Re: [PATCH] mtd: nand: atmel_nand: fix a possible NULL dereference

[LABBE Corentin <clabbe.montjoie@gmail.com>]

On Thu, Nov 12, 2015 at 09:29:57AM +0100, Joachim Eastwood wrote:
> On 12 November 2015 at 08:49, LABBE Corentin <clabbe.montjoie@gmail.com> wrote:
> > of_match_device could return NULL, and so cause a NULL pointer
> > dereference later.
> >
> > Signed-off-by: LABBE Corentin <clabbe.montjoie@gmail.com>
> > ---
> >  drivers/mtd/nand/atmel_nand.c | 7 +++++--
> >  1 file changed, 5 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/mtd/nand/atmel_nand.c b/drivers/mtd/nand/atmel_nand.c
> > index 475c938..f3cf68b 100644
> > --- a/drivers/mtd/nand/atmel_nand.c
> > +++ b/drivers/mtd/nand/atmel_nand.c
> > @@ -1495,9 +1495,12 @@ static int atmel_of_init_port(struct atmel_nand_host *host,
> >         int ecc_mode;
> >         struct atmel_nand_data *board = &host->board;
> >         enum of_gpio_flags flags = 0;
> > +       const struct of_device_id *of_id;
> >
> > -       host->caps = (struct atmel_nand_caps *)
> > -               of_match_device(atmel_nand_dt_ids, host->dev)->data;
> > +       of_id = of_match_device(atmel_nand_dt_ids, host->dev);
> > +       if (!of_id)
> > +               return -ENODEV;
> > +       host->caps = of_id->data;
> 
> It might be cleaner to use of_device_get_match_data() here.
> 

I agree, I will resend shortly with it.
And this time, I wont forgot the patch which permit to supress the kbuild_robot warning.

Regards

LABBE Corentin