On Fri, 2015-03-06 at 10:41 +0000, Will Deacon wrote: > > As a longer term solution, would it make sense to add a user flag at VFIO > > user API level to turn the IOMMU_CACHE on? > > I think userspace certainly needs a way to figure out if a device is > coherent or not, otherwise it can't generate the correct device-tree > properties for something like a KVM guest but the IOMMU_* setting should > remain in the kernel IMO. Similarly for things like MSI pages, which would > need to be mapped as device memory on ARM -- that should be exposed as a > higher level "please map my MSI page here" ioctl as opposed to requiring > userspace to supply the correct memory attributes. Userspace already has a way to determine if all the IOMMUs for a container support cache coherence, as does KVM via the kvm-vfio pseudo device. It's inadvisable to allow userspace to control whether IOMMU_CACHE is used on x86 as emulating cache sync instructions hurts the overall system performance. We don't want users to arbitrarily decide to enable those heavy-weight operations. Thanks, Alex _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
