IOMMU Fixes for Linux v6.11-rc5

Including:
 
 	- Fix a device-stall problem in bad io-page-fault setups (faults
 	  received from devices with no supporting domain attached).
 
 	- Context flush fix for Intel VT-d.
 
 	- Do not allow non-read+non-write mapping through iommufd as most
 	  implementations can not handle that.
 
 	- Fix a possible infinite-loop issue in map_pages() path.
 
 	- Add Jean-Philippe as reviewer for SMMUv3 SVA support
 -----BEGIN PGP SIGNATURE-----
 
 iQIzBAABCAAdFiEEr9jSbILcajRFYWYyK/BELZcBGuMFAmbRvfEACgkQK/BELZcB
 GuOB8w//WLapQpxMw9w+4l3Z3SqxB5gSPF6pdCJwYRrpFGBX1yNZ0vWtF2TpKtOC
 NaMa/EC1C2FWjcArCP21uFtDvN04FgXSVl6sjFUHsUf+YALrUfljQk/XFI4SenTq
 PtvPv8PVGbhqLtdJDXMlQWBN3RX0qK/PIFmuUX5ySBk7J7k5QyBi2HEuK2DbPM7j
 +LMnyTHj5Aa2jRz/NSCDIRKbSFJKgvd8apval2VX0zljjpyqk5KmHHjkLtiOiTTI
 G6ZJlRYCn98eTLU2ww8b7/y0vVYop7C1Q7Cyds/72xvW+a3jbSRIGf6yqtmdbMYd
 faxRng5rWHWsq3XMZC+Ts9k2FA3pUIvOmfptCFfrQYYXvZI6dD6o7uMko6SF82n4
 xEy+H6AEWZXF70xaJDp1cn1PpURJgJly/l/6qAIB746qNT7j/CcOOha1bpbCy81x
 EIOl0B4wyJGjQnxjKsH01K9ec3uT6rugbpFEE9PL8l25khhyweBwuQWc2EVxRZgH
 ICH4pCmvU9Wy6mpXL2R/SyzECWjgg0oJr+pq3Yxv7xufSGQswWJ/StFozSBHnH01
 OGGA/2xMrNeRzlm4PZfRzdAiCfYX9kEodiF1jGLA4B1V5Tx/y1LSX7W/nCeZmlRz
 /OhEC07DWZumeSCTe5I+BmZwiXh/DEAlUypDQkVKaaeGltlyvl8=
 =8XuD
 -----END PGP SIGNATURE-----

Merge tag 'iommu-fixes-v6.11-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/iommu/linux

Pull iommu fixes from Joerg Roedel:

 - Fix a device-stall problem in bad io-page-fault setups (faults
   received from devices with no supporting domain attached).

 - Context flush fix for Intel VT-d.

 - Do not allow non-read+non-write mapping through iommufd as most
   implementations can not handle that.

 - Fix a possible infinite-loop issue in map_pages() path.

 - Add Jean-Philippe as reviewer for SMMUv3 SVA support

* tag 'iommu-fixes-v6.11-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/iommu/linux:
  MAINTAINERS: Add Jean-Philippe as SMMUv3 SVA reviewer
  iommu: Do not return 0 from map_pages if it doesn't do anything
  iommufd: Do not allow creating areas without READ or WRITE
  iommu/vt-d: Fix incorrect domain ID in context flush helper
  iommu: Handle iommu faults for a bad iopf setup

MAINTAINERS

@@ -1880,6 +1880,10 @@ F:	Documentation/devicetree/bindings/iommu/arm,smmu*
 F:	drivers/iommu/arm/
 F:	drivers/iommu/io-pgtable-arm*
 
+ARM SMMU SVA SUPPORT
+R:	Jean-Philippe Brucker <jean-philippe@linaro.org>
+F:	drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3-sva.c
+
 ARM SUB-ARCHITECTURES
 L:	linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
 S:	Maintained

drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c

@@ -1777,7 +1777,7 @@ static int arm_smmu_handle_evt(struct arm_smmu_device *smmu, u64 *evt)
 		goto out_unlock;
 	}
 
-	iommu_report_device_fault(master->dev, &fault_evt);
+	ret = iommu_report_device_fault(master->dev, &fault_evt);
 out_unlock:
 	mutex_unlock(&smmu->streams_mutex);
 	return ret;

drivers/iommu/intel/iommu.c

@@ -1944,6 +1944,7 @@ static void domain_context_clear_one(struct device_domain_info *info, u8 bus, u8
 {
 	struct intel_iommu *iommu = info->iommu;
 	struct context_entry *context;
+	u16 did;
 
 	spin_lock(&iommu->lock);
 	context = iommu_context_addr(iommu, bus, devfn, 0);
@@ -1952,10 +1953,11 @@ static void domain_context_clear_one(struct device_domain_info *info, u8 bus, u8
 		return;
 	}
 
+	did = context_domain_id(context);
 	context_clear_entry(context);
 	__iommu_flush_cache(iommu, context, sizeof(*context));
 	spin_unlock(&iommu->lock);
-	intel_context_flush_present(info, context, true);
+	intel_context_flush_present(info, context, did, true);
 }
 
 static int domain_setup_first_level(struct intel_iommu *iommu,
@@ -4249,6 +4251,7 @@ static int context_flip_pri(struct device_domain_info *info, bool enable)
 	struct intel_iommu *iommu = info->iommu;
 	u8 bus = info->bus, devfn = info->devfn;
 	struct context_entry *context;
+	u16 did;
 
 	spin_lock(&iommu->lock);
 	if (context_copied(iommu, bus, devfn)) {
@@ -4261,6 +4264,7 @@ static int context_flip_pri(struct device_domain_info *info, bool enable)
 		spin_unlock(&iommu->lock);
 		return -ENODEV;
 	}
+	did = context_domain_id(context);
 
 	if (enable)
 		context_set_sm_pre(context);
@@ -4269,7 +4273,7 @@ static int context_flip_pri(struct device_domain_info *info, bool enable)
 
 	if (!ecap_coherent(iommu->ecap))
 		clflush_cache_range(context, sizeof(*context));
-	intel_context_flush_present(info, context, true);
+	intel_context_flush_present(info, context, did, true);
 	spin_unlock(&iommu->lock);
 
 	return 0;

drivers/iommu/intel/iommu.h

@@ -1154,7 +1154,7 @@ void cache_tag_flush_range_np(struct dmar_domain *domain, unsigned long start,
 
 void intel_context_flush_present(struct device_domain_info *info,
 				 struct context_entry *context,
-				 bool affect_domains);
+				 u16 did, bool affect_domains);
 
 #ifdef CONFIG_INTEL_IOMMU_SVM
 void intel_svm_check(struct intel_iommu *iommu);

drivers/iommu/intel/pasid.c

@@ -683,6 +683,7 @@ static void device_pasid_table_teardown(struct device *dev, u8 bus, u8 devfn)
 	struct device_domain_info *info = dev_iommu_priv_get(dev);
 	struct intel_iommu *iommu = info->iommu;
 	struct context_entry *context;
+	u16 did;
 
 	spin_lock(&iommu->lock);
 	context = iommu_context_addr(iommu, bus, devfn, false);
@@ -691,10 +692,11 @@ static void device_pasid_table_teardown(struct device *dev, u8 bus, u8 devfn)
 		return;
 	}
 
+	did = context_domain_id(context);
 	context_clear_entry(context);
 	__iommu_flush_cache(iommu, context, sizeof(*context));
 	spin_unlock(&iommu->lock);
-	intel_context_flush_present(info, context, false);
+	intel_context_flush_present(info, context, did, false);
 }
 
 static int pci_pasid_table_teardown(struct pci_dev *pdev, u16 alias, void *data)
@@ -885,10 +887,9 @@ static void __context_flush_dev_iotlb(struct device_domain_info *info)
  */
 void intel_context_flush_present(struct device_domain_info *info,
 				 struct context_entry *context,
-				 bool flush_domains)
+				 u16 did, bool flush_domains)
 {
 	struct intel_iommu *iommu = info->iommu;
-	u16 did = context_domain_id(context);
 	struct pasid_entry *pte;
 	int i;
 

drivers/iommu/io-pgfault.c

@@ -115,6 +115,59 @@ static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param,
 	return group;
 }
 
+static struct iommu_attach_handle *find_fault_handler(struct device *dev,
+						     struct iopf_fault *evt)
+{
+	struct iommu_fault *fault = &evt->fault;
+	struct iommu_attach_handle *attach_handle;
+
+	if (fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_PASID_VALID) {
+		attach_handle = iommu_attach_handle_get(dev->iommu_group,
+				fault->prm.pasid, 0);
+		if (IS_ERR(attach_handle)) {
+			const struct iommu_ops *ops = dev_iommu_ops(dev);
+
+			if (!ops->user_pasid_table)
+				return NULL;
+			/*
+			 * The iommu driver for this device supports user-
+			 * managed PASID table. Therefore page faults for
+			 * any PASID should go through the NESTING domain
+			 * attached to the device RID.
+			 */
+			attach_handle = iommu_attach_handle_get(
+					dev->iommu_group, IOMMU_NO_PASID,
+					IOMMU_DOMAIN_NESTED);
+			if (IS_ERR(attach_handle))
+				return NULL;
+		}
+	} else {
+		attach_handle = iommu_attach_handle_get(dev->iommu_group,
+				IOMMU_NO_PASID, 0);
+
+		if (IS_ERR(attach_handle))
+			return NULL;
+	}
+
+	if (!attach_handle->domain->iopf_handler)
+		return NULL;
+
+	return attach_handle;
+}
+
+static void iopf_error_response(struct device *dev, struct iopf_fault *evt)
+{
+	const struct iommu_ops *ops = dev_iommu_ops(dev);
+	struct iommu_fault *fault = &evt->fault;
+	struct iommu_page_response resp = {
+		.pasid = fault->prm.pasid,
+		.grpid = fault->prm.grpid,
+		.code = IOMMU_PAGE_RESP_INVALID
+	};
+
+	ops->page_response(dev, evt, &resp);
+}
+
 /**
  * iommu_report_device_fault() - Report fault event to device driver
  * @dev: the device
@@ -153,24 +206,39 @@ static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param,
  * handling framework should guarantee that the iommu domain could only be
  * freed after the device has stopped generating page faults (or the iommu
  * hardware has been set to block the page faults) and the pending page faults
- * have been flushed.
+ * have been flushed. In case no page fault handler is attached or no iopf params
+ * are setup, then the ops->page_response() is called to complete the evt.
+ *
+ * Returns 0 on success, or an error in case of a bad/failed iopf setup.
  */
-void iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
+int iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
 {
+	struct iommu_attach_handle *attach_handle;
 	struct iommu_fault *fault = &evt->fault;
 	struct iommu_fault_param *iopf_param;
 	struct iopf_group abort_group = {};
 	struct iopf_group *group;
 
+	attach_handle = find_fault_handler(dev, evt);
+	if (!attach_handle)
+		goto err_bad_iopf;
+
+	/*
+	 * Something has gone wrong if a fault capable domain is attached but no
+	 * iopf_param is setup
+	 */
 	iopf_param = iopf_get_dev_fault_param(dev);
 	if (WARN_ON(!iopf_param))
-		return;
+		goto err_bad_iopf;
 
 	if (!(fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_LAST_PAGE)) {
-		report_partial_fault(iopf_param, fault);
+		int ret;
+
+		ret = report_partial_fault(iopf_param, fault);
 		iopf_put_dev_fault_param(iopf_param);
 		/* A request that is not the last does not need to be ack'd */
-		return;
+
+		return ret;
 	}
 
 	/*
@@ -185,38 +253,7 @@ void iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
 	if (group == &abort_group)
 		goto err_abort;
 
-	if (fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_PASID_VALID) {
+	group->attach_handle = attach_handle;
-		group->attach_handle = iommu_attach_handle_get(dev->iommu_group,
-							       fault->prm.pasid,
-							       0);
-		if (IS_ERR(group->attach_handle)) {
-			const struct iommu_ops *ops = dev_iommu_ops(dev);
-
-			if (!ops->user_pasid_table)
-				goto err_abort;
-
-			/*
-			 * The iommu driver for this device supports user-
-			 * managed PASID table. Therefore page faults for
-			 * any PASID should go through the NESTING domain
-			 * attached to the device RID.
-			 */
-			group->attach_handle =
-				iommu_attach_handle_get(dev->iommu_group,
-							IOMMU_NO_PASID,
-							IOMMU_DOMAIN_NESTED);
-			if (IS_ERR(group->attach_handle))
-				goto err_abort;
-		}
-	} else {
-		group->attach_handle =
-			iommu_attach_handle_get(dev->iommu_group, IOMMU_NO_PASID, 0);
-		if (IS_ERR(group->attach_handle))
-			goto err_abort;
-	}
-
-	if (!group->attach_handle->domain->iopf_handler)
-		goto err_abort;
 
 	/*
 	 * On success iopf_handler must call iopf_group_response() and
@@ -225,7 +262,7 @@ void iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
 	if (group->attach_handle->domain->iopf_handler(group))
 		goto err_abort;
 
-	return;
+	return 0;
 
 err_abort:
 	dev_warn_ratelimited(dev, "iopf with pasid %d aborted\n",
@@ -235,6 +272,14 @@ err_abort:
 		__iopf_free_group(group);
 	else
 		iopf_free_group(group);
+
+	return 0;
+
+err_bad_iopf:
+	if (fault->type == IOMMU_FAULT_PAGE_REQ)
+		iopf_error_response(dev, evt);
+
+	return -EINVAL;
 }
 EXPORT_SYMBOL_GPL(iommu_report_device_fault);
 

drivers/iommu/io-pgtable-arm-v7s.c

@@ -552,9 +552,8 @@ static int arm_v7s_map_pages(struct io_pgtable_ops *ops, unsigned long iova,
 		    paddr >= (1ULL << data->iop.cfg.oas)))
 		return -ERANGE;
 
-	/* If no access, then nothing to do */
 	if (!(prot & (IOMMU_READ | IOMMU_WRITE)))
-		return 0;
+		return -EINVAL;
 
 	while (pgcount--) {
 		ret = __arm_v7s_map(data, iova, paddr, pgsize, prot, 1, data->pgd,

drivers/iommu/io-pgtable-arm.c

@@ -515,9 +515,8 @@ static int arm_lpae_map_pages(struct io_pgtable_ops *ops, unsigned long iova,
 	if (WARN_ON(iaext || paddr >> cfg->oas))
 		return -ERANGE;
 
-	/* If no access, then nothing to do */
 	if (!(iommu_prot & (IOMMU_READ | IOMMU_WRITE)))
-		return 0;
+		return -EINVAL;
 
 	prot = arm_lpae_prot_to_pte(data, iommu_prot);
 	ret = __arm_lpae_map(data, iova, paddr, pgsize, pgcount, prot, lvl,

drivers/iommu/io-pgtable-dart.c

@@ -245,9 +245,8 @@ static int dart_map_pages(struct io_pgtable_ops *ops, unsigned long iova,
 	if (WARN_ON(paddr >> cfg->oas))
 		return -ERANGE;
 
-	/* If no access, then nothing to do */
 	if (!(iommu_prot & (IOMMU_READ | IOMMU_WRITE)))
-		return 0;
+		return -EINVAL;
 
 	tbl = dart_get_table(data, iova);
 

drivers/iommu/iommufd/ioas.c

@@ -213,6 +213,10 @@ int iommufd_ioas_map(struct iommufd_ucmd *ucmd)
 	if (cmd->iova >= ULONG_MAX || cmd->length >= ULONG_MAX)
 		return -EOVERFLOW;
 
+	if (!(cmd->flags &
+	      (IOMMU_IOAS_MAP_WRITEABLE | IOMMU_IOAS_MAP_READABLE)))
+		return -EINVAL;
+
 	ioas = iommufd_get_ioas(ucmd->ictx, cmd->ioas_id);
 	if (IS_ERR(ioas))
 		return PTR_ERR(ioas);
@@ -253,6 +257,10 @@ int iommufd_ioas_copy(struct iommufd_ucmd *ucmd)
 	    cmd->dst_iova >= ULONG_MAX)
 		return -EOVERFLOW;
 
+	if (!(cmd->flags &
+	      (IOMMU_IOAS_MAP_WRITEABLE | IOMMU_IOAS_MAP_READABLE)))
+		return -EINVAL;
+
 	src_ioas = iommufd_get_ioas(ucmd->ictx, cmd->src_ioas_id);
 	if (IS_ERR(src_ioas))
 		return PTR_ERR(src_ioas);

include/linux/iommu.h

@@ -1563,7 +1563,7 @@ struct iopf_queue *iopf_queue_alloc(const char *name);
 void iopf_queue_free(struct iopf_queue *queue);
 int iopf_queue_discard_partial(struct iopf_queue *queue);
 void iopf_free_group(struct iopf_group *group);
-void iommu_report_device_fault(struct device *dev, struct iopf_fault *evt);
+int iommu_report_device_fault(struct device *dev, struct iopf_fault *evt);
 void iopf_group_response(struct iopf_group *group,
 			 enum iommu_page_response_code status);
 #else
@@ -1601,9 +1601,10 @@ static inline void iopf_free_group(struct iopf_group *group)
 {
 }
 
-static inline void
+static inline int
 iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
 {
+	return -ENODEV;
 }
 
 static inline void iopf_group_response(struct iopf_group *group,

tools/testing/selftests/iommu/iommufd.c

@@ -825,7 +825,7 @@ TEST_F(iommufd_ioas, copy_area)
 {
 	struct iommu_ioas_copy copy_cmd = {
 		.size = sizeof(copy_cmd),
-		.flags = IOMMU_IOAS_MAP_FIXED_IOVA,
+		.flags = IOMMU_IOAS_MAP_FIXED_IOVA | IOMMU_IOAS_MAP_WRITEABLE,
 		.dst_ioas_id = self->ioas_id,
 		.src_ioas_id = self->ioas_id,
 		.length = PAGE_SIZE,
@@ -1318,7 +1318,7 @@ TEST_F(iommufd_ioas, copy_sweep)
 {
 	struct iommu_ioas_copy copy_cmd = {
 		.size = sizeof(copy_cmd),
-		.flags = IOMMU_IOAS_MAP_FIXED_IOVA,
+		.flags = IOMMU_IOAS_MAP_FIXED_IOVA | IOMMU_IOAS_MAP_WRITEABLE,
 		.src_ioas_id = self->ioas_id,
 		.dst_iova = MOCK_APERTURE_START,
 		.length = MOCK_PAGE_SIZE,
@@ -1608,7 +1608,7 @@ TEST_F(iommufd_mock_domain, user_copy)
 	};
 	struct iommu_ioas_copy copy_cmd = {
 		.size = sizeof(copy_cmd),
-		.flags = IOMMU_IOAS_MAP_FIXED_IOVA,
+		.flags = IOMMU_IOAS_MAP_FIXED_IOVA | IOMMU_IOAS_MAP_WRITEABLE,
 		.dst_ioas_id = self->ioas_id,
 		.dst_iova = MOCK_APERTURE_START,
 		.length = BUFFER_SIZE,