On Tue, Apr 12, 2022, Jue Wang wrote:
> This feature is only enabled when the vCPU has opted in to enable
> MCG_CMCI_P.
Again, waaaay too terse. What is CMCI? What does "support" mean since an astute
reader will notice that it's impossible for MCG_CMCI_P to be set. How/when will
the vCPU (which is wrong no? doesn't userspace do the write?) be able to opt-in?
> Signed-off-by: Jue Wang <juew@xxxxxxxxxx>
> ---
> arch/x86/kvm/lapic.c | 33 ++++++++++++++++++++++++++-------
> arch/x86/kvm/lapic.h | 7 ++++++-
> 2 files changed, 32 insertions(+), 8 deletions(-)
>
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 2c770e4c0e6c..0b370ccd11a1 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -27,6 +27,7 @@
> #include <linux/math64.h>
> #include <linux/slab.h>
> #include <asm/processor.h>
> +#include <asm/mce.h>
> #include <asm/msr.h>
> #include <asm/page.h>
> #include <asm/current.h>
> @@ -364,9 +365,14 @@ static inline int apic_lvt_nmi_mode(u32 lvt_val)
> return (lvt_val & (APIC_MODE_MASK | APIC_LVT_MASKED)) == APIC_DM_NMI;
> }
>
> +static inline bool kvm_is_cmci_supported(struct kvm_vcpu *vcpu)
> +{
> + return vcpu->arch.mcg_cap & MCG_CMCI_P;
> +}
> +
> static inline int kvm_apic_get_nr_lvt_entries(struct kvm_vcpu *vcpu)
I think it makes sense to take @apic here, not @vcpu, since this is an APIC-specific
helper. kvm_apic_set_version() will need to be modified to not call
kvm_apic_get_nr_lvt_entries() until after it has verified the local APIC is in-kernel,
but IMO that's a good thing.
> {
> - return KVM_APIC_MAX_NR_LVT_ENTRIES;
> + return KVM_APIC_MAX_NR_LVT_ENTRIES - !kvm_is_cmci_supported(vcpu);
> }
>
> void kvm_apic_set_version(struct kvm_vcpu *vcpu)
