All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jason Gunthorpe <jgg@nvidia.com>
To: iommu@lists.linux.dev, Kevin Tian <kevin.tian@intel.com>,
	linux-kselftest@vger.kernel.org
Cc: Lu Baolu <baolu.lu@linux.intel.com>,
	kvm@vger.kernel.org, Nicolin Chen <nicolinc@nvidia.com>,
	Yi Liu <yi.l.liu@intel.com>
Subject: [PATCH v3 03/17] iommufd: Replace the hwpt->devices list with iommufd_group
Date: Tue, 21 Mar 2023 16:14:37 -0300	[thread overview]
Message-ID: <3-v3-61d41fd9e13e+1f5-iommufd_alloc_jgg@nvidia.com> (raw)
In-Reply-To: <0-v3-61d41fd9e13e+1f5-iommufd_alloc_jgg@nvidia.com>

The devices list was used as a simple way to avoid having per-group
information. Now that this seems to be unavoidable, just commit to
per-group information fully and remove the devices list from the HWPT.

The iommufd_group stores the currently assigned HWPT for the entire group
and we can manage the per-device attach/detach with a list in the
iommufd_group.

For destruction the flow is organized to make the following patches
easier, the actual call to iommufd_object_destroy_user() is done at the
top of the call chain without holding any locks. The HWPT to be destroyed
is returned out from the locked region to make this possible. Later
patches create locking that requires this.

Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
---
 drivers/iommu/iommufd/device.c          | 78 ++++++++++++-------------
 drivers/iommu/iommufd/hw_pagetable.c    | 23 +++-----
 drivers/iommu/iommufd/iommufd_private.h | 13 ++---
 3 files changed, 51 insertions(+), 63 deletions(-)

diff --git a/drivers/iommu/iommufd/device.c b/drivers/iommu/iommufd/device.c
index 3fd623208c691f..66de0274d65629 100644
--- a/drivers/iommu/iommufd/device.c
+++ b/drivers/iommu/iommufd/device.c
@@ -20,9 +20,12 @@ static void iommufd_group_release(struct kref *kref)
 	struct iommufd_group *igroup =
 		container_of(kref, struct iommufd_group, ref);
 
+	WARN_ON(igroup->hwpt || !list_empty(&igroup->device_list));
+
 	xa_cmpxchg(&igroup->ictx->groups, iommu_group_id(igroup->group), igroup,
 		   NULL, GFP_KERNEL);
 	iommu_group_put(igroup->group);
+	mutex_destroy(&igroup->lock);
 	kfree(igroup);
 }
 
@@ -83,6 +86,8 @@ static struct iommufd_group *iommufd_get_group(struct iommufd_ctx *ictx,
 	}
 
 	kref_init(&new_igroup->ref);
+	mutex_init(&new_igroup->lock);
+	INIT_LIST_HEAD(&new_igroup->device_list);
 	/* group reference moves into new_igroup */
 	new_igroup->group = group;
 
@@ -277,28 +282,15 @@ static int iommufd_device_setup_msi(struct iommufd_device *idev,
 	return 0;
 }
 
-static bool iommufd_hw_pagetable_has_group(struct iommufd_hw_pagetable *hwpt,
-					   struct iommufd_group *igroup)
-{
-	struct iommufd_device *cur_dev;
-
-	lockdep_assert_held(&hwpt->devices_lock);
-
-	list_for_each_entry(cur_dev, &hwpt->devices, devices_item)
-		if (cur_dev->igroup->group == igroup->group)
-			return true;
-	return false;
-}
-
 int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 				struct iommufd_device *idev)
 {
 	phys_addr_t sw_msi_start = PHYS_ADDR_MAX;
 	int rc;
 
-	lockdep_assert_held(&hwpt->devices_lock);
+	lockdep_assert_held(&idev->igroup->lock);
 
-	if (WARN_ON(idev->hwpt))
+	if (idev->igroup->hwpt != NULL && idev->igroup->hwpt != hwpt)
 		return -EINVAL;
 
 	/*
@@ -313,7 +305,7 @@ int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 				hwpt->domain->ops->enforce_cache_coherency(
 					hwpt->domain);
 		if (!hwpt->enforce_cache_coherency) {
-			WARN_ON(list_empty(&hwpt->devices));
+			WARN_ON(list_empty(&idev->igroup->device_list));
 			return -EINVAL;
 		}
 	}
@@ -329,26 +321,40 @@ int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 		goto err_unresv;
 
 	/*
-	 * FIXME: Hack around missing a device-centric iommu api, only attach to
-	 * the group once for the first device that is in the group.
+	 * Only attach to the group once for the first device that is in the
+	 * group. All the other devices will follow this attachment. The user
+	 * should attach every device individually to as the per-device reserved
+	 * regions are only updated during individual device attachment.
 	 */
-	if (!iommufd_hw_pagetable_has_group(hwpt, idev->igroup)) {
+	if (list_empty(&idev->igroup->device_list)) {
 		rc = iommu_attach_group(hwpt->domain, idev->igroup->group);
 		if (rc)
 			goto err_unresv;
+		idev->igroup->hwpt = hwpt;
 	}
+	refcount_inc(&hwpt->obj.users);
+	list_add_tail(&idev->group_item, &idev->igroup->device_list);
 	return 0;
 err_unresv:
 	iopt_remove_reserved_iova(&hwpt->ioas->iopt, idev->dev);
 	return rc;
 }
 
-void iommufd_hw_pagetable_detach(struct iommufd_hw_pagetable *hwpt,
-				 struct iommufd_device *idev)
+struct iommufd_hw_pagetable *
+iommufd_hw_pagetable_detach(struct iommufd_device *idev)
 {
-	if (!iommufd_hw_pagetable_has_group(hwpt, idev->igroup))
+	struct iommufd_hw_pagetable *hwpt = idev->igroup->hwpt;
+
+	lockdep_assert_held(&idev->igroup->lock);
+
+	list_del(&idev->group_item);
+	if (list_empty(&idev->igroup->device_list)) {
 		iommu_detach_group(hwpt->domain, idev->igroup->group);
+		idev->igroup->hwpt = NULL;
+	}
 	iopt_remove_reserved_iova(&hwpt->ioas->iopt, idev->dev);
+	/* Caller must destroy hwpt */
+	return hwpt;
 }
 
 static int iommufd_device_do_attach(struct iommufd_device *idev,
@@ -356,16 +362,9 @@ static int iommufd_device_do_attach(struct iommufd_device *idev,
 {
 	int rc;
 
-	mutex_lock(&hwpt->devices_lock);
+	mutex_lock(&idev->igroup->lock);
 	rc = iommufd_hw_pagetable_attach(hwpt, idev);
-	if (rc)
-		goto out_unlock;
-
-	idev->hwpt = hwpt;
-	refcount_inc(&hwpt->obj.users);
-	list_add(&idev->devices_item, &hwpt->devices);
-out_unlock:
-	mutex_unlock(&hwpt->devices_lock);
+	mutex_unlock(&idev->igroup->lock);
 	return rc;
 }
 
@@ -375,7 +374,7 @@ static int iommufd_device_do_attach(struct iommufd_device *idev,
  * Automatic domain selection will never pick a manually created domain.
  */
 static int iommufd_device_auto_get_domain(struct iommufd_device *idev,
-					  struct iommufd_ioas *ioas)
+					  struct iommufd_ioas *ioas, u32 *pt_id)
 {
 	struct iommufd_hw_pagetable *hwpt;
 	int rc;
@@ -402,6 +401,7 @@ static int iommufd_device_auto_get_domain(struct iommufd_device *idev,
 		 */
 		if (rc == -EINVAL)
 			continue;
+		*pt_id = hwpt->obj.id;
 		goto out_unlock;
 	}
 
@@ -411,6 +411,7 @@ static int iommufd_device_auto_get_domain(struct iommufd_device *idev,
 		goto out_unlock;
 	}
 	hwpt->auto_domain = true;
+	*pt_id = hwpt->obj.id;
 
 	mutex_unlock(&ioas->mutex);
 	iommufd_object_finalize(idev->ictx, &hwpt->obj);
@@ -455,7 +456,7 @@ int iommufd_device_attach(struct iommufd_device *idev, u32 *pt_id)
 		struct iommufd_ioas *ioas =
 			container_of(pt_obj, struct iommufd_ioas, obj);
 
-		rc = iommufd_device_auto_get_domain(idev, ioas);
+		rc = iommufd_device_auto_get_domain(idev, ioas, pt_id);
 		if (rc)
 			goto out_put_pt_obj;
 		break;
@@ -466,7 +467,6 @@ int iommufd_device_attach(struct iommufd_device *idev, u32 *pt_id)
 	}
 
 	refcount_inc(&idev->obj.users);
-	*pt_id = idev->hwpt->obj.id;
 	rc = 0;
 
 out_put_pt_obj:
@@ -484,13 +484,11 @@ EXPORT_SYMBOL_NS_GPL(iommufd_device_attach, IOMMUFD);
  */
 void iommufd_device_detach(struct iommufd_device *idev)
 {
-	struct iommufd_hw_pagetable *hwpt = idev->hwpt;
+	struct iommufd_hw_pagetable *hwpt;
 
-	mutex_lock(&hwpt->devices_lock);
-	list_del(&idev->devices_item);
-	idev->hwpt = NULL;
-	iommufd_hw_pagetable_detach(hwpt, idev);
-	mutex_unlock(&hwpt->devices_lock);
+	mutex_lock(&idev->igroup->lock);
+	hwpt = iommufd_hw_pagetable_detach(idev);
+	mutex_unlock(&idev->igroup->lock);
 
 	if (hwpt->auto_domain)
 		iommufd_object_destroy_user(idev->ictx, &hwpt->obj);
diff --git a/drivers/iommu/iommufd/hw_pagetable.c b/drivers/iommu/iommufd/hw_pagetable.c
index 6cdb6749d359f3..566eba0cd9b917 100644
--- a/drivers/iommu/iommufd/hw_pagetable.c
+++ b/drivers/iommu/iommufd/hw_pagetable.c
@@ -11,8 +11,6 @@ void iommufd_hw_pagetable_destroy(struct iommufd_object *obj)
 	struct iommufd_hw_pagetable *hwpt =
 		container_of(obj, struct iommufd_hw_pagetable, obj);
 
-	WARN_ON(!list_empty(&hwpt->devices));
-
 	if (!list_empty(&hwpt->hwpt_item)) {
 		mutex_lock(&hwpt->ioas->mutex);
 		list_del(&hwpt->hwpt_item);
@@ -25,7 +23,6 @@ void iommufd_hw_pagetable_destroy(struct iommufd_object *obj)
 		iommu_domain_free(hwpt->domain);
 
 	refcount_dec(&hwpt->ioas->obj.users);
-	mutex_destroy(&hwpt->devices_lock);
 }
 
 /**
@@ -52,9 +49,7 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
 	if (IS_ERR(hwpt))
 		return hwpt;
 
-	INIT_LIST_HEAD(&hwpt->devices);
 	INIT_LIST_HEAD(&hwpt->hwpt_item);
-	mutex_init(&hwpt->devices_lock);
 	/* Pairs with iommufd_hw_pagetable_destroy() */
 	refcount_inc(&ioas->obj.users);
 	hwpt->ioas = ioas;
@@ -65,13 +60,16 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
 		goto out_abort;
 	}
 
-	mutex_lock(&hwpt->devices_lock);
+	mutex_lock(&idev->igroup->lock);
 
 	/*
 	 * immediate_attach exists only to accommodate iommu drivers that cannot
 	 * directly allocate a domain. These drivers do not finish creating the
 	 * domain until attach is completed. Thus we must have this call
 	 * sequence. Once those drivers are fixed this should be removed.
+	 *
+	 * Note we hold the igroup->lock here which prevents any other thread
+	 * from observing igroup->hwpt until we finish setting it up.
 	 */
 	if (immediate_attach) {
 		rc = iommufd_hw_pagetable_attach(hwpt, idev);
@@ -84,21 +82,14 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
 		goto out_detach;
 	list_add_tail(&hwpt->hwpt_item, &hwpt->ioas->hwpt_list);
 
-	if (immediate_attach) {
-		/* See iommufd_device_do_attach() */
-		refcount_inc(&hwpt->obj.users);
-		idev->hwpt = hwpt;
-		list_add(&idev->devices_item, &hwpt->devices);
-	}
-
-	mutex_unlock(&hwpt->devices_lock);
+	mutex_unlock(&idev->igroup->lock);
 	return hwpt;
 
 out_detach:
 	if (immediate_attach)
-		iommufd_hw_pagetable_detach(hwpt, idev);
+		iommufd_hw_pagetable_detach(idev);
 out_unlock:
-	mutex_unlock(&hwpt->devices_lock);
+	mutex_unlock(&idev->igroup->lock);
 out_abort:
 	iommufd_object_abort_and_destroy(ictx, &hwpt->obj);
 	return ERR_PTR(rc);
diff --git a/drivers/iommu/iommufd/iommufd_private.h b/drivers/iommu/iommufd/iommufd_private.h
index 2544f10dae9aef..2ff192777f27d3 100644
--- a/drivers/iommu/iommufd/iommufd_private.h
+++ b/drivers/iommu/iommufd/iommufd_private.h
@@ -250,8 +250,6 @@ struct iommufd_hw_pagetable {
 	bool msi_cookie : 1;
 	/* Head at iommufd_ioas::hwpt_list */
 	struct list_head hwpt_item;
-	struct mutex devices_lock;
-	struct list_head devices;
 };
 
 struct iommufd_hw_pagetable *
@@ -259,14 +257,17 @@ iommufd_hw_pagetable_alloc(struct iommufd_ctx *ictx, struct iommufd_ioas *ioas,
 			   struct iommufd_device *idev, bool immediate_attach);
 int iommufd_hw_pagetable_attach(struct iommufd_hw_pagetable *hwpt,
 				struct iommufd_device *idev);
-void iommufd_hw_pagetable_detach(struct iommufd_hw_pagetable *hwpt,
-				 struct iommufd_device *idev);
+struct iommufd_hw_pagetable *
+iommufd_hw_pagetable_detach(struct iommufd_device *idev);
 void iommufd_hw_pagetable_destroy(struct iommufd_object *obj);
 
 struct iommufd_group {
 	struct kref ref;
+	struct mutex lock;
 	struct iommufd_ctx *ictx;
 	struct iommu_group *group;
+	struct iommufd_hw_pagetable *hwpt;
+	struct list_head device_list;
 };
 
 /*
@@ -278,9 +279,7 @@ struct iommufd_device {
 	struct iommufd_object obj;
 	struct iommufd_ctx *ictx;
 	struct iommufd_group *igroup;
-	struct iommufd_hw_pagetable *hwpt;
-	/* Head at iommufd_hw_pagetable::devices */
-	struct list_head devices_item;
+	struct list_head group_item;
 	/* always the physical device */
 	struct device *dev;
 	bool enforce_cache_coherency;
-- 
2.40.0


  parent reply	other threads:[~2023-03-21 19:16 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-03-21 19:14 [PATCH v3 00/17] Add iommufd physical device operations for replace and alloc hwpt Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 01/17] iommufd: Move isolated msi enforcement to iommufd_device_bind() Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 02/17] iommufd: Add iommufd_group Jason Gunthorpe
2023-03-21 19:14 ` Jason Gunthorpe [this message]
2023-03-23  7:21   ` [PATCH v3 03/17] iommufd: Replace the hwpt->devices list with iommufd_group Tian, Kevin
2023-03-23 14:23     ` Jason Gunthorpe
2023-03-24  1:37       ` Tian, Kevin
2023-03-24 15:02         ` Jason Gunthorpe
2023-03-28  2:32           ` Tian, Kevin
2023-03-28 11:38             ` Jason Gunthorpe
2023-03-29  3:03               ` Tian, Kevin
2023-04-11 14:31     ` Jason Gunthorpe
2023-04-12  8:27       ` Tian, Kevin
2023-04-12 11:17         ` Jason Gunthorpe
2023-04-13  2:52           ` Tian, Kevin
2023-04-14 13:31             ` Jason Gunthorpe
2023-04-20  6:15               ` Tian, Kevin
2023-04-20 15:34                 ` Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 04/17] iommu: Export iommu_get_resv_regions() Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 05/17] iommufd: Keep track of each device's reserved regions instead of groups Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 06/17] iommufd: Use the iommufd_group to avoid duplicate MSI setup Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 07/17] iommufd: Make sw_msi_start a group global Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 08/17] iommufd: Move putting a hwpt to a helper function Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 09/17] iommufd: Add enforced_cache_coherency to iommufd_hw_pagetable_alloc() Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 10/17] iommufd: Reorganize iommufd_device_attach into iommufd_device_change_pt Jason Gunthorpe
2023-03-23  7:25   ` Tian, Kevin
2023-03-23 14:26     ` Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 11/17] iommu: Introduce a new iommu_group_replace_domain() API Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 12/17] iommufd: Add iommufd_device_replace() Jason Gunthorpe
2023-03-23  7:31   ` Tian, Kevin
2023-03-23 14:30     ` Jason Gunthorpe
2023-03-24  1:42       ` Tian, Kevin
2023-03-24 15:03         ` Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 13/17] iommufd: Make destroy_rwsem use a lock class per object type Jason Gunthorpe
2023-03-23  7:54   ` Tian, Kevin
2023-03-21 19:14 ` [PATCH v3 14/17] iommufd/selftest: Test iommufd_device_replace() Jason Gunthorpe
2023-03-23  7:57   ` Tian, Kevin
2023-03-23 14:32     ` Jason Gunthorpe
2023-03-21 19:14 ` [PATCH v3 15/17] iommufd: Add IOMMU_HWPT_ALLOC Jason Gunthorpe
2023-03-23  8:00   ` Tian, Kevin
2023-03-21 19:14 ` [PATCH v3 16/17] iommufd/selftest: Return the real idev id from selftest mock_domain Jason Gunthorpe
2023-03-23  8:02   ` Tian, Kevin
2023-03-21 19:14 ` [PATCH v3 17/17] iommufd/selftest: Add a selftest for IOMMU_HWPT_ALLOC Jason Gunthorpe
2023-03-23  8:03   ` Tian, Kevin
2023-03-23  8:04 ` [PATCH v3 00/17] Add iommufd physical device operations for replace and alloc hwpt Tian, Kevin
2023-03-23 14:35   ` Jason Gunthorpe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3-v3-61d41fd9e13e+1f5-iommufd_alloc_jgg@nvidia.com \
    --to=jgg@nvidia.com \
    --cc=baolu.lu@linux.intel.com \
    --cc=iommu@lists.linux.dev \
    --cc=kevin.tian@intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=nicolinc@nvidia.com \
    --cc=yi.l.liu@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.