for-6.5-rc5-tag

-----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCgAdFiEE8rQSAMVO+zA4DBdWxWXV+ddtWDsFAmTXzuUACgkQxWXV+ddt
 WDvQVg/+PwDYtfsFBBxWboR/Ehu+nGj+PGRGH5kUumCt03760GtVMYqJzakinAoA
 TUg7N+SvC0i6STUQ1LxkqdyU+eHxk0D1qwK7HJtbqNQJ+kwaEPlilHwsptMmuuM/
 xaei+C8gLmQreL5ZH6ZsnLfV4aaFR7Ur8KSiAq28H6dKXGnh4q9yio2BspeFoc4Y
 8cD2Q8eOUxLBbGkAy9RHeMWf6OMOv2jyzdA761NZrjxUe23bDWSdRM6cRhfdJIh+
 gfwW1IVH2EVOwo+FeaIpMSf4dpnenOYOKOftTncrz7XS0VEN/wJYQXGjNbLa7u4d
 RxV2RujzRPePAUKDbLRakfXotcuKdSQuX2epLSYkQTfGQ0KRYu5YIDQgkm3r7Yky
 cF5mkyEyI8lFCiop7Bgi3MqnzoY5ZgWAkWSy9/TzjQ4yRhjiZ3fmk5JgoJ8gwUc3
 Fle4czcmKvk6ZqQAn90b0qGtW9FXzVAekZjLAH26O7+dgEn+CCAfwT9GuG7h+ATM
 9Bh+5U5PWxWmNPTYU8Sn+WR9HpVL6+1maxrax/Ftb8/FuFlQXFHxK+OnTcKx9K+y
 OGsv0r/4Zv517k1qqlHvf397Jvz7MmYLyOwkqu5xyomCGtrKIBkkEGF/9sHrZJVM
 YokgphDZL8AILrnnPwCOgt4lsph1VKS/Sgvu7XKovnZbvvh8S+M=
 =csAj
 -----END PGP SIGNATURE-----

Merge tag 'for-6.5-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux

Pull btrfs fixes from David Sterba:
 "More fixes, some of them going back to older releases and there are
  fixes for hangs in stress tests regarding space caching:

   - fixes and progress tracking for hangs in free space caching, found
     by test generic/475

   - writeback fixes, write pages in integrity mode and skip writing
     pages that have been written meanwhile

   - properly clear end of extent range after an error

   - relocation fixes:
      - fix race betwen qgroup tree creation and relocation
      - detect and report invalid reloc roots"

* tag 'for-6.5-rc5-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
  btrfs: set cache_block_group_error if we find an error
  btrfs: reject invalid reloc tree root keys with stack dump
  btrfs: exit gracefully if reloc roots don't match
  btrfs: avoid race between qgroup tree creation and relocation
  btrfs: properly clear end of the unreserved range in cow_file_range
  btrfs: don't wait for writeback on clean pages in extent_write_cache_pages
  btrfs: don't stop integrity writeback too early
  btrfs: wait for actual caching progress during allocation

fs/btrfs/block-group.c

@@ -441,13 +441,23 @@ void btrfs_wait_block_group_cache_progress(struct btrfs_block_group *cache,
 					   u64 num_bytes)
 {
 	struct btrfs_caching_control *caching_ctl;
+	int progress;
 
 	caching_ctl = btrfs_get_caching_control(cache);
 	if (!caching_ctl)
 		return;
 
+	/*
+	 * We've already failed to allocate from this block group, so even if
+	 * there's enough space in the block group it isn't contiguous enough to
+	 * allow for an allocation, so wait for at least the next wakeup tick,
+	 * or for the thing to be done.
+	 */
+	progress = atomic_read(&caching_ctl->progress);
+
 	wait_event(caching_ctl->wait, btrfs_block_group_done(cache) ||
-		   (cache->free_space_ctl->free_space >= num_bytes));
+		   (progress != atomic_read(&caching_ctl->progress) &&
+		    (cache->free_space_ctl->free_space >= num_bytes)));
 
 	btrfs_put_caching_control(caching_ctl);
 }
@@ -802,8 +812,10 @@ static int load_extent_tree_free(struct btrfs_caching_control *caching_ctl)
 
 			if (total_found > CACHING_CTL_WAKE_UP) {
 				total_found = 0;
-				if (wakeup)
+				if (wakeup) {
+					atomic_inc(&caching_ctl->progress);
 					wake_up(&caching_ctl->wait);
+				}
 			}
 		}
 		path->slots[0]++;
@@ -910,6 +922,7 @@ int btrfs_cache_block_group(struct btrfs_block_group *cache, bool wait)
 	init_waitqueue_head(&caching_ctl->wait);
 	caching_ctl->block_group = cache;
 	refcount_set(&caching_ctl->count, 2);
+	atomic_set(&caching_ctl->progress, 0);
 	btrfs_init_work(&caching_ctl->work, caching_thread, NULL, NULL);
 
 	spin_lock(&cache->lock);

fs/btrfs/block-group.h

@@ -90,6 +90,8 @@ struct btrfs_caching_control {
 	wait_queue_head_t wait;
 	struct btrfs_work work;
 	struct btrfs_block_group *block_group;
+	/* Track progress of caching during allocation. */
+	atomic_t progress;
 	refcount_t count;
 };
 

fs/btrfs/disk-io.c

@@ -1103,7 +1103,8 @@ static int btrfs_init_fs_root(struct btrfs_root *root, dev_t anon_dev)
 	btrfs_drew_lock_init(&root->snapshot_lock);
 
 	if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID &&
-	    !btrfs_is_data_reloc_root(root)) {
+	    !btrfs_is_data_reloc_root(root) &&
+	    is_fstree(root->root_key.objectid)) {
 		set_bit(BTRFS_ROOT_SHAREABLE, &root->state);
 		btrfs_check_and_init_root_item(&root->root_item);
 	}
@@ -1300,6 +1301,16 @@ static struct btrfs_root *btrfs_get_root_ref(struct btrfs_fs_info *fs_info,
 	root = btrfs_get_global_root(fs_info, objectid);
 	if (root)
 		return root;
+
+	/*
+	 * If we're called for non-subvolume trees, and above function didn't
+	 * find one, do not try to read it from disk.
+	 *
+	 * This is namely for free-space-tree and quota tree, which can change
+	 * at runtime and should only be grabbed from fs_info.
+	 */
+	if (!is_fstree(objectid) && objectid != BTRFS_DATA_RELOC_TREE_OBJECTID)
+		return ERR_PTR(-ENOENT);
 again:
 	root = btrfs_lookup_fs_root(fs_info, objectid);
 	if (root) {

fs/btrfs/extent-tree.c

@@ -4310,8 +4310,11 @@ static noinline int find_free_extent(struct btrfs_root *root,
 			ret = 0;
 		}
 
-		if (unlikely(block_group->cached == BTRFS_CACHE_ERROR))
+		if (unlikely(block_group->cached == BTRFS_CACHE_ERROR)) {
+			if (!cache_block_group_error)
+				cache_block_group_error = -EIO;
 			goto loop;
+		}
 
 		if (!find_free_extent_check_size_class(ffe_ctl, block_group))
 			goto loop;

fs/btrfs/extent_io.c

@@ -2145,6 +2145,12 @@ static int extent_write_cache_pages(struct address_space *mapping,
 				continue;
 			}
 
+			if (!folio_test_dirty(folio)) {
+				/* Someone wrote it for us. */
+				folio_unlock(folio);
+				continue;
+			}
+
 			if (wbc->sync_mode != WB_SYNC_NONE) {
 				if (folio_test_writeback(folio))
 					submit_write_bio(bio_ctrl, 0);
@@ -2164,11 +2170,12 @@ static int extent_write_cache_pages(struct address_space *mapping,
 			}
 
 			/*
-			 * the filesystem may choose to bump up nr_to_write.
+			 * The filesystem may choose to bump up nr_to_write.
 			 * We have to make sure to honor the new nr_to_write
-			 * at any time
+			 * at any time.
 			 */
-			nr_to_write_done = wbc->nr_to_write <= 0;
+			nr_to_write_done = (wbc->sync_mode == WB_SYNC_NONE &&
+					    wbc->nr_to_write <= 0);
 		}
 		folio_batch_release(&fbatch);
 		cond_resched();

fs/btrfs/inode.c

@@ -1654,8 +1654,6 @@ static noinline int cow_file_range(struct btrfs_inode *inode,
 					     clear_bits,
 					     page_ops);
 		start += cur_alloc_size;
-		if (start >= end)
-			return ret;
 	}
 
 	/*
@@ -1664,9 +1662,11 @@ static noinline int cow_file_range(struct btrfs_inode *inode,
 	 * space_info's bytes_may_use counter, reserved in
 	 * btrfs_check_data_free_space().
 	 */
-	extent_clear_unlock_delalloc(inode, start, end, locked_page,
-				     clear_bits | EXTENT_CLEAR_DATA_RESV,
-				     page_ops);
+	if (start < end) {
+		clear_bits |= EXTENT_CLEAR_DATA_RESV;
+		extent_clear_unlock_delalloc(inode, start, end, locked_page,
+					     clear_bits, page_ops);
+	}
 	return ret;
 }
 

fs/btrfs/relocation.c

@@ -1916,7 +1916,39 @@ int prepare_to_merge(struct reloc_control *rc, int err)
 				err = PTR_ERR(root);
 			break;
 		}
-		ASSERT(root->reloc_root == reloc_root);
+
+		if (unlikely(root->reloc_root != reloc_root)) {
+			if (root->reloc_root) {
+				btrfs_err(fs_info,
+"reloc tree mismatch, root %lld has reloc root key (%lld %u %llu) gen %llu, expect reloc root key (%lld %u %llu) gen %llu",
+					  root->root_key.objectid,
+					  root->reloc_root->root_key.objectid,
+					  root->reloc_root->root_key.type,
+					  root->reloc_root->root_key.offset,
+					  btrfs_root_generation(
+						  &root->reloc_root->root_item),
+					  reloc_root->root_key.objectid,
+					  reloc_root->root_key.type,
+					  reloc_root->root_key.offset,
+					  btrfs_root_generation(
+						  &reloc_root->root_item));
+			} else {
+				btrfs_err(fs_info,
+"reloc tree mismatch, root %lld has no reloc root, expect reloc root key (%lld %u %llu) gen %llu",
+					  root->root_key.objectid,
+					  reloc_root->root_key.objectid,
+					  reloc_root->root_key.type,
+					  reloc_root->root_key.offset,
+					  btrfs_root_generation(
+						  &reloc_root->root_item));
+			}
+			list_add(&reloc_root->root_list, &reloc_roots);
+			btrfs_put_root(root);
+			btrfs_abort_transaction(trans, -EUCLEAN);
+			if (!err)
+				err = -EUCLEAN;
+			break;
+		}
 
 		/*
 		 * set reference count to 1, so btrfs_recover_relocation
@@ -1989,7 +2021,7 @@ void merge_reloc_roots(struct reloc_control *rc)
 		root = btrfs_get_fs_root(fs_info, reloc_root->root_key.offset,
 					 false);
 		if (btrfs_root_refs(&reloc_root->root_item) > 0) {
-			if (IS_ERR(root)) {
+			if (WARN_ON(IS_ERR(root))) {
 				/*
 				 * For recovery we read the fs roots on mount,
 				 * and if we didn't find the root then we marked
@@ -1998,17 +2030,14 @@ void merge_reloc_roots(struct reloc_control *rc)
 				 * memory.  However there's no reason we can't
 				 * handle the error properly here just in case.
 				 */
-				ASSERT(0);
 				ret = PTR_ERR(root);
 				goto out;
 			}
-			if (root->reloc_root != reloc_root) {
+			if (WARN_ON(root->reloc_root != reloc_root)) {
 				/*
-				 * This is actually impossible without something
-				 * going really wrong (like weird race condition
-				 * or cosmic rays).
+				 * This can happen if on-disk metadata has some
+				 * corruption, e.g. bad reloc tree key offset.
 				 */
-				ASSERT(0);
 				ret = -EINVAL;
 				goto out;
 			}

fs/btrfs/tree-checker.c

@@ -446,6 +446,20 @@ static int check_root_key(struct extent_buffer *leaf, struct btrfs_key *key,
 	btrfs_item_key_to_cpu(leaf, &item_key, slot);
 	is_root_item = (item_key.type == BTRFS_ROOT_ITEM_KEY);
 
+	/*
+	 * Bad rootid for reloc trees.
+	 *
+	 * Reloc trees are only for subvolume trees, other trees only need
+	 * to be COWed to be relocated.
+	 */
+	if (unlikely(is_root_item && key->objectid == BTRFS_TREE_RELOC_OBJECTID &&
+		     !is_fstree(key->offset))) {
+		generic_err(leaf, slot,
+		"invalid reloc tree for root %lld, root id is not a subvolume tree",
+			    key->offset);
+		return -EUCLEAN;
+	}
+
 	/* No such tree id */
 	if (unlikely(key->objectid == 0)) {
 		if (is_root_item)