On Mon, 26 May 2025 at 02:21, <cem@xxxxxxxxxx> wrote:
>
> This is the first time I'm sending a pull-request with a conflict, let me know
> if you need any extra information beyond the one I added.
No, this is plenty fine. The main thing is the heads-up, which I want
mainly just as a sign that maintainers are aware of - and are tracking
- what gets reported in linux-next.
But also partly because it can then make me take those conflicts into
account when planning pull requests (ie I might end up delaying merge
conflicts that look complicated until I have more time to look into
them, although honestly, that kind of complex merge conflict is _very_
rare in practice).
So typically all I really want is just a "you'll get a conflict in
xyz" heads-up, although I do appreciate any details on what happened
and what the resolution is.
That said, I do note that your diffstat is also very messy. That's
"normal" when you have back-merges, and it's not some kind of fatal
thing, but when it happens I do tend to like seeing the result of a
test-merge.
IOW, those messy pull request diffstats only happen when you have
non-linear development, but that also indicates that you know how to
do merges.
So if you notice that your pull request diffstat is messy, and
particularly if you did a test-merge anyway due to known conflicts,
I'd suggest replacing the automatically generated diffstat with the
one generated after you've done the test merge.
In this case, the real diff summary (modulo differences in diff
algorithms) for your merge was
48 files changed, 1521 insertions(+), 188 deletions(-)
but your pull request ends up containing the diffstat for this horror:
1621 files changed, 14176 insertions(+), 7549 deletions(-)
and that is generally a good indication that "Oh, I had a back-merge,
so my diffstat shows the diff against some random upstream state that
contained lots of other changes too".
Linus
