Age | Commit message (Collapse) | Author | Files |
|
Arch Linux has unified /sbin and /usr/sbin into a single binaries
directory: /usr/bin
|
|
This repository has no official connection to Arch Linux, is not
endorsed by any Arch Linux developers, and the project is not
responsible for any unbootable systems that result from it.
|
|
|
|
|
|
|
|
Interpret exscape sequences in the filessytem target while unmounting.
Having recursive unmount would be probably better (see falconindy's
comment on the bug), but IMO this workaround is good enough.
[tomegun: recursive umount will almost certainly be in next util-linux,
but that is still some time away, so we take this for now.]
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Make it clear that rc.conf is the config file for initscripts.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
This should hopefully help against people accidentally both using legacy
units and native units for the same service at the same time.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
The user should pick what daemons to use, no need to provide default
ones.
In particular as this is also used by systemd, where the default units
no longer make sense.
Suggested-by: Thomas Bächler <thomas@archlinux.org>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
We now no longer support reading the MODULES array when booting
with systemd. The support is kept in initscripts (though a warning
is printed) for sake of compatibility.
Both initscripts and systemd should be using the new modules-load.d
syntax and we should remove all mention of the old from varios
documentation, for the sake of simplicity.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
This reflects the mainline PKGBUILD. Depend on systemd, rather than systemd-tools
and only optdepend on sysvinit.
|
|
No idea how I did not spot this obvious bug.
[tomegun: note to self, I didn't spot the bug because it is only
a problem when not booting with an initramfs...]
Fixes: FS#31765
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Also properly check for the deprecated use of rc.conf for this purpose,
as we were getting lots of false positives.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Conflicts:
functions
|
|
The new locale.sh makes DAEMON_LOCALE redundant, so we remove that
functionality. A post-upgrade notice will be added.
The new functionality will apply the settings to /etc/locale.conf
system-wide, and allow each user to override it in
$HOME/.config/locale.conf.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
As we are moving to systemd, alert initscripts users in case they use the legacy configuration options.
The legacy options will still work in initscripts, but eventually will not work with systemd. In other
words, most users will want to switch to the new settings.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
The point of the target is to allow it to be masked so DAEMONS will be ignored. Document this.
|
|
|
|
This setting is intended for dynamically created units.
Signed-off-by: Jan Alexander Steffens (heftig) <jan.steffens@gmail.com>
|
|
This (re)introduces arch-daemons.target, so a simple
systemctl mask arch-daemons.target
will suppress the starting of DAEMONS from rc.conf.
Signed-off-by: Jan Alexander Steffens (heftig) <jan.steffens@gmail.com>
|
|
Reported-by: Thomas Weißschuh <thomas_weissschuh@lavabit.com>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
|
|
|
|
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Abstract the unquoting logic from parse_envfile into unquote() and reuse
it for our crypttab unlock logic.
Signed-off-by: Dave Reisner <dreisner@archlinux.org>
|
|
As per crypttab(5), this is supported. This also adds forward looking support
for PARTUUID, which is valid and should "just work" for GPT partitioned disks.
Signed-off-by: Dave Reisner <dreisner@archlinux.org>
|
|
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Rather than pointing to the documentation online, include the
documentation here.
The main point with the documentation is to point out that these
things should not be used.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
The compat layer might attempt to start this, but that is never
necessary, as systemd deals with remote-fs internally.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Unify this as much as possible, write out MODULES to a
modules-load.d file in /run before calling systemd-modules-load,
rather than using modprobe directly.
This should not make any difference to the functionality.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Reported-by: Ali Kaasinen <ali.kaasinen@gmail.com>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Signed-off-by: Dave Reisner <dreisner@archlinux.org>
|
|
This is a bit different from the other instructions, so make it extra
clear.
Reported-by: Myra Nelson <myra.nelson@hughes.net>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
The check was inverted. Just do this unconditionally.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
This never worked.
Reported-by: James Gordon <drwitherstone@gmail.com>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Reported-by: Oded
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Reported-by: Eric Fernandez <zebcom@zebulon.org.uk>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Fixes FS#30891
Signed-off-by: Eric Bélanger <snowmaniscool@gmail.com>
|
|
Don't create wrapper services, for things that are not exectuable files.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
This makes sure that systemd supports some initscripts API's. With this
patch, systemd will:
* Parse and use DAEMONS and MODULES from rc.conf
* Run rc.local and rc.local.shutdown on boot and shutdown respectively
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Signed-off-by: Eric Bélanger <snowmaniscool@gmail.com>
|
|
Suggested-by: Pierre Schmitz <pierre@archlinux.org>
Suggested-by: Karol Blazewicz <karol.blazewicz@gmail.com>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
[00:31] <dgbaley27> tomegun: typo rc.conf:5
[00:31] <tomegun> dgbaley27: thanks
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
This outlines what files needs to be configured to get a well-behaved
system.
For now it is extremely terse, but the main point is to work as a
check-list, and to point people to the correct manpages to learn more.
I'll be happy to take patches to extend this in the future.
One might argue that this sholud not be part of initscripts, and we
might indeed move it to 'filesystem' or something like that in the
future. For the time being I'll keep it here for convenience.
Suggested-by: Thomas Bächler <thomas@archlinux.org>
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
Increase the max log level to avoid INFO messages.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|
|
|
|
This should be safe for freshly mapped swap space, especially since we
check that blkid returns 2.
|
|
This is now handled by udev. Leaving it set does no harm.
Signed-off-by: Tom Gundersen <teg@jklm.no>
|