Make of_overlay_remove() tolerate ovcs_id being 0 without logging an error.

Suggested-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
---
This is needed for drivers/misc/rp1/rp1_pci.c where things are taken from
DT or at runtime. In the former case, ovcs_id is unused and left to 0.
Being able to tolerate such cases simplify error handling.

This was suggested by Dan as a reply to patch [1].

[1]: https://lore.kernel.org/lkml/4e92a271fdb98560c4e659556a1f3e99e7d0d38e.1760987458.git.christophe.jaillet@wanadoo.fr/

---
 drivers/of/overlay.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
index 255e8362f600..5b4f42230e6c 100644
--- a/drivers/of/overlay.c
+++ b/drivers/of/overlay.c
@@ -1190,6 +1190,9 @@ int of_overlay_remove(int *ovcs_id)
 	struct overlay_changeset *ovcs;
 	int ret, ret_apply, ret_tmp;
 
+	if (*ovcs_id == 0)
+		return 0;
+
 	if (devicetree_corrupt()) {
 		pr_err("suspect devicetree state, refuse to remove overlay\n");
 		ret = -EBUSY;
-- 
2.51.0

On Fri, 24 Oct 2025 21:50:58 +0200, Christophe JAILLET wrote:
> Make of_overlay_remove() tolerate ovcs_id being 0 without logging an error.
> 
> Suggested-by: Dan Carpenter <dan.carpenter@linaro.org>
> Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
> ---
> This is needed for drivers/misc/rp1/rp1_pci.c where things are taken from
> DT or at runtime. In the former case, ovcs_id is unused and left to 0.
> Being able to tolerate such cases simplify error handling.
> 
> This was suggested by Dan as a reply to patch [1].
> 
> [1]: https://lore.kernel.org/lkml/4e92a271fdb98560c4e659556a1f3e99e7d0d38e.1760987458.git.christophe.jaillet@wanadoo.fr/
> 
> ---
>  drivers/of/overlay.c | 3 +++
>  1 file changed, 3 insertions(+)
> 

Applied, thanks!