Re: No-op case in ExecEvalConvertRowtype

On Fri, Apr 7, 2017 at 7:35 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> Amit Langote <Langote_Amit_f8(at)lab(dot)ntt(dot)co(dot)jp> writes:
>> ... One of
>> the earlier versions of that patch introduced a consider_typeid parameter
>> for which only ExecEvalConvertRowtype() passed true.
>
> Yeah, I was thinking of adding a flag along that line to fix this, but
> desisted after figuring out that ExecEvalConvertRowtype was the only
> candidate for using it. Ashutosh's patch had already shown that it'd
> be better to pass "false" there too, so we'd end up with no use cases
> at all.

Probably we should also add an assertion there to make sure
ExecEvalConvertRowtype never gets same input and output types. If
that's the case, we don't need the copy as well.

--
Best Wishes,
Ashutosh Bapat
EnterpriseDB Corporation
The Postgres Database Company