Re: [PATCH 4/4] KVM: hyper-v: Advertise support for fast XMM hypercalls

[Vitaly Kuznetsov]

Siddharth Chandrasekaran <sidcha@xxxxxxxxx> writes:

> On Thu, Apr 08, 2021 at 04:44:23PM +0200, Vitaly Kuznetsov wrote:
>> Siddharth Chandrasekaran <sidcha@xxxxxxxxx> writes:
>> > On Thu, Apr 08, 2021 at 02:05:53PM +0200, Vitaly Kuznetsov wrote:
>> >> Siddharth Chandrasekaran <sidcha@xxxxxxxxx> writes:
>> >> > Now that all extant hypercalls that can use XMM registers (based on
>> >> > spec) for input/outputs are patched to support them, we can start
>> >> > advertising this feature to guests.
>> >> >
>> >> > Cc: Alexander Graf <graf@xxxxxxxxxx>
>> >> > Cc: Evgeny Iakovlev <eyakovl@xxxxxxxxx>
>> >> > Signed-off-by: Siddharth Chandrasekaran <sidcha@xxxxxxxxx>
>> >> > ---
>> >> >  arch/x86/include/asm/hyperv-tlfs.h | 4 ++--
>> >> >  arch/x86/kvm/hyperv.c              | 1 +
>> >> >  2 files changed, 3 insertions(+), 2 deletions(-)
>> >> >
>> >> > diff --git a/arch/x86/include/asm/hyperv-tlfs.h b/arch/x86/include/asm/hyperv-tlfs.h
>> >> > index e6cd3fee562b..1f160ef60509 100644
>> >> > --- a/arch/x86/include/asm/hyperv-tlfs.h
>> >> > +++ b/arch/x86/include/asm/hyperv-tlfs.h
>> >> > @@ -49,10 +49,10 @@
>> >> >  /* Support for physical CPU dynamic partitioning events is available*/
>> >> >  #define HV_X64_CPU_DYNAMIC_PARTITIONING_AVAILABLE    BIT(3)
>> >> >  /*
>> >> > - * Support for passing hypercall input parameter block via XMM
>> >> > + * Support for passing hypercall input and output parameter block via XMM
>> >> >   * registers is available
>> >> >   */
>> >> > -#define HV_X64_HYPERCALL_PARAMS_XMM_AVAILABLE                BIT(4)
>> >> > +#define HV_X64_HYPERCALL_PARAMS_XMM_AVAILABLE                BIT(4) | BIT(15)
>> >>
>> >> TLFS 6.0b states that there are two distinct bits for input and output:
>> >>
>> >> CPUID Leaf 0x40000003.EDX:
>> >> Bit 4: support for passing hypercall input via XMM registers is available.
>> >> Bit 15: support for returning hypercall output via XMM registers is available.
>> >>
>> >> and HV_X64_HYPERCALL_PARAMS_XMM_AVAILABLE is not currently used
>> >> anywhere, I'd suggest we just rename
>> >>
>> >> HV_X64_HYPERCALL_PARAMS_XMM_AVAILABLE to HV_X64_HYPERCALL_XMM_INPUT_AVAILABLE
>> >> and add HV_X64_HYPERCALL_XMM_OUTPUT_AVAILABLE (bit 15).
>> >
>> > That is how I had it initially; but then noticed that we would never
>> > need to use either of them separately. So it seemed like a reasonable
>> > abstraction to put them together.
>> >
>> 
>> Actually, we may. In theory, KVM userspace may decide to expose just
>> one of these two to the guest as it is not obliged to copy everything
>> from KVM_GET_SUPPORTED_HV_CPUID so we will need separate
>> guest_cpuid_has() checks.
>
> Looks like guest_cpuid_has() check is for x86 CPU features only (if I'm
> not mistaken) and I don't see a suitable alternative that looks into
> vcpu->arch.cpuid_entries[]. So I plan to add a new method
> hv_guest_cpuid_has() in hyperv.c to have this check; does that sound
> right to you?
> If you can give a quick go-ahead, I'll make the changes requested so
> far and send v2 this series.

Sorry my mistake, guest_cpuid_has() was the wrong function to name. In the
meantime I started working on fine-grained access to the existing
Hyper-V enlightenments as well and I think the best approach would be to
cache CPUID 0x40000003 (EAX, EBX, EDX) in kvm_hv_set_cpuid()  to avoid
looping through all guest CPUID entries on every hypercall. Your check
will then look like

 if (hv_vcpu->cpuid_cache.features_edx & HV_X64_HYPERCALL_XMM_INPUT_AVAILABLE)
 ...


 if (hv_vcpu->cpuid_cache.features_edx & HV_X64_HYPERCALL_XMM_OUTPUT_AVAILABLE)
 ...

We can wrap this into a hv_guest_cpuid_has() helper indeed, it'll look like:

 if (hv_guest_cpuid_has(vcpu, HYPERV_CPUID_FEATURES, CPUID_EDX, HV_X64_HYPERCALL_XMM_INPUT_AVAILABLE))
 ...

but I'm not sure it's worth it, maybe raw check is shorter and better.

I plan to send something out in a day or two, I'll Cc: you. Feel free to
do v2 without this, if your series gets merged first I can just add the
'fine-grained access' to mine.

Thanks!

-- 
Vitaly