Re: [PATCH] i2c: ACPI: Replace acpi_bus_get_device()

[Andy Shevchenko <andriy.shevchenko@linux.intel.com>]

On Tue, Feb 01, 2022 at 08:01:46PM +0100, Rafael J. Wysocki wrote:
> On Tue, Feb 1, 2022 at 7:44 PM Andy Shevchenko
> <andriy.shevchenko@linux.intel.com> wrote:
> >
> > On Tue, Feb 01, 2022 at 07:00:42PM +0100, Rafael J. Wysocki wrote:
> > > From: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
> > >
> > > Replace acpi_bus_get_device() that is going to be dropped with
> > > acpi_fetch_acpi_dev().
> > >
> > > No intentional functional impact.
> >
> > ...
> >
> > > +     if (!adev || i2c_acpi_get_info(adev, &info, adapter, NULL))
> >
> > AFAICS the !adev check is redundant since acpi_device_enumerated() does it.
> 
> No.
> 
> acpi_device_enumerated() returns false if adev is NULL, so without
> this extra check i2c_acpi_get_info() will end up passing NULL to
> i2c_acpi_do_lookup().

I see now.

The patch LGTM,
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

> > >               return AE_OK;
> >
> > ...
> >
> > > +     struct acpi_device *adev = acpi_fetch_acpi_dev(handle);
> > >
> > > -     if (i2c_acpi_do_lookup(adev, lookup))
> > > +     if (!adev || i2c_acpi_do_lookup(adev, lookup))
> > >               return AE_OK;
> >
> > Here we need it indeed.
> > Dunno, if acpi_dev_ready_for_enumeration() can gain the check itself.
> 
> Well, acpi_bus_get_status() would need it too.

-- 
With Best Regards,
Andy Shevchenko