1 files changed, 33 insertions, 185 deletions

diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 3b71623..d00260b 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -417,16 +417,6 @@ should_free_lseg(struct pnfs_layout_range *lseg_range,
 	       lo_seg_intersecting(lseg_range, recall_range);
 }
 
-static bool pnfs_lseg_dec_and_remove_zero(struct pnfs_layout_segment *lseg,
-		struct list_head *tmp_list)
-{
-	if (!atomic_dec_and_test(&lseg->pls_refcount))
-		return false;
-	pnfs_layout_remove_lseg(lseg->pls_layout, lseg);
-	list_add(&lseg->pls_list, tmp_list);
-	return true;
-}
-
 /* Returns 1 if lseg is removed from list, 0 otherwise */
 static int mark_lseg_invalid(struct pnfs_layout_segment *lseg,
 			     struct list_head *tmp_list)
@@ -440,8 +430,11 @@ static int mark_lseg_invalid(struct pnfs_layout_segment *lseg,
 		 */
 		dprintk("%s: lseg %p ref %d\n", __func__, lseg,
 			atomic_read(&lseg->pls_refcount));
-		if (pnfs_lseg_dec_and_remove_zero(lseg, tmp_list))
+		if (atomic_dec_and_test(&lseg->pls_refcount)) {
+			pnfs_layout_remove_lseg(lseg->pls_layout, lseg);
+			list_add(&lseg->pls_list, tmp_list);
 			rv = 1;
+		}
 	}
 	return rv;
 }
@@ -512,147 +505,37 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
 }
 EXPORT_SYMBOL_GPL(pnfs_destroy_layout);
 
-static bool
-pnfs_layout_add_bulk_destroy_list(struct inode *inode,
-		struct list_head *layout_list)
-{
-	struct pnfs_layout_hdr *lo;
-	bool ret = false;
-
-	spin_lock(&inode->i_lock);
-	lo = NFS_I(inode)->layout;
-	if (lo != NULL && list_empty(&lo->plh_bulk_destroy)) {
-		pnfs_get_layout_hdr(lo);
-		list_add(&lo->plh_bulk_destroy, layout_list);
-		ret = true;
-	}
-	spin_unlock(&inode->i_lock);
-	return ret;
-}
-
-/* Caller must hold rcu_read_lock and clp->cl_lock */
-static int
-pnfs_layout_bulk_destroy_byserver_locked(struct nfs_client *clp,
-		struct nfs_server *server,
-		struct list_head *layout_list)
-{
-	struct pnfs_layout_hdr *lo, *next;
-	struct inode *inode;
-
-	list_for_each_entry_safe(lo, next, &server->layouts, plh_layouts) {
-		inode = igrab(lo->plh_inode);
-		if (inode == NULL)
-			continue;
-		list_del_init(&lo->plh_layouts);
-		if (pnfs_layout_add_bulk_destroy_list(inode, layout_list))
-			continue;
-		rcu_read_unlock();
-		spin_unlock(&clp->cl_lock);
-		iput(inode);
-		spin_lock(&clp->cl_lock);
-		rcu_read_lock();
-		return -EAGAIN;
-	}
-	return 0;
-}
-
-static int
-pnfs_layout_free_bulk_destroy_list(struct list_head *layout_list,
-		bool is_bulk_recall)
+/*
+ * Called by the state manger to remove all layouts established under an
+ * expired lease.
+ */
+void
+pnfs_destroy_all_layouts(struct nfs_client *clp)
 {
+	struct nfs_server *server;
 	struct pnfs_layout_hdr *lo;
-	struct inode *inode;
-	struct pnfs_layout_range range = {
-		.iomode = IOMODE_ANY,
-		.offset = 0,
-		.length = NFS4_MAX_UINT64,
-	};
-	LIST_HEAD(lseg_list);
-	int ret = 0;
-
-	while (!list_empty(layout_list)) {
-		lo = list_entry(layout_list->next, struct pnfs_layout_hdr,
-				plh_bulk_destroy);
-		dprintk("%s freeing layout for inode %lu\n", __func__,
-			lo->plh_inode->i_ino);
-		inode = lo->plh_inode;
-		spin_lock(&inode->i_lock);
-		list_del_init(&lo->plh_bulk_destroy);
-		lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */
-		if (is_bulk_recall)
-			set_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags);
-		if (pnfs_mark_matching_lsegs_invalid(lo, &lseg_list, &range))
-			ret = -EAGAIN;
-		spin_unlock(&inode->i_lock);
-		pnfs_free_lseg_list(&lseg_list);
-		pnfs_put_layout_hdr(lo);
-		iput(inode);
-	}
-	return ret;
-}
+	LIST_HEAD(tmp_list);
 
-int
-pnfs_destroy_layouts_byfsid(struct nfs_client *clp,
-		struct nfs_fsid *fsid,
-		bool is_recall)
-{
-	struct nfs_server *server;
-	LIST_HEAD(layout_list);
+	nfs4_deviceid_mark_client_invalid(clp);
+	nfs4_deviceid_purge_client(clp);
 
 	spin_lock(&clp->cl_lock);
 	rcu_read_lock();
-restart:
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		if (memcmp(&server->fsid, fsid, sizeof(*fsid)) != 0)
-			continue;
-		if (pnfs_layout_bulk_destroy_byserver_locked(clp,
-				server,
-				&layout_list) != 0)
-			goto restart;
+		if (!list_empty(&server->layouts))
+			list_splice_init(&server->layouts, &tmp_list);
 	}
 	rcu_read_unlock();
 	spin_unlock(&clp->cl_lock);
 
-	if (list_empty(&layout_list))
-		return 0;
-	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
-}
-
-int
-pnfs_destroy_layouts_byclid(struct nfs_client *clp,
-		bool is_recall)
-{
-	struct nfs_server *server;
-	LIST_HEAD(layout_list);
-
-	spin_lock(&clp->cl_lock);
-	rcu_read_lock();
-restart:
-	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
-		if (pnfs_layout_bulk_destroy_byserver_locked(clp,
-					server,
-					&layout_list) != 0)
-			goto restart;
+	while (!list_empty(&tmp_list)) {
+		lo = list_entry(tmp_list.next, struct pnfs_layout_hdr,
+				plh_layouts);
+		dprintk("%s freeing layout for inode %lu\n", __func__,
+			lo->plh_inode->i_ino);
+		list_del_init(&lo->plh_layouts);
+		pnfs_destroy_layout(NFS_I(lo->plh_inode));
 	}
-	rcu_read_unlock();
-	spin_unlock(&clp->cl_lock);
-
-	if (list_empty(&layout_list))
-		return 0;
-	return pnfs_layout_free_bulk_destroy_list(&layout_list, is_recall);
-}
-
-/*
- * Called by the state manger to remove all layouts established under an
- * expired lease.
- */
-void
-pnfs_destroy_all_layouts(struct nfs_client *clp)
-{
-	nfs4_deviceid_mark_client_invalid(clp);
-	nfs4_deviceid_purge_client(clp);
-
-	pnfs_destroy_layouts_byclid(clp, false);
 }
 
 /*
@@ -784,21 +667,6 @@ send_layoutget(struct pnfs_layout_hdr *lo,
 	return lseg;
 }
 
-static void pnfs_clear_layoutcommit(struct inode *inode,
-		struct list_head *head)
-{
-	struct nfs_inode *nfsi = NFS_I(inode);
-	struct pnfs_layout_segment *lseg, *tmp;
-
-	if (!test_and_clear_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags))
-		return;
-	list_for_each_entry_safe(lseg, tmp, &nfsi->layout->plh_segs, pls_list) {
-		if (!test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags))
-			continue;
-		pnfs_lseg_dec_and_remove_zero(lseg, head);
-	}
-}
-
 /*
  * Initiates a LAYOUTRETURN(FILE), and removes the pnfs_layout_hdr
  * when the layout segment list is empty.
@@ -830,7 +698,6 @@ _pnfs_return_layout(struct inode *ino)
 	/* Reference matched in nfs4_layoutreturn_release */
 	pnfs_get_layout_hdr(lo);
 	empty = list_empty(&lo->plh_segs);
-	pnfs_clear_layoutcommit(ino, &tmp_list);
 	pnfs_mark_matching_lsegs_invalid(lo, &tmp_list, NULL);
 	/* Don't send a LAYOUTRETURN if list was initially empty */
 	if (empty) {
@@ -843,6 +710,8 @@ _pnfs_return_layout(struct inode *ino)
 	spin_unlock(&ino->i_lock);
 	pnfs_free_lseg_list(&tmp_list);
 
+	WARN_ON(test_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags));
+
 	lrp = kzalloc(sizeof(*lrp), GFP_KERNEL);
 	if (unlikely(lrp == NULL)) {
 		status = -ENOMEM;
@@ -1019,7 +888,7 @@ alloc_init_layout_hdr(struct inode *ino,
 	atomic_set(&lo->plh_refcount, 1);
 	INIT_LIST_HEAD(&lo->plh_layouts);
 	INIT_LIST_HEAD(&lo->plh_segs);
-	INIT_LIST_HEAD(&lo->plh_bulk_destroy);
+	INIT_LIST_HEAD(&lo->plh_bulk_recall);
 	lo->plh_inode = ino;
 	lo->plh_lc_cred = get_rpccred(ctx->state->owner->so_cred);
 	return lo;
@@ -1443,15 +1312,13 @@ EXPORT_SYMBOL_GPL(pnfs_generic_pg_test);
 
 int pnfs_write_done_resend_to_mds(struct inode *inode,
 				struct list_head *head,
-				const struct nfs_pgio_completion_ops *compl_ops,
-				struct nfs_direct_req *dreq)
+				const struct nfs_pgio_completion_ops *compl_ops)
 {
 	struct nfs_pageio_descriptor pgio;
 	LIST_HEAD(failed);
 
 	/* Resend all requests through the MDS */
 	nfs_pageio_init_write(&pgio, inode, FLUSH_STABLE, compl_ops);
-	pgio.pg_dreq = dreq;
 	while (!list_empty(head)) {
 		struct nfs_page *req = nfs_list_entry(head->next);
 
@@ -1480,13 +1347,13 @@ static void pnfs_ld_handle_write_error(struct nfs_write_data *data)
 	dprintk("pnfs write error = %d\n", hdr->pnfs_error);
 	if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags &
 	    PNFS_LAYOUTRET_ON_ERROR) {
+		clear_bit(NFS_INO_LAYOUTCOMMIT, &NFS_I(hdr->inode)->flags);
 		pnfs_return_layout(hdr->inode);
 	}
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags))
 		data->task.tk_status = pnfs_write_done_resend_to_mds(hdr->inode,
 							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+							hdr->completion_ops);
 }
 
 /*
@@ -1601,15 +1468,13 @@ EXPORT_SYMBOL_GPL(pnfs_generic_pg_writepages);
 
 int pnfs_read_done_resend_to_mds(struct inode *inode,
 				struct list_head *head,
-				const struct nfs_pgio_completion_ops *compl_ops,
-				struct nfs_direct_req *dreq)
+				const struct nfs_pgio_completion_ops *compl_ops)
 {
 	struct nfs_pageio_descriptor pgio;
 	LIST_HEAD(failed);
 
 	/* Resend all requests through the MDS */
 	nfs_pageio_init_read(&pgio, inode, compl_ops);
-	pgio.pg_dreq = dreq;
 	while (!list_empty(head)) {
 		struct nfs_page *req = nfs_list_entry(head->next);
 
@@ -1634,13 +1499,13 @@ static void pnfs_ld_handle_read_error(struct nfs_read_data *data)
 	dprintk("pnfs read error = %d\n", hdr->pnfs_error);
 	if (NFS_SERVER(hdr->inode)->pnfs_curr_ld->flags &
 	    PNFS_LAYOUTRET_ON_ERROR) {
+		clear_bit(NFS_INO_LAYOUTCOMMIT, &NFS_I(hdr->inode)->flags);
 		pnfs_return_layout(hdr->inode);
 	}
 	if (!test_and_set_bit(NFS_IOHDR_REDO, &hdr->flags))
 		data->task.tk_status = pnfs_read_done_resend_to_mds(hdr->inode,
 							&hdr->pages,
-							hdr->completion_ops,
-							hdr->dreq);
+							hdr->completion_ops);
 }
 
 /*
@@ -1766,27 +1631,11 @@ static void pnfs_list_write_lseg(struct inode *inode, struct list_head *listp)
 
 	list_for_each_entry(lseg, &NFS_I(inode)->layout->plh_segs, pls_list) {
 		if (lseg->pls_range.iomode == IOMODE_RW &&
-		    test_and_clear_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags))
+		    test_bit(NFS_LSEG_LAYOUTCOMMIT, &lseg->pls_flags))
 			list_add(&lseg->pls_lc_list, listp);
 	}
 }
 
-static void pnfs_list_write_lseg_done(struct inode *inode, struct list_head *listp)
-{
-	struct pnfs_layout_segment *lseg, *tmp;
-	unsigned long *bitlock = &NFS_I(inode)->flags;
-
-	/* Matched by references in pnfs_set_layoutcommit */
-	list_for_each_entry_safe(lseg, tmp, listp, pls_lc_list) {
-		list_del_init(&lseg->pls_lc_list);
-		pnfs_put_lseg(lseg);
-	}
-
-	clear_bit_unlock(NFS_INO_LAYOUTCOMMITTING, bitlock);
-	smp_mb__after_clear_bit();
-	wake_up_bit(bitlock, NFS_INO_LAYOUTCOMMITTING);
-}
-
 void pnfs_set_lo_fail(struct pnfs_layout_segment *lseg)
 {
 	pnfs_layout_io_set_failed(lseg->pls_layout, lseg->pls_range.iomode);
@@ -1831,7 +1680,6 @@ void pnfs_cleanup_layoutcommit(struct nfs4_layoutcommit_data *data)
 
 	if (nfss->pnfs_curr_ld->cleanup_layoutcommit)
 		nfss->pnfs_curr_ld->cleanup_layoutcommit(data);
-	pnfs_list_write_lseg_done(data->args.inode, &data->lseg_list);
 }
 
 /*