Hi Sean,
I finally got around to posting v3 of this patch series. I wanted to touch on a few issues you had mentioned in previous review.
On Fri, Aug 26, 2022 at 11:47 AM Ralph Siemsen ralph.siemsen@linaro.org wrote:
On Tue, Aug 23, 2022 at 12:14:31AM -0400, Sean Anderson wrote:
Regarding the unused fields (scon, mirack, mistat): I am not really sure what their purpose is. Maybe there is some value in having them. I'll try to find out more information about them. If we do decide to drop them, I would like to keep it synchronised with the Linux driver.
OK, well if you don't use them then perhaps you can just leave them in the macro but remove them from the struct. That way you can add support for them later if you need to, but they don't take up space in the mean time. A comment summarizing your explanation above would be helpful.
So I ended up doing as you suggested: keeping the values in the tables, in case they are needed in the future, but omitting them from the structure to save space.
DIV_ROUND_CLOSEST?
I'm hesitant to change the logic on this, as it could subtly alter the values.
Well if you have 2MHz divided by 3, the resulting rate is closer to 666667 kHz than 666666 Hz.
This could result in the chosen clock rate being higher than what was requested. And in turn that could result in violating the spec of a hardware device.
So I have kept the original logic here.
Regards Ralph