Re: Exposing the lock manager's WaitForLockers() to SQL

On Sun, Sep 3, 2023 at 11:16 PM Will Mortensen <will@extrahop.com> wrote:
> I realized that for our use case, we'd ideally wait for holders of
> RowExclusiveLock only, and not e.g. VACUUM holding
> ShareUpdateExclusiveLock. Waiting for lockers in a specific mode seems
> possible by generalizing/duplicating WaitForLockersMultiple() and
> GetLockConflicts(), but I'd love to have a sanity check before
> attempting that. Also, I imagine those semantics might be too
> different to make sense as part of the LOCK command.

Well I attempted it. :-) Here is a new series that refactors
GetLockConflicts(), generalizes WaitForLockersMultiple(), and adds a
new WAIT FOR LOCKERS command.

I first tried extending LOCK further, but the code became somewhat
unwieldy and the syntax became very confusing. I also thought again
about making new pg_foo() functions, but that would seemingly make it
harder to share code with LOCK, and sharing syntax (to the extent it
makes sense) feels very natural. Also, a new SQL command provides
plenty of doc space. :-) (I'll see about adding more examples later.)

I'll try to edit the title of the CF entry accordingly. Still looking
forward to any feedback. :-)