Re: reducing the footprint of ScanKeyword (was Re: Large writable variables)

Hi,

On 2019-01-09 14:44:24 -0500, Tom Lane wrote:
> I wrote:
> > Also, I fail to understand why fmgr_builtin_oid_index has 10000 entries
> > anyway. We could easily have fmgrtab.c expose the last actually assigned
> > builtin function OID (presently 6121) and make the index array only
> > that big, which just about eliminates the space advantage completely.
>
> Concretely, like the attached.

Seems like a good improvement.

> We could make the index table still smaller if we wanted to reassign
> a couple dozen high-numbered functions down to lower OIDs, but I dunno
> if it's worth the trouble. It certainly isn't from a performance
> standpoint, because those unused entry ranges will never be touched
> in normal usage; but it'd make the server executable a couple KB smaller.

Probably indeed not worth it. I'm not 100% convinced on the performance
POV, but in contrast to the earlier binary search either approach is
fast enough that it probably hard to measure any difference.

> diff --git a/src/backend/utils/fmgr/fmgindex b41649f..506eeef 100644
> --- a/src/backend/utils/fmgr/fmgr.c
> +++ b/src/backend/utils/fmgr/fmgr.c
> @@ -75,12 +75,12 @@ fmgr_isbuiltin(Oid id)
> uint16 index;
>
> /* fast lookup only possible if original oid still assigned */
> - if (id >= FirstGenbkiObjectId)
> + if (id > fmgr_last_builtin_oid)
> return NULL;

An extern reference here will make the code a bit less efficient, but
it's probably not worth generating a header with a define for it
instead...

Greetings,

Andres Freund