Gregory Stark <stark@enterprisedb.com> writes:
> I think we should go ahead and print a LOG message in the case where a
> deadlock is detected.

After chewing on it some more, I liked this patch better ... but I found
a couple other issues, particularly with the soft-deadlock-recovery
messaging sequence. Modified version committed.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend