Re: [RFC/RFT PATCH 1/4] drm/bridge: dw-hdmi: Switch to regmap for register access

[Laurent Pinchart]

Hi Neil,

Thank you for the patch.

On Tuesday 17 Jan 2017 13:31:31 Neil Armstrong wrote:
> The Synopsys Designware HDMI TX Controller does not enforce register access
> on platforms instanciating it.
> The current driver supports two different types of memory-mapped flat
> register access, but in order to support the Amlogic Meson SoCs integration,
> and provide a more generic way to handle all sorts of register mapping,
> switch the register access to use the regmap infrastructure.
> 
> In the case of the registers are not flat memory-mapped or does not conform

s/does/do/

> at the actual driver implementation, a regmap struct can be given in the

s/at the actual/to the current/ ?

> plat_data and be used at probe or bind.
> 
> Since the AHB audio driver only uses direct memory access, using regmap only
> allows the I2S audio driver to be registered.

This sounds a bit unclear to me, how about "[...], only allow the I2C audio 
driver to be registered if the device is directly memory-mapped." ?

> Signed-off-by: Neil Armstrong <narmstrong@xxxxxxxxxxxx>
> ---
>  drivers/gpu/drm/bridge/dw-hdmi.c | 105 +++++++++++++++++++-----------------
>  include/drm/bridge/dw_hdmi.h     |   1 +
>  2 files changed, 57 insertions(+), 49 deletions(-)
> 
> diff --git a/drivers/gpu/drm/bridge/dw-hdmi.c
> b/drivers/gpu/drm/bridge/dw-hdmi.c index ca9d0ce..13747fe 100644
> --- a/drivers/gpu/drm/bridge/dw-hdmi.c
> +++ b/drivers/gpu/drm/bridge/dw-hdmi.c
> @@ -20,6 +20,7 @@
>  #include <linux/mutex.h>
>  #include <linux/of_device.h>
>  #include <linux/spinlock.h>
> +#include <linux/regmap.h>

Could you please keep the headers alphabetically sorted ?

>  #include <drm/drm_of.h>
>  #include <drm/drmP.h>
> @@ -167,8 +168,7 @@ struct dw_hdmi {
>  	unsigned int audio_n;
>  	bool audio_enable;
> 
> -	void (*write)(struct dw_hdmi *hdmi, u8 val, int offset);
> -	u8 (*read)(struct dw_hdmi *hdmi, int offset);
> +	struct regmap *regm;
>  };
> 
>  #define HDMI_IH_PHY_STAT0_RX_SENSE \
> @@ -179,42 +179,23 @@ struct dw_hdmi {
>  	(HDMI_PHY_RX_SENSE0 | HDMI_PHY_RX_SENSE1 | \
>  	 HDMI_PHY_RX_SENSE2 | HDMI_PHY_RX_SENSE3)
> 
> -static void dw_hdmi_writel(struct dw_hdmi *hdmi, u8 val, int offset)
> -{
> -	writel(val, hdmi->regs + (offset << 2));
> -}
> -
> -static u8 dw_hdmi_readl(struct dw_hdmi *hdmi, int offset)
> -{
> -	return readl(hdmi->regs + (offset << 2));
> -}
> -
> -static void dw_hdmi_writeb(struct dw_hdmi *hdmi, u8 val, int offset)
> -{
> -	writeb(val, hdmi->regs + offset);
> -}
> -
> -static u8 dw_hdmi_readb(struct dw_hdmi *hdmi, int offset)
> -{
> -	return readb(hdmi->regs + offset);
> -}
> -
>  static inline void hdmi_writeb(struct dw_hdmi *hdmi, u8 val, int offset)

Not related to this patch, but the value, offset order of arguments to the 
write function has been making me cringe since the very first time I read the 
code. I wonder if modifying this would be accepted.

>  {
> -	hdmi->write(hdmi, val, offset);
> +	regmap_write(hdmi->regm, offset, val);
>  }
> 
>  static inline u8 hdmi_readb(struct dw_hdmi *hdmi, int offset)
>  {
> -	return hdmi->read(hdmi, offset);
> +	unsigned int val = 0;
> +
> +	regmap_read(hdmi->regm, offset, &val);
> +
> +	return val;
>  }
> 
>  static void hdmi_modb(struct dw_hdmi *hdmi, u8 data, u8 mask, unsigned reg)
> {
> -	u8 val = hdmi_readb(hdmi, reg) & ~mask;
> -
> -	val |= data & mask;
> -	hdmi_writeb(hdmi, val, reg);
> +	regmap_update_bits(hdmi->regm, reg, mask, data);
>  }
> 
>  static void hdmi_mask_writeb(struct dw_hdmi *hdmi, u8 data, unsigned int
> reg, @@ -1949,6 +1930,21 @@ static int dw_hdmi_detect_phy(struct dw_hdmi
> *hdmi) return 0;
>  }
> 
> +static struct regmap_config hdmi_regmap_8bit_config = {
> +	.reg_bits	= 32,
> +	.val_bits	= 8,
> +	.reg_stride	= 1,
> +	.max_register	= HDMI_I2CM_FS_SCL_LCNT_0_ADDR,
> +};
> +
> +static struct regmap_config hdmi_regmap_32bit_config = {
> +	.reg_bits	= 8,
> +	.pad_bits	= 24,
> +	.val_bits	= 32,
> +	.reg_stride	= 4,
> +	.max_register	= HDMI_I2CM_FS_SCL_LCNT_0_ADDR,
> +};

I believe you can make these const.

>  static struct dw_hdmi *
>  __dw_hdmi_probe(struct platform_device *pdev,
>  		const struct dw_hdmi_plat_data *plat_data)
> @@ -1958,7 +1954,8 @@ static int dw_hdmi_detect_phy(struct dw_hdmi *hdmi)
>  	struct platform_device_info pdevinfo;
>  	struct device_node *ddc_node;
>  	struct dw_hdmi *hdmi;
> -	struct resource *iores;
> +	struct regmap_config *reg_config = &hdmi_regmap_8bit_config;

No need to assign a value at declaration time (unless the compiler is not 
smart enough and complains, or is smarter than me and finds a problem where I 
don't).

> +	struct resource *iores = NULL;
>  	int irq;
>  	int ret;
>  	u32 val = 1;
> @@ -1982,20 +1979,21 @@ static int dw_hdmi_detect_phy(struct dw_hdmi *hdmi)
>  	mutex_init(&hdmi->audio_mutex);
>  	spin_lock_init(&hdmi->audio_lock);
> 
> -	of_property_read_u32(np, "reg-io-width", &val);
> -
> -	switch (val) {
> -	case 4:
> -		hdmi->write = dw_hdmi_writel;
> -		hdmi->read = dw_hdmi_readl;
> -		break;
> -	case 1:
> -		hdmi->write = dw_hdmi_writeb;
> -		hdmi->read = dw_hdmi_readb;
> -		break;
> -	default:
> -		dev_err(dev, "reg-io-width must be 1 or 4\n");
> -		return ERR_PTR(-EINVAL);
> +	if (plat_data->regm)
> +		hdmi->regm = plat_data->regm;

You need curly braces around this statement.

> +	else {
> +		of_property_read_u32(np, "reg-io-width", &val);
> +		switch (val) {
> +		case 4:
> +			reg_config = &hdmi_regmap_32bit_config;
> +			break;
> +		case 1:
> +			reg_config = &hdmi_regmap_8bit_config;
> +			break;
> +		default:
> +			dev_err(dev, "reg-io-width must be 1 or 4\n");
> +			return ERR_PTR(-EINVAL);
> +		}
>  	}
> 
>  	ddc_node = of_parse_phandle(np, "ddc-i2c-bus", 0);
> @@ -2011,11 +2009,20 @@ static int dw_hdmi_detect_phy(struct dw_hdmi *hdmi)
>  		dev_dbg(hdmi->dev, "no ddc property found\n");
>  	}
> 
> -	iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> -	hdmi->regs = devm_ioremap_resource(dev, iores);
> -	if (IS_ERR(hdmi->regs)) {
> -		ret = PTR_ERR(hdmi->regs);
> -		goto err_res;
> +	if (!plat_data->regm) {
> +		iores = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> +		hdmi->regs = devm_ioremap_resource(dev, iores);
> +		if (IS_ERR(hdmi->regs)) {
> +			ret = PTR_ERR(hdmi->regs);
> +			goto err_res;
> +		}
> +
> +		hdmi->regm = devm_regmap_init_mmio(dev, hdmi->regs, 
reg_config);
> +		if (IS_ERR(hdmi->regm)) {
> +			dev_err(dev, "Failed to configure regmap\n");
> +			ret = PTR_ERR(hdmi->regm);
> +			goto err_res;
> +		}
>  	}
> 
>  	hdmi->isfr_clk = devm_clk_get(hdmi->dev, "isfr");
> @@ -2123,7 +2130,7 @@ static int dw_hdmi_detect_phy(struct dw_hdmi *hdmi)
>  	config0 = hdmi_readb(hdmi, HDMI_CONFIG0_ID);
>  	config3 = hdmi_readb(hdmi, HDMI_CONFIG3_ID);
> 
> -	if (config3 & HDMI_CONFIG3_AHBAUDDMA) {
> +	if (iores && config3 & HDMI_CONFIG3_AHBAUDDMA) {

You test !plat->regm above, and iores here. How about standardizing that ? If 
you test for !plat->regm here, you won't have to initialize iores to NULL.

Apart from these small issues the patch looks good to me.

>  		struct dw_hdmi_audio_data audio;
> 
>  		audio.phys = iores->start;
> diff --git a/include/drm/bridge/dw_hdmi.h b/include/drm/bridge/dw_hdmi.h
> index 735a8ab..163842d 100644
> --- a/include/drm/bridge/dw_hdmi.h
> +++ b/include/drm/bridge/dw_hdmi.h
> @@ -60,6 +60,7 @@ struct dw_hdmi_plat_data {
>  			     unsigned long mpixelclock);
>  	enum drm_mode_status (*mode_valid)(struct drm_connector *connector,
>  					   struct drm_display_mode *mode);
> +	struct regmap *regm;
>  };
> 
>  int dw_hdmi_probe(struct platform_device *pdev,

-- 
Regards,

Laurent Pinchart