On Mon, Sep 02, 2019 at 03:26:27PM +0300, Andy Shevchenko wrote:
> On Mon, Sep 02, 2019 at 11:32:40AM +0300, Sakari Ailus wrote:
> > Add a test for the %pfw printk modifier using software nodes.
>
> > + const struct software_node softnodes[] = {
> > + { .name = "first", },
> > + { .name = "second", .parent = &softnodes[0], },
> > + { .name = "third", .parent = &softnodes[1], },
> > + { NULL /* Guardian */ },
> > + };
>
> > + test(full_name_second, "%pfw",
> > + software_node_fwnode(&softnodes[ARRAY_SIZE(softnodes) - 3]));
> > + test(full_name, "%pfw",
> > + software_node_fwnode(&softnodes[ARRAY_SIZE(softnodes) - 2]));
> > + test(full_name, "%pfwf",
> > + software_node_fwnode(&softnodes[ARRAY_SIZE(softnodes) - 2]));
> > + test(second_name, "%pfwP",
> > + software_node_fwnode(&softnodes[ARRAY_SIZE(softnodes) - 3]));
> > + test(third_name, "%pfwP",
> > + software_node_fwnode(&softnodes[ARRAY_SIZE(softnodes) - 2]));
>
>
> These can be shorted and easier to parse if you use absolute indexes.
The above doesn't end up accessing out-of-bounds memory without compiler
errors or warnings if the array is changed, therefore I'd prefer to keep it
as-is.
But I'll remove the comma from the guardian entry for v5. :-)
--
Sakari Ailus
[email protected]
