1 files changed, 533 insertions, 99 deletions

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 07b3ac662e19..632f8f3cc9db 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -256,8 +256,7 @@ void btrfs_csum_final(u32 crc, char *result)
 static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
                            int verify)
 {
-        u16 csum_size =
-                btrfs_super_csum_size(&root->fs_info->super_copy);
+        u16 csum_size = btrfs_super_csum_size(root->fs_info->super_copy);
         char *result = NULL;
         unsigned long len;
         unsigned long cur_len;
@@ -367,7 +366,8 @@ static int btree_read_extent_buffer_pages(struct btrfs_root *root,
         clear_bit(EXTENT_BUFFER_CORRUPT, &eb->bflags);
         io_tree = &BTRFS_I(root->fs_info->btree_inode)->io_tree;
         while (1) {
-                ret = read_extent_buffer_pages(io_tree, eb, start, 1,
+                ret = read_extent_buffer_pages(io_tree, eb, start,
+                                               WAIT_COMPLETE,
                                                btree_get_extent, mirror_num);
                 if (!ret &&
                     !verify_parent_transid(io_tree, eb, parent_transid))
@@ -608,11 +608,48 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
         end = min_t(u64, eb->len, PAGE_CACHE_SIZE);
         end = eb->start + end - 1;
 err:
+        if (test_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) {
+                clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags);
+                btree_readahead_hook(root, eb, eb->start, ret);
+        }
+
         free_extent_buffer(eb);
 out:
         return ret;
 }
 
+static int btree_io_failed_hook(struct bio *failed_bio,
+                         struct page *page, u64 start, u64 end,
+                         int mirror_num, struct extent_state *state)
+{
+        struct extent_io_tree *tree;
+        unsigned long len;
+        struct extent_buffer *eb;
+        struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
+
+        tree = &BTRFS_I(page->mapping->host)->io_tree;
+        if (page->private == EXTENT_PAGE_PRIVATE)
+                goto out;
+        if (!page->private)
+                goto out;
+
+        len = page->private >> 2;
+        WARN_ON(len == 0);
+
+        eb = alloc_extent_buffer(tree, start, len, page);
+        if (eb == NULL)
+                goto out;
+
+        if (test_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags)) {
+                clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags);
+                btree_readahead_hook(root, eb, eb->start, -EIO);
+        }
+        free_extent_buffer(eb);
+
+out:
+        return -EIO;    /* we fixed nothing */
+}
+
 static void end_workqueue_bio(struct bio *bio, int err)
 {
         struct end_io_wq *end_io_wq = bio->bi_private;
@@ -908,7 +945,7 @@ static int btree_readpage(struct file *file, struct page *page)
 {
         struct extent_io_tree *tree;
         tree = &BTRFS_I(page->mapping->host)->io_tree;
-        return extent_read_full_page(tree, page, btree_get_extent);
+        return extent_read_full_page(tree, page, btree_get_extent, 0);
 }
 
 static int btree_releasepage(struct page *page, gfp_t gfp_flags)
@@ -974,11 +1011,43 @@ int readahead_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize,
         if (!buf)
                 return 0;
         read_extent_buffer_pages(&BTRFS_I(btree_inode)->io_tree,
-                                 buf, 0, 0, btree_get_extent, 0);
+                                 buf, 0, WAIT_NONE, btree_get_extent, 0);
         free_extent_buffer(buf);
         return ret;
 }
 
+int reada_tree_block_flagged(struct btrfs_root *root, u64 bytenr, u32 blocksize,
+                         int mirror_num, struct extent_buffer **eb)
+{
+        struct extent_buffer *buf = NULL;
+        struct inode *btree_inode = root->fs_info->btree_inode;
+        struct extent_io_tree *io_tree = &BTRFS_I(btree_inode)->io_tree;
+        int ret;
+
+        buf = btrfs_find_create_tree_block(root, bytenr, blocksize);
+        if (!buf)
+                return 0;
+
+        set_bit(EXTENT_BUFFER_READAHEAD, &buf->bflags);
+
+        ret = read_extent_buffer_pages(io_tree, buf, 0, WAIT_PAGE_LOCK,
+                                       btree_get_extent, mirror_num);
+        if (ret) {
+                free_extent_buffer(buf);
+                return ret;
+        }
+
+        if (test_bit(EXTENT_BUFFER_CORRUPT, &buf->bflags)) {
+                free_extent_buffer(buf);
+                return -EIO;
+        } else if (extent_buffer_uptodate(io_tree, buf, NULL)) {
+                *eb = buf;
+        } else {
+                free_extent_buffer(buf);
+        }
+        return 0;
+}
+
 struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
                                             u64 bytenr, u32 blocksize)
 {
@@ -1135,10 +1204,12 @@ static int find_and_setup_root(struct btrfs_root *tree_root,
 
         generation = btrfs_root_generation(&root->root_item);
         blocksize = btrfs_level_size(root, btrfs_root_level(&root->root_item));
+        root->commit_root = NULL;
         root->node = read_tree_block(root, btrfs_root_bytenr(&root->root_item),
                                      blocksize, generation);
         if (!root->node || !btrfs_buffer_uptodate(root->node, generation)) {
                 free_extent_buffer(root->node);
+                root->node = NULL;
                 return -EIO;
         }
         root->commit_root = btrfs_root_node(root);
@@ -1577,6 +1648,235 @@ sleep:
         return 0;
 }
 
+/*
+ * this will find the highest generation in the array of
+ * root backups.  The index of the highest array is returned,
+ * or -1 if we can't find anything.
+ *
+ * We check to make sure the array is valid by comparing the
+ * generation of the latest  root in the array with the generation
+ * in the super block.  If they don't match we pitch it.
+ */
+static int find_newest_super_backup(struct btrfs_fs_info *info, u64 newest_gen)
+{
+        u64 cur;
+        int newest_index = -1;
+        struct btrfs_root_backup *root_backup;
+        int i;
+
+        for (i = 0; i < BTRFS_NUM_BACKUP_ROOTS; i++) {
+                root_backup = info->super_copy->super_roots + i;
+                cur = btrfs_backup_tree_root_gen(root_backup);
+                if (cur == newest_gen)
+                        newest_index = i;
+        }
+
+        /* check to see if we actually wrapped around */
+        if (newest_index == BTRFS_NUM_BACKUP_ROOTS - 1) {
+                root_backup = info->super_copy->super_roots;
+                cur = btrfs_backup_tree_root_gen(root_backup);
+                if (cur == newest_gen)
+                        newest_index = 0;
+        }
+        return newest_index;
+}
+
+
+/*
+ * find the oldest backup so we know where to store new entries
+ * in the backup array.  This will set the backup_root_index
+ * field in the fs_info struct
+ */
+static void find_oldest_super_backup(struct btrfs_fs_info *info,
+                                     u64 newest_gen)
+{
+        int newest_index = -1;
+
+        newest_index = find_newest_super_backup(info, newest_gen);
+        /* if there was garbage in there, just move along */
+        if (newest_index == -1) {
+                info->backup_root_index = 0;
+        } else {
+                info->backup_root_index = (newest_index + 1) % BTRFS_NUM_BACKUP_ROOTS;
+        }
+}
+
+/*
+ * copy all the root pointers into the super backup array.
+ * this will bump the backup pointer by one when it is
+ * done
+ */
+static void backup_super_roots(struct btrfs_fs_info *info)
+{
+        int next_backup;
+        struct btrfs_root_backup *root_backup;
+        int last_backup;
+
+        next_backup = info->backup_root_index;
+        last_backup = (next_backup + BTRFS_NUM_BACKUP_ROOTS - 1) %
+                BTRFS_NUM_BACKUP_ROOTS;
+
+        /*
+         * just overwrite the last backup if we're at the same generation
+         * this happens only at umount
+         */
+        root_backup = info->super_for_commit->super_roots + last_backup;
+        if (btrfs_backup_tree_root_gen(root_backup) ==
+            btrfs_header_generation(info->tree_root->node))
+                next_backup = last_backup;
+
+        root_backup = info->super_for_commit->super_roots + next_backup;
+
+        /*
+         * make sure all of our padding and empty slots get zero filled
+         * regardless of which ones we use today
+         */
+        memset(root_backup, 0, sizeof(*root_backup));
+
+        info->backup_root_index = (next_backup + 1) % BTRFS_NUM_BACKUP_ROOTS;
+
+        btrfs_set_backup_tree_root(root_backup, info->tree_root->node->start);
+        btrfs_set_backup_tree_root_gen(root_backup,
+                               btrfs_header_generation(info->tree_root->node));
+
+        btrfs_set_backup_tree_root_level(root_backup,
+                               btrfs_header_level(info->tree_root->node));
+
+        btrfs_set_backup_chunk_root(root_backup, info->chunk_root->node->start);
+        btrfs_set_backup_chunk_root_gen(root_backup,
+                               btrfs_header_generation(info->chunk_root->node));
+        btrfs_set_backup_chunk_root_level(root_backup,
+                               btrfs_header_level(info->chunk_root->node));
+
+        btrfs_set_backup_extent_root(root_backup, info->extent_root->node->start);
+        btrfs_set_backup_extent_root_gen(root_backup,
+                               btrfs_header_generation(info->extent_root->node));
+        btrfs_set_backup_extent_root_level(root_backup,
+                               btrfs_header_level(info->extent_root->node));
+
+        /*
+         * we might commit during log recovery, which happens before we set
+         * the fs_root.  Make sure it is valid before we fill it in.
+         */
+        if (info->fs_root && info->fs_root->node) {
+                btrfs_set_backup_fs_root(root_backup,
+                                         info->fs_root->node->start);
+                btrfs_set_backup_fs_root_gen(root_backup,
+                               btrfs_header_generation(info->fs_root->node));
+                btrfs_set_backup_fs_root_level(root_backup,
+                               btrfs_header_level(info->fs_root->node));
+        }
+
+        btrfs_set_backup_dev_root(root_backup, info->dev_root->node->start);
+        btrfs_set_backup_dev_root_gen(root_backup,
+                               btrfs_header_generation(info->dev_root->node));
+        btrfs_set_backup_dev_root_level(root_backup,
+                                       btrfs_header_level(info->dev_root->node));
+
+        btrfs_set_backup_csum_root(root_backup, info->csum_root->node->start);
+        btrfs_set_backup_csum_root_gen(root_backup,
+                               btrfs_header_generation(info->csum_root->node));
+        btrfs_set_backup_csum_root_level(root_backup,
+                               btrfs_header_level(info->csum_root->node));
+
+        btrfs_set_backup_total_bytes(root_backup,
+                             btrfs_super_total_bytes(info->super_copy));
+        btrfs_set_backup_bytes_used(root_backup,
+                             btrfs_super_bytes_used(info->super_copy));
+        btrfs_set_backup_num_devices(root_backup,
+                             btrfs_super_num_devices(info->super_copy));
+
+        /*
+         * if we don't copy this out to the super_copy, it won't get remembered
+         * for the next commit
+         */
+        memcpy(&info->super_copy->super_roots,
+               &info->super_for_commit->super_roots,
+               sizeof(*root_backup) * BTRFS_NUM_BACKUP_ROOTS);
+}
+
+/*
+ * this copies info out of the root backup array and back into
+ * the in-memory super block.  It is meant to help iterate through
+ * the array, so you send it the number of backups you've already
+ * tried and the last backup index you used.
+ *
+ * this returns -1 when it has tried all the backups
+ */
+static noinline int next_root_backup(struct btrfs_fs_info *info,
+                                     struct btrfs_super_block *super,
+                                     int *num_backups_tried, int *backup_index)
+{
+        struct btrfs_root_backup *root_backup;
+        int newest = *backup_index;
+
+        if (*num_backups_tried == 0) {
+                u64 gen = btrfs_super_generation(super);
+
+                newest = find_newest_super_backup(info, gen);
+                if (newest == -1)
+                        return -1;
+
+                *backup_index = newest;
+                *num_backups_tried = 1;
+        } else if (*num_backups_tried == BTRFS_NUM_BACKUP_ROOTS) {
+                /* we've tried all the backups, all done */
+                return -1;
+        } else {
+                /* jump to the next oldest backup */
+                newest = (*backup_index + BTRFS_NUM_BACKUP_ROOTS - 1) %
+                        BTRFS_NUM_BACKUP_ROOTS;
+                *backup_index = newest;
+                *num_backups_tried += 1;
+        }
+        root_backup = super->super_roots + newest;
+
+        btrfs_set_super_generation(super,
+                                   btrfs_backup_tree_root_gen(root_backup));
+        btrfs_set_super_root(super, btrfs_backup_tree_root(root_backup));
+        btrfs_set_super_root_level(super,
+                                   btrfs_backup_tree_root_level(root_backup));
+        btrfs_set_super_bytes_used(super, btrfs_backup_bytes_used(root_backup));
+
+        /*
+         * fixme: the total bytes and num_devices need to match or we should
+         * need a fsck
+         */
+        btrfs_set_super_total_bytes(super, btrfs_backup_total_bytes(root_backup));
+        btrfs_set_super_num_devices(super, btrfs_backup_num_devices(root_backup));
+        return 0;
+}
+
+/* helper to cleanup tree roots */
+static void free_root_pointers(struct btrfs_fs_info *info, int chunk_root)
+{
+        free_extent_buffer(info->tree_root->node);
+        free_extent_buffer(info->tree_root->commit_root);
+        free_extent_buffer(info->dev_root->node);
+        free_extent_buffer(info->dev_root->commit_root);
+        free_extent_buffer(info->extent_root->node);
+        free_extent_buffer(info->extent_root->commit_root);
+        free_extent_buffer(info->csum_root->node);
+        free_extent_buffer(info->csum_root->commit_root);
+
+        info->tree_root->node = NULL;
+        info->tree_root->commit_root = NULL;
+        info->dev_root->node = NULL;
+        info->dev_root->commit_root = NULL;
+        info->extent_root->node = NULL;
+        info->extent_root->commit_root = NULL;
+        info->csum_root->node = NULL;
+        info->csum_root->commit_root = NULL;
+
+        if (chunk_root) {
+                free_extent_buffer(info->chunk_root->node);
+                free_extent_buffer(info->chunk_root->commit_root);
+                info->chunk_root->node = NULL;
+                info->chunk_root->commit_root = NULL;
+        }
+}
+
+
 struct btrfs_root *open_ctree(struct super_block *sb,
                               struct btrfs_fs_devices *fs_devices,
                               char *options)
@@ -1590,29 +1890,32 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         u64 features;
         struct btrfs_key location;
         struct buffer_head *bh;
-        struct btrfs_root *extent_root = kzalloc(sizeof(struct btrfs_root),
-                                                 GFP_NOFS);
-        struct btrfs_root *csum_root = kzalloc(sizeof(struct btrfs_root),
-                                                 GFP_NOFS);
+        struct btrfs_super_block *disk_super;
         struct btrfs_root *tree_root = btrfs_sb(sb);
-        struct btrfs_fs_info *fs_info = NULL;
-        struct btrfs_root *chunk_root = kzalloc(sizeof(struct btrfs_root),
-                                                GFP_NOFS);
-        struct btrfs_root *dev_root = kzalloc(sizeof(struct btrfs_root),
-                                              GFP_NOFS);
+        struct btrfs_fs_info *fs_info = tree_root->fs_info;
+        struct btrfs_root *extent_root;
+        struct btrfs_root *csum_root;
+        struct btrfs_root *chunk_root;
+        struct btrfs_root *dev_root;
         struct btrfs_root *log_tree_root;
-
         int ret;
         int err = -EINVAL;
-
-        struct btrfs_super_block *disk_super;
-
-        if (!extent_root || !tree_root || !tree_root->fs_info ||
-            !chunk_root || !dev_root || !csum_root) {
+        int num_backups_tried = 0;
+        int backup_index = 0;
+
+        extent_root = fs_info->extent_root =
+                kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
+        csum_root = fs_info->csum_root =
+                kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
+        chunk_root = fs_info->chunk_root =
+                kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
+        dev_root = fs_info->dev_root =
+                kzalloc(sizeof(struct btrfs_root), GFP_NOFS);
+
+        if (!extent_root || !csum_root || !chunk_root || !dev_root) {
                 err = -ENOMEM;
                 goto fail;
         }
-        fs_info = tree_root->fs_info;
 
         ret = init_srcu_struct(&fs_info->subvol_srcu);
         if (ret) {
@@ -1648,15 +1951,10 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         spin_lock_init(&fs_info->fs_roots_radix_lock);
         spin_lock_init(&fs_info->delayed_iput_lock);
         spin_lock_init(&fs_info->defrag_inodes_lock);
+        spin_lock_init(&fs_info->free_chunk_lock);
         mutex_init(&fs_info->reloc_mutex);
 
         init_completion(&fs_info->kobj_unregister);
-        fs_info->tree_root = tree_root;
-        fs_info->extent_root = extent_root;
-        fs_info->csum_root = csum_root;
-        fs_info->chunk_root = chunk_root;
-        fs_info->dev_root = dev_root;
-        fs_info->fs_devices = fs_devices;
         INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
         INIT_LIST_HEAD(&fs_info->space_info);
         btrfs_mapping_init(&fs_info->mapping_tree);
@@ -1665,8 +1963,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         btrfs_init_block_rsv(&fs_info->trans_block_rsv);
         btrfs_init_block_rsv(&fs_info->chunk_block_rsv);
         btrfs_init_block_rsv(&fs_info->empty_block_rsv);
-        INIT_LIST_HEAD(&fs_info->durable_block_rsv_list);
-        mutex_init(&fs_info->durable_block_rsv_mutex);
+        btrfs_init_block_rsv(&fs_info->delayed_block_rsv);
         atomic_set(&fs_info->nr_async_submits, 0);
         atomic_set(&fs_info->async_delalloc_pages, 0);
         atomic_set(&fs_info->async_submit_draining, 0);
@@ -1677,6 +1974,11 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         fs_info->metadata_ratio = 0;
         fs_info->defrag_inodes = RB_ROOT;
         fs_info->trans_no_join = 0;
+        fs_info->free_chunk_space = 0;
+
+        /* readahead state */
+        INIT_RADIX_TREE(&fs_info->reada_tree, GFP_NOFS & ~__GFP_WAIT);
+        spin_lock_init(&fs_info->reada_lock);
 
         fs_info->thread_pool_size = min_t(unsigned long,
                                           num_online_cpus() + 2, 8);
@@ -1705,7 +2007,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         sb->s_bdi = &fs_info->bdi;
 
         fs_info->btree_inode->i_ino = BTRFS_BTREE_INODE_OBJECTID;
-        fs_info->btree_inode->i_nlink = 1;
+        set_nlink(fs_info->btree_inode, 1);
         /*
          * we set the i_size on the btree inode to the max possible int.
          * the real end of the address space is determined by all of
@@ -1766,14 +2068,14 @@ struct btrfs_root *open_ctree(struct super_block *sb,
                 goto fail_alloc;
         }
 
-        memcpy(&fs_info->super_copy, bh->b_data, sizeof(fs_info->super_copy));
-        memcpy(&fs_info->super_for_commit, &fs_info->super_copy,
-               sizeof(fs_info->super_for_commit));
+        memcpy(fs_info->super_copy, bh->b_data, sizeof(*fs_info->super_copy));
+        memcpy(fs_info->super_for_commit, fs_info->super_copy,
+               sizeof(*fs_info->super_for_commit));
         brelse(bh);
 
-        memcpy(fs_info->fsid, fs_info->super_copy.fsid, BTRFS_FSID_SIZE);
+        memcpy(fs_info->fsid, fs_info->super_copy->fsid, BTRFS_FSID_SIZE);
 
-        disk_super = &fs_info->super_copy;
+        disk_super = fs_info->super_copy;
         if (!btrfs_super_root(disk_super))
                 goto fail_alloc;
 
@@ -1783,6 +2085,13 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         btrfs_check_super_valid(fs_info, sb->s_flags & MS_RDONLY);
 
         /*
+         * run through our array of backup supers and setup
+         * our ring pointer to the oldest one
+         */
+        generation = btrfs_super_generation(disk_super);
+        find_oldest_super_backup(fs_info, generation);
+
+        /*
          * In the long term, we'll store the compression type in the super
          * block, and it'll be used for per file compression control.
          */
@@ -1870,6 +2179,9 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         btrfs_init_workers(&fs_info->delayed_workers, "delayed-meta",
                            fs_info->thread_pool_size,
                            &fs_info->generic_worker);
+        btrfs_init_workers(&fs_info->readahead_workers, "readahead",
+                           fs_info->thread_pool_size,
+                           &fs_info->generic_worker);
 
         /*
          * endios are largely parallel and should have a very
@@ -1880,6 +2192,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 
         fs_info->endio_write_workers.idle_thresh = 2;
         fs_info->endio_meta_write_workers.idle_thresh = 2;
+        fs_info->readahead_workers.idle_thresh = 2;
 
         btrfs_start_workers(&fs_info->workers, 1);
         btrfs_start_workers(&fs_info->generic_worker, 1);
@@ -1893,6 +2206,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         btrfs_start_workers(&fs_info->endio_freespace_worker, 1);
         btrfs_start_workers(&fs_info->delayed_workers, 1);
         btrfs_start_workers(&fs_info->caching_workers, 1);
+        btrfs_start_workers(&fs_info->readahead_workers, 1);
 
         fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
         fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
@@ -1939,7 +2253,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         if (!test_bit(EXTENT_BUFFER_UPTODATE, &chunk_root->node->bflags)) {
                 printk(KERN_WARNING "btrfs: failed to read chunk root on %s\n",
                        sb->s_id);
-                goto fail_chunk_root;
+                goto fail_tree_roots;
         }
         btrfs_set_root_node(&chunk_root->root_item, chunk_root->node);
         chunk_root->commit_root = btrfs_root_node(chunk_root);
@@ -1954,11 +2268,12 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         if (ret) {
                 printk(KERN_WARNING "btrfs: failed to read chunk tree on %s\n",
                        sb->s_id);
-                goto fail_chunk_root;
+                goto fail_tree_roots;
         }
 
         btrfs_close_extra_devices(fs_devices);
 
+retry_root_backup:
         blocksize = btrfs_level_size(tree_root,
                                      btrfs_super_root_level(disk_super));
         generation = btrfs_super_generation(disk_super);
@@ -1966,32 +2281,33 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         tree_root->node = read_tree_block(tree_root,
                                           btrfs_super_root(disk_super),
                                           blocksize, generation);
-        if (!tree_root->node)
-                goto fail_chunk_root;
-        if (!test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) {
+        if (!tree_root->node ||
+            !test_bit(EXTENT_BUFFER_UPTODATE, &tree_root->node->bflags)) {
                 printk(KERN_WARNING "btrfs: failed to read tree root on %s\n",
                        sb->s_id);
-                goto fail_tree_root;
+
+                goto recovery_tree_root;
         }
+
         btrfs_set_root_node(&tree_root->root_item, tree_root->node);
         tree_root->commit_root = btrfs_root_node(tree_root);
 
         ret = find_and_setup_root(tree_root, fs_info,
                                   BTRFS_EXTENT_TREE_OBJECTID, extent_root);
         if (ret)
-                goto fail_tree_root;
+                goto recovery_tree_root;
         extent_root->track_dirty = 1;
 
         ret = find_and_setup_root(tree_root, fs_info,
                                   BTRFS_DEV_TREE_OBJECTID, dev_root);
         if (ret)
-                goto fail_extent_root;
+                goto recovery_tree_root;
         dev_root->track_dirty = 1;
 
         ret = find_and_setup_root(tree_root, fs_info,
                                   BTRFS_CSUM_TREE_OBJECTID, csum_root);
         if (ret)
-                goto fail_dev_root;
+                goto recovery_tree_root;
 
         csum_root->track_dirty = 1;
 
@@ -2124,22 +2440,13 @@ fail_cleaner:
 
 fail_block_groups:
         btrfs_free_block_groups(fs_info);
-        free_extent_buffer(csum_root->node);
-        free_extent_buffer(csum_root->commit_root);
-fail_dev_root:
-        free_extent_buffer(dev_root->node);
-        free_extent_buffer(dev_root->commit_root);
-fail_extent_root:
-        free_extent_buffer(extent_root->node);
-        free_extent_buffer(extent_root->commit_root);
-fail_tree_root:
-        free_extent_buffer(tree_root->node);
-        free_extent_buffer(tree_root->commit_root);
-fail_chunk_root:
-        free_extent_buffer(chunk_root->node);
-        free_extent_buffer(chunk_root->commit_root);
+
+fail_tree_roots:
+        free_root_pointers(fs_info, 1);
+
 fail_sb_buffer:
         btrfs_stop_workers(&fs_info->generic_worker);
+        btrfs_stop_workers(&fs_info->readahead_workers);
         btrfs_stop_workers(&fs_info->fixup_workers);
         btrfs_stop_workers(&fs_info->delalloc_workers);
         btrfs_stop_workers(&fs_info->workers);
@@ -2152,25 +2459,37 @@ fail_sb_buffer:
         btrfs_stop_workers(&fs_info->delayed_workers);
         btrfs_stop_workers(&fs_info->caching_workers);
 fail_alloc:
-        kfree(fs_info->delayed_root);
 fail_iput:
+        btrfs_mapping_tree_free(&fs_info->mapping_tree);
+
         invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
         iput(fs_info->btree_inode);
-
-        btrfs_close_devices(fs_info->fs_devices);
-        btrfs_mapping_tree_free(&fs_info->mapping_tree);
 fail_bdi:
         bdi_destroy(&fs_info->bdi);
 fail_srcu:
         cleanup_srcu_struct(&fs_info->subvol_srcu);
 fail:
-        kfree(extent_root);
-        kfree(tree_root);
-        kfree(fs_info);
-        kfree(chunk_root);
-        kfree(dev_root);
-        kfree(csum_root);
+        btrfs_close_devices(fs_info->fs_devices);
+        free_fs_info(fs_info);
         return ERR_PTR(err);
+
+recovery_tree_root:
+        if (!btrfs_test_opt(tree_root, RECOVERY))
+                goto fail_tree_roots;
+
+        free_root_pointers(fs_info, 0);
+
+        /* don't use the log in recovery mode, it won't be valid */
+        btrfs_set_super_log_root(disk_super, 0);
+
+        /* we can't trust the free space cache either */
+        btrfs_set_opt(fs_info->mount_opt, CLEAR_CACHE);
+
+        ret = next_root_backup(fs_info, fs_info->super_copy,
+                               &num_backups_tried, &backup_index);
+        if (ret == -1)
+                goto fail_block_groups;
+        goto retry_root_backup;
 }
 
 static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
@@ -2254,22 +2573,10 @@ static int write_dev_supers(struct btrfs_device *device,
         int errors = 0;
         u32 crc;
         u64 bytenr;
-        int last_barrier = 0;
 
         if (max_mirrors == 0)
                 max_mirrors = BTRFS_SUPER_MIRROR_MAX;
 
-        /* make sure only the last submit_bh does a barrier */
-        if (do_barriers) {
-                for (i = 0; i < max_mirrors; i++) {
-                        bytenr = btrfs_sb_offset(i);
-                        if (bytenr + BTRFS_SUPER_INFO_SIZE >=
-                            device->total_bytes)
-                                break;
-                        last_barrier = i;
-                }
-        }
-
         for (i = 0; i < max_mirrors; i++) {
                 bytenr = btrfs_sb_offset(i);
                 if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes)
@@ -2315,17 +2622,136 @@ static int write_dev_supers(struct btrfs_device *device,
                         bh->b_end_io = btrfs_end_buffer_write_sync;
                 }
 
-                if (i == last_barrier && do_barriers)
-                        ret = submit_bh(WRITE_FLUSH_FUA, bh);
-                else
-                        ret = submit_bh(WRITE_SYNC, bh);
-
+                /*
+                 * we fua the first super.  The others we allow
+                 * to go down lazy.
+                 */
+                ret = submit_bh(WRITE_FUA, bh);
                 if (ret)
                         errors++;
         }
         return errors < i ? 0 : -1;
 }
 
+/*
+ * endio for the write_dev_flush, this will wake anyone waiting
+ * for the barrier when it is done
+ */
+static void btrfs_end_empty_barrier(struct bio *bio, int err)
+{
+        if (err) {
+                if (err == -EOPNOTSUPP)
+                        set_bit(BIO_EOPNOTSUPP, &bio->bi_flags);
+                clear_bit(BIO_UPTODATE, &bio->bi_flags);
+        }
+        if (bio->bi_private)
+                complete(bio->bi_private);
+        bio_put(bio);
+}
+
+/*
+ * trigger flushes for one the devices.  If you pass wait == 0, the flushes are
+ * sent down.  With wait == 1, it waits for the previous flush.
+ *
+ * any device where the flush fails with eopnotsupp are flagged as not-barrier
+ * capable
+ */
+static int write_dev_flush(struct btrfs_device *device, int wait)
+{
+        struct bio *bio;
+        int ret = 0;
+
+        if (device->nobarriers)
+                return 0;
+
+        if (wait) {
+                bio = device->flush_bio;
+                if (!bio)
+                        return 0;
+
+                wait_for_completion(&device->flush_wait);
+
+                if (bio_flagged(bio, BIO_EOPNOTSUPP)) {
+                        printk("btrfs: disabling barriers on dev %s\n",
+                               device->name);
+                        device->nobarriers = 1;
+                }
+                if (!bio_flagged(bio, BIO_UPTODATE)) {
+                        ret = -EIO;
+                }
+
+                /* drop the reference from the wait == 0 run */
+                bio_put(bio);
+                device->flush_bio = NULL;
+
+                return ret;
+        }
+
+        /*
+         * one reference for us, and we leave it for the
+         * caller
+         */
+        device->flush_bio = NULL;;
+        bio = bio_alloc(GFP_NOFS, 0);
+        if (!bio)
+                return -ENOMEM;
+
+        bio->bi_end_io = btrfs_end_empty_barrier;
+        bio->bi_bdev = device->bdev;
+        init_completion(&device->flush_wait);
+        bio->bi_private = &device->flush_wait;
+        device->flush_bio = bio;
+
+        bio_get(bio);
+        submit_bio(WRITE_FLUSH, bio);
+
+        return 0;
+}
+
+/*
+ * send an empty flush down to each device in parallel,
+ * then wait for them
+ */
+static int barrier_all_devices(struct btrfs_fs_info *info)
+{
+        struct list_head *head;
+        struct btrfs_device *dev;
+        int errors = 0;
+        int ret;
+
+        /* send down all the barriers */
+        head = &info->fs_devices->devices;
+        list_for_each_entry_rcu(dev, head, dev_list) {
+                if (!dev->bdev) {
+                        errors++;
+                        continue;
+                }
+                if (!dev->in_fs_metadata || !dev->writeable)
+                        continue;
+
+                ret = write_dev_flush(dev, 0);
+                if (ret)
+                        errors++;
+        }
+
+        /* wait for all the barriers */
+        list_for_each_entry_rcu(dev, head, dev_list) {
+                if (!dev->bdev) {
+                        errors++;
+                        continue;
+                }
+                if (!dev->in_fs_metadata || !dev->writeable)
+                        continue;
+
+                ret = write_dev_flush(dev, 1);
+                if (ret)
+                        errors++;
+        }
+        if (errors)
+                return -EIO;
+        return 0;
+}
+
 int write_all_supers(struct btrfs_root *root, int max_mirrors)
 {
         struct list_head *head;
@@ -2338,14 +2764,19 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
         int total_errors = 0;
         u64 flags;
 
-        max_errors = btrfs_super_num_devices(&root->fs_info->super_copy) - 1;
+        max_errors = btrfs_super_num_devices(root->fs_info->super_copy) - 1;
         do_barriers = !btrfs_test_opt(root, NOBARRIER);
+        backup_super_roots(root->fs_info);
 
-        sb = &root->fs_info->super_for_commit;
+        sb = root->fs_info->super_for_commit;
         dev_item = &sb->dev_item;
 
         mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
         head = &root->fs_info->fs_devices->devices;
+
+        if (do_barriers)
+                barrier_all_devices(root->fs_info);
+
         list_for_each_entry_rcu(dev, head, dev_list) {
                 if (!dev->bdev) {
                         total_errors++;
@@ -2545,8 +2976,6 @@ int close_ctree(struct btrfs_root *root)
         /* clear out the rbtree of defraggable inodes */
         btrfs_run_defrag_inodes(root->fs_info);
 
-        btrfs_put_block_group_cache(fs_info);
-
         /*
          * Here come 2 situations when btrfs is broken to flip readonly:
          *
@@ -2572,6 +3001,8 @@ int close_ctree(struct btrfs_root *root)
                         printk(KERN_ERR "btrfs: commit super ret %d\n", ret);
         }
 
+        btrfs_put_block_group_cache(fs_info);
+
         kthread_stop(root->fs_info->transaction_kthread);
         kthread_stop(root->fs_info->cleaner_kthread);
 
@@ -2603,7 +3034,6 @@ int close_ctree(struct btrfs_root *root)
         del_fs_roots(fs_info);
 
         iput(fs_info->btree_inode);
-        kfree(fs_info->delayed_root);
 
         btrfs_stop_workers(&fs_info->generic_worker);
         btrfs_stop_workers(&fs_info->fixup_workers);
@@ -2617,6 +3047,7 @@ int close_ctree(struct btrfs_root *root)
         btrfs_stop_workers(&fs_info->submit_workers);
         btrfs_stop_workers(&fs_info->delayed_workers);
         btrfs_stop_workers(&fs_info->caching_workers);
+        btrfs_stop_workers(&fs_info->readahead_workers);
 
         btrfs_close_devices(fs_info->fs_devices);
         btrfs_mapping_tree_free(&fs_info->mapping_tree);
@@ -2624,12 +3055,7 @@ int close_ctree(struct btrfs_root *root)
         bdi_destroy(&fs_info->bdi);
         cleanup_srcu_struct(&fs_info->subvol_srcu);
 
-        kfree(fs_info->extent_root);
-        kfree(fs_info->tree_root);
-        kfree(fs_info->chunk_root);
-        kfree(fs_info->dev_root);
-        kfree(fs_info->csum_root);
-        kfree(fs_info);
+        free_fs_info(fs_info);
 
         return 0;
 }
@@ -2735,7 +3161,8 @@ int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid)
         return ret;
 }
 
-int btree_lock_page_hook(struct page *page)
+static int btree_lock_page_hook(struct page *page, void *data,
+                                void (*flush_fn)(void *))
 {
         struct inode *inode = page->mapping->host;
         struct btrfs_root *root = BTRFS_I(inode)->root;
@@ -2752,7 +3179,10 @@ int btree_lock_page_hook(struct page *page)
         if (!eb)
                 goto out;
 
-        btrfs_tree_lock(eb);
+        if (!btrfs_try_tree_write_lock(eb)) {
+                flush_fn(data);
+                btrfs_tree_lock(eb);
+        }
         btrfs_set_header_flag(eb, BTRFS_HEADER_FLAG_WRITTEN);
 
         if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &eb->bflags)) {
@@ -2767,7 +3197,10 @@ int btree_lock_page_hook(struct page *page)
         btrfs_tree_unlock(eb);
         free_extent_buffer(eb);
 out:
-        lock_page(page);
+        if (!trylock_page(page)) {
+                flush_fn(data);
+                lock_page(page);
+        }
         return 0;
 }
 
@@ -3123,6 +3556,7 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root)
 static struct extent_io_ops btree_extent_io_ops = {
         .write_cache_pages_lock_hook = btree_lock_page_hook,
         .readpage_end_io_hook = btree_readpage_end_io_hook,
+        .readpage_io_failed_hook = btree_io_failed_hook,
         .submit_bio_hook = btree_submit_bio_hook,
         /* note we're sharing with inode.c for the merge bio hook */
         .merge_bio_hook = btrfs_merge_bio_hook,