On Tue, May 12, 2026 at 6:31 PM Danilo Krummrich <[email protected]> wrote: > > On Tue May 12, 2026 at 1:59 PM CEST, Bartosz Golaszewski wrote: > > +int platform_device_add_software_node(struct platform_device *pdev, > > + const struct software_node *swnode) > > +{ > > + return device_add_software_node(&pdev->dev, swnode); > > I think this should be device_create_managed_software_node(), see also [1]. > > That said, now that I see the resulting code and the corresponding subleties, > I > think we should reconsider your (well predicted) "temporary > platform_device_release_full()" idea for -fixes, while iterating this for > -next. >
Agreed, I think it's much safer this way. Bart
