On Mon, 05 Aug 2019 12:13:18 -0500, Larry Finger said: > > The kernel patches causing the issue are a set of 7 a in the linux-next tree > > and targeting the 5.4 kernel . As such, Cyrax's patch is incorrect, for > > production > > it should be comparing >= 5.4.0. It will however DTRT on a linux-next > > kernel > > that calls itself 5.3.something, but will *not* work on a mainline Linus > > tree. > > What I do during the merge of a new kernel, before -rc1 is released is to > manually edit the Makefile to increment the kernel version and make the > EXTRA_VERSION be -rc0. That way the conditional statements work with mainline.
You missed the point - if you apply the patch as it stands to a Linus 5.3.0-rc3 tree, it won't compile correctly, because it will pick up the wrong side of the ifdef. That's why I said "in production" - if Cyrax's patch goes into the virtualbox source tree, it needs to compare against 5.4.0 not 5.3.0. Anybody running linux-next or monkeying around with the innards of the Makefile is on their own on this one because they should have enough clue to hand-correct the patch to suit their actual tree. :) (And it's even more fun doing a git bisect that flips back and forth across version labels, because some of the time LINUX_VERSION says 5.1 and has the old API and other times it says 5.1 and has the new API, or you're looking at a 5.1 that has the new API and the next bisect says 5.2 but still has the old API... That's the sort of thing that causes programmers to seek strong beverages. :)
pgpmH7jrMfiij.pgp
Description: PGP signature
_______________________________________________ vbox-dev mailing list vbox-dev@virtualbox.org https://www.virtualbox.org/mailman/listinfo/vbox-dev
