Adapted changes of 'Ignore carrier gainloss' #10187
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
When loading units, sometimes we'd first encounter a unit from .wants or .requires directory. A typical case would be when multi-user.target.wants/ contains a symlink to some unit. We would prepare to load this unit using /etc/systemd/system/multi-user.target.wants/foo.service as the fragment path. This is always wrong. Instead, let's use NULL as the path and let manager_load_unit() figure out the path on its own. Fixes #9921. path=0x5625ed9b01a0 "/usr/lib/systemd/system/local-fs.target.wants/systemd-remount-fs.service", e=0x0, _ret=0x7ffe64645000) at ../src/core/manager.c:1887 name=0x5625ed9b01ce "systemd-remount-fs.service", path=0x5625ed9b01a0 "/usr/lib/systemd/system/local-fs.target.wants/systemd-remount-fs.service", e=0x0, _ret=0x7ffe64645000) at ../src/core/manager.c:1961 name=0x5625ed9b01ce "systemd-remount-fs.service", path=0x5625ed9b01a0 "/usr/lib/systemd/system/local-fs.target.wants/systemd-remount-fs.service", add_reference=true, mask=UNIT_DEPENDENCY_FILE) at ../src/core/unit.c:2946 dir_suffix=0x5625ebb179ed ".wants") at ../src/core/load-dropin.c:95 path=0x0, e=0x0, _ret=0x7ffe646452c0) at ../src/core/manager.c:1965 name=0x5625ebb186f8 "local-fs.target", path=0x0, add_reference=true, mask=UNIT_DEPENDENCY_MOUNTINFO_IMPLICIT) at ../src/core/unit.c:2946 where=0x5625ed9b3cc0 "/tmp", options=0x5625ed947110 "rw,nosuid,nodev,seclabel", fstype=0x5625ed95be90 "tmpfs", flags=0x7ffe64645395) at ../src/core/mount.c:1439 where=0x5625ed9b3cc0 "/tmp", options=0x5625ed947110 "rw,nosuid,nodev,seclabel", fstype=0x5625ed95be90 "tmpfs", set_flags=false) at ../src/core/mount.c:1567 at ../src/core/mount.c:1635 ret_retval=0x7ffe64645660, ret_shutdown_verb=0x7ffe646456c0, ret_fds=0x7ffe646456d8, ret_switch_root_dir=0x7ffe646456b0, ret_switch_root_init=0x7ffe646456b8, ret_error_message=0x7ffe646456c8) at ../src/core/main.c:1669
This reverts commit be80154.
This reverts commit 48d3e88. I kept the follow-symlink=false → follow-symlink=true change instact, since we're likely to have existing installations with a symlink now.
This reverts commit d4e9e57. (systemd.conf.m4 part was already reverted in 5b5d826.) Together those reverts should "fix" #10025 and #10011. ("fix" is in quotes because this doesn't really fix the underlying issue, which is combining DynamicUser= with strict container sandbox, but it avoids the problem by not using that feature in our default installation.) Dynamic users don't work well if the service requires matching configuration in other places, for example dbus policy. This is true for those three services. In effect, distros create the user statically [1, 2]. Dynamic users make more sense for "add-on" services where not creating the user, or more precisely, creating the user lazily, can save resources. For "basic" services, if we are going to create the user on package installation anyway, setting DynamicUser= just creates unneeded confusion. The only case where it is actually used is when somebody forgets to do system configuration. But it's better to have the service fail cleanly in this case too. If we want to turn on some side-effect of DynamicUser=yes for those services, we should just do that directly through fine-grained options. By not using DynamicUser= we also avoid the need to restart dbus. [1] https://salsa.debian.org/systemd-team/systemd/commit/bd9bf307274faca24699c0c2d67cb86f18c0b2cb [2] https://src.fedoraproject.org/rpms/systemd/blob/48ac1cebdedb055d9daf3dfe28c7bde80103f7a1/f/systemd.spec#_473 (Fedora does not create systemd-timesync user.)
"because blacklisted" is not grammatical, it should be something like "becuase it is blacklisted", which in turn in too verbose. Let's use a shorter form.
We have "installed tests", but don't provide an easy way to run them. The protocol is very simple: each test must return 0 for success, 77 means "skipped", anything else is an error. In addition, we want to print test output only if the test failed. I wrote this simple script. It is pretty basic, but implements the functions listed above. Since it is written in python it should be easy to add option parsing (like running only specific tests, or running unsafe tests, etc.) I looked at the following alternatives: - Ubuntu root-unittests: this works, but just dumps all output to the terminal, has no coloring. - @ssahani's test runner [2] It uses the unittest library and the test suite was implented as a class, and doesn't implement any of the functions listed above. - cram [3,4] cram runs our tests, but does not understand the "ignore the output" part, has not support for our magic skip code (it uses hardcoded 80 instead), and seems dead upstream. - meson test Here the idea would be to provide an almost-empty meson.build file under /usr/lib/systemd/tests/ that would just define all the tests. This would allow us to reuse the test runner we use normally. Unfortunately meson requires a build directory and configuration to be done before running tests. This would be possible, but seems a lot of effort to just run a few binaries. [1] https://salsa.debian.org/systemd-team/systemd/blob/242c96addb06480ec9cd75248a5660f37a17b4b9/debian/tests/root-unittests [2] https://github.com/systemd/systemd-fedora-ci/blob/master/upstream/systemd-upstream-tests.py [3] https://bitheap.org/cram/ [4] https://pypi.org/project/pytest-cram/ Fixes #10069.
…g too Apparently the new "bpfilter" subsystem otherwise confuses us. See: https://lists.freedesktop.org/archives/systemd-devel/2018-September/041392.html
we might as well filter these too since negative PIDs have special semantics in kill(), and we should never trigger that...
If SPLIT_RELAX is specified, then it accepts unfinished quotes or missing separator after right quote.
This intends to clarify the rule udev_build_argv() uses.
This also modernize udev_event_spawn() a bit.
boot_loader_read_conf(), boot_entries_find(), boot_entries_load_config() all log their errors internally, hence no need to log a second or third time about the same error when they return.
Let's internalize these four calls as noone else calls them.
We shouldn't generate an error in that case, as the file is optional.
This is really confusing, let's try to clean this up a bit, in particular as there are two very similar concepts: 1. The boot loaders, i.e. the category you find systemd-boot, the Windows and Apple boot loaders in. These may typically be listed in the firmware's EFI variables. 2. The boot loader entries, as defined by the Boot Loader Spec. In this category you find the various Linux kernels that are installed, i.e. the stuff systemd-boot shows on screen. To make things confusing, the Windows and Apple boot loaders can appear both as boot loaders and as boot loader entries. This tries to establish the following nomenclature: "boot loaders" and "boot loader entries" for these two concepts.
Also, let's move the --help and --version items to the end of the list.
let's add an env var for this, as this really shouldn't be a top-level feature, as it turning off the validity checks certainly isn't advisable. Fixes: #4925
On EFI variables that aren't whitelisted in the kernel the FS_IMMUTABLE_FL is set, as protection against accidental removal/modification. Since our own variables do not appear in those whielists, and we are not changing these variables, let's unset the flag temporarily when needed. We restore the flag after all writes, just in case.
…n $BOOT in the "list" output Let's use the new LoaderEntries efi var for this, and show all entries we couldn't find via the config snippets.
the bootctl changes from PR #9437 (the boot counting PR)
Some uses of n_netif already assumed it had time size_t. Others were a bit sloppy. Let's clean this up.
…10314) Let's make sure the integers we parse out are not larger than USHRT_MAX. This is a good idea as the kernel's TIOCSWINSZ ioctl for sizing terminals can't take larger values, and we shouldn't risk an overflow.
CID #1396013.
Until a core dump handler is installed by systemd-sysctl, the generation of core dump for services is turned OFF which can make the debugging of the early boot process harder especially since there's no easy way to restore the core dump generation. This patch introduces a new kernel command line option which specifies an absolute path where the kernel should write the core dump file when an early process crashes. This will take effect until systemd-coredump (or any other handlers) takes over.
…d $COLUMNS This way users can directly influence the tty size if they like when nspawn is invoked as a service and thus stdin/stdout/stderr are not connected to a TTY.
Small cleanup for test-udev
Adapted changes of #7403 to merge with currect state.
@poettering thats because @ssahani did not like DHCP being touched and I added checks. (may be to keep him happy and have some progress in PR). I would have preferred code to not check anything at all because administrator is not supposed to set IgnoreCarrierGainLoss=true anyway for cases where it makes no sense. See: #7403 (comment) I kept asking other developers to consider my point of view because there were differences with opinions of @ssahani but there was almost no response. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Adapted changes of #7403 to merge with currect state.