Best shown with an example
Session 1:
==========
SQL> create table PAR ( p int primary key, d date);
Table created.
SQL>
SQL> create table CHD ( p int primary key references PAR(p), d date);
Table created.
SQL>
SQL> insert into PAR values (1,sysdate);
1 row created.
SQL> insert into PAR values (2,sysdate);
1 row created.
SQL>
SQL> insert into CHD values (1,sysdate);
1 row created.
SQL>
SQL> commit;
Commit complete.
SQL> update CHD set d = sysdate + 1;
1 row updated.
Session 2
=========
SQL> lock table PAR in exclusive mode nowait;
Table(s) Locked.
Session 1:
==========
SQL> update CHD set p = 2;
[blocked, ie, waits]
So an exclusive lock on the parent stops the primary key update on CHD
But less extreme (and more likely usage) will be fine, eg (picking up from the point of Session 2 above)
Session 2
=========
SQL> update PAR set d = d + 1;
2 rows updated.
Session 1
=========
SQL> update CHD set p = 2;
1 row updated.
Hope this helps