Re: [PATCH] remoteproc: mediatek: Dereferencing a pointer that might be NULL

[Jason-ch Chen (陳建豪)]

Hi Mathieu,

On Fri, 2023-03-03 at 11:20 -0700, Mathieu Poirier wrote:
> On Fri, 3 Mar 2023 at 03:28, Jason-ch Chen <
> jason-ch.chen@xxxxxxxxxxxx> wrote:
> > 
> > From: jason-ch chen <Jason-ch.Chen@xxxxxxxxxxxx>
> > 
> > The res might be NULL when calling resource_size.
> > 
> 
> "resource_size" ?  What is that?
> 
> > Signed-off-by: jason-ch chen <Jason-ch.Chen@xxxxxxxxxxxx>
> > ---
> >  drivers/remoteproc/mtk_scp.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/drivers/remoteproc/mtk_scp.c
> > b/drivers/remoteproc/mtk_scp.c
> > index eacdf241f4ef..863d18f63f58 100644
> > --- a/drivers/remoteproc/mtk_scp.c
> > +++ b/drivers/remoteproc/mtk_scp.c
> > @@ -829,6 +829,9 @@ static int scp_probe(struct platform_device
> > *pdev)
> >         platform_set_drvdata(pdev, scp);
> > 
> >         res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
> > "sram");
> > +       if (!res)
> > +               return -ENODEV;
> > +
> 
> Please have a look at the implementation of devm_ioremap_resource()
> along with the function it calls and let me know if your patch is
> needed.
> 
Thanks for your reply.
It seems is patch is not needed.
> >         scp->sram_base = devm_ioremap_resource(dev, res);
> >         if (IS_ERR((__force void *)scp->sram_base)) {
> >                 dev_err(dev, "Failed to parse and map sram
> > memory\n");
> > --
> > 2.37.3
> >