Re: [PATCHv4 8/9] ASoC: fsl-asoc-card: add DT property "cpu-system-clock-direction-out"

[Elinor Montmasson <elinor.montmasson@savoirfairelinux.com>]

From: "Mark Brown" <broonie@kernel.org>
Sent: Thursday, 16 May, 2024 14:18:00
> On Wed, May 15, 2024 at 03:54:10PM +0200, Elinor Montmasson wrote:
>> Add new optional DT property "cpu-system-clock-direction-out" to set
>> sysclk direction as "out" for the CPU DAI when using the generic codec.
>> It is set for both Tx and Rx.
>> If not set, the direction is "in".
>> The way the direction value is used is up to the CPU DAI driver
>> implementation.
> 
> This feels like we should be using the clock bindings to specify the
> clock input of whatever is using the output from the SoC, though that's
> a lot more work.

Similarly to patch 7/9, I exposed this parameter because the driver has it, and
because there might be CPU DAIs needing this parameter.
Otherwise the cpu sysclk direction will always be IN as a default.
This parameter could be needed for cases with CPU DAIs, such as an SAI,
which can provide or consume Tx/Rx clocks.
For these devices I know the sysclk direction should correspond to
what was set in the dai format.

This new compatible is intended to be used when there is no codec
device/driver. There is technically no codec device/driver for which
the clock input can be set.

Is it a bad idea to allow setting the cpu sysclk direction only ?
Should the compatible be limited to use-cases where the cpu sysclk
direction cannot be set by the machine driver ?

Best regards,
Elinor Montmasson