-----Original Message-----
From: Andy Shevchenko <redacted>
Sent: Tuesday, July 20, 2021 5:08 PM
To: Miclaus, Antoniu <Antoniu.Miclaus@analog.com>
Cc: linux-iio@vger.kernel.org; linux-kernel@vger.kernel.org;
jic23@kernel.org; devicetree@vger.kernel.org
Subject: Re: [PATCH v6 1/2] iio: frequency: adrf6780: add support for
ADRF6780

[External]

On Tue, Jul 20, 2021 at 4:17 PM Miclaus, Antoniu
[off-list ref] wrote:

quoted

quoted

From: Andy Shevchenko <redacted>
Sent: Friday, July 16, 2021 5:53 PM
On Fri, Jul 16, 2021 at 2:43 PM Antoniu Miclaus
[off-list ref] wrote:

...

quoted

quoted

First question is why not to use the regmap API (I have heard it has
gained support of 17 bit)?

Initially that was the plan, but after this patch:

https://urldefense.com/v3/__https://github.com/torvalds/linux/commit/41
91f19792bf91267835eb090d970e9cd6277a65__;!!A3Ni8CS0y2Y!ptUV0YC2nfD
6AdH_y5U0ziyELl4B9pDL0ubkdpFHFtrFNE_NqUS_TWm_gE-SlHV315Ak$

quoted

the custom write formats for regmap allow the read only via cached

registers.

quoted

Therefore, I preferred using spi transfers for write/read to/from the

device.

Not sure I follow you. That patch is upstream. Does it prevent you
from switching to regmap SPI API?

It does not prevent me from switching to regmap SPI API.
It will prevent me from using regmap_read to read directly from device.
https://github.com/torvalds/linux/commit/4191f19792bf91267835eb090d970e9cd6277a65#diff-3e0b7d2f0a55adb6573693a514cb8024a81a55da848cb22632ab5fd4dd6dd4e4R39

...

quoted

quoted

quoted

+        depends on COMMON_CLK

Is it mandatory for any function inside the device?

Yes. It will serve as LO input to the device.

But can the device work without it (with limited functionality)?

This aspect is not clearly stated in the datasheet, but since the Enable register has the LO enabled by default, I considered this aspect as being mandatory.
I will try to find out more information from the hardware guys who developed the part.

--
With Best Regards,
Andy Shevchenko