s3:registry: eliminate race condition in creating/scanning sorted subkeys

Called, from key_exists, scan_sorted_subkeys re-creates the sorted
subkeys record of the given key and then searches through it.
The race is that between creation and parsing of the sorted subkey
record, another process that stores some other subkey of the same
parent key will delete the sorted subkey record, resulting in an
WERR_BADFILE of an operation that should actually succeed.

This patch fixes the issue by wrapping the creation and parsing
into a transaction.

Michael

1 files changed, 17 insertions, 0 deletions

diff --git a/source3/registry/reg_backend_db.c b/source3/registry/reg_backend_db.c
index 43963db3c2..fe15b4e461 100644
--- a/source3/registry/reg_backend_db.c
+++ b/source3/registry/reg_backend_db.c
@@ -1368,14 +1368,31 @@ static bool scan_parent_subkeys(struct db_context *db, const char *parent,
 	if (state.scanned) {
 		result = state.found;
 	} else {
+		res = db->transaction_start(db);
+		if (res != 0) {
+			DEBUG(0, ("error starting transacion\n"));
+			goto fail;
+		}
+
 		if (!create_sorted_subkeys(path, key)) {
+			res = db->transaction_cancel(db);
+			if (res != 0) {
+				smb_panic("Failed to cancel transaction.");
+			}
 			goto fail;
 		}
+
 		res = db->parse_record(db, string_term_tdb_data(key),
 				       parent_subkey_scanner, &state);
 		if ((res == 0) && (state.scanned)) {
 			result = state.found;
 		}
+
+		res = db->transaction_commit(db);
+		if (res != 0) {
+			DEBUG(0, ("error committing transaction\n"));
+			result = false;
+		}
 	}
 
  fail: