Conflicts:
	auto/config
	config/amnesia
	config/chroot_local-hooks/51-module-assistant
	config/chroot_local-includes/etc/default/kexec
	config/chroot_local-includes/etc/memlockd.cfg

While probably suboptimal, doing like so ensures the tails-reconfigure-memlockd
initscript does not start memlockd too early, i.e. before other needed parts are
in place.

Avoid breaking the build by adding too much scripting to /etc/default/kexec,
that is sourced by the kexec-tools.config maintainer script.

When this happens (i.e. the boot medium was removed), we don't want the shell to
need to look everywhere in the $PATH for a place where the eject command can be
found: directories in the $PATH are probably not listable anymore.

This script has too many dependencies we should list and have cached by
memlockd. Rather kill the main gdm3 process, and move to first virtual terminal.

These symlinks are now broken.

Since we don't ship those in /boot anymore, the symlinks in / are broken.
Moreover, once we ship several pairs of kernel/initrd, the running pair
must be cached (and is already used by our kexec configuration).

Our upcoming memlockd reconfiguration initscript will need this as well.

We'll soon ship several kernels for PAE / NX bit support: 486, 686-pae, amd64.

Therefore, if the system was booted on a 64-bit kernel, plenty of memory may
have been used, that need to be wiped on shutdown; in such a situation, the
memory wipe kexec'd kernel must be 64-bit too: a 32-bit kernel may be unable to
access all memory that may have been used during the runtime of Tails.

Those files are already shipped where they are needed, that is in the ISO
filesystem.

This reverts commit 73397e5e.

The Torbutton "New identity" feature does not support ControlSocket yet:
https://trac.torproject.org/projects/tor/ticket/3967

This saves 3MB on the final ISO (SquashFS compressed with XZ).

The runtime performance hit is probably very small, and I found it pretty
convincing that Ubuntu decided to take it in their Live systems.

This saves 6MB on the final ISO (with XZ SquashFS compression),
and changes nothing I could notice at runtime.

It's not as good as if Claws Mail supported SOCKS proxies itself,
but still better than relying on the transparent netfilter torification.

It hides the explanation and warning about sdmem we want to display.

plymouth.postinst creates links using update-rc.d, so the links it creates
cannot be disabled using LSB headers.

We only want hdparm so that laptop-mode-tools can use it.

No commit message

No commit message

No commit message