Skip to contentSkip to conversation

genirq: call cancel_work_sync from irq_set_affinity_notifier

PATCH In review Prasad Sodagudi 3 participants 5 messages Mar 20, 2019 – Mar 24, 2019 IRQ

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 genirq: call cancel_work_sync from irq_set_affinity_notifier

When ever notification of IRQ affinity changes, call cancel_work_sync from irq_set_affinity_notifier to cancel all pending works to avoid work list corruption.

---
 kernel/irq/manage.c | 3 +++
 1 file changed, 3 insertions(+)
kernel/irq/manage.c +3 -0
index 9ec34a2..da8b2ee 100644
@@ -356,6 +356,9 @@ static void irq_affinity_notify(struct work_struct *work)
 	desc->affinity_notify = notify;
 	raw_spin_unlock_irqrestore(&desc->lock, flags);
 
+	if (!notify && old_notify)
+		cancel_work_sync(&old_notify->work);
+
 	if (old_notify)
 		kref_put(&old_notify->kref, old_notify->release);
 
Signature · 1 line
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,\na Linux Foundation Collaborative Project
#1 lore raw

Prasad,

Quoted 1 line · Prasad Sodagudi

Subject: [PATCH] genirq: call cancel_work_sync from irq_set_affinity_notifier

Please do not decribe WHAT the code change is. Give a consice explanation WHY this change is done. The above is like '[PATCH] foo: Increment bar by 5'.

[PATCH] genirq: Prevent UAF and work list corruption

Quoted 3 lines

When ever notification of IRQ affinity changes, call cancel_work_sync from irq_set_affinity_notifier to cancel all pending works to avoid work list corruption.

Again, you describe first WHAT you are doing instead of telling WHY.

When irq_set_affinity_notifier() replaces the notifier, then the reference count on the old notifier is dropped which causes it to be freed. But nothing ensures that the old notifier is not longer queued in the work list. If it is queued this results in a use after free and possibly in work list corruption.

Ensure that the work is canceled before the reference is dropped.

See?

This gives precise context first and then describes the cure.

Also it is completely irrelevant whether this is achieved by calling cancel_work_sync() or by something else. What matters is that it's canceled. Changelogs describe context and concepts not implementation details. The implementation details are in the patch itself.

Quoted 10 lines
---
 kernel/irq/manage.c | 3 +++
 1 file changed, 3 insertions(+)
kernel/irq/manage.c +0 -0
index 9ec34a2..da8b2ee 100644

kernel/irq/manage.c static void irq_affinity_notify(struct work_struct *work) · line 360

Quoted 5 lines
kernel/irq/manage.c +2 -0
@@ -356,6 +356,9 @@ static void irq_affinity_notify(struct work_struct *work)
 	desc->affinity_notify = notify;
 	raw_spin_unlock_irqrestore(&desc->lock, flags);
 
+	if (!notify && old_notify)
+		cancel_work_sync(&old_notify->work);

That '!notify' doesn't make any sense.

Thanks,

tglx

#2 lore raw
Quoted 24 lines · Thomas Gleixner

Prasad,

Quoted 1 line · Prasad Sodagudi

Subject: [PATCH] genirq: call cancel_work_sync from irq_set_affinity_notifier

Please do not decribe WHAT the code change is. Give a consice explanation WHY this change is done. The above is like '[PATCH] foo: Increment bar by 5'.

[PATCH] genirq: Prevent UAF and work list corruption

Quoted 3 lines

When ever notification of IRQ affinity changes, call cancel_work_sync from irq_set_affinity_notifier to cancel all pending works to avoid work list corruption.

Again, you describe first WHAT you are doing instead of telling WHY.

When irq_set_affinity_notifier() replaces the notifier, then the reference count on the old notifier is dropped which causes it to be freed. But nothing ensures that the old notifier is not longer queued
in
the work list. If it is queued this results in a use after free and possibly in work list corruption.

Quoted 3 lines

Ensure that the work is canceled before the reference is dropped.

See?

Hi Tglx,

Thanks for suggesting commit text and modifications.

Quoted 22 lines

This gives precise context first and then describes the cure.

Also it is completely irrelevant whether this is achieved by calling cancel_work_sync() or by something else. What matters is that it's canceled. Changelogs describe context and concepts not implementation details. The implementation details are in the patch itself.

Quoted 14 lines
---
 kernel/irq/manage.c | 3 +++
 1 file changed, 3 insertions(+)
kernel/irq/manage.c +0 -0
index 9ec34a2..da8b2ee 100644
@@ -356,6 +356,9 @@ static void irq_affinity_notify(struct work_struct 
 *work)
 	desc->affinity_notify = notify;
 	raw_spin_unlock_irqrestore(&desc->lock, flags);
 
Quoted 4 lines

kernel/irq/manage.c static void irq_affinity_notify(struct work_struct *work) · line 360

Quoted 2 lines
kernel/irq/manage.c +2 -0
+	if (!notify && old_notify)
+		cancel_work_sync(&old_notify->work);

That '!notify' doesn't make any sense.

Yes. I will remove this in the next patch set. Thanks for reviewing.

-thanks, Prasad

Quoted 4 lines

Thanks,

tglx

Signature · 2 lines
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
Linux Foundation Collaborative Project
#3 lore raw

PATCH v2 1/1 genirq: Prevent use-after-free and work list corruption

When irq_set_affinity_notifier() replaces the notifier, then the reference count on the old notifier is dropped which causes it to be freed. But nothing ensures that the old notifier is not longer queued in the work list. If it is queued this results in a use after free and possibly in work list corruption.

Ensure that the work is canceled before the reference is dropped.

---
 kernel/irq/manage.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
kernel/irq/manage.c +3 -1
index 9ec34a2..1a1ac84 100644
@@ -356,8 +356,10 @@ static void irq_affinity_notify(struct work_struct *work)
 	desc->affinity_notify = notify;
 	raw_spin_unlock_irqrestore(&desc->lock, flags);
 
-	if (old_notify)
+	if (old_notify) {
+		cancel_work_sync(&old_notify->work);
 		kref_put(&old_notify->kref, old_notify->release);
+	}
 
 	return 0;
 }
Signature · 1 line
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,\na Linux Foundation Collaborative Project
#4 lore raw

[tip:irq/core] genirq: Prevent use-after-free and work list corruption

genirq: Prevent use-after-free and work list corruption
Commit-ID:  59c39840f5abf4a71e1810a8da71aaccd6c17d26
Gitweb:     https://git.kernel.org/tip/59c39840f5abf4a71e1810a8da71aaccd6c17d26
Author:     Prasad Sodagudi <[email protected]>
AuthorDate: Sun, 24 Mar 2019 07:57:04 -0700
Committer:  Thomas Gleixner <[email protected]>
CommitDate: Sun, 24 Mar 2019 22:13:17 +0100

genirq: Prevent use-after-free and work list corruption

When irq_set_affinity_notifier() replaces the notifier, then the reference count on the old notifier is dropped which causes it to be freed. But nothing ensures that the old notifier is not longer queued in the work list. If it is queued this results in a use after free and possibly in work list corruption.

Ensure that the work is canceled before the reference is dropped.

---
 kernel/irq/manage.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
kernel/irq/manage.c +3 -1
index 1401afa0d58a..53a081392115 100644
@@ -357,8 +357,10 @@ irq_set_affinity_notifier(unsigned int irq, struct irq_affinity_notify *notify)
 	desc->affinity_notify = notify;
 	raw_spin_unlock_irqrestore(&desc->lock, flags);
 
-	if (old_notify)
+	if (old_notify) {
+		cancel_work_sync(&old_notify->work);
 		kref_put(&old_notify->kref, old_notify->release);
+	}
 
 	return 0;
 }