USB: EHCI: fix up incorrect merge resolution
authorAlan Stern <stern@rowland.harvard.edu>
Thu, 21 Mar 2013 16:48:42 +0000 (12:48 -0400)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 21 Mar 2013 22:00:18 +0000 (15:00 -0700)
This patch (as1671) fixes up an incorrect resolution of a merge
conflict between Greg KH's usb-linus branch and his usb-next branch.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/usb/host/ehci-timer.c

index dca8fc42b8d77f11928b5f796af86bb1b72dab86..e7363332887efd634b3f5adc9b9ec66830346a33 100644 (file)
@@ -297,6 +297,15 @@ static void ehci_iaa_watchdog(struct ehci_hcd *ehci)
 {
        u32 cmd, status;
 
+       /*
+        * Lost IAA irqs wedge things badly; seen first with a vt8235.
+        * So we need this watchdog, but must protect it against both
+        * (a) SMP races against real IAA firing and retriggering, and
+        * (b) clean HC shutdown, when IAA watchdog was pending.
+        */
+       if (ehci->rh_state != EHCI_RH_RUNNING)
+               return;
+
        /* If we get here, IAA is *REALLY* late.  It's barely
         * conceivable that the system is so busy that CMD_IAAD
         * is still legitimately set, so let's be sure it's