Thomas Munro wrote:
> While trying to produce the heap_lock_updated_tuple_rec case you
> describe (so far unsuccessfully), I discovered I could make SELECT ...
> FOR UPDATE NOWAIT block indefinitely on unpatched 9.3 in a different
> code path after heap_lock_tuple returns: in another session, UPDATE,
> COMMIT, then UPDATE, all after the first session has taken its
> snapshot but before it tries to lock a given row. The code in
> EvalPlanQualFetch (reached from the HeapTupleUpdated case in
> ExecLockRow) finishes up waiting for the uncommitted transaction.
Interesting -- perhaps this helps explain the deadlock issue reported in
http://www.postgresql.org/message-id/CAKrjmhdN+GhAjNwqfHsOtGp+7YN27zR79m99RcAJMNazt5NJrA@mail.gmail.com
> I think I see how to teach EvalPlanQualFetch how to handle wait
> policies: for NOWAIT it should ereport (fixing a pre-existing bug
> (?)), and I guess it should handle SKIP LOCKED by returning NULL,
> similarly to the way it handles deleted rows, and of course in all
> cases passing the wait policy forward to heap_lock_tuple, which it
> eventually calls.
>
> Still looking at heap_lock_updated_tuple.
>
> The difficulty of course will be testing all these racy cases reproducibly...
Does this help?
http://www.postgresql.org/message-id/51FB4305.3070600@2ndquadrant.com
The useful trick there is forcing a query to get its snapshot and then
go to sleep before actually doing anything, by way of an advisory lock.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services
