---

yaml
---
r: 223088
b: refs/heads/master
c: d5f04ff
h: refs/heads/master
v: v3

[refs]

@@ -1,2 +1,2 @@
 ---
-refs/heads/master: 8b0f1840a46449e1946fc88860ef3ec8d6b1c2c7
+refs/heads/master: d5f04ff5fba75e3e9607a65f46cfbfbdf8d69ce4

trunk/Documentation/filesystems/Locking

@@ -173,12 +173,13 @@ prototypes:
 	sector_t (*bmap)(struct address_space *, sector_t);
 	int (*invalidatepage) (struct page *, unsigned long);
 	int (*releasepage) (struct page *, int);
+	void (*freepage)(struct page *);
 	int (*direct_IO)(int, struct kiocb *, const struct iovec *iov,
 			loff_t offset, unsigned long nr_segs);
 	int (*launder_page) (struct page *);
 
 locking rules:
-	All except set_page_dirty may block
+	All except set_page_dirty and freepage may block
 
 			BKL	PageLocked(page)	i_mutex
 writepage:		no	yes, unlocks (see below)
@@ -193,6 +194,7 @@ perform_write:		no	n/a			yes
 bmap:			no
 invalidatepage:		no	yes
 releasepage:		no	yes
+freepage:		no	yes
 direct_IO:		no
 launder_page:		no	yes
 
@@ -288,6 +290,9 @@ buffers from the page in preparation for freeing it.  It returns zero to
 indicate that the buffers are (or may be) freeable.  If ->releasepage is zero,
 the kernel assumes that the fs has no private interest in the buffers.
 
+	->freepage() is called when the kernel is done dropping the page
+from the page cache.
+
 	->launder_page() may be called prior to releasing a page if
 it is still found to be dirty. It returns zero if the page was successfully
 cleaned, or an error value if not. Note that in order to prevent the page

trunk/Documentation/filesystems/vfs.txt

@@ -534,6 +534,7 @@ struct address_space_operations {
 	sector_t (*bmap)(struct address_space *, sector_t);
 	int (*invalidatepage) (struct page *, unsigned long);
 	int (*releasepage) (struct page *, int);
+	void (*freepage)(struct page *);
 	ssize_t (*direct_IO)(int, struct kiocb *, const struct iovec *iov,
 			loff_t offset, unsigned long nr_segs);
 	struct page* (*get_xip_page)(struct address_space *, sector_t,
@@ -678,6 +679,12 @@ struct address_space_operations {
         need to ensure this.  Possibly it can clear the PageUptodate
         bit if it cannot free private data yet.
 
+  freepage: freepage is called once the page is no longer visible in
+        the page cache in order to allow the cleanup of any private
+	data. Since it may be called by the memory reclaimer, it
+	should not assume that the original address_space mapping still
+	exists, and it should not block.
+
   direct_IO: called by the generic read/write routines to perform
         direct_IO - that is IO requests which bypass the page cache
         and transfer data directly between the storage and the

trunk/fs/btrfs/disk-io.c

@@ -696,6 +696,7 @@ static int btree_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
 				   __btree_submit_bio_done);
 }
 
+#ifdef CONFIG_MIGRATION
 static int btree_migratepage(struct address_space *mapping,
 			struct page *newpage, struct page *page)
 {
@@ -712,12 +713,9 @@ static int btree_migratepage(struct address_space *mapping,
 	if (page_has_private(page) &&
 	    !try_to_release_page(page, GFP_KERNEL))
 		return -EAGAIN;
-#ifdef CONFIG_MIGRATION
 	return migrate_page(mapping, newpage, page);
-#else
-	return -ENOSYS;
-#endif
 }
+#endif
 
 static int btree_writepage(struct page *page, struct writeback_control *wbc)
 {
@@ -1009,7 +1007,10 @@ static int find_and_setup_root(struct btrfs_root *tree_root,
 	blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
 	root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
 				     blocksize, generation);
-	BUG_ON(!root->node);
+	if (!root->node || !btrfs_buffer_uptodate(root->node, generation)) {
+		free_extent_buffer(root->node);
+		return -EIO;
+	}
 	root->commit_root = btrfs_root_node(root);
 	return 0;
 }

trunk/fs/btrfs/extent-tree.c

@@ -429,6 +429,7 @@ static int caching_kthread(void *data)
 
 static int cache_block_group(struct btrfs_block_group_cache *cache,
 			     struct btrfs_trans_handle *trans,
+			     struct btrfs_root *root,
 			     int load_cache_only)
 {
 	struct btrfs_fs_info *fs_info = cache->fs_info;
@@ -442,9 +443,12 @@ static int cache_block_group(struct btrfs_block_group_cache *cache,
 
 	/*
 	 * We can't do the read from on-disk cache during a commit since we need
-	 * to have the normal tree locking.
+	 * to have the normal tree locking.  Also if we are currently trying to
+	 * allocate blocks for the tree root we can't do the fast caching since
+	 * we likely hold important locks.
 	 */
-	if (!trans->transaction->in_commit) {
+	if (!trans->transaction->in_commit &&
+	    (root && root != root->fs_info->tree_root)) {
 		spin_lock(&cache->lock);
 		if (cache->cached != BTRFS_CACHE_NO) {
 			spin_unlock(&cache->lock);
@@ -2741,6 +2745,7 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 	struct btrfs_root *root = block_group->fs_info->tree_root;
 	struct inode *inode = NULL;
 	u64 alloc_hint = 0;
+	int dcs = BTRFS_DC_ERROR;
 	int num_pages = 0;
 	int retries = 0;
 	int ret = 0;
@@ -2795,6 +2800,8 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 
 	spin_lock(&block_group->lock);
 	if (block_group->cached != BTRFS_CACHE_FINISHED) {
+		/* We're not cached, don't bother trying to write stuff out */
+		dcs = BTRFS_DC_WRITTEN;
 		spin_unlock(&block_group->lock);
 		goto out_put;
 	}
@@ -2821,17 +2828,16 @@ static int cache_save_setup(struct btrfs_block_group_cache *block_group,
 	ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, num_pages,
 					      num_pages, num_pages,
 					      &alloc_hint);
+	if (!ret)
+		dcs = BTRFS_DC_SETUP;
 	btrfs_free_reserved_data_space(inode, num_pages);
 out_put:
 	iput(inode);
 out_free:
 	btrfs_release_path(root, path);
 out:
 	spin_lock(&block_group->lock);
-	if (ret)
-		block_group->disk_cache_state = BTRFS_DC_ERROR;
-	else
-		block_group->disk_cache_state = BTRFS_DC_SETUP;
+	block_group->disk_cache_state = dcs;
 	spin_unlock(&block_group->lock);
 
 	return ret;
@@ -3037,7 +3043,13 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
 
 u64 btrfs_reduce_alloc_profile(struct btrfs_root *root, u64 flags)
 {
-	u64 num_devices = root->fs_info->fs_devices->rw_devices;
+	/*
+	 * we add in the count of missing devices because we want
+	 * to make sure that any RAID levels on a degraded FS
+	 * continue to be honored.
+	 */
+	u64 num_devices = root->fs_info->fs_devices->rw_devices +
+		root->fs_info->fs_devices->missing_devices;
 
 	if (num_devices == 1)
 		flags &= ~(BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID0);
@@ -4080,7 +4092,7 @@ static int update_block_group(struct btrfs_trans_handle *trans,
 		 * space back to the block group, otherwise we will leak space.
 		 */
 		if (!alloc && cache->cached == BTRFS_CACHE_NO)
-			cache_block_group(cache, trans, 1);
+			cache_block_group(cache, trans, NULL, 1);
 
 		byte_in_group = bytenr - cache->key.objectid;
 		WARN_ON(byte_in_group > cache->key.offset);
@@ -4930,11 +4942,31 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 		btrfs_get_block_group(block_group);
 		search_start = block_group->key.objectid;
 
+		/*
+		 * this can happen if we end up cycling through all the
+		 * raid types, but we want to make sure we only allocate
+		 * for the proper type.
+		 */
+		if (!block_group_bits(block_group, data)) {
+		    u64 extra = BTRFS_BLOCK_GROUP_DUP |
+				BTRFS_BLOCK_GROUP_RAID1 |
+				BTRFS_BLOCK_GROUP_RAID10;
+
+			/*
+			 * if they asked for extra copies and this block group
+			 * doesn't provide them, bail.  This does allow us to
+			 * fill raid0 from raid1.
+			 */
+			if ((data & extra) && !(block_group->flags & extra))
+				goto loop;
+		}
+
 have_block_group:
 		if (unlikely(block_group->cached == BTRFS_CACHE_NO)) {
 			u64 free_percent;
 
-			ret = cache_block_group(block_group, trans, 1);
+			ret = cache_block_group(block_group, trans,
+						orig_root, 1);
 			if (block_group->cached == BTRFS_CACHE_FINISHED)
 				goto have_block_group;
 
@@ -4958,7 +4990,8 @@ static noinline int find_free_extent(struct btrfs_trans_handle *trans,
 			if (loop > LOOP_CACHING_NOWAIT ||
 			    (loop > LOOP_FIND_IDEAL &&
 			     atomic_read(&space_info->caching_threads) < 2)) {
-				ret = cache_block_group(block_group, trans, 0);
+				ret = cache_block_group(block_group, trans,
+							orig_root, 0);
 				BUG_ON(ret);
 			}
 			found_uncached_bg = true;
@@ -5515,7 +5548,7 @@ int btrfs_alloc_logged_file_extent(struct btrfs_trans_handle *trans,
 	u64 num_bytes = ins->offset;
 
 	block_group = btrfs_lookup_block_group(root->fs_info, ins->objectid);
-	cache_block_group(block_group, trans, 0);
+	cache_block_group(block_group, trans, NULL, 0);
 	caching_ctl = get_caching_control(block_group);
 
 	if (!caching_ctl) {
@@ -6300,9 +6333,13 @@ int btrfs_drop_snapshot(struct btrfs_root *root,
 					   NULL, NULL);
 		BUG_ON(ret < 0);
 		if (ret > 0) {
-			ret = btrfs_del_orphan_item(trans, tree_root,
-						    root->root_key.objectid);
-			BUG_ON(ret);
+			/* if we fail to delete the orphan item this time
+			 * around, it'll get picked up the next time.
+			 *
+			 * The most common failure here is just -ENOENT.
+			 */
+			btrfs_del_orphan_item(trans, tree_root,
+					      root->root_key.objectid);
 		}
 	}
 
@@ -7878,7 +7915,14 @@ static u64 update_block_group_flags(struct btrfs_root *root, u64 flags)
 	u64 stripped = BTRFS_BLOCK_GROUP_RAID0 |
 		BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_RAID10;
 
-	num_devices = root->fs_info->fs_devices->rw_devices;
+	/*
+	 * we add in the count of missing devices because we want
+	 * to make sure that any RAID levels on a degraded FS
+	 * continue to be honored.
+	 */
+	num_devices = root->fs_info->fs_devices->rw_devices +
+		root->fs_info->fs_devices->missing_devices;
+
 	if (num_devices == 1) {
 		stripped |= BTRFS_BLOCK_GROUP_DUP;
 		stripped = flags & ~stripped;
@@ -8247,7 +8291,6 @@ int btrfs_read_block_groups(struct btrfs_root *root)
 			break;
 		if (ret != 0)
 			goto error;
-
 		leaf = path->nodes[0];
 		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
 		cache = kzalloc(sizeof(*cache), GFP_NOFS);