Horton

Re: WIP: Avoid creation of the free space map for small tables

Amit Kapila <amit.kapila16@gmail.com>

View thread View on postgr.es
From: Amit Kapila <amit.kapila16@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: John Naylor <john.naylor@2ndquadrant.com>, Mithun Cy <mithun.cy@enterprisedb.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2019-01-17T03:08:00Z
Lists: pgsql-hackers 
On Wed, Jan 16, 2019 at 11:25 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>
> John Naylor <john.naylor@2ndquadrant.com> writes:
> > It just occured to me that the style FSM_LOCAL_MAP_EXISTS seems more
> > common for macros that refer to constants, and FSMLocalMapExists for
> > expressions, but I've only seen a small amount of the code base. Do we
> > have a style preference here, or is it more a matter of matching the
> > surrounding code?
>

I am fine with the style (FSMLocalMapExists) you are suggesting, but
see the similar macros in nearby code like:

#define FSM_TREE_DEPTH ((SlotsPerFSMPage >= 1626) ? 3 : 4)

I think the above is not an exact match.  So, I have looked around and
found few other macros which serve a somewhat similar purpose, see
below:

#define ATT_IS_PACKABLE(att) \
((att)->attlen == -1 && (att)->attstorage != 'p')

#define VARLENA_ATT_IS_PACKABLE(att) \
((att)->attstorage != 'p')

#define CHECK_REL_PROCEDURE(pname)

#define SPTEST(f, x, y) \
DatumGetBool(DirectFunctionCall2(f, PointPGetDatum(x), PointPGetDatum(y)))

> I believe there's a pretty longstanding tradition in C coding to use
> all-caps names for macros representing constants.  Some people think
> that goes for all macros period, but I'm not on board with that for
> function-like macros.
>
> Different parts of the PG code base make different choices between
> camel-case and underscore-separation for multiword function names.
> For that, I'd say match the style of nearby code.
>

Yes, that is what we normally do.  However, in some cases, we might
need to refer to other places as well which I think is the case here.

-- 
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com

Commits

  1. Improve code comments in b0eaa4c51b.

  2. During pg_upgrade, conditionally skip transfer of FSMs.

  3. Add more tests for FSM.

  4. Doc: Update the documentation for FSM behavior for small tables.

  5. Make FSM test portable.

  6. Avoid creation of the free space map for small heap relations, take 2.

  7. Move page initialization from RelationAddExtraBlocks() to use, take 2.

  8. Avoid creation of the free space map for small heap relations.

  9. In bootstrap mode, don't allow the creation of files if they don't already