Re: [PATCH 3/6] drm/sun4i: sun8i-hdmi-phy: Used device-managed clocks/resets

[Maxime Ripard]

Hi Samuel,

Sorry for the (very) late answer

On Tue, Apr 12, 2022 at 06:34:40PM -0500, Samuel Holland wrote:
> On 4/12/22 8:23 AM, Maxime Ripard wrote:
> > Hi,
> > 
> > On Mon, Apr 11, 2022 at 11:35:08PM -0500, Samuel Holland wrote:
> >> Now that the HDMI PHY is using a platform driver, it can use device-
> >> managed resources. Use these, as well as the dev_err_probe helper, to
> >> simplify the probe function and get rid of the remove function.
> >>
> >> Signed-off-by: Samuel Holland <samuel@xxxxxxxxxxxx>
> >> ---
> >>
> >>  drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c | 100 ++++++++-----------------
> >>  1 file changed, 30 insertions(+), 70 deletions(-)
> >>
> >> diff --git a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
> >> index 1effa30bfe62..1351e633d485 100644
> >> --- a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
> >> +++ b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
> >> @@ -673,10 +673,8 @@ int sun8i_hdmi_phy_get(struct sun8i_dw_hdmi *hdmi, struct device_node *node)
> >>  static int sun8i_hdmi_phy_probe(struct platform_device *pdev)
> >>  {
> >>  	struct device *dev = &pdev->dev;
> >> -	struct device_node *node = dev->of_node;
> >>  	struct sun8i_hdmi_phy *phy;
> >>  	void __iomem *regs;
> >> -	int ret;
> >>  
> >>  	phy = devm_kzalloc(dev, sizeof(*phy), GFP_KERNEL);
> >>  	if (!phy)
> >> @@ -686,88 +684,50 @@ static int sun8i_hdmi_phy_probe(struct platform_device *pdev)
> >>  	phy->dev = dev;
> >>  
> >>  	regs = devm_platform_ioremap_resource(pdev, 0);
> >> -	if (IS_ERR(regs)) {
> >> -		dev_err(dev, "Couldn't map the HDMI PHY registers\n");
> >> -		return PTR_ERR(regs);
> >> -	}
> >> +	if (IS_ERR(regs))
> >> +		return dev_err_probe(dev, PTR_ERR(regs),
> >> +				     "Couldn't map the HDMI PHY registers\n");
> >>  
> >>  	phy->regs = devm_regmap_init_mmio(dev, regs,
> >>  					  &sun8i_hdmi_phy_regmap_config);
> >> -	if (IS_ERR(phy->regs)) {
> >> -		dev_err(dev, "Couldn't create the HDMI PHY regmap\n");
> >> -		return PTR_ERR(phy->regs);
> >> -	}
> >> +	if (IS_ERR(phy->regs))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->regs),
> >> +				     "Couldn't create the HDMI PHY regmap\n");
> >>  
> >> -	phy->clk_bus = of_clk_get_by_name(node, "bus");
> >> -	if (IS_ERR(phy->clk_bus)) {
> >> -		dev_err(dev, "Could not get bus clock\n");
> >> -		return PTR_ERR(phy->clk_bus);
> >> -	}
> >> -
> >> -	phy->clk_mod = of_clk_get_by_name(node, "mod");
> >> -	if (IS_ERR(phy->clk_mod)) {
> >> -		dev_err(dev, "Could not get mod clock\n");
> >> -		ret = PTR_ERR(phy->clk_mod);
> >> -		goto err_put_clk_bus;
> >> -	}
> >> +	phy->clk_bus = devm_clk_get(dev, "bus");
> >> +	if (IS_ERR(phy->clk_bus))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->clk_bus),
> >> +				     "Could not get bus clock\n");
> >>  
> >> -	if (phy->variant->has_phy_clk) {
> >> -		phy->clk_pll0 = of_clk_get_by_name(node, "pll-0");
> >> -		if (IS_ERR(phy->clk_pll0)) {
> >> -			dev_err(dev, "Could not get pll-0 clock\n");
> >> -			ret = PTR_ERR(phy->clk_pll0);
> >> -			goto err_put_clk_mod;
> >> -		}
> >> -
> >> -		if (phy->variant->has_second_pll) {
> >> -			phy->clk_pll1 = of_clk_get_by_name(node, "pll-1");
> >> -			if (IS_ERR(phy->clk_pll1)) {
> >> -				dev_err(dev, "Could not get pll-1 clock\n");
> >> -				ret = PTR_ERR(phy->clk_pll1);
> >> -				goto err_put_clk_pll0;
> >> -			}
> >> -		}
> >> -	}
> >> +	phy->clk_mod = devm_clk_get(dev, "mod");
> >> +	if (IS_ERR(phy->clk_mod))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->clk_mod),
> >> +				     "Could not get mod clock\n");
> >>  
> >> -	phy->rst_phy = of_reset_control_get_shared(node, "phy");
> >> -	if (IS_ERR(phy->rst_phy)) {
> >> -		dev_err(dev, "Could not get phy reset control\n");
> >> -		ret = PTR_ERR(phy->rst_phy);
> >> -		goto err_put_clk_pll1;
> >> -	}
> >> +	if (phy->variant->has_phy_clk)
> >> +		phy->clk_pll0 = devm_clk_get(dev, "pll-0");
> >> +	if (IS_ERR(phy->clk_pll0))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->clk_pll0),
> >> +				     "Could not get pll-0 clock\n");
> >> +
> >> +	if (phy->variant->has_second_pll)
> >> +		phy->clk_pll1 = devm_clk_get(dev, "pll-1");
> >> +	if (IS_ERR(phy->clk_pll1))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->clk_pll1),
> >> +				     "Could not get pll-1 clock\n");
> >> +
> >> +	phy->rst_phy = devm_reset_control_get_shared(dev, "phy");
> >> +	if (IS_ERR(phy->rst_phy))
> >> +		return dev_err_probe(dev, PTR_ERR(phy->rst_phy),
> >> +				     "Could not get phy reset control\n");
> > 
> > I find the old construct clearer with the imbricated blocks.
> 
> I'm not sure what you mean here. Are you suggesting braces around the
> dev_err_probe statements? Or do you want me to put the error handling for
> variant-specific resources inside the variant checks? Please clarify.

I meant that we went, for example, from:

if (phy->variant->has_phy_clk) {
	phy->clk_pll0 = devm_clk_get(dev, "pll-0");
	if (IS_ERR(phy->clk_pll0)) {
		...
	}
}

to

if (phy->variant->has_phy_clk)
	phy->clk_pll0 = devm_clk_get(dev, "pll-0");
if (IS_ERR(phy->clk_pll0)) {
		...
}

Which relies on the fact that phy->clk_pll0 is initialized to !IS_ERR(...), which we never
explicitly enforced. I find the first and original code clearer for that aspect (since we only use
that value if it was set), and less fragile.

Maxime