Re: [PATCH v9 07/14] vfio/migration: Block multiple devices migration

[Avihai Horon]

On 08/02/2023 18:44, Cédric Le Goater wrote:
> External email: Use caution opening links or attachments
>
>
> On 2/8/23 14:08, Avihai Horon wrote:
> > On 08/02/2023 0:34, Alex Williamson wrote:
> > > External email: Use caution opening links or attachments
> > >
> > >
> > > On Mon, 6 Feb 2023 14:31:30 +0200
> > > Avihai Horon <avihaih@nvidia.com> wrote:
> > >
> > > > Currently VFIO migration doesn't implement some kind of intermediate
> > > > quiescent state in which P2P DMAs are quiesced before stopping or
> > > > running the device. This can cause problems in multi-device migration
> > > > where the devices are doing P2P DMAs, since the devices are not 
> > > > stopped
> > > > together at the same time.
> > > >
> > > > Until such support is added, block migration of multiple devices.
> > > >
> > > > Signed-off-by: Avihai Horon <avihaih@nvidia.com>
> > > > ---
> > > >   include/hw/vfio/vfio-common.h |  2 ++
> > > >   hw/vfio/common.c              | 51 
> > > > +++++++++++++++++++++++++++++++++++
> > > >   hw/vfio/migration.c           |  6 +++++
> > > >   3 files changed, 59 insertions(+)
> > > >
> > > > diff --git a/include/hw/vfio/vfio-common.h 
> > > > b/include/hw/vfio/vfio-common.h
> > > > index e573f5a9f1..56b1683824 100644
> > > > --- a/include/hw/vfio/vfio-common.h
> > > > +++ b/include/hw/vfio/vfio-common.h
> > > > @@ -218,6 +218,8 @@ typedef QLIST_HEAD(VFIOGroupList, VFIOGroup) 
> > > > VFIOGroupList;
> > > >   extern VFIOGroupList vfio_group_list;
> > > >
> > > >   bool vfio_mig_active(void);
> > > > +int vfio_block_multiple_devices_migration(Error **errp);
> > > > +void vfio_unblock_multiple_devices_migration(void);
> > > >   int64_t vfio_mig_bytes_transferred(void);
> > > >
> > > >   #ifdef CONFIG_LINUX
> > > > diff --git a/hw/vfio/common.c b/hw/vfio/common.c
> > > > index 3a35f4afad..01db41b735 100644
> > > > --- a/hw/vfio/common.c
> > > > +++ b/hw/vfio/common.c
> > > > @@ -41,6 +41,7 @@
> > > >   #include "qapi/error.h"
> > > >   #include "migration/migration.h"
> > > >   #include "migration/misc.h"
> > > > +#include "migration/blocker.h"
> > > >   #include "sysemu/tpm.h"
> > > >
> > > >   VFIOGroupList vfio_group_list =
> > > > @@ -337,6 +338,56 @@ bool vfio_mig_active(void)
> > > >       return true;
> > > >   }
> > > >
> > > > +Error *multiple_devices_migration_blocker;
>
> static ?
Yes, I will add.

> So we have two migration blockers, one per device and one global. I guess
> it is easier that way than trying to update the per device Error*.

Yes, because this blocker is not related to a specific device but rather 
to how many devices we have at a specific moment.

Thanks.