Re: [PATCH V2 1/7] i2c: sprd: Add configurations that support 1Mhz and 3.4Mhz frequencies

[Andi Shyti]

Hi Huangzheng,

On Mon, Oct 23, 2023 at 04:11:52PM +0800, Huangzheng Lai wrote:
> Add support for 1Mhz and 3.4Mhz frequency configuration.
> 
> Signed-off-by: Huangzheng Lai <Huangzheng.Lai@xxxxxxxxxx>
> Acked-by: Andi Shyti <andi.shyti@xxxxxxxxxx>

you can make this:

Reviewed-by: Andi Shyti <andi.shyti@xxxxxxxxxx>

> ---
>  drivers/i2c/busses/i2c-sprd.c | 25 ++++++++++++++++++++-----
>  1 file changed, 20 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/i2c/busses/i2c-sprd.c b/drivers/i2c/busses/i2c-sprd.c
> index ffc54fbf814d..b44916c6741d 100644
> --- a/drivers/i2c/busses/i2c-sprd.c
> +++ b/drivers/i2c/busses/i2c-sprd.c
> @@ -343,10 +343,23 @@ static void sprd_i2c_set_clk(struct sprd_i2c *i2c_dev, u32 freq)
>  	writel(div1, i2c_dev->base + ADDR_DVD1);
>  
>  	/* Start hold timing = hold time(us) * source clock */
> -	if (freq == I2C_MAX_FAST_MODE_FREQ)
> -		writel((6 * apb_clk) / 10000000, i2c_dev->base + ADDR_STA0_DVD);
> -	else if (freq == I2C_MAX_STANDARD_MODE_FREQ)
> +	switch (freq) {
> +	case I2C_MAX_STANDARD_MODE_FREQ:
>  		writel((4 * apb_clk) / 1000000, i2c_dev->base + ADDR_STA0_DVD);
> +		break;
> +	case I2C_MAX_FAST_MODE_FREQ:
> +		writel((6 * apb_clk) / 10000000, i2c_dev->base + ADDR_STA0_DVD);
> +		break;
> +	case I2C_MAX_FAST_MODE_PLUS_FREQ:
> +		writel((8 * apb_clk) / 10000000, i2c_dev->base + ADDR_STA0_DVD);
> +		break;
> +	case I2C_MAX_HIGH_SPEED_MODE_FREQ:
> +		writel((8 * apb_clk) / 10000000, i2c_dev->base + ADDR_STA0_DVD);
> +		break;

so these were defined but not used.

> +	default:
> +		dev_err(i2c_dev->dev, "Unsupported frequency: %d\n", freq);
> +		break;
> +	}
>  }
>  
>  static void sprd_i2c_enable(struct sprd_i2c *i2c_dev)
> @@ -519,9 +532,11 @@ static int sprd_i2c_probe(struct platform_device *pdev)
>  	if (!of_property_read_u32(dev->of_node, "clock-frequency", &prop))
>  		i2c_dev->bus_freq = prop;
>  
> -	/* We only support 100k and 400k now, otherwise will return error. */
> +	/* We only support 100k\400k\1m\3.4m now, otherwise will return error. */
>  	if (i2c_dev->bus_freq != I2C_MAX_STANDARD_MODE_FREQ &&
> -	    i2c_dev->bus_freq != I2C_MAX_FAST_MODE_FREQ)
> +			i2c_dev->bus_freq != I2C_MAX_FAST_MODE_FREQ &&
> +			i2c_dev->bus_freq != I2C_MAX_FAST_MODE_PLUS_FREQ &&
> +			i2c_dev->bus_freq != I2C_MAX_HIGH_SPEED_MODE_FREQ)
>  		return -EINVAL;

This is a bit of a weird management of the bus_freq because it
goes like this:

	i2c_dev->bus_freq = I2C_MAX_STANDARD_MODE_FREQ;
	...
	if (!of_property_read_u32(dev->of_node, "clock-frequency", &prop))
		i2c_dev->bus_freq = prop;

	if (...)
		return -EINVAL;

A follow-up cleanup can be removing the first assignement and
instead of returning -EINVAL, we can keep going:

	if (!of_property_read_u32(dev->of_node, "clock-frequency", &prop))
		i2c_dev->bus_freq = prop;

	if (...) {
		dev_warn("...");
		i2c_dev->bus_freq = I2C_MAX_STANDARD_MODE_FREQ;
	}

But this is topic for a different patch.

Andi

>  
>  	ret = sprd_i2c_clk_init(i2c_dev);
> -- 
> 2.17.1
>