|From:||Michael Paquier <michael(at)paquier(dot)xyz>|
|To:||Daniel Gustafsson <daniel(at)yesql(dot)se>|
|Cc:||Heikki Linnakangas <hlinnaka(at)iki(dot)fi>, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, Postgres hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>|
|Subject:||Re: scram-sha-256 broken with FIPS and OpenSSL 1.0.2|
|Views:||Raw Message | Whole Thread | Download mbox | Resend email|
On Mon, Nov 30, 2020 at 02:29:29PM +0100, Daniel Gustafsson wrote:
> Yeah, that's along the lines of what I was thinking of.
Hmm. I have looked at that, and thought first about having directly a
reference to the resowner directly in pg_cryptohash_ctx, but that's
not a good plan for two reasons:
- common/cryptohash.h would get knowledge of that, requiring bundling
in a bunch of dependencies.
- There is no need for that in the non-OpenSSL case.
So, instead, I have been thinking about using an extra context layer
only for cryptohash_openssl.c with a structure saved as
pg_cryptohash_context->data that stores the information about
EVP_MD_CTX* and the resource owner. Then, I was thinking about
storing directly pg_cryptohash_ctx in the resowner EVP array and just
call pg_cryptohash_free() from resowner.c without the need of an
extra routine. I have not tested this idea but that should work.
What's your take?
In parallel, I have spent more time today polishing and reviewing 0001
(indented, adjusted a couple of areas and added also brackets and
extra comments as you suggested) and tested it on Linux and Windows,
with and without OpenSSL down to 1.0.1, the oldest version supported
on HEAD. So I'd like to apply the attached first and sort out the
resowner stuff in a next step.
|Next Messagefirstname.lastname@example.org||2020-12-01 05:39:59||RE: [POC] Fast COPY FROM command for the table with foreign partitions|
|Previous Message||Fujii Masao||2020-12-01 05:01:41||Re: Add statistics to pg_stat_wal view for wal related parameter tuning|