Re: [PATCH V3] input: Fix USB autosuspend on bcm5974

On Wed, 12 Oct 2011, Matthew Garrett wrote:

On Wed, Oct 12, 2011 at 01:24:42PM -0400, Alan Stern wrote:

quoted

On Wed, 12 Oct 2011, Matthew Garrett wrote:

quoted

True. That makes things rather more awkward.

What about the second idea I mentioned?  If autosuspend fails with
-EBUSY or -EAGAIN, usbcore could schedule another autosuspend attempt.

Yeah, that sounds better.

Does this patch help?

Alan Stern


 drivers/usb/core/driver.c |    7 +++++++
 1 file changed, 7 insertions(+)

Index: usb-3.1/drivers/usb/core/driver.c
===================================================================

--- usb-3.1.orig/drivers/usb/core/driver.c
+++ usb-3.1/drivers/usb/core/driver.c

@@ -1667,6 +1667,13 @@ int usb_runtime_suspend(struct device *d
 		return -EAGAIN;
 
 	status = usb_suspend_both(udev, PMSG_AUTO_SUSPEND);
+
+	/* Retry if autosuspend failed temporarily */
+	if (status == -EBUSY || status == -EAGAIN) {
+		usb_mark_last_busy(udev);
+		pm_request_autosuspend(dev);
+	}
+
 	/* The PM core reacts badly unless the return code is 0,
 	 * -EAGAIN, or -EBUSY, so always return -EBUSY on an error.
 	 */