Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Mon, 13 Aug 2018 10:04:57 -0400 (EDT)
From:   Alan Stern <stern@rowland.harvard.edu>
To:     Anshuman Gupta <anshuman.gupta@intel.com>
cc:     oneukum@suse.com, <linux-usb@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>
Subject: Re: [Query] USB device autosuspend and its runtime usage count.
In-Reply-To: <1534157102-29526-1-git-send-email-anshuman.gupta@intel.com>
Message-ID: <Pine.LNX.4.44L0.1808131003060.1452-100000@iolanthe.rowland.org>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, 13 Aug 2018, Anshuman Gupta wrote:

> On Mon, Aug 13, 2018 at 01:19:26PM +0530, Oliver Neukum wrote:
> > On Mo, 2018-08-13 at 12:15 +0530, Anshuman Gupta wrote:
> > > Hi ,
> > > I need to test a functionality with USB autosuspend with latest kernel          
> > > Linux 4.18-rc8. I am trying to autosuspend a USB key board, i enabled       
> > > its autosuspend by doing echo "auto" to its "control" attributes.
> > > I am expecting USB keyboard to go to autosuspend after autosuspend_delay_ms.
> > > But USB key board is not going to autosuspend because its runtime usage
> > > count is not equal to zero.
> > > 
> > > Below are the log snippets:
> > > 
> > > root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/runtime_enabled
> > > enabled
> > > root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/control
> > > auto
> > > root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/autosuspend_delay_ms
> > > 2000
> > > root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/runtime_status
> > > active
> > > root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/runtime_usage
> > > 1
> > > root@intel-Kabylake-Client-platform:# sleep 60; cat /sys/bus/usb/devices/1-5/power/runtime_usage
> > > 1
> > > 
> > > My USB keyboard runtime usage reference count is not decreasing to zero.
> > > Here am i missing something regarding runtime usage count or is it some issue with
> > > my kernel or OS?
> > 
> > Very hard to say without further information. Is the HID device the
> > only interface of the whole device?
> yes, HID device is only the interface for given device, below here lsusb -t log for usb dev 1-5.
> 
> root@intel-Kabylake-Client-platform:# lsusb -t
> /:  Bus 01.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/12p, 480M
>     |__ Port 5: Dev 3, If 1, Class=Human Interface Device, Driver=usbhid, 1.5M
>     |__ Port 5: Dev 3, If 0, Class=Human Interface Device, Driver=usbhid, 1.5M
> 
> > Does the device support remote wakeup?
> yes, from lsusb logs it seems, it is supporting remote wakeup.
> 
> root@intel-Kabylake-Client-platform:# lsusb 
> Bus 001 Device 003: ID 04d9:1702 Holtek Semiconductor, Inc. Keyboard LKS02
> 
> root@intel-Kabylake-Client-platform:# lsusb -v -d 04d9:1702 | grep Remote
>       Remote Wakeup
> 
> > Is an LED on?
> Num lock led was on, but switching it off also does not help.
> > 
> > Usbhid does support autosuspend, but its criteria must be met.
> > And all interface drivers must call the device idle.
> All of its interfaces are in runtime suspend
> 
> root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5\:1.0/power/runtime_status 
> suspended
> root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5\:1.1/power/runtime_status 
> suspended
> root@intel-Kabylake-Client-platform:# cat /sys/bus/usb/devices/1-5/power/runtime_active_kids 
> 0

What matters is not the power/* settings for the interfaces themselves, 
but rather the power/* settings for the interfaces' children.

Alan Stern