Re: [PATCH 1/2] i2c: s3c24xx: Add device tree support

[Grant Likely <grant.likely-s3s/WqlpOiPyB63q8FvJNQ@public.gmane.org>]

On Mon, Jul 18, 2011 at 06:20:40AM +0530, Thomas Abraham wrote:
> Add device tree support for Samsung's s3c24xx i2c driver. It adds
> support for parsing platform data from device tree and initialize
> all the slave devices specified as child nodes.
> 
> Signed-off-by: Thomas Abraham <thomas.abraham-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
> ---
>  .../devicetree/bindings/i2c/samsung-i2c.txt        |   43 ++++++++++
>  drivers/i2c/busses/i2c-s3c2410.c                   |   87 ++++++++++++++++----
>  2 files changed, 115 insertions(+), 15 deletions(-)
>  create mode 100644 Documentation/devicetree/bindings/i2c/samsung-i2c.txt
> 
> diff --git a/Documentation/devicetree/bindings/i2c/samsung-i2c.txt b/Documentation/devicetree/bindings/i2c/samsung-i2c.txt
> new file mode 100644
> index 0000000..73191c3
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/i2c/samsung-i2c.txt
> @@ -0,0 +1,43 @@
> +* Samsung's I2C controller
> +
> +The Samsung's I2C controller is used to interface with I2C devices.
> +
> +Required properties:
> +  - compatible: value should be either of the following.
> +      (a) "samsung, s3c2410-i2c", for i2c compatible with s3c2410 i2c.
> +      (b) "samsung, s3c2440-i2c", for i2c compatible with s3c2440 i2c.
> +
> +  - reg: physical base address of the controller and length of memory mapped
> +    region.
> +
> +  - interrupts : interrupt number to the cpu.
> +
> +  - samsung,i2c-bus-number: Specifies the i2c bus number.

Drop this property.  i2c bus numbers are dynamically assigned when
using the DT.

> +
> +  - samsung,i2c-sda-delay: Delay (in ns) applied to data line (SDA) edges.
> +
> +Optional properties:
> +  - samsung,i2c-slave-addr: Slave address in multi-master enviroment. If not
> +    specified, default value is 0x10.
> +
> +  - samsung,i2c-max-bus-freq: Desired frequency in Hz of the bus. If not
> +    specified, the default value in Hz is 100000.
> +
> +Example:
> +
> +	i2c@13870000 {
> +		compatible = "samsung,s3c2440-i2c";
> +		reg = <0x13870000 0x100>;
> +		interrupts = <345>;
> +		samsung,i2c-bus-number = <1>;
> +		samsung,i2c-slave-addr = <16>;
> +		samsung,i2c-sda-delay = <100>;
> +		samsung,i2c-max-bus-freq = <100000>;
> +		#address-cells = <1>;
> +		#size-cells = <0>;
> +
> +		wm8994@1a {
> +			compatible = "wlf,wm8994";
> +			reg = <0x1a>;
> +		};
> +	};
> diff --git a/drivers/i2c/busses/i2c-s3c2410.c b/drivers/i2c/busses/i2c-s3c2410.c
> index 6c00c10..5a84a0b 100644
> --- a/drivers/i2c/busses/i2c-s3c2410.c
> +++ b/drivers/i2c/busses/i2c-s3c2410.c
> @@ -35,6 +35,9 @@
>  #include <linux/cpufreq.h>
>  #include <linux/slab.h>
>  #include <linux/io.h>
> +#include <linux/of.h>
> +#include <linux/of_i2c.h>
> +#include <linux/of_device.h>
>  
>  #include <asm/irq.h>
>  
> @@ -83,6 +86,33 @@ struct s3c24xx_i2c {
>  #endif
>  };
>  
> +#ifdef CONFIG_OF
> +static enum s3c24xx_i2c_type samsung_i2c_types[] = {
> +	TYPE_S3C2410,
> +	TYPE_S3C2440
> +};
> +
> +static const struct of_device_id s3c24xx_i2c_match[] = {
> +	{ .compatible = "samsung,s3c2410-i2c", .data = &samsung_i2c_types[0], },
> +	{ .compatible = "samsung,s3c2440-i2c", .data = &samsung_i2c_types[1], },
> +	{},
> +};
> +MODULE_DEVICE_TABLE(of, s3c24xx_i2c_match);

The samsung_i2c_types[] array is unnecessary.  You can simply do:

static const struct of_device_id s3c24xx_i2c_match[] = {
	{ .compatible = "samsung,s3c2410-i2c", .data = (void*)TYPE_S3C2410, },
	{ .compatible = "samsung,s3c2440-i2c", .data = (void*)TYPE_S3C2440, },
	{},
};
MODULE_DEVICE_TABLE(of, s3c24xx_i2c_match);

Or, drop the .data field entirely and call of_device_is_compatible()
directly in s3c24xx_i2c_is2440().

> +
> +static const struct
> +of_device_id *s3c24xx_i2c_get_of_device_id(struct platform_device *pdev)
> +{
> +	return of_match_device(s3c24xx_i2c_match, &pdev->dev);
> +}
> +#else
> +#define s3c24xx_i2c_match NULL
> +static const struct
> +of_device_id *s3c24xx_i2c_get_of_device_id(struct platform_device *pdev)
> +{
> +	return NULL;
> +}
> +#endif
> +
>  /* default platform data removed, dev should always carry data. */
>  
>  /* s3c24xx_i2c_is2440()
> @@ -93,9 +123,15 @@ struct s3c24xx_i2c {
>  static inline int s3c24xx_i2c_is2440(struct s3c24xx_i2c *i2c)
>  {
>  	struct platform_device *pdev = to_platform_device(i2c->dev);
> +	const struct of_device_id *i2c_of_id;
>  	enum s3c24xx_i2c_type type;
>  
> -	type = platform_get_device_id(pdev)->driver_data;
> +	i2c_of_id = s3c24xx_i2c_get_of_device_id(pdev);
> +	if (platform_get_device_id(pdev))
> +		type = platform_get_device_id(pdev)->driver_data;
> +	else
> +		type = *(enum s3c24xx_i2c_type *)i2c_of_id->data;
> +
>  	return type == TYPE_S3C2440;

In fact, most of these line changes go away by putting the folowing
lines above the platform_get_device_id() line:

>  }
>  
> @@ -630,16 +666,21 @@ static int s3c24xx_i2c_clockrate(struct s3c24xx_i2c *i2c, unsigned int *got)
>  	unsigned long clkin = clk_get_rate(i2c->clk);
>  	unsigned int divs, div1;
>  	unsigned long target_frequency;
> +	unsigned int max_freq = 0;
>  	u32 iiccon;
>  	int freq;
>  
>  	i2c->clkrate = clkin;
>  	clkin /= 1000;		/* clkin now in KHz */
>  
> -	dev_dbg(i2c->dev, "pdata desired frequency %lu\n", pdata->frequency);
> -
> -	target_frequency = pdata->frequency ? pdata->frequency : 100000;
> +	if (i2c->dev->of_node)
> +		of_property_read_u32(i2c->dev->of_node,
> +				"samsung,i2c-max-bus-freq", &max_freq);

No need to test for i2c->dev->of_node.  of_property_read_u32() will
fail gracefully and not modify max_freq if of_node is NULL.

> +	else
> +		max_freq = pdata->frequency;

When using the DT, the driver needs to make sure pdata is valid before
dereferencing it.

> +	dev_dbg(i2c->dev, "desired frequency %u\n", max_freq);
>  
> +	target_frequency = max_freq ? max_freq : 100000;
>  	target_frequency /= 1000; /* Target frequency now in KHz */
>  
>  	freq = s3c24xx_i2c_calcdivisor(clkin, target_frequency, &div1, &divs);
> @@ -663,19 +704,24 @@ static int s3c24xx_i2c_clockrate(struct s3c24xx_i2c *i2c, unsigned int *got)
>  	writel(iiccon, i2c->regs + S3C2410_IICCON);
>  
>  	if (s3c24xx_i2c_is2440(i2c)) {
> -		unsigned long sda_delay;
> +		unsigned int sda_delay = 0;

Are you sure that a 32bit integer will be able to hold the sda_delay
value?  Was there a reason it was an unsigned long in the first place?

>  
> -		if (pdata->sda_delay) {
> -			sda_delay = clkin * pdata->sda_delay;
> +		if (i2c->dev->of_node)
> +			of_property_read_u32(i2c->dev->of_node,
> +				"samsung,i2c-sda-delay", &sda_delay);

sda_delay must be a u32, not an unsigned int.

> +		else
> +			sda_delay = pdata->sda_delay;
> +
> +		if (sda_delay) {
> +			sda_delay *= clkin;
>  			sda_delay = DIV_ROUND_UP(sda_delay, 1000000);
>  			sda_delay = DIV_ROUND_UP(sda_delay, 5);
>  			if (sda_delay > 3)
>  				sda_delay = 3;
>  			sda_delay |= S3C2410_IICLC_FILTER_ON;
> -		} else
> -			sda_delay = 0;
> +		}
>  
> -		dev_dbg(i2c->dev, "IICLC=%08lx\n", sda_delay);
> +		dev_dbg(i2c->dev, "IICLC=%08x\n", sda_delay);
>  		writel(sda_delay, i2c->regs + S3C2440_IICLC);
>  	}
>  
> @@ -751,7 +797,7 @@ static int s3c24xx_i2c_init(struct s3c24xx_i2c *i2c)
>  {
>  	unsigned long iicon = S3C2410_IICCON_IRQEN | S3C2410_IICCON_ACKEN;
>  	struct s3c2410_platform_i2c *pdata;
> -	unsigned int freq;
> +	unsigned int freq, slave_addr = 0x10;
>  
>  	/* get the plafrom data */
>  
> @@ -763,10 +809,14 @@ static int s3c24xx_i2c_init(struct s3c24xx_i2c *i2c)
>  		pdata->cfg_gpio(to_platform_device(i2c->dev));
>  
>  	/* write slave address */
> +	if (i2c->dev->of_node)
> +		of_property_read_u32(i2c->dev->of_node,
> +				"samsung,i2c-slave-addr", &slave_addr);
> +	else
> +		slave_addr = pdata->slave_addr;
> +	writeb(slave_addr, i2c->regs + S3C2410_IICADD);
>  
> -	writeb(pdata->slave_addr, i2c->regs + S3C2410_IICADD);
> -
> -	dev_info(i2c->dev, "slave address 0x%02x\n", pdata->slave_addr);
> +	dev_info(i2c->dev, "slave address 0x%02x\n", slave_addr);
>  
>  	writel(iicon, i2c->regs + S3C2410_IICCON);
>  
> @@ -904,7 +954,12 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
>  	 * being bus 0.
>  	 */
>  
> -	i2c->adap.nr = pdata->bus_num;
> +	if (i2c->dev->of_node)
> +		of_property_read_u32(i2c->dev->of_node,
> +				"samsung,i2c-bus-number", &i2c->adap.nr);
> +	else
> +		i2c->adap.nr = pdata->bus_num;

Overall, this patch ends up being quite invasive.  You should refactor
to a separate function just for decoding device tree data, and
using it to populate a platform_data structure.  Then the patch
doesn't need to touch most of the existing platform_data code, and in
fact makes use of it.

One warning though, be careful to *not* modify the existing
pdev->dev.platform_data pointer.  If the driver needs to access
platform_data after probe() returns, then it will need to keep a copy
of it in the driver private data structure.

g.

> +	i2c->adap.dev.of_node = pdev->dev.of_node;
>  
>  	ret = i2c_add_numbered_adapter(&i2c->adap);
>  	if (ret < 0) {
> @@ -912,6 +967,7 @@ static int s3c24xx_i2c_probe(struct platform_device *pdev)
>  		goto err_cpufreq;
>  	}
>  
> +	of_i2c_register_devices(&i2c->adap);
>  	platform_set_drvdata(pdev, i2c);
>  
>  	dev_info(&pdev->dev, "%s: S3C I2C adapter\n", dev_name(&i2c->adap.dev));
> @@ -1021,6 +1077,7 @@ static struct platform_driver s3c24xx_i2c_driver = {
>  		.owner	= THIS_MODULE,
>  		.name	= "s3c-i2c",
>  		.pm	= S3C24XX_DEV_PM_OPS,
> +		.of_match_table	= s3c24xx_i2c_match,
>  	},
>  };
>  
> -- 
> 1.7.1
> 
> _______________________________________________
> devicetree-discuss mailing list
> devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
> https://lists.ozlabs.org/listinfo/devicetree-discuss