I2C: introduce and adapt new I2C interface (2nd attempt)#9548
I2C: introduce and adapt new I2C interface (2nd attempt)#9548
Conversation
|
@cladmi I have a huge problem here EDIT: both links were inverted |
|
My idea would be to try rebasing but forbid git to perform automatic merge, so that you still must verify each commit. |
|
Something that can help finding these, is https://stackoverflow.com/a/40183550/395687 And doing |
|
@cladmi The I2C refactoring is now over and ready to be merged. The decision to add it to the incoming release belongs to you :) |
|
Do you have somewhere the list of tests that was run so they could be re-run in master ? |
|
Question, what are all these "fix attr order" and "cleanup AF" ? are these meant to be squashed after ? I may do a commit by commit diff review on monday. |
|
@cladmi mentionned issues are CPP fixes because Murdock was complaining. |
|
Is there somewhere a summary of the actual API changes? |
|
@kaspar030 You can have a look at #6577 |
@dylad I'm looking for an API change description. There was a PR / issue somewhere where the discussion took place, but I think it is not #6577. |
kaspar030
left a comment
There was a problem hiding this comment.
(blocking so noone accidentally hits the green button).
|
@kaspar030 I think you're talking about #6576 I can open a wiki page for summarize all the changes adopted. |
That would be awesome! |
|
@kaspar030 I'll inform you when this wiki page will be written. I'll start this weekend if I find some time. |
|
@kaspar030 please see here this is the first round of the wiki page. I am not a good writer usually so don't hesitate to correct me or ask for more information. |
|
not blocking this PR, but I just want to say that I'm against merging this PR in the current state, because of the suboptimal changes applied to the STM32F1 boards. |
|
@dylad, for the wiki page, this PR also contains other changes, not directly related to the new API. These are the ones that come to my mind right now:
|
|
I think #9574 should be handle before this one. Then I could rebase this branch on master. |
all comments address, dismiss as suggested
|
@kaspar030 care to approve or dismiss your review as the release branch is in place for some time now. We have lot's of stuff waiting for this to be master. |
|
Alright, let's go. |
|
ACK and GO then! |
|
Woooow! Congratulations to all the people involved! Special mention to @aabadie @dylad and @MrKevinWeiss. Thanks a lot guys! |
|
Yes. Congrats to getting the biggest (I think) API-rework in RIOT ever merged! |
|
Wow, it's finally over ! I'm so relief ! |
|
Congrats and thanks to everyone! |
Contribution description
This PR introduces and adapts to all cpu/board/device drivers, the new I2C interface of RIOT from the feature branch. The branch used is a cloned and rebased version of new_i2c_if because I didn't want to rebase new_i2c_if yet.
All comments/reviews will be much appreciate as the PR is pretty big.
For the 2nd attempt, the new_i2c_if3 used branch is based on new_i2c_if which has been rebase without --preserve-merges option. This option seems to cause some trouble (see discussion #9539)
Issues/PRs references
Tracking issue of the I2C refactoring #6577
Discussion #9539