From mboxrd@z Thu Jan  1 00:00:00 1970
From: Mika Westerberg <mika.westerberg@linux.intel.com>
Subject: Re: ACPI: Can I use I2cSerialBus with a PCI I2C controller?
Date: Thu, 22 Oct 2015 11:01:57 +0300
Message-ID: <20151022080157.GI1526@lahna.fi.intel.com>
References: <CAE7DoPZs69A5vLzDfrRXuC3q9fpkaXRYna2NQwHZk_BK9gLnvQ@mail.gmail.com>
 <20151021085056.GV1526@lahna.fi.intel.com>
 <CAE7DoPak939WjCWaGQw7W6aR0dZOcJCjcQK8gMa9iu+WRmnDDw@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Return-path: <linux-acpi-owner@vger.kernel.org>
Received: from mga14.intel.com ([192.55.52.115]:6336 "EHLO mga14.intel.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1750993AbbJVIE2 (ORCPT <rfc822;linux-acpi@vger.kernel.org>);
	Thu, 22 Oct 2015 04:04:28 -0400
Content-Disposition: inline
In-Reply-To: <CAE7DoPak939WjCWaGQw7W6aR0dZOcJCjcQK8gMa9iu+WRmnDDw@mail.gmail.com>
Sender: linux-acpi-owner@vger.kernel.org
List-Id: linux-acpi@vger.kernel.org
To: Ben Gardner <gardner.ben@gmail.com>
Cc: linux-acpi@vger.kernel.org

On Wed, Oct 21, 2015 at 06:14:27PM -0500, Ben Gardner wrote:
> Thanks for the help!
> 
> >> So my questions are:
> >>   Can I use I2cSerialBus with a PCI I2C controller?
> >
> > Yes you can.
> >
> > That's what we do all the time for Intel hardware.
> 
> That is good to hear.
> ...
> 
> > So you can just drop the whole _CRS and make your I2C host controller
> > device to look like:
> >
> > Device (I2C3)
> > {
> >         Name (_ADR, 0x00180003)
> >
> >         /* Standard Mode: HCNT, LCNT, SDA Hold Time */
> >         Name (SSCN, Package () { 0x200, 0x200, 0x6 })
> >
> >         /* Fast Mode: HCNT, LCNT, SDA Hold Time */
> >         Name (FMCN, Package () { 0x55, 0x99, 0x6 })
> >
> >         Device (EEP0)
> >         {
> >                 Name (_CID, Package() { "24c02" })
> >                 Name (_CRS, ResourceTemplate () {
> >                         I2cSerialBus (0x0057, ControllerInitiated, 400000,
> >                                       AddressingMode7Bit, "\\_SB.I2C3", 0x00,
> >                                       ResourceConsumer,,)
> >                 })
> >         }
> > }
> 
> OK. I tried that.
> It looks like the device is nested properly in the ACPI tree:
> 
> I see this symlink:
> /sys/bus/acpi/devices/24C02:00 ->
> ../../../devices/LNXSYSTM:00/LNXSYBUS:00/device:16/24C02:00
> 
> # for x in /sys/bus/acpi/devices/* ; do if [ -e $x/path ] ; then
> printf "%-30.30s  %s\n" "$(cat $x/path)" "$x" ; fi ; done | sort |
> grep I2C3
> \_SB_.I2C3                      device:16
> \_SB_.I2C3.EEP0                 24C02:00
> 
> But no I2C device are created under /sys/bus/i2c/devices/.
> # ll /sys/bus/i2c/devices/
> lrwxrwxrwx    1 root     root             0 Oct 21 11:30 i2c-1 ->
> ../../../devices/pci0000:00/0000:00:18.1/i2c-1
> lrwxrwxrwx    1 root     root             0 Oct 21 11:30 i2c-2 ->
> ../../../devices/pci0000:00/0000:00:18.2/i2c-2
> lrwxrwxrwx    1 root     root             0 Oct 21 11:30 i2c-3 ->
> ../../../devices/pci0000:00/0000:00:18.3/i2c-3
> 
> I'll have to dig further to find where it is failing in i2c_core.c. I
> assume it will be in the same spot.
> Is there a way to see PCI to ACPI associations?

Yes, the PCI device has a symlink "firmware_node" which points to the
ACPI device.

Here an example from Skylake system (which uses PCI mode for LPSS
devices):

# ls -l /sys/bus/pci/devices/0000\:00\:15.1/firmware_node
lrwxrwxrwx    1 root     root             0 Oct 22 08:00 /sys/bus/pci/devices/0000:00:15.1/firmware_node -> ../../LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/device:6f

# cat /sys/bus/pci/devices/0000\:00\:15.1/firmware_node/path 
\_SB_.PCI0.I2C1