Remove the bme680 kernel driver initialization code from the mangOH
yellow module because the Bosch userspace driver is now being used
instead.

Use "full update" rather than "partial update" LUT every 10th update to
the display to help prevent "ghosting" display artifacts from
accumulating.

The BSEC temperature offset and ambient temperature required by the
bme680 driver library are now supplied dynamically. Additionally, an
openweathermap.org based ambient temperature API was added. This is of
course only useful if the device will be outdoors.

Add callinng init in ws_eink_display and sleep after displaying

The temperatures being read back are higher than expected.  Add some
logs until the issue is resolved.

Use the bme680 driver library provided by Bosch and package it as a
Legato component. This provides a more complete driver implementation
than the kernel module that was being used previously. Also integrate
the proprietary BSEC library from Bosch to get more intelligence out of
the sensor.

Add fb_ioctl to framebuffer ops help to control deepsleep mode
for waveshare eink display from user space.
In user space can use ioctl(myfd, command,.) to set sleep mode or wakeup

Update the driver to indicate that only the 2.13" display is supported
since this is the only display we have tested.

Store a pointer to the device properties struct in the device data
rather than setting module scoped variables in the probe. The benefit of
this is that the driver could conceivably be used to support two
displays with different display properties simultaneously.

Set the max SPI speed to 4MHz, remove the corresponding TODO entry and
also a TODO relating to previously completed endian changes.

Not merging the branch HEAD because the final commit is not required.

Call pcf85063_set_clkout_freq() rather than manually writing the
control_2 register in the probe function.

- Replace spaces used for indent with tabs
- Wrap lines only when necessary

There were two issues:
  - Boards with a matching DV6 EEPROM string were being marked as DV5.
  - Some early DV5 boards have EEPROM content indicating that they are
    DV4 boards, but they are actually DV5 boards. As a result, there are
    now two strings that are used to identify DV5 boards.

Fix a copy/paste error in the mangOH yellow module. Previously, the IMU
variable was being checked after calling i2c_new_device to create the
magnetometer. The impact of this bug is that a failed probe of the
magnetometer would not have been detected during probe of the mangOH
yellow driver.

Update the deffered IO delay to 1 HZ base on document from https://www.kernel.org/doc/Documentation/fb/deferred_io.txt

Remove the endian-specific code from driver. It is not nesesary because ARM processors were little endians  so dont need to check