1 files changed, 472 insertions, 453 deletions

diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 9a3b6a4cd6b..5e2f1030454 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -42,10 +42,11 @@
  * Local function declarations
  */
 static void nfs_redirty_request(struct nfs_page *req);
-static const struct rpc_call_ops nfs_write_common_ops;
 static const struct rpc_call_ops nfs_commit_ops;
 static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops;
 static const struct nfs_commit_completion_ops nfs_commit_completion_ops;
+static const struct nfs_rw_ops nfs_rw_write_ops;
+static void nfs_clear_request_commit(struct nfs_page *req);
 
 static struct kmem_cache *nfs_wdata_cachep;
 static mempool_t *nfs_wdata_mempool;
@@ -70,76 +71,19 @@ void nfs_commit_free(struct nfs_commit_data *p)
 }
 EXPORT_SYMBOL_GPL(nfs_commit_free);
 
-struct nfs_write_header *nfs_writehdr_alloc(void)
+static struct nfs_rw_header *nfs_writehdr_alloc(void)
 {
-	struct nfs_write_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
-
-	if (p) {
-		struct nfs_pgio_header *hdr = &p->header;
+	struct nfs_rw_header *p = mempool_alloc(nfs_wdata_mempool, GFP_NOIO);
 
+	if (p)
 		memset(p, 0, sizeof(*p));
-		INIT_LIST_HEAD(&hdr->pages);
-		INIT_LIST_HEAD(&hdr->rpc_list);
-		spin_lock_init(&hdr->lock);
-		atomic_set(&hdr->refcnt, 0);
-		hdr->verf = &p->verf;
-	}
 	return p;
 }
-EXPORT_SYMBOL_GPL(nfs_writehdr_alloc);
 
-static struct nfs_write_data *nfs_writedata_alloc(struct nfs_pgio_header *hdr,
-						  unsigned int pagecount)
+static void nfs_writehdr_free(struct nfs_rw_header *whdr)
 {
-	struct nfs_write_data *data, *prealloc;
-
-	prealloc = &container_of(hdr, struct nfs_write_header, header)->rpc_data;
-	if (prealloc->header == NULL)
-		data = prealloc;
-	else
-		data = kzalloc(sizeof(*data), GFP_KERNEL);
-	if (!data)
-		goto out;
-
-	if (nfs_pgarray_set(&data->pages, pagecount)) {
-		data->header = hdr;
-		atomic_inc(&hdr->refcnt);
-	} else {
-		if (data != prealloc)
-			kfree(data);
-		data = NULL;
-	}
-out:
-	return data;
-}
-
-void nfs_writehdr_free(struct nfs_pgio_header *hdr)
-{
-	struct nfs_write_header *whdr = container_of(hdr, struct nfs_write_header, header);
 	mempool_free(whdr, nfs_wdata_mempool);
 }
-EXPORT_SYMBOL_GPL(nfs_writehdr_free);
-
-void nfs_writedata_release(struct nfs_write_data *wdata)
-{
-	struct nfs_pgio_header *hdr = wdata->header;
-	struct nfs_write_header *write_header = container_of(hdr, struct nfs_write_header, header);
-
-	put_nfs_open_context(wdata->args.context);
-	if (wdata->pages.pagevec != wdata->pages.page_array)
-		kfree(wdata->pages.pagevec);
-	if (wdata == &write_header->rpc_data) {
-		wdata->header = NULL;
-		wdata = NULL;
-	}
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
-	/* Note: we only free the rpc_task after callbacks are done.
-	 * See the comment in rpc_free_task() for why
-	 */
-	kfree(wdata);
-}
-EXPORT_SYMBOL_GPL(nfs_writedata_release);
 
 static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error)
 {
@@ -148,8 +92,15 @@ static void nfs_context_set_write_error(struct nfs_open_context *ctx, int error)
 	set_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
 }
 
+/*
+ * nfs_page_find_head_request_locked - find head request associated with @page
+ *
+ * must be called while holding the inode lock.
+ *
+ * returns matching head request with reference held, or NULL if not found.
+ */
 static struct nfs_page *
-nfs_page_find_request_locked(struct nfs_inode *nfsi, struct page *page)
+nfs_page_find_head_request_locked(struct nfs_inode *nfsi, struct page *page)
 {
 	struct nfs_page *req = NULL;
 
@@ -161,25 +112,33 @@ nfs_page_find_request_locked(struct nfs_inode *nfsi, struct page *page)
 		/* Linearly search the commit list for the correct req */
 		list_for_each_entry_safe(freq, t, &nfsi->commit_info.list, wb_list) {
 			if (freq->wb_page == page) {
-				req = freq;
+				req = freq->wb_head;
 				break;
 			}
 		}
 	}
 
-	if (req)
+	if (req) {
+		WARN_ON_ONCE(req->wb_head != req);
+
 		kref_get(&req->wb_kref);
+	}
 
 	return req;
 }
 
-static struct nfs_page *nfs_page_find_request(struct page *page)
+/*
+ * nfs_page_find_head_request - find head request associated with @page
+ *
+ * returns matching head request with reference held, or NULL if not found.
+ */
+static struct nfs_page *nfs_page_find_head_request(struct page *page)
 {
 	struct inode *inode = page_file_mapping(page)->host;
 	struct nfs_page *req = NULL;
 
 	spin_lock(&inode->i_lock);
-	req = nfs_page_find_request_locked(NFS_I(inode), page);
+	req = nfs_page_find_head_request_locked(NFS_I(inode), page);
 	spin_unlock(&inode->i_lock);
 	return req;
 }
@@ -211,18 +170,78 @@ static void nfs_set_pageerror(struct page *page)
 	nfs_zap_mapping(page_file_mapping(page)->host, page_file_mapping(page));
 }
 
+/*
+ * nfs_page_group_search_locked
+ * @head - head request of page group
+ * @page_offset - offset into page
+ *
+ * Search page group with head @head to find a request that contains the
+ * page offset @page_offset.
+ *
+ * Returns a pointer to the first matching nfs request, or NULL if no
+ * match is found.
+ *
+ * Must be called with the page group lock held
+ */
+static struct nfs_page *
+nfs_page_group_search_locked(struct nfs_page *head, unsigned int page_offset)
+{
+	struct nfs_page *req;
+
+	WARN_ON_ONCE(head != head->wb_head);
+	WARN_ON_ONCE(!test_bit(PG_HEADLOCK, &head->wb_head->wb_flags));
+
+	req = head;
+	do {
+		if (page_offset >= req->wb_pgbase &&
+		    page_offset < (req->wb_pgbase + req->wb_bytes))
+			return req;
+
+		req = req->wb_this_page;
+	} while (req != head);
+
+	return NULL;
+}
+
+/*
+ * nfs_page_group_covers_page
+ * @head - head request of page group
+ *
+ * Return true if the page group with head @head covers the whole page,
+ * returns false otherwise
+ */
+static bool nfs_page_group_covers_page(struct nfs_page *req)
+{
+	struct nfs_page *tmp;
+	unsigned int pos = 0;
+	unsigned int len = nfs_page_length(req->wb_page);
+
+	nfs_page_group_lock(req);
+
+	do {
+		tmp = nfs_page_group_search_locked(req->wb_head, pos);
+		if (tmp) {
+			/* no way this should happen */
+			WARN_ON_ONCE(tmp->wb_pgbase != pos);
+			pos += tmp->wb_bytes - (pos - tmp->wb_pgbase);
+		}
+	} while (tmp && pos < len);
+
+	nfs_page_group_unlock(req);
+	WARN_ON_ONCE(pos > len);
+	return pos == len;
+}
+
 /* We can set the PG_uptodate flag if we see that a write request
  * covers the full page.
  */
-static void nfs_mark_uptodate(struct page *page, unsigned int base, unsigned int count)
+static void nfs_mark_uptodate(struct nfs_page *req)
 {
-	if (PageUptodate(page))
-		return;
-	if (base != 0)
+	if (PageUptodate(req->wb_page))
 		return;
-	if (count != nfs_page_length(page))
+	if (!nfs_page_group_covers_page(req))
 		return;
-	SetPageUptodate(page);
+	SetPageUptodate(req->wb_page);
 }
 
 static int wb_priority(struct writeback_control *wbc)
@@ -258,46 +277,259 @@ static void nfs_set_page_writeback(struct page *page)
 	}
 }
 
-static void nfs_end_page_writeback(struct page *page)
+static void nfs_end_page_writeback(struct nfs_page *req)
 {
-	struct inode *inode = page_file_mapping(page)->host;
+	struct inode *inode = page_file_mapping(req->wb_page)->host;
 	struct nfs_server *nfss = NFS_SERVER(inode);
 
-	end_page_writeback(page);
+	if (!nfs_page_group_sync_on_bit(req, PG_WB_END))
+		return;
+
+	end_page_writeback(req->wb_page);
 	if (atomic_long_dec_return(&nfss->writeback) < NFS_CONGESTION_OFF_THRESH)
 		clear_bdi_congested(&nfss->backing_dev_info, BLK_RW_ASYNC);
 }
 
-static struct nfs_page *nfs_find_and_lock_request(struct page *page, bool nonblock)
+
+/* nfs_page_group_clear_bits
+ *   @req - an nfs request
+ * clears all page group related bits from @req
+ */
+static void
+nfs_page_group_clear_bits(struct nfs_page *req)
+{
+	clear_bit(PG_TEARDOWN, &req->wb_flags);
+	clear_bit(PG_UNLOCKPAGE, &req->wb_flags);
+	clear_bit(PG_UPTODATE, &req->wb_flags);
+	clear_bit(PG_WB_END, &req->wb_flags);
+	clear_bit(PG_REMOVE, &req->wb_flags);
+}
+
+
+/*
+ * nfs_unroll_locks_and_wait -  unlock all newly locked reqs and wait on @req
+ *
+ * this is a helper function for nfs_lock_and_join_requests
+ *
+ * @inode - inode associated with request page group, must be holding inode lock
+ * @head  - head request of page group, must be holding head lock
+ * @req   - request that couldn't lock and needs to wait on the req bit lock
+ * @nonblock - if true, don't actually wait
+ *
+ * NOTE: this must be called holding page_group bit lock and inode spin lock
+ *       and BOTH will be released before returning.
+ *
+ * returns 0 on success, < 0 on error.
+ */
+static int
+nfs_unroll_locks_and_wait(struct inode *inode, struct nfs_page *head,
+			  struct nfs_page *req, bool nonblock)
+	__releases(&inode->i_lock)
+{
+	struct nfs_page *tmp;
+	int ret;
+
+	/* relinquish all the locks successfully grabbed this run */
+	for (tmp = head ; tmp != req; tmp = tmp->wb_this_page)
+		nfs_unlock_request(tmp);
+
+	WARN_ON_ONCE(test_bit(PG_TEARDOWN, &req->wb_flags));
+
+	/* grab a ref on the request that will be waited on */
+	kref_get(&req->wb_kref);
+
+	nfs_page_group_unlock(head);
+	spin_unlock(&inode->i_lock);
+
+	/* release ref from nfs_page_find_head_request_locked */
+	nfs_release_request(head);
+
+	if (!nonblock)
+		ret = nfs_wait_on_request(req);
+	else
+		ret = -EAGAIN;
+	nfs_release_request(req);
+
+	return ret;
+}
+
+/*
+ * nfs_destroy_unlinked_subrequests - destroy recently unlinked subrequests
+ *
+ * @destroy_list - request list (using wb_this_page) terminated by @old_head
+ * @old_head - the old head of the list
+ *
+ * All subrequests must be locked and removed from all lists, so at this point
+ * they are only "active" in this function, and possibly in nfs_wait_on_request
+ * with a reference held by some other context.
+ */
+static void
+nfs_destroy_unlinked_subrequests(struct nfs_page *destroy_list,
+				 struct nfs_page *old_head)
+{
+	while (destroy_list) {
+		struct nfs_page *subreq = destroy_list;
+
+		destroy_list = (subreq->wb_this_page == old_head) ?
+				   NULL : subreq->wb_this_page;
+
+		WARN_ON_ONCE(old_head != subreq->wb_head);
+
+		/* make sure old group is not used */
+		subreq->wb_head = subreq;
+		subreq->wb_this_page = subreq;
+
+		nfs_clear_request_commit(subreq);
+
+		/* subreq is now totally disconnected from page group or any
+		 * write / commit lists. last chance to wake any waiters */
+		nfs_unlock_request(subreq);
+
+		if (!test_bit(PG_TEARDOWN, &subreq->wb_flags)) {
+			/* release ref on old head request */
+			nfs_release_request(old_head);
+
+			nfs_page_group_clear_bits(subreq);
+
+			/* release the PG_INODE_REF reference */
+			if (test_and_clear_bit(PG_INODE_REF, &subreq->wb_flags))
+				nfs_release_request(subreq);
+			else
+				WARN_ON_ONCE(1);
+		} else {
+			WARN_ON_ONCE(test_bit(PG_CLEAN, &subreq->wb_flags));
+			/* zombie requests have already released the last
+			 * reference and were waiting on the rest of the
+			 * group to complete. Since it's no longer part of a
+			 * group, simply free the request */
+			nfs_page_group_clear_bits(subreq);
+			nfs_free_request(subreq);
+		}
+	}
+}
+
+/*
+ * nfs_lock_and_join_requests - join all subreqs to the head req and return
+ *                              a locked reference, cancelling any pending
+ *                              operations for this page.
+ *
+ * @page - the page used to lookup the "page group" of nfs_page structures
+ * @nonblock - if true, don't block waiting for request locks
+ *
+ * This function joins all sub requests to the head request by first
+ * locking all requests in the group, cancelling any pending operations
+ * and finally updating the head request to cover the whole range covered by
+ * the (former) group.  All subrequests are removed from any write or commit
+ * lists, unlinked from the group and destroyed.
+ *
+ * Returns a locked, referenced pointer to the head request - which after
+ * this call is guaranteed to be the only request associated with the page.
+ * Returns NULL if no requests are found for @page, or a ERR_PTR if an
+ * error was encountered.
+ */
+static struct nfs_page *
+nfs_lock_and_join_requests(struct page *page, bool nonblock)
 {
 	struct inode *inode = page_file_mapping(page)->host;
-	struct nfs_page *req;
+	struct nfs_page *head, *subreq;
+	struct nfs_page *destroy_list = NULL;
+	unsigned int total_bytes;
 	int ret;
 
+try_again:
+	total_bytes = 0;
+
+	WARN_ON_ONCE(destroy_list);
+
 	spin_lock(&inode->i_lock);
-	for (;;) {
-		req = nfs_page_find_request_locked(NFS_I(inode), page);
-		if (req == NULL)
-			break;
-		if (nfs_lock_request(req))
-			break;
-		/* Note: If we hold the page lock, as is the case in nfs_writepage,
-		 *	 then the call to nfs_lock_request() will always
-		 *	 succeed provided that someone hasn't already marked the
-		 *	 request as dirty (in which case we don't care).
-		 */
+
+	/*
+	 * A reference is taken only on the head request which acts as a
+	 * reference to the whole page group - the group will not be destroyed
+	 * until the head reference is released.
+	 */
+	head = nfs_page_find_head_request_locked(NFS_I(inode), page);
+
+	if (!head) {
 		spin_unlock(&inode->i_lock);
-		if (!nonblock)
-			ret = nfs_wait_on_request(req);
-		else
-			ret = -EAGAIN;
-		nfs_release_request(req);
-		if (ret != 0)
+		return NULL;
+	}
+
+	/* lock each request in the page group */
+	nfs_page_group_lock(head);
+	subreq = head;
+	do {
+		/*
+		 * Subrequests are always contiguous, non overlapping
+		 * and in order. If not, it's a programming error.
+		 */
+		WARN_ON_ONCE(subreq->wb_offset !=
+		     (head->wb_offset + total_bytes));
+
+		/* keep track of how many bytes this group covers */
+		total_bytes += subreq->wb_bytes;
+
+		if (!nfs_lock_request(subreq)) {
+			/* releases page group bit lock and
+			 * inode spin lock and all references */
+			ret = nfs_unroll_locks_and_wait(inode, head,
+				subreq, nonblock);
+
+			if (ret == 0)
+				goto try_again;
+
 			return ERR_PTR(ret);
-		spin_lock(&inode->i_lock);
+		}
+
+		subreq = subreq->wb_this_page;
+	} while (subreq != head);
+
+	/* Now that all requests are locked, make sure they aren't on any list.
+	 * Commit list removal accounting is done after locks are dropped */
+	subreq = head;
+	do {
+		nfs_list_remove_request(subreq);
+		subreq = subreq->wb_this_page;
+	} while (subreq != head);
+
+	/* unlink subrequests from head, destroy them later */
+	if (head->wb_this_page != head) {
+		/* destroy list will be terminated by head */
+		destroy_list = head->wb_this_page;
+		head->wb_this_page = head;
+
+		/* change head request to cover whole range that
+		 * the former page group covered */
+		head->wb_bytes = total_bytes;
 	}
+
+	/*
+	 * prepare head request to be added to new pgio descriptor
+	 */
+	nfs_page_group_clear_bits(head);
+
+	/*
+	 * some part of the group was still on the inode list - otherwise
+	 * the group wouldn't be involved in async write.
+	 * grab a reference for the head request, iff it needs one.
+	 */
+	if (!test_and_set_bit(PG_INODE_REF, &head->wb_flags))
+		kref_get(&head->wb_kref);
+
+	nfs_page_group_unlock(head);
+
+	/* drop lock to clear_request_commit the head req and clean up
+	 * requests on destroy list */
 	spin_unlock(&inode->i_lock);
-	return req;
+
+	nfs_destroy_unlinked_subrequests(destroy_list, head);
+
+	/* clean up commit list state */
+	nfs_clear_request_commit(head);
+
+	/* still holds ref on head from nfs_page_find_head_request_locked
+	 * and still has lock on head from lock loop */
+	return head;
 }
 
 /*
@@ -310,7 +542,7 @@ static int nfs_page_async_flush(struct nfs_pageio_descriptor *pgio,
 	struct nfs_page *req;
 	int ret = 0;
 
-	req = nfs_find_and_lock_request(page, nonblock);
+	req = nfs_lock_and_join_requests(page, nonblock);
 	if (!req)
 		goto out;
 	ret = PTR_ERR(req);
@@ -354,10 +586,8 @@ static int nfs_writepage_locked(struct page *page, struct writeback_control *wbc
 	struct nfs_pageio_descriptor pgio;
 	int err;
 
-	NFS_PROTO(page_file_mapping(page)->host)->write_pageio_init(&pgio,
-							  page->mapping->host,
-							  wb_priority(wbc),
-							  &nfs_async_write_completion_ops);
+	nfs_pageio_init_write(&pgio, page->mapping->host, wb_priority(wbc),
+				false, &nfs_async_write_completion_ops);
 	err = nfs_do_writepage(page, wbc, &pgio);
 	nfs_pageio_complete(&pgio);
 	if (err < 0)
@@ -400,12 +630,13 @@ int nfs_writepages(struct address_space *mapping, struct writeback_control *wbc)
 
 	nfs_inc_stats(inode, NFSIOS_VFSWRITEPAGES);
 
-	NFS_PROTO(inode)->write_pageio_init(&pgio, inode, wb_priority(wbc), &nfs_async_write_completion_ops);
+	nfs_pageio_init_write(&pgio, inode, wb_priority(wbc), false,
+				&nfs_async_write_completion_ops);
 	err = write_cache_pages(mapping, wbc, nfs_writepages_callback, &pgio);
 	nfs_pageio_complete(&pgio);
 
 	clear_bit_unlock(NFS_INO_FLUSHING, bitlock);
-	smp_mb__after_clear_bit();
+	smp_mb__after_atomic();
 	wake_up_bit(bitlock, NFS_INO_FLUSHING);
 
 	if (err < 0)
@@ -425,6 +656,8 @@ static void nfs_inode_add_request(struct inode *inode, struct nfs_page *req)
 {
 	struct nfs_inode *nfsi = NFS_I(inode);
 
+	WARN_ON_ONCE(req->wb_this_page != req);
+
 	/* Lock the request! */
 	nfs_lock_request(req);
 
@@ -441,6 +674,9 @@ static void nfs_inode_add_request(struct inode *inode, struct nfs_page *req)
 		set_page_private(req->wb_page, (unsigned long)req);
 	}
 	nfsi->npages++;
+	/* this a head request for a page group - mark it as having an
+	 * extra reference so sub groups can follow suit */
+	WARN_ON(test_and_set_bit(PG_INODE_REF, &req->wb_flags));
 	kref_get(&req->wb_kref);
 	spin_unlock(&inode->i_lock);
 }
@@ -452,16 +688,23 @@ static void nfs_inode_remove_request(struct nfs_page *req)
 {
 	struct inode *inode = req->wb_context->dentry->d_inode;
 	struct nfs_inode *nfsi = NFS_I(inode);
+	struct nfs_page *head;
 
-	spin_lock(&inode->i_lock);
-	if (likely(!PageSwapCache(req->wb_page))) {
-		set_page_private(req->wb_page, 0);
-		ClearPagePrivate(req->wb_page);
-		clear_bit(PG_MAPPED, &req->wb_flags);
+	if (nfs_page_group_sync_on_bit(req, PG_REMOVE)) {
+		head = req->wb_head;
+
+		spin_lock(&inode->i_lock);
+		if (likely(!PageSwapCache(head->wb_page))) {
+			set_page_private(head->wb_page, 0);
+			ClearPagePrivate(head->wb_page);
+			clear_bit(PG_MAPPED, &head->wb_flags);
+		}
+		nfsi->npages--;
+		spin_unlock(&inode->i_lock);
 	}
-	nfsi->npages--;
-	spin_unlock(&inode->i_lock);
-	nfs_release_request(req);
+
+	if (test_and_clear_bit(PG_INODE_REF, &req->wb_flags))
+		nfs_release_request(req);
 }
 
 static void
@@ -583,7 +826,7 @@ nfs_clear_request_commit(struct nfs_page *req)
 }
 
 static inline
-int nfs_write_need_commit(struct nfs_write_data *data)
+int nfs_write_need_commit(struct nfs_pgio_data *data)
 {
 	if (data->verf.committed == NFS_DATA_SYNC)
 		return data->header->lseg == NULL;
@@ -614,7 +857,7 @@ nfs_clear_request_commit(struct nfs_page *req)
 }
 
 static inline
-int nfs_write_need_commit(struct nfs_write_data *data)
+int nfs_write_need_commit(struct nfs_pgio_data *data)
 {
 	return 0;
 }
@@ -645,7 +888,7 @@ static void nfs_write_completion(struct nfs_pgio_header *hdr)
 			goto next;
 		}
 		if (test_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags)) {
-			memcpy(&req->wb_verf, &hdr->verf->verifier, sizeof(req->wb_verf));
+			memcpy(&req->wb_verf, &hdr->verf.verifier, sizeof(req->wb_verf));
 			nfs_mark_request_commit(req, hdr->lseg, &cinfo);
 			goto next;
 		}
@@ -653,7 +896,7 @@ remove_req:
 		nfs_inode_remove_request(req);
 next:
 		nfs_unlock_request(req);
-		nfs_end_page_writeback(req->wb_page);
+		nfs_end_page_writeback(req);
 		nfs_release_request(req);
 	}
 out:
@@ -661,7 +904,7 @@ out:
 }
 
 #if  IS_ENABLED(CONFIG_NFS_V3) || IS_ENABLED(CONFIG_NFS_V4)
-static unsigned long
+unsigned long
 nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
 {
 	return cinfo->mds->ncommit;
@@ -718,7 +961,7 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst,
 }
 
 #else
-static unsigned long nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
+unsigned long nfs_reqs_to_commit(struct nfs_commit_info *cinfo)
 {
 	return 0;
 }
@@ -754,10 +997,14 @@ static struct nfs_page *nfs_try_to_update_request(struct inode *inode,
 	spin_lock(&inode->i_lock);
 
 	for (;;) {
-		req = nfs_page_find_request_locked(NFS_I(inode), page);
+		req = nfs_page_find_head_request_locked(NFS_I(inode), page);
 		if (req == NULL)
 			goto out_unlock;
 
+		/* should be handled by nfs_flush_incompatible */
+		WARN_ON_ONCE(req->wb_head != req);
+		WARN_ON_ONCE(req->wb_this_page != req);
+
 		rqend = req->wb_offset + req->wb_bytes;
 		/*
 		 * Tell the caller to flush out the request if
@@ -819,7 +1066,7 @@ static struct nfs_page * nfs_setup_write_request(struct nfs_open_context* ctx,
 	req = nfs_try_to_update_request(inode, page, offset, bytes);
 	if (req != NULL)
 		goto out;
-	req = nfs_create_request(ctx, inode, page, offset, bytes);
+	req = nfs_create_request(ctx, page, NULL, offset, bytes);
 	if (IS_ERR(req))
 		goto out;
 	nfs_inode_add_request(inode, req);
@@ -837,7 +1084,7 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page,
 		return PTR_ERR(req);
 	/* Update file length */
 	nfs_grow_file(page, offset, count);
-	nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes);
+	nfs_mark_uptodate(req);
 	nfs_mark_request_dirty(req);
 	nfs_unlock_and_release_request(req);
 	return 0;
@@ -858,11 +1105,13 @@ int nfs_flush_incompatible(struct file *file, struct page *page)
 	 * dropped page.
 	 */
 	do {
-		req = nfs_page_find_request(page);
+		req = nfs_page_find_head_request(page);
 		if (req == NULL)
 			return 0;
 		l_ctx = req->wb_lock_context;
 		do_flush = req->wb_page != page || req->wb_context != ctx;
+		/* for now, flush if more than 1 request in page_group */
+		do_flush |= req->wb_this_page != req;
 		if (l_ctx && ctx->dentry->d_inode->i_flock != NULL) {
 			do_flush |= l_ctx->lockowner.l_owner != current->files
 				|| l_ctx->lockowner.l_pid != current->tgid;
@@ -913,12 +1162,14 @@ static bool nfs_write_pageuptodate(struct page *page, struct inode *inode)
 
 	if (nfs_have_delegated_attributes(inode))
 		goto out;
-	if (nfsi->cache_validity & (NFS_INO_INVALID_DATA|NFS_INO_REVAL_PAGECACHE))
+	if (nfsi->cache_validity & NFS_INO_REVAL_PAGECACHE)
 		return false;
 	smp_rmb();
 	if (test_bit(NFS_INO_INVALIDATING, &nfsi->flags))
 		return false;
 out:
+	if (nfsi->cache_validity & NFS_INO_INVALID_DATA)
+		return false;
 	return PageUptodate(page) != 0;
 }
 
@@ -990,126 +1241,17 @@ static int flush_task_priority(int how)
 	return RPC_PRIORITY_NORMAL;
 }
 
-int nfs_initiate_write(struct rpc_clnt *clnt,
-		       struct nfs_write_data *data,
-		       const struct rpc_call_ops *call_ops,
-		       int how, int flags)
+static void nfs_initiate_write(struct nfs_pgio_data *data, struct rpc_message *msg,
+			       struct rpc_task_setup *task_setup_data, int how)
 {
 	struct inode *inode = data->header->inode;
 	int priority = flush_task_priority(how);
-	struct rpc_task *task;
-	struct rpc_message msg = {
-		.rpc_argp = &data->args,
-		.rpc_resp = &data->res,
-		.rpc_cred = data->header->cred,
-	};
-	struct rpc_task_setup task_setup_data = {
-		.rpc_client = clnt,
-		.task = &data->task,
-		.rpc_message = &msg,
-		.callback_ops = call_ops,
-		.callback_data = data,
-		.workqueue = nfsiod_workqueue,
-		.flags = RPC_TASK_ASYNC | flags,
-		.priority = priority,
-	};
-	int ret = 0;
-
-	/* Set up the initial task struct.  */
-	NFS_PROTO(inode)->write_setup(data, &msg);
 
-	dprintk("NFS: %5u initiated write call "
-		"(req %s/%llu, %u bytes @ offset %llu)\n",
-		data->task.tk_pid,
-		inode->i_sb->s_id,
-		(unsigned long long)NFS_FILEID(inode),
-		data->args.count,
-		(unsigned long long)data->args.offset);
+	task_setup_data->priority = priority;
+	NFS_PROTO(inode)->write_setup(data, msg);
 
 	nfs4_state_protect_write(NFS_SERVER(inode)->nfs_client,
-				 &task_setup_data.rpc_client, &msg, data);
-
-	task = rpc_run_task(&task_setup_data);
-	if (IS_ERR(task)) {
-		ret = PTR_ERR(task);
-		goto out;
-	}
-	if (how & FLUSH_SYNC) {
-		ret = rpc_wait_for_completion_task(task);
-		if (ret == 0)
-			ret = task->tk_status;
-	}
-	rpc_put_task(task);
-out:
-	return ret;
-}
-EXPORT_SYMBOL_GPL(nfs_initiate_write);
-
-/*
- * Set up the argument/result storage required for the RPC call.
- */
-static void nfs_write_rpcsetup(struct nfs_write_data *data,
-		unsigned int count, unsigned int offset,
-		int how, struct nfs_commit_info *cinfo)
-{
-	struct nfs_page *req = data->header->req;
-
-	/* Set up the RPC argument and reply structs
-	 * NB: take care not to mess about with data->commit et al. */
-
-	data->args.fh     = NFS_FH(data->header->inode);
-	data->args.offset = req_offset(req) + offset;
-	/* pnfs_set_layoutcommit needs this */
-	data->mds_offset = data->args.offset;
-	data->args.pgbase = req->wb_pgbase + offset;
-	data->args.pages  = data->pages.pagevec;
-	data->args.count  = count;
-	data->args.context = get_nfs_open_context(req->wb_context);
-	data->args.lock_context = req->wb_lock_context;
-	data->args.stable  = NFS_UNSTABLE;
-	switch (how & (FLUSH_STABLE | FLUSH_COND_STABLE)) {
-	case 0:
-		break;
-	case FLUSH_COND_STABLE:
-		if (nfs_reqs_to_commit(cinfo))
-			break;
-	default:
-		data->args.stable = NFS_FILE_SYNC;
-	}
-
-	data->res.fattr   = &data->fattr;
-	data->res.count   = count;
-	data->res.verf    = &data->verf;
-	nfs_fattr_init(&data->fattr);
-}
-
-static int nfs_do_write(struct nfs_write_data *data,
-		const struct rpc_call_ops *call_ops,
-		int how)
-{
-	struct inode *inode = data->header->inode;
-
-	return nfs_initiate_write(NFS_CLIENT(inode), data, call_ops, how, 0);
-}
-
-static int nfs_do_multiple_writes(struct list_head *head,
-		const struct rpc_call_ops *call_ops,
-		int how)
-{
-	struct nfs_write_data *data;
-	int ret = 0;
-
-	while (!list_empty(head)) {
-		int ret2;
-
-		data = list_first_entry(head, struct nfs_write_data, list);
-		list_del_init(&data->list);
-		
-		ret2 = nfs_do_write(data, call_ops, how);
-		 if (ret == 0)
-			 ret = ret2;
-	}
-	return ret;
+				 &task_setup_data->rpc_client, msg, data);
 }
 
 /* If a nfs_flush_* function fails, it should remove reqs from @head and
@@ -1120,7 +1262,7 @@ static void nfs_redirty_request(struct nfs_page *req)
 {
 	nfs_mark_request_dirty(req);
 	nfs_unlock_request(req);
-	nfs_end_page_writeback(req->wb_page);
+	nfs_end_page_writeback(req);
 	nfs_release_request(req);
 }
 
@@ -1140,173 +1282,30 @@ static const struct nfs_pgio_completion_ops nfs_async_write_completion_ops = {
 	.completion = nfs_write_completion,
 };
 
-static void nfs_flush_error(struct nfs_pageio_descriptor *desc,
-		struct nfs_pgio_header *hdr)
-{
-	set_bit(NFS_IOHDR_REDO, &hdr->flags);
-	while (!list_empty(&hdr->rpc_list)) {
-		struct nfs_write_data *data = list_first_entry(&hdr->rpc_list,
-				struct nfs_write_data, list);
-		list_del(&data->list);
-		nfs_writedata_release(data);
-	}
-	desc->pg_completion_ops->error_cleanup(&desc->pg_list);
-}
-
-/*
- * Generate multiple small requests to write out a single
- * contiguous dirty area on one page.
- */
-static int nfs_flush_multi(struct nfs_pageio_descriptor *desc,
-			   struct nfs_pgio_header *hdr)
-{
-	struct nfs_page *req = hdr->req;
-	struct page *page = req->wb_page;
-	struct nfs_write_data *data;
-	size_t wsize = desc->pg_bsize, nbytes;
-	unsigned int offset;
-	int requests = 0;
-	struct nfs_commit_info cinfo;
-
-	nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
-
-	if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
-	    (desc->pg_moreio || nfs_reqs_to_commit(&cinfo) ||
-	     desc->pg_count > wsize))
-		desc->pg_ioflags &= ~FLUSH_COND_STABLE;
-
-
-	offset = 0;
-	nbytes = desc->pg_count;
-	do {
-		size_t len = min(nbytes, wsize);
-
-		data = nfs_writedata_alloc(hdr, 1);
-		if (!data) {
-			nfs_flush_error(desc, hdr);
-			return -ENOMEM;
-		}
-		data->pages.pagevec[0] = page;
-		nfs_write_rpcsetup(data, len, offset, desc->pg_ioflags, &cinfo);
-		list_add(&data->list, &hdr->rpc_list);
-		requests++;
-		nbytes -= len;
-		offset += len;
-	} while (nbytes != 0);
-	nfs_list_remove_request(req);
-	nfs_list_add_request(req, &hdr->pages);
-	desc->pg_rpc_callops = &nfs_write_common_ops;
-	return 0;
-}
-
-/*
- * Create an RPC task for the given write request and kick it.
- * The page must have been locked by the caller.
- *
- * It may happen that the page we're passed is not marked dirty.
- * This is the case if nfs_updatepage detects a conflicting request
- * that has been written but not committed.
- */
-static int nfs_flush_one(struct nfs_pageio_descriptor *desc,
-			 struct nfs_pgio_header *hdr)
-{
-	struct nfs_page		*req;
-	struct page		**pages;
-	struct nfs_write_data	*data;
-	struct list_head *head = &desc->pg_list;
-	struct nfs_commit_info cinfo;
-
-	data = nfs_writedata_alloc(hdr, nfs_page_array_len(desc->pg_base,
-							   desc->pg_count));
-	if (!data) {
-		nfs_flush_error(desc, hdr);
-		return -ENOMEM;
-	}
-
-	nfs_init_cinfo(&cinfo, desc->pg_inode, desc->pg_dreq);
-	pages = data->pages.pagevec;
-	while (!list_empty(head)) {
-		req = nfs_list_entry(head->next);
-		nfs_list_remove_request(req);
-		nfs_list_add_request(req, &hdr->pages);
-		*pages++ = req->wb_page;
-	}
-
-	if ((desc->pg_ioflags & FLUSH_COND_STABLE) &&
-	    (desc->pg_moreio || nfs_reqs_to_commit(&cinfo)))
-		desc->pg_ioflags &= ~FLUSH_COND_STABLE;
-
-	/* Set up the argument struct */
-	nfs_write_rpcsetup(data, desc->pg_count, 0, desc->pg_ioflags, &cinfo);
-	list_add(&data->list, &hdr->rpc_list);
-	desc->pg_rpc_callops = &nfs_write_common_ops;
-	return 0;
-}
-
-int nfs_generic_flush(struct nfs_pageio_descriptor *desc,
-		      struct nfs_pgio_header *hdr)
-{
-	if (desc->pg_bsize < PAGE_CACHE_SIZE)
-		return nfs_flush_multi(desc, hdr);
-	return nfs_flush_one(desc, hdr);
-}
-EXPORT_SYMBOL_GPL(nfs_generic_flush);
-
-static int nfs_generic_pg_writepages(struct nfs_pageio_descriptor *desc)
-{
-	struct nfs_write_header *whdr;
-	struct nfs_pgio_header *hdr;
-	int ret;
-
-	whdr = nfs_writehdr_alloc();
-	if (!whdr) {
-		desc->pg_completion_ops->error_cleanup(&desc->pg_list);
-		return -ENOMEM;
-	}
-	hdr = &whdr->header;
-	nfs_pgheader_init(desc, hdr, nfs_writehdr_free);
-	atomic_inc(&hdr->refcnt);
-	ret = nfs_generic_flush(desc, hdr);
-	if (ret == 0)
-		ret = nfs_do_multiple_writes(&hdr->rpc_list,
-					     desc->pg_rpc_callops,
-					     desc->pg_ioflags);
-	if (atomic_dec_and_test(&hdr->refcnt))
-		hdr->completion_ops->completion(hdr);
-	return ret;
-}
-
-static const struct nfs_pageio_ops nfs_pageio_write_ops = {
-	.pg_test = nfs_generic_pg_test,
-	.pg_doio = nfs_generic_pg_writepages,
-};
-
 void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio,
-			       struct inode *inode, int ioflags,
+			       struct inode *inode, int ioflags, bool force_mds,
 			       const struct nfs_pgio_completion_ops *compl_ops)
 {
-	nfs_pageio_init(pgio, inode, &nfs_pageio_write_ops, compl_ops,
-				NFS_SERVER(inode)->wsize, ioflags);
+	struct nfs_server *server = NFS_SERVER(inode);
+	const struct nfs_pageio_ops *pg_ops = &nfs_pgio_rw_ops;
+
+#ifdef CONFIG_NFS_V4_1
+	if (server->pnfs_curr_ld && !force_mds)
+		pg_ops = server->pnfs_curr_ld->pg_write_ops;
+#endif
+	nfs_pageio_init(pgio, inode, pg_ops, compl_ops, &nfs_rw_write_ops,
+			server->wsize, ioflags);
 }
 EXPORT_SYMBOL_GPL(nfs_pageio_init_write);
 
 void nfs_pageio_reset_write_mds(struct nfs_pageio_descriptor *pgio)
 {
-	pgio->pg_ops = &nfs_pageio_write_ops;
+	pgio->pg_ops = &nfs_pgio_rw_ops;
 	pgio->pg_bsize = NFS_SERVER(pgio->pg_inode)->wsize;
 }
 EXPORT_SYMBOL_GPL(nfs_pageio_reset_write_mds);
 
 
-void nfs_write_prepare(struct rpc_task *task, void *calldata)
-{
-	struct nfs_write_data *data = calldata;
-	int err;
-	err = NFS_PROTO(data->header->inode)->write_rpc_prepare(task, data);
-	if (err)
-		rpc_exit(task, err);
-}
-
 void nfs_commit_prepare(struct rpc_task *task, void *calldata)
 {
 	struct nfs_commit_data *data = calldata;
@@ -1314,23 +1313,8 @@ void nfs_commit_prepare(struct rpc_task *task, void *calldata)
 	NFS_PROTO(data->inode)->commit_rpc_prepare(task, data);
 }
 
-/*
- * Handle a write reply that flushes a whole page.
- *
- * FIXME: There is an inherent race with invalidate_inode_pages and
- *	  writebacks since the page->count is kept > 1 for as long
- *	  as the page has a write request pending.
- */
-static void nfs_writeback_done_common(struct rpc_task *task, void *calldata)
-{
-	struct nfs_write_data	*data = calldata;
-
-	nfs_writeback_done(task, data);
-}
-
-static void nfs_writeback_release_common(void *calldata)
+static void nfs_writeback_release_common(struct nfs_pgio_data *data)
 {
-	struct nfs_write_data	*data = calldata;
 	struct nfs_pgio_header *hdr = data->header;
 	int status = data->task.tk_status;
 
@@ -1339,34 +1323,46 @@ static void nfs_writeback_release_common(void *calldata)
 		if (test_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags))
 			; /* Do nothing */
 		else if (!test_and_set_bit(NFS_IOHDR_NEED_COMMIT, &hdr->flags))
-			memcpy(hdr->verf, &data->verf, sizeof(*hdr->verf));
-		else if (memcmp(hdr->verf, &data->verf, sizeof(*hdr->verf)))
+			memcpy(&hdr->verf, &data->verf, sizeof(hdr->verf));
+		else if (memcmp(&hdr->verf, &data->verf, sizeof(hdr->verf)))
 			set_bit(NFS_IOHDR_NEED_RESCHED, &hdr->flags);
 		spin_unlock(&hdr->lock);
 	}
-	nfs_writedata_release(data);
 }
 
-static const struct rpc_call_ops nfs_write_common_ops = {
-	.rpc_call_prepare = nfs_write_prepare,
-	.rpc_call_done = nfs_writeback_done_common,
-	.rpc_release = nfs_writeback_release_common,
-};
+/*
+ * Special version of should_remove_suid() that ignores capabilities.
+ */
+static int nfs_should_remove_suid(const struct inode *inode)
+{
+	umode_t mode = inode->i_mode;
+	int kill = 0;
 
+	/* suid always must be killed */
+	if (unlikely(mode & S_ISUID))
+		kill = ATTR_KILL_SUID;
+
+	/*
+	 * sgid without any exec bits is just a mandatory locking mark; leave
+	 * it alone.  If some exec bits are set, it's a real sgid; kill it.
+	 */
+	if (unlikely((mode & S_ISGID) && (mode & S_IXGRP)))
+		kill |= ATTR_KILL_SGID;
+
+	if (unlikely(kill && S_ISREG(mode)))
+		return kill;
+
+	return 0;
+}
 
 /*
  * This function is called when the WRITE call is complete.
  */
-void nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
+static int nfs_writeback_done(struct rpc_task *task, struct nfs_pgio_data *data,
+			      struct inode *inode)
 {
-	struct nfs_writeargs	*argp = &data->args;
-	struct nfs_writeres	*resp = &data->res;
-	struct inode		*inode = data->header->inode;
 	int status;
 
-	dprintk("NFS: %5u nfs_writeback_done (status %d)\n",
-		task->tk_pid, task->tk_status);
-
 	/*
 	 * ->write_done will attempt to use post-op attributes to detect
 	 * conflicting writes by other clients.  A strict interpretation
@@ -1376,11 +1372,11 @@ void nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
 	 */
 	status = NFS_PROTO(inode)->write_done(task, data);
 	if (status != 0)
-		return;
-	nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, resp->count);
+		return status;
+	nfs_add_stats(inode, NFSIOS_SERVERWRITTENBYTES, data->res.count);
 
 #if IS_ENABLED(CONFIG_NFS_V3) || IS_ENABLED(CONFIG_NFS_V4)
-	if (resp->verf->committed < argp->stable && task->tk_status >= 0) {
+	if (data->res.verf->committed < data->args.stable && task->tk_status >= 0) {
 		/* We tried a write call, but the server did not
 		 * commit data to stable storage even though we
 		 * requested it.
@@ -1396,18 +1392,31 @@ void nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data)
 			dprintk("NFS:       faulty NFS server %s:"
 				" (committed = %d) != (stable = %d)\n",
 				NFS_SERVER(inode)->nfs_client->cl_hostname,
-				resp->verf->committed, argp->stable);
+				data->res.verf->committed, data->args.stable);
 			complain = jiffies + 300 * HZ;
 		}
 	}
 #endif
-	if (task->tk_status < 0)
-		nfs_set_pgio_error(data->header, task->tk_status, argp->offset);
-	else if (resp->count < argp->count) {
+
+	/* Deal with the suid/sgid bit corner case */
+	if (nfs_should_remove_suid(inode))
+		nfs_mark_for_revalidate(inode);
+	return 0;
+}
+
+/*
+ * This function is called when the WRITE call is complete.
+ */
+static void nfs_writeback_result(struct rpc_task *task, struct nfs_pgio_data *data)
+{
+	struct nfs_pgio_args	*argp = &data->args;
+	struct nfs_pgio_res	*resp = &data->res;
+
+	if (resp->count < argp->count) {
 		static unsigned long    complain;
 
 		/* This a short write! */
-		nfs_inc_stats(inode, NFSIOS_SHORTWRITE);
+		nfs_inc_stats(data->header->inode, NFSIOS_SHORTWRITE);
 
 		/* Has the server at least made some progress? */
 		if (resp->count == 0) {
@@ -1458,7 +1467,7 @@ static int nfs_commit_set_lock(struct nfs_inode *nfsi, int may_wait)
 static void nfs_commit_clear_lock(struct nfs_inode *nfsi)
 {
 	clear_bit(NFS_INO_COMMIT, &nfsi->flags);
-	smp_mb__after_clear_bit();
+	smp_mb__after_atomic();
 	wake_up_bit(&nfsi->flags, NFS_INO_COMMIT);
 }
 
@@ -1788,27 +1797,28 @@ int nfs_wb_page_cancel(struct inode *inode, struct page *page)
 	struct nfs_page *req;
 	int ret = 0;
 
-	for (;;) {
-		wait_on_page_writeback(page);
-		req = nfs_page_find_request(page);
-		if (req == NULL)
-			break;
-		if (nfs_lock_request(req)) {
-			nfs_clear_request_commit(req);
-			nfs_inode_remove_request(req);
-			/*
-			 * In case nfs_inode_remove_request has marked the
-			 * page as being dirty
-			 */
-			cancel_dirty_page(page, PAGE_CACHE_SIZE);
-			nfs_unlock_and_release_request(req);
-			break;
-		}
-		ret = nfs_wait_on_request(req);
-		nfs_release_request(req);
-		if (ret < 0)
-			break;
+	wait_on_page_writeback(page);
+
+	/* blocking call to cancel all requests and join to a single (head)
+	 * request */
+	req = nfs_lock_and_join_requests(page, false);
+
+	if (IS_ERR(req)) {
+		ret = PTR_ERR(req);
+	} else if (req) {
+		/* all requests from this page have been cancelled by
+		 * nfs_lock_and_join_requests, so just remove the head
+		 * request from the inode / page_private pointer and
+		 * release it */
+		nfs_inode_remove_request(req);
+		/*
+		 * In case nfs_inode_remove_request has marked the
+		 * page as being dirty
+		 */
+		cancel_dirty_page(page, PAGE_CACHE_SIZE);
+		nfs_unlock_and_release_request(req);
 	}
+
 	return ret;
 }
 
@@ -1874,7 +1884,7 @@ int nfs_migrate_page(struct address_space *mapping, struct page *newpage,
 int __init nfs_init_writepagecache(void)
 {
 	nfs_wdata_cachep = kmem_cache_create("nfs_write_data",
-					     sizeof(struct nfs_write_header),
+					     sizeof(struct nfs_rw_header),
 					     0, SLAB_HWCACHE_ALIGN,
 					     NULL);
 	if (nfs_wdata_cachep == NULL)
@@ -1936,3 +1946,12 @@ void nfs_destroy_writepagecache(void)
 	kmem_cache_destroy(nfs_wdata_cachep);
 }
 
+static const struct nfs_rw_ops nfs_rw_write_ops = {
+	.rw_mode		= FMODE_WRITE,
+	.rw_alloc_header	= nfs_writehdr_alloc,
+	.rw_free_header		= nfs_writehdr_free,
+	.rw_release		= nfs_writeback_release_common,
+	.rw_done		= nfs_writeback_done,
+	.rw_result		= nfs_writeback_result,
+	.rw_initiate		= nfs_initiate_write,
+};