Re: [PATCH v2 0/9] iio: accel: bmc150: Add support for yoga's with dual accelerometers with an ACPI HID of DUAL250E

[Jonathan Cameron <jic23@kernel.org>]

On Sun, 23 May 2021 22:08:57 +0300
Andy Shevchenko <andy.shevchenko@gmail.com> wrote:

> On Sun, May 23, 2021 at 8:01 PM Hans de Goede <hdegoede@redhat.com> wrote:
> >
> > Hi all,
> >
> > Here is v2 of this series, addressing Andy's review remarks and
> > rewrap some comments at 80 chars limit.
> >
> > For more info here is the v1 cover-letter:
> >
> > Some 360 degree hinges (yoga) style 2-in-1 devices use 2 bmc150 accels
> > to allow the OS to determine the angle between the display and the base
> > of the device, so that the OS can determine if the 2-in-1 is in laptop
> > or in tablet-mode.
> >
> > We already support this setup on devices using a single ACPI node
> > with a HID of "BOSC0200" to describe both accelerometers. This patch
> > set extends this support to also support the same setup but then
> > using a HID of "DUAL250E".
> >
> > While testing this I found some crashes on rmmod, patches 1-2
> > fix those patches, patch 3 does some refactoring and patch 4
> > adds support for the "DUAL250E" HID.
> >
> > Unfortunately we need some more special handling though, which the
> > rest of the patches are for.
> >
> > On Windows both accelerometers are read (polled) by a special service
> > and this service calls a DSM (Device Specific Method), which in turn
> > translates the angles to one of laptop/tablet/tent/stand mode and then
> > notifies the EC about the new mode and the EC then enables or disables
> > the builtin keyboard and touchpad based in the mode.
> >
> > When the 2-in-1 is powered-on or resumed folded in tablet mode the
> > EC senses this independent of the DSM by using a HALL effect sensor
> > which senses that the keyboard has been folded away behind the display.
> >
> > At power-on or resume the EC disables the keyboard based on this and
> > the only way to get the keyboard to work after this is to call the
> > DSM to re-enable it (similar to how we also need to call a special
> > DSM in the kxcjk-1013.c accel driver to re-enable the keyboard).
> >
> > Patches 5-7 deal with the DSM mess and patch 8 adds labels to the
> > 2 accelerometers specifying which one is which.  
> 
> Thanks!
> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Whole series now applied to the togreg branch of iio.git and pushed out
as testing to give 0-day a chance to pick on it.

Thanks,

Jonathan

> 
> > Regards,
> >
> > Hans
> >
> >
> > Hans de Goede (9):
> >   iio: accel: bmc150: Fix dereferencing the wrong pointer in
> >     bmc150_get/set_second_device
> >   iio: accel: bmc150: Don't make the remove function of the second
> >     accelerometer unregister itself
> >   iio: accel: bmc150: Move check for second ACPI device into a separate
> >     function
> >   iio: accel: bmc150: Add support for dual-accelerometers with a
> >     DUAL250E HID
> >   iio: accel: bmc150: Move struct bmc150_accel_data definition to
> >     bmc150-accel.h
> >   iio: accel: bmc150: Remove bmc150_set/get_second_device() accessor
> >     functions
> >   iio: accel: bmc150: Add support for DUAL250E ACPI DSM for setting the
> >     hinge angle
> >   iio: accel: bmc150: Refactor bmc150_apply_acpi_orientation()
> >   iio: accel: bmc150: Set label based on accel-location for ACPI
> >     DUAL250E fwnodes
> >
> >  drivers/iio/accel/bmc150-accel-core.c | 110 +++++----------
> >  drivers/iio/accel/bmc150-accel-i2c.c  | 193 ++++++++++++++++++++++----
> >  drivers/iio/accel/bmc150-accel.h      |  66 ++++++++-
> >  3 files changed, 260 insertions(+), 109 deletions(-)
> >
> > --
> > 2.31.1
> >  
> 
>