Merge tag 'for-5.11-rc5-tag' of git://git.kernel.org/pub/scm/linux/ke…

…rnel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "A few more fixes for a late rc:

   - fix lockdep complaint on 32bit arches and also remove an unsafe
     memory use due to device vs filesystem lifetime

   - two fixes for free space tree:

      * race during log replay and cache rebuild, now more likely to
        happen due to changes in this dev cycle

      * possible free space tree corruption with online conversion
        during initial tree population"

* tag 'for-5.11-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: fix log replay failure due to race with space cache rebuild
  btrfs: fix lockdep warning due to seqcount_mutex on 32bit arch
  btrfs: fix possible free space tree corruption with online conversion

fs/btrfs/block-group.c

@@ -673,7 +673,15 @@ static noinline void caching_thread(struct btrfs_work *work)
 		wake_up(&caching_ctl->wait);
 	}
 
-	if (btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE))
+	/*
+	 * If we are in the transaction that populated the free space tree we
+	 * can't actually cache from the free space tree as our commit root and
+	 * real root are the same, so we could change the contents of the blocks
+	 * while caching.  Instead do the slow caching in this case, and after
+	 * the transaction has committed we will be safe.
+	 */
+	if (btrfs_fs_compat_ro(fs_info, FREE_SPACE_TREE) &&
+	    !(test_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags)))
 		ret = load_free_space_tree(caching_ctl);
 	else
 		ret = load_extent_tree_free(caching_ctl);

fs/btrfs/ctree.h

@@ -563,6 +563,9 @@ enum {
 
 	/* Indicate that we need to cleanup space cache v1 */
 	BTRFS_FS_CLEANUP_SPACE_CACHE_V1,
+
+	/* Indicate that we can't trust the free space tree for caching yet */
+	BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED,
 };
 
 /*

fs/btrfs/extent-tree.c

@@ -2602,8 +2602,6 @@ int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,
 	struct btrfs_block_group *cache;
 	int ret;
 
-	btrfs_add_excluded_extent(trans->fs_info, bytenr, num_bytes);
-
 	cache = btrfs_lookup_block_group(trans->fs_info, bytenr);
 	if (!cache)
 		return -EINVAL;
@@ -2615,11 +2613,19 @@ int btrfs_pin_extent_for_log_replay(struct btrfs_trans_handle *trans,
 	 * the pinned extents.
 	 */
 	btrfs_cache_block_group(cache, 1);
+	/*
+	 * Make sure we wait until the cache is completely built in case it is
+	 * missing or is invalid and therefore needs to be rebuilt.
+	 */
+	ret = btrfs_wait_block_group_cache_done(cache);
+	if (ret)
+		goto out;
 
 	pin_down_extent(trans, cache, bytenr, num_bytes, 0);
 
 	/* remove us from the free space cache (if we're there at all) */
 	ret = btrfs_remove_free_space(cache, bytenr, num_bytes);
+out:
 	btrfs_put_block_group(cache);
 	return ret;
 }
@@ -2629,50 +2635,22 @@ static int __exclude_logged_extent(struct btrfs_fs_info *fs_info,
 {
 	int ret;
 	struct btrfs_block_group *block_group;
-	struct btrfs_caching_control *caching_ctl;
 
 	block_group = btrfs_lookup_block_group(fs_info, start);
 	if (!block_group)
 		return -EINVAL;
 
-	btrfs_cache_block_group(block_group, 0);
-	caching_ctl = btrfs_get_caching_control(block_group);
-
-	if (!caching_ctl) {
-		/* Logic error */
-		BUG_ON(!btrfs_block_group_done(block_group));
-		ret = btrfs_remove_free_space(block_group, start, num_bytes);
-	} else {
-		/*
-		 * We must wait for v1 caching to finish, otherwise we may not
-		 * remove our space.
-		 */
-		btrfs_wait_space_cache_v1_finished(block_group, caching_ctl);
-		mutex_lock(&caching_ctl->mutex);
-
-		if (start >= caching_ctl->progress) {
-			ret = btrfs_add_excluded_extent(fs_info, start,
-							num_bytes);
-		} else if (start + num_bytes <= caching_ctl->progress) {
-			ret = btrfs_remove_free_space(block_group,
-						      start, num_bytes);
-		} else {
-			num_bytes = caching_ctl->progress - start;
-			ret = btrfs_remove_free_space(block_group,
-						      start, num_bytes);
-			if (ret)
-				goto out_lock;
+	btrfs_cache_block_group(block_group, 1);
+	/*
+	 * Make sure we wait until the cache is completely built in case it is
+	 * missing or is invalid and therefore needs to be rebuilt.
+	 */
+	ret = btrfs_wait_block_group_cache_done(block_group);
+	if (ret)
+		goto out;
 
-			num_bytes = (start + num_bytes) -
-				caching_ctl->progress;
-			start = caching_ctl->progress;
-			ret = btrfs_add_excluded_extent(fs_info, start,
-							num_bytes);
-		}
-out_lock:
-		mutex_unlock(&caching_ctl->mutex);
-		btrfs_put_caching_control(caching_ctl);
-	}
+	ret = btrfs_remove_free_space(block_group, start, num_bytes);
+out:
 	btrfs_put_block_group(block_group);
 	return ret;
 }
@@ -2863,9 +2841,6 @@ int btrfs_finish_extent_commit(struct btrfs_trans_handle *trans)
 			mutex_unlock(&fs_info->unused_bg_unpin_mutex);
 			break;
 		}
-		if (test_bit(BTRFS_FS_LOG_RECOVERING, &fs_info->flags))
-			clear_extent_bits(&fs_info->excluded_extents, start,
-					  end, EXTENT_UPTODATE);
 
 		if (btrfs_test_opt(fs_info, DISCARD_SYNC))
 			ret = btrfs_discard_extent(fs_info, start,

fs/btrfs/free-space-tree.c

@@ -1150,6 +1150,7 @@ int btrfs_create_free_space_tree(struct btrfs_fs_info *fs_info)
 		return PTR_ERR(trans);
 
 	set_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
+	set_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
 	free_space_root = btrfs_create_tree(trans,
 					    BTRFS_FREE_SPACE_TREE_OBJECTID);
 	if (IS_ERR(free_space_root)) {
@@ -1171,11 +1172,18 @@ int btrfs_create_free_space_tree(struct btrfs_fs_info *fs_info)
 	btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE);
 	btrfs_set_fs_compat_ro(fs_info, FREE_SPACE_TREE_VALID);
 	clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
+	ret = btrfs_commit_transaction(trans);
 
-	return btrfs_commit_transaction(trans);
+	/*
+	 * Now that we've committed the transaction any reading of our commit
+	 * root will be safe, so we can cache from the free space tree now.
+	 */
+	clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
+	return ret;
 
 abort:
 	clear_bit(BTRFS_FS_CREATING_FREE_SPACE_TREE, &fs_info->flags);
+	clear_bit(BTRFS_FS_FREE_SPACE_TREE_UNTRUSTED, &fs_info->flags);
 	btrfs_abort_transaction(trans, ret);
 	btrfs_end_transaction(trans);
 	return ret;

fs/btrfs/volumes.c

@@ -433,7 +433,7 @@ static struct btrfs_device *__alloc_device(struct btrfs_fs_info *fs_info)
 
 	atomic_set(&dev->reada_in_flight, 0);
 	atomic_set(&dev->dev_stats_ccnt, 0);
-	btrfs_device_data_ordered_init(dev, fs_info);
+	btrfs_device_data_ordered_init(dev);
 	INIT_RADIX_TREE(&dev->reada_zones, GFP_NOFS & ~__GFP_DIRECT_RECLAIM);
 	INIT_RADIX_TREE(&dev->reada_extents, GFP_NOFS & ~__GFP_DIRECT_RECLAIM);
 	extent_io_tree_init(fs_info, &dev->alloc_state,

fs/btrfs/volumes.h

@@ -39,10 +39,10 @@ struct btrfs_io_geometry {
 #if BITS_PER_LONG==32 && defined(CONFIG_SMP)
 #include <linux/seqlock.h>
 #define __BTRFS_NEED_DEVICE_DATA_ORDERED
-#define btrfs_device_data_ordered_init(device, info)				\
-	seqcount_mutex_init(&device->data_seqcount, &info->chunk_mutex)
+#define btrfs_device_data_ordered_init(device)	\
+	seqcount_init(&device->data_seqcount)
 #else
-#define btrfs_device_data_ordered_init(device, info) do { } while (0)
+#define btrfs_device_data_ordered_init(device) do { } while (0)
 #endif
 
 #define BTRFS_DEV_STATE_WRITEABLE	(0)
@@ -76,8 +76,7 @@ struct btrfs_device {
 	blk_status_t last_flush_error;
 
 #ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED
-	/* A seqcount_t with associated chunk_mutex (for lockdep) */
-	seqcount_mutex_t data_seqcount;
+	seqcount_t data_seqcount;
 #endif
 
 	/* the internal btrfs device id */
@@ -168,9 +167,11 @@ btrfs_device_get_##name(const struct btrfs_device *dev)			\
 static inline void							\
 btrfs_device_set_##name(struct btrfs_device *dev, u64 size)		\
 {									\
+	preempt_disable();						\
 	write_seqcount_begin(&dev->data_seqcount);			\
 	dev->name = size;						\
 	write_seqcount_end(&dev->data_seqcount);			\
+	preempt_enable();						\
 }
 #elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPTION)
 #define BTRFS_DEVICE_GETSET_FUNCS(name)					\