Re: servers/slapd/bind.c, development tree

["Kurt D. Zeilenga" <Kurt@OpenLDAP.org>]

Sorry for getting back to you sooner... but I've been busy.

I think this assert is fine... but I could be wrong.  If your hitting it,
I suggest posting to -devel with details.

        Kurt


At 08:14 PM 6/20/00 +0000, Mei-Hui Su wrote:

>Hi, 
>
> In servers/slapd/bind.c, for the case for doing sasl_bind, shouldn't the last
>assertion be "== NULL" instead of "!= NULL" because around #360 of sasl.c
>
>        if ( sc != SASL_CONTINUE && conn->c_sasl_bind_context != NULL ) {
>                sasl_dispose( &conn->c_sasl_bind_context );
>                conn->c_sasl_bind_context = NULL;
>        }
>
>
>servers/slapd/bind.c:
>                ldap_pvt_thread_mutex_unlock( &conn->c_mutex );
>
>                edn = NULL;
>                rc = sasl_bind( conn, op, dn, ndn, mech, &cred, &edn );
>
>                if( rc == LDAP_SUCCESS ) {
>                        ldap_pvt_thread_mutex_lock( &conn->c_mutex );
>#ifdef HAVE_CYRUS_SASL
>                        assert( conn->c_sasl_bind_context == NULL );
>#endif
>                        conn->c_dn = edn;
>                        if(dn!=NULL) {
>                          conn->c_cdn = ch_malloc( sizeof( "dn" ) );
>                          strcpy( conn->c_cdn, dn );
>                        }
>                        ldap_pvt_thread_mutex_unlock( &conn->c_mutex );
>
>                } else if ( rc == LDAP_SASL_BIND_IN_PROGRESS ) {
>#ifdef HAVE_CYRUS_SASL
>                        assert( conn->c_sasl_bind_context != NULL );
>#endif
>                        conn->c_sasl_bind_mech = mech;
>                        mech = NULL;
>
>#ifdef HAVE_CYRUS_SASL
>                } else {
>                        assert( conn->c_sasl_bind_context != NULL );  <<<----
>#endif
>                }
>
>                goto cleanup;
>
>mei