[v3,1/2] usbip: give back URBs for unsent unlink requests during cleanup

Commit Message

Anirudh Rayabharam Aug. 13, 2021, 6:25 p.m. UTC

In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
not given back. This sometimes causes usb_kill_urb to wait indefinitely
for that urb to be given back. syzbot has reported a hung task issue [1]
for this.

To fix this, give back the urbs corresponding to unsent unlink requests
(unlink_tx list) similar to how urbs corresponding to unanswered unlink
requests (unlink_rx list) are given back.

[1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76

Reported-by: [e-mail]
Tested-by: [e-mail]
Signed-off-by: Anirudh Rayabharam <[e-mail]>
---
 drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)

Comments

Shuah Khan Aug. 17, 2021, 11:16 p.m. UTC | #1

On 8/13/21 12:25 PM, Anirudh Rayabharam wrote:
> In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
> not given back. This sometimes causes usb_kill_urb to wait indefinitely
> for that urb to be given back. syzbot has reported a hung task issue [1]
> for this.
> 
> To fix this, give back the urbs corresponding to unsent unlink requests
> (unlink_tx list) similar to how urbs corresponding to unanswered unlink
> requests (unlink_rx list) are given back.
> 
> [1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76
> 
> Reported-by: [e-mail]
> Tested-by: [e-mail]
> Signed-off-by: Anirudh Rayabharam <[e-mail]>
> ---
>   drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
>   1 file changed, 26 insertions(+)
> 
> diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
> index 4ba6bcdaa8e9..6f3f374d4bbc 100644
> --- a/drivers/usb/usbip/vhci_hcd.c
> +++ b/drivers/usb/usbip/vhci_hcd.c
> @@ -957,8 +957,34 @@ static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
>   	spin_lock(&vdev->priv_lock);
>   
>   	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
> +		struct urb *urb;
> +
> +		/* give back URB of unsent unlink request */
>   		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);

I know this is an exiting one.
Let's make this pr_debug or remove it all together.

> +
> +		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
> +		if (!urb) {
> +			pr_info("the urb (seqnum %lu) was already given back\n",
> +				unlink->unlink_seqnum);

Let's make this pr_debug or remove it all together.

> +			list_del(&unlink->list);
> +			kfree(unlink);
> +			continue;
> +		}
> +
> +		urb->status = -ENODEV;
> +
> +		usb_hcd_unlink_urb_from_ep(hcd, urb);
> +
>   		list_del(&unlink->list);
> +
> +		spin_unlock(&vdev->priv_lock);
> +		spin_unlock_irqrestore(&vhci->lock, flags);
> +
> +		usb_hcd_giveback_urb(hcd, urb, urb->status);
> +
> +		spin_lock_irqsave(&vhci->lock, flags);
> +		spin_lock(&vdev->priv_lock);
> +
>   		kfree(unlink);
>   	}
>   
> 

thanks,
-- Shuah

Greg KH Aug. 18, 2021, 5:39 a.m. UTC | #2

On Tue, Aug 17, 2021 at 05:16:51PM -0600, Shuah Khan wrote:
> On 8/13/21 12:25 PM, Anirudh Rayabharam wrote:
> > In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
> > not given back. This sometimes causes usb_kill_urb to wait indefinitely
> > for that urb to be given back. syzbot has reported a hung task issue [1]
> > for this.
> > 
> > To fix this, give back the urbs corresponding to unsent unlink requests
> > (unlink_tx list) similar to how urbs corresponding to unanswered unlink
> > requests (unlink_rx list) are given back.
> > 
> > [1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76
> > 
> > Reported-by: [e-mail]
> > Tested-by: [e-mail]
> > Signed-off-by: Anirudh Rayabharam <[e-mail]>
> > ---
> >   drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
> >   1 file changed, 26 insertions(+)
> > 
> > diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
> > index 4ba6bcdaa8e9..6f3f374d4bbc 100644
> > --- a/drivers/usb/usbip/vhci_hcd.c
> > +++ b/drivers/usb/usbip/vhci_hcd.c
> > @@ -957,8 +957,34 @@ static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
> >   	spin_lock(&vdev->priv_lock);
> >   	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
> > +		struct urb *urb;
> > +
> > +		/* give back URB of unsent unlink request */
> >   		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);
> 
> I know this is an exiting one.
> Let's make this pr_debug or remove it all together.
> 
> > +
> > +		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
> > +		if (!urb) {
> > +			pr_info("the urb (seqnum %lu) was already given back\n",
> > +				unlink->unlink_seqnum);
> 
> Let's make this pr_debug or remove it all together.

As you have a struct device for all of these, please use dev_dbg() and
friends, not pr_*(), for all of these.

thanks,

greg k-h

Shuah Khan Aug. 18, 2021, 6:36 p.m. UTC | #3

On 8/17/21 11:39 PM, Greg KH wrote:
> On Tue, Aug 17, 2021 at 05:16:51PM -0600, Shuah Khan wrote:
>> On 8/13/21 12:25 PM, Anirudh Rayabharam wrote:
>>> In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
>>> not given back. This sometimes causes usb_kill_urb to wait indefinitely
>>> for that urb to be given back. syzbot has reported a hung task issue [1]
>>> for this.
>>>
>>> To fix this, give back the urbs corresponding to unsent unlink requests
>>> (unlink_tx list) similar to how urbs corresponding to unanswered unlink
>>> requests (unlink_rx list) are given back.
>>>
>>> [1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76
>>>
>>> Reported-by: [e-mail]
>>> Tested-by: [e-mail]
>>> Signed-off-by: Anirudh Rayabharam <[e-mail]>
>>> ---
>>>    drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
>>>    1 file changed, 26 insertions(+)
>>>
>>> diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
>>> index 4ba6bcdaa8e9..6f3f374d4bbc 100644
>>> --- a/drivers/usb/usbip/vhci_hcd.c
>>> +++ b/drivers/usb/usbip/vhci_hcd.c
>>> @@ -957,8 +957,34 @@ static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
>>>    	spin_lock(&vdev->priv_lock);
>>>    	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
>>> +		struct urb *urb;
>>> +
>>> +		/* give back URB of unsent unlink request */
>>>    		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);
>>
>> I know this is an exiting one.
>> Let's make this pr_debug or remove it all together.
>>
>>> +
>>> +		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
>>> +		if (!urb) {
>>> +			pr_info("the urb (seqnum %lu) was already given back\n",
>>> +				unlink->unlink_seqnum);
>>
>> Let's make this pr_debug or remove it all together.
> 
> As you have a struct device for all of these, please use dev_dbg() and
> friends, not pr_*(), for all of these.
> 

Yes. Makes perfect sense.

thanks,
-- Shuah

Anirudh Rayabharam Aug. 19, 2021, 6:09 p.m. UTC | #4

On Wed, Aug 18, 2021 at 12:36:11PM -0600, Shuah Khan wrote:
> On 8/17/21 11:39 PM, Greg KH wrote:
> > On Tue, Aug 17, 2021 at 05:16:51PM -0600, Shuah Khan wrote:
> > > On 8/13/21 12:25 PM, Anirudh Rayabharam wrote:
> > > > In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
> > > > not given back. This sometimes causes usb_kill_urb to wait indefinitely
> > > > for that urb to be given back. syzbot has reported a hung task issue [1]
> > > > for this.
> > > > 
> > > > To fix this, give back the urbs corresponding to unsent unlink requests
> > > > (unlink_tx list) similar to how urbs corresponding to unanswered unlink
> > > > requests (unlink_rx list) are given back.
> > > > 
> > > > [1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76
> > > > 
> > > > Reported-by: [e-mail]
> > > > Tested-by: [e-mail]
> > > > Signed-off-by: Anirudh Rayabharam <[e-mail]>
> > > > ---
> > > >    drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
> > > >    1 file changed, 26 insertions(+)
> > > > 
> > > > diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
> > > > index 4ba6bcdaa8e9..6f3f374d4bbc 100644
> > > > --- a/drivers/usb/usbip/vhci_hcd.c
> > > > +++ b/drivers/usb/usbip/vhci_hcd.c
> > > > @@ -957,8 +957,34 @@ static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
> > > >    	spin_lock(&vdev->priv_lock);
> > > >    	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
> > > > +		struct urb *urb;
> > > > +
> > > > +		/* give back URB of unsent unlink request */
> > > >    		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);
> > > 
> > > I know this is an exiting one.
> > > Let's make this pr_debug or remove it all together.
> > > 
> > > > +
> > > > +		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
> > > > +		if (!urb) {
> > > > +			pr_info("the urb (seqnum %lu) was already given back\n",
> > > > +				unlink->unlink_seqnum);
> > > 
> > > Let's make this pr_debug or remove it all together.
> > 
> > As you have a struct device for all of these, please use dev_dbg() and
> > friends, not pr_*(), for all of these.
> > 
> 
> Yes. Makes perfect sense.

Perhaps we should use usbip_dbg_vhci_hc() instead of dev_dbg()? It is
one of the custom macros defined by the usbip driver for printing debug
logs.

Thanks,

	Anirudh

Shuah Khan Aug. 19, 2021, 6:49 p.m. UTC | #5

On 8/19/21 12:09 PM, Anirudh Rayabharam wrote:
> On Wed, Aug 18, 2021 at 12:36:11PM -0600, Shuah Khan wrote:
>> On 8/17/21 11:39 PM, Greg KH wrote:
>>> On Tue, Aug 17, 2021 at 05:16:51PM -0600, Shuah Khan wrote:
>>>> On 8/13/21 12:25 PM, Anirudh Rayabharam wrote:
>>>>> In vhci_device_unlink_cleanup(), the URBs for unsent unlink requests are
>>>>> not given back. This sometimes causes usb_kill_urb to wait indefinitely
>>>>> for that urb to be given back. syzbot has reported a hung task issue [1]
>>>>> for this.
>>>>>
>>>>> To fix this, give back the urbs corresponding to unsent unlink requests
>>>>> (unlink_tx list) similar to how urbs corresponding to unanswered unlink
>>>>> requests (unlink_rx list) are given back.
>>>>>
>>>>> [1]: https://syzkaller.appspot.com/bug?id=08f12df95ae7da69814e64eb5515d5a85ed06b76
>>>>>
>>>>> Reported-by: [e-mail]
>>>>> Tested-by: [e-mail]
>>>>> Signed-off-by: Anirudh Rayabharam <[e-mail]>
>>>>> ---
>>>>>     drivers/usb/usbip/vhci_hcd.c | 26 ++++++++++++++++++++++++++
>>>>>     1 file changed, 26 insertions(+)
>>>>>
>>>>> diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
>>>>> index 4ba6bcdaa8e9..6f3f374d4bbc 100644
>>>>> --- a/drivers/usb/usbip/vhci_hcd.c
>>>>> +++ b/drivers/usb/usbip/vhci_hcd.c
>>>>> @@ -957,8 +957,34 @@ static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
>>>>>     	spin_lock(&vdev->priv_lock);
>>>>>     	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
>>>>> +		struct urb *urb;
>>>>> +
>>>>> +		/* give back URB of unsent unlink request */
>>>>>     		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);
>>>>
>>>> I know this is an exiting one.
>>>> Let's make this pr_debug or remove it all together.
>>>>
>>>>> +
>>>>> +		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
>>>>> +		if (!urb) {
>>>>> +			pr_info("the urb (seqnum %lu) was already given back\n",
>>>>> +				unlink->unlink_seqnum);
>>>>
>>>> Let's make this pr_debug or remove it all together.
>>>
>>> As you have a struct device for all of these, please use dev_dbg() and
>>> friends, not pr_*(), for all of these.
>>>
>>
>> Yes. Makes perfect sense.
> 
> Perhaps we should use usbip_dbg_vhci_hc() instead of dev_dbg()? It is
> one of the custom macros defined by the usbip driver for printing debug
> logs.
> 

Yes that macro could be used. However, let's just get rid of the messages.
I don't see much use for them.

thanks,
-- Shuah

Patch

diff --git a/drivers/usb/usbip/vhci_hcd.c b/drivers/usb/usbip/vhci_hcd.c
index 4ba6bcdaa8e9..6f3f374d4bbc 100644
--- a/drivers/usb/usbip/vhci_hcd.c
+++ b/drivers/usb/usbip/vhci_hcd.c
@@ -957,8 +957,34 @@  static void vhci_device_unlink_cleanup(struct vhci_device *vdev)
 	spin_lock(&vdev->priv_lock);
 
 	list_for_each_entry_safe(unlink, tmp, &vdev->unlink_tx, list) {
+		struct urb *urb;
+
+		/* give back URB of unsent unlink request */
 		pr_info("unlink cleanup tx %lu\n", unlink->unlink_seqnum);
+
+		urb = pickup_urb_and_free_priv(vdev, unlink->unlink_seqnum);
+		if (!urb) {
+			pr_info("the urb (seqnum %lu) was already given back\n",
+				unlink->unlink_seqnum);
+			list_del(&unlink->list);
+			kfree(unlink);
+			continue;
+		}
+
+		urb->status = -ENODEV;
+
+		usb_hcd_unlink_urb_from_ep(hcd, urb);
+
 		list_del(&unlink->list);
+
+		spin_unlock(&vdev->priv_lock);
+		spin_unlock_irqrestore(&vhci->lock, flags);
+
+		usb_hcd_giveback_urb(hcd, urb, urb->status);
+
+		spin_lock_irqsave(&vhci->lock, flags);
+		spin_lock(&vdev->priv_lock);
+
 		kfree(unlink);
 	}