Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Thu, 4 Oct 2018 17:16:30 +0200
From:   Andrew Lunn <andrew@lunn.ch>
To:     Quentin Schulz <quentin.schulz@bootlin.com>
Cc:     davem@davemloft.net, f.fainelli@gmail.com,
        allan.nielsen@microchip.com, linux-kernel@vger.kernel.org,
        netdev@vger.kernel.org, thomas.petazzoni@bootlin.com,
        alexandre.belloni@bootlin.com
Subject: Re: [PATCH net-next v2 1/6] net: phy: mscc: migrate to
 phy_select/restore_page functions
Message-ID: <20181004151630.GE4730@lunn.ch>
References: <20181004124728.9821-1-quentin.schulz@bootlin.com>
 <20181004124728.9821-2-quentin.schulz@bootlin.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20181004124728.9821-2-quentin.schulz@bootlin.com>
User-Agent: Mutt/1.5.23 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Oct 04, 2018 at 02:47:23PM +0200, Quentin Schulz wrote:
> @@ -197,25 +199,30 @@ static int vsc85xx_mdix_set(struct phy_device *phydev, u8 mdix)
>  	if (rc != 0)
>  		return rc;
>  
> -	rc = vsc85xx_phy_page_set(phydev, MSCC_PHY_PAGE_EXTENDED);
> -	if (rc != 0)
> -		return rc;
> +	oldpage = phy_select_page(phydev, MSCC_PHY_PAGE_EXTENDED);
> +	if (oldpage < 0) {
> +		rc = oldpage;
> +		goto out;
> +	}
>  
> -	reg_val = phy_read(phydev, MSCC_PHY_EXT_MODE_CNTL);
> +	reg_val = __phy_read(phydev, MSCC_PHY_EXT_MODE_CNTL);
>  	reg_val &= ~(FORCE_MDI_CROSSOVER_MASK);
>  	if (mdix == ETH_TP_MDI)
>  		reg_val |= FORCE_MDI_CROSSOVER_MDI;
>  	else if (mdix == ETH_TP_MDI_X)
>  		reg_val |= FORCE_MDI_CROSSOVER_MDIX;

Hi Quentin

Could you use phy_modify_paged() here? This function only accesses a
single register, so using the wrapper should not have any
disadvantages. The same should apply for any function modifying a single
register.

> -	rc = phy_write(phydev, MSCC_PHY_EXT_MODE_CNTL, reg_val);
> +	rc = __phy_write(phydev, MSCC_PHY_EXT_MODE_CNTL, reg_val);
>  	if (rc != 0)
> -		return rc;
> +		goto out;
>  
> -	rc = vsc85xx_phy_page_set(phydev, MSCC_PHY_PAGE_STANDARD);
> -	if (rc != 0)
> +	rc = phy_restore_page(phydev, oldpage, rc);
> +	if (rc < 0)
>  		return rc;
>  
>  	return genphy_restart_aneg(phydev);
> +
> +out:
> +	return phy_restore_page(phydev, oldpage, rc);
>  }
>  
>  static int vsc85xx_downshift_get(struct phy_device *phydev, u8 *count)
> @@ -223,25 +230,24 @@ static int vsc85xx_downshift_get(struct phy_device *phydev, u8 *count)
>  	int rc;
>  	u16 reg_val;
>  
> -	rc = vsc85xx_phy_page_set(phydev, MSCC_PHY_PAGE_EXTENDED);
> -	if (rc != 0)
> +	rc = phy_select_page(phydev, MSCC_PHY_PAGE_EXTENDED);
> +	if (rc < 0)
>  		goto out;
>  
> -	reg_val = phy_read(phydev, MSCC_PHY_ACTIPHY_CNTL);
> +	reg_val = __phy_read(phydev, MSCC_PHY_ACTIPHY_CNTL);
>  	reg_val &= DOWNSHIFT_CNTL_MASK;
>  	if (!(reg_val & DOWNSHIFT_EN))
>  		*count = DOWNSHIFT_DEV_DISABLE;
>  	else
>  		*count = ((reg_val & ~DOWNSHIFT_EN) >> DOWNSHIFT_CNTL_POS) + 2;

phy_read_paged()? Any function which just accesses a single register
should try to use phy_read_paged(), phy_write_paged() or
phy_modify_paged() to avoid a lot of boilerplate code.

	   Andrew