Re: [PATCH v2 02/10] iommu: Introduce a new iommu_group_replace_domain() API

[Baolu Lu]

On 2023/2/13 10:24, Tian, Kevin wrote:
> > From: Jason Gunthorpe<jgg@xxxxxxxxxx>
> > Sent: Saturday, February 11, 2023 8:45 AM
> >
> > On Fri, Feb 10, 2023 at 04:51:10PM -0700, Alex Williamson wrote:
> > > On Tue, 7 Feb 2023 13:17:54 -0800
> > > Nicolin Chen<nicolinc@xxxxxxxxxx>  wrote:
> > >
> > > > qemu has a need to replace the translations associated with a domain
> > > > when the guest does large-scale operations like switching between an
> > > > IDENTITY domain and, say, dma-iommu.c.
> > > >
> > > > Currently, it does this by replacing all the mappings in a single
> > > > domain, but this is very inefficient and means that domains have to be
> > > > per-device rather than per-translation.
> > > >
> > > > Provide a high-level API to allow replacements of one domain with
> > > > another. This is similar to a detach/attach cycle except it doesn't
> > > > force the group to go to the blocking domain in-between.
> > > >
> > > > By removing this forced blocking domain the iommu driver has the
> > > > opportunity to implement an atomic replacement of the domains to the
> > > > greatest extent its hardware allows.
> > > >
> > > > It could be possible to adderss this by simply removing the protection
> > > > from the iommu_attach_group(), but it is not so clear if that is safe
> > > > for the few users. Thus, add a new API to serve this new purpose.
> > > >
> > > > Atomic replacement allows the qemu emulation of the viommu to be
> > more
> > > > complete, as real hardware has this ability.
> > > I was under the impression that we could not atomically switch a
> > > device's domain relative to in-flight DMA.
> it's possible as long as the mappings for in-flight DMA don't change
> in the transition.

It also requires the mappings in old and new domains are identical. In
another word, any IOVA should be translated to a same result no matter
through the old domain, the new domain, or hardware caches.

A similar replacement has been implemented in the code. For example,
the Intel IOMMU driver dynamically transforms a large range (2M or 1G)
mapping from discrete 4k pages to a super pages to improve the paging
cache efficiency.

Best regards,
baolu