arm64: errata: Mitigate Ampere1 erratum AC03_CPU_38 at stage-2
authorOliver Upton <oliver.upton@linux.dev>
Fri, 9 Jun 2023 22:01:02 +0000 (22:01 +0000)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 3 Aug 2023 08:23:54 +0000 (10:23 +0200)
commitcde7f2fd6323e5054e1db2f1ee7231864cd5294c
treec106458d77d3a59789ba40ff7df0a2b0e7ce8139
parenteb1de0a2347cf1968bf0ee39cbed2239ea9f210a
arm64: errata: Mitigate Ampere1 erratum AC03_CPU_38 at stage-2

[ Upstream commit 6df696cd9bc1ceed0e92e36908f88bbd16d18255 ]

AmpereOne has an erratum in its implementation of FEAT_HAFDBS that
required disabling the feature on the design. This was done by reporting
the feature as not implemented in the ID register, although the
corresponding control bits were not actually RES0. This does not align
well with the requirements of the architecture, which mandates these
bits be RES0 if HAFDBS isn't implemented.

The kernel's use of stage-1 is unaffected, as the HA and HD bits are
only set if HAFDBS is detected in the ID register. KVM, on the other
hand, relies on the RES0 behavior at stage-2 to use the same value for
VTCR_EL2 on any cpu in the system. Mitigate the non-RES0 behavior by
leaving VTCR_EL2.HA clear on affected systems.

Cc: stable@vger.kernel.org
Cc: D Scott Phillips <scott@os.amperecomputing.com>
Cc: Darren Hart <darren@os.amperecomputing.com>
Acked-by: D Scott Phillips <scott@os.amperecomputing.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20230609220104.1836988-2-oliver.upton@linux.dev
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Documentation/arm64/silicon-errata.rst
arch/arm64/Kconfig
arch/arm64/kernel/cpu_errata.c
arch/arm64/kvm/hyp/pgtable.c
arch/arm64/tools/cpucaps