Re: [PATCH 0/3] IIO: st_sensors_i2c: improve device enumeration

On Fri, Jun 29, 2018 at 1:30 PM, Nikolaus Voss
[off-list ref] wrote:

When trying to instantiate a st_accel_i2c device from an ACPI based
system, I ran into some problems:

For my device, there is no ACPI match table entry, so rather than
creating /allocating a new ACPI HID for the device, I wanted to use an
existing DT table compatible entry via creating an ACPI_DT_NAMESPACE_HID
/PRP0001 HID ACPI entry (see Documentation/acpi/enumeration.txt).

This did not work because st_accel_i2c.c bails out if there is a ACPI
node but no ACPI table match instead of looking for a match from one of
the fallback mechanisms (patch 1).

Patch 2 removes an error message when a ACPI node exists but no table
entry is found (this doesn't need to be fatal because of the fallback).

Patch 3 syncs the strings in the I2C device table (which is used by the
I2C core for fallback matching) with the DT compatible strings, so a
PRP0001 entry can use the same compatible strings as a corresponding
DT entry. As far as I can see, the old I2C table strings aren't used
in mainline, so renaming them should be safe.

I'm not sure I understand how ->probe_new() is supposed to work
against i2c_id_table, but I don't care for legacy platform data
anyway.

What I would like to point to is device_get_match_data() API which
should simplify / unify the case how you get driver data.

Nikolaus Voss (3):
  IIO: st_accel_i2c.c: Use fallback if DT/ACPI enum failed
  IIO: st_sensors_i2c.c: Don't print error on failed ACPI match
  IIO: st_accel.h: sync DT and I2C device ID table strings

 drivers/iio/accel/st_accel.h                  | 32 +++++++++----------
 drivers/iio/accel/st_accel_i2c.c              | 21 ++++++------
 .../iio/common/st_sensors/st_sensors_i2c.c    |  5 ++-
 3 files changed, 30 insertions(+), 28 deletions(-)

--
2.17.1

-- 
With Best Regards,
Andy Shevchenko