On Tue, Jul 13, 2021 at 03:04:28PM -0400, Tony Krowiak wrote: > > > On 7/13/21 1:05 PM, Jason Gunthorpe wrote: > > On Tue, Jul 13, 2021 at 06:45:17PM +0200, Halil Pasic wrote: > > > > > Jason may give it another try to convince us that 0cc00c8d4050 only > > > silenced lockdep, but vfio_ap remained prone to deadlocks. To my best > > > knowledge using condition variable and a mutex is one of the well known > > > ways to implement an rwlock. > > The well known pattern is to use a rwsem. > > > > This: > > wait_event_cmd(matrix_mdev->wait_for_kvm, > > !matrix_mdev->kvm_busy, > > mutex_unlock(&matrix_dev->lock), > > mutex_lock(&matrix_dev->lock)); > > > > > > Is not really a rwsem, and is invsible to lockdep. > > The lockdep splat was due to holding the matrix_dev->lock > mutex while the kvm->lock was taken to plug the AP devices > into the guest. The same problem would occur whether an > rwsem or the mutex was used. See, everytime you say 'the same problem would occur with a rwsem' you don't get to go on and say everthing is fine if we open code a rwsem as kvm_busy This patch shows it works as a rwsem - look at what had to be changed to make it so and you will find some clue to where the problems are in kvm_busy version. In any event, I don't care anymore - please just get this merged, to AlexW's tree so I don't have conflicts with the rest of the ap changes for VFIO I've got queued up. Jason
