In one of the next commits we'd like to introduce a concept of
optionally hashing the hostname from the machine ID. For that we we need
to optionally back gethostname_full() by code involving sd-id128, hence
let's move it from src/basic/ to src/shared/, since only there we are
allowed to use our public APIs.
Chasing symlinks is a core function that's used in a lot of places
so it deservers a less verbose names so let's rename it to chase()
and chaseat().
We also slightly change the pattern used for the chaseat() helpers
so we get chase_and_openat() and similar.
In various tools and services we have a per-system and per-user concept.
So far we sometimes used a boolean indicating whether we are in system
mode, or a reversed boolean indicating whether we are in user mode, or
the LookupScope enum used by the lookup path logic.
Let's address that, in introduce a common enum for this, we can use all
across the board.
This is mostly just search/replace, no actual code changes.
-1 was used everywhere, but -EBADF or -EBADFD started being used in various
places. Let's make things consistent in the new style.
Note that there are two candidates:
EBADF 9 Bad file descriptor
EBADFD 77 File descriptor in bad state
Since we're initializating the fd, we're just assigning a value that means
"no fd yet", so it's just a bad file descriptor, and the first errno fits
better. If instead we had a valid file descriptor that became invalid because
of some operation or state change, the other errno would fit better.
In some places, initialization is dropped if unnecessary.
This doesn't change anything for real uses, because we'd initialize the
variable to NULL for _cleanup_ anyway, but let's follow our general pattern
of always setting the output on "success". (Even if that success is an empty
answer here.)
We would resolve those specifiers to the calling user/group. This is mostly OK
when done in the manager, because the manager generally operates as root
in system mode, and a non-root in user mode. It would still be wrong if
called with --test though. But in systemctl, this would be generally wrong,
since we can call 'systemctl --system' as a normal user, either for testing
or even for actual operation with '--root=…'.
When operating in --global mode, %u/%U/%g/%G should return an error.
The information whether we're operating in system mode, user mode, or global
mode is passed as the data pointer to specifier_group_name(), specifier_user_name(),
specifier_group_id(), specifier_user_id(). We can't use userdata, because
it's already used for other things.
ENOENT is easily confused with the file that we're working on not being
present, e.g. when the file contains %o or something else that requires
os-release to be present. Let's use -EUNATCH instead to reduce that chances of
confusion if the context of the error is lost.
And once we have pinpointed the reason, let's provide a proper error message:
+ build/systemctl --root=/tmp/systemctl-test.TO7Mcb enable some-some-link6@.socket
/tmp/systemctl-test.TO7Mcb/etc/systemd/system/some-some-link6@.socket: Failed to resolve alias "target@A:%A.socket": Protocol driver not attached
Failed to enable unit, cannot resolve specifiers in "target@A:%A.socket".
In systemd.unit we document that unset fields resolve to "". But we didn't
directly test this, so let's do that. Also, we return -ENOENT if the file
is missing, which we didn't document or test.
Fixes#6308: people want to be able to link a unit file via 'systemctl enable'
from a git checkout or such and refer to other files in the same repo.
The new specifiers make that easy.
%y/%Y is used because other more obvious choices like %d/%D or %p/%P are
not available because at least on of the two letters is already used.
The new specifiers are only available in units. Technically it would be
trivial to add then in [Install] too, but I don't see how they could be
useful, so I didn't do that.
I added both %y and %Y because both were requested in the issue, and because I
think both could be useful, depending on the case. %Y to refer to other files
in the same repo, and %y in the case where a single repo has multiple unit files,
and e.g. each unit has some corresponding asset named after the unit file.
We would busily allocate an empty string to concatenate all of it's
zero characters to the output. Let's make things a bit simpler by letting
the specifier functions return NULL to mean "nothing to append".
This fixes repart's, systemctl's, sysusers' and tmpfiles' specifier
expansion to honour the root dir specified with --root=. This is
relevant for specifiers such as %m, %o, … which are directly sourced
from files on disk.
This doesn't try to be overly smart: specifiers referring to runtime
concepts (i.e. boot ID, architecture, hostname) rather than files on the
medium are left as is. There's certainly a point to be made that they
should fail in case --root= is specified, but I am not entirely convinced
about that, and it's certainly something we can look into later if
there's reason to.
I wondered for a while how to hook this up best, but given that quite a
large number of specifiers resolve to data from files on disks, and most
of our tools needs this, I ultimately decided to make the root dir a
first class parameter to specifier_printf().
Replaces: #16187Fixes: #16183
This moves the definition of the specifier table consisting only of
system and /tmp specifiers into generic code so that we can share it.
This patch only adds one user of it for now. Follow-up patches will add
more.
We recently started making more use of malloc_usable_size() and rely on
it (see the string_erase() story). Given that we don't really support
sytems where malloc_usable_size() cannot be trusted beyond statistics
anyway, let's go fully in and rework GREEDY_REALLOC() on top of it:
instead of passing around and maintaining the currenly allocated size
everywhere, let's just derive it automatically from
malloc_usable_size().
I am mostly after this for the simplicity this brings. It also brings
minor efficiency improvements I guess, but things become so much nicer
to look at if we can avoid these allocation size variables everywhere.
Note that the malloc_usable_size() man page says relying on it wasn't
"good programming practice", but I think it does this for reasons that
don't apply here: the greedy realloc logic specifically doesn't rely on
the returned extra size, beyond the fact that it is equal or larger than
what was requested.
(This commit was supposed to be a quick patch btw, but apparently we use
the greedy realloc stuff quite a bit across the codebase, so this ends
up touching *a*lot* of code.)
We usually call specifier_printf() and then check the validity of
the result. In many cases, validity checkers, e.g. path_is_valid(),
refuse too long strings. This makes specifier_printf() refuse such
long results earlier.
Moreover, unit_full_string() and description field in sysuser now
refuse results longer than LONG_LINE_MAX. config_parse() already
refuses the line longer than LONG_LINE_MAX. Hence, it should be ok
to set the same value as the maximum length of the resolved string.
This specifes two new optional fields for /etc/os-release:
IMAGE_VERSION= and IMAGE_ID= that are supposed to identify the image of
the current booted system by name and version.
This is inspired by the versioning stuff in
https://github.com/systemd/mkosi/pull/683.
In environments where pre-built images are installed and updated as a
whole the existing os-release version/distro identifier are not
sufficient to describe the system's version, as they describe only the
distro an image is built from, but not the image itself, even if that
image is deployed many times on many systems, and even if that image
contains more resources than just the RPMs/DEBs.
In particular, "mkosi" is a tool for building disk images based on
distro RPMs with additional resources dropped in. The combination of all
of these together with their versions should also carry an identifier
and version, and that's what IMAGE_VERSION= and IMAGE_ID= is supposed to
be.
As described in #15603, it is a fairly common setup to use a fqdn as the
configured hostname. But it is often convenient to use just the actual
hostname, i.e. until the first dot. This adds support in tmpfiles, sysusers,
and unit files for %l which expands to that.
Fixes#15603.
It would be very wrong if any of the specfier printf calls modified
any of the objects or data being printed. Let's mark all arguments as const
(primarily to make it easier for the reader to see where modifications cannot
occur).
These lines are generally out-of-date, incomplete and unnecessary. With
SPDX and git repository much more accurate and fine grained information
about licensing and authorship is available, hence let's drop the
per-file copyright notice. Of course, removing copyright lines of others
is problematic, hence this commit only removes my own lines and leaves
all others untouched. It might be nicer if sooner or later those could
go away too, making git the only and accurate source of authorship
information.
This part of the copyright blurb stems from the GPL use recommendations:
https://www.gnu.org/licenses/gpl-howto.en.html
The concept appears to originate in times where version control was per
file, instead of per tree, and was a way to glue the files together.
Ultimately, we nowadays don't live in that world anymore, and this
information is entirely useless anyway, as people are very welcome to
copy these files into any projects they like, and they shouldn't have to
change bits that are part of our copyright header for that.
hence, let's just get rid of this old cruft, and shorten our codebase a
bit.
This corresponds nicely with the specifiers we already pass for
/var/lib, /var/cache, /run and so on.
This is particular useful to update the test-path service files to
operate without guessable files, thus allowing multiple parallel
test-path invocations to pass without issues (the idea is to set $TMPDIR
early on in the test to some private directory, and then only use the
new %T or %V specifier to refer to it).
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.
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 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)
```
$ ./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)
```
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
