KVM: arm64: Do not clear VM-wide SVE feature on vCPU init failure
Linux
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
|
Linux
| Product | Affected | Fixed in |
|---|---|---|
| Linux Linux |
| No fix yet |
| Linux Linux |
|
|
6.18.557.2.97.3In the Linux kernel, the following vulnerability has been resolved:
KVM: arm64: Do not clear VM-wide SVE feature on vCPU init failure
pkvm_vcpu_init_sve() clears KVM_ARM_VCPU_SVE in kvm->arch.vcpu_features when it fails, but vcpu_has_sve() tests KVM_ARCH_FLAG_GUEST_HAS_SVE, which is left set. Later vCPUs on that VM then skip the SVE setup and register with a NULL sve_state, which the guest's first FP access hands to sve_load_state().
Return the error without touching vcpu_features.
From the CVE record · Linux
A failed SVE initialization for a KVM arm64 vCPU can leave inconsistent VM-wide feature flags. Later vCPUs may register with a NULL SVE state, which is passed to sve_load_state() on the guest's first floating-point access. The input does not specify attacker access or authentication requirements.
Engineers running Linux kernels with KVM arm64 should check their systems; affected versions are not specified.