Re: [PATCH v3 1/1] device property: Allow error pointer to be passed to fwnode APIs

[Andy Shevchenko]

On Tue, Mar 08, 2022 at 09:25:07AM +0000, Sa, Nuno wrote:
> > -----Original Message-----
> > From: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> > Sent: Monday, March 7, 2022 9:30 PM

...

> > v3: dropped test of secondary fwnode (Nuno), added tag (Nuno),
> > amended commit message
> > v2: adjusted the entire fwnode API (Sakari)
> > 
> > Nuno, can you re-test this with the ltc2983 series to be sure it is still
> > okay?
> 
> Still works!

Thanks for confirming!

...

> > @@ -988,14 +998,14 @@ fwnode_graph_get_next_endpoint(const
> > struct fwnode_handle *fwnode,
> >  		parent = fwnode_graph_get_port_parent(prev);
> >  	else
> >  		parent = fwnode;
> > +	if (IS_ERR_OR_NULL(parent))
> > +		return NULL;

(1)

> >  	ep = fwnode_call_ptr_op(parent, graph_get_next_endpoint,
> > prev);
> > +	if (ep)
> > +		return ep;
> 
> I might be missing something but before the check being done was
> 'if (IS_ERR_OR_NULL(ep)'. Is there anyway for ep to be an error
> pointer? Looking at OF, It seems that only NULL or a valid pointer
> is being returned. Did not looked at others implementations of
> though...

Yes, the IS_ERR() part is redundant there. I was quite confused with
that code while working on this change. So, now it looks much clearer
what's going on and what kind of values are being expected. This also
justifies the choice of returned value in (1).

-- 
With Best Regards,
Andy Shevchenko