darwin/unicast-extensions.diff

	This patch is new as of 2021-07-30 and is very experimental (not
	yet actually tested as part of a Darwin kernel).  It will be
	tested informally by someone Dave knows.

	This patch is against xnu-7195.121.3.

freebsd/freebsd_e.diff

	This patch has not been merged as of 2020-09-25.

	It should probably be considered superseded by
	https://reviews.freebsd.org/D19316 (which adds the same
	functionality in a different way).

freebsd/in_zeronet.patch

	This patch has not been merged as of 2020-09-25.

	It should probably be considered superseded by
	https://reviews.freebsd.org/D19316 (which adds the same
	functionality in a different way).

isdn-utils/isdn4k-utils-3.27_20151118.patch

	I'm not sure what would be considered the upstream source for this
	package anymore.  There are a few related projects that appear to
	be almost abandoned.  I could not access the CVS server mentioned
	at https://www.isdn4linux.de/cvs.shtml.

	I think the official upstream might be

	git://git.misdn.eu/isdn4k-utils.git

	which isn't explicitly mentioned on the corresponding web site,
	and which I deduced by analogy to the CVS page.

	There's also a fork of the source at

	https://github.com/marschap/isdn4k-utils

	This patch has not been merged on either of these trees as of
	2020-09-28.  As noted below, the affected code originally came
	from pppd.

iproute2/0001-Use-SCOPE_HOST-default-only-for-127-16-not-127-8.patch

	This patch has not been merged as of 2021-01-08.

	This patch is meant to be used in conjunction with a Linux kernel
	patch.  It has been tested successfully.

linux/0001-Reduce-local-loopback-network-to-16-updated.patch

	This patch has not been merged as of 2021-01-08.  It is a more
	comprehensive version of 0002-Reduce-localhost-to-a-16.patch.

	In order to make the loopback network actually shrink to /16, this patch
	requires userspace changes in iproute2 and systemd.

linux/0001-Reduce-multicast-range.patch

	This patch has not been merged as of 2020-09-25.  It is a narrower
	part of the functionality of
	0001-Allow-0.0.0.0-8-and-reduce-localnet-and-enable-225-2.patch.

linux/0002-Reduce-localhost-to-a-16.patch

	This patch has not been merged as of 2020-09-25.  It is a narrower
	part of the functionality of
	0001-Allow-0.0.0.0-8-and-reduce-localnet-and-enable-225-2.patch.
	It has been superseded by 0001-Reduce-local-loopback-network-to-16-updated.patch.

netbsd/Allow-with-sysctl.patch

	This is a version of the other patch that puts this functionality
	behind a pair of sysctls (disabled by default).

netbsd/Allow-zero-and-240-nets.patch

	This patch is new as of 2023-05-26, and allows NetBSD to use 0/8
	and 240/4.

openwrt/0001-Reduce-loopback-netmask-netmask-to-16.patch

	This patch is new as of May 3, 2021, and has not been merged yet.  It
	needs testing to double-check that it has the expected effect in
	combination with a patched kernel.

	This is the equivalent in functionality for OpenWrt of the systemd
	patch

	systemd/0001-Change-localhost-prefix-from-8-to-16.patch

ppp/ppp-2.4.7.patch

	This patch has not been merged as of 2020-09-28.  The affected code
	is apparently the origin of the code in isdn-utils (not that
	isdn-utils is likely to be updating its PPP implementation to match
	the original upstream pppd).

	The author of this code is my former Linuxcare coworker.

pureftpd/pure-ftpd-1.0.46.patch

	This patch has not been merged as of 2020-09-28.

	The upstream developer independently made a related change that
	allows the software to use class E addresses, though still not
	class D addresses, on November 21, 2019.

	https://github.com/jedisct1/pure-ftpd/commit/2d7bff38b7b82ba49b3bf73b393c7e43a6e27406

routing/0001-Allow-Class-E-addresses-throughout-FRR.patch

	This patch has not been merged as of 2020-09-25.  The upstream project
	is looking for revisions or actions from us.  See

	https://github.com/FRRouting/frr/pull/3462

routing/0001-bird-treat-former-classe-addresses-as-valid-addresse.patch

	This patch has not been merged as of 2020-09-25.  I'm not sure
	whether it was submitted to the upstream project or not.  It is
	superseded by
	0001-Don-t-treat-0-8-and-240-4-specially-in-IPv4-classifi.patch.

routing/0001-Don-t-treat-0-8-and-240-4-specially-in-IPv4-classifi.patch

	This is an updated version of the former patch.  It was submitted
	upstream to bird on 2022-09-30 via a method suggested on their
	GitLab site that turns out not to work, and resubmited 2022-11-18
	via the project mailing list.

	The overall goal of improving routing of 240/4 and 0/8, either
	with this patch or a variant approach, was not received positively
	by a BIRD developer who reviewed the patch.  The patch has not
	been merged as of 2022-12-25.

routing/0001-martians_check-remove-all-but-0.0.0.0-32-from-0.0.0..patch

	This is apparently meant to be applied to babeld (the name doesn't
	make that explicit).  This patch has not been merged as of
	2020-09-28.  See, e.g.,

	https://github.com/jech/babeld/blob/master/util.c

systemd/0001-Change-localhost-prefix-from-8-to-16.patch

	This patch has not been merged as of 2021-01-08.

	This patch is meant to be used in conjunction with a Linux kernel
	patch.  In combination with that patch, it lets a Linux system
	come up with 127.0.0.1/16 configured on loopback at boot time, and
	other 127/8 address available for regular unicast routing.
