[Qemu-devel] [PATCH 0/2] i386/kvm: TSC page clocksource for Hyper-V-on-KVM fixes

[Qemu-devel] [PATCH 0/2] i386/kvm: TSC page clocksource for Hyper-V-on-KVM fixes

[Vitaly Kuznetsov]

Previously, Ladi was working on enabling TSC page clocksource for nested
Hyper-V-on-KVM workloads. He found out that if Hyper-V frequency MSRs are
exposed to L1 as well as INVTSC flag Hyper-V enables TSC page clocksource
to its guests. Qemu doesn't pass INVTSC by default as it is a migration
blocker.

I found out there's a different way to make Hyper-V like us: expose
Reenlightenment MSRs to it. KVM doesn't fully support the feature as
we're still unable to migrate nested environments but rudimentary support
we have there (kvm/queue only currently) is enough.

Enable Hyper-V reenlightenment MSRs and expose frequency MSRs even without
INVTSC to make things work.

[My first patches for Qemu, please be nice :-) ]

Vitaly Kuznetsov (2):
  i386/kvm: add support for Hyper-V reenlightenment MSRs
  i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

 target/i386/cpu.h          |  3 +++
 target/i386/hyperv-proto.h |  9 ++++++++-
 target/i386/kvm.c          | 35 ++++++++++++++++++++++++++++++++++-
 3 files changed, 45 insertions(+), 2 deletions(-)

-- 
2.14.3

[Qemu-devel] [PATCH 1/2] i386/kvm: add support for Hyper-V reenlightenment MSRs

[Vitaly Kuznetsov]

KVM recently gained support for Hyper-V Reenlightenment MSRs which are
required to make KVM-on-Hyper-V enable TSC page clocksource to its guests
when INVTSC is not passed to it (and it is not passed by default in Qemu
as it effectively blocks migration).

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
 target/i386/cpu.h          |  3 +++
 target/i386/hyperv-proto.h |  9 ++++++++-
 target/i386/kvm.c          | 33 +++++++++++++++++++++++++++++++++
 3 files changed, 44 insertions(+), 1 deletion(-)

diff --git a/target/i386/cpu.h b/target/i386/cpu.h
index faf39ec1ce..502b535be2 100644
--- a/target/i386/cpu.h
+++ b/target/i386/cpu.h
@@ -1152,6 +1152,9 @@ typedef struct CPUX86State {
     uint64_t msr_hv_synic_sint[HV_SINT_COUNT];
     uint64_t msr_hv_stimer_config[HV_STIMER_COUNT];
     uint64_t msr_hv_stimer_count[HV_STIMER_COUNT];
+    uint64_t msr_hv_reenlightenment_control;
+    uint64_t msr_hv_tsc_emulation_control;
+    uint64_t msr_hv_tsc_emulation_status;
 
     /* exception/interrupt handling */
     int error_code;
diff --git a/target/i386/hyperv-proto.h b/target/i386/hyperv-proto.h
index cb4d7f2b7a..93352ebd2a 100644
--- a/target/i386/hyperv-proto.h
+++ b/target/i386/hyperv-proto.h
@@ -35,7 +35,7 @@
 #define HV_RESET_AVAILABLE           (1u << 7)
 #define HV_REFERENCE_TSC_AVAILABLE   (1u << 9)
 #define HV_ACCESS_FREQUENCY_MSRS     (1u << 11)
-
+#define HV_ACCESS_REENLIGHTENMENTS_CONTROL  (1u << 13)
 
 /*
  * HV_CPUID_FEATURES.EDX bits
@@ -129,6 +129,13 @@
 #define HV_X64_MSR_CRASH_CTL                    0x40000105
 #define HV_CRASH_CTL_NOTIFY                     (1ull << 63)
 
+/*
+ * Reenlightenment notification MSRs
+ */
+#define HV_X64_MSR_REENLIGHTENMENT_CONTROL      0x40000106
+#define HV_X64_MSR_TSC_EMULATION_CONTROL        0x40000107
+#define HV_X64_MSR_TSC_EMULATION_STATUS         0x40000108
+
 /*
  * Hypercall status code
  */
diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index ad4b159b28..21e06deaf1 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -90,6 +90,7 @@ static bool has_msr_hv_runtime;
 static bool has_msr_hv_synic;
 static bool has_msr_hv_stimer;
 static bool has_msr_hv_frequencies;
+static bool has_msr_hv_reenlightenment;
 static bool has_msr_xss;
 static bool has_msr_spec_ctrl;
 
@@ -649,6 +650,11 @@ static int hyperv_handle_properties(CPUState *cs)
             env->features[FEAT_HYPERV_EAX] |= HV_ACCESS_FREQUENCY_MSRS;
             env->features[FEAT_HYPERV_EDX] |= HV_FREQUENCY_MSRS_AVAILABLE;
         }
+
+        if (has_msr_hv_reenlightenment) {
+            env->features[FEAT_HYPERV_EAX] |=
+                HV_ACCESS_REENLIGHTENMENTS_CONTROL;
+        }
     }
     if (cpu->hyperv_crash && has_msr_hv_crash) {
         env->features[FEAT_HYPERV_EDX] |= HV_GUEST_CRASH_MSR_AVAILABLE;
@@ -1154,6 +1160,9 @@ static int kvm_get_supported_msrs(KVMState *s)
                 case HV_X64_MSR_TSC_FREQUENCY:
                     has_msr_hv_frequencies = true;
                     break;
+                case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
+                    has_msr_hv_reenlightenment = true;
+                    break;
                 case MSR_IA32_SPEC_CTRL:
                     has_msr_spec_ctrl = true;
                     break;
@@ -1713,6 +1722,15 @@ static int kvm_put_msrs(X86CPU *cpu, int level)
             if (cpu->hyperv_time) {
                 kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC,
                                   env->msr_hv_tsc);
+
+                if (has_msr_hv_reenlightenment) {
+                    kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL,
+                                      env->msr_hv_reenlightenment_control);
+                    kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL,
+                                      env->msr_hv_tsc_emulation_control);
+                    kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS,
+                                      env->msr_hv_tsc_emulation_status);
+                }
             }
         }
         if (cpu->hyperv_vapic) {
@@ -2053,6 +2071,12 @@ static int kvm_get_msrs(X86CPU *cpu)
     }
     if (cpu->hyperv_time) {
         kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC, 0);
+
+        if (has_msr_hv_reenlightenment) {
+            kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL, 0);
+            kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL, 0);
+            kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS, 0);
+        }
     }
     if (has_msr_hv_crash) {
         int j;
@@ -2294,6 +2318,15 @@ static int kvm_get_msrs(X86CPU *cpu)
             env->msr_hv_stimer_count[(index - HV_X64_MSR_STIMER0_COUNT)/2] =
                                 msrs[i].data;
             break;
+        case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
+            env->msr_hv_reenlightenment_control = msrs[i].data;
+            break;
+        case HV_X64_MSR_TSC_EMULATION_CONTROL:
+            env->msr_hv_tsc_emulation_control = msrs[i].data;
+            break;
+        case HV_X64_MSR_TSC_EMULATION_STATUS:
+            env->msr_hv_tsc_emulation_status = msrs[i].data;
+            break;
         case MSR_MTRRdefType:
             env->mtrr_deftype = msrs[i].data;
             break;
-- 
2.14.3

[Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Vitaly Kuznetsov]

Requiring tsc_is_stable_and_known() is too restrictive: even without INVTCS
nested Hyper-V-on-KVM enables TSC pages for its guests e.g. when
Reenlightenment MSRs are present. Presence of frequency MSRs doesn't mean
these frequencies are stable, it just means they're available for reading.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
 target/i386/kvm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index 21e06deaf1..43c521f61a 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -646,7 +646,7 @@ static int hyperv_handle_properties(CPUState *cs)
         env->features[FEAT_HYPERV_EAX] |= HV_TIME_REF_COUNT_AVAILABLE;
         env->features[FEAT_HYPERV_EAX] |= HV_REFERENCE_TSC_AVAILABLE;
 
-        if (has_msr_hv_frequencies && tsc_is_stable_and_known(env)) {
+        if (has_msr_hv_frequencies && env->tsc_khz) {
             env->features[FEAT_HYPERV_EAX] |= HV_ACCESS_FREQUENCY_MSRS;
             env->features[FEAT_HYPERV_EDX] |= HV_FREQUENCY_MSRS_AVAILABLE;
         }
-- 
2.14.3

Re: [Qemu-devel] [PATCH 1/2] i386/kvm: add support for Hyper-V reenlightenment MSRs

[Paolo Bonzini]

On 12/03/2018 16:12, Vitaly Kuznetsov wrote:
> KVM recently gained support for Hyper-V Reenlightenment MSRs which are
> required to make KVM-on-Hyper-V enable TSC page clocksource to its guests
> when INVTSC is not passed to it (and it is not passed by default in Qemu
> as it effectively blocks migration).
> 
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> ---
>  target/i386/cpu.h          |  3 +++
>  target/i386/hyperv-proto.h |  9 ++++++++-
>  target/i386/kvm.c          | 33 +++++++++++++++++++++++++++++++++
>  3 files changed, 44 insertions(+), 1 deletion(-)
> 
> diff --git a/target/i386/cpu.h b/target/i386/cpu.h
> index faf39ec1ce..502b535be2 100644
> --- a/target/i386/cpu.h
> +++ b/target/i386/cpu.h
> @@ -1152,6 +1152,9 @@ typedef struct CPUX86State {
>      uint64_t msr_hv_synic_sint[HV_SINT_COUNT];
>      uint64_t msr_hv_stimer_config[HV_STIMER_COUNT];
>      uint64_t msr_hv_stimer_count[HV_STIMER_COUNT];
> +    uint64_t msr_hv_reenlightenment_control;
> +    uint64_t msr_hv_tsc_emulation_control;
> +    uint64_t msr_hv_tsc_emulation_status;
>  
>      /* exception/interrupt handling */
>      int error_code;
> diff --git a/target/i386/hyperv-proto.h b/target/i386/hyperv-proto.h
> index cb4d7f2b7a..93352ebd2a 100644
> --- a/target/i386/hyperv-proto.h
> +++ b/target/i386/hyperv-proto.h
> @@ -35,7 +35,7 @@
>  #define HV_RESET_AVAILABLE           (1u << 7)
>  #define HV_REFERENCE_TSC_AVAILABLE   (1u << 9)
>  #define HV_ACCESS_FREQUENCY_MSRS     (1u << 11)
> -
> +#define HV_ACCESS_REENLIGHTENMENTS_CONTROL  (1u << 13)
>  
>  /*
>   * HV_CPUID_FEATURES.EDX bits
> @@ -129,6 +129,13 @@
>  #define HV_X64_MSR_CRASH_CTL                    0x40000105
>  #define HV_CRASH_CTL_NOTIFY                     (1ull << 63)
>  
> +/*
> + * Reenlightenment notification MSRs
> + */
> +#define HV_X64_MSR_REENLIGHTENMENT_CONTROL      0x40000106
> +#define HV_X64_MSR_TSC_EMULATION_CONTROL        0x40000107
> +#define HV_X64_MSR_TSC_EMULATION_STATUS         0x40000108
> +
>  /*
>   * Hypercall status code
>   */
> diff --git a/target/i386/kvm.c b/target/i386/kvm.c
> index ad4b159b28..21e06deaf1 100644
> --- a/target/i386/kvm.c
> +++ b/target/i386/kvm.c
> @@ -90,6 +90,7 @@ static bool has_msr_hv_runtime;
>  static bool has_msr_hv_synic;
>  static bool has_msr_hv_stimer;
>  static bool has_msr_hv_frequencies;
> +static bool has_msr_hv_reenlightenment;
>  static bool has_msr_xss;
>  static bool has_msr_spec_ctrl;
>  
> @@ -649,6 +650,11 @@ static int hyperv_handle_properties(CPUState *cs)
>              env->features[FEAT_HYPERV_EAX] |= HV_ACCESS_FREQUENCY_MSRS;
>              env->features[FEAT_HYPERV_EDX] |= HV_FREQUENCY_MSRS_AVAILABLE;
>          }
> +
> +        if (has_msr_hv_reenlightenment) {
> +            env->features[FEAT_HYPERV_EAX] |=
> +                HV_ACCESS_REENLIGHTENMENTS_CONTROL;
> +        }
>      }
>      if (cpu->hyperv_crash && has_msr_hv_crash) {
>          env->features[FEAT_HYPERV_EDX] |= HV_GUEST_CRASH_MSR_AVAILABLE;
> @@ -1154,6 +1160,9 @@ static int kvm_get_supported_msrs(KVMState *s)
>                  case HV_X64_MSR_TSC_FREQUENCY:
>                      has_msr_hv_frequencies = true;
>                      break;
> +                case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
> +                    has_msr_hv_reenlightenment = true;
> +                    break;
>                  case MSR_IA32_SPEC_CTRL:
>                      has_msr_spec_ctrl = true;
>                      break;
> @@ -1713,6 +1722,15 @@ static int kvm_put_msrs(X86CPU *cpu, int level)
>              if (cpu->hyperv_time) {
>                  kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC,
>                                    env->msr_hv_tsc);
> +
> +                if (has_msr_hv_reenlightenment) {
> +                    kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL,
> +                                      env->msr_hv_reenlightenment_control);
> +                    kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL,
> +                                      env->msr_hv_tsc_emulation_control);
> +                    kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS,
> +                                      env->msr_hv_tsc_emulation_status);
> +                }
>              }
>          }
>          if (cpu->hyperv_vapic) {
> @@ -2053,6 +2071,12 @@ static int kvm_get_msrs(X86CPU *cpu)
>      }
>      if (cpu->hyperv_time) {
>          kvm_msr_entry_add(cpu, HV_X64_MSR_REFERENCE_TSC, 0);
> +
> +        if (has_msr_hv_reenlightenment) {
> +            kvm_msr_entry_add(cpu, HV_X64_MSR_REENLIGHTENMENT_CONTROL, 0);
> +            kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_CONTROL, 0);
> +            kvm_msr_entry_add(cpu, HV_X64_MSR_TSC_EMULATION_STATUS, 0);
> +        }
>      }
>      if (has_msr_hv_crash) {
>          int j;
> @@ -2294,6 +2318,15 @@ static int kvm_get_msrs(X86CPU *cpu)
>              env->msr_hv_stimer_count[(index - HV_X64_MSR_STIMER0_COUNT)/2] =
>                                  msrs[i].data;
>              break;
> +        case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
> +            env->msr_hv_reenlightenment_control = msrs[i].data;
> +            break;
> +        case HV_X64_MSR_TSC_EMULATION_CONTROL:
> +            env->msr_hv_tsc_emulation_control = msrs[i].data;
> +            break;
> +        case HV_X64_MSR_TSC_EMULATION_STATUS:
> +            env->msr_hv_tsc_emulation_status = msrs[i].data;
> +            break;
>          case MSR_MTRRdefType:
>              env->mtrr_deftype = msrs[i].data;
>              break;
> 

Doesn't this also need a new subsection in target/i386/machine.c?

Paolo

Re: [Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Paolo Bonzini]

On 12/03/2018 16:12, Vitaly Kuznetsov wrote:
>  
> -        if (has_msr_hv_frequencies && tsc_is_stable_and_known(env)) {
> +        if (has_msr_hv_frequencies && env->tsc_khz) {

Should this be

((env->tsc_khz && has_msr_hv_reenlightenment) ||
 tsc_is_stable_and_known(env))

so that you don't regress on older kernels?

Paolo

>              env->features[FEAT_HYPERV_EAX] |= HV_ACCESS_FREQUENCY_MSRS;
>              env->features[FEAT_HYPERV_EDX] |= HV_FREQUENCY_MSRS_AVAILABLE;
>          }

Re: [Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Vitaly Kuznetsov]

Paolo Bonzini <pbonzini@redhat.com> writes:

> On 12/03/2018 16:12, Vitaly Kuznetsov wrote:
>>  
>> -        if (has_msr_hv_frequencies && tsc_is_stable_and_known(env)) {
>> +        if (has_msr_hv_frequencies && env->tsc_khz) {
>
> Should this be
>
> ((env->tsc_khz && has_msr_hv_reenlightenment) ||
>  tsc_is_stable_and_known(env))
>
> so that you don't regress on older kernels?
>

I don't actually see where the regression might come from: frequency
MSRs are supported regardless or reenlightenment/invtsc and there's
nothing wrong with exposing them but I may be missing something..

-- 
  Vitaly

Re: [Qemu-devel] [PATCH 1/2] i386/kvm: add support for Hyper-V reenlightenment MSRs

[Vitaly Kuznetsov]

Paolo Bonzini <pbonzini@redhat.com> writes:

> On 12/03/2018 16:12, Vitaly Kuznetsov wrote:

[snip]

>> @@ -2294,6 +2318,15 @@ static int kvm_get_msrs(X86CPU *cpu)
>>              env->msr_hv_stimer_count[(index - HV_X64_MSR_STIMER0_COUNT)/2] =
>>                                  msrs[i].data;
>>              break;
>> +        case HV_X64_MSR_REENLIGHTENMENT_CONTROL:
>> +            env->msr_hv_reenlightenment_control = msrs[i].data;
>> +            break;
>> +        case HV_X64_MSR_TSC_EMULATION_CONTROL:
>> +            env->msr_hv_tsc_emulation_control = msrs[i].data;
>> +            break;
>> +        case HV_X64_MSR_TSC_EMULATION_STATUS:
>> +            env->msr_hv_tsc_emulation_status = msrs[i].data;
>> +            break;
>>          case MSR_MTRRdefType:
>>              env->mtrr_deftype = msrs[i].data;
>>              break;
>> 
>
> Doesn't this also need a new subsection in target/i386/machine.c?
>

Actually yes, missed that completely! Thanks!

-- 
  Vitaly

Re: [Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Paolo Bonzini]

On 16/03/2018 16:05, Vitaly Kuznetsov wrote:
>>>  
>>> -        if (has_msr_hv_frequencies && tsc_is_stable_and_known(env)) {
>>> +        if (has_msr_hv_frequencies && env->tsc_khz) {
>> Should this be
>>
>> ((env->tsc_khz && has_msr_hv_reenlightenment) ||
>>  tsc_is_stable_and_known(env))
>>
>> so that you don't regress on older kernels?
>>
> I don't actually see where the regression might come from: frequency
> MSRs are supported regardless or reenlightenment/invtsc and there's
> nothing wrong with exposing them but I may be missing something..

On older kernel without re-enlightenment support, you don't want to
expose the frequency MSRs unless invtsc is on, right?

Paolo

Re: [Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Vitaly Kuznetsov]

Paolo Bonzini <pbonzini@redhat.com> writes:

> On 16/03/2018 16:05, Vitaly Kuznetsov wrote:
>>>>  
>>>> -        if (has_msr_hv_frequencies && tsc_is_stable_and_known(env)) {
>>>> +        if (has_msr_hv_frequencies && env->tsc_khz) {
>>> Should this be
>>>
>>> ((env->tsc_khz && has_msr_hv_reenlightenment) ||
>>>  tsc_is_stable_and_known(env))
>>>
>>> so that you don't regress on older kernels?
>>>
>> I don't actually see where the regression might come from: frequency
>> MSRs are supported regardless or reenlightenment/invtsc and there's
>> nothing wrong with exposing them but I may be missing something..
>
> On older kernel without re-enlightenment support, you don't want to
> expose the frequency MSRs unless invtsc is on, right?
>

Actually no, I think it's OK to expose frequency MSRs even when we don't
have invtsc and don't support re-enlightenment. Nested Hyper-V won't
pass stable TSC pages to its guests unless it sees either invtsc or
reenlightenment. So as long as we have something to put to these MSRs
(env->tsc_khz) I *think* we can expose them.

I may actually be missing the reason why Ladi put
tsc_is_stable_and_known() here. In case we're running Windows (and not
Hyper-V) as a guest KVM will update TSC page on migration. And genuine
Hyper-V also exposes these MSRs without exposing INVTSC flag by
default.

-- 
  Vitaly

Re: [Qemu-devel] [PATCH 2/2] i386/kvm: lower requirements for Hyper-V frequency MSRs exposure

[Paolo Bonzini]

On 16/03/2018 16:40, Vitaly Kuznetsov wrote:
>> On older kernel without re-enlightenment support, you don't want to
>> expose the frequency MSRs unless invtsc is on, right?
>>
> Actually no, I think it's OK to expose frequency MSRs even when we don't
> have invtsc and don't support re-enlightenment. Nested Hyper-V won't
> pass stable TSC pages to its guests unless it sees either invtsc or
> reenlightenment. So as long as we have something to put to these MSRs
> (env->tsc_khz) I *think* we can expose them.
> 
> I may actually be missing the reason why Ladi put
> tsc_is_stable_and_known() here.

Probably because I asked him to. :)  It looks like Hyper-V knows that
you need re-enlightenment in order to really trust the frequency MSRs
(of course the TSC page is special because it has the sequence count).

So the patch is good.  Thanks!

Paolo

> In case we're running Windows (and not
> Hyper-V) as a guest KVM will update TSC page on migration. And genuine
> Hyper-V also exposes these MSRs without exposing INVTSC flag by
> default.