Re: [PATCH v1 3/3] kvm: svm: Use the hardware provided GPA instead of page walk

[Tom Lendacky <thomas.lendacky@amd.com>]

On 11/21/2016 9:07 AM, Paolo Bonzini wrote:
> 
> 
> On 14/11/2016 23:16, Brijesh Singh wrote:
>> From: Tom Lendacky <thomas.lendacky@amd.com>
>>
>> When a guest causes a NPF which requires emulation, KVM sometimes walks
>> the guest page tables to translate the GVA to a GPA. This is unnecessary
>> most of the time on AMD hardware since the hardware provides the GPA in
>> EXITINFO2.
>>
>> The only exception cases involve string operations involving rep or
>> operations that use two memory locations. With rep, the GPA will only be
>> the value of the initial NPF and with dual memory locations we won't know
>> which memory address was translated into EXITINFO2.
>>
>> Signed-off-by: Tom Lendacky <thomas.lendacky@amd.com>
>> Reviewed-by: Borislav Petkov <bp@suse.de>
>> Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
>> ---
>>  arch/x86/include/asm/kvm_emulate.h |    3 +++
>>  arch/x86/include/asm/kvm_host.h    |    3 +++
>>  arch/x86/kvm/svm.c                 |    9 ++++++++-
>>  arch/x86/kvm/x86.c                 |   17 ++++++++++++++++-
>>  4 files changed, 30 insertions(+), 2 deletions(-)
>>
>> diff --git a/arch/x86/include/asm/kvm_emulate.h b/arch/x86/include/asm/kvm_emulate.h
>> index e9cd7be..2d1ac09 100644
>> --- a/arch/x86/include/asm/kvm_emulate.h
>> +++ b/arch/x86/include/asm/kvm_emulate.h
>> @@ -344,6 +344,9 @@ struct x86_emulate_ctxt {
>>  	struct read_cache mem_read;
>>  };
>>  
>> +/* String operation identifier (matches the definition in emulate.c) */
>> +#define CTXT_STRING_OP	(1 << 13)
>> +
>>  /* Repeat String Operation Prefix */
>>  #define REPE_PREFIX	0xf3
>>  #define REPNE_PREFIX	0xf2
>> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
>> index 77cb3f9..fd5b1c8 100644
>> --- a/arch/x86/include/asm/kvm_host.h
>> +++ b/arch/x86/include/asm/kvm_host.h
>> @@ -668,6 +668,9 @@ struct kvm_vcpu_arch {
>>  
>>  	int pending_ioapic_eoi;
>>  	int pending_external_vector;
>> +
>> +	/* GPA available (AMD only) */
>> +	bool gpa_available;
>>  };
>>  
>>  struct kvm_lpage_info {
>> diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
>> index 5e64e656..b442c5a 100644
>> --- a/arch/x86/kvm/svm.c
>> +++ b/arch/x86/kvm/svm.c
>> @@ -275,6 +275,9 @@ static int avic;
>>  module_param(avic, int, S_IRUGO);
>>  #endif
>>  
>> +/* EXITINFO2 contains valid GPA */
>> +static bool gpa_avail = true;
>> +
>>  /* AVIC VM ID bit masks and lock */
>>  static DECLARE_BITMAP(avic_vm_id_bitmap, AVIC_VM_ID_NR);
>>  static DEFINE_SPINLOCK(avic_vm_id_lock);
>> @@ -1055,8 +1058,10 @@ static __init int svm_hardware_setup(void)
>>  			goto err;
>>  	}
>>  
>> -	if (!boot_cpu_has(X86_FEATURE_NPT))
>> +	if (!boot_cpu_has(X86_FEATURE_NPT)) {
>>  		npt_enabled = false;
>> +		gpa_avail = false;
> 
> This is not necessary, since you will never have exit_code ==
> SVM_EXIT_NPF && !gpa_avail.

Ah, yes, that is a bit redundant. We'll get rid of the gpa_avail
flag.

Thanks,
Tom

> 
>> +	}
>>  
>>  	if (npt_enabled && !npt) {
>>  		printk(KERN_INFO "kvm: Nested Paging disabled\n");
>> @@ -4192,6 +4197,8 @@ static int handle_exit(struct kvm_vcpu *vcpu)
>>  		vcpu->arch.cr0 = svm->vmcb->save.cr0;
>>  	if (npt_enabled)
>>  		vcpu->arch.cr3 = svm->vmcb->save.cr3;
>> +	if (gpa_avail)
>> +		vcpu->arch.gpa_available = (exit_code == SVM_EXIT_NPF);
> 
> The "if", and the body moved just before
> 
> 	return svm_exit_handlers[exit_code](svm);
> 
> I'll try doing a similar patch for Intel too, for better testing.
> 
>>  
>>  	if (unlikely(svm->nested.exit_required)) {
>>  		nested_svm_vmexit(svm);
>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>> index d02aeff..c290794 100644
>> --- a/arch/x86/kvm/x86.c
>> +++ b/arch/x86/kvm/x86.c
>> @@ -4420,7 +4420,19 @@ static int vcpu_mmio_gva_to_gpa(struct kvm_vcpu *vcpu, unsigned long gva,
>>  		return 1;
>>  	}
>>  
>> -	*gpa = vcpu->arch.walk_mmu->gva_to_gpa(vcpu, gva, access, exception);
>> +	/*
>> +	 * If the exit was due to a NPF we may already have a GPA.
>> +	 * If the GPA is present, use it to avoid the GVA to GPA table
>> +	 * walk. Note, this cannot be used on string operations since
>> +	 * string operation using rep will only have the initial GPA
>> +	 * from when the NPF occurred.
>> +	 */
>> +	if (vcpu->arch.gpa_available &&
>> +	    !(vcpu->arch.emulate_ctxt.d & CTXT_STRING_OP))
>> +		*gpa = exception->address;
>> +	else
>> +		*gpa = vcpu->arch.walk_mmu->gva_to_gpa(vcpu, gva, access,
>> +						       exception);
>>  
>>  	if (*gpa == UNMAPPED_GVA)
>>  		return -1;
>> @@ -5542,6 +5554,9 @@ int x86_emulate_instruction(struct kvm_vcpu *vcpu,
>>  	}
>>  
>>  restart:
>> +	/* Save the faulting GPA (cr2) in the address field */
>> +	ctxt->exception.address = cr2;
>> +
>>  	r = x86_emulate_insn(ctxt);
>>  
>>  	if (r == EMULATION_INTERCEPTED)
>>