[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <090bbaee-88e0-42fa-b43d-db80ec065d35@linaro.org>
Date: Thu, 13 Mar 2025 08:46:15 +0100
From: Krzysztof Kozlowski <krzysztof.kozlowski@...aro.org>
To: Jameson Thies <jthies@...gle.com>, tzungbi@...nel.org,
ukaszb@...omium.org, bleung@...omium.org, heikki.krogerus@...ux.intel.com,
robh+dt@...nel.org, krzysztof.kozlowski+dt@...aro.org, groeck@...omium.org,
swboyd@...omium.org, akuchynski@...omium.org
Cc: devicetree@...r.kernel.org, chrome-platform@...ts.linux.dev,
linux-kernel@...r.kernel.org, linux-usb@...r.kernel.org
Subject: Re: [PATCH v1 3/3] mfd: cros_ec: Don't add cros_ec_ucsi if it is
defined in OF or ACPI
On 13/03/2025 08:11, Krzysztof Kozlowski wrote:
> On 12/03/2025 20:59, Jameson Thies wrote:
>> Check for cros_ec_ucsi to be defined in the OF device tree or an ACPI
>> node. If it is defined by either OF or ACPI, it does not need to be
>> added as a subdevice of cros_ec_dev.
>
> No, it does not have to. You just populate the children and appropriate
Uh, I did not notice that it is !of_find_compatible_node(), so this
comment should be rephrased - you just add MFD children anyway and if
there is no node, they won't be created.
> devices will be created automatically. None of parent devices should
> ever check if the child exist to create a child - it makes no sense.
This is still valid - none of parents should be poking around to see if
there is a child or not. The core handles it, DT handles it etc.
Best regards,
Krzysztof
Powered by blists - more mailing lists