[PATCH 3.2 030/153] USB: serial: io_edgeport: fix possible sleep-in-atomic

[Ben Hutchings]

3.2.102-rc1 review patch.  If anyone has any objections, please let me know.

------------------

From: Jia-Ju Bai <baijiaju1990@xxxxxxxxx>

commit c7b8f77872c73f69a16528a9eb87afefcccdc18b upstream.

According to drivers/usb/serial/io_edgeport.c, the driver may sleep
under a spinlock.
The function call path is:
edge_bulk_in_callback (acquire the spinlock)
   process_rcvd_data
     process_rcvd_status
       change_port_settings
         send_iosp_ext_cmd
           write_cmd_usb
             usb_kill_urb --> may sleep

To fix it, the redundant usb_kill_urb() is removed from the error path
after usb_submit_urb() fails.

This possible bug is found by my static analysis tool (DSAC) and checked
by my code review.

Signed-off-by: Jia-Ju Bai <baijiaju1990@xxxxxxxxx>
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Johan Hovold <johan@xxxxxxxxxx>
[bwh: Backported to 3.2: adjust context]
Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx>
---
 drivers/usb/serial/io_edgeport.c | 1 -
 1 file changed, 1 deletion(-)

--- a/drivers/usb/serial/io_edgeport.c
+++ b/drivers/usb/serial/io_edgeport.c
@@ -2387,7 +2387,6 @@ static int write_cmd_usb(struct edgeport
 		dev_err(&edge_port->port->dev,
 		    "%s - usb_submit_urb(write command) failed, status = %d\n",
 							__func__, status);
-		usb_kill_urb(urb);
 		usb_free_urb(urb);
 		atomic_dec(&CmdUrbs);
 		return status;