This does not yet support directory images properly
as systemd itself does not support unpriv directory
images properly yet.
The user profiles are a copy of the system profiles but without
DynamicUser=yes (can't be used by user managers) and without
ProtectHome=yes (this masks /home which breaks StateDirectory= which
is lcoated inside /home)
In order to make "freeze" attacks against the update logic harder let's
add the ability to encode a "Best Before" date into SHA256SUMS directory
listings: if the current time is already beyond that time, we'll ignore
the SHA256SUMS as "stale" and fail the upgrade. Or in other words: the
freeze attack will now result in a client-side error eventually, instead
of success state.
The best before data is encoded in an optional pseudo-file listed in SHA256SUMS:
any file named BEST-BEFORE-YYYY-MM-DD.
Add an option to generate the expected volume key hash for LUKS volumes
by systemd-repart
and put it to crypttab, make systemd-cryptsetup check it upon attaching.
The format of the hash
matches what's currently being measured to TPM2 PCR with
tpm2-measure-pcr=.
Closes#40123
The expected hash (SHA265 HMAC signature) uses the exact same algorithm which
is used to calculate sha256 PCR bank digest when 'tpm2-measure-pcr=' is used.
When attaching images generate a policy in the portable drop-in
that matches the partition types and content found while dissecting,
so that it can no longer be changed later without a reattach.
Main focus was to not introduce breaking change or duplicated argument.
The `--limit-messages=` option that is used under the `monitor` verb is
reused here. However, both `wait` and `monitor` have contradicting
default behaviors, so it's not the cleanest thing to do.
There was some post-commit discussion about the API in #33961, but the
final name adopted in #34928 wasn't that flexible either to fit nicely
here in the `wait` verbe.
Additionally, there wasn't consideration in #34555 for having uniform
behavrious, so we ended with `wait` verb and `--limit-messages=` option,
rather than `receive` verb with default of continuously receving signals
withtout exit so `--limit-messages=` make more sence and be expectable.
RootImageOptions=/ExtensionImages=/MountImages= all support custom
mount options, but mountfsd does not support it. Add varlink
parameters to allow callers to specify mount options so that
those directives can work as expected. Require additional privs via
polkit.
Allow configuring the menu timeout used by sd-boot by setting
`io.systemd.boot.timeout=<value>` as SMBIOS type 11 string.
This takes precedence over the setting from the config file
and the EFI var.
Useful for e.g. integration tests that want to control that
a generic image boots faster.
The systemd-vmspawn man page has a note that the `--notify-ready`
defaults to "true" and then says that this is unlike "systemd-vmspawn"
where it defaults to "false". This appears to be a copy/paste issue
and here we want "unlike systemd-nspawn".
Setting Restart=0 seems reasonable to have no delay on the first
restart, if you do not realize this is impossible with an exponential
restart. So explicitly mention that RestartSec must be set.
RootImageOptions=/ExtensionImages=/MountImages= all support
custom mount options, use the new mountfsd parameters to
configure them if they are specified.
This requires additioanl privileges via polkit due to security
implications of mount options, so document an example policy
that allows to use the nosuid mount option.
Currently mount options are handled in such a way that the first
definition for a given partition wins, and documented as such.
Change them so that they behave like other options, and the
last specified wins.
Applies to RootImageOptions=, MountImages= and ExtensionImages=.
Switch from a linked list to an array indexed by the partition
specifier to store them.
Transparent Hugepages (THP) is a Linux kernel feature that manages
memory using larger pages (2MB on x86, compared to the default 4KB).
The main goal is to improve memory management efficiency and system
performance, especially for memory-intensive applications.
However, it can cause drawbacks in some scenarios, such as memory
regression and latency spikes. THP policy is governed for the entire
system via /sys/kernel/mm/transparent_hugepage/enabled.
However, it can be overridden for individual workloads via prctl(2)
call.
MemoryTHP= is used to disable THPs at exec-invoke to stop
providing THPs for workloads where the drawbacks outweigh the advantages.
When set to "disable", MemoryTHP= disables THPs completely for the
process, irrespecitive of global THP controls.
When set to "madvise", MemoryTHP= disables THPs for the process except
when specifically madvised by the process with MADV_HUGEPAGE or MADV_COLLAPSE.
As announced by a few releases now, finally drop support for
sysvinit scripts.
Keep rc-local generator for now, as it's really a distinct
feature even though from the same era.
They are separate and independent settings, so adjust meson rules
and unit files accordingly. It is possible to enable support for
rc-local script without support for sysvinit scripts, and viceversa.
This will become useful later when sysvinit scripts support is
removed.
The primary motivation for this is to allow users of ukify to build
UKI-like objects, without having them later be detected as a UKI by
tools like kernel-install and bootctl.
The common code used by these tools to determine if a PE binary is a UKI
checks that both .osrel and .linux sections are present. Hence, adding
a mechansim to skip .osrel provides a way to avoid being labeled a UKI.
This has been supported since systemd v242 (specifically commit
7619cb32f0 if I’m not mistaken; added to NEWS in commit 4107452e51), but
the man page still claimed otherwise.
MultiPathRoute= can now specify device-only nexthops without a gateway
address, e.g. MultiPathRoute=@wg0. This enables ECMP configurations over
interfaces that don't use gateway addresses, such as WireGuard tunnels.
The syntax is extended from "address[@device] [weight]" to
"[address]@device [weight]". The address is now optional, but at least
one of gateway or device must be specified. The @ symbol must still be
present for device-only routes, making the syntax unambiguous: @wg0
specifies a device, while a bare IP address specifies a gateway.
Device-only nexthops are only available for IPv4 routes. Device-only
multipath routes for IPv6 are not supported by the kernel's netlink
interface and will be rejected with a warning.
This change is fully backwards compatible. All existing configurations
continue to work unchanged, as they always included a gateway address.
Closes#39699.
The UX of registering with the user session machined
instance is much better as there won't be an authorization
prompt. To make that available for users, let's add --user
and --system switches for vmspawn. For backwards compat, we'll
still try to register with the system machined instance if the
user machined instance is not available.
Allow CPUSchedulingPolicy to be set to "ext". SCHED_EXT is a new
scheduling policy in Linux v6.12 that allows processes to be scheduled
using custom BPF schedulers instead of the default in-kernel ones.
Selectively setting the SCHED_EXT policy is useful for systems running
in "partial mode" where not all processes are run using a custom
scheduler.
Fallback to SCHED_OTHER and print an error message for systems where
SCHED_EXT isn't available.
First, don't require blocking SIGCHLD for WEXITED. We watch for WEXITED
via pidfd instead of signalfd, so no need to insist on blocking SIGCHLD
anymore if we're only watching for WEXITED.
Second, do a proper check to see if the kernel autoreaping logic is
enabled. That has nothing to do with SIGCHLD being blocked for the current
thread or not. Instead, the kernel autoreaping logic is enabled either if
the disposition is set to SIG_IGN or if the SA_NOCLDWAIT flag is enabled.
Loading images is, generally speaking, the slowest part of sd-executor
when spawning a service. This is due to multiple factors. dm-verity is
obviously a big part of the cost, but dissecting in general via libblkid
also can take a lot of time, due to probing the images and their
filesystems.
A performance test doing service restarts in a row shows these
results, ran on a production system (low power and slow ARM64 SOC) with
a real production service, show the following service interruption
intervals:
```
Count | P50 (ms) | P90 (ms) | P95 (ms) | P99 (ms) | P99.9 (ms)
507 | 1,532 | 1,726 | 2,548 | 4,112 | 4,592
```
One iteration is 507 restarts in this case, but this has ran hundreds
of times and the results are always in line within margin of error.
This also holds true for metrics from live systems, same numbers.
Between 1.0s and 1.2s can be attributed by profiling to the time needed
for the service code itself to start up and sd_notify, the rest is spent
inside systemd's code.
This means there is currently a tradeoff for services - either use
secure
images, or make restarting fast. Downtime of services is a very
important
metric, as for many cases this directly translates to outages, total or
partial (blackouot or greyout).
In order to facilitate using secure images without downsides, skip the
slow dissect steps (probing, loop devices, etc) when the configured
image is a single filesystem dm-verity image with a policy that pins it
to a single filesystem, and an already existing and open dm-verity
device
can be found and reused.
This allows orchestrators to pre-open images on download, before
restarting
services, to minimize downtimes.
Useful to extract a certificate from a hardware token to a file, for
example in mkosi to ship the certificate from a hardware token in
/usr/lib/verity.d in an image
Lennart Poettering
Daan De Meyer
Vitaly Kuznetsov
Luca Boccassi
Yu Watanabe
zefr0x
Mike Yuan
Luca Boccassi
Michael Vogt
Zbigniew Jędrzejewski-Szmek
Christian Hesse
Matthijs Kooijman
Usama Arif
Alexis-Emmanuel Haeringer
Tobias Stoeckmann
Nick Rosbrook
Lucas Werkmeister
Ulrich Ölmann
Luiz Amaral
Matt Fleming
seidlerv