On Thu, Jun 06, 2019 at 02:08:50PM -0500, Pierre-Louis Bossart wrote:
Please don't top-post.
On 6/6/19 1:25 PM, Rojewski, Cezary wrote:
Hmm, guess reviewing 001 proved redundant after all. Unless I got it wrong, you are removing code implemented in that very patch (the 001).
Any chance for eliminating ping-pong effect and doing the "right" changes from the get-go? Especially the renames are confusing here (s/cleanup_platform/cleanup_legacy/) if you intend to remove them soon after.
Please fix your mail client to word wrap within paragraphs at something substantially less than 80 columns. Doing this makes your messages much easier to read and reply to.
Using a ping-pong analogy for a 146-patch series is pushing it. It's first make then break to avoid bisect issues. And the names match what is used in the existing code. maybe the naming isn't to your liking but it's what has been used for a while.
Note that the last patch is going to break all the non-upstream machine drivers so you will have quite a bit of work to do on your own when you rebase.
Right, avoiding build breaks is important here - it helps future bisectability if we don't have commits (and especially long serieses of commits) that just randomly fail to build. That's *way* more useful than dropping the initial patch would've been.
If there is no other way around it and solution is accepted, a note, perhaps in 001 would be helpful for future readers.
There's no guarantee that a patch series will be applied at once, if there's a problem part way through the series the earlier bits might get applied to save future review.