static inline void xs_tcp_read_fraghdr(struct rpc_xprt *xprt, skb_reader_t *desc)
{
size_t len, used;
char *p;
p = ((char *) &xprt->tcp_recm) + xprt->tcp_offset;
len = sizeof(xprt->tcp_recm) - xprt->tcp_offset;
used = xs_tcp_copy_data(desc, p, len);
xprt->tcp_offset += used;
if (used != len)
return;
xprt->tcp_reclen = ntohl(xprt->tcp_recm);
if (xprt->tcp_reclen & RPC_LAST_STREAM_FRAGMENT)
xprt->tcp_flags |= XPRT_LAST_FRAG;
else
xprt->tcp_flags &= ~XPRT_LAST_FRAG;
xprt->tcp_reclen &= RPC_FRAGMENT_SIZE_MASK;
xprt->tcp_flags &= ~XPRT_COPY_RECM;
xprt->tcp_offset = 0;
/* Sanity check of the record length */
if (unlikely(xprt->tcp_reclen < 4)) {
dprintk("RPC: invalid TCP record fragment length\n");
xprt_disconnect(xprt);
return;
}
dprintk("RPC: reading TCP record fragment of length %d\n",
xprt->tcp_reclen);
}
/**
* xs_tcp_send_request - write an RPC request to a TCP socket
* @task: address of RPC task that manages the state of an RPC request
*
* Return values:
* 0: The request has been sent
* EAGAIN: The socket was blocked, please call again later to
* complete the request
* ENOTCONN: Caller needs to invoke connect logic then call again
* other: Some other error occured, the request was not sent
*
* XXX: In the case of soft timeouts, should we eventually give up
* if sendmsg is not able to make progress?
*/
static int xs_tcp_send_request(struct rpc_task *task)
{
struct rpc_rqst *req = task->tk_rqstp;
struct rpc_xprt *xprt = req->rq_xprt;
struct xdr_buf *xdr = &req->rq_snd_buf;
int status, retry = 0;
xs_encode_tcp_record_marker(&req->rq_snd_buf);
xs_pktdump("packet data:",
req->rq_svec->iov_base,
req->rq_svec->iov_len);
/* Continue transmitting the packet/record. We must be careful
* to cope with writespace callbacks arriving _after_ we have
* called sendmsg(). */
while (1) {
req->rq_xtime = jiffies;
status = xs_sendpages(xprt->sock, NULL, 0, xdr,
req->rq_bytes_sent);
dprintk("RPC: xs_tcp_send_request(%u) = %d\n",
xdr->len - req->rq_bytes_sent, status);
if (unlikely(status < 0))
break;
/* If we've sent the entire packet, immediately
* reset the count of bytes sent. */
req->rq_bytes_sent += status;
task->tk_bytes_sent += status;
if (likely(req->rq_bytes_sent >= req->rq_slen)) {
req->rq_bytes_sent = 0;
return 0;
}
status = -EAGAIN;
if (retry++ > XS_SENDMSG_RETRY)
break;
}
switch (status) {
case -EAGAIN:
xs_nospace(task);
break;
case -ECONNREFUSED:
case -ECONNRESET:
case -ENOTCONN:
case -EPIPE:
status = -ENOTCONN;
break;
default:
dprintk("RPC: sendmsg returned unrecognized error %d\n",
-status);
xprt_disconnect(xprt);
break;
}
return status;
}
/**
* xs_destroy - prepare to shutdown a transport
* @xprt: doomed transport
*
*/
static void xs_destroy(struct rpc_xprt *xprt)
{
dprintk("RPC: xs_destroy xprt %p\n", xprt);
cancel_delayed_work(&xprt->connect_worker);
flush_scheduled_work();
xprt_disconnect(xprt);
xs_close(xprt);
kfree(xprt->slot);
}
/**
* xs_destroy - prepare to shutdown a transport
* @xprt: doomed transport
*
*/
static void xs_destroy(struct rpc_xprt *xprt)
{
dprintk("RPC: xs_destroy xprt %p\n", xprt);
if (!cancel_delayed_work(&xprt->connect_worker))
wait_on_bit(&xprt->state, XPRT_CONNECTING,
xs_wait_bit_uninterruptible, TASK_UNINTERRUPTIBLE);
xprt_disconnect(xprt);
xs_close(xprt);
kfree(xprt->slot);
}
static int xs_tcp_data_recv(read_descriptor_t *rd_desc, struct sk_buff *skb, unsigned int offset, size_t len)
{
struct rpc_xprt *xprt = rd_desc->arg.data;
skb_reader_t desc = {
.skb = skb,
.offset = offset,
.count = len,
.csum = 0
};
dprintk("RPC: xs_tcp_data_recv started\n");
do {
/* Read in a new fragment marker if necessary */
/* Can we ever really expect to get completely empty fragments? */
if (xprt->tcp_flags & XPRT_COPY_RECM) {
xs_tcp_read_fraghdr(xprt, &desc);
continue;
}
/* Read in the xid if necessary */
if (xprt->tcp_flags & XPRT_COPY_XID) {
xs_tcp_read_xid(xprt, &desc);
continue;
}
/* Read in the request data */
if (xprt->tcp_flags & XPRT_COPY_DATA) {
xs_tcp_read_request(xprt, &desc);
continue;
}
/* Skip over any trailing bytes on short reads */
xs_tcp_read_discard(xprt, &desc);
} while (desc.count);
dprintk("RPC: xs_tcp_data_recv done\n");
return len - desc.count;
}
/**
* xs_tcp_data_ready - "data ready" callback for TCP sockets
* @sk: socket with data to read
* @bytes: how much data to read
*
*/
static void xs_tcp_data_ready(struct sock *sk, int bytes)
{
struct rpc_xprt *xprt;
read_descriptor_t rd_desc;
read_lock(&sk->sk_callback_lock);
dprintk("RPC: xs_tcp_data_ready...\n");
if (!(xprt = xprt_from_sock(sk)))
goto out;
if (xprt->shutdown)
goto out;
/* We use rd_desc to pass struct xprt to xs_tcp_data_recv */
rd_desc.arg.data = xprt;
rd_desc.count = 65536;
tcp_read_sock(sk, &rd_desc, xs_tcp_data_recv);
out:
read_unlock(&sk->sk_callback_lock);
}
/**
* xs_tcp_state_change - callback to handle TCP socket state changes
* @sk: socket whose state has changed
*
*/
static void xs_tcp_state_change(struct sock *sk)
{
struct rpc_xprt *xprt;
read_lock(&sk->sk_callback_lock);
if (!(xprt = xprt_from_sock(sk)))
goto out;
dprintk("RPC: xs_tcp_state_change client %p...\n", xprt);
dprintk("RPC: state %x conn %d dead %d zapped %d\n",
sk->sk_state, xprt_connected(xprt),
sock_flag(sk, SOCK_DEAD),
sock_flag(sk, SOCK_ZAPPED));
switch (sk->sk_state) {
case TCP_ESTABLISHED:
spin_lock_bh(&xprt->transport_lock);
if (!xprt_test_and_set_connected(xprt)) {
/* Reset TCP record info */
xprt->tcp_offset = 0;
xprt->tcp_reclen = 0;
xprt->tcp_copied = 0;
if (xprt->tcp_flags & XPRT_SRCADDR_PRESENT)
xprt->tcp_flags = XPRT_SRCADDR_PRESENT |
XPRT_COPY_RECM |
XPRT_COPY_XID;
else
xprt->tcp_flags = XPRT_COPY_RECM |
XPRT_COPY_XID;
xprt->reestablish_timeout = XS_TCP_INIT_REEST_TO;
xprt_wake_pending_tasks(xprt, 0);
}
spin_unlock_bh(&xprt->transport_lock);
break;
case TCP_SYN_SENT:
case TCP_SYN_RECV:
break;
case TCP_CLOSE_WAIT:
/* Try to schedule an autoclose RPC calls */
set_bit(XPRT_CLOSE_WAIT, &xprt->state);
if (test_and_set_bit(XPRT_LOCKED, &xprt->state) == 0)
queue_work(rpciod_workqueue, &xprt->task_cleanup);
default:
xprt_disconnect(xprt);
}
out:
read_unlock(&sk->sk_callback_lock);
}
/**
* xs_udp_write_space - callback invoked when socket buffer space
* becomes available
* @sk: socket whose state has changed
*
* Called when more output buffer space is available for this socket.
* We try not to wake our writers until they can make "significant"
* progress, otherwise we'll waste resources thrashing kernel_sendmsg
* with a bunch of small requests.
*/
static void xs_udp_write_space(struct sock *sk)
{
read_lock(&sk->sk_callback_lock);
/* from net/core/sock.c:sock_def_write_space */
if (sock_writeable(sk)) {
struct socket *sock;
struct rpc_xprt *xprt;
if (unlikely(!(sock = sk->sk_socket)))
goto out;
if (unlikely(!(xprt = xprt_from_sock(sk))))
goto out;
if (unlikely(!test_and_clear_bit(SOCK_NOSPACE, &sock->flags)))
goto out;
xprt_write_space(xprt);
}
out:
read_unlock(&sk->sk_callback_lock);
}
