Re: [Qemu-devel] [PATCH RFC] virtio: put last seen used index into ring itself

[Michael S. Tsirkin]

On Tue, May 11, 2010 at 10:27:22PM +0300, Avi Kivity wrote:
> On 05/07/2010 06:23 AM, Rusty Russell wrote:
>> On Thu, 6 May 2010 07:30:00 pm Avi Kivity wrote:
>>    
>>> On 05/05/2010 11:58 PM, Michael S. Tsirkin wrote:
>>>      
>>>> +  /* We publish the last-seen used index at the end of the available ring.
>>>> +   * It is at the end for backwards compatibility. */
>>>> +  vr->last_used_idx =&(vr)->avail->ring[num];
>>>> +  /* Verify that last used index does not spill over the used ring. */
>>>> +  BUG_ON((void *)vr->last_used_idx +
>>>> +         sizeof *vr->last_used_idx>   (void *)vr->used);
>>>>    }
>>>>
>>>>        
>>> Shouldn't this be on its own cache line?
>>>      
>> It's next to the available ring; because that's where the guest publishes
>> its data.  That whole page is guest-write, host-read.
>>
>> Putting it on a cacheline by itself would be a slight pessimization; the host
>> cpu would have to get the last_used_idx cacheline and the avail descriptor
>> cacheline every time.  This way, they are sometimes the same cacheline.
>>    
>
> If one peer writes the tail of the available ring, while the other reads  
> last_used_idx, it's a false bounce, no?
>
> Having things on the same cacheline is only worthwhile if they are  
> accessed at the same time.

Yes, this is what I was trying to say.
avail flags and used index *are* accessed at the same time, so
there could be an advantage to sharing a cache line there.

All this should be kept in mind if we ever do
VIRTIO_RING_F_NEW_LAYOUT.

-- 
MST