On Sat, Nov 16, 2013 at 4:36 PM, Peter Geoghegan <pg(at)heroku(dot)com> wrote:
> I'll post a revision with fixes for those. Another concern is that I
> see some race conditions that only occur when pg_stat_statements cache
> is under a lot of pressure with a lot of concurrency, that wasn't
> revealed by my original testing. I'm working on a fix for that.
Revision attached.
--
Peter Geoghegan