* [PATCH net-next 4/9] tipc: split up function tipc_msg_eval()
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, Paul Gortmaker, tipc-discussion
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
The function tipc_msg_eval() is in reality doing two related, but
different tasks. First it tries to find a new destination for named
messages, in case there was no first lookup, or if the first lookup
failed. Second, it does what its name suggests, evaluating the validity
of the message and its destination, and returning an appropriate error
code depending on the result.
This is confusing, and in this commit we choose to break it up into two
functions. A new function, tipc_msg_lookup_dest(), first attempts to find
a new destination, if the message is of the right type. If this lookup
fails, or if the message should not be subject to a second lookup, the
already existing tipc_msg_reverse() is called. This function performs
prepares the message for rejection, if applicable.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/msg.c | 42 +++++++++++++++++++++---------------------
net/tipc/msg.h | 9 +++++----
net/tipc/socket.c | 40 ++++++++++++++++++++++------------------
3 files changed, 48 insertions(+), 43 deletions(-)
diff --git a/net/tipc/msg.c b/net/tipc/msg.c
index 940d741..697223a 100644
--- a/net/tipc/msg.c
+++ b/net/tipc/msg.c
@@ -411,43 +411,43 @@ bool tipc_msg_reverse(u32 own_addr, struct sk_buff *buf, u32 *dnode,
return true;
exit:
kfree_skb(buf);
+ *dnode = 0;
return false;
}
/**
- * tipc_msg_eval: determine fate of message that found no destination
- * @buf: the buffer containing the message.
- * @dnode: return value: next-hop node, if message to be forwarded
- * @err: error code to use, if message to be rejected
- *
+ * tipc_msg_lookup_dest(): try to find new destination for named message
+ * @skb: the buffer containing the message.
+ * @dnode: return value: next-hop node, if destination found
+ * @err: return value: error code to use, if message to be rejected
* Does not consume buffer
- * Returns 0 (TIPC_OK) if message ok and we can try again, -TIPC error
- * code if message to be rejected
+ * Returns true if a destination is found, false otherwise
*/
-int tipc_msg_eval(struct net *net, struct sk_buff *buf, u32 *dnode)
+bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb,
+ u32 *dnode, int *err)
{
- struct tipc_msg *msg = buf_msg(buf);
+ struct tipc_msg *msg = buf_msg(skb);
u32 dport;
- if (msg_type(msg) != TIPC_NAMED_MSG)
- return -TIPC_ERR_NO_PORT;
- if (skb_linearize(buf))
- return -TIPC_ERR_NO_NAME;
- if (msg_data_sz(msg) > MAX_FORWARD_SIZE)
- return -TIPC_ERR_NO_NAME;
+ if (!msg_isdata(msg))
+ return false;
+ if (!msg_named(msg))
+ return false;
+ *err = -TIPC_ERR_NO_NAME;
+ if (skb_linearize(skb))
+ return false;
if (msg_reroute_cnt(msg) > 0)
- return -TIPC_ERR_NO_NAME;
-
+ return false;
*dnode = addr_domain(net, msg_lookup_scope(msg));
dport = tipc_nametbl_translate(net, msg_nametype(msg),
- msg_nameinst(msg),
- dnode);
+ msg_nameinst(msg), dnode);
if (!dport)
- return -TIPC_ERR_NO_NAME;
+ return false;
msg_incr_reroute_cnt(msg);
msg_set_destnode(msg, *dnode);
msg_set_destport(msg, dport);
- return TIPC_OK;
+ *err = TIPC_OK;
+ return true;
}
/* tipc_msg_reassemble() - clone a buffer chain of fragments and
diff --git a/net/tipc/msg.h b/net/tipc/msg.h
index f7ea954..6070299 100644
--- a/net/tipc/msg.h
+++ b/net/tipc/msg.h
@@ -750,18 +750,19 @@ static inline u32 msg_tot_origport(struct tipc_msg *m)
struct sk_buff *tipc_buf_acquire(u32 size);
bool tipc_msg_reverse(u32 own_addr, struct sk_buff *buf, u32 *dnode,
int err);
-int tipc_msg_eval(struct net *net, struct sk_buff *buf, u32 *dnode);
void tipc_msg_init(u32 own_addr, struct tipc_msg *m, u32 user, u32 type,
u32 hsize, u32 destnode);
-struct sk_buff *tipc_msg_create(uint user, uint type,
- uint hdr_sz, uint data_sz, u32 dnode,
- u32 onode, u32 dport, u32 oport, int errcode);
+struct sk_buff *tipc_msg_create(uint user, uint type, uint hdr_sz,
+ uint data_sz, u32 dnode, u32 onode,
+ u32 dport, u32 oport, int errcode);
int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf);
bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu);
bool tipc_msg_make_bundle(struct sk_buff_head *list,
struct sk_buff *skb, u32 mtu, u32 dnode);
int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
int offset, int dsz, int mtu, struct sk_buff_head *list);
+bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb, u32 *dnode,
+ int *err);
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list);
#endif
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 1d98bfc..e14b2ae 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1739,7 +1739,7 @@ static int filter_rcv(struct sock *sk, struct sk_buff **skb)
* @sk: socket
* @skb: message
*
- * Caller must hold socket lock, but not port lock.
+ * Caller must hold socket lock
*
* Returns 0
*/
@@ -1805,27 +1805,31 @@ int tipc_sk_rcv(struct net *net, struct sk_buff *skb)
struct tipc_net *tn;
struct sock *sk;
u32 dport = msg_destport(buf_msg(skb));
- int err;
+ int err = -TIPC_ERR_NO_PORT;
u32 dnode;
- /* Validate destination and message */
+ /* Find destination */
tsk = tipc_sk_lookup(net, dport);
- if (unlikely(!tsk)) {
- err = tipc_msg_eval(net, skb, &dnode);
- goto exit;
- }
- sk = &tsk->sk;
-
- spin_lock_bh(&sk->sk_lock.slock);
- err = tipc_sk_enqueue_skb(sk, &skb);
- spin_unlock_bh(&sk->sk_lock.slock);
- sock_put(sk);
-exit:
- if (unlikely(skb)) {
- tn = net_generic(net, tipc_net_id);
- if (!err || tipc_msg_reverse(tn->own_addr, skb, &dnode, -err))
- tipc_link_xmit_skb(net, skb, dnode, 0);
+ if (likely(tsk)) {
+ sk = &tsk->sk;
+ spin_lock_bh(&sk->sk_lock.slock);
+ err = tipc_sk_enqueue_skb(sk, &skb);
+ spin_unlock_bh(&sk->sk_lock.slock);
+ sock_put(sk);
}
+ if (likely(!skb))
+ return 0;
+ if (tipc_msg_lookup_dest(net, skb, &dnode, &err))
+ goto xmit;
+ if (!err) {
+ dnode = msg_destnode(buf_msg(skb));
+ goto xmit;
+ }
+ tn = net_generic(net, tipc_net_id);
+ if (!tipc_msg_reverse(tn->own_addr, skb, &dnode, -err))
+ return -EHOSTUNREACH;
+xmit:
+ tipc_link_xmit_skb(net, skb, dnode, dport);
return err ? -EHOSTUNREACH : 0;
}
--
1.9.1
------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
^ permalink raw reply related
* [PATCH net-next 5/9] tipc: use existing sk_write_queue for outgoing packet chain
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, Paul Gortmaker, tipc-discussion
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
The list for outgoing traffic buffers from a socket is currently
allocated on the stack. This forces us to initialize the queue for
each sent message, something costing extra CPU cycles in the most
critical data path. Later in this series we will introduce a new
safe input buffer queue, something that would force us to initialize
even the spinlock of the outgoing queue. A closer analysis reveals
that the queue always is filled and emptied within the same lock_sock()
session. It is therefore safe to use a queue aggregated in the socket
itself for this purpose. Since there already exists a queue for this
in struct sock, sk_write_queue, we introduce use of that queue in
this commit.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/socket.c | 31 +++++++++++++------------------
1 file changed, 13 insertions(+), 18 deletions(-)
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index e14b2ae..611a04f 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -69,8 +69,6 @@
* @pub_count: total # of publications port has made during its lifetime
* @probing_state:
* @probing_intv:
- * @port: port - interacts with 'sk' and with the rest of the TIPC stack
- * @peer_name: the peer of the connection, if any
* @conn_timeout: the time we can wait for an unresponded setup request
* @dupl_rcvcnt: number of bytes counted twice, in both backlog and rcv queue
* @link_cong: non-zero if owner must sleep because of link congestion
@@ -737,7 +735,7 @@ static int tipc_sendmcast(struct socket *sock, struct tipc_name_seq *seq,
struct tipc_sock *tsk = tipc_sk(sk);
struct net *net = sock_net(sk);
struct tipc_msg *mhdr = &tsk->phdr;
- struct sk_buff_head head;
+ struct sk_buff_head *pktchain = &sk->sk_write_queue;
struct iov_iter save = msg->msg_iter;
uint mtu;
int rc;
@@ -753,13 +751,12 @@ static int tipc_sendmcast(struct socket *sock, struct tipc_name_seq *seq,
new_mtu:
mtu = tipc_bclink_get_mtu();
- __skb_queue_head_init(&head);
- rc = tipc_msg_build(mhdr, msg, 0, dsz, mtu, &head);
+ rc = tipc_msg_build(mhdr, msg, 0, dsz, mtu, pktchain);
if (unlikely(rc < 0))
return rc;
do {
- rc = tipc_bclink_xmit(net, &head);
+ rc = tipc_bclink_xmit(net, pktchain);
if (likely(rc >= 0)) {
rc = dsz;
break;
@@ -773,7 +770,7 @@ new_mtu:
tipc_sk(sk)->link_cong = 1;
rc = tipc_wait_for_sndmsg(sock, &timeo);
if (rc)
- __skb_queue_purge(&head);
+ __skb_queue_purge(pktchain);
} while (!rc);
return rc;
}
@@ -897,7 +894,7 @@ static int tipc_sendmsg(struct kiocb *iocb, struct socket *sock,
struct net *net = sock_net(sk);
struct tipc_msg *mhdr = &tsk->phdr;
u32 dnode, dport;
- struct sk_buff_head head;
+ struct sk_buff_head *pktchain = &sk->sk_write_queue;
struct sk_buff *skb;
struct tipc_name_seq *seq = &dest->addr.nameseq;
struct iov_iter save;
@@ -972,15 +969,14 @@ static int tipc_sendmsg(struct kiocb *iocb, struct socket *sock,
save = m->msg_iter;
new_mtu:
mtu = tipc_node_get_mtu(net, dnode, tsk->portid);
- __skb_queue_head_init(&head);
- rc = tipc_msg_build(mhdr, m, 0, dsz, mtu, &head);
+ rc = tipc_msg_build(mhdr, m, 0, dsz, mtu, pktchain);
if (rc < 0)
goto exit;
do {
- skb = skb_peek(&head);
+ skb = skb_peek(pktchain);
TIPC_SKB_CB(skb)->wakeup_pending = tsk->link_cong;
- rc = tipc_link_xmit(net, &head, dnode, tsk->portid);
+ rc = tipc_link_xmit(net, pktchain, dnode, tsk->portid);
if (likely(rc >= 0)) {
if (sock->state != SS_READY)
sock->state = SS_CONNECTING;
@@ -996,7 +992,7 @@ new_mtu:
tsk->link_cong = 1;
rc = tipc_wait_for_sndmsg(sock, &timeo);
if (rc)
- __skb_queue_purge(&head);
+ __skb_queue_purge(pktchain);
} while (!rc);
exit:
if (iocb)
@@ -1054,7 +1050,7 @@ static int tipc_send_stream(struct kiocb *iocb, struct socket *sock,
struct net *net = sock_net(sk);
struct tipc_sock *tsk = tipc_sk(sk);
struct tipc_msg *mhdr = &tsk->phdr;
- struct sk_buff_head head;
+ struct sk_buff_head *pktchain = &sk->sk_write_queue;
DECLARE_SOCKADDR(struct sockaddr_tipc *, dest, m->msg_name);
u32 portid = tsk->portid;
int rc = -EINVAL;
@@ -1091,13 +1087,12 @@ next:
save = m->msg_iter;
mtu = tsk->max_pkt;
send = min_t(uint, dsz - sent, TIPC_MAX_USER_MSG_SIZE);
- __skb_queue_head_init(&head);
- rc = tipc_msg_build(mhdr, m, sent, send, mtu, &head);
+ rc = tipc_msg_build(mhdr, m, sent, send, mtu, pktchain);
if (unlikely(rc < 0))
goto exit;
do {
if (likely(!tsk_conn_cong(tsk))) {
- rc = tipc_link_xmit(net, &head, dnode, portid);
+ rc = tipc_link_xmit(net, pktchain, dnode, portid);
if (likely(!rc)) {
tsk->sent_unacked++;
sent += send;
@@ -1117,7 +1112,7 @@ next:
}
rc = tipc_wait_for_sndpkt(sock, &timeo);
if (rc)
- __skb_queue_purge(&head);
+ __skb_queue_purge(pktchain);
} while (!rc);
exit:
if (iocb)
--
1.9.1
------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
^ permalink raw reply related
* [PATCH net-next 7/9] tipc: simplify connection abort notifications when links break
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, Paul Gortmaker, tipc-discussion
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
The new input message queue in struct tipc_link can be used for
delivering connection abort messages to subscribing sockets. This
makes it possible to simplify the code for such cases.
This commit removes the temporary list in tipc_node_unlock()
used for transforming abort subscriptions to messages. Instead, the
abort messages are now created at the moment of lost contact, and
then added to the last failed link's generic input queue for delivery
to the sockets concerned.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/node.c | 69 ++++++++++++++++++++++++---------------------------------
1 file changed, 29 insertions(+), 40 deletions(-)
diff --git a/net/tipc/node.c b/net/tipc/node.c
index dcb83d9..c7fdf3d 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -194,28 +194,6 @@ void tipc_node_remove_conn(struct net *net, u32 dnode, u32 port)
tipc_node_unlock(node);
}
-void tipc_node_abort_sock_conns(struct net *net, struct list_head *conns)
-{
- struct tipc_net *tn = net_generic(net, tipc_net_id);
- struct tipc_sock_conn *conn, *safe;
- struct sk_buff *skb;
- struct sk_buff_head skbs;
-
- skb_queue_head_init(&skbs);
- list_for_each_entry_safe(conn, safe, conns, list) {
- skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
- TIPC_CONN_MSG, SHORT_H_SIZE, 0,
- tn->own_addr, conn->peer_node,
- conn->port, conn->peer_port,
- TIPC_ERR_NO_NODE);
- if (likely(skb))
- skb_queue_tail(&skbs, skb);
- list_del(&conn->list);
- kfree(conn);
- }
- tipc_sk_rcv(net, &skbs);
-}
-
/**
* tipc_node_link_up - handle addition of link
*
@@ -377,7 +355,11 @@ static void node_established_contact(struct tipc_node *n_ptr)
static void node_lost_contact(struct tipc_node *n_ptr)
{
char addr_string[16];
- u32 i;
+ struct tipc_sock_conn *conn, *safe;
+ struct list_head *conns = &n_ptr->conn_sks;
+ struct sk_buff *skb;
+ struct tipc_net *tn = net_generic(n_ptr->net, tipc_net_id);
+ uint i;
pr_debug("Lost contact with %s\n",
tipc_addr_string_fill(addr_string, n_ptr->addr));
@@ -413,11 +395,25 @@ static void node_lost_contact(struct tipc_node *n_ptr)
n_ptr->action_flags &= ~TIPC_WAIT_OWN_LINKS_DOWN;
- /* Notify subscribers and prevent re-contact with node until
- * cleanup is done.
- */
- n_ptr->action_flags |= TIPC_WAIT_PEER_LINKS_DOWN |
- TIPC_NOTIFY_NODE_DOWN;
+ /* Prevent re-contact with node until cleanup is done */
+ n_ptr->action_flags |= TIPC_WAIT_PEER_LINKS_DOWN;
+
+ /* Notify publications from this node */
+ n_ptr->action_flags |= TIPC_NOTIFY_NODE_DOWN;
+
+ /* Notify sockets connected to node */
+ list_for_each_entry_safe(conn, safe, conns, list) {
+ skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE, TIPC_CONN_MSG,
+ SHORT_H_SIZE, 0, tn->own_addr,
+ conn->peer_node, conn->port,
+ conn->peer_port, TIPC_ERR_NO_NODE);
+ if (likely(skb)) {
+ skb_queue_tail(n_ptr->inputq, skb);
+ n_ptr->action_flags |= TIPC_MSG_EVT;
+ }
+ list_del(&conn->list);
+ kfree(conn);
+ }
}
struct sk_buff *tipc_node_get_nodes(struct net *net, const void *req_tlv_area,
@@ -566,13 +562,12 @@ int tipc_node_get_linkname(struct net *net, u32 bearer_id, u32 addr,
void tipc_node_unlock(struct tipc_node *node)
{
struct net *net = node->net;
- LIST_HEAD(nsub_list);
- LIST_HEAD(conn_sks);
u32 addr = 0;
u32 flags = node->action_flags;
u32 link_id = 0;
+ struct list_head *publ_list;
struct sk_buff_head *inputq = node->inputq;
- struct sk_buff_head *namedq = node->inputq;
+ struct sk_buff_head *namedq;
if (likely(!flags || (flags == TIPC_MSG_EVT))) {
node->action_flags = 0;
@@ -585,11 +580,8 @@ void tipc_node_unlock(struct tipc_node *node)
addr = node->addr;
link_id = node->link_id;
namedq = node->namedq;
+ publ_list = &node->publ_list;
- if (flags & TIPC_NOTIFY_NODE_DOWN) {
- list_replace_init(&node->publ_list, &nsub_list);
- list_replace_init(&node->conn_sks, &conn_sks);
- }
node->action_flags &= ~(TIPC_MSG_EVT | TIPC_NOTIFY_NODE_DOWN |
TIPC_NOTIFY_NODE_UP | TIPC_NOTIFY_LINK_UP |
TIPC_NOTIFY_LINK_DOWN |
@@ -598,11 +590,8 @@ void tipc_node_unlock(struct tipc_node *node)
spin_unlock_bh(&node->lock);
- if (!list_empty(&conn_sks))
- tipc_node_abort_sock_conns(net, &conn_sks);
-
- if (!list_empty(&nsub_list))
- tipc_publ_notify(net, &nsub_list, addr);
+ if (flags & TIPC_NOTIFY_NODE_DOWN)
+ tipc_publ_notify(net, publ_list, addr);
if (flags & TIPC_WAKEUP_BCAST_USERS)
tipc_bclink_wakeup_users(net);
--
1.9.1
------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
^ permalink raw reply related
* [PATCH net-next 8/9] tipc: simplify socket multicast reception
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, Paul Gortmaker, tipc-discussion
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
The structure 'tipc_port_list' is used to collect port numbers
representing multicast destination socket on a receiving node.
The list is not based on a standard linked list, and is in reality
optimized for the uncommon case that there are more than one
multicast destinations per node. This makes the list handling
unecessarily complex, and as a consequence, even the socket
multicast reception becomes more complex.
In this commit, we replace 'tipc_port_list' with a new 'struct
tipc_plist', which is based on a standard list. We give the new
list stack (push/pop) semantics, someting that simplifies
the implementation of the function tipc_sk_mcast_rcv().
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/bcast.c | 49 +------------------------------------------------
net/tipc/bcast.h | 24 ++++--------------------
net/tipc/name_table.c | 44 +++++++++++++++++++++++++++++++++++++++++---
net/tipc/name_table.h | 18 ++++++++++++++++--
net/tipc/socket.c | 48 +++++++++++++++++++++---------------------------
net/tipc/socket.h | 2 +-
6 files changed, 84 insertions(+), 101 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 2dfaf27..3eaa931 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -1,7 +1,7 @@
/*
* net/tipc/bcast.c: TIPC broadcast code
*
- * Copyright (c) 2004-2006, 2014, Ericsson AB
+ * Copyright (c) 2004-2006, 2014-2015, Ericsson AB
* Copyright (c) 2004, Intel Corporation.
* Copyright (c) 2005, 2010-2011, Wind River Systems
* All rights reserved.
@@ -1037,50 +1037,3 @@ static void tipc_nmap_diff(struct tipc_node_map *nm_a,
}
}
}
-
-/**
- * tipc_port_list_add - add a port to a port list, ensuring no duplicates
- */
-void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port)
-{
- struct tipc_port_list *item = pl_ptr;
- int i;
- int item_sz = PLSIZE;
- int cnt = pl_ptr->count;
-
- for (; ; cnt -= item_sz, item = item->next) {
- if (cnt < PLSIZE)
- item_sz = cnt;
- for (i = 0; i < item_sz; i++)
- if (item->ports[i] == port)
- return;
- if (i < PLSIZE) {
- item->ports[i] = port;
- pl_ptr->count++;
- return;
- }
- if (!item->next) {
- item->next = kmalloc(sizeof(*item), GFP_ATOMIC);
- if (!item->next) {
- pr_warn("Incomplete multicast delivery, no memory\n");
- return;
- }
- item->next->next = NULL;
- }
- }
-}
-
-/**
- * tipc_port_list_free - free dynamically created entries in port_list chain
- *
- */
-void tipc_port_list_free(struct tipc_port_list *pl_ptr)
-{
- struct tipc_port_list *item;
- struct tipc_port_list *next;
-
- for (item = pl_ptr->next; item; item = next) {
- next = item->next;
- kfree(item);
- }
-}
diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h
index 6ea190d..8f4d4dc 100644
--- a/net/tipc/bcast.h
+++ b/net/tipc/bcast.h
@@ -1,7 +1,7 @@
/*
* net/tipc/bcast.h: Include file for TIPC broadcast code
*
- * Copyright (c) 2003-2006, 2014, Ericsson AB
+ * Copyright (c) 2003-2006, 2014-2015, Ericsson AB
* Copyright (c) 2005, 2010-2011, Wind River Systems
* All rights reserved.
*
@@ -41,22 +41,6 @@
#include "link.h"
#include "node.h"
-#define TIPC_BCLINK_RESET 1
-#define PLSIZE 32
-#define BCBEARER MAX_BEARERS
-
-/**
- * struct tipc_port_list - set of node local destination ports
- * @count: # of ports in set (only valid for first entry in list)
- * @next: pointer to next entry in list
- * @ports: array of port references
- */
-struct tipc_port_list {
- int count;
- struct tipc_port_list *next;
- u32 ports[PLSIZE];
-};
-
/**
* struct tipc_bcbearer_pair - a pair of bearers used by broadcast link
* @primary: pointer to primary bearer
@@ -71,6 +55,9 @@ struct tipc_bcbearer_pair {
struct tipc_bearer *secondary;
};
+#define TIPC_BCLINK_RESET 1
+#define BCBEARER MAX_BEARERS
+
/**
* struct tipc_bcbearer - bearer used by broadcast link
* @bearer: (non-standard) broadcast bearer structure
@@ -126,9 +113,6 @@ static inline int tipc_nmap_equal(struct tipc_node_map *nm_a,
return !memcmp(nm_a, nm_b, sizeof(*nm_a));
}
-void tipc_port_list_add(struct tipc_port_list *pl_ptr, u32 port);
-void tipc_port_list_free(struct tipc_port_list *pl_ptr);
-
int tipc_bclink_init(struct net *net);
void tipc_bclink_stop(struct net *net);
void tipc_bclink_set_flags(struct net *tn, unsigned int flags);
diff --git a/net/tipc/name_table.c b/net/tipc/name_table.c
index ce09b86..18a3d44 100644
--- a/net/tipc/name_table.c
+++ b/net/tipc/name_table.c
@@ -1,7 +1,7 @@
/*
* net/tipc/name_table.c: TIPC name table code
*
- * Copyright (c) 2000-2006, 2014, Ericsson AB
+ * Copyright (c) 2000-2006, 2014-2015, Ericsson AB
* Copyright (c) 2004-2008, 2010-2014, Wind River Systems
* All rights reserved.
*
@@ -618,7 +618,7 @@ not_found:
* Returns non-zero if any off-node ports overlap
*/
int tipc_nametbl_mc_translate(struct net *net, u32 type, u32 lower, u32 upper,
- u32 limit, struct tipc_port_list *dports)
+ u32 limit, struct tipc_plist *dports)
{
struct name_seq *seq;
struct sub_seq *sseq;
@@ -643,7 +643,7 @@ int tipc_nametbl_mc_translate(struct net *net, u32 type, u32 lower, u32 upper,
info = sseq->info;
list_for_each_entry(publ, &info->node_list, node_list) {
if (publ->scope <= limit)
- tipc_port_list_add(dports, publ->ref);
+ tipc_plist_push(dports, publ->ref);
}
if (info->cluster_list_size != info->node_list_size)
@@ -1212,3 +1212,41 @@ int tipc_nl_name_table_dump(struct sk_buff *skb, struct netlink_callback *cb)
return skb->len;
}
+
+void tipc_plist_push(struct tipc_plist *pl, u32 port)
+{
+ struct tipc_plist *nl;
+
+ if (likely(!pl->port)) {
+ pl->port = port;
+ return;
+ }
+ if (pl->port == port)
+ return;
+ list_for_each_entry(nl, &pl->list, list) {
+ if (nl->port == port)
+ return;
+ }
+ nl = kmalloc(sizeof(*nl), GFP_ATOMIC);
+ if (nl) {
+ nl->port = port;
+ list_add(&nl->list, &pl->list);
+ }
+}
+
+u32 tipc_plist_pop(struct tipc_plist *pl)
+{
+ struct tipc_plist *nl;
+ u32 port = 0;
+
+ if (likely(list_empty(&pl->list))) {
+ port = pl->port;
+ pl->port = 0;
+ return port;
+ }
+ nl = list_first_entry(&pl->list, typeof(*nl), list);
+ port = nl->port;
+ list_del(&nl->list);
+ kfree(nl);
+ return port;
+}
diff --git a/net/tipc/name_table.h b/net/tipc/name_table.h
index f67b3d8..52501fd 100644
--- a/net/tipc/name_table.h
+++ b/net/tipc/name_table.h
@@ -38,7 +38,7 @@
#define _TIPC_NAME_TABLE_H
struct tipc_subscription;
-struct tipc_port_list;
+struct tipc_plist;
/*
* TIPC name types reserved for internal TIPC use (both current and planned)
@@ -101,7 +101,7 @@ struct sk_buff *tipc_nametbl_get(struct net *net, const void *req_tlv_area,
int req_tlv_space);
u32 tipc_nametbl_translate(struct net *net, u32 type, u32 instance, u32 *node);
int tipc_nametbl_mc_translate(struct net *net, u32 type, u32 lower, u32 upper,
- u32 limit, struct tipc_port_list *dports);
+ u32 limit, struct tipc_plist *dports);
struct publication *tipc_nametbl_publish(struct net *net, u32 type, u32 lower,
u32 upper, u32 scope, u32 port_ref,
u32 key);
@@ -118,4 +118,18 @@ void tipc_nametbl_unsubscribe(struct tipc_subscription *s);
int tipc_nametbl_init(struct net *net);
void tipc_nametbl_stop(struct net *net);
+struct tipc_plist {
+ struct list_head list;
+ u32 port;
+};
+
+static inline void tipc_plist_init(struct tipc_plist *pl)
+{
+ INIT_LIST_HEAD(&pl->list);
+ pl->port = 0;
+}
+
+void tipc_plist_push(struct tipc_plist *pl, u32 port);
+u32 tipc_plist_pop(struct tipc_plist *pl);
+
#endif
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index c1a4611..26aec84 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -1,7 +1,7 @@
/*
* net/tipc/socket.c: TIPC socket API
*
- * Copyright (c) 2001-2007, 2012-2014, Ericsson AB
+ * Copyright (c) 2001-2007, 2012-2015, Ericsson AB
* Copyright (c) 2004-2008, 2010-2013, Wind River Systems
* All rights reserved.
*
@@ -778,48 +778,42 @@ new_mtu:
/* tipc_sk_mcast_rcv - Deliver multicast message to all destination sockets
*/
-void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *buf)
+void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *skb)
{
- struct tipc_msg *msg = buf_msg(buf);
- struct tipc_port_list dports = {0, NULL, };
- struct tipc_port_list *item;
- struct sk_buff *b;
- uint i, last, dst = 0;
+ struct tipc_msg *msg = buf_msg(skb);
+ struct tipc_plist dports;
+ struct sk_buff *cskb;
+ u32 portid;
u32 scope = TIPC_CLUSTER_SCOPE;
- struct sk_buff_head msgs;
+ struct sk_buff_head msgq;
+ uint hsz = skb_headroom(skb) + msg_hdr_sz(msg);
+
+ skb_queue_head_init(&msgq);
+ tipc_plist_init(&dports);
if (in_own_node(net, msg_orignode(msg)))
scope = TIPC_NODE_SCOPE;
if (unlikely(!msg_mcast(msg))) {
pr_warn("Received non-multicast msg in multicast\n");
- kfree_skb(buf);
goto exit;
}
/* Create destination port list: */
tipc_nametbl_mc_translate(net, msg_nametype(msg), msg_namelower(msg),
msg_nameupper(msg), scope, &dports);
- last = dports.count;
- if (!last) {
- kfree_skb(buf);
- return;
- }
-
- for (item = &dports; item; item = item->next) {
- for (i = 0; i < PLSIZE && ++dst <= last; i++) {
- b = (dst != last) ? skb_clone(buf, GFP_ATOMIC) : buf;
- if (!b) {
- pr_warn("Failed do clone mcast rcv buffer\n");
- continue;
- }
- msg_set_destport(msg, item->ports[i]);
- skb_queue_head_init(&msgs);
- skb_queue_tail(&msgs, b);
- tipc_sk_rcv(net, &msgs);
+ portid = tipc_plist_pop(&dports);
+ for (; portid; portid = tipc_plist_pop(&dports)) {
+ cskb = __pskb_copy(skb, hsz, GFP_ATOMIC);
+ if (!cskb) {
+ pr_warn("Failed do clone mcast rcv buffer\n");
+ continue;
}
+ msg_set_destport(buf_msg(cskb), portid);
+ skb_queue_tail(&msgq, cskb);
}
+ tipc_sk_rcv(net, &msgq);
exit:
- tipc_port_list_free(&dports);
+ kfree_skb(skb);
}
/**
diff --git a/net/tipc/socket.h b/net/tipc/socket.h
index e3dbdc0..95b0159 100644
--- a/net/tipc/socket.h
+++ b/net/tipc/socket.h
@@ -1,6 +1,6 @@
/* net/tipc/socket.h: Include file for TIPC socket code
*
- * Copyright (c) 2014, Ericsson AB
+ * Copyright (c) 2014-2015, Ericsson AB
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
--
1.9.1
------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
^ permalink raw reply related
* [PATCH net-next 9/9] tipc: eliminate race condition at multicast reception
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem; +Cc: Jon Maloy, netdev, Paul Gortmaker, tipc-discussion
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
In a previous commit in this series we resolved a race problem during
unicast message reception.
Here, we resolve the same problem at multicast reception. We apply the
same technique: an input queue serializing the delivery of arriving
buffers. The main difference is that here we do it in two steps.
First, the broadcast link feeds arriving buffers into the tail of an
arrival queue, which head is consumed at the socket level, and where
destination lookup is performed. Second, if the lookup is successful,
the resulting buffer clones are fed into a second queue, the input
queue. This queue is consumed at reception in the socket just like
in the unicast case. Both queues are protected by the same lock, -the
one of the input queue.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/bcast.c | 57 ++++++++++++++++++++++++----------------
net/tipc/bcast.h | 3 +++
net/tipc/msg.h | 17 ++++++++++++
net/tipc/name_table.h | 2 +-
net/tipc/node.c | 11 +++++---
net/tipc/node.h | 7 +++--
net/tipc/socket.c | 72 +++++++++++++++++++++++++++++++--------------------
net/tipc/socket.h | 4 +--
8 files changed, 114 insertions(+), 59 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 3eaa931..81b1fef 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -79,6 +79,13 @@ static void tipc_bclink_unlock(struct net *net)
tipc_link_reset_all(node);
}
+void tipc_bclink_input(struct net *net)
+{
+ struct tipc_net *tn = net_generic(net, tipc_net_id);
+
+ tipc_sk_mcast_rcv(net, &tn->bclink->arrvq, &tn->bclink->inputq);
+}
+
uint tipc_bclink_get_mtu(void)
{
return MAX_PKT_DEFAULT_MCAST;
@@ -356,7 +363,7 @@ static void bclink_peek_nack(struct net *net, struct tipc_msg *msg)
tipc_node_unlock(n_ptr);
}
-/* tipc_bclink_xmit - broadcast buffer chain to all nodes in cluster
+/* tipc_bclink_xmit - deliver buffer chain to all nodes in cluster
* and to identified node local sockets
* @net: the applicable net namespace
* @list: chain of buffers containing message
@@ -371,6 +378,8 @@ int tipc_bclink_xmit(struct net *net, struct sk_buff_head *list)
int rc = 0;
int bc = 0;
struct sk_buff *skb;
+ struct sk_buff_head arrvq;
+ struct sk_buff_head inputq;
/* Prepare clone of message for local node */
skb = tipc_msg_reassemble(list);
@@ -379,7 +388,7 @@ int tipc_bclink_xmit(struct net *net, struct sk_buff_head *list)
return -EHOSTUNREACH;
}
- /* Broadcast to all other nodes */
+ /* Broadcast to all nodes */
if (likely(bclink)) {
tipc_bclink_lock(net);
if (likely(bclink->bcast_nodes.count)) {
@@ -399,12 +408,15 @@ int tipc_bclink_xmit(struct net *net, struct sk_buff_head *list)
if (unlikely(!bc))
__skb_queue_purge(list);
- /* Deliver message clone */
- if (likely(!rc))
- tipc_sk_mcast_rcv(net, skb);
- else
+ if (unlikely(rc)) {
kfree_skb(skb);
-
+ return rc;
+ }
+ /* Deliver message clone */
+ __skb_queue_head_init(&arrvq);
+ skb_queue_head_init(&inputq);
+ __skb_queue_tail(&arrvq, skb);
+ tipc_sk_mcast_rcv(net, &arrvq, &inputq);
return rc;
}
@@ -449,7 +461,7 @@ void tipc_bclink_rcv(struct net *net, struct sk_buff *buf)
int deferred = 0;
int pos = 0;
struct sk_buff *iskb;
- struct sk_buff_head msgs;
+ struct sk_buff_head *arrvq, *inputq;
/* Screen out unwanted broadcast messages */
if (msg_mc_netid(msg) != tn->net_id)
@@ -486,6 +498,8 @@ void tipc_bclink_rcv(struct net *net, struct sk_buff *buf)
/* Handle in-sequence broadcast message */
seqno = msg_seqno(msg);
next_in = mod(node->bclink.last_in + 1);
+ arrvq = &tn->bclink->arrvq;
+ inputq = &tn->bclink->inputq;
if (likely(seqno == next_in)) {
receive:
@@ -493,21 +507,26 @@ receive:
if (likely(msg_isdata(msg))) {
tipc_bclink_lock(net);
bclink_accept_pkt(node, seqno);
+ spin_lock_bh(&inputq->lock);
+ __skb_queue_tail(arrvq, buf);
+ spin_unlock_bh(&inputq->lock);
+ node->action_flags |= TIPC_BCAST_MSG_EVT;
tipc_bclink_unlock(net);
tipc_node_unlock(node);
- if (likely(msg_mcast(msg)))
- tipc_sk_mcast_rcv(net, buf);
- else
- kfree_skb(buf);
} else if (msg_user(msg) == MSG_BUNDLER) {
tipc_bclink_lock(net);
bclink_accept_pkt(node, seqno);
bcl->stats.recv_bundles++;
bcl->stats.recv_bundled += msg_msgcnt(msg);
+ pos = 0;
+ while (tipc_msg_extract(buf, &iskb, &pos)) {
+ spin_lock_bh(&inputq->lock);
+ __skb_queue_tail(arrvq, iskb);
+ spin_unlock_bh(&inputq->lock);
+ }
+ node->action_flags |= TIPC_BCAST_MSG_EVT;
tipc_bclink_unlock(net);
tipc_node_unlock(node);
- while (tipc_msg_extract(buf, &iskb, &pos))
- tipc_sk_mcast_rcv(net, iskb);
} else if (msg_user(msg) == MSG_FRAGMENTER) {
tipc_buf_append(&node->bclink.reasm_buf, &buf);
if (unlikely(!buf && !node->bclink.reasm_buf))
@@ -523,14 +542,6 @@ receive:
}
tipc_bclink_unlock(net);
tipc_node_unlock(node);
- } else if (msg_user(msg) == NAME_DISTRIBUTOR) {
- tipc_bclink_lock(net);
- bclink_accept_pkt(node, seqno);
- tipc_bclink_unlock(net);
- tipc_node_unlock(node);
- skb_queue_head_init(&msgs);
- skb_queue_tail(&msgs, buf);
- tipc_named_rcv(net, &msgs);
} else {
tipc_bclink_lock(net);
bclink_accept_pkt(node, seqno);
@@ -950,6 +961,8 @@ int tipc_bclink_init(struct net *net)
skb_queue_head_init(&bcl->wakeupq);
bcl->next_out_no = 1;
spin_lock_init(&bclink->node.lock);
+ __skb_queue_head_init(&bclink->arrvq);
+ skb_queue_head_init(&bclink->inputq);
bcl->owner = &bclink->node;
bcl->owner->net = net;
bcl->max_pkt = MAX_PKT_DEFAULT_MCAST;
diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h
index 8f4d4dc..a910c0b 100644
--- a/net/tipc/bcast.h
+++ b/net/tipc/bcast.h
@@ -97,6 +97,8 @@ struct tipc_bclink {
struct tipc_link link;
struct tipc_node node;
unsigned int flags;
+ struct sk_buff_head arrvq;
+ struct sk_buff_head inputq;
struct tipc_node_map bcast_nodes;
struct tipc_node *retransmit_to;
};
@@ -134,5 +136,6 @@ uint tipc_bclink_get_mtu(void);
int tipc_bclink_xmit(struct net *net, struct sk_buff_head *list);
void tipc_bclink_wakeup_users(struct net *net);
int tipc_nl_add_bc_link(struct net *net, struct tipc_nl_msg *msg);
+void tipc_bclink_input(struct net *net);
#endif
diff --git a/net/tipc/msg.h b/net/tipc/msg.h
index ab46726..9ace47f 100644
--- a/net/tipc/msg.h
+++ b/net/tipc/msg.h
@@ -767,6 +767,23 @@ bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb, u32 *dnode,
int *err);
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list);
+/* tipc_skb_peek(): peek and reserve first buffer in list
+ * @list: list to be peeked in
+ * Returns pointer to first buffer in list, if any
+ */
+static inline struct sk_buff *tipc_skb_peek(struct sk_buff_head *list,
+ spinlock_t *lock)
+{
+ struct sk_buff *skb;
+
+ spin_lock_bh(lock);
+ skb = skb_peek(list);
+ if (skb)
+ skb_get(skb);
+ spin_unlock_bh(lock);
+ return skb;
+}
+
/* tipc_skb_peek_port(): find a destination port, ignoring all destinations
* up to and including 'filter'.
* Note: ignoring previously tried destinations minimizes the risk of
diff --git a/net/tipc/name_table.h b/net/tipc/name_table.h
index 52501fd..0304ddc 100644
--- a/net/tipc/name_table.h
+++ b/net/tipc/name_table.h
@@ -1,7 +1,7 @@
/*
* net/tipc/name_table.h: Include file for TIPC name table code
*
- * Copyright (c) 2000-2006, 2014, Ericsson AB
+ * Copyright (c) 2000-2006, 2014-2015, Ericsson AB
* Copyright (c) 2004-2005, 2010-2011, Wind River Systems
* All rights reserved.
*
diff --git a/net/tipc/node.c b/net/tipc/node.c
index c7fdf3d..5230849 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -582,10 +582,10 @@ void tipc_node_unlock(struct tipc_node *node)
namedq = node->namedq;
publ_list = &node->publ_list;
- node->action_flags &= ~(TIPC_MSG_EVT | TIPC_NOTIFY_NODE_DOWN |
- TIPC_NOTIFY_NODE_UP | TIPC_NOTIFY_LINK_UP |
- TIPC_NOTIFY_LINK_DOWN |
- TIPC_WAKEUP_BCAST_USERS |
+ node->action_flags &= ~(TIPC_MSG_EVT |
+ TIPC_NOTIFY_NODE_DOWN | TIPC_NOTIFY_NODE_UP |
+ TIPC_NOTIFY_LINK_DOWN | TIPC_NOTIFY_LINK_UP |
+ TIPC_WAKEUP_BCAST_USERS | TIPC_BCAST_MSG_EVT |
TIPC_NAMED_MSG_EVT);
spin_unlock_bh(&node->lock);
@@ -612,6 +612,9 @@ void tipc_node_unlock(struct tipc_node *node)
if (flags & TIPC_NAMED_MSG_EVT)
tipc_named_rcv(net, namedq);
+
+ if (flags & TIPC_BCAST_MSG_EVT)
+ tipc_bclink_input(net);
}
/* Caller should hold node lock for the passed node */
diff --git a/net/tipc/node.h b/net/tipc/node.h
index c2b0fcf..20ec13f 100644
--- a/net/tipc/node.h
+++ b/net/tipc/node.h
@@ -1,7 +1,7 @@
/*
* net/tipc/node.h: Include file for TIPC node management routines
*
- * Copyright (c) 2000-2006, 2014, Ericsson AB
+ * Copyright (c) 2000-2006, 2014-2015, Ericsson AB
* Copyright (c) 2005, 2010-2014, Wind River Systems
* All rights reserved.
*
@@ -63,7 +63,8 @@ enum {
TIPC_WAKEUP_BCAST_USERS = (1 << 5),
TIPC_NOTIFY_LINK_UP = (1 << 6),
TIPC_NOTIFY_LINK_DOWN = (1 << 7),
- TIPC_NAMED_MSG_EVT = (1 << 8)
+ TIPC_NAMED_MSG_EVT = (1 << 8),
+ TIPC_BCAST_MSG_EVT = (1 << 9)
};
/**
@@ -74,6 +75,7 @@ enum {
* @oos_state: state tracker for handling OOS b'cast messages
* @deferred_queue: deferred queue saved OOS b'cast message received from node
* @reasm_buf: broadcast reassembly queue head from node
+ * @inputq_map: bitmap indicating which inqueues should be kicked
* @recv_permitted: true if node is allowed to receive b'cast messages
*/
struct tipc_node_bclink {
@@ -84,6 +86,7 @@ struct tipc_node_bclink {
u32 deferred_size;
struct sk_buff_head deferred_queue;
struct sk_buff *reasm_buf;
+ int inputq_map;
bool recv_permitted;
};
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 26aec84..6666680 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -776,44 +776,60 @@ new_mtu:
return rc;
}
-/* tipc_sk_mcast_rcv - Deliver multicast message to all destination sockets
+/**
+ * tipc_sk_mcast_rcv - Deliver multicast messages to all destination sockets
+ * @arrvq: queue with arriving messages, to be cloned after destination lookup
+ * @inputq: queue with cloned messages, delivered to socket after dest lookup
+ *
+ * Multi-threaded: parallel calls with reference to same queues may occur
*/
-void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *skb)
+void tipc_sk_mcast_rcv(struct net *net, struct sk_buff_head *arrvq,
+ struct sk_buff_head *inputq)
{
- struct tipc_msg *msg = buf_msg(skb);
+ struct tipc_msg *msg;
struct tipc_plist dports;
- struct sk_buff *cskb;
u32 portid;
u32 scope = TIPC_CLUSTER_SCOPE;
- struct sk_buff_head msgq;
- uint hsz = skb_headroom(skb) + msg_hdr_sz(msg);
+ struct sk_buff_head tmpq;
+ uint hsz;
+ struct sk_buff *skb, *_skb;
- skb_queue_head_init(&msgq);
+ __skb_queue_head_init(&tmpq);
tipc_plist_init(&dports);
- if (in_own_node(net, msg_orignode(msg)))
- scope = TIPC_NODE_SCOPE;
-
- if (unlikely(!msg_mcast(msg))) {
- pr_warn("Received non-multicast msg in multicast\n");
- goto exit;
- }
- /* Create destination port list: */
- tipc_nametbl_mc_translate(net, msg_nametype(msg), msg_namelower(msg),
- msg_nameupper(msg), scope, &dports);
- portid = tipc_plist_pop(&dports);
- for (; portid; portid = tipc_plist_pop(&dports)) {
- cskb = __pskb_copy(skb, hsz, GFP_ATOMIC);
- if (!cskb) {
- pr_warn("Failed do clone mcast rcv buffer\n");
- continue;
+ skb = tipc_skb_peek(arrvq, &inputq->lock);
+ for (; skb; skb = tipc_skb_peek(arrvq, &inputq->lock)) {
+ msg = buf_msg(skb);
+ hsz = skb_headroom(skb) + msg_hdr_sz(msg);
+
+ if (in_own_node(net, msg_orignode(msg)))
+ scope = TIPC_NODE_SCOPE;
+
+ /* Create destination port list and message clones: */
+ tipc_nametbl_mc_translate(net,
+ msg_nametype(msg), msg_namelower(msg),
+ msg_nameupper(msg), scope, &dports);
+ portid = tipc_plist_pop(&dports);
+ for (; portid; portid = tipc_plist_pop(&dports)) {
+ _skb = __pskb_copy(skb, hsz, GFP_ATOMIC);
+ if (_skb) {
+ msg_set_destport(buf_msg(_skb), portid);
+ __skb_queue_tail(&tmpq, _skb);
+ continue;
+ }
+ pr_warn("Failed to clone mcast rcv buffer\n");
}
- msg_set_destport(buf_msg(cskb), portid);
- skb_queue_tail(&msgq, cskb);
+ /* Append to inputq if not already done by other thread */
+ spin_lock_bh(&inputq->lock);
+ if (skb_peek(arrvq) == skb) {
+ skb_queue_splice_tail_init(&tmpq, inputq);
+ kfree_skb(__skb_dequeue(arrvq));
+ }
+ spin_unlock_bh(&inputq->lock);
+ __skb_queue_purge(&tmpq);
+ kfree_skb(skb);
}
- tipc_sk_rcv(net, &msgq);
-exit:
- kfree_skb(skb);
+ tipc_sk_rcv(net, inputq);
}
/**
diff --git a/net/tipc/socket.h b/net/tipc/socket.h
index 95b0159..8be0da7 100644
--- a/net/tipc/socket.h
+++ b/net/tipc/socket.h
@@ -42,7 +42,6 @@
#define TIPC_FLOWCTRL_WIN (TIPC_CONNACK_INTV * 2)
#define TIPC_CONN_OVERLOAD_LIMIT ((TIPC_FLOWCTRL_WIN * 2 + 1) * \
SKB_TRUESIZE(TIPC_MAX_USER_MSG_SIZE))
-
int tipc_socket_init(void);
void tipc_socket_stop(void);
int tipc_sock_create_local(struct net *net, int type, struct socket **res);
@@ -51,7 +50,8 @@ int tipc_sock_accept_local(struct socket *sock, struct socket **newsock,
int flags);
int tipc_sk_rcv(struct net *net, struct sk_buff_head *inputq);
struct sk_buff *tipc_sk_socks_show(struct net *net);
-void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *buf);
+void tipc_sk_mcast_rcv(struct net *net, struct sk_buff_head *arrvq,
+ struct sk_buff_head *inputq);
void tipc_sk_reinit(struct net *net);
int tipc_sk_rht_init(struct net *net);
void tipc_sk_rht_destroy(struct net *net);
--
1.9.1
------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
^ permalink raw reply related
* [PATCH net-next 0/9] tipc: resolve message disordering problem
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem
Cc: netdev, Paul Gortmaker, erik.hugne, ying.xue, maloy,
tipc-discussion, Jon Maloy
When TIPC receives messages from multi-threaded device drivers it may
occasionally deliver messages to their destination sockets in the wrong
order. This happens despite correct resequencing at the link layer,
because the upcall path from link to socket is not protected by any
locks.
These commits solve this problem by introducing an 'input' message
queue in each link, through which messages must be delivered to the
upper layers.
Jon Maloy (9):
tipc: reduce usage of context info in socket and link
tipc: simplify message forwarding and rejection in socket layer
tipc: enqueue arrived buffers in socket in separate function
tipc: split up function tipc_msg_eval()
tipc: use existing sk_write_queue for outgoing packet chain
tipc: resolve race problem at unicast message reception
tipc: simplify connection abort notifications when links break
tipc: simplify socket multicast reception
tipc: eliminate race condition at multicast reception
net/tipc/bcast.c | 121 ++++++----------
net/tipc/bcast.h | 29 ++--
net/tipc/discover.c | 3 +-
net/tipc/link.c | 301 +++++++++++++++++----------------------
net/tipc/link.h | 14 +-
net/tipc/msg.c | 109 ++++++++------
net/tipc/msg.h | 107 ++++++++++++--
net/tipc/name_distr.c | 38 +++--
net/tipc/name_distr.h | 2 +-
net/tipc/name_table.c | 44 +++++-
net/tipc/name_table.h | 20 ++-
net/tipc/node.c | 103 +++++++-------
net/tipc/node.h | 22 ++-
net/tipc/socket.c | 382 +++++++++++++++++++++++++++++---------------------
net/tipc/socket.h | 8 +-
15 files changed, 740 insertions(+), 563 deletions(-)
--
1.9.1
^ permalink raw reply
* [PATCH net-next 1/9] tipc: reduce usage of context info in socket and link
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem
Cc: netdev, Paul Gortmaker, erik.hugne, ying.xue, maloy,
tipc-discussion, Jon Maloy
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
The most common usage of namespace information is when we fetch the
own node addess from the net structure. This leads to a lot of
passing around of a parameter of type 'struct net *' between
functions just to make them able to obtain this address.
However, in many cases this is unnecessary. The own node address
is readily available as a member of both struct tipc_sock and
tipc_link, and can be fetched from there instead.
The fact that the vast majority of functions in socket.c and link.c
anyway are maintaining a pointer to their respective base structures
makes this option even more compelling.
In this commit, we introduce the inline functions tsk_own_node()
and link_own_node() to make it easy for functions to fetch the node
address from those structs instead of having to pass along and
dereference the namespace struct.
In particular, we make calls to the msg_xx() functions in msg.{h,c}
context independent by directly passing them the own node address
as parameter when needed. Those functions should be regarded as
leaves in the code dependency tree, and it is hence desirable to
keep them namspace unaware.
Apart from a potential positive effect on cache behavior, these
changes make it easier to introduce the changes that will follow
later in this series.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/bcast.c | 7 ++++--
net/tipc/bcast.h | 2 +-
net/tipc/discover.c | 3 ++-
net/tipc/link.c | 56 ++++++++++++++++++++----------------------
net/tipc/link.h | 4 +++
net/tipc/msg.c | 33 ++++++++++++-------------
net/tipc/msg.h | 10 ++++----
net/tipc/name_distr.c | 5 ++--
net/tipc/node.c | 2 +-
net/tipc/socket.c | 67 +++++++++++++++++++++++++++------------------------
10 files changed, 98 insertions(+), 91 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 53f8bf0..3b886eb 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -283,10 +283,11 @@ exit:
*
* RCU and node lock set
*/
-void tipc_bclink_update_link_state(struct net *net, struct tipc_node *n_ptr,
+void tipc_bclink_update_link_state(struct tipc_node *n_ptr,
u32 last_sent)
{
struct sk_buff *buf;
+ struct net *net = n_ptr->net;
struct tipc_net *tn = net_generic(net, tipc_net_id);
/* Ignore "stale" link state info */
@@ -317,7 +318,7 @@ void tipc_bclink_update_link_state(struct net *net, struct tipc_node *n_ptr,
struct sk_buff *skb = skb_peek(&n_ptr->bclink.deferred_queue);
u32 to = skb ? buf_seqno(skb) - 1 : n_ptr->bclink.last_sent;
- tipc_msg_init(net, msg, BCAST_PROTOCOL, STATE_MSG,
+ tipc_msg_init(tn->own_addr, msg, BCAST_PROTOCOL, STATE_MSG,
INT_H_SIZE, n_ptr->addr);
msg_set_non_seq(msg, 1);
msg_set_mc_netid(msg, tn->net_id);
@@ -954,6 +955,8 @@ int tipc_bclink_init(struct net *net)
bcl->bearer_id = MAX_BEARERS;
rcu_assign_pointer(tn->bearer_list[MAX_BEARERS], &bcbearer->bearer);
bcl->state = WORKING_WORKING;
+ bcl->pmsg = (struct tipc_msg *)&bcl->proto_msg;
+ msg_set_prevnode(bcl->pmsg, tn->own_addr);
strlcpy(bcl->name, tipc_bclink_name, TIPC_MAX_LINK_NAME);
tn->bcbearer = bcbearer;
tn->bclink = bclink;
diff --git a/net/tipc/bcast.h b/net/tipc/bcast.h
index a4583a1..6ea190d 100644
--- a/net/tipc/bcast.h
+++ b/net/tipc/bcast.h
@@ -139,7 +139,7 @@ void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked);
void tipc_bclink_rcv(struct net *net, struct sk_buff *buf);
u32 tipc_bclink_get_last_sent(struct net *net);
u32 tipc_bclink_acks_missing(struct tipc_node *n_ptr);
-void tipc_bclink_update_link_state(struct net *net, struct tipc_node *n_ptr,
+void tipc_bclink_update_link_state(struct tipc_node *node,
u32 last_sent);
int tipc_bclink_stats(struct net *net, char *stats_buf, const u32 buf_size);
int tipc_bclink_reset_stats(struct net *net);
diff --git a/net/tipc/discover.c b/net/tipc/discover.c
index a580a40..feef375 100644
--- a/net/tipc/discover.c
+++ b/net/tipc/discover.c
@@ -85,7 +85,8 @@ static void tipc_disc_init_msg(struct net *net, struct sk_buff *buf, u32 type,
u32 dest_domain = b_ptr->domain;
msg = buf_msg(buf);
- tipc_msg_init(net, msg, LINK_CONFIG, type, INT_H_SIZE, dest_domain);
+ tipc_msg_init(tn->own_addr, msg, LINK_CONFIG, type,
+ INT_H_SIZE, dest_domain);
msg_set_non_seq(msg, 1);
msg_set_node_sig(msg, tn->random);
msg_set_dest_domain(msg, dest_domain);
diff --git a/net/tipc/link.c b/net/tipc/link.c
index 77c7ccd..41cb09a 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -101,13 +101,12 @@ static const struct nla_policy tipc_nl_prop_policy[TIPC_NLA_PROP_MAX + 1] = {
*/
#define START_CHANGEOVER 100000u
-static void link_handle_out_of_seq_msg(struct net *net,
- struct tipc_link *l_ptr,
- struct sk_buff *buf);
-static void tipc_link_proto_rcv(struct net *net, struct tipc_link *l_ptr,
- struct sk_buff *buf);
-static int tipc_link_tunnel_rcv(struct net *net, struct tipc_node *n_ptr,
- struct sk_buff **buf);
+static void link_handle_out_of_seq_msg(struct tipc_link *link,
+ struct sk_buff *skb);
+static void tipc_link_proto_rcv(struct tipc_link *link,
+ struct sk_buff *skb);
+static int tipc_link_tunnel_rcv(struct tipc_node *node,
+ struct sk_buff **skb);
static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol);
static void link_state_event(struct tipc_link *l_ptr, u32 event);
static void link_reset_statistics(struct tipc_link *l_ptr);
@@ -303,7 +302,7 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr,
l_ptr->pmsg = (struct tipc_msg *)&l_ptr->proto_msg;
msg = l_ptr->pmsg;
- tipc_msg_init(n_ptr->net, msg, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE,
+ tipc_msg_init(tn->own_addr, msg, LINK_PROTOCOL, RESET_MSG, INT_H_SIZE,
l_ptr->addr);
msg_set_size(msg, sizeof(l_ptr->proto_msg));
msg_set_session(msg, (tn->random & 0xffff));
@@ -379,12 +378,11 @@ void tipc_link_delete_list(struct net *net, unsigned int bearer_id,
static bool link_schedule_user(struct tipc_link *link, u32 oport,
uint chain_sz, uint imp)
{
- struct net *net = link->owner->net;
- struct tipc_net *tn = net_generic(net, tipc_net_id);
struct sk_buff *buf;
- buf = tipc_msg_create(net, SOCK_WAKEUP, 0, INT_H_SIZE, 0, tn->own_addr,
- tn->own_addr, oport, 0, 0);
+ buf = tipc_msg_create(SOCK_WAKEUP, 0, INT_H_SIZE, 0,
+ link_own_addr(link), link_own_addr(link),
+ oport, 0, 0);
if (!buf)
return false;
TIPC_SKB_CB(buf)->chain_sz = chain_sz;
@@ -778,7 +776,7 @@ int __tipc_link_xmit(struct net *net, struct tipc_link *link,
} else if (tipc_msg_bundle(outqueue, skb, mtu)) {
link->stats.sent_bundled++;
continue;
- } else if (tipc_msg_make_bundle(net, outqueue, skb, mtu,
+ } else if (tipc_msg_make_bundle(outqueue, skb, mtu,
link->addr)) {
link->stats.sent_bundled++;
link->stats.sent_bundles++;
@@ -877,7 +875,7 @@ static void tipc_link_sync_xmit(struct tipc_link *link)
return;
msg = buf_msg(skb);
- tipc_msg_init(link->owner->net, msg, BCAST_PROTOCOL, STATE_MSG,
+ tipc_msg_init(link_own_addr(link), msg, BCAST_PROTOCOL, STATE_MSG,
INT_H_SIZE, link->addr);
msg_set_last_bcast(msg, link->owner->bclink.acked);
__tipc_link_xmit_skb(link, skb);
@@ -1207,7 +1205,7 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
/* Process the incoming packet */
if (unlikely(!link_working_working(l_ptr))) {
if (msg_user(msg) == LINK_PROTOCOL) {
- tipc_link_proto_rcv(net, l_ptr, skb);
+ tipc_link_proto_rcv(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
tipc_node_unlock(n_ptr);
continue;
@@ -1227,7 +1225,7 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
/* Link is now in state WORKING_WORKING */
if (unlikely(seq_no != mod(l_ptr->next_in_no))) {
- link_handle_out_of_seq_msg(net, l_ptr, skb);
+ link_handle_out_of_seq_msg(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
tipc_node_unlock(n_ptr);
continue;
@@ -1275,7 +1273,7 @@ static int tipc_link_prepare_input(struct net *net, struct tipc_link *l,
msg = buf_msg(*buf);
switch (msg_user(msg)) {
case CHANGEOVER_PROTOCOL:
- if (tipc_link_tunnel_rcv(net, n, buf))
+ if (tipc_link_tunnel_rcv(n, buf))
res = 0;
break;
case MSG_FRAGMENTER:
@@ -1375,14 +1373,13 @@ u32 tipc_link_defer_pkt(struct sk_buff_head *list, struct sk_buff *skb)
/*
* link_handle_out_of_seq_msg - handle arrival of out-of-sequence packet
*/
-static void link_handle_out_of_seq_msg(struct net *net,
- struct tipc_link *l_ptr,
+static void link_handle_out_of_seq_msg(struct tipc_link *l_ptr,
struct sk_buff *buf)
{
u32 seq_no = buf_seqno(buf);
if (likely(msg_user(buf_msg(buf)) == LINK_PROTOCOL)) {
- tipc_link_proto_rcv(net, l_ptr, buf);
+ tipc_link_proto_rcv(l_ptr, buf);
return;
}
@@ -1507,10 +1504,9 @@ void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int probe_msg,
* Note that network plane id propagates through the network, and may
* change at any time. The node with lowest address rules
*/
-static void tipc_link_proto_rcv(struct net *net, struct tipc_link *l_ptr,
+static void tipc_link_proto_rcv(struct tipc_link *l_ptr,
struct sk_buff *buf)
{
- struct tipc_net *tn = net_generic(net, tipc_net_id);
u32 rec_gap = 0;
u32 max_pkt_info;
u32 max_pkt_ack;
@@ -1522,7 +1518,7 @@ static void tipc_link_proto_rcv(struct net *net, struct tipc_link *l_ptr,
goto exit;
if (l_ptr->net_plane != msg_net_plane(msg))
- if (tn->own_addr > msg_prevnode(msg))
+ if (link_own_addr(l_ptr) > msg_prevnode(msg))
l_ptr->net_plane = msg_net_plane(msg);
switch (msg_type(msg)) {
@@ -1625,7 +1621,7 @@ static void tipc_link_proto_rcv(struct net *net, struct tipc_link *l_ptr,
/* Protocol message before retransmits, reduce loss risk */
if (l_ptr->owner->bclink.recv_permitted)
- tipc_bclink_update_link_state(net, l_ptr->owner,
+ tipc_bclink_update_link_state(l_ptr->owner,
msg_last_bcast(msg));
if (rec_gap || (msg_probe(msg))) {
@@ -1690,7 +1686,7 @@ void tipc_link_failover_send_queue(struct tipc_link *l_ptr)
if (!tunnel)
return;
- tipc_msg_init(l_ptr->owner->net, &tunnel_hdr, CHANGEOVER_PROTOCOL,
+ tipc_msg_init(link_own_addr(l_ptr), &tunnel_hdr, CHANGEOVER_PROTOCOL,
ORIGINAL_MSG, INT_H_SIZE, l_ptr->addr);
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
msg_set_msgcnt(&tunnel_hdr, msgcount);
@@ -1748,7 +1744,7 @@ void tipc_link_dup_queue_xmit(struct tipc_link *l_ptr,
struct sk_buff *skb;
struct tipc_msg tunnel_hdr;
- tipc_msg_init(l_ptr->owner->net, &tunnel_hdr, CHANGEOVER_PROTOCOL,
+ tipc_msg_init(link_own_addr(l_ptr), &tunnel_hdr, CHANGEOVER_PROTOCOL,
DUPLICATE_MSG, INT_H_SIZE, l_ptr->addr);
msg_set_msgcnt(&tunnel_hdr, skb_queue_len(&l_ptr->outqueue));
msg_set_bearer_id(&tunnel_hdr, l_ptr->peer_bearer_id);
@@ -1802,7 +1798,7 @@ static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos)
/* tipc_link_dup_rcv(): Receive a tunnelled DUPLICATE_MSG packet.
* Owner node is locked.
*/
-static void tipc_link_dup_rcv(struct net *net, struct tipc_link *l_ptr,
+static void tipc_link_dup_rcv(struct tipc_link *l_ptr,
struct sk_buff *t_buf)
{
struct sk_buff *buf;
@@ -1817,7 +1813,7 @@ static void tipc_link_dup_rcv(struct net *net, struct tipc_link *l_ptr,
}
/* Add buffer to deferred queue, if applicable: */
- link_handle_out_of_seq_msg(net, l_ptr, buf);
+ link_handle_out_of_seq_msg(l_ptr, buf);
}
/* tipc_link_failover_rcv(): Receive a tunnelled ORIGINAL_MSG packet
@@ -1869,7 +1865,7 @@ exit:
* returned to the active link for delivery upwards.
* Owner node is locked.
*/
-static int tipc_link_tunnel_rcv(struct net *net, struct tipc_node *n_ptr,
+static int tipc_link_tunnel_rcv(struct tipc_node *n_ptr,
struct sk_buff **buf)
{
struct sk_buff *t_buf = *buf;
@@ -1887,7 +1883,7 @@ static int tipc_link_tunnel_rcv(struct net *net, struct tipc_node *n_ptr,
goto exit;
if (msg_type(t_msg) == DUPLICATE_MSG)
- tipc_link_dup_rcv(net, l_ptr, t_buf);
+ tipc_link_dup_rcv(l_ptr, t_buf);
else if (msg_type(t_msg) == ORIGINAL_MSG)
*buf = tipc_link_failover_rcv(l_ptr, t_buf);
else
diff --git a/net/tipc/link.h b/net/tipc/link.h
index 3e3432b..5b9a17f 100644
--- a/net/tipc/link.h
+++ b/net/tipc/link.h
@@ -278,6 +278,10 @@ static inline u32 lesser(u32 left, u32 right)
return less_eq(left, right) ? left : right;
}
+static inline u32 link_own_addr(struct tipc_link *l)
+{
+ return msg_prevnode(l->pmsg);
+}
/*
* Link status checking routines
diff --git a/net/tipc/msg.c b/net/tipc/msg.c
index da67c8d..940d741 100644
--- a/net/tipc/msg.c
+++ b/net/tipc/msg.c
@@ -70,25 +70,23 @@ struct sk_buff *tipc_buf_acquire(u32 size)
return skb;
}
-void tipc_msg_init(struct net *net, struct tipc_msg *m, u32 user, u32 type,
- u32 hsize, u32 destnode)
+void tipc_msg_init(u32 own_node, struct tipc_msg *m, u32 user, u32 type,
+ u32 hsize, u32 dnode)
{
- struct tipc_net *tn = net_generic(net, tipc_net_id);
-
memset(m, 0, hsize);
msg_set_version(m);
msg_set_user(m, user);
msg_set_hdr_sz(m, hsize);
msg_set_size(m, hsize);
- msg_set_prevnode(m, tn->own_addr);
+ msg_set_prevnode(m, own_node);
msg_set_type(m, type);
if (hsize > SHORT_H_SIZE) {
- msg_set_orignode(m, tn->own_addr);
- msg_set_destnode(m, destnode);
+ msg_set_orignode(m, own_node);
+ msg_set_destnode(m, dnode);
}
}
-struct sk_buff *tipc_msg_create(struct net *net, uint user, uint type,
+struct sk_buff *tipc_msg_create(uint user, uint type,
uint hdr_sz, uint data_sz, u32 dnode,
u32 onode, u32 dport, u32 oport, int errcode)
{
@@ -100,9 +98,8 @@ struct sk_buff *tipc_msg_create(struct net *net, uint user, uint type,
return NULL;
msg = buf_msg(buf);
- tipc_msg_init(net, msg, user, type, hdr_sz, dnode);
+ tipc_msg_init(onode, msg, user, type, hdr_sz, dnode);
msg_set_size(msg, hdr_sz + data_sz);
- msg_set_prevnode(msg, onode);
msg_set_origport(msg, oport);
msg_set_destport(msg, dport);
msg_set_errcode(msg, errcode);
@@ -195,7 +192,7 @@ err:
*
* Returns message data size or errno: -ENOMEM, -EFAULT
*/
-int tipc_msg_build(struct net *net, struct tipc_msg *mhdr, struct msghdr *m,
+int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
int offset, int dsz, int pktmax, struct sk_buff_head *list)
{
int mhsz = msg_hdr_sz(mhdr);
@@ -227,8 +224,8 @@ int tipc_msg_build(struct net *net, struct tipc_msg *mhdr, struct msghdr *m,
}
/* Prepare reusable fragment header */
- tipc_msg_init(net, &pkthdr, MSG_FRAGMENTER, FIRST_FRAGMENT, INT_H_SIZE,
- msg_destnode(mhdr));
+ tipc_msg_init(msg_prevnode(mhdr), &pkthdr, MSG_FRAGMENTER,
+ FIRST_FRAGMENT, INT_H_SIZE, msg_destnode(mhdr));
msg_set_size(&pkthdr, pktmax);
msg_set_fragm_no(&pkthdr, pktno);
@@ -338,7 +335,7 @@ bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu)
* Replaces buffer if successful
* Returns true if success, otherwise false
*/
-bool tipc_msg_make_bundle(struct net *net, struct sk_buff_head *list,
+bool tipc_msg_make_bundle(struct sk_buff_head *list,
struct sk_buff *skb, u32 mtu, u32 dnode)
{
struct sk_buff *bskb;
@@ -362,7 +359,8 @@ bool tipc_msg_make_bundle(struct net *net, struct sk_buff_head *list,
skb_trim(bskb, INT_H_SIZE);
bmsg = buf_msg(bskb);
- tipc_msg_init(net, bmsg, MSG_BUNDLER, 0, INT_H_SIZE, dnode);
+ tipc_msg_init(msg_prevnode(msg), bmsg, MSG_BUNDLER, 0,
+ INT_H_SIZE, dnode);
msg_set_seqno(bmsg, msg_seqno(msg));
msg_set_ack(bmsg, msg_ack(msg));
msg_set_bcast_ack(bmsg, msg_bcast_ack(msg));
@@ -379,10 +377,9 @@ bool tipc_msg_make_bundle(struct net *net, struct sk_buff_head *list,
* Consumes buffer if failure
* Returns true if success, otherwise false
*/
-bool tipc_msg_reverse(struct net *net, struct sk_buff *buf, u32 *dnode,
+bool tipc_msg_reverse(u32 own_addr, struct sk_buff *buf, u32 *dnode,
int err)
{
- struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_msg *msg = buf_msg(buf);
uint imp = msg_importance(msg);
struct tipc_msg ohdr;
@@ -402,7 +399,7 @@ bool tipc_msg_reverse(struct net *net, struct sk_buff *buf, u32 *dnode,
msg_set_errcode(msg, err);
msg_set_origport(msg, msg_destport(&ohdr));
msg_set_destport(msg, msg_origport(&ohdr));
- msg_set_prevnode(msg, tn->own_addr);
+ msg_set_prevnode(msg, own_addr);
if (!msg_short(msg)) {
msg_set_orignode(msg, msg_destnode(&ohdr));
msg_set_destnode(msg, msg_orignode(&ohdr));
diff --git a/net/tipc/msg.h b/net/tipc/msg.h
index 526ef34..f7ea954 100644
--- a/net/tipc/msg.h
+++ b/net/tipc/msg.h
@@ -748,19 +748,19 @@ static inline u32 msg_tot_origport(struct tipc_msg *m)
}
struct sk_buff *tipc_buf_acquire(u32 size);
-bool tipc_msg_reverse(struct net *net, struct sk_buff *buf, u32 *dnode,
+bool tipc_msg_reverse(u32 own_addr, struct sk_buff *buf, u32 *dnode,
int err);
int tipc_msg_eval(struct net *net, struct sk_buff *buf, u32 *dnode);
-void tipc_msg_init(struct net *net, struct tipc_msg *m, u32 user, u32 type,
+void tipc_msg_init(u32 own_addr, struct tipc_msg *m, u32 user, u32 type,
u32 hsize, u32 destnode);
-struct sk_buff *tipc_msg_create(struct net *net, uint user, uint type,
+struct sk_buff *tipc_msg_create(uint user, uint type,
uint hdr_sz, uint data_sz, u32 dnode,
u32 onode, u32 dport, u32 oport, int errcode);
int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf);
bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu);
-bool tipc_msg_make_bundle(struct net *net, struct sk_buff_head *list,
+bool tipc_msg_make_bundle(struct sk_buff_head *list,
struct sk_buff *skb, u32 mtu, u32 dnode);
-int tipc_msg_build(struct net *net, struct tipc_msg *mhdr, struct msghdr *m,
+int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
int offset, int dsz, int mtu, struct sk_buff_head *list);
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list);
diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c
index 7f31cd4..dd8564c 100644
--- a/net/tipc/name_distr.c
+++ b/net/tipc/name_distr.c
@@ -71,13 +71,14 @@ static void publ_to_item(struct distr_item *i, struct publication *p)
static struct sk_buff *named_prepare_buf(struct net *net, u32 type, u32 size,
u32 dest)
{
+ struct tipc_net *tn = net_generic(net, tipc_net_id);
struct sk_buff *buf = tipc_buf_acquire(INT_H_SIZE + size);
struct tipc_msg *msg;
if (buf != NULL) {
msg = buf_msg(buf);
- tipc_msg_init(net, msg, NAME_DISTRIBUTOR, type, INT_H_SIZE,
- dest);
+ tipc_msg_init(tn->own_addr, msg, NAME_DISTRIBUTOR, type,
+ INT_H_SIZE, dest);
msg_set_size(msg, INT_H_SIZE + size);
}
return buf;
diff --git a/net/tipc/node.c b/net/tipc/node.c
index 842bd7a..1c409c4 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -204,7 +204,7 @@ void tipc_node_abort_sock_conns(struct net *net, struct list_head *conns)
struct sk_buff *buf;
list_for_each_entry_safe(conn, safe, conns, list) {
- buf = tipc_msg_create(net, TIPC_CRITICAL_IMPORTANCE,
+ buf = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
TIPC_CONN_MSG, SHORT_H_SIZE, 0,
tn->own_addr, conn->peer_node,
conn->port, conn->peer_port,
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index caa4d66..b384e65 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -177,6 +177,11 @@ static const struct nla_policy tipc_nl_sock_policy[TIPC_NLA_SOCK_MAX + 1] = {
* - port reference
*/
+static u32 tsk_own_node(struct tipc_sock *tsk)
+{
+ return msg_prevnode(&tsk->phdr);
+}
+
static u32 tsk_peer_node(struct tipc_sock *tsk)
{
return msg_destnode(&tsk->phdr);
@@ -249,11 +254,11 @@ static void tsk_rej_rx_queue(struct sock *sk)
{
struct sk_buff *skb;
u32 dnode;
- struct net *net = sock_net(sk);
+ u32 own_node = tsk_own_node(tipc_sk(sk));
while ((skb = __skb_dequeue(&sk->sk_receive_queue))) {
- if (tipc_msg_reverse(net, skb, &dnode, TIPC_ERR_NO_PORT))
- tipc_link_xmit_skb(net, skb, dnode, 0);
+ if (tipc_msg_reverse(own_node, skb, &dnode, TIPC_ERR_NO_PORT))
+ tipc_link_xmit_skb(sock_net(sk), skb, dnode, 0);
}
}
@@ -305,6 +310,7 @@ static bool tsk_peer_msg(struct tipc_sock *tsk, struct tipc_msg *msg)
static int tipc_sk_create(struct net *net, struct socket *sock,
int protocol, int kern)
{
+ struct tipc_net *tn;
const struct proto_ops *ops;
socket_state state;
struct sock *sk;
@@ -346,7 +352,8 @@ static int tipc_sk_create(struct net *net, struct socket *sock,
tsk->max_pkt = MAX_PKT_DEFAULT;
INIT_LIST_HEAD(&tsk->publications);
msg = &tsk->phdr;
- tipc_msg_init(net, msg, TIPC_LOW_IMPORTANCE, TIPC_NAMED_MSG,
+ tn = net_generic(sock_net(sk), tipc_net_id);
+ tipc_msg_init(tn->own_addr, msg, TIPC_LOW_IMPORTANCE, TIPC_NAMED_MSG,
NAMED_H_SIZE, 0);
/* Finish initializing socket data structures */
@@ -471,7 +478,6 @@ static int tipc_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct net *net;
- struct tipc_net *tn;
struct tipc_sock *tsk;
struct sk_buff *skb;
u32 dnode, probing_state;
@@ -484,8 +490,6 @@ static int tipc_release(struct socket *sock)
return 0;
net = sock_net(sk);
- tn = net_generic(net, tipc_net_id);
-
tsk = tipc_sk(sk);
lock_sock(sk);
@@ -507,7 +511,7 @@ static int tipc_release(struct socket *sock)
tsk->connected = 0;
tipc_node_remove_conn(net, dnode, tsk->portid);
}
- if (tipc_msg_reverse(net, skb, &dnode,
+ if (tipc_msg_reverse(tsk_own_node(tsk), skb, &dnode,
TIPC_ERR_NO_PORT))
tipc_link_xmit_skb(net, skb, dnode, 0);
}
@@ -520,9 +524,9 @@ static int tipc_release(struct socket *sock)
sock_put(sk);
tipc_sk_remove(tsk);
if (tsk->connected) {
- skb = tipc_msg_create(net, TIPC_CRITICAL_IMPORTANCE,
+ skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
TIPC_CONN_MSG, SHORT_H_SIZE, 0, dnode,
- tn->own_addr, tsk_peer_port(tsk),
+ tsk_own_node(tsk), tsk_peer_port(tsk),
tsk->portid, TIPC_ERR_NO_PORT);
if (skb)
tipc_link_xmit_skb(net, skb, dnode, tsk->portid);
@@ -730,8 +734,9 @@ static int tipc_sendmcast(struct socket *sock, struct tipc_name_seq *seq,
struct msghdr *msg, size_t dsz, long timeo)
{
struct sock *sk = sock->sk;
+ struct tipc_sock *tsk = tipc_sk(sk);
struct net *net = sock_net(sk);
- struct tipc_msg *mhdr = &tipc_sk(sk)->phdr;
+ struct tipc_msg *mhdr = &tsk->phdr;
struct sk_buff_head head;
struct iov_iter save = msg->msg_iter;
uint mtu;
@@ -749,7 +754,7 @@ static int tipc_sendmcast(struct socket *sock, struct tipc_name_seq *seq,
new_mtu:
mtu = tipc_bclink_get_mtu();
__skb_queue_head_init(&head);
- rc = tipc_msg_build(net, mhdr, msg, 0, dsz, mtu, &head);
+ rc = tipc_msg_build(mhdr, msg, 0, dsz, mtu, &head);
if (unlikely(rc < 0))
return rc;
@@ -836,7 +841,7 @@ static int tipc_sk_proto_rcv(struct tipc_sock *tsk, u32 *dnode,
if (conn_cong)
tsk->sk.sk_write_space(&tsk->sk);
} else if (msg_type(msg) == CONN_PROBE) {
- if (!tipc_msg_reverse(sock_net(&tsk->sk), buf, dnode, TIPC_OK))
+ if (!tipc_msg_reverse(tsk_own_node(tsk), buf, dnode, TIPC_OK))
return TIPC_OK;
msg_set_type(msg, CONN_PROBE_REPLY);
return TIPC_FWD_MSG;
@@ -971,7 +976,7 @@ static int tipc_sendmsg(struct kiocb *iocb, struct socket *sock,
new_mtu:
mtu = tipc_node_get_mtu(net, dnode, tsk->portid);
__skb_queue_head_init(&head);
- rc = tipc_msg_build(net, mhdr, m, 0, dsz, mtu, &head);
+ rc = tipc_msg_build(mhdr, m, 0, dsz, mtu, &head);
if (rc < 0)
goto exit;
@@ -1090,7 +1095,7 @@ next:
mtu = tsk->max_pkt;
send = min_t(uint, dsz - sent, TIPC_MAX_USER_MSG_SIZE);
__skb_queue_head_init(&head);
- rc = tipc_msg_build(net, mhdr, m, sent, send, mtu, &head);
+ rc = tipc_msg_build(mhdr, m, sent, send, mtu, &head);
if (unlikely(rc < 0))
goto exit;
do {
@@ -1263,7 +1268,6 @@ static int tipc_sk_anc_data_recv(struct msghdr *m, struct tipc_msg *msg,
static void tipc_sk_send_ack(struct tipc_sock *tsk, uint ack)
{
struct net *net = sock_net(&tsk->sk);
- struct tipc_net *tn = net_generic(net, tipc_net_id);
struct sk_buff *skb = NULL;
struct tipc_msg *msg;
u32 peer_port = tsk_peer_port(tsk);
@@ -1271,9 +1275,9 @@ static void tipc_sk_send_ack(struct tipc_sock *tsk, uint ack)
if (!tsk->connected)
return;
- skb = tipc_msg_create(net, CONN_MANAGER, CONN_ACK, INT_H_SIZE, 0,
- dnode, tn->own_addr, peer_port, tsk->portid,
- TIPC_OK);
+ skb = tipc_msg_create(CONN_MANAGER, CONN_ACK, INT_H_SIZE, 0,
+ dnode, tsk_own_node(tsk), peer_port,
+ tsk->portid, TIPC_OK);
if (!skb)
return;
msg = buf_msg(skb);
@@ -1756,7 +1760,7 @@ static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *skb)
return 0;
}
- if ((rc < 0) && !tipc_msg_reverse(net, skb, &onode, -rc))
+ if ((rc < 0) && !tipc_msg_reverse(tsk_own_node(tsk), skb, &onode, -rc))
return 0;
tipc_link_xmit_skb(net, skb, onode, 0);
@@ -1773,6 +1777,7 @@ static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *skb)
int tipc_sk_rcv(struct net *net, struct sk_buff *skb)
{
struct tipc_sock *tsk;
+ struct tipc_net *tn;
struct sock *sk;
u32 dport = msg_destport(buf_msg(skb));
int rc = TIPC_OK;
@@ -1804,7 +1809,8 @@ int tipc_sk_rcv(struct net *net, struct sk_buff *skb)
if (likely(!rc))
return 0;
exit:
- if ((rc < 0) && !tipc_msg_reverse(net, skb, &dnode, -rc))
+ tn = net_generic(net, tipc_net_id);
+ if ((rc < 0) && !tipc_msg_reverse(tn->own_addr, skb, &dnode, -rc))
return -EHOSTUNREACH;
tipc_link_xmit_skb(net, skb, dnode, 0);
@@ -2065,7 +2071,6 @@ static int tipc_shutdown(struct socket *sock, int how)
{
struct sock *sk = sock->sk;
struct net *net = sock_net(sk);
- struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_sock *tsk = tipc_sk(sk);
struct sk_buff *skb;
u32 dnode;
@@ -2088,16 +2093,17 @@ restart:
kfree_skb(skb);
goto restart;
}
- if (tipc_msg_reverse(net, skb, &dnode,
+ if (tipc_msg_reverse(tsk_own_node(tsk), skb, &dnode,
TIPC_CONN_SHUTDOWN))
tipc_link_xmit_skb(net, skb, dnode,
tsk->portid);
tipc_node_remove_conn(net, dnode, tsk->portid);
} else {
dnode = tsk_peer_node(tsk);
- skb = tipc_msg_create(net, TIPC_CRITICAL_IMPORTANCE,
+
+ skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
TIPC_CONN_MSG, SHORT_H_SIZE,
- 0, dnode, tn->own_addr,
+ 0, dnode, tsk_own_node(tsk),
tsk_peer_port(tsk),
tsk->portid, TIPC_CONN_SHUTDOWN);
tipc_link_xmit_skb(net, skb, dnode, tsk->portid);
@@ -2129,10 +2135,9 @@ static void tipc_sk_timeout(unsigned long data)
{
struct tipc_sock *tsk = (struct tipc_sock *)data;
struct sock *sk = &tsk->sk;
- struct net *net = sock_net(sk);
- struct tipc_net *tn = net_generic(net, tipc_net_id);
struct sk_buff *skb = NULL;
u32 peer_port, peer_node;
+ u32 own_node = tsk_own_node(tsk);
bh_lock_sock(sk);
if (!tsk->connected) {
@@ -2144,13 +2149,13 @@ static void tipc_sk_timeout(unsigned long data)
if (tsk->probing_state == TIPC_CONN_PROBING) {
/* Previous probe not answered -> self abort */
- skb = tipc_msg_create(net, TIPC_CRITICAL_IMPORTANCE,
+ skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
TIPC_CONN_MSG, SHORT_H_SIZE, 0,
- tn->own_addr, peer_node, tsk->portid,
+ own_node, peer_node, tsk->portid,
peer_port, TIPC_ERR_NO_PORT);
} else {
- skb = tipc_msg_create(net, CONN_MANAGER, CONN_PROBE, INT_H_SIZE,
- 0, peer_node, tn->own_addr,
+ skb = tipc_msg_create(CONN_MANAGER, CONN_PROBE,
+ INT_H_SIZE, 0, peer_node, own_node,
peer_port, tsk->portid, TIPC_OK);
tsk->probing_state = TIPC_CONN_PROBING;
sk_reset_timer(sk, &sk->sk_timer, jiffies + tsk->probing_intv);
--
1.9.1
^ permalink raw reply related
* [PATCH net-next 6/9] tipc: resolve race problem at unicast message reception
From: Jon Maloy @ 2015-02-05 13:36 UTC (permalink / raw)
To: davem
Cc: netdev, Paul Gortmaker, erik.hugne, ying.xue, maloy,
tipc-discussion, Jon Maloy
In-Reply-To: <1423143404-23322-1-git-send-email-jon.maloy@ericsson.com>
TIPC handles message cardinality and sequencing at the link layer,
before passing messages upwards to the destination sockets. During the
upcall from link to socket no locks are held. It is therefore possible,
and we see it happen occasionally, that messages arriving in different
threads and delivered in sequence still bypass each other before they
reach the destination socket. This must not happen, since it violates
the sequentiality guarantee.
We solve this by adding a new input buffer queue to the link structure.
Arriving messages are added safely to the tail of that queue by the
link, while the head of the queue is consumed, also safely, by the
receiving socket. Sequentiality is secured per socket by only allowing
buffers to be dequeued inside the socket lock. Since there may be multiple
simultaneous readers of the queue, we use a 'filter' parameter to reduce
the risk that they peek the same buffer from the queue, hence also
reducing the risk of contention on the receiving socket locks.
This solves the sequentiality problem, and seems to cause no measurable
performance degradation.
A nice side effect of this change is that lock handling in the functions
tipc_rcv() and tipc_bcast_rcv() now becomes uniform, something that
will enable future simplifications of those functions.
Reviewed-by: Ying Xue <ying.xue@windriver.com>
Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
---
net/tipc/bcast.c | 20 ++--
net/tipc/link.c | 247 +++++++++++++++++++++-----------------------------
net/tipc/link.h | 10 +-
net/tipc/msg.c | 34 +++++++
net/tipc/msg.h | 73 +++++++++++++++
net/tipc/name_distr.c | 33 ++++---
net/tipc/name_distr.h | 2 +-
net/tipc/node.c | 43 +++++----
net/tipc/node.h | 17 ++--
net/tipc/socket.c | 132 +++++++++++++++++----------
net/tipc/socket.h | 2 +-
11 files changed, 372 insertions(+), 241 deletions(-)
diff --git a/net/tipc/bcast.c b/net/tipc/bcast.c
index 3b886eb..2dfaf27 100644
--- a/net/tipc/bcast.c
+++ b/net/tipc/bcast.c
@@ -189,10 +189,8 @@ static void bclink_retransmit_pkt(struct tipc_net *tn, u32 after, u32 to)
void tipc_bclink_wakeup_users(struct net *net)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
- struct sk_buff *skb;
- while ((skb = skb_dequeue(&tn->bclink->link.waiting_sks)))
- tipc_sk_rcv(net, skb);
+ tipc_sk_rcv(net, &tn->bclink->link.wakeupq);
}
/**
@@ -271,9 +269,8 @@ void tipc_bclink_acknowledge(struct tipc_node *n_ptr, u32 acked)
tipc_link_push_packets(tn->bcl);
bclink_set_last_sent(net);
}
- if (unlikely(released && !skb_queue_empty(&tn->bcl->waiting_sks)))
+ if (unlikely(released && !skb_queue_empty(&tn->bcl->wakeupq)))
n_ptr->action_flags |= TIPC_WAKEUP_BCAST_USERS;
-
exit:
tipc_bclink_unlock(net);
}
@@ -450,6 +447,9 @@ void tipc_bclink_rcv(struct net *net, struct sk_buff *buf)
u32 next_in;
u32 seqno;
int deferred = 0;
+ int pos = 0;
+ struct sk_buff *iskb;
+ struct sk_buff_head msgs;
/* Screen out unwanted broadcast messages */
if (msg_mc_netid(msg) != tn->net_id)
@@ -506,7 +506,8 @@ receive:
bcl->stats.recv_bundled += msg_msgcnt(msg);
tipc_bclink_unlock(net);
tipc_node_unlock(node);
- tipc_link_bundle_rcv(net, buf);
+ while (tipc_msg_extract(buf, &iskb, &pos))
+ tipc_sk_mcast_rcv(net, iskb);
} else if (msg_user(msg) == MSG_FRAGMENTER) {
tipc_buf_append(&node->bclink.reasm_buf, &buf);
if (unlikely(!buf && !node->bclink.reasm_buf))
@@ -527,7 +528,9 @@ receive:
bclink_accept_pkt(node, seqno);
tipc_bclink_unlock(net);
tipc_node_unlock(node);
- tipc_named_rcv(net, buf);
+ skb_queue_head_init(&msgs);
+ skb_queue_tail(&msgs, buf);
+ tipc_named_rcv(net, &msgs);
} else {
tipc_bclink_lock(net);
bclink_accept_pkt(node, seqno);
@@ -944,10 +947,9 @@ int tipc_bclink_init(struct net *net)
spin_lock_init(&bclink->lock);
__skb_queue_head_init(&bcl->outqueue);
__skb_queue_head_init(&bcl->deferred_queue);
- skb_queue_head_init(&bcl->waiting_sks);
+ skb_queue_head_init(&bcl->wakeupq);
bcl->next_out_no = 1;
spin_lock_init(&bclink->node.lock);
- __skb_queue_head_init(&bclink->node.waiting_sks);
bcl->owner = &bclink->node;
bcl->owner->net = net;
bcl->max_pkt = MAX_PKT_DEFAULT_MCAST;
diff --git a/net/tipc/link.c b/net/tipc/link.c
index 41cb09a..9424912 100644
--- a/net/tipc/link.c
+++ b/net/tipc/link.c
@@ -113,10 +113,8 @@ static void link_reset_statistics(struct tipc_link *l_ptr);
static void link_print(struct tipc_link *l_ptr, const char *str);
static void tipc_link_sync_xmit(struct tipc_link *l);
static void tipc_link_sync_rcv(struct tipc_node *n, struct sk_buff *buf);
-static int tipc_link_input(struct net *net, struct tipc_link *l,
- struct sk_buff *buf);
-static int tipc_link_prepare_input(struct net *net, struct tipc_link *l,
- struct sk_buff **buf);
+static void tipc_link_input(struct tipc_link *l, struct sk_buff *skb);
+static bool tipc_data_input(struct tipc_link *l, struct sk_buff *skb);
/*
* Simple link routines
@@ -318,8 +316,9 @@ struct tipc_link *tipc_link_create(struct tipc_node *n_ptr,
l_ptr->next_out_no = 1;
__skb_queue_head_init(&l_ptr->outqueue);
__skb_queue_head_init(&l_ptr->deferred_queue);
- skb_queue_head_init(&l_ptr->waiting_sks);
-
+ skb_queue_head_init(&l_ptr->wakeupq);
+ skb_queue_head_init(&l_ptr->inputq);
+ skb_queue_head_init(&l_ptr->namedq);
link_reset_statistics(l_ptr);
tipc_node_attach_link(n_ptr, l_ptr);
setup_timer(&l_ptr->timer, link_timeout, (unsigned long)l_ptr);
@@ -387,7 +386,7 @@ static bool link_schedule_user(struct tipc_link *link, u32 oport,
return false;
TIPC_SKB_CB(buf)->chain_sz = chain_sz;
TIPC_SKB_CB(buf)->chain_imp = imp;
- skb_queue_tail(&link->waiting_sks, buf);
+ skb_queue_tail(&link->wakeupq, buf);
link->stats.link_congs++;
return true;
}
@@ -398,17 +397,19 @@ static bool link_schedule_user(struct tipc_link *link, u32 oport,
* Move a number of waiting users, as permitted by available space in
* the send queue, from link wait queue to node wait queue for wakeup
*/
-static void link_prepare_wakeup(struct tipc_link *link)
+void link_prepare_wakeup(struct tipc_link *link)
{
uint pend_qsz = skb_queue_len(&link->outqueue);
struct sk_buff *skb, *tmp;
- skb_queue_walk_safe(&link->waiting_sks, skb, tmp) {
+ skb_queue_walk_safe(&link->wakeupq, skb, tmp) {
if (pend_qsz >= link->queue_limit[TIPC_SKB_CB(skb)->chain_imp])
break;
pend_qsz += TIPC_SKB_CB(skb)->chain_sz;
- skb_unlink(skb, &link->waiting_sks);
- skb_queue_tail(&link->owner->waiting_sks, skb);
+ skb_unlink(skb, &link->wakeupq);
+ skb_queue_tail(&link->inputq, skb);
+ link->owner->inputq = &link->inputq;
+ link->owner->action_flags |= TIPC_MSG_EVT;
}
}
@@ -461,13 +462,13 @@ void tipc_link_reset(struct tipc_link *l_ptr)
l_ptr->exp_msg_count = START_CHANGEOVER;
}
- /* Clean up all queues: */
+ /* Clean up all queues, except inputq: */
__skb_queue_purge(&l_ptr->outqueue);
__skb_queue_purge(&l_ptr->deferred_queue);
- if (!skb_queue_empty(&l_ptr->waiting_sks)) {
- skb_queue_splice_init(&l_ptr->waiting_sks, &owner->waiting_sks);
- owner->action_flags |= TIPC_WAKEUP_USERS;
- }
+ skb_queue_splice_init(&l_ptr->wakeupq, &l_ptr->inputq);
+ if (!skb_queue_empty(&l_ptr->inputq))
+ owner->action_flags |= TIPC_MSG_EVT;
+ owner->inputq = &l_ptr->inputq;
l_ptr->next_out = NULL;
l_ptr->unacked_window = 0;
l_ptr->checkpoint = 1;
@@ -795,7 +796,7 @@ int __tipc_link_xmit(struct net *net, struct tipc_link *link,
static void skb2list(struct sk_buff *skb, struct sk_buff_head *list)
{
- __skb_queue_head_init(list);
+ skb_queue_head_init(list);
__skb_queue_tail(list, skb);
}
@@ -841,19 +842,13 @@ int tipc_link_xmit(struct net *net, struct sk_buff_head *list, u32 dnode,
rc = __tipc_link_xmit(net, link, list);
tipc_node_unlock(node);
}
-
if (link)
return rc;
- if (likely(in_own_node(net, dnode))) {
- /* As a node local message chain never contains more than one
- * buffer, we just need to dequeue one SKB buffer from the
- * head list.
- */
- return tipc_sk_rcv(net, __skb_dequeue(list));
- }
- __skb_queue_purge(list);
+ if (likely(in_own_node(net, dnode)))
+ return tipc_sk_rcv(net, list);
+ __skb_queue_purge(list);
return rc;
}
@@ -1162,7 +1157,7 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
/* Locate unicast link endpoint that should handle message */
l_ptr = n_ptr->links[b_ptr->identity];
if (unlikely(!l_ptr))
- goto unlock_discard;
+ goto unlock;
/* Verify that communication with node is currently allowed */
if ((n_ptr->action_flags & TIPC_WAIT_PEER_LINKS_DOWN) &&
@@ -1173,7 +1168,7 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
n_ptr->action_flags &= ~TIPC_WAIT_PEER_LINKS_DOWN;
if (tipc_node_blocked(n_ptr))
- goto unlock_discard;
+ goto unlock;
/* Validate message sequence number info */
seq_no = msg_seqno(msg);
@@ -1197,18 +1192,16 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
if (unlikely(l_ptr->next_out))
tipc_link_push_packets(l_ptr);
- if (released && !skb_queue_empty(&l_ptr->waiting_sks)) {
+ if (released && !skb_queue_empty(&l_ptr->wakeupq))
link_prepare_wakeup(l_ptr);
- l_ptr->owner->action_flags |= TIPC_WAKEUP_USERS;
- }
/* Process the incoming packet */
if (unlikely(!link_working_working(l_ptr))) {
if (msg_user(msg) == LINK_PROTOCOL) {
tipc_link_proto_rcv(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
- tipc_node_unlock(n_ptr);
- continue;
+ skb = NULL;
+ goto unlock;
}
/* Traffic message. Conditionally activate link */
@@ -1217,18 +1210,18 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
if (link_working_working(l_ptr)) {
/* Re-insert buffer in front of queue */
__skb_queue_head(&head, skb);
- tipc_node_unlock(n_ptr);
- continue;
+ skb = NULL;
+ goto unlock;
}
- goto unlock_discard;
+ goto unlock;
}
/* Link is now in state WORKING_WORKING */
if (unlikely(seq_no != mod(l_ptr->next_in_no))) {
link_handle_out_of_seq_msg(l_ptr, skb);
link_retrieve_defq(l_ptr, &head);
- tipc_node_unlock(n_ptr);
- continue;
+ skb = NULL;
+ goto unlock;
}
l_ptr->next_in_no++;
if (unlikely(!skb_queue_empty(&l_ptr->deferred_queue)))
@@ -1238,97 +1231,102 @@ void tipc_rcv(struct net *net, struct sk_buff *skb, struct tipc_bearer *b_ptr)
l_ptr->stats.sent_acks++;
tipc_link_proto_xmit(l_ptr, STATE_MSG, 0, 0, 0, 0, 0);
}
-
- if (tipc_link_prepare_input(net, l_ptr, &skb)) {
- tipc_node_unlock(n_ptr);
- continue;
- }
- tipc_node_unlock(n_ptr);
-
- if (tipc_link_input(net, l_ptr, skb) != 0)
- goto discard;
- continue;
-unlock_discard:
+ tipc_link_input(l_ptr, skb);
+ skb = NULL;
+unlock:
tipc_node_unlock(n_ptr);
discard:
- kfree_skb(skb);
+ if (unlikely(skb))
+ kfree_skb(skb);
}
}
-/**
- * tipc_link_prepare_input - process TIPC link messages
- *
- * returns nonzero if the message was consumed
+/* tipc_data_input - deliver data and name distr msgs to upper layer
*
+ * Consumes buffer if message is of right type
* Node lock must be held
*/
-static int tipc_link_prepare_input(struct net *net, struct tipc_link *l,
- struct sk_buff **buf)
+static bool tipc_data_input(struct tipc_link *link, struct sk_buff *skb)
{
- struct tipc_node *n;
- struct tipc_msg *msg;
- int res = -EINVAL;
+ struct tipc_node *node = link->owner;
+ struct tipc_msg *msg = buf_msg(skb);
+ u32 dport = msg_destport(msg);
- n = l->owner;
- msg = buf_msg(*buf);
switch (msg_user(msg)) {
- case CHANGEOVER_PROTOCOL:
- if (tipc_link_tunnel_rcv(n, buf))
- res = 0;
- break;
- case MSG_FRAGMENTER:
- l->stats.recv_fragments++;
- if (tipc_buf_append(&l->reasm_buf, buf)) {
- l->stats.recv_fragmented++;
- res = 0;
- } else if (!l->reasm_buf) {
- tipc_link_reset(l);
+ case TIPC_LOW_IMPORTANCE:
+ case TIPC_MEDIUM_IMPORTANCE:
+ case TIPC_HIGH_IMPORTANCE:
+ case TIPC_CRITICAL_IMPORTANCE:
+ case CONN_MANAGER:
+ if (tipc_skb_queue_tail(&link->inputq, skb, dport)) {
+ node->inputq = &link->inputq;
+ node->action_flags |= TIPC_MSG_EVT;
}
- break;
- case MSG_BUNDLER:
- l->stats.recv_bundles++;
- l->stats.recv_bundled += msg_msgcnt(msg);
- res = 0;
- break;
+ return true;
case NAME_DISTRIBUTOR:
- n->bclink.recv_permitted = true;
- res = 0;
- break;
+ node->bclink.recv_permitted = true;
+ node->namedq = &link->namedq;
+ skb_queue_tail(&link->namedq, skb);
+ if (skb_queue_len(&link->namedq) == 1)
+ node->action_flags |= TIPC_NAMED_MSG_EVT;
+ return true;
+ case MSG_BUNDLER:
+ case CHANGEOVER_PROTOCOL:
+ case MSG_FRAGMENTER:
case BCAST_PROTOCOL:
- tipc_link_sync_rcv(n, *buf);
- break;
+ return false;
default:
- res = 0;
- }
- return res;
+ pr_warn("Dropping received illegal msg type\n");
+ kfree_skb(skb);
+ return false;
+ };
}
-/**
- * tipc_link_input - Deliver message too higher layers
+
+/* tipc_link_input - process packet that has passed link protocol check
+ *
+ * Consumes buffer
+ * Node lock must be held
*/
-static int tipc_link_input(struct net *net, struct tipc_link *l,
- struct sk_buff *buf)
+static void tipc_link_input(struct tipc_link *link, struct sk_buff *skb)
{
- struct tipc_msg *msg = buf_msg(buf);
- int res = 0;
+ struct tipc_node *node = link->owner;
+ struct tipc_msg *msg = buf_msg(skb);
+ struct sk_buff *iskb;
+ int pos = 0;
+
+ if (likely(tipc_data_input(link, skb)))
+ return;
switch (msg_user(msg)) {
- case TIPC_LOW_IMPORTANCE:
- case TIPC_MEDIUM_IMPORTANCE:
- case TIPC_HIGH_IMPORTANCE:
- case TIPC_CRITICAL_IMPORTANCE:
- case CONN_MANAGER:
- tipc_sk_rcv(net, buf);
+ case CHANGEOVER_PROTOCOL:
+ if (!tipc_link_tunnel_rcv(node, &skb))
+ break;
+ if (msg_user(buf_msg(skb)) != MSG_BUNDLER) {
+ tipc_data_input(link, skb);
+ break;
+ }
+ case MSG_BUNDLER:
+ link->stats.recv_bundles++;
+ link->stats.recv_bundled += msg_msgcnt(msg);
+
+ while (tipc_msg_extract(skb, &iskb, &pos))
+ tipc_data_input(link, iskb);
break;
- case NAME_DISTRIBUTOR:
- tipc_named_rcv(net, buf);
+ case MSG_FRAGMENTER:
+ link->stats.recv_fragments++;
+ if (tipc_buf_append(&link->reasm_buf, &skb)) {
+ link->stats.recv_fragmented++;
+ tipc_data_input(link, skb);
+ } else if (!link->reasm_buf) {
+ tipc_link_reset(link);
+ }
break;
- case MSG_BUNDLER:
- tipc_link_bundle_rcv(net, buf);
+ case BCAST_PROTOCOL:
+ tipc_link_sync_rcv(node, skb);
break;
default:
- res = -EINVAL;
- }
- return res;
+ break;
+ };
}
/**
@@ -1779,7 +1777,7 @@ void tipc_link_dup_queue_xmit(struct tipc_link *l_ptr,
* @from_pos: offset to extract from
*
* Returns a new message buffer containing an embedded message. The
- * encapsulating message itself is left unchanged.
+ * encapsulating buffer is left unchanged.
*/
static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos)
{
@@ -1793,8 +1791,6 @@ static struct sk_buff *buf_extract(struct sk_buff *skb, u32 from_pos)
return eb;
}
-
-
/* tipc_link_dup_rcv(): Receive a tunnelled DUPLICATE_MSG packet.
* Owner node is locked.
*/
@@ -1893,41 +1889,6 @@ exit:
return *buf != NULL;
}
-/*
- * Bundler functionality:
- */
-void tipc_link_bundle_rcv(struct net *net, struct sk_buff *buf)
-{
- u32 msgcount = msg_msgcnt(buf_msg(buf));
- u32 pos = INT_H_SIZE;
- struct sk_buff *obuf;
- struct tipc_msg *omsg;
-
- while (msgcount--) {
- obuf = buf_extract(buf, pos);
- if (obuf == NULL) {
- pr_warn("Link unable to unbundle message(s)\n");
- break;
- }
- omsg = buf_msg(obuf);
- pos += align(msg_size(omsg));
- if (msg_isdata(omsg)) {
- if (unlikely(msg_type(omsg) == TIPC_MCAST_MSG))
- tipc_sk_mcast_rcv(net, obuf);
- else
- tipc_sk_rcv(net, obuf);
- } else if (msg_user(omsg) == CONN_MANAGER) {
- tipc_sk_rcv(net, obuf);
- } else if (msg_user(omsg) == NAME_DISTRIBUTOR) {
- tipc_named_rcv(net, obuf);
- } else {
- pr_warn("Illegal bundled msg: %u\n", msg_user(omsg));
- kfree_skb(obuf);
- }
- }
- kfree_skb(buf);
-}
-
static void link_set_supervision_props(struct tipc_link *l_ptr, u32 tol)
{
unsigned long intv = ((tol / 4) > 500) ? 500 : tol / 4;
diff --git a/net/tipc/link.h b/net/tipc/link.h
index 5b9a17f..34d3f55 100644
--- a/net/tipc/link.h
+++ b/net/tipc/link.h
@@ -131,8 +131,10 @@ struct tipc_stats {
* @next_in_no: next sequence number to expect for inbound messages
* @deferred_queue: deferred queue saved OOS b'cast message received from node
* @unacked_window: # of inbound messages rx'd without ack'ing back to peer
+ * @inputq: buffer queue for messages to be delivered upwards
+ * @namedq: buffer queue for name table messages to be delivered upwards
* @next_out: ptr to first unsent outbound message in queue
- * @waiting_sks: linked list of sockets waiting for link congestion to abate
+ * @wakeupq: linked list of wakeup msgs waiting for link congestion to abate
* @long_msg_seq_no: next identifier to use for outbound fragmented messages
* @reasm_buf: head of partially reassembled inbound message fragments
* @stats: collects statistics regarding link activity
@@ -184,10 +186,12 @@ struct tipc_link {
u32 next_in_no;
struct sk_buff_head deferred_queue;
u32 unacked_window;
+ struct sk_buff_head inputq;
+ struct sk_buff_head namedq;
/* Congestion handling */
struct sk_buff *next_out;
- struct sk_buff_head waiting_sks;
+ struct sk_buff_head wakeupq;
/* Fragmentation/reassembly */
u32 long_msg_seq_no;
@@ -228,7 +232,6 @@ int tipc_link_xmit(struct net *net, struct sk_buff_head *list, u32 dest,
u32 selector);
int __tipc_link_xmit(struct net *net, struct tipc_link *link,
struct sk_buff_head *list);
-void tipc_link_bundle_rcv(struct net *net, struct sk_buff *buf);
void tipc_link_proto_xmit(struct tipc_link *l_ptr, u32 msg_typ, int prob,
u32 gap, u32 tolerance, u32 priority, u32 acked_mtu);
void tipc_link_push_packets(struct tipc_link *l_ptr);
@@ -244,6 +247,7 @@ int tipc_nl_link_get(struct sk_buff *skb, struct genl_info *info);
int tipc_nl_link_set(struct sk_buff *skb, struct genl_info *info);
int tipc_nl_link_reset_stats(struct sk_buff *skb, struct genl_info *info);
int tipc_nl_parse_link_prop(struct nlattr *prop, struct nlattr *props[]);
+void link_prepare_wakeup(struct tipc_link *l);
/*
* Link sequence number manipulation routines (uses modulo 2**16 arithmetic)
diff --git a/net/tipc/msg.c b/net/tipc/msg.c
index 697223a..b6eb90c 100644
--- a/net/tipc/msg.c
+++ b/net/tipc/msg.c
@@ -327,6 +327,40 @@ bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu)
}
/**
+ * tipc_msg_extract(): extract bundled inner packet from buffer
+ * @skb: linear outer buffer, to be extracted from.
+ * @iskb: extracted inner buffer, to be returned
+ * @pos: position of msg to be extracted. Returns with pointer of next msg
+ * Consumes outer buffer when last packet extracted
+ * Returns true when when there is an extracted buffer, otherwise false
+ */
+bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos)
+{
+ struct tipc_msg *msg = buf_msg(skb);
+ int imsz;
+ struct tipc_msg *imsg = (struct tipc_msg *)(msg_data(msg) + *pos);
+
+ /* Is there space left for shortest possible message? */
+ if (*pos > (msg_data_sz(msg) - SHORT_H_SIZE))
+ goto none;
+ imsz = msg_size(imsg);
+
+ /* Is there space left for current message ? */
+ if ((*pos + imsz) > msg_data_sz(msg))
+ goto none;
+ *iskb = tipc_buf_acquire(imsz);
+ if (!*iskb)
+ goto none;
+ skb_copy_to_linear_data(*iskb, imsg, imsz);
+ *pos += align(imsz);
+ return true;
+none:
+ kfree_skb(skb);
+ *iskb = NULL;
+ return false;
+}
+
+/**
* tipc_msg_make_bundle(): Create bundle buf and append message to its tail
* @list: the buffer chain
* @skb: buffer to be appended and replaced
diff --git a/net/tipc/msg.h b/net/tipc/msg.h
index 6070299..ab46726 100644
--- a/net/tipc/msg.h
+++ b/net/tipc/msg.h
@@ -45,6 +45,7 @@
* Note: Some items are also used with TIPC internal message headers
*/
#define TIPC_VERSION 2
+struct plist;
/*
* Payload message users are defined in TIPC's public API:
@@ -759,10 +760,82 @@ int tipc_buf_append(struct sk_buff **headbuf, struct sk_buff **buf);
bool tipc_msg_bundle(struct sk_buff_head *list, struct sk_buff *skb, u32 mtu);
bool tipc_msg_make_bundle(struct sk_buff_head *list,
struct sk_buff *skb, u32 mtu, u32 dnode);
+bool tipc_msg_extract(struct sk_buff *skb, struct sk_buff **iskb, int *pos);
int tipc_msg_build(struct tipc_msg *mhdr, struct msghdr *m,
int offset, int dsz, int mtu, struct sk_buff_head *list);
bool tipc_msg_lookup_dest(struct net *net, struct sk_buff *skb, u32 *dnode,
int *err);
struct sk_buff *tipc_msg_reassemble(struct sk_buff_head *list);
+/* tipc_skb_peek_port(): find a destination port, ignoring all destinations
+ * up to and including 'filter'.
+ * Note: ignoring previously tried destinations minimizes the risk of
+ * contention on the socket lock
+ * @list: list to be peeked in
+ * @filter: last destination to be ignored from search
+ * Returns a destination port number, of applicable.
+ */
+static inline u32 tipc_skb_peek_port(struct sk_buff_head *list, u32 filter)
+{
+ struct sk_buff *skb;
+ u32 dport = 0;
+ bool ignore = true;
+
+ spin_lock_bh(&list->lock);
+ skb_queue_walk(list, skb) {
+ dport = msg_destport(buf_msg(skb));
+ if (!filter || skb_queue_is_last(list, skb))
+ break;
+ if (dport == filter)
+ ignore = false;
+ else if (!ignore)
+ break;
+ }
+ spin_unlock_bh(&list->lock);
+ return dport;
+}
+
+/* tipc_skb_dequeue(): unlink first buffer with dest 'dport' from list
+ * @list: list to be unlinked from
+ * @dport: selection criteria for buffer to unlink
+ */
+static inline struct sk_buff *tipc_skb_dequeue(struct sk_buff_head *list,
+ u32 dport)
+{
+ struct sk_buff *_skb, *tmp, *skb = NULL;
+
+ spin_lock_bh(&list->lock);
+ skb_queue_walk_safe(list, _skb, tmp) {
+ if (msg_destport(buf_msg(_skb)) == dport) {
+ __skb_unlink(_skb, list);
+ skb = _skb;
+ break;
+ }
+ }
+ spin_unlock_bh(&list->lock);
+ return skb;
+}
+
+/* tipc_skb_queue_tail(): add buffer to tail of list;
+ * @list: list to be appended to
+ * @skb: buffer to append. Always appended
+ * @dport: the destination port of the buffer
+ * returns true if dport differs from previous destination
+ */
+static inline bool tipc_skb_queue_tail(struct sk_buff_head *list,
+ struct sk_buff *skb, u32 dport)
+{
+ struct sk_buff *_skb = NULL;
+ bool rv = false;
+
+ spin_lock_bh(&list->lock);
+ _skb = skb_peek_tail(list);
+ if (!_skb || (msg_destport(buf_msg(_skb)) != dport) ||
+ (skb_queue_len(list) > 32))
+ rv = true;
+ __skb_queue_tail(list, skb);
+ spin_unlock_bh(&list->lock);
+ return rv;
+}
+
#endif
diff --git a/net/tipc/name_distr.c b/net/tipc/name_distr.c
index dd8564c..fcb0791 100644
--- a/net/tipc/name_distr.c
+++ b/net/tipc/name_distr.c
@@ -381,25 +381,34 @@ void tipc_named_process_backlog(struct net *net)
}
/**
- * tipc_named_rcv - process name table update message sent by another node
+ * tipc_named_rcv - process name table update messages sent by another node
*/
-void tipc_named_rcv(struct net *net, struct sk_buff *buf)
+void tipc_named_rcv(struct net *net, struct sk_buff_head *inputq)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
- struct tipc_msg *msg = buf_msg(buf);
- struct distr_item *item = (struct distr_item *)msg_data(msg);
- u32 count = msg_data_sz(msg) / ITEM_SIZE;
- u32 node = msg_orignode(msg);
+ struct tipc_msg *msg;
+ struct distr_item *item;
+ uint count;
+ u32 node;
+ struct sk_buff *skb;
+ int mtype;
spin_lock_bh(&tn->nametbl_lock);
- while (count--) {
- if (!tipc_update_nametbl(net, item, node, msg_type(msg)))
- tipc_named_add_backlog(item, msg_type(msg), node);
- item++;
+ for (skb = skb_dequeue(inputq); skb; skb = skb_dequeue(inputq)) {
+ msg = buf_msg(skb);
+ mtype = msg_type(msg);
+ item = (struct distr_item *)msg_data(msg);
+ count = msg_data_sz(msg) / ITEM_SIZE;
+ node = msg_orignode(msg);
+ while (count--) {
+ if (!tipc_update_nametbl(net, item, node, mtype))
+ tipc_named_add_backlog(item, mtype, node);
+ item++;
+ }
+ kfree_skb(skb);
+ tipc_named_process_backlog(net);
}
- tipc_named_process_backlog(net);
spin_unlock_bh(&tn->nametbl_lock);
- kfree_skb(buf);
}
/**
diff --git a/net/tipc/name_distr.h b/net/tipc/name_distr.h
index 5ec10b5..dd2d9fd 100644
--- a/net/tipc/name_distr.h
+++ b/net/tipc/name_distr.h
@@ -71,7 +71,7 @@ struct sk_buff *tipc_named_publish(struct net *net, struct publication *publ);
struct sk_buff *tipc_named_withdraw(struct net *net, struct publication *publ);
void named_cluster_distribute(struct net *net, struct sk_buff *buf);
void tipc_named_node_up(struct net *net, u32 dnode);
-void tipc_named_rcv(struct net *net, struct sk_buff *buf);
+void tipc_named_rcv(struct net *net, struct sk_buff_head *msg_queue);
void tipc_named_reinit(struct net *net);
void tipc_named_process_backlog(struct net *net);
void tipc_publ_notify(struct net *net, struct list_head *nsub_list, u32 addr);
diff --git a/net/tipc/node.c b/net/tipc/node.c
index 1c409c4..dcb83d9 100644
--- a/net/tipc/node.c
+++ b/net/tipc/node.c
@@ -111,11 +111,8 @@ struct tipc_node *tipc_node_create(struct net *net, u32 addr)
INIT_LIST_HEAD(&n_ptr->list);
INIT_LIST_HEAD(&n_ptr->publ_list);
INIT_LIST_HEAD(&n_ptr->conn_sks);
- skb_queue_head_init(&n_ptr->waiting_sks);
__skb_queue_head_init(&n_ptr->bclink.deferred_queue);
-
hlist_add_head_rcu(&n_ptr->hash, &tn->node_htable[tipc_hashfn(addr)]);
-
list_for_each_entry_rcu(temp_node, &tn->node_list, list) {
if (n_ptr->addr < temp_node->addr)
break;
@@ -201,19 +198,22 @@ void tipc_node_abort_sock_conns(struct net *net, struct list_head *conns)
{
struct tipc_net *tn = net_generic(net, tipc_net_id);
struct tipc_sock_conn *conn, *safe;
- struct sk_buff *buf;
+ struct sk_buff *skb;
+ struct sk_buff_head skbs;
+ skb_queue_head_init(&skbs);
list_for_each_entry_safe(conn, safe, conns, list) {
- buf = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
+ skb = tipc_msg_create(TIPC_CRITICAL_IMPORTANCE,
TIPC_CONN_MSG, SHORT_H_SIZE, 0,
tn->own_addr, conn->peer_node,
conn->port, conn->peer_port,
TIPC_ERR_NO_NODE);
- if (likely(buf))
- tipc_sk_rcv(net, buf);
+ if (likely(skb))
+ skb_queue_tail(&skbs, skb);
list_del(&conn->list);
kfree(conn);
}
+ tipc_sk_rcv(net, &skbs);
}
/**
@@ -568,37 +568,36 @@ void tipc_node_unlock(struct tipc_node *node)
struct net *net = node->net;
LIST_HEAD(nsub_list);
LIST_HEAD(conn_sks);
- struct sk_buff_head waiting_sks;
u32 addr = 0;
- int flags = node->action_flags;
+ u32 flags = node->action_flags;
u32 link_id = 0;
+ struct sk_buff_head *inputq = node->inputq;
+ struct sk_buff_head *namedq = node->inputq;
- if (likely(!flags)) {
+ if (likely(!flags || (flags == TIPC_MSG_EVT))) {
+ node->action_flags = 0;
spin_unlock_bh(&node->lock);
+ if (flags == TIPC_MSG_EVT)
+ tipc_sk_rcv(net, inputq);
return;
}
addr = node->addr;
link_id = node->link_id;
- __skb_queue_head_init(&waiting_sks);
-
- if (flags & TIPC_WAKEUP_USERS)
- skb_queue_splice_init(&node->waiting_sks, &waiting_sks);
+ namedq = node->namedq;
if (flags & TIPC_NOTIFY_NODE_DOWN) {
list_replace_init(&node->publ_list, &nsub_list);
list_replace_init(&node->conn_sks, &conn_sks);
}
- node->action_flags &= ~(TIPC_WAKEUP_USERS | TIPC_NOTIFY_NODE_DOWN |
+ node->action_flags &= ~(TIPC_MSG_EVT | TIPC_NOTIFY_NODE_DOWN |
TIPC_NOTIFY_NODE_UP | TIPC_NOTIFY_LINK_UP |
TIPC_NOTIFY_LINK_DOWN |
- TIPC_WAKEUP_BCAST_USERS);
+ TIPC_WAKEUP_BCAST_USERS |
+ TIPC_NAMED_MSG_EVT);
spin_unlock_bh(&node->lock);
- while (!skb_queue_empty(&waiting_sks))
- tipc_sk_rcv(net, __skb_dequeue(&waiting_sks));
-
if (!list_empty(&conn_sks))
tipc_node_abort_sock_conns(net, &conn_sks);
@@ -618,6 +617,12 @@ void tipc_node_unlock(struct tipc_node *node)
if (flags & TIPC_NOTIFY_LINK_DOWN)
tipc_nametbl_withdraw(net, TIPC_LINK_STATE, addr,
link_id, addr);
+
+ if (flags & TIPC_MSG_EVT)
+ tipc_sk_rcv(net, inputq);
+
+ if (flags & TIPC_NAMED_MSG_EVT)
+ tipc_named_rcv(net, namedq);
}
/* Caller should hold node lock for the passed node */
diff --git a/net/tipc/node.h b/net/tipc/node.h
index 43ef88e..c2b0fcf 100644
--- a/net/tipc/node.h
+++ b/net/tipc/node.h
@@ -55,14 +55,15 @@
* TIPC_DISTRIBUTE_NAME: publish or withdraw link state name type
*/
enum {
+ TIPC_MSG_EVT = 1,
TIPC_WAIT_PEER_LINKS_DOWN = (1 << 1),
TIPC_WAIT_OWN_LINKS_DOWN = (1 << 2),
TIPC_NOTIFY_NODE_DOWN = (1 << 3),
TIPC_NOTIFY_NODE_UP = (1 << 4),
- TIPC_WAKEUP_USERS = (1 << 5),
- TIPC_WAKEUP_BCAST_USERS = (1 << 6),
- TIPC_NOTIFY_LINK_UP = (1 << 7),
- TIPC_NOTIFY_LINK_DOWN = (1 << 8)
+ TIPC_WAKEUP_BCAST_USERS = (1 << 5),
+ TIPC_NOTIFY_LINK_UP = (1 << 6),
+ TIPC_NOTIFY_LINK_DOWN = (1 << 7),
+ TIPC_NAMED_MSG_EVT = (1 << 8)
};
/**
@@ -92,6 +93,9 @@ struct tipc_node_bclink {
* @lock: spinlock governing access to structure
* @net: the applicable net namespace
* @hash: links to adjacent nodes in unsorted hash chain
+ * @inputq: pointer to input queue containing messages for msg event
+ * @namedq: pointer to name table input queue with name table messages
+ * @curr_link: the link holding the node lock, if any
* @active_links: pointers to active links to node
* @links: pointers to all links to node
* @action_flags: bit mask of different types of node actions
@@ -109,10 +113,12 @@ struct tipc_node {
spinlock_t lock;
struct net *net;
struct hlist_node hash;
+ struct sk_buff_head *inputq;
+ struct sk_buff_head *namedq;
struct tipc_link *active_links[2];
u32 act_mtus[2];
struct tipc_link *links[MAX_BEARERS];
- unsigned int action_flags;
+ int action_flags;
struct tipc_node_bclink bclink;
struct list_head list;
int link_cnt;
@@ -120,7 +126,6 @@ struct tipc_node {
u32 signature;
u32 link_id;
struct list_head publ_list;
- struct sk_buff_head waiting_sks;
struct list_head conn_sks;
struct rcu_head rcu;
};
diff --git a/net/tipc/socket.c b/net/tipc/socket.c
index 611a04f..c1a4611 100644
--- a/net/tipc/socket.c
+++ b/net/tipc/socket.c
@@ -41,6 +41,7 @@
#include "node.h"
#include "link.h"
#include "config.h"
+#include "name_distr.h"
#include "socket.h"
#define SS_LISTENING -1 /* socket is listening */
@@ -785,10 +786,16 @@ void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *buf)
struct sk_buff *b;
uint i, last, dst = 0;
u32 scope = TIPC_CLUSTER_SCOPE;
+ struct sk_buff_head msgs;
if (in_own_node(net, msg_orignode(msg)))
scope = TIPC_NODE_SCOPE;
+ if (unlikely(!msg_mcast(msg))) {
+ pr_warn("Received non-multicast msg in multicast\n");
+ kfree_skb(buf);
+ goto exit;
+ }
/* Create destination port list: */
tipc_nametbl_mc_translate(net, msg_nametype(msg), msg_namelower(msg),
msg_nameupper(msg), scope, &dports);
@@ -806,9 +813,12 @@ void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *buf)
continue;
}
msg_set_destport(msg, item->ports[i]);
- tipc_sk_rcv(net, b);
+ skb_queue_head_init(&msgs);
+ skb_queue_tail(&msgs, b);
+ tipc_sk_rcv(net, &msgs);
}
}
+exit:
tipc_port_list_free(&dports);
}
@@ -1760,71 +1770,99 @@ static int tipc_backlog_rcv(struct sock *sk, struct sk_buff *skb)
}
/**
- * tipc_sk_enqueue_skb - enqueue buffer to socket or backlog queue
- * @sk: socket
- * @skb: pointer to message. Set to NULL if buffer is consumed.
- * @dnode: if buffer should be forwarded/returned, send to this node
+ * tipc_sk_enqueue - extract all buffers with destination 'dport' from
+ * inputq and try adding them to socket or backlog queue
+ * @inputq: list of incoming buffers with potentially different destinations
+ * @sk: socket where the buffers should be enqueued
+ * @dport: port number for the socket
+ * @_skb: returned buffer to be forwarded or rejected, if applicable
*
* Caller must hold socket lock
*
- * Returns TIPC_OK (0) or -tipc error code
+ * Returns TIPC_OK if all buffers enqueued, otherwise -TIPC_ERR_OVERLOAD
+ * or -TIPC_ERR_NO_PORT
*/
-static int tipc_sk_enqueue_skb(struct sock *sk, struct sk_buff **skb)
+static int tipc_sk_enqueue(struct sk_buff_head *inputq, struct sock *sk,
+ u32 dport, struct sk_buff **_skb)
{
unsigned int lim;
atomic_t *dcnt;
-
- if (unlikely(!*skb))
- return TIPC_OK;
- if (!sock_owned_by_user(sk))
- return filter_rcv(sk, skb);
- dcnt = &tipc_sk(sk)->dupl_rcvcnt;
- if (sk->sk_backlog.len)
- atomic_set(dcnt, 0);
- lim = rcvbuf_limit(sk, *skb) + atomic_read(dcnt);
- if (unlikely(sk_add_backlog(sk, *skb, lim)))
+ int err;
+ struct sk_buff *skb;
+ unsigned long time_limit = jiffies + 2;
+
+ while (skb_queue_len(inputq)) {
+ skb = tipc_skb_dequeue(inputq, dport);
+ if (unlikely(!skb))
+ return TIPC_OK;
+ /* Return if softirq window exhausted */
+ if (unlikely(time_after_eq(jiffies, time_limit)))
+ return TIPC_OK;
+ if (!sock_owned_by_user(sk)) {
+ err = filter_rcv(sk, &skb);
+ if (likely(!skb))
+ continue;
+ *_skb = skb;
+ return err;
+ }
+ dcnt = &tipc_sk(sk)->dupl_rcvcnt;
+ if (sk->sk_backlog.len)
+ atomic_set(dcnt, 0);
+ lim = rcvbuf_limit(sk, skb) + atomic_read(dcnt);
+ if (likely(!sk_add_backlog(sk, skb, lim)))
+ continue;
+ *_skb = skb;
return -TIPC_ERR_OVERLOAD;
- *skb = NULL;
+ }
return TIPC_OK;
}
/**
- * tipc_sk_rcv - handle incoming message
- * @skb: buffer containing arriving message
- * Consumes buffer
- * Returns 0 if success, or errno: -EHOSTUNREACH
+ * tipc_sk_rcv - handle a chain of incoming buffers
+ * @inputq: buffer list containing the buffers
+ * Consumes all buffers in list until inputq is empty
+ * Note: may be called in multiple threads referring to the same queue
+ * Returns 0 if last buffer was accepted, otherwise -EHOSTUNREACH
+ * Only node local calls check the return value, sending single-buffer queues
*/
-int tipc_sk_rcv(struct net *net, struct sk_buff *skb)
+int tipc_sk_rcv(struct net *net, struct sk_buff_head *inputq)
{
+ u32 dnode, dport = 0;
+ int err = -TIPC_ERR_NO_PORT;
+ struct sk_buff *skb;
struct tipc_sock *tsk;
struct tipc_net *tn;
struct sock *sk;
- u32 dport = msg_destport(buf_msg(skb));
- int err = -TIPC_ERR_NO_PORT;
- u32 dnode;
- /* Find destination */
- tsk = tipc_sk_lookup(net, dport);
- if (likely(tsk)) {
- sk = &tsk->sk;
- spin_lock_bh(&sk->sk_lock.slock);
- err = tipc_sk_enqueue_skb(sk, &skb);
- spin_unlock_bh(&sk->sk_lock.slock);
- sock_put(sk);
- }
- if (likely(!skb))
- return 0;
- if (tipc_msg_lookup_dest(net, skb, &dnode, &err))
- goto xmit;
- if (!err) {
- dnode = msg_destnode(buf_msg(skb));
- goto xmit;
- }
- tn = net_generic(net, tipc_net_id);
- if (!tipc_msg_reverse(tn->own_addr, skb, &dnode, -err))
- return -EHOSTUNREACH;
+ while (skb_queue_len(inputq)) {
+ skb = NULL;
+ dport = tipc_skb_peek_port(inputq, dport);
+ tsk = tipc_sk_lookup(net, dport);
+ if (likely(tsk)) {
+ sk = &tsk->sk;
+ if (likely(spin_trylock_bh(&sk->sk_lock.slock))) {
+ err = tipc_sk_enqueue(inputq, sk, dport, &skb);
+ spin_unlock_bh(&sk->sk_lock.slock);
+ dport = 0;
+ }
+ sock_put(sk);
+ } else {
+ skb = tipc_skb_dequeue(inputq, dport);
+ }
+ if (likely(!skb))
+ continue;
+ if (tipc_msg_lookup_dest(net, skb, &dnode, &err))
+ goto xmit;
+ if (!err) {
+ dnode = msg_destnode(buf_msg(skb));
+ goto xmit;
+ }
+ tn = net_generic(net, tipc_net_id);
+ if (!tipc_msg_reverse(tn->own_addr, skb, &dnode, -err))
+ continue;
xmit:
- tipc_link_xmit_skb(net, skb, dnode, dport);
+ tipc_link_xmit_skb(net, skb, dnode, dport);
+ }
return err ? -EHOSTUNREACH : 0;
}
diff --git a/net/tipc/socket.h b/net/tipc/socket.h
index f56c3fd..e3dbdc0 100644
--- a/net/tipc/socket.h
+++ b/net/tipc/socket.h
@@ -49,7 +49,7 @@ int tipc_sock_create_local(struct net *net, int type, struct socket **res);
void tipc_sock_release_local(struct socket *sock);
int tipc_sock_accept_local(struct socket *sock, struct socket **newsock,
int flags);
-int tipc_sk_rcv(struct net *net, struct sk_buff *buf);
+int tipc_sk_rcv(struct net *net, struct sk_buff_head *inputq);
struct sk_buff *tipc_sk_socks_show(struct net *net);
void tipc_sk_mcast_rcv(struct net *net, struct sk_buff *buf);
void tipc_sk_reinit(struct net *net);
--
1.9.1
^ permalink raw reply related
* [PATCH] xen-netback: fix sparse warning
From: Lad Prabhakar @ 2015-02-05 13:38 UTC (permalink / raw)
To: Ian Campbell, Wei Liu, xen-devel; +Cc: netdev, linux-kernel, Lad, Prabhakar
From: "Lad, Prabhakar" <prabhakar.csengg@gmail.com>
this patch fixes following sparse warning:
interface.c:83:5: warning: symbol 'xenvif_poll' was not declared. Should it be static?
Signed-off-by: Lad, Prabhakar <prabhakar.csengg@gmail.com>
---
Found this issue on linux-next (gcc version 4.9.2,
sparse version 0.4.5-rc1)and applies on top linux-next.
drivers/net/xen-netback/interface.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index 12f9e27..f38227a 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -80,7 +80,7 @@ static irqreturn_t xenvif_tx_interrupt(int irq, void *dev_id)
return IRQ_HANDLED;
}
-int xenvif_poll(struct napi_struct *napi, int budget)
+static int xenvif_poll(struct napi_struct *napi, int budget)
{
struct xenvif_queue *queue =
container_of(napi, struct xenvif_queue, napi);
--
1.9.1
^ permalink raw reply related
* Re: Throughput regression with `tcp: refine TSO autosizing`
From: Michal Kazior @ 2015-02-05 13:44 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Neal Cardwell, linux-wireless, Network Development, eyalpe
In-Reply-To: <1423142342.31870.49.camel@edumazet-glaptop2.roam.corp.google.com>
On 5 February 2015 at 14:19, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Thu, 2015-02-05 at 04:57 -0800, Eric Dumazet wrote:
>
>> The intention is to control the queues to the following :
>>
>> 1 ms of buffering, but limited to a configurable value.
>>
>> On a 40Gbps flow, 1ms represents 5 MB, which is insane.
>>
>> We do not want to queue 5 MB of traffic, this would destroy latencies
>> for all concurrent flows. (Or would require having fq_codel or fq as
>> packet schedulers, instead of default pfifo_fast)
>>
>> This is why having 1.5 ms delay between the transmit and TX completion
>> is a problem in your case.
I do get your point. But 1.5ms is really tough on Wi-Fi.
Just look at this:
; ping 192.168.1.2 -c 3
PING 192.168.1.2 (192.168.1.2) 56(84) bytes of data.
64 bytes from 192.168.1.2: icmp_seq=1 ttl=64 time=1.83 ms
64 bytes from 192.168.1.2: icmp_seq=2 ttl=64 time=2.02 ms
64 bytes from 192.168.1.2: icmp_seq=3 ttl=64 time=1.98 ms
; ping 192.168.1.2 -c 3 -Q 224
PING 192.168.1.2 (192.168.1.2) 56(84) bytes of data.
64 bytes from 192.168.1.2: icmp_seq=1 ttl=64 time=0.939 ms
64 bytes from 192.168.1.2: icmp_seq=2 ttl=64 time=0.906 ms
64 bytes from 192.168.1.2: icmp_seq=3 ttl=64 time=0.946 ms
This was run with no load so batching code in the driver itself should
have no measurable effect. The channel was near-ideal: low noise
floor, cabled rf, no other traffic.
The lower latency ping is when 802.11 QoS Voice Access Category is
used. I also get 400mbps instead of 250mbps in this case with 5 flows
(net/master).
Dealing with black box firmware blobs is a pain.
> Note that TCP stack could detect when this happens, *if* ACK where
> delivered before the TX completions, or when TX completion happens,
> we could detect that the clone of the freed packet was freed.
>
> In my test, when I did "ethtool -C eth0 tx-usecs 1024 tx-frames 64", and
> disabling GSO, TCP stack sends a bunch of packets (a bit less than 64),
> blocks on tcp_limit_output_bytes.
>
> Then we receive 2 stretch ACKS after ~50 usec.
>
> TCP stack tries to push again some packets but blocks on
> tcp_limit_output_bytes again.
>
> 1ms later, TX completion happens, tcp_wfree() is called, and TCP stack
> push following ~60 packets.
>
>
> TCP could eventually dynamically adjust the tcp_limit_output_bytes,
> using a per flow dynamic value, but I would rather not add a kludge in
> TCP stack only to deal with a possible bug in ath10k driver.
>
> niu has a similar issue and simply had to call skb_orphan() :
>
> drivers/net/ethernet/sun/niu.c:6669: skb_orphan(skb);
Ok. I tried calling skb_orphan() right after I submit each Tx frame
(similar to niu which does this in start_xmit):
--- a/drivers/net/wireless/ath/ath10k/htt_tx.c
+++ b/drivers/net/wireless/ath/ath10k/htt_tx.c
@@ -564,6 +564,8 @@ int ath10k_htt_tx(struct ath10k_htt *htt, struct
sk_buff *msdu)
if (res)
goto err_unmap_msdu;
+ skb_orphan(msdu);
+
return 0;
err_unmap_msdu:
Now, with {net/master + ath10k GRO + the above} I get 620mbps on a
single flow (even better then before). Wow.
Does this look ok/safe as a solution to you?
Michał
^ permalink raw reply
* Re: [RFC PATCH 00/29] net: VRF support
From: Nicolas Dichtel @ 2015-02-05 13:44 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-1-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
[snip]
> This is accomplished by enhancing the current namespace checks to a
> broader network context that is both a namepsace and a VRF id. The VRF
> id is a tag applied to relevant structures, an integer between 1 and 4095
> which allows for 4095 VRFs (could have 0 be the default VRF and then the
> range is 0-4095 = 4096s VRFs). (The limitation is arguably artificial. It
> is based on the genid scheme for versioning networking data which is a
> 32-bit integer. The VRF id is the lower 12 bits of the genid's.)
Would it be possible to avoid this artificial limit?
There could be scenarii with more than 4096 vrf.
Do you plan to have a way to dump or monitor VRF via netlink?
^ permalink raw reply
* Re: [RFC PATCH 18/29] net: vrf: Plumbing for vrf context on a socket
From: Nicolas Dichtel @ 2015-02-05 13:44 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-19-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
> Sockets inherit the vrf context of the task opening it. The context can
> be read/changed via a socket option (IP_VRF_CONTEXT).
What about using a common socket option (SO_VRF_CONTEXT) instead of an ipv4
specific option?
^ permalink raw reply
* Re: [RFC PATCH 14/29] net: vrf: Introduce vrf header file
From: Nicolas Dichtel @ 2015-02-05 13:44 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-15-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
> Defines for min and max vrf id and helpers for examining
>
> Signed-off-by: David Ahern <dsahern@gmail.com>
> ---
> include/net/vrf.h | 36 ++++++++++++++++++++++++++++++++++++
> 1 file changed, 36 insertions(+)
> create mode 100644 include/net/vrf.h
>
> diff --git a/include/net/vrf.h b/include/net/vrf.h
> new file mode 100644
> index 000000000000..67bc2e465661
> --- /dev/null
> +++ b/include/net/vrf.h
> @@ -0,0 +1,36 @@
> +#ifndef _VRF_H_
> +#define _VRF_H_
> +
> +#define VRF_BITS 12
> +#define VRF_MIN 1
> +#define VRF_MAX ((1 << VRF_BITS) - 1)
> +#define VRF_MASK VRF_MAX
> +
> +#define VRF_DEFAULT 1
> +#define VRF_ANY 0xffff
It could be useful to expose this value to userland.
^ permalink raw reply
* Re: [ovs-dev] [RFC: add openvswitch actions using BPF 1/9] hack: Do not compile datapath
From: Thomas Graf @ 2015-02-05 13:44 UTC (permalink / raw)
To: Andy Zhou; +Cc: dev, netdev
In-Reply-To: <1423090163-19902-2-git-send-email-azhou@nicira.com>
On 02/04/15 at 02:49pm, Andy Zhou wrote:
> The OVS eBPF patch sets are developed against master 'net-next' tree,
> which is currently targeting for kernel version 3.19.
>
> The datapath in OVS mater does not currently support 3.19 kernel.
> But we should not really using OVS datapath, The changes in the
> follow patches are only changes in user space programs, which should
> work with the kernel module built from the 'net-next' tree.
>
> This patch disables building of the datapath and disables the
> configuration time kernel version check.
>
> Signed-off-by: Andy Zhou <azhou@nicira.com>
You can drop this when the datapath sync series is fully merged.
^ permalink raw reply
* Re: [RFC PATCH 19/29] net: vrf: Add vrf context to skb
From: Nicolas Dichtel @ 2015-02-05 13:45 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-20-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
> On ingress skb's inherit vrf context from the net_device. For TX skb's
> inherit the vrf context from the socket originating the packet. Update
> SKB related net_ctx macros to set vrf.
Is it really needed to have this skb->vrf?
Is it not possible to get the vrf from sockets, interfaces, etc? Like it's done
with netns?
^ permalink raw reply
* Re: [RFC PATCH 25/29] net: vrf: Handle VRF any context
From: Nicolas Dichtel @ 2015-02-05 13:46 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-26-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
> VRF any context applies only to tasks to and sockets. Devices are
> associated with a single VRF, and skb's by extension are connected to
> a single VRF.
>
> Listen sockets and unconnected sockets can be opened in a "VRF any"
> context allowing a single daemon to provide service across all VRFs
> in a namespace. Connected sockets must be in a specific vrf context.
> Accepted sockets acquire the VRF context from the device the packet
> enters (via the skb).
>
> "VRF any" context is also useful for tasks wanting to view L3/L4
> data for all VRFs.
>
> Signed-off-by: David Ahern <dsahern@gmail.com>
> ---
[snip]
> +static inline int neigh_parms_net_ctx_eq_any(const struct neigh_parms *parms,
> + const struct net_ctx *net_ctx)
> +{
> +#ifdef CONFIG_NET_NS
> + if (net_eq(neigh_parms_net(parms), net_ctx->net) &&
> + (vrf_eq(neigh_parms_vrf(parms), net_ctx->vrf) ||
> + vrf_is_any(net_ctx->vrf))) {
> + return 1;
> + }
> +
> + return 0;
> +#else
> + return 1;
> +#endif
If I understand well, the way the patch is done, VRF can be used only if
CONFIG_NET_NS is enabled.
But if I'm not wrong, it could be independent. Am I right?
^ permalink raw reply
* Re: [RFC PATCH 02/29] net: Flip net_device to use net_ctx
From: Nicolas Dichtel @ 2015-02-05 13:47 UTC (permalink / raw)
To: David Ahern, netdev; +Cc: ebiederm
In-Reply-To: <1423100070-31848-3-git-send-email-dsahern@gmail.com>
Le 05/02/2015 02:34, David Ahern a écrit :
> +static inline
> +int dev_net_ctx_eq(const struct net_device *dev, struct net_ctx *ctx)
> +{
> + if (net_eq(dev_net(dev), ctx->net))
> + return 1;
Why not using net_ctx_eq()?
^ permalink raw reply
* [PATCH] dsa: do not dereference non-existing routing table
From: Tobias Waldekranz @ 2015-02-05 13:52 UTC (permalink / raw)
To: netdev
In the case where there is only one switch, no routing table will have
been allocated, so do not dereference it in this case.
Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
---
drivers/net/dsa/mv88e6131.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/drivers/net/dsa/mv88e6131.c b/drivers/net/dsa/mv88e6131.c
index 1fc98e7..eadb27c 100644
--- a/drivers/net/dsa/mv88e6131.c
+++ b/drivers/net/dsa/mv88e6131.c
@@ -142,7 +142,8 @@ static int mv88e6131_setup_global(struct dsa_switch *ds)
int nexthop;
nexthop = 0x1f;
- if (i != ds->index && i < ds->dst->pd->nr_chips)
+ if (ds->pd->rtable &&
+ i != ds->index && i < ds->dst->pd->nr_chips)
nexthop = ds->pd->rtable[i] & 0x1f;
REG_WRITE(REG_GLOBAL2, 0x06, 0x8000 | (i << 8) | nexthop);
--
1.8.4.357.g8d83871.dirty
^ permalink raw reply related
* [PATCH net-next 1/3] net/bonding: Fix potential bad memory access during bonding events
From: Or Gerlitz @ 2015-02-05 13:52 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Amir Vadai, Tal Alon, Moni Shoua,
Or Gerlitz
In-Reply-To: <1423144366-11814-1-git-send-email-ogerlitz@mellanox.com>
From: Moni Shoua <monis@mellanox.com>
When queuing work to send the NETDEV_BONDING_INFO netdev event, it's
possible that when the work is executed, the pointer to the slave
becomes invalid. This can happen if between queuing the event and the
execution of the work, the net-device was un-ensvaled and re-enslaved.
Fix that by queuing a work with the data of the slave instead of the
slave structure.
Fixes: 69e6113343cf ('net/bonding: Notify state change on slaves')
Reported-by: Nikolay Aleksandrov <nikolay@redhat.com>
Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/net/bonding/bond_main.c | 28 +++++++++++-----------------
include/net/bonding.h | 2 +-
2 files changed, 12 insertions(+), 18 deletions(-)
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 679ef00..b979c26 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -1196,18 +1196,11 @@ static void bond_fill_ifslave(struct slave *slave, struct ifslave *info)
info->link_failure_count = slave->link_failure_count;
}
-static void bond_netdev_notify(struct slave *slave, struct net_device *dev)
+static void bond_netdev_notify(struct net_device *dev,
+ struct netdev_bonding_info *info)
{
- struct bonding *bond = slave->bond;
- struct netdev_bonding_info bonding_info;
-
rtnl_lock();
- /* make sure that slave is still valid */
- if (dev->priv_flags & IFF_BONDING) {
- bond_fill_ifslave(slave, &bonding_info.slave);
- bond_fill_ifbond(bond, &bonding_info.master);
- netdev_bonding_info_change(slave->dev, &bonding_info);
- }
+ netdev_bonding_info_change(dev, info);
rtnl_unlock();
}
@@ -1216,25 +1209,26 @@ static void bond_netdev_notify_work(struct work_struct *_work)
struct netdev_notify_work *w =
container_of(_work, struct netdev_notify_work, work.work);
- bond_netdev_notify(w->slave, w->dev);
+ bond_netdev_notify(w->dev, &w->bonding_info);
dev_put(w->dev);
+ kfree(w);
}
void bond_queue_slave_event(struct slave *slave)
{
+ struct bonding *bond = slave->bond;
struct netdev_notify_work *nnw = kzalloc(sizeof(*nnw), GFP_ATOMIC);
if (!nnw)
return;
- INIT_DELAYED_WORK(&nnw->work, bond_netdev_notify_work);
- nnw->slave = slave;
+ dev_hold(slave->dev);
nnw->dev = slave->dev;
+ bond_fill_ifslave(slave, &nnw->bonding_info.slave);
+ bond_fill_ifbond(bond, &nnw->bonding_info.master);
+ INIT_DELAYED_WORK(&nnw->work, bond_netdev_notify_work);
- if (queue_delayed_work(slave->bond->wq, &nnw->work, 0))
- dev_hold(slave->dev);
- else
- kfree(nnw);
+ queue_delayed_work(slave->bond->wq, &nnw->work, 0);
}
/* enslave device <slave> to bond device <master> */
diff --git a/include/net/bonding.h b/include/net/bonding.h
index 4e17095..fda6fee 100644
--- a/include/net/bonding.h
+++ b/include/net/bonding.h
@@ -152,8 +152,8 @@ struct bond_parm_tbl {
struct netdev_notify_work {
struct delayed_work work;
- struct slave *slave;
struct net_device *dev;
+ struct netdev_bonding_info bonding_info;
};
struct slave {
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 2/3] IB/mlx4: Always use the correct port for mirrored multicast attachments
From: Or Gerlitz @ 2015-02-05 13:52 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Amir Vadai, Tal Alon, Moni Shoua,
Or Gerlitz
In-Reply-To: <1423144366-11814-1-git-send-email-ogerlitz@mellanox.com>
From: Moni Shoua <monis@mellanox.com>
When attaching a QP to a multicast address in bonded mode, there was an
assumption that the port of the QP must be #1. This assumption isn't the
case under the flow which enables maximal usage of the physical ports.
Fix it by always checking the port of the original flow and create the
mirrored flow on the other port.
Fixes: c6215745b66a ("IB/mlx4: Load balance ports in port aggregation mode")
Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/infiniband/hw/mlx4/main.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 2ed5b99..271d3f1 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -1186,6 +1186,9 @@ static struct ib_flow *mlx4_ib_create_flow(struct ib_qp *qp,
goto err_create_flow;
i++;
if (is_bonded) {
+ /* Application always sees one port so the mirror rule
+ * must be on port #2
+ */
flow_attr->port = 2;
err = __mlx4_ib_create_flow(qp, flow_attr,
domain, type[j],
@@ -1279,14 +1282,15 @@ static int mlx4_ib_mcg_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw, mqp->port,
!!(mqp->flags &
- MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK),
+ MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK),
prot, ®_id.id);
if (err)
goto err_malloc;
reg_id.mirror = 0;
if (mlx4_is_bonded(dev)) {
- err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw, 2,
+ err = mlx4_multicast_attach(mdev->dev, &mqp->mqp, gid->raw,
+ (mqp->port == 1) ? 2 : 1,
!!(mqp->flags &
MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK),
prot, ®_id.mirror);
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 3/3] IB/mlx4_ib: Reset flow support for IB kernel ULPs
From: Or Gerlitz @ 2015-02-05 13:52 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Amir Vadai, Tal Alon, Yishai Hadas,
Or Gerlitz
In-Reply-To: <1423144366-11814-1-git-send-email-ogerlitz@mellanox.com>
From: Yishai Hadas <yishaih@mellanox.com>
The driver exposes interfaces that directly relate to HW state. Upon fatal
error, consumers of these interfaces (ULPs) that rely on completion of
all their posted work-request could hang, thereby introducing dependencies
in shutdown order. To prevent this from happening, we manage the
relevant resources (CQs, QPs) that are used by the device. Upon a fatal error,
we now generate simulated completions for outstanding WQEs that were not
completed at the time the HW was reset.
It includes invoking the completion event handler for all involved CQs so that
the ULPs will poll those CQs. When polled we return simulated CQEs with
IB_WC_WR_FLUSH_ERR return code enabling ULPs to clean up their resources and
not wait forever for completions upon receiving remove_one.
The above change requires an extra check in the data path to make sure that when
device is in error state, the simulated CQEs will be returned and no further
WQEs will be posted.
Signed-off-by: Yishai Hadas <yishaih@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/infiniband/hw/mlx4/cq.c | 57 ++++++++++++++++++++++++++++++
drivers/infiniband/hw/mlx4/main.c | 64 ++++++++++++++++++++++++++++++++++
drivers/infiniband/hw/mlx4/mlx4_ib.h | 9 +++++
drivers/infiniband/hw/mlx4/qp.c | 59 ++++++++++++++++++++++++++++---
drivers/infiniband/hw/mlx4/srq.c | 8 ++++
include/linux/mlx4/device.h | 2 +
6 files changed, 193 insertions(+), 6 deletions(-)
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index a3b70f6..543ecdd 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -188,6 +188,8 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
spin_lock_init(&cq->lock);
cq->resize_buf = NULL;
cq->resize_umem = NULL;
+ INIT_LIST_HEAD(&cq->send_qp_list);
+ INIT_LIST_HEAD(&cq->recv_qp_list);
if (context) {
struct mlx4_ib_create_cq ucmd;
@@ -594,6 +596,55 @@ static int use_tunnel_data(struct mlx4_ib_qp *qp, struct mlx4_ib_cq *cq, struct
return 0;
}
+static void mlx4_ib_qp_sw_comp(struct mlx4_ib_qp *qp, int num_entries,
+ struct ib_wc *wc, int *npolled, int is_send)
+{
+ struct mlx4_ib_wq *wq;
+ unsigned cur;
+ int i;
+
+ wq = is_send ? &qp->sq : &qp->rq;
+ cur = wq->head - wq->tail;
+
+ if (cur == 0)
+ return;
+
+ for (i = 0; i < cur && *npolled < num_entries; i++) {
+ wc->wr_id = wq->wrid[wq->tail & (wq->wqe_cnt - 1)];
+ wc->status = IB_WC_WR_FLUSH_ERR;
+ wc->vendor_err = MLX4_CQE_SYNDROME_WR_FLUSH_ERR;
+ wq->tail++;
+ (*npolled)++;
+ wc->qp = &qp->ibqp;
+ wc++;
+ }
+}
+
+static void mlx4_ib_poll_sw_comp(struct mlx4_ib_cq *cq, int num_entries,
+ struct ib_wc *wc, int *npolled)
+{
+ struct mlx4_ib_qp *qp;
+
+ *npolled = 0;
+ /* Find uncompleted WQEs belonging to that cq and retrun
+ * simulated FLUSH_ERR completions
+ */
+ list_for_each_entry(qp, &cq->send_qp_list, cq_send_list) {
+ mlx4_ib_qp_sw_comp(qp, num_entries, wc, npolled, 1);
+ if (*npolled >= num_entries)
+ goto out;
+ }
+
+ list_for_each_entry(qp, &cq->recv_qp_list, cq_recv_list) {
+ mlx4_ib_qp_sw_comp(qp, num_entries, wc + *npolled, npolled, 0);
+ if (*npolled >= num_entries)
+ goto out;
+ }
+
+out:
+ return;
+}
+
static int mlx4_ib_poll_one(struct mlx4_ib_cq *cq,
struct mlx4_ib_qp **cur_qp,
struct ib_wc *wc)
@@ -836,8 +887,13 @@ int mlx4_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc)
unsigned long flags;
int npolled;
int err = 0;
+ struct mlx4_ib_dev *mdev = to_mdev(cq->ibcq.device);
spin_lock_irqsave(&cq->lock, flags);
+ if (mdev->dev->persist->state & MLX4_DEVICE_STATE_INTERNAL_ERROR) {
+ mlx4_ib_poll_sw_comp(cq, num_entries, wc, &npolled);
+ goto out;
+ }
for (npolled = 0; npolled < num_entries; ++npolled) {
err = mlx4_ib_poll_one(cq, &cur_qp, wc + npolled);
@@ -847,6 +903,7 @@ int mlx4_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc)
mlx4_cq_set_ci(&cq->mcq);
+out:
spin_unlock_irqrestore(&cq->lock, flags);
if (err == 0 || err == -EAGAIN)
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 271d3f1..66cb011 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2308,6 +2308,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
spin_lock_init(&ibdev->sm_lock);
mutex_init(&ibdev->cap_mask_mutex);
+ INIT_LIST_HEAD(&ibdev->qp_list);
+ spin_lock_init(&ibdev->reset_flow_resource_lock);
if (ibdev->steering_support == MLX4_STEERING_MODE_DEVICE_MANAGED &&
ib_num_ports) {
@@ -2622,6 +2624,67 @@ out:
return;
}
+static void mlx4_ib_handle_catas_error(struct mlx4_ib_dev *ibdev)
+{
+ struct mlx4_ib_qp *mqp;
+ unsigned long flags_qp;
+ unsigned long flags_cq;
+ struct mlx4_ib_cq *send_mcq, *recv_mcq;
+ struct list_head cq_notify_list;
+ struct mlx4_cq *mcq;
+ unsigned long flags;
+
+ pr_warn("mlx4_ib_handle_catas_error was started\n");
+ INIT_LIST_HEAD(&cq_notify_list);
+
+ /* Go over qp list reside on that ibdev, sync with create/destroy qp.*/
+ spin_lock_irqsave(&ibdev->reset_flow_resource_lock, flags);
+
+ list_for_each_entry(mqp, &ibdev->qp_list, qps_list) {
+ spin_lock_irqsave(&mqp->sq.lock, flags_qp);
+ if (mqp->sq.tail != mqp->sq.head) {
+ send_mcq = to_mcq(mqp->ibqp.send_cq);
+ spin_lock_irqsave(&send_mcq->lock, flags_cq);
+ if (send_mcq->mcq.comp &&
+ mqp->ibqp.send_cq->comp_handler) {
+ if (!send_mcq->mcq.reset_notify_added) {
+ send_mcq->mcq.reset_notify_added = 1;
+ list_add_tail(&send_mcq->mcq.reset_notify,
+ &cq_notify_list);
+ }
+ }
+ spin_unlock_irqrestore(&send_mcq->lock, flags_cq);
+ }
+ spin_unlock_irqrestore(&mqp->sq.lock, flags_qp);
+ /* Now, handle the QP's receive queue */
+ spin_lock_irqsave(&mqp->rq.lock, flags_qp);
+ /* no handling is needed for SRQ */
+ if (!mqp->ibqp.srq) {
+ if (mqp->rq.tail != mqp->rq.head) {
+ recv_mcq = to_mcq(mqp->ibqp.recv_cq);
+ spin_lock_irqsave(&recv_mcq->lock, flags_cq);
+ if (recv_mcq->mcq.comp &&
+ mqp->ibqp.recv_cq->comp_handler) {
+ if (!recv_mcq->mcq.reset_notify_added) {
+ recv_mcq->mcq.reset_notify_added = 1;
+ list_add_tail(&recv_mcq->mcq.reset_notify,
+ &cq_notify_list);
+ }
+ }
+ spin_unlock_irqrestore(&recv_mcq->lock,
+ flags_cq);
+ }
+ }
+ spin_unlock_irqrestore(&mqp->rq.lock, flags_qp);
+ }
+
+ list_for_each_entry(mcq, &cq_notify_list, reset_notify) {
+ mcq->comp(mcq);
+ }
+ spin_unlock_irqrestore(&ibdev->reset_flow_resource_lock, flags);
+ pr_warn("mlx4_ib_handle_catas_error ended\n");
+}
+
static void handle_bonded_port_state_event(struct work_struct *work)
{
struct ib_event_work *ew =
@@ -2701,6 +2764,7 @@ static void mlx4_ib_event(struct mlx4_dev *dev, void *ibdev_ptr,
case MLX4_DEV_EVENT_CATASTROPHIC_ERROR:
ibdev->ib_active = false;
ibev.event = IB_EVENT_DEVICE_FATAL;
+ mlx4_ib_handle_catas_error(ibdev);
break;
case MLX4_DEV_EVENT_PORT_MGMT_CHANGE:
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 721540c..f829fd9 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -110,6 +110,9 @@ struct mlx4_ib_cq {
struct mutex resize_mutex;
struct ib_umem *umem;
struct ib_umem *resize_umem;
+ /* List of qps that it serves.*/
+ struct list_head send_qp_list;
+ struct list_head recv_qp_list;
};
struct mlx4_ib_mr {
@@ -300,6 +303,9 @@ struct mlx4_ib_qp {
struct mlx4_roce_smac_vlan_info pri;
struct mlx4_roce_smac_vlan_info alt;
u64 reg_id;
+ struct list_head qps_list;
+ struct list_head cq_recv_list;
+ struct list_head cq_send_list;
};
struct mlx4_ib_srq {
@@ -535,6 +541,9 @@ struct mlx4_ib_dev {
/* lock when destroying qp1_proxy and getting netdev events */
struct mutex qp1_proxy_lock[MLX4_MAX_PORTS];
u8 bond_next_port;
+ /* protect resources needed as part of reset flow */
+ spinlock_t reset_flow_resource_lock;
+ struct list_head qp_list;
};
struct ib_event_work {
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 792f9dc..dfc6ca1 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -46,6 +46,11 @@
#include "mlx4_ib.h"
#include "user.h"
+static void mlx4_ib_lock_cqs(struct mlx4_ib_cq *send_cq,
+ struct mlx4_ib_cq *recv_cq);
+static void mlx4_ib_unlock_cqs(struct mlx4_ib_cq *send_cq,
+ struct mlx4_ib_cq *recv_cq);
+
enum {
MLX4_IB_ACK_REQ_FREQ = 8,
};
@@ -618,6 +623,8 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
struct mlx4_ib_sqp *sqp;
struct mlx4_ib_qp *qp;
enum mlx4_ib_qp_type qp_type = (enum mlx4_ib_qp_type) init_attr->qp_type;
+ struct mlx4_ib_cq *mcq;
+ unsigned long flags;
/* When tunneling special qps, we use a plain UD qp */
if (sqpn) {
@@ -828,6 +835,24 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
qp->mqp.event = mlx4_ib_qp_event;
if (!*caller_qp)
*caller_qp = qp;
+
+ spin_lock_irqsave(&dev->reset_flow_resource_lock, flags);
+ mlx4_ib_lock_cqs(to_mcq(init_attr->send_cq),
+ to_mcq(init_attr->recv_cq));
+ /* Maintain device to QPs access, needed for further handling
+ * via reset flow
+ */
+ list_add_tail(&qp->qps_list, &dev->qp_list);
+ /* Maintain CQ to QPs access, needed for further handling
+ * via reset flow
+ */
+ mcq = to_mcq(init_attr->send_cq);
+ list_add_tail(&qp->cq_send_list, &mcq->send_qp_list);
+ mcq = to_mcq(init_attr->recv_cq);
+ list_add_tail(&qp->cq_recv_list, &mcq->recv_qp_list);
+ mlx4_ib_unlock_cqs(to_mcq(init_attr->send_cq),
+ to_mcq(init_attr->recv_cq));
+ spin_unlock_irqrestore(&dev->reset_flow_resource_lock, flags);
return 0;
err_qpn:
@@ -886,13 +911,13 @@ static void mlx4_ib_lock_cqs(struct mlx4_ib_cq *send_cq, struct mlx4_ib_cq *recv
__acquires(&send_cq->lock) __acquires(&recv_cq->lock)
{
if (send_cq == recv_cq) {
- spin_lock_irq(&send_cq->lock);
+ spin_lock(&send_cq->lock);
__acquire(&recv_cq->lock);
} else if (send_cq->mcq.cqn < recv_cq->mcq.cqn) {
- spin_lock_irq(&send_cq->lock);
+ spin_lock(&send_cq->lock);
spin_lock_nested(&recv_cq->lock, SINGLE_DEPTH_NESTING);
} else {
- spin_lock_irq(&recv_cq->lock);
+ spin_lock(&recv_cq->lock);
spin_lock_nested(&send_cq->lock, SINGLE_DEPTH_NESTING);
}
}
@@ -902,13 +927,13 @@ static void mlx4_ib_unlock_cqs(struct mlx4_ib_cq *send_cq, struct mlx4_ib_cq *re
{
if (send_cq == recv_cq) {
__release(&recv_cq->lock);
- spin_unlock_irq(&send_cq->lock);
+ spin_unlock(&send_cq->lock);
} else if (send_cq->mcq.cqn < recv_cq->mcq.cqn) {
spin_unlock(&recv_cq->lock);
- spin_unlock_irq(&send_cq->lock);
+ spin_unlock(&send_cq->lock);
} else {
spin_unlock(&send_cq->lock);
- spin_unlock_irq(&recv_cq->lock);
+ spin_unlock(&recv_cq->lock);
}
}
@@ -953,6 +978,7 @@ static void destroy_qp_common(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp,
int is_user)
{
struct mlx4_ib_cq *send_cq, *recv_cq;
+ unsigned long flags;
if (qp->state != IB_QPS_RESET) {
if (mlx4_qp_modify(dev->dev, NULL, to_mlx4_state(qp->state),
@@ -984,8 +1010,13 @@ static void destroy_qp_common(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp,
get_cqs(qp, &send_cq, &recv_cq);
+ spin_lock_irqsave(&dev->reset_flow_resource_lock, flags);
mlx4_ib_lock_cqs(send_cq, recv_cq);
+ /* del from lists under both locks above to protect reset flow paths */
+ list_del(&qp->qps_list);
+ list_del(&qp->cq_send_list);
+ list_del(&qp->cq_recv_list);
if (!is_user) {
__mlx4_ib_cq_clean(recv_cq, qp->mqp.qpn,
qp->ibqp.srq ? to_msrq(qp->ibqp.srq): NULL);
@@ -996,6 +1027,7 @@ static void destroy_qp_common(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp,
mlx4_qp_remove(dev->dev, &qp->mqp);
mlx4_ib_unlock_cqs(send_cq, recv_cq);
+ spin_unlock_irqrestore(&dev->reset_flow_resource_lock, flags);
mlx4_qp_free(dev->dev, &qp->mqp);
@@ -2618,8 +2650,15 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
__be32 uninitialized_var(lso_hdr_sz);
__be32 blh;
int i;
+ struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
spin_lock_irqsave(&qp->sq.lock, flags);
+ if (mdev->dev->persist->state & MLX4_DEVICE_STATE_INTERNAL_ERROR) {
+ err = -EIO;
+ *bad_wr = wr;
+ nreq = 0;
+ goto out;
+ }
ind = qp->sq_next_wqe;
@@ -2917,10 +2956,18 @@ int mlx4_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr,
int ind;
int max_gs;
int i;
+ struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
max_gs = qp->rq.max_gs;
spin_lock_irqsave(&qp->rq.lock, flags);
+ if (mdev->dev->persist->state & MLX4_DEVICE_STATE_INTERNAL_ERROR) {
+ err = -EIO;
+ *bad_wr = wr;
+ nreq = 0;
+ goto out;
+ }
+
ind = qp->rq.head & (qp->rq.wqe_cnt - 1);
for (nreq = 0; wr; ++nreq, wr = wr->next) {
diff --git a/drivers/infiniband/hw/mlx4/srq.c b/drivers/infiniband/hw/mlx4/srq.c
index 62d9285..dce5dfe 100644
--- a/drivers/infiniband/hw/mlx4/srq.c
+++ b/drivers/infiniband/hw/mlx4/srq.c
@@ -316,8 +316,15 @@ int mlx4_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
int err = 0;
int nreq;
int i;
+ struct mlx4_ib_dev *mdev = to_mdev(ibsrq->device);
spin_lock_irqsave(&srq->lock, flags);
+ if (mdev->dev->persist->state & MLX4_DEVICE_STATE_INTERNAL_ERROR) {
+ err = -EIO;
+ *bad_wr = wr;
+ nreq = 0;
+ goto out;
+ }
for (nreq = 0; wr; ++nreq, wr = wr->next) {
if (unlikely(wr->num_sge > srq->msrq.max_gs)) {
@@ -362,6 +369,7 @@ int mlx4_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
*srq->db.db = cpu_to_be32(srq->wqe_ctr);
}
+out:
spin_unlock_irqrestore(&srq->lock, flags);
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 977b0b1..e742201 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -689,6 +689,8 @@ struct mlx4_cq {
void (*comp)(struct mlx4_cq *);
void *priv;
} tasklet_ctx;
+ int reset_notify_added;
+ struct list_head reset_notify;
};
struct mlx4_qp {
--
1.7.1
^ permalink raw reply related
* [PATCH net-next 0/3] mlx4 driver fixes for the HA/LAG support and reset flow
From: Or Gerlitz @ 2015-02-05 13:52 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev, Roland Dreier, Amir Vadai, Tal Alon, Or Gerlitz
Hi Dave,
There are two fixes to mlx4 HA/LAG support from Moni and a patch from Yishai
which does further hardening of the mlx4 reset support for IB kernel ULPs.
Or.
Moni Shoua (2):
net/bonding: Fix potential bad memory access during bonding events
IB/mlx4: Always use the correct port for mirrored multicast attachments
Yishai Hadas (1):
IB/mlx4_ib: Reset flow support for IB kernel ULPs
drivers/infiniband/hw/mlx4/cq.c | 57 +++++++++++++++++++++++++++
drivers/infiniband/hw/mlx4/main.c | 72 +++++++++++++++++++++++++++++++++-
drivers/infiniband/hw/mlx4/mlx4_ib.h | 9 ++++
drivers/infiniband/hw/mlx4/qp.c | 59 +++++++++++++++++++++++++---
drivers/infiniband/hw/mlx4/srq.c | 8 ++++
drivers/net/bonding/bond_main.c | 28 +++++--------
include/linux/mlx4/device.h | 2 +
include/net/bonding.h | 2 +-
8 files changed, 211 insertions(+), 26 deletions(-)
^ permalink raw reply
* [PATCH] dsa: correctly determine the number of switches in a system
From: Tobias Waldekranz @ 2015-02-05 13:54 UTC (permalink / raw)
To: netdev
The number of connected switches was sourced from the number of
children to the DSA node, change it to the number of available
children, skipping any disabled switches.
Signed-off-by: Tobias Waldekranz <tobias@waldekranz.com>
---
net/dsa/dsa.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/dsa/dsa.c b/net/dsa/dsa.c
index 3731714..2173402 100644
--- a/net/dsa/dsa.c
+++ b/net/dsa/dsa.c
@@ -603,7 +603,7 @@ static int dsa_of_probe(struct platform_device *pdev)
pdev->dev.platform_data = pd;
pd->netdev = ðernet_dev->dev;
- pd->nr_chips = of_get_child_count(np);
+ pd->nr_chips = of_get_available_child_count(np);
if (pd->nr_chips > DSA_MAX_SWITCHES)
pd->nr_chips = DSA_MAX_SWITCHES;
--
1.8.4.357.g8d83871.dirty
^ permalink raw reply related
* [PATCH] vxge: fix sparse warning
From: Lad Prabhakar @ 2015-02-05 13:56 UTC (permalink / raw)
To: Jon Mason, David S. Miller, netdev; +Cc: linux-kernel, Lad, Prabhakar
From: "Lad, Prabhakar" <prabhakar.csengg@gmail.com>
this patch fixes following sparse warning:
vxge-config.c:4640:30: warning: Using plain integer as NULL pointer
Signed-off-by: Lad, Prabhakar <prabhakar.csengg@gmail.com>
---
Found this issue on linux-next (gcc version 4.9.2,
sparse version 0.4.5-rc1)and applies on top linux-next.
drivers/net/ethernet/neterion/vxge/vxge-config.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/neterion/vxge/vxge-config.c b/drivers/net/ethernet/neterion/vxge/vxge-config.c
index 2bbd01f..6223930 100644
--- a/drivers/net/ethernet/neterion/vxge/vxge-config.c
+++ b/drivers/net/ethernet/neterion/vxge/vxge-config.c
@@ -4637,7 +4637,7 @@ static void __vxge_hw_vp_terminate(struct __vxge_hw_device *hldev, u32 vp_id)
vpath->ringh = NULL;
vpath->fifoh = NULL;
memset(&vpath->vpath_handles, 0, sizeof(struct list_head));
- vpath->stats_block = 0;
+ vpath->stats_block = NULL;
vpath->hw_stats = NULL;
vpath->hw_stats_sav = NULL;
vpath->sw_stats = NULL;
--
1.9.1
^ permalink raw reply related
* Re: [ovs-dev] [RFC: add openvswitch actions using BPF 2/9] odp: add a new ODP action: OVS_ACTION_ATTR_BPF_PROG
From: Thomas Graf @ 2015-02-05 13:56 UTC (permalink / raw)
To: Andy Zhou; +Cc: dev, netdev
In-Reply-To: <1423090163-19902-3-git-send-email-azhou@nicira.com>
On 02/04/15 at 02:49pm, Andy Zhou wrote:
> /**
> + * struct ovs_action_bpf_prog - %OVS_ACTION_ATTR_BPF_PROG action argument.
> + *
> + * XXX provides bpf program id and execution context.
> + *
> + */
> +struct ovs_action_bpf_prog {
> + __be32 prog_fd;
> + __be32 arg0;
> + __be32 arg1;
> +};
I assume you intend to replace this with a nested Netlink attribute with a
series of U32 attributes to carry args in the future?
> +enum ovs_bpf_func_id {
> + OVS_BPF_FUNC_unspec,
> + OVS_BPF_FUNC_output, /* int ovs_bpf_output(ctxt) */
> + __OVS_BPF_FUNC_MAX_ID,
> +};
The mix of cap / no-cap looks a bit weird.
> +static void
> +format_bpf_prog_action(struct ds *ds, const struct nlattr *a)
> +{
> + struct ovs_action_bpf_prog *bpf_act = (struct ovs_action_bpf_prog *)
> + nl_attr_get(a);
> +
> + ds_put_cstr(ds, "bpf");
> + ds_put_format(ds, "(%u,%u,%u)", ntohl(bpf_act->prog_fd),
> + ntohl(bpf_act->arg0), ntohl(bpf_act->arg1));
> +}
Maybe ds_put_format(ds, "bpf(... ?
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox