Skip to contentSkip to conversation

iommu: Handle iommu faults for a bad iopf setup

PATCH Superseded Pranjal Shrivastava 2 participants 5 messages Aug 15, 2024

Subsystems (2)

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 iommu: Handle iommu faults for a bad iopf setup

The iommu_report_device_fault function was updated to return void while assuming that drivers only need to call iommu_report_device_fault() for reporting an iopf. This implementation causes following problems:

1. The drivers rely on the core code to call it's page_reponse, however, when a fault is received and no fault capable domain is attached / iopf_param is NULL, the ops->page_response is NOT called causing the device to stall in case the fault type was PAGE_REQ.

2. The arm_smmu_v3 driver relies on the returned value to log errors returning void from iommu_report_device_fault causes these events to be missed while logging.

Modify the iommu_report_device_fault function to return -EINVAL for cases where no fault capable domain is attached or iopf_param was NULL and calls back to the driver (ops->page_response) in case the fault type was IOMMU_FAULT_PAGE_REQ. The returned value can be used by the drivers to log the fault/event as needed.

Fixes 3dfa64aecbaf ("iommu: Make iommu_report_device_fault() return void")
---
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c |   2 +-
 drivers/iommu/io-pgfault.c                  | 116 ++++++++++++++------
 include/linux/iommu.h                       |   5 +-
 3 files changed, 84 insertions(+), 39 deletions(-)
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +1 -1
index 9bc50bded5af..8a6cd0adfcf2 100644
@@ -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/io-pgfault.c +80 -36

153 lines not on this page · raw message

include/linux/iommu.h +3 -2
index 04cbdae0052e..bd722f473635 100644
@@ -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,
Signature · 1 line
2.46.0.184.g6999bdac58-goog
#1 lore raw
Quoted 131 lines · Pranjal Shrivastava

The iommu_report_device_fault function was updated to return void while assuming that drivers only need to call iommu_report_device_fault() for reporting an iopf. This implementation causes following problems:

1. The drivers rely on the core code to call it's page_reponse, however, when a fault is received and no fault capable domain is attached / iopf_param is NULL, the ops->page_response is NOT called causing the device to stall in case the fault type was PAGE_REQ.

2. The arm_smmu_v3 driver relies on the returned value to log errors returning void from iommu_report_device_fault causes these events to be missed while logging.

Modify the iommu_report_device_fault function to return -EINVAL for cases where no fault capable domain is attached or iopf_param was NULL and calls back to the driver (ops->page_response) in case the fault type was IOMMU_FAULT_PAGE_REQ. The returned value can be used by the drivers to log the fault/event as needed.

Fixes 3dfa64aecbaf ("iommu: Make iommu_report_device_fault() return void")
---
 drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c |   2 +-
 drivers/iommu/io-pgfault.c                  | 116 ++++++++++++++------
 include/linux/iommu.h                       |   5 +-
 3 files changed, 84 insertions(+), 39 deletions(-)
drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c +1 -1
index 9bc50bded5af..8a6cd0adfcf2 100644
@@ -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/io-pgfault.c +66 -1
index 81e9cc6e3164..5a35da8fa78f 100644
@@ -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
@@ -155,16 +208,30 @@ static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param,
  * hardware has been set to block the page faults) and the pending page faults
  * have been flushed.
  */
-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;
+	int ret = 0;
+
+	attach_handle = find_fault_handler(dev, evt);
+	if (!attach_handle) {
+		ret = -EINVAL;
+		goto err_bad_iopf;
+	}
 
+	/*
+	 * Something has gone wrong if a fault capable domain is attached but no
+	 * iopf_param is setup
+	 */

drivers/iommu/io-pgfault.c static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param, · line 237

Quoted 10 lines
drivers/iommu/io-pgfault.c +4 -2
 	iopf_param = iopf_get_dev_fault_param(dev);
-	if (WARN_ON(!iopf_param))
-		return;
+	if (WARN_ON(!iopf_param)) {
+		ret = -EINVAL;
+		goto err_bad_iopf;
+	}
 
 	if (!(fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_LAST_PAGE)) {
 		report_partial_fault(iopf_param, fault);

Apologies, I sent out an older version that missed returning a value. Please ignore this email, I'll resend the updated version.

Quoted 92 lines
drivers/iommu/io-pgfault.c +10 -33
@@ -185,38 +252,7 @@ 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 = 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;
+	group->attach_handle = attach_handle;
 
 	/*
 	 * On success iopf_handler must call iopf_group_response() and
@@ -225,7 +261,7 @@ 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 +271,14 @@ void iommu_report_device_fault(struct device *dev, struct iopf_fault *evt)
 		__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 ret;
 }
 EXPORT_SYMBOL_GPL(iommu_report_device_fault);
 
include/linux/iommu.h +3 -2
index 04cbdae0052e..bd722f473635 100644
@@ -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,
Signature · 1 line
2.46.0.184.g6999bdac58-goog

Thanks,
Pranjal

Reviewed-by Lu Baolu
#2 lore raw
Quoted 5 lines · Pranjal Shrivastava

drivers/iommu/io-pgfault.c static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param, · line 237

Quoted 2 lines
drivers/iommu/io-pgfault.c +0 -0
 	if (!(fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_LAST_PAGE)) {
 		report_partial_fault(iopf_param, fault);

Apologies, I sent out an older version that missed returning a value. Please ignore this email, I'll resend the updated version.

It is probably Ok, the merge should resolve it.

The fix is already in linux-next:

commit fca5b78511e98bdff2cdd55c172b23200a7b3404

Author: Barak Biber <[email protected]>
Date:   Thu Aug 1 09:26:04 2024 -0300

iommu: Restore lost return in iommu_report_device_fault()

When iommu_report_device_fault gets called with a partial fault it is supposed to collect the fault into the group and then return.

Instead the return was accidently deleted which results in trying to process the fault and an eventual crash.

Deleting the return was a typo, put it back.

Fixes 3dfa64aecbaf ("iommu: Make iommu_report_device_fault() return void")

But doesn't look like it made it to rc3 yet

Jason

#3 lore raw
Quoted 34 lines · Jason Gunthorpe
Quoted 5 lines · Pranjal Shrivastava

drivers/iommu/io-pgfault.c static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param, · line 237

Quoted 2 lines
drivers/iommu/io-pgfault.c +0 -0
 	if (!(fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_LAST_PAGE)) {
 		report_partial_fault(iopf_param, fault);

Apologies, I sent out an older version that missed returning a value. Please ignore this email, I'll resend the updated version.

It is probably Ok, the merge should resolve it.

The fix is already in linux-next:

commit fca5b78511e98bdff2cdd55c172b23200a7b3404

Author: Barak Biber <[email protected]>
Date:   Thu Aug 1 09:26:04 2024 -0300

iommu: Restore lost return in iommu_report_device_fault()

When iommu_report_device_fault gets called with a partial fault it is supposed to collect the fault into the group and then return.

Instead the return was accidently deleted which results in trying to process the fault and an eventual crash.

Deleting the return was a typo, put it back.

Fixes 3dfa64aecbaf ("iommu: Make iommu_report_device_fault() return void")
Quoted 1 line

But doesn't look like it made it to rc3 yet

I guess it's there on as it gave me a build error as the return; statement was returning void for a non-void function. Fixed this and sent it as a v2 of this patch.

Quoted 2 lines

Jason

Thanks,
Pranjal

#4 lore raw
Quoted 35 lines · Pranjal Shrivastava
Quoted 34 lines · Jason Gunthorpe
Quoted 5 lines · Pranjal Shrivastava

drivers/iommu/io-pgfault.c static struct iopf_group *iopf_group_alloc(struct iommu_fault_param *iopf_param, · line 237

Quoted 2 lines
drivers/iommu/io-pgfault.c +0 -0
 	if (!(fault->prm.flags & IOMMU_FAULT_PAGE_REQUEST_LAST_PAGE)) {
 		report_partial_fault(iopf_param, fault);

Apologies, I sent out an older version that missed returning a value. Please ignore this email, I'll resend the updated version.

It is probably Ok, the merge should resolve it.

The fix is already in linux-next:

commit fca5b78511e98bdff2cdd55c172b23200a7b3404

Author: Barak Biber <[email protected]>
Date:   Thu Aug 1 09:26:04 2024 -0300

iommu: Restore lost return in iommu_report_device_fault()

When iommu_report_device_fault gets called with a partial fault it is supposed to collect the fault into the group and then return.

Instead the return was accidently deleted which results in trying to process the fault and an eventual crash.

Deleting the return was a typo, put it back.

Fixes 3dfa64aecbaf ("iommu: Make iommu_report_device_fault() return void")
Quoted 5 lines
Quoted 1 line

But doesn't look like it made it to rc3 yet

I guess it's there on as it gave me a build error as the return; statement was returning void for a non-void function. Fixed this and sent it as a v2 of this patch.

Oh I see now, yes, that is the right thing to do!

You should also put [PATCH rc] so Joerg knows this is for rc as we are fixing a regression in this merge window.

Jason