Merge tag 'v4.3-rc4' into for-4.4/core

Linux 4.3-rc4

Pulling in v4.3-rc4 to avoid conflicts with NVMe fixes that have gone
in since for-4.4/core was based.

12 files changed, 148 insertions, 83 deletions

diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 2714ef835bdd..be806ead7f4d 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -113,7 +113,8 @@ out:
 	return status;
 }
 
-static int nfs_delegation_claim_opens(struct inode *inode, const nfs4_stateid *stateid)
+static int nfs_delegation_claim_opens(struct inode *inode,
+		const nfs4_stateid *stateid, fmode_t type)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 	struct nfs_open_context *ctx;
@@ -140,7 +141,7 @@ again:
 		/* Block nfs4_proc_unlck */
 		mutex_lock(&sp->so_delegreturn_mutex);
 		seq = raw_seqcount_begin(&sp->so_reclaim_seqcount);
-		err = nfs4_open_delegation_recall(ctx, state, stateid);
+		err = nfs4_open_delegation_recall(ctx, state, stateid, type);
 		if (!err)
 			err = nfs_delegation_claim_locks(ctx, state, stateid);
 		if (!err && read_seqcount_retry(&sp->so_reclaim_seqcount, seq))
@@ -411,7 +412,8 @@ static int nfs_end_delegation_return(struct inode *inode, struct nfs_delegation
 	do {
 		if (test_bit(NFS_DELEGATION_REVOKED, &delegation->flags))
 			break;
-		err = nfs_delegation_claim_opens(inode, &delegation->stateid);
+		err = nfs_delegation_claim_opens(inode, &delegation->stateid,
+				delegation->type);
 		if (!issync || err != -EAGAIN)
 			break;
 		/*
diff --git a/fs/nfs/delegation.h b/fs/nfs/delegation.h
index a44829173e57..333063e032f0 100644
--- a/fs/nfs/delegation.h
+++ b/fs/nfs/delegation.h
@@ -54,7 +54,7 @@ void nfs_delegation_reap_unclaimed(struct nfs_client *clp);
 
 /* NFSv4 delegation-related procedures */
 int nfs4_proc_delegreturn(struct inode *inode, struct rpc_cred *cred, const nfs4_stateid *stateid, int issync);
-int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid);
+int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid, fmode_t type);
 int nfs4_lock_delegation_recall(struct file_lock *fl, struct nfs4_state *state, const nfs4_stateid *stateid);
 bool nfs4_copy_delegation_stateid(nfs4_stateid *dst, struct inode *inode, fmode_t flags);
 
diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
index 38678d9a5cc4..4b1d08f56aba 100644
--- a/fs/nfs/direct.c
+++ b/fs/nfs/direct.c
@@ -166,8 +166,11 @@ nfs_direct_select_verf(struct nfs_direct_req *dreq,
 	struct nfs_writeverf *verfp = &dreq->verf;
 
 #ifdef CONFIG_NFS_V4_1
-	if (ds_clp) {
-		/* pNFS is in use, use the DS verf */
+	/*
+	 * pNFS is in use, use the DS verf except commit_through_mds is set
+	 * for layout segment where nbuckets is zero.
+	 */
+	if (ds_clp && dreq->ds_cinfo.nbuckets > 0) {
 		if (commit_idx >= 0 && commit_idx < dreq->ds_cinfo.nbuckets)
 			verfp = &dreq->ds_cinfo.buckets[commit_idx].direct_verf;
 		else
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index b34f2e228601..02ec07973bc4 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -629,23 +629,18 @@ out_put:
 	goto out;
 }
 
-static void filelayout_free_fh_array(struct nfs4_filelayout_segment *fl)
+static void _filelayout_free_lseg(struct nfs4_filelayout_segment *fl)
 {
 	int i;
 
-	for (i = 0; i < fl->num_fh; i++) {
-		if (!fl->fh_array[i])
-			break;
-		kfree(fl->fh_array[i]);
+	if (fl->fh_array) {
+		for (i = 0; i < fl->num_fh; i++) {
+			if (!fl->fh_array[i])
+				break;
+			kfree(fl->fh_array[i]);
+		}
+		kfree(fl->fh_array);
 	}
-	kfree(fl->fh_array);
-	fl->fh_array = NULL;
-}
-
-static void
-_filelayout_free_lseg(struct nfs4_filelayout_segment *fl)
-{
-	filelayout_free_fh_array(fl);
 	kfree(fl);
 }
 
@@ -716,21 +711,21 @@ filelayout_decode_layout(struct pnfs_layout_hdr *flo,
 		/* Do we want to use a mempool here? */
 		fl->fh_array[i] = kmalloc(sizeof(struct nfs_fh), gfp_flags);
 		if (!fl->fh_array[i])
-			goto out_err_free;
+			goto out_err;
 
 		p = xdr_inline_decode(&stream, 4);
 		if (unlikely(!p))
-			goto out_err_free;
+			goto out_err;
 		fl->fh_array[i]->size = be32_to_cpup(p++);
 		if (sizeof(struct nfs_fh) < fl->fh_array[i]->size) {
 			printk(KERN_ERR "NFS: Too big fh %d received %d\n",
 			       i, fl->fh_array[i]->size);
-			goto out_err_free;
+			goto out_err;
 		}
 
 		p = xdr_inline_decode(&stream, fl->fh_array[i]->size);
 		if (unlikely(!p))
-			goto out_err_free;
+			goto out_err;
 		memcpy(fl->fh_array[i]->data, p, fl->fh_array[i]->size);
 		dprintk("DEBUG: %s: fh len %d\n", __func__,
 			fl->fh_array[i]->size);
@@ -739,8 +734,6 @@ filelayout_decode_layout(struct pnfs_layout_hdr *flo,
 	__free_page(scratch);
 	return 0;
 
-out_err_free:
-	filelayout_free_fh_array(fl);
 out_err:
 	__free_page(scratch);
 	return -EIO;
diff --git a/fs/nfs/nfs42proc.c b/fs/nfs/nfs42proc.c
index d731bbf974aa..0f020e4d8421 100644
--- a/fs/nfs/nfs42proc.c
+++ b/fs/nfs/nfs42proc.c
@@ -175,10 +175,12 @@ loff_t nfs42_proc_llseek(struct file *filep, loff_t offset, int whence)
 {
 	struct nfs_server *server = NFS_SERVER(file_inode(filep));
 	struct nfs4_exception exception = { };
-	int err;
+	loff_t err;
 
 	do {
 		err = _nfs42_proc_llseek(filep, offset, whence);
+		if (err >= 0)
+			break;
 		if (err == -ENOTSUPP)
 			return -EOPNOTSUPP;
 		err = nfs4_handle_exception(server, err, &exception);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 693b903b48bd..f93b9cdb4934 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -1127,6 +1127,21 @@ static int nfs4_wait_for_completion_rpc_task(struct rpc_task *task)
 	return ret;
 }
 
+static bool nfs4_mode_match_open_stateid(struct nfs4_state *state,
+		fmode_t fmode)
+{
+	switch(fmode & (FMODE_READ|FMODE_WRITE)) {
+	case FMODE_READ|FMODE_WRITE:
+		return state->n_rdwr != 0;
+	case FMODE_WRITE:
+		return state->n_wronly != 0;
+	case FMODE_READ:
+		return state->n_rdonly != 0;
+	}
+	WARN_ON_ONCE(1);
+	return false;
+}
+
 static int can_open_cached(struct nfs4_state *state, fmode_t mode, int open_mode)
 {
 	int ret = 0;
@@ -1571,17 +1586,13 @@ static struct nfs4_opendata *nfs4_open_recoverdata_alloc(struct nfs_open_context
 	return opendata;
 }
 
-static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmode, struct nfs4_state **res)
+static int nfs4_open_recover_helper(struct nfs4_opendata *opendata,
+		fmode_t fmode)
 {
 	struct nfs4_state *newstate;
 	int ret;
 
-	if ((opendata->o_arg.claim == NFS4_OPEN_CLAIM_DELEGATE_CUR ||
-	     opendata->o_arg.claim == NFS4_OPEN_CLAIM_DELEG_CUR_FH) &&
-	    (opendata->o_arg.u.delegation_type & fmode) != fmode)
-		/* This mode can't have been delegated, so we must have
-		 * a valid open_stateid to cover it - not need to reclaim.
-		 */
+	if (!nfs4_mode_match_open_stateid(opendata->state, fmode))
 		return 0;
 	opendata->o_arg.open_flags = 0;
 	opendata->o_arg.fmode = fmode;
@@ -1597,14 +1608,14 @@ static int nfs4_open_recover_helper(struct nfs4_opendata *opendata, fmode_t fmod
 	newstate = nfs4_opendata_to_nfs4_state(opendata);
 	if (IS_ERR(newstate))
 		return PTR_ERR(newstate);
+	if (newstate != opendata->state)
+		ret = -ESTALE;
 	nfs4_close_state(newstate, fmode);
-	*res = newstate;
-	return 0;
+	return ret;
 }
 
 static int nfs4_open_recover(struct nfs4_opendata *opendata, struct nfs4_state *state)
 {
-	struct nfs4_state *newstate;
 	int ret;
 
 	/* Don't trigger recovery in nfs_test_and_clear_all_open_stateid */
@@ -1615,27 +1626,15 @@ static int nfs4_open_recover(struct nfs4_opendata *opendata, struct nfs4_state *
 	clear_bit(NFS_DELEGATED_STATE, &state->flags);
 	clear_bit(NFS_OPEN_STATE, &state->flags);
 	smp_rmb();
-	if (state->n_rdwr != 0) {
-		ret = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE, &newstate);
-		if (ret != 0)
-			return ret;
-		if (newstate != state)
-			return -ESTALE;
-	}
-	if (state->n_wronly != 0) {
-		ret = nfs4_open_recover_helper(opendata, FMODE_WRITE, &newstate);
-		if (ret != 0)
-			return ret;
-		if (newstate != state)
-			return -ESTALE;
-	}
-	if (state->n_rdonly != 0) {
-		ret = nfs4_open_recover_helper(opendata, FMODE_READ, &newstate);
-		if (ret != 0)
-			return ret;
-		if (newstate != state)
-			return -ESTALE;
-	}
+	ret = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE);
+	if (ret != 0)
+		return ret;
+	ret = nfs4_open_recover_helper(opendata, FMODE_WRITE);
+	if (ret != 0)
+		return ret;
+	ret = nfs4_open_recover_helper(opendata, FMODE_READ);
+	if (ret != 0)
+		return ret;
 	/*
 	 * We may have performed cached opens for all three recoveries.
 	 * Check if we need to update the current stateid.
@@ -1759,18 +1758,32 @@ static int nfs4_handle_delegation_recall_error(struct nfs_server *server, struct
 	return err;
 }
 
-int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state *state, const nfs4_stateid *stateid)
+int nfs4_open_delegation_recall(struct nfs_open_context *ctx,
+		struct nfs4_state *state, const nfs4_stateid *stateid,
+		fmode_t type)
 {
 	struct nfs_server *server = NFS_SERVER(state->inode);
 	struct nfs4_opendata *opendata;
-	int err;
+	int err = 0;
 
 	opendata = nfs4_open_recoverdata_alloc(ctx, state,
 			NFS4_OPEN_CLAIM_DELEG_CUR_FH);
 	if (IS_ERR(opendata))
 		return PTR_ERR(opendata);
 	nfs4_stateid_copy(&opendata->o_arg.u.delegation, stateid);
-	err = nfs4_open_recover(opendata, state);
+	clear_bit(NFS_DELEGATED_STATE, &state->flags);
+	switch (type & (FMODE_READ|FMODE_WRITE)) {
+	case FMODE_READ|FMODE_WRITE:
+	case FMODE_WRITE:
+		err = nfs4_open_recover_helper(opendata, FMODE_READ|FMODE_WRITE);
+		if (err)
+			break;
+		err = nfs4_open_recover_helper(opendata, FMODE_WRITE);
+		if (err)
+			break;
+	case FMODE_READ:
+		err = nfs4_open_recover_helper(opendata, FMODE_READ);
+	}
 	nfs4_opendata_put(opendata);
 	return nfs4_handle_delegation_recall_error(server, state, stateid, err);
 }
@@ -2645,6 +2658,15 @@ out:
 	return err;
 }
 
+static bool
+nfs4_wait_on_layoutreturn(struct inode *inode, struct rpc_task *task)
+{
+	if (inode == NULL || !nfs_have_layout(inode))
+		return false;
+
+	return pnfs_wait_on_layoutreturn(inode, task);
+}
+
 struct nfs4_closedata {
 	struct inode *inode;
 	struct nfs4_state *state;
@@ -2763,6 +2785,11 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 		goto out_no_action;
 	}
 
+	if (nfs4_wait_on_layoutreturn(inode, task)) {
+		nfs_release_seqid(calldata->arg.seqid);
+		goto out_wait;
+	}
+
 	if (calldata->arg.fmode == 0)
 		task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE];
 	if (calldata->roc)
@@ -5308,6 +5335,9 @@ static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data)
 
 	d_data = (struct nfs4_delegreturndata *)data;
 
+	if (nfs4_wait_on_layoutreturn(d_data->inode, task))
+		return;
+
 	if (d_data->roc)
 		pnfs_roc_get_barrier(d_data->inode, &d_data->roc_barrier);
 
@@ -7800,39 +7830,46 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata)
 			dprintk("%s: NFS4ERR_RECALLCONFLICT waiting %lu\n",
 				__func__, delay);
 			rpc_delay(task, delay);
-			task->tk_status = 0;
-			rpc_restart_call_prepare(task);
-			goto out; /* Do not call nfs4_async_handle_error() */
+			/* Do not call nfs4_async_handle_error() */
+			goto out_restart;
 		}
 		break;
 	case -NFS4ERR_EXPIRED:
 	case -NFS4ERR_BAD_STATEID:
 		spin_lock(&inode->i_lock);
-		lo = NFS_I(inode)->layout;
-		if (!lo || list_empty(&lo->plh_segs)) {
+		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;
-		} else {
+			break;
+		}
+		lo = NFS_I(inode)->layout;
+		if (lo && nfs4_stateid_match(&lgp->args.stateid,
+					&lo->plh_stateid)) {
 			LIST_HEAD(head);
 
 			/*
 			 * Mark the bad layout state as invalid, then retry
 			 * with the current stateid.
 			 */
+			set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
 			pnfs_mark_matching_lsegs_invalid(lo, &head, NULL);
 			spin_unlock(&inode->i_lock);
 			pnfs_free_lseg_list(&head);
-	
-			task->tk_status = 0;
-			rpc_restart_call_prepare(task);
-		}
+		} else
+			spin_unlock(&inode->i_lock);
+		goto out_restart;
 	}
 	if (nfs4_async_handle_error(task, server, state, NULL) == -EAGAIN)
-		rpc_restart_call_prepare(task);
+		goto out_restart;
 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;
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index da73bc443238..5db324635e92 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -1481,7 +1481,7 @@ restart:
 					spin_unlock(&state->state_lock);
 				}
 				nfs4_put_open_state(state);
-				clear_bit(NFS4CLNT_RECLAIM_NOGRACE,
+				clear_bit(NFS_STATE_RECLAIM_NOGRACE,
 					&state->flags);
 				spin_lock(&sp->so_lock);
 				goto restart;
diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c
index 7c5718ba625e..fe3ddd20ff89 100644
--- a/fs/nfs/pagelist.c
+++ b/fs/nfs/pagelist.c
@@ -508,7 +508,7 @@ size_t nfs_generic_pg_test(struct nfs_pageio_descriptor *desc,
 	 * for it without upsetting the slab allocator.
 	 */
 	if (((mirror->pg_count + req->wb_bytes) >> PAGE_SHIFT) *
-			sizeof(struct page) > PAGE_SIZE)
+			sizeof(struct page *) > PAGE_SIZE)
 		return 0;
 
 	return min(mirror->pg_bsize - mirror->pg_count, (size_t)req->wb_bytes);
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index ba1246433794..8abe27165ad0 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -1104,20 +1104,15 @@ bool pnfs_roc(struct inode *ino)
 			mark_lseg_invalid(lseg, &tmp_list);
 			found = true;
 		}
-	/* pnfs_prepare_layoutreturn() grabs lo ref and it will be put
-	 * in pnfs_roc_release(). We don't really send a layoutreturn but
-	 * still want others to view us like we are sending one!
-	 *
-	 * If pnfs_prepare_layoutreturn() fails, it means someone else is doing
-	 * LAYOUTRETURN, so we proceed like there are no layouts to return.
-	 *
-	 * ROC in three conditions:
+	/* ROC in two conditions:
 	 * 1. there are ROC lsegs
 	 * 2. we don't send layoutreturn
-	 * 3. no others are sending layoutreturn
 	 */
-	if (found && !layoutreturn && pnfs_prepare_layoutreturn(lo))
+	if (found && !layoutreturn) {
+		/* lo ref dropped in pnfs_roc_release() */
+		pnfs_get_layout_hdr(lo);
 		roc = true;
+	}
 
 out_noroc:
 	spin_unlock(&ino->i_lock);
@@ -1172,6 +1167,26 @@ void pnfs_roc_get_barrier(struct inode *ino, u32 *barrier)
 	spin_unlock(&ino->i_lock);
 }
 
+bool pnfs_wait_on_layoutreturn(struct inode *ino, struct rpc_task *task)
+{
+	struct nfs_inode *nfsi = NFS_I(ino);
+        struct pnfs_layout_hdr *lo;
+        bool sleep = false;
+
+	/* we might not have grabbed lo reference. so need to check under
+	 * i_lock */
+        spin_lock(&ino->i_lock);
+        lo = nfsi->layout;
+        if (lo && test_bit(NFS_LAYOUT_RETURN, &lo->plh_flags))
+                sleep = true;
+        spin_unlock(&ino->i_lock);
+
+        if (sleep)
+                rpc_sleep_on(&NFS_SERVER(ino)->roc_rpcwaitq, task, NULL);
+
+        return sleep;
+}
+
 /*
  * Compare two layout segments for sorting into layout cache.
  * We want to preferentially return RW over RO layouts, so ensure those
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 78c9351ff117..d1990e90e7a0 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -270,6 +270,7 @@ bool pnfs_roc(struct inode *ino);
 void pnfs_roc_release(struct inode *ino);
 void pnfs_roc_set_barrier(struct inode *ino, u32 barrier);
 void pnfs_roc_get_barrier(struct inode *ino, u32 *barrier);
+bool pnfs_wait_on_layoutreturn(struct inode *ino, struct rpc_task *task);
 void pnfs_set_layoutcommit(struct inode *, struct pnfs_layout_segment *, loff_t);
 void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data);
 int pnfs_layoutcommit_inode(struct inode *inode, bool sync);
@@ -639,6 +640,12 @@ pnfs_roc_get_barrier(struct inode *ino, u32 *barrier)
 {
 }
 
+static inline bool
+pnfs_wait_on_layoutreturn(struct inode *ino, struct rpc_task *task)
+{
+	return false;
+}
+
 static inline void set_pnfs_layoutdriver(struct nfs_server *s,
 					 const struct nfs_fh *mntfh, u32 id)
 {
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index ae0ff7a11b40..01b8cc8e8cfc 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -72,6 +72,9 @@ void nfs_pageio_reset_read_mds(struct nfs_pageio_descriptor *pgio)
 {
 	struct nfs_pgio_mirror *mirror;
 
+	if (pgio->pg_ops && pgio->pg_ops->pg_cleanup)
+		pgio->pg_ops->pg_cleanup(pgio);
+
 	pgio->pg_ops = &nfs_pgio_rw_ops;
 
 	/* read path should never have more than one mirror */
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 388f48079c43..72624dc4a623 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1351,6 +1351,9 @@ void nfs_pageio_reset_write_mds(struct nfs_pageio_descriptor *pgio)
 {
 	struct nfs_pgio_mirror *mirror;
 
+	if (pgio->pg_ops && pgio->pg_ops->pg_cleanup)
+		pgio->pg_ops->pg_cleanup(pgio);
+
 	pgio->pg_ops = &nfs_pgio_rw_ops;
 
 	nfs_pageio_stop_mirroring(pgio);