Currently, the introspection data for GLib and its sub-libraries is
generated by gobject-introspection, to avoid the cyclic dependency
between the two projects.
Since gobject-introspection is generally available on installed systems,
we can check for its presence, and generate the introspection data
directly from GLib.
This does introduce a cyclic dependency, which is why it's possible to
build GLib without introspection, then build gobject-introspection, and
finally rebuild GLib.
By having introspection data available during the GLib build, we can do
things like generating documentation; validating newly added API; and
close the loop between adding new API and it becoming available to non-C
consumers of the C ABI (i.e. language bindings).
If `update-desktop-database` or `update-mime-database` are not
installed, there are a few `GAppInfo` methods for modifying file
associations which will print a `g_warning()` when called. Something
like:
```
GLib-GIO-FATAL-WARNING: Failed to execute child process ‘update-desktop-database’ (No such file or directory)
```
(Example: https://gitlab.gnome.org/zamaudio/glib/-/jobs/3190053)
This will cause the appinfo/associations test to fail, as warnings are
fatal.
If that’s going to happen, skip the test. We can’t hard-depend on these
tools as they are external to GLib and only needed for a few operations.
Instead we have a soft runtime dependency on them; that should be
reflected in the tests.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
Helps: #3148
Commit f6c40b1d fixed libelf detection on FreeBSD (where the library has
no pkg-config file and needs to be found via `find_library()`), but
broke `-Dlibelf=disabled` on Linux, as `get_option('libelf')` was no
longer checked.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #3120
Add a default modification timestamp flag to g_file_copy so that it
doesn't copy the modification time from the source file as it does by
default. Similarly to G_FILE_COPY_TARGET_DEFAULT_PERMS, this flag
overrides the G_FILE_COPY_ALL_METADATA flag.
Fixes: https://gitlab.gnome.org/GNOME/glib/-/issues/3140
There are some flavors of MIPS ABIs, such like r6 vs legacy,
nan2008 vs nan1985 etc.
The `cc -r` may not produce the correct elf binaries.
So let's skip this test for MIPS.
Many toolchain did not change the definition of NULL to avoid introducing
breaking changes in existing codebases. For example, on Windows NULL is
0 (int) regardless of the C++ standard in use.
Fixes the following warnings on CLang when compiling for Windows:
../glib/glib/tests/cxx.cpp:539:34: warning: missing sentinel in function call [-Wsentinel]
g_test_init (&argc, &argv, NULL);
^
, nullptr
../glib/glib/gtestutils.h:298:9: note: function has been explicitly marked sentinel here
void g_test_init (int *argc,
^
../glib/gio/tests/cxx.cpp:62:34: warning: missing sentinel in function call [-Wsentinel]
g_test_init (&argc, &argv, NULL);
^
, nullptr
../glib/glib/gtestutils.h:298:9: note: function has been explicitly marked sentinel here
void g_test_init (int *argc,
^
Fixes the following warnings on CLang:
../glib/gio/gregistrysettingsbackend.c:1503:32: warning: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Wsingle-bit-bitfield-constant-conversion]
child_item->readable = TRUE;
^ ~~~~
../glib/gio/gregistrysettingsbackend.c:1567:28: warning: implicit truncation from 'int' to a one-bit wide bit-field changes value from 1 to -1 [-Wsingle-bit-bitfield-constant-conversion]
child_item->readable = TRUE;
^ ~~~~
The C standard does not specify whether the underlying type of an enum
is signed or unsigned, and until C23 there was no way to control this
explicitly. GCC appears to make enums unsigned unless there is a
negative value among cases of the enum, in which case it becomes signed.
MSCV appears to make enums signed by default.
A bitfield of an enum type (which is not specificied in the C standard
either) behaves as if it was an instance of a numeric type with a
reduced value range. Specifically, a 'signed int val : 2;' bitfield will
have the possible values of -2, -1, 0, and 1, with the usual wraparound
behavior for the values that don't fit (although this too is
implementation-defined).
This causes the following issue, if we have:
typedef enum
{
G_ZERO,
G_ONE,
G_TWO
} GFoo;
struct _GBar
{
GFoo foo : 2;
};
and then assign bar.foo = G_TWO and read it back, it will have the
expected value of 2 (aka G_TWO) on GCC, but a value of -2 (not matching
any of the enum variants) on MSVC.
There does not seem to be any way to influence signedness of an enum
prior to C23, nor is there a 'unsigned GFoo foo : 2;' syntax. The only
remaining options seems to be never using enums in bitfields, which is
what this change implements.
This corresponds to https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/6467
in GTK.
Signed-off-by: Sergey Bugaev <bugaevc@gmail.com>
This triggered a warning from the CHERI compiler since the struct contains
a `void *` but `__attribute__((aligned(8))` reduced alignment to less than
the `void *` alignment (which is 16 for Arm Morello).
Helps: https://gitlab.gnome.org/GNOME/glib/-/issues/2842
This should not result in any functional changes, but will eventually
allow glib to be functional on CHERI-enabled systems such as Morello.
Helps: https://gitlab.gnome.org/GNOME/glib/-/issues/2842
It’s not actually needed on any platform, and causes compilation
problems on platforms where it’s not available.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #3111
The Comment field provides a user-visible description of the app,
which usually contains generic words ("and", "or", "not", "is", ...)
that add noise when used for search.
It made some sense to match against the field as a fallback for
Keywords, before that key was well established. However that key
has been around for years now, so hopefully every app where additional
terms are helpful uses it by now.
With that, the downside of added noise outweighs the benefit, so
it's time to stop matching on comments.
When setting the file time using utimensat, don't ignore
microseconds for access/modify times. By doing that, they're preserved
when using g_file_info_set_modification_date_time and then setting the file's
attributes from it.
Fixes#3116
Commit 9e2ad88455 improved app search results by allowing to differentiate
their match_type: prefix match or substring match; while giving more priority
to prefix matches over substring matches, but only when they are in the same
match_category[1].
This was a step forward but, as outlined in #3082, still not enough to get
most relevant results first to the user, because apparently (and for the
specific case of desktop app searching) a prefix match in a lower category
is more relevant to the user than a substring match in a higher category.
So that's what this commit implements, i.e. it makes sure prefix matches
are still preferred over substring matches but this time not only when
in the same category but also across different categories.
[1] Match category is the Desktop file key where the match happened.
They are shown below from top to lesser priority.
DESKTOP_KEY_Name
DESKTOP_KEY_Exec
DESKTOP_KEY_Keywords
DESKTOP_KEY_GenericName
DESKTOP_KEY_X_GNOME_FullName
DESKTOP_KEY_Comment
Fixes#3082
· Add a usage output that is printed when called with no argument
or with '--help' argument. This is helpful as it avoids having
to read the source code to know how to run the different options.
· Adds new '--should-show-only' option to 'search' command, to
better mimick the gnome-shell app search, by not returning
apps with NoDisplay=true.
Example for running a desktop app search with the new option from
inside the GLib build dir:
$ gio/tests/apps search --should-show-only settings
Make sure to fail consistently in case people created a GPropertyAction
with g_object_new() without passing a property name.
Bindings that construct objects with g_object_new() have no idea if a
property is mandatory.
See: #3130
This avoids a critical warning from trying to disconnect a signal
handler from a `NULL` object if `paction->object` is `NULL` for whatever
reason (see: the following commit).
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #3130
The `GTK_USE_PORTAL` environment variable has started to be misused by
users, which is causing deployment issues (such as portal services
themselves ending up being forced to use portals, which is never going
to work).
Try and sidestep users’ broken configurations by renaming the
environment variable, and also separating it from the old GTK
environment variable, since the GLib one affects a lot more processes.
This environment variable is meant to be used for
debugging and development, and never in production.
GTK already renamed their environment variable in
https://gitlab.gnome.org/GNOME/gtk/-/merge_requests/4829, so keeping the
`GTK_USE_PORTAL` name in GLib doesn’t make sense anyway.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #3107
I had thought that because `g_source_destroy()` was called for the two
sources (cancel and timeout) in the `GTask` finalize function for a
threaded resolver operation, that it would be fine to use a plain
pointer in the source callbacks to point to the `GTask`.
That turns out to not be true: because the source callbacks are executed
in the GLib worker thread, and the `GTask` can be finalized in another
thread, it’s possible for a source callback (e.g. `cancelled_cb()`) to
be scheduled in the worker thread, then for the `GTask` to be finalized,
and then the source callback to continue execution and find itself
doing a use-after-free.
Fix that by using a weak ref to the `GTask` in the source callbacks,
rather than a plain pointer.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
Fixes: #3105
The fields are fully validated in `validate_headers()` in
`gdbusmessage.c` now, so the connection code should be able to rely on
the required ones being non-`NULL`.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
Helps: #3061
`object_path` and `path` were doing exactly the same thing here.
This introduces no functional changes.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
We already validated that the required headers for each type of D-Bus
message were present. However, we didn’t validate that they contained a
variant of the right type. This could lead to functions like
`g_dbus_message_get_path()` returning `NULL` unexpectedly.
This failure could only be hit when using GDBus in peer-to-peer mode, or
with a D-Bus server which didn’t validate the headers itself. The
reference D-Bus server does validate the headers, and doesn’t forward
invalid messages to clients.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
Fixes: #3061
Add test cases that result in lookup of the port via
getservbyname().
As the result depends on "/etc/services", it's not reliably the same on
every system. It requires a workaround.
Commit cf55c31170 added a new test which
uses `ptrace()` to check some `GSubprocess` behaviour. FreeBSD uses
different symbol names for ptrace symbols, and we haven’t tested whether
the test works (and reproduces the failure) on FreeBSD, so skip the test
for now.
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
The tooling won’t pick them up unless they’re directly above the gettext
calls.
Spotted by Piotr Drąg in
ec03755355 (note_1808152).
Signed-off-by: Philip Withnall <philip@tecnocode.co.uk>
The test case will fail with the
g_assert_false (g_subprocess_get_successful (proc));
assert failing. Without the fix, it'll hit sometimes, but rather
unreliably. When running `meson test --repeat 100`, it'll reproduce
anywhere between the first or much later, but mostly before the 20th
iteration on my system.
Helps: #3071
It's not safe to use setlocale() to mutate the locale in a threaded
program. Lots of other tests still do this, and I'm not putting in the
effort to fix them comprehensively in the absense of actual failures on
CI, but I figured it'd be good to fix the tests that I was touching.
This definitely does not do anything on Linux. I bet it's not needed on
other platforms, either. It's unsafe and may crash; there is no safe way
to mutate the environment in threaded programs.
This is a copy of the existing test_l10n, modified to use LC_TIME
instead of LC_MESSAGES. It's not safe as each call to g_setenv() or
setlocale() could cause the test to crash; there is no safe way to
change a threaded process's environment, and a threaded process's locale
can only be safely changed using uselocale(), not with setlocale().
The calls to g_setenv() are definitely not needed on Linux. I wonder
whether removing these will break the test on other platforms?
The calls to setlocale() should be replaced by a dance of
uselocale() -> duplocale() -> newlocale() -> uselocale() on Linux. But
this is not portable and this is a cross-platform test. We would have to
make the test platform-specific to do this. macOS and at least FreeBSD
provide these functions via xlocale.h, but this isn't portable.
It's supposed to be possible to translate settings values using LC_TIME
rather than LC_MESSAGES to determine which translation to use, but
Sebastian Keller noticed that it's not working properly. I've
implemented his proposed solution, which is to actually temporarily
change LC_MESSAGES to match LC_TIME for just as long as necessary to
force gettext to use the desired message catalog.
Fixes#2575
Fix typo in gio/gappinfo (duplicated word)
Fix typo in gio/gtlsdatabase.c (duplicated word)
Fix typo in gio/gapplication.c (duplicated word, previous words rearranged to improve readability)
Fix typo in glib/tests/gvariant.c (duplicated word)
Fix typo in glib/win_iconv.c (duplicated word)
Fix typo in gio/gschema.dtd (meaning, missing letter)
Fix typo in gio/gdbusintrospection.c (duplicated word)
Fix typo in gio/gdbusintrospection.c (duplicated word).
I made the kqueue failure 100% reliable with `samu -j1` on FreeBSD,
and therefore confirmed this fixes that problem. Issue #2929 is
an identical failure on win32, so I assume this fixes that, too,
but I haven't confirmed.
Fixes: #2929
Following Emmanuele's instructions for use of introspection annotations:
https://www.bassi.io/articles/2023/02/20/bindable-api-2023/
I have audited all uses of the (closure) annotation in glib and
determined that only a handful are correct. This commit changes almost
all of our use of (closure) annotations to conform to Emmanuele's rules.
The test was passing fine when `bindir` was equal to `multiarch_bindir`,
but not when they differ.
For example, on a Debian system, `gio-querymodules` is installed to
`/usr/lib/x86_64-linux-gnu/glib-2.0/gio-querymodules` rather than
`/usr/bin/gio-querymodules` as it is on (say) Fedora.
This was causing the pkg-config tests to fail on Debian.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #3045
This reverts commit 004f48f4fc.
Per the discussion on #3356, this change was prompted by a
misunderstanding of ldflags/link_args, and it resulted in various other
packages using glib no longer getting symbols exported. This commit
restores the glib 2.76 behaviour.
Avoid generating more code than needed, so other than continuing using
the generic glib marshallers when possible, define once the custom ones
we need for each file we generate.
The marshallers are then re-used across all the interfaces defined
without duplicating the code size.
This is the same we're doing in code generated by glib-genmarshaller and
what gmarshal does internally.
Since the generated gdbus C code can be considered private too, this is
safe to do, and will allow faster access to GValue objects.
Get rid completely of the usage of the generic marshaller in gdbus
generated code, using instead specific marshallers.
Code is not yet optimized fully since we may still have duplicated
functions to be generated.
Closes: https://gitlab.gnome.org/GNOME/glib/-/issues/3028
We relied on g_cclosure_marshal_generic() to easily generate signal
marshallers, but this relies on inspecting each parameter type with ffi
and this implies a performance hit, other than breaking the stack-frame
unwinder used by Linux perf and so by sysprof.
Given that we know the types we work on, it's easy enough to generate
the marshallers ourself.
Helps with: https://gitlab.gnome.org/GNOME/glib/-/issues/3028
For some reason, `time_t` is defined as being 32 bits wide on that
platform, which causes truncation of the timestamps from `struct stat`.
Avoid that problem by consistently using a 64-bit return value from the
`struct stat` accessors.
Helps: #3039
We were using emit_by_name which implies looking up for the signal name,
while the generated code can easily remember the signal ID and use it
instead, allowing direct access to the signal emission.
`_g_stat_has_field (statbuf, G_LOCAL_FILE_STAT_FIELD_ATIME)` will always
return `TRUE` on Windows (since it uses a basic `struct stat`), so the
platform-inspecific code is equivalent to the Windows-specific code.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
`GLocalFileStat` is a platform-specific abstraction around `struct stat`
or `struct statx`. If `struct statx` is available, it will use that by
preference as it has more features.
`glocalfileinfo.c` was, in some places, incorrectly accessing the fields
of `GLocalFileStat` directly rather than through its `_g_stat_*()`
getters. While it correctly accounted for the platform-specific
differences between `st_mtimensec` and `st_mtim.tv_nsec`, it hadn’t been
updated to deal with `stx_mtime`.
On Android, `st_mtimensec` is defined as a fallback for
`st_mtim.tv_nsec` (even though it doesn’t need to be). This caused GLib
to take the `st_mtimensec` code path rather than the `stx_mtime` code
path, and hence try to dereference `st_mtim` in a `struct statx`.
Fix that by correctly using the `_g_stat_*()` getters consistently.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #3039
glib-compile-resources --dependency-file= currently generates a depfile
with rules that look like this:
foo.xml: resource1 resource2
This means that if any of the files listed in the GResource manifest
foo.xml change, rebuild foo.xml because foo.xml depends on those files.
This is not useful because the XML manifest is not expected to be a
generated dependency and even if it was, changes to the listed files
would not imply any need to regenerate the manifest. What we really do
need to regenerate is the C source file that is generated by
glib-compile-resources after processing the XML manifest and all the
resource files. That is, the rule should look like this:
foo.c: foo.xml resource1 resource2
as suggested by Hans Ulrich Niedermann in the issue report.
Fixes#2829
Currently we require explicitly specifying the port when configuring a
proxy server, which is seriously weird. I take the fact that nobody
reported a bug until 2022 to indicate that almost nobody is using
proxies. Whatever. Let's assume that if no port is provided, the default
port for the protocol should be used instead.
For example, you can now specify in GNOME settings that your proxy server
is https://example.com and it will work. Previously, you had to write
https://example.com:443. Yuck!
This was originally reported as GProxyResolver bug, but nothing is
actually wrong there. It's actually GProxyAddressEnumerator that gets
tripped up by URLs returned by GProxyResolver without a default port.
This breaks GSocketClient.
Fixing this requires exposing GUri's _default_scheme_port() function to
GIO. I considered copy/pasting it since it's not very much code, but I
figure the private call mechanism is probably not too expensive, and I
don't like code duplication.
Fixes#2832
On some platforms, pointer-sized reads are not necessarily atomic, so we
always need to use the correct atomic access primitives.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
When the gnome test runner executes the tests, the test appear to execute in disk
order. This means it sometimes works and sometimes we see breakage in portal-support-snap
and portal-support-snap-classic.
The issue is that some tests create config files but some don't. If they run
in the wrong order, tests see config files they shouldn't and break.
Fix this by deleting the files after each test run, properly cleaning up after
themselves. The cleanup code is based upon gtestutils.c:rm_rf().
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Since we have a convenience method to add actions let's allow to remove
them just as easily. This makes resource cleanup as simple as initially
adding the entries.
Makes the tests compile using clang with meson directly under
termux on android, this build environment does not approve of
overloading libc symbols.
Fixes: #3008
foo
This is a workaround for build conditions one ends up with under termux,
where the defined __ANDROID_API__ level is lower than what is provided
by gcc installed for it, the libc .so nevertheless contains these symbols
thus enabling the codepaths. This definition is only in use when meson
detected the presence of this symbol in the libc.
Fixes#3008
foo
In the typical `while (g_file_enumerator_next_file ())` patterns,
there is nothing much checking whether the operation was cancelled
on the GIO side. Unless the user checks for the case, this means
local enumerators always run to completion even if cancelled.
Fix this by checking the cancellable state explicitly for local
enumerators, so there are oportunities for bailing out early if
the enumerator is going through a very large directory.
Note that the prepare callback only has one caller, which pre-initializes
the timeout argument to -1. That may be an implementation detail and not
publicly promised, but it wouldn't make sense to do it any other way in
the caller.
Also, note that g_unix_signal_watch_prepare() and the UNIX branch of
g_child_watch_prepare() already relied on that.
When `copy_file_range()` support was added, I used the definition of
`copy_file_range()` from Linux, which uses `loff_t` to abstract the
different `off*_t` types.
`loff_t` doesn’t exist on FreeBSD, so this doesn’t compile, and was
caught in subsequent asynchronous CI.
Define `loff_t` with a fallback value if it’s not defined, which should
fix this and other uses of `loff_t` in `gfile.c` (for example, if
FreeBSD ever starts declaring `splice()`).
Fixes this CI failure: https://gitlab.gnome.org/GNOME/glib/-/jobs/2812302
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
GTK lost it's '+' suffix back in 2019, according to
<https://mail.gnome.org/archives/gtk-devel-list/2019-February/msg00000.html>
This commit can be re-generated with:
git grep -l GTK+ \
| grep -v -e ^NEWS -e ^glib/tests/collate.c \
| xargs sed -i 's/GTK+/GTK/g'
Most of the changes are in comments and documentation.
While it can’t be used in all situations, it is a little bit faster than
`splice()` in some situations, basically if the file system supports
copy on write. In other situations it’s no slower than `splice()`.
See `man copy_file_range` for the situations where it doesn’t work. In
all of these situations, it will return an error, and the GLib code will
fall through and try the existing `splice()` copy code instead.
From my testing of `time gio copy A B` with a 9GB file, the `splice()`
code path takes 22s, and the `copy_file_range()` code path takes 20s.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #2863
The start of the `g_file_copy()` implementation stats the source file to
find all the attributes to copy onto the destination file, so it makes
sense to get it to store the source file size at the same time.
This saves a subsequent `stat()` call on the source FD in the btrfs
reflink or splice code. Every little helps.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Something has changed recently which causes this error to now be emitted
when building on Windows msys2-mingw32:
```
../gio/gwin32networkmonitor.c: In function 'win_network_monitor_get_ip_info':
../gio/gwin32networkmonitor.c:92:15: error: storing the address of local variable 'prefix' in '*dest' [-Werror=dangling-pointer=]
92 | *dest = (guint8 *) &prefix.Prefix.Ipv4.sin_addr;
| ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
```
If `IP_ADDRESS_PREFIX` is defined as a scalar rather than a pointer,
that could explain the problem.
Change the function to always operate on a pointer to avoid any
potential such issues.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
The gio/tests/socket-client.c doesn't use GSocketClient, which makes the
filename confusing. What the file actually tests is the GSocket. Rename
it to socket-testclient.c
The corresponding GSocket server test file naming doesn't conflict with other
class names, but rename it to socket-testserver.c for consistency.
Closes#2855
The file was not listed in `POTFILES.in` (as pointed out by Piotr Drąg
in
fee0a7679a (note_1722885)),
so either it needs to be added to `POTFILES.in` or the translatable
strings need to be removed.
Recent prior art from GTK shows that there’s actually no longer any
point in setting the nick/blurb as no tools use them (and if they did,
it would result in a rubbish user experience). See
https://gitlab.gnome.org/GNOME/gtk/-/issues/4904.
Hence, drop the strings entirely.
See #2991 for tracking this across all of GIO.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #2991
If a timeout executes on the same main context iteration as completion
or cancellation of a resolver lookup, `has_returned` will be set
multiple times. That’s fine (the `GCond` will be notified multiple
times, but that’s fine). It was triggering an incorrect assertion, so
remove that.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
The default for the class is still to have no timeout, but it seems more
practical for most use cases to set a non-infinite timeout on the
default resolver.
If applications have a more specific use case, they can change the
timeout or replace the default resolver.
See https://gitlab.gnome.org/GNOME/glib/-/merge_requests/3397#note_1731387
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
If `async_cancel()` was invoked, it would remove the IO watch source,
which would cause the `g_source_remove()` call at the end of `main()` to
warn about an unknown source ID.
Fix that by handling the source as a pointer instead of a handle.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Rather than running lookups in the global shared thread pool belonging
to `GTask`, run them in a private thread pool.
This is needed because the global shared thread pool is constrained to
only 14 threads. If there are 14 ongoing calls to
`g_task_run_in_thread()` from any library/code in the process, and then
one of them asks to do a DNS lookup, the lookup will block forever.
Under certain circumstances, particularly where there are a couple of
deep chains of dependent tasks running with `g_task_run_in_thread()`,
this can livelock the program.
Since `GResolver` is likely to be called as a frequent leaf call in
certain workloads, and in particular there are likely to be several
lookups requested at the same time, it makes sense to move resolver
lookups to a private thread pool.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
This will make it simpler to handle timeouts and cancellation in future,
as all the logic for working out whether to return will all be in one
place, and all the lookup-specific code is now implemented in simple
sync functions which don’t need to care about `GTask`s.
This commit introduces no functional changes, it’s just setting up for
the following commit.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
This introduces no functional changes, but will make a reorganisation of
the code simpler in the next commit.
Rather than dealing with three different closure types, this changes the
code to deal with one which is a tagged union of the three.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
The class and its header are not public, so this should not be an API or
ABI break.
This just simplifies the code a little and allows for easy extension of
the object’s private data in future commits.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Without a timeout, some lookup requests can go on forever, typically due
to bugs in underlying systems.
This can have particularly significant effects on the Happy Eyeballs
algorithm in `GSocketClient`, which relies on multiple name lookups as
its first step.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #2866
Track the `GTask`s which are still alive (not finalised) in a shared
list, and provide a secret debugging function for printing that list.
Too often when debugging apps, I have found that a ‘leaked’ object is
actually still (validly) referenced by an ongoing `GTask` which hasn’t
completed for whatever reason. Or I have found that an operation has
obviously stalled, but there are no pointers available to the `GTask`
which is stalled, because it’s being tracked as a collection of closure
pointers from some `GSource` which is hard to get to in the debugger.
It will be very useful for debugging apps, if there’s a list of all the
still alive `GTask`s somewhere. This is that list.
The code is disabled if `G_ENABLE_DEBUG` is not defined, to avoid every
`GTask` construction/finalisation imposing a global locking penalty.
To use the new list, break in `gdb` while running your app, and call
`g_task_print_alive_tasks()`, or inspect the `task_list` manually:
```
(gdb) print g_task_print_alive_tasks()
16:44:17:788 GLib-GIO 5 GTasks still alive:
• GTask 0x6100000ac740, gs_plugin_appstream_setup_async, ref count: 1, ever_returned: 0, completed: 0
• GTask 0x6100000bf940, [gio] D-Bus read, ref count: 2, ever_returned: 0, completed: 0
• GTask 0x6100000aac40, gs_plugin_loader_setup_async, ref count: 1, ever_returned: 0, completed: 0
• GTask 0x61000006d940, gs_plugin_loader_job_process_async GsPluginJobRefine, ref count: 1, ever_returned: 0, completed: 0
• GTask 0x610000118c40, [gio] D-Bus read, ref count: 2, ever_returned: 0, completed: 0
```
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
This commit changes the use of `ngettext` with `g_dngettext`. The
project defined `g_dngettext` (with domain support) provides the same
functionality as `ngettext` with a NULL domain provided. The purpose of
this change is to help address a build error for certain compilers that
trigger a `format-nonliteral` error-promoted-warning when using
`ngettext` (see also [1][2]). The benefit of switching to use
`g_dngettext` is that the function is defined with `G_GNUC_FORMAT`. This
provides a hint to GNU GCC compilers to still sanity check these
arguments, but not generate a `format-nonliteral`.
[1]: 4ae8606b6f
[2]: 0ca660315a
Signed-off-by: James Knight <james.d.knight@live.com>
This reverts commit 4ae8606b6f. The idea
for the change [1] was to address a build error for certain compilers
that trigger a `format-nonliteral` error-promoted-warning since these
compilers do not gracefully support `ngettext` usage. The changes
following a pattern from an old commit [2]; however, James Hilliard has
pointed out these changes do not work as intended. A deeper inspection
of the commit showed that the commit was from an old merge request that
was not pulled in, detailing why the changes did not work (see also
[3][4]).
Manipulating the sockets unit test confirms that the format values no
longer get a proper value:
...
ok 9 /socket/address
ok 10 /socket/unix-from-fd
ok 11 /socket/unix-connection
**
GLib-GIO:ERROR:../gio/tests/socket.c:1493:test_unix_connection_ancillary_data: assertion failed (err == NULL): Expecting one fd, but got %d
(g-io-error-quark, 0)
...
And reverting this change restores the original functionality:
...
ok 9 /socket/address
ok 10 /socket/unix-from-fd
ok 11 /socket/unix-connection
**
GLib-GIO:ERROR:../gio/tests/socket.c:1493:test_unix_connection_ancillary_data: assertion failed (err == NULL): Expecting 1 control message, got 0 (g-io-error-quark, 0)
...
[1]: https://gitlab.gnome.org/GNOME/glib/-/merge_requests/3390
[2]: 44b3d5d80445234041f6c59feb89645f7102c3a4
[3]: https://gitlab.gnome.org/GNOME/glib/-/merge_requests/770
[4]: https://gitlab.gnome.org/GNOME/glib/-/issues/1744
Signed-off-by: James Knight <james.d.knight@live.com>
These make it a bit easier to track the ongoing resolver tasks, as the
tasks and/or their closures are not tracked in a big list somewhere.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
It’s a bad idea to use it without some care for how much it’s being
called in parallel, or dependencies between tasks. If the thread pool
gets exhausted by too many inter-dependent calls to
`g_task_run_in_thread()` then the process will livelock.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Update a series of error messages to use `g_set_error_literal` instead
of `g_set_error`. This should prevent `format-nonliteral` compiler
issues when `-Werror` is configured:
../gio/gunixconnection.c: In function ‘g_unix_connection_receive_fd’:
../gio/gunixconnection.c:183:9: error: format not a string literal, argument types not checked [-Werror=format-nonliteral]
183 | nscm);
| ^~~~
../gio/gunixconnection.c:217:20: error: format not a string literal, argument types not checked [-Werror=format-nonliteral]
217 | nfd);
| ^~~
../gio/gunixconnection.c: In function ‘g_unix_connection_receive_credentials’:
../gio/gunixconnection.c:601:24: error: format not a string literal, argument types not checked [-Werror=format-nonliteral]
601 | nscm);
| ^~~~
This is similar to a previous change [1] made to `gunixconnection.c`.
[1]: 44b3d5d80445234041f6c59feb89645f7102c3a4
Signed-off-by: James Knight <james.d.knight@live.com>
Since commit c0ca3f99 this test is strictly depending on GDesktopAppInfo
that is not defined or available in macos, so skip the test as we do for
windows.
We could have done this at meson level too, but keeping it this way is
probably a better reminder that this should be adapted for such scenario
one day™
See: https://gitlab.gnome.org/GNOME/glib/-/jobs/2753753
The merge request !2848 added code to automatically detect the module
prefix on macOS, with a test for the Mac #define TARGET_OS_OSX. However,
older versions of the SDK (at least 10.11) don't provide this #define,
leading to build failure. If the #define is missing, fall back to
checking TARGET_OS_MAC. On newer SDKs this symbol is also true for
watchOS, etc., but in those situations TARGET_OS_OSX is available.
Various gio modules include gmodule.h that requires the
gmodule-visibility.h to be already built.
To make this easier, just provide a dependency and use it where we are
building modules that do not depend on libgio_dep (that already includes
that).
Fixes: https://gitlab.gnome.org/GNOME/glib/-/issues/2982
The generated gir file marks the size parameter as "out" by default. This is wrong in the context of a caller allocated buffer with a given size. Explicitly marking the size parameter as (in) fixes the issue.
This reverts commit 4cad66580b.
Downgrading the criticals was only temporary. Now we’ve branched for
GLib 2.78, the criticals can be reinstated early this cycle, so people
have the maximum time to fix latent bugs in their code.
Fixes: #2951
The ref on the timeout source owned by `SendMessageData` was being
dropped just after attaching the source to the main context, leaving it
unowned in that struct. That meant the only ref on the source was held
by the `GMainContext` it was attached to.
This ref was dropped when returning `G_SOURCE_REMOVE` from
`send_message_with_reply_timeout_cb()`. Before that happens,
`send_message_data_deliver_error()` is called, which normally calls
`send_message_with_reply_cleanup()` and destroys the source.
However, if `send_message_data_deliver_error()` is called when the
message has already been delivered, calling
`send_message_with_reply_cleanup()` will be skipped. This leaves the
source pointer in `SendMessageData` dangling, which will cause problems
when `g_source_destroy()` is subsequently called on it.
I’m not sure if it’s possible in practice for this situation to occur,
but the code certainly does nothing to prevent it, and it’s easy enough
to avoid by keeping a strong ref on the source in `SendMessageData`.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #1264
It already implicitly held a strong ref on its `GTask` values, but
didn’t have a free function set so that they would be automatically
unreffed on removal from the map.
This meant that the functions handling removals from the map,
`on_worker_closed()` (via `cancel_method_on_close()`) and
`send_message_with_reply_cleanup()` had to call unref once more than
they would otherwise.
In `send_message_with_reply_cleanup()`, this behaviour depended on
whether it was called with `remove == TRUE`. If not, it was `(transfer
none)` not `(transfer full)`. This led to bugs in its callers.
For example, this led to a direct leak in `cancel_method_on_close()`, as
it needed to remove tasks from `map_method_serial_to_task`, but called
`send_message_with_reply_cleanup(remove = FALSE)` and erroneously didn’t
call unref an additional time.
Try and simplify it all by setting a `GDestroyNotify` on
`map_method_serial_to_task`’s values, and making the refcount handling
of `send_message_with_reply_cleanup()` not be conditional on its
arguments.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #1264
This is equivalent to the current behaviour, but a little clearer in its
meaning.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #1264
The ownership transfers in this code are a bit complex, so adding some
extra documentation and `g_steal_pointer()` calls should hopefully help
clarify things.
This doesn’t introduce any functional changes, just code documentation.
Another drive-by improvement in the quest for #1264.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #1264
The existing comment in the code was correct that `data` is freed when
the task callback is called, because `data` is also pointed to by the
`user_data` for the task, and that’s freed at the end of the callback.
So the existing code was correct to take a copy of `data->task` before
calling `g_task_return_*()`.
After calling `g_task_return_*()`, the existing code unreffed the task
(which is correct), but then didn’t clear the `data->task` pointer,
leaving `data->task` dangling. That could cause a use-after-free or a
double-unref.
Avoid that risk by explicitly clearing `data->task` before calling
`g_task_return_*()`.
After some testing, it turns out this doesn’t actually fix any bugs, but
it’s still a good robustness improvement.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #1264
The generated gir file marks the size parameter as "out" by default. This is wrong in the context of a caller allocated buffer with a given size. Explicitly marking the size parameter as (in) fixes the issue.
As explained in the previous commit, GNU Coreutils 9.2 changes the behaviour
of `du --bytes` to only count regular files and symlinks.
The previous commit makes the test pass with GNU Coreutils >=9.2, but the
machine running the tests may have an older version, or perhaps even a
reimplementation such as uutils. So we can't rely on the size returned by `du`
to be the consistent across systems any more.
However, the plus side of the new behaviour is that the size reported by `du`
/ `G_FILE_MEASURE_APPARENT_SIZE` is now well-defined across filesystems
(as the sum of the sizes of regular files & symlinks), so we can hardcode it.
Fixes: https://gitlab.gnome.org/GNOME/glib/-/issues/2965
Since GNU Coreutils 9.2 (commit 110bcd28386b1f47a4cd876098acb708fdcbbb25),
`du --apparent-size` (including `du --bytes`) no longer counts all kinds of
files (directories, FIFOs, etc.), but only those for which `st_size` in
`struct stat` is defined by POSIX, namely regular files and symlinks
(and also rarely supported memory objects).
This aligns the behaviour of GLib's `G_FILE_MEASURE_APPARENT_SIZE` flag
with the new GNU Coreutils `du` and correct POSIX use.
Note that this may be a breaking change for some uses.
Link: https://lists.gnu.org/archive/html/bug-coreutils/2023-03/msg00007.html
Fixes: https://gitlab.gnome.org/GNOME/glib/-/issues/2965
Warn when the boolean return isn't used, since we may not initialize
the out arguments in the FALSE case. Update all internal callers to
use the return value. They were already safe, but users outside GLib
may not be.
Previously, `-Wl,--export-dynamic` was in `Libs` key of `gmodule-2.0.pc`,
even though `-Wl` is a compiler flag, rather than a linker one.
This caused issues with API reference builds in evolution-data-server,
which passes the output of `pkg-config --libs` through `--ldflags`
argument of `gtkdoc-scan`, which are forwarded unchanged to `ld`:
ld: unrecognized option '-Wl,--export-dynamic'
Let’s move the flag to `Cflags` so that the compiler can deal with it.
https://gitlab.gnome.org/GNOME/evolution-data-server/-/issues/463
Add support for it, but don’t use it anywhere yet — this is an API
addition, but currently doesn’t cause any functional changes. It’ll be
used in the next commit.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
The `inotify_init1()` API has supported this flag for a long time
(possibly since it was first introduced, although I haven’t bothered
doing the archaeology).
This saves a syscall when first connecting to inotify.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Both `GPollableInputStream` and `GPollableOutputStream` are dynamic
interfaces, in that their implementation on a class may only be
functional if certain prerequisites are met at runtime. For example,
a `GConverterInputStream` is only pollable if its base stream is
pollable, and that’s determined at runtime rather than compile time.
As such, both interfaces have a `can_poll()` method. If that method
returns `FALSE`, the behaviour of all other methods on the interface is
undefined.
That was mentioned in the documentation for `can_poll()`, but not any of
the other documentation for the interfaces, which made it a bit hard to
find.
Mention it more widely.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Fixes: #2952
If the libc and kernel support `SOCK_NONBLOCK`, we can specify that in
the `socket()` flags, and avoid a subsequent call to `fcntl()` to set
`O_NONBLOCK`.
For modern Linux distributions, this will save a syscall when creating a
socket.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
This should split the code up into logical blocks a bit better, and make
it a bit easier to see what the test is doing at a glance.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
This makes the exit conditions for each main loop clearer, and
eliminates use of global variables. It introduces no functional changes.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
If the first part of the test takes less than 3s (which is normal), the
timeout for it is not removed, and could spuriously fire during the
second part of the test, causing a false failure.
Instead of relying on source IDs, just use (and explicitly destroy) a
`GSource` for the timeouts.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
The test thought that calling `g_app_info_get()` was a bit of a hack,
but actually it (or calling another `g_app_info_*()` function) is the
right way to use the `GAppInfoMonitor` API.
See the documentation improvements a couple of commits back for details.
The remaining FIXME higher up in the test should probably be fixed by
getting `g_app_info_monitor_get()` to arm the signal. That requires
changes in `g_app_info_monitor_get()` to call `desktop_file_dir_init()`.
That will have to happen another time.
Signed-off-by: Philip Withnall <pwithnall@endlessos.org>
Helps: #799
