Follow-up for ab73333c43
Methods that take numeric pid values use 0 to denote the peer,
hence let's log about 0 on PIDREF_AUTOMATIC, -1 if truly unset.
When user creates dropin files via API (e.g. systemctl set-property ...)
we put the dropin directory path into unit_path_cache. Drop those
directories from the cache in unit_free() and prevent memory leak.
Follow-up for fce94c5c56.
History has clearly shown that we're terrible at keeping
the size of the dont_close array up-to-date. Hence let's
step away from a hardcoded max size for that, instead
always collect all fds in the array initializer and
let compiler figure it out, taking advantage of the fact
that close_all_fds() gracefully handles invalid fds in
the 'except' array.
Follow-up for 12001b1bf0
It is part of the keep_fds array already, which is intended
for fds that need to survive until the very end of executor
runtime, i.e. across the second close_all_fds() call.
The owned fds should be closed before close_all_fds() along the lines
of exec_params/runtime_close(), otherwise they would be left dangling
afterwards.
After the update to 260-rc1 in Fedora Rawhide, we get a lot of messages like
this:
systemd[1]: systemd-resolved-monitor.socket:
Failed to create symlink /run/systemd/resolve/io.systemd.Resolve.Monitor →
/run/varlink/registry/io.systemd.Resolve.Monitor, ignoring: Permission denied
The actual issue will need to be fixed too, but let's fix the message first.
While at it, let's rename the variables to be meaningful.
With the previous changes the modes for /tmp/ and /var/tmp/
may come in arbitrary directions, i.e. we may downgrade one
and upgrade another. Hence let's refactor the setup logic
a bit so they're truly independent.
As in the previous commit, checking for both requirements and ordering seems
unnecessary. In practical cases, the mount will be pulled in by the rest
of the transaction, so ordering is the part that matters. (The setup is
racy without the ordering.) If we drop the second check, the admin can
just use After=tmp.mount to achieve the desired behaviour, without needing
to explicitly pull in the unit. This is easier to configure and more robust.
This changes the implementation introduced in
6156bec7a4.
Also actually describe the implemented behaviour in the man page.
In https://github.com/systemd/systemd/issues/28515, multiple people report that
services that have PrivateTmp=yes and DefaultDependencies=no fail to create the
temporary directories under /tmp, when /tmp is e.g. a bind mount or some other
kind of mount that takes more time.
Before PrivateTmp=disconnected was added, we didn't have a nice solution:
DefaultDependencies=no is used to start services very early, so we wouldn't
want to add a dependency on /tmp automatically. With PrivateTmp=disconnected we
have a fairly nice solution. Let's "upgrade" to this mode automatically.
Strictly speaking, it is a small compat break, but in practice it's unlikely to
matter for early-boot services whether their /tmp is private or disconnected.
The dependency on /tmp that is checked is After. I think this is enough,
since any tmp.mount would be pulled in by local-fs.target and the rest of
the transaction anyway, so we don't need to check more than After.
The asserts are relaxed, because now the two settings can now diverge
in either way.
Resolves https://github.com/systemd/systemd/issues/28515.
[yhndnzj: fix unit_add_exec_dependencies() to handle the new
combination; add a comment in exec_needs_sys_admin()]
The only reason why ExecSharedRuntime.(var_)tmp_dir might
be suppressed is when (/var)/tmp/ is listed in InaccessiblePaths=,
in which case mount namespace is needed anyways. Moreover,
the current logic imposes a chicken-and-egg problem:
at various places in pid1 we'd need to check whether
the unit runs in a mountns, but the ExecSharedRuntime
may not be allocated for it just yet.
* Make sure we're not passing NULL to rmdir()
* Remove the inner "tmp" subdir as well, so that callers
can handle this transparently (e.g. exec_shared_runtime_make()
is not aware of the nested rmdir() requirements)
While at it, remove unneeded PROTECT_ERRNO.
Useful for packaging scripts, when units are removed.
When multiple markers are assigned without +/-, the last one wins.
When using +/-, the job merging logic is followed to the extent possible.
ref_uid is initialized to invalid, and is only set in some
circumstances. The AttachProcesses will attempt to check it,
and assert that it is valid. Check beforehand.
Reported as YWH-PGM9780-89
Follow-up for 59857b672c
This adds the ability to download OCI images via importd.
Not a fan of the OCI format tbh, in particular its security properties
are a bit sad. But I guess it exists and is very popular, hence we might
as well add support for it, even if it comes at much weaker security
properties than DDIs.
Fixes#36447
This introduces PinnedResources as a structure combining pinned
references to a root directory, root image, or root mstack. This is not
only easier to work with, but essential to make certain unpriv things
work, as we need some mechanism to pin resources before we drop into a
userns which might possibly not provide access anymore to those
resources.
Hence this does two things: introduce the new structure, and immediately
hook it up so that we pin things properly before dropping into userns,
and then makes use of this after dropping the right way, and enables
unpriv userns operation.
The concept is generic enough to eventually implement extension images +
mount images with the same structure, but in order to keep the changes
managable this is left for another time.
(This also makes one further clean-up: client-side verity-reuse checks
are moved server side if we are unpriv. Previously we'd do them client
side, but they were doomed to fail because of lack of privs. Hence let's
drop the client side if we are unpriv and purely do them server-side in
that case.)
So far we opened a new Varlink connection for every mountfsd/nsresourced
method call. Given each tool only does a very small number of calls
(usually 1…5) on them and the connections are cheap this is not too
wasteful. Nonetheless, let's do something about it, and allow reusing
the connection for multiple calls.
This not only makes things a bit more efficient, but has one more
important benefit: Varlink connections pin the security context of the
client when connecting. This means that varlink method calls done with a
connection established while some code was privileged will still operate
as privieged once privs are dropped, until the connection is closed.
This pinning effect is really nice, as it gives us behaviour in a
"capability system" like scheme. Later code is going to use that to
continue doing certain priv userns ops even after unsharing userns and
becoming fully unpriv.
A later commit will add transient allocation of user namespaces with
dynamic UID range assignment. That creates certain permission issues.
Let's hence allow them to be handled gracefully in case the 'ignore'
field is set for a mount.
This is not just refactoring, but has the big benefit that it makes us
indepdendent from a temporary directory we might not have enough access
to create. (This matters with the new PrivateUsers=managed).
We have two very similar checks in place: in some contexts we want to
know if *any* RootDirectory= is configured, in the other we want to
suppress if it is configured to our regular root. Let's add a helper for
both (even if we only need it once), to make the mirrored behaviour
clear.
There's no need to pass in a boolean to decide whether we use
namespace_enter_delegated() or not. Instead, we can just check if we
have CAP_SYS_ADMIN in our own user namespace. If we don't, then we have
to insist on a child user namespace being passed in and we have to enter
it first to get CAP_SYS_ADMIN as without CAP_SYS_ADMIN we wouldn't be
able
to call setns() in the first place. If we do have CAP_SYS_ADMIN, we can
always enter the other namespaces first before entering the user
namespace.
Additionally, we don't fail anymore if we can't reset the UID/GID since
a
root user might not always be available in every user namespace we might
enter.
There's no need to pass in a boolean to decide whether we use
namespace_enter_delegated() or not. Instead, we can just check if we
have CAP_SYS_ADMIN in our own user namespace. If we don't, then we have
to insist on a child user namespace being passed in and we have to enter
it first to get CAP_SYS_ADMIN as without CAP_SYS_ADMIN we wouldn't be able
to call setns() in the first place. If we do have CAP_SYS_ADMIN, we can
always enter the other namespaces first before entering the user namespace.
Additionally, we don't fail anymore if we can't reset the UID/GID since a
root user might not always be available in every user namespace we might
enter.
Follow-up for bb1ef2edf7
The commit introduced a new "metrics" varlink server, but for
user scope stuff it is not bound anywhere. The copy-pasted
"fresh" handling for deserialization is also essentially
meaningless as metrics_setup_varlink_server() doesn't even report
whether the varlink server is fresh (let alone that no serialization
is being done at all right now). Moreover, currently the event
priority is hardcoded, while event loop and associated priority
assignment ought to be subject to each daemon.
While fixing the mentioned issues I took the chance to restructure
the existing code a bit for readability. Note that serialization
for the metrics server is still missing - it will be tackled
in subsequent commits.
Zbigniew Jędrzejewski-Szmek
Mike Yuan
Michal Sekletar
Frantisek Sumsal
noxiouz
Zbigniew Jędrzejewski-Szmek
Luca Boccassi
Daan De Meyer
Lennart Poettering
Yaping Li
Yu Watanabe