Merge tag 'nfs-for-4.7-1' of git://git.linux-nfs.org/projects/anna/linux-nfs

Pull NFS client updates from Anna Schumaker:
 "Highlights include:

  Features:
   - Add support for the NFS v4.2 COPY operation
   - Add support for NFS/RDMA over IPv6

  Bugfixes and cleanups:
   - Avoid race that crashes nfs_init_commit()
   - Fix oops in callback path
   - Fix LOCK/OPEN race when unlinking an open file
   - Choose correct stateids when using delegations in setattr, read and
     write
   - Don't send empty SETATTR after OPEN_CREATE
   - xprtrdma: Prevent server from writing a reply into memory client
     has released
   - xprtrdma: Support using Read list and Reply chunk in one RPC call"

* tag 'nfs-for-4.7-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (61 commits)
  pnfs: pnfs_update_layout needs to consider if strict iomode checking is on
  nfs/flexfiles: Use the layout segment for reading unless it a IOMODE_RW and reading is disabled
  nfs/flexfiles: Helper function to detect FF_FLAGS_NO_READ_IO
  nfs: avoid race that crashes nfs_init_commit
  NFS: checking for NULL instead of IS_ERR() in nfs_commit_file()
  pnfs: make pnfs_layout_process more robust
  pnfs: rework LAYOUTGET retry handling
  pnfs: lift retry logic from send_layoutget to pnfs_update_layout
  pnfs: fix bad error handling in send_layoutget
  flexfiles: add kerneldoc header to nfs4_ff_layout_prepare_ds
  flexfiles: remove pointless setting of NFS_LAYOUT_RETURN_REQUESTED
  pnfs: only tear down lsegs that precede seqid in LAYOUTRETURN args
  pnfs: keep track of the return sequence number in pnfs_layout_hdr
  pnfs: record sequence in pnfs_layout_segment when it's created
  pnfs: don't merge new ff lsegs with ones that have LAYOUTRETURN bit set
  pNFS/flexfiles: When initing reads or writes, we might have to retry connecting to DSes
  pNFS/flexfiles: When checking for available DSes, conditionally check for MDS io
  pNFS/flexfile: Fix erroneous fall back to read/write through the MDS
  NFS: Reclaim writes via writepage are opportunistic
  NFSv4: Use the right stateid for delegations in setattr, read and write
  ...

1 files changed, 99 insertions, 86 deletions

diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 084e8570da18..223982eb38c9 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -74,6 +74,17 @@
 #define NFS4_POLL_RETRY_MIN	(HZ/10)
 #define NFS4_POLL_RETRY_MAX	(15*HZ)
 
+/* file attributes which can be mapped to nfs attributes */
+#define NFS4_VALID_ATTRS (ATTR_MODE \
+	| ATTR_UID \
+	| ATTR_GID \
+	| ATTR_SIZE \
+	| ATTR_ATIME \
+	| ATTR_MTIME \
+	| ATTR_CTIME \
+	| ATTR_ATIME_SET \
+	| ATTR_MTIME_SET)
+
 struct nfs4_opendata;
 static int _nfs4_proc_open(struct nfs4_opendata *data);
 static int _nfs4_recover_proc_open(struct nfs4_opendata *data);
@@ -416,6 +427,7 @@ static int nfs4_do_handle_exception(struct nfs_server *server,
 		case -NFS4ERR_DELAY:
 			nfs_inc_server_stats(server, NFSIOS_DELAY);
 		case -NFS4ERR_GRACE:
+		case -NFS4ERR_RECALLCONFLICT:
 			exception->delay = 1;
 			return 0;
 
@@ -2558,15 +2570,20 @@ static int _nfs4_do_open(struct inode *dir,
 	if ((opendata->o_arg.open_flags & (O_CREAT|O_EXCL)) == (O_CREAT|O_EXCL) &&
 	    (opendata->o_arg.createmode != NFS4_CREATE_GUARDED)) {
 		nfs4_exclusive_attrset(opendata, sattr, &label);
-
-		nfs_fattr_init(opendata->o_res.f_attr);
-		status = nfs4_do_setattr(state->inode, cred,
-				opendata->o_res.f_attr, sattr,
-				state, label, olabel);
-		if (status == 0) {
-			nfs_setattr_update_inode(state->inode, sattr,
-					opendata->o_res.f_attr);
-			nfs_setsecurity(state->inode, opendata->o_res.f_attr, olabel);
+		/*
+		 * send create attributes which was not set by open
+		 * with an extra setattr.
+		 */
+		if (sattr->ia_valid & NFS4_VALID_ATTRS) {
+			nfs_fattr_init(opendata->o_res.f_attr);
+			status = nfs4_do_setattr(state->inode, cred,
+					opendata->o_res.f_attr, sattr,
+					state, label, olabel);
+			if (status == 0) {
+				nfs_setattr_update_inode(state->inode, sattr,
+						opendata->o_res.f_attr);
+				nfs_setsecurity(state->inode, opendata->o_res.f_attr, olabel);
+			}
 		}
 	}
 	if (opened && opendata->file_created)
@@ -2676,6 +2693,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 		.rpc_resp	= &res,
 		.rpc_cred	= cred,
         };
+	struct rpc_cred *delegation_cred = NULL;
 	unsigned long timestamp = jiffies;
 	fmode_t fmode;
 	bool truncate;
@@ -2691,7 +2709,7 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 	truncate = (sattr->ia_valid & ATTR_SIZE) ? true : false;
 	fmode = truncate ? FMODE_WRITE : FMODE_READ;
 
-	if (nfs4_copy_delegation_stateid(&arg.stateid, inode, fmode)) {
+	if (nfs4_copy_delegation_stateid(inode, fmode, &arg.stateid, &delegation_cred)) {
 		/* Use that stateid */
 	} else if (truncate && state != NULL) {
 		struct nfs_lockowner lockowner = {
@@ -2700,13 +2718,17 @@ static int _nfs4_do_setattr(struct inode *inode, struct rpc_cred *cred,
 		};
 		if (!nfs4_valid_open_stateid(state))
 			return -EBADF;
-		if (nfs4_select_rw_stateid(&arg.stateid, state, FMODE_WRITE,
-				&lockowner) == -EIO)
+		if (nfs4_select_rw_stateid(state, FMODE_WRITE, &lockowner,
+				&arg.stateid, &delegation_cred) == -EIO)
 			return -EBADF;
 	} else
 		nfs4_stateid_copy(&arg.stateid, &zero_stateid);
+	if (delegation_cred)
+		msg.rpc_cred = delegation_cred;
 
 	status = nfs4_call_sync(server->client, server, &msg, &arg.seq_args, &res.seq_res, 1);
+
+	put_rpccred(delegation_cred);
 	if (status == 0 && state != NULL)
 		renew_lease(server, timestamp);
 	trace_nfs4_setattr(inode, &arg.stateid, status);
@@ -4285,7 +4307,7 @@ int nfs4_set_rw_stateid(nfs4_stateid *stateid,
 
 	if (l_ctx != NULL)
 		lockowner = &l_ctx->lockowner;
-	return nfs4_select_rw_stateid(stateid, ctx->state, fmode, lockowner);
+	return nfs4_select_rw_stateid(ctx->state, fmode, lockowner, stateid, NULL);
 }
 EXPORT_SYMBOL_GPL(nfs4_set_rw_stateid);
 
@@ -6054,6 +6076,7 @@ static int nfs41_lock_expired(struct nfs4_state *state, struct file_lock *reques
 static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock *request)
 {
 	struct nfs_inode *nfsi = NFS_I(state->inode);
+	struct nfs4_state_owner *sp = state->owner;
 	unsigned char fl_flags = request->fl_flags;
 	int status = -ENOLCK;
 
@@ -6068,6 +6091,7 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock
 	status = do_vfs_lock(state->inode, request);
 	if (status < 0)
 		goto out;
+	mutex_lock(&sp->so_delegreturn_mutex);
 	down_read(&nfsi->rwsem);
 	if (test_bit(NFS_DELEGATED_STATE, &state->flags)) {
 		/* Yes: cache locks! */
@@ -6075,9 +6099,11 @@ static int _nfs4_proc_setlk(struct nfs4_state *state, int cmd, struct file_lock
 		request->fl_flags = fl_flags & ~FL_SLEEP;
 		status = do_vfs_lock(state->inode, request);
 		up_read(&nfsi->rwsem);
+		mutex_unlock(&sp->so_delegreturn_mutex);
 		goto out;
 	}
 	up_read(&nfsi->rwsem);
+	mutex_unlock(&sp->so_delegreturn_mutex);
 	status = _nfs4_do_setlk(state, cmd, request, NFS_LOCK_NEW);
 out:
 	request->fl_flags = fl_flags;
@@ -7351,9 +7377,11 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo)
  * always set csa_cachethis to FALSE because the current implementation
  * of the back channel DRC only supports caching the CB_SEQUENCE operation.
  */
-static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args)
+static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args,
+				    struct rpc_clnt *clnt)
 {
 	unsigned int max_rqst_sz, max_resp_sz;
+	unsigned int max_bc_payload = rpc_max_bc_payload(clnt);
 
 	max_rqst_sz = NFS_MAX_FILE_IO_SIZE + nfs41_maxwrite_overhead;
 	max_resp_sz = NFS_MAX_FILE_IO_SIZE + nfs41_maxread_overhead;
@@ -7371,8 +7399,8 @@ static void nfs4_init_channel_attrs(struct nfs41_create_session_args *args)
 		args->fc_attrs.max_ops, args->fc_attrs.max_reqs);
 
 	/* Back channel attributes */
-	args->bc_attrs.max_rqst_sz = PAGE_SIZE;
-	args->bc_attrs.max_resp_sz = PAGE_SIZE;
+	args->bc_attrs.max_rqst_sz = max_bc_payload;
+	args->bc_attrs.max_resp_sz = max_bc_payload;
 	args->bc_attrs.max_resp_sz_cached = 0;
 	args->bc_attrs.max_ops = NFS4_MAX_BACK_CHANNEL_OPS;
 	args->bc_attrs.max_reqs = NFS41_BC_MAX_CALLBACKS;
@@ -7476,7 +7504,7 @@ static int _nfs4_proc_create_session(struct nfs_client *clp,
 	};
 	int status;
 
-	nfs4_init_channel_attrs(&args);
+	nfs4_init_channel_attrs(&args, clp->cl_rpcclient);
 	args.flags = (SESSION4_PERSIST | SESSION4_BACK_CHAN);
 
 	status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT);
@@ -7820,40 +7848,34 @@ nfs4_layoutget_prepare(struct rpc_task *task, void *calldata)
 	struct nfs4_layoutget *lgp = calldata;
 	struct nfs_server *server = NFS_SERVER(lgp->args.inode);
 	struct nfs4_session *session = nfs4_get_session(server);
-	int ret;
 
 	dprintk("--> %s\n", __func__);
-	/* Note the is a race here, where a CB_LAYOUTRECALL can come in
-	 * right now covering the LAYOUTGET we are about to send.
-	 * However, that is not so catastrophic, and there seems
-	 * to be no way to prevent it completely.
-	 */
-	if (nfs41_setup_sequence(session, &lgp->args.seq_args,
-				&lgp->res.seq_res, task))
-		return;
-	ret = pnfs_choose_layoutget_stateid(&lgp->args.stateid,
-					  NFS_I(lgp->args.inode)->layout,
-					  &lgp->args.range,
-					  lgp->args.ctx->state);
-	if (ret < 0)
-		rpc_exit(task, ret);
+	nfs41_setup_sequence(session, &lgp->args.seq_args,
+				&lgp->res.seq_res, task);
+	dprintk("<-- %s\n", __func__);
 }
 
 static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
 {
 	struct nfs4_layoutget *lgp = calldata;
+
+	dprintk("--> %s\n", __func__);
+	nfs41_sequence_done(task, &lgp->res.seq_res);
+	dprintk("<-- %s\n", __func__);
+}
+
+static int
+nfs4_layoutget_handle_exception(struct rpc_task *task,
+		struct nfs4_layoutget *lgp, struct nfs4_exception *exception)
+{
 	struct inode *inode = lgp->args.inode;
 	struct nfs_server *server = NFS_SERVER(inode);
 	struct pnfs_layout_hdr *lo;
-	struct nfs4_state *state = NULL;
-	unsigned long timeo, now, giveup;
+	int status = task->tk_status;
 
 	dprintk("--> %s tk_status => %d\n", __func__, -task->tk_status);
 
-	if (!nfs41_sequence_done(task, &lgp->res.seq_res))
-		goto out;
-
-	switch (task->tk_status) {
+	switch (status) {
 	case 0:
 		goto out;
 
@@ -7863,57 +7885,43 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
 	 * retry go inband.
 	 */
 	case -NFS4ERR_LAYOUTUNAVAILABLE:
-		task->tk_status = -ENODATA;
+		status = -ENODATA;
 		goto out;
 	/*
 	 * NFS4ERR_BADLAYOUT means the MDS cannot return a layout of
 	 * length lgp->args.minlength != 0 (see RFC5661 section 18.43.3).
 	 */
 	case -NFS4ERR_BADLAYOUT:
-		goto out_overflow;
+		status = -EOVERFLOW;
+		goto out;
 	/*
 	 * NFS4ERR_LAYOUTTRYLATER is a conflict with another client
 	 * (or clients) writing to the same RAID stripe except when
 	 * the minlength argument is 0 (see RFC5661 section 18.43.3).
+	 *
+	 * Treat it like we would RECALLCONFLICT -- we retry for a little
+	 * while, and then eventually give up.
 	 */
 	case -NFS4ERR_LAYOUTTRYLATER:
-		if (lgp->args.minlength == 0)
-			goto out_overflow;
-	/*
-	 * NFS4ERR_RECALLCONFLICT is when conflict with self (must recall
-	 * existing layout before getting a new one).
-	 */
-	case -NFS4ERR_RECALLCONFLICT:
-		timeo = rpc_get_timeout(task->tk_client);
-		giveup = lgp->args.timestamp + timeo;
-		now = jiffies;
-		if (time_after(giveup, now)) {
-			unsigned long delay;
-
-			/* Delay for:
-			 * - Not less then NFS4_POLL_RETRY_MIN.
-			 * - One last time a jiffie before we give up
-			 * - exponential backoff (time_now minus start_attempt)
-			 */
-			delay = max_t(unsigned long, NFS4_POLL_RETRY_MIN,
-				    min((giveup - now - 1),
-					now - lgp->args.timestamp));
-
-			dprintk("%s: NFS4ERR_RECALLCONFLICT waiting %lu\n",
-				__func__, delay);
-			rpc_delay(task, delay);
-			/* Do not call nfs4_async_handle_error() */
-			goto out_restart;
+		if (lgp->args.minlength == 0) {
+			status = -EOVERFLOW;
+			goto out;
 		}
-		break;
+		/* Fallthrough */
+	case -NFS4ERR_RECALLCONFLICT:
+		nfs4_handle_exception(server, -NFS4ERR_RECALLCONFLICT,
+					exception);
+		status = -ERECALLCONFLICT;
+		goto out;
 	case -NFS4ERR_EXPIRED:
 	case -NFS4ERR_BAD_STATEID:
+		exception->timeout = 0;
 		spin_lock(&inode->i_lock);
 		if (nfs4_stateid_match(&lgp->args.stateid,
 					&lgp->args.ctx->state->stateid)) {
 			spin_unlock(&inode->i_lock);
 			/* If the open stateid was bad, then recover it. */
-			state = lgp->args.ctx->state;
+			exception->state = lgp->args.ctx->state;
 			break;
 		}
 		lo = NFS_I(inode)->layout;
@@ -7926,25 +7934,21 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
 			 * with the current stateid.
 			 */
 			set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
-			pnfs_mark_matching_lsegs_invalid(lo, &head, NULL);
+			pnfs_mark_matching_lsegs_invalid(lo, &head, NULL, 0);
 			spin_unlock(&inode->i_lock);
 			pnfs_free_lseg_list(&head);
 		} else
 			spin_unlock(&inode->i_lock);
-		goto out_restart;
+		status = -EAGAIN;
+		goto out;
 	}
-	if (nfs4_async_handle_error(task, server, state, &lgp->timeout) == -EAGAIN)
-		goto out_restart;
+
+	status = nfs4_handle_exception(server, status, exception);
+	if (exception->retry)
+		status = -EAGAIN;
 out:
 	dprintk("<-- %s\n", __func__);
-	return;
-out_restart:
-	task->tk_status = 0;
-	rpc_restart_call_prepare(task);
-	return;
-out_overflow:
-	task->tk_status = -EOVERFLOW;
-	goto out;
+	return status;
 }
 
 static size_t max_response_pages(struct nfs_server *server)
@@ -8013,7 +8017,7 @@ static const struct rpc_call_ops nfs4_layoutget_call_ops = {
 };
 
 struct pnfs_layout_segment *
-nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
+nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout, gfp_t gfp_flags)
 {
 	struct inode *inode = lgp->args.inode;
 	struct nfs_server *server = NFS_SERVER(inode);
@@ -8033,6 +8037,7 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 		.flags = RPC_TASK_ASYNC,
 	};
 	struct pnfs_layout_segment *lseg = NULL;
+	struct nfs4_exception exception = { .timeout = *timeout };
 	int status = 0;
 
 	dprintk("--> %s\n", __func__);
@@ -8046,7 +8051,6 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 		return ERR_PTR(-ENOMEM);
 	}
 	lgp->args.layout.pglen = max_pages * PAGE_SIZE;
-	lgp->args.timestamp = jiffies;
 
 	lgp->res.layoutp = &lgp->args.layout;
 	lgp->res.seq_res.sr_slot = NULL;
@@ -8056,13 +8060,17 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, gfp_t gfp_flags)
 	if (IS_ERR(task))
 		return ERR_CAST(task);
 	status = nfs4_wait_for_completion_rpc_task(task);
-	if (status == 0)
-		status = task->tk_status;
+	if (status == 0) {
+		status = nfs4_layoutget_handle_exception(task, lgp, &exception);
+		*timeout = exception.timeout;
+	}
+
 	trace_nfs4_layoutget(lgp->args.ctx,
 			&lgp->args.range,
 			&lgp->res.range,
 			&lgp->res.stateid,
 			status);
+
 	/* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
 	if (status == 0 && lgp->res.layoutp->len)
 		lseg = pnfs_layout_process(lgp);
@@ -8118,7 +8126,8 @@ static void nfs4_layoutreturn_release(void *calldata)
 
 	dprintk("--> %s\n", __func__);
 	spin_lock(&lo->plh_inode->i_lock);
-	pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range);
+	pnfs_mark_matching_lsegs_invalid(lo, &freeme, &lrp->args.range,
+			be32_to_cpu(lrp->args.stateid.seqid));
 	pnfs_mark_layout_returned_if_empty(lo);
 	if (lrp->res.lrs_present)
 		pnfs_set_layout_stateid(lo, &lrp->res.stateid, true);
@@ -8653,6 +8662,9 @@ nfs41_free_lock_state(struct nfs_server *server, struct nfs4_lock_state *lsp)
 static bool nfs41_match_stateid(const nfs4_stateid *s1,
 		const nfs4_stateid *s2)
 {
+	if (s1->type != s2->type)
+		return false;
+
 	if (memcmp(s1->other, s2->other, sizeof(s1->other)) != 0)
 		return false;
 
@@ -8793,6 +8805,7 @@ static const struct nfs4_minor_version_ops nfs_v4_2_minor_ops = {
 		| NFS_CAP_STATEID_NFSV41
 		| NFS_CAP_ATOMIC_OPEN_V1
 		| NFS_CAP_ALLOCATE
+		| NFS_CAP_COPY
 		| NFS_CAP_DEALLOCATE
 		| NFS_CAP_SEEK
 		| NFS_CAP_LAYOUTSTATS