Re: [PATCH v7 6/7] KVM: x86: Load Guest fpu state when accessing MSRs managed by XSAVES

[Jim Mattson <jmattson@google.com>]

On Thu, Sep 26, 2019 at 7:17 PM Yang Weijiang <weijiang.yang@intel.com> wrote:
>
> From: Sean Christopherson <sean.j.christopherson@intel.com>
>
> A handful of CET MSRs are not context switched through "traditional"
> methods, e.g. VMCS or manual switching, but rather are passed through
> to the guest and are saved and restored by XSAVES/XRSTORS, i.e. the
> guest's FPU state.
>
> Load the guest's FPU state if userspace is accessing MSRs whose values
> are managed by XSAVES so that the MSR helper, e.g. vmx_{get,set}_msr(),
> can simply do {RD,WR}MSR to access the guest's value.
>
> Note that guest_cpuid_has() is not queried as host userspace is allowed
> to access MSRs that have not been exposed to the guest, e.g. it might do
> KVM_SET_MSRS prior to KVM_SET_CPUID2.
>
> Signed-off-by: Sean Christopherson <sean.j.christopherson@intel.com>
> Co-developed-by: Yang Weijiang <weijiang.yang@intel.com>
> Signed-off-by: Yang Weijiang <weijiang.yang@intel.com>
> ---
>  arch/x86/kvm/x86.c | 22 +++++++++++++++++++++-
>  1 file changed, 21 insertions(+), 1 deletion(-)
>
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 290c3c3efb87..5b8116028a59 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -104,6 +104,8 @@ static void enter_smm(struct kvm_vcpu *vcpu);
>  static void __kvm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags);
>  static void store_regs(struct kvm_vcpu *vcpu);
>  static int sync_regs(struct kvm_vcpu *vcpu);
> +static void kvm_load_guest_fpu(struct kvm_vcpu *vcpu);
> +static void kvm_put_guest_fpu(struct kvm_vcpu *vcpu);
>
>  struct kvm_x86_ops *kvm_x86_ops __read_mostly;
>  EXPORT_SYMBOL_GPL(kvm_x86_ops);
> @@ -2999,6 +3001,12 @@ int kvm_get_msr_common(struct kvm_vcpu *vcpu, struct msr_data *msr_info)
>  }
>  EXPORT_SYMBOL_GPL(kvm_get_msr_common);
>
> +static bool is_xsaves_msr(u32 index)
> +{
> +       return index == MSR_IA32_U_CET ||
> +              (index >= MSR_IA32_PL0_SSP && index <= MSR_IA32_PL3_SSP);
> +}
> +
>  /*
>   * Read or write a bunch of msrs. All parameters are kernel addresses.
>   *
> @@ -3009,11 +3017,23 @@ static int __msr_io(struct kvm_vcpu *vcpu, struct kvm_msrs *msrs,
>                     int (*do_msr)(struct kvm_vcpu *vcpu,
>                                   unsigned index, u64 *data))
>  {
> +       bool fpu_loaded = false;
>         int i;
> +       const u64 cet_bits = XFEATURE_MASK_CET_USER | XFEATURE_MASK_CET_KERNEL;
> +       bool cet_xss = kvm_x86_ops->xsaves_supported() &&
> +                      (kvm_supported_xss() & cet_bits);

It seems like I've seen a lot of checks like this. Can this be
simplified (throughout this series) by sinking the
kvm_x86_ops->xsaves_supported() check into kvm_supported_xss()? That
is, shouldn't kvm_supported_xss() return 0 if
kvm_x86_ops->xsaves_supported() is false?

> -       for (i = 0; i < msrs->nmsrs; ++i)
> +       for (i = 0; i < msrs->nmsrs; ++i) {
> +               if (!fpu_loaded && cet_xss &&
> +                   is_xsaves_msr(entries[i].index)) {
> +                       kvm_load_guest_fpu(vcpu);
> +                       fpu_loaded = true;
> +               }
>                 if (do_msr(vcpu, entries[i].index, &entries[i].data))
>                         break;
> +       }
> +       if (fpu_loaded)
> +               kvm_put_guest_fpu(vcpu);
>
>         return i;
>  }
> --
> 2.17.2
>