line6_read_data() is never called in atomic context.
The call chains ending up at line6_read_data() are: [1] line6_read_data() <- line6_read_serial_number() -> pod_startup4() [2] line6_read_data() <- line6_read_serial_number() <- podhd_startup_workqueue() [3] line6_read_data() <- podhd_dev_start() <- podhd_startup_workqueue() [4] line6_read_data() <- line6_read_serial_number() <- toneport_init() <- line6_probe() [5] line6_read_data() <- toneport_init() <- line6_probe()
pod_startup4() and podhd_startup_workqueue() are set in INIT_WORK(). line6_probe() is called in toneport_probe(), pod_probe(), podhd_probe() and variax_probe(), which are set as ".probe" in usb_driver structure. These functions are not called in atomic context.
Despite never getting called from atomic context, line6_read_data() calls mdelay for busy wait. That is not necessary and can be replaced with msleep to avoid busy waiting.
This is found by a static analysis tool named DCNS written by myself.
Signed-off-by: Jia-Ju Bai baijiaju1990@gmail.com --- sound/usb/line6/driver.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/sound/usb/line6/driver.c b/sound/usb/line6/driver.c index 0ff5a7d..af99e73 100644 --- a/sound/usb/line6/driver.c +++ b/sound/usb/line6/driver.c @@ -356,7 +356,7 @@ int line6_read_data(struct usb_line6 *line6, unsigned address, void *data,
/* Wait for data length. We'll get 0xff until length arrives. */ for (count = 0; count < LINE6_READ_WRITE_MAX_RETRIES; count++) { - mdelay(LINE6_READ_WRITE_STATUS_DELAY); + msleep(LINE6_READ_WRITE_STATUS_DELAY);
ret = usb_control_msg(usbdev, usb_rcvctrlpipe(usbdev, 0), 0x67, USB_TYPE_VENDOR | USB_RECIP_DEVICE |