static int
__smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
struct smb_rqst *rqst)
{
int rc = 0;
struct kvec *iov;
int n_vec;
unsigned int send_length = 0;
unsigned int i, j;
sigset_t mask, oldmask;
size_t total_len = 0, sent, size;
struct socket *ssocket = server->ssocket;
struct msghdr smb_msg;
int val = 1;
__be32 rfc1002_marker;
if (cifs_rdma_enabled(server) && server->smbd_conn) {
rc = smbd_send(server, num_rqst, rqst);
goto smbd_done;
}
if (ssocket == NULL)
return -EAGAIN;
if (signal_pending(current)) {
cifs_dbg(FYI, "signal is pending before sending any data\n");
return -EINTR;
}
/* cork the socket */
kernel_setsockopt(ssocket, SOL_TCP, TCP_CORK,
(char *)&val, sizeof(val));
for (j = 0; j < num_rqst; j++)
send_length += smb_rqst_len(server, &rqst[j]);
rfc1002_marker = cpu_to_be32(send_length);
/*
* We should not allow signals to interrupt the network send because
* any partial send will cause session reconnects thus increasing
* latency of system calls and overload a server with unnecessary
* requests.
*/
sigfillset(&mask);
sigprocmask(SIG_BLOCK, &mask, &oldmask);
/* Generate a rfc1002 marker for SMB2+ */
if (server->vals->header_preamble_size == 0) {
struct kvec hiov = {
.iov_base = &rfc1002_marker,
.iov_len = 4
};
iov_iter_kvec(&smb_msg.msg_iter, WRITE, &hiov, 1, 4);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
goto unmask;
total_len += sent;
send_length += 4;
}
cifs_dbg(FYI, "Sending smb: smb_len=%u\n", send_length);
for (j = 0; j < num_rqst; j++) {
iov = rqst[j].rq_iov;
n_vec = rqst[j].rq_nvec;
size = 0;
for (i = 0; i < n_vec; i++) {
dump_smb(iov[i].iov_base, iov[i].iov_len);
size += iov[i].iov_len;
}
iov_iter_kvec(&smb_msg.msg_iter, WRITE, iov, n_vec, size);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
goto unmask;
total_len += sent;
/* now walk the page array and send each page in it */
for (i = 0; i < rqst[j].rq_npages; i++) {
struct bio_vec bvec;
bvec.bv_page = rqst[j].rq_pages[i];
rqst_page_get_length(&rqst[j], i, &bvec.bv_len,
&bvec.bv_offset);
iov_iter_bvec(&smb_msg.msg_iter, WRITE,
&bvec, 1, bvec.bv_len);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
break;
total_len += sent;
}
}
unmask:
sigprocmask(SIG_SETMASK, &oldmask, NULL);
/*
* If signal is pending but we have already sent the whole packet to
* the server we need to return success status to allow a corresponding
* mid entry to be kept in the pending requests queue thus allowing
* to handle responses from the server by the client.
*
* If only part of the packet has been sent there is no need to hide
* interrupt because the session will be reconnected anyway, so there
* won't be any response from the server to handle.
*/
if (signal_pending(current) && (total_len != send_length)) {
cifs_dbg(FYI, "signal is pending after attempt to send\n");
rc = -EINTR;
}
/* uncork it */
val = 0;
kernel_setsockopt(ssocket, SOL_TCP, TCP_CORK,
(char *)&val, sizeof(val));
if ((total_len > 0) && (total_len != send_length)) {
cifs_dbg(FYI, "partial send (wanted=%u sent=%zu): terminating session\n",
send_length, total_len);
/*
* If we have only sent part of an SMB then the next SMB could
* be taken as the remainder of this one. We need to kill the
* socket so the server throws away the partial SMB
*/
server->tcpStatus = CifsNeedReconnect;
trace_smb3_partial_send_reconnect(server->CurrentMid,
server->hostname);
}
smbd_done:
if (rc < 0 && rc != -EINTR)
cifs_dbg(VFS, "Error %d sending data on socket to server\n",
rc);
else if (rc > 0)
rc = 0;
return rc;
}
static int
smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
struct smb_rqst *rqst, int flags)
{
struct kvec iov;
struct smb2_transform_hdr tr_hdr;
struct smb_rqst cur_rqst[MAX_COMPOUND];
int rc;
if (!(flags & CIFS_TRANSFORM_REQ))
return __smb_send_rqst(server, num_rqst, rqst);
if (num_rqst > MAX_COMPOUND - 1)
return -ENOMEM;
memset(&cur_rqst[0], 0, sizeof(cur_rqst));
memset(&iov, 0, sizeof(iov));
memset(&tr_hdr, 0, sizeof(tr_hdr));
iov.iov_base = &tr_hdr;
iov.iov_len = sizeof(tr_hdr);
cur_rqst[0].rq_iov = &iov;
cur_rqst[0].rq_nvec = 1;
if (!server->ops->init_transform_rq) {
cifs_dbg(VFS, "Encryption requested but transform callback "
"is missing\n");
return -EIO;
}
rc = server->ops->init_transform_rq(server, num_rqst + 1,
&cur_rqst[0], rqst);
if (rc)
return rc;
rc = __smb_send_rqst(server, num_rqst + 1, &cur_rqst[0]);
smb3_free_compound_rqst(num_rqst, &cur_rqst[1]);
return rc;
}
int
smb_send(struct TCP_Server_Info *server, struct smb_hdr *smb_buffer,
unsigned int smb_buf_length)
{
struct kvec iov[2];
struct smb_rqst rqst = { .rq_iov = iov,
.rq_nvec = 2 };
iov[0].iov_base = smb_buffer;
iov[0].iov_len = 4;
iov[1].iov_base = (char *)smb_buffer + 4;
iov[1].iov_len = smb_buf_length;
return __smb_send_rqst(server, 1, &rqst);
}
static int
wait_for_free_credits(struct TCP_Server_Info *server, const int num_credits,
const int timeout, const int flags,
unsigned int *instance)
{
int rc;
int *credits;
int optype;
long int t;
if (timeout < 0)
t = MAX_JIFFY_OFFSET;
else
t = msecs_to_jiffies(timeout);
optype = flags & CIFS_OP_MASK;
*instance = 0;
credits = server->ops->get_credits_field(server, optype);
/* Since an echo is already inflight, no need to wait to send another */
if (*credits <= 0 && optype == CIFS_ECHO_OP)
return -EAGAIN;
spin_lock(&server->req_lock);
if ((flags & CIFS_TIMEOUT_MASK) == CIFS_NON_BLOCKING) {
/* oplock breaks must not be held up */
server->in_flight++;
*credits -= 1;
*instance = server->reconnect_instance;
spin_unlock(&server->req_lock);
return 0;
}
while (1) {
if (*credits < num_credits) {
spin_unlock(&server->req_lock);
cifs_num_waiters_inc(server);
rc = wait_event_killable_timeout(server->request_q,
has_credits(server, credits, num_credits), t);
cifs_num_waiters_dec(server);
if (!rc) {
trace_smb3_credit_timeout(server->CurrentMid,
server->hostname, num_credits);
cifs_dbg(VFS, "wait timed out after %d ms\n",
timeout);
return -ENOTSUPP;
}
if (rc == -ERESTARTSYS)
return -ERESTARTSYS;
spin_lock(&server->req_lock);
} else {
if (server->tcpStatus == CifsExiting) {
spin_unlock(&server->req_lock);
return -ENOENT;
}
/*
* For normal commands, reserve the last MAX_COMPOUND
* credits to compound requests.
* Otherwise these compounds could be permanently
* starved for credits by single-credit requests.
*
* To prevent spinning CPU, block this thread until
* there are >MAX_COMPOUND credits available.
* But only do this is we already have a lot of
* credits in flight to avoid triggering this check
* for servers that are slow to hand out credits on
* new sessions.
*/
if (!optype && num_credits == 1 &&
server->in_flight > 2 * MAX_COMPOUND &&
*credits <= MAX_COMPOUND) {
spin_unlock(&server->req_lock);
cifs_num_waiters_inc(server);
rc = wait_event_killable_timeout(
server->request_q,
has_credits(server, credits,
MAX_COMPOUND + 1),
t);
cifs_num_waiters_dec(server);
if (!rc) {
trace_smb3_credit_timeout(
server->CurrentMid,
server->hostname, num_credits);
cifs_dbg(VFS, "wait timed out after %d ms\n",
timeout);
return -ENOTSUPP;
}
if (rc == -ERESTARTSYS)
return -ERESTARTSYS;
spin_lock(&server->req_lock);
continue;
}
/*
* Can not count locking commands against total
* as they are allowed to block on server.
*/
/* update # of requests on the wire to server */
if ((flags & CIFS_TIMEOUT_MASK) != CIFS_BLOCKING_OP) {
*credits -= num_credits;
server->in_flight += num_credits;
*instance = server->reconnect_instance;
}
spin_unlock(&server->req_lock);
break;
}
}
return 0;
}
static int
wait_for_free_request(struct TCP_Server_Info *server, const int flags,
unsigned int *instance)
{
return wait_for_free_credits(server, 1, -1, flags,
instance);
}
static int
wait_for_compound_request(struct TCP_Server_Info *server, int num,
const int flags, unsigned int *instance)
{
int *credits;
credits = server->ops->get_credits_field(server, flags & CIFS_OP_MASK);
spin_lock(&server->req_lock);
if (*credits < num) {
/*
* Return immediately if not too many requests in flight since
* we will likely be stuck on waiting for credits.
*/
if (server->in_flight < num - *credits) {
spin_unlock(&server->req_lock);
return -ENOTSUPP;
}
}
spin_unlock(&server->req_lock);
return wait_for_free_credits(server, num, 60000, flags,
instance);
}
static int
__smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
struct smb_rqst *rqst)
{
int rc = 0;
struct kvec *iov;
int n_vec;
unsigned int send_length = 0;
unsigned int i, j;
size_t total_len = 0, sent, size;
struct socket *ssocket = server->ssocket;
struct msghdr smb_msg;
int val = 1;
__be32 rfc1002_marker;
if (cifs_rdma_enabled(server) && server->smbd_conn) {
rc = smbd_send(server, rqst);
goto smbd_done;
}
if (ssocket == NULL)
return -ENOTSOCK;
/* cork the socket */
kernel_setsockopt(ssocket, SOL_TCP, TCP_CORK,
(char *)&val, sizeof(val));
for (j = 0; j < num_rqst; j++)
send_length += smb_rqst_len(server, &rqst[j]);
rfc1002_marker = cpu_to_be32(send_length);
/* Generate a rfc1002 marker for SMB2+ */
if (server->vals->header_preamble_size == 0) {
struct kvec hiov = {
.iov_base = &rfc1002_marker,
.iov_len = 4
};
iov_iter_kvec(&smb_msg.msg_iter, WRITE, &hiov, 1, 4);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
goto uncork;
total_len += sent;
send_length += 4;
}
cifs_dbg(FYI, "Sending smb: smb_len=%u\n", send_length);
for (j = 0; j < num_rqst; j++) {
iov = rqst[j].rq_iov;
n_vec = rqst[j].rq_nvec;
size = 0;
for (i = 0; i < n_vec; i++) {
dump_smb(iov[i].iov_base, iov[i].iov_len);
size += iov[i].iov_len;
}
iov_iter_kvec(&smb_msg.msg_iter, WRITE, iov, n_vec, size);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
goto uncork;
total_len += sent;
/* now walk the page array and send each page in it */
for (i = 0; i < rqst[j].rq_npages; i++) {
struct bio_vec bvec;
bvec.bv_page = rqst[j].rq_pages[i];
rqst_page_get_length(&rqst[j], i, &bvec.bv_len,
&bvec.bv_offset);
iov_iter_bvec(&smb_msg.msg_iter, WRITE,
&bvec, 1, bvec.bv_len);
rc = smb_send_kvec(server, &smb_msg, &sent);
if (rc < 0)
break;
total_len += sent;
}
}
uncork:
/* uncork it */
val = 0;
kernel_setsockopt(ssocket, SOL_TCP, TCP_CORK,
(char *)&val, sizeof(val));
if ((total_len > 0) && (total_len != send_length)) {
cifs_dbg(FYI, "partial send (wanted=%u sent=%zu): terminating session\n",
send_length, total_len);
/*
* If we have only sent part of an SMB then the next SMB could
* be taken as the remainder of this one. We need to kill the
* socket so the server throws away the partial SMB
*/
server->tcpStatus = CifsNeedReconnect;
trace_smb3_partial_send_reconnect(server->CurrentMid,
server->hostname);
}
smbd_done:
if (rc < 0 && rc != -EINTR)
cifs_dbg(VFS, "Error %d sending data on socket to server\n",
rc);
else if (rc > 0)
rc = 0;
return rc;
}
static int
smb_send_rqst(struct TCP_Server_Info *server, int num_rqst,
struct smb_rqst *rqst, int flags)
{
struct kvec iov;
struct smb2_transform_hdr tr_hdr;
struct smb_rqst cur_rqst[MAX_COMPOUND];
int rc;
if (!(flags & CIFS_TRANSFORM_REQ))
return __smb_send_rqst(server, num_rqst, rqst);
if (num_rqst > MAX_COMPOUND - 1)
return -ENOMEM;
memset(&cur_rqst[0], 0, sizeof(cur_rqst));
memset(&iov, 0, sizeof(iov));
memset(&tr_hdr, 0, sizeof(tr_hdr));
iov.iov_base = &tr_hdr;
iov.iov_len = sizeof(tr_hdr);
cur_rqst[0].rq_iov = &iov;
cur_rqst[0].rq_nvec = 1;
if (!server->ops->init_transform_rq) {
cifs_dbg(VFS, "Encryption requested but transform callback "
"is missing\n");
return -EIO;
}
rc = server->ops->init_transform_rq(server, num_rqst + 1,
&cur_rqst[0], rqst);
if (rc)
return rc;
rc = __smb_send_rqst(server, num_rqst + 1, &cur_rqst[0]);
smb3_free_compound_rqst(num_rqst, &cur_rqst[1]);
return rc;
}
int
smb_send(struct TCP_Server_Info *server, struct smb_hdr *smb_buffer,
unsigned int smb_buf_length)
{
struct kvec iov[2];
struct smb_rqst rqst = { .rq_iov = iov,
.rq_nvec = 2 };
iov[0].iov_base = smb_buffer;
iov[0].iov_len = 4;
iov[1].iov_base = (char *)smb_buffer + 4;
iov[1].iov_len = smb_buf_length;
return __smb_send_rqst(server, 1, &rqst);
}
static int
wait_for_free_credits(struct TCP_Server_Info *server, const int timeout,
int *credits)
{
int rc;
spin_lock(&server->req_lock);
if (timeout == CIFS_ASYNC_OP) {
/* oplock breaks must not be held up */
server->in_flight++;
*credits -= 1;
spin_unlock(&server->req_lock);
return 0;
}
while (1) {
if (*credits <= 0) {
spin_unlock(&server->req_lock);
cifs_num_waiters_inc(server);
rc = wait_event_killable(server->request_q,
has_credits(server, credits));
cifs_num_waiters_dec(server);
if (rc)
return rc;
spin_lock(&server->req_lock);
} else {
if (server->tcpStatus == CifsExiting) {
spin_unlock(&server->req_lock);
return -ENOENT;
}
/*
* Can not count locking commands against total
* as they are allowed to block on server.
*/
/* update # of requests on the wire to server */
if (timeout != CIFS_BLOCKING_OP) {
*credits -= 1;
server->in_flight++;
}
spin_unlock(&server->req_lock);
break;
}
}
return 0;
}
static int
wait_for_free_request(struct TCP_Server_Info *server, const int timeout,
const int optype)
{
int *val;
val = server->ops->get_credits_field(server, optype);
/* Since an echo is already inflight, no need to wait to send another */
if (*val <= 0 && optype == CIFS_ECHO_OP)
return -EAGAIN;
return wait_for_free_credits(server, timeout, val);
}