summaryrefslogtreecommitdiffstats
path: root/net/rds
diff options
context:
space:
mode:
Diffstat (limited to 'net/rds')
-rw-r--r--net/rds/connection.c1
-rw-r--r--net/rds/ib.c10
-rw-r--r--net/rds/ib_cm.c47
-rw-r--r--net/rds/ib_mr.h2
-rw-r--r--net/rds/page.c29
-rw-r--r--net/rds/rds.h15
-rw-r--r--net/rds/tcp.c44
-rw-r--r--net/rds/tcp.h2
-rw-r--r--net/rds/tcp_listen.c11
-rw-r--r--net/rds/transport.c4
10 files changed, 76 insertions, 89 deletions
diff --git a/net/rds/connection.c b/net/rds/connection.c
index 0e04dcceb1d4..1fa75ab7b733 100644
--- a/net/rds/connection.c
+++ b/net/rds/connection.c
@@ -429,6 +429,7 @@ void rds_conn_destroy(struct rds_connection *conn)
*/
rds_cong_remove_conn(conn);
+ put_net(conn->c_net);
kmem_cache_free(rds_conn_slab, conn);
spin_lock_irqsave(&rds_conn_lock, flags);
diff --git a/net/rds/ib.c b/net/rds/ib.c
index 91fe46f1e4cc..7a64c8db81ab 100644
--- a/net/rds/ib.c
+++ b/net/rds/ib.c
@@ -45,8 +45,8 @@
#include "ib.h"
#include "ib_mr.h"
-unsigned int rds_ib_mr_1m_pool_size = RDS_MR_1M_POOL_SIZE;
-unsigned int rds_ib_mr_8k_pool_size = RDS_MR_8K_POOL_SIZE;
+static unsigned int rds_ib_mr_1m_pool_size = RDS_MR_1M_POOL_SIZE;
+static unsigned int rds_ib_mr_8k_pool_size = RDS_MR_8K_POOL_SIZE;
unsigned int rds_ib_retry_count = RDS_IB_DEFAULT_RETRY_COUNT;
module_param(rds_ib_mr_1m_pool_size, int, 0444);
@@ -438,16 +438,12 @@ int rds_ib_init(void)
if (ret)
goto out_sysctl;
- ret = rds_trans_register(&rds_ib_transport);
- if (ret)
- goto out_recv;
+ rds_trans_register(&rds_ib_transport);
rds_info_register_func(RDS_INFO_IB_CONNECTIONS, rds_ib_ic_info);
goto out;
-out_recv:
- rds_ib_recv_exit();
out_sysctl:
rds_ib_sysctl_exit();
out_ibreg:
diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c
index ce3775abc6e7..1c38d2c7caa8 100644
--- a/net/rds/ib_cm.c
+++ b/net/rds/ib_cm.c
@@ -442,7 +442,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
ic->i_send_cq = NULL;
ibdev_put_vector(rds_ibdev, ic->i_scq_vector);
rdsdebug("ib_create_cq send failed: %d\n", ret);
- goto out;
+ goto rds_ibdev_out;
}
ic->i_rcq_vector = ibdev_get_unused_vector(rds_ibdev);
@@ -456,19 +456,19 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
ic->i_recv_cq = NULL;
ibdev_put_vector(rds_ibdev, ic->i_rcq_vector);
rdsdebug("ib_create_cq recv failed: %d\n", ret);
- goto out;
+ goto send_cq_out;
}
ret = ib_req_notify_cq(ic->i_send_cq, IB_CQ_NEXT_COMP);
if (ret) {
rdsdebug("ib_req_notify_cq send failed: %d\n", ret);
- goto out;
+ goto recv_cq_out;
}
ret = ib_req_notify_cq(ic->i_recv_cq, IB_CQ_SOLICITED);
if (ret) {
rdsdebug("ib_req_notify_cq recv failed: %d\n", ret);
- goto out;
+ goto recv_cq_out;
}
/* XXX negotiate max send/recv with remote? */
@@ -494,7 +494,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
ret = rdma_create_qp(ic->i_cm_id, ic->i_pd, &attr);
if (ret) {
rdsdebug("rdma_create_qp failed: %d\n", ret);
- goto out;
+ goto recv_cq_out;
}
ic->i_send_hdrs = ib_dma_alloc_coherent(dev,
@@ -504,7 +504,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
if (!ic->i_send_hdrs) {
ret = -ENOMEM;
rdsdebug("ib_dma_alloc_coherent send failed\n");
- goto out;
+ goto qp_out;
}
ic->i_recv_hdrs = ib_dma_alloc_coherent(dev,
@@ -514,7 +514,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
if (!ic->i_recv_hdrs) {
ret = -ENOMEM;
rdsdebug("ib_dma_alloc_coherent recv failed\n");
- goto out;
+ goto send_hdrs_dma_out;
}
ic->i_ack = ib_dma_alloc_coherent(dev, sizeof(struct rds_header),
@@ -522,7 +522,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
if (!ic->i_ack) {
ret = -ENOMEM;
rdsdebug("ib_dma_alloc_coherent ack failed\n");
- goto out;
+ goto recv_hdrs_dma_out;
}
ic->i_sends = vzalloc_node(ic->i_send_ring.w_nr * sizeof(struct rds_ib_send_work),
@@ -530,7 +530,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
if (!ic->i_sends) {
ret = -ENOMEM;
rdsdebug("send allocation failed\n");
- goto out;
+ goto ack_dma_out;
}
ic->i_recvs = vzalloc_node(ic->i_recv_ring.w_nr * sizeof(struct rds_ib_recv_work),
@@ -538,7 +538,7 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
if (!ic->i_recvs) {
ret = -ENOMEM;
rdsdebug("recv allocation failed\n");
- goto out;
+ goto sends_out;
}
rds_ib_recv_init_ack(ic);
@@ -546,8 +546,33 @@ static int rds_ib_setup_qp(struct rds_connection *conn)
rdsdebug("conn %p pd %p cq %p %p\n", conn, ic->i_pd,
ic->i_send_cq, ic->i_recv_cq);
-out:
+ return ret;
+
+sends_out:
+ vfree(ic->i_sends);
+ack_dma_out:
+ ib_dma_free_coherent(dev, sizeof(struct rds_header),
+ ic->i_ack, ic->i_ack_dma);
+recv_hdrs_dma_out:
+ ib_dma_free_coherent(dev, ic->i_recv_ring.w_nr *
+ sizeof(struct rds_header),
+ ic->i_recv_hdrs, ic->i_recv_hdrs_dma);
+send_hdrs_dma_out:
+ ib_dma_free_coherent(dev, ic->i_send_ring.w_nr *
+ sizeof(struct rds_header),
+ ic->i_send_hdrs, ic->i_send_hdrs_dma);
+qp_out:
+ rdma_destroy_qp(ic->i_cm_id);
+recv_cq_out:
+ if (!ib_destroy_cq(ic->i_recv_cq))
+ ic->i_recv_cq = NULL;
+send_cq_out:
+ if (!ib_destroy_cq(ic->i_send_cq))
+ ic->i_send_cq = NULL;
+rds_ibdev_out:
+ rds_ib_remove_conn(rds_ibdev, conn);
rds_ib_dev_put(rds_ibdev);
+
return ret;
}
diff --git a/net/rds/ib_mr.h b/net/rds/ib_mr.h
index 24c086db4511..5d6e98a79a5e 100644
--- a/net/rds/ib_mr.h
+++ b/net/rds/ib_mr.h
@@ -107,8 +107,6 @@ struct rds_ib_mr_pool {
};
extern struct workqueue_struct *rds_ib_mr_wq;
-extern unsigned int rds_ib_mr_1m_pool_size;
-extern unsigned int rds_ib_mr_8k_pool_size;
extern bool prefer_frmr;
struct rds_ib_mr_pool *rds_ib_create_mr_pool(struct rds_ib_device *rds_dev,
diff --git a/net/rds/page.c b/net/rds/page.c
index e2b5a5832d3d..7cc57e098ddb 100644
--- a/net/rds/page.c
+++ b/net/rds/page.c
@@ -45,35 +45,6 @@ struct rds_page_remainder {
static
DEFINE_PER_CPU_SHARED_ALIGNED(struct rds_page_remainder, rds_page_remainders);
-/*
- * returns 0 on success or -errno on failure.
- *
- * We don't have to worry about flush_dcache_page() as this only works
- * with private pages. If, say, we were to do directed receive to pinned
- * user pages we'd have to worry more about cache coherence. (Though
- * the flush_dcache_page() in get_user_pages() would probably be enough).
- */
-int rds_page_copy_user(struct page *page, unsigned long offset,
- void __user *ptr, unsigned long bytes,
- int to_user)
-{
- unsigned long ret;
- void *addr;
-
- addr = kmap(page);
- if (to_user) {
- rds_stats_add(s_copy_to_user, bytes);
- ret = copy_to_user(ptr, addr + offset, bytes);
- } else {
- rds_stats_add(s_copy_from_user, bytes);
- ret = copy_from_user(addr + offset, ptr, bytes);
- }
- kunmap(page);
-
- return ret ? -EFAULT : 0;
-}
-EXPORT_SYMBOL_GPL(rds_page_copy_user);
-
/**
* rds_page_remainder_alloc - build up regions of a message.
*
diff --git a/net/rds/rds.h b/net/rds/rds.h
index 07fff73dd4f3..82d38ccf5e8b 100644
--- a/net/rds/rds.h
+++ b/net/rds/rds.h
@@ -147,7 +147,7 @@ struct rds_connection {
/* Protocol version */
unsigned int c_version;
- possible_net_t c_net;
+ struct net *c_net;
struct list_head c_map_item;
unsigned long c_map_queued;
@@ -162,13 +162,13 @@ struct rds_connection {
static inline
struct net *rds_conn_net(struct rds_connection *conn)
{
- return read_pnet(&conn->c_net);
+ return conn->c_net;
}
static inline
void rds_conn_net_set(struct rds_connection *conn, struct net *net)
{
- write_pnet(&conn->c_net, net);
+ conn->c_net = get_net(net);
}
#define RDS_FLAG_CONG_BITMAP 0x01
@@ -798,13 +798,6 @@ static inline int rds_message_verify_checksum(const struct rds_header *hdr)
/* page.c */
int rds_page_remainder_alloc(struct scatterlist *scat, unsigned long bytes,
gfp_t gfp);
-int rds_page_copy_user(struct page *page, unsigned long offset,
- void __user *ptr, unsigned long bytes,
- int to_user);
-#define rds_page_copy_to_user(page, offset, ptr, bytes) \
- rds_page_copy_user(page, offset, ptr, bytes, 1)
-#define rds_page_copy_from_user(page, offset, ptr, bytes) \
- rds_page_copy_user(page, offset, ptr, bytes, 0)
void rds_page_exit(void);
/* recv.c */
@@ -910,7 +903,7 @@ void rds_connect_path_complete(struct rds_conn_path *conn, int curr);
void rds_connect_complete(struct rds_connection *conn);
/* transport.c */
-int rds_trans_register(struct rds_transport *trans);
+void rds_trans_register(struct rds_transport *trans);
void rds_trans_unregister(struct rds_transport *trans);
struct rds_transport *rds_trans_get_preferred(struct net *net, __be32 addr);
void rds_trans_put(struct rds_transport *trans);
diff --git a/net/rds/tcp.c b/net/rds/tcp.c
index 5438f6725092..225690076773 100644
--- a/net/rds/tcp.c
+++ b/net/rds/tcp.c
@@ -484,9 +484,10 @@ static void __net_exit rds_tcp_exit_net(struct net *net)
* we do need to clean up the listen socket here.
*/
if (rtn->rds_tcp_listen_sock) {
- rds_tcp_listen_stop(rtn->rds_tcp_listen_sock);
+ struct socket *lsock = rtn->rds_tcp_listen_sock;
+
rtn->rds_tcp_listen_sock = NULL;
- flush_work(&rtn->rds_tcp_accept_w);
+ rds_tcp_listen_stop(lsock, &rtn->rds_tcp_accept_w);
}
}
@@ -523,13 +524,13 @@ static void rds_tcp_kill_sock(struct net *net)
struct rds_tcp_connection *tc, *_tc;
LIST_HEAD(tmp_list);
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
+ struct socket *lsock = rtn->rds_tcp_listen_sock;
- rds_tcp_listen_stop(rtn->rds_tcp_listen_sock);
rtn->rds_tcp_listen_sock = NULL;
- flush_work(&rtn->rds_tcp_accept_w);
+ rds_tcp_listen_stop(lsock, &rtn->rds_tcp_accept_w);
spin_lock_irq(&rds_tcp_conn_lock);
list_for_each_entry_safe(tc, _tc, &rds_tcp_conn_list, t_tcp_node) {
- struct net *c_net = read_pnet(&tc->t_cpath->cp_conn->c_net);
+ struct net *c_net = tc->t_cpath->cp_conn->c_net;
if (net != c_net || !tc->t_sock)
continue;
@@ -546,8 +547,12 @@ static void rds_tcp_kill_sock(struct net *net)
void *rds_tcp_listen_sock_def_readable(struct net *net)
{
struct rds_tcp_net *rtn = net_generic(net, rds_tcp_netid);
+ struct socket *lsock = rtn->rds_tcp_listen_sock;
+
+ if (!lsock)
+ return NULL;
- return rtn->rds_tcp_listen_sock->sk->sk_user_data;
+ return lsock->sk->sk_user_data;
}
static int rds_tcp_dev_event(struct notifier_block *this,
@@ -584,7 +589,7 @@ static void rds_tcp_sysctl_reset(struct net *net)
spin_lock_irq(&rds_tcp_conn_lock);
list_for_each_entry_safe(tc, _tc, &rds_tcp_conn_list, t_tcp_node) {
- struct net *c_net = read_pnet(&tc->t_cpath->cp_conn->c_net);
+ struct net *c_net = tc->t_cpath->cp_conn->c_net;
if (net != c_net || !tc->t_sock)
continue;
@@ -638,35 +643,30 @@ static int rds_tcp_init(void)
goto out;
}
- ret = register_netdevice_notifier(&rds_tcp_dev_notifier);
- if (ret) {
- pr_warn("could not register rds_tcp_dev_notifier\n");
+ ret = rds_tcp_recv_init();
+ if (ret)
goto out_slab;
- }
ret = register_pernet_subsys(&rds_tcp_net_ops);
if (ret)
- goto out_notifier;
+ goto out_recv;
- ret = rds_tcp_recv_init();
- if (ret)
+ ret = register_netdevice_notifier(&rds_tcp_dev_notifier);
+ if (ret) {
+ pr_warn("could not register rds_tcp_dev_notifier\n");
goto out_pernet;
+ }
- ret = rds_trans_register(&rds_tcp_transport);
- if (ret)
- goto out_recv;
+ rds_trans_register(&rds_tcp_transport);
rds_info_register_func(RDS_INFO_TCP_SOCKETS, rds_tcp_tc_info);
goto out;
-out_recv:
- rds_tcp_recv_exit();
out_pernet:
unregister_pernet_subsys(&rds_tcp_net_ops);
-out_notifier:
- if (unregister_netdevice_notifier(&rds_tcp_dev_notifier))
- pr_warn("could not unregister rds_tcp_dev_notifier\n");
+out_recv:
+ rds_tcp_recv_exit();
out_slab:
kmem_cache_destroy(rds_tcp_conn_slab);
out:
diff --git a/net/rds/tcp.h b/net/rds/tcp.h
index 9a1cc8906576..56ea6620fcf9 100644
--- a/net/rds/tcp.h
+++ b/net/rds/tcp.h
@@ -66,7 +66,7 @@ void rds_tcp_state_change(struct sock *sk);
/* tcp_listen.c */
struct socket *rds_tcp_listen_init(struct net *);
-void rds_tcp_listen_stop(struct socket *);
+void rds_tcp_listen_stop(struct socket *sock, struct work_struct *acceptor);
void rds_tcp_listen_data_ready(struct sock *sk);
int rds_tcp_accept_one(struct socket *sock);
int rds_tcp_keepalive(struct socket *sock);
diff --git a/net/rds/tcp_listen.c b/net/rds/tcp_listen.c
index 67d0929c7d3d..507678853e6c 100644
--- a/net/rds/tcp_listen.c
+++ b/net/rds/tcp_listen.c
@@ -133,7 +133,7 @@ int rds_tcp_accept_one(struct socket *sock)
new_sock->type = sock->type;
new_sock->ops = sock->ops;
- ret = sock->ops->accept(sock, new_sock, O_NONBLOCK);
+ ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, true);
if (ret < 0)
goto out;
@@ -223,6 +223,9 @@ void rds_tcp_listen_data_ready(struct sock *sk)
* before it has been accepted and the accepter has set up their
* data_ready.. we only want to queue listen work for our listening
* socket
+ *
+ * (*ready)() may be null if we are racing with netns delete, and
+ * the listen socket is being torn down.
*/
if (sk->sk_state == TCP_LISTEN)
rds_tcp_accept_work(sk);
@@ -231,7 +234,8 @@ void rds_tcp_listen_data_ready(struct sock *sk)
out:
read_unlock_bh(&sk->sk_callback_lock);
- ready(sk);
+ if (ready)
+ ready(sk);
}
struct socket *rds_tcp_listen_init(struct net *net)
@@ -271,7 +275,7 @@ out:
return NULL;
}
-void rds_tcp_listen_stop(struct socket *sock)
+void rds_tcp_listen_stop(struct socket *sock, struct work_struct *acceptor)
{
struct sock *sk;
@@ -292,5 +296,6 @@ void rds_tcp_listen_stop(struct socket *sock)
/* wait for accepts to stop and close the socket */
flush_workqueue(rds_wq);
+ flush_work(acceptor);
sock_release(sock);
}
diff --git a/net/rds/transport.c b/net/rds/transport.c
index 2ffd3e30c643..0b188dd0a344 100644
--- a/net/rds/transport.c
+++ b/net/rds/transport.c
@@ -40,7 +40,7 @@
static struct rds_transport *transports[RDS_TRANS_COUNT];
static DECLARE_RWSEM(rds_trans_sem);
-int rds_trans_register(struct rds_transport *trans)
+void rds_trans_register(struct rds_transport *trans)
{
BUG_ON(strlen(trans->t_name) + 1 > TRANSNAMSIZ);
@@ -55,8 +55,6 @@ int rds_trans_register(struct rds_transport *trans)
}
up_write(&rds_trans_sem);
-
- return 0;
}
EXPORT_SYMBOL_GPL(rds_trans_register);
OpenPOWER on IntegriCloud