Re: [PATCH v4 05/11] iommu/sva: Assign a PASID to mm on PASID allocation and free it on mm exit

From: Jean-Philippe Brucker
Date: Mon Apr 25 2022 - 12:41:06 EST

On Mon, Apr 25, 2022 at 08:55:46AM -0700, Dave Hansen wrote:
> On 4/25/22 07:26, Jean-Philippe Brucker wrote:
> >>
> >> How does the IOMMU hardware know that all activity to a given PASID is
> >> finished? That activity should, today, be independent of an mm or a
> >> fd's lifetime.
> > In the case of uacce, it's tied to the fd lifetime: opening an accelerator
> > queue calls iommu_sva_bind_device(), which sets up the PASID context in
> > the IOMMU. Closing the queue calls iommu_sva_unbind_device() which
> > destroys the PASID context (after the device driver stopped all DMA for
> > this PASID).
> Could this PASID context destruction move from being "fd-based" to
> happening under mm_pasid_drop()? Logically, it seems like that should
> work because mm_pasid_drop() happens after exit_mmap() where the VMAs
> (which hold references to 'struct file' via vma->vm_file) are torn down.

The problem is that we'd have to request the device driver to stop DMA
before we can destroy the context and free the PASID. We did consider
doing this in the release() MMU notifier, but there were concerns about
blocking mmput() for too long (for example
though I think there was a more recent discussion). We also need to drain
the PRI and fault queues to get rid of all references to that PASID.

At the moment we disable (but not destroy) the PASID context in release(),
so when the process gets killed pending DMA transactions are silently
ignored. Then the device driver informs us through unbind() that no DMA is
active anymore and we can finish cleaning up, then reuse the PASID.