Re: [PATCH] clk: si5351: Add setup steps

[Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>]

On 19.11.2015 14:40, Jacob Siverskog wrote:
> This is according to figure 12 ("I2C Programming Procedure") in
> "Si5351A/B/C Data Sheet"
> (https://www.silabs.com/Support%20Documents/TechnicalDocs/Si5351-B.pdf).
> 
> Without the PLL soft reset, we were unable to get three outputs
> working at the same time.
> 
> According to Silicon Labs support, performing PLL soft reset will only
> be noticable if the PLL parameters have been changed.
> 
> Signed-off-by: Jacob Siverskog <jacob@teenage.engineering>
> Signed-off-by: Jens Rudberg <jens@teenage.engineering>
> ---
>  drivers/clk/clk-si5351.c | 13 +++++++++++++
>  1 file changed, 13 insertions(+)

Jacob,

thanks for the patches! However, besides Mareks comments I have
some more below.

> diff --git a/drivers/clk/clk-si5351.c b/drivers/clk/clk-si5351.c
> index e346b22..110de35 100644
> --- a/drivers/clk/clk-si5351.c
> +++ b/drivers/clk/clk-si5351.c
> @@ -1091,6 +1091,11 @@ static int si5351_clkout_set_rate(struct clk_hw *hw, unsigned long rate,
>  	si5351_set_bits(hwdata->drvdata, SI5351_CLK0_CTRL + hwdata->num,
>  			SI5351_CLK_POWERDOWN, 0);
>  
> +	/* do a pll soft reset on both plls, needed in some cases to get all
> +	 * outputs running */
> +	si5351_reg_write(hwdata->drvdata, SI5351_PLL_RESET,
> +			 SI5351_PLL_RESET_A | SI5351_PLL_RESET_B);
> +
>  	dev_dbg(&hwdata->drvdata->client->dev,
>  		"%s - %s: rdiv = %u, parent_rate = %lu, rate = %lu\n",
>  		__func__, clk_hw_get_name(hw), (1 << rdiv),
> @@ -1359,6 +1364,14 @@ static int si5351_i2c_probe(struct i2c_client *client,
>  		return PTR_ERR(drvdata->regmap);
>  	}
>  
> +	/* Disable outputs */
> +	si5351_reg_write(drvdata, SI5351_OUTPUT_ENABLE_CTRL, 0xff);
> +
> +	/* Power down all output drivers */
> +	for (n = 0; n < 8; n++) {
> +		si5351_reg_write(drvdata, SI5351_CLK0_CTRL + n, 0x80);
> +	}

Is disabling outputs and clock drivers required?

If we disable the clocks here unconditionally, it will
break those systems that require specific outputs to be always
enabled. Consider one clock output driving your SoC or similar.

If it is really required, you'll have to mention that in the
patch description and we need to find a way to tag specific
outputs to be never disabled.

Sebastian

>  	/* Disable interrupts */
>  	si5351_reg_write(drvdata, SI5351_INTERRUPT_MASK, 0xf0);
>  	/* Ensure pll select is on XTAL for Si5351A/B */
>