Re: [PATCH v3 00/20] iio:adc:ad7280a Cleanup and proposed staging graduation.

[Marcelo Schmitt <marcelo.schmitt1@gmail.com>]

On 02/06, Jonathan Cameron wrote:
> From: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> 
> Changes since v2:
> All in response to Marcelo Schmitt doing a really detailed review and
> testing against the QEMU model.
> 
> - Moved elements of patch 2 to patch 5 to avoid missing defines.
> - Patch 6: Amend ordering to avoid updated software cached values until the
>   hardware write has succeeded.
> - Patch 12 Fixed wrong callbacks in no_irq case
> - Patch 14: Changed all numeric properties to u32 to avoid needing to
>   specify bit depth in dts.
> - Patch 14: Typo fix.
> - Patch 15: New Patch to drop the handling of partial aux alert channels as
>   it is buggy (and always was). Anyone wanting to bring this back will
>   hopefully have hardware to test as it will be fiddly.
> - Patch 15 (old number, now 16). Update dt bindings to drop the
>   adi,temp-alert-last-chan given new patch 15 drops the support.
> - New Patch 18: Change cb_mask software cache to not include the offset of 2
>   and instead apply that only when writing the registers.
> - New Patch 19: More conservative timings to allow 105 degree Centigrade
>   operation.
> - Patch 20 is moving the new code so obviously all the above apply to that
>   as well!

I had another look into this set (including the new patches), and it seems all good. :)
Also, thanks for clarifying my doubts about the driver on the previous thread.
So,
Reviewed-by: Marcelo Schmitt <marcelo.schmitt1@gmail.com>
for the whole series.

It will be nice to see another driver getting out of staging. :D

Thanks,

Marcelo

> 
> Hi All,
> 
> This one proved an interesting diversion.
> 
> Work done against a somewhat hacked up QEMU emulation of 3 daisy chained
> ad7280a devices (18 channels).  Note that the emulation isn't complete
> but does do chaining, CRC, and readout of channels etc in a fashion that
> worked with the original driver (up to the bug in patch 1) and continues
> to work with the updated version. I've not intention to upstream the
> emulation (as would need to make it more completed and flexible), but
> happy to share it with anyone who is interested.
> 
> I briefly flirted with posting a patch to just drop the driver entirely,
> but the part is still available and it looked like fun + isn't going
> to greatly impact maintainability of the subsystem long term so is low
> cost even if it becomes unavailable sometime soon.
> 
> There are lots of things we could do after this set to improved the driver
> and make things more flexible, but it should basically 'just work'
> 
> Anyhow, as normal for staging graduations, last patch has rename detection
> turned off so that people can easily see what I am proposing we move
> out of staging.
> 
> All comments welcome and thanks to Marcelo for reviewing this beast!
> 
> Thanks,
> 
> Jonathan
> 
> 
> Jonathan Cameron (20):
>   staging:iio:adc:ad7280a: Fix handing of device address bit reversing.
>   staging:iio:adc:ad7280a: Register define cleanup.
>   staging:iio:adc:ad7280a: rename _read() to _read_reg()
>   staging:iio:adc:ad7280a: Split buff[2] into tx and rx parts
>   staging:iio:adc:ad7280a: Use bitfield ops to managed fields in
>     transfers.
>   staging:iio:adc:ad7280a: Switch to standard event control
>   staging:iio:adc:ad7280a: Standardize extended ABI naming
>   staging:iio:adc:ad7280a: Drop unused timestamp channel.
>   staging:iio:adc:ad7280a: Trivial comment formatting cleanup
>   staging:iio:adc:ad7280a: Make oversampling_ratio a runtime control
>   staging:iio:adc:ad7280a: Cleanup includes
>   staging:iio:ad7280a: Reflect optionality of irq in ABI
>   staging:iio:adc:ad7280a: Use a local dev pointer to avoid &spi->dev
>   staging:iio:adc:ad7280a: Use device properties to replace platform
>     data.
>   staging:iio:adc:ad7280a: Drop buggy support for early termination of
>     AUX alert.
>   dt-bindings:iio:adc:ad7280a: Add binding
>   iio:adc:ad7280a: Document ABI for cell balance switches
>   staging:iio:adc:ad7280a: Remove shift from cb_mask state cache.
>   staging:iio:adc:ad7280a: Use more conservative delays to allow 105C
>     operation.
>   iio:adc:ad7280a: Move out of staging
> 
>  .../ABI/testing/sysfs-bus-iio-adc-ad7280a     |   13 +
>  .../bindings/iio/adc/adi,ad7280a.yaml         |   77 ++
>  drivers/iio/adc/Kconfig                       |   11 +
>  drivers/iio/adc/Makefile                      |    1 +
>  drivers/iio/adc/ad7280a.c                     | 1111 +++++++++++++++++
>  drivers/staging/iio/adc/Kconfig               |   11 -
>  drivers/staging/iio/adc/Makefile              |    1 -
>  drivers/staging/iio/adc/ad7280a.c             | 1044 ----------------
>  drivers/staging/iio/adc/ad7280a.h             |   37 -
>  9 files changed, 1213 insertions(+), 1093 deletions(-)
>  create mode 100644 Documentation/ABI/testing/sysfs-bus-iio-adc-ad7280a
>  create mode 100644 Documentation/devicetree/bindings/iio/adc/adi,ad7280a.yaml
>  create mode 100644 drivers/iio/adc/ad7280a.c
>  delete mode 100644 drivers/staging/iio/adc/ad7280a.c
>  delete mode 100644 drivers/staging/iio/adc/ad7280a.h
> 
> -- 
> 2.35.1
>