[PATCH] KVM: x86: get smi pending status correctly

[PATCH] KVM: x86: get smi pending status correctly

[Jay Zhou]

The injection process of smi has two steps:

    Qemu                        KVM
Step1:
    cpu->interrupt_request &= \
        ~CPU_INTERRUPT_SMI;
    kvm_vcpu_ioctl(cpu, KVM_SMI)

                                call kvm_vcpu_ioctl_smi() and
                                kvm_make_request(KVM_REQ_SMI, vcpu);

Step2:
    kvm_vcpu_ioctl(cpu, KVM_RUN, 0)

                                call process_smi() if
                                kvm_check_request(KVM_REQ_SMI, vcpu) is
                                true, mark vcpu->arch.smi_pending = true;

The vcpu->arch.smi_pending will be set true in step2, unfortunately if
vcpu paused between step1 and step2, the kvm_run->immediate_exit will be
set and vcpu has to exit to Qemu immediately during step2 before mark
vcpu->arch.smi_pending true.
During VM migration, Qemu will get the smi pending status from KVM using
KVM_GET_VCPU_EVENTS ioctl at the downtime, then the smi pending status
will be lost.

Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
Signed-off-by: Shengen Zhuang <zhuangshengen@huawei.com>
---
 arch/x86/kvm/x86.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9a8969a..9025c76 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -105,6 +105,7 @@
 
 static void update_cr8_intercept(struct kvm_vcpu *vcpu);
 static void process_nmi(struct kvm_vcpu *vcpu);
+static void process_smi(struct kvm_vcpu *vcpu);
 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);
@@ -4230,6 +4231,9 @@ static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu,
 {
 	process_nmi(vcpu);
 
+	if (kvm_check_request(KVM_REQ_SMI, vcpu))
+		process_smi(vcpu);
+
 	/*
 	 * In guest mode, payload delivery should be deferred,
 	 * so that the L1 hypervisor can intercept #PF before
-- 
1.8.3.1

Re: [PATCH] KVM: x86: get smi pending status correctly

[Paolo Bonzini]

On 18/01/21 09:47, Jay Zhou wrote:
> The injection process of smi has two steps:
> 
>      Qemu                        KVM
> Step1:
>      cpu->interrupt_request &= \
>          ~CPU_INTERRUPT_SMI;
>      kvm_vcpu_ioctl(cpu, KVM_SMI)
> 
>                                  call kvm_vcpu_ioctl_smi() and
>                                  kvm_make_request(KVM_REQ_SMI, vcpu);
> 
> Step2:
>      kvm_vcpu_ioctl(cpu, KVM_RUN, 0)
> 
>                                  call process_smi() if
>                                  kvm_check_request(KVM_REQ_SMI, vcpu) is
>                                  true, mark vcpu->arch.smi_pending = true;
> 
> The vcpu->arch.smi_pending will be set true in step2, unfortunately if
> vcpu paused between step1 and step2, the kvm_run->immediate_exit will be
> set and vcpu has to exit to Qemu immediately during step2 before mark
> vcpu->arch.smi_pending true.
> During VM migration, Qemu will get the smi pending status from KVM using
> KVM_GET_VCPU_EVENTS ioctl at the downtime, then the smi pending status
> will be lost.
> 
> Signed-off-by: Jay Zhou <jianjay.zhou@huawei.com>
> Signed-off-by: Shengen Zhuang <zhuangshengen@huawei.com>
> ---
>   arch/x86/kvm/x86.c | 4 ++++
>   1 file changed, 4 insertions(+)
> 
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 9a8969a..9025c76 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -105,6 +105,7 @@
>   
>   static void update_cr8_intercept(struct kvm_vcpu *vcpu);
>   static void process_nmi(struct kvm_vcpu *vcpu);
> +static void process_smi(struct kvm_vcpu *vcpu);
>   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);
> @@ -4230,6 +4231,9 @@ static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu,
>   {
>   	process_nmi(vcpu);
>   
> +	if (kvm_check_request(KVM_REQ_SMI, vcpu))
> +		process_smi(vcpu);
> +
>   	/*
>   	 * In guest mode, payload delivery should be deferred,
>   	 * so that the L1 hypervisor can intercept #PF before
> 

Queued, thanks.

Paolo