summaryrefslogtreecommitdiffstats
path: root/net/sunrpc
diff options
context:
space:
mode:
Diffstat (limited to 'net/sunrpc')
-rw-r--r--net/sunrpc/backchannel_rqst.c5
-rw-r--r--net/sunrpc/clnt.c15
-rw-r--r--net/sunrpc/rpcb_clnt.c8
-rw-r--r--net/sunrpc/sched.c7
-rw-r--r--net/sunrpc/stats.c26
-rw-r--r--net/sunrpc/svc.c4
-rw-r--r--net/sunrpc/svc_xprt.c3
-rw-r--r--net/sunrpc/xprt.c38
-rw-r--r--net/sunrpc/xprtrdma/svc_rdma_marshal.c16
-rw-r--r--net/sunrpc/xprtrdma/svc_rdma_recvfrom.c244
-rw-r--r--net/sunrpc/xprtrdma/svc_rdma_sendto.c46
-rw-r--r--net/sunrpc/xprtrdma/svc_rdma_transport.c47
-rw-r--r--net/sunrpc/xprtsock.c238
13 files changed, 372 insertions, 325 deletions
diff --git a/net/sunrpc/backchannel_rqst.c b/net/sunrpc/backchannel_rqst.c
index 651f49ab601f..9dd0ea8db463 100644
--- a/net/sunrpc/backchannel_rqst.c
+++ b/net/sunrpc/backchannel_rqst.c
@@ -309,12 +309,15 @@ void xprt_complete_bc_request(struct rpc_rqst *req, uint32_t copied)
struct rpc_xprt *xprt = req->rq_xprt;
struct svc_serv *bc_serv = xprt->bc_serv;
+ spin_lock(&xprt->bc_pa_lock);
+ list_del(&req->rq_bc_pa_list);
+ spin_unlock(&xprt->bc_pa_lock);
+
req->rq_private_buf.len = copied;
set_bit(RPC_BC_PA_IN_USE, &req->rq_bc_pa_state);
dprintk("RPC: add callback request to list\n");
spin_lock(&bc_serv->sv_cb_lock);
- list_del(&req->rq_bc_pa_list);
list_add(&req->rq_bc_list, &bc_serv->sv_cb_list);
wake_up(&bc_serv->sv_cb_waitq);
spin_unlock(&bc_serv->sv_cb_lock);
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 05da12a33945..612aa73bbc60 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -286,10 +286,8 @@ static struct rpc_xprt *rpc_clnt_set_transport(struct rpc_clnt *clnt,
static void rpc_clnt_set_nodename(struct rpc_clnt *clnt, const char *nodename)
{
- clnt->cl_nodelen = strlen(nodename);
- if (clnt->cl_nodelen > UNX_MAXNODENAME)
- clnt->cl_nodelen = UNX_MAXNODENAME;
- memcpy(clnt->cl_nodename, nodename, clnt->cl_nodelen);
+ clnt->cl_nodelen = strlcpy(clnt->cl_nodename,
+ nodename, sizeof(clnt->cl_nodename));
}
static int rpc_client_register(struct rpc_clnt *clnt,
@@ -365,6 +363,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args,
const struct rpc_version *version;
struct rpc_clnt *clnt = NULL;
const struct rpc_timeout *timeout;
+ const char *nodename = args->nodename;
int err;
/* sanity check the name before trying to print it */
@@ -420,8 +419,10 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args,
atomic_set(&clnt->cl_count, 1);
+ if (nodename == NULL)
+ nodename = utsname()->nodename;
/* save the nodename */
- rpc_clnt_set_nodename(clnt, utsname()->nodename);
+ rpc_clnt_set_nodename(clnt, nodename);
err = rpc_client_register(clnt, args->authflavor, args->client_name);
if (err)
@@ -576,6 +577,7 @@ static struct rpc_clnt *__rpc_clone_client(struct rpc_create_args *args,
if (xprt == NULL)
goto out_err;
args->servername = xprt->servername;
+ args->nodename = clnt->cl_nodename;
new = rpc_new_client(args, xprt, clnt);
if (IS_ERR(new)) {
@@ -1824,6 +1826,7 @@ call_connect_status(struct rpc_task *task)
case -ECONNABORTED:
case -ENETUNREACH:
case -EHOSTUNREACH:
+ case -EADDRINUSE:
case -ENOBUFS:
case -EPIPE:
if (RPC_IS_SOFTCONN(task))
@@ -1932,6 +1935,7 @@ call_transmit_status(struct rpc_task *task)
}
case -ECONNRESET:
case -ECONNABORTED:
+ case -EADDRINUSE:
case -ENOTCONN:
case -ENOBUFS:
case -EPIPE:
@@ -2051,6 +2055,7 @@ call_status(struct rpc_task *task)
case -ECONNRESET:
case -ECONNABORTED:
rpc_force_rebind(clnt);
+ case -EADDRINUSE:
case -ENOBUFS:
rpc_delay(task, 3*HZ);
case -EPIPE:
diff --git a/net/sunrpc/rpcb_clnt.c b/net/sunrpc/rpcb_clnt.c
index 05202012bcfc..cf5770d8f49a 100644
--- a/net/sunrpc/rpcb_clnt.c
+++ b/net/sunrpc/rpcb_clnt.c
@@ -355,7 +355,8 @@ out:
return result;
}
-static struct rpc_clnt *rpcb_create(struct net *net, const char *hostname,
+static struct rpc_clnt *rpcb_create(struct net *net, const char *nodename,
+ const char *hostname,
struct sockaddr *srvaddr, size_t salen,
int proto, u32 version)
{
@@ -365,6 +366,7 @@ static struct rpc_clnt *rpcb_create(struct net *net, const char *hostname,
.address = srvaddr,
.addrsize = salen,
.servername = hostname,
+ .nodename = nodename,
.program = &rpcb_program,
.version = version,
.authflavor = RPC_AUTH_UNIX,
@@ -740,7 +742,9 @@ void rpcb_getport_async(struct rpc_task *task)
dprintk("RPC: %5u %s: trying rpcbind version %u\n",
task->tk_pid, __func__, bind_version);
- rpcb_clnt = rpcb_create(xprt->xprt_net, xprt->servername, sap, salen,
+ rpcb_clnt = rpcb_create(xprt->xprt_net,
+ clnt->cl_nodename,
+ xprt->servername, sap, salen,
xprt->prot, bind_version);
if (IS_ERR(rpcb_clnt)) {
status = PTR_ERR(rpcb_clnt);
diff --git a/net/sunrpc/sched.c b/net/sunrpc/sched.c
index d20f2329eea3..b91fd9c597b4 100644
--- a/net/sunrpc/sched.c
+++ b/net/sunrpc/sched.c
@@ -844,10 +844,10 @@ static void rpc_async_schedule(struct work_struct *work)
void *rpc_malloc(struct rpc_task *task, size_t size)
{
struct rpc_buffer *buf;
- gfp_t gfp = GFP_NOWAIT | __GFP_NOWARN;
+ gfp_t gfp = GFP_NOIO | __GFP_NOWARN;
if (RPC_IS_SWAPPER(task))
- gfp |= __GFP_MEMALLOC;
+ gfp = __GFP_MEMALLOC | GFP_NOWAIT | __GFP_NOWARN;
size += sizeof(struct rpc_buffer);
if (size <= RPC_BUFFER_MAXSIZE)
@@ -1069,7 +1069,8 @@ static int rpciod_start(void)
* Create the rpciod thread and wait for it to start.
*/
dprintk("RPC: creating workqueue rpciod\n");
- wq = alloc_workqueue("rpciod", WQ_MEM_RECLAIM, 1);
+ /* Note: highpri because network receive is latency sensitive */
+ wq = alloc_workqueue("rpciod", WQ_MEM_RECLAIM | WQ_HIGHPRI, 0);
rpciod_workqueue = wq;
return rpciod_workqueue != NULL;
}
diff --git a/net/sunrpc/stats.c b/net/sunrpc/stats.c
index 9711a155bc50..2ecb994314c1 100644
--- a/net/sunrpc/stats.c
+++ b/net/sunrpc/stats.c
@@ -140,22 +140,20 @@ void rpc_free_iostats(struct rpc_iostats *stats)
EXPORT_SYMBOL_GPL(rpc_free_iostats);
/**
- * rpc_count_iostats - tally up per-task stats
+ * rpc_count_iostats_metrics - tally up per-task stats
* @task: completed rpc_task
- * @stats: array of stat structures
+ * @op_metrics: stat structure for OP that will accumulate stats from @task
*/
-void rpc_count_iostats(const struct rpc_task *task, struct rpc_iostats *stats)
+void rpc_count_iostats_metrics(const struct rpc_task *task,
+ struct rpc_iostats *op_metrics)
{
struct rpc_rqst *req = task->tk_rqstp;
- struct rpc_iostats *op_metrics;
ktime_t delta, now;
- if (!stats || !req)
+ if (!op_metrics || !req)
return;
now = ktime_get();
- op_metrics = &stats[task->tk_msg.rpc_proc->p_statidx];
-
spin_lock(&op_metrics->om_lock);
op_metrics->om_ops++;
@@ -175,6 +173,20 @@ void rpc_count_iostats(const struct rpc_task *task, struct rpc_iostats *stats)
spin_unlock(&op_metrics->om_lock);
}
+EXPORT_SYMBOL_GPL(rpc_count_iostats_metrics);
+
+/**
+ * rpc_count_iostats - tally up per-task stats
+ * @task: completed rpc_task
+ * @stats: array of stat structures
+ *
+ * Uses the statidx from @task
+ */
+void rpc_count_iostats(const struct rpc_task *task, struct rpc_iostats *stats)
+{
+ rpc_count_iostats_metrics(task,
+ &stats[task->tk_msg.rpc_proc->p_statidx]);
+}
EXPORT_SYMBOL_GPL(rpc_count_iostats);
static void _print_name(struct seq_file *seq, unsigned int op,
diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
index 91eaef1844c8..78974e4d9ad2 100644
--- a/net/sunrpc/svc.c
+++ b/net/sunrpc/svc.c
@@ -768,8 +768,8 @@ svc_set_num_threads(struct svc_serv *serv, struct svc_pool *pool, int nrservs)
EXPORT_SYMBOL_GPL(svc_set_num_threads);
/*
- * Called from a server thread as it's exiting. Caller must hold the BKL or
- * the "service mutex", whichever is appropriate for the service.
+ * Called from a server thread as it's exiting. Caller must hold the "service
+ * mutex" for the service.
*/
void
svc_exit_thread(struct svc_rqst *rqstp)
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index c69358b3cf7f..163ac45c3639 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -42,7 +42,7 @@ static LIST_HEAD(svc_xprt_class_list);
* svc_pool->sp_lock protects most of the fields of that pool.
* svc_serv->sv_lock protects sv_tempsocks, sv_permsocks, sv_tmpcnt.
* when both need to be taken (rare), svc_serv->sv_lock is first.
- * BKL protects svc_serv->sv_nrthread.
+ * The "service mutex" protects svc_serv->sv_nrthread.
* svc_sock->sk_lock protects the svc_sock->sk_deferred list
* and the ->sk_info_authunix cache.
*
@@ -67,7 +67,6 @@ static LIST_HEAD(svc_xprt_class_list);
* that no other thread will be using the transport or will
* try to set XPT_DEAD.
*/
-
int svc_reg_xprt_class(struct svc_xprt_class *xcl)
{
struct svc_xprt_class *cl;
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index ebbefad21a37..e3015aede0d9 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -683,13 +683,43 @@ xprt_init_autodisconnect(unsigned long data)
if (test_and_set_bit(XPRT_LOCKED, &xprt->state))
goto out_abort;
spin_unlock(&xprt->transport_lock);
- set_bit(XPRT_CONNECTION_CLOSE, &xprt->state);
queue_work(rpciod_workqueue, &xprt->task_cleanup);
return;
out_abort:
spin_unlock(&xprt->transport_lock);
}
+bool xprt_lock_connect(struct rpc_xprt *xprt,
+ struct rpc_task *task,
+ void *cookie)
+{
+ bool ret = false;
+
+ spin_lock_bh(&xprt->transport_lock);
+ if (!test_bit(XPRT_LOCKED, &xprt->state))
+ goto out;
+ if (xprt->snd_task != task)
+ goto out;
+ xprt->snd_task = cookie;
+ ret = true;
+out:
+ spin_unlock_bh(&xprt->transport_lock);
+ return ret;
+}
+
+void xprt_unlock_connect(struct rpc_xprt *xprt, void *cookie)
+{
+ spin_lock_bh(&xprt->transport_lock);
+ if (xprt->snd_task != cookie)
+ goto out;
+ if (!test_bit(XPRT_LOCKED, &xprt->state))
+ goto out;
+ xprt->snd_task =NULL;
+ xprt->ops->release_xprt(xprt, NULL);
+out:
+ spin_unlock_bh(&xprt->transport_lock);
+}
+
/**
* xprt_connect - schedule a transport connect operation
* @task: RPC task that is requesting the connect
@@ -712,9 +742,7 @@ void xprt_connect(struct rpc_task *task)
if (test_and_clear_bit(XPRT_CLOSE_WAIT, &xprt->state))
xprt->ops->close(xprt);
- if (xprt_connected(xprt))
- xprt_release_write(xprt, task);
- else {
+ if (!xprt_connected(xprt)) {
task->tk_rqstp->rq_bytes_sent = 0;
task->tk_timeout = task->tk_rqstp->rq_timeout;
rpc_sleep_on(&xprt->pending, task, xprt_connect_status);
@@ -726,6 +754,7 @@ void xprt_connect(struct rpc_task *task)
xprt->stat.connect_start = jiffies;
xprt->ops->connect(xprt, task);
}
+ xprt_release_write(xprt, task);
}
static void xprt_connect_status(struct rpc_task *task)
@@ -758,7 +787,6 @@ static void xprt_connect_status(struct rpc_task *task)
dprintk("RPC: %5u xprt_connect_status: error %d connecting to "
"server %s\n", task->tk_pid, -task->tk_status,
xprt->servername);
- xprt_release_write(xprt, task);
task->tk_status = -EIO;
}
}
diff --git a/net/sunrpc/xprtrdma/svc_rdma_marshal.c b/net/sunrpc/xprtrdma/svc_rdma_marshal.c
index 65b146297f5a..b681855cf970 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_marshal.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_marshal.c
@@ -71,22 +71,6 @@ static u32 *decode_read_list(u32 *va, u32 *vaend)
}
/*
- * Determine number of chunks and total bytes in chunk list. The chunk
- * list has already been verified to fit within the RPCRDMA header.
- */
-void svc_rdma_rcl_chunk_counts(struct rpcrdma_read_chunk *ch,
- int *ch_count, int *byte_count)
-{
- /* compute the number of bytes represented by read chunks */
- *byte_count = 0;
- *ch_count = 0;
- for (; ch->rc_discrim != 0; ch++) {
- *byte_count = *byte_count + ntohl(ch->rc_target.rs_length);
- *ch_count = *ch_count + 1;
- }
-}
-
-/*
* Decodes a write chunk list. The expected format is as follows:
* descrim : xdr_one
* nchunks : <count>
diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
index e0110270d650..f9f13a32ddb8 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
@@ -43,7 +43,6 @@
#include <linux/sunrpc/debug.h>
#include <linux/sunrpc/rpc_rdma.h>
#include <linux/spinlock.h>
-#include <linux/highmem.h>
#include <asm/unaligned.h>
#include <rdma/ib_verbs.h>
#include <rdma/rdma_cm.h>
@@ -60,6 +59,7 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
struct svc_rdma_op_ctxt *ctxt,
u32 byte_count)
{
+ struct rpcrdma_msg *rmsgp;
struct page *page;
u32 bc;
int sge_no;
@@ -82,7 +82,14 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
/* If data remains, store it in the pagelist */
rqstp->rq_arg.page_len = bc;
rqstp->rq_arg.page_base = 0;
- rqstp->rq_arg.pages = &rqstp->rq_pages[1];
+
+ /* RDMA_NOMSG: RDMA READ data should land just after RDMA RECV data */
+ rmsgp = (struct rpcrdma_msg *)rqstp->rq_arg.head[0].iov_base;
+ if (be32_to_cpu(rmsgp->rm_type) == RDMA_NOMSG)
+ rqstp->rq_arg.pages = &rqstp->rq_pages[0];
+ else
+ rqstp->rq_arg.pages = &rqstp->rq_pages[1];
+
sge_no = 1;
while (bc && sge_no < ctxt->count) {
page = ctxt->pages[sge_no];
@@ -95,14 +102,6 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
rqstp->rq_respages = &rqstp->rq_pages[sge_no];
rqstp->rq_next_page = rqstp->rq_respages + 1;
- /* We should never run out of SGE because the limit is defined to
- * support the max allowed RPC data length
- */
- BUG_ON(bc && (sge_no == ctxt->count));
- BUG_ON((rqstp->rq_arg.head[0].iov_len + rqstp->rq_arg.page_len)
- != byte_count);
- BUG_ON(rqstp->rq_arg.len != byte_count);
-
/* If not all pages were used from the SGL, free the remaining ones */
bc = sge_no;
while (sge_no < ctxt->count) {
@@ -125,26 +124,16 @@ static int rdma_read_max_sge(struct svcxprt_rdma *xprt, int sge_count)
return min_t(int, sge_count, xprt->sc_max_sge);
}
-typedef int (*rdma_reader_fn)(struct svcxprt_rdma *xprt,
- struct svc_rqst *rqstp,
- struct svc_rdma_op_ctxt *head,
- int *page_no,
- u32 *page_offset,
- u32 rs_handle,
- u32 rs_length,
- u64 rs_offset,
- int last);
-
/* Issue an RDMA_READ using the local lkey to map the data sink */
-static int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt,
- struct svc_rqst *rqstp,
- struct svc_rdma_op_ctxt *head,
- int *page_no,
- u32 *page_offset,
- u32 rs_handle,
- u32 rs_length,
- u64 rs_offset,
- int last)
+int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt,
+ struct svc_rqst *rqstp,
+ struct svc_rdma_op_ctxt *head,
+ int *page_no,
+ u32 *page_offset,
+ u32 rs_handle,
+ u32 rs_length,
+ u64 rs_offset,
+ bool last)
{
struct ib_send_wr read_wr;
int pages_needed = PAGE_ALIGN(*page_offset + rs_length) >> PAGE_SHIFT;
@@ -229,15 +218,15 @@ static int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt,
}
/* Issue an RDMA_READ using an FRMR to map the data sink */
-static int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
- struct svc_rqst *rqstp,
- struct svc_rdma_op_ctxt *head,
- int *page_no,
- u32 *page_offset,
- u32 rs_handle,
- u32 rs_length,
- u64 rs_offset,
- int last)
+int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
+ struct svc_rqst *rqstp,
+ struct svc_rdma_op_ctxt *head,
+ int *page_no,
+ u32 *page_offset,
+ u32 rs_handle,
+ u32 rs_length,
+ u64 rs_offset,
+ bool last)
{
struct ib_send_wr read_wr;
struct ib_send_wr inv_wr;
@@ -365,24 +354,84 @@ static int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt,
return ret;
}
+static unsigned int
+rdma_rcl_chunk_count(struct rpcrdma_read_chunk *ch)
+{
+ unsigned int count;
+
+ for (count = 0; ch->rc_discrim != xdr_zero; ch++)
+ count++;
+ return count;
+}
+
+/* If there was additional inline content, append it to the end of arg.pages.
+ * Tail copy has to be done after the reader function has determined how many
+ * pages are needed for RDMA READ.
+ */
+static int
+rdma_copy_tail(struct svc_rqst *rqstp, struct svc_rdma_op_ctxt *head,
+ u32 position, u32 byte_count, u32 page_offset, int page_no)
+{
+ char *srcp, *destp;
+ int ret;
+
+ ret = 0;
+ srcp = head->arg.head[0].iov_base + position;
+ byte_count = head->arg.head[0].iov_len - position;
+ if (byte_count > PAGE_SIZE) {
+ dprintk("svcrdma: large tail unsupported\n");
+ return 0;
+ }
+
+ /* Fit as much of the tail on the current page as possible */
+ if (page_offset != PAGE_SIZE) {
+ destp = page_address(rqstp->rq_arg.pages[page_no]);
+ destp += page_offset;
+ while (byte_count--) {
+ *destp++ = *srcp++;
+ page_offset++;
+ if (page_offset == PAGE_SIZE && byte_count)
+ goto more;
+ }
+ goto done;
+ }
+
+more:
+ /* Fit the rest on the next page */
+ page_no++;
+ destp = page_address(rqstp->rq_arg.pages[page_no]);
+ while (byte_count--)
+ *destp++ = *srcp++;
+
+ rqstp->rq_respages = &rqstp->rq_arg.pages[page_no+1];
+ rqstp->rq_next_page = rqstp->rq_respages + 1;
+
+done:
+ byte_count = head->arg.head[0].iov_len - position;
+ head->arg.page_len += byte_count;
+ head->arg.len += byte_count;
+ head->arg.buflen += byte_count;
+ return 1;
+}
+
static int rdma_read_chunks(struct svcxprt_rdma *xprt,
struct rpcrdma_msg *rmsgp,
struct svc_rqst *rqstp,
struct svc_rdma_op_ctxt *head)
{
- int page_no, ch_count, ret;
+ int page_no, ret;
struct rpcrdma_read_chunk *ch;
- u32 page_offset, byte_count;
+ u32 handle, page_offset, byte_count;
+ u32 position;
u64 rs_offset;
- rdma_reader_fn reader;
+ bool last;
/* If no read list is present, return 0 */
ch = svc_rdma_get_read_chunk(rmsgp);
if (!ch)
return 0;
- svc_rdma_rcl_chunk_counts(ch, &ch_count, &byte_count);
- if (ch_count > RPCSVC_MAXPAGES)
+ if (rdma_rcl_chunk_count(ch) > RPCSVC_MAXPAGES)
return -EINVAL;
/* The request is completed when the RDMA_READs complete. The
@@ -391,34 +440,41 @@ static int rdma_read_chunks(struct svcxprt_rdma *xprt,
*/
head->arg.head[0] = rqstp->rq_arg.head[0];
head->arg.tail[0] = rqstp->rq_arg.tail[0];
- head->arg.pages = &head->pages[head->count];
head->hdr_count = head->count;
head->arg.page_base = 0;
head->arg.page_len = 0;
head->arg.len = rqstp->rq_arg.len;
head->arg.buflen = rqstp->rq_arg.buflen;
- /* Use FRMR if supported */
- if (xprt->sc_dev_caps & SVCRDMA_DEVCAP_FAST_REG)
- reader = rdma_read_chunk_frmr;
- else
- reader = rdma_read_chunk_lcl;
+ ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
+ position = be32_to_cpu(ch->rc_position);
+
+ /* RDMA_NOMSG: RDMA READ data should land just after RDMA RECV data */
+ if (position == 0) {
+ head->arg.pages = &head->pages[0];
+ page_offset = head->byte_len;
+ } else {
+ head->arg.pages = &head->pages[head->count];
+ page_offset = 0;
+ }
- page_no = 0; page_offset = 0;
- for (ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
- ch->rc_discrim != 0; ch++) {
+ ret = 0;
+ page_no = 0;
+ for (; ch->rc_discrim != xdr_zero; ch++) {
+ if (be32_to_cpu(ch->rc_position) != position)
+ goto err;
+ handle = be32_to_cpu(ch->rc_target.rs_handle),
+ byte_count = be32_to_cpu(ch->rc_target.rs_length);
xdr_decode_hyper((__be32 *)&ch->rc_target.rs_offset,
&rs_offset);
- byte_count = ntohl(ch->rc_target.rs_length);
while (byte_count > 0) {
- ret = reader(xprt, rqstp, head,
- &page_no, &page_offset,
- ntohl(ch->rc_target.rs_handle),
- byte_count, rs_offset,
- ((ch+1)->rc_discrim == 0) /* last */
- );
+ last = (ch + 1)->rc_discrim == xdr_zero;
+ ret = xprt->sc_reader(xprt, rqstp, head,
+ &page_no, &page_offset,
+ handle, byte_count,
+ rs_offset, last);
if (ret < 0)
goto err;
byte_count -= ret;
@@ -426,7 +482,24 @@ static int rdma_read_chunks(struct svcxprt_rdma *xprt,
head->arg.buflen += ret;
}
}
+
+ /* Read list may need XDR round-up (see RFC 5666, s. 3.7) */
+ if (page_offset & 3) {
+ u32 pad = 4 - (page_offset & 3);
+
+ head->arg.page_len += pad;
+ head->arg.len += pad;
+ head->arg.buflen += pad;
+ page_offset += pad;
+ }
+
ret = 1;
+ if (position && position < head->arg.head[0].iov_len)
+ ret = rdma_copy_tail(rqstp, head, position,
+ byte_count, page_offset, page_no);
+ head->arg.head[0].iov_len = position;
+ head->position = position;
+
err:
/* Detach arg pages. svc_recv will replenish them */
for (page_no = 0;
@@ -436,47 +509,33 @@ static int rdma_read_chunks(struct svcxprt_rdma *xprt,
return ret;
}
-/*
- * To avoid a separate RDMA READ just for a handful of zero bytes,
- * RFC 5666 section 3.7 allows the client to omit the XDR zero pad
- * in chunk lists.
- */
-static void
-rdma_fix_xdr_pad(struct xdr_buf *buf)
-{
- unsigned int page_len = buf->page_len;
- unsigned int size = (XDR_QUADLEN(page_len) << 2) - page_len;
- unsigned int offset, pg_no;
- char *p;
-
- if (size == 0)
- return;
-
- pg_no = page_len >> PAGE_SHIFT;
- offset = page_len & ~PAGE_MASK;
- p = page_address(buf->pages[pg_no]);
- memset(p + offset, 0, size);
-
- buf->page_len += size;
- buf->buflen += size;
- buf->len += size;
-}
-
static int rdma_read_complete(struct svc_rqst *rqstp,
struct svc_rdma_op_ctxt *head)
{
int page_no;
int ret;
- BUG_ON(!head);
-
/* Copy RPC pages */
for (page_no = 0; page_no < head->count; page_no++) {
put_page(rqstp->rq_pages[page_no]);
rqstp->rq_pages[page_no] = head->pages[page_no];
}
+
+ /* Adjustments made for RDMA_NOMSG type requests */
+ if (head->position == 0) {
+ if (head->arg.len <= head->sge[0].length) {
+ head->arg.head[0].iov_len = head->arg.len -
+ head->byte_len;
+ head->arg.page_len = 0;
+ } else {
+ head->arg.head[0].iov_len = head->sge[0].length -
+ head->byte_len;
+ head->arg.page_len = head->arg.len -
+ head->sge[0].length;
+ }
+ }
+
/* Point rq_arg.pages past header */
- rdma_fix_xdr_pad(&head->arg);
rqstp->rq_arg.pages = &rqstp->rq_pages[head->hdr_count];
rqstp->rq_arg.page_len = head->arg.page_len;
rqstp->rq_arg.page_base = head->arg.page_base;
@@ -501,8 +560,8 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
ret = rqstp->rq_arg.head[0].iov_len
+ rqstp->rq_arg.page_len
+ rqstp->rq_arg.tail[0].iov_len;
- dprintk("svcrdma: deferred read ret=%d, rq_arg.len =%d, "
- "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
+ dprintk("svcrdma: deferred read ret=%d, rq_arg.len=%u, "
+ "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len=%zu\n",
ret, rqstp->rq_arg.len, rqstp->rq_arg.head[0].iov_base,
rqstp->rq_arg.head[0].iov_len);
@@ -558,7 +617,6 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
}
dprintk("svcrdma: processing ctxt=%p on xprt=%p, rqstp=%p, status=%d\n",
ctxt, rdma_xprt, rqstp, ctxt->wc_status);
- BUG_ON(ctxt->wc_status != IB_WC_SUCCESS);
atomic_inc(&rdma_stat_recv);
/* Build up the XDR from the receive buffers. */
@@ -591,8 +649,8 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
+ rqstp->rq_arg.tail[0].iov_len;
svc_rdma_put_context(ctxt, 0);
out:
- dprintk("svcrdma: ret = %d, rq_arg.len =%d, "
- "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
+ dprintk("svcrdma: ret=%d, rq_arg.len=%u, "
+ "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len=%zd\n",
ret, rqstp->rq_arg.len,
rqstp->rq_arg.head[0].iov_base,
rqstp->rq_arg.head[0].iov_len);
diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
index 9f1b50689c0f..7de33d1af9b6 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
@@ -60,8 +60,11 @@ static int map_xdr(struct svcxprt_rdma *xprt,
u32 page_off;
int page_no;
- BUG_ON(xdr->len !=
- (xdr->head[0].iov_len + xdr->page_len + xdr->tail[0].iov_len));
+ if (xdr->len !=
+ (xdr->head[0].iov_len + xdr->page_len + xdr->tail[0].iov_len)) {
+ pr_err("svcrdma: map_xdr: XDR buffer length error\n");
+ return -EIO;
+ }
/* Skip the first sge, this is for the RPCRDMA header */
sge_no = 1;
@@ -150,7 +153,11 @@ static int send_write(struct svcxprt_rdma *xprt, struct svc_rqst *rqstp,
int bc;
struct svc_rdma_op_ctxt *ctxt;
- BUG_ON(vec->count > RPCSVC_MAXPAGES);
+ if (vec->count > RPCSVC_MAXPAGES) {
+ pr_err("svcrdma: Too many pages (%lu)\n", vec->count);
+ return -EIO;
+ }
+
dprintk("svcrdma: RDMA_WRITE rmr=%x, to=%llx, xdr_off=%d, "
"write_len=%d, vec->sge=%p, vec->count=%lu\n",
rmr, (unsigned long long)to, xdr_off,
@@ -190,7 +197,10 @@ static int send_write(struct svcxprt_rdma *xprt, struct svc_rqst *rqstp,
sge_off = 0;
sge_no++;
xdr_sge_no++;
- BUG_ON(xdr_sge_no > vec->count);
+ if (xdr_sge_no > vec->count) {
+ pr_err("svcrdma: Too many sges (%d)\n", xdr_sge_no);
+ goto err;
+ }
bc -= sge_bytes;
if (sge_no == xprt->sc_max_sge)
break;
@@ -421,7 +431,10 @@ static int send_reply(struct svcxprt_rdma *rdma,
ctxt->sge[sge_no].lkey = rdma->sc_dma_lkey;
ctxt->sge[sge_no].length = sge_bytes;
}
- BUG_ON(byte_count != 0);
+ if (byte_count != 0) {
+ pr_err("svcrdma: Could not map %d bytes\n", byte_count);
+ goto err;
+ }
/* Save all respages in the ctxt and remove them from the
* respages array. They are our pages until the I/O
@@ -442,7 +455,10 @@ static int send_reply(struct svcxprt_rdma *rdma,
}
rqstp->rq_next_page = rqstp->rq_respages + 1;
- BUG_ON(sge_no > rdma->sc_max_sge);
+ if (sge_no > rdma->sc_max_sge) {
+ pr_err("svcrdma: Too many sges (%d)\n", sge_no);
+ goto err;
+ }
memset(&send_wr, 0, sizeof send_wr);
ctxt->wr_op = IB_WR_SEND;
send_wr.wr_id = (unsigned long)ctxt;
@@ -467,18 +483,6 @@ void svc_rdma_prep_reply_hdr(struct svc_rqst *rqstp)
{
}
-/*
- * Return the start of an xdr buffer.
- */
-static void *xdr_start(struct xdr_buf *xdr)
-{
- return xdr->head[0].iov_base -
- (xdr->len -
- xdr->page_len -
- xdr->tail[0].iov_len -
- xdr->head[0].iov_len);
-}
-
int svc_rdma_sendto(struct svc_rqst *rqstp)
{
struct svc_xprt *xprt = rqstp->rq_xprt;
@@ -496,8 +500,10 @@ int svc_rdma_sendto(struct svc_rqst *rqstp)
dprintk("svcrdma: sending response for rqstp=%p\n", rqstp);
- /* Get the RDMA request header. */
- rdma_argp = xdr_start(&rqstp->rq_arg);
+ /* Get the RDMA request header. The receive logic always
+ * places this at the start of page 0.
+ */
+ rdma_argp = page_address(rqstp->rq_pages[0]);
/* Build an req vec for the XDR */
ctxt = svc_rdma_get_context(rdma);
diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
index 4e618808bc98..f609c1c2d38d 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c
@@ -139,7 +139,6 @@ void svc_rdma_put_context(struct svc_rdma_op_ctxt *ctxt, int free_pages)
struct svcxprt_rdma *xprt;
int i;
- BUG_ON(!ctxt);
xprt = ctxt->xprt;
if (free_pages)
for (i = 0; i < ctxt->count; i++)
@@ -339,12 +338,14 @@ static void process_context(struct svcxprt_rdma *xprt,
switch (ctxt->wr_op) {
case IB_WR_SEND:
- BUG_ON(ctxt->frmr);
+ if (ctxt->frmr)
+ pr_err("svcrdma: SEND: ctxt->frmr != NULL\n");
svc_rdma_put_context(ctxt, 1);
break;
case IB_WR_RDMA_WRITE:
- BUG_ON(ctxt->frmr);
+ if (ctxt->frmr)
+ pr_err("svcrdma: WRITE: ctxt->frmr != NULL\n");
svc_rdma_put_context(ctxt, 0);
break;
@@ -353,19 +354,21 @@ static void process_context(struct svcxprt_rdma *xprt,
svc_rdma_put_frmr(xprt, ctxt->frmr);
if (test_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags)) {
struct svc_rdma_op_ctxt *read_hdr = ctxt->read_hdr;
- BUG_ON(!read_hdr);
- spin_lock_bh(&xprt->sc_rq_dto_lock);
- set_bit(XPT_DATA, &xprt->sc_xprt.xpt_flags);
- list_add_tail(&read_hdr->dto_q,
- &xprt->sc_read_complete_q);
- spin_unlock_bh(&xprt->sc_rq_dto_lock);
+ if (read_hdr) {
+ spin_lock_bh(&xprt->sc_rq_dto_lock);
+ set_bit(XPT_DATA, &xprt->sc_xprt.xpt_flags);
+ list_add_tail(&read_hdr->dto_q,
+ &xprt->sc_read_complete_q);
+ spin_unlock_bh(&xprt->sc_rq_dto_lock);
+ } else {
+ pr_err("svcrdma: ctxt->read_hdr == NULL\n");
+ }
svc_xprt_enqueue(&xprt->sc_xprt);
}
svc_rdma_put_context(ctxt, 0);
break;
default:
- BUG_ON(1);
printk(KERN_ERR "svcrdma: unexpected completion type, "
"opcode=%d\n",
ctxt->wr_op);
@@ -513,7 +516,10 @@ int svc_rdma_post_recv(struct svcxprt_rdma *xprt)
buflen = 0;
ctxt->direction = DMA_FROM_DEVICE;
for (sge_no = 0; buflen < xprt->sc_max_req_size; sge_no++) {
- BUG_ON(sge_no >= xprt->sc_max_sge);
+ if (sge_no >= xprt->sc_max_sge) {
+ pr_err("svcrdma: Too many sges (%d)\n", sge_no);
+ goto err_put_ctxt;
+ }
page = svc_rdma_get_page();
ctxt->pages[sge_no] = page;
pa = ib_dma_map_page(xprt->sc_cm_id->device,
@@ -687,7 +693,6 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv,
{
struct rdma_cm_id *listen_id;
struct svcxprt_rdma *cma_xprt;
- struct svc_xprt *xprt;
int ret;
dprintk("svcrdma: Creating RDMA socket\n");
@@ -698,7 +703,6 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv,
cma_xprt = rdma_create_xprt(serv, 1);
if (!cma_xprt)
return ERR_PTR(-ENOMEM);
- xprt = &cma_xprt->sc_xprt;
listen_id = rdma_create_id(rdma_listen_handler, cma_xprt, RDMA_PS_TCP,
IB_QPT_RC);
@@ -822,7 +826,7 @@ void svc_rdma_put_frmr(struct svcxprt_rdma *rdma,
if (frmr) {
frmr_unmap_dma(rdma, frmr);
spin_lock_bh(&rdma->sc_frmr_q_lock);
- BUG_ON(!list_empty(&frmr->frmr_list));
+ WARN_ON_ONCE(!list_empty(&frmr->frmr_list));
list_add(&frmr->frmr_list, &rdma->sc_frmr_q);
spin_unlock_bh(&rdma->sc_frmr_q_lock);
}
@@ -970,10 +974,12 @@ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt)
* NB: iWARP requires remote write access for the data sink
* of an RDMA_READ. IB does not.
*/
+ newxprt->sc_reader = rdma_read_chunk_lcl;
if (devattr.device_cap_flags & IB_DEVICE_MEM_MGT_EXTENSIONS) {
newxprt->sc_frmr_pg_list_len =
devattr.max_fast_reg_page_list_len;
newxprt->sc_dev_caps |= SVCRDMA_DEVCAP_FAST_REG;
+ newxprt->sc_reader = rdma_read_chunk_frmr;
}
/*
@@ -1125,7 +1131,9 @@ static void __svc_rdma_free(struct work_struct *work)
dprintk("svcrdma: svc_rdma_free(%p)\n", rdma);
/* We should only be called from kref_put */
- BUG_ON(atomic_read(&rdma->sc_xprt.xpt_ref.refcount) != 0);
+ if (atomic_read(&rdma->sc_xprt.xpt_ref.refcount) != 0)
+ pr_err("svcrdma: sc_xprt still in use? (%d)\n",
+ atomic_read(&rdma->sc_xprt.xpt_ref.refcount));
/*
* Destroy queued, but not processed read completions. Note
@@ -1153,8 +1161,12 @@ static void __svc_rdma_free(struct work_struct *work)
}
/* Warn if we leaked a resource or under-referenced */
- WARN_ON(atomic_read(&rdma->sc_ctxt_used) != 0);
- WARN_ON(atomic_read(&rdma->sc_dma_used) != 0);
+ if (atomic_read(&rdma->sc_ctxt_used) != 0)
+ pr_err("svcrdma: ctxt still in use? (%d)\n",
+ atomic_read(&rdma->sc_ctxt_used));
+ if (atomic_read(&rdma->sc_dma_used) != 0)
+ pr_err("svcrdma: dma still in use? (%d)\n",
+ atomic_read(&rdma->sc_dma_used));
/* De-allocate fastreg mr */
rdma_dealloc_frmr_q(rdma);
@@ -1254,7 +1266,6 @@ int svc_rdma_send(struct svcxprt_rdma *xprt, struct ib_send_wr *wr)
if (test_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags))
return -ENOTCONN;
- BUG_ON(wr->send_flags != IB_SEND_SIGNALED);
wr_count = 1;
for (n_wr = wr->next; n_wr; n_wr = n_wr->next)
wr_count++;
diff --git a/net/sunrpc/xprtsock.c b/net/sunrpc/xprtsock.c
index 87ce7e8bb8dc..66891e32c5e3 100644
--- a/net/sunrpc/xprtsock.c
+++ b/net/sunrpc/xprtsock.c
@@ -63,6 +63,8 @@ static unsigned int xprt_max_tcp_slot_table_entries = RPC_MAX_SLOT_TABLE;
static unsigned int xprt_min_resvport = RPC_DEF_MIN_RESVPORT;
static unsigned int xprt_max_resvport = RPC_DEF_MAX_RESVPORT;
+#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
+
#define XS_TCP_LINGER_TO (15U * HZ)
static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
@@ -75,8 +77,6 @@ static unsigned int xs_tcp_fin_timeout __read_mostly = XS_TCP_LINGER_TO;
* someone else's file names!
*/
-#if IS_ENABLED(CONFIG_SUNRPC_DEBUG)
-
static unsigned int min_slot_table_size = RPC_MIN_SLOT_TABLE;
static unsigned int max_slot_table_size = RPC_MAX_SLOT_TABLE;
static unsigned int max_tcp_slot_table_limit = RPC_MAX_SLOT_TABLE_LIMIT;
@@ -627,7 +627,7 @@ process_status:
* @xprt: transport
*
* Initiates a graceful shutdown of the TCP socket by calling the
- * equivalent of shutdown(SHUT_WR);
+ * equivalent of shutdown(SHUT_RDWR);
*/
static void xs_tcp_shutdown(struct rpc_xprt *xprt)
{
@@ -635,7 +635,7 @@ static void xs_tcp_shutdown(struct rpc_xprt *xprt)
struct socket *sock = transport->sock;
if (sock != NULL) {
- kernel_sock_shutdown(sock, SHUT_WR);
+ kernel_sock_shutdown(sock, SHUT_RDWR);
trace_rpc_socket_shutdown(xprt, sock);
}
}
@@ -718,9 +718,9 @@ static int xs_tcp_send_request(struct rpc_task *task)
dprintk("RPC: sendmsg returned unrecognized error %d\n",
-status);
case -ECONNRESET:
- xs_tcp_shutdown(xprt);
case -ECONNREFUSED:
case -ENOTCONN:
+ case -EADDRINUSE:
case -EPIPE:
clear_bit(SOCK_ASYNC_NOSPACE, &transport->sock->flags);
}
@@ -773,6 +773,21 @@ static void xs_restore_old_callbacks(struct sock_xprt *transport, struct sock *s
sk->sk_error_report = transport->old_error_report;
}
+static void xs_sock_reset_connection_flags(struct rpc_xprt *xprt)
+{
+ smp_mb__before_atomic();
+ clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
+ clear_bit(XPRT_CLOSING, &xprt->state);
+ smp_mb__after_atomic();
+}
+
+static void xs_sock_mark_closed(struct rpc_xprt *xprt)
+{
+ xs_sock_reset_connection_flags(xprt);
+ /* Mark transport as closed and wake up all pending tasks */
+ xprt_disconnect_done(xprt);
+}
+
/**
* xs_error_report - callback to handle TCP socket state errors
* @sk: socket
@@ -792,11 +807,12 @@ static void xs_error_report(struct sock *sk)
err = -sk->sk_err;
if (err == 0)
goto out;
+ /* Is this a reset event? */
+ if (sk->sk_state == TCP_CLOSE)
+ xs_sock_mark_closed(xprt);
dprintk("RPC: xs_error_report client %p, error=%d...\n",
xprt, -err);
trace_rpc_socket_error(xprt, sk->sk_socket, err);
- if (test_bit(XPRT_CONNECTION_REUSE, &xprt->state))
- goto out;
xprt_wake_pending_tasks(xprt, err);
out:
read_unlock_bh(&sk->sk_callback_lock);
@@ -806,12 +822,11 @@ static void xs_reset_transport(struct sock_xprt *transport)
{
struct socket *sock = transport->sock;
struct sock *sk = transport->inet;
+ struct rpc_xprt *xprt = &transport->xprt;
if (sk == NULL)
return;
- transport->srcport = 0;
-
write_lock_bh(&sk->sk_callback_lock);
transport->inet = NULL;
transport->sock = NULL;
@@ -820,8 +835,9 @@ static void xs_reset_transport(struct sock_xprt *transport)
xs_restore_old_callbacks(transport, sk);
write_unlock_bh(&sk->sk_callback_lock);
+ xs_sock_reset_connection_flags(xprt);
- trace_rpc_socket_close(&transport->xprt, sock);
+ trace_rpc_socket_close(xprt, sock);
sock_release(sock);
}
@@ -841,27 +857,12 @@ static void xs_close(struct rpc_xprt *xprt)
dprintk("RPC: xs_close xprt %p\n", xprt);
- cancel_delayed_work_sync(&transport->connect_worker);
-
xs_reset_transport(transport);
xprt->reestablish_timeout = 0;
- smp_mb__before_atomic();
- clear_bit(XPRT_CONNECTION_ABORT, &xprt->state);
- clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
- clear_bit(XPRT_CLOSING, &xprt->state);
- smp_mb__after_atomic();
xprt_disconnect_done(xprt);
}
-static void xs_tcp_close(struct rpc_xprt *xprt)
-{
- if (test_and_clear_bit(XPRT_CONNECTION_CLOSE, &xprt->state))
- xs_close(xprt);
- else
- xs_tcp_shutdown(xprt);
-}
-
static void xs_xprt_free(struct rpc_xprt *xprt)
{
xs_free_peer_addresses(xprt);
@@ -1032,7 +1033,6 @@ static void xs_udp_data_ready(struct sock *sk)
*/
static void xs_tcp_force_close(struct rpc_xprt *xprt)
{
- set_bit(XPRT_CONNECTION_CLOSE, &xprt->state);
xprt_force_disconnect(xprt);
}
@@ -1425,54 +1425,6 @@ out:
read_unlock_bh(&sk->sk_callback_lock);
}
-/*
- * Do the equivalent of linger/linger2 handling for dealing with
- * broken servers that don't close the socket in a timely
- * fashion
- */
-static void xs_tcp_schedule_linger_timeout(struct rpc_xprt *xprt,
- unsigned long timeout)
-{
- struct sock_xprt *transport;
-
- if (xprt_test_and_set_connecting(xprt))
- return;
- set_bit(XPRT_CONNECTION_ABORT, &xprt->state);
- transport = container_of(xprt, struct sock_xprt, xprt);
- queue_delayed_work(rpciod_workqueue, &transport->connect_worker,
- timeout);
-}
-
-static void xs_tcp_cancel_linger_timeout(struct rpc_xprt *xprt)
-{
- struct sock_xprt *transport;
-
- transport = container_of(xprt, struct sock_xprt, xprt);
-
- if (!test_bit(XPRT_CONNECTION_ABORT, &xprt->state) ||
- !cancel_delayed_work(&transport->connect_worker))
- return;
- clear_bit(XPRT_CONNECTION_ABORT, &xprt->state);
- xprt_clear_connecting(xprt);
-}
-
-static void xs_sock_reset_connection_flags(struct rpc_xprt *xprt)
-{
- smp_mb__before_atomic();
- clear_bit(XPRT_CONNECTION_ABORT, &xprt->state);
- clear_bit(XPRT_CONNECTION_CLOSE, &xprt->state);
- clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
- clear_bit(XPRT_CLOSING, &xprt->state);
- smp_mb__after_atomic();
-}
-
-static void xs_sock_mark_closed(struct rpc_xprt *xprt)
-{
- xs_sock_reset_connection_flags(xprt);
- /* Mark transport as closed and wake up all pending tasks */
- xprt_disconnect_done(xprt);
-}
-
/**
* xs_tcp_state_change - callback to handle TCP socket state changes
* @sk: socket whose state has changed
@@ -1521,7 +1473,6 @@ static void xs_tcp_state_change(struct sock *sk)
clear_bit(XPRT_CONNECTED, &xprt->state);
clear_bit(XPRT_CLOSE_WAIT, &xprt->state);
smp_mb__after_atomic();
- xs_tcp_schedule_linger_timeout(xprt, xs_tcp_fin_timeout);
break;
case TCP_CLOSE_WAIT:
/* The server initiated a shutdown of the socket */
@@ -1538,13 +1489,11 @@ static void xs_tcp_state_change(struct sock *sk)
break;
case TCP_LAST_ACK:
set_bit(XPRT_CLOSING, &xprt->state);
- xs_tcp_schedule_linger_timeout(xprt, xs_tcp_fin_timeout);
smp_mb__before_atomic();
clear_bit(XPRT_CONNECTED, &xprt->state);
smp_mb__after_atomic();
break;
case TCP_CLOSE:
- xs_tcp_cancel_linger_timeout(xprt);
xs_sock_mark_closed(xprt);
}
out:
@@ -1667,6 +1616,40 @@ static unsigned short xs_get_random_port(void)
}
/**
+ * xs_set_reuseaddr_port - set the socket's port and address reuse options
+ * @sock: socket
+ *
+ * Note that this function has to be called on all sockets that share the
+ * same port, and it must be called before binding.
+ */
+static void xs_sock_set_reuseport(struct socket *sock)
+{
+ int opt = 1;
+
+ kernel_setsockopt(sock, SOL_SOCKET, SO_REUSEPORT,
+ (char *)&opt, sizeof(opt));
+}
+
+static unsigned short xs_sock_getport(struct socket *sock)
+{
+ struct sockaddr_storage buf;
+ int buflen;
+ unsigned short port = 0;
+
+ if (kernel_getsockname(sock, (struct sockaddr *)&buf, &buflen) < 0)
+ goto out;
+ switch (buf.ss_family) {
+ case AF_INET6:
+ port = ntohs(((struct sockaddr_in6 *)&buf)->sin6_port);
+ break;
+ case AF_INET:
+ port = ntohs(((struct sockaddr_in *)&buf)->sin_port);
+ }
+out:
+ return port;
+}
+
+/**
* xs_set_port - reset the port number in the remote endpoint address
* @xprt: generic transport
* @port: new port number
@@ -1680,6 +1663,12 @@ static void xs_set_port(struct rpc_xprt *xprt, unsigned short port)
xs_update_peer_port(xprt);
}
+static void xs_set_srcport(struct sock_xprt *transport, struct socket *sock)
+{
+ if (transport->srcport == 0)
+ transport->srcport = xs_sock_getport(sock);
+}
+
static unsigned short xs_get_srcport(struct sock_xprt *transport)
{
unsigned short port = transport->srcport;
@@ -1833,7 +1822,8 @@ static void xs_dummy_setup_socket(struct work_struct *work)
}
static struct socket *xs_create_sock(struct rpc_xprt *xprt,
- struct sock_xprt *transport, int family, int type, int protocol)
+ struct sock_xprt *transport, int family, int type,
+ int protocol, bool reuseport)
{
struct socket *sock;
int err;
@@ -1846,6 +1836,9 @@ static struct socket *xs_create_sock(struct rpc_xprt *xprt,
}
xs_reclassify_socket(family, sock);
+ if (reuseport)
+ xs_sock_set_reuseport(sock);
+
err = xs_bind(transport, sock);
if (err) {
sock_release(sock);
@@ -1903,7 +1896,6 @@ static int xs_local_setup_socket(struct sock_xprt *transport)
struct socket *sock;
int status = -EIO;
- clear_bit(XPRT_CONNECTION_ABORT, &xprt->state);
status = __sock_create(xprt->xprt_net, AF_LOCAL,
SOCK_STREAM, 0, &sock, 1);
if (status < 0) {
@@ -2044,10 +2036,9 @@ static void xs_udp_setup_socket(struct work_struct *work)
struct socket *sock = transport->sock;
int status = -EIO;
- /* Start by resetting any existing state */
- xs_reset_transport(transport);
sock = xs_create_sock(xprt, transport,
- xs_addr(xprt)->sa_family, SOCK_DGRAM, IPPROTO_UDP);
+ xs_addr(xprt)->sa_family, SOCK_DGRAM,
+ IPPROTO_UDP, false);
if (IS_ERR(sock))
goto out;
@@ -2061,61 +2052,11 @@ static void xs_udp_setup_socket(struct work_struct *work)
trace_rpc_socket_connect(xprt, sock, 0);
status = 0;
out:
+ xprt_unlock_connect(xprt, transport);
xprt_clear_connecting(xprt);
xprt_wake_pending_tasks(xprt, status);
}
-/*
- * We need to preserve the port number so the reply cache on the server can
- * find our cached RPC replies when we get around to reconnecting.
- */
-static void xs_abort_connection(struct sock_xprt *transport)
-{
- int result;
- struct sockaddr any;
-
- dprintk("RPC: disconnecting xprt %p to reuse port\n", transport);
-
- /*
- * Disconnect the transport socket by doing a connect operation
- * with AF_UNSPEC. This should return immediately...
- */
- memset(&any, 0, sizeof(any));
- any.sa_family = AF_UNSPEC;
- result = kernel_connect(transport->sock, &any, sizeof(any), 0);
- trace_rpc_socket_reset_connection(&transport->xprt,
- transport->sock, result);
- if (!result)
- xs_sock_reset_connection_flags(&transport->xprt);
- dprintk("RPC: AF_UNSPEC connect return code %d\n", result);
-}
-
-static void xs_tcp_reuse_connection(struct sock_xprt *transport)
-{
- unsigned int state = transport->inet->sk_state;
-
- if (state == TCP_CLOSE && transport->sock->state == SS_UNCONNECTED) {
- /* we don't need to abort the connection if the socket
- * hasn't undergone a shutdown
- */
- if (transport->inet->sk_shutdown == 0)
- return;
- dprintk("RPC: %s: TCP_CLOSEd and sk_shutdown set to %d\n",
- __func__, transport->inet->sk_shutdown);
- }
- if ((1 << state) & (TCPF_ESTABLISHED|TCPF_SYN_SENT)) {
- /* we don't need to abort the connection if the socket
- * hasn't undergone a shutdown
- */
- if (transport->inet->sk_shutdown == 0)
- return;
- dprintk("RPC: %s: ESTABLISHED/SYN_SENT "
- "sk_shutdown set to %d\n",
- __func__, transport->inet->sk_shutdown);
- }
- xs_abort_connection(transport);
-}
-
static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
{
struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
@@ -2149,9 +2090,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
sk->sk_allocation = GFP_ATOMIC;
/* socket options */
- sk->sk_userlocks |= SOCK_BINDPORT_LOCK;
sock_reset_flag(sk, SOCK_LINGER);
- tcp_sk(sk)->linger2 = 0;
tcp_sk(sk)->nonagle |= TCP_NAGLE_OFF;
xprt_clear_connected(xprt);
@@ -2174,6 +2113,7 @@ static int xs_tcp_finish_connecting(struct rpc_xprt *xprt, struct socket *sock)
ret = kernel_connect(sock, xs_addr(xprt), xprt->addrlen, O_NONBLOCK);
switch (ret) {
case 0:
+ xs_set_srcport(transport, sock);
case -EINPROGRESS:
/* SYN_SENT! */
if (xprt->reestablish_timeout < XS_TCP_INIT_REEST_TO)
@@ -2200,25 +2140,13 @@ static void xs_tcp_setup_socket(struct work_struct *work)
int status = -EIO;
if (!sock) {
- clear_bit(XPRT_CONNECTION_ABORT, &xprt->state);
sock = xs_create_sock(xprt, transport,
- xs_addr(xprt)->sa_family, SOCK_STREAM, IPPROTO_TCP);
+ xs_addr(xprt)->sa_family, SOCK_STREAM,
+ IPPROTO_TCP, true);
if (IS_ERR(sock)) {
status = PTR_ERR(sock);
goto out;
}
- } else {
- int abort_and_exit;
-
- abort_and_exit = test_and_clear_bit(XPRT_CONNECTION_ABORT,
- &xprt->state);
- /* "close" the socket, preserving the local port */
- set_bit(XPRT_CONNECTION_REUSE, &xprt->state);
- xs_tcp_reuse_connection(transport);
- clear_bit(XPRT_CONNECTION_REUSE, &xprt->state);
-
- if (abort_and_exit)
- goto out_eagain;
}
dprintk("RPC: worker connecting xprt %p via %s to "
@@ -2245,6 +2173,7 @@ static void xs_tcp_setup_socket(struct work_struct *work)
case 0:
case -EINPROGRESS:
case -EALREADY:
+ xprt_unlock_connect(xprt, transport);
xprt_clear_connecting(xprt);
return;
case -EINVAL:
@@ -2254,13 +2183,15 @@ static void xs_tcp_setup_socket(struct work_struct *work)
case -ECONNREFUSED:
case -ECONNRESET:
case -ENETUNREACH:
+ case -EADDRINUSE:
case -ENOBUFS:
/* retry with existing socket, after a delay */
+ xs_tcp_force_close(xprt);
goto out;
}
-out_eagain:
status = -EAGAIN;
out:
+ xprt_unlock_connect(xprt, transport);
xprt_clear_connecting(xprt);
xprt_wake_pending_tasks(xprt, status);
}
@@ -2283,6 +2214,11 @@ static void xs_connect(struct rpc_xprt *xprt, struct rpc_task *task)
{
struct sock_xprt *transport = container_of(xprt, struct sock_xprt, xprt);
+ WARN_ON_ONCE(!xprt_lock_connect(xprt, task, transport));
+
+ /* Start by resetting any existing state */
+ xs_reset_transport(transport);
+
if (transport->sock != NULL && !RPC_IS_SOFTCONN(task)) {
dprintk("RPC: xs_connect delayed xprt %p for %lu "
"seconds\n",
@@ -2559,7 +2495,7 @@ static struct rpc_xprt_ops xs_tcp_ops = {
.buf_free = rpc_free,
.send_request = xs_tcp_send_request,
.set_retrans_timeout = xprt_set_retrans_timeout_def,
- .close = xs_tcp_close,
+ .close = xs_tcp_shutdown,
.destroy = xs_destroy,
.print_stats = xs_tcp_print_stats,
};
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-26 16:47:16 +0000