On Fri, Apr 14, 2017 at 2:34 PM, Tim Montague <tmontague@xxxxxxx> wrote:
> Hello dtc developers,
>
> My apologies if this is the incorrect place for this, I'm new here.
It is the right place.
>
> With the following device tree, dtc fails to find the /foo/baz node:
> $ cat broken.dts
> /dts-v1/;
> / {
> ref = <&{/foo/baz}>;
> food {
> };
> foo {
> baz {
> };
> };
> };
> $ dtc -I dts -O dtb -o broken.dtb broken.dts
> broken.dtb: ERROR (phandle_references): Reference to non-existent node or label "/foo/baz"
>
> ERROR: Input tree has errors, aborting (use -f to force output)
> $ dtc -v
> Version: DTC 1.4.4-g9067ee4b
>
> The issue is in get_node_by_path, where it recurses into a child node without properly checking that the path strings are equal. This change fixes the issue for me:
>
> diff --git a/livetree.c b/livetree.c
> index 3673de0..aecd278 100644
> --- a/livetree.c
> +++ b/livetree.c
> @@ -478,7 +478,8 @@ struct node *get_node_by_path(struct node *tree, const char *path)
> p = strchr(path, '/');
>
> for_each_child(tree, child) {
> - if (p && strneq(path, child->name, p-path))
> + if (p && (strlen(child->name) == p-path) &&
> + strneq(path, child->name, p-path))
> return get_node_by_path(child, p+1);
> else if (!p && streq(path, child->name))
> return child;
>
> Let me know if there is a better place to submit bug reports, or if you need anything more from me.
Can you send a proper patch for David to apply. The change looks good to me.
Rob
--
To unsubscribe from this list: send the line "unsubscribe devicetree-compiler" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
