Update the CHANGELOG.md and meson.

Another glib fix "gspawn: Report errors with closing file descriptors
between fork/exec" is breaking CI because glib now treats `EPERM`
returned from `close_range` differently which breaks running inside
docker which apparently returns `EPERM`.
Since fixing the CI env and also patching glib in debian (like Fedora
did) is both out of my control, the only option is to exclude Debian
from testing for now.
See also https://gitlab.gnome.org/Infrastructure/GitLab/-/issues/545

Apparently calling `g_log_set_writer_func` was never meant to be
called more than once, even though it worked just fine. Work
around this by introducing another level of indirection: A new
`TestContext` struct is passed to all tests as `user_data` and
in this we save keep track of the current logger. It is reset
to the standard logger for the tests in the test's setup
function.
NB: since the test data is passed via `gconstpointer` we need
to explicitly cast the const away. This is ok since the actual
`TestContext` object is non-const.

udev/systemd, starting version 251 (a791061) started to verify that
paths are actually on sysfs, i.e. the fs type is SYSFS_MAGIC. Since
we use umockdev to emulate sysfs on normal file system this check
will now fail. There is an exit hatch in form of an environment
variable called `SYSTEMD_DEVICE_VERIFY_SYSFS` that when set to `0`
makes systemd/udev forgo that check.  umockdev >= 0.17.10 does set
that variable itself but in Python changes to the env by libraries
does not effect the env of the running script. Thus we have to set
it ourselves.

Update the CHANGELOG.md and meson.

We were using features from 0.46.0, so let's require this and not
trigger a warning during configure anymore.

Fix the name and update the list of packages to better match
this project.

This was fixed in umockdev 0.16.3, but now causes warning messages due
to the existing directories. This at least breaks Debian/Ubuntu's
autopkgtests.

Always initialize the syspath member of BoltDevInfo in
info_for_device, so the struct has sane values even if
the function fails.

This got lost by when the other `only` keyword was added.

Create a devcontainer to use with vs code.

Only run the coverit job if the respective needed exist.

The version of uncrustify in F34 wants it that way, so be it.

Move g_autofree so that it is after g_auto, g_autoptr.

Declare that the integration tests depend on `boltd` and `boltctl`.

meson 0.59.0 breaks "contains" for dependency objects[1], which we
use to conditionally set the umockdev ld preload. So this breaks
the umockdev based tests. Since Fedora 33 and above updated to this
version it breaks bolt[2]. Reverse gating tests would be nice.

[1] https://github.com/mesonbuild/meson/issues/9047
[2] https://bugzilla.redhat.com/show_bug.cgi?id=1994006

Install glibc-langpack-en in the fedora based container.

Fedora 32 is EOL.

Fedora 32 is EOL.

Due to a bug in system [1], the unit tests currently fail on
arch, so allow this to fail for now.
[1] https://github.com/systemd/systemd/pull/19265

Apparently the arch container got renamed to archlinux/archlinux.
Also use the base-devel tag.
Drop the removal of package-cleanup.hook.

Use curl isntead of wget in the container to fetch coverity.

The license `GPL-2.1+` does not exist (confused it with LGPL-2.1+).
Use GPL 2.0 or later instead.

The checks for modifying the access time via bolt_fs_touch was
previously disabled because it did not work on file-systems
that have the noatime flag set (for very obvious reasons).
Separate these checks out and re-enable them but make their
execution depended on the file-system support for access times,
as reported via statvfs(2).

Access time is not always supported, so support for it is best-effort.
We disable such assertions so bolt can be unit tested on systems that have
"noatime" configured to speed up the filesystem.

This is only a bugfix release so we keep the 0.9 prefix.

Since bolt is also being used for Thunderbolt 4 and USB 4, put
less emphasis on the version, specifically version 3, in the
README.md and the spec file.

Add a .md file section and insist we use spaces there.

Describe how to use the Address Sanitizer in bolt.

Free the `idstr` for the MockDevice, free the `dmi` path variable
for the dmi device path and also free the return value from
umockdev_testbed_get_root_dir.

When the address sanitizer is enabled, skip the recover_guards_fail
test, since we fork() in there and that somehow trips up ASAN.

Use the compiler specific method to detect if the code was compiled
with the address sanitizer enabled. This might be needed to skip
some incompatible tests.

When removing the host device via mock_sysfs_host_remove, destroy
the MockDevice for the host.

The function 'bolt_flags_class_to_string' returns a dynamically
allocated string. This was previously not freed, creating a memory
leak. Its ownership is now correctly transferred to the GVariant
instead.

The memory for custom_id, to identify a custom conversion, is
dynamically allocated, but was not freed in bolt_wire_conv_unref.
Explicitly set it to NULL in the other constructor so we don't
attempt to free invalid memory.

Normally 'path' gets freed in bolt_guard_remove, which is called
from the dispose method of the BoltGuard. There is one exception
though: when the pipe (named fifo) is still active, we do not
unlink the path, since we want to recover that guard later. In
that case the memory for path will be leaked. Add an additional
g_clear_pointer for path in the 'finalize' method to ensure that
path always gets freed.

Ensure that we are not saved already in 'bolt_guard_save'.