Merge branch 'for-2.6.26' of git://linux-nfs.org/~bfields/linux

* 'for-2.6.26' of git://linux-nfs.org/~bfields/linux: (25 commits)
  svcrdma: Verify read-list fits within RPCSVC_MAXPAGES
  svcrdma: Change svc_rdma_send_error return type to void
  svcrdma: Copy transport address and arm CQ before calling rdma_accept
  svcrdma: Set rqstp transport address in rdma_read_complete function
  svcrdma: Use ib verbs version of dma_unmap
  svcrdma: Cleanup queued, but unprocessed I/O in svc_rdma_free
  svcrdma: Move the QP and cm_id destruction to svc_rdma_free
  svcrdma: Add reference for each SQ/RQ WR
  svcrdma: Move destroy to kernel thread
  svcrdma: Shrink scope of spinlock on RQ CQ
  svcrdma: Use standard Linux lists for context cache
  svcrdma: Simplify RDMA_READ deferral buffer management
  svcrdma: Remove unused READ_DONE context flags bit
  svcrdma: Return error from rdma_read_xdr so caller knows to free context
  svcrdma: Fix error handling during listening endpoint creation
  svcrdma: Free context on post_recv error in send_reply
  svcrdma: Free context on ib_post_recv error
  svcrdma: Add put of connection ESTABLISHED reference in rdma_cma_handler
  svcrdma: Fix return value in svc_rdma_send
  svcrdma: Fix race with dto_tasklet in svc_rdma_send
  ...

fs/nfsd/nfs4callback.c

@@ -419,9 +419,9 @@ static int do_probe_callback(void *data)
 out_release_client:
 	rpc_shutdown_client(client);
 out_err:
-	put_nfs4_client(clp);
 	dprintk("NFSD: warning: no callback path to client %.*s\n",
 		(int)clp->cl_name.len, clp->cl_name.data);
+	put_nfs4_client(clp);
 	return status;
 }
 

include/linux/sunrpc/svc_rdma.h

@@ -71,7 +71,8 @@ extern atomic_t rdma_stat_sq_prod;
  * completes.
  */
 struct svc_rdma_op_ctxt {
-	struct svc_rdma_op_ctxt *next;
+	struct svc_rdma_op_ctxt *read_hdr;
+	struct list_head free_list;
 	struct xdr_buf arg;
 	struct list_head dto_q;
 	enum ib_wr_opcode wr_op;
@@ -85,7 +86,6 @@ struct svc_rdma_op_ctxt {
 	struct page *pages[RPCSVC_MAXPAGES];
 };
 
-#define RDMACTXT_F_READ_DONE	1
 #define RDMACTXT_F_LAST_CTXT	2
 
 struct svcxprt_rdma {
@@ -104,7 +104,8 @@ struct svcxprt_rdma {
 
 	struct ib_pd         *sc_pd;
 
-	struct svc_rdma_op_ctxt  *sc_ctxt_head;
+	atomic_t	     sc_ctxt_used;
+	struct list_head     sc_ctxt_free;
 	int		     sc_ctxt_cnt;
 	int		     sc_ctxt_bump;
 	int		     sc_ctxt_max;
@@ -123,6 +124,7 @@ struct svcxprt_rdma {
 	struct list_head     sc_dto_q;		/* DTO tasklet I/O pending Q */
 	struct list_head     sc_read_complete_q;
 	spinlock_t           sc_read_complete_lock;
+	struct work_struct   sc_work;
 };
 /* sc_flags */
 #define RDMAXPRT_RQ_PENDING	1
@@ -164,8 +166,8 @@ extern int svc_rdma_sendto(struct svc_rqst *);
 
 /* svc_rdma_transport.c */
 extern int svc_rdma_send(struct svcxprt_rdma *, struct ib_send_wr *);
-extern int svc_rdma_send_error(struct svcxprt_rdma *, struct rpcrdma_msg *,
-			       enum rpcrdma_errcode);
+extern void svc_rdma_send_error(struct svcxprt_rdma *, struct rpcrdma_msg *,
+				enum rpcrdma_errcode);
 struct page *svc_rdma_get_page(void);
 extern int svc_rdma_post_recv(struct svcxprt_rdma *);
 extern int svc_rdma_create_listen(struct svc_serv *, int, struct sockaddr *);

net/sunrpc/svc_xprt.c

@@ -6,30 +6,9 @@
 
 #include <linux/sched.h>
 #include <linux/errno.h>
-#include <linux/fcntl.h>
-#include <linux/net.h>
-#include <linux/in.h>
-#include <linux/inet.h>
-#include <linux/udp.h>
-#include <linux/tcp.h>
-#include <linux/unistd.h>
-#include <linux/slab.h>
-#include <linux/netdevice.h>
-#include <linux/skbuff.h>
-#include <linux/file.h>
 #include <linux/freezer.h>
 #include <linux/kthread.h>
 #include <net/sock.h>
-#include <net/checksum.h>
-#include <net/ip.h>
-#include <net/ipv6.h>
-#include <net/tcp_states.h>
-#include <linux/uaccess.h>
-#include <asm/ioctls.h>
-
-#include <linux/sunrpc/types.h>
-#include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/xdr.h>
 #include <linux/sunrpc/stats.h>
 #include <linux/sunrpc/svc_xprt.h>
 
@@ -296,8 +275,6 @@ void svc_xprt_enqueue(struct svc_xprt *xprt)
 	if (!(xprt->xpt_flags &
 	      ((1<<XPT_CONN)|(1<<XPT_DATA)|(1<<XPT_CLOSE)|(1<<XPT_DEFERRED))))
 		return;
-	if (test_bit(XPT_DEAD, &xprt->xpt_flags))
-		return;
 
 	cpu = get_cpu();
 	pool = svc_pool_for_cpu(xprt->xpt_server, cpu);

net/sunrpc/svcauth_unix.c

@@ -278,7 +278,7 @@ static int ip_map_show(struct seq_file *m,
 		dom = im->m_client->h.name;
 
 	if (ipv6_addr_v4mapped(&addr)) {
-		seq_printf(m, "%s" NIPQUAD_FMT "%s\n",
+		seq_printf(m, "%s " NIPQUAD_FMT " %s\n",
 			im->m_class,
 			ntohl(addr.s6_addr32[3]) >> 24 & 0xff,
 			ntohl(addr.s6_addr32[3]) >> 16 & 0xff,
@@ -286,7 +286,7 @@ static int ip_map_show(struct seq_file *m,
 			ntohl(addr.s6_addr32[3]) >>  0 & 0xff,
 			dom);
 	} else {
-		seq_printf(m, "%s" NIP6_FMT "%s\n",
+		seq_printf(m, "%s " NIP6_FMT " %s\n",
 			im->m_class, NIP6(addr), dom);
 	}
 	return 0;

net/sunrpc/xprtrdma/svc_rdma_recvfrom.c

@@ -260,11 +260,16 @@ static int rdma_read_max_sge(struct svcxprt_rdma *xprt, int sge_count)
  * On our side, we need to read into a pagelist. The first page immediately
  * follows the RPC header.
  *
- * This function returns 1 to indicate success. The data is not yet in
+ * This function returns:
+ * 0 - No error and no read-list found.
+ *
+ * 1 - Successful read-list processing. The data is not yet in
  * the pagelist and therefore the RPC request must be deferred. The
  * I/O completion will enqueue the transport again and
  * svc_rdma_recvfrom will complete the request.
  *
+ * <0 - Error processing/posting read-list.
+ *
  * NOTE: The ctxt must not be touched after the last WR has been posted
  * because the I/O completion processing may occur on another
  * processor and free / modify the context. Ne touche pas!
@@ -284,7 +289,6 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 	u64 sgl_offset;
 	struct rpcrdma_read_chunk *ch;
 	struct svc_rdma_op_ctxt *ctxt = NULL;
-	struct svc_rdma_op_ctxt *head;
 	struct svc_rdma_op_ctxt *tmp_sge_ctxt;
 	struct svc_rdma_op_ctxt *tmp_ch_ctxt;
 	struct chunk_sge *ch_sge_ary;
@@ -302,25 +306,19 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 	ch_sge_ary = (struct chunk_sge *)tmp_ch_ctxt->sge;
 
 	svc_rdma_rcl_chunk_counts(ch, &ch_count, &byte_count);
+	if (ch_count > RPCSVC_MAXPAGES)
+		return -EINVAL;
 	sge_count = rdma_rcl_to_sge(xprt, rqstp, hdr_ctxt, rmsgp,
 				    sge, ch_sge_ary,
 				    ch_count, byte_count);
-	head = svc_rdma_get_context(xprt);
 	sgl_offset = 0;
 	ch_no = 0;
 
 	for (ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
 	     ch->rc_discrim != 0; ch++, ch_no++) {
 next_sge:
-		if (!ctxt)
-			ctxt = head;
-		else {
-			ctxt->next = svc_rdma_get_context(xprt);
-			ctxt = ctxt->next;
-		}
-		ctxt->next = NULL;
+		ctxt = svc_rdma_get_context(xprt);
 		ctxt->direction = DMA_FROM_DEVICE;
-		clear_bit(RDMACTXT_F_READ_DONE, &ctxt->flags);
 		clear_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
 
 		/* Prepare READ WR */
@@ -347,20 +345,15 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 			 * the client and the RPC needs to be enqueued.
 			 */
 			set_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
-			ctxt->next = hdr_ctxt;
-			hdr_ctxt->next = head;
+			ctxt->read_hdr = hdr_ctxt;
 		}
 		/* Post the read */
 		err = svc_rdma_send(xprt, &read_wr);
 		if (err) {
-			printk(KERN_ERR "svcrdma: Error posting send = %d\n",
+			printk(KERN_ERR "svcrdma: Error %d posting RDMA_READ\n",
 			       err);
-			/*
-			 * Break the circular list so free knows when
-			 * to stop if the error happened to occur on
-			 * the last read
-			 */
-			ctxt->next = NULL;
+			set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags);
+			svc_rdma_put_context(ctxt, 0);
 			goto out;
 		}
 		atomic_inc(&rdma_stat_read);
@@ -371,7 +364,7 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 			goto next_sge;
 		}
 		sgl_offset = 0;
-		err = 0;
+		err = 1;
 	}
 
  out:
@@ -389,25 +382,12 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 	while (rqstp->rq_resused)
 		rqstp->rq_respages[--rqstp->rq_resused] = NULL;
 
-	if (err) {
-		printk(KERN_ERR "svcrdma : RDMA_READ error = %d\n", err);
-		set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags);
-		/* Free the linked list of read contexts */
-		while (head != NULL) {
-			ctxt = head->next;
-			svc_rdma_put_context(head, 1);
-			head = ctxt;
-		}
-		return 0;
-	}
-
-	return 1;
+	return err;
 }
 
 static int rdma_read_complete(struct svc_rqst *rqstp,
-			      struct svc_rdma_op_ctxt *data)
+			      struct svc_rdma_op_ctxt *head)
 {
-	struct svc_rdma_op_ctxt *head = data->next;
 	int page_no;
 	int ret;
 
@@ -433,21 +413,12 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
 	rqstp->rq_arg.len = head->arg.len;
 	rqstp->rq_arg.buflen = head->arg.buflen;
 
+	/* Free the context */
+	svc_rdma_put_context(head, 0);
+
 	/* XXX: What should this be? */
 	rqstp->rq_prot = IPPROTO_MAX;
-
-	/*
-	 * Free the contexts we used to build the RDMA_READ. We have
-	 * to be careful here because the context list uses the same
-	 * next pointer used to chain the contexts associated with the
-	 * RDMA_READ
-	 */
-	data->next = NULL;	/* terminate circular list */
-	do {
-		data = head->next;
-		svc_rdma_put_context(head, 0);
-		head = data;
-	} while (head != NULL);
+	svc_xprt_copy_addrs(rqstp, rqstp->rq_xprt);
 
 	ret = rqstp->rq_arg.head[0].iov_len
 		+ rqstp->rq_arg.page_len
@@ -457,8 +428,6 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
 		ret, rqstp->rq_arg.len,	rqstp->rq_arg.head[0].iov_base,
 		rqstp->rq_arg.head[0].iov_len);
 
-	/* Indicate that we've consumed an RQ credit */
-	rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
 	svc_xprt_received(rqstp->rq_xprt);
 	return ret;
 }
@@ -480,13 +449,6 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 
 	dprintk("svcrdma: rqstp=%p\n", rqstp);
 
-	/*
-	 * The rq_xprt_ctxt indicates if we've consumed an RQ credit
-	 * or not. It is used in the rdma xpo_release_rqst function to
-	 * determine whether or not to return an RQ WQE to the RQ.
-	 */
-	rqstp->rq_xprt_ctxt = NULL;
-
 	spin_lock_bh(&rdma_xprt->sc_read_complete_lock);
 	if (!list_empty(&rdma_xprt->sc_read_complete_q)) {
 		ctxt = list_entry(rdma_xprt->sc_read_complete_q.next,
@@ -537,21 +499,22 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 	/* If the request is invalid, reply with an error */
 	if (len < 0) {
 		if (len == -ENOSYS)
-			(void)svc_rdma_send_error(rdma_xprt, rmsgp, ERR_VERS);
+			svc_rdma_send_error(rdma_xprt, rmsgp, ERR_VERS);
 		goto close_out;
 	}
 
-	/* Read read-list data. If we would need to wait, defer
-	 * it. Not that in this case, we don't return the RQ credit
-	 * until after the read completes.
-	 */
-	if (rdma_read_xdr(rdma_xprt, rmsgp, rqstp, ctxt)) {
+	/* Read read-list data. */
+	ret = rdma_read_xdr(rdma_xprt, rmsgp, rqstp, ctxt);
+	if (ret > 0) {
+		/* read-list posted, defer until data received from client. */
 		svc_xprt_received(xprt);
 		return 0;
 	}
-
-	/* Indicate we've consumed an RQ credit */
-	rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
+	if (ret < 0) {
+		/* Post of read-list failed, free context. */
+		svc_rdma_put_context(ctxt, 1);
+		return 0;
+	}
 
 	ret = rqstp->rq_arg.head[0].iov_len
 		+ rqstp->rq_arg.page_len
@@ -569,11 +532,8 @@ int svc_rdma_recvfrom(struct svc_rqst *rqstp)
 	return ret;
 
  close_out:
-	if (ctxt) {
+	if (ctxt)
 		svc_rdma_put_context(ctxt, 1);
-		/* Indicate we've consumed an RQ credit */
-		rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
-	}
 	dprintk("svcrdma: transport %p is closing\n", xprt);
 	/*
 	 * Set the close bit and enqueue it. svc_recv will see the

net/sunrpc/xprtrdma/svc_rdma_sendto.c

@@ -389,6 +389,17 @@ static int send_reply(struct svcxprt_rdma *rdma,
 	int page_no;
 	int ret;
 
+	/* Post a recv buffer to handle another request. */
+	ret = svc_rdma_post_recv(rdma);
+	if (ret) {
+		printk(KERN_INFO
+		       "svcrdma: could not post a receive buffer, err=%d."
+		       "Closing transport %p.\n", ret, rdma);
+		set_bit(XPT_CLOSE, &rdma->sc_xprt.xpt_flags);
+		svc_rdma_put_context(ctxt, 0);
+		return -ENOTCONN;
+	}
+
 	/* Prepare the context */
 	ctxt->pages[0] = page;
 	ctxt->count = 1;