This patch series is to fix 2 OF device node refcount leakage issues.

Signed-off-by: Zijun Hu <quic_zijuhu@quicinc.com>
---
Zijun Hu (2):
      net: pse-pd: tps23881: Fix device node refcount leakage in tps23881_get_of_channels()
      net: bcmasp: Fix device node refcount leakage in bcmasp_probe()

 drivers/net/ethernet/broadcom/asp2/bcmasp.c | 1 +
 drivers/net/pse-pd/tps23881.c               | 1 +
 2 files changed, 2 insertions(+)
---
base-commit: ff7afaeca1a15fbeaa2c4795ee806c0667bd77b2
change-id: 20241206-drivers_fix-735f56ed5ec7

Best regards,
-- 
Zijun Hu <quic_zijuhu@quicinc.com>

On Thu, 12 Dec 2024 23:06:53 +0800
Zijun Hu <zijun_hu@icloud.com> wrote:

> This patch series is to fix 2 OF device node refcount leakage issues.
> 
> Signed-off-by: Zijun Hu <quic_zijuhu@quicinc.com>
> ---
> Zijun Hu (2):
>       net: pse-pd: tps23881: Fix device node refcount leakage in
> tps23881_get_of_channels() net: bcmasp: Fix device node refcount leakage in
> bcmasp_probe()

Thanks for the patch. This fix was already sent by Zhang Zekun:
https://lore.kernel.org/netdev/20241024015909.58654-1-zhangzekun11@huawei.com/

net maintainers would prefer to have the API changed as calling of_node_get
before of_find_node_by_name is not intuitive.

Still, don't know if we should fix it until the API is changed?  

Regards,
-- 
Köry Maincent, Bootlin
Embedded Linux and kernel engineering
https://bootlin.com

On 2024/12/12 23:33, Kory Maincent wrote:
> On Thu, 12 Dec 2024 23:06:53 +0800
> Zijun Hu <zijun_hu@icloud.com> wrote:
> 
>> This patch series is to fix 2 OF device node refcount leakage issues.
>>
>> Signed-off-by: Zijun Hu <quic_zijuhu@quicinc.com>
>> ---
>> Zijun Hu (2):
>>       net: pse-pd: tps23881: Fix device node refcount leakage in
>> tps23881_get_of_channels() net: bcmasp: Fix device node refcount leakage in
>> bcmasp_probe()
> 
> Thanks for the patch. This fix was already sent by Zhang Zekun:
> https://lore.kernel.org/netdev/20241024015909.58654-1-zhangzekun11@huawei.com/
> 
thank you for sharing this info.

> net maintainers would prefer to have the API changed as calling of_node_get
> before of_find_node_by_name is not intuitive.
> 

agree.

> Still, don't know if we should fix it until the API is changed?  
> 

(^^)
> Regards,