USB: don't rebind drivers after failed resume or reset
authorAlan Stern <stern@rowland.harvard.edu>
Tue, 21 Oct 2008 19:40:03 +0000 (15:40 -0400)
committerGreg Kroah-Hartman <gregkh@suse.de>
Wed, 22 Oct 2008 17:05:29 +0000 (10:05 -0700)
This patch (as1152) may help prevent some problems associated with the
new policy of unbinding drivers that don't support suspend/resume or
pre_reset/post_reset.  If for any reason the resume or reset fails, and
the device is logically disconnected, there's no point in trying to
rebind the driver.  So the patch checks for success before carrying
out the unbind/rebind.

There was a report from one user that this fixed a problem he was
experiencing, but the details never became fully clear.  In any case,
adding these tests can't hurt.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
drivers/usb/core/driver.c
drivers/usb/core/hub.c

index e935be7eb468b7b70b08bbb4a16f9a7b5fb6546d..3d7793d93031525400d3ed746a60639ebed1d52f 100644 (file)
@@ -1610,7 +1610,8 @@ int usb_external_resume_device(struct usb_device *udev)
        status = usb_resume_both(udev);
        udev->last_busy = jiffies;
        usb_pm_unlock(udev);
-       do_unbind_rebind(udev, DO_REBIND);
+       if (status == 0)
+               do_unbind_rebind(udev, DO_REBIND);
 
        /* Now that the device is awake, we can start trying to autosuspend
         * it again. */
index d73ce262c3651446582b251c887b5f9af787f7ce..9b3f16bd12cb4508a51a63d936a527b02b152475 100644 (file)
@@ -3504,7 +3504,7 @@ int usb_reset_device(struct usb_device *udev)
                                                USB_INTERFACE_BOUND)
                                        rebind = 1;
                        }
-                       if (rebind)
+                       if (ret == 0 && rebind)
                                usb_rebind_intf(cintf);
                }
        }