consider discarding the linux-patch-inittmpfs

[dkg]

given that the nested approach is working well with stock debian kernels, is it worthwhile continuing to maintain the inittmpfs kernel patch?

[jrollins]

I was wondering this myself the other day. I don't know, though. I can see that it might be nice to have this package around, though, in case one did want to apply the patch for some reason (maybe not necessarily debirf related). i don't see that it needs to be associated with debirf, though.

Debirf includes fs/usr/bin/build-debirf-kernel. This is no longer necessary, so it should be removed from debirf. It might be nice to keep it's functionality somewhere else, though.

[dkg]

I'm glad to see you already moved build-debirf-kernel over to the linux-patch-inittmpfs source tree. I'd prefer not to get rid of it until we complete some benchmarks about the difference in memory consumption between nested and stupid_simple initramfs strategies. it would be great to not need the kernel hacking part of this project.