Re: [PATCH v8 7/9] vfio/mdev: Add iommu related member in mdev_device

[Jason Gunthorpe]

On Mon, Mar 25, 2019 at 09:30:34AM +0800, Lu Baolu wrote:
> A parent device might create different types of mediated
> devices. For example, a mediated device could be created
> by the parent device with full isolation and protection
> provided by the IOMMU. One usage case could be found on
> Intel platforms where a mediated device is an assignable
> subset of a PCI, the DMA requests on behalf of it are all
> tagged with a PASID. Since IOMMU supports PASID-granular
> translations (scalable mode in VT-d 3.0), this mediated
> device could be individually protected and isolated by an
> IOMMU.
> 
> This patch adds a new member in the struct mdev_device to
> indicate that the mediated device represented by mdev could
> be isolated and protected by attaching a domain to a device
> represented by mdev->iommu_device. It also adds a helper to
> add or set the iommu device.
> 
> * mdev_device->iommu_device
>   - This, if set, indicates that the mediated device could
>     be fully isolated and protected by IOMMU via attaching
>     an iommu domain to this device. If empty, it indicates
>     using vendor defined isolation, hence bypass IOMMU.
> 
> * mdev_set/get_iommu_device(dev, iommu_device)
>   - Set or get the iommu device which represents this mdev
>     in IOMMU's device scope. Drivers don't need to set the
>     iommu device if it uses vendor defined isolation.
> 
> Cc: Ashok Raj <ashok.raj@xxxxxxxxx>
> Cc: Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>
> Cc: Kevin Tian <kevin.tian@xxxxxxxxx>
> Cc: Liu Yi L <yi.l.liu@xxxxxxxxx>
> Suggested-by: Kevin Tian <kevin.tian@xxxxxxxxx>
> Suggested-by: Alex Williamson <alex.williamson@xxxxxxxxxx>
> Signed-off-by: Lu Baolu <baolu.lu@xxxxxxxxxxxxxxx>
> Reviewed-by: Jean-Philippe Brucker <jean-philippe.brucker@xxxxxxx>
> --- 
>  drivers/vfio/mdev/mdev_core.c    | 18 ++++++++++++++++++
>  drivers/vfio/mdev/mdev_private.h |  1 +
>  include/linux/mdev.h             | 14 ++++++++++++++
>  3 files changed, 33 insertions(+)
> 
> diff --git a/drivers/vfio/mdev/mdev_core.c b/drivers/vfio/mdev/mdev_core.c
> index b96fedc77ee5..1b6435529166 100644
> +++ b/drivers/vfio/mdev/mdev_core.c
> @@ -390,6 +390,24 @@ int mdev_device_remove(struct device *dev, bool force_remove)
>  	return 0;
>  }
>  
> +int mdev_set_iommu_device(struct device *dev, struct device *iommu_device)
> +{
> +	struct mdev_device *mdev = to_mdev_device(dev);
> +
> +	mdev->iommu_device = iommu_device;
> +
> +	return 0;
> +}
> +EXPORT_SYMBOL(mdev_set_iommu_device);

I was looking at these functions when touching the mdev stuff and I
have some concerns.

1) Please don't merge dead code. It is a year later and there is still
   no in-tree user for any of this. This is not our process. Even
   worse it was exported so it looks like this dead code is supporting
   out of tree modules.

2) Why is this like this? Every struct device already has a connection
   to the iommu layer and every mdev has a struct device all its own.

   Why did we need to add special 'if (mdev)' stuff all over the
   place? This smells like the same abuse Thomas
   and I pointed out for the interrupt domains.

   After my next series the mdev drivers will have direct access to
   the vfio_device. So an alternative to using the struct device, or
   adding 'if mdev' is to add an API to the vfio_device world to
   inject what iommu configuration is needed from that direction
   instead of trying to discover it from a struct device.

3) The vfio_bus_is_mdev() and related symbol_get() nonsense in
   drivers/vfio/vfio_iommu_type1.c has to go, for the same reasons
   it was not acceptable to do this for the interrupt side either.

4) It seems pretty clear to me this will be heavily impacted by the
   /dev/ioasid discussion. Please consider removing the dead code now.

Basically, please fix this before trying to get idxd mdev merged as
the first user.

Jason