On Wed, Mar 26, 2014 at 5:08 AM, Heikki Linnakangas
<hlinnakangas@vmware.com> wrote:
> Oh, no, there's no need to do it while holding WALInsertLock. It's quite
> trivial, actually, see attached. So it's not difficult to fix this if we
> want to.
Well is there any reason not to just commit that patch? I mean, it
seems odd to rip this out if the fix is straightforward and already
written.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company