Re: should INSERT SELECT use a BulkInsertState?

On Mon, Feb 22, 2021 at 02:25:22AM +0000, houzj(dot)fnst(at)fujitsu(dot)com wrote:
> > > Yes, you can see that I've copied the checks from copy.
> > > Like copy, some checks are done once, in ExecInitModifyTable, outside
> > > of the ExecModifyTable "loop".
> > >
> > > This squishes some commits together.
> > > And uses bistate for ON CONFLICT.
> > > And attempts to use memory context for tuple size.
> >
> > Rebased on 9dc718bdf2b1a574481a45624d42b674332e2903
> >
> > I guess my patch should/may be subsumed by this other one - I'm fine with
> > that.
> > https://commitfest.postgresql.org/31/2871/
> >
> > Note that my interest here is just in bistate, to avoid leaving behind many dirty
> > buffers, not improved performance of COPY.
>
> I am very interested in this patch, and I plan to do some experiments with the patch.
> Can you please rebase the patch because it seems can not applied to the master now.

Thanks for your interest.

I was sitting on a rebased version since the bulk FDW patch will cause
conflicts, and since this should maybe be built on top of the table-am patch
(2871). Have fun :)

--
Justin