The usual devm_regulator_get() call already handles "optional"
regulators by returning a valid dummy and printing a warning
that the dummy regulator should be described properly. This
code open coded the same behaviour, but masked any errors that
are not -EPROBE_DEFER and is quite noisy.

This change effectively unmasks and propagates regulators errors
not involving -ENODEV, downgrades the error print to warning level
if no regulator is specified and captures the probe defer message
for /sys/kernel/debug/devices_deferred.

Fixes: 2e12f536635f8 ("net: stmmac: dwmac-rk: Use standard devicetree property for phy regulator")
Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
---
 drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
index 6fdad0f10d6f..d9deba110d4b 100644
--- a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
+++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
@@ -1656,14 +1656,11 @@ static struct rk_priv_data *rk_gmac_setup(struct platform_device *pdev,
 		}
 	}
 
-	bsp_priv->regulator = devm_regulator_get_optional(dev, "phy");
+	bsp_priv->regulator = devm_regulator_get(dev, "phy");
 	if (IS_ERR(bsp_priv->regulator)) {
-		if (PTR_ERR(bsp_priv->regulator) == -EPROBE_DEFER) {
-			dev_err(dev, "phy regulator is not available yet, deferred probing\n");
-			return ERR_PTR(-EPROBE_DEFER);
-		}
-		dev_err(dev, "no regulator found\n");
-		bsp_priv->regulator = NULL;
+		ret = PTR_ERR(bsp_priv->regulator);
+		dev_err_probe(dev, ret, "failed to get phy regulator\n");
+		return ERR_PTR(ret);
 	}
 
 	ret = of_property_read_string(dev->of_node, "clock_in_out", &strings);
-- 
2.39.2

On Wed, Apr 05, 2023 at 06:10:43PM +0200, Sebastian Reichel wrote:
> The usual devm_regulator_get() call already handles "optional"
> regulators by returning a valid dummy and printing a warning
> that the dummy regulator should be described properly. This
> code open coded the same behaviour, but masked any errors that
> are not -EPROBE_DEFER and is quite noisy.
> 
> This change effectively unmasks and propagates regulators errors
> not involving -ENODEV, downgrades the error print to warning level
> if no regulator is specified and captures the probe defer message
> for /sys/kernel/debug/devices_deferred.
> 
> Fixes: 2e12f536635f8 ("net: stmmac: dwmac-rk: Use standard devicetree property for phy regulator")
> Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
> ---
>  drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 11 ++++-------
>  1 file changed, 4 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> index 6fdad0f10d6f..d9deba110d4b 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> @@ -1656,14 +1656,11 @@ static struct rk_priv_data *rk_gmac_setup(struct platform_device *pdev,
>  		}
>  	}
>  
> -	bsp_priv->regulator = devm_regulator_get_optional(dev, "phy");
> +	bsp_priv->regulator = devm_regulator_get(dev, "phy");
>  	if (IS_ERR(bsp_priv->regulator)) {
> -		if (PTR_ERR(bsp_priv->regulator) == -EPROBE_DEFER) {
> -			dev_err(dev, "phy regulator is not available yet, deferred probing\n");
> -			return ERR_PTR(-EPROBE_DEFER);
> -		}
> -		dev_err(dev, "no regulator found\n");
> -		bsp_priv->regulator = NULL;

Does phy_power_on() need to be updated for this change?
F.e. Does the bsp_priv->regulator == NULL still make sense?

> +		ret = PTR_ERR(bsp_priv->regulator);
> +		dev_err_probe(dev, ret, "failed to get phy regulator\n");
> +		return ERR_PTR(ret);
>  	}
>  
>  	ret = of_property_read_string(dev->of_node, "clock_in_out", &strings);
> -- 
> 2.39.2
>

Hi Simon,

On Wed, Apr 05, 2023 at 07:43:15PM +0200, Simon Horman wrote:
> On Wed, Apr 05, 2023 at 06:10:43PM +0200, Sebastian Reichel wrote:
> > The usual devm_regulator_get() call already handles "optional"
> > regulators by returning a valid dummy and printing a warning
> > that the dummy regulator should be described properly. This
> > code open coded the same behaviour, but masked any errors that
> > are not -EPROBE_DEFER and is quite noisy.
> > 
> > This change effectively unmasks and propagates regulators errors
> > not involving -ENODEV, downgrades the error print to warning level
> > if no regulator is specified and captures the probe defer message
> > for /sys/kernel/debug/devices_deferred.
> > 
> > Fixes: 2e12f536635f8 ("net: stmmac: dwmac-rk: Use standard devicetree property for phy regulator")
> > Signed-off-by: Sebastian Reichel <sebastian.reichel@collabora.com>
> > ---
> >  drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c | 11 ++++-------
> >  1 file changed, 4 insertions(+), 7 deletions(-)
> > 
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> > index 6fdad0f10d6f..d9deba110d4b 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/dwmac-rk.c
> > @@ -1656,14 +1656,11 @@ static struct rk_priv_data *rk_gmac_setup(struct platform_device *pdev,
> >  		}
> >  	}
> >  
> > -	bsp_priv->regulator = devm_regulator_get_optional(dev, "phy");
> > +	bsp_priv->regulator = devm_regulator_get(dev, "phy");
> >  	if (IS_ERR(bsp_priv->regulator)) {
> > -		if (PTR_ERR(bsp_priv->regulator) == -EPROBE_DEFER) {
> > -			dev_err(dev, "phy regulator is not available yet, deferred probing\n");
> > -			return ERR_PTR(-EPROBE_DEFER);
> > -		}
> > -		dev_err(dev, "no regulator found\n");
> > -		bsp_priv->regulator = NULL;
> 
> Does phy_power_on() need to be updated for this change?
> F.e. Does the bsp_priv->regulator == NULL still make sense?

Yes, it can be removed (but does not hurt). The regulator API
returns NULL for devm_regulator_get when CONFIG_REGULATOR is
not enabled. But regulator_enable/regulator_disable are just
'return 0;' stubs for that case anyways.

-- Sebastian

> > +		ret = PTR_ERR(bsp_priv->regulator);
> > +		dev_err_probe(dev, ret, "failed to get phy regulator\n");
> > +		return ERR_PTR(ret);
> >  	}
> >  
> >  	ret = of_property_read_string(dev->of_node, "clock_in_out", &strings);
> > -- 
> > 2.39.2
> > 
> 
> -- 
> To unsubscribe, send mail to kernel-unsubscribe@lists.collabora.co.uk.