glib/gkeyfile.c | 83 83 + 0 - 0 !
1 file changed, 83 insertions(+)

 call gettext if .desktop file does not have inline translations

Patch from OpenSUSE via Ubuntu, original author unknown. Martin Pitt and
Vincent Untz appear to be the main authors.

Reworked slightly by Philip Withnall to avoid exposing new public API
for the non-standard keys.

Bug: https://bugzilla.gnome.org/show_bug.cgi?id=569829
Bug-Ubuntu: https://launchpad.net/bugs/3935

glib/gkeyfile.c | 20 20 + 0 - 0 !
1 file changed, 20 insertions(+)

 provide backwards compatibility for 01_gettext-desktopfiles.patch
 for X-{Debian,Ubuntu}-Gettext-Domain

Ubuntu-specific. 01_gettext-desktopfiles.patch was changed to use
X-GNOME-, so this is necessary until all our .desktop files are converted.

gio/glib-compile-schemas.c | 4 4 + 0 - 0 !
1 file changed, 4 insertions(+)

 disable confusing (to users) warning about deprecated schema paths

Disable a warning when compiling schemas which are installed
into 'deprecated' locations. Users see this very often due to
glib-compile-schemas being called from libglib2.0-0's trigger and it is
not very useful for them.

gio/gdesktopappinfo.c | 1 1 + 0 - 0 !
1 file changed, 1 insertion(+)

 gdesktopappinfo: try using x-terminal-emulator for terminal=true
 apps

Debian Policy provides x-terminal-emulator as an interface for launching
a preferred terminal, for some definition of "preferred". However, the
x-terminal-emulator alternative is a system-wide choice, so in situations
where for example a GNOME user and a KDE user share a computer, only one
of them can have the terminal that will match their desktop environment's
appearance and behaviour conventions as the x-terminal-emulator.

As a result, we still try to use a GTK-based terminal as a higher
preference than x-terminal-emulator. This is done on the assumption
that when apps are launched using GLib interfaces, they are most likely
to have been launched from a GTK application or GTK-based desktop
environment, and therefore a GTK-based terminal will be the best fit for
the desktop environment's conventions and appearance.

I've somewhat arbitrarily sorted x-terminal-emulator as less preferred
than KDE's Konsole, but more preferred than rxvt and various xterm
variants which are not associated with a particular desktop environment.

Bug-Debian: https://bugs.debian.org/773915

glib/tests/timer.c | 4 2 + 2 - 0 !
1 file changed, 2 insertions(+), 2 deletions(-)

 timer test: use 'volatile' for locals

GCC seems to be failing to follow the letter of the C spec by allowing extra
precision in floating point values to persist across assignments which are
optimised away.

Force its hand by using 'volatile' on the locals in question.

Bug: https://gitlab.gnome.org/GNOME/glib/issues/820

glib/tests/gwakeuptest.c | 17 13 + 4 - 0 !
1 file changed, 13 insertions(+), 4 deletions(-)

 gwakeuptest: be less parallel unless invoked with -m slow

This is a workaround for test failures on the reproducible-builds
infrastructure, where a multi-threaded stress-test sometimes takes longer
to finish on x86_64 than it would have done on slow architectures like
arm and mips on the official Debian autobuilders. It is not clear why.

This change will make this test more likely to pass, but less likely to
detect bugs.

Signed-off-by: Simon McVittie <smcv@debian.org>
Bug-Debian: https://bugs.debian.org/884659

gobject/tests/closure-refcount.c | 8 8 + 0 - 0 !
1 file changed, 8 insertions(+)

 closures test: skip on arm* unless flaky tests are allowed

Choosing the right number of iterations to avoid either taking literally
hours on some hardware, or getting spurious failures when one thread
starves another, seems to be too hard to get right in practice.
Make this test opt-in so that its failures aren't release-critical.
We can run it as a separate autopkgtest that is marked flaky.

Signed-off-by: Simon McVittie <smcv@debian.org>
Bug-Debian: https://bugs.debian.org/880883
Bug-Debian: https://bugs.debian.org/917983

glib/tests/mainloop.c | 16 16 + 0 - 0 !
glib/tests/timeout.c | 9 9 + 0 - 0 !
gobject/tests/threadtests.c | 8 8 + 0 - 0 !
3 files changed, 33 insertions(+)

 disable some tests on slow architectures which keep failing the
 tests

[smcv: Modified to use g_test_skip() instead of omitting those test cases
completely, and allow them to be re-enabled with a Debian-specific
environment variable]

Co-authored-by: Simon McVittie <smcv@debian.org>

glib/tests/timer.c | 6 6 + 0 - 0 !
1 file changed, 6 insertions(+)

 skip test which performs some unreliable floating point comparisons

[smcv: Modified to use g_test_skip() instead of omitting those test cases
completely, and allow them to be re-enabled with a Debian-specific
environment variable]

Co-authored-by: Simon McVittie <smcv@debian.org>
Bug: https://gitlab.gnome.org/GNOME/glib/issues/820

gio/tests/gdbus-threading.c | 6 6 + 0 - 0 !
1 file changed, 6 insertions(+)

 skip unreliable gdbus-threading tests by default

test_threaded_singleton() test to reproduce a race condition between
last-unref of the global singleton GDBusConnection and g_bus_get_sync().

test_method_calls_in_thread() checks that multiple threads can all make
method calls to the same proxy.

However, test setup intermittently times out with:

    # GLib-GIO-DEBUG: run 0: refcount is 2, sleeping
    Bail out! GLib-GIO-FATAL-ERROR: connection had too many refs

The current theory upstream is that this might be a reference leak in
test_delivery_in_thread().

Furthermore, test teardown is now often failing when destroying the test
bus.

Demote these tests to be run as part of the "flaky" autopkgtests, but
not at build time or in the part of the autopkgtest run that gates
progress into testing.

Bug: https://gitlab.gnome.org/GNOME/glib/issues/1515

glib/tests/gvariant.c | 8 8 + 0 - 0 !
1 file changed, 8 insertions(+)

 gvariant test: don't run at build-time on mips

DEB_ALLOW_FLAKY_TESTS is not quite right here, because we don't know
that the test would fail if left for long enough - the problem is that
it doesn't get there, because generating random floating-point numbers
is very slow on some of our mips hardware. However, it has the right
practical effect.

Signed-off-by: Simon McVittie <smcv@debian.org>
Bug: https://bugs.debian.org/932678

gio/tests/gdbus-server-auth.c | 7 7 + 0 - 0 !
1 file changed, 7 insertions(+)

 gdbus-server-auth: normally skip flaky dbus_cookie_sha1 tests

These intermittently fail on the buildds, but the failure cannot be
reproduced in a debugging environment.

We do not expect to use D-Bus over TCP on non-Windows platforms: we use
an AF_UNIX socket, which is much more robust and secure. However, when
using AF_UNIX, DBUS_COOKIE_SHA1 is unnecessary, because we can use the
more reliable EXTERNAL authentication.

gio/tests/memory-monitor-dbus.py.in | 2 2 + 0 - 0 !
1 file changed, 2 insertions(+)

 skip memory-monitor-dbus test if not specifically requested

This seems to be unreliable, particularly on non-x86.

Bug-Debian: https://bugs.debian.org/995178

gio/tests/debugcontroller.c | 6 6 + 0 - 0 !
1 file changed, 6 insertions(+)

 tests: skip debugcontroller test

This is known to be flaky upstream.

gio/tests/testfilemonitor.c | 38 37 + 1 - 0 !
1 file changed, 37 insertions(+), 1 deletion(-)

 testfilemonitor: skip if we are avoiding flaky tests

See https://gitlab.gnome.org/GNOME/glib/issues/1634

Signed-off-by: Simon McVittie <smcv@debian.org>

docs/reference/girepository/gi-compile-repository.rst | 9 9 + 0 - 0 !
docs/reference/girepository/gi-decompile-typelib.rst | 9 9 + 0 - 0 !
docs/reference/girepository/gi-inspect-typelib.rst | 9 9 + 0 - 0 !
3 files changed, 27 insertions(+)

 girepository: describe the debian-specific cross-prefixed names

glib/gfileutils.c | 25 23 + 2 - 0 !
glib/tests/fileutils.c | 14 10 + 4 - 0 !
2 files changed, 33 insertions(+), 6 deletions(-)

 gfileutils: preserve mode during atomic updates

If g_file_set_contents{_full,} is replacing an existing file, require
that the tmpfile have the same mode as the existing file.

This prevents the umask from taking effect for consistent writes to
existing files.

Closes GNOME/dconf#76

(cherry picked from commit 3cc0c0de33bc4b461e89b05d142e1ecf5f474317)