/**
* xprt_rdma_close - close a transport connection
* @xprt: transport context
*
* Called during transport shutdown, reconnect, or device removal.
* Caller holds @xprt's send lock to prevent activity on this
* transport while the connection is torn down.
*/
static void
xprt_rdma_close(struct rpc_xprt *xprt)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
struct rpcrdma_ep *ep = &r_xprt->rx_ep;
struct rpcrdma_ia *ia = &r_xprt->rx_ia;
dprintk("RPC: %s: closing xprt %p\n", __func__, xprt);
if (test_and_clear_bit(RPCRDMA_IAF_REMOVING, &ia->ri_flags)) {
xprt_clear_connected(xprt);
rpcrdma_ia_remove(ia);
return;
}
if (ep->rep_connected == -ENODEV)
return;
if (ep->rep_connected > 0)
xprt->reestablish_timeout = 0;
xprt_disconnect_done(xprt);
rpcrdma_ep_disconnect(ep, ia);
/* Prepare @xprt for the next connection by reinitializing
* its credit grant to one (see RFC 8166, Section 3.3.3).
*/
r_xprt->rx_buf.rb_credits = 1;
xprt->cwnd = RPC_CWNDSHIFT;
}
/**
* xprt_rdma_allocate - allocate transport resources for an RPC
* @task: RPC task
*
* Return values:
* 0: Success; rq_buffer points to RPC buffer to use
* ENOMEM: Out of memory, call again later
* EIO: A permanent error occurred, do not retry
*/
static int
xprt_rdma_allocate(struct rpc_task *task)
{
struct rpc_rqst *rqst = task->tk_rqstp;
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(rqst->rq_xprt);
struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
gfp_t flags;
flags = RPCRDMA_DEF_GFP;
if (RPC_IS_SWAPPER(task))
flags = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
if (!rpcrdma_get_sendbuf(r_xprt, req, rqst->rq_callsize, flags))
goto out_fail;
if (!rpcrdma_get_recvbuf(r_xprt, req, rqst->rq_rcvsize, flags))
goto out_fail;
rqst->rq_buffer = req->rl_sendbuf->rg_base;
rqst->rq_rbuffer = req->rl_recvbuf->rg_base;
trace_xprtrdma_allocate(task, req);
return 0;
out_fail:
trace_xprtrdma_allocate(task, NULL);
return -ENOMEM;
}
static int
rpcrdma_bc_send_request(struct svcxprt_rdma *rdma, struct rpc_rqst *rqst)
{
struct rpc_xprt *xprt = rqst->rq_xprt;
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
__be32 *p;
int rc;
/* Space in the send buffer for an RPC/RDMA header is reserved
* via xprt->tsh_size.
*/
p = rqst->rq_buffer;
*p++ = rqst->rq_xid;
*p++ = rpcrdma_version;
*p++ = cpu_to_be32(r_xprt->rx_buf.rb_bc_max_requests);
*p++ = rdma_msg;
*p++ = xdr_zero;
*p++ = xdr_zero;
*p = xdr_zero;
#ifdef SVCRDMA_BACKCHANNEL_DEBUG
pr_info("%s: %*ph\n", __func__, 64, rqst->rq_buffer);
#endif
rc = svc_rdma_bc_sendto(rdma, rqst);
if (rc)
goto drop_connection;
return rc;
drop_connection:
dprintk("svcrdma: failed to send bc call\n");
xprt_disconnect_done(xprt);
return -ENOTCONN;
}
static void xprt_rdma_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
long idle_time = 0;
if (xprt_connected(xprt))
idle_time = (long)(jiffies - xprt->last_used) / HZ;
seq_printf(seq,
"\txprt:\trdma %u %lu %lu %lu %ld %lu %lu %lu %Lu %Lu "
"%lu %lu %lu %Lu %Lu %Lu %Lu %lu %lu %lu\n",
0, /* need a local port? */
xprt->stat.bind_count,
xprt->stat.connect_count,
xprt->stat.connect_time,
idle_time,
xprt->stat.sends,
xprt->stat.recvs,
xprt->stat.bad_xids,
xprt->stat.req_u,
xprt->stat.bklog_u,
r_xprt->rx_stats.read_chunk_count,
r_xprt->rx_stats.write_chunk_count,
r_xprt->rx_stats.reply_chunk_count,
r_xprt->rx_stats.total_rdma_request,
r_xprt->rx_stats.total_rdma_reply,
r_xprt->rx_stats.pullup_copy_count,
r_xprt->rx_stats.fixup_copy_count,
r_xprt->rx_stats.hardway_register_count,
r_xprt->rx_stats.failed_marshal_count,
r_xprt->rx_stats.bad_reply_count);
}
/* It shouldn't matter if the number of backchannel session slots
* doesn't match the number of RPC/RDMA credits. That just means
* one or the other will have extra slots that aren't used.
*/
static struct rpc_xprt *
xprt_setup_rdma_bc(struct xprt_create *args)
{
struct rpc_xprt *xprt;
struct rpcrdma_xprt *new_xprt;
if (args->addrlen > sizeof(xprt->addr)) {
dprintk("RPC: %s: address too large\n", __func__);
return ERR_PTR(-EBADF);
}
xprt = xprt_alloc(args->net, sizeof(*new_xprt),
RPCRDMA_MAX_BC_REQUESTS,
RPCRDMA_MAX_BC_REQUESTS);
if (!xprt) {
dprintk("RPC: %s: couldn't allocate rpc_xprt\n",
__func__);
return ERR_PTR(-ENOMEM);
}
xprt->timeout = &xprt_rdma_bc_timeout;
xprt_set_bound(xprt);
xprt_set_connected(xprt);
xprt->bind_timeout = RPCRDMA_BIND_TO;
xprt->reestablish_timeout = RPCRDMA_INIT_REEST_TO;
xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO;
xprt->prot = XPRT_TRANSPORT_BC_RDMA;
xprt->tsh_size = RPCRDMA_HDRLEN_MIN / sizeof(__be32);
xprt->ops = &xprt_rdma_bc_procs;
memcpy(&xprt->addr, args->dstaddr, args->addrlen);
xprt->addrlen = args->addrlen;
xprt_rdma_format_addresses(xprt, (struct sockaddr *)&xprt->addr);
xprt->resvport = 0;
xprt->max_payload = xprt_rdma_max_inline_read;
new_xprt = rpcx_to_rdmax(xprt);
new_xprt->rx_buf.rb_bc_max_requests = xprt->max_reqs;
xprt_get(xprt);
args->bc_xprt->xpt_bc_xprt = xprt;
xprt->bc_xprt = args->bc_xprt;
if (!try_module_get(THIS_MODULE))
goto out_fail;
/* Final put for backchannel xprt is in __svc_rdma_free */
xprt_get(xprt);
return xprt;
out_fail:
xprt_rdma_free_addresses(xprt);
args->bc_xprt->xpt_bc_xprt = NULL;
args->bc_xprt->xpt_bc_xps = NULL;
xprt_put(xprt);
xprt_free(xprt);
return ERR_PTR(-EINVAL);
}
/**
* xprt_rdma_inject_disconnect - inject a connection fault
* @xprt: transport context
*
* If @xprt is connected, disconnect it to simulate spurious connection
* loss.
*/
static void
xprt_rdma_inject_disconnect(struct rpc_xprt *xprt)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
trace_xprtrdma_inject_dsc(r_xprt);
rdma_disconnect(r_xprt->rx_ia.ri_id);
}
/*
* Close a connection, during shutdown or timeout/reconnect
*/
static void
xprt_rdma_close(struct rpc_xprt *xprt)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
dprintk("RPC: %s: closing\n", __func__);
if (r_xprt->rx_ep.rep_connected > 0)
xprt->reestablish_timeout = 0;
xprt_disconnect_done(xprt);
rpcrdma_ep_disconnect(&r_xprt->rx_ep, &r_xprt->rx_ia);
}
/**
* xprt_rdma_free - release resources allocated by xprt_rdma_allocate
* @task: RPC task
*
* Caller guarantees rqst->rq_buffer is non-NULL.
*/
static void
xprt_rdma_free(struct rpc_task *task)
{
struct rpc_rqst *rqst = task->tk_rqstp;
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(rqst->rq_xprt);
struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
if (test_bit(RPCRDMA_REQ_F_PENDING, &req->rl_flags))
rpcrdma_release_rqst(r_xprt, req);
trace_xprtrdma_rpc_done(task, req);
}
/**
* xprt_rdma_send_request - marshal and send an RPC request
* @rqst: RPC message in rq_snd_buf
*
* Caller holds the transport's write lock.
*
* Returns:
* %0 if the RPC message has been sent
* %-ENOTCONN if the caller should reconnect and call again
* %-EAGAIN if the caller should call again
* %-ENOBUFS if the caller should call again after a delay
* %-EIO if a permanent error occurred and the request was not
* sent. Do not try to send this message again.
*/
static int
xprt_rdma_send_request(struct rpc_rqst *rqst)
{
struct rpc_xprt *xprt = rqst->rq_xprt;
struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
int rc = 0;
#if defined(CONFIG_SUNRPC_BACKCHANNEL)
if (unlikely(!rqst->rq_buffer))
return xprt_rdma_bc_send_reply(rqst);
#endif /* CONFIG_SUNRPC_BACKCHANNEL */
if (!xprt_connected(xprt))
goto drop_connection;
if (!xprt_request_get_cong(xprt, rqst))
return -EBADSLT;
rc = rpcrdma_marshal_req(r_xprt, rqst);
if (rc < 0)
goto failed_marshal;
/* Must suppress retransmit to maintain credits */
if (rqst->rq_connect_cookie == xprt->connect_cookie)
goto drop_connection;
rqst->rq_xtime = ktime_get();
__set_bit(RPCRDMA_REQ_F_PENDING, &req->rl_flags);
if (rpcrdma_ep_post(&r_xprt->rx_ia, &r_xprt->rx_ep, req))
goto drop_connection;
rqst->rq_xmit_bytes_sent += rqst->rq_snd_buf.len;
rqst->rq_bytes_sent = 0;
/* An RPC with no reply will throw off credit accounting,
* so drop the connection to reset the credit grant.
*/
if (!rpc_reply_expected(rqst->rq_task))
goto drop_connection;
return 0;
failed_marshal:
if (rc != -ENOTCONN)
return rc;
drop_connection:
xprt_disconnect_done(xprt);
return -ENOTCONN; /* implies disconnect */
}
static int
xprt_rdma_send_request(struct rpc_task *task)
{
struct rpc_rqst *rqst = task->tk_rqstp;
struct rpc_xprt *xprt = rqst->rq_xprt;
struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
int rc = 0;
if (req->rl_niovs == 0)
rc = rpcrdma_marshal_req(rqst);
else if (r_xprt->rx_ia.ri_memreg_strategy != RPCRDMA_ALLPHYSICAL)
rc = rpcrdma_marshal_chunks(rqst, 0);
if (rc < 0)
goto failed_marshal;
if (req->rl_reply == NULL) /* e.g. reconnection */
rpcrdma_recv_buffer_get(req);
if (req->rl_reply) {
req->rl_reply->rr_func = rpcrdma_reply_handler;
/* this need only be done once, but... */
req->rl_reply->rr_xprt = xprt;
}
/* Must suppress retransmit to maintain credits */
if (req->rl_connect_cookie == xprt->connect_cookie)
goto drop_connection;
req->rl_connect_cookie = xprt->connect_cookie;
if (rpcrdma_ep_post(&r_xprt->rx_ia, &r_xprt->rx_ep, req))
goto drop_connection;
rqst->rq_xmit_bytes_sent += rqst->rq_snd_buf.len;
rqst->rq_bytes_sent = 0;
return 0;
failed_marshal:
r_xprt->rx_stats.failed_marshal_count++;
dprintk("RPC: %s: rpcrdma_marshal_req failed, status %i\n",
__func__, rc);
if (rc == -EIO)
return -EIO;
drop_connection:
xprt_disconnect_done(xprt);
return -ENOTCONN; /* implies disconnect */
}
/**
* xprt_rdma_alloc_slot - allocate an rpc_rqst
* @xprt: controlling RPC transport
* @task: RPC task requesting a fresh rpc_rqst
*
* tk_status values:
* %0 if task->tk_rqstp points to a fresh rpc_rqst
* %-EAGAIN if no rpc_rqst is available; queued on backlog
*/
static void
xprt_rdma_alloc_slot(struct rpc_xprt *xprt, struct rpc_task *task)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
struct rpcrdma_req *req;
req = rpcrdma_buffer_get(&r_xprt->rx_buf);
if (!req)
goto out_sleep;
task->tk_rqstp = &req->rl_slot;
task->tk_status = 0;
return;
out_sleep:
rpc_sleep_on(&xprt->backlog, task, NULL);
task->tk_status = -EAGAIN;
}
/**
* xprt_rdma_destroy - Full tear down of transport
* @xprt: doomed transport context
*
* Caller guarantees there will be no more calls to us with
* this @xprt.
*/
static void
xprt_rdma_destroy(struct rpc_xprt *xprt)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
trace_xprtrdma_destroy(r_xprt);
cancel_delayed_work_sync(&r_xprt->rx_connect_worker);
rpcrdma_ep_destroy(&r_xprt->rx_ep, &r_xprt->rx_ia);
rpcrdma_buffer_destroy(&r_xprt->rx_buf);
rpcrdma_ia_close(&r_xprt->rx_ia);
xprt_rdma_free_addresses(xprt);
xprt_free(xprt);
module_put(THIS_MODULE);
}
/**
* xprt_rdma_connect - try to establish a transport connection
* @xprt: transport state
* @task: RPC scheduler context
*
*/
static void
xprt_rdma_connect(struct rpc_xprt *xprt, struct rpc_task *task)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
if (r_xprt->rx_ep.rep_connected != 0) {
/* Reconnect */
schedule_delayed_work(&r_xprt->rx_connect_worker,
xprt->reestablish_timeout);
xprt->reestablish_timeout <<= 1;
if (xprt->reestablish_timeout > RPCRDMA_MAX_REEST_TO)
xprt->reestablish_timeout = RPCRDMA_MAX_REEST_TO;
else if (xprt->reestablish_timeout < RPCRDMA_INIT_REEST_TO)
xprt->reestablish_timeout = RPCRDMA_INIT_REEST_TO;
} else {
schedule_delayed_work(&r_xprt->rx_connect_worker, 0);
if (!RPC_IS_ASYNC(task))
flush_delayed_work(&r_xprt->rx_connect_worker);
}
}
void xprt_rdma_print_stats(struct rpc_xprt *xprt, struct seq_file *seq)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
long idle_time = 0;
if (xprt_connected(xprt))
idle_time = (long)(jiffies - xprt->last_used) / HZ;
seq_puts(seq, "\txprt:\trdma ");
seq_printf(seq, "%u %lu %lu %lu %ld %lu %lu %lu %llu %llu ",
0, /* need a local port? */
xprt->stat.bind_count,
xprt->stat.connect_count,
xprt->stat.connect_time / HZ,
idle_time,
xprt->stat.sends,
xprt->stat.recvs,
xprt->stat.bad_xids,
xprt->stat.req_u,
xprt->stat.bklog_u);
seq_printf(seq, "%lu %lu %lu %llu %llu %llu %llu %lu %lu %lu %lu ",
r_xprt->rx_stats.read_chunk_count,
r_xprt->rx_stats.write_chunk_count,
r_xprt->rx_stats.reply_chunk_count,
r_xprt->rx_stats.total_rdma_request,
r_xprt->rx_stats.total_rdma_reply,
r_xprt->rx_stats.pullup_copy_count,
r_xprt->rx_stats.fixup_copy_count,
r_xprt->rx_stats.hardway_register_count,
r_xprt->rx_stats.failed_marshal_count,
r_xprt->rx_stats.bad_reply_count,
r_xprt->rx_stats.nomsg_call_count);
seq_printf(seq, "%lu %lu %lu %lu %lu %lu\n",
r_xprt->rx_stats.mrs_recycled,
r_xprt->rx_stats.mrs_orphaned,
r_xprt->rx_stats.mrs_allocated,
r_xprt->rx_stats.local_inv_needed,
r_xprt->rx_stats.empty_sendctx_q,
r_xprt->rx_stats.reply_waits_for_send);
}
static int
xprt_rdma_send_request(struct rpc_task *task)
{
struct rpc_rqst *rqst = task->tk_rqstp;
struct rpc_xprt *xprt = rqst->rq_xprt;
struct rpcrdma_req *req = rpcr_to_rdmar(rqst);
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
int rc = 0;
rc = rpcrdma_marshal_req(rqst);
if (rc < 0)
goto failed_marshal;
if (req->rl_reply == NULL) /* e.g. reconnection */
rpcrdma_recv_buffer_get(req);
/* Must suppress retransmit to maintain credits */
if (req->rl_connect_cookie == xprt->connect_cookie)
goto drop_connection;
req->rl_connect_cookie = xprt->connect_cookie;
if (rpcrdma_ep_post(&r_xprt->rx_ia, &r_xprt->rx_ep, req))
goto drop_connection;
rqst->rq_xmit_bytes_sent += rqst->rq_snd_buf.len;
rqst->rq_bytes_sent = 0;
return 0;
failed_marshal:
r_xprt->rx_stats.failed_marshal_count++;
dprintk("RPC: %s: rpcrdma_marshal_req failed, status %i\n",
__func__, rc);
if (rc == -EIO)
return -EIO;
drop_connection:
xprt_disconnect_done(xprt);
return -ENOTCONN; /* implies disconnect */
}
static int
rpcrdma_bc_send_request(struct svcxprt_rdma *rdma, struct rpc_rqst *rqst)
{
struct rpc_xprt *xprt = rqst->rq_xprt;
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
struct svc_rdma_send_ctxt *ctxt;
__be32 *p;
int rc;
ctxt = svc_rdma_send_ctxt_get(rdma);
if (!ctxt)
goto drop_connection;
p = ctxt->sc_xprt_buf;
*p++ = rqst->rq_xid;
*p++ = rpcrdma_version;
*p++ = cpu_to_be32(r_xprt->rx_buf.rb_bc_max_requests);
*p++ = rdma_msg;
*p++ = xdr_zero;
*p++ = xdr_zero;
*p = xdr_zero;
svc_rdma_sync_reply_hdr(rdma, ctxt, RPCRDMA_HDRLEN_MIN);
#ifdef SVCRDMA_BACKCHANNEL_DEBUG
pr_info("%s: %*ph\n", __func__, 64, rqst->rq_buffer);
#endif
rc = svc_rdma_bc_sendto(rdma, rqst, ctxt);
if (rc) {
svc_rdma_send_ctxt_put(rdma, ctxt);
goto drop_connection;
}
return 0;
drop_connection:
dprintk("svcrdma: failed to send bc call\n");
return -ENOTCONN;
}
/*
* xprt_rdma_destroy
*
* Destroy the xprt.
* Free all memory associated with the object, including its own.
* NOTE: none of the *destroy methods free memory for their top-level
* objects, even though they may have allocated it (they do free
* private memory). It's up to the caller to handle it. In this
* case (RDMA transport), all structure memory is inlined with the
* struct rpcrdma_xprt.
*/
static void
xprt_rdma_destroy(struct rpc_xprt *xprt)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
dprintk("RPC: %s: called\n", __func__);
cancel_delayed_work_sync(&r_xprt->rx_connect_worker);
xprt_clear_connected(xprt);
rpcrdma_buffer_destroy(&r_xprt->rx_buf);
rpcrdma_ep_destroy(&r_xprt->rx_ep, &r_xprt->rx_ia);
rpcrdma_ia_close(&r_xprt->rx_ia);
xprt_rdma_free_addresses(xprt);
xprt_free(xprt);
dprintk("RPC: %s: returning\n", __func__);
module_put(THIS_MODULE);
}
static int
rpcrdma_conn_upcall(struct rdma_cm_id *id, struct rdma_cm_event *event)
{
struct rpcrdma_xprt *xprt = id->context;
struct rpcrdma_ia *ia = &xprt->rx_ia;
struct rpcrdma_ep *ep = &xprt->rx_ep;
struct sockaddr_in *addr = (struct sockaddr_in *) &ep->rep_remote_addr;
struct ib_qp_attr attr;
struct ib_qp_init_attr iattr;
int connstate = 0;
switch (event->event) {
case RDMA_CM_EVENT_ADDR_RESOLVED:
case RDMA_CM_EVENT_ROUTE_RESOLVED:
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ADDR_ERROR:
ia->ri_async_rc = -EHOSTUNREACH;
dprintk("RPC: %s: CM address resolution error, ep 0x%p\n",
__func__, ep);
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ROUTE_ERROR:
ia->ri_async_rc = -ENETUNREACH;
dprintk("RPC: %s: CM route resolution error, ep 0x%p\n",
__func__, ep);
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ESTABLISHED:
connstate = 1;
ib_query_qp(ia->ri_id->qp, &attr,
IB_QP_MAX_QP_RD_ATOMIC | IB_QP_MAX_DEST_RD_ATOMIC,
&iattr);
dprintk("RPC: %s: %d responder resources"
" (%d initiator)\n",
__func__, attr.max_dest_rd_atomic, attr.max_rd_atomic);
goto connected;
case RDMA_CM_EVENT_CONNECT_ERROR:
connstate = -ENOTCONN;
goto connected;
case RDMA_CM_EVENT_UNREACHABLE:
connstate = -ENETDOWN;
goto connected;
case RDMA_CM_EVENT_REJECTED:
connstate = -ECONNREFUSED;
goto connected;
case RDMA_CM_EVENT_DISCONNECTED:
connstate = -ECONNABORTED;
goto connected;
case RDMA_CM_EVENT_DEVICE_REMOVAL:
connstate = -ENODEV;
connected:
dprintk("RPC: %s: %s: %u.%u.%u.%u:%u"
" (ep 0x%p event 0x%x)\n",
__func__,
(event->event <= 11) ? conn[event->event] :
"unknown connection error",
NIPQUAD(addr->sin_addr.s_addr),
ntohs(addr->sin_port),
ep, event->event);
atomic_set(&rpcx_to_rdmax(ep->rep_xprt)->rx_buf.rb_credits, 1);
dprintk("RPC: %s: %sconnected\n",
__func__, connstate > 0 ? "" : "dis");
ep->rep_connected = connstate;
ep->rep_func(ep);
wake_up_all(&ep->rep_connect_wait);
break;
default:
ia->ri_async_rc = -EINVAL;
dprintk("RPC: %s: unexpected CM event %X\n",
__func__, event->event);
complete(&ia->ri_done);
break;
}
return 0;
}
/**
* xprt_setup_rdma - Set up transport to use RDMA
*
* @args: rpc transport arguments
*/
static struct rpc_xprt *
xprt_setup_rdma(struct xprt_create *args)
{
struct rpcrdma_create_data_internal cdata;
struct rpc_xprt *xprt;
struct rpcrdma_xprt *new_xprt;
struct rpcrdma_ep *new_ep;
struct sockaddr_in *sin;
int rc;
if (args->addrlen > sizeof(xprt->addr)) {
dprintk("RPC: %s: address too large\n", __func__);
return ERR_PTR(-EBADF);
}
xprt = xprt_alloc(args->net, sizeof(struct rpcrdma_xprt),
xprt_rdma_slot_table_entries,
xprt_rdma_slot_table_entries);
if (xprt == NULL) {
dprintk("RPC: %s: couldn't allocate rpcrdma_xprt\n",
__func__);
return ERR_PTR(-ENOMEM);
}
/* 60 second timeout, no retries */
xprt->timeout = &xprt_rdma_default_timeout;
xprt->bind_timeout = RPCRDMA_BIND_TO;
xprt->reestablish_timeout = RPCRDMA_INIT_REEST_TO;
xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO;
xprt->resvport = 0; /* privileged port not needed */
xprt->tsh_size = 0; /* RPC-RDMA handles framing */
xprt->ops = &xprt_rdma_procs;
/*
* Set up RDMA-specific connect data.
*/
/* Put server RDMA address in local cdata */
memcpy(&cdata.addr, args->dstaddr, args->addrlen);
/* Ensure xprt->addr holds valid server TCP (not RDMA)
* address, for any side protocols which peek at it */
xprt->prot = IPPROTO_TCP;
xprt->addrlen = args->addrlen;
memcpy(&xprt->addr, &cdata.addr, xprt->addrlen);
sin = (struct sockaddr_in *)&cdata.addr;
if (ntohs(sin->sin_port) != 0)
xprt_set_bound(xprt);
dprintk("RPC: %s: %pI4:%u\n",
__func__, &sin->sin_addr.s_addr, ntohs(sin->sin_port));
/* Set max requests */
cdata.max_requests = xprt->max_reqs;
/* Set some length limits */
cdata.rsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA write max */
cdata.wsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA read max */
cdata.inline_wsize = xprt_rdma_max_inline_write;
if (cdata.inline_wsize > cdata.wsize)
cdata.inline_wsize = cdata.wsize;
cdata.inline_rsize = xprt_rdma_max_inline_read;
if (cdata.inline_rsize > cdata.rsize)
cdata.inline_rsize = cdata.rsize;
cdata.padding = xprt_rdma_inline_write_padding;
/*
* Create new transport instance, which includes initialized
* o ia
* o endpoint
* o buffers
*/
new_xprt = rpcx_to_rdmax(xprt);
rc = rpcrdma_ia_open(new_xprt, (struct sockaddr *) &cdata.addr,
xprt_rdma_memreg_strategy);
if (rc)
goto out1;
/*
* initialize and create ep
*/
new_xprt->rx_data = cdata;
new_ep = &new_xprt->rx_ep;
new_ep->rep_remote_addr = cdata.addr;
rc = rpcrdma_ep_create(&new_xprt->rx_ep,
&new_xprt->rx_ia, &new_xprt->rx_data);
if (rc)
goto out2;
/*
* Allocate pre-registered send and receive buffers for headers and
* any inline data. Also specify any padding which will be provided
* from a preregistered zero buffer.
*/
rc = rpcrdma_buffer_create(new_xprt);
if (rc)
goto out3;
/*
* Register a callback for connection events. This is necessary because
* connection loss notification is async. We also catch connection loss
* when reaping receives.
*/
INIT_DELAYED_WORK(&new_xprt->rx_connect_worker,
xprt_rdma_connect_worker);
xprt_rdma_format_addresses(xprt);
xprt->max_payload = rpcrdma_max_payload(new_xprt);
dprintk("RPC: %s: transport data payload maximum: %zu bytes\n",
__func__, xprt->max_payload);
if (!try_module_get(THIS_MODULE))
goto out4;
return xprt;
out4:
xprt_rdma_free_addresses(xprt);
rc = -EINVAL;
out3:
rpcrdma_ep_destroy(new_ep, &new_xprt->rx_ia);
out2:
rpcrdma_ia_close(&new_xprt->rx_ia);
out1:
xprt_free(xprt);
return ERR_PTR(rc);
}
/*
* The RDMA allocate/free functions need the task structure as a place
* to hide the struct rpcrdma_req, which is necessary for the actual send/recv
* sequence.
*
* The RPC layer allocates both send and receive buffers in the same call
* (rq_send_buf and rq_rcv_buf are both part of a single contiguous buffer).
* We may register rq_rcv_buf when using reply chunks.
*/
static void *
xprt_rdma_allocate(struct rpc_task *task, size_t size)
{
struct rpc_xprt *xprt = task->tk_rqstp->rq_xprt;
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
struct rpcrdma_regbuf *rb;
struct rpcrdma_req *req;
size_t min_size;
gfp_t flags;
req = rpcrdma_buffer_get(&r_xprt->rx_buf);
if (req == NULL)
return NULL;
flags = GFP_NOIO | __GFP_NOWARN;
if (RPC_IS_SWAPPER(task))
flags = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
if (req->rl_rdmabuf == NULL)
goto out_rdmabuf;
if (req->rl_sendbuf == NULL)
goto out_sendbuf;
if (size > req->rl_sendbuf->rg_size)
goto out_sendbuf;
out:
dprintk("RPC: %s: size %zd, request 0x%p\n", __func__, size, req);
req->rl_connect_cookie = 0; /* our reserved value */
return req->rl_sendbuf->rg_base;
out_rdmabuf:
min_size = RPCRDMA_INLINE_WRITE_THRESHOLD(task->tk_rqstp);
rb = rpcrdma_alloc_regbuf(&r_xprt->rx_ia, min_size, flags);
if (IS_ERR(rb))
goto out_fail;
req->rl_rdmabuf = rb;
out_sendbuf:
/* XDR encoding and RPC/RDMA marshaling of this request has not
* yet occurred. Thus a lower bound is needed to prevent buffer
* overrun during marshaling.
*
* RPC/RDMA marshaling may choose to send payload bearing ops
* inline, if the result is smaller than the inline threshold.
* The value of the "size" argument accounts for header
* requirements but not for the payload in these cases.
*
* Likewise, allocate enough space to receive a reply up to the
* size of the inline threshold.
*
* It's unlikely that both the send header and the received
* reply will be large, but slush is provided here to allow
* flexibility when marshaling.
*/
min_size = RPCRDMA_INLINE_READ_THRESHOLD(task->tk_rqstp);
min_size += RPCRDMA_INLINE_WRITE_THRESHOLD(task->tk_rqstp);
if (size < min_size)
size = min_size;
rb = rpcrdma_alloc_regbuf(&r_xprt->rx_ia, size, flags);
if (IS_ERR(rb))
goto out_fail;
rb->rg_owner = req;
r_xprt->rx_stats.hardway_register_count += size;
rpcrdma_free_regbuf(&r_xprt->rx_ia, req->rl_sendbuf);
req->rl_sendbuf = rb;
goto out;
out_fail:
rpcrdma_buffer_put(req);
r_xprt->rx_stats.failed_marshal_count++;
return NULL;
}
int svc_rdma_handle_bc_reply(struct rpc_xprt *xprt, struct rpcrdma_msg *rmsgp,
struct xdr_buf *rcvbuf)
{
struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
struct kvec *dst, *src = &rcvbuf->head[0];
struct rpc_rqst *req;
unsigned long cwnd;
u32 credits;
size_t len;
__be32 xid;
__be32 *p;
int ret;
p = (__be32 *)src->iov_base;
len = src->iov_len;
xid = rmsgp->rm_xid;
#ifdef SVCRDMA_BACKCHANNEL_DEBUG
pr_info("%s: xid=%08x, length=%zu\n",
__func__, be32_to_cpu(xid), len);
pr_info("%s: RPC/RDMA: %*ph\n",
__func__, (int)RPCRDMA_HDRLEN_MIN, rmsgp);
pr_info("%s: RPC: %*ph\n",
__func__, (int)len, p);
#endif
ret = -EAGAIN;
if (src->iov_len < 24)
goto out_shortreply;
spin_lock_bh(&xprt->transport_lock);
req = xprt_lookup_rqst(xprt, xid);
if (!req)
goto out_notfound;
dst = &req->rq_private_buf.head[0];
memcpy(&req->rq_private_buf, &req->rq_rcv_buf, sizeof(struct xdr_buf));
if (dst->iov_len < len)
goto out_unlock;
memcpy(dst->iov_base, p, len);
credits = be32_to_cpu(rmsgp->rm_credit);
if (credits == 0)
credits = 1; /* don't deadlock */
else if (credits > r_xprt->rx_buf.rb_bc_max_requests)
credits = r_xprt->rx_buf.rb_bc_max_requests;
cwnd = xprt->cwnd;
xprt->cwnd = credits << RPC_CWNDSHIFT;
if (xprt->cwnd > cwnd)
xprt_release_rqst_cong(req->rq_task);
ret = 0;
xprt_complete_rqst(req->rq_task, rcvbuf->len);
rcvbuf->len = 0;
out_unlock:
spin_unlock_bh(&xprt->transport_lock);
out:
return ret;
out_shortreply:
dprintk("svcrdma: short bc reply: xprt=%p, len=%zu\n",
xprt, src->iov_len);
goto out;
out_notfound:
dprintk("svcrdma: unrecognized bc reply: xprt=%p, xid=%08x\n",
xprt, be32_to_cpu(xid));
goto out_unlock;
}
/**
* xprt_setup_rdma - Set up transport to use RDMA
*
* @args: rpc transport arguments
*/
static struct rpc_xprt *
xprt_setup_rdma(struct xprt_create *args)
{
struct rpcrdma_create_data_internal cdata;
struct rpc_xprt *xprt;
struct rpcrdma_xprt *new_xprt;
struct rpcrdma_ep *new_ep;
struct sockaddr *sap;
int rc;
if (args->addrlen > sizeof(xprt->addr)) {
dprintk("RPC: %s: address too large\n", __func__);
return ERR_PTR(-EBADF);
}
xprt = xprt_alloc(args->net, sizeof(struct rpcrdma_xprt), 0, 0);
if (xprt == NULL) {
dprintk("RPC: %s: couldn't allocate rpcrdma_xprt\n",
__func__);
return ERR_PTR(-ENOMEM);
}
/* 60 second timeout, no retries */
xprt->timeout = &xprt_rdma_default_timeout;
xprt->bind_timeout = RPCRDMA_BIND_TO;
xprt->reestablish_timeout = RPCRDMA_INIT_REEST_TO;
xprt->idle_timeout = RPCRDMA_IDLE_DISC_TO;
xprt->resvport = 0; /* privileged port not needed */
xprt->tsh_size = 0; /* RPC-RDMA handles framing */
xprt->ops = &xprt_rdma_procs;
/*
* Set up RDMA-specific connect data.
*/
sap = args->dstaddr;
/* Ensure xprt->addr holds valid server TCP (not RDMA)
* address, for any side protocols which peek at it */
xprt->prot = IPPROTO_TCP;
xprt->addrlen = args->addrlen;
memcpy(&xprt->addr, sap, xprt->addrlen);
if (rpc_get_port(sap))
xprt_set_bound(xprt);
xprt_rdma_format_addresses(xprt, sap);
cdata.max_requests = xprt_rdma_slot_table_entries;
cdata.rsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA write max */
cdata.wsize = RPCRDMA_MAX_SEGS * PAGE_SIZE; /* RDMA read max */
cdata.inline_wsize = xprt_rdma_max_inline_write;
if (cdata.inline_wsize > cdata.wsize)
cdata.inline_wsize = cdata.wsize;
cdata.inline_rsize = xprt_rdma_max_inline_read;
if (cdata.inline_rsize > cdata.rsize)
cdata.inline_rsize = cdata.rsize;
/*
* Create new transport instance, which includes initialized
* o ia
* o endpoint
* o buffers
*/
new_xprt = rpcx_to_rdmax(xprt);
rc = rpcrdma_ia_open(new_xprt);
if (rc)
goto out1;
/*
* initialize and create ep
*/
new_xprt->rx_data = cdata;
new_ep = &new_xprt->rx_ep;
rc = rpcrdma_ep_create(&new_xprt->rx_ep,
&new_xprt->rx_ia, &new_xprt->rx_data);
if (rc)
goto out2;
rc = rpcrdma_buffer_create(new_xprt);
if (rc)
goto out3;
INIT_DELAYED_WORK(&new_xprt->rx_connect_worker,
xprt_rdma_connect_worker);
xprt->max_payload = new_xprt->rx_ia.ri_ops->ro_maxpages(new_xprt);
if (xprt->max_payload == 0)
goto out4;
xprt->max_payload <<= PAGE_SHIFT;
dprintk("RPC: %s: transport data payload maximum: %zu bytes\n",
__func__, xprt->max_payload);
if (!try_module_get(THIS_MODULE))
goto out4;
dprintk("RPC: %s: %s:%s\n", __func__,
xprt->address_strings[RPC_DISPLAY_ADDR],
xprt->address_strings[RPC_DISPLAY_PORT]);
trace_xprtrdma_create(new_xprt);
return xprt;
out4:
rpcrdma_buffer_destroy(&new_xprt->rx_buf);
rc = -ENODEV;
out3:
rpcrdma_ep_destroy(new_ep, &new_xprt->rx_ia);
out2:
rpcrdma_ia_close(&new_xprt->rx_ia);
out1:
trace_xprtrdma_destroy(new_xprt);
xprt_rdma_free_addresses(xprt);
xprt_free(xprt);
return ERR_PTR(rc);
}
static int
rpcrdma_conn_upcall(struct rdma_cm_id *id, struct rdma_cm_event *event)
{
struct rpcrdma_xprt *xprt = id->context;
struct rpcrdma_ia *ia = &xprt->rx_ia;
struct rpcrdma_ep *ep = &xprt->rx_ep;
#ifdef RPC_DEBUG
struct sockaddr_in *addr = (struct sockaddr_in *) &ep->rep_remote_addr;
#endif
struct ib_qp_attr attr;
struct ib_qp_init_attr iattr;
int connstate = 0;
switch (event->event) {
case RDMA_CM_EVENT_ADDR_RESOLVED:
case RDMA_CM_EVENT_ROUTE_RESOLVED:
ia->ri_async_rc = 0;
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ADDR_ERROR:
ia->ri_async_rc = -EHOSTUNREACH;
dprintk("RPC: %s: CM address resolution error, ep 0x%p\n",
__func__, ep);
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ROUTE_ERROR:
ia->ri_async_rc = -ENETUNREACH;
dprintk("RPC: %s: CM route resolution error, ep 0x%p\n",
__func__, ep);
complete(&ia->ri_done);
break;
case RDMA_CM_EVENT_ESTABLISHED:
connstate = 1;
ib_query_qp(ia->ri_id->qp, &attr,
IB_QP_MAX_QP_RD_ATOMIC | IB_QP_MAX_DEST_RD_ATOMIC,
&iattr);
dprintk("RPC: %s: %d responder resources"
" (%d initiator)\n",
__func__, attr.max_dest_rd_atomic, attr.max_rd_atomic);
goto connected;
case RDMA_CM_EVENT_CONNECT_ERROR:
connstate = -ENOTCONN;
goto connected;
case RDMA_CM_EVENT_UNREACHABLE:
connstate = -ENETDOWN;
goto connected;
case RDMA_CM_EVENT_REJECTED:
connstate = -ECONNREFUSED;
goto connected;
case RDMA_CM_EVENT_DISCONNECTED:
connstate = -ECONNABORTED;
goto connected;
case RDMA_CM_EVENT_DEVICE_REMOVAL:
connstate = -ENODEV;
connected:
dprintk("RPC: %s: %s: %pI4:%u (ep 0x%p event 0x%x)\n",
__func__,
(event->event <= 11) ? conn[event->event] :
"unknown connection error",
&addr->sin_addr.s_addr,
ntohs(addr->sin_port),
ep, event->event);
atomic_set(&rpcx_to_rdmax(ep->rep_xprt)->rx_buf.rb_credits, 1);
dprintk("RPC: %s: %sconnected\n",
__func__, connstate > 0 ? "" : "dis");
ep->rep_connected = connstate;
ep->rep_func(ep);
wake_up_all(&ep->rep_connect_wait);
break;
default:
dprintk("RPC: %s: unexpected CM event %d\n",
__func__, event->event);
break;
}
#ifdef RPC_DEBUG
if (connstate == 1) {
int ird = attr.max_dest_rd_atomic;
int tird = ep->rep_remote_cma.responder_resources;
printk(KERN_INFO "rpcrdma: connection to %pI4:%u "
"on %s, memreg %d slots %d ird %d%s\n",
&addr->sin_addr.s_addr,
ntohs(addr->sin_port),
ia->ri_id->device->name,
ia->ri_memreg_strategy,
xprt->rx_buf.rb_max_requests,
ird, ird < 4 && ird < tird / 2 ? " (low!)" : "");
} else if (connstate < 0) {
printk(KERN_INFO "rpcrdma: connection to %pI4:%u closed (%d)\n",
&addr->sin_addr.s_addr,
ntohs(addr->sin_port),
connstate);
}
#endif
return 0;
}