| Message ID | 1547372691-28324-2-git-send-email-Julia.Lawall@lip6.fr (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | add missing of_node_put | expand |
Hi Julia, Thank you for the patch, On 13/01/2019 09:44, Julia Lawall wrote: > Add an of_node_put when the result of of_graph_get_remote_port_parent is > not available. > > The semantic match that finds this problem is as follows > (http://coccinelle.lip6.fr): > > // <smpl> > @r exists@ > local idexpression e; > expression x; > @@ > e = of_graph_get_remote_port_parent(...); > ... when != x = e > when != true e == NULL > when != of_node_put(e) > when != of_fwnode_handle(e) > ( > return e; > | > *return ...; > ) > // </smpl> > For this part at least: Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> But a little discussion below: > Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> This does indeed look like a missing of_node_put() but I don't think it will be the only one in that code. It gets a bit tangled - and I think there's possibly another missing put on the error path of if (!encoder) { dev_warn(rcdu->dev, "no encoder found for endpoint %pOF, skipping\n", ep->local_node); return -ENODEV; } because that implies that no encoder was found, and thus - encoder = entity; was not executed, and the tail of that function calls: of_node_put(encoder); which would have been the final of_node_put(entity) I think that part could be considered separate to this patch, but it's also quite closely related. Either way - I think it needs a second pair of eyes to see if I'm not crazy :) Regards Kieran > > --- > drivers/gpu/drm/rcar-du/rcar_du_kms.c | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c > index 9c7007d..bc3fcb3 100644 > --- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c > +++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c > @@ -331,6 +331,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, > dev_dbg(rcdu->dev, > "connected entity %pOF is disabled, skipping\n", > entity); > + of_node_put(entity); > return -ENODEV; > } > >
On Mon, 14 Jan 2019, Kieran Bingham wrote: > Hi Julia, > > Thank you for the patch, > > On 13/01/2019 09:44, Julia Lawall wrote: > > Add an of_node_put when the result of of_graph_get_remote_port_parent is > > not available. > > > > The semantic match that finds this problem is as follows > > (http://coccinelle.lip6.fr): > > > > // <smpl> > > @r exists@ > > local idexpression e; > > expression x; > > @@ > > e = of_graph_get_remote_port_parent(...); > > ... when != x = e > > when != true e == NULL > > when != of_node_put(e) > > when != of_fwnode_handle(e) > > ( > > return e; > > | > > *return ...; > > ) > > // </smpl> > > > > For this part at least: > > Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> > > But a little discussion below: > > > > Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> > > This does indeed look like a missing of_node_put() but I don't think it > will be the only one in that code. It gets a bit tangled - and I think > there's possibly another missing put on the error path of > > if (!encoder) { > dev_warn(rcdu->dev, > "no encoder found for endpoint %pOF, skipping\n", > ep->local_node); > return -ENODEV; > } > > because that implies that no encoder was found, and thus - > encoder = entity; > > was not executed, and the tail of that function calls: > > of_node_put(encoder); > > which would have been the final of_node_put(entity) Looking at it again, I agree. I can just extend this patch. julia > > > I think that part could be considered separate to this patch, but it's > also quite closely related. Either way - I think it needs a second pair > of eyes to see if I'm not crazy :) > > Regards > > Kieran > > > > > --- > > drivers/gpu/drm/rcar-du/rcar_du_kms.c | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c > > index 9c7007d..bc3fcb3 100644 > > --- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c > > +++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c > > @@ -331,6 +331,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, > > dev_dbg(rcdu->dev, > > "connected entity %pOF is disabled, skipping\n", > > entity); > > + of_node_put(entity); > > return -ENODEV; > > } > > > > > >
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c index 9c7007d..bc3fcb3 100644 --- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c +++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c @@ -331,6 +331,7 @@ static int rcar_du_encoders_init_one(struct rcar_du_device *rcdu, dev_dbg(rcdu->dev, "connected entity %pOF is disabled, skipping\n", entity); + of_node_put(entity); return -ENODEV; }
Add an of_node_put when the result of of_graph_get_remote_port_parent is not available. The semantic match that finds this problem is as follows (http://coccinelle.lip6.fr): // <smpl> @r exists@ local idexpression e; expression x; @@ e = of_graph_get_remote_port_parent(...); ... when != x = e when != true e == NULL when != of_node_put(e) when != of_fwnode_handle(e) ( return e; | *return ...; ) // </smpl> Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> --- drivers/gpu/drm/rcar-du/rcar_du_kms.c | 1 + 1 file changed, 1 insertion(+)