netfilter: ipset: Unified hash type generation
authorJozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Mon, 8 Apr 2013 19:05:44 +0000 (21:05 +0200)
committerPablo Neira Ayuso <pablo@netfilter.org>
Mon, 29 Apr 2013 18:08:56 +0000 (20:08 +0200)
Signed-off-by: Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
include/linux/netfilter/ipset/ip_set_ahash.h [deleted file]
net/netfilter/ipset/ip_set_hash_gen.h [new file with mode: 0644]

diff --git a/include/linux/netfilter/ipset/ip_set_ahash.h b/include/linux/netfilter/ipset/ip_set_ahash.h
deleted file mode 100644 (file)
index 0214c4c..0000000
+++ /dev/null
@@ -1,1241 +0,0 @@
-#ifndef _IP_SET_AHASH_H
-#define _IP_SET_AHASH_H
-
-#include <linux/rcupdate.h>
-#include <linux/jhash.h>
-#include <linux/netfilter/ipset/ip_set_timeout.h>
-
-#define CONCAT(a, b, c)                a##b##c
-#define TOKEN(a, b, c)         CONCAT(a, b, c)
-
-#define type_pf_next           TOKEN(TYPE, PF, _elem)
-
-/* Hashing which uses arrays to resolve clashing. The hash table is resized
- * (doubled) when searching becomes too long.
- * Internally jhash is used with the assumption that the size of the
- * stored data is a multiple of sizeof(u32). If storage supports timeout,
- * the timeout field must be the last one in the data structure - that field
- * is ignored when computing the hash key.
- *
- * Readers and resizing
- *
- * Resizing can be triggered by userspace command only, and those
- * are serialized by the nfnl mutex. During resizing the set is
- * read-locked, so the only possible concurrent operations are
- * the kernel side readers. Those must be protected by proper RCU locking.
- */
-
-/* Number of elements to store in an initial array block */
-#define AHASH_INIT_SIZE                        4
-/* Max number of elements to store in an array block */
-#define AHASH_MAX_SIZE                 (3*AHASH_INIT_SIZE)
-
-/* Max number of elements can be tuned */
-#ifdef IP_SET_HASH_WITH_MULTI
-#define AHASH_MAX(h)                   ((h)->ahash_max)
-
-static inline u8
-tune_ahash_max(u8 curr, u32 multi)
-{
-       u32 n;
-
-       if (multi < curr)
-               return curr;
-
-       n = curr + AHASH_INIT_SIZE;
-       /* Currently, at listing one hash bucket must fit into a message.
-        * Therefore we have a hard limit here.
-        */
-       return n > curr && n <= 64 ? n : curr;
-}
-#define TUNE_AHASH_MAX(h, multi)       \
-       ((h)->ahash_max = tune_ahash_max((h)->ahash_max, multi))
-#else
-#define AHASH_MAX(h)                   AHASH_MAX_SIZE
-#define TUNE_AHASH_MAX(h, multi)
-#endif
-
-/* A hash bucket */
-struct hbucket {
-       void *value;            /* the array of the values */
-       u8 size;                /* size of the array */
-       u8 pos;                 /* position of the first free entry */
-};
-
-/* The hash table: the table size stored here in order to make resizing easy */
-struct htable {
-       u8 htable_bits;         /* size of hash table == 2^htable_bits */
-       struct hbucket bucket[0]; /* hashtable buckets */
-};
-
-#define hbucket(h, i)          (&((h)->bucket[i]))
-
-/* Book-keeping of the prefixes added to the set */
-struct ip_set_hash_nets {
-       u8 cidr;                /* the different cidr values in the set */
-       u32 nets;               /* number of elements per cidr */
-};
-
-/* The generic ip_set hash structure */
-struct ip_set_hash {
-       struct htable *table;   /* the hash table */
-       u32 maxelem;            /* max elements in the hash */
-       u32 elements;           /* current element (vs timeout) */
-       u32 initval;            /* random jhash init value */
-       u32 timeout;            /* timeout value, if enabled */
-       struct timer_list gc;   /* garbage collection when timeout enabled */
-       struct type_pf_next next; /* temporary storage for uadd */
-#ifdef IP_SET_HASH_WITH_MULTI
-       u8 ahash_max;           /* max elements in an array block */
-#endif
-#ifdef IP_SET_HASH_WITH_NETMASK
-       u8 netmask;             /* netmask value for subnets to store */
-#endif
-#ifdef IP_SET_HASH_WITH_RBTREE
-       struct rb_root rbtree;
-#endif
-#ifdef IP_SET_HASH_WITH_NETS
-       struct ip_set_hash_nets nets[0]; /* book-keeping of prefixes */
-#endif
-};
-
-static size_t
-htable_size(u8 hbits)
-{
-       size_t hsize;
-
-       /* We must fit both into u32 in jhash and size_t */
-       if (hbits > 31)
-               return 0;
-       hsize = jhash_size(hbits);
-       if ((((size_t)-1) - sizeof(struct htable))/sizeof(struct hbucket)
-           < hsize)
-               return 0;
-
-       return hsize * sizeof(struct hbucket) + sizeof(struct htable);
-}
-
-/* Compute htable_bits from the user input parameter hashsize */
-static u8
-htable_bits(u32 hashsize)
-{
-       /* Assume that hashsize == 2^htable_bits */
-       u8 bits = fls(hashsize - 1);
-       if (jhash_size(bits) != hashsize)
-               /* Round up to the first 2^n value */
-               bits = fls(hashsize);
-
-       return bits;
-}
-
-#ifdef IP_SET_HASH_WITH_NETS
-#ifdef IP_SET_HASH_WITH_NETS_PACKED
-/* When cidr is packed with nomatch, cidr - 1 is stored in the entry */
-#define CIDR(cidr)     (cidr + 1)
-#else
-#define CIDR(cidr)     (cidr)
-#endif
-
-#define SET_HOST_MASK(family)  (family == AF_INET ? 32 : 128)
-#ifdef IP_SET_HASH_WITH_MULTI
-#define NETS_LENGTH(family)    (SET_HOST_MASK(family) + 1)
-#else
-#define NETS_LENGTH(family)    SET_HOST_MASK(family)
-#endif
-
-/* Network cidr size book keeping when the hash stores different
- * sized networks */
-static void
-add_cidr(struct ip_set_hash *h, u8 cidr, u8 nets_length)
-{
-       int i, j;
-
-       /* Add in increasing prefix order, so larger cidr first */
-       for (i = 0, j = -1; i < nets_length && h->nets[i].nets; i++) {
-               if (j != -1)
-                       continue;
-               else if (h->nets[i].cidr < cidr)
-                       j = i;
-               else if (h->nets[i].cidr == cidr) {
-                       h->nets[i].nets++;
-                       return;
-               }
-       }
-       if (j != -1) {
-               for (; i > j; i--) {
-                       h->nets[i].cidr = h->nets[i - 1].cidr;
-                       h->nets[i].nets = h->nets[i - 1].nets;
-               }
-       }
-       h->nets[i].cidr = cidr;
-       h->nets[i].nets = 1;
-}
-
-static void
-del_cidr(struct ip_set_hash *h, u8 cidr, u8 nets_length)
-{
-       u8 i, j;
-
-       for (i = 0; i < nets_length - 1 && h->nets[i].cidr != cidr; i++)
-               ;
-       h->nets[i].nets--;
-
-       if (h->nets[i].nets != 0)
-               return;
-
-       for (j = i; j < nets_length - 1 && h->nets[j].nets; j++) {
-               h->nets[j].cidr = h->nets[j + 1].cidr;
-               h->nets[j].nets = h->nets[j + 1].nets;
-       }
-}
-#else
-#define NETS_LENGTH(family)            0
-#endif
-
-/* Destroy the hashtable part of the set */
-static void
-ahash_destroy(struct htable *t)
-{
-       struct hbucket *n;
-       u32 i;
-
-       for (i = 0; i < jhash_size(t->htable_bits); i++) {
-               n = hbucket(t, i);
-               if (n->size)
-                       /* FIXME: use slab cache */
-                       kfree(n->value);
-       }
-
-       ip_set_free(t);
-}
-
-/* Calculate the actual memory size of the set data */
-static size_t
-ahash_memsize(const struct ip_set_hash *h, size_t dsize, u8 nets_length)
-{
-       u32 i;
-       struct htable *t = h->table;
-       size_t memsize = sizeof(*h)
-                        + sizeof(*t)
-#ifdef IP_SET_HASH_WITH_NETS
-                        + sizeof(struct ip_set_hash_nets) * nets_length
-#endif
-                        + jhash_size(t->htable_bits) * sizeof(struct hbucket);
-
-       for (i = 0; i < jhash_size(t->htable_bits); i++)
-                       memsize += t->bucket[i].size * dsize;
-
-       return memsize;
-}
-
-/* Flush a hash type of set: destroy all elements */
-static void
-ip_set_hash_flush(struct ip_set *set)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       struct hbucket *n;
-       u32 i;
-
-       for (i = 0; i < jhash_size(t->htable_bits); i++) {
-               n = hbucket(t, i);
-               if (n->size) {
-                       n->size = n->pos = 0;
-                       /* FIXME: use slab cache */
-                       kfree(n->value);
-               }
-       }
-#ifdef IP_SET_HASH_WITH_NETS
-       memset(h->nets, 0, sizeof(struct ip_set_hash_nets)
-                          * NETS_LENGTH(set->family));
-#endif
-       h->elements = 0;
-}
-
-/* Destroy a hash type of set */
-static void
-ip_set_hash_destroy(struct ip_set *set)
-{
-       struct ip_set_hash *h = set->data;
-
-       if (with_timeout(h->timeout))
-               del_timer_sync(&h->gc);
-
-       ahash_destroy(h->table);
-#ifdef IP_SET_HASH_WITH_RBTREE
-       rbtree_destroy(&h->rbtree);
-#endif
-       kfree(h);
-
-       set->data = NULL;
-}
-
-#endif /* _IP_SET_AHASH_H */
-
-#ifndef HKEY_DATALEN
-#define HKEY_DATALEN   sizeof(struct type_pf_elem)
-#endif
-
-#define HKEY(data, initval, htable_bits)                       \
-(jhash2((u32 *)(data), HKEY_DATALEN/sizeof(u32), initval)      \
-       & jhash_mask(htable_bits))
-
-/* Type/family dependent function prototypes */
-
-#define type_pf_data_equal     TOKEN(TYPE, PF, _data_equal)
-#define type_pf_data_isnull    TOKEN(TYPE, PF, _data_isnull)
-#define type_pf_data_copy      TOKEN(TYPE, PF, _data_copy)
-#define type_pf_data_zero_out  TOKEN(TYPE, PF, _data_zero_out)
-#define type_pf_data_netmask   TOKEN(TYPE, PF, _data_netmask)
-#define type_pf_data_list      TOKEN(TYPE, PF, _data_list)
-#define type_pf_data_tlist     TOKEN(TYPE, PF, _data_tlist)
-#define type_pf_data_next      TOKEN(TYPE, PF, _data_next)
-#define type_pf_data_flags     TOKEN(TYPE, PF, _data_flags)
-#define type_pf_data_reset_flags TOKEN(TYPE, PF, _data_reset_flags)
-#ifdef IP_SET_HASH_WITH_NETS
-#define type_pf_data_match     TOKEN(TYPE, PF, _data_match)
-#else
-#define type_pf_data_match(d)  1
-#endif
-
-#define type_pf_elem           TOKEN(TYPE, PF, _elem)
-#define type_pf_telem          TOKEN(TYPE, PF, _telem)
-#define type_pf_data_timeout   TOKEN(TYPE, PF, _data_timeout)
-#define type_pf_data_expired   TOKEN(TYPE, PF, _data_expired)
-#define type_pf_data_timeout_set TOKEN(TYPE, PF, _data_timeout_set)
-
-#define type_pf_elem_add       TOKEN(TYPE, PF, _elem_add)
-#define type_pf_add            TOKEN(TYPE, PF, _add)
-#define type_pf_del            TOKEN(TYPE, PF, _del)
-#define type_pf_test_cidrs     TOKEN(TYPE, PF, _test_cidrs)
-#define type_pf_test           TOKEN(TYPE, PF, _test)
-
-#define type_pf_elem_tadd      TOKEN(TYPE, PF, _elem_tadd)
-#define type_pf_del_telem      TOKEN(TYPE, PF, _ahash_del_telem)
-#define type_pf_expire         TOKEN(TYPE, PF, _expire)
-#define type_pf_tadd           TOKEN(TYPE, PF, _tadd)
-#define type_pf_tdel           TOKEN(TYPE, PF, _tdel)
-#define type_pf_ttest_cidrs    TOKEN(TYPE, PF, _ahash_ttest_cidrs)
-#define type_pf_ttest          TOKEN(TYPE, PF, _ahash_ttest)
-
-#define type_pf_resize         TOKEN(TYPE, PF, _resize)
-#define type_pf_tresize                TOKEN(TYPE, PF, _tresize)
-#define type_pf_flush          ip_set_hash_flush
-#define type_pf_destroy                ip_set_hash_destroy
-#define type_pf_head           TOKEN(TYPE, PF, _head)
-#define type_pf_list           TOKEN(TYPE, PF, _list)
-#define type_pf_tlist          TOKEN(TYPE, PF, _tlist)
-#define type_pf_same_set       TOKEN(TYPE, PF, _same_set)
-#define type_pf_kadt           TOKEN(TYPE, PF, _kadt)
-#define type_pf_uadt           TOKEN(TYPE, PF, _uadt)
-#define type_pf_gc             TOKEN(TYPE, PF, _gc)
-#define type_pf_gc_init                TOKEN(TYPE, PF, _gc_init)
-#define type_pf_variant                TOKEN(TYPE, PF, _variant)
-#define type_pf_tvariant       TOKEN(TYPE, PF, _tvariant)
-
-/* Flavour without timeout */
-
-/* Get the ith element from the array block n */
-#define ahash_data(n, i)       \
-       ((struct type_pf_elem *)((n)->value) + (i))
-
-/* Add an element to the hash table when resizing the set:
- * we spare the maintenance of the internal counters. */
-static int
-type_pf_elem_add(struct hbucket *n, const struct type_pf_elem *value,
-                u8 ahash_max, u32 cadt_flags)
-{
-       struct type_pf_elem *data;
-
-       if (n->pos >= n->size) {
-               void *tmp;
-
-               if (n->size >= ahash_max)
-                       /* Trigger rehashing */
-                       return -EAGAIN;
-
-               tmp = kzalloc((n->size + AHASH_INIT_SIZE)
-                             * sizeof(struct type_pf_elem),
-                             GFP_ATOMIC);
-               if (!tmp)
-                       return -ENOMEM;
-               if (n->size) {
-                       memcpy(tmp, n->value,
-                              sizeof(struct type_pf_elem) * n->size);
-                       kfree(n->value);
-               }
-               n->value = tmp;
-               n->size += AHASH_INIT_SIZE;
-       }
-       data = ahash_data(n, n->pos++);
-       type_pf_data_copy(data, value);
-#ifdef IP_SET_HASH_WITH_NETS
-       /* Resizing won't overwrite stored flags */
-       if (cadt_flags)
-               type_pf_data_flags(data, cadt_flags);
-#endif
-       return 0;
-}
-
-/* Resize a hash: create a new hash table with doubling the hashsize
- * and inserting the elements to it. Repeat until we succeed or
- * fail due to memory pressures. */
-static int
-type_pf_resize(struct ip_set *set, bool retried)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t, *orig = h->table;
-       u8 htable_bits = orig->htable_bits;
-       struct type_pf_elem *data;
-       struct hbucket *n, *m;
-       u32 i, j, flags = 0;
-       int ret;
-
-retry:
-       ret = 0;
-       htable_bits++;
-       pr_debug("attempt to resize set %s from %u to %u, t %p\n",
-                set->name, orig->htable_bits, htable_bits, orig);
-       if (!htable_bits) {
-               /* In case we have plenty of memory :-) */
-               pr_warning("Cannot increase the hashsize of set %s further\n",
-                          set->name);
-               return -IPSET_ERR_HASH_FULL;
-       }
-       t = ip_set_alloc(sizeof(*t)
-                        + jhash_size(htable_bits) * sizeof(struct hbucket));
-       if (!t)
-               return -ENOMEM;
-       t->htable_bits = htable_bits;
-
-       read_lock_bh(&set->lock);
-       for (i = 0; i < jhash_size(orig->htable_bits); i++) {
-               n = hbucket(orig, i);
-               for (j = 0; j < n->pos; j++) {
-                       data = ahash_data(n, j);
-#ifdef IP_SET_HASH_WITH_NETS
-                       flags = 0;
-                       type_pf_data_reset_flags(data, &flags);
-#endif
-                       m = hbucket(t, HKEY(data, h->initval, htable_bits));
-                       ret = type_pf_elem_add(m, data, AHASH_MAX(h), flags);
-                       if (ret < 0) {
-#ifdef IP_SET_HASH_WITH_NETS
-                               type_pf_data_flags(data, flags);
-#endif
-                               read_unlock_bh(&set->lock);
-                               ahash_destroy(t);
-                               if (ret == -EAGAIN)
-                                       goto retry;
-                               return ret;
-                       }
-               }
-       }
-
-       rcu_assign_pointer(h->table, t);
-       read_unlock_bh(&set->lock);
-
-       /* Give time to other readers of the set */
-       synchronize_rcu_bh();
-
-       pr_debug("set %s resized from %u (%p) to %u (%p)\n", set->name,
-                orig->htable_bits, orig, t->htable_bits, t);
-       ahash_destroy(orig);
-
-       return 0;
-}
-
-static inline void
-type_pf_data_next(struct ip_set_hash *h, const struct type_pf_elem *d);
-
-/* Add an element to a hash and update the internal counters when succeeded,
- * otherwise report the proper error code. */
-static int
-type_pf_add(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t;
-       const struct type_pf_elem *d = value;
-       struct hbucket *n;
-       int i, ret = 0;
-       u32 key, multi = 0;
-       u32 cadt_flags = flags >> 16;
-
-       if (h->elements >= h->maxelem) {
-               if (net_ratelimit())
-                       pr_warning("Set %s is full, maxelem %u reached\n",
-                                  set->name, h->maxelem);
-               return -IPSET_ERR_HASH_FULL;
-       }
-
-       rcu_read_lock_bh();
-       t = rcu_dereference_bh(h->table);
-       key = HKEY(value, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++)
-               if (type_pf_data_equal(ahash_data(n, i), d, &multi)) {
-#ifdef IP_SET_HASH_WITH_NETS
-                       if (flags & IPSET_FLAG_EXIST)
-                               /* Support overwriting just the flags */
-                               type_pf_data_flags(ahash_data(n, i),
-                                                  cadt_flags);
-#endif
-                       ret = -IPSET_ERR_EXIST;
-                       goto out;
-               }
-       TUNE_AHASH_MAX(h, multi);
-       ret = type_pf_elem_add(n, value, AHASH_MAX(h), cadt_flags);
-       if (ret != 0) {
-               if (ret == -EAGAIN)
-                       type_pf_data_next(h, d);
-               goto out;
-       }
-
-#ifdef IP_SET_HASH_WITH_NETS
-       add_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
-#endif
-       h->elements++;
-out:
-       rcu_read_unlock_bh();
-       return ret;
-}
-
-/* Delete an element from the hash: swap it with the last element
- * and free up space if possible.
- */
-static int
-type_pf_del(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       const struct type_pf_elem *d = value;
-       struct hbucket *n;
-       int i;
-       struct type_pf_elem *data;
-       u32 key, multi = 0;
-
-       key = HKEY(value, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++) {
-               data = ahash_data(n, i);
-               if (!type_pf_data_equal(data, d, &multi))
-                       continue;
-               if (i != n->pos - 1)
-                       /* Not last one */
-                       type_pf_data_copy(data, ahash_data(n, n->pos - 1));
-
-               n->pos--;
-               h->elements--;
-#ifdef IP_SET_HASH_WITH_NETS
-               del_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
-#endif
-               if (n->pos + AHASH_INIT_SIZE < n->size) {
-                       void *tmp = kzalloc((n->size - AHASH_INIT_SIZE)
-                                           * sizeof(struct type_pf_elem),
-                                           GFP_ATOMIC);
-                       if (!tmp)
-                               return 0;
-                       n->size -= AHASH_INIT_SIZE;
-                       memcpy(tmp, n->value,
-                              n->size * sizeof(struct type_pf_elem));
-                       kfree(n->value);
-                       n->value = tmp;
-               }
-               return 0;
-       }
-
-       return -IPSET_ERR_EXIST;
-}
-
-#ifdef IP_SET_HASH_WITH_NETS
-
-/* Special test function which takes into account the different network
- * sizes added to the set */
-static int
-type_pf_test_cidrs(struct ip_set *set, struct type_pf_elem *d, u32 timeout)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       struct hbucket *n;
-       const struct type_pf_elem *data;
-       int i, j = 0;
-       u32 key, multi = 0;
-       u8 nets_length = NETS_LENGTH(set->family);
-
-       pr_debug("test by nets\n");
-       for (; j < nets_length && h->nets[j].nets && !multi; j++) {
-               type_pf_data_netmask(d, h->nets[j].cidr);
-               key = HKEY(d, h->initval, t->htable_bits);
-               n = hbucket(t, key);
-               for (i = 0; i < n->pos; i++) {
-                       data = ahash_data(n, i);
-                       if (type_pf_data_equal(data, d, &multi))
-                               return type_pf_data_match(data);
-               }
-       }
-       return 0;
-}
-#endif
-
-/* Test whether the element is added to the set */
-static int
-type_pf_test(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       struct type_pf_elem *d = value;
-       struct hbucket *n;
-       const struct type_pf_elem *data;
-       int i;
-       u32 key, multi = 0;
-
-#ifdef IP_SET_HASH_WITH_NETS
-       /* If we test an IP address and not a network address,
-        * try all possible network sizes */
-       if (CIDR(d->cidr) == SET_HOST_MASK(set->family))
-               return type_pf_test_cidrs(set, d, timeout);
-#endif
-
-       key = HKEY(d, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++) {
-               data = ahash_data(n, i);
-               if (type_pf_data_equal(data, d, &multi))
-                       return type_pf_data_match(data);
-       }
-       return 0;
-}
-
-/* Reply a HEADER request: fill out the header part of the set */
-static int
-type_pf_head(struct ip_set *set, struct sk_buff *skb)
-{
-       const struct ip_set_hash *h = set->data;
-       struct nlattr *nested;
-       size_t memsize;
-
-       read_lock_bh(&set->lock);
-       memsize = ahash_memsize(h, with_timeout(h->timeout)
-                                       ? sizeof(struct type_pf_telem)
-                                       : sizeof(struct type_pf_elem),
-                               NETS_LENGTH(set->family));
-       read_unlock_bh(&set->lock);
-
-       nested = ipset_nest_start(skb, IPSET_ATTR_DATA);
-       if (!nested)
-               goto nla_put_failure;
-       if (nla_put_net32(skb, IPSET_ATTR_HASHSIZE,
-                         htonl(jhash_size(h->table->htable_bits))) ||
-           nla_put_net32(skb, IPSET_ATTR_MAXELEM, htonl(h->maxelem)))
-               goto nla_put_failure;
-#ifdef IP_SET_HASH_WITH_NETMASK
-       if (h->netmask != HOST_MASK &&
-           nla_put_u8(skb, IPSET_ATTR_NETMASK, h->netmask))
-               goto nla_put_failure;
-#endif
-       if (nla_put_net32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)) ||
-           nla_put_net32(skb, IPSET_ATTR_MEMSIZE, htonl(memsize)) ||
-           (with_timeout(h->timeout) &&
-            nla_put_net32(skb, IPSET_ATTR_TIMEOUT, htonl(h->timeout))))
-               goto nla_put_failure;
-       ipset_nest_end(skb, nested);
-
-       return 0;
-nla_put_failure:
-       return -EMSGSIZE;
-}
-
-/* Reply a LIST/SAVE request: dump the elements of the specified set */
-static int
-type_pf_list(const struct ip_set *set,
-            struct sk_buff *skb, struct netlink_callback *cb)
-{
-       const struct ip_set_hash *h = set->data;
-       const struct htable *t = h->table;
-       struct nlattr *atd, *nested;
-       const struct hbucket *n;
-       const struct type_pf_elem *data;
-       u32 first = cb->args[2];
-       /* We assume that one hash bucket fills into one page */
-       void *incomplete;
-       int i;
-
-       atd = ipset_nest_start(skb, IPSET_ATTR_ADT);
-       if (!atd)
-               return -EMSGSIZE;
-       pr_debug("list hash set %s\n", set->name);
-       for (; cb->args[2] < jhash_size(t->htable_bits); cb->args[2]++) {
-               incomplete = skb_tail_pointer(skb);
-               n = hbucket(t, cb->args[2]);
-               pr_debug("cb->args[2]: %lu, t %p n %p\n", cb->args[2], t, n);
-               for (i = 0; i < n->pos; i++) {
-                       data = ahash_data(n, i);
-                       pr_debug("list hash %lu hbucket %p i %u, data %p\n",
-                                cb->args[2], n, i, data);
-                       nested = ipset_nest_start(skb, IPSET_ATTR_DATA);
-                       if (!nested) {
-                               if (cb->args[2] == first) {
-                                       nla_nest_cancel(skb, atd);
-                                       return -EMSGSIZE;
-                               } else
-                                       goto nla_put_failure;
-                       }
-                       if (type_pf_data_list(skb, data))
-                               goto nla_put_failure;
-                       ipset_nest_end(skb, nested);
-               }
-       }
-       ipset_nest_end(skb, atd);
-       /* Set listing finished */
-       cb->args[2] = 0;
-
-       return 0;
-
-nla_put_failure:
-       nlmsg_trim(skb, incomplete);
-       ipset_nest_end(skb, atd);
-       if (unlikely(first == cb->args[2])) {
-               pr_warning("Can't list set %s: one bucket does not fit into "
-                          "a message. Please report it!\n", set->name);
-               cb->args[2] = 0;
-               return -EMSGSIZE;
-       }
-       return 0;
-}
-
-static int
-type_pf_kadt(struct ip_set *set, const struct sk_buff *skb,
-            const struct xt_action_param *par,
-            enum ipset_adt adt, const struct ip_set_adt_opt *opt);
-static int
-type_pf_uadt(struct ip_set *set, struct nlattr *tb[],
-            enum ipset_adt adt, u32 *lineno, u32 flags, bool retried);
-
-static const struct ip_set_type_variant type_pf_variant = {
-       .kadt   = type_pf_kadt,
-       .uadt   = type_pf_uadt,
-       .adt    = {
-               [IPSET_ADD] = type_pf_add,
-               [IPSET_DEL] = type_pf_del,
-               [IPSET_TEST] = type_pf_test,
-       },
-       .destroy = type_pf_destroy,
-       .flush  = type_pf_flush,
-       .head   = type_pf_head,
-       .list   = type_pf_list,
-       .resize = type_pf_resize,
-       .same_set = type_pf_same_set,
-};
-
-/* Flavour with timeout support */
-
-#define ahash_tdata(n, i) \
-       (struct type_pf_elem *)((struct type_pf_telem *)((n)->value) + (i))
-
-static inline u32
-type_pf_data_timeout(const struct type_pf_elem *data)
-{
-       const struct type_pf_telem *tdata =
-               (const struct type_pf_telem *) data;
-
-       return tdata->timeout;
-}
-
-static inline bool
-type_pf_data_expired(const struct type_pf_elem *data)
-{
-       const struct type_pf_telem *tdata =
-               (const struct type_pf_telem *) data;
-
-       return ip_set_timeout_expired(tdata->timeout);
-}
-
-static inline void
-type_pf_data_timeout_set(struct type_pf_elem *data, u32 timeout)
-{
-       struct type_pf_telem *tdata = (struct type_pf_telem *) data;
-
-       tdata->timeout = ip_set_timeout_set(timeout);
-}
-
-static int
-type_pf_elem_tadd(struct hbucket *n, const struct type_pf_elem *value,
-                 u8 ahash_max, u32 cadt_flags, u32 timeout)
-{
-       struct type_pf_elem *data;
-
-       if (n->pos >= n->size) {
-               void *tmp;
-
-               if (n->size >= ahash_max)
-                       /* Trigger rehashing */
-                       return -EAGAIN;
-
-               tmp = kzalloc((n->size + AHASH_INIT_SIZE)
-                             * sizeof(struct type_pf_telem),
-                             GFP_ATOMIC);
-               if (!tmp)
-                       return -ENOMEM;
-               if (n->size) {
-                       memcpy(tmp, n->value,
-                              sizeof(struct type_pf_telem) * n->size);
-                       kfree(n->value);
-               }
-               n->value = tmp;
-               n->size += AHASH_INIT_SIZE;
-       }
-       data = ahash_tdata(n, n->pos++);
-       type_pf_data_copy(data, value);
-       type_pf_data_timeout_set(data, timeout);
-#ifdef IP_SET_HASH_WITH_NETS
-       /* Resizing won't overwrite stored flags */
-       if (cadt_flags)
-               type_pf_data_flags(data, cadt_flags);
-#endif
-       return 0;
-}
-
-/* Delete expired elements from the hashtable */
-static void
-type_pf_expire(struct ip_set_hash *h, u8 nets_length)
-{
-       struct htable *t = h->table;
-       struct hbucket *n;
-       struct type_pf_elem *data;
-       u32 i;
-       int j;
-
-       for (i = 0; i < jhash_size(t->htable_bits); i++) {
-               n = hbucket(t, i);
-               for (j = 0; j < n->pos; j++) {
-                       data = ahash_tdata(n, j);
-                       if (type_pf_data_expired(data)) {
-                               pr_debug("expired %u/%u\n", i, j);
-#ifdef IP_SET_HASH_WITH_NETS
-                               del_cidr(h, CIDR(data->cidr), nets_length);
-#endif
-                               if (j != n->pos - 1)
-                                       /* Not last one */
-                                       type_pf_data_copy(data,
-                                               ahash_tdata(n, n->pos - 1));
-                               n->pos--;
-                               h->elements--;
-                       }
-               }
-               if (n->pos + AHASH_INIT_SIZE < n->size) {
-                       void *tmp = kzalloc((n->size - AHASH_INIT_SIZE)
-                                           * sizeof(struct type_pf_telem),
-                                           GFP_ATOMIC);
-                       if (!tmp)
-                               /* Still try to delete expired elements */
-                               continue;
-                       n->size -= AHASH_INIT_SIZE;
-                       memcpy(tmp, n->value,
-                              n->size * sizeof(struct type_pf_telem));
-                       kfree(n->value);
-                       n->value = tmp;
-               }
-       }
-}
-
-static int
-type_pf_tresize(struct ip_set *set, bool retried)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t, *orig = h->table;
-       u8 htable_bits = orig->htable_bits;
-       struct type_pf_elem *data;
-       struct hbucket *n, *m;
-       u32 i, j, flags = 0;
-       int ret;
-
-       /* Try to cleanup once */
-       if (!retried) {
-               i = h->elements;
-               write_lock_bh(&set->lock);
-               type_pf_expire(set->data, NETS_LENGTH(set->family));
-               write_unlock_bh(&set->lock);
-               if (h->elements <  i)
-                       return 0;
-       }
-
-retry:
-       ret = 0;
-       htable_bits++;
-       pr_debug("attempt to resize set %s from %u to %u, t %p\n",
-                set->name, orig->htable_bits, htable_bits, orig);
-       if (!htable_bits) {
-               /* In case we have plenty of memory :-) */
-               pr_warning("Cannot increase the hashsize of set %s further\n",
-                          set->name);
-               return -IPSET_ERR_HASH_FULL;
-       }
-       t = ip_set_alloc(sizeof(*t)
-                        + jhash_size(htable_bits) * sizeof(struct hbucket));
-       if (!t)
-               return -ENOMEM;
-       t->htable_bits = htable_bits;
-
-       read_lock_bh(&set->lock);
-       for (i = 0; i < jhash_size(orig->htable_bits); i++) {
-               n = hbucket(orig, i);
-               for (j = 0; j < n->pos; j++) {
-                       data = ahash_tdata(n, j);
-#ifdef IP_SET_HASH_WITH_NETS
-                       flags = 0;
-                       type_pf_data_reset_flags(data, &flags);
-#endif
-                       m = hbucket(t, HKEY(data, h->initval, htable_bits));
-                       ret = type_pf_elem_tadd(m, data, AHASH_MAX(h), flags,
-                               ip_set_timeout_get(type_pf_data_timeout(data)));
-                       if (ret < 0) {
-#ifdef IP_SET_HASH_WITH_NETS
-                               type_pf_data_flags(data, flags);
-#endif
-                               read_unlock_bh(&set->lock);
-                               ahash_destroy(t);
-                               if (ret == -EAGAIN)
-                                       goto retry;
-                               return ret;
-                       }
-               }
-       }
-
-       rcu_assign_pointer(h->table, t);
-       read_unlock_bh(&set->lock);
-
-       /* Give time to other readers of the set */
-       synchronize_rcu_bh();
-
-       ahash_destroy(orig);
-
-       return 0;
-}
-
-static int
-type_pf_tadd(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       const struct type_pf_elem *d = value;
-       struct hbucket *n;
-       struct type_pf_elem *data;
-       int ret = 0, i, j = AHASH_MAX(h) + 1;
-       bool flag_exist = flags & IPSET_FLAG_EXIST;
-       u32 key, multi = 0;
-       u32 cadt_flags = flags >> 16;
-
-       if (h->elements >= h->maxelem)
-               /* FIXME: when set is full, we slow down here */
-               type_pf_expire(h, NETS_LENGTH(set->family));
-       if (h->elements >= h->maxelem) {
-               if (net_ratelimit())
-                       pr_warning("Set %s is full, maxelem %u reached\n",
-                                  set->name, h->maxelem);
-               return -IPSET_ERR_HASH_FULL;
-       }
-
-       rcu_read_lock_bh();
-       t = rcu_dereference_bh(h->table);
-       key = HKEY(d, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++) {
-               data = ahash_tdata(n, i);
-               if (type_pf_data_equal(data, d, &multi)) {
-                       if (type_pf_data_expired(data) || flag_exist)
-                               /* Just timeout value may be updated */
-                               j = i;
-                       else {
-                               ret = -IPSET_ERR_EXIST;
-                               goto out;
-                       }
-               } else if (j == AHASH_MAX(h) + 1 &&
-                          type_pf_data_expired(data))
-                       j = i;
-       }
-       if (j != AHASH_MAX(h) + 1) {
-               data = ahash_tdata(n, j);
-#ifdef IP_SET_HASH_WITH_NETS
-               del_cidr(h, CIDR(data->cidr), NETS_LENGTH(set->family));
-               add_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
-#endif
-               type_pf_data_copy(data, d);
-               type_pf_data_timeout_set(data, timeout);
-#ifdef IP_SET_HASH_WITH_NETS
-               type_pf_data_flags(data, cadt_flags);
-#endif
-               goto out;
-       }
-       TUNE_AHASH_MAX(h, multi);
-       ret = type_pf_elem_tadd(n, d, AHASH_MAX(h), cadt_flags, timeout);
-       if (ret != 0) {
-               if (ret == -EAGAIN)
-                       type_pf_data_next(h, d);
-               goto out;
-       }
-
-#ifdef IP_SET_HASH_WITH_NETS
-       add_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
-#endif
-       h->elements++;
-out:
-       rcu_read_unlock_bh();
-       return ret;
-}
-
-static int
-type_pf_tdel(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       const struct type_pf_elem *d = value;
-       struct hbucket *n;
-       int i;
-       struct type_pf_elem *data;
-       u32 key, multi = 0;
-
-       key = HKEY(value, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++) {
-               data = ahash_tdata(n, i);
-               if (!type_pf_data_equal(data, d, &multi))
-                       continue;
-               if (type_pf_data_expired(data))
-                       return -IPSET_ERR_EXIST;
-               if (i != n->pos - 1)
-                       /* Not last one */
-                       type_pf_data_copy(data, ahash_tdata(n, n->pos - 1));
-
-               n->pos--;
-               h->elements--;
-#ifdef IP_SET_HASH_WITH_NETS
-               del_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
-#endif
-               if (n->pos + AHASH_INIT_SIZE < n->size) {
-                       void *tmp = kzalloc((n->size - AHASH_INIT_SIZE)
-                                           * sizeof(struct type_pf_telem),
-                                           GFP_ATOMIC);
-                       if (!tmp)
-                               return 0;
-                       n->size -= AHASH_INIT_SIZE;
-                       memcpy(tmp, n->value,
-                              n->size * sizeof(struct type_pf_telem));
-                       kfree(n->value);
-                       n->value = tmp;
-               }
-               return 0;
-       }
-
-       return -IPSET_ERR_EXIST;
-}
-
-#ifdef IP_SET_HASH_WITH_NETS
-static int
-type_pf_ttest_cidrs(struct ip_set *set, struct type_pf_elem *d, u32 timeout)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       struct type_pf_elem *data;
-       struct hbucket *n;
-       int i, j = 0;
-       u32 key, multi = 0;
-       u8 nets_length = NETS_LENGTH(set->family);
-
-       for (; j < nets_length && h->nets[j].nets && !multi; j++) {
-               type_pf_data_netmask(d, h->nets[j].cidr);
-               key = HKEY(d, h->initval, t->htable_bits);
-               n = hbucket(t, key);
-               for (i = 0; i < n->pos; i++) {
-                       data = ahash_tdata(n, i);
-#ifdef IP_SET_HASH_WITH_MULTI
-                       if (type_pf_data_equal(data, d, &multi)) {
-                               if (!type_pf_data_expired(data))
-                                       return type_pf_data_match(data);
-                               multi = 0;
-                       }
-#else
-                       if (type_pf_data_equal(data, d, &multi) &&
-                           !type_pf_data_expired(data))
-                               return type_pf_data_match(data);
-#endif
-               }
-       }
-       return 0;
-}
-#endif
-
-static int
-type_pf_ttest(struct ip_set *set, void *value, u32 timeout, u32 flags)
-{
-       struct ip_set_hash *h = set->data;
-       struct htable *t = h->table;
-       struct type_pf_elem *data, *d = value;
-       struct hbucket *n;
-       int i;
-       u32 key, multi = 0;
-
-#ifdef IP_SET_HASH_WITH_NETS
-       if (CIDR(d->cidr) == SET_HOST_MASK(set->family))
-               return type_pf_ttest_cidrs(set, d, timeout);
-#endif
-       key = HKEY(d, h->initval, t->htable_bits);
-       n = hbucket(t, key);
-       for (i = 0; i < n->pos; i++) {
-               data = ahash_tdata(n, i);
-               if (type_pf_data_equal(data, d, &multi) &&
-                   !type_pf_data_expired(data))
-                       return type_pf_data_match(data);
-       }
-       return 0;
-}
-
-static int
-type_pf_tlist(const struct ip_set *set,
-             struct sk_buff *skb, struct netlink_callback *cb)
-{
-       const struct ip_set_hash *h = set->data;
-       const struct htable *t = h->table;
-       struct nlattr *atd, *nested;
-       const struct hbucket *n;
-       const struct type_pf_elem *data;
-       u32 first = cb->args[2];
-       /* We assume that one hash bucket fills into one page */
-       void *incomplete;
-       int i;
-
-       atd = ipset_nest_start(skb, IPSET_ATTR_ADT);
-       if (!atd)
-               return -EMSGSIZE;
-       for (; cb->args[2] < jhash_size(t->htable_bits); cb->args[2]++) {
-               incomplete = skb_tail_pointer(skb);
-               n = hbucket(t, cb->args[2]);
-               for (i = 0; i < n->pos; i++) {
-                       data = ahash_tdata(n, i);
-                       pr_debug("list %p %u\n", n, i);
-                       if (type_pf_data_expired(data))
-                               continue;
-                       pr_debug("do list %p %u\n", n, i);
-                       nested = ipset_nest_start(skb, IPSET_ATTR_DATA);
-                       if (!nested) {
-                               if (cb->args[2] == first) {
-                                       nla_nest_cancel(skb, atd);
-                                       return -EMSGSIZE;
-                               } else
-                                       goto nla_put_failure;
-                       }
-                       if (type_pf_data_tlist(skb, data))
-                               goto nla_put_failure;
-                       ipset_nest_end(skb, nested);
-               }
-       }
-       ipset_nest_end(skb, atd);
-       /* Set listing finished */
-       cb->args[2] = 0;
-
-       return 0;
-
-nla_put_failure:
-       nlmsg_trim(skb, incomplete);
-       ipset_nest_end(skb, atd);
-       if (unlikely(first == cb->args[2])) {
-               pr_warning("Can't list set %s: one bucket does not fit into "
-                          "a message. Please report it!\n", set->name);
-               cb->args[2] = 0;
-               return -EMSGSIZE;
-       }
-       return 0;
-}
-
-static const struct ip_set_type_variant type_pf_tvariant = {
-       .kadt   = type_pf_kadt,
-       .uadt   = type_pf_uadt,
-       .adt    = {
-               [IPSET_ADD] = type_pf_tadd,
-               [IPSET_DEL] = type_pf_tdel,
-               [IPSET_TEST] = type_pf_ttest,
-       },
-       .destroy = type_pf_destroy,
-       .flush  = type_pf_flush,
-       .head   = type_pf_head,
-       .list   = type_pf_tlist,
-       .resize = type_pf_tresize,
-       .same_set = type_pf_same_set,
-};
-
-static void
-type_pf_gc(unsigned long ul_set)
-{
-       struct ip_set *set = (struct ip_set *) ul_set;
-       struct ip_set_hash *h = set->data;
-
-       pr_debug("called\n");
-       write_lock_bh(&set->lock);
-       type_pf_expire(h, NETS_LENGTH(set->family));
-       write_unlock_bh(&set->lock);
-
-       h->gc.expires = jiffies + IPSET_GC_PERIOD(h->timeout) * HZ;
-       add_timer(&h->gc);
-}
-
-static void
-type_pf_gc_init(struct ip_set *set)
-{
-       struct ip_set_hash *h = set->data;
-
-       init_timer(&h->gc);
-       h->gc.data = (unsigned long) set;
-       h->gc.function = type_pf_gc;
-       h->gc.expires = jiffies + IPSET_GC_PERIOD(h->timeout) * HZ;
-       add_timer(&h->gc);
-       pr_debug("gc initialized, run in every %u\n",
-                IPSET_GC_PERIOD(h->timeout));
-}
-
-#undef HKEY_DATALEN
-#undef HKEY
-#undef type_pf_data_equal
-#undef type_pf_data_isnull
-#undef type_pf_data_copy
-#undef type_pf_data_zero_out
-#undef type_pf_data_netmask
-#undef type_pf_data_list
-#undef type_pf_data_tlist
-#undef type_pf_data_next
-#undef type_pf_data_flags
-#undef type_pf_data_reset_flags
-#undef type_pf_data_match
-
-#undef type_pf_elem
-#undef type_pf_telem
-#undef type_pf_data_timeout
-#undef type_pf_data_expired
-#undef type_pf_data_timeout_set
-
-#undef type_pf_elem_add
-#undef type_pf_add
-#undef type_pf_del
-#undef type_pf_test_cidrs
-#undef type_pf_test
-
-#undef type_pf_elem_tadd
-#undef type_pf_del_telem
-#undef type_pf_expire
-#undef type_pf_tadd
-#undef type_pf_tdel
-#undef type_pf_ttest_cidrs
-#undef type_pf_ttest
-
-#undef type_pf_resize
-#undef type_pf_tresize
-#undef type_pf_flush
-#undef type_pf_destroy
-#undef type_pf_head
-#undef type_pf_list
-#undef type_pf_tlist
-#undef type_pf_same_set
-#undef type_pf_kadt
-#undef type_pf_uadt
-#undef type_pf_gc
-#undef type_pf_gc_init
-#undef type_pf_variant
-#undef type_pf_tvariant
diff --git a/net/netfilter/ipset/ip_set_hash_gen.h b/net/netfilter/ipset/ip_set_hash_gen.h
new file mode 100644 (file)
index 0000000..2ba7d4e
--- /dev/null
@@ -0,0 +1,1039 @@
+/* Copyright (C) 2013 Jozsef Kadlecsik <kadlec@blackhole.kfki.hu>
+ *
+ * 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.
+ */
+
+#ifndef _IP_SET_HASH_GEN_H
+#define _IP_SET_HASH_GEN_H
+
+#include <linux/rcupdate.h>
+#include <linux/jhash.h>
+#include <linux/netfilter/ipset/ip_set_timeout.h>
+#ifndef rcu_dereference_bh
+#define rcu_dereference_bh(p)  rcu_dereference(p)
+#endif
+
+#define CONCAT(a, b)           a##b
+#define TOKEN(a, b)            CONCAT(a, b)
+
+/* Hashing which uses arrays to resolve clashing. The hash table is resized
+ * (doubled) when searching becomes too long.
+ * Internally jhash is used with the assumption that the size of the
+ * stored data is a multiple of sizeof(u32). If storage supports timeout,
+ * the timeout field must be the last one in the data structure - that field
+ * is ignored when computing the hash key.
+ *
+ * Readers and resizing
+ *
+ * Resizing can be triggered by userspace command only, and those
+ * are serialized by the nfnl mutex. During resizing the set is
+ * read-locked, so the only possible concurrent operations are
+ * the kernel side readers. Those must be protected by proper RCU locking.
+ */
+
+/* Number of elements to store in an initial array block */
+#define AHASH_INIT_SIZE                        4
+/* Max number of elements to store in an array block */
+#define AHASH_MAX_SIZE                 (3*AHASH_INIT_SIZE)
+
+/* Max number of elements can be tuned */
+#ifdef IP_SET_HASH_WITH_MULTI
+#define AHASH_MAX(h)                   ((h)->ahash_max)
+
+static inline u8
+tune_ahash_max(u8 curr, u32 multi)
+{
+       u32 n;
+
+       if (multi < curr)
+               return curr;
+
+       n = curr + AHASH_INIT_SIZE;
+       /* Currently, at listing one hash bucket must fit into a message.
+        * Therefore we have a hard limit here.
+        */
+       return n > curr && n <= 64 ? n : curr;
+}
+#define TUNE_AHASH_MAX(h, multi)       \
+       ((h)->ahash_max = tune_ahash_max((h)->ahash_max, multi))
+#else
+#define AHASH_MAX(h)                   AHASH_MAX_SIZE
+#define TUNE_AHASH_MAX(h, multi)
+#endif
+
+/* A hash bucket */
+struct hbucket {
+       void *value;            /* the array of the values */
+       u8 size;                /* size of the array */
+       u8 pos;                 /* position of the first free entry */
+};
+
+/* The hash table: the table size stored here in order to make resizing easy */
+struct htable {
+       u8 htable_bits;         /* size of hash table == 2^htable_bits */
+       struct hbucket bucket[0]; /* hashtable buckets */
+};
+
+#define hbucket(h, i)          (&((h)->bucket[i]))
+
+/* Book-keeping of the prefixes added to the set */
+struct net_prefixes {
+       u8 cidr;                /* the different cidr values in the set */
+       u32 nets;               /* number of elements per cidr */
+};
+
+/* Compute the hash table size */
+static size_t
+htable_size(u8 hbits)
+{
+       size_t hsize;
+
+       /* We must fit both into u32 in jhash and size_t */
+       if (hbits > 31)
+               return 0;
+       hsize = jhash_size(hbits);
+       if ((((size_t)-1) - sizeof(struct htable))/sizeof(struct hbucket)
+           < hsize)
+               return 0;
+
+       return hsize * sizeof(struct hbucket) + sizeof(struct htable);
+}
+
+/* Compute htable_bits from the user input parameter hashsize */
+static u8
+htable_bits(u32 hashsize)
+{
+       /* Assume that hashsize == 2^htable_bits */
+       u8 bits = fls(hashsize - 1);
+       if (jhash_size(bits) != hashsize)
+               /* Round up to the first 2^n value */
+               bits = fls(hashsize);
+
+       return bits;
+}
+
+/* Destroy the hashtable part of the set */
+static void
+ahash_destroy(struct htable *t)
+{
+       struct hbucket *n;
+       u32 i;
+
+       for (i = 0; i < jhash_size(t->htable_bits); i++) {
+               n = hbucket(t, i);
+               if (n->size)
+                       /* FIXME: use slab cache */
+                       kfree(n->value);
+       }
+
+       ip_set_free(t);
+}
+
+static int
+hbucket_elem_add(struct hbucket *n, u8 ahash_max, size_t dsize)
+{
+       if (n->pos >= n->size) {
+               void *tmp;
+
+               if (n->size >= ahash_max)
+                       /* Trigger rehashing */
+                       return -EAGAIN;
+
+               tmp = kzalloc((n->size + AHASH_INIT_SIZE) * dsize,
+                             GFP_ATOMIC);
+               if (!tmp)
+                       return -ENOMEM;
+               if (n->size) {
+                       memcpy(tmp, n->value, n->size * dsize);
+                       kfree(n->value);
+               }
+               n->value = tmp;
+               n->size += AHASH_INIT_SIZE;
+       }
+       return 0;
+}
+
+#ifdef IP_SET_HASH_WITH_NETS
+#ifdef IP_SET_HASH_WITH_NETS_PACKED
+/* When cidr is packed with nomatch, cidr - 1 is stored in the entry */
+#define CIDR(cidr)             (cidr + 1)
+#else
+#define CIDR(cidr)             (cidr)
+#endif
+
+#define SET_HOST_MASK(family)  (family == AF_INET ? 32 : 128)
+
+#ifdef IP_SET_HASH_WITH_MULTI
+#define NETS_LENGTH(family)    (SET_HOST_MASK(family) + 1)
+#else
+#define NETS_LENGTH(family)    SET_HOST_MASK(family)
+#endif
+
+#else
+#define NETS_LENGTH(family)    0
+#endif /* IP_SET_HASH_WITH_NETS */
+
+#define ext_timeout(e, h)      \
+(unsigned long *)(((void *)(e)) + (h)->offset[IPSET_OFFSET_TIMEOUT])
+
+#endif /* _IP_SET_HASH_GEN_H */
+
+/* Family dependent templates */
+
+#undef ahash_data
+#undef mtype_data_equal
+#undef mtype_do_data_match
+#undef mtype_data_set_flags
+#undef mtype_data_reset_flags
+#undef mtype_data_netmask
+#undef mtype_data_list
+#undef mtype_data_next
+#undef mtype_elem
+
+#undef mtype_add_cidr
+#undef mtype_del_cidr
+#undef mtype_ahash_memsize
+#undef mtype_flush
+#undef mtype_destroy
+#undef mtype_gc_init
+#undef mtype_same_set
+#undef mtype_kadt
+#undef mtype_uadt
+#undef mtype
+
+#undef mtype_add
+#undef mtype_del
+#undef mtype_test_cidrs
+#undef mtype_test
+#undef mtype_expire
+#undef mtype_resize
+#undef mtype_head
+#undef mtype_list
+#undef mtype_gc
+#undef mtype_gc_init
+#undef mtype_variant
+#undef mtype_data_match
+
+#undef HKEY
+
+#define mtype_data_equal       TOKEN(MTYPE, _data_equal)
+#ifdef IP_SET_HASH_WITH_NETS
+#define mtype_do_data_match    TOKEN(MTYPE, _do_data_match)
+#else
+#define mtype_do_data_match(d) 1
+#endif
+#define mtype_data_set_flags   TOKEN(MTYPE, _data_set_flags)
+#define mtype_data_reset_flags TOKEN(MTYPE, _data_reset_flags)
+#define mtype_data_netmask     TOKEN(MTYPE, _data_netmask)
+#define mtype_data_list                TOKEN(MTYPE, _data_list)
+#define mtype_data_next                TOKEN(MTYPE, _data_next)
+#define mtype_elem             TOKEN(MTYPE, _elem)
+#define mtype_add_cidr         TOKEN(MTYPE, _add_cidr)
+#define mtype_del_cidr         TOKEN(MTYPE, _del_cidr)
+#define mtype_ahash_memsize    TOKEN(MTYPE, _ahash_memsize)
+#define mtype_flush            TOKEN(MTYPE, _flush)
+#define mtype_destroy          TOKEN(MTYPE, _destroy)
+#define mtype_gc_init          TOKEN(MTYPE, _gc_init)
+#define mtype_same_set         TOKEN(MTYPE, _same_set)
+#define mtype_kadt             TOKEN(MTYPE, _kadt)
+#define mtype_uadt             TOKEN(MTYPE, _uadt)
+#define mtype                  MTYPE
+
+#define mtype_elem             TOKEN(MTYPE, _elem)
+#define mtype_add              TOKEN(MTYPE, _add)
+#define mtype_del              TOKEN(MTYPE, _del)
+#define mtype_test_cidrs       TOKEN(MTYPE, _test_cidrs)
+#define mtype_test             TOKEN(MTYPE, _test)
+#define mtype_expire           TOKEN(MTYPE, _expire)
+#define mtype_resize           TOKEN(MTYPE, _resize)
+#define mtype_head             TOKEN(MTYPE, _head)
+#define mtype_list             TOKEN(MTYPE, _list)
+#define mtype_gc               TOKEN(MTYPE, _gc)
+#define mtype_variant          TOKEN(MTYPE, _variant)
+#define mtype_data_match       TOKEN(MTYPE, _data_match)
+
+#ifndef HKEY_DATALEN
+#define HKEY_DATALEN           sizeof(struct mtype_elem)
+#endif
+
+#define HKEY(data, initval, htable_bits)                       \
+(jhash2((u32 *)(data), HKEY_DATALEN/sizeof(u32), initval)      \
+       & jhash_mask(htable_bits))
+
+#ifndef htype
+#define htype                  HTYPE
+
+/* The generic hash structure */
+struct htype {
+       struct htable *table;   /* the hash table */
+       u32 maxelem;            /* max elements in the hash */
+       u32 elements;           /* current element (vs timeout) */
+       u32 initval;            /* random jhash init value */
+       u32 timeout;            /* timeout value, if enabled */
+       size_t dsize;           /* data struct size */
+       size_t offset[IPSET_OFFSET_MAX]; /* Offsets to extensions */
+       struct timer_list gc;   /* garbage collection when timeout enabled */
+       struct mtype_elem next; /* temporary storage for uadd */
+#ifdef IP_SET_HASH_WITH_MULTI
+       u8 ahash_max;           /* max elements in an array block */
+#endif
+#ifdef IP_SET_HASH_WITH_NETMASK
+       u8 netmask;             /* netmask value for subnets to store */
+#endif
+#ifdef IP_SET_HASH_WITH_RBTREE
+       struct rb_root rbtree;
+#endif
+#ifdef IP_SET_HASH_WITH_NETS
+       struct net_prefixes nets[0]; /* book-keeping of prefixes */
+#endif
+};
+#endif
+
+#ifdef IP_SET_HASH_WITH_NETS
+/* Network cidr size book keeping when the hash stores different
+ * sized networks */
+static void
+mtype_add_cidr(struct htype *h, u8 cidr, u8 nets_length)
+{
+       int i, j;
+
+       /* Add in increasing prefix order, so larger cidr first */
+       for (i = 0, j = -1; i < nets_length && h->nets[i].nets; i++) {
+               if (j != -1)
+                       continue;
+               else if (h->nets[i].cidr < cidr)
+                       j = i;
+               else if (h->nets[i].cidr == cidr) {
+                       h->nets[i].nets++;
+                       return;
+               }
+       }
+       if (j != -1) {
+               for (; i > j; i--) {
+                       h->nets[i].cidr = h->nets[i - 1].cidr;
+                       h->nets[i].nets = h->nets[i - 1].nets;
+               }
+       }
+       h->nets[i].cidr = cidr;
+       h->nets[i].nets = 1;
+}
+
+static void
+mtype_del_cidr(struct htype *h, u8 cidr, u8 nets_length)
+{
+       u8 i, j;
+
+       for (i = 0; i < nets_length - 1 && h->nets[i].cidr != cidr; i++)
+               ;
+       h->nets[i].nets--;
+
+       if (h->nets[i].nets != 0)
+               return;
+
+       for (j = i; j < nets_length - 1 && h->nets[j].nets; j++) {
+               h->nets[j].cidr = h->nets[j + 1].cidr;
+               h->nets[j].nets = h->nets[j + 1].nets;
+       }
+}
+#endif
+
+/* Calculate the actual memory size of the set data */
+static size_t
+mtype_ahash_memsize(const struct htype *h, u8 nets_length)
+{
+       u32 i;
+       struct htable *t = h->table;
+       size_t memsize = sizeof(*h)
+                        + sizeof(*t)
+#ifdef IP_SET_HASH_WITH_NETS
+                        + sizeof(struct net_prefixes) * nets_length
+#endif
+                        + jhash_size(t->htable_bits) * sizeof(struct hbucket);
+
+       for (i = 0; i < jhash_size(t->htable_bits); i++)
+               memsize += t->bucket[i].size * h->dsize;
+
+       return memsize;
+}
+
+/* Flush a hash type of set: destroy all elements */
+static void
+mtype_flush(struct ip_set *set)
+{
+       struct htype *h = set->data;
+       struct htable *t = h->table;
+       struct hbucket *n;
+       u32 i;
+
+       for (i = 0; i < jhash_size(t->htable_bits); i++) {
+               n = hbucket(t, i);
+               if (n->size) {
+                       n->size = n->pos = 0;
+                       /* FIXME: use slab cache */
+                       kfree(n->value);
+               }
+       }
+#ifdef IP_SET_HASH_WITH_NETS
+       memset(h->nets, 0, sizeof(struct net_prefixes)
+                          * NETS_LENGTH(set->family));
+#endif
+       h->elements = 0;
+}
+
+/* Destroy a hash type of set */
+static void
+mtype_destroy(struct ip_set *set)
+{
+       struct htype *h = set->data;
+
+       if (set->extensions & IPSET_EXT_TIMEOUT)
+               del_timer_sync(&h->gc);
+
+       ahash_destroy(h->table);
+#ifdef IP_SET_HASH_WITH_RBTREE
+       rbtree_destroy(&h->rbtree);
+#endif
+       kfree(h);
+
+       set->data = NULL;
+}
+
+static void
+mtype_gc_init(struct ip_set *set, void (*gc)(unsigned long ul_set))
+{
+       struct htype *h = set->data;
+
+       init_timer(&h->gc);
+       h->gc.data = (unsigned long) set;
+       h->gc.function = gc;
+       h->gc.expires = jiffies + IPSET_GC_PERIOD(h->timeout) * HZ;
+       add_timer(&h->gc);
+       pr_debug("gc initialized, run in every %u\n",
+                IPSET_GC_PERIOD(h->timeout));
+}
+
+static bool
+mtype_same_set(const struct ip_set *a, const struct ip_set *b)
+{
+       const struct htype *x = a->data;
+       const struct htype *y = b->data;
+
+       /* Resizing changes htable_bits, so we ignore it */
+       return x->maxelem == y->maxelem &&
+              x->timeout == y->timeout &&
+#ifdef IP_SET_HASH_WITH_NETMASK
+              x->netmask == y->netmask &&
+#endif
+              a->extensions == b->extensions;
+}
+
+/* Get the ith element from the array block n */
+#define ahash_data(n, i, dsize)        \
+       ((struct mtype_elem *)((n)->value + ((i) * (dsize))))
+
+/* Delete expired elements from the hashtable */
+static void
+mtype_expire(struct htype *h, u8 nets_length, size_t dsize)
+{
+       struct htable *t = h->table;
+       struct hbucket *n;
+       struct mtype_elem *data;
+       u32 i;
+       int j;
+
+       for (i = 0; i < jhash_size(t->htable_bits); i++) {
+               n = hbucket(t, i);
+               for (j = 0; j < n->pos; j++) {
+                       data = ahash_data(n, j, dsize);
+                       if (ip_set_timeout_expired(ext_timeout(data, h))) {
+                               pr_debug("expired %u/%u\n", i, j);
+#ifdef IP_SET_HASH_WITH_NETS
+                               mtype_del_cidr(h, CIDR(data->cidr),
+                                              nets_length);
+#endif
+                               if (j != n->pos - 1)
+                                       /* Not last one */
+                                       memcpy(data,
+                                              ahash_data(n, n->pos - 1, dsize),
+                                              dsize);
+                               n->pos--;
+                               h->elements--;
+                       }
+               }
+               if (n->pos + AHASH_INIT_SIZE < n->size) {
+                       void *tmp = kzalloc((n->size - AHASH_INIT_SIZE)
+                                           * dsize,
+                                           GFP_ATOMIC);
+                       if (!tmp)
+                               /* Still try to delete expired elements */
+                               continue;
+                       n->size -= AHASH_INIT_SIZE;
+                       memcpy(tmp, n->value, n->size * dsize);
+                       kfree(n->value);
+                       n->value = tmp;
+               }
+       }
+}
+
+static void
+mtype_gc(unsigned long ul_set)
+{
+       struct ip_set *set = (struct ip_set *) ul_set;
+       struct htype *h = set->data;
+
+       pr_debug("called\n");
+       write_lock_bh(&set->lock);
+       mtype_expire(h, NETS_LENGTH(set->family), h->dsize);
+       write_unlock_bh(&set->lock);
+
+       h->gc.expires = jiffies + IPSET_GC_PERIOD(h->timeout) * HZ;
+       add_timer(&h->gc);
+}
+
+/* Resize a hash: create a new hash table with doubling the hashsize
+ * and inserting the elements to it. Repeat until we succeed or
+ * fail due to memory pressures. */
+static int
+mtype_resize(struct ip_set *set, bool retried)
+{
+       struct htype *h = set->data;
+       struct htable *t, *orig = h->table;
+       u8 htable_bits = orig->htable_bits;
+#ifdef IP_SET_HASH_WITH_NETS
+       u8 flags;
+#endif
+       struct mtype_elem *data;
+       struct mtype_elem *d;
+       struct hbucket *n, *m;
+       u32 i, j;
+       int ret;
+
+       /* Try to cleanup once */
+       if (SET_WITH_TIMEOUT(set) && !retried) {
+               i = h->elements;
+               write_lock_bh(&set->lock);
+               mtype_expire(set->data, NETS_LENGTH(set->family),
+                            h->dsize);
+               write_unlock_bh(&set->lock);
+               if (h->elements < i)
+                       return 0;
+       }
+
+retry:
+       ret = 0;
+       htable_bits++;
+       pr_debug("attempt to resize set %s from %u to %u, t %p\n",
+                set->name, orig->htable_bits, htable_bits, orig);
+       if (!htable_bits) {
+               /* In case we have plenty of memory :-) */
+               pr_warning("Cannot increase the hashsize of set %s further\n",
+                          set->name);
+               return -IPSET_ERR_HASH_FULL;
+       }
+       t = ip_set_alloc(sizeof(*t)
+                        + jhash_size(htable_bits) * sizeof(struct hbucket));
+       if (!t)
+               return -ENOMEM;
+       t->htable_bits = htable_bits;
+
+       read_lock_bh(&set->lock);
+       for (i = 0; i < jhash_size(orig->htable_bits); i++) {
+               n = hbucket(orig, i);
+               for (j = 0; j < n->pos; j++) {
+                       data = ahash_data(n, j, h->dsize);
+#ifdef IP_SET_HASH_WITH_NETS
+                       flags = 0;
+                       mtype_data_reset_flags(data, &flags);
+#endif
+                       m = hbucket(t, HKEY(data, h->initval, htable_bits));
+                       ret = hbucket_elem_add(m, AHASH_MAX(h), h->dsize);
+                       if (ret < 0) {
+#ifdef IP_SET_HASH_WITH_NETS
+                               mtype_data_reset_flags(data, &flags);
+#endif
+                               read_unlock_bh(&set->lock);
+                               ahash_destroy(t);
+                               if (ret == -EAGAIN)
+                                       goto retry;
+                               return ret;
+                       }
+                       d = ahash_data(m, m->pos++, h->dsize);
+                       memcpy(d, data, h->dsize);
+#ifdef IP_SET_HASH_WITH_NETS
+                       mtype_data_reset_flags(d, &flags);
+#endif
+               }
+       }
+
+       rcu_assign_pointer(h->table, t);
+       read_unlock_bh(&set->lock);
+
+       /* Give time to other readers of the set */
+       synchronize_rcu_bh();
+
+       pr_debug("set %s resized from %u (%p) to %u (%p)\n", set->name,
+                orig->htable_bits, orig, t->htable_bits, t);
+       ahash_destroy(orig);
+
+       return 0;
+}
+
+/* Add an element to a hash and update the internal counters when succeeded,
+ * otherwise report the proper error code. */
+static int
+mtype_add(struct ip_set *set, void *value, const struct ip_set_ext *ext,
+         struct ip_set_ext *mext, u32 flags)
+{
+       struct htype *h = set->data;
+       struct htable *t;
+       const struct mtype_elem *d = value;
+       struct mtype_elem *data;
+       struct hbucket *n;
+       int i, ret = 0;
+       int j = AHASH_MAX(h) + 1;
+       bool flag_exist = flags & IPSET_FLAG_EXIST;
+       u32 key, multi = 0;
+
+       if (SET_WITH_TIMEOUT(set) && h->elements >= h->maxelem)
+               /* FIXME: when set is full, we slow down here */
+               mtype_expire(h, NETS_LENGTH(set->family), h->dsize);
+
+       if (h->elements >= h->maxelem) {
+               if (net_ratelimit())
+                       pr_warning("Set %s is full, maxelem %u reached\n",
+                                  set->name, h->maxelem);
+               return -IPSET_ERR_HASH_FULL;
+       }
+
+       rcu_read_lock_bh();
+       t = rcu_dereference_bh(h->table);
+       key = HKEY(value, h->initval, t->htable_bits);
+       n = hbucket(t, key);
+       for (i = 0; i < n->pos; i++) {
+               data = ahash_data(n, i, h->dsize);
+               if (mtype_data_equal(data, d, &multi)) {
+                       if (flag_exist ||
+                           (SET_WITH_TIMEOUT(set) &&
+                            ip_set_timeout_expired(ext_timeout(data, h)))) {
+                               /* Just the extensions could be overwritten */
+                               j = i;
+                               goto reuse_slot;
+                       } else {
+                               ret = -IPSET_ERR_EXIST;
+                               goto out;
+                       }
+               }
+               /* Reuse first timed out entry */
+               if (SET_WITH_TIMEOUT(set) &&
+                   ip_set_timeout_expired(ext_timeout(data, h)) &&
+                   j != AHASH_MAX(h) + 1)
+                       j = i;
+       }
+reuse_slot:
+       if (j != AHASH_MAX(h) + 1) {
+               /* Fill out reused slot */
+               data = ahash_data(n, j, h->dsize);
+#ifdef IP_SET_HASH_WITH_NETS
+               mtype_del_cidr(h, CIDR(data->cidr), NETS_LENGTH(set->family));
+               mtype_add_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
+#endif
+       } else {
+               /* Use/create a new slot */
+               TUNE_AHASH_MAX(h, multi);
+               ret = hbucket_elem_add(n, AHASH_MAX(h), h->dsize);
+               if (ret != 0) {
+                       if (ret == -EAGAIN)
+                               mtype_data_next(&h->next, d);
+                       goto out;
+               }
+               data = ahash_data(n, n->pos++, h->dsize);
+#ifdef IP_SET_HASH_WITH_NETS
+               mtype_add_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
+#endif
+               h->elements++;
+       }
+       memcpy(data, d, sizeof(struct mtype_elem));
+#ifdef IP_SET_HASH_WITH_NETS
+       mtype_data_set_flags(data, flags);
+#endif
+       if (SET_WITH_TIMEOUT(set))
+               ip_set_timeout_set(ext_timeout(data, h), ext->timeout);
+
+out:
+       rcu_read_unlock_bh();
+       return ret;
+}
+
+/* Delete an element from the hash: swap it with the last element
+ * and free up space if possible.
+ */
+static int
+mtype_del(struct ip_set *set, void *value, const struct ip_set_ext *ext,
+         struct ip_set_ext *mext, u32 flags)
+{
+       struct htype *h = set->data;
+       struct htable *t = h->table;
+       const struct mtype_elem *d = value;
+       struct mtype_elem *data;
+       struct hbucket *n;
+       int i;
+       u32 key, multi = 0;
+
+       key = HKEY(value, h->initval, t->htable_bits);
+       n = hbucket(t, key);
+       for (i = 0; i < n->pos; i++) {
+               data = ahash_data(n, i, h->dsize);
+               if (!mtype_data_equal(data, d, &multi))
+                       continue;
+               if (SET_WITH_TIMEOUT(set) &&
+                   ip_set_timeout_expired(ext_timeout(data, h)))
+                       return -IPSET_ERR_EXIST;
+               if (i != n->pos - 1)
+                       /* Not last one */
+                       memcpy(data, ahash_data(n, n->pos - 1, h->dsize),
+                              h->dsize);
+
+               n->pos--;
+               h->elements--;
+#ifdef IP_SET_HASH_WITH_NETS
+               mtype_del_cidr(h, CIDR(d->cidr), NETS_LENGTH(set->family));
+#endif
+               if (n->pos + AHASH_INIT_SIZE < n->size) {
+                       void *tmp = kzalloc((n->size - AHASH_INIT_SIZE)
+                                           * h->dsize,
+                                           GFP_ATOMIC);
+                       if (!tmp)
+                               return 0;
+                       n->size -= AHASH_INIT_SIZE;
+                       memcpy(tmp, n->value, n->size * h->dsize);
+                       kfree(n->value);
+                       n->value = tmp;
+               }
+               return 0;
+       }
+
+       return -IPSET_ERR_EXIST;
+}
+
+static inline int
+mtype_data_match(struct mtype_elem *data, const struct ip_set_ext *ext,
+                struct ip_set_ext *mext, struct ip_set *set, u32 flags)
+{
+       return mtype_do_data_match(data);
+}
+
+#ifdef IP_SET_HASH_WITH_NETS
+/* Special test function which takes into account the different network
+ * sizes added to the set */
+static int
+mtype_test_cidrs(struct ip_set *set, struct mtype_elem *d,
+                const struct ip_set_ext *ext,
+                struct ip_set_ext *mext, u32 flags)
+{
+       struct htype *h = set->data;
+       struct htable *t = h->table;
+       struct hbucket *n;
+       struct mtype_elem *data;
+       int i, j = 0;
+       u32 key, multi = 0;
+       u8 nets_length = NETS_LENGTH(set->family);
+
+       pr_debug("test by nets\n");
+       for (; j < nets_length && h->nets[j].nets && !multi; j++) {
+               mtype_data_netmask(d, h->nets[j].cidr);
+               key = HKEY(d, h->initval, t->htable_bits);
+               n = hbucket(t, key);
+               for (i = 0; i < n->pos; i++) {
+                       data = ahash_data(n, i, h->dsize);
+                       if (!mtype_data_equal(data, d, &multi))
+                               continue;
+                       if (SET_WITH_TIMEOUT(set)) {
+                               if (!ip_set_timeout_expired(
+                                                       ext_timeout(data, h)))
+                                       return mtype_data_match(data, ext,
+                                                               mext, set,
+                                                               flags);
+#ifdef IP_SET_HASH_WITH_MULTI
+                               multi = 0;
+#endif
+                       } else
+                               return mtype_data_match(data, ext,
+                                                       mext, set, flags);
+               }
+       }
+       return 0;
+}
+#endif
+
+/* Test whether the element is added to the set */
+static int
+mtype_test(struct ip_set *set, void *value, const struct ip_set_ext *ext,
+          struct ip_set_ext *mext, u32 flags)
+{
+       struct htype *h = set->data;
+       struct htable *t = h->table;
+       struct mtype_elem *d = value;
+       struct hbucket *n;
+       struct mtype_elem *data;
+       int i;
+       u32 key, multi = 0;
+
+#ifdef IP_SET_HASH_WITH_NETS
+       /* If we test an IP address and not a network address,
+        * try all possible network sizes */
+       if (CIDR(d->cidr) == SET_HOST_MASK(set->family))
+               return mtype_test_cidrs(set, d, ext, mext, flags);
+#endif
+
+       key = HKEY(d, h->initval, t->htable_bits);
+       n = hbucket(t, key);
+       for (i = 0; i < n->pos; i++) {
+               data = ahash_data(n, i, h->dsize);
+               if (mtype_data_equal(data, d, &multi) &&
+                   !(SET_WITH_TIMEOUT(set) &&
+                     ip_set_timeout_expired(ext_timeout(data, h))))
+                       return mtype_data_match(data, ext, mext, set, flags);
+       }
+       return 0;
+}
+
+/* Reply a HEADER request: fill out the header part of the set */
+static int
+mtype_head(struct ip_set *set, struct sk_buff *skb)
+{
+       const struct htype *h = set->data;
+       struct nlattr *nested;
+       size_t memsize;
+
+       read_lock_bh(&set->lock);
+       memsize = mtype_ahash_memsize(h, NETS_LENGTH(set->family));
+       read_unlock_bh(&set->lock);
+
+       nested = ipset_nest_start(skb, IPSET_ATTR_DATA);
+       if (!nested)
+               goto nla_put_failure;
+       if (nla_put_net32(skb, IPSET_ATTR_HASHSIZE,
+                         htonl(jhash_size(h->table->htable_bits))) ||
+           nla_put_net32(skb, IPSET_ATTR_MAXELEM, htonl(h->maxelem)))
+               goto nla_put_failure;
+#ifdef IP_SET_HASH_WITH_NETMASK
+       if (h->netmask != HOST_MASK &&
+           nla_put_u8(skb, IPSET_ATTR_NETMASK, h->netmask))
+               goto nla_put_failure;
+#endif
+       if (nla_put_net32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)) ||
+           nla_put_net32(skb, IPSET_ATTR_MEMSIZE, htonl(memsize)) ||
+           ((set->extensions & IPSET_EXT_TIMEOUT) &&
+            nla_put_net32(skb, IPSET_ATTR_TIMEOUT, htonl(h->timeout))))
+               goto nla_put_failure;
+       ipset_nest_end(skb, nested);
+
+       return 0;
+nla_put_failure:
+       return -EMSGSIZE;
+}
+
+/* Reply a LIST/SAVE request: dump the elements of the specified set */
+static int
+mtype_list(const struct ip_set *set,
+          struct sk_buff *skb, struct netlink_callback *cb)
+{
+       const struct htype *h = set->data;
+       const struct htable *t = h->table;
+       struct nlattr *atd, *nested;
+       const struct hbucket *n;
+       const struct mtype_elem *e;
+       u32 first = cb->args[2];
+       /* We assume that one hash bucket fills into one page */
+       void *incomplete;
+       int i;
+
+       atd = ipset_nest_start(skb, IPSET_ATTR_ADT);
+       if (!atd)
+               return -EMSGSIZE;
+       pr_debug("list hash set %s\n", set->name);
+       for (; cb->args[2] < jhash_size(t->htable_bits); cb->args[2]++) {
+               incomplete = skb_tail_pointer(skb);
+               n = hbucket(t, cb->args[2]);
+               pr_debug("cb->args[2]: %lu, t %p n %p\n", cb->args[2], t, n);
+               for (i = 0; i < n->pos; i++) {
+                       e = ahash_data(n, i, h->dsize);
+                       if (SET_WITH_TIMEOUT(set) &&
+                           ip_set_timeout_expired(ext_timeout(e, h)))
+                               continue;
+                       pr_debug("list hash %lu hbucket %p i %u, data %p\n",
+                                cb->args[2], n, i, e);
+                       nested = ipset_nest_start(skb, IPSET_ATTR_DATA);
+                       if (!nested) {
+                               if (cb->args[2] == first) {
+                                       nla_nest_cancel(skb, atd);
+                                       return -EMSGSIZE;
+                               } else
+                                       goto nla_put_failure;
+                       }
+                       if (mtype_data_list(skb, e))
+                               goto nla_put_failure;
+                       if (SET_WITH_TIMEOUT(set) &&
+                           nla_put_net32(skb, IPSET_ATTR_TIMEOUT,
+                                         htonl(ip_set_timeout_get(
+                                               ext_timeout(e, h)))))
+                               goto nla_put_failure;
+                       ipset_nest_end(skb, nested);
+               }
+       }
+       ipset_nest_end(skb, atd);
+       /* Set listing finished */
+       cb->args[2] = 0;
+
+       return 0;
+
+nla_put_failure:
+       nlmsg_trim(skb, incomplete);
+       ipset_nest_end(skb, atd);
+       if (unlikely(first == cb->args[2])) {
+               pr_warning("Can't list set %s: one bucket does not fit into "
+                          "a message. Please report it!\n", set->name);
+               cb->args[2] = 0;
+               return -EMSGSIZE;
+       }
+       return 0;
+}
+
+static int
+TOKEN(MTYPE, _kadt)(struct ip_set *set, const struct sk_buff *skb,
+             const struct xt_action_param *par,
+             enum ipset_adt adt, struct ip_set_adt_opt *opt);
+
+static int
+TOKEN(MTYPE, _uadt)(struct ip_set *set, struct nlattr *tb[],
+             enum ipset_adt adt, u32 *lineno, u32 flags, bool retried);
+
+static const struct ip_set_type_variant mtype_variant = {
+       .kadt   = mtype_kadt,
+       .uadt   = mtype_uadt,
+       .adt    = {
+               [IPSET_ADD] = mtype_add,
+               [IPSET_DEL] = mtype_del,
+               [IPSET_TEST] = mtype_test,
+       },
+       .destroy = mtype_destroy,
+       .flush  = mtype_flush,
+       .head   = mtype_head,
+       .list   = mtype_list,
+       .resize = mtype_resize,
+       .same_set = mtype_same_set,
+};
+
+#ifdef IP_SET_EMIT_CREATE
+static int
+TOKEN(HTYPE, _create)(struct ip_set *set, struct nlattr *tb[], u32 flags)
+{
+       u32 hashsize = IPSET_DEFAULT_HASHSIZE, maxelem = IPSET_DEFAULT_MAXELEM;
+       u8 hbits;
+#ifdef IP_SET_HASH_WITH_NETMASK
+       u8 netmask;
+#endif
+       size_t hsize;
+       struct HTYPE *h;
+
+       if (!(set->family == NFPROTO_IPV4 || set->family == NFPROTO_IPV6))
+               return -IPSET_ERR_INVALID_FAMILY;
+#ifdef IP_SET_HASH_WITH_NETMASK
+       netmask = set->family == NFPROTO_IPV4 ? 32 : 128;
+       pr_debug("Create set %s with family %s\n",
+                set->name, set->family == NFPROTO_IPV4 ? "inet" : "inet6");
+#endif
+
+       if (unlikely(!ip_set_optattr_netorder(tb, IPSET_ATTR_HASHSIZE) ||
+                    !ip_set_optattr_netorder(tb, IPSET_ATTR_MAXELEM) ||
+                    !ip_set_optattr_netorder(tb, IPSET_ATTR_TIMEOUT) ||
+                    !ip_set_optattr_netorder(tb, IPSET_ATTR_CADT_FLAGS)))
+               return -IPSET_ERR_PROTOCOL;
+
+       if (tb[IPSET_ATTR_HASHSIZE]) {
+               hashsize = ip_set_get_h32(tb[IPSET_ATTR_HASHSIZE]);
+               if (hashsize < IPSET_MIMINAL_HASHSIZE)
+                       hashsize = IPSET_MIMINAL_HASHSIZE;
+       }
+
+       if (tb[IPSET_ATTR_MAXELEM])
+               maxelem = ip_set_get_h32(tb[IPSET_ATTR_MAXELEM]);
+
+#ifdef IP_SET_HASH_WITH_NETMASK
+       if (tb[IPSET_ATTR_NETMASK]) {
+               netmask = nla_get_u8(tb[IPSET_ATTR_NETMASK]);
+
+               if ((set->family == NFPROTO_IPV4 && netmask > 32) ||
+                   (set->family == NFPROTO_IPV6 && netmask > 128) ||
+                   netmask == 0)
+                       return -IPSET_ERR_INVALID_NETMASK;
+       }
+#endif
+
+       hsize = sizeof(*h);
+#ifdef IP_SET_HASH_WITH_NETS
+       hsize += sizeof(struct net_prefixes) *
+               (set->family == NFPROTO_IPV4 ? 32 : 128);
+#endif
+       h = kzalloc(hsize, GFP_KERNEL);
+       if (!h)
+               return -ENOMEM;
+
+       h->maxelem = maxelem;
+#ifdef IP_SET_HASH_WITH_NETMASK
+       h->netmask = netmask;
+#endif
+       get_random_bytes(&h->initval, sizeof(h->initval));
+       h->timeout = IPSET_NO_TIMEOUT;
+
+       hbits = htable_bits(hashsize);
+       hsize = htable_size(hbits);
+       if (hsize == 0) {
+               kfree(h);
+               return -ENOMEM;
+       }
+       h->table = ip_set_alloc(hsize);
+       if (!h->table) {
+               kfree(h);
+               return -ENOMEM;
+       }
+       h->table->htable_bits = hbits;
+
+       set->data = h;
+       if (set->family ==  NFPROTO_IPV4)
+               set->variant = &TOKEN(HTYPE, 4_variant);
+       else
+               set->variant = &TOKEN(HTYPE, 6_variant);
+
+       if (tb[IPSET_ATTR_TIMEOUT]) {
+               h->timeout = ip_set_timeout_uget(tb[IPSET_ATTR_TIMEOUT]);
+               set->extensions |= IPSET_EXT_TIMEOUT;
+               if (set->family == NFPROTO_IPV4) {
+                       h->dsize = sizeof(struct TOKEN(HTYPE, 4t_elem));
+                       h->offset[IPSET_OFFSET_TIMEOUT] =
+                               offsetof(struct TOKEN(HTYPE, 4t_elem),
+                                        timeout);
+                       TOKEN(HTYPE, 4_gc_init)(set, TOKEN(HTYPE, 4_gc));
+               } else {
+                       h->dsize = sizeof(struct TOKEN(HTYPE, 6t_elem));
+                       h->offset[IPSET_OFFSET_TIMEOUT] =
+                               offsetof(struct TOKEN(HTYPE, 6t_elem),
+                                        timeout);
+                       TOKEN(HTYPE, 6_gc_init)(set, TOKEN(HTYPE, 6_gc));
+               }
+       } else {
+               if (set->family == NFPROTO_IPV4)
+                       h->dsize = sizeof(struct TOKEN(HTYPE, 4_elem));
+               else
+                       h->dsize = sizeof(struct TOKEN(HTYPE, 6_elem));
+       }
+
+       pr_debug("create %s hashsize %u (%u) maxelem %u: %p(%p)\n",
+                set->name, jhash_size(h->table->htable_bits),
+                h->table->htable_bits, h->maxelem, set->data, h->table);
+
+       return 0;
+}
+#endif /* IP_SET_EMIT_CREATE */