Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Wed, 28 Jun 2023 10:50:41 +0100
From:   "Russell King (Oracle)" <linux@armlinux.org.uk>
To:     wuych <yunchuan@nfschina.com>
Cc:     iyappan@os.amperecomputing.com, keyur@os.amperecomputing.com,
        quan@os.amperecomputing.com, andrew@lunn.ch, hkallweit1@gmail.com,
        davem@davemloft.net, edumazet@google.com, kuba@kernel.org,
        pabeni@redhat.com, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org
Subject: Re: [PATCH net-next 08/10] net: mdio: Remove unnecessary (void*)
 conversions
Message-ID: <ZJwCcWgi0d6kEepI@shell.armlinux.org.uk>
References: <20230628024517.1440644-1-yunchuan@nfschina.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230628024517.1440644-1-yunchuan@nfschina.com>
Sender: Russell King (Oracle) <linux@armlinux.org.uk>
Precedence: bulk

Hi,

I think you missed one case:

        if (mdio_id == XGENE_MDIO_RGMII) {
                mdio_bus->read = xgene_mdio_rgmii_read;
                mdio_bus->write = xgene_mdio_rgmii_write;
                mdio_bus->priv = (void __force *)pdata;

This cast using __force is also not required.

On Wed, Jun 28, 2023 at 10:45:17AM +0800, wuych wrote:
> @@ -211,7 +211,7 @@ static void xgene_enet_wr_mdio_csr(void __iomem *base_addr,
>  static int xgene_xfi_mdio_write(struct mii_bus *bus, int phy_id,
>  				int reg, u16 data)
>  {
> -	void __iomem *addr = (void __iomem *)bus->priv;
> +	void __iomem *addr = bus->priv;
>  	int timeout = 100;
>  	u32 status, val;
>  
> @@ -234,7 +234,7 @@ static int xgene_xfi_mdio_write(struct mii_bus *bus, int phy_id,
>  
>  static int xgene_xfi_mdio_read(struct mii_bus *bus, int phy_id, int reg)
>  {
> -	void __iomem *addr = (void __iomem *)bus->priv;
> +	void __iomem *addr = bus->priv;
>  	u32 data, status, val;
>  	int timeout = 100;

These probably cause Sparse to warn whether or not the cast is there.

Given that in this case, bus->priv is initialised via:

                mdio_bus->priv = (void __force *)pdata->mdio_csr_addr;

I think the simple thing is to _always_ initialise mdio_bus->priv
to point at pdata, and have xgene_xfi_mdio_*() always do:

	struct xgene_mdio_pdata *pdata = bus->priv;
	void __iomem *addr = pdata->mdio_csr_addr;

The extra access will be dwarfed by the time taken to perform the
access.

This change should be made with a separate patch and not combined with
the patch removing the casts in xgene_mdio_rgmii_*().

Thanks.

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