Merge branch 'cleanups' of git://repo.or.cz/linux-2.6/btrfs-unstable into inode_numbers

[karo-tx-linux.git] / fs / btrfs / disk-io.c

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index cb9d1b8bfe745291edb1bc5c7a937011a570bdc9..a2eb3a3755db5fef162093b4e2d58652f2b25bc0 100644 (file)
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -42,6 +42,7 @@
 #include "locking.h"
 #include "tree-log.h"
 #include "free-space-cache.h"
+#include "inode-map.h"
 
 static struct extent_io_ops btree_extent_io_ops;
 static void end_workqueue_fn(struct btrfs_work *work);
@@ -1045,6 +1046,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
        root->name = NULL;
        root->in_sysfs = 0;
        root->inode_tree = RB_ROOT;
+       INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC);
        root->block_rsv = NULL;
        root->orphan_block_rsv = NULL;
 
@@ -1298,6 +1300,19 @@ again:
        if (IS_ERR(root))
                return root;
 
+       root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS);
+       if (!root->free_ino_ctl)
+               goto fail;
+       root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned),
+                                       GFP_NOFS);
+       if (!root->free_ino_pinned)
+               goto fail;
+
+       btrfs_init_free_ino_ctl(root);
+       mutex_init(&root->fs_commit_mutex);
+       spin_lock_init(&root->cache_lock);
+       init_waitqueue_head(&root->cache_wait);
+
        set_anon_super(&root->anon_super, NULL);
 
        if (btrfs_root_refs(&root->root_item) == 0) {
@@ -1631,6 +1646,13 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 
        INIT_LIST_HEAD(&fs_info->ordered_extents);
        spin_lock_init(&fs_info->ordered_extent_lock);
+       fs_info->delayed_root = kmalloc(sizeof(struct btrfs_delayed_root),
+                                       GFP_NOFS);
+       if (!fs_info->delayed_root) {
+               err = -ENOMEM;
+               goto fail_iput;
+       }
+       btrfs_init_delayed_root(fs_info->delayed_root);
 
        sb->s_blocksize = 4096;
        sb->s_blocksize_bits = blksize_bits(4096);
@@ -1696,7 +1718,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
        bh = btrfs_read_dev_super(fs_devices->latest_bdev);
        if (!bh) {
                err = -EINVAL;
-               goto fail_iput;
+               goto fail_alloc;
        }
 
        memcpy(&fs_info->super_copy, bh->b_data, sizeof(fs_info->super_copy));
@@ -1708,7 +1730,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 
        disk_super = &fs_info->super_copy;
        if (!btrfs_super_root(disk_super))
-               goto fail_iput;
+               goto fail_alloc;
 
        /* check FS state, whether FS is broken. */
        fs_info->fs_state |= btrfs_super_flags(disk_super);
@@ -1724,7 +1746,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
        ret = btrfs_parse_options(tree_root, options);
        if (ret) {
                err = ret;
-               goto fail_iput;
+               goto fail_alloc;
        }
 
        features = btrfs_super_incompat_flags(disk_super) &
@@ -1734,7 +1756,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
                       "unsupported optional features (%Lx).\n",
                       (unsigned long long)features);
                err = -EINVAL;
-               goto fail_iput;
+               goto fail_alloc;
        }
 
        features = btrfs_super_incompat_flags(disk_super);
@@ -1750,7 +1772,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
                       "unsupported option features (%Lx).\n",
                       (unsigned long long)features);
                err = -EINVAL;
-               goto fail_iput;
+               goto fail_alloc;
        }
 
        btrfs_init_workers(&fs_info->generic_worker,
@@ -1797,6 +1819,9 @@ struct btrfs_root *open_ctree(struct super_block *sb,
                           &fs_info->generic_worker);
        btrfs_init_workers(&fs_info->endio_freespace_worker, "freespace-write",
                           1, &fs_info->generic_worker);
+       btrfs_init_workers(&fs_info->delayed_workers, "delayed-meta",
+                          fs_info->thread_pool_size,
+                          &fs_info->generic_worker);
 
        /*
         * endios are largely parallel and should have a very
@@ -1818,6 +1843,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
        btrfs_start_workers(&fs_info->endio_meta_write_workers, 1);
        btrfs_start_workers(&fs_info->endio_write_workers, 1);
        btrfs_start_workers(&fs_info->endio_freespace_worker, 1);
+       btrfs_start_workers(&fs_info->delayed_workers, 1);
 
        fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super);
        fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages,
@@ -2074,6 +2100,9 @@ fail_sb_buffer:
        btrfs_stop_workers(&fs_info->endio_write_workers);
        btrfs_stop_workers(&fs_info->endio_freespace_worker);
        btrfs_stop_workers(&fs_info->submit_workers);
+       btrfs_stop_workers(&fs_info->delayed_workers);
+fail_alloc:
+       kfree(fs_info->delayed_root);
 fail_iput:
        invalidate_inode_pages2(fs_info->btree_inode->i_mapping);
        iput(fs_info->btree_inode);
@@ -2338,12 +2367,15 @@ int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root)
        if (btrfs_root_refs(&root->root_item) == 0)
                synchronize_srcu(&fs_info->subvol_srcu);
 
+       __btrfs_remove_free_space_cache(root->free_ino_pinned);
+       __btrfs_remove_free_space_cache(root->free_ino_ctl);
        free_fs_root(root);
        return 0;
 }
 
 static void free_fs_root(struct btrfs_root *root)
 {
+       iput(root->cache_inode);
        WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree));
        if (root->anon_super.s_dev) {
                down_write(&root->anon_super.s_umount);
@@ -2351,6 +2383,8 @@ static void free_fs_root(struct btrfs_root *root)
        }
        free_extent_buffer(root->node);
        free_extent_buffer(root->commit_root);
+       kfree(root->free_ino_ctl);
+       kfree(root->free_ino_pinned);
        kfree(root->name);
        kfree(root);
 }
@@ -2512,6 +2546,7 @@ 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);
@@ -2523,6 +2558,7 @@ int close_ctree(struct btrfs_root *root)
        btrfs_stop_workers(&fs_info->endio_write_workers);
        btrfs_stop_workers(&fs_info->endio_freespace_worker);
        btrfs_stop_workers(&fs_info->submit_workers);
+       btrfs_stop_workers(&fs_info->delayed_workers);
 
        btrfs_close_devices(fs_info->fs_devices);
        btrfs_mapping_tree_free(&fs_info->mapping_tree);
@@ -2596,6 +2632,29 @@ void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr)
        u64 num_dirty;
        unsigned long thresh = 32 * 1024 * 1024;
 
+       if (current->flags & PF_MEMALLOC)
+               return;
+
+       btrfs_balance_delayed_items(root);
+
+       num_dirty = root->fs_info->dirty_metadata_bytes;
+
+       if (num_dirty > thresh) {
+               balance_dirty_pages_ratelimited_nr(
+                                  root->fs_info->btree_inode->i_mapping, 1);
+       }
+       return;
+}
+
+void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr)
+{
+       /*
+        * looks as though older kernels can get into trouble with
+        * this code, they end up stuck in balance_dirty_pages forever
+        */
+       u64 num_dirty;
+       unsigned long thresh = 32 * 1024 * 1024;
+
        if (current->flags & PF_MEMALLOC)
                return;