Re: [PATCH v2 5/7] clk: tegra124: Add support for Tegra124 clocks

[Stephen Warren <swarren-3lzwWm7+Weoh9ZMKESR00Q@public.gmane.org>]

On 10/15/2013 09:14 AM, Peter De Schrijver wrote:
> Implement clock support for Tegra124.

> diff --git a/drivers/clk/tegra/clk-tegra124.c b/drivers/clk/tegra/clk-tegra124.c

> +static struct pdiv_map pll12g_ssd_esd_p[] = {
> +	{ .pdiv = 1, .hw_val = 0 },
> +	{ .pdiv = 2, .hw_val = 1 },
...
> +	{ .pdiv = 24, .hw_val = 13 },
> +	{ .pdiv = 32, .hw_val = 14 },
> +	{ .pdiv = 0, .hw_val = 15 },
> +};

I'm curious why that last entry doesn't have .hw_val = 0, since I think
it's a sentinel value. For example, the last entry in pllxc_p[] does
have .hw_val = 0.

> diff --git a/include/dt-bindings/clock/tegra124-car.h b/include/dt-bindings/clock/tegra124-car.h

I think you also need to create
Documentation/devicetree/bindings/clock/nvidia,tegra124-car.txt in this
patch. All the other SoCs have their own binding file. That said, given
all the clock IDs have moved into <dt-bindings/clock/tegraNNN-car.h>,
perhaps we should just have a separate patch that removes the separate
bindings for Tegra30 and Tegra114, and simply add their compatible
values into the existing nvidia,tegra20-car.txt (and also make the
header file reference in that file more generic so it applies to any
Tegra SoC)?

> @@ -0,0 +1,341 @@
> +/*
> + * This header provides constants for binding nvidia,tegra124-car.
> + *
> + * The first 185 clocks are numbered to match the bits in the CAR's CLK_OUT_ENB
> + * registers.

Shouldn't that be 192 (== 185 * 32)? Perhaps some bits aren't allocated,
but in previous SoCs, I rounded the ID space for peripheral clocks up to
whole registers. I see that as far as the clock IDs, this is already
fine; it's just this description that says 185 not 192.