Обсуждение: operator family changes, sinval bug?

Поиск
Список
Период
Сортировка

operator family changes, sinval bug?

От
Jeff Janes
Дата:
When I drop the optional triconsistent support function for a gin index
operator family, the system continues to use the function to support index
queries anyway, until I close the session and restart it, at which point it
reverts to using the binary consistent function.

Is this a bug in the cache invalidation mechanism?  Or am I just expecting
too much of it?

On one occasion, I got a
ERROR:  cache lookup failed for function 16442
when trying to do the query using the gin index, but usually it just
continues to work as expected even though it is invoking a function which
no longer exists.

I see the same behavior in both 9.4.4 and in HEAD.

Cheers,

Jeff



alter operator family gin_foobar_ops using gin drop function 6 (text,text);
drop function gin_foobar_triconsistent ( internal, smallint, text, integer,
internal, internal, internal) ;

Re: operator family changes, sinval bug?

От
Tom Lane
Дата:
Jeff Janes <jeff.janes@gmail.com> writes:
> When I drop the optional triconsistent support function for a gin index
> operator family, the system continues to use the function to support index
> queries anyway, until I close the session and restart it, at which point it
> reverts to using the binary consistent function.

> Is this a bug in the cache invalidation mechanism?  Or am I just expecting
> too much of it?

There is no invalidation support for mucking with operator class contents.

            regards, tom lane