| From: | Michael Paquier <michael(at)paquier(dot)xyz> |
|---|---|
| To: | Mark Atwood <mark(at)reviewcommit(dot)com> |
| Cc: | pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: [PATCH v1 0/3] Route crypto through the OpenSSL 3 provider API |
| Date: | 2026-08-05 08:26:13 |
| Message-ID: | anLzpT8OqzwPgNfG@paquier.xyz |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Tue, Aug 04, 2026 at 05:48:02PM -0700, Mark Atwood wrote:
> When built against OpenSSL 3.0 or newer, PostgreSQL's OpenSSL crypto backend
> still uses pre-3.0 interfaces that do not go through the provider framework,
> so a loaded provider's implementation is not used:
>
> - HMAC uses HMAC_CTX / HMAC_Init_ex, deprecated since 3.0;
> - cryptohash and channel binding initialize the digest with the implicit
> static MDs (EVP_sha256() etc.).
Please note patches should be attached to the emails sent to the
community mailing lists. We do not use commands like `git send-mail`.
FYI, I am interested in what you are doing here for the HMAC and
channel binding parts, at least, having committed the code we have in
the tree and that you are updating here. Just make sure to begin one
thread for each patch proposed, with the patch attached, so as we are
able to discuss and review each item separately. Generating the
patches with format-patch is a common practice.
Please see also:
https://wiki.postgresql.org/wiki/Submitting_a_Patch#Patch_submission
Thanks,
--
Michael
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Dinesh Salve | 2026-08-05 08:35:32 | Re: Add test_bufmgr - algorithm-agnostic tests for the buffer manager |
| Previous Message | Akshay Joshi | 2026-08-05 08:21:52 | Re: [PATCH] Add pg_get_table_ddl() to reconstruct CREATE TABLE statements |