[Linux-stm32] [PATCH v2] leds: trigger: netdev: Check offload ability on interface up

[Marek Vasut]

The trigger_data->hw_control indicates whether the LED is controlled by HW
offload, i.e. the PHY. The trigger_data->hw_control = can_hw_control() is
currently called only from netdev_led_attr_store(), i.e. when writing any
sysfs attribute of the netdev trigger instance associated with a PHY LED.

The can_hw_control() calls validate_net_dev() which internally calls
led_cdev->hw_control_get_device(), which is phy_led_hw_control_get_device()
for PHY LEDs. The phy_led_hw_control_get_device() returns NULL if the PHY
is not attached.

At least in case of DWMAC (STM32MP, iMX8M, ...), the PHY device is attached
only when the interface is brought up and is detached again when the
interface is brought down. In case e.g. udev rules configure the netdev
LED trigger sysfs attributes before the interface is brought up, then when
the interface is brought up, the LEDs are not blinking.

This is because trigger_data->hw_control = can_hw_control() was called
when udev wrote the sysfs attribute files, before the interface was up,
so can_hw_control() resp. validate_net_dev() returned false, and the
trigger_data->hw_control = can_hw_control() was never called again to
update the trigger_data->hw_control content and let the offload take
over the LED blinking.

Call data->hw_control = can_hw_control() from netdev_trig_notify() to
update the offload capability of the LED when the UP notification arrives.
This makes the LEDs blink after the interface is brought up.

On STM32MP13xx with RTL8211F, it is enough to have the following udev rule
in place, boot the machine with cable plugged in, and the LEDs won't work
without this patch once the interface is brought up, even if they should:
"
ACTION=="add", SUBSYSTEM=="leds", KERNEL=="stmmac-0:01:green:wan", ATTR{trigger}="netdev", ATTR{link_10}="1", ATTR{link_100}="1", ATTR{link_1000}="1", ATTR{device_name}="end0"
ACTION=="add", SUBSYSTEM=="leds", KERNEL=="stmmac-0:01:yellow:wan", ATTR{trigger}="netdev", ATTR{rx}="1", ATTR{tx}="1", ATTR{device_name}="end0"
"

Reviewed-by: Andrew Lunn <andrew at lunn.ch>
Signed-off-by: Marek Vasut <marex at denx.de>
---
Cc: Alexandre Torgue <alexandre.torgue at foss.st.com>
Cc: Andrew Lunn <andrew at lunn.ch>
Cc: Christian Marangi <ansuelsmth at gmail.com>
Cc: Christophe Roullier <christophe.roullier at foss.st.com>
Cc: Daniel Golle <daniel at makrotopia.org>
Cc: Heiner Kallweit <hkallweit1 at gmail.com>
Cc: Lee Jones <lee at kernel.org>
Cc: Lukasz Majewski <lukma at denx.de>
Cc: Pavel Machek <pavel at ucw.cz>
Cc: kernel at dh-electronics.com
Cc: linux-leds at vger.kernel.org
Cc: linux-stm32 at st-md-mailman.stormreply.com
Cc: netdev at vger.kernel.org
---
V2: Add RB from Andrew
---
 drivers/leds/trigger/ledtrig-netdev.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/leds/trigger/ledtrig-netdev.c b/drivers/leds/trigger/ledtrig-netdev.c
index 4b0863db901a9..c15efe3e50780 100644
--- a/drivers/leds/trigger/ledtrig-netdev.c
+++ b/drivers/leds/trigger/ledtrig-netdev.c
@@ -605,6 +605,8 @@ static int netdev_trig_notify(struct notifier_block *nb,
 		trigger_data->net_dev = NULL;
 		break;
 	case NETDEV_UP:
+		trigger_data->hw_control = can_hw_control(trigger_data);
+		fallthrough;
 	case NETDEV_CHANGE:
 		get_device_state(trigger_data);
 		/* Refresh link_speed visibility */
-- 
2.45.2