Re: [PATCH 08/10] md, cipher: allow internal users to skip fast random poll

Werner Koch via Gcrypt-devel <[email protected]> Mon, 03 Aug 2026 14:47:53 +0200
Newsgroups gmane.comp.encryption.gpg.libgcrypt.devel
Message-ID <[email protected]>
--===============8968175467161535939==
Content-Type: multipart/signed; boundary="=E._Coli_Chemical_fire_Treasury_H1N1_NIJ_BLU-97_A/B_SRI_Yakima_Drug=E";
	micalg=pgp-sha512; protocol="application/pgp-signature"

--=E._Coli_Chemical_fire_Treasury_H1N1_NIJ_BLU-97_A/B_SRI_Yakima_Drug=E
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Sun,  2 Aug 2026 12:55, Jussi Kivilinna said:

> 'md_open' and 'cipher_open' call '_gcry_fast_random_poll', which takes
> pool lock, reads RDRAND and mixes CSPRNG pool. ML-KEM and ML-DSA
> route SHAKE through the md interface and open one handle per polynomial,

Sounds okay for me.  The random poll things used to be effective when
random bytes were scarce which is for most platforms not anymore the
case.


Shalom-Salam,

   Werner

=2D-=20
The pioneers of a warless world are the youth that
refuse military service.             - A. Einstein

--=E._Coli_Chemical_fire_Treasury_H1N1_NIJ_BLU-97_A/B_SRI_Yakima_Drug=E
Content-Type: application/pgp-signature; name="openpgp-digital-signature.asc"

-----BEGIN PGP SIGNATURE-----

iJ8EARYKAEcWIQSHd0YfKgdOvEgNNZQZzByeCFsQegUCanCN+RsUgAAAAAAEAA5t
YW51MiwyLjUrMS4xMiwyLDINHHdrQGdudXBnLm9yZwAKCRAZzByeCFsQeh4qAQCr
hfVVG3f1MflrmsFy6GAOXWIcTz15eyQsj04XkJPcWgEA91kV3QWes8HUU9KQ0rVH
gUAS4cnYtYkBwB/BUzUPzgw=
=1kZY
-----END PGP SIGNATURE-----
--=E._Coli_Chemical_fire_Treasury_H1N1_NIJ_BLU-97_A/B_SRI_Yakima_Drug=E--



--===============8968175467161535939==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Gcrypt-devel mailing list
[email protected]
https://lists.gnupg.org/mailman/listinfo/gcrypt-devel

--===============8968175467161535939==--