Re: [PATCH] soc/tegra: add of_node_put()

[Frank Lee]

On Wed, Nov 21, 2018 at 10:41 PM Jon Hunter <jonathanh@xxxxxxxxxx> wrote:
>
>
> On 21/11/2018 13:31, Jon Hunter wrote:
> >
> > On 21/11/2018 12:49, Yangtao Li wrote:
> >> of_find_node_by_path() acquires a reference to the node
> >> returned by it and that reference needs to be dropped by its caller.
> >> bl_idle_init() doesn't do that, so fix it.
>
> The bl_idle_init() looks like a typo here.
Hi Jon:

Thanks. :-)

Yous,
    Yangtao
>
> >>
> >> Signed-off-by: Yangtao Li <tiny.windzz@xxxxxxxxx>
> >> ---
> >>  drivers/soc/tegra/common.c | 5 ++++-
> >>  1 file changed, 4 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/soc/tegra/common.c b/drivers/soc/tegra/common.c
> >> index cd8f41351add..2e30604c9658 100644
> >> --- a/drivers/soc/tegra/common.c
> >> +++ b/drivers/soc/tegra/common.c
> >> @@ -23,10 +23,13 @@ static const struct of_device_id tegra_machine_match[] = {
> >>  bool soc_is_tegra(void)
> >>  {
> >>      struct device_node *root;
> >> +    bool rc;
> >>
> >>      root = of_find_node_by_path("/");
> >>      if (!root)
> >>              return false;
> >>
> >> -    return of_match_node(tegra_machine_match, root) != NULL;
> >> +    rc = of_match_node(tegra_machine_match, root) != NULL;
> >> +    of_node_put(root);
> >> +    return rc;
> >>  }
> >
> > Thanks for the fix, however, I think it would be even simpler just to
> > call of_machine_is_compatible() here instead.
>
> Right so I was wrong above about using of_machine_is_compatible, so with
> the typo above fixed ...
>
> Acked-by: Jon Hunter <jonathanh@xxxxxxxxxx>
>
> Cheers
> Jon
>
> --
> nvpublic