void
xprt_rdma_format_addresses(struct rpc_xprt *xprt, struct sockaddr *sap)
{
char buf[128];
switch (sap->sa_family) {
case AF_INET:
xprt_rdma_format_addresses4(xprt, sap);
break;
case AF_INET6:
xprt_rdma_format_addresses6(xprt, sap);
break;
default:
pr_err("rpcrdma: Unrecognized address family\n");
return;
}
(void)rpc_ntop(sap, buf, sizeof(buf));
xprt->address_strings[RPC_DISPLAY_ADDR] = kstrdup(buf, GFP_KERNEL);
snprintf(buf, sizeof(buf), "%u", rpc_get_port(sap));
xprt->address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL);
snprintf(buf, sizeof(buf), "%4hx", rpc_get_port(sap));
xprt->address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL);
xprt->address_strings[RPC_DISPLAY_PROTO] = "rdma";
}
static void
xprt_rdma_format_addresses(struct rpc_xprt *xprt)
{
struct sockaddr *sap = (struct sockaddr *)
&rpcx_to_rdmad(xprt).addr;
struct sockaddr_in *sin = (struct sockaddr_in *)sap;
char buf[64];
(void)rpc_ntop(sap, buf, sizeof(buf));
xprt->address_strings[RPC_DISPLAY_ADDR] = kstrdup(buf, GFP_KERNEL);
snprintf(buf, sizeof(buf), "%u", rpc_get_port(sap));
xprt->address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL);
xprt->address_strings[RPC_DISPLAY_PROTO] = "rdma";
snprintf(buf, sizeof(buf), "%08x", ntohl(sin->sin_addr.s_addr));
xprt->address_strings[RPC_DISPLAY_HEX_ADDR] = kstrdup(buf, GFP_KERNEL);
snprintf(buf, sizeof(buf), "%4hx", rpc_get_port(sap));
xprt->address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL);
/* netid */
xprt->address_strings[RPC_DISPLAY_NETID] = "rdma";
}
/**
* 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;
#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
struct sockaddr *sap = (struct sockaddr *)&ep->rep_remote_addr;
#endif
struct ib_qp_attr *attr = &ia->ri_qp_attr;
struct ib_qp_init_attr *iattr = &ia->ri_qp_init_attr;
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: %sconnected\n",
__func__, connstate > 0 ? "" : "dis");
atomic_set(&xprt->rx_buf.rb_credits, 1);
ep->rep_connected = connstate;
rpcrdma_conn_func(ep);
wake_up_all(&ep->rep_connect_wait);
/*FALLTHROUGH*/
default:
dprintk("RPC: %s: %pIS:%u (ep 0x%p): %s\n",
__func__, sap, rpc_get_port(sap), ep,
rdma_event_msg(event->event));
break;
}
#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
if (connstate == 1) {
int ird = attr->max_dest_rd_atomic;
int tird = ep->rep_remote_cma.responder_resources;
pr_info("rpcrdma: connection to %pIS:%u on %s, memreg '%s', %d credits, %d responders%s\n",
sap, rpc_get_port(sap),
ia->ri_device->name,
ia->ri_ops->ro_displayname,
xprt->rx_buf.rb_max_requests,
ird, ird < 4 && ird < tird / 2 ? " (low!)" : "");
} else if (connstate < 0) {
pr_info("rpcrdma: connection to %pIS:%u closed (%d)\n",
sap, rpc_get_port(sap), connstate);
}
#endif
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 *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),
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.
*/
sap = (struct sockaddr *)&cdata.addr;
memcpy(sap, 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, sap, xprt->addrlen);
if (rpc_get_port(sap))
xprt_set_bound(xprt);
cdata.max_requests = xprt->max_reqs;
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, sap, 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, sap);
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]);
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);
}
