mbox series

[0/2] arm64: zynqmp: Move firmware constants from binding to platform

Message ID cover.1738600745.git.michal.simek@amd.com (mailing list archive)
Headers show
Series arm64: zynqmp: Move firmware constants from binding to platform | expand

Message

Michal Simek Feb. 3, 2025, 4:39 p.m. UTC 
Hi,

Based on discussion done long time ago
https://lore.kernel.org/all/5353872c-56a3-98f9-7f22-ec1f6c2ccdc8@linaro.org/
it is better to deprecate firmware contants which are not used in any code
and it's only purpose is to use macros in dt files.

There is no reason to be the part of dt bindings but it should be kept
there with adding deprecated warning.

I want to see if this is the right way to go. If yes, I will also do the
same changes in other headers which contain only constans which are not
used in any code.

Thanks,
Michal


Michal Simek (2):
  arm64: zynqmp: Use DT header for firmware constants
  dt-bindings: xilinx: Deprecate header with firmware constants

 .../bindings/ata/ceva,ahci-1v84.yaml          |   4 +-
 .../dma/xilinx/xlnx,zynqmp-dma-1.0.yaml       |   3 +-
 .../bindings/iio/adc/xlnx,zynqmp-ams.yaml     |   3 +-
 .../devicetree/bindings/net/cdns,macb.yaml    |   7 +-
 .../bindings/spi/spi-zynqmp-qspi.yaml         |   3 +-
 .../devicetree/bindings/usb/dwc3-xilinx.yaml  |   3 +-
 arch/arm64/boot/dts/xilinx/xlnx-zynqmp-clk.h  | 126 ++++++++++++++++++
 .../arm64/boot/dts/xilinx/zynqmp-clk-ccf.dtsi |   2 +-
 include/dt-bindings/clock/xlnx-zynqmp-clk.h   |   7 +
 9 files changed, 142 insertions(+), 16 deletions(-)
 create mode 100644 arch/arm64/boot/dts/xilinx/xlnx-zynqmp-clk.h

Comments

Conor Dooley Feb. 4, 2025, 7:43 p.m. UTC | #1 
On Mon, Feb 03, 2025 at 05:39:09PM +0100, Michal Simek wrote:
> Hi,
> 
> Based on discussion done long time ago
> https://lore.kernel.org/all/5353872c-56a3-98f9-7f22-ec1f6c2ccdc8@linaro.org/
> it is better to deprecate firmware contants which are not used in any code
> and it's only purpose is to use macros in dt files.
> 
> There is no reason to be the part of dt bindings but it should be kept
> there with adding deprecated warning.
> 
> I want to see if this is the right way to go. If yes, I will also do the
> same changes in other headers which contain only constans which are not
> used in any code.

Ye, I think this is a good approach to take.
Acked-by: Conor Dooley <conor.dooley@microchip.com>