[PATCH 0/1] Avoid triggering an fTPM bug from kernel

From: Mario Limonciello <mario.limonciello@amd.com>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jarkko Sakkinen <jarkko@kernel.org>,
	Thorsten Leemhuis <regressions@leemhuis.info>,
	James Bottomley <James.Bottomley@hansenpartnership.com>,
	<Jason@zx2c4.com>, <linux-integrity@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>,
	"Mario Limonciello" <mario.limonciello@amd.com>
Subject: [PATCH 0/1] Avoid triggering an fTPM bug from kernel
Date: Tue, 14 Feb 2023 14:19:54 -0600	[thread overview]
Message-ID: <20230214201955.7461-1-mario.limonciello@amd.com> (raw)

AMD issued an advisory that having fTPM enabled and utilizing certain
functionality can cause stuttering in the OS. This was a Windows specific
problem initially, but commit b006c439d58db ("hwrng: core - start
hwrng kthread also for untrusted sources") exposed it for Linux as well.

This issue has been fixed by newer fTPM firmware, but not many system
designers have rolled out the fix, so to avoid triggering it check
AMD's fTPM implementation version to decide whether to register the fTPM
RNG.

As this regression was reported by many users and escalated by Thorsten
as tracking the regressions it's being sent directly to Linus to try to
catch 6.2 release.

Mario Limonciello (1):
  tpm: disable hwrng for fTPM on some AMD designs

 drivers/char/tpm/tpm-chip.c | 62 ++++++++++++++++++++++++++++++-
 drivers/char/tpm/tpm.h      | 73 +++++++++++++++++++++++++++++++++++++
 2 files changed, 134 insertions(+), 1 deletion(-)

-- 
2.25.1