Am Sonntag, dem 28.11.2021 um 07:18 -0600 schrieb Adam Ford:
> The i.MX8M-Nano features a GC7000. The Etnaviv driver detects it as:
>
> etnaviv-gpu 38000000.gpu: model: GC7000, revision: 6203
>
> Signed-off-by: Adam Ford <aford173@xxxxxxxxx>
> ---
> arch/arm64/boot/dts/freescale/imx8mn.dtsi | 25 +++++++++++++++++++++++
> 1 file changed, 25 insertions(+)
>
> diff --git a/arch/arm64/boot/dts/freescale/imx8mn.dtsi b/arch/arm64/boot/dts/freescale/imx8mn.dtsi
> index d8726d0ce326..5b8f8488e362 100644
> --- a/arch/arm64/boot/dts/freescale/imx8mn.dtsi
> +++ b/arch/arm64/boot/dts/freescale/imx8mn.dtsi
> @@ -1117,6 +1117,31 @@ gpmi: nand-controller@33002000 {
> status = "disabled";
> };
>
> + gpu: gpu@38000000 {
> + compatible = "vivante,gc";
> + reg = <0x38000000 0x8000>;
> + interrupts = <GIC_SPI 3 IRQ_TYPE_LEVEL_HIGH>;
> + clocks = <&clk IMX8MN_CLK_GPU_AHB>,
> + <&clk IMX8MN_CLK_GPU_BUS_ROOT>,
> + <&clk IMX8MN_CLK_GPU_CORE_ROOT>,
> + <&clk IMX8MN_CLK_GPU_SHADER>;
> + clock-names = "reg", "bus", "core", "shader";
> + assigned-clocks = <&clk IMX8MN_CLK_GPU_CORE>,
> + <&clk IMX8MN_CLK_GPU_SHADER>,
> + <&clk IMX8MN_CLK_GPU_AXI>,
> + <&clk IMX8MN_CLK_GPU_AHB>,
> + <&clk IMX8MN_GPU_PLL>,
> + <&clk IMX8MN_CLK_GPU_CORE>,
> + <&clk IMX8MN_CLK_GPU_SHADER>;
This repeated CORE and SHADER clock looks odd. Wouldn't it be possible
to avoid this by reordering the assigned-clocks?
Regards,
Lucas
> + assigned-clock-parents = <&clk IMX8MN_GPU_PLL_OUT>,
> + <&clk IMX8MN_GPU_PLL_OUT>,
> + <&clk IMX8MN_SYS_PLL1_800M>,
> + <&clk IMX8MN_SYS_PLL1_800M>;
> + assigned-clock-rates = <0>, <0>, <800000000>, <400000000>, <1200000000>,
> + <400000000>, <400000000>;
> + power-domains = <&pgc_gpumix>;
> + };
> +
> gic: interrupt-controller@38800000 {
> compatible = "arm,gic-v3";
> reg = <0x38800000 0x10000>,
> --
> 2.32.0
>
