[RFC PATCH v2 4/7] iommufd: Associate kvm pointer to iommufd ctx

[Jason Gunthorpe]

On Mon, Jun 24, 2024 at 12:51:34PM -0700, Oliver Upton wrote:

> Having the SMMU participate in broadcast TLB maintenance means TLBI
> instructions on the CPU invalidate translations in the SMMU, which is
> useful for SVA usecases. However, if the host is using different VMIDs
> for the CPU and SMMU, then guest TLBIs no longer match the guest's SMMU
> context...
> 
> Pinning / sharing the VMID between CPU and SMMU is a hard requirement if
> you advertise BTM support to the guest.

Yes, and the extra detail that the SMMU *must* always have a unique
VMID for the specific VM, we can't leave it in a state without one.

Meaning there is little purpose to changing the VMID at runtime as you
can't multiplex more VM's than you have IDs once the SMMU is using it
anyhow.

Jason