Re: BUG: pg_stat_statements query normalization issues with combined queries

On 13 Jan. 2017 16:35, "Kyotaro HORIGUCHI" <horiguchi.kyotaro@lab.ntt.co.jp>
wrote:

Hello,

At Thu, 12 Jan 2017 20:08:54 +0100 (CET), Fabien COELHO <coelho@cri.ensmp.fr>
wrote in <alpine.DEB.2.20.1701122004190.3788@lancre>
>
> About having a pointer to the initial string from RawStmt, Query &
> PlannedStmt:
>
> > I remembered one reason why we haven't done this: it's unclear how
> > we'd handle copying if we do it. If, say, Query contains a "char *"
> > pointer then you'd expect copyObject() to pstrdup that string, [...,
> > So] We'd need to work out a way of managing multiple Queries carrying
> > references to the same source string, and it's not clear how to do
> > that reasonably.
>
> For me it would be shared, but then it may break some memory
> management hypothesis downstream.

+1 to they have a pointer to the shared query string. But doing
that without some measure like reference counting seems
difficult..



Sounds like it'd be better as a separate change so as not to block this one.

I really like what you have done Tom, though I'm about to travel so I
haven't read it in full detail. Like Fabien I would've been certain that
it'd be rejected if I tried it, but I sure am glad you did it.