Merge branch 'for-3.15' of git://linux-nfs.org/~bfields/linux

Pull nfsd updates from Bruce Fields:
 "Highlights:
   - server-side nfs/rdma fixes from Jeff Layton and Tom Tucker
   - xdr fixes (a larger xdr rewrite has been posted but I decided it
     would be better to queue it up for 3.16).
   - miscellaneous fixes and cleanup from all over (thanks especially to
     Kinglong Mee)"

* 'for-3.15' of git://linux-nfs.org/~bfields/linux: (36 commits)
  nfsd4: don't create unnecessary mask acl
  nfsd: revert v2 half of "nfsd: don't return high mode bits"
  nfsd4: fix memory leak in nfsd4_encode_fattr()
  nfsd: check passed socket's net matches NFSd superblock's one
  SUNRPC: Clear xpt_bc_xprt if xs_setup_bc_tcp failed
  NFSD/SUNRPC: Check rpc_xprt out of xs_setup_bc_tcp
  SUNRPC: New helper for creating client with rpc_xprt
  NFSD: Free backchannel xprt in bc_destroy
  NFSD: Clear wcc data between compound ops
  nfsd: Don't return NFS4ERR_STALE_STATEID for NFSv4.1+
  nfsd4: fix nfs4err_resource in 4.1 case
  nfsd4: fix setclientid encode size
  nfsd4: remove redundant check from nfsd4_check_resp_size
  nfsd4: use more generous NFS4_ACL_MAX
  nfsd4: minor nfsd4_replay_cache_entry cleanup
  nfsd4: nfsd4_replay_cache_entry should be static
  nfsd4: update comments with obsolete function name
  rpc: Allow xdr_buf_subsegment to operate in-place
  NFSD: Using free_conn free connection
  SUNRPC: fix memory leak of peer addresses in XPRT
  ...

fs/lockd/svc.c

@@ -235,6 +235,7 @@ static int make_socks(struct svc_serv *serv, struct net *net)
 	if (warned++ == 0)
 		printk(KERN_WARNING
 			"lockd_up: makesock failed, error=%d\n", err);
+	svc_shutdown_net(serv, net);
 	return err;
 }
 

fs/nfsd/acl.h

@@ -39,9 +39,13 @@ struct nfs4_acl;
 struct svc_fh;
 struct svc_rqst;
 
-/* Maximum ACL we'll accept from client; chosen (somewhat arbitrarily) to
- * fit in a page: */
-#define NFS4_ACL_MAX 170
+/*
+ * Maximum ACL we'll accept from a client; chosen (somewhat
+ * arbitrarily) so that kmalloc'ing the ACL shouldn't require a
+ * high-order allocation.  This allows 204 ACEs on x86_64:
+ */
+#define NFS4_ACL_MAX ((PAGE_SIZE - sizeof(struct nfs4_acl)) \
+			/ sizeof(struct nfs4_ace))
 
 struct nfs4_acl *nfs4_acl_new(int);
 int nfs4_acl_get_whotype(char *, u32);

fs/nfsd/nfs4acl.c

@@ -542,6 +542,9 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags)
 	 * up setting a 3-element effective posix ACL with all
 	 * permissions zero.
 	 */
+	if (!state->users->n && !state->groups->n)
+		nace = 3;
+	else /* Note we also include a MASK ACE in this case: */
 		nace = 4 + state->users->n + state->groups->n;
 	pacl = posix_acl_alloc(nace, GFP_KERNEL);
 	if (!pacl)
@@ -586,9 +589,11 @@ posix_state_to_acl(struct posix_acl_state *state, unsigned int flags)
 		add_to_mask(state, &state->groups->aces[i].perms);
 	}
 
+	if (!state->users->n && !state->groups->n) {
 		pace++;
 		pace->e_tag = ACL_MASK;
 		low_mode_from_nfs4(state->mask.allow, &pace->e_perm, flags);
+	}
 
 	pace++;
 	pace->e_tag = ACL_OTHER;

fs/nfsd/nfs4callback.c

@@ -32,6 +32,7 @@
  */
 
 #include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/xprt.h>
 #include <linux/sunrpc/svc_xprt.h>
 #include <linux/slab.h>
 #include "nfsd.h"
@@ -635,6 +636,22 @@ static struct rpc_cred *get_backchannel_cred(struct nfs4_client *clp, struct rpc
 	}
 }
 
+static struct rpc_clnt *create_backchannel_client(struct rpc_create_args *args)
+{
+	struct rpc_xprt *xprt;
+
+	if (args->protocol != XPRT_TRANSPORT_BC_TCP)
+		return rpc_create(args);
+
+	xprt = args->bc_xprt->xpt_bc_xprt;
+	if (xprt) {
+		xprt_get(xprt);
+		return rpc_create_xprt(args, xprt);
+	}
+
+	return rpc_create(args);
+}
+
 static int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn, struct nfsd4_session *ses)
 {
 	struct rpc_timeout	timeparms = {
@@ -674,7 +691,7 @@ static int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *c
 		args.authflavor = ses->se_cb_sec.flavor;
 	}
 	/* Create RPC client */
-	client = rpc_create(&args);
+	client = create_backchannel_client(&args);
 	if (IS_ERR(client)) {
 		dprintk("NFSD: couldn't create callback client: %ld\n",
 			PTR_ERR(client));

fs/nfsd/nfs4proc.c

@@ -1273,6 +1273,8 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 	struct nfsd4_op	*op;
 	struct nfsd4_operation *opdesc;
 	struct nfsd4_compound_state *cstate = &resp->cstate;
+	struct svc_fh *current_fh = &cstate->current_fh;
+	struct svc_fh *save_fh = &cstate->save_fh;
 	int		slack_bytes;
 	u32		plen = 0;
 	__be32		status;
@@ -1288,11 +1290,11 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 	resp->tag = args->tag;
 	resp->opcnt = 0;
 	resp->rqstp = rqstp;
-	resp->cstate.minorversion = args->minorversion;
-	resp->cstate.replay_owner = NULL;
-	resp->cstate.session = NULL;
-	fh_init(&resp->cstate.current_fh, NFS4_FHSIZE);
-	fh_init(&resp->cstate.save_fh, NFS4_FHSIZE);
+	cstate->minorversion = args->minorversion;
+	cstate->replay_owner = NULL;
+	cstate->session = NULL;
+	fh_init(current_fh, NFS4_FHSIZE);
+	fh_init(save_fh, NFS4_FHSIZE);
 	/*
 	 * Don't use the deferral mechanism for NFSv4; compounds make it
 	 * too hard to avoid non-idempotency problems.
@@ -1345,20 +1347,28 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 
 		opdesc = OPDESC(op);
 
-		if (!cstate->current_fh.fh_dentry) {
+		if (!current_fh->fh_dentry) {
 			if (!(opdesc->op_flags & ALLOWED_WITHOUT_FH)) {
 				op->status = nfserr_nofilehandle;
 				goto encode_op;
 			}
-		} else if (cstate->current_fh.fh_export->ex_fslocs.migrated &&
+		} else if (current_fh->fh_export->ex_fslocs.migrated &&
 			  !(opdesc->op_flags & ALLOWED_ON_ABSENT_FS)) {
 			op->status = nfserr_moved;
 			goto encode_op;
 		}
 
+		fh_clear_wcc(current_fh);
+
 		/* If op is non-idempotent */
 		if (opdesc->op_flags & OP_MODIFIES_SOMETHING) {
 			plen = opdesc->op_rsize_bop(rqstp, op);
+			/*
+			 * If there's still another operation, make sure
+			 * we'll have space to at least encode an error:
+			 */
+			if (resp->opcnt < args->opcnt)
+				plen += COMPOUND_ERR_SLACK_SPACE;
 			op->status = nfsd4_check_resp_size(resp, plen);
 		}
 
@@ -1377,12 +1387,12 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 				clear_current_stateid(cstate);
 
 			if (need_wrongsec_check(rqstp))
-				op->status = check_nfsd_access(cstate->current_fh.fh_export, rqstp);
+				op->status = check_nfsd_access(current_fh->fh_export, rqstp);
 		}
 
 encode_op:
 		/* Only from SEQUENCE */
-		if (resp->cstate.status == nfserr_replay_cache) {
+		if (cstate->status == nfserr_replay_cache) {
 			dprintk("%s NFS4.1 replay from cache\n", __func__);
 			status = op->status;
 			goto out;
@@ -1411,10 +1421,10 @@ nfsd4_proc_compound(struct svc_rqst *rqstp,
 		nfsd4_increment_op_stats(op->opnum);
 	}
 
-	resp->cstate.status = status;
-	fh_put(&resp->cstate.current_fh);
-	fh_put(&resp->cstate.save_fh);
-	BUG_ON(resp->cstate.replay_owner);
+	cstate->status = status;
+	fh_put(current_fh);
+	fh_put(save_fh);
+	BUG_ON(cstate->replay_owner);
 out:
 	/* Reset deferral mechanism for RPC deferrals */
 	rqstp->rq_usedeferral = 1;
@@ -1523,7 +1533,8 @@ static inline u32 nfsd4_setattr_rsize(struct svc_rqst *rqstp, struct nfsd4_op *o
 
 static inline u32 nfsd4_setclientid_rsize(struct svc_rqst *rqstp, struct nfsd4_op *op)
 {
-	return (op_encode_hdr_size + 2 + 1024) * sizeof(__be32);
+	return (op_encode_hdr_size + 2 + XDR_QUADLEN(NFS4_VERIFIER_SIZE)) *
+								sizeof(__be32);
 }
 
 static inline u32 nfsd4_write_rsize(struct svc_rqst *rqstp, struct nfsd4_op *op)

fs/nfsd/nfs4state.c

@@ -1538,7 +1538,7 @@ gen_callback(struct nfs4_client *clp, struct nfsd4_setclientid *se, struct svc_r
 }
 
 /*
- * Cache a reply. nfsd4_check_drc_limit() has bounded the cache size.
+ * Cache a reply. nfsd4_check_resp_size() has bounded the cache size.
  */
 void
 nfsd4_store_cache_entry(struct nfsd4_compoundres *resp)
@@ -1596,7 +1596,7 @@ nfsd4_enc_sequence_replay(struct nfsd4_compoundargs *args,
  * The sequence operation is not cached because we can use the slot and
  * session values.
  */
-__be32
+static __be32
 nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp,
 			 struct nfsd4_sequence *seq)
 {
@@ -1605,9 +1605,8 @@ nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp,
 
 	dprintk("--> %s slot %p\n", __func__, slot);
 
-	/* Either returns 0 or nfserr_retry_uncached */
 	status = nfsd4_enc_sequence_replay(resp->rqstp->rq_argp, resp);
-	if (status == nfserr_retry_uncached_rep)
+	if (status)
 		return status;
 
 	/* The sequence operation has been encoded, cstate->datap set. */
@@ -2287,7 +2286,8 @@ nfsd4_sequence(struct svc_rqst *rqstp,
 	if (!list_empty(&clp->cl_revoked))
 		seq->status_flags |= SEQ4_STATUS_RECALLABLE_STATE_REVOKED;
 out_no_session:
-	kfree(conn);
+	if (conn)
+		free_conn(conn);
 	spin_unlock(&nn->client_lock);
 	return status;
 out_put_session:
@@ -3627,8 +3627,11 @@ static __be32 nfsd4_lookup_stateid(stateid_t *stateid, unsigned char typemask,
 		return nfserr_bad_stateid;
 	status = lookup_clientid(&stateid->si_opaque.so_clid, sessions,
 							nn, &cl);
-	if (status == nfserr_stale_clientid)
+	if (status == nfserr_stale_clientid) {
+		if (sessions)
+			return nfserr_bad_stateid;
 		return nfserr_stale_stateid;
+	}
 	if (status)
 		return status;
 	*s = find_stateid_by_type(cl, stateid, typemask);
@@ -5062,7 +5065,6 @@ nfs4_state_destroy_net(struct net *net)
 	int i;
 	struct nfs4_client *clp = NULL;
 	struct nfsd_net *nn = net_generic(net, nfsd_net_id);
-	struct rb_node *node, *tmp;
 
 	for (i = 0; i < CLIENT_HASH_SIZE; i++) {
 		while (!list_empty(&nn->conf_id_hashtbl[i])) {
@@ -5071,14 +5073,12 @@ nfs4_state_destroy_net(struct net *net)
 		}
 	}
 
-	node = rb_first(&nn->unconf_name_tree);
-	while (node != NULL) {
-		tmp = node;
-		node = rb_next(tmp);
-		clp = rb_entry(tmp, struct nfs4_client, cl_namenode);
-		rb_erase(tmp, &nn->unconf_name_tree);
+	for (i = 0; i < CLIENT_HASH_SIZE; i++) {
+		while (!list_empty(&nn->unconf_id_hashtbl[i])) {
+			clp = list_entry(nn->unconf_id_hashtbl[i].next, struct nfs4_client, cl_idhash);
 			destroy_client(clp);
 		}
+	}
 
 	kfree(nn->sessionid_hashtbl);
 	kfree(nn->lockowner_ino_hashtbl);

fs/nfsd/nfs4xdr.c

@@ -294,7 +294,7 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval,
 		READ32(nace);
 
 		if (nace > NFS4_ACL_MAX)
-			return nfserr_resource;
+			return nfserr_fbig;
 
 		*acl = nfs4_acl_new(nace);
 		if (*acl == NULL)
@@ -1222,7 +1222,6 @@ nfsd4_decode_write(struct nfsd4_compoundargs *argp, struct nfsd4_write *write)
 	}
 	write->wr_head.iov_base = p;
 	write->wr_head.iov_len = avail;
-	WARN_ON(avail != (XDR_QUADLEN(avail) << 2));
 	write->wr_pagelist = argp->pagelist;
 
 	len = XDR_QUADLEN(write->wr_buflen) << 2;
@@ -2483,6 +2482,8 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 			goto out;
 	}
 	if (bmval2 & FATTR4_WORD2_SUPPATTR_EXCLCREAT) {
+		if ((buflen -= 16) < 0)
+			goto out_resource;
 		WRITE32(3);
 		WRITE32(NFSD_SUPPATTR_EXCLCREAT_WORD0);
 		WRITE32(NFSD_SUPPATTR_EXCLCREAT_WORD1);
@@ -2499,8 +2500,10 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp,
 		security_release_secctx(context, contextlen);
 #endif /* CONFIG_NFSD_V4_SECURITY_LABEL */
 	kfree(acl);
-	if (tempfh)
+	if (tempfh) {
 		fh_put(tempfh);
+		kfree(tempfh);
+	}
 	return status;
 out_nfserr:
 	status = nfserrno(err);
@@ -3471,6 +3474,9 @@ nfsd4_encode_test_stateid(struct nfsd4_compoundres *resp, __be32 nfserr,
 	struct nfsd4_test_stateid_id *stateid, *next;
 	__be32 *p;
 
+	if (nfserr)
+		return nfserr;
+
 	RESERVE_SPACE(4 + (4 * test_stateid->ts_num_ids));
 	*p++ = htonl(test_stateid->ts_num_ids);
 
@@ -3579,8 +3585,6 @@ __be32 nfsd4_check_resp_size(struct nfsd4_compoundres *resp, u32 pad)
 		return 0;
 
 	session = resp->cstate.session;
-	if (session == NULL)
-		return 0;
 
 	if (xb->page_len == 0) {
 		length = (char *)resp->p - (char *)xb->head[0].iov_base + pad;
@@ -3620,9 +3624,17 @@ nfsd4_encode_operation(struct nfsd4_compoundres *resp, struct nfsd4_op *op)
 	BUG_ON(op->opnum < 0 || op->opnum >= ARRAY_SIZE(nfsd4_enc_ops) ||
 	       !nfsd4_enc_ops[op->opnum]);
 	op->status = nfsd4_enc_ops[op->opnum](resp, op->status, &op->u);
-	/* nfsd4_check_drc_limit guarantees enough room for error status */
+	/* nfsd4_check_resp_size guarantees enough room for error status */
 	if (!op->status)
 		op->status = nfsd4_check_resp_size(resp, 0);
+	if (op->status == nfserr_resource && nfsd4_has_session(&resp->cstate)) {
+		struct nfsd4_slot *slot = resp->cstate.slot;
+
+		if (slot->sl_flags & NFSD4_SLOT_CACHETHIS)
+			op->status = nfserr_rep_too_big_to_cache;
+		else
+			op->status = nfserr_rep_too_big;
+	}
 	if (so) {
 		so->so_replay.rp_status = op->status;
 		so->so_replay.rp_buflen = (char *)resp->p - (char *)(statp+1);
@@ -3691,6 +3703,12 @@ int nfsd4_release_compoundargs(void *rq, __be32 *p, void *resp)
 int
 nfs4svc_decode_compoundargs(struct svc_rqst *rqstp, __be32 *p, struct nfsd4_compoundargs *args)
 {
+	if (rqstp->rq_arg.head[0].iov_len % 4) {
+		/* client is nuts */
+		dprintk("%s: compound not properly padded! (peeraddr=%pISc xid=0x%x)",
+			__func__, svc_addr(rqstp), be32_to_cpu(rqstp->rq_xid));
+		return 0;
+	}
 	args->p = p;
 	args->end = rqstp->rq_arg.head[0].iov_base + rqstp->rq_arg.head[0].iov_len;
 	args->pagelist = rqstp->rq_arg.pages;

fs/nfsd/nfsctl.c

@@ -699,6 +699,11 @@ static ssize_t __write_ports_addfd(char *buf, struct net *net)
 	if (err != 0 || fd < 0)
 		return -EINVAL;
 
+	if (svc_alien_sock(net, fd)) {
+		printk(KERN_ERR "%s: socket net is different to NFSd's one\n", __func__);
+		return -EINVAL;
+	}
+
 	err = nfsd_create_serv(net);
 	if (err != 0)
 		return err;

fs/nfsd/nfsd.h

@@ -282,7 +282,7 @@ void		nfsd_lockd_shutdown(void);
  * reason.
  */
 #define	COMPOUND_SLACK_SPACE		140    /* OP_GETFH */
-#define COMPOUND_ERR_SLACK_SPACE	12     /* OP_SETATTR */
+#define COMPOUND_ERR_SLACK_SPACE	16     /* OP_SETATTR */
 
 #define NFSD_LAUNDROMAT_MINTIMEOUT      1   /* seconds */
 

fs/nfsd/nfsfh.h

@@ -132,6 +132,17 @@ fh_init(struct svc_fh *fhp, int maxsize)
 }
 
 #ifdef CONFIG_NFSD_V3
+/*
+ * The wcc data stored in current_fh should be cleared
+ * between compound ops.
+ */
+static inline void
+fh_clear_wcc(struct svc_fh *fhp)
+{
+	fhp->fh_post_saved = 0;
+	fhp->fh_pre_saved = 0;
+}
+
 /*
  * Fill in the pre_op attr for the wcc data
  */
@@ -152,6 +163,7 @@ fill_pre_wcc(struct svc_fh *fhp)
 
 extern void fill_post_wcc(struct svc_fh *);
 #else
+#define fh_clear_wcc(ignored)
 #define fill_pre_wcc(ignored)
 #define fill_post_wcc(notused)
 #endif /* CONFIG_NFSD_V3 */

fs/nfsd/nfsxdr.c

@@ -152,7 +152,7 @@ encode_fattr(struct svc_rqst *rqstp, __be32 *p, struct svc_fh *fhp,
 	type = (stat->mode & S_IFMT);
 
 	*p++ = htonl(nfs_ftypes[type >> 12]);
-	*p++ = htonl((u32) (stat->mode & S_IALLUGO));
+	*p++ = htonl((u32) stat->mode);
 	*p++ = htonl((u32) stat->nlink);
 	*p++ = htonl((u32) from_kuid(&init_user_ns, stat->uid));
 	*p++ = htonl((u32) from_kgid(&init_user_ns, stat->gid));

fs/nfsd/vfs.c

@@ -404,6 +404,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 	umode_t		ftype = 0;
 	__be32		err;
 	int		host_err;
+	bool		get_write_count;
 	int		size_change = 0;
 
 	if (iap->ia_valid & (ATTR_ATIME | ATTR_MTIME | ATTR_SIZE))
@@ -411,10 +412,18 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, struct iattr *iap,
 	if (iap->ia_valid & ATTR_SIZE)
 		ftype = S_IFREG;
 
+	/* Callers that do fh_verify should do the fh_want_write: */
+	get_write_count = !fhp->fh_dentry;
+
 	/* Get inode */
 	err = fh_verify(rqstp, fhp, ftype, accmode);
 	if (err)
 		goto out;
+	if (get_write_count) {
+		host_err = fh_want_write(fhp);
+		if (host_err)
+			return nfserrno(host_err);
+	}
 
 	dentry = fhp->fh_dentry;
 	inode = dentry->d_inode;
@@ -1706,10 +1715,10 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen,
 	dput(odentry);
  out_nfserr:
 	err = nfserrno(host_err);
-
-	/* we cannot reply on fh_unlock on the two filehandles,
+	/*
+	 * We cannot rely on fh_unlock on the two filehandles,
 	 * as that would do the wrong thing if the two directories
-	 * were the same, so again we do it by hand
+	 * were the same, so again we do it by hand.
 	 */
 	fill_post_wcc(ffhp);
 	fill_post_wcc(tfhp);

fs/nfsd/xdr4.h

@@ -574,8 +574,6 @@ extern __be32 nfsd4_setclientid_confirm(struct svc_rqst *rqstp,
 		struct nfsd4_compound_state *,
 		struct nfsd4_setclientid_confirm *setclientid_confirm);
 extern void nfsd4_store_cache_entry(struct nfsd4_compoundres *resp);
-extern __be32 nfsd4_replay_cache_entry(struct nfsd4_compoundres *resp,
-		struct nfsd4_sequence *seq);
 extern __be32 nfsd4_exchange_id(struct svc_rqst *rqstp,
 		struct nfsd4_compound_state *, struct nfsd4_exchange_id *);
 extern __be32 nfsd4_backchannel_ctl(struct svc_rqst *, struct nfsd4_compound_state *, struct nfsd4_backchannel_ctl *);

include/linux/sunrpc/clnt.h

@@ -130,6 +130,8 @@ struct rpc_create_args {
 #define RPC_CLNT_CREATE_NO_RETRANS_TIMEOUT	(1UL << 9)
 
 struct rpc_clnt *rpc_create(struct rpc_create_args *args);
+struct rpc_clnt *rpc_create_xprt(struct rpc_create_args *args,
+					struct rpc_xprt *xprt);
 struct rpc_clnt	*rpc_bind_new_program(struct rpc_clnt *,
 				const struct rpc_program *, u32);
 void rpc_task_reset_client(struct rpc_task *task, struct rpc_clnt *clnt);

include/linux/sunrpc/svcsock.h

@@ -56,6 +56,7 @@ int		svc_recv(struct svc_rqst *, long);
 int		svc_send(struct svc_rqst *);
 void		svc_drop(struct svc_rqst *);
 void		svc_sock_update_bufs(struct svc_serv *serv);
+bool		svc_alien_sock(struct net *net, int fd);
 int		svc_addsock(struct svc_serv *serv, const int fd,
 					char *name_return, const size_t len);
 void		svc_init_xprt_sock(void);

include/linux/sunrpc/xprt.h

@@ -295,13 +295,24 @@ int			xprt_adjust_timeout(struct rpc_rqst *req);
 void			xprt_release_xprt(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_release_xprt_cong(struct rpc_xprt *xprt, struct rpc_task *task);
 void			xprt_release(struct rpc_task *task);
-struct rpc_xprt *	xprt_get(struct rpc_xprt *xprt);
 void			xprt_put(struct rpc_xprt *xprt);
 struct rpc_xprt *	xprt_alloc(struct net *net, size_t size,
 				unsigned int num_prealloc,
 				unsigned int max_req);
 void			xprt_free(struct rpc_xprt *);
 
+/**
+ * xprt_get - return a reference to an RPC transport.
+ * @xprt: pointer to the transport
+ *
+ */
+static inline struct rpc_xprt *xprt_get(struct rpc_xprt *xprt)
+{
+	if (atomic_inc_not_zero(&xprt->count))
+		return xprt;
+	return NULL;
+}
+
 static inline __be32 *xprt_skip_transport_header(struct rpc_xprt *xprt, __be32 *p)
 {
 	return p + xprt->tsh_size;

net/sunrpc/Kconfig

@@ -9,19 +9,6 @@ config SUNRPC_BACKCHANNEL
 	bool
 	depends on SUNRPC
 
-config SUNRPC_XPRT_RDMA
-	tristate
-	depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS
-	default SUNRPC && INFINIBAND
-	help
-	  This option allows the NFS client and server to support
-	  an RDMA-enabled transport.
-
-	  To compile RPC client RDMA transport support as a module,
-	  choose M here: the module will be called xprtrdma.
-
-	  If unsure, say N.
-
 config SUNRPC_SWAP
 	bool
 	depends on SUNRPC
@@ -57,3 +44,29 @@ config SUNRPC_DEBUG
 	  but makes troubleshooting NFS issues significantly harder.
 
 	  If unsure, say Y.
+
+config SUNRPC_XPRT_RDMA_CLIENT
+	tristate "RPC over RDMA Client Support"
+	depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS
+	default SUNRPC && INFINIBAND
+	help
+	  This option allows the NFS client to support an RDMA-enabled
+	  transport.
+
+	  To compile RPC client RDMA transport support as a module,
+	  choose M here: the module will be called xprtrdma.
+
+	  If unsure, say N.
+
+config SUNRPC_XPRT_RDMA_SERVER
+	tristate "RPC over RDMA Server Support"
+	depends on SUNRPC && INFINIBAND && INFINIBAND_ADDR_TRANS
+	default SUNRPC && INFINIBAND
+	help
+	  This option allows the NFS server to support an RDMA-enabled
+	  transport.
+
+	  To compile RPC server RDMA transport support as a module,
+	  choose M here: the module will be called svcrdma.
+
+	  If unsure, say N.

net/sunrpc/Makefile

@@ -5,7 +5,8 @@
 
 obj-$(CONFIG_SUNRPC) += sunrpc.o
 obj-$(CONFIG_SUNRPC_GSS) += auth_gss/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA) += xprtrdma/
+
+obj-y += xprtrdma/
 
 sunrpc-y := clnt.o xprt.o socklib.o xprtsock.o sched.o \
 	    auth.o auth_null.o auth_unix.o auth_generic.o \

net/sunrpc/clnt.c

@@ -438,6 +438,38 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args,
 	return ERR_PTR(err);
 }
 
+struct rpc_clnt *rpc_create_xprt(struct rpc_create_args *args,
+					struct rpc_xprt *xprt)
+{
+	struct rpc_clnt *clnt = NULL;
+
+	clnt = rpc_new_client(args, xprt, NULL);
+	if (IS_ERR(clnt))
+		return clnt;
+
+	if (!(args->flags & RPC_CLNT_CREATE_NOPING)) {
+		int err = rpc_ping(clnt);
+		if (err != 0) {
+			rpc_shutdown_client(clnt);
+			return ERR_PTR(err);
+		}
+	}
+
+	clnt->cl_softrtry = 1;
+	if (args->flags & RPC_CLNT_CREATE_HARDRTRY)
+		clnt->cl_softrtry = 0;
+
+	if (args->flags & RPC_CLNT_CREATE_AUTOBIND)
+		clnt->cl_autobind = 1;
+	if (args->flags & RPC_CLNT_CREATE_DISCRTRY)
+		clnt->cl_discrtry = 1;
+	if (!(args->flags & RPC_CLNT_CREATE_QUIET))
+		clnt->cl_chatty = 1;
+
+	return clnt;
+}
+EXPORT_SYMBOL_GPL(rpc_create_xprt);
+
 /**
  * rpc_create - create an RPC client and transport with one call
  * @args: rpc_clnt create argument structure
@@ -451,7 +483,6 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args,
 struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 {
 	struct rpc_xprt *xprt;
-	struct rpc_clnt *clnt;
 	struct xprt_create xprtargs = {
 		.net = args->net,
 		.ident = args->protocol,
@@ -515,30 +546,7 @@ struct rpc_clnt *rpc_create(struct rpc_create_args *args)
 	if (args->flags & RPC_CLNT_CREATE_NONPRIVPORT)
 		xprt->resvport = 0;
 
-	clnt = rpc_new_client(args, xprt, NULL);
-	if (IS_ERR(clnt))
-		return clnt;
-
-	if (!(args->flags & RPC_CLNT_CREATE_NOPING)) {
-		int err = rpc_ping(clnt);
-		if (err != 0) {
-			rpc_shutdown_client(clnt);
-			return ERR_PTR(err);
-		}
-	}
-
-	clnt->cl_softrtry = 1;
-	if (args->flags & RPC_CLNT_CREATE_HARDRTRY)
-		clnt->cl_softrtry = 0;
-
-	if (args->flags & RPC_CLNT_CREATE_AUTOBIND)
-		clnt->cl_autobind = 1;
-	if (args->flags & RPC_CLNT_CREATE_DISCRTRY)
-		clnt->cl_discrtry = 1;
-	if (!(args->flags & RPC_CLNT_CREATE_QUIET))
-		clnt->cl_chatty = 1;
-
-	return clnt;
+	return rpc_create_xprt(args, xprt);
 }
 EXPORT_SYMBOL_GPL(rpc_create);
 

net/sunrpc/svcsock.c

@@ -1397,6 +1397,22 @@ static struct svc_sock *svc_setup_socket(struct svc_serv *serv,
 	return svsk;
 }
 
+bool svc_alien_sock(struct net *net, int fd)
+{
+	int err;
+	struct socket *sock = sockfd_lookup(fd, &err);
+	bool ret = false;
+
+	if (!sock)
+		goto out;
+	if (sock_net(sock->sk) != net)
+		ret = true;
+	sockfd_put(sock);
+out:
+	return ret;
+}
+EXPORT_SYMBOL_GPL(svc_alien_sock);
+
 /**
  * svc_addsock - add a listener socket to an RPC service
  * @serv: pointer to RPC service to which to add a new listener

net/sunrpc/xdr.c

@@ -833,8 +833,20 @@ xdr_buf_from_iov(struct kvec *iov, struct xdr_buf *buf)
 }
 EXPORT_SYMBOL_GPL(xdr_buf_from_iov);
 
-/* Sets subbuf to the portion of buf of length len beginning base bytes
- * from the start of buf. Returns -1 if base of length are out of bounds. */
+/**
+ * xdr_buf_subsegment - set subbuf to a portion of buf
+ * @buf: an xdr buffer
+ * @subbuf: the result buffer
+ * @base: beginning of range in bytes
+ * @len: length of range in bytes
+ *
+ * sets @subbuf to an xdr buffer representing the portion of @buf of
+ * length @len starting at offset @base.
+ *
+ * @buf and @subbuf may be pointers to the same struct xdr_buf.
+ *
+ * Returns -1 if base of length are out of bounds.
+ */
 int
 xdr_buf_subsegment(struct xdr_buf *buf, struct xdr_buf *subbuf,
 			unsigned int base, unsigned int len)
@@ -847,9 +859,8 @@ xdr_buf_subsegment(struct xdr_buf *buf, struct xdr_buf *subbuf,
 		len -= subbuf->head[0].iov_len;
 		base = 0;
 	} else {
-		subbuf->head[0].iov_base = NULL;
-		subbuf->head[0].iov_len = 0;
 		base -= buf->head[0].iov_len;
+		subbuf->head[0].iov_len = 0;
 	}
 
 	if (base < buf->page_len) {
@@ -871,9 +882,8 @@ xdr_buf_subsegment(struct xdr_buf *buf, struct xdr_buf *subbuf,
 		len -= subbuf->tail[0].iov_len;
 		base = 0;
 	} else {
-		subbuf->tail[0].iov_base = NULL;
-		subbuf->tail[0].iov_len = 0;
 		base -= buf->tail[0].iov_len;
+		subbuf->tail[0].iov_len = 0;
 	}
 
 	if (base || len)

net/sunrpc/xprt.c

@@ -1383,15 +1383,3 @@ void xprt_put(struct rpc_xprt *xprt)
 	if (atomic_dec_and_test(&xprt->count))
 		xprt_destroy(xprt);
 }
-
-/**
- * xprt_get - return a reference to an RPC transport.
- * @xprt: pointer to the transport
- *
- */
-struct rpc_xprt *xprt_get(struct rpc_xprt *xprt)
-{
-	if (atomic_inc_not_zero(&xprt->count))
-		return xprt;
-	return NULL;
-}

net/sunrpc/xprtrdma/Makefile

-obj-$(CONFIG_SUNRPC_XPRT_RDMA) += xprtrdma.o
+obj-$(CONFIG_SUNRPC_XPRT_RDMA_CLIENT) += xprtrdma.o
 
 xprtrdma-y := transport.o rpc_rdma.o verbs.o
 
-obj-$(CONFIG_SUNRPC_XPRT_RDMA) += svcrdma.o
+obj-$(CONFIG_SUNRPC_XPRT_RDMA_SERVER) += svcrdma.o
 
 svcrdma-y := svc_rdma.o svc_rdma_transport.o \
 	svc_rdma_marshal.o svc_rdma_sendto.o svc_rdma_recvfrom.o

net/sunrpc/xprtrdma/svc_rdma_recvfrom.c

@@ -90,6 +90,7 @@ static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
 		sge_no++;
 	}
 	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
@@ -169,6 +170,7 @@ static int map_read_chunks(struct svcxprt_rdma *xprt,
 		 */
 		head->arg.pages[page_no] = rqstp->rq_arg.pages[page_no];
 		rqstp->rq_respages = &rqstp->rq_arg.pages[page_no+1];
+		rqstp->rq_next_page = rqstp->rq_respages + 1;
 
 		byte_count -= sge_bytes;
 		ch_bytes -= sge_bytes;
@@ -276,6 +278,7 @@ static int fast_reg_read_chunks(struct svcxprt_rdma *xprt,
 
 	/* rq_respages points one past arg pages */
 	rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];
+	rqstp->rq_next_page = rqstp->rq_respages + 1;
 
 	/* Create the reply and chunk maps */
 	offset = 0;
@@ -520,13 +523,6 @@ static int rdma_read_xdr(struct svcxprt_rdma *xprt,
 	for (ch_no = 0; &rqstp->rq_pages[ch_no] < rqstp->rq_respages; ch_no++)
 		rqstp->rq_pages[ch_no] = NULL;
 
-	/*
-	 * Detach res pages. If svc_release sees any it will attempt to
-	 * put them.
-	 */
-	while (rqstp->rq_next_page != rqstp->rq_respages)
-		*(--rqstp->rq_next_page) = NULL;
-
 	return err;
 }
 
@@ -550,7 +546,7 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
 
 	/* rq_respages starts after the last arg page */
 	rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];
-	rqstp->rq_next_page = &rqstp->rq_arg.pages[page_no];
+	rqstp->rq_next_page = rqstp->rq_respages + 1;
 
 	/* Rebuild rq_arg head and tail. */
 	rqstp->rq_arg.head[0] = head->arg.head[0];

net/sunrpc/xprtrdma/svc_rdma_sendto.c

@@ -265,6 +265,7 @@ static dma_addr_t dma_map_xdr(struct svcxprt_rdma *xprt,
 		xdr_off -= xdr->head[0].iov_len;
 		if (xdr_off < xdr->page_len) {
 			/* This offset is in the page list */
+			xdr_off += xdr->page_base;
 			page = xdr->pages[xdr_off >> PAGE_SHIFT];
 			xdr_off &= ~PAGE_MASK;
 		} else {
@@ -625,6 +626,7 @@ static int send_reply(struct svcxprt_rdma *rdma,
 		if (page_no+1 >= sge_no)
 			ctxt->sge[page_no+1].length = 0;
 	}
+	rqstp->rq_next_page = rqstp->rq_respages + 1;
 	BUG_ON(sge_no > rdma->sc_max_sge);
 	memset(&send_wr, 0, sizeof send_wr);
 	ctxt->wr_op = IB_WR_SEND;

net/sunrpc/xprtrdma/svc_rdma_transport.c

@@ -477,8 +477,7 @@ struct page *svc_rdma_get_page(void)
 
 	while ((page = alloc_page(GFP_KERNEL)) == NULL) {
 		/* If we can't get memory, wait a bit and try again */
-		printk(KERN_INFO "svcrdma: out of memory...retrying in 1000 "
-		       "jiffies.\n");
+		printk(KERN_INFO "svcrdma: out of memory...retrying in 1s\n");
 		schedule_timeout_uninterruptible(msecs_to_jiffies(1000));
 	}
 	return page;

net/sunrpc/xprtsock.c

@@ -909,6 +909,12 @@ static void xs_tcp_close(struct rpc_xprt *xprt)
 		xs_tcp_shutdown(xprt);
 }
 
+static void xs_xprt_free(struct rpc_xprt *xprt)
+{
+	xs_free_peer_addresses(xprt);
+	xprt_free(xprt);
+}
+
 /**
  * xs_destroy - prepare to shutdown a transport
  * @xprt: doomed transport
@@ -919,8 +925,7 @@ static void xs_destroy(struct rpc_xprt *xprt)
 	dprintk("RPC:       xs_destroy xprt %p\n", xprt);
 
 	xs_close(xprt);
-	xs_free_peer_addresses(xprt);
-	xprt_free(xprt);
+	xs_xprt_free(xprt);
 	module_put(THIS_MODULE);
 }
 
@@ -2532,6 +2537,10 @@ static void bc_close(struct rpc_xprt *xprt)
 
 static void bc_destroy(struct rpc_xprt *xprt)
 {
+	dprintk("RPC:       bc_destroy xprt %p\n", xprt);
+
+	xs_xprt_free(xprt);
+	module_put(THIS_MODULE);
 }
 
 static struct rpc_xprt_ops xs_local_ops = {
@@ -2732,7 +2741,7 @@ static struct rpc_xprt *xs_setup_local(struct xprt_create *args)
 		return xprt;
 	ret = ERR_PTR(-EINVAL);
 out_err:
-	xprt_free(xprt);
+	xs_xprt_free(xprt);
 	return ret;
 }
 
@@ -2810,7 +2819,7 @@ static struct rpc_xprt *xs_setup_udp(struct xprt_create *args)
 		return xprt;
 	ret = ERR_PTR(-EINVAL);
 out_err:
-	xprt_free(xprt);
+	xs_xprt_free(xprt);
 	return ret;
 }
 
@@ -2885,12 +2894,11 @@ static struct rpc_xprt *xs_setup_tcp(struct xprt_create *args)
 				xprt->address_strings[RPC_DISPLAY_ADDR],
 				xprt->address_strings[RPC_DISPLAY_PROTO]);
 
-
 	if (try_module_get(THIS_MODULE))
 		return xprt;
 	ret = ERR_PTR(-EINVAL);
 out_err:
-	xprt_free(xprt);
+	xs_xprt_free(xprt);
 	return ret;
 }
 
@@ -2907,15 +2915,6 @@ static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args)
 	struct svc_sock *bc_sock;
 	struct rpc_xprt *ret;
 
-	if (args->bc_xprt->xpt_bc_xprt) {
-		/*
-		 * This server connection already has a backchannel
-		 * transport; we can't create a new one, as we wouldn't
-		 * be able to match replies based on xid any more.  So,
-		 * reuse the already-existing one:
-		 */
-		 return args->bc_xprt->xpt_bc_xprt;
-	}
 	xprt = xs_setup_xprt(args, xprt_tcp_slot_table_entries,
 			xprt_tcp_slot_table_entries);
 	if (IS_ERR(xprt))
@@ -2973,13 +2972,14 @@ static struct rpc_xprt *xs_setup_bc_tcp(struct xprt_create *args)
 	 */
 	xprt_set_connected(xprt);
 
-
 	if (try_module_get(THIS_MODULE))
 		return xprt;
+
+	args->bc_xprt->xpt_bc_xprt = NULL;
 	xprt_put(xprt);
 	ret = ERR_PTR(-EINVAL);
 out_err:
-	xprt_free(xprt);
+	xs_xprt_free(xprt);
 	return ret;
 }