[HACKERS] some question about deadlock

[ipig]

Hi,   Below is the notes from postgresql-8.1.3/src/backend/storage/lmgr/README:     Lock acquisition (routines LockAcquire and ProcSleep) follows these rules:   1. A lock request is granted immediately if it does not conflict with any existing or waiting lock request, or if the process already holds an instance of the same lock type (eg, there's no penalty to acquire a read lock twice).  Note that a process never conflicts with itself, eg one can obtain read lock when one already holds exclusive lock.   2. Otherwise the process joins the lock's wait queue.  Normally it will be added to the end of the queue, but there is an exception: if the process already holds locks on this same lockable object that conflict with the request of any pending waiter, then the process will be inserted in the wait queue just ahead of the first such waiter.  (If we did not make this check, the deadlock detection code would adjust the queue order to resolve the conflict, but it's relatively cheap to make the check in ProcSleep and avoid a deadlock timeout delay in this case.)  Note special case when inserting before the end of the queue: if the process's request does not conflict with any existing lock nor any waiting request before its insertion point, then go ahead and grant the lock without waiting.     I am confused with that exception(in bold), could some one give me an example?     Best regards.