diff options
| author |   Oliver Upton <oliver.upton@linux.dev> | 2023-07-28 00:08:24 +0000 |
|---|---|---|
| committer | Oliver Upton <oliver.upton@linux.dev> | 2023-07-28 17:11:23 +0000 |
| commit | 74158a8cad79d2f5dcf71508993664c5cfcbfa3c (patch) | |
| tree | 4f41122ca99ab2ee5fa4bebb23e22e3723fd4aba /tools/perf/scripts/python/export-to-postgresql.py | |
| parent | KVM: arm64: fix __kvm_host_psci_cpu_entry() prototype (diff) | |
| download | wireguard-linux-74158a8cad79d2f5dcf71508993664c5cfcbfa3c.tar.xz wireguard-linux-74158a8cad79d2f5dcf71508993664c5cfcbfa3c.zip | |
KVM: arm64: Skip instruction after emulating write to TCR_EL1
Whelp, this is embarrassing. Since commit 082fdfd13841 ("KVM: arm64:
Prevent guests from enabling HA/HD on Ampere1") KVM traps writes to
TCR_EL1 on AmpereOne to work around an erratum in the unadvertised
HAFDBS implementation, preventing the guest from enabling the feature.
Unfortunately, I failed virtualization 101 when working on that change,
and forgot to advance PC after instruction emulation.
Do the right thing and skip the MSR instruction after emulating the
write.
Fixes: 082fdfd13841 ("KVM: arm64: Prevent guests from enabling HA/HD on Ampere1")
Acked-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20230728000824.3848025-1-oliver.upton@linux.dev
Signed-off-by: Oliver Upton <oliver.upton@linux.dev>
Diffstat (limited to '')
0 files changed, 0 insertions, 0 deletions