Re: 12.1 not useable: clientlib fails after a dozen queries (GSSAPI ?)

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Peter <pmc(at)citylink(dot)dinoex(dot)sub(dot)org>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org, Stephen Frost <sfrost(at)snowman(dot)net>
Subject: Re: 12.1 not useable: clientlib fails after a dozen queries (GSSAPI ?)
Date: 2020-01-11 03:51:50
Message-ID: 26433.1578714710@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general pgsql-hackers

Here's a draft patch that cleans up all the logic errors I could find.

I also expanded the previous patch for the kerberos test so that it
verifies that we can upload a nontrivial amount of data to the server,
as well as download.

I also spent a fair amount of effort on removing cosmetic differences
between the comparable logic in be-secure-gssapi.c and fe-secure-gssapi.c,
such that the two files can now be diff'd to confirm that be_gssapi_write
and be_gssapi_read implement identical logic to pg_GSS_write and
pg_GSS_read. (They did not, before :-(.)

This does not deal with the problem that libpq shouldn't be using
static data space for this purpose. It seems reasonable to me to
leave that for a separate patch.

This passes tests for me, on my FreeBSD build with lo0 mtu = 1500.
It wouldn't hurt to get some more mileage on it though. Peter,
I didn't follow how to set up the "packet radio speed" environment
that you mentioned, but if you could beat on this with that setup
it would surely be useful.

regards, tom lane

Attachment Content-Type Size
fix-assorted-gssapi-bugs-1.patch text/x-diff 34.4 KB

In response to

Responses

Browse pgsql-general by date

  From Date Subject
Next Message Jernigan, Kevin 2020-01-11 13:01:14 Re: Upgrade PostgreSQL 9.6 to 10.6
Previous Message Adrian Klaver 2020-01-11 03:20:59 Re: pg_repack in cluster

Browse pgsql-hackers by date

  From Date Subject
Next Message Masahiko Sawada 2020-01-11 03:53:05 Re: [HACKERS] Block level parallel vacuum
Previous Message Peter Geoghegan 2020-01-11 02:49:33 Re: Amcheck: do rightlink verification with lock coupling