diff mbox series

[v10,38/50] KVM: SEV: Add support for GHCB-based termination requests

Message ID 20231016132819.1002933-39-michael.roth@amd.com (mailing list archive)
State New, archived
Headers show
Series Add AMD Secure Nested Paging (SEV-SNP) Hypervisor Support | expand

Commit Message

Michael Roth Oct. 16, 2023, 1:28 p.m. UTC
GHCB version 2 adds support for a GHCB-based termination request that
a guest can issue when it reaches an error state and wishes to inform
the hypervisor that it should be terminated. Implement support for that
similarly to GHCB MSR-based termination requests that are already
available to SEV-ES guests via earlier versions of the GHCB protocol.

Signed-off-by: Michael Roth <michael.roth@amd.com>
---
 arch/x86/kvm/svm/sev.c | 9 +++++++++
 1 file changed, 9 insertions(+)

Comments

Liam Merwick Oct. 19, 2023, 12:20 p.m. UTC | #1
On 16/10/2023 14:28, Michael Roth wrote:
> GHCB version 2 adds support for a GHCB-based termination request that
> a guest can issue when it reaches an error state and wishes to inform
> the hypervisor that it should be terminated. Implement support for that
> similarly to GHCB MSR-based termination requests that are already
> available to SEV-ES guests via earlier versions of the GHCB protocol.


Maybe add

See 'Termination Request' in the 'Invoking VMGEXIT' section of AMD's 
GHCB spec for more details.

> 
> Signed-off-by: Michael Roth <michael.roth@amd.com>
> ---
>   arch/x86/kvm/svm/sev.c | 9 +++++++++
>   1 file changed, 9 insertions(+)
> 
> diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
> index e547adddacfa..9c38fe796e00 100644
> --- a/arch/x86/kvm/svm/sev.c
> +++ b/arch/x86/kvm/svm/sev.c
> @@ -3094,6 +3094,7 @@ static int sev_es_validate_vmgexit(struct vcpu_svm *svm)
>   	case SVM_VMGEXIT_UNSUPPORTED_EVENT:
>   	case SVM_VMGEXIT_HV_FEATURES:
>   	case SVM_VMGEXIT_PSC:
> +	case SVM_VMGEXIT_TERM_REQUEST:
>   		break;
>   	default:
>   		reason = GHCB_ERR_INVALID_EVENT;
> @@ -3762,6 +3763,14 @@ int sev_handle_vmgexit(struct kvm_vcpu *vcpu)
>   
>   		ret = 1;
>   		break;
> +	case SVM_VMGEXIT_TERM_REQUEST:
> +		pr_info("SEV-ES guess requested termination: reason %#llx info %#llx\n",
> +			control->exit_info_1, control->exit_info_1);

typo: "guess" -> "guest"
It prints exit_info_1 twice - was one of those meant to be exit_info_2?

Otherwise
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>


> +		vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
> +		vcpu->run->system_event.type = KVM_SYSTEM_EVENT_SEV_TERM;
> +		vcpu->run->system_event.ndata = 1;
> +		vcpu->run->system_event.data[0] = control->ghcb_gpa;
> +		break;
>   	case SVM_VMGEXIT_UNSUPPORTED_EVENT:
>   		vcpu_unimpl(vcpu,
>   			    "vmgexit: unsupported event - exit_info_1=%#llx, exit_info_2=%#llx\n",
diff mbox series

Patch

diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
index e547adddacfa..9c38fe796e00 100644
--- a/arch/x86/kvm/svm/sev.c
+++ b/arch/x86/kvm/svm/sev.c
@@ -3094,6 +3094,7 @@  static int sev_es_validate_vmgexit(struct vcpu_svm *svm)
 	case SVM_VMGEXIT_UNSUPPORTED_EVENT:
 	case SVM_VMGEXIT_HV_FEATURES:
 	case SVM_VMGEXIT_PSC:
+	case SVM_VMGEXIT_TERM_REQUEST:
 		break;
 	default:
 		reason = GHCB_ERR_INVALID_EVENT;
@@ -3762,6 +3763,14 @@  int sev_handle_vmgexit(struct kvm_vcpu *vcpu)
 
 		ret = 1;
 		break;
+	case SVM_VMGEXIT_TERM_REQUEST:
+		pr_info("SEV-ES guess requested termination: reason %#llx info %#llx\n",
+			control->exit_info_1, control->exit_info_1);
+		vcpu->run->exit_reason = KVM_EXIT_SYSTEM_EVENT;
+		vcpu->run->system_event.type = KVM_SYSTEM_EVENT_SEV_TERM;
+		vcpu->run->system_event.ndata = 1;
+		vcpu->run->system_event.data[0] = control->ghcb_gpa;
+		break;
 	case SVM_VMGEXIT_UNSUPPORTED_EVENT:
 		vcpu_unimpl(vcpu,
 			    "vmgexit: unsupported event - exit_info_1=%#llx, exit_info_2=%#llx\n",