bcachefs: Flag btrees with missing data

We need this to know when we should attempt to reconstruct the snapshots
btree

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

fs/bcachefs/bcachefs.h

@@ -797,6 +797,7 @@ struct bch_fs {
 		u64		features;
 		u64		compat;
 		unsigned long	errors_silent[BITS_TO_LONGS(BCH_SB_ERR_MAX)];
+		u64		btrees_lost_data;
 	}			sb;
 
 

fs/bcachefs/bcachefs_format.h

@@ -818,6 +818,7 @@ struct bch_sb_field_ext {
 	struct bch_sb_field	field;
 	__le64			recovery_passes_required[2];
 	__le64			errors_silent[8];
+	__le64			btrees_lost_data;
 };
 
 struct bch_sb_field_downgrade_entry {

fs/bcachefs/btree_io.c

@@ -1264,10 +1264,12 @@ out:
 	return retry_read;
 fsck_err:
 	if (ret == -BCH_ERR_btree_node_read_err_want_retry ||
-	    ret == -BCH_ERR_btree_node_read_err_must_retry)
+	    ret == -BCH_ERR_btree_node_read_err_must_retry) {
 		retry_read = 1;
-	else
+	} else {
 		set_btree_node_read_error(b);
+		bch2_btree_lost_data(c, b->c.btree_id);
+	}
 	goto out;
 }
 
@@ -1328,6 +1330,7 @@ start:
 
 		if (!can_retry) {
 			set_btree_node_read_error(b);
+			bch2_btree_lost_data(c, b->c.btree_id);
 			break;
 		}
 	}
@@ -1527,9 +1530,10 @@ fsck_err:
 		ret = -1;
 	}
 
-	if (ret)
+	if (ret) {
 		set_btree_node_read_error(b);
-	else if (*saw_error)
+		bch2_btree_lost_data(c, b->c.btree_id);
+	} else if (*saw_error)
 		bch2_btree_node_rewrite_async(c, b);
 
 	for (i = 0; i < ra->nr; i++) {
@@ -1665,6 +1669,7 @@ void bch2_btree_node_read(struct btree_trans *trans, struct btree *b,
 			bch2_fatal_error(c);
 
 		set_btree_node_read_error(b);
+		bch2_btree_lost_data(c, b->c.btree_id);
 		clear_btree_node_read_in_flight(b);
 		wake_up_bit(&b->flags, BTREE_NODE_read_in_flight);
 		printbuf_exit(&buf);

fs/bcachefs/recovery.c

@@ -33,6 +33,20 @@
 
 #define QSTR(n) { { { .len = strlen(n) } }, .name = n }
 
+void bch2_btree_lost_data(struct bch_fs *c, enum btree_id btree)
+{
+	u64 b = BIT_ULL(btree);
+
+	if (!(c->sb.btrees_lost_data & b)) {
+		bch_err(c, "flagging btree %s lost data", bch2_btree_id_str(btree));
+
+		mutex_lock(&c->sb_lock);
+		bch2_sb_field_get(c->disk_sb.sb, ext)->btrees_lost_data |= cpu_to_le64(b);
+		bch2_write_super(c);
+		mutex_unlock(&c->sb_lock);
+	}
+}
+
 static bool btree_id_is_alloc(enum btree_id id)
 {
 	switch (id) {
@@ -470,6 +484,7 @@ static int read_btree_roots(struct bch_fs *c)
 			}
 
 			ret = 0;
+			bch2_btree_lost_data(c, i);
 		}
 	}
 
@@ -848,6 +863,14 @@ use_clean:
 		write_sb = true;
 	}
 
+	if (c->opts.fsck &&
+	    !test_bit(BCH_FS_error, &c->flags) &&
+	    c->recovery_pass_done == BCH_RECOVERY_PASS_NR - 1 &&
+	    ext->btrees_lost_data) {
+		ext->btrees_lost_data = 0;
+		write_sb = true;
+	}
+
 	if (c->opts.fsck &&
 	    !test_bit(BCH_FS_error, &c->flags) &&
 	    !test_bit(BCH_FS_errors_not_fixed, &c->flags)) {

fs/bcachefs/recovery.h

@@ -2,6 +2,8 @@
 #ifndef _BCACHEFS_RECOVERY_H
 #define _BCACHEFS_RECOVERY_H
 
+void bch2_btree_lost_data(struct bch_fs *, enum btree_id);
+
 int bch2_journal_replay(struct bch_fs *);
 
 int bch2_fs_recovery(struct bch_fs *);

fs/bcachefs/super-io.c

@@ -527,9 +527,11 @@ static void bch2_sb_update(struct bch_fs *c)
 	memset(c->sb.errors_silent, 0, sizeof(c->sb.errors_silent));
 
 	struct bch_sb_field_ext *ext = bch2_sb_field_get(src, ext);
-	if (ext)
+	if (ext) {
 		le_bitvector_to_cpu(c->sb.errors_silent, (void *) ext->errors_silent,
 				    sizeof(c->sb.errors_silent) * 8);
+		c->sb.btrees_lost_data = le64_to_cpu(ext->btrees_lost_data);
+	}
 
 	for_each_member_device(c, ca) {
 		struct bch_member m = bch2_sb_member_get(src, ca->dev_idx);
@@ -1162,6 +1164,11 @@ static void bch2_sb_ext_to_text(struct printbuf *out, struct bch_sb *sb,
 
 		kfree(errors_silent);
 	}
+
+	prt_printf(out, "Btrees with missing data:");
+	prt_tab(out);
+	prt_bitflags(out, __bch2_btree_ids, le64_to_cpu(e->btrees_lost_data));
+	prt_newline(out);
 }
 
 static const struct bch_sb_field_ops bch_sb_field_ops_ext = {