While investigating a problem building sbcl (which stopped building out of the blue after make was updated to 4.4, but it took me a while to find this out), I found a regression in make 4.4 for which I opened a bug report here:

https://savannah.gnu.org/bugs/?63552

I'm mentioning it in case some packages started mysteriously failing to build (or did not build properly) after make was updated to v4.4.