Fri, Sep 27, 2024 at 01:20:51AM +0200, Javier Carrasco kirjoitti:
> This series switches from the device_for_each_child_node() macro to its
> scoped variant, which in general makes the code more robust if new early
> exits are added to the loops, because there is no need for explicit
> calls to fwnode_handle_put(). Depending on the complexity of the loop
> and its error handling, the code gets simplified and it gets easier to
> follow.
>
> The non-scoped variant of the macro is error-prone, and it has been the
> source of multiple bugs where the child node refcount was not
> decremented accordingly in error paths within the loops. The first patch
> of this series is a good example, which fixes that kind of bug that is
> regularly found in node iterators.
>
> The uses of device_for_each_child_node() with no early exits have been
> left untouched because their simpilicty justifies the non-scoped
> variant.
>
> Note that the child node is now declared in the macro, and therefore the
> explicit declaration is no longer required.
>
> The general functionality should not be affected by this modification.
> If functional changes are found, please report them back as errors.
Thank you for this series. It may now benefit from the next steps:
1) converting to return dev_err_probe() or dev_warn_probe() directly since
the goto had been replaced by direct return, hence saving even more LoCs;
2) dropping or refactoring the complex conditions due to your nice series
being applied.
I'll comment individually on some to show what I meant by the above.
--
With Best Regards,
Andy Shevchenko