From stern@rowland.harvard.edu Tue May 8 08:08:55 2007 Date: Fri, 4 May 2007 11:51:25 -0400 (EDT) From: Alan Stern To: Greg KH Subject: [PATCH 02/16] USB: Implement PM FREEZE and PRETHAW Message-ID: This patch (as884) finally implements the time-saving semantics possible with the Power Management FREEZE and PRETHAW events. Their proper handling requires only that devices be quiesced, with interrupts and DMA turned off; non-root USB devices don't actually need to be put in a suspended state. The patch checks and avoids doing the suspend call when possible. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman --- drivers/usb/core/driver.c | 9 ++++++++- drivers/usb/core/generic.c | 4 ---- 2 files changed, 8 insertions(+), 5 deletions(-) --- a/drivers/usb/core/driver.c +++ b/drivers/usb/core/driver.c @@ -1050,8 +1050,15 @@ static int usb_suspend_both(struct usb_d break; } } - if (status == 0) + if (status == 0) { + + /* Non-root devices don't need to do anything for FREEZE + * or PRETHAW. */ + if (udev->parent && (msg.event == PM_EVENT_FREEZE || + msg.event == PM_EVENT_PRETHAW)) + goto done; status = usb_suspend_device(udev, msg); + } /* If the suspend failed, resume interfaces that did get suspended */ if (status != 0) { --- a/drivers/usb/core/generic.c +++ b/drivers/usb/core/generic.c @@ -193,10 +193,6 @@ static void generic_disconnect(struct us static int generic_suspend(struct usb_device *udev, pm_message_t msg) { - /* USB devices enter SUSPEND state through their hubs, but can be - * marked for FREEZE as soon as their children are already idled. - * But those semantics are useless, so we equate the two (sigh). - */ return usb_port_suspend(udev); }