[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <2f61ae5c-0658-e5c9-754c-9ca80148a54d@amd.com>
Date: Mon, 4 Nov 2019 18:45:48 +0000
From: "Moger, Babu" <Babu.Moger@....com>
To: Paolo Bonzini <pbonzini@...hat.com>,
"tglx@...utronix.de" <tglx@...utronix.de>,
"mingo@...hat.com" <mingo@...hat.com>,
"bp@...en8.de" <bp@...en8.de>, "hpa@...or.com" <hpa@...or.com>,
"rkrcmar@...hat.com" <rkrcmar@...hat.com>,
"sean.j.christopherson@...el.com" <sean.j.christopherson@...el.com>,
"vkuznets@...hat.com" <vkuznets@...hat.com>,
"wanpengli@...cent.com" <wanpengli@...cent.com>,
"jmattson@...gle.com" <jmattson@...gle.com>
CC: "x86@...nel.org" <x86@...nel.org>,
"joro@...tes.org" <joro@...tes.org>,
"luto@...nel.org" <luto@...nel.org>,
"zohar@...ux.ibm.com" <zohar@...ux.ibm.com>,
"yamada.masahiro@...ionext.com" <yamada.masahiro@...ionext.com>,
"nayna@...ux.ibm.com" <nayna@...ux.ibm.com>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
"kvm@...r.kernel.org" <kvm@...r.kernel.org>
Subject: Re: [PATCH 2/4] kvm: svm: Enable UMIP feature on AMD
On 11/4/19 5:54 AM, Paolo Bonzini wrote:
> On 01/11/19 18:33, Moger, Babu wrote:
>> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
>> index 4153ca8cddb7..79abbdeca148 100644
>> --- a/arch/x86/kvm/svm.c
>> +++ b/arch/x86/kvm/svm.c
>> @@ -2533,6 +2533,11 @@ static void svm_decache_cr4_guest_bits(struct kvm_vcpu *vcpu)
>> {
>> }
>>
>> +static bool svm_umip_emulated(void)
>> +{
>> + return boot_cpu_has(X86_FEATURE_UMIP);
>> +}
>
> For hardware that supports UMIP, this is only needed because of your
> patch 1. Without it, X86_FEATURE_UMIP should already be enabled on
> processors that natively support UMIP.
Yes, That is correct. Will remove the patch #1. Intention was to enable
UMIP for the hardware that supports it. Will send out only the config
changes(Patch #4). Also there is a complexity with supporting emulation
on SEV guest.
>
> If you want UMIP *emulation* instead, this should become "return true".
>
>> static void update_cr0_intercept(struct vcpu_svm *svm)
>> {
>> ulong gcr0 = svm->vcpu.arch.cr0;
>> @@ -4438,6 +4443,13 @@ static int interrupt_window_interception(struct vcpu_svm *svm)
>> return 1;
>> }
>>
>> +static int umip_interception(struct vcpu_svm *svm)
>> +{
>> + struct kvm_vcpu *vcpu = &svm->vcpu;
>> +
>> + return kvm_emulate_instruction(vcpu, 0);
>> +}
>> +
>> static int pause_interception(struct vcpu_svm *svm)
>> {
>> struct kvm_vcpu *vcpu = &svm->vcpu;
>> @@ -4775,6 +4787,10 @@ static int (*const svm_exit_handlers[])(struct vcpu_svm *svm) = {
>> [SVM_EXIT_SMI] = nop_on_interception,
>> [SVM_EXIT_INIT] = nop_on_interception,
>> [SVM_EXIT_VINTR] = interrupt_window_interception,
>> + [SVM_EXIT_IDTR_READ] = umip_interception,
>> + [SVM_EXIT_GDTR_READ] = umip_interception,
>> + [SVM_EXIT_LDTR_READ] = umip_interception,
>> + [SVM_EXIT_TR_READ] = umip_interception,
>
> This is missing enabling the intercepts. Also, this can be just
> emulate_on_interception instead of a new function.
>
> Paolo
>
>> [SVM_EXIT_RDPMC] = rdpmc_interception,
>> [SVM_EXIT_CPUID] = cpuid_interception,
>> [SVM_EXIT_IRET] = iret_interception,
>> @@ -5976,11 +5992,6 @@ static bool svm_xsaves_supported(void)
>> return boot_cpu_has(X86_FEATURE_XSAVES);
>> }
>>
>> -static bool svm_umip_emulated(void)
>> -{
>> - return false;
>> -}
>> -
>> static bool svm_pt_supported(void)
>> {
>> return false;
>>
>
Powered by blists - more mailing lists