| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
| |
The runtime could be in an indistinct state when the signal was
emitted. Now it is possible to shutdown the runtime in the signal
handler. Also fixed object lifetime in the in-process runtime.
Fixes: AUTOSUITE-1431
Change-Id: I2782f4d91872c1037462bbce171fa432ed43fb6e
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
| |
Change-Id: I0926e68e411c4285d70144b2db4b6a848b39a784
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
MingW ships with lower-case header names instead of the camel-case names that
MSVC is providing. This is not a problem when building on Windows, but
including a camel-cased header fails when doing a MingW cross-build on a
Linux machine with a case-sensitive filesystem.
Fixes: AUTOSUITE-1484
Change-Id: Iebfc9b58956eb274c3cccee070b4ea4f69e105a3
Reviewed-by: Dominik Holland <dominik.holland@qt.io>
|
|
|
|
|
|
|
| |
This adds a Windows version to the already existing Unix mechanism.
Change-Id: I7d92699875730322507d19a26cce794918caf267
Reviewed-by: Dominik Holland <dominik.holland@qt.io>
|
|
|
|
|
|
|
|
|
| |
Right now we cannot create a backtrace, but at least we can use
the crashhandler to kill all child processes and also print a
QML backtrace and other useful information
Change-Id: I865e4d33dce5f6e032f550cd680176e1b3b54e19
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
This will handle all the usual suspects, but it will not be able to handle
SIGKILL.
Please note: this is Unix-only for now.
Change-Id: I1fcac4dd8ddf2ef12d988012b04c2cb74b6f5f35
Reviewed-by: Dominik Holland <dominik.holland@qt.io>
|
|
|
|
|
|
|
|
|
|
|
| |
QRegExp is not thread-safe, but we are using it in multiple threads when
parsing config files on startup.
Backported the 5.14 implementation and added docs on why it works now.
Also backported and extended the 5.14 autotest.
Change-Id: I09d93b9783c8af4dffe0b3b37efc0754d293c955
Fixes: AUTOSUITE-1410
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
|
|
|
|
| |
Change-Id: I5c08b3e266a57fe113d59ad0781a33250ab6f327
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
IPC extensions did not work in an application, when it was started in
a quick-launched runtime, because the extension interfaces were not
registered.
Besides, a crash is avoided, should an ApplicationInterfaceExtension
type be used in quicklaunchQml code (although the interface will not
work).
Change-Id: Ie6a86d72fd41b401ce8754d0b6f4379749b56ea7
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
Get rid of wanring "load() is deprecated". Besides, one mutex should
perform better than ten atomic ints. On my machine it's roughly twice
as fast.
Change-Id: I35254604b7739f44011b3799c5400a5d0140949a
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
| |
Show critical message, if an application's QML cannot be loaded.
Change-Id: Ie1a6f0632861cb17af91c5ead297f1b6512fd3c0
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
Function __cxa_demangle will reallocate the output buffer, if it is
too short. To avoid a potential dangling buffer (pointer), it is
always updated. Increased the initial buffer size, as well.
Change-Id: I11f331e51fa28f940d4843740c0b0137d8d47dc4
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
Privatization of /tmp through mount namespaces (tmpfs)
and bind mount the P2P sockets per tools with static definition, like systemd,
requires a statically known path deeper than simply /tmp
Change-Id: I422a04753a84c27127ffd8d5e5c7ace5797d46c1
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
| |
- Moved smaps parsing to dedicated thread
- Made memory reporting optional
- Support CPU loads greater than 1.0
Task-number: AUTOSUITE-1310
Change-Id: Ic7e647a5270b1be2e3ec19c5122dd29156ea0a88
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
| |
Both tools are native tools and are not supposed to be cross-compiled.
To support the yocto SDK creation the tools can be forced to be
build by using the 'tools-only' config option.
Change-Id: I89e34b3bb18b59dde1a807949eb41d1ba535da7b
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
| |
The client error string was not updated when the server was called
directly (short-circuite).
Change-Id: Ie4ed81008f16eb89e3397a195f4864300abfa18b
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
| |
Change-Id: Ibb9acd12252d7a5489b6ccf7919d1ba938eb62e8
Fixes: AUTOSUITE-1330
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
|
|
| |
Currently every project which uses the am-app feature fails with
an error if the appman-packager exectuable is not available.
Instead check for the executable before and return early with
a warning.
Change-Id: I6725d6d801baa3a60bbe381b3b42fe680dcbff3e
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
|
|
|
| |
This fixes the file listing in the custom-appman example and also makes
sure that the example is properly listed in the QtCreator example panel.
Change-Id: I2dd5e72ac3c93a410aa6f9229b18504add8db821
Reviewed-by: Robert Griebl <robert.griebl@qt.io>
|
|
|
|
|
| |
Change-Id: Ifb641df2440db29c0970a636cab970754de6efdf
Reviewed-by: Dominik Holland <dominik.holland@qt.io>
|
|\
| |
| |
| | |
Change-Id: Ice9f495adf5d9b464a2567a0346d68c528157639
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The messagePattern given in the am-config.yaml is set as env variable
QT_MESSAGE_PATTERN. However this might be too late for the System UI
(the env variable is evaluated by Qt before it is set by appman), so
the pattern will be set explicitly with qSetMessagePattern.
Change-Id: Ie8c9858feb600cd2b03ed5d6f77874235e0a05df
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Canceling an update left the app in an inconsistent state. There is
still a known issues: canceling the update of an already updated
built-in app will revert back to the original app. This will be
fixed in 5.14
Also downgrading a built-in app didn't work as expected (the app would
be blocked, the icon etc. wouldn't be updated).
Change-Id: I6c72e7f87e993a6839e97ceb6cea443f07e1fd77
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: I2c0fd0ec92cc06475b37ee157f95f2bff8a1c851
Reviewed-by: Kavindra Palaraja <kpalaraja@luxoft.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This adds support to define pluginPaths for the system-ui,
the qml and qml-in-process runtimes in am-config.yaml.
In addition the apps can now also define their own pluginPaths
in their info.yaml
Fixes: AUTOSUITE-1219
Change-Id: I1e1ae4d8d7f53dbfb3b679910d7ec57c50f90004
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
|\ \ |
|
|/| |
| |/
| |
| | |
Change-Id: I7bacb588ea023b8110449b5313d517089f799d56
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
passed to DLT"
This reverts commit 78beb6acca90fb2f2490e2b60d2103a25f1508fe.
This change was breaking SC and can't be released with a
patch release. The change will be put into the dev branch
instead.
Task-number: AUTOSUITE-1187
Change-Id: Iff07e290308c2464ab25c5ef17b1749ddfd85723
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
This uses a new config option in QtGeniviExtras and makes it possible
to set this in the am-config.yaml
Change-Id: I7e13e9114dd0e281cdd8bf85e517bd5654f49c64
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* Introduce an option to configure the qtMessagePattern in the config.yaml
* Add an option to configure whether the AM internal console logging
function should be used.
Cherry-picked from 5.13: 09f0e58
Change-Id: Ied5b08766072431384c0718529600b767d200bf7
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
Reviewed-by: Kavindra Palaraja <kpalaraja@luxoft.com>
|
| |
| |
| |
| |
| | |
Change-Id: I6a9e2004c4eedaf7c136679c7cbf44a7a94d57e5
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| | |
Task-number: AUTOSUITE-1111
Change-Id: Ia19034c07b59b33cce4576f3d8c57fba33838ac5
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
* Minor fixes to the "Write Applications" topic that I've been looking at
Change-Id: I372729f3e08340d3a78569cb15ee05f2cc4c2e6e
Fixes: AUTOSUITE-867
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
|\ \ |
|
|/| |
| |/
| |
| | |
Change-Id: Id223c3a0679698f4d90715d3e43a76fc68a52d18
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
In singel-process mode the ApplicatonObject could still have an
associated runtime object, although the runtime had been terminated.
This sometimes caused a runtime object to be reused by mistake.
Change-Id: Ib4184281a04b7da8f08a848177cf87bace1ce30a
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| | |
Change-Id: I21e3974701df3ad5c6a655f3ab97b79cff2202d4
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Normally the AM would honor the $WAYLAND_DISPLAY environment variable, but
this is always set to "wayland-0" on Wayland desktop environments which
messes up the AM.
We now ignore the env. variable if the the platform is "wayland" and the
value of the variable is "wayland-0".
Change-Id: I831660879b877ecfa2b03e8b01b4fa501a90517f
Reviewed-by: Vladimir Minenko <vladimir.minenko@pelagicore.com>
Reviewed-by: Bernd Weimer <bernd.weimer@pelagicore.com>
|
|\ \ |
|
|/| |
| |/
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Conflicts:
.qmake.conf
qmake-features/am-config.prf
src/main-lib/main.cpp
src/manager-lib/applicationmanager.cpp
src/monitor-lib/monitor-lib.pro
Change-Id: I4aa0f2c901480d1f04f4260fe4ab1e005369432e
|
| |
| |
| |
| |
| | |
Change-Id: Ib1dfa2b4d80d5bbda9848be9870e4a02c14d7bab
Reviewed-by: Dominik Holland <dominik.holland@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If an app has been started and immediately stopped again, the next
time the app is started the assert would fail, that checks that the
app's root object is null. The stop signal would be handled before
the InProcessWindow could be created and hence the app's window
surface would never be released. The assert is in
QmlInProcessRuntime::start().
Cherry-picked from dev: cada823
Change-Id: I0711416a0fa4582bdaaa0a7f9b0c081eb6539c01
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Cherry-picked from dev: 742f39b
Change-Id: Id4e38169185e6a99a38294d7c5146efcf5db4716
Task-number: AUTOSUITE-951
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Cherry-picked from dev: d1ac8c9
Change-Id: I784772e66afe55434297b89669d2555318887dde
Task-number: AUTOSUITE-925
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The algorithm for finding a free 'qtam-wayland-%d' Wayland socket was
modelled after the code in libwayland-server, but due to a mixup between
the socket filename and the lock filename it was not working correctly.
Cherry-picked from dev: 4d01c85
Change-Id: Iac204c31fd0e6c239e44acefb5d65afe414461a8
Task-number: AUTOSUITE-857
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Cherry-picked from dev: 6be19e2
Task-number: AUTOSUITE-1079
Change-Id: I11fb811af92d782815659ce41c50008f9b336a71
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| | |
Cherry-picked from dev: f6a895c
Change-Id: Iaf4b3da0c622079ccb79886875134af58175bd43
Task-number: AUTOSUITE-1075
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Otherwise we might end up with a 0-byte app db, if the AM gets killed before
the OS (notably Windows) has synced the data to disk.
Cherry-picked from dev: 071799b
Task-number: AUTOSUITE-1096
Change-Id: Ifacf4854ab3f2b454d098a7b503b3922f72f9742
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| | |
Cherry-picked from dev: 6482d3d
Change-Id: If5f52f8fcf6f06bb0a4b9820c48b2ae9123f1e90
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
If no other configuration was chosen appman calculates its hardwareId
using the available network interfaces.
Instead of iterating and sorting them every time the hardwareId is
now saved once calculated.
This is also a workaround for QTBUG-76408 on macOS which reports
a wrong intereface type on the second run and cause the appman to
stop without hardwareId.
Cherry-picked from dev: c82f080
Task-number: AUTOSUITE-1019
Change-Id: I521245987486b09fce5c410918d0a35be37b1e10
Reviewed-by: Robert Griebl <robert.griebl@pelagicore.com>
|