1 files changed, 114 insertions, 96 deletions

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 228cf36ece83..98b6a71decba 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -29,6 +29,7 @@
 #include <linux/crc32c.h>
 #include <linux/slab.h>
 #include <linux/migrate.h>
+#include <linux/ratelimit.h>
 #include <asm/unaligned.h>
 #include "compat.h"
 #include "ctree.h"
@@ -41,6 +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);
@@ -137,7 +139,7 @@ static const char *btrfs_eb_name[BTRFS_MAX_LEVEL + 1] = {
  * that covers the entire device
  */
 static struct extent_map *btree_get_extent(struct inode *inode,
-                struct page *page, size_t page_offset, u64 start, u64 len,
+                struct page *page, size_t pg_offset, u64 start, u64 len,
                 int create)
 {
         struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
@@ -154,7 +156,7 @@ static struct extent_map *btree_get_extent(struct inode *inode,
         }
         read_unlock(&em_tree->lock);
 
-        em = alloc_extent_map(GFP_NOFS);
+        em = alloc_extent_map();
         if (!em) {
                 em = ERR_PTR(-ENOMEM);
                 goto out;
@@ -254,14 +256,12 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf,
                         memcpy(&found, result, csum_size);
 
                         read_extent_buffer(buf, &val, 0, csum_size);
-                        if (printk_ratelimit()) {
-                                printk(KERN_INFO "btrfs: %s checksum verify "
+                        printk_ratelimited(KERN_INFO "btrfs: %s checksum verify "
                                        "failed on %llu wanted %X found %X "
                                        "level %d\n",
                                        root->fs_info->sb->s_id,
                                        (unsigned long long)buf->start, val, found,
                                        btrfs_header_level(buf));
-                        }
                         if (result != (char *)&inline_result)
                                 kfree(result);
                         return 1;
@@ -296,13 +296,11 @@ static int verify_parent_transid(struct extent_io_tree *io_tree,
                 ret = 0;
                 goto out;
         }
-        if (printk_ratelimit()) {
-                printk("parent transid verify failed on %llu wanted %llu "
+        printk_ratelimited("parent transid verify failed on %llu wanted %llu "
                        "found %llu\n",
                        (unsigned long long)eb->start,
                        (unsigned long long)parent_transid,
                        (unsigned long long)btrfs_header_generation(eb));
-        }
         ret = 1;
         clear_extent_buffer_uptodate(io_tree, eb, &cached_state);
 out:
@@ -380,7 +378,7 @@ static int csum_dirty_buffer(struct btrfs_root *root, struct page *page)
         len = page->private >> 2;
         WARN_ON(len == 0);
 
-        eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS);
+        eb = alloc_extent_buffer(tree, start, len, page);
         if (eb == NULL) {
                 WARN_ON(1);
                 goto out;
@@ -525,7 +523,7 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
         len = page->private >> 2;
         WARN_ON(len == 0);
 
-        eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS);
+        eb = alloc_extent_buffer(tree, start, len, page);
         if (eb == NULL) {
                 ret = -EIO;
                 goto out;
@@ -533,12 +531,10 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
 
         found_start = btrfs_header_bytenr(eb);
         if (found_start != start) {
-                if (printk_ratelimit()) {
-                        printk(KERN_INFO "btrfs bad tree block start "
+                printk_ratelimited(KERN_INFO "btrfs bad tree block start "
                                "%llu %llu\n",
                                (unsigned long long)found_start,
                                (unsigned long long)eb->start);
-                }
                 ret = -EIO;
                 goto err;
         }
@@ -550,10 +546,8 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end,
                 goto err;
         }
         if (check_tree_block_fsid(root, eb)) {
-                if (printk_ratelimit()) {
-                        printk(KERN_INFO "btrfs bad fsid on block %llu\n",
+                printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n",
                                (unsigned long long)eb->start);
-                }
                 ret = -EIO;
                 goto err;
         }
@@ -650,12 +644,6 @@ unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info)
         return 256 * limit;
 }
 
-int btrfs_congested_async(struct btrfs_fs_info *info, int iodone)
-{
-        return atomic_read(&info->nr_async_bios) >
-                btrfs_async_submit_limit(info);
-}
-
 static void run_one_async_start(struct btrfs_work *work)
 {
         struct async_submit_bio *async;
@@ -963,7 +951,7 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root,
         struct inode *btree_inode = root->fs_info->btree_inode;
         struct extent_buffer *eb;
         eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
-                                bytenr, blocksize, GFP_NOFS);
+                                bytenr, blocksize);
         return eb;
 }
 
@@ -974,7 +962,7 @@ struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root,
         struct extent_buffer *eb;
 
         eb = alloc_extent_buffer(&BTRFS_I(btree_inode)->io_tree,
-                                 bytenr, blocksize, NULL, GFP_NOFS);
+                                 bytenr, blocksize, NULL);
         return eb;
 }
 
@@ -1058,13 +1046,13 @@ 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;
 
         INIT_LIST_HEAD(&root->dirty_list);
         INIT_LIST_HEAD(&root->orphan_list);
         INIT_LIST_HEAD(&root->root_list);
-        spin_lock_init(&root->node_lock);
         spin_lock_init(&root->orphan_lock);
         spin_lock_init(&root->inode_lock);
         spin_lock_init(&root->accounting_lock);
@@ -1080,7 +1068,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize,
         root->log_transid = 0;
         root->last_log_commit = 0;
         extent_io_tree_init(&root->dirty_log_pages,
-                             fs_info->btree_inode->i_mapping, GFP_NOFS);
+                             fs_info->btree_inode->i_mapping);
 
         memset(&root->root_key, 0, sizeof(root->root_key));
         memset(&root->root_item, 0, sizeof(root->root_item));
@@ -1283,21 +1271,6 @@ out:
         return root;
 }
 
-struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info,
-                                        u64 root_objectid)
-{
-        struct btrfs_root *root;
-
-        if (root_objectid == BTRFS_ROOT_TREE_OBJECTID)
-                return fs_info->tree_root;
-        if (root_objectid == BTRFS_EXTENT_TREE_OBJECTID)
-                return fs_info->extent_root;
-
-        root = radix_tree_lookup(&fs_info->fs_roots_radix,
-                                 (unsigned long)root_objectid);
-        return root;
-}
-
 struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info,
                                               struct btrfs_key *location)
 {
@@ -1326,6 +1299,19 @@ 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) {
@@ -1369,41 +1355,6 @@ fail:
         return ERR_PTR(ret);
 }
 
-struct btrfs_root *btrfs_read_fs_root(struct btrfs_fs_info *fs_info,
-                                      struct btrfs_key *location,
-                                      const char *name, int namelen)
-{
-        return btrfs_read_fs_root_no_name(fs_info, location);
-#if 0
-        struct btrfs_root *root;
-        int ret;
-
-        root = btrfs_read_fs_root_no_name(fs_info, location);
-        if (!root)
-                return NULL;
-
-        if (root->in_sysfs)
-                return root;
-
-        ret = btrfs_set_root_name(root, name, namelen);
-        if (ret) {
-                free_extent_buffer(root->node);
-                kfree(root);
-                return ERR_PTR(ret);
-        }
-
-        ret = btrfs_sysfs_add_root(root);
-        if (ret) {
-                free_extent_buffer(root->node);
-                kfree(root->name);
-                kfree(root);
-                return ERR_PTR(ret);
-        }
-        root->in_sysfs = 1;
-        return root;
-#endif
-}
-
 static int btrfs_congested_fn(void *congested_data, int bdi_bits)
 {
         struct btrfs_fs_info *info = (struct btrfs_fs_info *)congested_data;
@@ -1411,7 +1362,8 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
         struct btrfs_device *device;
         struct backing_dev_info *bdi;
 
-        list_for_each_entry(device, &info->fs_devices->devices, dev_list) {
+        rcu_read_lock();
+        list_for_each_entry_rcu(device, &info->fs_devices->devices, dev_list) {
                 if (!device->bdev)
                         continue;
                 bdi = blk_get_backing_dev_info(device->bdev);
@@ -1420,6 +1372,7 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits)
                         break;
                 }
         }
+        rcu_read_unlock();
         return ret;
 }
 
@@ -1522,6 +1475,7 @@ static int cleaner_kthread(void *arg)
                         btrfs_run_delayed_iputs(root);
                         btrfs_clean_old_snapshots(root);
                         mutex_unlock(&root->fs_info->cleaner_mutex);
+                        btrfs_run_defrag_inodes(root->fs_info);
                 }
 
                 if (freezing(current)) {
@@ -1611,7 +1565,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         struct btrfs_root *csum_root = kzalloc(sizeof(struct btrfs_root),
                                                  GFP_NOFS);
         struct btrfs_root *tree_root = btrfs_sb(sb);
-        struct btrfs_fs_info *fs_info = tree_root->fs_info;
+        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),
@@ -1623,11 +1577,12 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 
         struct btrfs_super_block *disk_super;
 
-        if (!extent_root || !tree_root || !fs_info ||
+        if (!extent_root || !tree_root || !tree_root->fs_info ||
             !chunk_root || !dev_root || !csum_root) {
                 err = -ENOMEM;
                 goto fail;
         }
+        fs_info = tree_root->fs_info;
 
         ret = init_srcu_struct(&fs_info->subvol_srcu);
         if (ret) {
@@ -1662,6 +1617,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         spin_lock_init(&fs_info->ref_cache_lock);
         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);
 
         init_completion(&fs_info->kobj_unregister);
         fs_info->tree_root = tree_root;
@@ -1684,15 +1640,35 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         atomic_set(&fs_info->async_delalloc_pages, 0);
         atomic_set(&fs_info->async_submit_draining, 0);
         atomic_set(&fs_info->nr_async_bios, 0);
+        atomic_set(&fs_info->defrag_running, 0);
         fs_info->sb = sb;
         fs_info->max_inline = 8192 * 1024;
         fs_info->metadata_ratio = 0;
+        fs_info->defrag_inodes = RB_ROOT;
 
         fs_info->thread_pool_size = min_t(unsigned long,
                                           num_online_cpus() + 2, 8);
 
         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);
+
+        mutex_init(&fs_info->scrub_lock);
+        atomic_set(&fs_info->scrubs_running, 0);
+        atomic_set(&fs_info->scrub_pause_req, 0);
+        atomic_set(&fs_info->scrubs_paused, 0);
+        atomic_set(&fs_info->scrub_cancel_req, 0);
+        init_waitqueue_head(&fs_info->scrub_pause_wait);
+        init_rwsem(&fs_info->scrub_super_lock);
+        fs_info->scrub_workers_refcnt = 0;
+        btrfs_init_workers(&fs_info->scrub_workers, "scrub",
+                           fs_info->thread_pool_size, &fs_info->generic_worker);
 
         sb->s_blocksize = 4096;
         sb->s_blocksize_bits = blksize_bits(4096);
@@ -1711,10 +1687,8 @@ struct btrfs_root *open_ctree(struct super_block *sb,
 
         RB_CLEAR_NODE(&BTRFS_I(fs_info->btree_inode)->rb_node);
         extent_io_tree_init(&BTRFS_I(fs_info->btree_inode)->io_tree,
-                             fs_info->btree_inode->i_mapping,
-                             GFP_NOFS);
-        extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree,
-                             GFP_NOFS);
+                             fs_info->btree_inode->i_mapping);
+        extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree);
 
         BTRFS_I(fs_info->btree_inode)->io_tree.ops = &btree_extent_io_ops;
 
@@ -1728,9 +1702,9 @@ struct btrfs_root *open_ctree(struct super_block *sb,
         fs_info->block_group_cache_tree = RB_ROOT;
 
         extent_io_tree_init(&fs_info->freed_extents[0],
-                             fs_info->btree_inode->i_mapping, GFP_NOFS);
+                             fs_info->btree_inode->i_mapping);
         extent_io_tree_init(&fs_info->freed_extents[1],
-                             fs_info->btree_inode->i_mapping, GFP_NOFS);
+                             fs_info->btree_inode->i_mapping);
         fs_info->pinned_extents = &fs_info->freed_extents[0];
         fs_info->do_barriers = 1;
 
@@ -1760,7 +1734,7 @@ 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));
@@ -1772,7 +1746,7 @@ 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);
@@ -1788,7 +1762,7 @@ 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) &
@@ -1798,7 +1772,7 @@ 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);
@@ -1814,7 +1788,7 @@ 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,
@@ -1861,6 +1835,9 @@ 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
@@ -1882,6 +1859,7 @@ 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,
@@ -2138,6 +2116,9 @@ 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);
@@ -2165,11 +2146,9 @@ static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate)
         if (uptodate) {
                 set_buffer_uptodate(bh);
         } else {
-                if (printk_ratelimit()) {
-                        printk(KERN_WARNING "lost page write due to "
+                printk_ratelimited(KERN_WARNING "lost page write due to "
                                         "I/O error on %s\n",
                                        bdevname(bh->b_bdev, b));
-                }
                 /* note, we dont' set_buffer_write_io_error because we have
                  * our own ways of dealing with the IO errors
                  */
@@ -2333,7 +2312,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
 
         mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
         head = &root->fs_info->fs_devices->devices;
-        list_for_each_entry(dev, head, dev_list) {
+        list_for_each_entry_rcu(dev, head, dev_list) {
                 if (!dev->bdev) {
                         total_errors++;
                         continue;
@@ -2366,7 +2345,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors)
         }
 
         total_errors = 0;
-        list_for_each_entry(dev, head, dev_list) {
+        list_for_each_entry_rcu(dev, head, dev_list) {
                 if (!dev->bdev)
                         continue;
                 if (!dev->in_fs_metadata || !dev->writeable)
@@ -2404,12 +2383,15 @@ 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);
@@ -2417,6 +2399,8 @@ 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);
 }
@@ -2520,6 +2504,15 @@ int close_ctree(struct btrfs_root *root)
         fs_info->closing = 1;
         smp_mb();
 
+        btrfs_scrub_cancel(root);
+
+        /* wait for any defraggers to finish */
+        wait_event(fs_info->transaction_wait,
+                   (atomic_read(&fs_info->defrag_running) == 0));
+
+        /* clear out the rbtree of defraggable inodes */
+        btrfs_run_defrag_inodes(root->fs_info);
+
         btrfs_put_block_group_cache(fs_info);
 
         /*
@@ -2578,6 +2571,7 @@ 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);
@@ -2589,6 +2583,7 @@ 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);
@@ -2665,6 +2660,29 @@ void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr)
         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;
+
         num_dirty = root->fs_info->dirty_metadata_bytes;
 
         if (num_dirty > thresh) {
@@ -2697,7 +2715,7 @@ int btree_lock_page_hook(struct page *page)
                 goto out;
 
         len = page->private >> 2;
-        eb = find_extent_buffer(io_tree, bytenr, len, GFP_NOFS);
+        eb = find_extent_buffer(io_tree, bytenr, len);
         if (!eb)
                 goto out;