It has the wrong path, doesn't work at all...
Change-Id: I28a5fbb599e4b2c696898ca0260102e3f69ed92a
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@theqtcompany.com>
This is covered now be the debuggee member.
Change-Id: Id2983e2b8d74236249263fdd4b867bd310bf20cc
Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
Lots of stuff is still horribly broken due to the weird confusion
between local/remote and "has project"/"does not have project" that
crept in at some point, but this at least makes remote, project-specific
memcheck work again.
Still broken:
- Remote callgrind
- All remote valgrinds for external projects.
- Local remote valgrind is not even available (anymore), as
the respective menu entries have disappeared at some point.
Change-Id: Iae6b60a0508d1b1b64fd0e0e5f776ea2c2e23598
Reviewed-by: hjk <hjk@theqtcompany.com>
Sprinkle overrides over code derived from classes in ProjectExplorer
Change-Id: Ia4cc25649f7dc00b0ea126d8176a59afbc5ed574
Reviewed-by: hjk <hjk@theqtcompany.com>
Using the AndroidRunnable is easy for other plugins to run custom adb
commands before intent is started and after it's stopped.
Change-Id: I012ae87c92cea16aa8074dce2dc6f2b0c4ebeb30
Reviewed-by: hjk <hjk@theqtcompany.com>
If the current instance of Creator is crashing, it might not be a good idea to
use it for debugging...
Change-Id: I4c28b3960c0722184fc92a6d20fbff8badb1fea3
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@theqtcompany.com>
deleteLater schedules the event in the current thread, so we
may not use it directly on a thread from within the thread itself...
Instead connect the threads finished signal to its deleteLater slot.
To avoid problems if runAsync is not run from the main thread,
move the thread to the main thread before starting it.
Task-number: QTCREATORBUG-15688
Change-Id: Iea003d00c58d1f921fb0b4ddf9cf67dcb1379833
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
These re-implementations are conceptually "too specific". Rather let
the indiviual tools handle there expectations themselves.
Change-Id: I0bbea407b2241816a40d19eb1dbb0a7589cbda7b
Reviewed-by: Christian Stenger <christian.stenger@theqtcompany.com>
...if possible. If passing tests had output beside the result
it had been omitted so far.
Change-Id: I37fd14714d00fe6a28f079f3fe2ca6e82606fa31
Reviewed-by: Niels Weber <niels.weber@theqtcompany.com>
The TextEditor settings do not remember a global Color Scheme setting,
anymore, but rather a Color Scheme ber Theme. A .creatortheme can define
a default TextEditor Color Scheme (overridable in the settings). This
makes switching of schemes more pleasant.
Taks-number: QTCREATORBUG-15229
Change-Id: I3bd36a4dfa23feea2254be2df50fce064e8fe2af
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@theqtcompany.com>
Reviewed-by: David Schulz <david.schulz@theqtcompany.com>
No worries, this is not supposed to stay as default. I just enabled falt
icons in the default theme in order to get feedback about technical
issues as-well as comments about the look-and-feel. Send feedback to the
patch author's email.
The plan so far is to revert this mid-February, and just have flat mode
icons enabled for some alternative themes.
Change-Id: I371f4ede432597ad7f1bc4b9ae661c9f2bc8df3e
Reviewed-by: Alessandro Portale <alessandro.portale@theqtcompany.com>
Timers/singleShot etc can only be used in QThreads (they require the
event dispatcher). So we have to use QThreads instead of std::threads to
nicely interoperate with Qt features.
Task-number: QTCREATORBUG-15681
Change-Id: I851e2f102e15ccd3347d455cc9d8b7df935c2d5e
Reviewed-by: Eike Ziller <eike.ziller@theqtcompany.com>
This is a mechanical replacement for the former executable, processArgs,
inferiorEnvironment and workingDirectory members.
Change-Id: I4160e01427ed801df9b729f1f31d0a2ca48159b5
Reviewed-by: Christian Stenger <christian.stenger@theqtcompany.com>
Using a global thread pool does not really make sense, since we do not
want to block any completely unrelated task. But the qmake project
manager still wants to limit the number of simultaneous threads without
managing that itself. A shared thread pool in project explorer sounds
like a sensible middle ground.
Change-Id: Ide6fd546a56e8f4896c387168513b608dfe7e3e8
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
To be used as type safe wrapper around a 16 bit port number.
Change-Id: I20a26be69e8b6b7dc9a7975fbe41c76ef89afbea
Reviewed-by: Eike Ziller <eike.ziller@theqtcompany.com>
QTimer::singleShot with a member function pointer or a lambda does
not schedule the invocation directly on the receivers event loop,
even if the timeout is 0. It actually creates a timer in the thread
where QTimer::singleShot is called, which requires an event dispatcher
to be available. Only the timer's timeout() signal then schedules the
invocation on the receivers thread.
We do not want to actually run an event loop here until a timer fires.
Even if a timer with timeout 0 might directly schedule the timeout
event on the event dispatcher without involving actual timers, and
even if that event dispatcher makes sure all events are handled before
deleting itself, it would feel like relying on an implementation
detail.
This reverts commit a2e19ba1be.
Task-number: QTCREATORBUG-15681
Change-Id: I19846ce089d72c971d3cc5927c2ddf563e06de14
Reviewed-by: Eike Ziller <eike.ziller@theqtcompany.com>
...mainly to not show the old zoom factor anymore.
Change-Id: I01c3050ff38b0b328774e6d05bcc2fe913ee4f56
Reviewed-by: Eike Ziller <eike.ziller@theqtcompany.com>
Instead, use the one in its (Standard)Runnable.
Change-Id: I14dbe91c50d083d11e18d514ec391875c64e3851
Reviewed-by: Christian Stenger <christian.stenger@theqtcompany.com>
...since that's
1. easier to handle (see e.g. QTCREATORBUG-15630, where a "&" was
not properly auto-detected).
2. consistent with the tooltip that is used for function signatures
(FunctionHintProposalWidgetPrivate)
Change-Id: I66aab253e13c5c7f63a0d9ddaf297b890567f335
Reviewed-by: David Schulz <david.schulz@theqtcompany.com>
Instead, use the one in its (Standard)Runnable.
The change is mechanical except to the fallback to the project directory
which was introduced in 8150209f to keep some compatibility with the previous
setup of the Android runner which created and copied the fallback value around
and ignored it in the end. Not producing the fallback value is less effort.
Change-Id: Ie04da570f0f9fbc1d048f2eacaed522f7253afa3
Reviewed-by: hjk <hjk@theqtcompany.com>
This reverts commit c5f70a3bad1ee2457741459cafb6419c67c417ad.
We will set the rich text unconditionally in a follow-up change to
simplify things.
Change-Id: I59aad8e33011ef68aa7c32ec80bb02edfd29c6a6
Reviewed-by: Marco Bubke <marco.bubke@theqtcompany.com>
... with a lot potential to code consolidation.
Change-Id: I4d3a7fcc1cc6ae8763799f18cf9701695f387791
Reviewed-by: Christian Stenger <christian.stenger@theqtcompany.com>
Do not run cmake 5 times during startup. Delay that as far as possible.
Also add a supportedGenerators() method while visiting the code anyway.
Fix up and simplify the other cmake help output parsers.
Change-Id: I6622d552ffe559bf099b4b278618676a045e350e
Reviewed-by: Tobias Hunger <tobias.hunger@theqtcompany.com>
runAsync avoids the globally shared thread pool.
updateCppQmlTypes missed some kind of progress information.
Change-Id: I8739761326c2f2a5364c0c1552b3c43b00f1ff28
Reviewed-by: Marco Benelli <marco.benelli@theqtcompany.com>
Reviewed-by: Eike Ziller <eike.ziller@theqtcompany.com>
Both runner tests fail on Windows at
QVERIFY(st.expectStartedSignal());
because QProcess::started() is emitted before ClangStaticAnalyzerRunner::run()
returns and thus before QSignalSpy::wait() is reached.
Watch out for the signal manually before calling run().
Change-Id: Ia786bfc6c859a4a4b420aeccc7e7f7bb0b4c47c0
Reviewed-by: Christian Kandeler <christian.kandeler@theqtcompany.com>
...this was wrong in cases where we have additional output before
this information.
Change-Id: I6f55c48815217da98078d40e5a0f1f4ce4ad8a77
Reviewed-by: Niels Weber <niels.weber@theqtcompany.com>
