int psock_tcp_accept(FAR struct socket *psock, FAR struct sockaddr *addr,
FAR socklen_t *addrlen, FAR void **newconn)
{
FAR struct tcp_conn_s *conn;
struct accept_s state;
int ret;
DEBUGASSERT(psock && newconn);
/* Check the backlog to see if there is a connection already pending for
* this listener.
*/
conn = (FAR struct tcp_conn_s *)psock->s_conn;
#ifdef CONFIG_NET_TCPBACKLOG
state.acpt_newconn = tcp_backlogremove(conn);
if (state.acpt_newconn)
{
/* Yes... get the address of the connected client */
ninfo("Pending conn=%p\n", state.acpt_newconn);
accept_tcpsender(psock, state.acpt_newconn, addr, addrlen);
}
/* In general, this implementation will not support non-blocking socket
* operations... except in a few cases: Here for TCP accept with
* backlog enabled. If this socket is configured as non-blocking then
* return EAGAIN if there is no pending connection in the backlog.
*/
else if (_SS_ISNONBLOCK(psock->s_flags))
{
return -EAGAIN;
}
else
#endif
{
/* Set the socket state to accepting */
psock->s_flags = _SS_SETSTATE(psock->s_flags, _SF_ACCEPT);
/* Perform the TCP accept operation */
/* Initialize the state structure. This is done with the network
* locked because we don't want anything to happen until we are
* ready.
*/
state.acpt_sock = psock;
state.acpt_addr = addr;
state.acpt_addrlen = addrlen;
state.acpt_newconn = NULL;
state.acpt_result = OK;
/* This semaphore is used for signaling and, hence, should not have
* priority inheritance enabled.
*/
nxsem_init(&state.acpt_sem, 0, 0);
nxsem_setprotocol(&state.acpt_sem, SEM_PRIO_NONE);
/* Set up the callback in the connection */
conn->accept_private = (FAR void *)&state;
conn->accept = accept_eventhandler;
/* Wait for the send to complete or an error to occur: NOTES:
* net_lockedwait will also terminate if a signal is received.
*/
ret = net_lockedwait(&state.acpt_sem);
/* Make sure that no further events are processed */
conn->accept_private = NULL;
conn->accept = NULL;
nxsem_destroy(&state. acpt_sem);
/* Set the socket state to idle */
psock->s_flags = _SS_SETSTATE(psock->s_flags, _SF_IDLE);
/* Check for a errors. Errors are signalled by negative errno values
* for the send length.
*/
if (state.acpt_result != 0)
{
DEBUGASSERT(state.acpt_result > 0);
return -state.acpt_result;
}
/* If net_lockedwait failed, then we were probably reawakened by a
* signal. In this case, net_lockedwait will have returned negated
* errno appropriately.
*/
if (ret < 0)
{
return ret;
}
}
*newconn = (FAR void *)state.acpt_newconn;
return OK;
}
int psock_tcp_accept(FAR struct socket *psock, FAR struct sockaddr *addr,
FAR socklen_t *addrlen, FAR void **newconn)
{
FAR struct tcp_conn_s *conn;
struct accept_s state;
int ret;
DEBUGASSERT(psock && newconn);
/* Check the backlog to see if there is a connection already pending for
* this listener.
*/
conn = (FAR struct tcp_conn_s *)psock->s_conn;
#ifdef CONFIG_NET_TCPBACKLOG
state.acpt_newconn = tcp_backlogremove(conn);
if (state.acpt_newconn)
{
/* Yes... get the address of the connected client */
nvdbg("Pending conn=%p\n", state.acpt_newconn);
accept_tcpsender(psock, state.acpt_newconn, addr, addrlen);
}
/* In general, this uIP-based implementation will not support non-blocking
* socket operations... except in a few cases: Here for TCP accept with
* backlog enabled. If this socket is configured as non-blocking then
* return EAGAIN if there is no pending connection in the backlog.
*/
else if (_SS_ISNONBLOCK(psock->s_flags))
{
return -EAGAIN;
}
else
#endif
{
/* Set the socket state to accepting */
psock->s_flags = _SS_SETSTATE(psock->s_flags, _SF_ACCEPT);
/* Perform the TCP accept operation */
/* Initialize the state structure. This is done with interrupts
* disabled because we don't want anything to happen until we
* are ready.
*/
state.acpt_sock = psock;
state.acpt_addr = addr;
state.acpt_addrlen = addrlen;
state.acpt_newconn = NULL;
state.acpt_result = OK;
sem_init(&state.acpt_sem, 0, 0);
/* Set up the callback in the connection */
conn->accept_private = (FAR void *)&state;
conn->accept = accept_interrupt;
/* Wait for the send to complete or an error to occur: NOTES: (1)
* net_lockedwait will also terminate if a signal is received, (2)
* interrupts may be disabled! They will be re-enabled while the
* task sleeps and automatically re-enabled when the task restarts.
*/
ret = net_lockedwait(&state.acpt_sem);
if (ret < 0)
{
/* The value returned by net_lockedwait() the same as the value
* returned by sem_wait(): Zero (OK) is returned on success; -1
* (ERROR) is returned on a failure with the errno value set
* appropriately.
*
* We have to preserve the errno value here because it may be
* altered by intervening operations.
*/
ret = -get_errno();
DEBUGASSERT(ret < 0);
}
/* Make sure that no further interrupts are processed */
conn->accept_private = NULL;
conn->accept = NULL;
sem_destroy(&state. acpt_sem);
/* Set the socket state to idle */
psock->s_flags = _SS_SETSTATE(psock->s_flags, _SF_IDLE);
/* Check for a errors. Errors are signalled by negative errno values
* for the send length.
*/
if (state.acpt_result != 0)
{
DEBUGASSERT(state.acpt_result > 0);
return -state.acpt_result;
}
/* If net_lockedwait failed, then we were probably reawakened by a
* signal. In this case, logic above will have set 'ret' to the
* errno value returned by net_lockedwait().
*/
if (ret < 0)
{
return ret;
}
}
*newconn = (FAR void *)state.acpt_newconn;
return OK;
}
