On Thu, Mar 05, 2020 at 09:58:29PM +0000, Peter Rosin wrote: > On 2020-02-27 13:21, Andy Shevchenko wrote: > > There are few maximum bus frequencies being used in the I²C core code. > > Provide generic definitions for bus frequencies and use them in the core. > > > > The drivers may use predefined constants where it is appropriate. > > Some of them are already using these under slightly different names. > > We will convert them later to use newly introduced defines. > > > > These definitions will also help to avoid typos in the numbers that > > may lead to subtle errors. ... > > +/* I2C Frequency Modes */ > > +#define I2C_MAX_STANDARD_MODE_FREQ 100000 > > +#define I2C_MAX_FAST_MODE_FREQ 400000 > > +#define I2C_MAX_FAST_PLUS_MODE_FREQ 1000000 > > +#define I2C_MAX_TURBO_MODE_FREQ 1400000 > > +#define I2C_MAX_HIGH_SPEED_MODE_FREQ 3400000 > > +#define I2C_MAX_ULTRA_SPEED_MODE_FREQ 5000000 > > Am I the only one who do /not/ find these names readable? Seems so :-) > I can't seem to > remember what frequency TURBO, HIGH, etc are and their ordering is > difficult. It's all sounds like marketing buzz to me and my brain shuts > off instantly. > > This feels a lot like moving away from octal permissions for files, which > is frowned upon... > > Can we include some kind of indication of the actual frequency in the > names as well please? > > Perhaps something like I2C_MAX_STANDARD_MODE_100KHZ? It's definitely a duplicate information here. In I²C we have quite established range of frequencies (4 out of 6 in above list), so, their names already implying very well defined frequencies. For the rest 2 perhaps it's good to have something like your proposal, but it will be inconsistent with the rest and with many controller specifications. Wolfram, what do you think about Peter's idea? -- With Best Regards, Andy Shevchenko
