Re: xHCI and suspend/resume

On Wednesday, May 18, 2011, Dwight Schauer wrote:

On Tue, May 17, 2011 at 8:59 PM, Xu, Andiry [off-list ref] wrote:

quoted

quoted

-----Original Message-----
From: linux-usb-owner@vger.kernel.org [mailto:linux-usb-
owner@vger.kernel.org] On Behalf Of Dwight Schauer
Sent: Wednesday, May 18, 2011 7:02 AM
To: Rafael J. Wysocki
Cc: linux-pm@lists.linux-foundation.org; Sarah Sharp; dschauer@ti.com;
USB list
Subject: Re: [linux-pm] xHCI and suspend/resume

On Tue, May 17, 2011 at 5:56 PM, Rafael J. Wysocki [off-list ref] wrote:

quoted

On Wednesday, May 18, 2011, Dwight Schauer wrote:

quoted

On Tue, May 17, 2011 at 4:54 PM, Rafael J. Wysocki [off-list ref]

wrote:

quoted

quoted

quoted

On Tuesday, May 17, 2011, Dwight Schauer wrote:

quoted

On Tue, May 17, 2011 at 4:43 PM, Rafael J. Wysocki [off-list ref]

wrote:

quoted

quoted

quoted

...

quoted

quoted

OK, one more test, please.

Try to do

# echo core > /sys/power/pm_test
# echo mem > /sys/power/state

(that should simulate suspend, but without going into the BIOS,

and it

quoted

quoted

quoted

quoted

quoted

should return do the command prompt after 5-10 sec.) and check

if the

quoted

quoted

quoted

quoted

quoted

USB3 controllers work after that ("echo none >

/sys/power/pm_test" resets

quoted

quoted

quoted

quoted

quoted

to the normal suspend behavior).

Thanks,
Rafael

No problem.

The simulated suspend works fine.

Good.

quoted

Also, waking up from S3 via a PS/2 keyboard works fine.

Hmm.  Do you mean that the USB3 controllers work after the resume

if

quoted

quoted

quoted

the box has been woken up via the keyboard and they don't work

when it

quoted

quoted

quoted

has been woken up via a power button?

Rafael

They work either way. I'm just not able to get the USB keyboard that
is connected to the USB3 controller to perform the wakeup. (Which is
what I thought I made clear from the beginning and what Alan Stern

had

quoted

quoted

reiterated for new readers further into the thread).

Oh, I must have missed that information.  Sorry about that.

So, the situation is that if you set up the USB3 controllers to wake

up

quoted

and next you wake up the system from S3 using a USB device connected

to one

quoted

of those controllers, then they appear to be in D3 after the resume

and

quoted

apparently cannot be put into D0.  However, if the wakeup is done in

any

quoted

different way, they work correctly after the resume, right?

Is that the case on both the affected systems?

On both systems the USB3 controllers plugged into PCIe slots work fine
after the system has been woken up from S3 suspend state.

The USB3 keyboard plugged into the PCIe USB3 controller is immediately
avaiable after the system has woken up. It is just that in Linux a
keyboard plugged into the USB3 controller can not be used to wake the
system up when it has gone into S3 suspend. With Windows 7 on our test
systems our USB3 contoller (the Texas Instruments Inc. PCEe xHCI
device) a USB keyboard plugged into it can wake it up from an S3
suspend.

I've verified that USB keyboard under xHCI controller can wakeup system
from S3, either by press the keyboard or plug it in during suspend. But
you need to enable wakeup in /proc/acpi/wakeup for the corresponding
xHC controller.

Thanks,
Andiry

I was told /proc/acpi/wakeup was deprecated, and to use
/sys/devices/pci.../power/wakeup instead.

That is correct, but /proc/acpi/wakeup still contains valid information
and is generally useful for debugging, so I'd appraciate it if you could
tell me what's in there.

Anyways, the controllers in question are not showing up in
/proc/acpi/wakeup. These are add on cards, not devices included on the
motherboard.

In that case you may try to enable wakeup for the bridges they are
attached to.  Hard to say without any more information.

Thanks,
Rafael