Re: Anonymous code blocks

Dimitri Fontaine napsal(a):
> Hi,
>
> Dimitri Fontaine <dfontaine(at)hi-media(dot)com> writes:
>
>> Patch applies cleanly and build cleanly too, basic examples are working
>> fine.
>>
>
> I've been reading through the code and am going to mark it as ready for
> commiter, as only remarks I have are probably because I do not know
> enough about PostgreSQL internals, and the one I missed are in the same
> category.
>
> The patch is easy to read and all it does looks straightforward, even
> for me :)
>
> Here we go:
>
> *** a/src/backend/tcop/utility.c
> --- b/src/backend/tcop/utility.c
> ...
> *************** UtilityReturnsTuples(Node *parsetree)
> *** 1147,1155 ****
> ...
> - case T_ExplainStmt:
> - return true;
> -
>
> Is this not a oversight in the final patch?
>

It is. I attached patch which does not have this part.

>
> + /* This is short-lived, so needn't allocate in function's cxt */
> + plpgsql_Datums = palloc(sizeof(PLpgSQL_datum *) * datums_alloc);
> ...
> + compile_tmp_cxt = MemoryContextSwitchTo(func_cxt);
>
> I wonder why not having the datums into the func_cxt too.
>

Actually I think we might not need that function memory context for
anonymous code blocks at all since we don't cache compiled functions.
But I am not sure so I basically copied it from standard function
compiler to be on safe side. I am sure commiter will comment on this :)

--
Regards
Petr Jelinek (PJMODOS)