Re: Consider pipeline implicit transaction as a transaction block

On Wed, Nov 27, 2024 at 04:20:10PM -0500, Bruce Momjian wrote:
> Yeah, I was surprised too, even though the author was clear they wanted
> to backpatch. I couldn't figure out why it was being backpatched, so I
> figured I was missing something.

The set of inconsistencies of how we decide if one command or the
other should behave differently depending on the use of a pipeline,
on the transaction block state, on the code paths involved and on
MyXactFlags in this case qualified as good enough to me for a
backpatch, in the same lines as 20432f873140 & friends. Also worth
noting is the lack of regression tests back then, some could have been
introduced through pgbench's meta-commands as Anthonin has done here
to provide some combination checks even if the error messages
generated cannot be directly looked at. pgbench was mentioned on the
original thread leading to this commit and the result not include
anything. Anyway, what's done is done..

I don't mind being more careful here based on your concerns, so I'll
go remove that in the stable branches.
--
Michael