On Mon, Jul 4, 2016 at 6:12 PM, Randy Dunlap [off-list ref] wrote:

quoted

I merged it I think, Alex made a long series enabling compile
testing and I started to cherry-pick the first commits to let
them trickle in.

I guess that when you do a git pull of a series of patches, you
sign the pull commit but not each patch in the series?
That could explain it.

I think I cherry-picked it actually. But yeah.

I agree with Arnd that this particular commit should be reverted
(1e4a80640338924b9f9fd7a121ac31d08134410a).

I've reverted it for fixes.

Yours,
Linus Walleij