linux-kernel - [BUG] Non working HWRNG on AMD Ryzen 5 5500GT

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-Id: <D3GJCRNY4KDK.3SPJB5WP8Z7DK@cknow.org>
Date: Thu, 15 Aug 2024 15:56:26 +0200
From: "Diederik de Haas" <didi.debian@...ow.org>
To: "Tom Lendacky" <thomas.lendacky@....com>, "John Allen"
 <john.allen@....com>
Cc: "Herbert Xu" <herbert@...dor.apana.org.au>, "David S. Miller"
 <davem@...emloft.net>, <linux-crypto@...r.kernel.org>,
 <linux-kernel@...r.kernel.org>, "Diederik de Haas" <didi.debian@...ow.org>
Subject: [BUG] Non working HWRNG on AMD Ryzen 5 5500GT

Hi,

I recently bought an Asus ROG STRIX B550-F GAMING MB with an
AMD Ryzen 5 5500GT CPU (and installed the latest BIOS: 3607).
I'm running Debian Testing/Sid on it with kernel 6.9 and now 6.10
and it seems to work great.
I've been doing some (unrelated) tests with `rngtest` from the
`rng-tools5` package and wondered how it would fare on my AMD CPU.

And I found out it doesn't work at all!
But on another system I have (Asus ROG CROSSHAIR VII HERO MB +
AMD Ryzen 1800X CPU) it works absolutely fine.

# dmesg | grep ccp
[    5.399853] ccp 0000:07:00.2: ccp: unable to access the device: you might
be running a broken BIOS.
[    5.401031] ccp 0000:07:00.2: tee enabled
[    5.401113] ccp 0000:07:00.2: psp enabled

Found an article [1] which could be relevant and downloaded and ran the
accompanying test program (written by Jason Donenfeld):
# ./amd-rdrand-bug
Your RDRAND() does not have the AMD bug.
# ./test-rdrand
RDRAND() = 0x47c993c0
RDRAND() = 0xec7c697d
... (more seemingly random numbers)
RDRAND() = 0xba858101

I tried it with the latest microcode dd 2024-07-10, but that didn't make
a difference.

So I'd like to know if this may actually be a bug on the kernel side.

Happy to provide additional information or run tests or try patches.

Cheers,
  Diederik

[1] https://arstechnica.com/gadgets/2019/10/how-a-months-old-amd-microcode-bug-destroyed-my-weekend/

Download attachment "signature.asc" of type "application/pgp-signature" (229 bytes)