Today listen file descriptors created by socket unit don't get passed to
commands in Exec{Start,Stop}{Pre,Post}= socket options.
This prevents ExecXYZ= commands from accessing the created socket FDs to do
any kind of system setup which involves the socket but is not covered by
existing socket unit options.
One concrete example is to insert a socket FD into a BPF map capable of
holding socket references, such as BPF sockmap/sockhash [1] or
reuseport_sockarray [2]. Or, similarly, send the file descriptor with
SCM_RIGHTS to another process, which has access to a BPF map for storing
sockets.
To unblock this use case, pass ListenXYZ= file descriptors to ExecXYZ=
commands as listen FDs [4]. As an exception, ExecStartPre= command does not
inherit any file descriptors because it gets invoked before the listen FDs
are created.
This new behavior can potentially break existing configurations. Commands
invoked from ExecXYZ= might not expect to inherit file descriptors through
sd_listen_fds protocol.
To prevent breakage, add a new socket unit parameter,
PassFileDescriptorsToExec=, to control whether ExecXYZ= programs inherit
listen FDs.
[1] https://docs.kernel.org/bpf/map_sockmap.html
[2] https://lore.kernel.org/r/20180808075917.3009181-1-kafai@fb.com
[3] https://man.archlinux.org/man/socket.7#SO_INCOMING_CPU
[4] https://www.freedesktop.org/software/systemd/man/latest/sd_listen_fds.html
And rename it to id128_get_boot_for_machine().
This also splits out id128_get_boot() from sd_id128_get_boot(), and
make id128_get_boot_for_machine() use it.
- split it into small pieces to reduce indentation,
- return earlier if no filter specified,
- use journal_add_matchf() and journal_add_match_pair(),
chase_and_stat(), is_symlink(), and sd_device_new_from_devname(),
- replace sd_device_get_devname() + stat() with sd_device_get_devnum(),
- refuse to install filter by path with --root, --image, or --machine
at least now, as previously it has not worked anyway.
- add missing assertions,
- use set_ensure_consume() and strv_consume(),
- use string_hash_ops_free and _cleanup_set_free_,
- use strv_fnmatch_full(),
- replace unused counter with a boolean flag,
- return earlier if no unit filtering is requested.
Drop connections and caches and reload config from files, to allow
for low-interruptions updates, and hook up to the usual SIGHUP and
ExecReload=. Mark servers and services configured directly via D-Bus
so that they can be kept around, and only the configuration file
settings are dropped and reloaded.
Fixes https://github.com/systemd/systemd/issues/17503
Fixes https://github.com/systemd/systemd/issues/20604
* efi: Measure into both CC and TPM if available.
It's possible that both measurement protocols are made available, so
instead of assuming only one or the other are available, measure into
both to avoid a problem like CVE-2021-42299.
Signed-off-by: Dionna Glaze <dionnaglaze@google.com>
In case kernel-install is called to install what's already a UKI, 60-ukify
must be skipped. E.g. when distro shipped and signed UKI is being installed,
it is counter-productive to try to rebuild it with ukify. Also, the existing
script is not ready to handle UKIs with embedded initramfs.
Note, it is already possible to disable 60-ukify by setting
KERNEL_INSTALL_UKI_GENERATOR to something but in case it is not set, 60-ukify
assumes it should run.
The testuser user is only needed for integration tests,
which are used in the system user and this config
can be provided as drop-ins instead of inline in postinst scripts.
Jakub Sitnicki
Yu Watanabe
Antonio Alvarez Feijoo
Luxiter
Luca Boccassi
Mike Yuan
Zbigniew Jędrzejewski-Szmek
dependabot[bot]
Dionna Amalie Glaze
Frantisek Sumsal
Gaël Donval
Vitaly Kuznetsov
Daan De Meyer
Richard Maw