summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--include/net/netfilter/nf_conntrack_core.h4
-rw-r--r--include/net/netfilter/nf_conntrack_tuple.h3
-rw-r--r--net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c17
-rw-r--r--net/netfilter/nf_conntrack_core.c100
-rw-r--r--net/netfilter/nf_conntrack_helper.c5
-rw-r--r--net/netfilter/nf_conntrack_netlink.c6
-rw-r--r--net/netfilter/nf_conntrack_standalone.c17
7 files changed, 83 insertions, 69 deletions
diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h
index 3bf7d05ea64d..6351948654b3 100644
--- a/include/net/netfilter/nf_conntrack_core.h
+++ b/include/net/netfilter/nf_conntrack_core.h
@@ -84,9 +84,9 @@ print_tuple(struct seq_file *s, const struct nf_conntrack_tuple *tuple,
struct nf_conntrack_l3proto *l3proto,
struct nf_conntrack_l4proto *proto);
-extern struct list_head *nf_conntrack_hash;
+extern struct hlist_head *nf_conntrack_hash;
extern struct list_head nf_conntrack_expect_list;
extern rwlock_t nf_conntrack_lock ;
-extern struct list_head unconfirmed;
+extern struct hlist_head unconfirmed;
#endif /* _NF_CONNTRACK_CORE_H */
diff --git a/include/net/netfilter/nf_conntrack_tuple.h b/include/net/netfilter/nf_conntrack_tuple.h
index 5d72b16e876f..d02ce876b4ca 100644
--- a/include/net/netfilter/nf_conntrack_tuple.h
+++ b/include/net/netfilter/nf_conntrack_tuple.h
@@ -125,8 +125,7 @@ DEBUGP("tuple %p: %u %u " NIP6_FMT " %hu -> " NIP6_FMT " %hu\n", \
/* Connections have two entries in the hash table: one for each way */
struct nf_conntrack_tuple_hash
{
- struct list_head list;
-
+ struct hlist_node hnode;
struct nf_conntrack_tuple tuple;
};
diff --git a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c
index 89f933e81035..888f27fd884f 100644
--- a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c
+++ b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c
@@ -41,35 +41,36 @@ struct ct_iter_state {
unsigned int bucket;
};
-static struct list_head *ct_get_first(struct seq_file *seq)
+static struct hlist_node *ct_get_first(struct seq_file *seq)
{
struct ct_iter_state *st = seq->private;
for (st->bucket = 0;
st->bucket < nf_conntrack_htable_size;
st->bucket++) {
- if (!list_empty(&nf_conntrack_hash[st->bucket]))
- return nf_conntrack_hash[st->bucket].next;
+ if (!hlist_empty(&nf_conntrack_hash[st->bucket]))
+ return nf_conntrack_hash[st->bucket].first;
}
return NULL;
}
-static struct list_head *ct_get_next(struct seq_file *seq, struct list_head *head)
+static struct hlist_node *ct_get_next(struct seq_file *seq,
+ struct hlist_node *head)
{
struct ct_iter_state *st = seq->private;
head = head->next;
- while (head == &nf_conntrack_hash[st->bucket]) {
+ while (head == NULL) {
if (++st->bucket >= nf_conntrack_htable_size)
return NULL;
- head = nf_conntrack_hash[st->bucket].next;
+ head = nf_conntrack_hash[st->bucket].first;
}
return head;
}
-static struct list_head *ct_get_idx(struct seq_file *seq, loff_t pos)
+static struct hlist_node *ct_get_idx(struct seq_file *seq, loff_t pos)
{
- struct list_head *head = ct_get_first(seq);
+ struct hlist_node *head = ct_get_first(seq);
if (head)
while (pos && (head = ct_get_next(seq, head)))
diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c
index 54acac5c6ea7..992d0ef31fa3 100644
--- a/net/netfilter/nf_conntrack_core.c
+++ b/net/netfilter/nf_conntrack_core.c
@@ -59,14 +59,14 @@ EXPORT_SYMBOL_GPL(nf_conntrack_htable_size);
int nf_conntrack_max __read_mostly;
EXPORT_SYMBOL_GPL(nf_conntrack_max);
-struct list_head *nf_conntrack_hash __read_mostly;
+struct hlist_head *nf_conntrack_hash __read_mostly;
EXPORT_SYMBOL_GPL(nf_conntrack_hash);
struct nf_conn nf_conntrack_untracked __read_mostly;
EXPORT_SYMBOL_GPL(nf_conntrack_untracked);
unsigned int nf_ct_log_invalid __read_mostly;
-LIST_HEAD(unconfirmed);
+HLIST_HEAD(unconfirmed);
static int nf_conntrack_vmalloc __read_mostly;
static struct kmem_cache *nf_conntrack_cachep __read_mostly;
static unsigned int nf_conntrack_next_id;
@@ -142,8 +142,8 @@ static void
clean_from_lists(struct nf_conn *ct)
{
DEBUGP("clean_from_lists(%p)\n", ct);
- list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
- list_del(&ct->tuplehash[IP_CT_DIR_REPLY].list);
+ hlist_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnode);
+ hlist_del(&ct->tuplehash[IP_CT_DIR_REPLY].hnode);
/* Destroy all pending expectations */
nf_ct_remove_expectations(ct);
@@ -184,8 +184,8 @@ destroy_conntrack(struct nf_conntrack *nfct)
/* We overload first tuple to link into unconfirmed list. */
if (!nf_ct_is_confirmed(ct)) {
- BUG_ON(list_empty(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list));
- list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+ BUG_ON(hlist_unhashed(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnode));
+ hlist_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnode);
}
NF_CT_STAT_INC(delete);
@@ -226,9 +226,10 @@ __nf_conntrack_find(const struct nf_conntrack_tuple *tuple,
const struct nf_conn *ignored_conntrack)
{
struct nf_conntrack_tuple_hash *h;
+ struct hlist_node *n;
unsigned int hash = hash_conntrack(tuple);
- list_for_each_entry(h, &nf_conntrack_hash[hash], list) {
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[hash], hnode) {
if (nf_ct_tuplehash_to_ctrack(h) != ignored_conntrack &&
nf_ct_tuple_equal(tuple, &h->tuple)) {
NF_CT_STAT_INC(found);
@@ -263,10 +264,10 @@ static void __nf_conntrack_hash_insert(struct nf_conn *ct,
unsigned int repl_hash)
{
ct->id = ++nf_conntrack_next_id;
- list_add(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list,
- &nf_conntrack_hash[hash]);
- list_add(&ct->tuplehash[IP_CT_DIR_REPLY].list,
- &nf_conntrack_hash[repl_hash]);
+ hlist_add_head(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnode,
+ &nf_conntrack_hash[hash]);
+ hlist_add_head(&ct->tuplehash[IP_CT_DIR_REPLY].hnode,
+ &nf_conntrack_hash[repl_hash]);
}
void nf_conntrack_hash_insert(struct nf_conn *ct)
@@ -290,6 +291,7 @@ __nf_conntrack_confirm(struct sk_buff **pskb)
struct nf_conntrack_tuple_hash *h;
struct nf_conn *ct;
struct nf_conn_help *help;
+ struct hlist_node *n;
enum ip_conntrack_info ctinfo;
ct = nf_ct_get(*pskb, &ctinfo);
@@ -319,17 +321,17 @@ __nf_conntrack_confirm(struct sk_buff **pskb)
/* See if there's one in the list already, including reverse:
NAT could have grabbed it without realizing, since we're
not in the hash. If there is, we lost race. */
- list_for_each_entry(h, &nf_conntrack_hash[hash], list)
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[hash], hnode)
if (nf_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple,
&h->tuple))
goto out;
- list_for_each_entry(h, &nf_conntrack_hash[repl_hash], list)
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[repl_hash], hnode)
if (nf_ct_tuple_equal(&ct->tuplehash[IP_CT_DIR_REPLY].tuple,
&h->tuple))
goto out;
/* Remove from unconfirmed list */
- list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list);
+ hlist_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].hnode);
__nf_conntrack_hash_insert(ct, hash, repl_hash);
/* Timer relative to confirmation time, not original
@@ -378,22 +380,22 @@ EXPORT_SYMBOL_GPL(nf_conntrack_tuple_taken);
/* There's a small race here where we may free a just-assured
connection. Too bad: we're in trouble anyway. */
-static int early_drop(struct list_head *chain)
+static int early_drop(struct hlist_head *chain)
{
- /* Traverse backwards: gives us oldest, which is roughly LRU */
+ /* Use oldest entry, which is roughly LRU */
struct nf_conntrack_tuple_hash *h;
struct nf_conn *ct = NULL, *tmp;
+ struct hlist_node *n;
int dropped = 0;
read_lock_bh(&nf_conntrack_lock);
- list_for_each_entry_reverse(h, chain, list) {
+ hlist_for_each_entry(h, n, chain, hnode) {
tmp = nf_ct_tuplehash_to_ctrack(h);
- if (!test_bit(IPS_ASSURED_BIT, &tmp->status)) {
+ if (!test_bit(IPS_ASSURED_BIT, &tmp->status))
ct = tmp;
- atomic_inc(&ct->ct_general.use);
- break;
- }
}
+ if (ct)
+ atomic_inc(&ct->ct_general.use);
read_unlock_bh(&nf_conntrack_lock);
if (!ct)
@@ -535,7 +537,8 @@ init_conntrack(const struct nf_conntrack_tuple *tuple,
}
/* Overload tuple linked list to put us in unconfirmed list. */
- list_add(&conntrack->tuplehash[IP_CT_DIR_ORIGINAL].list, &unconfirmed);
+ hlist_add_head(&conntrack->tuplehash[IP_CT_DIR_ORIGINAL].hnode,
+ &unconfirmed);
write_unlock_bh(&nf_conntrack_lock);
@@ -873,16 +876,17 @@ get_next_corpse(int (*iter)(struct nf_conn *i, void *data),
{
struct nf_conntrack_tuple_hash *h;
struct nf_conn *ct;
+ struct hlist_node *n;
write_lock_bh(&nf_conntrack_lock);
for (; *bucket < nf_conntrack_htable_size; (*bucket)++) {
- list_for_each_entry(h, &nf_conntrack_hash[*bucket], list) {
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[*bucket], hnode) {
ct = nf_ct_tuplehash_to_ctrack(h);
if (iter(ct, data))
goto found;
}
}
- list_for_each_entry(h, &unconfirmed, list) {
+ hlist_for_each_entry(h, n, &unconfirmed, hnode) {
ct = nf_ct_tuplehash_to_ctrack(h);
if (iter(ct, data))
set_bit(IPS_DYING_BIT, &ct->status);
@@ -917,13 +921,14 @@ static int kill_all(struct nf_conn *i, void *data)
return 1;
}
-static void free_conntrack_hash(struct list_head *hash, int vmalloced, int size)
+static void free_conntrack_hash(struct hlist_head *hash, int vmalloced,
+ int size)
{
if (vmalloced)
vfree(hash);
else
free_pages((unsigned long)hash,
- get_order(sizeof(struct list_head) * size));
+ get_order(sizeof(struct hlist_head) * size));
}
void nf_conntrack_flush(void)
@@ -965,26 +970,26 @@ void nf_conntrack_cleanup(void)
nf_conntrack_helper_fini();
}
-static struct list_head *alloc_hashtable(int *sizep, int *vmalloced)
+static struct hlist_head *alloc_hashtable(int *sizep, int *vmalloced)
{
- struct list_head *hash;
+ struct hlist_head *hash;
unsigned int size, i;
*vmalloced = 0;
- size = *sizep = roundup(*sizep, PAGE_SIZE / sizeof(struct list_head));
+ size = *sizep = roundup(*sizep, PAGE_SIZE / sizeof(struct hlist_head));
hash = (void*)__get_free_pages(GFP_KERNEL,
- get_order(sizeof(struct list_head)
+ get_order(sizeof(struct hlist_head)
* size));
if (!hash) {
*vmalloced = 1;
printk(KERN_WARNING "nf_conntrack: falling back to vmalloc.\n");
- hash = vmalloc(sizeof(struct list_head) * size);
+ hash = vmalloc(sizeof(struct hlist_head) * size);
}
if (hash)
for (i = 0; i < size; i++)
- INIT_LIST_HEAD(&hash[i]);
+ INIT_HLIST_HEAD(&hash[i]);
return hash;
}
@@ -994,7 +999,7 @@ int set_hashsize(const char *val, struct kernel_param *kp)
int i, bucket, hashsize, vmalloced;
int old_vmalloced, old_size;
int rnd;
- struct list_head *hash, *old_hash;
+ struct hlist_head *hash, *old_hash;
struct nf_conntrack_tuple_hash *h;
/* On boot, we can set this without any fancy locking. */
@@ -1015,12 +1020,12 @@ int set_hashsize(const char *val, struct kernel_param *kp)
write_lock_bh(&nf_conntrack_lock);
for (i = 0; i < nf_conntrack_htable_size; i++) {
- while (!list_empty(&nf_conntrack_hash[i])) {
- h = list_entry(nf_conntrack_hash[i].next,
- struct nf_conntrack_tuple_hash, list);
- list_del(&h->list);
+ while (!hlist_empty(&nf_conntrack_hash[i])) {
+ h = hlist_entry(nf_conntrack_hash[i].first,
+ struct nf_conntrack_tuple_hash, hnode);
+ hlist_del(&h->hnode);
bucket = __hash_conntrack(&h->tuple, hashsize, rnd);
- list_add_tail(&h->list, &hash[bucket]);
+ hlist_add_head(&h->hnode, &hash[bucket]);
}
}
old_size = nf_conntrack_htable_size;
@@ -1042,18 +1047,25 @@ module_param_call(hashsize, set_hashsize, param_get_uint,
int __init nf_conntrack_init(void)
{
+ int max_factor = 8;
int ret;
/* Idea from tcp.c: use 1/16384 of memory. On i386: 32MB
- * machine has 256 buckets. >= 1GB machines have 8192 buckets. */
+ * machine has 512 buckets. >= 1GB machines have 16384 buckets. */
if (!nf_conntrack_htable_size) {
nf_conntrack_htable_size
= (((num_physpages << PAGE_SHIFT) / 16384)
- / sizeof(struct list_head));
+ / sizeof(struct hlist_head));
if (num_physpages > (1024 * 1024 * 1024 / PAGE_SIZE))
- nf_conntrack_htable_size = 8192;
- if (nf_conntrack_htable_size < 16)
- nf_conntrack_htable_size = 16;
+ nf_conntrack_htable_size = 16384;
+ if (nf_conntrack_htable_size < 32)
+ nf_conntrack_htable_size = 32;
+
+ /* Use a max. factor of four by default to get the same max as
+ * with the old struct list_heads. When a table size is given
+ * we use the old value of 8 to avoid reducing the max.
+ * entries. */
+ max_factor = 4;
}
nf_conntrack_hash = alloc_hashtable(&nf_conntrack_htable_size,
&nf_conntrack_vmalloc);
@@ -1062,7 +1074,7 @@ int __init nf_conntrack_init(void)
goto err_out;
}
- nf_conntrack_max = 8 * nf_conntrack_htable_size;
+ nf_conntrack_max = max_factor * nf_conntrack_htable_size;
printk("nf_conntrack version %s (%u buckets, %d max)\n",
NF_CONNTRACK_VERSION, nf_conntrack_htable_size,
diff --git a/net/netfilter/nf_conntrack_helper.c b/net/netfilter/nf_conntrack_helper.c
index dc352f509477..3fc6e9f0de1a 100644
--- a/net/netfilter/nf_conntrack_helper.c
+++ b/net/netfilter/nf_conntrack_helper.c
@@ -116,6 +116,7 @@ void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me)
unsigned int i;
struct nf_conntrack_tuple_hash *h;
struct nf_conntrack_expect *exp, *tmp;
+ struct hlist_node *n;
/* Need write lock here, to delete helper. */
write_lock_bh(&nf_conntrack_lock);
@@ -132,10 +133,10 @@ void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me)
}
/* Get rid of expecteds, set helpers to NULL. */
- list_for_each_entry(h, &unconfirmed, list)
+ hlist_for_each_entry(h, n, &unconfirmed, hnode)
unhelp(h, me);
for (i = 0; i < nf_conntrack_htable_size; i++) {
- list_for_each_entry(h, &nf_conntrack_hash[i], list)
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[i], hnode)
unhelp(h, me);
}
write_unlock_bh(&nf_conntrack_lock);
diff --git a/net/netfilter/nf_conntrack_netlink.c b/net/netfilter/nf_conntrack_netlink.c
index 3d56f36074f7..0627559ca470 100644
--- a/net/netfilter/nf_conntrack_netlink.c
+++ b/net/netfilter/nf_conntrack_netlink.c
@@ -428,7 +428,7 @@ ctnetlink_dump_table(struct sk_buff *skb, struct netlink_callback *cb)
{
struct nf_conn *ct, *last;
struct nf_conntrack_tuple_hash *h;
- struct list_head *i;
+ struct hlist_node *n;
struct nfgenmsg *nfmsg = NLMSG_DATA(cb->nlh);
u_int8_t l3proto = nfmsg->nfgen_family;
@@ -436,8 +436,8 @@ ctnetlink_dump_table(struct sk_buff *skb, struct netlink_callback *cb)
last = (struct nf_conn *)cb->args[1];
for (; cb->args[0] < nf_conntrack_htable_size; cb->args[0]++) {
restart:
- list_for_each_prev(i, &nf_conntrack_hash[cb->args[0]]) {
- h = (struct nf_conntrack_tuple_hash *) i;
+ hlist_for_each_entry(h, n, &nf_conntrack_hash[cb->args[0]],
+ hnode) {
if (NF_CT_DIRECTION(h) != IP_CT_DIR_ORIGINAL)
continue;
ct = nf_ct_tuplehash_to_ctrack(h);
diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
index 45baeb0e30f9..fe536b20408b 100644
--- a/net/netfilter/nf_conntrack_standalone.c
+++ b/net/netfilter/nf_conntrack_standalone.c
@@ -60,35 +60,36 @@ struct ct_iter_state {
unsigned int bucket;
};
-static struct list_head *ct_get_first(struct seq_file *seq)
+static struct hlist_node *ct_get_first(struct seq_file *seq)
{
struct ct_iter_state *st = seq->private;
for (st->bucket = 0;
st->bucket < nf_conntrack_htable_size;
st->bucket++) {
- if (!list_empty(&nf_conntrack_hash[st->bucket]))
- return nf_conntrack_hash[st->bucket].next;
+ if (!hlist_empty(&nf_conntrack_hash[st->bucket]))
+ return nf_conntrack_hash[st->bucket].first;
}
return NULL;
}
-static struct list_head *ct_get_next(struct seq_file *seq, struct list_head *head)
+static struct hlist_node *ct_get_next(struct seq_file *seq,
+ struct hlist_node *head)
{
struct ct_iter_state *st = seq->private;
head = head->next;
- while (head == &nf_conntrack_hash[st->bucket]) {
+ while (head == NULL) {
if (++st->bucket >= nf_conntrack_htable_size)
return NULL;
- head = nf_conntrack_hash[st->bucket].next;
+ head = nf_conntrack_hash[st->bucket].first;
}
return head;
}
-static struct list_head *ct_get_idx(struct seq_file *seq, loff_t pos)
+static struct hlist_node *ct_get_idx(struct seq_file *seq, loff_t pos)
{
- struct list_head *head = ct_get_first(seq);
+ struct hlist_node *head = ct_get_first(seq);
if (head)
while (pos && (head = ct_get_next(seq, head)))