This makes most header files easier to look at. Also Emacs gets really
slow when browsing through large sections of overly long prototypes,
which is much improved by this macro.
We should probably not do something similar with too many other cases,
as macros like this might help readability for some, but make it worse
for others. But I think given the complexity of this specific prototype
and how often we use it, it's worth doing.
This builds on the previous GENERIC_PARSER_ARGS macro work. I think in
general it is a better idea to declare macros that generate full C
statements instead of just parts of them, hence, let's introduce
CONFIG_PARSER_PROTOTYPE() which defines a full C function prototype,
instead of the pre-existing way of defining the C function prototype
manually, but then using GENERIC_PARSER_ARGS to define its arguments.
This doesn't drop GENERIC_PARSER_ARGS though, but renames it to
CONFIG_PARSER_ARGUMENTS, and changes the ConfigParserCallback function
type to use it. The new name follows more closely how the other symbols
in the header are named.
That way we can use it in nspawn.
Also, while we are at it, let's rename the call config_parse_rlimit(),
i.e. insert the "r", to clarify what kind of limit this is about.
This makes it behave the same whether there is a blank line or not at
the end of the file. This is also consistent with the behavior of the
shell on a shell script that ends on a trailing backslash at the last
line.
Added tests to test_config_parse(), which only pass if the corresponding
change to config_parse() is included.
This means that when those targets are built, all the sources are built again,
instead of reusing the work done to create libbasic.a and other convenience static
libraries. It would be nice to not do this, but there seems to be no support in
our toolchain for joining multiple static libraries into one. When linking
a static library, any -l arguments are simply ignored by ar/gcc-ar, and .a
libraries given as positional arguments are copied verbatim into the archive
so they objects in them cannot be accessed.
https://stackoverflow.com/questions/2157629/linking-static-libraries-to-other-static-libraries
suggests either unzipping all the archives and putting them back togather,
or using a linker script. Unzipping and zipping back together seems ugly.
The other option is not very nice. The linker script language does not
allow "+" to appear in the filenames, and filenames that meson generates
use that, so files would have to be renamed before a linker script was used.
And we would have to generate the linker script on the fly. Either way, this
doesn't seem attractive. Since those static libraries are a niche use case,
it seems reasonable to just go with the easiest and safest solution and
recompile all the source files. Thanks to ccache, this is probably almost as
cheap as actually reusing the convenience .a libraries.
test-libsystemd-sym.c and test-libudev-sym.c compile fine with the generated
static libs, so it seems that they indeed provide all the symbols they should.
Scope units are populated from PIDs specified by the bus client. We do
that when a scope is started. We really shouldn't allow scopes to be
started multiple times, as the PIDs then might be heavily out of date.
Moreover, clients should have the guarantee that any scope they allocate
has a clear runtime cycle which is not repetitive.
Previously we were a bit sloppy with the index and size types of arrays,
we'd regularly use unsigned. While I don't think this ever resulted in
real issues I think we should be more careful there and follow a
stricter regime: unless there's a strong reason not to use size_t for
array sizes and indexes, size_t it should be. Any allocations we do
ultimately will use size_t anyway, and converting forth and back between
unsigned and size_t will always be a source of problems.
Note that on 32bit machines "unsigned" and "size_t" are equivalent, and
on 64bit machines our arrays shouldn't grow that large anyway, and if
they do we have a problem, however that kind of overly large allocation
we have protections for usually, but for overflows we do not have that
so much, hence let's add it.
So yeah, it's a story of the current code being already "good enough",
but I think some extra type hygiene is better.
This patch tries to be comprehensive, but it probably isn't and I missed
a few cases. But I guess we can cover that later as we notice it. Among
smaller fixes, this changes:
1. strv_length()' return type becomes size_t
2. the unit file changes array size becomes size_t
3. DNS answer and query array sizes become size_t
Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=76745
It's mostly a wrapper around parse_mtu() but with some nicer logging.
The address family is initialized from the "ltype" parameter, so that
configuration file parser tables can be easily declare it.
This drops a good number of type-specific _cleanup_ macros, and patches
all users to just use the generic ones.
In most recent code we abstained from defining type-specific macros, and
this basically removes all those added already, with the exception of
the really low-level ones.
Having explicit macros for this is not too useful, as the expression
without the extra macro is generally just 2ch wider. We should generally
emphesize generic code, unless there are really good reasons for
specific code, hence let's follow this in this case too.
Note that _cleanup_free_ and similar really low-level, libc'ish, Linux
API'ish macros continue to be defined, only the really high-level OO
ones are dropped. From now on this should really be the rule: for really
low-level stuff, such as memory allocation, fd handling and so one, go
ahead and define explicit per-type macros, but for high-level, specific
program code, just use the generic _cleanup_() macro directly, in order
to keep things simple and as readable as possible for the uninitiated.
Note that before this patch some of the APIs (notable libudev ones) were
already used with the high-level macros at some places and with the
generic _cleanup_ macro at others. With this patch we hence unify on the
latter.
Those are quite similar to %i/%I, but refer to the last dash-separated
component of the name prefix.
The new functionality of dash-dropins could largely supersede the template
functionality, so it would be tempting to overload %i/%I. But that would
not be backwards compatible. So let's add the two new letters instead.
Double newlines (i.e. one empty lines) are great to structure code. But
let's avoid triple newlines (i.e. two empty lines), quadruple newlines,
quintuple newlines, …, that's just spurious whitespace.
It's an easy way to drop 121 lines of code, and keeps the coding style
of our sources a bit tigther.
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
pager.[ch] doesn't use any APIs from src/libsystemd/ or src/shared/
hence there's no reason for it to be in src/shared/, let's move it to
src/basic/ instead.
This enables us to use pager.[ch] APIs from other code in src/basic/,
for example pager_have() and suchlike.
This extends the logic by which we look for drop-ins for unit files when
loading them. Previously for a unit "foo-quux-bar.service" we'd look in
a directory "foo-quux-bar.service.d" accompanying it for extension
dropins. With this change we'll additionally look in:
"foo-quux-.service.d" and "foo-.service.d", i.e. we'll truncate the unit
name after every dash.
This is an alternative to templating for many services, as it permits
configuring defaults for sets of units that all use the same prefix in
the unit name. This is particularly useful in slice, mount and
automount units which reflect a hierarchy of concepts, as it permits
setting defaults for specific subsets of the tree. For example, in order
to provide every user with a memory of 1G it's now possible to do:
# mkdir -p /etc/systemd/system/user-.slice.d
# cat > /etc/systemd/system/user-.slice.d/50-memory.conf << EOF
[Slice]
MemoryMax=1G
EOF
# systemctl daemon-reload
This makes use of the fact that every user gets his own slice unit when
logging in, named "user-$UID.slice".
This doesn't precisely provide what is requested in #2556, but it does
provide equivalent functionality.
Fixes: #2556
See: #3504#7599
$ sudo swapoff -av
swapoff /dev/vda4
$ sudo systemctl hibernate
Failed to hibernate system via logind: Not enough swap space for hibernation
Fixes#6729.
can_sleep() returns 0 if the operation is impossible, but
the code assumed that negative is returned in that case,
in effect reporting s2h was possible even if hibernation or
suspend were not possible.
The Linux kernel is adding support for configuring the offset
into a disk. This allows swapfiles to be more usable as users
will no longer need to set the offset on their kernel command
line.
Use this API in systemd when hibernating as well.
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
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.
When we try to read meta-data from an image, don't bother with mounting
/home or the ESP, as that's not where the metadata is. This not only
speeds things up a bit, but also has the benefit that setups where an
unencrypted root is mixed with an encrypted /home (which I have on one
of my own systems) won't result in errors that the crypto key is needed.
This extends on #8609, and makes two changes:
1. We'll now explicitly check that the child devices of a block device
we are interested in (i.e. the partitions) are block devices themselves.
On newer kernels the mmc rpmb stuff is actually exposed as char rather
than block device as before, and they probably should have been that in
the first place. By adding this check we'll hence filter out these weird
devices through a second rule too, that hopefully makes things a bit
more future-proof, should more devices like this be added eventually,
or other subsystems do a similar thing.
2. When counting partitions we'll now also check the devnum of the
device being non-null, which we already do when matching up the devices
in the second iteration. This should make things more robust, and
prevent other kinds of miscounting, which after all was the main
issue #8609 fixed.
Filter-out RPMB partitions and boot partitions from MMC devices when
counting partitions enumerated by the kernel. Also factor out the now
duplicated code into a separate function.
This complement the previous fixes to the problem reported in
https://github.com/systemd/systemd/issues/5806
The names of drop-in files can be anything as long as they are suffixed
in ".conf", hence don't be stricter than necessary when validating the
names used in symlink chains of such drop-in files.
Also, drop-in files should not be ale to change the type of unit file
itself, i.e. not affect whether it is considered masked or an alias as a
whole.
This adds a flag SEARCH_DROPIN that is passed whenever we load a drop-in
rather the main unit file, and in that case loosen checks and behaviour
we otherwise enforce for the unit file itself. Specifically:
1. If SEARCH_DROPIN is passed we won't change the unit's info->type
field anymore, as that field (which can be REGULAR, MASKED, SYMLINK)
should not be affected by drop-ins, but only by the unit file itself.
2. If SEARCH_DROPIN is passed we will shortcut following of symlink
chains, and not validate the naming of each element in the chain,
since that's irrelevant for drop-ins, and only matters for the unit
file itself.
Or in other words, without this:
1. A symlink /etc/systemd/system/foobar.service.d/20-quux.conf →
/dev/null might have caused the whole of foobar.service to be
considered "masked".
2. A symlink /etc/systemd/system/foobar.service.d/20-quux.conf →
/tmp/miepf might have caused the whole loading of foobar.service to
fail as EINVAL, as "miepf" is not a valid unit name.
This adds flags BUS_MAP_STRDUP and BUS_MAP_BOOLEAN_AS_BOOL.
If BUS_MAP_STRDUP is set, then each "s" message is duplicated.
If BUS_MAP_BOOLEAN_AS_BOOL is set, then each "b" message is
written to a bool pointer.
Follow-up for #8488.
See https://github.com/systemd/systemd/pull/8488#discussion_r175816270.
This is a separate commit only because it actually *increases* memory allocations:
==3256== total heap usage: 100,120 allocs, 100,120 frees, 13,097,140 bytes allocated
to
==4690== total heap usage: 100,121 allocs, 100,121 frees, 14,198,329 bytes allocated
Essentially, we do a little more work to reduce the memory footprint a bit. For a
test where we just allocate the memory and drop it soon afterwards, this is not
beneficial, but it should still be useful for a long running program.
ubsan times out because we do too many allocations:
$ valgrind build/fuzz-unit-file test/fuzz-regressions/fuzz-unit-file/oss-fuzz-6977-full
...
test/fuzz-regressions/fuzz-unit-file/oss-fuzz-6977-full... ok
==1757==
==1757== HEAP SUMMARY:
==1757== in use at exit: 0 bytes in 0 blocks
==1757== total heap usage: 199,997 allocs, 199,997 frees, 90,045,318,585 bytes allocated
...
==3256== total heap usage: 100,120 allocs, 100,120 frees, 13,097,140 bytes allocated
https://oss-fuzz.com/v2/issue/4651449704251392/6977 should now be really fixed.
e3c3d6761b was the first attempt, but even with this change, e3c3d6761b
still makes sense.
This reverts the mmap parts of f5aeac1439,
but keeps the part which restricts address families which works
correctly.
Unfortunately the MIPS toolchains still do not implement PT_GNU_STACK.
This means that while the commit to restrict mmap on MIPS was "correct",
it had the side effect of causing pthread_create to fail because glibc tries
to allocate an executable stack for new threads in the absense of
PT_GNU_STACK. We should wait until PT_GNU_STACK is implemented in all
the relevant parts of the toolchain (at least gcc and glibc) before
enabling this again.
This is similar to TAKE_PTR() but operates on file descriptors, and thus
assigns -1 to the fd parameter after returning it.
Removes 60 lines from our codebase. Pretty good too I think.
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)
This rearranges chase_symlinks() a bit: if no special flags are
specified it will now revert to behaviour before
b12d25a8d6. However, if the new
CHASE_TRAIL_SLASH flag is specified it will follow the behaviour
introduced by that commit.
I wasn't sure which one to make the beaviour that requires specification
of a flag to enable. I opted to make the "append trailing slash"
behaviour the one to enable by a flag, following the thinking that the
function should primarily be used to generate a normalized path, and I
am pretty sure a path without trailing slash is the more "normalized"
one, as the trailing slash is not really a part of it, but merely a
"decorator" that tells various system calls to generate ENOTDIR if the
path doesn't refer to a path.
Or to say this differently: if the slash was part of normalization then
we really should add it in all cases when the final path is a directory,
not just when the user originally specified it.
Fixes: #8544
Replaces: #8545
Let's use first_word() instead of startswith(), it's more explanatory
and a bit more correct. Also, let's use the return value instead of
adding +9 when looking for the second part of the directive.
.include lines are already deprecated somewhat, and for example
explicitly not mentioned in the documentation for this reason. Let's get
one step further and generatea warning when we encounter them (but still
process them).
Why are they deprecated? Because they are semantically awful — they
complicate stat() based mtime checks for configuration files and they
allow arbitrary loops we currently have zero protection against and
really shouldn't have to have.
In commit da1921a5c3 ppc64/ppc64el were added as supported architectures for
socketcall() for the POWER family. Extend the support for the 32bits
architectures.
Support was killed in kernel 4.15 as well as ethtool 4.13.
Justification was lack of use by drivers and too much of a maintenance burden.
https://www.spinics.net/lists/netdev/msg443815.html
Also moved config_parse_warn_compat to conf-parser.[ch] to fix compile errors.
"noreturn" is reserved and can be used in other header files we include:
[ 16s] In file included from /usr/include/gcrypt.h:30:0,
[ 16s] from ../src/journal/journal-file.h:26,
[ 16s] from ../src/journal/journal-vacuum.c:31:
[ 16s] /usr/include/gpg-error.h:1544:46: error: expected ‘,’ or ‘;’ before ‘)’ token
[ 16s] void gpgrt_log_bug (const char *fmt, ...) GPGRT_ATTR_NR_PRINTF(1,2);
Here we include grcrypt.h (which in turns include gpg-error.h) *after* we
"noreturn" was defined in macro.h.
Suspend to Hibernate is a new sleep method that invokes suspend
for a predefined period of time before automatically waking up
and hibernating the system.
It's similar to HybridSleep however there isn't a performance
impact on every suspend cycle.
It's intended to use with systems that may have a higher power
drain in their supported suspend states to prevent battery and
data loss over an extended suspend cycle.
Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>
ISO C does not allow empty statements outside of functions, and gcc
will warn the trailing semicolons when compiling with -pedantic:
warning: ISO C does not allow extra ‘;’ outside of a function [-Wpedantic]
But our code cannot compile with -pedantic anyway, at least because
warning: ISO C does not support ‘__PRETTY_FUNCTION__’ predefined identifier [-Wpedantic]
Without -pedatnic, clang and even old gcc (3.4) generate no warnings about
those semicolons, so let's just drop __useless_struct_to_allow_trailing_semicolon__.
This reworks system call filter parsing, and replaces a couple of "bool"
function arguments by a single flags parameter.
This shouldn't change behaviour, except for one case: when we
recursively call our parsing function on our own syscall list, then
we'll lower the log level to LOG_DEBUG from LOG_WARNING, because at that
point things are just a problem in our own code rather than in the user
configuration we are parsing, and we shouldn't hence generate confusing
warnings about syntax errors.
Fixes: #8261
There isn't much difference, but in general we prefer to use the standard
functions. glibc provides reallocarray since version 2.26.
I moved explicit_bzero is configure test to the bottom, so that the two stdlib
functions are at the bottom.
When running journalctl --user-unit=foo as an unprivileged user we could get
the usual hint:
Hint: You are currently not seeing messages from the system and other users.
Users in groups 'adm', 'systemd-journal', 'wheel' can see all messages.
...
But with --user-unit our filter is:
(((_UID=0 OR _UID=1000) AND OBJECT_SYSTEMD_USER_UNIT=foo.service) OR
((_UID=0 OR _UID=1000) AND COREDUMP_USER_UNIT=foo.service) OR
(_UID=1000 AND USER_UNIT=foo.service) OR
(_UID=1000 AND _SYSTEMD_USER_UNIT=foo.service))
so we would never see messages from other users.
We could still see messages from the system. In fact, on my machine the
only messages with OBJECT_SYSTEMD_USER_UNIT= are from the system:
journalctl $(journalctl -F OBJECT_SYSTEMD_USER_UNIT|sed 's/.*/OBJECT_SYSTEMD_USER_UNIT=\0/')
Thus, a more correct hint is that we cannot see messages from the system.
Make it so.
Fixes#7887.
Let's add a common implementation for regular file checks, that are
careful to return the right error code (EISDIR/EISLNK/EBADFD) when we
are encountering a wrong file node.
config_parse_join_controllers would free the destination argument on failure,
which is contrary to our normal style, where failed parsing has no effect.
Moving it to shared also allows a test to be added.
The arguments have to be indentical everywhere, so let's use a macro to
make things more readable. But only in the headers, in the .c files let's
keep them verbose so that it's easy to see the argument list.
In the parse logic `line_get_key_value()` in sd-boot treats spaces
and tabs are valid spacing between key and value in the line.
So, let's use the same logic for `bootctl` and the others which read
sd-boot configs.
Fixes#8154.
A couple of fixes:
1. always bzero_explicit() away what we remove from the passphrase
buffer. The UTF-8 code assumes the string remains NUL-terminated, and
we hence should enforce that. memzero() would do too here, but let's
be paranoid after all this is key material.
2. when clearing '*' characters from string, do so counting UTF-8
codepoints properly. We already have code in place to count UTF-8
codepoints when generating '*' characters, hence we should take the
same care when clearing them again.
3. Treat NUL on input as an alternative terminator to newline or EOF.
4. When removing characters from the password always also reset the
"codepoint" index properly.
We already have the terminal open, hence pass the fd we got to
ask_password_tty(), so that it doesn't have to reopen it a second time.
This is mostly an optimization, but it has the nice benefit of making us
independent from RLIMIT_NOFILE issues and so on, as we don't need to
allocate another fd needlessly.
We should rather sleep to much than too little. This otherwise might
result in a busy loop, because we slept too little and then recheck
again coming to the conclusion we need to go to sleep again, and so on.
$ diff -u <(old/systemd-analyze --user unit-paths) <(new/systemd-analyze --user unit-paths)|colordiff
--- /proc/self/fd/14 2018-02-08 14:36:34.190046129 +0100
+++ /proc/self/fd/15 2018-02-08 14:36:34.190046129 +0100
@@ -1,5 +1,5 @@
-/home/zbyszek/.config/systemd/system.control
-/run/user/1000/systemd/system.control
+/home/zbyszek/.config/systemd/user.control
+/run/user/1000/systemd/user.control
/run/user/1000/systemd/transient
...
Strictly speaking, online upgrades of user instances through daemon-reexec will
be broken. We can get away with this since
a) reexecs of the user instance are not commonly done, at least package upgrade
scripts don't do this afawk.
b) cgroups aren't delegateable on cgroupsv1 there's little reason to use "systemctl
set-property" for --user mode
It's not good if the paths are in different order. With --user, we expect
more paths, but it must be a strict superset, and the order for the ones
that appear in both sets must be the same.
$ diff -u <(build/systemd-analyze --global unit-paths) <(build/systemd-analyze --user unit-paths)|colordiff
--- /proc/self/fd/14 2018-02-08 14:11:45.425353107 +0100
+++ /proc/self/fd/15 2018-02-08 14:11:45.426353116 +0100
@@ -1,6 +1,17 @@
+/home/zbyszek/.config/systemd/system.control
+/run/user/1000/systemd/system.control
+/run/user/1000/systemd/transient
+/run/user/1000/systemd/generator.early
+/home/zbyszek/.config/systemd/user
/etc/systemd/user
+/run/user/1000/systemd/user
/run/systemd/user
+/run/user/1000/systemd/generator
+/home/zbyszek/.local/share/systemd/user
+/home/zbyszek/.local/share/flatpak/exports/share/systemd/user
+/var/lib/flatpak/exports/share/systemd/user
/usr/local/share/systemd/user
/usr/share/systemd/user
/usr/local/lib/systemd/user
/usr/lib/systemd/user
+/run/user/1000/systemd/generator.late
A test is added so that we don't regress on this.
This doesn't matter that much, because set-property --global does not work,
so at least those paths wouldn't be used automatically. It is still possible
to create such snippets manually, so we better fix this.
The VDSO provided by the kernel for x32, uses x86-64 syscalls instead of
x32 ones.
I think we can safely allow this; the set of x86-64 syscalls should be
very similar to the x32 ones. The real point is not to allow *x86*
syscalls, because some of those are inconveniently multiplexed and we're
apparently not able to block the specific actions we want to.
Booting with `systemd.log_level=debug` and looking in `dmesg -u` showed
messages like this:
systemd[433]: Failed to add rule for system call n/a() / 156, ignoring:
Numerical argument out of domain
This commit fixes it to:
systemd[449]: Failed to add rule for system call _sysctl() / 156,
ignoring: Numerical argument out of domain
Some of the messages could be even more misleading, e.g. we were reporting
that utimensat() / 320 was skipped as non-existent on x86, when actually
the syscall number 320 is kexec_file_load() on x86 .
The problem was that syscall NRs are looked up (and correctly passed to
libseccomp) as native syscall NRs. But we forgot that when we tried to
go back from the syscall NR to the name.
I think the natural way to write this would be
seccomp_syscall_resolve_num(nr), however there is no such function.
I couldn't work out a short comment that would make this clearer. FWIW
I wrote it up as a ticket for libseccomp instead.
https://github.com/seccomp/libseccomp/issues/104
Hiding automatic entries allows for giving custom entry names like
"Windows 10" instead of "Windows Boot Manager" by creating an appropriate
loader file in the loader/entries folder.
Note that it is already doable by renaming bootmgfw.efi (or the other auto-detected
boot loaders) and then using the renamed file for a custom entry. But windows will
automatically recreate the boot loader on updates, including the default EFI
bootloader entry if that one is missing.
Make hiding EFI reboot a separate option because there is no simple way to create
it with a custom loader entry and people may still want that around while still
hiding the other auto entries.
Also, turn no_editor into a positive boolean name while we're touching this code.
Red is used for highligting, the same as grep does. Except when the line is
highlighted red already, because it has high priority, in which case plain ansi
highlight is used for the matched substring.
Coloring is implemented for short and cat outputs, and not for other types.
I guess we could also add it for verbose output in the future.
Before this, `systemctl show` for calendar type timer unit outputs
something like below.
```
NextElapseUSecRealtime=48y 3w 3d 15h
NextElapseUSecMonotonic=0
LastTriggerUSec=48y 3w 3d 3h 41min 44.093095s
LastTriggerUSecMonotonic=0
```
As both NextElapseUSecRealtime= and LastTriggerUSec= are not timespan
but timestamp, this makes format these values by `format_timestamp()`.
/dev/mmcblk0boot0 is a partition found in eMMC
This is not relevant for mounting
This complement the previous fix as reported in
https://github.com/systemd/systemd/issues/5806
Signed-off-by: Nicolas Chauvet <kwizart@gmail.com>
Apparently O_NONBLOCK is the modern name used in most documentation and
for most cases in our sources. Let's hence replace the old alias
O_NDELAY and stick to O_NONBLOCK everywhere.
Incorporating the fix from d00f1d57 into other output formats of journalctl.
If journal files are corrupted, e.g. not cleanly closed, some journal
entries can not be read by output options other than 'short' (default).
If such entries has been identified, they will now just be skipped.
log.h really should only include the bare minimum of other headers, as
it is really pulled into pretty much everything else and already in
itself one of the most basic pieces of code we have.
Let's hence drop inclusion of:
1. sd-id128.h because it's entirely unneeded in current log.h
2. errno.h, dito.
3. sys/signalfd.h which we can replace by a simple struct forward
declaration
4. process-util.h which was needed for getpid_cached() which we now hide
in a funciton log_emergency_level() instead, which nicely abstracts
the details away.
5. sys/socket.h which was needed for struct iovec, but a simple struct
forward declaration suffices for that too.
Ultimately this actually makes our source tree larger (since users of
the functionality above must now include it themselves, log.h won't do
that for them), but I think it helps to untangle our web of includes a
tiny bit.
(Background: I'd like to isolate the generic bits of src/basic/ enough
so that we can do a git submodule import into casync for it)
The changes both networkd and resolved to make use of the watch_bind
feature of sd-bus to connect to the system bus. This way, both daemons
can be started during early boot, and automatically and instantly
connect to the system bus as it becomes available.
This replaces prior code that used a time-based retry logic to connect
to the bus.
Let's remove a number of synchronization points from our service
startups: let's drop synchronous match installation, and let's opt for
asynchronous instead.
Also, let's use sd_bus_match_signal() instead of sd_bus_add_match()
where we can.
This is useful so that callers know whether anything at all and how much
was flushed.
This patches through users of this functions to ensure that the return
values > 0 which may be returned now are not propagated in public APIs.
Also, users that ignore the return value are changed to do so explicitly
now.
We maintain static process-wide variables in these subsystems without
locking, hence let's refuse operation unless we are called from the main
thread (which we do anyway) just as a safety precaution.
Using wait_for_terminate_and_check() instead of wait_for_terminate()
let's us simplify, shorten and unify the return value checking and
logging of waitid(). Hence, let's use it all over the place.
Instead of compiling those files twice, once for libsystemd and once for
libshared, compile once as a static archive and then link into both.
This reduce the meson target for man=no compile to 1291.
We were including gcrypt-util.[ch] by hand in the few places where it
was used. Create a convenience library to avoid compiling the same
files multiple times.
v2:
- use a separate static library instead of mergin into libbasic
gcrypt_util_sources had to be moved because otherwise they appeared twice
in libshared.so halfproducts, causing an error.
-fvisibility=default is added to libbasic, libshared_static so that the symbols
appear properly in the exported symbol list in libshared.
The advantage is that files are not compiled twice. When configured with -Dman=false,
the ninja target list is reduced from 1588 to 1347 targets. The difference in compilation
time is small (<10%). I think this is because of -O0 and ccache and multiple cores, and
in different settings the compilation time could be reduced. The main advantage is that
errors and warnings are not reported twice.
This adds a simple condition/assert/match to the service manager, to
udev's .link handling and to networkd, for matching the kernel version
string.
In this version we only do fnmatch() based globbing, but we might want
to extend that to version comparisons later on, if we like, by slightly
extending the syntax with ">=", "<=", ">", "<" and "==" expressions.
The boot loader systemd-boot removes ".conf" from file name of entry
configs, and determine which entry is the default entry.
However, bootspec, which is used by systemctl and bootctl did not
remove ".conf", then sometimes bootctl marks wrong entry as default.
This fixes the logic to choose the default entry in bootspec, to
match the logic used in systemd-boot boot loader.
Fixes#7727.
It's a relatively small wrapper around safe_fork() now, hence let's move
it over, and make its signature even more alike. Also, set a different
process name for the polkit and askpw agents.
This adds a new safe_fork() wrapper around fork() and makes use of it
everywhere. The new wrapper does a couple of things we previously did
manually and separately in a safer, more correct and automatic way:
1. Optionally resets signal handlers/mask in the child
2. Sets a name on all processes we fork off right after forking off (and
the patch assigns useful names for all processes we fork off now,
following a systematic naming scheme: always enclosed in () – in order
to indicate that these are not proper, exec()ed processes, but only
forked off children, and if the process is long-running with only our
own code, without execve()'ing something else, it gets am "sd-" prefix.)
3. Optionally closes all file descriptors in the child
4. Optionally sets a PR_SET_DEATHSIG to SIGTERM in the child, in a safe
way so that the parent dying before this happens being handled
safely.
5. Optionally reopens the logs
6. Optionally connects stdin/stdout/stderr to /dev/null
7. Debug logs about the forked off processes.
If multiple SystemCallFilter= settings, some of them are whitelist
and the others are blacklist, are sent to bus, then the parse
result was corrupted.
This fixes the parse logic, now it is the same as one used in
load-fragment.c
Also remove the warning:
./src/shared/seccomp-util.c:1414:2: warning: #warning "Consider adding the right mmap() syscall definitions here!" [-Wcpp]
#warning "Consider adding the right mmap() syscall definitions here!"
We already use the "_static" suffix for libshared_static ("shared" is the name
of the library, "static" is the format) and other libs, so let's rename for
consistency.
Also change libsystemd_static_sources to libsystemd_sources, since the same
list is used for both and shorter is better.
Up until now, the behaviour in systemd has (mostly) been to silently
ignore failures to action unit directives that refer to an unavailble
controller. The addition of AssertControlGroupController and its
conditional counterpart allow explicit specification of the desired
behaviour when such a situation occurs.
As for how this can happen, it is possible that a particular controller
is not available in the cgroup hierarchy. One possible reason for this
is that, in the running kernel, the controller simply doesn't exist --
for example, the CPU controller in cgroup v2 has only recently been
merged and was out of tree until then. Another possibility is that the
controller exists, but has been forcibly disabled by `cgroup_disable=`
on the kernel command line.
In future this will also support whatever comes out of issue #7624,
`DefaultXAccounting=never`, or similar.
This makes things a bit easier to read I think, and also makes sure we
always use the _unlikely_ wrapper around it, which so far we used
sometimes and other times we didn't. Let's clean that up.
This changes dns_name_between() to deal properly with checking whether B
is between A and C if A and C are equal. Previously we simply returned
-EINVAL in this case, refusing checking. With this change we correct
behaviour: if A and C are equal, then B is "between" both if it is
different from them. That's logical, since we do < and > comparisons, not
<= and >=, and that means that anything "right of A" and "left of C"
lies in between with wrap-around at the ends. And if A and C are equal
that means everything lies between, except for A itself.
This fixes handling of domains using NSEC3 "white lies", for example the
.it TLD.
Fixes: #7421
fputs() writes only first 2048 bytes and fails
to write to /proc when values are larger than that.
This patch adds a new flag to WriteStringFileFlags
that make it possible to disable the buffer under
specific cases.
This renames find_esp() to find_esp_and_warn() and tries to normalize its
behaviour:
1. Change the error that is returned when we can't find the ESP to
ENOKEY (from ENOENT). This way the error code can only mean one
thing: that our search loop didn't find a good candidate.
2. Really log about all errors, except for ENOKEY and EACCES, and
document the letter cases.
3. Normalize parameters to the call: separate out the path parameter in
two: an input path and an output path. That way the memory management
is clear: we will access the input parameter only for reading, and
only write out the output parameter, using malloc() memory.
Before the calling convention were quire surprising for internal API
code, as the path parameter had to be malloc() memory and might and
might not have changed.
4. Rename bootctl's find_esp_warn() to acquire_esp(), and make it a
simple wrapper around find_esp_warn(), that basically just adds the
friendly logging for the ENOKEY case. This rework removes double
logging in a number of error cases, as we no longer log here in
anything but ENOKEY, and leave that entirely to find_esp_warn().
5. find_esp_and_warn() now takes a bool flag parameter
"unprivileged_mode", which disables logging in the EACCES case, and
skips privileged validation of the path. This makes the function less
magic, and doesn't hide this internal silencing automatism from the
caller anymore.
With all that in place "bootctl list" and "bootctl status" work properly
(or as good as they can) when I invoke the tools whithout privileges on
my system where /boot is not world-readable
Currently systemd hardcodes the use of /dev/watchdog. This is a legacy
chardev that points to watchdog0 in the system.
Modify the watchdog API to allow a different device path to be passed
and stored. Opening the watchdog defaults to /dev/watchdog, maintaining
existing behavior.
When no entries matches with entry_oneshot, entry_default and
default_pattern, then log message shows a wrong entry.
Moreover, if none of entry_oneshot, entry_default and default_pattern
are set, then the index `i` is uninitialized.
This fixes such problem.
```
$ ./src/test/test-systemd-tmpfiles.py valgrind --leak-check=full --error-exitcode=1 ./build/systemd-tmpfiles
...
Running valgrind --leak-check=full --error-exitcode=1 ./build/systemd-tmpfiles on 'w /unresolved/argument - - - - "%Y"'
...
[<stdin>:1] Failed to substitute specifiers in argument: Invalid slot
...
==22602== 5 bytes in 1 blocks are definitely lost in loss record 1 of 2
==22602== at 0x4C2DB8F: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==22602== by 0x4ECA7D4: malloc_multiply (alloc-util.h:74)
==22602== by 0x4ECA909: specifier_printf (specifier.c:59)
==22602== by 0x113490: specifier_expansion_from_arg (tmpfiles.c:1923)
==22602== by 0x1144E7: parse_line (tmpfiles.c:2159)
==22602== by 0x11551C: read_config_file (tmpfiles.c:2425)
==22602== by 0x115AB0: main (tmpfiles.c:2529)
```
This adds uid_is_system() and gid_is_system(), similar in style to
uid_is_dynamic(). That a helper like this is useful is illustrated by
the fact that test-condition.c didn't get the check right so far, which
this patch fixes.
The code intentionally ignored unknown specifiers, treating them as text. This
needs to change because otherwise we can never add a new specifier in a backwards
compatible way. So just treat an unknown (potential) specifier as an error.
In principle this is a break of backwards compatibility, but the previous
behaviour was pretty much useless, since the expanded value could change every
time we add new specifiers, which we do all the time.
As a compromise for backwards compatibility, only fail on alphanumerical
characters. This should cover the most cases where an unescaped percent
character is used, like size=5% and such, which behave the same as before with
this patch. OTOH, this means that we will not be able to use non-alphanumerical
specifiers without breaking backwards compatibility again. I think that's an
acceptable compromise.
v2:
- add NEWS entry
v3:
- only fail on alphanumerical
Before this CPUAffinity= requires a valid cpu set, and the setting
cannot be reset. Moreover, if CPUAffinity= with empty string is passed,
then message container is closed without no values appended, thus
we get error.
This makes CPUAffinity= accepts empty string to reset the setting
and avoid error.
Apparently there's no guarantee that EPOLLIN is immediately propagated
from a pty slave to the master when data is written to it, hence it's
not sufficient to check EPOLLIN to decide whether the pty device is
drained.
Let's fix this by asking the kernel directly through SIOCINQ + SIOCOUTQ,
if there's anything buffered left.
Fixes: #7531
In -U mode we might need to re-chown() all files and directories to
match the UID shift we want for the image. That's problematic on fat
partitions, such as the ESP (and which is generated by mkosi's
--bootable switch), because fat of course knows no UID/GID file
ownership natively.
With this change we take benefit of the uid= and gid= mount options FAT
knows: instead of chown()ing all files and directories we can just
specify the right UID/GID to use at mount time.
This beefs up the image dissection logic in two ways:
1. First of all support for mounting relevant file systems with
uid=/gid= is added: when a UID is specified during mount it is used for
all applicable file systems.
2. Secondly, two new mount flags are added:
DISSECT_IMAGE_MOUNT_ROOT_ONLY and DISSECT_IMAGE_MOUNT_NON_ROOT_ONLY.
If one is specified the mount routine will either only mount the root
partition of an image, or all partitions except the root partition.
This is used by nspawn: first the root partition is mounted, so that
we can determine the UID shift in use so far, based on ownership of
the image's root directory. Then, we mount the remaining partitions
in a second go, this time with the right UID/GID information.
Let's optimize things a bit, and instead of having to strip whitespace
first before decoding base64, let's do that implicitly while doing so.
Given that base64 was designed the way it was designed specifically to
be tolerant to whitespace changes, it's a good idea to do this
automatically and implicitly.
This usually happens for device units with long
path in /sys. But users can't even create such drop-ins,
so lets just ignore the error here.
Fixes#6867
We let the caller make the decision. Existing callers are OK with treating an
ambiguous result the same as no content, but makefs and growfs should refuse such
partitions.
I opted to completely generate a unit for both mount points and swaps. For
swaps, it would be possible to use fixed template unit like systemd-mkswap@.service,
because there's no information passed except the device name. For mount points,
that's not possible because both the device name and file system type need to
be passed. Nevertheless, I expect that options will need to passed to both mkfs
and mkswap, in which case it'll be necessary to create units of both types
anyway.
Also do not include libcryptsetup.h directly, but only through crypt-util.h.
This way we do not have to repeat the define in every file where it is used.
