Merge branch 'for-linus-4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs

Pull btrfs fixes from Chris Mason:
 "My patch fixes the btrfs list_head abuse that we tracked down during
  Dave Jones' memory corruption investigation. With both Jens and my
  patches in place, I'm no longer able to trigger problems.

  Filipe is fixing a difficult old bug between snapshots, balance and
  send. Dave is cooking a few more for the next rc, but these are tested
  and ready"

* 'for-linus-4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  btrfs: fix races on root_log_ctx lists
  btrfs: fix incremental send failure caused by balance

2 files changed, 64 insertions, 14 deletions

diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index 01bc36cec26e..71261b459863 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -5805,6 +5805,64 @@ static int changed_extent(struct send_ctx *sctx,
 	int ret = 0;
 
 	if (sctx->cur_ino != sctx->cmp_key->objectid) {
+
+		if (result == BTRFS_COMPARE_TREE_CHANGED) {
+			struct extent_buffer *leaf_l;
+			struct extent_buffer *leaf_r;
+			struct btrfs_file_extent_item *ei_l;
+			struct btrfs_file_extent_item *ei_r;
+
+			leaf_l = sctx->left_path->nodes[0];
+			leaf_r = sctx->right_path->nodes[0];
+			ei_l = btrfs_item_ptr(leaf_l,
+					      sctx->left_path->slots[0],
+					      struct btrfs_file_extent_item);
+			ei_r = btrfs_item_ptr(leaf_r,
+					      sctx->right_path->slots[0],
+					      struct btrfs_file_extent_item);
+
+			/*
+			 * We may have found an extent item that has changed
+			 * only its disk_bytenr field and the corresponding
+			 * inode item was not updated. This case happens due to
+			 * very specific timings during relocation when a leaf
+			 * that contains file extent items is COWed while
+			 * relocation is ongoing and its in the stage where it
+			 * updates data pointers. So when this happens we can
+			 * safely ignore it since we know it's the same extent,
+			 * but just at different logical and physical locations
+			 * (when an extent is fully replaced with a new one, we
+			 * know the generation number must have changed too,
+			 * since snapshot creation implies committing the current
+			 * transaction, and the inode item must have been updated
+			 * as well).
+			 * This replacement of the disk_bytenr happens at
+			 * relocation.c:replace_file_extents() through
+			 * relocation.c:btrfs_reloc_cow_block().
+			 */
+			if (btrfs_file_extent_generation(leaf_l, ei_l) ==
+			    btrfs_file_extent_generation(leaf_r, ei_r) &&
+			    btrfs_file_extent_ram_bytes(leaf_l, ei_l) ==
+			    btrfs_file_extent_ram_bytes(leaf_r, ei_r) &&
+			    btrfs_file_extent_compression(leaf_l, ei_l) ==
+			    btrfs_file_extent_compression(leaf_r, ei_r) &&
+			    btrfs_file_extent_encryption(leaf_l, ei_l) ==
+			    btrfs_file_extent_encryption(leaf_r, ei_r) &&
+			    btrfs_file_extent_other_encoding(leaf_l, ei_l) ==
+			    btrfs_file_extent_other_encoding(leaf_r, ei_r) &&
+			    btrfs_file_extent_type(leaf_l, ei_l) ==
+			    btrfs_file_extent_type(leaf_r, ei_r) &&
+			    btrfs_file_extent_disk_bytenr(leaf_l, ei_l) !=
+			    btrfs_file_extent_disk_bytenr(leaf_r, ei_r) &&
+			    btrfs_file_extent_disk_num_bytes(leaf_l, ei_l) ==
+			    btrfs_file_extent_disk_num_bytes(leaf_r, ei_r) &&
+			    btrfs_file_extent_offset(leaf_l, ei_l) ==
+			    btrfs_file_extent_offset(leaf_r, ei_r) &&
+			    btrfs_file_extent_num_bytes(leaf_l, ei_l) ==
+			    btrfs_file_extent_num_bytes(leaf_r, ei_r))
+				return 0;
+		}
+
 		inconsistent_snapshot_error(sctx, result, "extent");
 		return -EIO;
 	}
diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c
index 528cae123dc9..3d33c4e41e5f 100644
--- a/fs/btrfs/tree-log.c
+++ b/fs/btrfs/tree-log.c
@@ -2713,14 +2713,12 @@ static inline void btrfs_remove_all_log_ctxs(struct btrfs_root *root,
 					     int index, int error)
 {
 	struct btrfs_log_ctx *ctx;
+	struct btrfs_log_ctx *safe;
 
-	if (!error) {
-		INIT_LIST_HEAD(&root->log_ctxs[index]);
-		return;
-	}
-
-	list_for_each_entry(ctx, &root->log_ctxs[index], list)
+	list_for_each_entry_safe(ctx, safe, &root->log_ctxs[index], list) {
+		list_del_init(&ctx->list);
 		ctx->log_ret = error;
+	}
 
 	INIT_LIST_HEAD(&root->log_ctxs[index]);
 }
@@ -2961,13 +2959,9 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	mutex_unlock(&root->log_mutex);
 
 out_wake_log_root:
-	/*
-	 * We needn't get log_mutex here because we are sure all
-	 * the other tasks are blocked.
-	 */
+	mutex_lock(&log_root_tree->log_mutex);
 	btrfs_remove_all_log_ctxs(log_root_tree, index2, ret);
 
-	mutex_lock(&log_root_tree->log_mutex);
 	log_root_tree->log_transid_committed++;
 	atomic_set(&log_root_tree->log_commit[index2], 0);
 	mutex_unlock(&log_root_tree->log_mutex);
@@ -2978,10 +2972,8 @@ out_wake_log_root:
 	if (waitqueue_active(&log_root_tree->log_commit_wait[index2]))
 		wake_up(&log_root_tree->log_commit_wait[index2]);
 out:
-	/* See above. */
-	btrfs_remove_all_log_ctxs(root, index1, ret);
-
 	mutex_lock(&root->log_mutex);
+	btrfs_remove_all_log_ctxs(root, index1, ret);
 	root->log_transid_committed++;
 	atomic_set(&root->log_commit[index1], 0);
 	mutex_unlock(&root->log_mutex);