lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [day] [month] [year] [list]
Message-ID: <20150615124152.GA3311637@phare.normalesup.org>
Date:	Mon, 15 Jun 2015 14:41:53 +0200
From:	Nicolas George <george@...p.org>
To:	linux-kernel@...r.kernel.org
Subject: Help configuring kernel for Baytrail system

Hi.

I am trying to configure a kernel for a Baytrail SoC-based hybrid laptop (a
Lenovo Miix 3-1030). Amongst other things, the battery gauge is not
detected. I suspect it may be related to the following commit:

# commit 894acb2f823b13afacfe40b02efbd9146af58586
# Author: David Box <david.e.box@...ux.intel.com>
# Date:   Thu Jan 15 01:12:17 2015 -0800
# 
#     i2c: designware: Add Intel Baytrail PMIC I2C bus support

It is very likely that I am missing a simple compile-time or run-time
option, and in that case, on top of getting the configuration right, I have
a second question: how could I find the solution by myself?

Here is the problem data:

- The battery gauge works with the kernel from the Ubuntu 15.4 live image.
  Quite a few other things fail to work with that kernel, so using it is not
  a solution, but it is proof that it can work.

- I have tried with 3.16 from Debian, 4.0.1 and 4.1-rc* with custom
  configuration, and the battery is completely invisible. Nothing in dmesg.

- The sysfs entry for the battery is this:
  /sys/devices/LNXSYSTM:00/LNXSYBUS:00/80860F41:00/PNP0C0A:00
  It should contain a driver link to bus/acpi/drivers/battery, but it does
  not.

- Adding custom debug messages, I was able to find that
  acpi_bus_get_status_handle() evaluates the _STA attribute of the BATC
  device to 0, meaning the device is not present.

- The corresponding code in the ACPI tables is this:

        Device (BATC)
        {
            [...]
            Method (_STA, 0, NotSerialized)  // _STA: Status
            {
                If ((AVBL == One))
                {
                    BUF7 = CMD7 /* \_SB_.I2C1.CMD7 */
                    If ((STAT != Zero))
                    {
                        Return (Zero)
                    }

                    Return (0x1F)
                }

                Return (Zero)
            }

  but I do not know how to debug further. Enabling full ACPI debug messages
  gives so much output that it is unusable.

- I also suspect that the issue comes from the "dev->pm_runtime_disabled"
  flag on the corresponding I2C bus entry: it is enabled by
  i2c_dw_eval_lock_support() with a message "I2C bus managed by PUNIT", and
  inhibits probing the device. I did not manage to find how to re-enable it.


Please let me know if you have any advice on this issue in particular or any
more general documentation on how to investigate I2C and PLATFORM/ACPI-based
systems.

In the meantime, I have summarized what I was able to get working on this
laptop on the following webpage:
http://nsup.org/~george/comp/linux_lenovo_miix3/

Regards,

-- 
  Nicolas George

Download attachment "signature.asc" of type "application/pgp-signature" (820 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ