diff options
Diffstat (limited to 'net/ipv4')
42 files changed, 2759 insertions, 768 deletions
diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index bf147f8db39..a9d84f93442 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1248,11 +1248,6 @@ module_init(inet_init); /* ------------------------------------------------------------------------ */ #ifdef CONFIG_PROC_FS -#ifdef CONFIG_IP_FIB_TRIE -extern int fib_stat_proc_init(void); -extern void fib_stat_proc_exit(void); -#endif - static int __init ipv4_proc_init(void) { int rc = 0; @@ -1265,19 +1260,11 @@ static int __init ipv4_proc_init(void) goto out_udp; if (fib_proc_init()) goto out_fib; -#ifdef CONFIG_IP_FIB_TRIE - if (fib_stat_proc_init()) - goto out_fib_stat; -#endif if (ip_misc_proc_init()) goto out_misc; out: return rc; out_misc: -#ifdef CONFIG_IP_FIB_TRIE - fib_stat_proc_exit(); -out_fib_stat: -#endif fib_proc_exit(); out_fib: udp4_proc_exit(); diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c index 514c85b2631..035ad2c9e1b 100644 --- a/net/ipv4/ah4.c +++ b/net/ipv4/ah4.c @@ -263,10 +263,8 @@ static int ah_init_state(struct xfrm_state *x) error: if (ahp) { - if (ahp->work_icv) - kfree(ahp->work_icv); - if (ahp->tfm) - crypto_free_tfm(ahp->tfm); + kfree(ahp->work_icv); + crypto_free_tfm(ahp->tfm); kfree(ahp); } return -EINVAL; @@ -279,14 +277,10 @@ static void ah_destroy(struct xfrm_state *x) if (!ahp) return; - if (ahp->work_icv) { - kfree(ahp->work_icv); - ahp->work_icv = NULL; - } - if (ahp->tfm) { - crypto_free_tfm(ahp->tfm); - ahp->tfm = NULL; - } + kfree(ahp->work_icv); + ahp->work_icv = NULL; + crypto_free_tfm(ahp->tfm); + ahp->tfm = NULL; kfree(ahp); } diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index b31ffc5053d..1b5a09d1b90 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -343,22 +343,14 @@ static void esp_destroy(struct xfrm_state *x) if (!esp) return; - if (esp->conf.tfm) { - crypto_free_tfm(esp->conf.tfm); - esp->conf.tfm = NULL; - } - if (esp->conf.ivec) { - kfree(esp->conf.ivec); - esp->conf.ivec = NULL; - } - if (esp->auth.tfm) { - crypto_free_tfm(esp->auth.tfm); - esp->auth.tfm = NULL; - } - if (esp->auth.work_icv) { - kfree(esp->auth.work_icv); - esp->auth.work_icv = NULL; - } + crypto_free_tfm(esp->conf.tfm); + esp->conf.tfm = NULL; + kfree(esp->conf.ivec); + esp->conf.ivec = NULL; + crypto_free_tfm(esp->auth.tfm); + esp->auth.tfm = NULL; + kfree(esp->auth.work_icv); + esp->auth.work_icv = NULL; kfree(esp); } diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index b2dea4e5da7..50c0519cd70 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -43,7 +43,7 @@ * 2 of the License, or (at your option) any later version. */ -#define VERSION "0.402" +#define VERSION "0.404" #include <linux/config.h> #include <asm/uaccess.h> @@ -164,7 +164,6 @@ static struct node *resize(struct trie *t, struct tnode *tn); static struct tnode *inflate(struct trie *t, struct tnode *tn); static struct tnode *halve(struct trie *t, struct tnode *tn); static void tnode_free(struct tnode *tn); -static void trie_dump_seq(struct seq_file *seq, struct trie *t); static kmem_cache_t *fn_alias_kmem __read_mostly; static struct trie *trie_local = NULL, *trie_main = NULL; @@ -225,7 +224,7 @@ static inline int tkey_mismatch(t_key a, int offset, t_key b) Consider a node 'n' and its parent 'tp'. If n is a leaf, every bit in its key is significant. Its presence is - necessitaded by path compression, since during a tree traversal (when + necessitated by path compression, since during a tree traversal (when searching for a leaf - unless we are doing an insertion) we will completely ignore all skipped bits we encounter. Thus we need to verify, at the end of a potentially successful search, that we have indeed been walking the @@ -837,11 +836,12 @@ static void trie_init(struct trie *t) #endif } -/* readside most use rcu_read_lock currently dump routines +/* readside must use rcu_read_lock currently dump routines via get_fa_head and dump */ -static struct leaf_info *find_leaf_info(struct hlist_head *head, int plen) +static struct leaf_info *find_leaf_info(struct leaf *l, int plen) { + struct hlist_head *head = &l->list; struct hlist_node *node; struct leaf_info *li; @@ -854,7 +854,7 @@ static struct leaf_info *find_leaf_info(struct hlist_head *head, int plen) static inline struct list_head * get_fa_head(struct leaf *l, int plen) { - struct leaf_info *li = find_leaf_info(&l->list, plen); + struct leaf_info *li = find_leaf_info(l, plen); if (!li) return NULL; @@ -1086,7 +1086,7 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) } if (tp && tp->pos + tp->bits > 32) - printk("ERROR tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", + printk(KERN_WARNING "fib_trie tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", tp, tp->pos, tp->bits, key, plen); /* Rebalance the trie */ @@ -1249,7 +1249,7 @@ err: } -/* should be clalled with rcu_read_lock */ +/* should be called with rcu_read_lock */ static inline int check_leaf(struct trie *t, struct leaf *l, t_key key, int *plen, const struct flowi *flp, struct fib_result *res) @@ -1591,7 +1591,7 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); l = fib_find_node(t, key); - li = find_leaf_info(&l->list, plen); + li = find_leaf_info(l, plen); list_del_rcu(&fa->fa_list); @@ -1715,7 +1715,6 @@ static int fn_trie_flush(struct fib_table *tb) t->revision++; - rcu_read_lock(); for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { found += trie_flush_leaf(t, l); @@ -1723,7 +1722,6 @@ static int fn_trie_flush(struct fib_table *tb) trie_leaf_remove(t, ll->key); ll = l; } - rcu_read_unlock(); if (ll && hlist_empty(&ll->list)) trie_leaf_remove(t, ll->key); @@ -1834,16 +1832,7 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi i++; continue; } - if (fa->fa_info->fib_nh == NULL) { - printk("Trie error _fib_nh=NULL in fa[%d] k=%08x plen=%d\n", i, key, plen); - i++; - continue; - } - if (fa->fa_info == NULL) { - printk("Trie error fa_info=NULL in fa[%d] k=%08x plen=%d\n", i, key, plen); - i++; - continue; - } + BUG_ON(!fa->fa_info); if (fib_dump_info(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, @@ -1966,563 +1955,531 @@ struct fib_table * __init fib_hash_init(int id) trie_main = t; if (id == RT_TABLE_LOCAL) - printk("IPv4 FIB: Using LC-trie version %s\n", VERSION); + printk(KERN_INFO "IPv4 FIB: Using LC-trie version %s\n", VERSION); return tb; } -/* Trie dump functions */ +#ifdef CONFIG_PROC_FS +/* Depth first Trie walk iterator */ +struct fib_trie_iter { + struct tnode *tnode; + struct trie *trie; + unsigned index; + unsigned depth; +}; -static void putspace_seq(struct seq_file *seq, int n) +static struct node *fib_trie_get_next(struct fib_trie_iter *iter) { - while (n--) - seq_printf(seq, " "); -} + struct tnode *tn = iter->tnode; + unsigned cindex = iter->index; + struct tnode *p; -static void printbin_seq(struct seq_file *seq, unsigned int v, int bits) -{ - while (bits--) - seq_printf(seq, "%s", (v & (1<<bits))?"1":"0"); -} + pr_debug("get_next iter={node=%p index=%d depth=%d}\n", + iter->tnode, iter->index, iter->depth); +rescan: + while (cindex < (1<<tn->bits)) { + struct node *n = tnode_get_child(tn, cindex); -static void printnode_seq(struct seq_file *seq, int indent, struct node *n, - int pend, int cindex, int bits) -{ - putspace_seq(seq, indent); - if (IS_LEAF(n)) - seq_printf(seq, "|"); - else - seq_printf(seq, "+"); - if (bits) { - seq_printf(seq, "%d/", cindex); - printbin_seq(seq, cindex, bits); - seq_printf(seq, ": "); - } else - seq_printf(seq, "<root>: "); - seq_printf(seq, "%s:%p ", IS_LEAF(n)?"Leaf":"Internal node", n); + if (n) { + if (IS_LEAF(n)) { + iter->tnode = tn; + iter->index = cindex + 1; + } else { + /* push down one level */ + iter->tnode = (struct tnode *) n; + iter->index = 0; + ++iter->depth; + } + return n; + } - if (IS_LEAF(n)) { - struct leaf *l = (struct leaf *)n; - struct fib_alias *fa; - int i; + ++cindex; + } - seq_printf(seq, "key=%d.%d.%d.%d\n", - n->key >> 24, (n->key >> 16) % 256, (n->key >> 8) % 256, n->key % 256); - - for (i = 32; i >= 0; i--) - if (find_leaf_info(&l->list, i)) { - struct list_head *fa_head = get_fa_head(l, i); - - if (!fa_head) - continue; - - if (list_empty(fa_head)) - continue; - - putspace_seq(seq, indent+2); - seq_printf(seq, "{/%d...dumping}\n", i); - - list_for_each_entry_rcu(fa, fa_head, fa_list) { - putspace_seq(seq, indent+2); - if (fa->fa_info == NULL) { - seq_printf(seq, "Error fa_info=NULL\n"); - continue; - } - if (fa->fa_info->fib_nh == NULL) { - seq_printf(seq, "Error _fib_nh=NULL\n"); - continue; - } - - seq_printf(seq, "{type=%d scope=%d TOS=%d}\n", - fa->fa_type, - fa->fa_scope, - fa->fa_tos); - } - } - } else { - struct tnode *tn = (struct tnode *)n; - int plen = ((struct tnode *)n)->pos; - t_key prf = MASK_PFX(n->key, plen); - - seq_printf(seq, "key=%d.%d.%d.%d/%d\n", - prf >> 24, (prf >> 16) % 256, (prf >> 8) % 256, prf % 256, plen); - - putspace_seq(seq, indent); seq_printf(seq, "| "); - seq_printf(seq, "{key prefix=%08x/", tn->key & TKEY_GET_MASK(0, tn->pos)); - printbin_seq(seq, tkey_extract_bits(tn->key, 0, tn->pos), tn->pos); - seq_printf(seq, "}\n"); - putspace_seq(seq, indent); seq_printf(seq, "| "); - seq_printf(seq, "{pos=%d", tn->pos); - seq_printf(seq, " (skip=%d bits)", tn->pos - pend); - seq_printf(seq, " bits=%d (%u children)}\n", tn->bits, (1 << tn->bits)); - putspace_seq(seq, indent); seq_printf(seq, "| "); - seq_printf(seq, "{empty=%d full=%d}\n", tn->empty_children, tn->full_children); + /* Current node exhausted, pop back up */ + p = NODE_PARENT(tn); + if (p) { + cindex = tkey_extract_bits(tn->key, p->pos, p->bits)+1; + tn = p; + --iter->depth; + goto rescan; } + + /* got root? */ + return NULL; } -static void trie_dump_seq(struct seq_file *seq, struct trie *t) +static struct node *fib_trie_get_first(struct fib_trie_iter *iter, + struct trie *t) { - struct node *n; - int cindex = 0; - int indent = 1; - int pend = 0; - int depth = 0; - struct tnode *tn; + struct node *n = rcu_dereference(t->trie); - rcu_read_lock(); - n = rcu_dereference(t->trie); - seq_printf(seq, "------ trie_dump of t=%p ------\n", t); - - if (!n) { - seq_printf(seq, "------ trie is empty\n"); - - rcu_read_unlock(); - return; - } - - printnode_seq(seq, indent, n, pend, cindex, 0); - - if (!IS_TNODE(n)) { - rcu_read_unlock(); - return; + if (n && IS_TNODE(n)) { + iter->tnode = (struct tnode *) n; + iter->trie = t; + iter->index = 0; + iter->depth = 1; + return n; } + return NULL; +} - tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - putspace_seq(seq, indent); seq_printf(seq, "\\--\n"); - indent += 3; - depth++; - - while (tn && cindex < (1 << tn->bits)) { - struct node *child = rcu_dereference(tn->child[cindex]); - if (!child) - cindex++; - else { - /* Got a child */ - printnode_seq(seq, indent, child, pend, - cindex, tn->bits); - - if (IS_LEAF(child)) - cindex++; - - else { - /* - * New tnode. Decend one level - */ - - depth++; - n = child; - tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - putspace_seq(seq, indent); - seq_printf(seq, "\\--\n"); - indent += 3; - cindex = 0; - } - } - - /* - * Test if we are done - */ - - while (cindex >= (1 << tn->bits)) { - /* - * Move upwards and test for root - * pop off all traversed nodes - */ +static void trie_collect_stats(struct trie *t, struct trie_stat *s) +{ + struct node *n; + struct fib_trie_iter iter; - if (NODE_PARENT(tn) == NULL) { - tn = NULL; - break; - } + memset(s, 0, sizeof(*s)); - cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); - cindex++; - tn = NODE_PARENT(tn); - pend = tn->pos + tn->bits; - indent -= 3; - depth--; + rcu_read_lock(); + for (n = fib_trie_get_first(&iter, t); n; + n = fib_trie_get_next(&iter)) { + if (IS_LEAF(n)) { + s->leaves++; + s->totdepth += iter.depth; + if (iter.depth > s->maxdepth) + s->maxdepth = iter.depth; + } else { + const struct tnode *tn = (const struct tnode *) n; + int i; + + s->tnodes++; + s->nodesizes[tn->bits]++; + for (i = 0; i < (1<<tn->bits); i++) + if (!tn->child[i]) + s->nullpointers++; } } rcu_read_unlock(); } -static struct trie_stat *trie_stat_new(void) +/* + * This outputs /proc/net/fib_triestats + */ +static void trie_show_stats(struct seq_file *seq, struct trie_stat *stat) { - struct trie_stat *s; - int i; + unsigned i, max, pointers, bytes, avdepth; - s = kmalloc(sizeof(struct trie_stat), GFP_KERNEL); - if (!s) - return NULL; + if (stat->leaves) + avdepth = stat->totdepth*100 / stat->leaves; + else + avdepth = 0; - s->totdepth = 0; - s->maxdepth = 0; - s->tnodes = 0; - s->leaves = 0; - s->nullpointers = 0; + seq_printf(seq, "\tAver depth: %d.%02d\n", avdepth / 100, avdepth % 100 ); + seq_printf(seq, "\tMax depth: %u\n", stat->maxdepth); - for (i = 0; i < MAX_CHILDS; i++) - s->nodesizes[i] = 0; + seq_printf(seq, "\tLeaves: %u\n", stat->leaves); - return s; -} + bytes = sizeof(struct leaf) * stat->leaves; + seq_printf(seq, "\tInternal nodes: %d\n\t", stat->tnodes); + bytes += sizeof(struct tnode) * stat->tnodes; -static struct trie_stat *trie_collect_stats(struct trie *t) -{ - struct node *n; - struct trie_stat *s = trie_stat_new(); - int cindex = 0; - int pend = 0; - int depth = 0; + max = MAX_CHILDS-1; + while (max >= 0 && stat->nodesizes[max] == 0) + max--; - if (!s) - return NULL; + pointers = 0; + for (i = 1; i <= max; i++) + if (stat->nodesizes[i] != 0) { + seq_printf(seq, " %d: %d", i, stat->nodesizes[i]); + pointers += (1<<i) * stat->nodesizes[i]; + } + seq_putc(seq, '\n'); + seq_printf(seq, "\tPointers: %d\n", pointers); - rcu_read_lock(); - n = rcu_dereference(t->trie); + bytes += sizeof(struct node *) * pointers; + seq_printf(seq, "Null ptrs: %d\n", stat->nullpointers); + seq_printf(seq, "Total size: %d kB\n", (bytes + 1023) / 1024); - if (!n) - return s; +#ifdef CONFIG_IP_FIB_TRIE_STATS + seq_printf(seq, "Counters:\n---------\n"); + seq_printf(seq,"gets = %d\n", t->stats.gets); + seq_printf(seq,"backtracks = %d\n", t->stats.backtrack); + seq_printf(seq,"semantic match passed = %d\n", t->stats.semantic_match_passed); + seq_printf(seq,"semantic match miss = %d\n", t->stats.semantic_match_miss); + seq_printf(seq,"null node hit= %d\n", t->stats.null_node_hit); + seq_printf(seq,"skipped node resize = %d\n", t->stats.resize_node_skipped); +#ifdef CLEAR_STATS + memset(&(t->stats), 0, sizeof(t->stats)); +#endif +#endif /* CONFIG_IP_FIB_TRIE_STATS */ +} - if (IS_TNODE(n)) { - struct tnode *tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - s->nodesizes[tn->bits]++; - depth++; - - while (tn && cindex < (1 << tn->bits)) { - struct node *ch = rcu_dereference(tn->child[cindex]); - if (ch) { - - /* Got a child */ - - if (IS_LEAF(tn->child[cindex])) { - cindex++; - - /* stats */ - if (depth > s->maxdepth) - s->maxdepth = depth; - s->totdepth += depth; - s->leaves++; - } else { - /* - * New tnode. Decend one level - */ - - s->tnodes++; - s->nodesizes[tn->bits]++; - depth++; - - n = ch; - tn = (struct tnode *)n; - pend = tn->pos+tn->bits; - - cindex = 0; - } - } else { - cindex++; - s->nullpointers++; - } +static int fib_triestat_seq_show(struct seq_file *seq, void *v) +{ + struct trie_stat *stat; - /* - * Test if we are done - */ + stat = kmalloc(sizeof(*stat), GFP_KERNEL); + if (!stat) + return -ENOMEM; - while (cindex >= (1 << tn->bits)) { - /* - * Move upwards and test for root - * pop off all traversed nodes - */ + seq_printf(seq, "Basic info: size of leaf: %Zd bytes, size of tnode: %Zd bytes.\n", + sizeof(struct leaf), sizeof(struct tnode)); - if (NODE_PARENT(tn) == NULL) { - tn = NULL; - n = NULL; - break; - } + if (trie_local) { + seq_printf(seq, "Local:\n"); + trie_collect_stats(trie_local, stat); + trie_show_stats(seq, stat); + } - cindex = tkey_extract_bits(tn->key, NODE_PARENT(tn)->pos, NODE_PARENT(tn)->bits); - tn = NODE_PARENT(tn); - cindex++; - n = (struct node *)tn; - pend = tn->pos+tn->bits; - depth--; - } - } + if (trie_main) { + seq_printf(seq, "Main:\n"); + trie_collect_stats(trie_main, stat); + trie_show_stats(seq, stat); } + kfree(stat); - rcu_read_unlock(); - return s; + return 0; } -#ifdef CONFIG_PROC_FS - -static struct fib_alias *fib_triestat_get_first(struct seq_file *seq) +static int fib_triestat_seq_open(struct inode *inode, struct file *file) { - return NULL; + return single_open(file, fib_triestat_seq_show, NULL); } -static struct fib_alias *fib_triestat_get_next(struct seq_file *seq) +static struct file_operations fib_triestat_fops = { + .owner = THIS_MODULE, + .open = fib_triestat_seq_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static struct node *fib_trie_get_idx(struct fib_trie_iter *iter, + loff_t pos) { + loff_t idx = 0; + struct node *n; + + for (n = fib_trie_get_first(iter, trie_local); + n; ++idx, n = fib_trie_get_next(iter)) { + if (pos == idx) + return n; + } + + for (n = fib_trie_get_first(iter, trie_main); + n; ++idx, n = fib_trie_get_next(iter)) { + if (pos == idx) + return n; + } return NULL; } -static void *fib_triestat_seq_start(struct seq_file *seq, loff_t *pos) +static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos) { - if (!ip_fib_main_table) - return NULL; - - if (*pos) - return fib_triestat_get_next(seq); - else + rcu_read_lock(); + if (*pos == 0) return SEQ_START_TOKEN; + return fib_trie_get_idx(seq->private, *pos - 1); } -static void *fib_triestat_seq_next(struct seq_file *seq, void *v, loff_t *pos) +static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos) { + struct fib_trie_iter *iter = seq->private; + void *l = v; + ++*pos; if (v == SEQ_START_TOKEN) - return fib_triestat_get_first(seq); - else - return fib_triestat_get_next(seq); -} + return fib_trie_get_idx(iter, 0); -static void fib_triestat_seq_stop(struct seq_file *seq, void *v) -{ + v = fib_trie_get_next(iter); + BUG_ON(v == l); + if (v) + return v; -} + /* continue scan in next trie */ + if (iter->trie == trie_local) + return fib_trie_get_first(iter, trie_main); -/* - * This outputs /proc/net/fib_triestats - * - * It always works in backward compatibility mode. - * The format of the file is not supposed to be changed. - */ + return NULL; +} -static void collect_and_show(struct trie *t, struct seq_file *seq) +static void fib_trie_seq_stop(struct seq_file *seq, void *v) { - int bytes = 0; /* How many bytes are used, a ref is 4 bytes */ - int i, max, pointers; - struct trie_stat *stat; - int avdepth; - - stat = trie_collect_stats(t); - - bytes = 0; - seq_printf(seq, "trie=%p\n", t); - - if (stat) { - if (stat->leaves) - avdepth = stat->totdepth*100 / stat->leaves; - else - avdepth = 0; - seq_printf(seq, "Aver depth: %d.%02d\n", avdepth / 100, avdepth % 100); - seq_printf(seq, "Max depth: %4d\n", stat->maxdepth); + rcu_read_unlock(); +} - seq_printf(seq, "Leaves: %d\n", stat->leaves); - bytes += sizeof(struct leaf) * stat->leaves; - seq_printf(seq, "Internal nodes: %d\n", stat->tnodes); - bytes += sizeof(struct tnode) * stat->tnodes; +static void seq_indent(struct seq_file *seq, int n) +{ + while (n-- > 0) seq_puts(seq, " "); +} - max = MAX_CHILDS-1; +static inline const char *rtn_scope(enum rt_scope_t s) +{ + static char buf[32]; - while (max >= 0 && stat->nodesizes[max] == 0) - max--; - pointers = 0; + switch(s) { + case RT_SCOPE_UNIVERSE: return "universe"; + case RT_SCOPE_SITE: return "site"; + case RT_SCOPE_LINK: return "link"; + case RT_SCOPE_HOST: return "host"; + case RT_SCOPE_NOWHERE: return "nowhere"; + default: + snprintf(buf, sizeof(buf), "scope=%d", s); + return buf; + } +} - for (i = 1; i <= max; i++) - if (stat->nodesizes[i] != 0) { - seq_printf(seq, " %d: %d", i, stat->nodesizes[i]); - pointers += (1<<i) * stat->nodesizes[i]; - } - seq_printf(seq, "\n"); - seq_printf(seq, "Pointers: %d\n", pointers); - bytes += sizeof(struct node *) * pointers; - seq_printf(seq, "Null ptrs: %d\n", stat->nullpointers); - seq_printf(seq, "Total size: %d kB\n", bytes / 1024); +static const char *rtn_type_names[__RTN_MAX] = { + [RTN_UNSPEC] = "UNSPEC", + [RTN_UNICAST] = "UNICAST", + [RTN_LOCAL] = "LOCAL", + [RTN_BROADCAST] = "BROADCAST", + [RTN_ANYCAST] = "ANYCAST", + [RTN_MULTICAST] = "MULTICAST", + [RTN_BLACKHOLE] = "BLACKHOLE", + [RTN_UNREACHABLE] = "UNREACHABLE", + [RTN_PROHIBIT] = "PROHIBIT", + [RTN_THROW] = "THROW", + [RTN_NAT] = "NAT", + [RTN_XRESOLVE] = "XRESOLVE", +}; - kfree(stat); - } +static inline const char *rtn_type(unsigned t) +{ + static char buf[32]; -#ifdef CONFIG_IP_FIB_TRIE_STATS - seq_printf(seq, "Counters:\n---------\n"); - seq_printf(seq,"gets = %d\n", t->stats.gets); - seq_printf(seq,"backtracks = %d\n", t->stats.backtrack); - seq_printf(seq,"semantic match passed = %d\n", t->stats.semantic_match_passed); - seq_printf(seq,"semantic match miss = %d\n", t->stats.semantic_match_miss); - seq_printf(seq,"null node hit= %d\n", t->stats.null_node_hit); - seq_printf(seq,"skipped node resize = %d\n", t->stats.resize_node_skipped); -#ifdef CLEAR_STATS - memset(&(t->stats), 0, sizeof(t->stats)); -#endif -#endif /* CONFIG_IP_FIB_TRIE_STATS */ + if (t < __RTN_MAX && rtn_type_names[t]) + return rtn_type_names[t]; + snprintf(buf, sizeof(buf), "type %d", t); + return buf; } -static int fib_triestat_seq_show(struct seq_file *seq, void *v) +/* Pretty print the trie */ +static int fib_trie_seq_show(struct seq_file *seq, void *v) { - char bf[128]; + const struct fib_trie_iter *iter = seq->private; + struct node *n = v; - if (v == SEQ_START_TOKEN) { - seq_printf(seq, "Basic info: size of leaf: %Zd bytes, size of tnode: %Zd bytes.\n", - sizeof(struct leaf), sizeof(struct tnode)); - if (trie_local) - collect_and_show(trie_local, seq); + if (v == SEQ_START_TOKEN) + return 0; - if (trie_main) - collect_and_show(trie_main, seq); - } else { - snprintf(bf, sizeof(bf), "*\t%08X\t%08X", 200, 400); + if (IS_TNODE(n)) { + struct tnode *tn = (struct tnode *) n; + t_key prf = ntohl(MASK_PFX(tn->key, tn->pos)); - seq_printf(seq, "%-127s\n", bf); + if (!NODE_PARENT(n)) { + if (iter->trie == trie_local) + seq_puts(seq, "<local>:\n"); + else + seq_puts(seq, "<main>:\n"); + } + seq_indent(seq, iter->depth-1); + seq_printf(seq, " +-- %d.%d.%d.%d/%d %d %d %d\n", + NIPQUAD(prf), tn->pos, tn->bits, tn->full_children, + tn->empty_children); + + } else { + struct leaf *l = (struct leaf *) n; + int i; + u32 val = ntohl(l->key); + + seq_indent(seq, iter->depth); + seq_printf(seq, " |-- %d.%d.%d.%d\n", NIPQUAD(val)); + for (i = 32; i >= 0; i--) { + struct leaf_info *li = find_leaf_info(l, i); + if (li) { + struct fib_alias *fa; + list_for_each_entry_rcu(fa, &li->falh, fa_list) { + seq_indent(seq, iter->depth+1); + seq_printf(seq, " /%d %s %s", i, + rtn_scope(fa->fa_scope), + rtn_type(fa->fa_type)); + if (fa->fa_tos) + seq_printf(seq, "tos =%d\n", + fa->fa_tos); + seq_putc(seq, '\n'); + } + } + } } + return 0; } -static struct seq_operations fib_triestat_seq_ops = { - .start = fib_triestat_seq_start, - .next = fib_triestat_seq_next, - .stop = fib_triestat_seq_stop, - .show = fib_triestat_seq_show, +static struct seq_operations fib_trie_seq_ops = { + .start = fib_trie_seq_start, + .next = fib_trie_seq_next, + .stop = fib_trie_seq_stop, + .show = fib_trie_seq_show, }; -static int fib_triestat_seq_open(struct inode *inode, struct file *file) +static int fib_trie_seq_open(struct inode *inode, struct file *file) { struct seq_file *seq; int rc = -ENOMEM; + struct fib_trie_iter *s = kmalloc(sizeof(*s), GFP_KERNEL); + + if (!s) + goto out; - rc = seq_open(file, &fib_triestat_seq_ops); + rc = seq_open(file, &fib_trie_seq_ops); if (rc) goto out_kfree; - seq = file->private_data; + seq = file->private_data; + seq->private = s; + memset(s, 0, sizeof(*s)); out: return rc; out_kfree: + kfree(s); goto out; } -static struct file_operations fib_triestat_seq_fops = { - .owner = THIS_MODULE, - .open = fib_triestat_seq_open, - .read = seq_read, - .llseek = seq_lseek, +static struct file_operations fib_trie_fops = { + .owner = THIS_MODULE, + .open = fib_trie_seq_open, + .read = seq_read, + .llseek = seq_lseek, .release = seq_release_private, }; -int __init fib_stat_proc_init(void) +static unsigned fib_flag_trans(int type, u32 mask, const struct fib_info *fi) { - if (!proc_net_fops_create("fib_triestat", S_IRUGO, &fib_triestat_seq_fops)) - return -ENOMEM; - return 0; -} + static unsigned type2flags[RTN_MAX + 1] = { + [7] = RTF_REJECT, [8] = RTF_REJECT, + }; + unsigned flags = type2flags[type]; -void __init fib_stat_proc_exit(void) -{ - proc_net_remove("fib_triestat"); + if (fi && fi->fib_nh->nh_gw) + flags |= RTF_GATEWAY; + if (mask == 0xFFFFFFFF) + flags |= RTF_HOST; + flags |= RTF_UP; + return flags; } -static struct fib_alias *fib_trie_get_first(struct seq_file *seq) +/* + * This outputs /proc/net/route. + * The format of the file is not supposed to be changed + * and needs to be same as fib_hash output to avoid breaking + * legacy utilities + */ +static int fib_route_seq_show(struct seq_file *seq, void *v) { - return NULL; -} + struct leaf *l = v; + int i; + char bf[128]; -static struct fib_alias *fib_trie_get_next(struct seq_file *seq) -{ - return NULL; -} + if (v == SEQ_START_TOKEN) { + seq_printf(seq, "%-127s\n", "Iface\tDestination\tGateway " + "\tFlags\tRefCnt\tUse\tMetric\tMask\t\tMTU" + "\tWindow\tIRTT"); + return 0; + } -static void *fib_trie_seq_start(struct seq_file *seq, loff_t *pos) -{ - if (!ip_fib_main_table) - return NULL; + if (IS_TNODE(l)) + return 0; - if (*pos) - return fib_trie_get_next(seq); - else - return SEQ_START_TOKEN; -} + for (i=32; i>=0; i--) { + struct leaf_info *li = find_leaf_info(l, i); + struct fib_alias *fa; + u32 mask, prefix; -static void *fib_trie_seq_next(struct seq_file *seq, void *v, loff_t *pos) -{ - ++*pos; - if (v == SEQ_START_TOKEN) - return fib_trie_get_first(seq); - else - return fib_trie_get_next(seq); + if (!li) + continue; -} + mask = inet_make_mask(li->plen); + prefix = htonl(l->key); -static void fib_trie_seq_stop(struct seq_file *seq, void *v) -{ -} - -/* - * This outputs /proc/net/fib_trie. - * - * It always works in backward compatibility mode. - * The format of the file is not supposed to be changed. - */ + list_for_each_entry_rcu(fa, &li->falh, fa_list) { + const struct fib_info *fi = rcu_dereference(fa->fa_info); + unsigned flags = fib_flag_trans(fa->fa_type, mask, fi); -static int fib_trie_seq_show(struct seq_file *seq, void *v) -{ - char bf[128]; + if (fa->fa_type == RTN_BROADCAST + || fa->fa_type == RTN_MULTICAST) + continue; - if (v == SEQ_START_TOKEN) { - if (trie_local) - trie_dump_seq(seq, trie_local); + if (fi) + snprintf(bf, sizeof(bf), + "%s\t%08X\t%08X\t%04X\t%d\t%u\t%d\t%08X\t%d\t%u\t%u", + fi->fib_dev ? fi->fib_dev->name : "*", + prefix, + fi->fib_nh->nh_gw, flags, 0, 0, + fi->fib_priority, + mask, + (fi->fib_advmss ? fi->fib_advmss + 40 : 0), + fi->fib_window, + fi->fib_rtt >> 3); + else + snprintf(bf, sizeof(bf), + "*\t%08X\t%08X\t%04X\t%d\t%u\t%d\t%08X\t%d\t%u\t%u", + prefix, 0, flags, 0, 0, 0, + mask, 0, 0, 0); - if (trie_main) - trie_dump_seq(seq, trie_main); - } else { - snprintf(bf, sizeof(bf), - "*\t%08X\t%08X", 200, 400); - seq_printf(seq, "%-127s\n", bf); + seq_printf(seq, "%-127s\n", bf); + } } return 0; } -static struct seq_operations fib_trie_seq_ops = { - .start = fib_trie_seq_start, - .next = fib_trie_seq_next, - .stop = fib_trie_seq_stop, - .show = fib_trie_seq_show, +static struct seq_operations fib_route_seq_ops = { + .start = fib_trie_seq_start, + .next = fib_trie_seq_next, + .stop = fib_trie_seq_stop, + .show = fib_route_seq_show, }; -static int fib_trie_seq_open(struct inode *inode, struct file *file) +static int fib_route_seq_open(struct inode *inode, struct file *file) { struct seq_file *seq; int rc = -ENOMEM; + struct fib_trie_iter *s = kmalloc(sizeof(*s), GFP_KERNEL); - rc = seq_open(file, &fib_trie_seq_ops); + if (!s) + goto out; + + rc = seq_open(file, &fib_route_seq_ops); if (rc) goto out_kfree; - seq = file->private_data; + seq = file->private_data; + seq->private = s; + memset(s, 0, sizeof(*s)); out: return rc; out_kfree: + kfree(s); goto out; } -static struct file_operations fib_trie_seq_fops = { - .owner = THIS_MODULE, - .open = fib_trie_seq_open, - .read = seq_read, - .llseek = seq_lseek, - .release= seq_release_private, +static struct file_operations fib_route_fops = { + .owner = THIS_MODULE, + .open = fib_route_seq_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, }; int __init fib_proc_init(void) { - if (!proc_net_fops_create("fib_trie", S_IRUGO, &fib_trie_seq_fops)) - return -ENOMEM; + if (!proc_net_fops_create("fib_trie", S_IRUGO, &fib_trie_fops)) + goto out1; + + if (!proc_net_fops_create("fib_triestat", S_IRUGO, &fib_triestat_fops)) + goto out2; + + if (!proc_net_fops_create("route", S_IRUGO, &fib_route_fops)) + goto out3; + return 0; + +out3: + proc_net_remove("fib_triestat"); +out2: + proc_net_remove("fib_trie"); +out1: + return -ENOMEM; } void __init fib_proc_exit(void) { proc_net_remove("fib_trie"); + proc_net_remove("fib_triestat"); + proc_net_remove("route"); } #endif /* CONFIG_PROC_FS */ diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index 44607f4767b..70c44e4c3ce 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -1603,7 +1603,7 @@ static void ip_mc_clear_src(struct ip_mc_list *pmc) } pmc->sources = NULL; pmc->sfmode = MCAST_EXCLUDE; - pmc->sfcount[MCAST_EXCLUDE] = 0; + pmc->sfcount[MCAST_INCLUDE] = 0; pmc->sfcount[MCAST_EXCLUDE] = 1; } diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index f84ba9c9655..2fc3fd38924 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -100,8 +100,7 @@ DEFINE_SPINLOCK(inet_peer_unused_lock); #define PEER_MAX_CLEANUP_WORK 30 static void peer_check_expire(unsigned long dummy); -static struct timer_list peer_periodic_timer = - TIMER_INITIALIZER(peer_check_expire, 0, 0); +static DEFINE_TIMER(peer_periodic_timer, peer_check_expire, 0, 0); /* Exported for sysctl_net_ipv4. */ int inet_peer_gc_mintime = 10 * HZ, diff --git a/net/ipv4/ip_fragment.c b/net/ipv4/ip_fragment.c index 9e6e683cc34..e7d26d9943c 100644 --- a/net/ipv4/ip_fragment.c +++ b/net/ipv4/ip_fragment.c @@ -457,7 +457,7 @@ static void ip_frag_queue(struct ipq *qp, struct sk_buff *skb) if (pskb_pull(skb, ihl) == NULL) goto err; - if (pskb_trim(skb, end-offset)) + if (pskb_trim_rcsum(skb, end-offset)) goto err; /* Find out which fragments are in front and at the back of us diff --git a/net/ipv4/ipcomp.c b/net/ipv4/ipcomp.c index dcb7ee6c485..fc718df17b4 100644 --- a/net/ipv4/ipcomp.c +++ b/net/ipv4/ipcomp.c @@ -345,8 +345,7 @@ static void ipcomp_free_tfms(struct crypto_tfm **tfms) for_each_cpu(cpu) { struct crypto_tfm *tfm = *per_cpu_ptr(tfms, cpu); - if (tfm) - crypto_free_tfm(tfm); + crypto_free_tfm(tfm); } free_percpu(tfms); } diff --git a/net/ipv4/ipconfig.c b/net/ipv4/ipconfig.c index 63e106605f2..e8674baaa8d 100644 --- a/net/ipv4/ipconfig.c +++ b/net/ipv4/ipconfig.c @@ -54,6 +54,7 @@ #include <linux/major.h> #include <linux/root_dev.h> #include <linux/delay.h> +#include <linux/nfs_fs.h> #include <net/arp.h> #include <net/ip.h> #include <net/ipconfig.h> @@ -1102,10 +1103,8 @@ static int __init ic_dynamic(void) #endif jiff = jiffies + (d->next ? CONF_INTER_TIMEOUT : timeout); - while (time_before(jiffies, jiff) && !ic_got_reply) { - set_current_state(TASK_UNINTERRUPTIBLE); - schedule_timeout(1); - } + while (time_before(jiffies, jiff) && !ic_got_reply) + schedule_timeout_uninterruptible(1); #ifdef IPCONFIG_DHCP /* DHCP isn't done until we get a DHCPACK. */ if ((ic_got_reply & IC_BOOTP) diff --git a/net/ipv4/ipvs/ip_vs_conn.c b/net/ipv4/ipvs/ip_vs_conn.c index e11952ea17a..f828fa2eb7d 100644 --- a/net/ipv4/ipvs/ip_vs_conn.c +++ b/net/ipv4/ipvs/ip_vs_conn.c @@ -196,6 +196,7 @@ static inline struct ip_vs_conn *__ip_vs_conn_in_get list_for_each_entry(cp, &ip_vs_conn_tab[hash], c_list) { if (s_addr==cp->caddr && s_port==cp->cport && d_port==cp->vport && d_addr==cp->vaddr && + ((!s_port) ^ (!(cp->flags & IP_VS_CONN_F_NO_CPORT))) && protocol==cp->protocol) { /* HIT */ atomic_inc(&cp->refcnt); @@ -227,6 +228,40 @@ struct ip_vs_conn *ip_vs_conn_in_get return cp; } +/* Get reference to connection template */ +struct ip_vs_conn *ip_vs_ct_in_get +(int protocol, __u32 s_addr, __u16 s_port, __u32 d_addr, __u16 d_port) +{ + unsigned hash; + struct ip_vs_conn *cp; + + hash = ip_vs_conn_hashkey(protocol, s_addr, s_port); + + ct_read_lock(hash); + + list_for_each_entry(cp, &ip_vs_conn_tab[hash], c_list) { + if (s_addr==cp->caddr && s_port==cp->cport && + d_port==cp->vport && d_addr==cp->vaddr && + cp->flags & IP_VS_CONN_F_TEMPLATE && + protocol==cp->protocol) { + /* HIT */ + atomic_inc(&cp->refcnt); + goto out; + } + } + cp = NULL; + + out: + ct_read_unlock(hash); + + IP_VS_DBG(7, "template lookup/in %s %u.%u.%u.%u:%d->%u.%u.%u.%u:%d %s\n", + ip_vs_proto_name(protocol), + NIPQUAD(s_addr), ntohs(s_port), + NIPQUAD(d_addr), ntohs(d_port), + cp?"hit":"not hit"); + + return cp; +} /* * Gets ip_vs_conn associated with supplied parameters in the ip_vs_conn_tab. @@ -367,7 +402,7 @@ ip_vs_bind_dest(struct ip_vs_conn *cp, struct ip_vs_dest *dest) atomic_read(&dest->refcnt)); /* Update the connection counters */ - if (cp->cport || (cp->flags & IP_VS_CONN_F_NO_CPORT)) { + if (!(cp->flags & IP_VS_CONN_F_TEMPLATE)) { /* It is a normal connection, so increase the inactive connection counter because it is in TCP SYNRECV state (inactive) or other protocol inacive state */ @@ -406,7 +441,7 @@ static inline void ip_vs_unbind_dest(struct ip_vs_conn *cp) atomic_read(&dest->refcnt)); /* Update the connection counters */ - if (cp->cport || (cp->flags & IP_VS_CONN_F_NO_CPORT)) { + if (!(cp->flags & IP_VS_CONN_F_TEMPLATE)) { /* It is a normal connection, so decrease the inactconns or activeconns counter */ if (cp->flags & IP_VS_CONN_F_INACTIVE) { @@ -467,7 +502,7 @@ int ip_vs_check_template(struct ip_vs_conn *ct) /* * Invalidate the connection template */ - if (ct->cport) { + if (ct->vport != 65535) { if (ip_vs_conn_unhash(ct)) { ct->dport = 65535; ct->vport = 65535; @@ -776,7 +811,7 @@ void ip_vs_random_dropentry(void) ct_write_lock_bh(hash); list_for_each_entry(cp, &ip_vs_conn_tab[hash], c_list) { - if (!cp->cport && !(cp->flags & IP_VS_CONN_F_NO_CPORT)) + if (cp->flags & IP_VS_CONN_F_TEMPLATE) /* connection template */ continue; diff --git a/net/ipv4/ipvs/ip_vs_core.c b/net/ipv4/ipvs/ip_vs_core.c index 3ac7eeca04a..981cc3244ef 100644 --- a/net/ipv4/ipvs/ip_vs_core.c +++ b/net/ipv4/ipvs/ip_vs_core.c @@ -243,10 +243,10 @@ ip_vs_sched_persist(struct ip_vs_service *svc, if (ports[1] == svc->port) { /* Check if a template already exists */ if (svc->port != FTPPORT) - ct = ip_vs_conn_in_get(iph->protocol, snet, 0, + ct = ip_vs_ct_in_get(iph->protocol, snet, 0, iph->daddr, ports[1]); else - ct = ip_vs_conn_in_get(iph->protocol, snet, 0, + ct = ip_vs_ct_in_get(iph->protocol, snet, 0, iph->daddr, 0); if (!ct || !ip_vs_check_template(ct)) { @@ -272,14 +272,14 @@ ip_vs_sched_persist(struct ip_vs_service *svc, iph->daddr, ports[1], dest->addr, dest->port, - 0, + IP_VS_CONN_F_TEMPLATE, dest); else ct = ip_vs_conn_new(iph->protocol, snet, 0, iph->daddr, 0, dest->addr, 0, - 0, + IP_VS_CONN_F_TEMPLATE, dest); if (ct == NULL) return NULL; @@ -298,10 +298,10 @@ ip_vs_sched_persist(struct ip_vs_service *svc, * port zero template: <protocol,caddr,0,vaddr,0,daddr,0> */ if (svc->fwmark) - ct = ip_vs_conn_in_get(IPPROTO_IP, snet, 0, + ct = ip_vs_ct_in_get(IPPROTO_IP, snet, 0, htonl(svc->fwmark), 0); else - ct = ip_vs_conn_in_get(iph->protocol, snet, 0, + ct = ip_vs_ct_in_get(iph->protocol, snet, 0, iph->daddr, 0); if (!ct || !ip_vs_check_template(ct)) { @@ -326,14 +326,14 @@ ip_vs_sched_persist(struct ip_vs_service *svc, snet, 0, htonl(svc->fwmark), 0, dest->addr, 0, - 0, + IP_VS_CONN_F_TEMPLATE, dest); else ct = ip_vs_conn_new(iph->protocol, snet, 0, iph->daddr, 0, dest->addr, 0, - 0, + IP_VS_CONN_F_TEMPLATE, dest); if (ct == NULL) return NULL; diff --git a/net/ipv4/ipvs/ip_vs_sync.c b/net/ipv4/ipvs/ip_vs_sync.c index 574d1f509b4..2e5ced3d806 100644 --- a/net/ipv4/ipvs/ip_vs_sync.c +++ b/net/ipv4/ipvs/ip_vs_sync.c @@ -297,16 +297,24 @@ static void ip_vs_process_message(const char *buffer, const size_t buflen) p = (char *)buffer + sizeof(struct ip_vs_sync_mesg); for (i=0; i<m->nr_conns; i++) { + unsigned flags; + s = (struct ip_vs_sync_conn *)p; - cp = ip_vs_conn_in_get(s->protocol, - s->caddr, s->cport, - s->vaddr, s->vport); + flags = ntohs(s->flags); + if (!(flags & IP_VS_CONN_F_TEMPLATE)) + cp = ip_vs_conn_in_get(s->protocol, + s->caddr, s->cport, + s->vaddr, s->vport); + else + cp = ip_vs_ct_in_get(s->protocol, + s->caddr, s->cport, + s->vaddr, s->vport); if (!cp) { cp = ip_vs_conn_new(s->protocol, s->caddr, s->cport, s->vaddr, s->vport, s->daddr, s->dport, - ntohs(s->flags), NULL); + flags, NULL); if (!cp) { IP_VS_ERR("ip_vs_conn_new failed\n"); return; @@ -315,11 +323,11 @@ static void ip_vs_process_message(const char *buffer, const size_t buflen) } else if (!cp->dest) { /* it is an entry created by the synchronization */ cp->state = ntohs(s->state); - cp->flags = ntohs(s->flags) | IP_VS_CONN_F_HASHED; + cp->flags = flags | IP_VS_CONN_F_HASHED; } /* Note that we don't touch its state and flags if it is a normal entry. */ - if (ntohs(s->flags) & IP_VS_CONN_F_SEQ_MASK) { + if (flags & IP_VS_CONN_F_SEQ_MASK) { opt = (struct ip_vs_sync_conn_options *)&s[1]; memcpy(&cp->in_seq, opt, sizeof(*opt)); p += FULL_CONN_SIZE; diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig index e046f552181..3cf9b451675 100644 --- a/net/ipv4/netfilter/Kconfig +++ b/net/ipv4/netfilter/Kconfig @@ -34,6 +34,7 @@ config IP_NF_CT_ACCT config IP_NF_CONNTRACK_MARK bool 'Connection mark tracking support' + depends on IP_NF_CONNTRACK help This option enables support for connection marks, used by the `CONNMARK' target and `connmark' match. Similar to the mark value @@ -50,6 +51,14 @@ config IP_NF_CONNTRACK_EVENTS IF unsure, say `N'. +config IP_NF_CONNTRACK_NETLINK + tristate 'Connection tracking netlink interface' + depends on IP_NF_CONNTRACK && NETFILTER_NETLINK + depends on IP_NF_CONNTRACK!=y || NETFILTER_NETLINK!=m + help + This option enables support for a netlink-based userspace interface + + config IP_NF_CT_PROTO_SCTP tristate 'SCTP protocol connection tracking support (EXPERIMENTAL)' depends on IP_NF_CONNTRACK && EXPERIMENTAL @@ -85,6 +94,25 @@ config IP_NF_IRC To compile it as a module, choose M here. If unsure, say Y. +config IP_NF_NETBIOS_NS + tristate "NetBIOS name service protocol support (EXPERIMENTAL)" + depends on IP_NF_CONNTRACK && EXPERIMENTAL + help + NetBIOS name service requests are sent as broadcast messages from an + unprivileged port and responded to with unicast messages to the + same port. This make them hard to firewall properly because connection + tracking doesn't deal with broadcasts. This helper tracks locally + originating NetBIOS name service requests and the corresponding + responses. It relies on correct IP address configuration, specifically + netmask and broadcast address. When properly configured, the output + of "ip address show" should look similar to this: + + $ ip -4 address show eth0 + 4: eth0: <BROADCAST,MULTICAST,UP> mtu 1500 qdisc pfifo_fast qlen 1000 + inet 172.16.2.252/24 brd 172.16.2.255 scope global eth0 + + To compile it as a module, choose M here. If unsure, say N. + config IP_NF_TFTP tristate "TFTP protocol support" depends on IP_NF_CONNTRACK @@ -109,6 +137,22 @@ config IP_NF_AMANDA To compile it as a module, choose M here. If unsure, say Y. +config IP_NF_PPTP + tristate 'PPTP protocol support' + help + This module adds support for PPTP (Point to Point Tunnelling + Protocol, RFC2637) conncection tracking and NAT. + + If you are running PPTP sessions over a stateful firewall or NAT + box, you may want to enable this feature. + + Please note that not all PPTP modes of operation are supported yet. + For more info, read top of the file + net/ipv4/netfilter/ip_conntrack_pptp.c + + If you want to compile it as a module, say M here and read + Documentation/modules.txt. If unsure, say `N'. + config IP_NF_QUEUE tristate "IP Userspace queueing via NETLINK (OBSOLETE)" help @@ -593,6 +637,12 @@ config IP_NF_NAT_AMANDA default IP_NF_NAT if IP_NF_AMANDA=y default m if IP_NF_AMANDA=m +config IP_NF_NAT_PPTP + tristate + depends on IP_NF_NAT!=n && IP_NF_PPTP!=n + default IP_NF_NAT if IP_NF_PPTP=y + default m if IP_NF_PPTP=m + # mangle + specific targets config IP_NF_MANGLE tristate "Packet mangling" @@ -754,11 +804,5 @@ config IP_NF_ARP_MANGLE Allows altering the ARP packet payload: source and destination hardware and network addresses. -config IP_NF_CONNTRACK_NETLINK - tristate 'Connection tracking netlink interface' - depends on IP_NF_CONNTRACK && NETFILTER_NETLINK - help - This option enables support for a netlink-based userspace interface - endmenu diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile index a7bd38f5052..3d45d3c0283 100644 --- a/net/ipv4/netfilter/Makefile +++ b/net/ipv4/netfilter/Makefile @@ -6,6 +6,9 @@ ip_conntrack-objs := ip_conntrack_standalone.o ip_conntrack_core.o ip_conntrack_proto_generic.o ip_conntrack_proto_tcp.o ip_conntrack_proto_udp.o ip_conntrack_proto_icmp.o iptable_nat-objs := ip_nat_standalone.o ip_nat_rule.o ip_nat_core.o ip_nat_helper.o ip_nat_proto_unknown.o ip_nat_proto_tcp.o ip_nat_proto_udp.o ip_nat_proto_icmp.o +ip_conntrack_pptp-objs := ip_conntrack_helper_pptp.o ip_conntrack_proto_gre.o +ip_nat_pptp-objs := ip_nat_helper_pptp.o ip_nat_proto_gre.o + # connection tracking obj-$(CONFIG_IP_NF_CONNTRACK) += ip_conntrack.o @@ -17,12 +20,15 @@ obj-$(CONFIG_IP_NF_CONNTRACK_NETLINK) += ip_conntrack_netlink.o obj-$(CONFIG_IP_NF_CT_PROTO_SCTP) += ip_conntrack_proto_sctp.o # connection tracking helpers +obj-$(CONFIG_IP_NF_PPTP) += ip_conntrack_pptp.o obj-$(CONFIG_IP_NF_AMANDA) += ip_conntrack_amanda.o obj-$(CONFIG_IP_NF_TFTP) += ip_conntrack_tftp.o obj-$(CONFIG_IP_NF_FTP) += ip_conntrack_ftp.o obj-$(CONFIG_IP_NF_IRC) += ip_conntrack_irc.o +obj-$(CONFIG_IP_NF_NETBIOS_NS) += ip_conntrack_netbios_ns.o # NAT helpers +obj-$(CONFIG_IP_NF_NAT_PPTP) += ip_nat_pptp.o obj-$(CONFIG_IP_NF_NAT_AMANDA) += ip_nat_amanda.o obj-$(CONFIG_IP_NF_NAT_TFTP) += ip_nat_tftp.o obj-$(CONFIG_IP_NF_NAT_FTP) += ip_nat_ftp.o diff --git a/net/ipv4/netfilter/ip_conntrack_amanda.c b/net/ipv4/netfilter/ip_conntrack_amanda.c index be4c9eb3243..fa3f914117e 100644 --- a/net/ipv4/netfilter/ip_conntrack_amanda.c +++ b/net/ipv4/netfilter/ip_conntrack_amanda.c @@ -65,7 +65,7 @@ static int help(struct sk_buff **pskb, /* increase the UDP timeout of the master connection as replies from * Amanda clients to the server can be quite delayed */ - ip_ct_refresh_acct(ct, ctinfo, NULL, master_timeout * HZ); + ip_ct_refresh(ct, *pskb, master_timeout * HZ); /* No data? */ dataoff = (*pskb)->nh.iph->ihl*4 + sizeof(struct udphdr); @@ -108,6 +108,7 @@ static int help(struct sk_buff **pskb, } exp->expectfn = NULL; + exp->flags = 0; exp->tuple.src.ip = ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip; exp->tuple.src.u.tcp.port = 0; diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c index a0648600190..ea65dd3e517 100644 --- a/net/ipv4/netfilter/ip_conntrack_core.c +++ b/net/ipv4/netfilter/ip_conntrack_core.c @@ -197,7 +197,7 @@ ip_ct_invert_tuple(struct ip_conntrack_tuple *inverse, /* ip_conntrack_expect helper functions */ -static void unlink_expect(struct ip_conntrack_expect *exp) +void ip_ct_unlink_expect(struct ip_conntrack_expect *exp) { ASSERT_WRITE_LOCK(&ip_conntrack_lock); IP_NF_ASSERT(!timer_pending(&exp->timeout)); @@ -207,18 +207,12 @@ static void unlink_expect(struct ip_conntrack_expect *exp) ip_conntrack_expect_put(exp); } -void __ip_ct_expect_unlink_destroy(struct ip_conntrack_expect *exp) -{ - unlink_expect(exp); - ip_conntrack_expect_put(exp); -} - static void expectation_timed_out(unsigned long ul_expect) { struct ip_conntrack_expect *exp = (void *)ul_expect; write_lock_bh(&ip_conntrack_lock); - unlink_expect(exp); + ip_ct_unlink_expect(exp); write_unlock_bh(&ip_conntrack_lock); ip_conntrack_expect_put(exp); } @@ -239,7 +233,7 @@ __ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple) /* Just find a expectation corresponding to a tuple. */ struct ip_conntrack_expect * -ip_conntrack_expect_find_get(const struct ip_conntrack_tuple *tuple) +ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple) { struct ip_conntrack_expect *i; @@ -264,10 +258,14 @@ find_expectation(const struct ip_conntrack_tuple *tuple) master ct never got confirmed, we'd hold a reference to it and weird things would happen to future packets). */ if (ip_ct_tuple_mask_cmp(tuple, &i->tuple, &i->mask) - && is_confirmed(i->master) - && del_timer(&i->timeout)) { - unlink_expect(i); - return i; + && is_confirmed(i->master)) { + if (i->flags & IP_CT_EXPECT_PERMANENT) { + atomic_inc(&i->use); + return i; + } else if (del_timer(&i->timeout)) { + ip_ct_unlink_expect(i); + return i; + } } } return NULL; @@ -284,7 +282,7 @@ void ip_ct_remove_expectations(struct ip_conntrack *ct) list_for_each_entry_safe(i, tmp, &ip_conntrack_expect_list, list) { if (i->master == ct && del_timer(&i->timeout)) { - unlink_expect(i); + ip_ct_unlink_expect(i); ip_conntrack_expect_put(i); } } @@ -925,7 +923,7 @@ void ip_conntrack_unexpect_related(struct ip_conntrack_expect *exp) /* choose the the oldest expectation to evict */ list_for_each_entry_reverse(i, &ip_conntrack_expect_list, list) { if (expect_matches(i, exp) && del_timer(&i->timeout)) { - unlink_expect(i); + ip_ct_unlink_expect(i); write_unlock_bh(&ip_conntrack_lock); ip_conntrack_expect_put(i); return; @@ -934,6 +932,9 @@ void ip_conntrack_unexpect_related(struct ip_conntrack_expect *exp) write_unlock_bh(&ip_conntrack_lock); } +/* We don't increase the master conntrack refcount for non-fulfilled + * conntracks. During the conntrack destruction, the expectations are + * always killed before the conntrack itself */ struct ip_conntrack_expect *ip_conntrack_expect_alloc(struct ip_conntrack *me) { struct ip_conntrack_expect *new; @@ -944,17 +945,14 @@ struct ip_conntrack_expect *ip_conntrack_expect_alloc(struct ip_conntrack *me) return NULL; } new->master = me; - atomic_inc(&new->master->ct_general.use); atomic_set(&new->use, 1); return new; } void ip_conntrack_expect_put(struct ip_conntrack_expect *exp) { - if (atomic_dec_and_test(&exp->use)) { - ip_conntrack_put(exp->master); + if (atomic_dec_and_test(&exp->use)) kmem_cache_free(ip_conntrack_expect_cachep, exp); - } } static void ip_conntrack_expect_insert(struct ip_conntrack_expect *exp) @@ -982,7 +980,7 @@ static void evict_oldest_expect(struct ip_conntrack *master) list_for_each_entry_reverse(i, &ip_conntrack_expect_list, list) { if (i->master == master) { if (del_timer(&i->timeout)) { - unlink_expect(i); + ip_ct_unlink_expect(i); ip_conntrack_expect_put(i); } break; @@ -1099,7 +1097,7 @@ void ip_conntrack_helper_unregister(struct ip_conntrack_helper *me) /* Get rid of expectations */ list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, list) { if (exp->master->helper == me && del_timer(&exp->timeout)) { - unlink_expect(exp); + ip_ct_unlink_expect(exp); ip_conntrack_expect_put(exp); } } @@ -1114,42 +1112,46 @@ void ip_conntrack_helper_unregister(struct ip_conntrack_helper *me) synchronize_net(); } -static inline void ct_add_counters(struct ip_conntrack *ct, - enum ip_conntrack_info ctinfo, - const struct sk_buff *skb) -{ -#ifdef CONFIG_IP_NF_CT_ACCT - if (skb) { - ct->counters[CTINFO2DIR(ctinfo)].packets++; - ct->counters[CTINFO2DIR(ctinfo)].bytes += - ntohs(skb->nh.iph->tot_len); - } -#endif -} - -/* Refresh conntrack for this many jiffies and do accounting (if skb != NULL) */ -void ip_ct_refresh_acct(struct ip_conntrack *ct, +/* Refresh conntrack for this many jiffies and do accounting if do_acct is 1 */ +void __ip_ct_refresh_acct(struct ip_conntrack *ct, enum ip_conntrack_info ctinfo, const struct sk_buff *skb, - unsigned long extra_jiffies) + unsigned long extra_jiffies, + int do_acct) { + int do_event = 0; + IP_NF_ASSERT(ct->timeout.data == (unsigned long)ct); + IP_NF_ASSERT(skb); + + write_lock_bh(&ip_conntrack_lock); /* If not in hash table, timer will not be active yet */ if (!is_confirmed(ct)) { ct->timeout.expires = extra_jiffies; - ct_add_counters(ct, ctinfo, skb); + do_event = 1; } else { - write_lock_bh(&ip_conntrack_lock); /* Need del_timer for race avoidance (may already be dying). */ if (del_timer(&ct->timeout)) { ct->timeout.expires = jiffies + extra_jiffies; add_timer(&ct->timeout); - ip_conntrack_event_cache(IPCT_REFRESH, skb); + do_event = 1; } - ct_add_counters(ct, ctinfo, skb); - write_unlock_bh(&ip_conntrack_lock); } + +#ifdef CONFIG_IP_NF_CT_ACCT + if (do_acct) { + ct->counters[CTINFO2DIR(ctinfo)].packets++; + ct->counters[CTINFO2DIR(ctinfo)].bytes += + ntohs(skb->nh.iph->tot_len); + } +#endif + + write_unlock_bh(&ip_conntrack_lock); + + /* must be unlocked when calling event cache */ + if (do_event) + ip_conntrack_event_cache(IPCT_REFRESH, skb); } #if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ diff --git a/net/ipv4/netfilter/ip_conntrack_ftp.c b/net/ipv4/netfilter/ip_conntrack_ftp.c index 3a2627db172..d77d6b3f5f8 100644 --- a/net/ipv4/netfilter/ip_conntrack_ftp.c +++ b/net/ipv4/netfilter/ip_conntrack_ftp.c @@ -29,9 +29,9 @@ static char *ftp_buffer; static DEFINE_SPINLOCK(ip_ftp_lock); #define MAX_PORTS 8 -static int ports[MAX_PORTS]; +static short ports[MAX_PORTS]; static int ports_c; -module_param_array(ports, int, &ports_c, 0400); +module_param_array(ports, short, &ports_c, 0400); static int loose; module_param(loose, int, 0600); @@ -421,6 +421,7 @@ static int help(struct sk_buff **pskb, { 0xFFFFFFFF, { .tcp = { 0xFFFF } }, 0xFF }}); exp->expectfn = NULL; + exp->flags = 0; /* Now, NAT might want to mangle the packet, and register the * (possibly changed) expectation itself. */ @@ -449,7 +450,7 @@ out_update_nl: } static struct ip_conntrack_helper ftp[MAX_PORTS]; -static char ftp_names[MAX_PORTS][10]; +static char ftp_names[MAX_PORTS][sizeof("ftp-65535")]; /* Not __exit: called from init() */ static void fini(void) diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c new file mode 100644 index 00000000000..926a6684643 --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c @@ -0,0 +1,806 @@ +/* + * ip_conntrack_pptp.c - Version 3.0 + * + * Connection tracking support for PPTP (Point to Point Tunneling Protocol). + * PPTP is a a protocol for creating virtual private networks. + * It is a specification defined by Microsoft and some vendors + * working with Microsoft. PPTP is built on top of a modified + * version of the Internet Generic Routing Encapsulation Protocol. + * GRE is defined in RFC 1701 and RFC 1702. Documentation of + * PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte <laforge@gnumonks.org> + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + * Limitations: + * - We blindly assume that control connections are always + * established in PNS->PAC direction. This is a violation + * of RFFC2673 + * - We can only support one single call within each session + * + * TODO: + * - testing of incoming PPTP calls + * + * Changes: + * 2002-02-05 - Version 1.3 + * - Call ip_conntrack_unexpect_related() from + * pptp_destroy_siblings() to destroy expectations in case + * CALL_DISCONNECT_NOTIFY or tcp fin packet was seen + * (Philip Craig <philipc@snapgear.com>) + * - Add Version information at module loadtime + * 2002-02-10 - Version 1.6 + * - move to C99 style initializers + * - remove second expectation if first arrives + * 2004-10-22 - Version 2.0 + * - merge Mandrake's 2.6.x port with recent 2.6.x API changes + * - fix lots of linear skb assumptions from Mandrake's port + * 2005-06-10 - Version 2.1 + * - use ip_conntrack_expect_free() instead of kfree() on the + * expect's (which are from the slab for quite some time) + * 2005-06-10 - Version 3.0 + * - port helper to post-2.6.11 API changes, + * funded by Oxcoda NetBox Blue (http://www.netboxblue.com/) + * 2005-07-30 - Version 3.1 + * - port helper to 2.6.13 API changes + * + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/netfilter.h> +#include <linux/ip.h> +#include <net/checksum.h> +#include <net/tcp.h> + +#include <linux/netfilter_ipv4/ip_conntrack.h> +#include <linux/netfilter_ipv4/ip_conntrack_core.h> +#include <linux/netfilter_ipv4/ip_conntrack_helper.h> +#include <linux/netfilter_ipv4/ip_conntrack_proto_gre.h> +#include <linux/netfilter_ipv4/ip_conntrack_pptp.h> + +#define IP_CT_PPTP_VERSION "3.1" + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@gnumonks.org>"); +MODULE_DESCRIPTION("Netfilter connection tracking helper module for PPTP"); + +static DEFINE_SPINLOCK(ip_pptp_lock); + +int +(*ip_nat_pptp_hook_outbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +int +(*ip_nat_pptp_hook_inbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +int +(*ip_nat_pptp_hook_exp_gre)(struct ip_conntrack_expect *expect_orig, + struct ip_conntrack_expect *expect_reply); + +void +(*ip_nat_pptp_hook_expectfn)(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp); + +#if 0 +/* PptpControlMessageType names */ +const char *pptp_msg_name[] = { + "UNKNOWN_MESSAGE", + "START_SESSION_REQUEST", + "START_SESSION_REPLY", + "STOP_SESSION_REQUEST", + "STOP_SESSION_REPLY", + "ECHO_REQUEST", + "ECHO_REPLY", + "OUT_CALL_REQUEST", + "OUT_CALL_REPLY", + "IN_CALL_REQUEST", + "IN_CALL_REPLY", + "IN_CALL_CONNECT", + "CALL_CLEAR_REQUEST", + "CALL_DISCONNECT_NOTIFY", + "WAN_ERROR_NOTIFY", + "SET_LINK_INFO" +}; +EXPORT_SYMBOL(pptp_msg_name); +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, __FUNCTION__, ## args) +#else +#define DEBUGP(format, args...) +#endif + +#define SECS *HZ +#define MINS * 60 SECS +#define HOURS * 60 MINS + +#define PPTP_GRE_TIMEOUT (10 MINS) +#define PPTP_GRE_STREAM_TIMEOUT (5 HOURS) + +static void pptp_expectfn(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp) +{ + DEBUGP("increasing timeouts\n"); + + /* increase timeout of GRE data channel conntrack entry */ + ct->proto.gre.timeout = PPTP_GRE_TIMEOUT; + ct->proto.gre.stream_timeout = PPTP_GRE_STREAM_TIMEOUT; + + /* Can you see how rusty this code is, compared with the pre-2.6.11 + * one? That's what happened to my shiny newnat of 2002 ;( -HW */ + + if (!ip_nat_pptp_hook_expectfn) { + struct ip_conntrack_tuple inv_t; + struct ip_conntrack_expect *exp_other; + + /* obviously this tuple inversion only works until you do NAT */ + invert_tuplepr(&inv_t, &exp->tuple); + DEBUGP("trying to unexpect other dir: "); + DUMP_TUPLE(&inv_t); + + exp_other = ip_conntrack_expect_find(&inv_t); + if (exp_other) { + /* delete other expectation. */ + DEBUGP("found\n"); + ip_conntrack_unexpect_related(exp_other); + ip_conntrack_expect_put(exp_other); + } else { + DEBUGP("not found\n"); + } + } else { + /* we need more than simple inversion */ + ip_nat_pptp_hook_expectfn(ct, exp); + } +} + +static int destroy_sibling_or_exp(const struct ip_conntrack_tuple *t) +{ + struct ip_conntrack_tuple_hash *h; + struct ip_conntrack_expect *exp; + + DEBUGP("trying to timeout ct or exp for tuple "); + DUMP_TUPLE(t); + + h = ip_conntrack_find_get(t, NULL); + if (h) { + struct ip_conntrack *sibling = tuplehash_to_ctrack(h); + DEBUGP("setting timeout of conntrack %p to 0\n", sibling); + sibling->proto.gre.timeout = 0; + sibling->proto.gre.stream_timeout = 0; + if (del_timer(&sibling->timeout)) + sibling->timeout.function((unsigned long)sibling); + ip_conntrack_put(sibling); + return 1; + } else { + exp = ip_conntrack_expect_find(t); + if (exp) { + DEBUGP("unexpect_related of expect %p\n", exp); + ip_conntrack_unexpect_related(exp); + ip_conntrack_expect_put(exp); + return 1; + } + } + + return 0; +} + + +/* timeout GRE data connections */ +static void pptp_destroy_siblings(struct ip_conntrack *ct) +{ + struct ip_conntrack_tuple t; + + /* Since ct->sibling_list has literally rusted away in 2.6.11, + * we now need another way to find out about our sibling + * contrack and expects... -HW */ + + /* try original (pns->pac) tuple */ + memcpy(&t, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, sizeof(t)); + t.dst.protonum = IPPROTO_GRE; + t.src.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id); + t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id); + + if (!destroy_sibling_or_exp(&t)) + DEBUGP("failed to timeout original pns->pac ct/exp\n"); + + /* try reply (pac->pns) tuple */ + memcpy(&t, &ct->tuplehash[IP_CT_DIR_REPLY].tuple, sizeof(t)); + t.dst.protonum = IPPROTO_GRE; + t.src.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id); + t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id); + + if (!destroy_sibling_or_exp(&t)) + DEBUGP("failed to timeout reply pac->pns ct/exp\n"); +} + +/* expect GRE connections (PNS->PAC and PAC->PNS direction) */ +static inline int +exp_gre(struct ip_conntrack *master, + u_int32_t seq, + __be16 callid, + __be16 peer_callid) +{ + struct ip_conntrack_tuple inv_tuple; + struct ip_conntrack_tuple exp_tuples[] = { + /* tuple in original direction, PNS->PAC */ + { .src = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip, + .u = { .gre = { .key = peer_callid } } + }, + .dst = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip, + .u = { .gre = { .key = callid } }, + .protonum = IPPROTO_GRE + }, + }, + /* tuple in reply direction, PAC->PNS */ + { .src = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip, + .u = { .gre = { .key = callid } } + }, + .dst = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip, + .u = { .gre = { .key = peer_callid } }, + .protonum = IPPROTO_GRE + }, + } + }; + struct ip_conntrack_expect *exp_orig, *exp_reply; + int ret = 1; + + exp_orig = ip_conntrack_expect_alloc(master); + if (exp_orig == NULL) + goto out; + + exp_reply = ip_conntrack_expect_alloc(master); + if (exp_reply == NULL) + goto out_put_orig; + + memcpy(&exp_orig->tuple, &exp_tuples[0], sizeof(exp_orig->tuple)); + + exp_orig->mask.src.ip = 0xffffffff; + exp_orig->mask.src.u.all = 0; + exp_orig->mask.dst.u.all = 0; + exp_orig->mask.dst.u.gre.key = htons(0xffff); + exp_orig->mask.dst.ip = 0xffffffff; + exp_orig->mask.dst.protonum = 0xff; + + exp_orig->master = master; + exp_orig->expectfn = pptp_expectfn; + exp_orig->flags = 0; + + exp_orig->dir = IP_CT_DIR_ORIGINAL; + + /* both expectations are identical apart from tuple */ + memcpy(exp_reply, exp_orig, sizeof(*exp_reply)); + memcpy(&exp_reply->tuple, &exp_tuples[1], sizeof(exp_reply->tuple)); + + exp_reply->dir = !exp_orig->dir; + + if (ip_nat_pptp_hook_exp_gre) + ret = ip_nat_pptp_hook_exp_gre(exp_orig, exp_reply); + else { + + DEBUGP("calling expect_related PNS->PAC"); + DUMP_TUPLE(&exp_orig->tuple); + + if (ip_conntrack_expect_related(exp_orig) != 0) { + DEBUGP("cannot expect_related()\n"); + goto out_put_both; + } + + DEBUGP("calling expect_related PAC->PNS"); + DUMP_TUPLE(&exp_reply->tuple); + + if (ip_conntrack_expect_related(exp_reply) != 0) { + DEBUGP("cannot expect_related()\n"); + goto out_unexpect_orig; + } + + /* Add GRE keymap entries */ + if (ip_ct_gre_keymap_add(master, &exp_reply->tuple, 0) != 0) { + DEBUGP("cannot keymap_add() exp\n"); + goto out_unexpect_both; + } + + invert_tuplepr(&inv_tuple, &exp_reply->tuple); + if (ip_ct_gre_keymap_add(master, &inv_tuple, 1) != 0) { + ip_ct_gre_keymap_destroy(master); + DEBUGP("cannot keymap_add() exp_inv\n"); + goto out_unexpect_both; + } + ret = 0; + } + +out_put_both: + ip_conntrack_expect_put(exp_reply); +out_put_orig: + ip_conntrack_expect_put(exp_orig); +out: + return ret; + +out_unexpect_both: + ip_conntrack_unexpect_related(exp_reply); +out_unexpect_orig: + ip_conntrack_unexpect_related(exp_orig); + goto out_put_both; +} + +static inline int +pptp_inbound_pkt(struct sk_buff **pskb, + struct tcphdr *tcph, + unsigned int nexthdr_off, + unsigned int datalen, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo) +{ + struct PptpControlHeader _ctlh, *ctlh; + unsigned int reqlen; + union pptp_ctrl_union _pptpReq, *pptpReq; + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + u_int16_t msg; + __be16 *cid, *pcid; + u_int32_t seq; + + ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); + if (!ctlh) { + DEBUGP("error during skb_header_pointer\n"); + return NF_ACCEPT; + } + nexthdr_off += sizeof(_ctlh); + datalen -= sizeof(_ctlh); + + reqlen = datalen; + if (reqlen > sizeof(*pptpReq)) + reqlen = sizeof(*pptpReq); + pptpReq = skb_header_pointer(*pskb, nexthdr_off, reqlen, &_pptpReq); + if (!pptpReq) { + DEBUGP("error during skb_header_pointer\n"); + return NF_ACCEPT; + } + + msg = ntohs(ctlh->messageType); + DEBUGP("inbound control message %s\n", pptp_msg_name[msg]); + + switch (msg) { + case PPTP_START_SESSION_REPLY: + if (reqlen < sizeof(_pptpReq.srep)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms new control session */ + if (info->sstate < PPTP_SESSION_REQUESTED) { + DEBUGP("%s without START_SESS_REQUEST\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->srep.resultCode == PPTP_START_OK) + info->sstate = PPTP_SESSION_CONFIRMED; + else + info->sstate = PPTP_SESSION_ERROR; + break; + + case PPTP_STOP_SESSION_REPLY: + if (reqlen < sizeof(_pptpReq.strep)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms end of control session */ + if (info->sstate > PPTP_SESSION_STOPREQ) { + DEBUGP("%s without STOP_SESS_REQUEST\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->strep.resultCode == PPTP_STOP_OK) + info->sstate = PPTP_SESSION_NONE; + else + info->sstate = PPTP_SESSION_ERROR; + break; + + case PPTP_OUT_CALL_REPLY: + if (reqlen < sizeof(_pptpReq.ocack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server accepted call, we now expect GRE frames */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + if (info->cstate != PPTP_CALL_OUT_REQ && + info->cstate != PPTP_CALL_OUT_CONF) { + DEBUGP("%s without OUTCALL_REQ\n", pptp_msg_name[msg]); + break; + } + if (pptpReq->ocack.resultCode != PPTP_OUTCALL_CONNECT) { + info->cstate = PPTP_CALL_NONE; + break; + } + + cid = &pptpReq->ocack.callID; + pcid = &pptpReq->ocack.peersCallID; + + info->pac_call_id = ntohs(*cid); + + if (htons(info->pns_call_id) != *pcid) { + DEBUGP("%s for unknown callid %u\n", + pptp_msg_name[msg], ntohs(*pcid)); + break; + } + + DEBUGP("%s, CID=%X, PCID=%X\n", pptp_msg_name[msg], + ntohs(*cid), ntohs(*pcid)); + + info->cstate = PPTP_CALL_OUT_CONF; + + seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr) + + sizeof(struct PptpControlHeader) + + ((void *)pcid - (void *)pptpReq); + + if (exp_gre(ct, seq, *cid, *pcid) != 0) + printk("ip_conntrack_pptp: error during exp_gre\n"); + break; + + case PPTP_IN_CALL_REQUEST: + if (reqlen < sizeof(_pptpReq.icack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server tells us about incoming call request */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + pcid = &pptpReq->icack.peersCallID; + DEBUGP("%s, PCID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + info->cstate = PPTP_CALL_IN_REQ; + info->pac_call_id = ntohs(*pcid); + break; + + case PPTP_IN_CALL_CONNECT: + if (reqlen < sizeof(_pptpReq.iccon)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server tells us about incoming call established */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + if (info->sstate != PPTP_CALL_IN_REP + && info->sstate != PPTP_CALL_IN_CONF) { + DEBUGP("%s but never sent IN_CALL_REPLY\n", + pptp_msg_name[msg]); + break; + } + + pcid = &pptpReq->iccon.peersCallID; + cid = &info->pac_call_id; + + if (info->pns_call_id != ntohs(*pcid)) { + DEBUGP("%s for unknown CallID %u\n", + pptp_msg_name[msg], ntohs(*pcid)); + break; + } + + DEBUGP("%s, PCID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + info->cstate = PPTP_CALL_IN_CONF; + + /* we expect a GRE connection from PAC to PNS */ + seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr) + + sizeof(struct PptpControlHeader) + + ((void *)pcid - (void *)pptpReq); + + if (exp_gre(ct, seq, *cid, *pcid) != 0) + printk("ip_conntrack_pptp: error during exp_gre\n"); + + break; + + case PPTP_CALL_DISCONNECT_NOTIFY: + if (reqlen < sizeof(_pptpReq.disc)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms disconnect */ + cid = &pptpReq->disc.callID; + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*cid)); + info->cstate = PPTP_CALL_NONE; + + /* untrack this call id, unexpect GRE packets */ + pptp_destroy_siblings(ct); + break; + + case PPTP_WAN_ERROR_NOTIFY: + break; + + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* I don't have to explain these ;) */ + break; + default: + DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX) + ? pptp_msg_name[msg]:pptp_msg_name[0], msg); + break; + } + + + if (ip_nat_pptp_hook_inbound) + return ip_nat_pptp_hook_inbound(pskb, ct, ctinfo, ctlh, + pptpReq); + + return NF_ACCEPT; + +} + +static inline int +pptp_outbound_pkt(struct sk_buff **pskb, + struct tcphdr *tcph, + unsigned int nexthdr_off, + unsigned int datalen, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo) +{ + struct PptpControlHeader _ctlh, *ctlh; + unsigned int reqlen; + union pptp_ctrl_union _pptpReq, *pptpReq; + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + u_int16_t msg; + __be16 *cid, *pcid; + + ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); + if (!ctlh) + return NF_ACCEPT; + nexthdr_off += sizeof(_ctlh); + datalen -= sizeof(_ctlh); + + reqlen = datalen; + if (reqlen > sizeof(*pptpReq)) + reqlen = sizeof(*pptpReq); + pptpReq = skb_header_pointer(*pskb, nexthdr_off, reqlen, &_pptpReq); + if (!pptpReq) + return NF_ACCEPT; + + msg = ntohs(ctlh->messageType); + DEBUGP("outbound control message %s\n", pptp_msg_name[msg]); + + switch (msg) { + case PPTP_START_SESSION_REQUEST: + /* client requests for new control session */ + if (info->sstate != PPTP_SESSION_NONE) { + DEBUGP("%s but we already have one", + pptp_msg_name[msg]); + } + info->sstate = PPTP_SESSION_REQUESTED; + break; + case PPTP_STOP_SESSION_REQUEST: + /* client requests end of control session */ + info->sstate = PPTP_SESSION_STOPREQ; + break; + + case PPTP_OUT_CALL_REQUEST: + if (reqlen < sizeof(_pptpReq.ocreq)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + /* FIXME: break; */ + } + + /* client initiating connection to server */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", + pptp_msg_name[msg]); + break; + } + info->cstate = PPTP_CALL_OUT_REQ; + /* track PNS call id */ + cid = &pptpReq->ocreq.callID; + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*cid)); + info->pns_call_id = ntohs(*cid); + break; + case PPTP_IN_CALL_REPLY: + if (reqlen < sizeof(_pptpReq.icack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* client answers incoming call */ + if (info->cstate != PPTP_CALL_IN_REQ + && info->cstate != PPTP_CALL_IN_REP) { + DEBUGP("%s without incall_req\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->icack.resultCode != PPTP_INCALL_ACCEPT) { + info->cstate = PPTP_CALL_NONE; + break; + } + pcid = &pptpReq->icack.peersCallID; + if (info->pac_call_id != ntohs(*pcid)) { + DEBUGP("%s for unknown call %u\n", + pptp_msg_name[msg], ntohs(*pcid)); + break; + } + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + /* part two of the three-way handshake */ + info->cstate = PPTP_CALL_IN_REP; + info->pns_call_id = ntohs(pptpReq->icack.callID); + break; + + case PPTP_CALL_CLEAR_REQUEST: + /* client requests hangup of call */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("CLEAR_CALL but no session\n"); + break; + } + /* FUTURE: iterate over all calls and check if + * call ID is valid. We don't do this without newnat, + * because we only know about last call */ + info->cstate = PPTP_CALL_CLEAR_REQ; + break; + case PPTP_SET_LINK_INFO: + break; + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* I don't have to explain these ;) */ + break; + default: + DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0], msg); + /* unknown: no need to create GRE masq table entry */ + break; + } + + if (ip_nat_pptp_hook_outbound) + return ip_nat_pptp_hook_outbound(pskb, ct, ctinfo, ctlh, + pptpReq); + + return NF_ACCEPT; +} + + +/* track caller id inside control connection, call expect_related */ +static int +conntrack_pptp_help(struct sk_buff **pskb, + struct ip_conntrack *ct, enum ip_conntrack_info ctinfo) + +{ + struct pptp_pkt_hdr _pptph, *pptph; + struct tcphdr _tcph, *tcph; + u_int32_t tcplen = (*pskb)->len - (*pskb)->nh.iph->ihl * 4; + u_int32_t datalen; + int dir = CTINFO2DIR(ctinfo); + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + unsigned int nexthdr_off; + + int oldsstate, oldcstate; + int ret; + + /* don't do any tracking before tcp handshake complete */ + if (ctinfo != IP_CT_ESTABLISHED + && ctinfo != IP_CT_ESTABLISHED+IP_CT_IS_REPLY) { + DEBUGP("ctinfo = %u, skipping\n", ctinfo); + return NF_ACCEPT; + } + + nexthdr_off = (*pskb)->nh.iph->ihl*4; + tcph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_tcph), &_tcph); + BUG_ON(!tcph); + nexthdr_off += tcph->doff * 4; + datalen = tcplen - tcph->doff * 4; + + if (tcph->fin || tcph->rst) { + DEBUGP("RST/FIN received, timeouting GRE\n"); + /* can't do this after real newnat */ + info->cstate = PPTP_CALL_NONE; + + /* untrack this call id, unexpect GRE packets */ + pptp_destroy_siblings(ct); + } + + pptph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_pptph), &_pptph); + if (!pptph) { + DEBUGP("no full PPTP header, can't track\n"); + return NF_ACCEPT; + } + nexthdr_off += sizeof(_pptph); + datalen -= sizeof(_pptph); + + /* if it's not a control message we can't do anything with it */ + if (ntohs(pptph->packetType) != PPTP_PACKET_CONTROL || + ntohl(pptph->magicCookie) != PPTP_MAGIC_COOKIE) { + DEBUGP("not a control packet\n"); + return NF_ACCEPT; + } + + oldsstate = info->sstate; + oldcstate = info->cstate; + + spin_lock_bh(&ip_pptp_lock); + + /* FIXME: We just blindly assume that the control connection is always + * established from PNS->PAC. However, RFC makes no guarantee */ + if (dir == IP_CT_DIR_ORIGINAL) + /* client -> server (PNS -> PAC) */ + ret = pptp_outbound_pkt(pskb, tcph, nexthdr_off, datalen, ct, + ctinfo); + else + /* server -> client (PAC -> PNS) */ + ret = pptp_inbound_pkt(pskb, tcph, nexthdr_off, datalen, ct, + ctinfo); + DEBUGP("sstate: %d->%d, cstate: %d->%d\n", + oldsstate, info->sstate, oldcstate, info->cstate); + spin_unlock_bh(&ip_pptp_lock); + + return ret; +} + +/* control protocol helper */ +static struct ip_conntrack_helper pptp = { + .list = { NULL, NULL }, + .name = "pptp", + .me = THIS_MODULE, + .max_expected = 2, + .timeout = 5 * 60, + .tuple = { .src = { .ip = 0, + .u = { .tcp = { .port = + __constant_htons(PPTP_CONTROL_PORT) } } + }, + .dst = { .ip = 0, + .u = { .all = 0 }, + .protonum = IPPROTO_TCP + } + }, + .mask = { .src = { .ip = 0, + .u = { .tcp = { .port = __constant_htons(0xffff) } } + }, + .dst = { .ip = 0, + .u = { .all = 0 }, + .protonum = 0xff + } + }, + .help = conntrack_pptp_help +}; + +extern void __exit ip_ct_proto_gre_fini(void); +extern int __init ip_ct_proto_gre_init(void); + +/* ip_conntrack_pptp initialization */ +static int __init init(void) +{ + int retcode; + + retcode = ip_ct_proto_gre_init(); + if (retcode < 0) + return retcode; + + DEBUGP(" registering helper\n"); + if ((retcode = ip_conntrack_helper_register(&pptp))) { + printk(KERN_ERR "Unable to register conntrack application " + "helper for pptp: %d\n", retcode); + ip_ct_proto_gre_fini(); + return retcode; + } + + printk("ip_conntrack_pptp version %s loaded\n", IP_CT_PPTP_VERSION); + return 0; +} + +static void __exit fini(void) +{ + ip_conntrack_helper_unregister(&pptp); + ip_ct_proto_gre_fini(); + printk("ip_conntrack_pptp version %s unloaded\n", IP_CT_PPTP_VERSION); +} + +module_init(init); +module_exit(fini); + +EXPORT_SYMBOL(ip_nat_pptp_hook_outbound); +EXPORT_SYMBOL(ip_nat_pptp_hook_inbound); +EXPORT_SYMBOL(ip_nat_pptp_hook_exp_gre); +EXPORT_SYMBOL(ip_nat_pptp_hook_expectfn); diff --git a/net/ipv4/netfilter/ip_conntrack_irc.c b/net/ipv4/netfilter/ip_conntrack_irc.c index 25438eec21a..15457415a4f 100644 --- a/net/ipv4/netfilter/ip_conntrack_irc.c +++ b/net/ipv4/netfilter/ip_conntrack_irc.c @@ -34,7 +34,7 @@ #include <linux/moduleparam.h> #define MAX_PORTS 8 -static int ports[MAX_PORTS]; +static short ports[MAX_PORTS]; static int ports_c; static int max_dcc_channels = 8; static unsigned int dcc_timeout = 300; @@ -52,7 +52,7 @@ EXPORT_SYMBOL_GPL(ip_nat_irc_hook); MODULE_AUTHOR("Harald Welte <laforge@netfilter.org>"); MODULE_DESCRIPTION("IRC (DCC) connection tracking helper"); MODULE_LICENSE("GPL"); -module_param_array(ports, int, &ports_c, 0400); +module_param_array(ports, short, &ports_c, 0400); MODULE_PARM_DESC(ports, "port numbers of IRC servers"); module_param(max_dcc_channels, int, 0400); MODULE_PARM_DESC(max_dcc_channels, "max number of expected DCC channels per IRC session"); @@ -221,6 +221,7 @@ static int help(struct sk_buff **pskb, { { 0, { 0 } }, { 0xFFFFFFFF, { .tcp = { 0xFFFF } }, 0xFF }}); exp->expectfn = NULL; + exp->flags = 0; if (ip_nat_irc_hook) ret = ip_nat_irc_hook(pskb, ctinfo, addr_beg_p - ib_ptr, @@ -239,7 +240,7 @@ static int help(struct sk_buff **pskb, } static struct ip_conntrack_helper irc_helpers[MAX_PORTS]; -static char irc_names[MAX_PORTS][10]; +static char irc_names[MAX_PORTS][sizeof("irc-65535")]; static void fini(void); diff --git a/net/ipv4/netfilter/ip_conntrack_netbios_ns.c b/net/ipv4/netfilter/ip_conntrack_netbios_ns.c new file mode 100644 index 00000000000..577bac22dcc --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_netbios_ns.c @@ -0,0 +1,142 @@ +/* + * NetBIOS name service broadcast connection tracking helper + * + * (c) 2005 Patrick McHardy <kaber@trash.net> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +/* + * This helper tracks locally originating NetBIOS name service + * requests by issuing permanent expectations (valid until + * timing out) matching all reply connections from the + * destination network. The only NetBIOS specific thing is + * actually the port number. + */ +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/init.h> +#include <linux/skbuff.h> +#include <linux/netdevice.h> +#include <linux/inetdevice.h> +#include <linux/in.h> +#include <linux/ip.h> +#include <net/route.h> + +#include <linux/netfilter.h> +#include <linux/netfilter_ipv4.h> +#include <linux/netfilter_ipv4/ip_conntrack.h> +#include <linux/netfilter_ipv4/ip_conntrack_helper.h> + +#define NMBD_PORT 137 + +MODULE_AUTHOR("Patrick McHardy <kaber@trash.net>"); +MODULE_DESCRIPTION("NetBIOS name service broadcast connection tracking helper"); +MODULE_LICENSE("GPL"); + +static unsigned int timeout = 3; +module_param(timeout, int, 0600); +MODULE_PARM_DESC(timeout, "timeout for master connection/replies in seconds"); + +static int help(struct sk_buff **pskb, + struct ip_conntrack *ct, enum ip_conntrack_info ctinfo) +{ + struct ip_conntrack_expect *exp; + struct iphdr *iph = (*pskb)->nh.iph; + struct rtable *rt = (struct rtable *)(*pskb)->dst; + struct in_device *in_dev; + u_int32_t mask = 0; + + /* we're only interested in locally generated packets */ + if ((*pskb)->sk == NULL) + goto out; + if (rt == NULL || !(rt->rt_flags & RTCF_BROADCAST)) + goto out; + if (CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL) + goto out; + + rcu_read_lock(); + in_dev = __in_dev_get(rt->u.dst.dev); + if (in_dev != NULL) { + for_primary_ifa(in_dev) { + if (ifa->ifa_broadcast == iph->daddr) { + mask = ifa->ifa_mask; + break; + } + } endfor_ifa(in_dev); + } + rcu_read_unlock(); + + if (mask == 0) + goto out; + + exp = ip_conntrack_expect_alloc(ct); + if (exp == NULL) + goto out; + + exp->tuple = ct->tuplehash[IP_CT_DIR_REPLY].tuple; + exp->tuple.src.u.udp.port = ntohs(NMBD_PORT); + + exp->mask.src.ip = mask; + exp->mask.src.u.udp.port = 0xFFFF; + exp->mask.dst.ip = 0xFFFFFFFF; + exp->mask.dst.u.udp.port = 0xFFFF; + exp->mask.dst.protonum = 0xFF; + + exp->expectfn = NULL; + exp->flags = IP_CT_EXPECT_PERMANENT; + + ip_conntrack_expect_related(exp); + ip_conntrack_expect_put(exp); + + ip_ct_refresh(ct, *pskb, timeout * HZ); +out: + return NF_ACCEPT; +} + +static struct ip_conntrack_helper helper = { + .name = "netbios-ns", + .tuple = { + .src = { + .u = { + .udp = { + .port = __constant_htons(NMBD_PORT), + } + } + }, + .dst = { + .protonum = IPPROTO_UDP, + }, + }, + .mask = { + .src = { + .u = { + .udp = { + .port = 0xFFFF, + } + } + }, + .dst = { + .protonum = 0xFF, + }, + }, + .max_expected = 1, + .me = THIS_MODULE, + .help = help, +}; + +static int __init init(void) +{ + helper.timeout = timeout; + return ip_conntrack_helper_register(&helper); +} + +static void __exit fini(void) +{ + ip_conntrack_helper_unregister(&helper); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ip_conntrack_netlink.c b/net/ipv4/netfilter/ip_conntrack_netlink.c index a4e9278db4e..b08a432efcf 100644 --- a/net/ipv4/netfilter/ip_conntrack_netlink.c +++ b/net/ipv4/netfilter/ip_conntrack_netlink.c @@ -1270,7 +1270,7 @@ ctnetlink_get_expect(struct sock *ctnl, struct sk_buff *skb, if (err < 0) return err; - exp = ip_conntrack_expect_find_get(&tuple); + exp = ip_conntrack_expect_find(&tuple); if (!exp) return -ENOENT; @@ -1318,7 +1318,7 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, return err; /* bump usage count to 2 */ - exp = ip_conntrack_expect_find_get(&tuple); + exp = ip_conntrack_expect_find(&tuple); if (!exp) return -ENOENT; @@ -1349,8 +1349,10 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, list) { if (exp->master->helper == h - && del_timer(&exp->timeout)) - __ip_ct_expect_unlink_destroy(exp); + && del_timer(&exp->timeout)) { + ip_ct_unlink_expect(exp); + ip_conntrack_expect_put(exp); + } } write_unlock(&ip_conntrack_lock); } else { @@ -1358,8 +1360,10 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, write_lock_bh(&ip_conntrack_lock); list_for_each_entry_safe(exp, tmp, &ip_conntrack_expect_list, list) { - if (del_timer(&exp->timeout)) - __ip_ct_expect_unlink_destroy(exp); + if (del_timer(&exp->timeout)) { + ip_ct_unlink_expect(exp); + ip_conntrack_expect_put(exp); + } } write_unlock_bh(&ip_conntrack_lock); } @@ -1413,6 +1417,7 @@ ctnetlink_create_expect(struct nfattr *cda[]) } exp->expectfn = NULL; + exp->flags = 0; exp->master = ct; memcpy(&exp->tuple, &tuple, sizeof(struct ip_conntrack_tuple)); memcpy(&exp->mask, &mask, sizeof(struct ip_conntrack_tuple)); diff --git a/net/ipv4/netfilter/ip_conntrack_proto_gre.c b/net/ipv4/netfilter/ip_conntrack_proto_gre.c new file mode 100644 index 00000000000..de3cb9db6f8 --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_proto_gre.c @@ -0,0 +1,327 @@ +/* + * ip_conntrack_proto_gre.c - Version 3.0 + * + * Connection tracking protocol helper module for GRE. + * + * GRE is a generic encapsulation protocol, which is generally not very + * suited for NAT, as it has no protocol-specific part as port numbers. + * + * It has an optional key field, which may help us distinguishing two + * connections between the same two hosts. + * + * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 + * + * PPTP is built on top of a modified version of GRE, and has a mandatory + * field called "CallID", which serves us for the same purpose as the key + * field in plain GRE. + * + * Documentation about PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte <laforge@gnumonks.org> + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/types.h> +#include <linux/timer.h> +#include <linux/netfilter.h> +#include <linux/ip.h> +#include <linux/in.h> +#include <linux/list.h> + +static DEFINE_RWLOCK(ip_ct_gre_lock); +#define ASSERT_READ_LOCK(x) +#define ASSERT_WRITE_LOCK(x) + +#include <linux/netfilter_ipv4/listhelp.h> +#include <linux/netfilter_ipv4/ip_conntrack_protocol.h> +#include <linux/netfilter_ipv4/ip_conntrack_helper.h> +#include <linux/netfilter_ipv4/ip_conntrack_core.h> + +#include <linux/netfilter_ipv4/ip_conntrack_proto_gre.h> +#include <linux/netfilter_ipv4/ip_conntrack_pptp.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@gnumonks.org>"); +MODULE_DESCRIPTION("netfilter connection tracking protocol helper for GRE"); + +/* shamelessly stolen from ip_conntrack_proto_udp.c */ +#define GRE_TIMEOUT (30*HZ) +#define GRE_STREAM_TIMEOUT (180*HZ) + +#if 0 +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, __FUNCTION__, ## args) +#define DUMP_TUPLE_GRE(x) printk("%u.%u.%u.%u:0x%x -> %u.%u.%u.%u:0x%x\n", \ + NIPQUAD((x)->src.ip), ntohs((x)->src.u.gre.key), \ + NIPQUAD((x)->dst.ip), ntohs((x)->dst.u.gre.key)) +#else +#define DEBUGP(x, args...) +#define DUMP_TUPLE_GRE(x) +#endif + +/* GRE KEYMAP HANDLING FUNCTIONS */ +static LIST_HEAD(gre_keymap_list); + +static inline int gre_key_cmpfn(const struct ip_ct_gre_keymap *km, + const struct ip_conntrack_tuple *t) +{ + return ((km->tuple.src.ip == t->src.ip) && + (km->tuple.dst.ip == t->dst.ip) && + (km->tuple.dst.protonum == t->dst.protonum) && + (km->tuple.dst.u.all == t->dst.u.all)); +} + +/* look up the source key for a given tuple */ +static u_int32_t gre_keymap_lookup(struct ip_conntrack_tuple *t) +{ + struct ip_ct_gre_keymap *km; + u_int32_t key = 0; + + read_lock_bh(&ip_ct_gre_lock); + km = LIST_FIND(&gre_keymap_list, gre_key_cmpfn, + struct ip_ct_gre_keymap *, t); + if (km) + key = km->tuple.src.u.gre.key; + read_unlock_bh(&ip_ct_gre_lock); + + DEBUGP("lookup src key 0x%x up key for ", key); + DUMP_TUPLE_GRE(t); + + return key; +} + +/* add a single keymap entry, associate with specified master ct */ +int +ip_ct_gre_keymap_add(struct ip_conntrack *ct, + struct ip_conntrack_tuple *t, int reply) +{ + struct ip_ct_gre_keymap **exist_km, *km, *old; + + if (!ct->helper || strcmp(ct->helper->name, "pptp")) { + DEBUGP("refusing to add GRE keymap to non-pptp session\n"); + return -1; + } + + if (!reply) + exist_km = &ct->help.ct_pptp_info.keymap_orig; + else + exist_km = &ct->help.ct_pptp_info.keymap_reply; + + if (*exist_km) { + /* check whether it's a retransmission */ + old = LIST_FIND(&gre_keymap_list, gre_key_cmpfn, + struct ip_ct_gre_keymap *, t); + if (old == *exist_km) { + DEBUGP("retransmission\n"); + return 0; + } + + DEBUGP("trying to override keymap_%s for ct %p\n", + reply? "reply":"orig", ct); + return -EEXIST; + } + + km = kmalloc(sizeof(*km), GFP_ATOMIC); + if (!km) + return -ENOMEM; + + memcpy(&km->tuple, t, sizeof(*t)); + *exist_km = km; + + DEBUGP("adding new entry %p: ", km); + DUMP_TUPLE_GRE(&km->tuple); + + write_lock_bh(&ip_ct_gre_lock); + list_append(&gre_keymap_list, km); + write_unlock_bh(&ip_ct_gre_lock); + + return 0; +} + +/* destroy the keymap entries associated with specified master ct */ +void ip_ct_gre_keymap_destroy(struct ip_conntrack *ct) +{ + DEBUGP("entering for ct %p\n", ct); + + if (!ct->helper || strcmp(ct->helper->name, "pptp")) { + DEBUGP("refusing to destroy GRE keymap to non-pptp session\n"); + return; + } + + write_lock_bh(&ip_ct_gre_lock); + if (ct->help.ct_pptp_info.keymap_orig) { + DEBUGP("removing %p from list\n", + ct->help.ct_pptp_info.keymap_orig); + list_del(&ct->help.ct_pptp_info.keymap_orig->list); + kfree(ct->help.ct_pptp_info.keymap_orig); + ct->help.ct_pptp_info.keymap_orig = NULL; + } + if (ct->help.ct_pptp_info.keymap_reply) { + DEBUGP("removing %p from list\n", + ct->help.ct_pptp_info.keymap_reply); + list_del(&ct->help.ct_pptp_info.keymap_reply->list); + kfree(ct->help.ct_pptp_info.keymap_reply); + ct->help.ct_pptp_info.keymap_reply = NULL; + } + write_unlock_bh(&ip_ct_gre_lock); +} + + +/* PUBLIC CONNTRACK PROTO HELPER FUNCTIONS */ + +/* invert gre part of tuple */ +static int gre_invert_tuple(struct ip_conntrack_tuple *tuple, + const struct ip_conntrack_tuple *orig) +{ + tuple->dst.u.gre.key = orig->src.u.gre.key; + tuple->src.u.gre.key = orig->dst.u.gre.key; + + return 1; +} + +/* gre hdr info to tuple */ +static int gre_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct ip_conntrack_tuple *tuple) +{ + struct gre_hdr_pptp _pgrehdr, *pgrehdr; + u_int32_t srckey; + struct gre_hdr _grehdr, *grehdr; + + /* first only delinearize old RFC1701 GRE header */ + grehdr = skb_header_pointer(skb, dataoff, sizeof(_grehdr), &_grehdr); + if (!grehdr || grehdr->version != GRE_VERSION_PPTP) { + /* try to behave like "ip_conntrack_proto_generic" */ + tuple->src.u.all = 0; + tuple->dst.u.all = 0; + return 1; + } + + /* PPTP header is variable length, only need up to the call_id field */ + pgrehdr = skb_header_pointer(skb, dataoff, 8, &_pgrehdr); + if (!pgrehdr) + return 1; + + if (ntohs(grehdr->protocol) != GRE_PROTOCOL_PPTP) { + DEBUGP("GRE_VERSION_PPTP but unknown proto\n"); + return 0; + } + + tuple->dst.u.gre.key = pgrehdr->call_id; + srckey = gre_keymap_lookup(tuple); + tuple->src.u.gre.key = srckey; + + return 1; +} + +/* print gre part of tuple */ +static int gre_print_tuple(struct seq_file *s, + const struct ip_conntrack_tuple *tuple) +{ + return seq_printf(s, "srckey=0x%x dstkey=0x%x ", + ntohs(tuple->src.u.gre.key), + ntohs(tuple->dst.u.gre.key)); +} + +/* print private data for conntrack */ +static int gre_print_conntrack(struct seq_file *s, + const struct ip_conntrack *ct) +{ + return seq_printf(s, "timeout=%u, stream_timeout=%u ", + (ct->proto.gre.timeout / HZ), + (ct->proto.gre.stream_timeout / HZ)); +} + +/* Returns verdict for packet, and may modify conntrack */ +static int gre_packet(struct ip_conntrack *ct, + const struct sk_buff *skb, + enum ip_conntrack_info conntrackinfo) +{ + /* If we've seen traffic both ways, this is a GRE connection. + * Extend timeout. */ + if (ct->status & IPS_SEEN_REPLY) { + ip_ct_refresh_acct(ct, conntrackinfo, skb, + ct->proto.gre.stream_timeout); + /* Also, more likely to be important, and not a probe. */ + set_bit(IPS_ASSURED_BIT, &ct->status); + } else + ip_ct_refresh_acct(ct, conntrackinfo, skb, + ct->proto.gre.timeout); + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int gre_new(struct ip_conntrack *ct, + const struct sk_buff *skb) +{ + DEBUGP(": "); + DUMP_TUPLE_GRE(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + + /* initialize to sane value. Ideally a conntrack helper + * (e.g. in case of pptp) is increasing them */ + ct->proto.gre.stream_timeout = GRE_STREAM_TIMEOUT; + ct->proto.gre.timeout = GRE_TIMEOUT; + + return 1; +} + +/* Called when a conntrack entry has already been removed from the hashes + * and is about to be deleted from memory */ +static void gre_destroy(struct ip_conntrack *ct) +{ + struct ip_conntrack *master = ct->master; + DEBUGP(" entering\n"); + + if (!master) + DEBUGP("no master !?!\n"); + else + ip_ct_gre_keymap_destroy(master); +} + +/* protocol helper struct */ +static struct ip_conntrack_protocol gre = { + .proto = IPPROTO_GRE, + .name = "gre", + .pkt_to_tuple = gre_pkt_to_tuple, + .invert_tuple = gre_invert_tuple, + .print_tuple = gre_print_tuple, + .print_conntrack = gre_print_conntrack, + .packet = gre_packet, + .new = gre_new, + .destroy = gre_destroy, + .me = THIS_MODULE, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .tuple_to_nfattr = ip_ct_port_tuple_to_nfattr, + .nfattr_to_tuple = ip_ct_port_nfattr_to_tuple, +#endif +}; + +/* ip_conntrack_proto_gre initialization */ +int __init ip_ct_proto_gre_init(void) +{ + return ip_conntrack_protocol_register(&gre); +} + +void __exit ip_ct_proto_gre_fini(void) +{ + struct list_head *pos, *n; + + /* delete all keymap entries */ + write_lock_bh(&ip_ct_gre_lock); + list_for_each_safe(pos, n, &gre_keymap_list) { + DEBUGP("deleting keymap %p at module unload time\n", pos); + list_del(pos); + kfree(pos); + } + write_unlock_bh(&ip_ct_gre_lock); + + ip_conntrack_protocol_unregister(&gre); +} + +EXPORT_SYMBOL(ip_ct_gre_keymap_add); +EXPORT_SYMBOL(ip_ct_gre_keymap_destroy); diff --git a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c index f23ef1f88c4..1985abc59d2 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_tcp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_tcp.c @@ -349,6 +349,7 @@ static int tcp_to_nfattr(struct sk_buff *skb, struct nfattr *nfa, return 0; nfattr_failure: + read_unlock_bh(&tcp_lock); return -1; } #endif diff --git a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c index ee5895afd0c..dd476b191f4 100644 --- a/net/ipv4/netfilter/ip_conntrack_standalone.c +++ b/net/ipv4/netfilter/ip_conntrack_standalone.c @@ -989,16 +989,16 @@ EXPORT_SYMBOL(need_ip_conntrack); EXPORT_SYMBOL(ip_conntrack_helper_register); EXPORT_SYMBOL(ip_conntrack_helper_unregister); EXPORT_SYMBOL(ip_ct_iterate_cleanup); -EXPORT_SYMBOL(ip_ct_refresh_acct); +EXPORT_SYMBOL(__ip_ct_refresh_acct); EXPORT_SYMBOL(ip_conntrack_expect_alloc); EXPORT_SYMBOL(ip_conntrack_expect_put); -EXPORT_SYMBOL_GPL(ip_conntrack_expect_find_get); +EXPORT_SYMBOL_GPL(__ip_conntrack_expect_find); +EXPORT_SYMBOL_GPL(ip_conntrack_expect_find); EXPORT_SYMBOL(ip_conntrack_expect_related); EXPORT_SYMBOL(ip_conntrack_unexpect_related); EXPORT_SYMBOL_GPL(ip_conntrack_expect_list); -EXPORT_SYMBOL_GPL(__ip_conntrack_expect_find); -EXPORT_SYMBOL_GPL(__ip_ct_expect_unlink_destroy); +EXPORT_SYMBOL_GPL(ip_ct_unlink_expect); EXPORT_SYMBOL(ip_conntrack_tuple_taken); EXPORT_SYMBOL(ip_ct_gather_frags); diff --git a/net/ipv4/netfilter/ip_conntrack_tftp.c b/net/ipv4/netfilter/ip_conntrack_tftp.c index f8ff170f390..a78736b8525 100644 --- a/net/ipv4/netfilter/ip_conntrack_tftp.c +++ b/net/ipv4/netfilter/ip_conntrack_tftp.c @@ -26,9 +26,9 @@ MODULE_DESCRIPTION("tftp connection tracking helper"); MODULE_LICENSE("GPL"); #define MAX_PORTS 8 -static int ports[MAX_PORTS]; +static short ports[MAX_PORTS]; static int ports_c; -module_param_array(ports, int, &ports_c, 0400); +module_param_array(ports, short, &ports_c, 0400); MODULE_PARM_DESC(ports, "port numbers of tftp servers"); #if 0 @@ -75,6 +75,7 @@ static int tftp_help(struct sk_buff **pskb, exp->mask.dst.u.udp.port = 0xffff; exp->mask.dst.protonum = 0xff; exp->expectfn = NULL; + exp->flags = 0; DEBUGP("expect: "); DUMP_TUPLE(&exp->tuple); @@ -99,7 +100,7 @@ static int tftp_help(struct sk_buff **pskb, } static struct ip_conntrack_helper tftp[MAX_PORTS]; -static char tftp_names[MAX_PORTS][10]; +static char tftp_names[MAX_PORTS][sizeof("tftp-65535")]; static void fini(void) { diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c index 1adedb743f6..c3ea891d38e 100644 --- a/net/ipv4/netfilter/ip_nat_core.c +++ b/net/ipv4/netfilter/ip_nat_core.c @@ -578,6 +578,8 @@ ip_nat_port_nfattr_to_range(struct nfattr *tb[], struct ip_nat_range *range) return ret; } +EXPORT_SYMBOL_GPL(ip_nat_port_nfattr_to_range); +EXPORT_SYMBOL_GPL(ip_nat_port_range_to_nfattr); #endif int __init ip_nat_init(void) diff --git a/net/ipv4/netfilter/ip_nat_helper_pptp.c b/net/ipv4/netfilter/ip_nat_helper_pptp.c new file mode 100644 index 00000000000..3cdd0684d30 --- /dev/null +++ b/net/ipv4/netfilter/ip_nat_helper_pptp.c @@ -0,0 +1,401 @@ +/* + * ip_nat_pptp.c - Version 3.0 + * + * NAT support for PPTP (Point to Point Tunneling Protocol). + * PPTP is a a protocol for creating virtual private networks. + * It is a specification defined by Microsoft and some vendors + * working with Microsoft. PPTP is built on top of a modified + * version of the Internet Generic Routing Encapsulation Protocol. + * GRE is defined in RFC 1701 and RFC 1702. Documentation of + * PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte <laforge@gnumonks.org> + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + * TODO: - NAT to a unique tuple, not to TCP source port + * (needs netfilter tuple reservation) + * + * Changes: + * 2002-02-10 - Version 1.3 + * - Use ip_nat_mangle_tcp_packet() because of cloned skb's + * in local connections (Philip Craig <philipc@snapgear.com>) + * - add checks for magicCookie and pptp version + * - make argument list of pptp_{out,in}bound_packet() shorter + * - move to C99 style initializers + * - print version number at module loadtime + * 2003-09-22 - Version 1.5 + * - use SNATed tcp sourceport as callid, since we get called before + * TCP header is mangled (Philip Craig <philipc@snapgear.com>) + * 2004-10-22 - Version 2.0 + * - kernel 2.6.x version + * 2005-06-10 - Version 3.0 + * - kernel >= 2.6.11 version, + * funded by Oxcoda NetBox Blue (http://www.netboxblue.com/) + * + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/ip.h> +#include <linux/tcp.h> +#include <net/tcp.h> + +#include <linux/netfilter_ipv4/ip_nat.h> +#include <linux/netfilter_ipv4/ip_nat_rule.h> +#include <linux/netfilter_ipv4/ip_nat_helper.h> +#include <linux/netfilter_ipv4/ip_nat_pptp.h> +#include <linux/netfilter_ipv4/ip_conntrack_core.h> +#include <linux/netfilter_ipv4/ip_conntrack_helper.h> +#include <linux/netfilter_ipv4/ip_conntrack_proto_gre.h> +#include <linux/netfilter_ipv4/ip_conntrack_pptp.h> + +#define IP_NAT_PPTP_VERSION "3.0" + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@gnumonks.org>"); +MODULE_DESCRIPTION("Netfilter NAT helper module for PPTP"); + + +#if 0 +extern const char *pptp_msg_name[]; +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, \ + __FUNCTION__, ## args) +#else +#define DEBUGP(format, args...) +#endif + +static void pptp_nat_expected(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp) +{ + struct ip_conntrack *master = ct->master; + struct ip_conntrack_expect *other_exp; + struct ip_conntrack_tuple t; + struct ip_ct_pptp_master *ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info; + + ct_pptp_info = &master->help.ct_pptp_info; + nat_pptp_info = &master->nat.help.nat_pptp_info; + + /* And here goes the grand finale of corrosion... */ + + if (exp->dir == IP_CT_DIR_ORIGINAL) { + DEBUGP("we are PNS->PAC\n"); + /* therefore, build tuple for PAC->PNS */ + t.src.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip; + t.src.u.gre.key = htons(master->help.ct_pptp_info.pac_call_id); + t.dst.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip; + t.dst.u.gre.key = htons(master->help.ct_pptp_info.pns_call_id); + t.dst.protonum = IPPROTO_GRE; + } else { + DEBUGP("we are PAC->PNS\n"); + /* build tuple for PNS->PAC */ + t.src.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip; + t.src.u.gre.key = + htons(master->nat.help.nat_pptp_info.pns_call_id); + t.dst.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip; + t.dst.u.gre.key = + htons(master->nat.help.nat_pptp_info.pac_call_id); + t.dst.protonum = IPPROTO_GRE; + } + + DEBUGP("trying to unexpect other dir: "); + DUMP_TUPLE(&t); + other_exp = ip_conntrack_expect_find(&t); + if (other_exp) { + ip_conntrack_unexpect_related(other_exp); + ip_conntrack_expect_put(other_exp); + DEBUGP("success\n"); + } else { + DEBUGP("not found!\n"); + } + + ip_nat_follow_master(ct, exp); +} + +/* outbound packets == from PNS to PAC */ +static int +pptp_outbound_pkt(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq) + +{ + struct ip_ct_pptp_master *ct_pptp_info = &ct->help.ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info; + + u_int16_t msg, *cid = NULL, new_callid; + + new_callid = htons(ct_pptp_info->pns_call_id); + + switch (msg = ntohs(ctlh->messageType)) { + case PPTP_OUT_CALL_REQUEST: + cid = &pptpReq->ocreq.callID; + /* FIXME: ideally we would want to reserve a call ID + * here. current netfilter NAT core is not able to do + * this :( For now we use TCP source port. This breaks + * multiple calls within one control session */ + + /* save original call ID in nat_info */ + nat_pptp_info->pns_call_id = ct_pptp_info->pns_call_id; + + /* don't use tcph->source since we are at a DSTmanip + * hook (e.g. PREROUTING) and pkt is not mangled yet */ + new_callid = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.tcp.port; + + /* save new call ID in ct info */ + ct_pptp_info->pns_call_id = ntohs(new_callid); + break; + case PPTP_IN_CALL_REPLY: + cid = &pptpReq->icreq.callID; + break; + case PPTP_CALL_CLEAR_REQUEST: + cid = &pptpReq->clrreq.callID; + break; + default: + DEBUGP("unknown outbound packet 0x%04x:%s\n", msg, + (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0]); + /* fall through */ + + case PPTP_SET_LINK_INFO: + /* only need to NAT in case PAC is behind NAT box */ + case PPTP_START_SESSION_REQUEST: + case PPTP_START_SESSION_REPLY: + case PPTP_STOP_SESSION_REQUEST: + case PPTP_STOP_SESSION_REPLY: + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* no need to alter packet */ + return NF_ACCEPT; + } + + /* only OUT_CALL_REQUEST, IN_CALL_REPLY, CALL_CLEAR_REQUEST pass + * down to here */ + + IP_NF_ASSERT(cid); + + DEBUGP("altering call id from 0x%04x to 0x%04x\n", + ntohs(*cid), ntohs(new_callid)); + + /* mangle packet */ + if (ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)cid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_callid), + (char *)&new_callid, + sizeof(new_callid)) == 0) + return NF_DROP; + + return NF_ACCEPT; +} + +static int +pptp_exp_gre(struct ip_conntrack_expect *expect_orig, + struct ip_conntrack_expect *expect_reply) +{ + struct ip_ct_pptp_master *ct_pptp_info = + &expect_orig->master->help.ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info = + &expect_orig->master->nat.help.nat_pptp_info; + + struct ip_conntrack *ct = expect_orig->master; + + struct ip_conntrack_tuple inv_t; + struct ip_conntrack_tuple *orig_t, *reply_t; + + /* save original PAC call ID in nat_info */ + nat_pptp_info->pac_call_id = ct_pptp_info->pac_call_id; + + /* alter expectation */ + orig_t = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; + reply_t = &ct->tuplehash[IP_CT_DIR_REPLY].tuple; + + /* alter expectation for PNS->PAC direction */ + invert_tuplepr(&inv_t, &expect_orig->tuple); + expect_orig->saved_proto.gre.key = htons(nat_pptp_info->pac_call_id); + expect_orig->tuple.src.u.gre.key = htons(nat_pptp_info->pns_call_id); + expect_orig->tuple.dst.u.gre.key = htons(ct_pptp_info->pac_call_id); + inv_t.src.ip = reply_t->src.ip; + inv_t.dst.ip = reply_t->dst.ip; + inv_t.src.u.gre.key = htons(nat_pptp_info->pac_call_id); + inv_t.dst.u.gre.key = htons(ct_pptp_info->pns_call_id); + + if (!ip_conntrack_expect_related(expect_orig)) { + DEBUGP("successfully registered expect\n"); + } else { + DEBUGP("can't expect_related(expect_orig)\n"); + return 1; + } + + /* alter expectation for PAC->PNS direction */ + invert_tuplepr(&inv_t, &expect_reply->tuple); + expect_reply->saved_proto.gre.key = htons(nat_pptp_info->pns_call_id); + expect_reply->tuple.src.u.gre.key = htons(nat_pptp_info->pac_call_id); + expect_reply->tuple.dst.u.gre.key = htons(ct_pptp_info->pns_call_id); + inv_t.src.ip = orig_t->src.ip; + inv_t.dst.ip = orig_t->dst.ip; + inv_t.src.u.gre.key = htons(nat_pptp_info->pns_call_id); + inv_t.dst.u.gre.key = htons(ct_pptp_info->pac_call_id); + + if (!ip_conntrack_expect_related(expect_reply)) { + DEBUGP("successfully registered expect\n"); + } else { + DEBUGP("can't expect_related(expect_reply)\n"); + ip_conntrack_unexpect_related(expect_orig); + return 1; + } + + if (ip_ct_gre_keymap_add(ct, &expect_reply->tuple, 0) < 0) { + DEBUGP("can't register original keymap\n"); + ip_conntrack_unexpect_related(expect_orig); + ip_conntrack_unexpect_related(expect_reply); + return 1; + } + + if (ip_ct_gre_keymap_add(ct, &inv_t, 1) < 0) { + DEBUGP("can't register reply keymap\n"); + ip_conntrack_unexpect_related(expect_orig); + ip_conntrack_unexpect_related(expect_reply); + ip_ct_gre_keymap_destroy(ct); + return 1; + } + + return 0; +} + +/* inbound packets == from PAC to PNS */ +static int +pptp_inbound_pkt(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq) +{ + struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info; + u_int16_t msg, new_cid = 0, new_pcid, *pcid = NULL, *cid = NULL; + + int ret = NF_ACCEPT, rv; + + new_pcid = htons(nat_pptp_info->pns_call_id); + + switch (msg = ntohs(ctlh->messageType)) { + case PPTP_OUT_CALL_REPLY: + pcid = &pptpReq->ocack.peersCallID; + cid = &pptpReq->ocack.callID; + break; + case PPTP_IN_CALL_CONNECT: + pcid = &pptpReq->iccon.peersCallID; + break; + case PPTP_IN_CALL_REQUEST: + /* only need to nat in case PAC is behind NAT box */ + break; + case PPTP_WAN_ERROR_NOTIFY: + pcid = &pptpReq->wanerr.peersCallID; + break; + case PPTP_CALL_DISCONNECT_NOTIFY: + pcid = &pptpReq->disc.callID; + break; + case PPTP_SET_LINK_INFO: + pcid = &pptpReq->setlink.peersCallID; + break; + + default: + DEBUGP("unknown inbound packet %s\n", (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0]); + /* fall through */ + + case PPTP_START_SESSION_REQUEST: + case PPTP_START_SESSION_REPLY: + case PPTP_STOP_SESSION_REQUEST: + case PPTP_STOP_SESSION_REPLY: + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* no need to alter packet */ + return NF_ACCEPT; + } + + /* only OUT_CALL_REPLY, IN_CALL_CONNECT, IN_CALL_REQUEST, + * WAN_ERROR_NOTIFY, CALL_DISCONNECT_NOTIFY pass down here */ + + /* mangle packet */ + IP_NF_ASSERT(pcid); + DEBUGP("altering peer call id from 0x%04x to 0x%04x\n", + ntohs(*pcid), ntohs(new_pcid)); + + rv = ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)pcid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_pcid), (char *)&new_pcid, + sizeof(new_pcid)); + if (rv != NF_ACCEPT) + return rv; + + if (new_cid) { + IP_NF_ASSERT(cid); + DEBUGP("altering call id from 0x%04x to 0x%04x\n", + ntohs(*cid), ntohs(new_cid)); + rv = ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)cid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_cid), + (char *)&new_cid, + sizeof(new_cid)); + if (rv != NF_ACCEPT) + return rv; + } + + /* check for earlier return value of 'switch' above */ + if (ret != NF_ACCEPT) + return ret; + + /* great, at least we don't need to resize packets */ + return NF_ACCEPT; +} + + +extern int __init ip_nat_proto_gre_init(void); +extern void __exit ip_nat_proto_gre_fini(void); + +static int __init init(void) +{ + int ret; + + DEBUGP("%s: registering NAT helper\n", __FILE__); + + ret = ip_nat_proto_gre_init(); + if (ret < 0) + return ret; + + BUG_ON(ip_nat_pptp_hook_outbound); + ip_nat_pptp_hook_outbound = &pptp_outbound_pkt; + + BUG_ON(ip_nat_pptp_hook_inbound); + ip_nat_pptp_hook_inbound = &pptp_inbound_pkt; + + BUG_ON(ip_nat_pptp_hook_exp_gre); + ip_nat_pptp_hook_exp_gre = &pptp_exp_gre; + + BUG_ON(ip_nat_pptp_hook_expectfn); + ip_nat_pptp_hook_expectfn = &pptp_nat_expected; + + printk("ip_nat_pptp version %s loaded\n", IP_NAT_PPTP_VERSION); + return 0; +} + +static void __exit fini(void) +{ + DEBUGP("cleanup_module\n" ); + + ip_nat_pptp_hook_expectfn = NULL; + ip_nat_pptp_hook_exp_gre = NULL; + ip_nat_pptp_hook_inbound = NULL; + ip_nat_pptp_hook_outbound = NULL; + + ip_nat_proto_gre_fini(); + /* Make sure noone calls it, meanwhile */ + synchronize_net(); + + printk("ip_nat_pptp version %s unloaded\n", IP_NAT_PPTP_VERSION); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ip_nat_proto_gre.c b/net/ipv4/netfilter/ip_nat_proto_gre.c new file mode 100644 index 00000000000..7c128540167 --- /dev/null +++ b/net/ipv4/netfilter/ip_nat_proto_gre.c @@ -0,0 +1,214 @@ +/* + * ip_nat_proto_gre.c - Version 2.0 + * + * NAT protocol helper module for GRE. + * + * GRE is a generic encapsulation protocol, which is generally not very + * suited for NAT, as it has no protocol-specific part as port numbers. + * + * It has an optional key field, which may help us distinguishing two + * connections between the same two hosts. + * + * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 + * + * PPTP is built on top of a modified version of GRE, and has a mandatory + * field called "CallID", which serves us for the same purpose as the key + * field in plain GRE. + * + * Documentation about PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte <laforge@gnumonks.org> + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + */ + +#include <linux/config.h> +#include <linux/module.h> +#include <linux/ip.h> +#include <linux/netfilter_ipv4/ip_nat.h> +#include <linux/netfilter_ipv4/ip_nat_rule.h> +#include <linux/netfilter_ipv4/ip_nat_protocol.h> +#include <linux/netfilter_ipv4/ip_conntrack_proto_gre.h> + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte <laforge@gnumonks.org>"); +MODULE_DESCRIPTION("Netfilter NAT protocol helper module for GRE"); + +#if 0 +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, \ + __FUNCTION__, ## args) +#else +#define DEBUGP(x, args...) +#endif + +/* is key in given range between min and max */ +static int +gre_in_range(const struct ip_conntrack_tuple *tuple, + enum ip_nat_manip_type maniptype, + const union ip_conntrack_manip_proto *min, + const union ip_conntrack_manip_proto *max) +{ + u_int32_t key; + + if (maniptype == IP_NAT_MANIP_SRC) + key = tuple->src.u.gre.key; + else + key = tuple->dst.u.gre.key; + + return ntohl(key) >= ntohl(min->gre.key) + && ntohl(key) <= ntohl(max->gre.key); +} + +/* generate unique tuple ... */ +static int +gre_unique_tuple(struct ip_conntrack_tuple *tuple, + const struct ip_nat_range *range, + enum ip_nat_manip_type maniptype, + const struct ip_conntrack *conntrack) +{ + static u_int16_t key; + u_int16_t *keyptr; + unsigned int min, i, range_size; + + if (maniptype == IP_NAT_MANIP_SRC) + keyptr = &tuple->src.u.gre.key; + else + keyptr = &tuple->dst.u.gre.key; + + if (!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED)) { + DEBUGP("%p: NATing GRE PPTP\n", conntrack); + min = 1; + range_size = 0xffff; + } else { + min = ntohl(range->min.gre.key); + range_size = ntohl(range->max.gre.key) - min + 1; + } + + DEBUGP("min = %u, range_size = %u\n", min, range_size); + + for (i = 0; i < range_size; i++, key++) { + *keyptr = htonl(min + key % range_size); + if (!ip_nat_used_tuple(tuple, conntrack)) + return 1; + } + + DEBUGP("%p: no NAT mapping\n", conntrack); + + return 0; +} + +/* manipulate a GRE packet according to maniptype */ +static int +gre_manip_pkt(struct sk_buff **pskb, + unsigned int iphdroff, + const struct ip_conntrack_tuple *tuple, + enum ip_nat_manip_type maniptype) +{ + struct gre_hdr *greh; + struct gre_hdr_pptp *pgreh; + struct iphdr *iph = (struct iphdr *)((*pskb)->data + iphdroff); + unsigned int hdroff = iphdroff + iph->ihl*4; + + /* pgreh includes two optional 32bit fields which are not required + * to be there. That's where the magic '8' comes from */ + if (!skb_make_writable(pskb, hdroff + sizeof(*pgreh)-8)) + return 0; + + greh = (void *)(*pskb)->data + hdroff; + pgreh = (struct gre_hdr_pptp *) greh; + + /* we only have destination manip of a packet, since 'source key' + * is not present in the packet itself */ + if (maniptype == IP_NAT_MANIP_DST) { + /* key manipulation is always dest */ + switch (greh->version) { + case 0: + if (!greh->key) { + DEBUGP("can't nat GRE w/o key\n"); + break; + } + if (greh->csum) { + /* FIXME: Never tested this code... */ + *(gre_csum(greh)) = + ip_nat_cheat_check(~*(gre_key(greh)), + tuple->dst.u.gre.key, + *(gre_csum(greh))); + } + *(gre_key(greh)) = tuple->dst.u.gre.key; + break; + case GRE_VERSION_PPTP: + DEBUGP("call_id -> 0x%04x\n", + ntohl(tuple->dst.u.gre.key)); + pgreh->call_id = htons(ntohl(tuple->dst.u.gre.key)); + break; + default: + DEBUGP("can't nat unknown GRE version\n"); + return 0; + break; + } + } + return 1; +} + +/* print out a nat tuple */ +static unsigned int +gre_print(char *buffer, + const struct ip_conntrack_tuple *match, + const struct ip_conntrack_tuple *mask) +{ + unsigned int len = 0; + + if (mask->src.u.gre.key) + len += sprintf(buffer + len, "srckey=0x%x ", + ntohl(match->src.u.gre.key)); + + if (mask->dst.u.gre.key) + len += sprintf(buffer + len, "dstkey=0x%x ", + ntohl(match->src.u.gre.key)); + + return len; +} + +/* print a range of keys */ +static unsigned int +gre_print_range(char *buffer, const struct ip_nat_range *range) +{ + if (range->min.gre.key != 0 + || range->max.gre.key != 0xFFFF) { + if (range->min.gre.key == range->max.gre.key) + return sprintf(buffer, "key 0x%x ", + ntohl(range->min.gre.key)); + else + return sprintf(buffer, "keys 0x%u-0x%u ", + ntohl(range->min.gre.key), + ntohl(range->max.gre.key)); + } else + return 0; +} + +/* nat helper struct */ +static struct ip_nat_protocol gre = { + .name = "GRE", + .protonum = IPPROTO_GRE, + .manip_pkt = gre_manip_pkt, + .in_range = gre_in_range, + .unique_tuple = gre_unique_tuple, + .print = gre_print, + .print_range = gre_print_range, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .range_to_nfattr = ip_nat_port_range_to_nfattr, + .nfattr_to_range = ip_nat_port_nfattr_to_range, +#endif +}; + +int __init ip_nat_proto_gre_init(void) +{ + return ip_nat_protocol_register(&gre); +} + +void __exit ip_nat_proto_gre_fini(void) +{ + ip_nat_protocol_unregister(&gre); +} diff --git a/net/ipv4/netfilter/ip_nat_rule.c b/net/ipv4/netfilter/ip_nat_rule.c index 60d70fa41a1..cb66b8bddeb 100644 --- a/net/ipv4/netfilter/ip_nat_rule.c +++ b/net/ipv4/netfilter/ip_nat_rule.c @@ -255,6 +255,27 @@ alloc_null_binding(struct ip_conntrack *conntrack, return ip_nat_setup_info(conntrack, &range, hooknum); } +unsigned int +alloc_null_binding_confirmed(struct ip_conntrack *conntrack, + struct ip_nat_info *info, + unsigned int hooknum) +{ + u_int32_t ip + = (HOOK2MANIP(hooknum) == IP_NAT_MANIP_SRC + ? conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip + : conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip); + u_int16_t all + = (HOOK2MANIP(hooknum) == IP_NAT_MANIP_SRC + ? conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.all + : conntrack->tuplehash[IP_CT_DIR_REPLY].tuple.src.u.all); + struct ip_nat_range range + = { IP_NAT_RANGE_MAP_IPS, ip, ip, { all }, { all } }; + + DEBUGP("Allocating NULL binding for confirmed %p (%u.%u.%u.%u)\n", + conntrack, NIPQUAD(ip)); + return ip_nat_setup_info(conntrack, &range, hooknum); +} + int ip_nat_rule_find(struct sk_buff **pskb, unsigned int hooknum, const struct net_device *in, diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c index 89db052add8..0ff368b131f 100644 --- a/net/ipv4/netfilter/ip_nat_standalone.c +++ b/net/ipv4/netfilter/ip_nat_standalone.c @@ -123,8 +123,12 @@ ip_nat_fn(unsigned int hooknum, if (!ip_nat_initialized(ct, maniptype)) { unsigned int ret; - /* LOCAL_IN hook doesn't have a chain! */ - if (hooknum == NF_IP_LOCAL_IN) + if (unlikely(is_confirmed(ct))) + /* NAT module was loaded late */ + ret = alloc_null_binding_confirmed(ct, info, + hooknum); + else if (hooknum == NF_IP_LOCAL_IN) + /* LOCAL_IN hook doesn't have a chain! */ ret = alloc_null_binding(ct, info, hooknum); else ret = ip_nat_rule_find(pskb, hooknum, diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c index 2d05cafec22..9bcb398fbc1 100644 --- a/net/ipv4/netfilter/ipt_CLUSTERIP.c +++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c @@ -13,6 +13,7 @@ #include <linux/config.h> #include <linux/proc_fs.h> #include <linux/jhash.h> +#include <linux/bitops.h> #include <linux/skbuff.h> #include <linux/ip.h> #include <linux/tcp.h> @@ -30,7 +31,7 @@ #include <linux/netfilter_ipv4/ipt_CLUSTERIP.h> #include <linux/netfilter_ipv4/ip_conntrack.h> -#define CLUSTERIP_VERSION "0.7" +#define CLUSTERIP_VERSION "0.8" #define DEBUG_CLUSTERIP @@ -49,13 +50,14 @@ MODULE_DESCRIPTION("iptables target for CLUSTERIP"); struct clusterip_config { struct list_head list; /* list of all configs */ atomic_t refcount; /* reference count */ + atomic_t entries; /* number of entries/rules + * referencing us */ u_int32_t clusterip; /* the IP address */ u_int8_t clustermac[ETH_ALEN]; /* the MAC address */ struct net_device *dev; /* device */ u_int16_t num_total_nodes; /* total number of nodes */ - u_int16_t num_local_nodes; /* number of local nodes */ - u_int16_t local_nodes[CLUSTERIP_MAX_NODES]; /* node number array */ + unsigned long local_nodes; /* node number array */ #ifdef CONFIG_PROC_FS struct proc_dir_entry *pde; /* proc dir entry */ @@ -66,8 +68,7 @@ struct clusterip_config { static LIST_HEAD(clusterip_configs); -/* clusterip_lock protects the clusterip_configs list _AND_ the configurable - * data within all structurses (num_local_nodes, local_nodes[]) */ +/* clusterip_lock protects the clusterip_configs list */ static DEFINE_RWLOCK(clusterip_lock); #ifdef CONFIG_PROC_FS @@ -76,23 +77,48 @@ static struct proc_dir_entry *clusterip_procdir; #endif static inline void -clusterip_config_get(struct clusterip_config *c) { +clusterip_config_get(struct clusterip_config *c) +{ atomic_inc(&c->refcount); } static inline void -clusterip_config_put(struct clusterip_config *c) { - if (atomic_dec_and_test(&c->refcount)) { +clusterip_config_put(struct clusterip_config *c) +{ + if (atomic_dec_and_test(&c->refcount)) + kfree(c); +} + +/* increase the count of entries(rules) using/referencing this config */ +static inline void +clusterip_config_entry_get(struct clusterip_config *c) +{ + atomic_inc(&c->entries); +} + +/* decrease the count of entries using/referencing this config. If last + * entry(rule) is removed, remove the config from lists, but don't free it + * yet, since proc-files could still be holding references */ +static inline void +clusterip_config_entry_put(struct clusterip_config *c) +{ + if (atomic_dec_and_test(&c->entries)) { write_lock_bh(&clusterip_lock); list_del(&c->list); write_unlock_bh(&clusterip_lock); + dev_mc_delete(c->dev, c->clustermac, ETH_ALEN, 0); dev_put(c->dev); - kfree(c); + + /* In case anyone still accesses the file, the open/close + * functions are also incrementing the refcount on their own, + * so it's safe to remove the entry even if it's in use. */ +#ifdef CONFIG_PROC_FS + remove_proc_entry(c->pde->name, c->pde->parent); +#endif } } - static struct clusterip_config * __clusterip_config_find(u_int32_t clusterip) { @@ -111,7 +137,7 @@ __clusterip_config_find(u_int32_t clusterip) } static inline struct clusterip_config * -clusterip_config_find_get(u_int32_t clusterip) +clusterip_config_find_get(u_int32_t clusterip, int entry) { struct clusterip_config *c; @@ -122,11 +148,24 @@ clusterip_config_find_get(u_int32_t clusterip) return NULL; } atomic_inc(&c->refcount); + if (entry) + atomic_inc(&c->entries); read_unlock_bh(&clusterip_lock); return c; } +static void +clusterip_config_init_nodelist(struct clusterip_config *c, + const struct ipt_clusterip_tgt_info *i) +{ + int n; + + for (n = 0; n < i->num_local_nodes; n++) { + set_bit(i->local_nodes[n] - 1, &c->local_nodes); + } +} + static struct clusterip_config * clusterip_config_init(struct ipt_clusterip_tgt_info *i, u_int32_t ip, struct net_device *dev) @@ -143,11 +182,11 @@ clusterip_config_init(struct ipt_clusterip_tgt_info *i, u_int32_t ip, c->clusterip = ip; memcpy(&c->clustermac, &i->clustermac, ETH_ALEN); c->num_total_nodes = i->num_total_nodes; - c->num_local_nodes = i->num_local_nodes; - memcpy(&c->local_nodes, &i->local_nodes, sizeof(&c->local_nodes)); + clusterip_config_init_nodelist(c, i); c->hash_mode = i->hash_mode; c->hash_initval = i->hash_initval; atomic_set(&c->refcount, 1); + atomic_set(&c->entries, 1); #ifdef CONFIG_PROC_FS /* create proc dir entry */ @@ -171,53 +210,28 @@ clusterip_config_init(struct ipt_clusterip_tgt_info *i, u_int32_t ip, static int clusterip_add_node(struct clusterip_config *c, u_int16_t nodenum) { - int i; - - write_lock_bh(&clusterip_lock); - if (c->num_local_nodes >= CLUSTERIP_MAX_NODES - || nodenum > CLUSTERIP_MAX_NODES) { - write_unlock_bh(&clusterip_lock); + if (nodenum == 0 || + nodenum > c->num_total_nodes) return 1; - } - - /* check if we alrady have this number in our array */ - for (i = 0; i < c->num_local_nodes; i++) { - if (c->local_nodes[i] == nodenum) { - write_unlock_bh(&clusterip_lock); - return 1; - } - } - c->local_nodes[c->num_local_nodes++] = nodenum; + /* check if we already have this number in our bitfield */ + if (test_and_set_bit(nodenum - 1, &c->local_nodes)) + return 1; - write_unlock_bh(&clusterip_lock); return 0; } static int clusterip_del_node(struct clusterip_config *c, u_int16_t nodenum) { - int i; - - write_lock_bh(&clusterip_lock); - - if (c->num_local_nodes <= 1 || nodenum > CLUSTERIP_MAX_NODES) { - write_unlock_bh(&clusterip_lock); + if (nodenum == 0 || + nodenum > c->num_total_nodes) return 1; - } - for (i = 0; i < c->num_local_nodes; i++) { - if (c->local_nodes[i] == nodenum) { - int size = sizeof(u_int16_t)*(c->num_local_nodes-(i+1)); - memmove(&c->local_nodes[i], &c->local_nodes[i+1], size); - c->num_local_nodes--; - write_unlock_bh(&clusterip_lock); - return 0; - } - } + if (test_and_clear_bit(nodenum - 1, &c->local_nodes)) + return 0; - write_unlock_bh(&clusterip_lock); return 1; } @@ -285,25 +299,7 @@ clusterip_hashfn(struct sk_buff *skb, struct clusterip_config *config) static inline int clusterip_responsible(struct clusterip_config *config, u_int32_t hash) { - int i; - - read_lock_bh(&clusterip_lock); - - if (config->num_local_nodes == 0) { - read_unlock_bh(&clusterip_lock); - return 0; - } - - for (i = 0; i < config->num_local_nodes; i++) { - if (config->local_nodes[i] == hash) { - read_unlock_bh(&clusterip_lock); - return 1; - } - } - - read_unlock_bh(&clusterip_lock); - - return 0; + return test_bit(hash - 1, &config->local_nodes); } /*********************************************************************** @@ -415,8 +411,26 @@ checkentry(const char *tablename, /* FIXME: further sanity checks */ - config = clusterip_config_find_get(e->ip.dst.s_addr); - if (!config) { + config = clusterip_config_find_get(e->ip.dst.s_addr, 1); + if (config) { + if (cipinfo->config != NULL) { + /* Case A: This is an entry that gets reloaded, since + * it still has a cipinfo->config pointer. Simply + * increase the entry refcount and return */ + if (cipinfo->config != config) { + printk(KERN_ERR "CLUSTERIP: Reloaded entry " + "has invalid config pointer!\n"); + return 0; + } + clusterip_config_entry_get(cipinfo->config); + } else { + /* Case B: This is a new rule referring to an existing + * clusterip config. */ + cipinfo->config = config; + clusterip_config_entry_get(cipinfo->config); + } + } else { + /* Case C: This is a completely new clusterip config */ if (!(cipinfo->flags & CLUSTERIP_FLAG_NEW)) { printk(KERN_WARNING "CLUSTERIP: no config found for %u.%u.%u.%u, need 'new'\n", NIPQUAD(e->ip.dst.s_addr)); return 0; @@ -443,10 +457,9 @@ checkentry(const char *tablename, } dev_mc_add(config->dev,config->clustermac, ETH_ALEN, 0); } + cipinfo->config = config; } - cipinfo->config = config; - return 1; } @@ -455,13 +468,10 @@ static void destroy(void *matchinfo, unsigned int matchinfosize) { struct ipt_clusterip_tgt_info *cipinfo = matchinfo; - /* we first remove the proc entry and then drop the reference - * count. In case anyone still accesses the file, the open/close - * functions are also incrementing the refcount on their own */ -#ifdef CONFIG_PROC_FS - remove_proc_entry(cipinfo->config->pde->name, - cipinfo->config->pde->parent); -#endif + /* if no more entries are referencing the config, remove it + * from the list and destroy the proc entry */ + clusterip_config_entry_put(cipinfo->config); + clusterip_config_put(cipinfo->config); } @@ -533,7 +543,7 @@ arp_mangle(unsigned int hook, /* if there is no clusterip configuration for the arp reply's * source ip, we don't want to mangle it */ - c = clusterip_config_find_get(payload->src_ip); + c = clusterip_config_find_get(payload->src_ip, 0); if (!c) return NF_ACCEPT; @@ -574,56 +584,69 @@ static struct nf_hook_ops cip_arp_ops = { #ifdef CONFIG_PROC_FS +struct clusterip_seq_position { + unsigned int pos; /* position */ + unsigned int weight; /* number of bits set == size */ + unsigned int bit; /* current bit */ + unsigned long val; /* current value */ +}; + static void *clusterip_seq_start(struct seq_file *s, loff_t *pos) { struct proc_dir_entry *pde = s->private; struct clusterip_config *c = pde->data; - unsigned int *nodeidx; - - read_lock_bh(&clusterip_lock); - if (*pos >= c->num_local_nodes) + unsigned int weight; + u_int32_t local_nodes; + struct clusterip_seq_position *idx; + + /* FIXME: possible race */ + local_nodes = c->local_nodes; + weight = hweight32(local_nodes); + if (*pos >= weight) return NULL; - nodeidx = kmalloc(sizeof(unsigned int), GFP_KERNEL); - if (!nodeidx) + idx = kmalloc(sizeof(struct clusterip_seq_position), GFP_KERNEL); + if (!idx) return ERR_PTR(-ENOMEM); - *nodeidx = *pos; - return nodeidx; + idx->pos = *pos; + idx->weight = weight; + idx->bit = ffs(local_nodes); + idx->val = local_nodes; + clear_bit(idx->bit - 1, &idx->val); + + return idx; } static void *clusterip_seq_next(struct seq_file *s, void *v, loff_t *pos) { - struct proc_dir_entry *pde = s->private; - struct clusterip_config *c = pde->data; - unsigned int *nodeidx = (unsigned int *)v; + struct clusterip_seq_position *idx = (struct clusterip_seq_position *)v; - *pos = ++(*nodeidx); - if (*pos >= c->num_local_nodes) { + *pos = ++idx->pos; + if (*pos >= idx->weight) { kfree(v); return NULL; } - return nodeidx; + idx->bit = ffs(idx->val); + clear_bit(idx->bit - 1, &idx->val); + return idx; } static void clusterip_seq_stop(struct seq_file *s, void *v) { kfree(v); - - read_unlock_bh(&clusterip_lock); } static int clusterip_seq_show(struct seq_file *s, void *v) { - struct proc_dir_entry *pde = s->private; - struct clusterip_config *c = pde->data; - unsigned int *nodeidx = (unsigned int *)v; + struct clusterip_seq_position *idx = (struct clusterip_seq_position *)v; - if (*nodeidx != 0) + if (idx->pos != 0) seq_putc(s, ','); - seq_printf(s, "%u", c->local_nodes[*nodeidx]); - if (*nodeidx == c->num_local_nodes-1) + seq_printf(s, "%u", idx->bit); + + if (idx->pos == idx->weight - 1) seq_putc(s, '\n'); return 0; diff --git a/net/ipv4/netfilter/ipt_MASQUERADE.c b/net/ipv4/netfilter/ipt_MASQUERADE.c index 2f3e181c8e9..275a174c6fe 100644 --- a/net/ipv4/netfilter/ipt_MASQUERADE.c +++ b/net/ipv4/netfilter/ipt_MASQUERADE.c @@ -90,6 +90,12 @@ masquerade_target(struct sk_buff **pskb, IP_NF_ASSERT(ct && (ctinfo == IP_CT_NEW || ctinfo == IP_CT_RELATED || ctinfo == IP_CT_RELATED + IP_CT_IS_REPLY)); + /* Source address is 0.0.0.0 - locally generated packet that is + * probably not supposed to be masqueraded. + */ + if (ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip == 0) + return NF_ACCEPT; + mr = targinfo; rt = (struct rtable *)(*pskb)->dst; newsrc = inet_select_addr(out, rt->rt_gateway, RT_SCOPE_UNIVERSE); diff --git a/net/ipv4/netfilter/ipt_REDIRECT.c b/net/ipv4/netfilter/ipt_REDIRECT.c index d2e13447678..715cb613405 100644 --- a/net/ipv4/netfilter/ipt_REDIRECT.c +++ b/net/ipv4/netfilter/ipt_REDIRECT.c @@ -88,14 +88,18 @@ redirect_target(struct sk_buff **pskb, newdst = htonl(0x7F000001); else { struct in_device *indev; + struct in_ifaddr *ifa; - /* Device might not have an associated in_device. */ - indev = (struct in_device *)(*pskb)->dev->ip_ptr; - if (indev == NULL || indev->ifa_list == NULL) - return NF_DROP; + newdst = 0; + + rcu_read_lock(); + indev = __in_dev_get((*pskb)->dev); + if (indev && (ifa = indev->ifa_list)) + newdst = ifa->ifa_local; + rcu_read_unlock(); - /* Grab first address on interface. */ - newdst = indev->ifa_list->ifa_local; + if (!newdst) + return NF_DROP; } /* Transfer from original range. */ diff --git a/net/ipv4/netfilter/ipt_REJECT.c b/net/ipv4/netfilter/ipt_REJECT.c index f115a84a4ac..f057025a719 100644 --- a/net/ipv4/netfilter/ipt_REJECT.c +++ b/net/ipv4/netfilter/ipt_REJECT.c @@ -92,10 +92,7 @@ static inline struct rtable *route_reverse(struct sk_buff *skb, fl.fl_ip_sport = tcph->dest; fl.fl_ip_dport = tcph->source; - if (xfrm_lookup((struct dst_entry **)&rt, &fl, NULL, 0)) { - dst_release(&rt->u.dst); - rt = NULL; - } + xfrm_lookup((struct dst_entry **)&rt, &fl, NULL, 0); return rt; } diff --git a/net/ipv4/netfilter/ipt_owner.c b/net/ipv4/netfilter/ipt_owner.c index c1889f88262..0cee2862ed8 100644 --- a/net/ipv4/netfilter/ipt_owner.c +++ b/net/ipv4/netfilter/ipt_owner.c @@ -11,6 +11,7 @@ #include <linux/module.h> #include <linux/skbuff.h> #include <linux/file.h> +#include <linux/rcupdate.h> #include <net/sock.h> #include <linux/netfilter_ipv4/ipt_owner.h> diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 304bb0a1d4f..4b0d7e4d626 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -361,7 +361,7 @@ static void raw_probe_proto_opt(struct flowi *fl, struct msghdr *msg) if (type && code) { get_user(fl->fl_icmp_type, type); - __get_user(fl->fl_icmp_code, code); + get_user(fl->fl_icmp_code, code); probed = 1; } break; diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 8c0b14e3bee..8549f26e249 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1760,6 +1760,7 @@ static inline int __mkroute_input(struct sk_buff *skb, goto cleanup; } + atomic_set(&rth->u.dst.__refcnt, 1); rth->u.dst.flags= DST_HOST; #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED if (res->fi->fib_nhs > 1) @@ -1820,7 +1821,6 @@ static inline int ip_mkroute_input_def(struct sk_buff *skb, err = __mkroute_input(skb, res, in_dev, daddr, saddr, tos, &rth); if (err) return err; - atomic_set(&rth->u.dst.__refcnt, 1); /* put it into the cache */ hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5), tos); @@ -1834,8 +1834,8 @@ static inline int ip_mkroute_input(struct sk_buff *skb, u32 daddr, u32 saddr, u32 tos) { #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED - struct rtable* rth = NULL; - unsigned char hop, hopcount, lasthop; + struct rtable* rth = NULL, *rtres; + unsigned char hop, hopcount; int err = -EINVAL; unsigned int hash; @@ -1844,8 +1844,6 @@ static inline int ip_mkroute_input(struct sk_buff *skb, else hopcount = 1; - lasthop = hopcount - 1; - /* distinguish between multipath and singlepath */ if (hopcount < 2) return ip_mkroute_input_def(skb, res, fl, in_dev, daddr, @@ -1855,6 +1853,10 @@ static inline int ip_mkroute_input(struct sk_buff *skb, for (hop = 0; hop < hopcount; hop++) { res->nh_sel = hop; + /* put reference to previous result */ + if (hop) + ip_rt_put(rtres); + /* create a routing cache entry */ err = __mkroute_input(skb, res, in_dev, daddr, saddr, tos, &rth); @@ -1863,7 +1865,7 @@ static inline int ip_mkroute_input(struct sk_buff *skb, /* put it into the cache */ hash = rt_hash_code(daddr, saddr ^ (fl->iif << 5), tos); - err = rt_intern_hash(hash, rth, (struct rtable**)&skb->dst); + err = rt_intern_hash(hash, rth, &rtres); if (err) return err; @@ -1873,13 +1875,8 @@ static inline int ip_mkroute_input(struct sk_buff *skb, FIB_RES_NETMASK(*res), res->prefixlen, &FIB_RES_NH(*res)); - - /* only for the last hop the reference count is handled - * outside - */ - if (hop == lasthop) - atomic_set(&(skb->dst->__refcnt), 1); } + skb->dst = &rtres->u.dst; return err; #else /* CONFIG_IP_ROUTE_MULTIPATH_CACHED */ return ip_mkroute_input_def(skb, res, fl, in_dev, daddr, saddr, tos); @@ -2208,6 +2205,7 @@ static inline int __mkroute_output(struct rtable **result, goto cleanup; } + atomic_set(&rth->u.dst.__refcnt, 1); rth->u.dst.flags= DST_HOST; #ifdef CONFIG_IP_ROUTE_MULTIPATH_CACHED if (res->fi) { @@ -2290,8 +2288,6 @@ static inline int ip_mkroute_output_def(struct rtable **rp, if (err == 0) { u32 tos = RT_FL_TOS(oldflp); - atomic_set(&rth->u.dst.__refcnt, 1); - hash = rt_hash_code(oldflp->fl4_dst, oldflp->fl4_src ^ (oldflp->oif << 5), tos); err = rt_intern_hash(hash, rth, rp); @@ -2326,6 +2322,10 @@ static inline int ip_mkroute_output(struct rtable** rp, dev2nexthop = FIB_RES_DEV(*res); dev_hold(dev2nexthop); + /* put reference to previous result */ + if (hop) + ip_rt_put(*rp); + err = __mkroute_output(&rth, res, fl, oldflp, dev2nexthop, flags); @@ -2350,7 +2350,6 @@ static inline int ip_mkroute_output(struct rtable** rp, if (err != 0) return err; } - atomic_set(&(*rp)->u.dst.__refcnt, 1); return err; } else { return ip_mkroute_output_def(rp, res, fl, oldflp, dev_out, diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 02fdda68718..f3f0013a958 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -552,8 +552,7 @@ new_segment: tcp_mark_push(tp, skb); goto new_segment; } - if (sk->sk_forward_alloc < copy && - !sk_stream_mem_schedule(sk, copy, 0)) + if (!sk_stream_wmem_schedule(sk, copy)) goto wait_for_memory; if (can_coalesce) { @@ -770,19 +769,23 @@ new_segment: if (off == PAGE_SIZE) { put_page(page); TCP_PAGE(sk) = page = NULL; + off = 0; } - } + } else + off = 0; + + if (copy > PAGE_SIZE - off) + copy = PAGE_SIZE - off; + + if (!sk_stream_wmem_schedule(sk, copy)) + goto wait_for_memory; if (!page) { /* Allocate new cache page. */ if (!(page = sk_stream_alloc_page(sk))) goto wait_for_memory; - off = 0; } - if (copy > PAGE_SIZE - off) - copy = PAGE_SIZE - off; - /* Time to copy data. We are close to * the end! */ err = skb_copy_to_page(sk, from, skb, page, diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 1afb080bdf0..a7537c7bbd0 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -923,14 +923,6 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ int flag = 0; int i; - /* So, SACKs for already sent large segments will be lost. - * Not good, but alternative is to resegment the queue. */ - if (sk->sk_route_caps & NETIF_F_TSO) { - sk->sk_route_caps &= ~NETIF_F_TSO; - sock_set_flag(sk, SOCK_NO_LARGESEND); - tp->mss_cache = tp->mss_cache; - } - if (!tp->sacked_out) tp->fackets_out = 0; prior_fackets = tp->fackets_out; @@ -978,20 +970,42 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ flag |= FLAG_DATA_LOST; sk_stream_for_retrans_queue(skb, sk) { - u8 sacked = TCP_SKB_CB(skb)->sacked; - int in_sack; + int in_sack, pcount; + u8 sacked; /* The retransmission queue is always in order, so * we can short-circuit the walk early. */ - if(!before(TCP_SKB_CB(skb)->seq, end_seq)) + if (!before(TCP_SKB_CB(skb)->seq, end_seq)) break; - fack_count += tcp_skb_pcount(skb); - in_sack = !after(start_seq, TCP_SKB_CB(skb)->seq) && !before(end_seq, TCP_SKB_CB(skb)->end_seq); + pcount = tcp_skb_pcount(skb); + + if (pcount > 1 && !in_sack && + after(TCP_SKB_CB(skb)->end_seq, start_seq)) { + unsigned int pkt_len; + + in_sack = !after(start_seq, + TCP_SKB_CB(skb)->seq); + + if (!in_sack) + pkt_len = (start_seq - + TCP_SKB_CB(skb)->seq); + else + pkt_len = (end_seq - + TCP_SKB_CB(skb)->seq); + if (tcp_fragment(sk, skb, pkt_len, skb_shinfo(skb)->tso_size)) + break; + pcount = tcp_skb_pcount(skb); + } + + fack_count += pcount; + + sacked = TCP_SKB_CB(skb)->sacked; + /* Account D-SACK for retransmitted packet. */ if ((dup_sack && in_sack) && (sacked & TCPCB_RETRANS) && diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index a88db28b0af..b1a63b2c6b4 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -384,7 +384,7 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, newtp->frto_counter = 0; newtp->frto_highmark = 0; - newicsk->icsk_ca_ops = &tcp_reno; + newicsk->icsk_ca_ops = &tcp_init_congestion_ops; tcp_set_ca_state(newsk, TCP_CA_Open); tcp_init_xmit_timers(newsk); diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 75b68116682..d6e3d269e90 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -428,13 +428,15 @@ static void tcp_set_skb_tso_segs(struct sock *sk, struct sk_buff *skb, unsigned * packet to the list. This won't be called frequently, I hope. * Remember, these are still headerless SKBs at this point. */ -static int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss_now) +int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss_now) { struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *buff; - int nsize; + int nsize, old_factor; u16 flags; + BUG_ON(len >= skb->len); + nsize = skb_headlen(skb) - len; if (nsize < 0) nsize = 0; @@ -459,9 +461,7 @@ static int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned flags = TCP_SKB_CB(skb)->flags; TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH); TCP_SKB_CB(buff)->flags = flags; - TCP_SKB_CB(buff)->sacked = - (TCP_SKB_CB(skb)->sacked & - (TCPCB_LOST | TCPCB_EVER_RETRANS | TCPCB_AT_TAIL)); + TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked; TCP_SKB_CB(skb)->sacked &= ~TCPCB_AT_TAIL; if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_HW) { @@ -485,23 +485,44 @@ static int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned TCP_SKB_CB(buff)->when = TCP_SKB_CB(skb)->when; buff->tstamp = skb->tstamp; - if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) { - tp->lost_out -= tcp_skb_pcount(skb); - tp->left_out -= tcp_skb_pcount(skb); - } + old_factor = tcp_skb_pcount(skb); /* Fix up tso_factor for both original and new SKB. */ tcp_set_skb_tso_segs(sk, skb, mss_now); tcp_set_skb_tso_segs(sk, buff, mss_now); - if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) { - tp->lost_out += tcp_skb_pcount(skb); - tp->left_out += tcp_skb_pcount(skb); - } + /* If this packet has been sent out already, we must + * adjust the various packet counters. + */ + if (!before(tp->snd_nxt, TCP_SKB_CB(buff)->end_seq)) { + int diff = old_factor - tcp_skb_pcount(skb) - + tcp_skb_pcount(buff); - if (TCP_SKB_CB(buff)->sacked&TCPCB_LOST) { - tp->lost_out += tcp_skb_pcount(buff); - tp->left_out += tcp_skb_pcount(buff); + tp->packets_out -= diff; + + if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED) + tp->sacked_out -= diff; + if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS) + tp->retrans_out -= diff; + + if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) { + tp->lost_out -= diff; + tp->left_out -= diff; + } + + if (diff > 0) { + /* Adjust Reno SACK estimate. */ + if (!tp->rx_opt.sack_ok) { + tp->sacked_out -= diff; + if ((int)tp->sacked_out < 0) + tp->sacked_out = 0; + tcp_sync_left_out(tp); + } + + tp->fackets_out -= diff; + if ((int)tp->fackets_out < 0) + tp->fackets_out = 0; + } } /* Link BUFF into the send queue. */ @@ -1350,12 +1371,6 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb) if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) { if (before(TCP_SKB_CB(skb)->end_seq, tp->snd_una)) BUG(); - - if (sk->sk_route_caps & NETIF_F_TSO) { - sk->sk_route_caps &= ~NETIF_F_TSO; - sock_set_flag(sk, SOCK_NO_LARGESEND); - } - if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq)) return -ENOMEM; } @@ -1370,22 +1385,8 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb) return -EAGAIN; if (skb->len > cur_mss) { - int old_factor = tcp_skb_pcount(skb); - int diff; - if (tcp_fragment(sk, skb, cur_mss, cur_mss)) return -ENOMEM; /* We'll try again later. */ - - /* New SKB created, account for it. */ - diff = old_factor - tcp_skb_pcount(skb) - - tcp_skb_pcount(skb->next); - tp->packets_out -= diff; - - if (diff > 0) { - tp->fackets_out -= diff; - if ((int)tp->fackets_out < 0) - tp->fackets_out = 0; - } } /* Collapse two adjacent packets if worthwhile and we can. */ @@ -1993,12 +1994,6 @@ int tcp_write_wakeup(struct sock *sk) TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_PSH; if (tcp_fragment(sk, skb, seg_size, mss)) return -1; - /* SWS override triggered forced fragmentation. - * Disable TSO, the connection is too sick. */ - if (sk->sk_route_caps & NETIF_F_TSO) { - sock_set_flag(sk, SOCK_NO_LARGESEND); - sk->sk_route_caps &= ~NETIF_F_TSO; - } } else if (!tcp_skb_pcount(skb)) tcp_set_skb_tso_segs(sk, skb, mss); diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index e5beca7de86..e0bd1013cb0 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1141,7 +1141,7 @@ int udp_rcv(struct sk_buff *skb) if (ulen > len || ulen < sizeof(*uh)) goto short_packet; - if (pskb_trim(skb, ulen)) + if (pskb_trim_rcsum(skb, ulen)) goto short_packet; if (udp_checksum_init(skb, uh, ulen, saddr, daddr) < 0) |