mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Paul Cercueil <paul@crapouillou.net>
To: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
Cc: mturquette@baylibre.com, sboyd@kernel.org, robh+dt@kernel.org,
	krzysztof.kozlowski+dt@linaro.org, zhouyu@wanyeetech.com,
	linux-mips@vger.kernel.org, linux-clk@vger.kernel.org,
	devicetree@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v1 5/5] clk: ingenic: Add X1000 audio clocks
Date: Sun, 23 Oct 2022 16:34:58 +0100	[thread overview]
Message-ID: <AYP7KR.ZQW0GW51WJAG@crapouillou.net> (raw)
In-Reply-To: <20221023145653.177234-6-aidanmacdonald.0x0@gmail.com>

Hi Aidan,

Le dim. 23 oct. 2022 à 15:56:53 +0100, Aidan MacDonald 
<aidanmacdonald.0x0@gmail.com> a écrit :
> The X1000's CGU supplies the I2S system clock to the AIC module
> and ultimately the audio codec, represented by the "i2s" clock.
> It is a simple mux which can either pass through EXCLK or a PLL
> multiplied by a fractional divider (the "i2s_pll" clock).
> 
> The AIC contains a separate 1/N divider controlled by the I2S
> driver, which generates the bit clock from the system clock.
> The frame clock is always fixed to 1/64th of the bit clock.
> 
> Signed-off-by: Aidan MacDonald <aidanmacdonald.0x0@gmail.com>
> ---
>  drivers/clk/ingenic/x1000-cgu.c | 69 
> +++++++++++++++++++++++++++++++++
>  1 file changed, 69 insertions(+)
> 
> diff --git a/drivers/clk/ingenic/x1000-cgu.c 
> b/drivers/clk/ingenic/x1000-cgu.c
> index b2ce3fb83f54..341276e5e1ef 100644
> --- a/drivers/clk/ingenic/x1000-cgu.c
> +++ b/drivers/clk/ingenic/x1000-cgu.c
> @@ -8,6 +8,7 @@
>  #include <linux/delay.h>
>  #include <linux/io.h>
>  #include <linux/of.h>
> +#include <linux/rational.h>
> 
>  #include <dt-bindings/clock/ingenic,x1000-cgu.h>
> 
> @@ -168,6 +169,37 @@ static const struct clk_ops x1000_otg_phy_ops = {
>  	.is_enabled	= x1000_usb_phy_is_enabled,
>  };
> 
> +static void
> +x1000_i2spll_calc_m_n_od(const struct ingenic_cgu_pll_info *pll_info,
> +			 unsigned long rate, unsigned long parent_rate,
> +			 unsigned int *pm, unsigned int *pn, unsigned int *pod)
> +{
> +	const unsigned long m_max = GENMASK(pll_info->m_bits - 1, 0);
> +	const unsigned long n_max = GENMASK(pll_info->n_bits - 1, 0);
> +	unsigned long m, n;
> +
> +	rational_best_approximation(rate, parent_rate, m_max, n_max, &m, 
> &n);
> +
> +	/* n should not be less than 2*m */
> +	if (n < 2 * m)
> +		n = 2 * m;
> +
> +	*pm = m;
> +	*pn = n;
> +	*pod = 1;
> +}
> +
> +static void
> +x1000_i2spll_set_rate_hook(const struct ingenic_cgu_pll_info 
> *pll_info,
> +			   unsigned long rate, unsigned long parent_rate)
> +{
> +	/*
> +	 * For some reason, the I2S divider doesn't work properly after
> +	 * updating I2SCDR unless I2SCDR1 is read & written back.
> +	 */
> +	writel(readl(cgu->base + CGU_REG_I2SCDR1), cgu->base + 
> CGU_REG_I2SCDR1);

Not fond of the nesting here, just use a variable.

Besides... According to the documentation, bits 31 and 30 of this 
register are misconnected: writing to bit 31 will be reflected in bit 
30, and vice-versa. So this would work only if the bits 30 and 31 have 
the same value.

And worse than that, where do you actually set the register's value? 
Because bits 30/31, if cleared, will automatically compute the M/N 
values to the I2SCDR fields, overriding what the driver's .set_rate() 
callback is doing.

Either we want that, and in that case the I2S clock should be a custom 
clock (since it wouldn't need to compute or write M/N), or we don't, 
and in this case bits 30/31 of this register should be set.

> +}
> +
>  static const s8 pll_od_encoding[8] = {
>  	0x0, 0x1, -1, 0x2, -1, -1, -1, 0x3,
>  };
> @@ -319,6 +351,37 @@ static const struct ingenic_cgu_clk_info 
> x1000_cgu_clocks[] = {
>  		.gate = { CGU_REG_CLKGR, 25 },
>  	},
> 
> +	[X1000_CLK_I2SPLLMUX] = {
> +		"i2s_pll_mux", CGU_CLK_MUX,
> +		.parents = { X1000_CLK_SCLKA, X1000_CLK_MPLL, -1, -1 },

If you have only 1 bit you can only have two parents, so you can remove 
the -1s.

> +		.mux = { CGU_REG_I2SCDR, 31, 1 },
> +	},
> +
> +	[X1000_CLK_I2SPLL] = {
> +		"i2s_pll", CGU_CLK_PLL,
> +		.parents = { X1000_CLK_I2SPLLMUX, -1, -1, -1 },

.parents = { X1000_CLK_I2SPLLMUX, },

> +		.pll = {
> +			.reg = CGU_REG_I2SCDR,
> +			.rate_multiplier = 1,
> +			.m_shift = 13,
> +			.m_bits = 9,
> +			.n_shift = 0,
> +			.n_bits = 13,
> +			.calc_m_n_od = x1000_i2spll_calc_m_n_od,
> +			.set_rate_hook = x1000_i2spll_set_rate_hook,
> +		},
> +	},
> +
> +	[X1000_CLK_I2S] = {
> +		"i2s", CGU_CLK_MUX,
> +		.parents = { X1000_CLK_EXCLK, -1, -1, X1000_CLK_I2SPLL },
> +		/*
> +		 * NOTE: the mux is at bit 30; bit 29 enables the M/N divider.
> +		 * Therefore, the divider is disabled when EXCLK is selected.
> +		 */
> +		.mux = { CGU_REG_I2SCDR, 29, 2 },
> +	},
> +
>  	[X1000_CLK_LCD] = {
>  		"lcd", CGU_CLK_MUX | CGU_CLK_DIV | CGU_CLK_GATE,
>  		.parents = { X1000_CLK_SCLKA, X1000_CLK_MPLL },
> @@ -426,6 +489,12 @@ static const struct ingenic_cgu_clk_info 
> x1000_cgu_clocks[] = {
>  		.gate = { CGU_REG_CLKGR, 9 },
>  	},
> 
> +	[X1000_CLK_AIC] = {
> +		"aic", CGU_CLK_GATE,
> +		.parents = { X1000_CLK_EXCLK, -1, -1, -1 },

.parents = { X1000_CLK_EXCLK, },

Cheers,
-Paul

> +		.gate = { CGU_REG_CLKGR, 11 },
> +	},
> +
>  	[X1000_CLK_UART0] = {
>  		"uart0", CGU_CLK_GATE,
>  		.parents = { X1000_CLK_EXCLK, -1, -1, -1 },
> --
> 2.38.1
> 



  reply	other threads:[~2022-10-23 15:35 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-23 14:56 [PATCH v1 0/5] Add support for " Aidan MacDonald
2022-10-23 14:56 ` [PATCH v1 1/5] clk: ingenic: Make PLL clock "od" field optional Aidan MacDonald
2022-10-23 15:20   ` Paul Cercueil
2022-10-23 22:45     ` Aidan MacDonald
2022-10-23 14:56 ` [PATCH v1 2/5] clk: ingenic: Make PLL clock enable_bit and stable_bit optional Aidan MacDonald
2022-10-23 14:56 ` [PATCH v1 3/5] clk: ingenic: Add .set_rate_hook() for PLL clocks Aidan MacDonald
2022-10-23 14:56 ` [PATCH v1 4/5] dt-bindings: ingenic,x1000-cgu: Add audio clocks Aidan MacDonald
2022-10-23 15:22   ` Krzysztof Kozlowski
2022-10-23 14:56 ` [PATCH v1 5/5] clk: ingenic: Add X1000 " Aidan MacDonald
2022-10-23 15:34   ` Paul Cercueil [this message]
2022-10-24 16:29     ` Aidan MacDonald

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=AYP7KR.ZQW0GW51WJAG@crapouillou.net \
    --to=paul@crapouillou.net \
    --cc=aidanmacdonald.0x0@gmail.com \
    --cc=devicetree@vger.kernel.org \
    --cc=krzysztof.kozlowski+dt@linaro.org \
    --cc=linux-clk@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mips@vger.kernel.org \
    --cc=mturquette@baylibre.com \
    --cc=robh+dt@kernel.org \
    --cc=sboyd@kernel.org \
    --cc=zhouyu@wanyeetech.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®