linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [Linux-kernel-mentees] [PATCH net] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt()
@ 2020-08-05 18:09 Peilin Ye
  2020-09-09  7:17 ` [Linux-kernel-mentees] [PATCH net v2] " Peilin Ye
  0 siblings, 1 reply; 4+ messages in thread
From: Peilin Ye @ 2020-08-05 18:09 UTC (permalink / raw)
  To: Marcel Holtmann, Johan Hedberg
  Cc: Peilin Ye, David S. Miller, Jakub Kicinski, Greg Kroah-Hartman,
	linux-bluetooth, netdev, linux-kernel-mentees, linux-kernel

`num_reports` is not being properly checked. A malformed event packet with
a large `num_reports` number makes hci_le_direct_adv_report_evt() read out
of bounds. Fix it.

Reported-and-tested-by: syzbot+24ebd650e20bd263ca01@syzkaller.appspotmail.com
Fixes: 2f010b55884e ("Bluetooth: Add support for handling LE Direct Advertising Report events")
Link: https://syzkaller.appspot.com/bug?extid=24ebd650e20bd263ca01
Signed-off-by: Peilin Ye <yepeilin.cs@gmail.com>
---
I moved the initialization of `ev` out of the loop and restructured the
function a bit, since otherwise the check would look like:

	if (!num_reports || skb->len < num_reports * sizeof(struct hci_ev_le_direct_adv_info) + 1)
		return;

Therefore I used the similar structure with hci_inquiry_result_evt() etc.

hci_le_adv_report_evt() and hci_le_ext_adv_report_evt() also have the
same issue with `num_reports`, but I'm not sure how to perform the check
for them, since they use variable-length reports. Should we do something
like this? (take hci_le_adv_report_evt() as example:)

	if (!num_reports ||
	    skb->len < num_reports * (sizeof(*ev) + HCI_MAX_AD_LENGTH + 1) + 1)
		return;

Then how about hci_le_ext_adv_report_evt()? There is no such
`HCI_MAX_AD_LENGTH` restrictions on `ev->length` for it, I assume?

Would like to hear your opinion. Thank you!

 net/bluetooth/hci_event.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
index 4b7fc430793c..aec43ae488d1 100644
--- a/net/bluetooth/hci_event.c
+++ b/net/bluetooth/hci_event.c
@@ -5863,21 +5863,19 @@ static void hci_le_direct_adv_report_evt(struct hci_dev *hdev,
 					 struct sk_buff *skb)
 {
 	u8 num_reports = skb->data[0];
-	void *ptr = &skb->data[1];
+	struct hci_ev_le_direct_adv_info *ev = (void *)&skb->data[1];
 
-	hci_dev_lock(hdev);
+	if (!num_reports || skb->len < num_reports * sizeof(*ev) + 1)
+		return;
 
-	while (num_reports--) {
-		struct hci_ev_le_direct_adv_info *ev = ptr;
+	hci_dev_lock(hdev);
 
+	for (; num_reports; num_reports--, ev++)
 		process_adv_report(hdev, ev->evt_type, &ev->bdaddr,
 				   ev->bdaddr_type, &ev->direct_addr,
 				   ev->direct_addr_type, ev->rssi, NULL, 0,
 				   false);
 
-		ptr += sizeof(*ev);
-	}
-
 	hci_dev_unlock(hdev);
 }
 
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 4+ messages in thread
* [Linux-kernel-mentees] [PATCH net v2] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt()
  2020-08-05 18:09 [Linux-kernel-mentees] [PATCH net] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt() Peilin Ye
@ 2020-09-09  7:17 ` Peilin Ye
  2020-11-09 12:16   ` Marcel Holtmann
  0 siblings, 1 reply; 4+ messages in thread
From: Peilin Ye @ 2020-09-09  7:17 UTC (permalink / raw)
  To: Marcel Holtmann, Johan Hedberg
  Cc: Peilin Ye, David S. Miller, Jakub Kicinski, Greg Kroah-Hartman,
	linux-bluetooth, netdev, linux-kernel-mentees, linux-kernel

`num_reports` is not being properly checked. A malformed event packet with
a large `num_reports` number makes hci_le_direct_adv_report_evt() read out
of bounds. Fix it.

Cc: stable@vger.kernel.org
Fixes: 2f010b55884e ("Bluetooth: Add support for handling LE Direct Advertising Report events")
Reported-and-tested-by: syzbot+24ebd650e20bd263ca01@syzkaller.appspotmail.com
Link: https://syzkaller.appspot.com/bug?extid=24ebd650e20bd263ca01
Signed-off-by: Peilin Ye <yepeilin.cs@gmail.com>
---
Change in v2:
    - add "Cc: stable@" tag.

 net/bluetooth/hci_event.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
index 4b7fc430793c..aec43ae488d1 100644
--- a/net/bluetooth/hci_event.c
+++ b/net/bluetooth/hci_event.c
@@ -5863,21 +5863,19 @@ static void hci_le_direct_adv_report_evt(struct hci_dev *hdev,
 					 struct sk_buff *skb)
 {
 	u8 num_reports = skb->data[0];
-	void *ptr = &skb->data[1];
+	struct hci_ev_le_direct_adv_info *ev = (void *)&skb->data[1];
 
-	hci_dev_lock(hdev);
+	if (!num_reports || skb->len < num_reports * sizeof(*ev) + 1)
+		return;
 
-	while (num_reports--) {
-		struct hci_ev_le_direct_adv_info *ev = ptr;
+	hci_dev_lock(hdev);
 
+	for (; num_reports; num_reports--, ev++)
 		process_adv_report(hdev, ev->evt_type, &ev->bdaddr,
 				   ev->bdaddr_type, &ev->direct_addr,
 				   ev->direct_addr_type, ev->rssi, NULL, 0,
 				   false);
 
-		ptr += sizeof(*ev);
-	}
-
 	hci_dev_unlock(hdev);
 }
 
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 4+ messages in thread
* Re: [Linux-kernel-mentees] [PATCH net v2] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt()
  2020-09-09  7:17 ` [Linux-kernel-mentees] [PATCH net v2] " Peilin Ye
@ 2020-11-09 12:16   ` Marcel Holtmann
  2020-11-11  5:18     ` Peilin Ye
  0 siblings, 1 reply; 4+ messages in thread
From: Marcel Holtmann @ 2020-11-09 12:16 UTC (permalink / raw)
  To: Peilin Ye
  Cc: Johan Hedberg, David S. Miller, Jakub Kicinski,
	Greg Kroah-Hartman, Bluez mailing list,
	open list:NETWORKING [GENERAL],
	linux-kernel-mentees, linux-kernel

Hi Peilin,

> `num_reports` is not being properly checked. A malformed event packet with
> a large `num_reports` number makes hci_le_direct_adv_report_evt() read out
> of bounds. Fix it.
> 
> Cc: stable@vger.kernel.org
> Fixes: 2f010b55884e ("Bluetooth: Add support for handling LE Direct Advertising Report events")
> Reported-and-tested-by: syzbot+24ebd650e20bd263ca01@syzkaller.appspotmail.com
> Link: https://syzkaller.appspot.com/bug?extid=24ebd650e20bd263ca01
> Signed-off-by: Peilin Ye <yepeilin.cs@gmail.com>
> ---
> Change in v2:
>    - add "Cc: stable@" tag.
> 
> net/bluetooth/hci_event.c | 12 +++++-------
> 1 file changed, 5 insertions(+), 7 deletions(-)
> 
> diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
> index 4b7fc430793c..aec43ae488d1 100644
> --- a/net/bluetooth/hci_event.c
> +++ b/net/bluetooth/hci_event.c
> @@ -5863,21 +5863,19 @@ static void hci_le_direct_adv_report_evt(struct hci_dev *hdev,
> 					 struct sk_buff *skb)
> {
> 	u8 num_reports = skb->data[0];
> -	void *ptr = &skb->data[1];
> +	struct hci_ev_le_direct_adv_info *ev = (void *)&skb->data[1];
> 
> -	hci_dev_lock(hdev);
> +	if (!num_reports || skb->len < num_reports * sizeof(*ev) + 1)
> +		return;
> 
> -	while (num_reports--) {
> -		struct hci_ev_le_direct_adv_info *ev = ptr;
> +	hci_dev_lock(hdev);
> 
> +	for (; num_reports; num_reports--, ev++)
> 		process_adv_report(hdev, ev->evt_type, &ev->bdaddr,
> 				   ev->bdaddr_type, &ev->direct_addr,
> 				   ev->direct_addr_type, ev->rssi, NULL, 0,
> 				   false);
> 
> -		ptr += sizeof(*ev);
> -	}
> -
> 	hci_dev_unlock(hdev);
> }

patch has been applied to bluetooth-next tree.

Regards

Marcel


^ permalink raw reply	[flat|nested] 4+ messages in thread
* Re: [Linux-kernel-mentees] [PATCH net v2] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt()
  2020-11-09 12:16   ` Marcel Holtmann
@ 2020-11-11  5:18     ` Peilin Ye
  0 siblings, 0 replies; 4+ messages in thread
From: Peilin Ye @ 2020-11-11  5:18 UTC (permalink / raw)
  To: Marcel Holtmann
  Cc: Johan Hedberg, David S. Miller, Jakub Kicinski,
	Greg Kroah-Hartman, Bluez mailing list,
	open list:NETWORKING [GENERAL],
	linux-kernel-mentees, linux-kernel

On Mon, Nov 09, 2020 at 01:16:53PM +0100, Marcel Holtmann wrote:
> Hi Peilin,
> 
> > `num_reports` is not being properly checked. A malformed event packet with
> > a large `num_reports` number makes hci_le_direct_adv_report_evt() read out
> > of bounds. Fix it.
> > 
> > Cc: stable@vger.kernel.org
> > Fixes: 2f010b55884e ("Bluetooth: Add support for handling LE Direct Advertising Report events")
> > Reported-and-tested-by: syzbot+24ebd650e20bd263ca01@syzkaller.appspotmail.com
> > Link: https://syzkaller.appspot.com/bug?extid=24ebd650e20bd263ca01
> > Signed-off-by: Peilin Ye <yepeilin.cs@gmail.com>
> > ---
> > Change in v2:
> >    - add "Cc: stable@" tag.
> > 
> > net/bluetooth/hci_event.c | 12 +++++-------
> > 1 file changed, 5 insertions(+), 7 deletions(-)
> > 
> > diff --git a/net/bluetooth/hci_event.c b/net/bluetooth/hci_event.c
> > index 4b7fc430793c..aec43ae488d1 100644
> > --- a/net/bluetooth/hci_event.c
> > +++ b/net/bluetooth/hci_event.c
> > @@ -5863,21 +5863,19 @@ static void hci_le_direct_adv_report_evt(struct hci_dev *hdev,
> > 					 struct sk_buff *skb)
> > {
> > 	u8 num_reports = skb->data[0];
> > -	void *ptr = &skb->data[1];
> > +	struct hci_ev_le_direct_adv_info *ev = (void *)&skb->data[1];
> > 
> > -	hci_dev_lock(hdev);
> > +	if (!num_reports || skb->len < num_reports * sizeof(*ev) + 1)
> > +		return;
> > 
> > -	while (num_reports--) {
> > -		struct hci_ev_le_direct_adv_info *ev = ptr;
> > +	hci_dev_lock(hdev);
> > 
> > +	for (; num_reports; num_reports--, ev++)
> > 		process_adv_report(hdev, ev->evt_type, &ev->bdaddr,
> > 				   ev->bdaddr_type, &ev->direct_addr,
> > 				   ev->direct_addr_type, ev->rssi, NULL, 0,
> > 				   false);
> > 
> > -		ptr += sizeof(*ev);
> > -	}
> > -
> > 	hci_dev_unlock(hdev);
> > }
> 
> patch has been applied to bluetooth-next tree.

Thank you for reviewing it,

Peilin Ye


^ permalink raw reply	[flat|nested] 4+ messages in thread
end of thread, other threads:[~2020-11-11  5:19 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-08-05 18:09 [Linux-kernel-mentees] [PATCH net] Bluetooth: Fix slab-out-of-bounds read in hci_le_direct_adv_report_evt() Peilin Ye
2020-09-09  7:17 ` [Linux-kernel-mentees] [PATCH net v2] " Peilin Ye
2020-11-09 12:16   ` Marcel Holtmann
2020-11-11  5:18     ` Peilin Ye

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).