Date: Fri, 10 Jan 2020 22:03:37 +0100
From: Peter <pmc@citylink.dinoex.sub.org>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers@lists.postgresql.org, Stephen Frost <sfrost@snowman.net>
Subject: Re: 12.1 not useable: clientlib fails after a dozen queries (GSSAPI
 ?)
Message-ID: <20200110210337.GA9580@gate.oper.dinoex.org>
References: <20200109181822.GA74698@gate.oper.dinoex.org>
 <7042.1578595681@sss.pgh.pa.us>
 <20200109205319.GA85142@gate.oper.dinoex.org>
 <15232.1578605504@sss.pgh.pa.us>
 <20200110022342.GA92515@gate.oper.dinoex.org>
 <12348.1578679149@sss.pgh.pa.us>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <12348.1578679149@sss.pgh.pa.us>
User-Agent: Mutt/1.12.2 (2019-09-21)
Precedence: bulk

On Fri, Jan 10, 2020 at 12:59:09PM -0500, Tom Lane wrote:
! [ redirecting to -hackers ]

! I modified the kerberos test so that it tries a query with a less
! negligible amount of data, and what I find is:
! 
! * passes on Fedora 30, with either default or 1500 mtu
! * passes on FreeBSD 12.0 with default mtu
! * FAILS on FreeBSD 12.0 with mtu = 1500

So, it is not related to only VIMAGE @ R11.3, and -more important to
me- it is not only happening in my kitchen. Thank You very much :)

! OTOH, I also find that there's some hysteresis in the behavior:
! once it's failed, reverting the mtu to the default setting doesn't
! necessarily make subsequent runs pass.  It's really hard to explain
! that behavior if it's our bug.

That's affirmative. Made me go astray a few times when trying to
isolate it.

On Fri, Jan 10, 2020 at 02:25:22PM -0500, Tom Lane wrote:
! Ah, I have it: whoever wrote pg_GSS_read() failed to pay attention
! to the fact that setting errno is a critical part of its API.
! Sometimes it returns -1 while leaving errno in a state that causes

Wow, that's fast. My probability-guess this morning was either some
hard-coded 8192-byte buffer, or something taking an [EWOULDBLOCK] for
OK. Then I decided to not look into the code, as You will be much
faster anyway, and there are other pieces of software where I do
not have such a competent peer to talk to...

Anyway, thanks a lot!
PMc