[v2,2/2] of: property: Do not link to disabled devices
diff mbox series

Message ID 20200417165442.1856-3-nsaenzjulienne@suse.de
State Superseded
Headers show
Series
  • of: property: fw_devlink misc fixes
Related show

Commit Message

Nicolas Saenz Julienne April 17, 2020, 4:54 p.m. UTC
When creating a consumer/supplier relationship between two devices,
make sure the supplier node is actually active. Otherwise this will
create a link relationship that will never be fulfilled. This, in the
worst case scenario, will hang the system during boot.

Note that, in practice, the fact that a device-tree represented
consumer/supplier relationship isn't fulfilled will not prevent devices
from successfully probing.

Fixes: a3e1d1a7f5fc ("of: property: Add functional dependency link from DT bindings")
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>

---

Changes since v1:
 - Move availability check into the compatible search routine and bail
   if device node disabled

 drivers/of/property.c | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

Comments

Saravana Kannan April 17, 2020, 9:08 p.m. UTC | #1
On Fri, Apr 17, 2020 at 9:54 AM Nicolas Saenz Julienne
<nsaenzjulienne@suse.de> wrote:
>
> When creating a consumer/supplier relationship between two devices,
> make sure the supplier node is actually active. Otherwise this will
> create a link relationship that will never be fulfilled. This, in the
> worst case scenario, will hang the system during boot.
>
> Note that, in practice, the fact that a device-tree represented
> consumer/supplier relationship isn't fulfilled will not prevent devices
> from successfully probing.
>
> Fixes: a3e1d1a7f5fc ("of: property: Add functional dependency link from DT bindings")
> Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
>
> ---
>
> Changes since v1:
>  - Move availability check into the compatible search routine and bail
>    if device node disabled
>
>  drivers/of/property.c | 19 ++++++++++++++++++-
>  1 file changed, 18 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/of/property.c b/drivers/of/property.c
> index dc034eb45defd..14b6266dd054b 100644
> --- a/drivers/of/property.c
> +++ b/drivers/of/property.c
> @@ -1045,8 +1045,25 @@ static int of_link_to_phandle(struct device *dev, struct device_node *sup_np,
>          * Find the device node that contains the supplier phandle.  It may be
>          * @sup_np or it may be an ancestor of @sup_np.
>          */
> -       while (sup_np && !of_find_property(sup_np, "compatible", NULL))
> +       while (sup_np) {
> +
> +               /*
> +                * Don't allow linking a device node as consumer of a disabled
> +                * node.
> +                */

Minor nit: I'd just say "Don't allow linking to a disabled supplier".

> +               if (!of_device_is_available(sup_np)) {
> +                       dev_dbg(dev, "Not linking to %pOFP - Not available\n",
> +                               sup_np);
> +                       of_node_put(sup_np);
> +                       return -ENODEV;
> +               }

This if block looks very similar to the one right after the loop.
Maybe there's a nice way to combine it?

If you replace this if block with this, it'll end up with the same result.
if (!of_device_is_available(sup_np)) {
        of_node_put(sup_np);
        sup_np = NULL;
}

of_get_next_parent() handles a NULL input properly. So that won't be a
problem. And "No device" is a valid statement for both cases I think.

> +
> +               if (of_find_property(sup_np, "compatible", NULL))
> +                       break;
> +
>                 sup_np = of_get_next_parent(sup_np);
> +       }
> +
>         if (!sup_np) {
>                 dev_dbg(dev, "Not linking to %pOFP - No device\n", tmp_np);
>                 return -ENODEV;

However, not against this patch as is if Rob/Frank like it as is.

-Saravana
Nicolas Saenz Julienne April 18, 2020, 9:20 a.m. UTC | #2
On Fri, 2020-04-17 at 14:08 -0700, Saravana Kannan wrote:
> On Fri, Apr 17, 2020 at 9:54 AM Nicolas Saenz Julienne
> <nsaenzjulienne@suse.de> wrote:
> > When creating a consumer/supplier relationship between two devices,
> > make sure the supplier node is actually active. Otherwise this will
> > create a link relationship that will never be fulfilled. This, in the
> > worst case scenario, will hang the system during boot.
> > 
> > Note that, in practice, the fact that a device-tree represented
> > consumer/supplier relationship isn't fulfilled will not prevent devices
> > from successfully probing.
> > 
> > Fixes: a3e1d1a7f5fc ("of: property: Add functional dependency link from DT
> > bindings")
> > Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
> > 
> > ---
> > 
> > Changes since v1:
> >  - Move availability check into the compatible search routine and bail
> >    if device node disabled
> > 
> >  drivers/of/property.c | 19 ++++++++++++++++++-
> >  1 file changed, 18 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/of/property.c b/drivers/of/property.c
> > index dc034eb45defd..14b6266dd054b 100644
> > --- a/drivers/of/property.c
> > +++ b/drivers/of/property.c
> > @@ -1045,8 +1045,25 @@ static int of_link_to_phandle(struct device *dev,
> > struct device_node *sup_np,
> >          * Find the device node that contains the supplier phandle.  It may
> > be
> >          * @sup_np or it may be an ancestor of @sup_np.
> >          */
> > -       while (sup_np && !of_find_property(sup_np, "compatible", NULL))
> > +       while (sup_np) {
> > +
> > +               /*
> > +                * Don't allow linking a device node as consumer of a
> > disabled
> > +                * node.
> > +                */
> 
> Minor nit: I'd just say "Don't allow linking to a disabled supplier".
> 
> > +               if (!of_device_is_available(sup_np)) {
> > +                       dev_dbg(dev, "Not linking to %pOFP - Not
> > available\n",
> > +                               sup_np);
> > +                       of_node_put(sup_np);
> > +                       return -ENODEV;
> > +               }
> 
> This if block looks very similar to the one right after the loop.
> Maybe there's a nice way to combine it?
> 
> If you replace this if block with this, it'll end up with the same result.
> if (!of_device_is_available(sup_np)) {
>         of_node_put(sup_np);
>         sup_np = NULL;
> }
> 
> of_get_next_parent() handles a NULL input properly. So that won't be a
> problem. And "No device" is a valid statement for both cases I think.
> 
> > +
> > +               if (of_find_property(sup_np, "compatible", NULL))
> > +                       break;
> > +
> >                 sup_np = of_get_next_parent(sup_np);
> > +       }
> > +
> >         if (!sup_np) {
> >                 dev_dbg(dev, "Not linking to %pOFP - No device\n", tmp_np);
> >                 return -ENODEV;
> 
> However, not against this patch as is if Rob/Frank like it as is.

Agree with your suggestions, I'll send an v3.

Regards,
Nicolas

Patch
diff mbox series

diff --git a/drivers/of/property.c b/drivers/of/property.c
index dc034eb45defd..14b6266dd054b 100644
--- a/drivers/of/property.c
+++ b/drivers/of/property.c
@@ -1045,8 +1045,25 @@  static int of_link_to_phandle(struct device *dev, struct device_node *sup_np,
 	 * Find the device node that contains the supplier phandle.  It may be
 	 * @sup_np or it may be an ancestor of @sup_np.
 	 */
-	while (sup_np && !of_find_property(sup_np, "compatible", NULL))
+	while (sup_np) {
+
+		/*
+		 * Don't allow linking a device node as consumer of a disabled
+		 * node.
+		 */
+		if (!of_device_is_available(sup_np)) {
+			dev_dbg(dev, "Not linking to %pOFP - Not available\n",
+				sup_np);
+			of_node_put(sup_np);
+			return -ENODEV;
+		}
+
+		if (of_find_property(sup_np, "compatible", NULL))
+			break;
+
 		sup_np = of_get_next_parent(sup_np);
+	}
+
 	if (!sup_np) {
 		dev_dbg(dev, "Not linking to %pOFP - No device\n", tmp_np);
 		return -ENODEV;