Otherwise we might still use the more restrictive limit when we get to
STDOUT_STREAM_RUNNING, causing unexpected line breaks in the journal
messages.
Follow-up to: 80e9720616Resolves: #29916
Currently empty epochs are not sealed. This allows an attacker to truncate
a sealed log and continue it without any problems showing when verifying the
log.
This partially addresses CVE-2023-31438. One way to extend this change to
address CVE-2023-31438 completely, would be to verify that there is exactly
one seal per epoch (and not sealing when the epoch has not ended yet).
the change also adds a journal-file flag: HEADER_COMPATIBLE_SEALED_CONTINUOUS
this flag indicates that a journal file is sealed continuously and decides whether
any missing crypto epochs should trigger a warning or an error.
Let's install a simple no-op signal handler without SA_RESTART for
SIGINT/SIGTERM, which allows us to interrupt read_one_char() and follow
it up with a proper cleanup, including restoring the vt to the original
state.
Resolves: #29478
We use it for more than just pipe() arrays. For example also for
socketpair(). Hence let's give it a generic name.
Also add EBADF_TRIPLET to mirror this for things like
stdin/stdout/stderr arrays, which we use a bunch of times.
The function isn't necessarily fast (it's O(n)), and there's no reason
to have it defined as inline function, since it's neither fast, nor
entirely trivial.
This is preparation for #28891, which adds a bunch more helpers around
"struct iovec", at which point this really deserves its own .c/.h file.
The idea is that we sooner or later can consider "struct iovec" as an
entirely generic mechanism to reference some binary blob, and is the
go-to type for this purpose whenever we need one.
This allows distros to install configuration file templates in /usr/lib/systemd
for example.
Currently we install "empty" config files in /etc/systemd/. They serve two
purposes:
- The file contains commented-out values that show the default settings.
- It is easier to edit the right file if it is already there, the user doesn't
have to type in the path correctly, and the basic file structure is already in
place so it's easier to edit.
Things that have happened since this approach was put in place:
- We started supporting drop-ins for config files, and drop-ins are the
recommended way to create local configuration overrides.
- We have systemd-analyze cat-config which takes care of iterating over
all possible locations (/etc, /run, /usr, /usr/local) and figuring out
the right file.
- Because of the first two points, systemd-analyze cat-config is much better,
because it takes care of finding all the drop-ins and figuring out the
precedence. Looking at files manually is still possible of course, but not
very convenient.
The disadvantages of the current approach with "empty" files in /etc:
- We clutter up /etc so it's harder to see what the local configuration actually is.
- If a user edits the file, package updates will not override the file (e.g.
systemd.rpm uses %config(noreplace). This means that the "documented defaults"
will become stale over time, if the user ever edits the main config file.
Thus, I think that it's reasonable to:
- Install the main config file to /usr/lib so that it serves as reference for
syntax and option names and default values and is properly updated on package
upgrades.
- Recommend to users to always use drop-ins for configuration and
systemd-analyze cat-config to view the documentation.
This setting makes this change opt-in.
Fixes#18420.
[zjs: add more text to the description]
systemd's own cgroup hierarchy is special to us, we use it to actually
manage processes. Because of that many calls tha apply to cgroups are
only ever called with the SYSTEMD_CGROUP_CONTROLLER as controller
argument. Let's hence remove the argument altogether.
This in particular touches the kill and xattr routines.
This changes no behaviour, we just drop an argument that is always set
to the same value anyway.
This is preparation to eventually getting rid of the cgroupvs1, because
on cgroupvs2 the cgroup paths do not change for different controllers,
there's only a single hierarchy there.
Because journal_file_next_entry_for_data() provides the first entry, while
journal_file_next_entry() actually provides the next entry of the input,
this also renames it to journal_file_move_to_entry_for_data().
Also, previously, on DIRECTION_UP the function did not fall back to the
'extra' entry when all entries linked in the chained array are broken.
This also fixes the issue, and now it fall back to the extra entry.
In 9289e093ae we started using install_emptydir().
When running unprivileged, 'DESTDIR=… meson install -C build --quiet --no-rebuild'
would emit two warnings:
'…/var/log/journal': Unable to set owner 'root' and group 'root': Operation not permitted, ignoring...
'…/var/log/journal/remote': Unable to set owner 'root' and group 'root': Operation not permitted, ignoring...
Those were the only two install_emptydir()s that specified ownership.
Let's drop the user/group specification to get rid of the warning.
When installing as root, we will create a root-owned directory anyway.
When not running as root, we cannot create a root-owned directory.
So this specification only makes a difference if we are running as root,
and the directory already existed, and was not owned by root. In that case,
I think it's actually better to leave the existing modification in place.
(E.g. maybe the admin chgrp'ed the ownership for whatever reason. We might
just as well leave that in place.)
Otherwise, sd_journal_previous() -> real_journal_next(DIRECTION_UP) ->
next_beyond_location() wrongly handles that previously we hit EOF of
the file, and returns 0 without finding a matching entry.
Fixes#29216.
The monotonic clock value consistency should only be checked if the boot
ID is the same as the last journal entry, and requiring that the current
boot ID be the same as the boot ID of the last entry is not necessary
for ensuring deterministic bisection as we already enforce strict
ordering of the real time clock value in a single journal file.
This fixes an issue where a new journal file is unnecessarily created
every boot, potentially wasting storage space.
We went back-and-forth a bit on this. Very old meson would print a message
about detecting the program if a quoted argument was used, leading to a lot of
noise. So we started to convert various places to use the variable, but then it
turned out that meson < 0.56.2 doesn't handle this correctly and we reverted to
using strings everywhere in 7c22f07cbd. Then at
some point we stopped supporting old meson and over time we started using the
variable in various places again, somewhat inconsistently. Then most calls to
'sh' were removed in 9289e093ae when
install_emptydir() builtin started being used.
Now meson allows either the string or variable to be used, and doesn't print a
message if the string is used. Let's use the variable everywhere. For 'sh', we
could do either, but for other variables, we _do_ want the detection to happen,
for example for git, find, awk, which might not be installed and we want to
detect that early, before we start the build. It would be ugly to use quotes
for some programs, but not for others. Also, a string is still refused for
test(), so we couldn't use the string version even if we didn't care about
detection.
- add reference to the service unit in the man page,
- fix several indentation and typos,
- replace '(uint64_t) -1' with 'UINT64_MAX',
- drop unnecessary 'continue'.
Frantisek Sumsal
Felix Dörre
Lennart Poettering
Yu Watanabe
Luca Boccassi
Franck Bui
Zbigniew Jędrzejewski-Szmek
Daan De Meyer
Joerg Behrmann
Mike Yuan
Thorsten Kukuk
msizanoen
OMOJOLA JOSHUA