Hi Andrew, I re-submitted the patch today with a more detailed explanation about the problem and the proposed solution. I would like to add that I tested this solution with iommuv2 driver and with AMD's amdkfd driver (which we currently try to upstream). If you could merge it for 3.17-rc1 as you said, that would be great. Oded On 29/07/14 23:48, Andrew Morton wrote: > On Tue, 29 Jul 2014 18:04:32 +0300 Oded Gabbay <oded.gabbay@xxxxxxx> wrote: > >> On 21/07/14 18:01, Oded Gabbay wrote: >>> On 17/07/14 21:46, j.glisse@xxxxxxxxx wrote: >>>> From: Peter Zijlstra <peterz@xxxxxxxxxxxxx> >>>> >>>> New mmu_notifier listener are eager to cleanup there structure after the >>>> mmu_notifier::release callback. In order to allow this the patch provide >>>> a function that allows to add a delayed call to the mmu_notifier srcu. It >>>> also add a function that will call barrier_srcu so those listener can sync >>>> with mmu_notifier. >>> >>> Tested with amdkfd and iommuv2 driver >>> So, >>> Tested-by: Oded Gabbay <oded.gabbay@xxxxxxx> >> >> akpm, any chance that only this specific patch from Peter.Z will get in 3.17 ? >> I must have it for amdkfd (HSA driver). Without it, I can't be in 3.17 either. > > I can send it in for 3.17-rc1. > > Can we get a better changelog please? "eager to cleanup there > structure after the mmu_notifier::release callback" is terribly vague > and brief. Fully describe the problem then describe the proposed > solution. Because others may be able to suggest alternative ways of > solving that problem, but this text simply doesn't provide enough details for > them to do so. > -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>