From: Tom Lane <tgl@sss.pgh.pa.us>
To: Peter Eisentraut <peter.eisentraut@enterprisedb.com>
cc: Alvaro Herrera <alvherre@alvh.no-ip.org>,
        Fabien COELHO <coelho@cri.ensmp.fr>,
        pgsql-hackers <pgsql-hackers@postgresql.org>
Subject: Re: libpq async duplicate error results
In-reply-to: <de57761c-b99b-3435-b0a6-474c72b1149a@enterprisedb.com>
References: <202202081500.xl3iwa7doscz@alvherre.pgsql>
 <394797.1645388816@sss.pgh.pa.us> <1262938.1645658107@sss.pgh.pa.us>
 <1421785.1645723238@sss.pgh.pa.us>
 <de57761c-b99b-3435-b0a6-474c72b1149a@enterprisedb.com>
Comments: In-reply-to Peter Eisentraut <peter.eisentraut@enterprisedb.com>
	message dated "Fri, 06 May 2022 15:49:28 +0200"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2889549.1651847333.1@sss.pgh.pa.us>
Date: Fri, 06 May 2022 10:28:53 -0400
Message-ID: <2889550.1651847333@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2889550.1651847333%40sss.pgh.pa.us>
Precedence: bulk

Peter Eisentraut <peter.eisentraut@enterprisedb.com> writes:
> I took another look at this.  The output from the test program at the 
> beginning of the thread is now (master branch):
> ...
> command = SELECT 'after';
> PQsendQuery() error: FATAL:  terminating connection due to administrator 
> command
> server closed the connection unexpectedly
> 	This probably means the server terminated abnormally
> 	before or while processing the request.
> no connection to the server

> It appears the "after" query is getting the error message from the 
> previous cycle somehow.

What is happening is that this bit in PQsendQueryStart is deciding
not to clear the message buffer because conn->cmd_queue_head isn't
NULL:

    /*
     * If this is the beginning of a query cycle, reset the error state.
     * However, in pipeline mode with something already queued, the error
     * buffer belongs to that command and we shouldn't clear it.
     */
    if (newQuery && conn->cmd_queue_head == NULL)
        pqClearConnErrorState(conn);

So apparently the fault is with the pipelined-query logic.  I'm
not sure what cleanup step is missing though.

			regards, tom lane