[PATCH net-next 0/2] net: phy: use generic PHY ability readers if callback get_features isn't set

[PATCH net-next 0/2] net: phy: use generic PHY ability readers if callback get_features isn't set

[Heiner Kallweit]

Meanwhile we have generic functions for reading the abilities of
Clause 22 / 45 PHY's. This allows to use them as fallback in case
callback get_features isn't set. Benefit is the reduction of
boilerplate code in PHY drivers.

Heiner Kallweit (2):
  net: phy: allow a PHY driver to define neither features nor
    get_features
  net: phy: realtek: remove setting callback get_features and use phylib
    fallback

 drivers/net/phy/phy_device.c | 17 +++++++++++------
 drivers/net/phy/realtek.c    | 10 ----------
 2 files changed, 11 insertions(+), 16 deletions(-)

-- 
2.21.0

[PATCH net-next 1/2] net: phy: allow a PHY driver to define neither features nor get_features

[Heiner Kallweit]

Meanwhile we have generic functions for reading the abilities of
Clause 22 / 45 PHY's. This allows to use them as fallback in case
callback get_features isn't set. Benefit is the reduction of
boilerplate code in PHY drivers.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
---
 drivers/net/phy/phy_device.c | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/drivers/net/phy/phy_device.c b/drivers/net/phy/phy_device.c
index 72fc714c9..3a3166a7d 100644
--- a/drivers/net/phy/phy_device.c
+++ b/drivers/net/phy/phy_device.c
@@ -2143,12 +2143,17 @@ static int phy_probe(struct device *dev)
 	 */
 	if (phydrv->features) {
 		linkmode_copy(phydev->supported, phydrv->features);
-	} else {
+	} else if (phydrv->get_features) {
 		err = phydrv->get_features(phydev);
-		if (err)
-			goto out;
+	} else if (phydev->is_c45) {
+		err = genphy_c45_pma_read_abilities(phydev);
+	} else {
+		err = genphy_read_abilities(phydev);
 	}
 
+	if (err)
+		goto out;
+
 	of_set_phy_supported(phydev);
 	linkmode_copy(phydev->advertising, phydev->supported);
 
@@ -2218,9 +2223,9 @@ int phy_driver_register(struct phy_driver *new_driver, struct module *owner)
 	/* Either the features are hard coded, or dynamically
 	 * determine. It cannot be both or neither
 	 */
-	if (WARN_ON((!new_driver->features && !new_driver->get_features) ||
-		    (new_driver->features && new_driver->get_features))) {
-		pr_err("%s: Driver features are missing\n", new_driver->name);
+	if (WARN_ON(new_driver->features && new_driver->get_features)) {
+		pr_err("%s: features and get_features must not both be set\n",
+		       new_driver->name);
 		return -EINVAL;
 	}
 
-- 
2.21.0

[PATCH net-next 2/2] net: phy: realtek: remove setting callback get_features and use phylib fallback

[Heiner Kallweit]

Now that phylib uses genphy_read_abilities() as fallback, we don't have
to set callback get_features any longer.

Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
---
 drivers/net/phy/realtek.c | 10 ----------
 1 file changed, 10 deletions(-)

diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
index 5ecbd41ed..d6a10f323 100644
--- a/drivers/net/phy/realtek.c
+++ b/drivers/net/phy/realtek.c
@@ -199,11 +199,9 @@ static struct phy_driver realtek_drvs[] = {
 	{
 		PHY_ID_MATCH_EXACT(0x00008201),
 		.name           = "RTL8201CP Ethernet",
-		.get_features	= genphy_read_abilities,
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc816),
 		.name		= "RTL8201F Fast Ethernet",
-		.get_features	= genphy_read_abilities,
 		.ack_interrupt	= &rtl8201_ack_interrupt,
 		.config_intr	= &rtl8201_config_intr,
 		.suspend	= genphy_suspend,
@@ -213,14 +211,12 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc910),
 		.name		= "RTL8211 Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.config_aneg	= rtl8211_config_aneg,
 		.read_mmd	= &genphy_read_mmd_unsupported,
 		.write_mmd	= &genphy_write_mmd_unsupported,
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc912),
 		.name		= "RTL8211B Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.ack_interrupt	= &rtl821x_ack_interrupt,
 		.config_intr	= &rtl8211b_config_intr,
 		.read_mmd	= &genphy_read_mmd_unsupported,
@@ -230,14 +226,12 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc913),
 		.name		= "RTL8211C Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.config_init	= rtl8211c_config_init,
 		.read_mmd	= &genphy_read_mmd_unsupported,
 		.write_mmd	= &genphy_write_mmd_unsupported,
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc914),
 		.name		= "RTL8211DN Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.ack_interrupt	= rtl821x_ack_interrupt,
 		.config_intr	= rtl8211e_config_intr,
 		.suspend	= genphy_suspend,
@@ -245,7 +239,6 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc915),
 		.name		= "RTL8211E Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.ack_interrupt	= &rtl821x_ack_interrupt,
 		.config_intr	= &rtl8211e_config_intr,
 		.suspend	= genphy_suspend,
@@ -253,7 +246,6 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc916),
 		.name		= "RTL8211F Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.config_init	= &rtl8211f_config_init,
 		.ack_interrupt	= &rtl8211f_ack_interrupt,
 		.config_intr	= &rtl8211f_config_intr,
@@ -264,7 +256,6 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc800),
 		.name		= "Generic Realtek PHY",
-		.get_features	= genphy_read_abilities,
 		.suspend	= genphy_suspend,
 		.resume		= genphy_resume,
 		.read_page	= rtl821x_read_page,
@@ -272,7 +263,6 @@ static struct phy_driver realtek_drvs[] = {
 	}, {
 		PHY_ID_MATCH_EXACT(0x001cc961),
 		.name		= "RTL8366RB Gigabit Ethernet",
-		.get_features	= genphy_read_abilities,
 		.config_init	= &rtl8366rb_config_init,
 		/* These interrupts are handled by the irq controller
 		 * embedded inside the RTL8366RB, they get unmasked when the
-- 
2.21.0

Re: [PATCH net-next 1/2] net: phy: allow a PHY driver to define neither features nor get_features

[Andrew Lunn]

> @@ -2218,9 +2223,9 @@ int phy_driver_register(struct phy_driver *new_driver, struct module *owner)
>  	/* Either the features are hard coded, or dynamically
>  	 * determine. It cannot be both or neither
>  	 */

Hi Heiner

The comment needs updating to match the code.

> -	if (WARN_ON((!new_driver->features && !new_driver->get_features) ||
> -		    (new_driver->features && new_driver->get_features))) {
> -		pr_err("%s: Driver features are missing\n", new_driver->name);
> +	if (WARN_ON(new_driver->features && new_driver->get_features)) {
> +		pr_err("%s: features and get_features must not both be set\n",
> +		       new_driver->name);
>  		return -EINVAL;

	Andrew

Re: [PATCH net-next 2/2] net: phy: realtek: remove setting callback get_features and use phylib fallback

[Andrew Lunn]

On Wed, Apr 03, 2019 at 08:20:09PM +0200, Heiner Kallweit wrote:
> Now that phylib uses genphy_read_abilities() as fallback, we don't have
> to set callback get_features any longer.

Hi Heiner

I wonder how many PHY drivers we will break if we just remove
.features everywhere?

Maybe we should create patches for the Marvell driver and the Broadcom
drivers. I and Florian can test them.

	 Andrew

Re: [PATCH net-next 1/2] net: phy: allow a PHY driver to define neither features nor get_features

[Heiner Kallweit]

On 03.04.2019 22:46, Andrew Lunn wrote:
>> @@ -2218,9 +2223,9 @@ int phy_driver_register(struct phy_driver *new_driver, struct module *owner)
>>  	/* Either the features are hard coded, or dynamically
>>  	 * determine. It cannot be both or neither
>>  	 */
> 
> Hi Heiner
> 
> The comment needs updating to match the code.
> 
Indeed, I have to fix this.

>> -	if (WARN_ON((!new_driver->features && !new_driver->get_features) ||
>> -		    (new_driver->features && new_driver->get_features))) {
>> -		pr_err("%s: Driver features are missing\n", new_driver->name);
>> +	if (WARN_ON(new_driver->features && new_driver->get_features)) {
>> +		pr_err("%s: features and get_features must not both be set\n",
>> +		       new_driver->name);
>>  		return -EINVAL;
> 
> 	Andrew
> 

Re: [PATCH net-next 2/2] net: phy: realtek: remove setting callback get_features and use phylib fallback

[Heiner Kallweit]

On 03.04.2019 22:55, Andrew Lunn wrote:
> On Wed, Apr 03, 2019 at 08:20:09PM +0200, Heiner Kallweit wrote:
>> Now that phylib uses genphy_read_abilities() as fallback, we don't have
>> to set callback get_features any longer.
> 
> Hi Heiner
> 
> I wonder how many PHY drivers we will break if we just remove
> .features everywhere?
> 
At least I have seen no C22 PHY yet that couldn't be handled by
genphy_read_abilities. Just for non-TP modes there might be an issue.
genphy_read_abilities only sets TP and MII, if a driver checks for
e.g. ETHTOOL_LINK_MODE_FIBRE_BIT we may have some work.

> Maybe we should create patches for the Marvell driver and the Broadcom
> drivers. I and Florian can test them.
> 
> 	 Andrew
> 
Heiner