Skip to contentSkip to conversation

iommu: Introduce reset_mutex to avoid circular locking dependency

PATCH In review David Matlack 1 participant 2 messages Aug 27 IOMMU

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 iommu: Introduce reset_mutex to avoid circular locking dependency

Introduce a dedicated reset_mutex inside struct iommu_group to serialize device reset operations and domain resetting state without acquiring group->mutex.

Commit f5b16b802174 ("PCI: Suspend iommu function prior to resetting a device") introduced pci_dev_reset_iommu_prepare() and pci_dev_reset_iommu_done(), which acquire group->mutex during PCI device resets. In drivers such as VFIO, device reset handlers (e.g. pci_try_reset_function()) are executed while holding driver locks such as vdev->memory_lock.

This introduces a circular locking dependency that can lead to tasks being permanently stuck in a deadlock:

[vdev->memory_lock] ──────(1. VFIO Reset)──────> [iommu_group->mutex]
▲
│ │
(3. VFIO BAR Page Fault) ... [*]

         │                                                    │
         │                                                    │
         └───────────  [mm->mmap_lock]  <─────────────────────┘

Fix this by using group->reset_mutex instead of group->mutex in pci_dev_reset_iommu_prepare() and pci_dev_reset_iommu_done(). To ensure other iommu_group operations stay properly synchronized with resets, also take group->reset_mutex when checking group->recovery_cnt.

[*] The iommu_group->mutex --> mm->mmap_lock dependency was reported as a transitive chain: iommu_group->mutex --> cpu_hotplug_lock --> i_mutex_dir_key --> mm->mmap_lock.

Fixes f5b16b802174 ("PCI: Suspend iommu function prior to resetting a device")

---
Cc: Alex Williamson <[email protected]>

 drivers/iommu/iommu.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/iommu/iommu.c +7 -2
index e8f13dcebbde..29e63697211f 100644
@@ -56,6 +56,7 @@ struct iommu_group {
 	struct list_head devices;
 	struct xarray pasid_array;
 	struct mutex mutex;
+	struct mutex reset_mutex;
 	void *iommu_data;
 	void (*iommu_data_release)(void *iommu_data);
 	char *name;
@@ -1080,6 +1081,7 @@ struct iommu_group *iommu_group_alloc(void)
 
 	group->kobj.kset = iommu_group_kset;
 	mutex_init(&group->mutex);
+	mutex_init(&group->reset_mutex);
 	INIT_LIST_HEAD(&group->devices);
 	INIT_LIST_HEAD(&group->entry);
 	xa_init(&group->pasid_array);
@@ -2476,6 +2478,7 @@ static int __iommu_group_set_domain_internal(struct iommu_group *group,
 	 * pci_dev_reset_iommu_done() attaches the device to group->domain, if
 	 * IOMMU_SET_DOMAIN_MUST_SUCCEED is not set.
 	 */
+	guard(mutex)(&group->reset_mutex);
 	if (group->recovery_cnt && !(flags & IOMMU_SET_DOMAIN_MUST_SUCCEED))
 		return -EBUSY;
 
@@ -3652,6 +3655,7 @@ int iommu_attach_device_pasid(struct iommu_domain *domain,
 	 * This is a concurrent attach during device recovery. Reject it until
 	 * pci_dev_reset_iommu_done() attaches the device to group->domain.
 	 */
+	guard(mutex)(&group->reset_mutex);
 	if (group->recovery_cnt) {
 		ret = -EBUSY;
 		goto out_unlock;
@@ -3745,6 +3749,7 @@ int iommu_replace_device_pasid(struct iommu_domain *domain,
 	 * This is a concurrent attach during device recovery. Reject it until
 	 * pci_dev_reset_iommu_done() attaches the device to group->domain.
 	 */
+	guard(mutex)(&group->reset_mutex);
 	if (group->recovery_cnt) {
 		ret = -EBUSY;
 		goto out_unlock;
@@ -4042,7 +4047,7 @@ int pci_dev_reset_iommu_prepare(struct pci_dev *pdev)
 	if (!pci_ats_supported(pdev) || !dev_has_iommu(&pdev->dev))
 		return 0;
 
-	guard(mutex)(&group->mutex);
+	guard(mutex)(&group->reset_mutex);
 
 	gdev = __dev_to_gdev(&pdev->dev);
 	if (WARN_ON(!gdev))
@@ -4153,7 +4158,7 @@ void pci_dev_reset_iommu_done(struct pci_dev *pdev)
 	if (!pci_ats_supported(pdev) || !dev_has_iommu(&pdev->dev))
 		return;
 
-	guard(mutex)(&group->mutex);
+	guard(mutex)(&group->reset_mutex);
 
 	gdev = __dev_to_gdev(&pdev->dev);
 	if (WARN_ON(!gdev))

base-commit: 37ffa24c9d07edcd414d34283e02af3f3866cf12

Signature · 1 line
2.55.0.897.gb25b4bd76c-goog
#1 lore raw
Quoted 96 lines · David Matlack

Introduce a dedicated reset_mutex inside struct iommu_group to serialize device reset operations and domain resetting state without acquiring group->mutex.

Commit f5b16b802174 ("PCI: Suspend iommu function prior to resetting a device") introduced pci_dev_reset_iommu_prepare() and pci_dev_reset_iommu_done(), which acquire group->mutex during PCI device resets. In drivers such as VFIO, device reset handlers (e.g. pci_try_reset_function()) are executed while holding driver locks such as vdev->memory_lock.

This introduces a circular locking dependency that can lead to tasks being permanently stuck in a deadlock:

[vdev->memory_lock] ──────(1. VFIO Reset)──────> [iommu_group->mutex]
▲
│ │
(3. VFIO BAR Page Fault) ... [*]

         │                                                    │
         │                                                    │
         └───────────  [mm->mmap_lock]  <─────────────────────┘

Fix this by using group->reset_mutex instead of group->mutex in pci_dev_reset_iommu_prepare() and pci_dev_reset_iommu_done(). To ensure other iommu_group operations stay properly synchronized with resets, also take group->reset_mutex when checking group->recovery_cnt.

[*] The iommu_group->mutex --> mm->mmap_lock dependency was reported as a transitive chain: iommu_group->mutex --> cpu_hotplug_lock --> i_mutex_dir_key --> mm->mmap_lock.

Fixes f5b16b802174 ("PCI: Suspend iommu function prior to resetting a device")

---
Cc: Alex Williamson <[email protected]>

 drivers/iommu/iommu.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/iommu/iommu.c +6 -1
index e8f13dcebbde..29e63697211f 100644
@@ -56,6 +56,7 @@ struct iommu_group {
        struct list_head devices;
        struct xarray pasid_array;
        struct mutex mutex;
+       struct mutex reset_mutex;
        void *iommu_data;
        void (*iommu_data_release)(void *iommu_data);
        char *name;
@@ -1080,6 +1081,7 @@ struct iommu_group *iommu_group_alloc(void)
 
        group->kobj.kset = iommu_group_kset;
        mutex_init(&group->mutex);
+       mutex_init(&group->reset_mutex);
        INIT_LIST_HEAD(&group->devices);
        INIT_LIST_HEAD(&group->entry);
        xa_init(&group->pasid_array);
@@ -2476,6 +2478,7 @@ static int __iommu_group_set_domain_internal(struct iommu_group *group,
         * pci_dev_reset_iommu_done() attaches the device to group->domain, if
         * IOMMU_SET_DOMAIN_MUST_SUCCEED is not set.
         */
+       guard(mutex)(&group->reset_mutex);
        if (group->recovery_cnt && !(flags & IOMMU_SET_DOMAIN_MUST_SUCCEED))
                return -EBUSY;
 
@@ -3652,6 +3655,7 @@ int iommu_attach_device_pasid(struct iommu_domain *domain,
         * This is a concurrent attach during device recovery. Reject it until
         * pci_dev_reset_iommu_done() attaches the device to group->domain.
         */
+       guard(mutex)(&group->reset_mutex);
        if (group->recovery_cnt) {
                ret = -EBUSY;
                goto out_unlock;
@@ -3745,6 +3749,7 @@ int iommu_replace_device_pasid(struct iommu_domain *domain,
         * This is a concurrent attach during device recovery. Reject it until
         * pci_dev_reset_iommu_done() attaches the device to group->domain.
         */
+       guard(mutex)(&group->reset_mutex);
        if (group->recovery_cnt) {
                ret = -EBUSY;
                goto out_unlock;
@@ -4042,7 +4047,7 @@ int pci_dev_reset_iommu_prepare(struct pci_dev *pdev)
        if (!pci_ats_supported(pdev) || !dev_has_iommu(&pdev->dev))
                return 0;
 
-       guard(mutex)(&group->mutex);
+       guard(mutex)(&group->reset_mutex);
 
        gdev = __dev_to_gdev(&pdev->dev);
        if (WARN_ON(!gdev))

drivers/iommu/iommu.c void pci_dev_reset_iommu_done(struct pci_dev *pdev) · line 4163

Quoted 7 lines
drivers/iommu/iommu.c +1 -1
@@ -4153,7 +4158,7 @@ void pci_dev_reset_iommu_done(struct pci_dev *pdev)
        if (!pci_ats_supported(pdev) || !dev_has_iommu(&pdev->dev))
                return;
 
-       guard(mutex)(&group->mutex);
+       guard(mutex)(&group->reset_mutex);
 
        gdev = __dev_to_gdev(&pdev->dev);

Ah, this does not work. __dev_to_gdev() has a lockdep assert that group->mutex must be held. I'm not sure why my test did not catch that...

Quoted 6 lines
drivers/iommu/iommu.c +0 -0
        if (WARN_ON(!gdev))

base-commit: 37ffa24c9d07edcd414d34283e02af3f3866cf12 --
2.55.0.897.gb25b4bd76c-goog