netdev - Re: [PATCH net-next v3 5/5] net-next: dsa: add dsa support for Mediatek MT7530 switch

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20170329231251.GH28922@lunn.ch>
Date:   Thu, 30 Mar 2017 01:12:51 +0200
From:   Andrew Lunn <andrew@...n.ch>
To:     sean.wang@...iatek.com
Cc:     f.fainelli@...il.com, vivien.didelot@...oirfairelinux.com,
        matthias.bgg@...il.com, robh+dt@...nel.org, mark.rutland@....com,
        devicetree@...r.kernel.org, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-mediatek@...ts.infradead.org,
        davem@...emloft.net, Landen.Chao@...iatek.com, keyhaede@...il.com,
        objelf@...il.com
Subject: Re: [PATCH net-next v3 5/5] net-next: dsa: add dsa support for
 Mediatek MT7530 switch

> +static struct mt7530_priv *lpriv;

Hi Sean

Why do you need this global variable? What if somebody has two
switches connected?

> +static void mt7530_port_disable(struct dsa_switch *ds, int port,
> +				struct phy_device *phy);
> +static int mt7530_cpu_port_enable(struct mt7530_priv *priv,
> +				  int port);

It is better to move the functions around than have forward
declarations.

> +static u32
> +_mt7530_read(u32 reg)
> +{
> +	struct mt7530_priv	*priv = lpriv;
> +	struct mii_bus		*bus = priv->bus;
> +	u32 val;
> +
> +	mutex_lock_nested(&bus->mdio_lock, MDIO_MUTEX_NESTED);
> +
> +	val = mt7530_mii_read(priv, reg);
> +
> +	mutex_unlock(&bus->mdio_lock);
> +
> +	return val;
> +}
> +
> +static u32
> +mt7530_read(struct mt7530_priv *priv, u32 reg)
> +{
> +	return _mt7530_read(reg);
> +}

So here you would pass priv to _mt7530_read().

> +static int
> +mt7530_fdb_cmd(struct mt7530_priv *priv, enum mt7530_fdb_cmd cmd, u32 *rsp)
> +{
> +	u32 val;
> +	int ret;
> +
> +	/* Set the command operating upon the MAC address entries */
> +	val = ATC_BUSY | ATC_MAT(0) | cmd;
> +	mt7530_write(priv, MT7530_ATC, val);
> +
> +	ret = readx_poll_timeout(_mt7530_read, MT7530_ATC, val,
> +				 !(val & ATC_BUSY), 20, 20000);

This is where your need for lpriv comes, you can only pass a single
argument. But you can work around it. readx_poll_timeout is a
#define. So there is no type check for _mt7530_read, or the arg passed
to it. So pass a struct containing the address and priv.

I think you can then kill of your global lpriv.

  Andrew