On 11/6/19 9:43 AM, Sean Christopherson wrote:
> On Wed, Nov 06, 2019 at 08:36:37AM -0500, Nitesh Narayan Lal wrote:
>> From: Radim Krčmář <rkrcmar@xxxxxxxxxx>
>>
>> Fetching an index for any vcpu in kvm->vcpus array by traversing
>> the entire array everytime is costly.
>> This patch remembers the position of each vcpu in kvm->vcpus array
>> by storing it in vcpus_idx under kvm_vcpu structure.
>>
>> Signed-off-by: Radim Krčmář <rkrcmar@xxxxxxxxxx>
>> Signed-off-by: Nitesh Narayan Lal <nitesh@xxxxxxxxxx>
>> ---
>> include/linux/kvm_host.h | 11 +++--------
>> virt/kvm/kvm_main.c | 5 ++++-
>> 2 files changed, 7 insertions(+), 9 deletions(-)
>>
>> diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
>> index 719fc3e..31c4fde 100644
>> --- a/include/linux/kvm_host.h
>> +++ b/include/linux/kvm_host.h
>> @@ -266,7 +266,8 @@ struct kvm_vcpu {
>> struct preempt_notifier preempt_notifier;
>> #endif
>> int cpu;
>> - int vcpu_id;
>> + int vcpu_id; /* id given by userspace at creation */
>> + int vcpus_idx; /* index in kvm->vcpus array */
> I'd probably prefer vcpu_idx or vcpu_index, but it's not a strong
> preference by any means.
Sure, I will probably replace it with vcpu_idx.
>
>> int srcu_idx;
>> int mode;
>> u64 requests;
>> @@ -571,13 +572,7 @@ static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id)
>>
>> static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
>> {
>> - struct kvm_vcpu *tmp;
>> - int idx;
>> -
>> - kvm_for_each_vcpu(idx, tmp, vcpu->kvm)
>> - if (tmp == vcpu)
>> - return idx;
>> - BUG();
>> + return vcpu->vcpus_idx;
>> }
>>
>> #define kvm_for_each_memslot(memslot, slots) \
>> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
>> index 67ef3f2..24ab711 100644
>> --- a/virt/kvm/kvm_main.c
>> +++ b/virt/kvm/kvm_main.c
>> @@ -2673,7 +2673,10 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, u32 id)
>> goto unlock_vcpu_destroy;
>> }
>>
>> - BUG_ON(kvm->vcpus[atomic_read(&kvm->online_vcpus)]);
>> + vcpu->vcpus_idx = atomic_read(&kvm->online_vcpus);
>> +
> Nit: I'd omit this newline since the assignment and BUG_ON() are directly
> related.
Makes sense to me.
>
>> + BUG_ON(kvm->vcpus[vcpu->vcpus_idx]);
> The assignment to kvm->vcpus a few lines below should be updated to use
> the new index.
Ah yes.
Thanks for pointing this out.
>
>> +
>>
>> /* Now it's all set up, let userspace reach it */
>> kvm_get_kvm(kvm);
>> --
>> 1.8.3.1
>>
--
Nitesh
Attachment:
signature.asc
Description: OpenPGP digital signature
