Re: [PATCH 1/13] KVM: Add tlb_remote_flush_with_range callback in kvm_x86_ops

[Tianyu Lan <Tianyu.Lan@microsoft.com>]

Hi Sean:
	Thanks for your review.

On 9/10/2018 10:21 PM, Sean Christopherson wrote:
> On Mon, 2018-09-10 at 08:38 +0000, Tianyu Lan wrote:
>> Add flush range call back in the kvm_x86_ops and platform can use it
>> to register its associated function. The parameter "kvm_tlb_range"
>> accepts a single range and flush list which contains a list of ranges.
>>
>> Signed-off-by: Lan Tianyu <Tianyu.Lan@microsoft.com>
>> ---
>>   arch/x86/include/asm/kvm_host.h | 8 ++++++++
>>   1 file changed, 8 insertions(+)
>>
>> diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
>> index e12916e7c2fb..dcdf8cc16388 100644
>> --- a/arch/x86/include/asm/kvm_host.h
>> +++ b/arch/x86/include/asm/kvm_host.h
>> @@ -402,6 +402,12 @@ struct kvm_mmu {
>>   	u64 pdptrs[4]; /* pae */
>>   };
>>   
>> +struct kvm_tlb_range {
>> +	u64 start_gfn;
>> +	u64 end_gfn;
> 
> IMO this struct and all functions should pass around the number of pages
> instead of end_gfn to avoid confusion as to whether end_gfn is inclusive
> or exlusive.

That's good suggestion. Thanks.

> 
>> +	struct list_head *flush_list;
>> +};
>> +
>>   enum pmc_type {
>>   	KVM_PMC_GP = 0,
>>   	KVM_PMC_FIXED,
>> @@ -991,6 +997,8 @@ struct kvm_x86_ops {
>>   
>>   	void (*tlb_flush)(struct kvm_vcpu *vcpu, bool invalidate_gpa);
>>   	int  (*tlb_remote_flush)(struct kvm *kvm);
>> +	int  (*tlb_remote_flush_with_range)(struct kvm *kvm,
>> +			struct kvm_tlb_range *range);
>>   
>>   	/*
>>   	 * Flush any TLB entries associated with the given GVA.