| Message ID | 20220722202303.391709-2-aaronlewis@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | MSR filtering / exiting flag cleanup | expand |
diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h index ee3896416c68..e6dd76c94d47 100644 --- a/arch/x86/include/uapi/asm/kvm.h +++ b/arch/x86/include/uapi/asm/kvm.h @@ -222,7 +222,9 @@ struct kvm_msr_filter_range { #define KVM_MSR_FILTER_MAX_RANGES 16 struct kvm_msr_filter { +#ifndef __KERNEL__ #define KVM_MSR_FILTER_DEFAULT_ALLOW (0 << 0) +#endif #define KVM_MSR_FILTER_DEFAULT_DENY (1 << 0) __u32 flags; struct kvm_msr_filter_range ranges[KVM_MSR_FILTER_MAX_RANGES];
Protect the kernel from using the flag KVM_MSR_FILTER_DEFAULT_ALLOW. Its value is 0, and using it incorrectly could have unintended consequences. E.g. prevent someone in the kernel from writing something like this. if (filter.flags & KVM_MSR_FILTER_DEFAULT_ALLOW) <allow the MSR> and getting confused when it doesn't work. It would be more ideal to remove this flag altogether, but userspace may already be using it, so protecting the kernel is all that can reasonably be done at this point. Suggested-by: Sean Christopherson <seanjc@google.com> Signed-off-by: Aaron Lewis <aaronlewis@google.com> --- Google's VMM is already using this flag, so we *know* that dropping the flag entirely will break userspace. All we can do at this point is prevent the kernel from using it. arch/x86/include/uapi/asm/kvm.h | 2 ++ 1 file changed, 2 insertions(+)