On 7/20/07, Heikki Linnakangas <[EMAIL PROTECTED]> wrote:
Pavan Deolasee wrote: > Are we leaking memory in vac_update_relstats ? > > /* Fetch a copy of the tuple to scribble on */ > ctup = SearchSysCacheCopy(RELOID, > ObjectIdGetDatum(relid), > 0, 0, 0); > > This copy is not subsequently freed in the function. It's palloc'd in the current memory context, so it's not serious. It'll be freed at the end of the transaction, if not before that. That's the beauty of memory contexts; no need to worry about small allocations like that.
Right. But may be for code completeness, we should add that missing heap_freetuple. Thanks, Pavan -- Pavan Deolasee EnterpriseDB http://www.enterprisedb.com
