netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Vladimir Oltean <olteanv@gmail.com>
To: Andrew Lunn <andrew@lunn.ch>
Cc: netdev <netdev@vger.kernel.org>,
	Sean Wang <sean.wang@mediatek.com>,
	Landen Chao <Landen.Chao@mediatek.com>,
	DENG Qingfang <dqfext@gmail.com>,
	Vivien Didelot <vivien.didelot@gmail.com>,
	Florian Fainelli <f.fainelli@gmail.com>,
	Matthias Brugger <matthias.bgg@gmail.com>,
	Joakim Zhang <qiangqing.zhang@nxp.com>,
	Sergey Shtylyov <s.shtylyov@omp.ru>,
	Heiner Kallweit <hkallweit1@gmail.com>,
	Russell King <linux@armlinux.org.uk>,
	Tobias Waldekranz <tobias@waldekranz.com>,
	Marcin Wojtas <mw@semihalf.com>,
	Calvin Johnson <calvin.johnson@oss.nxp.com>,
	Markus Koch <markus@notsyncing.net>,
	Geert Uytterhoeven <geert+renesas@glider.be>,
	Yang Yingliang <yangyingliang@huawei.com>,
	Hao Chen <chenhao288@hisilicon.com>
Subject: Re: [PATCH net-next 07/10] net: ethernet: freescale: xgmac: Separate C22 and C45 transactions for xgmac
Date: Tue, 10 May 2022 21:28:18 +0300	[thread overview]
Message-ID: <20220510182818.w7kl3vmlgvqjjj4u@skbuf> (raw)
In-Reply-To: <20220508153049.427227-8-andrew@lunn.ch>

On Sun, May 08, 2022 at 05:30:46PM +0200, Andrew Lunn wrote:
> The xgmac MDIO bus driver can perform both C22 and C45 transfers.
> Create separate functions for each and register the C45 versions using
> the new API calls where appropriate.
> 
> Signed-off-by: Andrew Lunn <andrew@lunn.ch>
> ---
>  drivers/net/ethernet/freescale/xgmac_mdio.c | 154 +++++++++++++++-----
>  1 file changed, 117 insertions(+), 37 deletions(-)
> 
> diff --git a/drivers/net/ethernet/freescale/xgmac_mdio.c b/drivers/net/ethernet/freescale/xgmac_mdio.c
> index ec90da1de030..ddfe6bf1f231 100644
> --- a/drivers/net/ethernet/freescale/xgmac_mdio.c
> +++ b/drivers/net/ethernet/freescale/xgmac_mdio.c
> @@ -128,30 +128,59 @@ static int xgmac_wait_until_done(struct device *dev,
>  	return 0;
>  }
>  
> -/*
> - * Write value to the PHY for this device to the register at regnum,waiting
> +/* Write value to the PHY for this device to the register at regnum,waiting
>   * until the write is done before it returns.  All PHY configuration has to be
>   * done through the TSEC1 MIIM regs.
>   */
> -static int xgmac_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 value)
> +static int xgmac_mdio_write_c22(struct mii_bus *bus, int phy_id, int regnum,
> +				u16 value)
>  {
>  	struct mdio_fsl_priv *priv = (struct mdio_fsl_priv *)bus->priv;
>  	struct tgec_mdio_controller __iomem *regs = priv->mdio_base;
> -	uint16_t dev_addr;
> +	bool endian = priv->is_little_endian;
>  	u32 mdio_ctl, mdio_stat;
> +	u16 dev_addr;
>  	int ret;
> +
> +	mdio_stat = xgmac_read32(&regs->mdio_stat, endian);
> +	dev_addr = regnum & 0x1f;

Please move this either to the variable declaration, or near the mdio_ctl write,
or just integrate it into the macro argument.

> +	mdio_stat &= ~MDIO_STAT_ENC;
> +

You can remove this empty line during read-modify-write patterns.

> +	xgmac_write32(mdio_stat, &regs->mdio_stat, endian);
> +
> +	ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> +	if (ret)
> +		return ret;
> +
> +	/* Set the port and dev addr */
> +	mdio_ctl = MDIO_CTL_PORT_ADDR(phy_id) | MDIO_CTL_DEV_ADDR(dev_addr);
> +	xgmac_write32(mdio_ctl, &regs->mdio_ctl, endian);
> +
> +	/* Write the value to the register */
> +	xgmac_write32(MDIO_DATA(value), &regs->mdio_data, endian);
> +
> +	ret = xgmac_wait_until_done(&bus->dev, regs, endian);
> +	if (ret)
> +		return ret;
> +
> +	return 0;
> +}
> +
> +/* Write value to the PHY for this device to the register at regnum,waiting
> + * until the write is done before it returns.  All PHY configuration has to be
> + * done through the TSEC1 MIIM regs.
> + */
> +static int xgmac_mdio_write_c45(struct mii_bus *bus, int phy_id, int dev_addr,
> +				int regnum, u16 value)
> +{
> +	struct mdio_fsl_priv *priv = (struct mdio_fsl_priv *)bus->priv;
> +	struct tgec_mdio_controller __iomem *regs = priv->mdio_base;
>  	bool endian = priv->is_little_endian;
> +	u32 mdio_ctl, mdio_stat;
> +	int ret;
>  
>  	mdio_stat = xgmac_read32(&regs->mdio_stat, endian);
> -	if (regnum & MII_ADDR_C45) {
> -		/* Clause 45 (ie 10G) */
> -		dev_addr = (regnum >> 16) & 0x1f;
> -		mdio_stat |= MDIO_STAT_ENC;
> -	} else {
> -		/* Clause 22 (ie 1G) */
> -		dev_addr = regnum & 0x1f;
> -		mdio_stat &= ~MDIO_STAT_ENC;
> -	}
> +	mdio_stat |= MDIO_STAT_ENC;
>  
>  	xgmac_write32(mdio_stat, &regs->mdio_stat, endian);
>  
> @@ -164,13 +193,11 @@ static int xgmac_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 val
>  	xgmac_write32(mdio_ctl, &regs->mdio_ctl, endian);
>  
>  	/* Set the register address */
> -	if (regnum & MII_ADDR_C45) {
> -		xgmac_write32(regnum & 0xffff, &regs->mdio_addr, endian);
> +	xgmac_write32(regnum & 0xffff, &regs->mdio_addr, endian);

Does regnum ever exceed 0xffff now that the MMD is no longer encoded into it?

>  
> -		ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> -		if (ret)
> -			return ret;
> -	}
> +	ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> +	if (ret)
> +		return ret;
>  
>  	/* Write the value to the register */
>  	xgmac_write32(MDIO_DATA(value), &regs->mdio_data, endian);
> @@ -182,31 +209,84 @@ static int xgmac_mdio_write(struct mii_bus *bus, int phy_id, int regnum, u16 val
>  	return 0;
>  }
>  
> -/*
> - * Reads from register regnum in the PHY for device dev, returning the value.
> +/* Reads from register regnum in the PHY for device dev, returning the value.
>   * Clears miimcom first.  All PHY configuration has to be done through the
>   * TSEC1 MIIM regs.
>   */
> -static int xgmac_mdio_read(struct mii_bus *bus, int phy_id, int regnum)
> +static int xgmac_mdio_read_c22(struct mii_bus *bus, int phy_id, int regnum)
>  {
>  	struct mdio_fsl_priv *priv = (struct mdio_fsl_priv *)bus->priv;
>  	struct tgec_mdio_controller __iomem *regs = priv->mdio_base;
> +	bool endian = priv->is_little_endian;
>  	unsigned long flags;
> -	uint16_t dev_addr;
>  	uint32_t mdio_stat;
>  	uint32_t mdio_ctl;
> +	u16 dev_addr;
>  	int ret;
> -	bool endian = priv->is_little_endian;
>  
>  	mdio_stat = xgmac_read32(&regs->mdio_stat, endian);
> -	if (regnum & MII_ADDR_C45) {
> -		dev_addr = (regnum >> 16) & 0x1f;
> -		mdio_stat |= MDIO_STAT_ENC;
> +	dev_addr = regnum & 0x1f;

I'm thinking we could just pass "regnum & 0x1f" (or just regnum) to
MDIO_CTL_DEV_ADDR() for the c22 functions.

> +	mdio_stat &= ~MDIO_STAT_ENC;
> +
> +	xgmac_write32(mdio_stat, &regs->mdio_stat, endian);
> +
> +	ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> +	if (ret)
> +		return ret;
> +
> +	/* Set the Port and Device Addrs */
> +	mdio_ctl = MDIO_CTL_PORT_ADDR(phy_id) | MDIO_CTL_DEV_ADDR(dev_addr);
> +	xgmac_write32(mdio_ctl, &regs->mdio_ctl, endian);
> +
> +	if (priv->has_a009885)
> +		/* Once the operation completes, i.e. MDIO_STAT_BSY clears, we
> +		 * must read back the data register within 16 MDC cycles.
> +		 */
> +		local_irq_save(flags);
> +
> +	/* Initiate the read */
> +	xgmac_write32(mdio_ctl | MDIO_CTL_READ, &regs->mdio_ctl, endian);
> +
> +	ret = xgmac_wait_until_done(&bus->dev, regs, endian);
> +	if (ret)
> +		goto irq_restore;
> +
> +	/* Return all Fs if nothing was there */
> +	if ((xgmac_read32(&regs->mdio_stat, endian) & MDIO_STAT_RD_ER) &&
> +	    !priv->has_a011043) {
> +		dev_dbg(&bus->dev,
> +			"Error while reading PHY%d reg at %d.%d\n",
> +			phy_id, dev_addr, regnum);
> +		ret = 0xffff;
>  	} else {
> -		dev_addr = regnum & 0x1f;
> -		mdio_stat &= ~MDIO_STAT_ENC;
> +		ret = xgmac_read32(&regs->mdio_data, endian) & 0xffff;
> +		dev_dbg(&bus->dev, "read %04x\n", ret);
>  	}
>  
> +irq_restore:
> +	if (priv->has_a009885)
> +		local_irq_restore(flags);
> +
> +	return ret;
> +}
> +
> +/* Reads from register regnum in the PHY for device dev, returning the value.
> + * Clears miimcom first.  All PHY configuration has to be done through the
> + * TSEC1 MIIM regs.
> + */
> +static int xgmac_mdio_read_c45(struct mii_bus *bus, int phy_id, int dev_addr,
> +			       int regnum)
> +{
> +	struct mdio_fsl_priv *priv = (struct mdio_fsl_priv *)bus->priv;
> +	struct tgec_mdio_controller __iomem *regs = priv->mdio_base;
> +	bool endian = priv->is_little_endian;
> +	u32 mdio_stat, mdio_ctl;
> +	unsigned long flags;
> +	int ret;
> +
> +	mdio_stat = xgmac_read32(&regs->mdio_stat, endian);
> +	mdio_stat |= MDIO_STAT_ENC;
> +
>  	xgmac_write32(mdio_stat, &regs->mdio_stat, endian);
>  
>  	ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> @@ -218,13 +298,11 @@ static int xgmac_mdio_read(struct mii_bus *bus, int phy_id, int regnum)
>  	xgmac_write32(mdio_ctl, &regs->mdio_ctl, endian);
>  
>  	/* Set the register address */
> -	if (regnum & MII_ADDR_C45) {
> -		xgmac_write32(regnum & 0xffff, &regs->mdio_addr, endian);
> +	xgmac_write32(regnum & 0xffff, &regs->mdio_addr, endian);
>  
> -		ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> -		if (ret)
> -			return ret;
> -	}
> +	ret = xgmac_wait_until_free(&bus->dev, regs, endian);
> +	if (ret)
> +		return ret;
>  
>  	if (priv->has_a009885)
>  		/* Once the operation completes, i.e. MDIO_STAT_BSY clears, we
> @@ -326,8 +404,10 @@ static int xgmac_mdio_probe(struct platform_device *pdev)
>  		return -ENOMEM;
>  
>  	bus->name = "Freescale XGMAC MDIO Bus";
> -	bus->read = xgmac_mdio_read;
> -	bus->write = xgmac_mdio_write;
> +	bus->read = xgmac_mdio_read_c22;
> +	bus->write = xgmac_mdio_write_c22;
> +	bus->read_c45 = xgmac_mdio_read_c45;
> +	bus->write_c45 = xgmac_mdio_write_c45;
>  	bus->parent = &pdev->dev;
>  	bus->probe_capabilities = MDIOBUS_C22_C45;
>  	snprintf(bus->id, MII_BUS_ID_SIZE, "%pa", &res->start);
> -- 
> 2.36.0
> 

  reply	other threads:[~2022-05-10 18:28 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-08 15:30 [PATCH net-next 00/10] net: mdio: Start separating C22 and C45 Andrew Lunn
2022-05-08 15:30 ` [PATCH net-next 01/10] net: mdio: Add dedicated C45 API to MDIO bus drivers Andrew Lunn
2022-05-08 15:30 ` [PATCH net-next 02/10] net: mdio: mdiobus_register: Update validation test Andrew Lunn
2022-05-09  7:49   ` Sergey Shtylyov
2022-05-08 15:30 ` [PATCH net-next 03/10] net: mdio: C22 is now optional, EOPNOTSUPP if not provided Andrew Lunn
2022-05-08 15:30 ` [PATCH net-next 04/10] net: mdio: Move mdiobus_c45_addr() next to users Andrew Lunn
2022-05-08 15:30 ` [PATCH net-next 05/10] net: mdio: mdio-bitbang: Separate C22 and C45 transactions Andrew Lunn
2022-05-09  7:52   ` Geert Uytterhoeven
2022-05-08 15:30 ` [PATCH net-next 06/10] net: mdio: mvmdio: Convert XSMI bus to new API Andrew Lunn
2022-05-08 15:30 ` [PATCH net-next 07/10] net: ethernet: freescale: xgmac: Separate C22 and C45 transactions for xgmac Andrew Lunn
2022-05-10 18:28   ` Vladimir Oltean [this message]
2022-05-10 19:01     ` Andrew Lunn
2022-05-10 19:09       ` Vladimir Oltean
2022-05-08 15:30 ` [PATCH net-next 08/10] net: ethernet: freescale: fec: " Andrew Lunn
2022-05-10 18:30   ` Vladimir Oltean
2022-05-08 15:30 ` [PATCH net-next 09/10] net: dsa: Separate C22 and C45 MDIO bus transaction methods Andrew Lunn
2022-05-08 17:36   ` kernel test robot
2022-05-08 19:19   ` kernel test robot
2022-05-10 19:05   ` Vladimir Oltean
2022-05-08 15:30 ` [PATCH net-next 10/10] net: dsa: mv88e6xxx: Separate C22 and C45 transactions Andrew Lunn

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220510182818.w7kl3vmlgvqjjj4u@skbuf \
    --to=olteanv@gmail.com \
    --cc=Landen.Chao@mediatek.com \
    --cc=andrew@lunn.ch \
    --cc=calvin.johnson@oss.nxp.com \
    --cc=chenhao288@hisilicon.com \
    --cc=dqfext@gmail.com \
    --cc=f.fainelli@gmail.com \
    --cc=geert+renesas@glider.be \
    --cc=hkallweit1@gmail.com \
    --cc=linux@armlinux.org.uk \
    --cc=markus@notsyncing.net \
    --cc=matthias.bgg@gmail.com \
    --cc=mw@semihalf.com \
    --cc=netdev@vger.kernel.org \
    --cc=qiangqing.zhang@nxp.com \
    --cc=s.shtylyov@omp.ru \
    --cc=sean.wang@mediatek.com \
    --cc=tobias@waldekranz.com \
    --cc=vivien.didelot@gmail.com \
    --cc=yangyingliang@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).