Re: [PATCHv2] kvm: remove in_range and switch to rwsem for iobus

["Michael S. Tsirkin" <mst@redhat.com>]

On Mon, Jun 29, 2009 at 12:44:53PM +0300, Avi Kivity wrote:
> On 06/29/2009 12:23 PM, Michael S. Tsirkin wrote:
>> On Mon, Jun 29, 2009 at 11:37:00AM +0300, Avi Kivity wrote:
>>    
>>> On 06/28/2009 10:34 PM, Michael S. Tsirkin wrote:
>>>      
>>>> This changes bus accesses to use high-level kvm_io_bus_read/kvm_io_bus_write
>>>> functions, which utilize read/write semaphore intead of mutex.  in_range now
>>>> becomes unused so it is removed from device ops in favor of read/write
>>>> callbacks performing range checks internally.
>>>>
>>>> This allows aliasing (mostly for in-kernel virtio), as well as better error
>>>> handling by making it possible to pass errors up to userspace. And it's enough
>>>> to look at the diffstat to see that it's a better API anyway.
>>>>
>>>> While we are at it, document locking rules for kvm_io_device_ops.
>>>>
>>>> Note: since the use of the new bus_lock is localized to a small number of
>>>> places, it will be easy to switch to srcu in the future if we so desire.
>>>>
>>>>        
>>> Looks good. But please split into a locking change patch and an API
>>> change patch (in whatever order makes more sense).
>>>
>>> I think you can reuse slots_lock instead of adding a new lock. IIRC
>>> slots_lock is already taken for read everywhere, so you only need to
>>> take it for write when registering things.
>>>      
>>
>> IMO this will make it harder to convert to rcu down the line.
>> As it is we just grep for bus_lock and replace with rcu.
>> While possibly slots_lock can be converted to rcu as well,
>> let's do it one thing at a time.
>>    
>
> We can convert it to rcu indepenently of other things protected by  
> slots_lock; no need to do everything at the same time.

Yes but once we merge locks, it will be harder to split them out.
I know I can now do grep bus_lock and find all places affected,
if I reuse slot_lock this information is lost. No?

-- 
MST