Copying an ASID into new vCPUs will not work for SEV-ES since the vCPUs VMSAs need to be setup and measured before SEV_LAUNCH_FINISH. Return an error if a users tries to KVM_CAP_VM_COPY_ENC_CONTEXT_FROM from an SEV-ES guest. The destination VM is already checked for SEV and SEV-ES with sev_guest(), so this ioctl already fails if the destination is SEV enabled.
Enabling mirroring a VM or copying its encryption context with an SEV-ES VM is more involved and should happen in its own feature patch if that's needed. This is because the vCPUs of SEV-ES VMs need to be updated with LAUNCH_UPDATE_VMSA before LAUNCH_FINISH. This needs KVM changes because the mirror VM has all its SEV ioctls blocked and the original VM doesn't know about the mirrors vCPUs.
Fixes: 54526d1fd593 ("KVM: x86: Support KVM VMs sharing SEV context")
V2: * Updated changelog with more information and added stable CC.
Signed-off-by: Peter Gonda pgonda@google.com Cc: Marc Orr marcorr@google.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: Sean Christopherson seanjc@google.com Cc: Nathan Tempelman natet@google.com Cc: Brijesh Singh brijesh.singh@amd.com Cc: kvm@vger.kernel.org Cc: stable@vger.kernel.org Cc: linux-kernel@vger.kernel.org --- arch/x86/kvm/svm/sev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 75e0b21ad07c..8a279027425f 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -1728,7 +1728,7 @@ int svm_vm_copy_asid_from(struct kvm *kvm, unsigned int source_fd) source_kvm = source_kvm_file->private_data; mutex_lock(&source_kvm->lock);
- if (!sev_guest(source_kvm)) { + if (!sev_guest(source_kvm) || sev_es_guest(source_kvm)) { ret = -EINVAL; goto e_source_unlock; }
On Tue, Sep 14, 2021, Peter Gonda wrote:
Copying an ASID into new vCPUs will not work for SEV-ES since the vCPUs VMSAs need to be setup and measured before SEV_LAUNCH_FINISH. Return an error if a users tries to KVM_CAP_VM_COPY_ENC_CONTEXT_FROM from an SEV-ES guest. The destination VM is already checked for SEV and SEV-ES with sev_guest(), so this ioctl already fails if the destination is SEV enabled.
Enabling mirroring a VM or copying its encryption context with an SEV-ES VM is more involved and should happen in its own feature patch if that's needed. This is because the vCPUs of SEV-ES VMs need to be updated with LAUNCH_UPDATE_VMSA before LAUNCH_FINISH. This needs KVM changes because the mirror VM has all its SEV ioctls blocked and the original VM doesn't know about the mirrors vCPUs.
mirror's, or I guess mirrors'? :-)
Fixes: 54526d1fd593 ("KVM: x86: Support KVM VMs sharing SEV context")
V2:
- Updated changelog with more information and added stable CC.
Nit, but this in the section ignored by git (below the ---) so that omitted from the committed changelog.
Signed-off-by: Peter Gonda pgonda@google.com Cc: Marc Orr marcorr@google.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: Sean Christopherson seanjc@google.com Cc: Nathan Tempelman natet@google.com Cc: Brijesh Singh brijesh.singh@amd.com Cc: kvm@vger.kernel.org Cc: stable@vger.kernel.org Cc: linux-kernel@vger.kernel.org
Reviewed-by: Sean Christopherson seanjc@google.com
Git ignores stuff in this section.
arch/x86/kvm/svm/sev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 75e0b21ad07c..8a279027425f 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -1728,7 +1728,7 @@ int svm_vm_copy_asid_from(struct kvm *kvm, unsigned int source_fd) source_kvm = source_kvm_file->private_data; mutex_lock(&source_kvm->lock);
- if (!sev_guest(source_kvm)) {
- if (!sev_guest(source_kvm) || sev_es_guest(source_kvm)) { ret = -EINVAL; goto e_source_unlock; }
-- 2.33.0.309.g3052b89438-goog
On Tue, Sep 14, 2021 at 12:01 PM Peter Gonda pgonda@google.com wrote:
Copying an ASID into new vCPUs will not work for SEV-ES since the vCPUs VMSAs need to be setup and measured before SEV_LAUNCH_FINISH. Return an error if a users tries to KVM_CAP_VM_COPY_ENC_CONTEXT_FROM from an SEV-ES guest. The destination VM is already checked for SEV and SEV-ES with sev_guest(), so this ioctl already fails if the destination is SEV enabled.
Enabling mirroring a VM or copying its encryption context with an SEV-ES VM is more involved and should happen in its own feature patch if that's needed. This is because the vCPUs of SEV-ES VMs need to be updated with LAUNCH_UPDATE_VMSA before LAUNCH_FINISH. This needs KVM changes because the mirror VM has all its SEV ioctls blocked and the original VM doesn't know about the mirrors vCPUs.
Fixes: 54526d1fd593 ("KVM: x86: Support KVM VMs sharing SEV context")
V2:
- Updated changelog with more information and added stable CC.
Signed-off-by: Peter Gonda pgonda@google.com Cc: Marc Orr marcorr@google.com Cc: Paolo Bonzini pbonzini@redhat.com Cc: Sean Christopherson seanjc@google.com Cc: Nathan Tempelman natet@google.com Cc: Brijesh Singh brijesh.singh@amd.com Cc: kvm@vger.kernel.org Cc: stable@vger.kernel.org Cc: linux-kernel@vger.kernel.org
arch/x86/kvm/svm/sev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index 75e0b21ad07c..8a279027425f 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -1728,7 +1728,7 @@ int svm_vm_copy_asid_from(struct kvm *kvm, unsigned int source_fd) source_kvm = source_kvm_file->private_data; mutex_lock(&source_kvm->lock);
if (!sev_guest(source_kvm)) {
if (!sev_guest(source_kvm) || sev_es_guest(source_kvm)) { ret = -EINVAL; goto e_source_unlock; }-- 2.33.0.309.g3052b89438-goog
Reviewed-by: Marc Orr marcorr@google.com
linux-stable-mirror@lists.linaro.org
-
Marc Orr -
Peter Gonda -
Sean Christopherson