On Thu, Nov 13, 2025 at 02:56:17PM -0800, Sean Christopherson wrote:
> Check for an unexpected/unhandled VM-Exit after the manual RETPOLINE=y
> handling. The entire point of the RETPOLINE checks is to optimize for
> common VM-Exits, i.e. checking for the rare case of an unsupported
> VM-Exit is counter-productive. This also aligns SVM and VMX exit handling.
>
> No functional change intended.
>
> Signed-off-by: Sean Christopherson <seanjc@google.com>
Reviewed-by: Yosry Ahmed <yosry.ahmed@linux.dev>
> ---
> arch/x86/kvm/svm/svm.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
> index 638a67ef0c37..202a4d8088a2 100644
> --- a/arch/x86/kvm/svm/svm.c
> +++ b/arch/x86/kvm/svm/svm.c
> @@ -3435,12 +3435,6 @@ static void dump_vmcb(struct kvm_vcpu *vcpu)
>
> int svm_invoke_exit_handler(struct kvm_vcpu *vcpu, u64 exit_code)
> {
> - if (exit_code >= ARRAY_SIZE(svm_exit_handlers))
> - goto unexpected_vmexit;
> -
> - if (!svm_exit_handlers[exit_code])
> - goto unexpected_vmexit;
> -
> #ifdef CONFIG_MITIGATION_RETPOLINE
> if (exit_code == SVM_EXIT_MSR)
> return msr_interception(vcpu);
> @@ -3457,6 +3451,12 @@ int svm_invoke_exit_handler(struct kvm_vcpu *vcpu, u64 exit_code)
> return sev_handle_vmgexit(vcpu);
> #endif
> #endif
> + if (exit_code >= ARRAY_SIZE(svm_exit_handlers))
> + goto unexpected_vmexit;
> +
> + if (!svm_exit_handlers[exit_code])
> + goto unexpected_vmexit;
> +
> return svm_exit_handlers[exit_code](vcpu);
>
> unexpected_vmexit:
> --
> 2.52.0.rc1.455.g30608eb744-goog
>