On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
> On 11/8/21 10:59, Kele Huang wrote:
> > Currently, AVIC is disabled if x2apic feature is exposed to guest
> > or in-kernel PIT is in re-injection mode.
> >
> > We can enable AVIC with options:
> >
> > Kmod args:
> > modprobe kvm_amd avic=1 nested=0 npt=1
> > QEMU args:
> > ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
> >
> > When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
> > can accelerate IPI operations for guest. However, the relationship
> > between AVIC and PV_SEND_IPI feature is not sorted out.
> >
> > In logical, AVIC accelerates most of frequently IPI operations
> > without VMM intervention, while the re-hooking of apic->send_IPI_xxx
> > from PV_SEND_IPI feature masks out it. People can get confused
> > if AVIC is enabled while getting lots of hypercall kvm_exits
> > from IPI.
> >
> > In performance, benchmark tool
> > https://lore.kernel.org/kvm/[email protected]/
> > shows below results:
> >
> > Test env:
> > CPU: AMD EPYC 7742 64-Core Processor
> > 2 vCPUs pinned 1:1
> > idle=poll
> >
> > Test result (average ns per IPI of lots of running):
> > PV_SEND_IPI : 1860
> > AVIC : 1390
> >
> > Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
> > do have some solid performance test results to this.
> >
> > This patch fixes this by masking out PV_SEND_IPI feature when
> > AVIC is enabled in setting up of guest vCPUs' CPUID.
> >
> > Signed-off-by: Kele Huang <[email protected]>
>
> AVIC can change across migration. I think we should instead use a new
> KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
> The KVM_HINTS_* bits are intended to be changeable across migration,
> even though we don't have for now anything equivalent to the Hyper-V
> reenlightenment interrupt.
Note that the same issue exists with HyperV. It also has PV APIC,
which is harmful when AVIC is enabled (that is guest uses it instead
of using AVIC, negating AVIC benefits).
Also note that Intel recently posted IPI virtualizaion, which
will make this issue relevant to APICv too soon.
I don't yet know if there is a solution to this which doesn't
involve some management software decision (e.g libvirt or higher).
Best regards,
Maxim Levitsky
>
> Paolo
>
> > ---
> > arch/x86/kvm/cpuid.c | 4 ++--
> > arch/x86/kvm/svm/svm.c | 13 +++++++++++++
> > 2 files changed, 15 insertions(+), 2 deletions(-)
> >
> > diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
> > index 2d70edb0f323..cc22975e2ac5 100644
> > --- a/arch/x86/kvm/cpuid.c
> > +++ b/arch/x86/kvm/cpuid.c
> > @@ -194,8 +194,6 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
> > best->ecx |= XFEATURE_MASK_FPSSE;
> > }
> >
> > - kvm_update_pv_runtime(vcpu);
> > -
> > vcpu->arch.maxphyaddr = cpuid_query_maxphyaddr(vcpu);
> > vcpu->arch.reserved_gpa_bits = kvm_vcpu_reserved_gpa_bits_raw(vcpu);
> >
> > @@ -208,6 +206,8 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
> > /* Invoke the vendor callback only after the above state is updated. */
> > static_call(kvm_x86_vcpu_after_set_cpuid)(vcpu);
> >
> > + kvm_update_pv_runtime(vcpu);
> > +
> > /*
> > * Except for the MMU, which needs to do its thing any vendor specific
> > * adjustments to the reserved GPA bits.
> > diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
> > index b36ca4e476c2..b13bcfb2617c 100644
> > --- a/arch/x86/kvm/svm/svm.c
> > +++ b/arch/x86/kvm/svm/svm.c
> > @@ -4114,6 +4114,19 @@ static void svm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
> > if (nested && guest_cpuid_has(vcpu, X86_FEATURE_SVM))
> > kvm_request_apicv_update(vcpu->kvm, false,
> > APICV_INHIBIT_REASON_NESTED);
> > +
> > + if (!guest_cpuid_has(vcpu, X86_FEATURE_X2APIC) &&
> > + !(nested && guest_cpuid_has(vcpu, X86_FEATURE_SVM))) {
> > + /*
> > + * PV_SEND_IPI feature masks out AVIC acceleration to IPI.
> > + * So, we do not expose PV_SEND_IPI feature to guest when
> > + * AVIC is enabled.
> > + */
> > + best = kvm_find_cpuid_entry(vcpu, KVM_CPUID_FEATURES, 0);
> > + if (best && enable_apicv &&
> > + (best->eax & (1 << KVM_FEATURE_PV_SEND_IPI)))
> > + best->eax &= ~(1 << KVM_FEATURE_PV_SEND_IPI);
> > + }
> > }
> > init_vmcb_after_set_cpuid(vcpu);
> > }
> >
On 11/8/21 7:08 PM, Maxim Levitsky wrote:
> On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
>> On 11/8/21 10:59, Kele Huang wrote:
>>> Currently, AVIC is disabled if x2apic feature is exposed to guest
>>> or in-kernel PIT is in re-injection mode.
>>>
>>> We can enable AVIC with options:
>>>
>>> Kmod args:
>>> modprobe kvm_amd avic=1 nested=0 npt=1
>>> QEMU args:
>>> ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
>>>
>>> When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
>>> can accelerate IPI operations for guest. However, the relationship
>>> between AVIC and PV_SEND_IPI feature is not sorted out.
>>>
>>> In logical, AVIC accelerates most of frequently IPI operations
>>> without VMM intervention, while the re-hooking of apic->send_IPI_xxx
>>> from PV_SEND_IPI feature masks out it. People can get confused
>>> if AVIC is enabled while getting lots of hypercall kvm_exits
>>> from IPI.
>>>
>>> In performance, benchmark tool
>>> https://lore.kernel.org/kvm/[email protected]/
>>> shows below results:
>>>
>>> Test env:
>>> CPU: AMD EPYC 7742 64-Core Processor
>>> 2 vCPUs pinned 1:1
>>> idle=poll
>>>
>>> Test result (average ns per IPI of lots of running):
>>> PV_SEND_IPI : 1860
>>> AVIC : 1390
>>>
>>> Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
>>> do have some solid performance test results to this.
>>>
>>> This patch fixes this by masking out PV_SEND_IPI feature when
>>> AVIC is enabled in setting up of guest vCPUs' CPUID.
>>>
>>> Signed-off-by: Kele Huang <[email protected]>
>>
>> AVIC can change across migration. I think we should instead use a new
>> KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
>> The KVM_HINTS_* bits are intended to be changeable across migration,
>> even though we don't have for now anything equivalent to the Hyper-V
>> reenlightenment interrupt.
>
> Note that the same issue exists with HyperV. It also has PV APIC,
> which is harmful when AVIC is enabled (that is guest uses it instead
> of using AVIC, negating AVIC benefits).
>
> Also note that Intel recently posted IPI virtualizaion, which
> will make this issue relevant to APICv too soon.
>
> I don't yet know if there is a solution to this which doesn't
> involve some management software decision (e.g libvirt or higher).
>
> Best regards,
> Maxim Levitsky
>
For QEMU, "-cpu host,kvm-pv-ipi=off" can disable kvm-pv-ipi.
And for libvirt, I posted a patch to disable kvm-pv-ipi by libvirt xml,
link:
https://github.com/libvirt/libvirt/commit/b2757b697e29fa86972a4638a5879dccc8add2ad
>>
>> Paolo
>>
>>> ---
>>> arch/x86/kvm/cpuid.c | 4 ++--
>>> arch/x86/kvm/svm/svm.c | 13 +++++++++++++
>>> 2 files changed, 15 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c
>>> index 2d70edb0f323..cc22975e2ac5 100644
>>> --- a/arch/x86/kvm/cpuid.c
>>> +++ b/arch/x86/kvm/cpuid.c
>>> @@ -194,8 +194,6 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
>>> best->ecx |= XFEATURE_MASK_FPSSE;
>>> }
>>>
>>> - kvm_update_pv_runtime(vcpu);
>>> -
>>> vcpu->arch.maxphyaddr = cpuid_query_maxphyaddr(vcpu);
>>> vcpu->arch.reserved_gpa_bits = kvm_vcpu_reserved_gpa_bits_raw(vcpu);
>>>
>>> @@ -208,6 +206,8 @@ static void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
>>> /* Invoke the vendor callback only after the above state is updated. */
>>> static_call(kvm_x86_vcpu_after_set_cpuid)(vcpu);
>>>
>>> + kvm_update_pv_runtime(vcpu);
>>> +
>>> /*
>>> * Except for the MMU, which needs to do its thing any vendor specific
>>> * adjustments to the reserved GPA bits.
>>> diff --git a/arch/x86/kvm/svm/svm.c b/arch/x86/kvm/svm/svm.c
>>> index b36ca4e476c2..b13bcfb2617c 100644
>>> --- a/arch/x86/kvm/svm/svm.c
>>> +++ b/arch/x86/kvm/svm/svm.c
>>> @@ -4114,6 +4114,19 @@ static void svm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu)
>>> if (nested && guest_cpuid_has(vcpu, X86_FEATURE_SVM))
>>> kvm_request_apicv_update(vcpu->kvm, false,
>>> APICV_INHIBIT_REASON_NESTED);
>>> +
>>> + if (!guest_cpuid_has(vcpu, X86_FEATURE_X2APIC) &&
>>> + !(nested && guest_cpuid_has(vcpu, X86_FEATURE_SVM))) {
>>> + /*
>>> + * PV_SEND_IPI feature masks out AVIC acceleration to IPI.
>>> + * So, we do not expose PV_SEND_IPI feature to guest when
>>> + * AVIC is enabled.
>>> + */
>>> + best = kvm_find_cpuid_entry(vcpu, KVM_CPUID_FEATURES, 0);
>>> + if (best && enable_apicv &&
>>> + (best->eax & (1 << KVM_FEATURE_PV_SEND_IPI)))
>>> + best->eax &= ~(1 << KVM_FEATURE_PV_SEND_IPI);
>>> + }
>>> }
>>> init_vmcb_after_set_cpuid(vcpu);
>>> }
>>>
>
>
--
zhenwei pi
On Mon, 8 Nov 2021 at 22:09, Maxim Levitsky <[email protected]> wrote:
>
> On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
> > On 11/8/21 10:59, Kele Huang wrote:
> > > Currently, AVIC is disabled if x2apic feature is exposed to guest
> > > or in-kernel PIT is in re-injection mode.
> > >
> > > We can enable AVIC with options:
> > >
> > > Kmod args:
> > > modprobe kvm_amd avic=1 nested=0 npt=1
> > > QEMU args:
> > > ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
> > >
> > > When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
> > > can accelerate IPI operations for guest. However, the relationship
> > > between AVIC and PV_SEND_IPI feature is not sorted out.
> > >
> > > In logical, AVIC accelerates most of frequently IPI operations
> > > without VMM intervention, while the re-hooking of apic->send_IPI_xxx
> > > from PV_SEND_IPI feature masks out it. People can get confused
> > > if AVIC is enabled while getting lots of hypercall kvm_exits
> > > from IPI.
> > >
> > > In performance, benchmark tool
> > > https://lore.kernel.org/kvm/[email protected]/
> > > shows below results:
> > >
> > > Test env:
> > > CPU: AMD EPYC 7742 64-Core Processor
> > > 2 vCPUs pinned 1:1
> > > idle=poll
> > >
> > > Test result (average ns per IPI of lots of running):
> > > PV_SEND_IPI : 1860
> > > AVIC : 1390
> > >
> > > Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
> > > do have some solid performance test results to this.
> > >
> > > This patch fixes this by masking out PV_SEND_IPI feature when
> > > AVIC is enabled in setting up of guest vCPUs' CPUID.
> > >
> > > Signed-off-by: Kele Huang <[email protected]>
> >
> > AVIC can change across migration. I think we should instead use a new
> > KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
> > The KVM_HINTS_* bits are intended to be changeable across migration,
> > even though we don't have for now anything equivalent to the Hyper-V
> > reenlightenment interrupt.
>
> Note that the same issue exists with HyperV. It also has PV APIC,
> which is harmful when AVIC is enabled (that is guest uses it instead
> of using AVIC, negating AVIC benefits).
>
> Also note that Intel recently posted IPI virtualizaion, which
> will make this issue relevant to APICv too soon.
The recently posted Intel IPI virtualization will accelerate unicast
ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
accelerates broadcast ipis worse than pv ipis. Could we just handle
unicast ipi here?
Wanpeng
On 11/16/21 10:48 AM, Wanpeng Li wrote:
> On Mon, 8 Nov 2021 at 22:09, Maxim Levitsky <[email protected]> wrote:
>>
>> On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
>>> On 11/8/21 10:59, Kele Huang wrote:
>>>> Currently, AVIC is disabled if x2apic feature is exposed to guest
>>>> or in-kernel PIT is in re-injection mode.
>>>>
>>>> We can enable AVIC with options:
>>>>
>>>> Kmod args:
>>>> modprobe kvm_amd avic=1 nested=0 npt=1
>>>> QEMU args:
>>>> ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
>>>>
>>>> When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
>>>> can accelerate IPI operations for guest. However, the relationship
>>>> between AVIC and PV_SEND_IPI feature is not sorted out.
>>>>
>>>> In logical, AVIC accelerates most of frequently IPI operations
>>>> without VMM intervention, while the re-hooking of apic->send_IPI_xxx
>>>> from PV_SEND_IPI feature masks out it. People can get confused
>>>> if AVIC is enabled while getting lots of hypercall kvm_exits
>>>> from IPI.
>>>>
>>>> In performance, benchmark tool
>>>> https://lore.kernel.org/kvm/[email protected]/
>>>> shows below results:
>>>>
>>>> Test env:
>>>> CPU: AMD EPYC 7742 64-Core Processor
>>>> 2 vCPUs pinned 1:1
>>>> idle=poll
>>>>
>>>> Test result (average ns per IPI of lots of running):
>>>> PV_SEND_IPI : 1860
>>>> AVIC : 1390
>>>>
>>>> Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
>>>> do have some solid performance test results to this.
>>>>
>>>> This patch fixes this by masking out PV_SEND_IPI feature when
>>>> AVIC is enabled in setting up of guest vCPUs' CPUID.
>>>>
>>>> Signed-off-by: Kele Huang <[email protected]>
>>>
>>> AVIC can change across migration. I think we should instead use a new
>>> KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
>>> The KVM_HINTS_* bits are intended to be changeable across migration,
>>> even though we don't have for now anything equivalent to the Hyper-V
>>> reenlightenment interrupt.
>>
>> Note that the same issue exists with HyperV. It also has PV APIC,
>> which is harmful when AVIC is enabled (that is guest uses it instead
>> of using AVIC, negating AVIC benefits).
>>
>> Also note that Intel recently posted IPI virtualizaion, which
>> will make this issue relevant to APICv too soon.
>
> The recently posted Intel IPI virtualization will accelerate unicast
> ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
> accelerates broadcast ipis worse than pv ipis. Could we just handle
> unicast ipi here?
>
> Wanpeng
>
Depend on the number of target vCPUs, broadcast IPIs gets unstable
performance on AVIC, and usually worse than PV Send IPI.
So agree with Wanpeng's point, is it possible to separate single IPI and
broadcast IPI on a hardware acceleration platform?
--
zhenwei pi
On Tue, Nov 16, 2021 at 10:56:25AM +0800, zhenwei pi wrote:
>
>
>On 11/16/21 10:48 AM, Wanpeng Li wrote:
>> On Mon, 8 Nov 2021 at 22:09, Maxim Levitsky <[email protected]> wrote:
>> >
>> > On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
>> > > On 11/8/21 10:59, Kele Huang wrote:
>> > > > Currently, AVIC is disabled if x2apic feature is exposed to guest
>> > > > or in-kernel PIT is in re-injection mode.
>> > > >
>> > > > We can enable AVIC with options:
>> > > >
>> > > > Kmod args:
>> > > > modprobe kvm_amd avic=1 nested=0 npt=1
>> > > > QEMU args:
>> > > > ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
>> > > >
>> > > > When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
>> > > > can accelerate IPI operations for guest. However, the relationship
>> > > > between AVIC and PV_SEND_IPI feature is not sorted out.
>> > > >
>> > > > In logical, AVIC accelerates most of frequently IPI operations
>> > > > without VMM intervention, while the re-hooking of apic->send_IPI_xxx
>> > > > from PV_SEND_IPI feature masks out it. People can get confused
>> > > > if AVIC is enabled while getting lots of hypercall kvm_exits
>> > > > from IPI.
>> > > >
>> > > > In performance, benchmark tool
>> > > > https://lore.kernel.org/kvm/[email protected]/
>> > > > shows below results:
>> > > >
>> > > > Test env:
>> > > > CPU: AMD EPYC 7742 64-Core Processor
>> > > > 2 vCPUs pinned 1:1
>> > > > idle=poll
>> > > >
>> > > > Test result (average ns per IPI of lots of running):
>> > > > PV_SEND_IPI : 1860
>> > > > AVIC : 1390
>> > > >
>> > > > Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
>> > > > do have some solid performance test results to this.
>> > > >
>> > > > This patch fixes this by masking out PV_SEND_IPI feature when
>> > > > AVIC is enabled in setting up of guest vCPUs' CPUID.
>> > > >
>> > > > Signed-off-by: Kele Huang <[email protected]>
>> > >
>> > > AVIC can change across migration. I think we should instead use a new
>> > > KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
>> > > The KVM_HINTS_* bits are intended to be changeable across migration,
>> > > even though we don't have for now anything equivalent to the Hyper-V
>> > > reenlightenment interrupt.
>> >
>> > Note that the same issue exists with HyperV. It also has PV APIC,
>> > which is harmful when AVIC is enabled (that is guest uses it instead
>> > of using AVIC, negating AVIC benefits).
>> >
>> > Also note that Intel recently posted IPI virtualizaion, which
>> > will make this issue relevant to APICv too soon.
>>
>> The recently posted Intel IPI virtualization will accelerate unicast
>> ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
>> accelerates broadcast ipis worse than pv ipis. Could we just handle
>> unicast ipi here?
>>
>> Wanpeng
>>
>Depend on the number of target vCPUs, broadcast IPIs gets unstable
>performance on AVIC, and usually worse than PV Send IPI.
>So agree with Wanpeng's point, is it possible to separate single IPI and
>broadcast IPI on a hardware acceleration platform?
Actually, this is how kernel works in x2apic mode: use PV interface
(hypercall) to send multi-cast IPIs and write ICR MSR directly to send
unicast IPIs.
But if guest works in xapic mode, both unicast and multi-cast are issued
via PV interface. It is a side-effect introduced by commit aaffcfd1e82d.
how about just correcting the logic for xapic:
From 13447b221252b64cd85ed1329f7d917afa54efc8 Mon Sep 17 00:00:00 2001
From: Jiaqing Zhao <[email protected]>
Date: Fri, 9 Apr 2021 13:53:39 +0800
Subject: [PATCH 1/2] x86/apic/flat: Add specific send IPI logic
Currently, apic_flat.send_IPI() uses default_send_IPI_single(), which
is a wrapper of apic->send_IPI_mask(). Since commit aaffcfd1e82d
("KVM: X86: Implement PV IPIs in linux guest"), KVM PV IPI driver will
override apic->send_IPI_mask(), and may cause unwated side effects.
This patch removes such side effects by creating a specific send_IPI
method.
Signed-off-by: Jiaqing Zhao <[email protected]>
---
arch/x86/kernel/apic/apic_flat_64.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
index 8f72b4351c9f..3196bf220230 100644
--- a/arch/x86/kernel/apic/apic_flat_64.c
+++ b/arch/x86/kernel/apic/apic_flat_64.c
@@ -64,6 +64,13 @@ static void flat_send_IPI_mask(const struct cpumask *cpumask, int vector)
_flat_send_IPI_mask(mask, vector);
}
+static void flat_send_IPI_single(int cpu, int vector)
+{
+ unsigned long mask = cpumask_bits(cpumask_of(cpu))[0];
+
+ _flat_send_IPI_mask(mask, vector);
+}
+
static void
flat_send_IPI_mask_allbutself(const struct cpumask *cpumask, int vector)
{
@@ -132,7 +139,7 @@ static struct apic apic_flat __ro_after_init = {
.calc_dest_apicid = apic_flat_calc_apicid,
- .send_IPI = default_send_IPI_single,
+ .send_IPI = flat_send_IPI_single,
.send_IPI_mask = flat_send_IPI_mask,
.send_IPI_mask_allbutself = flat_send_IPI_mask_allbutself,
.send_IPI_allbutself = default_send_IPI_allbutself,
--
2.27.0
> The recently posted Intel IPI virtualization will accelerate unicast
> ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
> accelerates broadcast ipis worse than pv ipis. Could we just handle
> unicast ipi here?
Thanks for the explanation! It is true that AVIC does not always perform better
than PV IPI, actually not even swx2apic.
> So agree with Wanpeng's point, is it possible to separate single IPI and
> broadcast IPI on a hardware acceleration platform?
> how about just correcting the logic for xapic:
> From 13447b221252b64cd85ed1329f7d917afa54efc8 Mon Sep 17 00:00:00 2001
> From: Jiaqing Zhao <[email protected]>
> Date: Fri, 9 Apr 2021 13:53:39 +0800
> Subject: [PATCH 1/2] x86/apic/flat: Add specific send IPI logic
> Currently, apic_flat.send_IPI() uses default_send_IPI_single(), which
> is a wrapper of apic->send_IPI_mask(). Since commit aaffcfd1e82d
> ("KVM: X86: Implement PV IPIs in linux guest"), KVM PV IPI driver will
> override apic->send_IPI_mask(), and may cause unwated side effects.
> This patch removes such side effects by creating a specific send_IPI
> method.
> Signed-off-by: Jiaqing Zhao <[email protected]>
Actually, I think this issue is more about how to sort out the relationship
between AVIC and PV IPI. As far as I understand, currently, no matter
the option from userspace or the determination made in kernel works
in some way, but not in the migration scenario. For instance, migration with
AVIC feature changes can make guests lose the PV IPI feature needlessly.
Besides, the current patch is not consistent with
KVM_CAP_ENFORCE_PV_FEATURE_CPUID.
Paolo's advice about using a new hint shall work well. Currently try
working on it.
Best regards,
Kele
On Tue, Nov 16, 2021 at 4:56 PM Chao Gao <[email protected]> wrote:
>
> On Tue, Nov 16, 2021 at 10:56:25AM +0800, zhenwei pi wrote:
> >
> >
> >On 11/16/21 10:48 AM, Wanpeng Li wrote:
> >> On Mon, 8 Nov 2021 at 22:09, Maxim Levitsky <[email protected]> wrote:
> >> >
> >> > On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
> >> > > On 11/8/21 10:59, Kele Huang wrote:
> >> > > > Currently, AVIC is disabled if x2apic feature is exposed to guest
> >> > > > or in-kernel PIT is in re-injection mode.
> >> > > >
> >> > > > We can enable AVIC with options:
> >> > > >
> >> > > > Kmod args:
> >> > > > modprobe kvm_amd avic=1 nested=0 npt=1
> >> > > > QEMU args:
> >> > > > ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
> >> > > >
> >> > > > When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
> >> > > > can accelerate IPI operations for guest. However, the relationship
> >> > > > between AVIC and PV_SEND_IPI feature is not sorted out.
> >> > > >
> >> > > > In logical, AVIC accelerates most of frequently IPI operations
> >> > > > without VMM intervention, while the re-hooking of apic->send_IPI_xxx
> >> > > > from PV_SEND_IPI feature masks out it. People can get confused
> >> > > > if AVIC is enabled while getting lots of hypercall kvm_exits
> >> > > > from IPI.
> >> > > >
> >> > > > In performance, benchmark tool
> >> > > > https://lore.kernel.org/kvm/[email protected]/
> >> > > > shows below results:
> >> > > >
> >> > > > Test env:
> >> > > > CPU: AMD EPYC 7742 64-Core Processor
> >> > > > 2 vCPUs pinned 1:1
> >> > > > idle=poll
> >> > > >
> >> > > > Test result (average ns per IPI of lots of running):
> >> > > > PV_SEND_IPI : 1860
> >> > > > AVIC : 1390
> >> > > >
> >> > > > Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
> >> > > > do have some solid performance test results to this.
> >> > > >
> >> > > > This patch fixes this by masking out PV_SEND_IPI feature when
> >> > > > AVIC is enabled in setting up of guest vCPUs' CPUID.
> >> > > >
> >> > > > Signed-off-by: Kele Huang <[email protected]>
> >> > >
> >> > > AVIC can change across migration. I think we should instead use a new
> >> > > KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
> >> > > The KVM_HINTS_* bits are intended to be changeable across migration,
> >> > > even though we don't have for now anything equivalent to the Hyper-V
> >> > > reenlightenment interrupt.
> >> >
> >> > Note that the same issue exists with HyperV. It also has PV APIC,
> >> > which is harmful when AVIC is enabled (that is guest uses it instead
> >> > of using AVIC, negating AVIC benefits).
> >> >
> >> > Also note that Intel recently posted IPI virtualizaion, which
> >> > will make this issue relevant to APICv too soon.
> >>
> >> The recently posted Intel IPI virtualization will accelerate unicast
> >> ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
> >> accelerates broadcast ipis worse than pv ipis. Could we just handle
> >> unicast ipi here?
> >>
> >> Wanpeng
> >>
> >Depend on the number of target vCPUs, broadcast IPIs gets unstable
> >performance on AVIC, and usually worse than PV Send IPI.
> >So agree with Wanpeng's point, is it possible to separate single IPI and
> >broadcast IPI on a hardware acceleration platform?
>
> Actually, this is how kernel works in x2apic mode: use PV interface
> (hypercall) to send multi-cast IPIs and write ICR MSR directly to send
> unicast IPIs.
>
> But if guest works in xapic mode, both unicast and multi-cast are issued
> via PV interface. It is a side-effect introduced by commit aaffcfd1e82d.
>
> how about just correcting the logic for xapic:
>
> From 13447b221252b64cd85ed1329f7d917afa54efc8 Mon Sep 17 00:00:00 2001
> From: Jiaqing Zhao <[email protected]>
> Date: Fri, 9 Apr 2021 13:53:39 +0800
> Subject: [PATCH 1/2] x86/apic/flat: Add specific send IPI logic
>
> Currently, apic_flat.send_IPI() uses default_send_IPI_single(), which
> is a wrapper of apic->send_IPI_mask(). Since commit aaffcfd1e82d
> ("KVM: X86: Implement PV IPIs in linux guest"), KVM PV IPI driver will
> override apic->send_IPI_mask(), and may cause unwated side effects.
>
> This patch removes such side effects by creating a specific send_IPI
> method.
>
> Signed-off-by: Jiaqing Zhao <[email protected]>
> ---
> arch/x86/kernel/apic/apic_flat_64.c | 9 ++++++++-
> 1 file changed, 8 insertions(+), 1 deletion(-)
>
> diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c
> index 8f72b4351c9f..3196bf220230 100644
> --- a/arch/x86/kernel/apic/apic_flat_64.c
> +++ b/arch/x86/kernel/apic/apic_flat_64.c
> @@ -64,6 +64,13 @@ static void flat_send_IPI_mask(const struct cpumask *cpumask, int vector)
> _flat_send_IPI_mask(mask, vector);
> }
>
> +static void flat_send_IPI_single(int cpu, int vector)
> +{
> + unsigned long mask = cpumask_bits(cpumask_of(cpu))[0];
> +
> + _flat_send_IPI_mask(mask, vector);
> +}
> +
> static void
> flat_send_IPI_mask_allbutself(const struct cpumask *cpumask, int vector)
> {
> @@ -132,7 +139,7 @@ static struct apic apic_flat __ro_after_init = {
>
> .calc_dest_apicid = apic_flat_calc_apicid,
>
> - .send_IPI = default_send_IPI_single,
> + .send_IPI = flat_send_IPI_single,
> .send_IPI_mask = flat_send_IPI_mask,
> .send_IPI_mask_allbutself = flat_send_IPI_mask_allbutself,
> .send_IPI_allbutself = default_send_IPI_allbutself,
> --
> 2.27.0
>
On Tue, 16 Nov 2021 at 16:56, Chao Gao <[email protected]> wrote:
>
> On Tue, Nov 16, 2021 at 10:56:25AM +0800, zhenwei pi wrote:
> >
> >
> >On 11/16/21 10:48 AM, Wanpeng Li wrote:
> >> On Mon, 8 Nov 2021 at 22:09, Maxim Levitsky <[email protected]> wrote:
> >> >
> >> > On Mon, 2021-11-08 at 11:30 +0100, Paolo Bonzini wrote:
> >> > > On 11/8/21 10:59, Kele Huang wrote:
> >> > > > Currently, AVIC is disabled if x2apic feature is exposed to guest
> >> > > > or in-kernel PIT is in re-injection mode.
> >> > > >
> >> > > > We can enable AVIC with options:
> >> > > >
> >> > > > Kmod args:
> >> > > > modprobe kvm_amd avic=1 nested=0 npt=1
> >> > > > QEMU args:
> >> > > > ... -cpu host,-x2apic -global kvm-pit.lost_tick_policy=discard ...
> >> > > >
> >> > > > When LAPIC works in xapic mode, both AVIC and PV_SEND_IPI feature
> >> > > > can accelerate IPI operations for guest. However, the relationship
> >> > > > between AVIC and PV_SEND_IPI feature is not sorted out.
> >> > > >
> >> > > > In logical, AVIC accelerates most of frequently IPI operations
> >> > > > without VMM intervention, while the re-hooking of apic->send_IPI_xxx
> >> > > > from PV_SEND_IPI feature masks out it. People can get confused
> >> > > > if AVIC is enabled while getting lots of hypercall kvm_exits
> >> > > > from IPI.
> >> > > >
> >> > > > In performance, benchmark tool
> >> > > > https://lore.kernel.org/kvm/[email protected]/
> >> > > > shows below results:
> >> > > >
> >> > > > Test env:
> >> > > > CPU: AMD EPYC 7742 64-Core Processor
> >> > > > 2 vCPUs pinned 1:1
> >> > > > idle=poll
> >> > > >
> >> > > > Test result (average ns per IPI of lots of running):
> >> > > > PV_SEND_IPI : 1860
> >> > > > AVIC : 1390
> >> > > >
> >> > > > Besides, disscussions in https://lkml.org/lkml/2021/10/20/423
> >> > > > do have some solid performance test results to this.
> >> > > >
> >> > > > This patch fixes this by masking out PV_SEND_IPI feature when
> >> > > > AVIC is enabled in setting up of guest vCPUs' CPUID.
> >> > > >
> >> > > > Signed-off-by: Kele Huang <[email protected]>
> >> > >
> >> > > AVIC can change across migration. I think we should instead use a new
> >> > > KVM_HINTS_* bit (KVM_HINTS_ACCELERATED_LAPIC or something like that).
> >> > > The KVM_HINTS_* bits are intended to be changeable across migration,
> >> > > even though we don't have for now anything equivalent to the Hyper-V
> >> > > reenlightenment interrupt.
> >> >
> >> > Note that the same issue exists with HyperV. It also has PV APIC,
> >> > which is harmful when AVIC is enabled (that is guest uses it instead
> >> > of using AVIC, negating AVIC benefits).
> >> >
> >> > Also note that Intel recently posted IPI virtualizaion, which
> >> > will make this issue relevant to APICv too soon.
> >>
> >> The recently posted Intel IPI virtualization will accelerate unicast
> >> ipi but not broadcast ipis, AMD AVIC accelerates unicast ipi well but
> >> accelerates broadcast ipis worse than pv ipis. Could we just handle
> >> unicast ipi here?
> >>
> >> Wanpeng
> >>
> >Depend on the number of target vCPUs, broadcast IPIs gets unstable
> >performance on AVIC, and usually worse than PV Send IPI.
> >So agree with Wanpeng's point, is it possible to separate single IPI and
> >broadcast IPI on a hardware acceleration platform?
>
> Actually, this is how kernel works in x2apic mode: use PV interface
> (hypercall) to send multi-cast IPIs and write ICR MSR directly to send
> unicast IPIs.
>
> But if guest works in xapic mode, both unicast and multi-cast are issued
> via PV interface. It is a side-effect introduced by commit aaffcfd1e82d.
>
> how about just correcting the logic for xapic:
>
> From 13447b221252b64cd85ed1329f7d917afa54efc8 Mon Sep 17 00:00:00 2001
> From: Jiaqing Zhao <[email protected]>
> Date: Fri, 9 Apr 2021 13:53:39 +0800
> Subject: [PATCH 1/2] x86/apic/flat: Add specific send IPI logic
>
> Currently, apic_flat.send_IPI() uses default_send_IPI_single(), which
> is a wrapper of apic->send_IPI_mask(). Since commit aaffcfd1e82d
> ("KVM: X86: Implement PV IPIs in linux guest"), KVM PV IPI driver will
> override apic->send_IPI_mask(), and may cause unwated side effects.
>
> This patch removes such side effects by creating a specific send_IPI
> method.
This looks reasonable to me.
Wanpeng