Mike Adelson <mike.adelson314@gmail.com> writes:
> Glad to have the race condition confirmed. In my view the “real” fix
> here would be to offer an advisory locking function that accepts a
> timeout.
How would that remove the low-level race condition?
regards, tom lane