[PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Jan Petrous (OSS)]

The helper rgmii_clock() implemented Russel's hint during stmmac
glue driver review:

---
We seem to have multiple cases of very similar logic in lots of stmmac
platform drivers, and I think it's about time we said no more to this.
So, what I think we should do is as follows:

add the following helper - either in stmmac, or more generically
(phylib? - in which case its name will need changing.)

static long stmmac_get_rgmii_clock(int speed)
{
	switch (speed) {
	case SPEED_10:
		return 2500000;

	case SPEED_100:
		return 25000000;

	case SPEED_1000:
		return 125000000;

	default:
		return -ENVAL;
	}
}

Then, this can become:

	long tx_clk_rate;

	...

	tx_clk_rate = stmmac_get_rgmii_clock(speed);
	if (tx_clk_rate < 0) {
		dev_err(gmac->dev, "Unsupported/Invalid speed: %d\n", speed);
		return;
	}

	ret = clk_set_rate(gmac->tx_clk, tx_clk_rate);
---

Signed-off-by: Jan Petrous (OSS) <jan.petrous@oss.nxp.com>
---
 include/linux/phy.h | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/include/linux/phy.h b/include/linux/phy.h
index 6b7d40d49129..bb797364d91c 100644
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -298,6 +298,27 @@ static inline const char *phy_modes(phy_interface_t interface)
 	}
 }
 
+/**
+ * rgmi_clock - map link speed to the clock rate
+ * @speed: link speed value
+ *
+ * Description: maps RGMII supported link speeds
+ * into the clock rates.
+ */
+static inline long rgmii_clock(int speed)
+{
+	switch (speed) {
+	case SPEED_10:
+		return 2500000;
+	case SPEED_100:
+		return 25000000;
+	case SPEED_1000:
+		return 125000000;
+	default:
+		return -EINVAL;
+	}
+}
+
 #define PHY_INIT_TIMEOUT	100000
 #define PHY_FORCE_TIMEOUT	10
 
-- 
2.46.0

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Simon Horman]

On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> The helper rgmii_clock() implemented Russel's hint during stmmac
> glue driver review:
> 
> ---
> We seem to have multiple cases of very similar logic in lots of stmmac
> platform drivers, and I think it's about time we said no more to this.
> So, what I think we should do is as follows:
> 
> add the following helper - either in stmmac, or more generically
> (phylib? - in which case its name will need changing.)
> 
> static long stmmac_get_rgmii_clock(int speed)
> {
> 	switch (speed) {
> 	case SPEED_10:
> 		return 2500000;
> 
> 	case SPEED_100:
> 		return 25000000;
> 
> 	case SPEED_1000:
> 		return 125000000;
> 
> 	default:
> 		return -ENVAL;
> 	}
> }
> 
> Then, this can become:
> 
> 	long tx_clk_rate;
> 
> 	...
> 
> 	tx_clk_rate = stmmac_get_rgmii_clock(speed);
> 	if (tx_clk_rate < 0) {
> 		dev_err(gmac->dev, "Unsupported/Invalid speed: %d\n", speed);
> 		return;
> 	}
> 
> 	ret = clk_set_rate(gmac->tx_clk, tx_clk_rate);
> ---
> 
> Signed-off-by: Jan Petrous (OSS) <jan.petrous@oss.nxp.com>
> ---
>  include/linux/phy.h | 21 +++++++++++++++++++++
>  1 file changed, 21 insertions(+)
> 
> diff --git a/include/linux/phy.h b/include/linux/phy.h
> index 6b7d40d49129..bb797364d91c 100644
> --- a/include/linux/phy.h
> +++ b/include/linux/phy.h
> @@ -298,6 +298,27 @@ static inline const char *phy_modes(phy_interface_t interface)
>  	}
>  }
>  
> +/**
> + * rgmi_clock - map link speed to the clock rate

nit: rgmii_clock

     Flagged by ./scripts/kernel-doc -none

> + * @speed: link speed value
> + *
> + * Description: maps RGMII supported link speeds
> + * into the clock rates.
> + */
> +static inline long rgmii_clock(int speed)
> +{
> +	switch (speed) {
> +	case SPEED_10:
> +		return 2500000;
> +	case SPEED_100:
> +		return 25000000;
> +	case SPEED_1000:
> +		return 125000000;
> +	default:
> +		return -EINVAL;
> +	}
> +}
> +
>  #define PHY_INIT_TIMEOUT	100000
>  #define PHY_FORCE_TIMEOUT	10
>  
> -- 
> 2.46.0
> 
> 

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Andrew Lunn]

On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> The helper rgmii_clock() implemented Russel's hint during stmmac
> glue driver review:
> 
> ---
> We seem to have multiple cases of very similar logic in lots of stmmac
> platform drivers, and I think it's about time we said no more to this.
> So, what I think we should do is as follows:
> 
> add the following helper - either in stmmac, or more generically
> (phylib? - in which case its name will need changing.)
> 
> static long stmmac_get_rgmii_clock(int speed)
> {
> 	switch (speed) {
> 	case SPEED_10:
> 		return 2500000;
> 
> 	case SPEED_100:
> 		return 25000000;
> 
> 	case SPEED_1000:
> 		return 125000000;
> 
> 	default:
> 		return -ENVAL;
> 	}
> }
> 
> Then, this can become:
> 
> 	long tx_clk_rate;
> 
> 	...
> 
> 	tx_clk_rate = stmmac_get_rgmii_clock(speed);
> 	if (tx_clk_rate < 0) {
> 		dev_err(gmac->dev, "Unsupported/Invalid speed: %d\n", speed);
> 		return;
> 	}
> 
> 	ret = clk_set_rate(gmac->tx_clk, tx_clk_rate);
> ---
> 
> Signed-off-by: Jan Petrous (OSS) <jan.petrous@oss.nxp.com>

This Signed-off-by: needs to be above the first ---, otherwise it gets
discard.

When you repost, please do try to get threading correct.

    Andrew

---
pw-bot: cr

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Andrew Lunn]

On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> The helper rgmii_clock() implemented Russel's hint during stmmac
> glue driver review:
> 
> ---
> We seem to have multiple cases of very similar logic in lots of stmmac
> platform drivers, and I think it's about time we said no more to this.
> So, what I think we should do is as follows:
> 
> add the following helper - either in stmmac, or more generically
> (phylib? - in which case its name will need changing.)

As Russell pointed out, this code appears a few times in the stmmac
driver. Please include patches changing the other instances to use
this helper.

It also looks like macb, and maybe xgene_enet_hw.c could use it as
well.

	Andrew

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Jan Petrous]

On Mon, Aug 19, 2024 at 05:49:58PM +0200, Andrew Lunn wrote:
> On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> > The helper rgmii_clock() implemented Russel's hint during stmmac
> > glue driver review:
> > 
> > ---
> > We seem to have multiple cases of very similar logic in lots of stmmac
> > platform drivers, and I think it's about time we said no more to this.
> > So, what I think we should do is as follows:
> > 
> > add the following helper - either in stmmac, or more generically
> > (phylib? - in which case its name will need changing.)
> > 
> > static long stmmac_get_rgmii_clock(int speed)
> > {
> > 	switch (speed) {
> > 	case SPEED_10:
> > 		return 2500000;
> > 
> > 	case SPEED_100:
> > 		return 25000000;
> > 
> > 	case SPEED_1000:
> > 		return 125000000;
> > 
> > 	default:
> > 		return -ENVAL;
> > 	}
> > }
> > 
> > Then, this can become:
> > 
> > 	long tx_clk_rate;
> > 
> > 	...
> > 
> > 	tx_clk_rate = stmmac_get_rgmii_clock(speed);
> > 	if (tx_clk_rate < 0) {
> > 		dev_err(gmac->dev, "Unsupported/Invalid speed: %d\n", speed);
> > 		return;
> > 	}
> > 
> > 	ret = clk_set_rate(gmac->tx_clk, tx_clk_rate);
> > ---
> > 
> > Signed-off-by: Jan Petrous (OSS) <jan.petrous@oss.nxp.com>
> 
> This Signed-off-by: needs to be above the first ---, otherwise it gets
> discard.
> 

I see, it is used as delimiter, my fault.
I will change formating for v3.

> When you repost, please do try to get threading correct.
> 

Yeh, I already got the same feedback from Krzysztof.
I'm switching to b4/lei/mutt for v3 what I hope fixed
the threading issue.

BR.
/Jan

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Jan Petrous]

On Mon, Aug 19, 2024 at 03:15:41PM +0100, Simon Horman wrote:
> On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> > The helper rgmii_clock() implemented Russel's hint during stmmac
> > glue driver review:
> > 
> > ---
> > We seem to have multiple cases of very similar logic in lots of stmmac
> > platform drivers, and I think it's about time we said no more to this.
> > So, what I think we should do is as follows:
> > 
> > add the following helper - either in stmmac, or more generically
> > (phylib? - in which case its name will need changing.)
> > 
> > static long stmmac_get_rgmii_clock(int speed)
> > {
> > 	switch (speed) {
> > 	case SPEED_10:
> > 		return 2500000;
> > 
> > 	case SPEED_100:
> > 		return 25000000;
> > 
> > 	case SPEED_1000:
> > 		return 125000000;
> > 
> > 	default:
> > 		return -ENVAL;
> > 	}
> > }
> > 
> > Then, this can become:
> > 
> > 	long tx_clk_rate;
> > 
> > 	...
> > 
> > 	tx_clk_rate = stmmac_get_rgmii_clock(speed);
> > 	if (tx_clk_rate < 0) {
> > 		dev_err(gmac->dev, "Unsupported/Invalid speed: %d\n", speed);
> > 		return;
> > 	}
> > 
> > 	ret = clk_set_rate(gmac->tx_clk, tx_clk_rate);
> > ---
> > 
> > Signed-off-by: Jan Petrous (OSS) <jan.petrous@oss.nxp.com>
> > ---
> >  include/linux/phy.h | 21 +++++++++++++++++++++
> >  1 file changed, 21 insertions(+)
> > 
> > diff --git a/include/linux/phy.h b/include/linux/phy.h
> > index 6b7d40d49129..bb797364d91c 100644
> > --- a/include/linux/phy.h
> > +++ b/include/linux/phy.h
> > @@ -298,6 +298,27 @@ static inline const char *phy_modes(phy_interface_t interface)
> >  	}
> >  }
> >  
> > +/**
> > + * rgmi_clock - map link speed to the clock rate
> 
> nit: rgmii_clock
> 
>      Flagged by ./scripts/kernel-doc -none
> 

Thanks. Fixed in v3.

BR.
/Jan

Re: [PATCH v2 4/7] net: phy: add helper for mapping RGMII link speed to clock rate

[Jan Petrous]

On Mon, Aug 19, 2024 at 05:57:11PM +0200, Andrew Lunn wrote:
> On Sun, Aug 18, 2024 at 09:50:46PM +0000, Jan Petrous (OSS) wrote:
> > The helper rgmii_clock() implemented Russel's hint during stmmac
> > glue driver review:
> > 
> > ---
> > We seem to have multiple cases of very similar logic in lots of stmmac
> > platform drivers, and I think it's about time we said no more to this.
> > So, what I think we should do is as follows:
> > 
> > add the following helper - either in stmmac, or more generically
> > (phylib? - in which case its name will need changing.)
> 
> As Russell pointed out, this code appears a few times in the stmmac
> driver. Please include patches changing the other instances to use
> this helper.
> 
> It also looks like macb, and maybe xgene_enet_hw.c could use it as
> well.
> 

OK, for v3 added patches for the following possible users:
dwmac-sti, xgene_enet, macb, dwmac-starfive, dwmac-rk,
dwmac-intel-plat, dwmac-imx, dwmac-dwc-qos-eth.

BR.
/Jan