linux-kernel - Re: [PATCH v11 1/5] can: m_can: Create a m

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] [day] [month] [year] [list]

Date:   Thu, 9 May 2019 08:38:29 +0200
From:   Marc Kleine-Budde <mkl@...gutronix.de>
To:     Dan Murphy <dmurphy@...com>, wg@...ndegger.com, davem@...emloft.net
Cc:     linux-can@...r.kernel.org, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v11 1/5] can: m_can: Create a m_can platform framework

On 5/8/19 9:54 PM, Dan Murphy wrote:
>>> -/* m_can private data structure */
>>> -struct m_can_priv {
>>> -	struct can_priv can;	/* must be the first member */
>>> -	struct napi_struct napi;
>>> -	struct net_device *dev;
>>> -	struct device *device;
>>> -	struct clk *hclk;
>>> -	struct clk *cclk;
>>> -	void __iomem *base;
>>> -	u32 irqstatus;
>>> -	int version;
>>> -
>>> -	/* message ram configuration */
>>> -	void __iomem *mram_base;
>>> -	struct mram_cfg mcfg[MRAM_CFG_NUM];
>>> -};
>>> +static u32 m_can_read(struct m_can_priv *priv, enum m_can_reg reg)
>>> +{
>>> +	if (priv->ops->read_reg)
>>> +		return priv->ops->read_reg(priv, reg);
>>> +	else
>>> +		return -EINVAL;
>>> +}
>>
>> How do you plan to check the return value here?
>> What's the difference between a register value of 0xffffffe9 and
>> returning -EINVAL?
> 
> Good point.  I could just inline this and return whatever is sent
> from the callback and as you said allow a backtrace to happen if
> read_reg is invalid.

ACK.
A minimal function will look like this:

static inline u32 m_can_read(struct m_can_priv *priv, enum m_can_reg reg)
{
	return priv->ops->read_reg(priv, reg);
}

And of ops or read_reg is NULL, the kernel will access a NULL pointer
that would then generate a backtrace.

>>> -static inline u32 m_can_read(const struct m_can_priv *priv, enum m_can_reg reg)
>>> +static int m_can_write(struct m_can_priv *priv, enum m_can_reg reg, u32 val)
>>>  {
>>> -	return readl(priv->base + reg);
>>> +	if (priv->ops->write_reg)
>>> +		return priv->ops->write_reg(priv, reg, val);
>>> +	else
>>> +		return -EINVAL;
>>>  }
>>
>> I don't see anyone checking the return value. Better just dereference
>> the pointer and the kernel will produce a nice backtrace.
>>
>> Same should be done for all read and write variants.
>>
> 
> I will need to go through this and see if there is any caller checking the return.  But
> I think you are correct.  If thats true I will just change this to a void, inline the function
> and allow a backtrace if the callback is null

ok

regards,
Marc

-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |



Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)