Patch 1 is brown paper bag fix for a copy+paste bug (thankfully from this
cycle).
Patch 2 fixes a nVMX + APICv bug where KVM essentially corrupts vmcs02 if
an APICv update arrives while L2 is running. Found when testing a slight
variation of patch 3.
Patch 3 fixes a race where an APICv update that occurs while a vCPU is
being created will fail to notify that vCPU due it not yet being visible
to the updater.
Patch 4 is a minor optimization/cleanup found by inspection when digging
into everything else.
v2:
- Collect reviews. [Maxim]
- Reword patch 2's changelog to make it clear that ignoring vmcs02
is perfectly ok. [Maxim]
v1: https://lore.kernel.org/all/20220416034249.2609491-3-seanjc@google.com
Sean Christopherson (4):
KVM: x86: Tag APICv DISABLE inhibit, not ABSENT, if APICv is disabled
KVM: nVMX: Defer APICv updates while L2 is active until L1 is active
KVM: x86: Pend KVM_REQ_APICV_UPDATE during vCPU creation to fix a race
KVM: x86: Skip KVM_GUESTDBG_BLOCKIRQ APICv update if APICv is disabled
arch/x86/kvm/vmx/nested.c | 5 +++++
arch/x86/kvm/vmx/vmx.c | 5 +++++
arch/x86/kvm/vmx/vmx.h | 1 +
arch/x86/kvm/x86.c | 20 ++++++++++++++++++--
4 files changed, 29 insertions(+), 2 deletions(-)
base-commit: 150866cd0ec871c765181d145aa0912628289c8a
--
2.36.0.rc0.470.gd361397f0d-goog