Aside from the usual boilerplate of moving the shared logic to shared/,
we also rework the implementation of --bind-user= to be similar to what
we'll do in systemd-vmspawn. Instead of messing with the nspawn container
user namespace, we use idmapped mounts to map the user's home directory on
the host to the mapped uid in the container.
Ideally we'd also use the "userdb.transient" credentials to provision the
user records, but this would only work for booted containers, whereas the
current logic works for non-booted containers as well.
Aside from being similar to how we'll implement --bind-user= in vmspawn,
using idmapped mounts also allows supporting --bind-user= without having to
use --private-users=.
We only need to clear the existing idmapping if we're going to be
replacing it with another idmapping. Otherwise we should keep the
existing idmapping in place.
This adds missing glue to reasonably allow unpriv users VMs/containers
to register with the system machined.
This primarily adds two things:
1. machined can now properly track VMs/containers residing in subcgroups
of units, because that's effectively what happens for per-user
VMs/containers: they are placed below the system unit `user@….service`
in some user unit.
2. machines registered with machined now have an owning UID: users can
operate on their own machines withour re-authentication, but not on
others.
Note that this is only a first step regarding machined's hookup of
nspawn/vmspawn in the long run for unpriv operation.
I think eventually we should make it so that there's both a per-user and
a per-system machined instance (so far, and even with this PR there's
still one per-system instance), and per-user containers/VMs would
registering with *both*. Having two instances makes sense I think,
because it would mean we can make machined reasonably manage the
per-user image discovery, and also do the per-system network/hostname
handling.
Let's also send out a STATUS= message when we get READY=1 if it didn't
come with a STATUS= message itself.
Also, let's initially say the container is "started", and only once the
READY=1 is seen claim it was "running".
This cleans up allocation of a scope unit for the container: when
invoked in user context we'll now allocate a scope through the per-user
service manager instead of the per-system manager. This makes a ton more
sense, since it's the user that invokes things after all. And given that
machined now can register containers in the user manager there's nothing
stopping us to clean this up.
Note that this means we'll connect to two busses if run unpriv: once to
the per-user bus to allocate the scope unit, and once to the per-system
bus to register it with machined.
Then, move syscall definitions to the wrapper, and prototypes are moved
to relevant headers.
This also adds checks for add_key() and request_key(), as one day
glibc may be going to add some of them separatedly.
The check for fspick in meson.build is dropped, as it is currently
unused in our code.
This also moves
- basic/missing_bpf.h -> include/override/linux/bpf.h,
- basic/missing_keyctl.h -> include/override/linux/keyctl.h.
- This introduces ERRNO_IS_FS_WRITE_REFUSED(), and apply it where
usable.
- This makes unexpected errors in access_fd() called by binfmt_mounted()
propagated to the caller.
- Renames binfmt_mounted() to binfmt_mounted_and_writable(), as it also
checks the fs is writable.
- Voidifies one disable_binfmt() call in shutdown.c.
Christian made this possible in Linux 6.15 with a new system call
open_tree_attr() that combines open_tree() and mount_setattr().
Because idmapped mounts are (rightfully) not nested, we have to do
some extra shenanigans to make source we're putting the right source
uid in the userns for any idmapped mounts that we do in nspawn.
Of course we also add the necessary boilerplate to make open_tree_attr()
available in our code and wrap open_tree_attr() and the corresponding
fallback in a new function which we then use everywhere else.
Prior to this change, no user shell can be specified in the user
records passed into a container via --bind-user=. This new option
allows users to:
1. When false (the default), continue to specify no user shell for
each bound user record, resulting in the use of the container's
default shell for bound users.
2. When true, include each host user's shell in the corresponding
user record passed into a container (via --bind-user=).
3. When an absolute path, set that path as the user shell for each
user record passed into a container (via --bind-user=).
This does not change the existing behavior, but allows users to
opt-in to either copy the shells specified by the host user records
or override the shell explicitly by path.
We use symbols provided by unistd.h without including it. E.g.
open(), close(), read(), write(), access(), symlink(), unlink(), rmdir(),
fsync(), syncfs(), lseek(), ftruncate(), fchown(), dup2(), pipe2(),
getuid(), getgid(), gettid(), getppid(), pipe2(), execv(), _exit(),
environ, STDIN_FILENO, STDOUT_FILENO, STDERR_FILENO, F_OK, and their
friends and variants, so on.
Currently, unistd.h is indirectly included mainly in the following two paths:
- through missing_syscall.h, which is planned to covert to .c file.
- through signal.h -> bits/sigstksz.h, which is new since glibc-2.34.
Note, signal.h is included by sd-eevent.h. So, many source files
indirectly include unistd.h if newer glibc is used.
Currently, our baseline on glibc is 2.31. We need to support glibc older
than 2.34, but unfortunately, we do not have any CI environments with
such old glibc. CIFuzz uses glibc-2.31, but it builds only fuzzers, and
many files are even not compiled.
[1] says:
> Since 0.60.0 the name argument is optional and defaults to the basename of
> the first output
We specify >= 0.62 as the supported version, so drop the duplicate name in all cases
where it is the same as outputs[0], i.e. almost all cases.
[1] https://mesonbuild.com/Reference-manual_functions.html#custom_target
Then, make parse_cpu_set() as a tiny wrapper of it.
Note, previously when an invalid CPU range, e.g. "3-0", is specified,
we ignore the range but allocate an empty set. But, with this commit,
now the conf parser simply ignore it without no side effect.
This potentially changes behavior of a system with such invalid setting,
but the change should be favorable for consistency with other parsers.
clang-include-cleaner has a hard time figuring out unused includes
with all the macro string concatentation we're doing in test_table().
Let's help it out a little by also passing the enum type into test_table().
While technically not needed, this makes the tests a bit more self-descriptive
and not really much more verbose so it feels like a good chance regardless,
even if the primary purpose is to help clang-include-cleaner figure out
used and unused includes.
Also, let's deflatten the lists of sources in preparation for the
next commit at the same time.
In systemctl, we split out systemctl-main.c to make sure the definition
of main() is in a separate object which allows us to extract the systemctl.c
object and link it in the fuzzer target without getting a multiple definition
error when linking.
Currently, when we want to add unit tests for code that is compiled into
an executable, we either compile the code at least twice (once for the
executable, and once for each test that uses it) or we create a static
library which is then used by both the executable and all the tests.
Both of these options are not ideal, compiling source files more than
once slows down the build for no reason and creating the intermediate
static libraries takes a lot of boilerplate.
Instead, let's use the extract_objects() method that meson exposes on
build targets. This allows us to extract the objects corresponding to
specific source files and use them in other executables. Because we
define all executables upfront into a dictionary, we integrate this into
the dictionary approach by adding two new fields:
- 'extract' takes a list of files for which objects should be extracted.
The extracted objects are stored in a dict keyed by the executable name
from which they were extracted.
- 'objects' takes the name of an executable from which the extracted
objects should be added to the current executable.
One side effect of this approach is that we can't build test executables
anymore without building the main executable, so we stop building test
executables unless we're also building the main executable. This allows
us to switch to using subdir_done() in all of these subdirectories to skip
parsing them if the corresponding component is disabled.
These changes get me down from 2439 => 2403 ninja targets on a full rebuild
from scratch.
Let's move some logic from _DEFINE_MAIN_FUNCTION() and other places
in main-func.h into functions that we implement in main-func.c to
allow moving some included headers from the header to the .c file.
