From e682adf021be796940be6cc10c07be7f7398c220 Mon Sep 17 00:00:00 2001 From: Fan Du Date: Thu, 7 Nov 2013 17:47:48 +0800 Subject: xfrm: Try to honor policy index if it's supplied by user xfrm code always searches for unused policy index for newly created policy regardless whether or not user space policy index hint supplied. This patch enables such feature so that using "ip xfrm ... index=xxx" can be used by user to set specific policy index. Currently this beahvior is broken, so this patch make it happen as expected. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_policy.c | 16 +++++++++++----- net/xfrm/xfrm_user.c | 10 +++++++++- 2 files changed, 20 insertions(+), 6 deletions(-) diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 9a91f74..a5bbdfb 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -538,7 +538,7 @@ static void xfrm_hash_resize(struct work_struct *work) /* Generate new index... KAME seems to generate them ordered by cost * of an absolute inpredictability of ordering of rules. This will not pass. */ -static u32 xfrm_gen_index(struct net *net, int dir) +static u32 xfrm_gen_index(struct net *net, int dir, u32 index) { static u32 idx_generator; @@ -548,8 +548,14 @@ static u32 xfrm_gen_index(struct net *net, int dir) u32 idx; int found; - idx = (idx_generator | dir); - idx_generator += 8; + if (!index) { + idx = (idx_generator | dir); + idx_generator += 8; + } else { + idx = index; + index = 0; + } + if (idx == 0) idx = 8; list = net->xfrm.policy_byidx + idx_hash(net, idx); @@ -672,7 +678,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl) xfrm_policy_requeue(delpol, policy); __xfrm_policy_unlink(delpol, dir); } - policy->index = delpol ? delpol->index : xfrm_gen_index(net, dir); + policy->index = delpol ? delpol->index : xfrm_gen_index(net, dir, policy->index); hlist_add_head(&policy->byidx, net->xfrm.policy_byidx+idx_hash(net, policy->index)); policy->curlft.add_time = get_seconds(); policy->curlft.use_time = 0; @@ -1192,7 +1198,7 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol) sk->sk_policy[dir] = pol; if (pol) { pol->curlft.add_time = get_seconds(); - pol->index = xfrm_gen_index(net, XFRM_POLICY_MAX+dir); + pol->index = xfrm_gen_index(net, XFRM_POLICY_MAX+dir, 0); __xfrm_policy_link(pol, XFRM_POLICY_MAX+dir); } if (old_pol) { diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index f964d4c..4e0546e 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -1189,6 +1189,8 @@ static int verify_policy_type(u8 type) static int verify_newpolicy_info(struct xfrm_userpolicy_info *p) { + int ret; + switch (p->share) { case XFRM_SHARE_ANY: case XFRM_SHARE_SESSION: @@ -1224,7 +1226,13 @@ static int verify_newpolicy_info(struct xfrm_userpolicy_info *p) return -EINVAL; } - return verify_policy_dir(p->dir); + ret = verify_policy_dir(p->dir); + if (ret) + return ret; + if (p->index && ((p->index & XFRM_POLICY_MAX) != p->dir)) + return -EINVAL; + + return 0; } static int copy_from_user_sec_ctx(struct xfrm_policy *pol, struct nlattr **attrs) -- cgit v1.1 From 8d549c4f5d92d80fc6f888fd314e10972ae0ec37 Mon Sep 17 00:00:00 2001 From: Fan Du Date: Thu, 7 Nov 2013 17:47:49 +0800 Subject: xfrm: Using the right namespace to migrate key info because the home agent could surely be run on a different net namespace other than init_net. The original behavior could lead into inconsistent of key info. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- include/net/xfrm.h | 2 +- net/key/af_key.c | 3 ++- net/xfrm/xfrm_policy.c | 10 +++++----- net/xfrm/xfrm_user.c | 3 ++- 4 files changed, 10 insertions(+), 8 deletions(-) diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 6b82fdf..5b522c5 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -1581,7 +1581,7 @@ struct xfrm_state *xfrm_state_migrate(struct xfrm_state *x, struct xfrm_migrate *m); int xfrm_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, struct xfrm_migrate *m, int num_bundles, - struct xfrm_kmaddress *k); + struct xfrm_kmaddress *k, struct net *net); #endif int km_new_mapping(struct xfrm_state *x, xfrm_address_t *ipaddr, __be16 sport); diff --git a/net/key/af_key.c b/net/key/af_key.c index 545f047..3fa811c 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -2485,6 +2485,7 @@ static int pfkey_migrate(struct sock *sk, struct sk_buff *skb, struct xfrm_selector sel; struct xfrm_migrate m[XFRM_MAX_DEPTH]; struct xfrm_kmaddress k; + struct net *net = sock_net(sk); if (!present_and_same_family(ext_hdrs[SADB_EXT_ADDRESS_SRC - 1], ext_hdrs[SADB_EXT_ADDRESS_DST - 1]) || @@ -2558,7 +2559,7 @@ static int pfkey_migrate(struct sock *sk, struct sk_buff *skb, } return xfrm_migrate(&sel, dir, XFRM_POLICY_TYPE_MAIN, m, i, - kma ? &k : NULL); + kma ? &k : NULL, net); out: return err; diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index a5bbdfb..907fd2f 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -3076,14 +3076,14 @@ static bool xfrm_migrate_selector_match(const struct xfrm_selector *sel_cmp, } static struct xfrm_policy * xfrm_migrate_policy_find(const struct xfrm_selector *sel, - u8 dir, u8 type) + u8 dir, u8 type, struct net *net) { struct xfrm_policy *pol, *ret = NULL; struct hlist_head *chain; u32 priority = ~0U; read_lock_bh(&xfrm_policy_lock); - chain = policy_hash_direct(&init_net, &sel->daddr, &sel->saddr, sel->family, dir); + chain = policy_hash_direct(net, &sel->daddr, &sel->saddr, sel->family, dir); hlist_for_each_entry(pol, chain, bydst) { if (xfrm_migrate_selector_match(sel, &pol->selector) && pol->type == type) { @@ -3092,7 +3092,7 @@ static struct xfrm_policy * xfrm_migrate_policy_find(const struct xfrm_selector break; } } - chain = &init_net.xfrm.policy_inexact[dir]; + chain = &net->xfrm.policy_inexact[dir]; hlist_for_each_entry(pol, chain, bydst) { if (xfrm_migrate_selector_match(sel, &pol->selector) && pol->type == type && @@ -3216,7 +3216,7 @@ static int xfrm_migrate_check(const struct xfrm_migrate *m, int num_migrate) int xfrm_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, struct xfrm_migrate *m, int num_migrate, - struct xfrm_kmaddress *k) + struct xfrm_kmaddress *k, struct net *net) { int i, err, nx_cur = 0, nx_new = 0; struct xfrm_policy *pol = NULL; @@ -3229,7 +3229,7 @@ int xfrm_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, goto out; /* Stage 1 - find policy */ - if ((pol = xfrm_migrate_policy_find(sel, dir, type)) == NULL) { + if ((pol = xfrm_migrate_policy_find(sel, dir, type, net)) == NULL) { err = -ENOENT; goto out; } diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 4e0546e..840cc8d 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -2137,6 +2137,7 @@ static int xfrm_do_migrate(struct sk_buff *skb, struct nlmsghdr *nlh, u8 type; int err; int n = 0; + struct net *net = sock_net(skb->sk); if (attrs[XFRMA_MIGRATE] == NULL) return -EINVAL; @@ -2154,7 +2155,7 @@ static int xfrm_do_migrate(struct sk_buff *skb, struct nlmsghdr *nlh, if (!n) return 0; - xfrm_migrate(&pi->sel, pi->dir, type, m, n, kmp); + xfrm_migrate(&pi->sel, pi->dir, type, m, n, kmp, net); return 0; } -- cgit v1.1 From 283bc9f35bbbcb0e9ab4e6d2427da7f9f710d52d Mon Sep 17 00:00:00 2001 From: Fan Du Date: Thu, 7 Nov 2013 17:47:50 +0800 Subject: xfrm: Namespacify xfrm state/policy locks By semantics, xfrm layer is fully name space aware, so will the locks, e.g. xfrm_state/pocliy_lock. Ensure exclusive access into state/policy link list for different name space with one global lock is not right in terms of semantics aspect at first place, as they are indeed mutually independent with each other, but also more seriously causes scalability problem. One practical scenario is on a Open Network Stack, more than hundreds of lxc tenants acts as routers within one host, a global xfrm_state/policy_lock becomes the bottleneck. But onces those locks are decoupled in a per-namespace fashion, locks contend is just with in specific name space scope, without causing additional SPD/SAD access delay for other name space. Also this patch improve scalability while as without changing original xfrm behavior. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- include/net/netns/xfrm.h | 4 ++ include/net/xfrm.h | 11 +++-- net/key/af_key.c | 15 ++++--- net/xfrm/xfrm_policy.c | 104 +++++++++++++++++++++++++---------------------- net/xfrm/xfrm_state.c | 104 ++++++++++++++++++++++++----------------------- net/xfrm/xfrm_user.c | 14 +++++-- 6 files changed, 138 insertions(+), 114 deletions(-) diff --git a/include/net/netns/xfrm.h b/include/net/netns/xfrm.h index 5299e69..ea28404 100644 --- a/include/net/netns/xfrm.h +++ b/include/net/netns/xfrm.h @@ -59,6 +59,10 @@ struct netns_xfrm { #if IS_ENABLED(CONFIG_IPV6) struct dst_ops xfrm6_dst_ops; #endif + spinlock_t xfrm_state_lock; + spinlock_t xfrm_policy_sk_bundle_lock; + rwlock_t xfrm_policy_lock; + struct mutex xfrm_cfg_mutex; }; #endif diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 5b522c5..59f5d0a 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -53,7 +53,6 @@ #define XFRM_INC_STATS_USER(net, field) ((void)(net)) #endif -extern struct mutex xfrm_cfg_mutex; /* Organization of SPD aka "XFRM rules" ------------------------------------ @@ -1409,7 +1408,7 @@ static inline void xfrm_sysctl_fini(struct net *net) void xfrm_state_walk_init(struct xfrm_state_walk *walk, u8 proto); int xfrm_state_walk(struct net *net, struct xfrm_state_walk *walk, int (*func)(struct xfrm_state *, int, void*), void *); -void xfrm_state_walk_done(struct xfrm_state_walk *walk); +void xfrm_state_walk_done(struct xfrm_state_walk *walk, struct net *net); struct xfrm_state *xfrm_state_alloc(struct net *net); struct xfrm_state *xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr, @@ -1436,12 +1435,12 @@ struct xfrm_state *xfrm_state_lookup_byaddr(struct net *net, u32 mark, unsigned short family); #ifdef CONFIG_XFRM_SUB_POLICY int xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n, - unsigned short family); + unsigned short family, struct net *net); int xfrm_state_sort(struct xfrm_state **dst, struct xfrm_state **src, int n, unsigned short family); #else static inline int xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, - int n, unsigned short family) + int n, unsigned short family, struct net *net) { return -ENOSYS; } @@ -1553,7 +1552,7 @@ void xfrm_policy_walk_init(struct xfrm_policy_walk *walk, u8 type); int xfrm_policy_walk(struct net *net, struct xfrm_policy_walk *walk, int (*func)(struct xfrm_policy *, int, int, void*), void *); -void xfrm_policy_walk_done(struct xfrm_policy_walk *walk); +void xfrm_policy_walk_done(struct xfrm_policy_walk *walk, struct net *net); int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl); struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type, int dir, @@ -1576,7 +1575,7 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol); int km_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, const struct xfrm_migrate *m, int num_bundles, const struct xfrm_kmaddress *k); -struct xfrm_state *xfrm_migrate_state_find(struct xfrm_migrate *m); +struct xfrm_state *xfrm_migrate_state_find(struct xfrm_migrate *m, struct net *net); struct xfrm_state *xfrm_state_migrate(struct xfrm_state *x, struct xfrm_migrate *m); int xfrm_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, diff --git a/net/key/af_key.c b/net/key/af_key.c index 3fa811c..9a039ac 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -1785,7 +1785,9 @@ static int pfkey_dump_sa(struct pfkey_sock *pfk) static void pfkey_dump_sa_done(struct pfkey_sock *pfk) { - xfrm_state_walk_done(&pfk->dump.u.state); + struct net *net = sock_net(&pfk->sk); + + xfrm_state_walk_done(&pfk->dump.u.state, net); } static int pfkey_dump(struct sock *sk, struct sk_buff *skb, const struct sadb_msg *hdr, void * const *ext_hdrs) @@ -1861,7 +1863,7 @@ static u32 gen_reqid(struct net *net) reqid = IPSEC_MANUAL_REQID_MAX+1; xfrm_policy_walk_init(&walk, XFRM_POLICY_TYPE_MAIN); rc = xfrm_policy_walk(net, &walk, check_reqid, (void*)&reqid); - xfrm_policy_walk_done(&walk); + xfrm_policy_walk_done(&walk, net); if (rc != -EEXIST) return reqid; } while (reqid != start); @@ -2660,7 +2662,9 @@ static int pfkey_dump_sp(struct pfkey_sock *pfk) static void pfkey_dump_sp_done(struct pfkey_sock *pfk) { - xfrm_policy_walk_done(&pfk->dump.u.policy); + struct net *net = sock_net((struct sock *)pfk); + + xfrm_policy_walk_done(&pfk->dump.u.policy, net); } static int pfkey_spddump(struct sock *sk, struct sk_buff *skb, const struct sadb_msg *hdr, void * const *ext_hdrs) @@ -3570,6 +3574,7 @@ static int pfkey_sendmsg(struct kiocb *kiocb, struct sk_buff *skb = NULL; struct sadb_msg *hdr = NULL; int err; + struct net *net = sock_net(sk); err = -EOPNOTSUPP; if (msg->msg_flags & MSG_OOB) @@ -3592,9 +3597,9 @@ static int pfkey_sendmsg(struct kiocb *kiocb, if (!hdr) goto out; - mutex_lock(&xfrm_cfg_mutex); + mutex_lock(&net->xfrm.xfrm_cfg_mutex); err = pfkey_process(sk, skb, hdr); - mutex_unlock(&xfrm_cfg_mutex); + mutex_unlock(&net->xfrm.xfrm_cfg_mutex); out: if (err && hdr && pfkey_error(hdr, err, sk) == 0) diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 907fd2f..73b04d3 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -39,12 +39,7 @@ #define XFRM_QUEUE_TMO_MAX ((unsigned)(60*HZ)) #define XFRM_MAX_QUEUE_LEN 100 -DEFINE_MUTEX(xfrm_cfg_mutex); -EXPORT_SYMBOL(xfrm_cfg_mutex); - -static DEFINE_SPINLOCK(xfrm_policy_sk_bundle_lock); static struct dst_entry *xfrm_policy_sk_bundles; -static DEFINE_RWLOCK(xfrm_policy_lock); static DEFINE_SPINLOCK(xfrm_policy_afinfo_lock); static struct xfrm_policy_afinfo __rcu *xfrm_policy_afinfo[NPROTO] @@ -438,7 +433,7 @@ static void xfrm_bydst_resize(struct net *net, int dir) if (!ndst) return; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); for (i = hmask; i >= 0; i--) xfrm_dst_hash_transfer(odst + i, ndst, nhashmask); @@ -446,7 +441,7 @@ static void xfrm_bydst_resize(struct net *net, int dir) net->xfrm.policy_bydst[dir].table = ndst; net->xfrm.policy_bydst[dir].hmask = nhashmask; - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); xfrm_hash_free(odst, (hmask + 1) * sizeof(struct hlist_head)); } @@ -463,7 +458,7 @@ static void xfrm_byidx_resize(struct net *net, int total) if (!nidx) return; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); for (i = hmask; i >= 0; i--) xfrm_idx_hash_transfer(oidx + i, nidx, nhashmask); @@ -471,7 +466,7 @@ static void xfrm_byidx_resize(struct net *net, int total) net->xfrm.policy_byidx = nidx; net->xfrm.policy_idx_hmask = nhashmask; - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); xfrm_hash_free(oidx, (hmask + 1) * sizeof(struct hlist_head)); } @@ -504,7 +499,7 @@ static inline int xfrm_byidx_should_resize(struct net *net, int total) void xfrm_spd_getinfo(struct net *net, struct xfrmk_spdinfo *si) { - read_lock_bh(&xfrm_policy_lock); + read_lock_bh(&net->xfrm.xfrm_policy_lock); si->incnt = net->xfrm.policy_count[XFRM_POLICY_IN]; si->outcnt = net->xfrm.policy_count[XFRM_POLICY_OUT]; si->fwdcnt = net->xfrm.policy_count[XFRM_POLICY_FWD]; @@ -513,7 +508,7 @@ void xfrm_spd_getinfo(struct net *net, struct xfrmk_spdinfo *si) si->fwdscnt = net->xfrm.policy_count[XFRM_POLICY_FWD+XFRM_POLICY_MAX]; si->spdhcnt = net->xfrm.policy_idx_hmask; si->spdhmcnt = xfrm_policy_hashmax; - read_unlock_bh(&xfrm_policy_lock); + read_unlock_bh(&net->xfrm.xfrm_policy_lock); } EXPORT_SYMBOL(xfrm_spd_getinfo); @@ -636,7 +631,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl) struct hlist_head *chain; struct hlist_node *newpos; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); chain = policy_hash_bysel(net, &policy->selector, policy->family, dir); delpol = NULL; newpos = NULL; @@ -647,7 +642,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl) xfrm_sec_ctx_match(pol->security, policy->security) && !WARN_ON(delpol)) { if (excl) { - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); return -EEXIST; } delpol = pol; @@ -685,7 +680,7 @@ int xfrm_policy_insert(int dir, struct xfrm_policy *policy, int excl) if (!mod_timer(&policy->timer, jiffies + HZ)) xfrm_pol_hold(policy); list_add(&policy->walk.all, &net->xfrm.policy_all); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); if (delpol) xfrm_policy_kill(delpol); @@ -705,7 +700,7 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type, struct hlist_head *chain; *err = 0; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); chain = policy_hash_bysel(net, sel, sel->family, dir); ret = NULL; hlist_for_each_entry(pol, chain, bydst) { @@ -718,7 +713,7 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type, *err = security_xfrm_policy_delete( pol->security); if (*err) { - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); return pol; } __xfrm_policy_unlink(pol, dir); @@ -727,7 +722,7 @@ struct xfrm_policy *xfrm_policy_bysel_ctx(struct net *net, u32 mark, u8 type, break; } } - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); if (ret && delete) xfrm_policy_kill(ret); @@ -746,7 +741,7 @@ struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8 type, return NULL; *err = 0; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); chain = net->xfrm.policy_byidx + idx_hash(net, id); ret = NULL; hlist_for_each_entry(pol, chain, byidx) { @@ -757,7 +752,7 @@ struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8 type, *err = security_xfrm_policy_delete( pol->security); if (*err) { - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); return pol; } __xfrm_policy_unlink(pol, dir); @@ -766,7 +761,7 @@ struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8 type, break; } } - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); if (ret && delete) xfrm_policy_kill(ret); @@ -829,7 +824,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) { int dir, err = 0, cnt = 0; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); err = xfrm_policy_flush_secctx_check(net, type, audit_info); if (err) @@ -845,7 +840,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) if (pol->type != type) continue; __xfrm_policy_unlink(pol, dir); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); cnt++; xfrm_audit_policy_delete(pol, 1, audit_info->loginuid, @@ -854,7 +849,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) xfrm_policy_kill(pol); - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); goto again1; } @@ -866,7 +861,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) if (pol->type != type) continue; __xfrm_policy_unlink(pol, dir); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); cnt++; xfrm_audit_policy_delete(pol, 1, @@ -875,7 +870,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) audit_info->secid); xfrm_policy_kill(pol); - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); goto again2; } } @@ -884,7 +879,7 @@ int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info) if (!cnt) err = -ESRCH; out: - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); return err; } EXPORT_SYMBOL(xfrm_policy_flush); @@ -904,7 +899,7 @@ int xfrm_policy_walk(struct net *net, struct xfrm_policy_walk *walk, if (list_empty(&walk->walk.all) && walk->seq != 0) return 0; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); if (list_empty(&walk->walk.all)) x = list_first_entry(&net->xfrm.policy_all, struct xfrm_policy_walk_entry, all); else @@ -930,7 +925,7 @@ int xfrm_policy_walk(struct net *net, struct xfrm_policy_walk *walk, } list_del_init(&walk->walk.all); out: - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); return error; } EXPORT_SYMBOL(xfrm_policy_walk); @@ -944,14 +939,14 @@ void xfrm_policy_walk_init(struct xfrm_policy_walk *walk, u8 type) } EXPORT_SYMBOL(xfrm_policy_walk_init); -void xfrm_policy_walk_done(struct xfrm_policy_walk *walk) +void xfrm_policy_walk_done(struct xfrm_policy_walk *walk, struct net *net) { if (list_empty(&walk->walk.all)) return; - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); /*FIXME where is net? */ list_del(&walk->walk.all); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); } EXPORT_SYMBOL(xfrm_policy_walk_done); @@ -996,7 +991,7 @@ static struct xfrm_policy *xfrm_policy_lookup_bytype(struct net *net, u8 type, if (unlikely(!daddr || !saddr)) return NULL; - read_lock_bh(&xfrm_policy_lock); + read_lock_bh(&net->xfrm.xfrm_policy_lock); chain = policy_hash_direct(net, daddr, saddr, family, dir); ret = NULL; hlist_for_each_entry(pol, chain, bydst) { @@ -1032,7 +1027,7 @@ static struct xfrm_policy *xfrm_policy_lookup_bytype(struct net *net, u8 type, if (ret) xfrm_pol_hold(ret); fail: - read_unlock_bh(&xfrm_policy_lock); + read_unlock_bh(&net->xfrm.xfrm_policy_lock); return ret; } @@ -1109,8 +1104,9 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, const struct flowi *fl) { struct xfrm_policy *pol; + struct net *net = sock_net(sk); - read_lock_bh(&xfrm_policy_lock); + read_lock_bh(&net->xfrm.xfrm_policy_lock); if ((pol = sk->sk_policy[dir]) != NULL) { bool match = xfrm_selector_match(&pol->selector, fl, sk->sk_family); @@ -1134,7 +1130,7 @@ static struct xfrm_policy *xfrm_sk_policy_lookup(struct sock *sk, int dir, pol = NULL; } out: - read_unlock_bh(&xfrm_policy_lock); + read_unlock_bh(&net->xfrm.xfrm_policy_lock); return pol; } @@ -1172,9 +1168,11 @@ static struct xfrm_policy *__xfrm_policy_unlink(struct xfrm_policy *pol, int xfrm_policy_delete(struct xfrm_policy *pol, int dir) { - write_lock_bh(&xfrm_policy_lock); + struct net *net = xp_net(pol); + + write_lock_bh(&net->xfrm.xfrm_policy_lock); pol = __xfrm_policy_unlink(pol, dir); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); if (pol) { xfrm_policy_kill(pol); return 0; @@ -1193,7 +1191,7 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol) return -EINVAL; #endif - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); old_pol = sk->sk_policy[dir]; sk->sk_policy[dir] = pol; if (pol) { @@ -1210,7 +1208,7 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol) */ __xfrm_policy_unlink(old_pol, XFRM_POLICY_MAX+dir); } - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); if (old_pol) { xfrm_policy_kill(old_pol); @@ -1221,6 +1219,7 @@ int xfrm_sk_policy_insert(struct sock *sk, int dir, struct xfrm_policy *pol) static struct xfrm_policy *clone_policy(const struct xfrm_policy *old, int dir) { struct xfrm_policy *newp = xfrm_policy_alloc(xp_net(old), GFP_ATOMIC); + struct net *net = xp_net(old); if (newp) { newp->selector = old->selector; @@ -1239,9 +1238,9 @@ static struct xfrm_policy *clone_policy(const struct xfrm_policy *old, int dir) newp->type = old->type; memcpy(newp->xfrm_vec, old->xfrm_vec, newp->xfrm_nr*sizeof(struct xfrm_tmpl)); - write_lock_bh(&xfrm_policy_lock); + write_lock_bh(&net->xfrm.xfrm_policy_lock); __xfrm_policy_link(newp, XFRM_POLICY_MAX+dir); - write_unlock_bh(&xfrm_policy_lock); + write_unlock_bh(&net->xfrm.xfrm_policy_lock); xfrm_pol_put(newp); } return newp; @@ -2112,10 +2111,10 @@ restart: dst_hold(&xdst->u.dst); - spin_lock_bh(&xfrm_policy_sk_bundle_lock); + spin_lock_bh(&net->xfrm.xfrm_policy_sk_bundle_lock); xdst->u.dst.next = xfrm_policy_sk_bundles; xfrm_policy_sk_bundles = &xdst->u.dst; - spin_unlock_bh(&xfrm_policy_sk_bundle_lock); + spin_unlock_bh(&net->xfrm.xfrm_policy_sk_bundle_lock); route = xdst->route; } @@ -2440,7 +2439,7 @@ int __xfrm_policy_check(struct sock *sk, int dir, struct sk_buff *skb, } xfrm_nr = ti; if (npols > 1) { - xfrm_tmpl_sort(stp, tpp, xfrm_nr, family); + xfrm_tmpl_sort(stp, tpp, xfrm_nr, family, net); tpp = stp; } @@ -2569,10 +2568,10 @@ static void __xfrm_garbage_collect(struct net *net) { struct dst_entry *head, *next; - spin_lock_bh(&xfrm_policy_sk_bundle_lock); + spin_lock_bh(&net->xfrm.xfrm_policy_sk_bundle_lock); head = xfrm_policy_sk_bundles; xfrm_policy_sk_bundles = NULL; - spin_unlock_bh(&xfrm_policy_sk_bundle_lock); + spin_unlock_bh(&net->xfrm.xfrm_policy_sk_bundle_lock); while (head) { next = head->next; @@ -2956,6 +2955,13 @@ static int __net_init xfrm_net_init(struct net *net) rv = xfrm_sysctl_init(net); if (rv < 0) goto out_sysctl; + + /* Initialize the per-net locks here */ + spin_lock_init(&net->xfrm.xfrm_state_lock); + rwlock_init(&net->xfrm.xfrm_policy_lock); + spin_lock_init(&net->xfrm.xfrm_policy_sk_bundle_lock); + mutex_init(&net->xfrm.xfrm_cfg_mutex); + return 0; out_sysctl: @@ -3082,7 +3088,7 @@ static struct xfrm_policy * xfrm_migrate_policy_find(const struct xfrm_selector struct hlist_head *chain; u32 priority = ~0U; - read_lock_bh(&xfrm_policy_lock); + read_lock_bh(&net->xfrm.xfrm_policy_lock); /*FIXME*/ chain = policy_hash_direct(net, &sel->daddr, &sel->saddr, sel->family, dir); hlist_for_each_entry(pol, chain, bydst) { if (xfrm_migrate_selector_match(sel, &pol->selector) && @@ -3105,7 +3111,7 @@ static struct xfrm_policy * xfrm_migrate_policy_find(const struct xfrm_selector if (ret) xfrm_pol_hold(ret); - read_unlock_bh(&xfrm_policy_lock); + read_unlock_bh(&net->xfrm.xfrm_policy_lock); return ret; } @@ -3236,7 +3242,7 @@ int xfrm_migrate(const struct xfrm_selector *sel, u8 dir, u8 type, /* Stage 2 - find and update state(s) */ for (i = 0, mp = m; i < num_migrate; i++, mp++) { - if ((x = xfrm_migrate_state_find(mp))) { + if ((x = xfrm_migrate_state_find(mp, net))) { x_cur[nx_cur] = x; nx_cur++; if ((xc = xfrm_state_migrate(x, mp))) { diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index 68c2f357..290479d 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -35,8 +35,6 @@ destination/tunnel endpoint. (output) */ -static DEFINE_SPINLOCK(xfrm_state_lock); - static unsigned int xfrm_state_hashmax __read_mostly = 1 * 1024 * 1024; static inline unsigned int xfrm_dst_hash(struct net *net, @@ -127,7 +125,7 @@ static void xfrm_hash_resize(struct work_struct *work) goto out_unlock; } - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); nhashmask = (nsize / sizeof(struct hlist_head)) - 1U; for (i = net->xfrm.state_hmask; i >= 0; i--) @@ -144,7 +142,7 @@ static void xfrm_hash_resize(struct work_struct *work) net->xfrm.state_byspi = nspi; net->xfrm.state_hmask = nhashmask; - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); osize = (ohashmask + 1) * sizeof(struct hlist_head); xfrm_hash_free(odst, osize); @@ -535,14 +533,14 @@ int __xfrm_state_delete(struct xfrm_state *x) if (x->km.state != XFRM_STATE_DEAD) { x->km.state = XFRM_STATE_DEAD; - spin_lock(&xfrm_state_lock); + spin_lock(&net->xfrm.xfrm_state_lock); list_del(&x->km.all); hlist_del(&x->bydst); hlist_del(&x->bysrc); if (x->id.spi) hlist_del(&x->byspi); net->xfrm.state_num--; - spin_unlock(&xfrm_state_lock); + spin_unlock(&net->xfrm.xfrm_state_lock); /* All xfrm_state objects are created by xfrm_state_alloc. * The xfrm_state_alloc call gives a reference, and that @@ -603,7 +601,7 @@ int xfrm_state_flush(struct net *net, u8 proto, struct xfrm_audit *audit_info) { int i, err = 0, cnt = 0; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); err = xfrm_state_flush_secctx_check(net, proto, audit_info); if (err) goto out; @@ -616,7 +614,7 @@ restart: if (!xfrm_state_kern(x) && xfrm_id_proto_match(x->id.proto, proto)) { xfrm_state_hold(x); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); err = xfrm_state_delete(x); xfrm_audit_state_delete(x, err ? 0 : 1, @@ -627,7 +625,7 @@ restart: if (!err) cnt++; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); goto restart; } } @@ -636,7 +634,7 @@ restart: err = 0; out: - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); wake_up(&net->xfrm.km_waitq); return err; } @@ -644,11 +642,11 @@ EXPORT_SYMBOL(xfrm_state_flush); void xfrm_sad_getinfo(struct net *net, struct xfrmk_sadinfo *si) { - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); si->sadcnt = net->xfrm.state_num; si->sadhcnt = net->xfrm.state_hmask; si->sadhmcnt = xfrm_state_hashmax; - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); } EXPORT_SYMBOL(xfrm_sad_getinfo); @@ -801,7 +799,7 @@ xfrm_state_find(const xfrm_address_t *daddr, const xfrm_address_t *saddr, to_put = NULL; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); h = xfrm_dst_hash(net, daddr, saddr, tmpl->reqid, encap_family); hlist_for_each_entry(x, net->xfrm.state_bydst+h, bydst) { if (x->props.family == encap_family && @@ -886,7 +884,7 @@ out: xfrm_state_hold(x); else *err = acquire_in_progress ? -EAGAIN : error; - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); if (to_put) xfrm_state_put(to_put); return x; @@ -900,7 +898,7 @@ xfrm_stateonly_find(struct net *net, u32 mark, unsigned int h; struct xfrm_state *rx = NULL, *x = NULL; - spin_lock(&xfrm_state_lock); + spin_lock(&net->xfrm.xfrm_state_lock); h = xfrm_dst_hash(net, daddr, saddr, reqid, family); hlist_for_each_entry(x, net->xfrm.state_bydst+h, bydst) { if (x->props.family == family && @@ -918,7 +916,7 @@ xfrm_stateonly_find(struct net *net, u32 mark, if (rx) xfrm_state_hold(rx); - spin_unlock(&xfrm_state_lock); + spin_unlock(&net->xfrm.xfrm_state_lock); return rx; @@ -957,7 +955,7 @@ static void __xfrm_state_insert(struct xfrm_state *x) xfrm_hash_grow_check(net, x->bydst.next != NULL); } -/* xfrm_state_lock is held */ +/* net->xfrm.xfrm_state_lock is held */ static void __xfrm_state_bump_genids(struct xfrm_state *xnew) { struct net *net = xs_net(xnew); @@ -980,14 +978,16 @@ static void __xfrm_state_bump_genids(struct xfrm_state *xnew) void xfrm_state_insert(struct xfrm_state *x) { - spin_lock_bh(&xfrm_state_lock); + struct net *net = xs_net(x); + + spin_lock_bh(&net->xfrm.xfrm_state_lock); __xfrm_state_bump_genids(x); __xfrm_state_insert(x); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); } EXPORT_SYMBOL(xfrm_state_insert); -/* xfrm_state_lock is held */ +/* net->xfrm.xfrm_state_lock is held */ static struct xfrm_state *__find_acq_core(struct net *net, const struct xfrm_mark *m, unsigned short family, u8 mode, @@ -1079,7 +1079,7 @@ int xfrm_state_add(struct xfrm_state *x) to_put = NULL; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x1 = __xfrm_state_locate(x, use_spi, family); if (x1) { @@ -1108,7 +1108,7 @@ int xfrm_state_add(struct xfrm_state *x) err = 0; out: - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); if (x1) { xfrm_state_delete(x1); @@ -1203,16 +1203,16 @@ out: return NULL; } -/* xfrm_state_lock is held */ -struct xfrm_state * xfrm_migrate_state_find(struct xfrm_migrate *m) +/* net->xfrm.xfrm_state_lock is held */ +struct xfrm_state *xfrm_migrate_state_find(struct xfrm_migrate *m, struct net *net) { unsigned int h; struct xfrm_state *x; if (m->reqid) { - h = xfrm_dst_hash(&init_net, &m->old_daddr, &m->old_saddr, + h = xfrm_dst_hash(net, &m->old_daddr, &m->old_saddr, m->reqid, m->old_family); - hlist_for_each_entry(x, init_net.xfrm.state_bydst+h, bydst) { + hlist_for_each_entry(x, net->xfrm.state_bydst+h, bydst) { if (x->props.mode != m->mode || x->id.proto != m->proto) continue; @@ -1227,9 +1227,9 @@ struct xfrm_state * xfrm_migrate_state_find(struct xfrm_migrate *m) return x; } } else { - h = xfrm_src_hash(&init_net, &m->old_daddr, &m->old_saddr, + h = xfrm_src_hash(net, &m->old_daddr, &m->old_saddr, m->old_family); - hlist_for_each_entry(x, init_net.xfrm.state_bysrc+h, bysrc) { + hlist_for_each_entry(x, net->xfrm.state_bysrc+h, bysrc) { if (x->props.mode != m->mode || x->id.proto != m->proto) continue; @@ -1283,10 +1283,11 @@ int xfrm_state_update(struct xfrm_state *x) struct xfrm_state *x1, *to_put; int err; int use_spi = xfrm_id_proto_match(x->id.proto, IPSEC_PROTO_ANY); + struct net *net = xs_net(x); to_put = NULL; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x1 = __xfrm_state_locate(x, use_spi, x->props.family); err = -ESRCH; @@ -1306,7 +1307,7 @@ int xfrm_state_update(struct xfrm_state *x) err = 0; out: - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); if (to_put) xfrm_state_put(to_put); @@ -1377,9 +1378,9 @@ xfrm_state_lookup(struct net *net, u32 mark, const xfrm_address_t *daddr, __be32 { struct xfrm_state *x; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x = __xfrm_state_lookup(net, mark, daddr, spi, proto, family); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); return x; } EXPORT_SYMBOL(xfrm_state_lookup); @@ -1391,9 +1392,9 @@ xfrm_state_lookup_byaddr(struct net *net, u32 mark, { struct xfrm_state *x; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x = __xfrm_state_lookup_byaddr(net, mark, daddr, saddr, proto, family); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); return x; } EXPORT_SYMBOL(xfrm_state_lookup_byaddr); @@ -1405,9 +1406,9 @@ xfrm_find_acq(struct net *net, const struct xfrm_mark *mark, u8 mode, u32 reqid, { struct xfrm_state *x; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x = __find_acq_core(net, mark, family, mode, reqid, proto, daddr, saddr, create); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); return x; } @@ -1416,17 +1417,17 @@ EXPORT_SYMBOL(xfrm_find_acq); #ifdef CONFIG_XFRM_SUB_POLICY int xfrm_tmpl_sort(struct xfrm_tmpl **dst, struct xfrm_tmpl **src, int n, - unsigned short family) + unsigned short family, struct net *net) { int err = 0; struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family); if (!afinfo) return -EAFNOSUPPORT; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); /*FIXME*/ if (afinfo->tmpl_sort) err = afinfo->tmpl_sort(dst, src, n); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); xfrm_state_put_afinfo(afinfo); return err; } @@ -1438,13 +1439,15 @@ xfrm_state_sort(struct xfrm_state **dst, struct xfrm_state **src, int n, { int err = 0; struct xfrm_state_afinfo *afinfo = xfrm_state_get_afinfo(family); + struct net *net = xs_net(*dst); + if (!afinfo) return -EAFNOSUPPORT; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); if (afinfo->state_sort) err = afinfo->state_sort(dst, src, n); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); xfrm_state_put_afinfo(afinfo); return err; } @@ -1476,9 +1479,9 @@ struct xfrm_state *xfrm_find_acq_byseq(struct net *net, u32 mark, u32 seq) { struct xfrm_state *x; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); x = __xfrm_find_acq_byseq(net, mark, seq); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); return x; } EXPORT_SYMBOL(xfrm_find_acq_byseq); @@ -1536,10 +1539,10 @@ int xfrm_alloc_spi(struct xfrm_state *x, u32 low, u32 high) } } if (x->id.spi) { - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); h = xfrm_spi_hash(net, &x->id.daddr, x->id.spi, x->id.proto, x->props.family); hlist_add_head(&x->byspi, net->xfrm.state_byspi+h); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); err = 0; } @@ -1562,7 +1565,7 @@ int xfrm_state_walk(struct net *net, struct xfrm_state_walk *walk, if (walk->seq != 0 && list_empty(&walk->all)) return 0; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); if (list_empty(&walk->all)) x = list_first_entry(&net->xfrm.state_all, struct xfrm_state_walk, all); else @@ -1586,7 +1589,7 @@ int xfrm_state_walk(struct net *net, struct xfrm_state_walk *walk, } list_del_init(&walk->all); out: - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); return err; } EXPORT_SYMBOL(xfrm_state_walk); @@ -1600,14 +1603,14 @@ void xfrm_state_walk_init(struct xfrm_state_walk *walk, u8 proto) } EXPORT_SYMBOL(xfrm_state_walk_init); -void xfrm_state_walk_done(struct xfrm_state_walk *walk) +void xfrm_state_walk_done(struct xfrm_state_walk *walk, struct net *net) { if (list_empty(&walk->all)) return; - spin_lock_bh(&xfrm_state_lock); + spin_lock_bh(&net->xfrm.xfrm_state_lock); list_del(&walk->all); - spin_unlock_bh(&xfrm_state_lock); + spin_unlock_bh(&net->xfrm.xfrm_state_lock); } EXPORT_SYMBOL(xfrm_state_walk_done); @@ -2026,6 +2029,7 @@ int __net_init xfrm_state_init(struct net *net) INIT_HLIST_HEAD(&net->xfrm.state_gc_list); INIT_WORK(&net->xfrm.state_gc_work, xfrm_state_gc_task); init_waitqueue_head(&net->xfrm.km_waitq); + spin_lock_init(&net->xfrm.xfrm_state_lock); return 0; out_byspi: diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 840cc8d..16c8460 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -877,7 +877,10 @@ static int dump_one_state(struct xfrm_state *x, int count, void *ptr) static int xfrm_dump_sa_done(struct netlink_callback *cb) { struct xfrm_state_walk *walk = (struct xfrm_state_walk *) &cb->args[1]; - xfrm_state_walk_done(walk); + struct sock *sk = cb->skb->sk; + struct net *net = sock_net(sk); + + xfrm_state_walk_done(walk, net); return 0; } @@ -1555,8 +1558,9 @@ static int dump_one_policy(struct xfrm_policy *xp, int dir, int count, void *ptr static int xfrm_dump_policy_done(struct netlink_callback *cb) { struct xfrm_policy_walk *walk = (struct xfrm_policy_walk *) &cb->args[1]; + struct net *net = sock_net(cb->skb->sk); - xfrm_policy_walk_done(walk); + xfrm_policy_walk_done(walk, net); return 0; } @@ -2403,9 +2407,11 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) static void xfrm_netlink_rcv(struct sk_buff *skb) { - mutex_lock(&xfrm_cfg_mutex); + struct net *net = sock_net(skb->sk); + + mutex_lock(&net->xfrm.xfrm_cfg_mutex); netlink_rcv_skb(skb, &xfrm_user_rcv_msg); - mutex_unlock(&xfrm_cfg_mutex); + mutex_unlock(&net->xfrm.xfrm_cfg_mutex); } static inline size_t xfrm_expire_msgsize(void) -- cgit v1.1 From 5b8ef3415a21f173ab115e90ec92c071a03f22d7 Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Tue, 27 Aug 2013 13:43:30 +0200 Subject: xfrm: Remove ancient sleeping when the SA is in acquire state We now queue packets to the policy if the states are not yet resolved, this replaces the ancient sleeping code. Also the sleeping can cause indefinite task hangs if the needed state does not get resolved. Signed-off-by: Steffen Klassert --- include/net/netns/xfrm.h | 2 -- net/key/af_key.c | 5 ++--- net/xfrm/xfrm_policy.c | 21 ++------------------- net/xfrm/xfrm_state.c | 21 +-------------------- 4 files changed, 5 insertions(+), 44 deletions(-) diff --git a/include/net/netns/xfrm.h b/include/net/netns/xfrm.h index ea28404..1006a26 100644 --- a/include/net/netns/xfrm.h +++ b/include/net/netns/xfrm.h @@ -33,8 +33,6 @@ struct netns_xfrm { struct hlist_head state_gc_list; struct work_struct state_gc_work; - wait_queue_head_t km_waitq; - struct list_head policy_all; struct hlist_head *policy_byidx; unsigned int policy_idx_hmask; diff --git a/net/key/af_key.c b/net/key/af_key.c index 9a039ac..5beabd8 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -1380,10 +1380,9 @@ static int pfkey_acquire(struct sock *sk, struct sk_buff *skb, const struct sadb return 0; spin_lock_bh(&x->lock); - if (x->km.state == XFRM_STATE_ACQ) { + if (x->km.state == XFRM_STATE_ACQ) x->km.state = XFRM_STATE_ERROR; - wake_up(&net->xfrm.km_waitq); - } + spin_unlock_bh(&x->lock); xfrm_state_put(x); return 0; diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 73b04d3..a7487f3 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -1901,8 +1901,7 @@ static struct xfrm_dst *xfrm_create_dummy_bundle(struct net *net, if (IS_ERR(xdst)) return xdst; - if (net->xfrm.sysctl_larval_drop || num_xfrms <= 0 || - (fl->flowi_flags & FLOWI_FLAG_CAN_SLEEP)) + if (net->xfrm.sysctl_larval_drop || num_xfrms <= 0) return xdst; dst1 = &xdst->u.dst; @@ -2077,7 +2076,6 @@ struct dst_entry *xfrm_lookup(struct net *net, struct dst_entry *dst_orig, u8 dir = policy_to_flow_dir(XFRM_POLICY_OUT); int i, err, num_pols, num_xfrms = 0, drop_pols = 0; -restart: dst = NULL; xdst = NULL; route = NULL; @@ -2157,23 +2155,8 @@ restart: return make_blackhole(net, family, dst_orig); } - if (fl->flowi_flags & FLOWI_FLAG_CAN_SLEEP) { - DECLARE_WAITQUEUE(wait, current); - add_wait_queue(&net->xfrm.km_waitq, &wait); - set_current_state(TASK_INTERRUPTIBLE); - schedule(); - set_current_state(TASK_RUNNING); - remove_wait_queue(&net->xfrm.km_waitq, &wait); - - if (!signal_pending(current)) { - dst_release(dst); - goto restart; - } - - err = -ERESTART; - } else - err = -EAGAIN; + err = -EAGAIN; XFRM_INC_STATS(net, LINUX_MIB_XFRMOUTNOSTATES); goto error; diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index 290479d..e845066 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -372,8 +372,6 @@ static void xfrm_state_gc_task(struct work_struct *work) hlist_for_each_entry_safe(x, tmp, &gc_list, gclist) xfrm_state_gc_destroy(x); - - wake_up(&net->xfrm.km_waitq); } static inline unsigned long make_jiffies(long secs) @@ -388,7 +386,6 @@ static enum hrtimer_restart xfrm_timer_handler(struct hrtimer * me) { struct tasklet_hrtimer *thr = container_of(me, struct tasklet_hrtimer, timer); struct xfrm_state *x = container_of(thr, struct xfrm_state, mtimer); - struct net *net = xs_net(x); unsigned long now = get_seconds(); long next = LONG_MAX; int warn = 0; @@ -458,12 +455,8 @@ resched: goto out; expired: - if (x->km.state == XFRM_STATE_ACQ && x->id.spi == 0) { + if (x->km.state == XFRM_STATE_ACQ && x->id.spi == 0) x->km.state = XFRM_STATE_EXPIRED; - wake_up(&net->xfrm.km_waitq); - next = 2; - goto resched; - } err = __xfrm_state_delete(x); if (!err) @@ -635,7 +628,6 @@ restart: out: spin_unlock_bh(&net->xfrm.xfrm_state_lock); - wake_up(&net->xfrm.km_waitq); return err; } EXPORT_SYMBOL(xfrm_state_flush); @@ -948,8 +940,6 @@ static void __xfrm_state_insert(struct xfrm_state *x) if (x->replay_maxage) mod_timer(&x->rtimer, jiffies + x->replay_maxage); - wake_up(&net->xfrm.km_waitq); - net->xfrm.state_num++; xfrm_hash_grow_check(net, x->bydst.next != NULL); @@ -1658,16 +1648,12 @@ EXPORT_SYMBOL(km_state_notify); void km_state_expired(struct xfrm_state *x, int hard, u32 portid) { - struct net *net = xs_net(x); struct km_event c; c.data.hard = hard; c.portid = portid; c.event = XFRM_MSG_EXPIRE; km_state_notify(x, &c); - - if (hard) - wake_up(&net->xfrm.km_waitq); } EXPORT_SYMBOL(km_state_expired); @@ -1710,16 +1696,12 @@ EXPORT_SYMBOL(km_new_mapping); void km_policy_expired(struct xfrm_policy *pol, int dir, int hard, u32 portid) { - struct net *net = xp_net(pol); struct km_event c; c.data.hard = hard; c.portid = portid; c.event = XFRM_MSG_POLEXPIRE; km_policy_notify(pol, dir, &c); - - if (hard) - wake_up(&net->xfrm.km_waitq); } EXPORT_SYMBOL(km_policy_expired); @@ -2028,7 +2010,6 @@ int __net_init xfrm_state_init(struct net *net) INIT_WORK(&net->xfrm.state_hash_work, xfrm_hash_resize); INIT_HLIST_HEAD(&net->xfrm.state_gc_list); INIT_WORK(&net->xfrm.state_gc_work, xfrm_state_gc_task); - init_waitqueue_head(&net->xfrm.km_waitq); spin_lock_init(&net->xfrm.xfrm_state_lock); return 0; -- cgit v1.1 From 0e0d44ab4275549998567cd4700b43f7496eb62b Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Wed, 28 Aug 2013 08:04:14 +0200 Subject: net: Remove FLOWI_FLAG_CAN_SLEEP FLOWI_FLAG_CAN_SLEEP was used to notify xfrm about the posibility to sleep until the needed states are resolved. This code is gone, so FLOWI_FLAG_CAN_SLEEP is not needed anymore. Signed-off-by: Steffen Klassert --- include/net/flow.h | 3 +-- include/net/ipv6.h | 6 ++---- include/net/route.h | 8 +++----- net/dccp/ipv4.c | 2 +- net/dccp/ipv6.c | 8 ++++---- net/decnet/dn_route.c | 2 -- net/ipv4/af_inet.c | 2 +- net/ipv4/datagram.c | 2 +- net/ipv4/raw.c | 2 +- net/ipv4/tcp_ipv4.c | 2 +- net/ipv4/udp.c | 2 +- net/ipv6/af_inet6.c | 2 +- net/ipv6/datagram.c | 2 +- net/ipv6/inet6_connection_sock.c | 4 ++-- net/ipv6/ip6_output.c | 12 ++---------- net/ipv6/ping.c | 2 +- net/ipv6/raw.c | 2 +- net/ipv6/syncookies.c | 2 +- net/ipv6/tcp_ipv6.c | 4 ++-- net/ipv6/udp.c | 2 +- net/l2tp/l2tp_ip6.c | 2 +- net/sctp/ipv6.c | 4 ++-- 22 files changed, 31 insertions(+), 46 deletions(-) diff --git a/include/net/flow.h b/include/net/flow.h index 65ce471..d23e7fa 100644 --- a/include/net/flow.h +++ b/include/net/flow.h @@ -20,8 +20,7 @@ struct flowi_common { __u8 flowic_proto; __u8 flowic_flags; #define FLOWI_FLAG_ANYSRC 0x01 -#define FLOWI_FLAG_CAN_SLEEP 0x02 -#define FLOWI_FLAG_KNOWN_NH 0x04 +#define FLOWI_FLAG_KNOWN_NH 0x02 __u32 flowic_secid; }; diff --git a/include/net/ipv6.h b/include/net/ipv6.h index eb198ac..26c4199 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -711,11 +711,9 @@ void ip6_flush_pending_frames(struct sock *sk); int ip6_dst_lookup(struct sock *sk, struct dst_entry **dst, struct flowi6 *fl6); struct dst_entry *ip6_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, - const struct in6_addr *final_dst, - bool can_sleep); + const struct in6_addr *final_dst); struct dst_entry *ip6_sk_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, - const struct in6_addr *final_dst, - bool can_sleep); + const struct in6_addr *final_dst); struct dst_entry *ip6_blackhole_route(struct net *net, struct dst_entry *orig_dst); diff --git a/include/net/route.h b/include/net/route.h index f68c167..638e3eb 100644 --- a/include/net/route.h +++ b/include/net/route.h @@ -239,14 +239,12 @@ static inline char rt_tos2priority(u8 tos) static inline void ip_route_connect_init(struct flowi4 *fl4, __be32 dst, __be32 src, u32 tos, int oif, u8 protocol, __be16 sport, __be16 dport, - struct sock *sk, bool can_sleep) + struct sock *sk) { __u8 flow_flags = 0; if (inet_sk(sk)->transparent) flow_flags |= FLOWI_FLAG_ANYSRC; - if (can_sleep) - flow_flags |= FLOWI_FLAG_CAN_SLEEP; flowi4_init_output(fl4, oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, protocol, flow_flags, dst, src, dport, sport); @@ -256,13 +254,13 @@ static inline struct rtable *ip_route_connect(struct flowi4 *fl4, __be32 dst, __be32 src, u32 tos, int oif, u8 protocol, __be16 sport, __be16 dport, - struct sock *sk, bool can_sleep) + struct sock *sk) { struct net *net = sock_net(sk); struct rtable *rt; ip_route_connect_init(fl4, dst, src, tos, oif, protocol, - sport, dport, sk, can_sleep); + sport, dport, sk); if (!dst || !src) { rt = __ip_route_output_key(net, fl4); diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index d9f65fc..88299c2 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -75,7 +75,7 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) rt = ip_route_connect(fl4, nexthop, inet->inet_saddr, RT_CONN_FLAGS(sk), sk->sk_bound_dev_if, IPPROTO_DCCP, - orig_sport, orig_dport, sk, true); + orig_sport, orig_dport, sk); if (IS_ERR(rt)) return PTR_ERR(rt); diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c index 4ac71ff..699a2d2 100644 --- a/net/dccp/ipv6.c +++ b/net/dccp/ipv6.c @@ -237,7 +237,7 @@ static int dccp_v6_send_response(struct sock *sk, struct request_sock *req) final_p = fl6_update_dst(&fl6, np->opt, &final); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); dst = NULL; @@ -301,7 +301,7 @@ static void dccp_v6_ctl_send_reset(struct sock *sk, struct sk_buff *rxskb) security_skb_classify_flow(rxskb, flowi6_to_flowi(&fl6)); /* sk = NULL, but it is safe for now. RST socket required. */ - dst = ip6_dst_lookup_flow(ctl_sk, &fl6, NULL, false); + dst = ip6_dst_lookup_flow(ctl_sk, &fl6, NULL); if (!IS_ERR(dst)) { skb_dst_set(skb, dst); ip6_xmit(ctl_sk, skb, &fl6, NULL, 0); @@ -512,7 +512,7 @@ static struct sock *dccp_v6_request_recv_sock(struct sock *sk, fl6.fl6_sport = htons(ireq->ir_num); security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) goto out; } @@ -932,7 +932,7 @@ static int dccp_v6_connect(struct sock *sk, struct sockaddr *uaddr, final_p = fl6_update_dst(&fl6, np->opt, &final); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); goto failure; diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c index fe32388..ad2efa5 100644 --- a/net/decnet/dn_route.c +++ b/net/decnet/dn_route.c @@ -1288,8 +1288,6 @@ int dn_route_output_sock(struct dst_entry __rcu **pprt, struct flowidn *fl, stru err = __dn_route_output_key(pprt, fl, flags & MSG_TRYHARD); if (err == 0 && fl->flowidn_proto) { - if (!(flags & MSG_DONTWAIT)) - fl->flowidn_flags |= FLOWI_FLAG_CAN_SLEEP; *pprt = xfrm_lookup(&init_net, *pprt, flowidn_to_flowi(fl), sk, 0); if (IS_ERR(*pprt)) { diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c index 70011e0..4f6d50b 100644 --- a/net/ipv4/af_inet.c +++ b/net/ipv4/af_inet.c @@ -1133,7 +1133,7 @@ static int inet_sk_reselect_saddr(struct sock *sk) fl4 = &inet->cork.fl.u.ip4; rt = ip_route_connect(fl4, daddr, 0, RT_CONN_FLAGS(sk), sk->sk_bound_dev_if, sk->sk_protocol, - inet->inet_sport, inet->inet_dport, sk, false); + inet->inet_sport, inet->inet_dport, sk); if (IS_ERR(rt)) return PTR_ERR(rt); diff --git a/net/ipv4/datagram.c b/net/ipv4/datagram.c index 19e3637..8b5134c 100644 --- a/net/ipv4/datagram.c +++ b/net/ipv4/datagram.c @@ -53,7 +53,7 @@ int ip4_datagram_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) rt = ip_route_connect(fl4, usin->sin_addr.s_addr, saddr, RT_CONN_FLAGS(sk), oif, sk->sk_protocol, - inet->inet_sport, usin->sin_port, sk, true); + inet->inet_sport, usin->sin_port, sk); if (IS_ERR(rt)) { err = PTR_ERR(rt); if (err == -ENETUNREACH) diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 23c3e5b..81e6cfd 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -575,7 +575,7 @@ static int raw_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, flowi4_init_output(&fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, inet->hdrincl ? IPPROTO_RAW : sk->sk_protocol, - inet_sk_flowi_flags(sk) | FLOWI_FLAG_CAN_SLEEP | + inet_sk_flowi_flags(sk) | (inet->hdrincl ? FLOWI_FLAG_KNOWN_NH : 0), daddr, saddr, 0, 0); diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 0672139..bbaf8cb 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -173,7 +173,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *uaddr, int addr_len) rt = ip_route_connect(fl4, nexthop, inet->inet_saddr, RT_CONN_FLAGS(sk), sk->sk_bound_dev_if, IPPROTO_TCP, - orig_sport, orig_dport, sk, true); + orig_sport, orig_dport, sk); if (IS_ERR(rt)) { err = PTR_ERR(rt); if (err == -ENETUNREACH) diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 44f6a20..480fb95df 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -990,7 +990,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, fl4 = &fl4_stack; flowi4_init_output(fl4, ipc.oif, sk->sk_mark, tos, RT_SCOPE_UNIVERSE, sk->sk_protocol, - inet_sk_flowi_flags(sk)|FLOWI_FLAG_CAN_SLEEP, + inet_sk_flowi_flags(sk), faddr, saddr, dport, inet->inet_sport); security_sk_classify_flow(sk, flowi4_to_flowi(fl4)); diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 4fbdb70..c33f98a 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -661,7 +661,7 @@ int inet6_sk_rebuild_header(struct sock *sk) final_p = fl6_update_dst(&fl6, np->opt, &final); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { sk->sk_route_caps = 0; sk->sk_err_soft = -PTR_ERR(dst); diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index 8dfe1f4..d7ef682 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -171,7 +171,7 @@ ipv4_connected: opt = flowlabel ? flowlabel->opt : np->opt; final_p = fl6_update_dst(&fl6, opt, &final); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); err = 0; if (IS_ERR(dst)) { err = PTR_ERR(dst); diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c index 77bb8af..c913818 100644 --- a/net/ipv6/inet6_connection_sock.c +++ b/net/ipv6/inet6_connection_sock.c @@ -86,7 +86,7 @@ struct dst_entry *inet6_csk_route_req(struct sock *sk, fl6->fl6_sport = htons(ireq->ir_num); security_req_classify_flow(req, flowi6_to_flowi(fl6)); - dst = ip6_dst_lookup_flow(sk, fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, fl6, final_p); if (IS_ERR(dst)) return NULL; @@ -216,7 +216,7 @@ static struct dst_entry *inet6_csk_route_socket(struct sock *sk, dst = __inet6_csk_dst_check(sk, np->dst_cookie); if (!dst) { - dst = ip6_dst_lookup_flow(sk, fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, fl6, final_p); if (!IS_ERR(dst)) __inet6_csk_dst_store(sk, dst, NULL, NULL); diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 4acdb63..3915cd8 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -938,7 +938,6 @@ EXPORT_SYMBOL_GPL(ip6_dst_lookup); * @sk: socket which provides route info * @fl6: flow to lookup * @final_dst: final destination address for ipsec lookup - * @can_sleep: we are in a sleepable context * * This function performs a route lookup on the given flow. * @@ -946,8 +945,7 @@ EXPORT_SYMBOL_GPL(ip6_dst_lookup); * error code. */ struct dst_entry *ip6_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, - const struct in6_addr *final_dst, - bool can_sleep) + const struct in6_addr *final_dst) { struct dst_entry *dst = NULL; int err; @@ -957,8 +955,6 @@ struct dst_entry *ip6_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, return ERR_PTR(err); if (final_dst) fl6->daddr = *final_dst; - if (can_sleep) - fl6->flowi6_flags |= FLOWI_FLAG_CAN_SLEEP; return xfrm_lookup(sock_net(sk), dst, flowi6_to_flowi(fl6), sk, 0); } @@ -969,7 +965,6 @@ EXPORT_SYMBOL_GPL(ip6_dst_lookup_flow); * @sk: socket which provides the dst cache and route info * @fl6: flow to lookup * @final_dst: final destination address for ipsec lookup - * @can_sleep: we are in a sleepable context * * This function performs a route lookup on the given flow with the * possibility of using the cached route in the socket if it is valid. @@ -980,8 +975,7 @@ EXPORT_SYMBOL_GPL(ip6_dst_lookup_flow); * error code. */ struct dst_entry *ip6_sk_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, - const struct in6_addr *final_dst, - bool can_sleep) + const struct in6_addr *final_dst) { struct dst_entry *dst = sk_dst_check(sk, inet6_sk(sk)->dst_cookie); int err; @@ -993,8 +987,6 @@ struct dst_entry *ip6_sk_dst_lookup_flow(struct sock *sk, struct flowi6 *fl6, return ERR_PTR(err); if (final_dst) fl6->daddr = *final_dst; - if (can_sleep) - fl6->flowi6_flags |= FLOWI_FLAG_CAN_SLEEP; return xfrm_lookup(sock_net(sk), dst, flowi6_to_flowi(fl6), sk, 0); } diff --git a/net/ipv6/ping.c b/net/ipv6/ping.c index a83243c..15d23b8 100644 --- a/net/ipv6/ping.c +++ b/net/ipv6/ping.c @@ -145,7 +145,7 @@ int ping_v6_sendmsg(struct kiocb *iocb, struct sock *sk, struct msghdr *msg, else if (!fl6.flowi6_oif) fl6.flowi6_oif = np->ucast_oif; - dst = ip6_sk_dst_lookup_flow(sk, &fl6, daddr, 1); + dst = ip6_sk_dst_lookup_flow(sk, &fl6, daddr); if (IS_ERR(dst)) return PTR_ERR(dst); rt = (struct rt6_info *) dst; diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 7fb4e14..4945fb6 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -865,7 +865,7 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, fl6.flowi6_oif = np->ucast_oif; security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); goto out; diff --git a/net/ipv6/syncookies.c b/net/ipv6/syncookies.c index 535a3ad..bb53a5e7 100644 --- a/net/ipv6/syncookies.c +++ b/net/ipv6/syncookies.c @@ -247,7 +247,7 @@ struct sock *cookie_v6_check(struct sock *sk, struct sk_buff *skb) fl6.fl6_sport = inet_sk(sk)->inet_sport; security_req_classify_flow(req, flowi6_to_flowi(&fl6)); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) goto out_free; } diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 0740f93..f14456f 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -258,7 +258,7 @@ static int tcp_v6_connect(struct sock *sk, struct sockaddr *uaddr, security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); goto failure; @@ -801,7 +801,7 @@ static void tcp_v6_send_response(struct sk_buff *skb, u32 seq, u32 ack, u32 win, * Underlying function will use this to retrieve the network * namespace */ - dst = ip6_dst_lookup_flow(ctl_sk, &fl6, NULL, false); + dst = ip6_dst_lookup_flow(ctl_sk, &fl6, NULL); if (!IS_ERR(dst)) { skb_dst_set(buff, dst); ip6_xmit(ctl_sk, buff, &fl6, NULL, tclass); diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index bcd5699..ec3eb7f 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -1221,7 +1221,7 @@ do_udp_sendmsg: security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); - dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_sk_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); dst = NULL; diff --git a/net/l2tp/l2tp_ip6.c b/net/l2tp/l2tp_ip6.c index d9b437e..0c231f6 100644 --- a/net/l2tp/l2tp_ip6.c +++ b/net/l2tp/l2tp_ip6.c @@ -598,7 +598,7 @@ static int l2tp_ip6_sendmsg(struct kiocb *iocb, struct sock *sk, security_sk_classify_flow(sk, flowi6_to_flowi(&fl6)); - dst = ip6_dst_lookup_flow(sk, &fl6, final_p, true); + dst = ip6_dst_lookup_flow(sk, &fl6, final_p); if (IS_ERR(dst)) { err = PTR_ERR(dst); goto out; diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c index 7567e6f..dd15cae 100644 --- a/net/sctp/ipv6.c +++ b/net/sctp/ipv6.c @@ -263,7 +263,7 @@ static void sctp_v6_get_dst(struct sctp_transport *t, union sctp_addr *saddr, } final_p = fl6_update_dst(fl6, np->opt, &final); - dst = ip6_dst_lookup_flow(sk, fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, fl6, final_p); if (!asoc || saddr) goto out; @@ -322,7 +322,7 @@ static void sctp_v6_get_dst(struct sctp_transport *t, union sctp_addr *saddr, fl6->saddr = baddr->v6.sin6_addr; fl6->fl6_sport = baddr->v6.sin6_port; final_p = fl6_update_dst(fl6, np->opt, &final); - dst = ip6_dst_lookup_flow(sk, fl6, final_p, false); + dst = ip6_dst_lookup_flow(sk, fl6, final_p); } out: -- cgit v1.1 From ea9884b3acf3311c8a11db67bfab21773f6f82ba Mon Sep 17 00:00:00 2001 From: Fan Du Date: Mon, 16 Dec 2013 18:47:48 +0800 Subject: xfrm: check user specified spi for IPComp IPComp connection between two hosts is broken if given spi bigger than 0xffff. OUTSPI=0x87 INSPI=0x11112 ip xfrm policy update dst 192.168.1.101 src 192.168.1.109 dir out action allow \ tmpl dst 192.168.1.101 src 192.168.1.109 proto comp spi $OUTSPI ip xfrm policy update src 192.168.1.101 dst 192.168.1.109 dir in action allow \ tmpl src 192.168.1.101 dst 192.168.1.109 proto comp spi $INSPI ip xfrm state add src 192.168.1.101 dst 192.168.1.109 proto comp spi $INSPI \ comp deflate ip xfrm state add dst 192.168.1.101 src 192.168.1.109 proto comp spi $OUTSPI \ comp deflate tcpdump can capture outbound ping packet, but inbound packet is dropped with XfrmOutNoStates errors. It looks like spi value used for IPComp is expected to be 16bits wide only. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_user.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 16c8460..4027c42 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -181,7 +181,9 @@ static int verify_newsa_info(struct xfrm_usersa_info *p, attrs[XFRMA_ALG_AEAD] || attrs[XFRMA_ALG_CRYPT] || attrs[XFRMA_ALG_COMP] || - attrs[XFRMA_TFCPAD]) + attrs[XFRMA_TFCPAD] || + (ntohl(p->id.spi) >= 0x10000)) + goto out; break; -- cgit v1.1 From 776e9dd90ca223b82166eb2835389493b5914cba Mon Sep 17 00:00:00 2001 From: Fan Du Date: Mon, 16 Dec 2013 18:47:49 +0800 Subject: xfrm: export verify_userspi_info for pkfey and netlink interface In order to check against valid IPcomp spi range, export verify_userspi_info for both pfkey and netlink interface. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- include/net/xfrm.h | 1 + net/key/af_key.c | 6 ++++++ net/xfrm/xfrm_state.c | 24 ++++++++++++++++++++++++ net/xfrm/xfrm_user.c | 25 +------------------------ 4 files changed, 32 insertions(+), 24 deletions(-) diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 59f5d0a..b7635ef 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -1563,6 +1563,7 @@ struct xfrm_policy *xfrm_policy_byid(struct net *net, u32 mark, u8, int dir, u32 id, int delete, int *err); int xfrm_policy_flush(struct net *net, u8 type, struct xfrm_audit *audit_info); u32 xfrm_get_acqseq(void); +int verify_spi_info(u8 proto, u32 min, u32 max); int xfrm_alloc_spi(struct xfrm_state *x, u32 minspi, u32 maxspi); struct xfrm_state *xfrm_find_acq(struct net *net, const struct xfrm_mark *mark, u8 mode, u32 reqid, u8 proto, diff --git a/net/key/af_key.c b/net/key/af_key.c index 5beabd8..1a04c13 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -1340,6 +1340,12 @@ static int pfkey_getspi(struct sock *sk, struct sk_buff *skb, const struct sadb_ max_spi = range->sadb_spirange_max; } + err = verify_spi_info(x->id.proto, min_spi, max_spi); + if (err) { + xfrm_state_put(x); + return err; + } + err = xfrm_alloc_spi(x, min_spi, max_spi); resp_skb = err ? ERR_PTR(err) : pfkey_xfrm_state2msg(x); diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index e845066..a62c25e 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -1489,6 +1489,30 @@ u32 xfrm_get_acqseq(void) } EXPORT_SYMBOL(xfrm_get_acqseq); +int verify_spi_info(u8 proto, u32 min, u32 max) +{ + switch (proto) { + case IPPROTO_AH: + case IPPROTO_ESP: + break; + + case IPPROTO_COMP: + /* IPCOMP spi is 16-bits. */ + if (max >= 0x10000) + return -EINVAL; + break; + + default: + return -EINVAL; + } + + if (min > max) + return -EINVAL; + + return 0; +} +EXPORT_SYMBOL(verify_spi_info); + int xfrm_alloc_spi(struct xfrm_state *x, u32 low, u32 high) { struct net *net = xs_net(x); diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 4027c42..97681a3 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -1079,29 +1079,6 @@ out_noput: return err; } -static int verify_userspi_info(struct xfrm_userspi_info *p) -{ - switch (p->info.id.proto) { - case IPPROTO_AH: - case IPPROTO_ESP: - break; - - case IPPROTO_COMP: - /* IPCOMP spi is 16-bits. */ - if (p->max >= 0x10000) - return -EINVAL; - break; - - default: - return -EINVAL; - } - - if (p->min > p->max) - return -EINVAL; - - return 0; -} - static int xfrm_alloc_userspi(struct sk_buff *skb, struct nlmsghdr *nlh, struct nlattr **attrs) { @@ -1116,7 +1093,7 @@ static int xfrm_alloc_userspi(struct sk_buff *skb, struct nlmsghdr *nlh, struct xfrm_mark m; p = nlmsg_data(nlh); - err = verify_userspi_info(p); + err = verify_spi_info(p->info.id.proto, p->min, p->max); if (err) goto out_noput; -- cgit v1.1 From b3c6efbc36e2c5ac820b1a800ac17cc3e040de0c Mon Sep 17 00:00:00 2001 From: Fan Du Date: Mon, 16 Dec 2013 18:47:50 +0800 Subject: xfrm: Add file to document IPsec corner case Create Documentation/networking/ipsec.txt to document IPsec corner issues and other info, which will be useful when user deploying IPsec. Signed-off-by: Fan Du Signed-off-by: Steffen Klassert --- Documentation/networking/ipsec.txt | 38 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 38 insertions(+) create mode 100644 Documentation/networking/ipsec.txt diff --git a/Documentation/networking/ipsec.txt b/Documentation/networking/ipsec.txt new file mode 100644 index 0000000..8dbc08b --- /dev/null +++ b/Documentation/networking/ipsec.txt @@ -0,0 +1,38 @@ + +Here documents known IPsec corner cases which need to be keep in mind when +deploy various IPsec configuration in real world production environment. + +1. IPcomp: Small IP packet won't get compressed at sender, and failed on + policy check on receiver. + +Quote from RFC3173: +2.2. Non-Expansion Policy + + If the total size of a compressed payload and the IPComp header, as + defined in section 3, is not smaller than the size of the original + payload, the IP datagram MUST be sent in the original non-compressed + form. To clarify: If an IP datagram is sent non-compressed, no + + IPComp header is added to the datagram. This policy ensures saving + the decompression processing cycles and avoiding incurring IP + datagram fragmentation when the expanded datagram is larger than the + MTU. + + Small IP datagrams are likely to expand as a result of compression. + Therefore, a numeric threshold should be applied before compression, + where IP datagrams of size smaller than the threshold are sent in the + original form without attempting compression. The numeric threshold + is implementation dependent. + +Current IPComp implementation is indeed by the book, while as in practice +when sending non-compressed packet to the peer(whether or not packet len +is smaller than the threshold or the compressed len is large than original +packet len), the packet is dropped when checking the policy as this packet +matches the selector but not coming from any XFRM layer, i.e., with no +security path. Such naked packet will not eventually make it to upper layer. +The result is much more wired to the user when ping peer with different +payload length. + +One workaround is try to set "level use" for each policy if user observed +above scenario. The consequence of doing so is small packet(uncompressed) +will skip policy checking on receiver side. -- cgit v1.1