On Tue, 2015-08-25 at 17:27 -0700, Paul E. McKenney wrote:

On Thu, Aug 20, 2015 at 04:56:04PM +0100, Will Deacon wrote:

quoted

On Thu, Aug 20, 2015 at 10:45:05AM +0100, Michael Ellerman wrote:

quoted

On Tue, 2015-08-18 at 09:37 +0100, Will Deacon wrote:

quoted

Thanks, that sounds great. FWIW, there are multiple ways of implementing
the patch (i.e. whether you strengthen lock or unlock). I had a crack at
something here, but it's not tested:

  http://marc.info/?l=linux-arch&m=143758379023849&w=2

I notice you are not changing PPC_RELEASE_BARRIER, but only the spin unlock
code. But from my reading of the docs we need to make sure any UNLOCK+LOCK is a
full barrier, not just spin unlock/lock?

So don't we need to worry about some of the other locks as well? At least
rwlock, and mutex fast path?

Hmm, that's a good question. I notice that you don't do any of the SYNC_IO
stuff for any locks other than spinlocks but I don't know whether
smp_mb__after_unlock_lock is similarly limited in scope.

Paul?

I would expect the various locks to have similar ordering characteristics.

Or am I missing something subtle here?

I don't think so.

The docs just talk about ACQUIRE/RELEASE, so I think it needs to apply to all
lock types. Or at least the list mentioned in the docs which is:

 (*) spin locks
 (*) R/W spin locks
 (*) mutexes
 (*) semaphores
 (*) R/W semaphores
 (*) RCU

cheers