Re: TupleTableSlot abstraction

I have only done the below two changes yet. After doing that and
rebasing with latest master, in the regression I got crashes, and I
suspect the reason being that I have used Virtual tuple slot for the
destination slot of execute_attr_map_slot(). I am analyzing it. I am
anyway attaching the patches (v12) to give you an idea of how I have
handled the below two items.

On Wed, 26 Sep 2018 at 05:09, Andres Freund <andres@anarazel.de> wrote:
>
> On 2018-09-04 18:35:34 +0530, Amit Khandekar wrote:
> > The attached v11 tar has the above set of changes.
>
> - I've pushed 0003 - although that commit seems to have included a lot
>   of things unrelated to the commit message. I guess two patches have
>   accidentally been merged?  Could you split out the part of the changes
>   that was mis-squashed?

Yeah, it indeed looks like it had unrelated things, mostly the changes
that moved the slot attribute functions into execTuples.c . I have
included this change as the very first patch in the patch series.


>> From 280eac5c6758061d0a46b7ef9dd324e9a23226b5 Mon Sep 17 00:00:00 2001
>> From: Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>
>> Date: Fri, 31 Aug 2018 10:53:42 +0530
>> Subject: [PATCH 08/14] Change tuple table slot creation routines to suite
>>  tuple table slot abstraction
>>
>> This change requires ExecInitResultTupleSlotTL, ExecInitScanTupleSlot,
>> ExecCreateScanSlotFromOuterPlan, ExecInitNullTupleSlot,
>> ExecInitExtraTupleSlot, MakeTupleTableSlot, ExecAllocTableSlot to
>> accept TupleTableSlotType as a new parameter. Change all their calls.
>>
>> Ashutosh Bapat and Andres Freund
>>
>> This by itself won't compile. Neither the tuple table slot abstraction
>> patch would compile and work without this change. Both of those need
>> to be committed together.
>
> I don't like this kind of split - all commits should individually
> compile. I think we should instead introduce dummy / empty structs for
> &TTSOpsHeapTuple etc, and add the parameters necessary to pass them
> through. And then move this patch to *before* the "core" abstract slot
> patch.  That way every commit, but the super verbose stuff is still
> split out.
>

Done. Moved this patch before the core one. In this patch, just
declared the global variables of type TupleTableSlotOps, without
initializing them.

I have tried to make sure individual patches compile successfully by
shuffling around the changes to the right patch, but there is one
particular patch that still gives error. Will fix that later.

I will handle the other review comments in the next patch series.