The systemd-modeset tool is meant to debug grdev issues. It simply
displays morphing colors on any found display. This is pretty handy to
look for tearing in the backends and debug hotplug issues.
Note that this tool requires systemd-logind to be compiled from git
(there're important fixes that haven't been released, yet).
The grdev-drm backend manages DRM cards for grdev. Any DRM card with
DUMB_BUFFER support can be used. So far, our policy is to configure all
available connectors, but keep pipes inactive as long as users don't
enable the displays on top.
We hard-code double-buffering so far, but can easily support
single-buffering or n-buffering. We also require XRGB8888 as format as
this is required to be supported by all DRM drivers and it is what VTs
use. This allows us to switch from VTs to grdev via page-flips instead of
deep modesets.
There is still a lot room for improvements in this backend, but it works
smoothly so far so more enhanced features can be added later.
The grdev layer provides graphics-device access via the
libsystemd-terminal library. It will be used by all terminal helpers to
actually access display hardware.
Like idev, the grdev layer is built around session objects. On each
session object you add/remove graphics devices as they appear and vanish.
Any device type can be supported via specific card-backends. The exported
grdev API hides any device details.
Graphics devices are represented by "cards". Those are hidden in the
session and any pipe-configuration is automatically applied. Out of those,
we configure displays which are then exported to the API user. Displays
are meant as lowest hardware entity available outside of grdev. The
underlying pipe configuration is fully hidden and not accessible from the
outside. The grdev tiling layer allows almost arbitrary setups out of
multiple pipes, but so far we only use a small subset of this. More will
follow.
A grdev-display is meant to represent real connected displays/monitors.
The upper level screen arrangements are user policy and not controlled by
grdev. Applications are free to apply any policy they want.
Real card-backends will follow in later patches.
Rather than forcing gcc to always produce colorized error messages
whether on tty or not, enable automatic colorization by ensuring
GCC_COLORS is set to a non-empty string.
Doing it this way removes the need for workarounds in ~/.emacs or
~/.vimrc for "M-x compile" or ":make", respectively, to work.
Like systemd-subterm, this new systemd-evcat tool should only be used to
debug libsystemd-terminal. systemd-evcat attaches to the running session
and pushes all evdev devices attached to the current session into an
idev-session. All events of the created idev-devices are then printed to
stdout for input-event debugging.
The idev-keyboard object provides keyboard devices to the idev interface.
It uses libxkbcommon to provide proper keymap support.
So far, the keyboard implementation is pretty straightforward with one
keyboard device per matching evdev element. We feed everything into the
system keymap and provide proper high-level keyboard events to the
application. Compose-features and IM need to be added later.
The evdev-element provides linux evdev interfaces as idev-elements. This
way, all real input hardware devices on linux can be used with the idev
interface.
We use libevdev to interface with the kernel. It's a simple wrapper
library around the kernel evdev API that takes care to resync devices
after kernel-queue overflows, which is a rather non-trivial task.
Furthermore, it's a well tested interface used by all other major input
users (Xorg, weston, libinput, ...).
Last but not least, it provides nice keycode to keyname lookup tables (and
vice versa), which is really nice for debugging input problems.
The idev-interface provides input drivers for all libsystemd-terminal
based applications. It is split into 4 main objects:
idev_context: The context object tracks global state of the input
interface. This will include data like system-keymaps,
xkb contexts and more.
idev_session: A session serves as controller for a set of devices.
Each session on an idev-context is independent of each
other. The session is also the main notification object.
All events raised via idev are reported through the
session interface. Apart of that, the session is a
pretty dumb object that just contains devices.
idev_element: Elements provide real hardware in the idev stack. For
each hardware device, one element is added. Elements
have no knowledge of higher-level device types, they
only provide raw input data to the upper levels. For
example, each evdev device is represented by a different
element in an idev session.
idev_device: Devices are objects that the application deals with. An
application is usually not interested in elements (and
those are hidden to applications), instead, they want
high-level input devices like keyboard, touchpads, mice
and more. Device are the high-level interface provided
by idev. Each device might be fed by a set of elements.
Elements drive the device. If elements are removed,
devices are destroyed. If elements are added, suitable
devices are created.
Applications should monitor the system for sessions and hardware devices.
For each session they want to operate on, they create an idev_session
object and add hardware to that object. The idev interface requires the
application to monitor the system (preferably via sysview_*, but not
required) for hardware devices. Whenever hardware is added to the idev
session, new devices *might* be created. The relationship between hardware
and high-level idev-devices is hidden in the idev-session and not exposed.
Internally, the idev elements and devices are virtual objects. Each real
hardware and device type inherits those virtual objects and provides real
elements and devices. Those types will be added in follow-up commits.
Data flow from hardware to the application is done via idev_*_feed()
functions. Data flow from applications to hardware is done via
idev_*_feedback() functions. Feedback is usually used for LEDs, FF and
similar operations.
We're going to need multiple binaries that provide session-services via
logind device management. To avoid re-writing the seat/session/device
scan/monitor interface for each of them, this commit adds a generic helper
to libsystemd-terminal:
The sysview interface scans and tracks seats, sessions and devices on a
system. It basically mirrors the state of logind on the application side.
Now, each session-service can listen for matching sessions and
attach to them. On each session, managed device access is provided. This
way, it is pretty simple to write session-services that attach to multiple
sessions (even split across seats).
hibernate-resume-generator understands resume= kernel command line parameter
and instantiates the systemd-resume@.service accordingly if it is passed.
This enables resume from hibernation using device specified on the kernel
command line, and it may be specified either as "/dev/disk/by-foo/bar"
or "FOO=bar", not only "/dev/sdXY" which is understood by the in-kernel
implementation.
So now resume= is brought on par with root= in terms of possible ways to
specify a device.
This can be used to initiate a resume from hibernation by path to a swap
device containing the hibernation image.
The respective templated unit is also added. It is instantiated using
path to the desired resume device.
When this system-wide start-up timeout is hit we execute one of the
failure actions already implemented for services that fail.
This should not only be useful on embedded devices, but also on laptops
which have the power-button reachable when the lid is closed. This
devices, when in a backpack might get powered on by accident due to the
easily reachable power button. We want to make sure that the system
turns itself off if it starts up due this after a while.
When the system manages to fully start-up logind will suspend the
machine by default if the lid is closed. However, in some cases we don't
even get as far as logind, and the boot hangs much earlier, for example
because we ask for a LUKS password that nobody ever enters.
Yeah, this is a real-life problem on my Yoga 13, which has one of those
easily accessible power buttons, even if the device is closed.
In contrast to the DHCP/IPv4LL/ICMP6 APIs sd-network is not a protocol
implementation but a client API for networkd, hence move it into
libsystemd proper.
In the long run this should become a full fledged client to networkd
(but not before networkd learns bus support). For now, just pull
interesting data out of networkd, udev, and rtnl and present it to the
user, in a simple but useful output.
We are unlikely to evert support most of them, but we can at least
display the types properly.
The list is taken from the IANA list.
The table of number->name mappings is converted to a switch
statement. gcc does a nice job of optimizing lookup (when optimization
is enabled).
systemd-resolve-host -t is now case insensitive.
We now maintain two lists of DNS servers: system servers and fallback
servers.
system servers are used in combination with any per-link servers.
fallback servers are only used if there are no system servers or
per-link servers configured.
The system server list is supposed to be populated from a foreign tool's
/etc/resolv.conf (not implemented yet).
Also adds a configuration switch for LLMNR, that allows configuring
whether LLMNR shall be used simply for resolving or also for responding.
Make sure we format UTF-8 labels as IDNA when writing them to DNS
packets, and as native UTF-8 when writing them to mDNS or LLMNR packets.
When comparing or processing labels always consider native UTF-8 and
IDNA formats equivalent.
LOC records have a version field. So far only version 0 has been
published, but if a record with a different version was encountered,
our only recourse is to treat it as an unknown type. This is
implemented with the 'unparseable' flag, which causes the
serialization/deserialization and printing function to cause the
record as a blob. The flag can be used if other packet types cannot be
parsed for whatever reason.
This tool will warn about misspelt directives, unknown sections, and
non-executable commands. It will also catch the common mistake of
using Accept=yes with a non-template unit and vice versa.
https://bugs.freedesktop.org/show_bug.cgi?id=56607
Commit 637f421e5c ("cgroups: always propagate controller membership
to siblings") changed the mask propagation logic, but the test wasn't
updated.
Move to normal tests from manual tests, it should not touch the system
anymore.
Also add a bit of debugging output to help diagnose problems,
add missing units, and simplify cppflags.
Move test-engine to normal tests from manual tests, it should now
work without destroying the system.
Our version has evolved independently of the original table
in systemd-config-keyboard, so it cannot be ever regenerated from
original upstream. Remove script to avoid confusion.
The unifont layer of libsystemd-terminal provides a fallback font for
situations where no system-fonts are available, or if you don't want to
deal with traditional font-formats for some reasons.
The unifont API mmaps a pre-compiled bitmap font that was generated out of
GNU-Unifont font-data. This guarantees, that all users of the font will
share the pages in memory. Furthermore, the layout of the binary file
allows accessing glyph data in O(1) without pre-rendering glyphs etc. That
is, the OS can skip loading pages for glyphs that we never access.
Note that this is currently a test-run and we want to include the binary
file in the GNU-Unifont package. However, until it was considered stable
and accepted by the maintainers, we will ship it as part of systemd. So
far it's only enabled with the experimental --enable-terminal, anyway.
The systemd-subterm example is a stacked terminal that shows how to
use sd-term. Instead of rendering images and displaying it via X11/etc.,
it uses its parent terminal to display the page (terminal-emulator inside
a terminal-emulator) (like GNU-screen and friends do).
This is only for testing and not installed system-wide!
The screen-layer represents the terminal-side (compared to the host-side).
It connects term_parser with term_page and implements all the required
control sequences.
We do not implement all available control sequences. Even though our
parser recognizes them, there is no need to handle them. Most of them are
legacy or unused. We try to be as compatible to xterm, so if we missed
something, we can implement it later. However, all the VT510 / VT440 stuff
can safely be skipped (who needs terminal macros? WTF?).
The keyboard-handling is still missing. It will be added once
systemd-console is available and we pulled in the key-definitions.
The term-parser is used to parse any input from TTY-clients. It reads CSI,
DCS, OSC and ST control sequences and normal escape sequences. It doesn't
do anything with the parsed data besides detecting the sequence and
returning it. The caller has to react to them.
The parser also comes with its own UTF-8 helpers. The reason for that is
that we don't want to assert() or hard-fail on parsing errors. Instead,
we treat any invalid UTF-8 sequences as ISO-8859-1. This allows pasting
invalid data into a terminal (which cannot be controlled through the TTY,
anyway) and we still deal with it in a proper manner.
This is _required_ for 8-bit and 7-bit DEC modes (including the g0-g3
mappings), so it's not just an ugly fallback because we can (it's still
horribly ugly but at least we have an excuse).
Rather than refetching the link information on ever event, we liston to
rtnl to track them. Much code stolen from resolved.
This will allow us to simplify the sd-network api and don't expose
information available over rtnl.
This commit introduces libsystemd-ui, a systemd-internal helper library
that will contain all the UI related functionality. It is going to be used
by systemd-welcomed, systemd-consoled, systemd-greeter and systemd-er.
Further use-cases may follow.
For now, this commit only adds terminal-page handling based on lines only.
Follow-up commits will add more functionality.
This Pty API wraps the ugliness that is POSIX PTY. It takes care of:
- edge-triggered HUP handling (avoid heavy CPU-usage on vhangup)
- HUP vs. input-queue draining (handle HUP _after_ draining the whole
input queue)
- SIGCHLD vs. HUP (HUP is no reliable way to catch PTY deaths, always
use SIGCHLD. Otherwise, vhangup() and friends will break.)
- Output queue buffering (async EPOLLOUT handling)
- synchronous setup (via Barrier API)
At the same time, the PTY API does not execve(). It simply fork()s and
leaves everything else to the caller. Usually, they execve() but we
support other setups, too.
This will be needed by multiple UI binaries (systemd-console, systemd-er,
...) so it's placed in src/shared/. It's not strictly related to
libsystemd-terminal, so it's not included there.
The Barrier-API simplifies cross-fork() synchronization a lot. Replace the
hard-coded eventfd-util implementation and drop it.
Compared to the old API, Barriers also handle exit() of the remote side as
abortion. This way, segfaults will not cause the parent to deadlock.
EINTR handling is currently ignored for any barrier-waits. This can easily
be added, but it isn't needed so far so I dropped it. EINTR handling in
general is ugly, anyway. You need to deal with pselect/ppoll/... variants
and make sure not to unblock signals at the wrong times. So genrally,
there's little use in adding it.
The "Barrier" object is a simple inter-process barrier implementation. It
allows placing synchronization points and waiting for the other side to
reach it. Additionally, it has an abortion-mechanism as second-layer
synchronization to send abortion-events asynchronously to the other side.
The API is usually used to synchronize processes during fork(). However,
it can be extended to pass state through execve() so you could synchronize
beyond execve().
Usually, it's used like this (error-handling replaced by assert() for
simplicity):
Barrier b;
r = barrier_init(&b);
assert_se(r >= 0);
pid = fork();
assert_se(pid >= 0);
if (pid == 0) {
barrier_set_role(&b, BARRIER_CHILD);
...do child post-setup...
if (CHILD_SETUP_FAILED)
exit(1);
...child setup done...
barrier_place(&b);
if (!barrier_sync(&b)) {
/* parent setup failed */
exit(1);
}
barrier_destroy(&b); /* redundant as execve() and exit() imply this */
/* parent & child setup successful */
execve(...);
}
barrier_set_role(&b, BARRIER_PARENT);
...do parent post-setup...
if (PARENT_SETUP_FAILED) {
barrier_abort(&b); /* send abortion event */
barrier_wait_abortion(&b); /* wait for child to abort (exit() implies abortion) */
barrier_destroy(&b);
...bail out...
}
...parent setup done...
barrier_place(&b);
if (!barrier_sync(&b)) {
...child setup failed... ;
barrier_destroy(&b);
...bail out...
}
barrier_destroy(&b);
...child setup successfull...
This is the most basic API. Using barrier_place() to place barriers and
barrier_sync() to perform a full synchronization between both processes.
barrier_abort() places an abortion barrier which superceeds any other
barriers, exit() (or barrier_destroy()) places an abortion-barrier that
queues behind existing barriers (thus *not* replacing existing barriers
unlike barrier_abort()).
This example uses hard-synchronization with wait_abortion(), sync() and
friends. These are all optional. Barriers are highly dynamic and can be
used for one-way synchronization or even no synchronization at all
(postponing it for later). The sync() call performs a full two-way
synchronization.
The API is documented and should be fairly self-explanatory. A test-suite
shows some special semantics regarding abortion, wait_next() and exit().
Internally, barriers use two eventfds and a pipe. The pipe is used to
detect exit()s of the remote side as eventfds do not allow that. The
eventfds are used to place barriers, one for each side. Barriers itself
are numbered, but the numbers are reused once both sides reached the same
barrier, thus you cannot address barriers by the index. Moreover, the
numbering is implicit and we only store a counter. This makes the
implementation itself very lightweight, which is probably negligible
considering that we need 3 FDs for a barrier..
Last but not least: This barrier implementation is quite heavy. It's
definitely not meant for fast IPC synchronization. However, it's very easy
to use. And given the *HUGE* overhead of fork(), the barrier-overhead
should be negligible.
Directory src/journal has become one of the largest directories,
and since systemd-journal-gatewayd, systemd-journal-remote, and
forthcoming systemd-journal-upload are all closely related, create
a separate directory for them.
Let's turn resolved into a something truly useful: a fully asynchronous
DNS stub resolver that subscribes to network changes.
(More to come: caching, LLMNR, mDNS/DNS-SD, DNSSEC, IDN, NSS module)
It is already in nodist_systemunit_DATA and if it is
shipped, it contains the hardcoded path to systemctl
which will cause it to fail to start when
rootprefix != prefix and rootbindir != bindir.
This reverts commit 0c26bfc3d2.
src/core/org.freedesktop.systemd1.policy.in.in depends on values which
are specified at configure time, so we cannot ship the corresponding
policy file in the tarball.
Since we need to regenerate one policy file, we might as well generate
them all.
The systemd-escape utility might be used during early boot (e.g. when
being triggered from udev rules), so move it to rootbindir to support
systems with a split /usr setup.
This patch adds supports networkd to configure bond mode
during creation via persistent conf. Mode can be configured
with conf param 'Mode'. A new section Bond is added to the
conf to support bond mode.
These modes can be configured now.
balance-rr
active-backup
balance-xor
broadcast
802.3ad
balance-tlb
balance-alb
Example conf file: test-bond.conf
[NetDev]
Name=bond1
Kind=bond
[Bond]
Mode=balance-xor
Test case:
1. start networkd service:
12: bond1: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN mode DEFAULT group default
link/ether 22:89:6c:47:23:d2 brd ff:ff:ff:ff:ff:ff
2. find bond mode:
cat /proc/net/bonding/bond1
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)
Bonding Mode: load balancing (xor)
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 0
Up Delay (ms): 0
Down Delay (ms): 0
Changes:
1. Added file networkd-bond.c
2. Bond mode enum BondMode
3. conf section [Bond]
[tomegun: whitespace]
As Zbigniew pointed out a new ConditionFirstBoot= appears like the nicer
way to hook in systemd-firstboot.service on first boots (those with /etc
unpopulated), so let's do this, and get rid of the generator again.
A new tool "systemd-firstboot" can be used either interactively on boot,
where it will query basic locale, timezone, hostname, root password
information and set it. Or it can be used non-interactively from the
command line when prepareing disk images for booting. When used
non-inertactively the tool can either copy settings from the host, or
take settings on the command line.
$ systemd-firstboot --root=/path/to/my/new/root --copy-locale --copy-root-password --hostname=waldi
The tool will be automatically invoked (interactively) now on first boot
if /etc is found unpopulated.
This also creates the infrastructure for generators to be notified via
an environment variable whether they are running on the first boot, or
not.
This is useful to test the behaviour of the compressor for various buffer
sizes.
Time is limited to a minute per compression, since otherwise, when LZ4
takes more than a second which is necessary to reduce the noise, XZ
takes more than 10 minutes.
% build/test-compress-benchmark (without time limit)
XZ: compressed & decompressed 2535300963 bytes in 794.57s (3.04MiB/s), mean compresion 99.95%, skipped 3570 bytes
LZ4: compressed & decompressed 2535303543 bytes in 1.56s (1550.07MiB/s), mean compresion 99.60%, skipped 990 bytes
% build/test-compress-benchmark (with time limit)
XZ: compressed & decompressed 174321481 bytes in 60.02s (2.77MiB/s), mean compresion 99.76%, skipped 3570 bytes
LZ4: compressed & decompressed 2535303543 bytes in 1.63s (1480.83MiB/s), mean compresion 99.60%, skipped 990 bytes
It appears that there's a bug in lzma_end where it leaks 32 bytes.
Add liblz4 as an optional dependency when requested with --enable-lz4,
and use it in preference to liblzma for journal blob and coredump
compression. To retain backwards compatibility, XZ is used to
decompress old blobs.
Things will function correctly only with lz4-119.
Based on the benchmarks found on the web, lz4 seems to be the best
choice for "quick" compressors atm.
For pkg-config status, see http://code.google.com/p/lz4/issues/detail?id=135.
This patch introduces TUN/TAP device creation support
to networkd.
Example conf to create a tap device:
file: tap.netdev
------------------
[NetDev]
Name=tap-test
Kind=tap
[Tap]
OneQueue=true
MultiQueue=true
PacketInfo=true
User=sus
Group=sus
------------------
Test:
1. output of ip link
tap-test: tap pi one_queue UNKNOWN_FLAGS:900 user 1000 group 1000
id:
uid=1000(sus) gid=10(wheel) groups=10(wheel),1000(sus)
context=unconfined_u:unconfined_r:unconfined_t:s0-s0:c0.c1023
Modifications:
Added:
1. file networkd-tuntap.c
3. netdev kind NETDEV_KIND_TUN and NETDEV_KIND_TAP
2. Tun and Tap Sections and config params to parse
conf and gperf conf parameters
[tomegun: tweak the 'kind' checking for received ifindex]
This new tool is based on "sd-path", a new (so far unexported) API for
libsystemd, that can hopefully grow into a workable API covering /opt
and more one day.
Instead of adjusting job timeouts in the core, let fstab-generator
write out a dropin snippet with the appropriate JobTimeout.
x-systemd-device.timeout option is removed from Options= line
in the generated unit.
The functions to write dropins are moved from core/unit.c to
shared/dropin.c, to make them available outside of core.
generator.c is moved to libsystemd-label, because it now uses
functions defined in dropin.c, which are in libsystemd-label.
When disk space taken up by coredumps grows beyond a configured limit
start removing the oldest coredump of the user with the most coredumps,
until we get below the limit again.
So that building from an archive works even if intltool is not present.
The README file already mentioned that intltool should only be required
when building from git.
Tested: Built it from the distribution archive on a host without intltool.
$ ./configure --enable-polkit
$ make
Running "make dist" requires --enable-compat-libs since DIST_SOURCES will list
generated files such as libsystemd-daemon.c.
Tested:
$ ./configure && make && make dist
*** compat-libs must be enabled in order to make dist
make: *** [dist-check-compat-libs] Error 1
Running "make dist" requires Python support since some of the man page sources
(such as man/systemd.index.xml and man/systemd.directives.xml) are generated by
Python scripts, so break "make dist" and give an useful error message when
Python or the Python lxml module is not available.
Tested:
$ ./configure --without-python && make && make dist
*** python and python-lxml module must be installed and enabled in order to make dist
make: *** [dist-check-python] Error 1
Python support is pretty much essential to create man pages, so we should make
sure that distcheck will request it during configure.
Tested: Successfully ran "make distcheck" and confirmed --with-python was
present in the ./configure run inside the unpacked distribution directory.
File src/python-systemd/id128-constants.h is auto generated and its generation
does not require special tools, only sed. There is no point in bundling it in
the distribution archive, so let's mark it as nodist_ to have it excluded.
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=80006
Tested: Successfully ran "make dist" after ./configure --without-python.
The sysusers.d/systemd.conf configuration file was originally introduced in
commit 1b99214789, but it was not marked for cleanup. This caused distcheck
to complain about the file not being removed by distcleam.
Tested: Successfully ran "make distcheck" with this patchset.
It was incorrectly looking for a file in src/libsystemd-network/ when the file was actually deployed to src/systemd/ instead. This broke "make dist".
Tested: "make dist" works again after this patchset is applied.
Fixes: f20a35cc0d
Makefile.am had a reference to it but it none of the sources included it.
Tested: "make dist" works again after this patchset is applied.
Fixes: 2ea8857eff
debug-generator can mask specific units if they are specified on the
kernel command line with systemd.mask=.
debug-generator can pull in debug-shell.service is systemd.debug-shell
is passed on the kernel command line.
Create a structure describing a DHCPv6 lease. Add internal functions
for creating a new lease and accessing the server ID, preference and
IAID. Provide functions for clearing addresses and associated timers.
External users are initially given only the capabilities of
referencing and unreferencing the lease structure.
Verify the Solicit message created by the DHCPv6 client code.
Provide local variants for detect_vm(), detect_container() and
detect_virtualization() defined in virt.h. This makes the DHCPv6
library believe it is run in a container and does not try to request
interface information from udev for the non-existing interface index
used by the test case code.
Add option appending and parsing. DHCPv6 options are not aligned, thus
the option handling code must be able to handle options starting at
any byte boundary.
Add a test case for the basic option handling.
Feed a Router Advertisement to the code and expect proper events
each time. The sending part is ignored, as all of it is static code
in the real dhcp_network_icmp6_send_rs() function.
Provide functions to bind the ICMPv6 socket to the approriate interface
and set multicast sending and receiving according to RFC 3493, section
5.2. and RFC 3542, sections 3. and 3.3. Filter out all ICMPv6 messages
except Router Advertisements for the socket in question according to
RFC 3542, section 3.2.
Send Router Solicitations to the all routers multicast group as
described in RFC 4861, section 6. and act on the received Router
Advertisments according to section 6.3.7.
Implement a similar API for ICMPv6 handling as is done for DHCPv4 and
DHCPv6.
Introduce a new configuration file /etc/systemd/coredump.conf to
configure when to place coredumps in the journal and when on disk.
Since the coredumps are quite large, default to storing them only on
disk.
When an address is configured to be all zeroes, networkd will now
automatically find a locally unused network of the right size from a
list of pre-configured pools. Currently those pools are 10.0.0.0/8,
172.16.0.0/12, 192.168.0.0/16 and fc00::/7, i.e. the network ranges for
private networks. They are compiled in, but should be configurable
eventually.
This allows applying the same configuration to a large number of
interfaces with each time a different IP range block, and management of
these IP ranges is fully automatic.
When allocating an address range from the pool it is made sure the range
is not used otherwise.
This is useful so that distros have something to base their own policy
of. It also useful to make sure that minimal installs always get useful
configuration in place.
With this in place RPMs can make sure that whatever they drop in is
immeidately applied, and not delayed until next reboot.
This also moves systemd-sysusers back to /usr/bin, since hardcoding the
path to /usr/lib in the macros would mean compatibility breaks in
future, should we turn sysusers into a command that is actually OK for
people to call directly. And given that that is quite likely to happen
(since it is useful to prepare images with its --root= switch), let's
just prepare for it.
In order to support offline updates to /usr, we need to be able to run
certain tasks on next boot-up to bring /etc and /var in line with the
updated /usr. Hence, let's devise a mechanism how we can detect whether
/etc or /var are not up-to-date with /usr anymore: we keep "touch
files" in /etc/.updated and /var/.updated that are mtime-compared with
/usr. This means:
Whenever the vendor OS tree in /usr is updated, and any services that
shall be executed at next boot shall be triggered, it is sufficient to
update the mtime of /usr itself. At next boot, if /etc/.updated and/or
/var/.updated is older than than /usr (or missing), we know we have to
run the update tools once. After that is completed we need to update the
mtime of these files to the one of /usr, to keep track that we made the
necessary updates, and won't repeat them on next reboot.
A subsequent commit adds a new ConditionNeedsUpdate= condition that
allows checking on boot whether /etc or /var are outdated and need
updating.
This is an early step to allow booting up with an empty /etc, with
automatic rebuilding of the necessary cache files or user databases
therein, as well as supporting later updates of /usr that then propagate
to /etc and /var again.
systemd-sysusers is a tool to reconstruct /etc/passwd and /etc/group
from static definition files that take a lot of inspiration from
tmpfiles snippets. These snippets should carry information about system
users only. To make sure it is not misused for normal users these
snippets only allow configuring UID and gecos field for each user, but
do not allow configuration of the home directory or shell, which is
necessary for real login users.
The purpose of this tool is to enable state-less systems that can
populate /etc with the minimal files necessary, solely from static data
in /usr. systemd-sysuser is additive only, and will never override
existing users.
This tool will create these files directly, and not via some user
database abtsraction layer. This is appropriate as this tool is supposed
to run really early at boot, and is only useful for creating system
users, and system users cannot be stored in remote databases anyway.
The tool is also useful to be invoked from RPM scriptlets, instead of
useradd. This allows moving from imperative user descriptions in RPM to
declarative descriptions.
The UID/GID for a user/group to be created can either be chosen dynamic,
or fixed, or be read from the owner of a file in the file system, in
order to support reconstructing the correct IDs for files that shall be
owned by them.
This also adds a minimal user definition file, that should be
sufficient for most basic systems. Distributions are expected to patch
these files and augment the contents, for example with fixed UIDs for
the users where that's necessary.
network-pre.target is a passive target that should be pulled in by
services that want to be executed before any network is configured (for
example: firewall scrips).
network-pre.target should be ordered before all network managemet
services (but not be pulled in by them).
network-pre.target should be order after all services that want to be
executed before any network is configured (and be pulled in by them).
Reuses logic from service.c and the rc-local generator.
Note that this drops reading of chkconfig entirely. It also drops reading
runlevels from the LSB headers. The runlevels were only used to check for
runlevels outside of the normal 1-5 range and then add special dependencies
and settings. Special runlevels were dropped in the past so it seemed to be
unused code.
The generator does not know about non-generated units with a value set with
SysVStartPriority=. These are therefor not taken into account when converting
start priority to before/after.
Either become uid/gid of the client we have been forked for, or become
the "systemd-bus-proxy" user if the client was root. We retain
CAP_IPC_OWNER so that we can tell kdbus we are actually our own client.
Rely on modules being built-in or autoloaded on-demand.
As networkd is a network facing service, we want to limits its capabilities,
as much as possible. Also, we may not have CAP_SYS_MODULE in a container,
and we want networkd to work the same there.
Module autoloading does not always work, but should be fixed by the kernel
patch f98f89a0104454f35a: 'net: tunnels - enable module autoloading', which
is currently in net-next and which people may consider backporting if they
want tunneling support without compiling in the modules.
Early adopters may also use a module-load.d snippet and order
systemd-modules-load.service before networkd to force the module
loading of tunneling modules.
This sholud fix the various build issues people have reported.
This allows us to run networkd mostly unpriviliged with the exception of
CAP_NET_* and CAP_SYS_MODULE. I'd really like to get rid of the latter
though...
Instead of accessing /proc/1/environ directly, trying to read the
$container variable from it, let's make PID 1 save the contents of that
variable to /run/systemd/container. This allows us to detect containers
without the need for CAP_SYS_PTRACE, which allows us to drop it from a
number of daemons and from the file capabilities of systemd-detect-virt.
Also, don't consider chroot a container technology anymore. After all,
we don't consider file system namespaces container technology anymore,
and hence chroot() should be considered a container even less.
nspawn and the container child use eventfd to wait and notify each other
that they are ready so the container setup can be completed.
However in its current form the wait/notify event ignore errors that
may especially affect the child (container).
On errors the child will jump to the "child_fail" label and terminate
with _exit(EXIT_FAILURE) without notifying the parent. Since the eventfd
is created without the "EFD_NONBLOCK" flag, this leaves the parent
blocking on the eventfd_read() call. The container can also be killed
at any moment before execv() and the parent will not receive
notifications.
We can fix this by using cheap mechanisms, the new high level eventfd
API and handle SIGCHLD signals:
* Keep the cheap eventfd and EFD_NONBLOCK flag.
* Introduce eventfd states for parent and child to sync.
Child notifies parent with EVENTFD_CHILD_SUCCEEDED on success or
EVENTFD_CHILD_FAILED on failure and before _exit(). This prevents the
parent from waiting on an event that will never come.
* If the child is killed before execv() or before notifying the parent,
we install a NOP handler for SIGCHLD which will interrupt blocking calls
with EINTR. This gives a chance to the parent to call wait() and
terminate in main().
* If there are no errors, parent will block SIGCHLD, restore default
handler and notify child which will do execv(), then parent will pass
control to process_pty() to do its magic.
This was exposed in part by:
https://bugs.freedesktop.org/show_bug.cgi?id=76193
Reported-by: Tobias Hunger tobias.hunger@gmail.com
On Thu, May 22, 2014 at 9:53 AM, Jan Engelhardt <jengelh@inai.de> wrote:
>
> If libsystemd-network.la is relying on that udev function, it ought
> to specify libudev(-internal).la in libsystemd_network_la_LIBADD.
The verbose link-time deprecation warnings are annoying. These libs
will never change or be extended; there is no need to test the list
of exported symbols.
./.libs/libsystemd-network.a(libsystemd_network_la-network-internal.o):
network-internal.c:function net_get_unique_predictable_data:
error: undefined reference to 'udev_device_get_property_value'
collect2: error: ld returned 1 exit status
We shouldn't destroy IPC objects of system users on logout.
http://lists.freedesktop.org/archives/systemd-devel/2014-April/018373.html
This introduces SYSTEM_UID_MAX defined to the maximum UID of system
users. This value is determined compile-time, either as configure switch
or from /etc/login.defs. (We don't read that file at runtime, since this
is really a choice for a system builder, not the end user.)
While we are at it we then also update journald to use SYSTEM_UID_MAX
when we decide whether to split out log data for a specific client.
This makes callback behaviour more like sd-event or sd-resolve, and
creates proper object for unregistering callbacks.
Taking the refernce to the slot is optional. If not taken life time of
the slot will be bound to the underlying bus object (or in the case of
an async call until the reply has been recieved).
New "struct ring" object that implements a basic ring buffer for arbitrary
byte-streams. A new basic runtime test is also added.
This will be needed for our pty helpers for systemd-console and friends.
As the operational state detection in sd-network is still too primitive, timesyncd
will likely try to connect a bit early, so the first attempt will fail.
- Also only allow positive ifindex on both dhcp and ipv4ll
[tomegun: the kernel always sets a positive ifindex, but some APIs accept
ifindex=0 with various meanings, so we should protect against
accidentally passing ifindex=0 along.]
To make sure we don't delay boot on systems where (some) network links are managed by someone else
we don't block if something else has successfully brought up a link.
We will still block until all links we are aware of that are managed by networkd have been
configured, but if no such links exist, and someone else have configured a link sufficiently
that it has a carrier, it may be that the link is ready so we should no longer block.
Note that in all likelyhood the link is not ready (no addresses/routes configured),
so whatever network managment daemon configured it should provide a similar wait-online
service to block network-online.target until it is ready.
The aim is to block as long as we know networking is not fully configured, but no longer. This
will allow systemd-networkd-wait-online.service to be enabled on any system, even if we don't
know whether networkd is the main/only network manager.
Even in the case networking is fully configured by networkd, the default behavior may not be
sufficient: if two links need to be configured, but the first is fully configured before the
second one appears we will assume the network is up. To work around that, we allow specifying
specific devices to wait for before considering the network up.
This unit is enabled by default, just like systemd-networkd, but will only be pulled in if
anyone pulls in network-online.target.
The way the kernel namespaces have been implemented breaks assumptions
udev made regarding uevent sequence numbers. Creating devices in a
namespace "steals" uevents and its sequence numbers from the host. It
confuses the "udevadmin settle" logic, which might block until util a
timeout is reached, even when no uevent is pending.
Remove any assumptions about sequence numbers and deprecate libudev's
API exposing these numbers; none of that can reliably be used anymore
when namespaces are involved.
Use a static table with all the typing information, rather than repeated
switch statements. This should make it a lot simpler to add new types.
We need to keep all the type info to be able to create containers
without exposing their implementation details to the users of the library.
As a freebee we verify the types of appended/read attributes.
The API is extended to nicely deal with unions of container types.
<kay> ssuominen: and drop --no-as-needed from the linkcheck?
<kay> ssuominen: i expect it all triggers without the gc-sections thing alone
<ssuominen> if the intention is to make it strict as possible,
to catch undefined references caused by missing -lfoo in linker line, then
LDFLAGS="-Wl,-fuse-ld=gold -Wl,--as-needed -Wl,--no-gc-sections"
Added support for tunneling netlink attrributes (ipip, gre, sit).
These works with kernel module ipip, gre and sit . The test cases are
moved to a separate file and manual test as well because they require
respective kernel modules as well.
tcpwrap is legacy code, that is barely maintained upstream. It's APIs
are awful, and the feature set it exposes (such as DNS and IDENT
access control) questionnable. We should not support this natively in
systemd.
Hence, let's remove the code. If people want to continue making use of
this, they can do so by plugging in "tcpd" for the processes they start.
With that scheme things are as well or badly supported as they were from
traditional inetd, hence no functionality is really lost.
This does not belong in shared as it is mostly a detail of our networking subsystem.
Moreover, now we can use libudev here, which will simplify things.
Add a new config 'Activating' directive which denotes whether a busname
is actually registered on the bus. It defaults to 'yes'.
If set to 'no', the .busname unit only uploads policy, which will remain
active as long as the unit is running.
After all, it is ultimately linked to libsystems.so anyway, thus belongs
there and shares very little with the rest of logind, hence let's move
this away.
The whole tool is made dependent on µhttpd availability. It should be
easy to make the µhttpd parts conditional, but since transfer over
HTTP seems to be the primary use case, currently this is not done.
Current implementation uses nested epoll loops: sd-event is used for
the external event loop, and µhttpd uses epoll in its own
loop. Unfortunately µhttpd does not expose enough information to add
the descriptors it uses to the external event loop. This means that
starvation of other events is possible, if one of the inner µhttpd
loops is constantly busy. This means that µhttpd servers should not
be mixed with other sources.
The TLS authentication parts haven't been really tested properly, and
should not be take too seriously.
Prefix "gnutls: " is added. Some semi-random mapping of gnutls levels
to syslog levels is done, but since gnutls levels seem to be used
rather loosely, most end up as debug.
This will let journald forward logs as messages sent to all logged in
users (like wall).
Two options are added:
* ForwardToWall (default yes)
* MaxLevelWall (default emerg)
'ForwardToWall' is overridable by kernel command line option
'systemd.journald.forward_to_wall'.
This is used to emulate the traditional syslogd behaviour of sending
emergency messages to all logged in users.
