On Thu, Mar 24, 2022 at 06:35:49PM -0700, Jakub Kicinski wrote:
On Fri, 25 Mar 2022 09:29:51 +0800 shenjian (K) wrote:
在 2022/3/25 9:03, Jakub Kicinski 写道:
I see you mention that the work is not complete in the cover letter. Either way this patch seems unnecessary, you can call the helpers for "active" features like you do, but don't start by renaming the existing field. The patch will be enormous. .
I agree that this patch will be enormous, I made this patch from suggestion from Andrew Lunn in RFCv3.[1] Willit make people confused for help name inconsistent with feature name ?
Thanks, not sure if I see a suggestion there from Andrew or just a question. Maybe you can add a comment instead to avoid surprising people?
Goes and looks at what i wrote.
Here _hw_ makes sense. But i think we need some sort of consistency. Either drop the _active_ from the function name, or rename the netdev field active_features.
So i suggested an either/or. In retrospect, the or seems like a bad idea, this patch will be enormous. So i would suggest the other option, netdev_set_active_features() gets renamed to netdev_set__features()
Andrew