It happens sometimes that the first processed total is 1, when this gets
downloaded we have 100% already and then the total changes making this
percentage faulty.
This makes the progress regress sometimes, but I'd say it's better than
a permanent, ficticious number.
Fixes#2428
When we're prompting the user to choose between a list of items but
there's only one item, use a yes/no prompt and different wording.
Closes: #2426
Approved by: matthiasclasen
This avoids a crash or deadlock if the timeout is reached during test
teardown, after the GMainLoop has already been freed. That appears to
be more likely to happen when run as an installed-test.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Resolves: https://github.com/flatpak/flatpak/issues/2422Closes: #2425
Approved by: matthiasclasen
We ran into some issues on flathub where a build with very large
files caused the parallel delta generation to use a lot of paging,
essentially blocking progress. This option can at least allow you
to make progress in that case, although we should ideally have
a better solution.
Closes: #2416
Approved by: alexlarsson
There was a typo here: (gpointer *) pid_data should have been
(gpointer *) &pid_data, so that g_hash_table_iter_next() would make
pid_data a pointer to a PidData struct. Instead, the previous
implementation left pid_data set to NULL, leading to a NULL dereference
and segmentation fault whenever a name fell off the bus while a watched
client existed.
Instead of directly inserting the missing "&", I've used a pattern
that avoids needing the cast, in an attempt to make it more obviously
correct.
Signed-off-by: Simon McVittie <smcv@collabora.com>
Closes: #2417
Approved by: matthiasclasen
Currently the flatpak install command reliably produces this error:
** (flatpak:21996): CRITICAL **: 22:38:52.472: polkit_unix_process_set_property: assertion 'val != -1' failed
This is caused by a recent patch to polkit[1] to address CVE-2018-19788,
which apparently means you can't pass -1 to
polkit_unix_process_new_for_owner() for the uid. So change Flatpak to
pass the actual UID instead.
[1] 2cb40c4d5fCloses: #2414
Approved by: alexlarsson
Capitalize Flatpak in text, add a paragraph about building
that references flatpak-builder, and add a reference to
www.flatpak.org.
Closes: #2410
Approved by: alexlarsson
The first element put in the variant created by
flatpak_dir_system_helper_call_deploy() is the repo path, but this is
being treated as the installation ID in
flatpak_authorize_method_handler(), which results in a seg fault when
dir_get_system() returns NULL and this NULL is passed to
dir_ref_is_installed(). Fix the seg fault by getting the correct element
from the variant.
Closes: #2411
Approved by: matthiasclasen
The uninstall command supports the syntax "flatpak uninstall
org.gnome.Builder stable" as an alternative to "flatpak uninstall
org.gnome.Builder//stable", but this support was broken by the fuzzy
matching (partial ref) feature. Since that syntax only supported full
app IDs not partial ones, assume it's being used if the first argument
is a full app ID and the second argument looks like a branch. This means
we don't support the syntax "flatpak uninstall builder stable" (unless
you're trying to uninstall an app called stable).
Similarly, fix the install command's support for the syntax "flatpak
install flathub org.gnome.Builder stable".
Also, add a unit test for both of the above.
Fixes https://github.com/flatpak/flatpak/issues/2395Closes: #2405
Approved by: matthiasclasen
Make sure we have g_autoptr support for libpolkit types even if we're
compiling against an old version of libpolkit, the same way we do in
system-helper/flatpak-system-helper.c.
Fixes https://github.com/flatpak/flatpak/issues/2401Closes: #2403
Approved by: matthiasclasen
We shouldn't put the unsightly dbus error wrapping in the
UI. The cient already strips the wrapping from the system
helper call, but it can't strip the second level itself.
Closes: #2399
Approved by: mwleeds
Comparing the code in flatpak-utils.c:progress_cb,
we need to set bytes-transferred for the total amount
of data that has been transferred so far. The value
we were setting so far, fetched-delta-part-size, refers
to the size of the objects we already have locally, and
is subtracted from the total, which explains oci progress
running backwards.
Closes: #2392Closes: #2400
Approved by: matthiasclasen
This means we pick up the built services. Also we set
FLATPAK_SYSTEM_HELPER_ON_SESSION so that the system-helper
runs on the session-bus just like in libtest.sh.
Closes: #2397
Approved by: matthiasclasen
GVariant doesn't take it lightly when you are trying to extract
an ay value with an s format. This was causing critical warnings.
Closes: #2393
Approved by: alexlarsson
Currently the update command only checks the default system and user
installations (unless overridden by a command line flag). This commit
makes it check custom installations as well, so they are not left out.
Care is taken so that an empty or broken installation doesn't cause the
whole update operation to fail.
Fixes https://github.com/flatpak/flatpak/issues/2006Closes: #2345
Approved by: matthiasclasen
Some commands depend on the first directory in the list returned by
flatpak_option_context_parse() being the default system directory, which
is true for FLATPAK_BUILTIN_FLAG_STANDARD_DIRS. This commit makes the
system directory first for FLATPAK_BUILTIN_FLAG_ALL_DIRS as well, in
case any code in the future uses that flag in combination with that
assumption.
There is one user of the ALL_DIRS flag that actually wants to keep the
user directory first in the list, the run command. This is so that a
user-installed app will be run even if the same app is installed
system-wide. So this commit modifies the run command to undo this
change after calling flatpak_option_context_parse().
Closes: #2345
Approved by: matthiasclasen
This lets us respect the fancy output setting, and
it lets us do some other things that make it better
integrated.
Closes: #2379
Approved by: alexlarsson
Arrange for progress to be done before we
ask for polkit auth, so we don't end up with
the text prompt mushed into the progress display.
Closes: #2379
Approved by: alexlarsson
Install a polkit text agent that can ask for
authorization when we are on the console.
This brings us closer to best practice for polkit use.
Closes: #2379
Approved by: alexlarsson
There was a bunch of handle methods which forgot to do
this manually, so call the flatpak_dir_set_source_pid()
method right when we create the FlatpakDir object.
Closes: #2384
Approved by: alexlarsson
gtk-doc complains if parameter names don't match between
headers and doc comments, so make them match, for a quieter build.
Closes: #2384
Approved by: alexlarsson
The Deploy method takes an 'update' flag from the caller
and picked the permission to check based on that, which
is not safe. Instead, we can check ourselves if the ref
we are given is already installed (in which case it is
an update, or a reinstall).
After this change, the update flag is ignored by the
system helper.
Closes: #2384
Approved by: alexlarsson
I realized that these messages are a bit confusing, with
keep permissions and implications. We end up asking a
question about one ref, but then use the permission we optained
to operate on other refs. This will get a bit worse with
the new hints, so go back to more generic messages that
don't talk about the details. The ultimate goal here is
less messages, anyway.
Closes: #2384
Approved by: alexlarsson
This is to avoid multiple polkit dialogs, regardless
of transaction ordering.
FlatpakTransaction calculates the 'strongest' op it has,
and passes the hints accordingly. FlatpakInstallation
doesn't pass hints, since it does individual operations.
The system helper uses the hints to determine which PolicyKit
permission to request. Since the policy typically has 'keep'
set, this mean that the following operations in the same
transaction will be able to reuse the permission obtained
for the first one.
Closes: #2384
Approved by: alexlarsson
