From: Sasha Levin <sashal@kernel.org>
To: linux-kernel@vger.kernel.org, stable@vger.kernel.org
Cc: Eric Dumazet <edumazet@google.com>,
Michal Kubecek <mkubecek@suse.cz>, Firo Yang <firo.yang@suse.com>,
Jakub Kicinski <jakub.kicinski@netronome.com>,
Sasha Levin <sashal@kernel.org>,
rcu@vger.kernel.org, netdev@vger.kernel.org
Subject: [PATCH AUTOSEL 5.4 103/187] tcp/dccp: fix possible race __inet_lookup_established()
Date: Fri, 27 Dec 2019 12:39:31 -0500 [thread overview]
Message-ID: <20191227174055.4923-103-sashal@kernel.org> (raw)
In-Reply-To: <20191227174055.4923-1-sashal@kernel.org>
From: Eric Dumazet <edumazet@google.com>
[ Upstream commit 8dbd76e79a16b45b2ccb01d2f2e08dbf64e71e40 ]
Michal Kubecek and Firo Yang did a very nice analysis of crashes
happening in __inet_lookup_established().
Since a TCP socket can go from TCP_ESTABLISH to TCP_LISTEN
(via a close()/socket()/listen() cycle) without a RCU grace period,
I should not have changed listeners linkage in their hash table.
They must use the nulls protocol (Documentation/RCU/rculist_nulls.txt),
so that a lookup can detect a socket in a hash list was moved in
another one.
Since we added code in commit d296ba60d8e2 ("soreuseport: Resolve
merge conflict for v4/v6 ordering fix"), we have to add
hlist_nulls_add_tail_rcu() helper.
Fixes: 3b24d854cb35 ("tcp/dccp: do not touch listener sk_refcnt under synflood")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Michal Kubecek <mkubecek@suse.cz>
Reported-by: Firo Yang <firo.yang@suse.com>
Reviewed-by: Michal Kubecek <mkubecek@suse.cz>
Link: https://lore.kernel.org/netdev/20191120083919.GH27852@unicorn.suse.cz/
Signed-off-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
include/linux/rculist_nulls.h | 37 +++++++++++++++++++++++++++++++++++
include/net/inet_hashtables.h | 12 +++++++++---
include/net/sock.h | 5 +++++
net/ipv4/inet_diag.c | 3 ++-
net/ipv4/inet_hashtables.c | 16 +++++++--------
net/ipv4/tcp_ipv4.c | 7 ++++---
6 files changed, 65 insertions(+), 15 deletions(-)
diff --git a/include/linux/rculist_nulls.h b/include/linux/rculist_nulls.h
index bc8206a8f30e..61974c4c566b 100644
--- a/include/linux/rculist_nulls.h
+++ b/include/linux/rculist_nulls.h
@@ -100,6 +100,43 @@ static inline void hlist_nulls_add_head_rcu(struct hlist_nulls_node *n,
first->pprev = &n->next;
}
+/**
+ * hlist_nulls_add_tail_rcu
+ * @n: the element to add to the hash list.
+ * @h: the list to add to.
+ *
+ * Description:
+ * Adds the specified element to the specified hlist_nulls,
+ * while permitting racing traversals.
+ *
+ * The caller must take whatever precautions are necessary
+ * (such as holding appropriate locks) to avoid racing
+ * with another list-mutation primitive, such as hlist_nulls_add_head_rcu()
+ * or hlist_nulls_del_rcu(), running on this same list.
+ * However, it is perfectly legal to run concurrently with
+ * the _rcu list-traversal primitives, such as
+ * hlist_nulls_for_each_entry_rcu(), used to prevent memory-consistency
+ * problems on Alpha CPUs. Regardless of the type of CPU, the
+ * list-traversal primitive must be guarded by rcu_read_lock().
+ */
+static inline void hlist_nulls_add_tail_rcu(struct hlist_nulls_node *n,
+ struct hlist_nulls_head *h)
+{
+ struct hlist_nulls_node *i, *last = NULL;
+
+ /* Note: write side code, so rcu accessors are not needed. */
+ for (i = h->first; !is_a_nulls(i); i = i->next)
+ last = i;
+
+ if (last) {
+ n->next = last->next;
+ n->pprev = &last->next;
+ rcu_assign_pointer(hlist_next_rcu(last), n);
+ } else {
+ hlist_nulls_add_head_rcu(n, h);
+ }
+}
+
/**
* hlist_nulls_for_each_entry_rcu - iterate over rcu list of given type
* @tpos: the type * to use as a loop cursor.
diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
index af2b4c065a04..d0019d3395cf 100644
--- a/include/net/inet_hashtables.h
+++ b/include/net/inet_hashtables.h
@@ -103,13 +103,19 @@ struct inet_bind_hashbucket {
struct hlist_head chain;
};
-/*
- * Sockets can be hashed in established or listening table
+/* Sockets can be hashed in established or listening table.
+ * We must use different 'nulls' end-of-chain value for all hash buckets :
+ * A socket might transition from ESTABLISH to LISTEN state without
+ * RCU grace period. A lookup in ehash table needs to handle this case.
*/
+#define LISTENING_NULLS_BASE (1U << 29)
struct inet_listen_hashbucket {
spinlock_t lock;
unsigned int count;
- struct hlist_head head;
+ union {
+ struct hlist_head head;
+ struct hlist_nulls_head nulls_head;
+ };
};
/* This is for listening sockets, thus all sockets which possess wildcards. */
diff --git a/include/net/sock.h b/include/net/sock.h
index 718e62fbe869..fe7cd6899cde 100644
--- a/include/net/sock.h
+++ b/include/net/sock.h
@@ -723,6 +723,11 @@ static inline void __sk_nulls_add_node_rcu(struct sock *sk, struct hlist_nulls_h
hlist_nulls_add_head_rcu(&sk->sk_nulls_node, list);
}
+static inline void __sk_nulls_add_node_tail_rcu(struct sock *sk, struct hlist_nulls_head *list)
+{
+ hlist_nulls_add_tail_rcu(&sk->sk_nulls_node, list);
+}
+
static inline void sk_nulls_add_node_rcu(struct sock *sk, struct hlist_nulls_head *list)
{
sock_hold(sk);
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index 7dc79b973e6e..6a4c82f96e78 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -914,11 +914,12 @@ void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
for (i = s_i; i < INET_LHTABLE_SIZE; i++) {
struct inet_listen_hashbucket *ilb;
+ struct hlist_nulls_node *node;
num = 0;
ilb = &hashinfo->listening_hash[i];
spin_lock(&ilb->lock);
- sk_for_each(sk, &ilb->head) {
+ sk_nulls_for_each(sk, node, &ilb->nulls_head) {
struct inet_sock *inet = inet_sk(sk);
if (!net_eq(sock_net(sk), net))
diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 83fb00153018..2bbaaf0c7176 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -516,10 +516,11 @@ static int inet_reuseport_add_sock(struct sock *sk,
struct inet_listen_hashbucket *ilb)
{
struct inet_bind_bucket *tb = inet_csk(sk)->icsk_bind_hash;
+ const struct hlist_nulls_node *node;
struct sock *sk2;
kuid_t uid = sock_i_uid(sk);
- sk_for_each_rcu(sk2, &ilb->head) {
+ sk_nulls_for_each_rcu(sk2, node, &ilb->nulls_head) {
if (sk2 != sk &&
sk2->sk_family == sk->sk_family &&
ipv6_only_sock(sk2) == ipv6_only_sock(sk) &&
@@ -555,9 +556,9 @@ int __inet_hash(struct sock *sk, struct sock *osk)
}
if (IS_ENABLED(CONFIG_IPV6) && sk->sk_reuseport &&
sk->sk_family == AF_INET6)
- hlist_add_tail_rcu(&sk->sk_node, &ilb->head);
+ __sk_nulls_add_node_tail_rcu(sk, &ilb->nulls_head);
else
- hlist_add_head_rcu(&sk->sk_node, &ilb->head);
+ __sk_nulls_add_node_rcu(sk, &ilb->nulls_head);
inet_hash2(hashinfo, sk);
ilb->count++;
sock_set_flag(sk, SOCK_RCU_FREE);
@@ -606,11 +607,9 @@ void inet_unhash(struct sock *sk)
reuseport_detach_sock(sk);
if (ilb) {
inet_unhash2(hashinfo, sk);
- __sk_del_node_init(sk);
- ilb->count--;
- } else {
- __sk_nulls_del_node_init_rcu(sk);
+ ilb->count--;
}
+ __sk_nulls_del_node_init_rcu(sk);
sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
unlock:
spin_unlock_bh(lock);
@@ -750,7 +749,8 @@ void inet_hashinfo_init(struct inet_hashinfo *h)
for (i = 0; i < INET_LHTABLE_SIZE; i++) {
spin_lock_init(&h->listening_hash[i].lock);
- INIT_HLIST_HEAD(&h->listening_hash[i].head);
+ INIT_HLIST_NULLS_HEAD(&h->listening_hash[i].nulls_head,
+ i + LISTENING_NULLS_BASE);
h->listening_hash[i].count = 0;
}
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 67b2dc7a1727..eda64871f983 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -2149,13 +2149,14 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
struct tcp_iter_state *st = seq->private;
struct net *net = seq_file_net(seq);
struct inet_listen_hashbucket *ilb;
+ struct hlist_nulls_node *node;
struct sock *sk = cur;
if (!sk) {
get_head:
ilb = &tcp_hashinfo.listening_hash[st->bucket];
spin_lock(&ilb->lock);
- sk = sk_head(&ilb->head);
+ sk = sk_nulls_head(&ilb->nulls_head);
st->offset = 0;
goto get_sk;
}
@@ -2163,9 +2164,9 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
++st->num;
++st->offset;
- sk = sk_next(sk);
+ sk = sk_nulls_next(sk);
get_sk:
- sk_for_each_from(sk) {
+ sk_nulls_for_each_from(sk, node) {
if (!net_eq(sock_net(sk), net))
continue;
if (sk->sk_family == afinfo->family)
--
2.20.1
next prev parent reply other threads:[~2019-12-27 17:52 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20191227174055.4923-1-sashal@kernel.org>
2019-12-27 17:37 ` [PATCH AUTOSEL 5.4 008/187] mwifiex: fix possible heap overflow in mwifiex_process_country_ie() Sasha Levin
2020-01-06 22:51 ` Brian Norris
2020-01-09 15:29 ` Sasha Levin
2020-01-09 17:33 ` Brian Norris
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 014/187] netfilter: ctnetlink: netns exit must wait for callbacks Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 017/187] netfilter: nf_tables_offload: Check for the NETDEV_UNREGISTER event Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 018/187] mwifiex: Fix heap overflow in mmwifiex_process_tdls_action_frame() Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 025/187] netfilter: nf_queue: enqueue skbs with NULL dst Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 033/187] netfilter: nft_set_rbtree: bogus lookup/get on consecutive elements in named sets Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 034/187] netfilter: nf_tables: validate NFT_SET_ELEM_INTERVAL_END Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 035/187] netfilter: nf_tables: validate NFT_DATA_VALUE after nft_data_init() Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 036/187] netfilter: nf_tables: skip module reference count bump on object updates Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 037/187] netfilter: nf_tables_offload: return EOPNOTSUPP if rule specifies no actions Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 038/187] netfilter: bridge: make sure to pull arp header in br_nf_forward_arp() Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 049/187] mlxsw: spectrum_router: Remove unlikely user-triggerable warning Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 050/187] selftests: forwarding: Delete IPv6 address at the end Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 052/187] net: ethernet: ti: davinci_cpdma: fix warning "device driver frees DMA memory with different size" Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 053/187] af_packet: set defaule value for tmo Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 054/187] fjes: fix missed check in fjes_acpi_add Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 055/187] NFC: nxp-nci: Fix probing without ACPI Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 056/187] Revert "iwlwifi: assign directly to iwl_trans->cfg in QuZ detection" Sasha Levin
2019-12-30 5:48 ` Anders Kaseorg
2020-01-05 15:42 ` Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 057/187] iwlwifi: pcie: move power gating workaround earlier in the flow Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 067/187] bnxt_en: Fix MSIX request logic for RDMA driver Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 068/187] bnxt_en: Return error if FW returns more data than dump length Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 069/187] net: ena: fix napi handler misbehavior when the napi budget is zero Sasha Levin
2019-12-27 17:38 ` [PATCH AUTOSEL 5.4 071/187] bpf, riscv: Limit to 33 tail calls Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 072/187] bpf, mips: " Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 073/187] bpftool: Don't crash on missing jited insns or ksyms Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 080/187] samples: bpf: Replace symbol compare of trace_event Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 081/187] samples: bpf: fix syscall_tp due to unused syscall Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 087/187] net: usb: lan78xx: Fix suspend/resume PHY register access error Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 088/187] qede: Fix multicast mac configuration Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 090/187] mac80211: fix TID field in monitor mode transmit Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 091/187] cfg80211: fix double-free after changing network namespace Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 100/187] bpf: Clear skb->tstamp in bpf_redirect when necessary Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 101/187] bnxt: apply computed clamp value for coalece parameter Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 102/187] ipv6/addrconf: only check invalid header values when NETLINK_F_STRICT_CHK is set Sasha Levin
2019-12-27 17:39 ` Sasha Levin [this message]
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 104/187] 6pack,mkiss: fix possible deadlock Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 107/187] net: marvell: mvpp2: phylink requires the link interrupt Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 108/187] hv_netvsc: Fix tx_table init in rndis_set_subchannel() Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 109/187] bnx2x: Do not handle requests from VFs after parity Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 110/187] bnx2x: Fix logic to get total no. of PFs per engine Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 111/187] bonding: fix active-backup transition after link failure Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 112/187] gtp: do not allow adding duplicate tid and ms_addr pdp context Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 113/187] gtp: fix wrong condition in gtp_genl_dump_pdp() Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 114/187] gtp: fix an use-after-free in ipv4_pdp_find() Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 115/187] gtp: avoid zero size hashtable Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 116/187] net: phylink: fix interface passed to mac_link_up Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 117/187] net/smc: add fallback check to connect() Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 118/187] cxgb4: Fix kernel panic while accessing sge_info Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 119/187] net: usb: lan78xx: Fix error message format specifier Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 122/187] rfkill: Fix incorrect check to avoid NULL pointer dereference Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 128/187] net: dsa: b53: Fix egress flooding settings Sasha Levin
2019-12-27 17:39 ` [PATCH AUTOSEL 5.4 129/187] net: gemini: Fix memory leak in gmac_setup_txqs Sasha Levin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191227174055.4923-103-sashal@kernel.org \
--to=sashal@kernel.org \
--cc=edumazet@google.com \
--cc=firo.yang@suse.com \
--cc=jakub.kicinski@netronome.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mkubecek@suse.cz \
--cc=netdev@vger.kernel.org \
--cc=rcu@vger.kernel.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).