On Tue, Feb 25, 2020 at 12:44:01PM +0100, Wolfram Sang wrote: > > Motivation is simple: > > - Standardize the (small) set of mostly used bus frequences > > - Get rid of repetition of (subset) of above in many drivers > > - Reduce amount of potential typos > > > > Let's discuss it here. I don't think new version of this would be good to have > > without initial settlement. > > Well, for me, this works. I agree to the typo thing and having less > magic values. It's all OK; I think it is just nice to have some things > in a coverletter. Since it looks like we will have only few patches at the end, I think the commit message for the first one would be good enough to describe this all. > > I aware about that, but I would like to avoid I²C subsystem storming for > > another change like this. So, let's consider this as a trampoline when in the > > future we will switch entire subsystem to Linux wide header at once. > > I can agree to that. > > > > Furthermore, I'd prefer to > > > have 'MAX' in there, e.g. I2C_MAX_STANDARD_MODE_FREQ etc. Just to make > > > clear that I2C can have other bus speeds as well. > > > > Works for me. > > Thanks, that's the most important point to me. > > > Btw, what about Vladimir's comment WRT STANDARD -> STD? My personal opinion > > that STD is a bit too short. > > No real opinion here. I think STD is understandable enough and I > encounter it regularly. However, I also don't think the saving is huge > enough to matter. Your call here. I prefer STANDARD over STD due to consistency (we don't have good abbreviations for Fast, Fast+, High Speed, etc, anyway). > > I'm fine with either. For reviewers it would be better I think to see only > > their portion. Since I got a lot of tags already I consider I may squash it > > together. So, what do you prefer? > > Sounds good to me. Keep collecting acks and squash all patches and tags > in v2. Good, thanks for review! -- With Best Regards, Andy Shevchenko
