Re: [PATCH] resource: handle wrong resource_size value on zero start/end resource

On Mon, Dec 08, 2025 at 01:57:37AM +0200, Andy Shevchenko wrote:

On Mon, Dec 08, 2025 at 12:33:03AM +0100, Christian Marangi wrote:

...

I think the better choice is to check all places where resource is assigned and convert that to a helper when the size can be or is 0. Definitely it's a lot of code to be audited.

But having something like

resource_set_size(res, 0) // note, this API is already in upstream or resource_set_range(res, start, 0)

instead of direct assignment of start and end is much simpler approach.

Just run

git grep -n -w resource_set_range

and you see that even OF uses it in once case already. And PCI core is full of the calls, that's why I believe PCI is not affected as it does the correct thing to begin with.