1 files changed, 130 insertions, 80 deletions

diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index 7782020996fe..e86ae04abe6a 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -35,6 +35,7 @@
 #include "volumes.h"
 #include "async-thread.h"
 #include "check-integrity.h"
+#include "rcu-string.h"
 
 static int init_first_rw_device(struct btrfs_trans_handle *trans,
 				struct btrfs_root *root,
@@ -64,7 +65,7 @@ static void free_fs_devices(struct btrfs_fs_devices *fs_devices)
 		device = list_entry(fs_devices->devices.next,
 				    struct btrfs_device, dev_list);
 		list_del(&device->dev_list);
-		kfree(device->name);
+		rcu_string_free(device->name);
 		kfree(device);
 	}
 	kfree(fs_devices);
@@ -334,8 +335,8 @@ static noinline int device_list_add(const char *path,
 {
 	struct btrfs_device *device;
 	struct btrfs_fs_devices *fs_devices;
+	struct rcu_string *name;
 	u64 found_transid = btrfs_super_generation(disk_super);
-	char *name;
 
 	fs_devices = find_fsid(disk_super->fsid);
 	if (!fs_devices) {
@@ -369,11 +370,13 @@ static noinline int device_list_add(const char *path,
 		memcpy(device->uuid, disk_super->dev_item.uuid,
 		       BTRFS_UUID_SIZE);
 		spin_lock_init(&device->io_lock);
-		device->name = kstrdup(path, GFP_NOFS);
-		if (!device->name) {
+
+		name = rcu_string_strdup(path, GFP_NOFS);
+		if (!name) {
 			kfree(device);
 			return -ENOMEM;
 		}
+		rcu_assign_pointer(device->name, name);
 		INIT_LIST_HEAD(&device->dev_alloc_list);
 
 		/* init readahead state */
@@ -390,12 +393,12 @@ static noinline int device_list_add(const char *path,
 
 		device->fs_devices = fs_devices;
 		fs_devices->num_devices++;
-	} else if (!device->name || strcmp(device->name, path)) {
-		name = kstrdup(path, GFP_NOFS);
+	} else if (!device->name || strcmp(device->name->str, path)) {
+		name = rcu_string_strdup(path, GFP_NOFS);
 		if (!name)
 			return -ENOMEM;
-		kfree(device->name);
-		device->name = name;
+		rcu_string_free(device->name);
+		rcu_assign_pointer(device->name, name);
 		if (device->missing) {
 			fs_devices->missing_devices--;
 			device->missing = 0;
@@ -426,19 +429,27 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig)
 	mutex_init(&fs_devices->device_list_mutex);
 	fs_devices->latest_devid = orig->latest_devid;
 	fs_devices->latest_trans = orig->latest_trans;
+	fs_devices->total_devices = orig->total_devices;
 	memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid));
 
 	/* We have held the volume lock, it is safe to get the devices. */
 	list_for_each_entry(orig_dev, &orig->devices, dev_list) {
+		struct rcu_string *name;
+
 		device = kzalloc(sizeof(*device), GFP_NOFS);
 		if (!device)
 			goto error;
 
-		device->name = kstrdup(orig_dev->name, GFP_NOFS);
-		if (!device->name) {
+		/*
+		 * This is ok to do without rcu read locked because we hold the
+		 * uuid mutex so nothing we touch in here is going to disappear.
+		 */
+		name = rcu_string_strdup(orig_dev->name->str, GFP_NOFS);
+		if (!name) {
 			kfree(device);
 			goto error;
 		}
+		rcu_assign_pointer(device->name, name);
 
 		device->devid = orig_dev->devid;
 		device->work.func = pending_bios_fn;
@@ -491,7 +502,7 @@ again:
 		}
 		list_del_init(&device->dev_list);
 		fs_devices->num_devices--;
-		kfree(device->name);
+		rcu_string_free(device->name);
 		kfree(device);
 	}
 
@@ -516,7 +527,7 @@ static void __free_device(struct work_struct *work)
 	if (device->bdev)
 		blkdev_put(device->bdev, device->mode);
 
-	kfree(device->name);
+	rcu_string_free(device->name);
 	kfree(device);
 }
 
@@ -540,6 +551,7 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 	mutex_lock(&fs_devices->device_list_mutex);
 	list_for_each_entry(device, &fs_devices->devices, dev_list) {
 		struct btrfs_device *new_device;
+		struct rcu_string *name;
 
 		if (device->bdev)
 			fs_devices->open_devices--;
@@ -555,8 +567,11 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices)
 		new_device = kmalloc(sizeof(*new_device), GFP_NOFS);
 		BUG_ON(!new_device); /* -ENOMEM */
 		memcpy(new_device, device, sizeof(*new_device));
-		new_device->name = kstrdup(device->name, GFP_NOFS);
-		BUG_ON(device->name && !new_device->name); /* -ENOMEM */
+
+		/* Safe because we are under uuid_mutex */
+		name = rcu_string_strdup(device->name->str, GFP_NOFS);
+		BUG_ON(device->name && !name); /* -ENOMEM */
+		rcu_assign_pointer(new_device->name, name);
 		new_device->bdev = NULL;
 		new_device->writeable = 0;
 		new_device->in_fs_metadata = 0;
@@ -621,9 +636,9 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices,
 		if (!device->name)
 			continue;
 
-		bdev = blkdev_get_by_path(device->name, flags, holder);
+		bdev = blkdev_get_by_path(device->name->str, flags, holder);
 		if (IS_ERR(bdev)) {
-			printk(KERN_INFO "open %s failed\n", device->name);
+			printk(KERN_INFO "open %s failed\n", device->name->str);
 			goto error;
 		}
 		filemap_write_and_wait(bdev->bd_inode->i_mapping);
@@ -725,6 +740,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	int ret;
 	u64 devid;
 	u64 transid;
+	u64 total_devices;
 
 	flags |= FMODE_EXCL;
 	bdev = blkdev_get_by_path(path, flags, holder);
@@ -746,6 +762,7 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	disk_super = (struct btrfs_super_block *)bh->b_data;
 	devid = btrfs_stack_device_id(&disk_super->dev_item);
 	transid = btrfs_super_generation(disk_super);
+	total_devices = btrfs_super_num_devices(disk_super);
 	if (disk_super->label[0])
 		printk(KERN_INFO "device label %s ", disk_super->label);
 	else
@@ -753,7 +770,8 @@ int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder,
 	printk(KERN_CONT "devid %llu transid %llu %s\n",
 	       (unsigned long long)devid, (unsigned long long)transid, path);
 	ret = device_list_add(path, disk_super, devid, fs_devices_ret);
-
+	if (!ret && fs_devices_ret)
+		(*fs_devices_ret)->total_devices = total_devices;
 	brelse(bh);
 error_close:
 	mutex_unlock(&uuid_mutex);
@@ -1419,6 +1437,7 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	list_del_rcu(&device->dev_list);
 
 	device->fs_devices->num_devices--;
+	device->fs_devices->total_devices--;
 
 	if (device->missing)
 		root->fs_info->fs_devices->missing_devices--;
@@ -1536,6 +1555,7 @@ static int btrfs_prepare_sprout(struct btrfs_root *root)
 	fs_devices->seeding = 0;
 	fs_devices->num_devices = 0;
 	fs_devices->open_devices = 0;
+	fs_devices->total_devices = 0;
 	fs_devices->seed = seed_devices;
 
 	generate_random_uuid(fs_devices->fsid);
@@ -1632,6 +1652,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	struct block_device *bdev;
 	struct list_head *devices;
 	struct super_block *sb = root->fs_info->sb;
+	struct rcu_string *name;
 	u64 total_bytes;
 	int seeding_dev = 0;
 	int ret = 0;
@@ -1671,23 +1692,24 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 		goto error;
 	}
 
-	device->name = kstrdup(device_path, GFP_NOFS);
-	if (!device->name) {
+	name = rcu_string_strdup(device_path, GFP_NOFS);
+	if (!name) {
 		kfree(device);
 		ret = -ENOMEM;
 		goto error;
 	}
+	rcu_assign_pointer(device->name, name);
 
 	ret = find_next_devid(root, &device->devid);
 	if (ret) {
-		kfree(device->name);
+		rcu_string_free(device->name);
 		kfree(device);
 		goto error;
 	}
 
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
-		kfree(device->name);
+		rcu_string_free(device->name);
 		kfree(device);
 		ret = PTR_ERR(trans);
 		goto error;
@@ -1722,10 +1744,6 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 
 	device->fs_devices = root->fs_info->fs_devices;
 
-	/*
-	 * we don't want write_supers to jump in here with our device
-	 * half setup
-	 */
 	mutex_lock(&root->fs_info->fs_devices->device_list_mutex);
 	list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices);
 	list_add(&device->dev_alloc_list,
@@ -1733,6 +1751,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	root->fs_info->fs_devices->num_devices++;
 	root->fs_info->fs_devices->open_devices++;
 	root->fs_info->fs_devices->rw_devices++;
+	root->fs_info->fs_devices->total_devices++;
 	if (device->can_discard)
 		root->fs_info->fs_devices->num_can_discard++;
 	root->fs_info->fs_devices->total_rw_bytes += device->total_bytes;
@@ -1796,7 +1815,7 @@ error_trans:
 	unlock_chunks(root);
 	btrfs_abort_transaction(trans, root, ret);
 	btrfs_end_transaction(trans, root);
-	kfree(device->name);
+	rcu_string_free(device->name);
 	kfree(device);
 error:
 	blkdev_put(bdev, FMODE_EXCL);
@@ -2829,31 +2848,48 @@ out:
 
 static int balance_kthread(void *data)
 {
-	struct btrfs_balance_control *bctl =
-			(struct btrfs_balance_control *)data;
-	struct btrfs_fs_info *fs_info = bctl->fs_info;
+	struct btrfs_fs_info *fs_info = data;
 	int ret = 0;
 
 	mutex_lock(&fs_info->volume_mutex);
 	mutex_lock(&fs_info->balance_mutex);
 
-	set_balance_control(bctl);
-
-	if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) {
-		printk(KERN_INFO "btrfs: force skipping balance\n");
-	} else {
+	if (fs_info->balance_ctl) {
 		printk(KERN_INFO "btrfs: continuing balance\n");
-		ret = btrfs_balance(bctl, NULL);
+		ret = btrfs_balance(fs_info->balance_ctl, NULL);
 	}
 
 	mutex_unlock(&fs_info->balance_mutex);
 	mutex_unlock(&fs_info->volume_mutex);
+
 	return ret;
 }
 
-int btrfs_recover_balance(struct btrfs_root *tree_root)
+int btrfs_resume_balance_async(struct btrfs_fs_info *fs_info)
 {
 	struct task_struct *tsk;
+
+	spin_lock(&fs_info->balance_lock);
+	if (!fs_info->balance_ctl) {
+		spin_unlock(&fs_info->balance_lock);
+		return 0;
+	}
+	spin_unlock(&fs_info->balance_lock);
+
+	if (btrfs_test_opt(fs_info->tree_root, SKIP_BALANCE)) {
+		printk(KERN_INFO "btrfs: force skipping balance\n");
+		return 0;
+	}
+
+	tsk = kthread_run(balance_kthread, fs_info, "btrfs-balance");
+	if (IS_ERR(tsk))
+		return PTR_ERR(tsk);
+
+	return 0;
+}
+
+int btrfs_recover_balance(struct btrfs_fs_info *fs_info)
+{
 	struct btrfs_balance_control *bctl;
 	struct btrfs_balance_item *item;
 	struct btrfs_disk_balance_args disk_bargs;
@@ -2866,29 +2902,30 @@ int btrfs_recover_balance(struct btrfs_root *tree_root)
 	if (!path)
 		return -ENOMEM;
 
-	bctl = kzalloc(sizeof(*bctl), GFP_NOFS);
-	if (!bctl) {
-		ret = -ENOMEM;
-		goto out;
-	}
-
 	key.objectid = BTRFS_BALANCE_OBJECTID;
 	key.type = BTRFS_BALANCE_ITEM_KEY;
 	key.offset = 0;
 
-	ret = btrfs_search_slot(NULL, tree_root, &key, path, 0, 0);
+	ret = btrfs_search_slot(NULL, fs_info->tree_root, &key, path, 0, 0);
 	if (ret < 0)
-		goto out_bctl;
+		goto out;
 	if (ret > 0) { /* ret = -ENOENT; */
 		ret = 0;
-		goto out_bctl;
+		goto out;
+	}
+
+	bctl = kzalloc(sizeof(*bctl), GFP_NOFS);
+	if (!bctl) {
+		ret = -ENOMEM;
+		goto out;
 	}
 
 	leaf = path->nodes[0];
 	item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_balance_item);
 
-	bctl->fs_info = tree_root->fs_info;
-	bctl->flags = btrfs_balance_flags(leaf, item) | BTRFS_BALANCE_RESUME;
+	bctl->fs_info = fs_info;
+	bctl->flags = btrfs_balance_flags(leaf, item);
+	bctl->flags |= BTRFS_BALANCE_RESUME;
 
 	btrfs_balance_data(leaf, item, &disk_bargs);
 	btrfs_disk_balance_args_to_cpu(&bctl->data, &disk_bargs);
@@ -2897,14 +2934,13 @@ int btrfs_recover_balance(struct btrfs_root *tree_root)
 	btrfs_balance_sys(leaf, item, &disk_bargs);
 	btrfs_disk_balance_args_to_cpu(&bctl->sys, &disk_bargs);
 
-	tsk = kthread_run(balance_kthread, bctl, "btrfs-balance");
-	if (IS_ERR(tsk))
-		ret = PTR_ERR(tsk);
-	else
-		goto out;
+	mutex_lock(&fs_info->volume_mutex);
+	mutex_lock(&fs_info->balance_mutex);
 
-out_bctl:
-	kfree(bctl);
+	set_balance_control(bctl);
+
+	mutex_unlock(&fs_info->balance_mutex);
+	mutex_unlock(&fs_info->volume_mutex);
 out:
 	btrfs_free_path(path);
 	return ret;
@@ -4045,16 +4081,18 @@ static void btrfs_end_bio(struct bio *bio, int err)
 
 			BUG_ON(stripe_index >= bbio->num_stripes);
 			dev = bbio->stripes[stripe_index].dev;
-			if (bio->bi_rw & WRITE)
-				btrfs_dev_stat_inc(dev,
-						   BTRFS_DEV_STAT_WRITE_ERRS);
-			else
-				btrfs_dev_stat_inc(dev,
-						   BTRFS_DEV_STAT_READ_ERRS);
-			if ((bio->bi_rw & WRITE_FLUSH) == WRITE_FLUSH)
-				btrfs_dev_stat_inc(dev,
-						   BTRFS_DEV_STAT_FLUSH_ERRS);
-			btrfs_dev_stat_print_on_error(dev);
+			if (dev->bdev) {
+				if (bio->bi_rw & WRITE)
+					btrfs_dev_stat_inc(dev,
+						BTRFS_DEV_STAT_WRITE_ERRS);
+				else
+					btrfs_dev_stat_inc(dev,
+						BTRFS_DEV_STAT_READ_ERRS);
+				if ((bio->bi_rw & WRITE_FLUSH) == WRITE_FLUSH)
+					btrfs_dev_stat_inc(dev,
+						BTRFS_DEV_STAT_FLUSH_ERRS);
+				btrfs_dev_stat_print_on_error(dev);
+			}
 		}
 	}
 
@@ -4204,10 +4242,17 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio,
 		bio->bi_sector = bbio->stripes[dev_nr].physical >> 9;
 		dev = bbio->stripes[dev_nr].dev;
 		if (dev && dev->bdev && (rw != WRITE || dev->writeable)) {
+#ifdef DEBUG
+			struct rcu_string *name;
+
+			rcu_read_lock();
+			name = rcu_dereference(dev->name);
 			pr_debug("btrfs_map_bio: rw %d, secor=%llu, dev=%lu "
 				 "(%s id %llu), size=%u\n", rw,
 				 (u64)bio->bi_sector, (u_long)dev->bdev->bd_dev,
-				 dev->name, dev->devid, bio->bi_size);
+				 name->str, dev->devid, bio->bi_size);
+			rcu_read_unlock();
+#endif
 			bio->bi_bdev = dev->bdev;
 			if (async_submit)
 				schedule_bio(root, dev, rw, bio);
@@ -4694,8 +4739,6 @@ int btrfs_init_dev_stats(struct btrfs_fs_info *fs_info)
 		key.offset = device->devid;
 		ret = btrfs_search_slot(NULL, dev_root, &key, path, 0, 0);
 		if (ret) {
-			printk(KERN_WARNING "btrfs: no dev_stats entry found for device %s (devid %llu) (OK on first mount after mkfs)\n",
-			       device->name, (unsigned long long)device->devid);
 			__btrfs_reset_dev_stats(device);
 			device->dev_stats_valid = 1;
 			btrfs_release_path(path);
@@ -4747,8 +4790,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 	BUG_ON(!path);
 	ret = btrfs_search_slot(trans, dev_root, &key, path, -1, 1);
 	if (ret < 0) {
-		printk(KERN_WARNING "btrfs: error %d while searching for dev_stats item for device %s!\n",
-		       ret, device->name);
+		printk_in_rcu(KERN_WARNING "btrfs: error %d while searching for dev_stats item for device %s!\n",
+			      ret, rcu_str_deref(device->name));
 		goto out;
 	}
 
@@ -4757,8 +4800,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		/* need to delete old one and insert a new one */
 		ret = btrfs_del_item(trans, dev_root, path);
 		if (ret != 0) {
-			printk(KERN_WARNING "btrfs: delete too small dev_stats item for device %s failed %d!\n",
-			       device->name, ret);
+			printk_in_rcu(KERN_WARNING "btrfs: delete too small dev_stats item for device %s failed %d!\n",
+				      rcu_str_deref(device->name), ret);
 			goto out;
 		}
 		ret = 1;
@@ -4770,8 +4813,8 @@ static int update_dev_stat_item(struct btrfs_trans_handle *trans,
 		ret = btrfs_insert_empty_item(trans, dev_root, path,
 					      &key, sizeof(*ptr));
 		if (ret < 0) {
-			printk(KERN_WARNING "btrfs: insert dev_stats item for device %s failed %d!\n",
-			       device->name, ret);
+			printk_in_rcu(KERN_WARNING "btrfs: insert dev_stats item for device %s failed %d!\n",
+				      rcu_str_deref(device->name), ret);
 			goto out;
 		}
 	}
@@ -4823,9 +4866,9 @@ void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
 {
 	if (!dev->dev_stats_valid)
 		return;
-	printk_ratelimited(KERN_ERR
+	printk_ratelimited_in_rcu(KERN_ERR
 			   "btrfs: bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
-			   dev->name,
+			   rcu_str_deref(dev->name),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
 			   btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_FLUSH_ERRS),
@@ -4837,8 +4880,16 @@ void btrfs_dev_stat_print_on_error(struct btrfs_device *dev)
 
 static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
 {
-	printk(KERN_INFO "btrfs: bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
-	       dev->name,
+	int i;
+
+	for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++)
+		if (btrfs_dev_stat_read(dev, i) != 0)
+			break;
+	if (i == BTRFS_DEV_STAT_VALUES_MAX)
+		return; /* all values == 0, suppress message */
+
+	printk_in_rcu(KERN_INFO "btrfs: bdev %s errs: wr %u, rd %u, flush %u, corrupt %u, gen %u\n",
+	       rcu_str_deref(dev->name),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_WRITE_ERRS),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_READ_ERRS),
 	       btrfs_dev_stat_read(dev, BTRFS_DEV_STAT_FLUSH_ERRS),
@@ -4847,8 +4898,7 @@ static void btrfs_dev_stat_print_on_load(struct btrfs_device *dev)
 }
 
 int btrfs_get_dev_stats(struct btrfs_root *root,
-			struct btrfs_ioctl_get_dev_stats *stats,
-			int reset_after_read)
+			struct btrfs_ioctl_get_dev_stats *stats)
 {
 	struct btrfs_device *dev;
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
@@ -4866,7 +4916,7 @@ int btrfs_get_dev_stats(struct btrfs_root *root,
 		printk(KERN_WARNING
 		       "btrfs: get dev_stats failed, not yet valid\n");
 		return -ENODEV;
-	} else if (reset_after_read) {
+	} else if (stats->flags & BTRFS_DEV_STATS_RESET) {
 		for (i = 0; i < BTRFS_DEV_STAT_VALUES_MAX; i++) {
 			if (stats->nr_items > i)
 				stats->values[i] =