qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH qemu v10 12/14] vfio: Unregister IOMMU notifiers


From: Alex Williamson
Subject: Re: [Qemu-devel] [PATCH qemu v10 12/14] vfio: Unregister IOMMU notifiers when container is destroyed
Date: Mon, 06 Jul 2015 07:45:00 -0600

On Mon, 2015-07-06 at 22:59 +1000, Alexey Kardashevskiy wrote:
> On 07/06/2015 10:49 PM, Alex Williamson wrote:
> > On Mon, 2015-07-06 at 20:33 +1000, David Gibson wrote:
> >> On Mon, Jul 06, 2015 at 12:11:08PM +1000, Alexey Kardashevskiy wrote:
> >>> On systems with guest visible IOMMU, adding a new memory region onto
> >>> PCI bus calls vfio_listener_region_add() for every DMA window. This
> >>> installs a notifier for IOMMU memory regions. The notifier is supposed
> >>> to be removed by vfio_listener_region_del(), however in the case of mixed
> >>> PHB (emulated + VFIO devices) when last VFIO device is unplugged and
> >>> container gets destroyed, all existing DMA windows stay alive altogether
> >>> with the notifiers which are on the linked list which head was in
> >>> the destroyed container.
> >>>
> >>> This unregisters IOMMU memory region notifier when a container is
> >>> destroyed.
> >>>
> >>> Signed-off-by: Alexey Kardashevskiy <address@hidden>
> >>
> >> Reviewed-by: David Gibson <address@hidden>
> >>
> >> Alex,
> >>
> >> I think this is correct, but you've probably got a better
> >> understanding of it.  Will you take this through your tree?
> >
> > Yes, confusingly this patch was sent twice yesterday, once in this
> > series and once separately.  AFAICT they're identical, so I'll add your
> > R-b and add the patch to my pull request for 2.4-rc0.  Thanks,
> 
> Yes, these are identical, sorry for the confusion. btw what was the right 
> to do with this patch?

The patch stands on its own and doesn't conflict or contribute
specifically to this series.  It should be left on its own.  The usual
practice should be to separate patches for different subsystems into
series that stand on their own, not to bundle everything together across
subsystems for convenience.  Thanks,

Alex

> >>> ---
> >>> Changes:
> >>> v10:
> >>> * new to the patchset
> >>> ---
> >>>   hw/vfio/common.c | 8 ++++++++
> >>>   1 file changed, 8 insertions(+)
> >>>
> >>> diff --git a/hw/vfio/common.c b/hw/vfio/common.c
> >>> index 89ef37b..8eacfd7 100644
> >>> --- a/hw/vfio/common.c
> >>> +++ b/hw/vfio/common.c
> >>> @@ -772,11 +772,19 @@ static void vfio_disconnect_container(VFIOGroup 
> >>> *group)
> >>>
> >>>       if (QLIST_EMPTY(&container->group_list)) {
> >>>           VFIOAddressSpace *space = container->space;
> >>> +        VFIOGuestIOMMU *giommu, *tmp;
> >>>
> >>>           if (container->iommu_data.release) {
> >>>               container->iommu_data.release(container);
> >>>           }
> >>>           QLIST_REMOVE(container, next);
> >>> +
> >>> +        QLIST_FOREACH_SAFE(giommu, &container->giommu_list, giommu_next, 
> >>> tmp) {
> >>> +            memory_region_unregister_iommu_notifier(&giommu->n);
> >>> +            QLIST_REMOVE(giommu, giommu_next);
> >>> +            g_free(giommu);
> >>> +        }
> >>> +
> >>>           trace_vfio_disconnect_container(container->fd);
> >>>           close(container->fd);
> >>>           g_free(container);
> >>
> >
> >
> >
> 
> 






reply via email to

[Prev in Thread] Current Thread [Next in Thread]