USB: UHCI: fix for suspend of virtual HP controller
authorAlan Stern <stern@rowland.harvard.edu>
Tue, 14 May 2013 17:55:29 +0000 (13:55 -0400)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 15 May 2013 17:41:40 +0000 (13:41 -0400)
HP's virtual UHCI host controller takes a long time to suspend
(several hundred microseconds), even when no devices are attached.
This provokes a warning message from uhci-hcd in the auto-stop case.

To prevent this from happening, this patch adds a test to avoid
performing an auto-stop when the wait_for_hp quirk flag is set.  The
controller will still suspend through the normal runtime PM mechanism.
And since that pathway includes a 1-ms delay, the slowness of the
virtual hardware won't matter.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-and-tested-by: ZhenHua <zhen-hual@hp.com>
CC: <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/usb/host/uhci-hub.c

index f87bee6d2789256fc8c8dcdde71a6e4510108397..9189bc984c98ce9804d51ec94c4aebe4147b8f3f 100644 (file)
@@ -225,7 +225,8 @@ static int uhci_hub_status_data(struct usb_hcd *hcd, char *buf)
                /* auto-stop if nothing connected for 1 second */
                if (any_ports_active(uhci))
                        uhci->rh_state = UHCI_RH_RUNNING;
-               else if (time_after_eq(jiffies, uhci->auto_stop_time))
+               else if (time_after_eq(jiffies, uhci->auto_stop_time) &&
+                               !uhci->wait_for_hp)
                        suspend_rh(uhci, UHCI_RH_AUTO_STOPPED);
                break;