We now have the "BPF" pseudo-controllers. These should never be assumed
to be accessible as /sys/fs/cgroup/<controller> and not through
"cgroup.subtree_control" either, hence always check explicitly before we
go to the file system. We do this through our new CGROUP_MASK_V1 and
CGROUP_MASK_V2 definitions.
Also, when we fail, don't clobber the return value.
This brings the call more in-line with our usual coding style, and
removes surprises.
None of the callers seemed to care about this behaviour.
Cgroup v2 provides the eBPF-based device controller, which isn't currently
supported by systemd. This commit aims to provide such support.
There are no user-visible changes, just the device policy and whitelist
start working if cgroup v2 is used.
Remove "arbitrary named hierarchies" from the list of things that
cg_kernel_controllers() might return, and clarify that "name="
pseudo-controllers are not included in the returned list.
/proc/cgroups does not contain "name=" pseudo-controllers, and
cg_kernel_controllers() makes no effort to enumerate them via a different
mechanism.
These lines are generally out-of-date, incomplete and unnecessary. With
SPDX and git repository much more accurate and fine grained information
about licensing and authorship is available, hence let's drop the
per-file copyright notice. Of course, removing copyright lines of others
is problematic, hence this commit only removes my own lines and leaves
all others untouched. It might be nicer if sooner or later those could
go away too, making git the only and accurate source of authorship
information.
This part of the copyright blurb stems from the GPL use recommendations:
https://www.gnu.org/licenses/gpl-howto.en.html
The concept appears to originate in times where version control was per
file, instead of per tree, and was a way to glue the files together.
Ultimately, we nowadays don't live in that world anymore, and this
information is entirely useless anyway, as people are very welcome to
copy these files into any projects they like, and they shouldn't have to
change bits that are part of our copyright header for that.
hence, let's just get rid of this old cruft, and shorten our codebase a
bit.
The function is similar to path_kill_slashes() but also removes
initial './', trailing '/.', and '/./' in the path.
When the second argument of path_simplify() is false, then it
behaves as the same as path_kill_slashes(). Hence, this also
replaces path_kill_slashes() with path_simplify().
If enabling controller for some reason fails we need to clear error
for the FILE stream. Enabling remaining controllers would otherwise
fail because write_string_stream_ts() checks for ferror(f) and returns
-EIO if there is one.
Broken by commit <77fa610b22>.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
We check the same condition at various places. Let's add a trivial,
common helper for this, and use it everywhere.
It's not going to make things much faster or much shorter, but I think a
lot more readable
Files which are installed as-is (any .service and other unit files, .conf
files, .policy files, etc), are left as is. My assumption is that SPDX
identifiers are not yet that well known, so it's better to retain the
extended header to avoid any doubt.
I also kept any copyright lines. We can probably remove them, but it'd nice to
obtain explicit acks from all involved authors before doing that.
https://clang.llvm.org/docs/MemorySanitizer.html#id5 documents this
check as the way to detect MemorySanitizer at compilation time. We
only need to skip the test if MemorySanitizer is used.
Also, use this condition in cg_slice_to_path(). There, the code that is
conditionalized is not harmful in any way (it's just unnecessary), so remove
the FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION condition.
Fixes#8482.
This macro will read a pointer of any type, return it, and set the
pointer to NULL. This is useful as an explicit concept of passing
ownership of a memory area between pointers.
This takes inspiration from Rust:
https://doc.rust-lang.org/std/option/enum.Option.html#method.take
and was suggested by Alan Jenkins (@sourcejedi).
It drops ~160 lines of code from our codebase, which makes me like it.
Also, I think it clarifies passing of ownership, and thus helps
readability a bit (at least for the initiated who know the new macro)
The orignal reproducer from oss-fuzz depends on the hostname (via %H and %c).
The hostname needs a dash for msan to report this, so a simpler case from
@evverx with the dash hardcoded is also added.
The issue is a false positive from msan, which does not instruct stpncpy
(https://github.com/google/sanitizers/issues/926). Let's add a work-around
until this is fixed.
Let's make sure we don't clobber the return parameter on failure, to
follow our coding style. Also, break the loop early if we have all
attributes we need.
This also changes the keys parameter to a simple char**, so that we can
use STRV_MAKE() for passing the list of attributes to read.
This also makes it possible to distuingish the case when the whole
attribute file doesn't exist from one key in it missing. In the former
case we return -ENOENT, in the latter we now return -ENXIO.
These helper calls are potentially called often, and allocate FILE*
objects internally for a very short period of time, let's turn off
locking for them too.
We never use these functions seperately, hence don't bother splitting
them into to.
Also, simplify things a bit, and maintain tables for the attribute files
to chown. Let's also update those tables a bit, and include thenew
"cgroup.threads" file in it, that needs to be delegated too, according
to the documentation.
When a process becomes a zombie its cgroup might be deleted. Let's add
some minimal code to detect cases like this, so that we can still
attribute this back to the original cgroup.
Already, path_is_safe() refused paths container the "." dir. Doing that
isn't strictly necessary to be "safe" by most definitions of the word.
But it is necessary in order to consider a path "normalized". Hence,
"path_is_safe()" is slightly misleading a name, but
"path_is_normalize()" is more descriptive, hence let's rename things
accordingly.
No functional changes.
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.
Fixes: #6216
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.
This moves pretty much all uses of getpid() over to getpid_raw(). I
didn't specifically check whether the optimization is worth it for each
replacement, but in order to keep things simple and systematic I
switched over everything at once.
Let's just check the unified level, directly. There's really no value in
wrapping cg_unified_controllers() with this, i.e. potentially do string
comparison when there's no reason to.
Also, this makes the clal more alike cg_hybrid_unified().
We use our cgroup APIs in various contexts, including from our libraries
sd-login, sd-bus. As we don#t control those environments we can't rely
that the unified cgroup setup logic succeeds, and hence really shouldn't
assert on it.
This more or less reverts 415fc41cea.
If we encounter an error in proc cmdline parsing, just treat that as permanent,
i.e. the same as if the option was not specified. Realistically, it is better
to use the same condition for all related mounts, then to have e.g.
/sys/fs/cgroup mounted and /sys/fs/cgroup/unified not. If we find something is
mounted and base our answer on that, cache that result too.
Fix the conditions so that if "unified" is used, make sure any "hybrid" mounts
are not mounted.
We need this to gracefully support older or strangely configured kernels.
v2:
- do not install a callback handler, just embed the right conditions into
cg_is_*_wanted()
v3:
- fix bug in cg_is_legacy_wanted()
The default default is set to "legacy", with "hybrid" and "unified"
being the other two alternatives.
There invert the behaviour for systemd.legacy_systemd_cgroup_controller:
if it is not specified on the kernel command line, "hybrid" is used if
selected as the default. If this option is specified, "hybrid" is used if false,
and full "legacy" if true.
Also make all fields in the configure summary lowercase (unless they are
capitalized names) for consistency.
v2:
- update for the fixed interpreation of systemd.legacy_systemd_cgroup_controller
v232's cgroup hybrid mode mounted v2 on /sys/fs/cgroup/systemd, which
unfortunately broke other tools which expect v1 there. From v233 on, hybrid
mode instead mounts and uses v2 on /sys/fs/cgroup/unified and keeps
/sys/fs/cgroup/systemd on v1 for compatibility with external tools. However,
to keep systemd live upgrades working, v233+ should be able to recognize v232
layout and keep using it.
This patch adds v232 hybrid mode support. If v232 layout is detected,
cg_unified(SYSTEMD_CGRouP_CONTROLLER) keeps returning %true but
cg_hybrid_unified() returns %false. This keeps process management on cgroup v2
but turns off the parallel layout.
Currently the hybrid mode mounts cgroup v2 on /sys/fs/cgroup instead of the v1
name=systemd hierarchy. While this works fine for systemd itself, it breaks
tools which expect cgroup v1 hierarchy on /sys/fs/cgroup/systemd.
This patch updates the hybrid mode so that it mounts v2 hierarchy on
/sys/fs/cgroup/unified and keeps v1 "name=systemd" hierarchy on
/sys/fs/cgroup/systemd for compatibility. systemd itself doesn't depend on the
"name=systemd" hierarchy at all. All operations take place on the v2 hierarchy
as before but the v1 hierarchy is kept in sync so that any tools which expect
it to be there can keep doing so. This allows systemd to take advantage of
cgroup v2 process management without requiring other tools to be aware of the
hybrid mode.
The hybrid mode is implemented by mapping the special systemd controller to
/sys/fs/cgroup/unified and making the basic cgroup utility operations -
cg_attach(), cg_create(), cg_rmdir() and cg_trim() - also operate on the
/sys/fs/cgroup/systemd hierarchy whenever the cgroup2 hierarchy is updated.
While a bit messy, this will allow dropping complications from using cgroup v1
for process management a lot sooner than otherwise possible which should make
it a net gain in terms of maintainability.
v2: Fixed !cgns breakage reported by @evverx and renamed the unified mount
point to /sys/fs/cgroup/unified as suggested by @brauner.
v3: chown the compat hierarchy too on delegation. Suggested by @evverx.
v4: [zj]
- drop the change to default, full "legacy" is still the default.
1d84ad9445 reversed the meaning of the option.
The kernel command line option has the opposite meaning to the function,
i.e. specifying "legacy=yes" means "unifed systemd controller=no".
SYSTEMD_CGROUP_CONTROLLER is currently defined as "name=systemd" which cgroup
utility functions interpret as a named cgroup hierarchy with the specified
named. With the planned cgroup hybrid mode changes, SYSTEMD_CGROUP_CONTROLLER
would map to different hierarchy names.
This patch makes SYSTEMD_CGROUP_CONTROLLER a special string "_systemd" which is
substituted to "name=systemd" by the cgroup utility functions. This allows the
callers to address the systemd hierarchy without actually specifying the
hierarchy name allowing the cgroup utility functions to map it to whatever is
appropriate.
Note that SYSTEMD_CGROUP_CONTROLLER was already special on full unified cgroup
hierarchy even before this patch.
cg_[all_]unified() test whether a specific controller or all controllers are on
the unified hierarchy. While what's being asked is a simple binary question,
the callers must assume that the functions may fail any time, which
unnecessarily complicates their usages. This complication is unnecessary.
Internally, the test result is cached anyway and there are only a few places
where the test actually needs to be performed.
This patch simplifies cg_[all_]unified().
* cg_[all_]unified() are updated to return bool. If the result can't be
decided, assertion failure is triggered. Error handlings from their callers
are dropped.
* cg_unified_flush() is updated to calculate the new result synchrnously and
return whether it succeeded or not. Places which need to flush the test
result are updated to test for failure. This ensures that all the following
cg_[all_]unified() tests succeed.
* Places which expected possible cg_[all_]unified() failures are updated to
call and test cg_unified_flush() before calling cg_[all_]unified(). This
includes functions used while setting up mounts during boot and
manager_setup_cgroup().
This improves kernel command line parsing in a number of ways:
a) An kernel option "foo_bar=xyz" is now considered equivalent to
"foo-bar-xyz", i.e. when comparing kernel command line option names "-" and
"_" are now considered equivalent (this only applies to the option names
though, not the option values!). Most of our kernel options used "-" as word
separator in kernel command line options so far, but some used "_". With
this change, which was a source of confusion for users (well, at least of
one user: myself, I just couldn't remember that it's systemd.debug-shell,
not systemd.debug_shell). Considering both as equivalent is inspired how
modern kernel module loading normalizes all kernel module names to use
underscores now too.
b) All options previously using a dash for separating words in kernel command
line options now use an underscore instead, in all documentation and in
code. Since a) has been implemented this should not create any compatibility
problems, but normalizes our documentation and our code.
c) All kernel command line options which take booleans (or are boolean-like)
have been reworked so that "foobar" (without argument) is now equivalent to
"foobar=1" (but not "foobar=0"), thus normalizing the handling of our
boolean arguments. Specifically this means systemd.debug-shell and
systemd_debug_shell=1 are now entirely equivalent.
d) All kernel command line options which take an argument, and where no
argument is specified will now result in a log message. e.g. passing just
"systemd.unit" will no result in a complain that it needs an argument. This
is implemented in the proc_cmdline_missing_value() function.
e) There's now a call proc_cmdline_get_bool() similar to proc_cmdline_get_key()
that parses booleans (following the logic explained in c).
f) The proc_cmdline_parse() call's boolean argument has been replaced by a new
flags argument that takes a common set of bits with proc_cmdline_get_key().
g) All kernel command line APIs now begin with the same "proc_cmdline_" prefix.
h) There are now tests for much of this. Yay!
We don't have plural in the name of any other -util files and this
inconsistency trips me up every time I try to type this file name
from memory. "formats-util" is even hard to pronounce.
This adds a new invocation ID concept to the service manager. The invocation ID
identifies each runtime cycle of a unit uniquely. A new randomized 128bit ID is
generated each time a unit moves from and inactive to an activating or active
state.
The primary usecase for this concept is to connect the runtime data PID 1
maintains about a service with the offline data the journal stores about it.
Previously we'd use the unit name plus start/stop times, which however is
highly racy since the journal will generally process log data after the service
already ended.
The "invocation ID" kinda matches the "boot ID" concept of the Linux kernel,
except that it applies to an individual unit instead of the whole system.
The invocation ID is passed to the activated processes as environment variable.
It is additionally stored as extended attribute on the cgroup of the unit. The
latter is used by journald to automatically retrieve it for each log logged
message and attach it to the log entry. The environment variable is very easily
accessible, even for unprivileged services. OTOH the extended attribute is only
accessible to privileged processes (this is because cgroupfs only supports the
"trusted." xattr namespace, not "user."). The environment variable may be
altered by services, the extended attribute may not be, hence is the better
choice for the journal.
Note that reading the invocation ID off the extended attribute from journald is
racy, similar to the way reading the unit name for a logging process is.
This patch adds APIs to read the invocation ID to sd-id128:
sd_id128_get_invocation() may be used in a similar fashion to
sd_id128_get_boot().
PID1's own logging is updated to always include the invocation ID when it logs
information about a unit.
A new bus call GetUnitByInvocationID() is added that allows retrieving a bus
path to a unit by its invocation ID. The bus path is built using the invocation
ID, thus providing a path for referring to a unit that is valid only for the
current runtime cycleof it.
Outlook for the future: should the kernel eventually allow passing of cgroup
information along AF_UNIX/SOCK_DGRAM messages via a unique cgroup id, then we
can alter the invocation ID to be generated as hash from that rather than
entirely randomly. This way we can derive the invocation race-freely from the
messages.
Currently, systemd uses either the legacy hierarchies or the unified hierarchy.
When the legacy hierarchies are used, systemd uses a named legacy hierarchy
mounted on /sys/fs/cgroup/systemd without any kernel controllers for process
management. Due to the shortcomings in the legacy hierarchy, this involves a
lot of workarounds and complexities.
Because the unified hierarchy can be mounted and used in parallel to legacy
hierarchies, there's no reason for systemd to use a legacy hierarchy for
management even if the kernel resource controllers need to be mounted on legacy
hierarchies. It can simply mount the unified hierarchy under
/sys/fs/cgroup/systemd and use it without affecting other legacy hierarchies.
This disables a significant amount of fragile workaround logics and would allow
using features which depend on the unified hierarchy membership such bpf cgroup
v2 membership test. In time, this would also allow deleting the said
complexities.
This patch updates systemd so that it prefers the unified hierarchy for the
systemd cgroup controller hierarchy when legacy hierarchies are used for kernel
resource controllers.
* cg_unified(@controller) is introduced which tests whether the specific
controller in on unified hierarchy and used to choose the unified hierarchy
code path for process and service management when available. Kernel
controller specific operations remain gated by cg_all_unified().
* "systemd.legacy_systemd_cgroup_controller" kernel argument can be used to
force the use of legacy hierarchy for systemd cgroup controller.
* nspawn: By default nspawn uses the same hierarchies as the host. If
UNIFIED_CGROUP_HIERARCHY is set to 1, unified hierarchy is used for all. If
0, legacy for all.
* nspawn: arg_unified_cgroup_hierarchy is made an enum and now encodes one of
three options - legacy, only systemd controller on unified, and unified. The
value is passed into mount setup functions and controls cgroup configuration.
* nspawn: Interpretation of SYSTEMD_CGROUP_CONTROLLER to the actual mount
option is moved to mount_legacy_cgroup_hierarchy() so that it can take an
appropriate action depending on the configuration of the host.
v2: - CGroupUnified enum replaces open coded integer values to indicate the
cgroup operation mode.
- Various style updates.
v3: Fixed a bug in detect_unified_cgroup_hierarchy() introduced during v2.
v4: Restored legacy container on unified host support and fixed another bug in
detect_unified_cgroup_hierarchy().
A following patch will update cgroup handling so that the systemd controller
(/sys/fs/cgroup/systemd) can use the unified hierarchy even if the kernel
resource controllers are on the legacy hierarchies. This would require
distinguishing whether all controllers are on cgroup v2 or only the systemd
controller is. In preparation, this patch renames cg_unified() to
cg_all_unified().
This patch doesn't cause any functional changes.
Unfortunately, due to the disagreements in the kernel development community,
CPU controller cgroup v2 support has not been merged and enabling it requires
applying two small out-of-tree kernel patches. The situation is explained in
the following documentation.
https://git.kernel.org/cgit/linux/kernel/git/tj/cgroup.git/tree/Documentation/cgroup-v2-cpu.txt?h=cgroup-v2-cpu
While it isn't clear what will happen with CPU controller cgroup v2 support,
there are critical features which are possible only on cgroup v2 such as
buffered write control making cgroup v2 essential for a lot of workloads. This
commit implements systemd CPU controller support on the unified hierarchy so
that users who choose to deploy CPU controller cgroup v2 support can easily
take advantage of it.
On the unified hierarchy, "cpu.weight" knob replaces "cpu.shares" and "cpu.max"
replaces "cpu.cfs_period_us" and "cpu.cfs_quota_us". [Startup]CPUWeight config
options are added with the usual compat translation. CPU quota settings remain
unchanged and apply to both legacy and unified hierarchies.
v2: - Error in man page corrected.
- CPU config application in cgroup_context_apply() refactored.
- CPU accounting now works on unified hierarchy.
Let's lot at LOG_NOTICE about any processes that we are going to
SIGKILL/SIGABRT because clean termination of them didn't work.
This turns the various boolean flag parameters to cg_kill(), cg_migrate() and
related calls into a single binary flags parameter, simply because the function
now gained even more parameters and the parameter listed shouldn't get too
long.
Logging for killing processes is done either when the kill signal is SIGABRT or
SIGKILL, or on explicit request if KILL_TERMINATE_AND_LOG instead of LOG_TERMINATE
is passed. This isn't used yet in this patch, but is made use of in a later
patch.
cgroup IO controller supports maximum limits for both bandwidth and IOPS but
systemd resource control currently only supports bandwidth limits. This patch
adds support for IOReadIOPSMax and IOWriteIOPSMax when unified cgroup hierarchy
is in use.
It isn't difficult to also add BlockIOReadIOPS and BlockIOWriteIOPS for legacy
hierarchies but IO control on legacy hierarchies is half-broken anyway, so
let's leave it alone for now.
Currently, there are two cgroup IO limits, bandwidth max for read and write,
and they are hard-coded in various places. This is fine for two limits but IO
is expected to grow more limits - low, high and max limits for bandwidth and
IOPS - and hard-coding each limit won't make sense.
This patch replaces hard-coded limits with an array indexed by
CGroupIOLimitType and accompanying string and default value tables so that new
limits can be added trivially.
On the unified hierarchy, blkio controller is renamed to io and the interface
is changed significantly.
* blkio.weight and blkio.weight_device are consolidated into io.weight which
uses the standardized weight range [1, 10000] with 100 as the default value.
* blkio.throttle.{read|write}_{bps|iops}_device are consolidated into io.max.
Expansion of throttling features is being worked on to support
work-conserving absolute limits (io.low and io.high).
* All stats are consolidated into io.stats.
This patchset adds support for the new interface. As the interface has been
revamped and new features are expected to be added, it seems best to treat it
as a separate controller rather than trying to expand the blkio settings
although we might add automatic translation if only blkio settings are
specified.
* io.weight handling is mostly identical to blkio.weight[_device] handling
except that the weight range is different.
* Both read and write bandwidth settings are consolidated into
CGroupIODeviceLimit which describes all limits applicable to the device.
This makes it less painful to add new limits.
* "max" can be used to specify the maximum limit which is equivalent to no
config for max limits and treated as such. If a given CGroupIODeviceLimit
doesn't contain any non-default configs, the config struct is discarded once
the no limit config is applied to cgroup.
* lookup_blkio_device() is renamed to lookup_block_device().
Signed-off-by: Tejun Heo <htejun@fb.com>
Earlier during the development of unified hierarchy, the populated event was
reported through by the dedicated "cgroup.populated" file; however, the
interface was updated so that it's reported through the "populated" field of
"cgroup.events" file. Update populated event handling logic accordingly.
Since Linux v4.4-rc1, __DEVEL__sane_behavior does not exist anymore and
is replaced by a new fstype "cgroup2".
With this patch, systemd no longer supports the old (unstable) way of
doing unified hierarchy with __DEVEL__sane_behavior and systemd now
requires Linux v4.4 for unified hierarchy.
Non-unified hierarchy is still the default and is unchanged by this
patch.
67e9c74b8a
Support for net_cls.class_id through the NetClass= configuration directive
has been added in v227 in preparation for a per-unit packet filter mechanism.
However, it turns out the kernel people have decided to deprecate the net_cls
and net_prio controllers in v2. Tejun provides a comprehensive justification
for this in his commit, which has landed during the merge window for kernel
v4.5:
https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=bd1060a1d671
As we're aiming for full support for the v2 cgroup hierarchy, we can no
longer support this feature. Userspace tool such as nftables are moving over
to setting rules that are specific to the full cgroup path of a task, which
obsoletes these controllers anyway.
This commit removes support for tweaking details in the net_cls controller,
but keeps the NetClass= directive around for legacy compatibility reasons.
gcc is confused by the common idiom of
return errno ? -errno : -ESOMETHING
and thinks a positive value may be returned. Replace this condition
with errno > 0 to help gcc and avoid many spurious warnings. I filed
a gcc rfe a long time ago, but it hard to say if it will ever be
implemented [1].
Both conventions were used in the codebase, this change makes things
more consistent. This is a follow up to bcb161b023.
[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61846
We need to check the same thing in multiple tests. Use a shared
macro to make it easier to update the list of errnos.
Change the errno code for "unitialized cgroup fs" for ENOMEDIUM.
Exec format error looks like something more serious.
This fixes test-execute invocation in mock.
Lennart Poettering