Skip to contentSkip to conversation

net: bluetooth: Fix null pointer deref in hci_phy_link_complete_evt

PATCH New Anmol Karn 2 participants 7 messages Aug 29, 2020 – Aug 30, 2020

Subsystems (4)
lore
Share & tools
lore

Anmol Karn

author
#0 lore raw

PATCH 1/1 net: bluetooth: Fix null pointer deref in hci_phy_link_complete_evt

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

---
 net/bluetooth/hci_core.c  | 4 ++++
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+)
net/bluetooth/hci_core.c +4 -0
index 68bfe57b6625..533048d2a624 100644
@@ -4922,6 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
+		if(hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}
+
 		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
net/bluetooth/hci_event.c +4 -0
index 4b7fc430793c..c621c8a20ea4 100644
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
 		return;
 	}
+	if(!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}
 
 	bredr_hcon = hcon->amp_mgr->l2cap_conn->hcon;
 
Signature · 6 lines
2.28.0

_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees
author
#1 lore raw

PATCH 1/1 net: bluetooth: Fix null pointer deref in hci_phy_link_complete_evt

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

---
 net/bluetooth/hci_core.c  | 5 ++++-
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+), 1 deletion(-)
net/bluetooth/hci_core.c +4 -1
index 68bfe57b6625..996efd654e7a 100644
@@ -4922,7 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
-		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		if (hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}
+
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
 				hci_dev_set_flag(hdev, HCI_CMD_PENDING);
net/bluetooth/hci_event.c +4 -0
index 4b7fc430793c..1e7d9bee9111 100644
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
 		return;
 	}
+	if (!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}
 
 	bredr_hcon = hcon->amp_mgr->l2cap_conn->hcon;
 
Signature · 6 lines
2.28.0

_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees
#2 lore raw
Quoted 9 lines · Anmol Karn

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

So this is a v2? That should go below the --- line, right? And you should have a v2 in the subject line like the documentation asks?

Can you redo this please?

thanks,

greg k-h
_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees

#3 lore raw
In reply to Anmol Karn
Quoted 23 lines · Anmol Karn

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

---
 net/bluetooth/hci_core.c  | 5 ++++-
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+), 1 deletion(-)
net/bluetooth/hci_core.c +0 -0
index 68bfe57b6625..996efd654e7a 100644

net/bluetooth/hci_core.c static void hci_cmd_work(struct work_struct *work) · line 4927

Quoted 7 lines
net/bluetooth/hci_core.c +3 -1
@@ -4922,7 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
-		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		if (hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}

How can sent_cmd be NULL here? Are you sure something previous to this shouldn't be fixed instead?

Quoted 9 lines
net/bluetooth/hci_core.c +1 -0
+
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
 				hci_dev_set_flag(hdev, HCI_CMD_PENDING);
net/bluetooth/hci_event.c +0 -0
index 4b7fc430793c..1e7d9bee9111 100644

net/bluetooth/hci_event.c static void hci_phy_link_complete_evt(struct hci_dev *hdev, · line 4947

Quoted 7 lines
net/bluetooth/hci_event.c +4 -0
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
 		return;
 	}
+	if (!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}

How can this be triggered?

thanks,

greg k-h

author
#4 lore raw
Quoted 27 lines · Greg KH
Quoted 26 lines · Anmol Karn

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

---
 net/bluetooth/hci_core.c  | 5 ++++-
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+), 1 deletion(-)
net/bluetooth/hci_core.c +0 -0
index 68bfe57b6625..996efd654e7a 100644
@@ -4922,7 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
Quoted 7 lines

net/bluetooth/hci_core.c static void hci_cmd_work(struct work_struct *work) · line 4927

Quoted 4 lines
net/bluetooth/hci_core.c +3 -1
-		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		if (hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}

How can sent_cmd be NULL here? Are you sure something previous to this shouldn't be fixed instead?

Sir, sent_cmd was freed before this condition check, thats why i checked it,

i think i should check it before the free of hdev->sent_cmd like,

if (hdev->sent_cmd)

	kfree_skb(hdev->sent_cmd);

what's your opininon on this.

Quoted 12 lines
Quoted 12 lines
net/bluetooth/hci_core.c +1 -0
+
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
 				hci_dev_set_flag(hdev, HCI_CMD_PENDING);
net/bluetooth/hci_event.c +0 -0
index 4b7fc430793c..1e7d9bee9111 100644
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
Quoted 8 lines

net/bluetooth/hci_event.c static void hci_phy_link_complete_evt(struct hci_dev *hdev, · line 4947

Quoted 6 lines
net/bluetooth/hci_event.c +4 -0
 		return;
 	}
+	if (!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}

How can this be triggered?

syzbot showed that this line is accessed irrespective of the null value it contains, so added a pointer check for that.

please give some opinions on this,

if (!bredr_hcon) {
	hci_dev_unlock(hdev);
        return;
}

Thanks,
Anmol Karn
_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees

#5 lore raw
Quoted 36 lines · Anmol Karn
Quoted 27 lines · Greg KH
Quoted 26 lines · Anmol Karn

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

---
 net/bluetooth/hci_core.c  | 5 ++++-
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+), 1 deletion(-)
net/bluetooth/hci_core.c +0 -0
index 68bfe57b6625..996efd654e7a 100644
@@ -4922,7 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
Quoted 7 lines

net/bluetooth/hci_core.c static void hci_cmd_work(struct work_struct *work) · line 4927

Quoted 4 lines
net/bluetooth/hci_core.c +3 -1
-		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		if (hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}

How can sent_cmd be NULL here? Are you sure something previous to this shouldn't be fixed instead?

Quoted 1 line

Sir, sent_cmd was freed before this condition check, thats why i checked it,

But it can not be NULL at that point in time, as nothing set it to NULL, correct?

Quoted 4 lines

i think i should check it before the free of hdev->sent_cmd like,

if (hdev->sent_cmd)

	kfree_skb(hdev->sent_cmd);

No, that's not needed.

What is the problem with these lines that you are trying to solve?

Quoted 19 lines
Quoted 10 lines
Quoted 10 lines
net/bluetooth/hci_core.c +1 -0
+
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
 				hci_dev_set_flag(hdev, HCI_CMD_PENDING);
net/bluetooth/hci_event.c +0 -0
index 4b7fc430793c..1e7d9bee9111 100644
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
Quoted 8 lines

net/bluetooth/hci_event.c static void hci_phy_link_complete_evt(struct hci_dev *hdev, · line 4947

Quoted 6 lines
net/bluetooth/hci_event.c +4 -0
 		return;
 	}
+	if (!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}

How can this be triggered?

Quoted 2 lines

syzbot showed that this line is accessed irrespective of the null value it contains, so added a pointer check for that.

But does hcon->amp_mgr->l2cap_conn->hcon become NULL here?

thanks,

greg k-h
_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees

author
#6 lore raw
Quoted 70 lines · Greg KH
Quoted 36 lines · Anmol Karn
Quoted 27 lines · Greg KH
Quoted 26 lines · Anmol Karn

Fix null pointer deref in hci_phy_link_complete_evt, there was no checking there for the hcon->amp_mgr->l2cap_conn->hconn, and also in hci_cmd_work, for hdev->sent_cmd.

To fix this issue Add pointer checking in hci_cmd_work and hci_phy_link_complete_evt.
[Linux-next-20200827]

This patch corrected some mistakes from previous patch.

---
 net/bluetooth/hci_core.c  | 5 ++++-
 net/bluetooth/hci_event.c | 4 ++++
 2 files changed, 8 insertions(+), 1 deletion(-)
net/bluetooth/hci_core.c +0 -0
index 68bfe57b6625..996efd654e7a 100644
@@ -4922,7 +4922,10 @@ static void hci_cmd_work(struct work_struct *work)
 
 		kfree_skb(hdev->sent_cmd);
 
Quoted 7 lines

net/bluetooth/hci_core.c static void hci_cmd_work(struct work_struct *work) · line 4927

Quoted 4 lines
net/bluetooth/hci_core.c +3 -1
-		hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		if (hdev->sent_cmd) {
+			hdev->sent_cmd = skb_clone(skb, GFP_KERNEL);
+		}

How can sent_cmd be NULL here? Are you sure something previous to this shouldn't be fixed instead?

Quoted 1 line

Sir, sent_cmd was freed before this condition check, thats why i checked it,

But it can not be NULL at that point in time, as nothing set it to NULL, correct?

Quoted 4 lines

i think i should check it before the free of hdev->sent_cmd like,

if (hdev->sent_cmd)

	kfree_skb(hdev->sent_cmd);

No, that's not needed.

What is the problem with these lines that you are trying to solve?

Quoted 19 lines
Quoted 18 lines
Quoted 9 lines
net/bluetooth/hci_core.c +1 -0
+
 		if (hdev->sent_cmd) {
 			if (hci_req_status_pend(hdev))
 				hci_dev_set_flag(hdev, HCI_CMD_PENDING);
net/bluetooth/hci_event.c +0 -0
index 4b7fc430793c..1e7d9bee9111 100644

net/bluetooth/hci_event.c static void hci_phy_link_complete_evt(struct hci_dev *hdev, · line 4947

Quoted 7 lines
net/bluetooth/hci_event.c +4 -0
@@ -4941,6 +4941,10 @@ static void hci_phy_link_complete_evt(struct hci_dev *hdev,
 		hci_dev_unlock(hdev);
 		return;
 	}
+	if (!(hcon->amp_mgr->l2cap_conn->hcon)) {
+		hci_dev_unlock(hdev);
+		return;
+	}

How can this be triggered?

Quoted 4 lines
Quoted 2 lines

syzbot showed that this line is accessed irrespective of the null value it contains, so added a pointer check for that.

But does hcon->amp_mgr->l2cap_conn->hcon become NULL here?

Sir, according to the report obtained by running decode_stacktrace on logs there is something getting null at this line, after verifying the buggy address i thought it would be better to check this whole line.

will dig more deeper into this and will make appropriate changes in the next version, thanks for review.

Anmol Karn

_______________________________________________
Linux-kernel-mentees mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/linux-kernel-mentees