Description of problem: Bugzilla Bug 199923: subtree search fails to find items under a db containing special characters (In reply to comment #11) > That's what Slapi_DN is supposed to do - normalize once, and be able to get > normalized and "raw" DN where needed. We should use Slapi_DN as much as > possible. We might want to schedule some time to do some much needed API > revisions to eliminate the use of char *dn in favor of Slapi_DN *dn. But not > for 7.2. In one operation, substr_dn_normalize could be called multiple times against the same dn). We should avoid it.
performance related work
Upstream ticket: https://fedorahosted.org/389/ticket/139