exit-on-close for a DBus connection is a completely normal thing. On
a regular GNOME login, gdm retains the X server, but terminates the
session login bus and associated helpers like gnome-settings-dameon,
the a11y tools, etc.
I've seen several downstream reports of confusion as to what these
apparent error messages mean in the system log. It doesn't help
that they're so obtuse.
We're also printing them to stderr, when this is not an error.
The reason this was introduced is presumably some people were confused
as to why their process exited when the system bus did. But the
solution for that I believe is documentation, not printing stuff to
everyone's system log in normal operation.
https://bugzilla.gnome.org/show_bug.cgi?id=742386
The previous commit introduced a possible memory leak in cases
where we get a G_IO_ERROR_CLOSED error. Make sure to always
free an error, if we got one.
https://bugzilla.gnome.org/show_bug.cgi?id=753278
There are real world cases where emitting signals can fail, such
as if the DBus connection closes. Asserting and aborting the process
in these cases is just plain lazy.
Ignore the errors when the connection is closed, and turn the
others into warnings.
https://bugzilla.gnome.org/show_bug.cgi?id=753278
Add a property to GNetworkMonitor indicating if the network
is metered, e.g. subject to limitations set by service providers.
The default value is FALSE
https://bugzilla.gnome.org/show_bug.cgi?id=750282
A signal accumulator can return TRUE to continue signal emission, and
FALSE to stop signal emission. handle-local-options callbacks can return
« return a non-negative option if you have handled your options and
want to exit the process ».
Currently, g_application_handle_local_options_accumulator (the
accumulator for the handle-local-options signal) returns TRUE on
non-negative return value (ie continue signal emission), and returns
FALSE on negative return values (ie when the default option processing
should continue).
This return value seems backward as on >= 0 values, subsequent
handle-local-options callbacks could overwrite the 'exit request' from
the handler, while on < 0 values, the handle-local-options processing
could end up early if several callbacks are listening for this signal.
In particular, the default handler for this signal
(g_application_real_handle_local_options) always returns -1 and will
overwrite >= 0 return values from other handlers.
This commit inverts the check so that signal emission stops early when
one of the handle-local-options callbacks indicates it wants processing
to stop and the process to exit.
https://bugzilla.gnome.org/show_bug.cgi?id=751598
We already have start, stop and is_active methods, but turning it
into a real property is useful for a few reasons:
- it allows us to bind the property to an UI or a setting
- it allows us to get notified when the state changes
- it allows us to instantiate objects directly in the stopped state
https://bugzilla.gnome.org/show_bug.cgi?id=752089
Also add g_app_info_get_fallback_for_type() and
g_app_info_get_recommended_for_type() as proxies for
g_app_info_get_all_for_type(), until gcontenttype support is improved.
This code was out of date with current coding practices.
Nowadays it's common to receive file descriptors over environment
variables from other processes like systemd. The unit files that
control these file descriptors are configurable by sysadmins.
It is not (necessarily) a programmer error when g_socket_details_from_fd()
is called with a file descriptor that is not a socket. It can also
be a system and/or configuration error.
https://bugzilla.gnome.org/show_bug.cgi?id=746339
In 4e7d22e268, deleting the file was moved
after the assertion which checks for the changed event that results from
it being deleted. This is the wrong way around and makes the assertion
fail.
Move the deletion back up before we check the condition. delete_app is
no longer an idle callback so it can be made void. The change
notification might come in when the loop isn't running now, so don't try
to quit if it isn't running. In this case we'll wait for the three
second timeout and the test will still pass.
https://bugzilla.gnome.org/show_bug.cgi?id=751737
When a task is cancelled, we want to move it to the front
of the queue - our sort function does that for us, but there
is no need to resort the entire queue here, we can just
move the one item and be done with it. This uses just-introduced
threadpool api for this purpose.
https://bugzilla.gnome.org/show_bug.cgi?id=751160
Make sure to initialize the notification backend in
g_application_withdraw_notification() the same way as is done in
g_application_send_notification().
This makes it possible for an app to withdraw notifications it has sent
in a previous execution of the application.
https://bugzilla.gnome.org/show_bug.cgi?id=750625
We need to be more careful when we try to assign values to gpointers, so
that means we have to assign the value to the properly-dereference
gpointer, so that the assigned value will be retained after the function
returns. This code will be dropped soon, but it is done for XP
compatibility's sake for 2.44.
Should fix the issue reported in bug 730352 comment #24.
For performance reasons we should always try to send or
receive our messages first and only wait for more space
or data to become available if we get an EAGAIN (and
are in blocking mode).
https://bugzilla.gnome.org/show_bug.cgi?id=751122
This can be handy when you want to change the sense of a toggle
in the UI without rewriting the underlying logic. Currently, this
is just exposed as a construct-only property. We may add a
convenience wrapper or a special !property syntax for this later.
https://bugzilla.gnome.org/show_bug.cgi?id=728489
This is the right thing to do for the "a session is a user-session"
model implemented in dbus 1.9.14, which is described in
<http://lists.freedesktop.org/archives/dbus/2015-January/016522.html>.
It also resembles sd-bus' behaviour, although sd-bus will only try
kdbus and XDG_RUNTIME_DIR/bus, and never runs dbus-launch.
On systems following the more traditional "a session is a login-session"
model, X_R_D/bus won't exist, so it is harmless to check for it before
falling back to X11 autolaunching. Again, this matches the behaviour
of current libdbus and sd-bus versions.
Now that we do this, g_test_dbus_unset() needs to clear XDG_RUNTIME_DIR
as well as everything else.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=747941
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Philip Withnall <philip.withnall@collabora.co.uk>
This only alters what happens if we specifically connect to
"autolaunch:", for instance via "DBUS_SESSION_BUS_ADDRESS=autolaunch:".
We will still potentially try other platform-specific things if
DBUS_SESSION_BUS_ADDRESS is unset. There are currently no other
platform-specific things, so there is no practical difference yet,
but I'm about to add a more-preferred fallback path before autolaunch.
This matches libdbus' behaviour and the D-Bus Specification, in which
the autolaunch: transport specifically means X11 autolaunch
(as implemented by "dbus-launch --autolaunch") on Unix, or a
shared-memory-based protocol on Windows. Other platform-specific
transports or default/fallback modes, including launchd on Mac OS X
and XDG_RUNTIME_DIR/bus on Unix, are not part of "autolaunch:".
It's rather unfortunate that the same name means two different
platform-specific mechanisms, specific to different platforms -
if they were added today I'd call them x11: and windows-shm: or
something - but it's been like this since 2007 so it's too late now.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=747941
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Philip Withnall <philip.withnall@collabora.co.uk>
Currently, applications using g_application_add_main_option_entries()
won't get translated entries in --help output. We need to call
g_option_group_set_translation_domain() with a NULL domain to ensure that the
default application gettext domain (ie the one passed to the
textdomain() call) will be used for the main entries passed by the
application.
If we want to allow more flexibility on which gettext domain should be
used for these entries, new API will be needed.
https://bugzilla.gnome.org/show_bug.cgi?id=750322
Add a new section to the main GSettings documentation which documents
the best practices for integrating GSettings into an autoconf/automake
build system using the GLIB_GSETTINGS macro.
Some of this material was adapted from the migrating-gconf.xml guide.
https://bugzilla.gnome.org/show_bug.cgi?id=741788
* Only check __OBJECT_ATTRIBUTES_DEFINED and __UNICODE_STRING_DEFINED
on MinGW (MSVC doesn't have these)
* MSVC: disable:4005 when including windows.h and ntstatus.h
* Move NTAPI cconv into the parens with the NtQueryKeyFunc
* Fix return values in some functions
https://bugzilla.gnome.org/show_bug.cgi?id=734888
- On first call scan the registry, collect information about URI protocols,
file extensions, applications and handlers, store that as a set of
interconnected structures in several hash tables
- Watch the registry keys, re-scan the registry when any one of them changes.
https://bugzilla.gnome.org/show_bug.cgi?id=666831
Use the newly added g_settings_schema_list_keys() API instead of
g_settings_list_keys() in order to list keys.
Doing this allows the 'list-keys' command to work without creating a
GSettings object, which is more efficient. It also means that we don't
have to provide a (meaningless and ignored) path when listing keys on
relocatable schemas.
While we're at it, update the 'range' command not to require creation of
a GSettings object, in a similar way.
https://bugzilla.gnome.org/show_bug.cgi?id=740308
The list of keys in a GSettings object depends entirely on the schema,
so it makes sense to expose this API there.
Move the implementation out of gsettings.c and into gsettingsschema.c,
replacing the earlier with a simple call to the new location.
We don't do the same for children because the children can change.
https://bugzilla.gnome.org/show_bug.cgi?id=740308
Cancellation of GPollFileMonitor is now handled correctly (in the sense
that no further signals will follow) but let's be extra paranoid and
disconnect our handler anyway, for good measure.
https://bugzilla.gnome.org/show_bug.cgi?id=739424
GPollFileMonitor emits CHANGES_DONE_HINT after CHANGED signals, but it
doesn't check to ensure that the file monitor wasn't cancelled before it
does that.
If the original signal caused the monitor to be unreffed, cancelled and
destroyed, we would still end up emitting an extra signal on it.
Avoid that by checking first for cancellation.
https://bugzilla.gnome.org/show_bug.cgi?id=739424
Removed all mentions of GLib file name encoding referring to
the environment strings. The env var content has no defined relation
to GLib's notion of filename encoding, or any encoding whatsoever.
It would be wrong to pass all UTF-8 strings through
g_filename_from_utf8() in order to put them into the environment,
for one thing.
https://bugzilla.gnome.org/show_bug.cgi?id=738185
DBus has recently introduced new message flag
DBUS_HEADER_FLAG_ALLOW_INTERACTIVE_AUTHORIZATION, which tells that
caller is willing to wait for unspecified amount of time for the call
to return, as the service may perform interactive authorization (e.g.
using polkit).
https://bugzilla.gnome.org/show_bug.cgi?id=739616
In order to maintain a logical stream of events, we need to make sure we
flush and queued change notifications before responding to any requests
for information from clients.
If we don't do this, it's possible that we emit an 'add' event that was
queued at the time of a 'DescribeAll' call _after_ the reply to that
call (which already contained the description of the new action).
In practice, this is not only logically incorrect, but it can also cause
problems. If a change to action 'state' or 'enabled' occurs after the
DescribeAll but before the signal has been dispatched, it will be
ignored because an 'add' signal is already pending. When that add
signal is sent, it will contain the correct data, but the receiver will
ignore it because it already saw the action in the DescribeAll reply.
https://bugzilla.gnome.org/show_bug.cgi?id=749693
This is a hack for GLocalFileInfo to correctly get icons for directories.
Without this change content type for any W32 directory is NULL
(because there's no registry entry for "inode/directory" by default,
and in any way there's no file extension that means "directory" to put there),
and GLocalFileInfo uses content type to grab icons.
https://bugzilla.gnome.org/show_bug.cgi?id=748727
The code here was returning gtk-directory and similar names as
fallback, with a comment claiming that these are 'builtin gtk'.
But they aren't, anymore, so just return the standard names.
This code used to look at the SCM_CREDENTIALS and ignore every message
not from uid 0. However, when user namespaces are in use this does not
work, as if uid 0 is not mapped you get overflowuid instead. Right now
this means we ignore all messages in such user namespaces and glib
apps hang on startup.
We can't look at pids either, as pid 0 is returned for processes
outside your pid namespace.
Instead the correct approach is to look at the sending sockaddr and
if the port id (nl_pid) is zero, then its from the kernel.
Source:
http://lists.linuxfoundation.org/pipermail/containers/2015-May/036032.htmlhttps://bugzilla.gnome.org/show_bug.cgi?id=750203
Instead of just dropping address types that we're not specifically
handling we return a GNativeSocketAddress which is just a dummy
container for the stuct sockaddr.
https://bugzilla.gnome.org/show_bug.cgi?id=750203
GListStore requires that item-type be derived from GObject, so specify
that the type of the item parameters is GObject so the functions can be
used via gobject-introspection.
Add a scope parameter for the callback used during insert_sorted.
Previously, we waited up to 0.5s, but that can fail on slow
architectures like ARM; now we wait up to 60s in 0.1s increments.
Patch originally by Simon McVittie <simon.mcvittie@collabora.co.uk>,
modified by Iain Lane to be called earlier, to catch all testcases in a
particular test.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=724113
Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Acked-by: Matthias Clasen <mclasen@redhat.com>
We previously waited 0.25s, which should be enough even on slow machines,
but you never know; but we also now wait in 0.1s increments, so this test
should actually be faster now.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=724113
Acked-by: Matthias Clasen <mclasen@redhat.com>
I searched all files that mention g_test_run, and replaced most
g_print() calls. This avoids interfering with TAP. Exceptions:
* gio/tests/network-monitor: a manual mode that is run by
"./network-monitor --watch" is unaffected
* glib/gtester.c: not a test
* glib/gtestutils.c: not a test
* glib/tests/logging.c: specifically exercising g_print()
* glib/tests/markup-parse.c: a manual mode that is run by
"./markup-parse --cdata-as-text" is unaffected
* glib/tests/testing.c: specifically exercising capture of stdout
in subprocesses
* glib/tests/utils.c: captures a subprocess's stdout
* glib/tests/testglib.c: exercises an assertion failure in g_print()
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=725981
Reviewed-by: Colin Walters <walters@verbum.org>
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
This stops it from interfering with structured stdout such as TAP.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=725981
Reviewed-by: Colin Walters <walters@verbum.org>
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
This avoids any possibility of interfering with test syntax (such as
TAP) on stdout. TAP specifically does not parse stderr.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=725981
Reviewed-by: Colin Walters <walters@verbum.org>
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
When running the nonce-tcp and tcp-anonymous tests in one run
of gdbus-peer, or running one of them twice via command-line options
"-p /gdbus/tcp-anonymous -p /gdbus/tcp-anonymous", the one run second
would sometimes fail to connect with ECONNRESET.
Adding more debug messages revealed that in the successful case,
g_main_loop_run() was executed in the server thread first:
# tcp-anonymous: server thread: listening on tcp:host=localhost,port=53517
# tcp-anonymous: server thread: starting server...
# tcp-anonymous: server thread: creating main loop...
# tcp-anonymous: server thread: running main loop...
# tcp-anonymous: main thread: trying tcp:host=localhost,port=53517...
# tcp-anonymous: main thread: waiting for server thread...
but in the failing case, the main thread attempted to connect
before the call to g_main_loop_run() in the server thread:
# tcp-anonymous: server thread: listening on tcp:host=localhost,port=40659
# tcp-anonymous: server thread: starting server...
# tcp-anonymous: server thread: creating main loop...
# tcp-anonymous: main thread: trying tcp:host=localhost,port=40659...
# tcp-anonymous: server thread: running main loop...
(The log message "creating main loop" was immediately before
create_service_loop(), and "running main loop" was immediately
before g_main_loop_run().)
To ensure that the GDBusServer has a chance to start accepting
connections before the main thread tries to connect to it, do not
tell the main thread about the service_loop immediately, but instead
defer it to an idle.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=749079
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Philip Withnall <philip.withnall@collabora.co.uk>
This test originally did not connect to the bus, which meant it was
omitted from commits like 415a8d81 that made sure none of GLib tests
rely on the presence of an existing session bus. (In particular,
Debian autobuilders don't have a session bus.)
When test_double_array() was added, environments like the Debian
autobuilders didn't catch the fact that this test relied on having a
session bus, because it is often skipped in minimal environments
due to its libdbus-1 dependency.
We don't actually need to connect to a dbus-daemon here: it's enough
to convert the message from GVariant to D-Bus serialization, and
back into an in-memory representation through libdbus. That's what
check_serialization() does, and I've verified that when I re-introduce
bug #732754 by reverting commits 627b49b and 2268628 locally, this
test still fails.
Bug: https://bugzilla.gnome.org/show_bug.cgi?id=744895
Signed-off-by: Simon McVittie <simon.mcvittie@collabora.co.uk>
Reviewed-by: Colin Walters <walters@verbum.org>
The changed variable was previously uninitialised in the path where the
rate limit was actually changed. This could result in the
GObject::notify signal not getting emitted.
Spotted by Coverity.
CID: #1296516https://bugzilla.gnome.org/show_bug.cgi?id=748834
The third parameter of the thumnail_verify() function had been updated to
const GLocalFileStat, so update the thumbnail-verification test likewise
so that the test works properly on all supported platforms.
https://bugzilla.gnome.org/show_bug.cgi?id=711547
Always run the full algorithm for a given mime type before considering
fallback types.
This includes considering installed applications capable of handling a
particular mimetype, even if such an app is not explicitly marked as
default, and there is a default app for a less-specific type.
Specifically, this often helps with cases of installing apps that can
handle a particular subtype of text/plain. We want to take those apps
in preference to a generic text editor, even if that editor is listed as
the default for text/plain and there is no default listed for the more
specific type.
Because of the more holistic approach taken by the algorithm, it is now
more complicated, but it also means that we can do more work while
holding the lock. In turn, that lets us avoid duplicating some strings,
which is nice.
https://bugzilla.gnome.org/show_bug.cgi?id=744282
If g_dbus_message_to_blob() fails at all, it will leak its mbuf. Spotted
by running the gdbus-serialization test under Valgrind — so there is a
justification for leak-free tests after all!
In path_rule_matches(), the given paths may be of 0-length. Do not
access memory before the array in those case. This is for example
triggered by:
test_match_rule (con, G_DBUS_SIGNAL_FLAGS_MATCH_ARG0_PATH, "/", "", FALSE);
in test_connection_signal_match_rules().
This bug was found thanks to GCC AddressSanitizer.
https://bugzilla.gnome.org/show_bug.cgi?id=745745
Passing an nonsense string for the --dest argument can lead
to a segfault of gdbus. Thats not nice, so use our existing
validation function for bus names here.
https://bugzilla.gnome.org/show_bug.cgi?id=747541
Make sure error handling on repeated <summary> and <description> is
being done properly, not resulting in glib-compile-schemas throwing a
critical.
https://bugzilla.gnome.org/show_bug.cgi?id=747542
Fix a couple of issues in error handling in glib-compile-schemas.
The first problem is that, in case of repeated <summary> or
<description> tags we were still allocating a GString which was never
being freed (due to the throwing of the error resulting in immediate
termination of the parse).
The second problem is that if the repeated <summary> tag also had
attributes, we would attempt to set the GError twice.
https://bugzilla.gnome.org/show_bug.cgi?id=747542
Bug 747209 introduced an error when multiple <summary> or <description>
tags are found for a single key in a GSettings schema. This check
should have been present from the start, but it was left out because the
schema compiler doesn't include these items in the cache file. Even
still -- part of the schema compiler's job is validation, and it should
be enforcing proper syntax here.
Repeated <summary> and <description> tags are a semi-common problem when
intltool has been misconfigured in the build system of a package, but
it's possible to imagine mistakes being made by hand as well.
The idea is that these problems would be caught during the build of a
package and maintainers would be forced to fix their build systems.
An unintended side-effect of this change, however, is that the schema
compiler started ignoring already-installed schemas that contained these
problems, when rebuilding the cache. This means that the installation
of _any_ application would cause the regeneration of the entire cache,
with these already-installed applications being excluded. Without the
schema in the cache, the application would crash on next startup.
The validation check in the gsettings m4 macro passes --strict to the
compiler, which is not used when rebuilding the cache after
installation. Pass this flag down into the parser and only throw the
error in case --strict was given. This will result in the (desired)
build failure without also causing already-installed apps to stop
functioning.
This means that we will not get even a warning about the invalid schema
file in the already-installed case, but that's fine. There is no sense
spamming the user with these messages when they are already quite fatal
for the developer at build time.
https://bugzilla.gnome.org/show_bug.cgi?id=747472
Commit f10b655 removed the inclusion of gasyncresult.h from gdbusproxy.c,
but gdbusproxy.c uses g_async_result_get_source_object(), which caused a
build warning/error. Fix that.
The gdbus GTask port introduced a deadlock because some code had been
using g_simple_async_result_complete_in_idle() to ensure that the
callback didn't run until after a mutex was unlocked, but in the gtask
version, the callback was being run immediately. Fix it to drop the
mutex before calling g_task_return*(). Also, tweak
tests/gdbus-connection to test this.
https://bugzilla.gnome.org/show_bug.cgi?id=747349
Add support for copying session data between client connections.
This is needed for implementing FTP over SSL. Most servers use a separate
session for each control connection and enforce sharing of each control
connection's session between the related data connection.
Copying session data between two connections is needed for two reasons:
1) The data connection runs on a separate port and so has a different
server_identity which means it would not normally share the session with
the control connection using the session caching currently implemented.
2) It is typical to have multiple control connections, each of which
uses a different session with the same server_identity, so only one of
these sessions gets stored in the cache. If a data connection is opened,
(ignoring the port issue) it may try and reuse the wrong control
connection's session, and fail.
This operation is conceptually the same as OpenSSL's SSL_copy_session_id
operation.
https://bugzilla.gnome.org/show_bug.cgi?id=745255
This allows the caller to know when a socket has been bound so that
it can for instance set the SO_SENDBUF and SO_RECVBUF socket options
before listen is called
https://bugzilla.gnome.org/show_bug.cgi?id=738207
GTask used a 10-thread thread pool for g_task_run_in_thread() /
g_task_run_in_thread_sync(), but this ran into problems when task
threads blocked waiting for another g_task_run_in_thread_sync()
operation to complete. Previously there was a workaround for this, by
bumping up the thread limit when that case was detected, but deadlocks
could still happen if there were non-GTask threads involved. (Eg, task
A sends a message to thread X and waits for a response, but thread X
needs to complete task B in a thread before returning the response to
task A.)
So, allow GTask's thread pool to be expanded dynamically, by watching
it from the glib worker thread, and growing it (at an
exponentially-decreasing rate) if too much time passes without any
tasks completing. This should solve the deadlocking problems without
causing sudden breakage in apps that assume they can queue huge
numbers of tasks at once without consequences.
https://bugzilla.gnome.org/show_bug.cgi?id=687223
This schema compiler was completely ignoring <summary> and
<description> tags. Unfortunately, there are modules out there
who merge translations for these back in, with xml:lang. And
this is giving dconf-editor a hard time. Since this is not
how translations of schemas are meant to be done, just
reject such schema files.
Also add tests exercising the new error handling.
https://bugzilla.gnome.org/show_bug.cgi?id=747209
glib-compile-resources was guessing a filename ending
in .c when generating sources, but did not do the same
for headers. Fix it so it generates a .h file when
guessing the filename for headers.
https://bugzilla.gnome.org/show_bug.cgi?id=746753
The hash table stores the list of unmatched IN_MOVE_FROM events, but we
were removing entries from it when popping IN_MOVE_TO events.
Fix that up to correct a crash in nautilus due to the assertion failure
below.
https://bugzilla.gnome.org/show_bug.cgi?id=746749
Due to a typo, a rename reported via a pair of delete/create events (due
to the watcher not giving the flag for moves to be paired) was
accidentally reported as being created with the old name instead of the
new name.
Fix that.
We declare the typedefs for GListModel and GListStore in giotypes.h, as
a matter of convention. This is not actually required, since the
typedef is emitted as part of the G_DECLARE_* macros.
The giotypes.h approach is only used to avoid cyclic dependencies
between headers, which is not a problem in this case.
Type redefinition is a C11 feature, and although it was around in some
compilers before then, gcc 4.2.1 (from 2007) is apparently still in wide
use, being the default compiler for OpenBSD.
Eventually, we will probably hit a case where we actually need to
redefine a type, but since we're not there yet, let's back off a bit.
This WIP patch moves the Windows Directory Monitoring code to the new
GLocalFileMonitor mechanism, and adds file monitoring in the process.
Progress from previous patch:
-File renames are now properly supported, but G_FILE_MONITOR_EVENT_MOVED_IN
and G_FILE_MONITOR_EVENT_MOVED_OUT needs to be investigated, as
ReadDirectoryChangesW() seems to send FILE_ACTION_REMOVED when a file is
moved out of a directory.
-Events are handled for both the long and short (8.3) variants of the
filenames, and files monitored will report changes when it is changed
via its short or long filenames.
Things to be done:
-Perhaps find out about attribute changes in files in a monitored
directory; if a file is monitored, attribute changes are correctly
handled.
-Investigate on G_FILE_MONITOR_EVENT_MOVED_OUT,
G_FILE_MONITOR_EVENT_MOVED_IN, G_FILE_MONITOR_EVENT_PRE_UNMOUNT,
G_FILE_MONITOR_EVENT_UNMOUNTED.
-Investigate on the "boredom" algoritm, and see how we can do it on
Windows.
https://bugzilla.gnome.org/show_bug.cgi?id=730116
Completely rewrite the FAM file monitor. Major changes:
- now runs in the worker thread
- dispatches events in a threadsafe way via GFileMonitorSource
- uses unix fd source instead of a GIOChannel
- is now simple enough to fit into one short file
This is the bare minimal effort. This seems not to crash immediately,
but it definitely needs some better testing.
The backend is not in good shape. It could use some serious work.
Use the "interesting" value from g_file_monitor_source_handle_event() to
decide if we're currently being flooded by a stream of boring events.
The main case here is when one or more files is being written to and the
change events are all being rate-limited in the GFileMonitor frontends.
In that case, we become "bored" with the event stream and add a backoff
timeout. In the case that it is exactly one large file being written
(which is the common case) then leaving the event in the queue also lets
the kernel perform merging on it, so when we wake up, we will only see
the one event. Even in the case that the kernel is unable to perform
merging, the context switch overhead will be vastly reduced.
In testing, this cuts down on the number of wake ups during a large file
copy, by a couple orders of magnitude (ie: less than 1% of the number of
wake ups).
Return an "interesting" boolean from the event handler function on
GFileMonitorSource.
An event was "interesting" if it will result in a signal actually being
dispatched to the user. It is "uninteresting" if it only hit an
already-dirty rate limiter.
We will use this information to do some backing off in the backends when
faced with a flood of uninteresting events.
We generally assume that an IN_CREATE event is the start of a series of
events in which another process is doing this:
fd = creat (...) -> IN_CREATE
write (fd, ..) -> IN_MODIFY
write (fd, ..) -> IN_MODIFY
close (fd) -> IN_CLOSE_WRITE
and as such, we use the CHANGES_DONE_HINT event after CREATED in order
to show when this sequence of events has completed (ie: when we receive
IN_CLOSE_WRITE when the user closes the file).
Renaming a file into place is handled by IN_MOVED_FROM so we don't have
to worry about that.
There are many other cases, however, where a new file 'appears' in a
directory in its completed form already, and the kernel reports
IN_CREATE. Examples include mkdir, mknod, and the creation of
hardlinks. In these cases, there is no corresponding IN_CLOSE_WRITE
event and the CHANGES_DONE_HINT will have to be emitted by an arbitrary
timeout.
Try to detect some of these cases and report CHANGES_DONE_HINT
immediately.
This is not perfect. There are some cases that will not be reliably
detected. An example is if the user makes a hardlink and then
immediately deletes the original (before we can stat the new file).
Another example is if the user creates a file with O_TMPFILE. In both
of these cases, CHANGES_DONE_HINT will still eventually be delivered via
the timeout.
Remove all event merging and dispatch logic from GFileMonitor. The only
implementation of GFileMonitor outside of glib is in gvfs and it already
does these things properly.
Get rid of GLocalDirectoryMonitor. We will use a single class,
GLocalFileMonitor, for both directory and file monitoring. This will
prevent every single backend from having to create two objects
separately (eg: ginotifydirectorymonitor.c and ginotifyfilemonitor.c).
Introduce GFileMonitorSource as a thread-safe cross-context dispatch
mechanism. Put it in GLocalFileMonitor. All backends will be expected
to dispatch via the source and not touch the GFileMonitor object at all
from the worker thread.
Remove all construct properties from GLocalFileMonitor and remove the
"context" construct property from GFileMonitor. All backends must now
get the information about what file to monitor from the ->start() call
which is mandatory to implement.
Remove the implementation of rate limiting in GFileMonitor and add an
implementation in GLocalFileMonitor. gvfs never did anything with this
anyway, but if it wanted to, it would have to implement it for itself.
This was done in order to get the rate_limit field into the
GFileMonitorSource so that it could be safely accessed from the worker
thread.
Expose g_local_file_is_remote() internally for NFS detection.
With the "is_remote" functionality exposed, we can now move all
functions for creating local file monitors to a proper location in
glocalfilemonitor.c
Port the inotify backend to adjust to the changes above. None of the
other backends are ported yet. Those will come in future commits.
Remove the hardwired 1 second event queue logic from inotify-kernel and
replace it with something vastly less complicated.
Events are now reported as soon as is possible instead of after a
delay.
We still must delay IN_MOVED_FROM events in order to look for the
matching IN_MOVED_TO events, and since we want to report events in order
this means that events behind those events can also be delayed. We
limit ourselves, however:
- no more than 100 events can be delayed at a time
- no event can be delayed by more than 10ms
https://bugzilla.gnome.org/show_bug.cgi?id=627285
Add a new internal constructor for GLocalFile (which itself is private).
This new constructor allows creating a GLocalFile from a dirname and a
basename, assuming that the dirname is already in canonical form and the
basename is a regular basename.
This will be used for creating GLocalFile instances from the file
monitoring code (for signal emissions).
For all of the effort spent ensuring that this algorithm would be
correctly threadsafe, I messed up the order of operations within a
single thread when porting to the new approach.
Fix that up.
Also: fix some overzealous asserting in the testcases. Since shutdown
is now lazy, we can never surely say !is_running at any particular point
in time.
This can be used to query whether the task has completed, in the sense
that it has had a result set on it, and has already – or will soon –
invoke its callback function.
Notifications for this property are emitted immediately after the task’s
main callback, in the same main context as that callback. This allows
for multiple bits of code to listen for completion of the GTask, which
opens the door for blocking on cancellation of the GTask and improved
handling of ‘pending’ behaviour.
https://bugzilla.gnome.org/show_bug.cgi?id=743636
g_socket_client_add_application_proxy() claimed "When the indicated
proxy protocol is returned by the #GProxyResolver, #GSocketClient will
consider this protocol as supported but will not try to find a #GProxy
instance to handle handshaking." But in fact, it did the checks in the
wrong order, so GProxy proxies ended up overriding
application-specified ones. Fix that.
Also, simplify the code a bit by making use of g_hash_table_add() and
g_hash_table_contains().
https://bugzilla.gnome.org/show_bug.cgi?id=733876
Currently, the Windows code use Winsock2-specific APIs to try to emulate
calls such as inet_pton(), inet_ntop() and if_nametoindex(), which may not
do the job all the time. On Vista and later, Winsock2 does provide a
proper implementation for these functions, so we can use them if they exist
on the system, by querying for them during g_networking_init(). Otherwise,
we continue to use the original code path for these, in the case of XP and
Server 2003.
This enables many of the network-address tests to pass on Windows as a
result, when the native Winsock2 implementations can be used.
https://bugzilla.gnome.org/show_bug.cgi?id=730352
Check the IPv6 addresses on Windows, as we need to reject those that have
brackets/ports around them as valid addresses in this form would have been
accepted during the call to g_inet_address_new_from_string ().
https://bugzilla.gnome.org/show_bug.cgi?id=730352
There was a theoretical deadlock between the worker trying to emit a
signal at the same time as we were waiting for it to shutdown the
notification (while holding the lock).
The deadlock was particularly annoying because we didn't really need to
wait for the shutdown and because it wasn't possible to signals to
arrive while waiting for a start. Attempting to deal with start and
stop in an asymmetric way could have lead to other weird situations,
however.
Drop the lock while waiting for the worker thread to start. This means
that we face the possibility of multiple waiters on the cond at the same
time, so we need to make more of a state machine.
https://bugzilla.gnome.org/show_bug.cgi?id=742599
GUnixMountMonitor was not threadsafe before. It was a global singleton
which emitted signals in the first thread that happened to construct it.
Move it to a per-context singleton model where each GMainContext gets
its own GUnixMountMonitor. Monitor for the changes from the GLib worker
thread and dispatch the results to each context with an active monitor.
https://bugzilla.gnome.org/show_bug.cgi?id=742599
Deprecate g_unix_mount_monitor_set_rate_limit() and turn it into a
no-op.
This function doesn't behave as advertised. It only controls rate
limiting for filesystem-based monitors. It has no impact over reporting
mount changes on Linux, for example, because those are based on polling
for changes in /proc (which doesn't use filesystem monitors). It also
has no impact on Mac OS because a library interface is used there.
This was added in https://bugzilla.gnome.org/show_bug.cgi?id=521946 in
order to be used by HAL, which is effectively dead. udisks no longer
uses this code at all.
https://bugzilla.gnome.org/show_bug.cgi?id=742599
This is a singleton, but we have a function called _new() to get it.
What's worse is that the documentation makes no mention of this, and
actually specifically says that a new monitor will be created each time.
https://bugzilla.gnome.org/show_bug.cgi?id=742599
Add a new internal helper called GContextSpecificGroup.
This is a mechanism for helping to maintain a group of context-specific
monitor objects (eg: GAppInfoMonitor, GUnixMountMonitor).
https://bugzilla.gnome.org/show_bug.cgi?id=742599
If someone explicitly calls g_application_quit() then don't attempt to
drain the mainloop of remaining sources.
This allows applications with 100% CPU utilisation to quit reliably.
https://bugzilla.gnome.org/show_bug.cgi?id=744876
We install win32-software/autorun.exe (as test data for mime scanning)
only on UNIX builds, so don't attempt to chmod it on 'make install'
unless we're on UNIX.
I love Emacs keyboard macros, used them to convert the list of
defines cleverly into a list of tests, then iterated and filled in
the necessary constructor arguments.
After ::shutdown, run the mainloop until all pending activity is
handled, before returning from run().
Among other things, this gives a chance for destroyed windows to be
properly withdrawn from the windowing system.
https://bugzilla.gnome.org/show_bug.cgi?id=744876
This is *significantly* more pleasant to use from C (while handling
errors and memory cleanup).
While we're here, change some ugly, leaky code in
tests/desktop-app-info.c to use it, in addition to a test case
in tests/file.c.
https://bugzilla.gnome.org/show_bug.cgi?id=661554
g_application_bind_busy_property() had the restriction that only one
property can be bound per object, so that NULL could be used to unbind.
Even though this is enough for most uses, it is a weird API.
Lift that restriction and add an explicit unbind function.
https://bugzilla.gnome.org/show_bug.cgi?id=744565
Add an implementation of non-thread-emulated async close of a GIOStream
if either of the underlying stream objects support it.
This prevents us from calling close() functions from another thread on
an object that may not be expecting that. It also allows us to skip the
thread entirely in case our objects support a pure async close.
https://bugzilla.gnome.org/show_bug.cgi?id=741630
Add an internal helper to find out if close_async() is implemented via
threads using the default implementation in the base class.
We will use this to decide if we should do a 'pure async' close of a
GIOStream or not.
https://bugzilla.gnome.org/show_bug.cgi?id=741630
delayed_close_free() calls g_object_unref() on a variable that is
expected to possibly contain NULL (as indicated by the fact that the
NULL case is handled in my_slow_close_output_stream_close_async()).
This is dead code right now (due to a bug in GDBus), which is why it
isn't actually causing a failure. It should still be fixed, however.
https://bugzilla.gnome.org/show_bug.cgi?id=743990
GApplication set the prgname to the application's id when it was running
in service mode. This broke with the addition of new --app-id option,
because g_set_prgname() was called before parsing the options. Calling
it after option parsing doesn't work, because GOptionContext sets
prgname to argv[0] unconditionally.
Instead of changing the semantics of GOptionContext, simply remove this
functionality from GApplication. It is very unusual to have the prgname
set to the app id instead of the binary's name and might confuse people
when looking at logs etc.
When overriding local_command_line() from a subclass,
g_option_context_parse() might never be invokded. Thus, continue setting
the prgname to argv[0] in GApplication.
https://bugzilla.gnome.org/show_bug.cgi?id=743933
Balancing g_application_{un,}mark_busy() is non-trivial in some cases.
Make it a bit more convenient by allowing to bind multiple boolean
properties (from different objects) to the busy state. As long as these
properties are true, the application is marked as busy.
https://bugzilla.gnome.org/show_bug.cgi?id=744565
Add g_list_store_insert_sorted() which takes a GCompareDataFunc to
decide where to insert. This ends up being a very trivial function,
thanks to GSequence.
https://bugzilla.gnome.org/show_bug.cgi?id=743927
Make sure that we only match the _get_type() function name by
restricting the regexp to matching [A-Za-z0-9_]. We were matching on .*
before which means that if we had two _get_type() functions appearing on
a single line then we would get everything in between them included (by
the default rule of '*' being greedy).
This affected G_DECLARE_*_TYPE which puts several uses of _get_type()
into a single line.
GListModel is an interface that represents a dynamic list of GObjects.
Also add GListStore, a simple implementation of GListModel that stores
all objects in memory, using a GSequence.
https://bugzilla.gnome.org/show_bug.cgi?id=729351
Currently the only way to set a state hint on an action is through a
subclass; add a g_simple_action_set_state_hint() method so that this
becomes easier for clients that already use GSimpleAction.
https://bugzilla.gnome.org/show_bug.cgi?id=743521
When implementing blocking operations on top of
nonblocking sockets we should always first try to
perform the operation and then if needed handle
EAGAIN and wait with g_socket_wait_condition.
This is an optimization since we avoid calling
wait condition when it is not needed, but most
importantly this fixes hangs on win32 where some
events (in particular FD_WRITE) are only emitted
after the operation fails with EWOULDBLOCK.
https://bugzilla.gnome.org/show_bug.cgi?id=732439https://bugzilla.gnome.org/show_bug.cgi?id=741707
Add a unit test that checks g_socket_new_from_fd by creating
a gsocket, obtaining its fd, duplicating the fd and then creating
a gsocket from the new fd. This shows a hang on win32 since the
gsocket created from the fd never receives the FD_WRITE event
because we wait for the condition without first trying to write
and windows signals the condition only after a EWOULDBLOCK error.
https://bugzilla.gnome.org/show_bug.cgi?id=741707
We were asking for properties on NM's dbus interface, but if NM is not
running then there won't be any. Check if the name has an owner before
doing anything to it.
https://bugzilla.gnome.org/show_bug.cgi?id=741653
