Cache invalidation was changed a little

Поиск
Список
Период
Сортировка
От Hiroshi Inoue
Тема Cache invalidation was changed a little
Дата
Msg-id 000501bf5b35$4d6f5740$2801007e@tpf.co.jp
обсуждение исходный текст
Список pgsql-hackers
Hi

I have just committed changes to cache invalidation stuff.
Maybe this would fix the following TODO.
* elog() flushes cache, try invalidating just entries from current xact, perhaps using invalidation cache 

1) In case of abort,catalog cache and relation cache will   be invalidated for system tuples marked by Relation-
Mark4RollbackHeapTuple().Both heap_insert() and   heap_update() call RelationMark4RollbackHeapTuple(). 
 

2) CommandCounterIncrement() calls AtCommit_LocalCache()   instead of AtCommit_Cache(). Registration of cache
invalidationfor other backends was postponed until commit.
 

3) The new function ImmediateSharedRelationCacheInvalidate()   is called from smgrunlink()/smgrtruncate() in order to
register  relation cache invalidation immediately.
 

Regards.

Hiroshi Inoue
Inoue@tpf.co.jp


В списке pgsql-hackers по дате отправления:

Предыдущее
От: Tom Lane
Дата:
Сообщение: Re: [HACKERS] Number of index fields configurable
Следующее
От: Stephen Birch
Дата:
Сообщение: Re: [HACKERS] Re:HEAP_MOVED_IN during vacuum - test case