Merge tag 'for-linus-iommufd' of git://git.kernel.org/pub/scm/linux/kernel/git/jgg/iommufd

Pull iommufd updates from Jason Gunthorpe:
 "Collection of small cleanup and one fix:

   - Sort headers and struct forward declarations

   - Fix random selftest failures in some cases due to dirty tracking
     tests

   - Have the reserved IOVA regions mechanism work when a HWPT is used
     as a nesting parent. This updates the nesting parent's IOAS with
     the reserved regions of the device and will also install the ITS
     doorbell page on ARM.

   - Add missed validation of parent domain ops against the current
     iommu

   - Fix a syzkaller bug related to integer overflow during ALIGN()

   - Tidy two iommu_domain attach paths"

* tag 'for-linus-iommufd' of git://git.kernel.org/pub/scm/linux/kernel/git/jgg/iommufd:
  iommu: Set iommu_attach_handle->domain in core
  iommufd: Avoid duplicated __iommu_group_set_core_domain() call
  iommufd: Protect against overflow of ALIGN() during iova allocation
  iommufd: Reorder struct forward declarations
  iommufd: Check the domain owner of the parent before creating a nesting domain
  iommufd/device: Enforce reserved IOVA also when attached to hwpt_nested
  iommufd/selftest: Fix buffer read overrrun in the dirty test
  iommufd: Reorder include files

Author: torvalds

drivers/iommu/iommu.c

@@ -3578,6 +3578,7 @@ int iommu_replace_group_handle(struct iommu_group *group,
 		ret = xa_reserve(&group->pasid_array, IOMMU_NO_PASID, GFP_KERNEL);
 		if (ret)
 			goto err_unlock;
+		handle->domain = new_domain;
 	}
 
 	ret = __iommu_group_set_domain(group, new_domain);

drivers/iommu/iommufd/device.c

@@ -1,12 +1,12 @@
 // SPDX-License-Identifier: GPL-2.0-only
 /* Copyright (c) 2021-2022, NVIDIA CORPORATION & AFFILIATES
  */
+#include <linux/iommu.h>
 #include <linux/iommufd.h>
 #include <linux/slab.h>
-#include <linux/iommu.h>
 #include <uapi/linux/iommufd.h>
-#include "../iommu-priv.h"
 
+#include "../iommu-priv.h"
 #include "io_pagetable.h"
 #include "iommufd_private.h"
 
@@ -327,8 +327,9 @@ static int iommufd_group_setup_msi(struct iommufd_group *igroup,
 	return 0;
 }
 
-static int iommufd_hwpt_paging_attach(struct iommufd_hwpt_paging *hwpt_paging,
-				      struct iommufd_device *idev)
+static int
+iommufd_device_attach_reserved_iova(struct iommufd_device *idev,
+				    struct iommufd_hwpt_paging *hwpt_paging)
 {
 	int rc;
 
@@ -354,6 +355,7 @@ static int iommufd_hwpt_paging_attach(struct iommufd_hwpt_paging *hwpt_paging,
 int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 				struct iommufd_device *idev)
 {
+	struct iommufd_hwpt_paging *hwpt_paging = find_hwpt_paging(hwpt);
 	int rc;
 
 	mutex_lock(&idev->igroup->lock);
@@ -363,8 +365,8 @@ int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 		goto err_unlock;
 	}
 
-	if (hwpt_is_paging(hwpt)) {
-		rc = iommufd_hwpt_paging_attach(to_hwpt_paging(hwpt), idev);
+	if (hwpt_paging) {
+		rc = iommufd_device_attach_reserved_iova(idev, hwpt_paging);
 		if (rc)
 			goto err_unlock;
 	}
@@ -387,9 +389,8 @@ int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 	mutex_unlock(&idev->igroup->lock);
 	return 0;
 err_unresv:
-	if (hwpt_is_paging(hwpt))
-		iopt_remove_reserved_iova(&to_hwpt_paging(hwpt)->ioas->iopt,
-					  idev->dev);
+	if (hwpt_paging)
+		iopt_remove_reserved_iova(&hwpt_paging->ioas->iopt, idev->dev);
 err_unlock:
 	mutex_unlock(&idev->igroup->lock);
 	return rc;
@@ -399,16 +400,16 @@ struct iommufd_hw_pagetable *
 iommufd_hw_pagetable_detach(struct iommufd_device *idev)
 {
 	struct iommufd_hw_pagetable *hwpt = idev->igroup->hwpt;
+	struct iommufd_hwpt_paging *hwpt_paging = find_hwpt_paging(hwpt);
 
 	mutex_lock(&idev->igroup->lock);
 	list_del(&idev->group_item);
 	if (list_empty(&idev->igroup->device_list)) {
 		iommufd_hwpt_detach_device(hwpt, idev);
 		idev->igroup->hwpt = NULL;
 	}
-	if (hwpt_is_paging(hwpt))
-		iopt_remove_reserved_iova(&to_hwpt_paging(hwpt)->ioas->iopt,
-					  idev->dev);
+	if (hwpt_paging)
+		iopt_remove_reserved_iova(&hwpt_paging->ioas->iopt, idev->dev);
 	mutex_unlock(&idev->igroup->lock);
 
 	/* Caller must destroy hwpt */
@@ -440,17 +441,17 @@ iommufd_group_remove_reserved_iova(struct iommufd_group *igroup,
 }
 
 static int
-iommufd_group_do_replace_paging(struct iommufd_group *igroup,
-				struct iommufd_hwpt_paging *hwpt_paging)
+iommufd_group_do_replace_reserved_iova(struct iommufd_group *igroup,
+				       struct iommufd_hwpt_paging *hwpt_paging)
 {
-	struct iommufd_hw_pagetable *old_hwpt = igroup->hwpt;
+	struct iommufd_hwpt_paging *old_hwpt_paging;
 	struct iommufd_device *cur;
 	int rc;
 
 	lockdep_assert_held(&igroup->lock);
 
-	if (!hwpt_is_paging(old_hwpt) ||
-	    hwpt_paging->ioas != to_hwpt_paging(old_hwpt)->ioas) {
+	old_hwpt_paging = find_hwpt_paging(igroup->hwpt);
+	if (!old_hwpt_paging || hwpt_paging->ioas != old_hwpt_paging->ioas) {
 		list_for_each_entry(cur, &igroup->device_list, group_item) {
 			rc = iopt_table_enforce_dev_resv_regions(
 				&hwpt_paging->ioas->iopt, cur->dev, NULL);
@@ -473,6 +474,8 @@ static struct iommufd_hw_pagetable *
 iommufd_device_do_replace(struct iommufd_device *idev,
 			  struct iommufd_hw_pagetable *hwpt)
 {
+	struct iommufd_hwpt_paging *hwpt_paging = find_hwpt_paging(hwpt);
+	struct iommufd_hwpt_paging *old_hwpt_paging;
 	struct iommufd_group *igroup = idev->igroup;
 	struct iommufd_hw_pagetable *old_hwpt;
 	unsigned int num_devices;
@@ -491,9 +494,8 @@ iommufd_device_do_replace(struct iommufd_device *idev,
 	}
 
 	old_hwpt = igroup->hwpt;
-	if (hwpt_is_paging(hwpt)) {
-		rc = iommufd_group_do_replace_paging(igroup,
-						     to_hwpt_paging(hwpt));
+	if (hwpt_paging) {
+		rc = iommufd_group_do_replace_reserved_iova(igroup, hwpt_paging);
 		if (rc)
 			goto err_unlock;
 	}
@@ -502,11 +504,10 @@ iommufd_device_do_replace(struct iommufd_device *idev,
 	if (rc)
 		goto err_unresv;
 
-	if (hwpt_is_paging(old_hwpt) &&
-	    (!hwpt_is_paging(hwpt) ||
-	     to_hwpt_paging(hwpt)->ioas != to_hwpt_paging(old_hwpt)->ioas))
-		iommufd_group_remove_reserved_iova(igroup,
-						   to_hwpt_paging(old_hwpt));
+	old_hwpt_paging = find_hwpt_paging(old_hwpt);
+	if (old_hwpt_paging &&
+	    (!hwpt_paging || hwpt_paging->ioas != old_hwpt_paging->ioas))
+		iommufd_group_remove_reserved_iova(igroup, old_hwpt_paging);
 
 	igroup->hwpt = hwpt;
 
@@ -524,9 +525,8 @@ iommufd_device_do_replace(struct iommufd_device *idev,
 	/* Caller must destroy old_hwpt */
 	return old_hwpt;
 err_unresv:
-	if (hwpt_is_paging(hwpt))
-		iommufd_group_remove_reserved_iova(igroup,
-						   to_hwpt_paging(hwpt));
+	if (hwpt_paging)
+		iommufd_group_remove_reserved_iova(igroup, hwpt_paging);
 err_unlock:
 	mutex_unlock(&idev->igroup->lock);
 	return ERR_PTR(rc);

drivers/iommu/iommufd/fault.c

@@ -3,14 +3,14 @@
  */
 #define pr_fmt(fmt) "iommufd: " fmt
 
+#include <linux/anon_inodes.h>
 #include <linux/file.h>
 #include <linux/fs.h>
+#include <linux/iommufd.h>
 #include <linux/module.h>
 #include <linux/mutex.h>
-#include <linux/iommufd.h>
 #include <linux/pci.h>
 #include <linux/poll.h>
-#include <linux/anon_inodes.h>
 #include <uapi/linux/iommufd.h>
 
 #include "../iommu-priv.h"
@@ -161,7 +161,6 @@ static int __fault_domain_replace_dev(struct iommufd_device *idev,
 		if (!handle)
 			return -ENOMEM;
 
-		handle->handle.domain = hwpt->domain;
 		handle->idev = idev;
 		ret = iommu_replace_group_handle(idev->igroup->group,
 						 hwpt->domain, &handle->handle);

drivers/iommu/iommufd/hw_pagetable.c

@@ -225,7 +225,8 @@ iommufd_hwpt_nested_alloc(struct iommufd_ctx *ictx,
 	if ((flags & ~IOMMU_HWPT_FAULT_ID_VALID) ||
 	    !user_data->len || !ops->domain_alloc_user)
 		return ERR_PTR(-EOPNOTSUPP);
-	if (parent->auto_domain || !parent->nest_parent)
+	if (parent->auto_domain || !parent->nest_parent ||
+	    parent->common.domain->owner != ops)
 		return ERR_PTR(-EINVAL);
 
 	hwpt_nested = __iommufd_object_alloc(

drivers/iommu/iommufd/io_pagetable.c

@@ -8,17 +8,17 @@
  * The datastructure uses the iopt_pages to optimize the storage of the PFNs
  * between the domains and xarray.
  */
+#include <linux/err.h>
+#include <linux/errno.h>
+#include <linux/iommu.h>
 #include <linux/iommufd.h>
 #include <linux/lockdep.h>
-#include <linux/iommu.h>
 #include <linux/sched/mm.h>
-#include <linux/err.h>
 #include <linux/slab.h>
-#include <linux/errno.h>
 #include <uapi/linux/iommufd.h>
 
-#include "io_pagetable.h"
 #include "double_span.h"
+#include "io_pagetable.h"
 
 struct iopt_pages_list {
 	struct iopt_pages *pages;
@@ -112,6 +112,7 @@ static int iopt_alloc_iova(struct io_pagetable *iopt, unsigned long *iova,
 	unsigned long page_offset = uptr % PAGE_SIZE;
 	struct interval_tree_double_span_iter used_span;
 	struct interval_tree_span_iter allowed_span;
+	unsigned long max_alignment = PAGE_SIZE;
 	unsigned long iova_alignment;
 
 	lockdep_assert_held(&iopt->iova_rwsem);
@@ -131,6 +132,13 @@ static int iopt_alloc_iova(struct io_pagetable *iopt, unsigned long *iova,
 				       roundup_pow_of_two(length),
 				       1UL << __ffs64(uptr));
 
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+	max_alignment = HPAGE_SIZE;
+#endif
+	/* Protect against ALIGN() overflow */
+	if (iova_alignment >= max_alignment)
+		iova_alignment = max_alignment;
+
 	if (iova_alignment < iopt->iova_alignment)
 		return -EINVAL;
 

drivers/iommu/iommufd/io_pagetable.h

@@ -6,8 +6,8 @@
 #define __IO_PAGETABLE_H
 
 #include <linux/interval_tree.h>
-#include <linux/mutex.h>
 #include <linux/kref.h>
+#include <linux/mutex.h>
 #include <linux/xarray.h>
 
 #include "iommufd_private.h"

drivers/iommu/iommufd/ioas.c

@@ -3,8 +3,8 @@
  * Copyright (c) 2021-2022, NVIDIA CORPORATION & AFFILIATES
  */
 #include <linux/interval_tree.h>
-#include <linux/iommufd.h>
 #include <linux/iommu.h>
+#include <linux/iommufd.h>
 #include <uapi/linux/iommufd.h>
 
 #include "io_pagetable.h"

drivers/iommu/iommufd/iommufd_private.h

@@ -4,13 +4,14 @@
 #ifndef __IOMMUFD_PRIVATE_H
 #define __IOMMUFD_PRIVATE_H
 
-#include <linux/rwsem.h>
-#include <linux/xarray.h>
-#include <linux/refcount.h>
-#include <linux/uaccess.h>
 #include <linux/iommu.h>
 #include <linux/iova_bitmap.h>
+#include <linux/refcount.h>
+#include <linux/rwsem.h>
+#include <linux/uaccess.h>
+#include <linux/xarray.h>
 #include <uapi/linux/iommufd.h>
+
 #include "../iommu-priv.h"
 
 struct iommu_domain;
@@ -324,6 +325,25 @@ to_hwpt_paging(struct iommufd_hw_pagetable *hwpt)
 	return container_of(hwpt, struct iommufd_hwpt_paging, common);
 }
 
+static inline struct iommufd_hwpt_nested *
+to_hwpt_nested(struct iommufd_hw_pagetable *hwpt)
+{
+	return container_of(hwpt, struct iommufd_hwpt_nested, common);
+}
+
+static inline struct iommufd_hwpt_paging *
+find_hwpt_paging(struct iommufd_hw_pagetable *hwpt)
+{
+	switch (hwpt->obj.type) {
+	case IOMMUFD_OBJ_HWPT_PAGING:
+		return to_hwpt_paging(hwpt);
+	case IOMMUFD_OBJ_HWPT_NESTED:
+		return to_hwpt_nested(hwpt)->parent;
+	default:
+		return NULL;
+	}
+}
+
 static inline struct iommufd_hwpt_paging *
 iommufd_get_hwpt_paging(struct iommufd_ucmd *ucmd, u32 id)
 {
@@ -490,8 +510,10 @@ static inline int iommufd_hwpt_attach_device(struct iommufd_hw_pagetable *hwpt,
 static inline void iommufd_hwpt_detach_device(struct iommufd_hw_pagetable *hwpt,
 					      struct iommufd_device *idev)
 {
-	if (hwpt->fault)
+	if (hwpt->fault) {
 		iommufd_fault_domain_detach_dev(hwpt, idev);
+		return;
+	}
 
 	iommu_detach_group(hwpt->domain, idev->igroup->group);
 }

drivers/iommu/iommufd/iommufd_test.h

@@ -4,8 +4,8 @@
 #ifndef _UAPI_IOMMUFD_TEST_H
 #define _UAPI_IOMMUFD_TEST_H
 
-#include <linux/types.h>
 #include <linux/iommufd.h>
+#include <linux/types.h>
 
 enum {
 	IOMMU_TEST_OP_ADD_RESERVED = 1,

drivers/iommu/iommufd/iova_bitmap.c

@@ -3,10 +3,10 @@
  * Copyright (c) 2022, Oracle and/or its affiliates.
  * Copyright (c) 2022, NVIDIA CORPORATION & AFFILIATES. All rights reserved
  */
+#include <linux/highmem.h>
 #include <linux/iova_bitmap.h>
 #include <linux/mm.h>
 #include <linux/slab.h>
-#include <linux/highmem.h>
 
 #define BITS_PER_PAGE (PAGE_SIZE * BITS_PER_BYTE)