Merge branch 'cachefs' into next

6 files changed, 152 insertions, 17 deletions

diff --git a/mm/filemap.c b/mm/filemap.c
index 23acefe51808..d5960666990a 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -565,6 +565,24 @@ void wait_on_page_bit(struct page *page, int bit_nr)
 EXPORT_SYMBOL(wait_on_page_bit);
 
 /**
+ * add_page_wait_queue - Add an arbitrary waiter to a page's wait queue
+ * @page - Page defining the wait queue of interest
+ * @waiter - Waiter to add to the queue
+ *
+ * Add an arbitrary @waiter to the wait queue for the nominated @page.
+ */
+void add_page_wait_queue(struct page *page, wait_queue_t *waiter)
+{
+	wait_queue_head_t *q = page_waitqueue(page);
+	unsigned long flags;
+
+	spin_lock_irqsave(&q->lock, flags);
+	__add_wait_queue(q, waiter);
+	spin_unlock_irqrestore(&q->lock, flags);
+}
+EXPORT_SYMBOL_GPL(add_page_wait_queue);
+
+/**
  * unlock_page - unlock a locked page
  * @page: the page
  *
@@ -603,6 +621,23 @@ void end_page_writeback(struct page *page)
 EXPORT_SYMBOL(end_page_writeback);
 
 /**
+ * end_page_owner_priv_2 - Clear PG_owner_priv_2 and wake up any waiters
+ * @page: the page
+ *
+ * Clear PG_owner_priv_2 and wake up any processes waiting for that event.
+ * This is used to indicate - using PG_fscache_write (an alternate name for the
+ * same bit) - that a page has finished being written to the local disk cache.
+ */
+void end_page_owner_priv_2(struct page *page)
+{
+	if (!TestClearPageOwnerPriv2(page))
+		BUG();
+	smp_mb__after_clear_bit();
+	wake_up_page(page, PG_owner_priv_2);
+}
+EXPORT_SYMBOL(end_page_owner_priv_2);
+
+/**
  * __lock_page - get a lock on the page, assuming we need to sleep to get it
  * @page: the page to lock
  *
@@ -2304,6 +2339,67 @@ generic_file_buffered_write(struct kiocb *iocb, const struct iovec *iov,
 }
 EXPORT_SYMBOL(generic_file_buffered_write);
 
+/**
+ * generic_file_buffered_write_one_page - Write a single page of data to an
+ *	inode
+ * @mapping - The address space of the target inode
+ * @index - The target page in the target inode to fill
+ * @source - The data to write into the target page
+ *
+ * Write the data from the source page to the page in the nominated address
+ * space at the @index specified.  Note that the file will not be extended if
+ * the page crosses the EOF marker, in which case only the first part of the
+ * page will be written.
+ *
+ * The @source page does not need to have any association with the file or the
+ * target page offset.
+ */
+int generic_file_buffered_write_one_page(struct address_space *mapping,
+					 pgoff_t index,
+					 struct page *source)
+{
+	const struct address_space_operations *a_ops = mapping->a_ops;
+	struct page *page;
+	unsigned len;
+	loff_t isize, pos;
+	void *fsdata;
+	int ret;
+
+	pos = index;
+	pos <<= PAGE_CACHE_SHIFT;
+
+	len = PAGE_CACHE_SIZE;
+	isize = i_size_read(mapping->host);
+	if ((isize >> PAGE_CACHE_SHIFT) == index)
+		len = isize & (PAGE_CACHE_SIZE - 1);
+
+	ret = pagecache_write_begin(NULL, mapping, pos, len,
+				    AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata);
+	if (ret < 0)
+		goto sync;
+
+	copy_highpage(page, source);
+
+	ret = pagecache_write_end(NULL, mapping, pos, len, len, page, fsdata);
+	if (ret < 0)
+		goto sync;
+
+	balance_dirty_pages_ratelimited(mapping);
+	cond_resched();
+
+sync:
+	/* the caller must handle O_SYNC themselves, but we handle S_SYNC and
+	 * MS_SYNCHRONOUS here */
+	if (unlikely(IS_SYNC(mapping->host)) && !a_ops->writepage)
+		ret = generic_osync_inode(mapping->host, mapping,
+					     OSYNC_METADATA | OSYNC_DATA);
+
+	/* the caller must handle O_DIRECT for themselves */
+
+	return ret;
+}
+EXPORT_SYMBOL(generic_file_buffered_write_one_page);
+
 static ssize_t
 __generic_file_aio_write_nolock(struct kiocb *iocb, const struct iovec *iov,
 				unsigned long nr_segs, loff_t *ppos)
@@ -2464,6 +2560,9 @@ EXPORT_SYMBOL(generic_file_aio_write);
  * (presumably at page->private).  If the release was successful, return `1'.
  * Otherwise return zero.
  *
+ * This may also be called if PG_fscache is set on a page, indicating that the
+ * page is known to the local caching routines.
+ *
  * The @gfp_mask argument specifies whether I/O may be performed to release
  * this page (__GFP_IO), and whether the call may block (__GFP_WAIT & __GFP_FS).
  *
diff --git a/mm/migrate.c b/mm/migrate.c
index a9eff3f092f6..068655d8f883 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -250,7 +250,7 @@ out:
  * The number of remaining references must be:
  * 1 for anonymous pages without a mapping
  * 2 for pages with a mapping
- * 3 for pages with a mapping and PagePrivate set.
+ * 3 for pages with a mapping and PagePrivate/PagePrivate2 set.
  */
 static int migrate_page_move_mapping(struct address_space *mapping,
 		struct page *newpage, struct page *page)
@@ -270,7 +270,7 @@ static int migrate_page_move_mapping(struct address_space *mapping,
 	pslot = radix_tree_lookup_slot(&mapping->page_tree,
  					page_index(page));
 
-	expected_count = 2 + !!PagePrivate(page);
+	expected_count = 2 + !!page_has_private(page);
 	if (page_count(page) != expected_count ||
 			(struct page *)radix_tree_deref_slot(pslot) != page) {
 		spin_unlock_irq(&mapping->tree_lock);
@@ -386,7 +386,7 @@ EXPORT_SYMBOL(fail_migrate_page);
 
 /*
  * Common logic to directly migrate a single page suitable for
- * pages that do not use PagePrivate.
+ * pages that do not use PagePrivate/PagePrivate2.
  *
  * Pages are locked upon entry and exit.
  */
@@ -522,7 +522,7 @@ static int fallback_migrate_page(struct address_space *mapping,
 	 * Buffers may be managed in a filesystem specific way.
 	 * We must have no buffers or drop them.
 	 */
-	if (PagePrivate(page) &&
+	if (page_has_private(page) &&
 	    !try_to_release_page(page, GFP_KERNEL))
 		return -EAGAIN;
 
@@ -655,7 +655,7 @@ static int unmap_and_move(new_page_t get_new_page, unsigned long private,
 	 * free the metadata, so the page can be freed.
 	 */
 	if (!page->mapping) {
-		if (!PageAnon(page) && PagePrivate(page)) {
+		if (!PageAnon(page) && page_has_private(page)) {
 			/*
 			 * Go direct to try_to_free_buffers() here because
 			 * a) that's what try_to_release_page() would do anyway
diff --git a/mm/readahead.c b/mm/readahead.c
index bec83c15a78f..e4e1050a291d 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -44,6 +44,42 @@ EXPORT_SYMBOL_GPL(file_ra_state_init);
 
 #define list_to_page(head) (list_entry((head)->prev, struct page, lru))
 
+/*
+ * see if a page needs releasing upon read_cache_pages() failure
+ * - the caller of read_cache_pages() may have set PG_private or PG_fscache
+ *   before calling, such as the NFS fs marking pages that are cached locally
+ *   on disk, thus we need to give the fs a chance to clean up in the event of
+ *   an error
+ */
+static void read_cache_pages_invalidate_page(struct address_space *mapping,
+					     struct page *page)
+{
+	if (page_has_private(page)) {
+		if (!trylock_page(page))
+			BUG();
+		page->mapping = mapping;
+		do_invalidatepage(page, 0);
+		page->mapping = NULL;
+		unlock_page(page);
+	}
+	page_cache_release(page);
+}
+
+/*
+ * release a list of pages, invalidating them first if need be
+ */
+static void read_cache_pages_invalidate_pages(struct address_space *mapping,
+					      struct list_head *pages)
+{
+	struct page *victim;
+
+	while (!list_empty(pages)) {
+		victim = list_to_page(pages);
+		list_del(&victim->lru);
+		read_cache_pages_invalidate_page(mapping, victim);
+	}
+}
+
 /**
  * read_cache_pages - populate an address space with some pages & start reads against them
  * @mapping: the address_space
@@ -65,14 +101,14 @@ int read_cache_pages(struct address_space *mapping, struct list_head *pages,
 		list_del(&page->lru);
 		if (add_to_page_cache_lru(page, mapping,
 					page->index, GFP_KERNEL)) {
-			page_cache_release(page);
+			read_cache_pages_invalidate_page(mapping, page);
 			continue;
 		}
 		page_cache_release(page);
 
 		ret = filler(data, page);
 		if (unlikely(ret)) {
-			put_pages_list(pages);
+			read_cache_pages_invalidate_pages(mapping, pages);
 			break;
 		}
 		task_io_account_read(PAGE_CACHE_SIZE);
diff --git a/mm/swap.c b/mm/swap.c
index 8adb9feb61e1..4d5a16548439 100644
--- a/mm/swap.c
+++ b/mm/swap.c
@@ -448,8 +448,8 @@ void pagevec_strip(struct pagevec *pvec)
 	for (i = 0; i < pagevec_count(pvec); i++) {
 		struct page *page = pvec->pages[i];
 
-		if (PagePrivate(page) && trylock_page(page)) {
-			if (PagePrivate(page))
+		if (page_has_private(page) && trylock_page(page)) {
+			if (page_has_private(page))
 				try_to_release_page(page, 0);
 			unlock_page(page);
 		}
diff --git a/mm/truncate.c b/mm/truncate.c
index 1229211104f8..55206fab7b99 100644
--- a/mm/truncate.c
+++ b/mm/truncate.c
@@ -50,7 +50,7 @@ void do_invalidatepage(struct page *page, unsigned long offset)
 static inline void truncate_partial_page(struct page *page, unsigned partial)
 {
 	zero_user_segment(page, partial, PAGE_CACHE_SIZE);
-	if (PagePrivate(page))
+	if (page_has_private(page))
 		do_invalidatepage(page, partial);
 }
 
@@ -99,7 +99,7 @@ truncate_complete_page(struct address_space *mapping, struct page *page)
 	if (page->mapping != mapping)
 		return;
 
-	if (PagePrivate(page))
+	if (page_has_private(page))
 		do_invalidatepage(page, 0);
 
 	cancel_dirty_page(page, PAGE_CACHE_SIZE);
@@ -126,7 +126,7 @@ invalidate_complete_page(struct address_space *mapping, struct page *page)
 	if (page->mapping != mapping)
 		return 0;
 
-	if (PagePrivate(page) && !try_to_release_page(page, 0))
+	if (page_has_private(page) && !try_to_release_page(page, 0))
 		return 0;
 
 	clear_page_mlock(page);
@@ -348,7 +348,7 @@ invalidate_complete_page2(struct address_space *mapping, struct page *page)
 	if (page->mapping != mapping)
 		return 0;
 
-	if (PagePrivate(page) && !try_to_release_page(page, GFP_KERNEL))
+	if (page_has_private(page) && !try_to_release_page(page, GFP_KERNEL))
 		return 0;
 
 	spin_lock_irq(&mapping->tree_lock);
@@ -356,7 +356,7 @@ invalidate_complete_page2(struct address_space *mapping, struct page *page)
 		goto failed;
 
 	clear_page_mlock(page);
-	BUG_ON(PagePrivate(page));
+	BUG_ON(page_has_private(page));
 	__remove_from_page_cache(page);
 	spin_unlock_irq(&mapping->tree_lock);
 	page_cache_release(page);	/* pagecache ref */
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 56ddf41149eb..801a55c716a6 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -276,7 +276,7 @@ static inline int page_mapping_inuse(struct page *page)
 
 static inline int is_page_cache_freeable(struct page *page)
 {
-	return page_count(page) - !!PagePrivate(page) == 2;
+	return page_count(page) - !!page_has_private(page) == 2;
 }
 
 static int may_write_to_queue(struct backing_dev_info *bdi)
@@ -360,7 +360,7 @@ static pageout_t pageout(struct page *page, struct address_space *mapping,
 		 * Some data journaling orphaned pages can have
 		 * page->mapping == NULL while being dirty with clean buffers.
 		 */
-		if (PagePrivate(page)) {
+		if (page_has_private(page)) {
 			if (try_to_free_buffers(page)) {
 				ClearPageDirty(page);
 				printk("%s: orphaned page\n", __func__);
@@ -720,7 +720,7 @@ static unsigned long shrink_page_list(struct list_head *page_list,
 		 * process address space (page_count == 1) it can be freed.
 		 * Otherwise, leave the page on the LRU so it is swappable.
 		 */
-		if (PagePrivate(page)) {
+		if (page_has_private(page)) {
 			if (!try_to_release_page(page, sc->gfp_mask))
 				goto activate_locked;
 			if (!mapping && page_count(page) == 1) {