Re: [mfd] question about potential null pointer dereference

[Lee Jones]

On Tue, 23 May 2017, Gustavo A. R. Silva wrote:

> 
> Hello everybody,
> 
> While looking into Coverity ID 1408830 I ran into the following piece of
> code at drivers/mfd/wm831x-spi.c:26
> 
>  26static int wm831x_spi_probe(struct spi_device *spi)
>  27{
>  28        struct wm831x_pdata *pdata = dev_get_platdata(&spi->dev);
>  29        const struct spi_device_id *id = spi_get_device_id(spi);
>  30        const struct of_device_id *of_id;
>  31        struct wm831x *wm831x;
>  32        enum wm831x_parent type;
>  33        int ret;
>  34
>  35        if (spi->dev.of_node) {
>  36                of_id = of_match_device(wm831x_of_match, &spi->dev);
>  37                type = (enum wm831x_parent)of_id->data;
>  38        } else {
>  39                type = (enum wm831x_parent)id->driver_data;
>  40        }
>  41
>  42        wm831x = devm_kzalloc(&spi->dev, sizeof(struct wm831x),
> GFP_KERNEL);
>  43        if (wm831x == NULL)
>  44                return -ENOMEM;
>  45
>  46        spi->mode = SPI_MODE_0;
>  47
>  48        spi_set_drvdata(spi, wm831x);
>  49        wm831x->dev = &spi->dev;
>  50        wm831x->type = type;
>  51
>  52        wm831x->regmap = devm_regmap_init_spi(spi, &wm831x_regmap_config);
>  53        if (IS_ERR(wm831x->regmap)) {
>  54                ret = PTR_ERR(wm831x->regmap);
>  55                dev_err(wm831x->dev, "Failed to allocate register map:
> %d\n",
>  56                        ret);
>  57                return ret;
>  58        }
>  59
>  60        if (pdata)
>  61                memcpy(&wm831x->pdata, pdata, sizeof(*pdata));
>  62
>  63        return wm831x_device_init(wm831x, spi->irq);
>  64}
> 
> The issue here is that there is a potential NULL pointer dereference at line
> 37, in case function of_match_device() returns NULL.
> 
> Maybe a patch like the following could be applied in order to avoid any
> chance of a NULL pointer dereference:

I do not believe it's possible for of_match_device() to return NULL in
this case.

However, if you wanted to submit a patch checking for it, it would not
be rejected.

> index c332e28..7b227c9 100644
> --- a/drivers/mfd/wm831x-spi.c
> +++ b/drivers/mfd/wm831x-spi.c
> @@ -34,6 +34,10 @@ static int wm831x_spi_probe(struct spi_device *spi)
> 
>         if (spi->dev.of_node) {
>                 of_id = of_match_device(wm831x_of_match, &spi->dev);
> +               if (!of_id) {
> +                       dev_err(&spi->dev, "Failed to find matching id\n");

"Failed to match device"

> +                       return -EINVAL;

-ENODEV

>                 type = (enum wm831x_parent)of_id->data;
>         } else {
>                 type = (enum wm831x_parent)id->driver_data;
> 
> What do you think?
> 
> I'd really appreciate any comment on this.
> 
> Thank you!

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org â Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog