This variable is not set by meson, so let's not try to use it.
We could use some more elaborate scheme (e.g. based on $MESON_BUILD_ROOT and
$MESON_SUBDIR) to find the path to systemd-sysv-generator, but it seems
that plain ./systemd-sysv-generator works just as well and has the advantage
that it's easy to invoke the test by hand (as long as one cd's to the
meson build dir).
fileio: try to read one byte too much in read_full_stream()
Let's read one byte more than the file size we read from stat() on the
first fread() invocation. That way, the first read() will already be
short and indicate eof to fread().
As it turns out LINE_MAX at 4096 is too short for many usecases. Since
the general concept of having a common maximum line length limit makes
sense let's add our own, and make it larger (1MB for now).
fileio: add new helper call read_line() as bounded getline() replacement
read_line() is much like getline(), and returns a line read from a
FILE*, of arbitrary sizes. In contrast to gets() it will grow the buffer
dynamically, and in contrast to getline() it will place a user-specified
boundary on the line.
socket: assign socket units to a default slice unconditionally
Due to the chown() logic socket units might end up with processes even
if no explicit command is defined for them, hence let's make sure these
processes are in the right cgroup, and that means within a slice.
Mount, swap and service units unconditionally are assigned to a slice
already, let's do the same here, too.
(This becomes more important as soon as the ebpf/firewall stuff is
merged, as there'll be another reason to fork off processes then)
cgroup: rework which files we chown() on delegation
On cgroupsv2 we should also chown()/chmod() the subtree_control file,
so that children can use controllers the way they like.
On cgroupsv1 we should also chown()/chmod() cgroups.clone_children, as
not setting this for new cgroups makes little sense, and hence delegated
clients should be able to write to it.
Note that error handling for both cases is different. subtree_control
matters so we check for errors, but the clone_children/tasks stuff
doesn't really, as it's legacy stuff. Hence we only log errors and
proceed.
cgroup-util: downgrade log messages from library code to LOG_DEBUG
These errors don't really matter, that's why we log and proceed in the
current code. However, we currently log at LOG_WARNING, but we really
shouldn't given that this is library code. Hence downgrade this to
LOG_DEBUG.
install: consider globally enabled units as "enabled" for the user
We would not consider symlinks in /etc/systemd/user/*.{wants,requires}/
towards the user unit being "enabled", because the symlinks were not
located in "config" paths. But this is confusing to users, since those units
are clearly enabled and will be started. So let's muddle the definition of
enablement a bit to include the paths only accessible to root when looking for
enabled user units.
install: consider non-Alias=/non-DefaultInstance= symlinks as "indirect" enablement
I think this matches the spirit of "indirect" well: the unit
*might* be active, even though it is not "installed" in the
sense of symlinks created based on the [Install] section.
The changes to test-install-root touch the same lines as in the previous
commit; the change in each case is from
assert_se(unit_file_get_state(...) >= 0 && state == UNIT_FILE_ENABLED)
to
assert_se(unit_file_get_state(...) >= 0 && state == UNIT_FILE_DISABLED)
to
assert_se(unit_file_get_state(...) >= 0 && state == UNIT_FILE_INDIRECT)
in the last two commits.
install: only consider names in Alias= as "enabling"
When a unit has a symlink that makes an alias in the filesystem,
but that name is not specified in [Install], it is confusing
is the unit is shown as "enabled". Look only for names specified
in Alias=.
Fixes #6338.
v2:
- Fix indentation.
- Fix checking for normal enablement, when the symlink name is the same as the
unit name. This case wasn't handled properly in v1.
v3:
- Rework the patch to also handle templates properly:
A template templ@.service with DefaultInstance=foo will be considered
enabled only when templ@foo.service symlink is found. Symlinks with
other instance names do not count, which matches the logic for aliases
to normal units. Tests are updated.
bpf: set BPF_F_ALLOW_OVERRIDE when attaching a cgroup program if Delegate=yes is set
Let's permit installing BPF programs in cgroup subtrees if
Delegeate=yes. Let's not document this precise behaviour for now though,
as most likely the logic here should become recursive, but that's only
going to happen if the kernel starts supporting that. Until then,
support this in a non-recursive fashion.
man: remove double newlines in systemd.special man page header
The <!-- --> comment lines resulted in double newlines in the man page
header, which looks quite ugly. Let's rearrange a bit so that these
comments don't result in changes in the output.
main: bump RLIMIT_NOFILE for the root user substantially
On current kernels BPF_MAP_TYPE_LPM_TRIE bpf maps are charged against
RLIMIT_MEMLOCK even for privileged users that have CAP_IPC_LOCK. Given
that mlock() generally ignores RLIMIT_MEMLOCK if CAP_IPC_LOCK is set
this appears to be an oversight in the kernel. Either way, until that's
fixed, let's just bump RLIMIT_MEMLOCK for the root user considerably, as
the default is quite limiting, and doesn't permit us to create more than
a few TRIE maps.
In times of seccomp it might very well fail, and given that we return
failures from this function anyway, let's also propagate getrlimit()
failures, just to be safe.
core: whenever a unit terminates, log its consumed resources to the journal
This adds a new recognizable log message for each unit invocation that
contains structured information about consumed resources of the unit as
a whole after it terminated. This is particular useful for apps that
want to figure out what the resource consumption of a unit given a
specific invocation ID was.
The log message is only generated for units that have at least one
XyzAccounting= property turned on, and currently only covers IP traffic and CPU
time metrics.
nspawn: set up a new session keyring for the container process
keyring material should not leak into the container. So far we relied on
seccomp to deny access to the keyring, but given that we now made the
seccomp configurable, and access to keyctl() and friends may optionally
be permitted to containers now let's make sure we disconnect the callers
keyring from the keyring of PID 1 in the container.
This adds IOVEC_INIT() and IOVEC_MAKE() for initializing iovec structures
from a pointer and a size. On top of these IOVEC_INIT_STRING() and
IOVEC_MAKE_STRING() are added which take a string and automatically
determine the size of the string using strlen().
This patch removes the old IOVEC_SET_STRING() macro, given that
IOVEC_MAKE_STRING() is now useful for similar purposes. Note that the
old IOVEC_SET_STRING() invocations were two characters shorter than the
new ones using IOVEC_MAKE_STRING(), but I think the new syntax is more
readable and more generic as it simply resolves to a C99 literal
structure initialization. Moreover, we can use very similar syntax now
for initializing strings and pointer+size iovec entries. We canalso use
the new macros to initialize function parameters on-the-fly or array
definitions. And given that we shouldn't have so many ways to do the
same stuff, let's just settle on the new macros.
(This also converts some code to use _cleanup_ where dynamically
allocated strings were using IOVEC_SET_STRING() before, to modernize
things a bit)
job: change result field for log message about job result RESULT= → JOB_RESULT=
So, currently, some of the structured log messages we generated based on
jobs carry the result in RESULT=, and others in JOB_RESULT=. Let's
streamline this, as stick to JOB_RESULT= in one place.
This is kind of an API break, but given that currently most software has
to check both fields anyway, I think we can get away with it.
Why unify on JOB_RESULT= rather than RESULT=? Well, we manage different
types of result codes in systemd. Most importanlty besides job results
there are also service results, and we should be explicit in what we
mean here.
cgroup: refuse to return accounting data if accounting isn't turned on
We used to be a bit sloppy on this, and handed out accounting data even
for units where accounting wasn't explicitly enabled. Let's be stricter
here, so that we know the accounting data is actually fully valid. This
is necessary, as the accounting data is no longer stored exclusively in
cgroupfs, but is partly maintained external of that, and flushed during
unit starts. We should hence only expose accounting data we really know
is fully current.
core: when coming back from reload/reexec, reapply all cgroup properties
With this change we'll invalidate all cgroup settings after coming back
from a daemon reload/reexec, so that the new settings are instantly
applied.
This is useful for the BPF case, because we don't serialize/deserialize
the BPF program fd, and hence have to install a new, updated BPF program
when coming back from the reload/reexec. However, this is also useful
for the rest of the cgroup settings, as it ensures that user
configuration really takes effect wherever we can.
core: serialize/deserialize IP accounting across daemon reload/reexec
Make sure the current IP accounting counters aren't lost during
reload/reexec.
Note that we destroy all BPF file objects during a reload: the BPF
programs, the access and the accounting maps. The former two need to be
regenerated anyway with the newly loaded configuration data, but the
latter one needs to survive reloads/reexec. In this implementation I
opted to only save/restore the accounting map content instead of the map
itself. While this opens a (theoretic) window where IP traffic is still
accounted to the old map after we read it out, and we thus miss a few
bytes this has the benefit that we can alter the map layout between
versions should the need arise.
core: when creating the socket fds for a socket unit, join socket's cgroup first
Let's make sure that a socket unit's IPAddressAllow=/IPAddressDeny=
settings are in effect on all socket fds associated with it. In order to
make this happen we need to make sure the cgroup the fds are associated
with are the socket unit's cgroup. The only way to do that is invoking
socket()+accept() in them. Since we really don't want to migrate PID 1
around we do this by forking off a helper process, which invokes
socket()/accept() and sends the newly created fd to PID 1. Ugly, but
works, and there's apparently no better way right now.
This generalizes forking off per-unit helper processes in a new function
unit_fork_helper_process(), which is then also used by the NSS chown()
code of socket units.
Daniel Mack [Thu, 3 Nov 2016 16:30:06 +0000 (17:30 +0100)]
Add IP address address ACL representation and parser
Add a config directive parser that takes multiple space separated IPv4
or IPv6 addresses with optional netmasks in CIDR notation rvalue and
puts a parsed version of it to linked list of IPAddressAccessItem objects.
The code actually using this will be added later.
manager: initialize timeouts when allocating a naked Manager object
This way we can safely run manager objects from tests and good timeouts
apply. Without this all timeouts are set 0, which means they fire
instantly, when run from tests which do not explicitly configure them
(the way main.c does).
cgroup: always invalidate "cpu" and "cpuacct" together
This doesn't really matter, as we never invalidate cpuacct explicitly,
and there's no real reason to care for it explicitly, however it's
prettier if we always treat cpu and cpuacct as belonging together, the
same way we conisder "io" and "blkio" to belong together.
core: make sure to dump cgroup context when unit_dump() is called for all unit types
For some reason we didn't dump the cgroup context for a number of unit
types, including service units. Not sure how this wasn't noticed
before... Add this in.
time-util: correctly handle the timezone when parsing
The timezone was cut off the string once the timezone was not UTC.
If it is not UTC but a other timezone that matches tzname[0] or
tzname[1], then we can leave it to the impl function to parse that
correctly. If not we can just fallback to whatever is the current
timezone is in the given t_timezone.
journald: make maximum size of stream log lines configurable and bump it to 48K (#6838)
This adds a new setting LineMax= to journald.conf, and sets it by
default to 48K. When we convert stream-based stdout/stderr logging into
record-based log entries, read up to the specified amount of bytes
before forcing a line-break.
This also makes three related changes:
- When a NUL byte is read we'll not recognize this as alternative line
break, instead of silently dropping everything after it. (see #4863)
- The reason for a line-break is now encoded in the log record, if it
wasn't a plain newline. Specifically, we distuingish "nul",
"line-max" and "eof", for line breaks due to NUL byte, due to the
maximum line length as configured with LineMax= or due to end of
stream. This data is stored in the new implicit _LINE_BREAK= field.
It's not synthesized for plain \n line breaks.
- A randomized 128bit ID is assigned to each log stream.
With these three changes in place it's (mostly) possible to reconstruct
the original byte streams from log data, as (most) of the context of
the conversion from the byte stream to log records is saved now. (So,
the only bits we still drop are empty lines. Which might be something to
look into in a future change, and which is outside of the scope of this
work)
Tommi Rantala [Tue, 19 Sep 2017 08:10:49 +0000 (11:10 +0300)]
journal: add object sanity check to journal_file_move_to_object()
Introduce journal_file_check_object(), which does lightweight object
sanity checks, and use it in journal_file_move_to_object(), so that we
will catch certain corrupted objects in the journal file.
This fixes #6447, where we had only partially written out OBJECT_ENTRY
(ObjectHeader written, but rest of object zero bytes), causing
"journalctl --list-boots" to fail.
$ builddir.vanilla/journalctl --list-boots -D bug6447/
Failed to determine boots: No data available
Note that journal_file_check_object() is similar to
journal_file_object_verify(). The most expensive checks are omitted, as
they would slow down every journal_file_move_to_object() call too much.
With this implementation, the added overhead is small, for example when
dumping some journal content to /dev/null
(built with -Dbuildtype=debugoptimized -Db_ndebug=true):
projects / thirdparty / systemd.git / log
