bcachefs: Fix subvol/snapshot deleting in recovery

fsck doesn't want to run while we're cleaning up deleted snapshots - if
that work needs to be done, we want it to have finished before fsck
runs, otherwise fsck will get confused when it finds multiple keys in
the same snapshot ID equivalence class (i.e. the mechanism that
snapshot deletion uses for cleaning up redundant keys).

Signed-off-by: Kent Overstreet <kent.overstreet@gmail.com>

5 files changed, 111 insertions, 88 deletions

diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h
index c07ea9af561d..7020eee5de21 100644
--- a/fs/bcachefs/bcachefs.h
+++ b/fs/bcachefs/bcachefs.h
@@ -514,6 +514,8 @@ enum {
 	BCH_FS_INITIAL_GC_UNFIXED,	/* kill when we enumerate fsck errors */
 	BCH_FS_NEED_ANOTHER_GC,
 
+	BCH_FS_HAVE_DELETED_SNAPSHOTS,
+
 	/* errors: */
 	BCH_FS_ERROR,
 	BCH_FS_TOPOLOGY_ERROR,
diff --git a/fs/bcachefs/fsck.c b/fs/bcachefs/fsck.c
index 609ac37ff1e9..eda6a6ac3c6e 100644
--- a/fs/bcachefs/fsck.c
+++ b/fs/bcachefs/fsck.c
@@ -953,66 +953,6 @@ static int check_inodes(struct bch_fs *c, bool full)
 	return ret;
 }
 
-static int check_subvol(struct btree_trans *trans,
-			struct btree_iter *iter)
-{
-	struct bkey_s_c k;
-	struct bkey_s_c_subvolume subvol;
-	int ret;
-
-	k = bch2_btree_iter_peek(iter);
-	if (!k.k)
-		return 0;
-
-	ret = bkey_err(k);
-	if (ret)
-		return ret;
-
-	if (k.k->type != KEY_TYPE_subvolume)
-		return 0;
-
-	subvol = bkey_s_c_to_subvolume(k);
-
-	if (BCH_SUBVOLUME_UNLINKED(subvol.v)) {
-		ret = bch2_subvolume_delete(trans, iter->pos.offset);
-		if (ret && ret != -EINTR)
-			bch_err(trans->c, "error deleting subvolume %llu: %i",
-				iter->pos.offset, ret);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-noinline_for_stack
-static int check_subvols(struct bch_fs *c)
-{
-	struct btree_trans trans;
-	struct btree_iter iter;
-	int ret;
-
-	bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
-
-	bch2_trans_iter_init(&trans, &iter, BTREE_ID_subvolumes,
-			     POS_MIN,
-			     BTREE_ITER_INTENT|
-			     BTREE_ITER_PREFETCH);
-
-	do {
-		ret = commit_do(&trans, NULL, NULL,
-				      BTREE_INSERT_LAZY_RW|
-				      BTREE_INSERT_NOFAIL,
-				      check_subvol(&trans, &iter));
-		if (ret)
-			break;
-	} while (bch2_btree_iter_advance(&iter));
-	bch2_trans_iter_exit(&trans, &iter);
-
-	bch2_trans_exit(&trans);
-	return ret;
-}
-
 /*
  * Checking for overlapping extents needs to be reimplemented
  */
@@ -2384,9 +2324,10 @@ static int fix_reflink_p(struct bch_fs *c)
  */
 int bch2_fsck_full(struct bch_fs *c)
 {
-	return  bch2_fs_snapshots_check(c) ?:
+	return  bch2_fs_check_snapshots(c) ?:
+		bch2_fs_check_subvols(c) ?:
+		bch2_delete_dead_snapshots(c) ?:
 		check_inodes(c, true) ?:
-		check_subvols(c) ?:
 		check_extents(c) ?:
 		check_dirents(c) ?:
 		check_xattrs(c) ?:
diff --git a/fs/bcachefs/recovery.c b/fs/bcachefs/recovery.c
index e6aed8d79bea..b7598e26c683 100644
--- a/fs/bcachefs/recovery.c
+++ b/fs/bcachefs/recovery.c
@@ -1396,6 +1396,12 @@ out:
 		bch2_journal_entries_free(c);
 	}
 	kfree(clean);
+
+	if (!ret && test_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags)) {
+		bch2_fs_read_write_early(c);
+		bch2_delete_dead_snapshots_async(c);
+	}
+
 	if (ret)
 		bch_err(c, "Error in recovery: %s (%i)", err, ret);
 	else
diff --git a/fs/bcachefs/subvolume.c b/fs/bcachefs/subvolume.c
index 1865c5b3a2c5..91133b3de325 100644
--- a/fs/bcachefs/subvolume.c
+++ b/fs/bcachefs/subvolume.c
@@ -9,9 +9,6 @@
 
 /* Snapshot tree: */
 
-static void bch2_delete_dead_snapshots_work(struct work_struct *);
-static void bch2_delete_dead_snapshots(struct bch_fs *);
-
 void bch2_snapshot_to_text(struct printbuf *out, struct bch_fs *c,
 			   struct bkey_s_c k)
 {
@@ -249,7 +246,7 @@ static int bch2_snapshot_check(struct btree_trans *trans,
 	return 0;
 }
 
-int bch2_fs_snapshots_check(struct bch_fs *c)
+int bch2_fs_check_snapshots(struct bch_fs *c)
 {
 	struct btree_trans trans;
 	struct btree_iter iter;
@@ -299,6 +296,66 @@ err:
 	return ret;
 }
 
+static int check_subvol(struct btree_trans *trans,
+			struct btree_iter *iter)
+{
+	struct bkey_s_c k;
+	struct bkey_s_c_subvolume subvol;
+	int ret;
+
+	k = bch2_btree_iter_peek(iter);
+	if (!k.k)
+		return 0;
+
+	ret = bkey_err(k);
+	if (ret)
+		return ret;
+
+	if (k.k->type != KEY_TYPE_subvolume)
+		return 0;
+
+	subvol = bkey_s_c_to_subvolume(k);
+
+	if (BCH_SUBVOLUME_UNLINKED(subvol.v)) {
+		ret = bch2_subvolume_delete(trans, iter->pos.offset);
+		if (ret && ret != -EINTR)
+			bch_err(trans->c, "error deleting subvolume %llu: %i",
+				iter->pos.offset, ret);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+int bch2_fs_check_subvols(struct bch_fs *c)
+{
+	struct btree_trans trans;
+	struct btree_iter iter;
+	int ret;
+
+	bch2_trans_init(&trans, c, BTREE_ITER_MAX, 0);
+
+	bch2_trans_iter_init(&trans, &iter, BTREE_ID_subvolumes,
+			     POS_MIN,
+			     BTREE_ITER_INTENT|
+			     BTREE_ITER_PREFETCH);
+
+	do {
+		ret = commit_do(&trans, NULL, NULL,
+				      BTREE_INSERT_LAZY_RW|
+				      BTREE_INSERT_NOFAIL,
+				      check_subvol(&trans, &iter));
+		if (ret)
+			break;
+	} while (bch2_btree_iter_advance(&iter));
+	bch2_trans_iter_exit(&trans, &iter);
+
+	bch2_trans_exit(&trans);
+
+	return ret;
+}
+
 void bch2_fs_snapshots_exit(struct bch_fs *c)
 {
 	genradix_free(&c->snapshots);
@@ -309,7 +366,6 @@ int bch2_fs_snapshots_start(struct bch_fs *c)
 	struct btree_trans trans;
 	struct btree_iter iter;
 	struct bkey_s_c k;
-	bool have_deleted = false;
 	int ret = 0;
 
 	bch2_trans_init(&trans, c, 0, 0);
@@ -326,7 +382,7 @@ int bch2_fs_snapshots_start(struct bch_fs *c)
 		}
 
 		if (BCH_SNAPSHOT_DELETED(bkey_s_c_to_snapshot(k).v))
-			have_deleted = true;
+			set_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags);
 
 		ret = bch2_mark_snapshot(&trans, bkey_s_c_null, k, 0);
 		if (ret)
@@ -342,16 +398,6 @@ int bch2_fs_snapshots_start(struct bch_fs *c)
 		goto err;
 err:
 	bch2_trans_exit(&trans);
-
-	if (!ret && have_deleted) {
-		bch_info(c, "restarting deletion of dead snapshots");
-		if (c->opts.fsck) {
-			bch2_delete_dead_snapshots_work(&c->snapshot_delete_work);
-		} else {
-			bch2_delete_dead_snapshots(c);
-		}
-	}
-
 	return ret;
 }
 
@@ -598,10 +644,6 @@ static int bch2_snapshot_delete_keys_btree(struct btree_trans *trans,
 
 		if (snapshot_list_has_id(deleted, k.k->p.snapshot) ||
 		    snapshot_list_has_id(&equiv_seen, equiv)) {
-			if (btree_id == BTREE_ID_inodes &&
-			    bch2_btree_key_cache_flush(trans, btree_id, iter.pos))
-				continue;
-
 			ret = commit_do(trans, NULL, NULL,
 					      BTREE_INSERT_NOFAIL,
 				bch2_btree_iter_traverse(&iter) ?:
@@ -624,9 +666,8 @@ static int bch2_snapshot_delete_keys_btree(struct btree_trans *trans,
 	return ret;
 }
 
-static void bch2_delete_dead_snapshots_work(struct work_struct *work)
+int bch2_delete_dead_snapshots(struct bch_fs *c)
 {
-	struct bch_fs *c = container_of(work, struct bch_fs, snapshot_delete_work);
 	struct btree_trans trans;
 	struct btree_iter iter;
 	struct bkey_s_c k;
@@ -635,6 +676,17 @@ static void bch2_delete_dead_snapshots_work(struct work_struct *work)
 	u32 i, id, children[2];
 	int ret = 0;
 
+	if (!test_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags))
+		return 0;
+
+	if (!test_bit(BCH_FS_STARTED, &c->flags)) {
+		ret = bch2_fs_read_write_early(c);
+		if (ret) {
+			bch_err(c, "error deleleting dead snapshots: error going rw: %i", ret);
+			return ret;
+		}
+	}
+
 	bch2_trans_init(&trans, c, 0, 0);
 
 	/*
@@ -718,15 +770,25 @@ static void bch2_delete_dead_snapshots_work(struct work_struct *work)
 			goto err;
 		}
 	}
+
+	clear_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags);
 err:
 	darray_exit(&deleted);
 	bch2_trans_exit(&trans);
+	return ret;
+}
+
+static void bch2_delete_dead_snapshots_work(struct work_struct *work)
+{
+	struct bch_fs *c = container_of(work, struct bch_fs, snapshot_delete_work);
+
+	bch2_delete_dead_snapshots(c);
 	percpu_ref_put(&c->writes);
 }
 
-static void bch2_delete_dead_snapshots(struct bch_fs *c)
+void bch2_delete_dead_snapshots_async(struct bch_fs *c)
 {
-	if (unlikely(!percpu_ref_tryget_live(&c->writes)))
+	if (!percpu_ref_tryget_live(&c->writes))
 		return;
 
 	if (!queue_work(system_long_wq, &c->snapshot_delete_work))
@@ -736,7 +798,14 @@ static void bch2_delete_dead_snapshots(struct bch_fs *c)
 static int bch2_delete_dead_snapshots_hook(struct btree_trans *trans,
 					   struct btree_trans_commit_hook *h)
 {
-	bch2_delete_dead_snapshots(trans->c);
+	struct bch_fs *c = trans->c;
+
+	set_bit(BCH_FS_HAVE_DELETED_SNAPSHOTS, &c->flags);
+
+	if (!test_bit(BCH_FS_FSCK_DONE, &c->flags))
+		return 0;
+
+	bch2_delete_dead_snapshots_async(c);
 	return 0;
 }
 
diff --git a/fs/bcachefs/subvolume.h b/fs/bcachefs/subvolume.h
index 28dbd0968f3d..7823040c9641 100644
--- a/fs/bcachefs/subvolume.h
+++ b/fs/bcachefs/subvolume.h
@@ -114,7 +114,9 @@ static inline int snapshot_list_add(struct bch_fs *c, snapshot_id_list *s, u32 i
 	return ret;
 }
 
-int bch2_fs_snapshots_check(struct bch_fs *);
+int bch2_fs_check_snapshots(struct bch_fs *);
+int bch2_fs_check_subvols(struct bch_fs *);
+
 void bch2_fs_snapshots_exit(struct bch_fs *);
 int bch2_fs_snapshots_start(struct bch_fs *);
 
@@ -137,6 +139,9 @@ int bch2_subvolume_get_snapshot(struct btree_trans *, u32, u32 *);
 int bch2_snapshot_node_create(struct btree_trans *, u32,
 			      u32 *, u32 *, unsigned);
 
+int bch2_delete_dead_snapshots(struct bch_fs *);
+void bch2_delete_dead_snapshots_async(struct bch_fs *);
+
 int bch2_subvolume_delete(struct btree_trans *, u32);
 int bch2_subvolume_unlink(struct btree_trans *, u32);
 int bch2_subvolume_create(struct btree_trans *, u64, u32,