diff mbox series

[v4] usb: core: Prevent null pointer dereference in update_port_device_state

Message ID 20240110091422.25347-1-quic_ugoswami@quicinc.com (mailing list archive)
State Superseded
Headers show
Series [v4] usb: core: Prevent null pointer dereference in update_port_device_state | expand

Commit Message

Udipto Goswami Jan. 10, 2024, 9:14 a.m. UTC
Currently,the function update_port_device_state gets the usb_hub from
udev->parent by calling usb_hub_to_struct_hub.
However, in case the actconfig or the maxchild is 0, the usb_hub would
be NULL and upon further accessing to get port_dev would result in null
pointer dereference.

Fix this by introducing an if check after the usb_hub is populated.

Fixes: 83cb2604f641 ("usb: core: add sysfs entry for usb device state")
Cc: stable@vger.kernel.org
Signed-off-by: Udipto Goswami <quic_ugoswami@quicinc.com>
---
v4: Fixed minor mistakes in the comment.
v3: Re-wrote the comment for better context.
v2: Introduced comment for the if check & CC'ed stable.

 drivers/usb/core/hub.c | 19 ++++++++++++++++---
 1 file changed, 16 insertions(+), 3 deletions(-)

Comments

Sergey Shtylyov Jan. 10, 2024, 9:22 a.m. UTC | #1
On 1/10/24 12:14 PM, Udipto Goswami wrote:

> Currently,the function update_port_device_state gets the usb_hub from

   Need space between comma and "the"...

> udev->parent by calling usb_hub_to_struct_hub.
> However, in case the actconfig or the maxchild is 0, the usb_hub would
> be NULL and upon further accessing to get port_dev would result in null
> pointer dereference.
> 
> Fix this by introducing an if check after the usb_hub is populated.
> 
> Fixes: 83cb2604f641 ("usb: core: add sysfs entry for usb device state")
> Cc: stable@vger.kernel.org
> Signed-off-by: Udipto Goswami <quic_ugoswami@quicinc.com>
[...]

> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
> index ffd7c99e24a3..5ba1875e6bf4 100644
> --- a/drivers/usb/core/hub.c
> +++ b/drivers/usb/core/hub.c
> @@ -2053,9 +2053,22 @@ static void update_port_device_state(struct usb_device *udev)
>  
>  	if (udev->parent) {
>  		hub = usb_hub_to_struct_hub(udev->parent);
> -		port_dev = hub->ports[udev->portnum - 1];
> -		WRITE_ONCE(port_dev->state, udev->state);
> -		sysfs_notify_dirent(port_dev->state_kn);
> +
> +		/*
> +		 * The Link Layer Validation System Driver (lvstest)
> +		 * has step to unbind the hub before running the rest
> +		 * of the procedure. This triggers hub_disconnect which
> +		 * will set the hub's maxchild to 0, further resulting

   Resulting in.

> +		 * usb_hub_to_struct_hub returning NULL.
> +		 *
> +		 * Add if check to avoid running into NULL pointer
> +		 * de-reference.

   This is obvious from the code below, I think...

> +		 */
> +		if (hub) {
> +			port_dev = hub->ports[udev->portnum - 1];
> +			WRITE_ONCE(port_dev->state, udev->state);
> +			sysfs_notify_dirent(port_dev->state_kn);
> +		}
>  	}
>  }
>  

MBR, Sergey
Udipto Goswami Jan. 10, 2024, 9:57 a.m. UTC | #2
On 1/10/2024 2:52 PM, Sergei Shtylyov wrote:
> On 1/10/24 12:14 PM, Udipto Goswami wrote:
> 
>> Currently,the function update_port_device_state gets the usb_hub from
> 
>     Need space between comma and "the"...
got it.
> 
>> udev->parent by calling usb_hub_to_struct_hub.
>> However, in case the actconfig or the maxchild is 0, the usb_hub would
>> be NULL and upon further accessing to get port_dev would result in null
>> pointer dereference.
>>
>> Fix this by introducing an if check after the usb_hub is populated.
>>
>> Fixes: 83cb2604f641 ("usb: core: add sysfs entry for usb device state")
>> Cc: stable@vger.kernel.org
>> Signed-off-by: Udipto Goswami <quic_ugoswami@quicinc.com>
> [...]
> 
>> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
>> index ffd7c99e24a3..5ba1875e6bf4 100644
>> --- a/drivers/usb/core/hub.c
>> +++ b/drivers/usb/core/hub.c
>> @@ -2053,9 +2053,22 @@ static void update_port_device_state(struct usb_device *udev)
>>   
>>   	if (udev->parent) {
>>   		hub = usb_hub_to_struct_hub(udev->parent);
>> -		port_dev = hub->ports[udev->portnum - 1];
>> -		WRITE_ONCE(port_dev->state, udev->state);
>> -		sysfs_notify_dirent(port_dev->state_kn);
>> +
>> +		/*
>> +		 * The Link Layer Validation System Driver (lvstest)
>> +		 * has step to unbind the hub before running the rest
>> +		 * of the procedure. This triggers hub_disconnect which
>> +		 * will set the hub's maxchild to 0, further resulting
> 
>     Resulting in.

got it.
> 
>> +		 * usb_hub_to_struct_hub returning NULL.
>> +		 *
>> +		 * Add if check to avoid running into NULL pointer
>> +		 * de-reference.
> 
>     This is obvious from the code below, I think...
ok will remove this.

Thanks,
-Udipto
diff mbox series

Patch

diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
index ffd7c99e24a3..5ba1875e6bf4 100644
--- a/drivers/usb/core/hub.c
+++ b/drivers/usb/core/hub.c
@@ -2053,9 +2053,22 @@  static void update_port_device_state(struct usb_device *udev)
 
 	if (udev->parent) {
 		hub = usb_hub_to_struct_hub(udev->parent);
-		port_dev = hub->ports[udev->portnum - 1];
-		WRITE_ONCE(port_dev->state, udev->state);
-		sysfs_notify_dirent(port_dev->state_kn);
+
+		/*
+		 * The Link Layer Validation System Driver (lvstest)
+		 * has step to unbind the hub before running the rest
+		 * of the procedure. This triggers hub_disconnect which
+		 * will set the hub's maxchild to 0, further resulting
+		 * usb_hub_to_struct_hub returning NULL.
+		 *
+		 * Add if check to avoid running into NULL pointer
+		 * de-reference.
+		 */
+		if (hub) {
+			port_dev = hub->ports[udev->portnum - 1];
+			WRITE_ONCE(port_dev->state, udev->state);
+			sysfs_notify_dirent(port_dev->state_kn);
+		}
 	}
 }