1 files changed, 130 insertions, 126 deletions

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index eb5c95569300..086dcbadce09 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -87,9 +87,8 @@ struct btrfs_end_io_wq {
 	bio_end_io_t *end_io;
 	void *private;
 	struct btrfs_fs_info *info;
-	int error;
+	blk_status_t status;
 	enum btrfs_wq_endio_type metadata;
-	struct list_head list;
 	struct btrfs_work work;
 };
 
@@ -118,9 +117,9 @@ void btrfs_end_io_wq_exit(void)
  * just before they are sent down the IO stack.
  */
 struct async_submit_bio {
-	struct inode *inode;
+	void *private_data;
+	struct btrfs_fs_info *fs_info;
 	struct bio *bio;
-	struct list_head list;
 	extent_submit_bio_hook_t *submit_bio_start;
 	extent_submit_bio_hook_t *submit_bio_done;
 	int mirror_num;
@@ -131,7 +130,7 @@ struct async_submit_bio {
 	 */
 	u64 bio_offset;
 	struct btrfs_work work;
-	int error;
+	blk_status_t status;
 };
 
 /*
@@ -762,7 +761,7 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio,
 err:
 	if (reads_done &&
 	    test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
-		btree_readahead_hook(fs_info, eb, ret);
+		btree_readahead_hook(eb, ret);
 
 	if (ret) {
 		/*
@@ -787,7 +786,7 @@ static int btree_io_failed_hook(struct page *page, int failed_mirror)
 	eb->read_mirror = failed_mirror;
 	atomic_dec(&eb->io_pages);
 	if (test_and_clear_bit(EXTENT_BUFFER_READAHEAD, &eb->bflags))
-		btree_readahead_hook(eb->fs_info, eb, -EIO);
+		btree_readahead_hook(eb, -EIO);
 	return -EIO;	/* we fixed nothing */
 }
 
@@ -799,7 +798,7 @@ static void end_workqueue_bio(struct bio *bio)
 	btrfs_work_func_t func;
 
 	fs_info = end_io_wq->info;
-	end_io_wq->error = bio->bi_error;
+	end_io_wq->status = bio->bi_status;
 
 	if (bio_op(bio) == REQ_OP_WRITE) {
 		if (end_io_wq->metadata == BTRFS_WQ_ENDIO_METADATA) {
@@ -836,19 +835,19 @@ static void end_workqueue_bio(struct bio *bio)
 	btrfs_queue_work(wq, &end_io_wq->work);
 }
 
-int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
+blk_status_t btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio,
 			enum btrfs_wq_endio_type metadata)
 {
 	struct btrfs_end_io_wq *end_io_wq;
 
 	end_io_wq = kmem_cache_alloc(btrfs_end_io_wq_cache, GFP_NOFS);
 	if (!end_io_wq)
-		return -ENOMEM;
+		return BLK_STS_RESOURCE;
 
 	end_io_wq->private = bio->bi_private;
 	end_io_wq->end_io = bio->bi_end_io;
 	end_io_wq->info = info;
-	end_io_wq->error = 0;
+	end_io_wq->status = 0;
 	end_io_wq->bio = bio;
 	end_io_wq->metadata = metadata;
 
@@ -868,14 +867,14 @@ unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info)
 static void run_one_async_start(struct btrfs_work *work)
 {
 	struct async_submit_bio *async;
-	int ret;
+	blk_status_t ret;
 
 	async = container_of(work, struct  async_submit_bio, work);
-	ret = async->submit_bio_start(async->inode, async->bio,
+	ret = async->submit_bio_start(async->private_data, async->bio,
 				      async->mirror_num, async->bio_flags,
 				      async->bio_offset);
 	if (ret)
-		async->error = ret;
+		async->status = ret;
 }
 
 static void run_one_async_done(struct btrfs_work *work)
@@ -885,7 +884,7 @@ static void run_one_async_done(struct btrfs_work *work)
 	int limit;
 
 	async = container_of(work, struct  async_submit_bio, work);
-	fs_info = BTRFS_I(async->inode)->root->fs_info;
+	fs_info = async->fs_info;
 
 	limit = btrfs_async_submit_limit(fs_info);
 	limit = limit * 2 / 3;
@@ -898,13 +897,13 @@ static void run_one_async_done(struct btrfs_work *work)
 		wake_up(&fs_info->async_submit_wait);
 
 	/* If an error occurred we just want to clean up the bio and move on */
-	if (async->error) {
-		async->bio->bi_error = async->error;
+	if (async->status) {
+		async->bio->bi_status = async->status;
 		bio_endio(async->bio);
 		return;
 	}
 
-	async->submit_bio_done(async->inode, async->bio, async->mirror_num,
+	async->submit_bio_done(async->private_data, async->bio, async->mirror_num,
 			       async->bio_flags, async->bio_offset);
 }
 
@@ -916,20 +915,20 @@ static void run_one_async_free(struct btrfs_work *work)
 	kfree(async);
 }
 
-int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
-			struct bio *bio, int mirror_num,
-			unsigned long bio_flags,
-			u64 bio_offset,
-			extent_submit_bio_hook_t *submit_bio_start,
-			extent_submit_bio_hook_t *submit_bio_done)
+blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
+				 int mirror_num, unsigned long bio_flags,
+				 u64 bio_offset, void *private_data,
+				 extent_submit_bio_hook_t *submit_bio_start,
+				 extent_submit_bio_hook_t *submit_bio_done)
 {
 	struct async_submit_bio *async;
 
 	async = kmalloc(sizeof(*async), GFP_NOFS);
 	if (!async)
-		return -ENOMEM;
+		return BLK_STS_RESOURCE;
 
-	async->inode = inode;
+	async->private_data = private_data;
+	async->fs_info = fs_info;
 	async->bio = bio;
 	async->mirror_num = mirror_num;
 	async->submit_bio_start = submit_bio_start;
@@ -941,7 +940,7 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 	async->bio_flags = bio_flags;
 	async->bio_offset = bio_offset;
 
-	async->error = 0;
+	async->status = 0;
 
 	atomic_inc(&fs_info->nr_async_submits);
 
@@ -959,7 +958,7 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode,
 	return 0;
 }
 
-static int btree_csum_one_bio(struct bio *bio)
+static blk_status_t btree_csum_one_bio(struct bio *bio)
 {
 	struct bio_vec *bvec;
 	struct btrfs_root *root;
@@ -972,12 +971,12 @@ static int btree_csum_one_bio(struct bio *bio)
 			break;
 	}
 
-	return ret;
+	return errno_to_blk_status(ret);
 }
 
-static int __btree_submit_bio_start(struct inode *inode, struct bio *bio,
-				    int mirror_num, unsigned long bio_flags,
-				    u64 bio_offset)
+static blk_status_t __btree_submit_bio_start(void *private_data, struct bio *bio,
+					     int mirror_num, unsigned long bio_flags,
+					     u64 bio_offset)
 {
 	/*
 	 * when we're called for a write, we're already in the async
@@ -986,11 +985,12 @@ static int __btree_submit_bio_start(struct inode *inode, struct bio *bio,
 	return btree_csum_one_bio(bio);
 }
 
-static int __btree_submit_bio_done(struct inode *inode, struct bio *bio,
-				 int mirror_num, unsigned long bio_flags,
-				 u64 bio_offset)
+static blk_status_t __btree_submit_bio_done(void *private_data, struct bio *bio,
+					    int mirror_num, unsigned long bio_flags,
+					    u64 bio_offset)
 {
-	int ret;
+	struct inode *inode = private_data;
+	blk_status_t ret;
 
 	/*
 	 * when we're called for a write, we're already in the async
@@ -998,7 +998,7 @@ static int __btree_submit_bio_done(struct inode *inode, struct bio *bio,
 	 */
 	ret = btrfs_map_bio(btrfs_sb(inode->i_sb), bio, mirror_num, 1);
 	if (ret) {
-		bio->bi_error = ret;
+		bio->bi_status = ret;
 		bio_endio(bio);
 	}
 	return ret;
@@ -1015,13 +1015,14 @@ static int check_async_write(unsigned long bio_flags)
 	return 1;
 }
 
-static int btree_submit_bio_hook(struct inode *inode, struct bio *bio,
-				 int mirror_num, unsigned long bio_flags,
-				 u64 bio_offset)
+static blk_status_t btree_submit_bio_hook(void *private_data, struct bio *bio,
+					  int mirror_num, unsigned long bio_flags,
+					  u64 bio_offset)
 {
+	struct inode *inode = private_data;
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	int async = check_async_write(bio_flags);
-	int ret;
+	blk_status_t ret;
 
 	if (bio_op(bio) != REQ_OP_WRITE) {
 		/*
@@ -1043,8 +1044,8 @@ static int btree_submit_bio_hook(struct inode *inode, struct bio *bio,
 		 * kthread helpers are used to submit writes so that
 		 * checksumming can happen in parallel across all CPUs
 		 */
-		ret = btrfs_wq_submit_bio(fs_info, inode, bio, mirror_num, 0,
-					  bio_offset,
+		ret = btrfs_wq_submit_bio(fs_info, bio, mirror_num, 0,
+					  bio_offset, private_data,
 					  __btree_submit_bio_start,
 					  __btree_submit_bio_done);
 	}
@@ -1054,7 +1055,7 @@ static int btree_submit_bio_hook(struct inode *inode, struct bio *bio,
 	return 0;
 
 out_w_error:
-	bio->bi_error = ret;
+	bio->bi_status = ret;
 	bio_endio(bio);
 	return ret;
 }
@@ -1222,10 +1223,10 @@ int btrfs_write_tree_block(struct extent_buffer *buf)
 					buf->start + buf->len - 1);
 }
 
-int btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
+void btrfs_wait_tree_block_writeback(struct extent_buffer *buf)
 {
-	return filemap_fdatawait_range(buf->pages[0]->mapping,
-				       buf->start, buf->start + buf->len - 1);
+	filemap_fdatawait_range(buf->pages[0]->mapping,
+			        buf->start, buf->start + buf->len - 1);
 }
 
 struct extent_buffer *read_tree_block(struct btrfs_fs_info *fs_info, u64 bytenr,
@@ -1340,15 +1341,14 @@ static void __setup_root(struct btrfs_root *root, struct btrfs_fs_info *fs_info,
 	atomic_set(&root->log_writers, 0);
 	atomic_set(&root->log_batch, 0);
 	atomic_set(&root->orphan_inodes, 0);
-	atomic_set(&root->refs, 1);
+	refcount_set(&root->refs, 1);
 	atomic_set(&root->will_be_snapshoted, 0);
 	atomic64_set(&root->qgroup_meta_rsv, 0);
 	root->log_transid = 0;
 	root->log_transid_committed = -1;
 	root->last_log_commit = 0;
 	if (!dummy)
-		extent_io_tree_init(&root->dirty_log_pages,
-				     fs_info->btree_inode->i_mapping);
+		extent_io_tree_init(&root->dirty_log_pages, NULL);
 
 	memset(&root->root_key, 0, sizeof(root->root_key));
 	memset(&root->root_item, 0, sizeof(root->root_item));
@@ -1820,7 +1820,7 @@ static void end_workqueue_fn(struct btrfs_work *work)
 	end_io_wq = container_of(work, struct btrfs_end_io_wq, work);
 	bio = end_io_wq->bio;
 
-	bio->bi_error = end_io_wq->error;
+	bio->bi_status = end_io_wq->status;
 	bio->bi_private = end_io_wq->private;
 	bio->bi_end_io = end_io_wq->end_io;
 	kmem_cache_free(btrfs_end_io_wq_cache, end_io_wq);
@@ -2309,7 +2309,7 @@ static void btrfs_init_btree_inode(struct btrfs_fs_info *fs_info)
 	inode->i_mapping->a_ops = &btree_aops;
 
 	RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
-	extent_io_tree_init(&BTRFS_I(inode)->io_tree, inode->i_mapping);
+	extent_io_tree_init(&BTRFS_I(inode)->io_tree, inode);
 	BTRFS_I(inode)->io_tree.track_uptodate = 0;
 	extent_map_tree_init(&BTRFS_I(inode)->extent_tree);
 
@@ -2626,7 +2626,6 @@ int 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);
 	spin_lock_init(&fs_info->tree_mod_seq_lock);
 	spin_lock_init(&fs_info->super_lock);
 	spin_lock_init(&fs_info->qgroup_op_lock);
@@ -2662,12 +2661,11 @@ int open_ctree(struct super_block *sb,
 	atomic_set(&fs_info->qgroup_op_seq, 0);
 	atomic_set(&fs_info->reada_works_cnt, 0);
 	atomic64_set(&fs_info->tree_mod_seq, 0);
-	fs_info->fs_frozen = 0;
 	fs_info->sb = sb;
 	fs_info->max_inline = BTRFS_DEFAULT_MAX_INLINE;
 	fs_info->metadata_ratio = 0;
 	fs_info->defrag_inodes = RB_ROOT;
-	fs_info->free_chunk_space = 0;
+	atomic64_set(&fs_info->free_chunk_space, 0);
 	fs_info->tree_mod_log = RB_ROOT;
 	fs_info->commit_interval = BTRFS_DEFAULT_COMMIT_INTERVAL;
 	fs_info->avg_delayed_ref_runtime = NSEC_PER_SEC >> 6; /* div by 64 */
@@ -2704,10 +2702,8 @@ int open_ctree(struct super_block *sb,
 	fs_info->block_group_cache_tree = RB_ROOT;
 	fs_info->first_logical_byte = (u64)-1;
 
-	extent_io_tree_init(&fs_info->freed_extents[0],
-			     fs_info->btree_inode->i_mapping);
-	extent_io_tree_init(&fs_info->freed_extents[1],
-			     fs_info->btree_inode->i_mapping);
+	extent_io_tree_init(&fs_info->freed_extents[0], NULL);
+	extent_io_tree_init(&fs_info->freed_extents[1], NULL);
 	fs_info->pinned_extents = &fs_info->freed_extents[0];
 	set_bit(BTRFS_FS_BARRIER, &fs_info->flags);
 
@@ -3467,10 +3463,12 @@ static int write_dev_supers(struct btrfs_device *device,
 		 * we fua the first super.  The others we allow
 		 * to go down lazy.
 		 */
-		if (i == 0)
-			ret = btrfsic_submit_bh(REQ_OP_WRITE, REQ_FUA, bh);
-		else
+		if (i == 0) {
+			ret = btrfsic_submit_bh(REQ_OP_WRITE,
+						REQ_SYNC | REQ_FUA, bh);
+		} else {
 			ret = btrfsic_submit_bh(REQ_OP_WRITE, REQ_SYNC, bh);
+		}
 		if (ret)
 			errors++;
 	}
@@ -3483,64 +3481,61 @@ static int write_dev_supers(struct btrfs_device *device,
  */
 static void btrfs_end_empty_barrier(struct bio *bio)
 {
-	if (bio->bi_private)
-		complete(bio->bi_private);
-	bio_put(bio);
+	complete(bio->bi_private);
 }
 
 /*
- * 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
+ * Submit a flush request to the device if it supports it. Error handling is
+ * done in the waiting counterpart.
  */
-static int write_dev_flush(struct btrfs_device *device, int wait)
+static void write_dev_flush(struct btrfs_device *device)
 {
-	struct bio *bio;
-	int ret = 0;
+	struct request_queue *q = bdev_get_queue(device->bdev);
+	struct bio *bio = device->flush_bio;
 
-	if (device->nobarriers)
-		return 0;
+	if (!test_bit(QUEUE_FLAG_WC, &q->queue_flags))
+		return;
 
-	if (wait) {
-		bio = device->flush_bio;
-		if (!bio)
-			return 0;
+	bio_reset(bio);
+	bio->bi_end_io = btrfs_end_empty_barrier;
+	bio->bi_bdev = device->bdev;
+	bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_PREFLUSH;
+	init_completion(&device->flush_wait);
+	bio->bi_private = &device->flush_wait;
 
-		wait_for_completion(&device->flush_wait);
+	submit_bio(bio);
+	device->flush_bio_sent = 1;
+}
 
-		if (bio->bi_error) {
-			ret = bio->bi_error;
-			btrfs_dev_stat_inc_and_print(device,
-				BTRFS_DEV_STAT_FLUSH_ERRS);
-		}
+/*
+ * If the flush bio has been submitted by write_dev_flush, wait for it.
+ */
+static blk_status_t wait_dev_flush(struct btrfs_device *device)
+{
+	struct bio *bio = device->flush_bio;
 
-		/* drop the reference from the wait == 0 run */
-		bio_put(bio);
-		device->flush_bio = NULL;
+	if (!device->flush_bio_sent)
+		return 0;
 
-		return ret;
-	}
+	device->flush_bio_sent = 0;
+	wait_for_completion_io(&device->flush_wait);
 
-	/*
-	 * one reference for us, and we leave it for the
-	 * caller
-	 */
-	device->flush_bio = NULL;
-	bio = btrfs_io_bio_alloc(GFP_NOFS, 0);
-	if (!bio)
-		return -ENOMEM;
+	return bio->bi_status;
+}
 
-	bio->bi_end_io = btrfs_end_empty_barrier;
-	bio->bi_bdev = device->bdev;
-	bio->bi_opf = REQ_OP_WRITE | REQ_PREFLUSH;
-	init_completion(&device->flush_wait);
-	bio->bi_private = &device->flush_wait;
-	device->flush_bio = bio;
+static int check_barrier_error(struct btrfs_fs_devices *fsdevs)
+{
+	int dev_flush_error = 0;
+	struct btrfs_device *dev;
+
+	list_for_each_entry_rcu(dev, &fsdevs->devices, dev_list) {
+		if (!dev->bdev || dev->last_flush_error)
+			dev_flush_error++;
+	}
 
-	bio_get(bio);
-	btrfsic_submit_bio(bio);
+	if (dev_flush_error >
+	    fsdevs->fs_info->num_tolerated_disk_barrier_failures)
+		return -EIO;
 
 	return 0;
 }
@@ -3553,25 +3548,21 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 {
 	struct list_head *head;
 	struct btrfs_device *dev;
-	int errors_send = 0;
 	int errors_wait = 0;
-	int ret;
+	blk_status_t ret;
 
 	/* send down all the barriers */
 	head = &info->fs_devices->devices;
 	list_for_each_entry_rcu(dev, head, dev_list) {
 		if (dev->missing)
 			continue;
-		if (!dev->bdev) {
-			errors_send++;
+		if (!dev->bdev)
 			continue;
-		}
 		if (!dev->in_fs_metadata || !dev->writeable)
 			continue;
 
-		ret = write_dev_flush(dev, 0);
-		if (ret)
-			errors_send++;
+		write_dev_flush(dev);
+		dev->last_flush_error = 0;
 	}
 
 	/* wait for all the barriers */
@@ -3585,13 +3576,23 @@ static int barrier_all_devices(struct btrfs_fs_info *info)
 		if (!dev->in_fs_metadata || !dev->writeable)
 			continue;
 
-		ret = write_dev_flush(dev, 1);
-		if (ret)
+		ret = wait_dev_flush(dev);
+		if (ret) {
+			dev->last_flush_error = ret;
+			btrfs_dev_stat_inc_and_print(dev,
+					BTRFS_DEV_STAT_FLUSH_ERRS);
 			errors_wait++;
+		}
+	}
+
+	if (errors_wait) {
+		/*
+		 * At some point we need the status of all disks
+		 * to arrive at the volume status. So error checking
+		 * is being pushed to a separate loop.
+		 */
+		return check_barrier_error(info->fs_devices);
 	}
-	if (errors_send > info->num_tolerated_disk_barrier_failures ||
-	    errors_wait > info->num_tolerated_disk_barrier_failures)
-		return -EIO;
 	return 0;
 }
 
@@ -4321,7 +4322,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans,
 		head = rb_entry(node, struct btrfs_delayed_ref_head,
 				href_node);
 		if (!mutex_trylock(&head->mutex)) {
-			atomic_inc(&head->node.refs);
+			refcount_inc(&head->node.refs);
 			spin_unlock(&delayed_refs->lock);
 
 			mutex_lock(&head->mutex);
@@ -4575,11 +4576,6 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 
 	cur_trans->state =TRANS_STATE_COMPLETED;
 	wake_up(&cur_trans->commit_wait);
-
-	/*
-	memset(cur_trans, 0, sizeof(*cur_trans));
-	kmem_cache_free(btrfs_transaction_cachep, cur_trans);
-	*/
 }
 
 static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info)
@@ -4593,7 +4589,7 @@ static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info)
 		t = list_first_entry(&fs_info->trans_list,
 				     struct btrfs_transaction, list);
 		if (t->state >= TRANS_STATE_COMMIT_START) {
-			atomic_inc(&t->use_count);
+			refcount_inc(&t->use_count);
 			spin_unlock(&fs_info->trans_lock);
 			btrfs_wait_for_commit(fs_info, t->transid);
 			btrfs_put_transaction(t);
@@ -4635,6 +4631,12 @@ static int btrfs_cleanup_transaction(struct btrfs_fs_info *fs_info)
 	return 0;
 }
 
+static struct btrfs_fs_info *btree_fs_info(void *private_data)
+{
+	struct inode *inode = private_data;
+	return btrfs_sb(inode->i_sb);
+}
+
 static const struct extent_io_ops btree_extent_io_ops = {
 	/* mandatory callbacks */
 	.submit_bio_hook = btree_submit_bio_hook,
@@ -4642,6 +4644,8 @@ static const struct extent_io_ops btree_extent_io_ops = {
 	/* note we're sharing with inode.c for the merge bio hook */
 	.merge_bio_hook = btrfs_merge_bio_hook,
 	.readpage_io_failed_hook = btree_io_failed_hook,
+	.set_range_writeback = btrfs_set_range_writeback,
+	.tree_fs_info = btree_fs_info,
 
 	/* optional callbacks */
 };