Re: [PATCH 3/3] net: phy: mediatek: add support for built-in 2.5G ethernet PHY on MT7988

From: "Russell King (Oracle)" <linux@armlinux.org.uk>
To: Sky Huang <SkyLake.Huang@mediatek.com>
Cc: Andrew Lunn <andrew@lunn.ch>,
	Heiner Kallweit <hkallweit1@gmail.com>,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Daniel Golle <daniel@makrotopia.org>,
	Qingfang Deng <dqfext@gmail.com>,
	Matthias Brugger <matthias.bgg@gmail.com>,
	AngeloGioacchino Del Regno
	<angelogioacchino.delregno@collabora.com>,
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
	linux-arm-kernel@lists.infradead.org,
	linux-mediatek@lists.infradead.org,
	Steven Liu <Steven.Liu@mediatek.com>
Subject: Re: [PATCH 3/3] net: phy: mediatek: add support for built-in 2.5G ethernet PHY on MT7988
Date: Thu, 25 Apr 2024 10:01:58 +0100	[thread overview]
Message-ID: <ZiocBmBWiNnbeyGq@shell.armlinux.org.uk> (raw)
In-Reply-To: <20240425023325.15586-4-SkyLake.Huang@mediatek.com>

On Thu, Apr 25, 2024 at 10:33:25AM +0800, Sky Huang wrote:
> +static int mt7988_2p5ge_phy_config_init(struct phy_device *phydev)
> +{
> +	int ret, i;
> +	const struct firmware *fw;
> +	struct device *dev = &phydev->mdio.dev;
> +	struct device_node *np;
> +	void __iomem *pmb_addr;
> +	void __iomem *md32_en_cfg_base;
> +	struct mtk_i2p5ge_phy_priv *priv = phydev->priv;
> +	u16 reg;
> +	struct pinctrl *pinctrl;
> +
> +	if (!priv->fw_loaded) {
> +		np = of_find_compatible_node(NULL, NULL, "mediatek,2p5gphy-fw");
> +		if (!np)
> +			return -ENOENT;
> +		pmb_addr = of_iomap(np, 0);
> +		if (!pmb_addr)
> +			return -ENOMEM;
> +		md32_en_cfg_base = of_iomap(np, 1);
> +		if (!md32_en_cfg_base)
> +			return -ENOMEM;

Wouldn't it be better to do this in the probe function rather than here?

> +
> +		ret = request_firmware(&fw, MT7988_2P5GE_PMB, dev);
> +		if (ret) {
> +			dev_err(dev, "failed to load firmware: %s, ret: %d\n",
> +				MT7988_2P5GE_PMB, ret);
> +			return ret;
> +		}

This will block for userspace while holding phydev->lock and the RTNL.
That blocks much of the networking APIs, which is not a good idea. If
you have a number of these PHYs, then the RTNL will serialise the
loading of firmware.

> +
> +		reg = readw(md32_en_cfg_base);
> +		if (reg & MD32_EN) {
> +			phy_set_bits(phydev, 0, BIT(15));

This is probably the BMCR, so if it is, please use the established
definitions.

> +			usleep_range(10000, 11000);
> +		}
> +		phy_set_bits(phydev, 0, BIT(11));

This also looks like it's probably the BMCR.

> +
> +		/* Write magic number to safely stall MCU */
> +		phy_write_mmd(phydev, MDIO_MMD_VEND1, 0x800e, 0x1100);
> +		phy_write_mmd(phydev, MDIO_MMD_VEND1, 0x800f, 0x00df);
> +
> +		for (i = 0; i < fw->size - 1; i += 4)
> +			writel(*((uint32_t *)(fw->data + i)), pmb_addr + i);
> +		release_firmware(fw);
> +
> +		writew(reg & ~MD32_EN, md32_en_cfg_base);
> +		writew(reg | MD32_EN, md32_en_cfg_base);
> +		phy_set_bits(phydev, 0, BIT(15));

And also probably the BMCR.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!