On 25/09/19 23:37, Sebastian Andrzej Siewior wrote: > I was surprised to see that the guest reported `fxsave_leak' while the > host did not. After digging deeper I noticed that the bits are simply > masked out during enumeration. > The XSAVEERPTR feature is actually a bug fix on AMD which means the > kernel can disable a workaround. > While here, I've seen that CLZERO is also masked out. This opcode is > unprivilged so exposing it to the guest should not make any difference. > > Pass CLZERO and XSAVEERPTR to the guest if available on the host. > > Signed-off-by: Sebastian Andrzej Siewior <bige...@linutronix.de> > --- > arch/x86/kvm/cpuid.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c > index 22c2720cd948e..0ae9194d0f4d2 100644 > --- a/arch/x86/kvm/cpuid.c > +++ b/arch/x86/kvm/cpuid.c > @@ -473,6 +473,7 @@ static inline int __do_cpuid_func(struct kvm_cpuid_entry2 > *entry, u32 function, > > /* cpuid 0x80000008.ebx */ > const u32 kvm_cpuid_8000_0008_ebx_x86_features = > + F(CLZERO) | F(XSAVEERPTR) | > F(WBNOINVD) | F(AMD_IBPB) | F(AMD_IBRS) | F(AMD_SSBD) | > F(VIRT_SSBD) | > F(AMD_SSB_NO) | F(AMD_STIBP) | F(AMD_STIBP_ALWAYS_ON); > >
Applied (on top of Jim's CLZERO patch, so removing the CLZERO reference in the commit message). Paolo
