Re: [PATCH 8/9] gpio: mockup: use the generic 'gpio-line-names' property

[Bartosz Golaszewski <brgl@bgdev.pl>]

On Fri, Sep 25, 2020 at 6:41 PM Andy Shevchenko
<andriy.shevchenko@linux.intel.com> wrote:
>
> On Fri, Sep 25, 2020 at 01:40:10PM +0200, Bartosz Golaszewski wrote:
> > On Fri, Sep 25, 2020 at 11:03 AM Andy Shevchenko
> > <andriy.shevchenko@linux.intel.com> wrote:
> > > On Thu, Sep 24, 2020 at 01:38:41PM +0200, Bartosz Golaszewski wrote:
> > > > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> > > >
> > > > GPIO line names are currently created by the driver from the chip label.
> > > > We'll want to support custom formats for line names (for instance: to
> > > > name all lines the same) for user-space tests so create them in the
> > > > module init function and pass them to the driver using the standard
> > > > 'gpio-line-names' property.
> > >
> > > ...
> > >
> > > > +             if (gpio_mockup_named_lines) {
> > > > +                     line_names = gpio_mockup_make_line_names(chip_label,
> > > > +                                                              ngpio);
> > > > +                     if (!line_names) {
> > > > +                             platform_driver_unregister(&gpio_mockup_driver);
> > > > +                             gpio_mockup_unregister_pdevs();
> > > > +                             return -ENOMEM;
> > > > +                     }
> > >
> > > > +                     properties[prop++] = PROPERTY_ENTRY_STRING_ARRAY_LEN(
> > > > +                                             "gpio-line-names",
> > > > +                                             line_names, ngpio);
> > >
> > > Forgot to update GPIO_MOCKUP_MAX_PROP?
> > >
> >
> > No, there are still three properties: chip-label, nr-gpios and
> > gpio-line-names. Same answer to patch 8/9.
> >
> > > > +             }
> > >
> > > ...
> > >
> > > > +             kfree_strarray(line_names, line_names ? ngpio : 0);
> > >
> > > Perhaps you may check for NULL pointer in the kfree_strarray() and drop ternary
> > > here?
> > >
> >
> > I did in the previous series and you told me to not to. :)
>
> Hmm... What was my argument? What was wrong with me? free() should be NULL-aware.
>

Well, it is - your just need to make sure ngpio is 0 too. :)

I'll revert back to having the NULL check.

Bartosz