Lennart Poettering [Thu, 31 Mar 2022 08:10:37 +0000 (10:10 +0200)]
udev: add /dev/disk/by-diskseq symlink for block devices
This adds another symlink for block devices:
/dev/disk/by-diskseq/<number>
where the number is the diskseq number as exposed by the kernel. It's
useful for apps because they can use it to open a device by diskseq, in
a way that is safe against device node reuse. I.e. if a device node path
like this is passed to an app it could open the device node via the
symlink and also parse the diskseq from the path. Once the device is
opened it could compare the parsed diskseq with the one returned by
BLKGETDISKSEQ on the open node, and if it matches they know they are
talking to the right device.
Fixes: #22906
Yu Watanabe [Fri, 1 Apr 2022 14:43:18 +0000 (23:43 +0900)]
Merge pull request #22926 from bluca/analyze_offline_filter
analyze: fix offline checks for syscall filter and 'native' architecture
Lennart Poettering [Fri, 1 Apr 2022 13:58:55 +0000 (15:58 +0200)]
Merge pull request #22921 from poettering/uid-range-tweaks
userns uid range tweaks: taint systemd if assigned userns uid range too short, and show userns uid range in userdbctl output
Lennart Poettering [Fri, 1 Apr 2022 12:26:50 +0000 (14:26 +0200)]
path-util: use STR_IN_SET() where appropriate
Lennart Poettering [Fri, 25 Mar 2022 11:03:17 +0000 (12:03 +0100)]
notify: remove spurious whitespace
Lennart Poettering [Fri, 1 Apr 2022 13:22:43 +0000 (15:22 +0200)]
Merge pull request #22934 from poettering/tls-test-fix-root
tests: make test-resolved-stream suceed even when run as root with restrictive access mode on build tree dir
Lennart Poettering [Fri, 1 Apr 2022 08:43:49 +0000 (10:43 +0200)]
process-util: refactor APIs for reading /proc/self/xyz symlinks
The three functions for reading cwd, exe and root symlinks of processes
already share a common core: get_process_link_contents(). Let's refactor
that a bit, and move formatting of the /proc/self/ path into this helper
function instead of doing that in the caller, thus sharing more code.
While we are at it, make the return parameters optional, in case the
information if the links are readable is interesting, but the contents
is not. (This also means safe_getcwd() and readlinkat_malloc() are
updated to make the return parameter optional, as these are called by
the relevant three functions)
Luca Boccassi [Thu, 31 Mar 2022 23:54:53 +0000 (00:54 +0100)]
analyze: fix offline check for syscal filter
The deny/allow list check was inverted, if we are deny listing and the
hashmap contains the syscall then that's good
Fixes https://github.com/systemd/systemd/issues/22914
Luca Boccassi [Thu, 31 Mar 2022 23:53:29 +0000 (00:53 +0100)]
analyze: fix offline check for 'native' syscall architecture
Enum values are stored in the set, not strings
Yu Watanabe [Fri, 1 Apr 2022 09:41:03 +0000 (18:41 +0900)]
Merge pull request #22872 from yuwata/udevadm-wait
udevadm: introduce 'wait' command
Lennart Poettering [Thu, 31 Mar 2022 13:22:33 +0000 (15:22 +0200)]
userdbctl: also show available UID range in current userns
Containers generally have a smaller UID range assigned than host
systems. Let's visualize this in the user/group tables. We insert
markers for unavailable regions. This way display is identical to status
quo ante on host systems, but in containers unavailable ranges will be
shown as that.
And while we are at it, also hide well-known UID ranges when they are
outside of userns uid_map range. This is mostly about the "container"
range. It's pointless showing the cotnainer range (i.e. a range UID >
65535) if that range isn#t available in the container anyway.
Lennart Poettering [Thu, 31 Mar 2022 12:39:24 +0000 (14:39 +0200)]
update TODO
Lennart Poettering [Thu, 31 Mar 2022 12:37:29 +0000 (14:37 +0200)]
pid1: add taint flag if uid/gid userns range too small
This will taint systemd if invoked in containers that do not have the
full 16bit range of UIDs defined.
we pretty much need uid root…nobody to be defined for a variety of
purposes, hence let's add this taint flag. Of course taints are
graceful, but it at least communicates the mess in some way...
Lennart Poettering [Thu, 31 Mar 2022 12:28:32 +0000 (14:28 +0200)]
uid-range: replace uid_range_contains() by more generalized uid_range_covers()
The former checks if one UID is inside the uid range set. The latter
checks if a full UID range is inside the uid range set. The former is
hence a special case of the latter.
Lennart Poettering [Thu, 31 Mar 2022 12:28:00 +0000 (14:28 +0200)]
uid-range: add some overflow checks
Lennart Poettering [Thu, 31 Mar 2022 11:28:54 +0000 (13:28 +0200)]
uid-range: add new uid_range_load_userns() for loading /proc/self/uid_map
Lennart Poettering [Thu, 31 Mar 2022 11:11:16 +0000 (13:11 +0200)]
test: port test-uid-range to tests.h
Lennart Poettering [Thu, 31 Mar 2022 10:56:24 +0000 (12:56 +0200)]
uid-range: use size_t for array size
Lennart Poettering [Fri, 1 Apr 2022 08:56:41 +0000 (10:56 +0200)]
test-resolved-stream: before entering user/network namespaces check if that's safe
I regularly run my tests also as root, since some of the tested code
uses privileged APIs. The test-resolved-stream so far tried to run its
tests in a user/network namespace if that can be allocated. This caused
the tests to fail on my system where once the user namespace is opened
access to the build tree in my $HOME is prohibited (due to restricted
access modes on my home dir). Let's add a check for that: before
actually isolating the test in a user/network namespace, let's see if
that would make it impossible for us to access the build tree (which we
need to do load the TLS certificates the test requires).
This should make the test pass when run as root from a build tree with
restrictive access mode.
Lennart Poettering [Fri, 1 Apr 2022 08:52:47 +0000 (10:52 +0200)]
tests: modernize load_testdata_env() a bit
Let's add assert() around everyhing we don't expect to fail.
Port to path_extract_directory().
Log errrors from load_env_file_pairs() which we ignore.
Lennart Poettering [Fri, 1 Apr 2022 09:03:15 +0000 (11:03 +0200)]
update TODO
Yu Watanabe [Wed, 30 Mar 2022 20:04:56 +0000 (05:04 +0900)]
udev: use sd_device_open() where appropriate
Yu Watanabe [Wed, 30 Mar 2022 18:54:41 +0000 (03:54 +0900)]
test: add test for sd_device_open()
Yu Watanabe [Wed, 30 Mar 2022 18:29:23 +0000 (03:29 +0900)]
sd-device: introduce sd_device_open()
We usually open() device node obtained by sd_device_get_devname().
However, the device node corresponds to the sd-device object may be
already removed, and another device node with the same path may be
created, hence an unexpected device may be opened.
The sd_device_open() opens device node, and checks the devnum and
diskseq of opened devnum, to avoid the above possibility.
Prompted by https://github.com/systemd/systemd/issues/22906#issuecomment-
1082736443.
Yu Watanabe [Wed, 30 Mar 2022 18:25:45 +0000 (03:25 +0900)]
fd-util: rename loop_get_diskseq() -> fd_get_diskseq()
And move it from loop-util.[ch] -> fd-util.[ch]
Yu Watanabe [Wed, 30 Mar 2022 18:23:33 +0000 (03:23 +0900)]
basic/missing: move BLKGETDISKSEQ to missing_fs.h
As it is defined at linux/fs.h.
Yu Watanabe [Fri, 25 Mar 2022 20:38:18 +0000 (05:38 +0900)]
test: replace helper_wait_for_dev() with 'udevadm wait'
Yu Watanabe [Fri, 25 Mar 2022 20:01:40 +0000 (05:01 +0900)]
udevadm: introduce new 'wait' command
Prompted by https://github.com/systemd/systemd/pull/22717#issuecomment-
1067348496.
The new command 'udevadm wait' waits for device or device symlink being
created. This may be useful to wait for a device is processed by udevd
after e.g. formatting or partitioning the device.
Yu Watanabe [Mon, 28 Mar 2022 18:57:49 +0000 (03:57 +0900)]
test: add more tests for sd_device_new_from_xxx()
Yu Watanabe [Sun, 27 Mar 2022 14:38:36 +0000 (23:38 +0900)]
sd-device: introduce sd_device_new_from_devname()
and sd_device_new_from_path() which takes devname or syspath.
Lennart Poettering [Wed, 30 Mar 2022 14:11:28 +0000 (16:11 +0200)]
systemctl: show tainted state
Lennart Poettering [Wed, 30 Mar 2022 08:46:16 +0000 (10:46 +0200)]
pid1: check for kernels older than baseline
Let's make this detectable explicitly.
Luca Boccassi [Thu, 31 Mar 2022 21:11:03 +0000 (22:11 +0100)]
Merge pull request #22923 from poettering/userns-check-refactor
virt: minor running_in_userns() modernizations
Lennart Poettering [Thu, 31 Mar 2022 14:50:37 +0000 (16:50 +0200)]
Merge pull request #22919 from poettering/cryptsetup-tweaks
various minor tweaks to cryptsetup/veritysetup/integritysetup
Lennart Poettering [Thu, 31 Mar 2022 11:28:18 +0000 (13:28 +0200)]
virt: use read_virtual_file() for reading /proc/self/setgroups
Lennart Poettering [Thu, 31 Mar 2022 11:27:21 +0000 (13:27 +0200)]
virt: simplify userns_has_mapping() by using fscanf() instead of scanf()
And while we are at it, also fix propagation of an uninitialized errno
error.
Antonio Alvarez Feijoo [Thu, 31 Mar 2022 08:09:29 +0000 (10:09 +0200)]
cryptsetup: fall back to traditional unlocking if any TPM2 operation fails
If any TPM2 operation fails, the boot process should continue and
prompt for a text password (if configured to do so).
Fixes #22870
Lennart Poettering [Thu, 31 Mar 2022 12:27:13 +0000 (14:27 +0200)]
update TODO
Martin Liska [Thu, 31 Mar 2022 08:27:45 +0000 (10:27 +0200)]
Support -D_FORTIFY_SOURCE=3 by using __builtin_dynamic_object_size.
As explained in the issue, -D_FORTIFY_SOURCE=3 requires usage
of __builtin_dynamic_object_size in MALLOC_SIZEOF_SAFE macro.
Fixes: #22801
Franck Bui [Thu, 31 Mar 2022 09:17:10 +0000 (11:17 +0200)]
meson: build kernel-install man page when necessary
Sebastian Pucilowski [Thu, 31 Mar 2022 05:31:28 +0000 (16:31 +1100)]
Fix "link-local" language inconsistencies
"Link-local" and "link local" are used throughout man pages and program
output, with the former used far more than the latter. This commit makes
it consistent throughout the project.
Lennart Poettering [Thu, 31 Mar 2022 09:22:07 +0000 (11:22 +0200)]
integritysetup: also validate volume name
Exactly like for veritysetup/cryptsetup
Lennart Poettering [Thu, 31 Mar 2022 09:21:37 +0000 (11:21 +0200)]
integritysetup: also port to mangle_none()
Let's make the tool work more like veritysetup/cryptsetup in this regard
too.
Lennart Poettering [Thu, 31 Mar 2022 09:20:25 +0000 (11:20 +0200)]
integritysetup: log when attempted to detach already detached volume
To make the tool behave more like cryptsetup/veritysetup
Lennart Poettering [Thu, 31 Mar 2022 09:20:01 +0000 (11:20 +0200)]
integritysetup: rename action → verb, to match other code
Lennart Poettering [Thu, 31 Mar 2022 09:09:48 +0000 (11:09 +0200)]
tree-wide: unify some code that looks for --help in the command line
Lennart Poettering [Thu, 31 Mar 2022 09:03:06 +0000 (11:03 +0200)]
veritysetup: do some superficial checking on volume name
cryptsetup does this too, so let's better be safe here, too.
Lennart Poettering [Thu, 31 Mar 2022 09:01:52 +0000 (11:01 +0200)]
veritysetup: mangle option strings like in cryptsetup
Lennart Poettering [Thu, 31 Mar 2022 08:52:50 +0000 (10:52 +0200)]
veritysetup: give command line parameters proper names
Accessing the various arguments always through argv[] is nasty, since
it's not obvious what we are talking about here. Let's give things nice
names.
We did the same in cryptsetup a while back.
Lennart Poettering [Thu, 31 Mar 2022 08:49:30 +0000 (10:49 +0200)]
cryptsetup: adjust some log levels
Let's upgrade log levels of some noteworthy messages from LOG_DEBUG to
LOG_NOTICE. These messages contain information that previous log
messages in the error path didn't say, namely that we'll now fall back
to traditional unlocking.
Note that this leaves similar log messages for cases where
TPM2/PKCS#11/FIDO2 support is disabled at build at LOG_DEBUG, since in
that case nothing really failed, we just systematically can't do
TPM2/PKCS#11/FIDO2 and hence it is pointless and not actionable for
users to do anything about it...
Lennart Poettering [Thu, 31 Mar 2022 08:48:37 +0000 (10:48 +0200)]
cryptsetup: add helper for mangling "none" option strings
let's unify some code here, and let's do so in cryptsetup-util.h so that
we can later reuse this in integritysetup/veritysetup
Lennart Poettering [Thu, 31 Mar 2022 08:47:24 +0000 (10:47 +0200)]
cryptsetup: rename functions that try to do FIDO2/TPM2/PKCS#11 via cryptsetup plugins to say so
The are so many different flavours of functions that attach volumes,
hence say explicitly that these are about libcryptsetup plugins, and
nothing else.
Just some renaming, no code changes beyond that.
Yu Watanabe [Thu, 31 Mar 2022 02:40:30 +0000 (11:40 +0900)]
Merge pull request #22899 from yuwata/network-ignore-carrier-loss
network: automatically determine timeout of waiting for carrier regain
Yu Watanabe [Tue, 29 Mar 2022 16:04:26 +0000 (01:04 +0900)]
network: shorten code a bit
Currently, there exist only two MTU sources, static and DHCPv4, and they
are exclusive. Hence, it is not necessary to check the existence of the
MTU option in the acquired DHCP lease. Let's unconditionally reset the
MTU. Note that, if the current and original MTU are equivalent, then
link_request_to_set_mtu() handles that gracefully.
Yu Watanabe [Tue, 29 Mar 2022 15:52:09 +0000 (00:52 +0900)]
network: automatically determine timeout of waiting for carrier regain
The commit
6706ce2fd2a13df0ae5e469b72d688eaf643dac4 made
IgnoreCarrierLoss= setting also take timespan, to make users handle
issues like #18738 or #20887. But still users needed to explicitly set
a timespan.
This makes networkd automatically determine the timeout when the
situations #18738 or #19832 is detected. Unfortunately, still users have
issue #20887 need to specify a value.
Closes #19832.
Yu Watanabe [Thu, 31 Mar 2022 00:24:38 +0000 (09:24 +0900)]
Merge pull request #22913 from yuwata/sd-device-cleanups
sd-device,udev: several cleanups
Zbigniew Jędrzejewski-Szmek [Wed, 30 Mar 2022 07:38:33 +0000 (09:38 +0200)]
veritysetup: fix parsing of root-hash-signature= option
The function was named confusingly and we managed to confused ourselves. The
parameter was assigned incorrectly and then reassigned correctly in the caller.
Let's simplify the whole thing by just saving the optarg param.
I considered moving the unhexmemming and/or reading of the file to the parse
function, but decided against it. I think it's nicer to parse all options
before opening external files.
Yu Watanabe [Wed, 30 Mar 2022 19:31:46 +0000 (04:31 +0900)]
udev: rename functions to emphasize whole disk is locked
Yu Watanabe [Wed, 30 Mar 2022 19:26:22 +0000 (04:26 +0900)]
udev: ignore one more error in device_get_block_device()
Yu Watanabe [Wed, 30 Mar 2022 19:14:49 +0000 (04:14 +0900)]
sd-device: do not ignore critical errors in device_new_from_child()
Yu Watanabe [Wed, 30 Mar 2022 19:11:30 +0000 (04:11 +0900)]
sd-device: use path_extract_directory() at one more place
Yu Watanabe [Wed, 30 Mar 2022 18:27:17 +0000 (03:27 +0900)]
sd-device: try to get DISKSEQ from uevent file
Otherwise, if the sd-device object is created from e.g. syspath, then
sd_device_get_diskseq() returns -ENOENT.
Yu Watanabe [Mon, 28 Mar 2022 19:04:54 +0000 (04:04 +0900)]
sd-device: drop /sys/subsystem support
Follow-ups for
37cf83d9bfdd9f6859b6f2654d8ec3bbb17873b2.
Gaël PORTAY [Thu, 25 Feb 2021 18:02:26 +0000 (13:02 -0500)]
man: update root-hash-signature option with value
This documents two possible values expected by the option
root-hash-signature for veritytab and veritysetup-generator.
Yu Watanabe [Wed, 30 Mar 2022 17:04:44 +0000 (02:04 +0900)]
udev: do not use sd_event_source_disable_unref() at more places
Fixes a bug introduced by
9612da361a825d70a9fd392f3ee5a53bf8896887.
Follow-up for
f777e745a7966ea52ef29f9e4edfdd16874cfe86.
Yu Watanabe [Wed, 30 Mar 2022 09:39:50 +0000 (18:39 +0900)]
udev: do not append unknown errno or signal name
Follow-up for
6467bda59d571696b645e8bbdf31926676890956.
Addresses https://github.com/systemd/systemd/pull/22871#discussion_r837705779.
Lennart Poettering [Wed, 30 Mar 2022 08:45:31 +0000 (10:45 +0200)]
update TODO
Frantisek Sumsal [Wed, 30 Mar 2022 09:32:31 +0000 (11:32 +0200)]
ci: drop clang 11 & add clang 14
Yu Watanabe [Wed, 30 Mar 2022 12:10:06 +0000 (21:10 +0900)]
fix typo
Luca Boccassi [Tue, 29 Mar 2022 20:54:01 +0000 (21:54 +0100)]
NEWS: specify that public headers are still C89
Luca Boccassi [Tue, 29 Mar 2022 20:52:21 +0000 (21:52 +0100)]
NEWS: mention that C11 is now used
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 18:28:47 +0000 (20:28 +0200)]
NEWS: add entry for the unit enablement stuff
It should be merged soon.
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 20:39:08 +0000 (22:39 +0200)]
test-systemctl-enable: skip test for %v if kver is not a valid instance
On arm, we'd fail with:
target@v:5.16.8-200.fc35.armv7hl+lpae.socket: not a valid unit name "target@v:5.16.8-200.fc35.armv7hl+lpae.socket": Invalid argument
наб [Tue, 29 Mar 2022 17:43:01 +0000 (19:43 +0200)]
test-copy: use non-0 data block in copy_holes
Some filesystems (e.g. zfs with compression!=off, which is the default
configuration) automatically hole-punch all-zero blocks ‒ write a block
full of ones instead
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 19:10:03 +0000 (21:10 +0200)]
Merge pull request #22649 from keszybz/symlink-enablement-yet-again-punish-me-harder
Fixups to the unit enablement logic
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 19:08:44 +0000 (21:08 +0200)]
Merge pull request #22871 from yuwata/udev-worker-error-code
udev: append error code on failure
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 17:46:47 +0000 (19:46 +0200)]
meson: bump numbers for v251-rc1
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 17:43:54 +0000 (19:43 +0200)]
NEWS: update contributor list
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 17:42:01 +0000 (19:42 +0200)]
tools/git-contrib: list contributions not only from Weblate
It seems that --invert-grep used to affect --author, but now it doesn't (with
git-2.35.1-1.fc36.x86_64), so effectively we would only show the one entry that
was supposed to be filtered out.
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 17:36:29 +0000 (19:36 +0200)]
NEWS: two more small features and some rewordings
наб [Tue, 29 Mar 2022 16:32:07 +0000 (18:32 +0200)]
Fix typos in user-util.c and dbus-unit.c
Yu Watanabe [Tue, 29 Mar 2022 05:58:58 +0000 (14:58 +0900)]
test: add tests for worker error code
Yu Watanabe [Fri, 25 Mar 2022 17:01:42 +0000 (02:01 +0900)]
udev: append error code in broadcasted message
So, libudev listners can easily detect errors.
Addresses https://github.com/systemd/systemd/pull/22717#discussion_r834420878.
Yu Watanabe [Fri, 25 Mar 2022 17:00:02 +0000 (02:00 +0900)]
sd-device: introduce device_add_propertyf()
Luca Boccassi [Tue, 29 Mar 2022 16:22:08 +0000 (17:22 +0100)]
NEWS: mention kernel requirement change 3.13 -> 3.15
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 16:17:32 +0000 (18:17 +0200)]
Merge pull request #22894 from keszybz/hwdb-update
Update hwdb again
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 10:22:21 +0000 (12:22 +0200)]
hwdb: update autosuspend entries
Zbigniew Jędrzejewski-Szmek [Thu, 9 Dec 2021 10:05:15 +0000 (11:05 +0100)]
hwdb: update for v251
As usual, there are mostly additions of new entries, and some spelling
correction and company renames, no big removals.
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 10:17:51 +0000 (12:17 +0200)]
hwdb: fix parser to work with newer pyparsing
The handling of whitespace in pyparsing is a bother. There's some
global state, and per-element state, and it's hard to get a handle on
things. With python3-pyparsing-2.4.7-10.fc36.noarch the grammar would
not match. After handling of tabs was fixed to not accept duplicate tabs,
the grammar passes.
It seems that the entry for usb:v8087p8087*
was generated incorrectly because we treated the interface line
(with two TABs) as a device line (with one TAB).
Zbigniew Jędrzejewski-Szmek [Fri, 25 Mar 2022 09:54:39 +0000 (10:54 +0100)]
cryptsetup: shorten message a bit
If it is reported as missing, we don't need to say that we assume
it is missing.
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 16:11:32 +0000 (18:11 +0200)]
Merge pull request #22843 from poettering/bootspec-json
bootctl: bootspec improvements and clean-ups
Zbigniew Jędrzejewski-Szmek [Tue, 29 Mar 2022 13:55:59 +0000 (15:55 +0200)]
Rename UnitFileScope to LookupScope
As suggested in
https://github.com/systemd/systemd/pull/22649/commits/
8b3ad3983f5440eef812b34e5ed862ca59fdf7f7#r837345892
The define is generalized and moved to path-lookup.h, where it seems to fit
better. This allows a recursive include to be removed and in general makes
things simpler.
Zbigniew Jędrzejewski-Szmek [Mon, 28 Mar 2022 18:20:09 +0000 (20:20 +0200)]
test-systemctl-enable: disable the test for %a for now
Zbigniew Jędrzejewski-Szmek [Mon, 28 Mar 2022 18:03:37 +0000 (20:03 +0200)]
test-systemctl-enable: also use freshly-built systemd-id128
Tests were failing on centos7 because systemd-id128 is not in path.
Zbigniew Jędrzejewski-Szmek [Fri, 25 Mar 2022 14:56:16 +0000 (15:56 +0100)]
test-systemctl-enable: use magic syntax to allow inverted tests
Inspired by
7910ec3bcde2ee0086b3e49f8aaa2a9f13f58d97.
'! true' passes, because it's a conditional expression.
But '( ! true )' fails, because '( … )' creates a subshell, i.e. a separate
program, and '! true' becomes the return value of that program, and the whole
thing apparently is not a conditional expression for the outer shell.
This is shorter, so let's just do this.
Zbigniew Jędrzejewski-Szmek [Fri, 25 Mar 2022 14:43:27 +0000 (15:43 +0100)]
shared/install: when creating symlinks, accept different but equivalent symlinks
We would only accept "identical" links, but having e.g. a symlink
/usr/lib/systemd/system/foo-alias.service → /usr/lib/systemd/system/foo.service
when we're trying to create /usr/lib/systemd/system/foo-alias.service →
./foo.service is OK. This fixes an issue found in ubuntuautopkg package
installation, where we'd fail when enabling systemd-resolved.service, because
the existing alias was absolute, and (with the recent patches) we were trying
to create a relative one.
A test is added.
(For .wants/.requires symlinks we were already doing OK. A test is also
added, to verify.)
Zbigniew Jędrzejewski-Szmek [Thu, 24 Mar 2022 10:52:35 +0000 (11:52 +0100)]
test-systemctl-enable: make shellcheck happy
Quoting is not necessary in many places, but I think it's nicer
to use it consistently.
Zbigniew Jędrzejewski-Szmek [Thu, 17 Mar 2022 15:02:10 +0000 (16:02 +0100)]
shared/install: fix handling of a linked unit file
When we have a symlink that goes outside of our search path, we should just
ignore the target file name. But we were verifying it, and rejecting in
the case where a symlink was created manually.
Zbigniew Jędrzejewski-Szmek [Thu, 17 Mar 2022 14:50:16 +0000 (15:50 +0100)]
shared/install: split UNIT_FILE_SYMLINK into two states
The two states are distinguished, but are treated everywhere identically,
so there is no difference in behaviour except for slighlty different log
output.
Zbigniew Jędrzejewski-Szmek [Thu, 17 Mar 2022 10:46:03 +0000 (11:46 +0100)]
basic/unit-file: reverse negative conditional
Having the reverse condition first makes changes that I want to do
later awkward, so reverse it as a separate step first.