Re: [PATCH 6/8] iommu/arm-smmu-v3: Implement IOMMU_HWPT_ALLOC_NEST_PARENT
From: Robin Murphy <robin.murphy@arm.com>
Date: 2024-08-09 15:07:23
Also in:
kvm, linux-acpi, linux-iommu, linux-patches
On 2024-08-07 12:41 am, Jason Gunthorpe wrote:
quoted hunk ↗ jump to hunk
For SMMUv3 the parent must be a S2 domain, which can be composed into a IOMMU_DOMAIN_NESTED. In future the S2 parent will also need a VMID linked to the VIOMMU and even to KVM. Signed-off-by: Jason Gunthorpe <jgg@nvidia.com> --- drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-)diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c index 6bbe4aa7b9511c..5faaccef707ef1 100644 --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c@@ -3103,7 +3103,8 @@ arm_smmu_domain_alloc_user(struct device *dev, u32 flags, const struct iommu_user_data *user_data) { struct arm_smmu_master *master = dev_iommu_priv_get(dev); - const u32 PAGING_FLAGS = IOMMU_HWPT_ALLOC_DIRTY_TRACKING; + const u32 PAGING_FLAGS = IOMMU_HWPT_ALLOC_DIRTY_TRACKING | + IOMMU_HWPT_ALLOC_NEST_PARENT; struct arm_smmu_domain *smmu_domain; int ret;@@ -3116,6 +3117,14 @@ arm_smmu_domain_alloc_user(struct device *dev, u32 flags, if (!smmu_domain) return ERR_PTR(-ENOMEM); + if (flags & IOMMU_HWPT_ALLOC_NEST_PARENT) { + if (!(master->smmu->features & ARM_SMMU_FEAT_TRANS_S2)) {
Nope, nesting needs to rely on FEAT_NESTING, that's why it exists. S2 alone isn't sufficient - without S1 there's nothing to expose to userspace, so zero point in having a "nested" domain with nothing to nest into it - but furthermore we need S2 *without* unsafe broken TLBs. Thanks, Robin.
+ ret = -EOPNOTSUPP; + goto err_free; + } + smmu_domain->stage = ARM_SMMU_DOMAIN_S2; + } + smmu_domain->domain.type = IOMMU_DOMAIN_UNMANAGED; smmu_domain->domain.ops = arm_smmu_ops.default_domain_ops; ret = arm_smmu_domain_finalise(smmu_domain, master->smmu, flags);