This fixes a regression introduced in 64e70e4 where the mount fails
when fstab is misconfigured with fs_passno > 0 on a virtual file
system like nfs, and the type is specified as "auto".
This allows the user to disable fsck's by masking.
If fsck fails, emergency target is started, the user might mount the
unit using mount and disable fsck by masking the unit. In this case,
.mount will be active because the mount is detect through
/proc/self/mountinfo, but systemd-fsck@.service will still be in
failed mode. This results in a funny situation where
$ systemctl show -p ActiveState local-fs.target yyy.mount
ActiveState=active
ActiveState=active
$ sudo systemctl start local-fs.target
[sudo] password for test:
Failed to start local-fs.target: Unit systemd-fsck@xxx.service is masked.
If fstab contains 1 for passno, treat this as an error, but only warn
briefly. If fstab doesn't contain this information, don't complain at
all.
Patch is complicated a bit by the fact that we might have the fstype specified
in fstab or on /proc/cmdline, in which case we can check if we have the appropriate
fsck tool, or not specified, or specified as auto, in which case we have to look
and check the type of the filesystem ourselves. It cannot be done before the
device appears, so it is too early in the generator phase, and it must be done
directly in fsck service.
Instead of individually checking for containers in each user do this
once in a new call proc_cmdline() that read the file only if we are not
in a container.
fsck-root is redundant in case an initrd is used, or in case the rootfs
is never remounted 'rw', so the new default is the correct behavior for
most users. For the rest, they should enable it in fstab.
[tomegun: without this we would never fsck the rootfs if it was directly
mounted 'rw' from the initrd. We now risk fsck'ing it twice in the case it
is mounted 'ro', so that should be addressed in a separate patch.]
As we load unit files lazily, we need to make sure something pulls in swap
units that should be started automatically, otherwise the default dependencies
will never be applied.
This partially reinstates code removed in
commit 64347fc2b9.
Also don't order swap devices after swap.target when they are 'nofail'.
This makes mount units work like swap units: when the backing device appears
the mount unit will be started.
v2: the device should want the mount unconditionally, not only for DefaultDependencies=yes
Currently we don't respect noauto/nofail root mount options (from
rootflags kernel cmdline). We should map these two flags to the
corresponding boolean variable noauto and nofail when calling
add_mount().
gcc thinks that errno might be negative, and functions could return
something positive on error (-errno). Should not matter in practice,
but makes an -O4 build much quieter.
x-initrd.mount now has different meanings, if fstab-generator is called
in the initramfs.
initrd:/etc/fstab and x-initrd.mount defines mounts for the
initrd-root-fs.target
initrd:/sysroot/etc/fstab and x-initrd.mount defines mounts for the
initrd-fs.target
This introduces remote-fs-setup.target independently of
remote-fs-pre.target. The former is only for pulling things in, the
latter only for ordering.
The new semantics:
remote-fs-setup.target: is pulled in automatically by all remote mounts.
Shall be used to pull in other units that want to run when at least one
remote mount is set up. Is not ordered against the actual mount units,
in order to allow activation of its dependencies even 'a posteriori',
i.e. when a mount is established outside of systemd and is only picked
up by it.
remote-fs-pre.target: needs to be pulled in automatically by the
implementing service, is otherwise not part of the initial transaction.
This is ordered before all remote mount units.
A service that wants to be pulled in and run before all remote mounts
should hence have:
a) WantedBy=remote-fs-setup.target -- so that it is pulled in
b) Wants=remote-fs-pre.target + Before=remote-fs-pre.target -- so that
it is ordered before the mount point, normally.
This changes the fstab mount option x-initrd-rootfs.mount to
x-initrd.rootfs, in order to only use a single namespace "x-initrd." for
all mount options of the initrd.
First, rename root-fs.target to initrd-root-fs.target to clarify its usage.
Mount units with "x-initrd-rootfs.mount" are now ordered before
initrd-root-fs.target. As we sometimes construct /sysroot mounts in
/etc/fstab in the initrd, we want these to be mounted before the
initrd-root-fs.target is active.
initrd.target can be the default target in the initrd.
(normal startup)
:
:
v
basic.target
|
______________________/|
/ |
| sysroot.mount
| |
| v
| initrd-root-fs.target
| |
| v
| initrd-parse-etc.service
(custom initrd services) |
| v
| (sysroot-usr.mount and
| various mounts marked
| with fstab option
| x-initrd.mount)
| |
| v
| initrd-fs.target
| |
\______________________ |
\|
v
initrd.target
|
v
initrd-cleanup.service
isolates to
initrd-switch-root.target
|
v
______________________/|
/ |
| initrd-udevadm-cleanup-db.service
| |
(custom initrd services) |
| |
\______________________ |
\|
v
initrd-switch-root.target
|
v
initrd-switch-root.service
|
v
switch-root
Instead of using local-fs*.target in the initrd, use root-fs.target for
sysroot.mount and initrd-fs.target for /sysroot/usr and friends.
Using local-fs.target would mean to carry over the activated
local-fs.target to the isolated initrd-switch-root.target and thus in
the real root. Having local-fs.target already active after
deserialization causes ordering problems with the real root services and
targets.
We better isolate to targets for initrd-switch-root.target, which are
only available in the initrd.
We only mount "/usr" and entries marked with "x-initrd.mount".
This (together with the right unit files) is needed in the initramfs in order to
natively support mounting /usr (and friends) from the initramfs.
The way it is meant to work is:
* wait for sysroot.mount to be mounted
* do a daemon-reload to generate sysroot-usr.mount (++) from /sysroot/etc/fstab
* wait for sysroot-usr.mount to be mounted
* switch-root
Cc: Harald Hoyer <harald.hoyer@gmail.com>
Cc: Dave Reisner <d@falconindy.com>
I originally added this to stay as compatible as possible with the kernel, but
as Lennart argued it is not really useful in the initramfs, so let's drop it (we
already don't support 'rootdealy').
The configuration is taken from /proc/cmdline, aiming at emulating the
behavior of the kernel when no initramfs is used.
The supported options are: root=, rootfstype=, rootwait=, rootflags=,
ro, and rw. rootdelay= was dropped, as it is not really useful in a
systemd world, but could easily be added.
v2: fix comments by Lennart, and complain loudly if root= can not be found
Cc: Harald Hoyer <harald@redhat.com>
Cc: Dave Reisner <dreisner@archlinux.org>
Traditional sysvinit systems would not complain about duplicates in
fstab. Rather it (through monut -a) would mount one fs on top of another,
in effect the last entry taking precedent.
In systemd, the first entry takes precedent, all subsequent ones are
ignored and an error is printed.
The change of behavior and the source of this error message was causing
some confusion, so give a hint what migt be wrong.
also a number of minor fixups and bug fixes: spelling, oom errors
that didn't print errors, not properly forwarding error codes,
few more consistency issues, et cetera
glibc/glib both use "out of memory" consistantly so maybe we should
consider that instead of this.
Eliminates one string out of a number of binaries. Also fixes extra newline
in udev/scsi_id
This is a valid source entry in /etc/fstab:
//192.168.6.10/data /data cifs noperm,auto
On Mon, Jun 4, 2012 at 2:04 PM, Dave Reisner <d@falconindy.com> wrote:
> On Mon, Jun 04, 2012 at 12:57:47PM +0200, Kay Sievers wrote:
>>
>> Changed it to use:
>> path_is_absolute()
>> instead of:
>> is_path(),
>> so that we still sanitize the input we might match against.
>>
>> Let me know, if you think that could still cause any problems?
> Yes, this will still break CIFS shares.
This can invalidate otherwise valid source paths with trailing slashes,
such as "host:/" in the case of a network mount.
Based on a patch from Dave Reisner <dreisner@archlinux.org>, which
removed the slash mangling entirely.
Changed it to match on the leading slash to exclude non-path values.
Previously generated units were always placed at the end of the search
path. With this change there will be three unit dirs instead of one, to
place generated entries at the beginning, in the middle and at the end
of the search path:
beginning: for units that need to override all configuration, regardless
of user or vendor. Example use: system-update-generator uses this to
temporarily redirect default.target.
middle: for units that need to override vendor configuration, but not
vendor configuration. Example use: /etc/fstab should override vendor
supplied configuration (think /tmp), but should not override native user
configuration.
end: does not override anything but is available as well. Possible usage
might be to convert D-Bus bus service files to native units but allowing
vendor supplied native units to win.
UnitPath= is also writable via native units and may be used by generators
to clarify from which file a unit is generated. This patch also hooks up
the cryptsetup and fstab generators to set UnitPath= accordingly.
Thomas Bächler