Re: [bug fix] Savepoint-related statements terminates connection

On 7 September 2017 at 11:07, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> I wrote:
>> Yeah, it seems like we have now made this behavior official enough that
>> it's time to document it better. My thought is to create a new subsection
>> in the FE/BE Protocol chapter that explains how multi-statement Query
>> messages are handled, and then to link to that from appropriate places
>> elsewhere. If anyone thinks the reference section would be better put
>> somewhere else than Protocol, please say where.
>
> I've pushed up an attempt at this:
>
> https://git.postgresql.org/gitweb/?p=postgresql.git;a=commitdiff;h=b976499480bdbab6d69a11e47991febe53865adc
>
> Feel free to suggest improvements.

Not so much an improvement as a follow-on thought:

All of this applies to simple queries.

At present we restrict using multi-statement requests in extended
protocol, saying that we don't allow it because of a protocol
restriction. The precise restriction is that we can't return more than
one reply. The restriction is implemented via this test
if (list_length(parsetree_list) > 1)
ereport(ERROR,
(errcode(ERRCODE_SYNTAX_ERROR),
errmsg("cannot insert multiple commands into a prepared
statement")));
at line 1277 of exec_parse_message()
which is actually more restrictive than it needs to be.

I would like to relax the restriction to allow this specific use case...
SET work_mem = X; SET max_parallel_workers = 4; SELECT ...
so we still have only one command (the last select), yet we have
multiple GUC settings beforehand.

Any reason to disallow that?

--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services