Re: [net-next,PATCH v2] net: phy: realtek: Add support for PHY LEDs on RTL8211F

[Marek Vasut]

On 6/28/24 4:27 PM, Simon Horman wrote:
> On Tue, Jun 25, 2024 at 10:42:17PM +0200, Marek Vasut wrote:
> > Realtek RTL8211F Ethernet PHY supports 3 LED pins which are used to
> > indicate link status and activity. Add minimal LED controller driver
> > supporting the most common uses with the 'netdev' trigger.
> >
> > Signed-off-by: Marek Vasut <marex@xxxxxxx>
> > ---
> > Cc: Alexandre Torgue <alexandre.torgue@xxxxxxxxxxx>
> > Cc: Andrew Lunn <andrew@xxxxxxx>
> > Cc: Christophe Roullier <christophe.roullier@xxxxxxxxxxx>
> > Cc: David S. Miller <davem@xxxxxxxxxxxxx>
> > Cc: Eric Dumazet <edumazet@xxxxxxxxxx>
> > Cc: Heiner Kallweit <hkallweit1@xxxxxxxxx>
> > Cc: Jakub Kicinski <kuba@xxxxxxxxxx>
> > Cc: Paolo Abeni <pabeni@xxxxxxxxxx>
> > Cc: Russell King <linux@xxxxxxxxxxxxxxx>
> > Cc: kernel@xxxxxxxxxxxxxxxxxx
> > Cc: linux-kernel@xxxxxxxxxxxxxxx
> > Cc: netdev@xxxxxxxxxxxxxxx
> > ---
> > V2: - RX and TX are not differentiated, either both are set or not set,
> >        filter this in rtl8211f_led_hw_is_supported()
> > ---
> >   drivers/net/phy/realtek.c | 106 ++++++++++++++++++++++++++++++++++++++
> >   1 file changed, 106 insertions(+)
> >
> > diff --git a/drivers/net/phy/realtek.c b/drivers/net/phy/realtek.c
> > index 2174893c974f3..bed839237fb55 100644
> > --- a/drivers/net/phy/realtek.c
> > +++ b/drivers/net/phy/realtek.c
> > @@ -32,6 +32,15 @@
> >   #define RTL8211F_PHYCR2				0x19
> >   #define RTL8211F_INSR				0x1d
> >   
> > +#define RTL8211F_LEDCR				0x10
> > +#define RTL8211F_LEDCR_MODE			BIT(15)
> > +#define RTL8211F_LEDCR_ACT_TXRX			BIT(4)
> > +#define RTL8211F_LEDCR_LINK_1000		BIT(3)
> > +#define RTL8211F_LEDCR_LINK_100			BIT(1)
> > +#define RTL8211F_LEDCR_LINK_10			BIT(0)
> > +#define RTL8211F_LEDCR_MASK			GENMASK(4, 0)
> > +#define RTL8211F_LEDCR_SHIFT			5
> > +
>
> Hi Marek,
>
> FWIIW, I think that if you use FIELD_PREP and FIELD_GET then
> RTL8211F_LEDCR_SHIFT can be removed.

FIELD_PREP/FIELD_GET only works for constant mask, in this case the mask 
is not constant but shifted by SHIFT*index .

Other drivers introduce workarounds like this for exactly this issue:

drivers/clk/at91/pmc.h:#define field_prep(_mask, _val) (((_val) << 
(ffs(_mask) - 1)) & (_mask))

I don't think it is worth perpetuating that.