Re: [PATCH v3] net: phy: add driver for Motorcomm yt8511 phy

From: Russell King (Oracle)
Date: Fri May 14 2021 - 09:09:19 EST


Hi!

On Fri, May 14, 2021 at 07:58:26AM -0400, Peter Geis wrote:
> + /* set rgmii delay mode */
> + val = __phy_read(phydev, YT8511_PAGE);
> +
> + switch (phydev->interface) {
> + case PHY_INTERFACE_MODE_RGMII:
> + val &= ~(YT8511_DELAY_RX | YT8511_DELAY_TX);
> + break;
> + case PHY_INTERFACE_MODE_RGMII_ID:
> + val |= YT8511_DELAY_RX | YT8511_DELAY_TX;
> + break;
> + case PHY_INTERFACE_MODE_RGMII_RXID:
> + val &= ~(YT8511_DELAY_TX);
> + val |= YT8511_DELAY_RX;
> + break;
> + case PHY_INTERFACE_MODE_RGMII_TXID:
> + val &= ~(YT8511_DELAY_RX);
> + val |= YT8511_DELAY_TX;
> + break;
> + default: /* leave everything alone in other modes */
> + break;
> + }
> +
> + ret = __phy_write(phydev, YT8511_PAGE, val);
> + if (ret < 0)
> + goto err_restore_page;

Another way of writing the above is to set "val" to be the value of the
YT8511_DELAY_RX and YT8511_DELAY_TX bits, and then do:

ret = __phy_modify(phydev, YT8511_PAGE,
(YT8511_DELAY_RX | YT8511_DELAY_TX), val);
if (ret < 0)
goto err_restore_page;

which moves the read-modify-write out of the driver into core code and
makes the driver code smaller. It also handles your missing error check
on __phy_read() above - would you want the above code to attempt to
write a -ve error number back to this register? I suspect not!

--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!