Re: RFC 9266: Channel Bindings for TLS 1.3 support - Mailing list pgsql-bugs

From Michael Paquier
Subject Re: RFC 9266: Channel Bindings for TLS 1.3 support
Date
Msg-id YuNzz767HkB79sys@paquier.xyz
Whole thread Raw
In response to RFC 9266: Channel Bindings for TLS 1.3 support  (* Neustradamus * <neustradamus@hotmail.com>)
Responses Re: RFC 9266: Channel Bindings for TLS 1.3 support  (Jacob Champion <jchampion@timescale.com>)
List pgsql-bugs
Hi,

On Thu, Jul 28, 2022 at 08:33:50PM +0000, * Neustradamus * wrote:
> Can you add the support of RFC 9266: Channel Bindings for TLS 1.3?
> - https://datatracker.ietf.org/doc/html/rfc9266
>
> Little details, to know easily:
> - tls-unique for TLS =< 1.2

tls-unique is not planned, as we have already tls-server-end-point for
TLS1.2 and Postgres requires a certificate, anyway.

> - tls-exporter for TLS = 1.3
>
> It is linked to:
> - https://github.com/postgres/postgres/search?q=tls-unique

So, tls-exporter has been made an official thing, finally.  I was
wondering when this was going to happen.  Jacob Champion has given me
a patch to support that, based on OpenSSL's SSL_export_keying_material()
to do the job.  The base integration is not complicated, but I still
need to think a bit more about it when it comes to the min/max TLS
protocols we allow in libpq, for example, and polish the whole with
tests.  We don't force any failures depending on the other connection
parameters for tls-server-end-point, so I suspect that we should be
fine with keeping things at their simplest.

I should be able to get something sent to the mailing lists for the
commit fest of September, so as we could have this feature in v16~.
--
Michael

Attachment

pgsql-bugs by date:

Previous
From: Tom Lane
Date:
Subject: Re: Statistics updates is delayed when using `commit and chain`
Next
From: Ajin Cherian
Date:
Subject: Re: Excessive number of replication slots for 12->14 logical replication