Ad7124 6.12 update

[thorenscientific]

Description

Brings the AD7124-8 class, examples, and dependent classes up-to-date with updated Linux driver, which in turn fixes a several bugs (like buffered channels becoming misaligned) and adds temperature sensor, calibration, and cleans up the filter type attributes.

Type of change

• Bug fix
• New device class interface
• New feature on an existing class
• Breaking change
• Documentation only
• Test or CI only

How has this been tested?

State explicitly whether this was tested against real hardware, an emulated context (iio-emu), or not run. Name the part(s) and the context URI or board used.

• Hardware:
  CN0554 (AD7124-8 and LTC2688) on Raspberry Pi 4

• Commands run (e.g. python3 -m pytest -k ad4080 --adi-hw-map):
  python -m pytest --color yes -vs test\test_ad7124_8.py --uri="ip:analog.local" --hw=ad7124-8

• Result:
  test/test_ad7124_8.py::test_ad7124_8_channels PASSED
  test/test_ad7124_8.py::test_ad7124_8_rx_data[channel=[1, 2, 3, 4]-classname=adi.ad7124] PASSED

• Emulation:
  CI emulated hardware test

• Result:
  Passes

Test configuration

• Kernel / libiio version: 6.12.41-v8+ / 0.26
• OS: ADI Kuiper Linux on Raspberry Pi 4, Windows 11 on remote host

Documentation

• No documentation change needed
• Updated existing docs under doc/source/
• Added a new device page and linked it from the relevant toctree

New device class interfaces

Skip this section if the PR does not add a new device class.

• Class extends one of the common base classes in adi.device_base (rx_chan_comp, tx_chan_comp, or a _no_buff variant). If not, explain why in the description. See the Device Base Classes developer doc page.
• compatible_parts lists every part number this class supports
• Part numbers added to supported_parts.md (verify with invoke checkparts)
• Emulation context (xml_gen, not iio_genxml) added under test/emu/ and referenced from a test
• Tests added that run against the emulation context in CI

Checklist

• invoke precommit passes locally
• All commits are signed off (Signed-off-by: Name <email>)
• No new warnings from the build or doc generation
• Dependent changes in libiio, kernel, or HDL are merged and referenced

[thorenscientific]

@tfcollins for the record I did not check the case where self._channel_def IS defined.

[thorenscientific]

@tfcollins the dict-based channel-definition checks out, thanks! Ran the ad7125_example, cn0554_loopback, and the pytest fixture. Anything else to address?