[PATCH v2 0/4] KVM: VMX: nVMX: Make eVMCS enablement more robust

From: Vitaly Kuznetsov <vkuznets@redhat.com>
To: kvm@vger.kernel.org, Paolo Bonzini <pbonzini@redhat.com>,
	Sean Christopherson <seanjc@google.com>
Cc: Wanpeng Li <wanpengli@tencent.com>,
	Jim Mattson <jmattson@google.com>,
	Maxim Levitsky <mlevitsk@redhat.com>,
	linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH v2 0/4] KVM: VMX: nVMX: Make eVMCS enablement more robust
Date: Fri,  4 Nov 2022 15:47:04 +0100	[thread overview]
Message-ID: <20221104144708.435865-1-vkuznets@redhat.com> (raw)

Changes since v1:
- Simplify PATCH4 by not using evmcs_get_supported_ctls() and doing more
  macro magic to avoid having two evmcs_check_vmcs_conf{u32, u64} variants
  [Sean]

This is a continuation of "KVM: VMX: Support updated eVMCSv1 revision + use
vmcs_config for L1 VMX MSRs" work:
https://lore.kernel.org/kvm/20220830133737.1539624-1-vkuznets@redhat.com/

and a preparation to enabling new eVMCS features for Hyper-V on KVM, namely
nested TSC scaling.

Future proof KVM against two scenarios:
- nVMX: A new feature which doesn't have a corresponding eVMCSv1 field gets
 implemented in KVM but EVMCS1_UNSUPPORTED_* defines are left unchanged.
- VMX: A new feature supported by KVM but currently missing in eVMCSv1 gets
 implemented in a future Hyper-V version breaking KVM.

Note: 'vmx/evmcs.{c,h}' are renamed to 'vmx/hyperv.{c,h}' in
https://lore.kernel.org/kvm/20221004123956.188909-7-vkuznets@redhat.com/

Vitaly Kuznetsov (4):
  KVM: nVMX: Sanitize primary processor-based VM-execution controls with
    eVMCS too
  KVM: nVMX: Invert 'unsupported by eVMCSv1' check
  KVM: nVMX: Prepare to sanitize tertiary execution controls with eVMCS
  KVM: VMX: Resurrect vmcs_conf sanitization for KVM-on-Hyper-V

 arch/x86/kvm/vmx/evmcs.c | 88 +++++++++++++++++++++++++++++--------
 arch/x86/kvm/vmx/evmcs.h | 93 +++++++++++++++++++++++++++++++++-------
 arch/x86/kvm/vmx/vmx.c   |  5 +++
 3 files changed, 152 insertions(+), 34 deletions(-)

-- 
2.38.1