Michael & Stephen,
Should we resubmit this series for 4.7 ?
Also to follow up on your previous request to use the platform driver
framework. We have noticed recent submissions use a style that mixes
early clocks registered with CLK_OF_DECLARE and the remaining clocks
registered through a platform driver. Is this the style we should
implement ?
Best Regards,
Lars Persson
On 02/25/2016 03:58 PM, Lars Persson wrote:
> Add clock support for the Artpec-6 SoC port. The ARM parts are in the series
> "arm: Add Artpec-6 SoC" and it goes through the arm-soc tree.
>
> Changes since v3:
> - The binding was corrected to handle two fractional divider clocks as input to
> the clock controller.
> - Updated clk-artpec6.c to handle a distinct fractional divider input for each
> i2s clock mux.
>
> Changes since v2:
> - Moved axis,artpec6-clkctrl.h to the first patch with the DT bindings.
>
> Changes since v1:
>
> - The driver now provides all clocks from the main clock controller block
> through one DT node.
> - Added a header file for the clock indexes.
> - Refer to clock-bindings.txt in the bindings document.
>
> A platform driver was not possible because the clocks are needed earlier in the
> kernel startup.
>
> Lars Persson (2):
> clk: add device tree binding for Artpec-6 clock controller
> clk: add artpec-6 clock controller
>
> .../devicetree/bindings/clock/artpec6.txt | 41 +++++
> drivers/clk/Makefile | 1 +
> drivers/clk/axis/Makefile | 1 +
> drivers/clk/axis/clk-artpec6.c | 189 +++++++++++++++++++++
> include/dt-bindings/clock/axis,artpec6-clkctrl.h | 38 +++++
> 5 files changed, 270 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/clock/artpec6.txt
> create mode 100644 drivers/clk/axis/Makefile
> create mode 100644 drivers/clk/axis/clk-artpec6.c
> create mode 100644 include/dt-bindings/clock/axis,artpec6-clkctrl.h
>
On 03/23, Lars Persson wrote:
> Michael & Stephen,
>
> Should we resubmit this series for 4.7 ?
>
> Also to follow up on your previous request to use the platform
> driver framework. We have noticed recent submissions use a style
> that mixes early clocks registered with CLK_OF_DECLARE and the
> remaining clocks registered through a platform driver. Is this the
> style we should implement ?
>
Yes that would be good. Thanks.
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project