USB: HCDs use the do_remote_wakeup flag
When a USB device is suspended, whether or not it is enabled for
remote wakeup depends on the device_may_wakeup() setting. The setting
is then saved in the do_remote_wakeup flag.
Later on, however, the device_may_wakeup() value can change because of
user activity. So when testing whether a suspended device is or
should be enabled for remote wakeup, we should always test
do_remote_wakeup instead of device_may_wakeup(). This patch (as1076)
makes that change for root hubs in several places.
The patch also adjusts uhci-hcd so that when an autostopped controller
is suspended, the remote wakeup setting agrees with the value recorded
in the root hub's do_remote_wakeup flag.
And the patch adjusts ehci-hcd so that wakeup events on selectively
suspended ports (i.e., the bus itself isn't suspended) don't turn on
the PME# wakeup signal.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
diff --git a/drivers/usb/host/isp116x-hcd.c b/drivers/usb/host/isp116x-hcd.c
index 66d773c..20b9a0d 100644
--- a/drivers/usb/host/isp116x-hcd.c
+++ b/drivers/usb/host/isp116x-hcd.c
@@ -1400,7 +1400,7 @@
spin_unlock_irqrestore(&isp116x->lock, flags);
val &= (~HCCONTROL_HCFS & ~HCCONTROL_RWE);
val |= HCCONTROL_USB_SUSPEND;
- if (device_may_wakeup(&hcd->self.root_hub->dev))
+ if (hcd->self.root_hub->do_remote_wakeup)
val |= HCCONTROL_RWE;
/* Wait for usb transfers to finish */
msleep(2);