[PATCH] usb: hso: obey DMA rules in tiocmget

[PATCH] usb: hso: obey DMA rules in tiocmget

[Oliver Neukum]

The serial state information must not be embedded into another
data structure, as this interferes with cache handling for DMA
on architectures without cache coherence..
That would result in data corruption on some architectures
Allocating it separately.

Signed-off-by: Oliver Neukum <oneukum@suse.com>
---
 drivers/net/usb/hso.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/net/usb/hso.c b/drivers/net/usb/hso.c
index a505b2ab88b8..b1d1d8ff1747 100644
--- a/drivers/net/usb/hso.c
+++ b/drivers/net/usb/hso.c
@@ -186,7 +186,7 @@ struct hso_tiocmget {
 	int    intr_completed;
 	struct usb_endpoint_descriptor *endp;
 	struct urb *urb;
-	struct hso_serial_state_notification serial_state_notification;
+	struct *hso_serial_state_notification serial_state_notification;
 	u16    prev_UART_state_bitmap;
 	struct uart_icount icount;
 };
@@ -1432,7 +1432,7 @@ static int tiocmget_submit_urb(struct hso_serial *serial,
 			 usb_rcvintpipe(usb,
 					tiocmget->endp->
 					bEndpointAddress & 0x7F),
-			 &tiocmget->serial_state_notification,
+			 tiocmget->serial_state_notification,
 			 sizeof(struct hso_serial_state_notification),
 			 tiocmget_intr_callback, serial,
 			 tiocmget->endp->bInterval);
@@ -1479,7 +1479,7 @@ static void tiocmget_intr_callback(struct urb *urb)
 	/* wIndex should be the USB interface number of the port to which the
 	 * notification applies, which should always be the Modem port.
 	 */
-	serial_state_notification = &tiocmget->serial_state_notification;
+	serial_state_notification = tiocmget->serial_state_notification;
 	if (serial_state_notification->bmRequestType != BM_REQUEST_TYPE ||
 	    serial_state_notification->bNotification != B_NOTIFICATION ||
 	    le16_to_cpu(serial_state_notification->wValue) != W_VALUE ||
@@ -2565,6 +2565,8 @@ static void hso_free_tiomget(struct hso_serial *serial)
 		usb_free_urb(tiocmget->urb);
 		tiocmget->urb = NULL;
 		serial->tiocmget = NULL;
+		kfree(tiocmget->serial_state_notification);
+		tiocmget->serial_state_notification = NULL;
 		kfree(tiocmget);
 	}
 }
@@ -2615,10 +2617,13 @@ static struct hso_device *hso_create_bulk_serial_device(
 		num_urbs = 2;
 		serial->tiocmget = kzalloc(sizeof(struct hso_tiocmget),
 					   GFP_KERNEL);
+		serial->tiocmget->serial_state_notification
+			= kzalloc(sizeof(struct hso_serial_state_notification),
+					   GFP_KERNEL);
 		/* it isn't going to break our heart if serial->tiocmget
 		 *  allocation fails don't bother checking this.
 		 */
-		if (serial->tiocmget) {
+		if (serial->tiocmget && serial->tiocmget->serial_state_notification) {
 			tiocmget = serial->tiocmget;
 			tiocmget->endp = hso_get_ep(interface,
 						    USB_ENDPOINT_XFER_INT,
-- 
2.16.4

Re: [PATCH] usb: hso: obey DMA rules in tiocmget

[David Miller]

From: Oliver Neukum <oneukum@suse.com>
Date: Thu, 17 Oct 2019 11:53:38 +0200

> The serial state information must not be embedded into another
> data structure, as this interferes with cache handling for DMA
> on architectures without cache coherence..
> That would result in data corruption on some architectures
> Allocating it separately.
> 
> Signed-off-by: Oliver Neukum <oneukum@suse.com>

Applied, thanks Oliver.

Re: [PATCH] usb: hso: obey DMA rules in tiocmget

[David Miller]

From: David Miller <davem@davemloft.net>
Date: Thu, 17 Oct 2019 14:19:55 -0400 (EDT)

> From: Oliver Neukum <oneukum@suse.com>
> Date: Thu, 17 Oct 2019 11:53:38 +0200
> 
>> The serial state information must not be embedded into another
>> data structure, as this interferes with cache handling for DMA
>> on architectures without cache coherence..
>> That would result in data corruption on some architectures
>> Allocating it separately.
>> 
>> Signed-off-by: Oliver Neukum <oneukum@suse.com>
> 
> Applied, thanks Oliver.

Ugh, Oliver did you even build test this?

drivers/net/usb/hso.c:189:9: error: expected ‘{’ before ‘*’ token
  189 |  struct *hso_serial_state_notification serial_state_notification;

Seriously, I expect better from an experienced developer such as
yourself.

Re: [PATCH] usb: hso: obey DMA rules in tiocmget

[Oliver Neukum]

Am Donnerstag, den 17.10.2019, 14:20 -0400 schrieb David Miller:
> From: David Miller <davem@davemloft.net>
> Date: Thu, 17 Oct 2019 14:19:55 -0400 (EDT)
> 
> > From: Oliver Neukum <oneukum@suse.com>
> > Date: Thu, 17 Oct 2019 11:53:38 +0200
> > 
> > > The serial state information must not be embedded into another
> > > data structure, as this interferes with cache handling for DMA
> > > on architectures without cache coherence..
> > > That would result in data corruption on some architectures
> > > Allocating it separately.
> > > 
> > > Signed-off-by: Oliver Neukum <oneukum@suse.com>
> > 
> > Applied, thanks Oliver.
> 
> Ugh, Oliver did you even build test this?
> 
> drivers/net/usb/hso.c:189:9: error: expected ‘{’ before ‘*’ token
>   189 |  struct *hso_serial_state_notification serial_state_notification;
> 
> Seriously, I expect better from an experienced developer such as
> yourself.

Sorry, that is why I sent a v2. I built it, but not alone. I have
a cleanup series for that driver in the works.

	Sorry
		Oliver