Since the file headers might be replaced by zeroed pages now due to
sigbus we should make sure we don't end up dividing by zero because we
don't check values read from journal file headers for changes.
This makes them robust regarding truncation. Ideally, we'd export this
as an API, but given how messy SIGBUS handling is, and the uncertain
ownership logic of signal handlers we should not do this (unless libc
one day invents a scheme how to sanely install SIGBUS handlers for
specific memory areas only). However, for now we can still make all our
own tools robust.
Note that external tools will only have read-access to the journal
anyway, where SIGBUS is much more unlikely, given that only writes are
subject to disk full problems.
Even though we use fallocate() it appears that file systems like btrfs
will trigger SIGBUS on certain low-disk-space situation. We should
handle that, hence catch the signal, add it to a list of invalidated
pages, and replace the page with an empty memory area. After each write
check if SIGBUS was triggered, and consider the write invalid if it was.
This should make journald a lot more robust with file systems where
fallocate() is not reliable, for example all CoW file systems
(btrfs...), where changing written data can fail with disk full errors.
https://bugzilla.redhat.com/show_bug.cgi?id=1045810
-n is only allowed for root. /etc/mtab is nowadays almost always a link to /proc/,
so in practice this does not really matter too much, but should allow .mount units
to work in --user mode.
https://bugs.freedesktop.org/show_bug.cgi?id=87602
Those values are based on a file we read from disk, so we should
verify everything we receive, and make sure everything we print
is sensible.
Also, print fractional seconds for TTL.
Fix a bunch of needless memzero() calls, a bunch of use-after-free
regarding _cleanup_free_ and drop unused variables.
Hint: Do NOT use _cleanup_free_ for temporary strappend() helpers that are
freed multiple times. All you safe is the last free() call, which is
really not worth the trouble resetting it to NULL all the time.
This partially reverts:
commit f131770b14
Author: Veres Lajos <vlajos@gmail.com>
Date: Mon Dec 29 09:45:58 2014 +0000
tree-wide: spelling fixes
The commit in question changed a binary file. I didn't look at the diff in
particular, so I have no idea what exactly was changed. However, the file
is generated and it looked highly suspiciuous. Therefore, I reverted that
part.
Note that this is generated by "make update-unifont" so really no reason
to touch at all.
If some sleep operation was not possible (e.g. because swap is missing),
we would try twice: once through logind, which would result in a clean error:
Failed to execute operation: Sleep verb not supported
and then second time by starting the appropriate unit directly, which is
more messy. If logind tells us that something is not possible (or already
in progress), report that to the user and quit. If logind is present and working
we should not try to work around it.
Loosely based on https://bugs.freedesktop.org/show_bug.cgi?id=87832.
ENOSYS is used to signify compiled-out functionality. Using it for
different kinds of error is misleading.
For BUS_ERROR_SLEEP_VERB_NOT_SUPPORTED, logind-action.c uses ENOTSUP
already, so changing it to ENOTSUP makes the dbus and action paths
behave the same.
This implements two new helpers, discussed on systemd-devel about 1 year
ago:
sd_bus_emit_object_added()
sd_bus_emit_object_removed()
Both calls are equivalent to their respective counterpart
sd_bus_emit_interfaces_{added/removed}(), but can figure out the list of
interfaces themselves, instead of requiring the caller to provide them.
Furthermore, both calls properly deal with builtin interfaces provided via
org.freedesktop.DBus.* and alike.
Both calls simply traverse a node and all its parent nodes to figure out a
list of all interfaces registered as vtable or fallback. It then appends
each of them, similar to the interfaces_{added/removed}() helpers.
Note that interfaces_{added/removed}() runs a parent traversal for *each*
passed interface. Therefore, it can simply bail out, once it found a
parent node that implements a given interface.
With object_{added/removed}() we cannot know the registered interfaces in
advance, thus, we cannot run one traversal per node. Instead, we run a
single traversal and remember all interfaces that we added. Therefore, a
child-interface overrides all conflicting parent-interfaces. We keep a
"Set *s" context to track those while climbing up the tree.
The kernel provides capabilities as a u32 array, sd-bus uses an u8 array.
This works fine on little-endian as both are encoded the same way.
However, this fails on big-endian if we do not perform sufficient
byte-swapping on each u32 entry.
This patch makes sd-bus use u32, too. We avoid changing any kernel
provided data so we can keep pointing into kdbus pool buffers which
contain u32 arrays.
The number of available caps can be read from
/proc/sys/kernel/cap_last_cap during runtime. Our helper cap_last_cap()
does that, so there's no reason to remember the size of any capability
cache. We can just pre-allocate arrays with a suitable size for all
available caps and reject any higher caps.
The kernel capability API uses u32 as base so make sure we do the same.
Note that this is specified by POSIX, so it's unlikely to change.
This macro calculates A / B but rounds up instead of down. We explicitly
do *NOT* use:
(A + B - 1) / A
as it suffers from an integer overflow, even though the passed values are
properly tested against overflow. Our test-cases show this behavior.
Instead, we use:
A / B + !!(A % B)
Note that on "Real CPUs" this does *NOT* result in two divisions. Instead,
instructions like idivl@x86 provide both, the quotient and the remainder.
Therefore, both algorithms should perform equally well (I didn't verify
this, though).
This reverts commit 206e7a5f7b.
We actually want to allow shutting down containers that use
RegisterMachine() rather than CreateMachine() to register their own
unit. It should be safe to do so, since the primary usecase for
RegisterMachine() are container managers that run only a single
container within their own unit, such as systemd-nspawn.
This file was introduced with linux-3.2, use it instead of probing for it
via prctl(PR_CAPBSET_READ).
For now, keep the old code for backwards compat. We can drop it once 3.2
is our lowest requirement.
The test-cap-list code is extended to verify cap_last_cap() is the same as
we'd get via prctl probing and /proc.
All we care about is that the kernel (pid==0) sent the message. Verifying the sender uid
seems to break when using userns.
Reported by Stéphane Graber.
Lennart Poettering
Thomas Hindoe Paaboel Andersen
Zbigniew Jędrzejewski-Szmek
Topi Miettinen
Robert Milasan
Piotr Drąg
David Herrmann
Veres Lajos
Tom Gundersen