static void __vsock_release(struct sock *sk)
{
if (sk) {
struct sk_buff *skb;
struct sock *pending;
struct vsock_sock *vsk;
vsk = vsock_sk(sk);
pending = NULL; /* Compiler warning. */
transport->release(vsk);
lock_sock(sk);
sock_orphan(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
while ((skb = skb_dequeue(&sk->sk_receive_queue)))
kfree_skb(skb);
/* Clean up any sockets that never were accepted. */
while ((pending = vsock_dequeue_accept(sk)) != NULL) {
__vsock_release(pending);
sock_put(pending);
}
release_sock(sk);
sock_put(sk);
}
}
static int l2cap_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct sock *srv_sk = NULL;
int err;
BT_DBG("sock %p, sk %p", sock, sk);
if (!sk)
return 0;
/* If this is an ATT Client socket, find the matching Server */
if (l2cap_pi(sk)->scid == L2CAP_CID_LE_DATA && !l2cap_pi(sk)->incoming)
srv_sk = l2cap_find_sock_by_fixed_cid_and_dir(L2CAP_CID_LE_DATA,
&bt_sk(sk)->src, &bt_sk(sk)->dst, 1);
/* If server socket found, request tear down */
BT_DBG("client:%p server:%p", sk, srv_sk);
if (srv_sk)
l2cap_sock_set_timer(srv_sk, 1);
err = l2cap_sock_shutdown(sock, 2);
sock_orphan(sk);
l2cap_sock_kill(sk);
return err;
}
static int hci_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct hci_dev *hdev = hci_pi(sk)->hdev;
DBG("sock %p sk %p", sock, sk);
if (!sk)
return 0;
bluez_sock_unlink(&hci_sk_list, sk);
if (hdev) {
if (!hci_sock_lookup(hdev))
hdev->flags &= ~HCI_SOCK;
hci_dev_put(hdev);
}
sock_orphan(sk);
skb_queue_purge(&sk->receive_queue);
skb_queue_purge(&sk->write_queue);
sock_put(sk);
MOD_DEC_USE_COUNT;
return 0;
}
static int l2cap_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct sock *sk2 = NULL;
int err;
BT_DBG("sock %p, sk %p", sock, sk);
if (!sk)
return 0;
/* If this is an ATT socket, find it's matching server/client */
if (l2cap_pi(sk)->scid == L2CAP_CID_LE_DATA)
sk2 = l2cap_find_sock_by_fixed_cid_and_dir(L2CAP_CID_LE_DATA,
&bt_sk(sk)->src, &bt_sk(sk)->dst,
l2cap_pi(sk)->incoming ? 0 : 1);
/* If matching socket found, request tear down */
BT_DBG("sock:%p companion:%p", sk, sk2);
if (sk2)
l2cap_sock_set_timer(sk2, 1);
err = l2cap_sock_shutdown(sock, 2);
sock_orphan(sk);
l2cap_sock_kill(sk);
return err;
}
static int netlink_release(struct socket *sock)
{
struct sock *sk = sock->sk;
if (!sk)
return 0;
netlink_remove(sk);
spin_lock(&sk->protinfo.af_netlink->cb_lock);
if (sk->protinfo.af_netlink->cb) {
sk->protinfo.af_netlink->cb->done(sk->protinfo.af_netlink->cb);
netlink_destroy_callback(sk->protinfo.af_netlink->cb);
sk->protinfo.af_netlink->cb = NULL;
__sock_put(sk);
}
spin_unlock(&sk->protinfo.af_netlink->cb_lock);
/* OK. Socket is unlinked, and, therefore,
no new packets will arrive */
sock_orphan(sk);
sock->sk = NULL;
wake_up_interruptible_all(&sk->protinfo.af_netlink->wait);
skb_queue_purge(&sk->write_queue);
sock_put(sk);
return 0;
}
static int hci_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct hci_dev *hdev = hci_pi(sk)->hdev;
BT_DBG("sock %p sk %p", sock, sk);
if (!sk)
return 0;
bluez_sock_unlink(&hci_sk_list, sk);
if (hdev) {
atomic_dec(&hdev->promisc);
hci_dev_put(hdev);
}
sock_orphan(sk);
skb_queue_purge(&sk->receive_queue);
skb_queue_purge(&sk->write_queue);
sock_put(sk);
MOD_DEC_USE_COUNT;
return 0;
}
static int packet_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct sock **skp;
if (!sk)
return 0;
write_lock_bh(&packet_sklist_lock);
for (skp = &packet_sklist; *skp; skp = &(*skp)->next) {
if (*skp == sk) {
*skp = sk->next;
__sock_put(sk);
break;
}
}
write_unlock_bh(&packet_sklist_lock);
/*
* Unhook packet receive handler.
*/
if (sk->protinfo.af_packet->running) {
/*
* Remove the protocol hook
*/
dev_remove_pack(&sk->protinfo.af_packet->prot_hook);
sk->protinfo.af_packet->running = 0;
__sock_put(sk);
}
#ifdef CONFIG_PACKET_MULTICAST
packet_flush_mclist(sk);
#endif
#ifdef CONFIG_PACKET_MMAP
if (sk->protinfo.af_packet->pg_vec) {
struct tpacket_req req;
memset(&req, 0, sizeof(req));
packet_set_ring(sk, &req, 1);
}
#endif
/*
* Now the socket is dead. No more input will appear.
*/
sock_orphan(sk);
sock->sk = NULL;
/* Purge queues */
skb_queue_purge(&sk->receive_queue);
sock_put(sk);
return 0;
}
static int rawsock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
nfc_dbg("sock=%p", sock);
sock_orphan(sk);
sock_put(sk);
return 0;
}
static int packet_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct packet_sock *po;
if (!sk)
return 0;
po = pkt_sk(sk);
write_lock_bh(&packet_sklist_lock);
sk_del_node_init(sk);
write_unlock_bh(&packet_sklist_lock);
/*
* Unhook packet receive handler.
*/
if (po->running) {
/*
* Remove the protocol hook
*/
dev_remove_pack(&po->prot_hook);
po->running = 0;
po->num = 0;
__sock_put(sk);
}
#ifdef CONFIG_PACKET_MULTICAST
packet_flush_mclist(sk);
#endif
#ifdef CONFIG_PACKET_MMAP
if (po->pg_vec) {
struct tpacket_req req;
memset(&req, 0, sizeof(req));
packet_set_ring(sk, &req, 1);
}
#endif
/*
* Now the socket is dead. No more input will appear.
*/
sock_orphan(sk);
sock->sk = NULL;
/* Purge queues */
skb_queue_purge(&sk->sk_receive_queue);
sock_put(sk);
return 0;
}
static int rawsock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
pr_debug("sock=%p\n", sock);
sock_orphan(sk);
sock_put(sk);
return 0;
}
/*
* release an RxRPC socket
*/
static int rxrpc_release_sock(struct sock *sk)
{
struct rxrpc_sock *rx = rxrpc_sk(sk);
struct rxrpc_net *rxnet = rxrpc_net(sock_net(&rx->sk));
_enter("%p{%d,%d}", sk, sk->sk_state, refcount_read(&sk->sk_refcnt));
/* declare the socket closed for business */
sock_orphan(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
/* We want to kill off all connections from a service socket
* as fast as possible because we can't share these; client
* sockets, on the other hand, can share an endpoint.
*/
switch (sk->sk_state) {
case RXRPC_SERVER_BOUND:
case RXRPC_SERVER_BOUND2:
case RXRPC_SERVER_LISTENING:
case RXRPC_SERVER_LISTEN_DISABLED:
rx->local->service_closed = true;
break;
}
spin_lock_bh(&sk->sk_receive_queue.lock);
sk->sk_state = RXRPC_CLOSE;
spin_unlock_bh(&sk->sk_receive_queue.lock);
if (rx->local && rcu_access_pointer(rx->local->service) == rx) {
write_lock(&rx->local->services_lock);
rcu_assign_pointer(rx->local->service, NULL);
write_unlock(&rx->local->services_lock);
}
/* try to flush out this socket */
rxrpc_discard_prealloc(rx);
rxrpc_release_calls_on_socket(rx);
flush_workqueue(rxrpc_workqueue);
rxrpc_purge_queue(&sk->sk_receive_queue);
rxrpc_queue_work(&rxnet->service_conn_reaper);
rxrpc_queue_work(&rxnet->client_conn_reaper);
rxrpc_put_local(rx->local);
rx->local = NULL;
key_put(rx->key);
rx->key = NULL;
key_put(rx->securities);
rx->securities = NULL;
sock_put(sk);
_leave(" = 0");
return 0;
}
static int bnep_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
BT_DBG("sock %p sk %p", sock, sk);
if (!sk)
return 0;
sock_orphan(sk);
sock_put(sk);
return 0;
}
static int l2cap_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
int err;
BT_DBG("sock %p, sk %p", sock, sk);
if (!sk) return 0;
err = l2cap_sock_shutdown(sock, 2);
sock_orphan(sk);
l2cap_sock_kill(sk);
return err;
}
static int cmtp_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
BT_DBG("sock %p sk %p", sock, sk);
if (!sk)
return 0;
bt_sock_unlink(&cmtp_sk_list, sk);
sock_orphan(sk);
sock_put(sk);
return 0;
}
static int
MksckRelease(struct socket *sock)
{
struct sock *sk = sock->sk;
if (sk) {
lock_sock(sk);
sock_orphan(sk);
release_sock(sk);
sock_put(sk);
}
sock->sk = NULL;
sock->state = SS_FREE;
return 0;
}
static int raw_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct raw_sock *ro;
if (!sk)
return 0;
ro = raw_sk(sk);
unregister_netdevice_notifier(&ro->notifier);
lock_sock(sk);
/* remove current filters & unregister */
if (ro->bound) {
if (ro->ifindex) {
struct net_device *dev;
dev = dev_get_by_index(&init_net, ro->ifindex);
if (dev) {
raw_disable_allfilters(dev, sk);
dev_put(dev);
}
} else
raw_disable_allfilters(NULL, sk);
}
if (ro->count > 1)
kfree(ro->filter);
ro->ifindex = 0;
ro->bound = 0;
ro->count = 0;
free_percpu(ro->uniq);
sock_orphan(sk);
sock->sk = NULL;
release_sock(sk);
sock_put(sk);
return 0;
}
static int sco_sock_release(struct socket *sock)
{
struct sock *sk = sock->sk;
int err = 0;
BT_DBG("sock %p, sk %p", sock, sk);
if (!sk)
return 0;
sco_sock_close(sk);
if (sk->linger) {
lock_sock(sk);
err = bluez_sock_wait_state(sk, BT_CLOSED, sk->lingertime);
release_sock(sk);
}
sock_orphan(sk);
sco_sock_kill(sk);
return err;
}
static int netlink_release(struct socket *sock)
{
struct sock *sk = sock->sk;
if (!sk)
return 0;
netlink_remove(sk);
spin_lock(&sk->protinfo.af_netlink->cb_lock);
if (sk->protinfo.af_netlink->cb) {
sk->protinfo.af_netlink->cb->done(sk->protinfo.af_netlink->cb);
netlink_destroy_callback(sk->protinfo.af_netlink->cb);
sk->protinfo.af_netlink->cb = NULL;
__sock_put(sk);
}
spin_unlock(&sk->protinfo.af_netlink->cb_lock);
/* OK. Socket is unlinked, and, therefore,
no new packets will arrive */
sock_orphan(sk);
sock->sk = NULL;
wake_up_interruptible_all(&sk->protinfo.af_netlink->wait);
skb_queue_purge(&sk->write_queue);
if (sk->protinfo.af_netlink->pid && !sk->protinfo.af_netlink->groups) {
struct netlink_notify n = { protocol:sk->protocol,
pid:sk->protinfo.af_netlink->pid };
notifier_call_chain(&netlink_chain, NETLINK_URELEASE, &n);
}
sock_put(sk);
return 0;
}
static int mptp_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct mptp_sock *ssk = mptp_sk(sk);
if (unlikely(!sk))
return 0;
mptp_unhash(ssk->src);
sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
synchronize_net();
sock_orphan(sk);
sock->sk = NULL;
skb_queue_purge(&sk->sk_receive_queue);
log_debug("mptp_release sock=%p\n", sk);
sock_put(sk);
return 0;
}
static int open_socket(cpt_object_t *obj, struct cpt_sock_image *si,
struct cpt_context *ctx)
{
int err;
struct socket *sock;
struct socket *sock2 = NULL;
struct file *file;
cpt_object_t *fobj;
cpt_object_t *pobj = NULL;
err = sock_create(si->cpt_family, si->cpt_type, si->cpt_protocol,
&sock);
if (err)
return err;
if (si->cpt_socketpair) {
err = sock_create(si->cpt_family, si->cpt_type,
si->cpt_protocol, &sock2);
if (err)
goto err_out;
err = sock->ops->socketpair(sock, sock2);
if (err < 0)
goto err_out;
/* Socketpair with a peer outside our environment.
* So, we create real half-open pipe and do not worry
* about dead end anymore. */
if (si->cpt_peer == -1) {
sock_release(sock2);
sock2 = NULL;
}
}
cpt_obj_setobj(obj, sock->sk, ctx);
if (si->cpt_file != CPT_NULL) {
file = sock_mapfile(sock);
err = PTR_ERR(file);
if (IS_ERR(file))
goto err_out;
err = -ENOMEM;
obj->o_parent = file;
if ((fobj = cpt_object_add(CPT_OBJ_FILE, file, ctx)) == NULL)
goto err_out;
cpt_obj_setpos(fobj, si->cpt_file, ctx);
cpt_obj_setindex(fobj, si->cpt_index, ctx);
}
if (sock2) {
struct file *file2;
pobj = lookup_cpt_obj_byindex(CPT_OBJ_SOCKET, si->cpt_peer, ctx);
if (!pobj) BUG();
if (pobj->o_obj) BUG();
cpt_obj_setobj(pobj, sock2->sk, ctx);
if (pobj->o_ppos != CPT_NULL) {
file2 = sock_mapfile(sock2);
err = PTR_ERR(file2);
if (IS_ERR(file2))
goto err_out;
err = -ENOMEM;
if ((fobj = cpt_object_add(CPT_OBJ_FILE, file2, ctx)) == NULL)
goto err_out;
cpt_obj_setpos(fobj, pobj->o_ppos, ctx);
cpt_obj_setindex(fobj, si->cpt_peer, ctx);
pobj->o_parent = file2;
}
}
setup_sock_common(sock->sk, si, obj->o_pos, ctx);
if (sock->sk->sk_family == AF_INET || sock->sk->sk_family == AF_INET6) {
int saved_reuse = sock->sk->sk_reuse;
inet_sk(sock->sk)->freebind = 1;
sock->sk->sk_reuse = 2;
if (si->cpt_laddrlen) {
err = sock->ops->bind(sock, (struct sockaddr *)&si->cpt_laddr, si->cpt_laddrlen);
if (err) {
dprintk_ctx("binding failed: %d, do not worry\n", err);
}
}
sock->sk->sk_reuse = saved_reuse;
rst_socket_in(si, obj->o_pos, sock->sk, ctx);
} else if (sock->sk->sk_family == AF_NETLINK) {
struct sockaddr_nl *nl = (struct sockaddr_nl *)&si->cpt_laddr;
if (nl->nl_pid) {
err = sock->ops->bind(sock, (struct sockaddr *)&si->cpt_laddr, si->cpt_laddrlen);
if (err) {
eprintk_ctx("AF_NETLINK binding failed: %d\n", err);
}
}
if (si->cpt_raddrlen && nl->nl_pid) {
err = sock->ops->connect(sock, (struct sockaddr *)&si->cpt_raddr, si->cpt_raddrlen, O_NONBLOCK);
if (err) {
eprintk_ctx("oops, AF_NETLINK connect failed: %d\n", err);
}
}
generic_restore_queues(sock->sk, si, obj->o_pos, ctx);
} else if (sock->sk->sk_family == PF_PACKET) {
struct sockaddr_ll *ll = (struct sockaddr_ll *)&si->cpt_laddr;
if (ll->sll_protocol || ll->sll_ifindex) {
int alen = si->cpt_laddrlen;
if (alen < sizeof(struct sockaddr_ll))
alen = sizeof(struct sockaddr_ll);
err = sock->ops->bind(sock, (struct sockaddr *)&si->cpt_laddr, alen);
if (err) {
eprintk_ctx("AF_PACKET binding failed: %d\n", err);
}
}
generic_restore_queues(sock->sk, si, obj->o_pos, ctx);
}
fixup_unix_address(sock, si, ctx);
if (sock2) {
err = rst_get_object(CPT_OBJ_SOCKET, pobj->o_pos, si, ctx);
if (err)
return err;
setup_sock_common(sock2->sk, si, pobj->o_pos, ctx);
fixup_unix_address(sock2, si, ctx);
}
if ((sock->sk->sk_family == AF_INET || sock->sk->sk_family == AF_INET6)
&& (int)si->cpt_parent != -1) {
cpt_object_t *lobj = lookup_cpt_obj_byindex(CPT_OBJ_SOCKET, si->cpt_parent, ctx);
if (lobj && cpt_attach_accept(lobj->o_obj, sock->sk, ctx) == 0)
sock->sk = NULL;
}
if (si->cpt_file == CPT_NULL && sock->sk &&
sock->sk->sk_family == AF_INET) {
struct sock *sk = sock->sk;
if (sk) {
sock->sk = NULL;
local_bh_disable();
bh_lock_sock(sk);
if (sock_owned_by_user(sk))
eprintk_ctx("oops, sock is locked by user\n");
sock_hold(sk);
sock_orphan(sk);
ub_inc_orphan_count(sk);
bh_unlock_sock(sk);
local_bh_enable();
sock_put(sk);
dprintk_ctx("orphaning socket %p\n", sk);
}
}
if (si->cpt_file == CPT_NULL && sock->sk == NULL)
sock_release(sock);
return 0;
err_out:
if (sock2)
sock_release(sock2);
sock_release(sock);
return err;
}
/**
* This is main body of the socket close function in Sync Sockets.
*
* inet_release() can sleep (as well as tcp_close()), so we make our own
* non-sleepable socket closing.
*
* This function must be used only for data sockets.
* Use standard sock_release() for listening sockets.
*
* In most cases it is called in softirq context and from ksoftirqd which
* processes data from the socket (RSS and RPS distribute packets that way).
*
* Note: it used to be called in process context as well, at the time when
* Tempesta starts or stops. That's not the case right now, but it may change.
*
* TODO In some cases we need to close socket agresively w/o FIN_WAIT_2 state,
* e.g. by sending RST. So we need to add second parameter to the function
* which says how to close the socket.
* One of the examples is rcl_req_limit() (it should reset connections).
* See tcp_sk(sk)->linger2 processing in standard tcp_close().
*
* Called with locked socket.
*/
static void
ss_do_close(struct sock *sk)
{
struct sk_buff *skb;
int data_was_unread = 0;
int state;
if (unlikely(!sk))
return;
SS_DBG("Close socket %p (%s): cpu=%d account=%d refcnt=%d\n",
sk, ss_statename[sk->sk_state], smp_processor_id(),
sk_has_account(sk), atomic_read(&sk->sk_refcnt));
assert_spin_locked(&sk->sk_lock.slock);
ss_sock_cpu_check(sk);
BUG_ON(sk->sk_state == TCP_LISTEN);
/* We must return immediately, so LINGER option is meaningless. */
WARN_ON(sock_flag(sk, SOCK_LINGER));
/* We don't support virtual containers, so TCP_REPAIR is prohibited. */
WARN_ON(tcp_sk(sk)->repair);
/* The socket must have atomic allocation mask. */
WARN_ON(!(sk->sk_allocation & GFP_ATOMIC));
/* The below is mostly copy-paste from tcp_close(). */
sk->sk_shutdown = SHUTDOWN_MASK;
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
u32 len = TCP_SKB_CB(skb)->end_seq - TCP_SKB_CB(skb)->seq -
tcp_hdr(skb)->fin;
data_was_unread += len;
SS_DBG("free rcv skb %p\n", skb);
__kfree_skb(skb);
}
sk_mem_reclaim(sk);
if (sk->sk_state == TCP_CLOSE)
goto adjudge_to_death;
if (data_was_unread) {
NET_INC_STATS_USER(sock_net(sk), LINUX_MIB_TCPABORTONCLOSE);
tcp_set_state(sk, TCP_CLOSE);
tcp_send_active_reset(sk, sk->sk_allocation);
}
else if (tcp_close_state(sk)) {
/* The code below is taken from tcp_send_fin(). */
struct tcp_sock *tp = tcp_sk(sk);
int mss_now = tcp_current_mss(sk);
skb = tcp_write_queue_tail(sk);
if (tcp_send_head(sk) != NULL) {
/* Send FIN with data if we have any. */
TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_FIN;
TCP_SKB_CB(skb)->end_seq++;
tp->write_seq++;
}
else {
/* No data to send in the socket, allocate new skb. */
skb = alloc_skb_fclone(MAX_TCP_HEADER,
sk->sk_allocation);
if (!skb) {
SS_WARN("can't send FIN due to bad alloc");
} else {
skb_reserve(skb, MAX_TCP_HEADER);
tcp_init_nondata_skb(skb, tp->write_seq,
TCPHDR_ACK | TCPHDR_FIN);
tcp_queue_skb(sk, skb);
}
}
__tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_OFF);
}
adjudge_to_death:
state = sk->sk_state;
sock_hold(sk);
sock_orphan(sk);
/*
* SS sockets are processed in softirq only,
* so backlog queue should be empty.
*/
WARN_ON(sk->sk_backlog.tail);
percpu_counter_inc(sk->sk_prot->orphan_count);
if (state != TCP_CLOSE && sk->sk_state == TCP_CLOSE)
return;
if (sk->sk_state == TCP_FIN_WAIT2) {
const int tmo = tcp_fin_time(sk);
if (tmo > TCP_TIMEWAIT_LEN) {
inet_csk_reset_keepalive_timer(sk,
tmo - TCP_TIMEWAIT_LEN);
} else {
tcp_time_wait(sk, TCP_FIN_WAIT2, tmo);
return;
}
}
if (sk->sk_state != TCP_CLOSE) {
sk_mem_reclaim(sk);
if (tcp_check_oom(sk, 0)) {
tcp_set_state(sk, TCP_CLOSE);
tcp_send_active_reset(sk, GFP_ATOMIC);
NET_INC_STATS_BH(sock_net(sk),
LINUX_MIB_TCPABORTONMEMORY);
}
}
if (sk->sk_state == TCP_CLOSE) {
struct request_sock *req = tcp_sk(sk)->fastopen_rsk;
if (req != NULL)
reqsk_fastopen_remove(sk, req, false);
inet_csk_destroy_sock(sk);
}
}
void dccp_close(struct sock *sk, long timeout)
{
struct dccp_sock *dp = dccp_sk(sk);
struct sk_buff *skb;
u32 data_was_unread = 0;
int state;
lock_sock(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
if (sk->sk_state == DCCP_LISTEN) {
dccp_set_state(sk, DCCP_CLOSED);
/* Special case. */
inet_csk_listen_stop(sk);
goto adjudge_to_death;
}
sk_stop_timer(sk, &dp->dccps_xmit_timer);
/*
* We need to flush the recv. buffs. We do this only on the
* descriptor close, not protocol-sourced closes, because the
*reader process may not have drained the data yet!
*/
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
data_was_unread += skb->len;
__kfree_skb(skb);
}
if (data_was_unread) {
/* Unread data was tossed, send an appropriate Reset Code */
DCCP_WARN("ABORT with %u bytes unread\n", data_was_unread);
dccp_send_reset(sk, DCCP_RESET_CODE_ABORTED);
dccp_set_state(sk, DCCP_CLOSED);
} else if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
/* Check zero linger _after_ checking for unread data. */
sk->sk_prot->disconnect(sk, 0);
} else if (sk->sk_state != DCCP_CLOSED) {
/*
* Normal connection termination. May need to wait if there are
* still packets in the TX queue that are delayed by the CCID.
*/
dccp_flush_write_queue(sk, &timeout);
dccp_terminate_connection(sk);
}
/*
* Flush write queue. This may be necessary in several cases:
* - we have been closed by the peer but still have application data;
* - abortive termination (unread data or zero linger time),
* - normal termination but queue could not be flushed within time limit
*/
__skb_queue_purge(&sk->sk_write_queue);
sk_stream_wait_close(sk, timeout);
adjudge_to_death:
state = sk->sk_state;
sock_hold(sk);
sock_orphan(sk);
/*
* It is the last release_sock in its life. It will remove backlog.
*/
release_sock(sk);
/*
* Now socket is owned by kernel and we acquire BH lock
* to finish close. No need to check for user refs.
*/
local_bh_disable();
bh_lock_sock(sk);
WARN_ON(sock_owned_by_user(sk));
percpu_counter_inc(sk->sk_prot->orphan_count);
/* Have we already been destroyed by a softirq or backlog? */
if (state != DCCP_CLOSED && sk->sk_state == DCCP_CLOSED)
goto out;
if (sk->sk_state == DCCP_CLOSED)
inet_csk_destroy_sock(sk);
/* Otherwise, socket is reprieved until protocol close. */
out:
bh_unlock_sock(sk);
local_bh_enable();
sock_put(sk);
}
void dccp_close(struct sock *sk, long timeout)
{
struct sk_buff *skb;
lock_sock(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
if (sk->sk_state == DCCP_LISTEN) {
dccp_set_state(sk, DCCP_CLOSED);
/* Special case. */
inet_csk_listen_stop(sk);
goto adjudge_to_death;
}
/*
* We need to flush the recv. buffs. We do this only on the
* descriptor close, not protocol-sourced closes, because the
*reader process may not have drained the data yet!
*/
/* FIXME: check for unread data */
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
__kfree_skb(skb);
}
if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
/* Check zero linger _after_ checking for unread data. */
sk->sk_prot->disconnect(sk, 0);
} else if (dccp_close_state(sk)) {
dccp_send_close(sk, 1);
}
sk_stream_wait_close(sk, timeout);
adjudge_to_death:
/*
* It is the last release_sock in its life. It will remove backlog.
*/
release_sock(sk);
/*
* Now socket is owned by kernel and we acquire BH lock
* to finish close. No need to check for user refs.
*/
local_bh_disable();
bh_lock_sock(sk);
BUG_TRAP(!sock_owned_by_user(sk));
sock_hold(sk);
sock_orphan(sk);
/*
* The last release_sock may have processed the CLOSE or RESET
* packet moving sock to CLOSED state, if not we have to fire
* the CLOSE/CLOSEREQ retransmission timer, see "8.3. Termination"
* in draft-ietf-dccp-spec-11. -acme
*/
if (sk->sk_state == DCCP_CLOSING) {
/* FIXME: should start at 2 * RTT */
/* Timer for repeating the CLOSE/CLOSEREQ until an answer. */
inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS,
inet_csk(sk)->icsk_rto,
DCCP_RTO_MAX);
#if 0
/* Yeah, we should use sk->sk_prot->orphan_count, etc */
dccp_set_state(sk, DCCP_CLOSED);
#endif
}
atomic_inc(sk->sk_prot->orphan_count);
if (sk->sk_state == DCCP_CLOSED)
inet_csk_destroy_sock(sk);
/* Otherwise, socket is reprieved until protocol close. */
bh_unlock_sock(sk);
local_bh_enable();
sock_put(sk);
}
static int netlink_release(struct socket *sock)
{
struct sock *sk = sock->sk;
struct netlink_opt *nlk;
if (!sk)
return 0;
netlink_remove(sk);
nlk = nlk_sk(sk);
spin_lock(&nlk->cb_lock);
if (nlk->cb) {
nlk->cb->done(nlk->cb);
netlink_destroy_callback(nlk->cb);
nlk->cb = NULL;
__sock_put(sk);
}
spin_unlock(&nlk->cb_lock);
/* OK. Socket is unlinked, and, therefore,
no new packets will arrive */
sock_orphan(sk);
sock->sk = NULL;
wake_up_interruptible_all(&nlk->wait);
skb_queue_purge(&sk->sk_write_queue);
if (nlk->pid && !nlk->groups) {
struct netlink_notify n = {
.protocol = sk->sk_protocol,
.pid = nlk->pid,
};
notifier_call_chain(&netlink_chain, NETLINK_URELEASE, &n);
}
sock_put(sk);
return 0;
}
static int netlink_autobind(struct socket *sock)
{
struct sock *sk = sock->sk;
struct nl_pid_hash *hash = &nl_table[sk->sk_protocol].hash;
struct hlist_head *head;
struct sock *osk;
struct hlist_node *node;
s32 pid = current->pid;
int err;
static s32 rover = -4097;
retry:
cond_resched();
netlink_table_grab();
head = nl_pid_hashfn(hash, pid);
sk_for_each(osk, node, head) {
if (nlk_sk(osk)->pid == pid) {
/* Bind collision, search negative pid values. */
pid = rover--;
if (rover > -4097)
rover = -4097;
netlink_table_ungrab();
goto retry;
}
}
netlink_table_ungrab();
err = netlink_insert(sk, pid);
if (err == -EADDRINUSE)
goto retry;
nlk_sk(sk)->groups = 0;
return 0;
}
static inline int netlink_capable(struct socket *sock, unsigned int flag)
{
return (nl_nonroot[sock->sk->sk_protocol] & flag) ||
capable(CAP_NET_ADMIN);
}
static int netlink_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
{
struct sock *sk = sock->sk;
struct netlink_opt *nlk = nlk_sk(sk);
struct sockaddr_nl *nladdr = (struct sockaddr_nl *)addr;
int err;
if (nladdr->nl_family != AF_NETLINK)
return -EINVAL;
/* Only superuser is allowed to listen multicasts */
if (nladdr->nl_groups && !netlink_capable(sock, NL_NONROOT_RECV))
return -EPERM;
if (nlk->pid) {
if (nladdr->nl_pid != nlk->pid)
return -EINVAL;
} else {
err = nladdr->nl_pid ?
netlink_insert(sk, nladdr->nl_pid) :
netlink_autobind(sock);
if (err)
return err;
}
if (!nladdr->nl_groups && !nlk->groups)
return 0;
netlink_table_grab();
if (nlk->groups && !nladdr->nl_groups)
__sk_del_bind_node(sk);
else if (!nlk->groups && nladdr->nl_groups)
sk_add_bind_node(sk, &nl_table[sk->sk_protocol].mc_list);
nlk->groups = nladdr->nl_groups;
netlink_table_ungrab();
return 0;
}
static int netlink_connect(struct socket *sock, struct sockaddr *addr,
int alen, int flags)
{
int err = 0;
struct sock *sk = sock->sk;
struct netlink_opt *nlk = nlk_sk(sk);
struct sockaddr_nl *nladdr=(struct sockaddr_nl*)addr;
if (addr->sa_family == AF_UNSPEC) {
sk->sk_state = NETLINK_UNCONNECTED;
nlk->dst_pid = 0;
nlk->dst_groups = 0;
return 0;
}
if (addr->sa_family != AF_NETLINK)
return -EINVAL;
/* Only superuser is allowed to send multicasts */
if (nladdr->nl_groups && !netlink_capable(sock, NL_NONROOT_SEND))
return -EPERM;
if (!nlk->pid)
err = netlink_autobind(sock);
if (err == 0) {
sk->sk_state = NETLINK_CONNECTED;
nlk->dst_pid = nladdr->nl_pid;
nlk->dst_groups = nladdr->nl_groups;
}
return err;
}
void dccp_close(struct sock *sk, long timeout)
{
struct dccp_sock *dp = dccp_sk(sk);
struct sk_buff *skb;
u32 data_was_unread = 0;
int state;
lock_sock(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
if (sk->sk_state == DCCP_LISTEN) {
dccp_set_state(sk, DCCP_CLOSED);
inet_csk_listen_stop(sk);
goto adjudge_to_death;
}
sk_stop_timer(sk, &dp->dccps_xmit_timer);
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
data_was_unread += skb->len;
__kfree_skb(skb);
}
if (data_was_unread) {
DCCP_WARN("DCCP: ABORT -- %u bytes unread\n", data_was_unread);
dccp_send_reset(sk, DCCP_RESET_CODE_ABORTED);
dccp_set_state(sk, DCCP_CLOSED);
} else if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
sk->sk_prot->disconnect(sk, 0);
} else if (sk->sk_state != DCCP_CLOSED) {
dccp_terminate_connection(sk);
}
sk_stream_wait_close(sk, timeout);
adjudge_to_death:
state = sk->sk_state;
sock_hold(sk);
sock_orphan(sk);
release_sock(sk);
local_bh_disable();
bh_lock_sock(sk);
WARN_ON(sock_owned_by_user(sk));
percpu_counter_inc(sk->sk_prot->orphan_count);
if (state != DCCP_CLOSED && sk->sk_state == DCCP_CLOSED)
goto out;
if (sk->sk_state == DCCP_CLOSED)
inet_csk_destroy_sock(sk);
out:
bh_unlock_sock(sk);
local_bh_enable();
sock_put(sk);
}
void dccp_close(struct sock *sk, long timeout)
{
struct dccp_sock *dp = dccp_sk(sk);
struct sk_buff *skb;
u32 data_was_unread = 0;
int state;
lock_sock(sk);
sk->sk_shutdown = SHUTDOWN_MASK;
if (sk->sk_state == DCCP_LISTEN) {
dccp_set_state(sk, DCCP_CLOSED);
/* Special case. */
inet_csk_listen_stop(sk);
goto adjudge_to_death;
}
sk_stop_timer(sk, &dp->dccps_xmit_timer);
/*
* We need to flush the recv. buffs. We do this only on the
* descriptor close, not protocol-sourced closes, because the
*reader process may not have drained the data yet!
*/
while ((skb = __skb_dequeue(&sk->sk_receive_queue)) != NULL) {
data_was_unread += skb->len;
__kfree_skb(skb);
}
if (data_was_unread) {
/* Unread data was tossed, send an appropriate Reset Code */
DCCP_WARN("DCCP: ABORT -- %u bytes unread\n", data_was_unread);
dccp_send_reset(sk, DCCP_RESET_CODE_ABORTED);
dccp_set_state(sk, DCCP_CLOSED);
} else if (sock_flag(sk, SOCK_LINGER) && !sk->sk_lingertime) {
/* Check zero linger _after_ checking for unread data. */
sk->sk_prot->disconnect(sk, 0);
} else if (sk->sk_state != DCCP_CLOSED) {
dccp_terminate_connection(sk);
}
sk_stream_wait_close(sk, timeout);
adjudge_to_death:
state = sk->sk_state;
sock_hold(sk);
sock_orphan(sk);
atomic_inc(sk->sk_prot->orphan_count);
/*
* It is the last release_sock in its life. It will remove backlog.
*/
release_sock(sk);
/*
* Now socket is owned by kernel and we acquire BH lock
* to finish close. No need to check for user refs.
*/
local_bh_disable();
bh_lock_sock(sk);
BUG_TRAP(!sock_owned_by_user(sk));
/* Have we already been destroyed by a softirq or backlog? */
if (state != DCCP_CLOSED && sk->sk_state == DCCP_CLOSED)
goto out;
if (sk->sk_state == DCCP_CLOSED)
inet_csk_destroy_sock(sk);
/* Otherwise, socket is reprieved until protocol close. */
out:
bh_unlock_sock(sk);
local_bh_enable();
sock_put(sk);
}
