Re: [PATCH v10 05/17] clk: tegra: Introduce ability for SoC-specific reset control callbacks

From: Thierry Reding <thierry.reding@gmail.com>
To: Mikko Perttunen <mikko.perttunen@kapsi.fi>
Cc: swarren@wwwdotorg.org, gnurou@gmail.com, pdeschrijver@nvidia.com,
	mturquette@linaro.org, rjw@rjwysocki.net,
	viresh.kumar@linaro.org, pwalmsley@nvidia.com, vinceh@nvidia.com,
	pgaikwad@nvidia.com, linux-kernel@vger.kernel.org,
	linux-pm@vger.kernel.org, linux-tegra@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org, tuomas.tynkkynen@iki.fi
Subject: Re: [PATCH v10 05/17] clk: tegra: Introduce ability for SoC-specific reset control callbacks
Date: Tue, 19 May 2015 16:59:23 +0200	[thread overview]
Message-ID: <20150519145922.GA28018@ulmo> (raw)
In-Reply-To: <1432035567-19008-1-git-send-email-mikko.perttunen@kapsi.fi>

[-- Attachment #1: Type: text/plain, Size: 3714 bytes --]

On Tue, May 19, 2015 at 02:39:27PM +0300, Mikko Perttunen wrote:
> This patch allows SoC-specific CAR initialization routines to register
> their own reset_assert and reset_deassert callbacks with the common Tegra
> CAR code. If defined, the common code will call these callbacks when a
> reset control with number >= num_periph_banks * 32 is attempted to be asserted 
> or deasserted respectively. Numbers greater than or equal to num_periph_banks * 32
> are used to avoid clashes with low numbers that are automatically mapped to
> standard CAR reset lines.
> 
> Each SoC with these special resets should specify the defined reset control
> numbers in a device tree header file.

This is looking pretty good, but I think we can simplify a wee bit
more...

> Signed-off-by: Mikko Perttunen <mikko.perttunen@kapsi.fi>
> Acked-by: Michael Turquette <mturquette@linaro.org>
> ---
>  drivers/clk/tegra/clk.c | 39 +++++++++++++++++++++++++++++++--------
>  drivers/clk/tegra/clk.h |  3 +++
>  2 files changed, 34 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/clk/tegra/clk.c b/drivers/clk/tegra/clk.c
> index 41cd87c..c093ed9 100644
> --- a/drivers/clk/tegra/clk.c
> +++ b/drivers/clk/tegra/clk.c
> @@ -49,7 +49,6 @@
>  #define RST_DEVICES_L			0x004
>  #define RST_DEVICES_H			0x008
>  #define RST_DEVICES_U			0x00C
> -#define RST_DFLL_DVCO			0x2F4
>  #define RST_DEVICES_V			0x358
>  #define RST_DEVICES_W			0x35C
>  #define RST_DEVICES_X			0x28C
> @@ -79,6 +78,11 @@ static struct clk **clks;
>  static int clk_num;
>  static struct clk_onecell_data clk_data;
>  
> +/* Handlers for SoC-specific reset lines */
> +static int (*special_reset_assert)(unsigned long);
> +static int (*special_reset_deassert)(unsigned long);
> +static int special_reset_num;

I think we can get rid of this if we...

> +
>  static struct tegra_clk_periph_regs periph_regs[] = {
>  	[0] = {
>  		.enb_reg = CLK_OUT_ENB_L,
> @@ -152,19 +156,29 @@ static int tegra_clk_rst_assert(struct reset_controller_dev *rcdev,
>  	 */
>  	tegra_read_chipid();
>  
> -	writel_relaxed(BIT(id % 32),
> -			clk_base + periph_regs[id / 32].rst_set_reg);
> +	if (id < periph_banks * 32) {
> +		writel_relaxed(BIT(id % 32),
> +			       clk_base + periph_regs[id / 32].rst_set_reg);
> +		return 0;
> +	} else if (id < periph_banks * 32 + special_reset_num) {
> +		return special_reset_assert(id);
> +	}

... pass id - periph_banks * 32 into special_reset_assert(). Oh, but
then...

> @@ -286,10 +300,19 @@ void __init tegra_add_of_provider(struct device_node *np)
>  	of_clk_add_provider(np, of_clk_src_onecell_get, &clk_data);
>  
>  	rst_ctlr.of_node = np;
> -	rst_ctlr.nr_resets = periph_banks * 32;
> +	rst_ctlr.nr_resets = periph_banks * 32 + special_reset_num;

... this is no longer going to work. We could keep special_reset_num,
though obviously it should be an unsigned int and renamed to
num_special_reset, yet still pass the relative ID into the SoC-specific
callbacks, after all that's what they care about and each implementation
would have to subtract periph_banks * 32 anyway.

>  	reset_controller_register(&rst_ctlr);
>  }
>  
> +void __init tegra_init_special_resets(int num,
> +				      int (*assert)(unsigned long),
> +				      int (*deassert)(unsigned long))
> +{
> +	special_reset_num = num;
> +	special_reset_assert = assert;
> +	special_reset_deassert = deassert;
> +}
> +

I think we could possibly improve this interface somewhat by turning it
upside-down, that is, make SoC-specific drivers call this in a helper
fashion. But this is good enough for now, I can always take a stab at
refactoring if I get bored.

Thierry

[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]