Re: [PATCH v5 11/11] vfio: unplug failover primary device before migration

[Alex Williamson]

On Wed, 23 Oct 2019 10:27:11 +0200
Jens Freimann <address@hidden> wrote:

> As usual block all vfio-pci devices from being migrated, but make an
> exception for failover primary devices. This is achieved by setting
> unmigratable to 0 but also add a migration blocker for all vfio-pci
> devices except failover primary devices. These will be unplugged before
> migration happens by the migration handler of the corresponding
> virtio-net standby device.
> 
> Signed-off-by: Jens Freimann <address@hidden>
> ---
>  hw/vfio/pci.c | 26 ++++++++++++++++++++------
>  hw/vfio/pci.h |  1 +
>  2 files changed, 21 insertions(+), 6 deletions(-)

I wonder if this might be cleaner if we add the migration
unconditionally initially, then remove it for specific cases, such as
this or when we start to see mdev devices that can actually support
migration.  That can be refined as we start to introduce the latter
though.

Acked-by: Alex Williamson <address@hidden>

> diff --git a/hw/vfio/pci.c b/hw/vfio/pci.c
> index 12fac39804..5dadfec6e8 100644
> --- a/hw/vfio/pci.c
> +++ b/hw/vfio/pci.c
> @@ -40,6 +40,7 @@
>  #include "pci.h"
>  #include "trace.h"
>  #include "qapi/error.h"
> +#include "migration/blocker.h"
>  
>  #define TYPE_VFIO_PCI "vfio-pci"
>  #define PCI_VFIO(obj)    OBJECT_CHECK(VFIOPCIDevice, obj, TYPE_VFIO_PCI)
> @@ -2732,6 +2733,17 @@ static void vfio_realize(PCIDevice *pdev, Error **errp)
>          return;
>      }
>  
> +    if (!pdev->net_failover_pair_id) {
> +        error_setg(&vdev->migration_blocker,
> +                "VFIO device doesn't support migration");
> +        ret = migrate_add_blocker(vdev->migration_blocker, &err);
> +        if (err) {
> +            error_propagate(errp, err);
> +            error_free(vdev->migration_blocker);
> +            return;
> +        }
> +    }
> +
>      vdev->vbasedev.name = g_path_get_basename(vdev->vbasedev.sysfsdev);
>      vdev->vbasedev.ops = &vfio_pci_ops;
>      vdev->vbasedev.type = VFIO_DEVICE_TYPE_PCI;
> @@ -3008,6 +3020,10 @@ out_teardown:
>      vfio_bars_exit(vdev);
>  error:
>      error_prepend(errp, VFIO_MSG_PREFIX, vdev->vbasedev.name);
> +    if (vdev->migration_blocker) {
> +        migrate_del_blocker(vdev->migration_blocker);
> +        error_free(vdev->migration_blocker);
> +    }
>  }
>  
>  static void vfio_instance_finalize(Object *obj)
> @@ -3019,6 +3035,10 @@ static void vfio_instance_finalize(Object *obj)
>      vfio_bars_finalize(vdev);
>      g_free(vdev->emulated_config_bits);
>      g_free(vdev->rom);
> +    if (vdev->migration_blocker) {
> +        migrate_del_blocker(vdev->migration_blocker);
> +        error_free(vdev->migration_blocker);
> +    }
>      /*
>       * XXX Leaking igd_opregion is not an oversight, we can't remove the
>       * fw_cfg entry therefore leaking this allocation seems like the safest
> @@ -3151,11 +3171,6 @@ static Property vfio_pci_dev_properties[] = {
>      DEFINE_PROP_END_OF_LIST(),
>  };
>  
> -static const VMStateDescription vfio_pci_vmstate = {
> -    .name = "vfio-pci",
> -    .unmigratable = 1,
> -};
> -
>  static void vfio_pci_dev_class_init(ObjectClass *klass, void *data)
>  {
>      DeviceClass *dc = DEVICE_CLASS(klass);
> @@ -3163,7 +3178,6 @@ static void vfio_pci_dev_class_init(ObjectClass *klass, 
> void *data)
>  
>      dc->reset = vfio_pci_reset;
>      dc->props = vfio_pci_dev_properties;
> -    dc->vmsd = &vfio_pci_vmstate;
>      dc->desc = "VFIO-based PCI device assignment";
>      set_bit(DEVICE_CATEGORY_MISC, dc->categories);
>      pdc->realize = vfio_realize;
> diff --git a/hw/vfio/pci.h b/hw/vfio/pci.h
> index 834a90d646..b329d50338 100644
> --- a/hw/vfio/pci.h
> +++ b/hw/vfio/pci.h
> @@ -168,6 +168,7 @@ typedef struct VFIOPCIDevice {
>      bool no_vfio_ioeventfd;
>      bool enable_ramfb;
>      VFIODisplay *dpy;
> +    Error *migration_blocker;
>  } VFIOPCIDevice;
>  
>  uint32_t vfio_pci_read_config(PCIDevice *pdev, uint32_t addr, int len);