On Mon, Nov 25, 2024 at 09:48:59AM +0100, Linus Walleij wrote:
On Sun, Nov 24, 2024 at 5:55 PM Sasha Levin <sashal@xxxxxxxxxx> wrote:
I've just hit the issue you've described in this PR:
(...)
Is effectively a revert of one of the commits that are part of this PR:
> pinctrl: aw9523: add missing mutex_destroy
Would it make more sense to just re-do this PR without the offending
commit? I understand that this is a fairly small fixup, but I'm
concerned that this will just create confusion later on...
I don't follow what you mean I should do. The offending commit is a
fix and it is already upstream since -rc4.
Oh, there's something off in the PR itself: it lists "pinctrl: aw9523:
add missing mutex_destroy" as a commit that is included in this PR, but
really it's already upstream.
Sorry, I got confused by that.
Torvalds could probably fix the issue by simply reverting
393c554093c0c4cbc8e2f178d36df169016384da
instead of applying the fixup though, it has the same textual and
semantic effect. I just tested it and it works fine.
^Torvalds: looks like revert on top is a better idea than fixups
so we don't upset the stable maintainer scripts.
Yes, a revert would be nicer as it'll make sure we can easily get it to
older stable trees.
--
Thanks,
Sasha
