipv4/netfilter/ip_queue.c b/net/ipv4/netfilter/ip_queue.c
index 78892980f42c..cfc886f382ac 100644
--- a/net/ipv4/netfilter/ip_queue.c
+++ b/net/ipv4/netfilter/ip_queue.c
@@ -280,7 +280,8 @@ nlmsg_failure:
 }
 
 static int
-ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data)
+ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info,
+                   unsigned int queuenum, void *data)
 {
         int status = -EINVAL;
         struct sk_buff *nskb;
diff --git a/net/ipv6/netfilter/ip6_queue.c b/net/ipv6/netfilter/ip6_queue.c
index c45d8f8815de..5af4cee93d9b 100644
--- a/net/ipv6/netfilter/ip6_queue.c
+++ b/net/ipv6/netfilter/ip6_queue.c
@@ -278,7 +278,8 @@ nlmsg_failure:
 }
 
 static int
-ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, void *data)
+ipq_enqueue_packet(struct sk_buff *skb, struct nf_info *info, 
+                   unsigned int queuenum, void *data)
 {
         int status = -EINVAL;
         struct sk_buff *nskb;
diff --git a/net/netfilter/nfnetlink.c b/net/netfilter/nfnetlink.c
index 6210ca42166c..30b25f47f7cc 100644
--- a/net/netfilter/nfnetlink.c
+++ b/net/netfilter/nfnetlink.c
@@ -44,7 +44,9 @@ MODULE_ALIAS_NET_PF_PROTO(PF_NETLINK, NETLINK_NETFILTER);
 static char __initdata nfversion[] = "0.30";
 
 #if 0
-#define DEBUGP printk
+#define DEBUGP(format, args...) \
+                printk(KERN_DEBUG "%s(%d):%s(): " format, __FILE__, \
+                        __LINE__, __FUNCTION__, ## args)
 #else
 #define DEBUGP(format, args...)
 #endif
@@ -67,11 +69,11 @@ int nfnetlink_subsys_register(struct nfnetlink_subsystem *n)
 {
         DEBUGP("registering subsystem ID %u\n", n->subsys_id);
 
-        /* If the netlink socket wasn't created, then fail */
-        if (!nfnl)
-                return -1;
-
         nfnl_lock();
+        if (subsys_table[n->subsys_id]) {
+                nfnl_unlock();
+                return -EBUSY;
+        }
         subsys_table[n->subsys_id] = n;
         nfnl_unlock();
 
@@ -227,8 +229,18 @@ static inline int nfnetlink_rcv_msg(struct sk_buff *skb,
 
         type = nlh->nlmsg_type;
         ss = nfnetlink_get_subsys(type);
-        if (!ss)
+        if (!ss) {
+#ifdef CONFIG_KMOD
+                /* don't call nfnl_shunlock, since it would reenter
+                 * with further packet processing */
+                up(&nfnl_sem);
+                request_module("nfnetlink-subsys-%d", NFNL_SUBSYS_ID(type));
+                nfnl_shlock();
+                ss = nfnetlink_get_subsys(type);
+                if (!ss)
+#endif
                 goto err_inval;
+        }
 
         nc = nfnetlink_find_client(type, ss);
         if (!nc) {
@@ -252,12 +264,14 @@ static inline int nfnetlink_rcv_msg(struct sk_buff *skb,
                 if (err < 0)
                         goto err_inval;
 
+                DEBUGP("calling handler\n");
                 err = nc->call(nfnl, skb, nlh, cda, errp);
                 *errp = err;
                 return err;
         }
 
 err_inval:
+        DEBUGP("returning -EINVAL\n");
         *errp = -EINVAL;
         return -1;
 }
@@ -311,6 +325,8 @@ static void nfnetlink_rcv(struct sock *sk, int len)
                         kfree_skb(skb);
                 }
 
+                /* don't call nfnl_shunlock, since it would reenter
+                 * with further packet processing */
                 up(&nfnl_sem);
         } while(nfnl && nfnl->sk_receive_queue.qlen);
 }