References: <AB5978B2-1772-4FEE-A245-74C91704ECB0@amazon.com>
 <87ftjifoql.fsf@ars-thinkpad>
 <20191024213157.7pm6niybfxgpvmgg@alap3.anarazel.de>
 <87eez1fh48.fsf@ars-thinkpad>
 <CAA4eK1L=MDbmGu5-+BmY7Svc07jr+ZabiH8C_qo3RSc8pgUpDQ@mail.gmail.com>
 <8736bs81sx.fsf@ars-thinkpad>
 <CAA4eK1LdNmrib1jub8b=KvYUrzXW0VT4P3MVPMyiMfMY3K64dA@mail.gmail.com>
 <871rrb942q.fsf@ars-thinkpad>
 <CAA4eK1LYzrZ_+8VhD_N_dsQwjxA9t+AyGKT-Wjnc8S7jCwAcBw@mail.gmail.com>
 <87zhdx76d5.fsf@ars-thinkpad>
 <CAA4eK1Jdh0zab=+D91MkFPbevzyjFCZsMwRsQxJh4F9+m_vCRA@mail.gmail.com>
 <CAA4eK1JnSKkNdgLAHWY+YCE_3Li454So_thyTyNJa_G3hSVscA@mail.gmail.com>
 <8736bjoiax.fsf@ars-thinkpad>
 <CAA4eK1Kcsib6UG7zFPrL-h2fvnByfKWxq8Xvzg=2hxUebYwt=g@mail.gmail.com>
 <87wo8ulhjl.fsf@ars-thinkpad>
 <CAA4eK1+zGWCTsPqpL0fON6xW8BWKp_JtHr72hio3RDLJnv0G-A@mail.gmail.com>
User-agent: mu4e 1.2.0; emacs 26.0.50
From: Arseny Sher <a.sher@postgrespro.ru>
To: Amit Kapila <amit.kapila16@gmail.com>
Cc: Andres Freund <andres@anarazel.de>, "Hsu\, John" <hsuchen@amazon.com>,
 "pgsql-bugs\@lists.postgresql.org" <pgsql-bugs@lists.postgresql.org>
Subject: Re: ERROR: subtransaction logged without previous top-level txn
 record
In-reply-to: 
 <CAA4eK1+zGWCTsPqpL0fON6xW8BWKp_JtHr72hio3RDLJnv0G-A@mail.gmail.com>
Date: Tue, 11 Feb 2020 07:32:22 +0300
Message-ID: <87lfp9u4kp.fsf@ars-thinkpad>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk


Amit Kapila <amit.kapila16@gmail.com> writes:

>> Attached is raw version of isolation test triggering false
>> 'subtransaction logged without...' (case (1)).
>>
>
> This didn't reproduce the desired error for me (tried without a
> patch).  I think you need to add two more steps ("s2_checkpoint"
> "s2_get_changes") at the end of the test to set the restart_lsn at the
> appropriate location.

That's weird, it reliably fails with expected error for me. There are
already two s2_checkpoint's: first establishes potential (broken)
restart_lsn (serializes snapshot after first xl_xact_assignment of s0
xact, but before first record of s1 xact), the second ensures
s2_get_changes directly following it will actually advance the slot,
making that potential restart_lsn real.

I don't see how adding s2_checkpoint and s2_get_changes helps here. Do
they really provoke error in your setup? Could you check with pg_waldump
what's going on?

>
>> However, frankly I don't
>> see much value in it, so I'm dubious whether it should be included in
>> the patch.
>>
>
> I think this will surely test some part of the system which was not
> tested before, mainly having some subxacts without top-xact getting
> decoded even though we don't need to send such a transaction.  Can you
> prepare a complete patch (for
> Stop-demanding-that-top-xact-must-be-seen-before-sub) having this test
> as part of it?

Ok, will do.


-- cheers, arseny