On Wed, Oct 02, 2024 at 09:36:59PM +0200, Kees Bakker wrote:
> Op 30-09-2024 om 21:50 schreef Drew Fustini:
> > From: Emil Renner Berthing <emil.renner.berthing@xxxxxxxxxxxxx>
> >
> > Add pinctrl driver for the T-Head TH1520 RISC-V SoC.
> >
> > Tested-by: Thomas Bonnefille <thomas.bonnefille@xxxxxxxxxxx>
> > Signed-off-by: Emil Renner Berthing <emil.renner.berthing@xxxxxxxxxxxxx>
> > [dfustini: use thead,pad-group to identify the pin controller instance]
> > Signed-off-by: Drew Fustini <dfustini@xxxxxxxxxxxxxxx>
> > ---
> > MAINTAINERS | 1 +
> > drivers/pinctrl/Kconfig | 13 +
> > drivers/pinctrl/Makefile | 1 +
> > drivers/pinctrl/pinctrl-th1520.c | 907 +++++++++++++++++++++++++++++++++++++++
> > 4 files changed, 922 insertions(+)
> >
> > [...]
> > +static int th1520_pinmux_set_mux(struct pinctrl_dev *pctldev,
> > + unsigned int fsel, unsigned int gsel)
> > +{
> > + struct th1520_pinctrl *thp = pinctrl_dev_get_drvdata(pctldev);
> > + const struct function_desc *func = pinmux_generic_get_function(pctldev, fsel);
> func can be NULL after calling pinmux_generic_get_function
> Please add something to avoid NULL pointer dereferencing in the next
> statement.
> All other callers of pinmux_generic_get_function have something like this:
> if (!func)
> return -EINVAL;
Thanks for catching this. I see that another person has posted a patch
as a result so I'll respond over in that thread [1].
Drew
[1] https://lore.kernel.org/lkml/20241003023307.2138695-1-clf700383@xxxxxxxxx/
